working with very large Finite State Machines

Steve Bradtke <sjbradtke@home.com>
22 Apr 2001 23:46:47 -0400

          From comp.compilers

Related articles
working with very large Finite State Machines sjbradtke@home.com (Steve Bradtke) (2001-04-22)
Re: working with very large Finite State Machines vannoord@let.rug.nl (2001-04-26)
Re: working with very large Finite State Machines chase@world.std.com (David Chase) (2001-04-26)
Re: working with very large Finite State Machines rboland@unb.ca (Ralph Boland) (2001-04-26)
Re: working with very large Finite State Machines rbates@southwind.net (Rodney M. Bates) (2001-04-29)
| List of all articles for this month |

From: Steve Bradtke <sjbradtke@home.com>
Newsgroups: comp.ai.nat-lang,comp.compilers
Date: 22 Apr 2001 23:46:47 -0400
Organization: Excite@Home - The Leader in Broadband http://home.com/faster
Keywords: lex, question
Posted-Date: 22 Apr 2001 23:46:47 EDT

Hi,


                I am working on a project that involves the construction of
very large Finite State Machines (up to approximately 10^7 states). I
am looking for software that will allow me to efficiently manipulate
these beasts. Starting from an initial non-deterministic version, I
need to be able to find the deterministic and minimal equivalents.
The automata will be used as components of a machine translation
project.


                The best I've been able to come up with so far is to modify
the various components of flex. However, this is still not really
fast enough, and flex has so many options and so little algorithmic
documentation that it is difficult to be sure that I'm not missing
important details.


                I would greatly appreciate any pointers to either source code
or papers that describe systems designed to handle very large Finite
State Machines.


Thanks in advance,
Steve
sjbradtke@home.com


Post a followup to this message

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