From wz_ at gmx.net  Fri Sep  1 01:07:57 2006
From: wz_ at gmx.net (Nicolas Tessore)
Date: Fri, 01 Sep 2006 01:07:57 +0200
Subject: [C++-sig] Lost Again! :)
In-Reply-To: <7465b6170608311200j41b6f088tb725d832253338b6@mail.gmail.com>
References: <44F6ECDB.3070104@gmx.net>	<7465b6170608311033j4c678819o3d91c804f9f2a156@mail.gmail.com>	<44F726AF.4010001@gmx.net>
	<7465b6170608311200j41b6f088tb725d832253338b6@mail.gmail.com>
Message-ID: <44F76BCD.7000405@gmx.net>

Roman Yakovenko wrote:
> This is what I understand until now:
> 
> struct MyClass{...};
> struct HisClass : MyClass
> {};
> struct HerClass : MyClass
> {};
> 
> And you want to cast from HisClass to HerClass, right?
> ( I don't want to know why you have to do this, really :) )

That's right. This is the way the lib is build.

> If HisClass derives from HerClass, than you don't need to introduce
> casting.

I have to, they are siblings, not ancestors. Even if, I have to use what 
I'm given by the library.

> from pygccxml import declarations
> 
> def rename_cast( cast ):
>     cls_to = declarations.class_traits.get_declaration( cast.return_type )
>     cast.rename( 'DynamicCastTo' + cls_to.name )
> 
> casts = mb.free_functions ( lambda decl: decl.name.startswith( 
> 'DynamicCast<' ) )
> map( rename_cast, casts )

This always leaves me with a "query returned no matches" error :(

> May be you can debug the code? I can do it too, but you will have to send me
> GCCXML generated xml file and script that reporoduce the error. C++ source
> code is not requiered.

I cannot even reproduce it anymore myself. I will try to go back to my 
original tests, and see if I can bring it up again.

Thanks for the continued effort! N.


From djowel at gmail.com  Fri Sep  1 01:36:32 2006
From: djowel at gmail.com (Joel de Guzman)
Date: Fri, 01 Sep 2006 07:36:32 +0800
Subject: [C++-sig] [tutorial] Auto-Overloading example bug?
In-Reply-To: <7465b6170608311021y2f5bebecr90ab1766c5b84ad2@mail.gmail.com>
References: <7465b6170608300429n564516bfi22c665cd6cbfca32@mail.gmail.com>	<44F666B8.9080609@boost-consulting.com>	<44F67B9F.5060802@boost-consulting.com>
	<7465b6170608311021y2f5bebecr90ab1766c5b84ad2@mail.gmail.com>
Message-ID: <44F77280.80103@boost-consulting.com>

Roman Yakovenko wrote:
> On 8/31/06, Joel de Guzman <djowel at gmail.com> wrote:
>> Ok, disregard that. I did check it out and the error is the def it
>> should be:
>>
>>      .def("foo", (void(*)(bool, int, char))0, foo_overloads());
>>
>> I fixed the docs and updated to CVS.
> 
> Do you want to consider to fix "def"?
> 
> .def("foo", foo_overloads() );
> 
> It is not clear why user should pass second argument.

I don't think it can be fixed. The compiler does not have enough
information in there to know the exact signature needed. I may
be wrong though, so I'll investigate. If you know how to implement
this, I'm all ears.

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


From dave at boost-consulting.com  Fri Sep  1 02:53:21 2006
From: dave at boost-consulting.com (David Abrahams)
Date: Thu, 31 Aug 2006 20:53:21 -0400
Subject: [C++-sig] boost:python::arg and default value
References: <7465b6170607041320j2f745b89xff3f8d44e475b1c0@mail.gmail.com>
	<u64g9rlbr.fsf@boost-consulting.com>
	<7465b6170608311008n5b7ec413p535eeb27e92b3249@mail.gmail.com>
Message-ID: <u1wqwpfum.fsf@boost-consulting.com>

"Roman Yakovenko" <roman.yakovenko at gmail.com> writes:

> On 8/31/06, David Abrahams <dave at boost-consulting.com> wrote:
>> > I see only one possible solution - to add to Boost.Python
>> > functionality, that will
>> > allow user to control argument default value conversion time.
>>
>> I disagree.  It should be lazy and do the conversion on demand.  Extra
>> control is not a significant benefit in this instance.
>
> Does "lazy and on demand" mean first time a user calls function?

Yes... or even the first time a user actually uses the default.

> If so, I will try to work on the patch, but I think I still will
> need help with boost::python::objects::function class.

I'll do my best to help

-- 
Dave Abrahams
Boost Consulting
www.boost-consulting.com



From wz_ at gmx.net  Fri Sep  1 09:39:17 2006
From: wz_ at gmx.net (Nicolas Tessore)
Date: Fri, 01 Sep 2006 09:39:17 +0200
Subject: [C++-sig] Lost Again! :)
In-Reply-To: <44F76BCD.7000405@gmx.net>
References: <44F6ECDB.3070104@gmx.net>	<7465b6170608311033j4c678819o3d91c804f9f2a156@mail.gmail.com>	<44F726AF.4010001@gmx.net>	<7465b6170608311200j41b6f088tb725d832253338b6@mail.gmail.com>
	<44F76BCD.7000405@gmx.net>
Message-ID: <44F7E3A5.6000604@gmx.net>

Nicolas Tessore wrote:
>> from pygccxml import declarations
>>
>> def rename_cast( cast ):
>>     cls_to = declarations.class_traits.get_declaration( cast.return_type )
>>     cast.rename( 'DynamicCastTo' + cls_to.name )
>>
>> casts = mb.free_functions ( lambda decl: decl.name.startswith( 
>> 'DynamicCast<' ) )
>> map( rename_cast, casts )
> 
> This always leaves me with a "query returned no matches" error :(
Ok, I figured this out myself. GCCXML does not save the template part of 
the name, so all those functions are called "DynamicCast". I slightly 
modified your approach:

   01:  #Find object type in Casts. Const T casts are ignored
   02:  pattern = "::Mylib::Ref<Mylib::(.*)>"
   03:  p = re.compile(pattern)
   04:
   05:  def rename_casts( cast ):
   06:      m = p.match(cast.return_type.decl_string)
   07:      if m:
   08:          cast.rename( cast.name + 'To' + m.groups()[0] )
   09:      else: #DynamicCast should be excluded
   10:          cast.exclude()
   11:
   12:  casts = mb.free_functions ( lambda decl: ( decl.name == 
'DynamicCast' or decl.name == 'StaticCast' ) )
   13:  map( rename_casts, casts )

This works beautifully :) But is it a correct approach? And it clearly 
shows I barely scratched the surface of the power that Py++/ pygccxml 
hold :/

Now while still giggling at the wonderful result I saw, it happened that 
I noticed my typedefs ? la

     typedef Mylib::Ref<Mylib::MyClass> MyClassRef;
     sizeof(MyClassRef);

are not automatically detected and thus, the Mylib::Ref<Mylib::MyClass> 
is still named Mylib_scope_Ref_less_than_Mylib_scope_MyClass_grate etc 
in the generated binding source. Any clues why that could be? It is not 
a problem though, I will just apply the same method as above to them.

The road goes ever on and on , N.


From roman.yakovenko at gmail.com  Fri Sep  1 09:55:25 2006
From: roman.yakovenko at gmail.com (Roman Yakovenko)
Date: Fri, 1 Sep 2006 10:55:25 +0300
Subject: [C++-sig] Lost Again! :)
In-Reply-To: <44F7E3A5.6000604@gmx.net>
References: <44F6ECDB.3070104@gmx.net>
	<7465b6170608311033j4c678819o3d91c804f9f2a156@mail.gmail.com>
	<44F726AF.4010001@gmx.net>
	<7465b6170608311200j41b6f088tb725d832253338b6@mail.gmail.com>
	<44F76BCD.7000405@gmx.net> <44F7E3A5.6000604@gmx.net>
Message-ID: <7465b6170609010055s371123daje97c9da8ff01b595@mail.gmail.com>

On 9/1/06, Nicolas Tessore <wz_ at gmx.net> wrote:
> Nicolas Tessore wrote:
> >> from pygccxml import declarations
> >>
> >> def rename_cast( cast ):
> >>     cls_to = declarations.class_traits.get_declaration( cast.return_type )
> >>     cast.rename( 'DynamicCastTo' + cls_to.name )
> >>
> >> casts = mb.free_functions ( lambda decl: decl.name.startswith(
> >> 'DynamicCast<' ) )
> >> map( rename_cast, casts )
> >
> > This always leaves me with a "query returned no matches" error :(
> Ok, I figured this out myself. GCCXML does not save the template part of
> the name, so all those functions are called "DynamicCast". I slightly
> modified your approach:
>
>    01:  #Find object type in Casts. Const T casts are ignored
>    02:  pattern = "::Mylib::Ref<Mylib::(.*)>"
>    03:  p = re.compile(pattern)
>    04:
>    05:  def rename_casts( cast ):
>    06:      m = p.match(cast.return_type.decl_string)
>    07:      if m:
>    08:          cast.rename( cast.name + 'To' + m.groups()[0] )
>    09:      else: #DynamicCast should be excluded
>    10:          cast.exclude()
>    11:
>    12:  casts = mb.free_functions ( lambda decl: ( decl.name ==
> 'DynamicCast' or decl.name == 'StaticCast' ) )
>    13:  map( rename_casts, casts )
>
> This works beautifully :) But is it a correct approach?

If it works, that the answer is yes.

> And it clearly
> shows I barely scratched the surface of the power that Py++/ pygccxml
> hold :/

Please take a look on
http://language-binding.net/pygccxml/apidocs/pygccxml.declarations.templates-module.html
The functionality provided by this module will do the job for you.

But you ignored my advice:

cls_to = declarations.class_traits.get_declaration( cast.return_type )
cast.rename( 'DynamicCastTo' + cls_to.name )

You don't have to parse, just extract class declaration from DynamicCast return
type.

> Now while still giggling at the wonderful result I saw, it happened that
> I noticed my typedefs ? la
>
>      typedef Mylib::Ref<Mylib::MyClass> MyClassRef;
>      sizeof(MyClassRef);
>
> are not automatically detected and thus, the Mylib::Ref<Mylib::MyClass>
> is still named Mylib_scope_Ref_less_than_Mylib_scope_MyClass_grate etc
> in the generated binding source. Any clues why that could be? It is not
> a problem though, I will just apply the same method as above to them.

Read this: http://language-binding.net/pyplusplus/documentation/hints.html
It should explain why Py++ does not work as you expects.

-- 
Roman Yakovenko
C++ Python language binding
http://www.language-binding.net/


From wz_ at gmx.net  Fri Sep  1 10:01:09 2006
From: wz_ at gmx.net (Nicolas Tessore)
Date: Fri, 01 Sep 2006 10:01:09 +0200
Subject: [C++-sig] Lost Again! :)
In-Reply-To: <7465b6170609010055s371123daje97c9da8ff01b595@mail.gmail.com>
References: <44F6ECDB.3070104@gmx.net>	<7465b6170608311033j4c678819o3d91c804f9f2a156@mail.gmail.com>	<44F726AF.4010001@gmx.net>	<7465b6170608311200j41b6f088tb725d832253338b6@mail.gmail.com>	<44F76BCD.7000405@gmx.net>
	<44F7E3A5.6000604@gmx.net>
	<7465b6170609010055s371123daje97c9da8ff01b595@mail.gmail.com>
Message-ID: <44F7E8C5.9030308@gmx.net>

Roman Yakovenko wrote:
> Please take a look on
> http://language-binding.net/pygccxml/apidocs/pygccxml.declarations.templates-module.html
> The functionality provided by this module will do the job for you.
I guess I should start searching both docs more. Thanks for the heads-up.

> But you ignored my advice:
> 
> cls_to = declarations.class_traits.get_declaration( cast.return_type )
> cast.rename( 'DynamicCastTo' + cls_to.name )
> 
> You don't have to parse, just extract class declaration from DynamicCast return
> type.

No, the return type is not the right kind. This is due to the fact that 
Mylib::Ref<T> is used as a pointer to T. Still the return type is T*.

> Read this: http://language-binding.net/pyplusplus/documentation/hints.html
> It should explain why Py++ does not work as you expects.
> 
Yes, I saw that and thus want to do it in the script :)

N.


From wz_ at gmx.net  Fri Sep  1 12:01:02 2006
From: wz_ at gmx.net (Nicolas Tessore)
Date: Fri, 01 Sep 2006 12:01:02 +0200
Subject: [C++-sig] Lost Again! :)
In-Reply-To: <7465b6170609010055s371123daje97c9da8ff01b595@mail.gmail.com>
References: <44F6ECDB.3070104@gmx.net>	<7465b6170608311033j4c678819o3d91c804f9f2a156@mail.gmail.com>	<44F726AF.4010001@gmx.net>	<7465b6170608311200j41b6f088tb725d832253338b6@mail.gmail.com>	<44F76BCD.7000405@gmx.net>
	<44F7E3A5.6000604@gmx.net>
	<7465b6170609010055s371123daje97c9da8ff01b595@mail.gmail.com>
Message-ID: <44F804DE.5020301@gmx.net>

Roman Yakovenko wrote:

> Please take a look on
> http://language-binding.net/pygccxml/apidocs/pygccxml.declarations.templates-module.html
> The functionality provided by this module will do the job for you.
> 
> But you ignored my advice:
> 
> cls_to = declarations.class_traits.get_declaration( cast.return_type )
> cast.rename( 'DynamicCastTo' + cls_to.name )
> 
> You don't have to parse, just extract class declaration from DynamicCast return
> type.

Ok, I even understood that now:

   01: def rename_casts( cast ):
   02:    cast_to = 
declarations.templates.args(cast.decl_string)[0].rsplit('::', 2)[-1]
   03:    cast.rename( cast.name + 'To' +  cast_to)
   04:
   05: def rename_refs( ref ):
   06:    ref_to = 
declarations.templates.args(ref.decl_string)[0].rsplit('::', 2)[-1]
   07:    ref.rename( ref_to + "Ref" )
   08:
   09: refs = mb.classes ( lambda decl: ( decl.name.startswith('Ref<') ) )
   10: map( rename_refs, refs )
   11:
   12: casts = mb.free_functions ( lambda decl: ( decl.name == 
'DynamicCast' or decl.name == 'StaticCast' ) )
   13: map( rename_casts, casts )

This leaves me with a build error, yelling at me that I should specify a 
call policy. Am I right this is not currently supported by Py++? At 
least my google searches told me so.

N.


From jorge.vargas at gmail.com  Fri Sep  1 16:04:38 2006
From: jorge.vargas at gmail.com (Jorge Vargas)
Date: Fri, 1 Sep 2006 10:04:38 -0400
Subject: [C++-sig] boost python mingw binaries
In-Reply-To: <32822fe60608290622l7c0382ffs87219f4532196fdf@mail.gmail.com>
References: <32822fe60608281258g669fb64fj23bcd6f46133b89c@mail.gmail.com>
	<200608291018.53778.achim-bpl@mol-net.com>
	<32822fe60608290622l7c0382ffs87219f4532196fdf@mail.gmail.com>
Message-ID: <32822fe60609010704p7c99ca67i14aa76e8313ccfc4@mail.gmail.com>

On 8/29/06, Jorge Vargas <jorge.vargas at gmail.com> wrote:
> On 8/29/06, Achim H. <achim-bpl at mol-net.com> wrote:
> > Am Montag, 28. August 2006 21:58 schrieb Jorge Vargas:
> > > so does anyone has boost python as a binary that works under mingw?
> Great thanks
> > Yes, I have binaries of 1.33.1 and I can even upload them if you tell me where
> > to.
> I guess the best place for that is the SF repository, since is the
> official site. So I guess we could wait for someone with access there
> to reply.
>
anyone is interested in posting this to SF?


From roman.yakovenko at gmail.com  Fri Sep  1 16:21:52 2006
From: roman.yakovenko at gmail.com (Roman Yakovenko)
Date: Fri, 1 Sep 2006 16:21:52 +0200
Subject: [C++-sig] Lost Again! :)
In-Reply-To: <44F7E8C5.9030308@gmx.net>
References: <44F6ECDB.3070104@gmx.net>
	<7465b6170608311033j4c678819o3d91c804f9f2a156@mail.gmail.com>
	<44F726AF.4010001@gmx.net>
	<7465b6170608311200j41b6f088tb725d832253338b6@mail.gmail.com>
	<44F76BCD.7000405@gmx.net> <44F7E3A5.6000604@gmx.net>
	<7465b6170609010055s371123daje97c9da8ff01b595@mail.gmail.com>
	<44F7E8C5.9030308@gmx.net>
Message-ID: <7465b6170609010721y26de3ci2f61af619ee7780f@mail.gmail.com>

On 9/1/06, Nicolas Tessore <wz_ at gmx.net> wrote:
> Roman Yakovenko wrote:
> > Please take a look on
> > http://language-binding.net/pygccxml/apidocs/pygccxml.declarations.templates-module.html
> > The functionality provided by this module will do the job for you.
> I guess I should start searching both docs more. Thanks for the heads-up.
>
> > But you ignored my advice:
> >
> > cls_to = declarations.class_traits.get_declaration( cast.return_type )
> > cast.rename( 'DynamicCastTo' + cls_to.name )
> >
> > You don't have to parse, just extract class declaration from DynamicCast return
> > type.
>
> No, the return type is not the right kind. This is due to the fact that
> Mylib::Ref<T> is used as a pointer to T. Still the return type is T*.

Take a look on
http://language-binding.net/pygccxml/apidocs/pygccxml.declarations.type_traits-module.html
P++ via pygccxml provides a lot of "type_traits" functionality found
in boost.type_traits
library:
http://language-binding.net/pygccxml/apidocs/pygccxml.declarations.type_traits-module.html

> > Read this: http://language-binding.net/pyplusplus/documentation/hints.html
> > It should explain why Py++ does not work as you expects.
> >
> Yes, I saw that and thus want to do it in the script :)

http://language-binding.net/pygccxml/apidocs/pygccxml.declarations.class_declaration.class_t-class.html#aliases

Take a look on the documentation. May be you will be able to implement
custom logic
that will assign aliases automatically

-- 
Roman Yakovenko
C++ Python language binding
http://www.language-binding.net/


From roman.yakovenko at gmail.com  Fri Sep  1 16:26:18 2006
From: roman.yakovenko at gmail.com (Roman Yakovenko)
Date: Fri, 1 Sep 2006 16:26:18 +0200
Subject: [C++-sig] Lost Again! :)
In-Reply-To: <44F804DE.5020301@gmx.net>
References: <44F6ECDB.3070104@gmx.net>
	<7465b6170608311033j4c678819o3d91c804f9f2a156@mail.gmail.com>
	<44F726AF.4010001@gmx.net>
	<7465b6170608311200j41b6f088tb725d832253338b6@mail.gmail.com>
	<44F76BCD.7000405@gmx.net> <44F7E3A5.6000604@gmx.net>
	<7465b6170609010055s371123daje97c9da8ff01b595@mail.gmail.com>
	<44F804DE.5020301@gmx.net>
Message-ID: <7465b6170609010726k6e2a78a1r124cf6e388a542c3@mail.gmail.com>

On 9/1/06, Nicolas Tessore <wz_ at gmx.net> wrote:
> Roman Yakovenko wrote:
>
> > Please take a look on
> > http://language-binding.net/pygccxml/apidocs/pygccxml.declarations.templates-module.html
> > The functionality provided by this module will do the job for you.
> >
> > But you ignored my advice:
> >
> > cls_to = declarations.class_traits.get_declaration( cast.return_type )
> > cast.rename( 'DynamicCastTo' + cls_to.name )
> >
> > You don't have to parse, just extract class declaration from DynamicCast return
> > type.
>
> Ok, I even understood that now:
>
>    01: def rename_casts( cast ):
>    02:    cast_to =
> declarations.templates.args(cast.decl_string)[0].rsplit('::', 2)[-1]
>    03:    cast.rename( cast.name + 'To' +  cast_to)
>    04:
>    05: def rename_refs( ref ):
>    06:    ref_to =
> declarations.templates.args(ref.decl_string)[0].rsplit('::', 2)[-1]
>    07:    ref.rename( ref_to + "Ref" )
>    08:
>    09: refs = mb.classes ( lambda decl: ( decl.name.startswith('Ref<') ) )
>    10: map( rename_refs, refs )
>    11:
>    12: casts = mb.free_functions ( lambda decl: ( decl.name ==
> 'DynamicCast' or decl.name == 'StaticCast' ) )
>    13: map( rename_casts, casts )
>
> This leaves me with a build error, yelling at me that I should specify a
> call policy. Am I right this is not currently supported by Py++? At
> least my google searches told me so.

API docs:
http://language-binding.net/pyplusplus/documentation/apidocs/pyplusplus.decl_wrappers.calldef_wrapper.free_function_t-class.html
http://language-binding.net/pyplusplus/documentation/apidocs/pyplusplus.decl_wrappers.call_policies-module.html

Tutorials:
http://language-binding.net/pyplusplus/documentation/tutorials/module_builder/module_builder.html
http://language-binding.net/pyplusplus/documentation/tutorials/module_builder/generate_code.html
-- 
Roman Yakovenko
C++ Python language binding
http://www.language-binding.net/


From lloyd at randombit.net  Fri Sep  1 22:25:36 2006
From: lloyd at randombit.net (Jack Lloyd)
Date: Fri, 1 Sep 2006 16:25:36 -0400
Subject: [C++-sig] Returning a Python object to C++ (is get_owner the right
	way?)
Message-ID: <20060901202535.GC28064@randombit.net>


I have a scenario where I'm wrapping a virtual base class that offers
cloning (here, sending a clone message to an object means "return a
newly constructed object of whatever type you are"), and want to
inherit from that type in Python. Simply implementing that function in
Python (eg:

class MySubclass(mylib.MyBaseClass):
   def clone(self):
      return MySubclass()

led to "returning a dangling reference" errors, so I'm using the
technique Dave suggested in
http://mail.python.org/pipermail/c++-sig/2004-December/008379.html
(basically, use get_owner, get ahold of the __class__ attribute, and
call the constructor, then keep a reference to the copy-constructed
Python object around so it is not destroyed until the lifetime of the
C++ object is over, all held inside another class that simply forwards
the appropriate messages to the Python object).

This has been working fine for me, however since get_owner is still
undocumented (AFAICT; couldn't find it anywhere in the 1.33.1 docs),
and generally the technique led to some rather ugly code, I wanted to
check and see if any new technique/method for dealing with this
situation had developed since that discussion?

I hope I've managed to explain what I'm doing clearly; if not, I'd
be happy to post a minimal example of my method.

Thanks,
  Jack


From pirx at cs.tu-berlin.de  Sun Sep  3 15:24:52 2006
From: pirx at cs.tu-berlin.de (Andreas)
Date: Sun, 03 Sep 2006 15:24:52 +0200
Subject: [C++-sig] problem with tkinter gui running in different interpreters
In-Reply-To: <cscblt$tfp$1@sea.gmane.org>
References: <002e01c4fa72$d7f3a060$1d02a8c0@wxp2000>	<003301c4fa85$0c9d8260$1d02a8c0@wxp2000><csablh$o4q$1@sea.gmane.org>	<csc112$66t$1@sea.gmane.org>	<csc833$lvj$1@sea.gmane.org>
	<cscblt$tfp$1@sea.gmane.org>
Message-ID: <44FAD7A4.2040908@cs.tu-berlin.de>

hello,

I'm not sure if this is the right place for my question.
I have a problem with tkinter running in different python interpreters.

I need to separate different python scripts as much as possible. 
Therefor each script runs in its own interpreter.
If only one of the scripts uses tkinter gui and runs the tkinter's 
mainloop in a created thread, everything is fine. But if more scripts 
uses this, the whole application with the created dialogs hangs.
To solve this only the first script must call tkinter's mainloop(). But 
thats not an option, because the scripts are independent and must not 
have dialogs. Thus a script does not know if it is the first which uses 
tkinter.

So the question is:
Is it possible to detect from different interpreters whether the 
tkinter's mainloop is already running or not?

Below is the running example code. In main() i want to start multiple 
dialogs calling startDialog(true), wich does not work.

(I'm using boost 1.33.1, MSVC7.1 and Python 2.4.3)

Any help would be appreciated,
andreas rose






void startDialog(bool bCallMainloop)
{
   PyEval_AcquireLock();
   PyThreadState* pThreadState = Py_NewInterpreter();
   PyThreadState_Swap(pThreadState);

   char pBuff[] = "Test.exe";
   char* Argv[1];
   Argv[0] = pBuff;
   PySys_SetArgv(1, Argv);

   PyObject* pDict = PyDict_New();
   PyDict_SetItemString(pDict, "__builtins__", PyEval_GetBuiltins());

   std::string strExec;
   if (bCallMainloop)
   {
   strExec =
"\
from threading import *                                       \n\
from Tkinter import *                                         \n\
                                                               \n\
class DialogThread(Thread):                                   \n\
     def run(self):                                            \n\
         label = Label(None, text='hello')                     \n\
         label.pack()                                          \n\
         label.mainloop()                                      \n\
                                                               \n\
DialogThread().start()                                        \n\
";
   }
   else
   {
   strExec =
"\
from threading import *                                       \n\
from Tkinter import *                                         \n\
                                                               \n\
class DialogThread(Thread):                                   \n\
     def run(self):                                            \n\
         label = Label(None, text='hello')                     \n\
         label.pack()                                          \n\
                                                               \n\
DialogThread().start()                                        \n\
";
   }

   boost::python::handle<> result(PyRun_String(strExec.c_str(), 
Py_file_input, pDict, pDict));
   result.reset();

   PyEval_ReleaseThread(pThreadState);
}


int main(int argc, char** argv)
{
   Py_Initialize();
   PyEval_InitThreads();
   PyThreadState* pyMainState = PyThreadState_Swap(NULL);
   PyEval_ReleaseLock();


   startDialog(true);

   startDialog(false);

   //wait little time to enable user interaction with the dialogs
   Sleep(10000);

   PyEval_AcquireLock();
   PyThreadState_Swap(pyMainState);
   Py_Finalize();
   return 0;
}


From Gottfried.Ganssauge at haufe.de  Sun Sep  3 16:07:14 2006
From: Gottfried.Ganssauge at haufe.de (Ganssauge, Gottfried)
Date: Sun, 3 Sep 2006 16:07:14 +0200
Subject: [C++-sig] Py++ wrapper generation and docstrings
In-Reply-To: <7465b6170609030548o13421632if3e04af9d35bd248@mail.gmail.com>
Message-ID: <51F4AD92DB06E9488BBC39B7B492409E0162AAB0@vg20ex1.grp.haufemg.com>

> -----Original Message-----
> From: Roman Yakovenko [mailto:roman.yakovenko at gmail.com] 
> Sent: Sunday, September 03, 2006 2:49 PM
> To: Ganssauge, Gottfried
> Subject: Re: Py++ wrapper generation and docstrings
> 
> On 9/3/06, Ganssauge, Gottfried <Gottfried.Ganssauge at haufe.de> wrote:
> > Hi Roman,
> >
> > I'm still working on my API with Py++ and there are a few bits left:
> 
> Is this because of Py++ or you have pretty big project :-)?
Neither nor I guess.
The project is quite old, it's got a C-API and the API is highly
irregular :-(

> 
> > 1.) my API contains several functions expecting pointers to callback
> > functions.
> >     Those callbacks get an instance pointer as one of their 
> arguments
> > which has to be supplied as well.
> >     An example:
> >        typedef int (*callback_t) (void *instance);
> >
> >        int search (void *instance, callback_t callback);
> >
> >     I have written a generator which generates a wrapper like this:
> >
> >         int wrap_callback_t (void *instance) {
> 
> You can use boost::python::object
No I can't :-(
Gccxml botches on the bpl headers:
$ cat x.hpp
# include <boost/python.hpp>

boost::python::object test ();
$ gccxml -I ../../../../../../FremdProdukte/boost2 -I
/usr/include/python2.4 x.hpp -fxml=test.xml
../../../../../../FremdProdukte/boost2/boost/python/object_core.hpp:285:
sorry, unimplemented: call_expr
   cannot be mangled due to a defect in the C++ ABI

This problem has been reported on the gccxml list already ...

> 
> >                 try {
> >                         return (boost::python::call<int>
> > (reinterpret_cast<PyObject *> (instance)));
> >                 }
> >
> >                 catch (...) {
> >                         return (-1);    // python 
> exceptions must not
> > disrupt program flow
> >                 }
> >           }
> >
> >         int search_wrapper (PyObject *p) {
> >                 return (search (p, &wrap_callback_t));
> >           }
> >
> >      This works quite well, but ...
> >      How do I make this generator interact with generators already
> > supplied by Py++
> >      For example:
> >         you're planning to support automatic conversion of 
> pointers to
> > C++ array arguments from python sequences
> 
> May be the next article will provide you with the answer.
> https://realityforge.vrsource.org/view/PyppApi/CodeInserter
> 
> I and Matthias work on this feature this days. Comments and test cases
> are welcome :-)
I'll have a look at it.

> 
> >      Now suppose the following:
> >          type char array_t [7];
> >
> >            int search2 (void *instance, callback_t cb, 
> array_t *result);
> 
> If you already create a wrapper, you can create  "better" wrapper:
> 
> boost::python::tuple search2( void* instance, callback_t cb ){
>     ...
>     return boost::python::make_tuple( .... )
> }
That's what I'd like to achieve - but automatically!

> 
> >      How would both our generators interact with each other?
> 
> I don't see the whole picture, so I can not answer your question.
> I assume that you have some functions that take pointer to 
> callback, right?
> You can not expose these functions as is, so you have to 
> write a wrapper for
> them. Basically you should exclude the function from being exported,
> and then you
> should add your declaration and registration code. Next link should
> help you with
> your task:
> 
>
http://www.language-binding.net/pyplusplus/documentation/inserting_code.
html
> 
> 
> >      Finally I'm not sure about when to run my generator.
> 
> In your case you may consider to run it twice:
> 
> 1. first time you can generate your own header file. This file will
> contain all wrappers.
> 2. second time you weel feed just generated header fileand original
> sources to Py++.
> 
> >      Currently I'm running it on every declaration 
> expecting callback-
> > and instance pointers.
> >      The declarations are then modified like this (f is a 
> declaration
> > for "search"):
> >         pyobject_type = pointer_t (dummy_type_t ("PyObject"))):
> >         wrapperargs   = [ argument_t (name = "instance", type =
> > pyobject_type) ]
> >         function      = f.name
> >         f.name        = "%s_wrapper" % function
> >         f.arguments   = wrapperargs
> >         f.exportable  = None # redo the check for exportability
> >           # rename f if this is the first wrapper
> >         if not f.alias or f.name == f.alias:
> >             f.rename (function)
> >
> >       Does this look reasonable? (at least it works :-0)
> 
> No. Too much work and too much details. Consider the way I described.
That was my first attempt, but it became difficult to chain several
generators together.
Eventually I should really consider doing several generator passes to a
temp file until the temp file doesn't change anymore.
This could take quite some time because gccxml must parse the
intermediate files...
Furthermore I'm not yet sure how to forward the original function names
...

> 
> >       I got another generator which generates wrappers for functions
> > expecting "const unsigned char *" arguments for strings ...
> >
> > 2.) docstring extraction is not really usable at the moment
> >     consider the attached test case
> >        [test1.hpp, pypp_test1.py]
> >
> >     When using the "latin1_doc_string" doc_extractor some unicode
> > exception is thrown.
> 
> I will take a look on this. I did not thought about unicodes :-(. It
> will take some time to
> convert Py++ to support it. I think, that doc_extractor will change,
> because you will
> have to provide encoding. Otherwise, I will not be able to 
> write code to files.
This could be a property on the doc_extractor.

> Also quick work around exists:
> 
> my_module
>     class_< X >( "X" )
>         .def( "do_smth", ... )
>    ;
> 
> import my_module
> my_module.X.__doc__ = ....
> my_module.X.do_smth.__doc__ = ...
> 
> This will work and it is simple to generate it.
Ok.

> 
> >     When using the "ascii_doc_string" doc_extractor invalid code is
> > generated
> >     Only when the doc extractor returns a string which is 
> conforming to
> > C++ syntax everything runs smooth.
> 
> This is how it was created. See explanation later.
> 
> >     Pragmatically I would request that doc extractors must return
> > docstrings conforming to C++ syntax.
> >     This means you wouldn't need to change Py++ - just the
> > documentation.
> >     You could supply some function which makes C++-syntax 
> from a normal
> > string like
> >         def make_cstring (s):
> >             return ('"%s"' % (repr (s) [1:-1]))
> 
> It is not as simple as you think :-). You have to escape " in the
> string and you have to do
> it in a smart way. That is why I leaved this to user. If you want we
> can work together
> to improve the situation. I will make Py++ to work with unicode and
> you will contribute
> better make_cstring function.
Currently I'm using
    string_escapes = re.compile (r'((\\x[a-f0-9][a-f0-9])|(\\*"))',
re.I)
    def make_cstring (s):
        def replace_escape(m):
            g = m.group(1)
            if g.startswith ('\\x'):
                return g + '""'
            elif g == '"':
                return '\\"'
            else:
                return g
            
        return '"%s"' % hex_escape.sub (replace_escape, repr (s) [1:-1])

This copes with the problem of hex escapes followed by valid hex
characters and with embedded quotes.

> 
> 
> Can I ask you something?
> 
> Could we switch to one of the mailing lists: pygccxml of boost.python?
> I'd like my answers
> to be available to other users too.
at your request :-)

> 
> It seems that you found work around to "pointer to function" problem.
> Can you create
> small example with explanation how it should be done. May be I will
> able to introduce
> this functionality to Py++.
Well not for any case.
Function pointers without instance pointers still can't be exposed for
the reasons given in the reference provided by Py++.
Function pointers _with_ instance pointers are a different kind of
beast.
Generally function declarations using this type of callback have exactly
one function pointer argument and exactly one void *argument which is
forwarded to the callback function.
	typedef callback_return_type (*callback_t) (callback_args, void
*instance, more callback_args);

	return_type function (args ..., callback_t, more args, void
*instance, more args)

Functions having this kind of signature may be wrapped like this:

	callback_return_type wrap_callback_t (callback_args, void
*instance, more callback_args) {
		PyObject *py_instance = reinterpret_cast<PyObject *>
(instance);

		if (py_instance) {
			try {
				return
(boost::python::call<callback_return_type> (py_instance, callback_args,
more callback_args))
			}

			except (...) {
			}
		}

		return some error value;
	}

	return_type function_wrapper (PyObject *instance, args,
more_args, ...) {
		return (function (args ..., &wrap_callback_t, more args,
instance, more args));
	}

I'm attaching the generators module I'm currently using ...
The generators are called like this:
    def wrap_functions (self, decls):
        """
        Generate wrappers for functions otherwise not exportable in
their
        current_form

        @param decls
        Function declarations to be wrapped
        """
        wrapper_generators = [
            callback_wrapper_generator (self),
            const_byte_wrapper_generator (self)
        ]
        finished = False
        while not finished:
            finished = True
            for d in decls:
                for g in wrapper_generators:
                    cookie = g.applies (d)
                    if cookie:
                        g (cookie)
                        finished = False
                        break


> 
> Thanks.
> 
> -- 
> Roman Yakovenko
> C++ Python language binding
> http://www.language-binding.net/
> 
Cheers,

Gottfried
-------------- next part --------------
A non-text attachment was scrubbed...
Name: generators.py
Type: application/octet-stream
Size: 9682 bytes
Desc: generators.py
URL: <http://mail.python.org/pipermail/cplusplus-sig/attachments/20060903/d49c6c54/attachment.obj>

From roman.yakovenko at gmail.com  Sun Sep  3 20:14:27 2006
From: roman.yakovenko at gmail.com (Roman Yakovenko)
Date: Sun, 3 Sep 2006 21:14:27 +0300
Subject: [C++-sig] Py++ wrapper generation and docstrings
In-Reply-To: <51F4AD92DB06E9488BBC39B7B492409E0162AAB0@vg20ex1.grp.haufemg.com>
References: <7465b6170609030548o13421632if3e04af9d35bd248@mail.gmail.com>
	<51F4AD92DB06E9488BBC39B7B492409E0162AAB0@vg20ex1.grp.haufemg.com>
Message-ID: <7465b6170609031114q4d906763y9b61fd5007df2597@mail.gmail.com>

On 9/3/06, Ganssauge, Gottfried <Gottfried.Ganssauge at haufe.de> wrote:
> > >      Now suppose the following:
> > >          type char array_t [7];
> > >
> > >            int search2 (void *instance, callback_t cb,
> > array_t *result);
> >
> > If you already create a wrapper, you can create  "better" wrapper:
> >
> > boost::python::tuple search2( void* instance, callback_t cb ){
> >     ...
> >     return boost::python::make_tuple( .... )
> > }
> That's what I'd like to achieve - but automatically!

Lets be more productive. Can you create small example, that you want to export
and Py++ script and I will make it work?

> >
> > >      How would both our generators interact with each other?
> >
> > I don't see the whole picture, so I can not answer your question.
> > I assume that you have some functions that take pointer to
> > callback, right?
> > You can not expose these functions as is, so you have to
> > write a wrapper for
> > them. Basically you should exclude the function from being exported,
> > and then you
> > should add your declaration and registration code. Next link should
> > help you with
> > your task:
> >
> >
> http://www.language-binding.net/pyplusplus/documentation/inserting_code.
> html
> >
> >
> > >      Finally I'm not sure about when to run my generator.
> >
> > In your case you may consider to run it twice:
> >
> > 1. first time you can generate your own header file. This file will
> > contain all wrappers.
> > 2. second time you weel feed just generated header fileand original
> > sources to Py++.
> >
> That was my first attempt, but it became difficult to chain several
> generators together.
> Eventually I should really consider doing several generator passes to a
> temp file until the temp file doesn't change anymore.
> This could take quite some time because gccxml must parse the
> intermediate files...
> Furthermore I'm not yet sure how to forward the original function names
> ...

I still fill that I don't have enough information to help you. For the
last problem I
can give you small hint: mangle generated function name.

Take a look on boost_random example:
http://tinyurl.com/gwc2t
http://tinyurl.com/kocap

I introduced small convention about struct names and than I reused this
information from the script.


> >
> > >       I got another generator which generates wrappers for functions
> > > expecting "const unsigned char *" arguments for strings ...
> > >
> > > 2.) docstring extraction is not really usable at the moment
> > >     consider the attached test case
> > >        [test1.hpp, pypp_test1.py]
> > >
> > >     When using the "latin1_doc_string" doc_extractor some unicode
> > > exception is thrown.
> >
> > I will take a look on this. I did not thought about unicodes :-(. It
> > will take some time to
> > convert Py++ to support it. I think, that doc_extractor will change,
> > because you will
> > have to provide encoding. Otherwise, I will not be able to
> > write code to files.
> This could be a property on the doc_extractor.

Yes I know, that is exactly what I meant by "doc_extractor" will change.

> Currently I'm using
>     string_escapes = re.compile (r'((\\x[a-f0-9][a-f0-9])|(\\*"))',
> re.I)
>     def make_cstring (s):
>         def replace_escape(m):
>             g = m.group(1)
>             if g.startswith ('\\x'):
>                 return g + '""'
>             elif g == '"':
>                 return '\\"'
>             else:
>                 return g
>
>         return '"%s"' % hex_escape.sub (replace_escape, repr (s) [1:-1])
>
> This copes with the problem of hex escapes followed by valid hex
> characters and with embedded quotes.

May I copy this code to Py++ contrib package under the license(
http://www.boost.org/more/license_info.html )

> > It seems that you found work around to "pointer to function" problem.
> > Can you create
> > small example with explanation how it should be done. May be I will
> > able to introduce
> > this functionality to Py++.
> Well not for any case.
> Function pointers without instance pointers still can't be exposed for
> the reasons given in the reference provided by Py++.
> Function pointers _with_ instance pointers are a different kind of
> beast.
> Generally function declarations using this type of callback have exactly
> one function pointer argument and exactly one void *argument which is
> forwarded to the callback function.
>         typedef callback_return_type (*callback_t) (callback_args, void
> *instance, more callback_args);
>
>         return_type function (args ..., callback_t, more args, void
> *instance, more args)
>
> Functions having this kind of signature may be wrapped like this:
>
>         callback_return_type wrap_callback_t (callback_args, void
> *instance, more callback_args) {
>                 PyObject *py_instance = reinterpret_cast<PyObject *>
> (instance);
>
>                 if (py_instance) {
>                         try {
>                                 return
> (boost::python::call<callback_return_type> (py_instance, callback_args,
> more callback_args))
>                         }
>
>                         except (...) {
>                         }
>                 }
>
>                 return some error value;
>         }
>
>         return_type function_wrapper (PyObject *instance, args,
> more_args, ...) {
>                 return (function (args ..., &wrap_callback_t, more args,
> instance, more args));
>         }
>
> I'm attaching the generators module I'm currently using ...
> The generators are called like this:
>     def wrap_functions (self, decls):
>         """
>         Generate wrappers for functions otherwise not exportable in
> their
>         current_form
>
>         @param decls
>         Function declarations to be wrapped
>         """
>         wrapper_generators = [
>             callback_wrapper_generator (self),
>             const_byte_wrapper_generator (self)
>         ]
>         finished = False
>         while not finished:
>             finished = True
>             for d in decls:
>                 for g in wrapper_generators:
>                     cookie = g.applies (d)
>                     if cookie:
>                         g (cookie)
>                         finished = False
>                         break
>

Thank you for explanation.

May I copy this code to Py++ contrib package under the license(
http://www.boost.org/more/license_info.html )


-- 
Roman Yakovenko
C++ Python language binding
http://www.language-binding.net/


From roman.yakovenko at gmail.com  Sun Sep  3 21:33:06 2006
From: roman.yakovenko at gmail.com (Roman Yakovenko)
Date: Sun, 3 Sep 2006 22:33:06 +0300
Subject: [C++-sig] Py++ wrapper generation and docstrings
In-Reply-To: <51F4AD92DB06E9488BBC39B7B492409E0162AAB0@vg20ex1.grp.haufemg.com>
References: <7465b6170609030548o13421632if3e04af9d35bd248@mail.gmail.com>
	<51F4AD92DB06E9488BBC39B7B492409E0162AAB0@vg20ex1.grp.haufemg.com>
Message-ID: <7465b6170609031233r7ffa8c38nd9091329d4b96aaa@mail.gmail.com>

On 9/3/06, Ganssauge, Gottfried <Gottfried.Ganssauge at haufe.de> wrote:
> > > 2.) docstring extraction is not really usable at the moment
> > >     consider the attached test case
> > >        [test1.hpp, pypp_test1.py]
> > >
> > >     When using the "latin1_doc_string" doc_extractor some unicode
> > > exception is thrown.

I added this to the project "todo" list. I hope it will be addressed
before next release.

-- 
Roman Yakovenko
C++ Python language binding
http://www.language-binding.net/


From ganssauge at gmx.de  Sun Sep  3 23:06:42 2006
From: ganssauge at gmx.de (=?ISO-8859-1?Q?Gottfried_Gan=DFauge?=)
Date: Sun, 03 Sep 2006 23:06:42 +0200
Subject: [C++-sig] Py++ wrapper generation and docstrings
In-Reply-To: <7465b6170609031114q4d906763y9b61fd5007df2597@mail.gmail.com>
References: <7465b6170609030548o13421632if3e04af9d35bd248@mail.gmail.com>	<51F4AD92DB06E9488BBC39B7B492409E0162AAB0@vg20ex1.grp.haufemg.com>
	<7465b6170609031114q4d906763y9b61fd5007df2597@mail.gmail.com>
Message-ID: <44FB43E2.8030408@gmx.de>

Roman Yakovenko schrieb:
> On 9/3/06, Ganssauge, Gottfried <Gottfried.Ganssauge at haufe.de> wrote:
>>>>      Now suppose the following:
>>>>          type char array_t [7];
>>>>
>>>>            int search2 (void *instance, callback_t cb,
>>> array_t *result);
>>>
>>> If you already create a wrapper, you can create  "better" wrapper:
>>>
>>> boost::python::tuple search2( void* instance, callback_t cb ){
>>>     ...
>>>     return boost::python::make_tuple( .... )
>>> }
>> That's what I'd like to achieve - but automatically!
> 
> Lets be more productive. Can you create small example, that you want to export
> and Py++ script and I will make it work?
Don't get me wrong, I don't want you to do my job - in principle it
already works using the generator module I posted.
What I wanted to achieve is interoperability with Py++'s infrastructure
(I figure I'm sharing goals with the pypp project here ...)

Please see test1* attachments.
What I have is:
	test1_bindings.cpp are generated
What I want is:
	test1_bindings.cpp.wanted

...
>> Eventually I should really consider doing several generator passes to a
>> temp file until the temp file doesn't change anymore.
>> This could take quite some time because gccxml must parse the
>> intermediate files...
>> Furthermore I'm not yet sure how to forward the original function names
>> ...
> 
> I still fill that I don't have enough information to help you. For the
> last problem I
> can give you small hint: mangle generated function name.
You can see that in action on my sample project already - I simply add
"_wrapper". Momentarily I'm not using that regularity but modify the
original declaration instead.

> 
> Take a look on boost_random example:
> http://tinyurl.com/gwc2t
> http://tinyurl.com/kocap
will do
...

>>>> 2.) docstring extraction is not really usable at the moment
>>>>     consider the attached test case
>>>>        [test1.hpp, pypp_test1.py]
...

> 
>> Currently I'm using
>>     string_escapes = re.compile (r'((\\x[a-f0-9][a-f0-9])|(\\*"))',
>> re.I)
>>     def make_cstring (s):
>>         def replace_escape(m):
>>             g = m.group(1)
>>             if g.startswith ('\\x'):
>>                 return g + '""'
>>             elif g == '"':
>>                 return '\\"'
>>             else:
>>                 return g
>>
>>         return '"%s"' % hex_escape.sub (replace_escape, repr (s) [1:-1])
>>
>> This copes with the problem of hex escapes followed by valid hex
>> characters and with embedded quotes.
> 
> May I copy this code to Py++ contrib package under the license(
> http://www.boost.org/more/license_info.html )
Please do!

> 
>>> It seems that you found work around to "pointer to function" problem.
>>> Can you create
>>> small example with explanation how it should be done. May be I will
>>> able to introduce
>>> this functionality to Py++.
...

> Thank you for explanation.
> 
> May I copy this code to Py++ contrib package under the license(
> http://www.boost.org/more/license_info.html )
Of course, please do so.

Cheers,

Gottfried
-------------- next part --------------
An embedded and charset-unspecified text was scrubbed...
Name: test1.hpp
URL: <http://mail.python.org/pipermail/cplusplus-sig/attachments/20060903/39161f6e/attachment.txt>
-------------- next part --------------
An embedded and charset-unspecified text was scrubbed...
Name: test1_bindings.cpp
URL: <http://mail.python.org/pipermail/cplusplus-sig/attachments/20060903/39161f6e/attachment-0001.txt>
-------------- next part --------------
An embedded and charset-unspecified text was scrubbed...
Name: test1_bindings.cpp.wanted
URL: <http://mail.python.org/pipermail/cplusplus-sig/attachments/20060903/39161f6e/attachment-0002.txt>

From roman.yakovenko at gmail.com  Mon Sep  4 08:09:36 2006
From: roman.yakovenko at gmail.com (Roman Yakovenko)
Date: Mon, 4 Sep 2006 09:09:36 +0300
Subject: [C++-sig] Py++ wrapper generation and docstrings
In-Reply-To: <44FB43E2.8030408@gmx.de>
References: <7465b6170609030548o13421632if3e04af9d35bd248@mail.gmail.com>
	<51F4AD92DB06E9488BBC39B7B492409E0162AAB0@vg20ex1.grp.haufemg.com>
	<7465b6170609031114q4d906763y9b61fd5007df2597@mail.gmail.com>
	<44FB43E2.8030408@gmx.de>
Message-ID: <7465b6170609032309vfab77bbp9ecbce481343395a@mail.gmail.com>

On 9/4/06, Gottfried Gan?auge <ganssauge at gmx.de> wrote:
> Don't get me wrong, I don't want you to do my job - in principle it
> already works using the generator module I posted.

There is an ongoing discussion about improving Py++ usability by
introducing DSL.
Problem definition:
http://www.language-binding.net/pyplusplus/peps/dsl_challenge.html
Wiki page:
https://realityforge.vrsource.org/view/PyppApi/DslDiscussion

So, I would like to see how you use Py++. It will help me to make it better.

> What I wanted to achieve is interoperability with Py++'s infrastructure
> (I figure I'm sharing goals with the pypp project here ...)
>
> Please see test1* attachments.
> What I have is:
>         test1_bindings.cpp are generated
> What I want is:
>         test1_bindings.cpp.wanted

You did not attached files :-(

P.S. Thank you for contribution


-- 
Roman Yakovenko
C++ Python language binding
http://www.language-binding.net/


From baas at ira.uka.de  Mon Sep  4 17:43:01 2006
From: baas at ira.uka.de (Matthias Baas)
Date: Mon, 04 Sep 2006 17:43:01 +0200
Subject: [C++-sig] What if a Python override returns several values?
Message-ID: <edhhnt$khq$1@sea.gmane.org>

Hi,

how can I obtain the result tuple from calling a Python override that 
returns several values?

The following example is from the tutorial and works fine as long as the 
Python implementation of f returns an integer.

struct BaseWrap : Base, wrapper<Base>
{
     int f()
     {
         return this->get_override("f")();
     }
};

But if I have another virtual member function "void g(int& a, int &b)" 
which returns two integers, then I'd like the Python version to take no 
arguments and return a tuple (a,b). But how does the wrapper look like 
in this case? I would have expected something like:

     void g(int& a, int& b)
     {
         boost::python::tuple tup = this->get_override("g")();

	a = ...extract an int from tup[0]...
         b = ...extract an int from tup[1]...
     }

But when I run this code I get the following error at runtime:

TypeError: No registered converter was able to extract a C++ reference 
to type boost::python::tuple from this Python object of type tuple


So how is this dealt with using the "wrapper" class?

(This is Boost.Python v1.33.1 on Linux)

- Matthias -



From kir at lapshin.net  Tue Sep  5 17:28:05 2006
From: kir at lapshin.net (Kirill Lapshin)
Date: Tue, 05 Sep 2006 19:28:05 +0400
Subject: [C++-sig] [Py++] First imressions
Message-ID: <44FD9785.9000204@lapshin.net>

Hello,

Recently I was charged with exposing our project to Python, and 
naturally went to Boost.Python documentation. After digging a bit, 
learning, etc, I decided to give some code generator a try -- so the 
choice was Pyste or Py++. I've decided to try the latter one first, 
since it seems to have quite powerful query language, and was offering a 
promise of doing certain operations in bulk. Another reason was that 
Py++ is actively maintained. Long story short, I dived in, and I haven't 
look back. I am quite happy with progress, albeit the initial 
development is not finished yet, so it is not a success story yet, but I 
am quite confident it will become one quite soon.

I would like to thank Roman for his hard work on Py++ and for being 
extremely helpful in our private e-mail discussion.

Just for the record I would like to post some of the problems I had and 
the way they were resolved.


1. The way overrides are generated does not compile on MSVC8 (2005). 
Looks like the problem is in conversion of return type if return type is 
bool or enum. For example


virtual bool is_representative( ) const {
bp::override func_is_representative = this->get_override(
"is_representative" );
return func_is_representative( ); // errors out
}

error C2440: 'return' : cannot convert from
'boost::python::detail::method_result' to 'bool'
No user-defined-conversion operator available that can perform this
conversion, or the operator cannot be called

Don't know why is this happening, but the good news, there is a 
workaround. One can use explicit conversion of return type, i.e.

return func_is_representative()..as(&boost::type<T>());

If someone could verify this problem, I guess the solution would be to 
patch code creator.


2. How to disable function overriding.

I don't really need ability to override virtual functions from Python 
anyway, and moreover using wrapper classes does not seem to work for me, 
since I am trying to expose abstract interfaces to python and hide 
concrete implementations (hence wrapper is abstract as well and can't be 
instantiated). Maybe there is still a way to expose it with wrapper, but 
I've decided not bother for now.

Roman suggested following way to suppress overriding:


from pygccxml import declarations

members = cls.decls( declarations.virtuality_type_matcher( 
declarations.VIRTUALITY_TYPES.VIRTUAL )
                                 , decl_type=pd.member_calldef_t
                                 , allow_empty=True)
members.set_virtuality( declarations.VIRTUALITY_TYPES.NOT_VIRTUAL )


Note that Py++ differentiates between VIRTUAL and PURE_VIRTUAL, so above 
listed code does not cover pure virtual functions.

3. Observation: Py++ generates wrappers even if there is no virtual 
functions, but there is nested class/struct. Nested enums are Ok.

4. Gccxml is buggy, especially with respect to default arguments. See 
http://language-binding.net/pygccxml/design.html#patchers. In my case I 
had problems with default std::vector<T> arguments. Those can be worked 
around manually with

mb = module_builder_t( ... )
Foo = mb.class_( 'Foo' )
Foo_constructor = Foo.constructor( select write constructor )
#help: http://language-binding.net/pygccxml/query_interface.html

Foo_constructor.arguments[0].default_value = fixed default value

5. Precompiled header support is not implemented even though 
documentation claims it is. Roman says it is in works. For a meantime 
one can use following workaround:

mb.code_creator.adopt_creator(
code_creators.include_t(header='_pch.hpp'), 0)
md.split_module(...)

6. shared_ptr<T> works out of the box, while shared_ptr<const T> does 
not. Roman gave me following pointers:

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

Also read this:
http://language-binding.net/pygccxml/apidocs/pygccxml.declarations.type_traits.smart_pointer_traits-class.html

Now you can implement some general solution similar to
http://python-ogre.python-hosting.com/file/trunk/python-ogre/shared_ptr.py

I did not have time to try it yet, but planning to do it soon.

Cheers,
Kirill



From kir at lapshin.net  Tue Sep  5 17:40:31 2006
From: kir at lapshin.net (Kirill Lapshin)
Date: Tue, 05 Sep 2006 19:40:31 +0400
Subject: [C++-sig] [Py++] Nested enum as default arg
Message-ID: <edk5pf$kha$1@sea.gmane.org>

Hi,

Suppose I have

struct Foo
{
     enum Bar {BAR1, BAR2};

     explicit Foo(Bar arg = BAR1);
};


Py++ generates code that tries to add constructor (and default argument) 
prior to enum being registered. Result is exception in module import.

Any ideas how to fix / work around? Other then moving enum outside the 
class?

Cheers,
Kirill



From amohr at pixar.com  Tue Sep  5 18:53:54 2006
From: amohr at pixar.com (Alex Mohr)
Date: Tue, 05 Sep 2006 09:53:54 -0700
Subject: [C++-sig] What if a Python override returns several values?
In-Reply-To: <edhhnt$khq$1@sea.gmane.org>
References: <edhhnt$khq$1@sea.gmane.org>
Message-ID: <44FDABA2.60809@pixar.com>

> But if I have another virtual member function "void g(int& a, int &b)" 
> which returns two integers, then I'd like the Python version to take no 
> arguments and return a tuple (a,b). But how does the wrapper look like 
> in this case? I would have expected something like:
> 
>      void g(int& a, int& b)
>      {
>          boost::python::tuple tup = this->get_override("g")();
> 
> 	a = ...extract an int from tup[0]...
>          b = ...extract an int from tup[1]...
>      }
> 
> But when I run this code I get the following error at runtime:
> 
> TypeError: No registered converter was able to extract a C++ reference 
> to type boost::python::tuple from this Python object of type tuple

Not sure exactly what's going on here, but you might be able to find out 
by looking at override.hpp.  Look at method_result -- that's what you're 
getting back from get_override.

Off the top of my head, I think this would probably work for you:

tuple tup = call<tuple>(this->get_override("g").ptr());
a = ...extract an int from tup[0]
b = ...extract an int from tup[1]

It's certainly not as nice as what you wrote, but I have a feeling it 
might make the conversions work right.  The idea here is that the 
boost::python::override object manages the python callable object, so 
you can get it (using ptr()) and call it any way you like.

Alex


From baas at ira.uka.de  Tue Sep  5 19:30:27 2006
From: baas at ira.uka.de (Matthias Baas)
Date: Tue, 05 Sep 2006 19:30:27 +0200
Subject: [C++-sig] What if a Python override returns several values?
In-Reply-To: <44FDABA2.60809@pixar.com>
References: <edhhnt$khq$1@sea.gmane.org> <44FDABA2.60809@pixar.com>
Message-ID: <edkcdo$f1i$1@sea.gmane.org>

Alex Mohr wrote:
>>      void g(int& a, int& b)
>>      {
>>          boost::python::tuple tup = this->get_override("g")();
>>
>> 	a = ...extract an int from tup[0]...
>>          b = ...extract an int from tup[1]...
>>      }
>>
>> But when I run this code I get the following error at runtime:
>>
>> TypeError: No registered converter was able to extract a C++ reference 
>> to type boost::python::tuple from this Python object of type tuple
> 
> [...]
> Off the top of my head, I think this would probably work for you:
> 
> tuple tup = call<tuple>(this->get_override("g").ptr());

Excellent! This works, thanks!

- Matthias -



From roman.yakovenko at gmail.com  Tue Sep  5 20:00:12 2006
From: roman.yakovenko at gmail.com (Roman Yakovenko)
Date: Tue, 5 Sep 2006 21:00:12 +0300
Subject: [C++-sig] [Py++] Nested enum as default arg
In-Reply-To: <edk5pf$kha$1@sea.gmane.org>
References: <edk5pf$kha$1@sea.gmane.org>
Message-ID: <7465b6170609051100x745f016fxde34980bc7021767@mail.gmail.com>

On 9/5/06, Kirill Lapshin <kir at lapshin.net> wrote:
> Hi,
>
> Suppose I have
>
> struct Foo
> {
>      enum Bar {BAR1, BAR2};
>
>      explicit Foo(Bar arg = BAR1);
> };
>
>
> Py++ generates code that tries to add constructor (and default argument)
> prior to enum being registered. Result is exception in module import.
>
> Any ideas how to fix / work around? Other then moving enum outside the
> class?

mb = module_builder_t( ... )
Foo = mb.constructor( 'Foo' ) #I assume that there is only one constructor
Foo.arguments[0].default_value = 'int(%s)' % Foo.arguments[0].default_value


You are not the first one who has this problem. Another work around is:
using boost::python;
class_< Foo > foo( "Foo", no_init );
scope foo_scope( foo );
enum_< ... >...;
foo_scope.def( init< ... >( ... ) );

If you search this mailing list you will find few discussions about
this problem.
I hope I will be able to submit patch to the library that will solve "default
arguments" dependency problem.


-- 
Roman Yakovenko
C++ Python language binding
http://www.language-binding.net/


From roman.yakovenko at gmail.com  Tue Sep  5 22:03:23 2006
From: roman.yakovenko at gmail.com (Roman Yakovenko)
Date: Tue, 5 Sep 2006 23:03:23 +0300
Subject: [C++-sig] custom smart pointers
In-Reply-To: <7465b6170608311358s1c02c6c9hc26222fae8cfdef6@mail.gmail.com>
References: <7465b6170608291048y6e2010d4ofa84fbd5760da4d3@mail.gmail.com>
	<uac5lvczx.fsf@boost-consulting.com>
	<7465b6170608311358s1c02c6c9hc26222fae8cfdef6@mail.gmail.com>
Message-ID: <7465b6170609051303w4132f343vc24d4cfd768408ed@mail.gmail.com>

On 8/31/06, Roman Yakovenko <roman.yakovenko at gmail.com> wrote:
> On 8/31/06, David Abrahams <dave at boost-consulting.com> wrote:
> > >From this I can't understand what you're doing or what errors you're
> > getting.  Please do what I tell everyone to do: make a _complete_,
> > _minimal_ example that reproduces your problem and post that.
>
> I attached 3 files. This is complete and minimal example.
>
> custom_smart_ptr.h
>   file contains smart pointer class definition.
>
> custom_smart_ptr_classes_to_be_exported.hpp
>    file contains definition of 2 classes:  "base" and "derived" classes.
>    This file also defines "get_value" function. this function is
> called from Python.
>
> custom_smart_ptr_classes.cpp
>    file contains Boost.Python bindings.
>
> Python code is next:
> import custom_smart_ptr_classes as cspc
> d = cspc.derived( 2 )
> cspc.get_value( d )
>
> The last line raises next exception:
>
> ArgumentError: Python argument types in
>     custom_smart_ptr_classes.get_value(derived)
> did not match C++ signature:
>     get_value(my_smart_ptr_t<base> a)
>
> Thank you.

Can somebody help me with this issue?

Thanks in advance.

-- 
Roman Yakovenko
C++ Python language binding
http://www.language-binding.net/


From amohr at pixar.com  Tue Sep  5 22:20:20 2006
From: amohr at pixar.com (Alex Mohr)
Date: Tue, 05 Sep 2006 13:20:20 -0700
Subject: [C++-sig] custom smart pointers
In-Reply-To: <7465b6170609051303w4132f343vc24d4cfd768408ed@mail.gmail.com>
References: <7465b6170608291048y6e2010d4ofa84fbd5760da4d3@mail.gmail.com>
	<uac5lvczx.fsf@boost-consulting.com>
	<7465b6170608311358s1c02c6c9hc26222fae8cfdef6@mail.gmail.com>
	<7465b6170609051303w4132f343vc24d4cfd768408ed@mail.gmail.com>
Message-ID: <44FDDC04.3010200@pixar.com>

>> Python code is next:
>> import custom_smart_ptr_classes as cspc
>> d = cspc.derived( 2 )
>> cspc.get_value( d )
>>
>> The last line raises next exception:
>>
>> ArgumentError: Python argument types in
>>     custom_smart_ptr_classes.get_value(derived)
>> did not match C++ signature:
>>     get_value(my_smart_ptr_t<base> a)
>>
>> Thank you.
> 
> Can somebody help me with this issue?
> 
> Thanks in advance.

In my experience, boost python needs a lot of hand holding to work well 
with custom smart pointers.

I think the first thing you need here is an rvalue from-python 
conversion for your smart pointers.  Register a conversion for each 
smart_ptr<T> that you expose to python.  And make sure you do it for 
smart_ptr<T>, not smart_ptr<wrapper<T> > in the case of polymorphic stuff.

I (semi)automate this in a custom def visitor.

Alex


From abierbaum at gmail.com  Wed Sep  6 01:09:12 2006
From: abierbaum at gmail.com (Allen Bierbaum)
Date: Tue, 5 Sep 2006 18:09:12 -0500
Subject: [C++-sig] Is there a way to automatically convert a smart_ptr
	to a held_type (try 2)
In-Reply-To: <44F5CD10.5010103@pixar.com>
References: <bea9869d0608300707r3009cdb2u101ab96b88ec9966@mail.gmail.com>
	<44F5CD10.5010103@pixar.com>
Message-ID: <bea9869d0609051609n7f6760f3jd43b7cea8c3d66dd@mail.gmail.com>

On 8/30/06, Alex Mohr <amohr at pixar.com> wrote:
> > What I need is a way to tell boost python that anytime it needs to
> > convert a ptr<Base> to python it should first convert it to a
> > ref_ptr<Base>  (using a standard copy constructor in ref_ptr<>) and
> > then use this ref_ptr as the held type like normal.
> >
> > At first I thought that implicitly_convertible<> would help:
> >
> > implicitly_convertible<ptr<Base>, ref_ptr<Base> >();
> >
> > But this doesn't seem to do anything for me. (this really isn't too
> > surprising since this isn't what implicitly_convertible is for, but it
> > seemed worth a try).
> >
> > Has anyone else ran into anything similar?  Any ideas?
>
> At first glance, it seems like you want to register a to-python
> conversion for all your ptr<T> types.  Note that all code I write in
> this message is untested:
>
> template <class T>
> struct ptr_to_python {
>      static PyObject *convert(ptr<T> const &p) {
>          return Py_INCREF(object(ref_ptr<T>(p)).ptr());
>      }
> };
>
> Then you need to do this for each class you register:
>
> to_python_converter<ptr<T>, ptr_to_python<T> >();
>

Thanks Alex, this worked great.  I had to change the Py_INCREF to a
bp::incref but otherwise things are all good.

