path: root/youtube/watch.py

import youtube
from youtube import yt_app
from youtube import util, comments, local_playlist, yt_data_extract
from youtube.util import time_utc_isoformat
import settings

from flask import request
import flask
import logging

logger = logging.getLogger(__name__)

import json
import gevent
import os
import math
import traceback
import urllib
import re
import urllib3.exceptions
from urllib.parse import parse_qs, urlencode
from types import SimpleNamespace
from math import ceil


try:
    with open(os.path.join(settings.data_dir, 'decrypt_function_cache.json'), 'r') as f:
        decrypt_cache = json.loads(f.read())['decrypt_cache']
except FileNotFoundError:
    decrypt_cache = {}


def codec_name(vcodec):
    if vcodec.startswith('avc'):
        return 'h264'
    elif vcodec.startswith('av01'):
        return 'av1'
    elif vcodec.startswith('vp'):
        return 'vp'
    else:
        return 'unknown'


def get_video_sources(info, target_resolution):
    '''return dict with organized sources'''
    audio_by_track = {}
    video_only_sources = {}
    uni_sources = []
    pair_sources = []

    for fmt in info['formats']:
        if not all(fmt[attr] for attr in ('ext', 'url', 'itag')):
            continue
        if fmt['acodec'] and fmt['vcodec']:
            if fmt.get('audio_track_is_default', True) is False:
                continue
            source = {'type': 'video/' + fmt['ext'],
                      'quality_string': short_video_quality_string(fmt)}
            source['quality_string'] += ' (integrated)'
            source.update(fmt)
            uni_sources.append(source)
            continue
        if not (fmt['init_range'] and fmt['index_range']):
            # Allow HLS-backed audio tracks (served locally, no init/index needed)
            if not fmt.get('url', '').startswith('http://127.') and not '/ytl-api/' in fmt.get('url', ''):
                continue
            # Mark as HLS for frontend
            fmt['is_hls'] = True
        if fmt['acodec'] and not fmt['vcodec'] and (fmt['audio_bitrate'] or fmt['bitrate']):
            if fmt['bitrate']:
                fmt['audio_bitrate'] = int(fmt['bitrate']/1000)
            source = {'type': 'audio/' + fmt['ext'],
                      'quality_string': audio_quality_string(fmt)}
            source.update(fmt)
            source['mime_codec'] = source['type'] + '; codecs="' + source['acodec'] + '"'
            tid = fmt.get('audio_track_id') or 'default'
            if tid not in audio_by_track:
                audio_by_track[tid] = {
                    'name': fmt.get('audio_track_name') or 'Default',
                    'is_default': fmt.get('audio_track_is_default', True),
                    'sources': [],
                }
            audio_by_track[tid]['sources'].append(source)
        elif all(fmt[attr] for attr in ('vcodec', 'quality', 'width', 'fps', 'file_size')):
            if codec_name(fmt['vcodec']) == 'unknown':
                continue
            source = {'type': 'video/' + fmt['ext'],
                      'quality_string': short_video_quality_string(fmt)}
            source.update(fmt)
            source['mime_codec'] = source['type'] + '; codecs="' + source['vcodec'] + '"'
            quality = str(fmt['quality']) + 'p' + str(fmt['fps'])
            video_only_sources.setdefault(quality, []).append(source)

    audio_tracks = []
    default_track_id = 'default'
    for tid, ti in audio_by_track.items():
        audio_tracks.append({'id': tid, 'name': ti['name'], 'is_default': ti['is_default']})
        if ti['is_default']:
            default_track_id = tid
    audio_tracks.sort(key=lambda t: (not t['is_default'], t['name']))

    default_audio = audio_by_track.get(default_track_id, {}).get('sources', [])
    default_audio.sort(key=lambda s: s['audio_bitrate'])
    uni_sources.sort(key=lambda src: src['quality'])
    webm_audios = [a for a in default_audio if a['ext'] == 'webm']
    mp4_audios = [a for a in default_audio if a['ext'] == 'mp4']

    for quality_string, sources in video_only_sources.items():
        # choose an audio source to go with it
        # 0.5 is semiarbitrary empirical constant to spread audio sources
        # between 144p and 1080p. Use something better eventually.
        quality, fps = map(int, quality_string.split('p'))
        target_audio_bitrate = quality*fps/30*0.5
        pair_info = {
            'quality_string': quality_string,
            'quality': quality,
            'height': sources[0]['height'],
            'width': sources[0]['width'],
            'fps': fps,
            'videos': sources,
            'audios': [],
        }
        for audio_choices in (webm_audios, mp4_audios):
            if not audio_choices:
                continue
            closest_audio_source = audio_choices[0]
            best_err = target_audio_bitrate - audio_choices[0]['audio_bitrate']
            best_err = abs(best_err)
            for audio_source in audio_choices[1:]:
                err = abs(audio_source['audio_bitrate'] - target_audio_bitrate)
                # once err gets worse we have passed the closest one
                if err > best_err:
                    break
                best_err = err
                closest_audio_source = audio_source
            pair_info['audios'].append(closest_audio_source)

        if not pair_info['audios']:
            continue

        def video_rank(src):
            ''' Sort by settings preference. Use file size as tiebreaker '''
            setting_name = 'codec_rank_' + codec_name(src['vcodec'])
            return (settings.current_settings_dict[setting_name],
                    src['file_size'])
        pair_info['videos'].sort(key=video_rank)

        pair_sources.append(pair_info)

    pair_sources.sort(key=lambda src: src['quality'])

    uni_idx = 0 if uni_sources else None
    for i, source in enumerate(uni_sources):
        if source['quality'] > target_resolution:
            break
        uni_idx = i

    pair_idx = 0 if pair_sources else None
    for i, pair_info in enumerate(pair_sources):
        if pair_info['quality'] > target_resolution:
            break
        pair_idx = i

    audio_track_sources = {}
    for tid, ti in audio_by_track.items():
        srcs = ti['sources']
        srcs.sort(key=lambda s: s.get('audio_bitrate', 0))
        audio_track_sources[tid] = srcs

    return {
        'uni_sources': uni_sources,
        'uni_idx': uni_idx,
        'pair_sources': pair_sources,
        'pair_idx': pair_idx,
        'audio_tracks': audio_tracks,
        'audio_track_sources': audio_track_sources,
    }


def make_caption_src(info, lang, auto=False, trans_lang=None):
    label = lang
    if auto:
        label += ' (Automatic)'
    if trans_lang:
        label += ' -> ' + trans_lang

