From gr at componic.co.nz  Wed Aug  1 02:07:08 2012
From: gr at componic.co.nz (Glenn Ramsey)
Date: Wed, 01 Aug 2012 12:07:08 +1200
Subject: [C++-sig] How to create a custom exception derived from
	Exception
In-Reply-To: <5017BA3F.6060403@gmail.com>
References: <5017A1FA.8050502@componic.co.nz> <5017BA3F.6060403@gmail.com>
Message-ID: <5018732C.3090309@componic.co.nz>

On 31/07/12 22:58, Jim Bosch wrote:
> On 07/31/2012 06:14 PM, Glenn Ramsey wrote:
>>
>> Using boost::python how can I create an a custom exception that is
>> derived from Python's Exception? The functionality I would like to get
>> by doing this is to provide additional methods for the catching code to
>> retrieve information from the exception.
>>
>
> The approach I've taken in the past (sorry, don't have an example right now) is
> to raise a true Python exception (from PyErr_NewException), and attach your
> Boost.Python-wrapped C++ exception to it as an instance variable.  Then put a
> __getattr__ in the true Python exception that forwards to the Boost.Python
> object it holds.
>
> I can try to reconstruct more details if needed, but hopefully that's enough to
> get you started.

Thanks, based on that I have done this, which seems to work, but I'm not sure if 
it is entirely correct. I didn't need to implement __getattr__ and would be 
interested in seeing some more details of your approach.

PyObject *exceptionType=NULL; // will be initialised inside 
BOOST_PYTHON_MODULE(...) using createExceptionClass(...)

void translator(const MyCPPException &x) {
	
	bp::object exc(x);

	bp::object exc_t(bp::handle<>(bp::borrowed(exceptionType)));
	exc_t.attr("cause") = exc;

	PyErr_SetString(exceptionType, x.what());
}

PyObject* createExceptionClass(const char* name, PyObject* baseTypeObj = 
PyExc_Exception)
{
     std::string scopeName = bp::extract<std::string (bp::scope().attr("__name__"));

     std::string qualifiedName0 = scopeName + "." + name;

     char* qualifiedName1 = const_cast<char*>(qualifiedName0.c_str());

     PyObject* typeObj = PyErr_NewException(qualifiedName1, baseTypeObj, 0);
     if(!typeObj)
     {
        bp::throw_error_already_set();
     }

     bp::scope().attr(name) = bp::handle<>(bp::borrowed(typeObj));
     return typeObj;
}

BOOST_PYTHON_MODULE(MyModule)
{
     bp::register_exception_translator<MyCPPException>(translator);
     ....
     bp::class_<MyCPPException>("MyCPPException")
         .def(...);
     exceptionType = createExceptionClass("MyCPPExceptionType");
     ....
}


In Python:

try:
     ...
except MyModule.MyCPPExceptionType as e:
    cause = e.cause # wrapped exception can be accessed here
     ...


Glenn



From talljimbo at gmail.com  Thu Aug  2 05:28:21 2012
From: talljimbo at gmail.com (Jim Bosch)
Date: Thu, 02 Aug 2012 12:28:21 +0900
Subject: [C++-sig] How to create a custom exception derived from
	Exception
In-Reply-To: <5018732C.3090309@componic.co.nz>
References: <5017A1FA.8050502@componic.co.nz> <5017BA3F.6060403@gmail.com>
	<5018732C.3090309@componic.co.nz>
Message-ID: <5019F3D5.4020700@gmail.com>

On 08/01/2012 09:07 AM, Glenn Ramsey wrote:
> On 31/07/12 22:58, Jim Bosch wrote:
>> On 07/31/2012 06:14 PM, Glenn Ramsey wrote:
>>>
>>> Using boost::python how can I create an a custom exception that is
>>> derived from Python's Exception? The functionality I would like to get
>>> by doing this is to provide additional methods for the catching code to
>>> retrieve information from the exception.
>>>
>>
>> The approach I've taken in the past (sorry, don't have an example
>> right now) is
>> to raise a true Python exception (from PyErr_NewException), and attach
>> your
>> Boost.Python-wrapped C++ exception to it as an instance variable.
>> Then put a
>> __getattr__ in the true Python exception that forwards to the
>> Boost.Python
>> object it holds.
>>
>> I can try to reconstruct more details if needed, but hopefully that's
>> enough to
>> get you started.
>
> Thanks, based on that I have done this, which seems to work, but I'm not
> sure if it is entirely correct. I didn't need to implement __getattr__
> and would be interested in seeing some more details of your approach.
>

Looks fine to me.  I'm afraid I don't really have any more details - I 
last did this a while ago, and I don't exactly remember where it was 
anymore.  But it seems pretty similar to what I did.

Jim



From josh.davidson at lmco.com  Sun Aug 12 23:55:40 2012
From: josh.davidson at lmco.com (Davidson, Josh)
Date: Sun, 12 Aug 2012 21:55:40 +0000
Subject: [C++-sig] pygccxml isn't working with Python 2.7 win64
Message-ID: <86299D4CFE2C1248AA41ACF782B0D143118FD575@HDXDSP33.us.lmco.com>

I recently updated to Python 2.7 from Python 2.6 and am running into issues with pygccxml.  Previously, I was on:
Python 2.6.6 (r266:84297, Aug 24 2010, 18:13:38) [MSC v.1500 64 bit (AMD64)] on win32
I'm now on:
Python 2.7.3 (default, Apr 10 2012, 23:24:47) [MSC v.1500 64 bit (AMD64)] on win32

On both systems, I have the latest versions of pygccxml/py++ installed from SVN.  I've diff'd the versions installed in both interpreter's site-packages directory and they match exactly.  When I run on Python 2.7.3, the suprocess call to gccxml barfs.  Here is an example:

INFO Parsing source file "D:\SimExternalInterfaces\src\C++\SimInterfaces\SimExecInterfaces\temp.executive.hpp" ...
INFO gccxml cmd: ""C:\MinGW\bin\gccxml.exe"  -DBOOST_NO_RVALUE_REFERENCES -D__mode__(x)= -DMS_WIN64   -I"." -I"D:\SimExternalInterfaces\src\C++" -I
ExternalInterfaces\src\C++\SimInterfaces\SimExecInterfaces" -I"C:\Python27\include" -I"D:\SimExternalInterfaces\src\C++\SimInterfaces\SimExecInterf
 "D:\SimExternalInterfaces\src\C++\SimInterfaces\SimExecInterfaces\temp.executive.hpp" -fxml="c:\users\josh\appdata\local\temp\tmp0h9nvf.xml""
COMMAND LINE*""C:\MinGW\bin\gccxml.exe"  -DBOOST_NO_RVALUE_REFERENCES -D__mode__(x)= -DMS_WIN64   -I"." -I"D:\SimExternalInterfaces\src\C++" -I"D:\
rnalInterfaces\src\C++\SimInterfaces\SimExecInterfaces" -I"C:\Python27\include" -I"D:\SimExternalInterfaces\src\C++\SimInterfaces\SimExecInterfaces
\SimExternalInterfaces\src\C++\SimInterfaces\SimExecInterfaces\temp.executive.hpp" -fxml="c:\users\josh\appdata\local\temp\tmp0h9nvf.xml""*
<type 'str'>
LENGTH 433
LINE The input line is too long.
LINE
LINE
Traceback (most recent call last):
  File "C:\Python27\lib\site-packages\pygccxml\parser\source_reader.py", line 229, in read_gccxml_file
    gccxml_file = self.create_xml_file( ffname )
  File "C:\Python27\lib\site-packages\pygccxml\parser\source_reader.py", line 185, in create_xml_file
    raise error
gccxml_runtime_error_t: Error occured while running GCC-XML: The input line is too long.
ffname D:\SimExternalInterfaces\src\C++\SimInterfaces\SimExecInterfaces\temp.executive.hpp

There are a few extra print statements at the end, which line up to some debugging code I inserted into source_reader.py:
            print "COMMAND LINE*%s*" % (command_line,)
            print type(command_line)
            print "LENGTH", len(command_line)

The few lines beginning with "LINE" are just prints from the loop that reads the stdout pipe after spawning gccxml.  As you can see, the line is only 433 characters which is well within Windows limits.  I've moved the project all over the place and no matter where it's at, I get this error.

If I switch back to Python26, this is what I get:
INFO Parsing source file "D:\SimExternalInterfaces\src\C++\SimInterfaces\SimExecInterfaces\temp.executive.hpp" ...
INFO gccxml cmd: ""C:\MinGW\bin\gccxml.exe"  -DBOOST_NO_RVALUE_REFERENCES -D__mode__(x)= -DMS_WIN64   -I"." -I"D:\SimExternalInterfaces\src\C++" -I
ExternalInterfaces\src\C++\SimInterfaces\SimExecInterfaces" -I"C:\Python26\include" -I"D:\SimExternalInterfaces\src\C++\SimInterfaces\SimExecInterf
 "D:\SimExternalInterfaces\src\C++\SimInterfaces\SimExecInterfaces\temp.executive.hpp" -fxml="c:\users\josh\appdata\local\temp\tmpdgchpg.xml""
COMMAND LINE*""C:\MinGW\bin\gccxml.exe"  -DBOOST_NO_RVALUE_REFERENCES -D__mode__(x)= -DMS_WIN64   -I"." -I"D:\SimExternalInterfaces\src\C++" -I"D:\
rnalInterfaces\src\C++\SimInterfaces\SimExecInterfaces" -I"C:\Python26\include" -I"D:\SimExternalInterfaces\src\C++\SimInterfaces\SimExecInterfaces
\SimExternalInterfaces\src\C++\SimInterfaces\SimExecInterfaces\temp.executive.hpp" -fxml="c:\users\josh\appdata\local\temp\tmpdgchpg.xml""*
<type 'str'>
LENGTH 433
LINE
INFO GCCXML version - 0.9( 1.135 ) 
<snip />

As you can see, the command line is basically the same save the references to the different Python installation.  Length is also the same.  It appears that someting in subprocess changed in 2.7 that doesn't play nicely with how gccxml is being launched.

From josh.davidson at lmco.com  Mon Aug 13 00:51:14 2012
From: josh.davidson at lmco.com (Davidson, Josh)
Date: Sun, 12 Aug 2012 22:51:14 +0000
Subject: [C++-sig] pygccxml isn't working with Python 2.7 win64
In-Reply-To: <86299D4CFE2C1248AA41ACF782B0D143118FD596@HDXDSP33.us.lmco.com>
References: <86299D4CFE2C1248AA41ACF782B0D143118FD575@HDXDSP33.us.lmco.com>
	<86299D4CFE2C1248AA41ACF782B0D143118FD596@HDXDSP33.us.lmco.com>
Message-ID: <86299D4CFE2C1248AA41ACF782B0D143118FD5A8@HDXDSP33.us.lmco.com>

Let's try this once more...

122c122
<             cmd_line = '"%s"' % cmd_line
---
>             cmd_line = '%s' % cmd_line
154c154
<                                         , shell=True
---
>                                         , shell=False if 'nt' == os.name else True

Command line isn't being split, so shell=True is required for *nix.
________________________________________
From: Davidson, Josh
Sent: Sunday, August 12, 2012 4:40 PM
To: cplusplus-sig at python.org
Subject: RE: pygccxml isn't working with Python 2.7 win64

Here is a patch that works for both 2.6 & 2.7:


D:\SimExternalInterfaces>diff C:\temp\pygccxml\pygccxml_dev\pygccxml\parser\source_reader.py C:\Python27\Lib\site-packages\pygccxml\parser\source_reader.py
122c122
<             cmd_line = '"%s"' % cmd_line
---
>             cmd_line = '%s' % cmd_line
154c154
<                                         , shell=True
---
>                                         , shell=False

From josh.davidson at lmco.com  Mon Aug 13 00:40:14 2012
From: josh.davidson at lmco.com (Davidson, Josh)
Date: Sun, 12 Aug 2012 22:40:14 +0000
Subject: [C++-sig] pygccxml isn't working with Python 2.7 win64
In-Reply-To: <86299D4CFE2C1248AA41ACF782B0D143118FD575@HDXDSP33.us.lmco.com>
References: <86299D4CFE2C1248AA41ACF782B0D143118FD575@HDXDSP33.us.lmco.com>
Message-ID: <86299D4CFE2C1248AA41ACF782B0D143118FD596@HDXDSP33.us.lmco.com>

Here is a patch that works for both 2.6 & 2.7:


D:\SimExternalInterfaces>diff C:\temp\pygccxml\pygccxml_dev\pygccxml\parser\source_reader.py C:\Python27\Lib\site-packages\pygccxml\parser\source_reader.py
122c122
<             cmd_line = '"%s"' % cmd_line
---
>             cmd_line = '%s' % cmd_line
154c154
<                                         , shell=True
---
>                                         , shell=False

________________________________________
From: Davidson, Josh
Sent: Sunday, August 12, 2012 3:55 PM
To: cplusplus-sig at python.org
Subject: pygccxml isn't working with Python 2.7 win64

I recently updated to Python 2.7 from Python 2.6 and am running into issues with pygccxml.  Previously, I was on:
Python 2.6.6 (r266:84297, Aug 24 2010, 18:13:38) [MSC v.1500 64 bit (AMD64)] on win32
I'm now on:
Python 2.7.3 (default, Apr 10 2012, 23:24:47) [MSC v.1500 64 bit (AMD64)] on win32

On both systems, I have the latest versions of pygccxml/py++ installed from SVN.  I've diff'd the versions installed in both interpreter's site-packages directory and they match exactly.  When I run on Python 2.7.3, the suprocess call to gccxml barfs.  Here is an example:

INFO Parsing source file "D:\SimExternalInterfaces\src\C++\SimInterfaces\SimExecInterfaces\temp.executive.hpp" ...
INFO gccxml cmd: ""C:\MinGW\bin\gccxml.exe"  -DBOOST_NO_RVALUE_REFERENCES -D__mode__(x)= -DMS_WIN64   -I"." -I"D:\SimExternalInterfaces\src\C++" -I
ExternalInterfaces\src\C++\SimInterfaces\SimExecInterfaces" -I"C:\Python27\include" -I"D:\SimExternalInterfaces\src\C++\SimInterfaces\SimExecInterf
 "D:\SimExternalInterfaces\src\C++\SimInterfaces\SimExecInterfaces\temp.executive.hpp" -fxml="c:\users\josh\appdata\local\temp\tmp0h9nvf.xml""
COMMAND LINE*""C:\MinGW\bin\gccxml.exe"  -DBOOST_NO_RVALUE_REFERENCES -D__mode__(x)= -DMS_WIN64   -I"." -I"D:\SimExternalInterfaces\src\C++" -I"D:\
rnalInterfaces\src\C++\SimInterfaces\SimExecInterfaces" -I"C:\Python27\include" -I"D:\SimExternalInterfaces\src\C++\SimInterfaces\SimExecInterfaces
\SimExternalInterfaces\src\C++\SimInterfaces\SimExecInterfaces\temp.executive.hpp" -fxml="c:\users\josh\appdata\local\temp\tmp0h9nvf.xml""*
<type 'str'>
LENGTH 433
LINE The input line is too long.
LINE
LINE
Traceback (most recent call last):
  File "C:\Python27\lib\site-packages\pygccxml\parser\source_reader.py", line 229, in read_gccxml_file
    gccxml_file = self.create_xml_file( ffname )
  File "C:\Python27\lib\site-packages\pygccxml\parser\source_reader.py", line 185, in create_xml_file
    raise error
gccxml_runtime_error_t: Error occured while running GCC-XML: The input line is too long.
ffname D:\SimExternalInterfaces\src\C++\SimInterfaces\SimExecInterfaces\temp.executive.hpp

There are a few extra print statements at the end, which line up to some debugging code I inserted into source_reader.py:
            print "COMMAND LINE*%s*" % (command_line,)
            print type(command_line)
            print "LENGTH", len(command_line)

The few lines beginning with "LINE" are just prints from the loop that reads the stdout pipe after spawning gccxml.  As you can see, the line is only 433 characters which is well within Windows limits.  I've moved the project all over the place and no matter where it's at, I get this error.

