X-Git-Url: http://git.bitcoin.ninja/index.cgi?a=blobdiff_plain;f=youtube_dl%2Fextractor%2Fbrightcove.py;h=558b3d0093975eff273d38a7a2bad9f1d4e36355;hb=8e4f824365543e394286742fcdb4c0a548becc8e;hp=68ee5292b61157b5bccbd78619a3da758fdf93dc;hpb=3fa95508373bb5813099c2e4ccad95638a506916;p=youtube-dl

diff --git a/youtube_dl/extractor/brightcove.py b/youtube_dl/extractor/brightcove.py
index 68ee5292b..558b3d009 100644
--- a/youtube_dl/extractor/brightcove.py
+++ b/youtube_dl/extractor/brightcove.py
@@ -1,3 +1,5 @@
+# encoding: utf-8
+
 import re
 import json
 import xml.etree.ElementTree
@@ -5,15 +7,41 @@ import xml.etree.ElementTree
 from .common import InfoExtractor
 from ..utils import (
     compat_urllib_parse,
+    find_xpath_attr,
+    compat_urlparse,
+
+    ExtractorError,
 )
 
 class BrightcoveIE(InfoExtractor):
     _VALID_URL = r'https?://.*brightcove\.com/(services|viewer).*\?(?P<query>.*)'
     _FEDERATED_URL_TEMPLATE = 'http://c.brightcove.com/services/viewer/htmlFederated?%s'
     _PLAYLIST_URL_TEMPLATE = 'http://c.brightcove.com/services/json/experience/runtime/?command=get_programming_for_experience&playerKey=%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
+
+    _TESTS = [
+        {
+            # From http://www.8tv.cat/8aldia/videos/xavier-sala-i-martin-aquesta-tarda-a-8-al-dia/
+            u'url': u'http://c.brightcove.com/services/viewer/htmlFederated?playerID=1654948606001&flashID=myExperience&%40videoPlayer=2371591881001',
+            u'file': u'2371591881001.mp4',
+            u'md5': u'9e80619e0a94663f0bdc849b4566af19',
+            u'note': u'Test Brightcove downloads and detection in GenericIE',
+            u'info_dict': {
+                u'title': u'Xavier Sala i Martín: “Un banc que no presta és un banc zombi que no serveix per a res”',
+                u'uploader': u'8TV',
+                u'description': u'md5:a950cc4285c43e44d763d036710cd9cd',
+            }
+        },
+        {
+            # From http://medianetwork.oracle.com/video/player/1785452137001
+            u'url': u'http://c.brightcove.com/services/viewer/htmlFederated?playerID=1217746023001&flashID=myPlayer&%40videoPlayer=1785452137001',
+            u'file': u'1785452137001.flv',
+            u'info_dict': {
+                u'title': u'JVMLS 2012: Arrays 2.0 - Opportunities and Challenges',
+                u'description': u'John Rose speaks at the JVM Language Summit, August 1, 2012.',
+                u'uploader': u'Oracle',
+            },
+        },
+    ]
 
     @classmethod
     def _build_brighcove_url(cls, object_str):
@@ -24,13 +52,13 @@ class BrightcoveIE(InfoExtractor):
         object_doc = xml.etree.ElementTree.fromstring(object_str)
         assert u'BrightcoveExperience' in object_doc.attrib['class']
         params = {'flashID': object_doc.attrib['id'],
-                  'playerID': object_doc.find('./param[@name="playerID"]').attrib['value'],
+                  'playerID': find_xpath_attr(object_doc, './param', 'name', 'playerID').attrib['value'],
                   }
-        playerKey = object_doc.find('./param[@name="playerKey"]')
+        playerKey = find_xpath_attr(object_doc, './param', 'name', 'playerKey')
         # Not all pages define this value
         if playerKey is not None:
             params['playerKey'] = playerKey.attrib['value']
-        videoPlayer = object_doc.find('./param[@name="@videoPlayer"]')
+        videoPlayer = find_xpath_attr(object_doc, './param', 'name', '@videoPlayer')
         if videoPlayer is not None:
             params['@videoPlayer'] = videoPlayer.attrib['value']
         data = compat_urllib_parse.urlencode(params)
@@ -38,15 +66,15 @@ class BrightcoveIE(InfoExtractor):
 
     def _real_extract(self, url):
         mobj = re.match(self._VALID_URL, url)
-        query = mobj.group('query')
+        query_str = mobj.group('query')
+        query = compat_urlparse.parse_qs(query_str)
 
-        m_video_id = re.search(r'videoPlayer=(\d+)', query)
-        if m_video_id is not None:
-            video_id = m_video_id.group(1)
-            return self._get_video_info(video_id, query)
+        videoPlayer = query.get('@videoPlayer')
+        if videoPlayer:
+            return self._get_video_info(videoPlayer[0], query_str)
         else:
-            player_key = self._search_regex(r'playerKey=(.+?)(&|$)', query, 'playlist_id')
-            return self._get_playlist_info(player_key)
+            player_key = query['playerKey']
+            return self._get_playlist_info(player_key[0])
 
     def _get_video_info(self, video_id, query):
         request_url = self._FEDERATED_URL_TEMPLATE % query
@@ -70,15 +98,27 @@ class BrightcoveIE(InfoExtractor):
                                     playlist_title=playlist_info['mediaCollectionDTO']['displayName'])
 
     def _extract_video_info(self, video_info):
-        renditions = video_info['renditions']
-        renditions = sorted(renditions, key=lambda r: r['size'])
-        best_format = renditions[-1]
+        info = {
+            'id': video_info['id'],
+            'title': video_info['displayName'],
+            'description': video_info.get('shortDescription'),
+            'thumbnail': video_info.get('videoStillURL') or video_info.get('thumbnailURL'),
+            'uploader': video_info.get('publisherName'),
+        }
 
-        return {'id': video_info['id'],
-                'title': video_info['displayName'],
-                'url': best_format['defaultURL'], 
+        renditions = video_info.get('renditions')
+        if renditions:
+            renditions = sorted(renditions, key=lambda r: r['size'])
+            best_format = renditions[-1]
+            info.update({
+                'url': best_format['defaultURL'],
                 'ext': 'mp4',
-                'description': video_info.get('shortDescription'),
-                'thumbnail': video_info.get('videoStillURL') or video_info.get('thumbnailURL'),
-                'uploader': video_info.get('publisherName'),
-                }
+            })
+        elif video_info.get('FLVFullLengthURL') is not None:
+            info.update({
+                'url': video_info['FLVFullLengthURL'],
+                'ext': 'flv',
+            })
+        else:
+            raise ExtractorError(u'Unable to extract video url for %s' % info['id'])
+        return info