4 from __future__
import unicode_literals
39 import xml
.etree
.ElementTree
44 compat_HTMLParseError
,
50 compat_ctypes_WINFUNCTYPE
,
51 compat_etree_fromstring
,
54 compat_html_entities_html5
,
67 compat_urllib_parse_urlencode
,
68 compat_urllib_parse_urlparse
,
69 compat_urllib_parse_urlunparse
,
70 compat_urllib_parse_quote
,
71 compat_urllib_parse_quote_plus
,
72 compat_urllib_parse_unquote_plus
,
73 compat_urllib_request
,
84 def register_socks_protocols():
85 # "Register" SOCKS protocols
86 # In Python < 2.6.5, urlsplit() suffers from bug https://bugs.python.org/issue7904
87 # URLs with protocols not in urlparse.uses_netloc are not handled correctly
88 for scheme
in ('socks', 'socks4', 'socks4a', 'socks5'):
89 if scheme
not in compat_urlparse
.uses_netloc
:
90 compat_urlparse
.uses_netloc
.append(scheme
)
93 # This is not clearly defined otherwise
94 compiled_regex_type
= type(re
.compile(''))
97 def random_user_agent():
98 _USER_AGENT_TPL
= 'Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/%s Safari/537.36'
139 return _USER_AGENT_TPL
% random
.choice(_CHROME_VERSIONS
)
143 'User-Agent': random_user_agent(),
144 'Accept': 'text/html,application/xhtml+xml,application/xml;q=0.9,*/*;q=0.8',
145 'Accept-Encoding': 'gzip, deflate',
146 'Accept-Language': 'en-us,en;q=0.5',
151 'Safari': 'Mozilla/5.0 (X11; Linux x86_64; rv:10.0) AppleWebKit/533.20.25 (KHTML, like Gecko) Version/5.0.4 Safari/533.20.27',
155 NO_DEFAULT
= object()
157 ENGLISH_MONTH_NAMES
= [
158 'January', 'February', 'March', 'April', 'May', 'June',
159 'July', 'August', 'September', 'October', 'November', 'December']
162 'en': ENGLISH_MONTH_NAMES
,
164 'janvier', 'février', 'mars', 'avril', 'mai', 'juin',
165 'juillet', 'août', 'septembre', 'octobre', 'novembre', 'décembre'],
169 'mp4', 'm4a', 'm4p', 'm4b', 'm4r', 'm4v', 'aac',
170 'flv', 'f4v', 'f4a', 'f4b',
171 'webm', 'ogg', 'ogv', 'oga', 'ogx', 'spx', 'opus',
172 'mkv', 'mka', 'mk3d',
181 'f4f', 'f4m', 'm3u8', 'smil')
183 # needed for sanitizing filenames in restricted mode
184 ACCENT_CHARS
= dict(zip('ÂÃÄÀÁÅÆÇÈÉÊËÌÍÎÏÐÑÒÓÔÕÖŐØŒÙÚÛÜŰÝÞßàáâãäåæçèéêëìíîïðñòóôõöőøœùúûüűýþÿ',
185 itertools
.chain('AAAAAA', ['AE'], 'CEEEEIIIIDNOOOOOOO', ['OE'], 'UUUUUY', ['TH', 'ss'],
186 'aaaaaa', ['ae'], 'ceeeeiiiionooooooo', ['oe'], 'uuuuuy', ['th'], 'y')))
215 '%Y-%m-%d %H:%M:%S.%f',
216 '%Y-%m-%d %H:%M:%S:%f',
219 '%Y-%m-%dT%H:%M:%SZ',
220 '%Y-%m-%dT%H:%M:%S.%fZ',
221 '%Y-%m-%dT%H:%M:%S.%f0Z',
223 '%Y-%m-%dT%H:%M:%S.%f',
226 '%b %d %Y at %H:%M:%S',
228 '%B %d %Y at %H:%M:%S',
232 DATE_FORMATS_DAY_FIRST
= list(DATE_FORMATS
)
233 DATE_FORMATS_DAY_FIRST
.extend([
242 DATE_FORMATS_MONTH_FIRST
= list(DATE_FORMATS
)
243 DATE_FORMATS_MONTH_FIRST
.extend([
251 PACKED_CODES_RE
= r
"}\('(.+)',(\d+),(\d+),'([^']+)'\.split\('\|'\)"
252 JSON_LD_RE
= r
'(?is)<script[^>]+type=(["\']?
)application
/ld\
+json\
1[^
>]*>(?P
<json_ld
>.+?
)</script
>'
255 def preferredencoding():
256 """Get preferred encoding.
258 Returns the best encoding scheme for the system, based on
259 locale.getpreferredencoding() and some further tweaks.
262 pref = locale.getpreferredencoding()
270 def write_json_file(obj, fn):
271 """ Encode obj as JSON and write it to fn, atomically if possible """
273 fn = encodeFilename(fn)
274 if sys.version_info < (3, 0) and sys.platform != 'win32
':
275 encoding = get_filesystem_encoding()
276 # os.path.basename returns a bytes object, but NamedTemporaryFile
277 # will fail if the filename contains non ascii characters unless we
278 # use a unicode object
279 path_basename = lambda f: os.path.basename(fn).decode(encoding)
280 # the same for os.path.dirname
281 path_dirname = lambda f: os.path.dirname(fn).decode(encoding)
283 path_basename = os.path.basename
284 path_dirname = os.path.dirname
288 'prefix
': path_basename(fn) + '.',
289 'dir': path_dirname(fn),
293 # In Python 2.x, json.dump expects a bytestream.
294 # In Python 3.x, it writes to a character stream
295 if sys.version_info < (3, 0):
303 tf = tempfile.NamedTemporaryFile(**compat_kwargs(args))
308 if sys.platform == 'win32
':
309 # Need to remove existing file on Windows, else os.rename raises
310 # WindowsError or FileExistsError.
318 os.chmod(tf.name, 0o666 & ~mask)
321 os.rename(tf.name, fn)
330 if sys.version_info >= (2, 7):
331 def find_xpath_attr(node, xpath, key, val=None):
332 """ Find the xpath xpath[@key=val] """
333 assert re.match(r'^
[a
-zA
-Z_
-]+$
', key)
334 expr = xpath + ('[@%s]' % key if val is None else "[@%s='%s']" % (key, val))
335 return node.find(expr)
337 def find_xpath_attr(node, xpath, key, val=None):
338 for f in node.findall(compat_xpath(xpath)):
339 if key not in f.attrib:
341 if val is None or f.attrib.get(key) == val:
345 # On python2.6 the xml.etree.ElementTree.Element methods don't support
346 # the namespace parameter
349 def xpath_with_ns(path
, ns_map
):
350 components
= [c
.split(':') for c
in path
.split('/')]
354 replaced
.append(c
[0])
357 replaced
.append('{%s}%s' % (ns_map
[ns
], tag
))
358 return '/'.join(replaced
)
361 def xpath_element(node
, xpath
, name
=None, fatal
=False, default
=NO_DEFAULT
):
362 def _find_xpath(xpath
):
363 return node
.find(compat_xpath(xpath
))
365 if isinstance(xpath
, (str, compat_str
)):
366 n
= _find_xpath(xpath
)
374 if default
is not NO_DEFAULT
:
377 name
= xpath
if name
is None else name
378 raise ExtractorError('Could not find XML element %s' % name
)
384 def xpath_text(node
, xpath
, name
=None, fatal
=False, default
=NO_DEFAULT
):
385 n
= xpath_element(node
, xpath
, name
, fatal
=fatal
, default
=default
)
386 if n
is None or n
== default
:
389 if default
is not NO_DEFAULT
:
392 name
= xpath
if name
is None else name
393 raise ExtractorError('Could not find XML element\'s text %s' % name
)
399 def xpath_attr(node
, xpath
, key
, name
=None, fatal
=False, default
=NO_DEFAULT
):
400 n
= find_xpath_attr(node
, xpath
, key
)
402 if default
is not NO_DEFAULT
:
405 name
= '%s[@%s]' % (xpath
, key
) if name
is None else name
406 raise ExtractorError('Could not find XML attribute %s' % name
)
412 def get_element_by_id(id, html
):
413 """Return the content of the tag with the specified ID in the passed HTML document"""
414 return get_element_by_attribute('id', id, html
)
417 def get_element_by_class(class_name
, html
):
418 """Return the content of the first tag with the specified class in the passed HTML document"""
419 retval
= get_elements_by_class(class_name
, html
)
420 return retval
[0] if retval
else None
423 def get_element_by_attribute(attribute
, value
, html
, escape_value
=True):
424 retval
= get_elements_by_attribute(attribute
, value
, html
, escape_value
)
425 return retval
[0] if retval
else None
428 def get_elements_by_class(class_name
, html
):
429 """Return the content of all tags with the specified class in the passed HTML document as a list"""
430 return get_elements_by_attribute(
431 'class', r
'[^\'"]*\b%s\b[^\'"]*' % re.escape(class_name),
432 html, escape_value=False)
435 def get_elements_by_attribute(attribute, value, html, escape_value=True):
436 """Return the content of the tag with the specified attribute in the passed HTML document"""
438 value = re.escape(value) if escape_value else value
441 for m in re.finditer(r'''(?xs)
443 (?:\s+[a-zA-Z0-9:._-]+(?:=[a-zA-Z0-9:._-]*|="[^"]*"|='[^
']*'|
))*?
445 (?:\s+[a-zA-Z0-9:._-]+(?:=[a-zA-Z0-9:._-]*|="[^
"]*"|
='[^']*'|))*?
449 ''' % (re.escape(attribute), value), html):
450 res = m.group('content
')
452 if res.startswith('"') or res.startswith("'"):
455 retlist.append(unescapeHTML(res))
460 class HTMLAttributeParser(compat_HTMLParser):
461 """Trivial HTML parser to gather the attributes for a single element"""
465 compat_HTMLParser.__init__(self)
467 def handle_starttag(self, tag, attrs):
468 self.attrs = dict(attrs)
471 class HTMLListAttrsParser(compat_HTMLParser):
472 """HTML parser to gather the attributes for the elements of a list"""
475 compat_HTMLParser.__init__(self)
479 def handle_starttag(self, tag, attrs):
480 if tag == 'li
' and self._level == 0:
481 self.items.append(dict(attrs))
484 def handle_endtag(self, tag):
488 def extract_attributes(html_element):
489 """Given a string for an HTML element such as
491 a="foo" B="bar" c="&98;az" d=boz
492 empty= noval entity="&"
495 Decode and return a dictionary of attributes.
497 'a
': 'foo
', 'b
': 'bar
', c: 'baz
', d: 'boz
',
498 'empty
': '', 'noval
': None, 'entity
': '&',
499 'sq
': '"', 'dq': '\''
501 NB HTMLParser is stricter in Python 2.6 & 3.2 than in later versions,
502 but the cases in the unit test will work for all of 2.6, 2.7, 3.2-3.5.
504 parser = HTMLAttributeParser()
506 parser.feed(html_element)
508 # Older Python may throw HTMLParseError in case of malformed HTML
509 except compat_HTMLParseError:
514 def parse_list(webpage):
515 """Given a string for an series of HTML <li> elements,
516 return a dictionary of their attributes"""
517 parser = HTMLListAttrsParser()
523 def clean_html(html):
524 """Clean an HTML snippet into a readable string"""
526 if html is None: # Convenience for sanitizing descriptions etc.
530 html = html.replace('\n', ' ')
531 html = re.sub(r'(?u)\s*<\s*br\s*/?\s*>\s*', '\n', html)
532 html = re.sub(r'(?u)<\s*/\s*p\s*>\s*<\s*p[^>]*>', '\n', html)
534 html = re.sub('<.*?>', '', html)
535 # Replace html entities
536 html = unescapeHTML(html)
540 def sanitize_open(filename, open_mode):
541 """Try to open the given filename, and slightly tweak it if this fails.
543 Attempts to open the given filename. If this fails, it tries to change
544 the filename slightly, step by step, until it's either able to open it
545 or it fails and raises a final exception, like the standard open()
548 It returns the tuple (stream, definitive_file_name).
552 if sys.platform == 'win32':
554 msvcrt.setmode(sys.stdout.fileno(), os.O_BINARY)
555 return (sys.stdout.buffer if hasattr(sys.stdout, 'buffer') else sys.stdout, filename)
556 stream = open(encodeFilename(filename), open_mode)
557 return (stream, filename)
558 except (IOError, OSError) as err:
559 if err.errno in (errno.EACCES,):
562 # In case of error, try to remove win32 forbidden chars
563 alt_filename = sanitize_path(filename)
564 if alt_filename == filename:
567 # An exception here should be caught in the caller
568 stream = open(encodeFilename(alt_filename), open_mode)
569 return (stream, alt_filename)
572 def timeconvert(timestr):
573 """Convert RFC 2822 defined time string into system timestamp"""
575 timetuple = email.utils.parsedate_tz(timestr)
576 if timetuple is not None:
577 timestamp = email.utils.mktime_tz(timetuple)
581 def sanitize_filename(s, restricted=False, is_id=False):
582 """Sanitizes a string so it could be used as part of a filename.
583 If restricted is set, use a stricter subset of allowed characters.
584 Set is_id if this is not an arbitrary string, but an ID that should be kept
587 def replace_insane(char):
588 if restricted and char in ACCENT_CHARS:
589 return ACCENT_CHARS[char]
590 elif not restricted and char == '\n':
592 elif char == '?' or ord(char) < 32 or ord(char) == 127:
595 return '' if restricted else '\''
597 return '_
-' if restricted else ' -'
598 elif char in '\\/|
*<>':
600 if restricted and (char in '!&\'()[]{}$
;`^
,#' or char.isspace()):
602 if restricted
and ord(char
) > 127:
609 s
= re
.sub(r
'[0-9]+(?::[0-9]+)+', lambda m
: m
.group(0).replace(':', '_'), s
)
610 result
= ''.join(map(replace_insane
, s
))
612 while '__' in result
:
613 result
= result
.replace('__', '_')
614 result
= result
.strip('_')
615 # Common case of "Foreign band name - English song title"
616 if restricted
and result
.startswith('-_'):
618 if result
.startswith('-'):
619 result
= '_' + result
[len('-'):]
620 result
= result
.lstrip('.')
626 def sanitize_path(s
, force
=False):
627 """Sanitizes and normalizes path on Windows"""
628 if sys
.platform
== 'win32':
630 drive_or_unc
, _
= os
.path
.splitdrive(s
)
631 if sys
.version_info
< (2, 7) and not drive_or_unc
:
632 drive_or_unc
, _
= os
.path
.splitunc(s
)
638 norm_path
= os
.path
.normpath(remove_start(s
, drive_or_unc
)).split(os
.path
.sep
)
642 path_part
if path_part
in ['.', '..'] else re
.sub(r
'(?:[/<>:"\|\\?\*]|[\s.]$)', '#', path_part
)
643 for path_part
in norm_path
]
645 sanitized_path
.insert(0, drive_or_unc
+ os
.path
.sep
)
646 elif force
and s
[0] == os
.path
.sep
:
647 sanitized_path
.insert(0, os
.path
.sep
)
648 return os
.path
.join(*sanitized_path
)
651 def sanitize_url(url
):
652 # Prepend protocol-less URLs with `http:` scheme in order to mitigate
653 # the number of unwanted failures due to missing protocol
654 if url
.startswith('//'):
655 return 'http:%s' % url
656 # Fix some common typos seen so far
658 # https://github.com/ytdl-org/youtube-dl/issues/15649
659 (r
'^httpss://', r
'https://'),
660 # https://bx1.be/lives/direct-tv/
661 (r
'^rmtp([es]?)://', r
'rtmp\1://'),
663 for mistake
, fixup
in COMMON_TYPOS
:
664 if re
.match(mistake
, url
):
665 return re
.sub(mistake
, fixup
, url
)
669 def extract_basic_auth(url
):
670 parts
= compat_urlparse
.urlsplit(url
)
671 if parts
.username
is None:
673 url
= compat_urlparse
.urlunsplit(parts
._replace
(netloc
=(
674 parts
.hostname
if parts
.port
is None
675 else '%s:%d' % (parts
.hostname
, parts
.port
))))
676 auth_payload
= base64
.b64encode(
677 ('%s:%s' % (parts
.username
, parts
.password
or '')).encode('utf-8'))
678 return url
, 'Basic ' + auth_payload
.decode('utf-8')
681 def sanitized_Request(url
, *args
, **kwargs
):
682 url
, auth_header
= extract_basic_auth(escape_url(sanitize_url(url
)))
683 if auth_header
is not None:
684 headers
= args
[1] if len(args
) >= 2 else kwargs
.setdefault('headers', {})
685 headers
['Authorization'] = auth_header
686 return compat_urllib_request
.Request(url
, *args
, **kwargs
)
690 """Expand shell variables and ~"""
691 return os
.path
.expandvars(compat_expanduser(s
))
694 def orderedSet(iterable
):
695 """ Remove all duplicates from the input iterable """
703 def _htmlentity_transform(entity_with_semicolon
):
704 """Transforms an HTML entity to a character."""
705 entity
= entity_with_semicolon
[:-1]
707 # Known non-numeric HTML entity
708 if entity
in compat_html_entities
.name2codepoint
:
709 return compat_chr(compat_html_entities
.name2codepoint
[entity
])
711 # TODO: HTML5 allows entities without a semicolon. For example,
712 # 'Éric' should be decoded as 'Éric'.
713 if entity_with_semicolon
in compat_html_entities_html5
:
714 return compat_html_entities_html5
[entity_with_semicolon
]
716 mobj
= re
.match(r
'#(x[0-9a-fA-F]+|[0-9]+)', entity
)
718 numstr
= mobj
.group(1)
719 if numstr
.startswith('x'):
721 numstr
= '0%s' % numstr
724 # See https://github.com/ytdl-org/youtube-dl/issues/7518
726 return compat_chr(int(numstr
, base
))
730 # Unknown entity in name, return its literal representation
731 return '&%s;' % entity
737 assert type(s
) == compat_str
740 r
'&([^&;]+;)', lambda m
: _htmlentity_transform(m
.group(1)), s
)
743 def escapeHTML(text
):
746 .replace('&', '&')
747 .replace('<', '<')
748 .replace('>', '>')
749 .replace('"', '"')
750 .replace("'", ''')
754 def process_communicate_or_kill(p
, *args
, **kwargs
):
756 return p
.communicate(*args
, **kwargs
)
757 except BaseException
: # Including KeyboardInterrupt
763 class Popen(subprocess
.Popen
):
764 if sys
.platform
== 'win32':
765 _startupinfo
= subprocess
.STARTUPINFO()
766 _startupinfo
.dwFlags |
= subprocess
.STARTF_USESHOWWINDOW
770 def __init__(self
, *args
, **kwargs
):
771 super(Popen
, self
).__init
__(*args
, **kwargs
, startupinfo
=self
._startupinfo
)
773 def communicate_or_kill(self
, *args
, **kwargs
):
774 return process_communicate_or_kill(self
, *args
, **kwargs
)
777 def get_subprocess_encoding():
778 if sys
.platform
== 'win32' and sys
.getwindowsversion()[0] >= 5:
779 # For subprocess calls, encode with locale encoding
780 # Refer to http://stackoverflow.com/a/9951851/35070
781 encoding
= preferredencoding()
783 encoding
= sys
.getfilesystemencoding()
789 def encodeFilename(s
, for_subprocess
=False):
791 @param s The name of the file
794 assert type(s
) == compat_str
796 # Python 3 has a Unicode API
797 if sys
.version_info
>= (3, 0):
800 # Pass '' directly to use Unicode APIs on Windows 2000 and up
801 # (Detecting Windows NT 4 is tricky because 'major >= 4' would
802 # match Windows 9x series as well. Besides, NT 4 is obsolete.)