If I switch back to Python26, this is what I get:
INFO Parsing source file "D:\SimExternalInterfaces\src\C++\SimInterfaces\SimExecInterfaces\temp.executive.hpp" ...
INFO gccxml cmd: ""C:\MinGW\bin\gccxml.exe"  -DBOOST_NO_RVALUE_REFERENCES -D__mode__(x)= -DMS_WIN64   -I"." -I"D:\SimExternalInterfaces\src\C++" -I
ExternalInterfaces\src\C++\SimInterfaces\SimExecInterfaces" -I"C:\Python26\include" -I"D:\SimExternalInterfaces\src\C++\SimInterfaces\SimExecInterf
 "D:\SimExternalInterfaces\src\C++\SimInterfaces\SimExecInterfaces\temp.executive.hpp" -fxml="c:\users\josh\appdata\local\temp\tmpdgchpg.xml""
COMMAND LINE*""C:\MinGW\bin\gccxml.exe"  -DBOOST_NO_RVALUE_REFERENCES -D__mode__(x)= -DMS_WIN64   -I"." -I"D:\SimExternalInterfaces\src\C++" -I"D:\
rnalInterfaces\src\C++\SimInterfaces\SimExecInterfaces" -I"C:\Python26\include" -I"D:\SimExternalInterfaces\src\C++\SimInterfaces\SimExecInterfaces
\SimExternalInterfaces\src\C++\SimInterfaces\SimExecInterfaces\temp.executive.hpp" -fxml="c:\users\josh\appdata\local\temp\tmpdgchpg.xml""*
<type 'str'>
LENGTH 433
LINE
INFO GCCXML version - 0.9( 1.135 )
<snip />

As you can see, the command line is basically the same save the references to the different Python installation.  Length is also the same.  It appears that someting in subprocess changed in 2.7 that doesn't play nicely with how gccxml is being launched.

From atalay at udel.edu  Tue Aug 14 01:05:25 2012
From: atalay at udel.edu (Murat Atalay)
Date: Mon, 13 Aug 2012 16:05:25 -0700
Subject: [C++-sig] Interact python with C++
Message-ID: <CAFBENeGxNQ-rDAhRJ0Z7zHi3h993ufq_N69hJQ61u9R18GQq5Q@mail.gmail.com>

Hello Everyone,

Currently, I am trying to call C++ classes and functions from python inside
my C++ program. I want to be able to edit the members and call functions
from python. I am using VS2010 with boost libraries. Boost comes with very
big embedding frame work and SWIG mix of C and C++ files and I don't want
to create DLLs. So, I was wondering if there is another tool to interact
python with C++ and vise versa.

I also have an issue when I use boost. I keep getting an python version
mismatch error. I was wondering what could cause this.

Regards,
Murat
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/cplusplus-sig/attachments/20120813/10dce138/attachment.html>

From j.reid at mail.cryst.bbk.ac.uk  Tue Aug 14 11:32:16 2012
From: j.reid at mail.cryst.bbk.ac.uk (John Reid)
Date: Tue, 14 Aug 2012 10:32:16 +0100
Subject: [C++-sig] Interact python with C++
In-Reply-To: <CAFBENeGxNQ-rDAhRJ0Z7zHi3h993ufq_N69hJQ61u9R18GQq5Q@mail.gmail.com>
References: <CAFBENeGxNQ-rDAhRJ0Z7zHi3h993ufq_N69hJQ61u9R18GQq5Q@mail.gmail.com>
Message-ID: <k0d5uv$dlg$1@dough.gmane.org>


On 14/08/12 00:05, Murat Atalay wrote:
> Hello Everyone,
> 
> Currently, I am trying to call C++ classes and functions from python
> inside my C++ program. I want to be able to edit the members and call
> functions from python. I am using VS2010 with boost libraries. Boost
> comes with very big embedding frame work and SWIG mix of C and C++ files
> and I don't want to create DLLs. So, I was wondering if there is another
> tool to interact python with C++ and vise versa. 

What do you mean "edit the members"? Also, I'm not sure how you can
interact with C++ from python without creating a DLL.

> 
> I also have an issue when I use boost. I keep getting an python version
> mismatch error. I was wondering what could cause this. 

It is hard to help without significantly more information.

Regards,
John.


From alfarobi0161 at yahoo.com  Tue Aug 14 13:25:25 2012
From: alfarobi0161 at yahoo.com (shahmi)
Date: Tue, 14 Aug 2012 19:25:25 +0800 (SGT)
Subject: [C++-sig] Accessing 2D array from python in C++
Message-ID: <1344943525.62630.YahooMailNeo@web190601.mail.sg3.yahoo.com>

Hi, 


I have tried to access/expose 2D array from python in C++ using 
boost::multi_array. Nevertheless, it seems that the 
boost::multi_array<double, 2>?datatype is not supported. Is there 
any way to resolve this? My way of accessing it as below: 

? ? ? ? ? ? boost::python::object Vp = main_namespace["Vp"]; 
? ? ? ? ? ? Vp_cpp = boost::python::extract<boost::multi_array<double, 2>?>(Vp); 


I've also tried to try to use 2D vector as below, nonetheless, it still doesn't work:

??? ??? boost::python::object Vp = main_namespace["Vp"];
??? ??? Vp_cpp = boost::python::extract<std::vector<std::vector<double> > >(Vp);

The error message I received is "TypeError: No registered converter was able to produce a C++ rvalue of type std::vector<std::vector<double, std::allocator<double> >, std::allocator<std::vector<double, std::allocator<double> > > > from this Python object of type matrix"

Thanks in advance.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/cplusplus-sig/attachments/20120814/0fb0bd43/attachment.html>

From s_sourceforge at nedprod.com  Tue Aug 14 13:34:33 2012
From: s_sourceforge at nedprod.com (Niall Douglas)
Date: Tue, 14 Aug 2012 12:34:33 +0100
Subject: [C++-sig] Interact python with C++
In-Reply-To: <CAFBENeGxNQ-rDAhRJ0Z7zHi3h993ufq_N69hJQ61u9R18GQq5Q@mail.gmail.com>
References: <CAFBENeGxNQ-rDAhRJ0Z7zHi3h993ufq_N69hJQ61u9R18GQq5Q@mail.gmail.com>
Message-ID: <502A37C9.3972.A295A98@s_sourceforge.nedprod.com>

Could use Cython. You're still going to have to use DLLs though. With 
Python it's very hard to avoid DLLs.

Niall


On 13 Aug 2012 at 16:05, Murat Atalay wrote:

> Hello Everyone,
> 
> Currently, I am trying to call C++ classes and functions from python inside
> my C++ program. I want to be able to edit the members and call functions
> from python. I am using VS2010 with boost libraries. Boost comes with very
> big embedding frame work and SWIG mix of C and C++ files and I don't want
> to create DLLs. So, I was wondering if there is another tool to interact
> python with C++ and vise versa.
> 
> I also have an issue when I use boost. I keep getting an python version
> mismatch error. I was wondering what could cause this.
> 
> Regards,
> Murat
> 


-- 
Technology & Consulting Services - ned Productions Limited.
http://www.nedproductions.biz/. VAT reg: IE 9708311Q.
Work Portfolio: http://careers.stackoverflow.com/nialldouglas/



-------------- next part --------------
A non-text attachment was scrubbed...
Name: SMime.p7s
Type: application/x-pkcs7-signature
Size: 6061 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/cplusplus-sig/attachments/20120814/f3e1c596/attachment.bin>

From talljimbo at gmail.com  Tue Aug 14 17:54:16 2012
From: talljimbo at gmail.com (Jim Bosch)
Date: Tue, 14 Aug 2012 08:54:16 -0700
Subject: [C++-sig] Accessing 2D array from python in C++
In-Reply-To: <1344943525.62630.YahooMailNeo@web190601.mail.sg3.yahoo.com>
References: <1344943525.62630.YahooMailNeo@web190601.mail.sg3.yahoo.com>
Message-ID: <CAN+0q26pWg94qrWh91YP=1XOG3xfLh+WAkL+pR6+x_f7nVpF-w@mail.gmail.com>

On Aug 14, 2012 4:25 AM, "shahmi" <alfarobi0161 at yahoo.com> wrote:
>
> Hi,
>
> I have tried to access/expose 2D array from python in C++ using
boost::multi_array. Nevertheless, it seems that the
boost::multi_array<double, 2> datatype is not supported.  Is there any way
to resolve this? My way of accessing it as below:
>
>             boost::python::object Vp = main_namespace["Vp"];
>             Vp_cpp = boost::python::extract<boost::multi_array<double,
2> >(Vp);
>
> I've also tried to try to use 2D vector as below, nonetheless, it still
doesn't work:
>
>         boost::python::object Vp = main_namespace["Vp"];
>         Vp_cpp = boost::python::extract<std::vector<std::vector<double> >
>(Vp);
>
> The error message I received is "TypeError: No registered converter was
able to produce a C++ rvalue of type std::vector<std::vector<double,
std::allocator<double> >, std::allocator<std::vector<double,
std::allocator<double> > > > from this Python object of type matrix"
>

In general Boost.Python doesn't provide converters for all the other boost
libraries.

The first question is what you want to return in Python; you could wrap
multi_array itself using class_.  You could also return a nested list or
tuple.
If you want to use numpy, you'll probably want a helper library like this
one:
https://github.com/ndarray/Boost.NumPy

In either of the last two cases, you can either write your own converter
and add it to the converter registry (harder, but reusable), or write a
custom wrapper for the function that returns a 2-d array (easier, but not
good if you have many such functions).

HTH

Jim
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/cplusplus-sig/attachments/20120814/c803f979/attachment.html>

From alfarobi0161 at yahoo.com  Wed Aug 15 10:47:30 2012
From: alfarobi0161 at yahoo.com (alfa)
Date: Wed, 15 Aug 2012 16:47:30 +0800 (SGT)
Subject: [C++-sig] Accessing 2D array from python in C++
In-Reply-To: <CAN+0q26pWg94qrWh91YP=1XOG3xfLh+WAkL+pR6+x_f7nVpF-w@mail.gmail.com>
References: <1344943525.62630.YahooMailNeo@web190601.mail.sg3.yahoo.com>
	<CAN+0q26pWg94qrWh91YP=1XOG3xfLh+WAkL+pR6+x_f7nVpF-w@mail.gmail.com>
Message-ID: <1345020450.71023.YahooMailNeo@web190606.mail.sg3.yahoo.com>

Hi Jim,

Do you mean, "you could wrap multi_array itself using class_", I should do as below:

boost::python::class_<std::vector<std::vector<double> > >("PyVec").def(boost::python::vector_indexing_suite<std::vector<std::vector<double> > >());


I also did as below to expose the data from python to C++:

main_namespace["Vp"] = boost::python::ptr(&Vp_cpp);

However, when I try to access the data to the std::cout, like the following, it complains:

std::cout << "Vp_cpp = " << Vp_cpp << std::endl;

What could be wrong or am I missing something? Thanks in advance.




________________________________
 From: Jim Bosch <talljimbo at gmail.com>
To: Development of Python/C++ integration <cplusplus-sig at python.org>; shahmi <alfarobi0161 at yahoo.com> 
Sent: Tuesday, August 14, 2012 5:54 PM
Subject: Re: [C++-sig] Accessing 2D array from python in C++
 


On Aug 14, 2012 4:25 AM, "shahmi" <alfarobi0161 at yahoo.com> wrote:
>
> Hi, 
>
> I have tried to access/expose 2D array from python in C++ using boost::multi_array. Nevertheless, it seems that the boost::multi_array<double, 2>?datatype is not supported.? Is there any way to resolve this? My way of accessing it as below: 
>
> ? ? ? ? ? ? boost::python::object Vp = main_namespace["Vp"]; 
> ? ? ? ? ? ? Vp_cpp = boost::python::extract<boost::multi_array<double, 2>?>(Vp); 
>
> I've also tried to try to use 2D vector as below, nonetheless, it still doesn't work:
>
> ??? ??? boost::python::object Vp = main_namespace["Vp"];
> ??? ??? Vp_cpp = boost::python::extract<std::vector<std::vector<double> > >(Vp);
>
> The error message I received is "TypeError: No registered converter was able to produce a C++ rvalue of type std::vector<std::vector<double, std::allocator<double> >, std::allocator<std::vector<double, std::allocator<double> > > > from this Python object of type matrix"
>
In general Boost.Python doesn't provide converters for all the other boost libraries.
The first question is what you want to return in Python; you could wrap multi_array itself using class_.? You could also return a nested list or tuple.
If you want to use numpy, you'll probably want a helper library like this one:
https://github.com/ndarray/Boost.NumPy
In either of the last two cases, you can either write your own converter and add it to the converter registry (harder, but reusable), or write a custom wrapper for the function that returns a 2-d 	array (easier, but not good if you have many such functions).
HTH
Jim
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/cplusplus-sig/attachments/20120815/bb42c7be/attachment-0001.html>

From alfarobi0161 at yahoo.com  Wed Aug 15 10:57:07 2012
From: alfarobi0161 at yahoo.com (alfa)
Date: Wed, 15 Aug 2012 16:57:07 +0800 (SGT)
Subject: [C++-sig] Accessing 2D array from python in C++
In-Reply-To: <CAN+0q26pWg94qrWh91YP=1XOG3xfLh+WAkL+pR6+x_f7nVpF-w@mail.gmail.com>
References: <1344943525.62630.YahooMailNeo@web190601.mail.sg3.yahoo.com>
	<CAN+0q26pWg94qrWh91YP=1XOG3xfLh+WAkL+pR6+x_f7nVpF-w@mail.gmail.com>
Message-ID: <1345021027.75677.YahooMailNeo@web190606.mail.sg3.yahoo.com>

I apologize for double posting this because it was spammed before.

