Merge remote-tracking branch 'jaimeMF/yt-playlists'

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

diff --git a/youtube_dl/extractor/generic.py b/youtube_dl/extractor/generic.py
index 76d3692736b58f77261d836b78e5a634c596cf2f..0b5f2b2bb92907c5489cca3c63643bf8f80dd4a9 100644 (file)
--- a/youtube_dl/extractor/generic.py
+++ b/youtube_dl/extractor/generic.py
@@ -54,6 +54,23 @@ class GenericIE(InfoExtractor):
             },
             u'skip': u'There is a limit of 200 free downloads / month for the test song',
         },
+        # embedded brightcove video
+        # it also tests brightcove videos that need to set the 'Referer' in the
+        # http requests
+        {
+            u'add_ie': ['Brightcove'],
+            u'url': u'http://www.bfmtv.com/video/bfmbusiness/cours-bourse/cours-bourse-l-analyse-technique-154522/',
+            u'info_dict': {
+                u'id': u'2765128793001',
+                u'ext': u'mp4',
+                u'title': u'Le cours de bourse : l’analyse technique',
+                u'description': u'md5:7e9ad046e968cb2d1114004aba466fd9',
+                u'uploader': u'BFM BUSINESS',
+            },
+            u'params': {
+                u'skip_download': True,
+            },
+        },
     ]
 
     def report_download_webpage(self, video_id):
@@ -145,11 +162,20 @@ class GenericIE(InfoExtractor):
             raise ExtractorError(u'Failed to download URL: %s' % url)
 
         self.report_extraction(video_id)
+
+        # it's tempting to parse this further, but you would
+        # have to take into account all the variations like
+        #   Video Title - Site Name
+        #   Site Name | Video Title
+        #   Video Title - Tagline | Site Name
+        # and so on and so forth; it's just not practical
+        video_title = self._html_search_regex(r'<title>(.*)</title>',
+            webpage, u'video title', default=u'video', flags=re.DOTALL)
+
         # Look for BrightCove:
-        m_brightcove = re.search(r'<object[^>]+?class=([\'"])[^>]*?BrightcoveExperience.*?\1.+?</object>', webpage, re.DOTALL)
-        if m_brightcove is not None:
+        bc_url = BrightcoveIE._extract_brightcove_url(webpage)
+        if bc_url is not None:
             self.to_screen(u'Brightcove video detected.')
-            bc_url = BrightcoveIE._build_brighcove_url(m_brightcove.group())
             return self.url_result(bc_url, 'Brightcove')
 
         # Look for embedded Vimeo player
@@ -161,17 +187,20 @@ class GenericIE(InfoExtractor):
             return self.url_result(surl, 'Vimeo')
 
         # Look for embedded YouTube player
-        mobj = re.search(
-            r'<iframe[^>]+?src=(["\'])(?P<url>https?://(?:www\.)?youtube.com/embed/.+?)\1', webpage)
-        if mobj:
-            surl = unescapeHTML(mobj.group(u'url'))
-            return self.url_result(surl, 'Youtube')
+        matches = re.findall(
+            r'<iframe[^>]+?src=(["\'])(?P<url>(?:https?:)?//(?:www\.)?youtube.com/embed/.+?)\1', webpage)
+        if matches:
+            urlrs = [self.url_result(unescapeHTML(tuppl[1]), 'Youtube')
+                     for tuppl in matches]
+            return self.playlist_result(
+                urlrs, playlist_id=video_id, playlist_title=video_title)
 
         # Look for Bandcamp pages with custom domain
         mobj = re.search(r'<meta property="og:url"[^>]*?content="(.*?bandcamp\.com.*?)"', webpage)
         if mobj is not None:
             burl = unescapeHTML(mobj.group(1))
-            return self.url_result(burl, 'Bandcamp')
+            # Don't set the extractor because it can be a track url or an album
+            return self.url_result(burl)
 
         # Start with something easy: JW Player in SWFObject
         mobj = re.search(r'flashvars: [\'"](?:.*&)?file=(http[^\'"&]*)', webpage)
@@ -210,15 +239,6 @@ class GenericIE(InfoExtractor):
         video_extension = os.path.splitext(video_id)[1][1:]
         video_id = os.path.splitext(video_id)[0]
 
-        # it's tempting to parse this further, but you would
-        # have to take into account all the variations like
-        #   Video Title - Site Name
-        #   Site Name | Video Title
-        #   Video Title - Tagline | Site Name
-        # and so on and so forth; it's just not practical
-        video_title = self._html_search_regex(r'<title>(.*)</title>',
-            webpage, u'video title', default=u'video', flags=re.DOTALL)
-
         # video uploader is domain name
         video_uploader = self._search_regex(r'(?:https?://)?([^/]*)/.*',
             url, u'video uploader')