[extractor/youtube] Extract concurrent view count for livestreams (#5152)

[yt-dlp.git] / yt_dlp / extractor / common.py

diff --git a/yt_dlp/extractor/common.py b/yt_dlp/extractor/common.py
index 30042d61feade16ca9bc48859576903f4b990c44..31a45b37a22ffd3992e7408c3e29185157f261f2 100644 (file)
--- a/yt_dlp/extractor/common.py
+++ b/yt_dlp/extractor/common.py
@@ -22,6 +22,7 @@
 
 from ..compat import functools  # isort: split
 from ..compat import compat_etree_fromstring, compat_expanduser, compat_os_name
+from ..cookies import LenientSimpleCookie
 from ..downloader import FileDownloader
 from ..downloader.f4m import get_base_url, remove_encrypted_media
 from ..utils import (
@@ -283,6 +284,7 @@ class InfoExtractor:
                     captions instead of normal subtitles
     duration:       Length of the video in seconds, as an integer or float.
     view_count:     How many users have watched the video on the platform.
+    concurrent_view_count: How many users are currently watching the video on the platform.
     like_count:     Number of positive ratings of the video
     dislike_count:  Number of negative ratings of the video
     repost_count:   Number of reposts of the video
@@ -1226,7 +1228,7 @@ def _search_regex(self, pattern, string, name, default=NO_DEFAULT, fatal=True, f
             return None
 
     def _search_json(self, start_pattern, string, name, video_id, *, end_pattern='',
-                     contains_pattern='(?s:.+)', fatal=True, default=NO_DEFAULT, **kwargs):
+                     contains_pattern=r'{(?s:.+)}', fatal=True, default=NO_DEFAULT, **kwargs):
         """Searches string for the JSON object specified by start_pattern"""
         # NB: end_pattern is only used to reduce the size of the initial match
         if default is NO_DEFAULT:
@@ -1235,7 +1237,7 @@ def _search_json(self, start_pattern, string, name, video_id, *, end_pattern='',
             fatal, has_default = False, True
 
         json_string = self._search_regex(
-            rf'{start_pattern}\s*(?P<json>{{\s*{contains_pattern}\s*}})\s*{end_pattern}',
+            rf'(?:{start_pattern})\s*(?P<json>{contains_pattern})\s*(?:{end_pattern})',
             string, name, group='json', fatal=fatal, default=None if has_default else NO_DEFAULT)
         if not json_string:
             return default
@@ -1535,10 +1537,10 @@ def extract_chapter_information(e):
                 info['chapters'] = chapters
 
         def extract_video_object(e):
-            assert is_type(e, 'VideoObject')
             author = e.get('author')
             info.update({
                 'url': url_or_none(e.get('contentUrl')),
+                'ext': mimetype2ext(e.get('encodingFormat')),
                 'title': unescapeHTML(e.get('name')),
                 'description': unescapeHTML(e.get('description')),
                 'thumbnails': [{'url': unescapeHTML(url)}
@@ -1551,12 +1553,19 @@ def extract_video_object(e):
                 # however some websites are using 'Text' type instead.
                 # 1. https://schema.org/VideoObject
                 'uploader': author.get('name') if isinstance(author, dict) else author if isinstance(author, str) else None,
+                'artist': traverse_obj(e, ('byArtist', 'name'), expected_type=str),
                 'filesize': int_or_none(float_or_none(e.get('contentSize'))),
                 'tbr': int_or_none(e.get('bitrate')),
                 'width': int_or_none(e.get('width')),
                 'height': int_or_none(e.get('height')),
                 'view_count': int_or_none(e.get('interactionCount')),
+                'tags': try_call(lambda: e.get('keywords').split(',')),
             })
+            if is_type(e, 'AudioObject'):
+                info.update({
+                    'vcodec': 'none',
+                    'abr': int_or_none(e.get('bitrate')),
+                })
             extract_interaction_statistic(e)
             extract_chapter_information(e)
 
