Re: eXene and openGL
Matthew Fluet <[email protected]>
| Newsgroups | gmane.comp.lang.sml.smlnj |
|---|---|
| Organization | Toyota Technological Institute at Chicago |
| Message-ID | <[email protected]> |
Vesa Karvonen wrote: > Quoting brian <briand-/[email protected]>: > [...] >> ml-ffigen is not robust enough to deal with the C include insanity >> and will often bail out. > [...] > > Yes, I've noticed this too although I've been mostly using the MLton > version (mlnlffigen). The problem seems to be that ml[-]nlffigen > (understandably) does not deal with non-standard C and compiler/OS > specific headers regularly contain non-standard C. The differences between the SML/NJ and MLton versions of ml[-]nlffigen are fairly minor in terms of the front-end (i.e., in what and how they process C code); the major differences are the back-end that produces SML files that are suitable for the different low-level FFI of the compilers. With regards to the front-end, ml[-]nlffigen uses the C Kit Library for parsing and analyzing C headers. And the C Kit only deals with pre-processed C code (i.e., gcc -E). > It might be possible to improve ml[-]nlffigen in this respect in various > ways (e.g. implement common extensions and make it deal more gracefully > with partial information and errors). You'd probably need to improve both the C Kit and ml[-]nlffigen to handle various extensions. ------------------------------------------------------------------------- Take Surveys. Earn Cash. Influence the Future of IT Join SourceForge.net's Techsay panel and you'll get the chance to share your opinions on IT & business topics through brief surveys - and earn cash http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV