This should be a fix for #33. All files will now be re-encoded to UTF-8 if any other encoding is found.