48 import xml
.etree
.ElementTree
51 from . import traversal
53 from ..compat
import functools
# isort: split
54 from ..compat
import (
55 compat_etree_fromstring
,
57 compat_HTMLParseError
,
61 from ..dependencies
import brotli
, certifi
, websockets
, xattr
62 from ..socks
import ProxyType
, sockssocket
64 __name__
= __name__
.rsplit('.', 1)[0] # Pretend to be the parent module
66 # This is not clearly defined otherwise
67 compiled_regex_type
= type(re
.compile(''))
70 def random_user_agent():
71 _USER_AGENT_TPL
= 'Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/%s Safari/537.36'
112 return _USER_AGENT_TPL
% random
.choice(_CHROME_VERSIONS
)
115 SUPPORTED_ENCODINGS
= [
119 SUPPORTED_ENCODINGS
.append('br')
122 'User-Agent': random_user_agent(),
123 'Accept': 'text/html,application/xhtml+xml,application/xml;q=0.9,*/*;q=0.8',
124 'Accept-Language': 'en-us,en;q=0.5',
125 'Sec-Fetch-Mode': 'navigate',
130 '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',
142 ENGLISH_MONTH_NAMES
= [
143 'January', 'February', 'March', 'April', 'May', 'June',
144 'July', 'August', 'September', 'October', 'November', 'December']
147 'en': ENGLISH_MONTH_NAMES
,
149 'janvier', 'février', 'mars', 'avril', 'mai', 'juin',
150 'juillet', 'août', 'septembre', 'octobre', 'novembre', 'décembre'],
151 # these follow the genitive grammatical case (dopełniacz)
152 # some websites might be using nominative, which will require another month list
153 # https://en.wikibooks.org/wiki/Polish/Noun_cases
154 'pl': ['stycznia', 'lutego', 'marca', 'kwietnia', 'maja', 'czerwca',
155 'lipca', 'sierpnia', 'września', 'października', 'listopada', 'grudnia'],
158 # From https://github.com/python/cpython/blob/3.11/Lib/email/_parseaddr.py#L36-L42
160 'UT': 0, 'UTC': 0, 'GMT': 0, 'Z': 0,
161 'AST': -4, 'ADT': -3, # Atlantic (used in Canada)
162 'EST': -5, 'EDT': -4, # Eastern
163 'CST': -6, 'CDT': -5, # Central
164 'MST': -7, 'MDT': -6, # Mountain
165 'PST': -8, 'PDT': -7 # Pacific
168 # needed for sanitizing filenames in restricted mode
169 ACCENT_CHARS
= dict(zip('ÂÃÄÀÁÅÆÇÈÉÊËÌÍÎÏÐÑÒÓÔÕÖŐØŒÙÚÛÜŰÝÞßàáâãäåæçèéêëìíîïðñòóôõöőøœùúûüűýþÿ',
170 itertools
.chain('AAAAAA', ['AE'], 'CEEEEIIIIDNOOOOOOO', ['OE'], 'UUUUUY', ['TH', 'ss'],
171 'aaaaaa', ['ae'], 'ceeeeiiiionooooooo', ['oe'], 'uuuuuy', ['th'], 'y')))
201 '%Y-%m-%d %H:%M:%S.%f',
202 '%Y-%m-%d %H:%M:%S:%f',
205 '%Y-%m-%dT%H:%M:%SZ',
206 '%Y-%m-%dT%H:%M:%S.%fZ',
207 '%Y-%m-%dT%H:%M:%S.%f0Z',
209 '%Y-%m-%dT%H:%M:%S.%f',
212 '%b %d %Y at %H:%M:%S',
214 '%B %d %Y at %H:%M:%S',
218 DATE_FORMATS_DAY_FIRST
= list(DATE_FORMATS
)
219 DATE_FORMATS_DAY_FIRST
.extend([
230 DATE_FORMATS_MONTH_FIRST
= list(DATE_FORMATS
)
231 DATE_FORMATS_MONTH_FIRST
.extend([
239 PACKED_CODES_RE
= r
"}\('(.+)',(\d+),(\d+),'([^']+)'\.split\('\|'\)"
240 JSON_LD_RE
= r
'(?is)<script[^>]+type=(["\']?
)application
/ld\
+json\
1[^
>]*>\s
*(?P
<json_ld
>{.+?}|\
[.+?\
])\s
*</script
>'
242 NUMBER_RE = r'\d
+(?
:\
.\d
+)?
'
246 def preferredencoding():
247 """Get preferred encoding.
249 Returns the best encoding scheme for the system, based on
250 locale.getpreferredencoding() and some further tweaks.
253 pref = locale.getpreferredencoding()
261 def write_json_file(obj, fn):
262 """ Encode obj as JSON and write it to fn, atomically if possible """
264 tf = tempfile.NamedTemporaryFile(
265 prefix=f'{os.path.basename(fn)}
.', dir=os.path.dirname(fn),
266 suffix='.tmp
', delete=False, mode='w
', encoding='utf
-8')
270 json.dump(obj, tf, ensure_ascii=False)
271 if sys.platform == 'win32
':
272 # Need to remove existing file on Windows, else os.rename raises
273 # WindowsError or FileExistsError.
274 with contextlib.suppress(OSError):
276 with contextlib.suppress(OSError):
279 os.chmod(tf.name, 0o666 & ~mask)
280 os.rename(tf.name, fn)
282 with contextlib.suppress(OSError):
287 def find_xpath_attr(node, xpath, key, val=None):
288 """ Find the xpath xpath[@key=val] """
289 assert re.match(r'^
[a
-zA
-Z_
-]+$
', key)
290 expr = xpath + ('[@%s]' % key if val is None else f"[@{key}='{val}
']")
291 return node.find(expr)
293 # On python2.6 the xml.etree.ElementTree.Element methods don't support
294 # the namespace parameter
297 def xpath_with_ns(path
, ns_map
):
298 components
= [c
.split(':') for c
in path
.split('/')]
302 replaced
.append(c
[0])
305 replaced
.append('{%s}%s' % (ns_map
[ns
], tag
))
306 return '/'.join(replaced
)
309 def xpath_element(node
, xpath
, name
=None, fatal
=False, default
=NO_DEFAULT
):
310 def _find_xpath(xpath
):
311 return node
.find(xpath
)
313 if isinstance(xpath
, str):
314 n
= _find_xpath(xpath
)
322 if default
is not NO_DEFAULT
:
325 name
= xpath
if name
is None else name
326 raise ExtractorError('Could not find XML element %s' % name
)
332 def xpath_text(node
, xpath
, name
=None, fatal
=False, default
=NO_DEFAULT
):
333 n
= xpath_element(node
, xpath
, name
, fatal
=fatal
, default
=default
)
334 if n
is None or n
== default
:
337 if default
is not NO_DEFAULT
:
340 name
= xpath
if name
is None else name
341 raise ExtractorError('Could not find XML element\'s text %s' % name
)
347 def xpath_attr(node
, xpath
, key
, name
=None, fatal
=False, default
=NO_DEFAULT
):
348 n
= find_xpath_attr(node
, xpath
, key
)
350 if default
is not NO_DEFAULT
:
353 name
= f
'{xpath}[@{key}]' if name
is None else name
354 raise ExtractorError('Could not find XML attribute %s' % name
)
360 def get_element_by_id(id, html
, **kwargs
):
361 """Return the content of the tag with the specified ID in the passed HTML document"""
362 return get_element_by_attribute('id', id, html
, **kwargs
)
365 def get_element_html_by_id(id, html
, **kwargs
):
366 """Return the html of the tag with the specified ID in the passed HTML document"""
367 return get_element_html_by_attribute('id', id, html
, **kwargs
)
370 def get_element_by_class(class_name
, html
):
371 """Return the content of the first tag with the specified class in the passed HTML document"""
372 retval
= get_elements_by_class(class_name
, html
)
373 return retval
[0] if retval
else None
376 def get_element_html_by_class(class_name
, html
):
377 """Return the html of the first tag with the specified class in the passed HTML document"""
378 retval
= get_elements_html_by_class(class_name
, html
)
379 return retval
[0] if retval
else None
382 def get_element_by_attribute(attribute
, value
, html
, **kwargs
):
383 retval
= get_elements_by_attribute(attribute
, value
, html
, **kwargs
)
384 return retval
[0] if retval
else None
387 def get_element_html_by_attribute(attribute
, value
, html
, **kargs
):
388 retval
= get_elements_html_by_attribute(attribute
, value
, html
, **kargs
)
389 return retval
[0] if retval
else None
392 def get_elements_by_class(class_name
, html
, **kargs
):
393 """Return the content of all tags with the specified class in the passed HTML document as a list"""
394 return get_elements_by_attribute(
395 'class', r
'[^\'"]*(?<=[\'"\s
])%s(?
=[\'"\s])[^\'"]*' % re.escape(class_name),
396 html, escape_value=False)
399 def get_elements_html_by_class(class_name, html):
400 """Return the html of all tags with the specified class in the passed HTML document as a list"""
401 return get_elements_html_by_attribute(
402 'class', r'[^
\'"]*(?<=[\'"\s
])%s(?
=[\'"\s])[^\'"]*' % re.escape(class_name),
403 html, escape_value=False)
406 def get_elements_by_attribute(*args, **kwargs):
407 """Return the content of the tag with the specified attribute in the passed HTML document"""
408 return [content for content, _ in get_elements_text_and_html_by_attribute(*args, **kwargs)]
411 def get_elements_html_by_attribute(*args, **kwargs):
412 """Return the html of the tag with the specified attribute in the passed HTML document"""
413 return [whole for _, whole in get_elements_text_and_html_by_attribute(*args, **kwargs)]
416 def get_elements_text_and_html_by_attribute(attribute, value, html, *, tag=r'[\w
:.-]+', escape_value=True):
418 Return the text (content) and the html (whole) of the tag with the specified
419 attribute in the passed HTML document
424 quote = '' if re.match(r'''[\s"'`
=<>]''', value) else '?'
426 value = re.escape(value) if escape_value else value
428 partial_element_re = rf'''(?x
)
430 (?
:\
s(?
:[^
>"']|"[^
"]*"|
'[^']*')*)?
431 \s{re.escape(attribute)}\s*=\s*(?P<_q>['"]{quote})(?-x:{value})(?P=_q)
434 for m in re.finditer(partial_element_re, html):
435 content, whole = get_element_text_and_html_by_tag(m.group('tag'), html[m.start():])
438 unescapeHTML(re.sub(r'^(?P<q>["\'])(?P
<content
>.*)(?P
=q
)$
', r'\g
<content
>', content, flags=re.DOTALL)),
443 class HTMLBreakOnClosingTagParser(html.parser.HTMLParser):
445 HTML parser which raises HTMLBreakOnClosingTagException upon reaching the
446 closing tag for the first opening tag it has encountered, and can be used
450 class HTMLBreakOnClosingTagException(Exception):
454 self.tagstack = collections.deque()
455 html.parser.HTMLParser.__init__(self)
460 def __exit__(self, *_):
464 # handle_endtag does not return upon raising HTMLBreakOnClosingTagException,
465 # so data remains buffered; we no longer have any interest in it, thus
466 # override this method to discard it
469 def handle_starttag(self, tag, _):
470 self.tagstack.append(tag)
472 def handle_endtag(self, tag):
473 if not self.tagstack:
474 raise compat_HTMLParseError('no tags
in the stack
')
476 inner_tag = self.tagstack.pop()
480 raise compat_HTMLParseError(f'matching opening tag
for closing {tag} tag
not found
')
481 if not self.tagstack:
482 raise self.HTMLBreakOnClosingTagException()
485 # XXX: This should be far less strict
486 def get_element_text_and_html_by_tag(tag, html):
488 For the first element with the specified tag in the passed HTML document
489 return its' content (text
) and the whole
element (html
)
491 def find_or_raise(haystack, needle, exc):
493 return haystack.index(needle)
496 closing_tag = f'</{tag}>'
497 whole_start = find_or_raise(
498 html, f'<{tag}', compat_HTMLParseError(f'opening {tag} tag not found'))
499 content_start = find_or_raise(
500 html[whole_start:], '>', compat_HTMLParseError(f'malformed opening {tag} tag'))
501 content_start += whole_start + 1
502 with HTMLBreakOnClosingTagParser() as parser:
503 parser.feed(html[whole_start:content_start])
504 if not parser.tagstack or parser.tagstack[0] != tag:
505 raise compat_HTMLParseError(f'parser did not match opening {tag} tag')
506 offset = content_start
507 while offset < len(html):
508 next_closing_tag_start = find_or_raise(
509 html[offset:], closing_tag,
510 compat_HTMLParseError(f'closing {tag} tag not found'))
511 next_closing_tag_end = next_closing_tag_start + len(closing_tag)
513 parser.feed(html[offset:offset + next_closing_tag_end])
514 offset += next_closing_tag_end
515 except HTMLBreakOnClosingTagParser.HTMLBreakOnClosingTagException:
516 return html[content_start:offset + next_closing_tag_start], \
517 html[whole_start:offset + next_closing_tag_end]
518 raise compat_HTMLParseError('unexpected end of html')
521 class HTMLAttributeParser(html.parser.HTMLParser):
522 """Trivial HTML parser to gather the attributes
for a single element
"""
526 html.parser.HTMLParser.__init__(self)
528 def handle_starttag(self, tag, attrs):
529 self.attrs = dict(attrs)
530 raise compat_HTMLParseError('done')
533 class HTMLListAttrsParser(html.parser.HTMLParser):
534 """HTML parser to gather the attributes
for the elements of a
list"""
537 html.parser.HTMLParser.__init__(self)
541 def handle_starttag(self, tag, attrs):
542 if tag == 'li' and self._level == 0:
543 self.items.append(dict(attrs))
546 def handle_endtag(self, tag):
550 def extract_attributes(html_element):
551 """Given a string
for an HTML element such
as
553 a
="foo" B
="bar" c
="&98;az" d
=boz
554 empty
= noval entity
="&"
557 Decode
and return a dictionary of attributes
.
559 'a': 'foo', 'b': 'bar', c
: 'baz', d
: 'boz',
560 'empty': '', 'noval': None, 'entity': '&',
561 'sq': '"', 'dq': '\''
564 parser = HTMLAttributeParser()
565 with contextlib.suppress(compat_HTMLParseError):
566 parser.feed(html_element)
571 def parse_list(webpage):
572 """Given a string
for an series of HTML
<li
> elements
,
573 return a dictionary of their attributes
"""
574 parser = HTMLListAttrsParser()
580 def clean_html(html):
581 """Clean an HTML snippet into a readable string
"""
583 if html is None: # Convenience for sanitizing descriptions etc.
586 html = re.sub(r'\s+', ' ', html)
587 html = re.sub(r'(?u)\s?<\s?br\s?/?\s?>\s?', '\n', html)
588 html = re.sub(r'(?u)<\s?/\s?p\s?>\s?<\s?p[^>]*>', '\n', html)
590 html = re.sub('<.*?>', '', html)
591 # Replace html entities
592 html = unescapeHTML(html)
596 class LenientJSONDecoder(json.JSONDecoder):
598 def __init__(self, *args, transform_source=None, ignore_extra=False, close_objects=0, **kwargs):
599 self.transform_source, self.ignore_extra = transform_source, ignore_extra
600 self._close_attempts = 2 * close_objects
601 super().__init__(*args, **kwargs)
604 def _close_object(err):
605 doc = err.doc[:err.pos]
606 # We need to add comma first to get the correct error message
607 if err.msg.startswith('Expecting \',\''):
609 elif not doc.endswith(','):
612 if err.msg.startswith('Expecting property name'):
613 return doc[:-1] + '}'
614 elif err.msg.startswith('Expecting value'):
615 return doc[:-1] + ']'
618 if self.transform_source:
619 s = self.transform_source(s)
620 for attempt in range(self._close_attempts + 1):
622 if self.ignore_extra:
623 return self.raw_decode(s.lstrip())[0]
624 return super().decode(s)
625 except json.JSONDecodeError as e:
628 elif attempt < self._close_attempts:
629 s = self._close_object(e)
632 raise type(e)(f'{e.msg} in {s[e.pos-10:e.pos+10]!r}', s, e.pos)
633 assert False, 'Too many attempts to decode JSON'
636 def sanitize_open(filename, open_mode):
637 """Try to
open the given filename
, and slightly tweak it
if this fails
.
639 Attempts to
open the given filename
. If this fails
, it tries to change
640 the filename slightly
, step by step
, until it
's either able to open it
641 or it fails and raises a final exception, like the standard open()
644 It returns the tuple (stream, definitive_file_name).
647 if sys.platform == 'win32
':
650 # stdout may be any IO stream, e.g. when using contextlib.redirect_stdout
651 with contextlib.suppress(io.UnsupportedOperation):
652 msvcrt.setmode(sys.stdout.fileno(), os.O_BINARY)
653 return (sys.stdout.buffer if hasattr(sys.stdout, 'buffer') else sys.stdout, filename)
655 for attempt in range(2):
658 if sys.platform == 'win32
':
659 # FIXME: An exclusive lock also locks the file from being read.
660 # Since windows locks are mandatory, don't lock the
file on
windows (for now
).
661 # Ref: https://github.com/yt-dlp/yt-dlp/issues/3124
662 raise LockingUnsupportedError()
663 stream
= locked_file(filename
, open_mode
, block
=False).__enter
__()
665 stream
= open(filename
, open_mode
)
666 return stream
, filename
667 except OSError as err
:
668 if attempt
or err
.errno
in (errno
.EACCES
,):
670 old_filename
, filename
= filename
, sanitize_path(filename
)
671 if old_filename
== filename
:
675 def timeconvert(timestr
):
676 """Convert RFC 2822 defined time string into system timestamp"""
678 timetuple
= email
.utils
.parsedate_tz(timestr
)
679 if timetuple
is not None:
680 timestamp
= email
.utils
.mktime_tz(timetuple
)
684 def sanitize_filename(s
, restricted
=False, is_id
=NO_DEFAULT
):
685 """Sanitizes a string so it could be used as part of a filename.
686 @param restricted Use a stricter subset of allowed characters
687 @param is_id Whether this is an ID that should be kept unchanged if possible.
688 If unset, yt-dlp's new sanitization rules are in effect
693 def replace_insane(char
):
694 if restricted
and char
in ACCENT_CHARS
:
695 return ACCENT_CHARS
[char
]
696 elif not restricted
and char
== '\n':
698 elif is_id
is NO_DEFAULT
and not restricted
and char
in '"*:<>?|/\\':
699 # Replace with their full-width unicode counterparts
700 return {'/': '\u29F8', '\\': '\u29f9'}
.get(char
, chr(ord(char
) + 0xfee0))
701 elif char
== '?' or ord(char
) < 32 or ord(char
) == 127:
704 return '' if restricted
else '\''
706 return '\0_\0-' if restricted
else '\0 \0-'
707 elif char
in '\\/|*<>':
709 if restricted
and (char
in '!&\'()[]{}$;`^,#' or char
.isspace() or ord(char
) > 127):
713 # Replace look-alike Unicode glyphs
714 if restricted
and (is_id
is NO_DEFAULT
or not is_id
):
715 s
= unicodedata
.normalize('NFKC', s
)
716 s
= re
.sub(r
'[0-9]+(?::[0-9]+)+', lambda m
: m
.group(0).replace(':', '_'), s
) # Handle timestamps
717 result
= ''.join(map(replace_insane
, s
))
718 if is_id
is NO_DEFAULT
:
719 result
= re
.sub(r
'(\0.)(?:(?=\1)..)+', r
'\1', result
) # Remove repeated substitute chars
720 STRIP_RE
= r
'(?:\0.|[ _-])*'
721 result
= re
.sub(f
'^\0.{STRIP_RE}|{STRIP_RE}\0.$', '', result
) # Remove substitute chars from start/end
722 result
= result
.replace('\0', '') or '_'
725 while '__' in result
:
726 result
= result
.replace('__', '_')
727 result
= result
.strip('_')
728 # Common case of "Foreign band name - English song title"
729 if restricted
and result
.startswith('-_'):
731 if result
.startswith('-'):
732 result
= '_' + result
[len('-'):]
733 result
= result
.lstrip('.')
739 def sanitize_path(s
, force
=False):
740 """Sanitizes and normalizes path on Windows"""
741 if sys
.platform
== 'win32':
743 drive_or_unc
, _
= os
.path
.splitdrive(s
)
749 norm_path
= os
.path
.normpath(remove_start(s
, drive_or_unc
)).split(os
.path
.sep
)
753 path_part
if path_part
in ['.', '..'] else re
.sub(r
'(?:[/<>:"\|\\?\*]|[\s.]$)', '#', path_part
)
754 for path_part
in norm_path
]
756 sanitized_path
.insert(0, drive_or_unc
+ os
.path
.sep
)
757 elif force
and s
and s
[0] == os
.path
.sep
:
758 sanitized_path
.insert(0, os
.path
.sep
)
759 return os
.path
.join(*sanitized_path
)
762 def sanitize_url(url
, *, scheme
='http'):
763 # Prepend protocol-less URLs with `http:` scheme in order to mitigate
764 # the number of unwanted failures due to missing protocol
767 elif url
.startswith('//'):
768 return f
'{scheme}:{url}'
769 # Fix some common typos seen so far
771 # https://github.com/ytdl-org/youtube-dl/issues/15649
772 (r
'^httpss://', r
'https://'),
773 # https://bx1.be/lives/direct-tv/
774 (r
'^rmtp([es]?)://', r
'rtmp\1://'),
776 for mistake
, fixup
in COMMON_TYPOS
:
777 if re
.match(mistake
, url
):
778 return re
.sub(mistake
, fixup
, url
)
782 def extract_basic_auth(url
):
783 parts
= urllib
.parse
.urlsplit(url
)
784 if parts
.username
is None:
786 url
= urllib
.parse
.urlunsplit(parts
._replace
(netloc
=(
787 parts
.hostname
if parts
.port
is None
788 else '%s:%d' % (parts
.hostname
, parts
.port
))))
789 auth_payload
= base64
.b64encode(
790 ('%s:%s' % (parts
.username
, parts
.password
or '')).encode())
791 return url
, f
'Basic {auth_payload.decode()}'
794 def sanitized_Request(url
, *args
, **kwargs
):
795 url
, auth_header
= extract_basic_auth(escape_url(sanitize_url(url
)))
796 if auth_header
is not None:
797 headers
= args
[1] if len(args
) >= 2 else kwargs
.setdefault('headers', {})
798 headers
['Authorization'] = auth_header
799 return urllib
.request
.Request(url
, *args
, **kwargs
)
803 """Expand shell variables and ~"""
804 return os
.path
.expandvars(compat_expanduser(s
))
807 def orderedSet(iterable
, *, lazy
=False):
808 """Remove all duplicates from the input iterable"""
810 seen
= [] # Do not use set since the items can be unhashable
816 return _iter() if lazy
else list(_iter())
819 def _htmlentity_transform(entity_with_semicolon
):
820 """Transforms an HTML entity to a character."""
821 entity
= entity_with_semicolon
[:-1]
823 # Known non-numeric HTML entity
824 if entity
in html
.entities
.name2codepoint
:
825 return chr(html
.entities
.name2codepoint
[entity
])
827 # TODO: HTML5 allows entities without a semicolon.
828 # E.g. 'Éric' should be decoded as 'Éric'.
