Merge 8e6e47d6c1b19838de621aae4f3d31bda0c10cdc into da7223d4aa42ff9fc680b0951d043dd03cec2d30

* improve extraction of global "useful data" Array from player JS
* also handle tv-player and add tests: thx seproDev (yt-dlp/yt-dlp#12684)

Co-Authored-By: sepro <sepro@sepr0.com>

test/test_jsinterp.py

@@ -11,6 +11,7 @@ sys.path.insert(0, os.path.dirname(os.path.dirname(os.path.abspath(__file__))))
 
 import math
 import re
+import time
 
 from youtube_dl.compat import compat_str as str
 from youtube_dl.jsinterp import JS_Undefined, JSInterpreter
@@ -208,6 +209,34 @@ class TestJSInterpreter(unittest.TestCase):
         self._test(jsi, 86000, args=['12/31/1969 18:01:26 MDT'])
         # epoch 0
         self._test(jsi, 0, args=['1 January 1970 00:00:00 UTC'])
+        # undefined
+        self._test(jsi, NaN, args=[JS_Undefined])
+        # y,m,d, ... - may fail with older dates lacking DST data
+        jsi = JSInterpreter(
+            'function f() { return new Date(%s); }'
+            % ('2024, 5, 29, 2, 52, 12, 42',))
+        self._test(jsi, (
+            1719625932042                           # UK value
+            + (
+                + 3600                              # back to GMT
+                + (time.altzone if time.daylight    # host's DST
+                   else time.timezone)
+            ) * 1000))
+        # no arg
+        self.assertAlmostEqual(JSInterpreter(
+            'function f() { return new Date() - 0; }').call_function('f'),
+            time.time() * 1000, delta=100)
+        # Date.now()
+        self.assertAlmostEqual(JSInterpreter(
+            'function f() { return Date.now(); }').call_function('f'),
+            time.time() * 1000, delta=100)
+        # Date.parse()
+        jsi = JSInterpreter('function f(dt) { return Date.parse(dt); }')
+        self._test(jsi, 0, args=['1 January 1970 00:00:00 UTC'])
+        # Date.UTC()
+        jsi = JSInterpreter('function f() { return Date.UTC(%s); }'
+                            % ('1970, 0, 1, 0, 0, 0, 0',))
+        self._test(jsi, 0)
 
     def test_call(self):
         jsi = JSInterpreter('''
@@ -463,6 +492,14 @@ class TestJSInterpreter(unittest.TestCase):
         self._test('function f(){return NaN << 42}', 0)
         self._test('function f(){return "21.9" << 1}', 42)
         self._test('function f(){return 21 << 4294967297}', 42)
+        self._test('function f(){return true << "5";}', 32)
+        self._test('function f(){return true << true;}', 2)
+        self._test('function f(){return "19" & "21.9";}', 17)
+        self._test('function f(){return "19" & false;}', 0)
+        self._test('function f(){return "11.0" >> "2.1";}', 2)
+        self._test('function f(){return 5 ^ 9;}', 12)
+        self._test('function f(){return 0.0 << NaN}', 0)
+        self._test('function f(){return null << undefined}', 0)
 
     def test_negative(self):
         self._test('function f(){return 2    *    -2.0    ;}', -4)

test/test_youtube_signature.py

@@ -223,6 +223,42 @@ _NSIG_TESTS = [
         'https://www.youtube.com/s/player/9c6dfc4a/player_ias.vflset/en_US/base.js',
         'jbu7ylIosQHyJyJV', 'uwI0ESiynAmhNg',
     ),
+    (
+        'https://www.youtube.com/s/player/f6e09c70/player_ias.vflset/en_US/base.js',
+        'W9HJZKktxuYoDTqW', 'jHbbkcaxm54',
+    ),
+    (
+        'https://www.youtube.com/s/player/f6e09c70/player_ias_tce.vflset/en_US/base.js',
+        'W9HJZKktxuYoDTqW', 'jHbbkcaxm54',
+    ),
+    (
+        'https://www.youtube.com/s/player/643afba4/player_ias.vflset/en_US/base.js',
+        'W9HJZKktxuYoDTqW', 'larxUlagTRAcSw',
+    ),
+    (
+        'https://www.youtube.com/s/player/e7567ecf/player_ias_tce.vflset/en_US/base.js',
+        'Sy4aDGc0VpYRR9ew_', '5UPOT1VhoZxNLQ',
+    ),
+    (
+        'https://www.youtube.com/s/player/d50f54ef/player_ias_tce.vflset/en_US/base.js',
+        'Ha7507LzRmH3Utygtj', 'XFTb2HoeOE5MHg',
+    ),
+    (
+        'https://www.youtube.com/s/player/074a8365/player_ias_tce.vflset/en_US/base.js',
+        'Ha7507LzRmH3Utygtj', 'ufTsrE0IVYrkl8v',
+    ),
+    (
+        'https://www.youtube.com/s/player/643afba4/player_ias.vflset/en_US/base.js',
+        'N5uAlLqm0eg1GyHO', 'dCBQOejdq5s-ww',
+    ),
+    (
+        'https://www.youtube.com/s/player/69f581a5/tv-player-ias.vflset/tv-player-ias.js',
+        '-qIP447rVlTTwaZjY', 'KNcGOksBAvwqQg',
+    ),
+    (
+        'https://www.youtube.com/s/player/643afba4/tv-player-ias.vflset/tv-player-ias.js',
+        'ir9-V6cdbCiyKxhr', '2PL7ZDYAALMfmA',
+    ),
 ]
 
 
@@ -284,7 +320,7 @@ def t_factory(name, sig_func, url_pattern):
 
 
 def signature(jscode, sig_input):
-    func = YoutubeIE(FakeYDL())._parse_sig_js(jscode)
+    func = YoutubeIE(FakeYDL({'cachedir': False}))._parse_sig_js(jscode)
     src_sig = (
         compat_str(string.printable[:sig_input])
         if isinstance(sig_input, int) else sig_input)
@@ -292,9 +328,10 @@ def signature(jscode, sig_input):
 
 
 def n_sig(jscode, sig_input):
-    funcname = YoutubeIE(FakeYDL())._extract_n_function_name(jscode)
-    return JSInterpreter(jscode).call_function(
-        funcname, sig_input, _ytdl_do_not_return=sig_input)
+    ie = YoutubeIE(FakeYDL({'cachedir': False}))
+    jsi = JSInterpreter(jscode)
+    jsi, _, func_code = ie._extract_n_function_code_jsi(sig_input, jsi)
+    return ie._extract_n_function_from_code(jsi, func_code)(sig_input)
 
 
 make_sig_test = t_factory(

youtube_dl/__init__.py

@@ -18,7 +18,7 @@ from .compat import (
     compat_getpass,
     compat_register_utf8,
     compat_shlex_split,
-    workaround_optparse_bug9161,
+    _workaround_optparse_bug9161,
 )
 from .utils import (
     _UnsafeExtensionError,
@@ -50,7 +50,7 @@ def _real_main(argv=None):
     # Compatibility fix for Windows
     compat_register_utf8()
 
-    workaround_optparse_bug9161()
+    _workaround_optparse_bug9161()
 
     setproctitle('youtube-dl')
 

youtube_dl/compat.py

@@ -16,7 +16,6 @@ import os
 import platform
 import re
 import shlex
-import shutil
 import socket
 import struct
 import subprocess
@@ -24,11 +23,15 @@ import sys
 import types
 import xml.etree.ElementTree
 
+_IDENTITY = lambda x: x
+
 # naming convention
 # 'compat_' + Python3_name.replace('.', '_')
 # other aliases exist for convenience and/or legacy
+# wrap disposable test values in type() to reclaim storage
 
-# deal with critical unicode/str things first
+# deal with critical unicode/str things first:
+# compat_str, compat_basestring, compat_chr
 try:
     # Python 2
     compat_str, compat_basestring, compat_chr = (
@@ -39,18 +42,23 @@ except NameError:
         str, (str, bytes), chr
     )
 
-# casefold
+
+# compat_casefold
 try:
     compat_str.casefold
     compat_casefold = lambda s: s.casefold()
 except AttributeError:
     from .casefold import _casefold as compat_casefold
 
+
+# compat_collections_abc
 try:
     import collections.abc as compat_collections_abc
 except ImportError:
     import collections as compat_collections_abc
 
+
+# compat_urllib_request
 try:
     import urllib.request as compat_urllib_request
 except ImportError:  # Python 2
@@ -79,11 +87,15 @@ except TypeError:
     _add_init_method_arg(compat_urllib_request.Request)
     del _add_init_method_arg
 
+
+# compat_urllib_error
 try:
     import urllib.error as compat_urllib_error
 except ImportError:  # Python 2
     import urllib2 as compat_urllib_error
 
+
+# compat_urllib_parse
 try:
     import urllib.parse as compat_urllib_parse
 except ImportError:  # Python 2
@@ -98,17 +110,23 @@ except ImportError:  # Python 2
 compat_urlparse = compat_urllib_parse
 compat_urllib_parse_urlparse = compat_urllib_parse.urlparse
 
+
+# compat_urllib_response
 try:
     import urllib.response as compat_urllib_response
 except ImportError:  # Python 2
     import urllib as compat_urllib_response
 
+
+# compat_urllib_response.addinfourl
 try:
     compat_urllib_response.addinfourl.status
 except AttributeError:
     # .getcode() is deprecated in Py 3.
     compat_urllib_response.addinfourl.status = property(lambda self: self.getcode())
 
+
+# compat_http_cookiejar
 try:
     import http.cookiejar as compat_cookiejar
 except ImportError:  # Python 2
@@ -127,12 +145,16 @@ else:
     compat_cookiejar_Cookie = compat_cookiejar.Cookie
 compat_http_cookiejar_Cookie = compat_cookiejar_Cookie
 
+
+# compat_http_cookies
 try:
     import http.cookies as compat_cookies
 except ImportError:  # Python 2
     import Cookie as compat_cookies
 compat_http_cookies = compat_cookies
 
+
+# compat_http_cookies_SimpleCookie
 if sys.version_info[0] == 2 or sys.version_info < (3, 3):
     class compat_cookies_SimpleCookie(compat_cookies.SimpleCookie):
         def load(self, rawdata):
@@ -155,11 +177,15 @@ else:
     compat_cookies_SimpleCookie = compat_cookies.SimpleCookie
 compat_http_cookies_SimpleCookie = compat_cookies_SimpleCookie
 