    # Try to use Android caption URL directly (no PO Token needed)
    caption_url = None
    for track in info.get('_android_caption_tracks', []):
        track_lang = track.get('languageCode', '')
        track_kind = track.get('kind', '')
        if track_lang == lang and (
            (auto and track_kind == 'asr') or
            (not auto and track_kind != 'asr')
        ):
            caption_url = track.get('baseUrl')
            break

    if caption_url:
        # Add format
        if '&fmt=' in caption_url:
            caption_url = re.sub(r'&fmt=[^&]*', '&fmt=vtt', caption_url)
        else:
            caption_url += '&fmt=vtt'
        if trans_lang:
            caption_url += '&tlang=' + trans_lang
        url = util.prefix_url(caption_url)
    else:
        # Fallback to old method
        url = util.prefix_url(yt_data_extract.get_caption_url(info, lang, 'vtt', auto, trans_lang))

    return {
        'url': url,
        'label': label,
        'srclang': trans_lang[0:2] if trans_lang else lang[0:2],
        'on': False,
    }


def lang_in(lang, sequence):
    '''Tests if the language is in sequence, with e.g. en and en-US considered the same'''
    if lang is None:
        return False
    lang = lang[0:2]
    return lang in (l[0:2] for l in sequence)


def lang_eq(lang1, lang2):
    '''Tests if two iso 639-1 codes are equal, with en and en-US considered the same.
       Just because the codes are equal does not mean the dialects are mutually intelligible, but this will have to do for now without a complex language model'''
    if lang1 is None or lang2 is None:
        return False
    return lang1[0:2] == lang2[0:2]


def equiv_lang_in(lang, sequence):
    '''Extracts a language in sequence which is equivalent to lang.
    e.g. if lang is en, extracts en-GB from sequence.
    Necessary because if only a specific variant like en-GB is available, can't ask YouTube for simply en. Need to get the available variant.'''
    lang = lang[0:2]
    for l in sequence:
        if l[0:2] == lang:
            return l
    return None


def get_subtitle_sources(info):
    '''Returns these sources, ordered from least to most intelligible:
    native_video_lang (Automatic)
    foreign_langs (Manual)
    native_video_lang (Automatic) -> pref_lang
    foreign_langs (Manual) -> pref_lang
    native_video_lang (Manual) -> pref_lang
    pref_lang (Automatic)
    pref_lang (Manual)'''
    sources = []
    if not yt_data_extract.captions_available(info):
        return []
    pref_lang = settings.subtitles_language
    native_video_lang = None
    if info['automatic_caption_languages']:
        native_video_lang = info['automatic_caption_languages'][0]

    highest_fidelity_is_manual = False

    # Sources are added in very specific order outlined above
    # More intelligible sources are put further down to avoid browser bug when there are too many languages
    # (in firefox, it is impossible to select a language near the top of the list because it is cut off)

    # native_video_lang (Automatic)
    if native_video_lang and not lang_eq(native_video_lang, pref_lang):
        sources.append(make_caption_src(info, native_video_lang, auto=True))

    # foreign_langs (Manual)
    for lang in info['manual_caption_languages']:
        if not lang_eq(lang, pref_lang):
            sources.append(make_caption_src(info, lang))

    if (lang_in(pref_lang, info['translation_languages'])
            and not lang_in(pref_lang, info['automatic_caption_languages'])
            and not lang_in(pref_lang, info['manual_caption_languages'])):
        # native_video_lang (Automatic) -> pref_lang
        if native_video_lang and not lang_eq(pref_lang, native_video_lang):
            sources.append(make_caption_src(info, native_video_lang, auto=True, trans_lang=pref_lang))

        # foreign_langs (Manual) -> pref_lang
        for lang in info['manual_caption_languages']:
            if not lang_eq(lang, native_video_lang) and not lang_eq(lang, pref_lang):
                sources.append(make_caption_src(info, lang, trans_lang=pref_lang))

        # native_video_lang (Manual) -> pref_lang
        if lang_in(native_video_lang, info['manual_caption_languages']):
            sources.append(make_caption_src(info, native_video_lang, trans_lang=pref_lang))

    # pref_lang (Automatic)
    if lang_in(pref_lang, info['automatic_caption_languages']):
        sources.append(make_caption_src(info, equiv_lang_in(pref_lang, info['automatic_caption_languages']), auto=True))

    # pref_lang (Manual)
    if lang_in(pref_lang, info['manual_caption_languages']):
        sources.append(make_caption_src(info, equiv_lang_in(pref_lang, info['manual_caption_languages'])))
        highest_fidelity_is_manual = True

    if sources and sources[-1]['srclang'] == pref_lang:
        # set as on by default since it's manual a default-on subtitles mode is in settings
        if highest_fidelity_is_manual and settings.subtitles_mode > 0:
            sources[-1]['on'] = True
        # set as on by default since settings indicate to set it as such even if it's not manual
        elif settings.subtitles_mode == 2:
            sources[-1]['on'] = True

    if len(sources) == 0:
        assert len(info['automatic_caption_languages']) == 0 and len(info['manual_caption_languages']) == 0

    return sources


def get_ordered_music_list_attributes(music_list):
    # get the set of attributes which are used by atleast 1 track
    # so there isn't an empty, extraneous album column which no tracks use, for example
    used_attributes = set()
    for track in music_list:
        used_attributes = used_attributes | track.keys()

    # now put them in the right order
    ordered_attributes = []
    for attribute in ('Artist', 'Title', 'Album'):
        if attribute.lower() in used_attributes:
            ordered_attributes.append(attribute)

    return ordered_attributes


def save_decrypt_cache():
    os.makedirs(settings.data_dir, exist_ok=True)
    f = open(os.path.join(settings.data_dir, 'decrypt_function_cache.json'), 'w')

    f.write(json.dumps({'version': 1, 'decrypt_cache':decrypt_cache}, indent=4, sort_keys=True))
    f.close()


def decrypt_signatures(info, video_id):
    '''return error string, or False if no errors'''
    if not yt_data_extract.requires_decryption(info):
        return False
    if not info['player_name']:
        return 'Could not find player name'

    player_name = info['player_name']
    if player_name in decrypt_cache:
        print('Using cached decryption function for: ' + player_name)
        info['decryption_function'] = decrypt_cache[player_name]
    else:
        base_js = util.fetch_url(info['base_js'], debug_name='base.js', report_text='Fetched player ' + player_name)
        base_js = base_js.decode('utf-8')
        err = yt_data_extract.extract_decryption_function(info, base_js)
        if err:
            return err
        decrypt_cache[player_name] = info['decryption_function']
        save_decrypt_cache()
    err = yt_data_extract.decrypt_signatures(info)
    return err


