home *** CD-ROM | disk | FTP | other *** search
- Path: sparky!uunet!gatech!rutgers!faatcrl!iecc!compilers-sender
- From: markh@csd4.csd.uwm.edu (Mark)
- Newsgroups: comp.compilers
- Subject: Re: LR(0) vs. LALR, and the Great Parsing War
- Keywords: parse, LALR
- Message-ID: <92-09-042@comp.compilers>
- Date: 5 Sep 92 15:43:56 GMT
- References: <92-08-179@comp.compilers> <92-09-018@comp.compilers>
- Sender: compilers-sender@iecc.cambridge.ma.us
- Reply-To: markh@csd4.csd.uwm.edu (Mark)
- Organization: Computing Services Division, University of Wisconsin - Milwaukee
- Lines: 17
- Approved: compilers@iecc.cambridge.ma.us
-
- goer@midway.uchicago.edu (Richard L. Goerwitz) writes:
- >>Yes and no. If this were true, then Tomita's algorithm would have a cubic
- >>worst-case time factor. He uses a graph-structured parse forest, though,
- >>and claims polynomial time.
- >Exponential time factor, I mean. Not "cubic"!
-
- This is what I understand to be the case:
-
- The later (1992) reference made a couple enhancements to the algorithm
- that gives it cubic worst case performance in both space and time. That
- means that every possible parse, even if there is an infinite number of
- them, is accepted within those space and time limits.
-
- And it is still linear on LR(1) grammars.
- --
- Send compilers articles to compilers@iecc.cambridge.ma.us or
- {ima | spdcc | world}!iecc!compilers. Meta-mail to compilers-request.
-