Merge e5bfed6c77b47bdf07713d8d88c8b21247999ae7 into 420d53387cff54ea1fccca061438d59bdb50a39c

* from yt-dlp/yt-dlp#12313
* also fix d7c2708

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 +223,18 @@ _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/91201489/player_ias_tce.vflset/en_US/base.js',
+        'W9HJZKktxuYoDTqW', 'U48vOZHaeYS6vO',
+    ),
 ]
 
 
@@ -284,7 +296,7 @@ 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 +304,10 @@ 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/pornhub.py

@@ -7,30 +7,35 @@ import operator
 import re
 
 from .common import InfoExtractor
+
 from ..compat import (
     compat_HTTPError,
-    compat_str,
-    compat_urllib_request,
 )
 from .openload import PhantomJSwrapper
 from ..utils import (
+    clean_html,
     determine_ext,
+    extract_attributes,
     ExtractorError,
+    get_element_by_class,
+    get_element_by_id,
     int_or_none,
     merge_dicts,
-    NO_DEFAULT,
-    orderedSet,
+    parse_count,
     remove_quotes,
-    str_to_int,
+    remove_start,
+    T,
+    traverse_obj,
     update_url_query,
-    urlencode_postdata,
     url_or_none,
+    urlencode_postdata,
+    urljoin,
 )
 
 
 class PornHubBaseIE(InfoExtractor):
     _NETRC_MACHINE = 'pornhub'
-    _PORNHUB_HOST_RE = r'(?:(?P<host>pornhub(?:premium)?\.(?:com|net|org))|pornhubthbh7ap3u\.onion)'
+    _PORNHUB_HOST_RE = r'(?:(?P<host>pornhub(?:premium)?\.(?:com|net|org))|pornhubvybmsymdol4iibwgwtkpwmeyd6luq2gxajgjzfjvotyt5zhyd\.onion)'
 
     def _download_webpage_handle(self, *args, **kwargs):
         def dl(*args, **kwargs):
@@ -47,10 +52,7 @@ class PornHubBaseIE(InfoExtractor):
                 r'<body\b[^>]+\bonload=["\']go\(\)',
                 r'document\.cookie\s*=\s*["\']RNKEY=',
                 r'document\.location\.reload\(true\)')):
-            url_or_request = args[0]
-            url = (url_or_request.get_full_url()
-                   if isinstance(url_or_request, compat_urllib_request.Request)
-                   else url_or_request)
+            url = urlh.geturl()
             phantom = PhantomJSwrapper(self, required_version='2.0')
             phantom.get(url, html=webpage)
             webpage, urlh = dl(*args, **kwargs)
@@ -60,11 +62,17 @@ class PornHubBaseIE(InfoExtractor):
     def _real_initialize(self):
         self._logged_in = False
 
+    def _set_age_cookies(self, host):
+        self._set_cookie(host, 'age_verified', '1')
+        self._set_cookie(host, 'accessAgeDisclaimerPH', '1')
+        self._set_cookie(host, 'accessAgeDisclaimerUK', '1')
+        self._set_cookie(host, 'accessPH', '1')
+
     def _login(self, host):
         if self._logged_in:
             return
 
-        site = host.split('.')[0]
+        site = host.split('.', 1)[0]
 
         # Both sites pornhub and pornhubpremium have separate accounts
         # so there should be an option to provide credentials for both.
@@ -81,9 +89,9 @@ class PornHubBaseIE(InfoExtractor):
             login_url, None, 'Downloading %s login page' % site)
 
         def is_logged(webpage):
-            return any(re.search(p, webpage) for p in (
-                r'class=["\']signOut',
-                r'>Sign\s+[Oo]ut\s*<'))
+            return bool(
+                get_element_by_id('profileMenuDropdown', webpage)
+                or get_element_by_class('ph-icon-logout', webpage))
 
         if is_logged(login_page):
             self._logged_in = True
@@ -92,12 +100,12 @@ class PornHubBaseIE(InfoExtractor):
         login_form = self._hidden_inputs(login_page)
 
         login_form.update({
-            'username': username,
+            'email': username,
             'password': password,
         })
 
         response = self._download_json(
-            'https://www.%s/front/authenticate' % host, None,
+            'https://www.%s/front/authenticate' % host, 'login',
             'Logging in to %s' % site,
             data=urlencode_postdata(login_form),
             headers={
@@ -119,17 +127,12 @@ class PornHubBaseIE(InfoExtractor):
 
