Merge tag 'upstream/2015.02.06'

[youtubedl] / youtube_dl / extractor / brightcove.py

diff --git a/youtube_dl/extractor/brightcove.py b/youtube_dl/extractor/brightcove.py
index ad22cbafd6a541599c289cfd7dcb1b27ec7a8e95..ea0969d4d259a99653bebbcabcebb0e1f87719f3 100644 (file)
--- a/youtube_dl/extractor/brightcove.py
+++ b/youtube_dl/extractor/brightcove.py
@@ -6,24 +6,26 @@ import json
 import xml.etree.ElementTree
 
 from .common import InfoExtractor
 import xml.etree.ElementTree
 
 from .common import InfoExtractor

-from ..utils import (
-    compat_urllib_parse,
-    find_xpath_attr,
-    fix_xml_ampersands,
-    compat_urlparse,
+from ..compat import (
+    compat_parse_qs,

     compat_str,
     compat_str,

+    compat_urllib_parse,
+    compat_urllib_parse_urlparse,

     compat_urllib_request,
     compat_urllib_request,

-    compat_parse_qs,
-
+    compat_urlparse,
+)
+from ..utils import (

     determine_ext,
     ExtractorError,
     determine_ext,
     ExtractorError,

-    unsmuggle_url,
+    find_xpath_attr,
+    fix_xml_ampersands,

     unescapeHTML,
     unescapeHTML,

+    unsmuggle_url,

 )
 
 
 class BrightcoveIE(InfoExtractor):
 )
 
 
 class BrightcoveIE(InfoExtractor):

-    _VALID_URL = r'https?://.*brightcove\.com/(services|viewer).*\?(?P<query>.*)'
+    _VALID_URL = r'(?:https?://.*brightcove\.com/(services|viewer).*?\?|brightcove:)(?P<query>.*)'

     _FEDERATED_URL_TEMPLATE = 'http://c.brightcove.com/services/viewer/htmlFederated?%s'
 
     _TESTS = [
     _FEDERATED_URL_TEMPLATE = 'http://c.brightcove.com/services/viewer/htmlFederated?%s'
 
     _TESTS = [


@@ -106,10 +108,12 @@ class BrightcoveIE(InfoExtractor):
         """
 
         # Fix up some stupid HTML, see https://github.com/rg3/youtube-dl/issues/1553
         """
 
         # Fix up some stupid HTML, see https://github.com/rg3/youtube-dl/issues/1553

-        object_str = re.sub(r'(<param name="[^"]+" value="[^"]+")>',
+        object_str = re.sub(r'(<param(?:\s+[a-zA-Z0-9_]+="[^"]*")*)>',

                             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('<--', '<!--')
                             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('<--', '<!--')

+        # remove namespace to simplify extraction
+        object_str = re.sub(r'(<object[^>]*)(xmlns=".*?")', r'\1', object_str)

         object_str = fix_xml_ampersands(object_str)
 
         object_doc = xml.etree.ElementTree.fromstring(object_str.encode('utf-8'))
         object_str = fix_xml_ampersands(object_str)
 
         object_doc = xml.etree.ElementTree.fromstring(object_str.encode('utf-8'))


@@ -218,7 +222,7 @@ class BrightcoveIE(InfoExtractor):
         webpage = self._download_webpage(req, video_id)
 
         error_msg = self._html_search_regex(
         webpage = self._download_webpage(req, video_id)
 
         error_msg = self._html_search_regex(

-            r"<h1>We're sorry.</h1>\s*<p>(.*?)</p>", webpage,
+            r"<h1>We're sorry.</h1>([\s\n]*<p>.*?</p>)+", webpage,

             'error message', default=None)
         if error_msg is not None:
             raise ExtractorError(
             'error message', default=None)
         if error_msg is not None:
             raise ExtractorError(


@@ -260,12 +264,21 @@ class BrightcoveIE(InfoExtractor):
             formats = []
             for rend in renditions:
                 url = rend['defaultURL']
             formats = []
             for rend in renditions:
                 url = rend['defaultURL']

+                if not url:
+                    continue
+                ext = None

                 if rend['remote']:
                 if rend['remote']:

-                    # This type of renditions are served through akamaihd.net,
-                    # but they don't use f4m manifests
-                    url = url.replace('control/', '') + '?&v=3.3.0&fp=13&r=FEEFJ&g=RTSJIMBMPFPB'
-                    ext = 'flv'
-                else:
+                    url_comp = compat_urllib_parse_urlparse(url)
+                    if url_comp.path.endswith('.m3u8'):
+                        formats.extend(
+                            self._extract_m3u8_formats(url, info['id'], 'mp4'))
+                        continue
+                    elif 'akamaihd.net' in url_comp.netloc:
+                        # This type of renditions are served through
+                        # akamaihd.net, but they don't use f4m manifests
+                        url = url.replace('control/', '') + '?&v=3.3.0&fp=13&r=FEEFJ&g=RTSJIMBMPFPB'
+                        ext = 'flv'
+                if ext is None:

                     ext = determine_ext(url)
                 size = rend.get('size')
                 formats.append({
                     ext = determine_ext(url)
                 size = rend.get('size')
                 formats.append({