X-Git-Url: https://jfr.im/git/yt-dlp.git/blobdiff_plain/9c42b7eef547e826e9fcc7beb6706a2523949d05..61edf57f8f13f6dfd81154174e647eb5fdd26089:/yt_dlp/extractor/common.py

diff --git a/yt_dlp/extractor/common.py b/yt_dlp/extractor/common.py
index a36a6187c..f63bd7825 100644
--- a/yt_dlp/extractor/common.py
+++ b/yt_dlp/extractor/common.py
@@ -1,5 +1,6 @@
 import base64
 import collections
+import functools
 import getpass
 import hashlib
 import http.client
@@ -21,7 +22,6 @@
 import urllib.request
 import xml.etree.ElementTree
 
-from ..compat import functools  # isort: split
 from ..compat import (
     compat_etree_fromstring,
     compat_expanduser,
@@ -37,6 +37,7 @@
     IncompleteRead,
     network_exceptions,
 )
+from ..networking.impersonate import ImpersonateTarget
 from ..utils import (
     IDENTITY,
     JSON_LD_RE,
@@ -59,7 +60,6 @@
     determine_ext,
     dict_get,
     encode_data_uri,
-    error_to_compat_str,
     extract_attributes,
     filter_dict,
     fix_xml_ampersands,
@@ -170,12 +170,12 @@ class InfoExtractor:
                                  Automatically calculated from width and height
                     * dynamic_range The dynamic range of the video. One of:
                                  "SDR" (None), "HDR10", "HDR10+, "HDR12", "HLG, "DV"
-                    * tbr        Average bitrate of audio and video in KBit/s
-                    * abr        Average audio bitrate in KBit/s
+                    * tbr        Average bitrate of audio and video in kbps (1000 bits/sec)
+                    * abr        Average audio bitrate in kbps (1000 bits/sec)
                     * acodec     Name of the audio codec in use
                     * asr        Audio sampling rate in Hertz
                     * audio_channels  Number of audio channels
-                    * vbr        Average video bitrate in KBit/s
+                    * vbr        Average video bitrate in kbps (1000 bits/sec)
                     * fps        Frame rate
                     * vcodec     Name of the video codec in use
                     * container  Name of the container format
@@ -234,7 +234,14 @@ class InfoExtractor:
                                  'maybe' if the format may have DRM and has to be tested before download.
                     * extra_param_to_segment_url  A query string to append to each
                                  fragment's URL, or to update each existing query string
-                                 with. Only applied by the native HLS/DASH downloaders.
+                                 with. If it is an HLS stream with an AES-128 decryption key,
+                                 the query paramaters will be passed to the key URI as well,
+                                 unless there is an `extra_param_to_key_url` given,
+                                 or unless an external key URI is provided via `hls_aes`.
+                                 Only applied by the native HLS/DASH downloaders.
+                    * extra_param_to_key_url  A query string to append to the URL
+                                 of the format's HLS AES-128 decryption key.
+                                 Only applied by the native HLS downloader.
                     * hls_aes    A dictionary of HLS AES-128 decryption information
                                  used by the native HLS downloader to override the
                                  values in the media playlist when an '#EXT-X-KEY' tag
@@ -766,8 +773,8 @@ def __maybe_fake_ip_and_retry(self, countries):
             self._x_forwarded_for_ip = GeoUtils.random_ipv4(country_code)
             if self._x_forwarded_for_ip:
                 self.report_warning(
-                    'Video is geo restricted. Retrying extraction with fake IP %s (%s) as X-Forwarded-For.'
-                    % (self._x_forwarded_for_ip, country_code.upper()))
+                    'Video is geo restricted. Retrying extraction with fake IP '
+                    f'{self._x_forwarded_for_ip} ({country_code.upper()}) as X-Forwarded-For.')
                 return True
         return False
 
@@ -818,7 +825,7 @@ def __can_accept_status_code(err, expected_status):
         else:
             return err.status in variadic(expected_status)
 
-    def _create_request(self, url_or_request, data=None, headers=None, query=None):
+    def _create_request(self, url_or_request, data=None, headers=None, query=None, extensions=None):
         if isinstance(url_or_request, urllib.request.Request):
             self._downloader.deprecation_warning(
                 'Passing a urllib.request.Request to _create_request() is deprecated. '
@@ -827,10 +834,11 @@ def _create_request(self, url_or_request, data=None, headers=None, query=None):
         elif not isinstance(url_or_request, Request):
             url_or_request = Request(url_or_request)
 
-        url_or_request.update(data=data, headers=headers, query=query)
+        url_or_request.update(data=data, headers=headers, query=query, extensions=extensions)
         return url_or_request
 
-    def _request_webpage(self, url_or_request, video_id, note=None, errnote=None, fatal=True, data=None, headers=None, query=None, expected_status=None):
+    def _request_webpage(self, url_or_request, video_id, note=None, errnote=None, fatal=True, data=None,
+                         headers=None, query=None, expected_status=None, impersonate=None, require_impersonation=False):
         """
         Return the response handle.
 
@@ -839,7 +847,7 @@ def _request_webpage(self, url_or_request, video_id, note=None, errnote=None, fa
         if not self._downloader._first_webpage_request:
             sleep_interval = self.get_param('sleep_interval_requests') or 0
             if sleep_interval > 0:
-                self.to_screen('Sleeping %s seconds ...' % sleep_interval)
+                self.to_screen(f'Sleeping {sleep_interval} seconds ...')
                 time.sleep(sleep_interval)
         else:
             self._downloader._first_webpage_request = False
@@ -861,8 +869,31 @@ def _request_webpage(self, url_or_request, video_id, note=None, errnote=None, fa
             headers = (headers or {}).copy()
             headers.setdefault('X-Forwarded-For', self._x_forwarded_for_ip)
 
+        extensions = {}
+
+        if impersonate in (True, ''):
+            impersonate = ImpersonateTarget()
+        requested_targets = [
+            t if isinstance(t, ImpersonateTarget) else ImpersonateTarget.from_str(t)
+            for t in variadic(impersonate)
+        ] if impersonate else []
+
+        available_target = next(filter(self._downloader._impersonate_target_available, requested_targets), None)
+        if available_target:
+            extensions['impersonate'] = available_target
+        elif requested_targets:
+            message = 'The extractor is attempting impersonation, but '
+            message += (
+                'no impersonate target is available' if not str(impersonate)
+                else f'none of these impersonate targets are available: "{", ".join(map(str, requested_targets))}"')
+            info_msg = ('see  https://github.com/yt-dlp/yt-dlp#impersonation  '
+                        'for information on installing the required dependencies')
+            if require_impersonation:
+                raise ExtractorError(f'{message}; {info_msg}', expected=True)
+            self.report_warning(f'{message}; if you encounter errors, then {info_msg}', only_once=True)
+
         try:
-            return self._downloader.urlopen(self._create_request(url_or_request, data, headers, query))
+            return self._downloader.urlopen(self._create_request(url_or_request, data, headers, query, extensions))
         except network_exceptions as err:
             if isinstance(err, HTTPError):
                 if self.__can_accept_status_code(err, expected_status):
@@ -873,7 +904,7 @@ def _request_webpage(self, url_or_request, video_id, note=None, errnote=None, fa
             if errnote is None:
                 errnote = 'Unable to download webpage'
 
-            errmsg = f'{errnote}: {error_to_compat_str(err)}'
+            errmsg = f'{errnote}: {err}'
             if fatal:
                 raise ExtractorError(errmsg, cause=err)
             else:
@@ -881,13 +912,14 @@ def _request_webpage(self, url_or_request, video_id, note=None, errnote=None, fa
                 return False
 
     def _download_webpage_handle(self, url_or_request, video_id, note=None, errnote=None, fatal=True,
-                                 encoding=None, data=None, headers={}, query={}, expected_status=None):
+                                 encoding=None, data=None, headers={}, query={}, expected_status=None,
+                                 impersonate=None, require_impersonation=False):
         """
         Return a tuple (page content as string, URL handle).
 
         Arguments:
         url_or_request -- plain text URL as a string or
-            a urllib.request.Request object
+            a yt_dlp.networking.Request object
         video_id -- Video/playlist/item identifier (string)
 
         Keyword arguments:
@@ -912,17 +944,27 @@ def _download_webpage_handle(self, url_or_request, video_id, note=None, errnote=
                   returning True if it should be accepted
             Note that this argument does not affect success status codes (2xx)
             which are always accepted.
+        impersonate -- the impersonate target. Can be any of the following entities:
+                - an instance of yt_dlp.networking.impersonate.ImpersonateTarget
+                - a string in the format of CLIENT[:OS]
+                - a list or a tuple of CLIENT[:OS] strings or ImpersonateTarget instances
+                - a boolean value; True means any impersonate target is sufficient
+        require_impersonation -- flag to toggle whether the request should raise an error
+            if impersonation is not possible (bool, default: False)
         """
 
         # Strip hashes from the URL (#1038)
         if isinstance(url_or_request, str):
             url_or_request = url_or_request.partition('#')[0]
 
-        urlh = self._request_webpage(url_or_request, video_id, note, errnote, fatal, data=data, headers=headers, query=query, expected_status=expected_status)
+        urlh = self._request_webpage(url_or_request, video_id, note, errnote, fatal, data=data,
+                                     headers=headers, query=query, expected_status=expected_status,
+                                     impersonate=impersonate, require_impersonation=require_impersonation)
         if urlh is False:
             assert not fatal
             return False
-        content = self._webpage_read_content(urlh, url_or_request, video_id, note, errnote, fatal, encoding=encoding)
+        content = self._webpage_read_content(urlh, url_or_request, video_id, note, errnote, fatal,
+                                             encoding=encoding, data=data)
         return (content, urlh)
 
     @staticmethod
@@ -951,7 +993,7 @@ def __check_blocked(self, content):
                 r'<iframe src="([^"]+)"', content,
                 'Websense information URL', default=None)
             if blocked_iframe:
-                msg += ' Visit %s for more details' % blocked_iframe
+                msg += f' Visit {blocked_iframe} for more details'
             raise ExtractorError(msg, expected=True)
         if '<title>The URL you requested has been blocked</title>' in first_block:
             msg = (
@@ -961,7 +1003,7 @@ def __check_blocked(self, content):
                 r'</h1><p>(.*?)</p>',
                 content, 'block message', default=None)
             if block_msg:
-                msg += ' (Message: "%s")' % block_msg.replace('\n', ' ')
+                msg += ' (Message: "{}")'.format(block_msg.replace('\n', ' '))
             raise ExtractorError(msg, expected=True)
         if ('<title>TTK :: Доступ к ресурсу ограничен</title>' in content
                 and 'blocklist.rkn.gov.ru' in content):
@@ -970,11 +1012,13 @@ def __check_blocked(self, content):
                 'Visit http://blocklist.rkn.gov.ru/ for a block reason.',
                 expected=True)
 
-    def _request_dump_filename(self, url, video_id):
-        basen = f'{video_id}_{url}'
+    def _request_dump_filename(self, url, video_id, data=None):
+        if data is not None:
+            data = hashlib.md5(data).hexdigest()
+        basen = join_nonempty(video_id, data, url, delim='_')
         trim_length = self.get_param('trim_file_name') or 240
         if len(basen) > trim_length:
-            h = '___' + hashlib.md5(basen.encode('utf-8')).hexdigest()
+            h = '___' + hashlib.md5(basen.encode()).hexdigest()
             basen = basen[:trim_length - len(h)] + h
         filename = sanitize_filename(f'{basen}.dump', restricted=True)
         # Working around MAX_PATH limitation on Windows (see
@@ -993,7 +1037,8 @@ def __decode_webpage(self, webpage_bytes, encoding, headers):
         except LookupError:
             return webpage_bytes.decode('utf-8', 'replace')
 
-    def _webpage_read_content(self, urlh, url_or_request, video_id, note=None, errnote=None, fatal=True, prefix=None, encoding=None):
+    def _webpage_read_content(self, urlh, url_or_request, video_id, note=None, errnote=None, fatal=True,
+                              prefix=None, encoding=None, data=None):
         webpage_bytes = urlh.read()
         if prefix is not None:
             webpage_bytes = prefix + webpage_bytes
@@ -1002,7 +1047,9 @@ def _webpage_read_content(self, urlh, url_or_request, video_id, note=None, errno
             dump = base64.b64encode(webpage_bytes).decode('ascii')
             self._downloader.to_screen(dump)
         if self.get_param('write_pages'):
-            filename = self._request_dump_filename(urlh.url, video_id)
+            if isinstance(url_or_request, Request):
+                data = self._create_request(url_or_request, data).data
+            filename = self._request_dump_filename(urlh.url, video_id, data)
             self.to_screen(f'Saving request to {filename}')
             with open(filename, 'wb') as outf:
                 outf.write(webpage_bytes)
@@ -1022,7 +1069,7 @@ def _parse_xml(self, xml_string, video_id, transform_source=None, fatal=True, er
         if transform_source:
             xml_string = transform_source(xml_string)
         try:
-            return compat_etree_fromstring(xml_string.encode('utf-8'))
+            return compat_etree_fromstring(xml_string.encode())
         except xml.etree.ElementTree.ParseError as ve:
             self.__print_error('Failed to parse XML' if errnote is None else errnote, fatal, video_id, ve)
 
@@ -1047,20 +1094,23 @@ def parse(ie, content, *args, errnote=errnote, **kwargs):
             return getattr(ie, parser)(content, *args, **kwargs)
 
         def download_handle(self, url_or_request, video_id, note=note, errnote=errnote, transform_source=None,
-                            fatal=True, encoding=None, data=None, headers={}, query={}, expected_status=None):
+                            fatal=True, encoding=None, data=None, headers={}, query={}, expected_status=None,
+                            impersonate=None, require_impersonation=False):
             res = self._download_webpage_handle(
                 url_or_request, video_id, note=note, errnote=errnote, fatal=fatal, encoding=encoding,
-                data=data, headers=headers, query=query, expected_status=expected_status)
+                data=data, headers=headers, query=query, expected_status=expected_status,
+                impersonate=impersonate, require_impersonation=require_impersonation)
             if res is False:
                 return res
             content, urlh = res
             return parse(self, content, video_id, transform_source=transform_source, fatal=fatal, errnote=errnote), urlh
 
         def download_content(self, url_or_request, video_id, note=note, errnote=errnote, transform_source=None,
-                             fatal=True, encoding=None, data=None, headers={}, query={}, expected_status=None):
+                             fatal=True, encoding=None, data=None, headers={}, query={}, expected_status=None,
+                             impersonate=None, require_impersonation=False):
             if self.get_param('load_pages'):
                 url_or_request = self._create_request(url_or_request, data, headers, query)
-                filename = self._request_dump_filename(url_or_request.url, video_id)
+                filename = self._request_dump_filename(url_or_request.url, video_id, url_or_request.data)
                 self.to_screen(f'Loading request from {filename}')
                 try:
                     with open(filename, 'rb') as dumpf:
@@ -1080,6 +1130,8 @@ def download_content(self, url_or_request, video_id, note=note, errnote=errnote,
                 'headers': headers,
                 'query': query,
                 'expected_status': expected_status,
+                'impersonate': impersonate,
+                'require_impersonation': require_impersonation,
             }
             if parser is None:
                 kwargs.pop('transform_source')
@@ -1168,11 +1220,11 @@ def report_drm(self, video_id, partial=NO_DEFAULT):
 
     def report_extraction(self, id_or_name):
         """Report information extraction."""
-        self.to_screen('%s: Extracting information' % id_or_name)
+        self.to_screen(f'{id_or_name}: Extracting information')
 
     def report_download_webpage(self, video_id):
         """Report webpage download."""
-        self.to_screen('%s: Downloading webpage' % video_id)
+        self.to_screen(f'{video_id}: Downloading webpage')
 
