[nhl] Add an extractor for videocenter's categories (#1586)

It downloads the last 12 videos.

test/test_playlists.py

@@ -16,6 +16,7 @@ from youtube_dl.extractor import (
     UstreamChannelIE,
     SoundcloudUserIE,
     LivestreamIE,
+    NHLVideocenterIE,
 )
 from youtube_dl.utils import *
 
@@ -74,5 +75,14 @@ class TestPlaylists(unittest.TestCase):
         self.assertEqual(result['title'], u'TEDCity2.0 (English)')
         self.assertTrue(len(result['entries']) >= 4)
 
+    def test_nhl_videocenter(self):
+        dl = FakeYDL()
+        ie = NHLVideocenterIE(dl)
+        result = ie.extract('http://video.canucks.nhl.com/videocenter/console?catid=999')
+        self.assertIsPlaylist(result)
+        self.assertEqual(result['id'], u'999')
+        self.assertEqual(result['title'], u'Highlights')
+        self.assertEqual(len(result['entries']), 12)
+
 if __name__ == '__main__':
     unittest.main()

youtube_dl/extractor/__init__.py

@@ -81,7 +81,7 @@ from .naver import NaverIE
 from .nba import NBAIE
 from .nbc import NBCNewsIE
 from .newgrounds import NewgroundsIE
-from .nhl import NHLIE
+from .nhl import NHLIE, NHLVideocenterIE
 from .ooyala import OoyalaIE
 from .orf import ORFIE
 from .pbs import PBSIE

youtube_dl/extractor/nhl.py

@@ -11,7 +11,40 @@ from ..utils import (
 )
 
 
-class NHLIE(InfoExtractor):
+class NHLBaseInfoExtractor(InfoExtractor):
+    @staticmethod
+    def _fix_json(json_string):
+        return json_string.replace('\\\'', '\'')
+
+    def _extract_video(self, info):
+        video_id = info['id']
+        self.report_extraction(video_id)
+
+        initial_video_url = info['publishPoint']
+        data = compat_urllib_parse.urlencode({
+            'type': 'fvod',
+            'path': initial_video_url.replace('.mp4', '_sd.mp4'),
+        })
+        path_url = 'http://video.nhl.com/videocenter/servlets/encryptvideopath?' + data
+        path_response = self._download_webpage(path_url, video_id,
+            u'Downloading final video url')
+        path_doc = xml.etree.ElementTree.fromstring(path_response)
+        video_url = path_doc.find('path').text
+
+        join = compat_urlparse.urljoin
+        return {
+            'id': video_id,
+            'title': info['name'],
+            'url': video_url,
+            'ext': determine_ext(video_url),
+            'description': info['description'],
+            'duration': int(info['duration']),
+            'thumbnail': join(join(video_url, '/u/'), info['bigImage']),
+            'upload_date': unified_strdate(info['releaseDate'].split('.')[0]),
+        }
+
+
+class NHLIE(NHLBaseInfoExtractor):
     IE_NAME = u'nhl.com'
     _VALID_URL = r'https?://video(?P<team>\.[^.]*)?\.nhl\.com/videocenter/console\?.*?(?<=[?&])id=(?P<id>\d+)'
 
@@ -32,28 +65,56 @@ class NHLIE(InfoExtractor):
         json_url = 'http://video.nhl.com/videocenter/servlets/playlist?ids=%s&format=json' % video_id
         info_json = self._download_webpage(json_url, video_id,
             u'Downloading info json')
-        info_json = info_json.replace('\\\'', '\'')
+        info_json = self._fix_json(info_json)
         info = json.loads(info_json)[0]
+        return self._extract_video(info)
+
+
+class NHLVideocenterIE(NHLBaseInfoExtractor):
+    IE_NAME = u'nhl.com:videocenter'
+    IE_DESC = u'Download the first 12 videos from a videocenter category'
+    _VALID_URL = r'https?://video\.(?P<team>[^.]*)\.nhl\.com/videocenter/(console\?.*?catid=(?P<catid>[^&]+))?'
+
+    @classmethod
+    def suitable(cls, url):
+        if NHLIE.suitable(url):
+            return False
+        return super(NHLVideocenterIE, cls).suitable(url)
+
+    def _real_extract(self, url):
+        mobj = re.match(self._VALID_URL, url)
+        team = mobj.group('team')
+        webpage = self._download_webpage(url, team)
+        cat_id = self._search_regex(
+            [r'var defaultCatId = "(.+?)";',
+             r'{statusIndex:0,index:0,.*?id:(.*?),'],
+            webpage, u'category id')
+        playlist_title = self._html_search_regex(
+            r'\?catid=%s">(.*?)</a>' % cat_id,
+            webpage, u'playlist title', flags=re.DOTALL)
 
-        initial_video_url = info['publishPoint']
         data = compat_urllib_parse.urlencode({
-            'type': 'fvod',
+            'cid': cat_id,
-            'path': initial_video_url.replace('.mp4', '_sd.mp4'),
+            # This is the default value
+            'count': 12,
+            'ptrs': 3,
+            'format': 'json',
         })
-        path_url = 'http://video.nhl.com/videocenter/servlets/encryptvideopath?' + data
+        path = '/videocenter/servlets/browse?' + data
-        path_response = self._download_webpage(path_url, video_id,
+        request_url = compat_urlparse.urljoin(url, path)
-            u'Downloading final video url')
+        response = self._download_webpage(request_url, playlist_title)
-        path_doc = xml.etree.ElementTree.fromstring(path_response)
+        response = self._fix_json(response)
-        video_url = path_doc.find('path').text
+        if not response.strip():
+            self._downloader.report_warning(u'Got an empty reponse, trying '
+                                            u'adding the "newvideos" parameter')
+            response = self._download_webpage(request_url + '&newvideos=true',
+                playlist_title)
+            response = self._fix_json(response)
+        videos = json.loads(response)
 
-        join = compat_urlparse.urljoin
         return {
-            'id': video_id,
+            '_type': 'playlist',
-            'title': info['name'],
+            'title': playlist_title,
-            'url': video_url,
+            'id': cat_id,
-            'ext': determine_ext(video_url),
+            'entries': [self._extract_video(i) for i in videos],
-            'description': info['description'],
-            'duration': int(info['duration']),
-            'thumbnail': join(join(video_url, '/u/'), info['bigImage']),
-            'upload_date': unified_strdate(info['releaseDate'].split('.')[0]),
         }