PEP8: applied even more rules

[youtube-dl] / youtube_dl / extractor / weibo.py

diff --git a/youtube_dl/extractor/weibo.py b/youtube_dl/extractor/weibo.py
index efcb5912b46e06126db4fcd9ed93eab55be81d5b..20bb039d38dd9c62ff6c38135f67f9aa41f484aa 100644 (file)
--- a/youtube_dl/extractor/weibo.py
+++ b/youtube_dl/extractor/weibo.py
@@ -1,9 +1,11 @@
 # coding: utf-8
+from __future__ import unicode_literals
 
 import re
 
 from .common import InfoExtractor
 
+
 class WeiboIE(InfoExtractor):
     """
     The videos in Weibo come from different sites, this IE just finds the link
@@ -12,15 +14,16 @@ class WeiboIE(InfoExtractor):
     _VALID_URL = r'https?://video\.weibo\.com/v/weishipin/t_(?P<id>.+?)\.htm'
 
     _TEST = {
-        u'url': u'http://video.weibo.com/v/weishipin/t_zjUw2kZ.htm',
-        u'file': u'98322879.flv',
-        u'info_dict': {
-            u'title': u'魔声耳机最新广告“All Eyes On Us”',
+        'url': 'http://video.weibo.com/v/weishipin/t_zjUw2kZ.htm',
+        'info_dict': {
+            'id': '98322879',
+            'ext': 'flv',
+            'title': '魔声耳机最新广告“All Eyes On Us”',
         },
-        u'note': u'Sina video',
-        u'params': {
-            u'skip_download': True,
+        'params': {
+            'skip_download': True,
         },
+        'add_ie': ['Sina'],
     }
 
     # Additional example videos from different sites
@@ -30,8 +33,17 @@ class WeiboIE(InfoExtractor):
     def _real_extract(self, url):
         mobj = re.match(self._VALID_URL, url, flags=re.VERBOSE)
         video_id = mobj.group('id')
-        webpage = self._download_webpage(url, video_id)
-        player_url = self._search_regex(r'var defaultPlayer="(.+?)"', webpage,
-                                        u'player url')
-        return self.url_result(player_url)
+        info_url = 'http://video.weibo.com/?s=v&a=play_list&format=json&mix_video_id=t_%s' % video_id
+        info = self._download_json(info_url, video_id)
 
+        videos_urls = map(lambda v: v['play_page_url'], info['result']['data'])
+        # Prefer sina video since they have thumbnails
+        videos_urls = sorted(videos_urls, key=lambda u: 'video.sina.com' in u)
+        player_url = videos_urls[-1]
+        m_sina = re.match(r'https?://video\.sina\.com\.cn/v/b/(\d+)-\d+\.html',
+                          player_url)
+        if m_sina is not None:
+            self.to_screen('Sina video detected')
+            sina_id = m_sina.group(1)
+            player_url = 'http://you.video.sina.com.cn/swf/quotePlayer.swf?vid=%s' % sina_id
+        return self.url_result(player_url)