Hi Jim,

Do you mean, "you could wrap multi_array itself using class_", I should do as below:

boost::python::class_<std::vector<std::vector<double> > 
>("PyVec").def(boost::python::vector_indexing_suite<std::vector<std::vector<double> > >());


I also did as below to expose the data from python to C++:

main_namespace["Vp"] = boost::python::ptr(&Vp_cpp);

However, when I try to access the data to the std::cout, like the following, it complains:

std::cout << "Vp_cpp = " << Vp_cpp << std::endl;
What could be wrong or am I missing something? Thanks in advance.


________________________________
 From: Jim Bosch <talljimbo at gmail.com>
To: Development of Python/C++ integration <cplusplus-sig at python.org>; shahmi <alfarobi0161 at yahoo.com> 
Sent: Tuesday, August 14, 2012 5:54 PM
Subject: Re: [C++-sig] Accessing 2D array from python in C++
 


On Aug 14, 2012 4:25 AM, "shahmi" <alfarobi0161 at yahoo.com> wrote:
>
> Hi, 
>
> I have tried to access/expose 2D array from python in C++ using boost::multi_array. Nevertheless, it seems that the boost::multi_array<double, 2>?datatype is not supported.? Is there any way to resolve this? My way of accessing it as below: 
>
> ? ? ? ? ? ? boost::python::object Vp = main_namespace["Vp"]; 
> ? ? ? ? ? ? Vp_cpp = boost::python::extract<boost::multi_array<double, 2>?>(Vp); 
>
> I've also tried to try to use 2D vector as below, nonetheless, it still doesn't work:
>
> ??? ??? boost::python::object Vp = main_namespace["Vp"];
> ??? ??? Vp_cpp = boost::python::extract<std::vector<std::vector<double> > >(Vp);
>
> The error message I received is "TypeError: No registered converter was able to produce a C++ rvalue of type std::vector<std::vector<double, std::allocator<double> >, std::allocator<std::vector<double, std::allocator<double> > > > from this Python object of type matrix"
>
In general Boost.Python doesn't provide converters for all the other boost libraries.
The first question is what you want to return in Python; you could wrap multi_array itself using class_.? You could also return a nested list or tuple.
If you want to use numpy, you'll probably want a helper library like this one:
https://github.com/ndarray/Boost.NumPy
In either of the last two cases, you can either write your own converter and add it to the converter registry (harder, but reusable), or write a custom wrapper for the function that returns a 2-d 	array (easier, but not good if you have many such functions).
HTH
Jim
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/cplusplus-sig/attachments/20120815/5555a485/attachment.html>

From alfarobi0161 at yahoo.com  Wed Aug 15 11:01:35 2012
From: alfarobi0161 at yahoo.com (alfa)
Date: Wed, 15 Aug 2012 17:01:35 +0800 (SGT)
Subject: [C++-sig] Fw:  Accessing 2D array from python in C++
In-Reply-To: <1345021027.75677.YahooMailNeo@web190606.mail.sg3.yahoo.com>
References: <1344943525.62630.YahooMailNeo@web190601.mail.sg3.yahoo.com>
	<CAN+0q26pWg94qrWh91YP=1XOG3xfLh+WAkL+pR6+x_f7nVpF-w@mail.gmail.com>
	<1345021027.75677.YahooMailNeo@web190606.mail.sg3.yahoo.com>
Message-ID: <1345021295.20509.YahooMailNeo@web190602.mail.sg3.yahoo.com>




----- Forwarded Message -----
From: alfa <alfarobi0161 at yahoo.com>
To: Jim Bosch <talljimbo at gmail.com>; Development of Python/C++ integration <cplusplus-sig at python.org> 
Sent: Wednesday, August 15, 2012 10:57 AM
Subject: Re: [C++-sig] Accessing 2D array from python in C++
 

I apologize for double posting this because it was spammed before.

Hi Jim,

Do you mean, "you could wrap multi_array itself using class_", I should do as below:

boost::python::class_<std::vector<std::vector<double> > 
>("PyVec").def(boost::python::vector_indexing_suite<std::vector<std::vector<double> > >());


I also did as below to expose the data from python to C++:

main_namespace["Vp"] = boost::python::ptr(&Vp_cpp);

However, when I try to access the data to the std::cout, like the following, it complains:

std::cout << "Vp_cpp = " << Vp_cpp << std::endl;
What could be wrong or am I missing something? Thanks in advance.


________________________________
 From: Jim Bosch <talljimbo at gmail.com>
To: Development of Python/C++ integration <cplusplus-sig at python.org>; shahmi <alfarobi0161 at yahoo.com> 
Sent: Tuesday, August 14, 2012 5:54 PM
Subject: Re: [C++-sig] Accessing 2D array from python in C++
 


On Aug 14, 2012 4:25 AM, "shahmi" <alfarobi0161 at yahoo.com> wrote:
>
> Hi, 
>
> I have tried to access/expose 2D array from python in C++ using boost::multi_array. Nevertheless, it seems that the boost::multi_array<double, 2>?datatype is not supported.? Is there any way to resolve this? My way of accessing it as below: 
>
> ? ? ? ? ? ? boost::python::object Vp = main_namespace["Vp"]; 
> ? ? ? ? ? ? Vp_cpp = boost::python::extract<boost::multi_array<double, 2>?>(Vp); 
>
> I've also tried to try to use 2D vector as below, nonetheless, it still doesn't work:
>
> ??? ??? boost::python::object Vp = main_namespace["Vp"];
> ??? ??? Vp_cpp = boost::python::extract<std::vector<std::vector<double> > >(Vp);
>
> The error message I received is "TypeError: No registered converter was able to produce a C++ rvalue of type std::vector<std::vector<double, std::allocator<double> >, std::allocator<std::vector<double, std::allocator<double> > > > from this Python object of type matrix"
>
In general Boost.Python doesn't provide converters for all the other boost libraries.
The first question is what you want to return in Python; you could wrap multi_array itself using class_.? You could also return a nested list or tuple.
If you want to use numpy, you'll probably want a helper library like this one:
https://github.com/ndarray/Boost.NumPy
In either of the last two cases, you can either write your own converter and add it to the converter registry (harder, but reusable), or write a custom wrapper for the function that returns a 2-d 	array (easier, but not good if you have many such functions).
HTH
Jim


_______________________________________________
Cplusplus-sig mailing list
Cplusplus-sig at python.org
http://mail.python.org/mailman/listinfo/cplusplus-sig
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/cplusplus-sig/attachments/20120815/4f3f1159/attachment.html>

From alfarobi0161 at yahoo.com  Wed Aug 15 15:05:17 2012
From: alfarobi0161 at yahoo.com (alfa)
Date: Wed, 15 Aug 2012 21:05:17 +0800 (SGT)
Subject: [C++-sig] Problem exposing 2D python array  in C++
Message-ID: <1345035917.24211.YahooMailNeo@web190604.mail.sg3.yahoo.com>

Hi,


Is the following way valid to expose a 2D array variable from python (Vp)? to c++(Vp_cpp)?
?? 

std::vector<std::vector<double> > Vp_cpp;
boost::python::class_<std::vector<std::vector<double> > >("PyVec").def(boost::python::vector_indexing_suite<std::vector<std::vector<double> > >());


main_namespace["Vp"] = boost::python::ptr(&Vp_cpp);

But, then if I tried to manipulate the 2D vector in C++ like the following, it doesn't work:

Vp_cpp[1][1] = 999;
std::cout << "Vp_cpp[1][1] = " << Vp_cpp[1][1] << std::endl;? // It doesn't print to


What could be wrong or am I missing something?

Thanks in advance.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/cplusplus-sig/attachments/20120815/b4ca7a4c/attachment.html>

From alfarobi0161 at yahoo.com  Wed Aug 15 15:11:49 2012
From: alfarobi0161 at yahoo.com (alfa)
Date: Wed, 15 Aug 2012 21:11:49 +0800 (SGT)
Subject: [C++-sig] Exposing 2D python array in C++
Message-ID: <1345036309.93831.YahooMailNeo@web190605.mail.sg3.yahoo.com>

Hi,


Is the following way valid to expose a 2D array variable from python (Vp)? to c++(Vp_cpp)?
?? 

std::vector<std::vector<double> > Vp_cpp;
boost::python::class_<std::vector<std::vector<double> > 
>("PyVec").def(boost::python::vector_indexing_suite<std::vector<std::vector<double> > >());


main_namespace["Vp"] = boost::python::ptr(&Vp_cpp);

But, then if I tried to manipulate the 2D vector in C++ like the following, it doesn't work:

Vp_cpp[1][1] = 999;
std::cout << "Vp_cpp[1][1] = " << Vp_cpp[1][1] << std::endl;? // It doesn't print to


What could be wrong or am I missing something?
Thanks in advance.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/cplusplus-sig/attachments/20120815/de14e8b9/attachment.html>

From talljimbo at gmail.com  Wed Aug 15 17:34:54 2012
From: talljimbo at gmail.com (Jim Bosch)
Date: Wed, 15 Aug 2012 08:34:54 -0700
Subject: [C++-sig] Accessing 2D array from python in C++
In-Reply-To: <1345021027.75677.YahooMailNeo@web190606.mail.sg3.yahoo.com>
References: <1344943525.62630.YahooMailNeo@web190601.mail.sg3.yahoo.com>
	<CAN+0q26pWg94qrWh91YP=1XOG3xfLh+WAkL+pR6+x_f7nVpF-w@mail.gmail.com>
	<1345021027.75677.YahooMailNeo@web190606.mail.sg3.yahoo.com>
Message-ID: <CAN+0q25-4vt2ooLyCpe=nKQ8bSUqfy2aY03v=nfgEuoSYEPxLQ@mail.gmail.com>

On Aug 15, 2012 1:57 AM, "alfa" <alfarobi0161 at yahoo.com> wrote:
>
> I apologize for double posting this because it was spammed before.
>
> Hi Jim,
>
> Do you mean, "you could wrap multi_array itself using class_", I should
do as below:
>
> boost::python::class_<std::vector<std::vector<double> >
>("PyVec").def(boost::python::vector_indexing_suite<std::vector<std::vector<double>
> >());
>

No, I don't think that will work; I meant something like this:

boost::python::class_< boost::multi_array<...> >(...)
    ...;

Where the last "..." is where you have to manually wrap the features of
multi_array you want to use.  I'm afraid there's nothing as easy as
vector_indexing_suite.

Jim
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/cplusplus-sig/attachments/20120815/76d5411f/attachment.html>

From alfarobi0161 at yahoo.com  Thu Aug 16 10:29:35 2012
From: alfarobi0161 at yahoo.com (alfa)
Date: Thu, 16 Aug 2012 16:29:35 +0800 (SGT)
Subject: [C++-sig] Accessing 2D array from python in C++
In-Reply-To: <CAN+0q25-4vt2ooLyCpe=nKQ8bSUqfy2aY03v=nfgEuoSYEPxLQ@mail.gmail.com>
References: <1344943525.62630.YahooMailNeo@web190601.mail.sg3.yahoo.com>
	<CAN+0q26pWg94qrWh91YP=1XOG3xfLh+WAkL+pR6+x_f7nVpF-w@mail.gmail.com>
	<1345021027.75677.YahooMailNeo@web190606.mail.sg3.yahoo.com>
	<CAN+0q25-4vt2ooLyCpe=nKQ8bSUqfy2aY03v=nfgEuoSYEPxLQ@mail.gmail.com>
Message-ID: <1345105775.68082.YahooMailNeo@web190606.mail.sg3.yahoo.com>

Hi,

Could you give more hints on this or any code snippet that I can implement this. I'm a bit naive in applying this. Please consider my 2D array of Vp_cpp(Vp equivalent in Python) is as below:


??? typedef boost::multi_array<double, 2> array_type;
??? typedef array_type::index index;
??? array_type Vp_cpp(boost::extents[X_cpp][X_cpp]);


I'm still puzzled how to "manually wrap the features" in the last "..." in:
??? boost::python::class_< boost::multi_array<...> >(...)

Thanks in advance.



________________________________
 From: Jim Bosch <talljimbo at gmail.com>
To: alfa <alfarobi0161 at yahoo.com> 
Cc: Development of Python/C++ integration <cplusplus-sig at python.org> 
Sent: Wednesday, August 15, 2012 5:34 PM
Subject: Re: [C++-sig] Accessing 2D array from python in C++
 

On Aug 15, 2012 1:57 AM, "alfa" <alfarobi0161 at yahoo.com> wrote:
>
> I apologize for double posting this because it was spammed before.
>
> Hi Jim,
>
> Do you mean, "you could wrap multi_array itself using class_", I should do as below:
>
> boost::python::class_<std::vector<std::vector<double> > >("PyVec").def(boost::python::vector_indexing_suite<std::vector<std::vector<double> > >());
>
No, I don't think that will work; I meant something like this:
boost::python::class_< boost::multi_array<...> >(...)
??? ...;
Where the last "..." is where you have to manually wrap the features of multi_array you want to use.? I'm afraid there's nothing as easy as vector_indexing_suite.

