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     url_or_none,
  19 )
  20
  21
  22 class TurnerBaseIE(AdobePassIE):
  23     _AKAMAI_SPE_TOKEN_CACHE = {}
  24
  25     def _extract_timestamp(self, video_data):
  26         return int_or_none(xpath_attr(video_data, 'dateCreated', 'uts'))
  27
  28     def _add_akamai_spe_token(self, tokenizer_src, video_url, content_id, ap_data, custom_tokenizer_query=None):
  29         secure_path = self._search_regex(r'https?://[^/]+(.+/)', video_url, 'secure path') + '*'
  30         token = self._AKAMAI_SPE_TOKEN_CACHE.get(secure_path)
  31         if not token:
  32             query = {
  33                 'path': secure_path,
  34             }
  35             if custom_tokenizer_query:
  36                 query.update(custom_tokenizer_query)
  37             else:
  38                 query['videoId'] = content_id
  39             if ap_data.get('auth_required'):
  40                 query['accessToken'] = self._extract_mvpd_auth(ap_data['url'], content_id, ap_data['site_name'], ap_data['site_name'])
  41             auth = self._download_xml(
  42                 tokenizer_src, content_id, query=query)
  43             error_msg = xpath_text(auth, 'error/msg')
  44             if error_msg:
  45                 raise ExtractorError(error_msg, expected=True)
  46             token = xpath_text(auth, 'token')
  47             if not token:
  48                 return video_url
  49             self._AKAMAI_SPE_TOKEN_CACHE[secure_path] = token
  50         return video_url + '?hdnea=' + token
  51
  52     def _extract_cvp_info(self, data_src, video_id, path_data={}, ap_data={}):
  53         video_data = self._download_xml(data_src, video_id)
  54         video_id = video_data.attrib['id']
  55         title = xpath_text(video_data, 'headline', fatal=True)
  56         content_id = xpath_text(video_data, 'contentId') or video_id
  57         # rtmp_src = xpath_text(video_data, 'akamai/src')
  58         # if rtmp_src:
  59         #     splited_rtmp_src = rtmp_src.split(',')
  60         #     if len(splited_rtmp_src) == 2:
  61         #         rtmp_src = splited_rtmp_src[1]
  62         # aifp = xpath_text(video_data, 'akamai/aifp', default='')
  63
  64         urls = []
  65         formats = []
  66         rex = re.compile(
  67             r'(?P<width>[0-9]+)x(?P<height>[0-9]+)(?:_(?P<bitrate>[0-9]+))?')
  68         # Possible formats locations: files/file, files/groupFiles/files
  69         # and maybe others
  70         for video_file in video_data.findall('.//file'):
  71             video_url = video_file.text.strip()
  72             if not video_url:
  73                 continue
  74             ext = determine_ext(video_url)
  75             if video_url.startswith('/mp4:protected/'):
  76                 continue
  77                 # TODO Correct extraction for these files
  78                 # protected_path_data = path_data.get('protected')
  79                 # if not protected_path_data or not rtmp_src:
  80                 #     continue
  81                 # protected_path = self._search_regex(
  82                 #     r'/mp4:(.+)\.[a-z0-9]', video_url, 'secure path')
  83                 # auth = self._download_webpage(
  84                 #     protected_path_data['tokenizer_src'], query={
  85                 #         'path': protected_path,
  86                 #         'videoId': content_id,
  87                 #         'aifp': aifp,
  88                 #     })
  89                 # token = xpath_text(auth, 'token')
  90                 # if not token:
  91                 #     continue
  92                 # video_url = rtmp_src + video_url + '?' + token
  93             elif video_url.startswith('/secure/'):
  94                 secure_path_data = path_data.get('secure')
  95                 if not secure_path_data:
  96                     continue
  97                 video_url = self._add_akamai_spe_token(
  98                     secure_path_data['tokenizer_src'],
  99                     secure_path_data['media_src'] + video_url,
 100                     content_id, ap_data)
 101             elif not re.match('https?://', video_url):
 102                 base_path_data = path_data.get(ext, path_data.get('default', {}))
 103                 media_src = base_path_data.get('media_src')
 104                 if not media_src:
 105                     continue
 106                 video_url = media_src + video_url
 107             if video_url in urls:
 108                 continue
 109             urls.append(video_url)
 110             format_id = video_file.get('bitrate')
 111             if ext == 'smil':
 112                 formats.extend(self._extract_smil_formats(
 113                     video_url, video_id, fatal=False))
 114             elif ext == 'm3u8':
 115                 m3u8_formats = self._extract_m3u8_formats(
 116                     video_url, video_id, 'mp4',
 117                     m3u8_id=format_id or 'hls', fatal=False)
 118                 if '/secure/' in video_url and '?hdnea=' in video_url:
 119                     for f in m3u8_formats:
 120                         f['_seekable'] = False
 121                 formats.extend(m3u8_formats)
 122             elif ext == 'f4m':
 123                 formats.extend(self._extract_f4m_formats(
