m (maybe implement Thompson NFA for regular expression library)
Revision as of 16:59, 24 January 2007
I just came across this article on Thompson Non-Finite Automata which presents an alternative implementation (to the one supposedly used in perl, ruby, python) which is MAAANY times faster.
Seing how badly we did in the shootout initially, wouldn't it be grand if "we" implemented this algorithm and could then outrun every other language WRT regular expressions, with a native haskell library...
This article is a stub. You can help by expanding it.