GenericIE: Detect videos from Brightcove

Brightcove videos info is usually found in an <object class="BrightcoveExperience"></object> node, this is passed to a new method of BrightcoveIE that builds a url to extract the video.

1 files changed, 28 insertions, 1 deletions

diff --git a/youtube_dl/extractor/brightcove.py b/youtube_dl/extractor/brightcove.py
index f85acbb5d..bf5916493 100644
--- a/youtube_dl/extractor/brightcove.py
+++ b/youtube_dl/extractor/brightcove.py
@@ -1,17 +1,44 @@
 import re
 import json
+import xml.etree.ElementTree
 
 from .common import InfoExtractor
+from ..utils import (
+    compat_urllib_parse,
+)
 
 class BrightcoveIE(InfoExtractor):
     _VALID_URL = r'http://.*brightcove\.com/.*\?(?P<query>.*videoPlayer=(?P<id>\d*).*)'
+    _FEDERATED_URL_TEMPLATE = 'http://c.brightcove.com/services/viewer/htmlFederated?%s'
+    
+    # There is a test for Brigtcove in GenericIE, that way we test both the download
+    # and the detection of videos, and we don't have to find an URL that is always valid
+
+    @classmethod
+    def _build_brighcove_url(cls, object_str):
+        """
+        Build a Brightcove url from a xml string containing
+        <object class="BrightcoveExperience">{params}</object>
+        """
+        object_doc = xml.etree.ElementTree.fromstring(object_str)
+        assert object_doc.attrib['class'] == u'BrightcoveExperience'
+        params = {'flashID': object_doc.attrib['id'],
+                  'playerID': object_doc.find('./param[@name="playerID"]').attrib['value'],
+                  '@videoPlayer': object_doc.find('./param[@name="@videoPlayer"]').attrib['value'],
+                  }
+        playerKey = object_doc.find('./param[@name="playerKey"]')
+        # Not all pages define this value
+        if playerKey is not None:
+            params['playerKey'] = playerKey.attrib['value']
+        data = compat_urllib_parse.urlencode(params)
+        return cls._FEDERATED_URL_TEMPLATE % data
 
     def _real_extract(self, url):
         mobj = re.match(self._VALID_URL, url)
         query = mobj.group('query')
         video_id = mobj.group('id')
 
-        request_url = 'http://c.brightcove.com/services/viewer/htmlFederated?%s' % query
+        request_url = self._FEDERATED_URL_TEMPLATE % query
         webpage = self._download_webpage(request_url, video_id)
 
         self.report_extraction(video_id)