But now I have a rather strange followup question.

It seems that Boost.Python is doing some magic behind the scenes to
convert an object coming from C++ as a smart pointer whose wrapped
pointer is NULL into a Python "None".  The is great, and was a
wonderful surprise.  But now I need to find a way to make it happen in
the other direction.

How does boost.python convert a parameter of "None" into an
object/smart_ptr to pass to C++?

I need to find a way to call a C++ method like:

void setObject(ptr<Base> obj);

from Python and call it in Python as "setObject(None)" and instruct
boost.python to convert this into a call behind the scenes in C++ to:

setObject(NullPtr);

where "NullPtr" is a global const value defined by the library that is
used to signify a NULL shared pointer throughout the system.  (the
reason it needs something like this is a little complex, but it is not
my library so I can't change this part much.)

Alternatively, I could probably make it work if I could get
boost.python to make a call like:

setObject(ref_ptr<Base>(NULL) )

as I think I could get the library to automatically convert
ref_ptr<Base>(NULL) into a NullPtr.

Any ideas?


-Allen


From amohr at pixar.com  Wed Sep  6 01:39:30 2006
From: amohr at pixar.com (Alex Mohr)
Date: Tue, 05 Sep 2006 16:39:30 -0700
Subject: [C++-sig] Is there a way to automatically convert a smart_ptr
 to a held_type (try 2)
In-Reply-To: <bea9869d0609051609n7f6760f3jd43b7cea8c3d66dd@mail.gmail.com>
References: <bea9869d0608300707r3009cdb2u101ab96b88ec9966@mail.gmail.com>
	<44F5CD10.5010103@pixar.com>
	<bea9869d0609051609n7f6760f3jd43b7cea8c3d66dd@mail.gmail.com>
Message-ID: <44FE0AB2.8010006@pixar.com>

> Thanks Alex, this worked great.  I had to change the Py_INCREF to a
> bp::incref but otherwise things are all good.
> 
> But now I have a rather strange followup question.
> 
> It seems that Boost.Python is doing some magic behind the scenes to
> convert an object coming from C++ as a smart pointer whose wrapped
> pointer is NULL into a Python "None".  The is great, and was a
> wonderful surprise.  But now I need to find a way to make it happen in
> the other direction.
> 
> How does boost.python convert a parameter of "None" into an
> object/smart_ptr to pass to C++?

It happens in the from_python conversion.  It just checks for None and 
(I believe) default-constructs a the pointer object.  Is it the case 
that default-constructing one of your smart pointers does not produce 
the intended NULL?  It probably wouldn't be hard to add a customization 
(paralleling get_pointer and pointee) to boost python.  But in the 
meantime...

> I need to find a way to call a C++ method like:
> 
> void setObject(ptr<Base> obj);
> 
> from Python and call it in Python as "setObject(None)" and instruct
> boost.python to convert this into a call behind the scenes in C++ to:
> 
> setObject(NullPtr);
> 
> where "NullPtr" is a global const value defined by the library that is
> used to signify a NULL shared pointer throughout the system.  (the
> reason it needs something like this is a little complex, but it is not
> my library so I can't change this part much.)
> 
> Alternatively, I could probably make it work if I could get
> boost.python to make a call like:
> 
> setObject(ref_ptr<Base>(NULL) )
> 
> as I think I could get the library to automatically convert
> ref_ptr<Base>(NULL) into a NullPtr.
> 
> Any ideas?

I think you can do either version by registering your own rvalue 
from_python converters for all your pointer types.  You can do this in 
the def_visitor along with the to_python conversion we discussed earlier.

Just replace the relevant bits below ("Deal with the "None" case") with 
whatever you want.

Again, I'm just typing this so it may not compile, and no guarantees of 
course...

template <class Ptr>
struct smart_ptr_from_python {
     typedef typename boost::python::pointee<Ptr>::type Pointee;
     smart_ptr_from_python() {
         converter::registry::insert(&convertible, &construct,
                                     type_id<Ptr>());
     }
   private:
     static void *convertible(PyObject *p) {
         // can always produce a pointer from None.
         if (p == Py_None)
             return p;
         // Otherwise, we can do it if we can get the pointee out.
         void *result = converter::get_lvalue_from_python
             (p, converter::registered<Pointee>::converters);
         return result;
     }

     static void construct(PyObject* source, converter::
                           rvalue_from_python_stage1_data* data) {
         void* const storage =
             ((converter::rvalue_from_python_storage<Ptr>*)data)->
             storage.bytes;
         // Deal with the "None" case.
         if (data->convertible == source)
             new (storage) Ptr();  // Or whatever you want.
         else
             new (storage) 
Ptr(static_cast<Pointee*>(data->convertible));
         data->convertible = storage;
     }
};


Alex


From abierbaum at gmail.com  Wed Sep  6 03:44:38 2006
From: abierbaum at gmail.com (Allen Bierbaum)
Date: Tue, 5 Sep 2006 20:44:38 -0500
Subject: [C++-sig] Is there a way to automatically convert a smart_ptr
	to a held_type (try 2)
In-Reply-To: <44FE0AB2.8010006@pixar.com>
References: <bea9869d0608300707r3009cdb2u101ab96b88ec9966@mail.gmail.com>
	<44F5CD10.5010103@pixar.com>
	<bea9869d0609051609n7f6760f3jd43b7cea8c3d66dd@mail.gmail.com>
	<44FE0AB2.8010006@pixar.com>
Message-ID: <bea9869d0609051844m5c400acdt8e4c53f8b608500f@mail.gmail.com>

On 9/5/06, Alex Mohr <amohr at pixar.com> wrote:
> > Thanks Alex, this worked great.  I had to change the Py_INCREF to a
> > bp::incref but otherwise things are all good.
> >
> > But now I have a rather strange followup question.
> >
> > It seems that Boost.Python is doing some magic behind the scenes to
> > convert an object coming from C++ as a smart pointer whose wrapped
> > pointer is NULL into a Python "None".  The is great, and was a
> > wonderful surprise.  But now I need to find a way to make it happen in
> > the other direction.
> >
> > How does boost.python convert a parameter of "None" into an
> > object/smart_ptr to pass to C++?
>
> It happens in the from_python conversion.  It just checks for None and
> (I believe) default-constructs a the pointer object.  Is it the case
> that default-constructing one of your smart pointers does not produce
> the intended NULL?  It probably wouldn't be hard to add a customization
> (paralleling get_pointer and pointee) to boost python.  But in the
> meantime...

I will try to find this in the boost.python code and verify the behavior.

>
> > I need to find a way to call a C++ method like:
> >
> > void setObject(ptr<Base> obj);
> >
> > from Python and call it in Python as "setObject(None)" and instruct
> > boost.python to convert this into a call behind the scenes in C++ to:
> >
> > setObject(NullPtr);
> >
> > where "NullPtr" is a global const value defined by the library that is
> > used to signify a NULL shared pointer throughout the system.  (the
> > reason it needs something like this is a little complex, but it is not
> > my library so I can't change this part much.)
> >
> > Alternatively, I could probably make it work if I could get
> > boost.python to make a call like:
> >
> > setObject(ref_ptr<Base>(NULL) )
> >
> > as I think I could get the library to automatically convert
> > ref_ptr<Base>(NULL) into a NullPtr.
> >
> > Any ideas?
>
> I think you can do either version by registering your own rvalue
> from_python converters for all your pointer types.  You can do this in
> the def_visitor along with the to_python conversion we discussed earlier.
>
> Just replace the relevant bits below ("Deal with the "None" case") with
> whatever you want.
>
> Again, I'm just typing this so it may not compile, and no guarantees of
> course...

All I can say is "wow!!".  You seem to understand the conversion
process inside boost.python much more then I do.  I will give these
methods a try tomorrow.  I will need to alter them slightly since I am
not using the def_visitor method you outlined.  (I am using Py++ to
generate the bindings so it really doesn't take any more typing to
just have it put the registration calls in the right place).  I will
let you know how it goes.

As a side note, if you don't mind me asking, how did you learn so much
about how the conversion process works.  I have always had trouble
understanding the full process and am looking for any advice about how
to learn it better.  (did you learn from the code, the API reference,
the mailing list, other??)  If you ever have any time or interest and
wanted to contribute some to a little documentation effort I am
working on in my spare time, take a look at http://tinyurl.com/zb76h .

Thanks again.

-Allen


>
> template <class Ptr>
> struct smart_ptr_from_python {
>      typedef typename boost::python::pointee<Ptr>::type Pointee;
>      smart_ptr_from_python() {
>          converter::registry::insert(&convertible, &construct,
>                                      type_id<Ptr>());
>      }
>    private:
>      static void *convertible(PyObject *p) {
>          // can always produce a pointer from None.
>          if (p == Py_None)
>              return p;
>          // Otherwise, we can do it if we can get the pointee out.
>          void *result = converter::get_lvalue_from_python
>              (p, converter::registered<Pointee>::converters);
>          return result;
>      }
>
>      static void construct(PyObject* source, converter::
>                            rvalue_from_python_stage1_data* data) {
>          void* const storage =
>              ((converter::rvalue_from_python_storage<Ptr>*)data)->
>              storage.bytes;
>          // Deal with the "None" case.
>          if (data->convertible == source)
>              new (storage) Ptr();  // Or whatever you want.
>          else
>              new (storage)
> Ptr(static_cast<Pointee*>(data->convertible));
>          data->convertible = storage;
>      }
> };
>
>
> Alex
> _______________________________________________
> C++-sig mailing list
> C++-sig at python.org
> http://mail.python.org/mailman/listinfo/c++-sig
>


From roman.yakovenko at gmail.com  Wed Sep  6 08:16:33 2006
From: roman.yakovenko at gmail.com (Roman Yakovenko)
Date: Wed, 6 Sep 2006 09:16:33 +0300
Subject: [C++-sig] custom smart pointers
In-Reply-To: <44FDDC04.3010200@pixar.com>
References: <7465b6170608291048y6e2010d4ofa84fbd5760da4d3@mail.gmail.com>
	<uac5lvczx.fsf@boost-consulting.com>
	<7465b6170608311358s1c02c6c9hc26222fae8cfdef6@mail.gmail.com>
	<7465b6170609051303w4132f343vc24d4cfd768408ed@mail.gmail.com>
	<44FDDC04.3010200@pixar.com>
Message-ID: <7465b6170609052316w887a0b5v3723de9d1833491e@mail.gmail.com>

On 9/5/06, Alex Mohr <amohr at pixar.com> wrote:
> In my experience, boost python needs a lot of hand holding to work well
> with custom smart pointers.
>
> I think the first thing you need here is an rvalue from-python
> conversion for your smart pointers.

I don't know and understand how I should do it, really :-(.

I took a look on shared_ptr_from_python class. I understand the code
which is written
in construct function.

static void construct(PyObject* source, rvalue_from_python_stage1_data* data)
{

...
new (storage) shared_ptr<T>(
                static_cast<T*>(data->convertible),
                shared_ptr_deleter(handle<>(borrowed(source)))
                );
...
}

Boost.Python creates new instance of shared_ptr with custom deleter. This way it
says to shared_ptr that the object is already managed.

The smart pointer, the project use, does not provide this
functionality. More over,
I don't understand what data->convertible will contain.


> Register a conversion for each
> smart_ptr<T> that you expose to python.  And make sure you do it for
> smart_ptr<T>, not smart_ptr<wrapper<T> > in the case of polymorphic stuff.

I think I do it:

register_ptr_to_python< my_smart_ptr_t< base > >();
register_ptr_to_python< my_smart_ptr_t< derived > >();
implicitly_convertible< my_smart_ptr_t< derived >, my_smart_ptr_t< base > >();

This is what yoy mean, right?


Many thanks for your help.

-- 
Roman Yakovenko
C++ Python language binding
http://www.language-binding.net/


From amohr at pixar.com  Wed Sep  6 18:36:08 2006
From: amohr at pixar.com (Alex Mohr)
Date: Wed, 06 Sep 2006 09:36:08 -0700
Subject: [C++-sig] custom smart pointers
In-Reply-To: <7465b6170609052316w887a0b5v3723de9d1833491e@mail.gmail.com>
References: <7465b6170608291048y6e2010d4ofa84fbd5760da4d3@mail.gmail.com>
	<uac5lvczx.fsf@boost-consulting.com>
	<7465b6170608311358s1c02c6c9hc26222fae8cfdef6@mail.gmail.com>
	<7465b6170609051303w4132f343vc24d4cfd768408ed@mail.gmail.com>
	<44FDDC04.3010200@pixar.com>
	<7465b6170609052316w887a0b5v3723de9d1833491e@mail.gmail.com>
Message-ID: <44FEF8F8.8080004@pixar.com>

> I don't know and understand how I should do it, really :-(.
> 
> Boost.Python creates new instance of shared_ptr with custom deleter. This way it
> says to shared_ptr that the object is already managed.

I believe it's just a hack to do what I call "object identity" -- which 
is just tracking the association between a C++ instance and a python 
object.  When converting a shared ptr to-python, boost python checks to 
see if it has its special deleter, and will use that to produce the 
python object rather than constructing a new one.

This is a fragile scheme.  It fails frequently for shared ptrs, as has 
been discussed on this list before, and it doesn't work at all for 
custom shared pointers.

Anyway, I consider object identity to be a different issue from the one 
we're currently discussing.  But I do think you should address it in 
py++.  We can talk about that later.

> The smart pointer, the project use, does not provide this
> functionality. More over,
> I don't understand what data->convertible will contain.

If I recall correctly (I'm not looking at the code) data->convertible is 
the result that the convertible method returned, so in this case it's a 
raw pointer to the C++ instance.

> I think I do it:
> 
> register_ptr_to_python< my_smart_ptr_t< base > >();
> register_ptr_to_python< my_smart_ptr_t< derived > >();
> implicitly_convertible< my_smart_ptr_t< derived >, my_smart_ptr_t< base > >();
> 
> This is what yoy mean, right?

No, these are to-python conversions.  I'm talking about rvalue 
from-python conversions.  Have a look at the last few messages in the 
concurrent thread on this list with subject "Is there a way to
automatically convert a smart_ptr to a held_type (try 2)".

I posted a message to that thread which has some (untested) smart 
pointer rvalue from-python conversion code in it.  Actually -- it 
probably looks something like shared_ptr_from_python.

> Many thanks for your help.

You're welcome.

Alex


From amohr at pixar.com  Wed Sep  6 18:42:37 2006
From: amohr at pixar.com (Alex Mohr)
Date: Wed, 06 Sep 2006 09:42:37 -0700
Subject: [C++-sig] Is there a way to automatically convert a smart_ptr
 to a held_type (try 2)
In-Reply-To: <bea9869d0609051844m5c400acdt8e4c53f8b608500f@mail.gmail.com>
References: <bea9869d0608300707r3009cdb2u101ab96b88ec9966@mail.gmail.com>
	<44F5CD10.5010103@pixar.com>
	<bea9869d0609051609n7f6760f3jd43b7cea8c3d66dd@mail.gmail.com>
	<44FE0AB2.8010006@pixar.com>
	<bea9869d0609051844m5c400acdt8e4c53f8b608500f@mail.gmail.com>
Message-ID: <44FEFA7D.8020606@pixar.com>

> As a side note, if you don't mind me asking, how did you learn so much
> about how the conversion process works.  I have always had trouble
> understanding the full process and am looking for any advice about how
> to learn it better.  (did you learn from the code, the API reference,
> the mailing list, other??)  If you ever have any time or interest and
> wanted to contribute some to a little documentation effort I am
> working on in my spare time, take a look at http://tinyurl.com/zb76h .

I learned from a combination of all the things you mention.  I currently 
find reading the code to be the most useful.  I don't know if there's a 
better resource, other than getting David to respond to your questions 
directly if you can. :)

Alex


From abierbaum at gmail.com  Wed Sep  6 19:02:06 2006
From: abierbaum at gmail.com (Allen Bierbaum)
Date: Wed, 6 Sep 2006 12:02:06 -0500
Subject: [C++-sig] Is there a way to automatically convert a smart_ptr
	to a held_type (try 2)
In-Reply-To: <44FEFA7D.8020606@pixar.com>
References: <bea9869d0608300707r3009cdb2u101ab96b88ec9966@mail.gmail.com>
	<44F5CD10.5010103@pixar.com>
	<bea9869d0609051609n7f6760f3jd43b7cea8c3d66dd@mail.gmail.com>
	<44FE0AB2.8010006@pixar.com>
	<bea9869d0609051844m5c400acdt8e4c53f8b608500f@mail.gmail.com>
	<44FEFA7D.8020606@pixar.com>
Message-ID: <bea9869d0609061002obe70af4qa02df83c7c60ff95@mail.gmail.com>

I've been trying to learn more about the calling and argument
conversion process today by stepping through the code with ddd/gdb.  I
am learning some things but I have to admit it is pretty difficult
with all the preprocessor library and MPL usage (this stuff confuses
the debugger quite a bit).

Hopefully by later today I will understand enough to know what you
code samples are trying to do. :)

-Allen

On 9/6/06, Alex Mohr <amohr at pixar.com> wrote:
> > As a side note, if you don't mind me asking, how did you learn so much
> > about how the conversion process works.  I have always had trouble
> > understanding the full process and am looking for any advice about how
> > to learn it better.  (did you learn from the code, the API reference,
> > the mailing list, other??)  If you ever have any time or interest and
> > wanted to contribute some to a little documentation effort I am
> > working on in my spare time, take a look at http://tinyurl.com/zb76h .
>
> I learned from a combination of all the things you mention.  I currently
> find reading the code to be the most useful.  I don't know if there's a
> better resource, other than getting David to respond to your questions
> directly if you can. :)
>
> Alex
> _______________________________________________
> C++-sig mailing list
> C++-sig at python.org
> http://mail.python.org/mailman/listinfo/c++-sig
>


From abierbaum at gmail.com  Wed Sep  6 21:17:12 2006
From: abierbaum at gmail.com (Allen Bierbaum)
Date: Wed, 6 Sep 2006 14:17:12 -0500
Subject: [C++-sig] Is there a way to automatically convert a smart_ptr
	to a held_type (try 2)
In-Reply-To: <44FE0AB2.8010006@pixar.com>
References: <bea9869d0608300707r3009cdb2u101ab96b88ec9966@mail.gmail.com>
	<44F5CD10.5010103@pixar.com>
	<bea9869d0609051609n7f6760f3jd43b7cea8c3d66dd@mail.gmail.com>
	<44FE0AB2.8010006@pixar.com>
Message-ID: <bea9869d0609061217r60fd8908w78154e5c4ea39bf8@mail.gmail.com>

On 9/5/06, Alex Mohr <amohr at pixar.com> wrote:
> > Thanks Alex, this worked great.  I had to change the Py_INCREF to a
> > bp::incref but otherwise things are all good.
> >
> > But now I have a rather strange followup question.
> >
> > It seems that Boost.Python is doing some magic behind the scenes to
> > convert an object coming from C++ as a smart pointer whose wrapped
> > pointer is NULL into a Python "None".  The is great, and was a
> > wonderful surprise.  But now I need to find a way to make it happen in
> > the other direction.
> >
> > How does boost.python convert a parameter of "None" into an
> > object/smart_ptr to pass to C++?
>
> It happens in the from_python conversion.  It just checks for None and
> (I believe) default-constructs a the pointer object.  Is it the case
> that default-constructing one of your smart pointers does not produce
> the intended NULL?  It probably wouldn't be hard to add a customization
> (paralleling get_pointer and pointee) to boost python.  But in the
> meantime...
>
> > I need to find a way to call a C++ method like:
> >
> > void setObject(ptr<Base> obj);
> >
> > from Python and call it in Python as "setObject(None)" and instruct
> > boost.python to convert this into a call behind the scenes in C++ to:
> >
> > setObject(NullPtr);
> >
> > where "NullPtr" is a global const value defined by the library that is
> > used to signify a NULL shared pointer throughout the system.  (the
> > reason it needs something like this is a little complex, but it is not
> > my library so I can't change this part much.)
> >
> > Alternatively, I could probably make it work if I could get
> > boost.python to make a call like:
> >
> > setObject(ref_ptr<Base>(NULL) )
> >
> > as I think I could get the library to automatically convert
> > ref_ptr<Base>(NULL) into a NullPtr.
> >
> > Any ideas?
>
> I think you can do either version by registering your own rvalue
> from_python converters for all your pointer types.  You can do this in
> the def_visitor along with the to_python conversion we discussed earlier.
>
> Just replace the relevant bits below ("Deal with the "None" case") with
> whatever you want.

I am sorting through this code right now.  Am I missing something or
is there no documentation in the reference guide about how to register
a converter directly with the registry?

Additionally, I was looking at the documentation for
implicitly_convertible and I am wondering why I can't just use that in
my case.

I want to call:

setObject(ptr<Base> val)

with a NoneType.  And I have all my objects (ie. Base) wrapped and set
with held types of ref_ptr<>'s.  RefPtr's are implicitly convertible
to ptr<>'s, so If I register this with boost.python using:

  bp::implicitly_convertible< ret_ptr<Base>, ptr<Base> >();

Then shouldn't boost.python be able to convert NoneType -->
ref_ptr<Base> --> ptr<Base>  ?  From the documentation in
implicitly_convertible it seems like this should be the case.  But I
have to admit I have not found the place in the boost.python code
where it handles the case of taking a NoneType and automatically
converting it into a NULL smart pointer to start with.

-Allen

>
> Again, I'm just typing this so it may not compile, and no guarantees of
> course...
>
> template <class Ptr>
> struct smart_ptr_from_python {
>      typedef typename boost::python::pointee<Ptr>::type Pointee;
>      smart_ptr_from_python() {
>          converter::registry::insert(&convertible, &construct,
>                                      type_id<Ptr>());
>      }
>    private:
>      static void *convertible(PyObject *p) {
>          // can always produce a pointer from None.
>          if (p == Py_None)
>              return p;
>          // Otherwise, we can do it if we can get the pointee out.
>          void *result = converter::get_lvalue_from_python
>              (p, converter::registered<Pointee>::converters);
>          return result;
>      }
>
>      static void construct(PyObject* source, converter::
>                            rvalue_from_python_stage1_data* data) {
>          void* const storage =
>              ((converter::rvalue_from_python_storage<Ptr>*)data)->
>              storage.bytes;
>          // Deal with the "None" case.
>          if (data->convertible == source)
>              new (storage) Ptr();  // Or whatever you want.
>          else
>              new (storage)
> Ptr(static_cast<Pointee*>(data->convertible));
>          data->convertible = storage;
>      }
> };
>
>
> Alex
> _______________________________________________
> C++-sig mailing list
> C++-sig at python.org
> http://mail.python.org/mailman/listinfo/c++-sig
>


From kir at lapshin.net  Wed Sep  6 21:20:02 2006
From: kir at lapshin.net (Kirill Lapshin)
Date: Wed, 06 Sep 2006 23:20:02 +0400
Subject: [C++-sig] [Py++] Nested enum as default arg
In-Reply-To: <7465b6170609051100x745f016fxde34980bc7021767@mail.gmail.com>
References: <edk5pf$kha$1@sea.gmane.org>
	<7465b6170609051100x745f016fxde34980bc7021767@mail.gmail.com>
Message-ID: <edn712$jc4$1@sea.gmane.org>

Roman Yakovenko wrote:
>> struct Foo
>> {
>>      enum Bar {BAR1, BAR2};
>>
>>      explicit Foo(Bar arg = BAR1);
>> };
>>
>>
>> Py++ generates code that tries to add constructor (and default argument)
>> prior to enum being registered. Result is exception in module import.
>>
>> Any ideas how to fix / work around? Other then moving enum outside the
>> class?
> 
> mb = module_builder_t( ... )
> Foo = mb.constructor( 'Foo' ) #I assume that there is only one constructor
> Foo.arguments[0].default_value = 'int(%s)' % Foo.arguments[0].default_value
> 

Oh, I see. But that would only work for enums, wouldn't it? I would 
prefer some more generic solution. In fact I've already encountered 
another similar problem. This time it is more along the lines of:

struct Foo
{
     explicit Foo( const shared_ptr<Foo>& foo = shared_ptr<Foo>() );
};


> You are not the first one who has this problem. Another work around is:
> using boost::python;
> class_< Foo > foo( "Foo", no_init );
> scope foo_scope( foo );
> enum_< ... >...;
> foo_scope.def( init< ... >( ... ) );

That makes a lot of sense, but that requires modification of Py++, 
doesn't it? Or is there a way to customize code generation to do it the 
way you've cited?

Seems to me if Py++ would always opt for code like that, it would 
already solve quite a few dependency problems.

I would be more then happy to implement this. Could you point me in 
right direction, please?

> If you search this mailing list you will find few discussions about
> this problem.
> I hope I will be able to submit patch to the library that will solve "default
> arguments" dependency problem.

Yep, I've found few messages. Looks like you are planning to patch 
boost.python. That would be the ultimate solution! Though it is a long 
term solution, and we also need something in short/mid term...

Kirill



From roman.yakovenko at gmail.com  Wed Sep  6 21:31:32 2006
From: roman.yakovenko at gmail.com (Roman Yakovenko)
Date: Wed, 6 Sep 2006 22:31:32 +0300
Subject: [C++-sig] [Py++] Nested enum as default arg
In-Reply-To: <edn712$jc4$1@sea.gmane.org>
References: <edk5pf$kha$1@sea.gmane.org>
	<7465b6170609051100x745f016fxde34980bc7021767@mail.gmail.com>
	<edn712$jc4$1@sea.gmane.org>
Message-ID: <7465b6170609061231s113a1440h8d435e861621a540@mail.gmail.com>

On 9/6/06, Kirill Lapshin <kir at lapshin.net> wrote:
> Oh, I see. But that would only work for enums, wouldn't it? I would
> prefer some more generic solution. In fact I've already encountered
> another similar problem. This time it is more along the lines of:
>
> struct Foo
> {
>      explicit Foo( const shared_ptr<Foo>& foo = shared_ptr<Foo>() );
> };
>
>
> > You are not the first one who has this problem. Another work around is:
> > using boost::python;
> > class_< Foo > foo( "Foo", no_init );
> > scope foo_scope( foo );
> > enum_< ... >...;
> > foo_scope.def( init< ... >( ... ) );
>
> That makes a lot of sense, but that requires modification of Py++,
> doesn't it? Or is there a way to customize code generation to do it the
> way you've cited?
>
> Seems to me if Py++ would always opt for code like that, it would
> already solve quite a few dependency problems.
>
> I would be more then happy to implement this. Could you point me in
> right direction, please?

Yes of course:

code_creators\class_declaration.py - class_t._generate_constructor
It should not be too difficult to change the code.

Next document will help you to understand design\architecture of Py++
http://language-binding.net/pyplusplus/documentation/architecture.html


-- 
Roman Yakovenko
C++ Python language binding
http://www.language-binding.net/


From amohr at pixar.com  Wed Sep  6 22:35:34 2006
From: amohr at pixar.com (Alex Mohr)
Date: Wed, 06 Sep 2006 13:35:34 -0700
Subject: [C++-sig] Is there a way to automatically convert a smart_ptr
 to a held_type (try 2)
In-Reply-To: <bea9869d0609061217r60fd8908w78154e5c4ea39bf8@mail.gmail.com>
References: <bea9869d0608300707r3009cdb2u101ab96b88ec9966@mail.gmail.com>
	<44F5CD10.5010103@pixar.com>
	<bea9869d0609051609n7f6760f3jd43b7cea8c3d66dd@mail.gmail.com>
	<44FE0AB2.8010006@pixar.com>
	<bea9869d0609061217r60fd8908w78154e5c4ea39bf8@mail.gmail.com>
Message-ID: <44FF3116.8070504@pixar.com>

> I am sorting through this code right now.  Am I missing something or
> is there no documentation in the reference guide about how to register
> a converter directly with the registry?

No, but the code I posted does it in the constructor.

> Additionally, I was looking at the documentation for
> implicitly_convertible and I am wondering why I can't just use that in
> my case.
> 
> I want to call:
> 
> setObject(ptr<Base> val)
> 
> with a NoneType.  And I have all my objects (ie. Base) wrapped and set
> with held types of ref_ptr<>'s.  RefPtr's are implicitly convertible
> to ptr<>'s, so If I register this with boost.python using:
> 
>   bp::implicitly_convertible< ret_ptr<Base>, ptr<Base> >();
> 
> Then shouldn't boost.python be able to convert NoneType -->
> ref_ptr<Base> --> ptr<Base>  ?  From the documentation in
> implicitly_convertible it seems like this should be the case.  But I
> have to admit I have not found the place in the boost.python code
> where it handles the case of taking a NoneType and automatically
> converting it into a NULL smart pointer to start with.

Sure -- if your types meet those requirements, then that should work. 
The rvalue converters I posted are more general.  implicitly_convertible 
just registers an rvalue converter using the same underlying mechanism.

 From your message, I was under the impression that default-constructing 
one of your smart pointers would not produce the desired "null" value.

Alex


From abierbaum at gmail.com  Wed Sep  6 22:50:56 2006
From: abierbaum at gmail.com (Allen Bierbaum)
Date: Wed, 6 Sep 2006 15:50:56 -0500
Subject: [C++-sig] Is there a way to automatically convert a smart_ptr
	to a held_type (try 2)
In-Reply-To: <44FF3116.8070504@pixar.com>
References: <bea9869d0608300707r3009cdb2u101ab96b88ec9966@mail.gmail.com>
	<44F5CD10.5010103@pixar.com>
	<bea9869d0609051609n7f6760f3jd43b7cea8c3d66dd@mail.gmail.com>
	<44FE0AB2.8010006@pixar.com>
	<bea9869d0609061217r60fd8908w78154e5c4ea39bf8@mail.gmail.com>
	<44FF3116.8070504@pixar.com>
Message-ID: <bea9869d0609061350t763da4cdw11941f97910862d7@mail.gmail.com>

On 9/6/06, Alex Mohr <amohr at pixar.com> wrote:
> > I am sorting through this code right now.  Am I missing something or
> > is there no documentation in the reference guide about how to register
> > a converter directly with the registry?
>
> No, but the code I posted does it in the constructor.
>
> > Additionally, I was looking at the documentation for
> > implicitly_convertible and I am wondering why I can't just use that in
> > my case.
> >
> > I want to call:
> >
> > setObject(ptr<Base> val)
> >
> > with a NoneType.  And I have all my objects (ie. Base) wrapped and set
> > with held types of ref_ptr<>'s.  RefPtr's are implicitly convertible
> > to ptr<>'s, so If I register this with boost.python using:
> >
> >   bp::implicitly_convertible< ret_ptr<Base>, ptr<Base> >();
> >
> > Then shouldn't boost.python be able to convert NoneType -->
> > ref_ptr<Base> --> ptr<Base>  ?  From the documentation in
> > implicitly_convertible it seems like this should be the case.  But I
> > have to admit I have not found the place in the boost.python code
> > where it handles the case of taking a NoneType and automatically
> > converting it into a NULL smart pointer to start with.
>
> Sure -- if your types meet those requirements, then that should work.
> The rvalue converters I posted are more general.  implicitly_convertible
> just registers an rvalue converter using the same underlying mechanism.
>
>  From your message, I was under the impression that default-constructing
> one of your smart pointers would not produce the desired "null" value.

A default construction of the RefPtr Type should, but using
implicitly_convertible didn't work.  It acted like it didn't know that
there was an option to convert through the implicit chains.

I have now implemented something based on the code you posted and am
working on testing it.  Hopefully it will work out.  I will let you
kno.

-Allen

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


From abierbaum at gmail.com  Wed Sep  6 23:06:42 2006
From: abierbaum at gmail.com (Allen Bierbaum)
Date: Wed, 6 Sep 2006 16:06:42 -0500
Subject: [C++-sig] Is there a way to automatically convert a smart_ptr
	to a held_type (try 2)
In-Reply-To: <44FE0AB2.8010006@pixar.com>
References: <bea9869d0608300707r3009cdb2u101ab96b88ec9966@mail.gmail.com>
	<44F5CD10.5010103@pixar.com>
	<bea9869d0609051609n7f6760f3jd43b7cea8c3d66dd@mail.gmail.com>
	<44FE0AB2.8010006@pixar.com>
Message-ID: <bea9869d0609061406lddb16fx5165bcc01e8814f3@mail.gmail.com>

On 9/5/06, Alex Mohr <amohr at pixar.com> wrote:
> > Thanks Alex, this worked great.  I had to change the Py_INCREF to a
> > bp::incref but otherwise things are all good.
> >
> > But now I have a rather strange followup question.
> >
> > It seems that Boost.Python is doing some magic behind the scenes to
> > convert an object coming from C++ as a smart pointer whose wrapped
> > pointer is NULL into a Python "None".  The is great, and was a
> > wonderful surprise.  But now I need to find a way to make it happen in
> > the other direction.
> >
> > How does boost.python convert a parameter of "None" into an
> > object/smart_ptr to pass to C++?
>
> It happens in the from_python conversion.  It just checks for None and
> (I believe) default-constructs a the pointer object.  Is it the case
> that default-constructing one of your smart pointers does not produce
> the intended NULL?  It probably wouldn't be hard to add a customization
> (paralleling get_pointer and pointee) to boost python.  But in the
> meantime...
>
> > I need to find a way to call a C++ method like:
> >
> > void setObject(ptr<Base> obj);
> >
> > from Python and call it in Python as "setObject(None)" and instruct
> > boost.python to convert this into a call behind the scenes in C++ to:
> >
> > setObject(NullPtr);
> >
> > where "NullPtr" is a global const value defined by the library that is
> > used to signify a NULL shared pointer throughout the system.  (the
> > reason it needs something like this is a little complex, but it is not
> > my library so I can't change this part much.)
> >
> > Alternatively, I could probably make it work if I could get
> > boost.python to make a call like:
> >
> > setObject(ref_ptr<Base>(NULL) )
> >
> > as I think I could get the library to automatically convert
> > ref_ptr<Base>(NULL) into a NullPtr.
> >
> > Any ideas?
>
> I think you can do either version by registering your own rvalue
> from_python converters for all your pointer types.  You can do this in
> the def_visitor along with the to_python conversion we discussed earlier.
>
> Just replace the relevant bits below ("Deal with the "None" case") with
> whatever you want.
>
> Again, I'm just typing this so it may not compile, and no guarantees of
> course...
>
> template <class Ptr>
> struct smart_ptr_from_python {
>      typedef typename boost::python::pointee<Ptr>::type Pointee;
>      smart_ptr_from_python() {
>          converter::registry::insert(&convertible, &construct,
>                                      type_id<Ptr>());
>      }
>    private:
>      static void *convertible(PyObject *p) {
>          // can always produce a pointer from None.
>          if (p == Py_None)
>              return p;
>          // Otherwise, we can do it if we can get the pointee out.
>          void *result = converter::get_lvalue_from_python
>              (p, converter::registered<Pointee>::converters);
>          return result;
>      }
>
>      static void construct(PyObject* source, converter::
>                            rvalue_from_python_stage1_data* data) {
>          void* const storage =
>              ((converter::rvalue_from_python_storage<Ptr>*)data)->
>              storage.bytes;
>          // Deal with the "None" case.
>          if (data->convertible == source)
>              new (storage) Ptr();  // Or whatever you want.
>          else
>              new (storage)
> Ptr(static_cast<Pointee*>(data->convertible));
>          data->convertible = storage;
>      }
> };

This code allowed the setObject(None) to work now.  Unfortunately it
broke the functionality that had been working where I could call
setObject(PythonObject).

More details (extracted from above):

I have:

// set method
void setObject(ptr<Base> obj);

// Implicit conversion for ref_ptr to ptr
bp::implicitly_convertible< ref_ptr<Base>, ptr<Base> );

// From original e-mail: this works to automatically convert
// ptr<Base> returns into ref_ptr<Base> as the held type
bp::to_pyton_converter< ptr<Base>, ptr_to_python< ptr<Base> >()

// Most recent discussion: way to get None from python as ptr<Base>
// - This works to let me pass None into wrapped setObject method
bp::converter::registry::insert(&convertible, &construct,
bp::type_id<ptr<Base> >());

The problem is that now when I try to call something like this in python:

new_obj = objtype.create()
setObject(None)
setObject(new_obj)

Both of these calls to setObject try to convert the object to a
ptr<Base> using the manually registered converter.  The converter only
really knows how to deal with the conversion from a None type.  In the
case of a real object I just want boost.python to convert the held
type (ref_ptr<Base> ) into a ptr<Base> using the implicit conversion.

How do I get this to work?  Can I have multiple converters registered
and if so how do I tell boost.python to only use the manually created
one for the NoneType?

-Allen


From amohr at pixar.com  Wed Sep  6 23:19:43 2006
From: amohr at pixar.com (Alex Mohr)
Date: Wed, 06 Sep 2006 14:19:43 -0700
Subject: [C++-sig] Is there a way to automatically convert a smart_ptr
 to a held_type (try 2)
In-Reply-To: <bea9869d0609061406lddb16fx5165bcc01e8814f3@mail.gmail.com>
References: <bea9869d0608300707r3009cdb2u101ab96b88ec9966@mail.gmail.com>
	<44F5CD10.5010103@pixar.com>
	<bea9869d0609051609n7f6760f3jd43b7cea8c3d66dd@mail.gmail.com>
	<44FE0AB2.8010006@pixar.com>
	<bea9869d0609061406lddb16fx5165bcc01e8814f3@mail.gmail.com>
Message-ID: <44FF3B6F.9030509@pixar.com>

> This code allowed the setObject(None) to work now.  Unfortunately it
> broke the functionality that had been working where I could call
> setObject(PythonObject).
> 
> More details (extracted from above):
> 
> I have:
> 
> // set method
> void setObject(ptr<Base> obj);
> 
> // Implicit conversion for ref_ptr to ptr
> bp::implicitly_convertible< ref_ptr<Base>, ptr<Base> );
> 
> // From original e-mail: this works to automatically convert
> // ptr<Base> returns into ref_ptr<Base> as the held type
> bp::to_pyton_converter< ptr<Base>, ptr_to_python< ptr<Base> >()
> 
> // Most recent discussion: way to get None from python as ptr<Base>
> // - This works to let me pass None into wrapped setObject method
> bp::converter::registry::insert(&convertible, &construct,
> bp::type_id<ptr<Base> >());
> 
> The problem is that now when I try to call something like this in python:
> 
> new_obj = objtype.create()
> setObject(None)
> setObject(new_obj)
> 
> Both of these calls to setObject try to convert the object to a
> ptr<Base> using the manually registered converter.  The converter only
> really knows how to deal with the conversion from a None type.  In the
> case of a real object I just want boost.python to convert the held
> type (ref_ptr<Base> ) into a ptr<Base> using the implicit conversion.
> 
> How do I get this to work?  Can I have multiple converters registered
> and if so how do I tell boost.python to only use the manually created
> one for the NoneType?


The rvalue converter should handle both cases.  In convertible, it 
checks if it is getting None -- if it is, it says yes, I can make a 
pointer.  If it gets something other than None, it says I can make a 
pointer iff I can get an lvalue pointer to the Pointee from the python 
object.  That's the code after the if.  It uses already registered 
lvalue converters to do this.

The construct method has a similar structure -- in the None case (first 
part of if) it will just placement new a default-constructed pointer. 
In the else part of the if, it will placement new a pointer pointing to 
the c++ instance.

Can you tell me what's going on in convertible and construct when you do 
this?

 > setObject(new_obj)

Alex


 >> template <class Ptr>
 >> struct smart_ptr_from_python {
 >>      typedef typename boost::python::pointee<Ptr>::type Pointee;
 >>      smart_ptr_from_python() {
 >>          converter::registry::insert(&convertible, &construct,
 >>                                      type_id<Ptr>());
 >>      }
 >>    private:
 >>      static void *convertible(PyObject *p) {
 >>          // can always produce a pointer from None.
 >>          if (p == Py_None)
 >>              return p;
 >>          // Otherwise, we can do it if we can get the pointee out.
 >>          void *result = converter::get_lvalue_from_python
 >>              (p, converter::registered<Pointee>::converters);
 >>          return result;
 >>      }
 >>
 >>      static void construct(PyObject* source, converter::
 >>                            rvalue_from_python_stage1_data* data) {
 >>          void* const storage =
 >>              ((converter::rvalue_from_python_storage<Ptr>*)data)->
 >>              storage.bytes;
 >>          // Deal with the "None" case.
 >>          if (data->convertible == source)
 >>              new (storage) Ptr();  // Or whatever you want.
 >>          else
 >>              new (storage)
 >> Ptr(static_cast<Pointee*>(data->convertible));
 >>          data->convertible = storage;
 >>      }
 >> };


From abierbaum at gmail.com  Wed Sep  6 23:32:12 2006
From: abierbaum at gmail.com (Allen Bierbaum)
Date: Wed, 6 Sep 2006 16:32:12 -0500
Subject: [C++-sig] Is there a way to automatically convert a smart_ptr
	to a held_type (try 2)
In-Reply-To: <44FF3B6F.9030509@pixar.com>
References: <bea9869d0608300707r3009cdb2u101ab96b88ec9966@mail.gmail.com>
	<44F5CD10.5010103@pixar.com>
	<bea9869d0609051609n7f6760f3jd43b7cea8c3d66dd@mail.gmail.com>
	<44FE0AB2.8010006@pixar.com>
	<bea9869d0609061406lddb16fx5165bcc01e8814f3@mail.gmail.com>
	<44FF3B6F.9030509@pixar.com>
Message-ID: <bea9869d0609061432k1c408ba5i8337efa421611360@mail.gmail.com>

On 9/6/06, Alex Mohr <amohr at pixar.com> wrote:
> > This code allowed the setObject(None) to work now.  Unfortunately it
> > broke the functionality that had been working where I could call
> > setObject(PythonObject).
> >
> > More details (extracted from above):
> >
> > I have:
> >
> > // set method
> > void setObject(ptr<Base> obj);
> >
> > // Implicit conversion for ref_ptr to ptr
> > bp::implicitly_convertible< ref_ptr<Base>, ptr<Base> );
> >
> > // From original e-mail: this works to automatically convert
> > // ptr<Base> returns into ref_ptr<Base> as the held type
> > bp::to_pyton_converter< ptr<Base>, ptr_to_python< ptr<Base> >()
> >
> > // Most recent discussion: way to get None from python as ptr<Base>
> > // - This works to let me pass None into wrapped setObject method
> > bp::converter::registry::insert(&convertible, &construct,
> > bp::type_id<ptr<Base> >());
> >
> > The problem is that now when I try to call something like this in python:
> >
> > new_obj = objtype.create()
> > setObject(None)
> > setObject(new_obj)
> >
> > Both of these calls to setObject try to convert the object to a
> > ptr<Base> using the manually registered converter.  The converter only
> > really knows how to deal with the conversion from a None type.  In the
> > case of a real object I just want boost.python to convert the held
> > type (ref_ptr<Base> ) into a ptr<Base> using the implicit conversion.
> >
> > How do I get this to work?  Can I have multiple converters registered
> > and if so how do I tell boost.python to only use the manually created
> > one for the NoneType?
>
>
> The rvalue converter should handle both cases.  In convertible, it
> checks if it is getting None -- if it is, it says yes, I can make a
> pointer.  If it gets something other than None, it says I can make a
> pointer iff I can get an lvalue pointer to the Pointee from the python
> object.  That's the code after the if.  It uses already registered
> lvalue converters to do this.
>
> The construct method has a similar structure -- in the None case (first
> part of if) it will just placement new a default-constructed pointer.
> In the else part of the if, it will placement new a pointer pointing to
> the c++ instance.
>
> Can you tell me what's going on in convertible and construct when you do
> this?

The problem for me was that the rvalue converter when passed a
non-None could not do the placement new created of the ptr<Base>
because the constructor is protected.  The library I am wrapping
requires all objects to be created using static factory methods:

ref_ptr<Base> obj = Class.create()

So there is something going on in the implicit conversion for the
smart pointers that just handles things better.

I ended up getting the code to work though by making it so the rvalue
convertible method return NULL if p != NonType.  This ends up meaning
that the rvalue converter is only used in the case of passing None,
but it works for me now.

Thanks for your help.  I will let you know if anything else crops up
but I think I am starting to understand this part of boost.python much
better now so hopefully I will be able to fix it myself. :)

Thanks,
Allen

>
>  > setObject(new_obj)
>
> Alex
>
>
>  >> template <class Ptr>
>  >> struct smart_ptr_from_python {
>  >>      typedef typename boost::python::pointee<Ptr>::type Pointee;
>  >>      smart_ptr_from_python() {
>  >>          converter::registry::insert(&convertible, &construct,
>  >>                                      type_id<Ptr>());
>  >>      }
>  >>    private:
>  >>      static void *convertible(PyObject *p) {
>  >>          // can always produce a pointer from None.
>  >>          if (p == Py_None)
>  >>              return p;
>  >>          // Otherwise, we can do it if we can get the pointee out.
>  >>          void *result = converter::get_lvalue_from_python
>  >>              (p, converter::registered<Pointee>::converters);
>  >>          return result;
>  >>      }
>  >>
>  >>      static void construct(PyObject* source, converter::
>  >>                            rvalue_from_python_stage1_data* data) {
>  >>          void* const storage =
>  >>              ((converter::rvalue_from_python_storage<Ptr>*)data)->
>  >>              storage.bytes;
>  >>          // Deal with the "None" case.
>  >>          if (data->convertible == source)
>  >>              new (storage) Ptr();  // Or whatever you want.
>  >>          else
>  >>              new (storage)
>  >> Ptr(static_cast<Pointee*>(data->convertible));
>  >>          data->convertible = storage;
>  >>      }
>  >> };
> _______________________________________________
> C++-sig mailing list
> C++-sig at python.org
> http://mail.python.org/mailman/listinfo/c++-sig
>


From amohr at pixar.com  Wed Sep  6 23:45:10 2006
From: amohr at pixar.com (Alex Mohr)
Date: Wed, 06 Sep 2006 14:45:10 -0700
Subject: [C++-sig] Is there a way to automatically convert a smart_ptr
 to a held_type (try 2)
In-Reply-To: <bea9869d0609061432k1c408ba5i8337efa421611360@mail.gmail.com>
References: <bea9869d0608300707r3009cdb2u101ab96b88ec9966@mail.gmail.com>
	<44F5CD10.5010103@pixar.com>
	<bea9869d0609051609n7f6760f3jd43b7cea8c3d66dd@mail.gmail.com>
	<44FE0AB2.8010006@pixar.com>
	<bea9869d0609061406lddb16fx5165bcc01e8814f3@mail.gmail.com>
	<44FF3B6F.9030509@pixar.com>
	<bea9869d0609061432k1c408ba5i8337efa421611360@mail.gmail.com>
Message-ID: <44FF4166.7040404@pixar.com>

> The problem for me was that the rvalue converter when passed a
> non-None could not do the placement new created of the ptr<Base>
> because the constructor is protected.  The library I am wrapping
> requires all objects to be created using static factory methods:
> 
> ref_ptr<Base> obj = Class.create()

But the converter's not trying to create a new *object* -- just a new 
smart pointer.  If you can write code that obtains a correct new smart 
pointer from a raw object pointer, then just fill in that code in the 
right place in the converter and it should work.  The only other thing 
that's required is that the smart pointer objects be copy-constructible, 
which I would think they ought to be.  Otherwise they're very strange.

> So there is something going on in the implicit conversion for the
> smart pointers that just handles things better.
> 
> I ended up getting the code to work though by making it so the rvalue
> convertible method return NULL if p != NonType.  This ends up meaning
> that the rvalue converter is only used in the case of passing None,
> but it works for me now.

Sure -- that sounds like it will be okay given your situation.

> Thanks for your help.  I will let you know if anything else crops up
> but I think I am starting to understand this part of boost.python much
> better now so hopefully I will be able to fix it myself. :)

No problem.

Alex


From abierbaum at gmail.com  Thu Sep  7 00:09:09 2006
From: abierbaum at gmail.com (Allen Bierbaum)
Date: Wed, 6 Sep 2006 17:09:09 -0500
Subject: [C++-sig] Is there a way to automatically convert a smart_ptr
	to a held_type (try 2)
In-Reply-To: <44FF4166.7040404@pixar.com>
References: <bea9869d0608300707r3009cdb2u101ab96b88ec9966@mail.gmail.com>
	<44F5CD10.5010103@pixar.com>
	<bea9869d0609051609n7f6760f3jd43b7cea8c3d66dd@mail.gmail.com>
	<44FE0AB2.8010006@pixar.com>
	<bea9869d0609061406lddb16fx5165bcc01e8814f3@mail.gmail.com>
	<44FF3B6F.9030509@pixar.com>
	<bea9869d0609061432k1c408ba5i8337efa421611360@mail.gmail.com>
	<44FF4166.7040404@pixar.com>
Message-ID: <bea9869d0609061509h7ef18c1doeb75f60b01f45d17@mail.gmail.com>

On 9/6/06, Alex Mohr <amohr at pixar.com> wrote:
> > The problem for me was that the rvalue converter when passed a
> > non-None could not do the placement new created of the ptr<Base>
> > because the constructor is protected.  The library I am wrapping
> > requires all objects to be created using static factory methods:
> >
> > ref_ptr<Base> obj = Class.create()
>
> But the converter's not trying to create a new *object* -- just a new
> smart pointer.  If you can write code that obtains a correct new smart
> pointer from a raw object pointer, then just fill in that code in the
> right place in the converter and it should work.

This is the part I can't do.  I can make a ref_ptr<> from a ptr<> as
vice versa, but I can't build either of them from a raw pointer.  I
may take a look at this later and see if I can make it work better so
I just have the one converter.

> The only other thing
> that's required is that the smart pointer objects be copy-constructible,
> which I would think they ought to be.  Otherwise they're very strange.
>
> > So there is something going on in the implicit conversion for the
> > smart pointers that just handles things better.
> >
> > I ended up getting the code to work though by making it so the rvalue
> > convertible method return NULL if p != NonType.  This ends up meaning
> > that the rvalue converter is only used in the case of passing None,
> > but it works for me now.
>
> Sure -- that sounds like it will be okay given your situation.

I agree.  NoneType --> ptr<> was always a special case for me since I
want everything held by ref_ptrs all the time anyway.

Thanks again.

-Allen

>
> > Thanks for your help.  I will let you know if anything else crops up
> > but I think I am starting to understand this part of boost.python much
> > better now so hopefully I will be able to fix it myself. :)
>
> No problem.
>
> Alex
> _______________________________________________
> C++-sig mailing list
> C++-sig at python.org
> http://mail.python.org/mailman/listinfo/c++-sig
>


From kir at lapshin.net  Thu Sep  7 12:04:19 2006
From: kir at lapshin.net (Kirill Lapshin)
Date: Thu, 07 Sep 2006 14:04:19 +0400
Subject: [C++-sig] [Py++] Nested enum as default arg
In-Reply-To: <7465b6170609061231s113a1440h8d435e861621a540@mail.gmail.com>
References: <edk5pf$kha$1@sea.gmane.org>	<7465b6170609051100x745f016fxde34980bc7021767@mail.gmail.com>	<edn712$jc4$1@sea.gmane.org>
	<7465b6170609061231s113a1440h8d435e861621a540@mail.gmail.com>
Message-ID: <edoqr3$au6$1@sea.gmane.org>

Roman Yakovenko wrote:
>> I would be more then happy to implement this. Could you point me in
>> right direction, please?
> 
> Yes of course:
> 
> code_creators\class_declaration.py - class_t._generate_constructor
> It should not be too difficult to change the code.
> 
> Next document will help you to understand design\architecture of Py++
> http://language-binding.net/pyplusplus/documentation/architecture.html

Thanks. Will take a look. Unfortunately I have to postpone it for a bit. 
It turns out the problem I have is a bit different, and at the moment 
the highest priority task is get the bindings working, I will try to 
make them nicer next.

The problem I have at the moment is along the lines of:

struct Base {};

struct Derived: public Base
{
     Derived(const vector<shared_ptr<Base> >& arg = 
vector<shared_ptr<Base> >() );
};

What happens is in mymodule.main.cpp vector<shared_ptr<Base> > gets 
registered after both Base and Derived, and obviously registration of 
Derived fails.

Is there any way to change the order in which various classes get 
registered?

Kirill



From ndbecker2 at gmail.com  Thu Sep  7 16:17:18 2006
From: ndbecker2 at gmail.com (Neal Becker)
Date: Thu, 07 Sep 2006 10:17:18 -0400
Subject: [C++-sig] boost::python tutorial on overloading not up-to-date?
Message-ID: <edp9mf$1rq$1@sea.gmane.org>

In boost::python tutorial there is a whole section on manual overloading. 
AFAICT, overloading "just works" without my doing anything.  Is this
information out-of-date?



From abierbaum at gmail.com  Thu Sep  7 17:09:52 2006
From: abierbaum at gmail.com (Allen Bierbaum)
Date: Thu, 7 Sep 2006 10:09:52 -0500
Subject: [C++-sig] Is there a way to automatically convert a smart_ptr
	to a held_type (try 2)
In-Reply-To: <bea9869d0609061432k1c408ba5i8337efa421611360@mail.gmail.com>
References: <bea9869d0608300707r3009cdb2u101ab96b88ec9966@mail.gmail.com>
	<44F5CD10.5010103@pixar.com>
	<bea9869d0609051609n7f6760f3jd43b7cea8c3d66dd@mail.gmail.com>
	<44FE0AB2.8010006@pixar.com>
	<bea9869d0609061406lddb16fx5165bcc01e8814f3@mail.gmail.com>
	<44FF3B6F.9030509@pixar.com>
	<bea9869d0609061432k1c408ba5i8337efa421611360@mail.gmail.com>
Message-ID: <bea9869d0609070809q13feb423yc9949b3206823625@mail.gmail.com>

>
> Thanks for your help.  I will let you know if anything else crops up
> but I think I am starting to understand this part of boost.python much
> better now so hopefully I will be able to fix it myself. :)

It looks like everything related to the auto-converting to ref_ptr
held type and handling of converting to and from None.

I have one remaining problem that I know of with this code.  How do I
make it possible to compare the wrapped ref_ptrs for equality?  I am
assuming that I need to add an __eq__ operator for the ref_ptr types
but since they are used as held_type I am not really exposing them in
the "normal" way.

This is how it is behaving right now (note: all the objects here have
ref_ptr held type)

>> n = osg.Node.create()
>> g = osg.Group.create()
>> n.setCore(g)
>> g
<osg._osg.Group object at 0xb785fcac>
>> n.getCore()
<osg._osg.Group object at 0xb785f16c>
>> g == n.getCore()
False


Does anyone know if this is also an issue when wrapping an API and
using boost::shared_ptr<> as the held type?  If so, how is that case
handled?

Thanks,
Allen


From amohr at pixar.com  Thu Sep  7 18:39:18 2006
From: amohr at pixar.com (Alex Mohr)
Date: Thu, 07 Sep 2006 09:39:18 -0700
Subject: [C++-sig] Is there a way to automatically convert a smart_ptr
 to a held_type (try 2)
In-Reply-To: <bea9869d0609070809q13feb423yc9949b3206823625@mail.gmail.com>
References: <bea9869d0608300707r3009cdb2u101ab96b88ec9966@mail.gmail.com>
	<44F5CD10.5010103@pixar.com>
	<bea9869d0609051609n7f6760f3jd43b7cea8c3d66dd@mail.gmail.com>
	<44FE0AB2.8010006@pixar.com>
	<bea9869d0609061406lddb16fx5165bcc01e8814f3@mail.gmail.com>
	<44FF3B6F.9030509@pixar.com>
	<bea9869d0609061432k1c408ba5i8337efa421611360@mail.gmail.com>
	<bea9869d0609070809q13feb423yc9949b3206823625@mail.gmail.com>
Message-ID: <45004B36.1020800@pixar.com>

> I have one remaining problem that I know of with this code.  How do I
> make it possible to compare the wrapped ref_ptrs for equality?  I am
> assuming that I need to add an __eq__ operator for the ref_ptr types
> but since they are used as held_type I am not really exposing them in
> the "normal" way.

Yup -- you'll have to implement __eq__.

> <osg._osg.Group object at 0xb785fcac>
>>> n.getCore()
> <osg._osg.Group object at 0xb785f16c>
>>> g == n.getCore()
> False
> 
> Does anyone know if this is also an issue when wrapping an API and
> using boost::shared_ptr<> as the held type?  If so, how is that case
> handled?

It works sometimes, and not others.  As you discovered, boost python 
will store the python object in a custom shared_ptr deleter.  Converting 
to-python, I believe it will see if the shared_ptr has its custom 
deleter and will use the python object in there if it is.

The problem is that if you're using your own shared_ptrs, then you may 
pass a shared_ptr to boost python which doesn't have the special custom 
deleter, and you'll end up with multiple python objects representing the 
same c++ object.

What you really want is to have just a single python object associated 
with each C++ instance.

I call this problem "object identity", and it's been discussed on this 
list before.

It is *possible* to solve this problem with boost python as is, but it 
requires doing *very* nasty things using internal api.

I think the real solution is to extend boost python so that:

- You can tell boost python that your c++ object can manage its python 
identity.

- As soon as boost python produces a python object for your c++ object, 
it calls you to establish the identity.

- Whenever (a pointer to) that c++ instance is to be converted 
to-python, boost python will consult you to find the python object.

Incidentally, wrapper<T> already does this stuff mostly right.  I hope 
to make a patch to support this someday, but it probably won't be that soon.

Anyway, in the meantime, if you just implement __eq__ in the obvious 
way, it should work.

Alex


From abierbaum at gmail.com  Thu Sep  7 18:46:37 2006
From: abierbaum at gmail.com (Allen Bierbaum)
Date: Thu, 7 Sep 2006 11:46:37 -0500
Subject: [C++-sig] Is there a way to automatically convert a smart_ptr
	to a held_type (try 2)
In-Reply-To: <45004B36.1020800@pixar.com>
References: <bea9869d0608300707r3009cdb2u101ab96b88ec9966@mail.gmail.com>
	<44F5CD10.5010103@pixar.com>
	<bea9869d0609051609n7f6760f3jd43b7cea8c3d66dd@mail.gmail.com>
	<44FE0AB2.8010006@pixar.com>
	<bea9869d0609061406lddb16fx5165bcc01e8814f3@mail.gmail.com>
	<44FF3B6F.9030509@pixar.com>
	<bea9869d0609061432k1c408ba5i8337efa421611360@mail.gmail.com>
	<bea9869d0609070809q13feb423yc9949b3206823625@mail.gmail.com>
	<45004B36.1020800@pixar.com>
Message-ID: <bea9869d0609070946x3589060bgcc1a539868451b4c@mail.gmail.com>

On 9/7/06, Alex Mohr <amohr at pixar.com> wrote:
> > I have one remaining problem that I know of with this code.  How do I
> > make it possible to compare the wrapped ref_ptrs for equality?  I am
> > assuming that I need to add an __eq__ operator for the ref_ptr types
> > but since they are used as held_type I am not really exposing them in
> > the "normal" way.
>
> Yup -- you'll have to implement __eq__.
>
> > <osg._osg.Group object at 0xb785fcac>
> >>> n.getCore()
> > <osg._osg.Group object at 0xb785f16c>
> >>> g == n.getCore()
> > False
> >
> > Does anyone know if this is also an issue when wrapping an API and
> > using boost::shared_ptr<> as the held type?  If so, how is that case
> > handled?
>
> It works sometimes, and not others.  As you discovered, boost python
> will store the python object in a custom shared_ptr deleter.  Converting
> to-python, I believe it will see if the shared_ptr has its custom
> deleter and will use the python object in there if it is.
>
> The problem is that if you're using your own shared_ptrs, then you may
> pass a shared_ptr to boost python which doesn't have the special custom
> deleter, and you'll end up with multiple python objects representing the
> same c++ object.
>
> What you really want is to have just a single python object associated
> with each C++ instance.
>
> I call this problem "object identity", and it's been discussed on this
> list before.
>
> It is *possible* to solve this problem with boost python as is, but it
> requires doing *very* nasty things using internal api.
>
> I think the real solution is to extend boost python so that:
>
> - You can tell boost python that your c++ object can manage its python
> identity.
>
> - As soon as boost python produces a python object for your c++ object,
> it calls you to establish the identity.
>
> - Whenever (a pointer to) that c++ instance is to be converted
> to-python, boost python will consult you to find the python object.
>
> Incidentally, wrapper<T> already does this stuff mostly right.  I hope
> to make a patch to support this someday, but it probably won't be that soon.
>
> Anyway, in the meantime, if you just implement __eq__ in the obvious
> way, it should work.

The non-obvious part of this for me is:

- Does this mean that I now need to explicitly wrap (expose using
bp::class_ ) all the ref_ptr<>  types?

I haven't had to do that thus far because it was handled internally
somehow because I used it has the held_type for all my classes.  If I
expose them manually do I run the risk of breaking any of the
held_type functionality?

-Allen


From amohr at pixar.com  Thu Sep  7 18:55:05 2006
From: amohr at pixar.com (Alex Mohr)
Date: Thu, 07 Sep 2006 09:55:05 -0700
Subject: [C++-sig] Is there a way to automatically convert a smart_ptr
 to a held_type (try 2)
In-Reply-To: <bea9869d0609070946x3589060bgcc1a539868451b4c@mail.gmail.com>
References: <bea9869d0608300707r3009cdb2u101ab96b88ec9966@mail.gmail.com>
	<44F5CD10.5010103@pixar.com>
	<bea9869d0609051609n7f6760f3jd43b7cea8c3d66dd@mail.gmail.com>
	<44FE0AB2.8010006@pixar.com>
	<bea9869d0609061406lddb16fx5165bcc01e8814f3@mail.gmail.com>
	<44FF3B6F.9030509@pixar.com>
	<bea9869d0609061432k1c408ba5i8337efa421611360@mail.gmail.com>
	<bea9869d0609070809q13feb423yc9949b3206823625@mail.gmail.com>
	<45004B36.1020800@pixar.com>
	<bea9869d0609070946x3589060bgcc1a539868451b4c@mail.gmail.com>
Message-ID: <45004EE9.7040604@pixar.com>

> The non-obvious part of this for me is:
> 
> - Does this mean that I now need to explicitly wrap (expose using
> bp::class_ ) all the ref_ptr<>  types?
> 
> I haven't had to do that thus far because it was handled internally
> somehow because I used it has the held_type for all my classes.  If I
> expose them manually do I run the risk of breaking any of the
> held_type functionality?

I'm not sure what you mean...  If ref_ptr<T> is the held type for T, 
then just provide __eq__ in terms of ref_ptr<T>...

template <class T>
static bool ref_ptr_eq(ref_ptr<T> const &self, ref_ptr<T> const &other)
{
     return self == other;
}

class_<T, ... ref_ptr<T> ... >(...)
     .def("__eq__", ref_ptr_eq<T>)
     ;

Am I misunderstanding your situation?