829 if entity_with_semicolon
in html
.entities
.html5
:
830 return html
.entities
.html5
[entity_with_semicolon
]
832 mobj
= re
.match(r
'#(x[0-9a-fA-F]+|[0-9]+)', entity
)
834 numstr
= mobj
.group(1)
835 if numstr
.startswith('x'):
837 numstr
= '0%s' % numstr
840 # See https://github.com/ytdl-org/youtube-dl/issues/7518
841 with contextlib
.suppress(ValueError):
842 return chr(int(numstr
, base
))
844 # Unknown entity in name, return its literal representation
845 return '&%s;' % entity
851 assert isinstance(s
, str)
854 r
'&([^&;]+;)', lambda m
: _htmlentity_transform(m
.group(1)), s
)
857 def escapeHTML(text
):
860 .replace('&', '&')
861 .replace('<', '<')
862 .replace('>', '>')
863 .replace('"', '"')
864 .replace("'", ''')
868 class netrc_from_content(netrc
.netrc
):
869 def __init__(self
, content
):
870 self
.hosts
, self
.macros
= {}, {}
871 with io
.StringIO(content
) as stream
:
872 self
._parse
('-', stream
, False)
875 class Popen(subprocess
.Popen
):
876 if sys
.platform
== 'win32':
877 _startupinfo
= subprocess
.STARTUPINFO()
878 _startupinfo
.dwFlags |
= subprocess
.STARTF_USESHOWWINDOW
883 def _fix_pyinstaller_ld_path(env
):
884 """Restore LD_LIBRARY_PATH when using PyInstaller
885 Ref: https://github.com/pyinstaller/pyinstaller/blob/develop/doc/runtime-information.rst#ld_library_path--libpath-considerations
886 https://github.com/yt-dlp/yt-dlp/issues/4573
888 if not hasattr(sys
, '_MEIPASS'):
892 orig
= env
.get(f
'{key}_ORIG')
898 _fix('LD_LIBRARY_PATH') # Linux
899 _fix('DYLD_LIBRARY_PATH') # macOS
901 def __init__(self
, *args
, env
=None, text
=False, **kwargs
):
903 env
= os
.environ
.copy()
904 self
._fix
_pyinstaller
_ld
_path
(env
)
906 self
.__text
_mode
= kwargs
.get('encoding') or kwargs
.get('errors') or text
or kwargs
.get('universal_newlines')
908 kwargs
['universal_newlines'] = True # For 3.6 compatibility
909 kwargs
.setdefault('encoding', 'utf-8')
910 kwargs
.setdefault('errors', 'replace')
911 super().__init
__(*args
, env
=env
, **kwargs
, startupinfo
=self
._startupinfo
)
913 def communicate_or_kill(self
, *args
, **kwargs
):
915 return self
.communicate(*args
, **kwargs
)
916 except BaseException
: # Including KeyboardInterrupt
917 self
.kill(timeout
=None)
920 def kill(self
, *, timeout
=0):
923 self
.wait(timeout
=timeout
)
926 def run(cls
, *args
, timeout
=None, **kwargs
):
927 with cls(*args
, **kwargs
) as proc
:
928 default
= '' if proc
.__text
_mode
else b
''
929 stdout
, stderr
= proc
.communicate_or_kill(timeout
=timeout
)
930 return stdout
or default
, stderr
or default
, proc
.returncode
933 def encodeArgument(s
):
934 # Legacy code that uses byte strings
935 # Uncomment the following line after fixing all post processors
936 # assert isinstance(s, str), 'Internal error: %r should be of type %r, is %r' % (s, str, type(s))
937 return s
if isinstance(s
, str) else s
.decode('ascii')
940 _timetuple
= collections
.namedtuple('Time', ('hours', 'minutes', 'seconds', 'milliseconds'))
943 def timetuple_from_msec(msec
):
944 secs
, msec
= divmod(msec
, 1000)
945 mins
, secs
= divmod(secs
, 60)
946 hrs
, mins
= divmod(mins
, 60)
947 return _timetuple(hrs
, mins
, secs
, msec
)
950 def formatSeconds(secs
, delim
=':', msec
=False):
951 time
= timetuple_from_msec(secs
* 1000)
953 ret
= '%d%s%02d%s%02d' % (time
.hours
, delim
, time
.minutes
, delim
, time
.seconds
)
955 ret
= '%d%s%02d' % (time
.minutes
, delim
, time
.seconds
)
957 ret
= '%d' % time
.seconds
958 return '%s.%03d' % (ret
, time
.milliseconds
) if msec
else ret
961 def _ssl_load_windows_store_certs(ssl_context
, storename
):
962 # Code adapted from _load_windows_store_certs in https://github.com/python/cpython/blob/main/Lib/ssl.py
964 certs
= [cert
for cert
, encoding
, trust
in ssl
.enum_certificates(storename
)
965 if encoding
== 'x509_asn' and (
966 trust
is True or ssl
.Purpose
.SERVER_AUTH
.oid
in trust
)]
967 except PermissionError
:
970 with contextlib
.suppress(ssl
.SSLError
):
971 ssl_context
.load_verify_locations(cadata
=cert
)
974 def make_HTTPS_handler(params
, **kwargs
):
975 opts_check_certificate
= not params
.get('nocheckcertificate')
976 context
= ssl
.SSLContext(ssl
.PROTOCOL_TLS_CLIENT
)
977 context
.check_hostname
= opts_check_certificate
978 if params
.get('legacyserverconnect'):
979 context
.options |
= 4 # SSL_OP_LEGACY_SERVER_CONNECT
980 # Allow use of weaker ciphers in Python 3.10+. See https://bugs.python.org/issue43998
981 context
.set_ciphers('DEFAULT')
983 sys
.version_info
< (3, 10)
984 and ssl
.OPENSSL_VERSION_INFO
>= (1, 1, 1)
985 and not ssl
.OPENSSL_VERSION
.startswith('LibreSSL')
987 # Backport the default SSL ciphers and minimum TLS version settings from Python 3.10 [1].
988 # This is to ensure consistent behavior across Python versions, and help avoid fingerprinting
989 # in some situations [2][3].
990 # Python 3.10 only supports OpenSSL 1.1.1+ [4]. Because this change is likely
991 # untested on older versions, we only apply this to OpenSSL 1.1.1+ to be safe.
992 # LibreSSL is excluded until further investigation due to cipher support issues [5][6].
993 # 1. https://github.com/python/cpython/commit/e983252b516edb15d4338b0a47631b59ef1e2536
994 # 2. https://github.com/yt-dlp/yt-dlp/issues/4627
995 # 3. https://github.com/yt-dlp/yt-dlp/pull/5294
996 # 4. https://peps.python.org/pep-0644/
997 # 5. https://peps.python.org/pep-0644/#libressl-support
998 # 6. https://github.com/yt-dlp/yt-dlp/commit/5b9f253fa0aee996cf1ed30185d4b502e00609c4#commitcomment-89054368
999 context
.set_ciphers('@SECLEVEL=2:ECDH+AESGCM:ECDH+CHACHA20:ECDH+AES:DHE+AES:!aNULL:!eNULL:!aDSS:!SHA1:!AESCCM')
1000 context
.minimum_version
= ssl
.TLSVersion
.TLSv1_2
1002 context
.verify_mode
= ssl
.CERT_REQUIRED
if opts_check_certificate
else ssl
.CERT_NONE
1003 if opts_check_certificate
:
1004 if certifi
and 'no-certifi' not in params
.get('compat_opts', []):
1005 context
.load_verify_locations(cafile
=certifi
.where())
1008 context
.load_default_certs()
1009 # Work around the issue in load_default_certs when there are bad certificates. See:
1010 # https://github.com/yt-dlp/yt-dlp/issues/1060,
1011 # https://bugs.python.org/issue35665, https://bugs.python.org/issue45312
1012 except ssl
.SSLError
:
1013 # enum_certificates is not present in mingw python. See https://github.com/yt-dlp/yt-dlp/issues/1151
1014 if sys
.platform
== 'win32' and hasattr(ssl
, 'enum_certificates'):
1015 for storename
in ('CA', 'ROOT'):
1016 _ssl_load_windows_store_certs(context
, storename
)
1017 context
.set_default_verify_paths()
1019 client_certfile
= params
.get('client_certificate')
1022 context
.load_cert_chain(
1023 client_certfile
, keyfile
=params
.get('client_certificate_key'),
1024 password
=params
.get('client_certificate_password'))
1025 except ssl
.SSLError
:
1026 raise YoutubeDLError('Unable to load client certificate')
1028 # Some servers may reject requests if ALPN extension is not sent. See:
1029 # https://github.com/python/cpython/issues/85140
1030 # https://github.com/yt-dlp/yt-dlp/issues/3878
1031 with contextlib
.suppress(NotImplementedError):
1032 context
.set_alpn_protocols(['http/1.1'])
1034 return YoutubeDLHTTPSHandler(params
, context
=context
, **kwargs
)
1037 def bug_reports_message(before
=';'):
1038 from ..update
import REPOSITORY
1040 msg
= (f
'please report this issue on https://github.com/{REPOSITORY}/issues?q= , '
1041 'filling out the appropriate issue template. Confirm you are on the latest version using yt-dlp -U')
1043 before
= before
.rstrip()
1044 if not before
or before
.endswith(('.', '!', '?')):
1045 msg
= msg
[0].title() + msg
[1:]
1047 return (before
+ ' ' if before
else '') + msg
1050 class YoutubeDLError(Exception):
1051 """Base exception for YoutubeDL errors."""
1054 def __init__(self
, msg
=None):
1057 elif self
.msg
is None:
1058 self
.msg
= type(self
).__name
__
1059 super().__init
__(self
.msg
)
1062 network_exceptions
= [urllib
.error
.URLError
, http
.client
.HTTPException
, socket
.error
]
1063 if hasattr(ssl
, 'CertificateError'):
1064 network_exceptions
.append(ssl
.CertificateError
)
1065 network_exceptions
= tuple(network_exceptions
)
1068 class ExtractorError(YoutubeDLError
):
1069 """Error during info extraction."""
1071 def __init__(self
, msg
, tb
=None, expected
=False, cause
=None, video_id
=None, ie
=None):
1072 """ tb, if given, is the original traceback (so that it can be printed out).
1073 If expected is set, this is a normal error message and most likely not a bug in yt-dlp.
1075 if sys
.exc_info()[0] in network_exceptions
:
1078 self
.orig_msg
= str(msg
)
1080 self
.expected
= expected
1082 self
.video_id
= video_id
1084 self
.exc_info
= sys
.exc_info() # preserve original exception
1085 if isinstance(self
.exc_info
[1], ExtractorError
):
1086 self
.exc_info
= self
.exc_info
[1].exc_info
1087 super().__init
__(self
.__msg
)
1092 format_field(self
.ie
, None, '[%s] '),
1093 format_field(self
.video_id
, None, '%s: '),
1095 format_field(self
.cause
, None, ' (caused by %r)'),
1096 '' if self
.expected
else bug_reports_message()))
1098 def format_traceback(self
):
1099 return join_nonempty(
1100 self
.traceback
and ''.join(traceback
.format_tb(self
.traceback
)),
1101 self
.cause
and ''.join(traceback
.format_exception(None, self
.cause
, self
.cause
.__traceback
__)[1:]),
1104 def __setattr__(self
, name
, value
):
1105 super().__setattr
__(name
, value
)
1106 if getattr(self
, 'msg', None) and name
not in ('msg', 'args'):
1107 self
.msg
= self
.__msg
or type(self
).__name
__
1108 self
.args
= (self
.msg
, ) # Cannot be property
1111 class UnsupportedError(ExtractorError
):
1112 def __init__(self
, url
):
1114 'Unsupported URL: %s' % url
, expected
=True)
1118 class RegexNotFoundError(ExtractorError
):
1119 """Error when a regex didn't match"""
1123 class GeoRestrictedError(ExtractorError
):
1124 """Geographic restriction Error exception.
1126 This exception may be thrown when a video is not available from your
1127 geographic location due to geographic restrictions imposed by a website.
1130 def __init__(self
, msg
, countries
=None, **kwargs
):
1131 kwargs
['expected'] = True
1132 super().__init
__(msg
, **kwargs
)
1133 self
.countries
= countries
1136 class UserNotLive(ExtractorError
):
1137 """Error when a channel/user is not live"""
1139 def __init__(self
, msg
=None, **kwargs
):
1140 kwargs
['expected'] = True
1141 super().__init
__(msg
or 'The channel is not currently live', **kwargs
)
1144 class DownloadError(YoutubeDLError
):
1145 """Download Error exception.
1147 This exception may be thrown by FileDownloader objects if they are not
1148 configured to continue on errors. They will contain the appropriate
1152 def __init__(self
, msg
, exc_info
=None):
1153 """ exc_info, if given, is the original exception that caused the trouble (as returned by sys.exc_info()). """
1154 super().__init
__(msg
)
1155 self
.exc_info
= exc_info
1158 class EntryNotInPlaylist(YoutubeDLError
):
1159 """Entry not in playlist exception.
1161 This exception will be thrown by YoutubeDL when a requested entry
1162 is not found in the playlist info_dict
1164 msg
= 'Entry not found in info'
1167 class SameFileError(YoutubeDLError
):
1168 """Same File exception.
1170 This exception will be thrown by FileDownloader objects if they detect
1171 multiple files would have to be downloaded to the same file on disk.
1173 msg
= 'Fixed output name but more than one file to download'
1175 def __init__(self
, filename
=None):
1176 if filename
is not None:
1177 self
.msg
+= f
': {filename}'
1178 super().__init
__(self
.msg
)
1181 class PostProcessingError(YoutubeDLError
):
1182 """Post Processing exception.
1184 This exception may be raised by PostProcessor's .run() method to
1185 indicate an error in the postprocessing task.
1189 class DownloadCancelled(YoutubeDLError
):
1190 """ Exception raised when the download queue should be interrupted """
1191 msg
= 'The download was cancelled'
1194 class ExistingVideoReached(DownloadCancelled
):
1195 """ --break-on-existing triggered """
1196 msg
= 'Encountered a video that is already in the archive, stopping due to --break-on-existing'
1199 class RejectedVideoReached(DownloadCancelled
):
1200 """ --break-match-filter triggered """
1201 msg
= 'Encountered a video that did not match filter, stopping due to --break-match-filter'
1204 class MaxDownloadsReached(DownloadCancelled
):
1205 """ --max-downloads limit has been reached. """
1206 msg
= 'Maximum number of downloads reached, stopping due to --max-downloads'
1209 class ReExtractInfo(YoutubeDLError
):
1210 """ Video info needs to be re-extracted. """
1212 def __init__(self
, msg
, expected
=False):
1213 super().__init
__(msg
)
1214 self
.expected
= expected
1217 class ThrottledDownload(ReExtractInfo
):
1218 """ Download speed below --throttled-rate. """
1219 msg
= 'The download speed is below throttle limit'
1222 super().__init
__(self
.msg
, expected
=False)
1225 class UnavailableVideoError(YoutubeDLError
):
1226 """Unavailable Format exception.
1228 This exception will be thrown when a video is requested
1229 in a format that is not available for that video.
1231 msg
= 'Unable to download video'
1233 def __init__(self
, err
=None):
1235 self
.msg
+= f
': {err}'
1236 super().__init
__(self
.msg
)
1239 class ContentTooShortError(YoutubeDLError
):
1240 """Content Too Short exception.
1242 This exception may be raised by FileDownloader objects when a file they
1243 download is too small for what the server announced first, indicating
1244 the connection was probably interrupted.
1247 def __init__(self
, downloaded
, expected
):
1248 super().__init
__(f
'Downloaded {downloaded} bytes, expected {expected} bytes')
1250 self
.downloaded
= downloaded
1251 self
.expected
= expected
1254 class XAttrMetadataError(YoutubeDLError
):
1255 def __init__(self
, code
=None, msg
='Unknown error'):
1256 super().__init
__(msg
)
1260 # Parsing code and msg
1261 if (self
.code
in (errno
.ENOSPC
, errno
.EDQUOT
)
1262 or 'No space left' in self
.msg
or 'Disk quota exceeded' in self
.msg
):
1263 self
.reason
= 'NO_SPACE'
1264 elif self
.code
== errno
.E2BIG
or 'Argument list too long' in self
.msg
:
1265 self
.reason
= 'VALUE_TOO_LONG'
1267 self
.reason
= 'NOT_SUPPORTED'
1270 class XAttrUnavailableError(YoutubeDLError
):
1274 def _create_http_connection(ydl_handler
, http_class
, is_https
, *args
, **kwargs
):
1275 hc
= http_class(*args
, **kwargs
)
1276 source_address
= ydl_handler
._params
.get('source_address')
1278 if source_address
is not None:
1279 # This is to workaround _create_connection() from socket where it will try all
1280 # address data from getaddrinfo() including IPv6. This filters the result from
1281 # getaddrinfo() based on the source_address value.
1282 # This is based on the cpython socket.create_connection() function.
1283 # https://github.com/python/cpython/blob/master/Lib/socket.py#L691
1284 def _create_connection(address
, timeout
=socket
._GLOBAL
_DEFAULT
_TIMEOUT
, source_address
=None):
1285 host
, port
= address
1287 addrs
= socket
.getaddrinfo(host
, port
, 0, socket
.SOCK_STREAM
)
1288 af
= socket
.AF_INET
if '.' in source_address
[0] else socket
.AF_INET6
1289 ip_addrs
= [addr
for addr
in addrs
if addr
[0] == af
]
1290 if addrs
and not ip_addrs
:
1291 ip_version
= 'v4' if af
== socket
.AF_INET
else 'v6'
1293 "No remote IP%s addresses available for connect, can't use '%s' as source address"
1294 % (ip_version
, source_address
[0]))
1295 for res
in ip_addrs
:
1296 af
, socktype
, proto
, canonname
, sa
= res
1299 sock
= socket
.socket(af
, socktype
, proto
)
1300 if timeout
is not socket
._GLOBAL
_DEFAULT
_TIMEOUT
:
1301 sock
.settimeout(timeout
)
1302 sock
.bind(source_address
)
1304 err
= None # Explicitly break reference cycle
1306 except OSError as _
:
1308 if sock
is not None:
1313 raise OSError('getaddrinfo returns an empty list')
1314 if hasattr(hc
, '_create_connection'):
1315 hc
._create
_connection
= _create_connection
1316 hc
.source_address
= (source_address
, 0)
1321 class YoutubeDLHandler(urllib
.request
.HTTPHandler
):
1322 """Handler for HTTP requests and responses.
1324 This class, when installed with an OpenerDirector, automatically adds
1325 the standard headers to every HTTP request and handles gzipped, deflated and
1326 brotli responses from web servers.
1328 Part of this code was copied from:
1330 http://techknack.net/python-urllib2-handlers/
1332 Andrew Rowls, the author of that code, agreed to release it to the
1336 def __init__(self
, params
, *args
, **kwargs
):
1337 urllib
.request
.HTTPHandler
.__init
__(self
, *args
, **kwargs
)
1338 self
._params
= params
1340 def http_open(self
, req
):
1341 conn_class
= http
.client
.HTTPConnection
1343 socks_proxy
= req
.headers
.get('Ytdl-socks-proxy')
1345 conn_class
= make_socks_conn_class(conn_class
, socks_proxy
)
1346 del req
.headers
['Ytdl-socks-proxy']
1348 return self
.do_open(functools
.partial(
1349 _create_http_connection
, self
, conn_class
, False),
1357 return zlib
.decompress(data
, -zlib
.MAX_WBITS
)
1359 return zlib
.decompress(data
)
1365 return brotli
.decompress(data
)
1369 gz
= gzip
.GzipFile(fileobj
=io
.BytesIO(data
), mode
='rb')
1372 except OSError as original_oserror
:
1373 # There may be junk add the end of the file
1374 # See http://stackoverflow.com/q/4928560/35070 for details
1375 for i
in range(1, 1024):
1377 gz
= gzip
.GzipFile(fileobj
=io
.BytesIO(data
[:-i
]), mode
='rb')
1382 raise original_oserror
1384 def http_request(self
, req
):
1385 # According to RFC 3986, URLs can not contain non-ASCII characters, however this is not
1386 # always respected by websites, some tend to give out URLs with non percent-encoded
1387 # non-ASCII characters (see telemb.py, ard.py [#3412])
1388 # urllib chokes on URLs with non-ASCII characters (see http://bugs.python.org/issue3991)
1389 # To work around aforementioned issue we will replace request's original URL with
1390 # percent-encoded one
1391 # Since redirects are also affected (e.g. http://www.southpark.de/alle-episoden/s18e09)
1392 # the code of this workaround has been moved here from YoutubeDL.urlopen()
1393 url
= req
.get_full_url()
1394 url_escaped
= escape_url(url
)
1396 # Substitute URL if any change after escaping
1397 if url
!= url_escaped
:
1398 req
= update_Request(req
, url
=url_escaped
)
1400 for h
, v
in self
._params
.get('http_headers', std_headers
).items():
1401 # Capitalize is needed because of Python bug 2275: http://bugs.python.org/issue2275
1402 # The dict keys are capitalized because of this bug by urllib
1403 if h
.capitalize() not in req
.headers
:
1404 req
.add_header(h
, v
)
1406 if 'Youtubedl-no-compression' in req
.headers
: # deprecated
1407 req
.headers
.pop('Youtubedl-no-compression', None)
1408 req
.add_header('Accept-encoding', 'identity')
1410 if 'Accept-encoding' not in req
.headers
:
1411 req
.add_header('Accept-encoding', ', '.join(SUPPORTED_ENCODINGS
))
1413 return super().do_request_(req
)
1415 def http_response(self
, req
, resp
):
1418 # Content-Encoding header lists the encodings in order that they were applied [1].
1419 # To decompress, we simply do the reverse.
1420 # [1]: https://datatracker.ietf.org/doc/html/rfc9110#name-content-encoding
1421 decoded_response
= None
1422 for encoding
in (e
.strip() for e
in reversed(resp
.headers
.get('Content-encoding', '').split(','))):
1423 if encoding
== 'gzip':
1424 decoded_response
= self
.gz(decoded_response
or resp
.read())
1425 elif encoding
== 'deflate':
1426 decoded_response
= self
.deflate(decoded_response
or resp
.read())
1427 elif encoding
== 'br' and brotli
:
1428 decoded_response
= self
.brotli(decoded_response
or resp
.read())
1430 if decoded_response
is not None:
1431 resp
= urllib
.request
.addinfourl(io
.BytesIO(decoded_response
), old_resp
.headers
, old_resp
.url
, old_resp
.code
)
1432 resp
.msg
= old_resp
.msg
1433 # Percent-encode redirect URL of Location HTTP header to satisfy RFC 3986 (see
1434 # https://github.com/ytdl-org/youtube-dl/issues/6457).
1435 if 300 <= resp
.code
< 400:
1436 location
= resp
.headers
.get('Location')
1438 # As of RFC 2616 default charset is iso-8859-1 that is respected by python 3
1439 location
= location
.encode('iso-8859-1').decode()
1440 location_escaped
= escape_url(location
)
1441 if location
!= location_escaped
:
1442 del resp
.headers
['Location']
1443 resp
.headers
['Location'] = location_escaped
1446 https_request
= http_request
1447 https_response
= http_response
1450 def make_socks_conn_class(base_class
, socks_proxy
):
1451 assert issubclass(base_class
, (
1452 http
.client
.HTTPConnection
, http
.client
.HTTPSConnection
))
1454 url_components
= urllib
.parse
.urlparse(socks_proxy
)
1455 if url_components
.scheme
.lower() == 'socks5':
1456 socks_type
= ProxyType
.SOCKS5
1457 elif url_components
.scheme
.lower() in ('socks', 'socks4'):
1458 socks_type
= ProxyType
.SOCKS4
1459 elif url_components
.scheme
.lower() == 'socks4a':
1460 socks_type
= ProxyType
.SOCKS4A
1462 def unquote_if_non_empty(s
):
1465 return urllib
.parse
.unquote_plus(s
)
1469 url_components
.hostname
, url_components
.port
or 1080,
1471 unquote_if_non_empty(url_components
.username
),
1472 unquote_if_non_empty(url_components
.password
),
1475 class SocksConnection(base_class
):
1477 self
.sock
= sockssocket()
1478 self
.sock
.setproxy(*proxy_args
)
1479 if isinstance(self
.timeout
, (int, float)):
1480 self
.sock
.settimeout(self
.timeout
)
1481 self
.sock
.connect((self
.host
, self
.port
))
1483 if isinstance(self
, http
.client
.HTTPSConnection
):
1484 if hasattr(self
, '_context'): # Python > 2.6
1485 self
.sock
= self
._context
.wrap_socket(
1486 self
.sock
, server_hostname
=self
.host
)
1488 self
.sock
= ssl
.wrap_socket(self
.sock
)
1490 return SocksConnection
1493 class YoutubeDLHTTPSHandler(urllib
.request
.HTTPSHandler
):
1494 def __init__(self
, params
, https_conn_class
=None, *args
, **kwargs
):
1495 urllib
.request
.HTTPSHandler
.__init
__(self
, *args
, **kwargs
)
1496 self
._https
_conn
_class
= https_conn_class
or http
.client
.HTTPSConnection
1497 self
._params
= params
1499 def https_open(self
, req
):
1501 conn_class
= self
._https
_conn
_class
1503 if hasattr(self
, '_context'): # python > 2.6
1504 kwargs
['context'] = self
._context
1505 if hasattr(self
, '_check_hostname'): # python 3.x
1506 kwargs
['check_hostname'] = self
._check
_hostname
1508 socks_proxy
= req
.headers
.get('Ytdl-socks-proxy')
1510 conn_class
= make_socks_conn_class(conn_class
, socks_proxy
)
1511 del req
.headers
['Ytdl-socks-proxy']
1514 return self
.do_open(
1515 functools
.partial(_create_http_connection
, self
, conn_class
, True), req
, **kwargs
)
1516 except urllib
.error
.URLError
as e
:
1517 if (isinstance(e
.reason
, ssl
.SSLError
)
1518 and getattr(e
.reason
, 'reason', None) == 'SSLV3_ALERT_HANDSHAKE_FAILURE'):
1519 raise YoutubeDLError('SSLV3_ALERT_HANDSHAKE_FAILURE: Try using --legacy-server-connect')
1523 def is_path_like(f
):
1524 return isinstance(f
, (str, bytes, os
.PathLike
))
1527 class YoutubeDLCookieProcessor(urllib
.request
.HTTPCookieProcessor
):
1528 def __init__(self
, cookiejar
=None):
1529 urllib
.request
.HTTPCookieProcessor
.__init
__(self
, cookiejar
)
1531 def http_response(self
, request
, response
):
1532 return urllib
.request
.HTTPCookieProcessor
.http_response(self
, request
, response
)
1534 https_request
= urllib
.request
.HTTPCookieProcessor
.http_request
1535 https_response
= http_response
1538 class YoutubeDLRedirectHandler(urllib
.request
.HTTPRedirectHandler
):
1539 """YoutubeDL redirect handler
1541 The code is based on HTTPRedirectHandler implementation from CPython [1].
1543 This redirect handler fixes and improves the logic to better align with RFC7261
1544 and what browsers tend to do [2][3]
1546 1. https://github.com/python/cpython/blob/master/Lib/urllib/request.py
1547 2. https://datatracker.ietf.org/doc/html/rfc7231
1548 3. https://github.com/python/cpython/issues/91306
1551 http_error_301
= http_error_303
= http_error_307
= http_error_308
= urllib
.request
.HTTPRedirectHandler
.http_error_302
1553 def redirect_request(self
, req
, fp
, code
, msg
, headers
, newurl
):
1554 if code
not in (301, 302, 303, 307, 308):
1555 raise urllib
.error
.HTTPError(req
.full_url
, code
, msg
, headers
, fp
)
1557 new_method
= req
.get_method()
1560 # Technically the Cookie header should be in unredirected_hdrs,
1561 # however in practice some may set it in normal headers anyway.