Jim
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/cplusplus-sig/attachments/20120816/c8999758/attachment.html>

From j.reid at mail.cryst.bbk.ac.uk  Thu Aug 16 17:02:03 2012
From: j.reid at mail.cryst.bbk.ac.uk (John Reid)
Date: Thu, 16 Aug 2012 16:02:03 +0100
Subject: [C++-sig] Accessing 2D array from python in C++
In-Reply-To: <1345105775.68082.YahooMailNeo@web190606.mail.sg3.yahoo.com>
References: <1344943525.62630.YahooMailNeo@web190601.mail.sg3.yahoo.com>
	<CAN+0q26pWg94qrWh91YP=1XOG3xfLh+WAkL+pR6+x_f7nVpF-w@mail.gmail.com>
	<1345021027.75677.YahooMailNeo@web190606.mail.sg3.yahoo.com>
	<CAN+0q25-4vt2ooLyCpe=nKQ8bSUqfy2aY03v=nfgEuoSYEPxLQ@mail.gmail.com>
	<1345105775.68082.YahooMailNeo@web190606.mail.sg3.yahoo.com>
Message-ID: <k0j21c$d1p$1@ger.gmane.org>

If you don't mind digging around in my code, you could download the
source code for the STEME python package from pypi and look in the
c++/myrrh directory. There is code to access multi_arrays from python in
there.

http://pypi.python.org/pypi/STEME/

In particular the file
c++/myrrh/myrrh/python/multi_array_to_numpy.h contains the bulk of the
implementation.

HTH,
John.



On 16/08/12 09:29, alfa wrote:
> Hi,
> 
> Could you give more hints on this or any code snippet that I can
> implement this. I'm a bit naive in applying this. Please consider my 2D
> array of Vp_cpp(Vp equivalent in Python) is as below:
> 
>     typedef boost::multi_array<double, 2> array_type;
>     typedef array_type::index index;
>     array_type Vp_cpp(boost::extents[X_cpp][X_cpp]);
> 
> 
> I'm still puzzled how to "manually wrap the features" in the last "..." in:
>     boost::python::class_< boost::multi_array<...> >(...)
> 
> Thanks in advance.
> 
> ------------------------------------------------------------------------
> *From:* Jim Bosch <talljimbo at gmail.com>
> *To:* alfa <alfarobi0161 at yahoo.com>
> *Cc:* Development of Python/C++ integration <cplusplus-sig at python.org>
> *Sent:* Wednesday, August 15, 2012 5:34 PM
> *Subject:* Re: [C++-sig] Accessing 2D array from python in C++
> 
> On Aug 15, 2012 1:57 AM, "alfa" <alfarobi0161 at yahoo.com
> <mailto:alfarobi0161 at yahoo.com>> wrote:
>>
>> I apologize for double posting this because it was spammed before.
>>
>> Hi Jim,
>>
>> Do you mean, "you could wrap multi_array itself using class_", I
> should do as below:
>>
>> boost::python::class_<std::vector<std::vector<double> >
>>("PyVec").def(boost::python::vector_indexing_suite<std::vector<std::vector<double>
>> >());
>>
> No, I don't think that will work; I meant something like this:
> boost::python::class_< boost::multi_array<...> >(...)
>     ...;
> Where the last "..." is where you have to manually wrap the features of
> multi_array you want to use.  I'm afraid there's nothing as easy as
> vector_indexing_suite.
> Jim
> 
> 
> 
> 
> _______________________________________________
> Cplusplus-sig mailing list
> Cplusplus-sig at python.org
> http://mail.python.org/mailman/listinfo/cplusplus-sig
> 


From trigves at yahoo.com  Fri Aug 17 14:58:34 2012
From: trigves at yahoo.com (Trigve Siver)
Date: Fri, 17 Aug 2012 05:58:34 -0700 (PDT)
Subject: [C++-sig] Member function of other class
Message-ID: <1345208314.56489.YahooMailNeo@web110413.mail.gq1.yahoo.com>

Hi,
I'm curios, why this piece of code does compile?


struct?BaseClass
{ void?func()?{}
}; struct?Wrap
{ void?func()?{}
}; BOOST_PYTHON_MODULE(main)
{ class_<BaseClass>("Base",?no_init) .def("func",?&Wrap::func) ;
}

I'm asking because a want to do something like this and want to know if it is UB or not:

struct?BaseClass
{ void?func()?{}
}; struct?Wrap
{ static?void?func(BaseClass?&Instance) { Instance.func(); }
}; BOOST_PYTHON_MODULE(main)
{ class_<BaseClass>("Base",?init<>()) .def("func",?&Wrap::func) ;
}

then in python:

import?main main.Base().func()

Thank you

Trigve


From trigves at yahoo.com  Fri Aug 17 15:07:03 2012
From: trigves at yahoo.com (Trigve Siver)
Date: Fri, 17 Aug 2012 06:07:03 -0700 (PDT)
Subject: [C++-sig] Member function of other class
In-Reply-To: <1345208314.56489.YahooMailNeo@web110413.mail.gq1.yahoo.com>
References: <1345208314.56489.YahooMailNeo@web110413.mail.gq1.yahoo.com>
Message-ID: <1345208823.63478.YahooMailNeo@web110402.mail.gq1.yahoo.com>


Code formatting went wrong


> ...
> struct?BaseClass
> { void?func()?{}
> }; struct?Wrap
> { void?func()?{}
> }; BOOST_PYTHON_MODULE(main)
> { class_<BaseClass>("Base",?no_init) 
> .def("func",?&Wrap::func) ;
> }

struct?BaseClass
{
? void?func()?{}
};
struct?Wrap
{
? void?func()?{}
};
BOOST_PYTHON_MODULE(main)
{
? class_<BaseClass>("Base",?no_init) 
? .def("func",?&Wrap::func) ;
}
?
> ...
> struct?BaseClass
> { void?func()?{}
> }; struct?Wrap
> { static?void?func(BaseClass?&Instance) { Instance.func(); }
> }; BOOST_PYTHON_MODULE(main)
> { class_<BaseClass>("Base",?init<>()) 
> .def("func",?&Wrap::func) ;
> }

struct?BaseClass
{
? void?func()?{}
};
struct?Wrap
{
? static?void?func(BaseClass?&Instance)
? {
??? Instance.func();
? }
};
BOOST_PYTHON_MODULE(main)
{
? class_<BaseClass>("Base",?init<>()) 
?.def("func",?&Wrap::func) ;
}

> ..
> import?main main.Base().func()
> 

import?main
main.Base().func()

Trigve


From alfarobi0161 at yahoo.com  Wed Aug 22 12:13:34 2012
From: alfarobi0161 at yahoo.com (alfa)
Date: Wed, 22 Aug 2012 18:13:34 +0800 (SGT)
Subject: [C++-sig] Accessing 2D array from python in C++
In-Reply-To: <k0j21c$d1p$1@ger.gmane.org>
References: <1344943525.62630.YahooMailNeo@web190601.mail.sg3.yahoo.com>
	<CAN+0q26pWg94qrWh91YP=1XOG3xfLh+WAkL+pR6+x_f7nVpF-w@mail.gmail.com>
	<1345021027.75677.YahooMailNeo@web190606.mail.sg3.yahoo.com>
	<CAN+0q25-4vt2ooLyCpe=nKQ8bSUqfy2aY03v=nfgEuoSYEPxLQ@mail.gmail.com>
	<1345105775.68082.YahooMailNeo@web190606.mail.sg3.yahoo.com>
	<k0j21c$d1p$1@ger.gmane.org>
Message-ID: <1345630414.39266.YahooMailNeo@web190602.mail.sg3.yahoo.com>

Hi John,

Thanks a lot for your implementation. In multi_array_to_numpy.h,? can you give more hints on which method/function that I can possibly use if I can include it in my implementation as a header file. 

Thanks in advance.




________________________________
 From: John Reid <j.reid at mail.cryst.bbk.ac.uk>
To: cplusplus-sig at python.org 
Sent: Thursday, August 16, 2012 5:02 PM
Subject: Re: [C++-sig] Accessing 2D array from python in C++
 
If you don't mind digging around in my code, you could download the
source code for the STEME python package from pypi and look in the
c++/myrrh directory. There is code to access multi_arrays from python in
there.

http://pypi.python.org/pypi/STEME/

In particular the file
c++/myrrh/myrrh/python/multi_array_to_numpy.h contains the bulk of the
implementation.

HTH,
John.



On 16/08/12 09:29, alfa wrote:
> Hi,
> 
> Could you give more hints on this or any code snippet that I can
> implement this. I'm a bit naive in applying this. Please consider my 2D
> array of Vp_cpp(Vp equivalent in Python) is as below:
> 
>? ?  typedef boost::multi_array<double, 2> array_type;
>? ?  typedef array_type::index index;
>? ?  array_type Vp_cpp(boost::extents[X_cpp][X_cpp]);
> 
> 
> I'm still puzzled how to "manually wrap the features" in the last "..." in:
>? ?  boost::python::class_< boost::multi_array<...> >(...)
> 
> Thanks in advance.
> 
> ------------------------------------------------------------------------
> *From:* Jim Bosch <talljimbo at gmail.com>
> *To:* alfa <alfarobi0161 at yahoo.com>
> *Cc:* Development of Python/C++ integration <cplusplus-sig at python.org>
> *Sent:* Wednesday, August 15, 2012 5:34 PM
> *Subject:* Re: [C++-sig] Accessing 2D array from python in C++
> 
> On Aug 15, 2012 1:57 AM, "alfa" <alfarobi0161 at yahoo.com
> <mailto:alfarobi0161 at yahoo.com>> wrote:
>>
>> I apologize for double posting this because it was spammed before.
>>
>> Hi Jim,
>>
>> Do you mean, "you could wrap multi_array itself using class_", I
> should do as below:
>>
>> boost::python::class_<std::vector<std::vector<double> >
>>("PyVec").def(boost::python::vector_indexing_suite<std::vector<std::vector<double>
>> >());
>>
> No, I don't think that will work; I meant something like this:
> boost::python::class_< boost::multi_array<...> >(...)
>? ?  ...;
> Where the last "..." is where you have to manually wrap the features of
> multi_array you want to use.? I'm afraid there's nothing as easy as
> vector_indexing_suite.
> Jim
> 
> 
> 
> 
> _______________________________________________
> Cplusplus-sig mailing list
> Cplusplus-sig at python.org
> http://mail.python.org/mailman/listinfo/cplusplus-sig
> 

_______________________________________________
Cplusplus-sig mailing list
Cplusplus-sig at python.org
http://mail.python.org/mailman/listinfo/cplusplus-sig
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/cplusplus-sig/attachments/20120822/c679fa61/attachment.html>

From j.reid at mail.cryst.bbk.ac.uk  Wed Aug 22 12:28:53 2012
From: j.reid at mail.cryst.bbk.ac.uk (John Reid)
Date: Wed, 22 Aug 2012 11:28:53 +0100
Subject: [C++-sig] Accessing 2D array from python in C++
In-Reply-To: <1345630414.39266.YahooMailNeo@web190602.mail.sg3.yahoo.com>
References: <1344943525.62630.YahooMailNeo@web190601.mail.sg3.yahoo.com>
	<CAN+0q26pWg94qrWh91YP=1XOG3xfLh+WAkL+pR6+x_f7nVpF-w@mail.gmail.com>
	<1345021027.75677.YahooMailNeo@web190606.mail.sg3.yahoo.com>
	<CAN+0q25-4vt2ooLyCpe=nKQ8bSUqfy2aY03v=nfgEuoSYEPxLQ@mail.gmail.com>
	<1345105775.68082.YahooMailNeo@web190606.mail.sg3.yahoo.com>
	<k0j21c$d1p$1@ger.gmane.org>
	<1345630414.39266.YahooMailNeo@web190602.mail.sg3.yahoo.com>
Message-ID: <k12c96$b12$1@ger.gmane.org>

On 22/08/12 11:13, alfa wrote:
> Hi John,
> 
> Thanks a lot for your implementation. In multi_array_to_numpy.h,  can
> you give more hints on which method/function that I can possibly use if
> I can include it in my implementation as a header file.
> 
> Thanks in advance.
> 
src/python/man_module.cpp and src/python/multi_array_to_numpy.cpp
implement a boost.python module that exposes the multi-array classes.
You have choose which data types you expose multiarrays for.

HTH,
John.


> 
> ------------------------------------------------------------------------
> *From:* John Reid <j.reid at mail.cryst.bbk.ac.uk>
> *To:* cplusplus-sig at python.org
> *Sent:* Thursday, August 16, 2012 5:02 PM
> *Subject:* Re: [C++-sig] Accessing 2D array from python in C++
> 
> If you don't mind digging around in my code, you could download the
> source code for the STEME python package from pypi and look in the
> c++/myrrh directory. There is code to access multi_arrays from python in
> there.
> 
> http://pypi.python.org/pypi/STEME/
> 
> In particular the file
> c++/myrrh/myrrh/python/multi_array_to_numpy.h contains the bulk of the
> implementation.
> 
> HTH,
> John.
> 
> 
> 
> On 16/08/12 09:29, alfa wrote:
>> Hi,
>>
>> Could you give more hints on this or any code snippet that I can
>> implement this. I'm a bit naive in applying this. Please consider my 2D
>> array of Vp_cpp(Vp equivalent in Python) is as below:
>>
>>    typedef boost::multi_array<double, 2> array_type;
>>    typedef array_type::index index;
>>    array_type Vp_cpp(boost::extents[X_cpp][X_cpp]);
>>
>>
>> I'm still puzzled how to "manually wrap the features" in the last
> "..." in:
>>    boost::python::class_< boost::multi_array<...> >(...)
>>
>> Thanks in advance.
>>
>> ------------------------------------------------------------------------
>> *From:* Jim Bosch <talljimbo at gmail.com <mailto:talljimbo at gmail.com>>
>> *To:* alfa <alfarobi0161 at yahoo.com <mailto:alfarobi0161 at yahoo.com>>
>> *Cc:* Development of Python/C++ integration <cplusplus-sig at python.org
> <mailto:cplusplus-sig at python.org>>
>> *Sent:* Wednesday, August 15, 2012 5:34 PM
>> *Subject:* Re: [C++-sig] Accessing 2D array from python in C++
>>
>> On Aug 15, 2012 1:57 AM, "alfa" <alfarobi0161 at yahoo.com
> <mailto:alfarobi0161 at yahoo.com>
>> <mailto:alfarobi0161 at yahoo.com <mailto:alfarobi0161 at yahoo.com>>> wrote:
>>>
>>> I apologize for double posting this because it was spammed before.
>>>
>>> Hi Jim,
>>>
>>> Do you mean, "you could wrap multi_array itself using class_", I
>> should do as below:
>>>
>>> boost::python::class_<std::vector<std::vector<double> >
>>>("PyVec").def(boost::python::vector_indexing_suite<std::vector<std::vector<double>
>>> >());
>>>
>> No, I don't think that will work; I meant something like this:
>> boost::python::class_< boost::multi_array<...> >(...)
>>    ...;
>> Where the last "..." is where you have to manually wrap the features of
>> multi_array you want to use.  I'm afraid there's nothing as easy as
>> vector_indexing_suite.
>> Jim
>>
>>
>>
>>
>> _______________________________________________
>> Cplusplus-sig mailing list
>> Cplusplus-sig at python.org <mailto:Cplusplus-sig at python.org>
>> http://mail.python.org/mailman/listinfo/cplusplus-sig
>>
> 
> _______________________________________________
> Cplusplus-sig mailing list
> Cplusplus-sig at python.org <mailto:Cplusplus-sig at python.org>
> http://mail.python.org/mailman/listinfo/cplusplus-sig
> 
> 
> 
> 
> _______________________________________________
> Cplusplus-sig mailing list
> Cplusplus-sig at python.org
> http://mail.python.org/mailman/listinfo/cplusplus-sig
> 


