Merge branch 'ytdl-org:master' into fix-npo-support

youtube_dl/YoutubeDL.py

@@ -7,6 +7,7 @@ import collections
 import copy
 import datetime
 import errno
+import functools
 import io
 import itertools
 import json
@@ -53,6 +54,7 @@ from .compat import (
     compat_urllib_request_DataHandler,
 )
 from .utils import (
+    _UnsafeExtensionError,
     age_restricted,
     args_to_str,
     bug_reports_message,
@@ -129,6 +131,20 @@ if compat_os_name == 'nt':
     import ctypes
 
 
+def _catch_unsafe_file_extension(func):
+    @functools.wraps(func)
+    def wrapper(self, *args, **kwargs):
+        try:
+            return func(self, *args, **kwargs)
+        except _UnsafeExtensionError as error:
+            self.report_error(
+                '{0} found; to avoid damaging your system, this value is disallowed.'
+                ' If you believe this is an error{1}'.format(
+                    error_to_compat_str(error), bug_reports_message(',')))
+
+    return wrapper
+
+
 class YoutubeDL(object):
     """YoutubeDL class.
 
@@ -1039,8 +1055,8 @@ class YoutubeDL(object):
         elif result_type in ('playlist', 'multi_video'):
             # Protect from infinite recursion due to recursively nested playlists
             # (see https://github.com/ytdl-org/youtube-dl/issues/27833)
-            webpage_url = ie_result['webpage_url']
-            if webpage_url in self._playlist_urls:
+            webpage_url = ie_result.get('webpage_url')  # not all pl/mv have this
+            if webpage_url and webpage_url in self._playlist_urls:
                 self.to_screen(
                     '[download] Skipping already downloaded playlist: %s'
                     % ie_result.get('title') or ie_result.get('id'))
@@ -1048,6 +1064,10 @@ class YoutubeDL(object):
 
             self._playlist_level += 1
             self._playlist_urls.add(webpage_url)
+            new_result = dict((k, v) for k, v in extra_info.items() if k not in ie_result)
+            if new_result:
+                new_result.update(ie_result)
+                ie_result = new_result
             try:
                 return self.__process_playlist(ie_result, download)
             finally:
@@ -1593,6 +1613,28 @@ class YoutubeDL(object):
         self.cookiejar.add_cookie_header(pr)
         return pr.get_header('Cookie')
 
+    def _fill_common_fields(self, info_dict, final=True):
+
+        for ts_key, date_key in (
+                ('timestamp', 'upload_date'),
+                ('release_timestamp', 'release_date'),
+        ):
+            if info_dict.get(date_key) is None and info_dict.get(ts_key) is not None:
+                # Working around out-of-range timestamp values (e.g. negative ones on Windows,
+                # see http://bugs.python.org/issue1646728)
+                try:
+                    upload_date = datetime.datetime.utcfromtimestamp(info_dict[ts_key])
+                    info_dict[date_key] = compat_str(upload_date.strftime('%Y%m%d'))
+                except (ValueError, OverflowError, OSError):
+                    pass
+
+        # Auto generate title fields corresponding to the *_number fields when missing
+        # in order to always have clean titles. This is very common for TV series.
+        if final:
+            for field in ('chapter', 'season', 'episode'):
+                if info_dict.get('%s_number' % field) is not None and not info_dict.get(field):
+                    info_dict[field] = '%s %d' % (field.capitalize(), info_dict['%s_number' % field])
+
     def process_video_result(self, info_dict, download=True):
         assert info_dict.get('_type', 'video') == 'video'
 
@@ -1660,24 +1702,7 @@ class YoutubeDL(object):
         if 'display_id' not in info_dict and 'id' in info_dict:
             info_dict['display_id'] = info_dict['id']
 
-        for ts_key, date_key in (
-                ('timestamp', 'upload_date'),
-                ('release_timestamp', 'release_date'),
-        ):
-            if info_dict.get(date_key) is None and info_dict.get(ts_key) is not None:
-                # Working around out-of-range timestamp values (e.g. negative ones on Windows,
-                # see http://bugs.python.org/issue1646728)
-                try:
-                    upload_date = datetime.datetime.utcfromtimestamp(info_dict[ts_key])
-                    info_dict[date_key] = compat_str(upload_date.strftime('%Y%m%d'))
-                except (ValueError, OverflowError, OSError):
-                    pass
-
-        # Auto generate title fields corresponding to the *_number fields when missing
-        # in order to always have clean titles. This is very common for TV series.
-        for field in ('chapter', 'season', 'episode'):
-            if info_dict.get('%s_number' % field) is not None and not info_dict.get(field):
-                info_dict[field] = '%s %d' % (field.capitalize(), info_dict['%s_number' % field])
+        self._fill_common_fields(info_dict)
 
         for cc_kind in ('subtitles', 'automatic_captions'):
             cc = info_dict.get(cc_kind)
@@ -1916,6 +1941,7 @@ class YoutubeDL(object):
         if self.params.get('forcejson', False):
             self.to_stdout(json.dumps(self.sanitize_info(info_dict)))
 
+    @_catch_unsafe_file_extension
     def process_info(self, info_dict):
         """Process a single resolved IE result."""
 
@@ -2088,18 +2114,26 @@ class YoutubeDL(object):
                         # TODO: Check acodec/vcodec
                         return False
 
-                    filename_real_ext = os.path.splitext(filename)[1][1:]
-                    filename_wo_ext = (
-                        os.path.splitext(filename)[0]
-                        if filename_real_ext == info_dict['ext']
-                        else filename)
+                    exts = [info_dict['ext']]
                     requested_formats = info_dict['requested_formats']
                     if self.params.get('merge_output_format') is None and not compatible_formats(requested_formats):
                         info_dict['ext'] = 'mkv'
                         self.report_warning(
                             'Requested formats are incompatible for merge and will be merged into mkv.')
+                    exts.append(info_dict['ext'])
+
                     # Ensure filename always has a correct extension for successful merge
-                    filename = '%s.%s' % (filename_wo_ext, info_dict['ext'])
+                    def correct_ext(filename, ext=exts[1]):
+                        if filename == '-':
+                            return filename
+                        f_name, f_real_ext = os.path.splitext(filename)
+                        f_real_ext = f_real_ext[1:]
+                        filename_wo_ext = f_name if f_real_ext in exts else filename
+                        if ext is None:
+                            ext = f_real_ext or None
+                        return join_nonempty(filename_wo_ext, ext, delim='.')
+
+                    filename = correct_ext(filename)
                     if os.path.exists(encodeFilename(filename)):
                         self.to_screen(
                             '[download] %s has already been downloaded and '
@@ -2109,8 +2143,9 @@ class YoutubeDL(object):
                             new_info = dict(info_dict)
                             new_info.update(f)
                             fname = prepend_extension(
-                                self.prepare_filename(new_info),
-                                'f%s' % f['format_id'], new_info['ext'])
+                                correct_ext(
+                                    self.prepare_filename(new_info), new_info['ext']),
+                                'f%s' % (f['format_id'],), new_info['ext'])
                             if not ensure_dir_exists(fname):
                                 return
                             downloaded.append(fname)

youtube_dl/__init__.py

@@ -21,6 +21,7 @@ from .compat import (
     workaround_optparse_bug9161,
 )
 from .utils import (
+    _UnsafeExtensionError,
     DateRange,
     decodeOption,
     DEFAULT_OUTTMPL,
@@ -173,6 +174,9 @@ def _real_main(argv=None):
     if opts.ap_mso and opts.ap_mso not in MSO_INFO:
         parser.error('Unsupported TV Provider, use --ap-list-mso to get a list of supported TV Providers')
 
+    if opts.no_check_extensions:
+        _UnsafeExtensionError.lenient = True
+
     def parse_retries(retries):
         if retries in ('inf', 'infinite'):
             parsed_retries = float('inf')

youtube_dl/compat.py

@@ -2421,29 +2421,26 @@ except ImportError:  # Python 2
 compat_urllib_request_urlretrieve = compat_urlretrieve
 
 try:
+    from HTMLParser import (
+        HTMLParser as compat_HTMLParser,
+        HTMLParseError as compat_HTMLParseError)
+except ImportError:  # Python 3
     from html.parser import HTMLParser as compat_HTMLParser
-except ImportError:  # Python 2
-    from HTMLParser import HTMLParser as compat_HTMLParser
-compat_html_parser_HTMLParser = compat_HTMLParser
-
-try:  # Python 2
-    from HTMLParser import HTMLParseError as compat_HTMLParseError
-except ImportError:  # Python <3.4
     try:
         from html.parser import HTMLParseError as compat_HTMLParseError
     except ImportError:  # Python >3.4
-
-        # HTMLParseError has been deprecated in Python 3.3 and removed in
+        # 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
 
 try:
-    from subprocess import DEVNULL
-    compat_subprocess_get_DEVNULL = lambda: DEVNULL
-except ImportError:
+    _DEVNULL = subprocess.DEVNULL
+    compat_subprocess_get_DEVNULL = lambda: _DEVNULL
+except AttributeError:
     compat_subprocess_get_DEVNULL = lambda: open(os.path.devnull, 'w')
 
 try:
@@ -2722,8 +2719,222 @@ if sys.version_info < (2, 7):
         if isinstance(xpath, compat_str):
             xpath = xpath.encode('ascii')
         return xpath
+
+    # further code below based on CPython 2.7 source
+    import functools
+
+    _xpath_tokenizer_re = re.compile(r'''(?x)
+        (                                   # (1)
+            '[^']*'|"[^"]*"|                # quoted strings, or
+            ::|//?|\.\.|\(\)|[/.*:[\]()@=]  # navigation specials
+        )|                                  # or (2)
+        ((?:\{[^}]+\})?[^/[\]()@=\s]+)|     # token: optional {ns}, no specials
+        \s+                                 # or white space
+    ''')
+
+    def _xpath_tokenizer(pattern, namespaces=None):
+        for token in _xpath_tokenizer_re.findall(pattern):
+            tag = token[1]
+            if tag and tag[0] != "{" and ":" in tag:
+                try:
+                    if not namespaces:
+                        raise KeyError
+                    prefix, uri = tag.split(":", 1)
+                    yield token[0], "{%s}%s" % (namespaces[prefix], uri)
+                except KeyError:
+                    raise SyntaxError("prefix %r not found in prefix map" % prefix)
+            else:
+                yield token
+
+    def _get_parent_map(context):
+        parent_map = context.parent_map
+        if parent_map is None:
+            context.parent_map = parent_map = {}
+            for p in context.root.getiterator():
+                for e in p:
+                    parent_map[e] = p
+        return parent_map
+
+    def _select(context, result, filter_fn=lambda *_: True):
+        for elem in result:
+            for e in elem:
+                if filter_fn(e, elem):
+                    yield e
+
+    def _prepare_child(next_, token):
+        tag = token[1]
+        return functools.partial(_select, filter_fn=lambda e, _: e.tag == tag)
+
+    def _prepare_star(next_, token):
+        return _select
+
+    def _prepare_self(next_, token):
+        return lambda _, result: (e for e in result)
+
+    def _prepare_descendant(next_, token):
+        token = next(next_)
+        if token[0] == "*":
+            tag = "*"
+        elif not token[0]:
+            tag = token[1]
+        else:
+            raise SyntaxError("invalid descendant")
+
+        def select(context, result):
+            for elem in result:
+                for e in elem.getiterator(tag):
+                    if e is not elem:
+                        yield e
+        return select
+
+    def _prepare_parent(next_, token):
+        def select(context, result):
+            # FIXME: raise error if .. is applied at toplevel?
+            parent_map = _get_parent_map(context)
+            result_map = {}
+            for elem in result:
+                if elem in parent_map:
+                    parent = parent_map[elem]
+                    if parent not in result_map:
+                        result_map[parent] = None
+                        yield parent
+        return select
+
+    def _prepare_predicate(next_, token):
+        signature = []
+        predicate = []
+        for token in next_:
+            if token[0] == "]":
+                break
+            if token[0] and token[0][:1] in "'\"":
+                token = "'", token[0][1:-1]
+            signature.append(token[0] or "-")
+            predicate.append(token[1])
+
+        def select(context, result, filter_fn=lambda _: True):
+            for elem in result:
+                if filter_fn(elem):
+                    yield elem
+
+        signature = "".join(signature)
+        # use signature to determine predicate type
+        if signature == "@-":
+            # [@attribute] predicate
+            key = predicate[1]
+            return functools.partial(
+                select, filter_fn=lambda el: el.get(key) is not None)
+        if signature == "@-='":
+            # [@attribute='value']
+            key = predicate[1]
+            value = predicate[-1]
+            return functools.partial(
+                select, filter_fn=lambda el: el.get(key) == value)
+        if signature == "-" and not re.match(r"\d+$", predicate[0]):
+            # [tag]
+            tag = predicate[0]
+            return functools.partial(
+                select, filter_fn=lambda el: el.find(tag) is not None)
+        if signature == "-='" and not re.match(r"\d+$", predicate[0]):
+            # [tag='value']
+            tag = predicate[0]
+            value = predicate[-1]
+
+            def itertext(el):
+                for e in el.getiterator():
+                    e = e.text
+                    if e:
+                        yield e
+
+            def select(context, result):
+                for elem in result:
+                    for e in elem.findall(tag):
+                        if "".join(itertext(e)) == value:
+                            yield elem
+                            break
+            return select
+        if signature == "-" or signature == "-()" or signature == "-()-":
+            # [index] or [last()] or [last()-index]
+            if signature == "-":
+                index = int(predicate[0]) - 1
+            else:
+                if predicate[0] != "last":
+                    raise SyntaxError("unsupported function")
+                if signature == "-()-":
+                    try:
+                        index = int(predicate[2]) - 1
+                    except ValueError:
+                        raise SyntaxError("unsupported expression")
+                else:
+                    index = -1
+
+            def select(context, result):
+                parent_map = _get_parent_map(context)
+                for elem in result:
+                    try:
+                        parent = parent_map[elem]
+                        # FIXME: what if the selector is "*" ?
+                        elems = list(parent.findall(elem.tag))
+                        if elems[index] is elem:
+                            yield elem
+                    except (IndexError, KeyError):
+                        pass
+            return select
+        raise SyntaxError("invalid predicate")
+
+    ops = {
+        "": _prepare_child,
+        "*": _prepare_star,
+        ".": _prepare_self,
+        "..": _prepare_parent,
+        "//": _prepare_descendant,
+        "[": _prepare_predicate,
+    }
+
+    _cache = {}
+
+    class _SelectorContext:
+        parent_map = None
+
+        def __init__(self, root):
+            self.root = root
+
+    ##
+    # Generate all matching objects.
+
+    def compat_etree_iterfind(elem, path, namespaces=None):
+        # compile selector pattern
+        if path[-1:] == "/":
+            path = path + "*"  # implicit all (FIXME: keep this?)
+        try:
+            selector = _cache[path]
+        except KeyError:
+            if len(_cache) > 100:
+                _cache.clear()
+            if path[:1] == "/":
+                raise SyntaxError("cannot use absolute path on element")
+            tokens = _xpath_tokenizer(path, namespaces)
+            selector = []
+            for token in tokens:
+                if token[0] == "/":
+                    continue
+                try:
+                    selector.append(ops[token[0]](tokens, token))
+                except StopIteration:
+                    raise SyntaxError("invalid path")
+            _cache[path] = selector
+        # execute selector pattern
+        result = [elem]
+        context = _SelectorContext(elem)
+        for select in selector:
+            result = select(context, result)
+        return result
+
+    # end of code based on CPython 2.7 source
+
+
 else:
     compat_xpath = lambda xpath: xpath
+    compat_etree_iterfind = lambda element, match: element.iterfind(match)
 
 
 compat_os_name = os._name if os.name == 'java' else os.name
@@ -2759,7 +2970,7 @@ except (AssertionError, UnicodeEncodeError):
 
 
 def compat_ord(c):
-    if type(c) is int:
+    if isinstance(c, int):
         return c
     else:
         return ord(c)
@@ -2943,6 +3154,51 @@ else:
     compat_socket_create_connection = socket.create_connection
 
 
+try:
+    from contextlib import suppress as compat_contextlib_suppress
+except ImportError:
+    class compat_contextlib_suppress(object):
+        _exceptions = None
+
+        def __init__(self, *exceptions):
+            super(compat_contextlib_suppress, self).__init__()
+            # TODO: [Base]ExceptionGroup (3.12+)
+            self._exceptions = exceptions
+
+        def __enter__(self):
+            return self
+
+        def __exit__(self, exc_type, exc_val, exc_tb):
+            return exc_type is not None and issubclass(exc_type, self._exceptions or tuple())
+
+
+# subprocess.Popen context manager
+# avoids leaking handles if .communicate() is not called
+try:
+    _Popen = subprocess.Popen
+    # check for required context manager attributes
+    _Popen.__enter__ and _Popen.__exit__
+    compat_subprocess_Popen = _Popen
+except AttributeError:
+    # not a context manager - make one
+    from contextlib import contextmanager
+
+    @contextmanager
+    def compat_subprocess_Popen(*args, **kwargs):
+        popen = None
+        try:
+            popen = _Popen(*args, **kwargs)
+            yield popen
+        finally:
+            if popen:
+                for f in (popen.stdin, popen.stdout, popen.stderr):
+                    if f:
+                        # repeated .close() is OK, but just in case
+                        with compat_contextlib_suppress(EnvironmentError):
+                            f.close()
+                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():
@@ -3263,8 +3519,10 @@ __all__ = [
     'compat_http_cookiejar_Cookie',
     'compat_http_cookies',
     'compat_http_cookies_SimpleCookie',
+    'compat_contextlib_suppress',
     'compat_ctypes_WINFUNCTYPE',
     'compat_etree_fromstring',
+    'compat_etree_iterfind',
     'compat_filter',
     'compat_get_terminal_size',
     'compat_getenv',
@@ -3298,6 +3556,7 @@ __all__ = [
     'compat_struct_pack',
     'compat_struct_unpack',
     'compat_subprocess_get_DEVNULL',
+    'compat_subprocess_Popen',
     'compat_tokenize_tokenize',
     'compat_urllib_error',
     'compat_urllib_parse',

youtube_dl/downloader/external.py

@@ -11,8 +11,14 @@ from .common import FileDownloader
 from ..compat import (
     compat_setenv,
     compat_str,
+    compat_subprocess_Popen,
 )
-from ..postprocessor.ffmpeg import FFmpegPostProcessor, EXT_TO_OUT_FORMATS
+
+try:
+    from ..postprocessor.ffmpeg import FFmpegPostProcessor, EXT_TO_OUT_FORMATS
+except ImportError:
+    FFmpegPostProcessor = None
+
 from ..utils import (
     cli_option,
     cli_valueless_option,
@@ -361,13 +367,14 @@ class FFmpegFD(ExternalFD):
 
     @classmethod
     def available(cls):
-        return FFmpegPostProcessor().available
+        # actual availability can only be confirmed for an instance
+        return bool(FFmpegPostProcessor)
 
     def _call_downloader(self, tmpfilename, info_dict):
-        url = info_dict['url']
-        ffpp = FFmpegPostProcessor(downloader=self)
+        # `downloader` means the parent `YoutubeDL`
+        ffpp = FFmpegPostProcessor(downloader=self.ydl)
         if not ffpp.available:
-            self.report_error('m3u8 download detected but ffmpeg or avconv could not be found. Please install one.')
+            self.report_error('ffmpeg required for download but no ffmpeg (nor avconv) executable could be found. Please install one.')
             return False
         ffpp.check_version()
 
@@ -396,6 +403,7 @@ class FFmpegFD(ExternalFD):
         # if end_time:
         #     args += ['-t', compat_str(end_time - start_time)]
 
+        url = info_dict['url']
         cookies = self.ydl.cookiejar.get_cookies_for_url(url)
         if cookies:
             args.extend(['-cookies', ''.join(
@@ -483,21 +491,25 @@ class FFmpegFD(ExternalFD):
 
         self._debug_cmd(args)
 
-        proc = subprocess.Popen(args, stdin=subprocess.PIPE, env=env)
-        try:
-            retval = proc.wait()
-        except BaseException as e:
-            # subprocess.run would send the SIGKILL signal to ffmpeg and the
-            # mp4 file couldn't be played, but if we ask ffmpeg to quit it
-            # produces a file that is playable (this is mostly useful for live
-            # streams). Note that Windows is not affected and produces playable
-            # files (see https://github.com/ytdl-org/youtube-dl/issues/8300).
-            if isinstance(e, KeyboardInterrupt) and sys.platform != 'win32':
-                process_communicate_or_kill(proc, b'q')
-            else:
-                proc.kill()
-                proc.wait()
-            raise
+        # From [1], a PIPE opened in Popen() should be closed, unless
+        # .communicate() is called. Avoid leaking any PIPEs by using Popen
+        # as a context manager (newer Python 3.x and compat)
+        # Fixes "Resource Warning" in test/test_downloader_external.py
+        # [1] https://devpress.csdn.net/python/62fde12d7e66823466192e48.html
+        with compat_subprocess_Popen(args, stdin=subprocess.PIPE, env=env) as proc:
+            try:
+                retval = proc.wait()
+            except BaseException as e:
+                # subprocess.run would send the SIGKILL signal to ffmpeg and the
+                # mp4 file couldn't be played, but if we ask ffmpeg to quit it
+                # produces a file that is playable (this is mostly useful for live
+                # streams). Note that Windows is not affected and produces playable
+                # files (see https://github.com/ytdl-org/youtube-dl/issues/8300).
+                if isinstance(e, KeyboardInterrupt) and sys.platform != 'win32':
+                    process_communicate_or_kill(proc, b'q')
+                else:
+                    proc.kill()
+                raise
         return retval
 
 

youtube_dl/extractor/common.py

@@ -1169,10 +1169,10 @@ class InfoExtractor(object):
     def _get_netrc_login_info(self, netrc_machine=None):
         username = None
         password = None
-        netrc_machine = netrc_machine or self._NETRC_MACHINE
 
         if self._downloader.params.get('usenetrc', False):
             try:
+                netrc_machine = netrc_machine or self._NETRC_MACHINE
                 info = netrc.netrc().authenticators(netrc_machine)
                 if info is not None:
                     username = info[0]
@@ -1180,7 +1180,7 @@ class InfoExtractor(object):
                 else:
                     raise netrc.NetrcParseError(
                         'No authenticators for %s' % netrc_machine)
-            except (IOError, netrc.NetrcParseError) as err:
+            except (AttributeError, IOError, netrc.NetrcParseError) as err:
                 self._downloader.report_warning(
                     'parsing .netrc: %s' % error_to_compat_str(err))
 
@@ -1490,14 +1490,18 @@ class InfoExtractor(object):
         return dict((k, v) for k, v in info.items() if v is not None)
 
     def _search_nextjs_data(self, webpage, video_id, **kw):
-        nkw = dict((k, v) for k, v in kw.items() if k in ('transform_source', 'fatal'))
-        kw.pop('transform_source', None)
-        next_data = self._search_regex(
-            r'''<script[^>]+\bid\s*=\s*('|")__NEXT_DATA__\1[^>]*>(?P<nd>[^<]+)</script>''',
-            webpage, 'next.js data', group='nd', **kw)
-        if not next_data:
-            return {}
-        return self._parse_json(next_data, video_id, **nkw)
+        # ..., *, transform_source=None, fatal=True, default=NO_DEFAULT
+
+        # TODO: remove this backward compat
+        default = kw.get('default', NO_DEFAULT)
+        if default == '{}':
+            kw['default'] = {}
+            kw = compat_kwargs(kw)
+
+        return self._search_json(
+            r'''<script\s[^>]*?\bid\s*=\s*('|")__NEXT_DATA__\1[^>]*>''',
+            webpage, 'next.js data', video_id, end_pattern='</script>',
+            **kw)
 
     def _search_nuxt_data(self, webpage, video_id, *args, **kwargs):
         """Parses Nuxt.js metadata. This works as long as the function __NUXT__ invokes is a pure function"""
@@ -3029,7 +3033,6 @@ class InfoExtractor(object):
             transform_source=transform_source, default=None)
 
     def _extract_jwplayer_data(self, webpage, video_id, *args, **kwargs):
-
         # allow passing `transform_source` through to _find_jwplayer_data()
         transform_source = kwargs.pop('transform_source', None)
         kwfind = compat_kwargs({'transform_source': transform_source}) if transform_source else {}
@@ -3296,12 +3299,16 @@ class InfoExtractor(object):
         return ret
 
     @classmethod
-    def _merge_subtitles(cls, subtitle_dict1, subtitle_dict2):
-        """ Merge two subtitle dictionaries, language by language. """
-        ret = dict(subtitle_dict1)
-        for lang in subtitle_dict2:
-            ret[lang] = cls._merge_subtitle_items(subtitle_dict1.get(lang, []), subtitle_dict2[lang])
-        return ret
+    def _merge_subtitles(cls, subtitle_dict1, *subtitle_dicts, **kwargs):
+        """ Merge subtitle dictionaries, language by language. """
+
+        # ..., * , target=None
+        target = kwargs.get('target') or dict(subtitle_dict1)
+
+        for subtitle_dict in subtitle_dicts:
+            for lang in subtitle_dict:
+                target[lang] = cls._merge_subtitle_items(target.get(lang, []), subtitle_dict[lang])
+        return target
 
     def extract_automatic_captions(self, *args, **kwargs):
         if (self._downloader.params.get('writeautomaticsub', False)
@@ -3334,6 +3341,29 @@ class InfoExtractor(object):
     def _generic_title(self, url):
         return compat_urllib_parse_unquote(os.path.splitext(url_basename(url))[0])
 
+    def _yes_playlist(self, playlist_id, video_id, *args, **kwargs):
+        # smuggled_data=None, *, playlist_label='playlist', video_label='video'
+        smuggled_data = args[0] if len(args) == 1 else kwargs.get('smuggled_data')
+        playlist_label = kwargs.get('playlist_label', 'playlist')
+        video_label = kwargs.get('video_label', 'video')
+
+        if not playlist_id or not video_id:
+            return not video_id
+
+        no_playlist = (smuggled_data or {}).get('force_noplaylist')
+        if no_playlist is not None:
+            return not no_playlist
+
+        video_id = '' if video_id is True else ' ' + video_id
+        noplaylist = self.get_param('noplaylist')
+        self.to_screen(
+            'Downloading just the {0}{1} because of --no-playlist'.format(video_label, video_id)
+            if noplaylist else
+            'Downloading {0}{1} - add --no-playlist to download just the {2}{3}'.format(
+                playlist_label, '' if playlist_id is True else ' ' + playlist_id,
+                video_label, video_id))
+        return not noplaylist
+
 
 class SearchInfoExtractor(InfoExtractor):
     """

youtube_dl/extractor/extractors.py

@@ -897,21 +897,13 @@ from .ooyala import (
 )
 from .ora import OraTVIE
 from .orf import (
-    ORFTVthekIE,
-    ORFFM4IE,
+    ORFONIE,
+    ORFONLiveIE,
     ORFFM4StoryIE,
-    ORFOE1IE,
-    ORFOE3IE,
-    ORFNOEIE,
-    ORFWIEIE,
-    ORFBGLIE,
-    ORFOOEIE,
-    ORFSTMIE,
-    ORFKTNIE,
-    ORFSBGIE,
-    ORFTIRIE,
-    ORFVBGIE,
     ORFIPTVIE,
+    ORFPodcastIE,
+    ORFRadioIE,
+    ORFRadioCollectionIE,
 )
 from .outsidetv import OutsideTVIE
 from .packtpub import (
@@ -1652,7 +1644,15 @@ from .younow import (
     YouNowChannelIE,
     YouNowMomentIE,
 )
-from .youporn import YouPornIE
+from .youporn import (
+    YouPornIE,
+    YouPornCategoryIE,
+    YouPornChannelIE,
+    YouPornCollectionIE,
+    YouPornStarIE,
+    YouPornTagIE,
+    YouPornVideosIE,
+)
 from .yourporn import YourPornIE
 from .yourupload import YourUploadIE
 from .youtube import (

