[imdb] Fix duplicated entries bug

[youtube-dl] / youtube_dl / extractor / yahoo.py

diff --git a/youtube_dl/extractor/yahoo.py b/youtube_dl/extractor/yahoo.py
index 617e3bb0641caf2452feb033be4d368b1f13b08a..e17a39782bd2e674855dff8a5ec3112bd40158c6 100644 (file)
--- a/youtube_dl/extractor/yahoo.py
+++ b/youtube_dl/extractor/yahoo.py
@@ -6,8 +6,8 @@ from .common import InfoExtractor, SearchInfoExtractor
 from ..utils import (
     compat_urllib_parse,
     compat_urlparse,
-    determine_ext,
     clean_html,
+    int_or_none,
 )
 
 
@@ -17,27 +17,21 @@ class YahooIE(InfoExtractor):
     _TESTS = [
         {
             u'url': u'http://screen.yahoo.com/julian-smith-travis-legg-watch-214727115.html',
-            u'file': u'214727115.flv',
+            u'file': u'214727115.mp4',
+            u'md5': u'4962b075c08be8690a922ee026d05e69',
             u'info_dict': {
                 u'title': u'Julian Smith & Travis Legg Watch Julian Smith',
                 u'description': u'Julian and Travis watch Julian Smith',
             },
-            u'params': {
-                # Requires rtmpdump
-                u'skip_download': True,
-            },
         },
         {
             u'url': u'http://screen.yahoo.com/wired/codefellas-s1-ep12-cougar-lies-103000935.html',
-            u'file': u'103000935.flv',
+            u'file': u'103000935.mp4',
+            u'md5': u'd6e6fc6e1313c608f316ddad7b82b306',
             u'info_dict': {
                 u'title': u'Codefellas - The Cougar Lies with Spanish Moss',
                 u'description': u'Agent Topple\'s mustache does its dirty work, and Nicole brokers a deal for peace. But why is the NSA collecting millions of Instagram brunch photos? And if your waffles have nothing to hide, what are they so worried about?',
             },
-            u'params': {
-                # Requires rtmpdump
-                u'skip_download': True,
-            },
         },
     ]
 
@@ -53,8 +47,12 @@ class YahooIE(InfoExtractor):
         # The 'meta' field is not always in the video webpage, we request it
         # from another page
         long_id = info['id']
+        return self._get_info(long_id, video_id)
+
+    def _get_info(self, long_id, video_id):
         query = ('SELECT * FROM yahoo.media.video.streams WHERE id="%s"'
-                 ' AND plrs="86Gj0vCaSzV_Iuf6hNylf2"' % long_id)
+                 ' AND plrs="86Gj0vCaSzV_Iuf6hNylf2" AND region="US"'
+                 ' AND protocol="http"' % long_id)
         data = compat_urllib_parse.urlencode({
             'q': query,
             'env': 'prod',
@@ -70,9 +68,9 @@ class YahooIE(InfoExtractor):
         formats = []
         for s in info['streams']:
             format_info = {
-                'width': s.get('width'),
-                'height': s.get('height'),
-                'bitrate': s.get('bitrate'),
+                'width': int_or_none(s.get('width')),
+                'height': int_or_none(s.get('height')),
+                'tbr': int_or_none(s.get('bitrate')),
             }
 
             host = s['host']
@@ -86,10 +84,10 @@ class YahooIE(InfoExtractor):
             else:
                 format_url = compat_urlparse.urljoin(host, path)
                 format_info['url'] = format_url
-                format_info['ext'] = determine_ext(format_url)
                 
             formats.append(format_info)
-        formats = sorted(formats, key=lambda f:(f['height'], f['width']))
+
+        self._sort_formats(formats)
 
         return {
             'id': video_id,
@@ -100,6 +98,32 @@ class YahooIE(InfoExtractor):
         }
 
 
+class YahooNewsIE(YahooIE):
+    IE_NAME = 'yahoo:news'
+    _VALID_URL = r'http://news\.yahoo\.com/video/.*?-(?P<id>\d*?)\.html'
+
+    _TEST = {
+        u'url': u'http://news.yahoo.com/video/china-moses-crazy-blues-104538833.html',
+        u'md5': u'67010fdf3a08d290e060a4dd96baa07b',
+        u'info_dict': {
+            u'id': u'104538833',
+            u'ext': u'mp4',
+            u'title': u'China Moses Is Crazy About the Blues',
+            u'description': u'md5:9900ab8cd5808175c7b3fe55b979bed0',
+        },
+    }
+
+    # Overwrite YahooIE properties we don't want
+    _TESTS = []
+
+    def _real_extract(self, url):
+        mobj = re.match(self._VALID_URL, url)
+        video_id = mobj.group('id')
+        webpage = self._download_webpage(url, video_id)
+        long_id = self._search_regex(r'contentId: \'(.+?)\',', webpage, u'long id')
+        return self._get_info(long_id, video_id)
+
+
 class YahooSearchIE(SearchInfoExtractor):
     IE_DESC = u'Yahoo screen search'
     _MAX_RESULTS = 1000