def _add_to_error(info, key, additional_message):
    if key in info and info[key]:
        info[key] += additional_message
    else:
        info[key] = additional_message


def fetch_player_response(client, video_id):
    return util.call_youtube_api(client, 'player', {
        'videoId': video_id,
    })


def fetch_watch_page_info(video_id, playlist_id, index):
    # bpctr=9999999999 will bypass are-you-sure dialogs for controversial
    # videos
    url = 'https://m.youtube.com/embed/' + video_id + '?bpctr=9999999999'
    if playlist_id:
        url += '&list=' + playlist_id
    if index:
        url += '&index=' + index

    headers = (
        ('Accept', '*/*'),
        ('Accept-Language', 'en-US,en;q=0.5'),
        ('X-YouTube-Client-Name', '2'),
        ('X-YouTube-Client-Version', '2.20180830'),
    ) + util.mobile_ua

    watch_page = util.fetch_url(url, headers=headers,
                                debug_name='watch')
    watch_page = watch_page.decode('utf-8')
    return yt_data_extract.extract_watch_info_from_html(watch_page)


def extract_info(video_id, use_invidious, playlist_id=None, index=None):
    primary_client = 'android_vr'
    fallback_client = 'ios'
    last_resort_client = 'tv_embedded'

    tasks = (
        # Get video metadata from here
        gevent.spawn(fetch_watch_page_info, video_id, playlist_id, index),
        gevent.spawn(fetch_player_response, primary_client, video_id)
    )
    gevent.joinall(tasks)
    util.check_gevent_exceptions(*tasks)

    info = tasks[0].value or {}
    player_response = tasks[1].value or {}

    # Save android_vr caption tracks (no PO Token needed for these URLs)
    if isinstance(player_response, str):
        try:
            pr_data = json.loads(player_response)
        except Exception:
            pr_data = {}
    else:
        pr_data = player_response or {}
    android_caption_tracks = yt_data_extract.deep_get(
        pr_data, 'captions', 'playerCaptionsTracklistRenderer',
        'captionTracks', default=[])
    info['_android_caption_tracks'] = android_caption_tracks

    # Save streamingData for multi-audio extraction
    pr_streaming_data = pr_data.get('streamingData', {})
    info['_streamingData'] = pr_streaming_data

    yt_data_extract.update_with_new_urls(info, player_response)

    # HLS manifest - try multiple clients in case one is blocked
    info['hls_manifest_url'] = None
    info['hls_audio_tracks'] = {}
    hls_data = None
    hls_client_used = None
    for hls_client in ('ios', 'ios_vr', 'android'):
        try:
            resp = fetch_player_response(hls_client, video_id) or {}
            hls_data = json.loads(resp) if isinstance(resp, str) else resp
            hls_manifest_url = (hls_data.get('streamingData') or {}).get('hlsManifestUrl', '')
            if hls_manifest_url:
                hls_client_used = hls_client
                break
        except Exception as e:
            print(f'HLS fetch with {hls_client} failed: {e}')

    if hls_manifest_url:
        info['hls_manifest_url'] = hls_manifest_url
        import re as _re
        from urllib.parse import urljoin
        hls_manifest = util.fetch_url(hls_manifest_url,
            headers=(('User-Agent', 'Mozilla/5.0'),),
            debug_name='hls_manifest').decode('utf-8')

        # Parse EXT-X-MEDIA audio tracks from HLS manifest
        for line in hls_manifest.split('\n'):
            if '#EXT-X-MEDIA' not in line or 'TYPE=AUDIO' not in line:
                continue
            name_m = _re.search(r'NAME="([^"]+)"', line)
            lang_m = _re.search(r'LANGUAGE="([^"]+)"', line)
            default_m = _re.search(r'DEFAULT=(YES|NO)', line)
            group_m = _re.search(r'GROUP-ID="([^"]+)"', line)
            uri_m = _re.search(r'URI="([^"]+)"', line)
            if not uri_m or not lang_m:
                continue
            lang = lang_m.group(1)
            is_default = default_m and default_m.group(1) == 'YES'
            group = group_m.group(1) if group_m else '0'
            key = lang
            absolute_hls_url = urljoin(hls_manifest_url, uri_m.group(1))
            if key not in info['hls_audio_tracks'] or group > info['hls_audio_tracks'][key].get('group', '0'):
                info['hls_audio_tracks'][key] = {
                    'name': name_m.group(1) if name_m else lang,
                    'lang': lang,
                    'hls_url': absolute_hls_url,
                    'group': group,
                    'is_default': is_default,
                }

        # Register HLS audio tracks for proxy access
        added = 0
        for lang, track in info['hls_audio_tracks'].items():
            ck = video_id + '_' + lang
            from youtube.hls_cache import register_track
            register_track(ck, track['hls_url'],
                           video_id=video_id, track_id=lang)

            fmt = {
                'audio_track_id': lang,
                'audio_track_name': track['name'],
                'audio_track_is_default': track['is_default'],
                'itag': 'hls_' + lang,
                'ext': 'mp4',
                'audio_bitrate': 128,
                'bitrate': 128000,
                'acodec': 'mp4a.40.2',
                'vcodec': None,
                'width': None,
                'height': None,
                'file_size': None,
                'audio_sample_rate': 44100,
                'duration_ms': None,
                'fps': None,
                'init_range': {'start': 0, 'end': 0},
                'index_range': {'start': 0, 'end': 0},
                'url': '/ytl-api/audio-track?id=' + urllib.parse.quote(ck),
                's': None,
                'sp': None,
                'quality': None,
                'type': 'audio/mp4',
                'quality_string': track['name'],
                'mime_codec': 'audio/mp4; codecs="mp4a.40.2"',
                'is_hls': True,
            }
            info['formats'].append(fmt)
            added += 1

        if added:
            print(f"Added {added} HLS audio tracks (via {hls_client_used})")
    else:
        print("No HLS manifest available from any client")
        info['hls_manifest_url'] = None
        info['hls_audio_tracks'] = {}
        info['hls_unavailable'] = True

    # Register HLS manifest for proxying
    if info['hls_manifest_url']:
        ck = video_id + '_video'
        from youtube.hls_cache import register_track
        register_track(ck, info['hls_manifest_url'], video_id=video_id, track_id='video')
        # Use proxy URL instead of direct Google Video URL
        info['hls_manifest_url'] = '/ytl-api/hls-manifest?id=' + urllib.parse.quote(ck)