youtube_dl/extractor/mixcloud.py

@@ -1,3 +1,4 @@
+# coding: utf-8
 from __future__ import unicode_literals
 
 import itertools
@@ -10,7 +11,7 @@ from ..compat import (
     compat_ord,
     compat_str,
     compat_urllib_parse_unquote,
-    compat_zip
+    compat_zip as zip,
 )
 from ..utils import (
     int_or_none,
@@ -24,7 +25,7 @@ class MixcloudBaseIE(InfoExtractor):
     def _call_api(self, object_type, object_fields, display_id, username, slug=None):
         lookup_key = object_type + 'Lookup'
         return self._download_json(
-            'https://www.mixcloud.com/graphql', display_id, query={
+            'https://app.mixcloud.com/graphql', display_id, query={
                 'query': '''{
   %s(lookup: {username: "%s"%s}) {
     %s
@@ -44,7 +45,7 @@ class MixcloudIE(MixcloudBaseIE):
             'ext': 'm4a',
             'title': 'Cryptkeeper',
             'description': 'After quite a long silence from myself, finally another Drum\'n\'Bass mix with my favourite current dance floor bangers.',
-            'uploader': 'Daniel Holbach',
+            'uploader': 'dholbach',  # was: 'Daniel Holbach',
             'uploader_id': 'dholbach',
             'thumbnail': r're:https?://.*\.jpg',
             'view_count': int,
@@ -57,7 +58,7 @@ class MixcloudIE(MixcloudBaseIE):
             'id': 'gillespeterson_caribou-7-inch-vinyl-mix-chat',
             'ext': 'mp3',
             'title': 'Caribou 7 inch Vinyl Mix & Chat',
-            'description': 'md5:2b8aec6adce69f9d41724647c65875e8',
+            'description': r're:Last week Dan Snaith aka Caribou swung by the Brownswood.{136}',
             'uploader': 'Gilles Peterson Worldwide',
             'uploader_id': 'gillespeterson',
             'thumbnail': 're:https?://.*',
@@ -65,6 +66,23 @@ class MixcloudIE(MixcloudBaseIE):
             'timestamp': 1422987057,
             'upload_date': '20150203',
         },
+        'params': {
+            'skip_download': '404 not found',
+        },
+    }, {
+        'url': 'https://www.mixcloud.com/gillespeterson/carnival-m%C3%BAsica-popular-brasileira-mix/',
+        'info_dict': {
+            'id': 'gillespeterson_carnival-música-popular-brasileira-mix',
+            'ext': 'm4a',
+            'title': 'Carnival Música Popular Brasileira Mix',
+            'description': r're:Gilles was recently in Brazil to play at Boiler Room.{208}',
+            'timestamp': 1454347174,
+            'upload_date': '20160201',
+            'uploader': 'Gilles Peterson Worldwide',
+            'uploader_id': 'gillespeterson',
+            'thumbnail': 're:https?://.*',
+            'view_count': int,
+        },
     }, {
         'url': 'https://beta.mixcloud.com/RedLightRadio/nosedrip-15-red-light-radio-01-18-2016/',
         'only_matching': True,
@@ -76,10 +94,10 @@ class MixcloudIE(MixcloudBaseIE):
         """Encrypt/Decrypt XOR cipher. Both ways are possible because it's XOR."""
         return ''.join([
             compat_chr(compat_ord(ch) ^ compat_ord(k))
-            for ch, k in compat_zip(ciphertext, itertools.cycle(key))])
+            for ch, k in zip(ciphertext, itertools.cycle(key))])
 
     def _real_extract(self, url):
-        username, slug = re.match(self._VALID_URL, url).groups()
+        username, slug = self._match_valid_url(url).groups()
         username, slug = compat_urllib_parse_unquote(username), compat_urllib_parse_unquote(slug)
         track_id = '%s_%s' % (username, slug)
 

youtube_dl/extractor/palcomp3.py

@@ -8,7 +8,7 @@ from ..compat import compat_str
 from ..utils import (
     int_or_none,
     str_or_none,
-    try_get,
+    traverse_obj,
 )
 
 
@@ -109,7 +109,7 @@ class PalcoMP3ArtistIE(PalcoMP3BaseIE):
     }
     name'''
 
-    @ classmethod
+    @classmethod
     def suitable(cls, url):
         return False if re.match(PalcoMP3IE._VALID_URL, url) else super(PalcoMP3ArtistIE, cls).suitable(url)
 
@@ -118,7 +118,8 @@ class PalcoMP3ArtistIE(PalcoMP3BaseIE):
         artist = self._call_api(artist_slug, self._ARTIST_FIELDS_TMPL)['artist']
 
         def entries():
-            for music in (try_get(artist, lambda x: x['musics']['nodes'], list) or []):
+            for music in traverse_obj(artist, (
+                    'musics', 'nodes', lambda _, m: m['musicID'])):
                 yield self._parse_music(music)
 
         return self.playlist_result(
@@ -137,7 +138,7 @@ class PalcoMP3VideoIE(PalcoMP3BaseIE):
             'title': 'Maiara e Maraisa - Você Faz Falta Aqui - DVD Ao Vivo Em Campo Grande',
             'description': 'md5:7043342c09a224598e93546e98e49282',
             'upload_date': '20161107',
-            'uploader_id': 'maiaramaraisaoficial',
+            'uploader_id': '@maiaramaraisaoficial',
             'uploader': 'Maiara e Maraisa',
         }
     }]

youtube_dl/extractor/vidlii.py

@@ -4,6 +4,7 @@ from __future__ import unicode_literals
 import re
 
 from .common import InfoExtractor
+
 from ..utils import (
     float_or_none,
     get_element_by_id,
@@ -11,6 +12,7 @@ from ..utils import (
     strip_or_none,
     unified_strdate,
     urljoin,
+    str_to_int,
 )
 
 
@@ -35,6 +37,26 @@ class VidLiiIE(InfoExtractor):
             'categories': ['News & Politics'],
             'tags': ['Vidlii', 'Jan', 'Videogames'],
         }
+    }, {
+        # HD
+        'url': 'https://www.vidlii.com/watch?v=2Ng8Abj2Fkl',
+        'md5': '450e7da379c884788c3a4fa02a3ce1a4',
+        'info_dict': {
+            'id': '2Ng8Abj2Fkl',
+            'ext': 'mp4',
+            'title': 'test',
+            'description': 'md5:cc55a86032a7b6b3cbfd0f6b155b52e9',
+            'thumbnail': 'https://www.vidlii.com/usfi/thmp/2Ng8Abj2Fkl.jpg',
+            'uploader': 'VidLii',
+            'uploader_url': 'https://www.vidlii.com/user/VidLii',
+            'upload_date': '20200927',
+            'duration': 5,
+            'view_count': int,
+            'comment_count': int,
+            'average_rating': float,
+            'categories': ['Film & Animation'],
+            'tags': list,
+        },
     }, {
         'url': 'https://www.vidlii.com/embed?v=tJluaH4BJ3v&a=0',
         'only_matching': True,
@@ -46,11 +68,32 @@ class VidLiiIE(InfoExtractor):
         webpage = self._download_webpage(
             'https://www.vidlii.com/watch?v=%s' % video_id, video_id)
 
-        video_url = self._search_regex(
-            r'src\s*:\s*(["\'])(?P<url>(?:https?://)?(?:(?!\1).)+)\1', webpage,
-            'video url', group='url')
+        formats = []
 
-        title = self._search_regex(
+        def add_format(format_url, height=None):
+            height = int(self._search_regex(r'(\d+)\.mp4',
+                         format_url, 'height', default=360))
+
+            formats.append({
+                'url': format_url,
+                'format_id': '%dp' % height if height else None,
+                'height': height,
+            })
+
+        sources = re.findall(
+            r'src\s*:\s*(["\'])(?P<url>(?:https?://)?(?:(?!\1).)+)\1',
+            webpage)
+
+        formats = []
+        if len(sources) > 1:
+            add_format(sources[1][1])
+            self._check_formats(formats, video_id)
+        if len(sources) > 0:
+            add_format(sources[0][1])
+
+        self._sort_formats(formats)
+
+        title = self._html_search_regex(
             (r'<h1>([^<]+)</h1>', r'<title>([^<]+) - VidLii<'), webpage,
             'title')
 
@@ -82,9 +125,9 @@ class VidLiiIE(InfoExtractor):
             default=None) or self._search_regex(
             r'duration\s*:\s*(\d+)', webpage, 'duration', fatal=False))
 
-        view_count = int_or_none(self._search_regex(
-            (r'<strong>(\d+)</strong> views',
-             r'Views\s*:\s*<strong>(\d+)</strong>'),
+        view_count = str_to_int(self._html_search_regex(
+            (r'<strong>([\d,.]+)</strong> views',
+             r'Views\s*:\s*<strong>([\d,.]+)</strong>'),
             webpage, 'view count', fatal=False))
 
         comment_count = int_or_none(self._search_regex(
@@ -109,7 +152,7 @@ class VidLiiIE(InfoExtractor):
 
         return {
             'id': video_id,
-            'url': video_url,
+            'formats': formats,
             'title': title,
             'description': description,
             'thumbnail': thumbnail,

youtube_dl/extractor/yandexmusic.py

@@ -106,6 +106,25 @@ class YandexMusicTrackIE(YandexMusicBaseIE):
     }, {
         'url': 'http://music.yandex.com/album/540508/track/4878838',
         'only_matching': True,
+    }, {
+        'url': 'https://music.yandex.ru/album/16302456/track/85430762',
+        'md5': '11b8d50ab03b57738deeaadf661a0a48',
+        'info_dict': {
+            'id': '85430762',
+            'ext': 'mp3',
+            'abr': 128,
+            'title': 'Haddadi Von Engst, Phonic Youth, Super Flu - Til The End (Super Flu Remix)',
+            'filesize': int,
+            'duration': 431.14,
+            'track': 'Til The End (Super Flu Remix)',
+            'album': 'Til The End',
+            'album_artist': 'Haddadi Von Engst, Phonic Youth',
+            'artist': 'Haddadi Von Engst, Phonic Youth, Super Flu',
+            'release_year': 2021,
+            'genre': 'house',
+            'disc_number': 1,
+            'track_number': 2,
+        }
     }]
 
     def _real_extract(self, url):
@@ -116,10 +135,14 @@ class YandexMusicTrackIE(YandexMusicBaseIE):
             'track', tld, url, track_id, 'Downloading track JSON',
             {'track': '%s:%s' % (track_id, album_id)})['track']
         track_title = track['title']
+        track_version = track.get('version')
+        if track_version:
+            track_title = '%s (%s)' % (track_title, track_version)
 
         download_data = self._download_json(
             'https://music.yandex.ru/api/v2.1/handlers/track/%s:%s/web-album_track-track-track-main/download/m' % (track_id, album_id),
             track_id, 'Downloading track location url JSON',
+            query={'hq': 1},
             headers={'X-Retpath-Y': url})
 
         fd_data = self._download_json(

youtube_dl/extractor/youporn.py

@@ -1,20 +1,38 @@
 # coding: utf-8
 from __future__ import unicode_literals
 
+import itertools
 import re
+from time import sleep
 
 from .common import InfoExtractor
 from ..utils import (
+    clean_html,
     extract_attributes,
+    ExtractorError,
+    get_element_by_class,
+    get_element_by_id,
     int_or_none,
-    str_to_int,
+    merge_dicts,
+    parse_count,
+    parse_qs,
+    T,
+    traverse_obj,
     unified_strdate,
     url_or_none,
+    urljoin,
 )
 
 
 class YouPornIE(InfoExtractor):
-    _VALID_URL = r'https?://(?:www\.)?youporn\.com/(?:watch|embed)/(?P<id>\d+)(?:/(?P<display_id>[^/?#&]+))?'
+    _VALID_URL = (
+        r'youporn:(?P<id>\d+)',
+        r'''(?x)
+            https?://(?:www\.)?youporn\.com/(?:watch|embed)/(?P<id>\d+)
+            (?:/(?:(?P<display_id>[^/?#&]+)/?)?)?(?:[#?]|$)
+    '''
+    )
+    _EMBED_REGEX = [r'<iframe[^>]+\bsrc=["\'](?P<url>(?:https?:)?//(?:www\.)?youporn\.com/embed/\d+)']
     _TESTS = [{
         'url': 'http://www.youporn.com/watch/505835/sex-ed-is-it-safe-to-masturbate-daily/',
         'md5': '3744d24c50438cf5b6f6d59feb5055c2',
@@ -34,7 +52,7 @@ class YouPornIE(InfoExtractor):
             'tags': list,
             'age_limit': 18,
         },
-        'skip': 'This video has been disabled',
+        'skip': 'This video has been deactivated',
     }, {
         # Unknown uploader
         'url': 'http://www.youporn.com/watch/561726/big-tits-awesome-brunette-on-amazing-webcam-show/?from=related3&al=2&from_id=561726&pos=4',
@@ -66,57 +84,104 @@ class YouPornIE(InfoExtractor):
     }, {
         'url': 'https://www.youporn.com/watch/13922959/femdom-principal/',
         'only_matching': True,
+    }, {
+        'url': 'https://www.youporn.com/watch/16290308/tinderspecial-trailer1/',
+        'info_dict': {
+            'id': '16290308',
+            'age_limit': 18,
+            'categories': [],
+            'description': None,  # SEO spam using title removed
+            'display_id': 'tinderspecial-trailer1',
+            'duration': 298.0,
+            'ext': 'mp4',
+            'upload_date': '20201123',
+            'uploader': 'Ersties',
+            'tags': [],
+            'thumbnail': 'https://fi1.ypncdn.com/m=eaSaaTbWx/202011/23/16290308/original/3.jpg',
+            'timestamp': 1606147564,
+            'title': 'Tinder In Real Life',
+            'view_count': int,
+        }
     }]
 
-    @staticmethod
-    def _extract_urls(webpage):
-        return re.findall(
-            r'<iframe[^>]+\bsrc=["\']((?:https?:)?//(?:www\.)?youporn\.com/embed/\d+)',
-            webpage)
+    @classmethod
+    def _extract_urls(cls, webpage):
+        def yield_urls():
+            for p in cls._EMBED_REGEX:
+                for m in re.finditer(p, webpage):
+                    yield m.group('url')
+
+        return list(yield_urls())
 
     def _real_extract(self, url):
-        mobj = re.match(self._VALID_URL, url)
-        video_id = mobj.group('id')
-        display_id = mobj.group('display_id') or video_id
+        # A different video ID (data-video-id) is hidden in the page but
+        # never seems to be used
+        video_id, display_id = self._match_valid_url(url).group('id', 'display_id')
+        url = 'http://www.youporn.com/watch/%s' % (video_id,)
+        webpage = self._download_webpage(
+            url, video_id, headers={'Cookie': 'age_verified=1'})
 