1562 # We will remove it here to prevent any leaks.
1563 remove_headers
= ['Cookie']
1565 # A 303 must either use GET or HEAD for subsequent request
1566 # https://datatracker.ietf.org/doc/html/rfc7231#section-6.4.4
1567 if code
== 303 and req
.get_method() != 'HEAD':
1569 # 301 and 302 redirects are commonly turned into a GET from a POST
1570 # for subsequent requests by browsers, so we'll do the same.
1571 # https://datatracker.ietf.org/doc/html/rfc7231#section-6.4.2
1572 # https://datatracker.ietf.org/doc/html/rfc7231#section-6.4.3
1573 elif code
in (301, 302) and req
.get_method() == 'POST':
1576 # only remove payload if method changed (e.g. POST to GET)
1577 if new_method
!= req
.get_method():
1579 remove_headers
.extend(['Content-Length', 'Content-Type'])
1581 new_headers
= {k: v for k, v in req.headers.items() if k.title() not in remove_headers}
1583 return urllib
.request
.Request(
1584 newurl
, headers
=new_headers
, origin_req_host
=req
.origin_req_host
,
1585 unverifiable
=True, method
=new_method
, data
=new_data
)
1588 def extract_timezone(date_str
):
1591 ^.{8,}? # >=8 char non-TZ prefix, if present
1592 (?P<tz>Z| # just the UTC Z, or
1593 (?:(?<=.\b\d{4}|\b\d{2}:\d\d)| # preceded by 4 digits or hh:mm or
1594 (?<!.\b[a-zA-Z]{3}|[a-zA-Z]{4}|..\b\d\d)) # not preceded by 3 alpha word or >= 4 alpha or 2 digits
1595 [ ]? # optional space
1596 (?P<sign>\+|-) # +/-
1597 (?P<hours>[0-9]{2}):?(?P<minutes>[0-9]{2}) # hh[:]mm
1601 m
= re
.search(r
'\d{1,2}:\d{1,2}(?:\.\d+)?(?P<tz>\s*[A-Z]+)$', date_str
)
1602 timezone
= TIMEZONE_NAMES
.get(m
and m
.group('tz').strip())
1603 if timezone
is not None:
1604 date_str
= date_str
[:-len(m
.group('tz'))]
1605 timezone
= datetime
.timedelta(hours
=timezone
or 0)
1607 date_str
= date_str
[:-len(m
.group('tz'))]
1608 if not m
.group('sign'):
1609 timezone
= datetime
.timedelta()
1611 sign
= 1 if m
.group('sign') == '+' else -1
1612 timezone
= datetime
.timedelta(
1613 hours
=sign
* int(m
.group('hours')),
1614 minutes
=sign
* int(m
.group('minutes')))
1615 return timezone
, date_str
1618 def parse_iso8601(date_str
, delimiter
='T', timezone
=None):
1619 """ Return a UNIX timestamp from the given date """
1621 if date_str
is None:
1624 date_str
= re
.sub(r
'\.[0-9]+', '', date_str
)
1626 if timezone
is None:
1627 timezone
, date_str
= extract_timezone(date_str
)
1629 with contextlib
.suppress(ValueError):
1630 date_format
= f
'%Y-%m-%d{delimiter}%H:%M:%S'
1631 dt
= datetime
.datetime
.strptime(date_str
, date_format
) - timezone
1632 return calendar
.timegm(dt
.timetuple())
1635 def date_formats(day_first
=True):
1636 return DATE_FORMATS_DAY_FIRST
if day_first
else DATE_FORMATS_MONTH_FIRST
1639 def unified_strdate(date_str
, day_first
=True):
1640 """Return a string with the date in the format YYYYMMDD"""
1642 if date_str
is None:
1646 date_str
= date_str
.replace(',', ' ')
1647 # Remove AM/PM + timezone
1648 date_str
= re
.sub(r
'(?i)\s*(?:AM|PM)(?:\s+[A-Z]+)?', '', date_str
)
1649 _
, date_str
= extract_timezone(date_str
)
1651 for expression
in date_formats(day_first
):
1652 with contextlib
.suppress(ValueError):
1653 upload_date
= datetime
.datetime
.strptime(date_str
, expression
).strftime('%Y%m%d')
1654 if upload_date
is None:
1655 timetuple
= email
.utils
.parsedate_tz(date_str
)
1657 with contextlib
.suppress(ValueError):
1658 upload_date
= datetime
.datetime(*timetuple
[:6]).strftime('%Y%m%d')
1659 if upload_date
is not None:
1660 return str(upload_date
)
1663 def unified_timestamp(date_str
, day_first
=True):
1664 if not isinstance(date_str
, str):
1667 date_str
= re
.sub(r
'\s+', ' ', re
.sub(
1668 r
'(?i)[,|]|(mon|tues?|wed(nes)?|thu(rs)?|fri|sat(ur)?)(day)?', '', date_str
))
1670 pm_delta
= 12 if re
.search(r
'(?i)PM', date_str
) else 0
1671 timezone
, date_str
= extract_timezone(date_str
)
1673 # Remove AM/PM + timezone
1674 date_str
= re
.sub(r
'(?i)\s*(?:AM|PM)(?:\s+[A-Z]+)?', '', date_str
)
1676 # Remove unrecognized timezones from ISO 8601 alike timestamps
1677 m
= re
.search(r
'\d{1,2}:\d{1,2}(?:\.\d+)?(?P<tz>\s*[A-Z]+)$', date_str
)
1679 date_str
= date_str
[:-len(m
.group('tz'))]
1681 # Python only supports microseconds, so remove nanoseconds
1682 m
= re
.search(r
'^([0-9]{4,}-[0-9]{1,2}-[0-9]{1,2}T[0-9]{1,2}:[0-9]{1,2}:[0-9]{1,2}\.[0-9]{6})[0-9]+$', date_str
)
1684 date_str
= m
.group(1)
1686 for expression
in date_formats(day_first
):
1687 with contextlib
.suppress(ValueError):
1688 dt
= datetime
.datetime
.strptime(date_str
, expression
) - timezone
+ datetime
.timedelta(hours
=pm_delta
)
1689 return calendar
.timegm(dt
.timetuple())
1691 timetuple
= email
.utils
.parsedate_tz(date_str
)
1693 return calendar
.timegm(timetuple
) + pm_delta
* 3600 - timezone
.total_seconds()
1696 def determine_ext(url
, default_ext
='unknown_video'):
1697 if url
is None or '.' not in url
:
1699 guess
= url
.partition('?')[0].rpartition('.')[2]
1700 if re
.match(r
'^[A-Za-z0-9]+$', guess
):
1702 # Try extract ext from URLs like http://example.com/foo/bar.mp4/?download
1703 elif guess
.rstrip('/') in KNOWN_EXTENSIONS
:
1704 return guess
.rstrip('/')
1709 def subtitles_filename(filename
, sub_lang
, sub_format
, expected_real_ext
=None):
1710 return replace_extension(filename
, sub_lang
+ '.' + sub_format
, expected_real_ext
)
1713 def datetime_from_str(date_str
, precision
='auto', format
='%Y%m%d'):
1715 Return a datetime object from a string.
1717 (now|today|yesterday|DATE)([+-]\d+(microsecond|second|minute|hour|day|week|month|year)s?)?
1719 @param format strftime format of DATE
1720 @param precision Round the datetime object: auto|microsecond|second|minute|hour|day
1721 auto: round to the unit provided in date_str (if applicable).
1723 auto_precision
= False
1724 if precision
== 'auto':
1725 auto_precision
= True
1726 precision
= 'microsecond'
1727 today
= datetime_round(datetime
.datetime
.utcnow(), precision
)
1728 if date_str
in ('now', 'today'):
1730 if date_str
== 'yesterday':
1731 return today
- datetime
.timedelta(days
=1)
1733 r
'(?P<start>.+)(?P<sign>[+-])(?P<time>\d+)(?P<unit>microsecond|second|minute|hour|day|week|month|year)s?',
1735 if match
is not None:
1736 start_time
= datetime_from_str(match
.group('start'), precision
, format
)
1737 time
= int(match
.group('time')) * (-1 if match
.group('sign') == '-' else 1)
1738 unit
= match
.group('unit')
1739 if unit
== 'month' or unit
== 'year':
1740 new_date
= datetime_add_months(start_time
, time
* 12 if unit
== 'year' else time
)
1746 delta
= datetime
.timedelta(**{unit + 's': time}
)
1747 new_date
= start_time
+ delta
1749 return datetime_round(new_date
, unit
)
1752 return datetime_round(datetime
.datetime
.strptime(date_str
, format
), precision
)
1755 def date_from_str(date_str
, format
='%Y%m%d', strict
=False):
1757 Return a date object from a string using datetime_from_str
1759 @param strict Restrict allowed patterns to "YYYYMMDD" and
1760 (now|today|yesterday)(-\d+(day|week|month|year)s?)?
1762 if strict
and not re
.fullmatch(r
'\d{8}|(now|today|yesterday)(-\d+(day|week|month|year)s?)?', date_str
):
1763 raise ValueError(f
'Invalid date format "{date_str}"')
1764 return datetime_from_str(date_str
, precision
='microsecond', format
=format
).date()
1767 def datetime_add_months(dt
, months
):
1768 """Increment/Decrement a datetime object by months."""
1769 month
= dt
.month
+ months
- 1
1770 year
= dt
.year
+ month
// 12
1771 month
= month
% 12 + 1
1772 day
= min(dt
.day
, calendar
.monthrange(year
, month
)[1])
1773 return dt
.replace(year
, month
, day
)
1776 def datetime_round(dt
, precision
='day'):
1778 Round a datetime object's time to a specific precision
1780 if precision
== 'microsecond':
1789 roundto
= lambda x
, n
: ((x
+ n
/ 2) // n
) * n
1790 timestamp
= calendar
.timegm(dt
.timetuple())
1791 return datetime
.datetime
.utcfromtimestamp(roundto(timestamp
, unit_seconds
[precision
]))
1794 def hyphenate_date(date_str
):
1796 Convert a date in 'YYYYMMDD' format to 'YYYY-MM-DD' format"""
1797 match
= re
.match(r
'^(\d\d\d\d)(\d\d)(\d\d)$', date_str
)
1798 if match
is not None:
1799 return '-'.join(match
.groups())
1805 """Represents a time interval between two dates"""
1807 def __init__(self
, start
=None, end
=None):
1808 """start and end must be strings in the format accepted by date"""
1809 if start
is not None:
1810 self
.start
= date_from_str(start
, strict
=True)
1812 self
.start
= datetime
.datetime
.min.date()
1814 self
.end
= date_from_str(end
, strict
=True)
1816 self
.end
= datetime
.datetime
.max.date()
1817 if self
.start
> self
.end
:
1818 raise ValueError('Date range: "%s" , the start date must be before the end date' % self
)
1822 """Returns a range that only contains the given day"""
1823 return cls(day
, day
)
1825 def __contains__(self
, date
):
1826 """Check if the date is in the range"""
1827 if not isinstance(date
, datetime
.date
):
1828 date
= date_from_str(date
)
1829 return self
.start
<= date
<= self
.end
1832 return f
'{__name__}.{type(self).__name__}({self.start.isoformat()!r}, {self.end.isoformat()!r})'
1834 def __eq__(self
, other
):
1835 return (isinstance(other
, DateRange
)
1836 and self
.start
== other
.start
and self
.end
== other
.end
)
1840 def system_identifier():
1841 python_implementation
= platform
.python_implementation()
1842 if python_implementation
== 'PyPy' and hasattr(sys
, 'pypy_version_info'):
1843 python_implementation
+= ' version %d.%d.%d' % sys
.pypy_version_info
[:3]
1845 with contextlib
.suppress(OSError): # We may not have access to the executable
1846 libc_ver
= platform
.libc_ver()
1848 return 'Python %s (%s %s %s) - %s (%s%s)' % (
1849 platform
.python_version(),
1850 python_implementation
,
1852 platform
.architecture()[0],
1853 platform
.platform(),
1854 ssl
.OPENSSL_VERSION
,
1855 format_field(join_nonempty(*libc_ver
, delim
=' '), None, ', %s'),
1860 def get_windows_version():
1861 ''' Get Windows version. returns () if it's not running on Windows '''
1862 if compat_os_name
== 'nt':
1863 return version_tuple(platform
.win32_ver()[1])
1868 def write_string(s
, out
=None, encoding
=None):
1869 assert isinstance(s
, str)
1870 out
= out
or sys
.stderr
1871 # `sys.stderr` might be `None` (Ref: https://github.com/pyinstaller/pyinstaller/pull/7217)
1875 if compat_os_name
== 'nt' and supports_terminal_sequences(out
):
1876 s
= re
.sub(r
'([\r\n]+)', r
' \1', s
)
1878 enc
, buffer = None, out
1879 if 'b' in getattr(out
, 'mode', ''):
1880 enc
= encoding
or preferredencoding()
1881 elif hasattr(out
, 'buffer'):
1883 enc
= encoding
or getattr(out
, 'encoding', None) or preferredencoding()
1885 buffer.write(s
.encode(enc
, 'ignore') if enc
else s
)
1889 def deprecation_warning(msg
, *, printer
=None, stacklevel
=0, **kwargs
):
1890 from .. import _IN_CLI
1892 if msg
in deprecation_warning
._cache
:
1894 deprecation_warning
._cache
.add(msg
)
1896 return printer(f
'{msg}{bug_reports_message()}', **kwargs
)
1897 return write_string(f
'ERROR: {msg}{bug_reports_message()}\n', **kwargs
)
1900 warnings
.warn(DeprecationWarning(msg
), stacklevel
=stacklevel
+ 3)
1903 deprecation_warning
._cache
= set()
1906 def bytes_to_intlist(bs
):
1909 if isinstance(bs
[0], int): # Python 3
1912 return [ord(c
) for c
in bs
]
1915 def intlist_to_bytes(xs
):
1918 return struct
.pack('%dB' % len(xs
), *xs
)
1921 class LockingUnsupportedError(OSError):
1922 msg
= 'File locking is not supported'
1925 super().__init
__(self
.msg
)
1928 # Cross-platform file locking
1929 if sys
.platform
== 'win32':
1931 import ctypes
.wintypes
1934 class OVERLAPPED(ctypes
.Structure
):
1936 ('Internal', ctypes
.wintypes
.LPVOID
),
1937 ('InternalHigh', ctypes
.wintypes
.LPVOID
),
1938 ('Offset', ctypes
.wintypes
.DWORD
),
1939 ('OffsetHigh', ctypes
.wintypes
.DWORD
),
1940 ('hEvent', ctypes
.wintypes
.HANDLE
),
1943 kernel32
= ctypes
.WinDLL('kernel32')
1944 LockFileEx
= kernel32
.LockFileEx
1945 LockFileEx
.argtypes
= [
1946 ctypes
.wintypes
.HANDLE
, # hFile
1947 ctypes
.wintypes
.DWORD
, # dwFlags
1948 ctypes
.wintypes
.DWORD
, # dwReserved
1949 ctypes
.wintypes
.DWORD
, # nNumberOfBytesToLockLow
1950 ctypes
.wintypes
.DWORD
, # nNumberOfBytesToLockHigh
1951 ctypes
.POINTER(OVERLAPPED
) # Overlapped
1953 LockFileEx
.restype
= ctypes
.wintypes
.BOOL
1954 UnlockFileEx
= kernel32
.UnlockFileEx
1955 UnlockFileEx
.argtypes
= [
1956 ctypes
.wintypes
.HANDLE
, # hFile
1957 ctypes
.wintypes
.DWORD
, # dwReserved
1958 ctypes
.wintypes
.DWORD
, # nNumberOfBytesToLockLow
1959 ctypes
.wintypes
.DWORD
, # nNumberOfBytesToLockHigh
1960 ctypes
.POINTER(OVERLAPPED
) # Overlapped
1962 UnlockFileEx
.restype
= ctypes
.wintypes
.BOOL
1963 whole_low
= 0xffffffff
1964 whole_high
= 0x7fffffff
1966 def _lock_file(f
, exclusive
, block
):
1967 overlapped
= OVERLAPPED()
1968 overlapped
.Offset
= 0
1969 overlapped
.OffsetHigh
= 0
1970 overlapped
.hEvent
= 0
1971 f
._lock
_file
_overlapped
_p
= ctypes
.pointer(overlapped
)
1973 if not LockFileEx(msvcrt
.get_osfhandle(f
.fileno()),
1974 (0x2 if exclusive
else 0x0) |
(0x0 if block
else 0x1),
1975 0, whole_low
, whole_high
, f
._lock
_file
_overlapped
_p
):
1976 # NB: No argument form of "ctypes.FormatError" does not work on PyPy
1977 raise BlockingIOError(f
'Locking file failed: {ctypes.FormatError(ctypes.GetLastError())!r}')
1979 def _unlock_file(f
):
1980 assert f
._lock
_file
_overlapped
_p
1981 handle
= msvcrt
.get_osfhandle(f
.fileno())
1982 if not UnlockFileEx(handle
, 0, whole_low
, whole_high
, f
._lock
_file
_overlapped
_p
):
1983 raise OSError('Unlocking file failed: %r' % ctypes
.FormatError())
1989 def _lock_file(f
, exclusive
, block
):
1990 flags
= fcntl
.LOCK_EX
if exclusive
else fcntl
.LOCK_SH
1992 flags |
= fcntl
.LOCK_NB
1994 fcntl
.flock(f
, flags
)
1995 except BlockingIOError
:
1997 except OSError: # AOSP does not have flock()
1998 fcntl
.lockf(f
, flags
)
2000 def _unlock_file(f
):
2001 with contextlib
.suppress(OSError):
2002 return fcntl
.flock(f
, fcntl
.LOCK_UN
)
2003 with contextlib
.suppress(OSError):
2004 return fcntl
.lockf(f
, fcntl
.LOCK_UN
) # AOSP does not have flock()
2005 return fcntl
.flock(f
, fcntl
.LOCK_UN | fcntl
.LOCK_NB
) # virtiofs needs LOCK_NB on unlocking
2009 def _lock_file(f
, exclusive
, block
):
2010 raise LockingUnsupportedError()
2012 def _unlock_file(f
):
2013 raise LockingUnsupportedError()
2019 def __init__(self
, filename
, mode
, block
=True, encoding
=None):
2020 if mode
not in {'r', 'rb', 'a', 'ab', 'w', 'wb'}
:
2021 raise NotImplementedError(mode
)
2022 self
.mode
, self
.block
= mode
, block
2024 writable
= any(f
in mode
for f
in 'wax+')
2025 readable
= any(f
in mode
for f
in 'r+')
2026 flags
= functools
.reduce(operator
.ior
, (
2027 getattr(os
, 'O_CLOEXEC', 0), # UNIX only
2028 getattr(os
, 'O_BINARY', 0), # Windows only
2029 getattr(os
, 'O_NOINHERIT', 0), # Windows only
2030 os
.O_CREAT
if writable
else 0, # O_TRUNC only after locking
2031 os
.O_APPEND
if 'a' in mode
else 0,
2032 os
.O_EXCL
if 'x' in mode
else 0,
2033 os
.O_RDONLY
if not writable
else os
.O_RDWR
if readable
else os
.O_WRONLY
,
2036 self
.f
= os
.fdopen(os
.open(filename
, flags
, 0o666), mode
, encoding
=encoding
)
2038 def __enter__(self
):
2039 exclusive
= 'r' not in self
.mode
2041 _lock_file(self
.f
, exclusive
, self
.block
)
2046 if 'w' in self
.mode
:
2049 except OSError as e
:
2051 errno
.ESPIPE
, # Illegal seek - expected for FIFO
2052 errno
.EINVAL
, # Invalid argument - expected for /dev/null
2061 _unlock_file(self
.f
)
2065 def __exit__(self
, *_
):
2074 def __getattr__(self
, attr
):
2075 return getattr(self
.f
, attr
)
2082 def get_filesystem_encoding():
2083 encoding
= sys
.getfilesystemencoding()
2084 return encoding
if encoding
is not None else 'utf-8'
2087 def shell_quote(args
):
2089 encoding
= get_filesystem_encoding()
2091 if isinstance(a
, bytes):
2092 # We may get a filename encoded with 'encodeFilename'
2093 a
= a
.decode(encoding
)
2094 quoted_args
.append(compat_shlex_quote(a
))
2095 return ' '.join(quoted_args
)
2098 def smuggle_url(url
, data
):
2099 """ Pass additional data in a URL for internal use. """
2101 url
, idata
= unsmuggle_url(url
, {})
2103 sdata
= urllib
.parse
.urlencode(
2104 {'__youtubedl_smuggle': json.dumps(data)}
)
2105 return url
+ '#' + sdata
2108 def unsmuggle_url(smug_url
, default
=None):
2109 if '#__youtubedl_smuggle' not in smug_url
:
2110 return smug_url
, default
2111 url
, _
, sdata
= smug_url
.rpartition('#')
2112 jsond
= urllib
.parse
.parse_qs(sdata
)['__youtubedl_smuggle'][0]
2113 data
= json
.loads(jsond
)
2117 def format_decimal_suffix(num
, fmt
='%d%s', *, factor
=1000):
2118 """ Formats numbers with decimal sufixes like K, M, etc """
2119 num
, factor
= float_or_none(num
), float(factor
)
2120 if num
is None or num
< 0:
2122 POSSIBLE_SUFFIXES
= 'kMGTPEZY'
2123 exponent
= 0 if num
== 0 else min(int(math
.log(num
, factor
)), len(POSSIBLE_SUFFIXES
))
2124 suffix
= ['', *POSSIBLE_SUFFIXES
][exponent
]
2126 suffix
= {'k': 'Ki', '': ''}
.get(suffix
, f
'{suffix}i')
2127 converted
= num
/ (factor
** exponent
)
2128 return fmt
% (converted
, suffix
)
2131 def format_bytes(bytes):
2132 return format_decimal_suffix(bytes, '%.2f%sB', factor
=1024) or 'N/A'
2135 def lookup_unit_table(unit_table
, s
, strict
=False):
2136 num_re
= NUMBER_RE
if strict
else NUMBER_RE
.replace(R
'\.', '[,.]')
2137 units_re
= '|'.join(re
.escape(u
) for u
in unit_table
)
2138 m
= (re
.fullmatch
if strict
else re
.match
)(
2139 rf
'(?P<num>{num_re})\s*(?P<unit>{units_re})\b', s
)
2143 num
= float(m
.group('num').replace(',', '.'))
