[Insight-developers] Re: SWIG 1.3.22 in CableSwig?

[Gaëtan Lehmann]

Le Thu, 10 Aug 2006 16:21:27 +0200, William A. Hoffman  
<billlist at nycap.rr.com> a écrit:

> At 04:36 AM 8/10/2006, Gaetan Lehmann wrote:
>
>> Hi,
>>
>> The old version of SWIG included in cableswig becomes more and more a
>> problem.
>> Python 2.5 for example have changed the signature of some functions, so  
>> we
>> now have to use the -fpermissive option to be able to build WrapITK with
>> that python release.
>> Cableswig can't stay with an old release of SWIG too much time. It now
>> becomes a real problem.
>> I know that the long term way to go is to drop cableswig and to  
>> implement
>> an xml parser in swig to parse gccxml output, but nobody seems to be  
>> ready
>> to do that for now, so wouldn't it be a good idea to update the SWIG
>> version of cableswig untill someone found enough time to implement the
>> swig parser ?
>
> Updating swig is fine.   I am sure there will be issues.  There were  
> several
> "fixes" to the swig code that were made.  Some of those fixes did not  
> make
> it back into the swig main tree, so they will have to be "fixed" again.   
> I tried
> to send all fixes back, but many times it was hard to give an example in  
> pure
> swig that showed the problem.   I would estimate the effort of updating  
> to
> a new swig at two weeks.    However, if we are luck all the problems  
> will be fixed
> and the API will not have changed significantly, and it will only take a  
> day or so.
>
> I would not expect anyone at Kitware to do this until Sept.
>

Great !
I hope it will not be too much difficult :-)

Gaetan



-- 
Gaëtan Lehmann
Biologie du Développement et de la Reproduction
INRA de Jouy-en-Josas (France)
tel: +33 1 34 65 29 66    fax: 01 34 65 29 09
http://voxel.jouy.inra.fr