Raphaël G. Git Repositories - youtubedl/blob - youtube_dl/extractor/turner.py

   1 # coding: utf-8
   2 from __future__ import unicode_literals
   3
   4 import re
   5
   6 from .adobepass import AdobePassIE
   7 from ..compat import compat_str
   8 from ..utils import (
   9     xpath_text,
  10     int_or_none,
  11     determine_ext,
  12     float_or_none,
  13     parse_duration,
  14     xpath_attr,
  15     update_url_query,
  16     ExtractorError,
  17     strip_or_none,
  18 )
  19
  20
  21 class TurnerBaseIE(AdobePassIE):
  22     _AKAMAI_SPE_TOKEN_CACHE = {}
  23
  24     def _extract_timestamp(self, video_data):
  25         return int_or_none(xpath_attr(video_data, 'dateCreated', 'uts'))
  26
  27     def _add_akamai_spe_token(self, tokenizer_src, video_url, content_id, ap_data, custom_tokenizer_query=None):
  28         secure_path = self._search_regex(r'https?://[^/]+(.+/)', video_url, 'secure path') + '*'
  29         token = self._AKAMAI_SPE_TOKEN_CACHE.get(secure_path)
  30         if not token:
  31             query = {
  32                 'path': secure_path,
  33             }
  34             if custom_tokenizer_query:
  35                 query.update(custom_tokenizer_query)
  36             else:
  37                 query['videoId'] = content_id
  38             if ap_data.get('auth_required'):
  39                 query['accessToken'] = self._extract_mvpd_auth(ap_data['url'], content_id, ap_data['site_name'], ap_data['site_name'])
  40             auth = self._download_xml(
  41                 tokenizer_src, content_id, query=query)
  42             error_msg = xpath_text(auth, 'error/msg')
  43             if error_msg:
  44                 raise ExtractorError(error_msg, expected=True)
  45             token = xpath_text(auth, 'token')
  46             if not token:
  47                 return video_url
  48             self._AKAMAI_SPE_TOKEN_CACHE[secure_path] = token
  49         return video_url + '?hdnea=' + token
  50
  51     def _extract_cvp_info(self, data_src, video_id, path_data={}, ap_data={}):
  52         video_data = self._download_xml(data_src, video_id)
  53         video_id = video_data.attrib['id']
  54         title = xpath_text(video_data, 'headline', fatal=True)
  55         content_id = xpath_text(video_data, 'contentId') or video_id
  56         # rtmp_src = xpath_text(video_data, 'akamai/src')
  57         # if rtmp_src:
  58         #     splited_rtmp_src = rtmp_src.split(',')
  59         #     if len(splited_rtmp_src) == 2:
  60         #         rtmp_src = splited_rtmp_src[1]
  61         # aifp = xpath_text(video_data, 'akamai/aifp', default='')
  62
  63         urls = []
  64         formats = []
  65         rex = re.compile(
  66             r'(?P<width>[0-9]+)x(?P<height>[0-9]+)(?:_(?P<bitrate>[0-9]+))?')
  67         # Possible formats locations: files/file, files/groupFiles/files
  68         # and maybe others
  69         for video_file in video_data.findall('.//file'):
  70             video_url = video_file.text.strip()
  71             if not video_url:
  72                 continue
  73             ext = determine_ext(video_url)
  74             if video_url.startswith('/mp4:protected/'):
  75                 continue
  76                 # TODO Correct extraction for these files
  77                 # protected_path_data = path_data.get('protected')
  78                 # if not protected_path_data or not rtmp_src:
  79                 #     continue
  80                 # protected_path = self._search_regex(
  81                 #     r'/mp4:(.+)\.[a-z0-9]', video_url, 'secure path')
  82                 # auth = self._download_webpage(
  83                 #     protected_path_data['tokenizer_src'], query={
  84                 #         'path': protected_path,
  85                 #         'videoId': content_id,
  86                 #         'aifp': aifp,
  87                 #     })
  88                 # token = xpath_text(auth, 'token')
  89                 # if not token:
  90                 #     continue
  91                 # video_url = rtmp_src + video_url + '?' + token
  92             elif video_url.startswith('/secure/'):
  93                 secure_path_data = path_data.get('secure')
  94                 if not secure_path_data:
  95                     continue
  96                 video_url = self._add_akamai_spe_token(
  97                     secure_path_data['tokenizer_src'],
  98                     secure_path_data['media_src'] + video_url,
  99                     content_id, ap_data)
 100             elif not re.match('https?://', video_url):
 101                 base_path_data = path_data.get(ext, path_data.get('default', {}))
 102                 media_src = base_path_data.get('media_src')
 103                 if not media_src:
 104                     continue
 105                 video_url = media_src + video_url
 106             if video_url in urls:
 107                 continue
 108             urls.append(video_url)
 109             format_id = video_file.get('bitrate')
 110             if ext == 'smil':
 111                 formats.extend(self._extract_smil_formats(
 112                     video_url, video_id, fatal=False))
 113             elif ext == 'm3u8':
 114                 m3u8_formats = self._extract_m3u8_formats(
 115                     video_url, video_id, 'mp4',
 116                     m3u8_id=format_id or 'hls', fatal=False)
 117                 if '/secure/' in video_url and '?hdnea=' in video_url:
 118                     for f in m3u8_formats:
