BasicURLFilter to remove URLs based on path repetition and max length

[jnioche]

The path repetition is very useful in most crawls there inevitably are sites that generate recursive URLs and rapidly take over everything else. We had a regex based filter inherited from Nutch but it is very slow. Instead we can put that functionality in a new BasicURLFilter which could also remove URLs based on their length.

These filters are meant to be super fast and so should be used early in the filtering chain.