Speed of lex generated scanners???

rtrigg@crl.com (Roger Trigg)
Thu, 26 May 1994 11:03:23 GMT

          From comp.compilers

Related articles
Speed of lex generated scanners??? rtrigg@crl.com (1994-05-26)
Re: Speed of lex generated scanners??? bazyar@netcom.com (1994-05-29)
Re: Speed of lex generated scanners??? wgsteven@undergrad.math.uwaterloo.ca (1994-05-30)
Re: Speed of lex generated scanners??? peter@csg.uwaterloo.ca (1994-05-30)
Re: Speed of lex generated scanners??? michi@km21.zfe.siemens.de (1994-06-01)
Re: Speed of lex generated scanners??? euambn@eua.ericsson.se (1994-06-06)
| List of all articles for this month |

Newsgroups: comp.compilers
From: rtrigg@crl.com (Roger Trigg)
Keywords: lex, flex, question, performance, comment
Organization: CRL Dialup Internet Access (415) 705-6060 [login: guest]
Date: Thu, 26 May 1994 11:03:23 GMT

I'm curious about the performance of lex generated scanners. I've done
some small preliminary benchmarks and the results were not to my liking. A
fairly simple table driven (large tables, greater than 5K), handwritten
scanner performed approximately 10 times faster than a lex generated
scanner. I'd rather not have to work out the full 5K tables by hand, so
I'm hoping I'm wrong, and actually lex generated scanners can be much
faster than my current tests have shown. So, I'm hoping someone can point
me to some references that could clear up a couple things for me. 1) How
do the tables in lex really work. (I've looked through the code, and it
didn't seem obvious; maybe I'm just dumb, or impatient)? 2) Benchmark
results for time required to scan some available C source on a particular
processor (preferably x86). 3) I've also heard that FLEX has a "fast"
option; does anyone have experience with this, and know how great of an
improvement can be expected using it?


    Many thanks in advance,
    Roger Trigg (rtrigg@crl.com)
[AT&T lex scanners are pretty slow, not to mention buggy. Everyone I know
uses flex which generates much faster scanners, particularly with the fast
option. -John]
--


Post a followup to this message

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