Merge tag 'upstream/2014.06.07'

[youtubedl] / youtube_dl / extractor / mtv.py

diff --git a/youtube_dl/extractor/mtv.py b/youtube_dl/extractor/mtv.py
index 5447b6c0cab098b895eda0e9f2b3b266fb65a7b0..e5ca41b4091698ad2a180f9d2ca00b4b96218c1e 100644 (file)
--- a/youtube_dl/extractor/mtv.py
+++ b/youtube_dl/extractor/mtv.py
@@ -5,9 +5,12 @@ import re
 from .common import InfoExtractor
 from ..utils import (
     compat_urllib_parse,
 from .common import InfoExtractor
 from ..utils import (
     compat_urllib_parse,

+    compat_urllib_request,

     ExtractorError,
     find_xpath_attr,
     fix_xml_ampersands,
     ExtractorError,
     find_xpath_attr,
     fix_xml_ampersands,

+    HEADRequest,
+    unescapeHTML,

     url_basename,
     RegexNotFoundError,
 )
     url_basename,
     RegexNotFoundError,
 )


@@ -18,6 +21,7 @@ def _media_xml_tag(tag):
 
 
 class MTVServicesInfoExtractor(InfoExtractor):
 
 
 class MTVServicesInfoExtractor(InfoExtractor):

+    _MOBILE_TEMPLATE = None

     @staticmethod
     def _id_from_uri(uri):
         return uri.split(':')[-1]
     @staticmethod
     def _id_from_uri(uri):
         return uri.split(':')[-1]


@@ -39,9 +43,29 @@ class MTVServicesInfoExtractor(InfoExtractor):
         else:
             return thumb_node.attrib['url']
 
         else:
             return thumb_node.attrib['url']
 

-    def _extract_video_formats(self, mdoc):
-        if re.match(r'.*/error_country_block\.swf$', mdoc.find('.//src').text) is not None:
-            raise ExtractorError('This video is not available from your country.', expected=True)
+    def _extract_mobile_video_formats(self, mtvn_id):
+        webpage_url = self._MOBILE_TEMPLATE % mtvn_id
+        req = compat_urllib_request.Request(webpage_url)
+        # Otherwise we get a webpage that would execute some javascript
+        req.add_header('Youtubedl-user-agent', 'curl/7')
+        webpage = self._download_webpage(req, mtvn_id,
+            'Downloading mobile page')
+        metrics_url = unescapeHTML(self._search_regex(r'<a href="(http://metrics.+?)"', webpage, 'url'))
+        req = HEADRequest(metrics_url)
+        response = self._request_webpage(req, mtvn_id, 'Resolving url')
+        url = response.geturl()
+        # Transform the url to get the best quality:
+        url = re.sub(r'.+pxE=mp4', 'http://mtvnmobile.vo.llnwd.net/kip0/_pxn=0+_pxK=18639+_pxE=mp4', url, 1)
+        return [{'url': url,'ext': 'mp4'}]
+
+    def _extract_video_formats(self, mdoc, mtvn_id):
+        if re.match(r'.*/(error_country_block\.swf|geoblock\.mp4)$', mdoc.find('.//src').text) is not None:
+            if mtvn_id is not None and self._MOBILE_TEMPLATE is not None:
+                self.to_screen('The normal version is not available from your '
+                    'country, trying with the mobile version')
+                return self._extract_mobile_video_formats(mtvn_id)
+            raise ExtractorError('This video is not available from your country.',
+                expected=True)

 
         formats = []
         for rendition in mdoc.findall('.//rendition'):
 
         formats = []
         for rendition in mdoc.findall('.//rendition'):


@@ -56,6 +80,7 @@ class MTVServicesInfoExtractor(InfoExtractor):
                                 })
             except (KeyError, TypeError):
                 raise ExtractorError('Invalid rendition field.')
                                 })
             except (KeyError, TypeError):
                 raise ExtractorError('Invalid rendition field.')

+        self._sort_formats(formats)

         return formats
 
     def _get_video_info(self, itemdoc):
         return formats
 
     def _get_video_info(self, itemdoc):


@@ -94,9 +119,16 @@ class MTVServicesInfoExtractor(InfoExtractor):
             raise ExtractorError('Could not find video title')
         title = title.strip()
 
             raise ExtractorError('Could not find video title')
         title = title.strip()
 

+        # This a short id that's used in the webpage urls
+        mtvn_id = None
+        mtvn_id_node = find_xpath_attr(itemdoc, './/{http://search.yahoo.com/mrss/}category',
+                'scheme', 'urn:mtvn:id')
+        if mtvn_id_node is not None:
+            mtvn_id = mtvn_id_node.text
+

         return {
             'title': title,
         return {
             'title': title,

-            'formats': self._extract_video_formats(mediagen_doc),
+            'formats': self._extract_video_formats(mediagen_doc, mtvn_id),

             'id': video_id,
             'thumbnail': self._get_thumbnail_url(uri, itemdoc),
             'description': description,
             'id': video_id,
             'thumbnail': self._get_thumbnail_url(uri, itemdoc),
             'description': description,