 119                         f['_seekable'] = False
 120                 formats.extend(m3u8_formats)
 121             elif ext == 'f4m':
 122                 formats.extend(self._extract_f4m_formats(
 123                     update_url_query(video_url, {'hdcore': '3.7.0'}),
 124                     video_id, f4m_id=format_id or 'hds', fatal=False))
 125             else:
 126                 f = {
 127                     'format_id': format_id,
 128                     'url': video_url,
 129                     'ext': ext,
 130                 }
 131                 mobj = rex.search(format_id + video_url)
 132                 if mobj:
 133                     f.update({
 134                         'width': int(mobj.group('width')),
 135                         'height': int(mobj.group('height')),
 136                         'tbr': int_or_none(mobj.group('bitrate')),
 137                     })
 138                 elif isinstance(format_id, compat_str):
 139                     if format_id.isdigit():
 140                         f['tbr'] = int(format_id)
 141                     else:
 142                         mobj = re.match(r'ios_(audio|[0-9]+)$', format_id)
 143                         if mobj:
 144                             if mobj.group(1) == 'audio':
 145                                 f.update({
 146                                     'vcodec': 'none',
 147                                     'ext': 'm4a',
 148                                 })
 149                             else:
 150                                 f['tbr'] = int(mobj.group(1))
 151                 formats.append(f)
 152         self._sort_formats(formats)
 153
 154         subtitles = {}
 155         for source in video_data.findall('closedCaptions/source'):
 156             for track in source.findall('track'):
 157                 track_url = track.get('url')
 158                 if not isinstance(track_url, compat_str) or track_url.endswith('/big'):
 159                     continue
 160                 lang = track.get('lang') or track.get('label') or 'en'
 161                 subtitles.setdefault(lang, []).append({
 162                     'url': track_url,
 163                     'ext': {
 164                         'scc': 'scc',
 165                         'webvtt': 'vtt',
 166                         'smptett': 'tt',
 167                     }.get(source.get('format'))
 168                 })
 169
 170         thumbnails = [{
 171             'id': image.get('cut'),
 172             'url': image.text,
 173             'width': int_or_none(image.get('width')),
 174             'height': int_or_none(image.get('height')),
 175         } for image in video_data.findall('images/image')]
 176
 177         is_live = xpath_text(video_data, 'isLive') == 'true'
 178
 179         return {
 180             'id': video_id,
 181             'title': self._live_title(title) if is_live else title,
 182             'formats': formats,
 183             'subtitles': subtitles,
 184             'thumbnails': thumbnails,
 185             'thumbnail': xpath_text(video_data, 'poster'),
 186             'description': strip_or_none(xpath_text(video_data, 'description')),
 187             'duration': parse_duration(xpath_text(video_data, 'length') or xpath_text(video_data, 'trt')),
 188             'timestamp': self._extract_timestamp(video_data),
 189             'upload_date': xpath_attr(video_data, 'metas', 'version'),
 190             'series': xpath_text(video_data, 'showTitle'),
 191             'season_number': int_or_none(xpath_text(video_data, 'seasonNumber')),
 192             'episode_number': int_or_none(xpath_text(video_data, 'episodeNumber')),
 193             'is_live': is_live,
 194         }
 195
 196     def _extract_ngtv_info(self, media_id, tokenizer_query, ap_data=None):
 197         streams_data = self._download_json(
 198             'http://medium.ngtv.io/media/%s/tv' % media_id,
 199             media_id)['media']['tv']
 200         duration = None
 201         chapters = []
 202         formats = []
 203         for supported_type in ('unprotected', 'bulkaes'):
 204             stream_data = streams_data.get(supported_type, {})
 205             m3u8_url = stream_data.get('secureUrl') or stream_data.get('url')
 206             if not m3u8_url:
 207                 continue
 208             if stream_data.get('playlistProtection') == 'spe':
 209                 m3u8_url = self._add_akamai_spe_token(
 210                     'http://token.ngtv.io/token/token_spe',
 211                     m3u8_url, media_id, ap_data or {}, tokenizer_query)
 212             formats.extend(self._extract_m3u8_formats(
 213                 m3u8_url, media_id, 'mp4', m3u8_id='hls', fatal=False))
 214
 215             duration = float_or_none(stream_data.get('totalRuntime'))
 216
 217             if not chapters:
 218                 for chapter in stream_data.get('contentSegments', []):
 219                     start_time = float_or_none(chapter.get('start'))
 220                     chapter_duration = float_or_none(chapter.get('duration'))
 221                     if start_time is None or chapter_duration is None:
 222                         continue
 223                     chapters.append({
 224                         'start_time': start_time,
 225                         'end_time': start_time + chapter_duration,
 226                     })
 227         self._sort_formats(formats)
 228
 229         return {
 230             'formats': formats,
 231             'chapters': chapters,
 232             'duration': duration,
 233         }