[Johnson_boy]

I am currently working on a new version which does not use the pawn preprocessor, but instead the files are parsed manually. The benefit is vastly improved performance on large files and the ability to include definitions as well. In addition, it allows you to parse includes in cases where the proprocessor would have left them out, like including both a_samp and a_npc.

The downside is that the output will be less accurate, as #if defined blocks for instance are no longer taken into account. Still, I believe that the positive effects outweight the negative ones in this case.