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')))
216 '%Y-%m-%d %H:%M:%S.%f',
217 '%Y-%m-%d %H:%M:%S:%f',
220 '%Y-%m-%dT%H:%M:%SZ',
221 '%Y-%m-%dT%H:%M:%S.%fZ',
222 '%Y-%m-%dT%H:%M:%S.%f0Z',
224 '%Y-%m-%dT%H:%M:%S.%f',
227 '%b %d %Y at %H:%M:%S',
229 '%B %d %Y at %H:%M:%S',
233 DATE_FORMATS_DAY_FIRST
= list(DATE_FORMATS
)
234 DATE_FORMATS_DAY_FIRST
.extend([
243 DATE_FORMATS_MONTH_FIRST
= list(DATE_FORMATS
)
244 DATE_FORMATS_MONTH_FIRST
.extend([
252 PACKED_CODES_RE
= r
"}\('(.+)',(\d+),(\d+),'([^']+)'\.split\('\|'\)"
253 JSON_LD_RE
= r
'(?is)<script[^>]+type=(["\']?
)application
/ld\
+json\
1[^
>]*>(?P
<json_ld
>.+?
)</script
>'
256 def preferredencoding():
257 """Get preferred encoding.
259 Returns the best encoding scheme for the system, based on
260 locale.getpreferredencoding() and some further tweaks.
263 pref = locale.getpreferredencoding()
271 def write_json_file(obj, fn):
272 """ Encode obj as JSON and write it to fn, atomically if possible """
274 fn = encodeFilename(fn)
275 if sys.version_info < (3, 0) and sys.platform != 'win32
':
276 encoding = get_filesystem_encoding()
277 # os.path.basename returns a bytes object, but NamedTemporaryFile
278 # will fail if the filename contains non ascii characters unless we
279 # use a unicode object
280 path_basename = lambda f: os.path.basename(fn).decode(encoding)
281 # the same for os.path.dirname
282 path_dirname = lambda f: os.path.dirname(fn).decode(encoding)
284 path_basename = os.path.basename
285 path_dirname = os.path.dirname
289 'prefix
': path_basename(fn) + '.',
290 'dir': path_dirname(fn),
294 # In Python 2.x, json.dump expects a bytestream.
295 # In Python 3.x, it writes to a character stream
296 if sys.version_info < (3, 0):
304 tf = tempfile.NamedTemporaryFile(**compat_kwargs(args))
309 if sys.platform == 'win32
':
310 # Need to remove existing file on Windows, else os.rename raises
311 # WindowsError or FileExistsError.
319 os.chmod(tf.name, 0o666 & ~mask)
322 os.rename(tf.name, fn)
331 if sys.version_info >= (2, 7):
332 def find_xpath_attr(node, xpath, key, val=None):
333 """ Find the xpath xpath[@key=val] """
334 assert re.match(r'^
[a
-zA
-Z_
-]+$
', key)
335 expr = xpath + ('[@%s]' % key if val is None else "[@%s='%s']" % (key, val))
336 return node.find(expr)
338 def find_xpath_attr(node, xpath, key, val=None):
339 for f in node.findall(compat_xpath(xpath)):
340 if key not in f.attrib:
342 if val is None or f.attrib.get(key) == val:
346 # On python2.6 the xml.etree.ElementTree.Element methods don't support
347 # the namespace parameter
350 def xpath_with_ns(path
, ns_map
):
351 components
= [c
.split(':') for c
in path
.split('/')]
355 replaced
.append(c
[0])
358 replaced
.append('{%s}%s' % (ns_map
[ns
], tag
))
359 return '/'.join(replaced
)
362 def xpath_element(node
, xpath
, name
=None, fatal
=False, default
=NO_DEFAULT
):
363 def _find_xpath(xpath
):
364 return node
.find(compat_xpath(xpath
))
366 if isinstance(xpath
, (str, compat_str
)):
367 n
= _find_xpath(xpath
)
375 if default
is not NO_DEFAULT
:
378 name
= xpath
if name
is None else name
379 raise ExtractorError('Could not find XML element %s' % name
)
385 def xpath_text(node
, xpath
, name
=None, fatal
=False, default
=NO_DEFAULT
):
386 n
= xpath_element(node
, xpath
, name
, fatal
=fatal
, default
=default
)
387 if n
is None or n
== default
:
390 if default
is not NO_DEFAULT
:
393 name
= xpath
if name
is None else name
394 raise ExtractorError('Could not find XML element\'s text %s' % name
)
400 def xpath_attr(node
, xpath
, key
, name
=None, fatal
=False, default
=NO_DEFAULT
):
401 n
= find_xpath_attr(node
, xpath
, key
)
403 if default
is not NO_DEFAULT
:
406 name
= '%s[@%s]' % (xpath
, key
) if name
is None else name
407 raise ExtractorError('Could not find XML attribute %s' % name
)
413 def get_element_by_id(id, html
):
414 """Return the content of the tag with the specified ID in the passed HTML document"""
415 return get_element_by_attribute('id', id, html
)
418 def get_element_by_class(class_name
, html
):
419 """Return the content of the first tag with the specified class in the passed HTML document"""
420 retval
= get_elements_by_class(class_name
, html
)
421 return retval
[0] if retval
else None
424 def get_element_by_attribute(attribute
, value
, html
, escape_value
=True):
425 retval
= get_elements_by_attribute(attribute
, value
, html
, escape_value
)
426 return retval
[0] if retval
else None
429 def get_elements_by_class(class_name
, html
):
430 """Return the content of all tags with the specified class in the passed HTML document as a list"""
431 return get_elements_by_attribute(
432 'class', r
'[^\'"]*\b%s\b[^\'"]*' % re.escape(class_name),
433 html, escape_value=False)
436 def get_elements_by_attribute(attribute, value, html, escape_value=True):
437 """Return the content of the tag with the specified attribute in the passed HTML document"""
439 value = re.escape(value) if escape_value else value
442 for m in re.finditer(r'''(?xs)
444 (?:\s+[a-zA-Z0-9:._-]+(?:=[a-zA-Z0-9:._-]*|="[^"]*"|='[^
']*'|
))*?
446 (?:\s+[a-zA-Z0-9:._-]+(?:=[a-zA-Z0-9:._-]*|="[^
"]*"|
='[^']*'|))*?
450 ''' % (re.escape(attribute), value), html):
451 res = m.group('content
')
453 if res.startswith('"') or res.startswith("'"):
456 retlist.append(unescapeHTML(res))
461 class HTMLAttributeParser(compat_HTMLParser):
462 """Trivial HTML parser to gather the attributes for a single element"""
466 compat_HTMLParser.__init__(self)
468 def handle_starttag(self, tag, attrs):
469 self.attrs = dict(attrs)
472 class HTMLListAttrsParser(compat_HTMLParser):
473 """HTML parser to gather the attributes for the elements of a list"""
476 compat_HTMLParser.__init__(self)
480 def handle_starttag(self, tag, attrs):
481 if tag == 'li
' and self._level == 0:
482 self.items.append(dict(attrs))
485 def handle_endtag(self, tag):
489 def extract_attributes(html_element):
490 """Given a string for an HTML element such as
492 a="foo" B="bar" c="&98;az" d=boz
493 empty= noval entity="&"
496 Decode and return a dictionary of attributes.
498 'a
': 'foo
', 'b
': 'bar
', c: 'baz
', d: 'boz
',
499 'empty
': '', 'noval
': None, 'entity
': '&',
500 'sq
': '"', 'dq': '\''
502 NB HTMLParser is stricter in Python 2.6 & 3.2 than in later versions,
503 but the cases in the unit test will work for all of 2.6, 2.7, 3.2-3.5.
505 parser = HTMLAttributeParser()
507 parser.feed(html_element)
509 # Older Python may throw HTMLParseError in case of malformed HTML
510 except compat_HTMLParseError:
515 def parse_list(webpage):
516 """Given a string for an series of HTML <li> elements,
517 return a dictionary of their attributes"""
518 parser = HTMLListAttrsParser()
524 def clean_html(html):
525 """Clean an HTML snippet into a readable string"""
527 if html is None: # Convenience for sanitizing descriptions etc.
531 html = html.replace('\n', ' ')
532 html = re.sub(r'(?u)\s*<\s*br\s*/?\s*>\s*', '\n', html)
533 html = re.sub(r'(?u)<\s*/\s*p\s*>\s*<\s*p[^>]*>', '\n', html)
535 html = re.sub('<.*?>', '', html)
536 # Replace html entities
537 html = unescapeHTML(html)
541 def sanitize_open(filename, open_mode):
542 """Try to open the given filename, and slightly tweak it if this fails.
544 Attempts to open the given filename. If this fails, it tries to change
545 the filename slightly, step by step, until it's either able to open it
546 or it fails and raises a final exception, like the standard open()
549 It returns the tuple (stream, definitive_file_name).
553 if sys.platform == 'win32':
555 msvcrt.setmode(sys.stdout.fileno(), os.O_BINARY)
556 return (sys.stdout.buffer if hasattr(sys.stdout, 'buffer') else sys.stdout, filename)
557 stream = open(encodeFilename(filename), open_mode)
558 return (stream, filename)
559 except (IOError, OSError) as err:
560 if err.errno in (errno.EACCES,):
563 # In case of error, try to remove win32 forbidden chars
564 alt_filename = sanitize_path(filename)
565 if alt_filename == filename:
568 # An exception here should be caught in the caller
569 stream = open(encodeFilename(alt_filename), open_mode)
570 return (stream, alt_filename)
573 def timeconvert(timestr):
574 """Convert RFC 2822 defined time string into system timestamp"""
576 timetuple = email.utils.parsedate_tz(timestr)
577 if timetuple is not None:
578 timestamp = email.utils.mktime_tz(timetuple)
582 def sanitize_filename(s, restricted=False, is_id=False):
583 """Sanitizes a string so it could be used as part of a filename.
584 If restricted is set, use a stricter subset of allowed characters.
585 Set is_id if this is not an arbitrary string, but an ID that should be kept
588 def replace_insane(char):
589 if restricted and char in ACCENT_CHARS:
590 return ACCENT_CHARS[char]
591 elif not restricted and char == '\n':
593 elif char == '?' or ord(char) < 32 or ord(char) == 127:
596 return '' if restricted else '\''
598 return '_
-' if restricted else ' -'
599 elif char in '\\/|
*<>':
601 if restricted and (char in '!&\'()[]{}$
;`^
,#' or char.isspace()):
603 if restricted
and ord(char
) > 127:
610 s
= re
.sub(r
'[0-9]+(?::[0-9]+)+', lambda m
: m
.group(0).replace(':', '_'), s
)
611 result
= ''.join(map(replace_insane
, s
))
613 while '__' in result
:
614 result
= result
.replace('__', '_')
615 result
= result
.strip('_')
616 # Common case of "Foreign band name - English song title"
617 if restricted
and result
.startswith('-_'):
619 if result
.startswith('-'):
620 result
= '_' + result
[len('-'):]
621 result
= result
.lstrip('.')
627 def sanitize_path(s
, force
=False):
628 """Sanitizes and normalizes path on Windows"""
629 if sys
.platform
== 'win32':
631 drive_or_unc
, _
= os
.path
.splitdrive(s
)
632 if sys
.version_info
< (2, 7) and not drive_or_unc
:
633 drive_or_unc
, _
= os
.path
.splitunc(s
)
639 norm_path
= os
.path
.normpath(remove_start(s
, drive_or_unc
)).split(os
.path
.sep
)
643 path_part
if path_part
in ['.', '..'] else re
.sub(r
'(?:[/<>:"\|\\?\*]|[\s.]$)', '#', path_part
)
644 for path_part
in norm_path
]
646 sanitized_path
.insert(0, drive_or_unc
+ os
.path
.sep
)
647 elif force
and s
[0] == os
.path
.sep
:
648 sanitized_path
.insert(0, os
.path
.sep
)
649 return os
.path
.join(*sanitized_path
)
652 def sanitize_url(url
):
653 # Prepend protocol-less URLs with `http:` scheme in order to mitigate
654 # the number of unwanted failures due to missing protocol
655 if url
.startswith('//'):
656 return 'http:%s' % url
657 # Fix some common typos seen so far
659 # https://github.com/ytdl-org/youtube-dl/issues/15649
660 (r
'^httpss://', r
'https://'),
661 # https://bx1.be/lives/direct-tv/
662 (r
'^rmtp([es]?)://', r
'rtmp\1://'),
664 for mistake
, fixup
in COMMON_TYPOS
:
665 if re
.match(mistake
, url
):
666 return re
.sub(mistake
, fixup
, url
)
670 def extract_basic_auth(url
):
671 parts
= compat_urlparse
.urlsplit(url
)
672 if parts
.username
is None:
674 url
= compat_urlparse
.urlunsplit(parts
._replace
(netloc
=(
675 parts
.hostname
if parts
.port
is None
676 else '%s:%d' % (parts
.hostname
, parts
.port
))))
677 auth_payload
= base64
.b64encode(
678 ('%s:%s' % (parts
.username
, parts
.password
or '')).encode('utf-8'))
679 return url
, 'Basic ' + auth_payload
.decode('utf-8')
682 def sanitized_Request(url
, *args
, **kwargs
):
683 url
, auth_header
= extract_basic_auth(escape_url(sanitize_url(url
)))
684 if auth_header
is not None:
685 headers
= args
[1] if len(args
) >= 2 else kwargs
.setdefault('headers', {})
686 headers
['Authorization'] = auth_header
687 return compat_urllib_request
.Request(url
, *args
, **kwargs
)
691 """Expand shell variables and ~"""
692 return os
.path
.expandvars(compat_expanduser(s
))
695 def orderedSet(iterable
):
696 """ Remove all duplicates from the input iterable """
704 def _htmlentity_transform(entity_with_semicolon
):
705 """Transforms an HTML entity to a character."""
706 entity
= entity_with_semicolon
[:-1]
708 # Known non-numeric HTML entity
709 if entity
in compat_html_entities
.name2codepoint
:
710 return compat_chr(compat_html_entities
.name2codepoint
[entity
])
712 # TODO: HTML5 allows entities without a semicolon. For example,
713 # 'Éric' should be decoded as 'Éric'.
714 if entity_with_semicolon
in compat_html_entities_html5
:
715 return compat_html_entities_html5
[entity_with_semicolon
]
717 mobj
= re
.match(r
'#(x[0-9a-fA-F]+|[0-9]+)', entity
)
719 numstr
= mobj
.group(1)
720 if numstr
.startswith('x'):
722 numstr
= '0%s' % numstr
725 # See https://github.com/ytdl-org/youtube-dl/issues/7518
727 return compat_chr(int(numstr
, base
))
731 # Unknown entity in name, return its literal representation
732 return '&%s;' % entity
738 assert type(s
) == compat_str
741 r
'&([^&;]+;)', lambda m
: _htmlentity_transform(m
.group(1)), s
)
744 def escapeHTML(text
):
747 .replace('&', '&')
748 .replace('<', '<')
749 .replace('>', '>')
750 .replace('"', '"')
751 .replace("'", ''')
755 def process_communicate_or_kill(p
, *args
, **kwargs
):
757 return p
.communicate(*args
, **kwargs
)
758 except BaseException
: # Including KeyboardInterrupt
764 class Popen(subprocess
.Popen
):
765 if sys
.platform
== 'win32':
766 _startupinfo
= subprocess
.STARTUPINFO()
767 _startupinfo
.dwFlags |
= subprocess
.STARTF_USESHOWWINDOW
771 def __init__(self
, *args
, **kwargs
):
772 super(Popen
, self
).__init
__(*args
, **kwargs
, startupinfo
=self
._startupinfo
)
774 def communicate_or_kill(self
, *args
, **kwargs
):
775 return process_communicate_or_kill(self
, *args
, **kwargs
)
778 def get_subprocess_encoding():
779 if sys
.platform
== 'win32' and sys
.getwindowsversion()[0] >= 5:
780 # For subprocess calls, encode with locale encoding
781 # Refer to http://stackoverflow.com/a/9951851/35070
782 encoding
= preferredencoding()
784 encoding
= sys
.getfilesystemencoding()
790 def encodeFilename(s
, for_subprocess
=False):
792 @param s The name of the file
795 assert type(s
) == compat_str
797 # Python 3 has a Unicode API
798 if sys
.version_info
>= (3, 0):
801 # Pass '' directly to use Unicode APIs on Windows 2000 and up
802 # (Detecting Windows NT 4 is tricky because 'major >= 4' would
803 # match Windows 9x series as well. Besides, NT 4 is obsolete.)
804 if not for_subprocess
and sys
.platform
== 'win32' and sys
.getwindowsversion()[0] >= 5:
807 # Jython assumes filenames are Unicode strings though reported as Python 2.x compatible
808 if sys
.platform
.startswith('java'):
811 return s
.encode(get_subprocess_encoding(), 'ignore')
814 def decodeFilename(b
, for_subprocess
=False):
816 if sys
.version_info
>= (3, 0):
819 if not isinstance(b
, bytes):
822 return b
.decode(get_subprocess_encoding(), 'ignore')
825 def encodeArgument(s
):
826 if not isinstance(s
, compat_str
):
827 # Legacy code that uses byte strings
828 # Uncomment the following line after fixing all post processors
829 # assert False, 'Internal error: %r should be of type %r, is %r' % (s, compat_str, type(s))
830 s
= s
.decode('ascii')
831 return encodeFilename(s
, True)
834 def decodeArgument(b
):
835 return decodeFilename(b
, True)
838 def decodeOption(optval
):
841 if isinstance(optval
, bytes):
842 optval
= optval
.decode(preferredencoding())
844 assert isinstance(optval
, compat_str
)
848 _timetuple
= collections
.namedtuple('Time', ('hours', 'minutes', 'seconds', 'milliseconds'))
851 def timetuple_from_msec(msec
):
852 secs
, msec
= divmod(msec
, 1000)
853 mins
, secs
= divmod(secs
, 60)
854 hrs
, mins
= divmod(mins
, 60)
855 return _timetuple(hrs
, mins
, secs
, msec
)
858 def formatSeconds(secs
, delim
=':', msec
=False):
859 time
= timetuple_from_msec(secs
* 1000)
861 ret
= '%d%s%02d%s%02d' % (time
.hours
, delim
, time
.minutes
, delim
, time
.seconds
)
863 ret
= '%d%s%02d' % (time
.minutes
, delim
, time
.seconds
)
865 ret
= '%d' % time
.seconds
866 return '%s.%03d' % (ret
, time
.milliseconds
) if msec
else ret
869 def _ssl_load_windows_store_certs(ssl_context
, storename
):
870 # Code adapted from _load_windows_store_certs in https://github.com/python/cpython/blob/main/Lib/ssl.py
872 certs
= [cert
for cert
, encoding
, trust
in ssl
.enum_certificates(storename
)
873 if encoding
== 'x509_asn' and (
874 trust
is True or ssl
.Purpose
.SERVER_AUTH
.oid
in trust
)]
875 except PermissionError
:
879 ssl_context
.load_verify_locations(cadata
=cert
)
884 def make_HTTPS_handler(params
, **kwargs
):
885 opts_check_certificate
= not params
.get('nocheckcertificate')
886 context
= ssl
.SSLContext(ssl
.PROTOCOL_TLS_CLIENT
)
887 context
.check_hostname
= opts_check_certificate
888 context
.verify_mode
= ssl
.CERT_REQUIRED
if opts_check_certificate
else ssl
.CERT_NONE
889 if opts_check_certificate
:
891 context
.load_default_certs()
892 # Work around the issue in load_default_certs when there are bad certificates. See:
893 # https://github.com/yt-dlp/yt-dlp/issues/1060,
894 # https://bugs.python.org/issue35665, https://bugs.python.org/issue45312
896 # enum_certificates is not present in mingw python. See https://github.com/yt-dlp/yt-dlp/issues/1151
897 if sys
.platform
== 'win32' and hasattr(ssl
, 'enum_certificates'):
898 # Create a new context to discard any certificates that were already loaded
899 context
= ssl
.SSLContext(ssl
.PROTOCOL_TLS_CLIENT
)
900 context
.check_hostname
, context
.verify_mode
= True, ssl
.CERT_REQUIRED
901 for storename
in ('CA', 'ROOT'):
902 _ssl_load_windows_store_certs(context
, storename
)
903 context
.set_default_verify_paths()
904 return YoutubeDLHTTPSHandler(params
, context
=context
, **kwargs
)
907 def bug_reports_message(before
=';'):
908 if ytdl_is_updateable():
909 update_cmd
= 'type yt-dlp -U to update'
911 update_cmd
= 'see https://github.com/yt-dlp/yt-dlp on how to update'
912 msg
= 'please report this issue on https://github.com/yt-dlp/yt-dlp .'