803 if not for_subprocess
and sys
.platform
== 'win32' and sys
.getwindowsversion()[0] >= 5:
806 # Jython assumes filenames are Unicode strings though reported as Python 2.x compatible
807 if sys
.platform
.startswith('java'):
810 return s
.encode(get_subprocess_encoding(), 'ignore')
813 def decodeFilename(b
, for_subprocess
=False):
815 if sys
.version_info
>= (3, 0):
818 if not isinstance(b
, bytes):
821 return b
.decode(get_subprocess_encoding(), 'ignore')
824 def encodeArgument(s
):
825 if not isinstance(s
, compat_str
):
826 # Legacy code that uses byte strings
827 # Uncomment the following line after fixing all post processors
828 # assert False, 'Internal error: %r should be of type %r, is %r' % (s, compat_str, type(s))
829 s
= s
.decode('ascii')
830 return encodeFilename(s
, True)
833 def decodeArgument(b
):
834 return decodeFilename(b
, True)
837 def decodeOption(optval
):
840 if isinstance(optval
, bytes):
841 optval
= optval
.decode(preferredencoding())
843 assert isinstance(optval
, compat_str
)
847 _timetuple
= collections
.namedtuple('Time', ('hours', 'minutes', 'seconds', 'milliseconds'))
850 def timetuple_from_msec(msec
):
851 secs
, msec
= divmod(msec
, 1000)
852 mins
, secs
= divmod(secs
, 60)
853 hrs
, mins
= divmod(mins
, 60)
854 return _timetuple(hrs
, mins
, secs
, msec
)
857 def formatSeconds(secs
, delim
=':', msec
=False):
858 time
= timetuple_from_msec(secs
* 1000)
860 ret
= '%d%s%02d%s%02d' % (time
.hours
, delim
, time
.minutes
, delim
, time
.seconds
)
862 ret
= '%d%s%02d' % (time
.minutes
, delim
, time
.seconds
)
864 ret
= '%d' % time
.seconds
865 return '%s.%03d' % (ret
, time
.milliseconds
) if msec
else ret
868 def _ssl_load_windows_store_certs(ssl_context
, storename
):
869 # Code adapted from _load_windows_store_certs in https://github.com/python/cpython/blob/main/Lib/ssl.py
871 certs
= [cert
for cert
, encoding
, trust
in ssl
.enum_certificates(storename
)
872 if encoding
== 'x509_asn' and (
873 trust
is True or ssl
.Purpose
.SERVER_AUTH
.oid
in trust
)]
874 except PermissionError
:
878 ssl_context
.load_verify_locations(cadata
=cert
)
883 def make_HTTPS_handler(params
, **kwargs
):
884 opts_check_certificate
= not params
.get('nocheckcertificate')
885 context
= ssl
.SSLContext(ssl
.PROTOCOL_TLS_CLIENT
)
886 context
.check_hostname
= opts_check_certificate
887 context
.verify_mode
= ssl
.CERT_REQUIRED
if opts_check_certificate
else ssl
.CERT_NONE
888 if opts_check_certificate
:
890 context
.load_default_certs()
891 # Work around the issue in load_default_certs when there are bad certificates. See:
892 # https://github.com/yt-dlp/yt-dlp/issues/1060,
893 # https://bugs.python.org/issue35665, https://bugs.python.org/issue45312
895 # enum_certificates is not present in mingw python. See https://github.com/yt-dlp/yt-dlp/issues/1151
896 if sys
.platform
== 'win32' and hasattr(ssl
, 'enum_certificates'):
897 # Create a new context to discard any certificates that were already loaded
898 context
= ssl
.SSLContext(ssl
.PROTOCOL_TLS_CLIENT
)
899 context
.check_hostname
, context
.verify_mode
= True, ssl
.CERT_REQUIRED
900 for storename
in ('CA', 'ROOT'):
901 _ssl_load_windows_store_certs(context
, storename
)
902 context
.set_default_verify_paths()
903 return YoutubeDLHTTPSHandler(params
, context
=context
, **kwargs
)
906 def bug_reports_message(before
=';'):
907 if ytdl_is_updateable():
908 update_cmd
= 'type yt-dlp -U to update'
910 update_cmd
= 'see https://github.com/yt-dlp/yt-dlp on how to update'
911 msg
= 'please report this issue on https://github.com/yt-dlp/yt-dlp .'
912 msg
+= ' Make sure you are using the latest version; %s.' % update_cmd
913 msg
+= ' Be sure to call yt-dlp with the --verbose flag and include its complete output.'
915 before
= before
.rstrip()
916 if not before
or before
.endswith(('.', '!', '?')):
917 msg
= msg
[0].title() + msg
[1:]
919 return (before
+ ' ' if before
else '') + msg
922 class YoutubeDLError(Exception):
923 """Base exception for YoutubeDL errors."""
926 def __init__(self
, msg
=None):
929 elif self
.msg
is None:
930 self
.msg
= type(self
).__name
__
931 super().__init
__(self
.msg
)
934 network_exceptions
= [compat_urllib_error
.URLError
, compat_http_client
.HTTPException
, socket
.error
]
935 if hasattr(ssl
, 'CertificateError'):
936 network_exceptions
.append(ssl
.CertificateError
)
937 network_exceptions
= tuple(network_exceptions
)
940 class ExtractorError(YoutubeDLError
):
941 """Error during info extraction."""
943 def __init__(self
, msg
, tb
=None, expected
=False, cause
=None, video_id
=None, ie
=None):
944 """ tb, if given, is the original traceback (so that it can be printed out).
945 If expected is set, this is a normal error message and most likely not a bug in yt-dlp.
947 if sys
.exc_info()[0] in network_exceptions
:
952 self
.expected
= expected
954 self
.video_id
= video_id
956 self
.exc_info
= sys
.exc_info() # preserve original exception
958 super(ExtractorError
, self
).__init
__(''.join((
959 format_field(ie
, template
='[%s] '),
960 format_field(video_id
, template
='%s: '),
962 format_field(cause
, template
=' (caused by %r)'),
963 '' if expected
else bug_reports_message())))
965 def format_traceback(self
):
966 if self
.traceback
is None:
968 return ''.join(traceback
.format_tb(self
.traceback
))
971 class UnsupportedError(ExtractorError
):
972 def __init__(self
, url
):
973 super(UnsupportedError
, self
).__init
__(
974 'Unsupported URL: %s' % url
, expected
=True)
978 class RegexNotFoundError(ExtractorError
):
979 """Error when a regex didn't match"""
983 class GeoRestrictedError(ExtractorError
):
984 """Geographic restriction Error exception.
986 This exception may be thrown when a video is not available from your
987 geographic location due to geographic restrictions imposed by a website.
990 def __init__(self
, msg
, countries
=None, **kwargs
):
991 kwargs
['expected'] = True
992 super(GeoRestrictedError
, self
).__init
__(msg
, **kwargs
)
993 self
.countries
= countries
996 class DownloadError(YoutubeDLError
):
997 """Download Error exception.
999 This exception may be thrown by FileDownloader objects if they are not
1000 configured to continue on errors. They will contain the appropriate
1004 def __init__(self
, msg
, exc_info
=None):
1005 """ exc_info, if given, is the original exception that caused the trouble (as returned by sys.exc_info()). """
1006 super(DownloadError
, self
).__init
__(msg
)
1007 self
.exc_info
= exc_info
1010 class EntryNotInPlaylist(YoutubeDLError
):
1011 """Entry not in playlist exception.
1013 This exception will be thrown by YoutubeDL when a requested entry
1014 is not found in the playlist info_dict
1016 msg
= 'Entry not found in info'
1019 class SameFileError(YoutubeDLError
):
1020 """Same File exception.
1022 This exception will be thrown by FileDownloader objects if they detect
1023 multiple files would have to be downloaded to the same file on disk.
1025 msg
= 'Fixed output name but more than one file to download'
1027 def __init__(self
, filename
=None):
1028 if filename
is not None:
1029 self
.msg
+= f
': {filename}'
1030 super().__init
__(self
.msg
)
1033 class PostProcessingError(YoutubeDLError
):
1034 """Post Processing exception.
1036 This exception may be raised by PostProcessor's .run() method to
1037 indicate an error in the postprocessing task.
1041 class DownloadCancelled(YoutubeDLError
):
1042 """ Exception raised when the download queue should be interrupted """
1043 msg
= 'The download was cancelled'
1046 class ExistingVideoReached(DownloadCancelled
):
1047 """ --break-on-existing triggered """
1048 msg
= 'Encountered a video that is already in the archive, stopping due to --break-on-existing'
1051 class RejectedVideoReached(DownloadCancelled
):
1052 """ --break-on-reject triggered """
1053 msg
= 'Encountered a video that did not match filter, stopping due to --break-on-reject'
1056 class MaxDownloadsReached(DownloadCancelled
):
1057 """ --max-downloads limit has been reached. """
1058 msg
= 'Maximum number of downloads reached, stopping due to --max-downloads'
1061 class ReExtractInfo(YoutubeDLError
):
1062 """ Video info needs to be re-extracted. """
1064 def __init__(self
, msg
, expected
=False):
1065 super().__init
__(msg
)
1066 self
.expected
= expected
1069 class ThrottledDownload(ReExtractInfo
):
1070 """ Download speed below --throttled-rate. """
1071 msg
= 'The download speed is below throttle limit'
1074 super().__init
__(self
.msg
, expected
=False)
1077 class UnavailableVideoError(YoutubeDLError
):
1078 """Unavailable Format exception.
1080 This exception will be thrown when a video is requested
1081 in a format that is not available for that video.
1083 msg
= 'Unable to download video'
1085 def __init__(self
, err
=None):
1087 self
.msg
+= f
': {err}'
1088 super().__init
__(self
.msg
)
1091 class ContentTooShortError(YoutubeDLError
):
1092 """Content Too Short exception.
1094 This exception may be raised by FileDownloader objects when a file they
1095 download is too small for what the server announced first, indicating
1096 the connection was probably interrupted.
1099 def __init__(self
, downloaded
, expected
):
1100 super(ContentTooShortError
, self
).__init
__(
1101 'Downloaded {0} bytes, expected {1} bytes'.format(downloaded
, expected
)
1104 self
.downloaded
= downloaded
1105 self
.expected
= expected
1108 class XAttrMetadataError(YoutubeDLError
):
1109 def __init__(self
, code
=None, msg
='Unknown error'):
1110 super(XAttrMetadataError
, self
).__init
__(msg
)
1114 # Parsing code and msg
1115 if (self
.code
in (errno
.ENOSPC
, errno
.EDQUOT
)
1116 or 'No space left' in self
.msg
or 'Disk quota exceeded' in self
.msg
):
1117 self
.reason
= 'NO_SPACE'
1118 elif self
.code
== errno
.E2BIG
or 'Argument list too long' in self
.msg
:
1119 self
.reason
= 'VALUE_TOO_LONG'
1121 self
.reason
= 'NOT_SUPPORTED'
1124 class XAttrUnavailableError(YoutubeDLError
):
1128 def _create_http_connection(ydl_handler
, http_class
, is_https
, *args
, **kwargs
):
1129 # Working around python 2 bug (see http://bugs.python.org/issue17849) by limiting
1130 # expected HTTP responses to meet HTTP/1.0 or later (see also
1131 # https://github.com/ytdl-org/youtube-dl/issues/6727)
1132 if sys
.version_info
< (3, 0):
1133 kwargs
['strict'] = True
1134 hc
= http_class(*args
, **compat_kwargs(kwargs
))
1135 source_address
= ydl_handler
._params
.get('source_address')
1137 if source_address
is not None:
1138 # This is to workaround _create_connection() from socket where it will try all
1139 # address data from getaddrinfo() including IPv6. This filters the result from
1140 # getaddrinfo() based on the source_address value.
1141 # This is based on the cpython socket.create_connection() function.
1142 # https://github.com/python/cpython/blob/master/Lib/socket.py#L691
1143 def _create_connection(address
, timeout
=socket
._GLOBAL
_DEFAULT
_TIMEOUT
, source_address
=None):
1144 host
, port
= address
1146 addrs
= socket
.getaddrinfo(host
, port
, 0, socket
.SOCK_STREAM
)
1147 af
= socket
.AF_INET
if '.' in source_address
[0] else socket
.AF_INET6
1148 ip_addrs
= [addr
for addr
in addrs
if addr
[0] == af
]
1149 if addrs
and not ip_addrs
:
1150 ip_version
= 'v4' if af
== socket
.AF_INET
else 'v6'
1152 "No remote IP%s addresses available for connect, can't use '%s' as source address"
1153 % (ip_version
, source_address
[0]))
1154 for res
in ip_addrs
:
1155 af
, socktype
, proto
, canonname
, sa
= res
1158 sock
= socket
.socket(af
, socktype
, proto
)
1159 if timeout
is not socket
._GLOBAL
_DEFAULT
_TIMEOUT
:
1160 sock
.settimeout(timeout
)
1161 sock
.bind(source_address
)
1163 err
= None # Explicitly break reference cycle
1165 except socket
.error
as _
:
1167 if sock
is not None:
1172 raise socket
.error('getaddrinfo returns an empty list')
1173 if hasattr(hc
, '_create_connection'):
1174 hc
._create
_connection
= _create_connection
1175 sa
= (source_address
, 0)
1176 if hasattr(hc
, 'source_address'): # Python 2.7+
1177 hc
.source_address
= sa
1179 def _hc_connect(self
, *args
, **kwargs
):
1180 sock
= _create_connection(
1181 (self
.host
, self
.port
), self
.timeout
, sa
)
1183 self
.sock
= ssl
.wrap_socket(
1184 sock
, self
.key_file
, self
.cert_file
,
1185 ssl_version
=ssl
.PROTOCOL_TLSv1
)
1188 hc
.connect
= functools
.partial(_hc_connect
, hc
)
1193 def handle_youtubedl_headers(headers
):
1194 filtered_headers
= headers
1196 if 'Youtubedl-no-compression' in filtered_headers
:
1197 filtered_headers
= dict((k
, v
) for k
, v
in filtered_headers
.items() if k
.lower() != 'accept-encoding')
1198 del filtered_headers
['Youtubedl-no-compression']
1200 return filtered_headers
1203 class YoutubeDLHandler(compat_urllib_request
.HTTPHandler
):
1204 """Handler for HTTP requests and responses.
1206 This class, when installed with an OpenerDirector, automatically adds
1207 the standard headers to every HTTP request and handles gzipped and
1208 deflated responses from web servers. If compression is to be avoided in
1209 a particular request, the original request in the program code only has
1210 to include the HTTP header "Youtubedl-no-compression", which will be
1211 removed before making the real request.
1213 Part of this code was copied from:
1215 http://techknack.net/python-urllib2-handlers/
1217 Andrew Rowls, the author of that code, agreed to release it to the
1221 def __init__(self
, params
, *args
, **kwargs
):
1222 compat_urllib_request
.HTTPHandler
.__init
__(self
, *args
, **kwargs
)
1223 self
._params
= params
1225 def http_open(self
, req
):
1226 conn_class
= compat_http_client
.HTTPConnection
1228 socks_proxy
= req
.headers
.get('Ytdl-socks-proxy')
1230 conn_class
= make_socks_conn_class(conn_class
, socks_proxy
)
1231 del req
.headers
['Ytdl-socks-proxy']
1233 return self
.do_open(functools
.partial(
1234 _create_http_connection
, self
, conn_class
, False),
1242 return zlib
.decompress(data
, -zlib
.MAX_WBITS
)
1244 return zlib
.decompress(data
)
1246 def http_request(self
, req
):
1247 # According to RFC 3986, URLs can not contain non-ASCII characters, however this is not
1248 # always respected by websites, some tend to give out URLs with non percent-encoded
1249 # non-ASCII characters (see telemb.py, ard.py [#3412])
1250 # urllib chokes on URLs with non-ASCII characters (see http://bugs.python.org/issue3991)
1251 # To work around aforementioned issue we will replace request's original URL with
1252 # percent-encoded one
1253 # Since redirects are also affected (e.g. http://www.southpark.de/alle-episoden/s18e09)
1254 # the code of this workaround has been moved here from YoutubeDL.urlopen()
1255 url
= req
.get_full_url()
1256 url_escaped
= escape_url(url
)
1258 # Substitute URL if any change after escaping
1259 if url
!= url_escaped
:
1260 req
= update_Request(req
, url
=url_escaped
)
1262 for h
, v
in std_headers
.items():
1263 # Capitalize is needed because of Python bug 2275: http://bugs.python.org/issue2275
1264 # The dict keys are capitalized because of this bug by urllib
1265 if h
.capitalize() not in req
.headers
:
1266 req
.add_header(h
, v
)
1268 req
.headers
= handle_youtubedl_headers(req
.headers
)
1270 if sys
.version_info
< (2, 7) and '#' in req
.get_full_url():
1271 # Python 2.6 is brain-dead when it comes to fragments
1272 req
._Request
__original
= req
._Request
__original
.partition('#')[0]
1273 req
._Request
__r
_type
= req
._Request
__r
_type
.partition('#')[0]
1277 def http_response(self
, req
, resp
):
1280 if resp
.headers
.get('Content-encoding', '') == 'gzip':
1281 content
= resp
.read()
1282 gz
= gzip
.GzipFile(fileobj
=io
.BytesIO(content
), mode
='rb')
1284 uncompressed
= io
.BytesIO(gz
.read())
1285 except IOError as original_ioerror
:
1286 # There may be junk add the end of the file
1287 # See http://stackoverflow.com/q/4928560/35070 for details
1288 for i
in range(1, 1024):
1290 gz
= gzip
.GzipFile(fileobj
=io
.BytesIO(content
[:-i
]), mode
='rb')
1291 uncompressed
= io
.BytesIO(gz
.read())
1296 raise original_ioerror
1297 resp
= compat_urllib_request
.addinfourl(uncompressed
, old_resp
.headers
, old_resp
.url
, old_resp
.code
)
1298 resp
.msg
= old_resp
.msg
1299 del resp
.headers
['Content-encoding']
1301 if resp
.headers
.get('Content-encoding', '') == 'deflate':
1302 gz
= io
.BytesIO(self
.deflate(resp
.read()))
1303 resp
= compat_urllib_request
.addinfourl(gz
, old_resp
.headers
, old_resp
.url
, old_resp
.code
)
1304 resp
.msg
= old_resp
.msg
1305 del resp
.headers
['Content-encoding']
1306 # Percent-encode redirect URL of Location HTTP header to satisfy RFC 3986 (see
1307 # https://github.com/ytdl-org/youtube-dl/issues/6457).