     def report_age_confirmation(self):
         """Report attempt to confirm age."""
@@ -1278,9 +1330,9 @@ def _search_regex(self, pattern, string, name, default=NO_DEFAULT, fatal=True, f
         elif default is not NO_DEFAULT:
             return default
         elif fatal:
-            raise RegexNotFoundError('Unable to extract %s' % _name)
+            raise RegexNotFoundError(f'Unable to extract {_name}')
         else:
-            self.report_warning('unable to extract %s' % _name + bug_reports_message())
+            self.report_warning(f'unable to extract {_name}' + bug_reports_message())
             return None
 
     def _search_json(self, start_pattern, string, name, video_id, *, end_pattern='',
@@ -1379,14 +1431,14 @@ def _get_tfa_info(self, note='two-factor verification code'):
         if tfa is not None:
             return tfa
 
-        return getpass.getpass('Type %s and press [Return]: ' % note)
+        return getpass.getpass(f'Type {note} and press [Return]: ')
 
     # Helper functions for extracting OpenGraph info
     @staticmethod
     def _og_regexes(prop):
         content_re = r'content=(?:"([^"]+?)"|\'([^\']+?)\'|\s*([^\s"\'=<>`]+?)(?=\s|/?>))'
-        property_re = (r'(?:name|property)=(?:\'og%(sep)s%(prop)s\'|"og%(sep)s%(prop)s"|\s*og%(sep)s%(prop)s\b)'
-                       % {'prop': re.escape(prop), 'sep': '(?:&#x3A;|[:-])'})
+        property_re = r'(?:name|property)=(?:\'og{sep}{prop}\'|"og{sep}{prop}"|\s*og{sep}{prop}\b)'.format(
+            prop=re.escape(prop), sep='(?:&#x3A;|[:-])')
         template = r'<meta[^>]+?%s[^>]+?%s'
         return [
             template % (property_re, content_re),
@@ -1395,14 +1447,14 @@ def _og_regexes(prop):
 
     @staticmethod
     def _meta_regex(prop):
-        return r'''(?isx)<meta
-                    (?=[^>]+(?:itemprop|name|property|id|http-equiv)=(["\']?)%s\1)
-                    [^>]+?content=(["\'])(?P<content>.*?)\2''' % re.escape(prop)
+        return rf'''(?isx)<meta
+                    (?=[^>]+(?:itemprop|name|property|id|http-equiv)=(["\']?){re.escape(prop)}\1)
+                    [^>]+?content=(["\'])(?P<content>.*?)\2'''
 
     def _og_search_property(self, prop, html, name=None, **kargs):
         prop = variadic(prop)
         if name is None:
-            name = 'OpenGraph %s' % prop[0]
+            name = f'OpenGraph {prop[0]}'
         og_regexes = []
         for p in prop:
             og_regexes.extend(self._og_regexes(p))
@@ -1525,7 +1577,7 @@ def _search_json_ld(self, html, video_id, expected_type=None, *, fatal=True, def
         elif fatal:
             raise RegexNotFoundError('Unable to extract JSON-LD')
         else:
-            self.report_warning('unable to extract JSON-LD %s' % bug_reports_message())
+            self.report_warning(f'unable to extract JSON-LD {bug_reports_message()}')
             return {}
 
     def _json_ld(self, json_ld, video_id, fatal=True, expected_type=None):
@@ -1547,8 +1599,8 @@ def _json_ld(self, json_ld, video_id, fatal=True, expected_type=None):
         }
 
         def is_type(e, *expected_types):
-            type = variadic(traverse_obj(e, '@type'))
-            return any(x in type for x in expected_types)
+            type_ = variadic(traverse_obj(e, '@type'))
+            return any(x in type_ for x in expected_types)
 
         def extract_interaction_type(e):
             interaction_type = e.get('interactionType')
@@ -1577,7 +1629,7 @@ def extract_interaction_statistic(e):
                 count_kind = INTERACTION_TYPE_MAP.get(interaction_type.split('/')[-1])
                 if not count_kind:
                     continue
-                count_key = '%s_count' % count_kind
+                count_key = f'{count_kind}_count'
                 if info.get(count_key) is not None:
                     continue
                 info[count_key] = interaction_count
@@ -1589,7 +1641,7 @@ def extract_chapter_information(e):
                 'end_time': part.get('endOffset'),
             } for part in variadic(e.get('hasPart') or []) if part.get('@type') == 'Clip']
             for idx, (last_c, current_c, next_c) in enumerate(zip(
-                    [{'end_time': 0}] + chapters, chapters, chapters[1:])):
+                    [{'end_time': 0}, *chapters], chapters, chapters[1:])):
                 current_c['end_time'] = current_c['end_time'] or next_c['start_time']
                 current_c['start_time'] = current_c['start_time'] or last_c['end_time']
                 if None in current_c.values():
@@ -1698,12 +1750,16 @@ def traverse_json_ld(json_ld, at_top_level=True):
         traverse_json_ld(json_ld)
         return filter_dict(info)
 
