Prepare to release.

[youtubedl] / youtube_dl / extractor / weibo.py

diff --git a/youtube_dl/extractor/weibo.py b/youtube_dl/extractor/weibo.py
index fa784ab994d2b8acede7e4b4496b12779a787de6..3cb4d71a6c2381fe1169b11785a2cc950d2b8437 100644 (file)
--- a/youtube_dl/extractor/weibo.py
+++ b/youtube_dl/extractor/weibo.py
@@ -1,49 +1,140 @@
 # coding: utf-8
 # coding: utf-8

+from __future__ import unicode_literals
+
+from .common import InfoExtractor

 
 

-import re

 import json
 import json

+import random
+import re

 
 

-from .common import InfoExtractor
+from ..compat import (
+    compat_parse_qs,
+    compat_str,
+)
+from ..utils import (
+    js_to_json,
+    strip_jsonp,
+    urlencode_postdata,
+)

 
 

-class WeiboIE(InfoExtractor):
-    """
-    The videos in Weibo come from different sites, this IE just finds the link
-    to the external video and returns it.
-    """
-    _VALID_URL = r'https?://video\.weibo\.com/v/weishipin/t_(?P<id>.+?)\.htm'

 
 

+class WeiboIE(InfoExtractor):
+    _VALID_URL = r'https?://weibo\.com/[0-9]+/(?P<id>[a-zA-Z0-9]+)'

     _TEST = {
     _TEST = {

-        u'add_ie': ['Sina'],
-        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”',
-        },
-        u'note': u'Sina video',
-        u'params': {
-            u'skip_download': True,
-        },
+        'url': 'https://weibo.com/6275294458/Fp6RGfbff?type=comment',
+        'info_dict': {
+            'id': 'Fp6RGfbff',
+            'ext': 'mp4',
+            'title': 'You should have servants to massage you,... 来自Hosico_猫 - 微博',
+        }

     }
 
     }
 

-    # Additional example videos from different sites
-    # Youku: http://video.weibo.com/v/weishipin/t_zQGDWQ8.htm
-    # 56.com: http://video.weibo.com/v/weishipin/t_zQ44HxN.htm
+    def _real_extract(self, url):
+        video_id = self._match_id(url)
+        # to get Referer url for genvisitor
+        webpage, urlh = self._download_webpage_handle(url, video_id)
+
+        visitor_url = urlh.geturl()
+
+        if 'passport.weibo.com' in visitor_url:
+            # first visit
+            visitor_data = self._download_json(
+                'https://passport.weibo.com/visitor/genvisitor', video_id,
+                note='Generating first-visit data',
+                transform_source=strip_jsonp,
+                headers={'Referer': visitor_url},
+                data=urlencode_postdata({
+                    'cb': 'gen_callback',
+                    'fp': json.dumps({
+                        'os': '2',
+                        'browser': 'Gecko57,0,0,0',
+                        'fonts': 'undefined',
+                        'screenInfo': '1440*900*24',
+                        'plugins': '',
+                    }),
+                }))
+
+            tid = visitor_data['data']['tid']
+            cnfd = '%03d' % visitor_data['data']['confidence']
+
+            self._download_webpage(
+                'https://passport.weibo.com/visitor/visitor', video_id,
+                note='Running first-visit callback',
+                query={
+                    'a': 'incarnate',
+                    't': tid,
+                    'w': 2,
+                    'c': cnfd,
+                    'cb': 'cross_domain',
+                    'from': 'weibo',
+                    '_rand': random.random(),
+                })
+
+            webpage = self._download_webpage(
+                url, video_id, note='Revisiting webpage')
+
+        title = self._html_search_regex(
+            r'<title>(.+?)</title>', webpage, 'title')
+
+        video_formats = compat_parse_qs(self._search_regex(
+            r'video-sources=\\\"(.+?)\"', webpage, 'video_sources'))
+
+        formats = []
+        supported_resolutions = (480, 720)
+        for res in supported_resolutions:
+            vid_urls = video_formats.get(compat_str(res))
+            if not vid_urls or not isinstance(vid_urls, list):
+                continue
+
+            vid_url = vid_urls[0]
+            formats.append({
+                'url': vid_url,
+                'height': res,
+            })
+
+        self._sort_formats(formats)
+
+        uploader = self._og_search_property(
+            'nick-name', webpage, 'uploader', default=None)
+
+        return {
+            'id': video_id,
+            'title': title,
+            'uploader': uploader,
+            'formats': formats
+        }
+
+
+class WeiboMobileIE(InfoExtractor):
+    _VALID_URL = r'https?://m\.weibo\.cn/status/(?P<id>[0-9]+)(\?.+)?'
+    _TEST = {
+        'url': 'https://m.weibo.cn/status/4189191225395228?wm=3333_2001&sourcetype=weixin&featurecode=newtitle&from=singlemessage&isappinstalled=0',
+        'info_dict': {
+            'id': '4189191225395228',
+            'ext': 'mp4',
+            'title': '午睡当然是要甜甜蜜蜜的啦',
+            'uploader': '柴犬柴犬'
+        }
+    }

 
     def _real_extract(self, url):
 
     def _real_extract(self, url):

-        mobj = re.match(self._VALID_URL, url, flags=re.VERBOSE)
-        video_id = mobj.group('id')
-        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)
+        video_id = self._match_id(url)
+        # to get Referer url for genvisitor
+        webpage = self._download_webpage(url, video_id, note='visit the page')
+
+        weibo_info = self._parse_json(self._search_regex(
+            r'var\s+\$render_data\s*=\s*\[({.*})\]\[0\]\s*\|\|\s*{};',
+            webpage, 'js_code', flags=re.DOTALL),
+            video_id, transform_source=js_to_json)
+
+        status_data = weibo_info.get('status', {})
+        page_info = status_data.get('page_info')
+        title = status_data['status_title']
+        uploader = status_data.get('user', {}).get('screen_name')

 
 

+        return {
+            'id': video_id,
+            'title': title,
+            'uploader': uploader,
+            'url': page_info['media_info']['stream_url']
+        }