Update the changelog.

[youtubedl] / youtube_dl / extractor / mtv.py

diff --git a/youtube_dl/extractor/mtv.py b/youtube_dl/extractor/mtv.py
index 302c9bf35bc6bb533c10f35ea11cd500012bba0a..dd06395891ac02ce81b40efd471776cb080d328d 100644 (file)
--- a/youtube_dl/extractor/mtv.py
+++ b/youtube_dl/extractor/mtv.py
@@ -4,18 +4,21 @@ import re
 
 from .common import InfoExtractor
 from ..compat import (
-    compat_urllib_parse,
-    compat_urllib_request,
+    compat_urllib_parse_urlencode,
     compat_str,
+    compat_xpath,
 )
 from ..utils import (
     ExtractorError,
     find_xpath_attr,
     fix_xml_ampersands,
+    float_or_none,
     HEADRequest,
+    sanitized_Request,
     unescapeHTML,
     url_basename,
     RegexNotFoundError,
+    xpath_text,
 )
 
 
@@ -53,7 +56,7 @@ class MTVServicesInfoExtractor(InfoExtractor):
 
     def _extract_mobile_video_formats(self, mtvn_id):
         webpage_url = self._MOBILE_TEMPLATE % mtvn_id
-        req = compat_urllib_request.Request(webpage_url)
+        req = sanitized_Request(webpage_url)
         # Otherwise we get a webpage that would execute some javascript
         req.add_header('User-Agent', 'curl/7')
         webpage = self._download_webpage(req, mtvn_id,
@@ -82,9 +85,10 @@ class MTVServicesInfoExtractor(InfoExtractor):
                 rtmp_video_url = rendition.find('./src').text
                 if rtmp_video_url.endswith('siteunavail.png'):
                     continue
+                new_url = self._transform_rtmp_url(rtmp_video_url)
                 formats.append({
-                    'ext': ext,
-                    'url': self._transform_rtmp_url(rtmp_video_url),
+                    'ext': 'flv' if new_url.startswith('rtmp') else ext,
+                    'url': new_url,
                     'format_id': rendition.get('bitrate'),
                     'width': int(rendition.get('width')),
                     'height': int(rendition.get('height')),
@@ -110,7 +114,8 @@ class MTVServicesInfoExtractor(InfoExtractor):
         uri = itemdoc.find('guid').text
         video_id = self._id_from_uri(uri)
         self.report_extraction(video_id)
-        mediagen_url = itemdoc.find('%s/%s' % (_media_xml_tag('group'), _media_xml_tag('content'))).attrib['url']
+        content_el = itemdoc.find('%s/%s' % (_media_xml_tag('group'), _media_xml_tag('content')))
+        mediagen_url = content_el.attrib['url']
         # Remove the templates, like &device={device}
         mediagen_url = re.sub(r'&[^=]*?={.*?}(?=(&|$))', '', mediagen_url)
         if 'acceptMethods' not in mediagen_url:
@@ -128,11 +133,7 @@ class MTVServicesInfoExtractor(InfoExtractor):
             message += item.text
             raise ExtractorError(message, expected=True)
 
-        description_node = itemdoc.find('description')
-        if description_node is not None:
-            description = description_node.text.strip()
-        else:
-            description = None
+        description = xpath_text(itemdoc, 'description')
 
         title_el = None
         if title_el is None:
@@ -140,9 +141,9 @@ class MTVServicesInfoExtractor(InfoExtractor):
                 itemdoc, './/{http://search.yahoo.com/mrss/}category',
                 'scheme', 'urn:mtvn:video_title')
         if title_el is None:
-            title_el = itemdoc.find('.//{http://search.yahoo.com/mrss/}title')
+            title_el = itemdoc.find(compat_xpath('.//{http://search.yahoo.com/mrss/}title'))
         if title_el is None:
-            title_el = itemdoc.find('.//title') or itemdoc.find('./title')
+            title_el = itemdoc.find(compat_xpath('.//title'))
             if title_el.text is None:
                 title_el = None
 
@@ -165,16 +166,19 @@ class MTVServicesInfoExtractor(InfoExtractor):
             'id': video_id,
             'thumbnail': self._get_thumbnail_url(uri, itemdoc),
             'description': description,
+            'duration': float_or_none(content_el.attrib.get('duration')),
         }
 
+    def _get_feed_query(self, uri):
+        data = {'uri': uri}
+        if self._LANG:
+            data['lang'] = self._LANG
+        return compat_urllib_parse_urlencode(data)
+
     def _get_videos_info(self, uri):
         video_id = self._id_from_uri(uri)
         feed_url = self._get_feed_url(uri)
-        data = compat_urllib_parse.urlencode({'uri': uri})
-        info_url = feed_url + '?'
-        if self._LANG:
-            info_url += 'lang=%s&' % self._LANG
-        info_url += data
+        info_url = feed_url + '?' + self._get_feed_query(uri)
         return self._get_videos_info_from_url(info_url, video_id)
 
     def _get_videos_info_from_url(self, url, video_id):
@@ -184,9 +188,7 @@ class MTVServicesInfoExtractor(InfoExtractor):
         return self.playlist_result(
             [self._get_video_info(item) for item in idoc.findall('.//item')])
 
-    def _real_extract(self, url):
-        title = url_basename(url)
-        webpage = self._download_webpage(url, title)
+    def _extract_mgid(self, webpage):
         try:
             # the url can be http://media.mtvnservices.com/fb/{mgid}.swf
             # or http://media.mtvnservices.com/{mgid}
@@ -207,7 +209,12 @@ class MTVServicesInfoExtractor(InfoExtractor):
                 'sm4:video:embed', webpage, 'sm4 embed', default='')
             mgid = self._search_regex(
                 r'embed/(mgid:.+?)["\'&?/]', sm4_embed, 'mgid')
+        return mgid
 
+    def _real_extract(self, url):
+        title = url_basename(url)
+        webpage = self._download_webpage(url, title)
+        mgid = self._extract_mgid(webpage)
         videos_info = self._get_videos_info(mgid)
         return videos_info