-    def _search_nextjs_data(self, webpage, video_id, *, transform_source=None, fatal=True, **kw):
-        return self._parse_json(
-            self._search_regex(
-                r'(?s)<script[^>]+id=[\'"]__NEXT_DATA__[\'"][^>]*>([^<]+)</script>',
-                webpage, 'next.js data', fatal=fatal, **kw),
-            video_id, transform_source=transform_source, fatal=fatal)
+    def _search_nextjs_data(self, webpage, video_id, *, fatal=True, default=NO_DEFAULT, **kw):
+        if default == '{}':
+            self._downloader.deprecation_warning('using `default=\'{}\'` is deprecated, use `default={}` instead')
+            default = {}
+        if default is not NO_DEFAULT:
+            fatal = False
+
+        return self._search_json(
+            r'<script[^>]+id=[\'"]__NEXT_DATA__[\'"][^>]*>', webpage, 'next.js data',
+            video_id, end_pattern='</script>', fatal=fatal, default=default, **kw)
 
     def _search_nuxt_data(self, webpage, video_id, context_name='__NUXT__', *, fatal=True, traverse=('data', 0)):
         """Parses Nuxt.js metadata. This works as long as the function __NUXT__ invokes is a pure function"""
@@ -1726,9 +1782,9 @@ def _search_nuxt_data(self, webpage, video_id, context_name='__NUXT__', *, fatal
     def _hidden_inputs(html):
         html = re.sub(r'<!--(?:(?!<!--).)*-->', '', html)
         hidden_inputs = {}
-        for input in re.findall(r'(?i)(<input[^>]+>)', html):
-            attrs = extract_attributes(input)
-            if not input:
+        for input_el in re.findall(r'(?i)(<input[^>]+>)', html):
+            attrs = extract_attributes(input_el)
+            if not input_el:
                 continue
             if attrs.get('type') not in ('hidden', 'submit'):
                 continue
@@ -1740,8 +1796,8 @@ def _hidden_inputs(html):
 
     def _form_hidden_inputs(self, form_id, html):
         form = self._search_regex(
-            r'(?is)<form[^>]+?id=(["\'])%s\1[^>]*>(?P<form>.+?)</form>' % form_id,
-            html, '%s form' % form_id, group='form')
+            rf'(?is)<form[^>]+?id=(["\']){form_id}\1[^>]*>(?P<form>.+?)</form>',
+            html, f'{form_id} form', group='form')
         return self._hidden_inputs(form)
 
     @classproperty(cache=True)
@@ -1771,7 +1827,7 @@ def _check_formats(self, formats, video_id):
             formats[:] = filter(
                 lambda f: self._is_valid_url(
                     f['url'], video_id,
-                    item='%s video format' % f.get('format_id') if f.get('format_id') else 'video'),
+                    item='{} video format'.format(f.get('format_id')) if f.get('format_id') else 'video'),
                 formats)
 
     @staticmethod
@@ -1787,15 +1843,14 @@ def _remove_duplicate_formats(formats):
     def _is_valid_url(self, url, video_id, item='video', headers={}):
         url = self._proto_relative_url(url, scheme='http:')
         # For now assume non HTTP(S) URLs always valid
-        if not (url.startswith('http://') or url.startswith('https://')):
+        if not url.startswith(('http://', 'https://')):
             return True
         try:
-            self._request_webpage(url, video_id, 'Checking %s URL' % item, headers=headers)
+            self._request_webpage(url, video_id, f'Checking {item} URL', headers=headers)
             return True
         except ExtractorError as e:
             self.to_screen(
-                '%s: %s URL is invalid, skipping: %s'
-                % (video_id, item, error_to_compat_str(e.cause)))
+                f'{video_id}: {item} URL is invalid, skipping: {e.cause!s}')
             return False
 
     def http_scheme(self):
@@ -1849,8 +1904,8 @@ def _parse_f4m_formats(self, manifest, manifest_url, video_id, preference=None,
         # currently yt-dlp cannot decode the playerVerificationChallenge as Akamai uses Adobe Alchemy
         akamai_pv = manifest.find('{http://ns.adobe.com/f4m/1.0}pv-2.0')
         if akamai_pv is not None and ';' in akamai_pv.text:
-            playerVerificationChallenge = akamai_pv.text.split(';')[0]
-            if playerVerificationChallenge.strip() != '':
+            player_verification_challenge = akamai_pv.text.split(';')[0]
+            if player_verification_challenge.strip() != '':
                 return []
 
         formats = []
@@ -1896,7 +1951,7 @@ def _parse_f4m_formats(self, manifest, manifest_url, video_id, preference=None,
                 if not media_url:
                     continue
                 manifest_url = (
-                    media_url if media_url.startswith('http://') or media_url.startswith('https://')
+                    media_url if media_url.startswith(('http://', 'https://'))
                     else ((manifest_base_url or '/'.join(manifest_url.split('/')[:-1])) + '/' + media_url))
                 # If media_url is itself a f4m manifest do the recursive extraction
                 # since bitrates in parent manifest (this one) and media_url manifest
@@ -1957,7 +2012,7 @@ def _m3u8_meta_format(self, m3u8_url, ext=None, preference=None, quality=None, m
     def _report_ignoring_subs(self, name):
         self.report_warning(bug_reports_message(
             f'Ignoring subtitle tracks found in the {name} manifest; '
-            'if any subtitle tracks are missing,'
+            'if any subtitle tracks are missing,',
         ), only_once=True)
 
