Working on a version of boilerpipe that supports using JSoup instead of xerces, etc and also supports extracting the HTML not just the text.

Also, this moves from ant to maven.

I haven't done much work here besides just getting it to work and getting maven setup.

TODO:

• Build out a LARGE number of tests (say 500-1000) that verify the output is correct. Do this SOON though because this way I can figure out if there's a regression easily.

• move to using multiple modules so that I can have one for nekohtml/xerces and another for jsoup

• tell the chromium guys about boilerpipe 2.0