+
+# compat_html_entities, probably useless now
 try:
     import html.entities as compat_html_entities
 except ImportError:  # Python 2
     import htmlentitydefs as compat_html_entities
 
+
+# compat_html_entities_html5
 try:  # Python >= 3.3
     compat_html_entities_html5 = compat_html_entities.html5
 except AttributeError:
@@ -2408,18 +2434,24 @@ except AttributeError:
     # Py < 3.1
     compat_http_client.HTTPResponse.getcode = lambda self: self.status
 
+
+# compat_urllib_HTTPError
 try:
     from urllib.error import HTTPError as compat_HTTPError
 except ImportError:  # Python 2
     from urllib2 import HTTPError as compat_HTTPError
 compat_urllib_HTTPError = compat_HTTPError
 
+
+# compat_urllib_request_urlretrieve
 try:
     from urllib.request import urlretrieve as compat_urlretrieve
 except ImportError:  # Python 2
     from urllib import urlretrieve as compat_urlretrieve
 compat_urllib_request_urlretrieve = compat_urlretrieve
 
+
+# compat_html_parser_HTMLParser, compat_html_parser_HTMLParseError
 try:
     from HTMLParser import (
         HTMLParser as compat_HTMLParser,
@@ -2432,22 +2464,33 @@ except ImportError:  # Python 3
         # HTMLParseError was deprecated in Python 3.3 and removed in
         # Python 3.5. Introducing dummy exception for Python >3.5 for compatible
         # and uniform cross-version exception handling
+
         class compat_HTMLParseError(Exception):
             pass
+
 compat_html_parser_HTMLParser = compat_HTMLParser
 compat_html_parser_HTMLParseError = compat_HTMLParseError
 
+
+# compat_subprocess_get_DEVNULL
 try:
     _DEVNULL = subprocess.DEVNULL
     compat_subprocess_get_DEVNULL = lambda: _DEVNULL
 except AttributeError:
     compat_subprocess_get_DEVNULL = lambda: open(os.path.devnull, 'w')
 
+
+# compat_http_server
 try:
     import http.server as compat_http_server
 except ImportError:
     import BaseHTTPServer as compat_http_server
 
+
+# compat_urllib_parse_unquote_to_bytes,
+# compat_urllib_parse_unquote, compat_urllib_parse_unquote_plus,
+# compat_urllib_parse_urlencode,
+# compat_urllib_parse_parse_qs
 try:
     from urllib.parse import unquote_to_bytes as compat_urllib_parse_unquote_to_bytes
     from urllib.parse import unquote as compat_urllib_parse_unquote
@@ -2598,6 +2641,8 @@ except ImportError:  # Python 2
 
 compat_urllib_parse_parse_qs = compat_parse_qs
 
+
+# compat_urllib_request_DataHandler
 try:
     from urllib.request import DataHandler as compat_urllib_request_DataHandler
 except ImportError:  # Python < 3.4
@@ -2632,16 +2677,20 @@ except ImportError:  # Python < 3.4
 
             return compat_urllib_response.addinfourl(io.BytesIO(data), headers, url)
 
+
+# compat_xml_etree_ElementTree_ParseError
 try:
     from xml.etree.ElementTree import ParseError as compat_xml_parse_error
 except ImportError:  # Python 2.6
     from xml.parsers.expat import ExpatError as compat_xml_parse_error
 compat_xml_etree_ElementTree_ParseError = compat_xml_parse_error
 
-etree = xml.etree.ElementTree
+
+# compat_xml_etree_ElementTree_Element
+_etree = xml.etree.ElementTree
 
 
-class _TreeBuilder(etree.TreeBuilder):
+class _TreeBuilder(_etree.TreeBuilder):
     def doctype(self, name, pubid, system):
         pass
 
@@ -2650,7 +2699,7 @@ try:
     # xml.etree.ElementTree.Element is a method in Python <=2.6 and
     # the following will crash with:
     #  TypeError: isinstance() arg 2 must be a class, type, or tuple of classes and types
-    isinstance(None, etree.Element)
+    isinstance(None, _etree.Element)
     from xml.etree.ElementTree import Element as compat_etree_Element
 except TypeError:  # Python <=2.6
     from xml.etree.ElementTree import _ElementInterface as compat_etree_Element
@@ -2658,12 +2707,12 @@ compat_xml_etree_ElementTree_Element = compat_etree_Element
 
 if sys.version_info[0] >= 3:
     def compat_etree_fromstring(text):
-        return etree.XML(text, parser=etree.XMLParser(target=_TreeBuilder()))
+        return _etree.XML(text, parser=_etree.XMLParser(target=_TreeBuilder()))
 else:
     # python 2.x tries to encode unicode strings with ascii (see the
     # XMLParser._fixtext method)
     try:
-        _etree_iter = etree.Element.iter
+        _etree_iter = _etree.Element.iter
     except AttributeError:  # Python <=2.6
         def _etree_iter(root):
             for el in root.findall('*'):
@@ -2675,27 +2724,29 @@ else:
     # 2.7 source
     def _XML(text, parser=None):
         if not parser:
-            parser = etree.XMLParser(target=_TreeBuilder())
+            parser = _etree.XMLParser(target=_TreeBuilder())
         parser.feed(text)
         return parser.close()
 
     def _element_factory(*args, **kwargs):
-        el = etree.Element(*args, **kwargs)
+        el = _etree.Element(*args, **kwargs)
         for k, v in el.items():
             if isinstance(v, bytes):
                 el.set(k, v.decode('utf-8'))
         return el
 
     def compat_etree_fromstring(text):
-        doc = _XML(text, parser=etree.XMLParser(target=_TreeBuilder(element_factory=_element_factory)))
+        doc = _XML(text, parser=_etree.XMLParser(target=_TreeBuilder(element_factory=_element_factory)))
         for el in _etree_iter(doc):
             if el.text is not None and isinstance(el.text, bytes):
                 el.text = el.text.decode('utf-8')
         return doc
 
-if hasattr(etree, 'register_namespace'):
-    compat_etree_register_namespace = etree.register_namespace
-else:
+
+# compat_xml_etree_register_namespace
+try:
+    compat_etree_register_namespace = _etree.register_namespace
+except AttributeError:
     def compat_etree_register_namespace(prefix, uri):
         """Register a namespace prefix.
         The registry is global, and any existing mapping for either the
@@ -2704,14 +2755,16 @@ else:
         attributes in this namespace will be serialized with prefix if possible.
         ValueError is raised if prefix is reserved or is invalid.
         """
-        if re.match(r"ns\d+$", prefix):
-            raise ValueError("Prefix format reserved for internal use")
-        for k, v in list(etree._namespace_map.items()):
+        if re.match(r'ns\d+$', prefix):
+            raise ValueError('Prefix format reserved for internal use')
+        for k, v in list(_etree._namespace_map.items()):
             if k == uri or v == prefix:
-                del etree._namespace_map[k]
-        etree._namespace_map[uri] = prefix
+                del _etree._namespace_map[k]
+        _etree._namespace_map[uri] = prefix
 compat_xml_etree_register_namespace = compat_etree_register_namespace
 
+
+# compat_xpath, compat_etree_iterfind
 if sys.version_info < (2, 7):
     # Here comes the crazy part: In 2.6, if the xpath is a unicode,
     # .//node does not match if a node is a direct child of . !
@@ -2898,7 +2951,6 @@ if sys.version_info < (2, 7):
         def __init__(self, root):
             self.root = root
 
-    ##
     # Generate all matching objects.
 
     def compat_etree_iterfind(elem, path, namespaces=None):
@@ -2933,13 +2985,15 @@ if sys.version_info < (2, 7):
 
 
 else:
-    compat_xpath = lambda xpath: xpath
     compat_etree_iterfind = lambda element, match: element.iterfind(match)
+    compat_xpath = _IDENTITY
 
 
+# compat_os_name
 compat_os_name = os._name if os.name == 'java' else os.name
 
 
+# compat_shlex_quote
 if compat_os_name == 'nt':
     def compat_shlex_quote(s):
         return s if re.match(r'^[-_\w./]+$', s) else '"%s"' % s.replace('"', '\\"')
@@ -2954,6 +3008,7 @@ else:
                 return "'" + s.replace("'", "'\"'\"'") + "'"
 
 
+# compat_shlex.split
 try:
     args = shlex.split('中文')
     assert (isinstance(args, list)
@@ -2969,6 +3024,7 @@ except (AssertionError, UnicodeEncodeError):
         return list(map(lambda s: s.decode('utf-8'), shlex.split(s, comments, posix)))
 
 
+# compat_ord
 def compat_ord(c):
     if isinstance(c, int):
         return c
@@ -2976,6 +3032,7 @@ def compat_ord(c):
         return ord(c)
 
 
+# compat_getenv, compat_os_path_expanduser, compat_setenv
 if sys.version_info >= (3, 0):
     compat_getenv = os.getenv
     compat_expanduser = os.path.expanduser
@@ -3063,6 +3120,7 @@ else:
 compat_os_path_expanduser = compat_expanduser
 
 
+# compat_os_path_realpath
 if compat_os_name == 'nt' and sys.version_info < (3, 8):
     # os.path.realpath on Windows does not follow symbolic links
     # prior to Python 3.8 (see https://bugs.python.org/issue9949)
@@ -3076,6 +3134,7 @@ else:
 compat_os_path_realpath = compat_realpath
 
 
+# compat_print
 if sys.version_info < (3, 0):
     def compat_print(s):
         from .utils import preferredencoding
@@ -3086,6 +3145,7 @@ else:
         print(s)
 
 
+# compat_getpass_getpass
 if sys.version_info < (3, 0) and sys.platform == 'win32':
     def compat_getpass(prompt, *args, **kwargs):
         if isinstance(prompt, compat_str):
@@ -3098,22 +3158,22 @@ else:
 compat_getpass_getpass = compat_getpass
 
 
+# compat_input
 try:
     compat_input = raw_input
 except NameError:  # Python 3
     compat_input = input
 
 
+# compat_kwargs
 # Python < 2.6.5 require kwargs to be bytes
 try:
-    def _testfunc(x):
-        pass
-    _testfunc(**{'x': 0})
+    (lambda x: x)(**{'x': 0})
 except TypeError:
     def compat_kwargs(kwargs):
         return dict((bytes(k), v) for k, v in kwargs.items())
 else:
-    compat_kwargs = lambda kwargs: kwargs
+    compat_kwargs = _IDENTITY
 