913 msg
+= ' Make sure you are using the latest version; %s.' % update_cmd
914 msg
+= ' Be sure to call yt-dlp with the --verbose flag and include its complete output.'
916 before
= before
.rstrip()
917 if not before
or before
.endswith(('.', '!', '?')):
918 msg
= msg
[0].title() + msg
[1:]
920 return (before
+ ' ' if before
else '') + msg
923 class YoutubeDLError(Exception):
924 """Base exception for YoutubeDL errors."""
927 def __init__(self
, msg
=None):
930 elif self
.msg
is None:
931 self
.msg
= type(self
).__name
__
932 super().__init
__(self
.msg
)
935 network_exceptions
= [compat_urllib_error
.URLError
, compat_http_client
.HTTPException
, socket
.error
]
936 if hasattr(ssl
, 'CertificateError'):
937 network_exceptions
.append(ssl
.CertificateError
)
938 network_exceptions
= tuple(network_exceptions
)
941 class ExtractorError(YoutubeDLError
):
942 """Error during info extraction."""
944 def __init__(self
, msg
, tb
=None, expected
=False, cause
=None, video_id
=None, ie
=None):
945 """ tb, if given, is the original traceback (so that it can be printed out).
946 If expected is set, this is a normal error message and most likely not a bug in yt-dlp.
948 if sys
.exc_info()[0] in network_exceptions
:
953 self
.expected
= expected
955 self
.video_id
= video_id
957 self
.exc_info
= sys
.exc_info() # preserve original exception
959 super(ExtractorError
, self
).__init
__(''.join((
960 format_field(ie
, template
='[%s] '),
961 format_field(video_id
, template
='%s: '),
963 format_field(cause
, template
=' (caused by %r)'),
964 '' if expected
else bug_reports_message())))
966 def format_traceback(self
):
967 if self
.traceback
is None:
969 return ''.join(traceback
.format_tb(self
.traceback
))
972 class UnsupportedError(ExtractorError
):
973 def __init__(self
, url
):
974 super(UnsupportedError
, self
).__init
__(
975 'Unsupported URL: %s' % url
, expected
=True)
979 class RegexNotFoundError(ExtractorError
):
980 """Error when a regex didn't match"""
984 class GeoRestrictedError(ExtractorError
):
985 """Geographic restriction Error exception.
987 This exception may be thrown when a video is not available from your
988 geographic location due to geographic restrictions imposed by a website.
991 def __init__(self
, msg
, countries
=None, **kwargs
):
992 kwargs
['expected'] = True
993 super(GeoRestrictedError
, self
).__init
__(msg
, **kwargs
)
994 self
.countries
= countries
997 class DownloadError(YoutubeDLError
):
998 """Download Error exception.
1000 This exception may be thrown by FileDownloader objects if they are not
1001 configured to continue on errors. They will contain the appropriate
1005 def __init__(self
, msg
, exc_info
=None):
1006 """ exc_info, if given, is the original exception that caused the trouble (as returned by sys.exc_info()). """
1007 super(DownloadError
, self
).__init
__(msg
)
1008 self
.exc_info
= exc_info
1011 class EntryNotInPlaylist(YoutubeDLError
):
1012 """Entry not in playlist exception.
1014 This exception will be thrown by YoutubeDL when a requested entry
1015 is not found in the playlist info_dict
1017 msg
= 'Entry not found in info'
1020 class SameFileError(YoutubeDLError
):
1021 """Same File exception.
1023 This exception will be thrown by FileDownloader objects if they detect
1024 multiple files would have to be downloaded to the same file on disk.
1026 msg
= 'Fixed output name but more than one file to download'
1028 def __init__(self
, filename
=None):
1029 if filename
is not None:
1030 self
.msg
+= f
': {filename}'
1031 super().__init
__(self
.msg
)
1034 class PostProcessingError(YoutubeDLError
):
1035 """Post Processing exception.
1037 This exception may be raised by PostProcessor's .run() method to
1038 indicate an error in the postprocessing task.
1042 class DownloadCancelled(YoutubeDLError
):
1043 """ Exception raised when the download queue should be interrupted """
1044 msg
= 'The download was cancelled'
1047 class ExistingVideoReached(DownloadCancelled
):
1048 """ --break-on-existing triggered """
1049 msg
= 'Encountered a video that is already in the archive, stopping due to --break-on-existing'
1052 class RejectedVideoReached(DownloadCancelled
):
1053 """ --break-on-reject triggered """
1054 msg
= 'Encountered a video that did not match filter, stopping due to --break-on-reject'
1057 class MaxDownloadsReached(DownloadCancelled
):
1058 """ --max-downloads limit has been reached. """
1059 msg
= 'Maximum number of downloads reached, stopping due to --max-downloads'
1062 class ReExtractInfo(YoutubeDLError
):
1063 """ Video info needs to be re-extracted. """
1065 def __init__(self
, msg
, expected
=False):
1066 super().__init
__(msg
)
1067 self
.expected
= expected
1070 class ThrottledDownload(ReExtractInfo
):
1071 """ Download speed below --throttled-rate. """
1072 msg
= 'The download speed is below throttle limit'
1075 super().__init
__(self
.msg
, expected
=False)
1078 class UnavailableVideoError(YoutubeDLError
):
1079 """Unavailable Format exception.
1081 This exception will be thrown when a video is requested
1082 in a format that is not available for that video.
1084 msg
= 'Unable to download video'
1086 def __init__(self
, err
=None):
1088 self
.msg
+= f
': {err}'
1089 super().__init
__(self
.msg
)
1092 class ContentTooShortError(YoutubeDLError
):
1093 """Content Too Short exception.
1095 This exception may be raised by FileDownloader objects when a file they
1096 download is too small for what the server announced first, indicating
1097 the connection was probably interrupted.
1100 def __init__(self
, downloaded
, expected
):
1101 super(ContentTooShortError
, self
).__init
__(
1102 'Downloaded {0} bytes, expected {1} bytes'.format(downloaded
, expected
)
1105 self
.downloaded
= downloaded
1106 self
.expected
= expected
1109 class XAttrMetadataError(YoutubeDLError
):
1110 def __init__(self
, code
=None, msg
='Unknown error'):
1111 super(XAttrMetadataError
, self
).__init
__(msg
)
1115 # Parsing code and msg
1116 if (self
.code
in (errno
.ENOSPC
, errno
.EDQUOT
)
1117 or 'No space left' in self
.msg
or 'Disk quota exceeded' in self
.msg
):
1118 self
.reason
= 'NO_SPACE'
1119 elif self
.code
== errno
.E2BIG
or 'Argument list too long' in self
.msg
:
1120 self
.reason
= 'VALUE_TOO_LONG'
1122 self
.reason
= 'NOT_SUPPORTED'
1125 class XAttrUnavailableError(YoutubeDLError
):
1129 def _create_http_connection(ydl_handler
, http_class
, is_https
, *args
, **kwargs
):
1130 # Working around python 2 bug (see http://bugs.python.org/issue17849) by limiting
1131 # expected HTTP responses to meet HTTP/1.0 or later (see also
1132 # https://github.com/ytdl-org/youtube-dl/issues/6727)
1133 if sys
.version_info
< (3, 0):
1134 kwargs
['strict'] = True
1135 hc
= http_class(*args
, **compat_kwargs(kwargs
))
1136 source_address
= ydl_handler
._params
.get('source_address')
1138 if source_address
is not None:
1139 # This is to workaround _create_connection() from socket where it will try all
1140 # address data from getaddrinfo() including IPv6. This filters the result from
1141 # getaddrinfo() based on the source_address value.
1142 # This is based on the cpython socket.create_connection() function.
1143 # https://github.com/python/cpython/blob/master/Lib/socket.py#L691
1144 def _create_connection(address
, timeout
=socket
._GLOBAL
_DEFAULT
_TIMEOUT
, source_address
=None):
1145 host
, port
= address
1147 addrs
= socket
.getaddrinfo(host
, port
, 0, socket
.SOCK_STREAM
)
1148 af
= socket
.AF_INET
if '.' in source_address
[0] else socket
.AF_INET6
1149 ip_addrs
= [addr
for addr
in addrs
if addr
[0] == af
]
1150 if addrs
and not ip_addrs
:
1151 ip_version
= 'v4' if af
== socket
.AF_INET
else 'v6'
1153 "No remote IP%s addresses available for connect, can't use '%s' as source address"
1154 % (ip_version
, source_address
[0]))
1155 for res
in ip_addrs
:
1156 af
, socktype
, proto
, canonname
, sa
= res
1159 sock
= socket
.socket(af
, socktype
, proto
)
1160 if timeout
is not socket
._GLOBAL
_DEFAULT
_TIMEOUT
:
1161 sock
.settimeout(timeout
)
1162 sock
.bind(source_address
)
1164 err
= None # Explicitly break reference cycle
1166 except socket
.error
as _
:
1168 if sock
is not None:
1173 raise socket
.error('getaddrinfo returns an empty list')
1174 if hasattr(hc
, '_create_connection'):
1175 hc
._create
_connection
= _create_connection
1176 sa
= (source_address
, 0)
1177 if hasattr(hc
, 'source_address'): # Python 2.7+
1178 hc
.source_address
= sa
1180 def _hc_connect(self
, *args
, **kwargs
):
1181 sock
= _create_connection(
1182 (self
.host
, self
.port
), self
.timeout
, sa
)
1184 self
.sock
= ssl
.wrap_socket(
1185 sock
, self
.key_file
, self
.cert_file
,
1186 ssl_version
=ssl
.PROTOCOL_TLSv1
)
1189 hc
.connect
= functools
.partial(_hc_connect
, hc
)
1194 def handle_youtubedl_headers(headers
):
1195 filtered_headers
= headers
1197 if 'Youtubedl-no-compression' in filtered_headers
:
1198 filtered_headers
= dict((k
, v
) for k
, v
in filtered_headers
.items() if k
.lower() != 'accept-encoding')
1199 del filtered_headers
['Youtubedl-no-compression']
1201 return filtered_headers
1204 class YoutubeDLHandler(compat_urllib_request
.HTTPHandler
):
1205 """Handler for HTTP requests and responses.
1207 This class, when installed with an OpenerDirector, automatically adds
1208 the standard headers to every HTTP request and handles gzipped and
1209 deflated responses from web servers. If compression is to be avoided in
1210 a particular request, the original request in the program code only has
1211 to include the HTTP header "Youtubedl-no-compression", which will be
1212 removed before making the real request.
1214 Part of this code was copied from:
1216 http://techknack.net/python-urllib2-handlers/
1218 Andrew Rowls, the author of that code, agreed to release it to the
1222 def __init__(self
, params
, *args
, **kwargs
):
1223 compat_urllib_request
.HTTPHandler
.__init
__(self
, *args
, **kwargs
)
1224 self
._params
= params
1226 def http_open(self
, req
):
1227 conn_class
= compat_http_client
.HTTPConnection
1229 socks_proxy
= req
.headers
.get('Ytdl-socks-proxy')
1231 conn_class
= make_socks_conn_class(conn_class
, socks_proxy
)
1232 del req
.headers
['Ytdl-socks-proxy']
1234 return self
.do_open(functools
.partial(
1235 _create_http_connection
, self
, conn_class
, False),
1243 return zlib
.decompress(data
, -zlib
.MAX_WBITS
)
1245 return zlib
.decompress(data
)
1247 def http_request(self
, req
):
1248 # According to RFC 3986, URLs can not contain non-ASCII characters, however this is not
1249 # always respected by websites, some tend to give out URLs with non percent-encoded
1250 # non-ASCII characters (see telemb.py, ard.py [#3412])
1251 # urllib chokes on URLs with non-ASCII characters (see http://bugs.python.org/issue3991)
1252 # To work around aforementioned issue we will replace request's original URL with
1253 # percent-encoded one
1254 # Since redirects are also affected (e.g. http://www.southpark.de/alle-episoden/s18e09)
1255 # the code of this workaround has been moved here from YoutubeDL.urlopen()
1256 url
= req
.get_full_url()
1257 url_escaped
= escape_url(url
)
1259 # Substitute URL if any change after escaping
1260 if url
!= url_escaped
:
1261 req
= update_Request(req
, url
=url_escaped
)
1263 for h
, v
in std_headers
.items():
1264 # Capitalize is needed because of Python bug 2275: http://bugs.python.org/issue2275
1265 # The dict keys are capitalized because of this bug by urllib
1266 if h
.capitalize() not in req
.headers
:
1267 req
.add_header(h
, v
)
1269 req
.headers
= handle_youtubedl_headers(req
.headers
)
1271 if sys
.version_info
< (2, 7) and '#' in req
.get_full_url():
1272 # Python 2.6 is brain-dead when it comes to fragments
1273 req
._Request
__original
= req
._Request
__original
.partition('#')[0]
1274 req
._Request
__r
_type
= req
._Request
__r
_type
.partition('#')[0]
1278 def http_response(self
, req
, resp
):
1281 if resp
.headers
.get('Content-encoding', '') == 'gzip':
1282 content
= resp
.read()
1283 gz
= gzip
.GzipFile(fileobj
=io
.BytesIO(content
), mode
='rb')
1285 uncompressed
= io
.BytesIO(gz
.read())
1286 except IOError as original_ioerror
:
1287 # There may be junk add the end of the file
1288 # See http://stackoverflow.com/q/4928560/35070 for details
1289 for i
in range(1, 1024):
1291 gz
= gzip
.GzipFile(fileobj
=io
.BytesIO(content
[:-i
]), mode
='rb')
1292 uncompressed
= io
.BytesIO(gz
.read())
1297 raise original_ioerror
1298 resp
= compat_urllib_request
.addinfourl(uncompressed
, old_resp
.headers
, old_resp
.url
, old_resp
.code
)
1299 resp
.msg
= old_resp
.msg
1300 del resp
.headers
['Content-encoding']
1302 if resp
.headers
.get('Content-encoding', '') == 'deflate':
1303 gz
= io
.BytesIO(self
.deflate(resp
.read()))
1304 resp
= compat_urllib_request
.addinfourl(gz
, old_resp
.headers
, old_resp
.url
, old_resp
.code
)
1305 resp
.msg
= old_resp
.msg
1306 del resp
.headers
['Content-encoding']
1307 # Percent-encode redirect URL of Location HTTP header to satisfy RFC 3986 (see
1308 # https://github.com/ytdl-org/youtube-dl/issues/6457).
1309 if 300 <= resp
.code
< 400:
1310 location
= resp
.headers
.get('Location')
1312 # As of RFC 2616 default charset is iso-8859-1 that is respected by python 3
1313 if sys
.version_info
>= (3, 0):
1314 location
= location
.encode('iso-8859-1').decode('utf-8')
1316 location
= location
.decode('utf-8')
1317 location_escaped
= escape_url(location
)
1318 if location
!= location_escaped
:
1319 del resp
.headers
['Location']
1320 if sys
.version_info
< (3, 0):
1321 location_escaped
= location_escaped
.encode('utf-8')
1322 resp
.headers
['Location'] = location_escaped
1325 https_request
= http_request
1326 https_response
= http_response
1329 def make_socks_conn_class(base_class
, socks_proxy
):
1330 assert issubclass(base_class
, (
1331 compat_http_client
.HTTPConnection
, compat_http_client
.HTTPSConnection
))
1333 url_components
= compat_urlparse
.urlparse(socks_proxy
)
1334 if url_components
.scheme
.lower() == 'socks5':
1335 socks_type
= ProxyType
.SOCKS5
1336 elif url_components
.scheme
.lower() in ('socks', 'socks4'):
1337 socks_type
= ProxyType
.SOCKS4
1338 elif url_components
.scheme
.lower() == 'socks4a':
1339 socks_type
= ProxyType
.SOCKS4A
1341 def unquote_if_non_empty(s
):
1344 return compat_urllib_parse_unquote_plus(s
)
1348 url_components
.hostname
, url_components
.port
or 1080,
1350 unquote_if_non_empty(url_components
.username
),
1351 unquote_if_non_empty(url_components
.password
),
1354 class SocksConnection(base_class
):
1356 self
.sock
= sockssocket()
1357 self
.sock
.setproxy(*proxy_args
)
1358 if type(self
.timeout
) in (int, float):
1359 self
.sock
.settimeout(self
.timeout
)
1360 self
.sock
.connect((self
.host
, self
.port
))
1362 if isinstance(self
, compat_http_client
.HTTPSConnection
):
1363 if hasattr(self
, '_context'): # Python > 2.6
1364 self
.sock
= self
._context
.wrap_socket(
1365 self
.sock
, server_hostname
=self
.host
)
1367 self
.sock
= ssl
.wrap_socket(self
.sock
)
1369 return SocksConnection
1372 class YoutubeDLHTTPSHandler(compat_urllib_request
.HTTPSHandler
):
1373 def __init__(self
, params
, https_conn_class
=None, *args
, **kwargs
):
1374 compat_urllib_request
.HTTPSHandler
.__init
__(self
, *args
, **kwargs
)
1375 self
._https
_conn
_class
= https_conn_class
or compat_http_client
.HTTPSConnection
1376 self
._params
= params
1378 def https_open(self
, req
):
1380 conn_class
= self
._https
_conn
_class
1382 if hasattr(self
, '_context'): # python > 2.6
1383 kwargs
['context'] = self
._context
1384 if hasattr(self
, '_check_hostname'): # python 3.x
1385 kwargs
['check_hostname'] = self
._check
_hostname
1387 socks_proxy
= req
.headers
.get('Ytdl-socks-proxy')
1389 conn_class
= make_socks_conn_class(conn_class
, socks_proxy
)
1390 del req
.headers
['Ytdl-socks-proxy']
1392 return self
.do_open(functools
.partial(
1393 _create_http_connection
, self
, conn_class
, True),
1397 class YoutubeDLCookieJar(compat_cookiejar
.MozillaCookieJar
):
1399 See [1] for cookie file format.
1401 1. https://curl.haxx.se/docs/http-cookies.html
1403 _HTTPONLY_PREFIX
= '#HttpOnly_'
1405 _HEADER
= '''# Netscape HTTP Cookie File
1406 # This file is generated by yt-dlp. Do not edit.
