Re: Typedefs of template instantiations not recognized properly?

Did you try the last CVS version?

This seem to work fine

------------- template_opaque.i -------------

%module template_opaque
%include std_vector.i

%{
 namespace A
 {
   struct OpaqueStruct
   {
   };
 }
%}


%inline {
namespace A {
 struct OpaqueStruct;
 typedef struct OpaqueStruct OpaqueType;
 typedef std::vector<OpaqueType> OpaqueVectorType;

 void FillVector(OpaqueVectorType& v)
 {
   for (int i = 0; i < v.size(); ++i) {
     v[i] = OpaqueStruct();
   }
 }
}
}

%template(OpaqueVectorType) std::vector<A::OpaqueType>;
---------------



--------------- template_opaque_runme.py ------

import template_opaque

v = template_opaque.OpaqueVectorType(10)

template_opaque.FillVector(v)

print v[0]

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

Marcelo


Zachary Pincus wrote:

> The typedef resolution mechanism seems (in my hands) to play poorly 
> with namespaces and/or the template mechanism.
>
> I've got a SWIG interface similar to:
> ----------------
>
> namespace A {
>     struct OpaqueStruct;
>     typedef struct OpaqueStruct OpaqueType;
>     typedef std::vector<OpaqueType> OpaqueVectorType;
>
>     void FillVector(OpaqueVectorType);
> }
>
> %template(OpaqueVectorType) std::vector<A::OpaqueType>;
>
> ----------------
>
> Now, when I use the wrapper as follows:
> >>> vec = SwigModule.OpaqueVectorType()
> >>> SwigModule.FillVector(vec)
>
> I get the following error:
> Type error. Got _p_std__vectorTA__OpaqueType_t, expected 
> _p_A__OpaqueVectorType
>
> That is, the wrapper code doesn't realize that std::vector<OpaqueType> 
> and A::OpaqueVectorType are the same thing!
>
> I can't really figure out if the problem is due to the template 
> mechanism, the typedef mechanism, or throwing namespaces into the mix. 
> (It doesn't matter whether the %template line is above or below the 
> namespace definition.)
>
> Any thoughts? Is this a known issue? Am I just doing something wrong?
> Basically, is there anything I can do, short or find/replace on the 
> interface files to get rid of the need for the typedefs?
>
> Thanks,
>
> Zach Pincus
>
> Department of Biochemistry and Program in Biomedical Informatics
> Stanford University School of Medicine
>
> _______________________________________________
> Swig maillist  -  Swig@xxxxxxxxxxxxxxx
> http://mailman.cs.uchicago.edu/mailman/listinfo/swig


_______________________________________________
Swig maillist  -  Swig@xxxxxxxxxxxxxxx
http://mailman.cs.uchicago.edu/mailman/listinfo/swig