Start work on parti extractor

yt_dlp/extractor/_extractors.py

@@ -1492,6 +1492,7 @@ from .paramountplus import (
 )
 from .parler import ParlerIE
 from .parlview import ParlviewIE
+from .parti import PartiIE
 from .patreon import (
     PatreonCampaignIE,
     PatreonIE,

yt_dlp/extractor/parti.py

@@ -0,0 +1,128 @@
+import datetime
+
+from ..utils import (
+    int_or_none,
+    traverse_obj,
+)
+from .common import InfoExtractor
+
+
+class PartiIE(InfoExtractor):
+    IE_DESC = 'Download a stream from parti.com'
+    _VALID_URL = r'https://parti\.com/creator/(parti|discord|telegram)/(?P<id>[\w-]+)'
+    _TESTS = [
+        {
+            'url': 'https://parti.com/creator/parti/ItZTMGG',
+            'info_dict': {
+                'id': 'ItZTMGG',
+                'ext': 'mp4',
+                'title': str,
+                'description': str,
+                'upload_date': str,
+                'is_live': False,
+            },
+            'params': {'skip_download': 'm3u8'},
+        },
+    ]
+    _CREATOR_API = 'https://api-backend.parti.com/parti_v2/profile/get_user_by_social_media/parti'
+    _GET_LIVESTREAM_API = 'https://api-backend.parti.com/parti_v2/profile/get_livestream_channel_info'
+    _GET_USER_FEED_API = 'https://api-backend.parti.com/parti_v2/profile/user_profile_feed/'
+    _RECORDING_BASE_URL = 'https://watch.parti.com'
+    _PLAYBACK_VERSION = '1.17.0'
+
+    def _get_creator_id(self, creator):
+        """ The creator ID is a number returned as plain text """
+        url = self._CREATOR_API + '/' + creator
+        page = self._download_webpage(url, None, 'Fetching creator id')
+        return str(page)
+
+    def _get_live_playback_data(self, creator_id):
+        """ If the stream is live, we can use this URL to download. """
+        url = self._GET_LIVESTREAM_API + '/' + creator_id
+        data = self._download_json(url, None, 'Fetching user profile feed')
+        if not data:
+            raise Exception('No data!')
+
+        is_live = False
+        viewer_count = 0
+        if 'stream' in data:
+            is_live = data['stream'] is not None
+            if is_live:
+                viewer_count = data['stream']['viewer_count']
+
+        channel = data['channel_info']['channel']
+        auth_token = channel['playback_auth_token']
+        base_url = channel['playback_url']
+
+        extracted = traverse_obj(data, {
+            'base_url': ('channel_info', 'channel', 'playback_url', {str}),
+            'auth_token': ('channel_info', 'channel', 'playback_auth_token', {str}),
+            'viewer_count': ('channel_info', 'stream', 'viewer_count', {int_or_none}),
+            'is_live': ('channel_info', 'stream', {lambda x: x is not None}),
+            })
+
+        base_url = extracted['base_url']
+        auth_token = extracted['auth_token']
+        url = None
+        if base_url and auth_token:
+            url = f'{base_url}?token={auth_token}&player_version={self._PLAYBACK_VERSION}'
+
+        return {
+            'url': url,
+            **extracted,
+        }
+
+    def _get_user_feed(self, creator_id):
+        """ The user feed are VODs listed below the main stream """
+        url = self._GET_USER_FEED_API + '/' + creator_id + '?limit=10'
+        vods = self._download_json(url, None, 'Fetching user feed')
+        if not vods:
+            raise Exception('No vods found!')
+        return list(vods)
+
+    def _download_vod(self, url, creator, creator_id):
+        """ Download the VOD visible on the creators feed """
+        feed = self._get_user_feed(creator_id)
+        vod = feed[0]
+        vod_url = self._RECORDING_BASE_URL + '/' + vod['livestream_recording']
+        created_at = datetime.date.fromtimestamp(vod['created_at'])
+        upload_date = str(created_at).replace('-', '')
+
+        formats = self._extract_m3u8_formats(vod_url, creator, 'mp4', live=False)
+        return {
+            'id': creator,
+            'url': url,
+            'title': f'{creator}\'s Parti VOD - {upload_date}',
+            'description': vod['post_content'],
+            'upload_date': upload_date,
+            'is_live': False,
+            'formats': formats,
+        }
+
+    def _download_livestream(self, url, creator, stream_url):
+        """ Download a currently active livestream """
+        formats = self._extract_m3u8_formats(stream_url, creator, 'mp4', live=True)
+
+        created_at = datetime.datetime.now()
+        upload_date = str(created_at.date()).replace('-', '')
+        pretty_timestamp = str(created_at).replace(':', '_')
+        return {
+            'id': creator,
+            'url': url,
+            'title': f'{creator}\'s Parti Live - {pretty_timestamp}',
+            'description': f'A livestream from {pretty_timestamp}',
+            'upload_date': upload_date,
+            'is_live': True,
+            'formats': formats,
+        }
+
+    def _real_extract(self, url):
+        creator = self._match_id(url)
+
+        creator_id = self._get_creator_id(creator)
+        playback_data = self._get_live_playback_data(creator_id)
+        if not playback_data['is_live']:
+            return self._download_vod(url, creator, creator_id)
+        else:
+            return self._download_livestream(url, creator, playback_data['url'])
+