1409 _CookieFileEntry
= collections
.namedtuple(
1411 ('domain_name', 'include_subdomains', 'path', 'https_only', 'expires_at', 'name', 'value'))
1413 def save(self
, filename
=None, ignore_discard
=False, ignore_expires
=False):
1415 Save cookies to a file.
1417 Most of the code is taken from CPython 3.8 and slightly adapted
1418 to support cookie files with UTF-8 in both python 2 and 3.
1420 if filename
is None:
1421 if self
.filename
is not None:
1422 filename
= self
.filename
1424 raise ValueError(compat_cookiejar
.MISSING_FILENAME_TEXT
)
1426 # Store session cookies with `expires` set to 0 instead of an empty
1429 if cookie
.expires
is None:
1432 with io
.open(filename
, 'w', encoding
='utf-8') as f
:
1433 f
.write(self
._HEADER
)
1436 if not ignore_discard
and cookie
.discard
:
1438 if not ignore_expires
and cookie
.is_expired(now
):
1444 if cookie
.domain
.startswith('.'):
1445 initial_dot
= 'TRUE'
1447 initial_dot
= 'FALSE'
1448 if cookie
.expires
is not None:
1449 expires
= compat_str(cookie
.expires
)
1452 if cookie
.value
is None:
1453 # cookies.txt regards 'Set-Cookie: foo' as a cookie
1454 # with no name, whereas http.cookiejar regards it as a
1455 # cookie with no value.
1460 value
= cookie
.value
1462 '\t'.join([cookie
.domain
, initial_dot
, cookie
.path
,
1463 secure
, expires
, name
, value
]) + '\n')
1465 def load(self
, filename
=None, ignore_discard
=False, ignore_expires
=False):
1466 """Load cookies from a file."""
1467 if filename
is None:
1468 if self
.filename
is not None:
1469 filename
= self
.filename
1471 raise ValueError(compat_cookiejar
.MISSING_FILENAME_TEXT
)
1473 def prepare_line(line
):
1474 if line
.startswith(self
._HTTPONLY
_PREFIX
):
1475 line
= line
[len(self
._HTTPONLY
_PREFIX
):]
1476 # comments and empty lines are fine
1477 if line
.startswith('#') or not line
.strip():
1479 cookie_list
= line
.split('\t')
1480 if len(cookie_list
) != self
._ENTRY
_LEN
:
1481 raise compat_cookiejar
.LoadError('invalid length %d' % len(cookie_list
))
1482 cookie
= self
._CookieFileEntry
(*cookie_list
)
1483 if cookie
.expires_at
and not cookie
.expires_at
.isdigit():
1484 raise compat_cookiejar
.LoadError('invalid expires at %s' % cookie
.expires_at
)
1488 with io
.open(filename
, encoding
='utf-8') as f
:
1491 cf
.write(prepare_line(line
))
1492 except compat_cookiejar
.LoadError
as e
:
1494 'WARNING: skipping cookie file entry due to %s: %r\n'
1495 % (e
, line
), sys
.stderr
)
1498 self
._really
_load
(cf
, filename
, ignore_discard
, ignore_expires
)
1499 # Session cookies are denoted by either `expires` field set to
1500 # an empty string or 0. MozillaCookieJar only recognizes the former
1501 # (see [1]). So we need force the latter to be recognized as session
1502 # cookies on our own.
1503 # Session cookies may be important for cookies-based authentication,
1504 # e.g. usually, when user does not check 'Remember me' check box while
1505 # logging in on a site, some important cookies are stored as session
1506 # cookies so that not recognizing them will result in failed login.
1507 # 1. https://bugs.python.org/issue17164
1509 # Treat `expires=0` cookies as session cookies
1510 if cookie
.expires
== 0:
1511 cookie
.expires
= None
1512 cookie
.discard
= True
1515 class YoutubeDLCookieProcessor(compat_urllib_request
.HTTPCookieProcessor
):
1516 def __init__(self
, cookiejar
=None):
1517 compat_urllib_request
.HTTPCookieProcessor
.__init
__(self
, cookiejar
)
1519 def http_response(self
, request
, response
):
1520 # Python 2 will choke on next HTTP request in row if there are non-ASCII
1521 # characters in Set-Cookie HTTP header of last response (see
1522 # https://github.com/ytdl-org/youtube-dl/issues/6769).
1523 # In order to at least prevent crashing we will percent encode Set-Cookie
1524 # header before HTTPCookieProcessor starts processing it.
1525 # if sys.version_info < (3, 0) and response.headers:
1526 # for set_cookie_header in ('Set-Cookie', 'Set-Cookie2'):
1527 # set_cookie = response.headers.get(set_cookie_header)
1529 # set_cookie_escaped = compat_urllib_parse.quote(set_cookie, b"%/;:@&=+$,!~*'()?#[] ")
1530 # if set_cookie != set_cookie_escaped:
1531 # del response.headers[set_cookie_header]
1532 # response.headers[set_cookie_header] = set_cookie_escaped
1533 return compat_urllib_request
.HTTPCookieProcessor
.http_response(self
, request
, response
)
1535 https_request
= compat_urllib_request
.HTTPCookieProcessor
.http_request
1536 https_response
= http_response
1539 class YoutubeDLRedirectHandler(compat_urllib_request
.HTTPRedirectHandler
):
1540 """YoutubeDL redirect handler
1542 The code is based on HTTPRedirectHandler implementation from CPython [1].
1544 This redirect handler solves two issues:
1545 - ensures redirect URL is always unicode under python 2
1546 - introduces support for experimental HTTP response status code
1547 308 Permanent Redirect [2] used by some sites [3]
1549 1. https://github.com/python/cpython/blob/master/Lib/urllib/request.py
1550 2. https://developer.mozilla.org/en-US/docs/Web/HTTP/Status/308
1551 3. https://github.com/ytdl-org/youtube-dl/issues/28768
1554 http_error_301
= http_error_303
= http_error_307
= http_error_308
= compat_urllib_request
.HTTPRedirectHandler
.http_error_302
1556 def redirect_request(self
, req
, fp
, code
, msg
, headers
, newurl
):
1557 """Return a Request or None in response to a redirect.
1559 This is called by the http_error_30x methods when a
1560 redirection response is received. If a redirection should
1561 take place, return a new Request to allow http_error_30x to
1562 perform the redirect. Otherwise, raise HTTPError if no-one
1563 else should try to handle this url. Return None if you can't
1564 but another Handler might.
1566 m
= req
.get_method()
1567 if (not (code
in (301, 302, 303, 307, 308) and m
in ("GET", "HEAD")
1568 or code
in (301, 302, 303) and m
== "POST")):
1569 raise compat_HTTPError(req
.full_url
, code
, msg
, headers
, fp
)
1570 # Strictly (according to RFC 2616), 301 or 302 in response to
1571 # a POST MUST NOT cause a redirection without confirmation
1572 # from the user (of urllib.request, in this case). In practice,
1573 # essentially all clients do redirect in this case, so we do
1576 # On python 2 urlh.geturl() may sometimes return redirect URL
1577 # as byte string instead of unicode. This workaround allows
1578 # to force it always return unicode.
1579 if sys
.version_info
[0] < 3:
1580 newurl
= compat_str(newurl
)
1582 # Be conciliant with URIs containing a space. This is mainly
1583 # redundant with the more complete encoding done in http_error_302(),
1584 # but it is kept for compatibility with other callers.
1585 newurl
= newurl
.replace(' ', '%20')
1587 CONTENT_HEADERS
= ("content-length", "content-type")
1588 # NB: don't use dict comprehension for python 2.6 compatibility
1589 newheaders
= dict((k
, v
) for k
, v
in req
.headers
.items()
1590 if k
.lower() not in CONTENT_HEADERS
)
1591 return compat_urllib_request
.Request(
1592 newurl
, headers
=newheaders
, origin_req_host
=req
.origin_req_host
,
1596 def extract_timezone(date_str
):
1599 ^.{8,}? # >=8 char non-TZ prefix, if present
1600 (?P<tz>Z| # just the UTC Z, or
1601 (?:(?<=.\b\d{4}|\b\d{2}:\d\d)| # preceded by 4 digits or hh:mm or
1602 (?<!.\b[a-zA-Z]{3}|[a-zA-Z]{4}|..\b\d\d)) # not preceded by 3 alpha word or >= 4 alpha or 2 digits
1603 [ ]? # optional space
1604 (?P<sign>\+|-) # +/-
1605 (?P<hours>[0-9]{2}):?(?P<minutes>[0-9]{2}) # hh[:]mm
1609 timezone
= datetime
.timedelta()
1611 date_str
= date_str
[:-len(m
.group('tz'))]
1612 if not m
.group('sign'):
1613 timezone
= datetime
.timedelta()
1615 sign
= 1 if m
.group('sign') == '+' else -1
1616 timezone
= datetime
.timedelta(
1617 hours
=sign
* int(m
.group('hours')),
1618 minutes
=sign
* int(m
.group('minutes')))
1619 return timezone
, date_str
1622 def parse_iso8601(date_str
, delimiter
='T', timezone
=None):
1623 """ Return a UNIX timestamp from the given date """
1625 if date_str
is None:
1628 date_str
= re
.sub(r
'\.[0-9]+', '', date_str
)
1630 if timezone
is None:
1631 timezone
, date_str
= extract_timezone(date_str
)
1634 date_format
= '%Y-%m-%d{0}%H:%M:%S'.format(delimiter
)
1635 dt
= datetime
.datetime
.strptime(date_str
, date_format
) - timezone
1636 return calendar
.timegm(dt
.timetuple())
1641 def date_formats(day_first
=True):
1642 return DATE_FORMATS_DAY_FIRST
if day_first
else DATE_FORMATS_MONTH_FIRST
1645 def unified_strdate(date_str
, day_first
=True):
1646 """Return a string with the date in the format YYYYMMDD"""
1648 if date_str
is None:
1652 date_str
= date_str
.replace(',', ' ')
1653 # Remove AM/PM + timezone
1654 date_str
= re
.sub(r
'(?i)\s*(?:AM|PM)(?:\s+[A-Z]+)?', '', date_str
)
1655 _
, date_str
= extract_timezone(date_str
)
1657 for expression
in date_formats(day_first
):
1659 upload_date
= datetime
.datetime
.strptime(date_str
, expression
).strftime('%Y%m%d')
1662 if upload_date
is None:
1663 timetuple
= email
.utils
.parsedate_tz(date_str
)
1666 upload_date
= datetime
.datetime(*timetuple
[:6]).strftime('%Y%m%d')
1669 if upload_date
is not None:
1670 return compat_str(upload_date
)
1673 def unified_timestamp(date_str
, day_first
=True):
1674 if date_str
is None:
1677 date_str
= re
.sub(r
'[,|]', '', date_str
)
1679 pm_delta
= 12 if re
.search(r
'(?i)PM', date_str
) else 0
1680 timezone
, date_str
= extract_timezone(date_str
)
1682 # Remove AM/PM + timezone
1683 date_str
= re
.sub(r
'(?i)\s*(?:AM|PM)(?:\s+[A-Z]+)?', '', date_str
)
1685 # Remove unrecognized timezones from ISO 8601 alike timestamps
1686 m
= re
.search(r
'\d{1,2}:\d{1,2}(?:\.\d+)?(?P<tz>\s*[A-Z]+)$', date_str
)
1688 date_str
= date_str
[:-len(m
.group('tz'))]
1690 # Python only supports microseconds, so remove nanoseconds
1691 m
= re
.search(r
'^([0-9]{4,}-[0-9]{1,2}-[0-9]{1,2}T[0-9]{1,2}:[0-9]{1,2}:[0-9]{1,2}\.[0-9]{6})[0-9]+$', date_str
)
1693 date_str
= m
.group(1)
1695 for expression
in date_formats(day_first
):
1697 dt
= datetime
.datetime
.strptime(date_str
, expression
) - timezone
+ datetime
.timedelta(hours
=pm_delta
)
1698 return calendar
.timegm(dt
.timetuple())
1701 timetuple
= email
.utils
.parsedate_tz(date_str
)
1703 return calendar
.timegm(timetuple
) + pm_delta
* 3600
1706 def determine_ext(url
, default_ext
='unknown_video'):
1707 if url
is None or '.' not in url
:
1709 guess
= url
.partition('?')[0].rpartition('.')[2]
1710 if re
.match(r
'^[A-Za-z0-9]+$', guess
):
1712 # Try extract ext from URLs like http://example.com/foo/bar.mp4/?download
1713 elif guess
.rstrip('/') in KNOWN_EXTENSIONS
:
1714 return guess
.rstrip('/')
1719 def subtitles_filename(filename
, sub_lang
, sub_format
, expected_real_ext
=None):
1720 return replace_extension(filename
, sub_lang
+ '.' + sub_format
, expected_real_ext
)
1723 def datetime_from_str(date_str
, precision
='auto', format
='%Y%m%d'):
1725 Return a datetime object from a string in the format YYYYMMDD or
1726 (now|today|date)[+-][0-9](microsecond|second|minute|hour|day|week|month|year)(s)?
1728 format: string date format used to return datetime object from
1729 precision: round the time portion of a datetime object.
1730 auto|microsecond|second|minute|hour|day.
1731 auto: round to the unit provided in date_str (if applicable).
1733 auto_precision
= False
1734 if precision
== 'auto':
1735 auto_precision
= True
1736 precision
= 'microsecond'
1737 today
= datetime_round(datetime
.datetime
.now(), precision
)
1738 if date_str
in ('now', 'today'):
1740 if date_str
== 'yesterday':
1741 return today
- datetime
.timedelta(days
=1)
1743 r
'(?P<start>.+)(?P<sign>[+-])(?P<time>\d+)(?P<unit>microsecond|second|minute|hour|day|week|month|year)(s)?',
1745 if match
is not None:
1746 start_time
= datetime_from_str(match
.group('start'), precision
, format
)
1747 time
= int(match
.group('time')) * (-1 if match
.group('sign') == '-' else 1)
1748 unit
= match
.group('unit')
1749 if unit
== 'month' or unit
== 'year':
1750 new_date
= datetime_add_months(start_time
, time
* 12 if unit
== 'year' else time
)
1756 delta
= datetime
.timedelta(**{unit + 's': time}
)
1757 new_date
= start_time
+ delta
1759 return datetime_round(new_date
, unit
)
1762 return datetime_round(datetime
.datetime
.strptime(date_str
, format
), precision
)
1765 def date_from_str(date_str
, format
='%Y%m%d'):
1767 Return a datetime object from a string in the format YYYYMMDD or
1768 (now|today|date)[+-][0-9](microsecond|second|minute|hour|day|week|month|year)(s)?
1770 format: string date format used to return datetime object from
1772 return datetime_from_str(date_str
, precision
='microsecond', format
=format
).date()
1775 def datetime_add_months(dt
, months
):
1776 """Increment/Decrement a datetime object by months."""
1777 month
= dt
.month
+ months
- 1
1778 year
= dt
.year
+ month
// 12
1779 month
= month
% 12 + 1
1780 day
= min(dt
.day
, calendar
.monthrange(year
, month
)[1])
1781 return dt
.replace(year
, month
, day
)
1784 def datetime_round(dt
, precision
='day'):
1786 Round a datetime object's time to a specific precision
1788 if precision
== 'microsecond':
1797 roundto
= lambda x
, n
: ((x
+ n
/ 2) // n
) * n
1798 timestamp
= calendar
.timegm(dt
.timetuple())
1799 return datetime
.datetime
.utcfromtimestamp(roundto(timestamp
, unit_seconds
[precision
]))
1802 def hyphenate_date(date_str
):
1804 Convert a date in 'YYYYMMDD' format to 'YYYY-MM-DD' format"""
1805 match
= re
.match(r
'^(\d\d\d\d)(\d\d)(\d\d)$', date_str
)
1806 if match
is not None:
1807 return '-'.join(match
.groups())
1812 class DateRange(object):
1813 """Represents a time interval between two dates"""
1815 def __init__(self
, start
=None, end
=None):
1816 """start and end must be strings in the format accepted by date"""
1817 if start
is not None:
1818 self
.start
= date_from_str(start
)
1820 self
.start
= datetime
.datetime
.min.date()
1822 self
.end
= date_from_str(end
)
1824 self
.end
= datetime
.datetime
.max.date()
1825 if self
.start
> self
.end
:
1826 raise ValueError('Date range: "%s" , the start date must be before the end date' % self
)
1830 """Returns a range that only contains the given day"""
1831 return cls(day
, day
)
1833 def __contains__(self
, date
):
1834 """Check if the date is in the range"""
1835 if not isinstance(date
, datetime
.date
):
1836 date
= date_from_str(date
)
1837 return self
.start
<= date
<= self
.end
1840 return '%s - %s' % (self
.start
.isoformat(), self
.end
.isoformat())
1843 def platform_name():
1844 """ Returns the platform name as a compat_str """
1845 res
= platform
.platform()
1846 if isinstance(res
, bytes):
1847 res
= res
.decode(preferredencoding())
1849 assert isinstance(res
, compat_str
)
1853 def get_windows_version():
1854 ''' Get Windows version. None if it's not running on Windows '''
1855 if compat_os_name
== 'nt':
1856 return version_tuple(platform
.win32_ver()[1])
1861 def _windows_write_string(s
, out
):
1862 """ Returns True if the string was written using special methods,
1863 False if it has yet to be written out."""
1864 # Adapted from http://stackoverflow.com/a/3259271/35070
1866 import ctypes
.wintypes
1874 fileno
= out
.fileno()
1875 except AttributeError:
1876 # If the output stream doesn't have a fileno, it's virtual
1878 except io
.UnsupportedOperation
:
1879 # Some strange Windows pseudo files?