2144 mult
= unit_table
[m
.group('unit')]
2145 return round(num
* mult
)
2149 """Parse a string indicating a byte quantity into an integer"""
2150 return lookup_unit_table(
2151 {u: 1024**i for i, u in enumerate(['', *'KMGTPEZY'])}
,
2152 s
.upper(), strict
=True)
2155 def parse_filesize(s
):
2159 # The lower-case forms are of course incorrect and unofficial,
2160 # but we support those too
2177 'megabytes': 1000 ** 2,
2178 'mebibytes': 1024 ** 2,
2184 'gigabytes': 1000 ** 3,
2185 'gibibytes': 1024 ** 3,
2191 'terabytes': 1000 ** 4,
2192 'tebibytes': 1024 ** 4,
2198 'petabytes': 1000 ** 5,
2199 'pebibytes': 1024 ** 5,
2205 'exabytes': 1000 ** 6,
2206 'exbibytes': 1024 ** 6,
2212 'zettabytes': 1000 ** 7,
2213 'zebibytes': 1024 ** 7,
2219 'yottabytes': 1000 ** 8,
2220 'yobibytes': 1024 ** 8,
2223 return lookup_unit_table(_UNIT_TABLE
, s
)
2230 s
= re
.sub(r
'^[^\d]+\s', '', s
).strip()
2232 if re
.match(r
'^[\d,.]+$', s
):
2233 return str_to_int(s
)
2246 ret
= lookup_unit_table(_UNIT_TABLE
, s
)
2250 mobj
= re
.match(r
'([\d,.]+)(?:$|\s)', s
)
2252 return str_to_int(mobj
.group(1))
2255 def parse_resolution(s
, *, lenient
=False):
2260 mobj
= re
.search(r
'(?P<w>\d+)\s*[xX×,]\s*(?P<h>\d+)', s
)
2262 mobj
= re
.search(r
'(?<![a-zA-Z0-9])(?P<w>\d+)\s*[xX×,]\s*(?P<h>\d+)(?![a-zA-Z0-9])', s
)
2265 'width': int(mobj
.group('w')),
2266 'height': int(mobj
.group('h')),
2269 mobj
= re
.search(r
'(?<![a-zA-Z0-9])(\d+)[pPiI](?![a-zA-Z0-9])', s
)
2271 return {'height': int(mobj.group(1))}
2273 mobj
= re
.search(r
'\b([48])[kK]\b', s
)
2275 return {'height': int(mobj.group(1)) * 540}
2280 def parse_bitrate(s
):
2281 if not isinstance(s
, str):
2283 mobj
= re
.search(r
'\b(\d+)\s*kbps', s
)
2285 return int(mobj
.group(1))
2288 def month_by_name(name
, lang
='en'):
2289 """ Return the number of a month by (locale-independently) English name """
2291 month_names
= MONTH_NAMES
.get(lang
, MONTH_NAMES
['en'])
2294 return month_names
.index(name
) + 1
2299 def month_by_abbreviation(abbrev
):
2300 """ Return the number of a month by (locale-independently) English
2304 return [s
[:3] for s
in ENGLISH_MONTH_NAMES
].index(abbrev
) + 1
2309 def fix_xml_ampersands(xml_str
):
2310 """Replace all the '&' by '&' in XML"""
2312 r
'&(?!amp;|lt;|gt;|apos;|quot;|#x[0-9a-fA-F]{,4};|#[0-9]{,4};)',
2317 def setproctitle(title
):
2318 assert isinstance(title
, str)
2320 # Workaround for https://github.com/yt-dlp/yt-dlp/issues/4541
2327 libc
= ctypes
.cdll
.LoadLibrary('libc.so.6')
2331 # LoadLibrary in Windows Python 2.7.13 only expects
2332 # a bytestring, but since unicode_literals turns
2333 # every string into a unicode string, it fails.
2335 title_bytes
= title
.encode()
2336 buf
= ctypes
.create_string_buffer(len(title_bytes
))
2337 buf
.value
= title_bytes
2339 libc
.prctl(15, buf
, 0, 0, 0)
2340 except AttributeError:
2341 return # Strange libc, just skip this
2344 def remove_start(s
, start
):
2345 return s
[len(start
):] if s
is not None and s
.startswith(start
) else s
2348 def remove_end(s
, end
):
2349 return s
[:-len(end
)] if s
is not None and s
.endswith(end
) else s
2352 def remove_quotes(s
):
2353 if s
is None or len(s
) < 2:
2355 for quote
in ('"', "'", ):
2356 if s
[0] == quote
and s
[-1] == quote
:
2361 def get_domain(url
):
2363 This implementation is inconsistent, but is kept for compatibility.
2364 Use this only for "webpage_url_domain"
2366 return remove_start(urllib
.parse
.urlparse(url
).netloc
, 'www.') or None
2369 def url_basename(url
):
2370 path
= urllib
.parse
.urlparse(url
).path
2371 return path
.strip('/').split('/')[-1]
2375 return re
.match(r
'https?://[^?#]+/', url
).group()
2378 def urljoin(base
, path
):
2379 if isinstance(path
, bytes):
2380 path
= path
.decode()
2381 if not isinstance(path
, str) or not path
:
2383 if re
.match(r
'^(?:[a-zA-Z][a-zA-Z0-9+-.]*:)?//', path
):
2385 if isinstance(base
, bytes):
2386 base
= base
.decode()
2387 if not isinstance(base
, str) or not re
.match(
2388 r
'^(?:https?:)?//', base
):
2390 return urllib
.parse
.urljoin(base
, path
)
2393 class HEADRequest(urllib
.request
.Request
):
2394 def get_method(self
):
2398 class PUTRequest(urllib
.request
.Request
):
2399 def get_method(self
):
2403 def int_or_none(v
, scale
=1, default
=None, get_attr
=None, invscale
=1):
2404 if get_attr
and v
is not None:
2405 v
= getattr(v
, get_attr
, None)
2407 return int(v
) * invscale
// scale
2408 except (ValueError, TypeError, OverflowError):
2412 def str_or_none(v
, default
=None):
2413 return default
if v
is None else str(v
)
2416 def str_to_int(int_str
):
2417 """ A more relaxed version of int_or_none """
2418 if isinstance(int_str
, int):
2420 elif isinstance(int_str
, str):
2421 int_str
= re
.sub(r
'[,\.\+]', '', int_str
)
2422 return int_or_none(int_str
)
2425 def float_or_none(v
, scale
=1, invscale
=1, default
=None):
2429 return float(v
) * invscale
/ scale
2430 except (ValueError, TypeError):
2434 def bool_or_none(v
, default
=None):
2435 return v
if isinstance(v
, bool) else default
2438 def strip_or_none(v
, default
=None):
2439 return v
.strip() if isinstance(v
, str) else default
2442 def url_or_none(url
):
2443 if not url
or not isinstance(url
, str):
2446 return url
if re
.match(r
'^(?:(?:https?|rt(?:m(?:pt?[es]?|fp)|sp[su]?)|mms|ftps?):)?//', url
) else None
2449 def request_to_url(req
):
2450 if isinstance(req
, urllib
.request
.Request
):
2451 return req
.get_full_url()
2456 def strftime_or_none(timestamp
, date_format
='%Y%m%d', default
=None):
2457 datetime_object
= None
2459 if isinstance(timestamp
, (int, float)): # unix timestamp
2460 # Using naive datetime here can break timestamp() in Windows
2461 # Ref: https://github.com/yt-dlp/yt-dlp/issues/5185, https://github.com/python/cpython/issues/94414
2462 # Also, datetime.datetime.fromtimestamp breaks for negative timestamps
2463 # Ref: https://github.com/yt-dlp/yt-dlp/issues/6706#issuecomment-1496842642
2464 datetime_object
= (datetime
.datetime
.fromtimestamp(0, datetime
.timezone
.utc
)
2465 + datetime
.timedelta(seconds
=timestamp
))
2466 elif isinstance(timestamp
, str): # assume YYYYMMDD
2467 datetime_object
= datetime
.datetime
.strptime(timestamp
, '%Y%m%d')
2468 date_format
= re
.sub( # Support %s on windows
2469 r
'(?<!%)(%%)*%s', rf
'\g<1>{int(datetime_object.timestamp())}', date_format
)
2470 return datetime_object
.strftime(date_format
)
2471 except (ValueError, TypeError, AttributeError):
2475 def parse_duration(s
):
2476 if not isinstance(s
, str):
2482 days
, hours
, mins
, secs
, ms
= [None] * 5
2483 m
= re
.match(r
'''(?x)
2485 (?:(?:(?P<days>[0-9]+):)?(?P<hours>[0-9]+):)?(?P<mins>[0-9]+):)?
2486 (?P<secs>(?(before_secs)[0-9]{1,2}|[0-9]+))
2487 (?P<ms>[.:][0-9]+)?Z?$
2490 days
, hours
, mins
, secs
, ms
= m
.group('days', 'hours', 'mins', 'secs', 'ms')
2495 [0-9]+\s*y(?:ears?)?,?\s*
2498 [0-9]+\s*m(?:onths?)?,?\s*
2501 [0-9]+\s*w(?:eeks?)?,?\s*
2504 (?P<days>[0-9]+)\s*d(?:ays?)?,?\s*
2508 (?P<hours>[0-9]+)\s*h(?:ours?)?,?\s*
2511 (?P<mins>[0-9]+)\s*m(?:in(?:ute)?s?)?,?\s*
2514 (?P<secs>[0-9]+)(?P<ms>\.[0-9]+)?\s*s(?:ec(?:ond)?s?)?\s*
2517 days
, hours
, mins
, secs
, ms
= m
.groups()
2519 m
= re
.match(r
'(?i)(?:(?P<hours>[0-9.]+)\s*(?:hours?)|(?P<mins>[0-9.]+)\s*(?:mins?\.?|minutes?)\s*)Z?$', s
)
2521 hours
, mins
= m
.groups()
2526 ms
= ms
.replace(':', '.')
2527 return sum(float(part
or 0) * mult
for part
, mult
in (
2528 (days
, 86400), (hours
, 3600), (mins
, 60), (secs
, 1), (ms
, 1)))
2531 def prepend_extension(filename
, ext
, expected_real_ext
=None):
2532 name
, real_ext
= os
.path
.splitext(filename
)
2534 f
'{name}.{ext}{real_ext}'
2535 if not expected_real_ext
or real_ext
[1:] == expected_real_ext
2536 else f
'{filename}.{ext}')
2539 def replace_extension(filename
, ext
, expected_real_ext
=None):
2540 name
, real_ext
= os
.path
.splitext(filename
)
2541 return '{}.{}'.format(
2542 name
if not expected_real_ext
or real_ext
[1:] == expected_real_ext
else filename
,
2546 def check_executable(exe
, args
=[]):
2547 """ Checks if the given binary is installed somewhere in PATH, and returns its name.
2548 args can be a list of arguments for a short output (like -version) """
2550 Popen
.run([exe
] + args
, stdout
=subprocess
.PIPE
, stderr
=subprocess
.PIPE
)
2556 def _get_exe_version_output(exe
, args
):
2558 # STDIN should be redirected too. On UNIX-like systems, ffmpeg triggers
2559 # SIGTTOU if yt-dlp is run in the background.
2560 # See https://github.com/ytdl-org/youtube-dl/issues/955#issuecomment-209789656
2561 stdout
, _
, ret
= Popen
.run([encodeArgument(exe
)] + args
, text
=True,
2562 stdin
=subprocess
.PIPE
, stdout
=subprocess
.PIPE
, stderr
=subprocess
.STDOUT
)
2570 def detect_exe_version(output
, version_re
=None, unrecognized
='present'):
2571 assert isinstance(output
, str)
2572 if version_re
is None:
2573 version_re
= r
'version\s+([-0-9._a-zA-Z]+)'
2574 m
= re
.search(version_re
, output
)
2581 def get_exe_version(exe
, args
=['--version'],
2582 version_re
=None, unrecognized
=('present', 'broken')):
2583 """ Returns the version of the specified executable,
2584 or False if the executable is not present """
2585 unrecognized
= variadic(unrecognized
)
2586 assert len(unrecognized
) in (1, 2)
2587 out
= _get_exe_version_output(exe
, args
)
2589 return unrecognized
[-1]
2590 return out
and detect_exe_version(out
, version_re
, unrecognized
[0])
2593 def frange(start
=0, stop
=None, step
=1):
2596 start
, stop
= 0, start
2597 sign
= [-1, 1][step
> 0] if step
else 0
2598 while sign
* start
< sign
* stop
:
2603 class LazyList(collections
.abc
.Sequence
):
2604 """Lazy immutable list from an iterable
2605 Note that slices of a LazyList are lists and not LazyList"""
2607 class IndexError(IndexError):
2610 def __init__(self
, iterable
, *, reverse
=False, _cache
=None):
2611 self
._iterable
= iter(iterable
)
2612 self
._cache
= [] if _cache
is None else _cache
2613 self
._reversed
= reverse
2617 # We need to consume the entire iterable to iterate in reverse
2618 yield from self
.exhaust()
2620 yield from self
._cache
2621 for item
in self
._iterable
:
2622 self
._cache
.append(item
)
2626 self
._cache
.extend(self
._iterable
)
2627 self
._iterable
= [] # Discard the emptied iterable to make it pickle-able
2631 """Evaluate the entire iterable"""
2632 return self
._exhaust
()[::-1 if self
._reversed
else 1]
2635 def _reverse_index(x
):
2636 return None if x
is None else ~x
2638 def __getitem__(self
, idx
):
2639 if isinstance(idx
, slice):
2641 idx
= slice(self
._reverse
_index
(idx
.start
), self
._reverse
_index
(idx
.stop
), -(idx
.step
or 1))
2642 start
, stop
, step
= idx
.start
, idx
.stop
, idx
.step
or 1
2643 elif isinstance(idx
, int):
2645 idx
= self
._reverse
_index
(idx
)
2646 start
, stop
, step
= idx
, idx
, 0
2648 raise TypeError('indices must be integers or slices')
2649 if ((start
or 0) < 0 or (stop
or 0) < 0
2650 or (start
is None and step
< 0)
2651 or (stop
is None and step
> 0)):
2652 # We need to consume the entire iterable to be able to slice from the end
2653 # Obviously, never use this with infinite iterables
2656 return self
._cache
[idx
]
2657 except IndexError as e
:
2658 raise self
.IndexError(e
) from e
2659 n
= max(start
or 0, stop
or 0) - len(self
._cache
) + 1
2661 self
._cache
.extend(itertools
.islice(self
._iterable
, n
))
2663 return self
._cache
[idx
]
2664 except IndexError as e
:
2665 raise self
.IndexError(e
) from e
2669 self
[-1] if self
._reversed
else self
[0]
2670 except self
.IndexError:
2676 return len(self
._cache
)
2678 def __reversed__(self
):
2679 return type(self
)(self
._iterable
, reverse
=not self
._reversed
, _cache
=self
._cache
)
2682 return type(self
)(self
._iterable
, reverse
=self
._reversed
, _cache
=self
._cache
)
2685 # repr and str should mimic a list. So we exhaust the iterable
2686 return repr(self
.exhaust())
2689 return repr(self
.exhaust())
2694 class IndexError(IndexError):
2698 # This is only useful for tests
2699 return len(self
.getslice())
2701 def __init__(self
, pagefunc
, pagesize
, use_cache
=True):
2702 self
._pagefunc
= pagefunc
2703 self
._pagesize
= pagesize
2704 self
._pagecount
= float('inf')
2705 self
._use
_cache
= use_cache
2708 def getpage(self
, pagenum
):
2709 page_results
= self
._cache
.get(pagenum
)
2710 if page_results
is None:
2711 page_results
= [] if pagenum
> self
._pagecount
else list(self
._pagefunc
(pagenum
))
2713 self
._cache
[pagenum
] = page_results
2716 def getslice(self
, start
=0, end
=None):
2717 return list(self
._getslice
(start
, end
))
2719 def _getslice(self
, start
, end
):
2720 raise NotImplementedError('This method must be implemented by subclasses')
2722 def __getitem__(self
, idx
):
2723 assert self
._use
_cache
, 'Indexing PagedList requires cache'
2724 if not isinstance(idx
, int) or idx
< 0:
2725 raise TypeError('indices must be non-negative integers')
2726 entries
= self
.getslice(idx
, idx
+ 1)
2728 raise self
.IndexError()
2732 class OnDemandPagedList(PagedList
):
2733 """Download pages until a page with less than maximum results"""
2735 def _getslice(self
, start
, end
):
2736 for pagenum
in itertools
.count(start
// self
._pagesize
):
2737 firstid
= pagenum
* self
._pagesize
2738 nextfirstid
= pagenum
* self
._pagesize
+ self
._pagesize
2739 if start
>= nextfirstid
:
2743 start
% self
._pagesize
2744 if firstid
<= start
< nextfirstid
2747 ((end
- 1) % self
._pagesize
) + 1
2748 if (end
is not None and firstid
<= end
<= nextfirstid
)
2752 page_results
= self
.getpage(pagenum
)
2754 self
._pagecount
= pagenum
- 1
2756 if startv
!= 0 or endv
is not None:
2757 page_results
= page_results
[startv
:endv
]
2758 yield from page_results
2760 # A little optimization - if current page is not "full", ie. does
2761 # not contain page_size videos then we can assume that this page
2762 # is the last one - there are no more ids on further pages -
2763 # i.e. no need to query again.
2764 if len(page_results
) + startv
< self
._pagesize
:
2767 # If we got the whole page, but the next page is not interesting,
2768 # break out early as well
2769 if end
== nextfirstid
:
2773 class InAdvancePagedList(PagedList
):
2774 """PagedList with total number of pages known in advance"""
2776 def __init__(self
, pagefunc
, pagecount
, pagesize
):
2777 PagedList
.__init
__(self
, pagefunc
, pagesize
, True)
2778 self
._pagecount
= pagecount
2780 def _getslice(self
, start
, end
):
2781 start_page
= start
// self
._pagesize
2782 end_page
= self
._pagecount
if end
is None else min(self
._pagecount
, end
// self
._pagesize
+ 1)
2783 skip_elems
= start
- start_page
* self
._pagesize
2784 only_more
= None if end
is None else end
- start
2785 for pagenum
in range(start_page
, end_page
):
2786 page_results
= self
.getpage(pagenum
)
2788 page_results
= page_results
[skip_elems
:]
2790 if only_more
is not None:
2791 if len(page_results
) < only_more
:
2792 only_more
-= len(page_results
)
2794 yield from page_results
[:only_more
]
2796 yield from page_results
2799 class PlaylistEntries
:
2800 MissingEntry
= object()
2801 is_exhausted
= False
2803 def __init__(self
, ydl
, info_dict
):
2806 # _entries must be assigned now since infodict can change during iteration
2807 entries
= info_dict
.get('entries')
2809 raise EntryNotInPlaylist('There are no entries')
2810 elif isinstance(entries
, list):
2811 self
.is_exhausted
= True
2813 requested_entries
= info_dict
.get('requested_entries')
2814 self
.is_incomplete
= requested_entries
is not None
2815 if self
.is_incomplete
:
2816 assert self
.is_exhausted
2817 self
._entries
= [self
.MissingEntry
] * max(requested_entries
or [0])
2818 for i
, entry
in zip(requested_entries
, entries
):
2819 self
._entries
[i
- 1] = entry
2820 elif isinstance(entries
, (list, PagedList
, LazyList
)):
2821 self
._entries
= entries
2823 self
._entries
= LazyList(entries
)
2825 PLAYLIST_ITEMS_RE
= re
.compile(r
'''(?x)
2826 (?P<start>[+-]?\d+)?
2828 (?P<end>[+-]?\d+|inf(?:inite)?)?
2829 (?::(?P<step>[+-]?\d+))?
2833 def parse_playlist_items(cls
, string
):
2834 for segment
in string
.split(','):
2836 raise ValueError('There is two or more consecutive commas')
2837 mobj
= cls
.PLAYLIST_ITEMS_RE
.fullmatch(segment
)
2839 raise ValueError(f
'{segment!r} is not a valid specification')
2840 start
, end
, step
, has_range
= mobj
.group('start', 'end', 'step', 'range')
2841 if int_or_none(step
) == 0:
2842 raise ValueError(f
'Step in {segment!r} cannot be zero')
2843 yield slice(int_or_none(start
), float_or_none(end
), int_or_none(step
)) if has_range
else int(start
)
2845 def get_requested_items(self
):
2846 playlist_items
= self
.ydl
.params
.get('playlist_items')
2847 playlist_start
= self
.ydl
.params
.get('playliststart', 1)
2848 playlist_end
= self
.ydl
.params
.get('playlistend')
2849 # For backwards compatibility, interpret -1 as whole list
2850 if playlist_end
in (-1, None):
2852 if not playlist_items
:
2853 playlist_items
= f
'{playlist_start}:{playlist_end}'
2854 elif playlist_start
!= 1 or playlist_end
:
2855 self
.ydl
.report_warning('Ignoring playliststart and playlistend because playlistitems was given', only_once
=True)
2857 for index
in self
.parse_playlist_items(playlist_items
):
2858 for i
, entry
in self
[index
]:
2863 # The item may have just been added to archive. Don't break due to it
2864 if not self
.ydl
.params
.get('lazy_playlist'):
2865 # TODO: Add auto-generated fields
2866 self
.ydl
._match
_entry
(entry
, incomplete
=True, silent
=True)
2867 except (ExistingVideoReached
, RejectedVideoReached
):
2870 def get_full_count(self
):
2871 if self
.is_exhausted
and not self
.is_incomplete
:
2873 elif isinstance(self
._entries
, InAdvancePagedList
):
2874 if self
._entries
._pagesize
== 1:
2875 return self
._entries
._pagecount
2877 @functools.cached_property
2879 if isinstance(self
._entries
, list):
2882 entry
= self
._entries
[i
]
2884 entry
= self
.MissingEntry
2885 if not self
.is_incomplete
:
2886 raise self
.IndexError()
2887 if entry
is self
.MissingEntry
:
2888 raise EntryNotInPlaylist(f
'Entry {i + 1} cannot be found')
2893 return type(self
.ydl
)._handle
_extraction
_exceptions
(lambda _
, i
: self
._entries
[i
])(self
.ydl
, i
)
2894 except (LazyList
.IndexError, PagedList
.IndexError):
2895 raise self
.IndexError()
2898 def __getitem__(self
, idx
):
2899 if isinstance(idx
, int):
2900 idx
= slice(idx
, idx
)
2902 # NB: PlaylistEntries[1:10] => (0, 1, ... 9)
2903 step
= 1 if idx
.step
is None else idx
.step
2904 if idx
.start
is None:
2905 start
= 0 if step
> 0 else len(self
) - 1
2907 start
= idx
.start
- 1 if idx
.start
>= 0 else len(self
) + idx
.start
2909 # NB: Do not call len(self) when idx == [:]
2910 if idx
.stop
is None:
2911 stop
= 0 if step
< 0 else float('inf')
2913 stop
= idx
.stop
- 1 if idx
.stop
>= 0 else len(self
) + idx
.stop
2914 stop
+= [-1, 1][step
> 0]
2916 for i
in frange(start
, stop
, step
):
2920 entry
= self
._getter
(i
)
2921 except self
.IndexError:
2922 self
.is_exhausted
= True
2929 return len(tuple(self
[:]))
2931 class IndexError(IndexError):
2935 def uppercase_escape(s
):
2936 unicode_escape
= codecs
.getdecoder('unicode_escape')
2938 r
'\\U[0-9a-fA-F]{8}',
2939 lambda m
: unicode_escape(m
.group(0))[0],
2943 def lowercase_escape(s
):
2944 unicode_escape
= codecs
.getdecoder('unicode_escape')
2946 r
'\\u[0-9a-fA-F]{4}',
2947 lambda m
: unicode_escape(m
.group(0))[0],
2951 def escape_rfc3986(s
):
2952 """Escape non-ASCII characters as suggested by RFC 3986"""
2953 return urllib
.parse
.quote(s
, b
"%/;:@&=+$,!~*'()?#[]")
2956 def escape_url(url
):
2957 """Escape URL as suggested by RFC 3986"""
2958 url_parsed
= urllib
.parse
.urlparse(url
)
2959 return url_parsed
._replace
(
2960 netloc
=url_parsed
.netloc
.encode('idna').decode('ascii'),
2961 path
=escape_rfc3986(url_parsed
.path
),
2962 params
=escape_rfc3986(url_parsed
.params
),
2963 query
=escape_rfc3986(url_parsed
.query
),
2964 fragment
=escape_rfc3986(url_parsed
.fragment
)
2968 def parse_qs(url
, **kwargs
):
2969 return urllib
.parse
.parse_qs(urllib
.parse
.urlparse(url
).query
, **kwargs
)
2972 def read_batch_urls(batch_fd
):
2974 if not isinstance(url
, str):
2975 url
= url
.decode('utf-8', 'replace')
2976 BOM_UTF8
= ('\xef\xbb\xbf', '\ufeff')
2977 for bom
in BOM_UTF8
:
2978 if url
.startswith(bom
):
2979 url
= url
[len(bom
):]
2981 if not url
or url
.startswith(('#', ';', ']')):
2983 # "#" cannot be stripped out since it is part of the URI
2984 # However, it can be safely stripped out if following a whitespace
2985 return re
.split(r
'\s#', url
, 1)[0].rstrip()
2987 with contextlib
.closing(batch_fd
) as fd
:
2988 return [url
for url
in map(fixup
, fd
) if url
]
2991 def urlencode_postdata(*args
, **kargs
):
2992 return urllib
.parse
.urlencode(*args
, **kargs
).encode('ascii')
2995 def update_url(url
, *, query_update
=None, **kwargs
):
2996 """Replace URL components specified by kwargs
2997 @param url str or parse url tuple
2998 @param query_update update query
3001 if isinstance(url
, str):
3002 if not kwargs
and not query_update
:
3005 url
= urllib
.parse
.urlparse(url
)
3007 assert 'query' not in kwargs
, 'query_update and query cannot be specified at the same time'
3008 kwargs
['query'] = urllib
.parse
.urlencode({
3009 **urllib
.parse
.parse_qs(url
.query
),
3012 return urllib
.parse
.urlunparse(url
._replace
(**kwargs
))
3015 def update_url_query(url
, query
):
3016 return update_url(url
, query_update
=query
)
3019 def update_Request(req
, url
=None, data
=None, headers
=None, query
=None):
3020 req_headers
= req
.headers
.copy()
3021 req_headers
.update(headers
or {})
3022 req_data
= data
or req
.data
3023 req_url
= update_url_query(url
or req
.get_full_url(), query
)
3024 req_get_method
= req
.get_method()
3025 if req_get_method
== 'HEAD':
3026 req_type
= HEADRequest
3027 elif req_get_method
== 'PUT':
3028 req_type
= PUTRequest
3030 req_type
= urllib
.request
.Request
3032 req_url
, data
=req_data
, headers
=req_headers
,
3033 origin_req_host
=req
.origin_req_host
, unverifiable
=req
.unverifiable
)
3034 if hasattr(req
, 'timeout'):
3035 new_req
.timeout
= req
.timeout
3039 def _multipart_encode_impl(data
, boundary
):
3040 content_type
= 'multipart/form-data; boundary=%s' % boundary
3043 for k
, v
in data
.items():
3044 out
+= b
'--' + boundary
.encode('ascii') + b
'\r\n'
3045 if isinstance(k
, str):
3047 if isinstance(v
, str):
3049 # RFC 2047 requires non-ASCII field names to be encoded, while RFC 7578
3050 # suggests sending UTF-8 directly. Firefox sends UTF-8, too
3051 content
= b
'Content-Disposition: form-data; name="' + k
+ b
'"\r\n\r\n' + v
+ b
'\r\n'
3052 if boundary
.encode('ascii') in content
:
3053 raise ValueError('Boundary overlaps with data')
3056 out
+= b
'--' + boundary
.encode('ascii') + b
'--\r\n'
3058 return out
, content_type
3061 def multipart_encode(data
, boundary
=None):
3063 Encode a dict to RFC 7578-compliant form-data
3066 A dict where keys and values can be either Unicode or bytes-like
3069 If specified a Unicode object, it's used as the boundary. Otherwise
3070 a random boundary is generated.
