[youtube] Fix extraction.

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

diff --git a/youtube_dl/extractor/sexu.py b/youtube_dl/extractor/sexu.py
index 45b33b3eafb01b32d6b2b6f70edce552e633f233..3df51520b06e8bc2da1b3afa3fee75db4e21be46 100644 (file)
--- a/youtube_dl/extractor/sexu.py
+++ b/youtube_dl/extractor/sexu.py
@@ -1,7 +1,5 @@
 from __future__ import unicode_literals
 
-import re
-
 from .common import InfoExtractor
 
 
@@ -14,38 +12,42 @@ class SexuIE(InfoExtractor):
             'id': '961791',
             'ext': 'mp4',
             'title': 'md5:4d05a19a5fc049a63dbbaf05fb71d91b',
-            'description': 'md5:c5ed8625eb386855d5a7967bd7b77a54',
+            'description': 'md5:2b75327061310a3afb3fbd7d09e2e403',
             'categories': list,  # NSFW
-            'thumbnail': 're:https?://.*\.jpg$',
+            'thumbnail': r're:https?://.*\.jpg$',
             'age_limit': 18,
         }
     }
 
     def _real_extract(self, url):
-        mobj = re.match(self._VALID_URL, url)
-        video_id = mobj.group('id')
-
+        video_id = self._match_id(url)
         webpage = self._download_webpage(url, video_id)
 
-        quality_arr = self._search_regex(r'sources:\s*\[([^\]]+)\]', webpage, 'quality formats')
+        jwvideo = self._parse_json(
+            self._search_regex(r'\.setup\(\s*({.+?})\s*\);', webpage, 'jwvideo'),
+            video_id)
+
+        sources = jwvideo['sources']
 
         formats = [{
-            'url': fmt[0].replace('\\', ''),
-            'format_id': fmt[1],
-            'height': int(fmt[1][:3]),
-        } for fmt in re.findall(r'"file":"([^"]+)","label":"([^"]+)"', quality_arr)]
+            'url': source['file'].replace('\\', ''),
+            'format_id': source.get('label'),
+            'height': int(self._search_regex(
+                r'^(\d+)[pP]', source.get('label', ''), 'height',
+                default=None)),
+        } for source in sources if source.get('file')]
         self._sort_formats(formats)
 
         title = self._html_search_regex(
-            r'<title>([^<]+)\s*-\s*Sexu.Com</title>', webpage, 'title')
+            r'<title>([^<]+)\s*-\s*Sexu\.Com</title>', webpage, 'title')
 
-        description = self._html_search_meta('description', webpage, 'description')
+        description = self._html_search_meta(
+            'description', webpage, 'description')
 
-        thumbnail = self._html_search_regex(
-            r'image:\s*"([^"]+)"',
-            webpage, 'thumbnail', fatal=False)
+        thumbnail = jwvideo.get('image')
 
-        categories_str = self._html_search_meta('keywords', webpage, 'categories', fatal=False)
+        categories_str = self._html_search_meta(
+            'keywords', webpage, 'categories')
         categories = (
             None if categories_str is None
             else categories_str.split(','))