Alex


From roman.yakovenko at gmail.com  Thu Sep  7 19:54:29 2006
From: roman.yakovenko at gmail.com (Roman Yakovenko)
Date: Thu, 7 Sep 2006 20:54:29 +0300
Subject: [C++-sig] [Py++] Nested enum as default arg
In-Reply-To: <edoqr3$au6$1@sea.gmane.org>
References: <edk5pf$kha$1@sea.gmane.org>
	<7465b6170609051100x745f016fxde34980bc7021767@mail.gmail.com>
	<edn712$jc4$1@sea.gmane.org>
	<7465b6170609061231s113a1440h8d435e861621a540@mail.gmail.com>
	<edoqr3$au6$1@sea.gmane.org>
Message-ID: <7465b6170609071054k49eb0cb0x53e243ee330fffec@mail.gmail.com>

On 9/7/06, Kirill Lapshin <kir at lapshin.net> wrote:
> Roman Yakovenko wrote:
> >> I would be more then happy to implement this. Could you point me in
> >> right direction, please?
> >
> > Yes of course:
> >
> > code_creators\class_declaration.py - class_t._generate_constructor
> > It should not be too difficult to change the code.
> >
> > Next document will help you to understand design\architecture of Py++
> > http://language-binding.net/pyplusplus/documentation/architecture.html
>
> Thanks. Will take a look. Unfortunately I have to postpone it for a bit.
> It turns out the problem I have is a bit different, and at the moment
> the highest priority task is get the bindings working, I will try to
> make them nicer next.
>
> The problem I have at the moment is along the lines of:
>
> struct Base {};
>
> struct Derived: public Base
> {
>      Derived(const vector<shared_ptr<Base> >& arg =
> vector<shared_ptr<Base> >() );
> };
>
> What happens is in mymodule.main.cpp vector<shared_ptr<Base> > gets
> registered after both Base and Derived, and obviously registration of
> Derived fails.
>
> Is there any way to change the order in which various classes get
> registered?

Yes of course. Take a look on this example( unit test ):
http://svn.sourceforge.net/viewvc/pygccxml/pyplusplus_dev/unittests/class_order_tester.py?view=markup

Then you can take a look on code_creators: compound_t, module_t,
module_body_t. This should help you.

-- 
Roman Yakovenko
C++ Python language binding
http://www.language-binding.net/


From abierbaum at gmail.com  Thu Sep  7 20:32:45 2006
From: abierbaum at gmail.com (Allen Bierbaum)
Date: Thu, 7 Sep 2006 13:32:45 -0500
Subject: [C++-sig] Is there a way to automatically convert a smart_ptr
	to a held_type (try 2)
In-Reply-To: <45004EE9.7040604@pixar.com>
References: <bea9869d0608300707r3009cdb2u101ab96b88ec9966@mail.gmail.com>
	<bea9869d0609051609n7f6760f3jd43b7cea8c3d66dd@mail.gmail.com>
	<44FE0AB2.8010006@pixar.com>
	<bea9869d0609061406lddb16fx5165bcc01e8814f3@mail.gmail.com>
	<44FF3B6F.9030509@pixar.com>
	<bea9869d0609061432k1c408ba5i8337efa421611360@mail.gmail.com>
	<bea9869d0609070809q13feb423yc9949b3206823625@mail.gmail.com>
	<45004B36.1020800@pixar.com>
	<bea9869d0609070946x3589060bgcc1a539868451b4c@mail.gmail.com>
	<45004EE9.7040604@pixar.com>
Message-ID: <bea9869d0609071132of696703qde7d7e1080868969@mail.gmail.com>

On 9/7/06, Alex Mohr <amohr at pixar.com> wrote:
> > The non-obvious part of this for me is:
> >
> > - Does this mean that I now need to explicitly wrap (expose using
> > bp::class_ ) all the ref_ptr<>  types?
> >
> > I haven't had to do that thus far because it was handled internally
> > somehow because I used it has the held_type for all my classes.  If I
> > expose them manually do I run the risk of breaking any of the
> > held_type functionality?
>
> I'm not sure what you mean...  If ref_ptr<T> is the held type for T,
> then just provide __eq__ in terms of ref_ptr<T>...
>
> template <class T>
> static bool ref_ptr_eq(ref_ptr<T> const &self, ref_ptr<T> const &other)
> {
>      return self == other;
> }
>
> class_<T, ... ref_ptr<T> ... >(...)
>      .def("__eq__", ref_ptr_eq<T>)
>      ;
>
> Am I misunderstanding your situation?

Ahhh....

No I see what you are meaning.  So I need the __eq__ op on the class
itself.  Ok.  That makes more sense.  I thought I may have to wrap the
ref_ptr<T> as a class a and put the __eq__ op on it.

Something like:

class_<ref_ptr<T>, ..>(..)
  .def("__eq__", ref_ptr_eq<T>)
;

Thanks for the feedback.  I will try your method out and reply to the
list with the results so other people following along can see what
ended up working.

Thanks again,
Allen

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


From kir at lapshin.net  Thu Sep  7 21:09:20 2006
From: kir at lapshin.net (Kirill Lapshin)
Date: Thu, 07 Sep 2006 23:09:20 +0400
Subject: [C++-sig] [Py++] Nested enum as default arg
In-Reply-To: <7465b6170609071054k49eb0cb0x53e243ee330fffec@mail.gmail.com>
References: <edk5pf$kha$1@sea.gmane.org>	<7465b6170609051100x745f016fxde34980bc7021767@mail.gmail.com>	<edn712$jc4$1@sea.gmane.org>	<7465b6170609061231s113a1440h8d435e861621a540@mail.gmail.com>	<edoqr3$au6$1@sea.gmane.org>
	<7465b6170609071054k49eb0cb0x53e243ee330fffec@mail.gmail.com>
Message-ID: <edpqp0$356$1@sea.gmane.org>

Roman Yakovenko wrote:
>> Is there any way to change the order in which various classes get
>> registered?
> 
> Yes of course. Take a look on this example( unit test ):
> http://svn.sourceforge.net/viewvc/pygccxml/pyplusplus_dev/unittests/class_order_tester.py?view=markup
> 
> Then you can take a look on code_creators: compound_t, module_t,
> module_body_t. This should help you.

Works like a charm! Thanks a bunch.

There is one tiny annoyance though -- match_declaration_t can't search 
declarations by alias, and searching by name is not that nice, since it 
requires spelling out full name such as std::vector< 
boost::shared_ptr<foo>, std::allocator< boost::shared_ptr<foo> >.
Not a big deal, I wrote my own matcher that does what I want.





From roman.yakovenko at gmail.com  Thu Sep  7 21:22:39 2006
From: roman.yakovenko at gmail.com (Roman Yakovenko)
Date: Thu, 7 Sep 2006 22:22:39 +0300
Subject: [C++-sig] [Py++] Nested enum as default arg
In-Reply-To: <edpqp0$356$1@sea.gmane.org>
References: <edk5pf$kha$1@sea.gmane.org>
	<7465b6170609051100x745f016fxde34980bc7021767@mail.gmail.com>
	<edn712$jc4$1@sea.gmane.org>
	<7465b6170609061231s113a1440h8d435e861621a540@mail.gmail.com>
	<edoqr3$au6$1@sea.gmane.org>
	<7465b6170609071054k49eb0cb0x53e243ee330fffec@mail.gmail.com>
	<edpqp0$356$1@sea.gmane.org>
Message-ID: <7465b6170609071222ob4cd8b3w53135eb4cdc0fed2@mail.gmail.com>

On 9/7/06, Kirill Lapshin <kir at lapshin.net> wrote:
> Roman Yakovenko wrote:
> >> Is there any way to change the order in which various classes get
> >> registered?
> >
> > Yes of course. Take a look on this example( unit test ):
> > http://svn.sourceforge.net/viewvc/pygccxml/pyplusplus_dev/unittests/class_order_tester.py?view=markup
> >
> > Then you can take a look on code_creators: compound_t, module_t,
> > module_body_t. This should help you.
>
> Works like a charm! Thanks a bunch.

Enjoy.

> There is one tiny annoyance though -- match_declaration_t can't search
> declarations by alias, and searching by name is not that nice, since it
> requires spelling out full name such as std::vector<
> boost::shared_ptr<foo>, std::allocator< boost::shared_ptr<foo> >.
> Not a big deal, I wrote my own matcher that does what I want.

Someday "code creators tree" will provide better( more user friendly )
interface. It will happen, I am sure.

-- 
Roman Yakovenko
C++ Python language binding
http://www.language-binding.net/


From kir at lapshin.net  Thu Sep  7 21:37:36 2006
From: kir at lapshin.net (Kirill Lapshin)
Date: Thu, 07 Sep 2006 23:37:36 +0400
Subject: [C++-sig] [boost.python] Indexing suite problems with convertable
	element type
Message-ID: <edpse0$8us$1@sea.gmane.org>

Hi,

I am just making my first step in boost.python world, so please forgive 
me for potential FAQ, but I can't find an answer in docs and by googling.

I have a custom date class, that I registered for conversion to Python's 
datetime.date object. Individual dates convert fine, but when I 
registered vector<date> with indexing suite, I have a problem that I can 
insert python date objects into vector, but can't extract them. E.g.

     >>> v = vector_day()
     >>> v.append( datetime.date(2006,1,1) )
     >>> len(v)
     1
     >>> v[0]
     Traceback (most recent call last):
       File "<stdin>", line 1, in ?
     TypeError: No Python class registered for C++ class class day

vector_day is exposed like this:

     bp::class_< std::vector<day, std::allocator<day> > >( "vector_day" 
)
         .def( bp::vector_indexing_suite<
             ::std::vector<day, std::allocator<day> > >() );

As I said, individual date objects convert just fine back and forth, 
i.e. function

     day test_day( day x ) { return x+1; }

exposed as

     typedef day ( *function_ptr_t )( day );

     bp::def( "test_day", function_ptr_t( &test_day ), ( bp::arg("x") ),
         bp::default_call_policies() );

can be used from Python without problems:

     >>> test_day( datetime.date(2006,1,1) )
     datetime.date(2006, 1, 2)

What am I doing wrong?

Regards,
Kirill Lapshin



From kir at lapshin.net  Thu Sep  7 21:52:48 2006
From: kir at lapshin.net (Kirill Lapshin)
Date: Thu, 07 Sep 2006 23:52:48 +0400
Subject: [C++-sig] [boost.python] Indexing suite problems with
 convertable element type
In-Reply-To: <edpse0$8us$1@sea.gmane.org>
References: <edpse0$8us$1@sea.gmane.org>
Message-ID: <edptag$bh2$1@sea.gmane.org>

> I have a custom date class, that I registered for conversion to Python's 
> datetime.date object. Individual dates convert fine, but when I 
> registered vector<date> with indexing suite, I have a problem that I can 
> insert python date objects into vector, but can't extract them.

Nevermind. A bit more searching around and I found this thread: 
http://thread.gmane.org/gmane.comp.python.c++/9846/focus=9846

So it can be solved by setting NoProxy to true.

Now I just have to figure out how to instruct Py++ to set NoProxy...



From roman.yakovenko at gmail.com  Thu Sep  7 21:59:18 2006
From: roman.yakovenko at gmail.com (Roman Yakovenko)
Date: Thu, 7 Sep 2006 22:59:18 +0300
Subject: [C++-sig] [boost.python] Indexing suite problems with
	convertable element type
In-Reply-To: <edptag$bh2$1@sea.gmane.org>
References: <edpse0$8us$1@sea.gmane.org> <edptag$bh2$1@sea.gmane.org>
Message-ID: <7465b6170609071259m90d780cn14adcdb992a02285@mail.gmail.com>

On 9/7/06, Kirill Lapshin <kir at lapshin.net> wrote:
> > I have a custom date class, that I registered for conversion to Python's
> > datetime.date object. Individual dates convert fine, but when I
> > registered vector<date> with indexing suite, I have a problem that I can
> > insert python date objects into vector, but can't extract them.
>
> Nevermind. A bit more searching around and I found this thread:
> http://thread.gmane.org/gmane.comp.python.c++/9846/focus=9846
>
> So it can be solved by setting NoProxy to true.
>
> Now I just have to figure out how to instruct Py++ to set NoProxy...

mb = module_builder_t( ... )
cls = mb.class_( ... )
cls.indexing_suite.no_proxy = True

Take a look on container_traits:
http://language-binding.net/pygccxml/apidocs/pygccxml.declarations.container_traits.vector_traits-class.html

-- 
Roman Yakovenko
C++ Python language binding
http://www.language-binding.net/


From abierbaum at gmail.com  Thu Sep  7 22:37:34 2006
From: abierbaum at gmail.com (Allen Bierbaum)
Date: Thu, 7 Sep 2006 15:37:34 -0500
Subject: [C++-sig] Is there a way to automatically convert a smart_ptr
	to a held_type (try 2)
In-Reply-To: <bea9869d0609071132of696703qde7d7e1080868969@mail.gmail.com>
References: <bea9869d0608300707r3009cdb2u101ab96b88ec9966@mail.gmail.com>
	<44FE0AB2.8010006@pixar.com>
	<bea9869d0609061406lddb16fx5165bcc01e8814f3@mail.gmail.com>
	<44FF3B6F.9030509@pixar.com>
	<bea9869d0609061432k1c408ba5i8337efa421611360@mail.gmail.com>
	<bea9869d0609070809q13feb423yc9949b3206823625@mail.gmail.com>
	<45004B36.1020800@pixar.com>
	<bea9869d0609070946x3589060bgcc1a539868451b4c@mail.gmail.com>
	<45004EE9.7040604@pixar.com>
	<bea9869d0609071132of696703qde7d7e1080868969@mail.gmail.com>
Message-ID: <bea9869d0609071337l26d001der8f815ecd30067301@mail.gmail.com>

On 9/7/06, Allen Bierbaum <abierbaum at gmail.com> wrote:
> On 9/7/06, Alex Mohr <amohr at pixar.com> wrote:
> > > The non-obvious part of this for me is:
> > >
> > > - Does this mean that I now need to explicitly wrap (expose using
> > > bp::class_ ) all the ref_ptr<>  types?
> > >
> > > I haven't had to do that thus far because it was handled internally
> > > somehow because I used it has the held_type for all my classes.  If I
> > > expose them manually do I run the risk of breaking any of the
> > > held_type functionality?
> >
> > I'm not sure what you mean...  If ref_ptr<T> is the held type for T,
> > then just provide __eq__ in terms of ref_ptr<T>...
> >
> > template <class T>
> > static bool ref_ptr_eq(ref_ptr<T> const &self, ref_ptr<T> const &other)
> > {
> >      return self == other;
> > }
> >
> > class_<T, ... ref_ptr<T> ... >(...)
> >      .def("__eq__", ref_ptr_eq<T>)
> >      ;
> >
> > Am I misunderstanding your situation?
>
> Ahhh....
>
> No I see what you are meaning.  So I need the __eq__ op on the class
> itself.  Ok.  That makes more sense.  I thought I may have to wrap the
> ref_ptr<T> as a class a and put the __eq__ op on it.
>
> Something like:
>
> class_<ref_ptr<T>, ..>(..)
>   .def("__eq__", ref_ptr_eq<T>)
> ;
>
> Thanks for the feedback.  I will try your method out and reply to the
> list with the results so other people following along can see what
> ended up working.

(I apologize for the length of this e-mail, I just want to make sure I
describe everything in case there is something basic I am missing)

This did not work out well.  It still has major problems because:
- Object allocated in python, passed to C++, and then returned from
C++ can't be compared
- Objects can't be compared to None anymore

Here is what I have done so far, if anyone has suggestions please let
me know (code generated from py++).

This is a combination of all the ideas discussed in this thread.  You
will notice that instead of defining an __eq__ I defined __cmp__ so
all the comparison operations would work.

typedef ::boost::python::class_< Group_wrapper,
::boost::python::bases< ::osg::GroupBase >, OSG::GroupRefPtr,
::boost::noncopyable > Group_exposer_t;
        Group_exposer_t Group_exposer = Group_exposer_t( "Group",
::boost::python::no_init );
...
        Group_exposer.def("__cmp__",pyopensg::cmp_ref_ptrs<OSG::GroupRefPtr>);
        Group_exposer.def("__hash__",pyopensg::hash_ref_ptr<OSG::GroupRefPtr>);
        register_fcptr_to_python<RefPtrType>::execute();
        fcptr_from_python<typename RefPtrType::FCPtrType>();
        bp::implicitly_convertible< OSG::GroupRefPtr, OSG::GroupPtr >();
        bp::implicitly_convertible< OSG::GroupRefPtr, OSG::NodeCoreRefPtr >();
    }
}

// Helpers are attached to the mail

There is also a class Node involved in the example code below.  In the
example I use two methods from there:

void Node::setCore(ptr<Base>);
ptr<Base> Node::getCore();

Here is some code that shows the problems

>>> g = osg.Group.create()
>>> g2 = osg.Group.create()
>>> g3 = g
>>> g == g2
False
>>> g == g3
True
>>> hash(g)
163
>>> hash(g2)
164
>>> n = osg.Node.create()
>>> n.setCore(g)
>>> core_g = n.getCore()
>>> core_g
<osg._osg.Group object at 0xafbf722c>
>>> g
<osg._osg.Group object at 0xb781fdac>

// Next line shows problem where the object coming back
// out of C++ can't be compared.  It is like it thinks it has
// a different type.
>>> g == core_g
Traceback (most recent call last):
  File "<input>", line 1, in ?
ArgumentError: Python argument types in
    Group.__cmp__(Group, Group)
did not match C++ signature:
    __cmp__(osg::RefPtr<osg::FCPtr<osg::FCPtr<osg::AttachmentContainerPtr,
osg::NodeCore>, osg::Group> >,
osg::RefPtr<osg::FCPtr<osg::FCPtr<osg::AttachmentContainerPtr,
osg::NodeCore>, osg::Group> >)

// This line shows the same problem with __hash__
>>> hash(core_g)
Traceback (most recent call last):
  File "<input>", line 1, in ?
ArgumentError: Python argument types in
    Group.__hash__(Group)
did not match C++ signature:
    __hash__(osg::RefPtr<osg::FCPtr<osg::FCPtr<osg::AttachmentContainerPtr,
osg::NodeCore>, osg::Group> >)

// And here is the problem comparing to None
>>> g == None
Traceback (most recent call last):
  File "<input>", line 1, in ?
ArgumentError: Python argument types in
    Group.__cmp__(Group, NoneType)
did not match C++ signature:
    __cmp__(osg::RefPtr<osg::FCPtr<osg::FCPtr<osg::AttachmentContainerPtr,
osg::NodeCore>, osg::Group> >,
osg::RefPtr<osg::FCPtr<osg::FCPtr<osg::AttachmentContainerPtr,
osg::NodeCore>, osg::Group> >)
>>>

I am at a loss to see what is causing this signature problem.  I don't
know of any way to ask boost.python to give me the "signature" of
python object at run-time, so I can't compare this signature with the
"good" signatures that work above.  From what I see though they look
very much the same.

I am going to try to trace back through the mailing list and look for
the object identity discussions you mentioned.  Maybe something in
those threads will provide a workaround that I can use.

As always, feel free to tell me if something looks way off.

Thanks,
Allen
-------------- next part --------------
A non-text attachment was scrubbed...
Name: bp_helpers.h
Type: text/x-chdr
Size: 2617 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/cplusplus-sig/attachments/20060907/c82f06a3/attachment.h>

From Gottfried.Ganssauge at haufe.de  Fri Sep  8 13:50:44 2006
From: Gottfried.Ganssauge at haufe.de (Ganssauge, Gottfried)
Date: Fri, 8 Sep 2006 13:50:44 +0200
Subject: [C++-sig] Bug in opaque_pointer_converter and patch
Message-ID: <51F4AD92DB06E9488BBC39B7B492409E027C1EB9@vg20ex1.grp.haufemg.com>

While debugging a python extension I found out that a segmentation fault
was produced by the python statement
	print type(some_opaque_pointer)

After some further debugging I found out that all instances created from
opaque_pointer_converter had a type object which itself didn't have a
type, and this finally lead to the crash.

Find attached a patch which corrects this behaviour by initialising the
type object using PyType_Ready().

Cheers,

Gottfried 
-------------- next part --------------
A non-text attachment was scrubbed...
Name: type_ready.patch
Type: application/octet-stream
Size: 850 bytes
Desc: type_ready.patch
URL: <http://mail.python.org/pipermail/cplusplus-sig/attachments/20060908/64f68afd/attachment.obj>

From ganssauge at gmx.de  Sat Sep  9 01:57:52 2006
From: ganssauge at gmx.de (=?ISO-8859-15?Q?Gottfried_Gan=DFauge?=)
Date: Sat, 09 Sep 2006 01:57:52 +0200
Subject: [C++-sig] New Patch for opaque_pointer_converter
Message-ID: <45020380.50906@gmx.de>

Please find attached a completely new patch against the current CVS Head.
This incorporates my patch from earlier today into the CVS Head.
Further I implemented cross module support in so far as the converter
registers itself only if not another module already has registered the
same converter before.
The opaque test case has been update to check for the segmentation fault
mentioned in my previous post.
A test case for the cross module support is included (crossmod_opaque).

I also updated the documentation for opaque.

I tested my patch with
 - GCC-4.1 running on GNU/Linux 2.6.16.21-0.13-default x86_64
 - Visual-C++-6.5 running on Windows XP
 - Visual-C++-7.1 running on Windows XP

The patch is applied by calling
	patch -p1 < opaque_patch
in the root of the boost hierarchy.

Files patched are:
boost_head/boost/python/opaque_pointer_converter.hpp
boost_head/libs/python/doc/v2/opaque.html
boost_head/libs/python/doc/v2/opaque_pointer_converter.html
boost_head/libs/python/doc/v2/reference.html
boost_head/libs/python/doc/v2/return_opaque_pointer.html
boost_head.orig/libs/python/test/crossmod_opaque_a.cpp
boost_head/libs/python/test/crossmod_opaque_a.cpp
boost_head/libs/python/test/crossmod_opaque_b.cpp
boost_head/libs/python/test/crossmod_opaque.py
boost_head/libs/python/test/Jamfile
boost_head/libs/python/test/Jamfile.v2
boost_head/libs/python/test/opaque.py

Enjoy ...

Gottfried
-------------- next part --------------
An embedded and charset-unspecified text was scrubbed...
Name: opaque_patch
URL: <http://mail.python.org/pipermail/cplusplus-sig/attachments/20060909/7cc88fe1/attachment.txt>

From roman.yakovenko at gmail.com  Sat Sep  9 22:29:32 2006
From: roman.yakovenko at gmail.com (Roman Yakovenko)
Date: Sat, 9 Sep 2006 23:29:32 +0300
Subject: [C++-sig] custom smart pointers
In-Reply-To: <44FEF8F8.8080004@pixar.com>
References: <7465b6170608291048y6e2010d4ofa84fbd5760da4d3@mail.gmail.com>
	<uac5lvczx.fsf@boost-consulting.com>
	<7465b6170608311358s1c02c6c9hc26222fae8cfdef6@mail.gmail.com>
	<7465b6170609051303w4132f343vc24d4cfd768408ed@mail.gmail.com>
	<44FDDC04.3010200@pixar.com>
	<7465b6170609052316w887a0b5v3723de9d1833491e@mail.gmail.com>
	<44FEF8F8.8080004@pixar.com>
Message-ID: <7465b6170609091329q19a74e78x90a9236d62d51527@mail.gmail.com>

On 9/6/06, Alex Mohr <amohr at pixar.com> wrote:
> > I don't know and understand how I should do it, really :-(.
> >
> > Boost.Python creates new instance of shared_ptr with custom deleter. This way it
> > says to shared_ptr that the object is already managed.
>
> I believe it's just a hack to do what I call "object identity" -- which
> is just tracking the association between a C++ instance and a python
> object.  When converting a shared ptr to-python, boost python checks to
> see if it has its special deleter, and will use that to produce the
> python object rather than constructing a new one.
>
> This is a fragile scheme.  It fails frequently for shared ptrs, as has
> been discussed on this list before, and it doesn't work at all for
> custom shared pointers.
>
> Anyway, I consider object identity to be a different issue from the one
> we're currently discussing.  But I do think you should address it in
> py++.  We can talk about that later.

I am not completely understand what you are talking about ( object identity ).
I will be glad to discuss it with you and add the appropriate functionality
to Py++, really.

> > The smart pointer, the project use, does not provide this
> > functionality. More over,
> > I don't understand what data->convertible will contain.
>
> If I recall correctly (I'm not looking at the code) data->convertible is
> the result that the convertible method returned, so in this case it's a
> raw pointer to the C++ instance.
>
> > I think I do it:
> >
> > register_ptr_to_python< my_smart_ptr_t< base > >();
> > register_ptr_to_python< my_smart_ptr_t< derived > >();
> > implicitly_convertible< my_smart_ptr_t< derived >, my_smart_ptr_t< base > >();
> >
> > This is what yoy mean, right?
>
> No, these are to-python conversions.  I'm talking about rvalue
> from-python conversions.  Have a look at the last few messages in the
> concurrent thread on this list with subject "Is there a way to
> automatically convert a smart_ptr to a held_type (try 2)".
>
> I posted a message to that thread which has some (untested) smart
> pointer rvalue from-python conversion code in it.  Actually -- it
> probably looks something like shared_ptr_from_python.

I read the thread "Is there a way to automatically convert a smart_ptr
to a held_type (try 2)" and still I am not able to get to a solution
:-(. I have to say that
I am lost.  I implemented rvalue converter,  I am registering custom
to_python_converter - nothing happens.

I attach the minimal example. Any help is welcome.

-- 
Roman Yakovenko
C++ Python language binding
http://www.language-binding.net/
-------------- next part --------------
A non-text attachment was scrubbed...
Name: classes.hpp
Type: text/x-c++hdr
Size: 1617 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/cplusplus-sig/attachments/20060909/746d8bfb/attachment.hpp>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: expose_classes.cpp
Type: text/x-c++src
Size: 4825 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/cplusplus-sig/attachments/20060909/746d8bfb/attachment.cpp>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: smart_ptr.h
Type: text/x-chdr
Size: 1843 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/cplusplus-sig/attachments/20060909/746d8bfb/attachment.h>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: test.py
Type: text/x-python
Size: 240 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/cplusplus-sig/attachments/20060909/746d8bfb/attachment.py>

From anderslanglands at gmail.com  Sun Sep 10 18:37:24 2006
From: anderslanglands at gmail.com (anders langlands)
Date: Sun, 10 Sep 2006 17:37:24 +0100
Subject: [C++-sig] Possible to inspect type of a boost::python::object
Message-ID: <57ffdb9c0609100937w3a144043k490470d5d82504a5@mail.gmail.com>

Currently I'm using extract<T>::check() over a bunch of different types to
decide what to do with python objects. Is there any way to inspect the type
directly?

A
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/cplusplus-sig/attachments/20060910/d1b25a78/attachment.htm>

From Gottfried.Ganssauge at haufe.de  Mon Sep 11 11:17:03 2006
From: Gottfried.Ganssauge at haufe.de (Ganssauge, Gottfried)
Date: Mon, 11 Sep 2006 11:17:03 +0200
Subject: [C++-sig] Patch for py++
Message-ID: <51F4AD92DB06E9488BBC39B7B492409E027C1F33@vg20ex1.grp.haufemg.com>

I had to make the following patch to
pyplusplus\code_repository\array_1.py in order to make it compile using
Visual-C++ 6.5:

Index: array_1.py
===================================================================
--- array_1.py	(Revision 494)
+++ array_1.py	(Arbeitskopie)
@@ -28,13 +28,14 @@
 
 inline void 
 raise_on_out_of_range( long unsigned int size, long unsigned int index
){
-    if( index < 0 || size <= index ){
+    if( size <= index ){
         throw std::out_of_range("index out of range");
     }
 }
 
 template< class TItemType, long unsigned int size >
 struct const_array_1_t{
+    typedef boost::call_traits<const TItemType>::param_type ref;
 
     const_array_1_t( TItemType const * const data )
     : m_data( data ){
@@ -47,8 +48,7 @@
         return size;
     }
     
-    typename boost::call_traits<const TItemType>::param_type
-    item_ref( long unsigned int index ) const{
+    ref item_ref( long unsigned int index ) const{
         raise_on_out_of_range( size, index );
         return m_data[index];
     }
@@ -61,6 +61,7 @@
 
 template< class TItemType, long unsigned int size >
 struct array_1_t{
+    typedef boost::call_traits<const TItemType>::param_type ref;
 
     array_1_t( TItemType * data )
     : m_data( data ){
@@ -73,15 +74,14 @@
         return size;
     }
     
-    typename boost::call_traits<const TItemType>::param_type
-    item_ref( long unsigned int index ) const{
+    ref item_ref( long unsigned int index ) const{
         raise_on_out_of_range( size, index );
         return m_data[index];
     }
 
     void
     set_item( long unsigned int index
-              , typename boost::call_traits<const
TItemType>::param_type new_value ){
+              ,  ref new_value ){
         raise_on_out_of_range( size, index );
         m_data[index] = new_value;
     }

Cheers,

Gottfried 


From roman.yakovenko at gmail.com  Mon Sep 11 11:54:34 2006
From: roman.yakovenko at gmail.com (Roman Yakovenko)
Date: Mon, 11 Sep 2006 12:54:34 +0300
Subject: [C++-sig] Patch for py++
In-Reply-To: <51F4AD92DB06E9488BBC39B7B492409E027C1F33@vg20ex1.grp.haufemg.com>
References: <51F4AD92DB06E9488BBC39B7B492409E027C1F33@vg20ex1.grp.haufemg.com>
Message-ID: <7465b6170609110254n69985a84n5f177305209ec94b@mail.gmail.com>

On 9/11/06, Ganssauge, Gottfried <Gottfried.Ganssauge at haufe.de> wrote:
> I had to make the following patch to
> pyplusplus\code_repository\array_1.py in order to make it compile using
> Visual-C++ 6.5:

Thank you. Unfortunately I was not able to apply the patch as is. It
breaks the compilation
on MSVC 7.1. I reworked the patch a little. I hope it will work on
your compiler too.
You can find the patch here.
http://svn.sourceforge.net/viewvc/pygccxml/pyplusplus_dev/pyplusplus/code_repository/array_1.py?view=markup

Can you test it? Thanks

-- 
Roman Yakovenko
C++ Python language binding
http://www.language-binding.net/


From dave at boost-consulting.com  Mon Sep 11 12:55:11 2006
From: dave at boost-consulting.com (David Abrahams)
Date: Mon, 11 Sep 2006 06:55:11 -0400
Subject: [C++-sig] array of strings in boost python
In-Reply-To: <MPG.1f6b9a371f63fd0e989681@news.gmane.org>
	(ctrembla@pcigeomatics.com's message of "Fri,
	8 Sep 2006 16:18:53 -0400")
References: <MPG.1f6b9a371f63fd0e989681@news.gmane.org>
Message-ID: <87k64aad0w.fsf@pereiro.luannocracy.com>

The following message is a courtesy copy of an article
that has been posted to gmane.comp.lib.boost.user as well.

<ctrembla at pcigeomatics.com> writes:

> Hi,
>
> newbie in python and boost::python

Please post  Boost.Python questions to the C++-sig:
http://www.boost.org/more/mailing_lists.htm#cplussig

> I have a method in c++ that returns a vector of strings, how can i 
> expose this method in python, and how should i use it in python

Sorry to turn the question around on you, but how do you /want/ to be
able to use it in Python?  That will determine how you should expose
it.

You might look into
http://boost.org/libs/python/doc/v2/indexing.html#vector_indexing_suite_class
to see if it meets your needs.


-- 
Dave Abrahams
Boost Consulting
www.boost-consulting.com


From anderslanglands at gmail.com  Mon Sep 11 12:56:51 2006
From: anderslanglands at gmail.com (anders langlands)
Date: Mon, 11 Sep 2006 11:56:51 +0100
Subject: [C++-sig] PyCode_New and new.code() functions
Message-ID: <57ffdb9c0609110356w609c73d3ic969cc28bcfe5dde@mail.gmail.com>

Does anyone know how to use either PyCode_New or new.code()?? Specifically I
want to use new.code() and new.function() to generate some python functions
from strings, which I can then call in python at some later time.

My trouble is I can find exactly zero documentation on the use of these
functions :( Any help would be greatly appreciated.

Cheers,

Anders
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/cplusplus-sig/attachments/20060911/69774fbd/attachment.htm>

From Gottfried.Ganssauge at haufe.de  Mon Sep 11 12:58:24 2006
From: Gottfried.Ganssauge at haufe.de (Ganssauge, Gottfried)
Date: Mon, 11 Sep 2006 12:58:24 +0200
Subject: [C++-sig] Patch for py++
In-Reply-To: <7465b6170609110254n69985a84n5f177305209ec94b@mail.gmail.com>
Message-ID: <51F4AD92DB06E9488BBC39B7B492409E027C1F67@vg20ex1.grp.haufemg.com>

> -----Original Message-----
> From: c++-sig-bounces at python.org 
> [mailto:c++-sig-bounces at python.org] On Behalf Of Roman Yakovenko
> Sent: Monday, September 11, 2006 11:55 AM
> To: Development of Python/C++ integration
> Subject: Re: [C++-sig] Patch for py++
> 
> On 9/11/06, Ganssauge, Gottfried <Gottfried.Ganssauge at haufe.de> wrote:
> > I had to make the following patch to
> > pyplusplus\code_repository\array_1.py in order to make it compile 
> > using
> > Visual-C++ 6.5:
> 
> Thank you. Unfortunately I was not able to apply the patch as 
> is. It breaks the compilation on MSVC 7.1. I reworked the 
> patch a little. I hope it will work on your compiler too.
> You can find the patch here.
> http://svn.sourceforge.net/viewvc/pygccxml/pyplusplus_dev/pypl
> usplus/code_repository/array_1.py?view=markup
> 
> Can you test it? Thanks
Works :-)

Cheers,

Gottfried  
> 
> --
> Roman Yakovenko
> C++ Python language binding
> http://www.language-binding.net/
> _______________________________________________
> C++-sig mailing list
> C++-sig at python.org
> http://mail.python.org/mailman/listinfo/c++-sig
> 


From seefeld at sympatico.ca  Mon Sep 11 13:40:59 2006
From: seefeld at sympatico.ca (Stefan Seefeld)
Date: Mon, 11 Sep 2006 07:40:59 -0400
Subject: [C++-sig] PyCode_New and new.code() functions
In-Reply-To: <57ffdb9c0609110356w609c73d3ic969cc28bcfe5dde@mail.gmail.com>
References: <57ffdb9c0609110356w609c73d3ic969cc28bcfe5dde@mail.gmail.com>
Message-ID: <45054B4B.1000303@sympatico.ca>

anders langlands wrote:
> Does anyone know how to use either PyCode_New or new.code()??
> Specifically I
> want to use new.code() and new.function() to generate some python functions
> from strings, which I can then call in python at some later time.

Not knowing what PyCode_New or new.code() do (do you have a pointer ?),
can't you use boost::python::exec() to convert strings to executable code ?

(David, http://www.boost.org/libs/python/doc/ doesn't seem to contain the
'new' documentation for exec and exec_file yet. Is that intentional ?)

Regards,
		Stefan

-- 

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


From anderslanglands at gmail.com  Mon Sep 11 14:24:31 2006
From: anderslanglands at gmail.com (anders langlands)
Date: Mon, 11 Sep 2006 13:24:31 +0100
Subject: [C++-sig] PyCode_New and new.code() functions
In-Reply-To: <45054B4B.1000303@sympatico.ca>
References: <57ffdb9c0609110356w609c73d3ic969cc28bcfe5dde@mail.gmail.com>
	<45054B4B.1000303@sympatico.ca>
Message-ID: <57ffdb9c0609110524w53fc9337ya212489ab859bda4@mail.gmail.com>

Hi Stefan,

Well PyCode_New is undocumented as far as I can see, and the header doesn't
even name the arguments!

new.code is documented in the library reference under module new. Its
documentation is equally illuminating: it simply says "An interface to
PyCode_New"...

boost::python::exec sounds like exactly what I want... but I can find no
documentation for this either, do you have I link? Nor can I find it in any
of the boost::python headers. we're using boost 1_33_1 here at the moment.

A

On 9/11/06, Stefan Seefeld <seefeld at sympatico.ca> wrote:
>
> anders langlands wrote:
> > Does anyone know how to use either PyCode_New or new.code()??
> > Specifically I
> > want to use new.code() and new.function() to generate some python
> functions
> > from strings, which I can then call in python at some later time.
>
> Not knowing what PyCode_New or new.code() do (do you have a pointer ?),
> can't you use boost::python::exec() to convert strings to executable code
> ?
>
> (David, http://www.boost.org/libs/python/doc/ doesn't seem to contain the
> 'new' documentation for exec and exec_file yet. Is that intentional ?)
>
> Regards,
>                 Stefan
>
> --
>
>       ...ich hab' noch einen Koffer in Berlin...
> _______________________________________________
> 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/20060911/3944662b/attachment.htm>

From seefeld at sympatico.ca  Mon Sep 11 15:24:14 2006
From: seefeld at sympatico.ca (Stefan Seefeld)
Date: Mon, 11 Sep 2006 09:24:14 -0400
Subject: [C++-sig] PyCode_New and new.code() functions
In-Reply-To: <57ffdb9c0609110524w53fc9337ya212489ab859bda4@mail.gmail.com>
References: <57ffdb9c0609110356w609c73d3ic969cc28bcfe5dde@mail.gmail.com>	<45054B4B.1000303@sympatico.ca>
	<57ffdb9c0609110524w53fc9337ya212489ab859bda4@mail.gmail.com>
Message-ID: <4505637E.5000706@sympatico.ca>

anders langlands wrote:

> boost::python::exec sounds like exactly what I want... but I can find no
> documentation for this either, do you have I link? Nor can I find it in any
> of the boost::python headers. we're using boost 1_33_1 here at the moment.

Ah, well, I added it after the 1_33_1 release. It will be included in the 1_34
release (in case that is of any help).

However, the implementation is really simple, so you may as well include the
code directory:


namespace boost
{
namespace python
{
object exec(str string, object global, object local)
{
  // should be 'char const *' but older python versions don't use 'const' yet.
  char *s = python::extract<char *>(string);
  PyObject* result = PyRun_String(s, Py_file_input, global.ptr(), local.ptr());
  if (!result) throw_error_already_set();
  return object(detail::new_reference(result));
}
}
}

This executes 'string' in the context of 'global' and 'local'. The latter two
are (potentially empty) dictionaries. Once execution has successfully terminated,
'global' will contain whatever objects 'string' injected into them such as type definitions
and other objects. You can then extract these from 'global' and call them from C++.
See for example
http://boost.cvs.sourceforge.net/*checkout*/boost/boost/libs/python/test/exec.cpp?revision=1.5.2.1

HTH,
		Stefan

-- 

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


From anderslanglands at gmail.com  Mon Sep 11 16:17:07 2006
From: anderslanglands at gmail.com (anders langlands)
Date: Mon, 11 Sep 2006 15:17:07 +0100
Subject: [C++-sig] PyCode_New and new.code() functions
In-Reply-To: <4505637E.5000706@sympatico.ca>
References: <57ffdb9c0609110356w609c73d3ic969cc28bcfe5dde@mail.gmail.com>
	<45054B4B.1000303@sympatico.ca>
	<57ffdb9c0609110524w53fc9337ya212489ab859bda4@mail.gmail.com>
	<4505637E.5000706@sympatico.ca>
Message-ID: <57ffdb9c0609110717u4c7650e1m9932a033742f05d@mail.gmail.com>

Thanks stefan, that does indeed look pretty simple!

However, is there any way I can do this with having my own interpreter?

The problem I am trying to solve is this:

I have a large class library I want to bind to python. Everything in the
library (Derived) is derived ultimately from a single class, Base. Now I
have functions to serialize and deserialize objects to disk. The deserialize
function returns a Derived object as a Base*. When this Base* is passed back
to python, python doesn't know that it's actually a Derived object, and it
would be nice if it did. I basically want to be able to automatically cast
the Base object to whichever flavour of Derived object it actually is in
python.

No all my objects support the functions typeId() and typename() to get
information about what they are. So the solution I've come up with is this.
I just store a big dict or list in the globals which maps typeId's to
functions. These functions do the conversion of the base type to the derived
type and the whole thing looks something like this:

-------

def castDerivedTypeA( obj ):
        return DerivedTypeA( obj )

def castDerivedTypeB( obj ):
        return DerivedTypeB( obj )

_globalCastsDict = {
        DerivedTypeATypeId : castDerivedTypeA,
        DerivedTypeBTypeId : castDerivedTypeB,
}

def cast( obj ):
        return _globalCastsDict[ obj.typeId() ]( obj )

--------

Now obviously with a large class library it's going to be tedious and
error-prone to maintain all those functions etc by hand. So what I'd like is
to automatically generate the code in C++ at binding time. What I'd like is
to do something like:

BOOST_PYTHON_MODULE( _mymodule )
{
        class_<DerivedA,bases<Base> >( "DerivedA" )
        ;

        exec( "code for castDerivedA function",
current_interpreter.globals(), current_interpreter.locals() );
}

Is it possible to get the globals from the currently running interpreter
without actually writing the interpreter executable yourself? I realise I
could just write my own interpreter and all would be fine and dandy, but I'd
like this to be able to work even if I'm running in the standard python
interrpreter

A




On 9/11/06, Stefan Seefeld <seefeld at sympatico.ca> wrote:
>
> anders langlands wrote:
>
> > boost::python::exec sounds like exactly what I want... but I can find no
> > documentation for this either, do you have I link? Nor can I find it in
> any
> > of the boost::python headers. we're using boost 1_33_1 here at the
> moment.
>
> Ah, well, I added it after the 1_33_1 release. It will be included in the
> 1_34
> release (in case that is of any help).
>
> However, the implementation is really simple, so you may as well include
> the
> code directory:
>
>
> namespace boost
> {
> namespace python
> {
> object exec(str string, object global, object local)
> {
>   // should be 'char const *' but older python versions don't use 'const'
> yet.
>   char *s = python::extract<char *>(string);
>   PyObject* result = PyRun_String(s, Py_file_input, global.ptr(),
> local.ptr());
>   if (!result) throw_error_already_set();
>   return object(detail::new_reference(result));
> }
> }
> }
>
> This executes 'string' in the context of 'global' and 'local'. The latter
> two
> are (potentially empty) dictionaries. Once execution has successfully
> terminated,
> 'global' will contain whatever objects 'string' injected into them such as
> type definitions
> and other objects. You can then extract these from 'global' and call them
> from C++.
> See for example
>
> http://boost.cvs.sourceforge.net/*checkout*/boost/boost/libs/python/test/exec.cpp?revision=1.5.2.1
>
> HTH,
>                 Stefan
>
> --
>
>       ...ich hab' noch einen Koffer in Berlin...
> _______________________________________________
> 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/20060911/5029b07c/attachment.htm>

From seefeld at sympatico.ca  Tue Sep 12 03:19:39 2006
From: seefeld at sympatico.ca (Stefan Seefeld)
Date: Mon, 11 Sep 2006 21:19:39 -0400
Subject: [C++-sig] PyCode_New and new.code() functions
In-Reply-To: <57ffdb9c0609110717u4c7650e1m9932a033742f05d@mail.gmail.com>
References: <57ffdb9c0609110356w609c73d3ic969cc28bcfe5dde@mail.gmail.com>	<45054B4B.1000303@sympatico.ca>	<57ffdb9c0609110524w53fc9337ya212489ab859bda4@mail.gmail.com>	<4505637E.5000706@sympatico.ca>
	<57ffdb9c0609110717u4c7650e1m9932a033742f05d@mail.gmail.com>
Message-ID: <45060B2B.1000703@sympatico.ca>

anders langlands wrote:
> Thanks stefan, that does indeed look pretty simple!
> 
> However, is there any way I can do this with having my own interpreter?
> 
> The problem I am trying to solve is this:
> 
> I have a large class library I want to bind to python. Everything in the
> library (Derived) is derived ultimately from a single class, Base. Now I
> have functions to serialize and deserialize objects to disk. The
> deserialize
> function returns a Derived object as a Base*. When this Base* is passed
> back
> to python, python doesn't know that it's actually a Derived object, and it
> would be nice if it did. I basically want to be able to automatically cast
> the Base object to whichever flavour of Derived object it actually is in
> python.
> 
> No all my objects support the functions typeId() and typename() to get
> information about what they are. So the solution I've come up with is this.
> I just store a big dict or list in the globals which maps typeId's to
> functions. These functions do the conversion of the base type to the
> derived
> type and the whole thing looks something like this:

I'm not sure I understand the purpose of your downcasting. Those derived
python classes don't offer any additional attributes over what 'Base'
already has, or do they ?
If not, what advantage do you get from having the type names differ,
instead of using your own type system (provided by 'typeId()' and 'typename()' ?

If you really want to generate python types for the many C++ classes,
and you can do the conversion at runtime (as opposed to compile-time),
you may consider doing all of this from within a python module, by means
of the 'type()' constructor.

But again, I don't yet understand what you are aiming at, since that
won't give you anything you couldn't get by directly querying an object's
typeId().

Thanks,
		Stefan

-- 

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


From dave at boost-consulting.com  Tue Sep 12 03:21:22 2006
From: dave at boost-consulting.com (David Abrahams)
Date: Mon, 11 Sep 2006 21:21:22 -0400
Subject: [C++-sig] Possible to inspect type of a boost::python::object
References: <57ffdb9c0609100937w3a144043k490470d5d82504a5@mail.gmail.com>
Message-ID: <8764ftrib1.fsf@pereiro.luannocracy.com>

"anders langlands" <anderslanglands at gmail.com> writes:

> Currently I'm using extract<T>::check() over a bunch of different types to
> decide what to do with python objects. Is there any way to inspect the type
> directly?

You can inspect the _Python_ type directly:

    x.ptr()->ob_type;

or if all the types have a common C++ base class with at least one
virtual function, you can write:

   typeid(extract<Base*>(x)())

HTH,

-- 
Dave Abrahams
Boost Consulting
www.boost-consulting.com



From dave at boost-consulting.com  Tue Sep 12 03:22:58 2006
From: dave at boost-consulting.com (David Abrahams)
Date: Mon, 11 Sep 2006 21:22:58 -0400
Subject: [C++-sig] New Patch for opaque_pointer_converter
References: <45020380.50906@gmx.de>
Message-ID: <87zmd5q3nx.fsf@pereiro.luannocracy.com>

Gottfried Gan?auge <ganssauge at gmx.de> writes:

> Please find attached a completely new patch against the current CVS Head.
> This incorporates my patch from earlier today into the CVS Head.
> Further I implemented cross module support in so far as the converter
> registers itself only if not another module already has registered the
> same converter before.
> The opaque test case has been update to check for the segmentation fault
> mentioned in my previous post.
> A test case for the cross module support is included (crossmod_opaque).
>
> I also updated the documentation for opaque.
>
> I tested my patch with
>  - GCC-4.1 running on GNU/Linux 2.6.16.21-0.13-default x86_64
>  - Visual-C++-6.5 running on Windows XP
>  - Visual-C++-7.1 running on Windows XP
>

Feel free to check it in.  If it's just a bug fix, you can merge it
into the release branch as well.

-- 
Dave Abrahams
Boost Consulting
www.boost-consulting.com



From dave at boost-consulting.com  Tue Sep 12 03:29:58 2006
From: dave at boost-consulting.com (David Abrahams)
Date: Mon, 11 Sep 2006 21:29:58 -0400
Subject: [C++-sig] [Py++] First imressions
References: <44FD9785.9000204@lapshin.net>
Message-ID: <87pse1q3c9.fsf@pereiro.luannocracy.com>

Kirill Lapshin <kir at lapshin.net> writes:

> 1. The way overrides are generated does not compile on MSVC8 (2005). 
> Looks like the problem is in conversion of return type if return type is 
> bool or enum. For example
>
>
> virtual bool is_representative( ) const {
> bp::override func_is_representative = this->get_override(
> "is_representative" );
> return func_is_representative( ); // errors out
> }
>
> error C2440: 'return' : cannot convert from
> 'boost::python::detail::method_result' to 'bool'
> No user-defined-conversion operator available that can perform this
> conversion, or the operator cannot be called

This is a compiler bug, which is worked around by the enclosed patch
to boost/python/override.hpp:

-------------- next part --------------
A non-text attachment was scrubbed...
Name: override.patch
Type: text/x-patch
Size: 638 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/cplusplus-sig/attachments/20060911/98705b3f/attachment.bin>
-------------- next part --------------

Note that your workaround would be more idiomatic if you wrote:

     return func_is_representative().as<bool>();

-- 
Dave Abrahams
Boost Consulting
www.boost-consulting.com

From dave at boost-consulting.com  Tue Sep 12 03:35:20 2006
From: dave at boost-consulting.com (David Abrahams)
Date: Mon, 11 Sep 2006 21:35:20 -0400
Subject: [C++-sig] What if a Python override returns several values?
References: <edhhnt$khq$1@sea.gmane.org> <44FDABA2.60809@pixar.com>
Message-ID: <87k649q33b.fsf@pereiro.luannocracy.com>

Alex Mohr <amohr at pixar.com> writes:

>> But if I have another virtual member function "void g(int& a, int &b)" 
>> which returns two integers, then I'd like the Python version to take no 
>> arguments and return a tuple (a,b). But how does the wrapper look like 
>> in this case? I would have expected something like:
>> 
>>      void g(int& a, int& b)
>>      {
>>          boost::python::tuple tup = this->get_override("g")();
>> 
>> 	a = ...extract an int from tup[0]...
>>          b = ...extract an int from tup[1]...
>>      }
>> 
>> But when I run this code I get the following error at runtime:
>> 
>> TypeError: No registered converter was able to extract a C++ reference 
>> to type boost::python::tuple from this Python object of type tuple
>
> Not sure exactly what's going on here, but you might be able to find out 
> by looking at override.hpp.  Look at method_result -- that's what you're 
> getting back from get_override.
>
> Off the top of my head, I think this would probably work for you:
>
> tuple tup = call<tuple>(this->get_override("g").ptr());

> a = ...extract an int from tup[0]
> b = ...extract an int from tup[1]


I don't know why people always use ugly low-level interfaces when they
can use beautiful high-level ones:

    python::tuple tup = this->get_override("g")();
    a = extract<t1>(tup[0]);
    .
    .
    .

There's almost never a reason to use call<> anymore!  I should
deprecate it.

I'm not trying to be critical, but this really concerns me.  I see
this sort of thing all the time.  If you could help me understand
why you were moved to suggest doing it the hard way, I would
appreciate it.

> It's certainly not as nice as what you wrote, 

But it could be. :)

-- 
Dave Abrahams
Boost Consulting
www.boost-consulting.com



From dave at boost-consulting.com  Tue Sep 12 03:53:18 2006
From: dave at boost-consulting.com (David Abrahams)
Date: Mon, 11 Sep 2006 21:53:18 -0400
Subject: [C++-sig] Returning a Python object to C++ (is get_owner the
	right way?)
References: <20060901202535.GC28064@randombit.net>
Message-ID: <87ejuhq29d.fsf@pereiro.luannocracy.com>

Jack Lloyd <lloyd at randombit.net> writes:

> I have a scenario where I'm wrapping a virtual base class 

You mean a base class with virtual functions, or an abstract base
class?  I'm pretty sure you don't mean

        class a : public virtual b { ... };

That's a virtual base class.  Pedantry aside...

> that offers cloning (here, sending a clone message to an object
> means "return a newly constructed object of whatever type you are"),
> and want to inherit from that type in Python. Simply implementing
> that function in Python (eg:>

> class MySubclass(mylib.MyBaseClass):
>    def clone(self):
>       return MySubclass()
>
> led to "returning a dangling reference" errors, so I'm using the
> technique Dave suggested in
> http://mail.python.org/pipermail/c++-sig/2004-December/008379.html
> (basically, use get_owner, get ahold of the __class__ attribute, and
> call the constructor, then keep a reference to the copy-constructed
> Python object around so it is not destroyed until the lifetime of the
> C++ object is over, all held inside another class that simply forwards
> the appropriate messages to the Python object).

       
>
> This has been working fine for me, however since get_owner is still
> undocumented (AFAICT; couldn't find it anywhere in the 1.33.1 docs),
> and generally the technique led to some rather ugly code, I wanted to
> check and see if any new technique/method for dealing with this
> situation had developed since that discussion?

Not unless you're willing to change the clone member function to
return shared_ptr<Base> instead of a raw pointer.

It would be fairly easy to factor out most of the interesting bits of
that posting into a new library component, or just put them in
wrapper<>, though.  I wish I had more time....

-- 
Dave Abrahams
Boost Consulting
www.boost-consulting.com



From dave at boost-consulting.com  Tue Sep 12 05:26:05 2006
From: dave at boost-consulting.com (David Abrahams)
Date: Mon, 11 Sep 2006 23:26:05 -0400
Subject: [C++-sig] custom smart pointers
References: <7465b6170608291048y6e2010d4ofa84fbd5760da4d3@mail.gmail.com>
	<uac5lvczx.fsf@boost-consulting.com>
	<7465b6170608311358s1c02c6c9hc26222fae8cfdef6@mail.gmail.com>
Message-ID: <878xkppxyq.fsf@pereiro.luannocracy.com>

"Roman Yakovenko" <roman.yakovenko at gmail.com> writes:

> On 8/31/06, David Abrahams <dave at boost-consulting.com> wrote:
>> >From this I can't understand what you're doing or what errors you're
>> getting.  Please do what I tell everyone to do: make a _complete_,
>> _minimal_ example that reproduces your problem and post that.
>
> I attached 3 files. This is complete and minimal example.

It's hardly minimal!

A minimal example would have no "documentation" strings, no use of
named arguments (bp::arg(...)) and no use of things that you get by
default anyway, like default_call_policies.  It would not contain
extraneous implicit conversions from int.  This is a problem with many
code generators: they create an unreadable mess for anyone who wants
to understand the result.  Part of the reason many people despise the
MFC Wizards.  I suggest you fix Py++ so it doesn't add vacuous
"documentation" and other needless cruft.

In a minimal example, there would be just one .cpp file, one .py file,
and if possible, a Jamfile I can use to build and run the test.

BTW, your #include guards use identifiers that are reserved to the C++
implementation.

I made your example work.  I had to add an operator* to your smart
pointer, although I think that uncovers a bug in Boost.Python: we
should be using get_pointer where it needed the operator*, instead.

I don't know if this meets your needs or not, but the result is
enclosed.

-------------- next part --------------
A non-text attachment was scrubbed...
Name: test.tar.gz
Type: application/octet-stream
Size: 1839 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/cplusplus-sig/attachments/20060911/c627c087/attachment.obj>
-------------- next part --------------

HTH,

-- 
Dave Abrahams
Boost Consulting
www.boost-consulting.com

From dave at boost-consulting.com  Tue Sep 12 05:29:24 2006
From: dave at boost-consulting.com (David Abrahams)
Date: Mon, 11 Sep 2006 23:29:24 -0400
Subject: [C++-sig] RuntimeError: unidentifiable C++ exception
References: <1f527df80608300109q5f9df417sf8d1a23023a0a3b7@mail.gmail.com>
Message-ID: <873baxpxt7.fsf@pereiro.luannocracy.com>

"Qun Cao" <quncao at gmail.com> writes:

> Hi Everyone,
>
> I created a simple class wrapper like this
>
> BOOST_PYTHON_MODULE(PythonTest) {
> 	class_<MyClass,boost::noncopyable>("pMyClass", init<char const*>());
> }
>
> When I try to initialize pMyClass in python, I got this error:
>
> Traceback (most recent call last):
>   File "<stdin>", line 1, in ?
> RuntimeError: unidentifiable C++ exception
>
> MyClass is part of a third party GameEngine, I can successfully use it in C++.
> Can anyone suggest what might went wrong here?

Either:

a. Your C++ class threw some exception for which there's no
   pre-registered Boost.Python exception translator, or

b. You're running on windows and your C++ code crashed.  To find out
   if this is the case, add the following code to your module source
   file:

  # ifdef _MSC_VER
  #  pragma warning(push)
  #  pragma warning(disable:4297)
  #  pragma warning(disable:4535)
  extern "C" void straight_to_debugger(unsigned int, EXCEPTION_POINTERS*)
  {
      throw;
  }
  extern "C" void (*old_translator)(unsigned, EXCEPTION_POINTERS*)
           = _set_se_translator(straight_to_debugger);
  #  pragma warning(pop)
  # endif

  That will cause crashes to look like crashes rather than exceptions.

-- 
Dave Abrahams
Boost Consulting
www.boost-consulting.com



From dave at boost-consulting.com  Tue Sep 12 05:30:24 2006
From: dave at boost-consulting.com (David Abrahams)
Date: Mon, 11 Sep 2006 23:30:24 -0400
Subject: [C++-sig] [tutorial] Auto-Overloading example bug?
References: <7465b6170608300429n564516bfi22c665cd6cbfca32@mail.gmail.com>
	<44F666B8.9080609@boost-consulting.com>
	<44F67B9F.5060802@boost-consulting.com>
	<7465b6170608311021y2f5bebecr90ab1766c5b84ad2@mail.gmail.com>
	<44F77280.80103@boost-consulting.com>
Message-ID: <87wt89oj73.fsf@pereiro.luannocracy.com>

Joel de Guzman <djowel at gmail.com> writes:

> Roman Yakovenko wrote:
>> On 8/31/06, Joel de Guzman <djowel at gmail.com> wrote:
>>> Ok, disregard that. I did check it out and the error is the def it
>>> should be:
>>>
>>>      .def("foo", (void(*)(bool, int, char))0, foo_overloads());
>>>
>>> I fixed the docs and updated to CVS.
>> 
>> Do you want to consider to fix "def"?
>> 
>> .def("foo", foo_overloads() );
>> 
>> It is not clear why user should pass second argument.
>
> I don't think it can be fixed. The compiler does not have enough
> information in there to know the exact signature needed. I may
> be wrong though, so I'll investigate. If you know how to implement
> this, I'm all ears.

I don't think you're wrong.

-- 
Dave Abrahams
Boost Consulting
www.boost-consulting.com



From dave at boost-consulting.com  Tue Sep 12 05:39:06 2006
From: dave at boost-consulting.com (David Abrahams)
Date: Mon, 11 Sep 2006 23:39:06 -0400
Subject: [C++-sig] Is there a way to automatically convert a smart_ptr
	to a held_type (try 2)
References: <bea9869d0608300707r3009cdb2u101ab96b88ec9966@mail.gmail.com>
	<44FE0AB2.8010006@pixar.com>
	<bea9869d0609061406lddb16fx5165bcc01e8814f3@mail.gmail.com>
	<44FF3B6F.9030509@pixar.com>
	<bea9869d0609061432k1c408ba5i8337efa421611360@mail.gmail.com>
	<bea9869d0609070809q13feb423yc9949b3206823625@mail.gmail.com>
	<45004B36.1020800@pixar.com>
	<bea9869d0609070946x3589060bgcc1a539868451b4c@mail.gmail.com>
	<45004EE9.7040604@pixar.com>
	<bea9869d0609071132of696703qde7d7e1080868969@mail.gmail.com>
	<bea9869d0609071337l26d001der8f815ecd30067301@mail.gmail.com>
Message-ID: <87r6yhoisl.fsf@pereiro.luannocracy.com>


I don't have time to do a complete analysis right now, but maybe I can
help understand these signatures:

"Allen Bierbaum" <abierbaum at gmail.com> writes:

> // Next line shows problem where the object coming back
> // out of C++ can't be compared.  It is like it thinks it has
> // a different type.
>>>> g == core_g
> Traceback (most recent call last):
>  File "<input>", line 1, in ?
> ArgumentError: Python argument types in
>    Group.__cmp__(Group, Group)
> did not match C++ signature:
>    __cmp__(osg::RefPtr<osg::FCPtr<osg::FCPtr<osg::AttachmentContainerPtr,
> osg::NodeCore>, osg::Group> >,
> osg::RefPtr<osg::FCPtr<osg::FCPtr<osg::AttachmentContainerPtr,
> osg::NodeCore>, osg::Group> >)

It's not that the Python types are different (both are Group).  It's
much more likely that they have a different way of holding their
internal C++ osg::Group object such that one can be converted to that
RefPtr type and the other can't.  What can be converted is up to the
converters you've registered.

> // This line shows the same problem with __hash__
>>>> hash(core_g)
> Traceback (most recent call last):
>  File "<input>", line 1, in ?
> ArgumentError: Python argument types in
>    Group.__hash__(Group)
> did not match C++ signature:
>    __hash__(osg::RefPtr<osg::FCPtr<osg::FCPtr<osg::AttachmentContainerPtr,
> osg::NodeCore>, osg::Group> >)
>
> // And here is the problem comparing to None
>>>> g == None
> Traceback (most recent call last):
>  File "<input>", line 1, in ?
> ArgumentError: Python argument types in
>    Group.__cmp__(Group, NoneType)
> did not match C++ signature:
>    __cmp__(osg::RefPtr<osg::FCPtr<osg::FCPtr<osg::AttachmentContainerPtr,
> osg::NodeCore>, osg::Group> >,
> osg::RefPtr<osg::FCPtr<osg::FCPtr<osg::AttachmentContainerPtr,
> osg::NodeCore>, osg::Group> >)
>>>>

This could be a case of the first (self) object not being convertible
to the RefPtr.

-- 
Dave Abrahams
Boost Consulting
www.boost-consulting.com



From roman.yakovenko at gmail.com  Tue Sep 12 07:03:00 2006
From: roman.yakovenko at gmail.com (Roman Yakovenko)
Date: Tue, 12 Sep 2006 08:03:00 +0300
Subject: [C++-sig] [tutorial] Auto-Overloading example bug?
In-Reply-To: <87wt89oj73.fsf@pereiro.luannocracy.com>
References: <7465b6170608300429n564516bfi22c665cd6cbfca32@mail.gmail.com>
	<44F666B8.9080609@boost-consulting.com>
	<44F67B9F.5060802@boost-consulting.com>
	<7465b6170608311021y2f5bebecr90ab1766c5b84ad2@mail.gmail.com>
	<44F77280.80103@boost-consulting.com>
	<87wt89oj73.fsf@pereiro.luannocracy.com>
Message-ID: <7465b6170609112203h69b9484bw63bca80b48a783bb@mail.gmail.com>

On 9/12/06, David Abrahams <dave at boost-consulting.com> wrote:
> Joel de Guzman <djowel at gmail.com> writes:
>
> > Roman Yakovenko wrote:
> >> On 8/31/06, Joel de Guzman <djowel at gmail.com> wrote:
> >>> Ok, disregard that. I did check it out and the error is the def it
> >>> should be:
> >>>
> >>>      .def("foo", (void(*)(bool, int, char))0, foo_overloads());
> >>>
> >>> I fixed the docs and updated to CVS.
> >>
> >> Do you want to consider to fix "def"?
> >>
> >> .def("foo", foo_overloads() );
> >>
> >> It is not clear why user should pass second argument.
> >
> > I don't think it can be fixed. The compiler does not have enough
> > information in there to know the exact signature needed. I may
> > be wrong though, so I'll investigate. If you know how to implement
> > this, I'm all ears.
>
> I don't think you're wrong.

