Hi Javier,

On 2011-Jun-20, at 18:51 , Javier Pimás wrote:

> Hi! It's nice to hear this. As Mariano said, I started to work in  
> something like this some months ago, but I couldn't make something  
> useful yet because I don't have much free time. I can give you a  
> hand if you need. I don't know about OS Process, but I have some  
> experience with FFI, so I'll look for some examples.

That would be great.  FFI, Alien, and NativeBoost evidence an itch...  
but actual uses that depend on the code seem to be harder to find.  I  
get the impression that many potential users have devolved to writing  
their own purpose-built plugins.

> I also have some questions,
>
> - how did you implement the parser, by hand or with a framework?

By hand.  It doesn't handle all of C, but that part you will find in  
include files.

> - did you implement a preprocessor or used the output of something  
> like CPP?

I'm using CPP at the moment.  The only serious downside that I see is  
that lots of C headers pre-date the introduction of enums into C, so  
constants are often #defined rather than being part of an enum.  So I  
may end up adding a pre-processor to recover these constants.

> - what about the output of the parser, do you create an AST?

No, I directly build the data structures I need to generate the FFI  
interfaces.

> On Fri, Jun 17, 2011 at 1:34 PM, Dave Mason <dmason at mason-rose.ca>  
> wrote:
> I'm working on making a clean interface layer on top of FFI (and  
> possibly Alien and NativeBoost).