Related articles |
---|
[3 earlier articles] |
Re: yacc parse tree bob@werken.com (1998-12-18) |
Re: yacc parse tree belinfan@cs.utwente.nl (1998-12-18) |
Re: yacc parse tree josh_adams@bmc.com (Josh Adams) (1998-12-19) |
Re: yacc parse tree joachim.durchholz@munich.netsurf.de (Joachim Durchholz) (1998-12-19) |
Re: yacc parse tree scinar@ug.bcc.bilkent.edu.tr (Sukru Cinar) (1998-12-21) |
Re: yacc parse tree scinar@ug.bcc.bilkent.edu.tr (Sukru Cinar) (1998-12-22) |
Re: yacc parse tree AMartin@ma.ultranet.com (Alan H. Martin) (1999-01-02) |
Re: yacc parse tree scinar@ug.bcc.bilkent.edu.tr (Sukru Cinar) (1999-01-02) |
Re: yacc parse tree buzzard@world.std.com (1999-01-02) |
From: | "Alan H. Martin" <AMartin@ma.ultranet.com> |
Newsgroups: | comp.compilers |
Date: | 2 Jan 1999 00:23:47 -0500 |
Organization: | UltraNet Communications , an RCN Company http://www.ultranet.com/ |
References: | 98-12-018 98-12-026 |
Keywords: | lex, comment |
> ... grep is unrelated to lex even though the expression languages
> are similar, lex compiles the patterns to a DFA while the various greps
> either use an NFA (egrep and plain grep) or ad-hoc fixed string matching
> (fgrep). ...
> -John]
Note that while fgrep's pattern language is less expressive than grep or
egrep's, the usual implementation is classical enough to have been published
in CACM:
A. V. Aho and M. J. Corasick.
Efficient string matching: an aid to bibliographic search.
Communications of the ACM, 18(6):333-340, 1975.
/AHM
--
Alan Howard Martin AMartin@MA.UltraNet.Com
[Now that he reminds me, the fgrep algorithm is quite clever, sometimes
requiring less than one comparison per character. -John]
Return to the
comp.compilers page.
Search the
comp.compilers archives again.