My current RegEx is 157k. It searches text for ancient references which means it has all the various abbreviations for those references. It resides on GoogleAppEngine and can index 100k in less than 2 seconds but the RegEx portion of that is under a few milliseconds the rest of the processing time is processing each of the found objects. It got so big to handle that I had to write code to generate the RegEx from a flat file.

Good to find a mirror in the comments. Funny thing, but I needed to find this regexp today.

I never claimed the regexp to be cover all possible schemes. It didn’t 10 years ago, and it certainly doesn’t now. Assuming you have the Regexp::Common module installed, you can get a similar regexp with the following line of code: perl -MRegexp::Common -E ‘say $RE{URI}’

Also, this doesn’t match irc:// or ssh:// or several others

link ahref is badly typed. Also, foad.org is either down or dead. The correct url is:
http://web.archive.org/web/20070302134659/foad.org/~abigail/Perl/url3. regex

broken link