Actually my idea was simple: foo_overloads class derives from
boost::python::detail::overloads_base class. I think you can use this
fact and to remove
the need to pass function signature to the "def".


-- 
Roman Yakovenko
C++ Python language binding
http://www.language-binding.net/


From anderslanglands at gmail.com  Tue Sep 12 10:50:47 2006
From: anderslanglands at gmail.com (anders langlands)
Date: Tue, 12 Sep 2006 09:50:47 +0100
Subject: [C++-sig] PyCode_New and new.code() functions
In-Reply-To: <45060B2B.1000703@sympatico.ca>
References: <57ffdb9c0609110356w609c73d3ic969cc28bcfe5dde@mail.gmail.com>
	<45054B4B.1000303@sympatico.ca>
	<57ffdb9c0609110524w53fc9337ya212489ab859bda4@mail.gmail.com>
	<4505637E.5000706@sympatico.ca>
	<57ffdb9c0609110717u4c7650e1m9932a033742f05d@mail.gmail.com>
	<45060B2B.1000703@sympatico.ca>
Message-ID: <57ffdb9c0609120150m3fc961fbg6c78df9c63370087@mail.gmail.com>

Sorry, I assumed it was clear since I know exactly what I mean!

All the derived types do indeed define many additional methods that I want
to use. By type() constructor, do you mean doing:

b = getDerivedObjectAsBasePointer
# This would fail as doSomeDerivedFunction is only defined for Derived, but
python thinks b is a Base*
# b.doSomeDerivedFunction()
d = Derived( b )
d.doSomeDerivedFunction()

This is basically what I want to do, but I want that to be hidden from the
user. I want to do the type-checking and construction automatically. So I'd
wrap getDerivedObjectAsBasePointer with a pure python function,
getDerivedObject, defined in my module's __init__.py.

I'd like to generate the code for this python function automatically from
C++ at module initialization time, since I know the typenames of all my
classes then, and each of them has to register itself anyway, so *if* I can
get access to the globals, I can define the dict which defines the functions
for doing run-time type() constructors on my base class pointers.

Does this make it any clearer? This doesn't gain me anything over doing the
type() constructors manually in python, but I'd like to save users this
extra leg work. It'll make the resulting scripts easier to read as well.

A

On 9/12/06, Stefan Seefeld <seefeld at sympatico.ca> wrote:
>
> anders langlands wrote:
> > Thanks stefan, that does indeed look pretty simple!
> >
> > However, is there any way I can do this with having my own interpreter?
> >
> > The problem I am trying to solve is this:
> >
> > I have a large class library I want to bind to python. Everything in the
> > library (Derived) is derived ultimately from a single class, Base. Now I
> > have functions to serialize and deserialize objects to disk. The
> > deserialize
> > function returns a Derived object as a Base*. When this Base* is passed
> > back
> > to python, python doesn't know that it's actually a Derived object, and
> it
> > would be nice if it did. I basically want to be able to automatically
> cast
> > the Base object to whichever flavour of Derived object it actually is in
> > python.
> >
> > No all my objects support the functions typeId() and typename() to get
> > information about what they are. So the solution I've come up with is
> this.
> > I just store a big dict or list in the globals which maps typeId's to
> > functions. These functions do the conversion of the base type to the
> > derived
> > type and the whole thing looks something like this:
>
> I'm not sure I understand the purpose of your downcasting. Those derived
> python classes don't offer any additional attributes over what 'Base'
> already has, or do they ?
> If not, what advantage do you get from having the type names differ,
> instead of using your own type system (provided by 'typeId()' and
> 'typename()' ?
>
> If you really want to generate python types for the many C++ classes,
> and you can do the conversion at runtime (as opposed to compile-time),
> you may consider doing all of this from within a python module, by means
> of the 'type()' constructor.
>
> But again, I don't yet understand what you are aiming at, since that
> won't give you anything you couldn't get by directly querying an object's
> typeId().
>
> Thanks,
>                 Stefan
>
> --
>
>       ...ich hab' noch einen Koffer in Berlin...
> _______________________________________________
> 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/20060912/e5a62518/attachment.htm>

From Gottfried.Ganssauge at haufe.de  Tue Sep 12 10:54:42 2006
From: Gottfried.Ganssauge at haufe.de (Ganssauge, Gottfried)
Date: Tue, 12 Sep 2006 10:54:42 +0200
Subject: [C++-sig] New Patch for opaque_pointer_converter
In-Reply-To: <87zmd5q3nx.fsf@pereiro.luannocracy.com>
Message-ID: <51F4AD92DB06E9488BBC39B7B492409E027C2016@vg20ex1.grp.haufemg.com>

> -----Original Message-----
> From: c++-sig-bounces at python.org 
> [mailto:c++-sig-bounces at python.org] On Behalf Of David Abrahams
> Sent: Tuesday, September 12, 2006 3:23 AM
> To: c++-sig at python.org
> Subject: Re: [C++-sig] New Patch for opaque_pointer_converter
> 
> Gottfried Gan?auge <ganssauge at gmx.de> writes:
> 
> > Please find attached a completely new patch against the 
> current CVS Head.
> > This incorporates my patch from earlier today into the CVS Head.
> > Further I implemented cross module support in so far as the 
> converter 
> > registers itself only if not another module already has 
> registered the 
> > same converter before.
> > The opaque test case has been update to check for the segmentation 
> > fault mentioned in my previous post.
> > A test case for the cross module support is included 
> (crossmod_opaque).
> >
> > I also updated the documentation for opaque.
> >
> > I tested my patch with
> >  - GCC-4.1 running on GNU/Linux 2.6.16.21-0.13-default x86_64
> >  - Visual-C++-6.5 running on Windows XP
> >  - Visual-C++-7.1 running on Windows XP
> >
> 
> Feel free to check it in.  If it's just a bug fix, you can 
> merge it into the release branch as well.
I would ... if I could :-)
I'm not an "authorized" boost developer (I'm _not_ planning to become one) so sourceforge won't let me in.

Cheers,

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


From Gottfried.Ganssauge at haufe.de  Tue Sep 12 12:49:24 2006
From: Gottfried.Ganssauge at haufe.de (Ganssauge, Gottfried)
Date: Tue, 12 Sep 2006 12:49:24 +0200
Subject: [C++-sig] Py++ - problem with split_module and
	return_opaque_pointer-policy
Message-ID: <51F4AD92DB06E9488BBC39B7B492409E029FDC9A@vg20ex1.grp.haufemg.com>

The return_value_policy<return_opaque_pointer> requires
an additional macro call (BOOST_PYTHON_OPAQUE_SPECIALIZED_TYPE_ID)
in the declaration section of a module.
This can be added e.g. using module_builder_t.add_declaration_code(),
and it works as long as you're generating the code using write_module().

In the moment you're using "split_module" to split your output, this
scheme breaks,
because BOOST_PYTHON_OPAQUE_SPECIALIZED_TYPE_ID is required in every
translation
unit using the opaque pointer type whose type_id is defined with that
macro.
Instead it is only generated into the main module.

An example:

test.hpp:
=========
typedef struct opaque_ *opaque_pointer;

opaque_pointer get_opaque(); 

test.py:
========
import os
from pyplusplus import module_builder
from pyplusplus.decl_wrappers import \
    return_value_policy, \
    return_opaque_pointer

mb = module_builder.module_builder_t( ["test.hpp"] )

mb.add_declaration_code
("BOOST_PYTHON_OPAQUE_SPECIALIZED_TYPE_ID(opaque_)")
mb.free_function (name = "get_opaque").call_policies = \
    return_value_policy (return_opaque_pointer)

mb.build_code_creator( module_name='test_opaque' )

mb.split_module( '.' ) 

When compiled with
	g++ -c -I/usr/include/python2.4 test_opaque*cpp
this produces the following sermon:
boost/python/type_id.hpp: In function `boost::python::type_info
boost::python::type_id() [with T = opaque_]':
boost/python/lvalue_from_pytype.hpp:23:   instantiated from
`boost::python::type_info
boost::python::detail::extractor_type_id(T&(*)(U)) [with T = opaque_, U
= boost::python::opaque_pointer_converter<opaque_*>::instance&]'
boost/python/lvalue_from_pytype.hpp:88:   instantiated from
`boost::python::lvalue_from_pytype<Extractor,
python_type>::lvalue_from_pytype() [with Extractor =
boost::python::opaque_pointer_converter<opaque_*>, const
PyTypeObject*python_type = ((const
PyTypeObject*)(&boost::python::opaque_pointer_converter<opaque_*>::type_
object))]'
boost/python/opaque_pointer_converter.hpp:116:   instantiated from `void
boost::python::opaque_pointer_converter<Pointer>::register_self() [with
Pointer = opaque_*]'
boost/python/opaque_pointer_converter.hpp:70:   instantiated from
`boost::python::opaque_pointer_converter<Pointer>::opaque_pointer_conver
ter(const char*) [with Pointer = opaque_*]'
boost/python/return_opaque_pointer.hpp:25:   instantiated from
`PyObject*
boost::python::detail::opaque_conversion_holder<Pointer>::operator()(Poi
nter) const [with Pointer = opaque_*]'
boost/python/detail/invoke.hpp:75:   instantiated from `PyObject*
boost::python::detail::invoke(boost::python::detail::invoke_tag_< false,
false>, const RC&, F&) [with RC =
boost::python::detail::opaque_conversion_holder<opaque_*>, F =
opaque_*(*)()]'
boost/python/detail/caller.hpp:199:   instantiated from `PyObject*
boost::python::detail::caller_arity<0u>::impl<F, Policies,
Sig>::operator()(PyObject*, PyObject*) [with F = opaque_*(*)(), Policies
=
boost::python::return_value_policy<boost::python::return_opaque_pointer,
boost::python::default_call_policies>, Sig =
boost::mpl::vector1<opaque_*>]'
boost/python/object/py_function.hpp:38:   instantiated from `PyObject*
boost::python::objects::caller_py_function_impl<Caller>::operator()(PyOb
ject*, PyObject*) [with Caller =
boost::python::detail::caller<opaque_*(*)(),
boost::python::return_value_policy<boost::python::return_opaque_pointer,
boost::python::default_call_policies>, boost::mpl::vector1<opaque_*> >]'
test_opaque_free_functions.pypp.cpp:21:   instantiated from here
boost/python/type_id.hpp:71: error: invalid use of undefined type
`struct opaque_'
test.hpp:1: error: forward declaration of `struct opaque_'

Cheers,

Gottfried 


From roman.yakovenko at gmail.com  Tue Sep 12 13:14:35 2006
From: roman.yakovenko at gmail.com (Roman Yakovenko)
Date: Tue, 12 Sep 2006 14:14:35 +0300
Subject: [C++-sig] Py++ - problem with split_module and
	return_opaque_pointer-policy
In-Reply-To: <51F4AD92DB06E9488BBC39B7B492409E029FDC9A@vg20ex1.grp.haufemg.com>
References: <51F4AD92DB06E9488BBC39B7B492409E029FDC9A@vg20ex1.grp.haufemg.com>
Message-ID: <7465b6170609120414r3935fd99mf0e957ad722f9542@mail.gmail.com>

On 9/12/06, Ganssauge, Gottfried <Gottfried.Ganssauge at haufe.de> wrote:
> The return_value_policy<return_opaque_pointer> requires
> an additional macro call (BOOST_PYTHON_OPAQUE_SPECIALIZED_TYPE_ID)
> in the declaration section of a module.
> This can be added e.g. using module_builder_t.add_declaration_code(),
> and it works as long as you're generating the code using write_module().
>
> In the moment you're using "split_module" to split your output, this
> scheme breaks,
> because BOOST_PYTHON_OPAQUE_SPECIALIZED_TYPE_ID is required in every
> translation
> unit using the opaque pointer type whose type_id is defined with that
> macro.
> Instead it is only generated into the main module.
>

Does "every translation unit" means "in every cpp file" ?
If not the quick work-around I can propose is to put this macro in
some header file
and to include it.

The solution to your problem requires small refactoring in Py++. I
will do that before next
release, but it will take some time.
The solution will be to add "add_declaration_code" to free function:

mb = module_builder.module_builder_t( ["test.hpp"] )

get_opaque = mb.free_function (name = "get_opaque")

get_opaque.add_declaration_code("BOOST_PYTHON_OPAQUE_SPECIALIZED_TYPE_ID(opaque_)")
get_opaque.call_policies = return_value_policy (return_opaque_pointer)

What do you think?

-- 
Roman Yakovenko
C++ Python language binding
http://www.language-binding.net/


From Gottfried.Ganssauge at haufe.de  Tue Sep 12 13:37:07 2006
From: Gottfried.Ganssauge at haufe.de (Ganssauge, Gottfried)
Date: Tue, 12 Sep 2006 13:37:07 +0200
Subject: [C++-sig] Py++ - problem with split_module
	andreturn_opaque_pointer-policy
In-Reply-To: <7465b6170609120414r3935fd99mf0e957ad722f9542@mail.gmail.com>
Message-ID: <51F4AD92DB06E9488BBC39B7B492409E029FDCA8@vg20ex1.grp.haufemg.com>

> -----Original Message-----
> From: c++-sig-bounces at python.org 
> [mailto:c++-sig-bounces at python.org] On Behalf Of Roman Yakovenko
> Sent: Tuesday, September 12, 2006 1:15 PM
> To: Development of Python/C++ integration
> Subject: Re: [C++-sig] Py++ - problem with split_module 
> andreturn_opaque_pointer-policy
> 
> On 9/12/06, Ganssauge, Gottfried <Gottfried.Ganssauge at haufe.de> wrote:
> > The return_value_policy<return_opaque_pointer> requires an 
> additional 
> > macro call (BOOST_PYTHON_OPAQUE_SPECIALIZED_TYPE_ID)
> > in the declaration section of a module.
> > This can be added e.g. using 
> module_builder_t.add_declaration_code(),
> > and it works as long as you're generating the code using 
> write_module().
> >
> > In the moment you're using "split_module" to split your 
> output, this 
> > scheme breaks, because BOOST_PYTHON_OPAQUE_SPECIALIZED_TYPE_ID is 
> > required in every translation unit using the opaque pointer 
> type whose 
> > type_id is defined with that macro.
> > Instead it is only generated into the main module.
> >
> 
> Does "every translation unit" means "in every cpp file" ?
Yes.

> If not the quick work-around I can propose is to put this macro in
> some header file
> and to include it.
:-(

> 
> The solution to your problem requires small refactoring in Py++. I
> will do that before next
> release, but it will take some time.
> The solution will be to add "add_declaration_code" to free function:
> 
> mb = module_builder.module_builder_t( ["test.hpp"] )
> 
> get_opaque = mb.free_function (name = "get_opaque")
> 
> get_opaque.add_declaration_code("BOOST_PYTHON_OPAQUE_SPECIALIZ
> ED_TYPE_ID(opaque_)")
> get_opaque.call_policies = return_value_policy (return_opaque_pointer)
> 
> What do you think?
Sounds not too bad, but how about member functions returning opaque
pointers?

What do you think about a code generator more "knowledgeable" about
return_value_policy?
One that could see "hey, I'm being ordered to return_opaque_pointer" and
then would call that macro wherever neccessary?
In the code generator it could go like:
  if calling_policy is return_value_policy(return_opaque_pointer):
    opaque_base = remove_pointer (current_decl.return_value.type)
    add_declaration_code ('BOOST_PYTHON_OPAQUE_SPECIALIZED_TYPE_ID(%s)'
% opaque_base)

> 
> -- 
> Roman Yakovenko
> C++ Python language binding
> http://www.language-binding.net/
> _______________________________________________
> C++-sig mailing list
> C++-sig at python.org
> http://mail.python.org/mailman/listinfo/c++-sig
> 

Cheers,

Gottfried


From roman.yakovenko at gmail.com  Tue Sep 12 13:43:59 2006
From: roman.yakovenko at gmail.com (Roman Yakovenko)
Date: Tue, 12 Sep 2006 14:43:59 +0300
Subject: [C++-sig] Py++ - problem with split_module
	andreturn_opaque_pointer-policy
In-Reply-To: <51F4AD92DB06E9488BBC39B7B492409E029FDCA8@vg20ex1.grp.haufemg.com>
References: <7465b6170609120414r3935fd99mf0e957ad722f9542@mail.gmail.com>
	<51F4AD92DB06E9488BBC39B7B492409E029FDCA8@vg20ex1.grp.haufemg.com>
Message-ID: <7465b6170609120443s7ef51998p425e7b70ada260f9@mail.gmail.com>

On 9/12/06, Ganssauge, Gottfried <Gottfried.Ganssauge at haufe.de> wrote:
> What do you think about a code generator more "knowledgeable" about
> return_value_policy?
> One that could see "hey, I'm being ordered to return_opaque_pointer" and
> then would call that macro wherever neccessary?
> In the code generator it could go like:
>   if calling_policy is return_value_policy(return_opaque_pointer):
>     opaque_base = remove_pointer (current_decl.return_value.type)
>     add_declaration_code ('BOOST_PYTHON_OPAQUE_SPECIALIZED_TYPE_ID(%s)'
> % opaque_base)

This is even better solution. I will implement it and will let you know.

-- 
Roman Yakovenko
C++ Python language binding
http://www.language-binding.net/


From dave at boost-consulting.com  Tue Sep 12 14:17:59 2006
From: dave at boost-consulting.com (David Abrahams)
Date: Tue, 12 Sep 2006 08:17:59 -0400
Subject: [C++-sig] [tutorial] Auto-Overloading example bug?
References: <7465b6170608300429n564516bfi22c665cd6cbfca32@mail.gmail.com>
	<44F666B8.9080609@boost-consulting.com>
	<44F67B9F.5060802@boost-consulting.com>
	<7465b6170608311021y2f5bebecr90ab1766c5b84ad2@mail.gmail.com>
	<44F77280.80103@boost-consulting.com>
	<87wt89oj73.fsf@pereiro.luannocracy.com>
	<7465b6170609112203h69b9484bw63bca80b48a783bb@mail.gmail.com>
Message-ID: <87odtlcm88.fsf@pereiro.luannocracy.com>

"Roman Yakovenko" <roman.yakovenko at gmail.com> writes:

> Actually my idea was simple: foo_overloads class derives from
> boost::python::detail::overloads_base class. I think you can use
> this fact and to remove the need to pass function signature to the
> "def".

How?  We need to be able to access the argument types at compile-time.
Anyway, if you think it can be done, please show us with actual
working code, because I'm pretty sure it's impossible.

-- 
Dave Abrahams
Boost Consulting
www.boost-consulting.com



From dave at boost-consulting.com  Tue Sep 12 14:20:12 2006
From: dave at boost-consulting.com (David Abrahams)
Date: Tue, 12 Sep 2006 08:20:12 -0400
Subject: [C++-sig] New Patch for opaque_pointer_converter
References: <87zmd5q3nx.fsf@pereiro.luannocracy.com>
	<51F4AD92DB06E9488BBC39B7B492409E027C2016@vg20ex1.grp.haufemg.com>
Message-ID: <87irjtcm4j.fsf@pereiro.luannocracy.com>

"Ganssauge, Gottfried" <Gottfried.Ganssauge at haufe.de> writes:

>> -----Original Message-----
>> From: c++-sig-bounces at python.org 
>> [mailto:c++-sig-bounces at python.org] On Behalf Of David Abrahams
>> Sent: Tuesday, September 12, 2006 3:23 AM
>> To: c++-sig at python.org
>> Subject: Re: [C++-sig] New Patch for opaque_pointer_converter
>> 
>> Gottfried Gan?auge <ganssauge at gmx.de> writes:
>> 
>> > Please find attached a completely new patch against the 
>> current CVS Head.
>> > This incorporates my patch from earlier today into the CVS Head.
>> > Further I implemented cross module support in so far as the 
>> converter 
>> > registers itself only if not another module already has 
>> registered the 
>> > same converter before.
>> > The opaque test case has been update to check for the segmentation 
>> > fault mentioned in my previous post.
>> > A test case for the cross module support is included 
>> (crossmod_opaque).
>> >
>> > I also updated the documentation for opaque.
>> >
>> > I tested my patch with
>> >  - GCC-4.1 running on GNU/Linux 2.6.16.21-0.13-default x86_64
>> >  - Visual-C++-6.5 running on Windows XP
>> >  - Visual-C++-7.1 running on Windows XP
>> >
>> 
>> Feel free to check it in.  If it's just a bug fix, you can 
>> merge it into the release branch as well.
> I would ... if I could :-)
> I'm not an "authorized" boost developer (I'm _not_ planning to become one) so sourceforge won't let me in.

Why not?

Anyway, I'm under a bit of pressure and I don't have time to merge
your patches right now.  If you don't want CVS access yourself and you
can't find anyone else to merge them, please stick them in the patches
tracker at SF so they'll be there when I _do_ find the time.

Thanks,

-- 
Dave Abrahams
Boost Consulting
www.boost-consulting.com



From seefeld at sympatico.ca  Tue Sep 12 15:24:00 2006
From: seefeld at sympatico.ca (Stefan Seefeld)
Date: Tue, 12 Sep 2006 09:24:00 -0400
Subject: [C++-sig] PyCode_New and new.code() functions
In-Reply-To: <57ffdb9c0609120150m3fc961fbg6c78df9c63370087@mail.gmail.com>
References: <57ffdb9c0609110356w609c73d3ic969cc28bcfe5dde@mail.gmail.com>	<45054B4B.1000303@sympatico.ca>	<57ffdb9c0609110524w53fc9337ya212489ab859bda4@mail.gmail.com>	<4505637E.5000706@sympatico.ca>	<57ffdb9c0609110717u4c7650e1m9932a033742f05d@mail.gmail.com>	<45060B2B.1000703@sympatico.ca>
	<57ffdb9c0609120150m3fc961fbg6c78df9c63370087@mail.gmail.com>
Message-ID: <4506B4F0.3010907@sympatico.ca>

anders langlands wrote:
> Sorry, I assumed it was clear since I know exactly what I mean!
> 
> All the derived types do indeed define many additional methods that I want
> to use.

While in python you can construct types at runtime (you don't even need the
python parser for that !), in C++ you can't. All your C++ API reflection has
to happen at compile-time. You may then rewrap the results however you want
from inside a running python instance.

> By type() constructor, do you mean doing:

I mean something like

----
from base import Base
Derived = type('Derived', (Base,), {})
----

which lets you create a new type 'Derived' in a functional way.

Regards,
		Stefan

-- 

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


From judog at campus.ie  Tue Sep 12 16:27:49 2006
From: judog at campus.ie (Gordon Blackshields)
Date: Tue, 12 Sep 2006 22:27:49 +0800
Subject: [C++-sig] ImportError for boost.python tutorial
Message-ID: <20060912142749.C02D7245E5@ws5-3.us4.outblaze.com>

Hi everyone, I found this post made back in march in an attempt to get started with boost.python using mingw, after unsuccessfully trying the seemingly very simple example given in the tutorial ("hello.cpp").

After following the advice given in the link (creating the libpython24.a file and so on using pexports and dlltool and so on) but  I'm still hitting a brick wall when trying to get the module imported in the interpreter. ("ImportError: dynamic module does not define init function (inithello)")

I have the .dll and .pyd files made ok, but pexports shows that hello.pyd isn't exporting inithello as expected. Are there other underlying causes for this import error that i'm just not seeing?

The original post that I found is below (the post was cunningly entitled "Problem")

Thanks in advance,
Gordon


Hi,

I am not sure this will solve your problem but just in case...
I also had similar problems at the beginning.  The problem was that the 
python library shipped
with Python installation only works with Microsoft compatible link and I 
was using, as you do, mingw (gcc).

In  http://www.boost.org/libs/python/doc/building.html#mingw you will 
find a detail explanation of what happens
and how to solve it. It's really simple.

Hope this will help...  It is really frustrating at the beginning but 
once it starts working... boost::python is wonderful!
Best Regards,
Pedro.

> Hi,
> I'm trying to build the basic boost.python tutorial that comes with the
> distribution.  I'm using bjam and mingw.  It builds the module, but
> upon import, it gives this error:
>
> ImportError: dynamic module does not define init function (inithello)
>
> I've found other threads with this problem as a topic, but no
> solutions.  Somebody has to have solved this....
>
> Any ideas?
>
> Thanks,
> -Justin
>
> OH yeah, I'm using Active Python (perhaps that's it?)
-- 
Pedro




-- 
_______________________________________________
For the largest FREE email in Ireland (25MB) and 20MB of online file storage space - Visit http://www.campus.ie


From jack833 at 163.com  Tue Sep 12 18:18:52 2006
From: jack833 at 163.com (jack833)
Date: Tue, 12 Sep 2006 09:18:52 -0700 (PDT)
Subject: [C++-sig] Help about warning C4541 in using VC7.1
Message-ID: <6269389.post@talk.nabble.com>


Here is My code . but show out warnning C4541 in VC7.1
namespace Unitest
{
	namespace UTester
	{
		class UTester
		{
		public:
			UTester(){}
			virtual	~UTester(){}
			std::string Foo(){return "FOO";}
		};
	}
}

warnning show:
E:\ProgramFiles\Microsoft Visual Studio.NET
2003\Vc7\Boost\include\boost\python\object\inheritance.hpp(43) : warning
C4541: ?dynamic_cast?using option  /GR-  at a Polymorphism 
?Unitest::UTester::UTester??May be result in unpredictable behavior.
        E:\ProgramFiles\Microsoft Visual Studio.NET
2003\Vc7\Boost\include\boost\python\object\inheritance.hpp(41) : while
compile template class member function?boost::python::objects::dynamic_id_t
boost::python::objects::polymorphic_id_generator<T>::execute(void *)?
        with
        [
            T=Unitest::UTester::UTester
        ]
 
can any one help me?
-- 
View this message in context: http://www.nabble.com/Help-about-warning-C4541-in-using-VC7.1-tf2259871.html#a6269389
Sent from the Python - c++-sig forum at Nabble.com.



From amohr at pixar.com  Tue Sep 12 18:36:06 2006
From: amohr at pixar.com (Alex Mohr)
Date: Tue, 12 Sep 2006 09:36:06 -0700
Subject: [C++-sig] What if a Python override returns several values?
In-Reply-To: <87k649q33b.fsf@pereiro.luannocracy.com>
References: <edhhnt$khq$1@sea.gmane.org> <44FDABA2.60809@pixar.com>
	<87k649q33b.fsf@pereiro.luannocracy.com>
Message-ID: <4506E1F6.5080604@pixar.com>

>>> But if I have another virtual member function "void g(int& a, int &b)" 
>>> which returns two integers, then I'd like the Python version to take no 
>>> arguments and return a tuple (a,b). But how does the wrapper look like 
>>> in this case? I would have expected something like:
>>>
>>>      void g(int& a, int& b)
>>>      {
>>>          boost::python::tuple tup = this->get_override("g")();
>>>
>>> 	a = ...extract an int from tup[0]...
>>>          b = ...extract an int from tup[1]...
>>>      }
>>>
>>> But when I run this code I get the following error at runtime:
>>>
>>> TypeError: No registered converter was able to extract a C++ reference 
>>> to type boost::python::tuple from this Python object of type tuple
>> Not sure exactly what's going on here, but you might be able to find out 
>> by looking at override.hpp.  Look at method_result -- that's what you're 
>> getting back from get_override.
>>
>> Off the top of my head, I think this would probably work for you:
>>
>> tuple tup = call<tuple>(this->get_override("g").ptr());
> 
>> a = ...extract an int from tup[0]
>> b = ...extract an int from tup[1]
> 
> 
> I don't know why people always use ugly low-level interfaces when they
> can use beautiful high-level ones:
> 
>     python::tuple tup = this->get_override("g")();
>     a = extract<t1>(tup[0]);
>     .
>     .
>     .

Isn't this what the original poster had, and claimed didn't work?  I 
agree that this is what should be written.  I was merely suggesting a 
workaround.  What am I missing?

> There's almost never a reason to use call<> anymore!  I should
> deprecate it.

Please don't -- I use it in some low-level code.

> I'm not trying to be critical, but this really concerns me.  I see
> this sort of thing all the time.  If you could help me understand
> why you were moved to suggest doing it the hard way, I would
> appreciate it.

Again, just because the original (which as far as I can see is the same 
as what you suggested) didn't work.  I was just suggesting a workaround. 
  I wouldn't suggest it as the "correct" solution.

Alex


From seefeld at sympatico.ca  Tue Sep 12 18:50:13 2006
From: seefeld at sympatico.ca (Stefan Seefeld)
Date: Tue, 12 Sep 2006 12:50:13 -0400
Subject: [C++-sig] Help about warning C4541 in using VC7.1
In-Reply-To: <6269389.post@talk.nabble.com>
References: <6269389.post@talk.nabble.com>
Message-ID: <4506E545.9040901@sympatico.ca>

jack833 wrote:
> Here is My code . but show out warnning C4541 in VC7.1
> namespace Unitest
> {
> 	namespace UTester
> 	{
> 		class UTester
> 		{
> 		public:
> 			UTester(){}
> 			virtual	~UTester(){}
> 			std::string Foo(){return "FOO";}
> 		};
> 	}
> }

This code doesn't appear to have anything to do with boost.python,
and neither with the error message from VS for that matter.

> warnning show:
> E:\ProgramFiles\Microsoft Visual Studio.NET
> 2003\Vc7\Boost\include\boost\python\object\inheritance.hpp(43) : warning
> C4541: ?dynamic_cast?using option  /GR-  at a Polymorphism 
> ?Unitest::UTester::UTester??May be result in unpredictable behavior.
>         E:\ProgramFiles\Microsoft Visual Studio.NET
> 2003\Vc7\Boost\include\boost\python\object\inheritance.hpp(41) : while
> compile template class member function?boost::python::objects::dynamic_id_t
> boost::python::objects::polymorphic_id_generator<T>::execute(void *)?
>         with
>         [
>             T=Unitest::UTester::UTester
>         ]
>  
> can any one help me?

Well, enable /GR (and /GX, while we are at it, since boost.python may throw
exceptions, too) !

Regards,
		Stefan

-- 

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


From Gottfried.Ganssauge at haufe.de  Tue Sep 12 19:09:53 2006
From: Gottfried.Ganssauge at haufe.de (Ganssauge, Gottfried)
Date: Tue, 12 Sep 2006 19:09:53 +0200
Subject: [C++-sig] New Patch for opaque_pointer_converter
In-Reply-To: <87irjtcm4j.fsf@pereiro.luannocracy.com>
Message-ID: <51F4AD92DB06E9488BBC39B7B492409E029FDD0D@vg20ex1.grp.haufemg.com>

> > I'm not an "authorized" boost developer (I'm _not_ planning 
> to become one) so sourceforge won't let me in.
> 
> Why not?
Because I couldn't easyly find out, how :-)
In earnest - I already got quite a few responsibilities and I don't take
things like that easy.

> 
> Anyway, I'm under a bit of pressure and I don't have time to 
> merge your patches right now.  If you don't want CVS access 
> yourself and you can't find anyone else to merge them, please 
> stick them in the patches tracker at SF so they'll be there 
> when I _do_ find the time.
OK.
Just put it in under ID boost-Patches-1557285.

> 
> Thanks,
Cheers,

Gottfried

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


From dave at boost-consulting.com  Tue Sep 12 19:36:34 2006
From: dave at boost-consulting.com (David Abrahams)
Date: Tue, 12 Sep 2006 13:36:34 -0400
Subject: [C++-sig] New Patch for opaque_pointer_converter
References: <87irjtcm4j.fsf@pereiro.luannocracy.com>
	<51F4AD92DB06E9488BBC39B7B492409E029FDD0D@vg20ex1.grp.haufemg.com>
Message-ID: <u8xkpyokd.fsf@boost-consulting.com>

"Ganssauge, Gottfried" <Gottfried.Ganssauge at haufe.de> writes:

>> > I'm not an "authorized" boost developer (I'm _not_ planning 
>> to become one) so sourceforge won't let me in.
>> 
>> Why not?
> Because I couldn't easyly find out, how :-)

You send your SourceForge userid to boost-owner at lists.boost.org and
ask.  Signing up at SourceForge is fairly trivial if you haven't done
it.

> In earnest - I already got quite a few responsibilities and I don't take
> things like that easy.
>
>> 
>> Anyway, I'm under a bit of pressure and I don't have time to 
>> merge your patches right now.  If you don't want CVS access 
>> yourself and you can't find anyone else to merge them, please 
>> stick them in the patches tracker at SF so they'll be there 
>> when I _do_ find the time.
> OK.
> Just put it in under ID boost-Patches-1557285.

Thanks, I appreciate it.

-- 
Dave Abrahams
Boost Consulting
www.boost-consulting.com



From dave at boost-consulting.com  Tue Sep 12 21:51:42 2006
From: dave at boost-consulting.com (David Abrahams)
Date: Tue, 12 Sep 2006 15:51:42 -0400
Subject: [C++-sig] What if a Python override returns several values?
References: <edhhnt$khq$1@sea.gmane.org> <44FDABA2.60809@pixar.com>
	<87k649q33b.fsf@pereiro.luannocracy.com> <4506E1F6.5080604@pixar.com>
Message-ID: <u3bawzwvl.fsf@boost-consulting.com>

Alex Mohr <amohr at pixar.com> writes:

>>>> But if I have another virtual member function "void g(int& a, int &b)" 
>>>> which returns two integers, then I'd like the Python version to take no 
>>>> arguments and return a tuple (a,b). But how does the wrapper look like 
>>>> in this case? I would have expected something like:
>>>>
>>>>      void g(int& a, int& b)
>>>>      {
>>>>          boost::python::tuple tup = this->get_override("g")();
>>>>
>>>> 	a = ...extract an int from tup[0]...
>>>>          b = ...extract an int from tup[1]...
>>>>      }
>>>>
>>>> But when I run this code I get the following error at runtime:
>>>>
>>>> TypeError: No registered converter was able to extract a C++ reference 
>>>> to type boost::python::tuple from this Python object of type tuple
>>> Not sure exactly what's going on here, but you might be able to find out 
>>> by looking at override.hpp.  Look at method_result -- that's what you're 
>>> getting back from get_override.
>>>
>>> Off the top of my head, I think this would probably work for you:
>>>
>>> tuple tup = call<tuple>(this->get_override("g").ptr());
>> 
>>> a = ...extract an int from tup[0]
>>> b = ...extract an int from tup[1]
>> 
>> 
>> I don't know why people always use ugly low-level interfaces when they
>> can use beautiful high-level ones:
>> 
>>     python::tuple tup = this->get_override("g")();
>>     a = extract<t1>(tup[0]);
>>     .
>>     .
>>     .
>
> Isn't this what the original poster had, and claimed didn't work?  

Nope.  

> I agree that this is what should be written.  I was merely
> suggesting a workaround.  What am I missing?

boost::tuple != boost::python::tuple.

>> There's almost never a reason to use call<> anymore!  I should
>> deprecate it.
>
> Please don't -- I use it in some low-level code.

Do you actually need it?

>> I'm not trying to be critical, but this really concerns me.  I see
>> this sort of thing all the time.  If you could help me understand
>> why you were moved to suggest doing it the hard way, I would
>> appreciate it.
>
> Again, just because the original (which as far as I can see is the same 
> as what you suggested) didn't work.  I was just suggesting a workaround. 
>   I wouldn't suggest it as the "correct" solution.

OK, thanks for the explanation.

-- 
Dave Abrahams
Boost Consulting
www.boost-consulting.com



From amohr at pixar.com  Tue Sep 12 22:32:22 2006
From: amohr at pixar.com (Alex Mohr)
Date: Tue, 12 Sep 2006 13:32:22 -0700
Subject: [C++-sig] What if a Python override returns several values?
In-Reply-To: <u3bawzwvl.fsf@boost-consulting.com>
References: <edhhnt$khq$1@sea.gmane.org> <44FDABA2.60809@pixar.com>
	<87k649q33b.fsf@pereiro.luannocracy.com>
	<4506E1F6.5080604@pixar.com> <u3bawzwvl.fsf@boost-consulting.com>
Message-ID: <45071956.2020104@pixar.com>

David Abrahams wrote:
>>>>> But if I have another virtual member function "void g(int& a, int &b)" 
>>>>> which returns two integers, then I'd like the Python version to take no 
>>>>> arguments and return a tuple (a,b). But how does the wrapper look like 
>>>>> in this case? I would have expected something like:
>>>>>
>>>>>      void g(int& a, int& b)
>>>>>      {
>>>>>          boost::python::tuple tup = this->get_override("g")();
>>>>>
>>>>> 	a = ...extract an int from tup[0]...
>>>>>          b = ...extract an int from tup[1]...
>>>>>      }
>>>>>
>>>>> But when I run this code I get the following error at runtime:
>>>>>
>>>>> TypeError: No registered converter was able to extract a C++ reference 
>>>>> to type boost::python::tuple from this Python object of type tuple
>>>> Not sure exactly what's going on here, but you might be able to find out 
>>>> by looking at override.hpp.  Look at method_result -- that's what you're 
>>>> getting back from get_override.
>>>>
>>>> Off the top of my head, I think this would probably work for you:
>>>>
>>>> tuple tup = call<tuple>(this->get_override("g").ptr());
>>>> a = ...extract an int from tup[0]
>>>> b = ...extract an int from tup[1]
>>>
>>> I don't know why people always use ugly low-level interfaces when they
>>> can use beautiful high-level ones:
>>>
>>>     python::tuple tup = this->get_override("g")();
>>>     a = extract<t1>(tup[0]);
>>>     .
>>>     .
>>>     .
>> Isn't this what the original poster had, and claimed didn't work?  
> 
> Nope.  
> 
>> I agree that this is what should be written.  I was merely
>> suggesting a workaround.  What am I missing?
> 
> boost::tuple != boost::python::tuple.
> 

Clearly that's a true statement, but I don't see where boost::tuple is 
coming from.  My eyes must be broken.  I'll copy-paste the original 
code, quoted earlier in this message.

void g(int& a, int& b) {
     boost::python::tuple tup = this->get_override("g")();
     a = ...extract an int from tup[0]...
     b = ...extract an int from tup[1]...
}

Compare this with the code you posted:

 >>     python::tuple tup = this->get_override("g")();
 >>     a = extract<t1>(tup[0]);

The poster says the error message he gets is:

TypeError: No registered converter was able to extract a C++ reference
to type boost::python::tuple from this Python object of type tuple

Where is boost::tuple coming into play?  I didn't see that at all 
anywhere in the thread.  In my reply to the poster, I removed the 
boost::python:: prefix from my code, so I mentioned "tuple" directly. 
Perhaps that's what was causing confusion?  I meant boost::python::tuple.

Regardless, unless I'm reading it wrong (please show me how!) I believe 
that the code you posted is identical to the original posted code.

>>> There's almost never a reason to use call<> anymore!  I should
>>> deprecate it.
>> Please don't -- I use it in some low-level code.
> 
> Do you actually need it?

I think so.  It's in some template code, where the return type is a 
template parameter, and is allowed to be void, and IIRC, the problem was 
that saying "return extract<void>(callable(...));" doesn't work. 
Whereas saying "return call<void>(callable.ptr(), ...));" does.  I guess 
I'd prefer to keep call<> around so I don't have to specialize the whole 
thing for the case of void.

I can try reworking it to avoid using call later in the week and get 
back to you.

Alex


From roman.yakovenko at gmail.com  Tue Sep 12 22:55:54 2006
From: roman.yakovenko at gmail.com (Roman Yakovenko)
Date: Tue, 12 Sep 2006 23:55:54 +0300
Subject: [C++-sig] custom smart pointers
In-Reply-To: <878xkppxyq.fsf@pereiro.luannocracy.com>
References: <7465b6170608291048y6e2010d4ofa84fbd5760da4d3@mail.gmail.com>
	<uac5lvczx.fsf@boost-consulting.com>
	<7465b6170608311358s1c02c6c9hc26222fae8cfdef6@mail.gmail.com>
	<878xkppxyq.fsf@pereiro.luannocracy.com>
Message-ID: <7465b6170609121355q42dc2dc7m526e6eaf5c3344ff@mail.gmail.com>

On 9/12/06, David Abrahams <dave at boost-consulting.com> wrote:
> I made your example work.  I had to add an operator* to your smart
> pointer, although I think that uncovers a bug in Boost.Python: we
> should be using get_pointer where it needed the operator*, instead.
>
> I don't know if this meets your needs or not, but the result is
> enclosed.

Thank you for your time and help.

Unfortunately it did not help :-(.
I reworked the example, base on your comments.

I had to add next conversion:
bp::implicitly_convertible< smart_ptr_t< base_wrapper_t >,
smart_ptr_t< base_i > >();

There are still 2 problems: I am not able to call functions that take
reference and const reference to smart_ptr_t< base > instance.

int ref_get_value( smart_ptr_t<base_i>& a ){
    return a->get_value();
}

I am unable to call this function, I've got ArgumentError exception

int const_ref_get_value( const smart_ptr_t<base_i>& a ){
    return a->get_value();
}

I am unable to call this function, I've got "segmentation fault" on Linux.

I attached test.py file that checks all the functions.

-- 
Roman Yakovenko
C++ Python language binding
http://www.language-binding.net/
-------------- next part --------------
A non-text attachment was scrubbed...
Name: cspc.cpp
Type: text/x-c++src
Size: 5276 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/cplusplus-sig/attachments/20060912/d37f6483/attachment.cpp>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: test.py
Type: text/x-python
Size: 1298 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/cplusplus-sig/attachments/20060912/d37f6483/attachment.py>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: Jamroot
Type: application/octet-stream
Size: 380 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/cplusplus-sig/attachments/20060912/d37f6483/attachment.obj>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: boost-build.jam
Type: application/octet-stream
Size: 283 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/cplusplus-sig/attachments/20060912/d37f6483/attachment-0001.obj>

From roman.yakovenko at gmail.com  Wed Sep 13 13:22:46 2006
From: roman.yakovenko at gmail.com (Roman Yakovenko)
Date: Wed, 13 Sep 2006 14:22:46 +0300
Subject: [C++-sig] Py++ generated code ( was custom smart pointers )
Message-ID: <7465b6170609130422i4f54ac36se9aa220e56a767da@mail.gmail.com>

On 9/12/06, David Abrahams <dave at boost-consulting.com> wrote:
> A minimal example would have no "documentation" stringsno use of
> named arguments (bp::arg(...)) and no use of things that you get by
> default anyway, like default_call_policies.  It would not contain
> extraneous implicit conversions from int.  This is a problem with many
> code generators: they create an unreadable mess for anyone who wants
> to understand the result.  Part of the reason many people despise the
> MFC Wizards.  I suggest you fix Py++ so it doesn't add vacuous
> "documentation" and other needless cruft.

1. Documentation was my fault. By default Py++ does not generate it.
2. default call policies is not generated anymore
3. implicit conversion - well, Py++ tries to generate as "complete
code" as possible.
    This is done for few reasons:
        * educational reasons, new users to Boost.Python can learn a
lot from Py++
          generated code. Py++ even has gui:
          http://language-binding.net/pyplusplus/tutorials/pyplusplus_demo.png
       * to minimize size of driving script
       * to minimize Py++ learning curve
    This case is a little bit special, because a "bug" in C++.
Constructors in C++ should be
    explicit by default. The bottom line is that I turned off
generation of "implicit conversion"
    for constructors.

In general, Py++ generates very readable code.

-- 
Roman Yakovenko
C++ Python language binding
http://www.language-binding.net/


From dave at boost-consulting.com  Wed Sep 13 13:45:01 2006
From: dave at boost-consulting.com (David Abrahams)
Date: Wed, 13 Sep 2006 07:45:01 -0400
Subject: [C++-sig] ImportError for boost.python tutorial
References: <20060912142749.C02D7245E5@ws5-3.us4.outblaze.com>
Message-ID: <87ejugvvlu.fsf@pereiro.luannocracy.com>

"Gordon Blackshields" <judog at campus.ie> writes:

> Hi everyone, I found this post made back in march in an attempt to
> get started with boost.python using mingw, after unsuccessfully
> trying the seemingly very simple example given in the tutorial
> ("hello.cpp").
>
> After following the advice given in the link (creating the
> libpython24.a file and so on using pexports and dlltool and so on)
> but I'm still hitting a brick wall when trying to get the module
> imported in the interpreter. ("ImportError: dynamic module does not
> define init function (inithello)")

Does the module source contain

  BOOST_PYTHON_MODULE(hello)

??

When you preprocess that file, you should see the necessary
decorations on the declaration to make the symbol exported.  If you're
using Boost.Build, it should supply the right compiler command-line
arguments to build a shared object.

If you're not using Boost.Build, all bets are off.

-- 
Dave Abrahams
Boost Consulting
www.boost-consulting.com



From dave at boost-consulting.com  Wed Sep 13 13:53:45 2006
From: dave at boost-consulting.com (David Abrahams)
Date: Wed, 13 Sep 2006 07:53:45 -0400
Subject: [C++-sig] What if a Python override returns several values?
References: <edhhnt$khq$1@sea.gmane.org> <44FDABA2.60809@pixar.com>
	<87k649q33b.fsf@pereiro.luannocracy.com>
	<4506E1F6.5080604@pixar.com> <u3bawzwvl.fsf@boost-consulting.com>
	<45071956.2020104@pixar.com>
Message-ID: <878xkovv7a.fsf@pereiro.luannocracy.com>

Alex Mohr <amohr at pixar.com> writes:

> Clearly that's a true statement, but I don't see where boost::tuple is 
> coming from.  My eyes must be broken.  I'll copy-paste the original 
> code, quoted earlier in this message.
>
> void g(int& a, int& b) {
>      boost::python::tuple tup = this->get_override("g")();
>      a = ...extract an int from tup[0]...
>      b = ...extract an int from tup[1]...
> }
>
> Compare this with the code you posted:
>
>  >>     python::tuple tup = this->get_override("g")();
>  >>     a = extract<t1>(tup[0]);
>
> The poster says the error message he gets is:
>
> TypeError: No registered converter was able to extract a C++ reference
> to type boost::python::tuple from this Python object of type tuple
>
> Where is boost::tuple coming into play?  

Nowhere; I think it's my eyes that are broken, sorry!

> I didn't see that at all 
> anywhere in the thread.  In my reply to the poster, I removed the 
> boost::python:: prefix from my code, so I mentioned "tuple" directly. 
> Perhaps that's what was causing confusion?  I meant boost::python::tuple.
>
> Regardless, unless I'm reading it wrong (please show me how!) I believe 
> that the code you posted is identical to the original posted code.

Okay.  Does it work if you don't use copy-initialization?

   boost::python::tuple tup(this->get_override("g")());

?

Does this work?

   boost::python::object t = this->get_override("g")();
   boost::python::tuple tup(t);

what about:

   boost::python::tuple tup 
     = this->get_override("g")().as<boost::python::tuple>();

??

>>>> There's almost never a reason to use call<> anymore!  I should
>>>> deprecate it.
>>> Please don't -- I use it in some low-level code.
>> 
>> Do you actually need it?
>
> I think so.  It's in some template code, where the return type is a 
> template parameter, and is allowed to be void, and IIRC, the problem was 
> that saying "return extract<void>(callable(...));" doesn't work. 
> Whereas saying "return call<void>(callable.ptr(), ...));" does.  I guess 
> I'd prefer to keep call<> around so I don't have to specialize the whole 
> thing for the case of void.
>
> I can try reworking it to avoid using call later in the week and get 
> back to you.

Don't bother; I understand your problem.  You need a facility
something more like get_override and method_result, which the library
doesn't supply for your case.

-- 
Dave Abrahams
Boost Consulting
www.boost-consulting.com



From dave at boost-consulting.com  Wed Sep 13 14:03:16 2006
From: dave at boost-consulting.com (David Abrahams)
Date: Wed, 13 Sep 2006 08:03:16 -0400
Subject: [C++-sig] Py++ generated code ( was custom smart pointers )
References: <7465b6170609130422i4f54ac36se9aa220e56a767da@mail.gmail.com>
Message-ID: <87zmd4ug6z.fsf@pereiro.luannocracy.com>

"Roman Yakovenko" <roman.yakovenko at gmail.com> writes:

> On 9/12/06, David Abrahams <dave at boost-consulting.com> wrote:

> 3. implicit conversion - well, Py++ tries to generate as "complete
> code" as possible.

I'm not criticizing that part of what Py++ does; it's probably the
right thing, but it shouldn't show up in a minimal example for tech
support purposes (unless it's relevant, of course)

> In general, Py++ generates very readable code.

It's not bad; it would probably help if you did something to control
the length of generated lines.

-- 
Dave Abrahams
Boost Consulting
www.boost-consulting.com



From dave at boost-consulting.com  Wed Sep 13 14:24:45 2006
From: dave at boost-consulting.com (David Abrahams)
Date: Wed, 13 Sep 2006 08:24:45 -0400
Subject: [C++-sig] custom smart pointers
References: <7465b6170608291048y6e2010d4ofa84fbd5760da4d3@mail.gmail.com>
	<uac5lvczx.fsf@boost-consulting.com>
	<7465b6170608311358s1c02c6c9hc26222fae8cfdef6@mail.gmail.com>
	<878xkppxyq.fsf@pereiro.luannocracy.com>
	<7465b6170609121355q42dc2dc7m526e6eaf5c3344ff@mail.gmail.com>
Message-ID: <87wt88uf76.fsf@pereiro.luannocracy.com>

"Roman Yakovenko" <roman.yakovenko at gmail.com> writes:

> On 9/12/06, David Abrahams <dave at boost-consulting.com> wrote:
>> I made your example work.  I had to add an operator* to your smart
>> pointer, although I think that uncovers a bug in Boost.Python: we
>> should be using get_pointer where it needed the operator*, instead.
>>
>> I don't know if this meets your needs or not, but the result is
>> enclosed.
>
> Thank you for your time and help.
>
> Unfortunately it did not help :-(.
> I reworked the example, base on your comments.
>
> I had to add next conversion:
> bp::implicitly_convertible< smart_ptr_t< base_wrapper_t >,
> smart_ptr_t< base_i > >();
>
> There are still 2 problems: I am not able to call functions that take
> reference and const reference to smart_ptr_t< base > instance.

FWIW, reference and const reference are totally separate cases.  const
reference is essentially the same as pass-by-value; it only requires
an rvalue conversion.  Mutable references require an lvalue conversion.

> int ref_get_value( smart_ptr_t<base_i>& a ){
>    return a->get_value();
> }
>
> I am unable to call this function, I've got ArgumentError exception

Naturally; there is no instance of smart_ptr_t<base_i> anywhere in the
Python object for the reference to bind to.   The rules are the same
as in C++: 

   int f(smart_ptr_t<base_i>& x) { return 0; }
   smart_ptr_t<base_wrapper_t> y;
   int z = f(y);                // fails to compile

> int const_ref_get_value( const smart_ptr_t<base_i>& a ){
>    return a->get_value();
> }
>
> I am unable to call this function, I've got "segmentation fault" on Linux.

Where does the segmentation fault occur?  Is 'a' a valid pointer
within const_ref_get_value?

> I attached test.py file that checks all the functions.

Try building a minimal example that does nothing more than to
reproduce your segmentation fault.  Then we'll see if the problem is
in Boost.Python or in your code.

-- 
Dave Abrahams
Boost Consulting
www.boost-consulting.com



From jack833 at 163.com  Wed Sep 13 14:44:46 2006
From: jack833 at 163.com (jack833)
Date: Wed, 13 Sep 2006 05:44:46 -0700 (PDT)
Subject: [C++-sig] Help about warning C4541 in using VC7.1
In-Reply-To: <4506E545.9040901@sympatico.ca>
References: <6269389.post@talk.nabble.com> <4506E545.9040901@sympatico.ca>
Message-ID: <6284711.post@talk.nabble.com>


Oh,I see,thank you for your help!

Stefan Seefeld wrote:
> 
> jack833 wrote:
>> Here is My code . but show out warnning C4541 in VC7.1
>> namespace Unitest
>> {
>> 	namespace UTester
>> 	{
>> 		class UTester
>> 		{
>> 		public:
>> 			UTester(){}
>> 			virtual	~UTester(){}
>> 			std::string Foo(){return "FOO";}
>> 		};
>> 	}
>> }
> 
> This code doesn't appear to have anything to do with boost.python,
> and neither with the error message from VS for that matter.
> 
>> warnning show:
>> E:\ProgramFiles\Microsoft Visual Studio.NET
>> 2003\Vc7\Boost\include\boost\python\object\inheritance.hpp(43) : warning
>> C4541: ?dynamic_cast?using option  /GR-  at a Polymorphism 
>> ?Unitest::UTester::UTester??May be result in unpredictable behavior.
>>         E:\ProgramFiles\Microsoft Visual Studio.NET
>> 2003\Vc7\Boost\include\boost\python\object\inheritance.hpp(41) : while
>> compile template class member
>> function?boost::python::objects::dynamic_id_t
>> boost::python::objects::polymorphic_id_generator<T>::execute(void *)?
>>         with
>>         [
>>             T=Unitest::UTester::UTester
>>         ]
>>  
>> can any one help me?
> 
> Well, enable /GR (and /GX, while we are at it, since boost.python may
> throw
> exceptions, too) !
> 
> Regards,
> 		Stefan
> 
> -- 
> 
>       ...ich hab' noch einen Koffer in Berlin...
> _______________________________________________
> C++-sig mailing list
> C++-sig at python.org
> http://mail.python.org/mailman/listinfo/c++-sig
> 
> 

-- 
View this message in context: http://www.nabble.com/Help-about-warning-C4541-in-using-VC7.1-tf2259871.html#a6284711
Sent from the Python - c++-sig forum at Nabble.com.



From eduardo.picado at gmail.com  Wed Sep 13 18:56:26 2006
From: eduardo.picado at gmail.com (Eduardo Picado)
Date: Wed, 13 Sep 2006 18:56:26 +0200
Subject: [C++-sig] Exception 'No to_python (by-value) converter found -
	'garbage collection'
Message-ID: <6c75ff8f0609130956s6b0e7c9m9fdb8d44840fd26b@mail.gmail.com>

Hi,

Im writing the following kind of python code using a C++ wrapped library:



system = TASRSystem.getASRSystem()
system.setParameter(TASRSystem.Option.asrServersListenPort, 32000)

try:
	system.initialize(1, 3000)
except Exception, e:
	sys.stderr.write("[E] : " +  e.__class__.__name__ + str(e) )
print "ok"



Running it I get:



ok
Exception exceptions.TypeError: 'No to_python (by-value) converter
found for C++ type:
class TASRSystem' in 'garbage collection' ignored
Fatal Python error: unexpected exception during garbage collection



The declaration of TASRSystem is like:



BOOST_PYTHON_MEMBER_FUNCTION_OVERLOADS(TASRSystem_initialize_overloads_0_2,
initialize, 0, 2)

class_< TASRSystem, boost::noncopyable >("TASRSystem", no_init)
	.def("getASRSystem", &TASRSystem::getASRSystem,
return_value_policy<reference_existing_object>())
	.def("initialize", &TASRSystem::initialize,
TASRSystem_initialize_overloads_0_2())
	.staticmethod("getASRSystem")


- TASRSystem.getASRSystem() is an static method, that creates and
returns an instance of an ASRSystem. Seems ok.
I've tried to use with_custodian_and_ward_postcall <1, 0>, to ensure
that the lifetime of the created instance is as long
as that of the Python object, but it didn't worked - gives "argument
index out of range".


- setParameter() seems good, as I can check with a .getParameter() call.

- system.initialize() It doesn't return anything, I think it just
tries to find/connect to remote ressources based on the parameters. It
seems to work, the "ok" is printed but after it gives the "No
to_python (by-value) converter found"... associated with the Python
garbage collection...


Some ideas?


Thanks very much,


Eduardo

-- 
Eduardo PICADO
MSN eduardo.picado at gmail.com
portable +33 6 32 47 10 10 (NEW)
fix +33 1 45 03 52 54
|??<?????>??|

From roman.yakovenko at gmail.com  Wed Sep 13 21:10:31 2006
From: roman.yakovenko at gmail.com (Roman Yakovenko)
Date: Wed, 13 Sep 2006 22:10:31 +0300
Subject: [C++-sig] custom smart pointers
In-Reply-To: <87wt88uf76.fsf@pereiro.luannocracy.com>
References: <7465b6170608291048y6e2010d4ofa84fbd5760da4d3@mail.gmail.com>
	<uac5lvczx.fsf@boost-consulting.com>
	<7465b6170608311358s1c02c6c9hc26222fae8cfdef6@mail.gmail.com>
	<878xkppxyq.fsf@pereiro.luannocracy.com>
	<7465b6170609121355q42dc2dc7m526e6eaf5c3344ff@mail.gmail.com>
	<87wt88uf76.fsf@pereiro.luannocracy.com>
Message-ID: <7465b6170609131210h30285ef8pd4c8b7e624ff82f3@mail.gmail.com>

On 9/13/06, David Abrahams <dave at boost-consulting.com> wrote:
> FWIW, reference and const reference are totally separate cases.  const
> reference is essentially the same as pass-by-value; it only requires
> an rvalue conversion.  Mutable references require an lvalue conversion.
>
> > int ref_get_value( smart_ptr_t<base_i>& a ){
> >    return a->get_value();
> > }
> >
> > I am unable to call this function, I've got ArgumentError exception
>
> Naturally; there is no instance of smart_ptr_t<base_i> anywhere in the
> Python object for the reference to bind to.   The rules are the same
> as in C++:
>
>    int f(smart_ptr_t<base_i>& x) { return 0; }
>    smart_ptr_t<base_wrapper_t> y;
>    int z = f(y);                // fails to compile

Thank you for explanation.

> > int const_ref_get_value( const smart_ptr_t<base_i>& a ){
> >    return a->get_value();
> > }
> >
> > I am unable to call this function, I've got "segmentation fault" on Linux.
>
> Where does the segmentation fault occur?  Is 'a' a valid pointer
> within const_ref_get_value?
>
>
> Try building a minimal example that does nothing more than to
> reproduce your segmentation fault.  Then we'll see if the problem is
> in Boost.Python or in your code.

I built one and I don't like what I see. Let me explain.

smart_ptr_t{...};

struct derived_t{
    virtual int get_value(){ return 11;}
};


int val_get_value( smart_ptr_t<derived_t> a ){

    return a->get_value();

}



int const_ref_get_value( const smart_ptr_t<derived_t>& a ){

    if( a.get() )

        return a->get_value();


    else

        return -1;


}


namespace boost{ namespace python{
    get_pointer definition
    pointee definition
} }

Registration is pretty simple:

    bp::class_< derived_t, smart_ptr_t<derived_t> >( "derived_t" )

        .def( "get_value", &::derived_t::get_value );



    bp::def( "const_ref_get_value", &::const_ref_get_value );

    bp::def( "val_get_value", &::val_get_value );


Python test code also simple:

inst = cspc_ext.derived_t()
cspc_ext.val_get_value(inst)
cspc_ext.const_ref_get_value(inst) # <= in this line I get segmentation fault.

Now the mistery I saw:
1. If I remove "virtual" from derived_t::get_value everything works as expected.
2. If I comment "cspc_ext.val_get_value(inst)" expression, than
    cspc_ext.const_ref_get_value(inst)
    works.

I attached the source code.
I am using compiler g++ 4.0.3 on Ubuntu, CVS version of boost.

-- 
Roman Yakovenko
C++ Python language binding
http://www.language-binding.net/
-------------- next part --------------
A non-text attachment was scrubbed...
Name: cspc.cpp
Type: text/x-c++src
Size: 2792 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/cplusplus-sig/attachments/20060913/aa66d950/attachment.cpp>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: test.py
Type: text/x-python
Size: 592 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/cplusplus-sig/attachments/20060913/aa66d950/attachment.py>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: Jamroot
Type: application/octet-stream
Size: 380 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/cplusplus-sig/attachments/20060913/aa66d950/attachment.obj>

From markus at relix.de  Thu Sep 14 04:39:19 2006
From: markus at relix.de (Markus Heller)
Date: Thu, 14 Sep 2006 04:39:19 +0200
Subject: [C++-sig] extract character from PyObject (c++)
Message-ID: <200609140439.20219.markus@relix.de>

Dear List,

I have this: a c++ PyObject which contains a unicode string. 

How can I isolate the individual unicode symbols from the string? Does anybody 
have sample code? 

Unfortunately I have only partly understood the definitions as given here:
http://docs.python.org/api/unicodeObjects.html

Thanks a lot! 

Markus





From shiqinfeng at gmail.com  Thu Sep 14 08:38:16 2006
From: shiqinfeng at gmail.com (Qinfeng(Javen) Shi )
Date: Thu, 14 Sep 2006 16:38:16 +1000
Subject: [C++-sig] how to wrap multi c++ files into python by pyste?
Message-ID: <fe76f74b0609132338v5f853669ge0a7738e579ae3f6@mail.gmail.com>

Dear all,

I know how to wrap a single c++ file into python by pyste? Does anybody how
to wrap many c++ files into several *.so?  We can't  wrap every single file
seperately because some types and functions used in it may be delared in
other file. And puting everything into a huge single file is a ugly thing.

Best wishes!

Javen

Research School of Information Sciences and Engineering
Australian National University
Locked Bag 8001
Canberra ACT 2601
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/cplusplus-sig/attachments/20060914/53bd2620/attachment.htm>

From roman.yakovenko at gmail.com  Thu Sep 14 08:55:37 2006
From: roman.yakovenko at gmail.com (Roman Yakovenko)
Date: Thu, 14 Sep 2006 09:55:37 +0300
Subject: [C++-sig] how to wrap multi c++ files into python by pyste?
In-Reply-To: <fe76f74b0609132338v5f853669ge0a7738e579ae3f6@mail.gmail.com>
References: <fe76f74b0609132338v5f853669ge0a7738e579ae3f6@mail.gmail.com>
Message-ID: <7465b6170609132355l2b6e50d6s99082dd8e03df5e3@mail.gmail.com>

On 9/14/06, Qinfeng(Javen) Shi <shiqinfeng at gmail.com> wrote:
> Dear all,
>
>  I know how to wrap a single c++ file into python by pyste? Does anybody how
> to wrap many c++ files into several *.so?

Take a look on this:
http://www.boost.org/libs/python/pyste/doc/running_pyste.html
It seems that you should combine "multiple" and "generate-main" options.

> We can't  wrap every single file
> seperately because some types and functions used in it may be delared in
> other file. And puting everything into a huge single file is a ugly thing.

I suggest you to look at Py++. Py++ is able to split generated code to
few files, more
over it is able to split generated code for single class to multiple
files. Take a look on
http://tinyurl.com/h33pn . Using Py++ you can generate code for a few
modules from the same Python script.


-- 
Roman Yakovenko
C++ Python language binding
http://www.language-binding.net/


From shiqinfeng at gmail.com  Thu Sep 14 09:31:01 2006
From: shiqinfeng at gmail.com (Qinfeng(Javen) Shi )
Date: Thu, 14 Sep 2006 17:31:01 +1000
Subject: [C++-sig] how to pass function pointer to a constructor of a
	template class when wrap it by pyste
Message-ID: <fe76f74b0609140031r2df83d32j73f93f4540c891e0@mail.gmail.com>

Dear all,

I have a class (in hello.cc file) whose constructor assgins a value to a
global function pointer. It is compiled by gcc and it works well. But it
always reports an error when I used bjam to build the hello.cpp file (which
is produced by pyste). If I remove the template, it works. But I do need the
template. Does anybody know how to wrap it correctly?

Any suggestion is appreciated.