    # Fallback to 'ios' if no valid URLs are found
    if not info.get('formats') or info.get('player_urls_missing'):
        print(f"No URLs found in '{primary_client}', attempting with '{fallback_client}'.")
        try:
            player_response = fetch_player_response(fallback_client, video_id) or {}
            yt_data_extract.update_with_new_urls(info, player_response)
        except util.FetchError as e:
            print(f"Fallback '{fallback_client}' failed: {e}")

    # Final attempt with 'tv_embedded' if there are still no URLs
    if not info.get('formats') or info.get('player_urls_missing'):
        print(f"No URLs found in '{fallback_client}', attempting with '{last_resort_client}'")
        try:
            player_response = fetch_player_response(last_resort_client, video_id) or {}
            yt_data_extract.update_with_new_urls(info, player_response)
        except util.FetchError as e:
            print(f"Fallback '{last_resort_client}' failed: {e}")

    # signature decryption
    if info.get('formats'):
        decryption_error = decrypt_signatures(info, video_id)
        if decryption_error:
            info['playability_error'] = 'Error decrypting url signatures: ' + decryption_error

    # check if urls ready (non-live format) in former livestream
    # urls not ready if all of them have no filesize
    if info['was_live']:
        info['urls_ready'] = False
        for fmt in info['formats']:
            if fmt['file_size'] is not None:
                info['urls_ready'] = True
    else:
        info['urls_ready'] = True

    # livestream urls
    # sometimes only the livestream urls work soon after the livestream is over
    info['hls_formats'] = []
    if info.get('hls_manifest_url') and (info.get('live') or not info.get('formats') or not info['urls_ready']):
        try:
            manifest = util.fetch_url(info['hls_manifest_url'],
                debug_name='hls_manifest.m3u8',
                report_text='Fetched hls manifest'
            ).decode('utf-8')
            info['hls_formats'], err = yt_data_extract.extract_hls_formats(manifest)
            if not err:
                info['playability_error'] = None
            for fmt in info['hls_formats']:
                fmt['video_quality'] = video_quality_string(fmt)
        except Exception as e:
            print(f"Error obteniendo HLS manifest: {e}")
            info['hls_formats'] = []

    # check for 403. Unnecessary for tor video routing b/c ip address is same
    info['invidious_used'] = False
    info['invidious_reload_button'] = False
    info['tor_bypass_used'] = False
    if (settings.route_tor == 1
            and info['formats'] and info['formats'][0]['url']):
        try:
            response = util.head(info['formats'][0]['url'],
                                 report_text='Checked for URL access')
        except urllib3.exceptions.HTTPError:
            print('Error while checking for URL access:\n')
            traceback.print_exc()
            return info

        if response.status == 403:
            print('Access denied (403) for video urls.')
            print('Routing video through Tor')
            info['tor_bypass_used'] = True
            for fmt in info['formats']:
                fmt['url'] += '&use_tor=1'
        elif 300 <= response.status < 400:
            print('Error: exceeded max redirects while checking video URL')
    return info


def video_quality_string(format):
    if format['vcodec']:
        result = str(format['width'] or '?') + 'x' + str(format['height'] or '?')
        if format['fps']:
            result += ' ' + str(format['fps']) + 'fps'
        return result
    elif format['acodec']:
        return 'audio only'

    return '?'


def short_video_quality_string(fmt):
    result = str(fmt['quality'] or '?') + 'p'
    if fmt['fps']:
        result += str(fmt['fps'])
    if fmt['vcodec'].startswith('av01'):
        result += ' AV1'
    elif fmt['vcodec'].startswith('avc'):
        result += ' h264'
    else:
        result += ' ' + fmt['vcodec']
    return result


def audio_quality_string(fmt):
    if fmt['acodec']:
        if fmt['audio_bitrate']:
            result = '%d' % fmt['audio_bitrate'] + 'k'
        else:
            result = '?k'
        if fmt['audio_sample_rate']:
            result += ' ' + '%.3G' % (fmt['audio_sample_rate']/1000) + 'kHz'
        return result
    elif fmt['vcodec']:
        return 'video only'
    return '?'


# from https://github.com/ytdl-org/youtube-dl/blob/master/youtube_dl/utils.py
def format_bytes(bytes):
    if bytes is None:
        return 'N/A'
    if type(bytes) is str:
        bytes = float(bytes)
    if bytes == 0.0:
        exponent = 0
    else:
        exponent = int(math.log(bytes, 1024.0))
    suffix = ['B', 'KiB', 'MiB', 'GiB', 'TiB', 'PiB', 'EiB', 'ZiB', 'YiB'][exponent]
    converted = float(bytes) / float(1024 ** exponent)
    return '%.2f%s' % (converted, suffix)


@yt_app.route('/ytl-api/audio-track-proxy')
def audio_track_proxy():
    """Proxy for DASH audio tracks to avoid throttling."""
    cache_key = request.args.get('id', '')
    audio_url = request.args.get('url', '')

    if not audio_url:
        flask.abort(400, 'Missing URL')

    try:
        headers = (
            ('User-Agent', 'Mozilla/5.0'),
            ('Accept', '*/*'),
        )
        content = util.fetch_url(audio_url, headers=headers,
            debug_name='audio_dash', report_text=None)
        return flask.Response(content, mimetype='audio/mp4',
            headers={'Access-Control-Allow-Origin': '*',
                     'Cache-Control': 'max-age=3600'})
    except Exception as e:
        flask.abort(502, f'Audio fetch failed: {e}')


@yt_app.route('/ytl-api/audio-track')
def get_audio_track():
    """Proxy HLS audio/video: playlist or individual segment."""
    from youtube.hls_cache import get_hls_url, _tracks

    cache_key = request.args.get('id', '')
    seg_url = request.args.get('seg', '')
    playlist_url = request.args.get('url', '')

    # Handle playlist/manifest URL (used for audio track playlists)
    if playlist_url:
        # Unwrap if double-proxied
        if '/ytl-api/audio-track' in playlist_url:
            import urllib.parse as _up
            parsed = _up.parse_qs(_up.urlparse(playlist_url).query)
            if 'url' in parsed:
                playlist_url = parsed['url'][0]

        try:
            playlist = util.fetch_url(playlist_url,
                headers=(('User-Agent', 'Mozilla/5.0'),),
                debug_name='audio_playlist').decode('utf-8')

