Video caption conversion script using the PBS pycaption library

convert-subtitles.py

#!/usr/bin/env python
# encoding: utf-8
"""Convert caption files to different formats

One or more caption files will be converted to the specified output format, saved next to the input file

Requires the pycaption library from PBS: https://pypi.python.org/pypi/pycaption/
"""

from __future__ import absolute_import, print_function, unicode_literals

import logging
import os
import sys
from argparse import ArgumentParser, FileType

try:
    import pycaption
except ImportError:
    print('Unable to import pycaption: have you installed it?', file=sys.stderr)
    raise


SUPPORTED_WRITERS = {
    'dfxp': pycaption.DFXPWriter,
    'sami': pycaption.SAMIWriter,
    'srt': pycaption.SRTWriter,
    'scc': pycaption.SCCWriter,
    'webvtt': pycaption.WebVTTWriter,
}

FILE_EXTENSIONS = {
    'dfxp': 'dfxp.xml',
    'sami': 'sami',
    'srt': 'srt',
    'scc': 'scc',
    'webvtt': 'vtt',
}


def convert_file(input_captions, output_writer):
    reader = pycaption.detect_format(input_captions)

    if not reader:
        raise RuntimeError('Unrecognized format')

    converter = pycaption.CaptionConverter()
    converter.read(input_captions, reader())
    return converter.write(output_writer)


if __name__ == "__main__":
    parser = ArgumentParser(description=__doc__.strip())
    parser.add_argument('--output-format', type=lambda i: i.lower(), metavar='FORMAT', default='WebVTT',
                        choices=SUPPORTED_WRITERS.keys(),
                        help='Output format: %(choices)s (default: %(default)s)')
    parser.add_argument('caption_file', type=FileType('r'), nargs='+',
                        help='Caption files to convert')
    args = parser.parse_args()

    if args.output_format not in SUPPORTED_WRITERS:
        parser.error('Output format must be one of %s' % ' '.join(SUPPORTED_WRITERS.keys()))
    else:
        output_writer_class = SUPPORTED_WRITERS[args.output_format]

    logging.basicConfig()

    for f in args.caption_file:
        output_file = '%s.%s' % (os.path.splitext(f.name)[0],
                                 FILE_EXTENSIONS[args.output_format])

        print(output_file)

        try:
            with open(output_file, 'wb') as output_f:
            	funicode = f.read()
                output_f.write(convert_file(funicode.decode('utf-8'), output_writer_class()))
        except Exception as exc:
            import pdb; pdb.post_mortem(sys.exc_info()[2])
            logging.error('Unable to convert %s: %s', f.name, exc)
            f.close()
            continue

Thanks a ton for this! I've gotten it to work for .vtt, but seems like I'm having some trouble with .dfxp:

/usr/local/lib/python2.7/site-packages/pycaption/dfxp/base.py(63)read()
-> raise InvalidInputError('The content is not a unicode string.')
(Pdb)

before I have to Ctrl-C out. I'm not sure what's wrong.

Im trying to do the same (read .dfxp using pycaption.DFXPReader) and get the same InvalidInputError: The content is not a unicode string. error. The string that I pass is also unicode type. And when I convert it to a str. It comes back with CaptionReadNoCaptions: CaptionReadNoCaptions(('empty caption file',)). Could really use some help :)