[youtube] Fix extraction (closes #20758, closes #20759, closes #20761, closes #20762...

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

diff --git a/youtube_dl/extractor/pornhub.py b/youtube_dl/extractor/pornhub.py
index 428324ef0b72366d4700372f826ad9c7c0fb9cb6..bf8f0be886eba04cad55e18d65dc3f3d1179d95d 100644 (file)
--- a/youtube_dl/extractor/pornhub.py
+++ b/youtube_dl/extractor/pornhub.py
@@ -14,6 +14,7 @@ from ..compat import (
 )
 from .openload import PhantomJSwrapper
 from ..utils import (
+    determine_ext,
     ExtractorError,
     int_or_none,
     orderedSet,
@@ -275,6 +276,10 @@ class PornHubIE(PornHubBaseIE):
                     r'/(\d{6}/\d{2})/', video_url, 'upload data', default=None)
                 if upload_date:
                     upload_date = upload_date.replace('/', '')
+            if determine_ext(video_url) == 'mpd':
+                formats.extend(self._extract_mpd_formats(
+                    video_url, video_id, mpd_id='dash', fatal=False))
+                continue
             tbr = None
             mobj = re.search(r'(?P<height>\d+)[pP]?_(?P<tbr>\d+)[kK]', video_url)
             if mobj:
@@ -302,17 +307,12 @@ class PornHubIE(PornHubBaseIE):
         comment_count = self._extract_count(
             r'All Comments\s*<span>\(([\d,.]+)\)', webpage, 'comment')
 
-        def _get_items(class_name):
+        def extract_list(meta_key):
             div = self._search_regex(
-                r'<div class="' + class_name + '">([\S\s]+?)</div>',
-                webpage, class_name, default=None)
+                r'(?s)<div[^>]+\bclass=["\'].*?\b%sWrapper[^>]*>(.+?)</div>'
+                % meta_key, webpage, meta_key, default=None)
             if div:
-                return [a for a in re.findall(r'<a href=[^>]+>([^<]+)', div)]
-            else:
-                return None
-
-        categories = _get_items('categoriesWrapper')
-        tags = _get_items('tagsWrapper')
+                return re.findall(r'<a[^>]+\bhref=[^>]+>([^<]+)', div)
 
         return {
             'id': video_id,
@@ -327,8 +327,8 @@ class PornHubIE(PornHubBaseIE):
             'comment_count': comment_count,
             'formats': formats,
             'age_limit': 18,
-            'tags': tags,
-            'categories': categories,
+            'tags': extract_list('tags'),
+            'categories': extract_list('categories'),
             'subtitles': subtitles,
         }
 
@@ -337,7 +337,7 @@ class PornHubPlaylistBaseIE(PornHubBaseIE):
     def _extract_entries(self, webpage, host):
         # Only process container div with main playlist content skipping
         # drop-down menu that uses similar pattern for videos (see
-        # https://github.com/rg3/youtube-dl/issues/11594).
+        # https://github.com/ytdl-org/youtube-dl/issues/11594).
         container = self._search_regex(
             r'(?s)(<div[^>]+class=["\']container.+)', webpage,
             'container', default=webpage)