From: MenTaLguY <mental@...>
Date: 2007-04-13T01:11:35+09:00
Subject: Re: Slow ruby regexes

On Thu, 12 Apr 2007 16:59:47 +0900, "Robert Dober" <robert.dober@gmail.com> wrote:
> Now there does not seem to be any practical need to do this
> transformation though as the Thompson NFA algorithm kinda does the
> same thing in runtime, not creating states that are sets of states but
> just keeping a set of states.

The Thompson algorithm essentially just does the NFA -> DFA conversion lazily.

-mental