45 import xml
.etree
.ElementTree
47 from . import traversal
49 from ..compat
import functools
# isort: split
50 from ..compat
import (
51 compat_etree_fromstring
,
53 compat_HTMLParseError
,
57 from ..dependencies
import websockets
, xattr
59 __name__
= __name__
.rsplit('.', 1)[0] # Pretend to be the parent module
61 # This is not clearly defined otherwise
62 compiled_regex_type
= type(re
.compile(''))
73 ENGLISH_MONTH_NAMES
= [
74 'January', 'February', 'March', 'April', 'May', 'June',
75 'July', 'August', 'September', 'October', 'November', 'December']
78 'en': ENGLISH_MONTH_NAMES
,
80 'janvier', 'février', 'mars', 'avril', 'mai', 'juin',
81 'juillet', 'août', 'septembre', 'octobre', 'novembre', 'décembre'],
82 # these follow the genitive grammatical case (dopełniacz)
83 # some websites might be using nominative, which will require another month list
84 # https://en.wikibooks.org/wiki/Polish/Noun_cases
85 'pl': ['stycznia', 'lutego', 'marca', 'kwietnia', 'maja', 'czerwca',
86 'lipca', 'sierpnia', 'września', 'października', 'listopada', 'grudnia'],
89 # From https://github.com/python/cpython/blob/3.11/Lib/email/_parseaddr.py#L36-L42
91 'UT': 0, 'UTC': 0, 'GMT': 0, 'Z': 0,
92 'AST': -4, 'ADT': -3, # Atlantic (used in Canada)
93 'EST': -5, 'EDT': -4, # Eastern
94 'CST': -6, 'CDT': -5, # Central
95 'MST': -7, 'MDT': -6, # Mountain
96 'PST': -8, 'PDT': -7 # Pacific
99 # needed for sanitizing filenames in restricted mode
100 ACCENT_CHARS
= dict(zip('ÂÃÄÀÁÅÆÇÈÉÊËÌÍÎÏÐÑÒÓÔÕÖŐØŒÙÚÛÜŰÝÞßàáâãäåæçèéêëìíîïðñòóôõöőøœùúûüűýþÿ',
101 itertools
.chain('AAAAAA', ['AE'], 'CEEEEIIIIDNOOOOOOO', ['OE'], 'UUUUUY', ['TH', 'ss'],
102 'aaaaaa', ['ae'], 'ceeeeiiiionooooooo', ['oe'], 'uuuuuy', ['th'], 'y')))
132 '%Y-%m-%d %H:%M:%S.%f',
133 '%Y-%m-%d %H:%M:%S:%f',
136 '%Y-%m-%dT%H:%M:%SZ',
137 '%Y-%m-%dT%H:%M:%S.%fZ',
138 '%Y-%m-%dT%H:%M:%S.%f0Z',
140 '%Y-%m-%dT%H:%M:%S.%f',
143 '%b %d %Y at %H:%M:%S',
145 '%B %d %Y at %H:%M:%S',
149 DATE_FORMATS_DAY_FIRST
= list(DATE_FORMATS
)
150 DATE_FORMATS_DAY_FIRST
.extend([
161 DATE_FORMATS_MONTH_FIRST
= list(DATE_FORMATS
)
162 DATE_FORMATS_MONTH_FIRST
.extend([
170 PACKED_CODES_RE
= r
"}\('(.+)',(\d+),(\d+),'([^']+)'\.split\('\|'\)"
171 JSON_LD_RE
= r
'(?is)<script[^>]+type=(["\']?
)application
/ld\
+json\
1[^
>]*>\s
*(?P
<json_ld
>{.+?}|\
[.+?\
])\s
*</script
>'
173 NUMBER_RE = r'\d
+(?
:\
.\d
+)?
'
177 def preferredencoding():
178 """Get preferred encoding.
180 Returns the best encoding scheme for the system, based on
181 locale.getpreferredencoding() and some further tweaks.
184 pref = locale.getpreferredencoding()
192 def write_json_file(obj, fn):
193 """ Encode obj as JSON and write it to fn, atomically if possible """
195 tf = tempfile.NamedTemporaryFile(
196 prefix=f'{os.path.basename(fn)}
.', dir=os.path.dirname(fn),
197 suffix='.tmp
', delete=False, mode='w
', encoding='utf
-8')
201 json.dump(obj, tf, ensure_ascii=False)
202 if sys.platform == 'win32
':
203 # Need to remove existing file on Windows, else os.rename raises
204 # WindowsError or FileExistsError.
205 with contextlib.suppress(OSError):
207 with contextlib.suppress(OSError):
210 os.chmod(tf.name, 0o666 & ~mask)
211 os.rename(tf.name, fn)
213 with contextlib.suppress(OSError):
218 def find_xpath_attr(node, xpath, key, val=None):
219 """ Find the xpath xpath[@key=val] """
220 assert re.match(r'^
[a
-zA
-Z_
-]+$
', key)
221 expr = xpath + ('[@%s]' % key if val is None else f"[@{key}='{val}
']")
222 return node.find(expr)
224 # On python2.6 the xml.etree.ElementTree.Element methods don't support
225 # the namespace parameter
228 def xpath_with_ns(path
, ns_map
):
229 components
= [c
.split(':') for c
in path
.split('/')]
233 replaced
.append(c
[0])
236 replaced
.append('{%s}%s' % (ns_map
[ns
], tag
))
237 return '/'.join(replaced
)
240 def xpath_element(node
, xpath
, name
=None, fatal
=False, default
=NO_DEFAULT
):
241 def _find_xpath(xpath
):
242 return node
.find(xpath
)
244 if isinstance(xpath
, str):
245 n
= _find_xpath(xpath
)
253 if default
is not NO_DEFAULT
:
256 name
= xpath
if name
is None else name
257 raise ExtractorError('Could not find XML element %s' % name
)
263 def xpath_text(node
, xpath
, name
=None, fatal
=False, default
=NO_DEFAULT
):
264 n
= xpath_element(node
, xpath
, name
, fatal
=fatal
, default
=default
)
265 if n
is None or n
== default
:
268 if default
is not NO_DEFAULT
:
271 name
= xpath
if name
is None else name
272 raise ExtractorError('Could not find XML element\'s text %s' % name
)
278 def xpath_attr(node
, xpath
, key
, name
=None, fatal
=False, default
=NO_DEFAULT
):
279 n
= find_xpath_attr(node
, xpath
, key
)
281 if default
is not NO_DEFAULT
:
284 name
= f
'{xpath}[@{key}]' if name
is None else name
285 raise ExtractorError('Could not find XML attribute %s' % name
)
291 def get_element_by_id(id, html
, **kwargs
):
292 """Return the content of the tag with the specified ID in the passed HTML document"""
293 return get_element_by_attribute('id', id, html
, **kwargs
)
296 def get_element_html_by_id(id, html
, **kwargs
):
297 """Return the html of the tag with the specified ID in the passed HTML document"""
298 return get_element_html_by_attribute('id', id, html
, **kwargs
)
301 def get_element_by_class(class_name
, html
):
302 """Return the content of the first tag with the specified class in the passed HTML document"""
303 retval
= get_elements_by_class(class_name
, html
)
304 return retval
[0] if retval
else None
307 def get_element_html_by_class(class_name
, html
):
308 """Return the html of the first tag with the specified class in the passed HTML document"""
309 retval
= get_elements_html_by_class(class_name
, html
)
310 return retval
[0] if retval
else None
313 def get_element_by_attribute(attribute
, value
, html
, **kwargs
):
314 retval
= get_elements_by_attribute(attribute
, value
, html
, **kwargs
)
315 return retval
[0] if retval
else None
318 def get_element_html_by_attribute(attribute
, value
, html
, **kargs
):
319 retval
= get_elements_html_by_attribute(attribute
, value
, html
, **kargs
)
320 return retval
[0] if retval
else None
323 def get_elements_by_class(class_name
, html
, **kargs
):
324 """Return the content of all tags with the specified class in the passed HTML document as a list"""
325 return get_elements_by_attribute(
326 'class', r
'[^\'"]*(?<=[\'"\s
])%s(?
=[\'"\s])[^\'"]*' % re.escape(class_name),
327 html, escape_value=False)
330 def get_elements_html_by_class(class_name, html):
331 """Return the html of all tags with the specified class in the passed HTML document as a list"""
332 return get_elements_html_by_attribute(
333 'class', r'[^
\'"]*(?<=[\'"\s
])%s(?
=[\'"\s])[^\'"]*' % re.escape(class_name),
334 html, escape_value=False)
337 def get_elements_by_attribute(*args, **kwargs):
338 """Return the content of the tag with the specified attribute in the passed HTML document"""
339 return [content for content, _ in get_elements_text_and_html_by_attribute(*args, **kwargs)]
342 def get_elements_html_by_attribute(*args, **kwargs):
343 """Return the html of the tag with the specified attribute in the passed HTML document"""
344 return [whole for _, whole in get_elements_text_and_html_by_attribute(*args, **kwargs)]
347 def get_elements_text_and_html_by_attribute(attribute, value, html, *, tag=r'[\w
:.-]+', escape_value=True):
349 Return the text (content) and the html (whole) of the tag with the specified
350 attribute in the passed HTML document
355 quote = '' if re.match(r'''[\s"'`
=<>]''', value) else '?'
357 value = re.escape(value) if escape_value else value
359 partial_element_re = rf'''(?x
)
361 (?
:\
s(?
:[^
>"']|"[^
"]*"|
'[^']*')*)?
362 \s{re.escape(attribute)}\s*=\s*(?P<_q>['"]{quote})(?-x:{value})(?P=_q)
365 for m in re.finditer(partial_element_re, html):
366 content, whole = get_element_text_and_html_by_tag(m.group('tag'), html[m.start():])
369 unescapeHTML(re.sub(r'^(?P<q>["\'])(?P
<content
>.*)(?P
=q
)$
', r'\g
<content
>', content, flags=re.DOTALL)),
374 class HTMLBreakOnClosingTagParser(html.parser.HTMLParser):
376 HTML parser which raises HTMLBreakOnClosingTagException upon reaching the
377 closing tag for the first opening tag it has encountered, and can be used
381 class HTMLBreakOnClosingTagException(Exception):
385 self.tagstack = collections.deque()
386 html.parser.HTMLParser.__init__(self)
391 def __exit__(self, *_):
395 # handle_endtag does not return upon raising HTMLBreakOnClosingTagException,
396 # so data remains buffered; we no longer have any interest in it, thus
397 # override this method to discard it
400 def handle_starttag(self, tag, _):
401 self.tagstack.append(tag)
403 def handle_endtag(self, tag):
404 if not self.tagstack:
405 raise compat_HTMLParseError('no tags
in the stack
')
407 inner_tag = self.tagstack.pop()
411 raise compat_HTMLParseError(f'matching opening tag
for closing {tag} tag
not found
')
412 if not self.tagstack:
413 raise self.HTMLBreakOnClosingTagException()
416 # XXX: This should be far less strict
417 def get_element_text_and_html_by_tag(tag, html):
419 For the first element with the specified tag in the passed HTML document
420 return its' content (text
) and the whole
element (html
)
422 def find_or_raise(haystack, needle, exc):
424 return haystack.index(needle)
427 closing_tag = f'</{tag}>'
428 whole_start = find_or_raise(
429 html, f'<{tag}', compat_HTMLParseError(f'opening {tag} tag not found'))
430 content_start = find_or_raise(
431 html[whole_start:], '>', compat_HTMLParseError(f'malformed opening {tag} tag'))
432 content_start += whole_start + 1
433 with HTMLBreakOnClosingTagParser() as parser:
434 parser.feed(html[whole_start:content_start])
435 if not parser.tagstack or parser.tagstack[0] != tag:
436 raise compat_HTMLParseError(f'parser did not match opening {tag} tag')
437 offset = content_start
438 while offset < len(html):
439 next_closing_tag_start = find_or_raise(
440 html[offset:], closing_tag,
441 compat_HTMLParseError(f'closing {tag} tag not found'))
442 next_closing_tag_end = next_closing_tag_start + len(closing_tag)
444 parser.feed(html[offset:offset + next_closing_tag_end])
445 offset += next_closing_tag_end
446 except HTMLBreakOnClosingTagParser.HTMLBreakOnClosingTagException:
447 return html[content_start:offset + next_closing_tag_start], \
448 html[whole_start:offset + next_closing_tag_end]
449 raise compat_HTMLParseError('unexpected end of html')
452 class HTMLAttributeParser(html.parser.HTMLParser):
453 """Trivial HTML parser to gather the attributes
for a single element
"""
457 html.parser.HTMLParser.__init__(self)
459 def handle_starttag(self, tag, attrs):
460 self.attrs = dict(attrs)
461 raise compat_HTMLParseError('done')
464 class HTMLListAttrsParser(html.parser.HTMLParser):
465 """HTML parser to gather the attributes
for the elements of a
list"""
468 html.parser.HTMLParser.__init__(self)
472 def handle_starttag(self, tag, attrs):
473 if tag == 'li' and self._level == 0:
474 self.items.append(dict(attrs))
477 def handle_endtag(self, tag):
481 def extract_attributes(html_element):
482 """Given a string
for an HTML element such
as
484 a
="foo" B
="bar" c
="&98;az" d
=boz
485 empty
= noval entity
="&"
488 Decode
and return a dictionary of attributes
.
490 'a': 'foo', 'b': 'bar', c
: 'baz', d
: 'boz',
491 'empty': '', 'noval': None, 'entity': '&',
492 'sq': '"', 'dq': '\''
495 parser = HTMLAttributeParser()
496 with contextlib.suppress(compat_HTMLParseError):
497 parser.feed(html_element)
502 def parse_list(webpage):
503 """Given a string
for an series of HTML
<li
> elements
,
504 return a dictionary of their attributes
"""
505 parser = HTMLListAttrsParser()
511 def clean_html(html):
512 """Clean an HTML snippet into a readable string
"""
514 if html is None: # Convenience for sanitizing descriptions etc.
517 html = re.sub(r'\s+', ' ', html)
518 html = re.sub(r'(?u)\s?<\s?br\s?/?\s?>\s?', '\n', html)
519 html = re.sub(r'(?u)<\s?/\s?p\s?>\s?<\s?p[^>]*>', '\n', html)
521 html = re.sub('<.*?>', '', html)
522 # Replace html entities
523 html = unescapeHTML(html)
527 class LenientJSONDecoder(json.JSONDecoder):
529 def __init__(self, *args, transform_source=None, ignore_extra=False, close_objects=0, **kwargs):
530 self.transform_source, self.ignore_extra = transform_source, ignore_extra
531 self._close_attempts = 2 * close_objects
532 super().__init__(*args, **kwargs)
535 def _close_object(err):
536 doc = err.doc[:err.pos]
537 # We need to add comma first to get the correct error message
538 if err.msg.startswith('Expecting \',\''):
540 elif not doc.endswith(','):
543 if err.msg.startswith('Expecting property name'):
544 return doc[:-1] + '}'
545 elif err.msg.startswith('Expecting value'):
546 return doc[:-1] + ']'
549 if self.transform_source:
550 s = self.transform_source(s)
551 for attempt in range(self._close_attempts + 1):
553 if self.ignore_extra:
554 return self.raw_decode(s.lstrip())[0]
555 return super().decode(s)
556 except json.JSONDecodeError as e:
559 elif attempt < self._close_attempts:
560 s = self._close_object(e)
563 raise type(e)(f'{e.msg} in {s[e.pos-10:e.pos+10]!r}', s, e.pos)
564 assert False, 'Too many attempts to decode JSON'
567 def sanitize_open(filename, open_mode):
568 """Try to
open the given filename
, and slightly tweak it
if this fails
.
570 Attempts to
open the given filename
. If this fails
, it tries to change
571 the filename slightly
, step by step
, until it
's either able to open it
572 or it fails and raises a final exception, like the standard open()
575 It returns the tuple (stream, definitive_file_name).
578 if sys.platform == 'win32
':
581 # stdout may be any IO stream, e.g. when using contextlib.redirect_stdout
582 with contextlib.suppress(io.UnsupportedOperation):
583 msvcrt.setmode(sys.stdout.fileno(), os.O_BINARY)
584 return (sys.stdout.buffer if hasattr(sys.stdout, 'buffer') else sys.stdout, filename)
586 for attempt in range(2):
589 if sys.platform == 'win32
':
590 # FIXME: An exclusive lock also locks the file from being read.
591 # Since windows locks are mandatory, don't lock the
file on
windows (for now
).
592 # Ref: https://github.com/yt-dlp/yt-dlp/issues/3124
593 raise LockingUnsupportedError()
594 stream
= locked_file(filename
, open_mode
, block
=False).__enter
__()
596 stream
= open(filename
, open_mode
)
597 return stream
, filename
598 except OSError as err
:
599 if attempt
or err
.errno
in (errno
.EACCES
,):
601 old_filename
, filename
= filename
, sanitize_path(filename
)
602 if old_filename
== filename
:
606 def timeconvert(timestr
):
607 """Convert RFC 2822 defined time string into system timestamp"""
609 timetuple
= email
.utils
.parsedate_tz(timestr
)
610 if timetuple
is not None:
611 timestamp
= email
.utils
.mktime_tz(timetuple
)
615 def sanitize_filename(s
, restricted
=False, is_id
=NO_DEFAULT
):
616 """Sanitizes a string so it could be used as part of a filename.
617 @param restricted Use a stricter subset of allowed characters
618 @param is_id Whether this is an ID that should be kept unchanged if possible.
