[brightcove] PEP 8

[youtube-dl] / youtube_dl / extractor / brightcove.py

diff --git a/youtube_dl/extractor/brightcove.py b/youtube_dl/extractor/brightcove.py
index c7059a12af287dd439ea342e5b18242abc2391a3..183ac7dd00e3563dde30df01493a3f533ec68ee6 100644 (file)
--- a/youtube_dl/extractor/brightcove.py
+++ b/youtube_dl/extractor/brightcove.py
@@ -131,6 +131,12 @@ class BrightcoveLegacyIE(InfoExtractor):
             },
             'playlist_mincount': 10,
         },
+        {
+            # playerID inferred from bcpid
+            # from http://www.un.org/chinese/News/story.asp?NewsID=27724
+            'url': 'https://link.brightcove.com/services/player/bcpid1722935254001/?bctid=5360463607001&autoStart=false&secureConnections=true&width=650&height=350',
+            'only_matching': True,  # Tested in GenericIE
+        }
     ]
     FLV_VCODECS = {
         1: 'SORENSON',
@@ -266,9 +272,13 @@ class BrightcoveLegacyIE(InfoExtractor):
         if matches:
             return list(filter(None, [cls._build_brighcove_url(m) for m in matches]))
 
-        return list(filter(None, [
-            cls._build_brighcove_url_from_js(custom_bc)
-            for custom_bc in re.findall(r'(customBC\.createVideo\(.+?\);)', webpage)]))
+        matches = re.findall(r'(customBC\.createVideo\(.+?\);)', webpage)
+        if matches:
+            return list(filter(None, [
+                cls._build_brighcove_url_from_js(custom_bc)
+                for custom_bc in matches]))
+        return [src for _, src in re.findall(
+            r'<iframe[^>]+src=([\'"])((?:https?:)?//link\.brightcove\.com/services/player/(?!\1).+)\1', webpage)]
 
     def _real_extract(self, url):
         url, smuggled_data = unsmuggle_url(url, {})
@@ -285,6 +295,10 @@ class BrightcoveLegacyIE(InfoExtractor):
         if videoPlayer:
             # We set the original url as the default 'Referer' header
             referer = smuggled_data.get('Referer', url)
+            if 'playerID' not in query:
+                mobj = re.search(r'/bcpid(\d+)', url)
+                if mobj is not None:
+                    query['playerID'] = [mobj.group(1)]
             return self._get_video_info(
                 videoPlayer[0], query, referer=referer)
         elif 'playerKey' in query:
@@ -484,12 +498,12 @@ class BrightcoveNewIE(InfoExtractor):
     }]
 
     @staticmethod
-    def _extract_url(webpage):
-        urls = BrightcoveNewIE._extract_urls(webpage)
+    def _extract_url(ie, webpage):
+        urls = BrightcoveNewIE._extract_urls(ie, webpage)
         return urls[0] if urls else None
 
     @staticmethod
-    def _extract_urls(webpage):
+    def _extract_urls(ie, webpage):
         # Reference:
         # 1. http://docs.brightcove.com/en/video-cloud/brightcove-player/guides/publish-video.html#setvideoiniframe
         # 2. http://docs.brightcove.com/en/video-cloud/brightcove-player/guides/publish-video.html#tag
@@ -506,37 +520,46 @@ class BrightcoveNewIE(InfoExtractor):
 
         # Look for <video> tags [2] and embed_in_page embeds [3]
         # [2] looks like:
-        # <video data-video-id="5320421710001" data-account="245991542" data-player="SJWAiyYWg" data-embed="default" class="video-js" controls itemscope itemtype="http://schema.org/VideoObject">
-
         for video, script_tag, account_id, player_id, embed in re.findall(
-            r'''(?isx)
-                (<video[^>]+>)
-                (?:.*?
-                  (<script[^>]+
-                    src=["\'](?:https?:)?//players\.brightcove\.net/
-                    (\d+)/([^/]+)_([^/]+)/index(?:\.min)?\.js
-                  )
-                )?
-            ''', webpage
-        ):
+                r'''(?isx)
+                    (<video\s+[^>]+>)
+                    (?:.*?
+                        (<script[^>]+
+                            src=["\'](?:https?:)?//players\.brightcove\.net/
+                            (\d+)/([^/]+)_([^/]+)/index(?:\.min)?\.js
+                        )
+                    )?
+                ''', webpage):
             attrs = extract_attributes(video)
 
             # According to examples from [4] it's unclear whether video id
             # may be optional and what to do when it is
             video_id = attrs.get('data-video-id')
-            # See PR#12099/bostonglobe.py for 'data-brightcove-video-id' variant
+            if not video_id:
+                continue
 
+            account_id = account_id or attrs.get('data-account')
             if not account_id:
-                account_id = attrs.get('data-account')
-            if not player_id:
-                player_id = attrs.get('data-player')
-            if not embed:
-                embed = attrs.get('data-embed')
-
-            if video_id and account_id and player_id and embed:
-                entries.append(
-                    'http://players.brightcove.net/%s/%s_%s/index.html?videoId=%s'
-                    % (account_id, player_id, embed, video_id))
+                continue
+
+            player_id = player_id or attrs.get('data-player') or 'default'
+            embed = embed or attrs.get('data-embed') or 'default'
+
+            bc_url = 'http://players.brightcove.net/%s/%s_%s/index.html?videoId=%s' % (
+                account_id, player_id, embed, video_id)
+
+            # Some brightcove videos may be embedded with video tag only and
+            # without script tag or any mentioning of brightcove at all. Such
+            # embeds are considered ambiguous since they are matched based only
+            # on data-video-id and data-account attributes and in the wild may
+            # not be brightcove embeds at all. Let's check reconstructed
+            # brightcove URLs in case of such embeds and only process valid
+            # ones. By this we ensure there is indeed a brightcove embed.
+            if not script_tag and not ie._is_valid_url(
+                    bc_url, video_id, 'possible brightcove video'):
+                continue
+
+            entries.append(bc_url)
 
         return entries