yt_dlp/extractor/whowatch.py

   1 from .common import InfoExtractor
   2 from ..utils import (
   3     ExtractorError,
   4     int_or_none,
   5     qualities,
   6     try_call,
   7     try_get,
   8 )
   9
  10
  11 class WhoWatchIE(InfoExtractor):
  12     IE_NAME = 'whowatch'
  13     _VALID_URL = r'https?://whowatch\.tv/viewer/(?P<id>\d+)'
  14
  15     _TESTS = [{
  16         'url': 'https://whowatch.tv/viewer/21450171',
  17         'only_matching': True,
  18     }]
  19
  20     def _real_extract(self, url):
  21         video_id = self._match_id(url)
  22         self._download_webpage(url, video_id)
  23         metadata = self._download_json(f'https://api.whowatch.tv/lives/{video_id}', video_id)
  24         live_data = self._download_json(f'https://api.whowatch.tv/lives/{video_id}/play', video_id)
  25
  26         title = try_call(
  27             lambda: live_data['share_info']['live_title'][1:-1],
  28             lambda: metadata['live']['title'],
  29             expected_type=str)
  30
  31         hls_url = live_data.get('hls_url')
  32         if not hls_url:
  33             raise ExtractorError(live_data.get('error_message') or 'The user is offline.', expected=True)
  34
  35         QUALITIES = qualities(['low', 'medium', 'high', 'veryhigh'])
  36         formats = []
  37
  38         for i, fmt in enumerate(live_data.get('streams') or []):
  39             name = fmt.get('quality') or fmt.get('name') or str(i)
  40             hls_url = fmt.get('hls_url')
  41             rtmp_url = fmt.get('rtmp_url')
  42             audio_only = fmt.get('audio_only')
  43             quality = QUALITIES(fmt.get('quality'))
  44
  45             if hls_url:
  46                 hls_fmts = self._extract_m3u8_formats(
  47                     hls_url, video_id, ext='mp4', m3u8_id=f'hls-{name}', quality=quality)
  48                 formats.extend(hls_fmts)
  49             else:
  50                 hls_fmts = []
  51
  52             # RTMP url for audio_only is same as high format, so skip it
  53             if rtmp_url and not audio_only:
  54                 formats.append({
  55                     'url': rtmp_url,
  56                     'format_id': f'rtmp-{name}',
  57                     'ext': 'mp4',
  58                     'protocol': 'rtmp_ffmpeg',  # ffmpeg can, while rtmpdump can't
  59                     'vcodec': 'h264',
  60                     'acodec': 'aac',
  61                     'quality': quality,
  62                     'format_note': fmt.get('label'),
  63                     # note: HLS and RTMP have same resolution for now, so it's acceptable
  64                     'width': try_get(hls_fmts, lambda x: x[0]['width'], int),
  65                     'height': try_get(hls_fmts, lambda x: x[0]['height'], int),
  66                 })
  67
  68         # This contains the same formats as the above manifests and is used only as a fallback
  69         formats.extend(self._extract_m3u8_formats(
  70             hls_url, video_id, ext='mp4', m3u8_id='hls'))
  71         self._remove_duplicate_formats(formats)
  72
  73         uploader_url = try_get(metadata, lambda x: x['live']['user']['user_path'], str)
  74         if uploader_url:
  75             uploader_url = f'https://whowatch.tv/profile/{uploader_url}'
  76         uploader_id = str(try_get(metadata, lambda x: x['live']['user']['id'], int))
  77         uploader = try_get(metadata, lambda x: x['live']['user']['name'], str)
  78         thumbnail = try_get(metadata, lambda x: x['live']['latest_thumbnail_url'], str)
  79         timestamp = int_or_none(try_get(metadata, lambda x: x['live']['started_at'], int), scale=1000)
  80         view_count = try_get(metadata, lambda x: x['live']['total_view_count'], int)
  81         comment_count = try_get(metadata, lambda x: x['live']['comment_count'], int)
  82
  83         return {
  84             'id': video_id,
  85             'title': title,
  86             'uploader_id': uploader_id,
  87             'uploader_url': uploader_url,
  88             'uploader': uploader,
  89             'formats': formats,
  90             'thumbnail': thumbnail,
  91             'timestamp': timestamp,
  92             'view_count': view_count,
  93             'comment_count': comment_count,
  94             'is_live': True,
  95         }