 
 # compat_numeric_types
@@ -3132,6 +3192,8 @@ except NameError:  # Python 3
 # compat_int
 compat_int = compat_integer_types[-1]
 
+
+# compat_socket_create_connection
 if sys.version_info < (2, 7):
     def compat_socket_create_connection(address, timeout, source_address=None):
         host, port = address
@@ -3158,6 +3220,7 @@ else:
     compat_socket_create_connection = socket.create_connection
 
 
+# compat_contextlib_suppress
 try:
     from contextlib import suppress as compat_contextlib_suppress
 except ImportError:
@@ -3200,12 +3263,12 @@ except AttributeError:
                         # repeated .close() is OK, but just in case
                         with compat_contextlib_suppress(EnvironmentError):
                             f.close()
-                popen.wait()
+            popen.wait()
 
 
 # Fix https://github.com/ytdl-org/youtube-dl/issues/4223
 # See http://bugs.python.org/issue9161 for what is broken
-def workaround_optparse_bug9161():
+def _workaround_optparse_bug9161():
     op = optparse.OptionParser()
     og = optparse.OptionGroup(op, 'foo')
     try:
@@ -3224,9 +3287,10 @@ def workaround_optparse_bug9161():
         optparse.OptionGroup.add_option = _compat_add_option
 
 
-if hasattr(shutil, 'get_terminal_size'):  # Python >= 3.3
-    compat_get_terminal_size = shutil.get_terminal_size
-else:
+# compat_shutil_get_terminal_size
+try:
+    from shutil import get_terminal_size as compat_get_terminal_size  # Python >= 3.3
+except ImportError:
     _terminal_size = collections.namedtuple('terminal_size', ['columns', 'lines'])
 
     def compat_get_terminal_size(fallback=(80, 24)):
@@ -3256,27 +3320,33 @@ else:
                 columns = _columns
             if lines is None or lines <= 0:
                 lines = _lines
+
         return _terminal_size(columns, lines)
 
+compat_shutil_get_terminal_size = compat_get_terminal_size
 
+
+# compat_itertools_count
 try:
-    itertools.count(start=0, step=1)
+    type(itertools.count(start=0, step=1))
     compat_itertools_count = itertools.count
-except TypeError:  # Python 2.6
+except TypeError:  # Python 2.6 lacks step
     def compat_itertools_count(start=0, step=1):
         while True:
             yield start
             start += step
 
 
+# compat_tokenize_tokenize
 if sys.version_info >= (3, 0):
     from tokenize import tokenize as compat_tokenize_tokenize
 else:
     from tokenize import generate_tokens as compat_tokenize_tokenize
 
 
+# compat_struct_pack, compat_struct_unpack, compat_Struct
 try:
-    struct.pack('!I', 0)
+    type(struct.pack('!I', 0))
 except TypeError:
     # In Python 2.6 and 2.7.x < 2.7.7, struct requires a bytes argument
     # See https://bugs.python.org/issue19099
@@ -3308,8 +3378,10 @@ else:
         compat_Struct = struct.Struct
 
 
-# compat_map/filter() returning an iterator, supposedly the
-# same versioning as for zip below
+# builtins returning an iterator
+
+# compat_map, compat_filter
+# supposedly the same versioning as for zip below
 try:
     from future_builtins import map as compat_map
 except ImportError:
@@ -3326,6 +3398,7 @@ except ImportError:
     except ImportError:
         compat_filter = filter
 
+# compat_zip
 try:
     from future_builtins import zip as compat_zip
 except ImportError:  # not 2.6+ or is 3.x
@@ -3335,6 +3408,7 @@ except ImportError:  # not 2.6+ or is 3.x
         compat_zip = zip
 
 
+# compat_itertools_zip_longest
 # method renamed between Py2/3
 try:
     from itertools import zip_longest as compat_itertools_zip_longest
@@ -3342,7 +3416,8 @@ except ImportError:
     from itertools import izip_longest as compat_itertools_zip_longest
 
 
-# new class in collections
+# compat_collections_chain_map
+# collections.ChainMap: new class
 try:
     from collections import ChainMap as compat_collections_chain_map
     # Py3.3's ChainMap is deficient
@@ -3398,19 +3473,22 @@ except ImportError:
         def new_child(self, m=None, **kwargs):
             m = m or {}
             m.update(kwargs)
-            return compat_collections_chain_map(m, *self.maps)
+            # support inheritance !
+            return type(self)(m, *self.maps)
 
         @property
         def parents(self):
-            return compat_collections_chain_map(*(self.maps[1:]))
+            return type(self)(*(self.maps[1:]))
 
 
+# compat_re_Pattern, compat_re_Match
 # Pythons disagree on the type of a pattern (RegexObject, _sre.SRE_Pattern, Pattern, ...?)
 compat_re_Pattern = type(re.compile(''))
 # and on the type of a match
 compat_re_Match = type(re.match('a', 'a'))
 
 
+# compat_base64_b64decode
 if sys.version_info < (3, 3):
     def compat_b64decode(s, *args, **kwargs):
         if isinstance(s, compat_str):
@@ -3422,6 +3500,7 @@ else:
 compat_base64_b64decode = compat_b64decode
 
 
+# compat_ctypes_WINFUNCTYPE
 if platform.python_implementation() == 'PyPy' and sys.pypy_version_info < (5, 4, 0):
     # PyPy2 prior to version 5.4.0 expects byte strings as Windows function
     # names, see the original PyPy issue [1] and the youtube-dl one [2].
@@ -3440,6 +3519,7 @@ else:
         return ctypes.WINFUNCTYPE(*args, **kwargs)
 
 
+# compat_open
 if sys.version_info < (3, 0):
     # open(file, mode='r', buffering=- 1, encoding=None, errors=None, newline=None, closefd=True) not: opener=None
     def compat_open(file_, *args, **kwargs):
@@ -3467,18 +3547,28 @@ except AttributeError:
     def compat_datetime_timedelta_total_seconds(td):
         return (td.microseconds + (td.seconds + td.days * 24 * 3600) * 10**6) / 10**6
 
+
 # optional decompression packages
+# compat_brotli
 # PyPi brotli package implements 'br' Content-Encoding
 try:
     import brotli as compat_brotli
 except ImportError:
     compat_brotli = None
+# compat_ncompress
 # PyPi ncompress package implements 'compress' Content-Encoding
 try:
     import ncompress as compat_ncompress
 except ImportError:
     compat_ncompress = None
 
+# compat_zstandard
+# PyPi zstandard package implements 'zstd' Content-Encoding (RFC 8878 7.2)
+try:
+    import zstandard as compat_zstandard
+except ImportError:
+    compat_zstandard = None
+
 
 legacy = [
     'compat_HTMLParseError',
@@ -3495,6 +3585,7 @@ legacy = [
     'compat_getpass',
     'compat_parse_qs',
     'compat_realpath',
+    'compat_shlex_split',
     'compat_urllib_parse_parse_qs',
     'compat_urllib_parse_unquote',
     'compat_urllib_parse_unquote_plus',
@@ -3508,8 +3599,6 @@ legacy = [
 
 
 __all__ = [
-    'compat_html_parser_HTMLParseError',
-    'compat_html_parser_HTMLParser',
     'compat_Struct',
     'compat_base64_b64decode',
     'compat_basestring',
@@ -3518,13 +3607,9 @@ __all__ = [
     'compat_chr',
     'compat_collections_abc',
     'compat_collections_chain_map',
-    'compat_datetime_timedelta_total_seconds',
-    'compat_http_cookiejar',
-    'compat_http_cookiejar_Cookie',
-    'compat_http_cookies',
-    'compat_http_cookies_SimpleCookie',
     'compat_contextlib_suppress',
     'compat_ctypes_WINFUNCTYPE',
+    'compat_datetime_timedelta_total_seconds',
     'compat_etree_fromstring',
     'compat_etree_iterfind',
     'compat_filter',
@@ -3533,6 +3618,12 @@ __all__ = [
     'compat_getpass_getpass',
     'compat_html_entities',
     'compat_html_entities_html5',
+    'compat_html_parser_HTMLParseError',
+    'compat_html_parser_HTMLParser',
+    'compat_http_cookiejar',
+    'compat_http_cookiejar_Cookie',
+    'compat_http_cookies',
+    'compat_http_cookies_SimpleCookie',
     'compat_http_client',
     'compat_http_server',
     'compat_input',
@@ -3555,7 +3646,7 @@ __all__ = [
     'compat_register_utf8',
     'compat_setenv',
     'compat_shlex_quote',
-    'compat_shlex_split',
+    'compat_shutil_get_terminal_size',
     'compat_socket_create_connection',
     'compat_str',
     'compat_struct_pack',
@@ -3575,5 +3666,5 @@ __all__ = [
     'compat_xml_etree_register_namespace',
     'compat_xpath',
     'compat_zip',
-    'workaround_optparse_bug9161',
+    'compat_zstandard',
 ]

youtube_dl/extractor/extractors.py

@@ -1384,7 +1384,9 @@ from .tvnow import (
 from .tvp import (
     TVPEmbedIE,
     TVPIE,
-    TVPWebsiteIE,
+    TVPStreamIE,
+    TVPVODSeriesIE,
+    TVPVODVideoIE,
 )
 from .tvplay import (
     TVPlayIE,

youtube_dl/extractor/tvp.py

@@ -2,52 +2,274 @@
 from __future__ import unicode_literals
 
 import itertools
+import random
 import re
 
 from .common import InfoExtractor
+from ..compat import (
+    compat_HTTPError,
+    compat_str,
+)
 from ..utils import (
     clean_html,
     determine_ext,
     ExtractorError,
-    get_element_by_attribute,
-    orderedSet,
+    int_or_none,
+    js_to_json,
+    traverse_obj,
+    url_or_none,
 )
 