619 If unset, yt-dlp's new sanitization rules are in effect
624 def replace_insane(char
):
625 if restricted
and char
in ACCENT_CHARS
:
626 return ACCENT_CHARS
[char
]
627 elif not restricted
and char
== '\n':
629 elif is_id
is NO_DEFAULT
and not restricted
and char
in '"*:<>?|/\\':
630 # Replace with their full-width unicode counterparts
631 return {'/': '\u29F8', '\\': '\u29f9'}
.get(char
, chr(ord(char
) + 0xfee0))
632 elif char
== '?' or ord(char
) < 32 or ord(char
) == 127:
635 return '' if restricted
else '\''
637 return '\0_\0-' if restricted
else '\0 \0-'
638 elif char
in '\\/|*<>':
640 if restricted
and (char
in '!&\'()[]{}$;`^,#' or char
.isspace() or ord(char
) > 127):
644 # Replace look-alike Unicode glyphs
645 if restricted
and (is_id
is NO_DEFAULT
or not is_id
):
646 s
= unicodedata
.normalize('NFKC', s
)
647 s
= re
.sub(r
'[0-9]+(?::[0-9]+)+', lambda m
: m
.group(0).replace(':', '_'), s
) # Handle timestamps
648 result
= ''.join(map(replace_insane
, s
))
649 if is_id
is NO_DEFAULT
:
650 result
= re
.sub(r
'(\0.)(?:(?=\1)..)+', r
'\1', result
) # Remove repeated substitute chars
651 STRIP_RE
= r
'(?:\0.|[ _-])*'
652 result
= re
.sub(f
'^\0.{STRIP_RE}|{STRIP_RE}\0.$', '', result
) # Remove substitute chars from start/end
653 result
= result
.replace('\0', '') or '_'
656 while '__' in result
:
657 result
= result
.replace('__', '_')
658 result
= result
.strip('_')
659 # Common case of "Foreign band name - English song title"
660 if restricted
and result
.startswith('-_'):
662 if result
.startswith('-'):
663 result
= '_' + result
[len('-'):]
664 result
= result
.lstrip('.')
670 def sanitize_path(s
, force
=False):
671 """Sanitizes and normalizes path on Windows"""
672 # XXX: this handles drive relative paths (c:sth) incorrectly
673 if sys
.platform
== 'win32':
675 drive_or_unc
, _
= os
.path
.splitdrive(s
)
681 norm_path
= os
.path
.normpath(remove_start(s
, drive_or_unc
)).split(os
.path
.sep
)
685 path_part
if path_part
in ['.', '..'] else re
.sub(r
'(?:[/<>:"\|\\?\*]|[\s.]$)', '#', path_part
)
686 for path_part
in norm_path
]
688 sanitized_path
.insert(0, drive_or_unc
+ os
.path
.sep
)
689 elif force
and s
and s
[0] == os
.path
.sep
:
690 sanitized_path
.insert(0, os
.path
.sep
)
691 # TODO: Fix behavioral differences <3.12
692 # The workaround using `normpath` only superficially passes tests
693 # Ref: https://github.com/python/cpython/pull/100351
694 return os
.path
.normpath(os
.path
.join(*sanitized_path
))
697 def sanitize_url(url
, *, scheme
='http'):
698 # Prepend protocol-less URLs with `http:` scheme in order to mitigate
699 # the number of unwanted failures due to missing protocol
702 elif url
.startswith('//'):
703 return f
'{scheme}:{url}'
704 # Fix some common typos seen so far
706 # https://github.com/ytdl-org/youtube-dl/issues/15649
707 (r
'^httpss://', r
'https://'),
708 # https://bx1.be/lives/direct-tv/
709 (r
'^rmtp([es]?)://', r
'rtmp\1://'),
711 for mistake
, fixup
in COMMON_TYPOS
:
712 if re
.match(mistake
, url
):
713 return re
.sub(mistake
, fixup
, url
)
717 def extract_basic_auth(url
):
718 parts
= urllib
.parse
.urlsplit(url
)
719 if parts
.username
is None:
721 url
= urllib
.parse
.urlunsplit(parts
._replace
(netloc
=(
722 parts
.hostname
if parts
.port
is None
723 else '%s:%d' % (parts
.hostname
, parts
.port
))))
724 auth_payload
= base64
.b64encode(
725 ('%s:%s' % (parts
.username
, parts
.password
or '')).encode())
726 return url
, f
'Basic {auth_payload.decode()}'
730 """Expand shell variables and ~"""
731 return os
.path
.expandvars(compat_expanduser(s
))
734 def orderedSet(iterable
, *, lazy
=False):
735 """Remove all duplicates from the input iterable"""
737 seen
= [] # Do not use set since the items can be unhashable
743 return _iter() if lazy
else list(_iter())
746 def _htmlentity_transform(entity_with_semicolon
):
747 """Transforms an HTML entity to a character."""
748 entity
= entity_with_semicolon
[:-1]
750 # Known non-numeric HTML entity
751 if entity
in html
.entities
.name2codepoint
:
752 return chr(html
.entities
.name2codepoint
[entity
])
754 # TODO: HTML5 allows entities without a semicolon.
755 # E.g. 'Éric' should be decoded as 'Éric'.
756 if entity_with_semicolon
in html
.entities
.html5
:
757 return html
.entities
.html5
[entity_with_semicolon
]
759 mobj
= re
.match(r
'#(x[0-9a-fA-F]+|[0-9]+)', entity
)
761 numstr
= mobj
.group(1)
762 if numstr
.startswith('x'):
764 numstr
= '0%s' % numstr
767 # See https://github.com/ytdl-org/youtube-dl/issues/7518
768 with contextlib
.suppress(ValueError):
769 return chr(int(numstr
, base
))
771 # Unknown entity in name, return its literal representation
772 return '&%s;' % entity
778 assert isinstance(s
, str)
781 r
'&([^&;]+;)', lambda m
: _htmlentity_transform(m
.group(1)), s
)
784 def escapeHTML(text
):
787 .replace('&', '&')
788 .replace('<', '<')
789 .replace('>', '>')
790 .replace('"', '"')
791 .replace("'", ''')
795 class netrc_from_content(netrc
.netrc
):
796 def __init__(self
, content
):
797 self
.hosts
, self
.macros
= {}, {}
798 with io
.StringIO(content
) as stream
:
799 self
._parse
('-', stream
, False)
802 class Popen(subprocess
.Popen
):
803 if sys
.platform
== 'win32':
804 _startupinfo
= subprocess
.STARTUPINFO()
805 _startupinfo
.dwFlags |
= subprocess
.STARTF_USESHOWWINDOW
810 def _fix_pyinstaller_ld_path(env
):
811 """Restore LD_LIBRARY_PATH when using PyInstaller
812 Ref: https://github.com/pyinstaller/pyinstaller/blob/develop/doc/runtime-information.rst#ld_library_path--libpath-considerations
813 https://github.com/yt-dlp/yt-dlp/issues/4573
815 if not hasattr(sys
, '_MEIPASS'):
819 orig
= env
.get(f
'{key}_ORIG')
825 _fix('LD_LIBRARY_PATH') # Linux
826 _fix('DYLD_LIBRARY_PATH') # macOS
828 def __init__(self
, args
, *remaining
, env
=None, text
=False, shell
=False, **kwargs
):
830 env
= os
.environ
.copy()
831 self
._fix
_pyinstaller
_ld
_path
(env
)
833 self
.__text
_mode
= kwargs
.get('encoding') or kwargs
.get('errors') or text
or kwargs
.get('universal_newlines')
835 kwargs
['universal_newlines'] = True # For 3.6 compatibility
836 kwargs
.setdefault('encoding', 'utf-8')
837 kwargs
.setdefault('errors', 'replace')
839 if shell
and compat_os_name
== 'nt' and kwargs
.get('executable') is None:
840 if not isinstance(args
, str):
841 args
= ' '.join(compat_shlex_quote(a
) for a
in args
)
843 args
= f
'{self.__comspec()} /Q /S /D /V:OFF /C "{args}"'
845 super().__init
__(args
, *remaining
, env
=env
, shell
=shell
, **kwargs
, startupinfo
=self
._startupinfo
)
848 comspec
= os
.environ
.get('ComSpec') or os
.path
.join(
849 os
.environ
.get('SystemRoot', ''), 'System32', 'cmd.exe')
850 if os
.path
.isabs(comspec
):
852 raise FileNotFoundError('shell not found: neither %ComSpec% nor %SystemRoot% is set')
854 def communicate_or_kill(self
, *args
, **kwargs
):
856 return self
.communicate(*args
, **kwargs
)
857 except BaseException
: # Including KeyboardInterrupt
858 self
.kill(timeout
=None)
861 def kill(self
, *, timeout
=0):
864 self
.wait(timeout
=timeout
)
867 def run(cls
, *args
, timeout
=None, **kwargs
):
868 with cls(*args
, **kwargs
) as proc
:
869 default
= '' if proc
.__text
_mode
else b
''
870 stdout
, stderr
= proc
.communicate_or_kill(timeout
=timeout
)
871 return stdout
or default
, stderr
or default
, proc
.returncode
874 def encodeArgument(s
):
875 # Legacy code that uses byte strings
876 # Uncomment the following line after fixing all post processors
877 # assert isinstance(s, str), 'Internal error: %r should be of type %r, is %r' % (s, str, type(s))
878 return s
if isinstance(s
, str) else s
.decode('ascii')
881 _timetuple
= collections
.namedtuple('Time', ('hours', 'minutes', 'seconds', 'milliseconds'))
884 def timetuple_from_msec(msec
):
885 secs
, msec
= divmod(msec
, 1000)
886 mins
, secs
= divmod(secs
, 60)
887 hrs
, mins
= divmod(mins
, 60)
888 return _timetuple(hrs
, mins
, secs
, msec
)
891 def formatSeconds(secs
, delim
=':', msec
=False):
892 time
= timetuple_from_msec(secs
* 1000)
894 ret
= '%d%s%02d%s%02d' % (time
.hours
, delim
, time
.minutes
, delim
, time
.seconds
)
896 ret
= '%d%s%02d' % (time
.minutes
, delim
, time
.seconds
)
898 ret
= '%d' % time
.seconds
899 return '%s.%03d' % (ret
, time
.milliseconds
) if msec
else ret
902 def bug_reports_message(before
=';'):
903 from ..update
import REPOSITORY
905 msg
= (f
'please report this issue on https://github.com/{REPOSITORY}/issues?q= , '
906 'filling out the appropriate issue template. Confirm you are on the latest version using yt-dlp -U')
908 before
= before
.rstrip()
909 if not before
or before
.endswith(('.', '!', '?')):
910 msg
= msg
[0].title() + msg
[1:]
912 return (before
+ ' ' if before
else '') + msg
915 class YoutubeDLError(Exception):
916 """Base exception for YoutubeDL errors."""
919 def __init__(self
, msg
=None):
922 elif self
.msg
is None:
923 self
.msg
= type(self
).__name
__
924 super().__init
__(self
.msg
)
927 class ExtractorError(YoutubeDLError
):
928 """Error during info extraction."""
930 def __init__(self
, msg
, tb
=None, expected
=False, cause
=None, video_id
=None, ie
=None):
931 """ tb, if given, is the original traceback (so that it can be printed out).
932 If expected is set, this is a normal error message and most likely not a bug in yt-dlp.
934 from ..networking
.exceptions
import network_exceptions
935 if sys
.exc_info()[0] in network_exceptions
:
938 self
.orig_msg
= str(msg
)
940 self
.expected
= expected
942 self
.video_id
= video_id
944 self
.exc_info
= sys
.exc_info() # preserve original exception
945 if isinstance(self
.exc_info
[1], ExtractorError
):
946 self
.exc_info
= self
.exc_info
[1].exc_info
947 super().__init
__(self
.__msg
)
952 format_field(self
.ie
, None, '[%s] '),
953 format_field(self
.video_id
, None, '%s: '),
955 format_field(self
.cause
, None, ' (caused by %r)'),
956 '' if self
.expected
else bug_reports_message()))
958 def format_traceback(self
):
959 return join_nonempty(
960 self
.traceback
and ''.join(traceback
.format_tb(self
.traceback
)),
961 self
.cause
and ''.join(traceback
.format_exception(None, self
.cause
, self
.cause
.__traceback
__)[1:]),
964 def __setattr__(self
, name
, value
):
965 super().__setattr
__(name
, value
)
966 if getattr(self
, 'msg', None) and name
not in ('msg', 'args'):
967 self
.msg
= self
.__msg
or type(self
).__name
__
968 self
.args
= (self
.msg
, ) # Cannot be property
971 class UnsupportedError(ExtractorError
):
972 def __init__(self
, url
):
974 'Unsupported URL: %s' % url
, expected
=True)
978 class RegexNotFoundError(ExtractorError
):
979 """Error when a regex didn't match"""
983 class GeoRestrictedError(ExtractorError
):
984 """Geographic restriction Error exception.
986 This exception may be thrown when a video is not available from your
987 geographic location due to geographic restrictions imposed by a website.
990 def __init__(self
, msg
, countries
=None, **kwargs
):
991 kwargs
['expected'] = True
992 super().__init
__(msg
, **kwargs
)
993 self
.countries
= countries
996 class UserNotLive(ExtractorError
):
997 """Error when a channel/user is not live"""
999 def __init__(self
, msg
=None, **kwargs
):
1000 kwargs
['expected'] = True
1001 super().__init
__(msg
or 'The channel is not currently live', **kwargs
)
1004 class DownloadError(YoutubeDLError
):
1005 """Download Error exception.
1007 This exception may be thrown by FileDownloader objects if they are not
1008 configured to continue on errors. They will contain the appropriate
1012 def __init__(self
, msg
, exc_info
=None):
1013 """ exc_info, if given, is the original exception that caused the trouble (as returned by sys.exc_info()). """
1014 super().__init
__(msg
)
1015 self
.exc_info
= exc_info
1018 class EntryNotInPlaylist(YoutubeDLError
):
1019 """Entry not in playlist exception.
1021 This exception will be thrown by YoutubeDL when a requested entry
1022 is not found in the playlist info_dict
1024 msg
= 'Entry not found in info'
1027 class SameFileError(YoutubeDLError
):
1028 """Same File exception.
1030 This exception will be thrown by FileDownloader objects if they detect
1031 multiple files would have to be downloaded to the same file on disk.
1033 msg
= 'Fixed output name but more than one file to download'
1035 def __init__(self
, filename
=None):
1036 if filename
is not None:
1037 self
.msg
+= f
': {filename}'
1038 super().__init
__(self
.msg
)
1041 class PostProcessingError(YoutubeDLError
):
1042 """Post Processing exception.
1044 This exception may be raised by PostProcessor's .run() method to
1045 indicate an error in the postprocessing task.
1049 class DownloadCancelled(YoutubeDLError
):
1050 """ Exception raised when the download queue should be interrupted """
1051 msg
= 'The download was cancelled'
1054 class ExistingVideoReached(DownloadCancelled
):
1055 """ --break-on-existing triggered """
1056 msg
= 'Encountered a video that is already in the archive, stopping due to --break-on-existing'
1059 class RejectedVideoReached(DownloadCancelled
):
1060 """ --break-match-filter triggered """
1061 msg
= 'Encountered a video that did not match filter, stopping due to --break-match-filter'
1064 class MaxDownloadsReached(DownloadCancelled
):
1065 """ --max-downloads limit has been reached. """
1066 msg
= 'Maximum number of downloads reached, stopping due to --max-downloads'
1069 class ReExtractInfo(YoutubeDLError
):
1070 """ Video info needs to be re-extracted. """
1072 def __init__(self
, msg
, expected
=False):
1073 super().__init
__(msg
)
1074 self
.expected
= expected
1077 class ThrottledDownload(ReExtractInfo
):
1078 """ Download speed below --throttled-rate. """
1079 msg
= 'The download speed is below throttle limit'
1082 super().__init
__(self
.msg
, expected
=False)
1085 class UnavailableVideoError(YoutubeDLError
):
1086 """Unavailable Format exception.
1088 This exception will be thrown when a video is requested
1089 in a format that is not available for that video.
1091 msg
= 'Unable to download video'
1093 def __init__(self
, err
=None):
1095 self
.msg
+= f
': {err}'
1096 super().__init
__(self
.msg
)
1099 class ContentTooShortError(YoutubeDLError
):
1100 """Content Too Short exception.
1102 This exception may be raised by FileDownloader objects when a file they
1103 download is too small for what the server announced first, indicating
1104 the connection was probably interrupted.