1881 if fileno
not in WIN_OUTPUT_IDS
:
1884 GetStdHandle
= compat_ctypes_WINFUNCTYPE(
1885 ctypes
.wintypes
.HANDLE
, ctypes
.wintypes
.DWORD
)(
1886 ('GetStdHandle', ctypes
.windll
.kernel32
))
1887 h
= GetStdHandle(WIN_OUTPUT_IDS
[fileno
])
1889 WriteConsoleW
= compat_ctypes_WINFUNCTYPE(
1890 ctypes
.wintypes
.BOOL
, ctypes
.wintypes
.HANDLE
, ctypes
.wintypes
.LPWSTR
,
1891 ctypes
.wintypes
.DWORD
, ctypes
.POINTER(ctypes
.wintypes
.DWORD
),
1892 ctypes
.wintypes
.LPVOID
)(('WriteConsoleW', ctypes
.windll
.kernel32
))
1893 written
= ctypes
.wintypes
.DWORD(0)
1895 GetFileType
= compat_ctypes_WINFUNCTYPE(ctypes
.wintypes
.DWORD
, ctypes
.wintypes
.DWORD
)(('GetFileType', ctypes
.windll
.kernel32
))
1896 FILE_TYPE_CHAR
= 0x0002
1897 FILE_TYPE_REMOTE
= 0x8000
1898 GetConsoleMode
= compat_ctypes_WINFUNCTYPE(
1899 ctypes
.wintypes
.BOOL
, ctypes
.wintypes
.HANDLE
,
1900 ctypes
.POINTER(ctypes
.wintypes
.DWORD
))(
1901 ('GetConsoleMode', ctypes
.windll
.kernel32
))
1902 INVALID_HANDLE_VALUE
= ctypes
.wintypes
.DWORD(-1).value
1904 def not_a_console(handle
):
1905 if handle
== INVALID_HANDLE_VALUE
or handle
is None:
1907 return ((GetFileType(handle
) & ~FILE_TYPE_REMOTE
) != FILE_TYPE_CHAR
1908 or GetConsoleMode(handle
, ctypes
.byref(ctypes
.wintypes
.DWORD())) == 0)
1910 if not_a_console(h
):
1913 def next_nonbmp_pos(s
):
1915 return next(i
for i
, c
in enumerate(s
) if ord(c
) > 0xffff)
1916 except StopIteration:
1920 count
= min(next_nonbmp_pos(s
), 1024)
1922 ret
= WriteConsoleW(
1923 h
, s
, count
if count
else 2, ctypes
.byref(written
), None)
1925 raise OSError('Failed to write string')
1926 if not count
: # We just wrote a non-BMP character
1927 assert written
.value
== 2
1930 assert written
.value
> 0
1931 s
= s
[written
.value
:]
1935 def write_string(s
, out
=None, encoding
=None):
1938 assert type(s
) == compat_str
1940 if sys
.platform
== 'win32' and encoding
is None and hasattr(out
, 'fileno'):
1941 if _windows_write_string(s
, out
):
1944 if ('b' in getattr(out
, 'mode', '')
1945 or sys
.version_info
[0] < 3): # Python 2 lies about mode of sys.stderr
1946 byt
= s
.encode(encoding
or preferredencoding(), 'ignore')
1948 elif hasattr(out
, 'buffer'):
1949 enc
= encoding
or getattr(out
, 'encoding', None) or preferredencoding()
1950 byt
= s
.encode(enc
, 'ignore')
1951 out
.buffer.write(byt
)
1957 def bytes_to_intlist(bs
):
1960 if isinstance(bs
[0], int): # Python 3
1963 return [ord(c
) for c
in bs
]
1966 def intlist_to_bytes(xs
):
1969 return compat_struct_pack('%dB' % len(xs
), *xs
)
1972 # Cross-platform file locking
1973 if sys
.platform
== 'win32':
1974 import ctypes
.wintypes
1977 class OVERLAPPED(ctypes
.Structure
):
1979 ('Internal', ctypes
.wintypes
.LPVOID
),
1980 ('InternalHigh', ctypes
.wintypes
.LPVOID
),
1981 ('Offset', ctypes
.wintypes
.DWORD
),
1982 ('OffsetHigh', ctypes
.wintypes
.DWORD
),
1983 ('hEvent', ctypes
.wintypes
.HANDLE
),
1986 kernel32
= ctypes
.windll
.kernel32
1987 LockFileEx
= kernel32
.LockFileEx
1988 LockFileEx
.argtypes
= [
1989 ctypes
.wintypes
.HANDLE
, # hFile
1990 ctypes
.wintypes
.DWORD
, # dwFlags
1991 ctypes
.wintypes
.DWORD
, # dwReserved
1992 ctypes
.wintypes
.DWORD
, # nNumberOfBytesToLockLow
1993 ctypes
.wintypes
.DWORD
, # nNumberOfBytesToLockHigh
1994 ctypes
.POINTER(OVERLAPPED
) # Overlapped
1996 LockFileEx
.restype
= ctypes
.wintypes
.BOOL
1997 UnlockFileEx
= kernel32
.UnlockFileEx
1998 UnlockFileEx
.argtypes
= [
1999 ctypes
.wintypes
.HANDLE
, # hFile
2000 ctypes
.wintypes
.DWORD
, # dwReserved
2001 ctypes
.wintypes
.DWORD
, # nNumberOfBytesToLockLow
2002 ctypes
.wintypes
.DWORD
, # nNumberOfBytesToLockHigh
2003 ctypes
.POINTER(OVERLAPPED
) # Overlapped
2005 UnlockFileEx
.restype
= ctypes
.wintypes
.BOOL
2006 whole_low
= 0xffffffff
2007 whole_high
= 0x7fffffff
2009 def _lock_file(f
, exclusive
):
2010 overlapped
= OVERLAPPED()
2011 overlapped
.Offset
= 0
2012 overlapped
.OffsetHigh
= 0
2013 overlapped
.hEvent
= 0
2014 f
._lock
_file
_overlapped
_p
= ctypes
.pointer(overlapped
)
2015 handle
= msvcrt
.get_osfhandle(f
.fileno())
2016 if not LockFileEx(handle
, 0x2 if exclusive
else 0x0, 0,
2017 whole_low
, whole_high
, f
._lock
_file
_overlapped
_p
):
2018 raise OSError('Locking file failed: %r' % ctypes
.FormatError())
2020 def _unlock_file(f
):
2021 assert f
._lock
_file
_overlapped
_p
2022 handle
= msvcrt
.get_osfhandle(f
.fileno())
2023 if not UnlockFileEx(handle
, 0,
2024 whole_low
, whole_high
, f
._lock
_file
_overlapped
_p
):
2025 raise OSError('Unlocking file failed: %r' % ctypes
.FormatError())
2028 # Some platforms, such as Jython, is missing fcntl
2032 def _lock_file(f
, exclusive
):
2033 fcntl
.flock(f
, fcntl
.LOCK_EX
if exclusive
else fcntl
.LOCK_SH
)
2035 def _unlock_file(f
):
2036 fcntl
.flock(f
, fcntl
.LOCK_UN
)
2038 UNSUPPORTED_MSG
= 'file locking is not supported on this platform'
2040 def _lock_file(f
, exclusive
):
2041 raise IOError(UNSUPPORTED_MSG
)
2043 def _unlock_file(f
):
2044 raise IOError(UNSUPPORTED_MSG
)
2047 class locked_file(object):
2048 def __init__(self
, filename
, mode
, encoding
=None):
2049 assert mode
in ['r', 'a', 'w']
2050 self
.f
= io
.open(filename
, mode
, encoding
=encoding
)
2053 def __enter__(self
):
2054 exclusive
= self
.mode
!= 'r'
2056 _lock_file(self
.f
, exclusive
)
2062 def __exit__(self
, etype
, value
, traceback
):
2064 _unlock_file(self
.f
)
2071 def write(self
, *args
):
2072 return self
.f
.write(*args
)
2074 def read(self
, *args
):
2075 return self
.f
.read(*args
)
2078 def get_filesystem_encoding():
2079 encoding
= sys
.getfilesystemencoding()
2080 return encoding
if encoding
is not None else 'utf-8'
2083 def shell_quote(args
):
2085 encoding
= get_filesystem_encoding()
2087 if isinstance(a
, bytes):
2088 # We may get a filename encoded with 'encodeFilename'
2089 a
= a
.decode(encoding
)
2090 quoted_args
.append(compat_shlex_quote(a
))
2091 return ' '.join(quoted_args
)
2094 def smuggle_url(url
, data
):
2095 """ Pass additional data in a URL for internal use. """
2097 url
, idata
= unsmuggle_url(url
, {})
2099 sdata
= compat_urllib_parse_urlencode(
2100 {'__youtubedl_smuggle': json.dumps(data)}
)
2101 return url
+ '#' + sdata
2104 def unsmuggle_url(smug_url
, default
=None):
2105 if '#__youtubedl_smuggle' not in smug_url
:
2106 return smug_url
, default
2107 url
, _
, sdata
= smug_url
.rpartition('#')
2108 jsond
= compat_parse_qs(sdata
)['__youtubedl_smuggle'][0]
2109 data
= json
.loads(jsond
)
2113 def format_decimal_suffix(num
, fmt
='%d%s', *, factor
=1000):
2114 """ Formats numbers with decimal sufixes like K, M, etc """
2115 num
, factor
= float_or_none(num
), float(factor
)
2118 exponent
= 0 if num
== 0 else int(math
.log(num
, factor
))
2119 suffix
= ['', *'KMGTPEZY'][exponent
]
2120 converted
= num
/ (factor
** exponent
)
2121 return fmt
% (converted
, suffix
)
2124 def format_bytes(bytes):
2125 return format_decimal_suffix(bytes, '%.2f%siB', factor
=1024) or 'N/A'
2128 def lookup_unit_table(unit_table
, s
):
2129 units_re
= '|'.join(re
.escape(u
) for u
in unit_table
)
2131 r
'(?P<num>[0-9]+(?:[,.][0-9]*)?)\s*(?P<unit>%s)\b' % units_re
, s
)
2134 num_str
= m
.group('num').replace(',', '.')
2135 mult
= unit_table
[m
.group('unit')]
2136 return int(float(num_str
) * mult
)
2139 def parse_filesize(s
):
2143 # The lower-case forms are of course incorrect and unofficial,
2144 # but we support those too
2161 'megabytes': 1000 ** 2,
2162 'mebibytes': 1024 ** 2,
2168 'gigabytes': 1000 ** 3,
2169 'gibibytes': 1024 ** 3,
2175 'terabytes': 1000 ** 4,
2176 'tebibytes': 1024 ** 4,
2182 'petabytes': 1000 ** 5,
2183 'pebibytes': 1024 ** 5,
2189 'exabytes': 1000 ** 6,
2190 'exbibytes': 1024 ** 6,
2196 'zettabytes': 1000 ** 7,
2197 'zebibytes': 1024 ** 7,
2203 'yottabytes': 1000 ** 8,
2204 'yobibytes': 1024 ** 8,
2207 return lookup_unit_table(_UNIT_TABLE
, s
)
2214 s
= re
.sub(r
'^[^\d]+\s', '', s
).strip()
2216 if re
.match(r
'^[\d,.]+$', s
):
2217 return str_to_int(s
)
2230 ret
= lookup_unit_table(_UNIT_TABLE
, s
)
2234 mobj
= re
.match(r
'([\d,.]+)(?:$|\s)', s
)
2236 return str_to_int(mobj
.group(1))
2239 def parse_resolution(s
):
2243 mobj
= re
.search(r
'(?<![a-zA-Z0-9])(?P<w>\d+)\s*[xX×,]\s*(?P<h>\d+)(?![a-zA-Z0-9])', s
)
2246 'width': int(mobj
.group('w')),
2247 'height': int(mobj
.group('h')),
2250 mobj
= re
.search(r
'(?<![a-zA-Z0-9])(\d+)[pPiI](?![a-zA-Z0-9])', s
)
2252 return {'height': int(mobj.group(1))}
2254 mobj
= re
.search(r
'\b([48])[kK]\b', s
)
2256 return {'height': int(mobj.group(1)) * 540}
2261 def parse_bitrate(s
):
2262 if not isinstance(s
, compat_str
):
2264 mobj
= re
.search(r
'\b(\d+)\s*kbps', s
)
2266 return int(mobj
.group(1))
2269 def month_by_name(name
, lang
='en'):
2270 """ Return the number of a month by (locale-independently) English name """
2272 month_names
= MONTH_NAMES
.get(lang
, MONTH_NAMES
['en'])
2275 return month_names
.index(name
) + 1
2280 def month_by_abbreviation(abbrev
):
2281 """ Return the number of a month by (locale-independently) English
2285 return [s
[:3] for s
in ENGLISH_MONTH_NAMES
].index(abbrev
) + 1
2290 def fix_xml_ampersands(xml_str
):
2291 """Replace all the '&' by '&' in XML"""
2293 r
'&(?!amp;|lt;|gt;|apos;|quot;|#x[0-9a-fA-F]{,4};|#[0-9]{,4};)',
2298 def setproctitle(title
):
2299 assert isinstance(title
, compat_str
)
2301 # ctypes in Jython is not complete
2302 # http://bugs.jython.org/issue2148
2303 if sys
.platform
.startswith('java'):
2307 libc
= ctypes
.cdll
.LoadLibrary('libc.so.6')
2311 # LoadLibrary in Windows Python 2.7.13 only expects
2312 # a bytestring, but since unicode_literals turns
2313 # every string into a unicode string, it fails.
2315 title_bytes
= title
.encode('utf-8')
2316 buf
= ctypes
.create_string_buffer(len(title_bytes
))
2317 buf
.value
= title_bytes
2319 libc
.prctl(15, buf
, 0, 0, 0)
2320 except AttributeError:
2321 return # Strange libc, just skip this
2324 def remove_start(s
, start
):
2325 return s
[len(start
):] if s
is not None and s
.startswith(start
) else s
2328 def remove_end(s
, end
):
2329 return s
[:-len(end
)] if s
is not None and s
.endswith(end
) else s
2332 def remove_quotes(s
):
2333 if s
is None or len(s
) < 2:
2335 for quote
in ('"', "'", ):
2336 if s
[0] == quote
and s
[-1] == quote
:
2341 def get_domain(url
):
2342 domain
= re
.match(r
'(?:https?:\/\/)?(?:www\.)?(?P<domain>[^\n\/]+\.[^\n\/]+)(?:\/(.*))?', url
)
2343 return domain
.group('domain') if domain
else None
2346 def url_basename(url
):
2347 path
= compat_urlparse
.urlparse(url
).path
2348 return path
.strip('/').split('/')[-1]
2352 return re
.match(r
'https?://[^?#&]+/', url
).group()
2355 def urljoin(base
, path
):
2356 if isinstance(path
, bytes):
2357 path
= path
.decode('utf-8')
2358 if not isinstance(path
, compat_str
) or not path
:
2360 if re
.match(r
'^(?:[a-zA-Z][a-zA-Z0-9+-.]*:)?//', path
):
2362 if isinstance(base
, bytes):
2363 base
= base
.decode('utf-8')
2364 if not isinstance(base
, compat_str
) or not re
.match(
2365 r
'^(?:https?:)?//', base
):
2367 return compat_urlparse
.urljoin(base
, path
)
2370 class HEADRequest(compat_urllib_request
.Request
):
2371 def get_method(self
):
2375 class PUTRequest(compat_urllib_request
.Request
):
2376 def get_method(self
):
2380 def int_or_none(v
, scale
=1, default
=None, get_attr
=None, invscale
=1):
2383 v
= getattr(v
, get_attr
, None)
2389 return int(v
) * invscale
// scale
2390 except (ValueError, TypeError, OverflowError):
2394 def str_or_none(v
, default
=None):
2395 return default
if v
is None else compat_str(v
)
2398 def str_to_int(int_str
):
2399 """ A more relaxed version of int_or_none """
2400 if isinstance(int_str
, compat_integer_types
):
2402 elif isinstance(int_str
, compat_str
):
2403 int_str
= re
.sub(r
'[,\.\+]', '', int_str
)
2404 return int_or_none(int_str
)
2407 def float_or_none(v
, scale
=1, invscale
=1, default
=None):
2411 return float(v
) * invscale
/ scale
2412 except (ValueError, TypeError):
2416 def bool_or_none(v
, default
=None):
2417 return v
if isinstance(v
, bool) else default
2420 def strip_or_none(v
, default
=None):
2421 return v
.strip() if isinstance(v
, compat_str
) else default
2424 def url_or_none(url
):
2425 if not url
or not isinstance(url
, compat_str
):
2428 return url
if re
.match(r
'^(?:(?:https?|rt(?:m(?:pt?[es]?|fp)|sp[su]?)|mms|ftps?):)?//', url
) else None
2431 def strftime_or_none(timestamp
, date_format
, default
=None):
2432 datetime_object
= None
2434 if isinstance(timestamp
, compat_numeric_types
): # unix timestamp
2435 datetime_object
= datetime
.datetime
.utcfromtimestamp(timestamp
)
2436 elif isinstance(timestamp
, compat_str
): # assume YYYYMMDD
2437 datetime_object
= datetime
.datetime
.strptime(timestamp
, '%Y%m%d')
2438 return datetime_object
.strftime(date_format
)
2439 except (ValueError, TypeError, AttributeError):
2443 def parse_duration(s
):
2444 if not isinstance(s
, compat_basestring
):
2450 days
, hours
, mins
, secs
, ms
= [None] * 5
2451 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?$', s
)
2453 days
, hours
, mins
, secs
, ms
= m
.groups()
2458 [0-9]+\s*y(?:ears?)?\s*
2461 [0-9]+\s*m(?:onths?)?\s*
2464 [0-9]+\s*w(?:eeks?)?\s*
2467 (?P<days>[0-9]+)\s*d(?:ays?)?\s*
2471 (?P<hours>[0-9]+)\s*h(?:ours?)?\s*
2474 (?P<mins>[0-9]+)\s*m(?:in(?:ute)?s?)?\s*
2477 (?P<secs>[0-9]+)(?P<ms>\.[0-9]+)?\s*s(?:ec(?:ond)?s?)?\s*
2480 days
, hours
, mins
, secs
, ms
= m
.groups()
2482 m
= re
.match(r
'(?i)(?:(?P<hours>[0-9.]+)\s*(?:hours?)|(?P<mins>[0-9.]+)\s*(?:mins?\.?|minutes?)\s*)Z?$', s
)
2484 hours
, mins
= m
.groups()
2490 duration
+= float(secs
)
2492 duration
+= float(mins
) * 60
2494 duration
+= float(hours
) * 60 * 60
2496 duration
+= float(days
) * 24 * 60 * 60
2498 duration
+= float(ms
)
2502 def prepend_extension(filename
, ext
, expected_real_ext
=None):
2503 name
, real_ext
= os
.path
.splitext(filename
)
2505 '{0}.{1}{2}'.format(name
, ext
, real_ext
)
2506 if not expected_real_ext
or real_ext
[1:] == expected_real_ext
2507 else '{0}.{1}'.format(filename
, ext
))
2510 def replace_extension(filename
, ext
, expected_real_ext
=None):
2511 name
, real_ext
= os
.path
.splitext(filename
)
2512 return '{0}.{1}'.format(
2513 name
if not expected_real_ext
or real_ext
[1:] == expected_real_ext
else filename
,
2517 def check_executable(exe
, args
=[]):
2518 """ Checks if the given binary is installed somewhere in PATH, and returns its name.