The hello.cc, hello.pyste, Jamfile are as following:
/********************hello.cc*************/
#include <stdio.h>
float add(float a, float b)
{
    return a+b;
}
float minus(float a, float b)
{
    return a-b;
}
float (*op)(float a, float b) = add;

template<class P>
class cover_tree {
P value;
public:
    cover_tree(float (* metric)(float a, float b))
    {
        op= metric;

    }
    cover_tree()
    {
    }
};

int main()
{
    float a,b,c;
    a = 1;
    b = 2;
    cover_tree<float> ct2(minus);
    c = op(a,b);
    printf("%f\n",c);

}

/************************hello.pyste*********/
cover_tree = Template('cover_tree','hello.cc')
cover = cover_tree("float")
Function("main", "hello.cc")

/************************Jamfile***********/

# This is the top of our own project tree
project-root ;

#   Include definitions needed for Python modules
import python ;

extension hello        # Declare a Python extension called hello
:   hello.cpp
    # requirements and dependencies for Boost.Python extensions
    <template>@boost/libs/python/build/extension
    ;

-- 
Javen

Research School of Information Sciences and Engineering
Australian National University
Locked Bag 8001
Canberra ACT 2601
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/cplusplus-sig/attachments/20060914/7ea24d9a/attachment.htm>

From baas at ira.uka.de  Thu Sep 14 16:44:19 2006
From: baas at ira.uka.de (Matthias Baas)
Date: Thu, 14 Sep 2006 16:44:19 +0200
Subject: [C++-sig] What if a Python override returns several values?
In-Reply-To: <878xkovv7a.fsf@pereiro.luannocracy.com>
References: <edhhnt$khq$1@sea.gmane.org>
	<44FDABA2.60809@pixar.com>	<87k649q33b.fsf@pereiro.luannocracy.com>	<4506E1F6.5080604@pixar.com>
	<u3bawzwvl.fsf@boost-consulting.com>	<45071956.2020104@pixar.com>
	<878xkovv7a.fsf@pereiro.luannocracy.com>
Message-ID: <eebq0u$bvi$1@sea.gmane.org>

David Abrahams wrote:
>> void g(int& a, int& b) {
>>      boost::python::tuple tup = this->get_override("g")();
>>      a = ...extract an int from tup[0]...
>>      b = ...extract an int from tup[1]...
>> }
>>
>> [...]
>> TypeError: No registered converter was able to extract a C++ reference
>> to type boost::python::tuple from this Python object of type tuple
>[...]
> Okay.  Does it work if you don't use copy-initialization?

Now that you mention it I actually have to confess that I was not aware 
that this could make a difference. My original code rather looked like this:

   boost::python::tuple tup;
   tup = this->get_override("g")();

And this results in the above TypeError exception.
I only noticed now that the following code actually does work (sorry):

   boost::python::tuple tup = this->get_override("g")();

I don't suppose this difference is intentional, is it?
Unfortunately, it's not that trivial for me to get to the latter version 
as the code I'm using is not written manually but is generated by a tool 
which separates the declaration of variables from their usage.

>    boost::python::tuple tup(this->get_override("g")());

No, this results in the following compile error:

call of overloaded `tuple(boost::python::detail::method_result)' is 
ambiguous
candidates are:
boost::python::tuple::tuple(const boost::python::tuple&)
boost::python::tuple::tuple(boost::python::detail::new_non_null_reference_t*)
boost::python::tuple::tuple(boost::python::detail::new_reference_t*)
boost::python::tuple::tuple(boost::python::detail::borrowed_reference_t*)
boost::python::tuple::tuple(const T&) [with T = 
boost::python::detail::method_result]

>    boost::python::object t = this->get_override("g")();
>    boost::python::tuple tup(t);

Yes, this works (see above).

> what about:
> 
>    boost::python::tuple tup 
>      = this->get_override("g")().as<boost::python::tuple>();

Yes (what's this anyway? I didn't see that in the reference manual).


- Matthias -



From amohr at pixar.com  Thu Sep 14 18:49:43 2006
From: amohr at pixar.com (Alex Mohr)
Date: Thu, 14 Sep 2006 09:49:43 -0700
Subject: [C++-sig] What if a Python override returns several values?
In-Reply-To: <eebq0u$bvi$1@sea.gmane.org>
References: <edhhnt$khq$1@sea.gmane.org> <44FDABA2.60809@pixar.com>
	<87k649q33b.fsf@pereiro.luannocracy.com>
	<4506E1F6.5080604@pixar.com> <u3bawzwvl.fsf@boost-consulting.com>
	<45071956.2020104@pixar.com>
	<878xkovv7a.fsf@pereiro.luannocracy.com>
	<eebq0u$bvi$1@sea.gmane.org>
Message-ID: <45098827.9030101@pixar.com>

>> what about:
>>
>>    boost::python::tuple tup 
>>      = this->get_override("g")().as<boost::python::tuple>();
> 
> Yes (what's this anyway? I didn't see that in the reference manual).

It's not documented.  It appears to be a convenience.  It's a method 
that appears to be equivalent to override::method_result::operator T().

Alex


From noah.schwartz1 at gmail.com  Thu Sep 14 20:02:19 2006
From: noah.schwartz1 at gmail.com (Noah Schwartz)
Date: Thu, 14 Sep 2006 14:02:19 -0400
Subject: [C++-sig] Complete confustion about using Boost.Python
Message-ID: <200609141402.19397.noah.schwartz1@gmail.com>

Hi,

????????I've recently built Boost based on the instructions at 
http://boost.org/more/getting_started.html. It found my python installation 
OK and so as far as I know that means Boost.Python has been installed on my 
system.

????????The issue I am having is that I can't seem to figure out how to build 
any 
modules that are not within the directory tree of the source I extracted from 
sourceforge. If I copy the exact tutorial out of libs/python/example/tutorial 
and try to build it elsewhere, it complains about not being able to find a 
project-root.jam file. Also, I have noticed that if I set the BOOST_ROOT path 
in say, for example, the Jamrules file in the that tutorial directory to 
where boost is now actually installed (/usr/local/include/boost-1_33_1), I 
can no longer build. What am I missing here? Right now it seems like I can 
only build projects in the directory tree of the source I extracted from the 
website.

Thanks,
Noah


From eduardo.picado at gmail.com  Thu Sep 14 20:45:44 2006
From: eduardo.picado at gmail.com (Eduardo Picado)
Date: Thu, 14 Sep 2006 20:45:44 +0200
Subject: [C++-sig] Complete confustion about using Boost.Python
In-Reply-To: <200609141402.19397.noah.schwartz1@gmail.com>
References: <200609141402.19397.noah.schwartz1@gmail.com>
Message-ID: <6c75ff8f0609141145s524ebbc6p1c1c4194adf1df5e@mail.gmail.com>

On 9/14/06, Noah Schwartz <noah.schwartz1 at gmail.com> wrote:
> Hi,
>

Hi,

> I've recently built Boost based on the instructions at
> http://boost.org/more/getting_started.html. It found my python installation
> OK and so as far as I know that means Boost.Python has been installed on my
> system.
>
> The issue I am having is that I can't seem to figure out how to build
> any
> modules that are not within the directory tree of the source I extracted from
> sourceforge. If I copy the exact tutorial out of libs/python/example/tutorial
> and try to build it elsewhere, it complains about not being able to find a
> project-root.jam file. Also, I have noticed that if I set the BOOST_ROOT path
> in say, for example, the Jamrules file in the that tutorial directory to
> where boost is now actually installed (/usr/local/include/boost-1_33_1), I
> can no longer build. What am I missing here? Right now it seems like I can
> only build projects in the directory tree of the source I extracted from the
> website.
>

What error did you get?
Can you show the Jamfile and Jamrules you are using?


Eduardo

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


-- 
Eduardo PICADO
MSN eduardo.picado at gmail.com
portable +33 6 32 47 10 10 (NEW)
fix +33 1 45 03 52 54
|??<?????>??|

From noah.schwartz1 at gmail.com  Thu Sep 14 20:57:08 2006
From: noah.schwartz1 at gmail.com (Noah Schwartz)
Date: Thu, 14 Sep 2006 14:57:08 -0400
Subject: [C++-sig] Complete confustion about using Boost.Python
In-Reply-To: <6c75ff8f0609141145s524ebbc6p1c1c4194adf1df5e@mail.gmail.com>
References: <200609141402.19397.noah.schwartz1@gmail.com>
	<6c75ff8f0609141145s524ebbc6p1c1c4194adf1df5e@mail.gmail.com>
Message-ID: <200609141457.08362.noah.schwartz1@gmail.com>

Good idea :D

Boost is installed on my machine at /usr/local/include/boost-1_33_1 (this is 
the directory with the boost subdirectory)

The libs are installed on my machine at /usr/local/lib


This is the Jamfile that I am using:
#   Hello World Example from the tutorial
#   Copyright Joel de Guzman 2002-2004. Distributed under the Boost
#   Software License, Version 1.0. (See accompanying file LICENSE_1_0.txt
#   or copy at http://www.boost.org/LICENSE_1_0.txt)
#   [Joel de Guzman 10/9/2002]

# This is the top of our own project tree
project-root ;

#   Include definitions needed for Python modules
import python ;

extension hello                     # Declare a Python extension called hello
:   hello.cpp                       # source
    # requirements and dependencies for Boost.Python extensions
    <template>@boost/libs/python/build/extension
    ;


This is the Jamrules file that I am using:
# Copyright David Abrahams 2003. See accompanying LICENSE for terms
# and conditions of use.

# If you move this example from its place in the Boost tree, edit this
# path to point at the root directory of your Boost installation (the
# one containing a subdirectory called "boost/" and a sub-subdirectory
# "boost/python/" full of .hpp files).  Absolute paths work, too.
#
path-global BOOST_ROOT : /usr/local/include/boost-1_33_1 ;

# Boost.Python configuration variables, as described in
# http://www.boost.org/libs/python/doc/building.html#configuration.
# Usually you don't need to set these; the defaults will work.  If you
# do set them, try to change as few of them as possible, starting with
# the first ones.

  # PYTHON_VERSION = <the two-part Major.Minor version number, e.g. 2.2> ;
  # PYTHON_ROOT = <root directory of your Python installation, e.g. /usr> ;
  # PYTHON_INCLUDES = <path to Python #include directories> ;
  # PYTHON_LIB_PATH = <path to Python library object> ;

# You may need to configure your compiler toolset, especially if you
# want to build with a compiler that is not the "system default" or if
# it is installed in a nonstandard place; see
# http://www.boost.org/more/getting_started.html#Configuring for
# details.

# Makes a project id for boost so that other Boost.Build projects can
# refer to it by name.
#
project boost : $(BOOST_ROOT) ;

# Change this setting to have all your built products placed under a
# single directory:
#
# ALL_LOCATE_TARGET = <root directory for all built products>

Like I said, copied from the examples directory in the extracted source with 
the exception of the change to BOOST_ROOT.


On Thursday 14 September 2006 14:45, Eduardo Picado wrote:
> On 9/14/06, Noah Schwartz <noah.schwartz1 at gmail.com> wrote:
> > Hi,
>
> Hi,
>
> > I've recently built Boost based on the instructions at
> > http://boost.org/more/getting_started.html. It found my python
> > installation OK and so as far as I know that means Boost.Python has been
> > installed on my system.
> >
> > The issue I am having is that I can't seem to figure out how to build
> > any
> > modules that are not within the directory tree of the source I extracted
> > from sourceforge. If I copy the exact tutorial out of
> > libs/python/example/tutorial and try to build it elsewhere, it complains
> > about not being able to find a project-root.jam file. Also, I have
> > noticed that if I set the BOOST_ROOT path in say, for example, the
> > Jamrules file in the that tutorial directory to where boost is now
> > actually installed (/usr/local/include/boost-1_33_1), I can no longer
> > build. What am I missing here? Right now it seems like I can only build
> > projects in the directory tree of the source I extracted from the
> > website.
>
> What error did you get?
> Can you show the Jamfile and Jamrules you are using?
>
>
> Eduardo
>
> > Thanks,
> > Noah
> > _______________________________________________
> > C++-sig mailing list
> > C++-sig at python.org
> > http://mail.python.org/mailman/listinfo/c++-sig


From noah.schwartz1 at gmail.com  Thu Sep 14 20:57:59 2006
From: noah.schwartz1 at gmail.com (Noah Schwartz)
Date: Thu, 14 Sep 2006 14:57:59 -0400
Subject: [C++-sig] Complete confustion about using Boost.Python
In-Reply-To: <6c75ff8f0609141145s524ebbc6p1c1c4194adf1df5e@mail.gmail.com>
References: <200609141402.19397.noah.schwartz1@gmail.com>
	<6c75ff8f0609141145s524ebbc6p1c1c4194adf1df5e@mail.gmail.com>
Message-ID: <200609141457.59649.noah.schwartz1@gmail.com>

And of course I forgot to mention the error message :D

Failed to find the project root for directory '.'.
Did not find a project-root.jam file there or in any of its parent 
directories.
Please consult the documentation at 'http://www.boost.org'.

On Thursday 14 September 2006 14:45, Eduardo Picado wrote:
> On 9/14/06, Noah Schwartz <noah.schwartz1 at gmail.com> wrote:
> > Hi,
>
> Hi,
>
> > I've recently built Boost based on the instructions at
> > http://boost.org/more/getting_started.html. It found my python
> > installation OK and so as far as I know that means Boost.Python has been
> > installed on my system.
> >
> > The issue I am having is that I can't seem to figure out how to build
> > any
> > modules that are not within the directory tree of the source I extracted
> > from sourceforge. If I copy the exact tutorial out of
> > libs/python/example/tutorial and try to build it elsewhere, it complains
> > about not being able to find a project-root.jam file. Also, I have
> > noticed that if I set the BOOST_ROOT path in say, for example, the
> > Jamrules file in the that tutorial directory to where boost is now
> > actually installed (/usr/local/include/boost-1_33_1), I can no longer
> > build. What am I missing here? Right now it seems like I can only build
> > projects in the directory tree of the source I extracted from the
> > website.
>
> What error did you get?
> Can you show the Jamfile and Jamrules you are using?
>
>
> Eduardo
>
> > Thanks,
> > Noah
> > _______________________________________________
> > C++-sig mailing list
> > C++-sig at python.org
> > http://mail.python.org/mailman/listinfo/c++-sig


From investtcartier at yahoo.com  Thu Sep 14 20:59:01 2006
From: investtcartier at yahoo.com (Kevin Jones)
Date: Thu, 14 Sep 2006 11:59:01 -0700 (PDT)
Subject: [C++-sig] memory leaks using the "manage_new_object" return
	policy...
Message-ID: <20060914185901.20076.qmail@web55507.mail.re4.yahoo.com>

Hi,

I am using the manage_new_object return policy to wrap
"newed" C++ object in an effort to conserve memory.

Even though I am deleting the returned object in
Python and invoking the Python garbage collector I
still have a leak. This is a long-running systems
adminstration app the read the contents of a very
large database.

Any suggestions on how to make manage_new_object
behave correctly?

Thanks in advance,

Kevin


__________________________________________________
Do You Yahoo!?
Tired of spam?  Yahoo! Mail has the best spam protection around 
http://mail.yahoo.com 


From roman.yakovenko at gmail.com  Thu Sep 14 21:11:14 2006
From: roman.yakovenko at gmail.com (Roman Yakovenko)
Date: Thu, 14 Sep 2006 22:11:14 +0300
Subject: [C++-sig] memory leaks using the "manage_new_object" return
	policy...
In-Reply-To: <20060914185901.20076.qmail@web55507.mail.re4.yahoo.com>
References: <20060914185901.20076.qmail@web55507.mail.re4.yahoo.com>
Message-ID: <7465b6170609141211u31957f7eif607f42d35396ef4@mail.gmail.com>

On 9/14/06, Kevin Jones <investtcartier at yahoo.com> wrote:
> Hi,
>
> I am using the manage_new_object return policy to wrap
> "newed" C++ object in an effort to conserve memory.
>
> Even though I am deleting the returned object in
> Python and invoking the Python garbage collector I
> still have a leak. This is a long-running systems
> adminstration app the read the contents of a very
> large database.
>
> Any suggestions on how to make manage_new_object
> behave correctly?

May be I can help. I am just going through "same" process. It is up to
you to prove
that boost.python does not free the memory. So if you create small test case
( one cpp, one Python script, Jamfile ) that reproduce the error, I
think you will get the answer\solution to your problem pretty quick.

-- 
Roman Yakovenko
C++ Python language binding
http://www.language-binding.net/


From seefeld at sympatico.ca  Thu Sep 14 21:09:21 2006
From: seefeld at sympatico.ca (Stefan Seefeld)
Date: Thu, 14 Sep 2006 15:09:21 -0400
Subject: [C++-sig] memory leaks using the "manage_new_object"
	return	policy...
In-Reply-To: <20060914185901.20076.qmail@web55507.mail.re4.yahoo.com>
References: <20060914185901.20076.qmail@web55507.mail.re4.yahoo.com>
Message-ID: <4509A8E1.7050601@sympatico.ca>

Kevin Jones wrote:
> Hi,
> 
> I am using the manage_new_object return policy to wrap
> "newed" C++ object in an effort to conserve memory.
> 
> Even though I am deleting the returned object in
> Python and invoking the Python garbage collector I
> still have a leak.

The obvious question: Are you sure your C++ objects are
not deleted, while the python (wrapper) objects are ?

Regards,
		Stefan


From investtcartier at yahoo.com  Thu Sep 14 23:09:19 2006
From: investtcartier at yahoo.com (Kevin Jones)
Date: Thu, 14 Sep 2006 21:09:19 +0000 (UTC)
Subject: [C++-sig] memory leaks using the
References: <20060914185901.20076.qmail@web55507.mail.re4.yahoo.com>
	<4509A8E1.7050601@sympatico.ca>
Message-ID: <loom.20060914T230502-21@post.gmane.org>

Stefan Seefeld <seefeld <at> sympatico.ca> writes:

> 
> Kevin Jones wrote:
> > Hi,
> > 
> > I am using the manage_new_object return policy to wrap
> > "newed" C++ object in an effort to conserve memory.
> > 
> > Even though I am deleting the returned object in
> > Python and invoking the Python garbage collector I
> > still have a leak.
> 
> The obvious question: Are you sure your C++ objects are
> not deleted, while the python (wrapper) objects are ?
> 
> Regards,
> 		Stefan
> 

Stefan,

In the python code I do the following:

    del x
    gc.collect()
    msvcrt.heapmin()    

I'm not sure what I could do beyond the above.





From investtcartier at yahoo.com  Thu Sep 14 23:13:55 2006
From: investtcartier at yahoo.com (Kevin Jones)
Date: Thu, 14 Sep 2006 21:13:55 +0000 (UTC)
Subject: [C++-sig]
	=?utf-8?q?memory_leaks_using_the_=22manage=5Fnew=5Fobje?=
	=?utf-8?b?Y3QicmV0dXJuCXBvbGljeS4uLg==?=
References: <20060914185901.20076.qmail@web55507.mail.re4.yahoo.com>
	<7465b6170609141211u31957f7eif607f42d35396ef4@mail.gmail.com>
Message-ID: <loom.20060914T230956-277@post.gmane.org>

Roman Yakovenko <roman.yakovenko <at> gmail.com> writes:

> 
> On 9/14/06, Kevin Jones <investtcartier <at> yahoo.com> wrote:
> > Hi,
> >
> > I am using the manage_new_object return policy to wrap
> > "newed" C++ object in an effort to conserve memory.
> >
> > Even though I am deleting the returned object in
> > Python and invoking the Python garbage collector I
> > still have a leak. This is a long-running systems
> > adminstration app the read the contents of a very
> > large database.
> >
> > Any suggestions on how to make manage_new_object
> > behave correctly?
> 
> May be I can help. I am just going through "same" process. It is up to
> you to prove
> that boost.python does not free the memory. So if you create small test case
> ( one cpp, one Python script, Jamfile ) that reproduce the error, I
> think you will get the answer\solution to your problem pretty quick.
> 
Roman,

One direct solution is the use of Purify which I get access next week. Another
consideration, is that when I changed the return policy to "return_by_value" the
same leak behavior was exhibited (N.B. - I call the python garbage collector
explcitly for each object).  Presumably, boost.python would delete an object
that it was converting to python as it would definitely be garbage otherwise.

I will work on the small test case.

Kevin




From seefeld at sympatico.ca  Thu Sep 14 23:22:23 2006
From: seefeld at sympatico.ca (Stefan Seefeld)
Date: Thu, 14 Sep 2006 17:22:23 -0400
Subject: [C++-sig] memory leaks using the
In-Reply-To: <loom.20060914T230502-21@post.gmane.org>
References: <20060914185901.20076.qmail@web55507.mail.re4.yahoo.com>	<4509A8E1.7050601@sympatico.ca>
	<loom.20060914T230502-21@post.gmane.org>
Message-ID: <4509C80F.40700@sympatico.ca>

Kevin Jones wrote:
> Stefan Seefeld <seefeld <at> sympatico.ca> writes:
> 
>> Kevin Jones wrote:
>>> Hi,
>>>
>>> I am using the manage_new_object return policy to wrap
>>> "newed" C++ object in an effort to conserve memory.
>>>
>>> Even though I am deleting the returned object in
>>> Python and invoking the Python garbage collector I
>>> still have a leak.
>> The obvious question: Are you sure your C++ objects are
>> not deleted, while the python (wrapper) objects are ?
>>
>> Regards,
>> 		Stefan
>>
> 
> Stefan,
> 
> In the python code I do the following:
> 
>     del x
>     gc.collect()
>     msvcrt.heapmin()    
> 
> I'm not sure what I could do beyond the above.

Insert print statements into the destructors of both,
the C++ as well as the python objects and see whether
they get triggered in pairs or not.

And only if you can prove that the python object got in
fact deleted but not the C++ object it is supposed to own
you should claim it is the manage_new_object that is doing
something wrong. (minimal test case etc....)

Regards,
		Stefan




From lakin.wecker at gmail.com  Sun Sep 17 18:44:03 2006
From: lakin.wecker at gmail.com (Lakin Wecker)
Date: Sun, 17 Sep 2006 10:44:03 -0600
Subject: [C++-sig] Problems Compiling Boost.Python w/ Indexing Suite v2
Message-ID: <f27c5670609170944k48e859f2se9cc84500e37dd88@mail.gmail.com>

Hi all,

I've been having problems getting boost.python from CVS with the v2
indexing suite.  I used the instructions from here:
http://www.boost.org/more/getting_started.html#CVS To get boost.python
from cvs.  Then I used the instructions from here:
http://mail.python.org/pipermail/c++-sig/2003-October/005802.html to
switch to v2 of the indexing suite.  Finally, I used
http://www.boost.org/libs/python/doc/building.html as instructions to
build boost.python.  Specifically, I executed "bjam -sTOOLS=gcc test"
from the libs/python/test directory.

The following is the log of the output from the above build command:
http://lakin.weckers.net/v2_log.txt

I'm running Ubuntu 6.06, which has g++ (GCC) 4.0.3 (Ubuntu 4.0.3-1ubuntu5)


Thanks in advance,

Lakin


From roman.yakovenko at gmail.com  Mon Sep 18 07:03:55 2006
From: roman.yakovenko at gmail.com (Roman Yakovenko)
Date: Mon, 18 Sep 2006 08:03:55 +0300
Subject: [C++-sig] Problems Compiling Boost.Python w/ Indexing Suite v2
In-Reply-To: <f27c5670609170944k48e859f2se9cc84500e37dd88@mail.gmail.com>
References: <f27c5670609170944k48e859f2se9cc84500e37dd88@mail.gmail.com>
Message-ID: <7465b6170609172203w6603d033n9f6fea01cb0ed711@mail.gmail.com>

On 9/17/06, Lakin Wecker <lakin.wecker at gmail.com> wrote:
>
> Hi all,
>
> I've been having problems getting boost.python from CVS with the v2
> indexing suite.  I used the instructions from here:
> http://www.boost.org/more/getting_started.html#CVS To get boost.python
> from cvs.  Then I used the instructions from here:
> http://mail.python.org/pipermail/c++-sig/2003-October/005802.html to
> switch to v2 of the indexing suite.  Finally, I used
> http://www.boost.org/libs/python/doc/building.html as instructions to
> build boost.python.  Specifically, I executed "bjam -sTOOLS=gcc test"
> from the libs/python/test directory.
>
> The following is the log of the output from the above build command:
> http://lakin.weckers.net/v2_log.txt
>
> I'm running Ubuntu 6.06, which has g++ (GCC) 4.0.3 (Ubuntu 4.0.3-1ubuntu5)
>

It has nothing to do with indexing suite v2. Just comment
BOOST_PYTHON_UNARY_OPERATOR(repr, lexical_cast<std::string>, repr)
line in operators.hpp in you will be okey. I found this yesterday too.

> Thanks in advance,
>
> Lakin


-- 
Roman Yakovenko
C++ Python language binding
http://www.language-binding.net/


From jack833 at 163.com  Mon Sep 18 13:52:08 2006
From: jack833 at 163.com (jack833)
Date: Mon, 18 Sep 2006 04:52:08 -0700 (PDT)
Subject: [C++-sig] how to use PyRun_SimpleFile in a multithread C++ program?
Message-ID: <6362248.post@talk.nabble.com>


Hi,Every one,
i need help about PyRun_simplefile function,the program i'm writing  is a
C++ program which will start serveral threads .each thread will use
PyRun_simplefile function to do something,but python script file is the
same.
now the program is offten crash down,but there is no problem while only
thread which use PyRun_simplefile,but that is not what i wanted. is there
anyone who can tell me how to resove it. 
-- 
View this message in context: http://www.nabble.com/how-to-use-PyRun_SimpleFile-in-a--multithread-C%2B%2B-program--tf2290802.html#a6362248
Sent from the Python - c++-sig mailing list archive at Nabble.com.



From seefeld at sympatico.ca  Mon Sep 18 14:41:33 2006
From: seefeld at sympatico.ca (Stefan Seefeld)
Date: Mon, 18 Sep 2006 08:41:33 -0400
Subject: [C++-sig] how to use PyRun_SimpleFile in a multithread C++
	program?
In-Reply-To: <6362248.post@talk.nabble.com>
References: <6362248.post@talk.nabble.com>
Message-ID: <450E93FD.4000102@sympatico.ca>

jack833 wrote:
> Hi,Every one,
> i need help about PyRun_simplefile function,the program i'm writing  is a
> C++ program which will start serveral threads .each thread will use
> PyRun_simplefile function to do something,but python script file is the
> same.
> now the program is offten crash down,but there is no problem while only
> thread which use PyRun_simplefile,but that is not what i wanted. is there
> anyone who can tell me how to resove it. 

Yes. You have to serialize access to the python runtime as it isn't fully
thread-safe: http://docs.python.org/api/threads.html

HTH,
		Stefan

-- 

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


From davidhuebel at gmail.com  Mon Sep 18 15:26:46 2006
From: davidhuebel at gmail.com (David Huebel)
Date: Mon, 18 Sep 2006 08:26:46 -0500
Subject: [C++-sig] Complete confustion about using Boost.Python
In-Reply-To: <200609141457.59649.noah.schwartz1@gmail.com>
References: <200609141402.19397.noah.schwartz1@gmail.com>
	<6c75ff8f0609141145s524ebbc6p1c1c4194adf1df5e@mail.gmail.com>
	<200609141457.59649.noah.schwartz1@gmail.com>
Message-ID: <50c6b7b60609180626u6e8e26c1s9eff269303087844@mail.gmail.com>

On 9/14/06, Noah Schwartz <noah.schwartz1 at gmail.com> wrote:
> Failed to find the project root for directory '.'.
> Did not find a project-root.jam file there or in any of its parent
> directories.
> Please consult the documentation at 'http://www.boost.org'.

The Boost directory that should be referenced by BOOST_ROOT is the
directory where Boost was extracted and built before installing.  The
files you need to develop Boost.Python extensions are not installed in
/usr/local/include.

On my machine, the BOOST_ROOT directory contains the following files:
configure, Jamfile, Jamfile.v2, and several others, and the following
directories:  boost, bin, doc, libs, tools, and several others.

This confused me when starting out with Boost.Python, because I
expected that everything I needed had been installed somewhere outside
this directory.  Actually, you will need to keep this directory around
indefinitely to keep developing Boost.Python extensions (unless I have
missed something.)

David Huebel


From npk.gangadhar at gmail.com  Mon Sep 18 16:20:21 2006
From: npk.gangadhar at gmail.com (Gangadhar NPK)
Date: Mon, 18 Sep 2006 19:50:21 +0530
Subject: [C++-sig] Assert fail during deletion for an obj containing
	external object ptr
Message-ID: <61b4e70f0609180720n112e03b9x47706e8d83086261@mail.gmail.com>

All,
I am a beginner for Boost.Python, so please be kind on the responses.
I am trying to create python wrappers for an existing library. I can't
change the source of the library ( I don't even have the src to
change). Also, since most of the classes are via factories, I can't
try to derive from the base classes, so am providing the wrappers by
containership. The python wrappers will contain the pointers to the
3rd party libs.

When I do this, I am able to get the interface to the third party
library's base interface, but during destruction (during the
interpretor exit), I get this assert failure:
---------------------------
Microsoft Visual C++ Debug Library
---------------------------
Debug Assertion Failed!
Program: c:\Python24\python.exe
File: dbgdel.cpp
Line: 51
Expression: _BLOCK_TYPE_IS_VALID(pHead->nBlockUse)
For information on how your program can cause an assertion
failure, see the Visual C++ documentation on asserts.
(Press Retry to debug the application)
---------------------------
Abort   Retry   Ignore
---------------------------
I understand that there might be a double delete or invalid deletion
that is happening, but I can't point exactly where. The code that
generates is attached as a CPP file.
This must be something basic that I am missing, but I can't figure out
from the boost.python documentation or the wiki. Can someone please
point at what I am doing wrong?

Thank You
Gangadhar
P.S:Also, ignore the fact that the 3rd party lib's static function is
not exported in the Wrapper class as a member function. That is
something that will be changed later.
-------------- next part --------------
#include <boost/python.hpp>
#include <boost/enable_shared_from_this.hpp>
#include <boost/shared_ptr.hpp>

using namespace boost::python;
using namespace boost;

/**
* The code that is a 3rd party lib
*/
class ExternalLib
{
protected:
   ExternalLib(){};
public:
   static ExternalLib* Create() {return new ExternalLib();}
   std::string Print() {return "ExternalLib::Print";}
};

/**
* The wrapper to the above class
*/
class Wrapper
{
public:
   typedef Wrapper* Wrapper_ptr;
   typedef ExternalLib* ExternalLib_ptr;
   Wrapper_ptr GetWrapper()
   {
      ExternalLib_ptr m_external(ExternalLib::Create());
      return Wrapper_ptr(this);
   }
   
   std::string Print()
   {
      return m_external->Print();
   }
   ExternalLib_ptr m_external;
};

BOOST_PYTHON_MODULE(BoostTest)
{
   class_<Wrapper>("Wrapper")
      .def("GetWrapper",&Wrapper::GetWrapper,return_value_policy<manage_new_object>())
      .def("Print",&Wrapper::Print)
      ;
}

From roman.yakovenko at gmail.com  Mon Sep 18 16:24:59 2006
From: roman.yakovenko at gmail.com (Roman Yakovenko)
Date: Mon, 18 Sep 2006 17:24:59 +0300
Subject: [C++-sig] Assert fail during deletion for an obj containing
	external object ptr
In-Reply-To: <61b4e70f0609180720n112e03b9x47706e8d83086261@mail.gmail.com>
References: <61b4e70f0609180720n112e03b9x47706e8d83086261@mail.gmail.com>
Message-ID: <7465b6170609180724r5bec6b3bj45777b97370d76d3@mail.gmail.com>

On 9/18/06, Gangadhar NPK <npk.gangadhar at gmail.com> wrote:
> All,
> I am a beginner for Boost.Python, so please be kind on the responses.
> I am trying to create python wrappers for an existing library. I can't
> change the source of the library ( I don't even have the src to
> change). Also, since most of the classes are via factories, I can't
> try to derive from the base classes, so am providing the wrappers by
> containership. The python wrappers will contain the pointers to the
> 3rd party libs.
>
> When I do this, I am able to get the interface to the third party
> library's base interface, but during destruction (during the
> interpretor exit), I get this assert failure:
> ---------------------------
> Microsoft Visual C++ Debug Library
> ---------------------------
> Debug Assertion Failed!
> Program: c:\Python24\python.exe
> File: dbgdel.cpp
> Line: 51
> Expression: _BLOCK_TYPE_IS_VALID(pHead->nBlockUse)
> For information on how your program can cause an assertion
> failure, see the Visual C++ documentation on asserts.
> (Press Retry to debug the application)
> ---------------------------
> Abort   Retry   Ignore
> ---------------------------
> I understand that there might be a double delete or invalid deletion
> that is happening, but I can't point exactly where. The code that
> generates is attached as a CPP file.
> This must be something basic that I am missing, but I can't figure out
> from the boost.python documentation or the wiki. Can someone please
> point at what I am doing wrong?
>
> Thank You
> Gangadhar
> P.S:Also, ignore the fact that the 3rd party lib's static function is
> not exported in the Wrapper class as a member function. That is
> something that will be changed later.

This is not a bug. If you want to debug your extension you will need
Python built
in "debug" mode. Try to compile your extension in release mode and it
should work.
( At least the example you attached )

-- 
Roman Yakovenko
C++ Python language binding
http://www.language-binding.net/


From roman.yakovenko at gmail.com  Mon Sep 18 16:32:52 2006
From: roman.yakovenko at gmail.com (Roman Yakovenko)
Date: Mon, 18 Sep 2006 17:32:52 +0300
Subject: [C++-sig] Assert fail during deletion for an obj containing
	external object ptr
In-Reply-To: <7465b6170609180724r5bec6b3bj45777b97370d76d3@mail.gmail.com>
References: <61b4e70f0609180720n112e03b9x47706e8d83086261@mail.gmail.com>
	<7465b6170609180724r5bec6b3bj45777b97370d76d3@mail.gmail.com>
Message-ID: <7465b6170609180732s683004b4sd59ac1962fb676e@mail.gmail.com>

On 9/18/06, Roman Yakovenko <roman.yakovenko at gmail.com> wrote:
> On 9/18/06, Gangadhar NPK <npk.gangadhar at gmail.com> wrote:
> > All,
> > I am a beginner for Boost.Python, so please be kind on the responses.
> > I am trying to create python wrappers for an existing library. I can't
> > change the source of the library ( I don't even have the src to
> > change). Also, since most of the classes are via factories, I can't
> > try to derive from the base classes, so am providing the wrappers by
> > containership. The python wrappers will contain the pointers to the
> > 3rd party libs.
> >
> > When I do this, I am able to get the interface to the third party
> > library's base interface, but during destruction (during the
> > interpretor exit), I get this assert failure:
> > ---------------------------
> > Microsoft Visual C++ Debug Library
> > ---------------------------
> > Debug Assertion Failed!
> > Program: c:\Python24\python.exe
> > File: dbgdel.cpp
> > Line: 51
> > Expression: _BLOCK_TYPE_IS_VALID(pHead->nBlockUse)
> > For information on how your program can cause an assertion
> > failure, see the Visual C++ documentation on asserts.
> > (Press Retry to debug the application)
> > ---------------------------
> > Abort   Retry   Ignore
> > ---------------------------
> > I understand that there might be a double delete or invalid deletion
> > that is happening, but I can't point exactly where. The code that
> > generates is attached as a CPP file.
> > This must be something basic that I am missing, but I can't figure out
> > from the boost.python documentation or the wiki. Can someone please
> > point at what I am doing wrong?
> >
> > Thank You
> > Gangadhar
> > P.S:Also, ignore the fact that the 3rd party lib's static function is
> > not exported in the Wrapper class as a member function. That is
> > something that will be changed later.
>
> This is not a bug. If you want to debug your extension you will need
> Python built in "debug" mode.

> Try to compile your extension in release mode and it should work.
> ( At least the example you attached )

I am not sure about last 2 sentence. The wrapper class is very very strange.


-- 
Roman Yakovenko
C++ Python language binding
http://www.language-binding.net/


From ctrembla at pcigeomatics.com  Mon Sep 18 16:38:14 2006
From: ctrembla at pcigeomatics.com (CT)
Date: Mon, 18 Sep 2006 10:38:14 -0400
Subject: [C++-sig] range and iterator
Message-ID: <MPG.1f78795ff2c28722989683@news.gmane.org>

Hi ,

Does someone can give me some help on how to expose correctly an array 
of pointers using boost.python ?


here is a small example of my code

i have a class 

class A
{
public:
	A(std::string &oName);
	~A(){}
	const std::string &GetName() const;
}


class B
{
	std::vector<*A> *poNames

public:	
	std::vector<*A> GetNames() { return *poNames }

}

assume that there is two instances of A in the container of the object B


now for the boost::python code. I would like to be able to expose class 
a and class b and the vector of A*


typedef std::vector<A*> array_proj;
array_proj::iterator begin_array_proj(array_proj &x) {return x.begin();}
array_proj::iterator end_array_proj(array_proj &x) {return x.end();}

BOOST_PYTHON_MODULE(pysession)
{

	class_<A>("object a",init<std::string &)
	.def("GetName",&A::GetName,return_value_policy	
	                <copy_const_reference>() )
	;
	
	class_<B>("object B")
	.def("GetNames", &B::GetNames,
	     return_value_policy<manage_new_object>() )
	;

	class_<std::vector<*A> >("array of pointer to objects A")
	.def("__iter__", range( begin_array_proj, end_array_proj) )
	;
	
}

Now in python for example I call

#create a new object B
#and iterate over all of the object A 

#assume that there are two object of type A in the vector
oNewSession = pysession.B()  
oVector = oNewSession.GetNames()        
        
for oA in oVector.__iter__():
    print (oA.GetName())       


i have this error when i start the python script

    for oA in oVector.__iter__():
TypeError: No to_python (by-value) converter found for C++ type: class 
::A *


Does someone can give me some help on how to expose correctly an array 
of pointers using boost.python ?





From thor.arne.johansen at ibas.no  Mon Sep 18 19:06:26 2006
From: thor.arne.johansen at ibas.no (thor.arne.johansen at ibas.no)
Date: Mon, 18 Sep 2006 19:06:26 +0200
Subject: [C++-sig] Py++ tutorial
In-Reply-To: <7465b6170609180732s683004b4sd59ac1962fb676e@mail.gmail.com>
Message-ID: <OFAB06BD97.A5ECF6AD-ONC12571ED.005CF67F-C12571ED.005DF8F7@ibas.no>


I've just tried the tutorial code for Py++. (hello_world.cpp)

The code seem to build OK, but the following fails:

>>> a = hw.animal('Dog')
>>> a.name()
'Dog'
>>> a.get_name_ptr()
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
TypeError: No Python class registered for C++ class class
std::basic_string<char
,struct std::char_traits<char>,class std::allocator<char> >
>>>

Is this the expected behavior? If so, how do I 'register' a Python class of
the correct type (using py++)

--
Thor Arne Johansen
R&D Director
Ibas AS



From roman.yakovenko at gmail.com  Mon Sep 18 21:28:49 2006
From: roman.yakovenko at gmail.com (Roman Yakovenko)
Date: Mon, 18 Sep 2006 22:28:49 +0300
Subject: [C++-sig] implicitly_convertible or order of init's problem
Message-ID: <7465b6170609181228g42d2625bk2c09c467a12d9e7e@mail.gmail.com>

Hi. I found some very very strange behaviour.
Please, consider next code:

struct vector{

    vector(){}

    vector( double ){}

    vector( const vector& ){}

};



struct float_vector{

    float_vector(){}

    float_vector( const float_vector& ){}

    float_vector( const vector& ){}

    float_vector( float ){}

};


namespace bp = boost::python;



BOOST_PYTHON_MODULE( ic_ext ){



    bp::class_< float_vector >( "float_vector" )

        .def( bp::init< >() )

        .def( bp::init< vector const & >() )

        .def( bp::init< float >() )

        .def( bp::init< const float_vector& >() );



    bp::implicitly_convertible< vector const &, float_vector >();



    bp::implicitly_convertible< float, float_vector >();



    bp::class_< vector >( "vector" )

        .def( bp::init< >() )

        .def( bp::init< double >() );



    bp::implicitly_convertible< double, vector >();



}


import ic_ext
ic_ext.float_vector( 5.0 )

The last line fails with the error:
TypeError: No registered converter was able to extract a C++ reference to type
vector from this Python object of type float

Question:
Why Boost.Python does not use float_vector( float ) constructor?

If I reorder the init methods "ic_ext.float_vector( 5.0 )" expression pass:

    bp::class_< float_vector >( "float_vector" )
        .def( bp::init< >() )
        .def( bp::init< vector const & >() )
        .def( bp::init< const float_vector& >() )
        .def( bp::init< float >() );

If I comment out "bp::implicitly_convertible< vector const &, float_vector >();"
than "ic_ext.float_vector( 5.0 )" expression pass.

It could be nice to understand why this happens. I attached small example
that reproduce the problem.( I hope this time I did it right :-) )

Thanks.

-- 
Roman Yakovenko
C++ Python language binding
http://www.language-binding.net/
-------------- next part --------------
A non-text attachment was scrubbed...
Name: ic.cpp
Type: text/x-c++src
Size: 864 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/cplusplus-sig/attachments/20060918/f91507a7/attachment.cpp>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: boost-build.jam
Type: application/octet-stream
Size: 283 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/cplusplus-sig/attachments/20060918/f91507a7/attachment.obj>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: Jamroot
Type: application/octet-stream
Size: 378 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/cplusplus-sig/attachments/20060918/f91507a7/attachment-0001.obj>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: test.py
Type: text/x-python
Size: 44 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/cplusplus-sig/attachments/20060918/f91507a7/attachment.py>

From roman.yakovenko at gmail.com  Mon Sep 18 21:46:43 2006
From: roman.yakovenko at gmail.com (Roman Yakovenko)
Date: Mon, 18 Sep 2006 22:46:43 +0300
Subject: [C++-sig] Py++ tutorial
In-Reply-To: <OFAB06BD97.A5ECF6AD-ONC12571ED.005CF67F-C12571ED.005DF8F7@ibas.no>
References: <7465b6170609180732s683004b4sd59ac1962fb676e@mail.gmail.com>
	<OFAB06BD97.A5ECF6AD-ONC12571ED.005CF67F-C12571ED.005DF8F7@ibas.no>
Message-ID: <7465b6170609181246o51a885ebg9ddaf3955765557@mail.gmail.com>

On 9/18/06, thor.arne.johansen at ibas.no <thor.arne.johansen at ibas.no> wrote:
>
> I've just tried the tutorial code for Py++. (hello_world.cpp)
>
> The code seem to build OK, but the following fails:
>
> >>> a = hw.animal('Dog')
> >>> a.name()
> 'Dog'
> >>> a.get_name_ptr()
> Traceback (most recent call last):
>   File "<stdin>", line 1, in ?
> TypeError: No Python class registered for C++ class class
> std::basic_string<char
> ,struct std::char_traits<char>,class std::allocator<char> >
> >>>
>
> Is this the expected behavior?

I think yes. The main purpose of the tutorials I built, was to
introduce Py++ and
not Boost.Python.

Here is the context we are talking about:

struct animal{

    animal( const std::string& name="" )
    : m_name( name )
    {}

    const std::string* get_name_ptr() const
    { return &m_name; }

private:
    std::string m_name;

};

I set return_internal_reference call policies for get_name_ptr member function.
Obviously, strings are immutable types in Python so they can not be changed
from C++.

> If so, how do I 'register' a Python class of the correct type (using py++)

In this case you don't. Boost.Python already did it for you. I think
that in this
case I should change the call policies ( or to provide better example ).

-- 
Roman Yakovenko
C++ Python language binding
http://www.language-binding.net/


From lakin.wecker at gmail.com  Mon Sep 18 22:23:53 2006
From: lakin.wecker at gmail.com (Lakin Wecker)
Date: Mon, 18 Sep 2006 14:23:53 -0600
Subject: [C++-sig] Fwd: Problems Compiling Boost.Python w/ Indexing Suite v2
In-Reply-To: <f27c5670609180930l73be50e4t775e5d91120e140@mail.gmail.com>
References: <f27c5670609170944k48e859f2se9cc84500e37dd88@mail.gmail.com>
	<7465b6170609172203w6603d033n9f6fea01cb0ed711@mail.gmail.com>
	<f27c5670609180930l73be50e4t775e5d91120e140@mail.gmail.com>
Message-ID: <f27c5670609181323r3e430a4dj4447eaa7af90ed57@mail.gmail.com>

Posting this back to the c++-sig mailing list, where it belongs. :)

Lakin

---------- Forwarded message ----------
From: Lakin Wecker <lakin.wecker at gmail.com>
Date: Sep 18, 2006 10:30 AM
Subject: Re: Problems Compiling Boost.Python w/ Indexing Suite v2
To: python-ogre-developers at googlegroups.com


I assume you meant boost/python/operators.hpp ... as that is the only
file where I could find that line.  I commented it out, did a bjam
clean, and then rebuilt using the same command.  Still no luck, but
different problems this time:

http://lakin.weckers.net/v2_log.2.txt

At this point, I haven't had time to look into the problem, so I hope
it isn't something simple.

Lakin

On 9/17/06, Roman Yakovenko <roman.yakovenko at gmail.com> wrote:
>
> On 9/17/06, Lakin Wecker <lakin.wecker at gmail.com> wrote:
> >
> > Hi all,
> >
> > I've been having problems getting boost.python from CVS with the v2
> > indexing suite.  I used the instructions from here:
> > http://www.boost.org/more/getting_started.html#CVS To get boost.python
> > from cvs.  Then I used the instructions from here:
> > http://mail.python.org/pipermail/c++-sig/2003-October/005802.html to
> > switch to v2 of the indexing suite.  Finally, I used
> > http://www.boost.org/libs/python/doc/building.html as instructions to
> > build boost.python.  Specifically, I executed "bjam -sTOOLS=gcc test"
> > from the libs/python/test directory.
> >
> > The following is the log of the output from the above build command:
> > http://lakin.weckers.net/v2_log.txt
> >
> > I'm running Ubuntu 6.06, which has g++ (GCC) 4.0.3 (Ubuntu 4.0.3-1ubuntu5)
> >
>
> It has nothing to do with indexing suite v2. Just comment
> BOOST_PYTHON_UNARY_OPERATOR(repr, lexical_cast<std::string>, repr)
> line in operators.hpp in you will be okey. I found this yesterday too.
>
> > Thanks in advance,
> >
> > Lakin
>
>
> --
> Roman Yakovenko
> C++ Python language binding
> http://www.language-binding.net/
>
> --~--~---------~--~----~------------~-------~--~----~
> You received this message because you are subscribed to the Google Groups "Python Ogre Developers" group.
> To post to this group, send email to python-ogre-developers at googlegroups.com
> To unsubscribe from this group, send email to python-ogre-developers-unsubscribe at googlegroups.com
> For more options, visit this group at http://groups.google.com/group/python-ogre-developers
> -~----------~----~----~----~------~----~------~--~---
>
>


From dave at boost-consulting.com  Mon Sep 18 22:45:00 2006
From: dave at boost-consulting.com (David Abrahams)
Date: Mon, 18 Sep 2006 16:45:00 -0400
Subject: [C++-sig] Exception 'No to_python (by-value) converter found -
	'garbage collection'
References: <6c75ff8f0609130956s6b0e7c9m9fdb8d44840fd26b@mail.gmail.com>
Message-ID: <87odtcgb03.fsf@pereiro.luannocracy.com>

"Eduardo Picado" <eduardo.picado at gmail.com> writes:

> Hi,
>
> Im writing the following kind of python code using a C++ wrapped library:
>
>
>
> system = TASRSystem.getASRSystem()
> system.setParameter(TASRSystem.Option.asrServersListenPort, 32000)
>
> try:
> 	system.initialize(1, 3000)
> except Exception, e:
> 	sys.stderr.write("[E] : " +  e.__class__.__name__ + str(e) )
> print "ok"
>
>
>
> Running it I get:
>
>
>
> ok
> Exception exceptions.TypeError: 'No to_python (by-value) converter
> found for C++ type:
> class TASRSystem' in 'garbage collection' ignored
> Fatal Python error: unexpected exception during garbage collection

This surely means you've wrapped a C++ function that returns a
TASRSystem and that function is getting invoked from some Python
class' __del__ method.

> The declaration of TASRSystem is like:
>
> BOOST_PYTHON_MEMBER_FUNCTION_OVERLOADS(TASRSystem_initialize_overloads_0_2,
> initialize, 0, 2)
>
> class_< TASRSystem, boost::noncopyable >("TASRSystem", no_init)
> 	.def("getASRSystem", &TASRSystem::getASRSystem,
> return_value_policy<reference_existing_object>())
> 	.def("initialize", &TASRSystem::initialize,
> TASRSystem_initialize_overloads_0_2())
> 	.staticmethod("getASRSystem")
>
>
> - TASRSystem.getASRSystem() is an static method, that creates and
> returns an instance of an ASRSystem. Seems ok.
> I've tried to use with_custodian_and_ward_postcall <1, 0>, to ensure
> that the lifetime of the created instance is as long
> as that of the Python object, but it didn't worked - gives "argument
> index out of range".

That's because the fewest number of arguments wrapped by your
invocation of BOOST_PYTHON_MEMBER_FUNCTION_OVERLOADS is zero.

-- 
Dave Abrahams
Boost Consulting
www.boost-consulting.com



From dave at boost-consulting.com  Mon Sep 18 22:49:09 2006
From: dave at boost-consulting.com (David Abrahams)
Date: Mon, 18 Sep 2006 16:49:09 -0400
Subject: [C++-sig] What if a Python override returns several values?
References: <edhhnt$khq$1@sea.gmane.org> <44FDABA2.60809@pixar.com>
	<87k649q33b.fsf@pereiro.luannocracy.com>
	<4506E1F6.5080604@pixar.com> <u3bawzwvl.fsf@boost-consulting.com>
	<45071956.2020104@pixar.com>
	<878xkovv7a.fsf@pereiro.luannocracy.com>
	<eebq0u$bvi$1@sea.gmane.org>
Message-ID: <87irjkgat6.fsf@pereiro.luannocracy.com>

Matthias Baas <baas at ira.uka.de> writes:

> David Abrahams wrote:
>>> void g(int& a, int& b) {
>>>      boost::python::tuple tup = this->get_override("g")();
>>>      a = ...extract an int from tup[0]...
>>>      b = ...extract an int from tup[1]...
>>> }
>>>
>>> [...]
>>> TypeError: No registered converter was able to extract a C++ reference
>>> to type boost::python::tuple from this Python object of type tuple
>>[...]
>> Okay.  Does it work if you don't use copy-initialization?
>
> Now that you mention it I actually have to confess that I was not aware 
> that this could make a difference. My original code rather looked like this:
>
>    boost::python::tuple tup;
>    tup = this->get_override("g")();

That's neither copy initialization nor direct initialization.  That's
assignment.

> And this results in the above TypeError exception.
> I only noticed now that the following code actually does work (sorry):
>
>    boost::python::tuple tup = this->get_override("g")();
>
> I don't suppose this difference is intentional, is it?

Yes.

> Unfortunately, it's not that trivial for me to get to the latter version 
> as the code I'm using is not written manually but is generated by a tool 
> which separates the declaration of variables from their usage.

That's almost always a bad idea in C++.  I think you should fix the
tool.

>>    boost::python::tuple tup(this->get_override("g")());
>
> No, this results in the following compile error:
>
> call of overloaded `tuple(boost::python::detail::method_result)' is 
> ambiguous
> candidates are:
> boost::python::tuple::tuple(const boost::python::tuple&)
> boost::python::tuple::tuple(boost::python::detail::new_non_null_reference_t*)
> boost::python::tuple::tuple(boost::python::detail::new_reference_t*)
> boost::python::tuple::tuple(boost::python::detail::borrowed_reference_t*)
> boost::python::tuple::tuple(const T&) [with T = 
> boost::python::detail::method_result]

Ah, right, that makes sense.

>>    boost::python::object t = this->get_override("g")();
>>    boost::python::tuple tup(t);
>
> Yes, this works (see above).
>
>> what about:
>> 
>>    boost::python::tuple tup 
>>      = this->get_override("g")().as<boost::python::tuple>();
>
> Yes (what's this anyway? I didn't see that in the reference manual).

It's a way to force the conversion to boost::python::tuple.

-- 
Dave Abrahams
Boost Consulting
www.boost-consulting.com



From dave at boost-consulting.com  Mon Sep 18 22:49:34 2006
From: dave at boost-consulting.com (David Abrahams)
Date: Mon, 18 Sep 2006 16:49:34 -0400
Subject: [C++-sig] What if a Python override returns several values?
References: <edhhnt$khq$1@sea.gmane.org> <44FDABA2.60809@pixar.com>
	<87k649q33b.fsf@pereiro.luannocracy.com>
	<4506E1F6.5080604@pixar.com> <u3bawzwvl.fsf@boost-consulting.com>
	<45071956.2020104@pixar.com>
	<878xkovv7a.fsf@pereiro.luannocracy.com>
	<eebq0u$bvi$1@sea.gmane.org> <45098827.9030101@pixar.com>
Message-ID: <87d59sgash.fsf@pereiro.luannocracy.com>

Alex Mohr <amohr at pixar.com> writes:

>>> what about:
>>>
>>>    boost::python::tuple tup 
>>>      = this->get_override("g")().as<boost::python::tuple>();
>> 
>> Yes (what's this anyway? I didn't see that in the reference manual).
>
> It's not documented.  It appears to be a convenience.  It's a method 
> that appears to be equivalent to override::method_result::operator T().

It's an inconvenience that might be useful as a workaround.

-- 
Dave Abrahams
Boost Consulting
www.boost-consulting.com



From dave at boost-consulting.com  Mon Sep 18 23:00:15 2006
From: dave at boost-consulting.com (David Abrahams)
Date: Mon, 18 Sep 2006 17:00:15 -0400
Subject: [C++-sig] Complete confustion about using Boost.Python
References: <200609141402.19397.noah.schwartz1@gmail.com>
Message-ID: <871wq8gaao.fsf@pereiro.luannocracy.com>

Noah Schwartz <noah.schwartz1 at gmail.com> writes:

> Hi,
>
> ????????I've recently built Boost based on the instructions at 
> http://boost.org/more/getting_started.html. It found my python installation 
> OK and so as far as I know that means Boost.Python has been installed on my 
> system.
>
> ????????The issue I am having is that I can't seem to figure out how to build 
> any 
> modules that are not within the directory tree of the source I extracted from 
> sourceforge. If I copy the exact tutorial out of libs/python/example/tutorial 
> and try to build it elsewhere, it complains about not being able to find a 
> project-root.jam file. 

Sounds like your boost-build.jam is pointing at the wrong directory.
It should be pointing at tools/build/v1, not tools/build/v2.

Did you copy libs/python/example/boost-build.jam and edit it appropriately?

> Also, I have noticed that if I set the BOOST_ROOT path 
> in say, for example, the Jamrules file in the that tutorial directory to 
> where boost is now actually installed (/usr/local/include/boost-1_33_1), I 
> can no longer build. 

Build what?  Aren't you already unable to build?

-- 
Dave Abrahams
Boost Consulting
www.boost-consulting.com



From dave at boost-consulting.com  Mon Sep 18 23:17:29 2006
From: dave at boost-consulting.com (David Abrahams)
Date: Mon, 18 Sep 2006 17:17:29 -0400
Subject: [C++-sig] custom smart pointers
References: <7465b6170608291048y6e2010d4ofa84fbd5760da4d3@mail.gmail.com>
	<uac5lvczx.fsf@boost-consulting.com>
	<7465b6170608311358s1c02c6c9hc26222fae8cfdef6@mail.gmail.com>
	<878xkppxyq.fsf@pereiro.luannocracy.com>
	<7465b6170609121355q42dc2dc7m526e6eaf5c3344ff@mail.gmail.com>
	<87wt88uf76.fsf@pereiro.luannocracy.com>
	<7465b6170609131210h30285ef8pd4c8b7e624ff82f3@mail.gmail.com>
Message-ID: <87venkeuxi.fsf@pereiro.luannocracy.com>

"Roman Yakovenko" <roman.yakovenko at gmail.com> writes:

> Python test code also simple:
>
> inst = cspc_ext.derived_t()
> cspc_ext.val_get_value(inst)
> cspc_ext.const_ref_get_value(inst) # <= in this line I get segmentation fault.

AFAICT this example doesn't work as packaged (the Jamroot doesn't look for
test.py).

Your smart pointer is buggy; it has only a default copy ctor.  The
enclosed works.

-------------- next part --------------
A non-text attachment was scrubbed...
Name: cspc.cpp
Type: text/x-c++src
Size: 3043 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/cplusplus-sig/attachments/20060918/5f025154/attachment.cpp>
-------------- next part --------------



-- 
Dave Abrahams
Boost Consulting
www.boost-consulting.com

From dave at boost-consulting.com  Mon Sep 18 23:21:21 2006
From: dave at boost-consulting.com (David Abrahams)
Date: Mon, 18 Sep 2006 17:21:21 -0400
Subject: [C++-sig] memory leaks using the
References: <20060914185901.20076.qmail@web55507.mail.re4.yahoo.com>
	<4509A8E1.7050601@sympatico.ca>
	<loom.20060914T230502-21@post.gmane.org>
	<4509C80F.40700@sympatico.ca>
Message-ID: <87k640eur2.fsf@pereiro.luannocracy.com>

Stefan Seefeld <seefeld at sympatico.ca> writes:

> Kevin Jones wrote:
>> Stefan Seefeld <seefeld <at> sympatico.ca> writes:
>> 
>>> Kevin Jones wrote:
>>>> Hi,
>>>>
>>>> I am using the manage_new_object return policy to wrap
>>>> "newed" C++ object in an effort to conserve memory.
>>>>
>>>> Even though I am deleting the returned object in
>>>> Python and invoking the Python garbage collector I
>>>> still have a leak.
>>> The obvious question: Are you sure your C++ objects are
>>> not deleted, while the python (wrapper) objects are ?
>>>
>>> Regards,
>>> 		Stefan
>>>
>> 
>> Stefan,
>> 
>> In the python code I do the following:
>> 
>>     del x
>>     gc.collect()
>>     msvcrt.heapmin()    
>> 
>> I'm not sure what I could do beyond the above.
>
> Insert print statements into the destructors of both,
> the C++ as well as the python objects and see whether
> they get triggered in pairs or not.

They don't always correspond in pairs, so that won't prove anything.

-- 
Dave Abrahams
Boost Consulting
www.boost-consulting.com



From roman.yakovenko at gmail.com  Mon Sep 18 23:26:39 2006
From: roman.yakovenko at gmail.com (Roman Yakovenko)
Date: Tue, 19 Sep 2006 00:26:39 +0300
Subject: [C++-sig] custom smart pointers
In-Reply-To: <87venkeuxi.fsf@pereiro.luannocracy.com>
References: <7465b6170608291048y6e2010d4ofa84fbd5760da4d3@mail.gmail.com>
	<uac5lvczx.fsf@boost-consulting.com>
	<7465b6170608311358s1c02c6c9hc26222fae8cfdef6@mail.gmail.com>
	<878xkppxyq.fsf@pereiro.luannocracy.com>
	<7465b6170609121355q42dc2dc7m526e6eaf5c3344ff@mail.gmail.com>
	<87wt88uf76.fsf@pereiro.luannocracy.com>
	<7465b6170609131210h30285ef8pd4c8b7e624ff82f3@mail.gmail.com>
	<87venkeuxi.fsf@pereiro.luannocracy.com>
Message-ID: <7465b6170609181426g3d5b115i1a254d8a6dc70884@mail.gmail.com>

On 9/19/06, David Abrahams <dave at boost-consulting.com> wrote:
> "Roman Yakovenko" <roman.yakovenko at gmail.com> writes:
>
> > Python test code also simple:
> >
> > inst = cspc_ext.derived_t()
> > cspc_ext.val_get_value(inst)
> > cspc_ext.const_ref_get_value(inst) # <= in this line I get segmentation fault.
>
> AFAICT this example doesn't work as packaged (the Jamroot doesn't look for
> test.py).

It is possible. I don't use bjam. It is much easier to use scons.
( At least I don't have to define twice the environment of the project )

> Your smart pointer is buggy; it has only a default copy ctor.  The
> enclosed works.

Thank you very much for your help. I will try to do much better job next time.

-- 
Roman Yakovenko
C++ Python language binding
http://www.language-binding.net/


From dave at boost-consulting.com  Mon Sep 18 23:21:44 2006
From: dave at boost-consulting.com (David Abrahams)
Date: Mon, 18 Sep 2006 17:21:44 -0400
Subject: [C++-sig] Problems Compiling Boost.Python w/ Indexing Suite v2
References: <f27c5670609170944k48e859f2se9cc84500e37dd88@mail.gmail.com>
	<7465b6170609172203w6603d033n9f6fea01cb0ed711@mail.gmail.com>
Message-ID: <87eju8euqf.fsf@pereiro.luannocracy.com>

"Roman Yakovenko" <roman.yakovenko at gmail.com> writes:

> On 9/17/06, Lakin Wecker <lakin.wecker at gmail.com> wrote:
>>
>> Hi all,
>>
>> I've been having problems getting boost.python from CVS with the v2
>> indexing suite.  I used the instructions from here:
>> http://www.boost.org/more/getting_started.html#CVS To get boost.python
>> from cvs.  Then I used the instructions from here:
>> http://mail.python.org/pipermail/c++-sig/2003-October/005802.html to
>> switch to v2 of the indexing suite.  Finally, I used
>> http://www.boost.org/libs/python/doc/building.html as instructions to
>> build boost.python.  Specifically, I executed "bjam -sTOOLS=gcc test"
>> from the libs/python/test directory.
>>
>> The following is the log of the output from the above build command:
>> http://lakin.weckers.net/v2_log.txt
>>
>> I'm running Ubuntu 6.06, which has g++ (GCC) 4.0.3 (Ubuntu 4.0.3-1ubuntu5)
>>
>
> It has nothing to do with indexing suite v2. Just comment
> BOOST_PYTHON_UNARY_OPERATOR(repr, lexical_cast<std::string>, repr)
> line in operators.hpp in you will be okey. I found this yesterday too.

Fixed now in CVS

-- 
Dave Abrahams
Boost Consulting
www.boost-consulting.com



From dave at boost-consulting.com  Mon Sep 18 23:20:34 2006
From: dave at boost-consulting.com (David Abrahams)
Date: Mon, 18 Sep 2006 17:20:34 -0400
Subject: [C++-sig] memory leaks using the
	"manage_new_object"return	policy...
References: <20060914185901.20076.qmail@web55507.mail.re4.yahoo.com>
	<7465b6170609141211u31957f7eif607f42d35396ef4@mail.gmail.com>
	<loom.20060914T230956-277@post.gmane.org>
Message-ID: <87psdseusd.fsf@pereiro.luannocracy.com>

Kevin Jones <investtcartier at yahoo.com> writes:

> One direct solution is the use of Purify which I get access next week. Another
> consideration, is that when I changed the return policy to "return_by_value" the
> same leak behavior was exhibited (N.B. - I call the python garbage collector
> explcitly for each object).  Presumably, boost.python would delete an object
> that it was converting to python as it would definitely be garbage otherwise.
>
> I will work on the small test case.

