"Nick Sabalausky" <a...@a.a> wrote in message news:ia01sk$1i7...@digitalmars.com... > "Nick Sabalausky" <a...@a.a> wrote in message > news:ia01q3$1i1...@digitalmars.com... >> "Walter Bright" <newshou...@digitalmars.com> wrote in message >> news:i9vn3l$bd...@digitalmars.com... >>> Nick Sabalausky wrote: >>>> What's wrong with regexes? >>> >>> They don't handle recursion. >> >> Neither do plain-old strings. But regexes will get you farther than plain >> strings before needing to resort to customized lexing. >> >> But I'm a big data-driven fan anyway. If you're not than I can see why it >> wouldn't seem as appealing as it does to me. >> >> In any case, if I have a chance I might see about adapting my Goldie ( >> www.dsource.org/projects/goldie ) library to more Phobos-friendly >> requirements. It's already a fully-usable lexer/parser (and the >> lexer/parser parts can be used independantly), with a complete grammar >> description language and I already have misc related tools written. And >> it's mostly working on D2 already (just need the next DMD because it has >> a fix for a bug that's a breaker for one of the tools). So if I can get >> it into a state more suitable for Phobos then that might end up putting >> things ahead of where they would be if someone just started from scratch. >> The initial versions might not be completely Phobos-ified, but it could >> definitely get there (especially if I had some guidance from people with >> more Phobos2 experience than me). Would Walter & co be interested in >> this? If not, I won't bother, but if so, then I may give it a shot. >> > > And FWIW, I was already thnking about making some improvements to Goldie's > API enyway. >
But that's all if you want generalized lexing or parsing though. If you just want "lexing D code"/"parsing D code", then IMO anything other than adapting parts of DDMD would be the wrong way to go.