-        definitions = self._download_json(
-            'https://www.youporn.com/api/video/media_definitions/%s/' % video_id,
-            display_id)
+        watchable = self._search_regex(
+            r'''(<div\s[^>]*\bid\s*=\s*('|")?watch-container(?(2)\2|(?!-)\b)[^>]*>)''',
+            webpage, 'watchability', default=None)
+        if not watchable:
+            msg = re.split(r'\s{4}', clean_html(get_element_by_id(
+                'mainContent', webpage)) or '')[0]
+            raise ExtractorError(
+                ('%s says: %s' % (self.IE_NAME, msg))
+                if msg else 'Video unavailable: no reason found',
+                expected=True)
+        # internal ID ?
+        # video_id = extract_attributes(watchable).get('data-video-id')
+
+        playervars = self._search_json(
+            r'\bplayervars\s*:', webpage, 'playervars', video_id)
+
+        def get_fmt(x):
+            v_url = url_or_none(x.get('videoUrl'))
+            if v_url:
+                x['videoUrl'] = v_url
+                return (x['format'], x)
+
+        defs_by_format = dict(traverse_obj(playervars, (
+            'mediaDefinitions', lambda _, v: v.get('format'), T(get_fmt))))
+
+        def get_format_data(f):
+            if f not in defs_by_format:
+                return []
+            return self._download_json(
+                defs_by_format[f]['videoUrl'], video_id, '{0}-formats'.format(f))
 
         formats = []
-        for definition in definitions:
-            if not isinstance(definition, dict):
-                continue
-            video_url = url_or_none(definition.get('videoUrl'))
-            if not video_url:
-                continue
-            f = {
-                'url': video_url,
-                'filesize': int_or_none(definition.get('videoSize')),
-            }
-            height = int_or_none(definition.get('quality'))
+        # Try to extract only the actual master m3u8 first, avoiding the duplicate single resolution "master" m3u8s
+        for hls_url in traverse_obj(
+                get_format_data('hls'),
+                (lambda _, v: not isinstance(v['defaultQuality'], bool), 'videoUrl'),
+                (Ellipsis, 'videoUrl')):
+            formats.extend(self._extract_m3u8_formats(
+                hls_url, video_id, 'mp4', fatal=False, m3u8_id='hls',
+                entry_protocol='m3u8_native'))
+
+        for f in traverse_obj(get_format_data('mp4'), (
+                lambda _, v: v.get('videoUrl'), {
+                    'url': ('videoUrl', T(url_or_none)),
+                    'filesize': ('videoSize', T(int_or_none)),
+                    'height': ('quality', T(int_or_none)),
+                }, T(lambda x: x.get('videoUrl') and x))):
             # Video URL's path looks like this:
             #  /201012/17/505835/720p_1500k_505835/YouPorn%20-%20Sex%20Ed%20Is%20It%20Safe%20To%20Masturbate%20Daily.mp4
             #  /201012/17/505835/vl_240p_240k_505835/YouPorn%20-%20Sex%20Ed%20Is%20It%20Safe%20To%20Masturbate%20Daily.mp4
             #  /videos/201703/11/109285532/1080P_4000K_109285532.mp4
             # We will benefit from it by extracting some metadata
-            mobj = re.search(r'(?P<height>\d{3,4})[pP]_(?P<bitrate>\d+)[kK]_\d+', video_url)
+            mobj = re.search(r'(?P<height>\d{3,4})[pP]_(?P<bitrate>\d+)[kK]_\d+', f['videoUrl'])
             if mobj:
-                if not height:
-                    height = int(mobj.group('height'))
-                bitrate = int(mobj.group('bitrate'))
-                f.update({
-                    'format_id': '%dp-%dk' % (height, bitrate),
-                    'tbr': bitrate,
-                })
-            f['height'] = height
+                if not f.get('height'):
+                    f['height'] = int(mobj.group('height'))
+                f['tbr'] = int(mobj.group('bitrate'))
+                f['format_id'] = '%dp-%dk' % (f['height'], f['tbr'])
             formats.append(f)
         self._sort_formats(formats)
 
-        webpage = self._download_webpage(
-            'http://www.youporn.com/watch/%s' % video_id, display_id,
-            headers={'Cookie': 'age_verified=1'})
-
         title = self._html_search_regex(
             r'(?s)<div[^>]+class=["\']watchVideoTitle[^>]+>(.+?)</div>',
             webpage, 'title', default=None) or self._og_search_title(
@@ -131,8 +196,10 @@ class YouPornIE(InfoExtractor):
         thumbnail = self._search_regex(
             r'(?:imageurl\s*=|poster\s*:)\s*(["\'])(?P<thumbnail>.+?)\1',
             webpage, 'thumbnail', fatal=False, group='thumbnail')
-        duration = int_or_none(self._html_search_meta(
-            'video:duration', webpage, 'duration', fatal=False))
+        duration = traverse_obj(playervars, ('duration', T(int_or_none)))
+        if duration is None:
+            duration = int_or_none(self._html_search_meta(
+                'video:duration', webpage, 'duration', fatal=False))
 
         uploader = self._html_search_regex(
             r'(?s)<div[^>]+class=["\']submitByLink["\'][^>]*>(.+?)</div>',
@@ -148,11 +215,11 @@ class YouPornIE(InfoExtractor):
 
         view_count = None
         views = self._search_regex(
-            r'(<div[^>]+\bclass=["\']js_videoInfoViews["\']>)', webpage,
-            'views', default=None)
+            r'(<div\s[^>]*\bdata-value\s*=[^>]+>)\s*<label>Views:</label>',
+            webpage, 'views', default=None)
         if views:
-            view_count = str_to_int(extract_attributes(views).get('data-value'))
-        comment_count = str_to_int(self._search_regex(
+            view_count = parse_count(extract_attributes(views).get('data-value'))
+        comment_count = parse_count(self._search_regex(
             r'>All [Cc]omments? \(([\d,.]+)\)',
             webpage, 'comment count', default=None))
 
@@ -168,7 +235,10 @@ class YouPornIE(InfoExtractor):
             r'(?s)Tags:.*?</div>\s*<div[^>]+class=["\']tagBoxContent["\'][^>]*>(.+?)</div>',
             'tags')
 
-        return {
+        data = self._search_json_ld(webpage, video_id, expected_type='VideoObject', fatal=False) or {}
+        data.pop('url', None)
+
+        result = merge_dicts(data, {
             'id': video_id,
             'display_id': display_id,
             'title': title,
@@ -183,4 +253,442 @@ class YouPornIE(InfoExtractor):
             'tags': tags,
             'age_limit': age_limit,
             'formats': formats,
-        }
+        })
+        # Remove promotional non-description
+        if result.get('description', '').startswith(
+                'Watch %s online' % (result['title'],)):
+            del result['description']
+        return result
+
+
+class YouPornListBase(InfoExtractor):
+    # pattern in '.title-text' element of page section containing videos
+    _PLAYLIST_TITLEBAR_RE = r'\s+[Vv]ideos\s*$'
+    _PAGE_RETRY_COUNT = 0  # ie, no retry
+    _PAGE_RETRY_DELAY = 2  # seconds
+
+    def _get_next_url(self, url, pl_id, html):
+        return urljoin(url, self._search_regex(
+            r'''<a\s[^>]*?\bhref\s*=\s*("|')(?P<url>(?:(?!\1)[^>])+)\1''',
+            get_element_by_id('next', html) or '', 'next page',
+            group='url', default=None))
+
+    @classmethod
+    def _get_title_from_slug(cls, title_slug):
+        return re.sub(r'[_-]', ' ', title_slug)
+
+    def _entries(self, url, pl_id, html=None, page_num=None):
+
+        # separates page sections
+        PLAYLIST_SECTION_RE = (
+            r'''<div\s[^>]*\bclass\s*=\s*('|")(?:[\w$-]+\s+|\s)*?title-bar(?:\s+[\w$-]+|\s)*\1[^>]*>'''
+        )
+        # contains video link
+        VIDEO_URL_RE = r'''(?x)
+            <div\s[^>]*\bdata-video-id\s*=\s*('|")\d+\1[^>]*>\s*
+            (?:<div\b[\s\S]+?</div>\s*)*
+            <a\s[^>]*\bhref\s*=\s*('|")(?P<url>(?:(?!\2)[^>])+)\2
+        '''
+
+        def yield_pages(url, html=html, page_num=page_num):
+            fatal = not html
+            for pnum in itertools.count(start=page_num or 1):
+                if not html:
+                    html = self._download_webpage(
+                        url, pl_id, note='Downloading page %d' % pnum,
+                        fatal=fatal)
+                if not html:
+                    break
+                fatal = False
+                yield (url, html, pnum)
+                # explicit page: extract just that page
+                if page_num is not None:
+                    break
+                next_url = self._get_next_url(url, pl_id, html)
+                if not next_url or next_url == url:
+                    break
+                url, html = next_url, None
+
+        def retry_page(msg, tries_left, page_data):
+            if tries_left <= 0:
+                return
+            self.report_warning(msg, pl_id)
+            sleep(self._PAGE_RETRY_DELAY)
+            return next(
+                yield_pages(page_data[0], page_num=page_data[2]), None)
+
+        def yield_entries(html):
+            for frag in re.split(PLAYLIST_SECTION_RE, html):
+                if not frag:
+                    continue
+                t_text = get_element_by_class('title-text', frag or '')
+                if not (t_text and re.search(self._PLAYLIST_TITLEBAR_RE, t_text)):
+                    continue
+                for m in re.finditer(VIDEO_URL_RE, frag):
+                    video_url = urljoin(url, m.group('url'))
+                    if video_url:
+                        yield self.url_result(video_url)
+
+        last_first_url = None
+        for page_data in yield_pages(url, html=html, page_num=page_num):
+            # page_data: url, html, page_num
+            first_url = None
+            tries_left = self._PAGE_RETRY_COUNT + 1
+            while tries_left > 0:
+                tries_left -= 1
+                for from_ in yield_entries(page_data[1]):
+                    # may get the same page twice instead of empty page
+                    # or (site bug) intead of actual next page
+                    if not first_url:
+                        first_url = from_['url']
+                        if first_url == last_first_url:
+                            # sometimes (/porntags/) the site serves the previous page
+                            # instead but may provide the correct page after a delay
+                            page_data = retry_page(
+                                'Retrying duplicate page...', tries_left, page_data)
+                            if page_data:
+                                first_url = None
+                                break
+                            continue
+                    yield from_
+                else:
+                    if not first_url and 'no-result-paragarph1' in page_data[1]:
+                        page_data = retry_page(
+                            'Retrying empty page...', tries_left, page_data)
+                        if page_data:
+                            continue
+                    else:
+                        # success/failure
+                        break
+            # may get an infinite (?) sequence of empty pages
+            if not first_url:
+                break
+            last_first_url = first_url
+
+    def _real_extract(self, url, html=None):
+        # exceptionally, id may be None
+        m_dict = self._match_valid_url(url).groupdict()
+        pl_id, page_type, sort = (m_dict.get(k) for k in ('id', 'type', 'sort'))
+
+        qs = parse_qs(url)
+        for q, v in qs.items():
+            if v:
+                qs[q] = v[-1]
+            else:
+                del qs[q]
+
+        base_id = pl_id or 'YouPorn'
+        title = self._get_title_from_slug(base_id)
+        if page_type:
+            title = '%s %s' % (page_type.capitalize(), title)
+        base_id = [base_id.lower()]
+        if sort is None:
+            title += ' videos'
+        else:
+            title = '%s videos by %s' % (title, re.sub(r'[_-]', ' ', sort))
+            base_id.append(sort)
+        if qs:
+            ps = ['%s=%s' % item for item in sorted(qs.items())]
+            title += ' (%s)' % ','.join(ps)
+            base_id.extend(ps)
+        pl_id = '/'.join(base_id)
+
+        return self.playlist_result(
+            self._entries(url, pl_id, html=html,
+                          page_num=int_or_none(qs.get('page'))),
+            playlist_id=pl_id, playlist_title=title)
+
+
+class YouPornCategoryIE(YouPornListBase):
+    IE_DESC = 'YouPorn category, with sorting, filtering and pagination'
+    _VALID_URL = r'''(?x)
+        https?://(?:www\.)?youporn\.com/
+        (?P<type>category)/(?P<id>[^/?#&]+)
+        (?:/(?P<sort>popular|views|rating|time|duration))?/?(?:[#?]|$)
+    '''
+    _TESTS = [{
+        'note': 'Full list with pagination',
+        'url': 'https://www.youporn.com/category/lingerie/popular/',
+        'info_dict': {
+            'id': 'lingerie/popular',
+            'title': 'Category lingerie videos by popular',
+        },
+        'playlist_mincount': 39,
+    }, {
+        'note': 'Filtered paginated list with single page result',
+        'url': 'https://www.youporn.com/category/lingerie/duration/?min_minutes=10',
+        'info_dict': {
+            'id': 'lingerie/duration/min_minutes=10',
+            'title': 'Category lingerie videos by duration (min_minutes=10)',
+        },
+        'playlist_maxcount': 30,
+    }, {
+        'note': 'Single page of full list',
+        'url': 'https://www.youporn.com/category/lingerie/popular?page=1',
+        'info_dict': {
+            'id': 'lingerie/popular/page=1',
+            'title': 'Category lingerie videos by popular (page=1)',
+        },
+        'playlist_count': 30,
+    }]
+
+
+class YouPornChannelIE(YouPornListBase):
+    IE_DESC = 'YouPorn channel, with sorting and pagination'
+    _VALID_URL = r'''(?x)
+        https?://(?:www\.)?youporn\.com/
+        (?P<type>channel)/(?P<id>[^/?#&]+)
+        (?:/(?P<sort>rating|views|duration))?/?(?:[#?]|$)
+    '''
+    _TESTS = [{
+        'note': 'Full list with pagination',
+        'url': 'https://www.youporn.com/channel/x-feeds/',
+        'info_dict': {
+            'id': 'x-feeds',
+            'title': 'Channel X-Feeds videos',
+        },
+        'playlist_mincount': 37,
+    }, {
+        'note': 'Single page of full list (no filters here)',
+        'url': 'https://www.youporn.com/channel/x-feeds/duration?page=1',
+        'info_dict': {
+            'id': 'x-feeds/duration/page=1',
+            'title': 'Channel X-Feeds videos by duration (page=1)',
+        },
+        'playlist_count': 24,
+    }]
+
+    @staticmethod
+    def _get_title_from_slug(title_slug):
+        return re.sub(r'_', ' ', title_slug).title()
+
+
+class YouPornCollectionIE(YouPornListBase):
+    IE_DESC = 'YouPorn collection (user playlist), with sorting and pagination'
+    _VALID_URL = r'''(?x)
+        https?://(?:www\.)?youporn\.com/
+        (?P<type>collection)s/videos/(?P<id>\d+)
+        (?:/(?P<sort>rating|views|time|duration))?/?(?:[#?]|$)
+    '''
+    _PLAYLIST_TITLEBAR_RE = r'^\s*Videos\s+in\s'
+    _TESTS = [{
+        'note': 'Full list with pagination',
+        'url': 'https://www.youporn.com/collections/videos/33044251/',
+        'info_dict': {
+            'id': '33044251',
+            'title': 'Collection Sexy Lips videos',
+            'uploader': 'ph-littlewillyb',
+        },
+        'playlist_mincount': 50,
+    }, {
+        'note': 'Single page of full list (no filters here)',
+        'url': 'https://www.youporn.com/collections/videos/33044251/time?page=1',
+        'info_dict': {
+            'id': '33044251/time/page=1',
+            'title': 'Collection Sexy Lips videos by time (page=1)',
+            'uploader': 'ph-littlewillyb',
+        },
+        'playlist_count': 20,
+    }]
+
+    def _real_extract(self, url):
+        pl_id = self._match_id(url)
+        html = self._download_webpage(url, pl_id)
+        playlist = super(YouPornCollectionIE, self)._real_extract(url, html=html)
+        infos = re.sub(r'\s+', ' ', clean_html(get_element_by_class(
+            'collection-infos', html)) or '')
+        title, uploader = self._search_regex(
+            r'^\s*Collection: (?P<title>.+?) \d+ VIDEOS \d+ VIEWS \d+ days LAST UPDATED From: (?P<uploader>[\w_-]+)',
+            infos, 'title/uploader', group=('title', 'uploader'), default=(None, None))
+
+        return merge_dicts({
+            'title': playlist['title'].replace(playlist['id'].split('/')[0], title),
+            'uploader': uploader,
+        }, playlist) if title else playlist
+
+
+class YouPornTagIE(YouPornListBase):
+    IE_DESC = 'YouPorn tag (porntags), with sorting, filtering and pagination'
+    _VALID_URL = r'''(?x)
+        https?://(?:www\.)?youporn\.com/
+        porn(?P<type>tag)s/(?P<id>[^/?#&]+)
+        (?:/(?P<sort>views|rating|time|duration))?/?(?:[#?]|$)
+    '''
+    _PLAYLIST_TITLEBAR_RE = r'^\s*Videos\s+tagged\s'
+    _PAGE_RETRY_COUNT = 1
+    _TESTS = [{
+        'note': 'Full list with pagination',
+        'url': 'https://www.youporn.com/porntags/austrian',
+        'info_dict': {
+            'id': 'austrian',
+            'title': 'Tag austrian videos',
+        },
+        'playlist_mincount': 35,
+        'expected_warnings': ['Retrying duplicate page'],
+    }, {
+        'note': 'Filtered paginated list with single page result',
+        'url': 'https://www.youporn.com/porntags/austrian/duration/?min_minutes=10',
+        'info_dict': {
+            'id': 'austrian/duration/min_minutes=10',
+            'title': 'Tag austrian videos by duration (min_minutes=10)',
+        },
+        # number of videos per page is (row x col) 2x3 + 6x4 + 2, or + 3,
+        # or more, varying with number of ads; let's set max as 9x4
+        # NB col 1 may not be shown in non-JS page with site CSS and zoom 100%
+        'playlist_maxcount': 32,
+        'expected_warnings': ['Retrying duplicate page', 'Retrying empty page'],
+    }, {
+        'note': 'Single page of full list',
+        'url': 'https://www.youporn.com/porntags/austrian/?page=1',
+        'info_dict': {
+            'id': 'austrian/page=1',
+            'title': 'Tag austrian videos (page=1)',
+        },
+        'playlist_mincount': 32,
+        'playlist_maxcount': 34,
+        'expected_warnings': ['Retrying duplicate page', 'Retrying empty page'],
+    }]
+
+    # YP tag navigation is broken, loses sort
+    def _get_next_url(self, url, pl_id, html):
+        next_url = super(YouPornTagIE, self)._get_next_url(url, pl_id, html)
+        if next_url:
+            n = self._match_valid_url(next_url)
+            if n:
+                s = n.groupdict().get('sort')
+            if s:
+                u = self._match_valid_url(url)
+                if u:
+                    u = u.groupdict().get('sort')
+                    if s and not u:
+                        n = n.end('sort')
+                        next_url = next_url[:n] + '/' + u + next_url[n:]
+        return next_url
+
+
+class YouPornStarIE(YouPornListBase):
+    IE_DESC = 'YouPorn Pornstar, with description, sorting and pagination'
+    _VALID_URL = r'''(?x)
+        https?://(?:www\.)?youporn\.com/
+        (?P<type>pornstar)/(?P<id>[^/?#&]+)
+        (?:/(?P<sort>rating|views|duration))?/?(?:[#?]|$)
+    '''
+    _PLAYLIST_TITLEBAR_RE = r'^\s*Videos\s+[fF]eaturing\s'
+    _TESTS = [{
+        'note': 'Full list with pagination',
+        'url': 'https://www.youporn.com/pornstar/daynia/',
+        'info_dict': {
+            'id': 'daynia',
+            'title': 'Pornstar Daynia videos',
+            'description': r're:Daynia Rank \d+ Videos \d+ Views [\d,.]+ .+ Subscribers \d+',
+        },
+        'playlist_mincount': 45,
+    }, {
+        'note': 'Single page of full list (no filters here)',
+        'url': 'https://www.youporn.com/pornstar/daynia/?page=1',
+        'info_dict': {
+            'id': 'daynia/page=1',
+            'title': 'Pornstar Daynia videos (page=1)',
+            'description': 're:.{180,}',
+        },
+        'playlist_count': 26,
+    }]
+
+    @staticmethod
+    def _get_title_from_slug(title_slug):
+        return re.sub(r'_', ' ', title_slug).title()
+
+    def _real_extract(self, url):
+        pl_id = self._match_id(url)
+        html = self._download_webpage(url, pl_id)
+        playlist = super(YouPornStarIE, self)._real_extract(url, html=html)
+        INFO_ELEMENT_RE = r'''(?x)
+            <div\s[^>]*\bclass\s*=\s*('|")(?:[\w$-]+\s+|\s)*?pornstar-info-wrapper(?:\s+[\w$-]+|\s)*\1[^>]*>
+            (?P<info>[\s\S]+?)(?:</div>\s*){6,}
+        '''
+
+        infos = self._search_regex(INFO_ELEMENT_RE, html, 'infos', group='info', default='')
+        if infos:
+            infos = re.sub(
+                r'(?:\s*nl=nl)+\s*', ' ',
+                re.sub(r'(?u)\s+', ' ', clean_html(
+                    re.sub('\n', 'nl=nl', infos)))).replace('ribe Subsc', '')
+
+        return merge_dicts({
+            'description': infos.strip() or None,
+        }, playlist)
+
+
+class YouPornVideosIE(YouPornListBase):
+    IE_DESC = 'YouPorn video (browse) playlists, with sorting, filtering and pagination'
+    _VALID_URL = r'''(?x)
+        https?://(?:www\.)?youporn\.com/
+            (?:(?P<id>browse)/)?
+            (?P<sort>(?(id)
+                (?:duration|rating|time|views)|
+                (?:most_(?:favou?rit|view)ed|recommended|top_rated)?))
+            (?:[/#?]|$)
+    '''
+    _PLAYLIST_TITLEBAR_RE = r'\s+(?:[Vv]ideos|VIDEOS)\s*$'
+    _TESTS = [{
+        'note': 'Full list with pagination (too long for test)',
+        'url': 'https://www.youporn.com/',
+        'info_dict': {
+            'id': 'youporn',
+            'title': 'YouPorn videos',
+        },
+        'only_matching': True,
+    }, {
+        'note': 'Full list with pagination (too long for test)',
+        'url': 'https://www.youporn.com/recommended',
+        'info_dict': {
+            'id': 'youporn/recommended',
+            'title': 'YouPorn videos by recommended',
+        },
+        'only_matching': True,
+    }, {
+        'note': 'Full list with pagination (too long for test)',
+        'url': 'https://www.youporn.com/top_rated',
+        'info_dict': {
+            'id': 'youporn/top_rated',
+            'title': 'YouPorn videos by top rated',
+        },
+        'only_matching': True,
+    }, {
+        'note': 'Full list with pagination (too long for test)',
+        'url': 'https://www.youporn.com/browse/time',
+        'info_dict': {
+            'id': 'browse/time',
+            'title': 'YouPorn videos by time',
+        },
+        'only_matching': True,
+    }, {
+        'note': 'Filtered paginated list with single page result',
+        'url': 'https://www.youporn.com/most_favorited/?res=VR&max_minutes=2',
+        'info_dict': {
+            'id': 'youporn/most_favorited/max_minutes=2/res=VR',
+            'title': 'YouPorn videos by most favorited (max_minutes=2,res=VR)',
+        },
+        'playlist_mincount': 10,
+        'playlist_maxcount': 28,
+    }, {
+        'note': 'Filtered paginated list with several pages',
+        'url': 'https://www.youporn.com/most_favorited/?res=VR&max_minutes=5',
+        'info_dict': {
+            'id': 'youporn/most_favorited/max_minutes=5/res=VR',
+            'title': 'YouPorn videos by most favorited (max_minutes=5,res=VR)',
+        },
+        'playlist_mincount': 45,
+    }, {
+        'note': 'Single page of full list',
+        'url': 'https://www.youporn.com/browse/time?page=1',
+        'info_dict': {
+            'id': 'browse/time/page=1',
+            'title': 'YouPorn videos by time (page=1)',
+        },
+        'playlist_count': 36,
+    }]
+
+    @staticmethod
+    def _get_title_from_slug(title_slug):
+        return 'YouPorn' if title_slug == 'browse' else title_slug

