yt_dlp/extractor/redge.py

   1
   2 from .common import InfoExtractor
   3 from ..networking import HEADRequest
   4 from ..utils import (
   5     float_or_none,
   6     int_or_none,
   7     join_nonempty,
   8     parse_qs,
   9     update_url_query,
  10 )
  11 from ..utils.traversal import traverse_obj
  12
  13
  14 class RedCDNLivxIE(InfoExtractor):
  15     _VALID_URL = r'https?://[^.]+\.(?:dcs\.redcdn|atmcdn)\.pl/(?:live(?:dash|hls|ss)|nvr)/o2/(?P<tenant>[^/?#]+)/(?P<id>[^?#]+)\.livx'
  16     IE_NAME = 'redcdnlivx'
  17
  18     _TESTS = [{
  19         'url': 'https://r.dcs.redcdn.pl/livedash/o2/senat/ENC02/channel.livx?indexMode=true&startTime=638272860000&stopTime=638292544000',
  20         'info_dict': {
  21             'id': 'ENC02-638272860000-638292544000',
  22             'ext': 'mp4',
  23             'title': 'ENC02',
  24             'duration': 19683.982,
  25             'live_status': 'was_live',
  26         },
  27     }, {
  28         'url': 'https://r.dcs.redcdn.pl/livedash/o2/sejm/ENC18/live.livx?indexMode=true&startTime=722333096000&stopTime=722335562000',
  29         'info_dict': {
  30             'id': 'ENC18-722333096000-722335562000',
  31             'ext': 'mp4',
  32             'title': 'ENC18',
  33             'duration': 2463.995,
  34             'live_status': 'was_live',
  35         },
  36     }, {
  37         'url': 'https://r.dcs.redcdn.pl/livehls/o2/sportevolution/live/triathlon2018/warsaw.livx/playlist.m3u8?startTime=550305000000&stopTime=550327620000',
  38         'info_dict': {
  39             'id': 'triathlon2018-warsaw-550305000000-550327620000',
  40             'ext': 'mp4',
  41             'title': 'triathlon2018/warsaw',
  42             'duration': 22619.98,
  43             'live_status': 'was_live',
  44         },
  45     }, {
  46         'url': 'https://n-25-12.dcs.redcdn.pl/nvr/o2/sejm/Migacz-ENC01/1.livx?startTime=722347200000&stopTime=722367345000',
  47         'only_matching': True,
  48     }, {
  49         'url': 'https://redir.atmcdn.pl/nvr/o2/sejm/ENC08/1.livx?startTime=503831270000&stopTime=503840040000',
  50         'only_matching': True,
  51     }]
  52
  53     '''
  54     Known methods (first in url path):
  55     - `livedash` - DASH MPD
  56     - `livehls` - HTTP Live Streaming
  57     - `livess` - IIS Smooth Streaming
  58     - `nvr` - CCTV mode, directly returns a file, typically flv, avc1, aac
  59     - `sc` - shoutcast/icecast (audio streams, like radio)
  60     '''
  61
  62     def _real_extract(self, url):
  63         tenant, path = self._match_valid_url(url).group('tenant', 'id')
  64         qs = parse_qs(url)
  65         start_time = traverse_obj(qs, ('startTime', 0, {int_or_none}))
  66         stop_time = traverse_obj(qs, ('stopTime', 0, {int_or_none}))
  67
  68         def livx_mode(mode):
  69             suffix = ''
  70             if mode == 'livess':
  71                 suffix = '/manifest'
  72             elif mode == 'livehls':
  73                 suffix = '/playlist.m3u8'
  74             file_qs = {}
  75             if start_time:
  76                 file_qs['startTime'] = start_time
  77             if stop_time:
  78                 file_qs['stopTime'] = stop_time
  79             if mode == 'nvr':
  80                 file_qs['nolimit'] = 1
  81             elif mode != 'sc':
  82                 file_qs['indexMode'] = 'true'
  83             return update_url_query(f'https://r.dcs.redcdn.pl/{mode}/o2/{tenant}/{path}.livx{suffix}', file_qs)
  84
  85         # no id or title for a transmission. making ones up.
  86         title = path \
  87             .replace('/live', '').replace('live/', '') \
  88             .replace('/channel', '').replace('channel/', '') \
  89             .strip('/')
  90         video_id = join_nonempty(title.replace('/', '-'), start_time, stop_time)
  91
  92         formats = []
  93         # downloading the manifest separately here instead of _extract_ism_formats to also get some stream metadata
  94         ism_res = self._download_xml_handle(
  95             livx_mode('livess'), video_id,
  96             note='Downloading ISM manifest',
  97             errnote='Failed to download ISM manifest',
  98             fatal=False)
  99         ism_doc = None
 100         if ism_res is not False:
 101             ism_doc, ism_urlh = ism_res
 102             formats, _ = self._parse_ism_formats_and_subtitles(ism_doc, ism_urlh.url, 'ss')
 103
 104         nvr_urlh = self._request_webpage(
 105             HEADRequest(livx_mode('nvr')), video_id, 'Follow flv file redirect', fatal=False,
 106             expected_status=lambda _: True)
 107         if nvr_urlh and nvr_urlh.status == 200:
 108             formats.append({
 109                 'url': nvr_urlh.url,
 110                 'ext': 'flv',
 111                 'format_id': 'direct-0',
 112                 'preference': -1,   # might be slow
 113             })
 114         formats.extend(self._extract_mpd_formats(livx_mode('livedash'), video_id, mpd_id='dash', fatal=False))
 115         formats.extend(self._extract_m3u8_formats(
 116             livx_mode('livehls'), video_id, m3u8_id='hls', ext='mp4', fatal=False))
 117
 118         time_scale = traverse_obj(ism_doc, ('@TimeScale', {int_or_none})) or 10000000
 119         duration = traverse_obj(
 120             ism_doc, ('@Duration', {float_or_none(scale=time_scale)})) or None
 121
 122         live_status = None
 123         if traverse_obj(ism_doc, '@IsLive') == 'TRUE':
 124             live_status = 'is_live'
 125         elif duration:
 126             live_status = 'was_live'
 127
 128         return {
 129             'id': video_id,
 130             'title': title,
 131             'formats': formats,
 132             'duration': duration,
 133             'live_status': live_status,
 134         }