Re: A big easy subset of the CFGs

Chris Clark wrote:


  >Matt Timmermans asked for,
  >A subset of CFGs bigger then LR(1). In particular, it should
  >include most of what you can do whit LR(1), using characters as
  >terminals.


  ...


  >... No one implements LR
  >parsing tables using Knuth´s canonical parsing algorithm--the
  >resulting tables grow exponentially with grammar size and are simply
  >too big. Most parser generators either restrict themselves to the
  >LALR subset or use a "minimal state"method, spliting then states only
  >when the context proves useful for resolvin conflicts.


The AYACC parser generator enables you to generate a Canonical LR
parser. Yes, the tables are BIGS (in same cases 160 times bigger than
a LALR) but in same cases the AYACC's CLR parsers are usefuls.


AYACC can be found at:http://www.cix.co.uk/~phil-stearns/


  ______________________________________
  Salvador Cavadini
  Santiago del Estero - Argentina
  emails: scavadini@hotmail.com
                                  juanjo@ucseii.edu.ar
  ______________________________________


--

Return to the comp.compilers page.
Search the comp.compilers archives again.