Thanks Jason.

Correct me if i'm wrong but doesn't yacc (or the parser) need something 
before it (like lex or a tokenizer) to 'tokenize' the text?

I can identify the tokens in parse.y but i was looking for a 
"specification" for them. Example: if i see "class" probably i want to 
create the token kCLASS, but i wanted to be sure..

Vasco Andrade e Silva
-- 
Posted via http://www.ruby-forum.com/.