1107 def __init__(self
, downloaded
, expected
):
1108 super().__init
__(f
'Downloaded {downloaded} bytes, expected {expected} bytes')
1110 self
.downloaded
= downloaded
1111 self
.expected
= expected
1114 class XAttrMetadataError(YoutubeDLError
):
1115 def __init__(self
, code
=None, msg
='Unknown error'):
1116 super().__init
__(msg
)
1120 # Parsing code and msg
1121 if (self
.code
in (errno
.ENOSPC
, errno
.EDQUOT
)
1122 or 'No space left' in self
.msg
or 'Disk quota exceeded' in self
.msg
):
1123 self
.reason
= 'NO_SPACE'
1124 elif self
.code
== errno
.E2BIG
or 'Argument list too long' in self
.msg
:
1125 self
.reason
= 'VALUE_TOO_LONG'
1127 self
.reason
= 'NOT_SUPPORTED'
1130 class XAttrUnavailableError(YoutubeDLError
):
1134 def is_path_like(f
):
1135 return isinstance(f
, (str, bytes, os
.PathLike
))
1138 def extract_timezone(date_str
):
1141 ^.{8,}? # >=8 char non-TZ prefix, if present
1142 (?P<tz>Z| # just the UTC Z, or
1143 (?:(?<=.\b\d{4}|\b\d{2}:\d\d)| # preceded by 4 digits or hh:mm or
1144 (?<!.\b[a-zA-Z]{3}|[a-zA-Z]{4}|..\b\d\d)) # not preceded by 3 alpha word or >= 4 alpha or 2 digits
1145 [ ]? # optional space
1146 (?P<sign>\+|-) # +/-
1147 (?P<hours>[0-9]{2}):?(?P<minutes>[0-9]{2}) # hh[:]mm
1151 m
= re
.search(r
'\d{1,2}:\d{1,2}(?:\.\d+)?(?P<tz>\s*[A-Z]+)$', date_str
)
1152 timezone
= TIMEZONE_NAMES
.get(m
and m
.group('tz').strip())
1153 if timezone
is not None:
1154 date_str
= date_str
[:-len(m
.group('tz'))]
1155 timezone
= datetime
.timedelta(hours
=timezone
or 0)
1157 date_str
= date_str
[:-len(m
.group('tz'))]
1158 if not m
.group('sign'):
1159 timezone
= datetime
.timedelta()
1161 sign
= 1 if m
.group('sign') == '+' else -1
1162 timezone
= datetime
.timedelta(
1163 hours
=sign
* int(m
.group('hours')),
1164 minutes
=sign
* int(m
.group('minutes')))
1165 return timezone
, date_str
1168 def parse_iso8601(date_str
, delimiter
='T', timezone
=None):
1169 """ Return a UNIX timestamp from the given date """
1171 if date_str
is None:
1174 date_str
= re
.sub(r
'\.[0-9]+', '', date_str
)
1176 if timezone
is None:
1177 timezone
, date_str
= extract_timezone(date_str
)
1179 with contextlib
.suppress(ValueError):
1180 date_format
= f
'%Y-%m-%d{delimiter}%H:%M:%S'
1181 dt
= datetime
.datetime
.strptime(date_str
, date_format
) - timezone
1182 return calendar
.timegm(dt
.timetuple())
1185 def date_formats(day_first
=True):
1186 return DATE_FORMATS_DAY_FIRST
if day_first
else DATE_FORMATS_MONTH_FIRST
1189 def unified_strdate(date_str
, day_first
=True):
1190 """Return a string with the date in the format YYYYMMDD"""
1192 if date_str
is None:
1196 date_str
= date_str
.replace(',', ' ')
1197 # Remove AM/PM + timezone
1198 date_str
= re
.sub(r
'(?i)\s*(?:AM|PM)(?:\s+[A-Z]+)?', '', date_str
)
1199 _
, date_str
= extract_timezone(date_str
)
1201 for expression
in date_formats(day_first
):
1202 with contextlib
.suppress(ValueError):
1203 upload_date
= datetime
.datetime
.strptime(date_str
, expression
).strftime('%Y%m%d')
1204 if upload_date
is None:
1205 timetuple
= email
.utils
.parsedate_tz(date_str
)
1207 with contextlib
.suppress(ValueError):
1208 upload_date
= datetime
.datetime(*timetuple
[:6]).strftime('%Y%m%d')
1209 if upload_date
is not None:
1210 return str(upload_date
)
1213 def unified_timestamp(date_str
, day_first
=True):
1214 if not isinstance(date_str
, str):
1217 date_str
= re
.sub(r
'\s+', ' ', re
.sub(
1218 r
'(?i)[,|]|(mon|tues?|wed(nes)?|thu(rs)?|fri|sat(ur)?)(day)?', '', date_str
))
1220 pm_delta
= 12 if re
.search(r
'(?i)PM', date_str
) else 0
1221 timezone
, date_str
= extract_timezone(date_str
)
1223 # Remove AM/PM + timezone
1224 date_str
= re
.sub(r
'(?i)\s*(?:AM|PM)(?:\s+[A-Z]+)?', '', date_str
)
1226 # Remove unrecognized timezones from ISO 8601 alike timestamps
1227 m
= re
.search(r
'\d{1,2}:\d{1,2}(?:\.\d+)?(?P<tz>\s*[A-Z]+)$', date_str
)
1229 date_str
= date_str
[:-len(m
.group('tz'))]
1231 # Python only supports microseconds, so remove nanoseconds
1232 m
= re
.search(r
'^([0-9]{4,}-[0-9]{1,2}-[0-9]{1,2}T[0-9]{1,2}:[0-9]{1,2}:[0-9]{1,2}\.[0-9]{6})[0-9]+$', date_str
)
1234 date_str
= m
.group(1)
1236 for expression
in date_formats(day_first
):
1237 with contextlib
.suppress(ValueError):
1238 dt
= datetime
.datetime
.strptime(date_str
, expression
) - timezone
+ datetime
.timedelta(hours
=pm_delta
)
1239 return calendar
.timegm(dt
.timetuple())
1241 timetuple
= email
.utils
.parsedate_tz(date_str
)
1243 return calendar
.timegm(timetuple
) + pm_delta
* 3600 - timezone
.total_seconds()
1246 def determine_ext(url
, default_ext
='unknown_video'):
1247 if url
is None or '.' not in url
:
1249 guess
= url
.partition('?')[0].rpartition('.')[2]
1250 if re
.match(r
'^[A-Za-z0-9]+$', guess
):
1252 # Try extract ext from URLs like http://example.com/foo/bar.mp4/?download
1253 elif guess
.rstrip('/') in KNOWN_EXTENSIONS
:
1254 return guess
.rstrip('/')
1259 def subtitles_filename(filename
, sub_lang
, sub_format
, expected_real_ext
=None):
1260 return replace_extension(filename
, sub_lang
+ '.' + sub_format
, expected_real_ext
)
1263 def datetime_from_str(date_str
, precision
='auto', format
='%Y%m%d'):
1265 Return a datetime object from a string.
1267 (now|today|yesterday|DATE)([+-]\d+(microsecond|second|minute|hour|day|week|month|year)s?)?
1269 @param format strftime format of DATE
1270 @param precision Round the datetime object: auto|microsecond|second|minute|hour|day
1271 auto: round to the unit provided in date_str (if applicable).
1273 auto_precision
= False
1274 if precision
== 'auto':
1275 auto_precision
= True
1276 precision
= 'microsecond'
1277 today
= datetime_round(datetime
.datetime
.now(datetime
.timezone
.utc
), precision
)
1278 if date_str
in ('now', 'today'):
1280 if date_str
== 'yesterday':
1281 return today
- datetime
.timedelta(days
=1)
1283 r
'(?P<start>.+)(?P<sign>[+-])(?P<time>\d+)(?P<unit>microsecond|second|minute|hour|day|week|month|year)s?',
1285 if match
is not None:
1286 start_time
= datetime_from_str(match
.group('start'), precision
, format
)
1287 time
= int(match
.group('time')) * (-1 if match
.group('sign') == '-' else 1)
1288 unit
= match
.group('unit')
1289 if unit
== 'month' or unit
== 'year':
1290 new_date
= datetime_add_months(start_time
, time
* 12 if unit
== 'year' else time
)
1296 delta
= datetime
.timedelta(**{unit + 's': time}
)
1297 new_date
= start_time
+ delta
1299 return datetime_round(new_date
, unit
)
1302 return datetime_round(datetime
.datetime
.strptime(date_str
, format
), precision
)
1305 def date_from_str(date_str
, format
='%Y%m%d', strict
=False):
1307 Return a date object from a string using datetime_from_str
1309 @param strict Restrict allowed patterns to "YYYYMMDD" and
1310 (now|today|yesterday)(-\d+(day|week|month|year)s?)?
1312 if strict
and not re
.fullmatch(r
'\d{8}|(now|today|yesterday)(-\d+(day|week|month|year)s?)?', date_str
):
1313 raise ValueError(f
'Invalid date format "{date_str}"')
1314 return datetime_from_str(date_str
, precision
='microsecond', format
=format
).date()
1317 def datetime_add_months(dt
, months
):
1318 """Increment/Decrement a datetime object by months."""
1319 month
= dt
.month
+ months
- 1
1320 year
= dt
.year
+ month
// 12
1321 month
= month
% 12 + 1
1322 day
= min(dt
.day
, calendar
.monthrange(year
, month
)[1])
1323 return dt
.replace(year
, month
, day
)
1326 def datetime_round(dt
, precision
='day'):
1328 Round a datetime object's time to a specific precision
1330 if precision
== 'microsecond':
1339 roundto
= lambda x
, n
: ((x
+ n
/ 2) // n
) * n
1340 timestamp
= roundto(calendar
.timegm(dt
.timetuple()), unit_seconds
[precision
])
1341 return datetime
.datetime
.fromtimestamp(timestamp
, datetime
.timezone
.utc
)
1344 def hyphenate_date(date_str
):
1346 Convert a date in 'YYYYMMDD' format to 'YYYY-MM-DD' format"""
1347 match
= re
.match(r
'^(\d\d\d\d)(\d\d)(\d\d)$', date_str
)
1348 if match
is not None:
1349 return '-'.join(match
.groups())
1355 """Represents a time interval between two dates"""
1357 def __init__(self
, start
=None, end
=None):
1358 """start and end must be strings in the format accepted by date"""
1359 if start
is not None:
1360 self
.start
= date_from_str(start
, strict
=True)
1362 self
.start
= datetime
.datetime
.min.date()
1364 self
.end
= date_from_str(end
, strict
=True)
1366 self
.end
= datetime
.datetime
.max.date()
1367 if self
.start
> self
.end
:
1368 raise ValueError('Date range: "%s" , the start date must be before the end date' % self
)
1372 """Returns a range that only contains the given day"""
1373 return cls(day
, day
)
1375 def __contains__(self
, date
):
1376 """Check if the date is in the range"""
1377 if not isinstance(date
, datetime
.date
):
1378 date
= date_from_str(date
)
1379 return self
.start
<= date
<= self
.end
1382 return f
'{__name__}.{type(self).__name__}({self.start.isoformat()!r}, {self.end.isoformat()!r})'
1384 def __eq__(self
, other
):
1385 return (isinstance(other
, DateRange
)
1386 and self
.start
== other
.start
and self
.end
== other
.end
)
1390 def system_identifier():
1391 python_implementation
= platform
.python_implementation()
1392 if python_implementation
== 'PyPy' and hasattr(sys
, 'pypy_version_info'):
1393 python_implementation
+= ' version %d.%d.%d' % sys
.pypy_version_info
[:3]
1395 with contextlib
.suppress(OSError): # We may not have access to the executable
1396 libc_ver
= platform
.libc_ver()
1398 return 'Python %s (%s %s %s) - %s (%s%s)' % (
1399 platform
.python_version(),
1400 python_implementation
,
1402 platform
.architecture()[0],
1403 platform
.platform(),
1404 ssl
.OPENSSL_VERSION
,
1405 format_field(join_nonempty(*libc_ver
, delim
=' '), None, ', %s'),
1410 def get_windows_version():
1411 ''' Get Windows version. returns () if it's not running on Windows '''
1412 if compat_os_name
== 'nt':
1413 return version_tuple(platform
.win32_ver()[1])
1418 def write_string(s
, out
=None, encoding
=None):
1419 assert isinstance(s
, str)
1420 out
= out
or sys
.stderr
1421 # `sys.stderr` might be `None` (Ref: https://github.com/pyinstaller/pyinstaller/pull/7217)
1425 if compat_os_name
== 'nt' and supports_terminal_sequences(out
):
1426 s
= re
.sub(r
'([\r\n]+)', r
' \1', s
)
1428 enc
, buffer = None, out
1429 if 'b' in getattr(out
, 'mode', ''):
1430 enc
= encoding
or preferredencoding()
1431 elif hasattr(out
, 'buffer'):
1433 enc
= encoding
or getattr(out
, 'encoding', None) or preferredencoding()
1435 buffer.write(s
.encode(enc
, 'ignore') if enc
else s
)
1439 # TODO: Use global logger
1440 def deprecation_warning(msg
, *, printer
=None, stacklevel
=0, **kwargs
):
1441 from .. import _IN_CLI
1443 if msg
in deprecation_warning
._cache
:
1445 deprecation_warning
._cache
.add(msg
)
1447 return printer(f
'{msg}{bug_reports_message()}', **kwargs
)
1448 return write_string(f
'ERROR: {msg}{bug_reports_message()}\n', **kwargs
)
1451 warnings
.warn(DeprecationWarning(msg
), stacklevel
=stacklevel
+ 3)
1454 deprecation_warning
._cache
= set()
1457 def bytes_to_intlist(bs
):
1460 if isinstance(bs
[0], int): # Python 3
1463 return [ord(c
) for c
in bs
]
1466 def intlist_to_bytes(xs
):
1469 return struct
.pack('%dB' % len(xs
), *xs
)
1472 class LockingUnsupportedError(OSError):
1473 msg
= 'File locking is not supported'
1476 super().__init
__(self
.msg
)
1479 # Cross-platform file locking
1480 if sys
.platform
== 'win32':
1482 import ctypes
.wintypes
1485 class OVERLAPPED(ctypes
.Structure
):
1487 ('Internal', ctypes
.wintypes
.LPVOID
),
1488 ('InternalHigh', ctypes
.wintypes
.LPVOID
),
1489 ('Offset', ctypes
.wintypes
.DWORD
),
1490 ('OffsetHigh', ctypes
.wintypes
.DWORD
),
1491 ('hEvent', ctypes
.wintypes
.HANDLE
),
1494 kernel32
= ctypes
.WinDLL('kernel32')
1495 LockFileEx
= kernel32
.LockFileEx
1496 LockFileEx
.argtypes
= [
1497 ctypes
.wintypes
.HANDLE
, # hFile
1498 ctypes
.wintypes
.DWORD
, # dwFlags
1499 ctypes
.wintypes
.DWORD
, # dwReserved
1500 ctypes
.wintypes
.DWORD
, # nNumberOfBytesToLockLow
1501 ctypes
.wintypes
.DWORD
, # nNumberOfBytesToLockHigh
1502 ctypes
.POINTER(OVERLAPPED
) # Overlapped
1504 LockFileEx
.restype
= ctypes
.wintypes
.BOOL
1505 UnlockFileEx
= kernel32
.UnlockFileEx
1506 UnlockFileEx
.argtypes
= [
1507 ctypes
.wintypes
.HANDLE
, # hFile
1508 ctypes
.wintypes
.DWORD
, # dwReserved
1509 ctypes
.wintypes
.DWORD
, # nNumberOfBytesToLockLow
1510 ctypes
.wintypes
.DWORD
, # nNumberOfBytesToLockHigh
1511 ctypes
.POINTER(OVERLAPPED
) # Overlapped
1513 UnlockFileEx
.restype
= ctypes
.wintypes
.BOOL
1514 whole_low
= 0xffffffff
1515 whole_high
= 0x7fffffff
1517 def _lock_file(f
, exclusive
, block
):
1518 overlapped
= OVERLAPPED()
1519 overlapped
.Offset
= 0
1520 overlapped
.OffsetHigh
= 0
1521 overlapped
.hEvent
= 0
1522 f
._lock
_file
_overlapped
_p
= ctypes
.pointer(overlapped
)
1524 if not LockFileEx(msvcrt
.get_osfhandle(f
.fileno()),
1525 (0x2 if exclusive
else 0x0) |
(0x0 if block
else 0x1),
1526 0, whole_low
, whole_high
, f
._lock
_file
_overlapped
_p
):
1527 # NB: No argument form of "ctypes.FormatError" does not work on PyPy
1528 raise BlockingIOError(f
'Locking file failed: {ctypes.FormatError(ctypes.GetLastError())!r}')
1530 def _unlock_file(f
):
1531 assert f
._lock
_file
_overlapped
_p
1532 handle
= msvcrt
.get_osfhandle(f
.fileno())
1533 if not UnlockFileEx(handle
, 0, whole_low
, whole_high
, f
._lock
_file
_overlapped
_p
):
1534 raise OSError('Unlocking file failed: %r' % ctypes
.FormatError())
1540 def _lock_file(f
, exclusive
, block
):
1541 flags
= fcntl
.LOCK_EX
if exclusive
else fcntl
.LOCK_SH
1543 flags |
= fcntl
.LOCK_NB
1545 fcntl
.flock(f
, flags
)
1546 except BlockingIOError
:
1548 except OSError: # AOSP does not have flock()
1549 fcntl
.lockf(f
, flags
)
1551 def _unlock_file(f
):
1552 with contextlib
.suppress(OSError):
1553 return fcntl
.flock(f
, fcntl
.LOCK_UN
)
1554 with contextlib
.suppress(OSError):
1555 return fcntl
.lockf(f
, fcntl
.LOCK_UN
) # AOSP does not have flock()
1556 return fcntl
.flock(f
, fcntl
.LOCK_UN | fcntl
.LOCK_NB
) # virtiofs needs LOCK_NB on unlocking
1560 def _lock_file(f
, exclusive
, block
):
1561 raise LockingUnsupportedError()
1563 def _unlock_file(f
):
1564 raise LockingUnsupportedError()
1570 def __init__(self
, filename
, mode
, block
=True, encoding
=None):
1571 if mode
not in {'r', 'rb', 'a', 'ab', 'w', 'wb'}
:
1572 raise NotImplementedError(mode
)
1573 self
.mode
, self
.block
= mode
, block
1575 writable
= any(f
in mode
for f
in 'wax+')
1576 readable
= any(f
in mode
for f
in 'r+')
1577 flags
= functools
.reduce(operator
.ior
, (
1578 getattr(os
, 'O_CLOEXEC', 0), # UNIX only
1579 getattr(os
, 'O_BINARY', 0), # Windows only
1580 getattr(os
, 'O_NOINHERIT', 0), # Windows only
1581 os
.O_CREAT
if writable
else 0, # O_TRUNC only after locking
1582 os
.O_APPEND
if 'a' in mode
else 0,
1583 os
.O_EXCL
if 'x' in mode
else 0,
1584 os
.O_RDONLY
if not writable
else os
.O_RDWR
if readable
else os
.O_WRONLY
,
1587 self
.f
= os
.fdopen(os
.open(filename
, flags
, 0o666), mode
, encoding
=encoding
)
1589 def __enter__(self
):
1590 exclusive
= 'r' not in self
.mode
1592 _lock_file(self
.f
, exclusive
, self
.block
)
1597 if 'w' in self
.mode
:
1600 except OSError as e
:
1602 errno
.ESPIPE
, # Illegal seek - expected for FIFO
1603 errno
.EINVAL
, # Invalid argument - expected for /dev/null
1612 _unlock_file(self
.f
)
1616 def __exit__(self
, *_
):
1625 def __getattr__(self
, attr
):
1626 return getattr(self
.f
, attr
)
1633 def get_filesystem_encoding():
1634 encoding
= sys
.getfilesystemencoding()
1635 return encoding
if encoding
is not None else 'utf-8'
1638 def shell_quote(args
):
1640 encoding
= get_filesystem_encoding()
1642 if isinstance(a
, bytes):
1643 # We may get a filename encoded with 'encodeFilename'
1644 a
= a
.decode(encoding
)
1645 quoted_args
.append(compat_shlex_quote(a
))
1646 return ' '.join(quoted_args
)
1649 def smuggle_url(url
, data
):
1650 """ Pass additional data in a URL for internal use. """
1652 url
, idata
= unsmuggle_url(url
, {})
1654 sdata
= urllib
.parse
.urlencode(
1655 {'__youtubedl_smuggle': json.dumps(data)}
)
1656 return url
+ '#' + sdata
1659 def unsmuggle_url(smug_url
, default
=None):
1660 if '#__youtubedl_smuggle' not in smug_url
:
1661 return smug_url
, default
1662 url
, _
, sdata
= smug_url
.rpartition('#')
1663 jsond
= urllib
.parse
.parse_qs(sdata
)['__youtubedl_smuggle'][0]
1664 data
= json
.loads(jsond
)
1668 def format_decimal_suffix(num
, fmt
='%d%s', *, factor
=1000):
1669 """ Formats numbers with decimal sufixes like K, M, etc """
1670 num
, factor
= float_or_none(num
), float(factor
)
1671 if num
is None or num
< 0:
1673 POSSIBLE_SUFFIXES
= 'kMGTPEZY'
1674 exponent
= 0 if num
== 0 else min(int(math
.log(num
, factor
)), len(POSSIBLE_SUFFIXES
))
1675 suffix
= ['', *POSSIBLE_SUFFIXES
][exponent
]
1677 suffix
= {'k': 'Ki', '': ''}
.get(suffix
, f
'{suffix}i')
1678 converted
= num
/ (factor
** exponent
)
1679 return fmt
% (converted
, suffix
)
1682 def format_bytes(bytes):
1683 return format_decimal_suffix(bytes, '%.2f%sB', factor
=1024) or 'N/A'
1686 def lookup_unit_table(unit_table
, s
, strict
=False):
1687 num_re
= NUMBER_RE
if strict
else NUMBER_RE
.replace(R
'\.', '[,.]')
1688 units_re
= '|'.join(re
.escape(u
) for u
in unit_table
)
1689 m
= (re
.fullmatch
if strict
else re
.match
)(
1690 rf
'(?P<num>{num_re})\s*(?P<unit>{units_re})\b', s
)
1694 num
= float(m
.group('num').replace(',', '.'))
1695 mult
= unit_table
[m
.group('unit')]
1696 return round(num
* mult
)
1700 """Parse a string indicating a byte quantity into an integer"""
1701 return lookup_unit_table(
1702 {u: 1024**i for i, u in enumerate(['', *'KMGTPEZY'])}
,
1703 s
.upper(), strict
=True)
1706 def parse_filesize(s
):
1710 # The lower-case forms are of course incorrect and unofficial,
1711 # but we support those too
1728 'megabytes': 1000 ** 2,
1729 'mebibytes': 1024 ** 2,
1735 'gigabytes': 1000 ** 3,
1736 'gibibytes': 1024 ** 3,
1742 'terabytes': 1000 ** 4,
1743 'tebibytes': 1024 ** 4,
1749 'petabytes': 1000 ** 5,
1750 'pebibytes': 1024 ** 5,
1756 'exabytes': 1000 ** 6,
1757 'exbibytes': 1024 ** 6,
1763 'zettabytes': 1000 ** 7,
1764 'zebibytes': 1024 ** 7,
1770 'yottabytes': 1000 ** 8,
1771 'yobibytes': 1024 ** 8,
1774 return lookup_unit_table(_UNIT_TABLE
, s
)
1781 s
= re
.sub(r
'^[^\d]+\s', '', s
).strip()
1783 if re
.match(r
'^[\d,.]+$', s
):
1784 return str_to_int(s
)
1797 ret
= lookup_unit_table(_UNIT_TABLE
, s
)
1801 mobj
= re
.match(r
'([\d,.]+)(?:$|\s)', s
)
1803 return str_to_int(mobj
.group(1))
1806 def parse_resolution(s
, *, lenient
=False):
1811 mobj
= re
.search(r
'(?P<w>\d+)\s*[xX×,]\s*(?P<h>\d+)', s
)
1813 mobj
= re
.search(r
'(?<![a-zA-Z0-9])(?P<w>\d+)\s*[xX×,]\s*(?P<h>\d+)(?![a-zA-Z0-9])', s
)
1816 'width': int(mobj
.group('w')),
1817 'height': int(mobj
.group('h')),
1820 mobj
= re
.search(r
'(?<![a-zA-Z0-9])(\d+)[pPiI](?![a-zA-Z0-9])', s
)
1822 return {'height': int(mobj.group(1))}
1824 mobj
= re
.search(r
'\b([48])[kK]\b', s
)
1826 return {'height': int(mobj.group(1)) * 540}
1831 def parse_bitrate(s
):
1832 if not isinstance(s
, str):
1834 mobj
= re
.search(r
'\b(\d+)\s*kbps', s
)
1836 return int(mobj
.group(1))
1839 def month_by_name(name
, lang
='en'):
1840 """ Return the number of a month by (locale-independently) English name """
1842 month_names
= MONTH_NAMES
.get(lang
, MONTH_NAMES
['en'])
1845 return month_names
.index(name
) + 1
1850 def month_by_abbreviation(abbrev
):
1851 """ Return the number of a month by (locale-independently) English
1855 return [s
[:3] for s
in ENGLISH_MONTH_NAMES
].index(abbrev
) + 1
1860 def fix_xml_ampersands(xml_str
):
1861 """Replace all the '&' by '&' in XML"""
1863 r
'&(?!amp;|lt;|gt;|apos;|quot;|#x[0-9a-fA-F]{,4};|#[0-9]{,4};)',
1868 def setproctitle(title
):
1869 assert isinstance(title
, str)
1871 # Workaround for https://github.com/yt-dlp/yt-dlp/issues/4541
1878 libc
= ctypes
.cdll
.LoadLibrary('libc.so.6')
1882 # LoadLibrary in Windows Python 2.7.13 only expects
1883 # a bytestring, but since unicode_literals turns
1884 # every string into a unicode string, it fails.
1886 title_bytes
= title
.encode()
1887 buf
= ctypes
.create_string_buffer(len(title_bytes
))
1888 buf
.value
= title_bytes
1890 libc
.prctl(15, buf
, 0, 0, 0)
1891 except AttributeError:
1892 return # Strange libc, just skip this
1895 def remove_start(s
, start
):
1896 return s
[len(start
):] if s
is not None and s
.startswith(start
) else s
1899 def remove_end(s
, end
):
1900 return s
[:-len(end
)] if s
is not None and s
.endswith(end
) else s
1903 def remove_quotes(s
):
1904 if s
is None or len(s
) < 2:
1906 for quote
in ('"', "'", ):
1907 if s
[0] == quote
and s
[-1] == quote
:
1912 def get_domain(url
):
1914 This implementation is inconsistent, but is kept for compatibility.
1915 Use this only for "webpage_url_domain"
1917 return remove_start(urllib
.parse
.urlparse(url
).netloc
, 'www.') or None
1920 def url_basename(url
):
1921 path
= urllib
.parse
.urlparse(url
).path
1922 return path
.strip('/').split('/')[-1]
1926 return re
.match(r
'https?://[^?#]+/', url
).group()
1929 def urljoin(base
, path
):
1930 if isinstance(path
, bytes):
1931 path
= path
.decode()
1932 if not isinstance(path
, str) or not path
:
1934 if re
.match(r
'^(?:[a-zA-Z][a-zA-Z0-9+-.]*:)?//', path
):
1936 if isinstance(base
, bytes):
1937 base
= base
.decode()
1938 if not isinstance(base
, str) or not re
.match(
1939 r
'^(?:https?:)?//', base
):
1941 return urllib
.parse
.urljoin(base
, path
)
1944 def int_or_none(v
, scale
=1, default
=None, get_attr
=None, invscale
=1):
1945 if get_attr
and v
is not None:
1946 v
= getattr(v
, get_attr
, None)
1948 return int(v
) * invscale
// scale
1949 except (ValueError, TypeError, OverflowError):
1953 def str_or_none(v
, default
=None):
1954 return default
if v
is None else str(v
)
1957 def str_to_int(int_str
):
1958 """ A more relaxed version of int_or_none """
1959 if isinstance(int_str
, int):
1961 elif isinstance(int_str
, str):
1962 int_str
= re
.sub(r
'[,\.\+]', '', int_str
)
1963 return int_or_none(int_str
)
1966 def float_or_none(v
, scale
=1, invscale
=1, default
=None):
1970 return float(v
) * invscale
/ scale
1971 except (ValueError, TypeError):
1975 def bool_or_none(v
, default
=None):
1976 return v
if isinstance(v
, bool) else default
1979 def strip_or_none(v
, default
=None):
1980 return v
.strip() if isinstance(v
, str) else default
1983 def url_or_none(url
):
1984 if not url
or not isinstance(url
, str):
1987 return url
if re
.match(r
'^(?:(?:https?|rt(?:m(?:pt?[es]?|fp)|sp[su]?)|mms|ftps?):)?//', url
) else None
1990 def strftime_or_none(timestamp
, date_format
='%Y%m%d', default
=None):
1991 datetime_object
= None
1993 if isinstance(timestamp
, (int, float)): # unix timestamp
1994 # Using naive datetime here can break timestamp() in Windows
1995 # Ref: https://github.com/yt-dlp/yt-dlp/issues/5185, https://github.com/python/cpython/issues/94414
1996 # Also, datetime.datetime.fromtimestamp breaks for negative timestamps
1997 # Ref: https://github.com/yt-dlp/yt-dlp/issues/6706#issuecomment-1496842642
1998 datetime_object
= (datetime
.datetime
.fromtimestamp(0, datetime
.timezone
.utc
)
1999 + datetime
.timedelta(seconds
=timestamp
))
2000 elif isinstance(timestamp
, str): # assume YYYYMMDD
2001 datetime_object
= datetime
.datetime
.strptime(timestamp
, '%Y%m%d')
2002 date_format
= re
.sub( # Support %s on windows
2003 r
'(?<!%)(%%)*%s', rf
'\g<1>{int(datetime_object.timestamp())}', date_format
)
2004 return datetime_object
.strftime(date_format
)
2005 except (ValueError, TypeError, AttributeError):
2009 def parse_duration(s
):
2010 if not isinstance(s
, str):
2016 days
, hours
, mins
, secs
, ms
= [None] * 5
2017 m
= re
.match(r
'''(?x)
2019 (?:(?:(?P<days>[0-9]+):)?(?P<hours>[0-9]+):)?(?P<mins>[0-9]+):)?
2020 (?P<secs>(?(before_secs)[0-9]{1,2}|[0-9]+))
2021 (?P<ms>[.:][0-9]+)?Z?$
2024 days
, hours
, mins
, secs
, ms
= m
.group('days', 'hours', 'mins', 'secs', 'ms')
2029 [0-9]+\s*y(?:ears?)?,?\s*
2032 [0-9]+\s*m(?:onths?)?,?\s*
2035 [0-9]+\s*w(?:eeks?)?,?\s*
2038 (?P<days>[0-9]+)\s*d(?:ays?)?,?\s*
2042 (?P<hours>[0-9]+)\s*h(?:(?:ou)?rs?)?,?\s*
2045 (?P<mins>[0-9]+)\s*m(?:in(?:ute)?s?)?,?\s*
2048 (?P<secs>[0-9]+)(?P<ms>\.[0-9]+)?\s*s(?:ec(?:ond)?s?)?\s*
2051 days
, hours
, mins
, secs
, ms
= m
.groups()
2053 m
= re
.match(r
'(?i)(?:(?P<hours>[0-9.]+)\s*(?:hours?)|(?P<mins>[0-9.]+)\s*(?:mins?\.?|minutes?)\s*)Z?$', s
)
2055 hours
, mins
= m
.groups()
2060 ms
= ms
.replace(':', '.')
2061 return sum(float(part
or 0) * mult
for part
, mult
in (
2062 (days
, 86400), (hours
, 3600), (mins
, 60), (secs
, 1), (ms
, 1)))
2065 def prepend_extension(filename
, ext
, expected_real_ext
=None):
2066 name
, real_ext
= os
.path
.splitext(filename
)
2068 f
'{name}.{ext}{real_ext}'
2069 if not expected_real_ext
or real_ext
[1:] == expected_real_ext
2070 else f
'{filename}.{ext}')
2073 def replace_extension(filename
, ext
, expected_real_ext
=None):
2074 name
, real_ext
= os
.path
.splitext(filename
)
2075 return '{}.{}'.format(
2076 name
if not expected_real_ext
or real_ext
[1:] == expected_real_ext
else filename
,
2080 def check_executable(exe
, args
=[]):
2081 """ Checks if the given binary is installed somewhere in PATH, and returns its name.
2082 args can be a list of arguments for a short output (like -version) """
2084 Popen
.run([exe
] + args
, stdout
=subprocess
.PIPE
, stderr
=subprocess
.PIPE
)
2090 def _get_exe_version_output(exe
, args
):
2092 # STDIN should be redirected too. On UNIX-like systems, ffmpeg triggers
2093 # SIGTTOU if yt-dlp is run in the background.
2094 # See https://github.com/ytdl-org/youtube-dl/issues/955#issuecomment-209789656
2095 stdout
, _
, ret
= Popen
.run([encodeArgument(exe
)] + args
, text
=True,
2096 stdin
=subprocess
.PIPE
, stdout
=subprocess
.PIPE
, stderr
=subprocess
.STDOUT
)
2104 def detect_exe_version(output
, version_re
=None, unrecognized
='present'):
2105 assert isinstance(output
, str)
2106 if version_re
is None:
2107 version_re
= r
'version\s+([-0-9._a-zA-Z]+)'
2108 m
= re
.search(version_re
, output
)
2115 def get_exe_version(exe
, args
=['--version'],
2116 version_re
=None, unrecognized
=('present', 'broken')):
2117 """ Returns the version of the specified executable,
2118 or False if the executable is not present """
2119 unrecognized
= variadic(unrecognized
)
2120 assert len(unrecognized
) in (1, 2)
2121 out
= _get_exe_version_output(exe
, args
)
2123 return unrecognized
[-1]
2124 return out
and detect_exe_version(out
, version_re
, unrecognized
[0])
2127 def frange(start
=0, stop
=None, step
=1):
2130 start
, stop
= 0, start
2131 sign
= [-1, 1][step
> 0] if step
else 0
2132 while sign
* start
< sign
* stop
:
2137 class LazyList(collections
.abc
.Sequence
):
2138 """Lazy immutable list from an iterable
2139 Note that slices of a LazyList are lists and not LazyList"""
2141 class IndexError(IndexError):
2144 def __init__(self
, iterable
, *, reverse
=False, _cache
=None):
2145 self
._iterable
= iter(iterable
)
2146 self
._cache
= [] if _cache
is None else _cache
2147 self
._reversed
= reverse
2151 # We need to consume the entire iterable to iterate in reverse
2152 yield from self
.exhaust()
2154 yield from self
._cache
2155 for item
in self
._iterable
:
2156 self
._cache
.append(item
)
2160 self
._cache
.extend(self
._iterable
)
2161 self
._iterable
= [] # Discard the emptied iterable to make it pickle-able
2165 """Evaluate the entire iterable"""
2166 return self
._exhaust
()[::-1 if self
._reversed
else 1]
2169 def _reverse_index(x
):
2170 return None if x
is None else ~x
2172 def __getitem__(self
, idx
):
2173 if isinstance(idx
, slice):
2175 idx
= slice(self
._reverse
_index
(idx
.start
), self
._reverse
_index
(idx
.stop
), -(idx
.step
or 1))
2176 start
, stop
, step
= idx
.start
, idx
.stop
, idx
.step
or 1
2177 elif isinstance(idx
, int):
2179 idx
= self
._reverse
_index
(idx
)
2180 start
, stop
, step
= idx
, idx
, 0
2182 raise TypeError('indices must be integers or slices')
2183 if ((start
or 0) < 0 or (stop
or 0) < 0
2184 or (start
is None and step
< 0)
2185 or (stop
is None and step
> 0)):
2186 # We need to consume the entire iterable to be able to slice from the end
2187 # Obviously, never use this with infinite iterables
2190 return self
._cache
[idx
]
2191 except IndexError as e
:
2192 raise self
.IndexError(e
) from e
2193 n
= max(start
or 0, stop
or 0) - len(self
._cache
) + 1
2195 self
._cache
.extend(itertools
.islice(self
._iterable
, n
))
2197 return self
._cache
[idx
]
2198 except IndexError as e
:
2199 raise self
.IndexError(e
) from e
2203 self
[-1] if self
._reversed
else self
[0]
2204 except self
.IndexError:
2210 return len(self
._cache
)
2212 def __reversed__(self
):
2213 return type(self
)(self
._iterable
, reverse
=not self
._reversed
, _cache
=self
._cache
)
2216 return type(self
)(self
._iterable
, reverse
=self
._reversed
, _cache
=self
._cache
)
2219 # repr and str should mimic a list. So we exhaust the iterable
2220 return repr(self
.exhaust())
2223 return repr(self
.exhaust())
2228 class IndexError(IndexError):
2232 # This is only useful for tests
2233 return len(self
.getslice())
2235 def __init__(self
, pagefunc
, pagesize
, use_cache
=True):
2236 self
._pagefunc
= pagefunc
2237 self
._pagesize
= pagesize
2238 self
._pagecount
= float('inf')
2239 self
._use
_cache
= use_cache
2242 def getpage(self
, pagenum
):
2243 page_results
= self
._cache
.get(pagenum
)
2244 if page_results
is None:
2245 page_results
= [] if pagenum
> self
._pagecount
else list(self
._pagefunc
(pagenum
))
2247 self
._cache
[pagenum
] = page_results
2250 def getslice(self
, start
=0, end
=None):
2251 return list(self
._getslice
(start
, end
))
2253 def _getslice(self
, start
, end
):
2254 raise NotImplementedError('This method must be implemented by subclasses')
2256 def __getitem__(self
, idx
):
2257 assert self
._use
_cache
, 'Indexing PagedList requires cache'
2258 if not isinstance(idx
, int) or idx
< 0:
2259 raise TypeError('indices must be non-negative integers')
2260 entries
= self
.getslice(idx
, idx
+ 1)
2262 raise self
.IndexError()
2266 class OnDemandPagedList(PagedList
):
2267 """Download pages until a page with less than maximum results"""
2269 def _getslice(self
, start
, end
):
2270 for pagenum
in itertools
.count(start
// self
._pagesize
):
2271 firstid
= pagenum
* self
._pagesize
2272 nextfirstid
= pagenum
* self
._pagesize
+ self
._pagesize
2273 if start
>= nextfirstid
:
2277 start
% self
._pagesize
2278 if firstid
<= start
< nextfirstid
2281 ((end
- 1) % self
._pagesize
) + 1
2282 if (end
is not None and firstid
<= end
<= nextfirstid
)
2286 page_results
= self
.getpage(pagenum
)
2288 self
._pagecount
= pagenum
- 1
2290 if startv
!= 0 or endv
is not None:
2291 page_results
= page_results
[startv
:endv
]
2292 yield from page_results
2294 # A little optimization - if current page is not "full", ie. does
2295 # not contain page_size videos then we can assume that this page
2296 # is the last one - there are no more ids on further pages -
2297 # i.e. no need to query again.
