[Scilab-users] using intersci in 6.0.0 (was: Matlab vs Scilab perf; calling a fortran routine.)

[Clément David]

Hi all,

I was probably not clear enough, sorry about that : It is still possible to call fortran from
Scilab, using :
   * call() and dynamic-link capabilities
   * or an hand-written API Scilab C interface
   * or using an API Scilab C interface generator called SWIG

If you want to use SWIG, you have to write a C-like interface file containing function definition
for your fortran code as in the attached file and use `swig -scilab DGEMM.i` to generate the Scilab
C API to access DGEMM in Scilab.

Regards,

--
Clément

Le lundi 06 mars 2017 à 14:03 -0500, shorne at energetiq.com a écrit :
> No.  Scilab used to be linked against BLAS.  I expect it still is.  I would guess V6 t was done
> using SWIG,  but I'm guessing. 
> I used to build scilab from the source distribution on linux, back around versions 3 and 4 - it's
> become just to hairy to do that - the prerequisites have grown enormously.
> Back then there were many working examples of calling fortran.
>  
> It's still possible to connect scilab to fortran using the "call" function.
> It's just clumsy.   Possibly slow as well.    I use Mingw.
> 
> There's a reason why all the "modern" languages brag that they are almost as fast as fortran...
> Even "modern" Fortran has managed to break itself, with the new "modules" thing...  old-fashoned
> makefiles no longer work.
> 
> F77 was powerful, succinct, and efficient.    I guess it just wasn't cool.
> 
> 
> 
> Adelson Oliveira ---03/06/2017 01:33:04 PM---Does it mean that scilab no longer carry about
> interfacing with Fortran (if ever did it)? Modern For
> Adelson Oliveira ---03/06/2017 01:33:04 PM---Does it mean that scilab no longer carry about
> interfacing with Fortran (if ever did it)? Modern For
> 
> From:	Adelson Oliveira <adelson.oliveira at gmail.com>
> To:	Users mailing list for Scilab <users at lists.scilab.org>
> Date:	03/06/2017 01:33 PM
> Subject:	Re: [Scilab-users] using intersci in 6.0.0 (was: Matlab vs Scilab perf; calling a
> fortran routine.)
> Sent by:	"users" <users-bounces at lists.scilab.org>
> 
> 
> 
> Does it mean that scilab no longer carry about interfacing with Fortran (if ever did it)? Modern
> Fortran is my prefered language for production codes.
> 
> Thanks
> 
> 2017-03-06 14:28 GMT-03:00 <shorne at energetiq.com>:
> I had looked up SWIG.  I have found no examples of connecting fortran to scilab using swig.
> Clément David ---03/06/2017 03:54:24 AM---> While I'm on the subject, the old  Intersci system was
> a very convenient way to automatically > ge
> Clément David ---03/06/2017 03:54:24 AM---> While I'm on the subject, the old  Intersci system was
> a very convenient way to automatically > ge
> 
> From: Clément David <clement.david at scilab-enterprises.com>
> To: Users mailing list for Scilab <users at lists.scilab.org>
> Cc: shorne at energetiq.com
> Date: 03/06/2017 03:54 AM
> Subject: using intersci in 6.0.0 (was: Matlab vs Scilab perf; calling a fortran routine.)
> 
> 
> 
> > While I'm on the subject, the old  Intersci system was a very convenient way to automatically
> > generate the interface routine between scilab and an arbitrary fortran subroutine.  There seems
> > not to be recent documentation on doing the same (specifically for fortran).  Or am I missing
> > something?   I've had to use the "call" interface to use old code.  Is there a better way?
> 
> In Scilab 6.0.0, we did not reproduce and intersci code generation as this is handled by another
> tool called `SWIG` [1] for multiple scripting (or not) languages. This idea is to let the tool
> parse
> an C interface description (similar to a .h file with directives) and generate the wrapper code
> for
> a specific language.
> 
> Scilab is natively supported and the tool generates API Scilab code. IMHO this is way simpler to
> define in interface just writing C code instead of guessing what's the intersci encoding scheme :)
> for a specific parameter.
> 
> [1]: http://swig.org/
> 
> Thanks,
> 
> --
> Clément
> 
> 
> 
> _______________________________________________
> users mailing list
> users at lists.scilab.org
> http://lists.scilab.org/mailman/listinfo/users
> _______________________________________________
> users mailing list
> users at lists.scilab.org
> http://lists.scilab.org/mailman/listinfo/users
> 
> 
> _______________________________________________
> users mailing list
> users at lists.scilab.org
> http://lists.scilab.org/mailman/listinfo/users
-------------- next part --------------
A non-text attachment was scrubbed...
Name: DGEMM.i
Type: text/x-matlab
Size: 541 bytes
Desc: not available
URL: <https://lists.scilab.org/pipermail/users/attachments/20170308/7191dc85/attachment.bin>