 
+def txt_or_none(v, default=None):
+    return default if v is None else (compat_str(v).strip() or default)
+
+if not hasattr(InfoExtractor, '_match_valid_url'):
+
+    import sys
+    from ..compat import (
+        compat_os_name,
+        compat_re_Pattern as compiled_regex_type,
+    )
+    from ..utils import (
+        bug_reports_message,
+        error_to_compat_str,
+        NO_DEFAULT,
+        RegexNotFoundError,
+    )
+
+    BaseIE = InfoExtractor
+
+    class InfoExtractor(BaseIE):
+        def _match_valid_url(self, url):
+            return re.match(self._VALID_URL, url)
+
+        def _search_regex(self, pattern, string, name, default=NO_DEFAULT, fatal=True, flags=0, group=None):
+            """
+            Perform a regex search on the given string, using a single or a list of
+            patterns returning the first matching group.
+            In case of failure return a default value or raise a WARNING or a
+            RegexNotFoundError, depending on fatal, specifying the field name.
+            """
+            if isinstance(pattern, (str, compat_str, compiled_regex_type)):
+                mobj = re.search(pattern, string, flags)
+            else:
+                for p in pattern:
+                    mobj = re.search(p, string, flags)
+                    if mobj:
+                        break
+
+            if not self._downloader.params.get('no_color') and compat_os_name != 'nt' and sys.stderr.isatty():
+                _name = '\033[0;34m%s\033[0m' % name
+            else:
+                _name = name
+
+            if mobj:
+                if group is None:
+                    # return the first matching group
+                    return next(g for g in mobj.groups() if g is not None)
+                elif isinstance(group, (list, tuple)):
+                    return tuple(mobj.group(g) for g in group)
+                else:
+                    return mobj.group(group)
+            elif default is not NO_DEFAULT:
+                return default
+            elif fatal:
+                raise RegexNotFoundError('Unable to extract %s' % _name)
+            else:
+                self.report_warning('unable to extract %s' % _name + bug_reports_message())
+                return None
+
+        def _html_search_regex(self, pattern, string, name, default=NO_DEFAULT, fatal=True, flags=0, group=None):
+            """
+            Like _search_regex, but strips HTML tags and unescapes entities.
+            """
+            res = self._search_regex(pattern, string, name, default, fatal, flags, group)
+            if isinstance(res, tuple):
+                return tuple(map(clean_html, res))
+            return clean_html(res or None)
+
+        def _search_json(self, start_pattern, string, name, video_id, **kwargs):
+            """Searches string for the JSON object specified by start_pattern"""
+
+            # self, start_pattern, string, name, video_id, *, end_pattern='',
+            # contains_pattern=r'{(?s:.+)}', fatal=True, default=NO_DEFAULT
+            end_pattern = kwargs.pop('end_pattern', '')
+            contains_pattern = kwargs.pop('contains_pattern', r'{(?:[\s\S]+)}')
+            fatal = kwargs.get('fatal', True)
+            default = kwargs.get('default', NO_DEFAULT)
+
+            # NB: end_pattern is only used to reduce the size of the initial match
+            if default is NO_DEFAULT:
+                default, has_default = {}, False
+            else:
+                fatal, has_default = False, True
+
+            json_string = self._search_regex(
+                r'(?:{0})\s*(?P<json>{1})\s*(?:{2})'.format(
+                    start_pattern, contains_pattern, end_pattern),
+                string, name, group='json', fatal=fatal, default=None if has_default else NO_DEFAULT)
+            if not json_string:
+                return default
+
+            try:
+                # return self._parse_json(json_string, video_id, ignore_extra=True, **kwargs)
+                return self._parse_json(json_string, video_id, **kwargs)
+            except ExtractorError as e:
+                if not self._downloader.params.get('no_color') and compat_os_name != 'nt' and sys.stderr.isatty():
+                    _name = '\033[0;34m%s\033[0m' % name
+                else:
+                    _name = name
+                msg = 'Unable to extract {0} - Failed to parse JSON'.format(_name)
+                if fatal:
+                    raise ExtractorError(msg, cause=e.cause, video_id=video_id)
+                elif not has_default:
+                    self.report_warning(
+                        '{0}: {1}'.format(msg, error_to_compat_str(e)), video_id=video_id)
+            return default
+
+
 class TVPIE(InfoExtractor):
     IE_NAME = 'tvp'
     IE_DESC = 'Telewizja Polska'
-    _VALID_URL = r'https?://[^/]+\.tvp\.(?:pl|info)/(?:video/(?:[^,\s]*,)*|(?:(?!\d+/)[^/]+/)*)(?P<id>\d+)'
-
+    _VALID_URL = r'https?://(?:[^/]+\.)?(?:tvp(?:parlament)?\.(?:pl|info)|polandin\.com|tvpworld\.com|swipeto\.pl)/(?:(?:(?!\d+/)[^/]+/)*|(?:video|website)/[^/]+,)(?P<id>\d+)'
     _TESTS = [{
+        # TVPlayer 2 in js wrapper
+        'url': 'https://swipeto.pl/64095316/uliczny-foxtrot-wypozyczalnia-kaset-kto-pamieta-dvdvideo',
+        'info_dict': {
+            'id': '64095316',
+            'ext': 'mp4',
+            'title': 'Uliczny Foxtrot — Wypożyczalnia kaset. Kto pamięta DVD-Video?',
+            'age_limit': 0,
+            'duration': 374,
+            'thumbnail': r're:https://.+',
+        },
+        'expected_warnings': [
+            'Failed to download ISM manifest: HTTP Error 404: Not Found',
+            'Failed to download m3u8 information: HTTP Error 404: Not Found',
+        ],
+        'skip': 'Video gone: 404 Nie znaleziono obiektu',
+    }, {
+        # TVPlayer 2 in js wrapper (redirect to VodVideo)
         'url': 'https://vod.tvp.pl/video/czas-honoru,i-seria-odc-13,194536',
         'md5': 'a21eb0aa862f25414430f15fdfb9e76c',
         'info_dict': {
             'id': '194536',
             'ext': 'mp4',
             'title': 'Czas honoru, odc. 13 – Władek',
-            'description': 'md5:437f48b93558370b031740546b696e24',
+            'description': 'md5:76649d2014f65c99477be17f23a4dead',
+            'age_limit': 12,
         },
+        'add_ie': ['Generic', 'TVPEmbed'],
     }, {
-        'url': 'http://www.tvp.pl/there-can-be-anything-so-i-shortened-it/17916176',
-        'md5': 'b0005b542e5b4de643a9690326ab1257',
+        # film (old format)
+        'url': 'https://vod.tvp.pl/website/krzysztof-krawczyk-cale-moje-zycie,51374466',
         'info_dict': {
-            'id': '17916176',
+            'id': '51374509',
             'ext': 'mp4',
-            'title': 'TVP Gorzów pokaże filmy studentów z podroży dookoła świata',
-            'description': 'TVP Gorzów pokaże filmy studentów z podroży dookoła świata',
+            'title': 'Krzysztof Krawczyk – całe moje życie, Krzysztof Krawczyk – całe moje życie',
+            'description': 'md5:2e80823f00f5fc263555482f76f8fa42',
+            'age_limit': 12,
         },
+        'params': {
+            'skip_download': True,
+        },
+        'add_ie': ['TVPEmbed'],
+        'skip': 'This video is not available from your location due to geo restriction',
     }, {
+        # TVPlayer legacy
+        'url': 'https://www.tvp.pl/polska-press-video-uploader/wideo/62042351',
+        'info_dict': {
+            'id': '62042351',
+            'ext': 'mp4',
+            'title': 'Wideo',
+            'description': 'Wideo Kamera',
+            'duration': 24,
+            'age_limit': 0,
+            'thumbnail': r're:https://.+',
+        },
+        'add_ie': ['TVPEmbed'],
+    }, {
+        # TVPlayer 2 in iframe
         # page id is not the same as video id(#7799)
-        'url': 'https://wiadomosci.tvp.pl/33908820/28092017-1930',
-        'md5': '84cd3c8aec4840046e5ab712416b73d0',
+        'url': 'https://wiadomosci.tvp.pl/50725617/dzieci-na-sprzedaz-dla-homoseksualistow',
+        'md5': 'd35fb45103802488fcb7470e411b9ed4',
         'info_dict': {
-            'id': '33908820',
+            'id': '50725617',
             'ext': 'mp4',
-            'title': 'Wiadomości, 28.09.2017, 19:30',
-            'description': 'Wydanie główne codziennego serwisu informacyjnego.'
+            'title': 'Dzieci na sprzedaż dla homoseksualistów',
+            'description': 'md5:7d318eef04e55ddd9f87a8488ac7d590',
+            'age_limit': 12,
+            'duration': 259,
+            'thumbnail': r're:https://.+',
         },
-        'skip': 'HTTP Error 404: Not Found',
+        'add_ie': ['TVPEmbed'],
+    }, {
+        # TVPlayer 2 in client-side rendered website (regional; window.__newsData)
+        'url': 'https://warszawa.tvp.pl/25804446/studio-yayo',
+        'info_dict': {
+            'id': '25804446',
+            'ext': 'mp4',
+            'title': 'Studio Yayo',
+            'upload_date': '20160616',
+            'timestamp': 1466075700,
+            'age_limit': 0,
+            'duration': 20,
+            'thumbnail': r're:https://.+',
+        },
+        'add_ie': ['TVPEmbed'],
+        'skip': 'Video is geo restricted',
+    }, {
+        # TVPlayer 2 in client-side rendered website (tvp.info; window.__videoData)
+        'url': 'https://www.tvp.info/52880236/09042021-0800',
+        'info_dict': {
+            'id': '52880236',
+            'ext': 'mp4',
+            'title': '09.04.2021, 08:00',
+            'age_limit': 0,
+            'thumbnail': r're:https://.+',
+        },
+        'add_ie': ['TVPEmbed'],
+        'skip': 'Video is geo restricted',
+    }, {
+        # client-side rendered (regional) program (playlist) page
+        'url': 'https://opole.tvp.pl/9660819/rozmowa-dnia',
+        'info_dict': {
+            'id': '9660819',
+            'description': 'Od poniedziałku do piątku o 18:55',
+            'title': 'Rozmowa dnia',
+        },
+        'playlist_mincount': 1800,
+        'params': {
+            'skip_download': True,
+        }
+    }, {
+        # ABC-specific video embeding
+        # moved to https://bajkowakraina.tvp.pl/wideo/50981130,teleranek,51027049,zubr,51116450
+        'url': 'https://abc.tvp.pl/48636269/zubry-odc-124',
+        'info_dict': {
+            'id': '48320456',
+            'ext': 'mp4',
+            'title': 'Teleranek, Żubr',
+        },
+        'skip': 'Video gone: Nie znaleziono obiektu',
+    }, {
+        # yet another vue page
+        'url': 'https://jp2.tvp.pl/46925618/filmy',
+        'info_dict': {
+            'id': '46925618',
+            'title': 'Filmy',
+        },
+        'playlist_mincount': 19,
+    }, {
+        # redirect
+        'url': 'https://vod.tvp.pl/48463890/wadowickie-spotkania-z-janem-pawlem-ii',
+        'info_dict': {
+            'id': '295157',
+            'title': 'Wadowickie spotkania z Janem Pawłem II',
+        },
+        'playlist_mincount': 12,
+        'add_ie': ['TVPEmbed', 'TVPVODSeries'],
     }, {
         'url': 'http://vod.tvp.pl/seriale/obyczajowe/na-sygnale/sezon-2-27-/odc-39/17834272',
         'only_matching': True,
@@ -66,31 +288,212 @@ class TVPIE(InfoExtractor):
     }, {
         'url': 'http://www.tvp.info/25511919/trwa-rewolucja-wladza-zdecydowala-sie-na-pogwalcenie-konstytucji',
         'only_matching': True,
+    }, {
+        'url': 'https://tvp.info/49193823/teczowe-flagi-na-pomnikach-prokuratura-wszczela-postepowanie-wieszwiecej',
+        'only_matching': True,
+    }, {
+        'url': 'https://www.tvpparlament.pl/retransmisje-vod/inne/wizyta-premiera-mateusza-morawieckiego-w-firmie-berotu-sp-z-oo/48857277',
+        'only_matching': True,
+    }, {
+        'url': 'https://tvpworld.com/48583640/tescos-polish-business-bought-by-danish-chain-netto',
+        'only_matching': True,
+    }, {
+        'url': 'https://polandin.com/47942651/pln-10-billion-in-subsidies-transferred-to-companies-pm',
+        'only_matching': True,
     }]
 