     def _extract_m3u8_formats(self, *args, **kwargs):
@@ -2048,7 +2103,7 @@ def _extract_m3u8_playlist_indices(*args, **kwargs):
             formats = [{
                 'format_id': join_nonempty(m3u8_id, idx),
                 'format_index': idx,
-                'url': m3u8_url or encode_data_uri(m3u8_doc.encode('utf-8'), 'application/x-mpegurl'),
+                'url': m3u8_url or encode_data_uri(m3u8_doc.encode(), 'application/x-mpegurl'),
                 'ext': ext,
                 'protocol': entry_protocol,
                 'preference': preference,
@@ -2167,6 +2222,11 @@ def build_stream_name():
                         'quality': quality,
                         'has_drm': has_drm,
                     }
+
+                    # YouTube-specific
+                    if yt_audio_content_id := last_stream_inf.get('YT-EXT-AUDIO-CONTENT-ID'):
+                        f['language'] = yt_audio_content_id.split('.')[0]
+
                     resolution = last_stream_inf.get('RESOLUTION')
                     if resolution:
                         mobj = re.search(r'(?P<width>\d+)[xX](?P<height>\d+)', resolution)
@@ -2260,7 +2320,7 @@ def _xpath_ns(path, namespace=None):
             if not c or c == '.':
                 out.append(c)
             else:
-                out.append('{%s}%s' % (namespace, c))
+                out.append(f'{{{namespace}}}{c}')
         return '/'.join(out)
 
     def _extract_smil_formats_and_subtitles(self, smil_url, video_id, fatal=True, f4m_params=None, transform_source=None):
@@ -2401,7 +2461,7 @@ def _parse_smil_formats_and_subtitles(
                     })
                 continue
 
-            src_url = src if src.startswith('http') else urllib.parse.urljoin(base, src)
+            src_url = src if src.startswith('http') else urllib.parse.urljoin(f'{base}/', src)
             src_url = src_url.strip()
 
             if proto == 'm3u8' or src_ext == 'm3u8':
@@ -2457,7 +2517,7 @@ def _parse_smil_formats_and_subtitles(
 
             imgs_count += 1
             formats.append({
-                'format_id': 'imagestream-%d' % (imgs_count),
+                'format_id': f'imagestream-{imgs_count}',
                 'url': src,
                 'ext': mimetype2ext(medium.get('type')),
                 'acodec': 'none',
@@ -2475,7 +2535,7 @@ def _parse_smil_formats_and_subtitles(
     def _parse_smil_subtitles(self, smil, namespace=None, subtitles_lang='en'):
         urls = []
         subtitles = {}
-        for num, textstream in enumerate(smil.findall(self._xpath_ns('.//textstream', namespace))):
+        for textstream in smil.findall(self._xpath_ns('.//textstream', namespace)):
             src = textstream.get('src')
             if not src or src in urls:
                 continue
@@ -2606,7 +2666,7 @@ def _merge_mpd_periods(self, periods):
             if subtitles and period['subtitles']:
                 self.report_warning(bug_reports_message(
                     'Found subtitles in multiple periods in the DASH manifest; '
-                    'if part of the subtitles are missing,'
+                    'if part of the subtitles are missing,',
                 ), only_once=True)
 
             for sub_lang, sub_info in period['subtitles'].items():
@@ -2732,7 +2792,7 @@ def extract_Initialization(source):
                         elif mimetype2ext(mime_type) in ('tt', 'dfxp', 'ttml', 'xml', 'json'):
                             content_type = 'text'
                         else:
-                            self.report_warning('Unknown MIME type %s in DASH manifest' % mime_type)
+                            self.report_warning(f'Unknown MIME type {mime_type} in DASH manifest')
                             continue
 
                     base_url = ''
@@ -2770,10 +2830,10 @@ def extract_Initialization(source):
                             'asr': int_or_none(representation_attrib.get('audioSamplingRate')),
                             'fps': int_or_none(representation_attrib.get('frameRate')),
                             'language': lang if lang not in ('mul', 'und', 'zxx', 'mis') else None,
-                            'format_note': 'DASH %s' % content_type,
+                            'format_note': f'DASH {content_type}',
                             'filesize': filesize,
                             'container': mimetype2ext(mime_type) + '_dash',
-                            **codecs
+                            **codecs,
                         }
                     elif content_type == 'text':
                         f = {
@@ -2814,8 +2874,8 @@ def prepare_template(template_name, identifiers):
                                 t += c
                         # Next, $...$ templates are translated to their
                         # %(...) counterparts to be used with % operator
-                        t = re.sub(r'\$(%s)\$' % '|'.join(identifiers), r'%(\1)d', t)
-                        t = re.sub(r'\$(%s)%%([^$]+)\$' % '|'.join(identifiers), r'%(\1)\2', t)
+                        t = re.sub(r'\$({})\$'.format('|'.join(identifiers)), r'%(\1)d', t)
+                        t = re.sub(r'\$({})%([^$]+)\$'.format('|'.join(identifiers)), r'%(\1)\2', t)
                         t.replace('$$', '$')
                         return t
 
@@ -2878,12 +2938,12 @@ def add_segment_url():
                                     'duration': float_or_none(segment_d, representation_ms_info['timescale']),
                                 })
 