            # Rewrite segment URLs
            import re as _re
            from urllib.parse import urljoin
            base_url = request.url_root.rstrip('/')
            playlist_base = playlist_url.rsplit('/', 1)[0] + '/'

            playlist_lines = []
            for line in playlist.split('\n'):
                line = line.strip()
                if not line or line.startswith('#'):
                    playlist_lines.append(line)
                    continue

                # Resolve and proxy segment URL
                seg = line if line.startswith('http') else urljoin(playlist_base, line)
                # Always use &seg= parameter, never &url= for segments
                playlist_lines.append(
                    base_url + '/ytl-api/audio-track?id='
                    + urllib.parse.quote(cache_key)
                    + '&seg=' + urllib.parse.quote(seg, safe='')
                )

            playlist = '\n'.join(playlist_lines)

            return flask.Response(playlist, mimetype='application/vnd.apple.mpegurl',
                headers={'Access-Control-Allow-Origin': '*'})
        except Exception as e:
            import traceback
            traceback.print_exc()
            flask.abort(502, f'Playlist fetch failed: {e}')

    # Handle individual segment or nested playlist
    if seg_url:
        # Check if seg_url is already a proxied URL
        if '/ytl-api/audio-track' in seg_url:
            import urllib.parse as _up
            parsed = _up.parse_qs(_up.urlparse(seg_url).query)
            if 'seg' in parsed:
                seg_url = parsed['seg'][0]
            elif 'url' in parsed:
                seg_url = parsed['url'][0]

        # Check if this is a nested playlist (m3u8) that needs rewriting
        # Playlists END with .m3u8 (optionally followed by query params)
        # Segments may contain /index.m3u8/ in their path but end with .ts or similar
        url_path = urllib.parse.urlparse(seg_url).path

        # Only treat as playlist if path ends with .m3u8
        # Don't use 'in' check because segments can have /index.m3u8/ in their path
        is_playlist = url_path.endswith('.m3u8')

        if is_playlist:
            # This is a variant playlist - fetch and rewrite it
            try:
                raw_content = util.fetch_url(seg_url,
                    headers=(('User-Agent', 'Mozilla/5.0'),),
                    debug_name='nested_playlist')

                # Check if this is actually binary data (segment) misidentified as playlist
                try:
                    playlist = raw_content.decode('utf-8')
                except UnicodeDecodeError:
                    is_playlist = False  # Fall through to segment handler

                if is_playlist:
                    # Rewrite segment URLs in this playlist
                    from urllib.parse import urljoin
                    import re as _re
                    base_url = request.url_root.rstrip('/')
                    playlist_base = seg_url.rsplit('/', 1)[0] + '/'

                    def proxy_url(url):
                        """Rewrite a single URL to go through the proxy"""
                        if not url or url.startswith('/ytl-api/'):
                            return url
                        if not url.startswith('http://') and not url.startswith('https://'):
                            url = urljoin(playlist_base, url)
                        return (base_url + '/ytl-api/audio-track?id='
                                + urllib.parse.quote(cache_key)
                                + '&seg=' + urllib.parse.quote(url, safe=''))

                    playlist_lines = []
                    for line in playlist.split('\n'):
                        line = line.strip()
                        if not line:
                            playlist_lines.append(line)
                            continue

                        # Handle tags with URI attributes (EXT-X-MAP, EXT-X-KEY, etc.)
                        if line.startswith('#') and 'URI=' in line:
                            def rewrite_uri_attr(match):
                                uri = match.group(1)
                                return 'URI="' + proxy_url(uri) + '"'
                            line = _re.sub(r'URI="([^"]+)"', rewrite_uri_attr, line)
                            playlist_lines.append(line)
                        elif line.startswith('#'):
                            # Other tags pass through unchanged
                            playlist_lines.append(line)
                        else:
                            # This is a segment URL line
                            seg = line if line.startswith('http') else urljoin(playlist_base, line)
                            playlist_lines.append(proxy_url(seg))

                    playlist = '\n'.join(playlist_lines)

                    return flask.Response(playlist, mimetype='application/vnd.apple.mpegurl',
                        headers={'Access-Control-Allow-Origin': '*'})
            except Exception as e:
                import traceback
                traceback.print_exc()
                flask.abort(502, f'Nested playlist fetch failed: {e}')

        # This is an actual segment - fetch and serve it
        try:
            headers = (
                ('User-Agent', 'Mozilla/5.0'),
                ('Accept', '*/*'),
            )
            content = util.fetch_url(seg_url, headers=headers,
                debug_name='hls_seg', report_text=None)

            # Determine content type based on URL or content
            # HLS segments are usually MPEG-TS (.ts) but can be MP4 (.mp4, .m4s)
            if '.mp4' in seg_url or '.m4s' in seg_url or seg_url.lower().endswith('.mp4'):
                content_type = 'video/mp4'
            elif '.webm' in seg_url or seg_url.lower().endswith('.webm'):
                content_type = 'video/webm'
            else:
                # Default to MPEG-TS for HLS
                content_type = 'video/mp2t'

            return flask.Response(content, mimetype=content_type,
                headers={
                    'Access-Control-Allow-Origin': '*',
                    'Access-Control-Allow-Methods': 'GET, OPTIONS',
                    'Access-Control-Allow-Headers': 'Range, Content-Type',
                    'Cache-Control': 'max-age=3600',
                    'Content-Type': content_type,
                })
        except Exception as e:
            import traceback
            traceback.print_exc()
            flask.abort(502, f'Segment fetch failed: {e}')

    # Legacy: Proxy the HLS playlist for audio tracks (using get_hls_url)
    hls_url = get_hls_url(cache_key)
    if not hls_url:
        flask.abort(404, 'Audio track not found')

    try:
        playlist = util.fetch_url(hls_url,
            headers=(('User-Agent', 'Mozilla/5.0'),),
            debug_name='audio_hls_playlist').decode('utf-8')

        # Rewrite segment URLs to go through our proxy endpoint
        import re as _re
        from urllib.parse import urljoin
        hls_base_url = hls_url.rsplit('/', 1)[0] + '/'

        def make_proxy_url(segment_url):
            if segment_url.startswith('/ytl-api/audio-track'):
                return segment_url
            base_url = request.url_root.rstrip('/')
            return (base_url + '/ytl-api/audio-track?id='
                    + urllib.parse.quote(cache_key)
                    + '&seg=' + urllib.parse.quote(segment_url))

        playlist_lines = []
        for line in playlist.split('\n'):
            line = line.strip()
            if not line or line.startswith('#'):
                playlist_lines.append(line)
                continue

            if line.startswith('http://') or line.startswith('https://'):
                segment_url = line
            else:
                segment_url = urljoin(hls_base_url, line)

            playlist_lines.append(make_proxy_url(segment_url))

        playlist = '\n'.join(playlist_lines)

        return flask.Response(playlist, mimetype='application/vnd.apple.mpegurl',
            headers={'Access-Control-Allow-Origin': '*'})
    except Exception as e:
        flask.abort(502, f'Playlist fetch failed: {e}')