1308 if 300 <= resp
.code
< 400:
1309 location
= resp
.headers
.get('Location')
1311 # As of RFC 2616 default charset is iso-8859-1 that is respected by python 3
1312 if sys
.version_info
>= (3, 0):
1313 location
= location
.encode('iso-8859-1').decode('utf-8')
1315 location
= location
.decode('utf-8')
1316 location_escaped
= escape_url(location
)
1317 if location
!= location_escaped
:
1318 del resp
.headers
['Location']
1319 if sys
.version_info
< (3, 0):
1320 location_escaped
= location_escaped
.encode('utf-8')
1321 resp
.headers
['Location'] = location_escaped
1324 https_request
= http_request
1325 https_response
= http_response
1328 def make_socks_conn_class(base_class
, socks_proxy
):
1329 assert issubclass(base_class
, (
1330 compat_http_client
.HTTPConnection
, compat_http_client
.HTTPSConnection
))
1332 url_components
= compat_urlparse
.urlparse(socks_proxy
)
1333 if url_components
.scheme
.lower() == 'socks5':
1334 socks_type
= ProxyType
.SOCKS5
1335 elif url_components
.scheme
.lower() in ('socks', 'socks4'):
1336 socks_type
= ProxyType
.SOCKS4
1337 elif url_components
.scheme
.lower() == 'socks4a':
1338 socks_type
= ProxyType
.SOCKS4A
1340 def unquote_if_non_empty(s
):
1343 return compat_urllib_parse_unquote_plus(s
)
1347 url_components
.hostname
, url_components
.port
or 1080,
1349 unquote_if_non_empty(url_components
.username
),
1350 unquote_if_non_empty(url_components
.password
),
1353 class SocksConnection(base_class
):
1355 self
.sock
= sockssocket()
1356 self
.sock
.setproxy(*proxy_args
)
1357 if type(self
.timeout
) in (int, float):
1358 self
.sock
.settimeout(self
.timeout
)
1359 self
.sock
.connect((self
.host
, self
.port
))
1361 if isinstance(self
, compat_http_client
.HTTPSConnection
):
1362 if hasattr(self
, '_context'): # Python > 2.6
1363 self
.sock
= self
._context
.wrap_socket(
1364 self
.sock
, server_hostname
=self
.host
)
1366 self
.sock
= ssl
.wrap_socket(self
.sock
)
1368 return SocksConnection
1371 class YoutubeDLHTTPSHandler(compat_urllib_request
.HTTPSHandler
):
1372 def __init__(self
, params
, https_conn_class
=None, *args
, **kwargs
):
1373 compat_urllib_request
.HTTPSHandler
.__init
__(self
, *args
, **kwargs
)
1374 self
._https
_conn
_class
= https_conn_class
or compat_http_client
.HTTPSConnection
1375 self
._params
= params
1377 def https_open(self
, req
):
1379 conn_class
= self
._https
_conn
_class
1381 if hasattr(self
, '_context'): # python > 2.6
1382 kwargs
['context'] = self
._context
1383 if hasattr(self
, '_check_hostname'): # python 3.x
1384 kwargs
['check_hostname'] = self
._check
_hostname
1386 socks_proxy
= req
.headers
.get('Ytdl-socks-proxy')
1388 conn_class
= make_socks_conn_class(conn_class
, socks_proxy
)
1389 del req
.headers
['Ytdl-socks-proxy']
1391 return self
.do_open(functools
.partial(
1392 _create_http_connection
, self
, conn_class
, True),
1396 class YoutubeDLCookieJar(compat_cookiejar
.MozillaCookieJar
):
1398 See [1] for cookie file format.
1400 1. https://curl.haxx.se/docs/http-cookies.html
1402 _HTTPONLY_PREFIX
= '#HttpOnly_'
1404 _HEADER
= '''# Netscape HTTP Cookie File
1405 # This file is generated by yt-dlp. Do not edit.
1408 _CookieFileEntry
= collections
.namedtuple(
1410 ('domain_name', 'include_subdomains', 'path', 'https_only', 'expires_at', 'name', 'value'))
1412 def save(self
, filename
=None, ignore_discard
=False, ignore_expires
=False):
1414 Save cookies to a file.
1416 Most of the code is taken from CPython 3.8 and slightly adapted
1417 to support cookie files with UTF-8 in both python 2 and 3.
1419 if filename
is None:
1420 if self
.filename
is not None:
1421 filename
= self
.filename
1423 raise ValueError(compat_cookiejar
.MISSING_FILENAME_TEXT
)
1425 # Store session cookies with `expires` set to 0 instead of an empty
1428 if cookie
.expires
is None:
1431 with io
.open(filename
, 'w', encoding
='utf-8') as f
:
1432 f
.write(self
._HEADER
)
1435 if not ignore_discard
and cookie
.discard
:
1437 if not ignore_expires
and cookie
.is_expired(now
):
1443 if cookie
.domain
.startswith('.'):
1444 initial_dot
= 'TRUE'
1446 initial_dot
= 'FALSE'
1447 if cookie
.expires
is not None:
1448 expires
= compat_str(cookie
.expires
)
1451 if cookie
.value
is None:
1452 # cookies.txt regards 'Set-Cookie: foo' as a cookie
1453 # with no name, whereas http.cookiejar regards it as a
1454 # cookie with no value.
1459 value
= cookie
.value
1461 '\t'.join([cookie
.domain
, initial_dot
, cookie
.path
,
1462 secure
, expires
, name
, value
]) + '\n')
1464 def load(self
, filename
=None, ignore_discard
=False, ignore_expires
=False):
1465 """Load cookies from a file."""
1466 if filename
is None:
1467 if self
.filename
is not None:
1468 filename
= self
.filename
1470 raise ValueError(compat_cookiejar
.MISSING_FILENAME_TEXT
)
1472 def prepare_line(line
):
1473 if line
.startswith(self
._HTTPONLY
_PREFIX
):
1474 line
= line
[len(self
._HTTPONLY
_PREFIX
):]
1475 # comments and empty lines are fine
1476 if line
.startswith('#') or not line
.strip():
1478 cookie_list
= line
.split('\t')
1479 if len(cookie_list
) != self
._ENTRY
_LEN
:
1480 raise compat_cookiejar
.LoadError('invalid length %d' % len(cookie_list
))
1481 cookie
= self
._CookieFileEntry
(*cookie_list
)
1482 if cookie
.expires_at
and not cookie
.expires_at
.isdigit():
1483 raise compat_cookiejar
.LoadError('invalid expires at %s' % cookie
.expires_at
)
1487 with io
.open(filename
, encoding
='utf-8') as f
:
1490 cf
.write(prepare_line(line
))
1491 except compat_cookiejar
.LoadError
as e
:
1493 'WARNING: skipping cookie file entry due to %s: %r\n'
1494 % (e
, line
), sys
.stderr
)
1497 self
._really
_load
(cf
, filename
, ignore_discard
, ignore_expires
)
1498 # Session cookies are denoted by either `expires` field set to
1499 # an empty string or 0. MozillaCookieJar only recognizes the former
1500 # (see [1]). So we need force the latter to be recognized as session
1501 # cookies on our own.
1502 # Session cookies may be important for cookies-based authentication,
1503 # e.g. usually, when user does not check 'Remember me' check box while
1504 # logging in on a site, some important cookies are stored as session
1505 # cookies so that not recognizing them will result in failed login.
1506 # 1. https://bugs.python.org/issue17164
1508 # Treat `expires=0` cookies as session cookies
1509 if cookie
.expires
== 0:
1510 cookie
.expires
= None
1511 cookie
.discard
= True
1514 class YoutubeDLCookieProcessor(compat_urllib_request
.HTTPCookieProcessor
):
1515 def __init__(self
, cookiejar
=None):
1516 compat_urllib_request
.HTTPCookieProcessor
.__init
__(self
, cookiejar
)
1518 def http_response(self
, request
, response
):
1519 # Python 2 will choke on next HTTP request in row if there are non-ASCII
1520 # characters in Set-Cookie HTTP header of last response (see
1521 # https://github.com/ytdl-org/youtube-dl/issues/6769).
1522 # In order to at least prevent crashing we will percent encode Set-Cookie
1523 # header before HTTPCookieProcessor starts processing it.
1524 # if sys.version_info < (3, 0) and response.headers:
1525 # for set_cookie_header in ('Set-Cookie', 'Set-Cookie2'):
1526 # set_cookie = response.headers.get(set_cookie_header)
1528 # set_cookie_escaped = compat_urllib_parse.quote(set_cookie, b"%/;:@&=+$,!~*'()?#[] ")
1529 # if set_cookie != set_cookie_escaped:
1530 # del response.headers[set_cookie_header]
1531 # response.headers[set_cookie_header] = set_cookie_escaped
1532 return compat_urllib_request
.HTTPCookieProcessor
.http_response(self
, request
, response
)
1534 https_request
= compat_urllib_request
.HTTPCookieProcessor
.http_request
1535 https_response
= http_response
1538 class YoutubeDLRedirectHandler(compat_urllib_request
.HTTPRedirectHandler
):
1539 """YoutubeDL redirect handler
1541 The code is based on HTTPRedirectHandler implementation from CPython [1].
1543 This redirect handler solves two issues:
1544 - ensures redirect URL is always unicode under python 2
1545 - introduces support for experimental HTTP response status code
1546 308 Permanent Redirect [2] used by some sites [3]
1548 1. https://github.com/python/cpython/blob/master/Lib/urllib/request.py
1549 2. https://developer.mozilla.org/en-US/docs/Web/HTTP/Status/308
1550 3. https://github.com/ytdl-org/youtube-dl/issues/28768
1553 http_error_301
= http_error_303
= http_error_307
= http_error_308
= compat_urllib_request
.HTTPRedirectHandler
.http_error_302
1555 def redirect_request(self
, req
, fp
, code
, msg
, headers
, newurl
):
1556 """Return a Request or None in response to a redirect.
1558 This is called by the http_error_30x methods when a
1559 redirection response is received. If a redirection should
1560 take place, return a new Request to allow http_error_30x to
1561 perform the redirect. Otherwise, raise HTTPError if no-one
1562 else should try to handle this url. Return None if you can't
1563 but another Handler might.
1565 m
= req
.get_method()
1566 if (not (code
in (301, 302, 303, 307, 308) and m
in ("GET", "HEAD")
1567 or code
in (301, 302, 303) and m
== "POST")):
1568 raise compat_HTTPError(req
.full_url
, code
, msg
, headers
, fp
)
1569 # Strictly (according to RFC 2616), 301 or 302 in response to
1570 # a POST MUST NOT cause a redirection without confirmation
1571 # from the user (of urllib.request, in this case). In practice,
1572 # essentially all clients do redirect in this case, so we do
1575 # On python 2 urlh.geturl() may sometimes return redirect URL
1576 # as byte string instead of unicode. This workaround allows
1577 # to force it always return unicode.
1578 if sys
.version_info
[0] < 3:
1579 newurl
= compat_str(newurl
)
1581 # Be conciliant with URIs containing a space. This is mainly
1582 # redundant with the more complete encoding done in http_error_302(),
1583 # but it is kept for compatibility with other callers.
1584 newurl
= newurl
.replace(' ', '%20')
1586 CONTENT_HEADERS
= ("content-length", "content-type")
1587 # NB: don't use dict comprehension for python 2.6 compatibility
1588 newheaders
= dict((k
, v
) for k
, v
in req
.headers
.items()
1589 if k
.lower() not in CONTENT_HEADERS
)
1590 return compat_urllib_request
.Request(
1591 newurl
, headers
=newheaders
, origin_req_host
=req
.origin_req_host
,
1595 def extract_timezone(date_str
):
1598 ^.{8,}? # >=8 char non-TZ prefix, if present
1599 (?P<tz>Z| # just the UTC Z, or
1600 (?:(?<=.\b\d{4}|\b\d{2}:\d\d)| # preceded by 4 digits or hh:mm or
1601 (?<!.\b[a-zA-Z]{3}|[a-zA-Z]{4}|..\b\d\d)) # not preceded by 3 alpha word or >= 4 alpha or 2 digits
1602 [ ]? # optional space
1603 (?P<sign>\+|-) # +/-
1604 (?P<hours>[0-9]{2}):?(?P<minutes>[0-9]{2}) # hh[:]mm
1608 timezone
= datetime
.timedelta()
1610 date_str
= date_str
[:-len(m
.group('tz'))]
1611 if not m
.group('sign'):
1612 timezone
= datetime
.timedelta()
1614 sign
= 1 if m
.group('sign') == '+' else -1
1615 timezone
= datetime
.timedelta(
1616 hours
=sign
* int(m
.group('hours')),
1617 minutes
=sign
* int(m
.group('minutes')))
1618 return timezone
, date_str
1621 def parse_iso8601(date_str
, delimiter
='T', timezone
=None):
1622 """ Return a UNIX timestamp from the given date """
1624 if date_str
is None:
1627 date_str
= re
.sub(r
'\.[0-9]+', '', date_str
)
1629 if timezone
is None:
1630 timezone
, date_str
= extract_timezone(date_str
)
1633 date_format
= '%Y-%m-%d{0}%H:%M:%S'.format(delimiter
)
1634 dt
= datetime
.datetime
.strptime(date_str
, date_format
) - timezone
1635 return calendar
.timegm(dt
.timetuple())
1640 def date_formats(day_first
=True):
1641 return DATE_FORMATS_DAY_FIRST
if day_first
else DATE_FORMATS_MONTH_FIRST
1644 def unified_strdate(date_str
, day_first
=True):
1645 """Return a string with the date in the format YYYYMMDD"""
1647 if date_str
is None:
1651 date_str
= date_str
.replace(',', ' ')
1652 # Remove AM/PM + timezone
1653 date_str
= re
.sub(r
'(?i)\s*(?:AM|PM)(?:\s+[A-Z]+)?', '', date_str
)
1654 _
, date_str
= extract_timezone(date_str
)
1656 for expression
in date_formats(day_first
):
1658 upload_date
= datetime
.datetime
.strptime(date_str
, expression
).strftime('%Y%m%d')
1661 if upload_date
is None:
1662 timetuple
= email
.utils
.parsedate_tz(date_str
)
1665 upload_date
= datetime
.datetime(*timetuple
[:6]).strftime('%Y%m%d')
1668 if upload_date
is not None:
1669 return compat_str(upload_date
)
1672 def unified_timestamp(date_str
, day_first
=True):
1673 if date_str
is None:
1676 date_str
= re
.sub(r
'[,|]', '', date_str
)
1678 pm_delta
= 12 if re
.search(r
'(?i)PM', date_str
) else 0
1679 timezone
, date_str
= extract_timezone(date_str
)
1681 # Remove AM/PM + timezone
1682 date_str
= re
.sub(r
'(?i)\s*(?:AM|PM)(?:\s+[A-Z]+)?', '', date_str
)
1684 # Remove unrecognized timezones from ISO 8601 alike timestamps
1685 m
= re
.search(r
'\d{1,2}:\d{1,2}(?:\.\d+)?(?P<tz>\s*[A-Z]+)$', date_str
)
1687 date_str
= date_str
[:-len(m
.group('tz'))]
1689 # Python only supports microseconds, so remove nanoseconds
1690 m
= re
.search(r
'^([0-9]{4,}-[0-9]{1,2}-[0-9]{1,2}T[0-9]{1,2}:[0-9]{1,2}:[0-9]{1,2}\.[0-9]{6})[0-9]+$', date_str
)
1692 date_str
= m
.group(1)
1694 for expression
in date_formats(day_first
):
1696 dt
= datetime
.datetime
.strptime(date_str
, expression
) - timezone
+ datetime
.timedelta(hours
=pm_delta
)
1697 return calendar
.timegm(dt
.timetuple())
1700 timetuple
= email
.utils
.parsedate_tz(date_str
)
1702 return calendar
.timegm(timetuple
) + pm_delta
* 3600
1705 def determine_ext(url
, default_ext
='unknown_video'):
1706 if url
is None or '.' not in url
:
1708 guess
= url
.partition('?')[0].rpartition('.')[2]
1709 if re
.match(r
'^[A-Za-z0-9]+$', guess
):
1711 # Try extract ext from URLs like http://example.com/foo/bar.mp4/?download
1712 elif guess
.rstrip('/') in KNOWN_EXTENSIONS
:
1713 return guess
.rstrip('/')
1718 def subtitles_filename(filename
, sub_lang
, sub_format
, expected_real_ext
=None):
1719 return replace_extension(filename
, sub_lang
+ '.' + sub_format
, expected_real_ext
)
1722 def datetime_from_str(date_str
, precision
='auto', format
='%Y%m%d'):
1724 Return a datetime object from a string in the format YYYYMMDD or
1725 (now|today|date)[+-][0-9](microsecond|second|minute|hour|day|week|month|year)(s)?
1727 format: string date format used to return datetime object from
1728 precision: round the time portion of a datetime object.
1729 auto|microsecond|second|minute|hour|day.
1730 auto: round to the unit provided in date_str (if applicable).
1732 auto_precision
= False
1733 if precision
== 'auto':
1734 auto_precision
= True
1735 precision
= 'microsecond'
1736 today
= datetime_round(datetime
.datetime
.now(), precision
)
1737 if date_str
in ('now', 'today'):
1739 if date_str
== 'yesterday':
1740 return today
- datetime
.timedelta(days
=1)
1742 r
'(?P<start>.+)(?P<sign>[+-])(?P<time>\d+)(?P<unit>microsecond|second|minute|hour|day|week|month|year)(s)?',
1744 if match
is not None:
1745 start_time
= datetime_from_str(match
.group('start'), precision
, format
)
1746 time
= int(match
.group('time')) * (-1 if match
.group('sign') == '-' else 1)
1747 unit
= match
.group('unit')
1748 if unit
== 'month' or unit
== 'year':
1749 new_date
= datetime_add_months(start_time
, time
* 12 if unit
== 'year' else time
)
1755 delta
= datetime
.timedelta(**{unit + 's': time}
)
1756 new_date
= start_time
+ delta
1758 return datetime_round(new_date
, unit
)
1761 return datetime_round(datetime
.datetime
.strptime(date_str
, format
), precision
)
1764 def date_from_str(date_str
, format
='%Y%m%d'):
1766 Return a datetime object from a string in the format YYYYMMDD or
1767 (now|today|date)[+-][0-9](microsecond|second|minute|hour|day|week|month|year)(s)?
1769 format: string date format used to return datetime object from
1771 return datetime_from_str(date_str
, precision
='microsecond', format
=format
).date()
1774 def datetime_add_months(dt
, months
):
1775 """Increment/Decrement a datetime object by months."""