youtube_dl/extractor/youtube.py

@@ -1636,7 +1636,7 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
         try:
             jsi, player_id, func_code = self._extract_n_function_code(video_id, player_url)
         except ExtractorError as e:
-            raise ExtractorError('Unable to extract nsig jsi, player_id, func_codefunction code', cause=e)
+            raise ExtractorError('Unable to extract nsig function code', cause=e)
         if self.get_param('youtube_print_sig_code'):
             self.to_screen('Extracted nsig function from {0}:\n{1}\n'.format(
                 player_id, func_code[1]))
@@ -1647,10 +1647,10 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
         except JSInterpreter.Exception as e:
             self.report_warning(
                 '%s (%s %s)' % (
-                    self.__ie_msg(
-                        'Unable to decode n-parameter: download likely to be throttled'),
+                    'Unable to decode n-parameter: expect download to be blocked or throttled',
                     error_to_compat_str(e),
-                    traceback.format_exc()))
+                    traceback.format_exc()),
+                video_id=video_id)
             return
 
         self.write_debug('Decrypted nsig {0} => {1}'.format(n, ret))
@@ -1658,13 +1658,52 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
 
     def _extract_n_function_name(self, jscode):
         func_name, idx = self._search_regex(
-            r'\.get\("n"\)\)&&\(b=(?P<nfunc>[a-zA-Z_$][\w$]*)(?:\[(?P<idx>\d+)\])?\([\w$]+\)',
-            jscode, 'Initial JS player n function name', group=('nfunc', 'idx'))
+            # new: (b=String.fromCharCode(110),c=a.get(b))&&c=nfunc[idx](c)
+            # or:  (b="nn"[+a.D],c=a.get(b))&&(c=nfunc[idx](c)
+            # or:  (PL(a),b=a.j.n||null)&&(b=nfunc[idx](b)
+            # or:  (b="nn"[+a.D],vL(a),c=a.j[b]||null)&&(c=narray[idx](c),a.set(b,c),narray.length||nfunc("")
+            # old: (b=a.get("n"))&&(b=nfunc[idx](b)(?P<c>[a-z])\s*=\s*[a-z]\s*
+            # older: (b=a.get("n"))&&(b=nfunc(b)
+            r'''(?x)
+                \((?:[\w$()\s]+,)*?\s*      # (
+                (?P<b>[a-z])\s*=\s*         # b=
+                (?:
+                    (?:                     # expect ,c=a.get(b) (etc)
+                        String\s*\.\s*fromCharCode\s*\(\s*110\s*\)|
+                        "n+"\[\s*\+?s*[\w$.]+\s*]
+                    )\s*(?:,[\w$()\s]+(?=,))*|
+                       (?P<old>[\w$]+)      # a (old[er])
+                   )\s*
+                   (?(old)
+                                            # b.get("n")
+                       (?:\.\s*[\w$]+\s*|\[\s*[\w$]+\s*]\s*)*?
+                       (?:\.\s*n|\[\s*"n"\s*]|\.\s*get\s*\(\s*"n"\s*\))
+                       |                    # ,c=a.get(b)
+                       ,\s*(?P<c>[a-z])\s*=\s*[a-z]\s*
+                       (?:\.\s*[\w$]+\s*|\[\s*[\w$]+\s*]\s*)*?
+                       (?:\[\s*(?P=b)\s*]|\.\s*get\s*\(\s*(?P=b)\s*\))
+                   )
+                                            # interstitial junk
+                   \s*(?:\|\|\s*null\s*)?(?:\)\s*)?&&\s*(?:\(\s*)?
+               (?(c)(?P=c)|(?P=b))\s*=\s*   # [c|b]=
+                                            # nfunc|nfunc[idx]
+                   (?P<nfunc>[a-zA-Z_$][\w$]*)(?:\s*\[(?P<idx>\d+)\])?\s*\(\s*[\w$]+\s*\)
+            ''', jscode, 'Initial JS player n function name', group=('nfunc', 'idx'),
+            default=(None, None))
+        # thx bashonly: yt-dlp/yt-dlp/pull/10611
+        if not func_name:
+            self.report_warning('Falling back to generic n function search')
+            return self._search_regex(
+                r'''(?xs)
+                    (?:(?<=[^\w$])|^)       # instead of \b, which ignores $
+                    (?P<name>(?!\d)[a-zA-Z\d_$]+)\s*=\s*function\((?!\d)[a-zA-Z\d_$]+\)
+                    \s*\{(?:(?!};).)+?["']enhanced_except_
+                ''', jscode, 'Initial JS player n function name', group='name')
         if not idx:
             return func_name
 
         return self._parse_json(self._search_regex(
-            r'var {0}\s*=\s*(\[.+?\])\s*[,;]'.format(re.escape(func_name)), jscode,
+            r'var\s+{0}\s*=\s*(\[.+?\])\s*[,;]'.format(re.escape(func_name)), jscode,
             'Initial JS player n function list ({0}.{1})'.format(func_name, idx)),
             func_name, transform_source=js_to_json)[int(idx)]
 
@@ -1679,17 +1718,7 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
 
         func_name = self._extract_n_function_name(jscode)
 
-        # For redundancy
-        func_code = self._search_regex(
-            r'''(?xs)%s\s*=\s*function\s*\((?P<var>[\w$]+)\)\s*
-                     # NB: The end of the regex is intentionally kept strict
-                     {(?P<code>.+?}\s*return\ [\w$]+.join\(""\))};''' % func_name,
-            jscode, 'nsig function', group=('var', 'code'), default=None)
-        if func_code:
-            func_code = ([func_code[0]], func_code[1])
-        else:
-            self.write_debug('Extracting nsig function with jsinterp')
-            func_code = jsi.extract_function_code(func_name)
+        func_code = jsi.extract_function_code(func_name)
 
         self.cache.store('youtube-nsig', player_id, func_code)
         return jsi, player_id, func_code

youtube_dl/jsinterp.py

@@ -14,12 +14,15 @@ from .utils import (
     remove_quotes,
     unified_timestamp,
     variadic,
+    write_string,
 )
 from .compat import (
     compat_basestring,
     compat_chr,
     compat_collections_chain_map as ChainMap,
+    compat_filter as filter,
     compat_itertools_zip_longest as zip_longest,
+    compat_map as map,
     compat_str,
 )
 
@@ -53,15 +56,16 @@ def wraps_op(op):
 
 # NB In principle NaN cannot be checked by membership.
 # Here all NaN values are actually this one, so _NaN is _NaN,
-# although _NaN != _NaN.
+# although _NaN != _NaN. Ditto Infinity.
 
