Andrew Pimlott <[EMAIL PROTECTED]> wrote: > On Thu, Feb 14, 2002 at 10:20:50PM -0500, Jeff 'japhy' Pinyan > wrote: > > while ($str =~ m{ ([^\0]{$len,}) (?= [^\0]* \0 [^\0]*? \1 ) }xg) { > > Thanks to the comma, ^ > > > $len = length($match = $1) + 1; > > } > > this is O(N^3) in the case of no common substring:
I don't think the /g is necessary either, unless there's a requirement to find the last substring if there's more than one of the same length. And without the /g there's no need to make the last part of the match a lookahead. -- Keith C. Ivey <[EMAIL PROTECTED]> Washington, DC