[youtube] Skip unsupported adaptive stream type (#18804)

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

diff --git a/youtube_dl/extractor/limelight.py b/youtube_dl/extractor/limelight.py
index 422be25288f66a83c32e370e833d6690d83ab54b..729d8de50fab70cd69bab41fae9db0cba4d7da9b 100644 (file)
--- a/youtube_dl/extractor/limelight.py
+++ b/youtube_dl/extractor/limelight.py
@@ -9,6 +9,8 @@ from ..utils import (
     determine_ext,
     float_or_none,
     int_or_none,
+    smuggle_url,
+    try_get,
     unsmuggle_url,
     ExtractorError,
 )
@@ -18,6 +20,49 @@ class LimelightBaseIE(InfoExtractor):
     _PLAYLIST_SERVICE_URL = 'http://production-ps.lvp.llnw.net/r/PlaylistService/%s/%s/%s'
     _API_URL = 'http://api.video.limelight.com/rest/organizations/%s/%s/%s/%s.json'
 
+    @classmethod
+    def _extract_urls(cls, webpage, source_url):
+        lm = {
+            'Media': 'media',
+            'Channel': 'channel',
+            'ChannelList': 'channel_list',
+        }
+
+        def smuggle(url):
+            return smuggle_url(url, {'source_url': source_url})
+
+        entries = []
+        for kind, video_id in re.findall(
+                r'LimelightPlayer\.doLoad(Media|Channel|ChannelList)\(["\'](?P<id>[a-z0-9]{32})',
+                webpage):
+            entries.append(cls.url_result(
+                smuggle('limelight:%s:%s' % (lm[kind], video_id)),
+                'Limelight%s' % kind, video_id))
+        for mobj in re.finditer(
+                # As per [1] class attribute should be exactly equal to
+                # LimelightEmbeddedPlayerFlash but numerous examples seen
+                # that don't exactly match it (e.g. [2]).
+                # 1. http://support.3playmedia.com/hc/en-us/articles/227732408-Limelight-Embedding-the-Captions-Plugin-with-the-Limelight-Player-on-Your-Webpage
+                # 2. http://www.sedona.com/FacilitatorTraining2017
+                r'''(?sx)
+                    <object[^>]+class=(["\'])(?:(?!\1).)*\bLimelightEmbeddedPlayerFlash\b(?:(?!\1).)*\1[^>]*>.*?
+                        <param[^>]+
+                            name=(["\'])flashVars\2[^>]+
+                            value=(["\'])(?:(?!\3).)*(?P<kind>media|channel(?:List)?)Id=(?P<id>[a-z0-9]{32})
+                ''', webpage):
+            kind, video_id = mobj.group('kind'), mobj.group('id')
+            entries.append(cls.url_result(
+                smuggle('limelight:%s:%s' % (kind, video_id)),
+                'Limelight%s' % kind.capitalize(), video_id))
+        # http://support.3playmedia.com/hc/en-us/articles/115009517327-Limelight-Embedding-the-Audio-Description-Plugin-with-the-Limelight-Player-on-Your-Web-Page)
+        for video_id in re.findall(
+                r'(?s)LimelightPlayerUtil\.embed\s*\(\s*{.*?\bmediaId["\']\s*:\s*["\'](?P<id>[a-z0-9]{32})',
+                webpage):
+            entries.append(cls.url_result(
+                smuggle('limelight:media:%s' % video_id),
+                LimelightMediaIE.ie_key(), video_id))
+        return entries
+
     def _call_playlist_service(self, item_id, method, fatal=True, referer=None):
         headers = {}
         if referer:
@@ -62,13 +107,21 @@ class LimelightBaseIE(InfoExtractor):
                 fmt = {
                     'url': stream_url,
                     'abr': float_or_none(stream.get('audioBitRate')),
-                    'vbr': float_or_none(stream.get('videoBitRate')),
                     'fps': float_or_none(stream.get('videoFrameRate')),
-                    'width': int_or_none(stream.get('videoWidthInPixels')),
-                    'height': int_or_none(stream.get('videoHeightInPixels')),
                     'ext': ext,
                 }
-                rtmp = re.search(r'^(?P<url>rtmpe?://(?P<host>[^/]+)/(?P<app>.+))/(?P<playpath>mp4:.+)$', stream_url)
+                width = int_or_none(stream.get('videoWidthInPixels'))
+                height = int_or_none(stream.get('videoHeightInPixels'))
+                vbr = float_or_none(stream.get('videoBitRate'))
+                if width or height or vbr:
+                    fmt.update({
+                        'width': width,
+                        'height': height,
+                        'vbr': vbr,
+                    })
+                else:
+                    fmt['vcodec'] = 'none'
+                rtmp = re.search(r'^(?P<url>rtmpe?://(?P<host>[^/]+)/(?P<app>.+))/(?P<playpath>mp[34]:.+)$', stream_url)
                 if rtmp:
                     format_id = 'rtmp'
                     if stream.get('videoBitRate'):
@@ -168,6 +221,12 @@ class LimelightBaseIE(InfoExtractor):
             'subtitles': subtitles,
         }
 
+    def _extract_info_helper(self, pc, mobile, i, metadata):
+        return self._extract_info(
+            try_get(pc, lambda x: x['playlistItems'][i]['streams'], list) or [],
+            try_get(mobile, lambda x: x['mediaList'][i]['mobileUrls'], list) or [],
+            metadata)
+
 
 class LimelightMediaIE(LimelightBaseIE):
     IE_NAME = 'limelight'
@@ -223,17 +282,16 @@ class LimelightMediaIE(LimelightBaseIE):
     def _real_extract(self, url):
         url, smuggled_data = unsmuggle_url(url, {})
         video_id = self._match_id(url)
-        self._initialize_geo_bypass(smuggled_data.get('geo_countries'))
+        self._initialize_geo_bypass({
+            'countries': smuggled_data.get('geo_countries'),
+        })
 
         pc, mobile, metadata = self._extract(
             video_id, 'getPlaylistByMediaId',
             'getMobilePlaylistByMediaId', 'properties',
             smuggled_data.get('source_url'))
 
-        return self._extract_info(
-            pc['playlistItems'][0].get('streams', []),
-            mobile['mediaList'][0].get('mobileUrls', []) if mobile else [],
-            metadata)
+        return self._extract_info_helper(pc, mobile, 0, metadata)
 
 
 class LimelightChannelIE(LimelightBaseIE):
@@ -274,10 +332,7 @@ class LimelightChannelIE(LimelightBaseIE):
             'media', smuggled_data.get('source_url'))
 
         entries = [
-            self._extract_info(
-                pc['playlistItems'][i].get('streams', []),
-                mobile['mediaList'][i].get('mobileUrls', []) if mobile else [],
-                medias['media_list'][i])
+            self._extract_info_helper(pc, mobile, i, medias['media_list'][i])
             for i in range(len(medias['media_list']))]
 
         return self.playlist_result(entries, channel_id, pc['title'])