2298 if len(page_results
) + startv
< self
._pagesize
:
2301 # If we got the whole page, but the next page is not interesting,
2302 # break out early as well
2303 if end
== nextfirstid
:
2307 class InAdvancePagedList(PagedList
):
2308 """PagedList with total number of pages known in advance"""
2310 def __init__(self
, pagefunc
, pagecount
, pagesize
):
2311 PagedList
.__init
__(self
, pagefunc
, pagesize
, True)
2312 self
._pagecount
= pagecount
2314 def _getslice(self
, start
, end
):
2315 start_page
= start
// self
._pagesize
2316 end_page
= self
._pagecount
if end
is None else min(self
._pagecount
, end
// self
._pagesize
+ 1)
2317 skip_elems
= start
- start_page
* self
._pagesize
2318 only_more
= None if end
is None else end
- start
2319 for pagenum
in range(start_page
, end_page
):
2320 page_results
= self
.getpage(pagenum
)
2322 page_results
= page_results
[skip_elems
:]
2324 if only_more
is not None:
2325 if len(page_results
) < only_more
:
2326 only_more
-= len(page_results
)
2328 yield from page_results
[:only_more
]
2330 yield from page_results
2333 class PlaylistEntries
:
2334 MissingEntry
= object()
2335 is_exhausted
= False
2337 def __init__(self
, ydl
, info_dict
):
2340 # _entries must be assigned now since infodict can change during iteration
2341 entries
= info_dict
.get('entries')
2343 raise EntryNotInPlaylist('There are no entries')
2344 elif isinstance(entries
, list):
2345 self
.is_exhausted
= True
2347 requested_entries
= info_dict
.get('requested_entries')
2348 self
.is_incomplete
= requested_entries
is not None
2349 if self
.is_incomplete
:
2350 assert self
.is_exhausted
2351 self
._entries
= [self
.MissingEntry
] * max(requested_entries
or [0])
2352 for i
, entry
in zip(requested_entries
, entries
):
2353 self
._entries
[i
- 1] = entry
2354 elif isinstance(entries
, (list, PagedList
, LazyList
)):
2355 self
._entries
= entries
2357 self
._entries
= LazyList(entries
)
2359 PLAYLIST_ITEMS_RE
= re
.compile(r
'''(?x)
2360 (?P<start>[+-]?\d+)?
2362 (?P<end>[+-]?\d+|inf(?:inite)?)?
2363 (?::(?P<step>[+-]?\d+))?
2367 def parse_playlist_items(cls
, string
):
2368 for segment
in string
.split(','):
2370 raise ValueError('There is two or more consecutive commas')
2371 mobj
= cls
.PLAYLIST_ITEMS_RE
.fullmatch(segment
)
2373 raise ValueError(f
'{segment!r} is not a valid specification')
2374 start
, end
, step
, has_range
= mobj
.group('start', 'end', 'step', 'range')
2375 if int_or_none(step
) == 0:
2376 raise ValueError(f
'Step in {segment!r} cannot be zero')
2377 yield slice(int_or_none(start
), float_or_none(end
), int_or_none(step
)) if has_range
else int(start
)
2379 def get_requested_items(self
):
2380 playlist_items
= self
.ydl
.params
.get('playlist_items')
2381 playlist_start
= self
.ydl
.params
.get('playliststart', 1)
2382 playlist_end
= self
.ydl
.params
.get('playlistend')
2383 # For backwards compatibility, interpret -1 as whole list
2384 if playlist_end
in (-1, None):
2386 if not playlist_items
:
2387 playlist_items
= f
'{playlist_start}:{playlist_end}'
2388 elif playlist_start
!= 1 or playlist_end
:
2389 self
.ydl
.report_warning('Ignoring playliststart and playlistend because playlistitems was given', only_once
=True)
2391 for index
in self
.parse_playlist_items(playlist_items
):
2392 for i
, entry
in self
[index
]:
2397 # The item may have just been added to archive. Don't break due to it
2398 if not self
.ydl
.params
.get('lazy_playlist'):
2399 # TODO: Add auto-generated fields
2400 self
.ydl
._match
_entry
(entry
, incomplete
=True, silent
=True)
2401 except (ExistingVideoReached
, RejectedVideoReached
):
2404 def get_full_count(self
):
2405 if self
.is_exhausted
and not self
.is_incomplete
:
2407 elif isinstance(self
._entries
, InAdvancePagedList
):
2408 if self
._entries
._pagesize
== 1:
2409 return self
._entries
._pagecount
2411 @functools.cached_property
2413 if isinstance(self
._entries
, list):
2416 entry
= self
._entries
[i
]
2418 entry
= self
.MissingEntry
2419 if not self
.is_incomplete
:
2420 raise self
.IndexError()
2421 if entry
is self
.MissingEntry
:
2422 raise EntryNotInPlaylist(f
'Entry {i + 1} cannot be found')
2427 return type(self
.ydl
)._handle
_extraction
_exceptions
(lambda _
, i
: self
._entries
[i
])(self
.ydl
, i
)
2428 except (LazyList
.IndexError, PagedList
.IndexError):
2429 raise self
.IndexError()
2432 def __getitem__(self
, idx
):
2433 if isinstance(idx
, int):
2434 idx
= slice(idx
, idx
)
2436 # NB: PlaylistEntries[1:10] => (0, 1, ... 9)
2437 step
= 1 if idx
.step
is None else idx
.step
2438 if idx
.start
is None:
2439 start
= 0 if step
> 0 else len(self
) - 1
2441 start
= idx
.start
- 1 if idx
.start
>= 0 else len(self
) + idx
.start
2443 # NB: Do not call len(self) when idx == [:]
2444 if idx
.stop
is None:
2445 stop
= 0 if step
< 0 else float('inf')
2447 stop
= idx
.stop
- 1 if idx
.stop
>= 0 else len(self
) + idx
.stop
2448 stop
+= [-1, 1][step
> 0]
2450 for i
in frange(start
, stop
, step
):
2454 entry
= self
._getter
(i
)
2455 except self
.IndexError:
2456 self
.is_exhausted
= True
2463 return len(tuple(self
[:]))
2465 class IndexError(IndexError):
2469 def uppercase_escape(s
):
2470 unicode_escape
= codecs
.getdecoder('unicode_escape')
2472 r
'\\U[0-9a-fA-F]{8}',
2473 lambda m
: unicode_escape(m
.group(0))[0],
2477 def lowercase_escape(s
):
2478 unicode_escape
= codecs
.getdecoder('unicode_escape')
2480 r
'\\u[0-9a-fA-F]{4}',
2481 lambda m
: unicode_escape(m
.group(0))[0],
2485 def parse_qs(url
, **kwargs
):
2486 return urllib
.parse
.parse_qs(urllib
.parse
.urlparse(url
).query
, **kwargs
)
2489 def read_batch_urls(batch_fd
):
2491 if not isinstance(url
, str):
2492 url
= url
.decode('utf-8', 'replace')
2493 BOM_UTF8
= ('\xef\xbb\xbf', '\ufeff')
2494 for bom
in BOM_UTF8
:
2495 if url
.startswith(bom
):
2496 url
= url
[len(bom
):]
2498 if not url
or url
.startswith(('#', ';', ']')):
2500 # "#" cannot be stripped out since it is part of the URI
2501 # However, it can be safely stripped out if following a whitespace
2502 return re
.split(r
'\s#', url
, 1)[0].rstrip()
2504 with contextlib
.closing(batch_fd
) as fd
:
2505 return [url
for url
in map(fixup
, fd
) if url
]
2508 def urlencode_postdata(*args
, **kargs
):
2509 return urllib
.parse
.urlencode(*args
, **kargs
).encode('ascii')
2512 def update_url(url
, *, query_update
=None, **kwargs
):
2513 """Replace URL components specified by kwargs
2514 @param url str or parse url tuple
2515 @param query_update update query
2518 if isinstance(url
, str):
2519 if not kwargs
and not query_update
:
2522 url
= urllib
.parse
.urlparse(url
)
2524 assert 'query' not in kwargs
, 'query_update and query cannot be specified at the same time'
2525 kwargs
['query'] = urllib
.parse
.urlencode({
2526 **urllib
.parse
.parse_qs(url
.query
),
2529 return urllib
.parse
.urlunparse(url
._replace
(**kwargs
))
2532 def update_url_query(url
, query
):
2533 return update_url(url
, query_update
=query
)
2536 def _multipart_encode_impl(data
, boundary
):
2537 content_type
= 'multipart/form-data; boundary=%s' % boundary
2540 for k
, v
in data
.items():
2541 out
+= b
'--' + boundary
.encode('ascii') + b
'\r\n'
2542 if isinstance(k
, str):
2544 if isinstance(v
, str):
2546 # RFC 2047 requires non-ASCII field names to be encoded, while RFC 7578
2547 # suggests sending UTF-8 directly. Firefox sends UTF-8, too
2548 content
= b
'Content-Disposition: form-data; name="' + k
+ b
'"\r\n\r\n' + v
+ b
'\r\n'
2549 if boundary
.encode('ascii') in content
:
2550 raise ValueError('Boundary overlaps with data')
2553 out
+= b
'--' + boundary
.encode('ascii') + b
'--\r\n'
2555 return out
, content_type
2558 def multipart_encode(data
, boundary
=None):
2560 Encode a dict to RFC 7578-compliant form-data
2563 A dict where keys and values can be either Unicode or bytes-like
2566 If specified a Unicode object, it's used as the boundary. Otherwise
2567 a random boundary is generated.
2569 Reference: https://tools.ietf.org/html/rfc7578
2571 has_specified_boundary
= boundary
is not None
2574 if boundary
is None:
2575 boundary
= '---------------' + str(random
.randrange(0x0fffffff, 0xffffffff))
2578 out
, content_type
= _multipart_encode_impl(data
, boundary
)
2581 if has_specified_boundary
:
2585 return out
, content_type
2588 def is_iterable_like(x
, allowed_types
=collections
.abc
.Iterable
, blocked_types
=NO_DEFAULT
):
2589 if blocked_types
is NO_DEFAULT
:
2590 blocked_types
= (str, bytes, collections
.abc
.Mapping
)
2591 return isinstance(x
, allowed_types
) and not isinstance(x
, blocked_types
)
2594 def variadic(x
, allowed_types
=NO_DEFAULT
):
2595 if not isinstance(allowed_types
, (tuple, type)):
2596 deprecation_warning('allowed_types should be a tuple or a type')
2597 allowed_types
= tuple(allowed_types
)
2598 return x
if is_iterable_like(x
, blocked_types
=allowed_types
) else (x
, )
2601 def try_call(*funcs
, expected_type
=None, args
=[], kwargs
={}):
2604 val
= f(*args
, **kwargs
)
2605 except (AttributeError, KeyError, TypeError, IndexError, ValueError, ZeroDivisionError):
2608 if expected_type
is None or isinstance(val
, expected_type
):
2612 def try_get(src
, getter
, expected_type
=None):
2613 return try_call(*variadic(getter
), args
=(src
,), expected_type
=expected_type
)
2616 def filter_dict(dct
, cndn
=lambda _
, v
: v
is not None):
2617 return {k: v for k, v in dct.items() if cndn(k, v)}
2620 def merge_dicts(*dicts
):
2622 for a_dict
in dicts
:
2623 for k
, v
in a_dict
.items():
2624 if (v
is not None and k
not in merged
2625 or isinstance(v
, str) and merged
[k
] == ''):
2630 def encode_compat_str(string
, encoding
=preferredencoding(), errors
='strict'):
2631 return string
if isinstance(string
, str) else str(string
, encoding
, errors
)
2643 TV_PARENTAL_GUIDELINES
= {
2653 def parse_age_limit(s
):
2654 # isinstance(False, int) is True. So type() must be used instead
2655 if type(s
) is int: # noqa: E721
2656 return s
if 0 <= s
<= 21 else None
2657 elif not isinstance(s
, str):
2659 m
= re
.match(r
'^(?P<age>\d{1,2})\+?$', s
)
2661 return int(m
.group('age'))
2664 return US_RATINGS
[s
]
2665 m
= re
.match(r
'^TV[_-]?(%s)$' % '|'.join(k
[3:] for k
in TV_PARENTAL_GUIDELINES
), s
)
2667 return TV_PARENTAL_GUIDELINES
['TV-' + m
.group(1)]
2671 def strip_jsonp(code
):
2674 (?:window\.)?(?P<func_name>[a-zA-Z0-9_.$]*)
2675 (?:\s*&&\s*(?P=func_name))?
2676 \s*\(\s*(?P<callback_data>.*)\);?
2677 \s*?(?://[^\n]*)*$''',
2678 r
'\g<callback_data>', code
)
2681 def js_to_json(code
, vars={}, *, strict
=False):
2682 # vars is a dict of var, val pairs to substitute
2683 STRING_QUOTES
= '\'"`'
2684 STRING_RE
= '|'.join(rf
'{q}(?:\\.|[^\\{q}])*{q}' for q
in STRING_QUOTES
)
2685 COMMENT_RE
= r
'/\*(?:(?!\*/).)*?\*/|//[^\n]*\n'
2686 SKIP_RE
= fr
'\s*(?:{COMMENT_RE})?\s*'
2688 (fr
'(?s)^(0[xX][0-9a-fA-F]+){SKIP_RE}:?$', 16),
2689 (fr
'(?s)^(0+[0-7]+){SKIP_RE}:?$', 8),
2692 def process_escape(match
):
2693 JSON_PASSTHROUGH_ESCAPES
= R
'"\bfnrtu'
2694 escape
= match
.group(1) or match
.group(2)
2696 return (Rf
'\{escape}' if escape
in JSON_PASSTHROUGH_ESCAPES
2697 else R
'\u00' if escape
== 'x'
2698 else '' if escape
== '\n'
2701 def template_substitute(match
):
2702 evaluated
= js_to_json(match
.group(1), vars, strict
=strict
)
2703 if evaluated
[0] == '"':
2704 return json
.loads(evaluated
)
2709 if v
in ('true', 'false', 'null'):
2711 elif v
in ('undefined', 'void 0'):
2713 elif v
.startswith('/*') or v
.startswith('//') or v
.startswith('!') or v
== ',':
2716 if v
[0] in STRING_QUOTES
:
2717 v
= re
.sub(r
'(?s)\${([^}]+)}', template_substitute
, v
[1:-1]) if v
[0] == '`' else v
[1:-1]
2718 escaped
= re
.sub(r
'(?s)(")|\\(.)', process_escape
, v
)
2719 return f
'"{escaped}"'
2721 for regex
, base
in INTEGER_TABLE
:
2722 im
= re
.match(regex
, v
)
2724 i
= int(im
.group(1), base
)
2725 return f
'"{i}":' if v
.endswith(':') else str(i
)
2731 except json
.JSONDecodeError
:
2732 return json
.dumps(vars[v
])
2739 raise ValueError(f
'Unknown value: {v}')
2741 def create_map(mobj
):
2742 return json
.dumps(dict(json
.loads(js_to_json(mobj
.group(1) or '[]', vars=vars))))
2744 code
= re
.sub(r
'(?:new\s+)?Array\((.*?)\)', r
'[\g<1>]', code
)
2745 code
= re
.sub(r
'new Map\((\[.*?\])?\)', create_map
, code
)
2747 code
= re
.sub(rf
'new Date\(({STRING_RE})\)', r
'\g<1>', code
)
2748 code
= re
.sub(r
'new \w+\((.*?)\)', lambda m
: json
.dumps(m
.group(0)), code
)
2749 code
= re
.sub(r
'parseInt\([^\d]+(\d+)[^\d]+\)', r
'\1', code
)
2750 code
= re
.sub(r
'\(function\([^)]*\)\s*\{[^}]*\}\s*\)\s*\(\s*(["\'][^
)]*["\'])\s*\)', r'\1', code)
2752 return re.sub(rf'''(?sx)
2754 {COMMENT_RE}|,(?={SKIP_RE}[\]}}])|
2755 void\s0|(?:(?<![0-9])[eE]|[a-df-zA-DF-Z_$])[.a-zA-Z_$0-9]*|
2756 \b(?:0[xX][0-9a-fA-F]+|0+[0-7]+)(?:{SKIP_RE}:)?|
2757 [0-9]+(?={SKIP_RE}:)|
2762 def qualities(quality_ids):
2763 """ Get a numeric quality value out of a list of possible values """
2766 return quality_ids.index(qid)
2772 POSTPROCESS_WHEN = ('pre_process', 'after_filter', 'video', 'before_dl', 'post_process', 'after_move', 'after_video', 'playlist')
2776 'default': '%(title)s [%(id)s].%(ext)s',
2777 'chapter': '%(title)s - %(section_number)03d %(section_title)s [%(id)s].%(ext)s',
2783 'description': 'description',
2784 'annotation': 'annotations.xml',
2785 'infojson': 'info.json',
2788 'pl_thumbnail': None,
2789 'pl_description': 'description',
2790 'pl_infojson': 'info.json',
2793 # As of [1] format syntax is:
2794 # %[mapping_key][conversion_flags][minimum_width][.precision][length_modifier]type
2795 # 1. https://docs.python.org/2/library/stdtypes.html#string-formatting
2796 STR_FORMAT_RE_TMPL = r'''(?x)
2797 (?<!%)(?P<prefix>(?:%%)*)
2799 (?P<has_key>\((?P<key>{0})\))?
