46 import xml
.etree
.ElementTree
49 from .compat
import asyncio
, functools
# isort: split
51 compat_etree_fromstring
,
53 compat_HTMLParseError
,
57 from .dependencies
import brotli
, certifi
, websockets
, xattr
58 from .socks
import ProxyType
, sockssocket
61 def register_socks_protocols():
62 # "Register" SOCKS protocols
63 # In Python < 2.6.5, urlsplit() suffers from bug https://bugs.python.org/issue7904
64 # URLs with protocols not in urlparse.uses_netloc are not handled correctly
65 for scheme
in ('socks', 'socks4', 'socks4a', 'socks5'):
66 if scheme
not in urllib
.parse
.uses_netloc
:
67 urllib
.parse
.uses_netloc
.append(scheme
)
70 # This is not clearly defined otherwise
71 compiled_regex_type
= type(re
.compile(''))
74 def random_user_agent():
75 _USER_AGENT_TPL
= 'Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/%s Safari/537.36'
116 return _USER_AGENT_TPL
% random
.choice(_CHROME_VERSIONS
)
119 SUPPORTED_ENCODINGS
= [
123 SUPPORTED_ENCODINGS
.append('br')
126 'User-Agent': random_user_agent(),
127 'Accept': 'text/html,application/xhtml+xml,application/xml;q=0.9,*/*;q=0.8',
128 'Accept-Language': 'en-us,en;q=0.5',
129 'Sec-Fetch-Mode': 'navigate',
134 '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',
138 NO_DEFAULT
= object()
139 IDENTITY
= lambda x
: x
141 ENGLISH_MONTH_NAMES
= [
142 'January', 'February', 'March', 'April', 'May', 'June',
143 'July', 'August', 'September', 'October', 'November', 'December']
146 'en': ENGLISH_MONTH_NAMES
,
148 'janvier', 'février', 'mars', 'avril', 'mai', 'juin',
149 'juillet', 'août', 'septembre', 'octobre', 'novembre', 'décembre'],
153 'mp4', 'm4a', 'm4p', 'm4b', 'm4r', 'm4v', 'aac',
154 'flv', 'f4v', 'f4a', 'f4b',
155 'webm', 'ogg', 'ogv', 'oga', 'ogx', 'spx', 'opus',
156 'mkv', 'mka', 'mk3d',
165 'f4f', 'f4m', 'm3u8', 'smil')
167 # needed for sanitizing filenames in restricted mode
168 ACCENT_CHARS
= dict(zip('ÂÃÄÀÁÅÆÇÈÉÊËÌÍÎÏÐÑÒÓÔÕÖŐØŒÙÚÛÜŰÝÞßàáâãäåæçèéêëìíîïðñòóôõöőøœùúûüűýþÿ',
169 itertools
.chain('AAAAAA', ['AE'], 'CEEEEIIIIDNOOOOOOO', ['OE'], 'UUUUUY', ['TH', 'ss'],
170 'aaaaaa', ['ae'], 'ceeeeiiiionooooooo', ['oe'], 'uuuuuy', ['th'], 'y')))
200 '%Y-%m-%d %H:%M:%S.%f',
201 '%Y-%m-%d %H:%M:%S:%f',
204 '%Y-%m-%dT%H:%M:%SZ',
205 '%Y-%m-%dT%H:%M:%S.%fZ',
206 '%Y-%m-%dT%H:%M:%S.%f0Z',
208 '%Y-%m-%dT%H:%M:%S.%f',
211 '%b %d %Y at %H:%M:%S',
213 '%B %d %Y at %H:%M:%S',
217 DATE_FORMATS_DAY_FIRST
= list(DATE_FORMATS
)
218 DATE_FORMATS_DAY_FIRST
.extend([
227 DATE_FORMATS_MONTH_FIRST
= list(DATE_FORMATS
)
228 DATE_FORMATS_MONTH_FIRST
.extend([
236 PACKED_CODES_RE
= r
"}\('(.+)',(\d+),(\d+),'([^']+)'\.split\('\|'\)"
237 JSON_LD_RE
= r
'(?is)<script[^>]+type=(["\']?
)application
/ld\
+json\
1[^
>]*>\s
*(?P
<json_ld
>{.+?}
)\s
*</script
>'
239 NUMBER_RE = r'\d
+(?
:\
.\d
+)?
'
243 def preferredencoding():
244 """Get preferred encoding.
246 Returns the best encoding scheme for the system, based on
247 locale.getpreferredencoding() and some further tweaks.
250 pref = locale.getpreferredencoding()
258 def write_json_file(obj, fn):
259 """ Encode obj as JSON and write it to fn, atomically if possible """
261 tf = tempfile.NamedTemporaryFile(
262 prefix=f'{os.path.basename(fn)}
.', dir=os.path.dirname(fn),
263 suffix='.tmp
', delete=False, mode='w
', encoding='utf
-8')
267 json.dump(obj, tf, ensure_ascii=False)
268 if sys.platform == 'win32
':
269 # Need to remove existing file on Windows, else os.rename raises
270 # WindowsError or FileExistsError.
271 with contextlib.suppress(OSError):
273 with contextlib.suppress(OSError):
276 os.chmod(tf.name, 0o666 & ~mask)
277 os.rename(tf.name, fn)
279 with contextlib.suppress(OSError):
284 def find_xpath_attr(node, xpath, key, val=None):
285 """ Find the xpath xpath[@key=val] """
286 assert re.match(r'^
[a
-zA
-Z_
-]+$
', key)
287 expr = xpath + ('[@%s]' % key if val is None else f"[@{key}='{val}
']")
288 return node.find(expr)
290 # On python2.6 the xml.etree.ElementTree.Element methods don't support
291 # the namespace parameter
294 def xpath_with_ns(path
, ns_map
):
295 components
= [c
.split(':') for c
in path
.split('/')]
299 replaced
.append(c
[0])
302 replaced
.append('{%s}%s' % (ns_map
[ns
], tag
))
303 return '/'.join(replaced
)
306 def xpath_element(node
, xpath
, name
=None, fatal
=False, default
=NO_DEFAULT
):
307 def _find_xpath(xpath
):
308 return node
.find(xpath
)
310 if isinstance(xpath
, str):
311 n
= _find_xpath(xpath
)
319 if default
is not NO_DEFAULT
:
322 name
= xpath
if name
is None else name
323 raise ExtractorError('Could not find XML element %s' % name
)
329 def xpath_text(node
, xpath
, name
=None, fatal
=False, default
=NO_DEFAULT
):
330 n
= xpath_element(node
, xpath
, name
, fatal
=fatal
, default
=default
)
331 if n
is None or n
== default
:
334 if default
is not NO_DEFAULT
:
337 name
= xpath
if name
is None else name
338 raise ExtractorError('Could not find XML element\'s text %s' % name
)
344 def xpath_attr(node
, xpath
, key
, name
=None, fatal
=False, default
=NO_DEFAULT
):
345 n
= find_xpath_attr(node
, xpath
, key
)
347 if default
is not NO_DEFAULT
:
350 name
= f
'{xpath}[@{key}]' if name
is None else name
351 raise ExtractorError('Could not find XML attribute %s' % name
)
357 def get_element_by_id(id, html
, **kwargs
):
358 """Return the content of the tag with the specified ID in the passed HTML document"""
359 return get_element_by_attribute('id', id, html
, **kwargs
)
362 def get_element_html_by_id(id, html
, **kwargs
):
363 """Return the html of the tag with the specified ID in the passed HTML document"""
364 return get_element_html_by_attribute('id', id, html
, **kwargs
)
367 def get_element_by_class(class_name
, html
):
368 """Return the content of the first tag with the specified class in the passed HTML document"""
369 retval
= get_elements_by_class(class_name
, html
)
370 return retval
[0] if retval
else None
373 def get_element_html_by_class(class_name
, html
):
374 """Return the html of the first tag with the specified class in the passed HTML document"""
375 retval
= get_elements_html_by_class(class_name
, html
)
376 return retval
[0] if retval
else None
379 def get_element_by_attribute(attribute
, value
, html
, **kwargs
):
380 retval
= get_elements_by_attribute(attribute
, value
, html
, **kwargs
)
381 return retval
[0] if retval
else None
384 def get_element_html_by_attribute(attribute
, value
, html
, **kargs
):
385 retval
= get_elements_html_by_attribute(attribute
, value
, html
, **kargs
)
386 return retval
[0] if retval
else None
389 def get_elements_by_class(class_name
, html
, **kargs
):
390 """Return the content of all tags with the specified class in the passed HTML document as a list"""
391 return get_elements_by_attribute(
392 'class', r
'[^\'"]*(?<=[\'"\s
])%s(?
=[\'"\s])[^\'"]*' % re.escape(class_name),
393 html, escape_value=False)
396 def get_elements_html_by_class(class_name, html):
397 """Return the html of all tags with the specified class in the passed HTML document as a list"""
398 return get_elements_html_by_attribute(
399 'class', r'[^
\'"]*(?<=[\'"\s
])%s(?
=[\'"\s])[^\'"]*' % re.escape(class_name),
400 html, escape_value=False)
403 def get_elements_by_attribute(*args, **kwargs):
404 """Return the content of the tag with the specified attribute in the passed HTML document"""
405 return [content for content, _ in get_elements_text_and_html_by_attribute(*args, **kwargs)]
408 def get_elements_html_by_attribute(*args, **kwargs):
409 """Return the html of the tag with the specified attribute in the passed HTML document"""
410 return [whole for _, whole in get_elements_text_and_html_by_attribute(*args, **kwargs)]
413 def get_elements_text_and_html_by_attribute(attribute, value, html, escape_value=True):
415 Return the text (content) and the html (whole) of the tag with the specified
416 attribute in the passed HTML document
419 quote = '' if re.match(r'''[\s"'`
=<>]''', value) else '?'
421 value = re.escape(value) if escape_value else value
423 partial_element_re = rf'''(?x
)
424 <(?P
<tag
>[a
-zA
-Z0
-9:._-]+)
425 (?
:\
s(?
:[^
>"']|"[^
"]*"|
'[^']*')*)?
426 \s{re.escape(attribute)}\s*=\s*(?P<_q>['"]{quote})(?-x:{value})(?P=_q)
429 for m in re.finditer(partial_element_re, html):
430 content, whole = get_element_text_and_html_by_tag(m.group('tag'), html[m.start():])
433 unescapeHTML(re.sub(r'^(?P<q>["\'])(?P
<content
>.*)(?P
=q
)$
', r'\g
<content
>', content, flags=re.DOTALL)),
438 class HTMLBreakOnClosingTagParser(html.parser.HTMLParser):
440 HTML parser which raises HTMLBreakOnClosingTagException upon reaching the
441 closing tag for the first opening tag it has encountered, and can be used
445 class HTMLBreakOnClosingTagException(Exception):
449 self.tagstack = collections.deque()
450 html.parser.HTMLParser.__init__(self)
455 def __exit__(self, *_):
459 # handle_endtag does not return upon raising HTMLBreakOnClosingTagException,
460 # so data remains buffered; we no longer have any interest in it, thus
461 # override this method to discard it
464 def handle_starttag(self, tag, _):
465 self.tagstack.append(tag)
467 def handle_endtag(self, tag):
468 if not self.tagstack:
469 raise compat_HTMLParseError('no tags
in the stack
')
471 inner_tag = self.tagstack.pop()
475 raise compat_HTMLParseError(f'matching opening tag
for closing {tag} tag
not found
')
476 if not self.tagstack:
477 raise self.HTMLBreakOnClosingTagException()
480 def get_element_text_and_html_by_tag(tag, html):
482 For the first element with the specified tag in the passed HTML document
483 return its' content (text
) and the whole
element (html
)
485 def find_or_raise(haystack, needle, exc):
487 return haystack.index(needle)
490 closing_tag = f'</{tag}>'
491 whole_start = find_or_raise(
492 html, f'<{tag}', compat_HTMLParseError(f'opening {tag} tag not found'))
493 content_start = find_or_raise(
494 html[whole_start:], '>', compat_HTMLParseError(f'malformed opening {tag} tag'))
495 content_start += whole_start + 1
496 with HTMLBreakOnClosingTagParser() as parser:
497 parser.feed(html[whole_start:content_start])
498 if not parser.tagstack or parser.tagstack[0] != tag:
499 raise compat_HTMLParseError(f'parser did not match opening {tag} tag')
500 offset = content_start
501 while offset < len(html):
502 next_closing_tag_start = find_or_raise(
503 html[offset:], closing_tag,
504 compat_HTMLParseError(f'closing {tag} tag not found'))
505 next_closing_tag_end = next_closing_tag_start + len(closing_tag)
507 parser.feed(html[offset:offset + next_closing_tag_end])
508 offset += next_closing_tag_end
509 except HTMLBreakOnClosingTagParser.HTMLBreakOnClosingTagException:
510 return html[content_start:offset + next_closing_tag_start], \
511 html[whole_start:offset + next_closing_tag_end]
512 raise compat_HTMLParseError('unexpected end of html')
515 class HTMLAttributeParser(html.parser.HTMLParser):
516 """Trivial HTML parser to gather the attributes
for a single element
"""
520 html.parser.HTMLParser.__init__(self)
522 def handle_starttag(self, tag, attrs):
523 self.attrs = dict(attrs)
526 class HTMLListAttrsParser(html.parser.HTMLParser):
527 """HTML parser to gather the attributes
for the elements of a
list"""
530 html.parser.HTMLParser.__init__(self)
534 def handle_starttag(self, tag, attrs):
535 if tag == 'li' and self._level == 0:
536 self.items.append(dict(attrs))
539 def handle_endtag(self, tag):
543 def extract_attributes(html_element):
544 """Given a string
for an HTML element such
as
546 a
="foo" B
="bar" c
="&98;az" d
=boz
547 empty
= noval entity
="&"
550 Decode
and return a dictionary of attributes
.
552 'a': 'foo', 'b': 'bar', c
: 'baz', d
: 'boz',
553 'empty': '', 'noval': None, 'entity': '&',
554 'sq': '"', 'dq': '\''
557 parser = HTMLAttributeParser()
558 with contextlib.suppress(compat_HTMLParseError):
559 parser.feed(html_element)
564 def parse_list(webpage):
565 """Given a string
for an series of HTML
<li
> elements
,
566 return a dictionary of their attributes
"""
567 parser = HTMLListAttrsParser()
573 def clean_html(html):
574 """Clean an HTML snippet into a readable string
"""
576 if html is None: # Convenience for sanitizing descriptions etc.
579 html = re.sub(r'\s+', ' ', html)
580 html = re.sub(r'(?u)\s?<\s?br\s?/?\s?>\s?', '\n', html)
581 html = re.sub(r'(?u)<\s?/\s?p\s?>\s?<\s?p[^>]*>', '\n', html)
583 html = re.sub('<.*?>', '', html)
584 # Replace html entities
585 html = unescapeHTML(html)
589 class LenientJSONDecoder(json.JSONDecoder):
590 def __init__(self, *args, transform_source=None, ignore_extra=False, **kwargs):
591 self.transform_source, self.ignore_extra = transform_source, ignore_extra
592 super().__init__(*args, **kwargs)
595 if self.transform_source:
596 s = self.transform_source(s)
597 if self.ignore_extra:
598 return self.raw_decode(s.lstrip())[0]
599 return super().decode(s)
602 def sanitize_open(filename, open_mode):
603 """Try to
open the given filename
, and slightly tweak it
if this fails
.
605 Attempts to
open the given filename
. If this fails
, it tries to change
606 the filename slightly
, step by step
, until it
's either able to open it
607 or it fails and raises a final exception, like the standard open()
610 It returns the tuple (stream, definitive_file_name).
613 if sys.platform == 'win32
':
615 msvcrt.setmode(sys.stdout.fileno(), os.O_BINARY)
616 return (sys.stdout.buffer if hasattr(sys.stdout, 'buffer') else sys.stdout, filename)
618 for attempt in range(2):
621 if sys.platform == 'win32
':
622 # FIXME: An exclusive lock also locks the file from being read.
623 # Since windows locks are mandatory, don't lock the
file on
windows (for now
).
624 # Ref: https://github.com/yt-dlp/yt-dlp/issues/3124
625 raise LockingUnsupportedError()
626 stream
= locked_file(filename
, open_mode
, block
=False).__enter
__()
628 stream
= open(filename
, open_mode
)
629 return stream
, filename
630 except OSError as err
:
631 if attempt
or err
.errno
in (errno
.EACCES
,):
633 old_filename
, filename
= filename
, sanitize_path(filename
)
634 if old_filename
== filename
:
638 def timeconvert(timestr
):
639 """Convert RFC 2822 defined time string into system timestamp"""
641 timetuple
= email
.utils
.parsedate_tz(timestr
)
642 if timetuple
is not None:
643 timestamp
= email
.utils
.mktime_tz(timetuple
)
647 def sanitize_filename(s
, restricted
=False, is_id
=NO_DEFAULT
):
648 """Sanitizes a string so it could be used as part of a filename.
649 @param restricted Use a stricter subset of allowed characters
650 @param is_id Whether this is an ID that should be kept unchanged if possible.
651 If unset, yt-dlp's new sanitization rules are in effect
656 def replace_insane(char
):
657 if restricted
and char
in ACCENT_CHARS
:
658 return ACCENT_CHARS
[char
]
659 elif not restricted
and char
== '\n':
661 elif char
== '?' or ord(char
) < 32 or ord(char
) == 127:
664 return '' if restricted
else '\''
666 return '\0_\0-' if restricted
else '\0 \0-'
667 elif char
in '\\/|*<>':
669 if restricted
and (char
in '!&\'()[]{}$;`^,#' or char
.isspace() or ord(char
) > 127):
673 s
= re
.sub(r
'[0-9]+(?::[0-9]+)+', lambda m
: m
.group(0).replace(':', '_'), s
) # Handle timestamps
674 result
= ''.join(map(replace_insane
, s
))
675 if is_id
is NO_DEFAULT
:
676 result
= re
.sub(r
'(\0.)(?:(?=\1)..)+', r
'\1', result
) # Remove repeated substitute chars
677 STRIP_RE
= r
'(?:\0.|[ _-])*'
678 result
= re
.sub(f
'^\0.{STRIP_RE}|{STRIP_RE}\0.$', '', result
) # Remove substitute chars from start/end
679 result
= result
.replace('\0', '') or '_'
682 while '__' in result
:
683 result
= result
.replace('__', '_')
684 result
= result
.strip('_')
685 # Common case of "Foreign band name - English song title"
686 if restricted
and result
.startswith('-_'):
688 if result
.startswith('-'):
689 result
= '_' + result
[len('-'):]
690 result
= result
.lstrip('.')
696 def sanitize_path(s
, force
=False):
697 """Sanitizes and normalizes path on Windows"""
698 if sys
.platform
== 'win32':
700 drive_or_unc
, _
= os
.path
.splitdrive(s
)
706 norm_path
= os
.path
.normpath(remove_start(s
, drive_or_unc
)).split(os
.path
.sep
)
710 path_part
if path_part
in ['.', '..'] else re
.sub(r
'(?:[/<>:"\|\\?\*]|[\s.]$)', '#', path_part
)
711 for path_part
in norm_path
]
713 sanitized_path
.insert(0, drive_or_unc
+ os
.path
.sep
)
714 elif force
and s
and s
[0] == os
.path
.sep
:
715 sanitized_path
.insert(0, os
.path
.sep
)
716 return os
.path
.join(*sanitized_path
)
719 def sanitize_url(url
):
720 # Prepend protocol-less URLs with `http:` scheme in order to mitigate
721 # the number of unwanted failures due to missing protocol
724 elif url
.startswith('//'):
725 return 'http:%s' % url
726 # Fix some common typos seen so far
728 # https://github.com/ytdl-org/youtube-dl/issues/15649
729 (r
'^httpss://', r
'https://'),
730 # https://bx1.be/lives/direct-tv/
731 (r
'^rmtp([es]?)://', r
'rtmp\1://'),
733 for mistake
, fixup
in COMMON_TYPOS
:
734 if re
.match(mistake
, url
):
735 return re
.sub(mistake
, fixup
, url
)
739 def extract_basic_auth(url
):
740 parts
= urllib
.parse
.urlsplit(url
)
741 if parts
.username
is None:
743 url
= urllib
.parse
.urlunsplit(parts
._replace
(netloc
=(
744 parts
.hostname
if parts
.port
is None
745 else '%s:%d' % (parts
.hostname
, parts
.port
))))
746 auth_payload
= base64
.b64encode(
747 ('%s:%s' % (parts
.username
, parts
.password
or '')).encode())
748 return url
, f
'Basic {auth_payload.decode()}'
751 def sanitized_Request(url
, *args
, **kwargs
):
752 url
, auth_header
= extract_basic_auth(escape_url(sanitize_url(url
)))
753 if auth_header
is not None:
754 headers
= args
[1] if len(args
) >= 2 else kwargs
.setdefault('headers', {})
755 headers
['Authorization'] = auth_header
756 return urllib
.request
.Request(url
, *args
, **kwargs
)
760 """Expand shell variables and ~"""
761 return os
.path
.expandvars(compat_expanduser(s
))
764 def orderedSet(iterable
, *, lazy
=False):
765 """Remove all duplicates from the input iterable"""
767 seen
= [] # Do not use set since the items can be unhashable
773 return _iter() if lazy
else list(_iter())
776 def _htmlentity_transform(entity_with_semicolon
):
777 """Transforms an HTML entity to a character."""
778 entity
= entity_with_semicolon
[:-1]
780 # Known non-numeric HTML entity
781 if entity
in html
.entities
.name2codepoint
:
782 return chr(html
.entities
.name2codepoint
[entity
])
784 # TODO: HTML5 allows entities without a semicolon. For example,
785 # 'Éric' should be decoded as 'Éric'.
