[Numpy-discussion] Access to C/C++, typemaps, pyrex...

Fernando Perez Fernando.Perez at colorado.edu
Mon Mar 13 10:20:05 CST 2006


Travis Oliphant wrote:
> Ed Schofield wrote:
> 
>>Fernando Perez wrote:
>>  
>>So, just to clarify, you think we should house the SWIG typemaps in the
>>NumPy tree temporarily, until they're vetted, and then move them over to
>>the SWIG tree?
>>  
> 
> 
> I think that's what he was saying.  I think it's a good idea.  

Yes, that was my point.  For example, the typemaps I sent were 1 and 2d only, 
and some people have posted 3-d and n-d ones as well, so this is not complete 
yet.  But having it in the core will make sure that the contributions are 
actually recorded and improve our common usage tools.  Once we feel they are 
solid enough, we can (if we think it's worthwhile) send them to the SWIG team. 
  Though frankly, I think the most important part was having /something/ 
official in the numpy distribution itself: anyone needing this by definition 
has numpy, so we may actually be better off controlling this code rather than 
relying on SWIG versions and their release schedule.

In any case, that can be discussed later, the point was to have this, and I'm 
very thankful to Travis for pulling it in.  I'll delete it from the wavelets 
repo so we only have one copy of this code lying around.

Cheers,

f





More information about the Numpy-discussion mailing list