yt_dlp/extractor/umg.py

   1 from .common import InfoExtractor
   2 from ..utils import (
   3     int_or_none,
   4     parse_filesize,
   5     parse_iso8601,
   6 )
   7
   8
   9 class UMGDeIE(InfoExtractor):
  10     IE_NAME = 'umg:de'
  11     IE_DESC = 'Universal Music Deutschland'
  12     _VALID_URL = r'https?://(?:www\.)?universal-music\.de/[^/]+/videos/[^/?#]+-(?P<id>\d+)'
  13     _TEST = {
  14         'url': 'https://www.universal-music.de/sido/videos/jedes-wort-ist-gold-wert-457803',
  15         'md5': 'ebd90f48c80dcc82f77251eb1902634f',
  16         'info_dict': {
  17             'id': '457803',
  18             'ext': 'mp4',
  19             'title': 'Jedes Wort ist Gold wert',
  20             'timestamp': 1513591800,
  21             'upload_date': '20171218',
  22         }
  23     }
  24
  25     def _real_extract(self, url):
  26         video_id = self._match_id(url)
  27         video_data = self._download_json(
  28             'https://graphql.universal-music.de/',
  29             video_id, query={
  30                 'query': '''{
  31   universalMusic(channel:16) {
  32     video(id:%s) {
  33       headline
  34       formats {
  35         formatId
  36         url
  37         type
  38         width
  39         height
  40         mimeType
  41         fileSize
  42       }
  43       duration
  44       createdDate
  45     }
  46   }
  47 }''' % video_id})['data']['universalMusic']['video']
  48
  49         title = video_data['headline']
  50         hls_url_template = 'http://mediadelivery.universal-music-services.de/vod/mp4:autofill/storage/' + '/'.join(list(video_id)) + '/content/%s/file/playlist.m3u8'
  51
  52         thumbnails = []
  53         formats = []
  54
  55         def add_m3u8_format(format_id):
  56             formats.extend(self._extract_m3u8_formats(
  57                 hls_url_template % format_id, video_id, 'mp4',
  58                 'm3u8_native', m3u8_id='hls', fatal=False))
  59
  60         for f in video_data.get('formats', []):
  61             f_url = f.get('url')
  62             mime_type = f.get('mimeType')
  63             if not f_url or mime_type == 'application/mxf':
  64                 continue
  65             fmt = {
  66                 'url': f_url,
  67                 'width': int_or_none(f.get('width')),
  68                 'height': int_or_none(f.get('height')),
  69                 'filesize': parse_filesize(f.get('fileSize')),
  70             }
  71             f_type = f.get('type')
  72             if f_type == 'Image':
  73                 thumbnails.append(fmt)
  74             elif f_type == 'Video':
  75                 format_id = f.get('formatId')
  76                 if format_id:
  77                     fmt['format_id'] = format_id
  78                     if mime_type == 'video/mp4':
  79                         add_m3u8_format(format_id)
  80                 urlh = self._request_webpage(f_url, video_id, fatal=False)
  81                 if urlh:
  82                     first_byte = urlh.read(1)
  83                     if first_byte not in (b'F', b'\x00'):
  84                         continue
  85                     formats.append(fmt)
  86         if not formats:
  87             for format_id in (867, 836, 940):
  88                 add_m3u8_format(format_id)
  89
  90         return {
  91             'id': video_id,
  92             'title': title,
  93             'duration': int_or_none(video_data.get('duration')),
  94             'timestamp': parse_iso8601(video_data.get('createdDate'), ' '),
  95             'thumbnails': thumbnails,
  96             'formats': formats,
  97         }