 _NaN = float('nan')
+_Infinity = float('inf')
 
 
 def _js_bit_op(op):
 
     def zeroise(x):
-        return 0 if x in (None, JS_Undefined, _NaN) else x
+        return 0 if x in (None, JS_Undefined, _NaN, _Infinity) else x
 
     @wraps_op(op)
     def wrapped(a, b):
@@ -84,7 +88,7 @@ def _js_arith_op(op):
 def _js_div(a, b):
     if JS_Undefined in (a, b) or not (a or b):
         return _NaN
-    return operator.truediv(a or 0, b) if b else float('inf')
+    return operator.truediv(a or 0, b) if b else _Infinity
 
 
 def _js_mod(a, b):
@@ -220,6 +224,42 @@ class LocalNameSpace(ChainMap):
         return 'LocalNameSpace%s' % (self.maps, )
 
 
+class Debugger(object):
+    ENABLED = False
+
+    @staticmethod
+    def write(*args, **kwargs):
+        level = kwargs.get('level', 100)
+
+        def truncate_string(s, left, right=0):
+            if s is None or len(s) <= left + right:
+                return s
+            return '...'.join((s[:left - 3], s[-right:] if right else ''))
+
+        write_string('[debug] JS: {0}{1}\n'.format(
+            '  ' * (100 - level),
+            ' '.join(truncate_string(compat_str(x), 50, 50) for x in args)))
+
+    @classmethod
+    def wrap_interpreter(cls, f):
+        def interpret_statement(self, stmt, local_vars, allow_recursion, *args, **kwargs):
+            if cls.ENABLED and stmt.strip():
+                cls.write(stmt, level=allow_recursion)
+            try:
+                ret, should_ret = f(self, stmt, local_vars, allow_recursion, *args, **kwargs)
+            except Exception as e:
+                if cls.ENABLED:
+                    if isinstance(e, ExtractorError):
+                        e = e.orig_msg
+                    cls.write('=> Raises:', e, '<-|', stmt, level=allow_recursion)
+                raise
+            if cls.ENABLED and stmt.strip():
+                if should_ret or repr(ret) != stmt:
+                    cls.write(['->', '=>'][should_ret], repr(ret), '<-|', stmt, level=allow_recursion)
+            return ret, should_ret
+        return interpret_statement
+
+
 class JSInterpreter(object):
     __named_object_counter = 0
 
@@ -307,8 +347,7 @@ class JSInterpreter(object):
     def __op_chars(cls):
         op_chars = set(';,[')
         for op in cls._all_operators():
-            for c in op[0]:
-                op_chars.add(c)
+            op_chars.update(op[0])
         return op_chars
 
     def _named_object(self, namespace, obj):
@@ -326,9 +365,10 @@ class JSInterpreter(object):
         # collections.Counter() is ~10% slower in both 2.7 and 3.9
         counters = dict((k, 0) for k in _MATCHING_PARENS.values())
         start, splits, pos, delim_len = 0, 0, 0, len(delim) - 1
-        in_quote, escaping, skipping = None, False, 0
-        after_op, in_regex_char_group = True, False
-
+        in_quote, escaping, after_op, in_regex_char_group = None, False, True, False
+        skipping = 0
+        if skip_delims:
+            skip_delims = variadic(skip_delims)
         for idx, char in enumerate(expr):
             paren_delta = 0
             if not in_quote:
@@ -355,7 +395,7 @@ class JSInterpreter(object):
                 continue
             elif pos == 0 and skip_delims:
                 here = expr[idx:]
-                for s in variadic(skip_delims):
+                for s in skip_delims:
                     if here.startswith(s) and s:
                         skipping = len(s) - 1
                         break
@@ -376,16 +416,17 @@ class JSInterpreter(object):
         if delim is None:
             delim = expr and _MATCHING_PARENS[expr[0]]
         separated = list(cls._separate(expr, delim, 1))
-
         if len(separated) < 2:
             raise cls.Exception('No terminating paren {delim} in {expr!r:.5500}'.format(**locals()))
         return separated[0][1:].strip(), separated[1].strip()
 
     @staticmethod
-    def _all_operators():
-        return itertools.chain(
-            # Ref: https://developer.mozilla.org/en-US/docs/Web/JavaScript/Reference/Operators/Operator_Precedence
-            _SC_OPERATORS, _LOG_OPERATORS, _COMP_OPERATORS, _OPERATORS)
+    def _all_operators(_cached=[]):
+        if not _cached:
+            _cached.extend(itertools.chain(
+                # Ref: https://developer.mozilla.org/en-US/docs/Web/JavaScript/Reference/Operators/Operator_Precedence
+                _SC_OPERATORS, _LOG_OPERATORS, _COMP_OPERATORS, _OPERATORS))
+        return _cached
 
     def _operator(self, op, left_val, right_expr, expr, local_vars, allow_recursion):
         if op in ('||', '&&'):
@@ -416,7 +457,7 @@ class JSInterpreter(object):
         except Exception as e:
             if allow_undefined:
                 return JS_Undefined
-            raise self.Exception('Cannot get index {idx:.100}'.format(**locals()), expr=repr(obj), cause=e)
+            raise self.Exception('Cannot get index {idx!r:.100}'.format(**locals()), expr=repr(obj), cause=e)
 
     def _dump(self, obj, namespace):
         try:
@@ -438,6 +479,7 @@ class JSInterpreter(object):
     _FINALLY_RE = re.compile(r'finally\s*\{')
     _SWITCH_RE = re.compile(r'switch\s*\(')
 
+    @Debugger.wrap_interpreter
     def interpret_statement(self, stmt, local_vars, allow_recursion=100):
         if allow_recursion < 0:
             raise self.Exception('Recursion limit reached')
@@ -448,6 +490,7 @@ class JSInterpreter(object):
         # fails on (eg) if (...) stmt1; else stmt2;
         sub_statements = list(self._separate(stmt, ';')) or ['']
         expr = stmt = sub_statements.pop().strip()
+
         for sub_stmt in sub_statements:
             ret, should_return = self.interpret_statement(sub_stmt, local_vars, allow_recursion)
             if should_return:
@@ -511,7 +554,6 @@ class JSInterpreter(object):
                 expr = self._dump(inner, local_vars) + outer
 
         if expr.startswith('('):
-
             m = re.match(r'\((?P<d>[a-z])%(?P<e>[a-z])\.length\+(?P=e)\.length\)%(?P=e)\.length', expr)
             if m:
                 # short-cut eval of frequently used `(d%e.length+e.length)%e.length`, worth ~6% on `pytest -k test_nsig`
@@ -588,8 +630,7 @@ class JSInterpreter(object):
                     if m.group('err'):
                         catch_vars[m.group('err')] = err.error if isinstance(err, JS_Throw) else err
                     catch_vars = local_vars.new_child(m=catch_vars)
-                    err = None
-                    pending = self.interpret_statement(sub_expr, catch_vars, allow_recursion)
+                    err, pending = None, self.interpret_statement(sub_expr, catch_vars, allow_recursion)
 
             m = self._FINALLY_RE.match(expr)
             if m:
@@ -693,7 +734,7 @@ class JSInterpreter(object):
                 (?P<op>{_OPERATOR_RE})?
                 =(?!=)(?P<expr>.*)$
             )|(?P<return>
-                (?!if|return|true|false|null|undefined)(?P<name>{_NAME_RE})$
+                (?!if|return|true|false|null|undefined|NaN|Infinity)(?P<name>{_NAME_RE})$
             )|(?P<indexing>
                 (?P<in>{_NAME_RE})\[(?P<idx>.+)\]$
             )|(?P<attribute>
@@ -727,11 +768,12 @@ class JSInterpreter(object):
             raise JS_Break()
         elif expr == 'continue':
             raise JS_Continue()
-
         elif expr == 'undefined':
             return JS_Undefined, should_return
         elif expr == 'NaN':
             return _NaN, should_return
+        elif expr == 'Infinity':
+            return _Infinity, should_return
 
         elif md.get('return'):
             return local_vars[m.group('name')], should_return
@@ -760,18 +802,31 @@ class JSInterpreter(object):
             right_expr = separated.pop()
             # handle operators that are both unary and binary, minimal BODMAS
             if op in ('+', '-'):
+                # simplify/adjust consecutive instances of these operators
                 undone = 0
-                while len(separated) > 1 and not separated[-1].strip():
+                separated = [s.strip() for s in separated]
+                while len(separated) > 1 and not separated[-1]:
                     undone += 1
                     separated.pop()
                 if op == '-' and undone % 2 != 0:
                     right_expr = op + right_expr
-                left_val = separated[-1]
+                elif op == '+':
+                    while len(separated) > 1 and set(separated[-1]) <= self.OP_CHARS:
+                        right_expr = separated.pop() + right_expr
+                    if separated[-1][-1:] in self.OP_CHARS:
+                        right_expr = separated.pop() + right_expr
+                # hanging op at end of left => unary + (strip) or - (push right)
+                left_val = separated[-1] if separated else ''
                 for dm_op in ('*', '%', '/', '**'):
                     bodmas = tuple(self._separate(left_val, dm_op, skip_delims=skip_delim))
                     if len(bodmas) > 1 and not bodmas[-1].strip():
                         expr = op.join(separated) + op + right_expr
-                        right_expr = None
+                        if len(separated) > 1:
+                            separated.pop()
+                            right_expr = op.join((left_val, right_expr))
+                        else:
+                            separated = [op.join((left_val, right_expr))]
+                            right_expr = None
                         break
                 if right_expr is None:
                     continue
@@ -795,12 +850,15 @@ class JSInterpreter(object):
                     memb = member
                     raise self.Exception('{memb} {msg}'.format(**locals()), expr=expr)
 
-            def eval_method():
+            def eval_method(variable, member):
                 if (variable, member) == ('console', 'debug'):
+                    if Debugger.ENABLED:
+                        Debugger.write(self.interpret_expression('[{}]'.format(arg_str), local_vars, allow_recursion))
                     return
                 types = {
                     'String': compat_str,
                     'Math': float,
+                    'Array': list,
                 }
                 obj = local_vars.get(variable)
                 if obj in (JS_Undefined, None):
@@ -826,12 +884,29 @@ class JSInterpreter(object):
                     self.interpret_expression(v, local_vars, allow_recursion)
                     for v in self._separate(arg_str)]
 
-                if obj == compat_str:
+                # Fixup prototype call
+                if isinstance(obj, type):
+                    new_member, rest = member.partition('.')[0::2]
+                    if new_member == 'prototype':
+                        new_member, func_prototype = rest.partition('.')[0::2]
+                        assertion(argvals, 'takes one or more arguments')
+                        assertion(isinstance(argvals[0], obj), 'must bind to type {0}'.format(obj))
+                        if func_prototype == 'call':
+                            obj = argvals.pop(0)
+                        elif func_prototype == 'apply':
+                            assertion(len(argvals) == 2, 'takes two arguments')
+                            obj, argvals = argvals
+                            assertion(isinstance(argvals, list), 'second argument must be a list')
+                        else:
+                            raise self.Exception('Unsupported Function method ' + func_prototype, expr)
+                        member = new_member
+
+                if obj is compat_str:
                     if member == 'fromCharCode':
                         assertion(argvals, 'takes one or more arguments')
                         return ''.join(map(compat_chr, argvals))
                     raise self.Exception('Unsupported string method ' + member, expr=expr)
-                elif obj == float:
+                elif obj is float:
                     if member == 'pow':
                         assertion(len(argvals) == 2, 'takes two arguments')
                         return argvals[0] ** argvals[1]
@@ -850,18 +925,25 @@ class JSInterpreter(object):
                     obj.reverse()
                     return obj
                 elif member == 'slice':
-                    assertion(isinstance(obj, list), 'must be applied on a list')
-                    assertion(len(argvals) == 1, 'takes exactly one argument')
-                    return obj[argvals[0]:]
+                    assertion(isinstance(obj, (list, compat_str)), 'must be applied on a list or string')
+                    # From [1]:
+                    # .slice() - like [:]
+                    # .slice(n) - like [n:] (not [slice(n)]
+                    # .slice(m, n) - like [m:n] or [slice(m, n)]
+                    # [1] https://developer.mozilla.org/en-US/docs/Web/JavaScript/Reference/Global_Objects/Array/slice
+                    assertion(len(argvals) <= 2, 'takes between 0 and 2 arguments')
+                    if len(argvals) < 2:
+                        argvals += (None,)
+                    return obj[slice(*argvals)]
                 elif member == 'splice':
                     assertion(isinstance(obj, list), 'must be applied on a list')
                     assertion(argvals, 'takes one or more arguments')
-                    index, howMany = map(int, (argvals + [len(obj)])[:2])
+                    index, how_many = map(int, (argvals + [len(obj)])[:2])
                     if index < 0:
                         index += len(obj)
                     add_items = argvals[2:]
                     res = []
-                    for i in range(index, min(index + howMany, len(obj))):
+                    for _ in range(index, min(index + how_many, len(obj))):
                         res.append(obj.pop(index))
                     for i, item in enumerate(add_items):
                         obj.insert(index + i, item)
@@ -919,11 +1001,11 @@ class JSInterpreter(object):
 
             if remaining:
                 ret, should_abort = self.interpret_statement(
-                    self._named_object(local_vars, eval_method()) + remaining,
+                    self._named_object(local_vars, eval_method(variable, member)) + remaining,
                     local_vars, allow_recursion)
                 return ret, should_return or should_abort
             else:
-                return eval_method(), should_return
+                return eval_method(variable, member), should_return
 
         elif md.get('function'):
             fname = m.group('fname')
@@ -951,28 +1033,25 @@ class JSInterpreter(object):
     def extract_object(self, objname):
         _FUNC_NAME_RE = r'''(?:[a-zA-Z$0-9]+|"[a-zA-Z$0-9]+"|'[a-zA-Z$0-9]+')'''
         obj = {}
-        fields = None
-        for obj_m in re.finditer(
+        fields = next(filter(None, (
+            obj_m.group('fields') for obj_m in re.finditer(
                 r'''(?xs)
                     {0}\s*\.\s*{1}|{1}\s*=\s*\{{\s*
                         (?P<fields>({2}\s*:\s*function\s*\(.*?\)\s*\{{.*?}}(?:,\s*)?)*)
                     }}\s*;
                 '''.format(_NAME_RE, re.escape(objname), _FUNC_NAME_RE),
-                self.code):
-            fields = obj_m.group('fields')
-            if fields:
-                break
-        else:
+                self.code))), None)
+        if not fields:
             raise self.Exception('Could not find object ' + objname)
         # Currently, it only supports function definitions
-        fields_m = re.finditer(
-            r'''(?x)
-                (?P<key>%s)\s*:\s*function\s*\((?P<args>(?:%s|,)*)\){(?P<code>[^}]+)}
-            ''' % (_FUNC_NAME_RE, _NAME_RE),
-            fields)
-        for f in fields_m:
+        for f in re.finditer(
+                r'''(?x)
+                    (?P<key>%s)\s*:\s*function\s*\((?P<args>(?:%s|,)*)\){(?P<code>[^}]+)}
+                ''' % (_FUNC_NAME_RE, _NAME_RE),
+                fields):
             argnames = self.build_arglist(f.group('args'))
-            obj[remove_quotes(f.group('key'))] = self.build_function(argnames, f.group('code'))
+            name = remove_quotes(f.group('key'))
+            obj[name] = function_with_repr(self.build_function(argnames, f.group('code')), 'F<{0}>'.format(name))
 
