Re: Backtracking yacc

sasghm@unx.sas.com (Gary Merrill) writes:
> andrewd@cs.adelaide.edu.au (Andrew Dunstan) writes:
> |> LALR(1) is powerful enough for most purposes. (In fact I
> |> am more taken with LL(1), for reasons I won't go into here.)
>
> Wait a minute! Please *do* go into these reasons here. I think they
> might very well be of general interest. I to am more "taken" will LL(1)
> grammars. They seem more "natural" to me. I think I could make this
> more precise, but I am quite interested in hearing the feelings of others.


Has anybody of the persons involved in this discussion actually tried
PCCTS (the Purdue Compiler Construction ToolSet)? It produces parsers for
LL(k) grammars.
        In addition, it has already some support routines for the creation of
ASTs at run time. The parser specification language resembles EBNF; the
lexeme specifications are similar to UNIX regexps. (That's most annoying
-- that they are only _similar_. :-( )
        The resulting C code is even debuggable...


I haven't yet used it for serious developments, so I would be interested
in comments about this tool. (Stability, maintainibility, etc.)


Oh yes,
  -- the reference manual was published in SIGPLAN 27(2), Feb 92, p. 88-165.
  -- anonymous ftp in the US:
en.ecn.purdue.edu
directory /carp/PCCTS
  -- European folks might want to try
ftp.th-darmstadt.de [130.83.55.75]
directory pub/programming/languages/compiler-compiler/pccts
        (This is mirrored automatically, so it's up to date.)
[You'll find also cocktail and eli there.
More contributions are always welcome. ;-)]
--
Joachim Schrod Email: schrod@iti.informatik.th-darmstadt.de
Computer Science Department
Technical University of Darmstadt, Germany
[PCCTS has been discussed before, most recently in message 92-05-006. -John]
--

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