786 if entity_with_semicolon
in html
.entities
.html5
:
787 return html
.entities
.html5
[entity_with_semicolon
]
789 mobj
= re
.match(r
'#(x[0-9a-fA-F]+|[0-9]+)', entity
)
791 numstr
= mobj
.group(1)
792 if numstr
.startswith('x'):
794 numstr
= '0%s' % numstr
797 # See https://github.com/ytdl-org/youtube-dl/issues/7518
798 with contextlib
.suppress(ValueError):
799 return chr(int(numstr
, base
))
801 # Unknown entity in name, return its literal representation
802 return '&%s;' % entity
808 assert isinstance(s
, str)
811 r
'&([^&;]+;)', lambda m
: _htmlentity_transform(m
.group(1)), s
)
814 def escapeHTML(text
):
817 .replace('&', '&')
818 .replace('<', '<')
819 .replace('>', '>')
820 .replace('"', '"')
821 .replace("'", ''')
825 def process_communicate_or_kill(p
, *args
, **kwargs
):
826 write_string('DeprecationWarning: yt_dlp.utils.process_communicate_or_kill is deprecated '
827 'and may be removed in a future version. Use yt_dlp.utils.Popen.communicate_or_kill instead')
828 return Popen
.communicate_or_kill(p
, *args
, **kwargs
)
831 class Popen(subprocess
.Popen
):
832 if sys
.platform
== 'win32':
833 _startupinfo
= subprocess
.STARTUPINFO()
834 _startupinfo
.dwFlags |
= subprocess
.STARTF_USESHOWWINDOW
838 def __init__(self
, *args
, text
=False, **kwargs
):
840 kwargs
['universal_newlines'] = True # For 3.6 compatibility
841 kwargs
.setdefault('encoding', 'utf-8')
842 kwargs
.setdefault('errors', 'replace')
843 super().__init
__(*args
, **kwargs
, startupinfo
=self
._startupinfo
)
845 def communicate_or_kill(self
, *args
, **kwargs
):
847 return self
.communicate(*args
, **kwargs
)
848 except BaseException
: # Including KeyboardInterrupt
849 self
.kill(timeout
=None)
852 def kill(self
, *, timeout
=0):
855 self
.wait(timeout
=timeout
)
858 def run(cls
, *args
, **kwargs
):
859 with cls(*args
, **kwargs
) as proc
:
860 stdout
, stderr
= proc
.communicate_or_kill()
861 return stdout
or '', stderr
or '', proc
.returncode
864 def get_subprocess_encoding():
865 if sys
.platform
== 'win32' and sys
.getwindowsversion()[0] >= 5:
866 # For subprocess calls, encode with locale encoding
867 # Refer to http://stackoverflow.com/a/9951851/35070
868 encoding
= preferredencoding()
870 encoding
= sys
.getfilesystemencoding()
876 def encodeFilename(s
, for_subprocess
=False):
877 assert isinstance(s
, str)
881 def decodeFilename(b
, for_subprocess
=False):
885 def encodeArgument(s
):
886 # Legacy code that uses byte strings
887 # Uncomment the following line after fixing all post processors
888 # assert isinstance(s, str), 'Internal error: %r should be of type %r, is %r' % (s, str, type(s))
889 return s
if isinstance(s
, str) else s
.decode('ascii')
892 def decodeArgument(b
):
896 def decodeOption(optval
):
899 if isinstance(optval
, bytes):
900 optval
= optval
.decode(preferredencoding())
902 assert isinstance(optval
, str)
906 _timetuple
= collections
.namedtuple('Time', ('hours', 'minutes', 'seconds', 'milliseconds'))
909 def timetuple_from_msec(msec
):
910 secs
, msec
= divmod(msec
, 1000)
911 mins
, secs
= divmod(secs
, 60)
912 hrs
, mins
= divmod(mins
, 60)
913 return _timetuple(hrs
, mins
, secs
, msec
)
916 def formatSeconds(secs
, delim
=':', msec
=False):
917 time
= timetuple_from_msec(secs
* 1000)
919 ret
= '%d%s%02d%s%02d' % (time
.hours
, delim
, time
.minutes
, delim
, time
.seconds
)
921 ret
= '%d%s%02d' % (time
.minutes
, delim
, time
.seconds
)
923 ret
= '%d' % time
.seconds
924 return '%s.%03d' % (ret
, time
.milliseconds
) if msec
else ret
927 def _ssl_load_windows_store_certs(ssl_context
, storename
):
928 # Code adapted from _load_windows_store_certs in https://github.com/python/cpython/blob/main/Lib/ssl.py
930 certs
= [cert
for cert
, encoding
, trust
in ssl
.enum_certificates(storename
)
931 if encoding
== 'x509_asn' and (
932 trust
is True or ssl
.Purpose
.SERVER_AUTH
.oid
in trust
)]
933 except PermissionError
:
936 with contextlib
.suppress(ssl
.SSLError
):
937 ssl_context
.load_verify_locations(cadata
=cert
)
940 def make_HTTPS_handler(params
, **kwargs
):
941 opts_check_certificate
= not params
.get('nocheckcertificate')
942 context
= ssl
.SSLContext(ssl
.PROTOCOL_TLS_CLIENT
)
943 context
.check_hostname
= opts_check_certificate
944 if params
.get('legacyserverconnect'):
945 context
.options |
= 4 # SSL_OP_LEGACY_SERVER_CONNECT
946 # Allow use of weaker ciphers in Python 3.10+. See https://bugs.python.org/issue43998
947 context
.set_ciphers('DEFAULT')
949 context
.verify_mode
= ssl
.CERT_REQUIRED
if opts_check_certificate
else ssl
.CERT_NONE
950 if opts_check_certificate
:
951 if has_certifi
and 'no-certifi' not in params
.get('compat_opts', []):
952 context
.load_verify_locations(cafile
=certifi
.where())
954 context
.load_default_certs()
955 # Work around the issue in load_default_certs when there are bad certificates. See:
956 # https://github.com/yt-dlp/yt-dlp/issues/1060,
957 # https://bugs.python.org/issue35665, https://bugs.python.org/issue45312
959 # enum_certificates is not present in mingw python. See https://github.com/yt-dlp/yt-dlp/issues/1151
960 if sys
.platform
== 'win32' and hasattr(ssl
, 'enum_certificates'):
961 for storename
in ('CA', 'ROOT'):
962 _ssl_load_windows_store_certs(context
, storename
)
963 context
.set_default_verify_paths()
965 client_certfile
= params
.get('client_certificate')
968 context
.load_cert_chain(
969 client_certfile
, keyfile
=params
.get('client_certificate_key'),
970 password
=params
.get('client_certificate_password'))
972 raise YoutubeDLError('Unable to load client certificate')
974 # Some servers may reject requests if ALPN extension is not sent. See:
975 # https://github.com/python/cpython/issues/85140
976 # https://github.com/yt-dlp/yt-dlp/issues/3878
977 with contextlib
.suppress(NotImplementedError):
978 context
.set_alpn_protocols(['http/1.1'])
980 return YoutubeDLHTTPSHandler(params
, context
=context
, **kwargs
)
983 def bug_reports_message(before
=';'):
984 from .update
import REPOSITORY
986 msg
= (f
'please report this issue on https://github.com/{REPOSITORY}/issues?q= , '
987 'filling out the appropriate issue template. Confirm you are on the latest version using yt-dlp -U')
989 before
= before
.rstrip()
990 if not before
or before
.endswith(('.', '!', '?')):
991 msg
= msg
[0].title() + msg
[1:]
993 return (before
+ ' ' if before
else '') + msg
996 class YoutubeDLError(Exception):
997 """Base exception for YoutubeDL errors."""
1000 def __init__(self
, msg
=None):
1003 elif self
.msg
is None:
1004 self
.msg
= type(self
).__name
__
1005 super().__init
__(self
.msg
)
1008 network_exceptions
= [urllib
.error
.URLError
, http
.client
.HTTPException
, socket
.error
]
1009 if hasattr(ssl
, 'CertificateError'):
1010 network_exceptions
.append(ssl
.CertificateError
)
1011 network_exceptions
= tuple(network_exceptions
)
1014 class ExtractorError(YoutubeDLError
):
1015 """Error during info extraction."""
1017 def __init__(self
, msg
, tb
=None, expected
=False, cause
=None, video_id
=None, ie
=None):
1018 """ tb, if given, is the original traceback (so that it can be printed out).
1019 If expected is set, this is a normal error message and most likely not a bug in yt-dlp.
1021 if sys
.exc_info()[0] in network_exceptions
:
1024 self
.orig_msg
= str(msg
)
1026 self
.expected
= expected
1028 self
.video_id
= video_id
1030 self
.exc_info
= sys
.exc_info() # preserve original exception
1031 if isinstance(self
.exc_info
[1], ExtractorError
):
1032 self
.exc_info
= self
.exc_info
[1].exc_info
1034 super().__init
__(''.join((
1035 format_field(ie
, None, '[%s] '),
1036 format_field(video_id
, None, '%s: '),
1038 format_field(cause
, None, ' (caused by %r)'),
1039 '' if expected
else bug_reports_message())))
1041 def format_traceback(self
):
1042 return join_nonempty(
1043 self
.traceback
and ''.join(traceback
.format_tb(self
.traceback
)),
1044 self
.cause
and ''.join(traceback
.format_exception(None, self
.cause
, self
.cause
.__traceback
__)[1:]),
1048 class UnsupportedError(ExtractorError
):
1049 def __init__(self
, url
):
1051 'Unsupported URL: %s' % url
, expected
=True)
1055 class RegexNotFoundError(ExtractorError
):
1056 """Error when a regex didn't match"""
1060 class GeoRestrictedError(ExtractorError
):
1061 """Geographic restriction Error exception.
1063 This exception may be thrown when a video is not available from your
1064 geographic location due to geographic restrictions imposed by a website.
1067 def __init__(self
, msg
, countries
=None, **kwargs
):
1068 kwargs
['expected'] = True
1069 super().__init
__(msg
, **kwargs
)
1070 self
.countries
= countries
1073 class DownloadError(YoutubeDLError
):
1074 """Download Error exception.
1076 This exception may be thrown by FileDownloader objects if they are not
1077 configured to continue on errors. They will contain the appropriate
1081 def __init__(self
, msg
, exc_info
=None):
1082 """ exc_info, if given, is the original exception that caused the trouble (as returned by sys.exc_info()). """
1083 super().__init
__(msg
)
1084 self
.exc_info
= exc_info
1087 class EntryNotInPlaylist(YoutubeDLError
):
1088 """Entry not in playlist exception.
1090 This exception will be thrown by YoutubeDL when a requested entry
1091 is not found in the playlist info_dict
1093 msg
= 'Entry not found in info'
1096 class SameFileError(YoutubeDLError
):
1097 """Same File exception.
1099 This exception will be thrown by FileDownloader objects if they detect
1100 multiple files would have to be downloaded to the same file on disk.
1102 msg
= 'Fixed output name but more than one file to download'
1104 def __init__(self
, filename
=None):
1105 if filename
is not None:
1106 self
.msg
+= f
': {filename}'
1107 super().__init
__(self
.msg
)
1110 class PostProcessingError(YoutubeDLError
):
1111 """Post Processing exception.
1113 This exception may be raised by PostProcessor's .run() method to
1114 indicate an error in the postprocessing task.
1118 class DownloadCancelled(YoutubeDLError
):
1119 """ Exception raised when the download queue should be interrupted """
1120 msg
= 'The download was cancelled'
1123 class ExistingVideoReached(DownloadCancelled
):
1124 """ --break-on-existing triggered """
1125 msg
= 'Encountered a video that is already in the archive, stopping due to --break-on-existing'
1128 class RejectedVideoReached(DownloadCancelled
):
1129 """ --break-on-reject triggered """
1130 msg
= 'Encountered a video that did not match filter, stopping due to --break-on-reject'
1133 class MaxDownloadsReached(DownloadCancelled
):
1134 """ --max-downloads limit has been reached. """
1135 msg
= 'Maximum number of downloads reached, stopping due to --max-downloads'
1138 class ReExtractInfo(YoutubeDLError
):
1139 """ Video info needs to be re-extracted. """
1141 def __init__(self
, msg
, expected
=False):
1142 super().__init
__(msg
)
1143 self
.expected
= expected
1146 class ThrottledDownload(ReExtractInfo
):
1147 """ Download speed below --throttled-rate. """
1148 msg
= 'The download speed is below throttle limit'
1151 super().__init
__(self
.msg
, expected
=False)
1154 class UnavailableVideoError(YoutubeDLError
):
1155 """Unavailable Format exception.
1157 This exception will be thrown when a video is requested
1158 in a format that is not available for that video.
1160 msg
= 'Unable to download video'
1162 def __init__(self
, err
=None):
1164 self
.msg
+= f
': {err}'
1165 super().__init
__(self
.msg
)
1168 class ContentTooShortError(YoutubeDLError
):
1169 """Content Too Short exception.
1171 This exception may be raised by FileDownloader objects when a file they
1172 download is too small for what the server announced first, indicating
1173 the connection was probably interrupted.
1176 def __init__(self
, downloaded
, expected
):
1177 super().__init
__(f
'Downloaded {downloaded} bytes, expected {expected} bytes')
1179 self
.downloaded
= downloaded
1180 self
.expected
= expected
1183 class XAttrMetadataError(YoutubeDLError
):
1184 def __init__(self
, code
=None, msg
='Unknown error'):
1185 super().__init
__(msg
)
1189 # Parsing code and msg
1190 if (self
.code
in (errno
.ENOSPC
, errno
.EDQUOT
)
1191 or 'No space left' in self
.msg
or 'Disk quota exceeded' in self
.msg
):
1192 self
.reason
= 'NO_SPACE'
1193 elif self
.code
== errno
.E2BIG
or 'Argument list too long' in self
.msg
:
1194 self
.reason
= 'VALUE_TOO_LONG'
1196 self
.reason
= 'NOT_SUPPORTED'
1199 class XAttrUnavailableError(YoutubeDLError
):
1203 def _create_http_connection(ydl_handler
, http_class
, is_https
, *args
, **kwargs
):
1204 hc
= http_class(*args
, **kwargs
)
1205 source_address
= ydl_handler
._params
.get('source_address')
1207 if source_address
is not None:
1208 # This is to workaround _create_connection() from socket where it will try all
1209 # address data from getaddrinfo() including IPv6. This filters the result from
1210 # getaddrinfo() based on the source_address value.
1211 # This is based on the cpython socket.create_connection() function.
1212 # https://github.com/python/cpython/blob/master/Lib/socket.py#L691
1213 def _create_connection(address
, timeout
=socket
._GLOBAL
_DEFAULT
_TIMEOUT
, source_address
=None):
1214 host
, port
= address
1216 addrs
= socket
.getaddrinfo(host
, port
, 0, socket
.SOCK_STREAM
)
1217 af
= socket
.AF_INET
if '.' in source_address
[0] else socket
.AF_INET6
1218 ip_addrs
= [addr
for addr
in addrs
if addr
[0] == af
]
1219 if addrs
and not ip_addrs
:
1220 ip_version
= 'v4' if af
== socket
.AF_INET
else 'v6'
1222 "No remote IP%s addresses available for connect, can't use '%s' as source address"
1223 % (ip_version
, source_address
[0]))
1224 for res
in ip_addrs
:
1225 af
, socktype
, proto
, canonname
, sa
= res
1228 sock
= socket
.socket(af
, socktype
, proto
)
1229 if timeout
is not socket
._GLOBAL
_DEFAULT
_TIMEOUT
:
1230 sock
.settimeout(timeout
)
1231 sock
.bind(source_address
)
1233 err
= None # Explicitly break reference cycle
1235 except OSError as _
:
1237 if sock
is not None:
1242 raise OSError('getaddrinfo returns an empty list')
1243 if hasattr(hc
, '_create_connection'):
1244 hc
._create
_connection
= _create_connection
1245 hc
.source_address
= (source_address
, 0)
1250 def handle_youtubedl_headers(headers
):
1251 filtered_headers
= headers
1253 if 'Youtubedl-no-compression' in filtered_headers
:
1254 filtered_headers
= {k: v for k, v in filtered_headers.items() if k.lower() != 'accept-encoding'}
1255 del filtered_headers
['Youtubedl-no-compression']
1257 return filtered_headers
1260 class YoutubeDLHandler(urllib
.request
.HTTPHandler
):
1261 """Handler for HTTP requests and responses.
1263 This class, when installed with an OpenerDirector, automatically adds
1264 the standard headers to every HTTP request and handles gzipped and
1265 deflated responses from web servers. If compression is to be avoided in
1266 a particular request, the original request in the program code only has
1267 to include the HTTP header "Youtubedl-no-compression", which will be
1268 removed before making the real request.
1270 Part of this code was copied from:
1272 http://techknack.net/python-urllib2-handlers/
1274 Andrew Rowls, the author of that code, agreed to release it to the
1278 def __init__(self
, params
, *args
, **kwargs
):
1279 urllib
.request
.HTTPHandler
.__init
__(self
, *args
, **kwargs
)
1280 self
._params
= params
1282 def http_open(self
, req
):
1283 conn_class
= http
.client
.HTTPConnection
1285 socks_proxy
= req
.headers
.get('Ytdl-socks-proxy')
1287 conn_class
= make_socks_conn_class(conn_class
, socks_proxy
)
1288 del req
.headers
['Ytdl-socks-proxy']
1290 return self
.do_open(functools
.partial(
1291 _create_http_connection
, self
, conn_class
, False),
1299 return zlib
.decompress(data
, -zlib
.MAX_WBITS
)
1301 return zlib
.decompress(data
)
1307 return brotli
.decompress(data
)
1309 def http_request(self
, req
):
1310 # According to RFC 3986, URLs can not contain non-ASCII characters, however this is not
1311 # always respected by websites, some tend to give out URLs with non percent-encoded
1312 # non-ASCII characters (see telemb.py, ard.py [#3412])
1313 # urllib chokes on URLs with non-ASCII characters (see http://bugs.python.org/issue3991)
1314 # To work around aforementioned issue we will replace request's original URL with
1315 # percent-encoded one
1316 # Since redirects are also affected (e.g. http://www.southpark.de/alle-episoden/s18e09)
1317 # the code of this workaround has been moved here from YoutubeDL.urlopen()
1318 url
= req
.get_full_url()
1319 url_escaped
= escape_url(url
)
1321 # Substitute URL if any change after escaping
1322 if url
!= url_escaped
:
1323 req
= update_Request(req
, url
=url_escaped
)
1325 for h
, v
in self
._params
.get('http_headers', std_headers
).items():
1326 # Capitalize is needed because of Python bug 2275: http://bugs.python.org/issue2275
1327 # The dict keys are capitalized because of this bug by urllib
1328 if h
.capitalize() not in req
.headers
:
1329 req
.add_header(h
, v
)
1331 if 'Accept-encoding' not in req
.headers
:
1332 req
.add_header('Accept-encoding', ', '.join(SUPPORTED_ENCODINGS
))
1334 req
.headers
= handle_youtubedl_headers(req
.headers
)
1336 return super().do_request_(req
)
1338 def http_response(self
, req
, resp
):
1341 if resp
.headers
.get('Content-encoding', '') == 'gzip':
1342 content
= resp
.read()
1343 gz
= gzip
.GzipFile(fileobj
=io
.BytesIO(content
), mode
='rb')
1345 uncompressed
= io
.BytesIO(gz
.read())
1346 except OSError as original_ioerror
:
1347 # There may be junk add the end of the file
1348 # See http://stackoverflow.com/q/4928560/35070 for details
1349 for i
in range(1, 1024):
1351 gz
= gzip
.GzipFile(fileobj
=io
.BytesIO(content
[:-i
]), mode
='rb')
1352 uncompressed
= io
.BytesIO(gz
.read())
1357 raise original_ioerror
1358 resp
= urllib
.request
.addinfourl(uncompressed
, old_resp
.headers
, old_resp
.url
, old_resp
.code
)
1359 resp
.msg
= old_resp
.msg
1360 del resp
.headers
['Content-encoding']
1362 if resp
.headers
.get('Content-encoding', '') == 'deflate':
1363 gz
= io
.BytesIO(self
.deflate(resp
.read()))
1364 resp
= urllib
.request
.addinfourl(gz
, old_resp
.headers
, old_resp
.url
, old_resp
.code
)
1365 resp
.msg
= old_resp
.msg
1366 del resp
.headers
['Content-encoding']
1368 if resp
.headers
.get('Content-encoding', '') == 'br':
1369 resp
= urllib
.request
.addinfourl(
1370 io
.BytesIO(self
.brotli(resp
.read())), old_resp
.headers
, old_resp
.url
, old_resp
.code
)
1371 resp
.msg
= old_resp
.msg
1372 del resp
.headers
['Content-encoding']
1373 # Percent-encode redirect URL of Location HTTP header to satisfy RFC 3986 (see
1374 # https://github.com/ytdl-org/youtube-dl/issues/6457).
