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     parse_duration,
  13     xpath_attr,
  14     update_url_query,
  15     ExtractorError,
  16 )
  17
  18
  19 class TurnerBaseIE(AdobePassIE):
  20     def _extract_timestamp(self, video_data):
  21         return int_or_none(xpath_attr(video_data, 'dateCreated', 'uts'))
  22
  23     def _extract_cvp_info(self, data_src, video_id, path_data={}, ap_data={}):
  24         video_data = self._download_xml(data_src, video_id)
  25         video_id = video_data.attrib['id']
  26         title = xpath_text(video_data, 'headline', fatal=True)
  27         content_id = xpath_text(video_data, 'contentId') or video_id
  28         # rtmp_src = xpath_text(video_data, 'akamai/src')
  29         # if rtmp_src:
  30         #     splited_rtmp_src = rtmp_src.split(',')
  31         #     if len(splited_rtmp_src) == 2:
  32         #         rtmp_src = splited_rtmp_src[1]
  33         # aifp = xpath_text(video_data, 'akamai/aifp', default='')
  34
  35         tokens = {}
  36         urls = []
  37         formats = []
  38         rex = re.compile(
  39             r'(?P<width>[0-9]+)x(?P<height>[0-9]+)(?:_(?P<bitrate>[0-9]+))?')
  40         # Possible formats locations: files/file, files/groupFiles/files
  41         # and maybe others
  42         for video_file in video_data.findall('.//file'):
  43             video_url = video_file.text.strip()
  44             if not video_url:
  45                 continue
  46             ext = determine_ext(video_url)
  47             if video_url.startswith('/mp4:protected/'):
  48                 continue
  49                 # TODO Correct extraction for these files
  50                 # protected_path_data = path_data.get('protected')
  51                 # if not protected_path_data or not rtmp_src:
  52                 #     continue
  53                 # protected_path = self._search_regex(
  54                 #     r'/mp4:(.+)\.[a-z0-9]', video_url, 'secure path')
  55                 # auth = self._download_webpage(
  56                 #     protected_path_data['tokenizer_src'], query={
  57                 #         'path': protected_path,
  58                 #         'videoId': content_id,
  59                 #         'aifp': aifp,
  60                 #     })
  61                 # token = xpath_text(auth, 'token')
  62                 # if not token:
  63                 #     continue
  64                 # video_url = rtmp_src + video_url + '?' + token
  65             elif video_url.startswith('/secure/'):
  66                 secure_path_data = path_data.get('secure')
  67                 if not secure_path_data:
  68                     continue
  69                 video_url = secure_path_data['media_src'] + video_url
  70                 secure_path = self._search_regex(r'https?://[^/]+(.+/)', video_url, 'secure path') + '*'
  71                 token = tokens.get(secure_path)
  72                 if not token:
  73                     query = {
  74                         'path': secure_path,
  75                         'videoId': content_id,
  76                     }
  77                     if ap_data.get('auth_required'):
  78                         query['accessToken'] = self._extract_mvpd_auth(ap_data['url'], video_id, ap_data['site_name'], ap_data['site_name'])
  79                     auth = self._download_xml(
  80                         secure_path_data['tokenizer_src'], video_id, query=query)
  81                     error_msg = xpath_text(auth, 'error/msg')
  82                     if error_msg:
  83                         raise ExtractorError(error_msg, expected=True)
  84                     token = xpath_text(auth, 'token')
  85                     if not token:
  86                         continue
  87                     tokens[secure_path] = token
  88                 video_url = video_url + '?hdnea=' + token
  89             elif not re.match('https?://', video_url):
  90                 base_path_data = path_data.get(ext, path_data.get('default', {}))
  91                 media_src = base_path_data.get('media_src')
  92                 if not media_src:
  93                     continue
  94                 video_url = media_src + video_url
  95             if video_url in urls:
  96                 continue
  97             urls.append(video_url)
  98             format_id = video_file.get('bitrate')
  99             if ext == 'smil':
 100                 formats.extend(self._extract_smil_formats(
 101                     video_url, video_id, fatal=False))
 102             elif ext == 'm3u8':
 103                 formats.extend(self._extract_m3u8_formats(
 104                     video_url, video_id, 'mp4',
 105                     m3u8_id=format_id or 'hls', fatal=False))
 106             elif ext == 'f4m':
 107                 formats.extend(self._extract_f4m_formats(
 108                     update_url_query(video_url, {'hdcore': '3.7.0'}),
 109                     video_id, f4m_id=format_id or 'hds', fatal=False))
 110             else:
 111                 f = {
 112                     'format_id': format_id,
 113                     'url': video_url,
 114                     'ext': ext,
 115                 }
 116                 mobj = rex.search(format_id + video_url)
 117                 if mobj:
 118                     f.update({
 119                         'width': int(mobj.group('width')),
 120                         'height': int(mobj.group('height')),
 121                         'tbr': int_or_none(mobj.group('bitrate')),
 122                     })
 123                 elif isinstance(format_id, compat_str):
 124                     if format_id.isdigit():
 125                         f['tbr'] = int(format_id)
 126                     else:
 127                         mobj = re.match(r'ios_(audio|[0-9]+)$', format_id)
 128                         if mobj:
 129                             if mobj.group(1) == 'audio':
 130                                 f.update({
 131                                     'vcodec': 'none',
 132                                     'ext': 'm4a',
 133                                 })
 134                             else:
 135                                 f['tbr'] = int(mobj.group(1))
 136                 formats.append(f)
 137         self._sort_formats(formats)
 138
 139         subtitles = {}
 140         for source in video_data.findall('closedCaptions/source'):
 141             for track in source.findall('track'):
 142                 track_url = track.get('url')
 143                 if not isinstance(track_url, compat_str) or track_url.endswith('/big'):
 144                     continue
 145                 lang = track.get('lang') or track.get('label') or 'en'
 146                 subtitles.setdefault(lang, []).append({
 147                     'url': track_url,
 148                     'ext': {
 149                         'scc': 'scc',
 150                         'webvtt': 'vtt',
 151                         'smptett': 'tt',
 152                     }.get(source.get('format'))
 153                 })
 154
 155         thumbnails = [{
 156             'id': image.get('cut'),
 157             'url': image.text,
 158             'width': int_or_none(image.get('width')),
 159             'height': int_or_none(image.get('height')),
 160         } for image in video_data.findall('images/image')]
 161
 162         return {
 163             'id': video_id,
 164             'title': title,
 165             'formats': formats,
 166             'subtitles': subtitles,
 167             'thumbnails': thumbnails,
 168             'description': xpath_text(video_data, 'description'),
 169             'duration': parse_duration(xpath_text(video_data, 'length') or xpath_text(video_data, 'trt')),
 170             'timestamp': self._extract_timestamp(video_data),
 171             'upload_date': xpath_attr(video_data, 'metas', 'version'),
 172             'series': xpath_text(video_data, 'showTitle'),
 173             'season_number': int_or_none(xpath_text(video_data, 'seasonNumber')),
 174             'episode_number': int_or_none(xpath_text(video_data, 'episodeNumber')),
 175         }