3072 Reference: https://tools.ietf.org/html/rfc7578
3074 has_specified_boundary
= boundary
is not None
3077 if boundary
is None:
3078 boundary
= '---------------' + str(random
.randrange(0x0fffffff, 0xffffffff))
3081 out
, content_type
= _multipart_encode_impl(data
, boundary
)
3084 if has_specified_boundary
:
3088 return out
, content_type
3091 def is_iterable_like(x
, allowed_types
=collections
.abc
.Iterable
, blocked_types
=NO_DEFAULT
):
3092 if blocked_types
is NO_DEFAULT
:
3093 blocked_types
= (str, bytes, collections
.abc
.Mapping
)
3094 return isinstance(x
, allowed_types
) and not isinstance(x
, blocked_types
)
3097 def variadic(x
, allowed_types
=NO_DEFAULT
):
3098 if not isinstance(allowed_types
, (tuple, type)):
3099 deprecation_warning('allowed_types should be a tuple or a type')
3100 allowed_types
= tuple(allowed_types
)
3101 return x
if is_iterable_like(x
, blocked_types
=allowed_types
) else (x
, )
3104 def try_call(*funcs
, expected_type
=None, args
=[], kwargs
={}):
3107 val
= f(*args
, **kwargs
)
3108 except (AttributeError, KeyError, TypeError, IndexError, ValueError, ZeroDivisionError):
3111 if expected_type
is None or isinstance(val
, expected_type
):
3115 def try_get(src
, getter
, expected_type
=None):
3116 return try_call(*variadic(getter
), args
=(src
,), expected_type
=expected_type
)
3119 def filter_dict(dct
, cndn
=lambda _
, v
: v
is not None):
3120 return {k: v for k, v in dct.items() if cndn(k, v)}
3123 def merge_dicts(*dicts
):
3125 for a_dict
in dicts
:
3126 for k
, v
in a_dict
.items():
3127 if (v
is not None and k
not in merged
3128 or isinstance(v
, str) and merged
[k
] == ''):
3133 def encode_compat_str(string
, encoding
=preferredencoding(), errors
='strict'):
3134 return string
if isinstance(string
, str) else str(string
, encoding
, errors
)
3146 TV_PARENTAL_GUIDELINES
= {
3156 def parse_age_limit(s
):
3157 # isinstance(False, int) is True. So type() must be used instead
3158 if type(s
) is int: # noqa: E721
3159 return s
if 0 <= s
<= 21 else None
3160 elif not isinstance(s
, str):
3162 m
= re
.match(r
'^(?P<age>\d{1,2})\+?$', s
)
3164 return int(m
.group('age'))
3167 return US_RATINGS
[s
]
3168 m
= re
.match(r
'^TV[_-]?(%s)$' % '|'.join(k
[3:] for k
in TV_PARENTAL_GUIDELINES
), s
)
3170 return TV_PARENTAL_GUIDELINES
['TV-' + m
.group(1)]
3174 def strip_jsonp(code
):
3177 (?:window\.)?(?P<func_name>[a-zA-Z0-9_.$]*)
3178 (?:\s*&&\s*(?P=func_name))?
3179 \s*\(\s*(?P<callback_data>.*)\);?
3180 \s*?(?://[^\n]*)*$''',
3181 r
'\g<callback_data>', code
)
3184 def js_to_json(code
, vars={}, *, strict
=False):
3185 # vars is a dict of var, val pairs to substitute
3186 STRING_QUOTES
= '\'"`'
3187 STRING_RE
= '|'.join(rf
'{q}(?:\\.|[^\\{q}])*{q}' for q
in STRING_QUOTES
)
3188 COMMENT_RE
= r
'/\*(?:(?!\*/).)*?\*/|//[^\n]*\n'
3189 SKIP_RE
= fr
'\s*(?:{COMMENT_RE})?\s*'
3191 (fr
'(?s)^(0[xX][0-9a-fA-F]+){SKIP_RE}:?$', 16),
3192 (fr
'(?s)^(0+[0-7]+){SKIP_RE}:?$', 8),
3195 def process_escape(match
):
3196 JSON_PASSTHROUGH_ESCAPES
= R
'"\bfnrtu'
3197 escape
= match
.group(1) or match
.group(2)
3199 return (Rf
'\{escape}' if escape
in JSON_PASSTHROUGH_ESCAPES
3200 else R
'\u00' if escape
== 'x'
3201 else '' if escape
== '\n'
3204 def template_substitute(match
):
3205 evaluated
= js_to_json(match
.group(1), vars, strict
=strict
)
3206 if evaluated
[0] == '"':
3207 return json
.loads(evaluated
)
3212 if v
in ('true', 'false', 'null'):
3214 elif v
in ('undefined', 'void 0'):
3216 elif v
.startswith('/*') or v
.startswith('//') or v
.startswith('!') or v
== ',':
3219 if v
[0] in STRING_QUOTES
:
3220 v
= re
.sub(r
'(?s)\${([^}]+)}', template_substitute
, v
[1:-1]) if v
[0] == '`' else v
[1:-1]
3221 escaped
= re
.sub(r
'(?s)(")|\\(.)', process_escape
, v
)
3222 return f
'"{escaped}"'
3224 for regex
, base
in INTEGER_TABLE
:
3225 im
= re
.match(regex
, v
)
3227 i
= int(im
.group(1), base
)
3228 return f
'"{i}":' if v
.endswith(':') else str(i
)
3234 except json
.JSONDecodeError
:
3235 return json
.dumps(vars[v
])
3242 raise ValueError(f
'Unknown value: {v}')
3244 def create_map(mobj
):
3245 return json
.dumps(dict(json
.loads(js_to_json(mobj
.group(1) or '[]', vars=vars))))
3247 code
= re
.sub(r
'new Map\((\[.*?\])?\)', create_map
, code
)
3249 code
= re
.sub(r
'new Date\((".+")\)', r
'\g<1>', code
)
3250 code
= re
.sub(r
'new \w+\((.*?)\)', lambda m
: json
.dumps(m
.group(0)), code
)
3251 code
= re
.sub(r
'parseInt\([^\d]+(\d+)[^\d]+\)', r
'\1', code
)
3252 code
= re
.sub(r
'\(function\([^)]*\)\s*\{[^}]*\}\s*\)\s*\(\s*(["\'][^
)]*["\'])\s*\)', r'\1', code)
3254 return re.sub(rf'''(?sx)
3256 {COMMENT_RE}|,(?={SKIP_RE}[\]}}])|
3257 void\s0|(?:(?<![0-9])[eE]|[a-df-zA-DF-Z_$])[.a-zA-Z_$0-9]*|
3258 \b(?:0[xX][0-9a-fA-F]+|0+[0-7]+)(?:{SKIP_RE}:)?|
3259 [0-9]+(?={SKIP_RE}:)|
3264 def qualities(quality_ids):
3265 """ Get a numeric quality value out of a list of possible values """
3268 return quality_ids.index(qid)
3274 POSTPROCESS_WHEN = ('pre_process', 'after_filter', 'video', 'before_dl', 'post_process', 'after_move', 'after_video', 'playlist')
3278 'default': '%(title)s [%(id)s].%(ext)s',
3279 'chapter': '%(title)s - %(section_number)03d %(section_title)s [%(id)s].%(ext)s',
3285 'description': 'description',
3286 'annotation': 'annotations.xml',
3287 'infojson': 'info.json',
3290 'pl_thumbnail': None,
3291 'pl_description': 'description',
3292 'pl_infojson': 'info.json',
3295 # As of [1] format syntax is:
3296 # %[mapping_key][conversion_flags][minimum_width][.precision][length_modifier]type
3297 # 1. https://docs.python.org/2/library/stdtypes.html#string-formatting
3298 STR_FORMAT_RE_TMPL = r'''(?x)
3299 (?<!%)(?P<prefix>(?:%%)*)
3301 (?P<has_key>\((?P<key>{0})\))?
3303 (?P<conversion>[#0\-+ ]+)?
3305 (?P<precision>\.\d+)?
3306 (?P<len_mod>[hlL])? # unused in python
3307 {1} # conversion type
3312 STR_FORMAT_TYPES = 'diouxXeEfFgGcrsa'
3315 def limit_length(s, length):
3316 """ Add ellipses to overly long strings """
3321 return s[:length - len(ELLIPSES)] + ELLIPSES
3325 def version_tuple(v):
3326 return tuple(int(e) for e in re.split(r'[-.]', v))
3329 def is_outdated_version(version, limit, assume_new=True):
3331 return not assume_new
3333 return version_tuple(version) < version_tuple(limit)
3335 return not assume_new
3338 def ytdl_is_updateable():
3339 """ Returns if yt-dlp can be updated with -U """
3341 from ..update import is_non_updateable
3343 return not is_non_updateable()
3346 def args_to_str(args):
3347 # Get a short string representation for a subprocess command
3348 return ' '.join(compat_shlex_quote(a) for a in args)
3351 def error_to_str(err):
3352 return f'{type(err).__name__}: {err}'
3355 def mimetype2ext(mt, default=NO_DEFAULT):
3356 if not isinstance(mt, str):
3357 if default is not NO_DEFAULT:
3373 'x-matroska': 'mkv',
3375 'x-mp4-fragmented': 'mp4',
3380 # application (streaming playlists)
3384 'vnd.apple.mpegurl': 'm3u8',
3385 'vnd.ms-sstr+xml': 'ism',
3386 'x-mpegurl': 'm3u8',
3390 # Per RFC 3003, audio/mpeg can be .mp1, .mp2 or .mp3.
3391 # Using .mp3 as it's the most popular one
3392 'audio/mpeg': 'mp3',
3393 'audio/webm': 'webm',
3394 'audio/x-matroska': 'mka',
3395 'audio/x-mpegurl': 'm3u',
3403 'x-realaudio': 'ra',
3414 'vnd.wap.wbmp': 'wbmp',
3421 'filmstrip+json': 'fs',
3422 'smptett+xml': 'tt',
3425 'x-ms-sami': 'sami',
3434 mimetype = mt.partition(';')[0].strip().lower()
3435 _, _, subtype = mimetype.rpartition('/')
3437 ext = traversal.traverse_obj(MAP, mimetype, subtype, subtype.rsplit('+')[-1])
3440 elif default is not NO_DEFAULT:
3442 return subtype.replace('+', '.')
3445 def ext2mimetype(ext_or_url):
3448 if '.' not in ext_or_url:
3449 ext_or_url = f'file.{ext_or_url}'
3450 return mimetypes.guess_type(ext_or_url)[0]
3453 def parse_codecs(codecs_str):
3454 # http://tools.ietf.org/html/rfc6381
3457 split_codecs = list(filter(None, map(
3458 str.strip, codecs_str.strip().strip(',').split(','))))
3459 vcodec, acodec, scodec, hdr = None, None, None, None
3460 for full_codec in split_codecs:
3461 parts = re.sub(r'0+(?=\d)', '', full_codec).split('.')
3462 if parts[0] in ('avc1', 'avc2', 'avc3', 'avc4', 'vp9', 'vp8', 'hev1', 'hev2',
3463 'h263', 'h264', 'mp4v', 'hvc1', 'av1', 'theora', 'dvh1', 'dvhe'):
3467 if parts[0] in ('dvh1', 'dvhe'):
3469 elif parts[0] == 'av1' and traversal.traverse_obj(parts, 3) == '10':
3471 elif parts[:2] == ['vp9', '2']:
3473 elif parts[0] in ('flac', 'mp4a', 'opus', 'vorbis', 'mp3', 'aac', 'ac-4',
3474 'ac-3', 'ec-3', 'eac3', 'dtsc', 'dtse', 'dtsh', 'dtsl'):
3475 acodec = acodec or full_codec
3476 elif parts[0] in ('stpp', 'wvtt'):
3477 scodec = scodec or full_codec
3479 write_string(f'WARNING: Unknown codec {full_codec}\n')
3480 if vcodec or acodec or scodec:
3482 'vcodec': vcodec or 'none',
3483 'acodec': acodec or 'none',
3484 'dynamic_range': hdr,
3485 **({'scodec': scodec} if scodec is not None else {}),
3487 elif len(split_codecs) == 2:
3489 'vcodec': split_codecs[0],
3490 'acodec': split_codecs[1],
3495 def get_compatible_ext(*, vcodecs, acodecs, vexts, aexts, preferences=None):
3496 assert len(vcodecs) == len(vexts) and len(acodecs) == len(aexts)
3498 allow_mkv = not preferences or 'mkv' in preferences
3500 if allow_mkv and max(len(acodecs), len(vcodecs)) > 1:
3501 return 'mkv' # TODO: any other format allows this?
3503 # TODO: All codecs supported by parse_codecs isn't handled here
3504 COMPATIBLE_CODECS = {
3506 'av1', 'hevc', 'avc1', 'mp4a', 'ac-4', # fourcc (m3u8, mpd)
3507 'h264', 'aacl', 'ec-3', # Set in ISM
3510 'av1', 'vp9', 'vp8', 'opus', 'vrbs',
3511 'vp9x', 'vp8x', # in the webm spec
3515 sanitize_codec = functools.partial(
3516 try_get, getter=lambda x: x[0].split('.')[0].replace('0', '').lower())
3517 vcodec, acodec = sanitize_codec(vcodecs), sanitize_codec(acodecs)
3519 for ext in preferences or COMPATIBLE_CODECS.keys():
3520 codec_set = COMPATIBLE_CODECS.get(ext, set())
3521 if ext == 'mkv' or codec_set.issuperset((vcodec, acodec)):
3525 {'mp3', 'mp4', 'm4a', 'm4p', 'm4b', 'm4r', 'm4v', 'ismv', 'isma', 'mov'},
3528 for ext in preferences or vexts:
3529 current_exts = {ext, *vexts, *aexts}
3530 if ext == 'mkv' or current_exts == {ext} or any(
3531 ext_sets.issuperset(current_exts) for ext_sets in COMPATIBLE_EXTS):
3533 return 'mkv' if allow_mkv else preferences[-1]
3536 def urlhandle_detect_ext(url_handle, default=NO_DEFAULT):
3537 getheader = url_handle.headers.get
3539 cd = getheader('Content-Disposition')
3541 m = re.match(r'attachment;\s*filename="(?P
<filename
>[^
"]+)"', cd)
3543 e = determine_ext(m.group('filename
'), default_ext=None)
3547 meta_ext = getheader('x
-amz
-meta
-name
')
3549 e = meta_ext.rpartition('.')[2]
3553 return mimetype2ext(getheader('Content
-Type
'), default=default)
3556 def encode_data_uri(data, mime_type):
3557 return 'data
:%s;base64
,%s' % (mime_type, base64.b64encode(data).decode('ascii
'))
3560 def age_restricted(content_limit, age_limit):
3561 """ Returns True iff the content should be blocked """
3563 if age_limit is None: # No limit set
3565 if content_limit is None:
3566 return False # Content available for everyone
3567 return age_limit < content_limit
3570 # List of known byte-order-marks (BOM)
3572 (b'\xef\xbb\xbf', 'utf
-8'),
3573 (b'\x00\x00\xfe\xff', 'utf
-32-be
'),
3574 (b'\xff\xfe\x00\x00', 'utf
-32-le
'),
3575 (b'\xff\xfe', 'utf
-16-le
'),
3576 (b'\xfe\xff', 'utf
-16-be
'),
3580 def is_html(first_bytes):
3581 """ Detect whether a file contains HTML by examining its first bytes. """
3584 for bom, enc in BOMS:
3585 while first_bytes.startswith(bom):
3586 encoding, first_bytes = enc, first_bytes[len(bom):]
3588 return re.match(r'^\s
*<', first_bytes.decode(encoding, 'replace
'))
3591 def determine_protocol(info_dict):
3592 protocol = info_dict.get('protocol
')
3593 if protocol is not None:
3596 url = sanitize_url(info_dict['url
'])
3597 if url.startswith('rtmp
'):
3599 elif url.startswith('mms
'):
3601 elif url.startswith('rtsp
'):
3604 ext = determine_ext(url)
3606 return 'm3u8
' if info_dict.get('is_live
') else 'm3u8_native
'
3610 return urllib.parse.urlparse(url).scheme
3613 def render_table(header_row, data, delim=False, extra_gap=0, hide_empty=False):
3614 """ Render a list of rows, each as a list of values.
3615 Text after a \t will be right aligned """
3617 return len(remove_terminal_sequences(string).replace('\t', ''))
3619 def get_max_lens(table):
3620 return [max(width(str(v)) for v in col) for col in zip(*table)]
3622 def filter_using_list(row, filterArray):
3623 return [col for take, col in itertools.zip_longest(filterArray, row, fillvalue=True) if take]
3625 max_lens = get_max_lens(data) if hide_empty else []
3626 header_row = filter_using_list(header_row, max_lens)
3627 data = [filter_using_list(row, max_lens) for row in data]
3629 table = [header_row] + data
3630 max_lens = get_max_lens(table)
3633 table = [header_row, [delim * (ml + extra_gap) for ml in max_lens]] + data
3634 table[1][-1] = table[1][-1][:-extra_gap * len(delim)] # Remove extra_gap from end of delimiter
3636 for pos, text in enumerate(map(str, row)):
3638 row[pos] = text.replace('\t', ' ' * (max_lens[pos] - width(text))) + ' ' * extra_gap
3640 row[pos] = text + ' ' * (max_lens[pos] - width(text) + extra_gap)
3641 ret = '\n'.join(''.join(row).rstrip() for row in table)
3645 def _match_one(filter_part, dct, incomplete):
3646 # TODO: Generalize code with YoutubeDL._build_format_filter
3647 STRING_OPERATORS = {
3648 '*=': operator.contains,
3649 '^
=': lambda attr, value: attr.startswith(value),
3650 '$
=': lambda attr, value: attr.endswith(value),
3651 '~
=': lambda attr, value: re.search(value, attr),
3653 COMPARISON_OPERATORS = {
3655 '<=': operator.le, # "<=" must be defined above "<"
3662 if isinstance(incomplete, bool):
3663 is_incomplete = lambda _: incomplete
3665 is_incomplete = lambda k: k in incomplete
3667 operator_rex = re.compile(r'''(?x)
3669 \s*(?P<negation>!\s*)?(?P<op>%s)(?P<none_inclusive>\s*\?)?\s*
3671 (?P<quote>["\'])(?P<quotedstrval>.+?)(?P=quote)|
3674 ''' % '|
'.join(map(re.escape, COMPARISON_OPERATORS.keys())))
3675 m = operator_rex.fullmatch(filter_part.strip())
3678 unnegated_op = COMPARISON_OPERATORS[m['op
']]
3680 op = lambda attr, value: not unnegated_op(attr, value)
3683 comparison_value = m['quotedstrval
'] or m['strval
'] or m['intval
']
3685 comparison_value = comparison_value.replace(r'\
%s' % m['quote
'], m['quote
'])
3686 actual_value = dct.get(m['key
'])
3687 numeric_comparison = None
3688 if isinstance(actual_value, (int, float)):
3689 # If the original field is a string and matching comparisonvalue is
3690 # a number we should respect the origin of the original field
3691 # and process comparison value as a string (see
3692 # https://github.com/ytdl-org/youtube-dl/issues/11082)
3694 numeric_comparison = int(comparison_value)
3696 numeric_comparison = parse_filesize(comparison_value)
3697 if numeric_comparison is None:
3698 numeric_comparison = parse_filesize(f'{comparison_value}B
')
3699 if numeric_comparison is None:
3700 numeric_comparison = parse_duration(comparison_value)
3701 if numeric_comparison is not None and m['op
'] in STRING_OPERATORS:
3702 raise ValueError('Operator
%s only supports string values
!' % m['op
'])
3703 if actual_value is None:
3704 return is_incomplete(m['key
']) or m['none_inclusive
']
3705 return op(actual_value, comparison_value if numeric_comparison is None else numeric_comparison)
3708 '': lambda v: (v is True) if isinstance(v, bool) else (v is not None),
3709 '!': lambda v: (v is False) if isinstance(v, bool) else (v is None),
3711 operator_rex = re.compile(r'''(?x)
3712 (?P<op>%s)\s*(?P<key>[a-z_]+)
3713 ''' % '|
'.join(map(re.escape, UNARY_OPERATORS.keys())))
3714 m = operator_rex.fullmatch(filter_part.strip())
3716 op = UNARY_OPERATORS[m.group('op
')]
3717 actual_value = dct.get(m.group('key
'))
3718 if is_incomplete(m.group('key
')) and actual_value is None:
3720 return op(actual_value)
3722 raise ValueError('Invalid
filter part
%r' % filter_part)
3725 def match_str(filter_str, dct, incomplete=False):
3726 """ Filter a dictionary with a simple string syntax.
3727 @returns Whether the filter passes
3728 @param incomplete Set of keys that is expected to be missing from dct.
3729 Can be True/False to indicate all/none of the keys may be missing.
3730 All conditions on incomplete keys pass if the key is missing
3733 _match_one(filter_part.replace(r'\
&', '&'), dct, incomplete)
3734 for filter_part in re.split(r'(?
<!\\)&', filter_str))
3737 def match_filter_func(filters, breaking_filters=None):
3738 if not filters and not breaking_filters:
3740 breaking_filters = match_filter_func(breaking_filters) or (lambda _, __: None)
3741 filters = set(variadic(filters or []))
3743 interactive = '-' in filters
3747 def _match_func(info_dict, incomplete=False):
3748 ret = breaking_filters(info_dict, incomplete)
3750 raise RejectedVideoReached(ret)
3752 if not filters or any(match_str(f, info_dict, incomplete) for f in filters):
3753 return NO_DEFAULT if interactive and not incomplete else None
3755 video_title = info_dict.get('title
') or info_dict.get('id') or 'entry
'
3756 filter_str = ') |
('.join(map(str.strip, filters))
3757 return f'{video_title} does
not pass filter ({filter_str}
), skipping
..'