1375 if 300 <= resp
.code
< 400:
1376 location
= resp
.headers
.get('Location')
1378 # As of RFC 2616 default charset is iso-8859-1 that is respected by python 3
1379 location
= location
.encode('iso-8859-1').decode()
1380 location_escaped
= escape_url(location
)
1381 if location
!= location_escaped
:
1382 del resp
.headers
['Location']
1383 resp
.headers
['Location'] = location_escaped
1386 https_request
= http_request
1387 https_response
= http_response
1390 def make_socks_conn_class(base_class
, socks_proxy
):
1391 assert issubclass(base_class
, (
1392 http
.client
.HTTPConnection
, http
.client
.HTTPSConnection
))
1394 url_components
= urllib
.parse
.urlparse(socks_proxy
)
1395 if url_components
.scheme
.lower() == 'socks5':
1396 socks_type
= ProxyType
.SOCKS5
1397 elif url_components
.scheme
.lower() in ('socks', 'socks4'):
1398 socks_type
= ProxyType
.SOCKS4
1399 elif url_components
.scheme
.lower() == 'socks4a':
1400 socks_type
= ProxyType
.SOCKS4A
1402 def unquote_if_non_empty(s
):
1405 return urllib
.parse
.unquote_plus(s
)
1409 url_components
.hostname
, url_components
.port
or 1080,
1411 unquote_if_non_empty(url_components
.username
),
1412 unquote_if_non_empty(url_components
.password
),
1415 class SocksConnection(base_class
):
1417 self
.sock
= sockssocket()
1418 self
.sock
.setproxy(*proxy_args
)
1419 if isinstance(self
.timeout
, (int, float)):
1420 self
.sock
.settimeout(self
.timeout
)
1421 self
.sock
.connect((self
.host
, self
.port
))
1423 if isinstance(self
, http
.client
.HTTPSConnection
):
1424 if hasattr(self
, '_context'): # Python > 2.6
1425 self
.sock
= self
._context
.wrap_socket(
1426 self
.sock
, server_hostname
=self
.host
)
1428 self
.sock
= ssl
.wrap_socket(self
.sock
)
1430 return SocksConnection
1433 class YoutubeDLHTTPSHandler(urllib
.request
.HTTPSHandler
):
1434 def __init__(self
, params
, https_conn_class
=None, *args
, **kwargs
):
1435 urllib
.request
.HTTPSHandler
.__init
__(self
, *args
, **kwargs
)
1436 self
._https
_conn
_class
= https_conn_class
or http
.client
.HTTPSConnection
1437 self
._params
= params
1439 def https_open(self
, req
):
1441 conn_class
= self
._https
_conn
_class
1443 if hasattr(self
, '_context'): # python > 2.6
1444 kwargs
['context'] = self
._context
1445 if hasattr(self
, '_check_hostname'): # python 3.x
1446 kwargs
['check_hostname'] = self
._check
_hostname
1448 socks_proxy
= req
.headers
.get('Ytdl-socks-proxy')
1450 conn_class
= make_socks_conn_class(conn_class
, socks_proxy
)
1451 del req
.headers
['Ytdl-socks-proxy']
1454 return self
.do_open(
1455 functools
.partial(_create_http_connection
, self
, conn_class
, True), req
, **kwargs
)
1456 except urllib
.error
.URLError
as e
:
1457 if (isinstance(e
.reason
, ssl
.SSLError
)
1458 and getattr(e
.reason
, 'reason', None) == 'SSLV3_ALERT_HANDSHAKE_FAILURE'):
1459 raise YoutubeDLError('SSLV3_ALERT_HANDSHAKE_FAILURE: Try using --legacy-server-connect')
1463 class YoutubeDLCookieJar(http
.cookiejar
.MozillaCookieJar
):
1465 See [1] for cookie file format.
1467 1. https://curl.haxx.se/docs/http-cookies.html
1469 _HTTPONLY_PREFIX
= '#HttpOnly_'
1471 _HEADER
= '''# Netscape HTTP Cookie File
1472 # This file is generated by yt-dlp. Do not edit.
1475 _CookieFileEntry
= collections
.namedtuple(
1477 ('domain_name', 'include_subdomains', 'path', 'https_only', 'expires_at', 'name', 'value'))
1479 def __init__(self
, filename
=None, *args
, **kwargs
):
1480 super().__init
__(None, *args
, **kwargs
)
1481 if self
.is_path(filename
):
1482 filename
= os
.fspath(filename
)
1483 self
.filename
= filename
1486 def _true_or_false(cndn
):
1487 return 'TRUE' if cndn
else 'FALSE'
1491 return isinstance(file, (str, bytes, os
.PathLike
))
1493 @contextlib.contextmanager
1494 def open(self
, file, *, write
=False):
1495 if self
.is_path(file):
1496 with open(file, 'w' if write
else 'r', encoding
='utf-8') as f
:
1503 def _really_save(self
, f
, ignore_discard
=False, ignore_expires
=False):
1506 if (not ignore_discard
and cookie
.discard
1507 or not ignore_expires
and cookie
.is_expired(now
)):
1509 name
, value
= cookie
.name
, cookie
.value
1511 # cookies.txt regards 'Set-Cookie: foo' as a cookie
1512 # with no name, whereas http.cookiejar regards it as a
1513 # cookie with no value.
1514 name
, value
= '', name
1515 f
.write('%s\n' % '\t'.join((
1517 self
._true
_or
_false
(cookie
.domain
.startswith('.')),
1519 self
._true
_or
_false
(cookie
.secure
),
1520 str_or_none(cookie
.expires
, default
=''),
1524 def save(self
, filename
=None, *args
, **kwargs
):
1526 Save cookies to a file.
1527 Code is taken from CPython 3.6
1528 https://github.com/python/cpython/blob/8d999cbf4adea053be6dbb612b9844635c4dfb8e/Lib/http/cookiejar.py#L2091-L2117 """
1530 if filename
is None:
1531 if self
.filename
is not None:
1532 filename
= self
.filename
1534 raise ValueError(http
.cookiejar
.MISSING_FILENAME_TEXT
)
1536 # Store session cookies with `expires` set to 0 instead of an empty string
1538 if cookie
.expires
is None:
1541 with self
.open(filename
, write
=True) as f
:
1542 f
.write(self
._HEADER
)
1543 self
._really
_save
(f
, *args
, **kwargs
)
1545 def load(self
, filename
=None, ignore_discard
=False, ignore_expires
=False):
1546 """Load cookies from a file."""
1547 if filename
is None:
1548 if self
.filename
is not None:
1549 filename
= self
.filename
1551 raise ValueError(http
.cookiejar
.MISSING_FILENAME_TEXT
)
1553 def prepare_line(line
):
1554 if line
.startswith(self
._HTTPONLY
_PREFIX
):
1555 line
= line
[len(self
._HTTPONLY
_PREFIX
):]
1556 # comments and empty lines are fine
1557 if line
.startswith('#') or not line
.strip():
1559 cookie_list
= line
.split('\t')
1560 if len(cookie_list
) != self
._ENTRY
_LEN
:
1561 raise http
.cookiejar
.LoadError('invalid length %d' % len(cookie_list
))
1562 cookie
= self
._CookieFileEntry
(*cookie_list
)
1563 if cookie
.expires_at
and not cookie
.expires_at
.isdigit():
1564 raise http
.cookiejar
.LoadError('invalid expires at %s' % cookie
.expires_at
)
1568 with self
.open(filename
) as f
:
1571 cf
.write(prepare_line(line
))
1572 except http
.cookiejar
.LoadError
as e
:
1573 if f
'{line.strip()} '[0] in '[{"':
1574 raise http
.cookiejar
.LoadError(
1575 'Cookies file must be Netscape formatted, not JSON. See '
1576 'https://github.com/ytdl-org/youtube-dl#how-do-i-pass-cookies-to-youtube-dl')
1577 write_string(f
'WARNING: skipping cookie file entry due to {e}: {line!r}\n')
1580 self
._really
_load
(cf
, filename
, ignore_discard
, ignore_expires
)
1581 # Session cookies are denoted by either `expires` field set to
1582 # an empty string or 0. MozillaCookieJar only recognizes the former
1583 # (see [1]). So we need force the latter to be recognized as session
1584 # cookies on our own.
1585 # Session cookies may be important for cookies-based authentication,
1586 # e.g. usually, when user does not check 'Remember me' check box while
1587 # logging in on a site, some important cookies are stored as session
1588 # cookies so that not recognizing them will result in failed login.
1589 # 1. https://bugs.python.org/issue17164
1591 # Treat `expires=0` cookies as session cookies
1592 if cookie
.expires
== 0:
1593 cookie
.expires
= None
1594 cookie
.discard
= True
1597 class YoutubeDLCookieProcessor(urllib
.request
.HTTPCookieProcessor
):
1598 def __init__(self
, cookiejar
=None):
1599 urllib
.request
.HTTPCookieProcessor
.__init
__(self
, cookiejar
)
1601 def http_response(self
, request
, response
):
1602 return urllib
.request
.HTTPCookieProcessor
.http_response(self
, request
, response
)
1604 https_request
= urllib
.request
.HTTPCookieProcessor
.http_request
1605 https_response
= http_response
1608 class YoutubeDLRedirectHandler(urllib
.request
.HTTPRedirectHandler
):
1609 """YoutubeDL redirect handler
1611 The code is based on HTTPRedirectHandler implementation from CPython [1].
1613 This redirect handler solves two issues:
1614 - ensures redirect URL is always unicode under python 2
1615 - introduces support for experimental HTTP response status code
1616 308 Permanent Redirect [2] used by some sites [3]
1618 1. https://github.com/python/cpython/blob/master/Lib/urllib/request.py
1619 2. https://developer.mozilla.org/en-US/docs/Web/HTTP/Status/308
1620 3. https://github.com/ytdl-org/youtube-dl/issues/28768
1623 http_error_301
= http_error_303
= http_error_307
= http_error_308
= urllib
.request
.HTTPRedirectHandler
.http_error_302
1625 def redirect_request(self
, req
, fp
, code
, msg
, headers
, newurl
):
1626 """Return a Request or None in response to a redirect.
1628 This is called by the http_error_30x methods when a
1629 redirection response is received. If a redirection should
1630 take place, return a new Request to allow http_error_30x to
1631 perform the redirect. Otherwise, raise HTTPError if no-one
1632 else should try to handle this url. Return None if you can't
1633 but another Handler might.
1635 m
= req
.get_method()
1636 if (not (code
in (301, 302, 303, 307, 308) and m
in ("GET", "HEAD")
1637 or code
in (301, 302, 303) and m
== "POST")):
1638 raise urllib
.error
.HTTPError(req
.full_url
, code
, msg
, headers
, fp
)
1639 # Strictly (according to RFC 2616), 301 or 302 in response to
1640 # a POST MUST NOT cause a redirection without confirmation
1641 # from the user (of urllib.request, in this case). In practice,
1642 # essentially all clients do redirect in this case, so we do
1645 # Be conciliant with URIs containing a space. This is mainly
1646 # redundant with the more complete encoding done in http_error_302(),
1647 # but it is kept for compatibility with other callers.
1648 newurl
= newurl
.replace(' ', '%20')
1650 CONTENT_HEADERS
= ("content-length", "content-type")
1651 # NB: don't use dict comprehension for python 2.6 compatibility
1652 newheaders
= {k: v for k, v in req.headers.items() if k.lower() not in CONTENT_HEADERS}
1654 # A 303 must either use GET or HEAD for subsequent request
1655 # https://datatracker.ietf.org/doc/html/rfc7231#section-6.4.4
1656 if code
== 303 and m
!= 'HEAD':
1658 # 301 and 302 redirects are commonly turned into a GET from a POST
1659 # for subsequent requests by browsers, so we'll do the same.
1660 # https://datatracker.ietf.org/doc/html/rfc7231#section-6.4.2
1661 # https://datatracker.ietf.org/doc/html/rfc7231#section-6.4.3
1662 if code
in (301, 302) and m
== 'POST':
1665 return urllib
.request
.Request(
1666 newurl
, headers
=newheaders
, origin_req_host
=req
.origin_req_host
,
1667 unverifiable
=True, method
=m
)
1670 def extract_timezone(date_str
):
1673 ^.{8,}? # >=8 char non-TZ prefix, if present
1674 (?P<tz>Z| # just the UTC Z, or
1675 (?:(?<=.\b\d{4}|\b\d{2}:\d\d)| # preceded by 4 digits or hh:mm or
1676 (?<!.\b[a-zA-Z]{3}|[a-zA-Z]{4}|..\b\d\d)) # not preceded by 3 alpha word or >= 4 alpha or 2 digits
1677 [ ]? # optional space
1678 (?P<sign>\+|-) # +/-
1679 (?P<hours>[0-9]{2}):?(?P<minutes>[0-9]{2}) # hh[:]mm
1683 timezone
= datetime
.timedelta()
1685 date_str
= date_str
[:-len(m
.group('tz'))]
1686 if not m
.group('sign'):
1687 timezone
= datetime
.timedelta()
1689 sign
= 1 if m
.group('sign') == '+' else -1
1690 timezone
= datetime
.timedelta(
1691 hours
=sign
* int(m
.group('hours')),
1692 minutes
=sign
* int(m
.group('minutes')))
1693 return timezone
, date_str
1696 def parse_iso8601(date_str
, delimiter
='T', timezone
=None):
1697 """ Return a UNIX timestamp from the given date """
1699 if date_str
is None:
1702 date_str
= re
.sub(r
'\.[0-9]+', '', date_str
)
1704 if timezone
is None:
1705 timezone
, date_str
= extract_timezone(date_str
)
1707 with contextlib
.suppress(ValueError):
1708 date_format
= f
'%Y-%m-%d{delimiter}%H:%M:%S'
1709 dt
= datetime
.datetime
.strptime(date_str
, date_format
) - timezone
1710 return calendar
.timegm(dt
.timetuple())
1713 def date_formats(day_first
=True):
1714 return DATE_FORMATS_DAY_FIRST
if day_first
else DATE_FORMATS_MONTH_FIRST
1717 def unified_strdate(date_str
, day_first
=True):
1718 """Return a string with the date in the format YYYYMMDD"""
1720 if date_str
is None:
1724 date_str
= date_str
.replace(',', ' ')
1725 # Remove AM/PM + timezone
1726 date_str
= re
.sub(r
'(?i)\s*(?:AM|PM)(?:\s+[A-Z]+)?', '', date_str
)
1727 _
, date_str
= extract_timezone(date_str
)
1729 for expression
in date_formats(day_first
):
1730 with contextlib
.suppress(ValueError):
1731 upload_date
= datetime
.datetime
.strptime(date_str
, expression
).strftime('%Y%m%d')
1732 if upload_date
is None:
1733 timetuple
= email
.utils
.parsedate_tz(date_str
)
1735 with contextlib
.suppress(ValueError):
1736 upload_date
= datetime
.datetime(*timetuple
[:6]).strftime('%Y%m%d')
1737 if upload_date
is not None:
1738 return str(upload_date
)
1741 def unified_timestamp(date_str
, day_first
=True):
1742 if date_str
is None:
1745 date_str
= re
.sub(r
'[,|]', '', date_str
)
1747 pm_delta
= 12 if re
.search(r
'(?i)PM', date_str
) else 0
1748 timezone
, date_str
= extract_timezone(date_str
)
1750 # Remove AM/PM + timezone
1751 date_str
= re
.sub(r
'(?i)\s*(?:AM|PM)(?:\s+[A-Z]+)?', '', date_str
)
1753 # Remove unrecognized timezones from ISO 8601 alike timestamps
1754 m
= re
.search(r
'\d{1,2}:\d{1,2}(?:\.\d+)?(?P<tz>\s*[A-Z]+)$', date_str
)
1756 date_str
= date_str
[:-len(m
.group('tz'))]
1758 # Python only supports microseconds, so remove nanoseconds
1759 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]+$', date_str
)
1761 date_str
= m
.group(1)
1763 for expression
in date_formats(day_first
):
1764 with contextlib
.suppress(ValueError):
1765 dt
= datetime
.datetime
.strptime(date_str
, expression
) - timezone
+ datetime
.timedelta(hours
=pm_delta
)
1766 return calendar
.timegm(dt
.timetuple())
1767 timetuple
= email
.utils
.parsedate_tz(date_str
)
1769 return calendar
.timegm(timetuple
) + pm_delta
* 3600
1772 def determine_ext(url
, default_ext
='unknown_video'):
1773 if url
is None or '.' not in url
:
1775 guess
= url
.partition('?')[0].rpartition('.')[2]
1776 if re
.match(r
'^[A-Za-z0-9]+$', guess
):
1778 # Try extract ext from URLs like http://example.com/foo/bar.mp4/?download
1779 elif guess
.rstrip('/') in KNOWN_EXTENSIONS
:
1780 return guess
.rstrip('/')
1785 def subtitles_filename(filename
, sub_lang
, sub_format
, expected_real_ext
=None):
1786 return replace_extension(filename
, sub_lang
+ '.' + sub_format
, expected_real_ext
)
1789 def datetime_from_str(date_str
, precision
='auto', format
='%Y%m%d'):
1791 Return a datetime object from a string.
1793 (now|today|yesterday|DATE)([+-]\d+(microsecond|second|minute|hour|day|week|month|year)s?)?
1795 @param format strftime format of DATE
1796 @param precision Round the datetime object: auto|microsecond|second|minute|hour|day
1797 auto: round to the unit provided in date_str (if applicable).
1799 auto_precision
= False
1800 if precision
== 'auto':
1801 auto_precision
= True
1802 precision
= 'microsecond'
1803 today
= datetime_round(datetime
.datetime
.utcnow(), precision
)
1804 if date_str
in ('now', 'today'):
1806 if date_str
== 'yesterday':
1807 return today
- datetime
.timedelta(days
=1)
1809 r
'(?P<start>.+)(?P<sign>[+-])(?P<time>\d+)(?P<unit>microsecond|second|minute|hour|day|week|month|year)s?',
1811 if match
is not None:
1812 start_time
= datetime_from_str(match
.group('start'), precision
, format
)
1813 time
= int(match
.group('time')) * (-1 if match
.group('sign') == '-' else 1)
1814 unit
= match
.group('unit')
1815 if unit
== 'month' or unit
== 'year':
1816 new_date
= datetime_add_months(start_time
, time
* 12 if unit
== 'year' else time
)
1822 delta
= datetime
.timedelta(**{unit + 's': time}
)
1823 new_date
= start_time
+ delta
1825 return datetime_round(new_date
, unit
)
1828 return datetime_round(datetime
.datetime
.strptime(date_str
, format
), precision
)
1831 def date_from_str(date_str
, format
='%Y%m%d', strict
=False):
1833 Return a date object from a string using datetime_from_str
1835 @param strict Restrict allowed patterns to "YYYYMMDD" and
1836 (now|today|yesterday)(-\d+(day|week|month|year)s?)?
1838 if strict
and not re
.fullmatch(r
'\d{8}|(now|today|yesterday)(-\d+(day|week|month|year)s?)?', date_str
):
1839 raise ValueError(f
'Invalid date format "{date_str}"')
1840 return datetime_from_str(date_str
, precision
='microsecond', format
=format
).date()
1843 def datetime_add_months(dt
, months
):
1844 """Increment/Decrement a datetime object by months."""