2519 args can be a list of arguments for a short output (like -version) """
2521 Popen([exe
] + args
, stdout
=subprocess
.PIPE
, stderr
=subprocess
.PIPE
).communicate_or_kill()
2527 def _get_exe_version_output(exe
, args
):
2529 # STDIN should be redirected too. On UNIX-like systems, ffmpeg triggers
2530 # SIGTTOU if yt-dlp is run in the background.
2531 # See https://github.com/ytdl-org/youtube-dl/issues/955#issuecomment-209789656
2533 [encodeArgument(exe
)] + args
, stdin
=subprocess
.PIPE
,
2534 stdout
=subprocess
.PIPE
, stderr
=subprocess
.STDOUT
).communicate_or_kill()
2537 if isinstance(out
, bytes): # Python 2.x
2538 out
= out
.decode('ascii', 'ignore')
2542 def detect_exe_version(output
, version_re
=None, unrecognized
='present'):
2543 assert isinstance(output
, compat_str
)
2544 if version_re
is None:
2545 version_re
= r
'version\s+([-0-9._a-zA-Z]+)'
2546 m
= re
.search(version_re
, output
)
2553 def get_exe_version(exe
, args
=['--version'],
2554 version_re
=None, unrecognized
='present'):
2555 """ Returns the version of the specified executable,
2556 or False if the executable is not present """
2557 out
= _get_exe_version_output(exe
, args
)
2558 return detect_exe_version(out
, version_re
, unrecognized
) if out
else False
2561 class LazyList(collections
.abc
.Sequence
):
2562 ''' Lazy immutable list from an iterable
2563 Note that slices of a LazyList are lists and not LazyList'''
2565 class IndexError(IndexError):
2568 def __init__(self
, iterable
, *, reverse
=False, _cache
=None):
2569 self
.__iterable
= iter(iterable
)
2570 self
.__cache
= [] if _cache
is None else _cache
2571 self
.__reversed
= reverse
2575 # We need to consume the entire iterable to iterate in reverse
2576 yield from self
.exhaust()
2578 yield from self
.__cache
2579 for item
in self
.__iterable
:
2580 self
.__cache
.append(item
)
2583 def __exhaust(self
):
2584 self
.__cache
.extend(self
.__iterable
)
2585 # Discard the emptied iterable to make it pickle-able
2586 self
.__iterable
= []
2590 ''' Evaluate the entire iterable '''
2591 return self
.__exhaust
()[::-1 if self
.__reversed
else 1]
2594 def __reverse_index(x
):
2595 return None if x
is None else -(x
+ 1)
2597 def __getitem__(self
, idx
):
2598 if isinstance(idx
, slice):
2600 idx
= slice(self
.__reverse
_index
(idx
.start
), self
.__reverse
_index
(idx
.stop
), -(idx
.step
or 1))
2601 start
, stop
, step
= idx
.start
, idx
.stop
, idx
.step
or 1
2602 elif isinstance(idx
, int):
2604 idx
= self
.__reverse
_index
(idx
)
2605 start
, stop
, step
= idx
, idx
, 0
2607 raise TypeError('indices must be integers or slices')
2608 if ((start
or 0) < 0 or (stop
or 0) < 0
2609 or (start
is None and step
< 0)
2610 or (stop
is None and step
> 0)):
2611 # We need to consume the entire iterable to be able to slice from the end
2612 # Obviously, never use this with infinite iterables
2615 return self
.__cache
[idx
]
2616 except IndexError as e
:
2617 raise self
.IndexError(e
) from e
2618 n
= max(start
or 0, stop
or 0) - len(self
.__cache
) + 1
2620 self
.__cache
.extend(itertools
.islice(self
.__iterable
, n
))
2622 return self
.__cache
[idx
]
2623 except IndexError as e
:
2624 raise self
.IndexError(e
) from e
2628 self
[-1] if self
.__reversed
else self
[0]
2629 except self
.IndexError:
2635 return len(self
.__cache
)
2637 def __reversed__(self
):
2638 return type(self
)(self
.__iterable
, reverse
=not self
.__reversed
, _cache
=self
.__cache
)
2641 return type(self
)(self
.__iterable
, reverse
=self
.__reversed
, _cache
=self
.__cache
)
2644 # repr and str should mimic a list. So we exhaust the iterable
2645 return repr(self
.exhaust())
2648 return repr(self
.exhaust())
2653 class IndexError(IndexError):
2657 # This is only useful for tests
2658 return len(self
.getslice())
2660 def __init__(self
, pagefunc
, pagesize
, use_cache
=True):
2661 self
._pagefunc
= pagefunc
2662 self
._pagesize
= pagesize
2663 self
._use
_cache
= use_cache
2666 def getpage(self
, pagenum
):
2667 page_results
= self
._cache
.get(pagenum
)
2668 if page_results
is None:
2669 page_results
= list(self
._pagefunc
(pagenum
))
2671 self
._cache
[pagenum
] = page_results
2674 def getslice(self
, start
=0, end
=None):
2675 return list(self
._getslice
(start
, end
))
2677 def _getslice(self
, start
, end
):
2678 raise NotImplementedError('This method must be implemented by subclasses')
2680 def __getitem__(self
, idx
):
2681 # NOTE: cache must be enabled if this is used
2682 if not isinstance(idx
, int) or idx
< 0:
2683 raise TypeError('indices must be non-negative integers')
2684 entries
= self
.getslice(idx
, idx
+ 1)
2686 raise self
.IndexError()
2690 class OnDemandPagedList(PagedList
):
2691 def _getslice(self
, start
, end
):
2692 for pagenum
in itertools
.count(start
// self
._pagesize
):
2693 firstid
= pagenum
* self
._pagesize
2694 nextfirstid
= pagenum
* self
._pagesize
+ self
._pagesize
2695 if start
>= nextfirstid
:
2699 start
% self
._pagesize
2700 if firstid
<= start
< nextfirstid
2703 ((end
- 1) % self
._pagesize
) + 1
2704 if (end
is not None and firstid
<= end
<= nextfirstid
)
2707 page_results
= self
.getpage(pagenum
)
2708 if startv
!= 0 or endv
is not None:
2709 page_results
= page_results
[startv
:endv
]
2710 yield from page_results
2712 # A little optimization - if current page is not "full", ie. does
2713 # not contain page_size videos then we can assume that this page
2714 # is the last one - there are no more ids on further pages -
2715 # i.e. no need to query again.
2716 if len(page_results
) + startv
< self
._pagesize
:
2719 # If we got the whole page, but the next page is not interesting,
2720 # break out early as well
2721 if end
== nextfirstid
:
2725 class InAdvancePagedList(PagedList
):
2726 def __init__(self
, pagefunc
, pagecount
, pagesize
):
2727 self
._pagecount
= pagecount
2728 PagedList
.__init
__(self
, pagefunc
, pagesize
, True)
2730 def _getslice(self
, start
, end
):
2731 start_page
= start
// self
._pagesize
2733 self
._pagecount
if end
is None else (end
// self
._pagesize
+ 1))
2734 skip_elems
= start
- start_page
* self
._pagesize
2735 only_more
= None if end
is None else end
- start
2736 for pagenum
in range(start_page
, end_page
):
2737 page_results
= self
.getpage(pagenum
)
2739 page_results
= page_results
[skip_elems
:]
2741 if only_more
is not None:
2742 if len(page_results
) < only_more
:
2743 only_more
-= len(page_results
)
2745 yield from page_results
[:only_more
]
2747 yield from page_results
2750 def uppercase_escape(s
):
2751 unicode_escape
= codecs
.getdecoder('unicode_escape')
2753 r
'\\U[0-9a-fA-F]{8}',
2754 lambda m
: unicode_escape(m
.group(0))[0],
2758 def lowercase_escape(s
):
2759 unicode_escape
= codecs
.getdecoder('unicode_escape')
2761 r
'\\u[0-9a-fA-F]{4}',
2762 lambda m
: unicode_escape(m
.group(0))[0],
2766 def escape_rfc3986(s
):
2767 """Escape non-ASCII characters as suggested by RFC 3986"""
2768 if sys
.version_info
< (3, 0) and isinstance(s
, compat_str
):
2769 s
= s
.encode('utf-8')
2770 return compat_urllib_parse
.quote(s
, b
"%/;:@&=+$,!~*'()?#[]")
2773 def escape_url(url
):
2774 """Escape URL as suggested by RFC 3986"""
2775 url_parsed
= compat_urllib_parse_urlparse(url
)
2776 return url_parsed
._replace
(
2777 netloc
=url_parsed
.netloc
.encode('idna').decode('ascii'),
2778 path
=escape_rfc3986(url_parsed
.path
),
2779 params
=escape_rfc3986(url_parsed
.params
),
2780 query
=escape_rfc3986(url_parsed
.query
),
2781 fragment
=escape_rfc3986(url_parsed
.fragment
)
2786 return compat_parse_qs(compat_urllib_parse_urlparse(url
).query
)
2789 def read_batch_urls(batch_fd
):
2791 if not isinstance(url
, compat_str
):
2792 url
= url
.decode('utf-8', 'replace')
2793 BOM_UTF8
= ('\xef\xbb\xbf', '\ufeff')
2794 for bom
in BOM_UTF8
:
2795 if url
.startswith(bom
):
2796 url
= url
[len(bom
):]
2798 if not url
or url
.startswith(('#', ';', ']')):
2800 # "#" cannot be stripped out since it is part of the URI
2801 # However, it can be safely stipped out if follwing a whitespace
2802 return re
.split(r
'\s#', url
, 1)[0].rstrip()
2804 with contextlib
.closing(batch_fd
) as fd
:
2805 return [url
for url
in map(fixup
, fd
) if url
]
2808 def urlencode_postdata(*args
, **kargs
):
2809 return compat_urllib_parse_urlencode(*args
, **kargs
).encode('ascii')
2812 def update_url_query(url
, query
):
2815 parsed_url
= compat_urlparse
.urlparse(url
)
2816 qs
= compat_parse_qs(parsed_url
.query
)
2818 return compat_urlparse
.urlunparse(parsed_url
._replace
(
2819 query
=compat_urllib_parse_urlencode(qs
, True)))
2822 def update_Request(req
, url
=None, data
=None, headers
={}, query={}
):
2823 req_headers
= req
.headers
.copy()
2824 req_headers
.update(headers
)
2825 req_data
= data
or req
.data
2826 req_url
= update_url_query(url
or req
.get_full_url(), query
)
2827 req_get_method
= req
.get_method()
2828 if req_get_method
== 'HEAD':
2829 req_type
= HEADRequest
2830 elif req_get_method
== 'PUT':
2831 req_type
= PUTRequest
2833 req_type
= compat_urllib_request
.Request
2835 req_url
, data
=req_data
, headers
=req_headers
,
2836 origin_req_host
=req
.origin_req_host
, unverifiable
=req
.unverifiable
)
2837 if hasattr(req
, 'timeout'):
2838 new_req
.timeout
= req
.timeout
2842 def _multipart_encode_impl(data
, boundary
):
2843 content_type
= 'multipart/form-data; boundary=%s' % boundary
2846 for k
, v
in data
.items():
2847 out
+= b
'--' + boundary
.encode('ascii') + b
'\r\n'
2848 if isinstance(k
, compat_str
):
2849 k
= k
.encode('utf-8')
2850 if isinstance(v
, compat_str
):
2851 v
= v
.encode('utf-8')
2852 # RFC 2047 requires non-ASCII field names to be encoded, while RFC 7578
2853 # suggests sending UTF-8 directly. Firefox sends UTF-8, too
2854 content
= b
'Content-Disposition: form-data; name="' + k
+ b
'"\r\n\r\n' + v
+ b
'\r\n'
2855 if boundary
.encode('ascii') in content
:
2856 raise ValueError('Boundary overlaps with data')
2859 out
+= b
'--' + boundary
.encode('ascii') + b
'--\r\n'
2861 return out
, content_type
2864 def multipart_encode(data
, boundary
=None):
2866 Encode a dict to RFC 7578-compliant form-data
2869 A dict where keys and values can be either Unicode or bytes-like
2872 If specified a Unicode object, it's used as the boundary. Otherwise
2873 a random boundary is generated.
2875 Reference: https://tools.ietf.org/html/rfc7578
2877 has_specified_boundary
= boundary
is not None
2880 if boundary
is None:
2881 boundary
= '---------------' + str(random
.randrange(0x0fffffff, 0xffffffff))
2884 out
, content_type
= _multipart_encode_impl(data
, boundary
)
2887 if has_specified_boundary
:
2891 return out
, content_type
2894 def dict_get(d
, key_or_keys
, default
=None, skip_false_values
=True):
2895 if isinstance(key_or_keys
, (list, tuple)):
2896 for key
in key_or_keys
:
2897 if key
not in d
or d
[key
] is None or skip_false_values
and not d
[key
]:
2901 return d
.get(key_or_keys
, default
)
2904 def try_get(src
, getter
, expected_type
=None):
2905 for get
in variadic(getter
):
2908 except (AttributeError, KeyError, TypeError, IndexError):
2911 if expected_type
is None or isinstance(v
, expected_type
):
2915 def merge_dicts(*dicts
):
2917 for a_dict
in dicts
:
2918 for k
, v
in a_dict
.items():
2922 or (isinstance(v
, compat_str
) and v
2923 and isinstance(merged
[k
], compat_str
)
2924 and not merged
[k
])):
2929 def encode_compat_str(string
, encoding
=preferredencoding(), errors
='strict'):
2930 return string
if isinstance(string
, compat_str
) else compat_str(string
, encoding
, errors
)
2942 TV_PARENTAL_GUIDELINES
= {
2952 def parse_age_limit(s
):
2954 return s
if 0 <= s
<= 21 else None
2955 if not isinstance(s
, compat_basestring
):
2957 m
= re
.match(r
'^(?P<age>\d{1,2})\+?$', s
)
2959 return int(m
.group('age'))
2962 return US_RATINGS
[s
]
2963 m
= re
.match(r
'^TV[_-]?(%s)$' % '|'.join(k
[3:] for k
in TV_PARENTAL_GUIDELINES
), s
)
2965 return TV_PARENTAL_GUIDELINES
['TV-' + m
.group(1)]
2969 def strip_jsonp(code
):
2972 (?:window\.)?(?P<func_name>[a-zA-Z0-9_.$]*)
2973 (?:\s*&&\s*(?P=func_name))?
2974 \s*\(\s*(?P<callback_data>.*)\);?
2975 \s*?(?://[^\n]*)*$''',
2976 r
'\g<callback_data>', code
)
2979 def js_to_json(code
, vars={}):
2980 # vars is a dict of var, val pairs to substitute
2981 COMMENT_RE
= r
'/\*(?:(?!\*/).)*?\*/|//[^\n]*\n'
2982 SKIP_RE
= r
'\s*(?:{comment})?\s*'.format(comment
=COMMENT_RE
)
2984 (r
'(?s)^(0[xX][0-9a-fA-F]+){skip}:?$'.format(skip
=SKIP_RE
), 16),
2985 (r
'(?s)^(0+[0-7]+){skip}:?$'.format(skip
=SKIP_RE
), 8),
2990 if v
in ('true', 'false', 'null'):
2992 elif v
in ('undefined', 'void 0'):
2994 elif v
.startswith('/*') or v
.startswith('//') or v
.startswith('!') or v
== ',':
2997 if v
[0] in ("'", '"'):
2998 v
= re
.sub(r
'(?s)\\.|"', lambda m
: {
3003 }.get(m
.group(0), m
.group(0)), v
[1:-1])
3005 for regex
, base
in INTEGER_TABLE
:
3006 im
= re
.match(regex
, v
)
3008 i
= int(im
.group(1), base
)
3009 return '"%d":' % i
if v
.endswith(':') else '%d' % i
3016 return re
.sub(r
'''(?sx)
3017 "(?:[^"\\]*(?:\\\\|\\['"nurtbfx/\n]))*[^"\\]*"|
3018 '(?:[^'\\]*(?:\\\\|\\['"nurtbfx/\n]))*[^'\\]*'|
3019 {comment}|,(?={skip}[\]}}])|
3020 void\s0|(?:(?<![0-9])[eE]|[a-df-zA-DF-Z_$])[.a-zA-Z_$0-9]*|
3021 \b(?:0[xX][0-9a-fA-F]+|0+[0-7]+)(?:{skip}:)?|
3024 '''.format(comment
=COMMENT_RE
, skip
=SKIP_RE
), fix_kv
, code
)
3027 def qualities(quality_ids
):
3028 """ Get a numeric quality value out of a list of possible values """
3031 return quality_ids
.index(qid
)
3038 'default': '%(title)s [%(id)s].%(ext)s',
3039 'chapter': '%(title)s - %(section_number)03d %(section_title)s [%(id)s].%(ext)s',
3045 'description': 'description',
3046 'annotation': 'annotations.xml',
3047 'infojson': 'info.json',
3049 'pl_thumbnail': None,
3050 'pl_description': 'description',
3051 'pl_infojson': 'info.json',
3054 # As of [1] format syntax is:
3055 # %[mapping_key][conversion_flags][minimum_width][.precision][length_modifier]type
3056 # 1. https://docs.python.org/2/library/stdtypes.html#string-formatting
3057 STR_FORMAT_RE_TMPL
= r
'''(?x)
3058 (?<!%)(?P<prefix>(?:%%)*)
3060 (?P<has_key>\((?P<key>{0})\))?
3062 (?P<conversion>[#0\-+ ]+)?
3064 (?P<precision>\.\d+)?