1776 month
= dt
.month
+ months
- 1
1777 year
= dt
.year
+ month
// 12
1778 month
= month
% 12 + 1
1779 day
= min(dt
.day
, calendar
.monthrange(year
, month
)[1])
1780 return dt
.replace(year
, month
, day
)
1783 def datetime_round(dt
, precision
='day'):
1785 Round a datetime object's time to a specific precision
1787 if precision
== 'microsecond':
1796 roundto
= lambda x
, n
: ((x
+ n
/ 2) // n
) * n
1797 timestamp
= calendar
.timegm(dt
.timetuple())
1798 return datetime
.datetime
.utcfromtimestamp(roundto(timestamp
, unit_seconds
[precision
]))
1801 def hyphenate_date(date_str
):
1803 Convert a date in 'YYYYMMDD' format to 'YYYY-MM-DD' format"""
1804 match
= re
.match(r
'^(\d\d\d\d)(\d\d)(\d\d)$', date_str
)
1805 if match
is not None:
1806 return '-'.join(match
.groups())
1811 class DateRange(object):
1812 """Represents a time interval between two dates"""
1814 def __init__(self
, start
=None, end
=None):
1815 """start and end must be strings in the format accepted by date"""
1816 if start
is not None:
1817 self
.start
= date_from_str(start
)
1819 self
.start
= datetime
.datetime
.min.date()
1821 self
.end
= date_from_str(end
)
1823 self
.end
= datetime
.datetime
.max.date()
1824 if self
.start
> self
.end
:
1825 raise ValueError('Date range: "%s" , the start date must be before the end date' % self
)
1829 """Returns a range that only contains the given day"""
1830 return cls(day
, day
)
1832 def __contains__(self
, date
):
1833 """Check if the date is in the range"""
1834 if not isinstance(date
, datetime
.date
):
1835 date
= date_from_str(date
)
1836 return self
.start
<= date
<= self
.end
1839 return '%s - %s' % (self
.start
.isoformat(), self
.end
.isoformat())
1842 def platform_name():
1843 """ Returns the platform name as a compat_str """
1844 res
= platform
.platform()
1845 if isinstance(res
, bytes):
1846 res
= res
.decode(preferredencoding())
1848 assert isinstance(res
, compat_str
)
1852 def get_windows_version():
1853 ''' Get Windows version. None if it's not running on Windows '''
1854 if compat_os_name
== 'nt':
1855 return version_tuple(platform
.win32_ver()[1])
1860 def _windows_write_string(s
, out
):
1861 """ Returns True if the string was written using special methods,
1862 False if it has yet to be written out."""
1863 # Adapted from http://stackoverflow.com/a/3259271/35070
1865 import ctypes
.wintypes
1873 fileno
= out
.fileno()
1874 except AttributeError:
1875 # If the output stream doesn't have a fileno, it's virtual
1877 except io
.UnsupportedOperation
:
1878 # Some strange Windows pseudo files?
1880 if fileno
not in WIN_OUTPUT_IDS
:
1883 GetStdHandle
= compat_ctypes_WINFUNCTYPE(
1884 ctypes
.wintypes
.HANDLE
, ctypes
.wintypes
.DWORD
)(
1885 ('GetStdHandle', ctypes
.windll
.kernel32
))
1886 h
= GetStdHandle(WIN_OUTPUT_IDS
[fileno
])
1888 WriteConsoleW
= compat_ctypes_WINFUNCTYPE(
1889 ctypes
.wintypes
.BOOL
, ctypes
.wintypes
.HANDLE
, ctypes
.wintypes
.LPWSTR
,
1890 ctypes
.wintypes
.DWORD
, ctypes
.POINTER(ctypes
.wintypes
.DWORD
),
1891 ctypes
.wintypes
.LPVOID
)(('WriteConsoleW', ctypes
.windll
.kernel32
))
1892 written
= ctypes
.wintypes
.DWORD(0)
1894 GetFileType
= compat_ctypes_WINFUNCTYPE(ctypes
.wintypes
.DWORD
, ctypes
.wintypes
.DWORD
)(('GetFileType', ctypes
.windll
.kernel32
))
1895 FILE_TYPE_CHAR
= 0x0002
1896 FILE_TYPE_REMOTE
= 0x8000
1897 GetConsoleMode
= compat_ctypes_WINFUNCTYPE(
1898 ctypes
.wintypes
.BOOL
, ctypes
.wintypes
.HANDLE
,
1899 ctypes
.POINTER(ctypes
.wintypes
.DWORD
))(
1900 ('GetConsoleMode', ctypes
.windll
.kernel32
))
1901 INVALID_HANDLE_VALUE
= ctypes
.wintypes
.DWORD(-1).value
1903 def not_a_console(handle
):
1904 if handle
== INVALID_HANDLE_VALUE
or handle
is None:
1906 return ((GetFileType(handle
) & ~FILE_TYPE_REMOTE
) != FILE_TYPE_CHAR
1907 or GetConsoleMode(handle
, ctypes
.byref(ctypes
.wintypes
.DWORD())) == 0)
1909 if not_a_console(h
):
1912 def next_nonbmp_pos(s
):
1914 return next(i
for i
, c
in enumerate(s
) if ord(c
) > 0xffff)
1915 except StopIteration:
1919 count
= min(next_nonbmp_pos(s
), 1024)
1921 ret
= WriteConsoleW(
1922 h
, s
, count
if count
else 2, ctypes
.byref(written
), None)
1924 raise OSError('Failed to write string')
1925 if not count
: # We just wrote a non-BMP character
1926 assert written
.value
== 2
1929 assert written
.value
> 0
1930 s
= s
[written
.value
:]
1934 def write_string(s
, out
=None, encoding
=None):
1937 assert type(s
) == compat_str
1939 if sys
.platform
== 'win32' and encoding
is None and hasattr(out
, 'fileno'):
1940 if _windows_write_string(s
, out
):
1943 if ('b' in getattr(out
, 'mode', '')
1944 or sys
.version_info
[0] < 3): # Python 2 lies about mode of sys.stderr
1945 byt
= s
.encode(encoding
or preferredencoding(), 'ignore')
1947 elif hasattr(out
, 'buffer'):
1948 enc
= encoding
or getattr(out
, 'encoding', None) or preferredencoding()
1949 byt
= s
.encode(enc
, 'ignore')
1950 out
.buffer.write(byt
)
1956 def bytes_to_intlist(bs
):
1959 if isinstance(bs
[0], int): # Python 3
1962 return [ord(c
) for c
in bs
]
1965 def intlist_to_bytes(xs
):
1968 return compat_struct_pack('%dB' % len(xs
), *xs
)
1971 # Cross-platform file locking
1972 if sys
.platform
== 'win32':
1973 import ctypes
.wintypes
1976 class OVERLAPPED(ctypes
.Structure
):
1978 ('Internal', ctypes
.wintypes
.LPVOID
),
1979 ('InternalHigh', ctypes
.wintypes
.LPVOID
),
1980 ('Offset', ctypes
.wintypes
.DWORD
),
1981 ('OffsetHigh', ctypes
.wintypes
.DWORD
),
1982 ('hEvent', ctypes
.wintypes
.HANDLE
),
1985 kernel32
= ctypes
.windll
.kernel32
1986 LockFileEx
= kernel32
.LockFileEx
1987 LockFileEx
.argtypes
= [
1988 ctypes
.wintypes
.HANDLE
, # hFile
1989 ctypes
.wintypes
.DWORD
, # dwFlags
1990 ctypes
.wintypes
.DWORD
, # dwReserved
1991 ctypes
.wintypes
.DWORD
, # nNumberOfBytesToLockLow
1992 ctypes
.wintypes
.DWORD
, # nNumberOfBytesToLockHigh
1993 ctypes
.POINTER(OVERLAPPED
) # Overlapped
1995 LockFileEx
.restype
= ctypes
.wintypes
.BOOL
1996 UnlockFileEx
= kernel32
.UnlockFileEx
1997 UnlockFileEx
.argtypes
= [
1998 ctypes
.wintypes
.HANDLE
, # hFile
1999 ctypes
.wintypes
.DWORD
, # dwReserved
2000 ctypes
.wintypes
.DWORD
, # nNumberOfBytesToLockLow
2001 ctypes
.wintypes
.DWORD
, # nNumberOfBytesToLockHigh
2002 ctypes
.POINTER(OVERLAPPED
) # Overlapped
2004 UnlockFileEx
.restype
= ctypes
.wintypes
.BOOL
2005 whole_low
= 0xffffffff
2006 whole_high
= 0x7fffffff
2008 def _lock_file(f
, exclusive
):
2009 overlapped
= OVERLAPPED()
2010 overlapped
.Offset
= 0
2011 overlapped
.OffsetHigh
= 0
2012 overlapped
.hEvent
= 0
2013 f
._lock
_file
_overlapped
_p
= ctypes
.pointer(overlapped
)
2014 handle
= msvcrt
.get_osfhandle(f
.fileno())
2015 if not LockFileEx(handle
, 0x2 if exclusive
else 0x0, 0,
2016 whole_low
, whole_high
, f
._lock
_file
_overlapped
_p
):
2017 raise OSError('Locking file failed: %r' % ctypes
.FormatError())
2019 def _unlock_file(f
):
2020 assert f
._lock
_file
_overlapped
_p
2021 handle
= msvcrt
.get_osfhandle(f
.fileno())
2022 if not UnlockFileEx(handle
, 0,
2023 whole_low
, whole_high
, f
._lock
_file
_overlapped
_p
):
2024 raise OSError('Unlocking file failed: %r' % ctypes
.FormatError())
2027 # Some platforms, such as Jython, is missing fcntl
2031 def _lock_file(f
, exclusive
):
2032 fcntl
.flock(f
, fcntl
.LOCK_EX
if exclusive
else fcntl
.LOCK_SH
)
2034 def _unlock_file(f
):
2035 fcntl
.flock(f
, fcntl
.LOCK_UN
)
2037 UNSUPPORTED_MSG
= 'file locking is not supported on this platform'
2039 def _lock_file(f
, exclusive
):
2040 raise IOError(UNSUPPORTED_MSG
)
2042 def _unlock_file(f
):
2043 raise IOError(UNSUPPORTED_MSG
)
2046 class locked_file(object):
2047 def __init__(self
, filename
, mode
, encoding
=None):
2048 assert mode
in ['r', 'a', 'w']
2049 self
.f
= io
.open(filename
, mode
, encoding
=encoding
)
2052 def __enter__(self
):
2053 exclusive
= self
.mode
!= 'r'
2055 _lock_file(self
.f
, exclusive
)
2061 def __exit__(self
, etype
, value
, traceback
):
2063 _unlock_file(self
.f
)
2070 def write(self
, *args
):
2071 return self
.f
.write(*args
)
2073 def read(self
, *args
):
2074 return self
.f
.read(*args
)
2077 def get_filesystem_encoding():
2078 encoding
= sys
.getfilesystemencoding()
2079 return encoding
if encoding
is not None else 'utf-8'
2082 def shell_quote(args
):
2084 encoding
= get_filesystem_encoding()
2086 if isinstance(a
, bytes):
2087 # We may get a filename encoded with 'encodeFilename'
2088 a
= a
.decode(encoding
)
2089 quoted_args
.append(compat_shlex_quote(a
))
2090 return ' '.join(quoted_args
)
2093 def smuggle_url(url
, data
):
2094 """ Pass additional data in a URL for internal use. """
2096 url
, idata
= unsmuggle_url(url
, {})
2098 sdata
= compat_urllib_parse_urlencode(
2099 {'__youtubedl_smuggle': json.dumps(data)}
)
2100 return url
+ '#' + sdata
2103 def unsmuggle_url(smug_url
, default
=None):
2104 if '#__youtubedl_smuggle' not in smug_url
:
2105 return smug_url
, default
2106 url
, _
, sdata
= smug_url
.rpartition('#')
2107 jsond
= compat_parse_qs(sdata
)['__youtubedl_smuggle'][0]
2108 data
= json
.loads(jsond
)
2112 def format_decimal_suffix(num
, fmt
='%d%s', *, factor
=1000):
2113 """ Formats numbers with decimal sufixes like K, M, etc """
2114 num
, factor
= float_or_none(num
), float(factor
)
2117 exponent
= 0 if num
== 0 else int(math
.log(num
, factor
))
2118 suffix
= ['', *'KMGTPEZY'][exponent
]
2119 converted
= num
/ (factor
** exponent
)
2120 return fmt
% (converted
, suffix
)
2123 def format_bytes(bytes):
2124 return format_decimal_suffix(bytes, '%.2f%siB', factor
=1024) or 'N/A'
2127 def lookup_unit_table(unit_table
, s
):
2128 units_re
= '|'.join(re
.escape(u
) for u
in unit_table
)
2130 r
'(?P<num>[0-9]+(?:[,.][0-9]*)?)\s*(?P<unit>%s)\b' % units_re
, s
)
2133 num_str
= m
.group('num').replace(',', '.')
2134 mult
= unit_table
[m
.group('unit')]
2135 return int(float(num_str
) * mult
)
2138 def parse_filesize(s
):
2142 # The lower-case forms are of course incorrect and unofficial,
2143 # but we support those too
2160 'megabytes': 1000 ** 2,
2161 'mebibytes': 1024 ** 2,
2167 'gigabytes': 1000 ** 3,
2168 'gibibytes': 1024 ** 3,
2174 'terabytes': 1000 ** 4,
2175 'tebibytes': 1024 ** 4,
2181 'petabytes': 1000 ** 5,
2182 'pebibytes': 1024 ** 5,
2188 'exabytes': 1000 ** 6,
2189 'exbibytes': 1024 ** 6,
2195 'zettabytes': 1000 ** 7,
2196 'zebibytes': 1024 ** 7,
2202 'yottabytes': 1000 ** 8,
2203 'yobibytes': 1024 ** 8,
2206 return lookup_unit_table(_UNIT_TABLE
, s
)
2215 if re
.match(r
'^[\d,.]+$', s
):
2216 return str_to_int(s
)
2227 return lookup_unit_table(_UNIT_TABLE
, s
)
2230 def parse_resolution(s
):
2234 mobj
= re
.search(r
'(?<![a-zA-Z0-9])(?P<w>\d+)\s*[xX×,]\s*(?P<h>\d+)(?![a-zA-Z0-9])', s
)
2237 'width': int(mobj
.group('w')),
2238 'height': int(mobj
.group('h')),
2241 mobj
= re
.search(r
'(?<![a-zA-Z0-9])(\d+)[pPiI](?![a-zA-Z0-9])', s
)
2243 return {'height': int(mobj.group(1))}
2245 mobj
= re
.search(r
'\b([48])[kK]\b', s
)
2247 return {'height': int(mobj.group(1)) * 540}
2252 def parse_bitrate(s
):
2253 if not isinstance(s
, compat_str
):
2255 mobj
= re
.search(r
'\b(\d+)\s*kbps', s
)
2257 return int(mobj
.group(1))
2260 def month_by_name(name
, lang
='en'):
2261 """ Return the number of a month by (locale-independently) English name """
2263 month_names
= MONTH_NAMES
.get(lang
, MONTH_NAMES
['en'])
2266 return month_names
.index(name
) + 1
2271 def month_by_abbreviation(abbrev
):
2272 """ Return the number of a month by (locale-independently) English
2276 return [s
[:3] for s
in ENGLISH_MONTH_NAMES
].index(abbrev
) + 1
2281 def fix_xml_ampersands(xml_str
):
2282 """Replace all the '&' by '&' in XML"""
2284 r
'&(?!amp;|lt;|gt;|apos;|quot;|#x[0-9a-fA-F]{,4};|#[0-9]{,4};)',
2289 def setproctitle(title
):
2290 assert isinstance(title
, compat_str
)
2292 # ctypes in Jython is not complete
2293 # http://bugs.jython.org/issue2148
2294 if sys
.platform
.startswith('java'):
2298 libc
= ctypes
.cdll
.LoadLibrary('libc.so.6')
2302 # LoadLibrary in Windows Python 2.7.13 only expects
2303 # a bytestring, but since unicode_literals turns
2304 # every string into a unicode string, it fails.
2306 title_bytes
= title
.encode('utf-8')
2307 buf
= ctypes
.create_string_buffer(len(title_bytes
))
2308 buf
.value
= title_bytes
2310 libc
.prctl(15, buf
, 0, 0, 0)
2311 except AttributeError:
2312 return # Strange libc, just skip this
2315 def remove_start(s
, start
):
2316 return s
[len(start
):] if s
is not None and s
.startswith(start
) else s
2319 def remove_end(s
, end
):
2320 return s
[:-len(end
)] if s
is not None and s
.endswith(end
) else s
2323 def remove_quotes(s
):
2324 if s
is None or len(s
) < 2:
2326 for quote
in ('"', "'", ):
2327 if s
[0] == quote
and s
[-1] == quote
:
2332 def get_domain(url
):
2333 domain
= re
.match(r
'(?:https?:\/\/)?(?:www\.)?(?P<domain>[^\n\/]+\.[^\n\/]+)(?:\/(.*))?', url
)
2334 return domain
.group('domain') if domain
else None
2337 def url_basename(url
):
2338 path
= compat_urlparse
.urlparse(url
).path
2339 return path
.strip('/').split('/')[-1]
2343 return re
.match(r
'https?://[^?#&]+/', url
).group()
2346 def urljoin(base
, path
):
2347 if isinstance(path
, bytes):
2348 path
= path
.decode('utf-8')
2349 if not isinstance(path
, compat_str
) or not path
:
2351 if re
.match(r
'^(?:[a-zA-Z][a-zA-Z0-9+-.]*:)?//', path
):
2353 if isinstance(base
, bytes):
2354 base
= base
.decode('utf-8')
2355 if not isinstance(base
, compat_str
) or not re
.match(
2356 r
'^(?:https?:)?//', base
):
2358 return compat_urlparse
.urljoin(base
, path
)
2361 class HEADRequest(compat_urllib_request
.Request
):
2362 def get_method(self
):
2366 class PUTRequest(compat_urllib_request
.Request
):
2367 def get_method(self
):
2371 def int_or_none(v
, scale
=1, default
=None, get_attr
=None, invscale
=1):
2374 v
= getattr(v
, get_attr
, None)
2380 return int(v
) * invscale
// scale
2381 except (ValueError, TypeError, OverflowError):
2385 def str_or_none(v
, default
=None):
2386 return default
if v
is None else compat_str(v
)
2389 def str_to_int(int_str
):
2390 """ A more relaxed version of int_or_none """
2391 if isinstance(int_str
, compat_integer_types
):
2393 elif isinstance(int_str
, compat_str
):
2394 int_str
= re
.sub(r
'[,\.\+]', '', int_str
)
2395 return int_or_none(int_str
)
2398 def float_or_none(v
, scale
=1, invscale
=1, default
=None):
2402 return float(v
) * invscale
/ scale
2403 except (ValueError, TypeError):
2407 def bool_or_none(v
, default
=None):
2408 return v
if isinstance(v
, bool) else default
2411 def strip_or_none(v
, default
=None):
2412 return v
.strip() if isinstance(v
, compat_str
) else default
2415 def url_or_none(url
):
2416 if not url
or not isinstance(url
, compat_str
):
2419 return url
if re
.match(r
'^(?:(?:https?|rt(?:m(?:pt?[es]?|fp)|sp[su]?)|mms|ftps?):)?//', url
) else None
2422 def strftime_or_none(timestamp
, date_format
, default
=None):
2423 datetime_object
= None
2425 if isinstance(timestamp
, compat_numeric_types
): # unix timestamp
2426 datetime_object
= datetime
.datetime
.utcfromtimestamp(timestamp
)
2427 elif isinstance(timestamp
, compat_str
): # assume YYYYMMDD
2428 datetime_object
= datetime
.datetime
.strptime(timestamp
, '%Y%m%d')
2429 return datetime_object
.strftime(date_format
)
2430 except (ValueError, TypeError, AttributeError):
2434 def parse_duration(s
):
2435 if not isinstance(s
, compat_basestring
):
2441 days
, hours
, mins
, secs
, ms
= [None] * 5
2442 m
= re
.match(r
'(?:(?:(?:(?P<days>[0-9]+):)?(?P<hours>[0-9]+):)?(?P<mins>[0-9]+):)?(?P<secs>[0-9]+)(?P<ms>\.[0-9]+)?Z?$', s
)
2444 days
, hours
, mins
, secs
, ms
= m
.groups()
2449 [0-9]+\s*y(?:ears?)?\s*
2452 [0-9]+\s*m(?:onths?)?\s*
2455 [0-9]+\s*w(?:eeks?)?\s*
2458 (?P<days>[0-9]+)\s*d(?:ays?)?\s*
2462 (?P<hours>[0-9]+)\s*h(?:ours?)?\s*
2465 (?P<mins>[0-9]+)\s*m(?:in(?:ute)?s?)?\s*
2468 (?P<secs>[0-9]+)(?P<ms>\.[0-9]+)?\s*s(?:ec(?:ond)?s?)?\s*
2471 days
, hours
, mins
, secs
, ms
= m
.groups()
2473 m
= re
.match(r
'(?i)(?:(?P<hours>[0-9.]+)\s*(?:hours?)|(?P<mins>[0-9.]+)\s*(?:mins?\.?|minutes?)\s*)Z?$', s
)
2475 hours
, mins
= m
.groups()
2481 duration
+= float(secs
)
2483 duration
+= float(mins
) * 60
2485 duration
+= float(hours
) * 60 * 60
2487 duration
+= float(days
) * 24 * 60 * 60
2489 duration
+= float(ms
)
2493 def prepend_extension(filename
, ext
, expected_real_ext
=None):
2494 name
, real_ext
= os
.path
.splitext(filename
)
2496 '{0}.{1}{2}'.format(name
, ext
, real_ext
)
2497 if not expected_real_ext
or real_ext
[1:] == expected_real_ext
2498 else '{0}.{1}'.format(filename
, ext
))
2501 def replace_extension(filename
, ext
, expected_real_ext
=None):
2502 name
, real_ext
= os
.path
.splitext(filename
)
2503 return '{0}.{1}'.format(
2504 name
if not expected_real_ext
or real_ext
[1:] == expected_real_ext
else filename
,
2508 def check_executable(exe
, args
=[]):
2509 """ Checks if the given binary is installed somewhere in PATH, and returns its name.