1845 month
= dt
.month
+ months
- 1
1846 year
= dt
.year
+ month
// 12
1847 month
= month
% 12 + 1
1848 day
= min(dt
.day
, calendar
.monthrange(year
, month
)[1])
1849 return dt
.replace(year
, month
, day
)
1852 def datetime_round(dt
, precision
='day'):
1854 Round a datetime object's time to a specific precision
1856 if precision
== 'microsecond':
1865 roundto
= lambda x
, n
: ((x
+ n
/ 2) // n
) * n
1866 timestamp
= calendar
.timegm(dt
.timetuple())
1867 return datetime
.datetime
.utcfromtimestamp(roundto(timestamp
, unit_seconds
[precision
]))
1870 def hyphenate_date(date_str
):
1872 Convert a date in 'YYYYMMDD' format to 'YYYY-MM-DD' format"""
1873 match
= re
.match(r
'^(\d\d\d\d)(\d\d)(\d\d)$', date_str
)
1874 if match
is not None:
1875 return '-'.join(match
.groups())
1881 """Represents a time interval between two dates"""
1883 def __init__(self
, start
=None, end
=None):
1884 """start and end must be strings in the format accepted by date"""
1885 if start
is not None:
1886 self
.start
= date_from_str(start
, strict
=True)
1888 self
.start
= datetime
.datetime
.min.date()
1890 self
.end
= date_from_str(end
, strict
=True)
1892 self
.end
= datetime
.datetime
.max.date()
1893 if self
.start
> self
.end
:
1894 raise ValueError('Date range: "%s" , the start date must be before the end date' % self
)
1898 """Returns a range that only contains the given day"""
1899 return cls(day
, day
)
1901 def __contains__(self
, date
):
1902 """Check if the date is in the range"""
1903 if not isinstance(date
, datetime
.date
):
1904 date
= date_from_str(date
)
1905 return self
.start
<= date
<= self
.end
1908 return f
'{self.start.isoformat()} - {self.end.isoformat()}'
1911 def platform_name():
1912 """ Returns the platform name as a str """
1913 write_string('DeprecationWarning: yt_dlp.utils.platform_name is deprecated, use platform.platform instead')
1914 return platform
.platform()
1918 def system_identifier():
1919 python_implementation
= platform
.python_implementation()
1920 if python_implementation
== 'PyPy' and hasattr(sys
, 'pypy_version_info'):
1921 python_implementation
+= ' version %d.%d.%d' % sys
.pypy_version_info
[:3]
1923 return 'Python %s (%s %s) - %s %s' % (
1924 platform
.python_version(),
1925 python_implementation
,
1926 platform
.architecture()[0],
1927 platform
.platform(),
1928 format_field(join_nonempty(*platform
.libc_ver(), delim
=' '), None, '(%s)'),
1933 def get_windows_version():
1934 ''' Get Windows version. returns () if it's not running on Windows '''
1935 if compat_os_name
== 'nt':
1936 return version_tuple(platform
.win32_ver()[1])
1941 def write_string(s
, out
=None, encoding
=None):
1942 assert isinstance(s
, str)
1943 out
= out
or sys
.stderr
1945 if compat_os_name
== 'nt' and supports_terminal_sequences(out
):
1946 s
= re
.sub(r
'([\r\n]+)', r
' \1', s
)
1948 enc
, buffer = None, out
1949 if 'b' in getattr(out
, 'mode', ''):
1950 enc
= encoding
or preferredencoding()
1951 elif hasattr(out
, 'buffer'):
1953 enc
= encoding
or getattr(out
, 'encoding', None) or preferredencoding()
1955 buffer.write(s
.encode(enc
, 'ignore') if enc
else s
)
1959 def bytes_to_intlist(bs
):
1962 if isinstance(bs
[0], int): # Python 3
1965 return [ord(c
) for c
in bs
]
1968 def intlist_to_bytes(xs
):
1971 return struct
.pack('%dB' % len(xs
), *xs
)
1974 class LockingUnsupportedError(OSError):
1975 msg
= 'File locking is not supported'
1978 super().__init
__(self
.msg
)
1981 # Cross-platform file locking
1982 if sys
.platform
== 'win32':
1983 import ctypes
.wintypes
1986 class OVERLAPPED(ctypes
.Structure
):
1988 ('Internal', ctypes
.wintypes
.LPVOID
),
1989 ('InternalHigh', ctypes
.wintypes
.LPVOID
),
1990 ('Offset', ctypes
.wintypes
.DWORD
),
1991 ('OffsetHigh', ctypes
.wintypes
.DWORD
),
1992 ('hEvent', ctypes
.wintypes
.HANDLE
),
1995 kernel32
= ctypes
.windll
.kernel32
1996 LockFileEx
= kernel32
.LockFileEx
1997 LockFileEx
.argtypes
= [
1998 ctypes
.wintypes
.HANDLE
, # hFile
1999 ctypes
.wintypes
.DWORD
, # dwFlags
2000 ctypes
.wintypes
.DWORD
, # dwReserved
2001 ctypes
.wintypes
.DWORD
, # nNumberOfBytesToLockLow
2002 ctypes
.wintypes
.DWORD
, # nNumberOfBytesToLockHigh
2003 ctypes
.POINTER(OVERLAPPED
) # Overlapped
2005 LockFileEx
.restype
= ctypes
.wintypes
.BOOL
2006 UnlockFileEx
= kernel32
.UnlockFileEx
2007 UnlockFileEx
.argtypes
= [
2008 ctypes
.wintypes
.HANDLE
, # hFile
2009 ctypes
.wintypes
.DWORD
, # dwReserved
2010 ctypes
.wintypes
.DWORD
, # nNumberOfBytesToLockLow
2011 ctypes
.wintypes
.DWORD
, # nNumberOfBytesToLockHigh
2012 ctypes
.POINTER(OVERLAPPED
) # Overlapped
2014 UnlockFileEx
.restype
= ctypes
.wintypes
.BOOL
2015 whole_low
= 0xffffffff
2016 whole_high
= 0x7fffffff
2018 def _lock_file(f
, exclusive
, block
):
2019 overlapped
= OVERLAPPED()
2020 overlapped
.Offset
= 0
2021 overlapped
.OffsetHigh
= 0
2022 overlapped
.hEvent
= 0
2023 f
._lock
_file
_overlapped
_p
= ctypes
.pointer(overlapped
)
2025 if not LockFileEx(msvcrt
.get_osfhandle(f
.fileno()),
2026 (0x2 if exclusive
else 0x0) |
(0x0 if block
else 0x1),
2027 0, whole_low
, whole_high
, f
._lock
_file
_overlapped
_p
):
2028 # NB: No argument form of "ctypes.FormatError" does not work on PyPy
2029 raise BlockingIOError(f
'Locking file failed: {ctypes.FormatError(ctypes.GetLastError())!r}')
2031 def _unlock_file(f
):
2032 assert f
._lock
_file
_overlapped
_p
2033 handle
= msvcrt
.get_osfhandle(f
.fileno())
2034 if not UnlockFileEx(handle
, 0, whole_low
, whole_high
, f
._lock
_file
_overlapped
_p
):
2035 raise OSError('Unlocking file failed: %r' % ctypes
.FormatError())
2041 def _lock_file(f
, exclusive
, block
):
2042 flags
= fcntl
.LOCK_EX
if exclusive
else fcntl
.LOCK_SH
2044 flags |
= fcntl
.LOCK_NB
2046 fcntl
.flock(f
, flags
)
2047 except BlockingIOError
:
2049 except OSError: # AOSP does not have flock()
2050 fcntl
.lockf(f
, flags
)
2052 def _unlock_file(f
):
2054 fcntl
.flock(f
, fcntl
.LOCK_UN
)
2056 fcntl
.lockf(f
, fcntl
.LOCK_UN
)
2060 def _lock_file(f
, exclusive
, block
):
2061 raise LockingUnsupportedError()
2063 def _unlock_file(f
):
2064 raise LockingUnsupportedError()
2070 def __init__(self
, filename
, mode
, block
=True, encoding
=None):
2071 if mode
not in {'r', 'rb', 'a', 'ab', 'w', 'wb'}
:
2072 raise NotImplementedError(mode
)
2073 self
.mode
, self
.block
= mode
, block
2075 writable
= any(f
in mode
for f
in 'wax+')
2076 readable
= any(f
in mode
for f
in 'r+')
2077 flags
= functools
.reduce(operator
.ior
, (
2078 getattr(os
, 'O_CLOEXEC', 0), # UNIX only
2079 getattr(os
, 'O_BINARY', 0), # Windows only
2080 getattr(os
, 'O_NOINHERIT', 0), # Windows only
2081 os
.O_CREAT
if writable
else 0, # O_TRUNC only after locking
2082 os
.O_APPEND
if 'a' in mode
else 0,
2083 os
.O_EXCL
if 'x' in mode
else 0,
2084 os
.O_RDONLY
if not writable
else os
.O_RDWR
if readable
else os
.O_WRONLY
,
2087 self
.f
= os
.fdopen(os
.open(filename
, flags
, 0o666), mode
, encoding
=encoding
)
2089 def __enter__(self
):
2090 exclusive
= 'r' not in self
.mode
2092 _lock_file(self
.f
, exclusive
, self
.block
)
2097 if 'w' in self
.mode
:
2100 except OSError as e
:
2102 errno
.ESPIPE
, # Illegal seek - expected for FIFO
2103 errno
.EINVAL
, # Invalid argument - expected for /dev/null
2112 _unlock_file(self
.f
)
2116 def __exit__(self
, *_
):
2125 def __getattr__(self
, attr
):
2126 return getattr(self
.f
, attr
)
2133 def get_filesystem_encoding():
2134 encoding
= sys
.getfilesystemencoding()
2135 return encoding
if encoding
is not None else 'utf-8'
2138 def shell_quote(args
):
2140 encoding
= get_filesystem_encoding()
2142 if isinstance(a
, bytes):
2143 # We may get a filename encoded with 'encodeFilename'
2144 a
= a
.decode(encoding
)
2145 quoted_args
.append(compat_shlex_quote(a
))
2146 return ' '.join(quoted_args
)
2149 def smuggle_url(url
, data
):
2150 """ Pass additional data in a URL for internal use. """
2152 url
, idata
= unsmuggle_url(url
, {})
2154 sdata
= urllib
.parse
.urlencode(
2155 {'__youtubedl_smuggle': json.dumps(data)}
)
2156 return url
+ '#' + sdata
2159 def unsmuggle_url(smug_url
, default
=None):
2160 if '#__youtubedl_smuggle' not in smug_url
:
2161 return smug_url
, default
2162 url
, _
, sdata
= smug_url
.rpartition('#')
2163 jsond
= urllib
.parse
.parse_qs(sdata
)['__youtubedl_smuggle'][0]
2164 data
= json
.loads(jsond
)
2168 def format_decimal_suffix(num
, fmt
='%d%s', *, factor
=1000):
2169 """ Formats numbers with decimal sufixes like K, M, etc """
2170 num
, factor
= float_or_none(num
), float(factor
)
2171 if num
is None or num
< 0:
2173 POSSIBLE_SUFFIXES
= 'kMGTPEZY'
2174 exponent
= 0 if num
== 0 else min(int(math
.log(num
, factor
)), len(POSSIBLE_SUFFIXES
))
2175 suffix
= ['', *POSSIBLE_SUFFIXES
][exponent
]
2177 suffix
= {'k': 'Ki', '': ''}
.get(suffix
, f
'{suffix}i')
2178 converted
= num
/ (factor
** exponent
)
2179 return fmt
% (converted
, suffix
)
2182 def format_bytes(bytes):
2183 return format_decimal_suffix(bytes, '%.2f%sB', factor
=1024) or 'N/A'
2186 def lookup_unit_table(unit_table
, s
):
2187 units_re
= '|'.join(re
.escape(u
) for u
in unit_table
)
2189 r
'(?P<num>[0-9]+(?:[,.][0-9]*)?)\s*(?P<unit>%s)\b' % units_re
, s
)
2192 num_str
= m
.group('num').replace(',', '.')
2193 mult
= unit_table
[m
.group('unit')]
2194 return int(float(num_str
) * mult
)
2197 def parse_filesize(s
):
2201 # The lower-case forms are of course incorrect and unofficial,
2202 # but we support those too
2219 'megabytes': 1000 ** 2,
2220 'mebibytes': 1024 ** 2,
2226 'gigabytes': 1000 ** 3,
2227 'gibibytes': 1024 ** 3,
2233 'terabytes': 1000 ** 4,
2234 'tebibytes': 1024 ** 4,
2240 'petabytes': 1000 ** 5,
2241 'pebibytes': 1024 ** 5,
2247 'exabytes': 1000 ** 6,
2248 'exbibytes': 1024 ** 6,
2254 'zettabytes': 1000 ** 7,
2255 'zebibytes': 1024 ** 7,
2261 'yottabytes': 1000 ** 8,
2262 'yobibytes': 1024 ** 8,
2265 return lookup_unit_table(_UNIT_TABLE
, s
)
2272 s
= re
.sub(r
'^[^\d]+\s', '', s
).strip()
2274 if re
.match(r
'^[\d,.]+$', s
):
2275 return str_to_int(s
)
2288 ret
= lookup_unit_table(_UNIT_TABLE
, s
)
2292 mobj
= re
.match(r
'([\d,.]+)(?:$|\s)', s
)
2294 return str_to_int(mobj
.group(1))
2297 def parse_resolution(s
, *, lenient
=False):
2302 mobj
= re
.search(r
'(?P<w>\d+)\s*[xX×,]\s*(?P<h>\d+)', s
)
2304 mobj
= re
.search(r
'(?<![a-zA-Z0-9])(?P<w>\d+)\s*[xX×,]\s*(?P<h>\d+)(?![a-zA-Z0-9])', s
)
2307 'width': int(mobj
.group('w')),
2308 'height': int(mobj
.group('h')),
2311 mobj
= re
.search(r
'(?<![a-zA-Z0-9])(\d+)[pPiI](?![a-zA-Z0-9])', s
)
2313 return {'height': int(mobj.group(1))}
2315 mobj
= re
.search(r
'\b([48])[kK]\b', s
)
2317 return {'height': int(mobj.group(1)) * 540}
2322 def parse_bitrate(s
):
2323 if not isinstance(s
, str):
2325 mobj
= re
.search(r
'\b(\d+)\s*kbps', s
)
2327 return int(mobj
.group(1))
2330 def month_by_name(name
, lang
='en'):
2331 """ Return the number of a month by (locale-independently) English name """
2333 month_names
= MONTH_NAMES
.get(lang
, MONTH_NAMES
['en'])
2336 return month_names
.index(name
) + 1
2341 def month_by_abbreviation(abbrev
):
2342 """ Return the number of a month by (locale-independently) English
2346 return [s
[:3] for s
in ENGLISH_MONTH_NAMES
].index(abbrev
) + 1
2351 def fix_xml_ampersands(xml_str
):
2352 """Replace all the '&' by '&' in XML"""
2354 r
'&(?!amp;|lt;|gt;|apos;|quot;|#x[0-9a-fA-F]{,4};|#[0-9]{,4};)',
2359 def setproctitle(title
):
2360 assert isinstance(title
, str)
2362 # ctypes in Jython is not complete
2363 # http://bugs.jython.org/issue2148
2364 if sys
.platform
.startswith('java'):
2368 libc
= ctypes
.cdll
.LoadLibrary('libc.so.6')
2372 # LoadLibrary in Windows Python 2.7.13 only expects
2373 # a bytestring, but since unicode_literals turns
2374 # every string into a unicode string, it fails.
2376 title_bytes
= title
.encode()
2377 buf
= ctypes
.create_string_buffer(len(title_bytes
))
2378 buf
.value
= title_bytes
2380 libc
.prctl(15, buf
, 0, 0, 0)
2381 except AttributeError:
2382 return # Strange libc, just skip this
2385 def remove_start(s
, start
):
2386 return s
[len(start
):] if s
is not None and s
.startswith(start
) else s
2389 def remove_end(s
, end
):
2390 return s
[:-len(end
)] if s
is not None and s
.endswith(end
) else s
2393 def remove_quotes(s
):
2394 if s
is None or len(s
) < 2:
2396 for quote
in ('"', "'", ):
2397 if s
[0] == quote
and s
[-1] == quote
:
2402 def get_domain(url
):
2403 return '.'.join(urllib
.parse
.urlparse(url
).netloc
.rsplit('.', 2)[-2:])
2406 def url_basename(url
):
2407 path
= urllib
.parse
.urlparse(url
).path
2408 return path
.strip('/').split('/')[-1]
2412 return re
.match(r
'https?://[^?#&]+/', url
).group()
2415 def urljoin(base
, path
):
2416 if isinstance(path
, bytes):
2417 path
= path
.decode()
2418 if not isinstance(path
, str) or not path
:
2420 if re
.match(r
'^(?:[a-zA-Z][a-zA-Z0-9+-.]*:)?//', path
):
2422 if isinstance(base
, bytes):
2423 base
= base
.decode()
2424 if not isinstance(base
, str) or not re
.match(
2425 r
'^(?:https?:)?//', base
):
2427 return urllib
.parse
.urljoin(base
, path
)
2430 class HEADRequest(urllib
.request
.Request
):
2431 def get_method(self
):
2435 class PUTRequest(urllib
.request
.Request
):
2436 def get_method(self
):
2440 def int_or_none(v
, scale
=1, default
=None, get_attr
=None, invscale
=1):
2441 if get_attr
and v
is not None:
2442 v
= getattr(v
, get_attr
, None)
2444 return int(v
) * invscale
// scale
2445 except (ValueError, TypeError, OverflowError):
2449 def str_or_none(v
, default
=None):
2450 return default
if v
is None else str(v
)
2453 def str_to_int(int_str
):
2454 """ A more relaxed version of int_or_none """
2455 if isinstance(int_str
, int):
2457 elif isinstance(int_str
, str):
2458 int_str
= re
.sub(r
'[,\.\+]', '', int_str
)
2459 return int_or_none(int_str
)
2462 def float_or_none(v
, scale
=1, invscale
=1, default
=None):
2466 return float(v
) * invscale
/ scale
2467 except (ValueError, TypeError):
2471 def bool_or_none(v
, default
=None):
2472 return v
if isinstance(v
, bool) else default
2475 def strip_or_none(v
, default
=None):
2476 return v
.strip() if isinstance(v
, str) else default
2479 def url_or_none(url
):
2480 if not url
or not isinstance(url
, str):
2483 return url
if re
.match(r
'^(?:(?:https?|rt(?:m(?:pt?[es]?|fp)|sp[su]?)|mms|ftps?):)?//', url
) else None
2486 def request_to_url(req
):
2487 if isinstance(req
, urllib
.request
.Request
):
2488 return req
.get_full_url()
2493 def strftime_or_none(timestamp
, date_format
, default
=None):
2494 datetime_object
= None
2496 if isinstance(timestamp
, (int, float)): # unix timestamp
2497 datetime_object
= datetime
.datetime
.utcfromtimestamp(timestamp
)
2498 elif isinstance(timestamp
, str): # assume YYYYMMDD
2499 datetime_object
= datetime
.datetime
.strptime(timestamp
, '%Y%m%d')
2500 return datetime_object
.strftime(date_format
)
2501 except (ValueError, TypeError, AttributeError):
2505 def parse_duration(s
):
2506 if not isinstance(s
, str):
2512 days
, hours
, mins
, secs
, ms
= [None] * 5
2513 m
= re
.match(r
'''(?x)
2515 (?:(?:(?P<days>[0-9]+):)?(?P<hours>[0-9]+):)?(?P<mins>[0-9]+):)?
2516 (?P<secs>(?(before_secs)[0-9]{1,2}|[0-9]+))
2517 (?P<ms>[.:][0-9]+)?Z?$
2520 days
, hours
, mins
, secs
, ms
= m
.group('days', 'hours', 'mins', 'secs', 'ms')
2525 [0-9]+\s*y(?:ears?)?,?\s*
2528 [0-9]+\s*m(?:onths?)?,?\s*
2531 [0-9]+\s*w(?:eeks?)?,?\s*
2534 (?P<days>[0-9]+)\s*d(?:ays?)?,?\s*
2538 (?P<hours>[0-9]+)\s*h(?:ours?)?,?\s*
2541 (?P<mins>[0-9]+)\s*m(?:in(?:ute)?s?)?,?\s*
2544 (?P<secs>[0-9]+)(?P<ms>\.[0-9]+)?\s*s(?:ec(?:ond)?s?)?\s*
2547 days
, hours
, mins
, secs
, ms
= m
.groups()
2549 m
= re
.match(r
'(?i)(?:(?P<hours>[0-9.]+)\s*(?:hours?)|(?P<mins>[0-9.]+)\s*(?:mins?\.?|minutes?)\s*)Z?$', s
)
2551 hours
, mins
= m
.groups()
2556 ms
= ms
.replace(':', '.')
2557 return sum(float(part
or 0) * mult
for part
, mult
in (
2558 (days
, 86400), (hours
, 3600), (mins
, 60), (secs
, 1), (ms
, 1)))
2561 def prepend_extension(filename
, ext
, expected_real_ext
=None):
2562 name
, real_ext
= os
.path
.splitext(filename
)
2564 f
'{name}.{ext}{real_ext}'
2565 if not expected_real_ext
or real_ext
[1:] == expected_real_ext
2566 else f
'{filename}.{ext}')
2569 def replace_extension(filename
, ext
, expected_real_ext
=None):
2570 name
, real_ext
= os
.path
.splitext(filename
)
2571 return '{}.{}'.format(
2572 name
if not expected_real_ext
or real_ext
[1:] == expected_real_ext
else filename
,
2576 def check_executable(exe
, args
=[]):
2577 """ Checks if the given binary is installed somewhere in PATH, and returns its name.
2578 args can be a list of arguments for a short output (like -version) """
2580 Popen
.run([exe
] + args
, stdout
=subprocess
.PIPE
, stderr
=subprocess
.PIPE
)
2586 def _get_exe_version_output(exe
, args
, *, to_screen
=None):
2588 to_screen(f
'Checking exe version: {shell_quote([exe] + args)}')
2590 # STDIN should be redirected too. On UNIX-like systems, ffmpeg triggers
2591 # SIGTTOU if yt-dlp is run in the background.
