[youtube] Add :ytnotifications extractor (#3347)

Authored by: krichbanana

README.md

@@ -79,7 +79,7 @@ yt-dlp is a [youtube-dl](https://github.com/ytdl-org/youtube-dl) fork based on t
 * **Merged with animelover1984/youtube-dl**: You get most of the features and improvements from [animelover1984/youtube-dl](https://github.com/animelover1984/youtube-dl) including `--write-comments`, `BiliBiliSearch`, `BilibiliChannel`, Embedding thumbnail in mp4/ogg/opus, playlist infojson etc. Note that the NicoNico livestreams are not available. See [#31](https://github.com/yt-dlp/yt-dlp/pull/31) for details.
 
 * **Youtube improvements**:
-    * All Feeds (`:ytfav`, `:ytwatchlater`, `:ytsubs`, `:ythistory`, `:ytrec`) and private playlists supports downloading multiple pages of content
+    * All Feeds (`:ytfav`, `:ytwatchlater`, `:ytsubs`, `:ythistory`, `:ytrec`, `:ytnotif`) and private playlists supports downloading multiple pages of content
     * Search (`ytsearch:`, `ytsearchdate:`), search URLs and in-channel search works
     * Mixes supports downloading multiple pages of content
     * Some (but not all) age-gated content can be downloaded without cookies

yt_dlp/extractor/extractors.py

@@ -2100,6 +2100,7 @@ from .youtube import (
     YoutubeIE,
     YoutubeClipIE,
     YoutubeFavouritesIE,
+    YoutubeNotificationsIE,
     YoutubeHistoryIE,
     YoutubeTabIE,
     YoutubeLivestreamEmbedIE,

yt_dlp/extractor/youtube.py

@@ -5526,6 +5526,95 @@ class YoutubeFavouritesIE(YoutubeBaseInfoExtractor):
             ie=YoutubeTabIE.ie_key())
 
 
+class YoutubeNotificationsIE(YoutubeTabBaseInfoExtractor):
+    IE_NAME = 'youtube:notif'
+    IE_DESC = 'YouTube notifications; ":ytnotif" keyword (requires cookies)'
+    _VALID_URL = r':ytnotif(?:ication)?s?'
+    _LOGIN_REQUIRED = True
+    _TESTS = [{
+        'url': ':ytnotif',
+        'only_matching': True,
+    }, {
+        'url': ':ytnotifications',
+        'only_matching': True,
+    }]
+
+    def _extract_notification_menu(self, response, continuation_list):
+        notification_list = traverse_obj(
+            response,
+            ('actions', 0, 'openPopupAction', 'popup', 'multiPageMenuRenderer', 'sections', 0, 'multiPageMenuNotificationSectionRenderer', 'items'),
+            ('actions', 0, 'appendContinuationItemsAction', 'continuationItems'),
+            expected_type=list) or []
+        continuation_list[0] = None
+        for item in notification_list:
+            entry = self._extract_notification_renderer(item.get('notificationRenderer'))
+            if entry:
+                yield entry
+            continuation = item.get('continuationItemRenderer')
+            if continuation:
+                continuation_list[0] = continuation
+
+    def _extract_notification_renderer(self, notification):
+        video_id = traverse_obj(
+            notification, ('navigationEndpoint', 'watchEndpoint', 'videoId'), expected_type=str)
+        url = f'https://www.youtube.com/watch?v={video_id}'
+        channel_id = None
+        if not video_id:
+            browse_ep = traverse_obj(
+                notification, ('navigationEndpoint', 'browseEndpoint'), expected_type=dict)
+            channel_id = traverse_obj(browse_ep, 'browseId', expected_type=str)
+            post_id = self._search_regex(
+                r'/post/(.+)', traverse_obj(browse_ep, 'canonicalBaseUrl', expected_type=str),
+                'post id', default=None)
+            if not channel_id or not post_id:
+                return
+            # The direct /post url redirects to this in the browser
+            url = f'https://www.youtube.com/channel/{channel_id}/community?lb={post_id}'
+
+        channel = traverse_obj(
+            notification, ('contextualMenu', 'menuRenderer', 'items', 1, 'menuServiceItemRenderer', 'text', 'runs', 1, 'text'),
+            expected_type=str)
+        title = self._search_regex(
+            rf'{re.escape(channel)} [^:]+: (.+)', self._get_text(notification, 'shortMessage'),
+            'video title', default=None)
+        if title:
+            title = title.replace('\xad', '')  # remove soft hyphens
+        upload_date = (strftime_or_none(self._extract_time_text(notification, 'sentTimeText')[0], '%Y%m%d')
+                       if self._configuration_arg('approximate_date', ie_key=YoutubeTabIE.ie_key())
+                       else None)
+        return {
+            '_type': 'url',
+            'url': url,
+            'ie_key': (YoutubeIE if video_id else YoutubeTabIE).ie_key(),
+            'video_id': video_id,
+            'title': title,
+            'channel_id': channel_id,
+            'channel': channel,
+            'thumbnails': self._extract_thumbnails(notification, 'videoThumbnail'),
+            'upload_date': upload_date,
+        }
+
+    def _notification_menu_entries(self, ytcfg):
+        continuation_list = [None]
+        response = None
+        for page in itertools.count(1):
+            ctoken = traverse_obj(
+                continuation_list, (0, 'continuationEndpoint', 'getNotificationMenuEndpoint', 'ctoken'), expected_type=str)
+            response = self._extract_response(
+                item_id=f'page {page}', query={'ctoken': ctoken} if ctoken else {}, ytcfg=ytcfg,
+                ep='notification/get_notification_menu', check_get_keys='actions',
+                headers=self.generate_api_headers(ytcfg=ytcfg, visitor_data=self._extract_visitor_data(response)))
+            yield from self._extract_notification_menu(response, continuation_list)
+            if not continuation_list[0]:
+                break
+
+    def _real_extract(self, url):
+        display_id = 'notifications'
+        ytcfg = self._download_ytcfg('web', display_id) if not self.skip_webpage else {}
+        self._report_playlist_authcheck(ytcfg)
+        return self.playlist_result(self._notification_menu_entries(ytcfg), display_id, display_id)
+
+
 class YoutubeSearchIE(YoutubeTabBaseInfoExtractor, SearchInfoExtractor):
     IE_DESC = 'YouTube search'
     IE_NAME = 'youtube:search'