The lack of
http://boost.org/libs/python/todo.html#pyfinalize-safety
can lead to symptoms that look like leaks.  You can't expect to start
and stop the python interpreter during a single process.

-- 
Dave Abrahams
Boost Consulting
www.boost-consulting.com



From noah.schwartz1 at gmail.com  Mon Sep 18 23:48:58 2006
From: noah.schwartz1 at gmail.com (Noah Schwartz)
Date: Mon, 18 Sep 2006 17:48:58 -0400
Subject: [C++-sig] Complete confustion about using Boost.Python
In-Reply-To: <50c6b7b60609180626u6e8e26c1s9eff269303087844@mail.gmail.com>
References: <200609141402.19397.noah.schwartz1@gmail.com>
	<6c75ff8f0609141145s524ebbc6p1c1c4194adf1df5e@mail.gmail.com>
	<200609141457.59649.noah.schwartz1@gmail.com>
	<50c6b7b60609180626u6e8e26c1s9eff269303087844@mail.gmail.com>
Message-ID: <c9c12f860609181448y68561cd2re7b020c8a613d317@mail.gmail.com>

Thanks David ;D

I had eventually figured this out when I realized that I didn't have a
build/tools/v1 directory installed anyone on my machine.

This is sort of confusing and isn't at all immediately obvious (to me
at least). Maybe the documentation should be updated?

Noah

On 9/18/06, David Huebel <davidhuebel at gmail.com> wrote:
> On 9/14/06, Noah Schwartz <noah.schwartz1 at gmail.com> wrote:
> > Failed to find the project root for directory '.'.
> > Did not find a project-root.jam file there or in any of its parent
> > directories.
> > Please consult the documentation at 'http://www.boost.org'.
>
> The Boost directory that should be referenced by BOOST_ROOT is the
> directory where Boost was extracted and built before installing.  The
> files you need to develop Boost.Python extensions are not installed in
> /usr/local/include.
>
> On my machine, the BOOST_ROOT directory contains the following files:
> configure, Jamfile, Jamfile.v2, and several others, and the following
> directories:  boost, bin, doc, libs, tools, and several others.
>
> This confused me when starting out with Boost.Python, because I
> expected that everything I needed had been installed somewhere outside
> this directory.  Actually, you will need to keep this directory around
> indefinitely to keep developing Boost.Python extensions (unless I have
> missed something.)
>
> David Huebel
> _______________________________________________
> C++-sig mailing list
> C++-sig at python.org
> http://mail.python.org/mailman/listinfo/c++-sig
>


From noah.schwartz1 at gmail.com  Mon Sep 18 23:50:35 2006
From: noah.schwartz1 at gmail.com (Noah Schwartz)
Date: Mon, 18 Sep 2006 17:50:35 -0400
Subject: [C++-sig] Complete confustion about using Boost.Python
In-Reply-To: <871wq8gaao.fsf@pereiro.luannocracy.com>
References: <200609141402.19397.noah.schwartz1@gmail.com>
	<871wq8gaao.fsf@pereiro.luannocracy.com>
Message-ID: <c9c12f860609181450x19b640bdue01769c7f41a923a@mail.gmail.com>

> > Also, I have noticed that if I set the BOOST_ROOT path
> > in say, for example, the Jamrules file in the that tutorial directory to
> > where boost is now actually installed (/usr/local/include/boost-1_33_1), I
> > can no longer build.
>
> Build what?  Aren't you already unable to build?
Yeah, I wasn't very clear. I was reffering to the tutorial that was in
the directory tree of the extracted source. That built fine until I
tried changing the BOOST_ROOT path in the Jamrules file to where I had
installed boost (/usr/local/include)

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


From eduardo.picado at gmail.com  Mon Sep 18 23:53:43 2006
From: eduardo.picado at gmail.com (Eduardo Picado)
Date: Mon, 18 Sep 2006 23:53:43 +0200
Subject: [C++-sig] Complete confustion about using Boost.Python
In-Reply-To: <c9c12f860609181448y68561cd2re7b020c8a613d317@mail.gmail.com>
References: <200609141402.19397.noah.schwartz1@gmail.com>
	<6c75ff8f0609141145s524ebbc6p1c1c4194adf1df5e@mail.gmail.com>
	<200609141457.59649.noah.schwartz1@gmail.com>
	<50c6b7b60609180626u6e8e26c1s9eff269303087844@mail.gmail.com>
	<c9c12f860609181448y68561cd2re7b020c8a613d317@mail.gmail.com>
Message-ID: <6c75ff8f0609181453hd32ec79ld84e1f800e8f9022@mail.gmail.com>

Now I remember I've had exactly the same problem when doing my first
extensions using BPL and bjam.
I think would be good to have at least a FAQ "How to compile
extensions out of the example directory?"

Eduardo


On 9/18/06, Noah Schwartz <noah.schwartz1 at gmail.com> wrote:
> Thanks David ;D
>
> I had eventually figured this out when I realized that I didn't have a
> build/tools/v1 directory installed anyone on my machine.
>
> This is sort of confusing and isn't at all immediately obvious (to me
> at least). Maybe the documentation should be updated?
>
> Noah
>
> On 9/18/06, David Huebel <davidhuebel at gmail.com> wrote:
> > On 9/14/06, Noah Schwartz <noah.schwartz1 at gmail.com> wrote:
> > > Failed to find the project root for directory '.'.
> > > Did not find a project-root.jam file there or in any of its parent
> > > directories.
> > > Please consult the documentation at 'http://www.boost.org'.
> >
> > The Boost directory that should be referenced by BOOST_ROOT is the
> > directory where Boost was extracted and built before installing.  The
> > files you need to develop Boost.Python extensions are not installed in
> > /usr/local/include.
> >
> > On my machine, the BOOST_ROOT directory contains the following files:
> > configure, Jamfile, Jamfile.v2, and several others, and the following
> > directories:  boost, bin, doc, libs, tools, and several others.
> >
> > This confused me when starting out with Boost.Python, because I
> > expected that everything I needed had been installed somewhere outside
> > this directory.  Actually, you will need to keep this directory around
> > indefinitely to keep developing Boost.Python extensions (unless I have
> > missed something.)
> >
> > David Huebel
> > _______________________________________________
> > 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
>


-- 
Eduardo PICADO
MSN eduardo.picado at gmail.com
portable +33 6 32 47 10 10 (NEW)
fix +33 1 45 03 52 54
|??<?????>??|

From eduardo.picado at gmail.com  Tue Sep 19 00:04:08 2006
From: eduardo.picado at gmail.com (Eduardo Picado)
Date: Tue, 19 Sep 2006 00:04:08 +0200
Subject: [C++-sig] Exception 'No to_python (by-value) converter found -
	'garbage collection'
In-Reply-To: <87odtcgb03.fsf@pereiro.luannocracy.com>
References: <6c75ff8f0609130956s6b0e7c9m9fdb8d44840fd26b@mail.gmail.com>
	<87odtcgb03.fsf@pereiro.luannocracy.com>
Message-ID: <6c75ff8f0609181504h7668cff9gaddcf653ed39221@mail.gmail.com>

On 9/18/06, David Abrahams <dave at boost-consulting.com> wrote:
> "Eduardo Picado" <eduardo.picado at gmail.com> writes:
>
> > Hi,
> >
> > Im writing the following kind of python code using a C++ wrapped library:
> >
> >
> >
> > system = TASRSystem.getASRSystem()
> > system.setParameter(TASRSystem.Option.asrServersListenPort, 32000)
> >
> > try:
> >       system.initialize(1, 3000)
> > except Exception, e:
> >       sys.stderr.write("[E] : " +  e.__class__.__name__ + str(e) )
> > print "ok"
> >
> >
> >
> > Running it I get:
> >
> >
> >
> > ok
> > Exception exceptions.TypeError: 'No to_python (by-value) converter
> > found for C++ type:
> > class TASRSystem' in 'garbage collection' ignored
> > Fatal Python error: unexpected exception during garbage collection
>
> This surely means you've wrapped a C++ function that returns a
> TASRSystem and that function is getting invoked from some Python
> class' __del__ method.
>

I suppose the __del__ method is gets invoked only by the garbage collector.

It's strange but I've got the reported error only when running my python script.
When typing in the comment line I get only:
exceptions.TypeError: 'No to_python (by-value) converter found for C++ type:
class TASRSystem

I think I have solved the problem by removing the "boost::noncopyable"
and commenting a private constructor in the header, to be able to
compile the module.


> > The declaration of TASRSystem is like:
> >
> > BOOST_PYTHON_MEMBER_FUNCTION_OVERLOADS(TASRSystem_initialize_overloads_0_2,
> > initialize, 0, 2)
> >
> > class_< TASRSystem, boost::noncopyable >("TASRSystem", no_init)
> >       .def("getASRSystem", &TASRSystem::getASRSystem,
> > return_value_policy<reference_existing_object>())
> >       .def("initialize", &TASRSystem::initialize,
> > TASRSystem_initialize_overloads_0_2())
> >       .staticmethod("getASRSystem")
> >
> >
> > - TASRSystem.getASRSystem() is an static method, that creates and
> > returns an instance of an ASRSystem. Seems ok.
> > I've tried to use with_custodian_and_ward_postcall <1, 0>, to ensure
> > that the lifetime of the created instance is as long
> > as that of the Python object, but it didn't worked - gives "argument
> > index out of range".
>
> That's because the fewest number of arguments wrapped by your
> invocation of BOOST_PYTHON_MEMBER_FUNCTION_OVERLOADS is zero.


Ok, it really makes senses.


>
> --
> Dave Abrahams
> Boost Consulting
> www.boost-consulting.com


Dave, thanks very much for your answers, I'm learning a lot with them.
And BPL is really very good.


Eduardo


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


-- 
Eduardo PICADO
MSN eduardo.picado at gmail.com
portable +33 6 32 47 10 10 (NEW)
fix +33 1 45 03 52 54
|??<?????>??|

From dave at boost-consulting.com  Tue Sep 19 00:23:23 2006
From: dave at boost-consulting.com (David Abrahams)
Date: Mon, 18 Sep 2006 18:23:23 -0400
Subject: [C++-sig] implicitly_convertible or order of init's problem
References: <7465b6170609181228g42d2625bk2c09c467a12d9e7e@mail.gmail.com>
Message-ID: <877j00ervo.fsf@pereiro.luannocracy.com>

"Roman Yakovenko" <roman.yakovenko at gmail.com> writes:

> If I comment out "bp::implicitly_convertible< vector const &, float_vector >();"
> than "ic_ext.float_vector( 5.0 )" expression pass.
>
> It could be nice to understand why this happens. I attached small example
> that reproduce the problem.( I hope this time I did it right :-) )

Please try the following patch to boost/python/converter/implicit.hpp.

-------------- next part --------------
A non-text attachment was scrubbed...
Name: implicit.patch
Type: text/x-patch
Size: 555 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/cplusplus-sig/attachments/20060918/5b7a57d4/attachment.bin>
-------------- next part --------------

-- 
Dave Abrahams
Boost Consulting
www.boost-consulting.com

From lloyd at randombit.net  Tue Sep 19 01:51:08 2006
From: lloyd at randombit.net (Jack Lloyd)
Date: Mon, 18 Sep 2006 19:51:08 -0400
Subject: [C++-sig] Returning a Python object to C++ (is get_owner the
	right way?)
In-Reply-To: <87ejuhq29d.fsf@pereiro.luannocracy.com>
References: <20060901202535.GC28064@randombit.net>
	<87ejuhq29d.fsf@pereiro.luannocracy.com>
Message-ID: <20060918235108.GG19138@randombit.net>

On Mon, Sep 11, 2006 at 09:53:18PM -0400, David Abrahams wrote:
> Jack Lloyd <lloyd at randombit.net> writes:
> 
> > I have a scenario where I'm wrapping a virtual base class 
> 
> You mean a base class with virtual functions, or an abstract base
> class?  I'm pretty sure you don't mean
> 
>         class a : public virtual b { ... };
> 
> That's a virtual base class.  Pedantry aside...

You are of course correct (virtual bases are involved in this
hierarchy, but what I meant was a class with pure virtual member
functions).

> Not unless you're willing to change the clone member function to
> return shared_ptr<Base> instead of a raw pointer.

That might well be feasible - as it is I'm having to write wrapper
classes all over the place (an exaggeration, 4 or 5 total, as I'm only
exposing objects fairly high up the inheritence tree here), so if I
have to replace my current wrappers with others, that's fine if it
gets me a little closer to easy to read/maintainable code. I'm not too
familiar with Boost's smart pointers, so I'm afraid I'm not
immediately seeing what else I would need to do for this to work. Has
this technique been discussed on the list before?

-Jack


From shiqinfeng at gmail.com  Tue Sep 19 02:17:18 2006
From: shiqinfeng at gmail.com (Qinfeng(Javen) Shi )
Date: Tue, 19 Sep 2006 10:17:18 +1000
Subject: [C++-sig] how to pass list to my c++ function which need an
	argument of array
Message-ID: <fe76f74b0609181717k509e8905xd2ef83200d007ac8@mail.gmail.com>

Dear all,

I wrapped a c++ function into python by pyste.
It is like:
func1(double[] a);
main()
{
   double a[]={1,2,3};
   func1(a);
}
I can run main in python.
But I what directly call func1 in python as well.
Python doesn't have a type called double[], but list(or numarray) has the
similar property to double[]. So if I can pass list to func1, it will be
fine.
Does anybody know how can I pass list to func1 in python?

Any suggestion is appreciated.

Javen.

Research School of Information Sciences and Engineering
Australian National University
Locked Bag 8001
Canberra ACT 2601
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/cplusplus-sig/attachments/20060919/3708bed9/attachment.htm>

From davidhuebel at gmail.com  Tue Sep 19 02:52:41 2006
From: davidhuebel at gmail.com (David Huebel)
Date: Mon, 18 Sep 2006 19:52:41 -0500
Subject: [C++-sig] Complete confustion about using Boost.Python
In-Reply-To: <6c75ff8f0609181453hd32ec79ld84e1f800e8f9022@mail.gmail.com>
References: <200609141402.19397.noah.schwartz1@gmail.com>
	<6c75ff8f0609141145s524ebbc6p1c1c4194adf1df5e@mail.gmail.com>
	<200609141457.59649.noah.schwartz1@gmail.com>
	<50c6b7b60609180626u6e8e26c1s9eff269303087844@mail.gmail.com>
	<c9c12f860609181448y68561cd2re7b020c8a613d317@mail.gmail.com>
	<6c75ff8f0609181453hd32ec79ld84e1f800e8f9022@mail.gmail.com>
Message-ID: <50c6b7b60609181752te818ddcja2ac3a852584ca72@mail.gmail.com>

On 9/18/06, Eduardo Picado <eduardo.picado at gmail.com> wrote:
> Now I remember I've had exactly the same problem when doing my first
> extensions using BPL and bjam.

I had this problem as well.

In my case, the unusual install setup for Boost.Python contributed to
the confusion.  The usual procedure for installing a package from
source is to extract to a directory, build, install to various
(usually system-wide) locations, and then delete the directory where
the source was extracted and built.   I did not expect the build
directories to be part of a permanent install, so naturally I assumed
it was possible, normal, and desirable to build extensions without any
reference to them.  I never doubted this assumption until I learned
enough Jam to figure out that the build process depended on files that
existed only in those directories and nowhere else in the system.

> I think would be good to have at least a FAQ "How to compile
> extensions out of the example directory?"
>
> Eduardo

I think that's an excellent idea.  An alternative would be for the
Boost install process to conform to expectations and actually install
everything needed to build Boost.Python extensions, but I don't know
whether that's possible or even desirable.  I may be out of line
suggesting it; please be gentle ;-)

-David


From npk.gangadhar at gmail.com  Tue Sep 19 06:12:16 2006
From: npk.gangadhar at gmail.com (Gangadhar NPK)
Date: Tue, 19 Sep 2006 09:42:16 +0530
Subject: [C++-sig] Assert fail during deletion for an obj containing
	external object ptr
In-Reply-To: <7465b6170609180732s683004b4sd59ac1962fb676e@mail.gmail.com>
References: <61b4e70f0609180720n112e03b9x47706e8d83086261@mail.gmail.com>
	<7465b6170609180724r5bec6b3bj45777b97370d76d3@mail.gmail.com>
	<7465b6170609180732s683004b4sd59ac1962fb676e@mail.gmail.com>
Message-ID: <61b4e70f0609182112i207f4333ubf8e75736471f6e3@mail.gmail.com>

Hi Romanov,
I am trying to learn Boost.python, so I might be wrong in the way I
have written code. How, do you think should the same thing be codified
?

Thank You
Gangadhar
On 9/18/06, Roman Yakovenko <roman.yakovenko at gmail.com> wrote:
> On 9/18/06, Roman Yakovenko <roman.yakovenko at gmail.com> wrote:
> > On 9/18/06, Gangadhar NPK <npk.gangadhar at gmail.com> wrote:
> > > All,
> > > I am a beginner for Boost.Python, so please be kind on the responses.
> > > I am trying to create python wrappers for an existing library. I can't
> > > change the source of the library ( I don't even have the src to
> > > change). Also, since most of the classes are via factories, I can't
> > > try to derive from the base classes, so am providing the wrappers by
> > > containership. The python wrappers will contain the pointers to the
> > > 3rd party libs.
> > >
> > > When I do this, I am able to get the interface to the third party
> > > library's base interface, but during destruction (during the
> > > interpretor exit), I get this assert failure:
> > > ---------------------------
> > > Microsoft Visual C++ Debug Library
> > > ---------------------------
> > > Debug Assertion Failed!
> > > Program: c:\Python24\python.exe
> > > File: dbgdel.cpp
> > > Line: 51
> > > Expression: _BLOCK_TYPE_IS_VALID(pHead->nBlockUse)
> > > For information on how your program can cause an assertion
> > > failure, see the Visual C++ documentation on asserts.
> > > (Press Retry to debug the application)
> > > ---------------------------
> > > Abort   Retry   Ignore
> > > ---------------------------
> > > I understand that there might be a double delete or invalid deletion
> > > that is happening, but I can't point exactly where. The code that
> > > generates is attached as a CPP file.
> > > This must be something basic that I am missing, but I can't figure out
> > > from the boost.python documentation or the wiki. Can someone please
> > > point at what I am doing wrong?
> > >
> > > Thank You
> > > Gangadhar
> > > P.S:Also, ignore the fact that the 3rd party lib's static function is
> > > not exported in the Wrapper class as a member function. That is
> > > something that will be changed later.
> >
> > This is not a bug. If you want to debug your extension you will need
> > Python built in "debug" mode.
>
> > Try to compile your extension in release mode and it should work.
> > ( At least the example you attached )
>
> I am not sure about last 2 sentence. The wrapper class is very very strange.
>
>
> --
> Roman Yakovenko
> C++ Python language binding
> http://www.language-binding.net/
> _______________________________________________
> C++-sig mailing list
> C++-sig at python.org
> http://mail.python.org/mailman/listinfo/c++-sig
>


From roman.yakovenko at gmail.com  Tue Sep 19 07:04:50 2006
From: roman.yakovenko at gmail.com (Roman Yakovenko)
Date: Tue, 19 Sep 2006 08:04:50 +0300
Subject: [C++-sig] how to pass list to my c++ function which need an
	argument of array
In-Reply-To: <fe76f74b0609181717k509e8905xd2ef83200d007ac8@mail.gmail.com>
References: <fe76f74b0609181717k509e8905xd2ef83200d007ac8@mail.gmail.com>
Message-ID: <7465b6170609182204m4a786fc7ma5a087b8acfb351@mail.gmail.com>

On 9/19/06, Qinfeng(Javen) Shi <shiqinfeng at gmail.com> wrote:
> Dear all,
>
> I wrapped a c++ function into python by pyste.
> It is like:
> func1(double[] a);
> main()
> {
>    double a[]={1,2,3};
>    func1(a);
> }
> I can run main in python.
> But I what directly call func1 in python as well.
> Python doesn't have a type called double[], but list(or numarray) has the
> similar property to double[]. So if I can pass list to func1, it will be
> fine.
> Does anybody know how can I pass list to func1 in python?
>
> Any suggestion is appreciated.

You have few choices:
1. To wrap this function manually using boost::python::list and
convert it to array of doubles
2. To use development version of Py++ :-).
    Matthias Baas added some functionality that allows you:

    mb  =module_builder_t(...)
    func1 = mb.free_function( 'func1' )
    func1.function_transformers.add( input_array_t(0,3) )
    #( arg 0 is array of size 3 )
    Py++ will automatically generate the wrapper. You can add more than 1
    transformation.

WARNING: The API is new and under development it could be change in any moment.


-- 
Roman Yakovenko
C++ Python language binding
http://www.language-binding.net/


From roman.yakovenko at gmail.com  Tue Sep 19 07:16:38 2006
From: roman.yakovenko at gmail.com (Roman Yakovenko)
Date: Tue, 19 Sep 2006 08:16:38 +0300
Subject: [C++-sig] Assert fail during deletion for an obj containing
	external object ptr
In-Reply-To: <61b4e70f0609182112i207f4333ubf8e75736471f6e3@mail.gmail.com>
References: <61b4e70f0609180720n112e03b9x47706e8d83086261@mail.gmail.com>
	<7465b6170609180724r5bec6b3bj45777b97370d76d3@mail.gmail.com>
	<7465b6170609180732s683004b4sd59ac1962fb676e@mail.gmail.com>
	<61b4e70f0609182112i207f4333ubf8e75736471f6e3@mail.gmail.com>
Message-ID: <7465b6170609182216m3877d895le8bd19d02b42b452@mail.gmail.com>

On 9/19/06, Gangadhar NPK <npk.gangadhar at gmail.com> wrote:
> Hi Romanov,

My name is Roman, not Romanov.

> I am trying to learn Boost.python, so I might be wrong in the way I
> have written code. How, do you think should the same thing be codified
> ?

Try Py++ GUI.
http://language-binding.net/pyplusplus/documentation/tutorials/tutorials.html#graphical-interface

You don't have to learn any API.
This is the code Py++ generated for your ExternalLib class:

namespace bp = boost::python;

struct ExternalLib_wrapper : ExternalLib, bp::wrapper< ExternalLib > {

    ExternalLib_wrapper( )
    : ExternalLib( )
      , bp::wrapper< ExternalLib >()
    {   // Normal constructor

    }

};

BOOST_PYTHON_MODULE(pyplusplus){
    bp::class_< ExternalLib_wrapper, boost::noncopyable >(
"ExternalLib", bp::init< >() )
        .def(
            "Create"
            , &::ExternalLib::Create
                /* undefined call policies */ )
        .def(
            "Print"
            , &::ExternalLib::Print )
        .staticmethod( "Create" );
}

-- 
Roman Yakovenko
C++ Python language binding
http://www.language-binding.net/


From roman.yakovenko at gmail.com  Tue Sep 19 07:36:18 2006
From: roman.yakovenko at gmail.com (Roman Yakovenko)
Date: Tue, 19 Sep 2006 08:36:18 +0300
Subject: [C++-sig] implicitly_convertible or order of init's problem
In-Reply-To: <877j00ervo.fsf@pereiro.luannocracy.com>
References: <7465b6170609181228g42d2625bk2c09c467a12d9e7e@mail.gmail.com>
	<877j00ervo.fsf@pereiro.luannocracy.com>
Message-ID: <7465b6170609182236i40d4ed8cm3355c46e59e1a429@mail.gmail.com>

On 9/19/06, David Abrahams <dave at boost-consulting.com> wrote:
> "Roman Yakovenko" <roman.yakovenko at gmail.com> writes:
>
> > If I comment out "bp::implicitly_convertible< vector const &, float_vector >();"
> > than "ic_ext.float_vector( 5.0 )" expression pass.
> >
> > It could be nice to understand why this happens. I attached small example
> > that reproduce the problem.( I hope this time I did it right :-) )
>
> Please try the following patch to boost/python/converter/implicit.hpp.

The patch corrects the problem. Thank you.


-- 
Roman Yakovenko
C++ Python language binding
http://www.language-binding.net/


From npk.gangadhar at gmail.com  Tue Sep 19 10:06:15 2006
From: npk.gangadhar at gmail.com (Gangadhar NPK)
Date: Tue, 19 Sep 2006 13:36:15 +0530
Subject: [C++-sig] Assert fail during deletion for an obj containing
	external object ptr
In-Reply-To: <7465b6170609182216m3877d895le8bd19d02b42b452@mail.gmail.com>
References: <61b4e70f0609180720n112e03b9x47706e8d83086261@mail.gmail.com>
	<7465b6170609180724r5bec6b3bj45777b97370d76d3@mail.gmail.com>
	<7465b6170609180732s683004b4sd59ac1962fb676e@mail.gmail.com>
	<61b4e70f0609182112i207f4333ubf8e75736471f6e3@mail.gmail.com>
	<7465b6170609182216m3877d895le8bd19d02b42b452@mail.gmail.com>
Message-ID: <61b4e70f0609190106o6b1c8a21n7007feec18be0302@mail.gmail.com>

Roman,
I'm sorry abt mistyping your name.
I will take a look at the py++ gui to generate the boost.python classes.

Thank You
Gangadhar

On 9/19/06, Roman Yakovenko <roman.yakovenko at gmail.com> wrote:
> On 9/19/06, Gangadhar NPK <npk.gangadhar at gmail.com> wrote:
> > Hi Romanov,
>
> My name is Roman, not Romanov.
>
> > I am trying to learn Boost.python, so I might be wrong in the way I
> > have written code. How, do you think should the same thing be codified
> > ?
>
> Try Py++ GUI.
> http://language-binding.net/pyplusplus/documentation/tutorials/tutorials.html#graphical-interface
>
> You don't have to learn any API.
> This is the code Py++ generated for your ExternalLib class:
>
> namespace bp = boost::python;
>
> struct ExternalLib_wrapper : ExternalLib, bp::wrapper< ExternalLib > {
>
>     ExternalLib_wrapper( )
>     : ExternalLib( )
>       , bp::wrapper< ExternalLib >()
>     {   // Normal constructor
>
>     }
>
> };
>
> BOOST_PYTHON_MODULE(pyplusplus){
>     bp::class_< ExternalLib_wrapper, boost::noncopyable >(
> "ExternalLib", bp::init< >() )
>         .def(
>             "Create"
>             , &::ExternalLib::Create
>                 /* undefined call policies */ )
>         .def(
>             "Print"
>             , &::ExternalLib::Print )
>         .staticmethod( "Create" );
> }
>
> --
> Roman Yakovenko
> C++ Python language binding
> http://www.language-binding.net/
> _______________________________________________
> C++-sig mailing list
> C++-sig at python.org
> http://mail.python.org/mailman/listinfo/c++-sig
>


From thor.arne.johansen at ibas.no  Tue Sep 19 16:08:29 2006
From: thor.arne.johansen at ibas.no (thor.arne.johansen at ibas.no)
Date: Tue, 19 Sep 2006 16:08:29 +0200
Subject: [C++-sig] Boost.Python and threads
Message-ID: <OFF5287AC2.17F0711D-ONC12571EE.004C7B72-C12571EE.004DAE70@ibas.no>


What is the status for threads and Boost.Python.

I wish to have multiple (Python) threads.
Each thread owns an instance of a wrapped C++ object doing extensive disk
IO.

Will Boost.Python release the GIL when executing the C++ code?
If I make a C++ wrapper which releases the GIL before executing the code,
will that be safe? (Can I be sure the internals of Boost.Python does not
call into Python Interpreter, without holding the GIL?)

Note I'm extending, and not embedding, so I will only have one Python
interpreter with multiple threads, each running C++ code wrapped with
Boost.Python.

--
Thor Arne Johansen
R&D Director
Ibas AS



From dave at boost-consulting.com  Tue Sep 19 16:23:05 2006
From: dave at boost-consulting.com (David Abrahams)
Date: Tue, 19 Sep 2006 10:23:05 -0400
Subject: [C++-sig] custom smart pointers
References: <7465b6170608291048y6e2010d4ofa84fbd5760da4d3@mail.gmail.com>
	<uac5lvczx.fsf@boost-consulting.com>
	<7465b6170608311358s1c02c6c9hc26222fae8cfdef6@mail.gmail.com>
	<878xkppxyq.fsf@pereiro.luannocracy.com>
	<7465b6170609121355q42dc2dc7m526e6eaf5c3344ff@mail.gmail.com>
	<87wt88uf76.fsf@pereiro.luannocracy.com>
	<7465b6170609131210h30285ef8pd4c8b7e624ff82f3@mail.gmail.com>
	<87venkeuxi.fsf@pereiro.luannocracy.com>
	<7465b6170609181426g3d5b115i1a254d8a6dc70884@mail.gmail.com>
Message-ID: <87bqpckkae.fsf@pereiro.luannocracy.com>

"Roman Yakovenko" <roman.yakovenko at gmail.com> writes:

> On 9/19/06, David Abrahams <dave at boost-consulting.com> wrote:
>> "Roman Yakovenko" <roman.yakovenko at gmail.com> writes:
>>
>> AFAICT this example doesn't work as packaged (the Jamroot doesn't look for
>> test.py).
>
> It is possible. I don't use bjam. It is much easier to use scons.

It would be a lot nicer for those giving you support if you could make
the effort to be sure that your test cases actually work as intended.

> ( At least I don't have to define twice the environment of the project )

Sorry, I don't know what that means.  Could you explain?

-- 
Dave Abrahams
Boost Consulting
www.boost-consulting.com



From dave at boost-consulting.com  Tue Sep 19 16:26:07 2006
From: dave at boost-consulting.com (David Abrahams)
Date: Tue, 19 Sep 2006 10:26:07 -0400
Subject: [C++-sig] Exception 'No to_python (by-value) converter found -
	'garbage collection'
References: <6c75ff8f0609130956s6b0e7c9m9fdb8d44840fd26b@mail.gmail.com>
	<87odtcgb03.fsf@pereiro.luannocracy.com>
	<6c75ff8f0609181504h7668cff9gaddcf653ed39221@mail.gmail.com>
Message-ID: <873baokk5c.fsf@pereiro.luannocracy.com>

"Eduardo Picado" <eduardo.picado at gmail.com> writes:

>> > Running it I get:
>> >
>> >
>> >
>> > ok
>> > Exception exceptions.TypeError: 'No to_python (by-value) converter
>> > found for C++ type:
>> > class TASRSystem' in 'garbage collection' ignored
>> > Fatal Python error: unexpected exception during garbage collection
>>
>> This surely means you've wrapped a C++ function that returns a
>> TASRSystem and that function is getting invoked from some Python
>> class' __del__ method.
>>
>
> I suppose the __del__ method is gets invoked only by the garbage collector.
>
> It's strange but I've got the reported error only when running my python script.
> When typing in the comment line I get only:
> exceptions.TypeError: 'No to_python (by-value) converter found for C++ type:
> class TASRSystem
>
> I think I have solved the problem by removing the "boost::noncopyable"
> and commenting a private constructor in the header, to be able to
> compile the module.

I don't understand; the original problem was a runtime error.  That
implies you were already able to compile.

-- 
Dave Abrahams
Boost Consulting
www.boost-consulting.com



From dave at boost-consulting.com  Tue Sep 19 16:31:59 2006
From: dave at boost-consulting.com (David Abrahams)
Date: Tue, 19 Sep 2006 10:31:59 -0400
Subject: [C++-sig] Complete confustion about using Boost.Python
References: <200609141402.19397.noah.schwartz1@gmail.com>
	<6c75ff8f0609141145s524ebbc6p1c1c4194adf1df5e@mail.gmail.com>
	<200609141457.59649.noah.schwartz1@gmail.com>
	<50c6b7b60609180626u6e8e26c1s9eff269303087844@mail.gmail.com>
	<c9c12f860609181448y68561cd2re7b020c8a613d317@mail.gmail.com>
	<6c75ff8f0609181453hd32ec79ld84e1f800e8f9022@mail.gmail.com>
	<50c6b7b60609181752te818ddcja2ac3a852584ca72@mail.gmail.com>
Message-ID: <87lkogj5b4.fsf@pereiro.luannocracy.com>

"David Huebel" <davidhuebel at gmail.com> writes:

> On 9/18/06, Eduardo Picado <eduardo.picado at gmail.com> wrote:
>> Now I remember I've had exactly the same problem when doing my first
>> extensions using BPL and bjam.
>
> I had this problem as well.
>
> In my case, the unusual install setup for Boost.Python contributed to
> the confusion.  The usual procedure for installing a package from
> source is to extract to a directory, build, install to various
> (usually system-wide) locations, and then delete the directory where
> the source was extracted and built.   I did not expect the build
> directories to be part of a permanent install

If you use bjam --install, they are not.

> , so naturally I assumed it was possible, normal, and desirable to
> build extensions without any reference to them.  I never doubted
> this assumption until I learned enough Jam to figure out that the
> build process depended on files that existed only in those
> directories and nowhere else in the system.

Which files are those?

>> I think would be good to have at least a FAQ "How to compile
>> extensions out of the example directory?"
>>
>> Eduardo
>
> I think that's an excellent idea.  An alternative would be for the
> Boost install process to conform to expectations and actually install
> everything needed to build Boost.Python extensions, 

AFAIK, it does.

> but I don't know
> whether that's possible or even desirable.  I may be out of line
> suggesting it; please be gentle ;-)

Please be specific ;-)

-- 
Dave Abrahams
Boost Consulting
www.boost-consulting.com



From dave at boost-consulting.com  Tue Sep 19 16:29:30 2006
From: dave at boost-consulting.com (David Abrahams)
Date: Tue, 19 Sep 2006 10:29:30 -0400
Subject: [C++-sig] Complete confustion about using Boost.Python
References: <200609141402.19397.noah.schwartz1@gmail.com>
	<6c75ff8f0609141145s524ebbc6p1c1c4194adf1df5e@mail.gmail.com>
	<200609141457.59649.noah.schwartz1@gmail.com>
	<50c6b7b60609180626u6e8e26c1s9eff269303087844@mail.gmail.com>
	<c9c12f860609181448y68561cd2re7b020c8a613d317@mail.gmail.com>
Message-ID: <87wt80j5f9.fsf@pereiro.luannocracy.com>

"Noah Schwartz" <noah.schwartz1 at gmail.com> writes:

> Thanks David ;D
>
> I had eventually figured this out when I realized that I didn't have a
> build/tools/v1 directory installed anyone on my machine.
>
> This is sort of confusing and isn't at all immediately obvious (to me
> at least). Maybe the documentation should be updated?

I'm not sure what to change.  Patches welcomed.  

FWIW, the build instructions for Boost.Python are going to be
rewritten for the next release anyway.

-- 
Dave Abrahams
Boost Consulting
www.boost-consulting.com



From dave at boost-consulting.com  Tue Sep 19 16:29:50 2006
From: dave at boost-consulting.com (David Abrahams)
Date: Tue, 19 Sep 2006 10:29:50 -0400
Subject: [C++-sig] Complete confustion about using Boost.Python
References: <200609141402.19397.noah.schwartz1@gmail.com>
	<6c75ff8f0609141145s524ebbc6p1c1c4194adf1df5e@mail.gmail.com>
	<200609141457.59649.noah.schwartz1@gmail.com>
	<50c6b7b60609180626u6e8e26c1s9eff269303087844@mail.gmail.com>
	<c9c12f860609181448y68561cd2re7b020c8a613d317@mail.gmail.com>
	<6c75ff8f0609181453hd32ec79ld84e1f800e8f9022@mail.gmail.com>
Message-ID: <87r6y8j5ep.fsf@pereiro.luannocracy.com>

"Eduardo Picado" <eduardo.picado at gmail.com> writes:

> Now I remember I've had exactly the same problem when doing my first
> extensions using BPL and bjam.
> I think would be good to have at least a FAQ "How to compile
> extensions out of the example directory?"

Patches welcomed.

-- 
Dave Abrahams
Boost Consulting
www.boost-consulting.com



From eduardo.picado at gmail.com  Tue Sep 19 17:00:04 2006
From: eduardo.picado at gmail.com (Eduardo Picado)
Date: Tue, 19 Sep 2006 17:00:04 +0200
Subject: [C++-sig] Exception 'No to_python (by-value) converter found -
	'garbage collection'
In-Reply-To: <873baokk5c.fsf@pereiro.luannocracy.com>
References: <6c75ff8f0609130956s6b0e7c9m9fdb8d44840fd26b@mail.gmail.com>
	<87odtcgb03.fsf@pereiro.luannocracy.com>
	<6c75ff8f0609181504h7668cff9gaddcf653ed39221@mail.gmail.com>
	<873baokk5c.fsf@pereiro.luannocracy.com>
Message-ID: <6c75ff8f0609190800y2b110d84t3ee650288fdfa3d3@mail.gmail.com>

On 9/19/06, David Abrahams <dave at boost-consulting.com> wrote:
> "Eduardo Picado" <eduardo.picado at gmail.com> writes:
>
> >> > Running it I get:
> >> >
> >> >
> >> >
> >> > ok
> >> > Exception exceptions.TypeError: 'No to_python (by-value) converter
> >> > found for C++ type:
> >> > class TASRSystem' in 'garbage collection' ignored
> >> > Fatal Python error: unexpected exception during garbage collection
> >>
> >> This surely means you've wrapped a C++ function that returns a
> >> TASRSystem and that function is getting invoked from some Python
> >> class' __del__ method.
> >>
> >
> > I suppose the __del__ method is gets invoked only by the garbage collector.
> >
> > It's strange but I've got the reported error only when running my python script.
> > When typing in the comment line I get only:
> > exceptions.TypeError: 'No to_python (by-value) converter found for C++ type:
> > class TASRSystem
> >
> > I think I have solved the problem by removing the "boost::noncopyable"
> > and commenting a private constructor in the header, to be able to
> > compile the module.
>
> I don't understand; the original problem was a runtime error.  That
> implies you were already able to compile.

Exactly, it was compiling but giving a runtime error (No to_python
(by-value) converter found for C++ type: class TASRSystem).

To workaround this, I removed "boost::noncopyable", to force BP to
create a converter.
With this modification it didn't compile. So I commented a private
constructor (that I suppose is unused) in the header.



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


-- 
Eduardo PICADO
MSN eduardo.picado at gmail.com
portable +33 6 32 47 10 10 (NEW)
fix +33 1 45 03 52 54
|??<?????>??|

From s_sourceforge at nedprod.com  Tue Sep 19 17:15:09 2006
From: s_sourceforge at nedprod.com (Niall Douglas)
Date: Tue, 19 Sep 2006 16:15:09 +0100
Subject: [C++-sig] Boost.Python and threads
In-Reply-To: <OFF5287AC2.17F0711D-ONC12571EE.004C7B72-C12571EE.004DAE70@ibas.no>
References: <OFF5287AC2.17F0711D-ONC12571EE.004C7B72-C12571EE.004DAE70@ibas.no>
Message-ID: <4510178D.24702.D6C09@s_sourceforge.nedprod.com>

On 19 Sep 2006 at 16:08, thor.arne.johansen at ibas.no wrote:

> What is the status for threads and Boost.Python.

BPL needs patching.

> I wish to have multiple (Python) threads.
> Each thread owns an instance of a wrapped C++ object doing extensive disk
> IO.
> 
> Will Boost.Python release the GIL when executing the C++ code?
> If I make a C++ wrapper which releases the GIL before executing the code,
> will that be safe? (Can I be sure the internals of Boost.Python does not
> call into Python Interpreter, without holding the GIL?)
> 
> Note I'm extending, and not embedding, so I will only have one Python
> interpreter with multiple threads, each running C++ code wrapped with
> Boost.Python.

Search the list archives. You'll find a patch integrating the 
necessary code, though you'll have to update it to match your 
threading system. Use TnFOX from SVN (http://www.nedprod.com/TnFOX/) 
as an example.

Cheers,
Niall





From abierbaum at gmail.com  Tue Sep 19 18:26:14 2006
From: abierbaum at gmail.com (Allen Bierbaum)
Date: Tue, 19 Sep 2006 11:26:14 -0500
Subject: [C++-sig] pointer_holder and pure virtual base classes
Message-ID: <bea9869d0609190926q3880b6f7jcebd82f0843a030b@mail.gmail.com>

This is a followup to my efforts in the thread here:
http://thread.gmane.org/gmane.comp.python.c++/10125/focus=10125

I seemed to have everything working but I just ran into a new problem.
 A simplified summary is:

class Base
{
public:
   int getSomething() = 0;
};

class DerivedA : public Base
{
public:
   int getSomething() { return 1; }
}

class DerivedB : public DerivedA
{
public:
   int getSomething() { return 1; }
}

I am wrapping these classes with a held_type of RefPtr<classType>
except for the Base class because it is pure virtual.  (this is a
custom ref counting pointer; see thread above for details)

I thought everything was good until I tried to wrap a factory class.
The class can return any class type but it returns it through a
pointer to the Base class like this:

class Store
{
public:
   RefPtr<Base>  getObject(int id);
};

The problem wrapping this is that since I do not have a held_type
registered for the base type as RefPtr<Base>, there is no to_python
(by-value) converter found for C++ type: RefPtr<Base>.

As an attempt to fix this I set the held type for Base to
RefPtr<Base>, but this fails because pointer_holder.hpp line 179 needs
to be able to alloate an object of the pointee type.  Since it is pure
virtual the compiler fails here.

So my question is, when dealing with a polymorphic hierarchy like this
is is possible to use a smart pointer as a held type for the most
basic types?  (ie. the pure virtual base).  If not, does anyone have
suggested workarounds for me to try?

Thanks,
Allen


Note: One simplification I made above is that there is another pointer
type in the mix.  So getObject is actually returning a BasePtr that
boost.python knows how to convert to a RefPtr<BasePtr>.  See the
previous thread for all the details.


From roman.yakovenko at gmail.com  Tue Sep 19 19:41:18 2006
From: roman.yakovenko at gmail.com (Roman Yakovenko)
Date: Tue, 19 Sep 2006 20:41:18 +0300
Subject: [C++-sig] pointer_holder and pure virtual base classes
In-Reply-To: <bea9869d0609190926q3880b6f7jcebd82f0843a030b@mail.gmail.com>
References: <bea9869d0609190926q3880b6f7jcebd82f0843a030b@mail.gmail.com>
Message-ID: <7465b6170609191041v52afa38fqba99519cbdc880af@mail.gmail.com>

On 9/19/06, Allen Bierbaum <abierbaum at gmail.com> wrote:
> This is a followup to my efforts in the thread here:
> http://thread.gmane.org/gmane.comp.python.c++/10125/focus=10125
>
> I seemed to have everything working but I just ran into a new problem.
>  A simplified summary is:
>
> class Base
> {
> public:
>    int getSomething() = 0;
> };
>
> class DerivedA : public Base
> {
> public:
>    int getSomething() { return 1; }
> }
>
> class DerivedB : public DerivedA
> {
> public:
>    int getSomething() { return 1; }
> }
>
> I am wrapping these classes with a held_type of RefPtr<classType>
> except for the Base class because it is pure virtual.  (this is a
> custom ref counting pointer; see thread above for details)
>
> I thought everything was good until I tried to wrap a factory class.
> The class can return any class type but it returns it through a
> pointer to the Base class like this:
>
> class Store
> {
> public:
>    RefPtr<Base>  getObject(int id);
> };
>
> The problem wrapping this is that since I do not have a held_type
> registered for the base type as RefPtr<Base>, there is no to_python
> (by-value) converter found for C++ type: RefPtr<Base>.
>
> As an attempt to fix this I set the held type for Base to
> RefPtr<Base>, but this fails because pointer_holder.hpp line 179 needs
> to be able to alloate an object of the pointee type.  Since it is pure
> virtual the compiler fails here.
>
> So my question is, when dealing with a polymorphic hierarchy like this
> is is possible to use a smart pointer as a held type for the most
> basic types?  (ie. the pure virtual base).  If not, does anyone have
> suggested workarounds for me to try?

David Abrahams solved this problem for my case. See "custom smart pointer"
thread.

Basically you will have to set as HeldType RefPtr<BaseWrapper>, where
BaseWrapper is a wrapper class, that defines "override" for pure
virtual functions.

Your smart pointer shoud support construction of RefPtr<Base> from
RefPtr<Derived>.

-- 
Roman Yakovenko
C++ Python language binding
http://www.language-binding.net/


From roman.yakovenko at gmail.com  Tue Sep 19 19:53:25 2006
From: roman.yakovenko at gmail.com (Roman Yakovenko)
Date: Tue, 19 Sep 2006 20:53:25 +0300
Subject: [C++-sig] custom smart pointers
In-Reply-To: <87bqpckkae.fsf@pereiro.luannocracy.com>
References: <7465b6170608291048y6e2010d4ofa84fbd5760da4d3@mail.gmail.com>
	<uac5lvczx.fsf@boost-consulting.com>
	<7465b6170608311358s1c02c6c9hc26222fae8cfdef6@mail.gmail.com>
	<878xkppxyq.fsf@pereiro.luannocracy.com>
	<7465b6170609121355q42dc2dc7m526e6eaf5c3344ff@mail.gmail.com>
	<87wt88uf76.fsf@pereiro.luannocracy.com>
	<7465b6170609131210h30285ef8pd4c8b7e624ff82f3@mail.gmail.com>
	<87venkeuxi.fsf@pereiro.luannocracy.com>
	<7465b6170609181426g3d5b115i1a254d8a6dc70884@mail.gmail.com>
	<87bqpckkae.fsf@pereiro.luannocracy.com>
Message-ID: <7465b6170609191053l650d27evaa9e9a3536246044@mail.gmail.com>

On 9/19/06, David Abrahams <dave at boost-consulting.com> wrote:
> "Roman Yakovenko" <roman.yakovenko at gmail.com> writes:
>
> > On 9/19/06, David Abrahams <dave at boost-consulting.com> wrote:
> >> "Roman Yakovenko" <roman.yakovenko at gmail.com> writes:
> >>
> >> AFAICT this example doesn't work as packaged (the Jamroot doesn't look for
> >> test.py).
> >
> > It is possible. I don't use bjam. It is much easier to use scons.
>
> It would be a lot nicer for those giving you support if you could make
> the effort to be sure that your test cases actually work as intended.

Sorry, I really tried to create test that you will have nothing to do,
but it is not that easy to use bjam. I think, I will prepare a
template, that will help me and other
people to report problems.

> > ( At least I don't have to define twice the environment of the project )
>
> Sorry, I don't know what that means.  Could you explain?

Using Python I can define "environment.py" file, where I can define
all relevant
settings: Boost.Python and Python header files directory, compilation
settings for
bindings and for the original project. Then I can reuse these configuration from
Py++ and from scons. It is very easy to setup the script on different machines
and different users.

Example:
http://svn.sourceforge.net/viewvc/pygccxml/pyplusplus_dev/environment.py?view=markup
http://svn.sourceforge.net/viewvc/pygccxml/pyplusplus_dev/examples/pyboost_dev/sconstruct?view=markup

-- 
Roman Yakovenko
C++ Python language binding
http://www.language-binding.net/


From lakin.wecker at gmail.com  Tue Sep 19 20:56:23 2006
From: lakin.wecker at gmail.com (Lakin Wecker)
Date: Tue, 19 Sep 2006 12:56:23 -0600
Subject: [C++-sig] Smart Pointers
Message-ID: <f27c5670609191156kc6a3e51uf1e51e6bdcc5d7bd@mail.gmail.com>

David Abrahams & boost.python/py++ contributors,

I just wanted to thank you for the time and effort you put forth helping
Roman with the custom smart pointers.  It has really helped the python-ogre
project. For those of us involved, it has saved us quite a bit of time.

I also just wanted to say kudos to everyone involved with Boost.python and
with py++.  I'm very happy with the ease that it allows wrapping C++ code,
and it keeps getting better everyday.

Cheers!

Lakin
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/cplusplus-sig/attachments/20060919/c0ec9b36/attachment.htm>

From roman.yakovenko at gmail.com  Tue Sep 19 22:59:50 2006
From: roman.yakovenko at gmail.com (Roman Yakovenko)
Date: Tue, 19 Sep 2006 23:59:50 +0300
Subject: [C++-sig] [solved]custom smart pointers
Message-ID: <7465b6170609191359ye7ee670od27fe48c3119c599@mail.gmail.com>

Hi. Few weeks ago I asked for help to solve problem with custom smart pointer.
David Abrahams helped me to solve all the problems I had. He pointed me to
the bugs in my code. Yes! There was no problem with Boost.Python, but with
my code.

I attach to this mail few files:
* smart_ptr.h - contains definition of my smart pointer
* classes.h - contains definition of base( abstract ) and derived classes
* bindings.cpp - contains code that exposes base and derived classes to Python
* test.py - test suite, defines 2 classes, one that derived from base
and another
                 one that derives from "derived" class
* sconstruct - scons build script

I hope you( Boost.Python user) will be able to learn from the code and
also to use
it to test your own smart pointer.

You can find the source code in Py++ repository:
http://svn.sourceforge.net/viewvc/pygccxml/pyplusplus_dev/unittests/smart_ptrs/

Comments, suggestions, new tests or just thoughts are welcome.

-- 
Roman Yakovenko
C++ Python language binding
http://www.language-binding.net/
-------------- next part --------------
A non-text attachment was scrubbed...
Name: bindings.cpp
Type: text/x-c++src
Size: 2412 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/cplusplus-sig/attachments/20060919/78c5cb73/attachment.cpp>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: classes.hpp
Type: text/x-c++hdr
Size: 935 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/cplusplus-sig/attachments/20060919/78c5cb73/attachment.hpp>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: sconstruct
Type: application/octet-stream
Size: 326 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/cplusplus-sig/attachments/20060919/78c5cb73/attachment.obj>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: smart_ptr.h
Type: text/x-chdr
Size: 2007 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/cplusplus-sig/attachments/20060919/78c5cb73/attachment.h>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: test.py
Type: text/x-python
Size: 1694 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/cplusplus-sig/attachments/20060919/78c5cb73/attachment.py>

From davidhuebel at gmail.com  Wed Sep 20 05:47:26 2006
From: davidhuebel at gmail.com (David Huebel)
Date: Tue, 19 Sep 2006 22:47:26 -0500
Subject: [C++-sig] Complete confustion about using Boost.Python
In-Reply-To: <87lkogj5b4.fsf@pereiro.luannocracy.com>
References: <200609141402.19397.noah.schwartz1@gmail.com>
	<6c75ff8f0609141145s524ebbc6p1c1c4194adf1df5e@mail.gmail.com>
	<200609141457.59649.noah.schwartz1@gmail.com>
	<50c6b7b60609180626u6e8e26c1s9eff269303087844@mail.gmail.com>
	<c9c12f860609181448y68561cd2re7b020c8a613d317@mail.gmail.com>
	<6c75ff8f0609181453hd32ec79ld84e1f800e8f9022@mail.gmail.com>
	<50c6b7b60609181752te818ddcja2ac3a852584ca72@mail.gmail.com>
	<87lkogj5b4.fsf@pereiro.luannocracy.com>
Message-ID: <50c6b7b60609192047o3b0a397ek9d6b886e37222f00@mail.gmail.com>

> Which files are those?

boost_1_33_1/libs/python/build/Jamfile, for one.

The tutorial Jamfile says:

# requirements and dependencies for Boost.Python extensions
<template>@boost/libs/python/build/extension
;

I extracted and built Boost in my home directory and set BOOST_ROOT to
/home/david/boost_1_33_1, so the line above directs bjam to the
directory /home/david/boost_1_33_1/libs/python/build.  If I remove
/home/david/boost_1_33_1/libs/python/build/Jamfile and try to build an
extension, I get:

/home/david/boost_1_33_1/libs/python/build/Jamfile: No such file or directory

I know this file isn't available anywhere else on the system, because
searching for all Jamfiles shows that they are all under the Boost
build tree or in my extension projects.  (The Boost include files were
installed to /usr/local/include/boost-1_33_1/boost/, and the Boost
libraries were installed to /usr/local/lib64.)  That means I can't
change BOOST_ROOT to point anywhere else, so I have to keep the build
tree around to build extensions.

>From your comments, I must be doing something wrong.  Can I do without
libs/python/build/Jamfile?  Did I miss an option that would have
installed it along with the rest of Boost?  Thanks for your patience.

Regards,
David Huebel


From undefined at pobox.com  Wed Sep 20 08:17:47 2006
From: undefined at pobox.com (Corey Wright)
Date: Wed, 20 Sep 2006 01:17:47 -0500
Subject: [C++-sig] Boost.Python, MinGW, & ImportError
Message-ID: <20060920011747.c857b35d.undefined@pobox.com>

this is in reference to the following postings to c++-sig and boost-users:
1. http://lists.boost.org/boost-users/2006/02/17367.php
2. http://mail.python.org/pipermail/c++-sig/2006-September/011202.html

the summary: individuals, including myself, are seeing "ImportError:
dynamic module does not define init function (inithello)" whenever trying
to import a module compiled with mingw against boost.python, specifically
the python examples and tutorial distributed with boost.

the symptom: g++ fails to properly link whenever a null library search path
(ie -L"") is specified (because maybe it stops parsing arguments there?).

the cause: MINGW_BIN_DIRECTORY is null

the short answer (for those newbies just trying to get through the
tutorial/example and don't care to dive into the details of boost.build):
bjam -sMINGW_BIN_DIRECTORY=C:\MinGW\bin (or where ever your mingw
installation is)

the longer answer: see attached patch

the explanation:

A. in gcc-tools.jam:
 1. the gcc-Link-action contains -L"$(STDLIBPATH:T)"
 2. STDLIBPATH is set based on following commands:
   a. flags gcc GCC_STDLIB_DIRECTORY : $(GCC_STDLIB_DIRECTORY) ;
   b. GCC_STDLIB_DIRECTORY ?= $(GCC_ROOT_DIRECTORY)$(SLASH)lib ;
   c. flags gcc STDLIBPATH : $(GCC_STDLIB_DIRECTORY) ;

but when using mingw, the values in gcc-tools.jam are set...

B. in mingw-tools.jam
 1. local GCC_BIN_DIRECTORY = $(MINGW_BIN_DIRECTORY) ;
 2. local GCC_STDLIB_DIRECTORY = $(MINGW_STDLIB_DIRECTORY) ;
 3. flags mingw STDLIBPATH : $(GCC_BIN_DIRECTORY) ;

so for mingw, STDLIBPATH is always composed of GCC_BIN_DIRECTORY (which is
really MINGW_BIN_DIRECTORY).

if you set MINGW_STDLIB_DIRECTORY, but not MINGW_BIN_DIRECTORY, you end up
with -L"mingw_stdlib_directory" -L"".  you must set MINGW_BIN_DIRECTORY to
get rid of the null string or insure the null string in MINGW_BIN_DIRECTORY
never gets added to STDLIBPATH.

i successfully compiled boost.python without needing to set
MINGW_BIN_DIRECTORY because the mingw tools are in my path and everything
"just works".  but if you try to compile python extensions (or at least the
tutorial and examples), then things start mysteriously failing.

i haven't had time to fully test my patch by building all of boost with it
applied, but i'm currently in the process and last i looked bjam had
compiled date and filesystem libraries, and was working on iostreams.

thanks for boost.python and the mingw support (it feels liberating to
compile free software, gratis & libre, with free software; and we'll
overlook the operating system ;-).
-- 
undefined at pobox.com
-------------- next part --------------
An embedded and charset-unspecified text was scrubbed...
Name: mingw-python-importerror.patch
URL: <http://mail.python.org/pipermail/cplusplus-sig/attachments/20060920/3ff31ce5/attachment.txt>

From eduardo.picado at gmail.com  Wed Sep 20 11:41:15 2006
From: eduardo.picado at gmail.com (Eduardo Picado)
Date: Wed, 20 Sep 2006 11:41:15 +0200
Subject: [C++-sig] Complete confustion about using Boost.Python
In-Reply-To: <50c6b7b60609192047o3b0a397ek9d6b886e37222f00@mail.gmail.com>
References: <200609141402.19397.noah.schwartz1@gmail.com>
	<6c75ff8f0609141145s524ebbc6p1c1c4194adf1df5e@mail.gmail.com>
	<200609141457.59649.noah.schwartz1@gmail.com>
	<50c6b7b60609180626u6e8e26c1s9eff269303087844@mail.gmail.com>
	<c9c12f860609181448y68561cd2re7b020c8a613d317@mail.gmail.com>
	<6c75ff8f0609181453hd32ec79ld84e1f800e8f9022@mail.gmail.com>
	<50c6b7b60609181752te818ddcja2ac3a852584ca72@mail.gmail.com>
	<87lkogj5b4.fsf@pereiro.luannocracy.com>
	<50c6b7b60609192047o3b0a397ek9d6b886e37222f00@mail.gmail.com>
Message-ID: <6c75ff8f0609200241qf61fcb7n3f2780dcf517e698@mail.gmail.com>

I understand that these instructions about building extensions are pretty
clear and complete:
"Building your Extension Module"
http://www.boost.org/libs/python/doc/building.html#building_ext


What may be giving trouble is the following use case:
People read the tutorial "Building Hello World" and successful build the
example in
C:\boost_1_33_1\libs\python\example\tutorial.

Then, they move this directory out of boost hierarchy to create their own
extension,
configure "Jamfile" and "Jamrules" as described in the tutorial but cannot
compile.
Gives: "Unable to load Boost.Build: could not find "boost-build.jam"

This is because "boost-build.jam" is missing in the directory -now out of
the boost folder-
and it is not mentioned in the tutorial page.

So, an obvious solution is to put "boost-build.jam" in the tutorial folder,
and comment it in the tutorial.


Eduardo



On 9/20/06, David Huebel <davidhuebel at gmail.com> wrote:
> > Which files are those?
>
> boost_1_33_1/libs/python/build/Jamfile, for one.
>
> The tutorial Jamfile says:
>
> # requirements and dependencies for Boost.Python extensions
> <template>@boost/libs/python/build/extension
> ;
>
> I extracted and built Boost in my home directory and set BOOST_ROOT to
> /home/david/boost_1_33_1, so the line above directs bjam to the
> directory /home/david/boost_1_33_1/libs/python/build.  If I remove
> /home/david/boost_1_33_1/libs/python/build/Jamfile and try to build an
> extension, I get:
>
> /home/david/boost_1_33_1/libs/python/build/Jamfile: No such file or
directory
>
> I know this file isn't available anywhere else on the system, because
> searching for all Jamfiles shows that they are all under the Boost
> build tree or in my extension projects.  (The Boost include files were
> installed to /usr/local/include/boost-1_33_1/boost/, and the Boost
> libraries were installed to /usr/local/lib64.)  That means I can't
> change BOOST_ROOT to point anywhere else, so I have to keep the build
> tree around to build extensions.
>
> >From your comments, I must be doing something wrong.  Can I do without
> libs/python/build/Jamfile?  Did I miss an option that would have
> installed it along with the rest of Boost?  Thanks for your patience.
>
> Regards,
> David Huebel
> _______________________________________________
> C++-sig mailing list
> C++-sig at python.org
> http://mail.python.org/mailman/listinfo/c++-sig
>


-- 
Eduardo PICADO
MSN eduardo.picado at gmail.com
portable +33 6 32 47 10 10 (NEW)
fix +33 1 45 03 52 54
|??<?????>??|
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/cplusplus-sig/attachments/20060920/70a627fc/attachment.htm>

From dave at boost-consulting.com  Wed Sep 20 18:01:46 2006
From: dave at boost-consulting.com (David Abrahams)
Date: Wed, 20 Sep 2006 12:01:46 -0400
Subject: [C++-sig] Boost.Python, MinGW, & ImportError
References: <20060920011747.c857b35d.undefined@pobox.com>
Message-ID: <87ac4uedcl.fsf@pereiro.luannocracy.com>

Corey Wright <undefined at pobox.com> writes:

> thanks for boost.python and the mingw support (it feels liberating to
> compile free software, gratis & libre, with free software; and we'll
> overlook the operating system ;-).

Corey,

Thank you very much for the very complete analysis; it is _much_
appreciated!  We are switching to Boost.Build v2 for the upcoming
release and I'd be grateful if you'd do a little testing to verify
that it already behaves correctly with MinGW.  If not, we'll have to
get it fixed.

Thanks again,

-- 
Dave Abrahams
Boost Consulting
www.boost-consulting.com



From investtcartier at yahoo.com  Wed Sep 20 18:22:49 2006
From: investtcartier at yahoo.com (Kevin Jones)
Date: Wed, 20 Sep 2006 16:22:49 +0000 (UTC)
Subject: [C++-sig]
	=?utf-8?q?memory_leaks_using_the_=22manage=5Fnew=5Fobje?=
	=?utf-8?b?Y3QicmV0dXJuCXBvbGljeS4uLg==?=
References: <20060914185901.20076.qmail@web55507.mail.re4.yahoo.com>
	<7465b6170609141211u31957f7eif607f42d35396ef4@mail.gmail.com>
Message-ID: <loom.20060920T181421-418@post.gmane.org>

Roman Yakovenko <roman.yakovenko <at> gmail.com> writes:

> 
> On 9/14/06, Kevin Jones <investtcartier <at> yahoo.com> wrote:
> > Hi,
> >
> > I am using the manage_new_object return policy to wrap
> > "newed" C++ object in an effort to conserve memory.
> >
> > Even though I am deleting the returned object in
> > Python and invoking the Python garbage collector I
> > still have a leak. This is a long-running systems
> > adminstration app the read the contents of a very
> > large database.
> >
> > Any suggestions on how to make manage_new_object
> > behave correctly?
> 
> May be I can help. I am just going through "same" process. It is up to
> you to prove
> that boost.python does not free the memory. So if you create small test case
> ( one cpp, one Python script, Jamfile ) that reproduce the error, I
> think you will get the answer\solution to your problem pretty quick.
> 

Roman,

It turns out that the leak is in the return_by_value policy and *not* the
manage_new_object policy.
I am converting a "newed" C++ instance to a Python object and the C++ instance
gets dropped on the floor (I.e. - not "deleted").

The sample source code is as follows in boosttest.py and BoostTest.cpp:
Note: The name of the generated DLL is _boosttest.dll in my example.

Kevin

boosttest.py:
-------------

import _boosttest
import os, sys, gc, random
import msvcrt

for i in range(100000):
    x = _boosttest.make(random.randint(1,1000))
    val = x[0]
    del x
    del val
    gc.collect()


BoostTest.cpp:

#include <vector>
#include "boostpython.h"

class BoostTest
{
    public:
        int getSize(){return v.size();}
        void append(int x){v.push_back(x);}
        int getItem(int i){return v[i];}
    private:
        std::vector<int> v;
};

// Factory function for free store instances of BoostTest
BoostTest* make(int n)
{
    BoostTest* b = new BoostTest();
    for(int i=0;i<n;i++)
    {
        b->append(i);
    }
    return b;
}

struct BoostTest_to_p
{
    static PyObject* convert(BoostTest* x);
};

PyObject* BoostTest_to_p::convert(BoostTest* x)
{
    PyObject* list = PyList_New(0);
    if(list == NULL) return Py_None;
    const int n = x->getSize();
    for(int i=0;i<n;i++)
    {
        PyObject* item = PyInt_FromLong(x->getItem(i));
        if( item == NULL || item == Py_None )
        {
            Py_DECREF(list);
            list = Py_None;
            break;
        }
        PyList_Append(list,item);
    }
    return list;
}

BOOST_PYTHON_MODULE(_boosttest)
{
    to_python_converter<BoostTest*, BoostTest_to_p>();
    def("make", make, return_value_policy<return_by_value>());

	class_<BoostTest>("BoostTest",no_init)
		.def("getSize", &BoostTest::getSize)
		.def("append", &BoostTest::append)
        .def("__getitem__",&BoostTest::getItem);
}





From roman.yakovenko at gmail.com  Wed Sep 20 20:57:33 2006
From: roman.yakovenko at gmail.com (Roman Yakovenko)
Date: Wed, 20 Sep 2006 21:57:33 +0300
Subject: [C++-sig] memory leaks using the "manage_new_object"return
	policy...
In-Reply-To: <loom.20060920T181421-418@post.gmane.org>
References: <20060914185901.20076.qmail@web55507.mail.re4.yahoo.com>
	<7465b6170609141211u31957f7eif607f42d35396ef4@mail.gmail.com>
	<loom.20060920T181421-418@post.gmane.org>
Message-ID: <7465b6170609201157p4f4c2816s67a0fc4a7c3c4f38@mail.gmail.com>

On 9/20/06, Kevin Jones <investtcartier at yahoo.com> wrote:
> It turns out that the leak is in the return_by_value policy and *not* the
> manage_new_object policy.

