Improving Parsing Quality

All of our information parsing mechanisms are quite "primitive" at the moment, a re-write could be useful for some of them. i.e. ytaudio, ytsubs(not too bad), pdf/doc/ppt extraction, web scraping,