_ Git - youtube-dl/blob - youtube_dl/extractor/turner.py

   1 # coding: utf-8
   2 from __future__ import unicode_literals
   3
   4 import re
   5
   6 from .common import InfoExtractor
   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     compat_urlparse,
  16 )
  17
  18
  19 class TurnerBaseIE(InfoExtractor):
  20     def _extract_cvp_info(self, data_src, video_id, path_data={}):
  21         video_data = self._download_xml(data_src, video_id)
  22         video_id = video_data.attrib['id'].split('/')[-1].split('.')[0]
  23         title = xpath_text(video_data, 'headline', fatal=True)
  24         # rtmp_src = xpath_text(video_data, 'akamai/src')
  25         # if rtmp_src:
  26         #     splited_rtmp_src = rtmp_src.split(',')
  27         #     if len(splited_rtmp_src) == 2:
  28         #         rtmp_src = splited_rtmp_src[1]
  29         # aifp = xpath_text(video_data, 'akamai/aifp', default='')
  30
  31         tokens = {}
  32         urls = []
  33         formats = []
  34         rex = re.compile(
  35             r'(?P<width>[0-9]+)x(?P<height>[0-9]+)(?:_(?P<bitrate>[0-9]+))?')
  36         # Possible formats locations: files/file, files/groupFiles/files
  37         # and maybe others
  38         for video_file in video_data.findall('.//file'):
  39             video_url = video_file.text.strip()
  40             if not video_url:
  41                 continue
  42             ext = determine_ext(video_url)
  43             if video_url.startswith('/mp4:protected/'):
  44                 continue
  45                 # TODO Correct extraction for these files
  46                 # protected_path_data = path_data.get('protected')
  47                 # if not protected_path_data or not rtmp_src:
  48                 #     continue
  49                 # protected_path = self._search_regex(
  50                 #     r'/mp4:(.+)\.[a-z0-9]', video_url, 'secure path')
  51                 # auth = self._download_webpage(
  52                 #     protected_path_data['tokenizer_src'], query={
  53                 #         'path': protected_path,
  54                 #         'videoId': video_id,
  55                 #         'aifp': aifp,
  56                 #     })
  57                 # token = xpath_text(auth, 'token')
  58                 # if not token:
  59                 #     continue
  60                 # video_url = rtmp_src + video_url + '?' + token
  61             elif video_url.startswith('/secure/'):
  62                 secure_path_data = path_data.get('secure')
  63                 if not secure_path_data:
  64                     continue
  65                 video_url = secure_path_data['media_src'] + video_url
  66                 secure_path = self._search_regex(r'https?://[^/]+(.+/)', video_url, 'secure path') + '*'
  67                 token = tokens.get(secure_path)
  68                 if not token:
  69                     auth = self._download_xml(
  70                         secure_path_data['tokenizer_src'], video_id, query={
  71                             'path': secure_path,
  72                             'videoId': video_id,
  73                         })
  74                     token = xpath_text(auth, 'token')
  75                     if not token:
  76                         continue
  77                     tokens[secure_path] = token
  78                 video_url = video_url + '?hdnea=' + token
  79             elif not re.match('https?://', video_url):
  80                 base_path_data = path_data.get(ext, path_data.get('default', {}))
  81                 media_src = base_path_data.get('media_src')
  82                 if not media_src:
  83                     continue
  84                 video_url = media_src + video_url
  85             if video_url in urls:
  86                 continue
  87             urls.append(video_url)
  88             format_id = video_file.get('bitrate')
  89             if ext == 'smil':
  90                 formats.extend(self._extract_smil_formats(
  91                     video_url, video_id, fatal=False))
  92             elif ext == 'm3u8':
  93                 m3u8_formats = self._extract_m3u8_formats(
  94                     video_url, video_id, 'mp4', m3u8_id=format_id or 'hls',
  95                     fatal=False)
  96                 if m3u8_formats:
  97                     # Sometimes final URLs inside m3u8 are unsigned, let's fix this
  98                     # ourselves
  99                     qs = compat_urlparse.urlparse(video_url).query
 100                     if qs:
 101                         query = compat_urlparse.parse_qs(qs)
 102                         for m3u8_format in m3u8_formats:
 103                             m3u8_format['url'] = update_url_query(m3u8_format['url'], query)
 104                             m3u8_format['extra_param_to_segment_url'] = qs
 105                     formats.extend(m3u8_formats)
 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         timestamp = None
 163         if 'cnn.com' not in data_src:
 164             timestamp = int_or_none(xpath_attr(video_data, 'dateCreated', 'uts'))
 165
 166         return {
 167             'id': video_id,
 168             'title': title,
 169             'formats': formats,
 170             'subtitles': subtitles,
 171             'thumbnails': thumbnails,
 172             'description': xpath_text(video_data, 'description'),
 173             'duration': parse_duration(xpath_text(video_data, 'length') or xpath_text(video_data, 'trt')),
 174             'timestamp': timestamp,
 175             'upload_date': xpath_attr(video_data, 'metas', 'version'),
 176             'series': xpath_text(video_data, 'showTitle'),
 177             'season_number': int_or_none(xpath_text(video_data, 'seasonNumber')),
 178             'episode_number': int_or_none(xpath_text(video_data, 'episodeNumber')),
 179         }