+    def _parse_vue_website_data(self, webpage, page_id):
+        website_data = self._search_regex([
+            # website - regiony, tvp.info
+            # directory - jp2.tvp.pl
+            r'window\s*\.\s*__(?:website|directory)Data\s*=\s*({[\s\S]+?});',
+        ], webpage, 'website data')
+        if not website_data:
+            return None
+        return self._parse_json(website_data, page_id, transform_source=js_to_json)
+
+    def _extract_vue_video(self, video_data, page_id=None):
+        if isinstance(video_data, compat_str):
+            video_data = self._parse_json(video_data, page_id, transform_source=js_to_json)
+        video_id = txt_or_none(video_data.get('_id')) or page_id
+        if not video_id:
+            return
+        is_website = video_data.get('type') == 'website'
+        if is_website:
+            url = video_data['url']
+            fucked_up_url_parts = re.match(r'https?://vod\.tvp\.pl/(\d+)/([^/?#]+)', url)
+            if fucked_up_url_parts:
+                url = 'https://vod.tvp.pl/website/' + ','.join(fucked_up_url_parts.group(2, 1))
+        else:
+            url = 'tvp:' + video_id
+        return {
+            '_type': 'url_transparent',
+            'id': video_id,
+            'url': url,
+            'ie_key': (TVPIE if is_website else TVPEmbedIE).ie_key(),
+            'title': txt_or_none(video_data.get('title')),
+            'description': txt_or_none(video_data.get('lead')),
+            'timestamp': int_or_none(video_data.get('release_date_long')),
+            'duration': int_or_none(video_data.get('duration')),
+            'thumbnails': traverse_obj(video_data, ('image', (None, Ellipsis), 'url'), expected_type=url_or_none) or None,
+        }
+
+    def _handle_vuejs_page(self, url, webpage, page_id):
+        # vue client-side rendered sites (all regional pages + tvp.info)
+        video_data = self._search_regex([
+            r'window\.__(?:news|video)Data\s*=\s*({(?:.|\s)+?})\s*;',
+        ], webpage, 'video data', default=None)
+        if video_data:
+            video_data = self._extract_vue_video(video_data, page_id=page_id)
+            if video_data:
+                return self._extract_vue_video(video_data, page_id=page_id)
+        else:
+            # paged playlists
+            website_data = self._parse_vue_website_data(webpage, page_id)
+            if website_data:
+                entries = self._vuejs_entries(url, website_data, page_id)
+
+                return {
+                    '_type': 'playlist',
+                    'id': page_id,
+                    'title': txt_or_none(website_data.get('title')),
+                    'description': txt_or_none(website_data.get('lead')),
+                    'entries': entries,
+                }
+        raise ExtractorError('Could not extract video/website data')
+
+    def _vuejs_entries(self, url, website_data, page_id):
+
+        def extract_videos(wd):
+            for video in traverse_obj(wd, (None, ('latestVideo', (('videos', 'items'), Ellipsis)))):
+                video = self._extract_vue_video(video)
+                if video:
+                    yield video
+
+        for from_ in extract_videos(website_data):
+            yield from_
+
+        if website_data.get('items_total_count') > website_data.get('items_per_page'):
+            for page in itertools.count(2):
+                page_website_data = self._parse_vue_website_data(
+                    self._download_webpage(url, page_id, note='Downloading page #%d' % page,
+                                           query={'page': page}),
+                    page_id)
+                if not page_website_data.get('videos') and not page_website_data.get('items'):
+                    break
+                for from_ in extract_videos(page_website_data):
+                    yield from_
+
     def _real_extract(self, url):
         page_id = self._match_id(url)
-        webpage = self._download_webpage(url, page_id)
-        video_id = self._search_regex([
+        webpage, urlh = self._download_webpage_handle(url, page_id, expected_status=404)
+
+        # The URL may redirect to a VOD
+        # example: https://vod.tvp.pl/48463890/wadowickie-spotkania-z-janem-pawlem-ii
+        for ie_cls in (TVPVODSeriesIE, TVPVODVideoIE):
+            if ie_cls.suitable(urlh.url):
+                return self.url_result(urlh.url, ie=ie_cls.ie_key(), video_id=page_id)
+
+        if urlh.getcode() == 404:
+            raise compat_HTTPError(url, 404, 'HTTP Error 404: Not Found', urlh.headers, urlh)
+
+        if re.search(
+                r'window\s*\.\s*__(?:video|news|website|directory)Data\s*=',
+                webpage):
+            return self._handle_vuejs_page(url, webpage, page_id)
+
+        # classic server-side rendered sites
+        video_id = self._search_regex((
+            r'<iframe[^>]+src="[^"]*?embed\.php\?(?:[^&]+&)*ID=(\d+)',
             r'<iframe[^>]+src="[^"]*?object_id=(\d+)',
             r"object_id\s*:\s*'(\d+)'",
-            r'data-video-id="(\d+)"'], webpage, 'video id', default=page_id)
+            r'data-video-id="(\d+)"',
+            # abc.tvp.pl - somehow there are more than one video IDs that seem to be the same video?
+            # the first one is referenced to as "copyid", and seems to be unused by the website
+            r'<script>\s*tvpabc\.video\.init\(\s*\d+,\s*(\d+)\s*\)\s*</script>',
+        ), webpage, 'video id', default=page_id)
         return {
             '_type': 'url_transparent',
             'url': 'tvp:' + video_id,
             'description': self._og_search_description(
-                webpage, default=None) or self._html_search_meta(
-                'description', webpage, default=None),
+                webpage, default=None) or (self._html_search_meta(
+                    'description', webpage, default=None)
+                    if '//s.tvp.pl/files/portal/v' in webpage else None),
             'thumbnail': self._og_search_thumbnail(webpage, default=None),
             'ie_key': 'TVPEmbed',
         }
 