 
 class PornHubIE(PornHubBaseIE):
-    IE_DESC = 'PornHub and Thumbzilla'
-    _VALID_URL = r'''(?x)
-                    https?://
-                        (?:
-                            (?:[^/]+\.)?
-                            %s
-                            /(?:(?:view_video\.php|video/show)\?viewkey=|embed/)|
-                            (?:www\.)?thumbzilla\.com/video/
-                        )
-                        (?P<id>[\da-z]+)
-                    ''' % PornHubBaseIE._PORNHUB_HOST_RE
+    IE_DESC = 'PornHub'  # Thumbzilla -> Redtube.com, Modelhub -> uviu.com
+    _PORNHUB_PATH_RE = r'/(?:(?:view_video\.php%s)\?(?:.+&)?viewkey=%s)(?P<id>[\da-z]+)'
+    _VALID_URL = r'https?://(?:[^/]+\.)?%s%s' % (
+        PornHubBaseIE._PORNHUB_HOST_RE, _PORNHUB_PATH_RE % ('|video/show', '|embed/'))
+    _PORNHUB_PATH_RE = _PORNHUB_PATH_RE % ('', '')
+    _EMBED_REGEX = [r'<iframe\s[^>]*?src=["\'](?P<url>(?:https?:)?//(?:www\.)?pornhub(?:premium)?\.(?:com|net|org)/embed/[\da-z]+)']
     _TESTS = [{
         'url': 'http://www.pornhub.com/view_video.php?viewkey=648719015',
         'md5': 'a6391306d050e4547f62b3f485dd9ba9',
@@ -138,6 +141,7 @@ class PornHubIE(PornHubBaseIE):
             'ext': 'mp4',
             'title': 'Seductive Indian beauty strips down and fingers her pink pussy',
             'uploader': 'Babes',
+            'uploader_id': '/users/babes-com',
             'upload_date': '20130628',
             'timestamp': 1372447216,
             'duration': 361,
@@ -148,6 +152,10 @@ class PornHubIE(PornHubBaseIE):
             'age_limit': 18,
             'tags': list,
             'categories': list,
+            'cast': list,
+        },
+        'params': {
+            'format': '[format_id!^=hls]',
         },
     }, {
         # non-ASCII title
@@ -189,14 +197,27 @@ class PornHubIE(PornHubBaseIE):
             'categories': list,
             'subtitles': {
                 'en': [{
-                    "ext": 'srt'
-                }]
+                    'ext': 'srt',
+                }],
             },
         },
         'params': {
             'skip_download': True,
         },
         'skip': 'This video has been disabled',
+    }, {
+        'url': 'http://www.pornhub.com/view_video.php?viewkey=ph601dc30bae19a',
+        'info_dict': {
+            'id': 'ph601dc30bae19a',
+            'ext': 'mp4',
+            'timestamp': 1612564932,
+            'age_limit': 18,
+            'uploader': 'Projekt Melody',
+            'uploader_id': 'projekt-melody',
+            'upload_date': '20210205',
+            'title': '"Welcome to My Pussy Mansion" - CB Stream (02/03/21)',
+            'thumbnail': r're:https?://.+',
+        },
     }, {
         'url': 'http://www.pornhub.com/view_video.php?viewkey=ph557bbb6676d2d',
         'only_matching': True,
@@ -216,9 +237,6 @@ class PornHubIE(PornHubBaseIE):
         # private video
         'url': 'http://www.pornhub.com/view_video.php?viewkey=ph56fd731fce6b7',
         'only_matching': True,
-    }, {
-        'url': 'https://www.thumbzilla.com/video/ph56c6114abd99a/horny-girlfriend-sex',
-        'only_matching': True,
     }, {
         'url': 'http://www.pornhub.com/video/show?viewkey=648719015',
         'only_matching': True,
@@ -244,28 +262,36 @@ class PornHubIE(PornHubBaseIE):
         'url': 'https://www.pornhub.com/view_video.php?viewkey=ph5a9813bfa7156',
         'only_matching': True,
     }, {
-        'url': 'http://pornhubthbh7ap3u.onion/view_video.php?viewkey=ph5a9813bfa7156',
+        'url': 'http://pornhubvybmsymdol4iibwgwtkpwmeyd6luq2gxajgjzfjvotyt5zhyd.onion/view_video.php?viewkey=ph5a9813bfa7156',
         'only_matching': True,
     }]
 