It seems to me that youare doing something very strange here

> I am converting a "newed" C++ instance to a Python object and the C++ instance

Boost.Python does not know about "newed" object, you did not tell this to it.

> #include <vector>
> #include "boostpython.h"
>
> class BoostTest
> {
>     public:
>         int getSize(){return v.size();}
>         void append(int x){v.push_back(x);}
>         int getItem(int i){return v[i];}
>     private:
>         std::vector<int> v;
> };
>
> // Factory function for free store instances of BoostTest
> BoostTest* make(int n)
> {
>     BoostTest* b = new BoostTest();
>     for(int i=0;i<n;i++)
>     {
>         b->append(i);
>     }
>     return b;
> }
>
> struct BoostTest_to_p
> {
>     static PyObject* convert(BoostTest* x);
> };
>
> PyObject* BoostTest_to_p::convert(BoostTest* x)
> {
>     PyObject* list = PyList_New(0);
>     if(list == NULL) return Py_None;
>     const int n = x->getSize();
>     for(int i=0;i<n;i++)
>     {
>         PyObject* item = PyInt_FromLong(x->getItem(i));
>         if( item == NULL || item == Py_None )
>         {
>             Py_DECREF(list);
>             list = Py_None;
>             break;
>         }
>         PyList_Append(list,item);
>     }
>     return list;
> }

I don't know why do you need this wrapper. Instead of it change you function
return value:

std::auto_ptr< BoostTest > make(int n);

Boost.Python will do the rest.

Also I don't understand why and when you expected Boost.Python to delete x?
It knows nothing about its lifetime, you didn't tell that to him.
I suggest you to read one more time section in tutorials about call policies.

-- 
Roman Yakovenko
C++ Python language binding
http://www.language-binding.net/


From investtcartier at yahoo.com  Thu Sep 21 00:41:08 2006
From: investtcartier at yahoo.com (Kevin Jones)
Date: Wed, 20 Sep 2006 22:41:08 +0000 (UTC)
Subject: [C++-sig] memory leaks using the
References: <20060914185901.20076.qmail@web55507.mail.re4.yahoo.com>
	<7465b6170609141211u31957f7eif607f42d35396ef4@mail.gmail.com>
	<loom.20060920T181421-418@post.gmane.org>
	<7465b6170609201157p4f4c2816s67a0fc4a7c3c4f38@mail.gmail.com>
Message-ID: <loom.20060920T230242-912@post.gmane.org>

Roman Yakovenko <roman.yakovenko <at> gmail.com> writes:

Roman,

The documentation that I have found for boost.python is rather sparse for call
policies. Can you point me to the documentation URL?

The reason that I expected boost.python to delete x is that given T, T& and T*,

I expect a value, a reference and an object from the free store.

Boost.Python does such an amazing job of compile-time type analysis that it was
perhaps wishful thinking on my part.

Kevin




From roman.yakovenko at gmail.com  Thu Sep 21 07:11:48 2006
From: roman.yakovenko at gmail.com (Roman Yakovenko)
Date: Thu, 21 Sep 2006 08:11:48 +0300
Subject: [C++-sig] memory leaks using the
In-Reply-To: <loom.20060920T230242-912@post.gmane.org>
References: <20060914185901.20076.qmail@web55507.mail.re4.yahoo.com>
	<7465b6170609141211u31957f7eif607f42d35396ef4@mail.gmail.com>
	<loom.20060920T181421-418@post.gmane.org>
	<7465b6170609201157p4f4c2816s67a0fc4a7c3c4f38@mail.gmail.com>
	<loom.20060920T230242-912@post.gmane.org>
Message-ID: <7465b6170609202211u57f7a41es58a1cdd8e3d47e67@mail.gmail.com>

On 9/21/06, Kevin Jones <investtcartier at yahoo.com> wrote:
> Roman,
>
> The documentation that I have found for boost.python is rather sparse for call
> policies. Can you point me to the documentation URL?

Py++ contains few links to help resources:
http://language-binding.net/pyplusplus/links.html#id4

Take a look on Boost.Python wiki and unit tests.

-- 
Roman Yakovenko
C++ Python language binding
http://www.language-binding.net/


From eilif at gmx.de  Thu Sep 21 14:10:39 2006
From: eilif at gmx.de (Eilif Mueller)
Date: Thu, 21 Sep 2006 14:10:39 +0200
Subject: [C++-sig] (no subject)
Message-ID: <20060921121039.39560@gmx.net>

Hi All,

maybe this is obvious to most, but...

the following code 

*****************

BOOST_PYTHON_MODULE(atest)
{
  using namespace numeric;
  using namespace boost::python;

  import_array();
  array::set_module_and_type("numpy", "ArrayType");
...

********************

works with numpy-0.9.5 but not with numpy-1.0b5 or up since it seems the 'ArrayType' attribute has been removed from the namespace.  While the module will build and import, the following messages occur when using functions that return arrays or have arrays as arguements:

***************

ImportError: No module named 'numpy' or its type 'ArrayType' did not follow the NumPy protocol

***************

It is sufficient to replace the above line with

array::set_module_and_type("numpy", "ndarray");


cheers,


Eilif

-- 
NEU: GMX DSL Sofort-Start-Set - blitzschnell ins Internet!
Echte DSL-Flatrate ab 0,- Euro* http://www.gmx.net/de/go/dsl


From shiqinfeng at gmail.com  Sat Sep 23 03:20:32 2006
From: shiqinfeng at gmail.com (Qinfeng(Javen) Shi )
Date: Sat, 23 Sep 2006 11:20:32 +1000
Subject: [C++-sig] numpy.ndarray did not match PyArrayObject *?
Message-ID: <fe76f74b0609221820i4930bef4vc39c338da6a8b39f@mail.gmail.com>

Dear All,

I wrapped a c++ code which need accept numpy.ndarray as argument.
Here numpy.version='1.0rc1'.
And to the best of my knowledge, numpy.ndarray's c type is PyArrayObject.

But why when I run it, it reports type did not match?

running result as following:
>>> b = numpy.array([1])
>>> b
array([1])
>>> hello.testArray(b)
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
Boost.Python.ArgumentError: Python argument types in
    hello.testArray(numpy.ndarray)
did not match C++ signature:
    testArray(PyArrayObject*)
>>>

Any suggestion?
Thanks!
-- 
Qinfeng(Javen) Shi

Research School of Information Sciences and Engineering
Australian National University
Locked Bag 8001
Canberra ACT 2601
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/cplusplus-sig/attachments/20060923/215f777a/attachment.htm>

From dave at boost-consulting.com  Sat Sep 23 15:56:56 2006
From: dave at boost-consulting.com (David Abrahams)
Date: Sat, 23 Sep 2006 09:56:56 -0400
Subject: [C++-sig] numpy.ndarray did not match PyArrayObject *?
References: <fe76f74b0609221820i4930bef4vc39c338da6a8b39f@mail.gmail.com>
Message-ID: <87zmcqy9cn.fsf@pereiro.luannocracy.com>

"Qinfeng(Javen) Shi " <shiqinfeng at gmail.com> writes:

> Dear All,
>
> I wrapped a c++ code which need accept numpy.ndarray as argument.
> Here numpy.version='1.0rc1'.
> And to the best of my knowledge, numpy.ndarray's c type is PyArrayObject.
>
> But why when I run it, it reports type did not match?
>
> running result as following:
>>>> b = numpy.array([1])
>>>> b
> array([1])
>>>> hello.testArray(b)
> Traceback (most recent call last):
>  File "<stdin>", line 1, in ?
> Boost.Python.ArgumentError: Python argument types in
>    hello.testArray(numpy.ndarray)
> did not match C++ signature:
>    testArray(PyArrayObject*)
>>>>
>
> Any suggestion?

You need to register an lvalue converter from Python numpy.ndarray to
PyArrayObject.

  struct extract_array_object
  {
      static PyArrayObject& execute(PyArrayObject& x)
      {
          return x;
      }
  };

  ...
  // in your module definition function
  boost::python::lvalue_from_pytype<extract_array_object,&PyArrayObject_Type>();
                                                          ^^^^^^^^^^^^^^^^^^

replace the indicated with the name of NumPy's python type object for ndarray.


-- 
Dave Abrahams
Boost Consulting
www.boost-consulting.com



From roman.yakovenko at gmail.com  Sun Sep 24 20:27:58 2006
From: roman.yakovenko at gmail.com (Roman Yakovenko)
Date: Sun, 24 Sep 2006 21:27:58 +0300
Subject: [C++-sig] Py++ - problem with split_module
	andreturn_opaque_pointer-policy
In-Reply-To: <7465b6170609120443s7ef51998p425e7b70ada260f9@mail.gmail.com>
References: <7465b6170609120414r3935fd99mf0e957ad722f9542@mail.gmail.com>
	<51F4AD92DB06E9488BBC39B7B492409E029FDCA8@vg20ex1.grp.haufemg.com>
	<7465b6170609120443s7ef51998p425e7b70ada260f9@mail.gmail.com>
Message-ID: <7465b6170609241127u3523e2aduf1c2ac4837210660@mail.gmail.com>

On 9/12/06, Roman Yakovenko <roman.yakovenko at gmail.com> wrote:
> On 9/12/06, Ganssauge, Gottfried <Gottfried.Ganssauge at haufe.de> wrote:
> > What do you think about a code generator more "knowledgeable" about
> > return_value_policy?
> > One that could see "hey, I'm being ordered to return_opaque_pointer" and
> > then would call that macro wherever neccessary?
> > In the code generator it could go like:
> >   if calling_policy is return_value_policy(return_opaque_pointer):
> >     opaque_base = remove_pointer (current_decl.return_value.type)
> >     add_declaration_code ('BOOST_PYTHON_OPAQUE_SPECIALIZED_TYPE_ID(%s)'
> > % opaque_base)
>
> This is even better solution. I will implement it and will let you know.

Good evening. It took some time, but I finally implemented this functionality.
Can you check-out  the source code from svn and test it?

Thanks.

-- 
Roman Yakovenko
C++ Python language binding
http://www.language-binding.net/


From chris207 at comcast.net  Sun Sep 24 21:51:04 2006
From: chris207 at comcast.net (CP)
Date: Sun, 24 Sep 2006 15:51:04 -0400
Subject: [C++-sig]  Opportunities at Naughty Dog...
Message-ID: <000501c6e012$c51308f0$6500a8c0@ttidownstairs>

Dear Mr. Summers,

    I hope to become one of Naughty Dogs employees one day sadly I am not old enought yet,
but before you delete this message like your company did when I e-mailed them 1-2 months
back just listen to my question and get back to me. I want to know if there is going to be another 
Jak game coming because the game in development "Big" just doesn't look like it is for me.
So is there goning to be another Jak game? That is all I want to know. 
    When I reach college I hope to become a content manager like Sony's own Gene Strocco.
I want to study at mabe USC and go on to work at Nauhty Dog after I get my degree. 
                
                                                                                                                        Thank you for listening,
                                                                                                                                                     Chris
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/cplusplus-sig/attachments/20060924/c461b1f3/attachment.htm>

From chris207 at comcast.net  Sun Sep 24 21:55:12 2006
From: chris207 at comcast.net (CP)
Date: Sun, 24 Sep 2006 15:55:12 -0400
Subject: [C++-sig]  Opportunities at Naughty Dog...
Message-ID: <000d01c6e013$58ff13b0$6500a8c0@ttidownstairs>

Dear Mr. Summers,

    I hope to become one of Naughty Dogs employees one day sadly I am not old enought yet,
but before you delete this message like your company did when I e-mailed them 1-2 months
back just listen to my question and get back to me. I want to know if there is going to be another 
Jak game coming because the game in development "Big" just doesn't look like it is for me.
So is there goning to be another Jak game? That is all I want to know. 
    When I reach college I hope to become a content manager like Sony's own Gene Strocco.
I want to study at mabe USC and go on to work at Nauhty Dog after I get my degree. 
                
                                                                                                                        Thank you for listening,
                                                                                                                                                     Chris
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/cplusplus-sig/attachments/20060924/aafb71a4/attachment.htm>

From undefined at pobox.com  Sun Sep 24 23:01:05 2006
From: undefined at pobox.com (Corey Wright)
Date: Sun, 24 Sep 2006 16:01:05 -0500
Subject: [C++-sig] boost.python w/ boost.build v2 (was Re: Boost.Python,
 MinGW, & ImportError)
In-Reply-To: <87ac4uedcl.fsf@pereiro.luannocracy.com>
References: <20060920011747.c857b35d.undefined@pobox.com>
	<87ac4uedcl.fsf@pereiro.luannocracy.com>
Message-ID: <20060924160105.92cfcf8a.undefined@pobox.com>

wow, it's hard finding straight forward documentation on building
boost.python (including examples) with boost.build v2.  lots of practical
stuff for v1, lots of general purpose documentation on v2, but google and i
had a hard finding any step-by-step instructions for using v2.  (maybe i
was looking too hard for something specific to boost.python.)

i started with boost.python from 1.33.1 and mingw and got nowhere.  i
peeked a little bit into boost.build v2, but the only reference i could
find for mingw was in a comment.

i decided to fall back to gcc on linux (a more popular, and hopefully
better tested, platform).  i got the boost.python library and the example
python modules to build, but the modules had the "lib" prefix which caused
errors on import (as the modules' init functions did not include the lib).

i thought to try the daily tarball.  boost.python, examples, and tests
built and executed after i applied the attached patch to Jamfile.v2.

i've also attached my changes to boost-build.jam to facilitate using v2
while still allowing v1, modeled after the top-level boost-build.jam.

my user-config.jam is attached as it took me 30 minutes to an hour of
searching and reading until i had constructed it.  hopefully it'll save
somebody from the same struggle.  i read the boost.build v2 documentation,
but it was so abstract and general purpose that it didn't help me.  i
finally found http://zigzag.cs.msu.su/boost.build/wiki/BoostRegressionsV2
which is very concrete with examples, but initially i wasn't sure how
applicable it was because it's for "boost regression testing".

anyways, i'll look at boost.python on mingw sometime this week, but wanted
to feed back my patch for Jamfile.v2 ("release early, release often").

again, thanks for boost (python, build, etc).

ps i'm posting this to c++-sig as this email is specific to boost.python (i
only previously posted to boost-users because my patch resolved a problem
previously discussed there) and i'm too lazy to subscribe to boost devel.
--
undefined at pobox.com

On Wed, 20 Sep 2006 12:01:46 -0400
David Abrahams <dave at boost-consulting.com> wrote:

> Corey Wright <undefined at pobox.com> writes:
> 
> > thanks for boost.python and the mingw support (it feels liberating to
> > compile free software, gratis & libre, with free software; and we'll
> > overlook the operating system ;-).
> 
> Corey,
> 
> Thank you very much for the very complete analysis; it is _much_
> appreciated!  We are switching to Boost.Build v2 for the upcoming
> release and I'd be grateful if you'd do a little testing to verify
> that it already behaves correctly with MinGW.  If not, we'll have to
> get it fixed.
-------------- next part --------------
An embedded and charset-unspecified text was scrubbed...
Name: Jamfile.v2.patch
URL: <http://mail.python.org/pipermail/cplusplus-sig/attachments/20060924/f5f3b072/attachment.txt>
-------------- next part --------------
An embedded and charset-unspecified text was scrubbed...
Name: boost-build.jam.patch
URL: <http://mail.python.org/pipermail/cplusplus-sig/attachments/20060924/f5f3b072/attachment-0001.txt>
-------------- next part --------------
An embedded and charset-unspecified text was scrubbed...
Name: user-config.jam
URL: <http://mail.python.org/pipermail/cplusplus-sig/attachments/20060924/f5f3b072/attachment-0002.txt>

From shiqinfeng at gmail.com  Mon Sep 25 08:12:33 2006
From: shiqinfeng at gmail.com (Qinfeng(Javen) Shi )
Date: Mon, 25 Sep 2006 16:12:33 +1000
Subject: [C++-sig] Pass the numpy.ndarray to c extension
Message-ID: <fe76f74b0609242312i37eded50k134ef2500145ab5e@mail.gmail.com>

Thank for Dave Abrahams's suggestion to "numpy.ndarray did not match
PyArrayObject *" problem. I didn't use the lvalue converter before. I will
look up how to do it.

Since what I really want to do is to pass the numpy.ndarray to c extension,
so I started a new title. I find a way to pass the argument from the
webpage: http://numpy.scipy.org/numpydoc/numpy-13.html . It uses
PyArg_ParseTuple to transfer PyObject into PyArrayObject.

I implemented two functions with PyArg_ParseTuple(), but they all reported
error.

/***************c extension*********/
void testArray(PyObject * args)
{

    PyObject *input;
    PyArrayObject *array;
    if (!PyArg_ParseTuple(args, "O", &input))
           printf("getting input failed!\n");
    else
        printf("getting input sucess!\n");

    array = (PyArrayObject *) PyArray_ContiguousFromObject(input,
PyArray_INT, 0, 3);
    if (array == NULL)
        printf("getting array failed!\n");
    else
        printf("getting array sucess!\n");
}


static PyObject * trace(PyObject *self, PyObject *args)
{
PyObject *input;
PyArrayObject *array;
double sum;
int i, n;

if (!PyArg_ParseTuple(args, "O", &input))
     return NULL;
array = (PyArrayObject *)
PyArray_ContiguousFromObject(input, PyArray_DOUBLE, 2, 2);
if (array == NULL)
      return NULL;
n = array->dimensions[0];
if (n > array->dimensions[1])
n = array->dimensions[1];
sum = 0.;
for (i = 0; i < n; i++)
sum += *(double *)(array->data + i*array->strides[0] + i*array->strides[1]);
Py_DECREF(array);
return PyFloat_FromDouble(sum);
}

/************error report************/

>>> import hello
>>> import numpy
>>> c = numpy.arange(25).reshape((5,5))
>>> print c
[[ 0  1  2  3  4]
 [ 5  6  7  8  9]
 [10 11 12 13 14]
 [15 16 17 18 19]
 [20 21 22 23 24]]
>>> b = hello.trace(c,c)
>>> b
SystemError: new style getargs format but argument is not a tuple
>>> d = hello.testArray(c);
getting input failed!
>>> d
SystemError: new style getargs format but argument is not a tuple
>>>


-- 
Qinfeng(Javen) Shi

Research School of Information Sciences and Engineering
Australian National University
Locked Bag 8001
Canberra ACT 2601
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/cplusplus-sig/attachments/20060925/dcf0ba58/attachment.htm>

From shiqinfeng at gmail.com  Mon Sep 25 09:30:23 2006
From: shiqinfeng at gmail.com (Qinfeng(Javen) Shi )
Date: Mon, 25 Sep 2006 17:30:23 +1000
Subject: [C++-sig] Stuck on accessing numpy.ndarray in c extension
Message-ID: <fe76f74b0609250030o57726b02jd8781f9606191359@mail.gmail.com>

Dear *All

 *I got stuck on it on wrapping c++ code which need accept numpy.ndarray as
argument. The numpy.version I am using is '1.0rc1'.* *(My previous question
is too massy, so I revised it. Sorry if I borthered you!)

I used two ways to access numpy.ndarray, but they all reported error.
-----------------------------------------------------------------------------------
1st way:
void testArray(PyArrayObject * args)
{
}

>>> hello.testArray(b)
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
Boost.Python.ArgumentError: Python argument types in
    hello.testArray(numpy.ndarray)
did not match C++ signature:
    testArray(PyArrayObject*)
>>>
------------------------------------------------------------------------------------
2nd way:
void testArray(PyObject * args)
{
    PyObject *input;
    PyArrayObject *array;
    if (!PyArg_ParseTuple(args, "O", &input))
           printf("getting input failed!\n");
    else
        printf("getting input sucess!\n");
}

>>> import hello
>>> import numpy
>>> c = numpy.arange(25).reshape((5,5))
>>> d = hello.testArray(c);
getting input failed!
>>> d
SystemError: new style getargs format but argument is not a tuple
>>>

Any suggestion will be appreciated.
Thanks!

-- 
Qinfeng(Javen) Shi

Research School of Information Sciences and Engineering
Australian National University
Locked Bag 8001
Canberra ACT 2601
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/cplusplus-sig/attachments/20060925/93aae473/attachment.htm>

From Gottfried.Ganssauge at haufe.de  Mon Sep 25 10:43:15 2006
From: Gottfried.Ganssauge at haufe.de (Ganssauge, Gottfried)
Date: Mon, 25 Sep 2006 10:43:15 +0200
Subject: [C++-sig] Py++ - problem with
	split_moduleandreturn_opaque_pointer-policy
In-Reply-To: <7465b6170609241127u3523e2aduf1c2ac4837210660@mail.gmail.com>
Message-ID: <51F4AD92DB06E9488BBC39B7B492409E029FE4FD@vg20ex1.grp.haufemg.com>

> -----Original Message-----
> From: c++-sig-bounces at python.org 
> [mailto:c++-sig-bounces at python.org] On Behalf Of Roman Yakovenko
> Sent: Sunday, September 24, 2006 8:28 PM
> To: Development of Python/C++ integration
> Subject: Re: [C++-sig] Py++ - problem with 
> split_moduleandreturn_opaque_pointer-policy
> 
> On 9/12/06, Roman Yakovenko <roman.yakovenko at gmail.com> wrote:
> > On 9/12/06, Ganssauge, Gottfried 
> <Gottfried.Ganssauge at haufe.de> wrote:
> > > What do you think about a code generator more 
> "knowledgeable" about 
> > > return_value_policy?
> > > One that could see "hey, I'm being ordered to 
> return_opaque_pointer" 
> > > and then would call that macro wherever neccessary?
> > > In the code generator it could go like:
> > >   if calling_policy is return_value_policy(return_opaque_pointer):
> > >     opaque_base = remove_pointer (current_decl.return_value.type)
> > >     add_declaration_code 
> ('BOOST_PYTHON_OPAQUE_SPECIALIZED_TYPE_ID(%s)'
> > > % opaque_base)
> >
> > This is even better solution. I will implement it and will 
> let you know.
> 
> Good evening. It took some time, but I finally implemented 
> this functionality.
> Can you check-out  the source code from svn and test it?
Well - I did.
The result still doesn't compile :-(
Do I have to call something in order to trigger your change?
I'm using the test case I sent in original message

Cheers,

Gottfried

> 
> Thanks.
> 
> --
> Roman Yakovenko
> C++ Python language binding
> http://www.language-binding.net/
> _______________________________________________
> C++-sig mailing list
> C++-sig at python.org
> http://mail.python.org/mailman/listinfo/c++-sig
> 


From meine at informatik.uni-hamburg.de  Mon Sep 25 10:30:16 2006
From: meine at informatik.uni-hamburg.de (Hans Meine)
Date: Mon, 25 Sep 2006 10:30:16 +0200
Subject: [C++-sig] Stuck on accessing numpy.ndarray in c extension
In-Reply-To: <fe76f74b0609250030o57726b02jd8781f9606191359@mail.gmail.com>
References: <fe76f74b0609250030o57726b02jd8781f9606191359@mail.gmail.com>
Message-ID: <200609251030.18079.meine@informatik.uni-hamburg.de>

On Monday, 25. September 2006 09:30, Qinfeng(Javen) Shi wrote:
> I used two ways to access numpy.ndarray, but they all reported error.
.. both ways look OK, but you only provided the raw functions, not the 
exporting code.  You should post the boost::python exporting code, too.

> Traceback (most recent call last):
>   File "<stdin>", line 1, in ?
> Boost.Python.ArgumentError: Python argument types in
>     hello.testArray(numpy.ndarray)
> did not match C++ signature:
>     testArray(PyArrayObject*)
To me, that looks like a missing (lvalue) converter.

> SystemError: new style getargs format but argument is not a tuple
Are you using raw_function for the export?  Otherwise, it naturally expects a 
tuple "args" with the array as first member. -> hello.testArray((c,))

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


From roman.yakovenko at gmail.com  Mon Sep 25 12:20:41 2006
From: roman.yakovenko at gmail.com (Roman Yakovenko)
Date: Mon, 25 Sep 2006 13:20:41 +0300
Subject: [C++-sig] Py++ - problem with
	split_moduleandreturn_opaque_pointer-policy
In-Reply-To: <51F4AD92DB06E9488BBC39B7B492409E029FE4FD@vg20ex1.grp.haufemg.com>
References: <7465b6170609241127u3523e2aduf1c2ac4837210660@mail.gmail.com>
	<51F4AD92DB06E9488BBC39B7B492409E029FE4FD@vg20ex1.grp.haufemg.com>
Message-ID: <7465b6170609250320r62668467i6e93874a9ae3f1fb@mail.gmail.com>

On 9/25/06, Ganssauge, Gottfried <Gottfried.Ganssauge at haufe.de> wrote:
> > > > What do you think about a code generator more
> > "knowledgeable" about
> > > > return_value_policy?
> > > > One that could see "hey, I'm being ordered to
> > return_opaque_pointer"
> > > > and then would call that macro wherever neccessary?
> > > > In the code generator it could go like:
> > > >   if calling_policy is return_value_policy(return_opaque_pointer):
> > > >     opaque_base = remove_pointer (current_decl.return_value.type)
> > > >     add_declaration_code
> > ('BOOST_PYTHON_OPAQUE_SPECIALIZED_TYPE_ID(%s)'
> > > > % opaque_base)
> > >
> > > This is even better solution. I will implement it and will
> > let you know.
> >
> > Good evening. It took some time, but I finally implemented
> > this functionality.
> > Can you check-out  the source code from svn and test it?
> Well - I did.
> The result still doesn't compile :-(
> Do I have to call something in order to trigger your change?
> I'm using the test case I sent in original message

Strange. I just commited unit test that test the code. All works as expected.
Can you post the error?
Also you don't have to add BOOST_PYTHON_OPAQUE_SPECIALIZED_TYPE_ID macro
Py++ will generate one.

-- 
Roman Yakovenko
C++ Python language binding
http://www.language-binding.net/


From Gottfried.Ganssauge at haufe.de  Mon Sep 25 12:38:49 2006
From: Gottfried.Ganssauge at haufe.de (Ganssauge, Gottfried)
Date: Mon, 25 Sep 2006 12:38:49 +0200
Subject: [C++-sig] Py++ - problem
	withsplit_moduleandreturn_opaque_pointer-policy
In-Reply-To: <7465b6170609250320r62668467i6e93874a9ae3f1fb@mail.gmail.com>
Message-ID: <51F4AD92DB06E9488BBC39B7B492409E029FE53A@vg20ex1.grp.haufemg.com>

> -----Original Message-----
> From: c++-sig-bounces at python.org 
> [mailto:c++-sig-bounces at python.org] On Behalf Of Roman Yakovenko
> Sent: Monday, September 25, 2006 12:21 PM
> To: Development of Python/C++ integration
> Subject: Re: [C++-sig] Py++ - problem 
> withsplit_moduleandreturn_opaque_pointer-policy
> 
> On 9/25/06, Ganssauge, Gottfried <Gottfried.Ganssauge at haufe.de> wrote:
...
> > Well - I did.
> > The result still doesn't compile :-(
> > Do I have to call something in order to trigger your change?
> > I'm using the test case I sent in original message
> 
> Strange. I just commited unit test that test the code. All 
> works as expected.
> Can you post the error?
Don't bother - it was my fault.

> Also you don't have to add 
> BOOST_PYTHON_OPAQUE_SPECIALIZED_TYPE_ID macro
To be exact: I _must_ not add ...

> Py++ will generate one.
Indeed.
It works quite nice :-)

Thanks,

Gottfried
> 
> --
> Roman Yakovenko
> C++ Python language binding
> http://www.language-binding.net/
> _______________________________________________
> C++-sig mailing list
> C++-sig at python.org
> http://mail.python.org/mailman/listinfo/c++-sig
> 


From shiqinfeng at gmail.com  Mon Sep 25 16:39:52 2006
From: shiqinfeng at gmail.com (Qinfeng(Javen) Shi )
Date: Tue, 26 Sep 2006 00:39:52 +1000
Subject: [C++-sig] Stuck on accessing numpy.ndarray in c extension
Message-ID: <fe76f74b0609250739s51267f1cx3fd57bfbcd433f19@mail.gmail.com>

No, I did not use raw_function for the export.I tried hello.testArray((c,)),
it crashed.

Then I used PyArray_Check(args) to check the input type, it crashed even
when I passed a numpy.ndarray to it.
void testArray(PyObject * args)
{
    if(!(PyArray_Check(args)))
    {
        printf("args is not list type!\n");
        return;
    }
}

>>> a = numpy.arange(25).reshape((5,5));
>>> rarray2 = hello.testArray(a,)
Segmentation fault

>>> a = numpy.arange(25).reshape((5,5));
>>> rarray2 = hello.testArray(a)
Segmentation fault

A *segmentation* *fault* occurs when a program attempts to access a memory
location that it is not allowed to access, or attempts to access a memory
location in a way that is not allowed.
It seems my PyObject * args can not neither be acessed by PyArray_Check()
nor by PyArg_ParseTuple().

I don't get it.
-- 
Qinfeng(Javen) Shi

Research School of Information Sciences and Engineering
Australian National University
Locked Bag 8001
Canberra ACT 2601
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/cplusplus-sig/attachments/20060926/195ab878/attachment.htm>

From shiqinfeng at gmail.com  Mon Sep 25 16:59:56 2006
From: shiqinfeng at gmail.com (Qinfeng(Javen) Shi )
Date: Tue, 26 Sep 2006 00:59:56 +1000
Subject: [C++-sig] Stuck on accessing numpy.ndarray in c extension
Message-ID: <fe76f74b0609250759w1a51a9ddp3d18fe391b46dbce@mail.gmail.com>

My boost::python exporting code is as following:

// Boost Includes
==============================================================
#include <boost/python.hpp>
#include <boost/cstdint.hpp>

// Includes
====================================================================
#include "hello.c"
// Using
=======================================================================
using namespace boost::python;
// Declarations
================================================================
// Module
======================================================================
BOOST_PYTHON_MODULE(hello)
{
    def("main", &main);
    def("testArray", &testArray);
}



-- 
Qinfeng(Javen) Shi

Research School of Information Sciences and Engineering
Australian National University
Locked Bag 8001
Canberra ACT 2601
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/cplusplus-sig/attachments/20060926/a9e9d4f4/attachment.htm>

From paustin at eos.ubc.ca  Mon Sep 25 20:11:14 2006
From: paustin at eos.ubc.ca (Philip Austin)
Date: Mon, 25 Sep 2006 11:11:14 -0700
Subject: [C++-sig] Stuck on accessing numpy.ndarray in c extension
In-Reply-To: <fe76f74b0609250759w1a51a9ddp3d18fe391b46dbce@mail.gmail.com>
References: <fe76f74b0609250759w1a51a9ddp3d18fe391b46dbce@mail.gmail.com>
Message-ID: <17688.7106.337767.907995@thrush.eos.ubc.ca>

Qinfeng(Javen) Shi  writes:
 > My boost::python exporting code is as following:
 > 
 > // Boost Includes
 > ==============================================================
 > #include <boost/python.hpp>
 > #include <boost/cstdint.hpp>
 > 
 > // Includes
 > ====================================================================
 > #include "hello.c"
 > // Using
 > =======================================================================
 > using namespace boost::python;
 > // Declarations
 > ================================================================
 > // Module
 > ======================================================================
 > BOOST_PYTHON_MODULE(hello)
 > {
 >     def("main", &main);
 >     def("testArray", &testArray);
 > }
 > 

You still have to initialize the function pointer array for the numpy
interface using import_array(), just as you do with a C extension.
For example, (using the boost::python::numeric):

namespace ar=boost::python::numeric;
namespace bp=boost::python;

BOOST_PYTHON_MODULE(simple_ext)
{ 
  import_array();
  ar::array::set_module_and_type("numpy", "ndarray");
  //global doc string
  bp::scope().attr("RCSID") = rcsid;
  std::string str1;
  str1="docstring for module\n";
  bp::scope().attr("__doc__") = str1.c_str();
  str1 = "Doc string for testToPython\n";
  def("testToPython", testToPython,str1.c_str());
  str1 = "Doc string for testFromPython\n";
  def("testFromPython", testFromPython,str1.c_str());
}



From shiqinfeng at gmail.com  Tue Sep 26 03:03:41 2006
From: shiqinfeng at gmail.com (Qinfeng(Javen) Shi )
Date: Tue, 26 Sep 2006 11:03:41 +1000
Subject: [C++-sig] return PyObject* failed.
Message-ID: <fe76f74b0609251803j2b08ba56q754c26c0d5c2acd2@mail.gmail.com>

Dear all,

I write a c extension function to access python list.
I can visit the list by PyObject* a, but failed when I tried to return
PyObject* to python.
Any suggestion will be appreciated.

>>> list1 = range(5);
>>> rlist = hello.testList(list1)
4 3 2 1 0 size:5
b is list type!                                   # seems right now.
>>> rlist                                         # but actually it
crashed when display the returned rlist
Segmentation fault


/**************c extension************/
static PyObject * testList(PyObject* a)
{
   if(!(PyList_Check(a)))
   {
	printf("not list type!\n");
	return NULL;
   }
   int num = PyList_Size(a);
   PyList_Reverse(a);
   // modify a.
    for(int i=0;i<num;i++)
    {
        printf("%d ",PyInt_AsLong(PyList_GetItem(a,i)));      // print
it out
    }
    printf("size:%d\n",num);

	PyObject * b = PyList_New(10);                             // create
a new list b
        if (b == NULL)
		return NULL;
    if(!(PyList_Check(b)))
     // check b
    {
	printf("b is not list type!\n");
	return NULL;
    }
    else
    {
         printf("b is list type!\n");
    }
    /**********************works untill now******************/
    return (PyObject *)b; /***********failed**/ // return b. in
fact,if return a, it also crashed
}


// Boost Includes ==============================================================
#include <boost/python.hpp>
#include <boost/cstdint.hpp>

// Includes ====================================================================
#include <hello.c>

// Using =======================================================================
using namespace boost::python;

// Declarations ================================================================
BOOST_PYTHON_OPAQUE_SPECIALIZED_TYPE_ID(PyObject)


// Module ======================================================================
BOOST_PYTHON_MODULE(hello)
{
    def("main", &main);
    def("testList", &testList, return_value_policy< return_opaque_pointer >());
}









-- 
Qinfeng(Javen) Shi

Research School of Information Sciences and Engineering
Australian National University
Locked Bag 8001
Canberra ACT 2601


From shiqinfeng at gmail.com  Tue Sep 26 03:19:00 2006
From: shiqinfeng at gmail.com (Qinfeng(Javen) Shi )
Date: Tue, 26 Sep 2006 11:19:00 +1000
Subject: [C++-sig] numpy.ndarray did not match PyArrayObject *
Message-ID: <fe76f74b0609251819w44b8ca84s97cd1a6edc388761@mail.gmail.com>

Hi David,
I did what you told me.
But it reported that PyArrayObject_Type undeclared though I included
"arrayobject.h".
Any important trick I missed?

-- 
Qinfeng(Javen) Shi

Research School of Information Sciences and Engineering
Australian National University
Locked Bag 8001
Canberra ACT 2601


From shiqinfeng at gmail.com  Tue Sep 26 04:15:12 2006
From: shiqinfeng at gmail.com (Qinfeng(Javen) Shi )
Date: Tue, 26 Sep 2006 12:15:12 +1000
Subject: [C++-sig] numpy.ndarray did not match PyArrayObject *
In-Reply-To: <fe76f74b0609251819w44b8ca84s97cd1a6edc388761@mail.gmail.com>
References: <fe76f74b0609251819w44b8ca84s97cd1a6edc388761@mail.gmail.com>
Message-ID: <fe76f74b0609251915s1776e90ek9a219b2e5dc9c4df@mail.gmail.com>