@yt_app.route('/ytl-api/hls-manifest')
def get_hls_manifest():
    """Proxy HLS video manifest, rewriting ALL URLs including audio tracks."""
    from youtube.hls_cache import get_hls_url

    cache_key = request.args.get('id', '')
    is_audio = '_audio_' in cache_key or cache_key.endswith('_audio')
    print(f'[hls-manifest] Request: id={cache_key[:40] if cache_key else ""}... (audio={is_audio})')

    hls_url = get_hls_url(cache_key)
    print(f'[hls-manifest] HLS URL: {hls_url[:80] if hls_url else None}...')
    if not hls_url:
        flask.abort(404, 'HLS manifest not found')

    try:
        print(f'[hls-manifest] Fetching HLS manifest...')
        manifest = util.fetch_url(hls_url,
            headers=(('User-Agent', 'Mozilla/5.0'),),
            debug_name='hls_manifest').decode('utf-8')
        print(f'[hls-manifest] Successfully fetched manifest ({len(manifest)} bytes)')

        # Rewrite all URLs in the manifest to go through our proxy
        import re as _re
        from urllib.parse import urljoin

        # Get the base URL for resolving relative URLs
        hls_base_url = hls_url.rsplit('/', 1)[0] + '/'
        base_url = request.url_root.rstrip('/')

        # Rewrite URLs - handle both segment URLs and audio track URIs
        def rewrite_url(url, is_audio_track=False):
            if not url or url.startswith('/ytl-api/'):
                return url

            # Resolve relative URLs
            if not url.startswith('http://') and not url.startswith('https://'):
                url = urljoin(hls_base_url, url)

            if is_audio_track:
                # Audio track playlist - proxy through audio-track endpoint
                return (base_url + '/ytl-api/audio-track?id='
                        + urllib.parse.quote(cache_key)
                        + '&url=' + urllib.parse.quote(url, safe=''))
            else:
                # Video segment or variant playlist - proxy through audio-track endpoint
                return (base_url + '/ytl-api/audio-track?id='
                        + urllib.parse.quote(cache_key)
                        + '&seg=' + urllib.parse.quote(url, safe=''))

        # Parse and rewrite the manifest
        manifest_lines = []
        rewritten_count = 0
        for line in manifest.split('\n'):
            line = line.strip()
            if not line:
                manifest_lines.append(line)
                continue

            # Handle EXT-X-MEDIA tags with URI (audio tracks)
            if line.startswith('#EXT-X-MEDIA:') and 'URI=' in line:
                # Extract and rewrite the URI attribute
                def rewrite_media_uri(match):
                    nonlocal rewritten_count
                    uri = match.group(1)
                    rewritten_count += 1
                    return 'URI="' + rewrite_url(uri, is_audio_track=True) + '"'
                line = _re.sub(r'URI="([^"]+)"', rewrite_media_uri, line)
                manifest_lines.append(line)
            elif line.startswith('#'):
                # Other tags pass through
                manifest_lines.append(line)
            else:
                # This is a URL (segment or variant playlist)
                if line.startswith('http://') or line.startswith('https://'):
                    url = line
                else:
                    url = urljoin(hls_base_url, line)
                rewritten_count += 1
                manifest_lines.append(rewrite_url(url))

        manifest = '\n'.join(manifest_lines)
        print(f'[hls-manifest] Rewrote manifest with {len(manifest_lines)} lines, {rewritten_count} URLs rewritten')

        return flask.Response(manifest, mimetype='application/vnd.apple.mpegurl',
            headers={
                'Access-Control-Allow-Origin': '*',
                'Access-Control-Allow-Methods': 'GET, OPTIONS',
                'Access-Control-Allow-Headers': 'Range, Content-Type',
                'Cache-Control': 'no-cache',
                'Content-Type': 'application/vnd.apple.mpegurl',
            })
    except Exception as e:
        print(f'[hls-manifest] Error: {e}')
        import traceback
        traceback.print_exc()
        flask.abort(502, f'Manifest fetch failed: {e}')


@yt_app.route('/ytl-api/storyboard.vtt')
def get_storyboard_vtt():
    """
    See:
        https://github.com/iv-org/invidious/blob/9a8b81fcbe49ff8d88f197b7f731d6bf79fc8087/src/invidious.cr#L3603
        https://github.com/iv-org/invidious/blob/3bb7fbb2f119790ee6675076b31cd990f75f64bb/src/invidious/videos.cr#L623
    """

    spec_url = request.args.get('spec_url')
    url, *boards = spec_url.split('|')
    base_url, q = url.split('?')
    q = parse_qs(q)  # for url query

    storyboard = None
    wanted_height = 90

    for i, board in enumerate(boards):
        *t, _, sigh = board.split("#")
        width, height, count, width_cnt, height_cnt, interval = map(int, t)
        if height != wanted_height: continue
        q['sigh'] = [sigh]
        url = f"{base_url}?{urlencode(q, doseq=True)}"
        storyboard = SimpleNamespace(
            url               = url.replace("$L", str(i)).replace("$N", "M$M"),
            width             = width,
            height            = height,
            interval          = interval,
            width_cnt         = width_cnt,
            height_cnt        = height_cnt,
            storyboard_count  = ceil(count / (width_cnt * height_cnt))
        )

    if not storyboard:
        flask.abort(404)

    def to_ts(ms):
        s, ms = divmod(ms, 1000)
        h, s = divmod(s, 3600)
        m, s = divmod(s, 60)
        return f"{h:02}:{m:02}:{s:02}.{ms:03}"

    r = "WEBVTT"  # result
    ts = 0  # current timestamp

    for i in range(storyboard.storyboard_count):
        url = '/' + storyboard.url.replace("$M", str(i))
        interval = storyboard.interval
        w, h = storyboard.width, storyboard.height
        w_cnt, h_cnt = storyboard.width_cnt, storyboard.height_cnt

        for j in range(h_cnt):
            for k in range(w_cnt):
                r += f"{to_ts(ts)} --> {to_ts(ts+interval)}\n"
                r += f"{url}#xywh={w * k},{h * j},{w},{h}\n\n"
                ts += interval

