Switch YTPlaylistIE to API (relevant: #586); fixes #651; fixes #673; fixes #661

test/test_youtube_lists.py

@@ -8,7 +8,7 @@ import json
 import os
 sys.path.append(os.path.dirname(os.path.dirname(os.path.abspath(__file__))))
 
-from youtube_dl.InfoExtractors import YoutubeUserIE,YoutubePlaylistIE
+from youtube_dl.InfoExtractors import YoutubeUserIE, YoutubePlaylistIE, YoutubeIE
 from youtube_dl.utils import *
 
 PARAMETERS_FILE = os.path.join(os.path.dirname(os.path.abspath(__file__)), "parameters.json")
@@ -38,11 +38,8 @@ class TestYoutubeLists(unittest.TestCase):
         DL = FakeDownloader()
         IE = YoutubePlaylistIE(DL)
         IE.extract('https://www.youtube.com/playlist?list=PLwiyx1dc3P2JR9N8gQaQN_BCvlSlap7re')
-        self.assertEqual(DL.result, [
+        self.assertEqual(map(lambda x: YoutubeIE()._extract_id(x[0]), DL.result),
-            ['http://www.youtube.com/watch?v=bV9L5Ht9LgY'],
+            [ 'bV9L5Ht9LgY', 'FXxLjLQi3Fg', 'tU3Bgo5qJZE' ])
-            ['http://www.youtube.com/watch?v=FXxLjLQi3Fg'],
-            ['http://www.youtube.com/watch?v=tU3Bgo5qJZE']
-        ])
 
     def test_youtube_playlist_long(self):
         DL = FakeDownloader()
@@ -50,14 +47,21 @@ class TestYoutubeLists(unittest.TestCase):
         IE.extract('https://www.youtube.com/playlist?list=UUBABnxM4Ar9ten8Mdjj1j0Q')
         self.assertTrue(len(DL.result) >= 799)
 
+    def test_youtube_playlist_with_deleted(self):
+        DL = FakeDownloader()
+        IE = YoutubePlaylistIE(DL)
+        IE.extract('https://www.youtube.com/playlist?list=PLwP_SiAcdui0KVebT0mU9Apz359a4ubsC')
+        self.assertFalse('pElCt5oNDuI' in map(lambda x: YoutubeIE()._extract_id(x[0]), DL.result))
+        self.assertFalse('KdPEApIVdWM' in map(lambda x: YoutubeIE()._extract_id(x[0]), DL.result))
+
     def test_youtube_course(self):
         DL = FakeDownloader()
         IE = YoutubePlaylistIE(DL)
         # TODO find a > 100 (paginating?) videos course
         IE.extract('https://www.youtube.com/course?list=ECUl4u3cNGP61MdtwGTqZA0MreSaDybji8')
-        self.assertEqual(DL.result[0], ['http://www.youtube.com/watch?v=j9WZyLZCBzs'])
+        self.assertEqual(YoutubeIE()._extract_id(DL.result[0][0]), 'j9WZyLZCBzs')
         self.assertEqual(len(DL.result), 25)
-        self.assertEqual(DL.result[-1], ['http://www.youtube.com/watch?v=rYefUsYuEp0'])
+        self.assertEqual(YoutubeIE()._extract_id(DL.result[-1][0]), 'rYefUsYuEp0')
 
     def test_youtube_channel(self):
         # I give up, please find a channel that does paginate and test this like test_youtube_playlist_long

youtube_dl/InfoExtractors.py

@@ -15,6 +15,7 @@ import email.utils
 import xml.etree.ElementTree
 import random
 import math
+import operator
 
 from .utils import *
 
@@ -1662,22 +1663,40 @@ class YahooSearchIE(InfoExtractor):
 class YoutubePlaylistIE(InfoExtractor):
     """Information Extractor for YouTube playlists."""
 
