[ie/patreon] Fix embedded HLS extraction (#8993)

Closes #8973
Authored by: johnvictorfs

yt_dlp/extractor/ard.py

@@ -8,6 +8,7 @@ from ..utils import (
     determine_ext,
     int_or_none,
     join_nonempty,
+    jwt_decode_hs256,
     make_archive_id,
     parse_duration,
     parse_iso8601,
@@ -238,6 +239,7 @@ class ARDBetaMediathekIE(InfoExtractor):
         (?P<id>[a-zA-Z0-9]+)
         /?(?:[?#]|$)'''
     _GEO_COUNTRIES = ['DE']
+    _TOKEN_URL = 'https://sso.ardmediathek.de/sso/token'
 
     _TESTS = [{
         'url': 'https://www.ardmediathek.de/video/filme-im-mdr/liebe-auf-vier-pfoten/mdr-fernsehen/Y3JpZDovL21kci5kZS9zZW5kdW5nLzI4MjA0MC80MjIwOTEtNDAyNTM0',
@@ -359,12 +361,27 @@ class ARDBetaMediathekIE(InfoExtractor):
 
     def _real_extract(self, url):
         display_id = self._match_id(url)
+        query = {'embedded': 'false', 'mcV6': 'true'}
+        headers = {}
+
+        if self._get_cookies(self._TOKEN_URL).get('ams'):
+            token = self._download_json(
+                self._TOKEN_URL, display_id, 'Fetching token for age verification',
+                'Unable to fetch age verification token', fatal=False)
+            id_token = traverse_obj(token, ('idToken', {str}))
+            decoded_token = traverse_obj(id_token, ({jwt_decode_hs256}, {dict}))
+            user_id = traverse_obj(decoded_token, (('user_id', 'sub'), {str}), get_all=False)
+            if not user_id:
+                self.report_warning('Unable to extract token, continuing without authentication')
+            else:
+                headers['x-authorization'] = f'Bearer {id_token}'
+                query['userId'] = user_id
+                if decoded_token.get('age_rating') != 18:
+                    self.report_warning('Account is not verified as 18+; video may be unavailable')
 
         page_data = self._download_json(
-            f'https://api.ardmediathek.de/page-gateway/pages/ard/item/{display_id}', display_id, query={
-                'embedded': 'false',
-                'mcV6': 'true',
-            })
+            f'https://api.ardmediathek.de/page-gateway/pages/ard/item/{display_id}',
+            display_id, query=query, headers=headers)
 
         # For user convenience we use the old contentId instead of the longer crid
         # Ref: https://github.com/yt-dlp/yt-dlp/issues/8731#issuecomment-1874398283
@@ -383,7 +400,7 @@ class ARDBetaMediathekIE(InfoExtractor):
         media_data = traverse_obj(player_data, ('mediaCollection', 'embedded', {dict}))
 
         if player_data.get('blockedByFsk'):
-            self.raise_no_formats('This video is only available after 22:00', expected=True)
+            self.raise_login_required('This video is only available for age verified users or after 22:00')
 
         formats = []
         subtitles = {}

yt_dlp/extractor/newgrounds.py

@@ -3,15 +3,15 @@ import re
 
 from .common import InfoExtractor
 from ..utils import (
+    OnDemandPagedList,
     clean_html,
     extract_attributes,
     get_element_by_id,
     int_or_none,
     parse_count,
     parse_duration,
+    traverse_obj,
     unified_timestamp,
-    OnDemandPagedList,
-    try_get,
 )
 
 
@@ -263,19 +263,16 @@ class NewgroundsUserIE(InfoExtractor):
     def _fetch_page(self, channel_id, url, page):
         page += 1
         posts_info = self._download_json(
-            f'{url}/page/{page}', channel_id,
+            f'{url}?page={page}', channel_id,
             note=f'Downloading page {page}', headers={
                 'Accept': 'application/json, text/javascript, */*; q = 0.01',
                 'X-Requested-With': 'XMLHttpRequest',
             })
-        sequence = posts_info.get('sequence', [])
-        for year in sequence:
-            posts = try_get(posts_info, lambda x: x['years'][str(year)]['items'])
-            for post in posts:
-                path, media_id = self._search_regex(
-                    r'<a[^>]+\bhref=["\'][^"\']+((?:portal/view|audio/listen)/(\d+))[^>]+>',
-                    post, 'url', group=(1, 2))
-                yield self.url_result(f'https://www.newgrounds.com/{path}', NewgroundsIE.ie_key(), media_id)
+        for post in traverse_obj(posts_info, ('items', ..., ..., {str})):
+            path, media_id = self._search_regex(
+                r'<a[^>]+\bhref=["\'][^"\']+((?:portal/view|audio/listen)/(\d+))[^>]+>',
+                post, 'url', group=(1, 2))
+            yield self.url_result(f'https://www.newgrounds.com/{path}', NewgroundsIE.ie_key(), media_id)
 
     def _real_extract(self, url):
         channel_id = self._match_id(url)

yt_dlp/extractor/patreon.py

@@ -275,7 +275,7 @@ class PatreonIE(PatreonBaseIE):
                     'ext': ext,
                     'url': post_file['url'],
                 }
-            elif name == 'video':
+            elif name == 'video' or determine_ext(post_file.get('url')) == 'm3u8':
                 formats, subtitles = self._extract_m3u8_formats_and_subtitles(post_file['url'], video_id)
                 return {
                     **info,

yt_dlp/extractor/rule34video.py

@@ -18,10 +18,10 @@ from ..utils.traversal import traverse_obj
 
 
 class Rule34VideoIE(InfoExtractor):
-    _VALID_URL = r'https?://(?:www\.)?rule34video\.com/videos/(?P<id>\d+)'
+    _VALID_URL = r'https?://(?:www\.)?rule34video\.com/videos?/(?P<id>\d+)'
     _TESTS = [
         {
-            'url': 'https://rule34video.com/videos/3065157/shot-it-mmd-hmv/',
+            'url': 'https://rule34video.com/video/3065157/shot-it-mmd-hmv/',
             'md5': 'ffccac2c23799dabbd192621ae4d04f3',
             'info_dict': {
                 'id': '3065157',