Currently, (because the unordered-list-parser function always consumes the next newline) you have to put two extra newlines after an unordered list in a paragraph. Otherwise if a header comes immediately after that paragraph, it will not be parsed properly. I don't really know how to fix this... yet!
Also, I should maybe change most of the parsers so that space is actually a token as well. This would make formatting a little more intuitive, and it would allow a bit saner behaviour as well.