-    _VALID_URL = r'(?:(?:https?://)?(?:\w+\.)?youtube\.com/(?:(?:course|view_play_list|my_playlists|artist|playlist)\?.*?(p|a|list)=|user/.*?/user/|p/|user/.*?#[pg]/c/)(?:PL|EC)?|PL|EC)([0-9A-Za-z-_]{10,})(?:/.*?/([0-9A-Za-z_-]+))?.*'
+    _VALID_URL = r"""(?:
-    _TEMPLATE_URL = 'http://www.youtube.com/%s?%s=%s&page=%s&gl=US&hl=en'
+                        (?:https?://)?
-    _VIDEO_INDICATOR_TEMPLATE = r'/watch\?v=(.+?)&([^&"]+&)*list=.*?%s'
+                        (?:\w+\.)?
-    _MORE_PAGES_INDICATOR = u"Next \N{RIGHT-POINTING DOUBLE ANGLE QUOTATION MARK}"
+                        youtube\.com/
+                        (?:
+                           (?:course|view_play_list|my_playlists|artist|playlist)
+                           \? .*? (p|a|list)=
+                        |  user/.*?/user/
+                        |  p/
+                        |  user/.*?#[pg]/c/
+                        )
+                        (?:PL|EC)?
+                     |PL|EC)
+                     ([0-9A-Za-z-_]{10,})
+                     (?:/.*?/([0-9A-Za-z_-]+))?
+                     .*"""
+    _TEMPLATE_URL = 'https://gdata.youtube.com/feeds/api/playlists/%s?max-results=%i&start-index=%i&v=2&alt=json'
+    _MAX_RESULTS = 50
     IE_NAME = u'youtube:playlist'
 
     def __init__(self, downloader=None):
         InfoExtractor.__init__(self, downloader)
 
+    def suitable(self, url):
+        """Receives a URL and returns True if suitable for this IE."""
+        return re.match(self._VALID_URL, url, re.VERBOSE) is not None
+
     def report_download_page(self, playlist_id, pagenum):
         """Report attempt to download playlist page with given number."""
         self._downloader.to_screen(u'[youtube] PL %s: Downloading page #%s' % (playlist_id, pagenum))
 
     def _real_extract(self, url):
         # Extract playlist id
-        mobj = re.match(self._VALID_URL, url)
+        mobj = re.match(self._VALID_URL, url, re.VERBOSE)
         if mobj is None:
             self._downloader.trouble(u'ERROR: invalid url: %s' % url)
             return
@@ -1687,55 +1706,51 @@ class YoutubePlaylistIE(InfoExtractor):
             self._downloader.download([mobj.group(3)])
             return
 
-        # Download playlist pages
+        # Download playlist videos from API
-        # prefix is 'p' as default for playlists but there are other types that need extra care
-        playlist_prefix = mobj.group(1)
-        if playlist_prefix == 'a':
-            playlist_access = 'artist'
-        else:
-            playlist_prefix = 'p'
-            playlist_access = 'view_play_list'
         playlist_id = mobj.group(2)
-        video_ids = []
+        page_num = 1
-        pagenum = 1
+        videos = []
 
         while True:
-            self.report_download_page(playlist_id, pagenum)
+            self.report_download_page(playlist_id, page_num)
-            url = self._TEMPLATE_URL % (playlist_access, playlist_prefix, playlist_id, pagenum)
+
-            request = compat_urllib_request.Request(url)
+            url = self._TEMPLATE_URL % (playlist_id, self._MAX_RESULTS, self._MAX_RESULTS * (page_num - 1) + 1)
             try:
-                page = compat_urllib_request.urlopen(request).read().decode('utf-8')
+                page = compat_urllib_request.urlopen(url).read().decode('utf8')
             except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
                 self._downloader.trouble(u'ERROR: unable to download webpage: %s' % compat_str(err))
                 return
 
-            # Extract video identifiers
+            try:
-            ids_in_page = []
+                response = json.loads(page)
-            for mobj in re.finditer(self._VIDEO_INDICATOR_TEMPLATE % playlist_id, page):
+            except ValueError as err:
-                if mobj.group(1) not in ids_in_page:
+                self._downloader.trouble(u'ERROR: Invalid JSON in API response: ' + compat_str(err))
-                    ids_in_page.append(mobj.group(1))
+                return
-            video_ids.extend(ids_in_page)
 
-            if self._MORE_PAGES_INDICATOR not in page:
+            videos += [(entry['yt$position']['$t'], entry['content']['src']) for entry in response['feed']['entry']]
+
+            if len(response['feed']['entry']) < self._MAX_RESULTS:
                 break
-            pagenum = pagenum + 1
+            page_num += 1
 
