Raphaël G. Git Repositories - youtubedl/blob - youtube_dl/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://api.universal-music.de/graphql',
  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             m3u8_formats = self._extract_m3u8_formats(
  60                 hls_url_template % format_id, video_id, 'mp4',
  61                 'm3u8_native', m3u8_id='hls', fatal='False')
  62             if m3u8_formats and m3u8_formats[0].get('height'):
  63                 formats.extend(m3u8_formats)
  64
  65         for f in video_data.get('formats', []):
  66             f_url = f.get('url')
  67             mime_type = f.get('mimeType')
  68             if not f_url or mime_type == 'application/mxf':
  69                 continue
  70             fmt = {
  71                 'url': f_url,
  72                 'width': int_or_none(f.get('width')),
  73                 'height': int_or_none(f.get('height')),
  74                 'filesize': parse_filesize(f.get('fileSize')),
  75             }
  76             f_type = f.get('type')
  77             if f_type == 'Image':
  78                 thumbnails.append(fmt)
  79             elif f_type == 'Video':
  80                 format_id = f.get('formatId')
  81                 if format_id:
  82                     fmt['format_id'] = format_id
  83                     if mime_type == 'video/mp4':
  84                         add_m3u8_format(format_id)
  85                 urlh = self._request_webpage(f_url, video_id, fatal=False)
  86                 if urlh:
  87                     first_byte = urlh.read(1)
  88                     if first_byte not in (b'F', b'\x00'):
  89                         continue
  90                     formats.append(fmt)
  91         if not formats:
  92             for format_id in (867, 836, 940):
  93                 add_m3u8_format(format_id)
  94         self._sort_formats(formats, ('width', 'height', 'filesize', 'tbr'))
  95
  96         return {
  97             'id': video_id,
  98             'title': title,
  99             'duration': int_or_none(video_data.get('duration')),
 100             'timestamp': parse_iso8601(video_data.get('createdDate'), ' '),
 101             'thumbnails': thumbnails,
 102             'formats': formats,
 103         }