         return obj
 
@@ -1007,7 +1086,7 @@ class JSInterpreter(object):
     def extract_function(self, funcname):
         return function_with_repr(
             self.extract_function_from_code(*self.extract_function_code(funcname)),
-            'F<%s>' % (funcname, ))
+            'F<%s>' % (funcname,))
 
     def extract_function_from_code(self, argnames, code, *global_stack):
         local_vars = {}
@@ -1016,7 +1095,7 @@ class JSInterpreter(object):
             if mobj is None:
                 break
             start, body_start = mobj.span()
-            body, remaining = self._separate_at_paren(code[body_start - 1:], '}')
+            body, remaining = self._separate_at_paren(code[body_start - 1:])
             name = self._named_object(local_vars, self.extract_function_from_code(
                 [x.strip() for x in mobj.group('args').split(',')],
                 body, local_vars, *global_stack))
@@ -1044,8 +1123,7 @@ class JSInterpreter(object):
         argnames = tuple(argnames)
 
         def resf(args, kwargs={}, allow_recursion=100):
-            global_stack[0].update(
-                zip_longest(argnames, args, fillvalue=None))
+            global_stack[0].update(zip_longest(argnames, args, fillvalue=None))
             global_stack[0].update(kwargs)
             var_stack = LocalNameSpace(*global_stack)
             ret, should_abort = self.interpret_statement(code.replace('\n', ' '), var_stack, allow_recursion - 1)

youtube_dl/options.py

@@ -533,6 +533,10 @@ def parseOpts(overrideArguments=None):
         '--no-check-certificate',
         action='store_true', dest='no_check_certificate', default=False,
         help='Suppress HTTPS certificate validation')
+    workarounds.add_option(
+        '--no-check-extensions',
+        action='store_true', dest='no_check_extensions', default=False,
+        help='Suppress file extension validation')
     workarounds.add_option(
         '--prefer-insecure',
         '--prefer-unsecure', action='store_true', dest='prefer_insecure',

youtube_dl/postprocessor/ffmpeg.py

@@ -74,8 +74,11 @@ class FFmpegPostProcessor(PostProcessor):
         return FFmpegPostProcessor(downloader)._versions
 
     def _determine_executables(self):
-        programs = ['avprobe', 'avconv', 'ffmpeg', 'ffprobe']
+        # ordered to match prefer_ffmpeg!
+        convs = ['ffmpeg', 'avconv']
+        probes = ['ffprobe', 'avprobe']
         prefer_ffmpeg = True
+        programs = convs + probes
 
         def get_ffmpeg_version(path):
             ver = get_exe_version(path, args=['-version'])
@@ -96,6 +99,7 @@ class FFmpegPostProcessor(PostProcessor):
 
         self._paths = None
         self._versions = None
+        location = None
         if self._downloader:
             prefer_ffmpeg = self._downloader.params.get('prefer_ffmpeg', True)
             location = self._downloader.params.get('ffmpeg_location')
@@ -118,33 +122,21 @@ class FFmpegPostProcessor(PostProcessor):
                     location = os.path.dirname(os.path.abspath(location))
                     if basename in ('ffmpeg', 'ffprobe'):
                         prefer_ffmpeg = True
+        self._paths = dict(
+            (p, p if location is None else os.path.join(location, p))
+            for p in programs)
+        self._versions = dict(
+            x for x in (
+                (p, get_ffmpeg_version(self._paths[p])) for p in programs)
+            if x[1] is not None)
 
-                self._paths = dict(
-                    (p, os.path.join(location, p)) for p in programs)
-                self._versions = dict(
-                    (p, get_ffmpeg_version(self._paths[p])) for p in programs)
-        if self._versions is None:
-            self._versions = dict(
-                (p, get_ffmpeg_version(p)) for p in programs)
-            self._paths = dict((p, p) for p in programs)
-
-        if prefer_ffmpeg is False:
-            prefs = ('avconv', 'ffmpeg')
-        else:
-            prefs = ('ffmpeg', 'avconv')
-        for p in prefs:
-            if self._versions[p]:
-                self.basename = p
-                break
-
-        if prefer_ffmpeg is False:
-            prefs = ('avprobe', 'ffprobe')
-        else:
-            prefs = ('ffprobe', 'avprobe')
-        for p in prefs:
-            if self._versions[p]:
-                self.probe_basename = p
-                break
+        basenames = [None, None]
+        for i, progs in enumerate((convs, probes)):
+            for p in progs[::-1 if prefer_ffmpeg is False else 1]:
+                if self._versions.get(p):
+                    basenames[i] = p
+                    break
+        self.basename, self.probe_basename = basenames
 
     @property
     def available(self):

youtube_dl/traversal.py

@@ -0,0 +1,10 @@
+# coding: utf-8
+
+# TODO: move these utils.fns here and move import to utils
+# flake8: noqa
+from .utils import (
+    dict_get,
+    get_first,
+    T,
+    traverse_obj,
+)

youtube_dl/utils.py

@@ -45,14 +45,18 @@ from .compat import (
     compat_casefold,
     compat_chr,
     compat_collections_abc,
+    compat_contextlib_suppress,
     compat_cookiejar,
     compat_ctypes_WINFUNCTYPE,
     compat_datetime_timedelta_total_seconds,
+    compat_etree_Element,
     compat_etree_fromstring,
+    compat_etree_iterfind,
     compat_expanduser,
     compat_html_entities,
     compat_html_entities_html5,
     compat_http_client,
+    compat_http_cookies,
     compat_integer_types,
     compat_kwargs,
     compat_ncompress as ncompress,
@@ -1713,21 +1717,6 @@ TIMEZONE_NAMES = {
     'PST': -8, 'PDT': -7   # Pacific
 }
 
-KNOWN_EXTENSIONS = (
-    'mp4', 'm4a', 'm4p', 'm4b', 'm4r', 'm4v', 'aac',
-    'flv', 'f4v', 'f4a', 'f4b',
-    'webm', 'ogg', 'ogv', 'oga', 'ogx', 'spx', 'opus',
-    'mkv', 'mka', 'mk3d',
-    'avi', 'divx',
-    'mov',
-    'asf', 'wmv', 'wma',
-    '3gp', '3g2',
-    'mp3',
-    'flac',
-    'ape',
-    'wav',
-    'f4f', 'f4m', 'm3u8', 'smil')
-
 # needed for sanitizing filenames in restricted mode
 ACCENT_CHARS = dict(zip('ÂÃÄÀÁÅÆÇÈÉÊËÌÍÎÏÐÑÒÓÔÕÖŐØŒÙÚÛÜŰÝÞßàáâãäåæçèéêëìíîïðñòóôõöőøœùúûüűýþÿ',
                         itertools.chain('AAAAAA', ['AE'], 'CEEEEIIIIDNOOOOOOO', ['OE'], 'UUUUUY', ['TH', 'ss'],
@@ -1855,25 +1844,18 @@ def write_json_file(obj, fn):
     try:
         with tf:
             json.dump(obj, tf)
-        if sys.platform == 'win32':
-            # Need to remove existing file on Windows, else os.rename raises
-            # WindowsError or FileExistsError.
-            try:
+        with compat_contextlib_suppress(OSError):
+            if sys.platform == 'win32':
+                # Need to remove existing file on Windows, else os.rename raises
+                # WindowsError or FileExistsError.
                 os.unlink(fn)
-            except OSError:
-                pass
-        try:
             mask = os.umask(0)
             os.umask(mask)
             os.chmod(tf.name, 0o666 & ~mask)
-        except OSError:
-            pass
         os.rename(tf.name, fn)
     except Exception:
-        try:
+        with compat_contextlib_suppress(OSError):
             os.remove(tf.name)
-        except OSError:
-            pass
         raise
 
 
@@ -2033,14 +2015,13 @@ def extract_attributes(html_element):
     NB HTMLParser is stricter in Python 2.6 & 3.2 than in later versions,
     but the cases in the unit test will work for all of 2.6, 2.7, 3.2-3.5.
     """
-    parser = HTMLAttributeParser()
-    try:
-        parser.feed(html_element)
-        parser.close()
-    # Older Python may throw HTMLParseError in case of malformed HTML
-    except compat_HTMLParseError:
-        pass
-    return parser.attrs
+    ret = None
+    # Older Python may throw HTMLParseError in case of malformed HTML (and on .close()!)
+    with compat_contextlib_suppress(compat_HTMLParseError):
+        with contextlib.closing(HTMLAttributeParser()) as parser:
+            parser.feed(html_element)
+            ret = parser.attrs
+    return ret or {}
 
 
 def clean_html(html):
@@ -2241,7 +2222,8 @@ def _htmlentity_transform(entity_with_semicolon):
             numstr = '0%s' % numstr
         else:
             base = 10
-        # See https://github.com/ytdl-org/youtube-dl/issues/7518
+        # See https://github.com/ytdl-org/youtube-dl/issues/7518\
+        # Also, weirdly, compat_contextlib_suppress fails here in 2.6
         try:
             return compat_chr(int(numstr, base))
         except ValueError:
@@ -2348,11 +2330,9 @@ def make_HTTPS_handler(params, **kwargs):
         # Some servers may (wrongly) reject requests if ALPN extension is not sent. See:
         # https://github.com/python/cpython/issues/85140
         # https://github.com/yt-dlp/yt-dlp/issues/3878
-        try:
+        with compat_contextlib_suppress(AttributeError, NotImplementedError):
+            # fails for Python < 2.7.10, not ssl.HAS_ALPN
             ctx.set_alpn_protocols(ALPN_PROTOCOLS)
-        except (AttributeError, NotImplementedError):
-            # Python < 2.7.10, not ssl.HAS_ALPN
-            pass
 
     opts_no_check_certificate = params.get('nocheckcertificate', False)
     if hasattr(ssl, 'create_default_context'):  # Python >= 3.4 or 2.7.9
@@ -2362,12 +2342,10 @@ def make_HTTPS_handler(params, **kwargs):
             context.check_hostname = False
             context.verify_mode = ssl.CERT_NONE
 
-        try:
+        with compat_contextlib_suppress(TypeError):
+            # Fails with Python 2.7.8 (create_default_context present
+            # but HTTPSHandler has no context=)
             return YoutubeDLHTTPSHandler(params, context=context, **kwargs)
-        except TypeError:
-            # Python 2.7.8
-            # (create_default_context present but HTTPSHandler has no context=)
-            pass
 
     if sys.version_info < (3, 2):
         return YoutubeDLHTTPSHandler(params, **kwargs)
@@ -2381,15 +2359,24 @@ def make_HTTPS_handler(params, **kwargs):
         return YoutubeDLHTTPSHandler(params, context=context, **kwargs)
 
 
-def bug_reports_message():
+def bug_reports_message(before=';'):
     if ytdl_is_updateable():
         update_cmd = 'type  youtube-dl -U  to update'
     else:
-        update_cmd = 'see  https://yt-dl.org/update  on how to update'
-    msg = '; please report this issue on https://yt-dl.org/bug .'
-    msg += ' Make sure you are using the latest version; %s.' % update_cmd
-    msg += ' Be sure to call youtube-dl with the --verbose flag and include its complete output.'
-    return msg
+        update_cmd = 'see  https://github.com/ytdl-org/youtube-dl/#user-content-installation  on how to update'
+
+    msg = (
+        'please report this issue on https://github.com/ytdl-org/youtube-dl/issues ,'
+        ' using the appropriate issue template.'
+        ' Make sure you are using the latest version; %s.'
+        ' Be sure to call youtube-dl with the --verbose option and include the complete output.'
+    ) % update_cmd
+
+    before = (before or '').rstrip()
+    if not before or before.endswith(('.', '!', '?')):
+        msg = msg[0].title() + msg[1:]
+
+    return (before + ' ' if before else '') + msg
 
 
 class YoutubeDLError(Exception):
@@ -2404,7 +2391,7 @@ class ExtractorError(YoutubeDLError):
         """ tb, if given, is the original traceback (so that it can be printed out).
         If expected is set, this is a normal error message and most likely not a bug in youtube-dl.
         """
-
+        self.orig_msg = msg
         if sys.exc_info()[0] in (compat_urllib_error.URLError, socket.timeout, UnavailableVideoError):
             expected = True
         if video_id is not None:
@@ -3176,12 +3163,10 @@ def parse_iso8601(date_str, delimiter='T', timezone=None):
     if timezone is None:
         timezone, date_str = extract_timezone(date_str)
 
-    try:
+    with compat_contextlib_suppress(ValueError):
         date_format = '%Y-%m-%d{0}%H:%M:%S'.format(delimiter)
         dt = datetime.datetime.strptime(date_str, date_format) - timezone
         return calendar.timegm(dt.timetuple())
-    except ValueError:
-        pass
 
 
 def date_formats(day_first=True):
@@ -3201,17 +3186,13 @@ def unified_strdate(date_str, day_first=True):
     _, date_str = extract_timezone(date_str)
 
     for expression in date_formats(day_first):
-        try:
+        with compat_contextlib_suppress(ValueError):
             upload_date = datetime.datetime.strptime(date_str, expression).strftime('%Y%m%d')
-        except ValueError:
-            pass
     if upload_date is None:
         timetuple = email.utils.parsedate_tz(date_str)
         if timetuple:
-            try:
+            with compat_contextlib_suppress(ValueError):
                 upload_date = datetime.datetime(*timetuple[:6]).strftime('%Y%m%d')
-            except ValueError:
-                pass
     if upload_date is not None:
         return compat_str(upload_date)
 
@@ -3240,11 +3221,9 @@ def unified_timestamp(date_str, day_first=True):
         date_str = m.group(1)
 
     for expression in date_formats(day_first):
-        try:
+        with compat_contextlib_suppress(ValueError):
             dt = datetime.datetime.strptime(date_str, expression) - timezone + datetime.timedelta(hours=pm_delta)
             return calendar.timegm(dt.timetuple())
-        except ValueError:
-            pass
     timetuple = email.utils.parsedate_tz(date_str)
     if timetuple:
         return calendar.timegm(timetuple) + pm_delta * 3600 - compat_datetime_timedelta_total_seconds(timezone)
@@ -3965,19 +3944,22 @@ def parse_duration(s):
     return duration
 
 
-def prepend_extension(filename, ext, expected_real_ext=None):
+def _change_extension(prepend, filename, ext, expected_real_ext=None):
     name, real_ext = os.path.splitext(filename)
-    return (
-        '{0}.{1}{2}'.format(name, ext, real_ext)
-        if not expected_real_ext or real_ext[1:] == expected_real_ext
-        else '{0}.{1}'.format(filename, ext))
+    sanitize_extension = _UnsafeExtensionError.sanitize_extension
+
+    if not expected_real_ext or real_ext.partition('.')[0::2] == ('', expected_real_ext):
+        filename = name
+        if prepend and real_ext:
+            sanitize_extension(ext, prepend=prepend)
+            return ''.join((filename, '.', ext, real_ext))
+
+    # Mitigate path traversal and file impersonation attacks
+    return '.'.join((filename, sanitize_extension(ext)))
 
 
-def replace_extension(filename, ext, expected_real_ext=None):
-    name, real_ext = os.path.splitext(filename)
-    return '{0}.{1}'.format(
-        name if not expected_real_ext or real_ext[1:] == expected_real_ext else filename,
-        ext)
+prepend_extension = functools.partial(_change_extension, True)
+replace_extension = functools.partial(_change_extension, False)
 
 
 def check_executable(exe, args=[]):
@@ -6262,15 +6244,16 @@ if __debug__:
 
 def traverse_obj(obj, *paths, **kwargs):
     """
-    Safely traverse nested `dict`s and `Iterable`s
+    Safely traverse nested `dict`s and `Iterable`s, etc
 
     >>> obj = [{}, {"key": "value"}]
     >>> traverse_obj(obj, (1, "key"))
-    "value"
+    'value'
 
     Each of the provided `paths` is tested and the first producing a valid result will be returned.
     The next path will also be tested if the path branched but no results could be found.
-    Supported values for traversal are `Mapping`, `Iterable` and `re.Match`.
+    Supported values for traversal are `Mapping`, `Iterable`, `re.Match`, `xml.etree.ElementTree`
+    (xpath) and `http.cookies.Morsel`.
     Unhelpful values (`{}`, `None`) are treated as the absence of a value and discarded.
 
