Hi Jochen, I remember you well too. Frankly, I wasn't very interrested in your module because It gave me the feeling we wanted to run before we knwew how to walk. I think the idea is great but I a conservative bore! What I wanted to have base tools and a system to make them work together. There is also the problem of multiple threading in yacc that must, IMO, be solved (by another tool). What I wanted was a drop in replacement for lexers and parser that I or other might already have not a new perlish way to do things.
I'd like to the base modules in place: Flex Re2c Yacc Pbyacc Lemon ... might be other stuff that you know about that might be fun to learn. Could we collaborate to get those working and inter-working? Then (parallely in fact) I'd be very glad to help with your module (which I believe is a sort of Inline insde a language inside an Inline, see why I'm carefull but facinated?). We must also find a solution for the P::RD slowness because generated C code can be real huge. Brian talked about the Regex based solution and I had a very crude solution base on listing what we want exported (which is often one or two functions). I also proposed a solution base on tagging the functions and we talked about using 'nm'. I also thought about 'ctags' and there is a Parse::ExCtags module already. I'd be glad to receive your module so I can look at it and I can send my 2 or 3 hacked modules. I would be good that other voices are heard. I have a certain need but creating, documenting, testing and releasing modules is expensive (intime and energy) and borring when no one seem to be interrested to use them or give feed back. Cheers, Nadim. On Thursday 27 May 2004 08:08, Jochen Stenzel wrote: > Hi, Nadim, > > > Anyone interrested in Inline::Flex, yacc, pbyacc? > > I worked on an Inline::JACC years ago (we had a discussion that time, so > I remember you well). I stopped working on the project for private > reasons. It's still on my wishlist, or todo project list, but I'm > unsure when I will have the time to complete it. What I had so far was > a prototype which is still on my disk, working (as I remember) in some > areas, but having problems in others. There was still work to do. > > My goal was/is to produce a flexible solution, being capable to > integrate various compiler compilers (yacc, lemon, ...) and to write > semantic actions in whatever Inline language or Perl (Perl already > working), various (specifically named) parsers being usable in parallel > (worked, as I remember, but see below), with the yacc variant instantly > compatible to Parse::Yacc files, so that people could easily switch > from Parse::Yapp to Inline::JACC, reusing their grammar *and* semantic > actions, to speed up parsing. > > I'm not sure if my prototype still works with up to date Inline > versions. I remember problems with P::RD, possibly these were solved > when using the "new" regexp engine. > > Well ... I just run the recent test script. Having installed Inline 0.44 > now, it basically seems to work, but currently fails to start the > generated yyparse() (which is generated under another name, of course). > According to the error message, it cannot find the <parserfunction>.al > module, which seems to me like a modification in the way these modules > are produced, since I remember this part worked before - but it's a > long time that passed by, so my assumption may be wrong of course. Was > there a significant change in the interface in the last two years or > so? > > Well ... having played with my module now it seems to me that I'm still > interested in the project ;-) > > Greetings > > Jochen