2592 # See https://github.com/ytdl-org/youtube-dl/issues/955#issuecomment-209789656
2593 stdout
, _
, _
= Popen
.run([encodeArgument(exe
)] + args
, text
=True,
2594 stdin
=subprocess
.PIPE
, stdout
=subprocess
.PIPE
, stderr
=subprocess
.STDOUT
)
2600 def detect_exe_version(output
, version_re
=None, unrecognized
='present'):
2601 assert isinstance(output
, str)
2602 if version_re
is None:
2603 version_re
= r
'version\s+([-0-9._a-zA-Z]+)'
2604 m
= re
.search(version_re
, output
)
2611 def get_exe_version(exe
, args
=['--version'],
2612 version_re
=None, unrecognized
='present'):
2613 """ Returns the version of the specified executable,
2614 or False if the executable is not present """
2615 out
= _get_exe_version_output(exe
, args
)
2616 return detect_exe_version(out
, version_re
, unrecognized
) if out
else False
2619 def frange(start
=0, stop
=None, step
=1):
2622 start
, stop
= 0, start
2623 sign
= [-1, 1][step
> 0] if step
else 0
2624 while sign
* start
< sign
* stop
:
2629 class LazyList(collections
.abc
.Sequence
):
2630 """Lazy immutable list from an iterable
2631 Note that slices of a LazyList are lists and not LazyList"""
2633 class IndexError(IndexError):
2636 def __init__(self
, iterable
, *, reverse
=False, _cache
=None):
2637 self
._iterable
= iter(iterable
)
2638 self
._cache
= [] if _cache
is None else _cache
2639 self
._reversed
= reverse
2643 # We need to consume the entire iterable to iterate in reverse
2644 yield from self
.exhaust()
2646 yield from self
._cache
2647 for item
in self
._iterable
:
2648 self
._cache
.append(item
)
2652 self
._cache
.extend(self
._iterable
)
2653 self
._iterable
= [] # Discard the emptied iterable to make it pickle-able
2657 """Evaluate the entire iterable"""
2658 return self
._exhaust
()[::-1 if self
._reversed
else 1]
2661 def _reverse_index(x
):
2662 return None if x
is None else -(x
+ 1)
2664 def __getitem__(self
, idx
):
2665 if isinstance(idx
, slice):
2667 idx
= slice(self
._reverse
_index
(idx
.start
), self
._reverse
_index
(idx
.stop
), -(idx
.step
or 1))
2668 start
, stop
, step
= idx
.start
, idx
.stop
, idx
.step
or 1
2669 elif isinstance(idx
, int):
2671 idx
= self
._reverse
_index
(idx
)
2672 start
, stop
, step
= idx
, idx
, 0
2674 raise TypeError('indices must be integers or slices')
2675 if ((start
or 0) < 0 or (stop
or 0) < 0
2676 or (start
is None and step
< 0)
2677 or (stop
is None and step
> 0)):
2678 # We need to consume the entire iterable to be able to slice from the end
2679 # Obviously, never use this with infinite iterables
2682 return self
._cache
[idx
]
2683 except IndexError as e
:
2684 raise self
.IndexError(e
) from e
2685 n
= max(start
or 0, stop
or 0) - len(self
._cache
) + 1
2687 self
._cache
.extend(itertools
.islice(self
._iterable
, n
))
2689 return self
._cache
[idx
]
2690 except IndexError as e
:
2691 raise self
.IndexError(e
) from e
2695 self
[-1] if self
._reversed
else self
[0]
2696 except self
.IndexError:
2702 return len(self
._cache
)
2704 def __reversed__(self
):
2705 return type(self
)(self
._iterable
, reverse
=not self
._reversed
, _cache
=self
._cache
)
2708 return type(self
)(self
._iterable
, reverse
=self
._reversed
, _cache
=self
._cache
)
2711 # repr and str should mimic a list. So we exhaust the iterable
2712 return repr(self
.exhaust())
2715 return repr(self
.exhaust())
2720 class IndexError(IndexError):
2724 # This is only useful for tests
2725 return len(self
.getslice())
2727 def __init__(self
, pagefunc
, pagesize
, use_cache
=True):
2728 self
._pagefunc
= pagefunc
2729 self
._pagesize
= pagesize
2730 self
._pagecount
= float('inf')
2731 self
._use
_cache
= use_cache
2734 def getpage(self
, pagenum
):
2735 page_results
= self
._cache
.get(pagenum
)
2736 if page_results
is None:
2737 page_results
= [] if pagenum
> self
._pagecount
else list(self
._pagefunc
(pagenum
))
2739 self
._cache
[pagenum
] = page_results
2742 def getslice(self
, start
=0, end
=None):
2743 return list(self
._getslice
(start
, end
))
2745 def _getslice(self
, start
, end
):
2746 raise NotImplementedError('This method must be implemented by subclasses')
2748 def __getitem__(self
, idx
):
2749 assert self
._use
_cache
, 'Indexing PagedList requires cache'
2750 if not isinstance(idx
, int) or idx
< 0:
2751 raise TypeError('indices must be non-negative integers')
2752 entries
= self
.getslice(idx
, idx
+ 1)
2754 raise self
.IndexError()
2758 class OnDemandPagedList(PagedList
):
2759 """Download pages until a page with less than maximum results"""
2761 def _getslice(self
, start
, end
):
2762 for pagenum
in itertools
.count(start
// self
._pagesize
):
2763 firstid
= pagenum
* self
._pagesize
2764 nextfirstid
= pagenum
* self
._pagesize
+ self
._pagesize
2765 if start
>= nextfirstid
:
2769 start
% self
._pagesize
2770 if firstid
<= start
< nextfirstid
2773 ((end
- 1) % self
._pagesize
) + 1
2774 if (end
is not None and firstid
<= end
<= nextfirstid
)
2778 page_results
= self
.getpage(pagenum
)
2780 self
._pagecount
= pagenum
- 1
2782 if startv
!= 0 or endv
is not None:
2783 page_results
= page_results
[startv
:endv
]
2784 yield from page_results
2786 # A little optimization - if current page is not "full", ie. does
2787 # not contain page_size videos then we can assume that this page
2788 # is the last one - there are no more ids on further pages -
2789 # i.e. no need to query again.
2790 if len(page_results
) + startv
< self
._pagesize
:
2793 # If we got the whole page, but the next page is not interesting,
2794 # break out early as well
2795 if end
== nextfirstid
:
2799 class InAdvancePagedList(PagedList
):
2800 """PagedList with total number of pages known in advance"""
2802 def __init__(self
, pagefunc
, pagecount
, pagesize
):
2803 PagedList
.__init
__(self
, pagefunc
, pagesize
, True)
2804 self
._pagecount
= pagecount
2806 def _getslice(self
, start
, end
):
2807 start_page
= start
// self
._pagesize
2808 end_page
= self
._pagecount
if end
is None else min(self
._pagecount
, end
// self
._pagesize
+ 1)
2809 skip_elems
= start
- start_page
* self
._pagesize
2810 only_more
= None if end
is None else end
- start
2811 for pagenum
in range(start_page
, end_page
):
2812 page_results
= self
.getpage(pagenum
)
2814 page_results
= page_results
[skip_elems
:]
2816 if only_more
is not None:
2817 if len(page_results
) < only_more
:
2818 only_more
-= len(page_results
)
2820 yield from page_results
[:only_more
]
2822 yield from page_results
2825 class PlaylistEntries
:
2826 MissingEntry
= object()
2827 is_exhausted
= False
2829 def __init__(self
, ydl
, info_dict
):
2832 # _entries must be assigned now since infodict can change during iteration
2833 entries
= info_dict
.get('entries')
2835 raise EntryNotInPlaylist('There are no entries')
2836 elif isinstance(entries
, list):
2837 self
.is_exhausted
= True
2839 requested_entries
= info_dict
.get('requested_entries')
2840 self
.is_incomplete
= bool(requested_entries
)
2841 if self
.is_incomplete
:
2842 assert self
.is_exhausted
2843 self
._entries
= [self
.MissingEntry
] * max(requested_entries
)
2844 for i
, entry
in zip(requested_entries
, entries
):
2845 self
._entries
[i
- 1] = entry
2846 elif isinstance(entries
, (list, PagedList
, LazyList
)):
2847 self
._entries
= entries
2849 self
._entries
= LazyList(entries
)
2851 PLAYLIST_ITEMS_RE
= re
.compile(r
'''(?x)
2852 (?P<start>[+-]?\d+)?
2854 (?P<end>[+-]?\d+|inf(?:inite)?)?
2855 (?::(?P<step>[+-]?\d+))?
2859 def parse_playlist_items(cls
, string
):
2860 for segment
in string
.split(','):
2862 raise ValueError('There is two or more consecutive commas')
2863 mobj
= cls
.PLAYLIST_ITEMS_RE
.fullmatch(segment
)
2865 raise ValueError(f
'{segment!r} is not a valid specification')
2866 start
, end
, step
, has_range
= mobj
.group('start', 'end', 'step', 'range')
2867 if int_or_none(step
) == 0:
2868 raise ValueError(f
'Step in {segment!r} cannot be zero')
2869 yield slice(int_or_none(start
), float_or_none(end
), int_or_none(step
)) if has_range
else int(start
)
2871 def get_requested_items(self
):
2872 playlist_items
= self
.ydl
.params
.get('playlist_items')
2873 playlist_start
= self
.ydl
.params
.get('playliststart', 1)
2874 playlist_end
= self
.ydl
.params
.get('playlistend')
2875 # For backwards compatibility, interpret -1 as whole list
2876 if playlist_end
in (-1, None):
2878 if not playlist_items
:
2879 playlist_items
= f
'{playlist_start}:{playlist_end}'
2880 elif playlist_start
!= 1 or playlist_end
:
2881 self
.ydl
.report_warning('Ignoring playliststart and playlistend because playlistitems was given', only_once
=True)
2883 for index
in self
.parse_playlist_items(playlist_items
):
2884 for i
, entry
in self
[index
]:
2889 # TODO: Add auto-generated fields
2890 self
.ydl
._match
_entry
(entry
, incomplete
=True, silent
=True)
2891 except (ExistingVideoReached
, RejectedVideoReached
):
2894 def get_full_count(self
):
2895 if self
.is_exhausted
and not self
.is_incomplete
:
2897 elif isinstance(self
._entries
, InAdvancePagedList
):
2898 if self
._entries
._pagesize
== 1:
2899 return self
._entries
._pagecount
2901 @functools.cached_property
2903 if isinstance(self
._entries
, list):
2906 entry
= self
._entries
[i
]
2908 entry
= self
.MissingEntry
2909 if not self
.is_incomplete
:
2910 raise self
.IndexError()
2911 if entry
is self
.MissingEntry
:
2912 raise EntryNotInPlaylist(f
'Entry {i} cannot be found')
2917 return type(self
.ydl
)._handle
_extraction
_exceptions
(lambda _
, i
: self
._entries
[i
])(self
.ydl
, i
)
2918 except (LazyList
.IndexError, PagedList
.IndexError):
2919 raise self
.IndexError()
2922 def __getitem__(self
, idx
):
2923 if isinstance(idx
, int):
2924 idx
= slice(idx
, idx
)
2926 # NB: PlaylistEntries[1:10] => (0, 1, ... 9)
2927 step
= 1 if idx
.step
is None else idx
.step
2928 if idx
.start
is None:
2929 start
= 0 if step
> 0 else len(self
) - 1
2931 start
= idx
.start
- 1 if idx
.start
>= 0 else len(self
) + idx
.start
2933 # NB: Do not call len(self) when idx == [:]
2934 if idx
.stop
is None:
2935 stop
= 0 if step
< 0 else float('inf')
2937 stop
= idx
.stop
- 1 if idx
.stop
>= 0 else len(self
) + idx
.stop
2938 stop
+= [-1, 1][step
> 0]
2940 for i
in frange(start
, stop
, step
):
2944 entry
= self
._getter
(i
)
2945 except self
.IndexError:
2946 self
.is_exhausted
= True
2953 return len(tuple(self
[:]))
2955 class IndexError(IndexError):
2959 def uppercase_escape(s
):
2960 unicode_escape
= codecs
.getdecoder('unicode_escape')
2962 r
'\\U[0-9a-fA-F]{8}',
2963 lambda m
: unicode_escape(m
.group(0))[0],
2967 def lowercase_escape(s
):
2968 unicode_escape
= codecs
.getdecoder('unicode_escape')
2970 r
'\\u[0-9a-fA-F]{4}',
2971 lambda m
: unicode_escape(m
.group(0))[0],
2975 def escape_rfc3986(s
):
2976 """Escape non-ASCII characters as suggested by RFC 3986"""
2977 return urllib
.parse
.quote(s
, b
"%/;:@&=+$,!~*'()?#[]")
2980 def escape_url(url
):
2981 """Escape URL as suggested by RFC 3986"""
2982 url_parsed
= urllib
.parse
.urlparse(url
)
2983 return url_parsed
._replace
(
2984 netloc
=url_parsed
.netloc
.encode('idna').decode('ascii'),
2985 path
=escape_rfc3986(url_parsed
.path
),
2986 params
=escape_rfc3986(url_parsed
.params
),
2987 query
=escape_rfc3986(url_parsed
.query
),
2988 fragment
=escape_rfc3986(url_parsed
.fragment
)
2993 return urllib
.parse
.parse_qs(urllib
.parse
.urlparse(url
).query
)
2996 def read_batch_urls(batch_fd
):
2998 if not isinstance(url
, str):
2999 url
= url
.decode('utf-8', 'replace')
3000 BOM_UTF8
= ('\xef\xbb\xbf', '\ufeff')
3001 for bom
in BOM_UTF8
:
3002 if url
.startswith(bom
):
3003 url
= url
[len(bom
):]
3005 if not url
or url
.startswith(('#', ';', ']')):
3007 # "#" cannot be stripped out since it is part of the URI
3008 # However, it can be safely stripped out if following a whitespace
3009 return re
.split(r
'\s#', url
, 1)[0].rstrip()
3011 with contextlib
.closing(batch_fd
) as fd
:
3012 return [url
for url
in map(fixup
, fd
) if url
]
3015 def urlencode_postdata(*args
, **kargs
):
3016 return urllib
.parse
.urlencode(*args
, **kargs
).encode('ascii')
3019 def update_url_query(url
, query
):
3022 parsed_url
= urllib
.parse
.urlparse(url
)
3023 qs
= urllib
.parse
.parse_qs(parsed_url
.query
)
3025 return urllib
.parse
.urlunparse(parsed_url
._replace
(
3026 query
=urllib
.parse
.urlencode(qs
, True)))
3029 def update_Request(req
, url
=None, data
=None, headers
=None, query
=None):
3030 req_headers
= req
.headers
.copy()
3031 req_headers
.update(headers
or {})
3032 req_data
= data
or req
.data
3033 req_url
= update_url_query(url
or req
.get_full_url(), query
)
3034 req_get_method
= req
.get_method()
3035 if req_get_method
== 'HEAD':
3036 req_type
= HEADRequest
3037 elif req_get_method
== 'PUT':
3038 req_type
= PUTRequest
3040 req_type
= urllib
.request
.Request
3042 req_url
, data
=req_data
, headers
=req_headers
,
3043 origin_req_host
=req
.origin_req_host
, unverifiable
=req
.unverifiable
)
3044 if hasattr(req
, 'timeout'):
3045 new_req
.timeout
= req
.timeout
3049 def _multipart_encode_impl(data
, boundary
):
3050 content_type
= 'multipart/form-data; boundary=%s' % boundary
3053 for k
, v
in data
.items():
3054 out
+= b
'--' + boundary
.encode('ascii') + b
'\r\n'
3055 if isinstance(k
, str):
3057 if isinstance(v
, str):
3059 # RFC 2047 requires non-ASCII field names to be encoded, while RFC 7578
3060 # suggests sending UTF-8 directly. Firefox sends UTF-8, too
3061 content
= b
'Content-Disposition: form-data; name="' + k
+ b
'"\r\n\r\n' + v
+ b
'\r\n'
3062 if boundary
.encode('ascii') in content
:
3063 raise ValueError('Boundary overlaps with data')
3066 out
+= b
'--' + boundary
.encode('ascii') + b
'--\r\n'
3068 return out
, content_type
3071 def multipart_encode(data
, boundary
=None):
3073 Encode a dict to RFC 7578-compliant form-data
3076 A dict where keys and values can be either Unicode or bytes-like
3079 If specified a Unicode object, it's used as the boundary. Otherwise
3080 a random boundary is generated.
3082 Reference: https://tools.ietf.org/html/rfc7578
3084 has_specified_boundary
= boundary
is not None
3087 if boundary
is None:
3088 boundary
= '---------------' + str(random
.randrange(0x0fffffff, 0xffffffff))
3091 out
, content_type
= _multipart_encode_impl(data
, boundary
)
3094 if has_specified_boundary
:
3098 return out
, content_type
3101 def dict_get(d
, key_or_keys
, default
=None, skip_false_values
=True):
3102 for val
in map(d
.get
, variadic(key_or_keys
)):
3103 if val
is not None and (val
or not skip_false_values
):
3108 def try_call(*funcs
, expected_type
=None, args
=[], kwargs
={}):
3111 val
= f(*args
, **kwargs
)
3112 except (AttributeError, KeyError, TypeError, IndexError, ZeroDivisionError):
3115 if expected_type
is None or isinstance(val
, expected_type
):
3119 def try_get(src
, getter
, expected_type
=None):
3120 return try_call(*variadic(getter
), args
=(src
,), expected_type
=expected_type
)
3123 def filter_dict(dct
, cndn
=lambda _
, v
: v
is not None):
3124 return {k: v for k, v in dct.items() if cndn(k, v)}
3127 def merge_dicts(*dicts
):
3129 for a_dict
in dicts
:
3130 for k
, v
in a_dict
.items():
3131 if (v
is not None and k
not in merged
3132 or isinstance(v
, str) and merged
[k
] == ''):
3137 def encode_compat_str(string
, encoding
=preferredencoding(), errors
='strict'):
3138 return string
if isinstance(string
, str) else str(string
, encoding
, errors
)
3150 TV_PARENTAL_GUIDELINES
= {
3160 def parse_age_limit(s
):
3161 # isinstance(False, int) is True. So type() must be used instead
3162 if type(s
) is int: # noqa: E721
3163 return s
if 0 <= s
<= 21 else None
3164 elif not isinstance(s
, str):
3166 m
= re
.match(r
'^(?P<age>\d{1,2})\+?$', s
)
3168 return int(m
.group('age'))
3171 return US_RATINGS
[s
]
3172 m
= re
.match(r
'^TV[_-]?(%s)$' % '|'.join(k
[3:] for k
in TV_PARENTAL_GUIDELINES
), s
)
3174 return TV_PARENTAL_GUIDELINES
['TV-' + m
.group(1)]
3178 def strip_jsonp(code
):
3181 (?:window\.)?(?P<func_name>[a-zA-Z0-9_.$]*)
3182 (?:\s*&&\s*(?P=func_name))?
3183 \s*\(\s*(?P<callback_data>.*)\);?
3184 \s*?(?://[^\n]*)*$''',
3185 r
'\g<callback_data>', code
)
3188 def js_to_json(code
, vars={}):
3189 # vars is a dict of var, val pairs to substitute
3190 COMMENT_RE
= r
'/\*(?:(?!\*/).)*?\*/|//[^\n]*\n'
3191 SKIP_RE
= fr
'\s*(?:{COMMENT_RE})?\s*'
3193 (fr
'(?s)^(0[xX][0-9a-fA-F]+){SKIP_RE}:?$', 16),
3194 (fr
'(?s)^(0+[0-7]+){SKIP_RE}:?$', 8),
3199 if v
in ('true', 'false', 'null'):
3201 elif v
in ('undefined', 'void 0'):
3203 elif v
.startswith('/*') or v
.startswith('//') or v
.startswith('!') or v
== ',':
3206 if v
[0] in ("'", '"'):
3207 v
= re
.sub(r
'(?s)\\.|"', lambda m
: {
3212 }.get(m
.group(0), m
.group(0)), v
[1:-1])
3214 for regex
, base
in INTEGER_TABLE
:
3215 im
= re
.match(regex
, v
)
3217 i
= int(im
.group(1), base
)
3218 return '"%d":' % i
if v
.endswith(':') else '%d' % i
3225 def create_map(mobj
):
3226 return json
.dumps(dict(json
.loads(js_to_json(mobj
.group(1) or '[]', vars=vars))))
3228 code
= re
.sub(r
'new Date\((".+")\)', r
'\g<1>', code
)
3229 code
= re
.sub(r
'new Map\((\[.*?\])?\)', create_map
, code
)
3231 return re
.sub(r
'''(?sx)
3232 "(?:[^"\\]*(?:\\\\|\\['"nurtbfx/\n]))*[^"\\]*"|
3233 '(?:[^'\\]*(?:\\\\|\\['"nurtbfx/\n]))*[^'\\]*'|
3234 {comment}|,(?={skip}[\]}}])|
3235 void\s0|(?:(?<![0-9])[eE]|[a-df-zA-DF-Z_$])[.a-zA-Z_$0-9]*|
3236 \b(?:0[xX][0-9a-fA-F]+|0+[0-7]+)(?:{skip}:)?|
3239 '''.format(comment
=COMMENT_RE
, skip
=SKIP_RE
), fix_kv
, code
)
3242 def qualities(quality_ids
):
3243 """ Get a numeric quality value out of a list of possible values """
3246 return quality_ids
.index(qid
)
3252 POSTPROCESS_WHEN
= ('pre_process', 'after_filter', 'before_dl', 'post_process', 'after_move', 'after_video', 'playlist')
3256 'default': '%(title)s [%(id)s].%(ext)s',
3257 'chapter': '%(title)s - %(section_number)03d %(section_title)s [%(id)s].%(ext)s',
3263 'description': 'description',
3264 'annotation': 'annotations.xml',
3265 'infojson': 'info.json',
3268 'pl_thumbnail': None,
3269 'pl_description': 'description',
3270 'pl_infojson': 'info.json',
3273 # As of [1] format syntax is:
3274 # %[mapping_key][conversion_flags][minimum_width][.precision][length_modifier]type
3275 # 1. https://docs.python.org/2/library/stdtypes.html#string-formatting
3276 STR_FORMAT_RE_TMPL
= r
'''(?x)
3277 (?<!%)(?P<prefix>(?:%%)*)
3279 (?P<has_key>\((?P<key>{0})\))?
3281 (?P<conversion>[#0\-+ ]+)?
3283 (?P<precision>\.\d+)?
3284 (?P<len_mod>[hlL])? # unused in python
3285 {1} # conversion type
3290 STR_FORMAT_TYPES
= 'diouxXeEfFgGcrs'
3293 def limit_length(s
, length
):
3294 """ Add ellipses to overly long strings """
3299 return s
[:length
- len(ELLIPSES
)] + ELLIPSES
3303 def version_tuple(v
):
3304 return tuple(int(e
) for e
in re
.split(r
'[-.]', v
))
3307 def is_outdated_version(version
, limit
, assume_new
=True):
3309 return not assume_new
3311 return version_tuple(version
) < version_tuple(limit
)
3313 return not assume_new
3316 def ytdl_is_updateable():
3317 """ Returns if yt-dlp can be updated with -U """
3319 from .update
import is_non_updateable
3321 return not is_non_updateable()
3324 def args_to_str(args
):
3325 # Get a short string representation for a subprocess command
3326 return ' '.join(compat_shlex_quote(a
) for a
in args
)
3329 def error_to_compat_str(err
):
3333 def error_to_str(err
):
3334 return f
'{type(err).__name__}: {err}'
3337 def mimetype2ext(mt
):
3341 mt
, _
, params
= mt
.partition(';')
3346 # Per RFC 3003, audio/mpeg can be .mp1, .mp2 or .mp3. Here use .mp3 as
3347 # it's the most popular one
3348 'audio/mpeg': 'mp3',
3349 'audio/x-wav': 'wav',
3351 'audio/wave': 'wav',
3354 ext
= FULL_MAP
.get(mt
)
3360 'smptett+xml': 'tt',
3364 'x-mp4-fragmented': 'mp4',
3365 'x-ms-sami': 'sami',
3368 'x-mpegurl': 'm3u8',
3369 'vnd.apple.mpegurl': 'm3u8',
3373 'vnd.ms-sstr+xml': 'ism',
3377 'filmstrip+json': 'fs',
3381 _
, _
, subtype
= mt
.rpartition('/')
3382 ext
= SUBTYPE_MAP
.get(subtype
.lower())
3393 _
, _
, suffix
= subtype
.partition('+')
3394 ext
= SUFFIX_MAP
.get(suffix
)
3398 return subtype
.replace('+', '.')