2510 args can be a list of arguments for a short output (like -version) """
2512 Popen([exe
] + args
, stdout
=subprocess
.PIPE
, stderr
=subprocess
.PIPE
).communicate_or_kill()
2518 def _get_exe_version_output(exe
, args
):
2520 # STDIN should be redirected too. On UNIX-like systems, ffmpeg triggers
2521 # SIGTTOU if yt-dlp is run in the background.
2522 # See https://github.com/ytdl-org/youtube-dl/issues/955#issuecomment-209789656
2524 [encodeArgument(exe
)] + args
, stdin
=subprocess
.PIPE
,
2525 stdout
=subprocess
.PIPE
, stderr
=subprocess
.STDOUT
).communicate_or_kill()
2528 if isinstance(out
, bytes): # Python 2.x
2529 out
= out
.decode('ascii', 'ignore')
2533 def detect_exe_version(output
, version_re
=None, unrecognized
='present'):
2534 assert isinstance(output
, compat_str
)
2535 if version_re
is None:
2536 version_re
= r
'version\s+([-0-9._a-zA-Z]+)'
2537 m
= re
.search(version_re
, output
)
2544 def get_exe_version(exe
, args
=['--version'],
2545 version_re
=None, unrecognized
='present'):
2546 """ Returns the version of the specified executable,
2547 or False if the executable is not present """
2548 out
= _get_exe_version_output(exe
, args
)
2549 return detect_exe_version(out
, version_re
, unrecognized
) if out
else False
2552 class LazyList(collections
.abc
.Sequence
):
2553 ''' Lazy immutable list from an iterable
2554 Note that slices of a LazyList are lists and not LazyList'''
2556 class IndexError(IndexError):
2559 def __init__(self
, iterable
, *, reverse
=False, _cache
=None):
2560 self
.__iterable
= iter(iterable
)
2561 self
.__cache
= [] if _cache
is None else _cache
2562 self
.__reversed
= reverse
2566 # We need to consume the entire iterable to iterate in reverse
2567 yield from self
.exhaust()
2569 yield from self
.__cache
2570 for item
in self
.__iterable
:
2571 self
.__cache
.append(item
)
2574 def __exhaust(self
):
2575 self
.__cache
.extend(self
.__iterable
)
2576 # Discard the emptied iterable to make it pickle-able
2577 self
.__iterable
= []
2581 ''' Evaluate the entire iterable '''
2582 return self
.__exhaust
()[::-1 if self
.__reversed
else 1]
2585 def __reverse_index(x
):
2586 return None if x
is None else -(x
+ 1)
2588 def __getitem__(self
, idx
):
2589 if isinstance(idx
, slice):
2591 idx
= slice(self
.__reverse
_index
(idx
.start
), self
.__reverse
_index
(idx
.stop
), -(idx
.step
or 1))
2592 start
, stop
, step
= idx
.start
, idx
.stop
, idx
.step
or 1
2593 elif isinstance(idx
, int):
2595 idx
= self
.__reverse
_index
(idx
)
2596 start
, stop
, step
= idx
, idx
, 0
2598 raise TypeError('indices must be integers or slices')
2599 if ((start
or 0) < 0 or (stop
or 0) < 0
2600 or (start
is None and step
< 0)
2601 or (stop
is None and step
> 0)):
2602 # We need to consume the entire iterable to be able to slice from the end
2603 # Obviously, never use this with infinite iterables
2606 return self
.__cache
[idx
]
2607 except IndexError as e
:
2608 raise self
.IndexError(e
) from e
2609 n
= max(start
or 0, stop
or 0) - len(self
.__cache
) + 1
2611 self
.__cache
.extend(itertools
.islice(self
.__iterable
, n
))
2613 return self
.__cache
[idx
]
2614 except IndexError as e
:
2615 raise self
.IndexError(e
) from e
2619 self
[-1] if self
.__reversed
else self
[0]
2620 except self
.IndexError:
2626 return len(self
.__cache
)
2628 def __reversed__(self
):
2629 return type(self
)(self
.__iterable
, reverse
=not self
.__reversed
, _cache
=self
.__cache
)
2632 return type(self
)(self
.__iterable
, reverse
=self
.__reversed
, _cache
=self
.__cache
)
2635 # repr and str should mimic a list. So we exhaust the iterable
2636 return repr(self
.exhaust())
2639 return repr(self
.exhaust())
2644 class IndexError(IndexError):
2648 # This is only useful for tests
2649 return len(self
.getslice())
2651 def __init__(self
, pagefunc
, pagesize
, use_cache
=True):
2652 self
._pagefunc
= pagefunc
2653 self
._pagesize
= pagesize
2654 self
._use
_cache
= use_cache
2657 def getpage(self
, pagenum
):
2658 page_results
= self
._cache
.get(pagenum
)
2659 if page_results
is None:
2660 page_results
= list(self
._pagefunc
(pagenum
))
2662 self
._cache
[pagenum
] = page_results
2665 def getslice(self
, start
=0, end
=None):
2666 return list(self
._getslice
(start
, end
))
2668 def _getslice(self
, start
, end
):
2669 raise NotImplementedError('This method must be implemented by subclasses')
2671 def __getitem__(self
, idx
):
2672 # NOTE: cache must be enabled if this is used
2673 if not isinstance(idx
, int) or idx
< 0:
2674 raise TypeError('indices must be non-negative integers')
2675 entries
= self
.getslice(idx
, idx
+ 1)
2677 raise self
.IndexError()
2681 class OnDemandPagedList(PagedList
):
2682 def _getslice(self
, start
, end
):
2683 for pagenum
in itertools
.count(start
// self
._pagesize
):
2684 firstid
= pagenum
* self
._pagesize
2685 nextfirstid
= pagenum
* self
._pagesize
+ self
._pagesize
2686 if start
>= nextfirstid
:
2690 start
% self
._pagesize
2691 if firstid
<= start
< nextfirstid
2694 ((end
- 1) % self
._pagesize
) + 1
2695 if (end
is not None and firstid
<= end
<= nextfirstid
)
2698 page_results
= self
.getpage(pagenum
)
2699 if startv
!= 0 or endv
is not None:
2700 page_results
= page_results
[startv
:endv
]
2701 yield from page_results
2703 # A little optimization - if current page is not "full", ie. does
2704 # not contain page_size videos then we can assume that this page
2705 # is the last one - there are no more ids on further pages -
2706 # i.e. no need to query again.
2707 if len(page_results
) + startv
< self
._pagesize
:
2710 # If we got the whole page, but the next page is not interesting,
2711 # break out early as well
2712 if end
== nextfirstid
:
2716 class InAdvancePagedList(PagedList
):
2717 def __init__(self
, pagefunc
, pagecount
, pagesize
):
2718 self
._pagecount
= pagecount
2719 PagedList
.__init
__(self
, pagefunc
, pagesize
, True)
2721 def _getslice(self
, start
, end
):
2722 start_page
= start
// self
._pagesize
2724 self
._pagecount
if end
is None else (end
// self
._pagesize
+ 1))
2725 skip_elems
= start
- start_page
* self
._pagesize
2726 only_more
= None if end
is None else end
- start
2727 for pagenum
in range(start_page
, end_page
):
2728 page_results
= self
.getpage(pagenum
)
2730 page_results
= page_results
[skip_elems
:]
2732 if only_more
is not None:
2733 if len(page_results
) < only_more
:
2734 only_more
-= len(page_results
)
2736 yield from page_results
[:only_more
]
2738 yield from page_results
2741 def uppercase_escape(s
):
2742 unicode_escape
= codecs
.getdecoder('unicode_escape')
2744 r
'\\U[0-9a-fA-F]{8}',
2745 lambda m
: unicode_escape(m
.group(0))[0],
2749 def lowercase_escape(s
):
2750 unicode_escape
= codecs
.getdecoder('unicode_escape')
2752 r
'\\u[0-9a-fA-F]{4}',
2753 lambda m
: unicode_escape(m
.group(0))[0],
2757 def escape_rfc3986(s
):
2758 """Escape non-ASCII characters as suggested by RFC 3986"""
2759 if sys
.version_info
< (3, 0) and isinstance(s
, compat_str
):
2760 s
= s
.encode('utf-8')
2761 return compat_urllib_parse
.quote(s
, b
"%/;:@&=+$,!~*'()?#[]")
2764 def escape_url(url
):
2765 """Escape URL as suggested by RFC 3986"""
2766 url_parsed
= compat_urllib_parse_urlparse(url
)
2767 return url_parsed
._replace
(
2768 netloc
=url_parsed
.netloc
.encode('idna').decode('ascii'),
2769 path
=escape_rfc3986(url_parsed
.path
),
2770 params
=escape_rfc3986(url_parsed
.params
),
2771 query
=escape_rfc3986(url_parsed
.query
),
2772 fragment
=escape_rfc3986(url_parsed
.fragment
)
2777 return compat_parse_qs(compat_urllib_parse_urlparse(url
).query
)
2780 def read_batch_urls(batch_fd
):
2782 if not isinstance(url
, compat_str
):
2783 url
= url
.decode('utf-8', 'replace')
2784 BOM_UTF8
= ('\xef\xbb\xbf', '\ufeff')
2785 for bom
in BOM_UTF8
:
2786 if url
.startswith(bom
):
2787 url
= url
[len(bom
):]
2789 if not url
or url
.startswith(('#', ';', ']')):
2791 # "#" cannot be stripped out since it is part of the URI
2792 # However, it can be safely stipped out if follwing a whitespace
2793 return re
.split(r
'\s#', url
, 1)[0].rstrip()
2795 with contextlib
.closing(batch_fd
) as fd
:
2796 return [url
for url
in map(fixup
, fd
) if url
]
2799 def urlencode_postdata(*args
, **kargs
):
2800 return compat_urllib_parse_urlencode(*args
, **kargs
).encode('ascii')
2803 def update_url_query(url
, query
):
2806 parsed_url
= compat_urlparse
.urlparse(url
)
2807 qs
= compat_parse_qs(parsed_url
.query
)
2809 return compat_urlparse
.urlunparse(parsed_url
._replace
(
2810 query
=compat_urllib_parse_urlencode(qs
, True)))
2813 def update_Request(req
, url
=None, data
=None, headers
={}, query={}
):
2814 req_headers
= req
.headers
.copy()
2815 req_headers
.update(headers
)
2816 req_data
= data
or req
.data
2817 req_url
= update_url_query(url
or req
.get_full_url(), query
)
2818 req_get_method
= req
.get_method()
2819 if req_get_method
== 'HEAD':
2820 req_type
= HEADRequest
2821 elif req_get_method
== 'PUT':
2822 req_type
= PUTRequest
2824 req_type
= compat_urllib_request
.Request
2826 req_url
, data
=req_data
, headers
=req_headers
,
2827 origin_req_host
=req
.origin_req_host
, unverifiable
=req
.unverifiable
)
2828 if hasattr(req
, 'timeout'):
2829 new_req
.timeout
= req
.timeout
2833 def _multipart_encode_impl(data
, boundary
):
2834 content_type
= 'multipart/form-data; boundary=%s' % boundary
2837 for k
, v
in data
.items():
2838 out
+= b
'--' + boundary
.encode('ascii') + b
'\r\n'
2839 if isinstance(k
, compat_str
):
2840 k
= k
.encode('utf-8')
2841 if isinstance(v
, compat_str
):
2842 v
= v
.encode('utf-8')
2843 # RFC 2047 requires non-ASCII field names to be encoded, while RFC 7578
2844 # suggests sending UTF-8 directly. Firefox sends UTF-8, too
2845 content
= b
'Content-Disposition: form-data; name="' + k
+ b
'"\r\n\r\n' + v
+ b
'\r\n'
2846 if boundary
.encode('ascii') in content
:
2847 raise ValueError('Boundary overlaps with data')
2850 out
+= b
'--' + boundary
.encode('ascii') + b
'--\r\n'
2852 return out
, content_type
2855 def multipart_encode(data
, boundary
=None):
2857 Encode a dict to RFC 7578-compliant form-data
2860 A dict where keys and values can be either Unicode or bytes-like
2863 If specified a Unicode object, it's used as the boundary. Otherwise
2864 a random boundary is generated.
2866 Reference: https://tools.ietf.org/html/rfc7578
2868 has_specified_boundary
= boundary
is not None
2871 if boundary
is None:
2872 boundary
= '---------------' + str(random
.randrange(0x0fffffff, 0xffffffff))
2875 out
, content_type
= _multipart_encode_impl(data
, boundary
)
2878 if has_specified_boundary
:
2882 return out
, content_type
2885 def dict_get(d
, key_or_keys
, default
=None, skip_false_values
=True):
2886 if isinstance(key_or_keys
, (list, tuple)):
2887 for key
in key_or_keys
:
2888 if key
not in d
or d
[key
] is None or skip_false_values
and not d
[key
]:
2892 return d
.get(key_or_keys
, default
)
2895 def try_get(src
, getter
, expected_type
=None):
2896 for get
in variadic(getter
):
2899 except (AttributeError, KeyError, TypeError, IndexError):
2902 if expected_type
is None or isinstance(v
, expected_type
):
2906 def merge_dicts(*dicts
):
2908 for a_dict
in dicts
:
2909 for k
, v
in a_dict
.items():
2913 or (isinstance(v
, compat_str
) and v
2914 and isinstance(merged
[k
], compat_str
)
2915 and not merged
[k
])):
2920 def encode_compat_str(string
, encoding
=preferredencoding(), errors
='strict'):
2921 return string
if isinstance(string
, compat_str
) else compat_str(string
, encoding
, errors
)
2933 TV_PARENTAL_GUIDELINES
= {
2943 def parse_age_limit(s
):
2945 return s
if 0 <= s
<= 21 else None
2946 if not isinstance(s
, compat_basestring
):
2948 m
= re
.match(r
'^(?P<age>\d{1,2})\+?$', s
)
2950 return int(m
.group('age'))
2953 return US_RATINGS
[s
]
2954 m
= re
.match(r
'^TV[_-]?(%s)$' % '|'.join(k
[3:] for k
in TV_PARENTAL_GUIDELINES
), s
)
2956 return TV_PARENTAL_GUIDELINES
['TV-' + m
.group(1)]
2960 def strip_jsonp(code
):
2963 (?:window\.)?(?P<func_name>[a-zA-Z0-9_.$]*)
2964 (?:\s*&&\s*(?P=func_name))?
2965 \s*\(\s*(?P<callback_data>.*)\);?
2966 \s*?(?://[^\n]*)*$''',
2967 r
'\g<callback_data>', code
)
2970 def js_to_json(code
, vars={}):
2971 # vars is a dict of var, val pairs to substitute
2972 COMMENT_RE
= r
'/\*(?:(?!\*/).)*?\*/|//[^\n]*\n'
2973 SKIP_RE
= r
'\s*(?:{comment})?\s*'.format(comment
=COMMENT_RE
)
2975 (r
'(?s)^(0[xX][0-9a-fA-F]+){skip}:?$'.format(skip
=SKIP_RE
), 16),
2976 (r
'(?s)^(0+[0-7]+){skip}:?$'.format(skip
=SKIP_RE
), 8),
2981 if v
in ('true', 'false', 'null'):
2983 elif v
in ('undefined', 'void 0'):
2985 elif v
.startswith('/*') or v
.startswith('//') or v
.startswith('!') or v
== ',':
2988 if v
[0] in ("'", '"'):
2989 v
= re
.sub(r
'(?s)\\.|"', lambda m
: {
2994 }.get(m
.group(0), m
.group(0)), v
[1:-1])
2996 for regex
, base
in INTEGER_TABLE
:
2997 im
= re
.match(regex
, v
)
2999 i
= int(im
.group(1), base
)
3000 return '"%d":' % i
if v
.endswith(':') else '%d' % i
3007 return re
.sub(r
'''(?sx)
3008 "(?:[^"\\]*(?:\\\\|\\['"nurtbfx/\n]))*[^"\\]*"|
3009 '(?:[^'\\]*(?:\\\\|\\['"nurtbfx/\n]))*[^'\\]*'|
3010 {comment}|,(?={skip}[\]}}])|
3011 void\s0|(?:(?<![0-9])[eE]|[a-df-zA-DF-Z_$])[.a-zA-Z_$0-9]*|
3012 \b(?:0[xX][0-9a-fA-F]+|0+[0-7]+)(?:{skip}:)?|
3015 '''.format(comment
=COMMENT_RE
, skip
=SKIP_RE
), fix_kv
, code
)
3018 def qualities(quality_ids
):
3019 """ Get a numeric quality value out of a list of possible values """
3022 return quality_ids
.index(qid
)
3029 'default': '%(title)s [%(id)s].%(ext)s',
3030 'chapter': '%(title)s - %(section_number)03d %(section_title)s [%(id)s].%(ext)s',
3036 'description': 'description',
3037 'annotation': 'annotations.xml',
3038 'infojson': 'info.json',
3040 'pl_thumbnail': None,
3041 'pl_description': 'description',
3042 'pl_infojson': 'info.json',
3045 # As of [1] format syntax is:
3046 # %[mapping_key][conversion_flags][minimum_width][.precision][length_modifier]type
3047 # 1. https://docs.python.org/2/library/stdtypes.html#string-formatting
3048 STR_FORMAT_RE_TMPL
= r
'''(?x)
3049 (?<!%)(?P<prefix>(?:%%)*)
3051 (?P<has_key>\((?P<key>{0})\))?
3053 (?P<conversion>[#0\-+ ]+)?
3055 (?P<precision>\.\d+)?