Hi David,

Since bjam can't find PyArrayObject_Type, I also tried
boost::python::lvalue_from_pytype< extract_array_object,const &PyArrayObject
>(); //converter. line:23.

~~~~~~~~~~
instead of PyArrayObject_Type.
bjam reports that "hello.cpp:23: error: parse error before `>'
token".hello.cpp:23 is the line of converter.

Any suggestion?


// Boost Includes
==============================================================
#include <boost/python.hpp>
#include <boost/cstdint.hpp>

// Includes
====================================================================
#include "hello.c"

// Using
=======================================================================
using namespace boost::python;

// Declarations
================================================================
BOOST_PYTHON_OPAQUE_SPECIALIZED_TYPE_ID(PyObject)


// Module
======================================================================
BOOST_PYTHON_MODULE(hello)
{
    def("main", &main);
    def("testList", &testList, return_value_policy< return_opaque_pointer
>());
    def("testArray", &testArray);
    def("trace", &trace, return_value_policy< return_opaque_pointer >());
    boost::python::lvalue_from_pytype<extract_array_object,const
&PyArrayObject>();
}
--
Qinfeng(Javen) Shi

Research School of Information Sciences and Engineering
Australian National University
Locked Bag 8001
Canberra ACT 2601


-- 
Qinfeng(Javen) Shi

Research School of Information Sciences and Engineering
Australian National University
Locked Bag 8001
Canberra ACT 2601
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/cplusplus-sig/attachments/20060926/9f9a7fe1/attachment.htm>

From dave at boost-consulting.com  Tue Sep 26 06:28:26 2006
From: dave at boost-consulting.com (David Abrahams)
Date: Tue, 26 Sep 2006 00:28:26 -0400
Subject: [C++-sig] numpy.ndarray did not match PyArrayObject *
References: <fe76f74b0609251819w44b8ca84s97cd1a6edc388761@mail.gmail.com>
Message-ID: <87hcyvnted.fsf@pereiro.luannocracy.com>

"Qinfeng(Javen) Shi " <shiqinfeng at gmail.com> writes:

> Hi David,
> I did what you told me.
> But it reported that PyArrayObject_Type undeclared though I included
> "arrayobject.h".
> Any important trick I missed?

Yes, the very subtle trick at the bottom of my posting

  "replace the indicated with the name of NumPy's python type object
  for ndarray."

:)
-- 
Dave Abrahams
Boost Consulting
www.boost-consulting.com



From shiqinfeng at gmail.com  Tue Sep 26 08:02:48 2006
From: shiqinfeng at gmail.com (Qinfeng(Javen) Shi )
Date: Tue, 26 Sep 2006 16:02:48 +1000
Subject: [C++-sig] to Philip Austin:Re: Stuck on accessing numpy.ndarray in
	c extension
Message-ID: <fe76f74b0609252302i66ad1e5du60769b3576f00883@mail.gmail.com>

Hi Philip Austin,

Thanks! I added following lines as you told me to hello.cpp file. It didn't
crash but still had error. The interesting thing is that now running it
twice would get two different error informations.

/************hello.cpp**************/
namespace ar=boost::python::numeric;
namespace bp=boost::python;
import_array();
ar::array::set_module_and_type("numpy", "ndarray");

########## error report##########
>>> import hello
>>> import numpy
>>> a = numpy.arange(25).reshape((5,5));
>>> hello.testArray(a)                                    // first time
SystemError: new style getargs format but argument is not a tuple
>>> hello.testArray(a)                                   // second time
args is array type!                         //it means PyArray_Check(args)
successed this time.
getting input failed!                        //PyArg_ParseTuple() failed but
not crashed

Any idea?
-- 
Qinfeng(Javen) Shi

Research School of Information Sciences and Engineering
Australian National University
Locked Bag 8001
Canberra ACT 2601
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/cplusplus-sig/attachments/20060926/4a3ad2ac/attachment.htm>

From Gottfried.Ganssauge at haufe.de  Tue Sep 26 09:48:27 2006
From: Gottfried.Ganssauge at haufe.de (Ganssauge, Gottfried)
Date: Tue, 26 Sep 2006 09:48:27 +0200
Subject: [C++-sig] Py++ - problem
	withsplit_moduleandreturn_opaque_pointer-policy
Message-ID: <51F4AD92DB06E9488BBC39B7B492409E029FE600@vg20ex1.grp.haufemg.com>

> -----Original Message-----
> From: Ganssauge, Gottfried 
> Sent: Monday, September 25, 2006 12:39 PM
> To: 'Development of Python/C++ integration'
> Subject: RE: [C++-sig] Py++ - problem 
> withsplit_moduleandreturn_opaque_pointer-policy
...
> > Also you don't have to add
> > BOOST_PYTHON_OPAQUE_SPECIALIZED_TYPE_ID macro
> To be exact: I _must_ not add ...
> 
> > Py++ will generate one.
> Indeed.
> It works quite nice :-)
One grain of salt though:
Py++ will happily generate more than one if there are several functions
returning the same opaque pointer type.
This leads to a compilation error: redefinition of
'boost::python::type_info boost::python::type_id() [with T = <opaque
pointee type>]
In the following test the generated file
test_opaque_free_functions.pypp.cpp will contain two instantiations of 
BOOST_PYTHON_OPAQUE_SPECIALIZED_TYPE_ID( opaque_ ).

test.hpp
---8><---><8---
typedef struct opaque_ *opaque_pointer;

opaque_pointer get_opaque(); 
opaque_pointer get2_opaque();  
---8><---><8---

test.py:
---8><---><8---
import os
from pyplusplus import module_builder
from pyplusplus.decl_wrappers import \
    return_value_policy, \
    return_opaque_pointer

mb = module_builder.module_builder_t( ["test.hpp"] )

mb.free_function (name = "get_opaque").call_policies = \
    return_value_policy (return_opaque_pointer)
mb.free_function (name = "get2_opaque").call_policies = \
    return_value_policy (return_opaque_pointer)

mb.build_code_creator( module_name='test_opaque' )

mb.split_module( '.' )  
---8><---><8---

Cheers,

Gottfried
> > 
> > --
> > Roman Yakovenko
> > C++ Python language binding
> > http://www.language-binding.net/
> > _______________________________________________
> > C++-sig mailing list
> > C++-sig at python.org
> > http://mail.python.org/mailman/listinfo/c++-sig
> > 
> 


From Gottfried.Ganssauge at haufe.de  Tue Sep 26 10:20:15 2006
From: Gottfried.Ganssauge at haufe.de (Ganssauge, Gottfried)
Date: Tue, 26 Sep 2006 10:20:15 +0200
Subject: [C++-sig] Another problem with split_module
Message-ID: <51F4AD92DB06E9488BBC39B7B492409E029FE612@vg20ex1.grp.haufemg.com>

Hi Roman,

as you may remember I'm generating function wrappers for functions
expecting pointers to callback functions along with an instance pointer
of type void.
How do I make sure these wrappers are generated to the same file as the
registration of the wrapper?

Consider the following test:
In single file mode this produces a working test module.
In split file mode the WRAPPER_CODE is generated to
test_wrapper.main.cpp whereas the registration is generated to
test_wrapper_free_functions.pypp.cpp.

test.hpp
--- 8>< ------ ><8 ---
void test (void *instance, void (*callback)(void *));
--- 8>< ------ ><8 ---

test.py
--- 8>< ------ ><8 ---
#
# Wrapper code for void test (void (*test)(void *));
#
WRAPPER_CODE = """
void wrap_test_callback (void *instance) {
    boost::python::call<void> (reinterpret_cast<PyObject *> (instance));

static
void test_wrapper (PyObject *callback) {
    test (
        callback && (callback != Py_None)? wrap_test_callback: 0,
        callback);
}
"""

import os
from pyplusplus import \
    module_builder
from pygccxml.declarations.calldef import \
     argument_t
from pygccxml.declarations.cpptypes import \
    dummy_type_t, \
    pointer_t

# PyObject *
pyobject_type = pointer_t (dummy_type_t ("PyObject"))

mb = module_builder.module_builder_t( ["test.hpp"] )

test = mb.free_function (name = "test")
test.arguments = [ argument_t (name = "callback", type = pyobject_type)
]
test.name      = "test_wrapper"
test.rename ("test")

# this goes to the wrong source file :-(
mb.add_declaration_code (WRAPPER_CODE)

mb.build_code_creator( module_name='test_wrapper' )

mb.split_module( '.' )
--- 8>< ------ ><8 ---

Cheers,

Gottfried 


From roman.yakovenko at gmail.com  Tue Sep 26 10:51:00 2006
From: roman.yakovenko at gmail.com (Roman Yakovenko)
Date: Tue, 26 Sep 2006 11:51:00 +0300
Subject: [C++-sig] Another problem with split_module
In-Reply-To: <51F4AD92DB06E9488BBC39B7B492409E029FE612@vg20ex1.grp.haufemg.com>
References: <51F4AD92DB06E9488BBC39B7B492409E029FE612@vg20ex1.grp.haufemg.com>
Message-ID: <7465b6170609260151v20548d9gb434225caddef276@mail.gmail.com>

On 9/26/06, Ganssauge, Gottfried <Gottfried.Ganssauge at haufe.de> wrote:
> Hi Roman,
>
> as you may remember I'm generating function wrappers for functions
> expecting pointers to callback functions along with an instance pointer
> of type void.
> How do I make sure these wrappers are generated to the same file as the
> registration of the wrapper?

The short answer is: you can not.
The long answer: I made some refactoring in order to introduce the
"register opaque type"
functionality and it should be very easy to implement what you want to do.

I will let you know when this functionality will be ready ( 3 -4  days ).

-- 
Roman Yakovenko
C++ Python language binding
http://www.language-binding.net/


From roman.yakovenko at gmail.com  Tue Sep 26 10:51:44 2006
From: roman.yakovenko at gmail.com (Roman Yakovenko)
Date: Tue, 26 Sep 2006 11:51:44 +0300
Subject: [C++-sig] Py++ - problem
	withsplit_moduleandreturn_opaque_pointer-policy
In-Reply-To: <51F4AD92DB06E9488BBC39B7B492409E029FE600@vg20ex1.grp.haufemg.com>
References: <51F4AD92DB06E9488BBC39B7B492409E029FE600@vg20ex1.grp.haufemg.com>
Message-ID: <7465b6170609260151u249d9d51u18124e17d92dafbd@mail.gmail.com>

On 9/26/06, Ganssauge, Gottfried <Gottfried.Ganssauge at haufe.de> wrote:
> One grain of salt though:

I just commited the fix to the repository. Can you try it?
( You need to update multiple_files.py only )

-- 
Roman Yakovenko
C++ Python language binding
http://www.language-binding.net/


From Gottfried.Ganssauge at haufe.de  Tue Sep 26 11:05:06 2006
From: Gottfried.Ganssauge at haufe.de (Ganssauge, Gottfried)
Date: Tue, 26 Sep 2006 11:05:06 +0200
Subject: [C++-sig] Py++ -
	problemwithsplit_moduleandreturn_opaque_pointer-policy
In-Reply-To: <7465b6170609260151u249d9d51u18124e17d92dafbd@mail.gmail.com>
Message-ID: <51F4AD92DB06E9488BBC39B7B492409E029FE63D@vg20ex1.grp.haufemg.com>

> -----Original Message-----
> From: c++-sig-bounces at python.org 
> [mailto:c++-sig-bounces at python.org] On Behalf Of Roman Yakovenko
> Sent: Tuesday, September 26, 2006 10:52 AM
> To: Development of Python/C++ integration
> Subject: Re: [C++-sig] Py++ - 
> problemwithsplit_moduleandreturn_opaque_pointer-policy
> 
> On 9/26/06, Ganssauge, Gottfried <Gottfried.Ganssauge at haufe.de> wrote:
> > One grain of salt though:
> 
> I just commited the fix to the repository. Can you try it?
> ( You need to update multiple_files.py only )
Brilliant!
It works.

Thanks,

Gottfried

> 
> --
> Roman Yakovenko
> C++ Python language binding
> http://www.language-binding.net/
> _______________________________________________
> C++-sig mailing list
> C++-sig at python.org
> http://mail.python.org/mailman/listinfo/c++-sig
> 


From j.reid at mail.cryst.bbk.ac.uk  Tue Sep 26 13:29:58 2006
From: j.reid at mail.cryst.bbk.ac.uk (John Reid)
Date: Tue, 26 Sep 2006 12:29:58 +0100
Subject: [C++-sig] [Graph] BGL python extension SVN version didn't link
Message-ID: <efb304$hin$1@sea.gmane.org>

Hi,

I tried to build the BGL from the latest SVN version and had some link 
errors to do with expat. My changes to fix these were the following:

Add the following to the $(EXTRA_REQUIREMENTS) in bgl-python\build\Jamfile
		<define>XML_STATIC

and add

	<define>XML_STATIC

to the requirements for lib expat in bgl-python\vendor\expat-2.0.0\Jamfile

I hope that helps someone or else someone can point out what I was doing 
wrong in the first place.

FWIW I'm using MSVC8 with boost RC_1_34_0

Cheers,
John.



From shiqinfeng at gmail.com  Tue Sep 26 17:53:07 2006
From: shiqinfeng at gmail.com (Qinfeng(Javen) Shi )
Date: Wed, 27 Sep 2006 01:53:07 +1000
Subject: [C++-sig] (David Abrahams)numpy.ndarray did not match
	PyArrayObject *
Message-ID: <fe76f74b0609260853v3095166fxae2f4651f2a65ee8@mail.gmail.com>

Hi David,


>boost::python::lvalue_from_pytype<extract_array_object,&PyArrayObject_Type>();
>struct extract_array_object
> {
>     static PyArrayObject& execute(PyArrayObject& x)
>     {
>         return x;
>     }
> };
> Yes, the very subtle trick at the bottom of my posting
> "replace the indicated with the name of NumPy's python type object
> for ndarray."

Sorry, I don't get you.
The Python type of the ndarray is PyArray_Type. InC, every ndarray is a
pointer to a PyArrayObject structure.
So what is the extract_array_object and what is PyArrayObject_Type?
I tried
boost::python::lvalue_from_pytype<PyArray_Type,&PyArrayObject>();
boost::python::lvalue_from_pytype<extract_array_object,&PyArray_Type>();
boost::python::lvalue_from_pytype<ndarray,&PyArray_Type>();
boost::python::lvalue_from_pytype<extract_array_object,&PyArrayObject_Type>();
boost::python::lvalue_from_pytype<extract_array_object,&ndarray>();
boost::python::lvalue_from_pytype<PyArray_Type,&PyArray_Type>();
boost::python::lvalue_from_pytype<ndarray,&ndarray>();
boost::python::lvalue_from_pytype<PyArray_Type,&ndarray>();
They all failed.

Could you please tell me exactly?
-- 
Qinfeng(Javen) Shi

Research School of Information Sciences and Engineering
Australian National University
Locked Bag 8001
Canberra ACT 2601
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/cplusplus-sig/attachments/20060927/37cf6634/attachment.htm>

From kir at lapshin.net  Tue Sep 26 18:20:53 2006
From: kir at lapshin.net (Kirill Lapshin)
Date: Tue, 26 Sep 2006 20:20:53 +0400
Subject: [C++-sig] Py++ and caching
Message-ID: <efbk16$nkg$1@sea.gmane.org>

Hi Roman,

Is there a way to not regenerate source code if cache is valid? I.e. 
currently cache saves us from running gccxml, but the whole code 
generation phase is still carried out even though it does not really 
changes anything (unless there was a change in driver .py script).

Ideally I think what has to be done is: add an option to Py++ that will 
say these files are also a dependency (in addition to header(s) we are 
parsing with gccxml and the whole tree of headers #included from there); 
Py++ will run gccxml as before, only when needed, and will run code 
generation only if gccxml has been run or any additional dependencies 
have been modified since last time. We'll list driver script as 
additional dependency.

This will reduce runtime of Py++ to almost zero if no changes are needed.

I think I should be able to do something similar externally by tracking 
additional dependencies myself and asking cache prior to instantiating 
module builder. However I would imagine others may benefit from such 
functionality as well. Unless it already exists that is. :)

Best regards,

Kirill



From paustin at eos.ubc.ca  Tue Sep 26 19:50:39 2006
From: paustin at eos.ubc.ca (Philip Austin)
Date: Tue, 26 Sep 2006 10:50:39 -0700
Subject: [C++-sig] (David Abrahams)numpy.ndarray did not match
	PyArrayObject *
In-Reply-To: <fe76f74b0609260853v3095166fxae2f4651f2a65ee8@mail.gmail.com>
References: <fe76f74b0609260853v3095166fxae2f4651f2a65ee8@mail.gmail.com>
Message-ID: <17689.26735.386649.800216@thrush.eos.ubc.ca>

Qinfeng(Javen) Shi  writes:
 > 
 > Sorry, I don't get you.
 > The Python type of the ndarray is PyArray_Type. InC, every ndarray is a
 > pointer to a PyArrayObject structure.
 > So what is the extract_array_object and what is PyArrayObject_Type?
 > I tried

[snip]

 > They all failed.
 > 
 > Could you please tell me exactly?

He wasn't sure of the exact type name, so he put in a place holder.

The actual type name is PyArray_Type, defined in 

http://svn.scipy.org/svn/numpy/trunk/numpy/core/src/arrayobject.c

As you note, this struct is compiled and passed to extension modules
using the PyArray_API array in __multiarray_api.c, which is
generated together with __multiarray_api.h by the script
http://svn.scipy.org/svn/numpy/trunk/numpy/core/code_generators/generate_array_api.py

so instead of getting PyArray_Type directly, as in the noddy
example in boost_1_33_1/libs/python/test/m1.cpp, it's provided by 
the line:

#define PyArray_Type (*(PyTypeObject *)PyArray_API[2])
which you get when you include __multiarray_api.h

I think this indirection is causing problems for 
lvalue_from_pytype<extract_array_object, &PyArray_Type>();
since I can't get this to work either.

Specifically, I edited

/home/phil/install/boost_1_33_1/libs/python/test/m1.cpp  
(diff is appended below)

to add 

  struct extract_array_object
  {
      static PyArrayObject& execute(PyArrayObject& x)
      {
          return x;
      }
  };

and at line 223:

    lvalue_from_pytype<extract_array_object, &PyArray_Type>();


but g++ 3.4.4 is unhappy with the second argument:

pam1.cpp: In function `void init_module_m1()':
pam1.cpp:223: error: missing `>' to terminate the template argument list
pam1.cpp:223: error: template argument 2 is invalid
pam1.cpp:223: error: expected primary-expression before ')' token

I'm out of time to work on this, but what I was going to try next was
to switch to Numeric-24.2 (because it's a little simpler 
to install and understand), and  reproduce the problem.

In Numeric the pertinent files are

Numeric-24.2/Src/_numpymodule.c  which has:

    PyArray_API[PyArray_Type_NUM] = (void *)&PyArray_Type;

and Include/arrayobject.h which has the PyArray_Type define

#define PyArray_Type *(PyTypeObject *)PyArray_API[PyArray_Type_NUM]

To convince myself that it's a C++ problem I was going to try to
create an array using

PyObject_NEW(PyArrayObject, &PyArray_Type) as in Src/arrayobject.c

and then put out a a working example with a failing lvalue converter
and an appeal to people who know more about C++ than I do 
-- regards, Phil

~/install/boost_1_33_1/libs/python/test phil at gull% diff -crN m1.cpp new_m1.cpp
*** m1.cpp	2004-08-20 04:09:19.000000000 -0700
--- new_m1.cpp	2006-09-26 09:50:42.000000000 -0700
***************
*** 16,21 ****
--- 16,23 ----
  #include <string.h>
  #include "simple_type.hpp"
  #include "complicated.hpp"
+ //#include <numpy/noprefix.h>
+ #include <numpy/arrayobject.h>
  
  // Declare some straightforward extension types
  extern "C" void
***************
*** 48,53 ****
--- 50,57 ----
      0,          /*tp_hash */
  };
  
+ 
+ 
  // Create a Noddy containing 42
  PyObject* new_noddy()
  {
***************
*** 200,212 ****
  
  Unregistered* make_unregistered2(int) { return new Unregistered; }
  
! BOOST_PYTHON_MODULE(m1)
  {
      using namespace boost::python;
      using boost::shared_ptr;
      
      simple_to_python();
  
      lvalue_from_pytype<int_from_noddy,&NoddyType>();
  
      lvalue_from_pytype<
--- 204,227 ----
  
  Unregistered* make_unregistered2(int) { return new Unregistered; }
  
!   struct extract_array_object
!   {
!       static PyArrayObject& execute(PyArrayObject& x)
!       {
!           return x;
!       }
!   };
! 
! 
! BOOST_PYTHON_MODULE(pam1)
  {
      using namespace boost::python;
      using boost::shared_ptr;
      
      simple_to_python();
  
+     lvalue_from_pytype<extract_array_object, &PyArray_Type>();
+ 
      lvalue_from_pytype<int_from_noddy,&NoddyType>();
  
      lvalue_from_pytype<
~/install/boost_1_33_1/libs/python/test phil at gull% 



From roman.yakovenko at gmail.com  Tue Sep 26 20:21:23 2006
From: roman.yakovenko at gmail.com (Roman Yakovenko)
Date: Tue, 26 Sep 2006 21:21:23 +0300
Subject: [C++-sig] Py++ and caching
In-Reply-To: <efbk16$nkg$1@sea.gmane.org>
References: <efbk16$nkg$1@sea.gmane.org>
Message-ID: <7465b6170609261121j34cdb478o67bb18a2ee67b03b@mail.gmail.com>

On 9/26/06, Kirill Lapshin <kir at lapshin.net> wrote:
> Hi Roman,
>
> Is there a way to not regenerate source code if cache is valid?

The short answer is - no.
Before we proceed I would like to know the time you will save, can you post it?
Actually few other metrics like size of the project, size of gccxml
generated file
b will be good.

> I.e.
> currently cache saves us from running gccxml, but the whole code
> generation phase is still carried out even though it does not really
> changes anything (unless there was a change in driver .py script).
> Ideally I think what has to be done is: add an option to Py++ that will
> say these files are also a dependency (in addition to header(s) we are
> parsing with gccxml and the whole tree of headers #included from there);
> Py++ will run gccxml as before, only when needed, and will run code
> generation only if gccxml has been run or any additional dependencies
> have been modified since last time. We'll list driver script as
> additional dependency.
>
> This will reduce runtime of Py++ to almost zero if no changes are needed.
>
> I think I should be able to do something similar externally by tracking
> additional dependencies myself and asking cache prior to instantiating
> module builder. However I would imagine others may benefit from such
> functionality as well. Unless it already exists that is. :)

This is not a new idea. It has been raised few times on pygccxml development
mailing list. I don't actually understand how it should work and more over in
my opinion this functionality should be implemented outside of the code
generation script. Like make utility, you describe the dependencies and
it decide what actions should be run.

I understand that it is not possible to implement such functionality without
some help from Py++. I think you need a function that will check whether
header files are different from those Py++ has in cache.

Py++ has goodies package. This package contains few utilities\convenience
function that were contributed by other users.  Can you create an
initial version
of the script and define missing functionality, than we will work
together to improve it and to add it to the goodies package?

-- 
Roman Yakovenko
C++ Python language binding
http://www.language-binding.net/


From roman.yakovenko at gmail.com  Tue Sep 26 22:14:59 2006
From: roman.yakovenko at gmail.com (Roman Yakovenko)
Date: Tue, 26 Sep 2006 23:14:59 +0300
Subject: [C++-sig] Difference between old and new Boost.Python interface.
Message-ID: <7465b6170609261314q4c558617k53f0af3f2f4c00f5@mail.gmail.com>

Good evening. I think I found a bug.
I expose a class that defines operator+. If I use old Boost.Python interface
for wrapper construction all works as expected. If I use new interface, then
I can not invoke it.

Please consider next code:

namespace bp = boost::python;

struct vector_new{
    vector_new(){}
    vector_new operator+( const vector_new& x ) const;
    static const vector_new v_new;
};

struct vector_old{
    vector_old(){}
    vector_old operator+( const vector_old& x ) const;
    static const vector_old v_old;
};

struct vector_new_wrapper : vector_new, bp::wrapper< vector_new > {
    vector_new_wrapper()
    : vector_new(), bp::wrapper< vector_new >()
    {}

    vector_new_wrapper(vector_new const & arg )
    : vector_new(), bp::wrapper< vector_new >()
    {}
};

struct vector_old_wrapper : vector_old, bp::wrapper< vector_old > {
    vector_old_wrapper(PyObject* )
    : vector_old(), bp::wrapper< vector_old >()
    {}

    vector_old_wrapper(PyObject*, vector_old const & arg )
    : vector_old(), bp::wrapper< vector_old >()
    {}
};

BOOST_PYTHON_MODULE( operators_ext ){
    bp::class_< vector_new_wrapper >( "vector_new" )
        .def( bp::self + bp::self )
        .def_readonly( "v_new", vector_new::v_new);

    bp::class_< vector_old, vector_old_wrapper >( "vector_old" )
        .def( bp::self + bp::self )
        .def_readonly( "v_old", vector_old::v_old );
}


I can invoke operator+ for vector_old class, but I can not invoke operator+ for
vector_new class:

import operators_ext
old = operators_ext.vector_old()
tmp = old + operators_ext.vector_old.v_old #this will work

new = operators_ext.vector_new()
tmp = new + operators_ext.vector_new.v_new
#Previous line will raise  TypeError:
#TypeError: unsupported operand type(s) for +: 'vector_new' and 'vector_new'

I attached minimal test case that reproduce the problem to the mail.

Thank you.

-- 
Roman Yakovenko
C++ Python language binding
http://www.language-binding.net/
-------------- next part --------------
A non-text attachment was scrubbed...
Name: operators_ext.cpp
Type: text/x-c++src
Size: 1403 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/cplusplus-sig/attachments/20060926/23e0a9fe/attachment.cpp>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: test.py
Type: text/x-python
Size: 386 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/cplusplus-sig/attachments/20060926/23e0a9fe/attachment.py>

From boost.python at cgrealy.mailcan.com  Wed Sep 27 00:46:33 2006
From: boost.python at cgrealy.mailcan.com (Colin Grealy)
Date: Wed, 27 Sep 2006 10:46:33 +1200
Subject: [C++-sig] [boost][python2.5] "with" support
Message-ID: <1159310793.25764.271924376@webmail.messagingengine.com>


Now that python 2.5 has been released, are there any plans for adding
support to bpl for the new language features?
Specifically will there be support for the __enter__() and __exit__()
special "with" methods? I think that most C++ developers would embrace
this feature as it's close to the RAII idiom that modern C++ uses.

Colin


From amohr at pixar.com  Wed Sep 27 01:44:50 2006
From: amohr at pixar.com (Alex Mohr)
Date: Tue, 26 Sep 2006 16:44:50 -0700
Subject: [C++-sig] [boost][python2.5] "with" support
In-Reply-To: <1159310793.25764.271924376@webmail.messagingengine.com>
References: <1159310793.25764.271924376@webmail.messagingengine.com>
Message-ID: <4519BB72.9010501@pixar.com>

What sort of support are you looking for?  You can certainly def() those 
methods on your wrapped classes easily with no changes to bpl.  Are you 
thinking that one might want some sort of mapping from C++ 
constructors/destructors to these methods?  I don't see a way that that 
would work off the top of my head.

Alex

Colin Grealy wrote:
> Now that python 2.5 has been released, are there any plans for adding
> support to bpl for the new language features?
> Specifically will there be support for the __enter__() and __exit__()
> special "with" methods? I think that most C++ developers would embrace
> this feature as it's close to the RAII idiom that modern C++ uses.
> 
> Colin
> _______________________________________________
> C++-sig mailing list
> C++-sig at python.org
> http://mail.python.org/mailman/listinfo/c++-sig


From shiqinfeng at gmail.com  Wed Sep 27 02:11:33 2006
From: shiqinfeng at gmail.com (Qinfeng(Javen) Shi )
Date: Wed, 27 Sep 2006 10:11:33 +1000
Subject: [C++-sig] (Philip Austin) (David Abrahams)numpy.ndarray did not
	match PyArrayObject
Message-ID: <fe76f74b0609261711r4951df51i7a03723b185d7dd0@mail.gmail.com>

Hi Philip and David

What I really want to do is to access numpy.ndarray in c extension. I tried
2 ways and Philip told me another possible way but I failed.

----------------------------------------------------- hello.c
1st way:
void testArray(PyArrayObject * args)
{
}

>>> hello.testArray(b)
Traceback (most recent call last):
 File "<stdin>", line 1, in ?
Boost.Python.ArgumentError: Python argument types in
   hello.testArray(numpy.ndarray)
did not match C++ signature:
   testArray(PyArrayObject*)
>>>
------------------------------------------------------hello.c
2nd way:
void testArray(PyObject * args)
{
   PyObject *input;
   PyArrayObject *array;
   if (!PyArg_ParseTuple(args, "O", &input))
          printf("getting input failed!\n");
   else
       printf("getting input sucess!\n");
}

>>> import hello
>>> import numpy
>>> c = numpy.arange(25).reshape((5,5))
>>> d = hello.testArray(c);
getting input failed!
>>> d
SystemError: new style getargs format but argument is not a tuple
>>>

then I tried adding following lines asPhilip told me to hello.cpp file.
------------------------------------------------hello.cpp
namespace ar=boost::python::numeric;
namespace bp=boost::python;
import_array();
ar::array::set_module_and_type ("numpy", "ndarray");

It didn't crash but still had error. The interesting thing is that now
running it
twice would get two different error informations.
---------------------------------------------------error report in
python
>>> import hello
>>> import numpy
>>> a = numpy.arange(25).reshape((5,5));
>>> hello.testArray(a)                                    // first time
SystemError: new style getargs format but argument is not a tuple
>>> hello.testArray(a)                                   // second time
args is array type!                         //it means PyArray_Check(args)
successed this time.
getting input failed!           //PyArg_ParseTuple() failed but not crashed

I am upset why I can't access numpy.ndarray.
Any suggestion?
Thanks!

-- 
Qinfeng(Javen) Shi

Research School of Information Sciences and Engineering
Australian National University
Locked Bag 8001
Canberra ACT 2601
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/cplusplus-sig/attachments/20060927/08f51f5b/attachment.htm>

From paustin at eos.ubc.ca  Wed Sep 27 03:27:15 2006
From: paustin at eos.ubc.ca (Philip Austin)
Date: Tue, 26 Sep 2006 18:27:15 -0700
Subject: [C++-sig] (no subject)
Message-ID: <17689.54131.291628.14500@thrush.eos.ubc.ca>

Qinfeng(Javen) Shi  writes:
 > Hi Philip and David
 > 
 > What I really want to do is to access numpy.ndarray in c extension. I tried
 > 2 ways and Philip told me another possible way but I failed.

The two appended files work for me with FC3 Linux -- Phil

-------------- next part --------------
An embedded and charset-unspecified text was scrubbed...
Name: numtest.cpp
URL: <http://mail.python.org/pipermail/cplusplus-sig/attachments/20060926/f5f6dcbb/attachment.txt>
-------------- next part --------------
An embedded and charset-unspecified text was scrubbed...
Name: test.py
URL: <http://mail.python.org/pipermail/cplusplus-sig/attachments/20060926/f5f6dcbb/attachment-0001.txt>

From cgrealy at mailcan.com  Wed Sep 27 07:08:10 2006
From: cgrealy at mailcan.com (Colin Grealy)
Date: Wed, 27 Sep 2006 17:08:10 +1200
Subject: [C++-sig] [boost][python2.5] "with" support
In-Reply-To: <4519BB72.9010501@pixar.com>
References: <1159310793.25764.271924376@webmail.messagingengine.com>
	<4519BB72.9010501@pixar.com>
Message-ID: <1159333690.22178.271945212@webmail.messagingengine.com>


On Tue, 26 Sep 2006 16:44:50 -0700, "Alex Mohr" <amohr at pixar.com> said:
> What sort of support are you looking for?  You can certainly def() those 
> methods on your wrapped classes easily with no changes to bpl.  Are you 
> thinking that one might want some sort of mapping from C++ 
> constructors/destructors to these methods?  I don't see a way that that 
> would work off the top of my head.
> 
> Alex

True, but to define the __exit__ method you need to write a C++ method
with the following signature

void func(boost::python::object exc_type,
          boost::python::object exc_val,
          boost::python::object exc_tb)

Most of the C++ objects I wrap are not python aware. I'd like some way
of wrapping a call to a simple method (e.g. void cleanup()). I suppose
I'm looking for a generic way to discard arguements from python.



From shiqinfeng at gmail.com  Wed Sep 27 08:47:42 2006
From: shiqinfeng at gmail.com (Qinfeng(Javen) Shi )
Date: Wed, 27 Sep 2006 16:47:42 +1000
Subject: [C++-sig] PyArg_ParseTuple() failed in parsing args
Message-ID: <fe76f74b0609262347p66d18a41q785cca47dba038ae@mail.gmail.com>

Dear all,

According the knowledge from http://numpy.scipy.org/numpydoc/numpy-13.html
It seems when input args is numpy.ndarray, it uses
PyArg_ParseTuple(args, "O!", &PyArray_Type, &array)) to convert PyObject*
args into PyArrayObject* array. Then we can handle the array in c extension.
But this PyArg_ParseTuple( ) failed(not crashed) when I used it in python
environment(even when I put a real tuple into args).

I have #define PY_ARRAY_UNIQUE_SYMBOL  Javen
and  import_array();
and even #include "ufuncobject.h";

Is that possible to make it work? I guess I missed some subtle thing.

Thanks!


-- 
Qinfeng(Javen) Shi

Research School of Information Sciences and Engineering
Australian National University
Locked Bag 8001
Canberra ACT 2601
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/cplusplus-sig/attachments/20060927/ce28ab57/attachment.htm>

From kir at lapshin.net  Wed Sep 27 16:48:40 2006
From: kir at lapshin.net (Kirill Lapshin)
Date: Wed, 27 Sep 2006 18:48:40 +0400
Subject: [C++-sig] Py++ and caching
In-Reply-To: <7465b6170609261121j34cdb478o67bb18a2ee67b03b@mail.gmail.com>
References: <efbk16$nkg$1@sea.gmane.org>
	<7465b6170609261121j34cdb478o67bb18a2ee67b03b@mail.gmail.com>
Message-ID: <efe308$mgj$1@sea.gmane.org>

Roman Yakovenko wrote:

>> Is there a way to not regenerate source code if cache is valid?
> The short answer is - no.
> Before we proceed I would like to know the time you will save, can you post it?
> Actually few other metrics like size of the project, size of gccxml
> generated file
> b will be good.

The whole process with parsing and generation takes about 40 sec, if 
gccxml output is cached, then runtime is reduced to 15 sec. I don't have 
hard numbers on how long does it take to validate cache, but from my 
debugging sessions it seemed to be almost instantaneous. The cache file 
is 4.6 MB.

I would love to get rid of these remaining 15 secs if possible.


> This is not a new idea. It has been raised few times on pygccxml development
> mailing list. I don't actually understand how it should work

I think it is hard to make bulletproof, but relatively easy to get 99% 
there. The hard part is already done -- you already cache gccxml 
results. If we hit the cache, then chances are the generated code 
already exists and there is no need to update it. There are only two 
reasons why that may be not the case: there are no generated files (user 
have deleted them), or python code describing what has to be exported 
has changed. Arguably first option is quite rare -- if there is cache, 
but no generated files then user should've deleted them manually. If we 
disregard this case then all we have to check is whether python code 
have been changed since last run of Py++. So we basically have to know 
the name(s) of python driver script(s) and record somewhere when last 
time Py++ has been run.
Accounting for the possibility of tampering with generated files is much 
more complicated and arguably is not possible to do reliably for the 
case of split files, since without loading gccxml output we can't tell 
what are the names of generated files. We can cover one common case of 
all generated files being deleted by checking whether main generated 
.cpp file is present. Not a bulletproof solution but certainly better 
then nothing.


> and more over in
> my opinion this functionality should be implemented outside of the code
> generation script. Like make utility, you describe the dependencies and
> it decide what actions should be run.

Well in some sense what I am asking for is more of build system domain. 
In ideal world build system would know that there is no need to run Py++ 
at all if nothing have changed. Unfortunately we are not living in ideal 
world and not all build systems are flexible enough to account for cases 
like this.

For example I was struggling with plugging Py++ into MSVC 
project/solution build. I've tried quite a few options, and finally 
settled on dummy project for code generation that runs Py++ in post 
build step. As a result code generation is started every time I run 
build, even if I haven't touched anything. Not nice, but other options 
were even uglier. No you may see why I would like to save the 15 sec.

> I understand that it is not possible to implement such functionality without
> some help from Py++. I think you need a function that will check whether
> header files are different from those Py++ has in cache.

That should be really simple, no? When I instantiate cache, but before 
instantiating module_builder, I can ask cache whether it is stale or 
not. Right?

What is needed is some way to quickly devise the names of all output 
files. Not sure if it is possible without extending data saved in cache 
though.

> 
> Py++ has goodies package. This package contains few utilities\convenience
> function that were contributed by other users.  Can you create an
> initial version
> of the script and define missing functionality, than we will work
> together to improve it and to add it to the goodies package?
> 

Sounds like a good idea. Will give it a try. Though I am bit tied up 
with other stuff, and it is not a high priority thing on my todo list. 
So don't expect anything soon, but hopefully I will dig into this 
eventually.


Kirill



From dave at boost-consulting.com  Wed Sep 27 16:58:50 2006
From: dave at boost-consulting.com (David Abrahams)
Date: Wed, 27 Sep 2006 10:58:50 -0400
Subject: [C++-sig] (David Abrahams)numpy.ndarray did not match
	PyArrayObject *
References: <fe76f74b0609260853v3095166fxae2f4651f2a65ee8@mail.gmail.com>
Message-ID: <87r6xxjqz9.fsf@pereiro.luannocracy.com>

"Qinfeng(Javen) Shi " <shiqinfeng at gmail.com> writes:

> Hi David,
>
>
>>boost::python::lvalue_from_pytype<extract_array_object,&PyArrayObject_Type>();
>>struct extract_array_object
>> {
>>     static PyArrayObject& execute(PyArrayObject& x)
>>     {
>>         return x;
>>     }
>> };
>> Yes, the very subtle trick at the bottom of my posting
>> "replace the indicated with the name of NumPy's python type object
>> for ndarray."
>
> Sorry, I don't get you.
> The Python type of the ndarray is PyArray_Type. 

Good.  In my original example, replace PyArrayObject_Type with
PyArray_Type.

-- 
Dave Abrahams
Boost Consulting
www.boost-consulting.com



From dave at boost-consulting.com  Wed Sep 27 17:12:04 2006
From: dave at boost-consulting.com (David Abrahams)
Date: Wed, 27 Sep 2006 11:12:04 -0400
Subject: [C++-sig] (David Abrahams)numpy.ndarray did not match
	PyArrayObject *
References: <fe76f74b0609260853v3095166fxae2f4651f2a65ee8@mail.gmail.com>
	<17689.26735.386649.800216@thrush.eos.ubc.ca>
Message-ID: <87lko5jqd7.fsf@pereiro.luannocracy.com>

Philip Austin <paustin at eos.ubc.ca> writes:

> so instead of getting PyArray_Type directly, as in the noddy
> example in boost_1_33_1/libs/python/test/m1.cpp, it's provided by 
> the line:
>
> #define PyArray_Type (*(PyTypeObject *)PyArray_API[2])
> which you get when you include __multiarray_api.h
>
> I think this indirection is causing problems for 
> lvalue_from_pytype<extract_array_object, &PyArray_Type>();
> since I can't get this to work either.

Yeah, that would be a problem.  lvalue_from_pytype needs a
compile-time constant for the Python type.  

Something like this will work, though:

   //
   // untested
   //

   void* extract_pyarray(PyObject* x)
   {
       return PyObject_TypeCheck(x, &PyArray_Type) ? x : 0;
   }

   BOOST_PYTHON_MODULE( /*your module name here*/ )
   {
       boost::python::converter::registry::insert(
             &extract_pyarray, type_id<PyArrayObject>());
   }

HTH,

-- 
Dave Abrahams
Boost Consulting
www.boost-consulting.com



From dave at boost-consulting.com  Wed Sep 27 17:13:51 2006
From: dave at boost-consulting.com (David Abrahams)
Date: Wed, 27 Sep 2006 11:13:51 -0400
Subject: [C++-sig] [boost][python2.5] "with" support
References: <1159310793.25764.271924376@webmail.messagingengine.com>
	<4519BB72.9010501@pixar.com>
	<1159333690.22178.271945212@webmail.messagingengine.com>
Message-ID: <87fyedjqa8.fsf@pereiro.luannocracy.com>

"Colin Grealy" <cgrealy at mailcan.com> writes:

>
> True, but to define the __exit__ method you need to write a C++ method
> with the following signature
>
> void func(boost::python::object exc_type,
>           boost::python::object exc_val,
>           boost::python::object exc_tb)
>
> Most of the C++ objects I wrap are not python aware. I'd like some way
> of wrapping a call to a simple method (e.g. void cleanup()). 

We need a design before it can be implemented.

> I suppose I'm looking for a generic way to discard arguements from
> python.

I don't understand the relationship between discarding arguments and
the rest of this thread.

-- 
Dave Abrahams
Boost Consulting
www.boost-consulting.com



From dave at boost-consulting.com  Wed Sep 27 17:17:07 2006
From: dave at boost-consulting.com (David Abrahams)
Date: Wed, 27 Sep 2006 11:17:07 -0400
Subject: [C++-sig] Difference between old and new Boost.Python interface.
References: <7465b6170609261314q4c558617k53f0af3f2f4c00f5@mail.gmail.com>
Message-ID: <87ac4ljq4s.fsf@pereiro.luannocracy.com>

"Roman Yakovenko" <roman.yakovenko at gmail.com> writes:

> I attached minimal test case that reproduce the problem to the mail.

It's not minimal.  I'm sorry, but I'm extremely pressed for time right
now.  Could you please make more of an effort to minimize it?  For
example, you can eliminate all explicit invocation of default
constructors.  You can also elminate the case that "works" and leave
only the case that "doesn't work."

-- 
Dave Abrahams
Boost Consulting
www.boost-consulting.com



From roman.yakovenko at gmail.com  Wed Sep 27 20:04:11 2006
From: roman.yakovenko at gmail.com (Roman Yakovenko)
Date: Wed, 27 Sep 2006 21:04:11 +0300
Subject: [C++-sig] Difference between old and new Boost.Python interface.
In-Reply-To: <87ac4ljq4s.fsf@pereiro.luannocracy.com>
References: <7465b6170609261314q4c558617k53f0af3f2f4c00f5@mail.gmail.com>
	<87ac4ljq4s.fsf@pereiro.luannocracy.com>
Message-ID: <7465b6170609271104l75de2cf8jec0eada52d461368@mail.gmail.com>

On 9/27/06, David Abrahams <dave at boost-consulting.com> wrote:
> "Roman Yakovenko" <roman.yakovenko at gmail.com> writes:
>
> > I attached minimal test case that reproduce the problem to the mail.
>
> It's not minimal.  I'm sorry, but I'm extremely pressed for time right
> now.  Could you please make more of an effort to minimize it?  For
> example, you can eliminate all explicit invocation of default
> constructors.  You can also elminate the case that "works" and leave
> only the case that "doesn't work."

I thought that you would like to see the "new" and "old" version. This
is not an urgent
bug, not at all. I just wanted to report it and it's work around.

I attached modified version of the files.

-- 
Roman Yakovenko
C++ Python language binding
http://www.language-binding.net/
-------------- next part --------------
A non-text attachment was scrubbed...
Name: operators_ext.cpp
Type: text/x-c++src
Size: 522 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/cplusplus-sig/attachments/20060927/4ea566d8/attachment.cpp>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: test.py
Type: text/x-python
Size: 127 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/cplusplus-sig/attachments/20060927/4ea566d8/attachment.py>

From roman.yakovenko at gmail.com  Wed Sep 27 20:22:01 2006
From: roman.yakovenko at gmail.com (Roman Yakovenko)
Date: Wed, 27 Sep 2006 21:22:01 +0300
Subject: [C++-sig] Py++ and caching
In-Reply-To: <efe308$mgj$1@sea.gmane.org>
References: <efbk16$nkg$1@sea.gmane.org>
	<7465b6170609261121j34cdb478o67bb18a2ee67b03b@mail.gmail.com>
	<efe308$mgj$1@sea.gmane.org>
Message-ID: <7465b6170609271122p60028458yd64ecc1079f88703@mail.gmail.com>

On 9/27/06, Kirill Lapshin <kir at lapshin.net> wrote:
>
> > This is not a new idea. It has been raised few times on pygccxml development
> > mailing list. I don't actually understand how it should work
>
> I think it is hard to make bulletproof, but relatively easy to get 99%
> there. The hard part is already done -- you already cache gccxml
> results. If we hit the cache, then chances are the generated code
> already exists and there is no need to update it. There are only two
> reasons why that may be not the case: there are no generated files (user
> have deleted them), or python code describing what has to be exported
> has changed. Arguably first option is quite rare -- if there is cache,
> but no generated files then user should've deleted them manually. If we
> disregard this case then all we have to check is whether python code
> have been changed since last run of Py++. So we basically have to know
> the name(s) of python driver script(s) and record somewhere when last
> time Py++ has been run.
> Accounting for the possibility of tampering with generated files is much
> more complicated and arguably is not possible to do reliably for the
> case of split files, since without loading gccxml output we can't tell
> what are the names of generated files. We can cover one common case of
> all generated files being deleted by checking whether main generated
> .cpp file is present. Not a bulletproof solution but certainly better
> then nothing.

This is what I thought. I don't like to have "holes" in the solution,
when we talk
about source code. It is too dangerous.

> For example I was struggling with plugging Py++ into MSVC
> project/solution build. I've tried quite a few options, and finally
> settled on dummy project for code generation that runs Py++ in post
> build step. As a result code generation is started every time I run
> build, even if I haven't touched anything. Not nice, but other options
> were even uglier. No you may see why I would like to save the 15 sec.

At work we had similar problem with "gettext" utility *.pi and *.mo files.
I will take a look how we solved it and let you know.

> > I understand that it is not possible to implement such functionality without
> > some help from Py++. I think you need a function that will check whether
> > header files are different from those Py++ has in cache.
>
> That should be really simple, no? When I instantiate cache, but before
> instantiating module_builder, I can ask cache whether it is stale or
> not. Right?

The functionality does not exists, but it should be simple to add it.

> > Py++ has goodies package. This package contains few utilities\convenience
> > function that were contributed by other users.  Can you create an
> > initial version
> > of the script and define missing functionality, than we will work
> > together to improve it and to add it to the goodies package?
> >
>
> Sounds like a good idea. Will give it a try. Though I am bit tied up
> with other stuff, and it is not a high priority thing on my todo list.
> So don't expect anything soon, but hopefully I will dig into this
> eventually.

When you have time, let me know, we will work together on it.

-- 
Roman Yakovenko
C++ Python language binding
http://www.language-binding.net/


From amohr at pixar.com  Wed Sep 27 20:31:24 2006
From: amohr at pixar.com (Alex Mohr)
Date: Wed, 27 Sep 2006 11:31:24 -0700
Subject: [C++-sig] Difference between old and new Boost.Python interface.
In-Reply-To: <7465b6170609271104l75de2cf8jec0eada52d461368@mail.gmail.com>
References: <7465b6170609261314q4c558617k53f0af3f2f4c00f5@mail.gmail.com>
	<87ac4ljq4s.fsf@pereiro.luannocracy.com>
	<7465b6170609271104l75de2cf8jec0eada52d461368@mail.gmail.com>
Message-ID: <451AC37C.10304@pixar.com>

My guess is that the operators stuff should be using 
detail::unwrap_wrapper to get the type for self rather than just using 
the wrapped_type.

This looks like a variant of the problem where if you are wrapping a 
polymorphic thing, you need to provide method pointers to the original 
class, not the polymorphic wrapper.  That's because boost python uses 
the type of the method pointer to figure out the type of the 'self' 
argument.

Looks like the bug might be in operators.hpp, where the actual def 
happens in the def_visitor -- it looks like it just uses 
ClassT::wrapped_type.  It should probably unwrap that.

This might be way off though -- David can correct me.

Alex


Roman Yakovenko wrote:
> On 9/27/06, David Abrahams <dave at boost-consulting.com> wrote:
>> "Roman Yakovenko" <roman.yakovenko at gmail.com> writes:
>>
>> > I attached minimal test case that reproduce the problem to the mail.
>>
>> It's not minimal.  I'm sorry, but I'm extremely pressed for time right
>> now.  Could you please make more of an effort to minimize it?  For
>> example, you can eliminate all explicit invocation of default
>> constructors.  You can also elminate the case that "works" and leave
>> only the case that "doesn't work."
> 
> I thought that you would like to see the "new" and "old" version. This
> is not an urgent
> bug, not at all. I just wanted to report it and it's work around.
> 
> I attached modified version of the files.
> 
> 
> ------------------------------------------------------------------------
> 
> #include "boost/python.hpp"
> 
> struct vector{
> 
>     vector(){}
> 
>     vector operator+( const vector& x ) const
>     { return vector(); }
> 
>     static const vector v;
> };
> 
> const vector vector::v;
> 
> using namespace boost::python;
> 
> struct vector_wrapper : vector, wrapper< vector > {
> 
>     vector_wrapper(){}
> 
>     vector_wrapper(vector const & arg ){}
> };
> 
> BOOST_PYTHON_MODULE( operators_ext ){
> 
>     class_< vector_wrapper >( "vector" )
>         .def( self + self )
>         .def_readonly( "v", vector::v);
> }
> 
> 
> ------------------------------------------------------------------------
> 
> from operators_ext import *
> 
> tmp = vector() + vector.v
> 
> print 'If you see this message than the bug was fixed. Thank you.'
> 
> 
> ------------------------------------------------------------------------
> 
> _______________________________________________
> C++-sig mailing list
> C++-sig at python.org
> http://mail.python.org/mailman/listinfo/c++-sig


From shiqinfeng at gmail.com  Wed Sep 27 14:52:45 2006
From: shiqinfeng at gmail.com (Qinfeng(Javen) Shi )
Date: Wed, 27 Sep 2006 22:52:45 +1000
Subject: [C++-sig] (Philip Austin) numpy.ndarray did not match
	PyArrayObject
Message-ID: <fe76f74b0609270552nbeab259o847d76140cf8ae10@mail.gmail.com>

Hi Philip,

It is GREAT!. You code works!

It is quite different from the numpybook 2006 or some online resources which
I can find.
Could please you provide some good stuffs about it here so that I can study
it?

Thanks you so much!

Cheers!
-- 
Qinfeng(Javen) Shi

Research School of Information Sciences and Engineering
Australian National University
Locked Bag 8001
Canberra ACT 2601
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/cplusplus-sig/attachments/20060927/93af4ea8/attachment.htm>

From paustin at eos.ubc.ca  Wed Sep 27 21:00:52 2006
From: paustin at eos.ubc.ca (Philip Austin)
Date: Wed, 27 Sep 2006 12:00:52 -0700
Subject: [C++-sig] (Philip Austin) numpy.ndarray did not
	match	PyArrayObject
In-Reply-To: <fe76f74b0609270552nbeab259o847d76140cf8ae10@mail.gmail.com>
References: <fe76f74b0609270552nbeab259o847d76140cf8ae10@mail.gmail.com>
Message-ID: <17690.51812.921459.340939@thrush.eos.ubc.ca>

Qinfeng(Javen) Shi  writes:

 > Could please you provide some good stuffs about it here so that I can study
 > it?

See:
http://www.eos.ubc.ca/research/clouds/software/pythonlibs/num_util/num_util_release2/Readme.html
(or http://tinyurl.com/pdqql)

for more routines in the general style of that example  -- cheers, Phil



From roman.yakovenko at gmail.com  Wed Sep 27 21:36:06 2006
From: roman.yakovenko at gmail.com (Roman Yakovenko)
Date: Wed, 27 Sep 2006 22:36:06 +0300
Subject: [C++-sig] Difference between old and new Boost.Python interface.
In-Reply-To: <451AC37C.10304@pixar.com>
References: <7465b6170609261314q4c558617k53f0af3f2f4c00f5@mail.gmail.com>
	<87ac4ljq4s.fsf@pereiro.luannocracy.com>
	<7465b6170609271104l75de2cf8jec0eada52d461368@mail.gmail.com>
	<451AC37C.10304@pixar.com>
Message-ID: <7465b6170609271236u20b918a1x3f890e5ca2666704@mail.gmail.com>

On 9/27/06, Alex Mohr <amohr at pixar.com> wrote:
> My guess is that the operators stuff should be using
> detail::unwrap_wrapper to get the type for self rather than just using
> the wrapped_type.
>
> This looks like a variant of the problem where if you are wrapping a
> polymorphic thing, you need to provide method pointers to the original
> class, not the polymorphic wrapper.  That's because boost python uses
> the type of the method pointer to figure out the type of the 'self'
> argument.
>
> Looks like the bug might be in operators.hpp, where the actual def
> happens in the def_visitor -- it looks like it just uses
> ClassT::wrapped_type.  It should probably unwrap that.

I think you are right. I tried to fix it, but run into problem.
It seems that Boost.Python does not contains helper class that extracts
"wrapped" class. ( I can not find it ). So I wrote next code that does not
compiles:

typedef

boost::type_traits::remove_pointer<

    boost::type_traits::function_traits<

        &detail::unwrap_wrapper<ClassT::wrapped_type>

     >::result_type

>::type unwrapped_type;



expected initializer before '<' token. I think that the problem with the way I
pass pointer to unwrap_wrapper function. Also I could be wrong. Do you have
an idea how I should fix the code or may be you can propose another way
to fix the problem?

Thanks

-- 
Roman Yakovenko
C++ Python language binding
http://www.language-binding.net/


From amohr at pixar.com  Wed Sep 27 22:26:02 2006
From: amohr at pixar.com (Alex Mohr)
Date: Wed, 27 Sep 2006 13:26:02 -0700
Subject: [C++-sig] Difference between old and new Boost.Python interface.
In-Reply-To: <7465b6170609271236u20b918a1x3f890e5ca2666704@mail.gmail.com>
References: <7465b6170609261314q4c558617k53f0af3f2f4c00f5@mail.gmail.com>
	<87ac4ljq4s.fsf@pereiro.luannocracy.com>
	<7465b6170609271104l75de2cf8jec0eada52d461368@mail.gmail.com>
	<451AC37C.10304@pixar.com>
	<7465b6170609271236u20b918a1x3f890e5ca2666704@mail.gmail.com>
Message-ID: <451ADE5A.9000907@pixar.com>

> I think you are right. I tried to fix it, but run into problem.
> It seems that Boost.Python does not contains helper class that extracts
> "wrapped" class. ( I can not find it ). So I wrote next code that does not
> compiles:
> 
> typedef
> 
> boost::type_traits::remove_pointer<
> 
>     boost::type_traits::function_traits<
> 
>         &detail::unwrap_wrapper<ClassT::wrapped_type>
> 
>      >::result_type
> 
>> ::type unwrapped_type;
> 
> expected initializer before '<' token. I think that the problem with the way I
> pass pointer to unwrap_wrapper function. Also I could be wrong. Do you have
> an idea how I should fix the code or may be you can propose another way
> to fix the problem?
> 
> Thanks

Try using the unwrapped type when you def the operators -- for instance, 
use:

.def(T() + T())

instead of

.def(self + self)

I think there's a way to avoid the construction too.. but the boost.org 
page is sluggish right now so I can't look it up easily.

Alex


From amohr at pixar.com  Wed Sep 27 22:41:40 2006
From: amohr at pixar.com (Alex Mohr)
Date: Wed, 27 Sep 2006 13:41:40 -0700
Subject: [C++-sig] Difference between old and new Boost.Python interface.
In-Reply-To: <451ADE5A.9000907@pixar.com>
References: <7465b6170609261314q4c558617k53f0af3f2f4c00f5@mail.gmail.com>
	<87ac4ljq4s.fsf@pereiro.luannocracy.com>
	<7465b6170609271104l75de2cf8jec0eada52d461368@mail.gmail.com>
	<451AC37C.10304@pixar.com>
	<7465b6170609271236u20b918a1x3f890e5ca2666704@mail.gmail.com>
	<451ADE5A.9000907@pixar.com>
Message-ID: <451AE204.90007@pixar.com>

> Try using the unwrapped type when you def the operators -- for instance, 
> use:
> 
> .def(T() + T())
> 
> instead of
> 
> .def(self + self)
> 
> I think there's a way to avoid the construction too.. but the boost.org 
> page is sluggish right now so I can't look it up easily.

Sorry -- nevermind.  The thing I was thinking of is bp::other<T>, but 
you need to use self in order to use the operators stuff...  So I don't 
think there's any other workaround for you for now other than to just 
def the operator methods directly...

Alex


From undefined at pobox.com  Thu Sep 28 06:58:01 2006
From: undefined at pobox.com (Corey Wright)
Date: Wed, 27 Sep 2006 23:58:01 -0500
Subject: [C++-sig] boost.python w/ boost.build v2 (was Re: Boost.Python,
 MinGW, & ImportError)
In-Reply-To: <20060924160105.92cfcf8a.undefined@pobox.com>
References: <20060920011747.c857b35d.undefined@pobox.com>
	<87ac4uedcl.fsf@pereiro.luannocracy.com>
	<20060924160105.92cfcf8a.undefined@pobox.com>
Message-ID: <20060927235801.17c8ed0e.undefined@pobox.com>

On Sun, 24 Sep 2006 16:01:05 -0500
Corey Wright <undefined at pobox.com> wrote:

> anyways, i'll look at boost.python on mingw sometime this week, but wanted
> to feed back my patch for Jamfile.v2 ("release early, release often").

with:
- the 06-09-27-1700 tarball
- the previously attached patches (fix Jamfile.v2, tweak boost.build.jam)
- similar user-config.jam (but targeting python on windows)
boost.python and the examples compiled under mingw and passed the tests.

with the attached Jamfile.v2 and test_hello.py (modeled after
test_std_pair.py) the tutorial compiled and tested fine (and due to the
boost.build framework, typing "bjam --toolset=mingw --v2 test" is sure
easier than manually setting PATH to include boost_python.dll, setting
PYTHONPATH to include boost.python module, etc).

thanks for boost.python.
-- 
undefined at pobox.com
-------------- next part --------------
An embedded and charset-unspecified text was scrubbed...
Name: Jamfile.v2
URL: <http://mail.python.org/pipermail/cplusplus-sig/attachments/20060927/c1bc603a/attachment.txt>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: test_hello.py
Type: text/x-python
Size: 257 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/cplusplus-sig/attachments/20060927/c1bc603a/attachment.py>

From thilinasepala at gmail.com  Thu Sep 28 09:22:38 2006
From: thilinasepala at gmail.com (thilina sepala dahanayake)
Date: Thu, 28 Sep 2006 13:22:38 +0600
Subject: [C++-sig] please send me basic
Message-ID: <e554344e0609280022w75f730eq122e94ffdf6080b6@mail.gmail.com>

I'm a new to c++, so I can't understand the hard things of c++.I want only
basic of c++
I'll give you what I want

   - about Loops
   - what is arrys
   - what can i do from classes and objects
   - basic of oop

I'll thank you very much if you send me above topics lectures
thank you
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/cplusplus-sig/attachments/20060928/277f3092/attachment.htm>

From roman.yakovenko at gmail.com  Thu Sep 28 09:26:31 2006
From: roman.yakovenko at gmail.com (Roman Yakovenko)
Date: Thu, 28 Sep 2006 10:26:31 +0300
Subject: [C++-sig] please send me basic
In-Reply-To: <e554344e0609280022w75f730eq122e94ffdf6080b6@mail.gmail.com>
References: <e554344e0609280022w75f730eq122e94ffdf6080b6@mail.gmail.com>
Message-ID: <7465b6170609280026x48d13351ra30a7a8b10ba537a@mail.gmail.com>

On 9/28/06, thilina sepala dahanayake <thilinasepala at gmail.com> wrote:
> I'm a new to c++, so I can't understand the hard things of c++.I want only
> basic of c++
> I'll give you what I want
>
> about Loops
> what is arrys
> what can i do from classes and objects
> basic of oop I'll thank you very much if you send me above topics lectures
> thank you

You asking for help on a wrong group. Before you post to another group
consider to read this tutorial -
"http://www.catb.org/~esr/faqs/smart-questions.html"

-- 
Roman Yakovenko
C++ Python language binding
http://www.language-binding.net/


From dave at boost-consulting.com  Thu Sep 28 15:42:31 2006
From: dave at boost-consulting.com (David Abrahams)
Date: Thu, 28 Sep 2006 09:42:31 -0400
Subject: [C++-sig] Difference between old and new Boost.Python interface.
References: <7465b6170609261314q4c558617k53f0af3f2f4c00f5@mail.gmail.com>
	<87ac4ljq4s.fsf@pereiro.luannocracy.com>
	<7465b6170609271104l75de2cf8jec0eada52d461368@mail.gmail.com>
Message-ID: <87d59gxg3c.fsf@pereiro.luannocracy.com>

"Roman Yakovenko" <roman.yakovenko at gmail.com> writes:

> I attached modified version of the files.

Thanks, that's better.  Still not minimal (why are you explicitly
defining the default and copy ctors?) but it is close enough.

-- 
Dave Abrahams
Boost Consulting
www.boost-consulting.com



From dave at boost-consulting.com  Thu Sep 28 16:41:42 2006
From: dave at boost-consulting.com (David Abrahams)
Date: Thu, 28 Sep 2006 10:41:42 -0400
Subject: [C++-sig] Difference between old and new Boost.Python interface.
References: <7465b6170609261314q4c558617k53f0af3f2f4c00f5@mail.gmail.com>
	<87ac4ljq4s.fsf@pereiro.luannocracy.com>
	<7465b6170609271104l75de2cf8jec0eada52d461368@mail.gmail.com>
	<451AC37C.10304@pixar.com>
Message-ID: <877izoxdcp.fsf@pereiro.luannocracy.com>

Alex Mohr <amohr at pixar.com> writes:

> My guess is that the operators stuff should be using 
> detail::unwrap_wrapper to get the type for self rather than just using 
> the wrapped_type.
>
> This looks like a variant of the problem where if you are wrapping a 
> polymorphic thing, you need to provide method pointers to the original 
> class, not the polymorphic wrapper.  That's because boost python uses 
> the type of the method pointer to figure out the type of the 'self' 
> argument.
>
> Looks like the bug might be in operators.hpp, where the actual def 
> happens in the def_visitor -- it looks like it just uses 
> ClassT::wrapped_type.  It should probably unwrap that.
>
> This might be way off though -- David can correct me.

Dead on, Alex.

Roman, I've checked in a fix (plus tests) to the CVS HEAD.  Could you
let me know if it works for you?  If deemed good, I'll transfer these
fixes to the release branch.

-- 
Dave Abrahams
Boost Consulting
www.boost-consulting.com



From roman.yakovenko at gmail.com  Thu Sep 28 20:04:54 2006
From: roman.yakovenko at gmail.com (Roman Yakovenko)
Date: Thu, 28 Sep 2006 21:04:54 +0300
Subject: [C++-sig] Difference between old and new Boost.Python interface.
In-Reply-To: <877izoxdcp.fsf@pereiro.luannocracy.com>
References: <7465b6170609261314q4c558617k53f0af3f2f4c00f5@mail.gmail.com>
	<87ac4ljq4s.fsf@pereiro.luannocracy.com>
	<7465b6170609271104l75de2cf8jec0eada52d461368@mail.gmail.com>
	<451AC37C.10304@pixar.com> <877izoxdcp.fsf@pereiro.luannocracy.com>
Message-ID: <7465b6170609281104j68ba81ccr64788f08cd0545c3@mail.gmail.com>

On 9/28/06, David Abrahams <dave at boost-consulting.com> wrote:
> Roman, I've checked in a fix (plus tests) to the CVS HEAD.  Could you
> let me know if it works for you?  If deemed good, I'll transfer these
> fixes to the release branch.

I saw the message :-).
It works, thank you. I tested it on Ubuntu, gcc 4.0.3.

-- 
Roman Yakovenko
C++ Python language binding
http://www.language-binding.net/


From jon.kristensen at dedikerad.se  Thu Sep 28 20:23:35 2006
From: jon.kristensen at dedikerad.se (Jon Kristensen)
Date: Thu, 28 Sep 2006 20:23:35 +0200
Subject: [C++-sig] Problem with PyMethodDef and non-static functions
Message-ID: <1159467816.14049.9.camel@jon>

Hello list!

I'm using Python/C to be able to have simple Python scripts in my
application. I was free of problems until I switched from using static
and global functions to using non-static class methods:

python.cpp:101: error: argument of type ?PyObject* (Python::)(PyObject*,
PyObject*)? does not match ?PyObject* (*)(PyObject*, PyObject*)?

After some time I understood that this was a problem that a little
tweaking around wouldn't solve. So I want to ask you guys if you can
describe how to get around this. I wouldn't want to add any additional
dependencies such as SWIG or Boost. Please also note that I'm not that
experienced with Python.

And I want to keep this as clean/simple as possible! :-)

Thanks in advance!

-- 
Jon Kristensen



From seefeld at sympatico.ca  Thu Sep 28 20:50:20 2006
From: seefeld at sympatico.ca (Stefan Seefeld)
Date: Thu, 28 Sep 2006 14:50:20 -0400
Subject: [C++-sig] Problem with PyMethodDef and non-static functions
In-Reply-To: <1159467816.14049.9.camel@jon>
References: <1159467816.14049.9.camel@jon>
Message-ID: <451C196C.7040101@sympatico.ca>

Jon Kristensen wrote:
> Hello list!
> 
> I'm using Python/C to be able to have simple Python scripts in my
> application. I was free of problems until I switched from using static
> and global functions to using non-static class methods:
> 
> python.cpp:101: error: argument of type ?PyObject* (Python::)(PyObject*,
> PyObject*)? does not match ?PyObject* (*)(PyObject*, PyObject*)?
> 
> After some time I understood that this was a problem that a little
> tweaking around wouldn't solve. So I want to ask you guys if you can
> describe how to get around this. I wouldn't want to add any additional
> dependencies such as SWIG or Boost. Please also note that I'm not that
> experienced with Python.

As you have found out, ordinary pointers and pointers to members are
fundamentally incompatible with each other, i.e. there is no way to
cast from one to the other.
You have to implement a wrapper around your method call. Something like:

struct MyType
{
  int func(int);
};

PyObject *my_method(PyObject *self, PyObject *args)
{
  CxxType *object = extract_object_from_self(self);
  int arg = extract_arg(args);
  int result = object->func(arg);
  return convert_result_to_python(result);
}

But of course, all this forward and backward extraction / conversion
business makes the above highly inconvenient and non-scalable, so you may
indeed want to give boost.python a try as your project grows. :-)


HTH,
		Stefan

-- 

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


From jon.kristensen at dedikerad.se  Thu Sep 28 21:07:28 2006
From: jon.kristensen at dedikerad.se (Jon Kristensen)
Date: Thu, 28 Sep 2006 21:07:28 +0200
Subject: [C++-sig] Problem with PyMethodDef and non-static functions
In-Reply-To: <451C196C.7040101@sympatico.ca>
References: <1159467816.14049.9.camel@jon>  <451C196C.7040101@sympatico.ca>
Message-ID: <1159470448.14049.15.camel@jon>

tor 2006-09-28 klockan 14:50 -0400 skrev Stefan Seefeld:
> Jon Kristensen wrote:
> > Hello list!
> > 
> > I'm using Python/C to be able to have simple Python scripts in my
> > application. I was free of problems until I switched from using static
> > and global functions to using non-static class methods:
> > 
> > python.cpp:101: error: argument of type ?PyObject* (Python::)(PyObject*,
> > PyObject*)? does not match ?PyObject* (*)(PyObject*, PyObject*)?
> > 
> > After some time I understood that this was a problem that a little
> > tweaking around wouldn't solve. So I want to ask you guys if you can
> > describe how to get around this. I wouldn't want to add any additional
> > dependencies such as SWIG or Boost. Please also note that I'm not that
> > experienced with Python.
> 
> As you have found out, ordinary pointers and pointers to members are
> fundamentally incompatible with each other, i.e. there is no way to
> cast from one to the other.
> You have to implement a wrapper around your method call. Something like:
> 
> struct MyType
> {
>   int func(int);
> };
> 
> PyObject *my_method(PyObject *self, PyObject *args)
> {
>   CxxType *object = extract_object_from_self(self);
>   int arg = extract_arg(args);
>   int result = object->func(arg);
>   return convert_result_to_python(result);
> }

Thanks for your reply. I don't understand the following:

      * What purpose does the struct MyType serve?
      * What's extract_object_from_self? (Don't find anything on
        Google.)

> But of course, all this forward and backward extraction / conversion
> business makes the above highly inconvenient and non-scalable, so you may
> indeed want to give boost.python a try as your project grows. :-)

OK, perhaps I will! :-)

> HTH,
> 		Stefan
> 
-- 
Jon Kristensen



From seefeld at sympatico.ca  Thu Sep 28 21:19:22 2006
From: seefeld at sympatico.ca (Stefan Seefeld)
Date: Thu, 28 Sep 2006 15:19:22 -0400
Subject: [C++-sig] Problem with PyMethodDef and non-static functions
In-Reply-To: <1159470448.14049.15.camel@jon>
References: <1159467816.14049.9.camel@jon> <451C196C.7040101@sympatico.ca>
	<1159470448.14049.15.camel@jon>
Message-ID: <451C203A.2090604@sympatico.ca>

Jon Kristensen wrote:

> Thanks for your reply. I don't understand the following:
> 
>       * What purpose does the struct MyType serve?

You ask for a way to bind a C++ member function to a python method.
But a member is always a member of some object, so MyType provides one.
Both, MyType as well as MyType::func are placeholders for your own code.

>       * What's extract_object_from_self? (Don't find anything on
>         Google.)

Indeed. You have to write that yourself, i.e. a function to extract
a MyType pointer (again, placeholder for your own object pointer)
from the python wrapper through which the member function is to be invoked.

Boost.python does all that for you automagically...

Regards,
		Stefan

-- 

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


From roman.yakovenko at gmail.com  Thu Sep 28 22:25:52 2006
From: roman.yakovenko at gmail.com (Roman Yakovenko)
Date: Thu, 28 Sep 2006 23:25:52 +0300
Subject: [C++-sig] Another problem with split_module
In-Reply-To: <7465b6170609260151v20548d9gb434225caddef276@mail.gmail.com>
References: <51F4AD92DB06E9488BBC39B7B492409E029FE612@vg20ex1.grp.haufemg.com>
	<7465b6170609260151v20548d9gb434225caddef276@mail.gmail.com>
Message-ID: <7465b6170609281325j790f8ba9t40fe37811788843c@mail.gmail.com>

On 9/26/06, Roman Yakovenko <roman.yakovenko at gmail.com> wrote:
> On 9/26/06, Ganssauge, Gottfried <Gottfried.Ganssauge at haufe.de> wrote:
> > Hi Roman,
> >
> > as you may remember I'm generating function wrappers for functions
> > expecting pointers to callback functions along with an instance pointer
> > of type void.
> > How do I make sure these wrappers are generated to the same file as the
> > registration of the wrapper?
>
> The short answer is: you can not.
> The long answer: I made some refactoring in order to introduce the
> "register opaque type"
> functionality and it should be very easy to implement what you want to do.
>
> I will let you know when this functionality will be ready ( 3 -4  days ).

Well, it took me less time than I thought. The functionality is ready.

mb = module_builder_t( ... )
mb.free_function( name of the function ).add_declaration_code( your code here )

Enjoy.

P.S. This time svn contains a lot of changes, also I tested them pretty well
I expect few bugs here and their. It could be nice if you can test the latest
code.

Thank you.

-- 
Roman Yakovenko
C++ Python language binding
http://www.language-binding.net/


From sergeymak at pisem.net  Fri Sep 29 09:25:18 2006
From: sergeymak at pisem.net (Sergey Makovkin)
Date: Fri, 29 Sep 2006 11:25:18 +0400
Subject: [C++-sig] boost::python problem with std::string
Message-ID: <efihos$k2o$1@sea.gmane.org>

    Hello!

I want to use Boost.python in my program but I have a problem. When I call a 
C++ function with std::string argument (or const std::string& ) from python, 
boost.python crashes! I don't know what to do and I hope you can help me.

Here is a link to the source code which demonstrates the problem:
http://www.powerfly.ru/bug.zip
To compile it, you need MS Visual Studio 2005, Python 2.4 and Boost.Python 
1.33 or later.

-- 
Sergey A. Makovkin 





From Gottfried.Ganssauge at haufe.de  Fri Sep 29 10:58:25 2006
From: Gottfried.Ganssauge at haufe.de (Ganssauge, Gottfried)
Date: Fri, 29 Sep 2006 10:58:25 +0200
Subject: [C++-sig] Another problem with split_module
In-Reply-To: <7465b6170609281325j790f8ba9t40fe37811788843c@mail.gmail.com>
Message-ID: <51F4AD92DB06E9488BBC39B7B492409E029FE9C9@vg20ex1.grp.haufemg.com>

> -----Original Message-----
> From: c++-sig-bounces at python.org 
> [mailto:c++-sig-bounces at python.org] On Behalf Of Roman Yakovenko
> Sent: Thursday, September 28, 2006 10:26 PM
> To: Development of Python/C++ integration
> Subject: Re: [C++-sig] Another problem with split_module
> 
> On 9/26/06, Roman Yakovenko <roman.yakovenko at gmail.com> wrote:
> > On 9/26/06, Ganssauge, Gottfried 
> <Gottfried.Ganssauge at haufe.de> wrote:
> > > Hi Roman,
> > >
> > > as you may remember I'm generating function wrappers for 
> functions 
> > > expecting pointers to callback functions along with an instance 
> > > pointer of type void.
> > > How do I make sure these wrappers are generated to the 
> same file as 
> > > the registration of the wrapper?
> >
> > The short answer is: you can not.
> > The long answer: I made some refactoring in order to introduce the 
> > "register opaque type"
> > functionality and it should be very easy to implement what 
> you want to do.
> >
> > I will let you know when this functionality will be ready ( 
> 3 -4  days ).
> 
> Well, it took me less time than I thought. The functionality is ready.
Great.

> 
> mb = module_builder_t( ... )
> mb.free_function( name of the function 
> ).add_declaration_code( your code here )
Is there a reason you omitted the 'tail' argument from this variant of
the function?
I certainly would need it e.g.
	- to make sure includes appear at the head of the generated
module
	- to make sure forward declarations appear before their use
	- ... (you name it :-)
> 
> Enjoy.
I do :-)

> 
> P.S. This time svn contains a lot of changes, also I tested 
> them pretty well I expect few bugs here and their. It could 
> be nice if you can test the latest code.
When running my build script I get the following error:
build.py:346: DeprecationWarning: create_casting_constructor argument is
deprecated and should not be used.
If you still want Py++ to generate implicitly_convertible code, consider
to use allow_implicit_conversion constructor property
mb = module_builder_t(...)
mb.constructors().allow_implicit_conversion = True
  doc_extractor = \
Traceback (most recent call last):
  File "build.py", line 557, in ?
    main (args, of)
  File "build.py", line 535, in main
    mb.build_code_creator ()
  File "build.py", line 346, in build_code_creator
    doc_extractor = \
  File
"/usr/lib/python2.4/site-packages/pyplusplus/module_builder/builder.py",
line 246, in build_code_creator
    , doc_extractor)
  File
"/usr/lib/python2.4/site-packages/pyplusplus/module_creator/creator.py",
line 126, in __init__
    prepared_decls = self._prepare_decls( decls, doc_extractor )
  File
"/usr/lib/python2.4/site-packages/pyplusplus/module_creator/creator.py",
line 160, in _prepare_decls
    readme = filter( lambda msg: msg not in DO_NOT_REPORT_MSGS,
decl.readme() )
  File
"/usr/lib/python2.4/site-packages/pyplusplus/decl_wrappers/decl_wrapper.
py", line 125, in readme
    text.extend( self._readme_impl() )
  File
"/usr/lib/python2.4/site-packages/pyplusplus/decl_wrappers/class_wrapper
.py", line 456, in _readme_impl
    return self.is_wrapper_needed()
  File
"/usr/lib/python2.4/site-packages/pyplusplus/decl_wrappers/class_wrapper
.py", line 427, in is_wrapper_needed
    if self.redefined_funcs():
  File
"/usr/lib/python2.4/site-packages/pyplusplus/decl_wrappers/class_wrapper
.py", line 385, in redefined_funcs
    is_same_function = declarations.is_same_function
AttributeError: 'module' object has no attribute 'is_same_function'

The call in line 346 of build.py is:
        module_builder.module_builder_t.build_code_creator (
            self,
            module_name   = self.module_name,
            doc_extractor = \
                lambda decl, extractor = doc_extractor(): \
                    make_cstring (extractor (decl))

Self is an instance derived from module_builder.module_builder_t.

I don't understand the DeprecationWarning - there is nothing like a
create_casting_constructor argument being used ...

Cheers,

Gottfried

> 
> Thank you.
> 
> --
> Roman Yakovenko
> C++ Python language binding
> http://www.language-binding.net/
> _______________________________________________
> C++-sig mailing list
> C++-sig at python.org
> http://mail.python.org/mailman/listinfo/c++-sig
> 


From shiqinfeng at gmail.com  Fri Sep 29 15:13:05 2006
From: shiqinfeng at gmail.com (Qinfeng(Javen) Shi )
Date: Fri, 29 Sep 2006 23:13:05 +1000
Subject: [C++-sig] PyArray_SimpleNewFromData(, PyArray_FLOAT,
	) can't get float c array in c extension?
Message-ID: <fe76f74b0609290613w6d865f95t8ccf4c8c88a9343e@mail.gmail.com>

Dear All,

I used PyArray_SimpleNewFromData to transfer a float c array data into
numpy.ndarray.
But it got the wrong array values. Any suggestion?
Thanks in advance!

---------------------------hello.c-------------
PyObject * testCreatArray()
{
    float fArray[5] = {0,1,2,3,4};
    intp m = 5;
    PyObject * c = PyArray_SimpleNewFromData(1,&m,PyArray_FLOAT,fArray);
    return c;
}


--------------------------result--------------------
>>> import hello
>>> import numpy
>>> list1 = range(5);
>>> alist = hello.testCreatArray()
>>> alist
array([  0.00000000e+00,  -2.65695271e-05,   4.64483736e-34,
         4.34402524e-44,  -8.06840762e-06], dtype=float32)
>>>




-- 
Qinfeng(Javen) Shi

Research School of Information Sciences and Engineering
Australian National University
Locked Bag 8001
Canberra ACT 2601
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/cplusplus-sig/attachments/20060929/f54bdbc6/attachment.htm>

From dave at boost-consulting.com  Fri Sep 29 17:26:22 2006
From: dave at boost-consulting.com (David Abrahams)
Date: Fri, 29 Sep 2006 11:26:22 -0400
Subject: [C++-sig] boost::python problem with std::string
References: <efihos$k2o$1@sea.gmane.org>
Message-ID: <87fyear8wx.fsf@pereiro.luannocracy.com>

"Sergey Makovkin" <sergeymak at pisem.net> writes:

>     Hello!
>
> I want to use Boost.python in my program but I have a problem. When I call a 
> C++ function with std::string argument

You mean parameter

> (or const std::string& ) from python, 

With a Python str argument?

> boost.python crashes! I don't know what to do and I hope you can help me.
>
> Here is a link to the source code which demonstrates the problem:
> http://www.powerfly.ru/bug.zip
> To compile it, you need MS Visual Studio 2005, Python 2.4 and Boost.Python 
> 1.33 or later.

Because we test that capability all the time, I can only conclude that
your problem is either:

  1) a bug somewhere in your code
  2) a configuration error in your build

I suspect #2, but to rule out #1 you should try a minimal, complete
test case that does nothing but call such a function from Python.  If
that still fails, you have a configuration error in your build.  To
correct that problem you should follow the official Boost.Python build
instructions using bjam (at http://www.boost.org/libs/python/doc/building.html) 
and add the -d+2 option to examine the command lines generated,
comparing them with what your IDE generates.


-- 
Dave Abrahams
Boost Consulting
www.boost-consulting.com



From dgregor at cs.indiana.edu  Fri Sep 29 17:54:11 2006
From: dgregor at cs.indiana.edu (Doug Gregor)
Date: Fri, 29 Sep 2006 11:54:11 -0400
Subject: [C++-sig] [Graph] BGL python extension SVN version didn't link
In-Reply-To: <efb304$hin$1@sea.gmane.org>
References: <efb304$hin$1@sea.gmane.org>
Message-ID: <E5205ADB-6A64-448E-A6D7-AC11DAD9E693@cs.indiana.edu>


On Sep 26, 2006, at 7:29 AM, John Reid wrote:

> Hi,
>
> I tried to build the BGL from the latest SVN version and had some link
> errors to do with expat. My changes to fix these were the following:
>
> Add the following to the $(EXTRA_REQUIREMENTS) in bgl-python\build 
> \Jamfile
> 		<define>XML_STATIC
>
> and add
>
> 	<define>XML_STATIC
>
> to the requirements for lib expat in bgl-python\vendor\expat-2.0.0 
> \Jamfile
>
> I hope that helps someone or else someone can point out what I was  
> doing
> wrong in the first place.

I think this is the right fix. Thank you! It's in SVN now.

	Cheers,
	Doug


From paustin at eos.ubc.ca  Fri Sep 29 19:33:07 2006
From: paustin at eos.ubc.ca (Philip Austin)
Date: Fri, 29 Sep 2006 10:33:07 -0700
Subject: [C++-sig] PyArray_SimpleNewFromData(, PyArray_FLOAT,
	_) can't get float c array in c extension?
In-Reply-To: <fe76f74b0609290613w6d865f95t8ccf4c8c88a9343e@mail.gmail.com>
References: <fe76f74b0609290613w6d865f95t8ccf4c8c88a9343e@mail.gmail.com>
Message-ID: <17693.22739.616286.409993@thrush.eos.ubc.ca>

Qinfeng(Javen) Shi  writes:
 > Dear All,
 > 
 > I used PyArray_SimpleNewFromData to transfer a float c array data into
 > numpy.ndarray.
 > But it got the wrong array values. Any suggestion?
 > Thanks in advance!
 > 
 > ---------------------------hello.c-------------
 > PyObject * testCreatArray()
 > {
 >     float fArray[5] = {0,1,2,3,4};
 >     intp m = 5;
 >     PyObject * c = PyArray_SimpleNewFromData(1,&m,PyArray_FLOAT,fArray);

You want PyArrayObject* c  here, then cast it

       return (PyObject*)c;
 > }



From roman.yakovenko at gmail.com  Fri Sep 29 19:39:59 2006
From: roman.yakovenko at gmail.com (Roman Yakovenko)
Date: Fri, 29 Sep 2006 20:39:59 +0300
Subject: [C++-sig] Another problem with split_module
In-Reply-To: <51F4AD92DB06E9488BBC39B7B492409E029FE9C9@vg20ex1.grp.haufemg.com>
References: <7465b6170609281325j790f8ba9t40fe37811788843c@mail.gmail.com>
	<51F4AD92DB06E9488BBC39B7B492409E029FE9C9@vg20ex1.grp.haufemg.com>
Message-ID: <7465b6170609291039q738e9de1p9c592e039a28c36b@mail.gmail.com>

On 9/29/06, Ganssauge, Gottfried <Gottfried.Ganssauge at haufe.de> wrote:
> > P.S. This time svn contains a lot of changes, also I tested
> > them pretty well I expect few bugs here and their. It could
> > be nice if you can test the latest code.
> When running my build script I get the following error:
> build.py:346: DeprecationWarning: create_casting_constructor argument is
> deprecated and should not be used.
> If you still want Py++ to generate implicitly_convertible code, consider
> to use allow_implicit_conversion constructor property
> mb = module_builder_t(...)
> mb.constructors().allow_implicit_conversion = True
>   doc_extractor = \
> Traceback (most recent call last):
>   File "build.py", line 557, in ?
>     main (args, of)
>   File "build.py", line 535, in main
>     mb.build_code_creator ()
>   File "build.py", line 346, in build_code_creator
>     doc_extractor = \
>   File
> "/usr/lib/python2.4/site-packages/pyplusplus/module_builder/builder.py",
> line 246, in build_code_creator
>     , doc_extractor)
>   File
> "/usr/lib/python2.4/site-packages/pyplusplus/module_creator/creator.py",
> line 126, in __init__
>     prepared_decls = self._prepare_decls( decls, doc_extractor )
>   File
> "/usr/lib/python2.4/site-packages/pyplusplus/module_creator/creator.py",
> line 160, in _prepare_decls
>     readme = filter( lambda msg: msg not in DO_NOT_REPORT_MSGS,
> decl.readme() )
>   File
> "/usr/lib/python2.4/site-packages/pyplusplus/decl_wrappers/decl_wrapper.
> py", line 125, in readme
>     text.extend( self._readme_impl() )
>   File
> "/usr/lib/python2.4/site-packages/pyplusplus/decl_wrappers/class_wrapper
> .py", line 456, in _readme_impl
>     return self.is_wrapper_needed()
>   File
> "/usr/lib/python2.4/site-packages/pyplusplus/decl_wrappers/class_wrapper
> .py", line 427, in is_wrapper_needed
>     if self.redefined_funcs():
>   File
> "/usr/lib/python2.4/site-packages/pyplusplus/decl_wrappers/class_wrapper
> .py", line 385, in redefined_funcs
>     is_same_function = declarations.is_same_function
> AttributeError: 'module' object has no attribute 'is_same_function'
>
> The call in line 346 of build.py is:
>         module_builder.module_builder_t.build_code_creator (
>             self,
>             module_name   = self.module_name,
>             doc_extractor = \
>                 lambda decl, extractor = doc_extractor(): \
>                     make_cstring (extractor (decl))
>
> Self is an instance derived from module_builder.module_builder_t.
>
> I don't understand the DeprecationWarning - there is nothing like a
> create_casting_constructor argument being used ...

I am going to deprecate some feature. So you can safety ignore the warning.
The error is my fault. I forgot to mention that you need to update pygccxml too.
Sorry for inconvenience.

-- 
Roman Yakovenko
C++ Python language binding
http://www.language-binding.net/


From shiqinfeng at gmail.com  Sat Sep 30 17:32:01 2006
From: shiqinfeng at gmail.com (Qinfeng(Javen) Shi )
Date: Sun, 1 Oct 2006 01:32:01 +1000
Subject: [C++-sig] (Philip Austin) PyArray_SimpleNewFromData(,
	PyArray_FLOAT, ) can't get c array in c extension?
Message-ID: <fe76f74b0609300832o34a679a3uc8009a53c6e732b6@mail.gmail.com>

Hi Philip,

I did as following, but it did not work.

--------------------------------c extension-----------------------
PyObject * testCreatArray()
{
    float fArray[5] = {0,1,2,3,4};
    PyArrayObject * c =
(PyArrayObject*)PyArray_SimpleNewFromData(1,&m,PyArray_FLOAT,fArray);
    return (PyObject *)c;
}

---------------------------result-------------
>>> import hello
>>> import numpy
>>> list1 = range(5);
>>> alist = hello.testCreatArray()
>>> print alist
[  4.83833291e-34  -2.52239988e-05  -2.52361860e-05   0.00000000e+00
   0.00000000e+00]
>>>


-- 
Qinfeng(Javen) Shi

Research School of Information Sciences and Engineering
Australian National University
Locked Bag 8001
Canberra ACT 2601
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/cplusplus-sig/attachments/20061001/a882b978/attachment.htm>

From paustin at eos.ubc.ca  Sat Sep 30 20:41:53 2006
From: paustin at eos.ubc.ca (Philip Austin)
Date: Sat, 30 Sep 2006 11:41:53 -0700
Subject: [C++-sig] PyArray_SimpleNewFromData
In-Reply-To: <fe76f74b0609300832o34a679a3uc8009a53c6e732b6@mail.gmail.com>
References: <fe76f74b0609300832o34a679a3uc8009a53c6e732b6@mail.gmail.com>
Message-ID: <17694.47729.698210.988905@thrush.eos.ubc.ca>

Qinfeng(Javen) Shi  writes:
 > Hi Philip,
 > 
 > I did as following, but it did not work.
 > 
 > --------------------------------c extension-----------------------
 > PyObject * testCreatArray()
 > {
 >     float fArray[5] = {0,1,2,3,4};
 >     PyArrayObject * c =
 > (PyArrayObject*)PyArray_SimpleNewFromData(1,&m,PyArray_FLOAT,fArray);
 >     return (PyObject *)c;
 > }

As Travis notes (on p. 187 of my copy of the numpy book), you have to
guarantee that fArray lasts at least as long as the numpy array
that contains it.  fArray is destroyed when the function returns --
the easiest way to get around this is to let python allocate the
memory and copy fArray into it:

something like (untested)

int nd=1;
intp m=5;

PyObject* c=PyArray_SimpleNew(nd, &m, PyArray_Float);

then get the pointer to the data:

void *arr_data = PyArray_DATA((PyArrayObject*)c);

and copy

memcpy(arr_data, fArray, PyArray_ITEMSIZE((PyArrayObject*) c) * m);

if this copy is too expensive, boost has ways of linking the lifetime
of the numpy array to a pointer to the array data (scan the boost
reference for with_custodian_and_ward), so you could for example create
a shared pointer to the data, skip the copy and use
PyArray_SimpleNewFromData, but so far I haven't needed to do that for
my applications.

 -- Phil






From jlisee at gmail.com  Fri Sep 29 23:54:10 2006
From: jlisee at gmail.com (Joseph Lisee)
Date: Fri, 29 Sep 2006 17:54:10 -0400
Subject: [C++-sig] Translating "type()"
Message-ID: <pan.2006.09.29.21.54.09.939778@gmail.com>

I have a line of python code that generates a new class:

class_name = "NewClass"
NewClass = type(class_name, (object,), {})

I have tried to replicate this in Boost.Python with this:

object NewClass = class_<object>(class_name);

This give warnings about registering a to-python convert more than once. 
What is the proper way to do this, ie: Generate a new class with the given
name.

-Joe