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