3401 def ext2mimetype(ext_or_url
):
3404 if '.' not in ext_or_url
:
3405 ext_or_url
= f
'file.{ext_or_url}'
3406 return mimetypes
.guess_type(ext_or_url
)[0]
3409 def parse_codecs(codecs_str
):
3410 # http://tools.ietf.org/html/rfc6381
3413 split_codecs
= list(filter(None, map(
3414 str.strip
, codecs_str
.strip().strip(',').split(','))))
3415 vcodec
, acodec
, scodec
, hdr
= None, None, None, None
3416 for full_codec
in split_codecs
:
3417 parts
= full_codec
.split('.')
3418 codec
= parts
[0].replace('0', '')
3419 if codec
in ('avc1', 'avc2', 'avc3', 'avc4', 'vp9', 'vp8', 'hev1', 'hev2',
3420 'h263', 'h264', 'mp4v', 'hvc1', 'av1', 'theora', 'dvh1', 'dvhe'):
3422 vcodec
= '.'.join(parts
[:4]) if codec
in ('vp9', 'av1', 'hvc1') else full_codec
3423 if codec
in ('dvh1', 'dvhe'):
3425 elif codec
== 'av1' and len(parts
) > 3 and parts
[3] == '10':
3427 elif full_codec
.replace('0', '').startswith('vp9.2'):
3429 elif codec
in ('flac', 'mp4a', 'opus', 'vorbis', 'mp3', 'aac', 'ac-3', 'ec-3', 'eac3', 'dtsc', 'dtse', 'dtsh', 'dtsl'):
3432 elif codec
in ('stpp', 'wvtt',):
3436 write_string(f
'WARNING: Unknown codec {full_codec}\n')
3437 if vcodec
or acodec
or scodec
:
3439 'vcodec': vcodec
or 'none',
3440 'acodec': acodec
or 'none',
3441 'dynamic_range': hdr
,
3442 **({'scodec': scodec}
if scodec
is not None else {}),
3444 elif len(split_codecs
) == 2:
3446 'vcodec': split_codecs
[0],
3447 'acodec': split_codecs
[1],
3452 def urlhandle_detect_ext(url_handle
):
3453 getheader
= url_handle
.headers
.get
3455 cd
= getheader('Content-Disposition')
3457 m
= re
.match(r
'attachment;\s*filename="(?P<filename>[^"]+)"', cd
)
3459 e
= determine_ext(m
.group('filename'), default_ext
=None)
3463 return mimetype2ext(getheader('Content-Type'))
3466 def encode_data_uri(data
, mime_type
):
3467 return 'data:%s;base64,%s' % (mime_type
, base64
.b64encode(data
).decode('ascii'))
3470 def age_restricted(content_limit
, age_limit
):
3471 """ Returns True iff the content should be blocked """
3473 if age_limit
is None: # No limit set
3475 if content_limit
is None:
3476 return False # Content available for everyone
3477 return age_limit
< content_limit
3480 def is_html(first_bytes
):
3481 """ Detect whether a file contains HTML by examining its first bytes. """
3484 (b
'\xef\xbb\xbf', 'utf-8'),
3485 (b
'\x00\x00\xfe\xff', 'utf-32-be'),
3486 (b
'\xff\xfe\x00\x00', 'utf-32-le'),
3487 (b
'\xff\xfe', 'utf-16-le'),
3488 (b
'\xfe\xff', 'utf-16-be'),
3492 for bom
, enc
in BOMS
:
3493 while first_bytes
.startswith(bom
):
3494 encoding
, first_bytes
= enc
, first_bytes
[len(bom
):]
3496 return re
.match(r
'^\s*<', first_bytes
.decode(encoding
, 'replace'))
3499 def determine_protocol(info_dict
):
3500 protocol
= info_dict
.get('protocol')
3501 if protocol
is not None:
3504 url
= sanitize_url(info_dict
['url'])
3505 if url
.startswith('rtmp'):
3507 elif url
.startswith('mms'):
3509 elif url
.startswith('rtsp'):
3512 ext
= determine_ext(url
)
3518 return urllib
.parse
.urlparse(url
).scheme
3521 def render_table(header_row
, data
, delim
=False, extra_gap
=0, hide_empty
=False):
3522 """ Render a list of rows, each as a list of values.
3523 Text after a \t will be right aligned """
3525 return len(remove_terminal_sequences(string
).replace('\t', ''))
3527 def get_max_lens(table
):
3528 return [max(width(str(v
)) for v
in col
) for col
in zip(*table
)]
3530 def filter_using_list(row
, filterArray
):
3531 return [col
for take
, col
in itertools
.zip_longest(filterArray
, row
, fillvalue
=True) if take
]
3533 max_lens
= get_max_lens(data
) if hide_empty
else []
3534 header_row
= filter_using_list(header_row
, max_lens
)
3535 data
= [filter_using_list(row
, max_lens
) for row
in data
]
3537 table
= [header_row
] + data
3538 max_lens
= get_max_lens(table
)
3541 table
= [header_row
, [delim
* (ml
+ extra_gap
) for ml
in max_lens
]] + data
3542 table
[1][-1] = table
[1][-1][:-extra_gap
* len(delim
)] # Remove extra_gap from end of delimiter
3544 for pos
, text
in enumerate(map(str, row
)):
3546 row
[pos
] = text
.replace('\t', ' ' * (max_lens
[pos
] - width(text
))) + ' ' * extra_gap
3548 row
[pos
] = text
+ ' ' * (max_lens
[pos
] - width(text
) + extra_gap
)
3549 ret
= '\n'.join(''.join(row
).rstrip() for row
in table
)
3553 def _match_one(filter_part
, dct
, incomplete
):
3554 # TODO: Generalize code with YoutubeDL._build_format_filter
3555 STRING_OPERATORS
= {
3556 '*=': operator
.contains
,
3557 '^=': lambda attr
, value
: attr
.startswith(value
),
3558 '$=': lambda attr
, value
: attr
.endswith(value
),
3559 '~=': lambda attr
, value
: re
.search(value
, attr
),
3561 COMPARISON_OPERATORS
= {
3563 '<=': operator
.le
, # "<=" must be defined above "<"
3570 if isinstance(incomplete
, bool):
3571 is_incomplete
= lambda _
: incomplete
3573 is_incomplete
= lambda k
: k
in incomplete
3575 operator_rex
= re
.compile(r
'''(?x)
3577 \s*(?P<negation>!\s*)?(?P<op>%s)(?P<none_inclusive>\s*\?)?\s*
3579 (?P<quote>["\'])(?P
<quotedstrval
>.+?
)(?P
=quote
)|
3582 ''' % '|'.join(map(re.escape, COMPARISON_OPERATORS.keys())))
3583 m = operator_rex.fullmatch(filter_part.strip())
3586 unnegated_op = COMPARISON_OPERATORS[m['op']]
3588 op = lambda attr, value: not unnegated_op(attr, value)
3591 comparison_value = m['quotedstrval'] or m['strval'] or m['intval']
3593 comparison_value = comparison_value.replace(r'\%s' % m['quote'], m['quote'])
3594 actual_value = dct.get(m['key'])
3595 numeric_comparison = None
3596 if isinstance(actual_value, (int, float)):
3597 # If the original field is a string and matching comparisonvalue is
3598 # a number we should respect the origin of the original field
3599 # and process comparison value as a string (see
3600 # https://github.com/ytdl-org/youtube-dl/issues/11082)
3602 numeric_comparison = int(comparison_value)
3604 numeric_comparison = parse_filesize(comparison_value)
3605 if numeric_comparison is None:
3606 numeric_comparison = parse_filesize(f'{comparison_value}B')
3607 if numeric_comparison is None:
3608 numeric_comparison = parse_duration(comparison_value)
3609 if numeric_comparison is not None and m['op'] in STRING_OPERATORS:
3610 raise ValueError('Operator %s only supports string values!' % m['op'])
3611 if actual_value is None:
3612 return is_incomplete(m['key']) or m['none_inclusive']
3613 return op(actual_value, comparison_value if numeric_comparison is None else numeric_comparison)
3616 '': lambda v: (v is True) if isinstance(v, bool) else (v is not None),
3617 '!': lambda v: (v is False) if isinstance(v, bool) else (v is None),
3619 operator_rex = re.compile(r'''(?x
)
3620 (?P
<op
>%s)\s
*(?P
<key
>[a
-z_
]+)
3621 ''' % '|'.join(map(re.escape, UNARY_OPERATORS.keys())))
3622 m = operator_rex.fullmatch(filter_part.strip())
3624 op = UNARY_OPERATORS[m.group('op')]
3625 actual_value = dct.get(m.group('key'))
3626 if is_incomplete(m.group('key')) and actual_value is None:
3628 return op(actual_value)
3630 raise ValueError('Invalid filter part %r' % filter_part)
3633 def match_str(filter_str, dct, incomplete=False):
3634 """ Filter a dictionary with a simple string syntax.
3635 @returns Whether the filter passes
3636 @param incomplete Set of keys that is expected to be missing from dct.
3637 Can be True/False to indicate all/none of the keys may be missing.
3638 All conditions on incomplete keys pass if the key is missing
3641 _match_one(filter_part.replace(r'\&', '&'), dct, incomplete)
3642 for filter_part in re.split(r'(?<!\\)&', filter_str))
3645 def match_filter_func(filters):
3648 filters = set(variadic(filters))
3650 interactive = '-' in filters
3654 def _match_func(info_dict, incomplete=False):
3655 if not filters or any(match_str(f, info_dict, incomplete) for f in filters):
3656 return NO_DEFAULT if interactive and not incomplete else None
3658 video_title = info_dict.get('title') or info_dict.get('id') or 'video'
3659 filter_str = ') | ('.join(map(str.strip, filters))
3660 return f'{video_title} does not pass filter ({filter_str}), skipping ..'
3664 def download_range_func(chapters, ranges):
3665 def inner(info_dict, ydl):
3666 warning = ('There are no chapters matching the regex' if info_dict.get('chapters')
3667 else 'Cannot match chapters since chapter information is unavailable')
3668 for regex in chapters or []:
3669 for i, chapter in enumerate(info_dict.get('chapters') or []):
3670 if re.search(regex, chapter['title']):
3672 yield {**chapter, 'index': i}
3673 if chapters and warning:
3674 ydl.to_screen(f'[info] {info_dict["id"]}: {warning}')
3676 yield from ({'start_time': start, 'end_time': end} for start, end in ranges or [])
3681 def parse_dfxp_time_expr(time_expr):
3685 mobj = re.match(rf'^(?P<time_offset>{NUMBER_RE})s?$', time_expr)
3687 return float(mobj.group('time_offset'))
3689 mobj = re.match(r'^(\d+):(\d\d):(\d\d(?:(?:\.|:)\d+)?)$', time_expr)
3691 return 3600 * int(mobj.group(1)) + 60 * int(mobj.group(2)) + float(mobj.group(3).replace(':', '.'))
3694 def srt_subtitles_timecode(seconds):
3695 return '%02d:%02d:%02d,%03d' % timetuple_from_msec(seconds * 1000)
3698 def ass_subtitles_timecode(seconds):
3699 time = timetuple_from_msec(seconds * 1000)
3700 return '%01d:%02d:%02d.%02d' % (*time[:-1], time.milliseconds / 10)
3703 def dfxp2srt(dfxp_data):
3705 @param dfxp_data A
bytes-like
object containing DFXP data
3706 @returns A
unicode object containing converted SRT data
3708 LEGACY_NAMESPACES = (
3709 (b'http://www.w3.org/ns/ttml', [
3710 b'http://www.w3.org/2004/11/ttaf1',
3711 b'http://www.w3.org/2006/04/ttaf1',
3712 b'http://www.w3.org/2006/10/ttaf1',
3714 (b'http://www.w3.org/ns/ttml#styling', [
3715 b'http://www.w3.org/ns/ttml#style',
3719 SUPPORTED_STYLING = [
3728 _x = functools.partial(xpath_with_ns, ns_map={
3729 'xml': 'http://www.w3.org/XML/1998/namespace',
3730 'ttml': 'http://www.w3.org/ns/ttml',
3731 'tts': 'http://www.w3.org/ns/ttml#styling',
3737 class TTMLPElementParser:
3739 _unclosed_elements = []
3740 _applied_styles = []
3742 def start(self, tag, attrib):
3743 if tag in (_x('ttml:br'), 'br'):
3746 unclosed_elements = []
3748 element_style_id = attrib.get('style')
3750 style.update(default_style)
3751 if element_style_id:
3752 style.update(styles.get(element_style_id, {}))
3753 for prop in SUPPORTED_STYLING:
3754 prop_val = attrib.get(_x('tts:' + prop))
3756 style[prop] = prop_val
3759 for k, v in sorted(style.items()):
3760 if self._applied_styles and self._applied_styles[-1].get(k) == v:
3763 font += ' color="%s"' % v
3764 elif k == 'fontSize':
3765 font += ' size="%s"' % v
3766 elif k == 'fontFamily':
3767 font += ' face="%s"' % v
3768 elif k == 'fontWeight' and v == 'bold':
3770 unclosed_elements.append('b')
3771 elif k == 'fontStyle' and v == 'italic':
3773 unclosed_elements.append('i')
3774 elif k == 'textDecoration' and v == 'underline':
3776 unclosed_elements.append('u')
3778 self._out += '<font' + font + '>'
3779 unclosed_elements.append('font')
3781 if self._applied_styles:
3782 applied_style.update(self._applied_styles[-1])
3783 applied_style.update(style)
3784 self._applied_styles.append(applied_style)
3785 self._unclosed_elements.append(unclosed_elements)
3788 if tag not in (_x('ttml:br'), 'br'):
3789 unclosed_elements = self._unclosed_elements.pop()
3790 for element in reversed(unclosed_elements):
3791 self._out += '</%s>' % element
3792 if unclosed_elements and self._applied_styles:
3793 self._applied_styles.pop()
3795 def data(self, data):
3799 return self._out.strip()
3801 def parse_node(node):
3802 target = TTMLPElementParser()
3803 parser = xml.etree.ElementTree.XMLParser(target=target)
3804 parser.feed(xml.etree.ElementTree.tostring(node))
3805 return parser.close()
3807 for k, v in LEGACY_NAMESPACES:
3809 dfxp_data = dfxp_data.replace(ns, k)
3811 dfxp = compat_etree_fromstring(dfxp_data)
3813 paras = dfxp.findall(_x('.//ttml:p')) or dfxp.findall('.//p')
3816 raise ValueError('Invalid dfxp/TTML subtitle')
3820 for style in dfxp.findall(_x('.//ttml:style')):
3821 style_id = style.get('id') or style.get(_x('xml:id'))
3824 parent_style_id = style.get('style')
3826 if parent_style_id not in styles:
3829 styles[style_id] = styles[parent_style_id].copy()
3830 for prop in SUPPORTED_STYLING:
3831 prop_val = style.get(_x('tts:' + prop))
3833 styles.setdefault(style_id, {})[prop] = prop_val
3839 for p in ('body', 'div'):
3840 ele = xpath_element(dfxp, [_x('.//ttml:' + p), './/' + p])
3843 style = styles.get(ele.get('style'))
3846 default_style.update(style)
3848 for para, index in zip(paras, itertools.count(1)):
3849 begin_time = parse_dfxp_time_expr(para.attrib.get('begin'))
3850 end_time = parse_dfxp_time_expr(para.attrib.get('end'))
3851 dur = parse_dfxp_time_expr(para.attrib.get('dur'))
3852 if begin_time is None:
3857 end_time = begin_time + dur
3858 out.append('%d\n%s --> %s\n%s\n\n' % (
3860 srt_subtitles_timecode(begin_time),
3861 srt_subtitles_timecode(end_time),
3867 def cli_option(params, command_option, param, separator=None):
3868 param = params.get(param)
3869 return ([] if param is None
3870 else [command_option, str(param)] if separator is None
3871 else [f'{command_option}{separator}{param}'])
3874 def cli_bool_option(params, command_option, param, true_value='true', false_value='false', separator=None):
3875 param = params.get(param)
3876 assert param in (True, False, None)
3877 return cli_option({True: true_value, False: false_value}, command_option, param, separator)
3880 def cli_valueless_option(params, command_option, param, expected_value=True):
3881 return [command_option] if params.get(param) == expected_value else []
3884 def cli_configuration_args(argdict, keys, default=[], use_compat=True):
3885 if isinstance(argdict, (list, tuple)): # for backward compatibility
3892 assert isinstance(argdict, dict)
3894 assert isinstance(keys, (list, tuple))
3895 for key_list in keys:
3896 arg_list = list(filter(
3897 lambda x: x is not None,
3898 [argdict.get(key.lower()) for key in variadic(key_list)]))
3900 return [arg for args in arg_list for arg in args]
3904 def _configuration_args(main_key, argdict, exe, keys=None, default=[], use_compat=True):
3905 main_key, exe = main_key.lower(), exe.lower()
3906 root_key = exe if main_key == exe else f'{main_key}+{exe}'
3907 keys = [f'{root_key}{k}' for k in (keys or [''])]
3908 if root_key in keys:
3910 keys.append((main_key, exe))
3911 keys.append('default')
3914 return cli_configuration_args(argdict, keys, default, use_compat)
3918 # See http://www.loc.gov/standards/iso639-2/ISO-639-2_utf-8.txt
3977 'iw': 'heb', # Replaced by he in 1989 revision
3987 'in': 'ind', # Replaced by id in 1989 revision
4102 'ji': 'yid', # Replaced by yi in 1989 revision
4110 def short2long(cls, code):
4111 """Convert language code from ISO 639-1 to ISO 639-2/T"""
4112 return cls._lang_map.get(code[:2])
4115 def long2short(cls, code):
4116 """Convert language code from ISO 639-2/T to ISO 639-1"""
4117 for short_name, long_name in cls._lang_map.items():
4118 if long_name == code:
4123 # From http://data.okfn.org/data/core/country-list
4125 'AF': 'Afghanistan',
4126 'AX': 'Åland Islands',
4129 'AS': 'American Samoa',
4134 'AG': 'Antigua and Barbuda',
4151 'BO': 'Bolivia, Plurinational State of',
4152 'BQ': 'Bonaire, Sint Eustatius and Saba',
4153 'BA': 'Bosnia and Herzegovina',
4155 'BV': 'Bouvet Island',
4157 'IO': 'British Indian Ocean Territory',
4158 'BN': 'Brunei Darussalam',
4160 'BF': 'Burkina Faso',
4166 'KY': 'Cayman Islands',
4167 'CF': 'Central African Republic',
4171 'CX': 'Christmas Island',
4172 'CC': 'Cocos (Keeling) Islands',
4176 'CD': 'Congo, the Democratic Republic of the',
4177 'CK': 'Cook Islands',
4179 'CI': 'Côte d\'Ivoire',
4184 'CZ': 'Czech Republic',
4188 'DO': 'Dominican Republic',
4191 'SV': 'El Salvador',
4192 'GQ': 'Equatorial Guinea',
4196 'FK': 'Falkland Islands (Malvinas)',
4197 'FO': 'Faroe Islands',
4201 'GF': 'French Guiana',
4202 'PF': 'French Polynesia',
4203 'TF': 'French Southern Territories',
4218 'GW': 'Guinea-Bissau',
4221 'HM': 'Heard Island and McDonald Islands',
4222 'VA': 'Holy See (Vatican City State)',
4229 'IR': 'Iran, Islamic Republic of',
4232 'IM': 'Isle of Man',
4242 'KP': 'Korea, Democratic People\'s Republic of',
4243 'KR': 'Korea, Republic of',
4246 'LA': 'Lao People\'s Democratic Republic',
4252 'LI': 'Liechtenstein',
4256 'MK': 'Macedonia, the Former Yugoslav Republic of',
4263 'MH': 'Marshall Islands',
4269 'FM': 'Micronesia, Federated States of',
4270 'MD': 'Moldova, Republic of',
4281 'NL': 'Netherlands',
4282 'NC': 'New Caledonia',
4283 'NZ': 'New Zealand',
4288 'NF': 'Norfolk Island',
4289 'MP': 'Northern Mariana Islands',
4294 'PS': 'Palestine, State of',
4296 'PG': 'Papua New Guinea',
4299 'PH': 'Philippines',
4303 'PR': 'Puerto Rico',
4307 'RU': 'Russian Federation',
4309 'BL': 'Saint Barthélemy',
4310 'SH': 'Saint Helena, Ascension and Tristan da Cunha',
4311 'KN': 'Saint Kitts and Nevis',
4312 'LC': 'Saint Lucia',
4313 'MF': 'Saint Martin (French part)',
4314 'PM': 'Saint Pierre and Miquelon',
4315 'VC': 'Saint Vincent and the Grenadines',
4318 'ST': 'Sao Tome and Principe',
4319 'SA': 'Saudi Arabia',
4323 'SL': 'Sierra Leone',
4325 'SX': 'Sint Maarten (Dutch part)',
4328 'SB': 'Solomon Islands',
