I would also like to note that with files that small, it should be possible to load the entire contents of both into memory. Processing via arrays instead of file access should significantly speed up your find.