Stripping out audio tracks from MKV files

/ February 11, 2017/ code

I have some MKV files where the default audio track is non-English, though there is an English audio track in the file. Some research led me to MKVMerge, part of the MKVToolNix collection. A bit more research led me to some Python2 code that’s able to identify the tracks in the file, apply a filter, and instruct mkvmerge to write out a new file with just the chosen language tracks and subtitles.

I made it a bit more Python3 friendly, and way, way longer. Far too much logger.critical() used, because I couldn’t be bothered to fix the loglevel; print() would have been fine I suppose. Too used to using the logger in code at work, where there’s a whole infrastructure behind it to vacuum up all the logs and make them available centrally. But hey, it works for me, and it might be useful to someone searching Google some day.

This is all on Windows; get Python at python.org.

#!/usr/bin/python

from collections import OrderedDict
from hashlib import md5
from io import StringIO
from os.path import isfile, join
from tabulate import tabulate

import argparse
import logging
import os
import re
import subprocess
import sys


parser = argparse.ArgumentParser(
    description='Strip MKV files to a single language')
parser.add_argument(
    '--input-directory',
    help='Directory with source MKV files to be processed',
    required=True,
)
parser.add_argument(
    '--audio-language',
    help='Audio language to keep. All other audio languages will be stripped',
)
parser.add_argument(
    '--subtitle-language',
    help='Subtitle language to keep. All other subtitle languages will be stripped',
    required=False,
)
parser.add_argument(
    '--list-tracks',
    help='List tracks in the files in the input directory',
    action='store_true'
)

args = parser.parse_args()
logging.basicConfig(format='%(asctime)s %(message)s')
logger = logging.getLogger('converter')
file_to_tracks = {}

# set this to the path for mkvmerge
MKVMERGE = "C:/Program Files/MKVtoolnix/mkvmerge.exe"

AUDIO_RE = re.compile(r"Track ID (\d+): audio .*language:([a-z]{3})")
SUBTITLE_RE = re.compile(r"Track ID (\d+): subtitles .*language:([a-z]{3})")


def _get_file_list(root=None):
    logger.critical('Processing {}'.format(root))
    dir_contents = os.listdir(root)
    paths = [f for f in dir_contents 
                if isfile(join(root, f)) and f.endswith('.mkv')]
    return paths


def _mkvmerge_identify(root=None, filename=None):
    logger.critical('[{}] Identifying'.format(filename))
    cmd = [MKVMERGE, "--identify-verbose", join(root, filename)]
    mkvmerge = subprocess.Popen(
        cmd, stdout=subprocess.PIPE, stderr=subprocess.PIPE)
    stdout, stderr = mkvmerge.communicate()
    if mkvmerge.returncode != 0:
        logger.critical(
            '[{}] mkvmerge failed to identify the file'.format(filename))
        raise Exception(stdout.decode('utf-8'))
    return stdout.decode('utf-8')


def _extract_tracks(mkvmerge_output=None, filename=None):
    logger.critical('[{}] Extracting tracks'.format(filename))
    audio = []
    subtitle = []

    for line in StringIO(content):
        m = AUDIO_RE.match(line)
        if m:
            audio.append(m.groups())
        else:
            m = SUBTITLE_RE.match(line)
            if m:
                subtitle.append(m.groups())
    logger.critical(
        '[{}] Found {} audio and {} subtitle tracks'.format(
            filename, len(audio), len(subtitle)))
    return audio, subtitle


def _list_tracks(file_tracks=None):
    for key, rec in file_tracks.items():
        print('{}'.format(rec.filename))
        print(' Audio tracks')
        if rec.audio:
            for t in rec.audio:
                print('  {} is {}'.format(t[0], t[1]))
        if rec.subtitle:
            for t in rec.subtitle:
                print('  {} is {}'.format(t[0], t[1]))
        print("")


def _audio_check(file_tracks=None):
    for key, rec in file_tracks.items():
        if len(rec.audio) < 2:
            logger.critical(
                '[{}] At least 1 audio track required'.format(rec.filename))
            del file_tracks[key]
    return file_tracks
    