2801 (?P<conversion>[#0\-+ ]+)?
2803 (?P<precision>\.\d+)?
2804 (?P<len_mod>[hlL])? # unused in python
2805 {1} # conversion type
2810 STR_FORMAT_TYPES = 'diouxXeEfFgGcrsa'
2813 def limit_length(s, length):
2814 """ Add ellipses to overly long strings """
2819 return s[:length - len(ELLIPSES)] + ELLIPSES
2823 def version_tuple(v):
2824 return tuple(int(e) for e in re.split(r'[-.]', v))
2827 def is_outdated_version(version, limit, assume_new=True):
2829 return not assume_new
2831 return version_tuple(version) < version_tuple(limit)
2833 return not assume_new
2836 def ytdl_is_updateable():
2837 """ Returns if yt-dlp can be updated with -U """
2839 from ..update import is_non_updateable
2841 return not is_non_updateable()
2844 def args_to_str(args):
2845 # Get a short string representation for a subprocess command
2846 return ' '.join(compat_shlex_quote(a) for a in args)
2849 def error_to_str(err):
2850 return f'{type(err).__name__}: {err}'
2853 def mimetype2ext(mt, default=NO_DEFAULT):
2854 if not isinstance(mt, str):
2855 if default is not NO_DEFAULT:
2872 'x-matroska': 'mkv',
2874 'x-mp4-fragmented': 'mp4',
2879 # application (streaming playlists)
2883 'vnd.apple.mpegurl': 'm3u8',
2884 'vnd.ms-sstr+xml': 'ism',
2885 'x-mpegurl': 'm3u8',
2889 # Per RFC 3003, audio/mpeg can be .mp1, .mp2 or .mp3.
2890 # Using .mp3 as it's the most popular one
2891 'audio/mpeg': 'mp3',
2892 'audio/webm': 'webm',
2893 'audio/x-matroska': 'mka',
2894 'audio/x-mpegurl': 'm3u',
2902 'x-realaudio': 'ra',
2913 'vnd.wap.wbmp': 'wbmp',
2920 'filmstrip+json': 'fs',
2921 'smptett+xml': 'tt',
2924 'x-ms-sami': 'sami',
2933 mimetype = mt.partition(';')[0].strip().lower()
2934 _, _, subtype = mimetype.rpartition('/')
2936 ext = traversal.traverse_obj(MAP, mimetype, subtype, subtype.rsplit('+')[-1])
2939 elif default is not NO_DEFAULT:
2941 return subtype.replace('+', '.')
2944 def ext2mimetype(ext_or_url):
2947 if '.' not in ext_or_url:
2948 ext_or_url = f'file.{ext_or_url}'
2949 return mimetypes.guess_type(ext_or_url)[0]
2952 def parse_codecs(codecs_str):
2953 # http://tools.ietf.org/html/rfc6381
2956 split_codecs = list(filter(None, map(
2957 str.strip, codecs_str.strip().strip(',').split(','))))
2958 vcodec, acodec, scodec, hdr = None, None, None, None
2959 for full_codec in split_codecs:
2960 parts = re.sub(r'0+(?=\d)', '', full_codec).split('.')
2961 if parts[0] in ('avc1', 'avc2', 'avc3', 'avc4', 'vp9', 'vp8', 'hev1', 'hev2',
2962 'h263', 'h264', 'mp4v', 'hvc1', 'av1', 'theora', 'dvh1', 'dvhe'):
2966 if parts[0] in ('dvh1', 'dvhe'):
2968 elif parts[0] == 'av1' and traversal.traverse_obj(parts, 3) == '10':
2970 elif parts[:2] == ['vp9', '2']:
2972 elif parts[0] in ('flac', 'mp4a', 'opus', 'vorbis', 'mp3', 'aac', 'ac-4',
2973 'ac-3', 'ec-3', 'eac3', 'dtsc', 'dtse', 'dtsh', 'dtsl'):
2974 acodec = acodec or full_codec
2975 elif parts[0] in ('stpp', 'wvtt'):
2976 scodec = scodec or full_codec
2978 write_string(f'WARNING: Unknown codec {full_codec}\n')
2979 if vcodec or acodec or scodec:
2981 'vcodec': vcodec or 'none',
2982 'acodec': acodec or 'none',
2983 'dynamic_range': hdr,
2984 **({'scodec': scodec} if scodec is not None else {}),
2986 elif len(split_codecs) == 2:
2988 'vcodec': split_codecs[0],
2989 'acodec': split_codecs[1],
2994 def get_compatible_ext(*, vcodecs, acodecs, vexts, aexts, preferences=None):
2995 assert len(vcodecs) == len(vexts) and len(acodecs) == len(aexts)
2997 allow_mkv = not preferences or 'mkv' in preferences
2999 if allow_mkv and max(len(acodecs), len(vcodecs)) > 1:
3000 return 'mkv' # TODO: any other format allows this?
3002 # TODO: All codecs supported by parse_codecs isn't handled here
3003 COMPATIBLE_CODECS = {
3005 'av1', 'hevc', 'avc1', 'mp4a', 'ac-4', # fourcc (m3u8, mpd)
3006 'h264', 'aacl', 'ec-3', # Set in ISM
3009 'av1', 'vp9', 'vp8', 'opus', 'vrbs',
3010 'vp9x', 'vp8x', # in the webm spec
3014 sanitize_codec = functools.partial(
3015 try_get, getter=lambda x: x[0].split('.')[0].replace('0', '').lower())
3016 vcodec, acodec = sanitize_codec(vcodecs), sanitize_codec(acodecs)
3018 for ext in preferences or COMPATIBLE_CODECS.keys():
3019 codec_set = COMPATIBLE_CODECS.get(ext, set())
3020 if ext == 'mkv' or codec_set.issuperset((vcodec, acodec)):
3024 {'mp3', 'mp4', 'm4a', 'm4p', 'm4b', 'm4r', 'm4v', 'ismv', 'isma', 'mov'},
3027 for ext in preferences or vexts:
3028 current_exts = {ext, *vexts, *aexts}
3029 if ext == 'mkv' or current_exts == {ext} or any(
3030 ext_sets.issuperset(current_exts) for ext_sets in COMPATIBLE_EXTS):
3032 return 'mkv' if allow_mkv else preferences[-1]
3035 def urlhandle_detect_ext(url_handle, default=NO_DEFAULT):
3036 getheader = url_handle.headers.get
3038 cd = getheader('Content-Disposition')
3040 m = re.match(r'attachment;\s*filename="(?P
<filename
>[^
"]+)"', cd)
3042 e = determine_ext(m.group('filename
'), default_ext=None)
3046 meta_ext = getheader('x
-amz
-meta
-name
')
3048 e = meta_ext.rpartition('.')[2]
3052 return mimetype2ext(getheader('Content
-Type
'), default=default)
3055 def encode_data_uri(data, mime_type):
3056 return 'data
:%s;base64
,%s' % (mime_type, base64.b64encode(data).decode('ascii
'))
3059 def age_restricted(content_limit, age_limit):
3060 """ Returns True iff the content should be blocked """
3062 if age_limit is None: # No limit set
3064 if content_limit is None:
3065 return False # Content available for everyone
3066 return age_limit < content_limit
3069 # List of known byte-order-marks (BOM)
3071 (b'\xef\xbb\xbf', 'utf
-8'),
3072 (b'\x00\x00\xfe\xff', 'utf
-32-be
'),
3073 (b'\xff\xfe\x00\x00', 'utf
-32-le
'),
3074 (b'\xff\xfe', 'utf
-16-le
'),
3075 (b'\xfe\xff', 'utf
-16-be
'),
3079 def is_html(first_bytes):
3080 """ Detect whether a file contains HTML by examining its first bytes. """
3083 for bom, enc in BOMS:
3084 while first_bytes.startswith(bom):
3085 encoding, first_bytes = enc, first_bytes[len(bom):]
3087 return re.match(r'^\s
*<', first_bytes.decode(encoding, 'replace
'))
3090 def determine_protocol(info_dict):
3091 protocol = info_dict.get('protocol
')
3092 if protocol is not None:
3095 url = sanitize_url(info_dict['url
'])
3096 if url.startswith('rtmp
'):
3098 elif url.startswith('mms
'):
3100 elif url.startswith('rtsp
'):
3103 ext = determine_ext(url)
3105 return 'm3u8
' if info_dict.get('is_live
') else 'm3u8_native
'
3109 return urllib.parse.urlparse(url).scheme
3112 def render_table(header_row, data, delim=False, extra_gap=0, hide_empty=False):
3113 """ Render a list of rows, each as a list of values.
3114 Text after a \t will be right aligned """
3116 return len(remove_terminal_sequences(string).replace('\t', ''))
3118 def get_max_lens(table):
3119 return [max(width(str(v)) for v in col) for col in zip(*table)]
3121 def filter_using_list(row, filterArray):
3122 return [col for take, col in itertools.zip_longest(filterArray, row, fillvalue=True) if take]
3124 max_lens = get_max_lens(data) if hide_empty else []
3125 header_row = filter_using_list(header_row, max_lens)
3126 data = [filter_using_list(row, max_lens) for row in data]
3128 table = [header_row] + data
3129 max_lens = get_max_lens(table)
3132 table = [header_row, [delim * (ml + extra_gap) for ml in max_lens]] + data
3133 table[1][-1] = table[1][-1][:-extra_gap * len(delim)] # Remove extra_gap from end of delimiter
3135 for pos, text in enumerate(map(str, row)):
3137 row[pos] = text.replace('\t', ' ' * (max_lens[pos] - width(text))) + ' ' * extra_gap
3139 row[pos] = text + ' ' * (max_lens[pos] - width(text) + extra_gap)
3140 ret = '\n'.join(''.join(row).rstrip() for row in table)
3144 def _match_one(filter_part, dct, incomplete):
3145 # TODO: Generalize code with YoutubeDL._build_format_filter
3146 STRING_OPERATORS = {
3147 '*=': operator.contains,
3148 '^
=': lambda attr, value: attr.startswith(value),
3149 '$
=': lambda attr, value: attr.endswith(value),
3150 '~
=': lambda attr, value: re.search(value, attr),
3152 COMPARISON_OPERATORS = {
3154 '<=': operator.le, # "<=" must be defined above "<"
3161 if isinstance(incomplete, bool):
3162 is_incomplete = lambda _: incomplete
3164 is_incomplete = lambda k: k in incomplete
3166 operator_rex = re.compile(r'''(?x)
3168 \s*(?P<negation>!\s*)?(?P<op>%s)(?P<none_inclusive>\s*\?)?\s*
3170 (?P<quote>["\'])(?P<quotedstrval>.+?)(?P=quote)|
3173 ''' % '|
'.join(map(re.escape, COMPARISON_OPERATORS.keys())))
3174 m = operator_rex.fullmatch(filter_part.strip())
3177 unnegated_op = COMPARISON_OPERATORS[m['op
']]
3179 op = lambda attr, value: not unnegated_op(attr, value)
3182 comparison_value = m['quotedstrval
'] or m['strval
'] or m['intval
']
3184 comparison_value = comparison_value.replace(r'\
%s' % m['quote
'], m['quote
'])
3185 actual_value = dct.get(m['key
'])
3186 numeric_comparison = None
3187 if isinstance(actual_value, (int, float)):
3188 # If the original field is a string and matching comparisonvalue is
3189 # a number we should respect the origin of the original field
3190 # and process comparison value as a string (see
3191 # https://github.com/ytdl-org/youtube-dl/issues/11082)
3193 numeric_comparison = int(comparison_value)
3195 numeric_comparison = parse_filesize(comparison_value)
3196 if numeric_comparison is None:
3197 numeric_comparison = parse_filesize(f'{comparison_value}B
')
3198 if numeric_comparison is None:
3199 numeric_comparison = parse_duration(comparison_value)
3200 if numeric_comparison is not None and m['op
'] in STRING_OPERATORS:
3201 raise ValueError('Operator
%s only supports string values
!' % m['op
'])
3202 if actual_value is None:
3203 return is_incomplete(m['key
']) or m['none_inclusive
']
3204 return op(actual_value, comparison_value if numeric_comparison is None else numeric_comparison)
3207 '': lambda v: (v is True) if isinstance(v, bool) else (v is not None),
3208 '!': lambda v: (v is False) if isinstance(v, bool) else (v is None),
3210 operator_rex = re.compile(r'''(?x)
3211 (?P<op>%s)\s*(?P<key>[a-z_]+)
3212 ''' % '|
'.join(map(re.escape, UNARY_OPERATORS.keys())))
3213 m = operator_rex.fullmatch(filter_part.strip())
3215 op = UNARY_OPERATORS[m.group('op
')]
3216 actual_value = dct.get(m.group('key
'))
3217 if is_incomplete(m.group('key
')) and actual_value is None:
3219 return op(actual_value)
3221 raise ValueError('Invalid
filter part
%r' % filter_part)
3224 def match_str(filter_str, dct, incomplete=False):
3225 """ Filter a dictionary with a simple string syntax.
3226 @returns Whether the filter passes
3227 @param incomplete Set of keys that is expected to be missing from dct.
3228 Can be True/False to indicate all/none of the keys may be missing.
3229 All conditions on incomplete keys pass if the key is missing
3232 _match_one(filter_part.replace(r'\
&', '&'), dct, incomplete)
3233 for filter_part in re.split(r'(?
<!\\)&', filter_str))
3236 def match_filter_func(filters, breaking_filters=None):
3237 if not filters and not breaking_filters:
3239 breaking_filters = match_filter_func(breaking_filters) or (lambda _, __: None)
3240 filters = set(variadic(filters or []))
3242 interactive = '-' in filters
3246 def _match_func(info_dict, incomplete=False):
3247 ret = breaking_filters(info_dict, incomplete)
3249 raise RejectedVideoReached(ret)
3251 if not filters or any(match_str(f, info_dict, incomplete) for f in filters):
3252 return NO_DEFAULT if interactive and not incomplete else None
3254 video_title = info_dict.get('title
') or info_dict.get('id') or 'entry
'
3255 filter_str = ') |
('.join(map(str.strip, filters))
3256 return f'{video_title} does
not pass filter ({filter_str}
), skipping
..'
3260 class download_range_func:
3261 def __init__(self, chapters, ranges, from_info=False):
3262 self.chapters, self.ranges, self.from_info = chapters, ranges, from_info
3264 def __call__(self, info_dict, ydl):
3266 warning = ('There are no chapters matching the regex
' if info_dict.get('chapters
')
3267 else 'Cannot match chapters since chapter information
is unavailable
')
3268 for regex in self.chapters or []:
3269 for i, chapter in enumerate(info_dict.get('chapters
') or []):
3270 if re.search(regex, chapter['title
']):
3272 yield {**chapter, 'index': i}
3273 if self.chapters and warning:
3274 ydl.to_screen(f'[info
] {info_dict["id"]}
: {warning}
')
3276 for start, end in self.ranges or []:
3278 'start_time
': self._handle_negative_timestamp(start, info_dict),
3279 'end_time
': self._handle_negative_timestamp(end, info_dict),
3282 if self.from_info and (info_dict.get('start_time
') or info_dict.get('end_time
')):
3284 'start_time
': info_dict.get('start_time
') or 0,
3285 'end_time
': info_dict.get('end_time
') or float('inf
'),
3287 elif not self.ranges and not self.chapters:
3291 def _handle_negative_timestamp(time, info):
3292 return max(info['duration
'] + time, 0) if info.get('duration
') and time < 0 else time
3294 def __eq__(self, other):
3295 return (isinstance(other, download_range_func)
3296 and self.chapters == other.chapters and self.ranges == other.ranges)
3299 return f'{__name__}
.{type(self).__name__}
({self.chapters}
, {self.ranges}
)'
3302 def parse_dfxp_time_expr(time_expr):
3306 mobj = re.match(rf'^
(?P
<time_offset
>{NUMBER_RE}
)s?$
', time_expr)
3308 return float(mobj.group('time_offset
'))
3310 mobj = re.match(r'^
(\d
+):(\d\d
):(\d\
d(?
