_ 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_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={}):
  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         # rtmp_src = xpath_text(video_data, 'akamai/src')
  28         # if rtmp_src:
  29         #     splited_rtmp_src = rtmp_src.split(',')
  30         #     if len(splited_rtmp_src) == 2:
  31         #         rtmp_src = splited_rtmp_src[1]
  32         # aifp = xpath_text(video_data, 'akamai/aifp', default='')
  33
  34         tokens = {}
  35         urls = []
  36         formats = []
  37         rex = re.compile(
  38             r'(?P<width>[0-9]+)x(?P<height>[0-9]+)(?:_(?P<bitrate>[0-9]+))?')
  39         # Possible formats locations: files/file, files/groupFiles/files
  40         # and maybe others
  41         for video_file in video_data.findall('.//file'):
  42             video_url = video_file.text.strip()
  43             if not video_url:
  44                 continue
  45             ext = determine_ext(video_url)
  46             if video_url.startswith('/mp4:protected/'):
  47                 continue
  48                 # TODO Correct extraction for these files
  49                 # protected_path_data = path_data.get('protected')
  50                 # if not protected_path_data or not rtmp_src:
  51                 #     continue
  52                 # protected_path = self._search_regex(
  53                 #     r'/mp4:(.+)\.[a-z0-9]', video_url, 'secure path')
  54                 # auth = self._download_webpage(
  55                 #     protected_path_data['tokenizer_src'], query={
  56                 #         'path': protected_path,
  57                 #         'videoId': video_id,
  58                 #         'aifp': aifp,
  59                 #     })
  60                 # token = xpath_text(auth, 'token')
  61                 # if not token:
  62                 #     continue
  63                 # video_url = rtmp_src + video_url + '?' + token
  64             elif video_url.startswith('/secure/'):
  65                 secure_path_data = path_data.get('secure')
  66                 if not secure_path_data:
  67                     continue
  68                 video_url = secure_path_data['media_src'] + video_url
  69                 secure_path = self._search_regex(r'https?://[^/]+(.+/)', video_url, 'secure path') + '*'
  70                 token = tokens.get(secure_path)
  71                 if not token:
  72                     auth = self._download_xml(
  73                         secure_path_data['tokenizer_src'], video_id, query={
  74                             'path': secure_path,
  75                             'videoId': video_id,
  76                         })
  77                     token = xpath_text(auth, 'token')
  78                     if not token:
  79                         continue
  80                     tokens[secure_path] = token
  81                 video_url = video_url + '?hdnea=' + token
  82             elif not re.match('https?://', video_url):
  83                 base_path_data = path_data.get(ext, path_data.get('default', {}))
  84                 media_src = base_path_data.get('media_src')
  85                 if not media_src:
  86                     continue
  87                 video_url = media_src + video_url
  88             if video_url in urls:
  89                 continue
  90             urls.append(video_url)
  91             format_id = video_file.get('bitrate')
  92             if ext == 'smil':
  93                 formats.extend(self._extract_smil_formats(
  94                     video_url, video_id, fatal=False))
  95             elif ext == 'm3u8':
  96                 m3u8_formats = self._extract_m3u8_formats(
  97                     video_url, video_id, 'mp4', m3u8_id=format_id or 'hls',
  98                     fatal=False)
  99                 if m3u8_formats:
 100                     # Sometimes final URLs inside m3u8 are unsigned, let's fix this
 101                     # ourselves
 102                     qs = compat_urlparse.urlparse(video_url).query
 103                     if qs:
 104                         query = compat_urlparse.parse_qs(qs)
 105                         for m3u8_format in m3u8_formats:
 106                             m3u8_format['url'] = update_url_query(m3u8_format['url'], query)
 107                             m3u8_format['extra_param_to_segment_url'] = qs
 108                     formats.extend(m3u8_formats)
 109             elif ext == 'f4m':
 110                 formats.extend(self._extract_f4m_formats(
 111                     update_url_query(video_url, {'hdcore': '3.7.0'}),
 112                     video_id, f4m_id=format_id or 'hds', fatal=False))
 113             else:
 114                 f = {
 115                     'format_id': format_id,
 116                     'url': video_url,
 117                     'ext': ext,
 118                 }
 119                 mobj = rex.search(format_id + video_url)
 120                 if mobj:
 121                     f.update({
 122                         'width': int(mobj.group('width')),
 123                         'height': int(mobj.group('height')),
 124                         'tbr': int_or_none(mobj.group('bitrate')),
 125                     })
 126                 elif isinstance(format_id, compat_str):
 127                     if format_id.isdigit():
 128                         f['tbr'] = int(format_id)
 129                     else:
 130                         mobj = re.match(r'ios_(audio|[0-9]+)$', format_id)
 131                         if mobj:
 132                             if mobj.group(1) == 'audio':
 133                                 f.update({
 134                                     'vcodec': 'none',
 135                                     'ext': 'm4a',
 136                                 })
 137                             else:
 138                                 f['tbr'] = int(mobj.group(1))
 139                 formats.append(f)
 140         self._sort_formats(formats)
 141
 142         subtitles = {}
 143         for source in video_data.findall('closedCaptions/source'):
 144             for track in source.findall('track'):
 145                 track_url = track.get('url')
 146                 if not isinstance(track_url, compat_str) or track_url.endswith('/big'):
 147                     continue
 148                 lang = track.get('lang') or track.get('label') or 'en'
 149                 subtitles.setdefault(lang, []).append({
 150                     'url': track_url,
 151                     'ext': {
 152                         'scc': 'scc',
 153                         'webvtt': 'vtt',
 154                         'smptett': 'tt',
 155                     }.get(source.get('format'))
 156                 })
 157
 158         thumbnails = [{
 159             'id': image.get('cut'),
 160             'url': image.text,
 161             'width': int_or_none(image.get('width')),
 162             'height': int_or_none(image.get('height')),
 163         } for image in video_data.findall('images/image')]
 164
 165         return {
 166             'id': video_id,
 167             'title': title,
 168             'formats': formats,
 169             'subtitles': subtitles,
 170             'thumbnails': thumbnails,
 171             'description': xpath_text(video_data, 'description'),
 172             'duration': parse_duration(xpath_text(video_data, 'length') or xpath_text(video_data, 'trt')),
 173             'timestamp': self._extract_timestamp(video_data),
 174             'upload_date': xpath_attr(video_data, 'metas', 'version'),
 175             'series': xpath_text(video_data, 'showTitle'),
 176             'season_number': int_or_none(xpath_text(video_data, 'seasonNumber')),
 177             'episode_number': int_or_none(xpath_text(video_data, 'episodeNumber')),
 178         }