YoutubeIE: new algo for length 81 (fixes #1127)

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

diff --git a/youtube_dl/extractor/youtube.py b/youtube_dl/extractor/youtube.py
index 02f215605e72174a1a45cd8b17e5df3196501e99..f86929e06f2591f569c77b166fd7e1769831a772 100644 (file)
--- a/youtube_dl/extractor/youtube.py
+++ b/youtube_dl/extractor/youtube.py
@@ -296,7 +296,7 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
         elif len(s) == 82:
             return s[36] + s[79:67:-1] + s[81] + s[66:40:-1] + s[33] + s[39:36:-1] + s[40] + s[35] + s[0] + s[67] + s[32:0:-1] + s[34]
         elif len(s) == 81:
-            return s[6] + s[3:6] + s[33] + s[7:24] + s[0] + s[25:33] + s[2] + s[34:53] + s[24] + s[54:81]
+            return s[56] + s[79:56:-1] + s[41] + s[55:41:-1] + s[80] + s[40:34:-1] + s[0] + s[33:29:-1] + s[34] + s[28:9:-1] + s[29] + s[8:0:-1] + s[9]
 
         else:
             raise ExtractorError(u'Unable to decrypt signature, key length %d not supported; retrying might work' % (len(s)))
@@ -705,10 +705,9 @@ class YoutubePlaylistIE(InfoExtractor):
 
         # Download playlist videos from API
         playlist_id = mobj.group(1) or mobj.group(2)
-        page_num = 1
         videos = []
 
-        while True:
+        for page_num in itertools.count(1):
             start_index = self._MAX_RESULTS * (page_num - 1) + 1
             if start_index >= 1000:
                 self._downloader.report_warning(u'Max number of results reached')
@@ -732,7 +731,6 @@ class YoutubePlaylistIE(InfoExtractor):
                 index = entry['yt$position']['$t']
                 if 'media$group' in entry and 'media$player' in entry['media$group']:
                     videos.append((index, entry['media$group']['media$player']['url']))
-            page_num += 1
 
         videos = [v[1] for v in sorted(videos)]
 
@@ -776,9 +774,7 @@ class YoutubeChannelIE(InfoExtractor):
 
         # Download any subsequent channel pages using the json-based channel_ajax query
         if self._MORE_PAGES_INDICATOR in page:
-            while True:
-                pagenum = pagenum + 1
-
+            for pagenum in itertools.count(1):
                 url = self._MORE_PAGES_URL % (pagenum, channel_id)
                 page = self._download_webpage(url, channel_id,
                                               u'Downloading page #%s' % pagenum)
@@ -821,9 +817,8 @@ class YoutubeUserIE(InfoExtractor):
         # all of them.
 
         video_ids = []
-        pagenum = 0
 
-        while True:
+        for pagenum in itertools.count(0):
             start_index = pagenum * self._GDATA_PAGE_SIZE + 1
 
             gdata_url = self._GDATA_URL % (username, self._GDATA_PAGE_SIZE, start_index)
@@ -848,8 +843,6 @@ class YoutubeUserIE(InfoExtractor):
             if len(ids_in_page) < self._GDATA_PAGE_SIZE:
                 break
 
-            pagenum += 1
-
         urls = ['http://www.youtube.com/watch?v=%s' % video_id for video_id in video_ids]
         url_results = [self.url_result(rurl, 'Youtube') for rurl in urls]
         return [self.playlist_result(url_results, playlist_title = username)]
@@ -920,10 +913,15 @@ class YoutubeFeedsInfoExtractor(YoutubeBaseInfoExtractor):
     """
     _LOGIN_REQUIRED = True
     _PAGING_STEP = 30
+    # use action_load_personal_feed instead of action_load_system_feed
+    _PERSONAL_FEED = False
 
     @property
     def _FEED_TEMPLATE(self):
-        return 'http://www.youtube.com/feed_ajax?action_load_system_feed=1&feed_name=%s&paging=%%s' % self._FEED_NAME
+        action = 'action_load_system_feed'
+        if self._PERSONAL_FEED:
+            action = 'action_load_personal_feed'
+        return 'http://www.youtube.com/feed_ajax?%s=1&feed_name=%s&paging=%%s' % (action, self._FEED_NAME)
 
     @property
     def IE_NAME(self):
@@ -942,7 +940,7 @@ class YoutubeFeedsInfoExtractor(YoutubeBaseInfoExtractor):
                                           u'Downloading page %s' % i)
             info = json.loads(info)
             feed_html = info['feed_html']
-            m_ids = re.finditer(r'"/watch\?v=(.*?)"', feed_html)
+            m_ids = re.finditer(r'"/watch\?v=(.*?)["&]', feed_html)
             ids = orderedSet(m.group(1) for m in m_ids)
             feed_entries.extend(self.url_result(id, 'Youtube') for id in ids)
             if info['paging'] is None:
@@ -961,6 +959,13 @@ class YoutubeRecommendedIE(YoutubeFeedsInfoExtractor):
     _FEED_NAME = 'recommended'
     _PLAYLIST_TITLE = u'Youtube Recommended videos'
 
+class YoutubeWatchLaterIE(YoutubeFeedsInfoExtractor):
+    IE_DESC = u'Youtube watch later list, "ytwatchlater" keyword (requires authentication)'
+    _VALID_URL = r'https?://www\.youtube\.com/feed/watch_later|:ytwatchlater'
+    _FEED_NAME = 'watch_later'
+    _PLAYLIST_TITLE = u'Youtube Watch Later'
+    _PAGING_STEP = 100
+    _PERSONAL_FEED = True
 
 class YoutubeFavouritesIE(YoutubeBaseInfoExtractor):
     IE_NAME = u'youtube:favorites'