4330 'ZA': 'South Africa',
4331 'GS': 'South Georgia and the South Sandwich Islands',
4332 'SS': 'South Sudan',
4337 'SJ': 'Svalbard and Jan Mayen',
4340 'CH': 'Switzerland',
4341 'SY': 'Syrian Arab Republic',
4342 'TW': 'Taiwan, Province of China',
4344 'TZ': 'Tanzania, United Republic of',
4346 'TL': 'Timor-Leste',
4350 'TT': 'Trinidad and Tobago',
4353 'TM': 'Turkmenistan',
4354 'TC': 'Turks and Caicos Islands',
4358 'AE': 'United Arab Emirates',
4359 'GB': 'United Kingdom',
4360 'US': 'United States',
4361 'UM': 'United States Minor Outlying Islands',
4365 'VE': 'Venezuela, Bolivarian Republic of',
4367 'VG': 'Virgin Islands, British',
4368 'VI': 'Virgin Islands, U.S.',
4369 'WF': 'Wallis and Futuna',
4370 'EH': 'Western Sahara',
4374 # Not ISO 3166 codes, but used for IP blocks
4375 'AP': 'Asia/Pacific Region',
4380 def short2full(cls, code):
4381 """Convert an ISO 3166-2 country code to the corresponding full name"""
4382 return cls._country_map.get(code.upper())
4386 # Major IPv4 address blocks per country
4388 'AD': '46.172.224.0/19',
4389 'AE': '94.200.0.0/13',
4390 'AF': '149.54.0.0/17',
4391 'AG': '209.59.64.0/18',
4392 'AI': '204.14.248.0/21',
4393 'AL': '46.99.0.0/16',
4394 'AM': '46.70.0.0/15',
4395 'AO': '105.168.0.0/13',
4396 'AP': '182.50.184.0/21',
4397 'AQ': '23.154.160.0/24',
4398 'AR': '181.0.0.0/12',
4399 'AS': '202.70.112.0/20',
4400 'AT': '77.116.0.0/14',
4401 'AU': '1.128.0.0/11',
4402 'AW': '181.41.0.0/18',
4403 'AX': '185.217.4.0/22',
4404 'AZ': '5.197.0.0/16',
4405 'BA': '31.176.128.0/17',
4406 'BB': '65.48.128.0/17',
4407 'BD': '114.130.0.0/16',
4409 'BF': '102.178.0.0/15',
4410 'BG': '95.42.0.0/15',
4411 'BH': '37.131.0.0/17',
4412 'BI': '154.117.192.0/18',
4413 'BJ': '137.255.0.0/16',
4414 'BL': '185.212.72.0/23',
4415 'BM': '196.12.64.0/18',
4416 'BN': '156.31.0.0/16',
4417 'BO': '161.56.0.0/16',
4418 'BQ': '161.0.80.0/20',
4419 'BR': '191.128.0.0/12',
4420 'BS': '24.51.64.0/18',
4421 'BT': '119.2.96.0/19',
4422 'BW': '168.167.0.0/16',
4423 'BY': '178.120.0.0/13',
4424 'BZ': '179.42.192.0/18',
4425 'CA': '99.224.0.0/11',
4426 'CD': '41.243.0.0/16',
4427 'CF': '197.242.176.0/21',
4428 'CG': '160.113.0.0/16',
4429 'CH': '85.0.0.0/13',
4430 'CI': '102.136.0.0/14',
4431 'CK': '202.65.32.0/19',
4432 'CL': '152.172.0.0/14',
4433 'CM': '102.244.0.0/14',
4434 'CN': '36.128.0.0/10',
4435 'CO': '181.240.0.0/12',
4436 'CR': '201.192.0.0/12',
4437 'CU': '152.206.0.0/15',
4438 'CV': '165.90.96.0/19',
4439 'CW': '190.88.128.0/17',
4440 'CY': '31.153.0.0/16',
4441 'CZ': '88.100.0.0/14',
4443 'DJ': '197.241.0.0/17',
4444 'DK': '87.48.0.0/12',
4445 'DM': '192.243.48.0/20',
4446 'DO': '152.166.0.0/15',
4447 'DZ': '41.96.0.0/12',
4448 'EC': '186.68.0.0/15',
4449 'EE': '90.190.0.0/15',
4450 'EG': '156.160.0.0/11',
4451 'ER': '196.200.96.0/20',
4452 'ES': '88.0.0.0/11',
4453 'ET': '196.188.0.0/14',
4454 'EU': '2.16.0.0/13',
4455 'FI': '91.152.0.0/13',
4456 'FJ': '144.120.0.0/16',
4457 'FK': '80.73.208.0/21',
4458 'FM': '119.252.112.0/20',
4459 'FO': '88.85.32.0/19',
4461 'GA': '41.158.0.0/15',
4463 'GD': '74.122.88.0/21',
4464 'GE': '31.146.0.0/16',
4465 'GF': '161.22.64.0/18',
4466 'GG': '62.68.160.0/19',
4467 'GH': '154.160.0.0/12',
4468 'GI': '95.164.0.0/16',
4469 'GL': '88.83.0.0/19',
4470 'GM': '160.182.0.0/15',
4471 'GN': '197.149.192.0/18',
4472 'GP': '104.250.0.0/19',
4473 'GQ': '105.235.224.0/20',
4474 'GR': '94.64.0.0/13',
4475 'GT': '168.234.0.0/16',
4476 'GU': '168.123.0.0/16',
4477 'GW': '197.214.80.0/20',
4478 'GY': '181.41.64.0/18',
4479 'HK': '113.252.0.0/14',
4480 'HN': '181.210.0.0/16',
4481 'HR': '93.136.0.0/13',
4482 'HT': '148.102.128.0/17',
4483 'HU': '84.0.0.0/14',
4484 'ID': '39.192.0.0/10',
4485 'IE': '87.32.0.0/12',
4486 'IL': '79.176.0.0/13',
4487 'IM': '5.62.80.0/20',
4488 'IN': '117.192.0.0/10',
4489 'IO': '203.83.48.0/21',
4490 'IQ': '37.236.0.0/14',
4491 'IR': '2.176.0.0/12',
4492 'IS': '82.221.0.0/16',
4493 'IT': '79.0.0.0/10',
4494 'JE': '87.244.64.0/18',
4495 'JM': '72.27.0.0/17',
4496 'JO': '176.29.0.0/16',
4497 'JP': '133.0.0.0/8',
4498 'KE': '105.48.0.0/12',
4499 'KG': '158.181.128.0/17',
4500 'KH': '36.37.128.0/17',
4501 'KI': '103.25.140.0/22',
4502 'KM': '197.255.224.0/20',
4503 'KN': '198.167.192.0/19',
4504 'KP': '175.45.176.0/22',
4505 'KR': '175.192.0.0/10',
4506 'KW': '37.36.0.0/14',
4507 'KY': '64.96.0.0/15',
4508 'KZ': '2.72.0.0/13',
4509 'LA': '115.84.64.0/18',
4510 'LB': '178.135.0.0/16',
4511 'LC': '24.92.144.0/20',
4512 'LI': '82.117.0.0/19',
4513 'LK': '112.134.0.0/15',
4514 'LR': '102.183.0.0/16',
4515 'LS': '129.232.0.0/17',
4516 'LT': '78.56.0.0/13',
4517 'LU': '188.42.0.0/16',
4518 'LV': '46.109.0.0/16',
4519 'LY': '41.252.0.0/14',
4520 'MA': '105.128.0.0/11',
4521 'MC': '88.209.64.0/18',
4522 'MD': '37.246.0.0/16',
4523 'ME': '178.175.0.0/17',
4524 'MF': '74.112.232.0/21',
4525 'MG': '154.126.0.0/17',
4526 'MH': '117.103.88.0/21',
4527 'MK': '77.28.0.0/15',
4528 'ML': '154.118.128.0/18',
4529 'MM': '37.111.0.0/17',
4530 'MN': '49.0.128.0/17',
4531 'MO': '60.246.0.0/16',
4532 'MP': '202.88.64.0/20',
4533 'MQ': '109.203.224.0/19',
4534 'MR': '41.188.64.0/18',
4535 'MS': '208.90.112.0/22',
4536 'MT': '46.11.0.0/16',
4537 'MU': '105.16.0.0/12',
4538 'MV': '27.114.128.0/18',
4539 'MW': '102.70.0.0/15',
4540 'MX': '187.192.0.0/11',
4541 'MY': '175.136.0.0/13',
4542 'MZ': '197.218.0.0/15',
4543 'NA': '41.182.0.0/16',
4544 'NC': '101.101.0.0/18',
4545 'NE': '197.214.0.0/18',
4546 'NF': '203.17.240.0/22',
4547 'NG': '105.112.0.0/12',
4548 'NI': '186.76.0.0/15',
4549 'NL': '145.96.0.0/11',
4550 'NO': '84.208.0.0/13',
4551 'NP': '36.252.0.0/15',
4552 'NR': '203.98.224.0/19',
4553 'NU': '49.156.48.0/22',
4554 'NZ': '49.224.0.0/14',
4555 'OM': '5.36.0.0/15',
4556 'PA': '186.72.0.0/15',
4557 'PE': '186.160.0.0/14',
4558 'PF': '123.50.64.0/18',
4559 'PG': '124.240.192.0/19',
4560 'PH': '49.144.0.0/13',
4561 'PK': '39.32.0.0/11',
4562 'PL': '83.0.0.0/11',
4563 'PM': '70.36.0.0/20',
4564 'PR': '66.50.0.0/16',
4565 'PS': '188.161.0.0/16',
4566 'PT': '85.240.0.0/13',
4567 'PW': '202.124.224.0/20',
4568 'PY': '181.120.0.0/14',
4569 'QA': '37.210.0.0/15',
4570 'RE': '102.35.0.0/16',
4571 'RO': '79.112.0.0/13',
4572 'RS': '93.86.0.0/15',
4573 'RU': '5.136.0.0/13',
4574 'RW': '41.186.0.0/16',
4575 'SA': '188.48.0.0/13',
4576 'SB': '202.1.160.0/19',
4577 'SC': '154.192.0.0/11',
4578 'SD': '102.120.0.0/13',
4579 'SE': '78.64.0.0/12',
4580 'SG': '8.128.0.0/10',
4581 'SI': '188.196.0.0/14',
4582 'SK': '78.98.0.0/15',
4583 'SL': '102.143.0.0/17',
4584 'SM': '89.186.32.0/19',
4585 'SN': '41.82.0.0/15',
4586 'SO': '154.115.192.0/18',
4587 'SR': '186.179.128.0/17',
4588 'SS': '105.235.208.0/21',
4589 'ST': '197.159.160.0/19',
4590 'SV': '168.243.0.0/16',
4591 'SX': '190.102.0.0/20',
4593 'SZ': '41.84.224.0/19',
4594 'TC': '65.255.48.0/20',
4595 'TD': '154.68.128.0/19',
4596 'TG': '196.168.0.0/14',
4597 'TH': '171.96.0.0/13',
4598 'TJ': '85.9.128.0/18',
4599 'TK': '27.96.24.0/21',
4600 'TL': '180.189.160.0/20',
4601 'TM': '95.85.96.0/19',
4602 'TN': '197.0.0.0/11',
4603 'TO': '175.176.144.0/21',
4604 'TR': '78.160.0.0/11',
4605 'TT': '186.44.0.0/15',
4606 'TV': '202.2.96.0/19',
4607 'TW': '120.96.0.0/11',
4608 'TZ': '156.156.0.0/14',
4609 'UA': '37.52.0.0/14',
4610 'UG': '102.80.0.0/13',
4612 'UY': '167.56.0.0/13',
4613 'UZ': '84.54.64.0/18',
4614 'VA': '212.77.0.0/19',
4615 'VC': '207.191.240.0/21',
4616 'VE': '186.88.0.0/13',
4617 'VG': '66.81.192.0/20',
4618 'VI': '146.226.0.0/16',
4619 'VN': '14.160.0.0/11',
4620 'VU': '202.80.32.0/20',
4621 'WF': '117.20.32.0/21',
4622 'WS': '202.4.32.0/19',
4623 'YE': '134.35.0.0/16',
4624 'YT': '41.242.116.0/22',
4625 'ZA': '41.0.0.0/11',
4626 'ZM': '102.144.0.0/13',
4627 'ZW': '102.177.192.0/18',
4631 def random_ipv4(cls, code_or_block):
4632 if len(code_or_block) == 2:
4633 block = cls._country_ip_map.get(code_or_block.upper())
4637 block = code_or_block
4638 addr, preflen = block.split('/')
4639 addr_min = struct.unpack('!L', socket.inet_aton(addr))[0]
4640 addr_max = addr_min | (0xffffffff >> int(preflen))
4641 return str(socket.inet_ntoa(
4642 struct.pack('!L', random.randint(addr_min, addr_max))))
4645 class PerRequestProxyHandler(urllib.request.ProxyHandler):
4646 def __init__(self, proxies=None):
4647 # Set default handlers
4648 for type in ('http', 'https'):
4649 setattr(self, '%s_open' % type,
4650 lambda r, proxy='__noproxy__', type=type, meth=self.proxy_open:
4651 meth(r, proxy, type))
4652 urllib.request.ProxyHandler.__init__(self, proxies)
4654 def proxy_open(self, req, proxy, type):
4655 req_proxy = req.headers.get('Ytdl-request-proxy')
4656 if req_proxy is not None:
4658 del req.headers['Ytdl-request-proxy']
4660 if proxy == '__noproxy__':
4661 return None # No Proxy
4662 if urllib.parse.urlparse(proxy).scheme.lower() in ('socks', 'socks4', 'socks4a', 'socks5'):
4663 req.add_header('Ytdl-socks-proxy', proxy)
4664 # yt-dlp's http/https handlers do wrapping the socket with socks
4666 return urllib.request.ProxyHandler.proxy_open(
4667 self, req, proxy, type)
4670 # Both long_to_bytes and bytes_to_long are adapted from PyCrypto, which is
4671 # released into Public Domain
4672 # https://github.com/dlitz/pycrypto/blob/master/lib/Crypto/Util/number.py#L387
4674 def long_to_bytes(n, blocksize=0):
4675 """long_to_bytes(n:long, blocksize:int) : string
4676 Convert a long integer to a byte string.
4678 If optional blocksize is given and greater than zero, pad the front of the
4679 byte string with binary zeros so that the length is a multiple of
4682 # after much testing, this algorithm was deemed to be the fastest
4686 s = struct.pack('>I', n & 0xffffffff) + s
4688 # strip off leading zeros
4689 for i in range(len(s)):
4690 if s[i] != b'\000'[0]:
4693 # only happens when n == 0
4697 # add back some pad bytes. this could be done more efficiently w.r.t. the
4698 # de-padding being done above, but sigh...
4699 if blocksize > 0 and len(s) % blocksize:
4700 s = (blocksize - len(s) % blocksize) * b'\000' + s
4704 def bytes_to_long(s):
4705 """bytes_to_long(string) : long
4706 Convert a byte string to a long integer.
4708 This is (essentially) the inverse of long_to_bytes().
4713 extra = (4 - length % 4)
4714 s = b'\000' * extra + s
4715 length = length + extra
4716 for i in range(0, length, 4):
4717 acc = (acc << 32) + struct.unpack('>I', s[i:i + 4])[0]
4721 def ohdave_rsa_encrypt(data, exponent, modulus):
4723 Implement OHDave
's RSA algorithm. See http://www.ohdave.com/rsa/
4726 data: data to encrypt, bytes-like object
4727 exponent, modulus: parameter e and N of RSA algorithm, both integer
4728 Output: hex string of encrypted data
4730 Limitation: supports one block encryption only
4733 payload = int(binascii.hexlify(data[::-1]), 16)
4734 encrypted = pow(payload, exponent, modulus)
4735 return '%x' % encrypted
4738 def pkcs1pad(data, length):
4740 Padding input data with PKCS#1 scheme
4742 @param {int[]} data input data
4743 @param {int} length target length
4744 @returns {int[]} padded data
4746 if len(data) > length - 11:
4747 raise ValueError('Input data too
long for PKCS
#1 padding')
4749 pseudo_random
= [random
.randint(0, 254) for _
in range(length
- len(data
) - 3)]
4750 return [0, 2] + pseudo_random
+ [0] + data
4753 def _base_n_table(n
, table
):
4754 if not table
and not n
:
4755 raise ValueError('Either table or n must be specified')
4756 table
= (table
or '0123456789abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ')[:n
]
4759 raise ValueError(f
'base {n} exceeds table length {len(table)}')
4763 def encode_base_n(num
, n
=None, table
=None):
4764 """Convert given int to a base-n string"""
4765 table
= _base_n_table(n
, table
)
4769 result
, base
= '', len(table
)
4771 result
= table
[num
% base
] + result
4776 def decode_base_n(string
, n
=None, table
=None):
4777 """Convert given base-n string to int"""
4778 table
= {char: index for index, char in enumerate(_base_n_table(n, table))}
4779 result
, base
= 0, len(table
)
4781 result
= result
* base
+ table
[char
]
4785 def decode_base(value
, digits
):
4786 write_string('DeprecationWarning: yt_dlp.utils.decode_base is deprecated '
4787 'and may be removed in a future version. Use yt_dlp.decode_base_n instead')
4788 return decode_base_n(value
, table
=digits
)
4791 def decode_packed_codes(code
):
4792 mobj
= re
.search(PACKED_CODES_RE
, code
)
4793 obfuscated_code
, base
, count
, symbols
= mobj
.groups()
4796 symbols
= symbols
.split('|')
4801 base_n_count
= encode_base_n(count
, base
)
4802 symbol_table
[base_n_count
] = symbols
[count
] or base_n_count
4805 r
'\b(\w+)\b', lambda mobj
: symbol_table
[mobj
.group(0)],
4809 def caesar(s
, alphabet
, shift
):
4814 alphabet
[(alphabet
.index(c
) + shift
) % l
] if c
in alphabet
else c
4819 return caesar(s
, r
'''!"#$%&'()*+,-./0123456789:;<=>?@ABCDEFGHIJKLMNOPQRSTUVWXYZ[\]^_`abcdefghijklmnopqrstuvwxyz{|}~''', 47)
4822 def parse_m3u8_attributes(attrib
):
4824 for (key
, val
) in re
.findall(r
'(?P<key>[A-Z0-9-]+)=(?P<val>"[^"]+"|[^",]+)(?:,|$)', attrib
):
4825 if val
.startswith('"'):
4831 def urshift(val
, n
):
4832 return val
>> n
if val
>= 0 else (val
+ 0x100000000) >> n
4835 # Based on png2str() written by @gdkchan and improved by @yokrysty
4836 # Originally posted at https://github.com/ytdl-org/youtube-dl/issues/9706
4837 def decode_png(png_data
):
4838 # Reference: https://www.w3.org/TR/PNG/
4839 header
= png_data
[8:]
4841 if png_data
[:8] != b
'\x89PNG\x0d\x0a\x1a\x0a' or header
[4:8] != b
'IHDR':
4842 raise OSError('Not a valid PNG file.')
4844 int_map
= {1: '>B', 2: '>H', 4: '>I'}
4845 unpack_integer
= lambda x
: struct
.unpack(int_map
[len(x
)], x
)[0]
4850 length
= unpack_integer(header
[:4])
4853 chunk_type
= header
[:4]
4856 chunk_data
= header
[:length
]
4857 header
= header
[length
:]
4859 header
= header
[4:] # Skip CRC
4867 ihdr
= chunks
[0]['data']
4869 width
= unpack_integer(ihdr
[:4])
4870 height
= unpack_integer(ihdr
[4:8])
4874 for chunk
in chunks
:
4875 if chunk
['type'] == b
'IDAT':
4876 idat
+= chunk
['data']
4879 raise OSError('Unable to read PNG data.')
4881 decompressed_data
= bytearray(zlib
.decompress(idat
))
4886 def _get_pixel(idx
):
4891 for y
in range(height
):
4892 basePos
= y
* (1 + stride
)
4893 filter_type
= decompressed_data
[basePos
]
4897 pixels
.append(current_row
)
4899 for x
in range(stride
):
4900 color
= decompressed_data
[1 + basePos
+ x
]
4901 basex
= y
* stride
+ x
4906 left
= _get_pixel(basex
- 3)
4908 up
= _get_pixel(basex
- stride
)
4910 if filter_type
== 1: # Sub
4911 color
= (color
+ left
) & 0xff
4912 elif filter_type
== 2: # Up
4913 color
= (color
+ up
) & 0xff
4914 elif filter_type
== 3: # Average
4915 color
= (color
+ ((left
+ up
) >> 1)) & 0xff
4916 elif filter_type
== 4: # Paeth
4922 c
= _get_pixel(basex
- stride
- 3)
4930 if pa
<= pb
and pa
<= pc
:
4931 color
= (color
+ a
) & 0xff
4933 color
= (color
+ b
) & 0xff
4935 color
= (color
+ c
) & 0xff
4937 current_row
.append(color
)
4939 return width
, height
, pixels
4942 def write_xattr(path
, key
, value
):
4943 # Windows: Write xattrs to NTFS Alternate Data Streams:
4944 # http://en.wikipedia.org/wiki/NTFS#Alternate_data_streams_.28ADS.29
4945 if compat_os_name
== 'nt':
4946 assert ':' not in key
4947 assert os
.path
.exists(path
)
4950 with open(f
'{path}:{key}', 'wb') as f
:
4952 except OSError as e
:
4953 raise XAttrMetadataError(e
.errno
, e
.strerror
)
4956 # UNIX Method 1. Use xattrs/pyxattrs modules
4959 if getattr(xattr
, '_yt_dlp__identifier', None) == 'pyxattr':
4960 # Unicode arguments are not supported in pyxattr until version 0.5.0
4961 # See https://github.com/ytdl-org/youtube-dl/issues/5498
4962 if version_tuple(xattr
.__version
__) >= (0, 5, 0):
4963 setxattr
= xattr
.set
4965 setxattr
= xattr
.setxattr
4969 setxattr(path
, key
, value
)
4970 except OSError as e
:
4971 raise XAttrMetadataError(e
.errno
, e
.strerror
)
4974 # UNIX Method 2. Use setfattr/xattr executables
4975 exe
= ('setfattr' if check_executable('setfattr', ['--version'])
4976 else 'xattr' if check_executable('xattr', ['-h']) else None)
4978 raise XAttrUnavailableError(
4979 'Couldn\'t find a tool to set the xattrs. Install either the python "xattr" or "pyxattr" modules or the '
4980 + ('"xattr" binary' if sys
.platform
!= 'linux' else 'GNU "attr" package (which contains the "setfattr" tool)'))
4982 value
= value
.decode()
4984 _
, stderr
, returncode
= Popen
.run(
4985 [exe
, '-w', key
, value
, path
] if exe
== 'xattr' else [exe
, '-n', key
, '-v', value
, path
],
4986 text
=True, stdout
=subprocess
.PIPE
, stderr
=subprocess
.PIPE
, stdin
=subprocess
.PIPE
)
4987 except OSError as e
:
4988 raise XAttrMetadataError(e
.errno
, e
.strerror
)
4990 raise XAttrMetadataError(returncode
, stderr
)
4993 def random_birthday(year_field
, month_field
, day_field
):
4994 start_date
= datetime
.date(1950, 1, 1)
4995 end_date
= datetime
.date(1995, 12, 31)
4996 offset
= random
.randint(0, (end_date
- start_date
).days
)
4997 random_date
= start_date
+ datetime
.timedelta(offset
)
4999 year_field
: str(random_date
.year
),
5000 month_field
: str(random_date
.month
),
5001 day_field
: str(random_date
.day
),
5005 # Templates for internet shortcut files, which are plain text files.
