in reply to Re: Re (tilly) 3: Research ideas
in thread Research ideas

A NFA is the opposite, it may (often) be much faster than a DFA,

I assume you mean this in the context of including init times? Excluding the DFA table build phase the DFA should be faster or equivelant to a NFA.

And I would absolutely welcome a DFA engine in addition to the existant NFA engine in perl. Especially if the DFA engine had a way of saving its state table. The possibilities are endless....

Yves / DeMerphq
---
Writing a good benchmark isnt as easy as it might look.