Dag Sverre Seljebotn <[EMAIL PROTECTED]> writes:

> The way I understood it (though I might be wrong) is that Andrey is 
> already using pyparsing to parse the .h file and produce a tree, and 
> then emits a string from that tree. Replacing the string-writing 
> commands with node-emitting commands shouldn't be such a pain (but it 
> depends on the rigour to which the .h files are parsed).
> 

Yeap I use pyparsing, because I don't know how PLEX works and how to use it 
with C/C++ sources. My program is written very crudely, firstly because of lack 
of skills with writing compilers and such things, secondly just because it 
already suits my needs.

> Of course, if the types are just treated as string identifier nodes of 
> some kind without any real interpretation (since Cython as a C-like 
> syntax and one can just shuffle things around) then one would need to 
> chain in some parts of the Cython parser. Difficult to tell without 
> seeing any code though.
> 
> Andrey: I think this discussion would be easier if we could see any 
> code, just to have a look at the approach etc, not necesarrily to use it.
> 

Sorry for long pause, I've been reading your opinions while seeking hosting for 
project. I've just got hosting on SF and uploaded my project there: http://
sourceforge.net/projects/ccython


_______________________________________________
Cython-dev mailing list
[email protected]
http://codespeak.net/mailman/listinfo/cython-dev

Reply via email to