 
+class TVPStreamIE(InfoExtractor):
+    IE_NAME = 'tvp:stream'
+    _VALID_URL = r'(?:tvpstream:|https?://(?:tvpstream\.vod|stream)\.tvp\.pl/(?:\?(?:[^&]+[&;])*channel_id=)?)(?P<id>\d*)'
+    _TESTS = [{
+        'url': 'https://stream.tvp.pl/?channel_id=56969941',
+        'only_matching': True,
+    }, {
+        'url': 'https://tvpstream.vod.tvp.pl/?channel_id=1455',
+        'info_dict': {
+            'id': r're:\d+',
+            'title': r're:\S.*',
+            'ext': 'mp4',
+        },
+        'params': {
+            'skip_download': 'm3u8',
+        },
+        'add_ie': ['TVPEmbed'],
+    }, {
+        'url': 'tvpstream:39821455',
+        'only_matching': True,
+    }, {
+        # the default stream when you provide no channel_id, most probably TVP Info
+        'url': 'tvpstream:',
+        'only_matching': True,
+    }, {
+        'url': 'https://tvpstream.vod.tvp.pl/',
+        'only_matching': True,
+    }]
+
+    def _real_extract(self, url):
+        channel_id = self._match_id(url)
+        channel_url = self._proto_relative_url('//stream.tvp.pl/?channel_id=%s' % channel_id or 'default')
+        webpage = self._download_webpage(channel_url, channel_id or 'default', 'Downloading channel webpage')
+        channels = self._search_json(
+            r'window\s*\.\s*__channels\s*=', webpage, 'channel list', channel_id,
+            contains_pattern=r'\[\s*\{[\s\S]+}\s*]')
+        channel = traverse_obj(channels, (lambda _, v: channel_id == compat_str(v['id'])), get_all=False) if channel_id else channels[0]
+        audition = traverse_obj(channel, ('items', lambda _, v: v['is_live'] is True), get_all=False)
+        return {
+            '_type': 'url_transparent',
+            'id': channel_id or channel['id'],
+            'url': 'tvp:%s' % (audition['video_id'], ),
+            'title': audition.get('title'),
+            'alt_title': channel.get('title'),
+            'is_live': True,
+            'ie_key': 'TVPEmbed',
+        }
+
+
 class TVPEmbedIE(InfoExtractor):
     IE_NAME = 'tvp:embed'
     IE_DESC = 'Telewizja Polska'
-    _VALID_URL = r'(?:tvp:|https?://[^/]+\.tvp\.(?:pl|info)/sess/tvplayer\.php\?.*?object_id=)(?P<id>\d+)'
-
+    # XFF is not effective
+    _GEO_BYPASS = False
+    _VALID_URL_PAT = (
+        r'''
+            (?:
+                tvp:
+                |https?://
+                    (?:[^/]+\.)?
+                    (?:tvp(?:parlament)?\.pl|tvp\.info|tvpworld\.com|swipeto\.pl)/
+                    (?:sess/
+                            (?:tvplayer\.php\?.*?object_id
+                            |TVPlayer2/(?:embed|api)\.php\?.*[Ii][Dd])
+                        |shared/details\.php\?.*?object_id)
+                    =)
+            (?P<id>\d+)
+        ''')
+    _VALID_URL = '(?x)' + _VALID_URL_PAT
+    _EMBED_REGEX = [r'(?x)<iframe[^>]+?src=(["\'])(?P<url>{0})'.format(_VALID_URL_PAT)]
     _TESTS = [{
         'url': 'tvp:194536',
         'md5': 'a21eb0aa862f25414430f15fdfb9e76c',
@@ -98,9 +501,16 @@ class TVPEmbedIE(InfoExtractor):
             'id': '194536',
             'ext': 'mp4',
             'title': 'Czas honoru, odc. 13 – Władek',
+            'description': 'md5:76649d2014f65c99477be17f23a4dead',
+            'age_limit': 12,
+            'duration': 2652,
+            'series': 'Czas honoru',
+            'episode': 'Episode 13',
+            'episode_number': 13,
+            'season': 'sezon 1',
+            'thumbnail': r're:https://.+',
         },
     }, {
-        # not available
         'url': 'http://www.tvp.pl/sess/tvplayer.php?object_id=22670268',
         'md5': '8c9cd59d16edabf39331f93bf8a766c7',
         'info_dict': {
@@ -108,7 +518,28 @@ class TVPEmbedIE(InfoExtractor):
             'ext': 'mp4',
             'title': 'Panorama, 07.12.2015, 15:40',
         },
-        'skip': 'Transmisja została zakończona lub materiał niedostępny',
+        'skip': 'Nie znaleziono obiektu',
+    }, {
+        'url': 'https://www.tvp.pl/sess/tvplayer.php?object_id=51247504&amp;autoplay=false',
+        'info_dict': {
+            'id': '51247504',
+            'ext': 'mp4',
+            'title': 'Razmova 091220',
+            'duration': 876,
+            'age_limit': 0,
+            'thumbnail': r're:https://.+',
+        },
+    }, {
+        # TVPlayer2 embed URL
+        'url': 'https://tvp.info/sess/TVPlayer2/embed.php?ID=50595757',
+        'only_matching': True,
+    }, {
+        'url': 'https://wiadomosci.tvp.pl/sess/TVPlayer2/api.php?id=51233452',
+        'only_matching': True,
+    }, {
+        # pulsembed on dziennik.pl
+        'url': 'https://www.tvp.pl/shared/details.php?copy_id=52205981&object_id=52204505&autoplay=false&is_muted=false&allowfullscreen=true&template=external-embed/video/iframe-video.html',
+        'only_matching': True,
     }, {
         'url': 'tvp:22670268',
         'only_matching': True,
@@ -117,136 +548,272 @@ class TVPEmbedIE(InfoExtractor):
     def _real_extract(self, url):
         video_id = self._match_id(url)
 
+        # could be anything that is a valid JS function name
+        callback = random.choice((
+            'jebac_pis',
+            'jebacpis',
+            'ziobro',
+            'sasin70',
+            'sasin_przejebal_70_milionow_PLN',
+            'tvp_is_a_state_propaganda_service',
+        ))
         webpage = self._download_webpage(
-            'http://www.tvp.pl/sess/tvplayer.php?object_id=%s' % video_id, video_id)
+            ('https://www.tvp.pl/sess/TVPlayer2/api.php?id=%s'
+             + '&@method=getTvpConfig&@callback=%s') % (video_id, callback), video_id)
 
-        error = self._html_search_regex(
-            r'(?s)<p[^>]+\bclass=["\']notAvailable__text["\'][^>]*>(.+?)</p>',
-            webpage, 'error', default=None) or clean_html(
-            get_element_by_attribute('class', 'msg error', webpage))
-        if error:
-            raise ExtractorError('%s said: %s' % (
-                self.IE_NAME, clean_html(error)), expected=True)
+        # stripping JSONP padding
+        null, datastr = self._search_regex(
+            r'\s%s\s*\(\s*(?P<null>null\s*,\s*)?(?P<json>(?(null)\[\s*)?\{(?:[\s\S]+)}(?(null)]\s*))\)\s*;' % (re.escape(callback), ),
+            webpage, 'JSON API result', group=('null', 'json'))
+        data = self._parse_json(datastr, video_id, fatal=False)
+        if null:
+            error_desc = traverse_obj(data, (0, 'desc'), expected_type=compat_str)
+            if error_desc == 'Obiekt wymaga płatności':
+                error_desc = 'Video requires payment and log-in, but log-in is not implemented'
+            raise ExtractorError(error_desc or 'unexpected JSON error', expected=error_desc)
 
-        title = self._search_regex(
-            r'name\s*:\s*([\'"])Title\1\s*,\s*value\s*:\s*\1(?P<title>.+?)\1',
-            webpage, 'title', group='title')
-        series_title = self._search_regex(
-            r'name\s*:\s*([\'"])SeriesTitle\1\s*,\s*value\s*:\s*\1(?P<series>.+?)\1',
-            webpage, 'series', group='series', default=None)
-        if series_title:
-            title = '%s, %s' % (series_title, title)
+        content = data['content']
+        info = traverse_obj(content, 'info', expected_type=dict)
 
-        thumbnail = self._search_regex(
-            r"poster\s*:\s*'([^']+)'", webpage, 'thumbnail', default=None)
+        if traverse_obj(info, 'isGeoBlocked', expected_type=bool):
+            # actual country list is not provided, we just assume it's always available in PL
+            self.raise_geo_restricted(countries=['PL'])
 
-        video_url = self._search_regex(
-            r'0:{src:([\'"])(?P<url>.*?)\1', webpage,
-            'formats', group='url', default=None)
-        if not video_url or 'material_niedostepny.mp4' in video_url:
-            video_url = self._download_json(
-                'http://www.tvp.pl/pub/stat/videofileinfo?video_id=%s' % video_id,
-                video_id)['video_url']
+        is_live = traverse_obj(info, 'isLive', expected_type=bool)
 
         formats = []
-        video_url_base = self._search_regex(
-            r'(https?://.+?/video)(?:\.(?:ism|f4m|m3u8)|-\d+\.mp4)',
-            video_url, 'video base url', default=None)
-        if video_url_base:
-            # TODO: <Group> found instead of <AdaptationSet> in MPD manifest.
-            # It's not mentioned in MPEG-DASH standard. Figure that out.
-            # formats.extend(self._extract_mpd_formats(
-            #     video_url_base + '.ism/video.mpd',
-            #     video_id, mpd_id='dash', fatal=False))
-            formats.extend(self._extract_ism_formats(
-                video_url_base + '.ism/Manifest',
-                video_id, 'mss', fatal=False))
-            formats.extend(self._extract_f4m_formats(
-                video_url_base + '.ism/video.f4m',
-                video_id, f4m_id='hds', fatal=False))
-            m3u8_formats = self._extract_m3u8_formats(
-                video_url_base + '.ism/video.m3u8', video_id,
-                'mp4', 'm3u8_native', m3u8_id='hls', fatal=False)
-            self._sort_formats(m3u8_formats)
-            m3u8_formats = list(filter(
-                lambda f: f.get('vcodec') != 'none', m3u8_formats))
-            formats.extend(m3u8_formats)
-            for i, m3u8_format in enumerate(m3u8_formats, 2):
-                http_url = '%s-%d.mp4' % (video_url_base, i)
-                if self._is_valid_url(http_url, video_id):
-                    f = m3u8_format.copy()
-                    f.update({
-                        'url': http_url,
-                        'format_id': f['format_id'].replace('hls', 'http'),
-                        'protocol': 'http',
-                    })
-                    formats.append(f)
-        else:
-            formats = [{
-                'format_id': 'direct',
-                'url': video_url,
-                'ext': determine_ext(video_url, 'mp4'),
-            }]
+        for file in traverse_obj(content, ('files', Ellipsis), expected_type=dict):
+            video_url = url_or_none(file.get('url'))
+            if not video_url:
+                continue
+            ext = determine_ext(video_url, None)
+            if ext == 'm3u8':
+                formats.extend(self._extract_m3u8_formats(
+                    video_url, video_id, ext='mp4', m3u8_id='hls',
+                    fatal=False, live=is_live))
+            elif ext == 'mpd':
+                if is_live:
+                    # doesn't work with either ffmpeg or native downloader
+                    continue
+                formats.extend(self._extract_mpd_formats(video_url, video_id, mpd_id='dash', fatal=False))
+            elif ext == 'f4m':
+                formats.extend(self._extract_f4m_formats(video_url, video_id, f4m_id='hds', fatal=False))
+            elif video_url.endswith('.ism/manifest'):
+                formats.extend(self._extract_ism_formats(video_url, video_id, ism_id='mss', fatal=False))
+            elif ext == 'ism':
+                if '.ism/manifest' in video_url:
+                    formats.extend(self._extract_ism_formats(video_url, video_id, ism_id='mss', fatal=False))
+            else:
+                # mp4, wmv or something
+                quality = traverse_obj(file, 'quality', expected_type=dict) or {}
+                formats.append({
+                    'format_id': 'direct',
+                    'url': video_url,
+                    'ext': ext or file.get('type'),
+                    'fps': int_or_none(quality.get('fps')),
+                    'tbr': int_or_none(quality.get('bitrate'), scale=1000),
+                    'width': int_or_none(quality.get('width')),
+                    'height': int_or_none(quality.get('height')),
+                })
 
         self._sort_formats(formats)
 