5006 DOT_URL_LINK_TEMPLATE
= '''\
5011 DOT_WEBLOC_LINK_TEMPLATE
= '''\
5012 <?xml version="1.0" encoding="UTF-8"?>
5013 <!DOCTYPE plist PUBLIC "-//Apple//DTD PLIST 1.0//EN" "http://www.apple.com/DTDs/PropertyList-1.0.dtd">
5014 <plist version="1.0">
5017 \t<string>%(url)s</string>
5022 DOT_DESKTOP_LINK_TEMPLATE
= '''\
5032 'url': DOT_URL_LINK_TEMPLATE
,
5033 'desktop': DOT_DESKTOP_LINK_TEMPLATE
,
5034 'webloc': DOT_WEBLOC_LINK_TEMPLATE
,
5038 def iri_to_uri(iri
):
5040 Converts an IRI (Internationalized Resource Identifier, allowing Unicode characters) to a URI (Uniform Resource Identifier, ASCII-only).
5042 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.
5045 iri_parts
= urllib
.parse
.urlparse(iri
)
5047 if '[' in iri_parts
.netloc
:
5048 raise ValueError('IPv6 URIs are not, yet, supported.')
5049 # Querying `.netloc`, when there's only one bracket, also raises a ValueError.
5051 # 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.
5054 if iri_parts
.username
:
5055 net_location
+= urllib
.parse
.quote(iri_parts
.username
, safe
=r
"!$%&'()*+,~")
5056 if iri_parts
.password
is not None:
5057 net_location
+= ':' + urllib
.parse
.quote(iri_parts
.password
, safe
=r
"!$%&'()*+,~")
5060 net_location
+= iri_parts
.hostname
.encode('idna').decode() # Punycode for Unicode hostnames.
5061 # The 'idna' encoding produces ASCII text.
5062 if iri_parts
.port
is not None and iri_parts
.port
!= 80:
5063 net_location
+= ':' + str(iri_parts
.port
)
5065 return urllib
.parse
.urlunparse(
5069 urllib
.parse
.quote_plus(iri_parts
.path
, safe
=r
"!$%&'()*+,/:;=@|~"),
5071 # Unsure about the `safe` argument, since this is a legacy way of handling parameters.
5072 urllib
.parse
.quote_plus(iri_parts
.params
, safe
=r
"!$%&'()*+,/:;=@|~"),
5074 # Not totally sure about the `safe` argument, since the source does not explicitly mention the query URI component.
5075 urllib
.parse
.quote_plus(iri_parts
.query
, safe
=r
"!$%&'()*+,/:;=?@{|}~"),
5077 urllib
.parse
.quote_plus(iri_parts
.fragment
, safe
=r
"!#$%&'()*+,/:;=?@{|}~")))
5079 # Source for `safe` arguments: https://url.spec.whatwg.org/#percent-encoded-bytes.
5082 def to_high_limit_path(path
):
5083 if sys
.platform
in ['win32', 'cygwin']:
5084 # Work around MAX_PATH limitation on Windows. The maximum allowed length for the individual path segments may still be quite limited.
5085 return '\\\\?\\' + os
.path
.abspath(path
)
5090 def format_field(obj
, field
=None, template
='%s', ignore
=NO_DEFAULT
, default
='', func
=IDENTITY
):
5091 val
= traverse_obj(obj
, *variadic(field
))
5092 if (not val
and val
!= 0) if ignore
is NO_DEFAULT
else val
in variadic(ignore
):
5094 return template
% func(val
)
5097 def clean_podcast_url(url
):
5098 return re
.sub(r
'''(?x)
5102 media\.blubrry\.com| # https://create.blubrry.com/resources/podcast-media-download-statistics/getting-started/
5105 (?:dts|www)\.podtrac\.com/(?:pts/)?redirect\.[0-9a-z]{3,4}| # http://analytics.podtrac.com/how-to-measure
5108 cn\.co| # https://podcorn.com/analytics-prefix/
5109 st\.fm # https://podsights.com/docs/
5114 _HEX_TABLE
= '0123456789abcdef'
5117 def random_uuidv4():
5118 return re
.sub(r
'[xy]', lambda x
: _HEX_TABLE
[random
.randint(0, 15)], 'xxxxxxxx-xxxx-4xxx-yxxx-xxxxxxxxxxxx')
5121 def make_dir(path
, to_screen
=None):
5123 dn
= os
.path
.dirname(path
)
5124 if dn
and not os
.path
.exists(dn
):
5127 except OSError as err
:
5128 if callable(to_screen
) is not None:
5129 to_screen('unable to create directory ' + error_to_compat_str(err
))
5133 def get_executable_path():
5134 from .update
import _get_variant_and_executable_path
5136 return os
.path
.dirname(os
.path
.abspath(_get_variant_and_executable_path()[1]))
5139 def load_plugins(name
, suffix
, namespace
):
5141 with contextlib
.suppress(FileNotFoundError
):
5142 plugins_spec
= importlib
.util
.spec_from_file_location(
5143 name
, os
.path
.join(get_executable_path(), 'ytdlp_plugins', name
, '__init__.py'))
5144 plugins
= importlib
.util
.module_from_spec(plugins_spec
)
5145 sys
.modules
[plugins_spec
.name
] = plugins
5146 plugins_spec
.loader
.exec_module(plugins
)
5147 for name
in dir(plugins
):
5148 if name
in namespace
:
5150 if not name
.endswith(suffix
):
5152 klass
= getattr(plugins
, name
)
5153 classes
[name
] = namespace
[name
] = klass
5158 obj
, *path_list
, default
=None, expected_type
=None, get_all
=True,
5159 casesense
=True, is_user_input
=False, traverse_string
=False):
5160 ''' Traverse nested list/dict/tuple
5161 @param path_list A list of paths which are checked one by one.
5162 Each path is a list of keys where each key is a:
5164 - string: A dictionary key
5165 - int: An index into a list
5166 - tuple: A list of keys all of which will be traversed
5167 - Ellipsis: Fetch all values in the object
5168 - Function: Takes the key and value as arguments
5169 and returns whether the key matches or not
5170 @param default Default value to return
5171 @param expected_type Only accept final value of this type (Can also be any callable)
5172 @param get_all Return all the values obtained from a path or only the first one
5173 @param casesense Whether to consider dictionary keys as case sensitive
5174 @param is_user_input Whether the keys are generated from user input. If True,
5175 strings are converted to int/slice if necessary
5176 @param traverse_string Whether to traverse inside strings. If True, any
5177 non-compatible object will also be converted into a string
5181 _lower
= lambda k
: (k
.lower() if isinstance(k
, str) else k
)
5182 path_list
= (map(_lower
, variadic(path
)) for path
in path_list
)
5184 def _traverse_obj(obj
, path
, _current_depth
=0):
5186 path
= tuple(variadic(path
))
5187 for i
, key
in enumerate(path
):
5188 if None in (key
, obj
):
5190 if isinstance(key
, (list, tuple)):
5191 obj
= [_traverse_obj(obj
, sub_key
, _current_depth
) for sub_key
in key
]
5194 obj
= (obj
.values() if isinstance(obj
, dict)
5195 else obj
if isinstance(obj
, (list, tuple, LazyList
))
5196 else str(obj
) if traverse_string
else [])
5198 depth
= max(depth
, _current_depth
)
5199 return [_traverse_obj(inner_obj
, path
[i
+ 1:], _current_depth
) for inner_obj
in obj
]
5201 if isinstance(obj
, (list, tuple, LazyList
)):
5202 obj
= enumerate(obj
)
5203 elif isinstance(obj
, dict):
5206 if not traverse_string
:
5210 depth
= max(depth
, _current_depth
)
5211 return [_traverse_obj(v
, path
[i
+ 1:], _current_depth
) for k
, v
in obj
if try_call(key
, args
=(k
, v
))]
5212 elif isinstance(obj
, dict) and not (is_user_input
and key
== ':'):
5213 obj
= (obj
.get(key
) if casesense
or (key
in obj
)
5214 else next((v
for k
, v
in obj
.items() if _lower(k
) == key
), None))
5217 key
= (int_or_none(key
) if ':' not in key
5218 else slice(*map(int_or_none
, key
.split(':'))))
5219 if key
== slice(None):
5220 return _traverse_obj(obj
, (..., *path
[i
+ 1:]), _current_depth
)
5221 if not isinstance(key
, (int, slice)):
5223 if not isinstance(obj
, (list, tuple, LazyList
)):
5224 if not traverse_string
:
5233 if isinstance(expected_type
, type):
5234 type_test
= lambda val
: val
if isinstance(val
, expected_type
) else None
5236 type_test
= expected_type
or IDENTITY
5238 for path
in path_list
:
5240 val
= _traverse_obj(obj
, path
)
5243 for _
in range(depth
- 1):
5244 val
= itertools
.chain
.from_iterable(v
for v
in val
if v
is not None)
5245 val
= [v
for v
in map(type_test
, val
) if v
is not None]
5247 return val
if get_all
else val
[0]
5249 val
= type_test(val
)
5255 def traverse_dict(dictn
, keys
, casesense
=True):
5256 write_string('DeprecationWarning: yt_dlp.utils.traverse_dict is deprecated '
5257 'and may be removed in a future version. Use yt_dlp.utils.traverse_obj instead')
5258 return traverse_obj(dictn
, keys
, casesense
=casesense
, is_user_input
=True, traverse_string
=True)
5261 def get_first(obj
, keys
, **kwargs
):
5262 return traverse_obj(obj
, (..., *variadic(keys
)), **kwargs
, get_all
=False)
5265 def variadic(x
, allowed_types
=(str, bytes, dict)):
5266 return x
if isinstance(x
, collections
.abc
.Iterable
) and not isinstance(x
, allowed_types
) else (x
,)
5269 def time_seconds(**kwargs
):
5270 t
= datetime
.datetime
.now(datetime
.timezone(datetime
.timedelta(**kwargs
)))
5271 return t
.timestamp()
5274 # create a JSON Web Signature (jws) with HS256 algorithm
5275 # the resulting format is in JWS Compact Serialization
5276 # implemented following JWT https://www.rfc-editor.org/rfc/rfc7519.html
5277 # implemented following JWS https://www.rfc-editor.org/rfc/rfc7515.html
5278 def jwt_encode_hs256(payload_data
, key
, headers
={}):
5284 header_data
.update(headers
)
5285 header_b64
= base64
.b64encode(json
.dumps(header_data
).encode())
5286 payload_b64
= base64
.b64encode(json
.dumps(payload_data
).encode())
5287 h
= hmac
.new(key
.encode(), header_b64
+ b
'.' + payload_b64
, hashlib
.sha256
)
5288 signature_b64
= base64
.b64encode(h
.digest())
5289 token
= header_b64
+ b
'.' + payload_b64
+ b
'.' + signature_b64
5293 # can be extended in future to verify the signature and parse header and return the algorithm used if it's not HS256
5294 def jwt_decode_hs256(jwt
):
5295 header_b64
, payload_b64
, signature_b64
= jwt
.split('.')
5296 payload_data
= json
.loads(base64
.urlsafe_b64decode(payload_b64
))
5300 WINDOWS_VT_MODE
= False if compat_os_name
== 'nt' else None
5304 def supports_terminal_sequences(stream
):
5305 if compat_os_name
== 'nt':
5306 if not WINDOWS_VT_MODE
:
5308 elif not os
.getenv('TERM'):
5311 return stream
.isatty()
5312 except BaseException
:
5316 def windows_enable_vt_mode(): # TODO: Do this the proper way https://bugs.python.org/issue30075
5317 if get_windows_version() < (10, 0, 10586):
5319 global WINDOWS_VT_MODE
5321 Popen
.run('', shell
=True)
5325 WINDOWS_VT_MODE
= True
5326 supports_terminal_sequences
.cache_clear()
5329 _terminal_sequences_re
= re
.compile('\033\\[[^m]+m')
5332 def remove_terminal_sequences(string
):
5333 return _terminal_sequences_re
.sub('', string
)
5336 def number_of_digits(number
):
5337 return len('%d' % number
)
5340 def join_nonempty(*values
, delim
='-', from_dict
=None):
5341 if from_dict
is not None:
5342 values
= (traverse_obj(from_dict
, variadic(v
)) for v
in values
)
5343 return delim
.join(map(str, filter(None, values
)))
5346 def scale_thumbnails_to_max_format_width(formats
, thumbnails
, url_width_re
):
5348 Find the largest format dimensions in terms of video width and, for each thumbnail:
5349 * Modify the URL: Match the width with the provided regex and replace with the former width
5352 This function is useful with video services that scale the provided thumbnails on demand
5354 _keys
= ('width', 'height')
5355 max_dimensions
= max(
5356 (tuple(format
.get(k
) or 0 for k
in _keys
) for format
in formats
),
5358 if not max_dimensions
[0]:
5362 {'url': re.sub(url_width_re, str(max_dimensions[0]), thumbnail['url'])}
,
5363 dict(zip(_keys
, max_dimensions
)), thumbnail
)
5364 for thumbnail
in thumbnails
5368 def parse_http_range(range):
5369 """ Parse value of "Range" or "Content-Range" HTTP header into tuple. """
5371 return None, None, None
5372 crg
= re
.search(r
'bytes[ =](\d+)-(\d+)?(?:/(\d+))?', range)
5374 return None, None, None
5375 return int(crg
.group(1)), int_or_none(crg
.group(2)), int_or_none(crg
.group(3))
5378 def read_stdin(what
):
5379 eof
= 'Ctrl+Z' if compat_os_name
== 'nt' else 'Ctrl+D'
5380 write_string(f
'Reading {what} from STDIN - EOF ({eof}) to end:\n')
5388 __initialized
= False
5390 def __init__(self
, parser
, label
=None):
5391 self
.parser
, self
.label
= parser
, label
5392 self
._loaded
_paths
, self
.configs
= set(), []
5394 def init(self
, args
=None, filename
=None):
5395 assert not self
.__initialized
5398 location
= os
.path
.realpath(filename
)
5399 directory
= os
.path
.dirname(location
)
5400 if location
in self
._loaded
_paths
:
5402 self
._loaded
_paths
.add(location
)
5404 self
.own_args
, self
.__initialized
= args
, True
5405 opts
, _
= self
.parser
.parse_known_args(args
)
5406 self
.parsed_args
, self
.filename
= args
, filename
5408 for location
in opts
.config_locations
or []:
5410 self
.append_config(shlex
.split(read_stdin('options'), comments
=True), label
='stdin')
5412 location
= os
.path
.join(directory
, expand_path(location
))
5413 if os
.path
.isdir(location
):
5414 location
= os
.path
.join(location
, 'yt-dlp.conf')
5415 if not os
.path
.exists(location
):
5416 self
.parser
.error(f
'config location {location} does not exist')
5417 self
.append_config(self
.read_file(location
), location
)
5421 label
= join_nonempty(
5422 self
.label
, 'config', f
'"{self.filename}"' if self
.filename
else '',
5424 return join_nonempty(
5425 self
.own_args
is not None and f
'{label[0].upper()}{label[1:]}: {self.hide_login_info(self.own_args)}',
5426 *(f
'\n{c}'.replace('\n', '\n| ')[1:] for c
in self
.configs
),
5430 def read_file(filename
, default
=[]):
5432 optionf
= open(filename
)
5434 return default
# silently skip if file is not present
5436 # FIXME: https://github.com/ytdl-org/youtube-dl/commit/dfe5fa49aed02cf36ba9f743b11b0903554b5e56
5437 contents
= optionf
.read()
5438 res
= shlex
.split(contents
, comments
=True)
5439 except Exception as err
:
5440 raise ValueError(f
'Unable to parse "{filename}": {err}')
5446 def hide_login_info(opts
):
5447 PRIVATE_OPTS
= {'-p', '--password', '-u', '--username', '--video-password', '--ap-password', '--ap-username'}
5448 eqre
= re
.compile('^(?P<key>' + ('|'.join(re
.escape(po
) for po
in PRIVATE_OPTS
)) + ')=.+$')
5453 return m
.group('key') + '=PRIVATE'
5457 opts
= list(map(_scrub_eq
, opts
))
5458 for idx
, opt
in enumerate(opts
):
5459 if opt
in PRIVATE_OPTS
and idx
+ 1 < len(opts
):
5460 opts
[idx
+ 1] = 'PRIVATE'
5463 def append_config(self
, *args
, label
=None):
5464 config
= type(self
)(self
.parser
, label
)
5465 config
._loaded
_paths
= self
._loaded
_paths
5466 if config
.init(*args
):
5467 self
.configs
.append(config
)
5471 for config
in reversed(self
.configs
):
5472 yield from config
.all_args
5473 yield from self
.parsed_args
or []
5475 def parse_known_args(self
, **kwargs
):
5476 return self
.parser
.parse_known_args(self
.all_args
, **kwargs
)
5478 def parse_args(self
):
5479 return self
.parser
.parse_args(self
.all_args
)
5482 class WebSocketsWrapper():
5483 """Wraps websockets module to use in non-async scopes"""
5486 def __init__(self
, url
, headers
=None, connect
=True):
5487 self
.loop
= asyncio
.new_event_loop()
5488 # XXX: "loop" is deprecated
5489 self
.conn
= websockets
.connect(
5490 url
, extra_headers
=headers
, ping_interval
=None,
5491 close_timeout
=float('inf'), loop
=self
.loop
, ping_timeout
=float('inf'))
5494 atexit
.register(self
.__exit
__, None, None, None)
5496 def __enter__(self
):
5498 self
.pool
= self
.run_with_loop(self
.conn
.__aenter
__(), self
.loop
)
5501 def send(self
, *args
):
5502 self
.run_with_loop(self
.pool
.send(*args
), self
.loop
)
5504 def recv(self
, *args
):
5505 return self
.run_with_loop(self
.pool
.recv(*args
), self
.loop
)
5507 def __exit__(self
, type, value
, traceback
):
5509 return self
.run_with_loop(self
.conn
.__aexit
__(type, value
, traceback
), self
.loop
)
5512 self
._cancel
_all
_tasks
(self
.loop
)
5514 # taken from https://github.com/python/cpython/blob/3.9/Lib/asyncio/runners.py with modifications
5515 # for contributors: If there's any new library using asyncio needs to be run in non-async, move these function out of this class
5517 def run_with_loop(main
, loop
):
5518 if not asyncio
.iscoroutine(main
):
5519 raise ValueError(f
'a coroutine was expected, got {main!r}')
5522 return loop
.run_until_complete(main
)
5524 loop
.run_until_complete(loop
.shutdown_asyncgens())
5525 if hasattr(loop
, 'shutdown_default_executor'):
5526 loop
.run_until_complete(loop
.shutdown_default_executor())
5529 def _cancel_all_tasks(loop
):
5530 to_cancel
= asyncio
.all_tasks(loop
)
5535 for task
in to_cancel
:
5538 # XXX: "loop" is removed in python 3.10+
5539 loop
.run_until_complete(
5540 asyncio
.gather(*to_cancel
, loop
=loop
, return_exceptions
=True))
5542 for task
in to_cancel
:
5543 if task
.cancelled():
5545 if task
.exception() is not None:
5546 loop
.call_exception_handler({
5547 'message': 'unhandled exception during asyncio.run() shutdown',
5548 'exception': task
.exception(),
5553 def merge_headers(*dicts
):
5554 """Merge dicts of http headers case insensitively, prioritizing the latter ones"""
5555 return {k.title(): v for k, v in itertools.chain.from_iterable(map(dict.items, dicts))}
5558 def cached_method(f
):
5559 """Cache a method"""
5560 signature
= inspect
.signature(f
)
5563 def wrapper(self
, *args
, **kwargs
):
5564 bound_args
= signature
.bind(self
, *args
, **kwargs
)
5565 bound_args
.apply_defaults()
5566 key
= tuple(bound_args
.arguments
.values())
5568 if not hasattr(self
, '__cached_method__cache'):
5569 self
.__cached
_method
__cache
= {}
5570 cache
= self
.__cached
_method
__cache
.setdefault(f
.__name
__, {})
5571 if key
not in cache
:
5572 cache
[key
] = f(self
, *args
, **kwargs
)
5577 class classproperty
:
5578 """property access for class methods"""
5580 def __init__(self
, func
):
5581 functools
.update_wrapper(self
, func
)
5584 def __get__(self
, _
, cls
):
5585 return self
.func(cls
)
5588 class Namespace(types
.SimpleNamespace
):
5589 """Immutable namespace"""
5592 return iter(self
.__dict
__.values())
5596 return self
.__dict
__.items()
5600 has_certifi
= bool(certifi
)
5601 has_websockets
= bool(websockets
)