3056 (?P<len_mod>[hlL])? # unused in python
3057 {1} # conversion type
3062 STR_FORMAT_TYPES
= 'diouxXeEfFgGcrs'
3065 def limit_length(s
, length
):
3066 """ Add ellipses to overly long strings """
3071 return s
[:length
- len(ELLIPSES
)] + ELLIPSES
3075 def version_tuple(v
):
3076 return tuple(int(e
) for e
in re
.split(r
'[-.]', v
))
3079 def is_outdated_version(version
, limit
, assume_new
=True):
3081 return not assume_new
3083 return version_tuple(version
) < version_tuple(limit
)
3085 return not assume_new
3088 def ytdl_is_updateable():
3089 """ Returns if yt-dlp can be updated with -U """
3091 from .update
import is_non_updateable
3093 return not is_non_updateable()
3096 def args_to_str(args
):
3097 # Get a short string representation for a subprocess command
3098 return ' '.join(compat_shlex_quote(a
) for a
in args
)
3101 def error_to_compat_str(err
):
3103 # On python 2 error byte string must be decoded with proper
3104 # encoding rather than ascii
3105 if sys
.version_info
[0] < 3:
3106 err_str
= err_str
.decode(preferredencoding())
3110 def mimetype2ext(mt
):
3114 mt
, _
, params
= mt
.partition(';')
3119 # Per RFC 3003, audio/mpeg can be .mp1, .mp2 or .mp3. Here use .mp3 as
3120 # it's the most popular one
3121 'audio/mpeg': 'mp3',
3122 'audio/x-wav': 'wav',
3124 'audio/wave': 'wav',
3127 ext
= FULL_MAP
.get(mt
)
3133 'smptett+xml': 'tt',
3137 'x-mp4-fragmented': 'mp4',
3138 'x-ms-sami': 'sami',
3141 'x-mpegurl': 'm3u8',
3142 'vnd.apple.mpegurl': 'm3u8',
3146 'vnd.ms-sstr+xml': 'ism',
3150 'filmstrip+json': 'fs',
3154 _
, _
, subtype
= mt
.rpartition('/')
3155 ext
= SUBTYPE_MAP
.get(subtype
.lower())
3166 _
, _
, suffix
= subtype
.partition('+')
3167 ext
= SUFFIX_MAP
.get(suffix
)
3171 return subtype
.replace('+', '.')
3174 def ext2mimetype(ext_or_url
):
3177 if '.' not in ext_or_url
:
3178 ext_or_url
= f
'file.{ext_or_url}'
3179 return mimetypes
.guess_type(ext_or_url
)[0]
3182 def parse_codecs(codecs_str
):
3183 # http://tools.ietf.org/html/rfc6381
3186 split_codecs
= list(filter(None, map(
3187 str.strip
, codecs_str
.strip().strip(',').split(','))))
3188 vcodec
, acodec
, hdr
= None, None, None
3189 for full_codec
in split_codecs
:
3190 parts
= full_codec
.split('.')
3191 codec
= parts
[0].replace('0', '')
3192 if codec
in ('avc1', 'avc2', 'avc3', 'avc4', 'vp9', 'vp8', 'hev1', 'hev2',
3193 'h263', 'h264', 'mp4v', 'hvc1', 'av1', 'theora', 'dvh1', 'dvhe'):
3195 vcodec
= '.'.join(parts
[:4]) if codec
in ('vp9', 'av1', 'hvc1') else full_codec
3196 if codec
in ('dvh1', 'dvhe'):
3198 elif codec
== 'av1' and len(parts
) > 3 and parts
[3] == '10':
3200 elif full_codec
.replace('0', '').startswith('vp9.2'):
3202 elif codec
in ('flac', 'mp4a', 'opus', 'vorbis', 'mp3', 'aac', 'ac-3', 'ec-3', 'eac3', 'dtsc', 'dtse', 'dtsh', 'dtsl'):
3206 write_string('WARNING: Unknown codec %s\n' % full_codec
, sys
.stderr
)
3207 if vcodec
or acodec
:
3209 'vcodec': vcodec
or 'none',
3210 'acodec': acodec
or 'none',
3211 'dynamic_range': hdr
,
3213 elif len(split_codecs
) == 2:
3215 'vcodec': split_codecs
[0],
3216 'acodec': split_codecs
[1],
3221 def urlhandle_detect_ext(url_handle
):
3222 getheader
= url_handle
.headers
.get
3224 cd
= getheader('Content-Disposition')
3226 m
= re
.match(r
'attachment;\s*filename="(?P<filename>[^"]+)"', cd
)
3228 e
= determine_ext(m
.group('filename'), default_ext
=None)
3232 return mimetype2ext(getheader('Content-Type'))
3235 def encode_data_uri(data
, mime_type
):
3236 return 'data:%s;base64,%s' % (mime_type
, base64
.b64encode(data
).decode('ascii'))
3239 def age_restricted(content_limit
, age_limit
):
3240 """ Returns True iff the content should be blocked """
3242 if age_limit
is None: # No limit set
3244 if content_limit
is None:
3245 return False # Content available for everyone
3246 return age_limit
< content_limit
3249 def is_html(first_bytes
):
3250 """ Detect whether a file contains HTML by examining its first bytes. """
3253 (b
'\xef\xbb\xbf', 'utf-8'),
3254 (b
'\x00\x00\xfe\xff', 'utf-32-be'),
3255 (b
'\xff\xfe\x00\x00', 'utf-32-le'),
3256 (b
'\xff\xfe', 'utf-16-le'),
3257 (b
'\xfe\xff', 'utf-16-be'),
3259 for bom
, enc
in BOMS
:
3260 if first_bytes
.startswith(bom
):
3261 s
= first_bytes
[len(bom
):].decode(enc
, 'replace')
3264 s
= first_bytes
.decode('utf-8', 'replace')
3266 return re
.match(r
'^\s*<', s
)
3269 def determine_protocol(info_dict
):
3270 protocol
= info_dict
.get('protocol')
3271 if protocol
is not None:
3274 url
= sanitize_url(info_dict
['url'])
3275 if url
.startswith('rtmp'):
3277 elif url
.startswith('mms'):
3279 elif url
.startswith('rtsp'):
3282 ext
= determine_ext(url
)
3288 return compat_urllib_parse_urlparse(url
).scheme
3291 def render_table(header_row
, data
, delim
=False, extra_gap
=0, hide_empty
=False):
3292 """ Render a list of rows, each as a list of values.
3293 Text after a \t will be right aligned """
3295 return len(remove_terminal_sequences(string
).replace('\t', ''))
3297 def get_max_lens(table
):
3298 return [max(width(str(v
)) for v
in col
) for col
in zip(*table
)]
3300 def filter_using_list(row
, filterArray
):
3301 return [col
for (take
, col
) in zip(filterArray
, row
) if take
]
3304 max_lens
= get_max_lens(data
)
3305 header_row
= filter_using_list(header_row
, max_lens
)
3306 data
= [filter_using_list(row
, max_lens
) for row
in data
]
3308 table
= [header_row
] + data
3309 max_lens
= get_max_lens(table
)
3312 table
= [header_row
, [delim
* (ml
+ extra_gap
) for ml
in max_lens
]] + data
3313 table
[1][-1] = table
[1][-1][:-extra_gap
] # Remove extra_gap from end of delimiter
3315 for pos
, text
in enumerate(map(str, row
)):
3317 row
[pos
] = text
.replace('\t', ' ' * (max_lens
[pos
] - width(text
))) + ' ' * extra_gap
3319 row
[pos
] = text
+ ' ' * (max_lens
[pos
] - width(text
) + extra_gap
)
3320 ret
= '\n'.join(''.join(row
).rstrip() for row
in table
)
3324 def _match_one(filter_part
, dct
, incomplete
):
3325 # TODO: Generalize code with YoutubeDL._build_format_filter
3326 STRING_OPERATORS
= {
3327 '*=': operator
.contains
,
3328 '^=': lambda attr
, value
: attr
.startswith(value
),
3329 '$=': lambda attr
, value
: attr
.endswith(value
),
3330 '~=': lambda attr
, value
: re
.search(value
, attr
),
3332 COMPARISON_OPERATORS
= {
3334 '<=': operator
.le
, # "<=" must be defined above "<"
3341 operator_rex
= re
.compile(r
'''(?x)\s*
3343 \s*(?P<negation>!\s*)?(?P<op>%s)(?P<none_inclusive>\s*\?)?\s*
3345 (?P<quote>["\'])(?P
<quotedstrval
>.+?
)(?P
=quote
)|
3349 ''' % '|'.join(map(re.escape, COMPARISON_OPERATORS.keys())))
3350 m = operator_rex.search(filter_part)
3353 unnegated_op = COMPARISON_OPERATORS[m['op']]
3355 op = lambda attr, value: not unnegated_op(attr, value)
3358 comparison_value = m['quotedstrval'] or m['strval'] or m['intval']
3360 comparison_value = comparison_value.replace(r'\%s' % m['quote'], m['quote'])
3361 actual_value = dct.get(m['key'])
3362 numeric_comparison = None
3363 if isinstance(actual_value, compat_numeric_types):
3364 # If the original field is a string and matching comparisonvalue is
3365 # a number we should respect the origin of the original field
3366 # and process comparison value as a string (see
3367 # https://github.com/ytdl-org/youtube-dl/issues/11082)
3369 numeric_comparison = int(comparison_value)
3371 numeric_comparison = parse_filesize(comparison_value)
3372 if numeric_comparison is None:
3373 numeric_comparison = parse_filesize(f'{comparison_value}B')
3374 if numeric_comparison is None:
3375 numeric_comparison = parse_duration(comparison_value)
3376 if numeric_comparison is not None and m['op'] in STRING_OPERATORS:
3377 raise ValueError('Operator %s only supports string values!' % m['op'])
3378 if actual_value is None:
3379 return incomplete or m['none_inclusive']
3380 return op(actual_value, comparison_value if numeric_comparison is None else numeric_comparison)
3383 '': lambda v: (v is True) if isinstance(v, bool) else (v is not None),
3384 '!': lambda v: (v is False) if isinstance(v, bool) else (v is None),
3386 operator_rex = re.compile(r'''(?x
)\s
*
3387 (?P
<op
>%s)\s
*(?P
<key
>[a
-z_
]+)
3389 ''' % '|'.join(map(re.escape, UNARY_OPERATORS.keys())))
3390 m = operator_rex.search(filter_part)
3392 op = UNARY_OPERATORS[m.group('op')]
3393 actual_value = dct.get(m.group('key'))
3394 if incomplete and actual_value is None:
3396 return op(actual_value)
3398 raise ValueError('Invalid filter part %r' % filter_part)
3401 def match_str(filter_str, dct, incomplete=False):
3402 """ Filter a dictionary with a simple string syntax. Returns True (=passes filter) or false
3403 When incomplete, all conditions passes on missing fields
3406 _match_one(filter_part.replace(r'\&', '&'), dct, incomplete)
3407 for filter_part in re.split(r'(?<!\\)&', filter_str))
3410 def match_filter_func(filter_str):
3411 def _match_func(info_dict, *args, **kwargs):
3412 if match_str(filter_str, info_dict, *args, **kwargs):
3415 video_title = info_dict.get('title', info_dict.get('id', 'video'))
3416 return '%s does not pass filter %s, skipping ..' % (video_title, filter_str)
3420 def parse_dfxp_time_expr(time_expr):
3424 mobj = re.match(r'^(?P<time_offset>\d+(?:\.\d+)?)s?$', time_expr)
3426 return float(mobj.group('time_offset'))
3428 mobj = re.match(r'^(\d+):(\d\d):(\d\d(?:(?:\.|:)\d+)?)$', time_expr)
3430 return 3600 * int(mobj.group(1)) + 60 * int(mobj.group(2)) + float(mobj.group(3).replace(':', '.'))
3433 def srt_subtitles_timecode(seconds):
3434 return '%02d:%02d:%02d,%03d' % timetuple_from_msec(seconds * 1000)
3437 def ass_subtitles_timecode(seconds):
3438 time = timetuple_from_msec(seconds * 1000)
3439 return '%01d:%02d:%02d.%02d' % (*time[:-1], time.milliseconds / 10)
3442 def dfxp2srt(dfxp_data):
3444 @param dfxp_data A
bytes-like
object containing DFXP data
3445 @returns A
unicode object containing converted SRT data
3447 LEGACY_NAMESPACES = (
3448 (b'http://www.w3.org/ns/ttml', [
3449 b'http://www.w3.org/2004/11/ttaf1',
3450 b'http://www.w3.org/2006/04/ttaf1',
3451 b'http://www.w3.org/2006/10/ttaf1',
3453 (b'http://www.w3.org/ns/ttml#styling', [
3454 b'http://www.w3.org/ns/ttml#style',
3458 SUPPORTED_STYLING = [
3467 _x = functools.partial(xpath_with_ns, ns_map={
3468 'xml': 'http://www.w3.org/XML/1998/namespace',
3469 'ttml': 'http://www.w3.org/ns/ttml',
3470 'tts': 'http://www.w3.org/ns/ttml#styling',
3476 class TTMLPElementParser(object):
3478 _unclosed_elements = []
3479 _applied_styles = []
3481 def start(self, tag, attrib):
3482 if tag in (_x('ttml:br'), 'br'):
3485 unclosed_elements = []
3487 element_style_id = attrib.get('style')
3489 style.update(default_style)
3490 if element_style_id:
3491 style.update(styles.get(element_style_id, {}))
3492 for prop in SUPPORTED_STYLING:
3493 prop_val = attrib.get(_x('tts:' + prop))
3495 style[prop] = prop_val
3498 for k, v in sorted(style.items()):
3499 if self._applied_styles and self._applied_styles[-1].get(k) == v:
3502 font += ' color="%s"' % v
3503 elif k == 'fontSize':
3504 font += ' size="%s"' % v
3505 elif k == 'fontFamily':
3506 font += ' face="%s"' % v
3507 elif k == 'fontWeight' and v == 'bold':
3509 unclosed_elements.append('b')
3510 elif k == 'fontStyle' and v == 'italic':
3512 unclosed_elements.append('i')
3513 elif k == 'textDecoration' and v == 'underline':
3515 unclosed_elements.append('u')
3517 self._out += '<font' + font + '>'
3518 unclosed_elements.append('font')
3520 if self._applied_styles:
3521 applied_style.update(self._applied_styles[-1])
3522 applied_style.update(style)
3523 self._applied_styles.append(applied_style)
3524 self._unclosed_elements.append(unclosed_elements)
3527 if tag not in (_x('ttml:br'), 'br'):
3528 unclosed_elements = self._unclosed_elements.pop()
3529 for element in reversed(unclosed_elements):
3530 self._out += '</%s>' % element
3531 if unclosed_elements and self._applied_styles:
3532 self._applied_styles.pop()
3534 def data(self, data):
3538 return self._out.strip()
3540 def parse_node(node):
3541 target = TTMLPElementParser()
3542 parser = xml.etree.ElementTree.XMLParser(target=target)
3543 parser.feed(xml.etree.ElementTree.tostring(node))
3544 return parser.close()
3546 for k, v in LEGACY_NAMESPACES:
3548 dfxp_data = dfxp_data.replace(ns, k)
3550 dfxp = compat_etree_fromstring(dfxp_data)
3552 paras = dfxp.findall(_x('.//ttml:p')) or dfxp.findall('.//p')
3555 raise ValueError('Invalid dfxp/TTML subtitle')
3559 for style in dfxp.findall(_x('.//ttml:style')):
3560 style_id = style.get('id') or style.get(_x('xml:id'))
3563 parent_style_id = style.get('style')
3565 if parent_style_id not in styles:
3568 styles[style_id] = styles[parent_style_id].copy()
3569 for prop in SUPPORTED_STYLING:
3570 prop_val = style.get(_x('tts:' + prop))
3572 styles.setdefault(style_id, {})[prop] = prop_val
3578 for p in ('body', 'div'):
3579 ele = xpath_element(dfxp, [_x('.//ttml:' + p), './/' + p])
3582 style = styles.get(ele.get('style'))
3585 default_style.update(style)
3587 for para, index in zip(paras, itertools.count(1)):
3588 begin_time = parse_dfxp_time_expr(para.attrib.get('begin'))
3589 end_time = parse_dfxp_time_expr(para.attrib.get('end'))
3590 dur = parse_dfxp_time_expr(para.attrib.get('dur'))
3591 if begin_time is None:
3596 end_time = begin_time + dur
3597 out.append('%d\n%s --> %s\n%s\n\n' % (
3599 srt_subtitles_timecode(begin_time),
3600 srt_subtitles_timecode(end_time),
3606 def cli_option(params, command_option, param):
3607 param = params.get(param)
3609 param = compat_str(param)
3610 return [command_option, param] if param is not None else []
3613 def cli_bool_option(params, command_option, param, true_value='true', false_value='false', separator=None):
3614 param = params.get(param)
3617 assert isinstance(param, bool)
3619 return [command_option + separator + (true_value if param else false_value)]
3620 return [command_option, true_value if param else false_value]
3623 def cli_valueless_option(params, command_option, param, expected_value=True):
3624 param = params.get(param)
3625 return [command_option] if param == expected_value else []
3628 def cli_configuration_args(argdict, keys, default=[], use_compat=True):
3629 if isinstance(argdict, (list, tuple)): # for backward compatibility
3636 assert isinstance(argdict, dict)
3638 assert isinstance(keys, (list, tuple))
3639 for key_list in keys:
3640 arg_list = list(filter(
3641 lambda x: x is not None,
3642 [argdict.get(key.lower()) for key in variadic(key_list)]))
3644 return [arg for args in arg_list for arg in args]
3648 def _configuration_args(main_key, argdict, exe, keys=None, default=[], use_compat=True):
3649 main_key, exe = main_key.lower(), exe.lower()
3650 root_key = exe if main_key == exe else f'{main_key}+{exe}'
3651 keys = [f'{root_key}{k}' for k in (keys or [''])]
3652 if root_key in keys:
3654 keys.append((main_key, exe))
3655 keys.append('default')
3658 return cli_configuration_args(argdict, keys, default, use_compat)
3661 class ISO639Utils(object):
3662 # See http://www.loc.gov/standards/iso639-2/ISO-639-2_utf-8.txt
3721 'iw': 'heb', # Replaced by he in 1989 revision
3731 'in': 'ind', # Replaced by id in 1989 revision
3846 'ji': 'yid', # Replaced by yi in 1989 revision
3854 def short2long(cls, code):
3855 """Convert language code from ISO 639-1 to ISO 639-2/T"""