-    @staticmethod
-    def _extract_urls(webpage):
-        return re.findall(
-            r'<iframe[^>]+?src=["\'](?P<url>(?:https?:)?//(?:www\.)?pornhub(?:premium)?\.(?:com|net|org)/embed/[\da-z]+)',
-            webpage)
+    @classmethod
+    def _extract_urls(cls, webpage):
+        def yield_urls():
+            for p in cls._EMBED_REGEX:
+                for from_ in re.finditer(p, webpage):
+                    yield from_.group('url')
+
+        return list(yield_urls())
 
     def _extract_count(self, pattern, webpage, name):
-        return str_to_int(self._search_regex(
+        return parse_count(self._search_regex(
             pattern, webpage, '%s count' % name, fatal=False))
 
     def _real_extract(self, url):
-        mobj = re.match(self._VALID_URL, url)
+        for _ in range(2):
+            mobj = self._match_valid_url(url)
+            video_id = mobj.group('id') if mobj else self._generic_id(url)
+            _, urlh = self._download_webpage_handle(url, video_id)
+            if url == urlh.geturl():
+                break
+            url = urlh.geturl()
+
         host = mobj.group('host') or 'pornhub.com'
-        video_id = mobj.group('id')
 
         self._login(host)
-
-        self._set_cookie(host, 'age_verified', '1')
+        self._set_age_cookies(host)
 
         def dl_webpage(platform):
             self._set_cookie(host, 'platform', platform)
@@ -276,7 +302,7 @@ class PornHubIE(PornHubBaseIE):
         webpage = dl_webpage('pc')
 
         error_msg = self._html_search_regex(
-            (r'(?s)<div[^>]+class=(["\'])(?:(?!\1).)*\b(?:removed|userMessageSection)\b(?:(?!\1).)*\1[^>]*>(?P<error>.+?)</div>',
+            (r'(?s)<div[^>]+class=("|\')(?:(?!\1).)*\b(?:removed|userMessageSection)\b(?:(?!\1).)*\1[^>]*>(?P<error>.+?)</div>',
              r'(?s)<section[^>]+class=["\']noVideo["\'][^>]*>(?P<error>.+?)</section>'),
             webpage, 'error message', default=None, group='error')
         if error_msg:
@@ -285,9 +311,9 @@ class PornHubIE(PornHubBaseIE):
                 'PornHub said: %s' % error_msg,
                 expected=True, video_id=video_id)
 
-        if any(re.search(p, webpage) for p in (
-                r'class=["\']geoBlocked["\']',
-                r'>\s*This content is unavailable in your country')):
+        if bool(get_element_by_class('geoBlocked', webpage)
+                or self._search_regex(
+                    r'>\s*This content is (unavailable) in your country', webpage, 'geo-restriction', default=False)):
             self.raise_geo_restricted()
 
         # video_title from flashvars contains whitespace instead of non-ASCII (see
@@ -304,36 +330,34 @@ class PornHubIE(PornHubBaseIE):
         video_urls_set = set()
         subtitles = {}
 
-        flashvars = self._parse_json(
-            self._search_regex(
-                r'var\s+flashvars_\d+\s*=\s*({.+?});', webpage, 'flashvars', default='{}'),
-            video_id)
-        if flashvars:
-            subtitle_url = url_or_none(flashvars.get('closedCaptionsFile'))
-            if subtitle_url:
-                subtitles.setdefault('en', []).append({
-                    'url': subtitle_url,
-                    'ext': 'srt',
-                })
-            thumbnail = flashvars.get('image_url')
-            duration = int_or_none(flashvars.get('video_duration'))
-            media_definitions = flashvars.get('mediaDefinitions')
-            if isinstance(media_definitions, list):
-                for definition in media_definitions:
-                    if not isinstance(definition, dict):
-                        continue
-                    video_url = definition.get('videoUrl')
-                    if not video_url or not isinstance(video_url, compat_str):
-                        continue
-                    if video_url in video_urls_set:
-                        continue
-                    video_urls_set.add(video_url)
-                    video_urls.append(
-                        (video_url, int_or_none(definition.get('quality'))))
-        else:
-            thumbnail, duration = [None] * 2
+        def add_video_url(video_url, quality=None):
+            v_url = url_or_none(video_url)
+            if not v_url:
+                return
+            if v_url in video_urls_set:
+                return
+            video_urls.append((v_url, quality))
+            video_urls_set.add(v_url)
 
