youtube_dl/extractor/sina.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     HEADRequest,
   9     ExtractorError,
  10     int_or_none,
  11     update_url_query,
  12     qualities,
  13     get_element_by_attribute,
  14     clean_html,
  15 )
  16
  17
  18 class SinaIE(InfoExtractor):
  19     _VALID_URL = r'''(?x)https?://(?:.*?\.)?video\.sina\.com\.cn/
  20                         (?:
  21                             (?:view/|.*\#)(?P<video_id>\d+)|
  22                             .+?/(?P<pseudo_id>[^/?#]+)(?:\.s?html)|
  23                             # This is used by external sites like Weibo
  24                             api/sinawebApi/outplay.php/(?P<token>.+?)\.swf
  25                         )
  26                   '''
  27
  28     _TESTS = [
  29         {
  30             'url': 'http://video.sina.com.cn/news/spj/topvideoes20160504/?opsubject_id=top1#250576622',
  31             'md5': 'd38433e2fc886007729735650ae4b3e9',
  32             'info_dict': {
  33                 'id': '250576622',
  34                 'ext': 'mp4',
  35                 'title': '现场:克鲁兹宣布退选 特朗普将稳获提名',
  36             }
  37         },
  38         {
  39             'url': 'http://video.sina.com.cn/v/b/101314253-1290078633.html',
  40             'info_dict': {
  41                 'id': '101314253',
  42                 'ext': 'flv',
  43                 'title': '军方提高对朝情报监视级别',
  44             },
  45             'skip': 'the page does not exist or has been deleted',
  46         },
  47         {
  48             'url': 'http://video.sina.com.cn/view/250587748.html',
  49             'md5': '3d1807a25c775092aab3bc157fff49b4',
  50             'info_dict': {
  51                 'id': '250587748',
  52                 'ext': 'mp4',
  53                 'title': '瞬间泪目：8年前汶川地震珍贵视频首曝光',
  54             },
  55         },
  56     ]
  57
  58     def _real_extract(self, url):
  59         mobj = re.match(self._VALID_URL, url)
  60
  61         video_id = mobj.group('video_id')
  62         if not video_id:
  63             if mobj.group('token') is not None:
  64                 # The video id is in the redirected url
  65                 self.to_screen('Getting video id')
  66                 request = HEADRequest(url)
  67                 _, urlh = self._download_webpage_handle(request, 'NA', False)
  68                 return self._real_extract(urlh.geturl())
  69             else:
  70                 pseudo_id = mobj.group('pseudo_id')
  71                 webpage = self._download_webpage(url, pseudo_id)
  72                 error = get_element_by_attribute('class', 'errtitle', webpage)
  73                 if error:
  74                     raise ExtractorError('%s said: %s' % (
  75                         self.IE_NAME, clean_html(error)), expected=True)
  76                 video_id = self._search_regex(
  77                     r"video_id\s*:\s*'(\d+)'", webpage, 'video id')
  78
  79         video_data = self._download_json(
  80             'http://s.video.sina.com.cn/video/h5play',
  81             video_id, query={'video_id': video_id})
  82         if video_data['code'] != 1:
  83             raise ExtractorError('%s said: %s' % (
  84                 self.IE_NAME, video_data['message']), expected=True)
  85         else:
  86             video_data = video_data['data']
  87             title = video_data['title']
  88             description = video_data.get('description')
  89             if description:
  90                 description = description.strip()
  91
  92             preference = qualities(['cif', 'sd', 'hd', 'fhd', 'ffd'])
  93             formats = []
  94             for quality_id, quality in video_data.get('videos', {}).get('mp4', {}).items():
  95                 file_api = quality.get('file_api')
  96                 file_id = quality.get('file_id')
  97                 if not file_api or not file_id:
  98                     continue
  99                 formats.append({
 100                     'format_id': quality_id,
 101                     'url': update_url_query(file_api, {'vid': file_id}),
 102                     'preference': preference(quality_id),
 103                     'ext': 'mp4',
 104                 })
 105             self._sort_formats(formats)
 106
 107             return {
 108                 'id': video_id,
 109                 'title': title,
 110                 'description': description,
 111                 'thumbnail': video_data.get('image'),
 112                 'duration': int_or_none(video_data.get('length')),
 113                 'timestamp': int_or_none(video_data.get('create_time')),
 114                 'formats': formats,
 115             }