youtube-dl

commit 652cdaa269725dfbf9effdc18a8fd0b369100399
parent e26f8712289c727a43d74a4669aee4924b9f75f2
Author: Jaime Marquínez Ferrándiz <jaime.marquinez.ferrandiz@gmail.com>
Date:   Tue, 26 Nov 2013 21:35:03 +0100

[youtube:playlist] Add support for YouTube mixes (fixes #1839)

Diffstat:
M
test/test_youtube_lists.py
 | 
9
+++++++++
M
youtube_dl/extractor/youtube.py
 | 
24
++++++++++++++++++++++--

2 files changed, 31 insertions(+), 2 deletions(-)
diff --git a/test/test_youtube_lists.py b/test/test_youtube_lists.py
@@ -107,5 +107,14 @@ class TestYoutubeLists(unittest.TestCase):
         result = ie.extract('http://www.youtube.com/show/airdisasters')
         self.assertTrue(len(result) >= 3)
 
+    def test_youtube_mix(self):
+        dl = FakeYDL()
+        ie = YoutubePlaylistIE(dl)
+        result = ie.extract('http://www.youtube.com/watch?v=lLJf9qJHR3E&list=RDrjFaenf1T-Y')
+        entries = result['entries']
+        self.assertTrue(len(entries) >= 20)
+        original_video = entries[0]
+        self.assertEqual(original_video['id'], 'rjFaenf1T-Y')
+
 if __name__ == '__main__':
     unittest.main()
diff --git a/youtube_dl/extractor/youtube.py b/youtube_dl/extractor/youtube.py
@@ -28,6 +28,7 @@ from ..utils import (
     clean_html,
     get_cachedir,
     get_element_by_id,
+    get_element_by_attribute,
     ExtractorError,
     unescapeHTML,
     unified_strdate,
@@ -1537,6 +1538,22 @@ class YoutubePlaylistIE(YoutubeBaseInfoExtractor):
     def _real_initialize(self):
         self._login()
 
+    def _ids_to_results(self, ids):
+        return [self.url_result(vid_id, 'Youtube', video_id=vid_id)
+                       for vid_id in ids]
+
+    def _extract_mix(self, playlist_id):
+        # The mixes are generated from a a single video
+        # the id of the playlist is just 'RD' + video_id
+        url = 'https://youtube.com/watch?v=%s&list=%s' % (playlist_id[2:], playlist_id)
+        webpage = self._download_webpage(url, playlist_id, u'Downloading Youtube mix')
+        title = clean_html(get_element_by_attribute('class', 'title long-title', webpage))
+        video_re = r'data-index="\d+".*?href="/watch\?v=([0-9A-Za-z_-]{11})&[^"]*?list=%s' % re.escape(playlist_id)
+        ids = orderedSet(re.findall(video_re, webpage))
+        url_results = self._ids_to_results(ids)
+
+        return self.playlist_result(url_results, playlist_id, title)
+
     def _real_extract(self, url):
         # Extract playlist id
         mobj = re.match(self._VALID_URL, url, re.VERBOSE)
@@ -1554,6 +1571,10 @@ class YoutubePlaylistIE(YoutubeBaseInfoExtractor):
             else:
                 self.to_screen(u'Downloading playlist PL%s - add --no-playlist to just download video %s' % (playlist_id, video_id))
 
+        if len(playlist_id) == 13:  # 'RD' + 11 characters for the video id
+            # Mixes require a custom extraction process
+            return self._extract_mix(playlist_id)
+
         # Extract the video ids from the playlist pages
         ids = []
 
@@ -1571,8 +1592,7 @@ class YoutubePlaylistIE(YoutubeBaseInfoExtractor):
 
         playlist_title = self._og_search_title(page)
 
-        url_results = [self.url_result(vid_id, 'Youtube', video_id=vid_id)
-                       for vid_id in ids]
+        url_results = self._ids_to_results(ids)
         return self.playlist_result(url_results, playlist_id, playlist_title)