From de7a91bfe3f761940bf7697653c505dbc229ce6b Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Jaime=20Marqui=CC=81nez=20Ferra=CC=81ndiz?= Date: Fri, 19 Jul 2013 20:43:44 +0200 Subject: [PATCH] WeiboIE: extract the player urls from a json webpage Also extract a Sina url that doesn't require to follow a redirection. --- youtube_dl/extractor/weibo.py | 17 ++++++++++++++--- 1 file changed, 14 insertions(+), 3 deletions(-) diff --git a/youtube_dl/extractor/weibo.py b/youtube_dl/extractor/weibo.py index efcb5912b..0757495bd 100644 --- a/youtube_dl/extractor/weibo.py +++ b/youtube_dl/extractor/weibo.py @@ -1,6 +1,7 @@ # coding: utf-8 import re +import json from .common import InfoExtractor @@ -30,8 +31,18 @@ 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') + info_url = 'http://video.weibo.com/?s=v&a=play_list&format=json&mix_video_id=t_%s' % video_id + info_page = self._download_webpage(info_url, video_id) + info = json.loads(info_page) + + 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: 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)