3761 class download_range_func:
3762 def __init__(self, chapters, ranges, from_info=False):
3763 self.chapters, self.ranges, self.from_info = chapters, ranges, from_info
3765 def __call__(self, info_dict, ydl):
3767 warning = ('There are no chapters matching the regex
' if info_dict.get('chapters
')
3768 else 'Cannot match chapters since chapter information
is unavailable
')
3769 for regex in self.chapters or []:
3770 for i, chapter in enumerate(info_dict.get('chapters
') or []):
3771 if re.search(regex, chapter['title
']):
3773 yield {**chapter, 'index': i}
3774 if self.chapters and warning:
3775 ydl.to_screen(f'[info
] {info_dict["id"]}
: {warning}
')
3777 for start, end in self.ranges or []:
3779 'start_time
': self._handle_negative_timestamp(start, info_dict),
3780 'end_time
': self._handle_negative_timestamp(end, info_dict),
3783 if self.from_info and (info_dict.get('start_time
') or info_dict.get('end_time
')):
3785 'start_time
': info_dict.get('start_time
') or 0,
3786 'end_time
': info_dict.get('end_time
') or float('inf
'),
3788 elif not self.ranges and not self.chapters:
3792 def _handle_negative_timestamp(time, info):
3793 return max(info['duration
'] + time, 0) if info.get('duration
') and time < 0 else time
3795 def __eq__(self, other):
3796 return (isinstance(other, download_range_func)
3797 and self.chapters == other.chapters and self.ranges == other.ranges)
3800 return f'{__name__}
.{type(self).__name__}
({self.chapters}
, {self.ranges}
)'
3803 def parse_dfxp_time_expr(time_expr):
3807 mobj = re.match(rf'^
(?P
<time_offset
>{NUMBER_RE}
)s?$
', time_expr)
3809 return float(mobj.group('time_offset
'))
3811 mobj = re.match(r'^
(\d
+):(\d\d
):(\d\
d(?
:(?
:\
.|
:)\d
+)?
)$
', time_expr)
3813 return 3600 * int(mobj.group(1)) + 60 * int(mobj.group(2)) + float(mobj.group(3).replace(':', '.'))
3816 def srt_subtitles_timecode(seconds):
3817 return '%02d
:%02d
:%02d
,%03d
' % timetuple_from_msec(seconds * 1000)
3820 def ass_subtitles_timecode(seconds):
3821 time = timetuple_from_msec(seconds * 1000)
3822 return '%01d
:%02d
:%02d
.%02d
' % (*time[:-1], time.milliseconds / 10)
3825 def dfxp2srt(dfxp_data):
3827 @param dfxp_data A bytes-like object containing DFXP data
3828 @returns A unicode object containing converted SRT data
3830 LEGACY_NAMESPACES = (
3831 (b'http
://www
.w3
.org
/ns
/ttml
', [
3832 b'http
://www
.w3
.org
/2004/11/ttaf1
',
3833 b'http
://www
.w3
.org
/2006/04/ttaf1
',
3834 b'http
://www
.w3
.org
/2006/10/ttaf1
',
3836 (b'http
://www
.w3
.org
/ns
/ttml
#styling', [
3837 b
'http://www.w3.org/ns/ttml#style',
3841 SUPPORTED_STYLING
= [
3850 _x
= functools
.partial(xpath_with_ns
, ns_map
={
3851 'xml': 'http://www.w3.org/XML/1998/namespace',
3852 'ttml': 'http://www.w3.org/ns/ttml',
3853 'tts': 'http://www.w3.org/ns/ttml#styling',
3859 class TTMLPElementParser
:
3861 _unclosed_elements
= []
3862 _applied_styles
= []
3864 def start(self
, tag
, attrib
):
3865 if tag
in (_x('ttml:br'), 'br'):
3868 unclosed_elements
= []
3870 element_style_id
= attrib
.get('style')
3872 style
.update(default_style
)
3873 if element_style_id
:
3874 style
.update(styles
.get(element_style_id
, {}))
3875 for prop
in SUPPORTED_STYLING
:
3876 prop_val
= attrib
.get(_x('tts:' + prop
))
3878 style
[prop
] = prop_val
3881 for k
, v
in sorted(style
.items()):
3882 if self
._applied
_styles
and self
._applied
_styles
[-1].get(k
) == v
:
3885 font
+= ' color="%s"' % v
3886 elif k
== 'fontSize':
3887 font
+= ' size="%s"' % v
3888 elif k
== 'fontFamily':
3889 font
+= ' face="%s"' % v
3890 elif k
== 'fontWeight' and v
== 'bold':
3892 unclosed_elements
.append('b')
3893 elif k
== 'fontStyle' and v
== 'italic':
3895 unclosed_elements
.append('i')
3896 elif k
== 'textDecoration' and v
== 'underline':
3898 unclosed_elements
.append('u')
3900 self
._out
+= '<font' + font
+ '>'
3901 unclosed_elements
.append('font')
3903 if self
._applied
_styles
:
3904 applied_style
.update(self
._applied
_styles
[-1])
3905 applied_style
.update(style
)
3906 self
._applied
_styles
.append(applied_style
)
3907 self
._unclosed
_elements
.append(unclosed_elements
)
3910 if tag
not in (_x('ttml:br'), 'br'):
3911 unclosed_elements
= self
._unclosed
_elements
.pop()
3912 for element
in reversed(unclosed_elements
):
3913 self
._out
+= '</%s>' % element
3914 if unclosed_elements
and self
._applied
_styles
:
3915 self
._applied
_styles
.pop()
3917 def data(self
, data
):
3921 return self
._out
.strip()
3923 # Fix UTF-8 encoded file wrongly marked as UTF-16. See https://github.com/yt-dlp/yt-dlp/issues/6543#issuecomment-1477169870
3924 # This will not trigger false positives since only UTF-8 text is being replaced
3925 dfxp_data
= dfxp_data
.replace(b
'encoding=\'UTF-16\'', b
'encoding=\'UTF-8\'')
3927 def parse_node(node
):
3928 target
= TTMLPElementParser()
3929 parser
= xml
.etree
.ElementTree
.XMLParser(target
=target
)
3930 parser
.feed(xml
.etree
.ElementTree
.tostring(node
))
3931 return parser
.close()
3933 for k
, v
in LEGACY_NAMESPACES
:
3935 dfxp_data
= dfxp_data
.replace(ns
, k
)
3937 dfxp
= compat_etree_fromstring(dfxp_data
)
3939 paras
= dfxp
.findall(_x('.//ttml:p')) or dfxp
.findall('.//p')
3942 raise ValueError('Invalid dfxp/TTML subtitle')
3946 for style
in dfxp
.findall(_x('.//ttml:style')):
3947 style_id
= style
.get('id') or style
.get(_x('xml:id'))
3950 parent_style_id
= style
.get('style')
3952 if parent_style_id
not in styles
:
3955 styles
[style_id
] = styles
[parent_style_id
].copy()
3956 for prop
in SUPPORTED_STYLING
:
3957 prop_val
= style
.get(_x('tts:' + prop
))
3959 styles
.setdefault(style_id
, {})[prop
] = prop_val
3965 for p
in ('body', 'div'):
3966 ele
= xpath_element(dfxp
, [_x('.//ttml:' + p
), './/' + p
])
3969 style
= styles
.get(ele
.get('style'))
3972 default_style
.update(style
)
3974 for para
, index
in zip(paras
, itertools
.count(1)):
3975 begin_time
= parse_dfxp_time_expr(para
.attrib
.get('begin'))
3976 end_time
= parse_dfxp_time_expr(para
.attrib
.get('end'))
3977 dur
= parse_dfxp_time_expr(para
.attrib
.get('dur'))
3978 if begin_time
is None:
3983 end_time
= begin_time
+ dur
3984 out
.append('%d\n%s --> %s\n%s\n\n' % (
3986 srt_subtitles_timecode(begin_time
),
3987 srt_subtitles_timecode(end_time
),
3993 def cli_option(params
, command_option
, param
, separator
=None):
3994 param
= params
.get(param
)
3995 return ([] if param
is None
3996 else [command_option
, str(param
)] if separator
is None
3997 else [f
'{command_option}{separator}{param}'])
4000 def cli_bool_option(params
, command_option
, param
, true_value
='true', false_value
='false', separator
=None):
4001 param
= params
.get(param
)
4002 assert param
in (True, False, None)
4003 return cli_option({True: true_value, False: false_value}
, command_option
, param
, separator
)
4006 def cli_valueless_option(params
, command_option
, param
, expected_value
=True):
4007 return [command_option
] if params
.get(param
) == expected_value
else []
4010 def cli_configuration_args(argdict
, keys
, default
=[], use_compat
=True):
4011 if isinstance(argdict
, (list, tuple)): # for backward compatibility
4018 assert isinstance(argdict
, dict)
4020 assert isinstance(keys
, (list, tuple))
4021 for key_list
in keys
:
4022 arg_list
= list(filter(
4023 lambda x
: x
is not None,
4024 [argdict
.get(key
.lower()) for key
in variadic(key_list
)]))
4026 return [arg
for args
in arg_list
for arg
in args
]
4030 def _configuration_args(main_key
, argdict
, exe
, keys
=None, default
=[], use_compat
=True):
4031 main_key
, exe
= main_key
.lower(), exe
.lower()
4032 root_key
= exe
if main_key
== exe
else f
'{main_key}+{exe}'
4033 keys
= [f
'{root_key}{k}' for k
in (keys
or [''])]
4034 if root_key
in keys
:
4036 keys
.append((main_key
, exe
))
4037 keys
.append('default')
4040 return cli_configuration_args(argdict
, keys
, default
, use_compat
)
4044 # See http://www.loc.gov/standards/iso639-2/ISO-639-2_utf-8.txt
4103 'iw': 'heb', # Replaced by he in 1989 revision
4113 'in': 'ind', # Replaced by id in 1989 revision
4229 'ji': 'yid', # Replaced by yi in 1989 revision
4237 def short2long(cls
, code
):
4238 """Convert language code from ISO 639-1 to ISO 639-2/T"""
4239 return cls
._lang
_map
.get(code
[:2])
4242 def long2short(cls
, code
):
4243 """Convert language code from ISO 639-2/T to ISO 639-1"""
4244 for short_name
, long_name
in cls
._lang
_map
.items():
4245 if long_name
== code
:
4250 # From http://data.okfn.org/data/core/country-list
4252 'AF': 'Afghanistan',
4253 'AX': 'Åland Islands',
4256 'AS': 'American Samoa',
4261 'AG': 'Antigua and Barbuda',
4278 'BO': 'Bolivia, Plurinational State of',
4279 'BQ': 'Bonaire, Sint Eustatius and Saba',
4280 'BA': 'Bosnia and Herzegovina',
4282 'BV': 'Bouvet Island',
4284 'IO': 'British Indian Ocean Territory',
4285 'BN': 'Brunei Darussalam',
4287 'BF': 'Burkina Faso',
4293 'KY': 'Cayman Islands',
4294 'CF': 'Central African Republic',
4298 'CX': 'Christmas Island',
4299 'CC': 'Cocos (Keeling) Islands',
4303 'CD': 'Congo, the Democratic Republic of the',
4304 'CK': 'Cook Islands',
4306 'CI': 'Côte d\'Ivoire',
4311 'CZ': 'Czech Republic',
4315 'DO': 'Dominican Republic',
4318 'SV': 'El Salvador',
4319 'GQ': 'Equatorial Guinea',
4323 'FK': 'Falkland Islands (Malvinas)',
4324 'FO': 'Faroe Islands',
4328 'GF': 'French Guiana',
4329 'PF': 'French Polynesia',
4330 'TF': 'French Southern Territories',
4345 'GW': 'Guinea-Bissau',
4348 'HM': 'Heard Island and McDonald Islands',
4349 'VA': 'Holy See (Vatican City State)',
4356 'IR': 'Iran, Islamic Republic of',
4359 'IM': 'Isle of Man',
4369 'KP': 'Korea, Democratic People\'s Republic of',
4370 'KR': 'Korea, Republic of',
4373 'LA': 'Lao People\'s Democratic Republic',
4379 'LI': 'Liechtenstein',
4383 'MK': 'Macedonia, the Former Yugoslav Republic of',
4390 'MH': 'Marshall Islands',
4396 'FM': 'Micronesia, Federated States of',
4397 'MD': 'Moldova, Republic of',
4408 'NL': 'Netherlands',
4409 'NC': 'New Caledonia',
4410 'NZ': 'New Zealand',
4415 'NF': 'Norfolk Island',
4416 'MP': 'Northern Mariana Islands',
4421 'PS': 'Palestine, State of',
4423 'PG': 'Papua New Guinea',
4426 'PH': 'Philippines',
4430 'PR': 'Puerto Rico',
4434 'RU': 'Russian Federation',
4436 'BL': 'Saint Barthélemy',
4437 'SH': 'Saint Helena, Ascension and Tristan da Cunha',
4438 'KN': 'Saint Kitts and Nevis',
4439 'LC': 'Saint Lucia',
4440 'MF': 'Saint Martin (French part)',
4441 'PM': 'Saint Pierre and Miquelon',
4442 'VC': 'Saint Vincent and the Grenadines',
4445 'ST': 'Sao Tome and Principe',
4446 'SA': 'Saudi Arabia',
4450 'SL': 'Sierra Leone',
4452 'SX': 'Sint Maarten (Dutch part)',
4455 'SB': 'Solomon Islands',
4457 'ZA': 'South Africa',
4458 'GS': 'South Georgia and the South Sandwich Islands',
4459 'SS': 'South Sudan',
4464 'SJ': 'Svalbard and Jan Mayen',
4467 'CH': 'Switzerland',
4468 'SY': 'Syrian Arab Republic',
4469 'TW': 'Taiwan, Province of China',
4471 'TZ': 'Tanzania, United Republic of',
4473 'TL': 'Timor-Leste',
4477 'TT': 'Trinidad and Tobago',
4480 'TM': 'Turkmenistan',
4481 'TC': 'Turks and Caicos Islands',
4485 'AE': 'United Arab Emirates',
4486 'GB': 'United Kingdom',
4487 'US': 'United States',
4488 'UM': 'United States Minor Outlying Islands',
4492 'VE': 'Venezuela, Bolivarian Republic of',
4494 'VG': 'Virgin Islands, British',
4495 'VI': 'Virgin Islands, U.S.',
4496 'WF': 'Wallis and Futuna',
4497 'EH': 'Western Sahara',
4501 # Not ISO 3166 codes, but used for IP blocks
4502 'AP': 'Asia/Pacific Region',
4507 def short2full(cls
, code
):
4508 """Convert an ISO 3166-2 country code to the corresponding full name"""
4509 return cls
._country
_map
.get(code
.upper())
4513 # Major IPv4 address blocks per country
4515 'AD': '46.172.224.0/19',
4516 'AE': '94.200.0.0/13',
4517 'AF': '149.54.0.0/17',
4518 'AG': '209.59.64.0/18',
4519 'AI': '204.14.248.0/21',
4520 'AL': '46.99.0.0/16',
4521 'AM': '46.70.0.0/15',
4522 'AO': '105.168.0.0/13',
4523 'AP': '182.50.184.0/21',
4524 'AQ': '23.154.160.0/24',
4525 'AR': '181.0.0.0/12',
4526 'AS': '202.70.112.0/20',
4527 'AT': '77.116.0.0/14',
4528 'AU': '1.128.0.0/11',
4529 'AW': '181.41.0.0/18',
4530 'AX': '185.217.4.0/22',
4531 'AZ': '5.197.0.0/16',
4532 'BA': '31.176.128.0/17',
4533 'BB': '65.48.128.0/17',
4534 'BD': '114.130.0.0/16',
4536 'BF': '102.178.0.0/15',
4537 'BG': '95.42.0.0/15',
4538 'BH': '37.131.0.0/17',
4539 'BI': '154.117.192.0/18',
4540 'BJ': '137.255.0.0/16',
4541 'BL': '185.212.72.0/23',
4542 'BM': '196.12.64.0/18',
4543 'BN': '156.31.0.0/16',
4544 'BO': '161.56.0.0/16',
4545 'BQ': '161.0.80.0/20',
4546 'BR': '191.128.0.0/12',
4547 'BS': '24.51.64.0/18',
4548 'BT': '119.2.96.0/19',
4549 'BW': '168.167.0.0/16',
4550 'BY': '178.120.0.0/13',
4551 'BZ': '179.42.192.0/18',
4552 'CA': '99.224.0.0/11',
4553 'CD': '41.243.0.0/16',
4554 'CF': '197.242.176.0/21',
4555 'CG': '160.113.0.0/16',
4556 'CH': '85.0.0.0/13',
4557 'CI': '102.136.0.0/14',
4558 'CK': '202.65.32.0/19',
4559 'CL': '152.172.0.0/14',
4560 'CM': '102.244.0.0/14',
4561 'CN': '36.128.0.0/10',
4562 'CO': '181.240.0.0/12',
4563 'CR': '201.192.0.0/12',
4564 'CU': '152.206.0.0/15',
4565 'CV': '165.90.96.0/19',
4566 'CW': '190.88.128.0/17',
4567 'CY': '31.153.0.0/16',
4568 'CZ': '88.100.0.0/14',
4570 'DJ': '197.241.0.0/17',
4571 'DK': '87.48.0.0/12',
4572 'DM': '192.243.48.0/20',
4573 'DO': '152.166.0.0/15',
4574 'DZ': '41.96.0.0/12',
4575 'EC': '186.68.0.0/15',
4576 'EE': '90.190.0.0/15',
4577 'EG': '156.160.0.0/11',
4578 'ER': '196.200.96.0/20',
4579 'ES': '88.0.0.0/11',
4580 'ET': '196.188.0.0/14',
4581 'EU': '2.16.0.0/13',
4582 'FI': '91.152.0.0/13',
4583 'FJ': '144.120.0.0/16',
4584 'FK': '80.73.208.0/21',
4585 'FM': '119.252.112.0/20',
4586 'FO': '88.85.32.0/19',
4588 'GA': '41.158.0.0/15',
4590 'GD': '74.122.88.0/21',
4591 'GE': '31.146.0.0/16',
4592 'GF': '161.22.64.0/18',
4593 'GG': '62.68.160.0/19',
4594 'GH': '154.160.0.0/12',
4595 'GI': '95.164.0.0/16',
4596 'GL': '88.83.0.0/19',
4597 'GM': '160.182.0.0/15',
4598 'GN': '197.149.192.0/18',
4599 'GP': '104.250.0.0/19',
4600 'GQ': '105.235.224.0/20',
4601 'GR': '94.64.0.0/13',
4602 'GT': '168.234.0.0/16',
4603 'GU': '168.123.0.0/16',
4604 'GW': '197.214.80.0/20',
4605 'GY': '181.41.64.0/18',
4606 'HK': '113.252.0.0/14',
4607 'HN': '181.210.0.0/16',
4608 'HR': '93.136.0.0/13',
4609 'HT': '148.102.128.0/17',
4610 'HU': '84.0.0.0/14',
4611 'ID': '39.192.0.0/10',
4612 'IE': '87.32.0.0/12',
4613 'IL': '79.176.0.0/13',
4614 'IM': '5.62.80.0/20',
4615 'IN': '117.192.0.0/10',
4616 'IO': '203.83.48.0/21',
4617 'IQ': '37.236.0.0/14',
4618 'IR': '2.176.0.0/12',
4619 'IS': '82.221.0.0/16',
4620 'IT': '79.0.0.0/10',
4621 'JE': '87.244.64.0/18',
4622 'JM': '72.27.0.0/17',
4623 'JO': '176.29.0.0/16',
4624 'JP': '133.0.0.0/8',
4625 'KE': '105.48.0.0/12',
4626 'KG': '158.181.128.0/17',
4627 'KH': '36.37.128.0/17',
4628 'KI': '103.25.140.0/22',
4629 'KM': '197.255.224.0/20',
4630 'KN': '198.167.192.0/19',
4631 'KP': '175.45.176.0/22',
4632 'KR': '175.192.0.0/10',
4633 'KW': '37.36.0.0/14',
4634 'KY': '64.96.0.0/15',
4635 'KZ': '2.72.0.0/13',
4636 'LA': '115.84.64.0/18',
4637 'LB': '178.135.0.0/16',
4638 'LC': '24.92.144.0/20',
4639 'LI': '82.117.0.0/19',
4640 'LK': '112.134.0.0/15',
4641 'LR': '102.183.0.0/16',
4642 'LS': '129.232.0.0/17',
4643 'LT': '78.56.0.0/13',
4644 'LU': '188.42.0.0/16',
4645 'LV': '46.109.0.0/16',
4646 'LY': '41.252.0.0/14',
4647 'MA': '105.128.0.0/11',
4648 'MC': '88.209.64.0/18',
4649 'MD': '37.246.0.0/16',
4650 'ME': '178.175.0.0/17',
4651 'MF': '74.112.232.0/21',
4652 'MG': '154.126.0.0/17',
4653 'MH': '117.103.88.0/21',
4654 'MK': '77.28.0.0/15',
4655 'ML': '154.118.128.0/18',
4656 'MM': '37.111.0.0/17',
4657 'MN': '49.0.128.0/17',
4658 'MO': '60.246.0.0/16',
4659 'MP': '202.88.64.0/20',
4660 'MQ': '109.203.224.0/19',
4661 'MR': '41.188.64.0/18',
4662 'MS': '208.90.112.0/22',
4663 'MT': '46.11.0.0/16',
4664 'MU': '105.16.0.0/12',
4665 'MV': '27.114.128.0/18',
4666 'MW': '102.70.0.0/15',
4667 'MX': '187.192.0.0/11',
4668 'MY': '175.136.0.0/13',
4669 'MZ': '197.218.0.0/15',
4670 'NA': '41.182.0.0/16',
4671 'NC': '101.101.0.0/18',
4672 'NE': '197.214.0.0/18',
4673 'NF': '203.17.240.0/22',
4674 'NG': '105.112.0.0/12',
4675 'NI': '186.76.0.0/15',
4676 'NL': '145.96.0.0/11',
4677 'NO': '84.208.0.0/13',
4678 'NP': '36.252.0.0/15',
4679 'NR': '203.98.224.0/19',
4680 'NU': '49.156.48.0/22',
4681 'NZ': '49.224.0.0/14',
4682 'OM': '5.36.0.0/15',
4683 'PA': '186.72.0.0/15',
4684 'PE': '186.160.0.0/14',
4685 'PF': '123.50.64.0/18',
4686 'PG': '124.240.192.0/19',
4687 'PH': '49.144.0.0/13',
4688 'PK': '39.32.0.0/11',
4689 'PL': '83.0.0.0/11',
4690 'PM': '70.36.0.0/20',
4691 'PR': '66.50.0.0/16',
4692 'PS': '188.161.0.0/16',
4693 'PT': '85.240.0.0/13',
4694 'PW': '202.124.224.0/20',
4695 'PY': '181.120.0.0/14',
4696 'QA': '37.210.0.0/15',
4697 'RE': '102.35.0.0/16',
4698 'RO': '79.112.0.0/13',
4699 'RS': '93.86.0.0/15',
4700 'RU': '5.136.0.0/13',
4701 'RW': '41.186.0.0/16',
4702 'SA': '188.48.0.0/13',
4703 'SB': '202.1.160.0/19',
4704 'SC': '154.192.0.0/11',
4705 'SD': '102.120.0.0/13',
4706 'SE': '78.64.0.0/12',
4707 'SG': '8.128.0.0/10',
4708 'SI': '188.196.0.0/14',
4709 'SK': '78.98.0.0/15',
4710 'SL': '102.143.0.0/17',
4711 'SM': '89.186.32.0/19',
4712 'SN': '41.82.0.0/15',
4713 'SO': '154.115.192.0/18',
4714 'SR': '186.179.128.0/17',
4715 'SS': '105.235.208.0/21',
4716 'ST': '197.159.160.0/19',
4717 'SV': '168.243.0.0/16',
4718 'SX': '190.102.0.0/20',
4720 'SZ': '41.84.224.0/19',
4721 'TC': '65.255.48.0/20',
4722 'TD': '154.68.128.0/19',
4723 'TG': '196.168.0.0/14',
4724 'TH': '171.96.0.0/13',
4725 'TJ': '85.9.128.0/18',
4726 'TK': '27.96.24.0/21',
4727 'TL': '180.189.160.0/20',
4728 'TM': '95.85.96.0/19',
4729 'TN': '197.0.0.0/11',
4730 'TO': '175.176.144.0/21',
4731 'TR': '78.160.0.0/11',
4732 'TT': '186.44.0.0/15',
4733 'TV': '202.2.96.0/19',
4734 'TW': '120.96.0.0/11',
4735 'TZ': '156.156.0.0/14',
4736 'UA': '37.52.0.0/14',
4737 'UG': '102.80.0.0/13',
4739 'UY': '167.56.0.0/13',
4740 'UZ': '84.54.64.0/18',
4741 'VA': '212.77.0.0/19',
4742 'VC': '207.191.240.0/21',
4743 'VE': '186.88.0.0/13',
4744 'VG': '66.81.192.0/20',
4745 'VI': '146.226.0.0/16',
4746 'VN': '14.160.0.0/11',
4747 'VU': '202.80.32.0/20',
4748 'WF': '117.20.32.0/21',
4749 'WS': '202.4.32.0/19',
4750 'YE': '134.35.0.0/16',
4751 'YT': '41.242.116.0/22',
4752 'ZA': '41.0.0.0/11',
4753 'ZM': '102.144.0.0/13',
4754 'ZW': '102.177.192.0/18',
4758 def random_ipv4(cls
, code_or_block
):
4759 if len(code_or_block
) == 2:
4760 block
= cls
._country
_ip
_map
.get(code_or_block
.upper())
4764 block
= code_or_block
4765 addr
, preflen
= block
.split('/')
4766 addr_min
= struct
.unpack('!L', socket
.inet_aton(addr
))[0]
4767 addr_max
= addr_min |
(0xffffffff >> int(preflen
))
4768 return str(socket
.inet_ntoa(
4769 struct
.pack('!L', random
.randint(addr_min
, addr_max
))))
4772 class PerRequestProxyHandler(urllib
.request
.ProxyHandler
):
4773 def __init__(self
, proxies
=None):
4774 # Set default handlers
4775 for type in ('http', 'https'):
4776 setattr(self
, '%s_open' % type,
4777 lambda r
, proxy
='__noproxy__', type=type, meth
=self
.proxy_open
:
4778 meth(r
, proxy
, type))
4779 urllib
.request
.ProxyHandler
.__init
__(self
, proxies
)
4781 def proxy_open(self
, req
, proxy
, type):
4782 req_proxy
= req
.headers
.get('Ytdl-request-proxy')
4783 if req_proxy
is not None:
4785 del req
.headers
['Ytdl-request-proxy']
4787 if proxy
== '__noproxy__':
4788 return None # No Proxy
4789 if urllib
.parse
.urlparse(proxy
).scheme
.lower() in ('socks', 'socks4', 'socks4a', 'socks5'):
4790 req
.add_header('Ytdl-socks-proxy', proxy
)
4791 # yt-dlp's http/https handlers do wrapping the socket with socks
4793 return urllib
.request
.ProxyHandler
.proxy_open(
4794 self
, req
, proxy
, type)
4797 # Both long_to_bytes and bytes_to_long are adapted from PyCrypto, which is
4798 # released into Public Domain
4799 # https://github.com/dlitz/pycrypto/blob/master/lib/Crypto/Util/number.py#L387
4801 def long_to_bytes(n
, blocksize
=0):
4802 """long_to_bytes(n:long, blocksize:int) : string
4803 Convert a long integer to a byte string.
4805 If optional blocksize is given and greater than zero, pad the front of the
4806 byte string with binary zeros so that the length is a multiple of
4809 # after much testing, this algorithm was deemed to be the fastest
4813 s
= struct
.pack('>I', n
& 0xffffffff) + s
4815 # strip off leading zeros
4816 for i
in range(len(s
)):
4817 if s
[i
] != b
'\000'[0]:
4820 # only happens when n == 0
4824 # add back some pad bytes. this could be done more efficiently w.r.t. the
4825 # de-padding being done above, but sigh...