    return flask.Response(r, mimetype='text/vtt')


time_table = {'h': 3600, 'm': 60, 's': 1}
@yt_app.route('/watch')
@yt_app.route('/embed')
@yt_app.route('/embed/<video_id>')
@yt_app.route('/shorts')
@yt_app.route('/shorts/<video_id>')
def get_watch_page(video_id=None):
    video_id = request.args.get('v') or video_id
    if not video_id:
        return flask.render_template('error.html', error_message='Missing video id'), 404
    if len(video_id) < 11:
        return flask.render_template('error.html', error_message='Incomplete video id (too short): ' + video_id), 404

    time_start_str = request.args.get('t', '0s')
    time_start = 0
    if re.fullmatch(r'(\d+(h|m|s))+', time_start_str):
        for match in re.finditer(r'(\d+)(h|m|s)', time_start_str):
            time_start += int(match.group(1))*time_table[match.group(2)]
    elif re.fullmatch(r'\d+', time_start_str):
        time_start = int(time_start_str)

    lc = request.args.get('lc', '')
    playlist_id = request.args.get('list')
    index = request.args.get('index')
    use_invidious = bool(int(request.args.get('use_invidious', '1')))
    if request.path.startswith('/embed') and settings.embed_page_mode:
        tasks = (
            gevent.spawn((lambda: {})),
            gevent.spawn(extract_info, video_id, use_invidious,
                         playlist_id=playlist_id, index=index),
        )
    else:
        tasks = (
            gevent.spawn(comments.video_comments, video_id,
                         int(settings.default_comment_sorting), lc=lc),
            gevent.spawn(extract_info, video_id, use_invidious,
                         playlist_id=playlist_id, index=index),
        )
    gevent.joinall(tasks)
    util.check_gevent_exceptions(tasks[1])
    comments_info, info = tasks[0].value, tasks[1].value

    if info['error']:
        return flask.render_template('error.html', error_message=info['error'])

    video_info = {
        'duration':  util.seconds_to_timestamp(info['duration'] or 0),
        'id':        info['id'],
        'title':     info['title'],
        'author':    info['author'],
        'author_id': info['author_id'],
    }

    # prefix urls, and other post-processing not handled by yt_data_extract
    for item in info['related_videos']:
        # Only set thumbnail if YouTube didn't provide one
        if not item.get('thumbnail'):
            if item.get('type') == 'playlist' and item.get('first_video_id'):
                item['thumbnail'] = "https://i.ytimg.com/vi/{}/hqdefault.jpg".format(item['first_video_id'])
            elif item.get('type') == 'video' and item.get('id'):
                item['thumbnail'] = "https://i.ytimg.com/vi/{}/hqdefault.jpg".format(item['id'])
        util.prefix_urls(item)
        util.add_extra_html_info(item)
    for song in info['music_list']:
        song['url'] = util.prefix_url(song['url'])
    if info['playlist']:
        playlist_id = info['playlist']['id']
        for item in info['playlist']['items']:
            # Only set thumbnail if YouTube didn't provide one
            if not item.get('thumbnail') and item.get('type') == 'video' and item.get('id'):
                item['thumbnail'] = "https://i.ytimg.com/vi/{}/hqdefault.jpg".format(item['id'])
            util.prefix_urls(item)
            util.add_extra_html_info(item)
            if playlist_id:
                item['url'] += '&list=' + playlist_id
            if item['index']:
                item['url'] += '&index=' + str(item['index'])
        info['playlist']['author_url'] = util.prefix_url(
            info['playlist']['author_url'])
    if settings.img_prefix:
        # Don't prefix hls_formats for now because the urls inside the manifest
        # would need to be prefixed as well.
        for fmt in info['formats']:
            fmt['url'] = util.prefix_url(fmt['url'])

    # Add video title to end of url path so it has a filename other than just
    # "videoplayback" when downloaded
    title = urllib.parse.quote(util.to_valid_filename(info['title'] or ''))
    for fmt in info['formats']:
        filename = title
        ext = fmt.get('ext')
        if ext:
            filename += '.' + ext
        fmt['url'] = fmt['url'].replace(
            '/videoplayback',
            '/videoplayback/name/' + filename)

    download_formats = []

    for format in (info['formats'] + info['hls_formats']):
        if format['acodec'] and format['vcodec']:
            codecs_string = format['acodec'] + ', ' + format['vcodec']
        else:
            codecs_string = format['acodec'] or format['vcodec'] or '?'
        download_formats.append({
            'url': format['url'],
            'ext': format['ext'] or '?',
            'audio_quality': audio_quality_string(format),
            'video_quality': video_quality_string(format),
            'file_size': format_bytes(format['file_size']),
            'codecs': codecs_string,
        })

    if (settings.route_tor == 2) or info['tor_bypass_used']:
        target_resolution = 240
    else:
        res = settings.default_resolution
        target_resolution = 1080 if res == 'auto' else int(res)

    # Get video sources for no-JS fallback and DASH (av-merge) fallback
    video_sources = get_video_sources(info, target_resolution)
    uni_sources = video_sources['uni_sources']
    pair_sources = video_sources['pair_sources']
    pair_idx = video_sources['pair_idx']
    audio_track_sources = video_sources['audio_track_sources']

    # Build audio tracks list from HLS
    audio_tracks = []
    hls_audio_tracks = info.get('hls_audio_tracks', {})
    hls_manifest_url = info.get('hls_manifest_url')
    if hls_audio_tracks:
        # Prefer "original" audio track
        original_lang = None
        for lang, track in hls_audio_tracks.items():
            if 'original' in (track.get('name') or '').lower():
                original_lang = lang
                break

        # Add tracks, preferring original as default
        for lang, track in hls_audio_tracks.items():
            is_default = (lang == original_lang) if original_lang else track['is_default']
            if is_default:
                audio_tracks.insert(0, {
                    'id': lang,
                    'name': track['name'],
                    'is_default': True,
                })
            else:
                audio_tracks.append({
                    'id': lang,
                    'name': track['name'],
                    'is_default': False,
                })
    else:
        # Fallback: single default audio track
        audio_tracks = [{'id': 'default', 'name': 'Default', 'is_default': True}]

    # Get video dimensions
    video_height = info.get('height') or 360
    video_width = info.get('width') or 640



    # 1 second per pixel, or the actual video width
    theater_video_target_width = max(640, info['duration'] or 0, video_width)

