[xnxx] Fix extraction (closes #15817)

youtube_dl/extractor/xnxx.py

@@ -1,19 +1,29 @@
 # coding: utf-8
 from __future__ import unicode_literals
 
+import re
+
 from .common import InfoExtractor
-from ..compat import compat_urllib_parse_unquote
+from ..utils import (
+    determine_ext,
+    int_or_none,
+    NO_DEFAULT,
+    str_to_int,
+)
 
 
 class XNXXIE(InfoExtractor):
     _VALID_URL = r'https?://(?:video|www)\.xnxx\.com/video-?(?P<id>[0-9a-z]+)/'
     _TESTS = [{
         'url': 'http://www.xnxx.com/video-55awb78/skyrim_test_video',
-        'md5': 'ef7ecee5af78f8b03dca2cf31341d3a0',
+        'md5': '7583e96c15c0f21e9da3453d9920fbba',
         'info_dict': {
             'id': '55awb78',
-            'ext': 'flv',
+            'ext': 'mp4',
             'title': 'Skyrim Test Video',
+            'thumbnail': r're:^https?://.*\.jpg',
+            'duration': 469,
+            'view_count': int,
             'age_limit': 18,
         },
     }, {
@@ -26,23 +36,49 @@ class XNXXIE(InfoExtractor):
 
     def _real_extract(self, url):
         video_id = self._match_id(url)
+
         webpage = self._download_webpage(url, video_id)
 
-        video_url = self._search_regex(r'flv_url=(.*?)&amp;',
+        def get(meta, default=NO_DEFAULT, fatal=True):
-                                       webpage, 'video URL')
+            return self._search_regex(
-        video_url = compat_urllib_parse_unquote(video_url)
+                r'set%s\s*\(\s*(["\'])(?P<value>(?:(?!\1).)+)\1' % meta,
+                webpage, meta, default=default, fatal=fatal, group='value')
+
+        title = self._og_search_title(
+            webpage, default=None) or get('VideoTitle')
 
-        video_title = self._html_search_regex(r'<title>(.*?)\s+-\s+XNXX.COM',
+        formats = []
-                                              webpage, 'title')
+        for mobj in re.finditer(
+                r'setVideo(?:Url(?P<id>Low|High)|HLS)\s*\(\s*(?P<q>["\'])(?P<url>(?:https?:)?//.+?)(?P=q)', webpage):
+            format_url = mobj.group('url')
+            if determine_ext(format_url) == 'm3u8':
+                formats.extend(self._extract_m3u8_formats(
+                    format_url, video_id, 'mp4', entry_protocol='m3u8_native',
+                    preference=1, m3u8_id='hls', fatal=False))
+            else:
+                format_id = mobj.group('id')
+                if format_id:
+                    format_id = format_id.lower()
+                formats.append({
+                    'url': format_url,
+                    'format_id': format_id,
+                    'quality': -1 if format_id == 'low' else 0,
+                })
+        self._sort_formats(formats)
 
-        video_thumbnail = self._search_regex(r'url_bigthumb=(.*?)&amp;',
+        thumbnail = self._og_search_thumbnail(webpage, default=None) or get(
-                                             webpage, 'thumbnail', fatal=False)
+            'ThumbUrl', fatal=False) or get('ThumbUrl169', fatal=False)
+        duration = int_or_none(self._og_search_property('duration', webpage))
+        view_count = str_to_int(self._search_regex(
+            r'id=["\']nb-views-number[^>]+>([\d,.]+)', webpage, 'view count',
+            default=None))
 
         return {
             'id': video_id,
-            'url': video_url,
+            'title': title,
-            'title': video_title,
+            'thumbnail': thumbnail,
-            'ext': 'flv',
+            'duration': duration,
-            'thumbnail': video_thumbnail,
+            'view_count': view_count,
             'age_limit': 18,
+            'formats': formats,
         }