     The paths will be wrapped in `variadic`, so that `'key'` is conveniently the same as `('key', )`.
@@ -6278,8 +6261,9 @@ def traverse_obj(obj, *paths, **kwargs):
     The keys in the path can be one of:
         - `None`:           Return the current object.
         - `set`:            Requires the only item in the set to be a type or function,
-                            like `{type}`/`{func}`. If a `type`, returns only values
-                            of this type. If a function, returns `func(obj)`.
+                            like `{type}`/`{type, type, ...}`/`{func}`. If one or more `type`s,
+                            return only values that have one of the types. If a function,
+                            return `func(obj)`.
         - `str`/`int`:      Return `obj[key]`. For `re.Match`, return `obj.group(key)`.
         - `slice`:          Branch out and return all values in `obj[key]`.
         - `Ellipsis`:       Branch out and return a list of all values.
@@ -6291,8 +6275,10 @@ def traverse_obj(obj, *paths, **kwargs):
                             For `Iterable`s, `key` is the enumeration count of the value.
                             For `re.Match`es, `key` is the group number (0 = full match)
                             as well as additionally any group names, if given.
-        - `dict`            Transform the current object and return a matching dict.
+        - `dict`:           Transform the current object and return a matching dict.
                             Read as: `{key: traverse_obj(obj, path) for key, path in dct.items()}`.
+        - `any`-builtin:    Take the first matching object and return it, resetting branching.
+        - `all`-builtin:    Take all matching objects and return them as a list, resetting branching.
 
         `tuple`, `list`, and `dict` all support nested paths and branches.
 
@@ -6308,10 +6294,8 @@ def traverse_obj(obj, *paths, **kwargs):
     @param get_all          If `False`, return the first matching result, otherwise all matching ones.
     @param casesense        If `False`, consider string dictionary keys as case insensitive.
 
-    The following are only meant to be used by YoutubeDL.prepare_outtmpl and are not part of the API
+    The following is only meant to be used by YoutubeDL.prepare_outtmpl and is not part of the API
 
-    @param _is_user_input    Whether the keys are generated from user input.
-                            If `True` strings get converted to `int`/`slice` if needed.
     @param _traverse_string  Whether to traverse into objects as strings.
                             If `True`, any non-compatible object will first be
                             converted into a string and then traversed into.
@@ -6331,7 +6315,6 @@ def traverse_obj(obj, *paths, **kwargs):
     expected_type = kwargs.get('expected_type')
     get_all = kwargs.get('get_all', True)
     casesense = kwargs.get('casesense', True)
-    _is_user_input = kwargs.get('_is_user_input', False)
     _traverse_string = kwargs.get('_traverse_string', False)
 
     # instant compat
@@ -6345,10 +6328,8 @@ def traverse_obj(obj, *paths, **kwargs):
         type_test = lambda val: try_call(expected_type or IDENTITY, args=(val,))
 
     def lookup_or_none(v, k, getter=None):
-        try:
+        with compat_contextlib_suppress(LookupError):
             return getter(v, k) if getter else v[k]
-        except IndexError:
-            return None
 
     def from_iterable(iterables):
         # chain.from_iterable(['ABC', 'DEF']) --> A B C D E F
@@ -6370,12 +6351,13 @@ def traverse_obj(obj, *paths, **kwargs):
             result = obj
 
         elif isinstance(key, set):
-            assert len(key) == 1, 'Set should only be used to wrap a single item'
-            item = next(iter(key))
-            if isinstance(item, type):
-                result = obj if isinstance(obj, item) else None
+            assert len(key) >= 1, 'At least one item is required in a `set` key'
+            if all(isinstance(item, type) for item in key):
+                result = obj if isinstance(obj, tuple(key)) else None
             else:
-                result = try_call(item, args=(obj,))
+                item = next(iter(key))
+                assert len(key) == 1, 'Multiple items in a `set` key must all be types'
+                result = try_call(item, args=(obj,)) if not isinstance(item, type) else None
 
         elif isinstance(key, (list, tuple)):
             branching = True
@@ -6384,9 +6366,11 @@ def traverse_obj(obj, *paths, **kwargs):
 
         elif key is Ellipsis:
             branching = True
+            if isinstance(obj, compat_http_cookies.Morsel):
+                obj = dict(obj, key=obj.key, value=obj.value)
             if isinstance(obj, compat_collections_abc.Mapping):
                 result = obj.values()
-            elif is_iterable_like(obj):
+            elif is_iterable_like(obj, (compat_collections_abc.Iterable, compat_etree_Element)):
                 result = obj
             elif isinstance(obj, compat_re_Match):
                 result = obj.groups()
@@ -6398,9 +6382,11 @@ def traverse_obj(obj, *paths, **kwargs):
 
         elif callable(key):
             branching = True
+            if isinstance(obj, compat_http_cookies.Morsel):
+                obj = dict(obj, key=obj.key, value=obj.value)
             if isinstance(obj, compat_collections_abc.Mapping):
                 iter_obj = obj.items()
-            elif is_iterable_like(obj):
+            elif is_iterable_like(obj, (compat_collections_abc.Iterable, compat_etree_Element)):
                 iter_obj = enumerate(obj)
             elif isinstance(obj, compat_re_Match):
                 iter_obj = itertools.chain(
@@ -6422,6 +6408,8 @@ def traverse_obj(obj, *paths, **kwargs):
                           if v is not None or default is not NO_DEFAULT) or None
 
         elif isinstance(obj, compat_collections_abc.Mapping):
+            if isinstance(obj, compat_http_cookies.Morsel):
+                obj = dict(obj, key=obj.key, value=obj.value)
             result = (try_call(obj.get, args=(key,))
                       if casesense or try_call(obj.__contains__, args=(key,))
                       else next((v for k, v in obj.items() if casefold(k) == key), None))
@@ -6439,12 +6427,40 @@ def traverse_obj(obj, *paths, **kwargs):
         else:
             result = None
             if isinstance(key, (int, slice)):
-                if is_iterable_like(obj, compat_collections_abc.Sequence):
+                if is_iterable_like(obj, (compat_collections_abc.Sequence, compat_etree_Element)):
                     branching = isinstance(key, slice)
                     result = lookup_or_none(obj, key)
                 elif _traverse_string:
                     result = lookup_or_none(str(obj), key)
 
+            elif isinstance(obj, compat_etree_Element) and isinstance(key, str):
+                xpath, _, special = key.rpartition('/')
+                if not special.startswith('@') and not special.endswith('()'):
+                    xpath = key
+                    special = None
+
+                # Allow abbreviations of relative paths, absolute paths error
+                if xpath.startswith('/'):
+                    xpath = '.' + xpath
+                elif xpath and not xpath.startswith('./'):
+                    xpath = './' + xpath
+
+                def apply_specials(element):
+                    if special is None:
+                        return element
+                    if special == '@':
+                        return element.attrib
+                    if special.startswith('@'):
+                        return try_call(element.attrib.get, args=(special[1:],))
+                    if special == 'text()':
+                        return element.text
+                    raise SyntaxError('apply_specials is missing case for {0!r}'.format(special))
+
+                if xpath:
+                    result = list(map(apply_specials, compat_etree_iterfind(obj, xpath)))
+                else:
+                    result = apply_specials(obj)
+
         return branching, result if branching else (result,)
 
     def lazy_last(iterable):
@@ -6465,17 +6481,18 @@ def traverse_obj(obj, *paths, **kwargs):
 
         key = None
         for last, key in lazy_last(variadic(path, (str, bytes, dict, set))):
-            if _is_user_input and isinstance(key, str):
-                if key == ':':
-                    key = Ellipsis
-                elif ':' in key:
-                    key = slice(*map(int_or_none, key.split(':')))
-                elif int_or_none(key) is not None:
-                    key = int(key)
-
             if not casesense and isinstance(key, str):
                 key = compat_casefold(key)
 
+            if key in (any, all):
+                has_branched = False
+                filtered_objs = (obj for obj in objs if obj not in (None, {}))
+                if key is any:
+                    objs = (next(filtered_objs, None),)
+                else:
+                    objs = (list(filtered_objs),)
+                continue
+
             if __debug__ and callable(key):
                 # Verify function signature
                 _try_bind_args(key, None, None)
@@ -6514,9 +6531,9 @@ def traverse_obj(obj, *paths, **kwargs):
     return None if default is NO_DEFAULT else default
 
 
-def T(x):
-    """ For use in yt-dl instead of {type} or set((type,)) """
-    return set((x,))
+def T(*x):
+    """ For use in yt-dl instead of {type, ...} or set((type, ...)) """
+    return set(x)
 
 
 def get_first(obj, keys, **kwargs):
@@ -6532,3 +6549,169 @@ def join_nonempty(*values, **kwargs):
     if from_dict is not None:
         values = (traverse_obj(from_dict, variadic(v)) for v in values)
     return delim.join(map(compat_str, filter(None, values)))
+
+
+class Namespace(object):
+    """Immutable namespace"""
+
+    def __init__(self, **kw_attr):
+        self.__dict__.update(kw_attr)
+
+    def __iter__(self):
+        return iter(self.__dict__.values())
+
+    @property
+    def items_(self):
+        return self.__dict__.items()
+
+
+MEDIA_EXTENSIONS = Namespace(
+    common_video=('avi', 'flv', 'mkv', 'mov', 'mp4', 'webm'),
+    video=('3g2', '3gp', 'f4v', 'mk3d', 'divx', 'mpg', 'ogv', 'm4v', 'wmv'),
+    common_audio=('aiff', 'alac', 'flac', 'm4a', 'mka', 'mp3', 'ogg', 'opus', 'wav'),
+    audio=('aac', 'ape', 'asf', 'f4a', 'f4b', 'm4b', 'm4p', 'm4r', 'oga', 'ogx', 'spx', 'vorbis', 'wma', 'weba'),
+    thumbnails=('jpg', 'png', 'webp'),
+    # storyboards=('mhtml', ),
+    subtitles=('srt', 'vtt', 'ass', 'lrc', 'ttml'),
+    manifests=('f4f', 'f4m', 'm3u8', 'smil', 'mpd'),
+)
+MEDIA_EXTENSIONS.video = MEDIA_EXTENSIONS.common_video + MEDIA_EXTENSIONS.video
+MEDIA_EXTENSIONS.audio = MEDIA_EXTENSIONS.common_audio + MEDIA_EXTENSIONS.audio
+
+KNOWN_EXTENSIONS = (
+    MEDIA_EXTENSIONS.video + MEDIA_EXTENSIONS.audio
+    + MEDIA_EXTENSIONS.manifests
+)
+
+
+class _UnsafeExtensionError(Exception):
+    """
+    Mitigation exception for unwanted file overwrite/path traversal
+
+    Ref: https://github.com/yt-dlp/yt-dlp/security/advisories/GHSA-79w7-vh3h-8g4j
+    """
+    _ALLOWED_EXTENSIONS = frozenset(itertools.chain(
+        (   # internal
+            'description',
+            'json',
+            'meta',
+            'orig',
+            'part',
+            'temp',
+            'uncut',
+            'unknown_video',
+            'ytdl',
+        ),
+        # video
+        MEDIA_EXTENSIONS.video, (
+            'asx',
+            'ismv',
+            'm2t',
+            'm2ts',
+            'm2v',
+            'm4s',
+            'mng',
+            'mp2v',
+            'mp4v',
+            'mpe',
+            'mpeg',
+            'mpeg1',
+            'mpeg2',
+            'mpeg4',
+            'mxf',
+            'ogm',
+            'qt',
+            'rm',
+            'swf',
+            'ts',
+            'vob',
+            'vp9',
+        ),
+        # audio
+        MEDIA_EXTENSIONS.audio, (
+            '3ga',
+            'ac3',
+            'adts',
+            'aif',
+            'au',
+            'dts',
+            'isma',
+            'it',
+            'mid',
+            'mod',
+            'mpga',
+            'mp1',
+            'mp2',
+            'mp4a',
+            'mpa',
+            'ra',
+            'shn',
+            'xm',
+        ),
+        # image
+        MEDIA_EXTENSIONS.thumbnails, (
+            'avif',
+            'bmp',
+            'gif',
+            'ico',
+            'heic',
+            'jng',
+            'jpeg',
+            'jxl',
+            'svg',
+            'tif',
+            'tiff',
+            'wbmp',
+        ),
+        # subtitle
+        MEDIA_EXTENSIONS.subtitles, (
+            'dfxp',
+            'fs',
+            'ismt',
+            'json3',
+            'sami',
+            'scc',
+            'srv1',
+            'srv2',
+            'srv3',
+            'ssa',
+            'tt',
+            'xml',
+        ),
+        # others
+        MEDIA_EXTENSIONS.manifests,
+        (
+            # not used in yt-dl
+            # *MEDIA_EXTENSIONS.storyboards,
+            # 'desktop',
+            # 'ism',
+            # 'm3u',
+            # 'sbv',
+            # 'swp',
+            # 'url',
+            # 'webloc',
+        )))
+
+    def __init__(self, extension):
+        super(_UnsafeExtensionError, self).__init__('unsafe file extension: {0!r}'.format(extension))
+        self.extension = extension
+
+    # support --no-check-extensions
+    lenient = False
+
+    @classmethod
+    def sanitize_extension(cls, extension, **kwargs):
+        # ... /, *, prepend=False
+        prepend = kwargs.get('prepend', False)
+
+        if '/' in extension or '\\' in extension:
+            raise cls(extension)
+
+        if not prepend:
+            last = extension.rpartition('.')[-1]
+            if last == 'bin':
+                extension = last = 'unknown_video'
+            if not (cls.lenient or last.lower() in cls._ALLOWED_EXTENSIONS):
+                raise cls(extension)
+
+        return extension