I took the code from http://ryanmerl.com/ (thanks dude) and fixed the broken regexp.
It's the simplest web crawler I found out there and actually pretty neat for some projects.
You can check theanti9 repo (or mine cause I forked it) to get a more complex web crawler.