X-Git-Url: http://git.bitcoin.ninja/index.cgi?a=blobdiff_plain;f=youtube_dl%2Fextractor%2Fgeneric.py;h=dfc2ef4e72f0804bc0a9e8f0099252a1f89a83c3;hb=823f1e015a8880442f9b9fcf2943f647f5cf8039;hp=263aa8579a47b0943546ecc5c104b7af1a8b1810;hpb=a8eb5a8e610a2b90eac2789d5b5f3cda81f543bb;p=youtube-dl

diff --git a/youtube_dl/extractor/generic.py b/youtube_dl/extractor/generic.py
index 263aa8579..dfc2ef4e7 100644
--- a/youtube_dl/extractor/generic.py
+++ b/youtube_dl/extractor/generic.py
@@ -155,7 +155,6 @@ class GenericIE(InfoExtractor):
         # funnyordie embed
         {
             'url': 'http://www.theguardian.com/world/2014/mar/11/obama-zach-galifianakis-between-two-ferns',
-            'md5': '7cf780be104d40fea7bae52eed4a470e',
             'info_dict': {
                 'id': '18e820ec3f',
                 'ext': 'mp4',
@@ -226,21 +225,6 @@ class GenericIE(InfoExtractor):
                 'skip_download': 'Requires rtmpdump'
             }
         },
-        # smotri embed
-        {
-            'url': 'http://rbctv.rbc.ru/archive/news/562949990879132.shtml',
-            'md5': 'ec40048448e9284c9a1de77bb188108b',
-            'info_dict': {
-                'id': 'v27008541fad',
-                'ext': 'mp4',
-                'title': 'Крым и Севастополь вошли в состав России',
-                'description': 'md5:fae01b61f68984c7bd2fa741e11c3175',
-                'duration': 900,
-                'upload_date': '20140318',
-                'uploader': 'rbctv_2012_4',
-                'uploader_id': 'rbctv_2012_4',
-            },
-        },
         # Condé Nast embed
         {
             'url': 'http://www.wired.com/2014/04/honda-asimo/',
@@ -655,6 +639,16 @@ class GenericIE(InfoExtractor):
             return _playlist_from_matches(
                 matches, lambda m: unescapeHTML(m[1]))
 
+        # Look for embedded Dailymotion playlist player (#3822)
+        m = re.search(
+            r'<iframe[^>]+?src=(["\'])(?P<url>(?:https?:)?//(?:www\.)?dailymotion\.[a-z]{2,3}/widget/jukebox\?.+?)\1', webpage)
+        if m:
+            playlists = re.findall(
+                r'list\[\]=/playlist/([^/]+)/', unescapeHTML(m.group('url')))
+            if playlists:
+                return _playlist_from_matches(
+                    playlists, lambda p: '//dailymotion.com/playlist/%s' % p)
+
         # Look for embedded Wistia player
         match = re.search(
             r'<iframe[^>]+?src=(["\'])(?P<url>(?:https?:)?//(?:fast\.)?wistia\.net/embed/iframe/.+?)\1', webpage)
@@ -853,47 +847,51 @@ class GenericIE(InfoExtractor):
         if mobj is not None:
             return self.url_result(mobj.group('url'), 'MLB')
 
+        def check_video(vurl):
+            vpath = compat_urlparse.urlparse(vurl).path
+            vext = determine_ext(vpath)
+            return '.' in vpath and vext not in ('swf', 'png', 'jpg', 'srt', 'sbv', 'sub', 'vtt', 'ttml')
+
+        def filter_video(urls):
+            return list(filter(check_video, urls))
+
         # Start with something easy: JW Player in SWFObject
-        found = re.findall(r'flashvars: [\'"](?:.*&)?file=(http[^\'"&]*)', webpage)
+        found = filter_video(re.findall(r'flashvars: [\'"](?:.*&)?file=(http[^\'"&]*)', webpage))
         if not found:
             # Look for gorilla-vid style embedding
-            found = re.findall(r'''(?sx)
+            found = filter_video(re.findall(r'''(?sx)
                 (?:
                     jw_plugins|
                     JWPlayerOptions|
                     jwplayer\s*\(\s*["'][^'"]+["']\s*\)\s*\.setup
                 )
-                .*?file\s*:\s*["\'](.*?)["\']''', webpage)
+                .*?file\s*:\s*["\'](.*?)["\']''', webpage))
         if not found:
             # Broaden the search a little bit
-            found = re.findall(r'[^A-Za-z0-9]?(?:file|source)=(http[^\'"&]*)', webpage)
+            found = filter_video(re.findall(r'[^A-Za-z0-9]?(?:file|source)=(http[^\'"&]*)', webpage))
         if not found:
             # Broaden the findall a little bit: JWPlayer JS loader
-            found = re.findall(r'[^A-Za-z0-9]?file["\']?:\s*["\'](http(?![^\'"]+\.[0-9]+[\'"])[^\'"]+)["\']', webpage)
+            found = filter_video(re.findall(
+                r'[^A-Za-z0-9]?file["\']?:\s*["\'](http(?![^\'"]+\.[0-9]+[\'"])[^\'"]+)["\']', webpage))
         if not found:
             # Flow player
-            found = re.findall(r'''(?xs)
+            found = filter_video(re.findall(r'''(?xs)
                 flowplayer\("[^"]+",\s*
                     \{[^}]+?\}\s*,
                     \s*{[^}]+? ["']?clip["']?\s*:\s*\{\s*
                         ["']?url["']?\s*:\s*["']([^"']+)["']
-            ''', webpage)
+            ''', webpage))
         if not found:
             # Try to find twitter cards info
-            found = re.findall(r'<meta (?:property|name)="twitter:player:stream" (?:content|value)="(.+?)"', webpage)
+            found = filter_video(re.findall(
+                r'<meta (?:property|name)="twitter:player:stream" (?:content|value)="(.+?)"', webpage))
         if not found:
             # We look for Open Graph info:
             # We have to match any number spaces between elements, some sites try to align them (eg.: statigr.am)
             m_video_type = re.findall(r'<meta.*?property="og:video:type".*?content="video/(.*?)"', webpage)
             # We only look in og:video if the MIME type is a video, don't try if it's a Flash player:
             if m_video_type is not None:
-                def check_video(vurl):
-                    vpath = compat_urlparse.urlparse(vurl).path
-                    vext = determine_ext(vpath)
-                    return '.' in vpath and vext not in ('swf', 'png', 'jpg')
-                found = list(filter(
-                    check_video,
-                    re.findall(r'<meta.*?property="og:video".*?content="(.*?)"', webpage)))
+                found = filter_video(re.findall(r'<meta.*?property="og:video".*?content="(.*?)"', webpage))
         if not found:
             # HTML5 video
             found = re.findall(r'(?s)<video[^<]*(?:>.*?<source[^>]+)? src="([^"]+)"', webpage)