-                            for num, s in enumerate(representation_ms_info['s']):
+                            for s in representation_ms_info['s']:
                                 segment_time = s.get('t') or segment_time
                                 segment_d = s['d']
                                 add_segment_url()
                                 segment_number += 1
-                                for r in range(s.get('r', 0)):
+                                for _ in range(s.get('r', 0)):
                                     segment_time += segment_d
                                     add_segment_url()
                                     segment_number += 1
@@ -2897,7 +2957,7 @@ def add_segment_url():
                         timescale = representation_ms_info['timescale']
                         for s in representation_ms_info['s']:
                             duration = float_or_none(s['d'], timescale)
-                            for r in range(s.get('r', 0) + 1):
+                            for _ in range(s.get('r', 0) + 1):
                                 segment_uri = representation_ms_info['segment_urls'][segment_index]
                                 fragments.append({
                                     location_key(segment_uri): segment_uri,
@@ -3005,7 +3065,7 @@ def _parse_ism_formats_and_subtitles(self, ism_doc, ism_url, ism_id=None):
                 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', 'EC-3'):
-                    self.report_warning('%s is not a supported codec' % fourcc)
+                    self.report_warning(f'{fourcc} is not a supported codec')
                     continue
                 tbr = int(track.attrib['Bitrate']) // 1000
                 # [1] does not mention Width and Height attributes. However,
@@ -3054,7 +3114,7 @@ def _parse_ism_formats_and_subtitles(self, ism_doc, ism_url, ism_id=None):
                             'fourcc': fourcc,
                             'language': stream_language,
                             'codec_private_data': track.get('CodecPrivateData'),
-                        }
+                        },
                     })
                 elif stream_type in ('video', 'audio'):
                     formats.append({
@@ -3136,13 +3196,13 @@ def _media_formats(src, cur_media_type, type_info=None):
         _MEDIA_TAG_NAME_RE = r'(?:(?:amp|dl8(?:-live)?)-)?(video|audio)'
         media_tags = [(media_tag, media_tag_name, media_type, '')
                       for media_tag, media_tag_name, media_type
-                      in re.findall(r'(?s)(<(%s)[^>]*/>)' % _MEDIA_TAG_NAME_RE, webpage)]
+                      in re.findall(rf'(?s)(<({_MEDIA_TAG_NAME_RE})[^>]*/>)', webpage)]
         media_tags.extend(re.findall(
             # We only allow video|audio followed by a whitespace or '>'.
             # Allowing more characters may end up in significant slow down (see
             # https://github.com/ytdl-org/youtube-dl/issues/11979,
             # e.g. http://www.porntrex.com/maps/videositemap.xml).
-            r'(?s)(<(?P<tag>%s)(?:\s+[^>]*)?>)(.*?)</(?P=tag)>' % _MEDIA_TAG_NAME_RE, webpage))
+            rf'(?s)(<(?P<tag>{_MEDIA_TAG_NAME_RE})(?:\s+[^>]*)?>)(.*?)</(?P=tag)>', webpage))
         for media_tag, _, media_type, media_content in media_tags:
             media_info = {
                 'formats': [],
@@ -3286,13 +3346,13 @@ def _extract_wowza_formats(self, url, video_id, m3u8_entry_protocol='m3u8_native
         mobj = re.search(
             r'(?:(?:http|rtmp|rtsp)(?P<s>s)?:)?(?P<url>//[^?]+)', url)
         url_base = mobj.group('url')
-        http_base_url = '%s%s:%s' % ('http', mobj.group('s') or '', url_base)
+        http_base_url = '{}{}:{}'.format('http', mobj.group('s') or '', url_base)
         formats = []
 
         def manifest_url(manifest):
             m_url = f'{http_base_url}/{manifest}'
             if query:
-                m_url += '?%s' % query
+                m_url += f'?{query}'
             return m_url
 
         if 'm3u8' not in skip_protocols:
@@ -3314,7 +3374,7 @@ def manifest_url(manifest):
                     video_id, fatal=False)
                 for rtmp_format in rtmp_formats:
                     rtsp_format = rtmp_format.copy()
-                    rtsp_format['url'] = '%s/%s' % (rtmp_format['url'], rtmp_format['play_path'])
+                    rtsp_format['url'] = '{}/{}'.format(rtmp_format['url'], rtmp_format['play_path'])
                     del rtsp_format['play_path']
                     del rtsp_format['ext']
                     rtsp_format.update({
@@ -3334,23 +3394,16 @@ def manifest_url(manifest):
         return formats
 
     def _find_jwplayer_data(self, webpage, video_id=None, transform_source=js_to_json):
-        mobj = re.search(
-            r'''(?s)jwplayer\s*\(\s*(?P<q>'|")(?!(?P=q)).+(?P=q)\s*\)(?!</script>).*?\.\s*setup\s*\(\s*(?P<options>(?:\([^)]*\)|[^)])+)\s*\)''',
-            webpage)
-        if mobj:
-            try:
-                jwplayer_data = self._parse_json(mobj.group('options'),
-                                                 video_id=video_id,
-                                                 transform_source=transform_source)
-            except ExtractorError:
-                pass
-            else:
-                if isinstance(jwplayer_data, dict):
-                    return jwplayer_data
-
-    def _extract_jwplayer_data(self, webpage, video_id, *args, **kwargs):
+        return self._search_json(
+            r'''(?<!-)\bjwplayer\s*\(\s*(?P<q>'|")(?!(?P=q)).+(?P=q)\s*\)(?:(?!</script>).)*?\.\s*(?:setup\s*\(|(?P<load>load)\s*\(\s*\[)''',
+            webpage, 'JWPlayer data', video_id,
+            # must be a {...} or sequence, ending
+            contains_pattern=r'\{(?s:.*)}(?(load)(?:\s*,\s*\{(?s:.*)})*)', end_pattern=r'(?(load)\]|\))',
+            transform_source=transform_source, default=None)
+
+    def _extract_jwplayer_data(self, webpage, video_id, *args, transform_source=js_to_json, **kwargs):
         jwplayer_data = self._find_jwplayer_data(
-            webpage, video_id, transform_source=js_to_json)
+            webpage, video_id, transform_source=transform_source)
         return self._parse_jwplayer_data(
             jwplayer_data, video_id, *args, **kwargs)
 