    # Check for false determination of disabled comments, which comes from
    # the watch page. But if we got comments in the separate request for those,
    # then the determination is wrong.
    if info['comments_disabled'] and comments_info.get('comments'):
        info['comments_disabled'] = False
        print('Warning: False determination that comments are disabled')
        print('Comment count:', info['comment_count'])
        info['comment_count'] = None # hack to make it obvious there's a bug

    # captions and transcript
    subtitle_sources = get_subtitle_sources(info)
    other_downloads = []
    for source in subtitle_sources:
        best_caption_parse = urllib.parse.urlparse(
            source['url'].lstrip('/'))
        transcript_url = (util.URL_ORIGIN
            + '/watch/transcript'
            + best_caption_parse.path
            + '?' + best_caption_parse.query)
        other_downloads.append({
            'label': 'Video Transcript: ' + source['label'],
            'ext': 'txt',
            'url': transcript_url
        })

    if request.path.startswith('/embed') and settings.embed_page_mode:
        template_name = 'embed.html'
    else:
        template_name = 'watch.html'
    return flask.render_template(template_name,
        header_playlist_names   = local_playlist.get_playlist_names(),
        uploader_channel_url    = ('/' + info['author_url']) if info['author_url'] else '',
        time_published             = info['time_published'],
        view_count    = (lambda x: '{:,}'.format(x) if x is not None else "")(info.get("view_count", None)),
        like_count    = (lambda x: '{:,}'.format(x) if x is not None else "")(info.get("like_count", None)),
        dislike_count = (lambda x: '{:,}'.format(x) if x is not None else "")(info.get("dislike_count", None)),
        download_formats        = download_formats,
        other_downloads         = other_downloads,
        video_info              = json.dumps(video_info),
        hls_formats             = info['hls_formats'],
        hls_manifest_url        = hls_manifest_url,
        audio_tracks            = audio_tracks,
        subtitle_sources        = subtitle_sources,
        uni_sources             = uni_sources,
        pair_sources            = pair_sources,
        pair_idx                = pair_idx,
        hls_unavailable         = info.get('hls_unavailable', False),
        playback_mode           = settings.playback_mode,
        related                 = info['related_videos'],
        playlist                = info['playlist'],
        music_list              = info['music_list'],
        music_attributes        = get_ordered_music_list_attributes(info['music_list']),
        comments_info           = comments_info,
        comment_count           = info['comment_count'],
        comments_disabled       = info['comments_disabled'],

        video_height            = video_height,
        video_width             = video_width,
        theater_video_target_width = theater_video_target_width,

        title       = info['title'],
        uploader    = info['author'],
        description = info['description'],
        unlisted    = info['unlisted'],
        limited_state = info['limited_state'],
        age_restricted    = info['age_restricted'],
        live              = info['live'],
        playability_error = info['playability_error'],

        allowed_countries = info['allowed_countries'],
        ip_address   = info['ip_address'] if settings.route_tor else None,
        invidious_used    = info['invidious_used'],
        invidious_reload_button = info['invidious_reload_button'],
        video_url = util.URL_ORIGIN + '/watch?v=' + video_id,
        video_id = video_id,
        storyboard_url = (util.URL_ORIGIN + '/ytl-api/storyboard.vtt?' +
            urlencode([('spec_url', info['storyboard_spec_url'])])
            if info['storyboard_spec_url'] else None),

        js_data = {
            'video_id': info['id'],
            'video_duration': info['duration'],
            'settings': settings.current_settings_dict,
            'has_manual_captions': any(s.get('on') for s in subtitle_sources),
            'audio_tracks': audio_tracks,
            'hls_manifest_url': hls_manifest_url,
            'time_start': time_start,
            'playlist': info['playlist'],
            'related': info['related_videos'],
            'playability_error': info['playability_error'],
            'hls_unavailable': info.get('hls_unavailable', False),
            'pair_sources': pair_sources,
            'pair_idx': pair_idx,
            'uni_sources': uni_sources,
            'uni_idx': video_sources['uni_idx'],
            'using_pair_sources': bool(pair_sources),
        },
        font_family = youtube.font_choices[settings.font], # for embed page
    )


@yt_app.route('/api/<path:dummy>')
def get_captions(dummy):
    url = 'https://www.youtube.com' + request.full_path
    try:
        result = util.fetch_url(url, headers=util.mobile_ua)
        result = result.replace(b"align:start position:0%", b"")
        return flask.Response(result, mimetype='text/vtt')
    except Exception as e:
        logger.debug(f'Caption fetch failed: {e}')
        return flask.Response(b'WEBVTT\n\n', mimetype='text/vtt', status=200)


times_reg = re.compile(r'^\d\d:\d\d:\d\d\.\d\d\d --> \d\d:\d\d:\d\d\.\d\d\d.*$')
inner_timestamp_removal_reg = re.compile(r'<[^>]+>')
@yt_app.route('/watch/transcript/<path:caption_path>')
def get_transcript(caption_path):
    try:
        captions = util.fetch_url('https://www.youtube.com/'
            + caption_path
            + '?' + request.environ['QUERY_STRING']).decode('utf-8')
    except util.FetchError as e:
        msg = ('Error retrieving captions: ' + str(e) + '\n\n'
            + 'The caption url may have expired.')
        print(msg)
        return flask.Response(
            msg,
            status=e.code,
            mimetype='text/plain;charset=UTF-8')

    lines = captions.splitlines()
    segments = []

    # skip captions file header
    i = 0
    while lines[i] != '':
        i += 1

    current_segment = None
    while i < len(lines):
        line = lines[i]
        if line == '':
            if ((current_segment is not None)
                    and (current_segment['begin'] is not None)):
                segments.append(current_segment)
            current_segment = {
                'begin': None,
                'end': None,
                'lines': [],
            }
        elif times_reg.fullmatch(line.rstrip()):
            current_segment['begin'], current_segment['end'] = line.split(' --> ')
        else:
            current_segment['lines'].append(
                inner_timestamp_removal_reg.sub('', line))
        i += 1

    # if automatic captions, but not translated
    if request.args.get('kind') == 'asr' and not request.args.get('tlang'):
        # Automatic captions repeat content. The new segment is displayed
        # on the bottom row; the old one is displayed on the top row.
        # So grab the bottom row only
        for seg in segments:
            seg['text'] = seg['lines'][1]
    else:
        for seg in segments:
            seg['text'] = ' '.join(map(str.rstrip, seg['lines']))

    result = ''
    for seg in segments:
        if seg['text'] != ' ':
            result += seg['begin'] + ' ' + seg['text'] + '\r\n'

    return flask.Response(result.encode('utf-8'),
                          mimetype='text/plain;charset=UTF-8')