-        def extract_js_vars(webpage, pattern, default=NO_DEFAULT):
+        flashvars = self._search_json(r'var\s+flashvars_\d+\s*=', webpage, 'flashvars', video_id)
+        flashvars = traverse_obj(flashvars, {
+            'closedCaptionsFile': ('closedCaptionsFile', T(url_or_none)),
+            'image_url': ('image_url', T(url_or_none)),
+            'video_duration': ('video_duration', T(int_or_none)),
+            'mediaDefinitions': ('mediaDefinitions', lambda _, v: v['videoUrl']),
+        }) or {}
+        subtitle_url = flashvars.get('closedCaptionsFile')
+        if subtitle_url:
+            subtitles.setdefault('en', []).append({
+                'url': subtitle_url,
+                'ext': 'srt',
+            })
+        thumbnail = flashvars.get('image_url')
+        duration = flashvars.get('video_duration')
+        for definition in flashvars.get('mediaDefinitions') or []:
+            add_video_url(definition['videoUrl'], int_or_none(definition.get('quality')))
+
+        def extract_js_vars(webpage, pattern, default=None):
             assignments = self._search_regex(
                 pattern, webpage, 'encoded url', default=default)
             if not assignments:
@@ -363,51 +387,33 @@ class PornHubIE(PornHubBaseIE):
                 js_vars[vname] = parse_js_value(value)
             return js_vars
 
-        def add_video_url(video_url):
-            v_url = url_or_none(video_url)
-            if not v_url:
-                return
-            if v_url in video_urls_set:
-                return
-            video_urls.append((v_url, None))
-            video_urls_set.add(v_url)
-
         def parse_quality_items(quality_items):
             q_items = self._parse_json(quality_items, video_id, fatal=False)
-            if not isinstance(q_items, list):
-                return
-            for item in q_items:
-                if isinstance(item, dict):
-                    add_video_url(item.get('url'))
+            for v_url in traverse_obj(q_items, (Ellipsis, 'url')):
+                add_video_url(v_url)
 
         if not video_urls:
             FORMAT_PREFIXES = ('media', 'quality', 'qualityItems')
             js_vars = extract_js_vars(
-                webpage, r'(var\s+(?:%s)_.+)' % '|'.join(FORMAT_PREFIXES),
-                default=None)
-            if js_vars:
-                for key, format_url in js_vars.items():
-                    if key.startswith(FORMAT_PREFIXES[-1]):
-                        parse_quality_items(format_url)
-                    elif any(key.startswith(p) for p in FORMAT_PREFIXES[:2]):
-                        add_video_url(format_url)
-            if not video_urls and re.search(
-                    r'<[^>]+\bid=["\']lockedPlayer', webpage):
+                webpage, r'(var\s+(?:%s)_.+)' % '|'.join(FORMAT_PREFIXES))
+            for key, format_url in js_vars.items():
+                if key.startswith(FORMAT_PREFIXES[-1]):
+                    parse_quality_items(format_url)
+                elif any(key.startswith(p) for p in FORMAT_PREFIXES[:2]):
+                    add_video_url(format_url)
+            if not video_urls and get_element_by_id('lockedPlayer', webpage):
                 raise ExtractorError(
                     'Video %s is locked' % video_id, expected=True)
 
         if not video_urls:
             js_vars = extract_js_vars(
                 dl_webpage('tv'), r'(var.+?mediastring.+?)</script>')
-            add_video_url(js_vars['mediastring'])
+            add_video_url(traverse_obj(js_vars, 'mediastring'))
 
         for mobj in re.finditer(
                 r'<a[^>]+\bclass=["\']downloadBtn\b[^>]+\bhref=(["\'])(?P<url>(?:(?!\1).)+)\1',
                 webpage):
-            video_url = mobj.group('url')
-            if video_url not in video_urls_set:
-                video_urls.append((video_url, None))
-                video_urls_set.add(video_url)
+            add_video_url(mobj.group('url'))
 
         upload_date = None
         formats = []
@@ -433,6 +439,13 @@ class PornHubIE(PornHubBaseIE):
                 'height': height,
             })
 
+        if not video_urls:
+            # import here to avoid mutually recursive dependency
+            from .generic import GenericIE
+            ret = GenericIE.generic_url_result(url, video_id=video_id, video_title=title, force_videoid=True)
+            ret['_type'] = 'url_transparent'
+            return ret
+
         for video_url, height in video_urls:
             if not upload_date:
                 upload_date = self._search_regex(
@@ -440,52 +453,55 @@ class PornHubIE(PornHubBaseIE):
                 if upload_date:
                     upload_date = upload_date.replace('/', '')
             if '/video/get_media' in video_url:
+                # self._set_cookie(host, 'platform', 'tv')
                 medias = self._download_json(video_url, video_id, fatal=False)
-                if isinstance(medias, list):
-                    for media in medias:
-                        if not isinstance(media, dict):
-                            continue
-                        video_url = url_or_none(media.get('videoUrl'))
-                        if not video_url:
-                            continue
-                        height = int_or_none(media.get('quality'))
-                        add_format(video_url, height)
+                for media in traverse_obj(medias, lambda _, v: v['videoUrl']):
+                    video_url = url_or_none(media['videoUrl'])
+                    if not video_url:
+                        continue
+                    height = int_or_none(media.get('quality'))
+                    add_format(video_url, height)
                 continue
             add_format(video_url)
+
         self._sort_formats(
             formats, field_preference=('height', 'width', 'fps', 'format_id'))
 
+        model_profile = self._search_json(
+            r'var\s+MODEL_PROFILE\s*=', webpage, 'model profile', video_id, fatal=False)
         video_uploader = self._html_search_regex(
             r'(?s)From:&nbsp;.+?<(?:a\b[^>]+\bhref=["\']/(?:(?:user|channel)s|model|pornstar)/|span\b[^>]+\bclass=["\']username)[^>]+>(.+?)<',
-            webpage, 'uploader', default=None)
+            webpage, 'uploader', default=None) or model_profile.get('username')
 
         def extract_vote_count(kind, name):
             return self._extract_count(
-                (r'<span[^>]+\bclass="votes%s"[^>]*>([\d,\.]+)</span>' % kind,
+                (r'<span[^>]+\bclass="votes%s"[^>]*>(\d[\d,\.]*[kKmM]?)</span>' % kind,
                  r'<span[^>]+\bclass=["\']votes%s["\'][^>]*\bdata-rating=["\'](\d+)' % kind),
                 webpage, name)
 
         view_count = self._extract_count(
-            r'<span class="count">([\d,\.]+)</span> [Vv]iews', webpage, 'view')
+            r'<span class="count">(\d[\d,\.]*[kKmM]?)</span> [Vv]iews', webpage, 'view')
         like_count = extract_vote_count('Up', 'like')
         dislike_count = extract_vote_count('Down', 'dislike')
         comment_count = self._extract_count(
-            r'All Comments\s*<span>\(([\d,.]+)\)', webpage, 'comment')
+            r'All Comments\s*<span>\((\d[\d,\.]*[kKmM]?)\)', webpage, 'comment')
 
         def extract_list(meta_key):
             div = self._search_regex(
                 r'(?s)<div[^>]+\bclass=["\'].*?\b%sWrapper[^>]*>(.+?)</div>'
                 % meta_key, webpage, meta_key, default=None)
             if div:
-                return re.findall(r'<a[^>]+\bhref=[^>]+>([^<]+)', div)
+                return [clean_html(x) for x in re.findall(r'(?s)<a[^>]+\bhref=[^>]+>.+?</a>', div)]
 
         info = self._search_json_ld(webpage, video_id, default={})
         # description provided in JSON-LD is irrelevant
-        info['description'] = None
+        for k in ('url', 'description'):
+            info.pop(k, None)
 
-        return merge_dicts({
+        return merge_dicts(info, {
             'id': video_id,
             'uploader': video_uploader,
+            'uploader_id': remove_start(model_profile.get('modelProfileLink'), '/model/'),
             'upload_date': upload_date,
             'title': title,
             'thumbnail': thumbnail,
@@ -498,8 +514,9 @@ class PornHubIE(PornHubBaseIE):
             'age_limit': 18,
             'tags': extract_list('tags'),
             'categories': extract_list('categories'),
+            'cast': extract_list('pornstars'),
             'subtitles': subtitles,
-        }, info)
+        })
 
 
 class PornHubPlaylistBaseIE(PornHubBaseIE):
@@ -512,65 +529,28 @@ class PornHubPlaylistBaseIE(PornHubBaseIE):
         # drop-down menu that uses similar pattern for videos (see
         # https://github.com/ytdl-org/youtube-dl/issues/11594).
         container = self._search_regex(
-            r'(?s)(<div[^>]+class=["\']container.+)', webpage,
+            r'(?s)(<div\s[^>]*class=["\']container.+)', webpage,
             'container', default=webpage)
 
+        def entries():
+            seen_ids = set()
+            for m in re.finditer(r'<\w+\s[^>]*(?<!-)\bhref\s*=\s*.("|\'|\b)%s\1[^>]*>' % (PornHubIE._PORNHUB_PATH_RE,), container):
+                video_id = m.group('id')
+                if video_id:
+                    if video_id in seen_ids:
+                        continue
+                    seen_ids.add(video_id)
+                elt = extract_attributes(m.group(0))
+                video_url = urljoin(host, elt.get('href'))
+                yield video_url, video_id, elt.get('title')
+
         return [
             self.url_result(
-                'http://www.%s/%s' % (host, video_url),
-                PornHubIE.ie_key(), video_title=title)
-            for video_url, title in orderedSet(re.findall(
-                r'href="/?(view_video\.php\?.*\bviewkey=[\da-z]+[^"]*)"[^>]*\s+title="([^"]+)"',
-                container))
+                video_url, PornHubIE.ie_key(), video_title=title, video_id=video_id)
+            for video_url, video_id, title in entries()
         ]
 
 
-class PornHubUserIE(PornHubPlaylistBaseIE):
-    _VALID_URL = r'(?P<url>https?://(?:[^/]+\.)?%s/(?:(?:user|channel)s|model|pornstar)/(?P<id>[^/?#&]+))(?:[?#&]|/(?!videos)|$)' % PornHubBaseIE._PORNHUB_HOST_RE
-    _TESTS = [{
-        'url': 'https://www.pornhub.com/model/zoe_ph',
-        'playlist_mincount': 118,
-    }, {
-        'url': 'https://www.pornhub.com/pornstar/liz-vicious',
-        'info_dict': {
-            'id': 'liz-vicious',
-        },
-        'playlist_mincount': 118,
-    }, {
-        'url': 'https://www.pornhub.com/users/russianveet69',
-        'only_matching': True,
-    }, {
-        'url': 'https://www.pornhub.com/channels/povd',
-        'only_matching': True,
-    }, {
-        'url': 'https://www.pornhub.com/model/zoe_ph?abc=1',
-        'only_matching': True,
-    }, {
-        # Unavailable via /videos page, but available with direct pagination
-        # on pornstar page (see [1]), requires premium
-        # 1. https://github.com/ytdl-org/youtube-dl/issues/27853
-        'url': 'https://www.pornhubpremium.com/pornstar/sienna-west',
-        'only_matching': True,
-    }, {
-        # Same as before, multi page
-        'url': 'https://www.pornhubpremium.com/pornstar/lily-labeau',
-        'only_matching': True,
-    }, {
-        'url': 'https://pornhubthbh7ap3u.onion/model/zoe_ph',
-        'only_matching': True,
-    }]
-
-    def _real_extract(self, url):
-        mobj = re.match(self._VALID_URL, url)
-        user_id = mobj.group('id')
-        videos_url = '%s/videos' % mobj.group('url')
-        page = self._extract_page(url)
-        if page:
-            videos_url = update_url_query(videos_url, {'page': page})
-        return self.url_result(
-            videos_url, ie=PornHubPagedVideoListIE.ie_key(), video_id=user_id)
-
-
 class PornHubPagedPlaylistBaseIE(PornHubPlaylistBaseIE):
     @staticmethod
     def _has_more(webpage):
@@ -617,23 +597,77 @@ class PornHubPagedPlaylistBaseIE(PornHubPlaylistBaseIE):
             page_entries = self._extract_entries(webpage, host)
             if not page_entries:
                 break
-            for e in page_entries:
-                yield e
+            for from_ in page_entries:
+                yield from_
             if not self._has_more(webpage):
                 break
 
     def _real_extract(self, url):
-        mobj = re.match(self._VALID_URL, url)
+        mobj = self._match_valid_url(url)
         host = mobj.group('host')
         item_id = mobj.group('id')
 
         self._login(host)
+        self._set_age_cookies(host)
 
         return self.playlist_result(self._entries(url, host, item_id), item_id)
 
 
+class PornHubUserIE(PornHubPagedPlaylistBaseIE):
+    _VALID_URL = r'(?P<url>https?://(?:[^/]+\.)?%s/(?P<id>(?:(?:user|channel)s|model|pornstar)/[^/?#&]+))(?:[?#&]|/(?!videos)|$)' % PornHubBaseIE._PORNHUB_HOST_RE
+    _TESTS = [{
+        'url': 'https://www.pornhub.com/model/zoe_ph',
+        'info_dict': {
+            'id': 'zoe_ph',
+        },
+        'playlist_mincount': 118,
+    }, {
+        'url': 'https://www.pornhub.com/pornstar/liz-vicious',
+        'info_dict': {
+            'id': 'liz-vicious',
+        },
+        'playlist_mincount': 118,
+    }, {
+        'url': 'https://www.pornhub.com/users/russianveet69',
+        'only_matching': True,
+    }, {
+        'url': 'https://www.pornhub.com/channels/povd',
+        'only_matching': True,
+    }, {
+        'url': 'https://www.pornhub.com/model/zoe_ph?abc=1',
+        'only_matching': True,
+    }, {
+        # Unavailable via /videos page, but available with direct pagination
+        # on pornstar page (see [1]), requires premium
+        # 1. https://github.com/ytdl-org/youtube-dl/issues/27853
+        'url': 'https://www.pornhubpremium.com/pornstar/sienna-west',
+        'only_matching': True,
+    }, {
+        # Same as before, multi page
+        'url': 'https://www.pornhubpremium.com/pornstar/lily-labeau',
+        'only_matching': True,
+    }, {
+        'url': 'https://pornhubvybmsymdol4iibwgwtkpwmeyd6luq2gxajgjzfjvotyt5zhyd.onion/model/zoe_ph',
+        'only_matching': True,
+    }]
+
+    def _real_extract(self, url):
+        mobj = self._match_valid_url(url)
+        user_id, host = mobj.group('id', 'host')
+        videos_url = '%s/videos' % mobj.group('url')
+        page = self._extract_page(url)
+        if page:
+            videos_url = update_url_query(videos_url, {'page': page})
+
+        self._login(host)
+
+        return self.playlist_result(self._entries(videos_url, host, user_id), user_id.split('/')[-1])
+        # return self.url_result(
+        #     videos_url, ie=PornHubPagedVideoListIE.ie_key(), video_id=user_id)
+
+
 class PornHubPagedVideoListIE(PornHubPagedPlaylistBaseIE):
-    _VALID_URL = r'https?://(?:[^/]+\.)?%s/(?P<id>(?:[^/]+/)*[^/?#&]+)' % PornHubBaseIE._PORNHUB_HOST_RE
+    _VALID_URL = r'https?://(?:[^/]+\.)?%s/(?!playlist/|gif/)(?P<id>(?:[^/]+/)*[^/?#&]+)' % PornHubBaseIE._PORNHUB_HOST_RE
     _TESTS = [{
         'url': 'https://www.pornhub.com/model/zoe_ph/videos',
         'only_matching': True,
@@ -642,16 +676,20 @@ class PornHubPagedVideoListIE(PornHubPagedPlaylistBaseIE):
         'only_matching': True,
     }, {
         'url': 'https://www.pornhub.com/pornstar/jenny-blighe/videos',
-        'info_dict': {
-            'id': 'pornstar/jenny-blighe/videos',
-        },
-        'playlist_mincount': 149,
+        'only_matching': True,
     }, {
-        'url': 'https://www.pornhub.com/pornstar/jenny-blighe/videos?page=3',
+        'url': 'https://www.pornhub.com/pornstar/kylie-quinn/videos',
         'info_dict': {
-            'id': 'pornstar/jenny-blighe/videos',
+            'id': 'pornstar/kylie-quinn/videos',
         },
-        'playlist_mincount': 40,
+        'playlist_mincount': 80,
+    }, {
+        'url': 'https://www.pornhub.com/pornstar/kylie-quinn/videos?page=2',
+        'info_dict': {
+            'id': 'pornstar/kylie-quinn/videos',
+        },
+        # specific page: process just that page
+        'playlist_count': 40,
     }, {
         # default sorting as Top Rated Videos
         'url': 'https://www.pornhub.com/channels/povd/videos',
@@ -727,27 +765,14 @@ class PornHubPagedVideoListIE(PornHubPagedPlaylistBaseIE):
         'url': 'https://www.pornhub.com/video/incategories/60fps-1/hd-porn',
         'only_matching': True,
     }, {
-        'url': 'https://www.pornhub.com/playlist/44121572',
-        'info_dict': {
-            'id': 'playlist/44121572',
-        },
-        'playlist_mincount': 132,
-    }, {
-        'url': 'https://www.pornhub.com/playlist/4667351',
-        'only_matching': True,
-    }, {
-        'url': 'https://de.pornhub.com/playlist/4667351',
-        'only_matching': True,
-    }, {
-        'url': 'https://pornhubthbh7ap3u.onion/model/zoe_ph/videos',
+        'url': 'https://pornhubvybmsymdol4iibwgwtkpwmeyd6luq2gxajgjzfjvotyt5zhyd.onion/model/zoe_ph/videos',
         'only_matching': True,
     }]
 
     @classmethod
     def suitable(cls, url):
-        return (False
-                if PornHubIE.suitable(url) or PornHubUserIE.suitable(url) or PornHubUserVideosUploadIE.suitable(url)
-                else super(PornHubPagedVideoListIE, cls).suitable(url))
+        return (not any(ph.suitable(url) for ph in (PornHubIE, PornHubUserIE, PornHubUserVideosUploadIE))
+                and super(PornHubPagedVideoListIE, cls).suitable(url))
 
 
 class PornHubUserVideosUploadIE(PornHubPagedPlaylistBaseIE):
@@ -762,6 +787,62 @@ class PornHubUserVideosUploadIE(PornHubPagedPlaylistBaseIE):
         'url': 'https://www.pornhub.com/model/zoe_ph/videos/upload',
         'only_matching': True,
     }, {
-        'url': 'http://pornhubthbh7ap3u.onion/pornstar/jenny-blighe/videos/upload',
+        'url': 'http://pornhubvybmsymdol4iibwgwtkpwmeyd6luq2gxajgjzfjvotyt5zhyd.onion/pornstar/jenny-blighe/videos/upload',
         'only_matching': True,
     }]
+
+
+class PornHubPlaylistIE(PornHubPlaylistBaseIE):
+    _VALID_URL = r'(?P<url>https?://(?:[^/]+\.)?%s/playlist/(?P<id>[^/?#&]+))' % PornHubBaseIE._PORNHUB_HOST_RE
+    _TESTS = [{
+        'url': 'https://www.pornhub.com/playlist/44121572',
+        'info_dict': {
+            'id': '44121572',
+        },
+        'playlist_mincount': 55,
+    }, {
+        'url': 'https://www.pornhub.com/playlist/4667351',
+        'only_matching': True,
+    }, {
+        'url': 'https://de.pornhub.com/playlist/4667351',
+        'only_matching': True,
+    }, {
+        'url': 'https://de.pornhub.com/playlist/4667351?page=2',
+        'only_matching': True,
+    }]
+
+    def _entries(self, url, host, item_id):
+        webpage = self._download_webpage(url, item_id, 'Downloading page 1')
+        playlist_id = self._search_regex(r'var\s+playlistId\s*=\s*"([^"]+)"', webpage, 'playlist_id')
+        video_count = int_or_none(
+            self._search_regex(r'var\s+itemsCount\s*=\s*([0-9]+)\s*\|\|', webpage, 'video_count'))
+        token = self._search_regex(r'var\s+token\s*=\s*"([^"]+)"', webpage, 'token')
+        page_count = (video_count - 36 + 39) // 40 + 1
+        page_entries = self._extract_entries(webpage, host)
+
+        def download_page(page_num):
+            note = 'Downloading page {0}'.format(page_num)
+            page_url = 'https://www.{0}/playlist/viewChunked'.format(host)
+            return self._download_webpage(page_url, item_id, note, query={
+                'id': playlist_id,
+                'page': page_num,
+                'token': token,
+            })
+
+        for page_num in range(1, page_count + 1):
+            if page_num > 1:
+                webpage = download_page(page_num)
+                page_entries = self._extract_entries(webpage, host)
+            if not page_entries:
+                break
+            for from_ in page_entries:
+                yield from_
+
+    def _real_extract(self, url):
+        mobj = self._match_valid_url(url)
+        host, item_id = mobj.group('host', 'id')
+
+        self._login(host)
+        self._set_age_cookies(host)
+
+        return self.playlist_result(self._entries(mobj.group('url'), host, item_id), item_id)

youtube_dl/extractor/youtube.py

@@ -122,7 +122,8 @@ class YoutubeBaseInfoExtractor(InfoExtractor):
             'INNERTUBE_CONTEXT': {
                 'client': {
                     'clientName': 'TVHTML5',
-                    'clientVersion': '7.20241201.18.00',
+                    'clientVersion': '7.20250120.19.00',
+                    'userAgent': 'Mozilla/5.0 (ChromiumStylePlatform) Cobalt/Version',
                 },
             },
             'INNERTUBE_CONTEXT_CLIENT_NAME': 7,
@@ -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\("\{"\))(?=\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')