def _subtitle_check(file_tracks=None):
    for key, rec in file_tracks.items():
        if len(rec.subtitle) < 2:
            logger.critical(
                '[{}] At least 1 subtitle track required'.format(rec.filename))
            del file_tracks[key]
    return file_tracks


def _build_args(langtype=None):
    if langtype not in ['audio', 'subtitle']:
        Exception('Type must be audio or subtitle')

    for key, rec in file_to_tracks.items():
        logger.critical(
            '[{}] Building CLI arguments for {} tracks'.format(
                rec.filename, langtype))
        if 'audio' in langtype:
            lang = args.audio_language
            field = rec.audio
        if 'language' in langtype:
            lang = args.audio_language
            field = rec.audio
        cmd = []
        filterlang = list(filter(lambda a: a[1]==lang, field))
        if lang and len(filterlang) == 0:
            logger.critical(
                '[{}] No {} tracks with language {} in {}'.format(
                    rec.filename, langtype, lang, path))
            continue
        if len(filterlang) > 1:
            logger.critical(
                '[{}] More than one {} track matching {}. Skipping'.format(
                    rec.filename, langtype, lang))
            continue
        if len(filterlang):
            cmd = ["--{}-tracks".format(langtype),
                    ",".join([str(a[0]) for a in filterlang])]
            for i in range(len(filterlang)):
                cmd += ["--default-track"]
                cmd += [":".join([filterlang[i][0], "0" if i else "1"])]
        if 'audio' in langtype:
            rec.audio_args = cmd
        if 'audio' in langtype:
            rec.subtitle_args = cmd
        file_to_tracks[key] = rec


class MKVFile:
    filename = None
    audio = None
    subtitle = None
    root = None
    audio_args = []
    subtitle_args = []

    def __init__(self, root=None, filename=None, audio=None, subtitle=None):
        self.filename = filename
        self.audio = audio
        self.subtitle = subtitle
        self.root = root

    def __repr__(self):
        return '{filename} {audio_args} {subtitle_args}'.format(
            filename=self.filename, audio_args=self.audio_args,
            subtitle_args=self.subtitle_args)


files = _get_file_list(root=args.input_directory)
for filename in files:
    content = _mkvmerge_identify(
        root=args.input_directory, filename=filename)
    audio_tracks, subtitle_tracks = _extract_tracks(
        mkvmerge_output=content, filename=filename)
    file_to_tracks[md5(filename.encode()).hexdigest()] = MKVFile(
        root=args.input_directory,
        filename=filename, 
        audio=audio_tracks, 
        subtitle=subtitle_tracks)

if args.list_tracks:
    _list_tracks(file_tracks=file_to_tracks)

if args.audio_language:
    file_to_tracks = _audio_check(file_tracks=file_to_tracks)

if args.subtitle_language:
    file_to_tracks = _subtitle_check(file_tracks=file_to_tracks)

if args.audio_language:
    _build_args('audio')
        
if args.subtitle_language:
    _build_args('subtitle')

for key, rec in file_to_tracks.items():
    if not rec.audio_args and not rec.subtitle_args:
        logger.critical('[{}] No work to be done'.format(rec.filename))
        continue

    path = join(rec.root, rec.filename)
    old_file = 'old' + rec.filename
    cmd = [MKVMERGE, "-o", path + ".temp"]
    cmd += rec.audio_args
    cmd += rec.subtitle_args
    cmd += [path]
    logger.critical('[{}] Processing ...'.format(rec.filename))

    mkvmerge = subprocess.Popen(cmd, stdout=subprocess.PIPE, stderr=subprocess.PIPE)
    stdout, stderr = mkvmerge.communicate()
    if mkvmerge.returncode != 0:
        logger.critical('Failed')
        raise Exception(stdout)

    logger.critical('Success')

    os.rename(path, join(rec.root, old_file))
    os.rename(path + ".temp", path)

Leave a Comment

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>
*
*