:(?
:\
.|
:)\d
+)?
)$
', time_expr)
3312 return 3600 * int(mobj.group(1)) + 60 * int(mobj.group(2)) + float(mobj.group(3).replace(':', '.'))
3315 def srt_subtitles_timecode(seconds):
3316 return '%02d
:%02d
:%02d
,%03d
' % timetuple_from_msec(seconds * 1000)
3319 def ass_subtitles_timecode(seconds):
3320 time = timetuple_from_msec(seconds * 1000)
3321 return '%01d
:%02d
:%02d
.%02d
' % (*time[:-1], time.milliseconds / 10)
3324 def dfxp2srt(dfxp_data):
3326 @param dfxp_data A bytes-like object containing DFXP data
3327 @returns A unicode object containing converted SRT data
3329 LEGACY_NAMESPACES = (
3330 (b'http
://www
.w3
.org
/ns
/ttml
', [
3331 b'http
://www
.w3
.org
/2004/11/ttaf1
',
3332 b'http
://www
.w3
.org
/2006/04/ttaf1
',
3333 b'http
://www
.w3
.org
/2006/10/ttaf1
',
3335 (b'http
://www
.w3
.org
/ns
/ttml
#styling', [
3336 b
'http://www.w3.org/ns/ttml#style',
3340 SUPPORTED_STYLING
= [
3349 _x
= functools
.partial(xpath_with_ns
, ns_map
={
3350 'xml': 'http://www.w3.org/XML/1998/namespace',
3351 'ttml': 'http://www.w3.org/ns/ttml',
3352 'tts': 'http://www.w3.org/ns/ttml#styling',
3358 class TTMLPElementParser
:
3360 _unclosed_elements
= []
3361 _applied_styles
= []
3363 def start(self
, tag
, attrib
):
3364 if tag
in (_x('ttml:br'), 'br'):
3367 unclosed_elements
= []
3369 element_style_id
= attrib
.get('style')
3371 style
.update(default_style
)
3372 if element_style_id
:
3373 style
.update(styles
.get(element_style_id
, {}))
3374 for prop
in SUPPORTED_STYLING
:
3375 prop_val
= attrib
.get(_x('tts:' + prop
))
3377 style
[prop
] = prop_val
3380 for k
, v
in sorted(style
.items()):
3381 if self
._applied
_styles
and self
._applied
_styles
[-1].get(k
) == v
:
3384 font
+= ' color="%s"' % v
3385 elif k
== 'fontSize':
3386 font
+= ' size="%s"' % v
3387 elif k
== 'fontFamily':
3388 font
+= ' face="%s"' % v
3389 elif k
== 'fontWeight' and v
== 'bold':
3391 unclosed_elements
.append('b')
3392 elif k
== 'fontStyle' and v
== 'italic':
3394 unclosed_elements
.append('i')
3395 elif k
== 'textDecoration' and v
== 'underline':
3397 unclosed_elements
.append('u')
3399 self
._out
+= '<font' + font
+ '>'
3400 unclosed_elements
.append('font')
3402 if self
._applied
_styles
:
3403 applied_style
.update(self
._applied
_styles
[-1])
3404 applied_style
.update(style
)
3405 self
._applied
_styles
.append(applied_style
)
3406 self
._unclosed
_elements
.append(unclosed_elements
)
3409 if tag
not in (_x('ttml:br'), 'br'):
3410 unclosed_elements
= self
._unclosed
_elements
.pop()
3411 for element
in reversed(unclosed_elements
):
3412 self
._out
+= '</%s>' % element
3413 if unclosed_elements
and self
._applied
_styles
:
3414 self
._applied
_styles
.pop()
3416 def data(self
, data
):
3420 return self
._out
.strip()
3422 # Fix UTF-8 encoded file wrongly marked as UTF-16. See https://github.com/yt-dlp/yt-dlp/issues/6543#issuecomment-1477169870
3423 # This will not trigger false positives since only UTF-8 text is being replaced
3424 dfxp_data
= dfxp_data
.replace(b
'encoding=\'UTF-16\'', b
'encoding=\'UTF-8\'')
3426 def parse_node(node
):
3427 target
= TTMLPElementParser()
3428 parser
= xml
.etree
.ElementTree
.XMLParser(target
=target
)
3429 parser
.feed(xml
.etree
.ElementTree
.tostring(node
))
3430 return parser
.close()
3432 for k
, v
in LEGACY_NAMESPACES
:
3434 dfxp_data
= dfxp_data
.replace(ns
, k
)
3436 dfxp
= compat_etree_fromstring(dfxp_data
)
3438 paras
= dfxp
.findall(_x('.//ttml:p')) or dfxp
.findall('.//p')
3441 raise ValueError('Invalid dfxp/TTML subtitle')
3445 for style
in dfxp
.findall(_x('.//ttml:style')):
3446 style_id
= style
.get('id') or style
.get(_x('xml:id'))
3449 parent_style_id
= style
.get('style')
3451 if parent_style_id
not in styles
:
3454 styles
[style_id
] = styles
[parent_style_id
].copy()
3455 for prop
in SUPPORTED_STYLING
:
3456 prop_val
= style
.get(_x('tts:' + prop
))
3458 styles
.setdefault(style_id
, {})[prop
] = prop_val
3464 for p
in ('body', 'div'):
3465 ele
= xpath_element(dfxp
, [_x('.//ttml:' + p
), './/' + p
])
3468 style
= styles
.get(ele
.get('style'))
3471 default_style
.update(style
)
3473 for para
, index
in zip(paras
, itertools
.count(1)):
3474 begin_time
= parse_dfxp_time_expr(para
.attrib
.get('begin'))
3475 end_time
= parse_dfxp_time_expr(para
.attrib
.get('end'))
3476 dur
= parse_dfxp_time_expr(para
.attrib
.get('dur'))
3477 if begin_time
is None:
3482 end_time
= begin_time
+ dur
3483 out
.append('%d\n%s --> %s\n%s\n\n' % (
3485 srt_subtitles_timecode(begin_time
),
3486 srt_subtitles_timecode(end_time
),
3492 def cli_option(params
, command_option
, param
, separator
=None):
3493 param
= params
.get(param
)
3494 return ([] if param
is None
3495 else [command_option
, str(param
)] if separator
is None
3496 else [f
'{command_option}{separator}{param}'])
3499 def cli_bool_option(params
, command_option
, param
, true_value
='true', false_value
='false', separator
=None):
3500 param
= params
.get(param
)
3501 assert param
in (True, False, None)
3502 return cli_option({True: true_value, False: false_value}
, command_option
, param
, separator
)
3505 def cli_valueless_option(params
, command_option
, param
, expected_value
=True):
3506 return [command_option
] if params
.get(param
) == expected_value
else []
3509 def cli_configuration_args(argdict
, keys
, default
=[], use_compat
=True):
3510 if isinstance(argdict
, (list, tuple)): # for backward compatibility
3517 assert isinstance(argdict
, dict)
3519 assert isinstance(keys
, (list, tuple))
3520 for key_list
in keys
:
3521 arg_list
= list(filter(
3522 lambda x
: x
is not None,
3523 [argdict
.get(key
.lower()) for key
in variadic(key_list
)]))
3525 return [arg
for args
in arg_list
for arg
in args
]
3529 def _configuration_args(main_key
, argdict
, exe
, keys
=None, default
=[], use_compat
=True):
3530 main_key
, exe
= main_key
.lower(), exe
.lower()
3531 root_key
= exe
if main_key
== exe
else f
'{main_key}+{exe}'
3532 keys
= [f
'{root_key}{k}' for k
in (keys
or [''])]
3533 if root_key
in keys
:
3535 keys
.append((main_key
, exe
))
3536 keys
.append('default')
3539 return cli_configuration_args(argdict
, keys
, default
, use_compat
)
3543 # See http://www.loc.gov/standards/iso639-2/ISO-639-2_utf-8.txt
3602 'iw': 'heb', # Replaced by he in 1989 revision
3612 'in': 'ind', # Replaced by id in 1989 revision
3728 'ji': 'yid', # Replaced by yi in 1989 revision
3736 def short2long(cls
, code
):
3737 """Convert language code from ISO 639-1 to ISO 639-2/T"""
3738 return cls
._lang
_map
.get(code
[:2])
3741 def long2short(cls
, code
):
3742 """Convert language code from ISO 639-2/T to ISO 639-1"""
3743 for short_name
, long_name
in cls
._lang
_map
.items():
3744 if long_name
== code
:
3749 # From http://data.okfn.org/data/core/country-list
3751 'AF': 'Afghanistan',
3752 'AX': 'Åland Islands',
3755 'AS': 'American Samoa',
3760 'AG': 'Antigua and Barbuda',
3777 'BO': 'Bolivia, Plurinational State of',
3778 'BQ': 'Bonaire, Sint Eustatius and Saba',
3779 'BA': 'Bosnia and Herzegovina',
3781 'BV': 'Bouvet Island',
3783 'IO': 'British Indian Ocean Territory',
3784 'BN': 'Brunei Darussalam',
3786 'BF': 'Burkina Faso',
3792 'KY': 'Cayman Islands',
3793 'CF': 'Central African Republic',
3797 'CX': 'Christmas Island',
3798 'CC': 'Cocos (Keeling) Islands',
3802 'CD': 'Congo, the Democratic Republic of the',
3803 'CK': 'Cook Islands',
3805 'CI': 'Côte d\'Ivoire',
3810 'CZ': 'Czech Republic',
3814 'DO': 'Dominican Republic',
3817 'SV': 'El Salvador',
3818 'GQ': 'Equatorial Guinea',
3822 'FK': 'Falkland Islands (Malvinas)',
3823 'FO': 'Faroe Islands',
3827 'GF': 'French Guiana',
3828 'PF': 'French Polynesia',
3829 'TF': 'French Southern Territories',
3844 'GW': 'Guinea-Bissau',
3847 'HM': 'Heard Island and McDonald Islands',
3848 'VA': 'Holy See (Vatican City State)',
3855 'IR': 'Iran, Islamic Republic of',
3858 'IM': 'Isle of Man',
3868 'KP': 'Korea, Democratic People\'s Republic of',
3869 'KR': 'Korea, Republic of',
3872 'LA': 'Lao People\'s Democratic Republic',
3878 'LI': 'Liechtenstein',
3882 'MK': 'Macedonia, the Former Yugoslav Republic of',
3889 'MH': 'Marshall Islands',
3895 'FM': 'Micronesia, Federated States of',
3896 'MD': 'Moldova, Republic of',
3907 'NL': 'Netherlands',
3908 'NC': 'New Caledonia',
3909 'NZ': 'New Zealand',
3914 'NF': 'Norfolk Island',
3915 'MP': 'Northern Mariana Islands',
3920 'PS': 'Palestine, State of',
3922 'PG': 'Papua New Guinea',
3925 'PH': 'Philippines',
3929 'PR': 'Puerto Rico',
3933 'RU': 'Russian Federation',
3935 'BL': 'Saint Barthélemy',
3936 'SH': 'Saint Helena, Ascension and Tristan da Cunha',
3937 'KN': 'Saint Kitts and Nevis',
3938 'LC': 'Saint Lucia',
3939 'MF': 'Saint Martin (French part)',
3940 'PM': 'Saint Pierre and Miquelon',
3941 'VC': 'Saint Vincent and the Grenadines',
3944 'ST': 'Sao Tome and Principe',
3945 'SA': 'Saudi Arabia',
3949 'SL': 'Sierra Leone',
3951 'SX': 'Sint Maarten (Dutch part)',
3954 'SB': 'Solomon Islands',
3956 'ZA': 'South Africa',
3957 'GS': 'South Georgia and the South Sandwich Islands',
3958 'SS': 'South Sudan',
3963 'SJ': 'Svalbard and Jan Mayen',
3966 'CH': 'Switzerland',
3967 'SY': 'Syrian Arab Republic',
3968 'TW': 'Taiwan, Province of China',
3970 'TZ': 'Tanzania, United Republic of',
3972 'TL': 'Timor-Leste',
3976 'TT': 'Trinidad and Tobago',
3979 'TM': 'Turkmenistan',
3980 'TC': 'Turks and Caicos Islands',
3984 'AE': 'United Arab Emirates',
3985 'GB': 'United Kingdom',
3986 'US': 'United States',
3987 'UM': 'United States Minor Outlying Islands',
3991 'VE': 'Venezuela, Bolivarian Republic of',
3993 'VG': 'Virgin Islands, British',
3994 'VI': 'Virgin Islands, U.S.',
3995 'WF': 'Wallis and Futuna',
3996 'EH': 'Western Sahara',
4000 # Not ISO 3166 codes, but used for IP blocks
4001 'AP': 'Asia/Pacific Region',
4006 def short2full(cls
, code
):
4007 """Convert an ISO 3166-2 country code to the corresponding full name"""
4008 return cls
._country
_map
.get(code
.upper())
4012 # Major IPv4 address blocks per country
4014 'AD': '46.172.224.0/19',
4015 'AE': '94.200.0.0/13',
4016 'AF': '149.54.0.0/17',
4017 'AG': '209.59.64.0/18',
4018 'AI': '204.14.248.0/21',
4019 'AL': '46.99.0.0/16',
4020 'AM': '46.70.0.0/15',
4021 'AO': '105.168.0.0/13',
4022 'AP': '182.50.184.0/21',
4023 'AQ': '23.154.160.0/24',
4024 'AR': '181.0.0.0/12',
4025 'AS': '202.70.112.0/20',
4026 'AT': '77.116.0.0/14',
4027 'AU': '1.128.0.0/11',
4028 'AW': '181.41.0.0/18',
4029 'AX': '185.217.4.0/22',
4030 'AZ': '5.197.0.0/16',
4031 'BA': '31.176.128.0/17',
4032 'BB': '65.48.128.0/17',
4033 'BD': '114.130.0.0/16',
4035 'BF': '102.178.0.0/15',
4036 'BG': '95.42.0.0/15',
4037 'BH': '37.131.0.0/17',
4038 'BI': '154.117.192.0/18',
4039 'BJ': '137.255.0.0/16',
4040 'BL': '185.212.72.0/23',
4041 'BM': '196.12.64.0/18',
4042 'BN': '156.31.0.0/16',
4043 'BO': '161.56.0.0/16',
4044 'BQ': '161.0.80.0/20',
4045 'BR': '191.128.0.0/12',
4046 'BS': '24.51.64.0/18',
4047 'BT': '119.2.96.0/19',
4048 'BW': '168.167.0.0/16',
4049 'BY': '178.120.0.0/13',
4050 'BZ': '179.42.192.0/18',
4051 'CA': '99.224.0.0/11',
4052 'CD': '41.243.0.0/16',
4053 'CF': '197.242.176.0/21',
4054 'CG': '160.113.0.0/16',
4055 'CH': '85.0.0.0/13',
4056 'CI': '102.136.0.0/14',
4057 'CK': '202.65.32.0/19',
4058 'CL': '152.172.0.0/14',
4059 'CM': '102.244.0.0/14',
4060 'CN': '36.128.0.0/10',
4061 'CO': '181.240.0.0/12',
4062 'CR': '201.192.0.0/12',
4063 'CU': '152.206.0.0/15',
4064 'CV': '165.90.96.0/19',
4065 'CW': '190.88.128.0/17',
4066 'CY': '31.153.0.0/16',
4067 'CZ': '88.100.0.0/14',
4069 'DJ': '197.241.0.0/17',
4070 'DK': '87.48.0.0/12',
4071 'DM': '192.243.48.0/20',
4072 'DO': '152.166.0.0/15',
4073 'DZ': '41.96.0.0/12',
4074 'EC': '186.68.0.0/15',
4075 'EE': '90.190.0.0/15',
4076 'EG': '156.160.0.0/11',
4077 'ER': '196.200.96.0/20',
4078 'ES': '88.0.0.0/11',
4079 'ET': '196.188.0.0/14',
4080 'EU': '2.16.0.0/13',
4081 'FI': '91.152.0.0/13',
4082 'FJ': '144.120.0.0/16',
4083 'FK': '80.73.208.0/21',
4084 'FM': '119.252.112.0/20',
4085 'FO': '88.85.32.0/19',
4087 'GA': '41.158.0.0/15',
4089 'GD': '74.122.88.0/21',
4090 'GE': '31.146.0.0/16',
4091 'GF': '161.22.64.0/18',
4092 'GG': '62.68.160.0/19',
4093 'GH': '154.160.0.0/12',
4094 'GI': '95.164.0.0/16',
4095 'GL': '88.83.0.0/19',
4096 'GM': '160.182.0.0/15',
4097 'GN': '197.149.192.0/18',
4098 'GP': '104.250.0.0/19',
4099 'GQ': '105.235.224.0/20',
4100 'GR': '94.64.0.0/13',
4101 'GT': '168.234.0.0/16',
4102 'GU': '168.123.0.0/16',
4103 'GW': '197.214.80.0/20',
4104 'GY': '181.41.64.0/18',
4105 'HK': '113.252.0.0/14',
4106 'HN': '181.210.0.0/16',
4107 'HR': '93.136.0.0/13',
4108 'HT': '148.102.128.0/17',
4109 'HU': '84.0.0.0/14',
4110 'ID': '39.192.0.0/10',
4111 'IE': '87.32.0.0/12',
4112 'IL': '79.176.0.0/13',
4113 'IM': '5.62.80.0/20',
4114 'IN': '117.192.0.0/10',
4115 'IO': '203.83.48.0/21',
4116 'IQ': '37.236.0.0/14',
4117 'IR': '2.176.0.0/12',
4118 'IS': '82.221.0.0/16',
4119 'IT': '79.0.0.0/10',
4120 'JE': '87.244.64.0/18',
4121 'JM': '72.27.0.0/17',
4122 'JO': '176.29.0.0/16',
4123 'JP': '133.0.0.0/8',
4124 'KE': '105.48.0.0/12',
