Merge branch 'atomicparsley' (closes #2436)

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

diff --git a/youtube_dl/extractor/brightcove.py b/youtube_dl/extractor/brightcove.py
index 717e151d93b6d0717c65cddf91c63e84da2dc0ae..3c02c297a58a32cf536e8ccc972dea68021f650b 100644 (file)
--- a/youtube_dl/extractor/brightcove.py
+++ b/youtube_dl/extractor/brightcove.py
@@ -17,6 +17,7 @@ from ..utils import (
 
     ExtractorError,
     unsmuggle_url,
+    unescapeHTML,
 )
 
 
@@ -69,17 +70,6 @@ class BrightcoveIE(InfoExtractor):
                 'description': 'md5:363109c02998fee92ec02211bd8000df',
                 'uploader': 'National Ballet of Canada',
             },
-        },
-        {
-            # https://github.com/rg3/youtube-dl/issues/2253
-            'url': 'http://v.thestar.com/services/player/bcpid2071349530001?bckey=AQ~~,AAAAuO4KaJE~,gatFNwSKdGDmDpIYqNJ-fTHn_c4z_LH_&bctid=3101154703001',
-            'file': '3101154703001.mp4',
-            'md5': '0ba9446db037002366bab3b3eb30c88c',
-            'info_dict': {
-                'title': 'Still no power',
-                'uploader': 'thestar.com',
-                'description': 'Mississauga resident David Farmer is still out of power as a result of the ice storm a month ago. To keep the house warm, Farmer cuts wood from his property for a wood burning stove downstairs.',
-            }
         }
     ]
 
@@ -97,7 +87,7 @@ class BrightcoveIE(InfoExtractor):
         object_str = object_str.replace('<--', '<!--')
         object_str = fix_xml_ampersands(object_str)
 
-        object_doc = xml.etree.ElementTree.fromstring(object_str)
+        object_doc = xml.etree.ElementTree.fromstring(object_str.encode('utf-8'))
 
         fv_el = find_xpath_attr(object_doc, './param', 'name', 'flashVars')
         if fv_el is not None:
@@ -138,25 +128,32 @@ class BrightcoveIE(InfoExtractor):
 
     @classmethod
     def _extract_brightcove_url(cls, webpage):
-        """Try to extract the brightcove url from the wepbage, returns None
+        """Try to extract the brightcove url from the webpage, returns None
         if it can't be found
         """
+        urls = cls._extract_brightcove_urls(webpage)
+        return urls[0] if urls else None
+
+    @classmethod
+    def _extract_brightcove_urls(cls, webpage):
+        """Return a list of all Brightcove URLs from the webpage """
 
         url_m = re.search(r'<meta\s+property="og:video"\s+content="(http://c.brightcove.com/[^"]+)"', webpage)
         if url_m:
-            return url_m.group(1)
+            url = unescapeHTML(url_m.group(1))
+            # Some sites don't add it, we can't download with this url, for example:
+            # http://www.ktvu.com/videos/news/raw-video-caltrain-releases-video-of-man-almost/vCTZdY/
+            if 'playerKey' in url:
+                return [url]
 
-        m_brightcove = re.search(
+        matches = re.findall(
             r'''(?sx)<object
             (?:
-                [^>]+?class=([\'"])[^>]*?BrightcoveExperience.*?\1 |
+                [^>]+?class=[\'"][^>]*?BrightcoveExperience.*?[\'"] |
                 [^>]*?>\s*<param\s+name="movie"\s+value="https?://[^/]*brightcove\.com/
             ).+?</object>''',
             webpage)
-        if m_brightcove is not None:
-            return cls._build_brighcove_url(m_brightcove.group())
-        else:
-            return None
+        return [cls._build_brighcove_url(m) for m in matches]
 
     def _real_extract(self, url):
         url, smuggled_data = unsmuggle_url(url, {})