_ 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 ..utils import (
   8     xpath_text,
   9     int_or_none,
  10     determine_ext,
  11     parse_duration,
  12     xpath_attr,
  13     update_url_query,
  14     compat_urlparse,
  15 )
  16
  17
  18 class TurnerBaseIE(InfoExtractor):
  19     def _extract_cvp_info(self, data_src, video_id, path_data={}):
  20         video_data = self._download_xml(data_src, video_id)
  21         video_id = video_data.attrib['id'].split('/')[-1].split('.')[0]
  22         title = xpath_text(video_data, 'headline', fatal=True)
  23         # rtmp_src = xpath_text(video_data, 'akamai/src')
  24         # if rtmp_src:
  25         #     splited_rtmp_src = rtmp_src.split(',')
  26         #     if len(splited_rtmp_src) == 2:
  27         #         rtmp_src = splited_rtmp_src[1]
  28         # aifp = xpath_text(video_data, 'akamai/aifp', default='')
  29
  30         tokens = {}
  31         urls = []
  32         formats = []
  33         rex = re.compile(r'''(?x)
  34             (?P<width>[0-9]+)x(?P<height>[0-9]+)
  35             (?:_(?P<bitrate>[0-9]+))?
  36         ''')
  37         for video_file in video_data.findall('files/file'):
  38             video_url = video_file.text.strip()
  39             if not video_url:
  40                 continue
  41             ext = determine_ext(video_url)
  42             if video_url.startswith('/mp4:protected/'):
  43                 continue
  44                 # TODO Correct extraction for these files
  45                 # protected_path_data = path_data.get('protected')
  46                 # if not protected_path_data or not rtmp_src:
  47                 #     continue
  48                 # protected_path = self._search_regex(
  49                 #     r'/mp4:(.+)\.[a-z0-9]', video_url, 'secure path')
  50                 # auth = self._download_webpage(
  51                 #     protected_path_data['tokenizer_src'], query={
  52                 #         'path': protected_path,
  53                 #         'videoId': video_id,
  54                 #         'aifp': aifp,
  55                 #     })
  56                 # token = xpath_text(auth, 'token')
  57                 # if not token:
  58                 #     continue
  59                 # video_url = rtmp_src + video_url + '?' + token
  60             elif video_url.startswith('/secure/'):
  61                 secure_path_data = path_data.get('secure')
  62                 if not secure_path_data:
  63                     continue
  64                 video_url = secure_path_data['media_src'] + video_url
  65                 secure_path = self._search_regex(r'https?://[^/]+(.+/)', video_url, 'secure path') + '*'
  66                 token = tokens.get(secure_path)
  67                 if not token:
  68                     auth = self._download_xml(
  69                         secure_path_data['tokenizer_src'], video_id, query={
  70                             'path': secure_path,
  71                             'videoId': video_id,
  72                         })
  73                     token = xpath_text(auth, 'token')
  74                     if not token:
  75                         continue
  76                     tokens[secure_path] = token
  77                 video_url = video_url + '?hdnea=' + token
  78             elif not re.match('https?://', video_url):
  79                 base_path_data = path_data.get(ext, path_data.get('default', {}))
  80                 media_src = base_path_data.get('media_src')
  81                 if not media_src:
  82                     continue
  83                 video_url = media_src + video_url
  84             if video_url in urls:
  85                 continue
  86             urls.append(video_url)
  87             format_id = video_file.attrib['bitrate']
  88             if ext == 'smil':
  89                 formats.extend(self._extract_smil_formats(video_url, video_id, fatal=False))
  90             elif ext == 'm3u8':
  91                 m3u8_formats = self._extract_m3u8_formats(
  92                     video_url, video_id, 'mp4', m3u8_id=format_id, fatal=False)
  93                 if m3u8_formats:
  94                     # Sometimes final URLs inside m3u8 are unsigned, let's fix this
  95                     # ourselves
  96                     qs = compat_urlparse.urlparse(video_url).query
  97                     if qs:
  98                         query = compat_urlparse.parse_qs(qs)
  99                         for m3u8_format in m3u8_formats:
 100                             m3u8_format['url'] = update_url_query(m3u8_format['url'], query)
 101                             m3u8_format['extra_param_to_segment_url'] = qs
 102                     formats.extend(m3u8_formats)
 103             elif ext == 'f4m':
 104                 formats.extend(self._extract_f4m_formats(
 105                     update_url_query(video_url, {'hdcore': '3.7.0'}),
 106                     video_id, f4m_id=format_id, fatal=False))
 107             else:
 108                 f = {
 109                     'format_id': format_id,
 110                     'url': video_url,
 111                     'ext': ext,
 112                 }
 113                 mobj = rex.search(format_id + video_url)
 114                 if mobj:
 115                     f.update({
 116                         'width': int(mobj.group('width')),
 117                         'height': int(mobj.group('height')),
 118                         'tbr': int_or_none(mobj.group('bitrate')),
 119                     })
 120                 elif format_id.isdigit():
 121                     f['tbr'] = int(format_id)
 122                 else:
 123                     mobj = re.match(r'ios_(audio|[0-9]+)$', format_id)
 124                     if mobj:
 125                         if mobj.group(1) == 'audio':
 126                             f.update({
 127                                 'vcodec': 'none',
 128                                 'ext': 'm4a',
 129                             })
 130                         else:
 131                             f['tbr'] = int(mobj.group(1))
 132                 formats.append(f)
 133         self._sort_formats(formats)
 134
 135         subtitles = {}
 136         for source in video_data.findall('closedCaptions/source'):
 137             for track in source.findall('track'):
 138                 track_url = track.get('url')
 139                 if not track_url:
 140                     continue
 141                 lang = track.get('lang') or track.get('label') or 'en'
 142                 subtitles.setdefault(lang, []).append({
 143                     'url': track_url,
 144                     'ext': {
 145                         'scc': 'scc',
 146                         'webvtt': 'vtt',
 147                         'smptett': 'tt',
 148                     }.get(source.get('format'))
 149                 })
 150
 151         thumbnails = [{
 152             'id': image.get('cut'),
 153             'url': image.text,
 154             'width': int_or_none(image.get('width')),
 155             'height': int_or_none(image.get('height')),
 156         } for image in video_data.findall('images/image')]
 157
 158         timestamp = None
 159         if 'cnn.com' not in data_src:
 160             timestamp = int_or_none(xpath_attr(video_data, 'dateCreated', 'uts'))
 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': timestamp,
 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         }