3856 return cls._lang_map.get(code[:2])
3859 def long2short(cls, code):
3860 """Convert language code from ISO 639-2/T to ISO 639-1"""
3861 for short_name, long_name in cls._lang_map.items():
3862 if long_name == code:
3866 class ISO3166Utils(object):
3867 # From http://data.okfn.org/data/core/country-list
3869 'AF': 'Afghanistan',
3870 'AX': 'Åland Islands',
3873 'AS': 'American Samoa',
3878 'AG': 'Antigua and Barbuda',
3895 'BO': 'Bolivia, Plurinational State of',
3896 'BQ': 'Bonaire, Sint Eustatius and Saba',
3897 'BA': 'Bosnia and Herzegovina',
3899 'BV': 'Bouvet Island',
3901 'IO': 'British Indian Ocean Territory',
3902 'BN': 'Brunei Darussalam',
3904 'BF': 'Burkina Faso',
3910 'KY': 'Cayman Islands',
3911 'CF': 'Central African Republic',
3915 'CX': 'Christmas Island',
3916 'CC': 'Cocos (Keeling) Islands',
3920 'CD': 'Congo, the Democratic Republic of the',
3921 'CK': 'Cook Islands',
3923 'CI': 'Côte d\'Ivoire',
3928 'CZ': 'Czech Republic',
3932 'DO': 'Dominican Republic',
3935 'SV': 'El Salvador',
3936 'GQ': 'Equatorial Guinea',
3940 'FK': 'Falkland Islands (Malvinas)',
3941 'FO': 'Faroe Islands',
3945 'GF': 'French Guiana',
3946 'PF': 'French Polynesia',
3947 'TF': 'French Southern Territories',
3962 'GW': 'Guinea-Bissau',
3965 'HM': 'Heard Island and McDonald Islands',
3966 'VA': 'Holy See (Vatican City State)',
3973 'IR': 'Iran, Islamic Republic of',
3976 'IM': 'Isle of Man',
3986 'KP': 'Korea, Democratic People\'s Republic of',
3987 'KR': 'Korea, Republic of',
3990 'LA': 'Lao People\'s Democratic Republic',
3996 'LI': 'Liechtenstein',
4000 'MK': 'Macedonia, the Former Yugoslav Republic of',
4007 'MH': 'Marshall Islands',
4013 'FM': 'Micronesia, Federated States of',
4014 'MD': 'Moldova, Republic of',
4025 'NL': 'Netherlands',
4026 'NC': 'New Caledonia',
4027 'NZ': 'New Zealand',
4032 'NF': 'Norfolk Island',
4033 'MP': 'Northern Mariana Islands',
4038 'PS': 'Palestine, State of',
4040 'PG': 'Papua New Guinea',
4043 'PH': 'Philippines',
4047 'PR': 'Puerto Rico',
4051 'RU': 'Russian Federation',
4053 'BL': 'Saint Barthélemy',
4054 'SH': 'Saint Helena, Ascension and Tristan da Cunha',
4055 'KN': 'Saint Kitts and Nevis',
4056 'LC': 'Saint Lucia',
4057 'MF': 'Saint Martin (French part)',
4058 'PM': 'Saint Pierre and Miquelon',
4059 'VC': 'Saint Vincent and the Grenadines',
4062 'ST': 'Sao Tome and Principe',
4063 'SA': 'Saudi Arabia',
4067 'SL': 'Sierra Leone',
4069 'SX': 'Sint Maarten (Dutch part)',
4072 'SB': 'Solomon Islands',
4074 'ZA': 'South Africa',
4075 'GS': 'South Georgia and the South Sandwich Islands',
4076 'SS': 'South Sudan',
4081 'SJ': 'Svalbard and Jan Mayen',
4084 'CH': 'Switzerland',
4085 'SY': 'Syrian Arab Republic',
4086 'TW': 'Taiwan, Province of China',
4088 'TZ': 'Tanzania, United Republic of',
4090 'TL': 'Timor-Leste',
4094 'TT': 'Trinidad and Tobago',
4097 'TM': 'Turkmenistan',
4098 'TC': 'Turks and Caicos Islands',
4102 'AE': 'United Arab Emirates',
4103 'GB': 'United Kingdom',
4104 'US': 'United States',
4105 'UM': 'United States Minor Outlying Islands',
4109 'VE': 'Venezuela, Bolivarian Republic of',
4111 'VG': 'Virgin Islands, British',
4112 'VI': 'Virgin Islands, U.S.',
4113 'WF': 'Wallis and Futuna',
4114 'EH': 'Western Sahara',
4121 def short2full(cls, code):
4122 """Convert an ISO 3166-2 country code to the corresponding full name"""
4123 return cls._country_map.get(code.upper())
4126 class GeoUtils(object):
4127 # Major IPv4 address blocks per country
4129 'AD': '46.172.224.0/19',
4130 'AE': '94.200.0.0/13',
4131 'AF': '149.54.0.0/17',
4132 'AG': '209.59.64.0/18',
4133 'AI': '204.14.248.0/21',
4134 'AL': '46.99.0.0/16',
4135 'AM': '46.70.0.0/15',
4136 'AO': '105.168.0.0/13',
4137 'AP': '182.50.184.0/21',
4138 'AQ': '23.154.160.0/24',
4139 'AR': '181.0.0.0/12',
4140 'AS': '202.70.112.0/20',
4141 'AT': '77.116.0.0/14',
4142 'AU': '1.128.0.0/11',
4143 'AW': '181.41.0.0/18',
4144 'AX': '185.217.4.0/22',
4145 'AZ': '5.197.0.0/16',
4146 'BA': '31.176.128.0/17',
4147 'BB': '65.48.128.0/17',
4148 'BD': '114.130.0.0/16',
4150 'BF': '102.178.0.0/15',
4151 'BG': '95.42.0.0/15',
4152 'BH': '37.131.0.0/17',
4153 'BI': '154.117.192.0/18',
4154 'BJ': '137.255.0.0/16',
4155 'BL': '185.212.72.0/23',
4156 'BM': '196.12.64.0/18',
4157 'BN': '156.31.0.0/16',
4158 'BO': '161.56.0.0/16',
4159 'BQ': '161.0.80.0/20',
4160 'BR': '191.128.0.0/12',
4161 'BS': '24.51.64.0/18',
4162 'BT': '119.2.96.0/19',
4163 'BW': '168.167.0.0/16',
4164 'BY': '178.120.0.0/13',
4165 'BZ': '179.42.192.0/18',
4166 'CA': '99.224.0.0/11',
4167 'CD': '41.243.0.0/16',
4168 'CF': '197.242.176.0/21',
4169 'CG': '160.113.0.0/16',
4170 'CH': '85.0.0.0/13',
4171 'CI': '102.136.0.0/14',
4172 'CK': '202.65.32.0/19',
4173 'CL': '152.172.0.0/14',
4174 'CM': '102.244.0.0/14',
4175 'CN': '36.128.0.0/10',
4176 'CO': '181.240.0.0/12',
4177 'CR': '201.192.0.0/12',
4178 'CU': '152.206.0.0/15',
4179 'CV': '165.90.96.0/19',
4180 'CW': '190.88.128.0/17',
4181 'CY': '31.153.0.0/16',
4182 'CZ': '88.100.0.0/14',
4184 'DJ': '197.241.0.0/17',
4185 'DK': '87.48.0.0/12',
4186 'DM': '192.243.48.0/20',
4187 'DO': '152.166.0.0/15',
4188 'DZ': '41.96.0.0/12',
4189 'EC': '186.68.0.0/15',
4190 'EE': '90.190.0.0/15',
4191 'EG': '156.160.0.0/11',
4192 'ER': '196.200.96.0/20',
4193 'ES': '88.0.0.0/11',
4194 'ET': '196.188.0.0/14',
4195 'EU': '2.16.0.0/13',
4196 'FI': '91.152.0.0/13',
4197 'FJ': '144.120.0.0/16',
4198 'FK': '80.73.208.0/21',
4199 'FM': '119.252.112.0/20',
4200 'FO': '88.85.32.0/19',
4202 'GA': '41.158.0.0/15',
4204 'GD': '74.122.88.0/21',
4205 'GE': '31.146.0.0/16',
4206 'GF': '161.22.64.0/18',
4207 'GG': '62.68.160.0/19',
4208 'GH': '154.160.0.0/12',
4209 'GI': '95.164.0.0/16',
4210 'GL': '88.83.0.0/19',
4211 'GM': '160.182.0.0/15',
4212 'GN': '197.149.192.0/18',
4213 'GP': '104.250.0.0/19',
4214 'GQ': '105.235.224.0/20',
4215 'GR': '94.64.0.0/13',
4216 'GT': '168.234.0.0/16',
4217 'GU': '168.123.0.0/16',
4218 'GW': '197.214.80.0/20',
4219 'GY': '181.41.64.0/18',
4220 'HK': '113.252.0.0/14',
4221 'HN': '181.210.0.0/16',
4222 'HR': '93.136.0.0/13',
4223 'HT': '148.102.128.0/17',
4224 'HU': '84.0.0.0/14',
4225 'ID': '39.192.0.0/10',
4226 'IE': '87.32.0.0/12',
4227 'IL': '79.176.0.0/13',
4228 'IM': '5.62.80.0/20',
4229 'IN': '117.192.0.0/10',
4230 'IO': '203.83.48.0/21',
4231 'IQ': '37.236.0.0/14',
4232 'IR': '2.176.0.0/12',
4233 'IS': '82.221.0.0/16',
4234 'IT': '79.0.0.0/10',
4235 'JE': '87.244.64.0/18',
4236 'JM': '72.27.0.0/17',
4237 'JO': '176.29.0.0/16',
4238 'JP': '133.0.0.0/8',
4239 'KE': '105.48.0.0/12',
4240 'KG': '158.181.128.0/17',
4241 'KH': '36.37.128.0/17',
4242 'KI': '103.25.140.0/22',
4243 'KM': '197.255.224.0/20',
4244 'KN': '198.167.192.0/19',
4245 'KP': '175.45.176.0/22',
4246 'KR': '175.192.0.0/10',
4247 'KW': '37.36.0.0/14',
4248 'KY': '64.96.0.0/15',
4249 'KZ': '2.72.0.0/13',
4250 'LA': '115.84.64.0/18',
4251 'LB': '178.135.0.0/16',
4252 'LC': '24.92.144.0/20',
4253 'LI': '82.117.0.0/19',
4254 'LK': '112.134.0.0/15',
4255 'LR': '102.183.0.0/16',
4256 'LS': '129.232.0.0/17',
4257 'LT': '78.56.0.0/13',
4258 'LU': '188.42.0.0/16',
4259 'LV': '46.109.0.0/16',
4260 'LY': '41.252.0.0/14',
4261 'MA': '105.128.0.0/11',
4262 'MC': '88.209.64.0/18',
4263 'MD': '37.246.0.0/16',
4264 'ME': '178.175.0.0/17',
4265 'MF': '74.112.232.0/21',
4266 'MG': '154.126.0.0/17',
4267 'MH': '117.103.88.0/21',
4268 'MK': '77.28.0.0/15',
4269 'ML': '154.118.128.0/18',
4270 'MM': '37.111.0.0/17',
4271 'MN': '49.0.128.0/17',
4272 'MO': '60.246.0.0/16',
4273 'MP': '202.88.64.0/20',
4274 'MQ': '109.203.224.0/19',
4275 'MR': '41.188.64.0/18',
4276 'MS': '208.90.112.0/22',
4277 'MT': '46.11.0.0/16',
4278 'MU': '105.16.0.0/12',
4279 'MV': '27.114.128.0/18',
4280 'MW': '102.70.0.0/15',
4281 'MX': '187.192.0.0/11',
4282 'MY': '175.136.0.0/13',
4283 'MZ': '197.218.0.0/15',
4284 'NA': '41.182.0.0/16',
4285 'NC': '101.101.0.0/18',
4286 'NE': '197.214.0.0/18',
4287 'NF': '203.17.240.0/22',
4288 'NG': '105.112.0.0/12',
4289 'NI': '186.76.0.0/15',
4290 'NL': '145.96.0.0/11',
4291 'NO': '84.208.0.0/13',
4292 'NP': '36.252.0.0/15',
4293 'NR': '203.98.224.0/19',
4294 'NU': '49.156.48.0/22',
4295 'NZ': '49.224.0.0/14',
4296 'OM': '5.36.0.0/15',
4297 'PA': '186.72.0.0/15',
4298 'PE': '186.160.0.0/14',
4299 'PF': '123.50.64.0/18',
4300 'PG': '124.240.192.0/19',
4301 'PH': '49.144.0.0/13',
4302 'PK': '39.32.0.0/11',
4303 'PL': '83.0.0.0/11',
4304 'PM': '70.36.0.0/20',
4305 'PR': '66.50.0.0/16',
4306 'PS': '188.161.0.0/16',
4307 'PT': '85.240.0.0/13',
4308 'PW': '202.124.224.0/20',
4309 'PY': '181.120.0.0/14',
4310 'QA': '37.210.0.0/15',
4311 'RE': '102.35.0.0/16',
4312 'RO': '79.112.0.0/13',
4313 'RS': '93.86.0.0/15',
4314 'RU': '5.136.0.0/13',
4315 'RW': '41.186.0.0/16',
4316 'SA': '188.48.0.0/13',
4317 'SB': '202.1.160.0/19',
4318 'SC': '154.192.0.0/11',
4319 'SD': '102.120.0.0/13',
4320 'SE': '78.64.0.0/12',
4321 'SG': '8.128.0.0/10',
4322 'SI': '188.196.0.0/14',
4323 'SK': '78.98.0.0/15',
4324 'SL': '102.143.0.0/17',
4325 'SM': '89.186.32.0/19',
4326 'SN': '41.82.0.0/15',
4327 'SO': '154.115.192.0/18',
4328 'SR': '186.179.128.0/17',
4329 'SS': '105.235.208.0/21',
4330 'ST': '197.159.160.0/19',
4331 'SV': '168.243.0.0/16',
4332 'SX': '190.102.0.0/20',
4334 'SZ': '41.84.224.0/19',
4335 'TC': '65.255.48.0/20',
4336 'TD': '154.68.128.0/19',
4337 'TG': '196.168.0.0/14',
4338 'TH': '171.96.0.0/13',
4339 'TJ': '85.9.128.0/18',
4340 'TK': '27.96.24.0/21',
4341 'TL': '180.189.160.0/20',
4342 'TM': '95.85.96.0/19',
4343 'TN': '197.0.0.0/11',
4344 'TO': '175.176.144.0/21',
4345 'TR': '78.160.0.0/11',
4346 'TT': '186.44.0.0/15',
4347 'TV': '202.2.96.0/19',
4348 'TW': '120.96.0.0/11',
4349 'TZ': '156.156.0.0/14',
4350 'UA': '37.52.0.0/14',
4351 'UG': '102.80.0.0/13',
4353 'UY': '167.56.0.0/13',
4354 'UZ': '84.54.64.0/18',
4355 'VA': '212.77.0.0/19',
4356 'VC': '207.191.240.0/21',
4357 'VE': '186.88.0.0/13',
4358 'VG': '66.81.192.0/20',
4359 'VI': '146.226.0.0/16',
4360 'VN': '14.160.0.0/11',
4361 'VU': '202.80.32.0/20',
4362 'WF': '117.20.32.0/21',
4363 'WS': '202.4.32.0/19',
4364 'YE': '134.35.0.0/16',
4365 'YT': '41.242.116.0/22',
4366 'ZA': '41.0.0.0/11',
4367 'ZM': '102.144.0.0/13',
4368 'ZW': '102.177.192.0/18',
4372 def random_ipv4(cls, code_or_block):
4373 if len(code_or_block) == 2:
4374 block = cls._country_ip_map.get(code_or_block.upper())
4378 block = code_or_block
4379 addr, preflen = block.split('/')
4380 addr_min = compat_struct_unpack('!L', socket.inet_aton(addr))[0]
4381 addr_max = addr_min | (0xffffffff >> int(preflen))
4382 return compat_str(socket.inet_ntoa(
4383 compat_struct_pack('!L', random.randint(addr_min, addr_max))))
4386 class PerRequestProxyHandler(compat_urllib_request.ProxyHandler):
4387 def __init__(self, proxies=None):
4388 # Set default handlers
4389 for type in ('http', 'https'):
4390 setattr(self, '%s_open' % type,
4391 lambda r, proxy='__noproxy__', type=type, meth=self.proxy_open:
4392 meth(r, proxy, type))
4393 compat_urllib_request.ProxyHandler.__init__(self, proxies)
4395 def proxy_open(self, req, proxy, type):
4396 req_proxy = req.headers.get('Ytdl-request-proxy')
4397 if req_proxy is not None:
4399 del req.headers['Ytdl-request-proxy']
4401 if proxy == '__noproxy__':
4402 return None # No Proxy
4403 if compat_urlparse.urlparse(proxy).scheme.lower() in ('socks', 'socks4', 'socks4a', 'socks5'):
4404 req.add_header('Ytdl-socks-proxy', proxy)
4405 # yt-dlp's http/https handlers do wrapping the socket with socks
4407 return compat_urllib_request.ProxyHandler.proxy_open(
4408 self, req, proxy, type)
4411 # Both long_to_bytes and bytes_to_long are adapted from PyCrypto, which is
4412 # released into Public Domain
4413 # https://github.com/dlitz/pycrypto/blob/master/lib/Crypto/Util/number.py#L387
4415 def long_to_bytes(n, blocksize=0):
4416 """long_to_bytes(n:long, blocksize:int) : string
4417 Convert a long integer to a byte string.
4419 If optional blocksize is given and greater than zero, pad the front of the
4420 byte string with binary zeros so that the length is a multiple of
4423 # after much testing, this algorithm was deemed to be the fastest
4427 s = compat_struct_pack('>I', n & 0xffffffff) + s
4429 # strip off leading zeros
4430 for i in range(len(s)):
4431 if s[i] != b'\000'[0]:
4434 # only happens when n == 0
4438 # add back some pad bytes. this could be done more efficiently w.r.t. the
4439 # de-padding being done above, but sigh...
4440 if blocksize > 0 and len(s) % blocksize:
4441 s = (blocksize - len(s) % blocksize) * b'\000' + s
4445 def bytes_to_long(s):
4446 """bytes_to_long(string) : long
4447 Convert a byte string to a long integer.
4449 This is (essentially) the inverse of long_to_bytes().
4454 extra = (4 - length % 4)
4455 s = b'\000' * extra + s
4456 length = length + extra
4457 for i in range(0, length, 4):
4458 acc = (acc << 32) + compat_struct_unpack('>I', s[i:i + 4])[0]
4462 def ohdave_rsa_encrypt(data, exponent, modulus):
4464 Implement OHDave
's RSA algorithm. See http://www.ohdave.com/rsa/
4467 data: data to encrypt, bytes-like object
4468 exponent, modulus: parameter e and N of RSA algorithm, both integer
4469 Output: hex string of encrypted data
4471 Limitation: supports one block encryption only
4474 payload = int(binascii.hexlify(data[::-1]), 16)
4475 encrypted = pow(payload, exponent, modulus)
4476 return '%x' % encrypted
4479 def pkcs1pad(data, length):
4481 Padding input data with PKCS#1 scheme
4483 @param {int[]} data input data
4484 @param {int} length target length
4485 @returns {int[]} padded data
4487 if len(data) > length - 11:
4488 raise ValueError('Input data too
long for PKCS
#1 padding')
4490 pseudo_random
= [random
.randint(0, 254) for _
in range(length
- len(data
) - 3)]
4491 return [0, 2] + pseudo_random
+ [0] + data
4494 def encode_base_n(num
, n
, table
=None):
4495 FULL_TABLE
= '0123456789abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ'
4497 table
= FULL_TABLE
[:n
]
4500 raise ValueError('base %d exceeds table length %d' % (n
, len(table
)))
4507 ret
= table
[num
% n
] + ret
4512 def decode_packed_codes(code
):
4513 mobj
= re
.search(PACKED_CODES_RE
, code
)
4514 obfuscated_code
, base
, count
, symbols
= mobj
.groups()
4517 symbols
= symbols
.split('|')
4522 base_n_count
= encode_base_n(count
, base
)
4523 symbol_table
[base_n_count
] = symbols
[count
] or base_n_count
4526 r
'\b(\w+)\b', lambda mobj
: symbol_table
[mobj
.group(0)],
4530 def caesar(s
, alphabet
, shift
):
4535 alphabet
[(alphabet
.index(c
) + shift
) % l
] if c
in alphabet
else c
4540 return caesar(s
, r
'''!"#$%&'()*+,-./0123456789:;<=>?@ABCDEFGHIJKLMNOPQRSTUVWXYZ[\]^_`abcdefghijklmnopqrstuvwxyz{|}~''', 47)
4543 def parse_m3u8_attributes(attrib
):
4545 for (key
, val
) in re
.findall(r
'(?P<key>[A-Z0-9-]+)=(?P<val>"[^"]+"|[^",]+)(?:,|$)', attrib
):
4546 if val
.startswith('"'):
4552 def urshift(val
, n
):
4553 return val
>> n
if val
>= 0 else (val
+ 0x100000000) >> n
4556 # Based on png2str() written by @gdkchan and improved by @yokrysty
4557 # Originally posted at https://github.com/ytdl-org/youtube-dl/issues/9706
4558 def decode_png(png_data
):
4559 # Reference: https://www.w3.org/TR/PNG/
4560 header
= png_data
[8:]
4562 if png_data
[:8] != b
'\x89PNG\x0d\x0a\x1a\x0a' or header
[4:8] != b
'IHDR':
4563 raise IOError('Not a valid PNG file.')
