[Insight-users] Help getting simple WrapITK external project going

Charl Botha c.p.botha at tudelft.nl
Tue Aug 4 04:57:14 EDT 2009


Dear Gaëtan,

2009/8/3 Gaëtan Lehmann <gaetan.lehmann at jouy.inra.fr>:
>> svn checkout http://itktudoss.googlecode.com/svn/trunk/ itktudoss
>>
>> This used to work with the built-in ITK 3.14 WrapITK.  I've made some
>> changes to better reflect what I see in the new wrapitk source tree,
>> but I can't get it working.
>
>
> I'm also having strange errors. I can't yet see the cause.
> I'll investigate tomorrow.

Some more info: The errors I'm seeing are in the generated CPP file
and look like this:
itktudossPython.cpp: In function ‘void
SWIG_Python_FixMethods(PyMethodDef*, swig_const_info*,
swig_type_info**, swig_type_info**)’:
itktudossPython.cpp:3338: error: expected unqualified-id before numeric constant

Googling a bit, I stumbled on this:
http://groups.google.com/group/comp.lang.c++/browse_thread/thread/69415ed5446f0e12

So I went to line 3338 of the generated CPP file above and manually
changed occurrences of the variable "c" into "c1".  Hey, it compiled!
It then breaked on the next generated file
wrap_itkTPGACLevelSetImageFilterPython.cpp on a similar error.  Lo and
behold, more "c".  It seems that this symbol is causing a name
collision with perhaps something from the preprocessor.

If you have any ideas on where this might be coming from, I'd be very
grateful to hear!

Thanks again,
Charl

P.S. I have a patch to the wrapitk cmakelists.txt for turning on
/bigobj on VS2005 and later.  Will send in another mail when i know
for sure that it works.


More information about the Insight-users mailing list