4826 if blocksize
> 0 and len(s
) % blocksize
:
4827 s
= (blocksize
- len(s
) % blocksize
) * b
'\000' + s
4831 def bytes_to_long(s
):
4832 """bytes_to_long(string) : long
4833 Convert a byte string to a long integer.
4835 This is (essentially) the inverse of long_to_bytes().
4840 extra
= (4 - length
% 4)
4841 s
= b
'\000' * extra
+ s
4842 length
= length
+ extra
4843 for i
in range(0, length
, 4):
4844 acc
= (acc
<< 32) + struct
.unpack('>I', s
[i
:i
+ 4])[0]
4848 def ohdave_rsa_encrypt(data
, exponent
, modulus
):
4850 Implement OHDave's RSA algorithm. See http://www.ohdave.com/rsa/
4853 data: data to encrypt, bytes-like object
4854 exponent, modulus: parameter e and N of RSA algorithm, both integer
4855 Output: hex string of encrypted data
4857 Limitation: supports one block encryption only
4860 payload
= int(binascii
.hexlify(data
[::-1]), 16)
4861 encrypted
= pow(payload
, exponent
, modulus
)
4862 return '%x' % encrypted
4865 def pkcs1pad(data
, length
):
4867 Padding input data with PKCS#1 scheme
4869 @param {int[]} data input data
4870 @param {int} length target length
4871 @returns {int[]} padded data
4873 if len(data
) > length
- 11:
4874 raise ValueError('Input data too long for PKCS#1 padding')
4876 pseudo_random
= [random
.randint(0, 254) for _
in range(length
- len(data
) - 3)]
4877 return [0, 2] + pseudo_random
+ [0] + data
4880 def _base_n_table(n
, table
):
4881 if not table
and not n
:
4882 raise ValueError('Either table or n must be specified')
4883 table
= (table
or '0123456789abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ')[:n
]
4885 if n
and n
!= len(table
):
4886 raise ValueError(f
'base {n} exceeds table length {len(table)}')
4890 def encode_base_n(num
, n
=None, table
=None):
4891 """Convert given int to a base-n string"""
4892 table
= _base_n_table(n
, table
)
4896 result
, base
= '', len(table
)
4898 result
= table
[num
% base
] + result
4903 def decode_base_n(string
, n
=None, table
=None):
4904 """Convert given base-n string to int"""
4905 table
= {char: index for index, char in enumerate(_base_n_table(n, table))}
4906 result
, base
= 0, len(table
)
4908 result
= result
* base
+ table
[char
]
4912 def decode_packed_codes(code
):
4913 mobj
= re
.search(PACKED_CODES_RE
, code
)
4914 obfuscated_code
, base
, count
, symbols
= mobj
.groups()
4917 symbols
= symbols
.split('|')
4922 base_n_count
= encode_base_n(count
, base
)
4923 symbol_table
[base_n_count
] = symbols
[count
] or base_n_count
4926 r
'\b(\w+)\b', lambda mobj
: symbol_table
[mobj
.group(0)],
4930 def caesar(s
, alphabet
, shift
):
4935 alphabet
[(alphabet
.index(c
) + shift
) % l
] if c
in alphabet
else c
4940 return caesar(s
, r
'''!"#$%&'()*+,-./0123456789:;<=>?@ABCDEFGHIJKLMNOPQRSTUVWXYZ[\]^_`abcdefghijklmnopqrstuvwxyz{|}~''', 47)
4943 def parse_m3u8_attributes(attrib
):
4945 for (key
, val
) in re
.findall(r
'(?P<key>[A-Z0-9-]+)=(?P<val>"[^"]+"|[^",]+)(?:,|$)', attrib
):
4946 if val
.startswith('"'):
4952 def urshift(val
, n
):
4953 return val
>> n
if val
>= 0 else (val
+ 0x100000000) >> n
4956 def write_xattr(path
, key
, value
):
4957 # Windows: Write xattrs to NTFS Alternate Data Streams:
4958 # http://en.wikipedia.org/wiki/NTFS#Alternate_data_streams_.28ADS.29
4959 if compat_os_name
== 'nt':
4960 assert ':' not in key
4961 assert os
.path
.exists(path
)
4964 with open(f
'{path}:{key}', 'wb') as f
:
4966 except OSError as e
:
4967 raise XAttrMetadataError(e
.errno
, e
.strerror
)
4970 # UNIX Method 1. Use xattrs/pyxattrs modules
4973 if getattr(xattr
, '_yt_dlp__identifier', None) == 'pyxattr':
4974 # Unicode arguments are not supported in pyxattr until version 0.5.0
4975 # See https://github.com/ytdl-org/youtube-dl/issues/5498
4976 if version_tuple(xattr
.__version
__) >= (0, 5, 0):
4977 setxattr
= xattr
.set
4979 setxattr
= xattr
.setxattr
4983 setxattr(path
, key
, value
)
4984 except OSError as e
:
4985 raise XAttrMetadataError(e
.errno
, e
.strerror
)
4988 # UNIX Method 2. Use setfattr/xattr executables
4989 exe
= ('setfattr' if check_executable('setfattr', ['--version'])
4990 else 'xattr' if check_executable('xattr', ['-h']) else None)
4992 raise XAttrUnavailableError(
4993 'Couldn\'t find a tool to set the xattrs. Install either the python "xattr" or "pyxattr" modules or the '
4994 + ('"xattr" binary' if sys
.platform
!= 'linux' else 'GNU "attr" package (which contains the "setfattr" tool)'))
4996 value
= value
.decode()
4998 _
, stderr
, returncode
= Popen
.run(
4999 [exe
, '-w', key
, value
, path
] if exe
== 'xattr' else [exe
, '-n', key
, '-v', value
, path
],
5000 text
=True, stdout
=subprocess
.PIPE
, stderr
=subprocess
.PIPE
, stdin
=subprocess
.PIPE
)
5001 except OSError as e
:
5002 raise XAttrMetadataError(e
.errno
, e
.strerror
)
5004 raise XAttrMetadataError(returncode
, stderr
)
5007 def random_birthday(year_field
, month_field
, day_field
):
5008 start_date
= datetime
.date(1950, 1, 1)
5009 end_date
= datetime
.date(1995, 12, 31)
5010 offset
= random
.randint(0, (end_date
- start_date
).days
)
5011 random_date
= start_date
+ datetime
.timedelta(offset
)
5013 year_field
: str(random_date
.year
),
5014 month_field
: str(random_date
.month
),
5015 day_field
: str(random_date
.day
),
5019 def find_available_port(interface
=''):
5021 with socket
.socket() as sock
:
5022 sock
.bind((interface
, 0))
5023 return sock
.getsockname()[1]
5028 # Templates for internet shortcut files, which are plain text files.
5029 DOT_URL_LINK_TEMPLATE
= '''\
5034 DOT_WEBLOC_LINK_TEMPLATE
= '''\
5035 <?xml version="1.0" encoding="UTF-8"?>
5036 <!DOCTYPE plist PUBLIC "-//Apple//DTD PLIST 1.0//EN" "http://www.apple.com/DTDs/PropertyList-1.0.dtd">
5037 <plist version="1.0">
5040 \t<string>%(url)s</string>
5045 DOT_DESKTOP_LINK_TEMPLATE
= '''\
5055 'url': DOT_URL_LINK_TEMPLATE
,
5056 'desktop': DOT_DESKTOP_LINK_TEMPLATE
,
5057 'webloc': DOT_WEBLOC_LINK_TEMPLATE
,
5061 def iri_to_uri(iri
):
5063 Converts an IRI (Internationalized Resource Identifier, allowing Unicode characters) to a URI (Uniform Resource Identifier, ASCII-only).
5065 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.
5068 iri_parts
= urllib
.parse
.urlparse(iri
)
5070 if '[' in iri_parts
.netloc
:
5071 raise ValueError('IPv6 URIs are not, yet, supported.')
5072 # Querying `.netloc`, when there's only one bracket, also raises a ValueError.
5074 # 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.
5077 if iri_parts
.username
:
5078 net_location
+= urllib
.parse
.quote(iri_parts
.username
, safe
=r
"!$%&'()*+,~")
5079 if iri_parts
.password
is not None:
5080 net_location
+= ':' + urllib
.parse
.quote(iri_parts
.password
, safe
=r
"!$%&'()*+,~")
5083 net_location
+= iri_parts
.hostname
.encode('idna').decode() # Punycode for Unicode hostnames.
5084 # The 'idna' encoding produces ASCII text.
5085 if iri_parts
.port
is not None and iri_parts
.port
!= 80:
5086 net_location
+= ':' + str(iri_parts
.port
)
5088 return urllib
.parse
.urlunparse(
5092 urllib
.parse
.quote_plus(iri_parts
.path
, safe
=r
"!$%&'()*+,/:;=@|~"),
5094 # Unsure about the `safe` argument, since this is a legacy way of handling parameters.
5095 urllib
.parse
.quote_plus(iri_parts
.params
, safe
=r
"!$%&'()*+,/:;=@|~"),
5097 # Not totally sure about the `safe` argument, since the source does not explicitly mention the query URI component.
5098 urllib
.parse
.quote_plus(iri_parts
.query
, safe
=r
"!$%&'()*+,/:;=?@{|}~"),
5100 urllib
.parse
.quote_plus(iri_parts
.fragment
, safe
=r
"!#$%&'()*+,/:;=?@{|}~")))
5102 # Source for `safe` arguments: https://url.spec.whatwg.org/#percent-encoded-bytes.
5105 def to_high_limit_path(path
):
5106 if sys
.platform
in ['win32', 'cygwin']:
5107 # Work around MAX_PATH limitation on Windows. The maximum allowed length for the individual path segments may still be quite limited.
5108 return '\\\\?\\' + os
.path
.abspath(path
)
5113 def format_field(obj
, field
=None, template
='%s', ignore
=NO_DEFAULT
, default
='', func
=IDENTITY
):
5114 val
= traversal
.traverse_obj(obj
, *variadic(field
))
5115 if not val
if ignore
is NO_DEFAULT
else val
in variadic(ignore
):
5117 return template
% func(val
)
5120 def clean_podcast_url(url
):
5121 url
= re
.sub(r
'''(?x)
5125 media\.blubrry\.com| # https://create.blubrry.com/resources/podcast-media-download-statistics/getting-started/
5130 (?:dts|www)\.podtrac\.com/(?:pts/)?redirect\.[0-9a-z]{3,4}| # http://analytics.podtrac.com/how-to-measure
5133 cn\.co| # https://podcorn.com/analytics-prefix/
5134 st\.fm # https://podsights.com/docs/
5139 return re
.sub(r
'^\w+://(\w+://)', r
'\1', url
)
5142 _HEX_TABLE
= '0123456789abcdef'
5145 def random_uuidv4():
5146 return re
.sub(r
'[xy]', lambda x
: _HEX_TABLE
[random
.randint(0, 15)], 'xxxxxxxx-xxxx-4xxx-yxxx-xxxxxxxxxxxx')
5149 def make_dir(path
, to_screen
=None):
5151 dn
= os
.path
.dirname(path
)
5153 os
.makedirs(dn
, exist_ok
=True)
5155 except OSError as err
:
5156 if callable(to_screen
) is not None:
5157 to_screen(f
'unable to create directory {err}')
5161 def get_executable_path():
5162 from ..update
import _get_variant_and_executable_path
5164 return os
.path
.dirname(os
.path
.abspath(_get_variant_and_executable_path()[1]))
5167 def get_user_config_dirs(package_name
):
5168 # .config (e.g. ~/.config/package_name)
5169 xdg_config_home
= os
.getenv('XDG_CONFIG_HOME') or compat_expanduser('~/.config')
5170 yield os
.path
.join(xdg_config_home
, package_name
)
5172 # appdata (%APPDATA%/package_name)
5173 appdata_dir
= os
.getenv('appdata')
5175 yield os
.path
.join(appdata_dir
, package_name
)
5177 # home (~/.package_name)
5178 yield os
.path
.join(compat_expanduser('~'), f
'.{package_name}')
5181 def get_system_config_dirs(package_name
):
5183 yield os
.path
.join('/etc', package_name
)
5186 def time_seconds(**kwargs
):
5188 Returns TZ-aware time in seconds since the epoch (1970-01-01T00:00:00Z)
5190 return time
.time() + datetime
.timedelta(**kwargs
).total_seconds()
5193 # create a JSON Web Signature (jws) with HS256 algorithm
5194 # the resulting format is in JWS Compact Serialization
5195 # implemented following JWT https://www.rfc-editor.org/rfc/rfc7519.html
5196 # implemented following JWS https://www.rfc-editor.org/rfc/rfc7515.html
5197 def jwt_encode_hs256(payload_data
, key
, headers
={}):
5203 header_data
.update(headers
)
5204 header_b64
= base64
.b64encode(json
.dumps(header_data
).encode())
5205 payload_b64
= base64
.b64encode(json
.dumps(payload_data
).encode())
5206 h
= hmac
.new(key
.encode(), header_b64
+ b
'.' + payload_b64
, hashlib
.sha256
)
5207 signature_b64
= base64
.b64encode(h
.digest())
5208 token
= header_b64
+ b
'.' + payload_b64
+ b
'.' + signature_b64
5212 # can be extended in future to verify the signature and parse header and return the algorithm used if it's not HS256
5213 def jwt_decode_hs256(jwt
):
5214 header_b64
, payload_b64
, signature_b64
= jwt
.split('.')
5215 # add trailing ='s that may have been stripped, superfluous ='s are ignored
5216 payload_data
= json
.loads(base64
.urlsafe_b64decode(f
'{payload_b64}==='))
5220 WINDOWS_VT_MODE
= False if compat_os_name
== 'nt' else None
5224 def supports_terminal_sequences(stream
):
5225 if compat_os_name
== 'nt':
5226 if not WINDOWS_VT_MODE
:
5228 elif not os
.getenv('TERM'):
5231 return stream
.isatty()
5232 except BaseException
:
5236 def windows_enable_vt_mode():
5237 """Ref: https://bugs.python.org/issue30075 """
5238 if get_windows_version() < (10, 0, 10586):
5242 import ctypes
.wintypes
5245 ENABLE_VIRTUAL_TERMINAL_PROCESSING
= 0x0004
5247 dll
= ctypes
.WinDLL('kernel32', use_last_error
=False)
5248 handle
= os
.open('CONOUT$', os
.O_RDWR
)
5250 h_out
= ctypes
.wintypes
.HANDLE(msvcrt
.get_osfhandle(handle
))
5251 dw_original_mode
= ctypes
.wintypes
.DWORD()
5252 success
= dll
.GetConsoleMode(h_out
, ctypes
.byref(dw_original_mode
))
5254 raise Exception('GetConsoleMode failed')
5256 success
= dll
.SetConsoleMode(h_out
, ctypes
.wintypes
.DWORD(
5257 dw_original_mode
.value | ENABLE_VIRTUAL_TERMINAL_PROCESSING
))
5259 raise Exception('SetConsoleMode failed')
5263 global WINDOWS_VT_MODE
5264 WINDOWS_VT_MODE
= True
5265 supports_terminal_sequences
.cache_clear()
5268 _terminal_sequences_re
= re
.compile('\033\\[[^m]+m')
5271 def remove_terminal_sequences(string
):
5272 return _terminal_sequences_re
.sub('', string
)
5275 def number_of_digits(number
):
5276 return len('%d' % number
)
5279 def join_nonempty(*values
, delim
='-', from_dict
=None):
5280 if from_dict
is not None:
5281 values
= (traversal
.traverse_obj(from_dict
, variadic(v
)) for v
in values
)
5282 return delim
.join(map(str, filter(None, values
)))
5285 def scale_thumbnails_to_max_format_width(formats
, thumbnails
, url_width_re
):
5287 Find the largest format dimensions in terms of video width and, for each thumbnail:
5288 * Modify the URL: Match the width with the provided regex and replace with the former width
5291 This function is useful with video services that scale the provided thumbnails on demand
5293 _keys
= ('width', 'height')
5294 max_dimensions
= max(
5295 (tuple(format
.get(k
) or 0 for k
in _keys
) for format
in formats
),
5297 if not max_dimensions
[0]:
5301 {'url': re.sub(url_width_re, str(max_dimensions[0]), thumbnail['url'])}
,
5302 dict(zip(_keys
, max_dimensions
)), thumbnail
)
5303 for thumbnail
in thumbnails
5307 def parse_http_range(range):
5308 """ Parse value of "Range" or "Content-Range" HTTP header into tuple. """
5310 return None, None, None
5311 crg
= re
.search(r
'bytes[ =](\d+)-(\d+)?(?:/(\d+))?', range)
5313 return None, None, None
5314 return int(crg
.group(1)), int_or_none(crg
.group(2)), int_or_none(crg
.group(3))
5317 def read_stdin(what
):
5318 eof
= 'Ctrl+Z' if compat_os_name
== 'nt' else 'Ctrl+D'
5319 write_string(f
'Reading {what} from STDIN - EOF ({eof}) to end:\n')
5323 def determine_file_encoding(data
):
5325 Detect the text encoding used
5326 @returns (encoding, bytes to skip)
5329 # BOM marks are given priority over declarations
5330 for bom
, enc
in BOMS
:
5331 if data
.startswith(bom
):
5332 return enc
, len(bom
)
5334 # Strip off all null bytes to match even when UTF-16 or UTF-32 is used.
