yt_dlp/extractor/whowatch.py

   1 # coding: utf-8
   2 from __future__ import unicode_literals
   3
   4 from .common import InfoExtractor
   5 from ..utils import (
   6     int_or_none,
   7     qualities,
   8     try_call,
   9     try_get,
  10     ExtractorError,
  11 )
  12 from ..compat import compat_str
  13
  14
  15 class WhoWatchIE(InfoExtractor):
  16     IE_NAME = 'whowatch'
  17     _VALID_URL = r'https?://whowatch\.tv/viewer/(?P<id>\d+)'
  18
  19     _TESTS = [{
  20         'url': 'https://whowatch.tv/viewer/21450171',
  21         'only_matching': True,
  22     }]
  23
  24     def _real_extract(self, url):
  25         video_id = self._match_id(url)
  26         self._download_webpage(url, video_id)
  27         metadata = self._download_json('https://api.whowatch.tv/lives/%s' % video_id, video_id)
  28         live_data = self._download_json('https://api.whowatch.tv/lives/%s/play' % video_id, video_id)
  29
  30         title = try_call(
  31             lambda: live_data['share_info']['live_title'][1:-1],
  32             lambda: metadata['live']['title'],
  33             expected_type=str)
  34
  35         hls_url = live_data.get('hls_url')
  36         if not hls_url:
  37             raise ExtractorError(live_data.get('error_message') or 'The user is offline.', expected=True)
  38
  39         QUALITIES = qualities(['low', 'medium', 'high', 'veryhigh'])
  40         formats = []
  41
  42         for i, fmt in enumerate(live_data.get('streams') or []):
  43             name = fmt.get('quality') or fmt.get('name') or compat_str(i)
  44             hls_url = fmt.get('hls_url')
  45             rtmp_url = fmt.get('rtmp_url')
  46             audio_only = fmt.get('audio_only')
  47             quality = QUALITIES(fmt.get('quality'))
  48
  49             if hls_url:
  50                 hls_fmts = self._extract_m3u8_formats(
  51                     hls_url, video_id, ext='mp4', m3u8_id='hls-%s' % name, quality=quality)
  52                 formats.extend(hls_fmts)
  53             else:
  54                 hls_fmts = []
  55
  56             # RTMP url for audio_only is same as high format, so skip it
  57             if rtmp_url and not audio_only:
  58                 formats.append({
  59                     'url': rtmp_url,
  60                     'format_id': 'rtmp-%s' % name,
  61                     'ext': 'mp4',
  62                     'protocol': 'rtmp_ffmpeg',  # ffmpeg can, while rtmpdump can't
  63                     'vcodec': 'h264',
  64                     'acodec': 'aac',
  65                     'quality': quality,
  66                     'format_note': fmt.get('label'),
  67                     # note: HLS and RTMP have same resolution for now, so it's acceptable
  68                     'width': try_get(hls_fmts, lambda x: x[0]['width'], int),
  69                     'height': try_get(hls_fmts, lambda x: x[0]['height'], int),
  70                 })
  71
  72         # This contains the same formats as the above manifests and is used only as a fallback
  73         formats.extend(self._extract_m3u8_formats(
  74             hls_url, video_id, ext='mp4', m3u8_id='hls'))
  75         self._remove_duplicate_formats(formats)
  76         self._sort_formats(formats)
  77
  78         uploader_url = try_get(metadata, lambda x: x['live']['user']['user_path'], compat_str)
  79         if uploader_url:
  80             uploader_url = 'https://whowatch.tv/profile/%s' % uploader_url
  81         uploader_id = compat_str(try_get(metadata, lambda x: x['live']['user']['id'], int))
  82         uploader = try_get(metadata, lambda x: x['live']['user']['name'], compat_str)
  83         thumbnail = try_get(metadata, lambda x: x['live']['latest_thumbnail_url'], compat_str)
  84         timestamp = int_or_none(try_get(metadata, lambda x: x['live']['started_at'], int), scale=1000)
  85         view_count = try_get(metadata, lambda x: x['live']['total_view_count'], int)
  86         comment_count = try_get(metadata, lambda x: x['live']['comment_count'], int)
  87
  88         return {
  89             'id': video_id,
  90             'title': title,
  91             'uploader_id': uploader_id,
  92             'uploader_url': uploader_url,
  93             'uploader': uploader,
  94             'formats': formats,
  95             'thumbnail': thumbnail,
  96             'timestamp': timestamp,
  97             'view_count': view_count,
  98             'comment_count': comment_count,
  99             'is_live': True,
 100         }