From alfarobi0161 at yahoo.com  Wed Aug 22 12:45:56 2012
From: alfarobi0161 at yahoo.com (alfa)
Date: Wed, 22 Aug 2012 18:45:56 +0800 (SGT)
Subject: [C++-sig] Accessing 2D array from python in C++
In-Reply-To: <k12c96$b12$1@ger.gmane.org>
References: <1344943525.62630.YahooMailNeo@web190601.mail.sg3.yahoo.com>
	<CAN+0q26pWg94qrWh91YP=1XOG3xfLh+WAkL+pR6+x_f7nVpF-w@mail.gmail.com>
	<1345021027.75677.YahooMailNeo@web190606.mail.sg3.yahoo.com>
	<CAN+0q25-4vt2ooLyCpe=nKQ8bSUqfy2aY03v=nfgEuoSYEPxLQ@mail.gmail.com>
	<1345105775.68082.YahooMailNeo@web190606.mail.sg3.yahoo.com>
	<k0j21c$d1p$1@ger.gmane.org>
	<1345630414.39266.YahooMailNeo@web190602.mail.sg3.yahoo.com>
	<k12c96$b12$1@ger.gmane.org>
Message-ID: <1345632356.46129.YahooMailNeo@web190605.mail.sg3.yahoo.com>

Hi John,

I'm sorry, I'm a bit confused because previously you said that the relevant implementation resides in "c++/myrrh/myrrh/python/multi_array_to_numpy.h"but now, you point to "src/python/man_module.cpp and src/python/multi_array_to_numpy.cpp". Hope you understand my problem of having 2D array in python that needs to be exposed/converted to C++. It would be great If I know any method/function that I can invoke from any of your implementations.

Thanks in advance.




________________________________
 From: John Reid <j.reid at mail.cryst.bbk.ac.uk>
To: cplusplus-sig at python.org 
Sent: Wednesday, August 22, 2012 12:28 PM
Subject: Re: [C++-sig] Accessing 2D array from python in C++
 
On 22/08/12 11:13, alfa wrote:
> Hi John,
> 
> Thanks a lot for your implementation. In multi_array_to_numpy.h,? can
> you give more hints on which method/function that I can possibly use if
> I can include it in my implementation as a header file.
> 
> Thanks in advance.
> 
src/python/man_module.cpp and src/python/multi_array_to_numpy.cpp
implement a boost.python module that exposes the multi-array classes.
You have choose which data types you expose multiarrays for.

HTH,
John.


> 
> ------------------------------------------------------------------------
> *From:* John Reid <j.reid at mail.cryst.bbk.ac.uk>
> *To:* cplusplus-sig at python.org
> *Sent:* Thursday, August 16, 2012 5:02 PM
> *Subject:* Re: [C++-sig] Accessing 2D array from python in C++
> 
> If you don't mind digging around in my code, you could download the
> source code for the STEME python package from pypi and look in the
> c++/myrrh directory. There is code to access multi_arrays from python in
> there.
> 
> http://pypi.python.org/pypi/STEME/
> 
> In particular the file
> c++/myrrh/myrrh/python/multi_array_to_numpy.h contains the bulk of the
> implementation.
> 
> HTH,
> John.
> 
> 
> 
> On 16/08/12 09:29, alfa wrote:
>> Hi,
>>
>> Could you give more hints on this or any code snippet that I can
>> implement this. I'm a bit naive in applying this. Please consider my 2D
>> array of Vp_cpp(Vp equivalent in Python) is as below:
>>
>>? ? typedef boost::multi_array<double, 2> array_type;
>>? ? typedef array_type::index index;
>>? ? array_type Vp_cpp(boost::extents[X_cpp][X_cpp]);
>>
>>
>> I'm still puzzled how to "manually wrap the features" in the last
> "..." in:
>>? ? boost::python::class_< boost::multi_array<...> >(...)
>>
>> Thanks in advance.
>>
>> ------------------------------------------------------------------------
>> *From:* Jim Bosch <talljimbo at gmail.com <mailto:talljimbo at gmail.com>>
>> *To:* alfa <alfarobi0161 at yahoo.com <mailto:alfarobi0161 at yahoo.com>>
>> *Cc:* Development of Python/C++ integration <cplusplus-sig at python.org
> <mailto:cplusplus-sig at python.org>>
>> *Sent:* Wednesday, August 15, 2012 5:34 PM
>> *Subject:* Re: [C++-sig] Accessing 2D array from python in C++
>>
>> On Aug 15, 2012 1:57 AM, "alfa" <alfarobi0161 at yahoo.com
> <mailto:alfarobi0161 at yahoo.com>
>> <mailto:alfarobi0161 at yahoo.com <mailto:alfarobi0161 at yahoo.com>>> wrote:
>>>
>>> I apologize for double posting this because it was spammed before.
>>>
>>> Hi Jim,
>>>
>>> Do you mean, "you could wrap multi_array itself using class_", I
>> should do as below:
>>>
>>> boost::python::class_<std::vector<std::vector<double> >
>>>("PyVec").def(boost::python::vector_indexing_suite<std::vector<std::vector<double>
>>> >());
>>>
>> No, I don't think that will work; I meant something like this:
>> boost::python::class_< boost::multi_array<...> >(...)
>>? ? ...;
>> Where the last "..." is where you have to manually wrap the features of
>> multi_array you want to use.? I'm afraid there's nothing as easy as
>> vector_indexing_suite.
>> Jim
>>
>>
>>
>>
>> _______________________________________________
>> Cplusplus-sig mailing list
>> Cplusplus-sig at python.org <mailto:Cplusplus-sig at python.org>
>> http://mail.python.org/mailman/listinfo/cplusplus-sig
>>
> 
> _______________________________________________
> Cplusplus-sig mailing list
> Cplusplus-sig at python.org <mailto:Cplusplus-sig at python.org>
> http://mail.python.org/mailman/listinfo/cplusplus-sig
> 
> 
> 
> 
> _______________________________________________
> Cplusplus-sig mailing list
> Cplusplus-sig at python.org
> http://mail.python.org/mailman/listinfo/cplusplus-sig
> 

_______________________________________________
Cplusplus-sig mailing list
Cplusplus-sig at python.org
http://mail.python.org/mailman/listinfo/cplusplus-sig
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/cplusplus-sig/attachments/20120822/ba618b3f/attachment-0001.html>

From j.reid at mail.cryst.bbk.ac.uk  Wed Aug 22 12:58:27 2012
From: j.reid at mail.cryst.bbk.ac.uk (John Reid)
Date: Wed, 22 Aug 2012 11:58:27 +0100
Subject: [C++-sig] Accessing 2D array from python in C++
In-Reply-To: <1345632356.46129.YahooMailNeo@web190605.mail.sg3.yahoo.com>
References: <1344943525.62630.YahooMailNeo@web190601.mail.sg3.yahoo.com>
	<CAN+0q26pWg94qrWh91YP=1XOG3xfLh+WAkL+pR6+x_f7nVpF-w@mail.gmail.com>
	<1345021027.75677.YahooMailNeo@web190606.mail.sg3.yahoo.com>
	<CAN+0q25-4vt2ooLyCpe=nKQ8bSUqfy2aY03v=nfgEuoSYEPxLQ@mail.gmail.com>
	<1345105775.68082.YahooMailNeo@web190606.mail.sg3.yahoo.com>
	<k0j21c$d1p$1@ger.gmane.org>
	<1345630414.39266.YahooMailNeo@web190602.mail.sg3.yahoo.com>
	<k12c96$b12$1@ger.gmane.org>
	<1345632356.46129.YahooMailNeo@web190605.mail.sg3.yahoo.com>
Message-ID: <k12e0k$poi$1@ger.gmane.org>



On 22/08/12 11:45, alfa wrote:
> Hi John,
> 
> I'm sorry, I'm a bit confused because previously you said that the
> relevant implementation resides in
> "c++/myrrh/myrrh/python/multi_array_to_numpy.h"but now, you point to
> "src/python/man_module.cpp and src/python/multi_array_to_numpy.cpp".
> Hope you understand my problem of having 2D array in python that needs
> to be exposed/converted to C++. It would be great If I know any
> method/function that I can invoke from any of your implementations.
> 
> Thanks in advance.

Hi Alfa,

You can look at all of the code I've mentioned to work out how it works.
If it does not make sense perhaps it might be worth working through the
boost.python tutorial to get up to speed on how boost.python works.

You can call the expose_converters< npy_double >() function to expose
multiarrays of doubles to python as numpy arrays.

Please ask if you have any specific questions.

Regards,
John.


> 
> 
> ------------------------------------------------------------------------
> *From:* John Reid <j.reid at mail.cryst.bbk.ac.uk>
> *To:* cplusplus-sig at python.org
> *Sent:* Wednesday, August 22, 2012 12:28 PM
> *Subject:* Re: [C++-sig] Accessing 2D array from python in C++
> 
> On 22/08/12 11:13, alfa wrote:
>> Hi John,
>>
>> Thanks a lot for your implementation. In multi_array_to_numpy.h,  can
>> you give more hints on which method/function that I can possibly use if
>> I can include it in my implementation as a header file.
>>
>> Thanks in advance.
>>
> src/python/man_module.cpp and src/python/multi_array_to_numpy.cpp
> implement a boost.python module that exposes the multi-array classes.
> You have choose which data types you expose multiarrays for.
> 
> HTH,
> John.
> 
> 
>>
>> ------------------------------------------------------------------------
>> *From:* John Reid <j.reid at mail.cryst.bbk.ac.uk
> <mailto:j.reid at mail.cryst.bbk.ac.uk>>
>> *To:* cplusplus-sig at python.org <mailto:cplusplus-sig at python.org>
>> *Sent:* Thursday, August 16, 2012 5:02 PM
>> *Subject:* Re: [C++-sig] Accessing 2D array from python in C++
>>
>> If you don't mind digging around in my code, you could download the
>> source code for the STEME python package from pypi and look in the
>> c++/myrrh directory. There is code to access multi_arrays from python in
>> there.
>>
>> http://pypi.python.org/pypi/STEME/
>>
>> In particular the file
>> c++/myrrh/myrrh/python/multi_array_to_numpy.h contains the bulk of the
>> implementation.
>>
>> HTH,
>> John.
>>
>>
>>
>> On 16/08/12 09:29, alfa wrote:
>>> Hi,
>>>
>>> Could you give more hints on this or any code snippet that I can
>>> implement this. I'm a bit naive in applying this. Please consider my 2D
>>> array of Vp_cpp(Vp equivalent in Python) is as below:
>>>
>>>    typedef boost::multi_array<double, 2> array_type;
>>>    typedef array_type::index index;
>>>    array_type Vp_cpp(boost::extents[X_cpp][X_cpp]);
>>>
>>>
>>> I'm still puzzled how to "manually wrap the features" in the last
>> "..." in:
>>>    boost::python::class_< boost::multi_array<...> >(...)
>>>
>>> Thanks in advance.
>>>
>>> ------------------------------------------------------------------------
>>> *From:* Jim Bosch <talljimbo at gmail.com <mailto:talljimbo at gmail.com>
> <mailto:talljimbo at gmail.com <mailto:talljimbo at gmail.com>>>
>>> *To:* alfa <alfarobi0161 at yahoo.com <mailto:alfarobi0161 at yahoo.com>
> <mailto:alfarobi0161 at yahoo.com <mailto:alfarobi0161 at yahoo.com>>>
>>> *Cc:* Development of Python/C++ integration <cplusplus-sig at python.org
> <mailto:cplusplus-sig at python.org>
>> <mailto:cplusplus-sig at python.org <mailto:cplusplus-sig at python.org>>>
>>> *Sent:* Wednesday, August 15, 2012 5:34 PM
>>> *Subject:* Re: [C++-sig] Accessing 2D array from python in C++
>>>
>>> On Aug 15, 2012 1:57 AM, "alfa" <alfarobi0161 at yahoo.com
> <mailto:alfarobi0161 at yahoo.com>
>> <mailto:alfarobi0161 at yahoo.com <mailto:alfarobi0161 at yahoo.com>>
>>> <mailto:alfarobi0161 at yahoo.com <mailto:alfarobi0161 at yahoo.com>
> <mailto:alfarobi0161 at yahoo.com <mailto:alfarobi0161 at yahoo.com>>>> wrote:
>>>>
>>>> I apologize for double posting this because it was spammed before.
>>>>
>>>> Hi Jim,
>>>>
>>>> Do you mean, "you could wrap multi_array itself using class_", I
>>> should do as below:
>>>>
>>>> boost::python::class_<std::vector<std::vector<double> >
>>>>("PyVec").def(boost::python::vector_indexing_suite<std::vector<std::vector<double>
>>>> >());
>>>>
>>> No, I don't think that will work; I meant something like this:
>>> boost::python::class_< boost::multi_array<...> >(...)
>>>    ...;
>>> Where the last "..." is where you have to manually wrap the features of
>>> multi_array you want to use.  I'm afraid there's nothing as easy as
>>> vector_indexing_suite.
>>> Jim
>>>
>>>
>>>
>>>
>>> _______________________________________________
>>> Cplusplus-sig mailing list
>>> Cplusplus-sig at python.org <mailto:Cplusplus-sig at python.org>
> <mailto:Cplusplus-sig at python.org <mailto:Cplusplus-sig at python.org>>
>>> http://mail.python.org/mailman/listinfo/cplusplus-sig
>>>
>>
>> _______________________________________________
>> Cplusplus-sig mailing list
>> Cplusplus-sig at python.org <mailto:Cplusplus-sig at python.org>
> <mailto:Cplusplus-sig at python.org <mailto:Cplusplus-sig at python.org>>
>> http://mail.python.org/mailman/listinfo/cplusplus-sig
>>
>>
>>
>>
>> _______________________________________________
>> Cplusplus-sig mailing list
>> Cplusplus-sig at python.org <mailto:Cplusplus-sig at python.org>
>> http://mail.python.org/mailman/listinfo/cplusplus-sig
>>
> 
> _______________________________________________
> Cplusplus-sig mailing list
> Cplusplus-sig at python.org <mailto:Cplusplus-sig at python.org>
> http://mail.python.org/mailman/listinfo/cplusplus-sig
> 
> 
> 
> 
> _______________________________________________
> Cplusplus-sig mailing list
> Cplusplus-sig at python.org
> http://mail.python.org/mailman/listinfo/cplusplus-sig
> 


