Last month I had to do a PDF parser and searched a lot for a solution like Tika, but strangely this didn't come up. Cool, I'll test.