Add a comprehensive test corpus for the parser

[xeno-by]

I think we should do the same thing that we do in scalameta/scalameta - hand-pick a huge corpus of Scala files and write a test that ensures that our scanner and parser work for them in a reasonable fashion.

[xeno-by]

One cool thing that I liked about our tests is the fact that they verify that all characters in all sources are covered by tokens, with every character covered exactly by one token.

Another useful thing to check is the fact that all trees have positions.

[xeno-by]

checkparse introduced in #85 is pretty thorough, so that's the half of the ticket. The other half will be figured out and ticketed once we tackle #88.