@@ -1607,7 +1616,7 @@ def traverse_json_ld(json_ld, at_top_level=True):
                         extract_video_object(e['video'][0])
                     elif is_type(traverse_obj(e, ('subjectOf', 0)), 'VideoObject'):
                         extract_video_object(e['subjectOf'][0])
-                elif is_type(e, 'VideoObject'):
+                elif is_type(e, 'VideoObject', 'AudioObject'):
                     extract_video_object(e)
                     if expected_type is None:
                         continue
@@ -1854,7 +1863,7 @@ def add_item(field, reverse, closest, limit_text):
                     alias, field = field, self._get_field_setting(field, 'field')
                     if self._get_field_setting(alias, 'deprecated'):
                         self.ydl.deprecated_feature(f'Format sorting alias {alias} is deprecated and may '
-                                                    'be removed in a future version. Please use {field} instead')
+                                                    f'be removed in a future version. Please use {field} instead')
                 reverse = match.group('reverse') is not None
                 closest = match.group('separator') == '~'
                 limit_text = match.group('limit')
@@ -3116,9 +3125,10 @@ def _parse_ism_formats_and_subtitles(self, ism_doc, ism_url, ism_id=None):
             stream_name = stream.get('Name')
             stream_language = stream.get('Language', 'und')
             for track in stream.findall('QualityLevel'):
-                fourcc = track.get('FourCC') or ('AACL' if track.get('AudioTag') == '255' else None)
+                KNOWN_TAGS = {'255': 'AACL', '65534': 'EC-3'}
+                fourcc = track.get('FourCC') or KNOWN_TAGS.get(track.get('AudioTag'))
                 # TODO: add support for WVC1 and WMAP
-                if fourcc not in ('H264', 'AVC1', 'AACL', 'TTML'):
+                if fourcc not in ('H264', 'AVC1', 'AACL', 'TTML', 'EC-3'):
                     self.report_warning('%s is not a supported codec' % fourcc)
                     continue
                 tbr = int(track.attrib['Bitrate']) // 1000
@@ -3578,7 +3588,8 @@ def _parse_jwplayer_formats(self, jwplayer_sources_data, video_id=None,
                     'url': source_url,
                     'width': int_or_none(source.get('width')),
                     'height': height,
-                    'tbr': int_or_none(source.get('bitrate')),
+                    'tbr': int_or_none(source.get('bitrate'), scale=1000),
+                    'filesize': int_or_none(source.get('filesize')),
                     'ext': ext,
                 }
                 if source_url.startswith('rtmp'):
@@ -3632,7 +3643,7 @@ def _set_cookie(self, domain, name, value, expire_time=None, port=None,
 
     def _get_cookies(self, url):
         """ Return a http.cookies.SimpleCookie with the cookies for the url """
-        return http.cookies.SimpleCookie(self._downloader._calc_cookies(url))
+        return LenientSimpleCookie(self._downloader._calc_cookies(url))
 
     def _apply_first_set_cookie_header(self, url_handle, cookie):
         """
@@ -3856,8 +3867,10 @@ def _yes_playlist(self, playlist_id, video_id, smuggled_data=None, *, playlist_l
         return True
 
     def _error_or_warning(self, err, _count=None, _retries=0, *, fatal=True):
-        RetryManager.report_retry(err, _count or int(fatal), _retries, info=self.to_screen, warn=self.report_warning,
-                                  sleep_func=self.get_param('retry_sleep_functions', {}).get('extractor'))
+        RetryManager.report_retry(
+            err, _count or int(fatal), _retries,
+            info=self.to_screen, warn=self.report_warning, error=None if fatal else self.report_warning,
+            sleep_func=self.get_param('retry_sleep_functions', {}).get('extractor'))
 
     def RetryManager(self, **kwargs):
         return RetryManager(self.get_param('extractor_retries', 3), self._error_or_warning, **kwargs)