[Porn91] Update extractor

youtube_dl/extractor/porn91.py

@@ -1,11 +1,21 @@
 # coding: utf-8
 from __future__ import unicode_literals
 
+import re
+
 from .common import InfoExtractor
+from ..compat import (
+    compat_urllib_parse_unquote,
+)
 from ..utils import (
+    clean_html,
+    extract_attributes,
+    get_elements_by_class,
     parse_duration,
-    int_or_none,
     ExtractorError,
+    str_to_int,
+    strip_or_none,
+    unified_strdate,
 )
 
 
@@ -15,13 +25,15 @@ class Porn91IE(InfoExtractor):
 
     _TEST = {
         'url': 'http://91porn.com/view_video.php?viewkey=7e42283b4f5ab36da134',
-        'md5': '7fcdb5349354f40d41689bd0fa8db05a',
+        'md5': 'd869db281402e0ef4ddef3c38b866f86',
         'info_dict': {
             'id': '7e42283b4f5ab36da134',
             'title': '18岁大一漂亮学妹，水嫩性感，再爽一次！',
             'ext': 'mp4',
             'duration': 431,
             'age_limit': 18,
+            'upload_date': '20150520',
+            'uploader': '千岁九王爷',
         }
     }
 
@@ -35,29 +47,44 @@ class Porn91IE(InfoExtractor):
         if '作为游客，你每天只可观看10个视频' in webpage:
             raise ExtractorError('91 Porn says: Daily limit 10 videos exceeded', expected=True)
 
-        title = self._search_regex(
+        title = self._html_search_regex(
-            r'<div id="viewvideo-title">([^<]+)</div>', webpage, 'title')
+            r'(?s)<title\b[^>]*>\s*(.+?)\s*(?:Chinese\s+homemade\s+video\s*)?</title', webpage, 'title')
-        title = title.replace('\n', '')
 
         video_link_url = self._search_regex(
-            r'<textarea[^>]+id=["\']fm-video_link[^>]+>([^<]+)</textarea>',
+            r'''document\s*\.\s*write\s*\(\s*strencode2\s*\((?P<q>"|')(?P<enc_str>[%\da-fA-F]+)(?P=q)\s*\)''',
-            webpage, 'video link')
+            webpage, 'video link', group='enc_str')
-        videopage = self._download_webpage(video_link_url, video_id)
+        video_link_url = compat_urllib_parse_unquote(video_link_url)
 
-        info_dict = self._parse_html5_media_entries(url, videopage, video_id)[0]
+        info_dict = self._parse_html5_media_entries(url, '<video>%s</video>' % (video_link_url, ), video_id)[0]
 
-        duration = parse_duration(self._search_regex(
+        FIELD_MAP = {
-            r'时长:\s*</span>\s*(\d+:\d+)', webpage, 'duration', fatal=False))
+            '时长': ('duration', parse_duration, 'Runtime', ),
+            '查看': ('view_count', str_to_int, 'Views', ),
+            '留言': ('comment_count', str_to_int, 'Comments', ),
+            '收藏': ('like_count', str_to_int, 'Favorites', ),
+            '添加时间': ('upload_date', unified_strdate, 'Added', ),
+            # same as title for en, not description for cn_CN
+            '__ignore__': ('description', strip_or_none, 'Description', ),
+            '作者': ('uploader', strip_or_none, 'From', ),
+        }
+        for elt in get_elements_by_class('info', re.sub(r'</span>\s*<span\b[^>]*?>', '', webpage)) or []:
+            elt = re.split(r':\s*', clean_html(elt), 1)
+            if len(elt) != 2 or elt[1] == '':
+                continue
+            parm = FIELD_MAP.get(elt[0].strip())
+            if parm and elt[1] is not None:
+                info_dict[parm[0]] = parm[1](elt[1]) if parm[1] else elt[1]
 
-        comment_count = int_or_none(self._search_regex(
+        thumbnail = extract_attributes(
-            r'留言:\s*</span>\s*(\d+)', webpage, 'comment count', fatal=False))
+            self._search_regex(
+                r'''(<video\b[^>]*\bid\s*=\s*(?P<q>"|')player_one(?P=q)[^>]*>)''',
+                webpage, 'poster', default='')).get('poster')
 
         info_dict.update({
             'id': video_id,
             'title': title,
-            'duration': duration,
-            'comment_count': comment_count,
             'age_limit': self._rta_search(webpage),
+            'thumbnail': thumbnail,
         })
 
         return info_dict