W
Walter Roberson
A few years ago, someone posted here with a reference to a module that
compiled standard regular expressions into a finite state machine
that could then be invoked. Sort of the reverse of DFA::Kleene,
and not requiring the full power of Parse::Lex because the only
necessary output is match / nomatch. Standard grep type regexps
are sufficient -- just the | [] . + * operators together
with ^ $ anchoring.
Would someone be able to steer me to the appropriate module?
(perl's pattern matcher does backtracking because it needs to be
more expressive than standard RE's. Regexp::Assemble helps some,
but still requires that perl's pattern matter be used; for what
I am doing, I am hoping for something that is time dependant
only on the length of the input string and independant of the
number of simple RE's being alternated.)
compiled standard regular expressions into a finite state machine
that could then be invoked. Sort of the reverse of DFA::Kleene,
and not requiring the full power of Parse::Lex because the only
necessary output is match / nomatch. Standard grep type regexps
are sufficient -- just the | [] . + * operators together
with ^ $ anchoring.
Would someone be able to steer me to the appropriate module?
(perl's pattern matcher does backtracking because it needs to be
more expressive than standard RE's. Regexp::Assemble helps some,
but still requires that perl's pattern matter be used; for what
I am doing, I am hoping for something that is time dependant
only on the length of the input string and independant of the
number of simple RE's being alternated.)