> -----Original Message----- > From: Adam D. Lopresto [mailto:[EMAIL PROTECTED] > Sent: Friday, March 26, 2004 3:38 PM > To: Chris Santerre > Cc: 'Al Danks'; [EMAIL PROTECTED] > Subject: RE: Working rule to catch multiple occurrences of text > > > On Fri, 26 Mar 2004, Chris Santerre wrote: > > > > > > -----Original Message----- > > > From: Al Danks [mailto:[EMAIL PROTECTED] > > > Sent: Friday, March 26, 2004 11:56 AM > > > To: [EMAIL PROTECTED] > > > Subject: Working rule to catch multiple occurrences of text > > > > > > > > > full L_TEST > > > /(\b(best|most|more|((best|bett|bigg|larg|fast|quick)(er|est)) > > > )\b.*){3}/is > > > > > > > Well can I add a final silly tweak? :-) > > > > full L_TEST > > > /(\b?:(?:best|most|more|((?:best|bett|bigg|larg|fast|quick)(?: > er|est)))\b.*) > > {3}/is > > > > It is just quicker. > > Hmmm. Is "bestest" actually a spam sign? How about > "Bester". I mean, sure > he's not the greatest guy in the world, but do you really > want to annoy the > psi-corp? I'd probably write it as (untested) > > full L_TEST > /(?:\b(?:be(?:st|tter)|mo(?:st|re)|(?:bigg|larg|fast|quick)e(? > :r|st))\b.*){3}/is > -- > Adam Lopresto > http://cec.wustl.edu/~adam/ > > "Verbing weirds language." > --Bill Waterson, "Calvin and Hobbes"
Yeah that looks good. I think I had the first '?:' wrong. Al, basically the magic of the question mark is this: "I don't care which of these match. I don't really care to ever use the info again. So If one matchs, then good for you. You get a cookie, lets move on." And so the computers says, "Gee, I have a cookie!" and forgets which pattern matched, but just that somethng matched the regex. It is a bit faster, but your computer tends to get a sugar high. Which isn't bad, its the coming down from it that hurts ;) --Chris
