Use original Referer URL in Brightcove requests (Fixes #2110)

youtube_dl/extractor/brightcove.py

@@ -14,6 +14,7 @@
     compat_urllib_request,
 
     ExtractorError,
+    unsmuggle_url,
 )
 
 
@@ -121,6 +122,8 @@ def _extract_brightcove_url(cls, webpage):
             return None
 
     def _real_extract(self, url):
+        url, smuggled_data = unsmuggle_url(url, {})
+
         # Change the 'videoId' and others field to '@videoPlayer'
         url = re.sub(r'(?<=[?&])(videoI(d|D)|bctid)', '%40videoPlayer', url)
         # Change bckey (used by bcove.me urls) to playerKey
@@ -131,9 +134,10 @@ def _real_extract(self, url):
 
         videoPlayer = query.get('@videoPlayer')
         if videoPlayer:
-            return self._get_video_info(videoPlayer[0], query_str, query,
-                # We set the original url as the default 'Referer' header
-                referer=url)
+            # We set the original url as the default 'Referer' header
+            referer = smuggled_data.get('Referer', url)
+            return self._get_video_info(
+                videoPlayer[0], query_str, query, referer=referer)
         else:
             player_key = query['playerKey']
             return self._get_playlist_info(player_key[0])

youtube_dl/extractor/generic.py

@@ -225,7 +225,8 @@ def _real_extract(self, url):
         bc_url = BrightcoveIE._extract_brightcove_url(webpage)
         if bc_url is not None:
             self.to_screen('Brightcove video detected.')
-            return self.url_result(bc_url, 'Brightcove')
+            surl = smuggle_url(bc_url, {'Referer': url})
+            return self.url_result(surl, 'Brightcove')
 
         # Look for embedded (iframe) Vimeo player
         mobj = re.search(

youtube_dl/utils.py

@@ -1027,9 +1027,9 @@ def smuggle_url(url, data):
     return url + u'#' + sdata
 
 
-def unsmuggle_url(smug_url):
+def unsmuggle_url(smug_url, default=None):
     if not '#__youtubedl_smuggle' in smug_url:
-        return smug_url, None
+        return smug_url, default
     url, _, sdata = smug_url.rpartition(u'#')
     jsond = compat_parse_qs(sdata)[u'__youtubedl_smuggle'][0]
     data = json.loads(jsond)