3065 (?P<len_mod>[hlL])? # unused in python
3066 {1} # conversion type
3071 STR_FORMAT_TYPES
= 'diouxXeEfFgGcrs'
3074 def limit_length(s
, length
):
3075 """ Add ellipses to overly long strings """
3080 return s
[:length
- len(ELLIPSES
)] + ELLIPSES
3084 def version_tuple(v
):
3085 return tuple(int(e
) for e
in re
.split(r
'[-.]', v
))
3088 def is_outdated_version(version
, limit
, assume_new
=True):
3090 return not assume_new
3092 return version_tuple(version
) < version_tuple(limit
)
3094 return not assume_new
3097 def ytdl_is_updateable():
3098 """ Returns if yt-dlp can be updated with -U """
3100 from .update
import is_non_updateable
3102 return not is_non_updateable()
3105 def args_to_str(args
):
3106 # Get a short string representation for a subprocess command
3107 return ' '.join(compat_shlex_quote(a
) for a
in args
)
3110 def error_to_compat_str(err
):
3112 # On python 2 error byte string must be decoded with proper
3113 # encoding rather than ascii
3114 if sys
.version_info
[0] < 3:
3115 err_str
= err_str
.decode(preferredencoding())
3119 def mimetype2ext(mt
):
3123 mt
, _
, params
= mt
.partition(';')
3128 # Per RFC 3003, audio/mpeg can be .mp1, .mp2 or .mp3. Here use .mp3 as
3129 # it's the most popular one
3130 'audio/mpeg': 'mp3',
3131 'audio/x-wav': 'wav',
3133 'audio/wave': 'wav',
3136 ext
= FULL_MAP
.get(mt
)
3142 'smptett+xml': 'tt',
3146 'x-mp4-fragmented': 'mp4',
3147 'x-ms-sami': 'sami',
3150 'x-mpegurl': 'm3u8',
3151 'vnd.apple.mpegurl': 'm3u8',
3155 'vnd.ms-sstr+xml': 'ism',
3159 'filmstrip+json': 'fs',
3163 _
, _
, subtype
= mt
.rpartition('/')
3164 ext
= SUBTYPE_MAP
.get(subtype
.lower())
3175 _
, _
, suffix
= subtype
.partition('+')
3176 ext
= SUFFIX_MAP
.get(suffix
)
3180 return subtype
.replace('+', '.')
3183 def ext2mimetype(ext_or_url
):
3186 if '.' not in ext_or_url
:
3187 ext_or_url
= f
'file.{ext_or_url}'
3188 return mimetypes
.guess_type(ext_or_url
)[0]
3191 def parse_codecs(codecs_str
):
3192 # http://tools.ietf.org/html/rfc6381
3195 split_codecs
= list(filter(None, map(
3196 str.strip
, codecs_str
.strip().strip(',').split(','))))
3197 vcodec
, acodec
, hdr
= None, None, None
3198 for full_codec
in split_codecs
:
3199 parts
= full_codec
.split('.')
3200 codec
= parts
[0].replace('0', '')
3201 if codec
in ('avc1', 'avc2', 'avc3', 'avc4', 'vp9', 'vp8', 'hev1', 'hev2',
3202 'h263', 'h264', 'mp4v', 'hvc1', 'av1', 'theora', 'dvh1', 'dvhe'):
3204 vcodec
= '.'.join(parts
[:4]) if codec
in ('vp9', 'av1', 'hvc1') else full_codec
3205 if codec
in ('dvh1', 'dvhe'):
3207 elif codec
== 'av1' and len(parts
) > 3 and parts
[3] == '10':
3209 elif full_codec
.replace('0', '').startswith('vp9.2'):
3211 elif codec
in ('flac', 'mp4a', 'opus', 'vorbis', 'mp3', 'aac', 'ac-3', 'ec-3', 'eac3', 'dtsc', 'dtse', 'dtsh', 'dtsl'):
3215 write_string('WARNING: Unknown codec %s\n' % full_codec
, sys
.stderr
)
3216 if vcodec
or acodec
:
3218 'vcodec': vcodec
or 'none',
3219 'acodec': acodec
or 'none',
3220 'dynamic_range': hdr
,
3222 elif len(split_codecs
) == 2:
3224 'vcodec': split_codecs
[0],
3225 'acodec': split_codecs
[1],
3230 def urlhandle_detect_ext(url_handle
):
3231 getheader
= url_handle
.headers
.get
3233 cd
= getheader('Content-Disposition')
3235 m
= re
.match(r
'attachment;\s*filename="(?P<filename>[^"]+)"', cd
)
3237 e
= determine_ext(m
.group('filename'), default_ext
=None)
3241 return mimetype2ext(getheader('Content-Type'))
3244 def encode_data_uri(data
, mime_type
):
3245 return 'data:%s;base64,%s' % (mime_type
, base64
.b64encode(data
).decode('ascii'))
3248 def age_restricted(content_limit
, age_limit
):
3249 """ Returns True iff the content should be blocked """
3251 if age_limit
is None: # No limit set
3253 if content_limit
is None:
3254 return False # Content available for everyone
3255 return age_limit
< content_limit
3258 def is_html(first_bytes
):
3259 """ Detect whether a file contains HTML by examining its first bytes. """
3262 (b
'\xef\xbb\xbf', 'utf-8'),
3263 (b
'\x00\x00\xfe\xff', 'utf-32-be'),
3264 (b
'\xff\xfe\x00\x00', 'utf-32-le'),
3265 (b
'\xff\xfe', 'utf-16-le'),
3266 (b
'\xfe\xff', 'utf-16-be'),
3268 for bom
, enc
in BOMS
:
3269 if first_bytes
.startswith(bom
):
3270 s
= first_bytes
[len(bom
):].decode(enc
, 'replace')
3273 s
= first_bytes
.decode('utf-8', 'replace')
3275 return re
.match(r
'^\s*<', s
)
3278 def determine_protocol(info_dict
):
3279 protocol
= info_dict
.get('protocol')
3280 if protocol
is not None:
3283 url
= sanitize_url(info_dict
['url'])
3284 if url
.startswith('rtmp'):
3286 elif url
.startswith('mms'):
3288 elif url
.startswith('rtsp'):
3291 ext
= determine_ext(url
)
3297 return compat_urllib_parse_urlparse(url
).scheme
3300 def render_table(header_row
, data
, delim
=False, extra_gap
=0, hide_empty
=False):
3301 """ Render a list of rows, each as a list of values.
3302 Text after a \t will be right aligned """
3304 return len(remove_terminal_sequences(string
).replace('\t', ''))
3306 def get_max_lens(table
):
3307 return [max(width(str(v
)) for v
in col
) for col
in zip(*table
)]
3309 def filter_using_list(row
, filterArray
):
3310 return [col
for (take
, col
) in zip(filterArray
, row
) if take
]
3313 max_lens
= get_max_lens(data
)
3314 header_row
= filter_using_list(header_row
, max_lens
)
3315 data
= [filter_using_list(row
, max_lens
) for row
in data
]
3317 table
= [header_row
] + data
3318 max_lens
= get_max_lens(table
)
3321 table
= [header_row
, [delim
* (ml
+ extra_gap
) for ml
in max_lens
]] + data
3322 table
[1][-1] = table
[1][-1][:-extra_gap
] # Remove extra_gap from end of delimiter
3324 for pos
, text
in enumerate(map(str, row
)):
3326 row
[pos
] = text
.replace('\t', ' ' * (max_lens
[pos
] - width(text
))) + ' ' * extra_gap
3328 row
[pos
] = text
+ ' ' * (max_lens
[pos
] - width(text
) + extra_gap
)
3329 ret
= '\n'.join(''.join(row
).rstrip() for row
in table
)
3333 def _match_one(filter_part
, dct
, incomplete
):
3334 # TODO: Generalize code with YoutubeDL._build_format_filter
3335 STRING_OPERATORS
= {
3336 '*=': operator
.contains
,
3337 '^=': lambda attr
, value
: attr
.startswith(value
),
3338 '$=': lambda attr
, value
: attr
.endswith(value
),
3339 '~=': lambda attr
, value
: re
.search(value
, attr
),
3341 COMPARISON_OPERATORS
= {
3343 '<=': operator
.le
, # "<=" must be defined above "<"
3350 operator_rex
= re
.compile(r
'''(?x)\s*
3352 \s*(?P<negation>!\s*)?(?P<op>%s)(?P<none_inclusive>\s*\?)?\s*
3354 (?P<quote>["\'])(?P
<quotedstrval
>.+?
)(?P
=quote
)|
3358 ''' % '|'.join(map(re.escape, COMPARISON_OPERATORS.keys())))
3359 m = operator_rex.search(filter_part)
3362 unnegated_op = COMPARISON_OPERATORS[m['op']]
3364 op = lambda attr, value: not unnegated_op(attr, value)
3367 comparison_value = m['quotedstrval'] or m['strval'] or m['intval']
3369 comparison_value = comparison_value.replace(r'\%s' % m['quote'], m['quote'])
3370 actual_value = dct.get(m['key'])
3371 numeric_comparison = None
3372 if isinstance(actual_value, compat_numeric_types):
3373 # If the original field is a string and matching comparisonvalue is
3374 # a number we should respect the origin of the original field
3375 # and process comparison value as a string (see
3376 # https://github.com/ytdl-org/youtube-dl/issues/11082)
3378 numeric_comparison = int(comparison_value)
3380 numeric_comparison = parse_filesize(comparison_value)
3381 if numeric_comparison is None:
3382 numeric_comparison = parse_filesize(f'{comparison_value}B')
3383 if numeric_comparison is None:
3384 numeric_comparison = parse_duration(comparison_value)
3385 if numeric_comparison is not None and m['op'] in STRING_OPERATORS:
3386 raise ValueError('Operator %s only supports string values!' % m['op'])
3387 if actual_value is None:
3388 return incomplete or m['none_inclusive']
3389 return op(actual_value, comparison_value if numeric_comparison is None else numeric_comparison)
3392 '': lambda v: (v is True) if isinstance(v, bool) else (v is not None),
3393 '!': lambda v: (v is False) if isinstance(v, bool) else (v is None),
3395 operator_rex = re.compile(r'''(?x
)\s
*
3396 (?P
<op
>%s)\s
*(?P
<key
>[a
-z_
]+)
3398 ''' % '|'.join(map(re.escape, UNARY_OPERATORS.keys())))
3399 m = operator_rex.search(filter_part)
3401 op = UNARY_OPERATORS[m.group('op')]
3402 actual_value = dct.get(m.group('key'))
3403 if incomplete and actual_value is None:
3405 return op(actual_value)
3407 raise ValueError('Invalid filter part %r' % filter_part)
3410 def match_str(filter_str, dct, incomplete=False):
3411 """ Filter a dictionary with a simple string syntax. Returns True (=passes filter) or false
3412 When incomplete, all conditions passes on missing fields
3415 _match_one(filter_part.replace(r'\&', '&'), dct, incomplete)
3416 for filter_part in re.split(r'(?<!\\)&', filter_str))
3419 def match_filter_func(filter_str):
3420 def _match_func(info_dict, *args, **kwargs):
3421 if match_str(filter_str, info_dict, *args, **kwargs):
3424 video_title = info_dict.get('title', info_dict.get('id', 'video'))
3425 return '%s does not pass filter %s, skipping ..' % (video_title, filter_str)
3429 def parse_dfxp_time_expr(time_expr):
3433 mobj = re.match(r'^(?P<time_offset>\d+(?:\.\d+)?)s?$', time_expr)
3435 return float(mobj.group('time_offset'))
3437 mobj = re.match(r'^(\d+):(\d\d):(\d\d(?:(?:\.|:)\d+)?)$', time_expr)
3439 return 3600 * int(mobj.group(1)) + 60 * int(mobj.group(2)) + float(mobj.group(3).replace(':', '.'))
3442 def srt_subtitles_timecode(seconds):
3443 return '%02d:%02d:%02d,%03d' % timetuple_from_msec(seconds * 1000)
3446 def ass_subtitles_timecode(seconds):
3447 time = timetuple_from_msec(seconds * 1000)
3448 return '%01d:%02d:%02d.%02d' % (*time[:-1], time.milliseconds / 10)
3451 def dfxp2srt(dfxp_data):
3453 @param dfxp_data A
bytes-like
object containing DFXP data
3454 @returns A
unicode object containing converted SRT data
3456 LEGACY_NAMESPACES = (
3457 (b'http://www.w3.org/ns/ttml', [
3458 b'http://www.w3.org/2004/11/ttaf1',
3459 b'http://www.w3.org/2006/04/ttaf1',
3460 b'http://www.w3.org/2006/10/ttaf1',
3462 (b'http://www.w3.org/ns/ttml#styling', [
3463 b'http://www.w3.org/ns/ttml#style',
3467 SUPPORTED_STYLING = [
3476 _x = functools.partial(xpath_with_ns, ns_map={
3477 'xml': 'http://www.w3.org/XML/1998/namespace',
3478 'ttml': 'http://www.w3.org/ns/ttml',
3479 'tts': 'http://www.w3.org/ns/ttml#styling',
3485 class TTMLPElementParser(object):
3487 _unclosed_elements = []
3488 _applied_styles = []
3490 def start(self, tag, attrib):
3491 if tag in (_x('ttml:br'), 'br'):
3494 unclosed_elements = []
3496 element_style_id = attrib.get('style')
3498 style.update(default_style)
3499 if element_style_id:
3500 style.update(styles.get(element_style_id, {}))
3501 for prop in SUPPORTED_STYLING:
3502 prop_val = attrib.get(_x('tts:' + prop))
3504 style[prop] = prop_val
3507 for k, v in sorted(style.items()):
3508 if self._applied_styles and self._applied_styles[-1].get(k) == v:
3511 font += ' color="%s"' % v
3512 elif k == 'fontSize':
3513 font += ' size="%s"' % v
3514 elif k == 'fontFamily':
3515 font += ' face="%s"' % v
3516 elif k == 'fontWeight' and v == 'bold':
3518 unclosed_elements.append('b')
3519 elif k == 'fontStyle' and v == 'italic':
3521 unclosed_elements.append('i')
3522 elif k == 'textDecoration' and v == 'underline':
3524 unclosed_elements.append('u')
3526 self._out += '<font' + font + '>'
3527 unclosed_elements.append('font')
3529 if self._applied_styles:
3530 applied_style.update(self._applied_styles[-1])
3531 applied_style.update(style)
3532 self._applied_styles.append(applied_style)
3533 self._unclosed_elements.append(unclosed_elements)
3536 if tag not in (_x('ttml:br'), 'br'):
3537 unclosed_elements = self._unclosed_elements.pop()
3538 for element in reversed(unclosed_elements):
3539 self._out += '</%s>' % element
3540 if unclosed_elements and self._applied_styles:
3541 self._applied_styles.pop()
3543 def data(self, data):
3547 return self._out.strip()
3549 def parse_node(node):
3550 target = TTMLPElementParser()
3551 parser = xml.etree.ElementTree.XMLParser(target=target)
3552 parser.feed(xml.etree.ElementTree.tostring(node))
3553 return parser.close()
3555 for k, v in LEGACY_NAMESPACES:
3557 dfxp_data = dfxp_data.replace(ns, k)
3559 dfxp = compat_etree_fromstring(dfxp_data)
3561 paras = dfxp.findall(_x('.//ttml:p')) or dfxp.findall('.//p')
3564 raise ValueError('Invalid dfxp/TTML subtitle')
3568 for style in dfxp.findall(_x('.//ttml:style')):
3569 style_id = style.get('id') or style.get(_x('xml:id'))
3572 parent_style_id = style.get('style')
3574 if parent_style_id not in styles:
3577 styles[style_id] = styles[parent_style_id].copy()
3578 for prop in SUPPORTED_STYLING:
3579 prop_val = style.get(_x('tts:' + prop))
3581 styles.setdefault(style_id, {})[prop] = prop_val
3587 for p in ('body', 'div'):
3588 ele = xpath_element(dfxp, [_x('.//ttml:' + p), './/' + p])
3591 style = styles.get(ele.get('style'))
3594 default_style.update(style)
3596 for para, index in zip(paras, itertools.count(1)):
3597 begin_time = parse_dfxp_time_expr(para.attrib.get('begin'))
3598 end_time = parse_dfxp_time_expr(para.attrib.get('end'))
3599 dur = parse_dfxp_time_expr(para.attrib.get('dur'))
3600 if begin_time is None:
3605 end_time = begin_time + dur
3606 out.append('%d\n%s --> %s\n%s\n\n' % (
3608 srt_subtitles_timecode(begin_time),
3609 srt_subtitles_timecode(end_time),
3615 def cli_option(params, command_option, param):
3616 param = params.get(param)
3618 param = compat_str(param)
3619 return [command_option, param] if param is not None else []
3622 def cli_bool_option(params, command_option, param, true_value='true', false_value='false', separator=None):
3623 param = params.get(param)
3626 assert isinstance(param, bool)
3628 return [command_option + separator + (true_value if param else false_value)]
3629 return [command_option, true_value if param else false_value]
3632 def cli_valueless_option(params, command_option, param, expected_value=True):
3633 param = params.get(param)
3634 return [command_option] if param == expected_value else []
3637 def cli_configuration_args(argdict, keys, default=[], use_compat=True):
3638 if isinstance(argdict, (list, tuple)): # for backward compatibility
3645 assert isinstance(argdict, dict)
3647 assert isinstance(keys, (list, tuple))
3648 for key_list in keys:
3649 arg_list = list(filter(
3650 lambda x: x is not None,
3651 [argdict.get(key.lower()) for key in variadic(key_list)]))
3653 return [arg for args in arg_list for arg in args]
3657 def _configuration_args(main_key, argdict, exe, keys=None, default=[], use_compat=True):
3658 main_key, exe = main_key.lower(), exe.lower()
3659 root_key = exe if main_key == exe else f'{main_key}+{exe}'
3660 keys = [f'{root_key}{k}' for k in (keys or [''])]
3661 if root_key in keys:
3663 keys.append((main_key, exe))
3664 keys.append('default')
3667 return cli_configuration_args(argdict, keys, default, use_compat)
3670 class ISO639Utils(object):
3671 # See http://www.loc.gov/standards/iso639-2/ISO-639-2_utf-8.txt
3730 'iw': 'heb', # Replaced by he in 1989 revision
3740 'in': 'ind', # Replaced by id in 1989 revision
3855 'ji': 'yid', # Replaced by yi in 1989 revision
3863 def short2long(cls, code):
3864 """Convert language code from ISO 639-1 to ISO 639-2/T"""