4565 int_map
= {1: '>B', 2: '>H', 4: '>I'}
4566 unpack_integer
= lambda x
: compat_struct_unpack(int_map
[len(x
)], x
)[0]
4571 length
= unpack_integer(header
[:4])
4574 chunk_type
= header
[:4]
4577 chunk_data
= header
[:length
]
4578 header
= header
[length
:]
4580 header
= header
[4:] # Skip CRC
4588 ihdr
= chunks
[0]['data']
4590 width
= unpack_integer(ihdr
[:4])
4591 height
= unpack_integer(ihdr
[4:8])
4595 for chunk
in chunks
:
4596 if chunk
['type'] == b
'IDAT':
4597 idat
+= chunk
['data']
4600 raise IOError('Unable to read PNG data.')
4602 decompressed_data
= bytearray(zlib
.decompress(idat
))
4607 def _get_pixel(idx
):
4612 for y
in range(height
):
4613 basePos
= y
* (1 + stride
)
4614 filter_type
= decompressed_data
[basePos
]
4618 pixels
.append(current_row
)
4620 for x
in range(stride
):
4621 color
= decompressed_data
[1 + basePos
+ x
]
4622 basex
= y
* stride
+ x
4627 left
= _get_pixel(basex
- 3)
4629 up
= _get_pixel(basex
- stride
)
4631 if filter_type
== 1: # Sub
4632 color
= (color
+ left
) & 0xff
4633 elif filter_type
== 2: # Up
4634 color
= (color
+ up
) & 0xff
4635 elif filter_type
== 3: # Average
4636 color
= (color
+ ((left
+ up
) >> 1)) & 0xff
4637 elif filter_type
== 4: # Paeth
4643 c
= _get_pixel(basex
- stride
- 3)
4651 if pa
<= pb
and pa
<= pc
:
4652 color
= (color
+ a
) & 0xff
4654 color
= (color
+ b
) & 0xff
4656 color
= (color
+ c
) & 0xff
4658 current_row
.append(color
)
4660 return width
, height
, pixels
4663 def write_xattr(path
, key
, value
):
4664 # This mess below finds the best xattr tool for the job
4666 # try the pyxattr module...
4669 if hasattr(xattr
, 'set'): # pyxattr
4670 # Unicode arguments are not supported in python-pyxattr until
4672 # See https://github.com/ytdl-org/youtube-dl/issues/5498
4673 pyxattr_required_version
= '0.5.0'
4674 if version_tuple(xattr
.__version
__) < version_tuple(pyxattr_required_version
):
4675 # TODO: fallback to CLI tools
4676 raise XAttrUnavailableError(
4677 'python-pyxattr is detected but is too old. '
4678 'yt-dlp requires %s or above while your version is %s. '
4679 'Falling back to other xattr implementations' % (
4680 pyxattr_required_version
, xattr
.__version
__))
4682 setxattr
= xattr
.set
4684 setxattr
= xattr
.setxattr
4687 setxattr(path
, key
, value
)
4688 except EnvironmentError as e
:
4689 raise XAttrMetadataError(e
.errno
, e
.strerror
)
4692 if compat_os_name
== 'nt':
4693 # Write xattrs to NTFS Alternate Data Streams:
4694 # http://en.wikipedia.org/wiki/NTFS#Alternate_data_streams_.28ADS.29
4695 assert ':' not in key
4696 assert os
.path
.exists(path
)
4698 ads_fn
= path
+ ':' + key
4700 with open(ads_fn
, 'wb') as f
:
4702 except EnvironmentError as e
:
4703 raise XAttrMetadataError(e
.errno
, e
.strerror
)
4705 user_has_setfattr
= check_executable('setfattr', ['--version'])
4706 user_has_xattr
= check_executable('xattr', ['-h'])
4708 if user_has_setfattr
or user_has_xattr
:
4710 value
= value
.decode('utf-8')
4711 if user_has_setfattr
:
4712 executable
= 'setfattr'
4713 opts
= ['-n', key
, '-v', value
]
4714 elif user_has_xattr
:
4715 executable
= 'xattr'
4716 opts
= ['-w', key
, value
]
4718 cmd
= ([encodeFilename(executable
, True)]
4719 + [encodeArgument(o
) for o
in opts
]
4720 + [encodeFilename(path
, True)])
4724 cmd
, stdout
=subprocess
.PIPE
, stderr
=subprocess
.PIPE
, stdin
=subprocess
.PIPE
)
4725 except EnvironmentError as e
:
4726 raise XAttrMetadataError(e
.errno
, e
.strerror
)
4727 stdout
, stderr
= p
.communicate_or_kill()
4728 stderr
= stderr
.decode('utf-8', 'replace')
4729 if p
.returncode
!= 0:
4730 raise XAttrMetadataError(p
.returncode
, stderr
)
4733 # On Unix, and can't find pyxattr, setfattr, or xattr.
4734 if sys
.platform
.startswith('linux'):
4735 raise XAttrUnavailableError(
4736 "Couldn't find a tool to set the xattrs. "
4737 "Install either the python 'pyxattr' or 'xattr' "
4738 "modules, or the GNU 'attr' package "
4739 "(which contains the 'setfattr' tool).")
4741 raise XAttrUnavailableError(
4742 "Couldn't find a tool to set the xattrs. "
4743 "Install either the python 'xattr' module, "
4744 "or the 'xattr' binary.")
4747 def random_birthday(year_field
, month_field
, day_field
):
4748 start_date
= datetime
.date(1950, 1, 1)
4749 end_date
= datetime
.date(1995, 12, 31)
4750 offset
= random
.randint(0, (end_date
- start_date
).days
)
4751 random_date
= start_date
+ datetime
.timedelta(offset
)
4753 year_field
: str(random_date
.year
),
4754 month_field
: str(random_date
.month
),
4755 day_field
: str(random_date
.day
),
4759 # Templates for internet shortcut files, which are plain text files.
4760 DOT_URL_LINK_TEMPLATE
= '''
4765 DOT_WEBLOC_LINK_TEMPLATE
= '''
4766 <?xml version="1.0" encoding="UTF-8"?>
4767 <!DOCTYPE plist PUBLIC "-//Apple//DTD PLIST 1.0//EN" "http://www.apple.com/DTDs/PropertyList-1.0.dtd">
4768 <plist version="1.0">
4771 \t<string>%(url)s</string>
4776 DOT_DESKTOP_LINK_TEMPLATE
= '''
4786 'url': DOT_URL_LINK_TEMPLATE
,
4787 'desktop': DOT_DESKTOP_LINK_TEMPLATE
,
4788 'webloc': DOT_WEBLOC_LINK_TEMPLATE
,
4792 def iri_to_uri(iri
):
4794 Converts an IRI (Internationalized Resource Identifier, allowing Unicode characters) to a URI (Uniform Resource Identifier, ASCII-only).
4796 The function doesn't add an additional layer of escaping; e.g., it doesn't escape `%3C` as `%253C`. Instead, it percent-escapes characters with an underlying UTF-8 encoding *besides* those already escaped, leaving the URI intact.
4799 iri_parts
= compat_urllib_parse_urlparse(iri
)
4801 if '[' in iri_parts
.netloc
:
4802 raise ValueError('IPv6 URIs are not, yet, supported.')
4803 # Querying `.netloc`, when there's only one bracket, also raises a ValueError.
4805 # The `safe` argument values, that the following code uses, contain the characters that should not be percent-encoded. Everything else but letters, digits and '_.-' will be percent-encoded with an underlying UTF-8 encoding. Everything already percent-encoded will be left as is.
4808 if iri_parts
.username
:
4809 net_location
+= compat_urllib_parse_quote(iri_parts
.username
, safe
=r
"!$%&'()*+,~")
4810 if iri_parts
.password
is not None:
4811 net_location
+= ':' + compat_urllib_parse_quote(iri_parts
.password
, safe
=r
"!$%&'()*+,~")
4814 net_location
+= iri_parts
.hostname
.encode('idna').decode('utf-8') # Punycode for Unicode hostnames.
4815 # The 'idna' encoding produces ASCII text.
4816 if iri_parts
.port
is not None and iri_parts
.port
!= 80:
4817 net_location
+= ':' + str(iri_parts
.port
)
4819 return compat_urllib_parse_urlunparse(
4823 compat_urllib_parse_quote_plus(iri_parts
.path
, safe
=r
"!$%&'()*+,/:;=@|~"),
4825 # Unsure about the `safe` argument, since this is a legacy way of handling parameters.
4826 compat_urllib_parse_quote_plus(iri_parts
.params
, safe
=r
"!$%&'()*+,/:;=@|~"),
4828 # Not totally sure about the `safe` argument, since the source does not explicitly mention the query URI component.
4829 compat_urllib_parse_quote_plus(iri_parts
.query
, safe
=r
"!$%&'()*+,/:;=?@{|}~"),
4831 compat_urllib_parse_quote_plus(iri_parts
.fragment
, safe
=r
"!#$%&'()*+,/:;=?@{|}~")))
4833 # Source for `safe` arguments: https://url.spec.whatwg.org/#percent-encoded-bytes.
4836 def to_high_limit_path(path
):
4837 if sys
.platform
in ['win32', 'cygwin']:
4838 # Work around MAX_PATH limitation on Windows. The maximum allowed length for the individual path segments may still be quite limited.
4839 return r
'\\?\ '.rstrip() + os
.path
.abspath(path
)
4844 def format_field(obj
, field
=None, template
='%s', ignore
=(None, ''), default
='', func
=None):
4846 val
= obj
if obj
is not None else default
4848 val
= obj
.get(field
, default
)
4849 if func
and val
not in ignore
:
4851 return template
% val
if val
not in ignore
else default
4854 def clean_podcast_url(url
):
4855 return re
.sub(r
'''(?x)
4859 media\.blubrry\.com| # https://create.blubrry.com/resources/podcast-media-download-statistics/getting-started/
4862 (?:dts|www)\.podtrac\.com/(?:pts/)?redirect\.[0-9a-z]{3,4}| # http://analytics.podtrac.com/how-to-measure
4865 cn\.co| # https://podcorn.com/analytics-prefix/
4866 st\.fm # https://podsights.com/docs/
4871 _HEX_TABLE
= '0123456789abcdef'
4874 def random_uuidv4():
4875 return re
.sub(r
'[xy]', lambda x
: _HEX_TABLE
[random
.randint(0, 15)], 'xxxxxxxx-xxxx-4xxx-yxxx-xxxxxxxxxxxx')
4878 def make_dir(path
, to_screen
=None):
4880 dn
= os
.path
.dirname(path
)
4881 if dn
and not os
.path
.exists(dn
):
4884 except (OSError, IOError) as err
:
4885 if callable(to_screen
) is not None:
4886 to_screen('unable to create directory ' + error_to_compat_str(err
))
4890 def get_executable_path():
4891 from zipimport
import zipimporter
4892 if hasattr(sys
, 'frozen'): # Running from PyInstaller
4893 path
= os
.path
.dirname(sys
.executable
)
4894 elif isinstance(globals().get('__loader__'), zipimporter
): # Running from ZIP
4895 path
= os
.path
.join(os
.path
.dirname(__file__
), '../..')
4897 path
= os
.path
.join(os
.path
.dirname(__file__
), '..')
4898 return os
.path
.abspath(path
)
4901 def load_plugins(name
, suffix
, namespace
):
4904 plugins_spec
= importlib
.util
.spec_from_file_location(
4905 name
, os
.path
.join(get_executable_path(), 'ytdlp_plugins', name
, '__init__.py'))
4906 plugins
= importlib
.util
.module_from_spec(plugins_spec
)
4907 sys
.modules
[plugins_spec
.name
] = plugins
4908 plugins_spec
.loader
.exec_module(plugins
)
4909 for name
in dir(plugins
):
4910 if name
in namespace
:
4912 if not name
.endswith(suffix
):
4914 klass
= getattr(plugins
, name
)
4915 classes
[name
] = namespace
[name
] = klass
4916 except FileNotFoundError
:
4922 obj
, *path_list
, default
=None, expected_type
=None, get_all
=True,
4923 casesense
=True, is_user_input
=False, traverse_string
=False):
4924 ''' Traverse nested list/dict/tuple
4925 @param path_list A list of paths which are checked one by one.
4926 Each path is a list of keys where each key is a string,
4927 a function, a tuple of strings or "...".
4928 When a fuction is given, it takes the key as argument and
4929 returns whether the key matches or not. When a tuple is given,
4930 all the keys given in the tuple are traversed, and
4931 "..." traverses all the keys in the object
4932 @param default Default value to return
4933 @param expected_type Only accept final value of this type (Can also be any callable)
4934 @param get_all Return all the values obtained from a path or only the first one
4935 @param casesense Whether to consider dictionary keys as case sensitive
4936 @param is_user_input Whether the keys are generated from user input. If True,
4937 strings are converted to int/slice if necessary
4938 @param traverse_string Whether to traverse inside strings. If True, any
4939 non-compatible object will also be converted into a string
4943 _lower
= lambda k
: (k
.lower() if isinstance(k
, str) else k
)
4944 path_list
= (map(_lower
, variadic(path
)) for path
in path_list
)
4946 def _traverse_obj(obj
, path
, _current_depth
=0):
4948 path
= tuple(variadic(path
))
4949 for i
, key
in enumerate(path
):
4952 if isinstance(key
, (list, tuple)):
4953 obj
= [_traverse_obj(obj
, sub_key
, _current_depth
) for sub_key
in key
]
4956 obj
= (obj
.values() if isinstance(obj
, dict)
4957 else obj
if isinstance(obj
, (list, tuple, LazyList
))
4958 else str(obj
) if traverse_string
else [])
4960 depth
= max(depth
, _current_depth
)
4961 return [_traverse_obj(inner_obj
, path
[i
+ 1:], _current_depth
) for inner_obj
in obj
]
4963 if isinstance(obj
, (list, tuple, LazyList
)):
4964 obj
= enumerate(obj
)
4965 elif isinstance(obj
, dict):
4968 if not traverse_string
:
4972 depth
= max(depth
, _current_depth
)
4973 return [_traverse_obj(v
, path
[i
+ 1:], _current_depth
) for k
, v
in obj
if key(k
)]
4974 elif isinstance(obj
, dict) and not (is_user_input
and key
== ':'):
4975 obj
= (obj
.get(key
) if casesense
or (key
in obj
)
4976 else next((v
for k
, v
in obj
.items() if _lower(k
) == key
), None))
4979 key
= (int_or_none(key
) if ':' not in key
4980 else slice(*map(int_or_none
, key
.split(':'))))
4981 if key
== slice(None):
4982 return _traverse_obj(obj
, (..., *path
[i
+ 1:]), _current_depth
)
4983 if not isinstance(key
, (int, slice)):
4985 if not isinstance(obj
, (list, tuple, LazyList
)):
4986 if not traverse_string
:
4995 if isinstance(expected_type
, type):
4996 type_test
= lambda val
: val
if isinstance(val
, expected_type
) else None
4997 elif expected_type
is not None:
4998 type_test
= expected_type
5000 type_test
= lambda val
: val
5002 for path
in path_list
:
5004 val
= _traverse_obj(obj
, path
)
5007 for _
in range(depth
- 1):
5008 val
= itertools
.chain
.from_iterable(v
for v
in val
if v
is not None)
5009 val
= [v
for v
in map(type_test
, val
) if v
is not None]
5011 return val
if get_all
else val
[0]
5013 val
= type_test(val
)
5020 def traverse_dict(dictn
, keys
, casesense
=True):
5021 write_string('DeprecationWarning: yt_dlp.utils.traverse_dict is deprecated '
5022 'and may be removed in a future version. Use yt_dlp.utils.traverse_obj instead')
5023 return traverse_obj(dictn
, keys
, casesense
=casesense
, is_user_input
=True, traverse_string
=True)
5026 def variadic(x
, allowed_types
=(str, bytes)):
5027 return x
if isinstance(x
, collections
.abc
.Iterable
) and not isinstance(x
, allowed_types
) else (x
,)
5030 # create a JSON Web Signature (jws) with HS256 algorithm
5031 # the resulting format is in JWS Compact Serialization
5032 # implemented following JWT https://www.rfc-editor.org/rfc/rfc7519.html
5033 # implemented following JWS https://www.rfc-editor.org/rfc/rfc7515.html
5034 def jwt_encode_hs256(payload_data
, key
, headers
={}):
5040 header_data
.update(headers
)
5041 header_b64
= base64
.b64encode(json
.dumps(header_data
).encode('utf-8'))
5042 payload_b64
= base64
.b64encode(json
.dumps(payload_data
).encode('utf-8'))
5043 h
= hmac
.new(key
.encode('utf-8'), header_b64
+ b
'.' + payload_b64
, hashlib
.sha256
)
5044 signature_b64
= base64
.b64encode(h
.digest())
5045 token
= header_b64
+ b
'.' + payload_b64
+ b
'.' + signature_b64
5049 # can be extended in future to verify the signature and parse header and return the algorithm used if it's not HS256
5050 def jwt_decode_hs256(jwt
):
5051 header_b64
, payload_b64
, signature_b64
= jwt
.split('.')
5052 payload_data
= json
.loads(base64
.urlsafe_b64decode(payload_b64
))
5056 def supports_terminal_sequences(stream
):
5057 if compat_os_name
== 'nt':
5058 from .compat
import WINDOWS_VT_MODE
# Must be imported locally
5059 if not WINDOWS_VT_MODE
or get_windows_version() < (10, 0, 10586):
5061 elif not os
.getenv('TERM'):
5064 return stream
.isatty()
5065 except BaseException
:
5069 _terminal_sequences_re
= re
.compile('\033\\[[^m]+m')
5072 def remove_terminal_sequences(string
):
5073 return _terminal_sequences_re
.sub('', string
)
5076 def number_of_digits(number
):
5077 return len('%d' % number
)
5080 def join_nonempty(*values
, delim
='-', from_dict
=None):
5081 if from_dict
is not None:
5082 values
= map(from_dict
.get
, values
)
5083 return delim
.join(map(str, filter(None, values
)))