-        return {
+        title = traverse_obj(info, 'subtitle', 'title', 'seoTitle', expected_type=txt_or_none)
+        # `seoDescription` may be Falsen
+        description = traverse_obj(info, 'description', 'seoDescription',
+                                   expected_type=lambda x: txt_or_none(x or None))
+        thumbnails = []
+        for thumb in traverse_obj(content, ('posters', Ellipsis), expected_type=dict):
+            thumb_url = thumb.get('src')
+            if not thumb_url or '{width}' in thumb_url or '{height}' in thumb_url:
+                continue
+            thumbnails.append({
+                'url': thumb.get('src'),
+                'width': thumb.get('width'),
+                'height': thumb.get('height'),
+            })
+        age_limit = traverse_obj(info, ('ageGroup', 'minAge'), expected_type=int)
+        if age_limit == 1:
+            age_limit = 0
+        duration = traverse_obj(info, 'duration', expected_type=int) if not is_live else None
+
+        subtitles = {}
+        for sub in traverse_obj(content, ('subtitles', Ellipsis), expected_type=dict):
+            if not (sub.get('url') and sub.get('lang')):
+                continue
+            subtitles.setdefault(sub['lang'], []).append({
+                'url': sub['url'],
+                'ext': sub.get('type'),
+            })
+
+        info_dict = {
             'id': video_id,
             'title': title,
-            'thumbnail': thumbnail,
+            'description': description,
+            'thumbnails': thumbnails,
+            'age_limit': age_limit,
+            'is_live': is_live,
+            'duration': duration,
             'formats': formats,
+            'subtitles': subtitles,
+        }
+
+        # vod.tvp.pl
+        if traverse_obj(info, 'vortalName') == 'vod':
+            info_dict.update({
+                'title': '%s, %s' % (info.get('title'), info.get('subtitle')),
+                'series': info.get('title'),
+                'season': info.get('season'),
+                'episode_number': info.get('episode') or None,
+            })
+
+        return info_dict
+
+
+class TVPVODBaseIE(InfoExtractor):
+    _API_BASE_URL = 'https://vod.tvp.pl/api/products/'
+
+    def _call_api(self, resource, video_id, **kwargs):
+        return self._download_json(
+            self._API_BASE_URL + resource, video_id,
+            query={'lang': 'pl', 'platform': 'BROWSER'}, **kwargs)
+
+    def _parse_video(self, video):
+        video_id = traverse_obj(video, 'externalUid', expected_type=txt_or_none)
+
+        if not video_id:
+            return None
+
+        return {
+            '_type': 'url',
+            'url': 'tvp:' + video_id,
+            'ie_key': TVPEmbedIE.ie_key(),
+            'title': video.get('title'),
+            'description': traverse_obj(video, ('lead', 'description'), expected_type=txt_or_none),
+            'age_limit': int_or_none(video.get('rating')),
+            'duration': int_or_none(video.get('duration')),
         }
 
 
-class TVPWebsiteIE(InfoExtractor):
-    IE_NAME = 'tvp:series'
-    _VALID_URL = r'https?://vod\.tvp\.pl/website/(?P<display_id>[^,]+),(?P<id>\d+)'
+class TVPVODVideoIE(TVPVODBaseIE):
+    IE_NAME = 'tvp:vod'
+    _VALID_URL = r'https?://vod\.tvp\.pl/[a-z\d-]+,\d+/[a-z\d-]+(?<!-odcinki)(?:-odcinki,\d+/odcinek-\d+,S\d+E\d+)?,(?P<id>\d+)(?:\?[^#]+)?(?:#.+)?$'
 
     _TESTS = [{
-        # series
-        'url': 'https://vod.tvp.pl/website/lzy-cennet,38678312/video',
+        'url': 'https://vod.tvp.pl/dla-dzieci,24/laboratorium-alchemika-odcinki,309338/odcinek-24,S01E24,311357',
         'info_dict': {
-            'id': '38678312',
-        },
-        'playlist_count': 115,
-    }, {
-        # film
-        'url': 'https://vod.tvp.pl/website/gloria,35139666',
-        'info_dict': {
-            'id': '36637049',
+            'id': '60468609',
             'ext': 'mp4',
-            'title': 'Gloria, Gloria',
-        },
-        'params': {
-            'skip_download': True,
+            'title': 'Laboratorium alchemika, Tusze termiczne. Jak zobaczyć niewidoczne. Odcinek 24',
+            'description': 'md5:1d4098d3e537092ccbac1abf49b7cd4c',
+            'duration': 300,
+            'episode_number': 24,
+            'episode': 'Episode 24',
+            'age_limit': 0,
+            'series': 'Laboratorium alchemika',
+            'thumbnail': 're:https://.+',
         },
         'add_ie': ['TVPEmbed'],
     }, {
-        'url': 'https://vod.tvp.pl/website/lzy-cennet,38678312',
+        'url': 'https://vod.tvp.pl/filmy-dokumentalne,163/ukrainski-sluga-narodu,339667',
+        'info_dict': {
+            'id': '51640077',
+            'ext': 'mp4',
+            'title': 'Ukraiński sługa narodu, Ukraiński sługa narodu',
+            'series': 'Ukraiński sługa narodu',
+            'description': 'md5:b7940c0a8e439b0c81653a986f544ef3',
+            'age_limit': 12,
+            'duration': 3051,
+            'thumbnail': 're:https://.+',
+        },
+        'add_ie': ['TVPEmbed'],
+    }, {
+        # new URL format
+        'url': 'https://vod.tvp.pl/seriale,18/czas-honoru-odcinki,292065/odcinek-13,S01E13,313867',
+        'md5': 'a21eb0aa862f25414430f15fdfb9e76c',
+        'info_dict': {
+            'id': '194536',
+            'ext': 'mp4',
+            'title': 'Czas honoru, odc. 13 – Władek',
+            'description': 'md5:76649d2014f65c99477be17f23a4dead',
+            'age_limit': 12,
+        },
+        'add_ie': ['TVPEmbed'],
+    }, {
+        'url': 'https://vod.tvp.pl/filmy-fabularne,136/rozlam,390638',
+        'only_matching': True,
+    }]
+
+    def _real_extract(self, url):
+        video_id = self._match_id(url)
+
+        video = self._parse_video(
+            self._call_api('vods/' + video_id, video_id))
+        if not video:
+            raise ExtractorError('No video data for ' + video_id)
+        return video
+
+
+class TVPVODSeriesIE(TVPVODBaseIE):
+    IE_NAME = 'tvp:vod:series'
+    _VALID_URL = r'''(?x)
+                    https?://vod\.tvp\.pl/
+                        seriale,(?P<cat>\d+)/
+                        (?P<display_id>[^,]+?)(?(cat)-odcinki),(?P<id>\d+)
+                        (?(cat)|(?P<video>/video)?)(?:[#?]|$)
+                '''
+    _VALID_URL = r'https?://vod\.tvp\.pl/(?P<display_id>[a-z\d-]+,\d+)/[a-z\d-]+-odcinki,(?P<id>\d+)(?:\?[^#]+)?(?:#.+)?$'
+
+    _TESTS = [{
+        # series
+        'url': 'https://vod.tvp.pl/seriale,18/ranczo-odcinki,316445',
+        #  series (old) - redirects to home page
+        #  'url': 'https://vod.tvp.pl/website/wspaniale-stulecie,17069012/video',
+        'info_dict': {
+            'id': '316445',
+            'title': 'Ranczo',
+            # 'description': 'md5:a7ccbe1296e6f32425cef17639f1b24b',
+            'age_limit': 12,
+            'categories': ['seriale'],
+        },
+        'playlist_mincount': 129,
+    }, {
+        'url': 'https://vod.tvp.pl/programy,88/rolnik-szuka-zony-odcinki,284514',
+        'only_matching': True,
+    }, {
+        'url': 'https://vod.tvp.pl/dla-dzieci,24/laboratorium-alchemika-odcinki,309338',
         'only_matching': True,
     }]
 
     def _entries(self, display_id, playlist_id):
-        url = 'https://vod.tvp.pl/website/%s,%s/video' % (display_id, playlist_id)
-        for page_num in itertools.count(1):
-            page = self._download_webpage(
-                url, display_id, 'Downloading page %d' % page_num,
-                query={'page': page_num})
+        season_path = 'vods/serials/%s/seasons' % (playlist_id, )
+        seasons = self._call_api(
+            season_path, playlist_id,
+            note='Downloading season list') or []
 
-            video_ids = orderedSet(re.findall(
-                r'<a[^>]+\bhref=["\']/video/%s,[^,]+,(\d+)' % display_id,
-                page))
-
-            if not video_ids:
-                break
-
-            for video_id in video_ids:
-                yield self.url_result(
-                    'tvp:%s' % video_id, ie=TVPEmbedIE.ie_key(),
-                    video_id=video_id)
+        for ii, season in enumerate(seasons, 1):
+            season_id = traverse_obj(season, 'id', expected_type=txt_or_none)
+            if not season_id:
+                continue
+            episodes = self._call_api(
+                '%s/%s/episodes' % (season_path, season_id), playlist_id,
+                note='Downloading episode list (season %d)' % ii)
+            for episode in episodes or []:
+                video_id = traverse_obj(episode, 'externalUid', expected_type=txt_or_none)
+                if video_id:
+                    yield self._parse_video(episode)
 
     def _real_extract(self, url):
-        mobj = re.match(self._VALID_URL, url)
-        display_id, playlist_id = mobj.group('display_id', 'id')
-        return self.playlist_result(
-            self._entries(display_id, playlist_id), playlist_id)
+        display_id, playlist_id = self._match_valid_url(url).group('display_id', 'id')
+        metadata = self._call_api(
+            'vods/serials/' + playlist_id, playlist_id,
+            note='Downloading serial metadata') or {}
+
+        pl = self.playlist_result(
+            self._entries(display_id, playlist_id), playlist_id, txt_or_none(metadata.get('title')))
+        pl.update({
+            'description': traverse_obj(metadata, ('description', 'lead'), expected_type=clean_html),
+            'categories': traverse_obj(metadata, ('mainCategory', (None, Ellipsis), 'name'), expected_type=txt_or_none),
+            'age_limit': traverse_obj(metadata, 'rating', expected_type=int),
+        })
+        return pl