3865 return cls._lang_map.get(code[:2])
3868 def long2short(cls, code):
3869 """Convert language code from ISO 639-2/T to ISO 639-1"""
3870 for short_name, long_name in cls._lang_map.items():
3871 if long_name == code:
3875 class ISO3166Utils(object):
3876 # From http://data.okfn.org/data/core/country-list
3878 'AF': 'Afghanistan',
3879 'AX': 'Åland Islands',
3882 'AS': 'American Samoa',
3887 'AG': 'Antigua and Barbuda',
3904 'BO': 'Bolivia, Plurinational State of',
3905 'BQ': 'Bonaire, Sint Eustatius and Saba',
3906 'BA': 'Bosnia and Herzegovina',
3908 'BV': 'Bouvet Island',
3910 'IO': 'British Indian Ocean Territory',
3911 'BN': 'Brunei Darussalam',
3913 'BF': 'Burkina Faso',
3919 'KY': 'Cayman Islands',
3920 'CF': 'Central African Republic',
3924 'CX': 'Christmas Island',
3925 'CC': 'Cocos (Keeling) Islands',
3929 'CD': 'Congo, the Democratic Republic of the',
3930 'CK': 'Cook Islands',
3932 'CI': 'Côte d\'Ivoire',
3937 'CZ': 'Czech Republic',
3941 'DO': 'Dominican Republic',
3944 'SV': 'El Salvador',
3945 'GQ': 'Equatorial Guinea',
3949 'FK': 'Falkland Islands (Malvinas)',
3950 'FO': 'Faroe Islands',
3954 'GF': 'French Guiana',
3955 'PF': 'French Polynesia',
3956 'TF': 'French Southern Territories',
3971 'GW': 'Guinea-Bissau',
3974 'HM': 'Heard Island and McDonald Islands',
3975 'VA': 'Holy See (Vatican City State)',
3982 'IR': 'Iran, Islamic Republic of',
3985 'IM': 'Isle of Man',
3995 'KP': 'Korea, Democratic People\'s Republic of',
3996 'KR': 'Korea, Republic of',
3999 'LA': 'Lao People\'s Democratic Republic',
4005 'LI': 'Liechtenstein',
4009 'MK': 'Macedonia, the Former Yugoslav Republic of',
4016 'MH': 'Marshall Islands',
4022 'FM': 'Micronesia, Federated States of',
4023 'MD': 'Moldova, Republic of',
4034 'NL': 'Netherlands',
4035 'NC': 'New Caledonia',
4036 'NZ': 'New Zealand',
4041 'NF': 'Norfolk Island',
4042 'MP': 'Northern Mariana Islands',
4047 'PS': 'Palestine, State of',
4049 'PG': 'Papua New Guinea',
4052 'PH': 'Philippines',
4056 'PR': 'Puerto Rico',
4060 'RU': 'Russian Federation',
4062 'BL': 'Saint Barthélemy',
4063 'SH': 'Saint Helena, Ascension and Tristan da Cunha',
4064 'KN': 'Saint Kitts and Nevis',
4065 'LC': 'Saint Lucia',
4066 'MF': 'Saint Martin (French part)',
4067 'PM': 'Saint Pierre and Miquelon',
4068 'VC': 'Saint Vincent and the Grenadines',
4071 'ST': 'Sao Tome and Principe',
4072 'SA': 'Saudi Arabia',
4076 'SL': 'Sierra Leone',
4078 'SX': 'Sint Maarten (Dutch part)',
4081 'SB': 'Solomon Islands',
4083 'ZA': 'South Africa',
4084 'GS': 'South Georgia and the South Sandwich Islands',
4085 'SS': 'South Sudan',
4090 'SJ': 'Svalbard and Jan Mayen',
4093 'CH': 'Switzerland',
4094 'SY': 'Syrian Arab Republic',
4095 'TW': 'Taiwan, Province of China',
4097 'TZ': 'Tanzania, United Republic of',
4099 'TL': 'Timor-Leste',
4103 'TT': 'Trinidad and Tobago',
4106 'TM': 'Turkmenistan',
4107 'TC': 'Turks and Caicos Islands',
4111 'AE': 'United Arab Emirates',
4112 'GB': 'United Kingdom',
4113 'US': 'United States',
4114 'UM': 'United States Minor Outlying Islands',
4118 'VE': 'Venezuela, Bolivarian Republic of',
4120 'VG': 'Virgin Islands, British',
4121 'VI': 'Virgin Islands, U.S.',
4122 'WF': 'Wallis and Futuna',
4123 'EH': 'Western Sahara',
4130 def short2full(cls, code):
4131 """Convert an ISO 3166-2 country code to the corresponding full name"""
4132 return cls._country_map.get(code.upper())
4135 class GeoUtils(object):
4136 # Major IPv4 address blocks per country
4138 'AD': '46.172.224.0/19',
4139 'AE': '94.200.0.0/13',
4140 'AF': '149.54.0.0/17',
4141 'AG': '209.59.64.0/18',
4142 'AI': '204.14.248.0/21',
4143 'AL': '46.99.0.0/16',
4144 'AM': '46.70.0.0/15',
4145 'AO': '105.168.0.0/13',
4146 'AP': '182.50.184.0/21',
4147 'AQ': '23.154.160.0/24',
4148 'AR': '181.0.0.0/12',
4149 'AS': '202.70.112.0/20',
4150 'AT': '77.116.0.0/14',
4151 'AU': '1.128.0.0/11',
4152 'AW': '181.41.0.0/18',
4153 'AX': '185.217.4.0/22',
4154 'AZ': '5.197.0.0/16',
4155 'BA': '31.176.128.0/17',
4156 'BB': '65.48.128.0/17',
4157 'BD': '114.130.0.0/16',
4159 'BF': '102.178.0.0/15',
4160 'BG': '95.42.0.0/15',
4161 'BH': '37.131.0.0/17',
4162 'BI': '154.117.192.0/18',
4163 'BJ': '137.255.0.0/16',
4164 'BL': '185.212.72.0/23',
4165 'BM': '196.12.64.0/18',
4166 'BN': '156.31.0.0/16',
4167 'BO': '161.56.0.0/16',
4168 'BQ': '161.0.80.0/20',
4169 'BR': '191.128.0.0/12',
4170 'BS': '24.51.64.0/18',
4171 'BT': '119.2.96.0/19',
4172 'BW': '168.167.0.0/16',
4173 'BY': '178.120.0.0/13',
4174 'BZ': '179.42.192.0/18',
4175 'CA': '99.224.0.0/11',
4176 'CD': '41.243.0.0/16',
4177 'CF': '197.242.176.0/21',
4178 'CG': '160.113.0.0/16',
4179 'CH': '85.0.0.0/13',
4180 'CI': '102.136.0.0/14',
4181 'CK': '202.65.32.0/19',
4182 'CL': '152.172.0.0/14',
4183 'CM': '102.244.0.0/14',
4184 'CN': '36.128.0.0/10',
4185 'CO': '181.240.0.0/12',
4186 'CR': '201.192.0.0/12',
4187 'CU': '152.206.0.0/15',
4188 'CV': '165.90.96.0/19',
4189 'CW': '190.88.128.0/17',
4190 'CY': '31.153.0.0/16',
4191 'CZ': '88.100.0.0/14',
4193 'DJ': '197.241.0.0/17',
4194 'DK': '87.48.0.0/12',
4195 'DM': '192.243.48.0/20',
4196 'DO': '152.166.0.0/15',
4197 'DZ': '41.96.0.0/12',
4198 'EC': '186.68.0.0/15',
4199 'EE': '90.190.0.0/15',
4200 'EG': '156.160.0.0/11',
4201 'ER': '196.200.96.0/20',
4202 'ES': '88.0.0.0/11',
4203 'ET': '196.188.0.0/14',
4204 'EU': '2.16.0.0/13',
4205 'FI': '91.152.0.0/13',
4206 'FJ': '144.120.0.0/16',
4207 'FK': '80.73.208.0/21',
4208 'FM': '119.252.112.0/20',
4209 'FO': '88.85.32.0/19',
4211 'GA': '41.158.0.0/15',
4213 'GD': '74.122.88.0/21',
4214 'GE': '31.146.0.0/16',
4215 'GF': '161.22.64.0/18',
4216 'GG': '62.68.160.0/19',
4217 'GH': '154.160.0.0/12',
4218 'GI': '95.164.0.0/16',
4219 'GL': '88.83.0.0/19',
4220 'GM': '160.182.0.0/15',
4221 'GN': '197.149.192.0/18',
4222 'GP': '104.250.0.0/19',
4223 'GQ': '105.235.224.0/20',
4224 'GR': '94.64.0.0/13',
4225 'GT': '168.234.0.0/16',
4226 'GU': '168.123.0.0/16',
4227 'GW': '197.214.80.0/20',
4228 'GY': '181.41.64.0/18',
4229 'HK': '113.252.0.0/14',
4230 'HN': '181.210.0.0/16',
4231 'HR': '93.136.0.0/13',
4232 'HT': '148.102.128.0/17',
4233 'HU': '84.0.0.0/14',
4234 'ID': '39.192.0.0/10',
4235 'IE': '87.32.0.0/12',
4236 'IL': '79.176.0.0/13',
4237 'IM': '5.62.80.0/20',
4238 'IN': '117.192.0.0/10',
4239 'IO': '203.83.48.0/21',
4240 'IQ': '37.236.0.0/14',
4241 'IR': '2.176.0.0/12',
4242 'IS': '82.221.0.0/16',
4243 'IT': '79.0.0.0/10',
4244 'JE': '87.244.64.0/18',
4245 'JM': '72.27.0.0/17',
4246 'JO': '176.29.0.0/16',
4247 'JP': '133.0.0.0/8',
4248 'KE': '105.48.0.0/12',
4249 'KG': '158.181.128.0/17',
4250 'KH': '36.37.128.0/17',
4251 'KI': '103.25.140.0/22',
4252 'KM': '197.255.224.0/20',
4253 'KN': '198.167.192.0/19',
4254 'KP': '175.45.176.0/22',
4255 'KR': '175.192.0.0/10',
4256 'KW': '37.36.0.0/14',
4257 'KY': '64.96.0.0/15',
4258 'KZ': '2.72.0.0/13',
4259 'LA': '115.84.64.0/18',
4260 'LB': '178.135.0.0/16',
4261 'LC': '24.92.144.0/20',
4262 'LI': '82.117.0.0/19',
4263 'LK': '112.134.0.0/15',
4264 'LR': '102.183.0.0/16',
4265 'LS': '129.232.0.0/17',
4266 'LT': '78.56.0.0/13',
4267 'LU': '188.42.0.0/16',
4268 'LV': '46.109.0.0/16',
4269 'LY': '41.252.0.0/14',
4270 'MA': '105.128.0.0/11',
4271 'MC': '88.209.64.0/18',
4272 'MD': '37.246.0.0/16',
4273 'ME': '178.175.0.0/17',
4274 'MF': '74.112.232.0/21',
4275 'MG': '154.126.0.0/17',
4276 'MH': '117.103.88.0/21',
4277 'MK': '77.28.0.0/15',
4278 'ML': '154.118.128.0/18',
4279 'MM': '37.111.0.0/17',
4280 'MN': '49.0.128.0/17',
4281 'MO': '60.246.0.0/16',
4282 'MP': '202.88.64.0/20',
4283 'MQ': '109.203.224.0/19',
4284 'MR': '41.188.64.0/18',
4285 'MS': '208.90.112.0/22',
4286 'MT': '46.11.0.0/16',
4287 'MU': '105.16.0.0/12',
4288 'MV': '27.114.128.0/18',
4289 'MW': '102.70.0.0/15',
4290 'MX': '187.192.0.0/11',
4291 'MY': '175.136.0.0/13',
4292 'MZ': '197.218.0.0/15',
4293 'NA': '41.182.0.0/16',
4294 'NC': '101.101.0.0/18',
4295 'NE': '197.214.0.0/18',
4296 'NF': '203.17.240.0/22',
4297 'NG': '105.112.0.0/12',
4298 'NI': '186.76.0.0/15',
4299 'NL': '145.96.0.0/11',
4300 'NO': '84.208.0.0/13',
4301 'NP': '36.252.0.0/15',
4302 'NR': '203.98.224.0/19',
4303 'NU': '49.156.48.0/22',
4304 'NZ': '49.224.0.0/14',
4305 'OM': '5.36.0.0/15',
4306 'PA': '186.72.0.0/15',
4307 'PE': '186.160.0.0/14',
4308 'PF': '123.50.64.0/18',
4309 'PG': '124.240.192.0/19',
4310 'PH': '49.144.0.0/13',
4311 'PK': '39.32.0.0/11',
4312 'PL': '83.0.0.0/11',
4313 'PM': '70.36.0.0/20',
4314 'PR': '66.50.0.0/16',
4315 'PS': '188.161.0.0/16',
4316 'PT': '85.240.0.0/13',
4317 'PW': '202.124.224.0/20',
4318 'PY': '181.120.0.0/14',
4319 'QA': '37.210.0.0/15',
4320 'RE': '102.35.0.0/16',
4321 'RO': '79.112.0.0/13',
4322 'RS': '93.86.0.0/15',
4323 'RU': '5.136.0.0/13',
4324 'RW': '41.186.0.0/16',
4325 'SA': '188.48.0.0/13',
4326 'SB': '202.1.160.0/19',
4327 'SC': '154.192.0.0/11',
4328 'SD': '102.120.0.0/13',
4329 'SE': '78.64.0.0/12',
4330 'SG': '8.128.0.0/10',
4331 'SI': '188.196.0.0/14',
4332 'SK': '78.98.0.0/15',
4333 'SL': '102.143.0.0/17',
4334 'SM': '89.186.32.0/19',
4335 'SN': '41.82.0.0/15',
4336 'SO': '154.115.192.0/18',
4337 'SR': '186.179.128.0/17',
4338 'SS': '105.235.208.0/21',
4339 'ST': '197.159.160.0/19',
4340 'SV': '168.243.0.0/16',
4341 'SX': '190.102.0.0/20',
4343 'SZ': '41.84.224.0/19',
4344 'TC': '65.255.48.0/20',
4345 'TD': '154.68.128.0/19',
4346 'TG': '196.168.0.0/14',
4347 'TH': '171.96.0.0/13',
4348 'TJ': '85.9.128.0/18',
4349 'TK': '27.96.24.0/21',
4350 'TL': '180.189.160.0/20',
4351 'TM': '95.85.96.0/19',
4352 'TN': '197.0.0.0/11',
4353 'TO': '175.176.144.0/21',
4354 'TR': '78.160.0.0/11',
4355 'TT': '186.44.0.0/15',
4356 'TV': '202.2.96.0/19',
4357 'TW': '120.96.0.0/11',
4358 'TZ': '156.156.0.0/14',
4359 'UA': '37.52.0.0/14',
4360 'UG': '102.80.0.0/13',
4362 'UY': '167.56.0.0/13',
4363 'UZ': '84.54.64.0/18',
4364 'VA': '212.77.0.0/19',
4365 'VC': '207.191.240.0/21',
4366 'VE': '186.88.0.0/13',
4367 'VG': '66.81.192.0/20',
4368 'VI': '146.226.0.0/16',
4369 'VN': '14.160.0.0/11',
4370 'VU': '202.80.32.0/20',
4371 'WF': '117.20.32.0/21',
4372 'WS': '202.4.32.0/19',
4373 'YE': '134.35.0.0/16',
4374 'YT': '41.242.116.0/22',
4375 'ZA': '41.0.0.0/11',
4376 'ZM': '102.144.0.0/13',
4377 'ZW': '102.177.192.0/18',
4381 def random_ipv4(cls, code_or_block):
4382 if len(code_or_block) == 2:
4383 block = cls._country_ip_map.get(code_or_block.upper())
4387 block = code_or_block
4388 addr, preflen = block.split('/')
4389 addr_min = compat_struct_unpack('!L', socket.inet_aton(addr))[0]
4390 addr_max = addr_min | (0xffffffff >> int(preflen))
4391 return compat_str(socket.inet_ntoa(
4392 compat_struct_pack('!L', random.randint(addr_min, addr_max))))
4395 class PerRequestProxyHandler(compat_urllib_request.ProxyHandler):
4396 def __init__(self, proxies=None):
4397 # Set default handlers
4398 for type in ('http', 'https'):
4399 setattr(self, '%s_open' % type,
4400 lambda r, proxy='__noproxy__', type=type, meth=self.proxy_open:
4401 meth(r, proxy, type))
4402 compat_urllib_request.ProxyHandler.__init__(self, proxies)
4404 def proxy_open(self, req, proxy, type):
4405 req_proxy = req.headers.get('Ytdl-request-proxy')
4406 if req_proxy is not None:
4408 del req.headers['Ytdl-request-proxy']
4410 if proxy == '__noproxy__':
4411 return None # No Proxy
4412 if compat_urlparse.urlparse(proxy).scheme.lower() in ('socks', 'socks4', 'socks4a', 'socks5'):
4413 req.add_header('Ytdl-socks-proxy', proxy)
4414 # yt-dlp's http/https handlers do wrapping the socket with socks
4416 return compat_urllib_request.ProxyHandler.proxy_open(
4417 self, req, proxy, type)
4420 # Both long_to_bytes and bytes_to_long are adapted from PyCrypto, which is
4421 # released into Public Domain
4422 # https://github.com/dlitz/pycrypto/blob/master/lib/Crypto/Util/number.py#L387
4424 def long_to_bytes(n, blocksize=0):
4425 """long_to_bytes(n:long, blocksize:int) : string
4426 Convert a long integer to a byte string.
4428 If optional blocksize is given and greater than zero, pad the front of the
4429 byte string with binary zeros so that the length is a multiple of
4432 # after much testing, this algorithm was deemed to be the fastest
4436 s = compat_struct_pack('>I', n & 0xffffffff) + s
4438 # strip off leading zeros
4439 for i in range(len(s)):
4440 if s[i] != b'\000'[0]:
4443 # only happens when n == 0
4447 # add back some pad bytes. this could be done more efficiently w.r.t. the
4448 # de-padding being done above, but sigh...
4449 if blocksize > 0 and len(s) % blocksize:
4450 s = (blocksize - len(s) % blocksize) * b'\000' + s
4454 def bytes_to_long(s):
4455 """bytes_to_long(string) : long
4456 Convert a byte string to a long integer.
4458 This is (essentially) the inverse of long_to_bytes().
4463 extra = (4 - length % 4)
4464 s = b'\000' * extra + s
4465 length = length + extra
4466 for i in range(0, length, 4):
4467 acc = (acc << 32) + compat_struct_unpack('>I', s[i:i + 4])[0]
4471 def ohdave_rsa_encrypt(data, exponent, modulus):
4473 Implement OHDave
's RSA algorithm. See http://www.ohdave.com/rsa/
4476 data: data to encrypt, bytes-like object
4477 exponent, modulus: parameter e and N of RSA algorithm, both integer
4478 Output: hex string of encrypted data
4480 Limitation: supports one block encryption only
4483 payload = int(binascii.hexlify(data[::-1]), 16)
4484 encrypted = pow(payload, exponent, modulus)
4485 return '%x' % encrypted
4488 def pkcs1pad(data, length):
4490 Padding input data with PKCS#1 scheme
4492 @param {int[]} data input data
4493 @param {int} length target length
4494 @returns {int[]} padded data
4496 if len(data) > length - 11:
4497 raise ValueError('Input data too
long for PKCS
#1 padding')
4499 pseudo_random
= [random
.randint(0, 254) for _
in range(length
- len(data
) - 3)]
4500 return [0, 2] + pseudo_random
+ [0] + data
4503 def encode_base_n(num
, n
, table
=None):
4504 FULL_TABLE
= '0123456789abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ'
4506 table
= FULL_TABLE
[:n
]
4509 raise ValueError('base %d exceeds table length %d' % (n
, len(table
)))
4516 ret
= table
[num
% n
] + ret
4521 def decode_packed_codes(code
):
4522 mobj
= re
.search(PACKED_CODES_RE
, code
)
4523 obfuscated_code
, base
, count
, symbols
= mobj
.groups()
4526 symbols
= symbols
.split('|')
4531 base_n_count
= encode_base_n(count
, base
)
4532 symbol_table
[base_n_count
] = symbols
[count
] or base_n_count
4535 r
'\b(\w+)\b', lambda mobj
: symbol_table
[mobj
.group(0)],
4539 def caesar(s
, alphabet
, shift
):
4544 alphabet
[(alphabet
.index(c
) + shift
) % l
] if c
in alphabet
else c
4549 return caesar(s
, r
'''!"#$%&'()*+,-./0123456789:;<=>?@ABCDEFGHIJKLMNOPQRSTUVWXYZ[\]^_`abcdefghijklmnopqrstuvwxyz{|}~''', 47)
4552 def parse_m3u8_attributes(attrib
):
4554 for (key
, val
) in re
.findall(r
'(?P<key>[A-Z0-9-]+)=(?P<val>"[^"]+"|[^",]+)(?:,|$)', attrib
):
4555 if val
.startswith('"'):
4561 def urshift(val
, n
):
4562 return val
>> n
if val
>= 0 else (val
+ 0x100000000) >> n
4565 # Based on png2str() written by @gdkchan and improved by @yokrysty
4566 # Originally posted at https://github.com/ytdl-org/youtube-dl/issues/9706
4567 def decode_png(png_data
):
4568 # Reference: https://www.w3.org/TR/PNG/
4569 header
= png_data
[8:]
4571 if png_data
[:8] != b
'\x89PNG\x0d\x0a\x1a\x0a' or header
[4:8] != b
'IHDR':
4572 raise IOError('Not a valid PNG file.')
