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