[nhk] fix video extraction(closes #22249)(closes #22353)

pareronia · Jun 22, 2020 · 1ae8d85 · 1ae8d85
1 parent aed6466
commit 1ae8d85
Show file tree

Hide file tree

Showing 2 changed files with 13 additions and 5 deletions.
diff --git a/youtube_dl/extractor/nhk.py b/youtube_dl/extractor/nhk.py
@@ -60,8 +60,8 @@ def get_clean_field(key):
         if is_video:
             info.update({
                 '_type': 'url_transparent',
-                'ie_key': 'Ooyala',
-                'url': 'ooyala:' + episode['vod_id'],
+                'ie_key': 'Piksel',
+                'url': 'https://player.piksel.com/v/refid/nhkworld/prefid/' + episode['vod_id'],
             })
         else:
             audio = episode['audio']

diff --git a/youtube_dl/extractor/piksel.py b/youtube_dl/extractor/piksel.py
@@ -15,7 +15,7 @@
 
 
 class PikselIE(InfoExtractor):
-    _VALID_URL = r'https?://player\.piksel\.com/v/(?P<id>[a-z0-9]+)'
+    _VALID_URL = r'https?://player\.piksel\.com/v/(?:refid/[^/]+/prefid/)?(?P<id>[a-z0-9_]+)'
     _TESTS = [
         {
             'url': 'http://player.piksel.com/v/ums2867l',
@@ -40,6 +40,11 @@ class PikselIE(InfoExtractor):
                 'timestamp': 1486171129,
                 'upload_date': '20170204'
             }
+        },
+        {
+            # https://www3.nhk.or.jp/nhkworld/en/ondemand/video/2019240/
+            'url': 'http://player.piksel.com/v/refid/nhkworld/prefid/nw_vod_v_en_2019_240_20190823233000_02_1566873477',
+            'only_matching': True,
         }
     ]
 
@@ -52,8 +57,11 @@ def _extract_url(webpage):
             return mobj.group('url')
 
     def _real_extract(self, url):
-        video_id = self._match_id(url)
-        webpage = self._download_webpage(url, video_id)
+        display_id = self._match_id(url)
+        webpage = self._download_webpage(url, display_id)
+        video_id = self._search_regex(
+            r'data-de-program-uuid=[\'"]([a-z0-9]+)',
+            webpage, 'program uuid', default=display_id)
         app_token = self._search_regex([
             r'clientAPI\s*:\s*"([^"]+)"',
             r'data-de-api-key\s*=\s*"([^"]+)"'