4574 int_map
= {1: '>B', 2: '>H', 4: '>I'}
4575 unpack_integer
= lambda x
: compat_struct_unpack(int_map
[len(x
)], x
)[0]
4580 length
= unpack_integer(header
[:4])
4583 chunk_type
= header
[:4]
4586 chunk_data
= header
[:length
]
4587 header
= header
[length
:]
4589 header
= header
[4:] # Skip CRC
4597 ihdr
= chunks
[0]['data']
4599 width
= unpack_integer(ihdr
[:4])
4600 height
= unpack_integer(ihdr
[4:8])
4604 for chunk
in chunks
:
4605 if chunk
['type'] == b
'IDAT':
4606 idat
+= chunk
['data']
4609 raise IOError('Unable to read PNG data.')
4611 decompressed_data
= bytearray(zlib
.decompress(idat
))
4616 def _get_pixel(idx
):
4621 for y
in range(height
):
4622 basePos
= y
* (1 + stride
)
4623 filter_type
= decompressed_data
[basePos
]
4627 pixels
.append(current_row
)
4629 for x
in range(stride
):
4630 color
= decompressed_data
[1 + basePos
+ x
]
4631 basex
= y
* stride
+ x
4636 left
= _get_pixel(basex
- 3)
4638 up
= _get_pixel(basex
- stride
)
4640 if filter_type
== 1: # Sub
4641 color
= (color
+ left
) & 0xff
4642 elif filter_type
== 2: # Up
4643 color
= (color
+ up
) & 0xff
4644 elif filter_type
== 3: # Average
4645 color
= (color
+ ((left
+ up
) >> 1)) & 0xff
4646 elif filter_type
== 4: # Paeth
4652 c
= _get_pixel(basex
- stride
- 3)
4660 if pa
<= pb
and pa
<= pc
:
4661 color
= (color
+ a
) & 0xff
4663 color
= (color
+ b
) & 0xff
4665 color
= (color
+ c
) & 0xff
4667 current_row
.append(color
)
4669 return width
, height
, pixels
4672 def write_xattr(path
, key
, value
):
4673 # This mess below finds the best xattr tool for the job
4675 # try the pyxattr module...
4678 if hasattr(xattr
, 'set'): # pyxattr
4679 # Unicode arguments are not supported in python-pyxattr until
4681 # See https://github.com/ytdl-org/youtube-dl/issues/5498
4682 pyxattr_required_version
= '0.5.0'
4683 if version_tuple(xattr
.__version
__) < version_tuple(pyxattr_required_version
):
4684 # TODO: fallback to CLI tools
4685 raise XAttrUnavailableError(
4686 'python-pyxattr is detected but is too old. '
4687 'yt-dlp requires %s or above while your version is %s. '
4688 'Falling back to other xattr implementations' % (
4689 pyxattr_required_version
, xattr
.__version
__))
4691 setxattr
= xattr
.set
4693 setxattr
= xattr
.setxattr
4696 setxattr(path
, key
, value
)
4697 except EnvironmentError as e
:
4698 raise XAttrMetadataError(e
.errno
, e
.strerror
)
4701 if compat_os_name
== 'nt':
4702 # Write xattrs to NTFS Alternate Data Streams:
4703 # http://en.wikipedia.org/wiki/NTFS#Alternate_data_streams_.28ADS.29
4704 assert ':' not in key
4705 assert os
.path
.exists(path
)
4707 ads_fn
= path
+ ':' + key
4709 with open(ads_fn
, 'wb') as f
:
4711 except EnvironmentError as e
:
4712 raise XAttrMetadataError(e
.errno
, e
.strerror
)
4714 user_has_setfattr
= check_executable('setfattr', ['--version'])
4715 user_has_xattr
= check_executable('xattr', ['-h'])
4717 if user_has_setfattr
or user_has_xattr
:
4719 value
= value
.decode('utf-8')
4720 if user_has_setfattr
:
4721 executable
= 'setfattr'
4722 opts
= ['-n', key
, '-v', value
]
4723 elif user_has_xattr
:
4724 executable
= 'xattr'
4725 opts
= ['-w', key
, value
]
4727 cmd
= ([encodeFilename(executable
, True)]
4728 + [encodeArgument(o
) for o
in opts
]
4729 + [encodeFilename(path
, True)])
4733 cmd
, stdout
=subprocess
.PIPE
, stderr
=subprocess
.PIPE
, stdin
=subprocess
.PIPE
)
4734 except EnvironmentError as e
:
4735 raise XAttrMetadataError(e
.errno
, e
.strerror
)
4736 stdout
, stderr
= p
.communicate_or_kill()
4737 stderr
= stderr
.decode('utf-8', 'replace')
4738 if p
.returncode
!= 0:
4739 raise XAttrMetadataError(p
.returncode
, stderr
)
4742 # On Unix, and can't find pyxattr, setfattr, or xattr.
4743 if sys
.platform
.startswith('linux'):
4744 raise XAttrUnavailableError(
4745 "Couldn't find a tool to set the xattrs. "
4746 "Install either the python 'pyxattr' or 'xattr' "
4747 "modules, or the GNU 'attr' package "
4748 "(which contains the 'setfattr' tool).")
4750 raise XAttrUnavailableError(
4751 "Couldn't find a tool to set the xattrs. "
4752 "Install either the python 'xattr' module, "
4753 "or the 'xattr' binary.")
4756 def random_birthday(year_field
, month_field
, day_field
):
4757 start_date
= datetime
.date(1950, 1, 1)
4758 end_date
= datetime
.date(1995, 12, 31)
4759 offset
= random
.randint(0, (end_date
- start_date
).days
)
4760 random_date
= start_date
+ datetime
.timedelta(offset
)
4762 year_field
: str(random_date
.year
),
4763 month_field
: str(random_date
.month
),
4764 day_field
: str(random_date
.day
),
4768 # Templates for internet shortcut files, which are plain text files.
4769 DOT_URL_LINK_TEMPLATE
= '''
4774 DOT_WEBLOC_LINK_TEMPLATE
= '''
4775 <?xml version="1.0" encoding="UTF-8"?>
4776 <!DOCTYPE plist PUBLIC "-//Apple//DTD PLIST 1.0//EN" "http://www.apple.com/DTDs/PropertyList-1.0.dtd">
4777 <plist version="1.0">
4780 \t<string>%(url)s</string>
4785 DOT_DESKTOP_LINK_TEMPLATE
= '''
4795 'url': DOT_URL_LINK_TEMPLATE
,
4796 'desktop': DOT_DESKTOP_LINK_TEMPLATE
,
4797 'webloc': DOT_WEBLOC_LINK_TEMPLATE
,
4801 def iri_to_uri(iri
):
4803 Converts an IRI (Internationalized Resource Identifier, allowing Unicode characters) to a URI (Uniform Resource Identifier, ASCII-only).
4805 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.
4808 iri_parts
= compat_urllib_parse_urlparse(iri
)
4810 if '[' in iri_parts
.netloc
:
4811 raise ValueError('IPv6 URIs are not, yet, supported.')
4812 # Querying `.netloc`, when there's only one bracket, also raises a ValueError.
4814 # 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.
4817 if iri_parts
.username
:
4818 net_location
+= compat_urllib_parse_quote(iri_parts
.username
, safe
=r
"!$%&'()*+,~")
4819 if iri_parts
.password
is not None:
4820 net_location
+= ':' + compat_urllib_parse_quote(iri_parts
.password
, safe
=r
"!$%&'()*+,~")
4823 net_location
+= iri_parts
.hostname
.encode('idna').decode('utf-8') # Punycode for Unicode hostnames.
4824 # The 'idna' encoding produces ASCII text.
4825 if iri_parts
.port
is not None and iri_parts
.port
!= 80:
4826 net_location
+= ':' + str(iri_parts
.port
)
4828 return compat_urllib_parse_urlunparse(
4832 compat_urllib_parse_quote_plus(iri_parts
.path
, safe
=r
"!$%&'()*+,/:;=@|~"),
4834 # Unsure about the `safe` argument, since this is a legacy way of handling parameters.
4835 compat_urllib_parse_quote_plus(iri_parts
.params
, safe
=r
"!$%&'()*+,/:;=@|~"),
4837 # Not totally sure about the `safe` argument, since the source does not explicitly mention the query URI component.
4838 compat_urllib_parse_quote_plus(iri_parts
.query
, safe
=r
"!$%&'()*+,/:;=?@{|}~"),
4840 compat_urllib_parse_quote_plus(iri_parts
.fragment
, safe
=r
"!#$%&'()*+,/:;=?@{|}~")))
4842 # Source for `safe` arguments: https://url.spec.whatwg.org/#percent-encoded-bytes.
4845 def to_high_limit_path(path
):
4846 if sys
.platform
in ['win32', 'cygwin']:
4847 # Work around MAX_PATH limitation on Windows. The maximum allowed length for the individual path segments may still be quite limited.
4848 return r
'\\?\ '.rstrip() + os
.path
.abspath(path
)
4853 def format_field(obj
, field
=None, template
='%s', ignore
=(None, ''), default
='', func
=None):
4855 val
= obj
if obj
is not None else default
4857 val
= obj
.get(field
, default
)
4858 if func
and val
not in ignore
:
4860 return template
% val
if val
not in ignore
else default
4863 def clean_podcast_url(url
):
4864 return re
.sub(r
'''(?x)
4868 media\.blubrry\.com| # https://create.blubrry.com/resources/podcast-media-download-statistics/getting-started/
4871 (?:dts|www)\.podtrac\.com/(?:pts/)?redirect\.[0-9a-z]{3,4}| # http://analytics.podtrac.com/how-to-measure
4874 cn\.co| # https://podcorn.com/analytics-prefix/
4875 st\.fm # https://podsights.com/docs/
4880 _HEX_TABLE
= '0123456789abcdef'
4883 def random_uuidv4():
4884 return re
.sub(r
'[xy]', lambda x
: _HEX_TABLE
[random
.randint(0, 15)], 'xxxxxxxx-xxxx-4xxx-yxxx-xxxxxxxxxxxx')
4887 def make_dir(path
, to_screen
=None):
4889 dn
= os
.path
.dirname(path
)
4890 if dn
and not os
.path
.exists(dn
):
4893 except (OSError, IOError) as err
:
4894 if callable(to_screen
) is not None:
4895 to_screen('unable to create directory ' + error_to_compat_str(err
))
4899 def get_executable_path():
4900 from zipimport
import zipimporter
4901 if hasattr(sys
, 'frozen'): # Running from PyInstaller
4902 path
= os
.path
.dirname(sys
.executable
)
4903 elif isinstance(globals().get('__loader__'), zipimporter
): # Running from ZIP
4904 path
= os
.path
.join(os
.path
.dirname(__file__
), '../..')
4906 path
= os
.path
.join(os
.path
.dirname(__file__
), '..')
4907 return os
.path
.abspath(path
)
4910 def load_plugins(name
, suffix
, namespace
):
4913 plugins_spec
= importlib
.util
.spec_from_file_location(
4914 name
, os
.path
.join(get_executable_path(), 'ytdlp_plugins', name
, '__init__.py'))
4915 plugins
= importlib
.util
.module_from_spec(plugins_spec
)
4916 sys
.modules
[plugins_spec
.name
] = plugins
4917 plugins_spec
.loader
.exec_module(plugins
)
4918 for name
in dir(plugins
):
4919 if name
in namespace
:
4921 if not name
.endswith(suffix
):
4923 klass
= getattr(plugins
, name
)
4924 classes
[name
] = namespace
[name
] = klass
4925 except FileNotFoundError
:
4931 obj
, *path_list
, default
=None, expected_type
=None, get_all
=True,
4932 casesense
=True, is_user_input
=False, traverse_string
=False):
4933 ''' Traverse nested list/dict/tuple
4934 @param path_list A list of paths which are checked one by one.
4935 Each path is a list of keys where each key is a string,
4936 a function, a tuple of strings or "...".
4937 When a fuction is given, it takes the key as argument and
4938 returns whether the key matches or not. When a tuple is given,
4939 all the keys given in the tuple are traversed, and
4940 "..." traverses all the keys in the object
4941 @param default Default value to return
4942 @param expected_type Only accept final value of this type (Can also be any callable)
4943 @param get_all Return all the values obtained from a path or only the first one
4944 @param casesense Whether to consider dictionary keys as case sensitive
4945 @param is_user_input Whether the keys are generated from user input. If True,
4946 strings are converted to int/slice if necessary
4947 @param traverse_string Whether to traverse inside strings. If True, any
4948 non-compatible object will also be converted into a string
4952 _lower
= lambda k
: (k
.lower() if isinstance(k
, str) else k
)
4953 path_list
= (map(_lower
, variadic(path
)) for path
in path_list
)
4955 def _traverse_obj(obj
, path
, _current_depth
=0):
4957 path
= tuple(variadic(path
))
4958 for i
, key
in enumerate(path
):
4961 if isinstance(key
, (list, tuple)):
4962 obj
= [_traverse_obj(obj
, sub_key
, _current_depth
) for sub_key
in key
]
4965 obj
= (obj
.values() if isinstance(obj
, dict)
4966 else obj
if isinstance(obj
, (list, tuple, LazyList
))
4967 else str(obj
) if traverse_string
else [])
4969 depth
= max(depth
, _current_depth
)
4970 return [_traverse_obj(inner_obj
, path
[i
+ 1:], _current_depth
) for inner_obj
in obj
]
4972 if isinstance(obj
, (list, tuple, LazyList
)):
4973 obj
= enumerate(obj
)
4974 elif isinstance(obj
, dict):
4977 if not traverse_string
:
4981 depth
= max(depth
, _current_depth
)
4982 return [_traverse_obj(v
, path
[i
+ 1:], _current_depth
) for k
, v
in obj
if key(k
)]
4983 elif isinstance(obj
, dict) and not (is_user_input
and key
== ':'):
4984 obj
= (obj
.get(key
) if casesense
or (key
in obj
)
4985 else next((v
for k
, v
in obj
.items() if _lower(k
) == key
), None))
4988 key
= (int_or_none(key
) if ':' not in key
4989 else slice(*map(int_or_none
, key
.split(':'))))
4990 if key
== slice(None):
4991 return _traverse_obj(obj
, (..., *path
[i
+ 1:]), _current_depth
)
4992 if not isinstance(key
, (int, slice)):
4994 if not isinstance(obj
, (list, tuple, LazyList
)):
4995 if not traverse_string
:
5004 if isinstance(expected_type
, type):
5005 type_test
= lambda val
: val
if isinstance(val
, expected_type
) else None
5006 elif expected_type
is not None:
5007 type_test
= expected_type
5009 type_test
= lambda val
: val
5011 for path
in path_list
:
5013 val
= _traverse_obj(obj
, path
)
5016 for _
in range(depth
- 1):
5017 val
= itertools
.chain
.from_iterable(v
for v
in val
if v
is not None)
5018 val
= [v
for v
in map(type_test
, val
) if v
is not None]
5020 return val
if get_all
else val
[0]
5022 val
= type_test(val
)
5029 def traverse_dict(dictn
, keys
, casesense
=True):
5030 write_string('DeprecationWarning: yt_dlp.utils.traverse_dict is deprecated '
5031 'and may be removed in a future version. Use yt_dlp.utils.traverse_obj instead')
5032 return traverse_obj(dictn
, keys
, casesense
=casesense
, is_user_input
=True, traverse_string
=True)
5035 def variadic(x
, allowed_types
=(str, bytes, dict)):
5036 return x
if isinstance(x
, collections
.abc
.Iterable
) and not isinstance(x
, allowed_types
) else (x
,)
5039 # create a JSON Web Signature (jws) with HS256 algorithm
5040 # the resulting format is in JWS Compact Serialization
5041 # implemented following JWT https://www.rfc-editor.org/rfc/rfc7519.html
5042 # implemented following JWS https://www.rfc-editor.org/rfc/rfc7515.html
5043 def jwt_encode_hs256(payload_data
, key
, headers
={}):
5049 header_data
.update(headers
)
5050 header_b64
= base64
.b64encode(json
.dumps(header_data
).encode('utf-8'))
5051 payload_b64
= base64
.b64encode(json
.dumps(payload_data
).encode('utf-8'))
5052 h
= hmac
.new(key
.encode('utf-8'), header_b64
+ b
'.' + payload_b64
, hashlib
.sha256
)
5053 signature_b64
= base64
.b64encode(h
.digest())
5054 token
= header_b64
+ b
'.' + payload_b64
+ b
'.' + signature_b64
5058 # can be extended in future to verify the signature and parse header and return the algorithm used if it's not HS256
5059 def jwt_decode_hs256(jwt
):
5060 header_b64
, payload_b64
, signature_b64
= jwt
.split('.')
5061 payload_data
= json
.loads(base64
.urlsafe_b64decode(payload_b64
))
5065 def supports_terminal_sequences(stream
):
5066 if compat_os_name
== 'nt':
5067 from .compat
import WINDOWS_VT_MODE
# Must be imported locally
5068 if not WINDOWS_VT_MODE
or get_windows_version() < (10, 0, 10586):
5070 elif not os
.getenv('TERM'):
5073 return stream
.isatty()
5074 except BaseException
:
5078 _terminal_sequences_re
= re
.compile('\033\\[[^m]+m')
5081 def remove_terminal_sequences(string
):
5082 return _terminal_sequences_re
.sub('', string
)
5085 def number_of_digits(number
):
5086 return len('%d' % number
)
5089 def join_nonempty(*values
, delim
='-', from_dict
=None):
5090 if from_dict
is not None:
5091 values
= map(from_dict
.get
, values
)
5092 return delim
.join(map(str, filter(None, values
)))