-        total = len(video_ids)
+        videos = map(operator.itemgetter(1), sorted(videos))
+
+        total = len(videos)
 
         playliststart = self._downloader.params.get('playliststart', 1) - 1
         playlistend = self._downloader.params.get('playlistend', -1)
         if playlistend == -1:
-            video_ids = video_ids[playliststart:]
+            videos = videos[playliststart:]
         else:
-            video_ids = video_ids[playliststart:playlistend]
+            videos = videos[playliststart:playlistend]
 
-        if len(video_ids) == total:
+        if len(videos) == total:
             self._downloader.to_screen(u'[youtube] PL %s: Found %i videos' % (playlist_id, total))
         else:
-            self._downloader.to_screen(u'[youtube] PL %s: Found %i videos, downloading %i' % (playlist_id, total, len(video_ids)))
+            self._downloader.to_screen(u'[youtube] PL %s: Found %i videos, downloading %i' % (playlist_id, total, len(videos)))
 
-        for id in video_ids:
+        for video in videos:
-            self._downloader.download(['http://www.youtube.com/watch?v=%s' % id])
+            self._downloader.download([video])
         return
 
 
@@ -3605,9 +3620,9 @@ class TweetReelIE(InfoExtractor):
             'upload_date': upload_date
         }
         return [info]
-        
+
 class SteamIE(InfoExtractor):
-    _VALID_URL = r"""http://store.steampowered.com/ 
+    _VALID_URL = r"""http://store.steampowered.com/
                 (?P<urltype>video|app)/ #If the page is only for videos or for a game
                 (?P<gameID>\d+)/?
                 (?P<videoID>\d*)(?P<extra>\??) #For urltype == video we sometimes get the videoID
@@ -3707,7 +3722,7 @@ class RBMARadioIE(InfoExtractor):
 class YouPornIE(InfoExtractor):
     """Information extractor for youporn.com."""
     _VALID_URL = r'^(?:https?://)?(?:\w+\.)?youporn\.com/watch/(?P<videoid>[0-9]+)/(?P<title>[^/]+)'
-   
+
     def _print_formats(self, formats):
         """Print all available formats"""
         print(u'Available formats:')
@@ -3769,8 +3784,8 @@ class YouPornIE(InfoExtractor):
         links = re.findall(LINK_RE, download_list_html)
         if(len(links) == 0):
             raise ExtractorError(u'ERROR: no known formats available for video')
-        
+
-        self._downloader.to_screen(u'[youporn] Links found: %d' % len(links))   
+        self._downloader.to_screen(u'[youporn] Links found: %d' % len(links))
 
         formats = []
         for link in links:
@@ -3821,7 +3836,7 @@ class YouPornIE(InfoExtractor):
                 return
             return [format]
 
-        
+
 
 class PornotubeIE(InfoExtractor):
     """Information extractor for pornotube.com."""
@@ -3893,7 +3908,7 @@ class YouJizzIE(InfoExtractor):
 
         embed_page_url = result.group(0).strip()
         video_id = result.group('videoid')
-    
+
         webpage = self._download_webpage(embed_page_url, video_id)
 
         # Get the video URL
@@ -4053,7 +4068,7 @@ class TEDIE(InfoExtractor):
 
 class MySpassIE(InfoExtractor):
     _VALID_URL = r'http://www.myspass.de/.*'
-    
+
     def _real_extract(self, url):
         META_DATA_URL_TEMPLATE = 'http://www.myspass.de/myspass/includes/apps/video/getvideometadataxml.php?id=%s'
 
@@ -4063,12 +4078,12 @@ class MySpassIE(InfoExtractor):
         url_parent_path, video_id = os.path.split(url_path)
         if not video_id:
             _, video_id = os.path.split(url_parent_path)
-        
+
         # get metadata
         metadata_url = META_DATA_URL_TEMPLATE % video_id
         metadata_text = self._download_webpage(metadata_url, video_id)
         metadata = xml.etree.ElementTree.fromstring(metadata_text.encode('utf-8'))
-        
+
         # extract values from metadata
         url_flv_el = metadata.find('url_flv')
         if url_flv_el is None: