Merge 8651fb67757da79d1f9bd3401b67a1e61506fa7c 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)
+    ie = YoutubeIE(FakeYDL({'cachedir': False}))
-    return JSInterpreter(jscode).call_function(
+    jsi = JSInterpreter(jscode)
-        funcname, sig_input, _ytdl_do_not_return=sig_input)
+    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
+# compat_xml_etree_register_namespace
-else:
+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+$", prefix):
+        if re.match(r'ns\d+$', prefix):
-            raise ValueError("Prefix format reserved for internal use")
+            raise ValueError('Prefix format reserved for internal use')
-        for k, v in list(etree._namespace_map.items()):
+        for k, v in list(_etree._namespace_map.items()):
             if k == uri or v == prefix:
-                del etree._namespace_map[k]
+                del _etree._namespace_map[k]
-        etree._namespace_map[uri] = prefix
+        _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):
+    (lambda x: x)(**{'x': 0})
-        pass
-    _testfunc(**{'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_shutil_get_terminal_size
-    compat_get_terminal_size = shutil.get_terminal_size
+try:
-else:
+    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
+# builtins returning an iterator
-# same versioning as for zip below
+
+# 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

@@ -1163,6 +1163,7 @@ from .soundcloud import (
     SoundcloudUserIE,
     SoundcloudTrackStationIE,
     SoundcloudPlaylistIE,
+    SoundcloudRelatedIE,
     SoundcloudSearchIE,
 )
 from .soundgasm import (

youtube_dl/extractor/soundcloud.py

@@ -3,6 +3,7 @@ from __future__ import unicode_literals
 
 import itertools
 import re
+import json
 
 from .common import (
     InfoExtractor,
@@ -22,12 +23,14 @@ from ..utils import (
     int_or_none,
     KNOWN_EXTENSIONS,
     mimetype2ext,
+    remove_end,
     str_or_none,
     try_get,
     unified_timestamp,
     update_url_query,
     url_or_none,
     urlhandle_detect_ext,
+    sanitized_Request,
 )
 
 
@@ -55,7 +58,152 @@ class SoundcloudEmbedIE(InfoExtractor):
         return self.url_result(api_url)
 
 
-class SoundcloudIE(InfoExtractor):
+class SoundcloudBaseIE(InfoExtractor):
+    _NETRC_MACHINE = 'soundcloud'
+
+    _API_V2_BASE = 'https://api-v2.soundcloud.com/'
+    _BASE_URL = 'https://soundcloud.com/'
+    _USER_AGENT = 'Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/84.0.4147.105 Safari/537.36'
+    _API_AUTH_QUERY_TEMPLATE = '?client_id=%s'
+    _API_AUTH_URL_PW = 'https://api-auth.soundcloud.com/web-auth/sign-in/password%s'
+    _API_VERIFY_AUTH_TOKEN = 'https://api-auth.soundcloud.com/connect/session%s'
+    _access_token = None
+    _HEADERS = {}
+
+    def _store_client_id(self, client_id):
+        self._downloader.cache.store('soundcloud', 'client_id', client_id)
+
+    def _update_client_id(self):
+        webpage = self._download_webpage('https://soundcloud.com/', None)
+        for src in reversed(re.findall(r'<script[^>]+src="([^"]+)"', webpage)):
+            script = self._download_webpage(src, None, fatal=False)
+            if script:
+                client_id = self._search_regex(
+                    r'client_id\s*:\s*"([0-9a-zA-Z]{32})"',
+                    script, 'client id', default=None)
+                if client_id:
+                    self._CLIENT_ID = client_id
+                    self._store_client_id(client_id)
+                    return
+        raise ExtractorError('Unable to extract client id')
+
+    def _download_json(self, *args, **kwargs):
+        non_fatal = kwargs.get('fatal') is False
+        if non_fatal:
+            del kwargs['fatal']
+        query = kwargs.get('query', {}).copy()
+        for _ in range(2):
+            query['client_id'] = self._CLIENT_ID
+            kwargs['query'] = query
+            try:
+                return super(SoundcloudBaseIE, self)._download_json(*args, **compat_kwargs(kwargs))
+            except ExtractorError as e:
+                if isinstance(e.cause, compat_HTTPError) and e.cause.code in (401, 403):
+                    self._store_client_id(None)
+                    self._update_client_id()
+                    continue
+                elif non_fatal:
+                    self.report_warning(error_to_compat_str(e))
+                    return False
+                raise
+
+    def _initialize_pre_login(self):
+        self._CLIENT_ID = self._downloader.cache.load('soundcloud', 'client_id') or 'a3e059563d7fd3372b49b37f00a00bcf'
+
+    def _perform_login(self, username, password):
+        if username != 'oauth':
+            self.report_warning(
+                'Login using username and password is not currently supported. '
+                'Use "--username oauth --password <oauth_token>" to login using an oauth token')
+        self._access_token = password
+        query = self._API_AUTH_QUERY_TEMPLATE % self._CLIENT_ID
+        payload = {'session': {'access_token': self._access_token}}
+        token_verification = sanitized_Request(self._API_VERIFY_AUTH_TOKEN % query, json.dumps(payload).encode('utf-8'))
+        response = self._download_json(token_verification, None, note='Verifying login token...', fatal=False)
+        if response is not False:
+            self._HEADERS = {'Authorization': 'OAuth ' + self._access_token}
+            self.report_login()
+        else:
+            self.report_warning('Provided authorization token seems to be invalid. Continue as guest')
+
+        r'''
+        def genDevId():
+            def genNumBlock():
+                return ''.join([str(random.randrange(10)) for i in range(6)])
+            return '-'.join([genNumBlock() for i in range(4)])
+
+        payload = {
+            'client_id': self._CLIENT_ID,
+            'recaptcha_pubkey': 'null',
+            'recaptcha_response': 'null',
+            'credentials': {
+                'identifier': username,
+                'password': password
+            },
+            'signature': self.sign(username, password, self._CLIENT_ID),
+            'device_id': genDevId(),
+            'user_agent': self._USER_AGENT
+        }
+
+        query = self._API_AUTH_QUERY_TEMPLATE % self._CLIENT_ID
+        login = sanitized_Request(self._API_AUTH_URL_PW % query, json.dumps(payload).encode('utf-8'))
+        response = self._download_json(login, None)
+        self._access_token = response.get('session').get('access_token')
+        if not self._access_token:
+            self.report_warning('Unable to get access token, login may has failed')
+        else:
+            self._HEADERS = {'Authorization': 'OAuth ' + self._access_token}
+        '''
+
+    def _real_initialize(self):
+        self._initialize_pre_login()
+        username, password = self._get_login_info()
+        if username:
+            self._perform_login(username, password)
+
+        super(SoundcloudBaseIE, self)._real_initialize()
+
+    # signature generation
+    def sign(self, user, pw, clid):
+        a = 33
+        i = 1
+        s = 440123
+        w = 117
+        u = 1800000
+        l = 1042
+        b = 37
+        k = 37
+        c = 5
+        n = '0763ed7314c69015fd4a0dc16bbf4b90'  # _KEY
+        y = '8'  # _REV
+        r = self._USER_AGENT
+        e = user  # _USERNAME
+        t = clid  # _CLIENT_ID
+
+        d = '-'.join([compat_str(mInt) for mInt in [a, i, s, w, u, l, b, k]])
+        p = n + y + d + r + e + t + d + n
+        h = p
+
+        m = 8011470
+        f = 0
+
+        for f in range(f, len(h)):
+            m = (m >> 1) + ((1 & m) << 23)
+            m += ord(h[f])
+            m &= 16777215
+
+        # c is not even needed
+        # out = str(y) + ':' + str(d) + ':' + format(m, 'x') + ':' + str(c)
+        out = '%s:%s:%x:%d' % (y, d, m, c)
+
+        return out
+
+    @classmethod
+    def _resolv_url(cls, url):
+        return cls._API_V2_BASE + 'resolve?url=' + url
+
+
+class SoundcloudIE(SoundcloudBaseIE):
     """Information extractor for soundcloud.com
        To access the media, the uid of the song and a stream token
        must be extracted from the page source and the script must make
@@ -69,8 +217,9 @@ class SoundcloudIE(InfoExtractor):
                             (?!stations/track)
                             (?P<uploader>[\w\d-]+)/
                             (?!(?:tracks|albums|sets(?:/.+?)?|reposts|likes|spotlight)/?(?:$|[?#]))
-                            (?P<title>[\w\d-]+)/?
+                            (?P<title>[\w\d-]+)
-                            (?P<token>[^?]+?)?(?:[?].*)?$)
+                            (?:/(?P<token>(?!(?:albums|sets|recommended))[^?]+?))?
+                            (?:[?].*)?$)
                        |(?:api(?:-v2)?\.soundcloud\.com/tracks/(?P<track_id>\d+)
                           (?:/?\?secret_token=(?P<secret_token>[^&]+))?)
                     )
@@ -161,23 +310,16 @@ class SoundcloudIE(InfoExtractor):
         },
         # downloadable song
         {
-            'url': 'https://soundcloud.com/oddsamples/bus-brakes',
+            'url': 'https://soundcloud.com/the80m/the-following',
-            'md5': '7624f2351f8a3b2e7cd51522496e7631',
+            'md5': '9ffcddb08c87d74fb5808a3c183a1d04',
             'info_dict': {
-                'id': '128590877',
+                'id': '343609555',
-                'ext': 'mp3',
+                'ext': 'wav',
-                'title': 'Bus Brakes',
+                'title': 'The Following',
-                'description': 'md5:0053ca6396e8d2fd7b7e1595ef12ab66',
+                'timestamp': 1506120436,
-                'uploader': 'oddsamples',
+                'upload_date': '20170922',
-                'uploader_id': '73680509',
+                'uploader_id': '312384765',
-                'timestamp': 1389232924,
+                'uploader': '80M',
-                'upload_date': '20140109',
-                'duration': 17.346,
-                'license': 'cc-by-sa',
-                'view_count': int,
-                'like_count': int,
-                'comment_count': int,
-                'repost_count': int,
             },
         },
         # private link, downloadable format
@@ -233,8 +375,8 @@ class SoundcloudIE(InfoExtractor):
                 'id': '583011102',
                 'ext': 'mp3',
                 'title': 'Mezzo Valzer',
-                'description': 'md5:4138d582f81866a530317bae316e8b61',
+                'description': 'md5:8de664f12895716c8ac6f1da6348eb8e',
-                'uploader': 'Micronie',
+                'uploader': 'Giovanni Sarani',
                 'uploader_id': '3352531',
                 'timestamp': 1551394171,
                 'upload_date': '20190228',
@@ -248,14 +390,17 @@ class SoundcloudIE(InfoExtractor):
             },
         },
         {
-            # with AAC HQ format available via OAuth token
+            # with AAC HQ format available via OAuth token (account with active subscription needed)
             'url': 'https://soundcloud.com/wandw/the-chainsmokers-ft-daya-dont-let-me-down-ww-remix-1',
             'only_matching': True,
         },
+        {
+            # Go+ (account with active subscription needed)
+            'url': 'https://soundcloud.com/taylorswiftofficial/look-what-you-made-me-do',
+            'only_matching': True,
+        },
     ]
 
-    _API_V2_BASE = 'https://api-v2.soundcloud.com/'
-    _BASE_URL = 'https://soundcloud.com/'
     _IMAGE_REPL_RE = r'-([0-9a-z]+)\.jpg'
 
     _ARTWORK_MAP = {
@@ -271,50 +416,6 @@ class SoundcloudIE(InfoExtractor):
         'original': 0,
     }
 
-    def _store_client_id(self, client_id):
-        self._downloader.cache.store('soundcloud', 'client_id', client_id)
-
-    def _update_client_id(self):
-        webpage = self._download_webpage('https://soundcloud.com/', None)
-        for src in reversed(re.findall(r'<script[^>]+src="([^"]+)"', webpage)):
-            script = self._download_webpage(src, None, fatal=False)
-            if script:
-                client_id = self._search_regex(
-                    r'client_id\s*:\s*"([0-9a-zA-Z]{32})"',
-                    script, 'client id', default=None)
-                if client_id:
-                    self._CLIENT_ID = client_id
-                    self._store_client_id(client_id)
-                    return
-        raise ExtractorError('Unable to extract client id')
-
-    def _download_json(self, *args, **kwargs):
-        non_fatal = kwargs.get('fatal') is False
-        if non_fatal:
-            del kwargs['fatal']
-        query = kwargs.get('query', {}).copy()
-        for _ in range(2):
-            query['client_id'] = self._CLIENT_ID
-            kwargs['query'] = query
-            try:
-                return super(SoundcloudIE, self)._download_json(*args, **compat_kwargs(kwargs))
-            except ExtractorError as e:
-                if isinstance(e.cause, compat_HTTPError) and e.cause.code == 401:
-                    self._store_client_id(None)
-                    self._update_client_id()
-                    continue
-                elif non_fatal:
-                    self._downloader.report_warning(error_to_compat_str(e))
-                    return False
-                raise
-
-    def _real_initialize(self):
-        self._CLIENT_ID = self._downloader.cache.load('soundcloud', 'client_id') or 'YUKXoArFcqrlQn9tfNHvvyfnDISj04zk'
-
-    @classmethod
-    def _resolv_url(cls, url):
-        return SoundcloudIE._API_V2_BASE + 'resolve?url=' + url
-
     def _extract_info_dict(self, info, full_title=None, secret_token=None):
         track_id = compat_str(info['id'])
         title = info['title']
@@ -389,7 +490,7 @@ class SoundcloudIE(InfoExtractor):
             if not format_url:
                 continue
             stream = self._download_json(
-                format_url, track_id, query=query, fatal=False)
+                format_url, track_id, query=query, fatal=False, headers=self._HEADERS)
             if not isinstance(stream, dict):
                 continue
             stream_url = url_or_none(stream.get('url'))
@@ -423,8 +524,8 @@ class SoundcloudIE(InfoExtractor):
 
         thumbnails = []
         artwork_url = info.get('artwork_url')
-        thumbnail = artwork_url or user.get('avatar_url')
+        thumbnail = url_or_none(artwork_url or user.get('avatar_url'))
-        if isinstance(thumbnail, compat_str):
+        if thumbnail:
             if re.search(self._IMAGE_REPL_RE, thumbnail):
                 for image_id, size in self._ARTWORK_MAP.items():
                     i = {
@@ -487,12 +588,12 @@ class SoundcloudIE(InfoExtractor):
             info_json_url = self._resolv_url(self._BASE_URL + resolve_title)
 
         info = self._download_json(
-            info_json_url, full_title, 'Downloading info JSON', query=query)
+            info_json_url, full_title, 'Downloading info JSON', query=query, headers=self._HEADERS)
 
         return self._extract_info_dict(info, full_title, token)
 
 
-class SoundcloudPlaylistBaseIE(SoundcloudIE):
+class SoundcloudPlaylistBaseIE(SoundcloudBaseIE):
     def _extract_set(self, playlist, token=None):
         playlist_id = compat_str(playlist['id'])
         tracks = playlist.get('tracks') or []
@@ -503,7 +604,7 @@ class SoundcloudPlaylistBaseIE(SoundcloudIE):
                     'ids': ','.join([compat_str(t['id']) for t in tracks]),
                     'playlistId': playlist_id,
                     'playlistSecretToken': token,
-                })
+                }, headers=self._HEADERS)
         entries = []
         for track in tracks:
             track_id = str_or_none(track.get('id'))
@@ -523,7 +624,7 @@ class SoundcloudPlaylistBaseIE(SoundcloudIE):
 
 
 class SoundcloudSetIE(SoundcloudPlaylistBaseIE):
-    _VALID_URL = r'https?://(?:(?:www|m)\.)?soundcloud\.com/(?P<uploader>[\w\d-]+)/sets/(?P<slug_title>[\w\d-]+)(?:/(?P<token>[^?/]+))?'
+    _VALID_URL = r'https?://(?:(?:www|m)\.)?soundcloud\.com/(?P<uploader>[\w\d-]+)/sets/(?P<slug_title>[:\w\d-]+)(?:/(?P<token>[^?/]+))?'
     IE_NAME = 'soundcloud:set'
     _TESTS = [{
         'url': 'https://soundcloud.com/the-concept-band/sets/the-royal-concept-ep',
@@ -536,6 +637,15 @@ class SoundcloudSetIE(SoundcloudPlaylistBaseIE):
     }, {
         'url': 'https://soundcloud.com/the-concept-band/sets/the-royal-concept-ep/token',
         'only_matching': True,
+    }, {
+        'url': 'https://soundcloud.com/discover/sets/weekly::flacmatic',
+        'only_matching': True,
+    }, {
+        'url': 'https://soundcloud.com/discover/sets/charts-top:all-music:de',
+        'only_matching': True,
+    }, {
+        'url': 'https://soundcloud.com/discover/sets/charts-top:hiphoprap:kr',
+        'only_matching': True,
     }]
 
     def _real_extract(self, url):
@@ -547,7 +657,7 @@ class SoundcloudSetIE(SoundcloudPlaylistBaseIE):
             full_title += '/' + token
 
         info = self._download_json(self._resolv_url(
-            self._BASE_URL + full_title), full_title)
+            self._BASE_URL + full_title), full_title, headers=self._HEADERS)
 
         if 'errors' in info:
             msgs = (compat_str(err['error_message']) for err in info['errors'])
@@ -556,66 +666,65 @@ class SoundcloudSetIE(SoundcloudPlaylistBaseIE):
         return self._extract_set(info, token)
 
 
-class SoundcloudPagedPlaylistBaseIE(SoundcloudIE):
+class SoundcloudPagedPlaylistBaseIE(SoundcloudBaseIE):
     def _extract_playlist(self, base_url, playlist_id, playlist_title):
+        return {
+            '_type': 'playlist',
+            'id': playlist_id,
+            'title': playlist_title,
+            'entries': self._entries(base_url, playlist_id),
+        }
+
+    def _entries(self, url, playlist_id):
         # Per the SoundCloud documentation, the maximum limit for a linked partitioning query is 200.
         # https://developers.soundcloud.com/blog/offset-pagination-deprecated
         COMMON_QUERY = {
             'limit': 200,
             'linked_partitioning': '1',
+            'offset': 0,
         }
 
         query = COMMON_QUERY.copy()
-        query['offset'] = 0
+        retries = 3
 
-        next_href = base_url
+        def resolve_entry(*candidates):
-
+            for cand in candidates:
-        entries = []
+                if not isinstance(cand, dict):
-        for i in itertools.count():
+                    continue
-            response = self._download_json(
+                permalink_url = url_or_none(cand.get('permalink_url'))
-                next_href, playlist_id,
+                if permalink_url:
-                'Downloading track page %s' % (i + 1), query=query)
-
-            collection = response['collection']
-
-            if not isinstance(collection, list):
-                collection = []
-
-            # Empty collection may be returned, in this case we proceed
-            # straight to next_href
-
-            def resolve_entry(candidates):
-                for cand in candidates:
-                    if not isinstance(cand, dict):
-                        continue
-                    permalink_url = url_or_none(cand.get('permalink_url'))
-                    if not permalink_url:
-                        continue
                     return self.url_result(
                         permalink_url,
                         SoundcloudIE.ie_key() if SoundcloudIE.suitable(permalink_url) else None,
                         str_or_none(cand.get('id')), cand.get('title'))
 
-            for e in collection:
+        for i in itertools.count():
-                entry = resolve_entry((e, e.get('track'), e.get('playlist')))
+            last_error = None
-                if entry:
+            for attempt in range(0, retries + 1):
-                    entries.append(entry)
+                if last_error:
+                    self._downloader.report_warning('%s. Retrying ...' % remove_end(last_error, '.'), playlist_id)
+                try:
+                    response = self._download_json(
+                        url, playlist_id, query=query, headers=self._HEADERS,
+                        note='Downloading track page %s%s' % (i + 1, ' (retry #%d)' % (attempt, ) if attempt > 0 else ''))
+                    if not isinstance(response, dict):
+                        response = {}
+                    break
+                except ExtractorError as e:
+                    # Downloading page may result in intermittent 502 HTTP error
+                    # See https://github.com/yt-dlp/yt-dlp/issues/872
+                    if attempt >= retries or not isinstance(e.cause, compat_HTTPError) or e.cause.code != 502:
+                        raise
+                    last_error = compat_str(e.cause or e.msg)
 
-            next_href = response.get('next_href')
+            for e in try_get(response, lambda x: x['collection'], list) or []:
-            if not next_href:
+                yield resolve_entry(e, e.get('track'), e.get('playlist'))
+
+            url = url_or_none(response.get('next_href'))
+            if not url:
                 break
 
-            next_href = response['next_href']
+            query.pop('offset', None)
-            parsed_next_href = compat_urlparse.urlparse(next_href)
-            query = compat_urlparse.parse_qs(parsed_next_href.query)
-            query.update(COMMON_QUERY)
-
-        return {
-            '_type': 'playlist',
-            'id': playlist_id,
-            'title': playlist_title,
-            'entries': entries,
-        }
 
 
 class SoundcloudUserIE(SoundcloudPagedPlaylistBaseIE):
@@ -696,7 +805,7 @@ class SoundcloudUserIE(SoundcloudPagedPlaylistBaseIE):
 
         user = self._download_json(
             self._resolv_url(self._BASE_URL + uploader),
-            uploader, 'Downloading user info')
+            uploader, 'Downloading user info', headers=self._HEADERS)
 
         resource = mobj.group('rsrc') or 'all'
 
@@ -721,7 +830,7 @@ class SoundcloudTrackStationIE(SoundcloudPagedPlaylistBaseIE):
     def _real_extract(self, url):
         track_name = self._match_id(url)
 
-        track = self._download_json(self._resolv_url(url), track_name)
+        track = self._download_json(self._resolv_url(url), track_name, headers=self._HEADERS)
         track_id = self._search_regex(
             r'soundcloud:track-stations:(\d+)', track['id'], 'track id')
 
@@ -730,6 +839,54 @@ class SoundcloudTrackStationIE(SoundcloudPagedPlaylistBaseIE):
             track_id, 'Track station: %s' % track['title'])
 
 
+class SoundcloudRelatedIE(SoundcloudPagedPlaylistBaseIE):
+    _VALID_URL = r'https?://(?:(?:www|m)\.)?soundcloud\.com/(?P<slug>[\w\d-]+/[\w\d-]+)/(?P<relation>albums|sets|recommended)'
+    IE_NAME = 'soundcloud:related'
+    _TESTS = [{
+        'url': 'https://soundcloud.com/wajang/sexapil-pingers-5/recommended',
+        'info_dict': {
+            'id': '1084577272',
+            'title': 'Sexapil - Pingers 5 (Recommended)',
+        },
+        'playlist_mincount': 50,
+    }, {
+        'url': 'https://soundcloud.com/wajang/sexapil-pingers-5/albums',
+        'info_dict': {
+            'id': '1084577272',
+            'title': 'Sexapil - Pingers 5 (Albums)',
+        },
+        'playlist_mincount': 1,
+    }, {
+        'url': 'https://soundcloud.com/wajang/sexapil-pingers-5/sets',
+        'info_dict': {
+            'id': '1084577272',
+            'title': 'Sexapil - Pingers 5 (Sets)',
+        },
+        'playlist_mincount': 4,
+    }]
+
+    _BASE_URL_MAP = {
+        'albums': 'tracks/%s/albums',
+        'sets': 'tracks/%s/playlists_without_albums',
+        'recommended': 'tracks/%s/related',
+    }
+
+    def _real_extract(self, url):
+        slug, relation = re.match(self._VALID_URL, url).group('slug', 'relation')
+
+        track = self._download_json(
+            self._resolv_url(self._BASE_URL + slug),
+            slug, 'Downloading track info', headers=self._HEADERS)
+
+        if track.get('errors'):
+            raise ExtractorError('%s said: %s' % (self.IE_NAME, ','.join(
+                compat_str(err['error_message']) for err in track['errors'])), expected=True)
+
+        return self._extract_playlist(
+            self._API_V2_BASE + self._BASE_URL_MAP[relation] % track['id'], compat_str(track['id']),
+            '%s (%s)' % (track.get('title') or slug, relation.capitalize()))
+
+
 class SoundcloudPlaylistIE(SoundcloudPlaylistBaseIE):
     _VALID_URL = r'https?://api(?:-v2)?\.soundcloud\.com/playlists/(?P<id>[0-9]+)(?:/?\?secret_token=(?P<token>[^&]+?))?$'
     IE_NAME = 'soundcloud:playlist'
@@ -754,23 +911,24 @@ class SoundcloudPlaylistIE(SoundcloudPlaylistBaseIE):
 
         data = self._download_json(
             self._API_V2_BASE + 'playlists/' + playlist_id,
-            playlist_id, 'Downloading playlist', query=query)
+            playlist_id, 'Downloading playlist', query=query, headers=self._HEADERS)
 
         return self._extract_set(data, token)
 
 
-class SoundcloudSearchIE(SearchInfoExtractor, SoundcloudIE):
+class SoundcloudSearchIE(SoundcloudBaseIE, SearchInfoExtractor):
     IE_NAME = 'soundcloud:search'
     IE_DESC = 'Soundcloud search'
-    _MAX_RESULTS = float('inf')
     _TESTS = [{
         'url': 'scsearch15:post-avant jazzcore',
         'info_dict': {
+            'id': 'post-avant jazzcore',
             'title': 'post-avant jazzcore',
         },
         'playlist_count': 15,
     }]
 
+    _MAX_RESULTS = float('inf')
     _SEARCH_KEY = 'scsearch'
     _MAX_RESULTS_PER_PAGE = 200
     _DEFAULT_RESULTS_PER_PAGE = 50
@@ -786,30 +944,20 @@ class SoundcloudSearchIE(SearchInfoExtractor, SoundcloudIE):
         })
         next_url = update_url_query(self._API_V2_BASE + endpoint, query)
 
-        collected_results = 0
-
         for i in itertools.count(1):
             response = self._download_json(
                 next_url, collection_id, 'Downloading page {0}'.format(i),
                 'Unable to download API page')
 
-            collection = response.get('collection', [])
+            for item in try_get(response, lambda x: x['collection'], list) or []:
-            if not collection:
+                if item:
-                break
+                    yield self.url_result(item['uri'], SoundcloudIE.ie_key())
-
-            collection = list(filter(bool, collection))
-            collected_results += len(collection)
-
-            for item in collection:
-                yield self.url_result(item['uri'], SoundcloudIE.ie_key())
-
-            if not collection or collected_results >= limit:
-                break
 
             next_url = response.get('next_href')
             if not next_url:
                 break
 
     def _get_n_results(self, query, n):
-        tracks = self._get_collection('search/tracks', query, limit=n, q=query)
+        return self.playlist_result(itertools.islice(
-        return self.playlist_result(tracks, playlist_title=query)
+            self._get_collection('search/tracks', query, limit=n, q=query),
+            0, None if n == float('inf') else n), query, query)

youtube_dl/extractor/youtube.py

@@ -27,6 +27,7 @@ from ..compat import (
 )
 from ..jsinterp import JSInterpreter
 from ..utils import (
+    bug_reports_message,
     clean_html,
     dict_get,
     error_to_compat_str,
@@ -65,6 +66,7 @@ from ..utils import (
     url_or_none,
     urlencode_postdata,
     urljoin,
+    variadic,
 )
 
 
@@ -120,7 +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,
@@ -460,6 +463,26 @@ class YoutubeBaseInfoExtractor(InfoExtractor):
             'uploader': uploader,
         }
 
+    @staticmethod
+    def _extract_thumbnails(data, *path_list, **kw_final_key):
+        """
+        Extract thumbnails from thumbnails dict
+        @param path_list: path list to level that contains 'thumbnails' key
+        """
+        final_key = kw_final_key.get('final_key', 'thumbnails')
+
+        return traverse_obj(data, ((
+            tuple(variadic(path) + (final_key, Ellipsis)
+                  for path in path_list or [()])), {
+            'url': ('url', T(url_or_none),
+                    # Sometimes youtube gives a wrong thumbnail URL. See:
+                    # https://github.com/yt-dlp/yt-dlp/issues/233
+                    # https://github.com/ytdl-org/youtube-dl/issues/28023
+                    T(lambda u: update_url(u, query=None) if u and 'maxresdefault' in u else u)),
+            'height': ('height', T(int_or_none)),
+            'width': ('width', T(int_or_none)),
+        }, T(lambda t: t if t.get('url') else None)))
+
     def _search_results(self, query, params):
         data = {
             'context': {
@@ -1829,12 +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):
@@ -3183,8 +3216,12 @@ class YoutubeTabIE(YoutubeBaseInfoExtractor):
             expected_type=txt_or_none)
 
     def _grid_entries(self, grid_renderer):
-        for item in grid_renderer['items']:
+        for item in traverse_obj(grid_renderer, ('items', Ellipsis, T(dict))):
-            if not isinstance(item, dict):
+            lockup_view_model = traverse_obj(item, ('lockupViewModel', T(dict)))
+            if lockup_view_model:
+                entry = self._extract_lockup_view_model(lockup_view_model)
+                if entry:
+                    yield entry
                 continue
             renderer = self._extract_grid_item_renderer(item)
             if not isinstance(renderer, dict):
@@ -3268,6 +3305,25 @@ class YoutubeTabIE(YoutubeBaseInfoExtractor):
                 continue
             yield self._extract_video(renderer)
 
+    def _extract_lockup_view_model(self, view_model):
+        content_id = view_model.get('contentId')
+        if not content_id:
+            return
+        content_type = view_model.get('contentType')
+        if content_type not in ('LOCKUP_CONTENT_TYPE_PLAYLIST', 'LOCKUP_CONTENT_TYPE_PODCAST'):
+            self.report_warning(
+                'Unsupported lockup view model content type "{0}"{1}'.format(content_type, bug_reports_message()), only_once=True)
+            return
+        return merge_dicts(self.url_result(
+            update_url_query('https://www.youtube.com/playlist', {'list': content_id}),
+            ie=YoutubeTabIE.ie_key(), video_id=content_id), {
+                'title': traverse_obj(view_model, (
+                    'metadata', 'lockupMetadataViewModel', 'title', 'content', T(compat_str))),
+                'thumbnails': self._extract_thumbnails(view_model, (
+                    'contentImage', 'collectionThumbnailViewModel', 'primaryThumbnail',
+                    'thumbnailViewModel', 'image'), final_key='sources'),
+        })
+
     def _video_entry(self, video_renderer):
         video_id = video_renderer.get('videoId')
         if video_id:

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:
+            # allow_undefined is None gives correct behaviour
-                # when is not allowed?
+            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')