From j.reid at mail.cryst.bbk.ac.uk  Wed Aug 22 13:11:00 2012
From: j.reid at mail.cryst.bbk.ac.uk (John Reid)
Date: Wed, 22 Aug 2012 12:11:00 +0100
Subject: [C++-sig] Accessing 2D array from python in C++
In-Reply-To: <k12e0k$poi$1@ger.gmane.org>
References: <1344943525.62630.YahooMailNeo@web190601.mail.sg3.yahoo.com>
	<CAN+0q26pWg94qrWh91YP=1XOG3xfLh+WAkL+pR6+x_f7nVpF-w@mail.gmail.com>
	<1345021027.75677.YahooMailNeo@web190606.mail.sg3.yahoo.com>
	<CAN+0q25-4vt2ooLyCpe=nKQ8bSUqfy2aY03v=nfgEuoSYEPxLQ@mail.gmail.com>
	<1345105775.68082.YahooMailNeo@web190606.mail.sg3.yahoo.com>
	<k0j21c$d1p$1@ger.gmane.org>
	<1345630414.39266.YahooMailNeo@web190602.mail.sg3.yahoo.com>
	<k12c96$b12$1@ger.gmane.org>
	<1345632356.46129.YahooMailNeo@web190605.mail.sg3.yahoo.com>
	<k12e0k$poi$1@ger.gmane.org>
Message-ID: <k12eo5$v1l$1@ger.gmane.org>



On 22/08/12 11:58, John Reid wrote:
> 
> 
> On 22/08/12 11:45, alfa wrote:
>> Hi John,
>>
>> I'm sorry, I'm a bit confused because previously you said that the
>> relevant implementation resides in
>> "c++/myrrh/myrrh/python/multi_array_to_numpy.h"but now, you point to
>> "src/python/man_module.cpp and src/python/multi_array_to_numpy.cpp".
>> Hope you understand my problem of having 2D array in python that needs
>> to be exposed/converted to C++. It would be great If I know any
>> method/function that I can invoke from any of your implementations.
>>
>> Thanks in advance.
> 
> Hi Alfa,
> 
> You can look at all of the code I've mentioned to work out how it works.
> If it does not make sense perhaps it might be worth working through the
> boost.python tutorial to get up to speed on how boost.python works.
> 
> You can call the expose_converters< npy_double >() function to expose
> multiarrays of doubles to python as numpy arrays.

Also look in python_test/test_man.py for code that tests the
implementation from the python side if it helps.


From alfarobi0161 at yahoo.com  Wed Aug 22 13:20:00 2012
From: alfarobi0161 at yahoo.com (alfa)
Date: Wed, 22 Aug 2012 19:20:00 +0800 (SGT)
Subject: [C++-sig] Accessing 2D array from python in C++
In-Reply-To: <k12e0k$poi$1@ger.gmane.org>
References: <1344943525.62630.YahooMailNeo@web190601.mail.sg3.yahoo.com>
	<CAN+0q26pWg94qrWh91YP=1XOG3xfLh+WAkL+pR6+x_f7nVpF-w@mail.gmail.com>
	<1345021027.75677.YahooMailNeo@web190606.mail.sg3.yahoo.com>
	<CAN+0q25-4vt2ooLyCpe=nKQ8bSUqfy2aY03v=nfgEuoSYEPxLQ@mail.gmail.com>
	<1345105775.68082.YahooMailNeo@web190606.mail.sg3.yahoo.com>
	<k0j21c$d1p$1@ger.gmane.org>
	<1345630414.39266.YahooMailNeo@web190602.mail.sg3.yahoo.com>
	<k12c96$b12$1@ger.gmane.org>
	<1345632356.46129.YahooMailNeo@web190605.mail.sg3.yahoo.com>
	<k12e0k$poi$1@ger.gmane.org>
Message-ID: <1345634400.17336.YahooMailNeo@web190604.mail.sg3.yahoo.com>

Hi John,

""



________________________________
 From: John Reid <j.reid at mail.cryst.bbk.ac.uk>
To: cplusplus-sig at python.org 
Sent: Wednesday, August 22, 2012 12:58 PM
Subject: Re: [C++-sig] Accessing 2D array from python in C++
 Hi John,

"You can call the expose_converters< npy_double >() function to expose
multiarrays of doubles to python as numpy arrays."

I'm afraid you understood that I have C++ array to be exposed as? numpy arrays in Python. My problem is actually the other way around, in which I have a 2D python array/matrix that I want to convert to/access in C++.

Thanks in advance.



On 22/08/12 11:45, alfa wrote:
> Hi John,
> 
> I'm sorry, I'm a bit confused because previously you said that the
> relevant implementation resides in
> "c++/myrrh/myrrh/python/multi_array_to_numpy.h"but now, you point to
> "src/python/man_module.cpp and src/python/multi_array_to_numpy.cpp".
> Hope you understand my problem of having 2D array in python that needs
> to be exposed/converted to C++. It would be great If I know any
> method/function that I can invoke from any of your implementations.
> 
> Thanks in advance.

Hi Alfa,

You can look at all of the code I've mentioned to work out how it works.
If it does not make sense perhaps it might be worth working through the
boost.python tutorial to get up to speed on how boost.python works.

You can call the expose_converters< npy_double >() function to expose
multiarrays of doubles to python as numpy arrays.

Please ask if you have any specific questions.

Regards,
John.


> 
> 
> ------------------------------------------------------------------------
> *From:* John Reid <j.reid at mail.cryst.bbk.ac.uk>
> *To:* cplusplus-sig at python.org
> *Sent:* Wednesday, August 22, 2012 12:28 PM
> *Subject:* Re: [C++-sig] Accessing 2D array from python in C++
> 
> On 22/08/12 11:13, alfa wrote:
>> Hi John,
>>
>> Thanks a lot for your implementation. In multi_array_to_numpy.h,? can
>> you give more hints on which method/function that I can possibly use if
>> I can include it in my implementation as a header file.
>>
>> Thanks in advance.
>>
> src/python/man_module.cpp and src/python/multi_array_to_numpy.cpp
> implement a boost.python module that exposes the multi-array classes.
> You have choose which data types you expose multiarrays for.
> 
> HTH,
> John.
> 
> 
>>
>> ------------------------------------------------------------------------
>> *From:* John Reid <j.reid at mail.cryst.bbk.ac.uk
> <mailto:j.reid at mail.cryst.bbk.ac.uk>>
>> *To:* cplusplus-sig at python.org <mailto:cplusplus-sig at python.org>
>> *Sent:* Thursday, August 16, 2012 5:02 PM
>> *Subject:* Re: [C++-sig] Accessing 2D array from python in C++
>>
>> If you don't mind digging around in my code, you could download the
>> source code for the STEME python package from pypi and look in the
>> c++/myrrh directory. There is code to access multi_arrays from python in
>> there.
>>
>> http://pypi.python.org/pypi/STEME/
>>
>> In particular the file
>> c++/myrrh/myrrh/python/multi_array_to_numpy.h contains the bulk of the
>> implementation.
>>
>> HTH,
>> John.
>>
>>
>>
>> On 16/08/12 09:29, alfa wrote:
>>> Hi,
>>>
>>> Could you give more hints on this or any code snippet that I can
>>> implement this. I'm a bit naive in applying this. Please consider my 2D
>>> array of Vp_cpp(Vp equivalent in Python) is as below:
>>>
>>>? ? typedef boost::multi_array<double, 2> array_type;
>>>? ? typedef array_type::index index;
>>>? ? array_type Vp_cpp(boost::extents[X_cpp][X_cpp]);
>>>
>>>
>>> I'm still puzzled how to "manually wrap the features" in the last
>> "..." in:
>>>? ? boost::python::class_< boost::multi_array<...> >(...)
>>>
>>> Thanks in advance.
>>>
>>> ------------------------------------------------------------------------
>>> *From:* Jim Bosch <talljimbo at gmail.com <mailto:talljimbo at gmail.com>
> <mailto:talljimbo at gmail.com <mailto:talljimbo at gmail.com>>>
>>> *To:* alfa <alfarobi0161 at yahoo.com <mailto:alfarobi0161 at yahoo.com>
> <mailto:alfarobi0161 at yahoo.com <mailto:alfarobi0161 at yahoo.com>>>
>>> *Cc:* Development of Python/C++ integration <cplusplus-sig at python.org
> <mailto:cplusplus-sig at python.org>
>> <mailto:cplusplus-sig at python.org <mailto:cplusplus-sig at python.org>>>
>>> *Sent:* Wednesday, August 15, 2012 5:34 PM
>>> *Subject:* Re: [C++-sig] Accessing 2D array from python in C++
>>>
>>> On Aug 15, 2012 1:57 AM, "alfa" <alfarobi0161 at yahoo.com
> <mailto:alfarobi0161 at yahoo.com>
>> <mailto:alfarobi0161 at yahoo.com <mailto:alfarobi0161 at yahoo.com>>
>>> <mailto:alfarobi0161 at yahoo.com <mailto:alfarobi0161 at yahoo.com>
> <mailto:alfarobi0161 at yahoo.com <mailto:alfarobi0161 at yahoo.com>>>> wrote:
>>>>
>>>> I apologize for double posting this because it was spammed before.
>>>>
>>>> Hi Jim,
>>>>
>>>> Do you mean, "you could wrap multi_array itself using class_", I
>>> should do as below:
>>>>
>>>> boost::python::class_<std::vector<std::vector<double> >
>>>>("PyVec").def(boost::python::vector_indexing_suite<std::vector<std::vector<double>
>>>> >());
>>>>
>>> No, I don't think that will work; I meant something like this:
>>> boost::python::class_< boost::multi_array<...> >(...)
>>>? ? ...;
>>> Where the last "..." is where you have to manually wrap the features of
>>> multi_array you want to use.? I'm afraid there's nothing as easy as
>>> vector_indexing_suite.
>>> Jim
>>>
>>>
>>>
>>>
>>> _______________________________________________
>>> Cplusplus-sig mailing list
>>> Cplusplus-sig at python.org <mailto:Cplusplus-sig at python.org>
> <mailto:Cplusplus-sig at python.org <mailto:Cplusplus-sig at python.org>>
>>> http://mail.python.org/mailman/listinfo/cplusplus-sig
>>>
>>
>> _______________________________________________
>> Cplusplus-sig mailing list
>> Cplusplus-sig at python.org <mailto:Cplusplus-sig at python.org>
> <mailto:Cplusplus-sig at python.org <mailto:Cplusplus-sig at python.org>>
>> http://mail.python.org/mailman/listinfo/cplusplus-sig
>>
>>
>>
>>
>> _______________________________________________
>> Cplusplus-sig mailing list
>> Cplusplus-sig at python.org <mailto:Cplusplus-sig at python.org>
>> http://mail.python.org/mailman/listinfo/cplusplus-sig
>>
> 
> _______________________________________________
> Cplusplus-sig mailing list
> Cplusplus-sig at python.org <mailto:Cplusplus-sig at python.org>
> http://mail.python.org/mailman/listinfo/cplusplus-sig
> 
> 
> 
> 
> _______________________________________________
> Cplusplus-sig mailing list
> Cplusplus-sig at python.org
> http://mail.python.org/mailman/listinfo/cplusplus-sig
> 

_______________________________________________
Cplusplus-sig mailing list
Cplusplus-sig at python.org
http://mail.python.org/mailman/listinfo/cplusplus-sig
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/cplusplus-sig/attachments/20120822/0007f6a7/attachment-0001.html>

From alfarobi0161 at yahoo.com  Wed Aug 22 13:28:18 2012
From: alfarobi0161 at yahoo.com (alfa)
Date: Wed, 22 Aug 2012 19:28:18 +0800 (SGT)
Subject: [C++-sig] Accessing 2D array from python in C++
In-Reply-To: <k12eo5$v1l$1@ger.gmane.org>
References: <1344943525.62630.YahooMailNeo@web190601.mail.sg3.yahoo.com>
	<CAN+0q26pWg94qrWh91YP=1XOG3xfLh+WAkL+pR6+x_f7nVpF-w@mail.gmail.com>
	<1345021027.75677.YahooMailNeo@web190606.mail.sg3.yahoo.com>
	<CAN+0q25-4vt2ooLyCpe=nKQ8bSUqfy2aY03v=nfgEuoSYEPxLQ@mail.gmail.com>
	<1345105775.68082.YahooMailNeo@web190606.mail.sg3.yahoo.com>
	<k0j21c$d1p$1@ger.gmane.org>
	<1345630414.39266.YahooMailNeo@web190602.mail.sg3.yahoo.com>
	<k12c96$b12$1@ger.gmane.org>
	<1345632356.46129.YahooMailNeo@web190605.mail.sg3.yahoo.com>
	<k12e0k$poi$1@ger.gmane.org> <k12eo5$v1l$1@ger.gmane.org>
Message-ID: <1345634898.12641.YahooMailNeo@web190603.mail.sg3.yahoo.com>

Hi John,

"You can call the expose_converters< npy_double >() function to expose multiarrays of doubles to python as numpy arrays."

I'm afraid you understood that I have a C++ array that is to be exposed in Python. Actually, my problem is the other way around in which I have a Python array that needs to be converted to/ exposed in C++.

Thanks in advance.




