Comment 8 for bug 557804

Thanks for this update! it is indeed the same problem. It looks like optparse.py uses:

        return textwrap.fill(text,
                             text_width,
                             initial_indent=indent,
                             subsequent_indent=indent)

For ja.po this will be a utf8 encoded python str type. But textwrap needs a "unicode" type to wrap it correctly.