5335 # We ignore the endianness to get a good enough match
5336 data
= data
.replace(b
'\0', b
'')
5337 mobj
= re
.match(rb
'(?m)^#\s*coding\s*:\s*(\S+)\s*$', data
)
5338 return mobj
.group(1).decode() if mobj
else None, 0
5345 __initialized
= False
5347 def __init__(self
, parser
, label
=None):
5348 self
.parser
, self
.label
= parser
, label
5349 self
._loaded
_paths
, self
.configs
= set(), []
5351 def init(self
, args
=None, filename
=None):
5352 assert not self
.__initialized
5353 self
.own_args
, self
.filename
= args
, filename
5354 return self
.load_configs()
5356 def load_configs(self
):
5359 location
= os
.path
.realpath(self
.filename
)
5360 directory
= os
.path
.dirname(location
)
5361 if location
in self
._loaded
_paths
:
5363 self
._loaded
_paths
.add(location
)
5365 self
.__initialized
= True
5366 opts
, _
= self
.parser
.parse_known_args(self
.own_args
)
5367 self
.parsed_args
= self
.own_args
5368 for location
in opts
.config_locations
or []:
5370 if location
in self
._loaded
_paths
:
5372 self
._loaded
_paths
.add(location
)
5373 self
.append_config(shlex
.split(read_stdin('options'), comments
=True), label
='stdin')
5375 location
= os
.path
.join(directory
, expand_path(location
))
5376 if os
.path
.isdir(location
):
5377 location
= os
.path
.join(location
, 'yt-dlp.conf')
5378 if not os
.path
.exists(location
):
5379 self
.parser
.error(f
'config location {location} does not exist')
5380 self
.append_config(self
.read_file(location
), location
)
5384 label
= join_nonempty(
5385 self
.label
, 'config', f
'"{self.filename}"' if self
.filename
else '',
5387 return join_nonempty(
5388 self
.own_args
is not None and f
'{label[0].upper()}{label[1:]}: {self.hide_login_info(self.own_args)}',
5389 *(f
'\n{c}'.replace('\n', '\n| ')[1:] for c
in self
.configs
),
5393 def read_file(filename
, default
=[]):
5395 optionf
= open(filename
, 'rb')
5397 return default
# silently skip if file is not present
5399 enc
, skip
= determine_file_encoding(optionf
.read(512))
5400 optionf
.seek(skip
, io
.SEEK_SET
)
5402 enc
= None # silently skip read errors
5404 # FIXME: https://github.com/ytdl-org/youtube-dl/commit/dfe5fa49aed02cf36ba9f743b11b0903554b5e56
5405 contents
= optionf
.read().decode(enc
or preferredencoding())
5406 res
= shlex
.split(contents
, comments
=True)
5407 except Exception as err
:
5408 raise ValueError(f
'Unable to parse "{filename}": {err}')
5414 def hide_login_info(opts
):
5415 PRIVATE_OPTS
= {'-p', '--password', '-u', '--username', '--video-password', '--ap-password', '--ap-username'}
5416 eqre
= re
.compile('^(?P<key>' + ('|'.join(re
.escape(po
) for po
in PRIVATE_OPTS
)) + ')=.+$')
5421 return m
.group('key') + '=PRIVATE'
5425 opts
= list(map(_scrub_eq
, opts
))
5426 for idx
, opt
in enumerate(opts
):
5427 if opt
in PRIVATE_OPTS
and idx
+ 1 < len(opts
):
5428 opts
[idx
+ 1] = 'PRIVATE'
5431 def append_config(self
, *args
, label
=None):
5432 config
= type(self
)(self
.parser
, label
)
5433 config
._loaded
_paths
= self
._loaded
_paths
5434 if config
.init(*args
):
5435 self
.configs
.append(config
)
5439 for config
in reversed(self
.configs
):
5440 yield from config
.all_args
5441 yield from self
.parsed_args
or []
5443 def parse_known_args(self
, **kwargs
):
5444 return self
.parser
.parse_known_args(self
.all_args
, **kwargs
)
5446 def parse_args(self
):
5447 return self
.parser
.parse_args(self
.all_args
)
5450 class WebSocketsWrapper
:
5451 """Wraps websockets module to use in non-async scopes"""
5454 def __init__(self
, url
, headers
=None, connect
=True):
5455 self
.loop
= asyncio
.new_event_loop()
5456 # XXX: "loop" is deprecated
5457 self
.conn
= websockets
.connect(
5458 url
, extra_headers
=headers
, ping_interval
=None,
5459 close_timeout
=float('inf'), loop
=self
.loop
, ping_timeout
=float('inf'))
5462 atexit
.register(self
.__exit
__, None, None, None)
5464 def __enter__(self
):
5466 self
.pool
= self
.run_with_loop(self
.conn
.__aenter
__(), self
.loop
)
5469 def send(self
, *args
):
5470 self
.run_with_loop(self
.pool
.send(*args
), self
.loop
)
5472 def recv(self
, *args
):
5473 return self
.run_with_loop(self
.pool
.recv(*args
), self
.loop
)
5475 def __exit__(self
, type, value
, traceback
):
5477 return self
.run_with_loop(self
.conn
.__aexit
__(type, value
, traceback
), self
.loop
)
5480 self
._cancel
_all
_tasks
(self
.loop
)
5482 # taken from https://github.com/python/cpython/blob/3.9/Lib/asyncio/runners.py with modifications
5483 # for contributors: If there's any new library using asyncio needs to be run in non-async, move these function out of this class
5485 def run_with_loop(main
, loop
):
5486 if not asyncio
.iscoroutine(main
):
5487 raise ValueError(f
'a coroutine was expected, got {main!r}')
5490 return loop
.run_until_complete(main
)
5492 loop
.run_until_complete(loop
.shutdown_asyncgens())
5493 if hasattr(loop
, 'shutdown_default_executor'):
5494 loop
.run_until_complete(loop
.shutdown_default_executor())
5497 def _cancel_all_tasks(loop
):
5498 to_cancel
= asyncio
.all_tasks(loop
)
5503 for task
in to_cancel
:
5506 # XXX: "loop" is removed in python 3.10+
5507 loop
.run_until_complete(
5508 asyncio
.gather(*to_cancel
, loop
=loop
, return_exceptions
=True))
5510 for task
in to_cancel
:
5511 if task
.cancelled():
5513 if task
.exception() is not None:
5514 loop
.call_exception_handler({
5515 'message': 'unhandled exception during asyncio.run() shutdown',
5516 'exception': task
.exception(),
5521 def merge_headers(*dicts
):
5522 """Merge dicts of http headers case insensitively, prioritizing the latter ones"""
5523 return {k.title(): v for k, v in itertools.chain.from_iterable(map(dict.items, dicts))}
5526 def cached_method(f
):
5527 """Cache a method"""
5528 signature
= inspect
.signature(f
)
5531 def wrapper(self
, *args
, **kwargs
):
5532 bound_args
= signature
.bind(self
, *args
, **kwargs
)
5533 bound_args
.apply_defaults()
5534 key
= tuple(bound_args
.arguments
.values())[1:]
5536 cache
= vars(self
).setdefault('_cached_method__cache', {}).setdefault(f.__name__, {}
)
5537 if key
not in cache
:
5538 cache
[key
] = f(self
, *args
, **kwargs
)
5543 class classproperty
:
5544 """property access for class methods with optional caching"""
5545 def __new__(cls
, func
=None, *args
, **kwargs
):
5547 return functools
.partial(cls
, *args
, **kwargs
)
5548 return super().__new
__(cls
)
5550 def __init__(self
, func
, *, cache
=False):
5551 functools
.update_wrapper(self
, func
)
5553 self
._cache
= {} if cache
else None
5555 def __get__(self
, _
, cls
):
5556 if self
._cache
is None:
5557 return self
.func(cls
)
5558 elif cls
not in self
._cache
:
5559 self
._cache
[cls
] = self
.func(cls
)
5560 return self
._cache
[cls
]
5563 class function_with_repr
:
5564 def __init__(self
, func
, repr_
=None):
5565 functools
.update_wrapper(self
, func
)
5566 self
.func
, self
.__repr
= func
, repr_
5568 def __call__(self
, *args
, **kwargs
):
5569 return self
.func(*args
, **kwargs
)
5574 return f
'{self.func.__module__}.{self.func.__qualname__}'
5577 class Namespace(types
.SimpleNamespace
):
5578 """Immutable namespace"""
5581 return iter(self
.__dict
__.values())
5585 return self
.__dict
__.items()
5588 MEDIA_EXTENSIONS
= Namespace(
5589 common_video
=('avi', 'flv', 'mkv', 'mov', 'mp4', 'webm'),
5590 video
=('3g2', '3gp', 'f4v', 'mk3d', 'divx', 'mpg', 'ogv', 'm4v', 'wmv'),
5591 common_audio
=('aiff', 'alac', 'flac', 'm4a', 'mka', 'mp3', 'ogg', 'opus', 'wav'),
5592 audio
=('aac', 'ape', 'asf', 'f4a', 'f4b', 'm4b', 'm4p', 'm4r', 'oga', 'ogx', 'spx', 'vorbis', 'wma', 'weba'),
5593 thumbnails
=('jpg', 'png', 'webp'),
5594 storyboards
=('mhtml', ),
5595 subtitles
=('srt', 'vtt', 'ass', 'lrc'),
5596 manifests
=('f4f', 'f4m', 'm3u8', 'smil', 'mpd'),
5598 MEDIA_EXTENSIONS
.video
+= MEDIA_EXTENSIONS
.common_video
5599 MEDIA_EXTENSIONS
.audio
+= MEDIA_EXTENSIONS
.common_audio
5601 KNOWN_EXTENSIONS
= (*MEDIA_EXTENSIONS
.video
, *MEDIA_EXTENSIONS
.audio
, *MEDIA_EXTENSIONS
.manifests
)
5606 for retry in RetryManager(...):
5609 except SomeException as err:
5613 attempt
, _error
= 0, None
5615 def __init__(self
, _retries
, _error_callback
, **kwargs
):
5616 self
.retries
= _retries
or 0
5617 self
.error_callback
= functools
.partial(_error_callback
, **kwargs
)
5619 def _should_retry(self
):
5620 return self
._error
is not NO_DEFAULT
and self
.attempt
<= self
.retries
5624 if self
._error
is NO_DEFAULT
:
5629 def error(self
, value
):
5633 while self
._should
_retry
():
5634 self
.error
= NO_DEFAULT
5638 self
.error_callback(self
.error
, self
.attempt
, self
.retries
)
5641 def report_retry(e
, count
, retries
, *, sleep_func
, info
, warn
, error
=None, suffix
=None):
5642 """Utility function for reporting retries"""
5645 return error(f
'{e}. Giving up after {count - 1} retries') if count
> 1 else error(str(e
))
5650 elif isinstance(e
, ExtractorError
):
5651 e
= remove_end(str_or_none(e
.cause
) or e
.orig_msg
, '.')
5652 warn(f
'{e}. Retrying{format_field(suffix, None, " %s")} ({count}/{retries})...')
5654 delay
= float_or_none(sleep_func(n
=count
- 1)) if callable(sleep_func
) else sleep_func
5656 info(f
'Sleeping {delay:.2f} seconds ...')
5660 def make_archive_id(ie
, video_id
):
5661 ie_key
= ie
if isinstance(ie
, str) else ie
.ie_key()
5662 return f
'{ie_key.lower()} {video_id}'
5665 def truncate_string(s
, left
, right
=0):
5666 assert left
> 3 and right
>= 0
5667 if s
is None or len(s
) <= left
+ right
:
5669 return f
'{s[:left-3]}...{s[-right:] if right else ""}'
5672 def orderedSet_from_options(options
, alias_dict
, *, use_regex
=False, start
=None):
5673 assert 'all' in alias_dict
, '"all" alias is required'
5674 requested
= list(start
or [])
5676 discard
= val
.startswith('-')
5680 if val
in alias_dict
:
5681 val
= alias_dict
[val
] if not discard
else [
5682 i
[1:] if i
.startswith('-') else f
'-{i}' for i
in alias_dict
[val
]]
5683 # NB: Do not allow regex in aliases for performance
5684 requested
= orderedSet_from_options(val
, alias_dict
, start
=requested
)
5687 current
= (filter(re
.compile(val
, re
.I
).fullmatch
, alias_dict
['all']) if use_regex
5688 else [val
] if val
in alias_dict
['all'] else None)
5690 raise ValueError(val
)
5693 for item
in current
:
5694 while item
in requested
:
5695 requested
.remove(item
)
5697 requested
.extend(current
)
5699 return orderedSet(requested
)
5704 regex
= r
' *((?P<reverse>\+)?(?P<field>[a-zA-Z0-9_]+)((?P<separator>[~:])(?P<limit>.*?))?)? *$'
5706 default
= ('hidden', 'aud_or_vid', 'hasvid', 'ie_pref', 'lang', 'quality',
5707 'res', 'fps', 'hdr:12', 'vcodec:vp9.2', 'channels', 'acodec',
5708 'size', 'br', 'asr', 'proto', 'ext', 'hasaud', 'source', 'id') # These must not be aliases
5709 ytdl_default
= ('hasaud', 'lang', 'quality', 'tbr', 'filesize', 'vbr',
5710 'height', 'width', 'proto', 'vext', 'abr', 'aext',
5711 'fps', 'fs_approx', 'source', 'id')
5714 'vcodec': {'type': 'ordered', 'regex': True,
5715 'order': ['av0?1', 'vp0?9.2', 'vp0?9', '[hx]265|he?vc?', '[hx]264|avc', 'vp0?8', 'mp4v|h263', 'theora', '', None, 'none']},
5716 'acodec': {'type': 'ordered', 'regex': True,
5717 'order': ['[af]lac', 'wav|aiff', 'opus', 'vorbis|ogg', 'aac', 'mp?4a?', 'mp3', 'ac-?4', 'e-?a?c-?3', 'ac-?3', 'dts', '', None, 'none']},
5718 'hdr': {'type': 'ordered', 'regex': True, 'field': 'dynamic_range',
5719 'order': ['dv', '(hdr)?12', r
'(hdr)?10\+', '(hdr)?10', 'hlg', '', 'sdr', None]},
5720 'proto': {'type': 'ordered', 'regex': True, 'field': 'protocol',
5721 'order': ['(ht|f)tps', '(ht|f)tp$', 'm3u8.*', '.*dash', 'websocket_frag', 'rtmpe?', '', 'mms|rtsp', 'ws|websocket', 'f4']},
5722 'vext': {'type': 'ordered', 'field': 'video_ext',
5723 'order': ('mp4', 'mov', 'webm', 'flv', '', 'none'),
5724 'order_free': ('webm', 'mp4', 'mov', 'flv', '', 'none')},
5725 'aext': {'type': 'ordered', 'regex': True, 'field': 'audio_ext',
5726 'order': ('m4a', 'aac', 'mp3', 'ogg', 'opus', 'web[am]', '', 'none'),
5727 'order_free': ('ogg', 'opus', 'web[am]', 'mp3', 'm4a', 'aac', '', 'none')},
5728 'hidden': {'visible': False, 'forced': True, 'type': 'extractor', 'max': -1000}
,
5729 'aud_or_vid': {'visible': False, 'forced': True, 'type': 'multiple',
5730 'field': ('vcodec', 'acodec'),
5731 'function': lambda it
: int(any(v
!= 'none' for v
in it
))},
5732 'ie_pref': {'priority': True, 'type': 'extractor'}
,
5733 'hasvid': {'priority': True, 'field': 'vcodec', 'type': 'boolean', 'not_in_list': ('none',)}
,
5734 'hasaud': {'field': 'acodec', 'type': 'boolean', 'not_in_list': ('none',)}
,
5735 'lang': {'convert': 'float', 'field': 'language_preference', 'default': -1}
,
5736 'quality': {'convert': 'float', 'default': -1}
,
5737 'filesize': {'convert': 'bytes'}
,
5738 'fs_approx': {'convert': 'bytes', 'field': 'filesize_approx'}
,
5739 'id': {'convert': 'string', 'field': 'format_id'}
,
5740 'height': {'convert': 'float_none'}
,
5741 'width': {'convert': 'float_none'}
,
5742 'fps': {'convert': 'float_none'}
,
5743 'channels': {'convert': 'float_none', 'field': 'audio_channels'}
,
5744 'tbr': {'convert': 'float_none'}
,
5745 'vbr': {'convert': 'float_none'}
,
5746 'abr': {'convert': 'float_none'}
,
5747 'asr': {'convert': 'float_none'}
,
5748 'source': {'convert': 'float', 'field': 'source_preference', 'default': -1}
,
5750 'codec': {'type': 'combined', 'field': ('vcodec', 'acodec')}
,
5751 'br': {'type': 'multiple', 'field': ('tbr', 'vbr', 'abr'), 'convert': 'float_none',
5752 'function': lambda it
: next(filter(None, it
), None)},
5753 'size': {'type': 'multiple', 'field': ('filesize', 'fs_approx'), 'convert': 'bytes',
5754 'function': lambda it
: next(filter(None, it
), None)},
5755 'ext': {'type': 'combined', 'field': ('vext', 'aext')}
,
5756 'res': {'type': 'multiple', 'field': ('height', 'width'),
5757 'function': lambda it
: (lambda l
: min(l
) if l
else 0)(tuple(filter(None, it
)))},
5759 # Actual field names
5760 'format_id': {'type': 'alias', 'field': 'id'}
,
5761 'preference': {'type': 'alias', 'field': 'ie_pref'}
,
5762 'language_preference': {'type': 'alias', 'field': 'lang'}
,
5763 'source_preference': {'type': 'alias', 'field': 'source'}
,
5764 'protocol': {'type': 'alias', 'field': 'proto'}
,
5765 'filesize_approx': {'type': 'alias', 'field': 'fs_approx'}
,
5766 'audio_channels': {'type': 'alias', 'field': 'channels'}
,
5769 'dimension': {'type': 'alias', 'field': 'res', 'deprecated': True}
,
5770 'resolution': {'type': 'alias', 'field': 'res', 'deprecated': True}
,
5771 'extension': {'type': 'alias', 'field': 'ext', 'deprecated': True}
,
5772 'bitrate': {'type': 'alias', 'field': 'br', 'deprecated': True}
,
5773 'total_bitrate': {'type': 'alias', 'field': 'tbr', 'deprecated': True}
,
5774 'video_bitrate': {'type': 'alias', 'field': 'vbr', 'deprecated': True}
,
5775 'audio_bitrate': {'type': 'alias', 'field': 'abr', 'deprecated': True}
,
5776 'framerate': {'type': 'alias', 'field': 'fps', 'deprecated': True}
,
5777 'filesize_estimate': {'type': 'alias', 'field': 'size', 'deprecated': True}
,
5778 'samplerate': {'type': 'alias', 'field': 'asr', 'deprecated': True}
,
5779 'video_ext': {'type': 'alias', 'field': 'vext', 'deprecated': True}
,
5780 'audio_ext': {'type': 'alias', 'field': 'aext', 'deprecated': True}
,
5781 'video_codec': {'type': 'alias', 'field': 'vcodec', 'deprecated': True}
,
5782 'audio_codec': {'type': 'alias', 'field': 'acodec', 'deprecated': True}
,
5783 'video': {'type': 'alias', 'field': 'hasvid', 'deprecated': True}
,
5784 'has_video': {'type': 'alias', 'field': 'hasvid', 'deprecated': True}
,
5785 'audio': {'type': 'alias', 'field': 'hasaud', 'deprecated': True}
,
5786 'has_audio': {'type': 'alias', 'field': 'hasaud', 'deprecated': True}
,
5787 'extractor': {'type': 'alias', 'field': 'ie_pref', 'deprecated': True}
,
5788 'extractor_preference': {'type': 'alias', 'field': 'ie_pref', 'deprecated': True}
,
5791 def __init__(self
, ydl
, field_preference
):
5794 self
.evaluate_params(self
.ydl
.params
, field_preference
)
5795 if ydl
.params
.get('verbose'):
5796 self
.print_verbose_info(self
.ydl
.write_debug
)
5798 def _get_field_setting(self
, field
, key
):
5799 if field
not in self
.settings
:
5800 if key
in ('forced', 'priority'):
5802 self
.ydl
.deprecated_feature(f
'Using arbitrary fields ({field}) for format sorting is '
5803 'deprecated and may be removed in a future version')
5804 self
.settings
[field
] = {}
5805 propObj
= self
.settings
[field
]
5806 if key
not in propObj
:
5807 type = propObj
.get('type')
5809 default
= 'preference' if type == 'extractor' else (field
,) if type in ('combined', 'multiple') else field
5810 elif key
== 'convert':
5811 default
= 'order' if type == 'ordered' else 'float_string' if field
else 'ignore'
5813 default
= {'type': 'field', 'visible': True, 'order': [], 'not_in_list': (None,)}
.get(key
, None)
5814 propObj
[key
] = default
5817 def _resolve_field_value(self
, field
, value
, convertNone
=False):
5822 value
= value
.lower()
5823 conversion
= self
._get
_field
_setting
(field
, 'convert')
5824 if conversion
== 'ignore':
5826 if conversion
== 'string':
5828 elif conversion
== 'float_none':
5829 return float_or_none(value
)
5830 elif conversion
== 'bytes':
5831 return parse_bytes(value
)
5832 elif conversion
== 'order':
5833 order_list
= (self
._use
_free
_order
and self
._get
_field
_setting
(field
, 'order_free')) or self
._get
_field
_setting
(field
, 'order')
5834 use_regex
= self
._get
_field
_setting
(field
, 'regex')
5835 list_length
= len(order_list
)
5836 empty_pos
= order_list
.index('') if '' in order_list
else list_length
+ 1
5837 if use_regex
and value
is not None:
5838 for i
, regex
in enumerate(order_list
):
5839 if regex
and re
.match(regex
, value
):
5840 return list_length
- i
5841 return list_length
- empty_pos
# not in list
5842 else: # not regex or value = None
5843 return list_length
- (order_list
.index(value
) if value
in order_list
else empty_pos
)
5845 if value
.isnumeric():
5848 self
.settings
[field
]['convert'] = 'string'
5851 def evaluate_params(self
, params
, sort_extractor
):
5852 self
._use
_free
_order
= params
.get('prefer_free_formats', False)
5853 self
._sort
_user
= params
.get('format_sort', [])
5854 self
._sort
_extractor
= sort_extractor
5856 def add_item(field
, reverse
, closest
, limit_text
):
5857 field
= field
.lower()
5858 if field
in self
._order
:
5860 self
._order
.append(field
)
5861 limit
= self
._resolve
_field
_value
(field
, limit_text
)
5864 'closest': False if limit
is None else closest
,
5865 'limit_text': limit_text
,
5867 if field
in self
.settings
:
5868 self
.settings
[field
].update(data
)
5870 self
.settings
[field
] = data
5873 tuple(field
for field
in self
.default
if self
._get
_field
_setting
(field
, 'forced'))
5874 + (tuple() if params
.get('format_sort_force', False)
5875 else tuple(field
for field
in self
.default
if self
._get
_field
_setting
(field
, 'priority')))
5876 + tuple(self
._sort
_user
) + tuple(sort_extractor
) + self
.default
)
5878 for item
in sort_list
:
5879 match
= re
.match(self
.regex
, item
)
5881 raise ExtractorError('Invalid format sort string "%s" given by extractor' % item
)
5882 field
= match
.group('field')
5885 if self
._get
_field
_setting
(field
, 'type') == 'alias':
5886 alias
, field
= field
, self
._get
_field
_setting
(field
, 'field')
5887 if self
._get
_field
_setting
(alias
, 'deprecated'):
5888 self
.ydl
.deprecated_feature(f
'Format sorting alias {alias} is deprecated and may '
5889 f
'be removed in a future version. Please use {field} instead')
5890 reverse
= match
.group('reverse') is not None
5891 closest
= match
.group('separator') == '~'
5892 limit_text
= match
.group('limit')
5894 has_limit
= limit_text
is not None
5895 has_multiple_fields
= self
._get
_field
_setting
(field
, 'type') == 'combined'
5896 has_multiple_limits
= has_limit
and has_multiple_fields
and not self
._get
_field
_setting
(field
, 'same_limit')
5898 fields
= self
._get
_field
_setting
(field
, 'field') if has_multiple_fields
else (field
,)
5899 limits
= limit_text
.split(':') if has_multiple_limits
else (limit_text
,) if has_limit
else tuple()
5900 limit_count
= len(limits
)
5901 for (i
, f
) in enumerate(fields
):
5902 add_item(f
, reverse
, closest
,
5903 limits
[i
] if i
< limit_count
5904 else limits
[0] if has_limit
and not has_multiple_limits
5907 def print_verbose_info(self
, write_debug
):
5909 write_debug('Sort order given by user: %s' % ', '.join(self
._sort
_user
))
5910 if self
._sort
_extractor
:
5911 write_debug('Sort order given by extractor: %s' % ', '.join(self
._sort
_extractor
))
5912 write_debug('Formats sorted by: %s' % ', '.join(['%s%s%s' % (
5913 '+' if self
._get
_field
_setting
(field
, 'reverse') else '', field
,
5914 '%s%s(%s)' % ('~' if self
._get
_field
_setting
(field
, 'closest') else ':',
5915 self
._get
_field
_setting
(field
, 'limit_text'),
5916 self
._get
_field
_setting
(field
, 'limit'))
5917 if self
._get
_field
_setting
(field
, 'limit_text') is not None else '')
5918 for field
in self
._order
if self
._get
_field
_setting
(field
, 'visible')]))
5920 def _calculate_field_preference_from_value(self
, format
, field
, type, value
):
5921 reverse
= self
._get
_field
_setting
(field
, 'reverse')
5922 closest
= self
._get
_field
_setting
(field
, 'closest')
5923 limit
= self
._get
_field
_setting
(field
, 'limit')
5925 if type == 'extractor':
5926 maximum
= self
._get
_field
_setting
(field
, 'max')
5927 if value
is None or (maximum
is not None and value
>= maximum
):
5929 elif type == 'boolean':
5930 in_list
= self
._get
_field
_setting
(field
, 'in_list')
5931 not_in_list
= self
._get
_field
_setting
(field
, 'not_in_list')
5932 value
= 0 if ((in_list
is None or value
in in_list
) and (not_in_list
is None or value
not in not_in_list
)) else -1
5933 elif type == 'ordered':
5934 value
= self
._resolve
_field
_value
(field
, value
, True)
5936 # try to convert to number
5937 val_num
= float_or_none(value
, default
=self
._get
_field
_setting
(field
, 'default'))
5938 is_num
= self
._get
_field
_setting
(field
, 'convert') != 'string' and val_num
is not None
5942 return ((-10, 0) if value
is None
5943 else (1, value
, 0) if not is_num
# if a field has mixed strings and numbers, strings are sorted higher
5944 else (0, -abs(value
- limit
), value
- limit
if reverse
else limit
- value
) if closest
5945 else (0, value
, 0) if not reverse
and (limit
is None or value
<= limit
)
5946 else (0, -value
, 0) if limit
is None or (reverse
and value
== limit
) or value
> limit
5947 else (-1, value
, 0))
5949 def _calculate_field_preference(self
, format
, field
):
5950 type = self
._get
_field
_setting
(field
, 'type') # extractor, boolean, ordered, field, multiple
5951 get_value
= lambda f
: format
.get(self
._get
_field
_setting
(f
, 'field'))
5952 if type == 'multiple':
5953 type = 'field' # Only 'field' is allowed in multiple for now
5954 actual_fields
= self
._get
_field
_setting
(field
, 'field')
5956 value
= self
._get
_field
_setting
(field
, 'function')(get_value(f
) for f
in actual_fields
)
5958 value
= get_value(field
)
5959 return self
._calculate
_field
_preference
_from
_value
(format
, field
, type, value
)
5961 def calculate_preference(self
, format
):
5962 # Determine missing protocol
5963 if not format
.get('protocol'):
5964 format
['protocol'] = determine_protocol(format
)
5966 # Determine missing ext
5967 if not format
.get('ext') and 'url' in format
:
5968 format
['ext'] = determine_ext(format
['url'])
5969 if format
.get('vcodec') == 'none':
5970 format
['audio_ext'] = format
['ext'] if format
.get('acodec') != 'none' else 'none'
5971 format
['video_ext'] = 'none'
5973 format
['video_ext'] = format
['ext']
5974 format
['audio_ext'] = 'none'
5975 # if format.get('preference') is None and format.get('ext') in ('f4f', 'f4m'): # Not supported?
5976 # format['preference'] = -1000
5978 if format
.get('preference') is None and format
.get('ext') == 'flv' and re
.match('[hx]265|he?vc?', format
.get('vcodec') or ''):
5979 # HEVC-over-FLV is out-of-spec by FLV's original spec
5980 # ref. https://trac.ffmpeg.org/ticket/6389
5981 # ref. https://github.com/yt-dlp/yt-dlp/pull/5821
5982 format
['preference'] = -100
5984 # Determine missing bitrates
5985 if format
.get('vcodec') == 'none':
5987 if format
.get('acodec') == 'none':
5989 if not format
.get('vbr') and format
.get('vcodec') != 'none':
5990 format
['vbr'] = try_call(lambda: format
['tbr'] - format
['abr']) or None
5991 if not format
.get('abr') and format
.get('acodec') != 'none':
5992 format
['abr'] = try_call(lambda: format
['tbr'] - format
['vbr']) or None
5993 if not format
.get('tbr'):
5994 format
['tbr'] = try_call(lambda: format
['vbr'] + format
['abr']) or None
5996 return tuple(self
._calculate
_field
_preference
(format
, field
) for field
in self
._order
)
6001 def __init__(self
, ydl
=None):
6004 def debug(self
, message
):
6006 self
._ydl
.write_debug(message
)
6008 def info(self
, message
):
6010 self
._ydl
.to_screen(message
)
6012 def warning(self
, message
, *, once
=False):
6014 self
._ydl
.report_warning(message
, only_once
=once
)
6016 def error(self
, message
, *, is_error
=True):
6018 self
._ydl
.report_error(message
, is_error
=is_error
)
6020 def stdout(self
, message
):
6022 self
._ydl
.to_stdout(message
)
6024 def stderr(self
, message
):
6026 self
._ydl
.to_stderr(message
)