________________________________
 From: John Reid <j.reid at mail.cryst.bbk.ac.uk>
To: cplusplus-sig at python.org 
Sent: Wednesday, August 22, 2012 1:11 PM
Subject: Re: [C++-sig] Accessing 2D array from python in C++
 


On 22/08/12 11:58, John Reid wrote:
> 
> 
> On 22/08/12 11:45, alfa wrote:
>> Hi John,
>>
>> I'm sorry, I'm a bit confused because previously you said that the
>> relevant implementation resides in
>> "c++/myrrh/myrrh/python/multi_array_to_numpy.h"but now, you point to
>> "src/python/man_module.cpp and src/python/multi_array_to_numpy.cpp".
>> Hope you understand my problem of having 2D array in python that needs
>> to be exposed/converted to C++. It would be great If I know any
>> method/function that I can invoke from any of your implementations.
>>
>> Thanks in advance.
> 
> Hi Alfa,
> 
> You can look at all of the code I've mentioned to work out how it works.
> If it does not make sense perhaps it might be worth working through the
> boost.python tutorial to get up to speed on how boost.python works.
> 
> You can call the expose_converters< npy_double >() function to expose
> multiarrays of doubles to python as numpy arrays.

Also look in python_test/test_man.py for code that tests the
implementation from the python side if it helps.

_______________________________________________
Cplusplus-sig mailing list
Cplusplus-sig at python.org
http://mail.python.org/mailman/listinfo/cplusplus-sig
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/cplusplus-sig/attachments/20120822/7dea11b7/attachment.html>

From j.reid at mail.cryst.bbk.ac.uk  Wed Aug 22 14:03:01 2012
From: j.reid at mail.cryst.bbk.ac.uk (John Reid)
Date: Wed, 22 Aug 2012 13:03:01 +0100
Subject: [C++-sig] Accessing 2D array from python in C++
In-Reply-To: <1345634898.12641.YahooMailNeo@web190603.mail.sg3.yahoo.com>
References: <1344943525.62630.YahooMailNeo@web190601.mail.sg3.yahoo.com>
	<CAN+0q26pWg94qrWh91YP=1XOG3xfLh+WAkL+pR6+x_f7nVpF-w@mail.gmail.com>
	<1345021027.75677.YahooMailNeo@web190606.mail.sg3.yahoo.com>
	<CAN+0q25-4vt2ooLyCpe=nKQ8bSUqfy2aY03v=nfgEuoSYEPxLQ@mail.gmail.com>
	<1345105775.68082.YahooMailNeo@web190606.mail.sg3.yahoo.com>
	<k0j21c$d1p$1@ger.gmane.org>
	<1345630414.39266.YahooMailNeo@web190602.mail.sg3.yahoo.com>
	<k12c96$b12$1@ger.gmane.org>
	<1345632356.46129.YahooMailNeo@web190605.mail.sg3.yahoo.com>
	<k12e0k$poi$1@ger.gmane.org> <k12eo5$v1l$1@ger.gmane.org>
	<1345634898.12641.YahooMailNeo@web190603.mail.sg3.yahoo.com>
Message-ID: <k12hpm$o15$1@ger.gmane.org>



On 22/08/12 12:28, alfa wrote:
> Hi John,
> 
> "You can call the expose_converters< npy_double >() function to expose
> multiarrays of doubles to python as numpy arrays."
> 
> I'm afraid you understood that I have a C++ array that is to be exposed
> in Python. Actually, my problem is the other way around in which I have
> a Python array that needs to be converted to/ exposed in C++.

Exposing works both ways.

> 
> Thanks in advance.
> 
> 
> ------------------------------------------------------------------------
> *From:* John Reid <j.reid at mail.cryst.bbk.ac.uk>
> *To:* cplusplus-sig at python.org
> *Sent:* Wednesday, August 22, 2012 1:11 PM
> *Subject:* Re: [C++-sig] Accessing 2D array from python in C++
> 
> 
> 
> On 22/08/12 11:58, John Reid wrote:
>>
>>
>> On 22/08/12 11:45, alfa wrote:
>>> Hi John,
>>>
>>> I'm sorry, I'm a bit confused because previously you said that the
>>> relevant implementation resides in
>>> "c++/myrrh/myrrh/python/multi_array_to_numpy.h"but now, you point to
>>> "src/python/man_module.cpp and src/python/multi_array_to_numpy.cpp".
>>> Hope you understand my problem of having 2D array in python that needs
>>> to be exposed/converted to C++. It would be great If I know any
>>> method/function that I can invoke from any of your implementations.
>>>
>>> Thanks in advance.
>>
>> Hi Alfa,
>>
>> You can look at all of the code I've mentioned to wor k out how it works.
>> If it does not make sense perhaps it might be worth working through the
>> boost.python tutorial to get up to speed on how boost.python works.
>>
>> You can call the expose_converters< npy_double >() function to expose
>> multiarrays of doubles to python as numpy arrays.
> 
> Also look in python_test/test_man.py for code that tests the
> implementation from the python side if it helps.
> 
> _______________________________________________
> Cplusplus-sig mailing list
> Cplusplus-sig at python.org <mailto:Cplusplus-sig at python.org>
> http://mail.python.org/mailman/listinfo/cplusplus-sig
> 
> 
> 
> 
> _______________________________________________
> Cplusplus-sig mailing list
> Cplusplus-sig at python.org
> http://mail.python.org/mailman/listinfo/cplusplus-sig
> 


From herzog at mpi-cbg.de  Thu Aug 23 16:53:55 2012
From: herzog at mpi-cbg.de (Ronny Herzog)
Date: Thu, 23 Aug 2012 16:53:55 +0200
Subject: [C++-sig] issues with loading the boost_python dll on different
	computers
Message-ID: <k15g65$5ka$1@ger.gmane.org>

Dear all,

I installed and used boost (1.50.0) to build a python extension on my 
Windows 7 / 32 Bit. Everything worked fine.

Now I tried the extension module on other Computers with Windows 7 and 
Windows XP and the same Python version and it gave me an error 
concerning the DLL:

** ImportError : DLL load failed: This application has failed to start
because the application configuration is incorrect. Reinstalling the
application may fix this problem. **

Would anyone have a hint on were this problem could come from. I am 
totally clueless.

Thanks a lot,
Ronny


From trigves at yahoo.com  Thu Aug 23 18:50:12 2012
From: trigves at yahoo.com (Trigve Siver)
Date: Thu, 23 Aug 2012 09:50:12 -0700 (PDT)
Subject: [C++-sig] issues with loading the boost_python dll on different
	computers
In-Reply-To: <k15g65$5ka$1@ger.gmane.org>
References: <k15g65$5ka$1@ger.gmane.org>
Message-ID: <1345740612.35396.YahooMailNeo@web110413.mail.gq1.yahoo.com>

> From: Ronny Herzog <herzog at mpi-cbg.de>
> To: cplusplus-sig at python.org
> Cc: 
> Sent: Thursday, August 23, 2012 4:53 PM
> Subject: [C++-sig] issues with loading the boost_python dll on different computers
> 
> Dear all,
> 
> I installed and used boost (1.50.0) to build a python extension on my Windows 7 
> / 32 Bit. Everything worked fine.
> 
> Now I tried the extension module on other Computers with Windows 7 and Windows 
> XP and the same Python version and it gave me an error concerning the DLL:
> 
> ** ImportError : DLL load failed: This application has failed to start
> because the application configuration is incorrect. Reinstalling the
> application may fix this problem. **
> 
> Would anyone have a hint on were this problem could come from. I am totally 
> clueless.

Hi, two possible reason from top of my mind:

* Is it possible that you are using debug version of your extension?
* If you are using VS, have you installed VC redist of *your's VS version* on target PC?

Also try depends.exe to find if you aren't missing any dependencies.
?
> Thanks a lot,
> Ronny

Trigve


From herzog at mpi-cbg.de  Fri Aug 24 14:27:35 2012
From: herzog at mpi-cbg.de (Ronny Herzog)
Date: Fri, 24 Aug 2012 14:27:35 +0200
Subject: [C++-sig] issues with loading the boost_python dll on different
 computers
In-Reply-To: <1345740612.35396.YahooMailNeo@web110413.mail.gq1.yahoo.com>
References: <k15g65$5ka$1@ger.gmane.org>
	<1345740612.35396.YahooMailNeo@web110413.mail.gq1.yahoo.com>
Message-ID: <k17rvo$6he$1@ger.gmane.org>

Thanks Trigve,

I used depends.exe and found out that it was looking for the debug 
version of the msvc runtime library. Compiling my extension in 'release' 
mode - everything worked. :-)

Ronny

Am 8/23/2012 6:50 PM, schrieb Trigve Siver:
>> From: Ronny Herzog <herzog at mpi-cbg.de>
>> To: cplusplus-sig at python.org
>> Cc:
>> Sent: Thursday, August 23, 2012 4:53 PM
>> Subject: [C++-sig] issues with loading the boost_python dll on different computers
>>
>> Dear all,
>>
>> I installed and used boost (1.50.0) to build a python extension on my Windows 7
>> / 32 Bit. Everything worked fine.
>>
>> Now I tried the extension module on other Computers with Windows 7 and Windows
>> XP and the same Python version and it gave me an error concerning the DLL:
>>
>> ** ImportError : DLL load failed: This application has failed to start
>> because the application configuration is incorrect. Reinstalling the
>> application may fix this problem. **
>>
>> Would anyone have a hint on were this problem could come from. I am totally
>> clueless.
>
> Hi, two possible reason from top of my mind:
>
> * Is it possible that you are using debug version of your extension?
> * If you are using VS, have you installed VC redist of *your's VS version* on target PC?
>
> Also try depends.exe to find if you aren't missing any dependencies.
>
>> Thanks a lot,
>> Ronny
>
> Trigve
>



From herzog at mpi-cbg.de  Fri Aug 24 14:31:19 2012
From: herzog at mpi-cbg.de (Ronny Herzog)
Date: Fri, 24 Aug 2012 14:31:19 +0200
Subject: [C++-sig] building 32 Bit and 64 Bit extensions on the same machine
Message-ID: <k17s6o$8rd$1@ger.gmane.org>

Dear all,

I want to build 32 bit extensions and 64 bit extensions on my Windows 7 
/ 32 Bit with boost. After two days of checking google I am a bit 
frustrated. I am new to Boost and it is a bit like a jungle.

Can anyone help on what I need to do to set up Boost in the right way 
and have a Jamroot which allows me to compile my Python extensions in 32 
Bit and 64 Bit at the same time (with MSVC 2008).

Thanks a lot,
Ronny


From wmamrak at gmail.com  Fri Aug 24 16:49:03 2012
From: wmamrak at gmail.com (Wojciech Mamrak)
Date: Fri, 24 Aug 2012 16:49:03 +0200
Subject: [C++-sig] building 32 Bit and 64 Bit extensions on the same
	machine
In-Reply-To: <k17s6o$8rd$1@ger.gmane.org>
References: <k17s6o$8rd$1@ger.gmane.org>
Message-ID: <CAMhLx01j7_NF=q=2YjZmtoRRs6S5UQCKKBfM-dbBMxB1yOz8vA@mail.gmail.com>

Hi,

add address-model=64 to Boost::Python bjam compilation command and
that's all. Be aware, that not all MSVS versions (e.g. Express
Edition) are equipped with 64 Bit compiler.

2012/8/24 Ronny Herzog <herzog at mpi-cbg.de>:
> Dear all,
>
> I want to build 32 bit extensions and 64 bit extensions on my Windows 7 / 32
> Bit with boost. After two days of checking google I am a bit frustrated. I
> am new to Boost and it is a bit like a jungle.
>
> Can anyone help on what I need to do to set up Boost in the right way and
> have a Jamroot which allows me to compile my Python extensions in 32 Bit and
> 64 Bit at the same time (with MSVC 2008).
>
> Thanks a lot,
> Ronny
>
> _______________________________________________
> Cplusplus-sig mailing list
> Cplusplus-sig at python.org
> http://mail.python.org/mailman/listinfo/cplusplus-sig

From hesam.ilati2 at gmail.com  Fri Aug 24 17:32:11 2012
From: hesam.ilati2 at gmail.com (Hesam)
Date: Fri, 24 Aug 2012 11:32:11 -0400
Subject: [C++-sig] Boost::Python and Functions with String Inputs
Message-ID: <CAO6XJGUqFCYYTLztqUd7xqfBQjkpBShWOO06KPRjEEO+aZp4Yw@mail.gmail.com>

*Dear Sirs,*

*Goal:* I'm trying to use python interactively in my c++ code using
Boost::Python library. My goal is to change variables of some class which I
have defined in c++ from the python interpreter.

The code is attached below. Actually this comes from Boost::Python
tutorials
http://www.boost.org/doc/libs/1_51_0/libs/python/doc/tutorial/doc/html/python/exposing.html

*Problem:* I can load the library in python interface (i.e. load hello) and
make an object out of it (obj = hello.World()). I even can access functions
( obj.greet() ) but when I want to access functions with string input
(obj.set("Hello") ) I get memory access violation ("Access violation
reading location 0xffffffffffffffff"). *The problem is just with string
inputs. i.e. if I change function set(string) to set(char*) it works fine. *

If instead of interpreter I use PyRun_SimpleString("import hello;
w=hello.World(); w.set('Hi there');\n"); I get error "Expression
_BLOCK_TYPE_IS_VALID(pHead->nBlockUse)". The point is when I make
breakpoint inside the function value of string is not what I want.

