On Jan 21, 2013, at 10:31 PM, Karl Rupp <rupp at mcs.anl.gov> wrote:
> Hi Barry, > > >>>> Well I think it is insane in 2012 that we all need to stare at the exact >>>> same ASCII strings with the exact same placement (each of you thinks is >>>> wrong) when we develop shared code, so please teach me how. Can this be >>>> done through some bridge that uses the current hg PETSc repository? >>> >>> Yeah, even in *2013* it feels so wrong. >>> I'll see whether I can use Clang to provide this bridge. >> >> Please do, don't take my previous email attacking the Clang pretty >> printers too seriously, perhaps your keen eye can figure out how to get them >> to do everything we'd like to have. > > Will do. > > >>>> Unfortunately the CPP crap in PETSc messes things up a bit, curse that >>>> CPP wish we could eliminate it all. And yes CPP is C preprocessor not C++. >>> >>> Does it? For the formatting itself it's just another layer? >> >> Yes but it doesn't fit into the C AST. > > If I'm not mistaken, one can attach a preprocessor to a compiler object and > generate some kind of syntax tree as well as a full C AST. This allows some > additional back-and-forth operations. There is the theory, and then there is the practice :-( Barry > > Best regards, > Karli >