 124                     update_url_query(video_url, {'hdcore': '3.7.0'}),
 125                     video_id, f4m_id=format_id or 'hds', fatal=False))
 126             else:
 127                 f = {
 128                     'format_id': format_id,
 129                     'url': video_url,
 130                     'ext': ext,
 131                 }
 132                 mobj = rex.search(format_id + video_url)
 133                 if mobj:
 134                     f.update({
 135                         'width': int(mobj.group('width')),
 136                         'height': int(mobj.group('height')),
 137                         'tbr': int_or_none(mobj.group('bitrate')),
 138                     })
 139                 elif isinstance(format_id, compat_str):
 140                     if format_id.isdigit():
 141                         f['tbr'] = int(format_id)
 142                     else:
 143                         mobj = re.match(r'ios_(audio|[0-9]+)$', format_id)
 144                         if mobj:
 145                             if mobj.group(1) == 'audio':
 146                                 f.update({
 147                                     'vcodec': 'none',
 148                                     'ext': 'm4a',
 149                                 })
 150                             else:
 151                                 f['tbr'] = int(mobj.group(1))
 152                 formats.append(f)
 153         self._sort_formats(formats)
 154
 155         subtitles = {}
 156         for source in video_data.findall('closedCaptions/source'):
 157             for track in source.findall('track'):
 158                 track_url = url_or_none(track.get('url'))
 159                 if not track_url or track_url.endswith('/big'):
 160                     continue
 161                 lang = track.get('lang') or track.get('label') or 'en'
 162                 subtitles.setdefault(lang, []).append({
 163                     'url': track_url,
 164                     'ext': {
 165                         'scc': 'scc',
 166                         'webvtt': 'vtt',
 167                         'smptett': 'tt',
 168                     }.get(source.get('format'))
 169                 })
 170
 171         thumbnails = [{
 172             'id': image.get('cut'),
 173             'url': image.text,
 174             'width': int_or_none(image.get('width')),
 175             'height': int_or_none(image.get('height')),
 176         } for image in video_data.findall('images/image')]
 177
 178         is_live = xpath_text(video_data, 'isLive') == 'true'
 179
 180         return {
 181             'id': video_id,
 182             'title': self._live_title(title) if is_live else title,
 183             'formats': formats,
 184             'subtitles': subtitles,
 185             'thumbnails': thumbnails,
 186             'thumbnail': xpath_text(video_data, 'poster'),
 187             'description': strip_or_none(xpath_text(video_data, 'description')),
 188             'duration': parse_duration(xpath_text(video_data, 'length') or xpath_text(video_data, 'trt')),
 189             'timestamp': self._extract_timestamp(video_data),
 190             'upload_date': xpath_attr(video_data, 'metas', 'version'),
 191             'series': xpath_text(video_data, 'showTitle'),
 192             'season_number': int_or_none(xpath_text(video_data, 'seasonNumber')),
 193             'episode_number': int_or_none(xpath_text(video_data, 'episodeNumber')),
 194             'is_live': is_live,
 195         }
 196
 197     def _extract_ngtv_info(self, media_id, tokenizer_query, ap_data=None):
 198         streams_data = self._download_json(
 199             'http://medium.ngtv.io/media/%s/tv' % media_id,
 200             media_id)['media']['tv']
 201         duration = None
 202         chapters = []
 203         formats = []
 204         for supported_type in ('unprotected', 'bulkaes'):
 205             stream_data = streams_data.get(supported_type, {})
 206             m3u8_url = stream_data.get('secureUrl') or stream_data.get('url')
 207             if not m3u8_url:
 208                 continue
 209             if stream_data.get('playlistProtection') == 'spe':
 210                 m3u8_url = self._add_akamai_spe_token(
 211                     'http://token.ngtv.io/token/token_spe',
 212                     m3u8_url, media_id, ap_data or {}, tokenizer_query)
 213             formats.extend(self._extract_m3u8_formats(
 214                 m3u8_url, media_id, 'mp4', m3u8_id='hls', fatal=False))
 215
 216             duration = float_or_none(stream_data.get('totalRuntime'))
 217
 218             if not chapters:
 219                 for chapter in stream_data.get('contentSegments', []):
 220                     start_time = float_or_none(chapter.get('start'))
 221                     chapter_duration = float_or_none(chapter.get('duration'))
 222                     if start_time is None or chapter_duration is None:
 223                         continue
 224                     chapters.append({
 225                         'start_time': start_time,
 226                         'end_time': start_time + chapter_duration,
 227                     })
 228         self._sort_formats(formats)
 229
 230         return {
 231             'formats': formats,
 232             'chapters': chapters,
 233             'duration': duration,
 234         }