I'm sorry I'm responding so late to this thread. I don't read this list very often. However, I would like to chime in and cast a vote for lemon. I've used it, and it's a great utility (better than bison IMO).
Having used yacc, bison, and now lemon, I can say the latter is a big improvement. Lemon being reentrant is a big win. Furthermore, I'd like to mention a scanner I've also used: re2c. Re2c is a much more flexible lexer (than flex or lex). It also can generate scanners w/o global variables, and it is EXTREMELY small compared to other lexers. I have a post on comp.compilers which says it all: http://groups.google.com/groups?q=group:comp.compilers+author:clint+author:olsen&hl=en&selm=01-02-123%40comp.compilers&rnum=5 See also http://www.tildeslash.org/re2c/index.html One of my big ideas was to link a C parser for verilog I wrote into Perl since many of our folks here are not C saavy. Unfortunately, I never found a nice way to pass multi-level data structures between the two domains (even Inline won't solve that), so I sort of gave up on it... Anyway, this type of development is certainly in the right direction. -Clint On Dec 11, Jochen Stenzel wrote: > > In my eyes, the problem is existing code and parser generator > installation. It would be fine to be able to reuse existing yacc grammars > as well as Parse::Yapp ones. Yacc is very popular and always there (on > UNIX) - regardless of lemons popularity which I cannot estimate. I do not > vote against lemon - it would be great to support it as well. I'm just > saying it would be pity if yacc would be missed.