@@ -3382,22 +3435,14 @@ def _parse_jwplayer_data(self, jwplayer_data, video_id=None, require_title=True,
                 mpd_id=mpd_id, rtmp_params=rtmp_params, base_url=base_url)
 
             subtitles = {}
-            tracks = video_data.get('tracks')
-            if tracks and isinstance(tracks, list):
-                for track in tracks:
-                    if not isinstance(track, dict):
-                        continue
-                    track_kind = track.get('kind')
-                    if not track_kind or not isinstance(track_kind, str):
-                        continue
-                    if track_kind.lower() not in ('captions', 'subtitles'):
-                        continue
-                    track_url = urljoin(base_url, track.get('file'))
-                    if not track_url:
-                        continue
-                    subtitles.setdefault(track.get('label') or 'en', []).append({
-                        'url': self._proto_relative_url(track_url)
-                    })
+            for track in traverse_obj(video_data, (
+                    'tracks', lambda _, v: v['kind'].lower() in ('captions', 'subtitles'))):
+                track_url = urljoin(base_url, track.get('file'))
+                if not track_url:
+                    continue
+                subtitles.setdefault(track.get('label') or 'en', []).append({
+                    'url': self._proto_relative_url(track_url),
+                })
 
             entry = {
                 'id': this_video_id,
@@ -3475,14 +3520,14 @@ def _parse_jwplayer_formats(self, jwplayer_sources_data, video_id=None,
                     'tbr': int_or_none(source.get('bitrate'), scale=1000),
                     'filesize': int_or_none(source.get('filesize')),
                     'ext': ext,
-                    'format_id': format_id
+                    'format_id': format_id,
                 }
                 if source_url.startswith('rtmp'):
                     a_format['ext'] = 'flv'
                     # See com/longtailvideo/jwplayer/media/RTMPMediaProvider.as
                     # of jwplayer.flash.swf
                     rtmp_url_parts = re.split(
-                        r'((?:mp4|mp3|flv):)', source_url, 1)
+                        r'((?:mp4|mp3|flv):)', source_url, maxsplit=1)
                     if len(rtmp_url_parts) == 3:
                         rtmp_url, prefix, play_path = rtmp_url_parts
                         a_format.update({
@@ -3549,7 +3594,7 @@ def _apply_first_set_cookie_header(self, url_handle, cookie):
                 continue
             cookies = cookies.encode('iso-8859-1').decode('utf-8')
             cookie_value = re.search(
-                r'%s=(.+?);.*?\b[Dd]omain=(.+?)(?:[,;]|$)' % cookie, cookies)
+                rf'{cookie}=(.+?);.*?\b[Dd]omain=(.+?)(?:[,;]|$)', cookies)
             if cookie_value:
                 value, domain = cookie_value.groups()
                 self._set_cookie(domain, cookie, value)
@@ -3633,7 +3678,7 @@ def description(cls, *, markdown=True, search_examples=None):
             desc += ' (**Currently broken**)' if markdown else ' (Currently broken)'
 
         # Escape emojis. Ref: https://github.com/github/markup/issues/1153
-        name = (' - **%s**' % re.sub(r':(\w+:)', ':\u200B\\g<1>', cls.IE_NAME)) if markdown else cls.IE_NAME
+        name = (' - **{}**'.format(re.sub(r':(\w+:)', ':\u200B\\g<1>', cls.IE_NAME))) if markdown else cls.IE_NAME
         return f'{name}:{desc}' if desc else name
 
     def extract_subtitles(self, *args, **kwargs):
@@ -3673,7 +3718,7 @@ def extractor():
             self.to_screen(f'Extracted {comment_count} comments')
             return {
                 'comments': comments,
-                'comment_count': None if interrupted else comment_count
+                'comment_count': None if interrupted else comment_count,
             }
         return extractor
 
@@ -3777,9 +3822,9 @@ def _extract_chapters_from_description(self, description, duration):
 
     @staticmethod
     def _availability(is_private=None, needs_premium=None, needs_subscription=None, needs_auth=None, is_unlisted=None):
-        all_known = all(map(
-            lambda x: x is not None,
-            (is_private, needs_premium, needs_subscription, needs_auth, is_unlisted)))
+        all_known = all(
+            x is not None for x in
+            (is_private, needs_premium, needs_subscription, needs_auth, is_unlisted))
         return (
             'private' if is_private
             else 'premium_only' if needs_premium
@@ -3899,7 +3944,7 @@ class SearchInfoExtractor(InfoExtractor):
 
     @classproperty
     def _VALID_URL(cls):
-        return r'%s(?P<prefix>|[1-9][0-9]*|all):(?P<query>[\s\S]+)' % cls._SEARCH_KEY
+        return rf'{cls._SEARCH_KEY}(?P<prefix>|[1-9][0-9]*|all):(?P<query>[\s\S]+)'
 
     def _real_extract(self, query):
         prefix, query = self._match_valid_url(query).group('prefix', 'query')