yt_dlp/extractor/umg.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     parse_filesize,
   8     parse_iso8601,
   9 )
  10
  11
  12 class UMGDeIE(InfoExtractor):
  13     IE_NAME = 'umg:de'
  14     IE_DESC = 'Universal Music Deutschland'
  15     _VALID_URL = r'https?://(?:www\.)?universal-music\.de/[^/]+/videos/[^/?#]+-(?P<id>\d+)'
  16     _TEST = {
  17         'url': 'https://www.universal-music.de/sido/videos/jedes-wort-ist-gold-wert-457803',
  18         'md5': 'ebd90f48c80dcc82f77251eb1902634f',
  19         'info_dict': {
  20             'id': '457803',
  21             'ext': 'mp4',
  22             'title': 'Jedes Wort ist Gold wert',
  23             'timestamp': 1513591800,
  24             'upload_date': '20171218',
  25         }
  26     }
  27
  28     def _real_extract(self, url):
  29         video_id = self._match_id(url)
  30         video_data = self._download_json(
  31             'https://graphql.universal-music.de/',
  32             video_id, query={
  33                 'query': '''{
  34   universalMusic(channel:16) {
  35     video(id:%s) {
  36       headline
  37       formats {
  38         formatId
  39         url
  40         type
  41         width
  42         height
  43         mimeType
  44         fileSize
  45       }
  46       duration
  47       createdDate
  48     }
  49   }
  50 }''' % video_id})['data']['universalMusic']['video']
  51
  52         title = video_data['headline']
  53         hls_url_template = 'http://mediadelivery.universal-music-services.de/vod/mp4:autofill/storage/' + '/'.join(list(video_id)) + '/content/%s/file/playlist.m3u8'
  54
  55         thumbnails = []
  56         formats = []
  57
  58         def add_m3u8_format(format_id):
  59             formats.extend(self._extract_m3u8_formats(
  60                 hls_url_template % format_id, video_id, 'mp4',
  61                 'm3u8_native', m3u8_id='hls', fatal=False))
  62
  63         for f in video_data.get('formats', []):
  64             f_url = f.get('url')
  65             mime_type = f.get('mimeType')
  66             if not f_url or mime_type == 'application/mxf':
  67                 continue
  68             fmt = {
  69                 'url': f_url,
  70                 'width': int_or_none(f.get('width')),
  71                 'height': int_or_none(f.get('height')),
  72                 'filesize': parse_filesize(f.get('fileSize')),
  73             }
  74             f_type = f.get('type')
  75             if f_type == 'Image':
  76                 thumbnails.append(fmt)
  77             elif f_type == 'Video':
  78                 format_id = f.get('formatId')
  79                 if format_id:
  80                     fmt['format_id'] = format_id
  81                     if mime_type == 'video/mp4':
  82                         add_m3u8_format(format_id)
  83                 urlh = self._request_webpage(f_url, video_id, fatal=False)
  84                 if urlh:
  85                     first_byte = urlh.read(1)
  86                     if first_byte not in (b'F', b'\x00'):
  87                         continue
  88                     formats.append(fmt)
  89         if not formats:
  90             for format_id in (867, 836, 940):
  91                 add_m3u8_format(format_id)
  92         self._sort_formats(formats)
  93
  94         return {
  95             'id': video_id,
  96             'title': title,
  97             'duration': int_or_none(video_data.get('duration')),
  98             'timestamp': parse_iso8601(video_data.get('createdDate'), ' '),
  99             'thumbnails': thumbnails,
 100             'formats': formats,
 101         }