youtube_dl/extractor/sina.py

   1 # coding: utf-8
   2
   3 import re
   4 import xml.etree.ElementTree
   5
   6 from .common import InfoExtractor
   7 from ..utils import (
   8     compat_urllib_request,
   9     compat_urllib_parse,
  10 )
  11
  12
  13 class SinaIE(InfoExtractor):
  14     _VALID_URL = r'''https?://(.*?\.)?video\.sina\.com\.cn/
  15                         (
  16                             (.+?/(((?P<pseudo_id>\d+).html)|(.*?(\#|(vid=))(?P<id>\d+?)($|&))))
  17                             |
  18                             # This is used by external sites like Weibo
  19                             (api/sinawebApi/outplay.php/(?P<token>.+?)\.swf)
  20                         )
  21                   '''
  22
  23     _TEST = {
  24         u'url': u'http://video.sina.com.cn/news/vlist/zt/chczlj2013/?opsubject_id=top12#110028898',
  25         u'file': u'110028898.flv',
  26         u'md5': u'd65dd22ddcf44e38ce2bf58a10c3e71f',
  27         u'info_dict': {
  28             u'title': u'《中国新闻》 朝鲜要求巴拿马立即释放被扣船员',
  29         }
  30     }
  31
  32     @classmethod
  33     def suitable(cls, url):
  34         return re.match(cls._VALID_URL, url, flags=re.VERBOSE) is not None
  35
  36     def _extract_video(self, video_id):
  37         data = compat_urllib_parse.urlencode({'vid': video_id})
  38         url_page = self._download_webpage('http://v.iask.com/v_play.php?%s' % data,
  39             video_id, u'Downloading video url')
  40         image_page = self._download_webpage(
  41             'http://interface.video.sina.com.cn/interface/common/getVideoImage.php?%s' % data,
  42             video_id, u'Downloading thumbnail info')
  43         url_doc = xml.etree.ElementTree.fromstring(url_page.encode('utf-8'))
  44
  45         return {'id': video_id,
  46                 'url': url_doc.find('./durl/url').text,
  47                 'ext': 'flv',
  48                 'title': url_doc.find('./vname').text,
  49                 'thumbnail': image_page.split('=')[1],
  50                 }
  51
  52     def _real_extract(self, url):
  53         mobj = re.match(self._VALID_URL, url, flags=re.VERBOSE)
  54         video_id = mobj.group('id')
  55         if mobj.group('token') is not None:
  56             # The video id is in the redirected url
  57             self.to_screen(u'Getting video id')
  58             request = compat_urllib_request.Request(url)
  59             request.get_method = lambda: 'HEAD'
  60             (_, urlh) = self._download_webpage_handle(request, 'NA', False)
  61             return self._real_extract(urlh.geturl())
  62         elif video_id is None:
  63             pseudo_id = mobj.group('pseudo_id')
  64             webpage = self._download_webpage(url, pseudo_id)
  65             video_id = self._search_regex(r'vid:\'(\d+?)\'', webpage, u'video id')
  66
  67         return self._extract_video(video_id)