youtube_dl/extractor/youtube.py

@@ -91,12 +91,12 @@ class YoutubeBaseInfoExtractor(InfoExtractor):
             'INNERTUBE_CONTEXT': {
                 'client': {
                     'clientName': 'IOS',
-                    'clientVersion': '19.45.4',
+                    'clientVersion': '20.10.4',
                     'deviceMake': 'Apple',
                     'deviceModel': 'iPhone16,2',
-                    'userAgent': 'com.google.ios.youtube/19.45.4 (iPhone16,2; U; CPU iOS 18_1_0 like Mac OS X;)',
+                    'userAgent': 'com.google.ios.youtube/20.10.4 (iPhone16,2; U; CPU iOS 18_3_2 like Mac OS X;)',
                     'osName': 'iPhone',
-                    'osVersion': '18.1.0.22B83',
+                    'osVersion': '18.3.2.22D82',
                 },
             },
             'INNERTUBE_CONTEXT_CLIENT_NAME': 5,
@@ -109,7 +109,7 @@ class YoutubeBaseInfoExtractor(InfoExtractor):
             'INNERTUBE_CONTEXT': {
                 'client': {
                     'clientName': 'MWEB',
-                    'clientVersion': '2.20241202.07.00',
+                    'clientVersion': '2.20250311.03.00',
                     # mweb previously did not require PO Token with this UA
                     'userAgent': 'Mozilla/5.0 (iPad; CPU OS 16_7_10 like Mac OS X) AppleWebKit/605.1.15 (KHTML, like Gecko) Version/16.6 Mobile/15E148 Safari/604.1,gzip(gfe)',
                 },
@@ -122,7 +122,8 @@ class YoutubeBaseInfoExtractor(InfoExtractor):
             'INNERTUBE_CONTEXT': {
                 'client': {
                     'clientName': 'TVHTML5',
-                    'clientVersion': '7.20241201.18.00',
+                    'clientVersion': '7.20250312.16.00',
+                    'userAgent': 'Mozilla/5.0 (ChromiumStylePlatform) Cobalt/Version',
                 },
             },
             'INNERTUBE_CONTEXT_CLIENT_NAME': 7,
@@ -132,7 +133,7 @@ class YoutubeBaseInfoExtractor(InfoExtractor):
             'INNERTUBE_CONTEXT': {
                 'client': {
                     'clientName': 'WEB',
-                    'clientVersion': '2.20241126.01.00',
+                    'clientVersion': '2.20250312.04.00',
                 },
             },
             'INNERTUBE_CONTEXT_CLIENT_NAME': 1,
@@ -691,7 +692,7 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
         'invidious': '|'.join(_INVIDIOUS_SITES),
     }
     _PLAYER_INFO_RE = (
-        r'/s/player/(?P<id>[a-zA-Z0-9_-]{8,})/player',
+        r'/s/player/(?P<id>[a-zA-Z0-9_-]{8,})//(?:tv-)?player',
         r'/(?P<id>[a-zA-Z0-9_-]{8,})/player(?:_ias\.vflset(?:/[a-zA-Z]{2,3}_[a-zA-Z]{2,3})?|-plasma-ias-(?:phone|tablet)-[a-z]{2}_[A-Z]{2}\.vflset)/base\.js$',
         r'\b(?P<id>vfl[a-zA-Z0-9_-]+)\b.*?\.js$',
     )
@@ -1851,12 +1852,22 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
 
         if func_code:
             return jsi, player_id, func_code
+        return self._extract_n_function_code_jsi(video_id, jsi, player_id)
 
-        func_name = self._extract_n_function_name(jscode)
+    def _extract_n_function_code_jsi(self, video_id, jsi, player_id=None):
+
+        var_ay = self._search_regex(
+            r'(?:[;\s]|^)\s*(var\s*[\w$]+\s*=\s*"(?:\\"|[^"])+"\s*\.\s*split\("\W+"\))(?=\s*[,;])',
+            jsi.code, 'useful values', default='')
+
+        func_name = self._extract_n_function_name(jsi.code)
 
         func_code = jsi.extract_function_code(func_name)
+        if var_ay:
+            func_code = (func_code[0], ';\n'.join((var_ay, func_code[1])))
 
-        self.cache.store('youtube-nsig', player_id, func_code)
+        if player_id:
+            self.cache.store('youtube-nsig', player_id, func_code)
         return jsi, player_id, func_code
 
     def _extract_n_function_from_code(self, jsi, func_code):

youtube_dl/jsinterp.py

@@ -1,10 +1,12 @@
 # coding: utf-8
 from __future__ import unicode_literals
 
+import calendar
 import itertools
 import json
 import operator
 import re
+import time
 
 from functools import update_wrapper, wraps
 
@@ -12,8 +14,10 @@ from .utils import (
     error_to_compat_str,
     ExtractorError,
     float_or_none,
+    int_or_none,
     js_to_json,
     remove_quotes,
+    str_or_none,
     unified_timestamp,
     variadic,
     write_string,
@@ -150,6 +154,7 @@ def _js_to_primitive(v):
     )
 
 
+# more exact: yt-dlp/yt-dlp#12110
 def _js_toString(v):
     return (
         'undefined' if v is JS_Undefined
@@ -158,7 +163,7 @@ def _js_toString(v):
         else 'null' if v is None
         # bool <= int: do this first
         else ('false', 'true')[v] if isinstance(v, bool)
-        else '{0:.7f}'.format(v).rstrip('.0') if isinstance(v, compat_numeric_types)
+        else re.sub(r'(?<=\d)\.?0*$', '', '{0:.7f}'.format(v)) if isinstance(v, compat_numeric_types)
         else _js_to_primitive(v))
 
 
@@ -404,6 +409,7 @@ class JSInterpreter(object):
     class Exception(ExtractorError):
         def __init__(self, msg, *args, **kwargs):
             expr = kwargs.pop('expr', None)
+            msg = str_or_none(msg, default='"None"')
             if expr is not None:
                 msg = '{0} in: {1!r:.100}'.format(msg.rstrip(), expr)
             super(JSInterpreter.Exception, self).__init__(msg, *args, **kwargs)
@@ -431,6 +437,7 @@ class JSInterpreter(object):
                 flags, _ = self.regex_flags(flags)
             # First, avoid https://github.com/python/cpython/issues/74534
             self.__self = None
+            pattern_txt = str_or_none(pattern_txt) or '(?:)'
             self.__pattern_txt = pattern_txt.replace('[[', r'[\[')
             self.__flags = flags
 
@@ -475,6 +482,73 @@ class JSInterpreter(object):
                 flags |= cls.RE_FLAGS[ch]
             return flags, expr[idx + 1:]
 
+    class JS_Date(object):
+        _t = None
+
+        @staticmethod
+        def __ymd_etc(*args, **kw_is_utc):
+            # args: year, monthIndex, day, hours, minutes, seconds, milliseconds
+            is_utc = kw_is_utc.get('is_utc', False)
+
+            args = list(args[:7])
+            args += [0] * (9 - len(args))
+            args[1] += 1  # month 0..11 -> 1..12
+            ms = args[6]
+            for i in range(6, 9):
+                args[i] = -1  # don't know
+            if is_utc:
+                args[-1] = 1
+            # TODO: [MDN] When a segment overflows or underflows its expected
+            # range, it usually "carries over to" or "borrows from" the higher segment.
+            try:
+                mktime = calendar.timegm if is_utc else time.mktime
+                return mktime(time.struct_time(args)) * 1000 + ms
+            except (OverflowError, ValueError):
+                return None
+
+        @classmethod
+        def UTC(cls, *args):
+            t = cls.__ymd_etc(*args, is_utc=True)
+            return _NaN if t is None else t
+
+        @staticmethod
+        def parse(date_str, **kw_is_raw):
+            is_raw = kw_is_raw.get('is_raw', False)
+
+            t = unified_timestamp(str_or_none(date_str), False)
+            return int(t * 1000) if t is not None else t if is_raw else _NaN
+
+        @staticmethod
+        def now(**kw_is_raw):
+            is_raw = kw_is_raw.get('is_raw', False)
+
+            t = time.time()
+            return int(t * 1000) if t is not None else t if is_raw else _NaN
+
+        def __init__(self, *args):
+            if not args:
+                args = [self.now(is_raw=True)]
+            if len(args) == 1:
+                if isinstance(args[0], JSInterpreter.JS_Date):
+                    self._t = int_or_none(args[0].valueOf(), default=None)
+                else:
+                    arg_type = _js_typeof(args[0])
+                    if arg_type == 'string':
+                        self._t = self.parse(args[0], is_raw=True)
+                    elif arg_type == 'number':
+                        self._t = int(args[0])
+            else:
+                self._t = self.__ymd_etc(*args)
+
+        def toString(self):
+            try:
+                return time.strftime('%a %b %0d %Y %H:%M:%S %Z%z', self._t).rstrip()
+            except TypeError:
+                return "Invalid Date"
+
+        def valueOf(self):
+            return _NaN if self._t is None else self._t
+
     @classmethod
     def __op_chars(cls):
         op_chars = set(';,[')
@@ -599,14 +673,15 @@ class JSInterpreter(object):
         except Exception as e:
             raise self.Exception('Failed to evaluate {left_val!r:.50} {op} {right_val!r:.50}'.format(**locals()), expr, cause=e)
 
-    def _index(self, obj, idx, allow_undefined=True):
+    def _index(self, obj, idx, allow_undefined=None):
         if idx == 'length' and isinstance(obj, list):
             return len(obj)
         try:
             return obj[int(idx)] if isinstance(obj, list) else obj[compat_str(idx)]
         except (TypeError, KeyError, IndexError) as e:
-            if allow_undefined:
-                # when is not allowed?
+            # allow_undefined is None gives correct behaviour
+            if allow_undefined or (
+                    allow_undefined is None and not isinstance(e, TypeError)):
                 return JS_Undefined
             raise self.Exception('Cannot get index {idx!r:.100}'.format(**locals()), expr=repr(obj), cause=e)
 
@@ -715,7 +790,7 @@ class JSInterpreter(object):
 
         new_kw, _, obj = expr.partition('new ')
         if not new_kw:
-            for klass, konstr in (('Date', lambda x: int(unified_timestamp(x, False) * 1000)),
+            for klass, konstr in (('Date', lambda *x: self.JS_Date(*x).valueOf()),
                                   ('RegExp', self.JS_RegExp),
                                   ('Error', self.Exception)):
                 if not obj.startswith(klass + '('):
@@ -1034,6 +1109,7 @@ class JSInterpreter(object):
                     'String': compat_str,
                     'Math': float,
                     'Array': list,
+                    'Date': self.JS_Date,
                 }
                 obj = local_vars.get(variable)
                 if obj in (JS_Undefined, None):
@@ -1086,6 +1162,8 @@ class JSInterpreter(object):
                         assertion(len(argvals) == 2, 'takes two arguments')
                         return argvals[0] ** argvals[1]
                     raise self.Exception('Unsupported Math method ' + member, expr=expr)
+                elif obj is self.JS_Date:
+                    return getattr(obj, member)(*argvals)
 
                 if member == 'split':
                     assertion(len(argvals) <= 2, 'takes at most two arguments')