*Details*: Visual studio 2010 (MDd mode), Debug Mode, x64, Win7

struct World
{

    void set(string _msg) {}

    string greet() { return msg; }

    string msg;
};
typedef boost::shared_ptr<World> World_ptr;

BOOST_PYTHON_MODULE(hello)
{

    bp::class_<World, World_ptr>("World")

        .def("greet", &World::greet)

        .def("set", &World::set)

    ;
}
int main(int argc, char **argv)
{

    Py_Initialize();

    bp::object main =
bp::object(bp::handle<>(bp::borrowed(PyImport_AddModule("__main__"))));

    bp::object global(main.attr("__dict__"));

    inithello();

    // Bring up Python interpreter

    Py_Main(argc, argv);

    Py_Finalize();

    return 0;
}

Actually this comes from Boost::Python tutorials with some modification to
use Python interpreter
http://www.boost.org/doc/libs/1_51_0/libs/python/doc/tutorial/doc/html/python/exposing.html

Many thanks

[image: Inline image 1]
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/cplusplus-sig/attachments/20120824/432935c3/attachment-0001.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: image.png
Type: image/png
Size: 42055 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/cplusplus-sig/attachments/20120824/432935c3/attachment-0001.png>

From wmamrak at gmail.com  Sun Aug 26 13:56:27 2012
From: wmamrak at gmail.com (Wojciech Mamrak)
Date: Sun, 26 Aug 2012 13:56:27 +0200
Subject: [C++-sig] building 32 Bit and 64 Bit extensions on the same
	machine
Message-ID: 27156



From hesam.ilati2 at gmail.com  Sun Aug 26 13:56:28 2012
From: hesam.ilati2 at gmail.com (Hesam)
Date: Sun, 26 Aug 2012 13:56:28 +0200
Subject: [C++-sig]  Boost::Python and Functions with String Inputs
Message-ID: 27157



From brandsmeier at gmx.de  Sun Aug 26 22:04:07 2012
From: brandsmeier at gmx.de (Holger Brandsmeier)
Date: Sun, 26 Aug 2012 22:04:07 +0200
Subject: [C++-sig] `cout << tuple` is ambiguous
Message-ID: <CAFP7zVsCWz2jsjO-6c72WXmsgBpyQmkrx4fTi5i2FyogjCfcGw@mail.gmail.com>

Dear list,

I get the following error
 error: use of overloaded operator '<<' is ambiguous (with operand
types 'basic_ostream<char, std::char_traits<char> >' and
      'const boost::python::tuple')

The two candidates are an output operator for tuple, which I want to use

/usr/include/boost-1_48/boost/python/object_operators.hpp:105:1: note:
candidate function [with L =
      std::basic_ostream<char>, R = boost::python::tuple]
BOOST_PYTHON_BINARY_OPERATOR(<<)

/usr/include/boost-1_48/boost/python/object_operators.hpp:90:36: note:
expanded from macro
      'BOOST_PYTHON_BINARY_OPERATOR'
BOOST_PYTHON_BINARY_RETURN(object) operator op(L const& l, R const& r)  \


And an output operator for bools

/usr/lib/gcc/x86_64-pc-linux-gnu/4.5.3/include/g++-v4/ostream:172:7:
note: candidate function
      operator<<(bool __n)


How do I avoid this ambiguity with booleans? And how can I output tuples?

Btw. the compiler that I used is `clang version 3.1 (trunk 154212)`

Best regards,
Holger Brandsmeier

From herzog at mpi-cbg.de  Mon Aug 27 16:23:14 2012
From: herzog at mpi-cbg.de (Ronny Herzog)
Date: Mon, 27 Aug 2012 16:23:14 +0200
Subject: [C++-sig] building 32 Bit and 64 Bit extensions on the same
	machine
In-Reply-To: <CAMhLx01j7_NF=q=2YjZmtoRRs6S5UQCKKBfM-dbBMxB1yOz8vA@mail.gmail.com>
References: <k17s6o$8rd$1@ger.gmane.org>
	<CAMhLx01j7_NF=q=2YjZmtoRRs6S5UQCKKBfM-dbBMxB1yOz8vA@mail.gmail.com>
Message-ID: <k1fvsk$mmg$1@ger.gmane.org>

Thank you,

but this does not seem to work. If I do this with
"bjam --address-model=64 -n"
I can see that it does compile it with my 32Bit Python and without 
setting the x86_AMD64 flag:
"[...]
file 
C:\boost_1_50_0\bin.v2\libs\python\build\msvc-9.0\release\threading-multi\o
bject\iterator.obj.rsp
"C:\boost_1_50_0\libs\python\src\object\iterator.cpp" 
-Fo"C:\boost_1_50_0\bin.v2
\libs\python\build\msvc-9.0\release\threading-multi\object\iterator.obj" 
    -TP
/O2 /Ob2 /W3 /GR /MD /Zc:forScope /Zc:wchar_t /wd4675 /EHs -c
-DBOOST_ALL_NO_LIB=1
-DBOOST_PYTHON_SOURCE
-DNDEBUG
"-IC:\Python26\include"
"-IC:\boost_1_50_0"
compile-c-c++ 
C:\boost_1_50_0\bin.v2\libs\python\build\msvc-9.0\release\threadin
g-multi\object\iterator.obj

     call "C:\Program Files\Microsoft Visual Studio 
9.0\VC\vcvarsall.bat" x86 >nu
l
cl /Zm800 -nologo 
@"C:\boost_1_50_0\bin.v2\libs\python\build\msvc-9.0\release\th
reading-multi\object\iterator.obj.rsp"
[...]"

I am using MS Visual Studio 2008.

Unfortunately, I cannot test it on an 64Bit machine right now.

I have Python26 (32Bit) and Python26 (64Bit) installed and I guess I 
have to inform bjam somewhere which one to use.

Thanks a lot,
Ronny

Am 8/24/2012 4:49 PM, schrieb Wojciech Mamrak:
> Hi,
>
> add address-model=64 to Boost::Python bjam compilation command and
> that's all. Be aware, that not all MSVS versions (e.g. Express
> Edition) are equipped with 64 Bit compiler.
>
> 2012/8/24 Ronny Herzog <herzog at mpi-cbg.de>:
>> Dear all,
>>
>> I want to build 32 bit extensions and 64 bit extensions on my Windows 7 / 32
>> Bit with boost. After two days of checking google I am a bit frustrated. I
>> am new to Boost and it is a bit like a jungle.
>>
>> Can anyone help on what I need to do to set up Boost in the right way and
>> have a Jamroot which allows me to compile my Python extensions in 32 Bit and
>> 64 Bit at the same time (with MSVC 2008).
>>
>> Thanks a lot,
>> Ronny
>>
>> _______________________________________________
>> Cplusplus-sig mailing list
>> Cplusplus-sig at python.org
>> http://mail.python.org/mailman/listinfo/cplusplus-sig



From phil at philipdunstan.com  Fri Aug 31 03:00:24 2012
From: phil at philipdunstan.com (Philip Dunstan)
Date: Fri, 31 Aug 2012 03:00:24 +0200
Subject: [C++-sig] boost-python: Passing an object held in a
 boost::shared_ptr from C++ to python to C++
Message-ID: <CACiFdsfzyKYS74oR8Sw1eGdKCzjy3631XLLs9Qe=HUJFac8rqQ@mail.gmail.com>

Hi

I have some code that creates an object in a shared_ptr in C++. The object
is then used within python and passed from Python to another C++ function
that tries to keep hold of it. It appears that when the object goes from
C++ to Python it is converted from a shared_ptr<X> into a python object.
Then when it goes back into C++ it is converted from the python object into
a new shared_ptr<X>, however this shared_ptr is unrelated to the initial
holding shared_ptr.

Is it possible to set up the boost-python bindings so that the conversion
from python object to shared_ptr references the original shared_ptr
internals?

Below is the abridged code I have used to show the problem. I have attached
the full reproducible to this email. In this example an object is initially
held in a shared_ptr named s_iniital. It is grabbed from within python
through the getSharedPtr function, then pushed back into C++ through the
putSharedPtr function. Inside putSharedPtr it is copied into the weak_ptr
s_copied. Inspecting the pointers in the debugger shows that the shared_ptr
used in putSharedPtr does not have the same reference counting internals as
s_initial. The final assert will fire because the weak pointer s_copied was
only related to a single strong pointer (the pointer used in putSharedPtr)
and that pointer was destructed once putSharedPtr was finished.


static shared_ptr<Captured> s_initial;
static weak_ptr<Captured> s_copied;

class UseSharedPtr
{
public:
    shared_ptr<Captured> getSharedPtr()
    {
        return s_initial;
    }

    void putSharedPtr(shared_ptr<Captured> ptr)
    {
        s_copied = ptr;
    }
};


BOOST_PYTHON_MODULE(test)
{
    class_<Captured, shared_ptr<Captured>, boost::noncopyable>("Captured",
no_init);
    class_<UseSharedPtr, boost::noncopyable>("UseSharedPtr", init<>())
        .def("getSharedPtr", &UseSharedPtr::getSharedPtr)
        .def("putSharedPtr", &UseSharedPtr::putSharedPtr)
        ;
}



        s_initial = make_shared<Captured>();

        const char* chunk = "\
from test import UseSharedPtr \n\
x = UseSharedPtr() \n\
ptr = x.getSharedPtr() \n\
x.putSharedPtr(ptr)\n\
del x \n\
del ptr \n\
";
        object result = exec(chunk, mainNamespace, mainNamespace);

        assert(s_copied.lock());



thanks,
Phil
--
Philip Dunstan
phil at philipdunstan.com
www.philipdunstan.com
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/cplusplus-sig/attachments/20120831/eff1300d/attachment.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: test.cpp
Type: text/x-c++src
Size: 1862 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/cplusplus-sig/attachments/20120831/eff1300d/attachment.cpp>

From zao at acc.umu.se  Fri Aug 31 03:34:10 2012
From: zao at acc.umu.se (Lars Viklund)
Date: Fri, 31 Aug 2012 03:34:10 +0200
Subject: [C++-sig] building 32 Bit and 64 Bit extensions on the
	same	machine
In-Reply-To: <k1fvsk$mmg$1@ger.gmane.org>
References: <k17s6o$8rd$1@ger.gmane.org>
	<CAMhLx01j7_NF=q=2YjZmtoRRs6S5UQCKKBfM-dbBMxB1yOz8vA@mail.gmail.com>
	<k1fvsk$mmg$1@ger.gmane.org>
Message-ID: <20120831013410.GK622624@lenin.acc.umu.se>

On Mon, Aug 27, 2012 at 04:23:14PM +0200, Ronny Herzog wrote:
> Thank you,
>
> but this does not seem to work. If I do this with
> "bjam --address-model=64 -n"

Note that the previous poster did not say "--address-model=64".
He said "address-model=64". It's a feature, not an option.
Features do not have a leading dashdash.
You may of course have to point out a suitable Python path anyway, but
it helps if you actually instruct the build to be 64-bit in the first
place.

-- 
Lars Viklund | zao at acc.umu.se

From brandsmeier at gmx.de  Fri Aug 31 10:09:36 2012
From: brandsmeier at gmx.de (Holger Brandsmeier)
Date: Fri, 31 Aug 2012 10:09:36 +0200
Subject: [C++-sig] boost-python: Passing an object held in a
 boost::shared_ptr from C++ to python to C++
In-Reply-To: <CACiFdsfzyKYS74oR8Sw1eGdKCzjy3631XLLs9Qe=HUJFac8rqQ@mail.gmail.com>
References: <CACiFdsfzyKYS74oR8Sw1eGdKCzjy3631XLLs9Qe=HUJFac8rqQ@mail.gmail.com>
Message-ID: <CAFP7zVubfU_4aXCHg5KQbjKBOTRKTfv6cVxOHPjZJH9G6cdxuA@mail.gmail.com>

Dear Phil,

the problem you are phasing is because you use a weak pointer. I had a
very similar problem in April, you may read it up here
  http://mail.python.org/pipermail/cplusplus-sig/2012-April/016524.html
There I also tried to explain why this happens, I'm not sure if I did
a good job, just ask.

To be able to help you here, can you explain a little bit more about
your usecase?

In my case I was able to exploit that the weak_ptr that I needed to
access, was a pointer to the class itself. And I only ran into
trouble, that I wanted to call an unsafe function like your
`putSharedPtr` from python, when I extended this class from python.
Whenever I needed to use `s_copied` I call a function, say
`getPointer()`, that in pure C++ would just turn `s_copied` from
weak_ptr into shared_ptr. In python extended classes I implement
`getPointer()` not in terms of `s_copied`, but by just returning
`self`.

-Holger



On Fri, Aug 31, 2012 at 3:00 AM, Philip Dunstan <phil at philipdunstan.com> wrote:
> Hi
>
> I have some code that creates an object in a shared_ptr in C++. The object
> is then used within python and passed from Python to another C++ function
> that tries to keep hold of it. It appears that when the object goes from C++
> to Python it is converted from a shared_ptr<X> into a python object. Then
> when it goes back into C++ it is converted from the python object into a new
> shared_ptr<X>, however this shared_ptr is unrelated to the initial holding
> shared_ptr.
>
> Is it possible to set up the boost-python bindings so that the conversion
> from python object to shared_ptr references the original shared_ptr
> internals?
>
> Below is the abridged code I have used to show the problem. I have attached
> the full reproducible to this email. In this example an object is initially
> held in a shared_ptr named s_iniital. It is grabbed from within python
> through the getSharedPtr function, then pushed back into C++ through the
> putSharedPtr function. Inside putSharedPtr it is copied into the weak_ptr
> s_copied. Inspecting the pointers in the debugger shows that the shared_ptr
> used in putSharedPtr does not have the same reference counting internals as
> s_initial. The final assert will fire because the weak pointer s_copied was
> only related to a single strong pointer (the pointer used in putSharedPtr)
> and that pointer was destructed once putSharedPtr was finished.
>
>
> static shared_ptr<Captured> s_initial;
> static weak_ptr<Captured> s_copied;
>
> class UseSharedPtr
> {
> public:
>     shared_ptr<Captured> getSharedPtr()
>     {
>         return s_initial;
>     }
>
>     void putSharedPtr(shared_ptr<Captured> ptr)
>     {
>         s_copied = ptr;
>     }
> };
>
>
> BOOST_PYTHON_MODULE(test)
> {
>     class_<Captured, shared_ptr<Captured>, boost::noncopyable>("Captured",
> no_init);
>     class_<UseSharedPtr, boost::noncopyable>("UseSharedPtr", init<>())
>         .def("getSharedPtr", &UseSharedPtr::getSharedPtr)
>         .def("putSharedPtr", &UseSharedPtr::putSharedPtr)
>         ;
> }
>
>
>
>         s_initial = make_shared<Captured>();
>
>         const char* chunk = "\
> from test import UseSharedPtr \n\
> x = UseSharedPtr() \n\
> ptr = x.getSharedPtr() \n\
> x.putSharedPtr(ptr)\n\
> del x \n\
> del ptr \n\
> ";
>         object result = exec(chunk, mainNamespace, mainNamespace);
>
>         assert(s_copied.lock());
>
>
>
> thanks,
> Phil
> --
> Philip Dunstan
> phil at philipdunstan.com
> www.philipdunstan.com
>
> _______________________________________________
> Cplusplus-sig mailing list
> Cplusplus-sig at python.org
> http://mail.python.org/mailman/listinfo/cplusplus-sig