youtube-dl

commit: 152f22920d73bb0dc24fa357d5904a8dd97a5bf6
parent 20cc7c082b82e82050a4e1f1bb815fee51f6c1c2
Author: Remita Amine <remitamine@gmail.com>
Date:   Fri,  1 Nov 2019 17:44:34 +0100

[wistia] reduce embed extraction false positives and support inline embeds(closes #22931)

Diffstat:
M
youtube_dl/extractor/wistia.py
15
++++++++-------

1 file changed, 8 insertions(+), 7 deletions(-)
diff --git a/youtube_dl/extractor/wistia.py b/youtube_dl/extractor/wistia.py
@@ -12,7 +12,7 @@ from ..utils import (
 
 
 class WistiaIE(InfoExtractor):
-    _VALID_URL = r'(?:wistia:|https?://(?:fast\.)?wistia\.(?:net|com)/embed/(?:iframe|medias)/)(?P<id>[a-z0-9]+)'
+    _VALID_URL = r'(?:wistia:|https?://(?:fast\.)?wistia\.(?:net|com)/embed/(?:iframe|medias)/)(?P<id>[a-z0-9]{10})'
     _API_URL = 'http://fast.wistia.com/embed/medias/%s.json'
     _IFRAME_URL = 'http://fast.wistia.net/embed/iframe/%s'
 
@@ -43,25 +43,26 @@ class WistiaIE(InfoExtractor):
         'only_matching': True,
     }]
 
+    # https://wistia.com/support/embed-and-share/video-on-your-website
     @staticmethod
     def _extract_url(webpage):
         match = re.search(
-            r'<(?:meta[^>]+?content|iframe[^>]+?src)=(["\'])(?P<url>(?:https?:)?//(?:fast\.)?wistia\.(?:net|com)/embed/iframe/.+?)\1', webpage)
+            r'<(?:meta[^>]+?content|(?:iframe|script)[^>]+?src)=["\'](?P<url>(?:https?:)?//(?:fast\.)?wistia\.(?:net|com)/embed/(?:iframe|medias)/[a-z0-9]{10})', webpage)
         if match:
             return unescapeHTML(match.group('url'))
 
-        match = re.search(r'(?:id=["\']wistia_|data-wistia-?id=["\']|Wistia\.embed\(["\'])(?P<id>[^"\']+)', webpage)
-        if match:
-            return 'wistia:%s' % match.group('id')
-
         match = re.search(
             r'''(?sx)
                 <script[^>]+src=(["'])(?:https?:)?//fast\.wistia\.com/assets/external/E-v1\.js\1[^>]*>.*?
-                <div[^>]+class=(["']).*?\bwistia_async_(?P<id>[a-z0-9]+)\b.*?\2
+                <div[^>]+class=(["']).*?\bwistia_async_(?P<id>[a-z0-9]{10})\b.*?\2
             ''', webpage)
         if match:
             return 'wistia:%s' % match.group('id')
 
+        match = re.search(r'(?:data-wistia-?id=["\']|Wistia\.embed\(["\']|id=["\']wistia_)(?P<id>[a-z0-9]{10})', webpage)
+        if match:
+            return 'wistia:%s' % match.group('id')
+
     def _real_extract(self, url):
         video_id = self._match_id(url)