debian/NEWS: Update with notes about split audio/video downloads.

[youtubedl] / youtube_dl / extractor / brightcove.py

diff --git a/youtube_dl/extractor/brightcove.py b/youtube_dl/extractor/brightcove.py
index 71e3c7883338154eea0c3d369a0fdd0bee828e26..1392f382a24c273604f0c67db7afafefbcec85b8 100644 (file)
--- 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
 import re
 import json
 import xml.etree.ElementTree


@@ -7,15 +9,39 @@ from ..utils import (
     compat_urllib_parse,
     find_xpath_attr,
     compat_urlparse,
     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'
 )
 
 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):
 
     @classmethod
     def _build_brighcove_url(cls, object_str):


@@ -23,6 +49,13 @@ class BrightcoveIE(InfoExtractor):
         Build a Brightcove url from a xml string containing
         <object class="BrightcoveExperience">{params}</object>
         """
         Build a Brightcove url from a xml string containing
         <object class="BrightcoveExperience">{params}</object>
         """

+
+        # Fix up some stupid HTML, see https://github.com/rg3/youtube-dl/issues/1553
+        object_str = re.sub(r'(<param name="[^"]+" value="[^"]+")>',
+                            lambda m: m.group(1) + '/>', object_str)
+        # Fix up some stupid XML, see https://github.com/rg3/youtube-dl/issues/1608
+        object_str = object_str.replace(u'<--', u'<!--')
+

         object_doc = xml.etree.ElementTree.fromstring(object_str)
         assert u'BrightcoveExperience' in object_doc.attrib['class']
         params = {'flashID': object_doc.attrib['id'],
         object_doc = xml.etree.ElementTree.fromstring(object_str)
         assert u'BrightcoveExperience' in object_doc.attrib['class']
         params = {'flashID': object_doc.attrib['id'],


@@ -65,22 +98,37 @@ class BrightcoveIE(InfoExtractor):
         playlist_info = self._download_webpage(self._PLAYLIST_URL_TEMPLATE % player_key,
                                                player_key, u'Downloading playlist information')
 
         playlist_info = self._download_webpage(self._PLAYLIST_URL_TEMPLATE % player_key,
                                                player_key, u'Downloading playlist information')
 

-        playlist_info = json.loads(playlist_info)['videoList']
+        json_data = json.loads(playlist_info)
+        if 'videoList' not in json_data:
+            raise ExtractorError(u'Empty playlist')
+        playlist_info = json_data['videoList']

         videos = [self._extract_video_info(video_info) for video_info in playlist_info['mediaCollectionDTO']['videoDTOs']]
 
         return self.playlist_result(videos, playlist_id=playlist_info['id'],
                                     playlist_title=playlist_info['mediaCollectionDTO']['displayName'])
 
     def _extract_video_info(self, video_info):
         videos = [self._extract_video_info(video_info) for video_info in playlist_info['mediaCollectionDTO']['videoDTOs']]
 
         return self.playlist_result(videos, playlist_id=playlist_info['id'],
                                     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',
                 '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