[yourporn] added support for posts with multiple videos

closes #27554

youtube_dl/extractor/yourporn.py

@@ -30,38 +30,71 @@ class YourPornIE(InfoExtractor):
     }]
 
     def _real_extract(self, url):
-        video_id = self._match_id(url)
+        get_duration = lambda webpage: parse_duration(self._search_regex(
-
-        webpage = self._download_webpage(url, video_id)
-
-        parts = self._parse_json(
-            self._search_regex(
-                r'data-vnfo=(["\'])(?P<data>{.+?})\1', webpage, 'data info',
-                group='data'),
-            video_id)[video_id].split('/')
-
-        num = 0
-        for c in parts[6] + parts[7]:
-            if c.isnumeric():
-                num += int(c)
-        parts[5] = compat_str(int(parts[5]) - num)
-        parts[1] += '8'
-        video_url = urljoin(url, '/'.join(parts))
-
-        title = (self._search_regex(
-            r'<[^>]+\bclass=["\']PostEditTA[^>]+>([^<]+)', webpage, 'title',
-            default=None) or self._og_search_description(webpage)).strip()
-        thumbnail = self._og_search_thumbnail(webpage)
-        duration = parse_duration(self._search_regex(
             r'duration\s*:\s*<[^>]+>([\d:]+)', webpage, 'duration',
             default=None))
 
-        return {
+        # Only for posts containing a single video is the post_id equal to
-            'id': video_id,
+        # video_id. If there a multiple videos there also exists a post with the
-            'url': video_url,
+        # video_id and we use this page to fetch the video title.
-            'title': title,
+        post_id = self._match_id(url)
-            'thumbnail': thumbnail,
+
-            'duration': duration,
+        webpage = self._download_webpage(url, post_id)
-            'age_limit': 18,
+
-            'ext': 'mp4',
+        videos = self._parse_json(
+            self._search_regex(
+                r'data-vnfo=(["\'])(?P<data>{.+?})\1', webpage, 'data info',
+                group='data'),
+            post_id)
+
+        for video_id in videos:
+            parts = videos[video_id].split('/')
+            num = 0
+            for c in parts[6] + parts[7]:
+                if c.isnumeric():
+                    num += int(c)
+            parts[5] = compat_str(int(parts[5]) - num)
+            parts[1] += '8'
+            videos[video_id] = urljoin(url, '/'.join(parts))
+
+        # If there ist only one video in the post (as is most likely) the
+        # video_id is equal to post_id and we save us from re-fetching the page
+        # to obtain the meta data.
+        # This may fail but if needed will be obtained in the next step.
+        titles = {
+            post_id: self._og_search_description(webpage, default=None)
         }
+        thumbnails = {
+            post_id: self._og_search_thumbnail(webpage, default=None)
+        }
+        durations = {
+            post_id: get_duration(webpage)
+        }
+
+        # obtain missing metadata for all videos in the post
+        for video_id in videos:
+            if not titles.get(video_id) or not thumbnails.get(video_id) or video_id not in durations:
+                webpage = self._download_webpage('https://sxyprn.com/post/%s.html' % video_id, video_id)
+            if not titles.get(video_id):
+                titles[video_id] = self._og_search_description(webpage)
+            if not thumbnails.get(video_id):
+                thumbnails[video_id] = self._og_search_thumbnail(webpage)
+            if video_id not in durations:
+                durations[video_id] = get_duration(webpage)
+
+        entries = []
+        for video_id in videos:
+            entries.append({
+                'id': video_id,
+                'url': videos[video_id],
+                'title': titles[video_id],
+                'thumbnail': thumbnails[video_id],
+                'duration': durations[video_id],
+                'age_limit': 18,
+                'ext': 'mp4',
+            })
+
+        if len(entries) == 1:
+            return entries[0]
+        else:
+            return self.playlist_result(entries, post_id, titles[post_id])