4125 'KG': '158.181.128.0/17',
4126 'KH': '36.37.128.0/17',
4127 'KI': '103.25.140.0/22',
4128 'KM': '197.255.224.0/20',
4129 'KN': '198.167.192.0/19',
4130 'KP': '175.45.176.0/22',
4131 'KR': '175.192.0.0/10',
4132 'KW': '37.36.0.0/14',
4133 'KY': '64.96.0.0/15',
4134 'KZ': '2.72.0.0/13',
4135 'LA': '115.84.64.0/18',
4136 'LB': '178.135.0.0/16',
4137 'LC': '24.92.144.0/20',
4138 'LI': '82.117.0.0/19',
4139 'LK': '112.134.0.0/15',
4140 'LR': '102.183.0.0/16',
4141 'LS': '129.232.0.0/17',
4142 'LT': '78.56.0.0/13',
4143 'LU': '188.42.0.0/16',
4144 'LV': '46.109.0.0/16',
4145 'LY': '41.252.0.0/14',
4146 'MA': '105.128.0.0/11',
4147 'MC': '88.209.64.0/18',
4148 'MD': '37.246.0.0/16',
4149 'ME': '178.175.0.0/17',
4150 'MF': '74.112.232.0/21',
4151 'MG': '154.126.0.0/17',
4152 'MH': '117.103.88.0/21',
4153 'MK': '77.28.0.0/15',
4154 'ML': '154.118.128.0/18',
4155 'MM': '37.111.0.0/17',
4156 'MN': '49.0.128.0/17',
4157 'MO': '60.246.0.0/16',
4158 'MP': '202.88.64.0/20',
4159 'MQ': '109.203.224.0/19',
4160 'MR': '41.188.64.0/18',
4161 'MS': '208.90.112.0/22',
4162 'MT': '46.11.0.0/16',
4163 'MU': '105.16.0.0/12',
4164 'MV': '27.114.128.0/18',
4165 'MW': '102.70.0.0/15',
4166 'MX': '187.192.0.0/11',
4167 'MY': '175.136.0.0/13',
4168 'MZ': '197.218.0.0/15',
4169 'NA': '41.182.0.0/16',
4170 'NC': '101.101.0.0/18',
4171 'NE': '197.214.0.0/18',
4172 'NF': '203.17.240.0/22',
4173 'NG': '105.112.0.0/12',
4174 'NI': '186.76.0.0/15',
4175 'NL': '145.96.0.0/11',
4176 'NO': '84.208.0.0/13',
4177 'NP': '36.252.0.0/15',
4178 'NR': '203.98.224.0/19',
4179 'NU': '49.156.48.0/22',
4180 'NZ': '49.224.0.0/14',
4181 'OM': '5.36.0.0/15',
4182 'PA': '186.72.0.0/15',
4183 'PE': '186.160.0.0/14',
4184 'PF': '123.50.64.0/18',
4185 'PG': '124.240.192.0/19',
4186 'PH': '49.144.0.0/13',
4187 'PK': '39.32.0.0/11',
4188 'PL': '83.0.0.0/11',
4189 'PM': '70.36.0.0/20',
4190 'PR': '66.50.0.0/16',
4191 'PS': '188.161.0.0/16',
4192 'PT': '85.240.0.0/13',
4193 'PW': '202.124.224.0/20',
4194 'PY': '181.120.0.0/14',
4195 'QA': '37.210.0.0/15',
4196 'RE': '102.35.0.0/16',
4197 'RO': '79.112.0.0/13',
4198 'RS': '93.86.0.0/15',
4199 'RU': '5.136.0.0/13',
4200 'RW': '41.186.0.0/16',
4201 'SA': '188.48.0.0/13',
4202 'SB': '202.1.160.0/19',
4203 'SC': '154.192.0.0/11',
4204 'SD': '102.120.0.0/13',
4205 'SE': '78.64.0.0/12',
4206 'SG': '8.128.0.0/10',
4207 'SI': '188.196.0.0/14',
4208 'SK': '78.98.0.0/15',
4209 'SL': '102.143.0.0/17',
4210 'SM': '89.186.32.0/19',
4211 'SN': '41.82.0.0/15',
4212 'SO': '154.115.192.0/18',
4213 'SR': '186.179.128.0/17',
4214 'SS': '105.235.208.0/21',
4215 'ST': '197.159.160.0/19',
4216 'SV': '168.243.0.0/16',
4217 'SX': '190.102.0.0/20',
4219 'SZ': '41.84.224.0/19',
4220 'TC': '65.255.48.0/20',
4221 'TD': '154.68.128.0/19',
4222 'TG': '196.168.0.0/14',
4223 'TH': '171.96.0.0/13',
4224 'TJ': '85.9.128.0/18',
4225 'TK': '27.96.24.0/21',
4226 'TL': '180.189.160.0/20',
4227 'TM': '95.85.96.0/19',
4228 'TN': '197.0.0.0/11',
4229 'TO': '175.176.144.0/21',
4230 'TR': '78.160.0.0/11',
4231 'TT': '186.44.0.0/15',
4232 'TV': '202.2.96.0/19',
4233 'TW': '120.96.0.0/11',
4234 'TZ': '156.156.0.0/14',
4235 'UA': '37.52.0.0/14',
4236 'UG': '102.80.0.0/13',
4238 'UY': '167.56.0.0/13',
4239 'UZ': '84.54.64.0/18',
4240 'VA': '212.77.0.0/19',
4241 'VC': '207.191.240.0/21',
4242 'VE': '186.88.0.0/13',
4243 'VG': '66.81.192.0/20',
4244 'VI': '146.226.0.0/16',
4245 'VN': '14.160.0.0/11',
4246 'VU': '202.80.32.0/20',
4247 'WF': '117.20.32.0/21',
4248 'WS': '202.4.32.0/19',
4249 'YE': '134.35.0.0/16',
4250 'YT': '41.242.116.0/22',
4251 'ZA': '41.0.0.0/11',
4252 'ZM': '102.144.0.0/13',
4253 'ZW': '102.177.192.0/18',
4257 def random_ipv4(cls
, code_or_block
):
4258 if len(code_or_block
) == 2:
4259 block
= cls
._country
_ip
_map
.get(code_or_block
.upper())
4263 block
= code_or_block
4264 addr
, preflen
= block
.split('/')
4265 addr_min
= struct
.unpack('!L', socket
.inet_aton(addr
))[0]
4266 addr_max
= addr_min |
(0xffffffff >> int(preflen
))
4267 return str(socket
.inet_ntoa(
4268 struct
.pack('!L', random
.randint(addr_min
, addr_max
))))
4271 # Both long_to_bytes and bytes_to_long are adapted from PyCrypto, which is
4272 # released into Public Domain
4273 # https://github.com/dlitz/pycrypto/blob/master/lib/Crypto/Util/number.py#L387
4275 def long_to_bytes(n
, blocksize
=0):
4276 """long_to_bytes(n:long, blocksize:int) : string
4277 Convert a long integer to a byte string.
4279 If optional blocksize is given and greater than zero, pad the front of the
4280 byte string with binary zeros so that the length is a multiple of
4283 # after much testing, this algorithm was deemed to be the fastest
4287 s
= struct
.pack('>I', n
& 0xffffffff) + s
4289 # strip off leading zeros
4290 for i
in range(len(s
)):
4291 if s
[i
] != b
'\000'[0]:
4294 # only happens when n == 0
4298 # add back some pad bytes. this could be done more efficiently w.r.t. the
4299 # de-padding being done above, but sigh...
4300 if blocksize
> 0 and len(s
) % blocksize
:
4301 s
= (blocksize
- len(s
) % blocksize
) * b
'\000' + s
4305 def bytes_to_long(s
):
4306 """bytes_to_long(string) : long
4307 Convert a byte string to a long integer.
4309 This is (essentially) the inverse of long_to_bytes().
4314 extra
= (4 - length
% 4)
4315 s
= b
'\000' * extra
+ s
4316 length
= length
+ extra
4317 for i
in range(0, length
, 4):
4318 acc
= (acc
<< 32) + struct
.unpack('>I', s
[i
:i
+ 4])[0]
4322 def ohdave_rsa_encrypt(data
, exponent
, modulus
):
4324 Implement OHDave's RSA algorithm. See http://www.ohdave.com/rsa/
4327 data: data to encrypt, bytes-like object
4328 exponent, modulus: parameter e and N of RSA algorithm, both integer
4329 Output: hex string of encrypted data
4331 Limitation: supports one block encryption only
4334 payload
= int(binascii
.hexlify(data
[::-1]), 16)
4335 encrypted
= pow(payload
, exponent
, modulus
)
4336 return '%x' % encrypted
4339 def pkcs1pad(data
, length
):
4341 Padding input data with PKCS#1 scheme
4343 @param {int[]} data input data
4344 @param {int} length target length
4345 @returns {int[]} padded data
4347 if len(data
) > length
- 11:
4348 raise ValueError('Input data too long for PKCS#1 padding')
4350 pseudo_random
= [random
.randint(0, 254) for _
in range(length
- len(data
) - 3)]
4351 return [0, 2] + pseudo_random
+ [0] + data
4354 def _base_n_table(n
, table
):
4355 if not table
and not n
:
4356 raise ValueError('Either table or n must be specified')
4357 table
= (table
or '0123456789abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ')[:n
]
4359 if n
and n
!= len(table
):
4360 raise ValueError(f
'base {n} exceeds table length {len(table)}')
4364 def encode_base_n(num
, n
=None, table
=None):
4365 """Convert given int to a base-n string"""
4366 table
= _base_n_table(n
, table
)
4370 result
, base
= '', len(table
)
4372 result
= table
[num
% base
] + result
4377 def decode_base_n(string
, n
=None, table
=None):
4378 """Convert given base-n string to int"""
4379 table
= {char: index for index, char in enumerate(_base_n_table(n, table))}
4380 result
, base
= 0, len(table
)
4382 result
= result
* base
+ table
[char
]
4386 def decode_packed_codes(code
):
4387 mobj
= re
.search(PACKED_CODES_RE
, code
)
4388 obfuscated_code
, base
, count
, symbols
= mobj
.groups()
4391 symbols
= symbols
.split('|')
4396 base_n_count
= encode_base_n(count
, base
)
4397 symbol_table
[base_n_count
] = symbols
[count
] or base_n_count
4400 r
'\b(\w+)\b', lambda mobj
: symbol_table
[mobj
.group(0)],
4404 def caesar(s
, alphabet
, shift
):
4409 alphabet
[(alphabet
.index(c
) + shift
) % l
] if c
in alphabet
else c
4414 return caesar(s
, r
'''!"#$%&'()*+,-./0123456789:;<=>?@ABCDEFGHIJKLMNOPQRSTUVWXYZ[\]^_`abcdefghijklmnopqrstuvwxyz{|}~''', 47)
4417 def parse_m3u8_attributes(attrib
):
4419 for (key
, val
) in re
.findall(r
'(?P<key>[A-Z0-9-]+)=(?P<val>"[^"]+"|[^",]+)(?:,|$)', attrib
):
4420 if val
.startswith('"'):
4426 def urshift(val
, n
):
4427 return val
>> n
if val
>= 0 else (val
+ 0x100000000) >> n
4430 def write_xattr(path
, key
, value
):
4431 # Windows: Write xattrs to NTFS Alternate Data Streams:
4432 # http://en.wikipedia.org/wiki/NTFS#Alternate_data_streams_.28ADS.29
4433 if compat_os_name
== 'nt':
4434 assert ':' not in key
4435 assert os
.path
.exists(path
)
4438 with open(f
'{path}:{key}', 'wb') as f
:
4440 except OSError as e
:
4441 raise XAttrMetadataError(e
.errno
, e
.strerror
)
4444 # UNIX Method 1. Use os.setxattr/xattrs/pyxattrs modules
4447 if callable(getattr(os
, 'setxattr', None)):
4448 setxattr
= os
.setxattr
4449 elif getattr(xattr
, '_yt_dlp__identifier', None) == 'pyxattr':
4450 # Unicode arguments are not supported in pyxattr until version 0.5.0
4451 # See https://github.com/ytdl-org/youtube-dl/issues/5498
4452 if version_tuple(xattr
.__version
__) >= (0, 5, 0):
4453 setxattr
= xattr
.set
4455 setxattr
= xattr
.setxattr
4459 setxattr(path
, key
, value
)
4460 except OSError as e
:
4461 raise XAttrMetadataError(e
.errno
, e
.strerror
)
4464 # UNIX Method 2. Use setfattr/xattr executables
4465 exe
= ('setfattr' if check_executable('setfattr', ['--version'])
4466 else 'xattr' if check_executable('xattr', ['-h']) else None)
4468 raise XAttrUnavailableError(
4469 'Couldn\'t find a tool to set the xattrs. Install either the python "xattr" or "pyxattr" modules or the '
4470 + ('"xattr" binary' if sys
.platform
!= 'linux' else 'GNU "attr" package (which contains the "setfattr" tool)'))
4472 value
= value
.decode()
4474 _
, stderr
, returncode
= Popen
.run(
4475 [exe
, '-w', key
, value
, path
] if exe
== 'xattr' else [exe
, '-n', key
, '-v', value
, path
],
4476 text
=True, stdout
=subprocess
.PIPE
, stderr
=subprocess
.PIPE
, stdin
=subprocess
.PIPE
)
4477 except OSError as e
:
4478 raise XAttrMetadataError(e
.errno
, e
.strerror
)
4480 raise XAttrMetadataError(returncode
, stderr
)
4483 def random_birthday(year_field
, month_field
, day_field
):
4484 start_date
= datetime
.date(1950, 1, 1)
4485 end_date
= datetime
.date(1995, 12, 31)
4486 offset
= random
.randint(0, (end_date
- start_date
).days
)
4487 random_date
= start_date
+ datetime
.timedelta(offset
)
4489 year_field
: str(random_date
.year
),
4490 month_field
: str(random_date
.month
),
4491 day_field
: str(random_date
.day
),
4495 def find_available_port(interface
=''):
4497 with socket
.socket() as sock
:
4498 sock
.bind((interface
, 0))
4499 return sock
.getsockname()[1]
4504 # Templates for internet shortcut files, which are plain text files.
4505 DOT_URL_LINK_TEMPLATE
= '''\
4510 DOT_WEBLOC_LINK_TEMPLATE
= '''\
4511 <?xml version="1.0" encoding="UTF-8"?>
4512 <!DOCTYPE plist PUBLIC "-//Apple//DTD PLIST 1.0//EN" "http://www.apple.com/DTDs/PropertyList-1.0.dtd">
4513 <plist version="1.0">
4516 \t<string>%(url)s</string>
4521 DOT_DESKTOP_LINK_TEMPLATE
= '''\
4531 'url': DOT_URL_LINK_TEMPLATE
,
4532 'desktop': DOT_DESKTOP_LINK_TEMPLATE
,
4533 'webloc': DOT_WEBLOC_LINK_TEMPLATE
,
4537 def iri_to_uri(iri
):
4539 Converts an IRI (Internationalized Resource Identifier, allowing Unicode characters) to a URI (Uniform Resource Identifier, ASCII-only).
4541 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.
4544 iri_parts
= urllib
.parse
.urlparse(iri
)
4546 if '[' in iri_parts
.netloc
:
4547 raise ValueError('IPv6 URIs are not, yet, supported.')
4548 # Querying `.netloc`, when there's only one bracket, also raises a ValueError.
4550 # 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.
4553 if iri_parts
.username
:
4554 net_location
+= urllib
.parse
.quote(iri_parts
.username
, safe
=r
"!$%&'()*+,~")
4555 if iri_parts
.password
is not None:
4556 net_location
+= ':' + urllib
.parse
.quote(iri_parts
.password
, safe
=r
"!$%&'()*+,~")
4559 net_location
+= iri_parts
.hostname
.encode('idna').decode() # Punycode for Unicode hostnames.
4560 # The 'idna' encoding produces ASCII text.
4561 if iri_parts
.port
is not None and iri_parts
.port
!= 80:
4562 net_location
+= ':' + str(iri_parts
.port
)
4564 return urllib
.parse
.urlunparse(
4568 urllib
.parse
.quote_plus(iri_parts
.path
, safe
=r
"!$%&'()*+,/:;=@|~"),
4570 # Unsure about the `safe` argument, since this is a legacy way of handling parameters.
4571 urllib
.parse
.quote_plus(iri_parts
.params
, safe
=r
"!$%&'()*+,/:;=@|~"),
4573 # Not totally sure about the `safe` argument, since the source does not explicitly mention the query URI component.
4574 urllib
.parse
.quote_plus(iri_parts
.query
, safe
=r
"!$%&'()*+,/:;=?@{|}~"),
4576 urllib
.parse
.quote_plus(iri_parts
.fragment
, safe
=r
"!#$%&'()*+,/:;=?@{|}~")))
4578 # Source for `safe` arguments: https://url.spec.whatwg.org/#percent-encoded-bytes.
4581 def to_high_limit_path(path
):
4582 if sys
.platform
in ['win32', 'cygwin']:
4583 # Work around MAX_PATH limitation on Windows. The maximum allowed length for the individual path segments may still be quite limited.
4584 return '\\\\?\\' + os
.path
.abspath(path
)
4589 def format_field(obj
, field
=None, template
='%s', ignore
=NO_DEFAULT
, default
='', func
=IDENTITY
):
4590 val
= traversal
.traverse_obj(obj
, *variadic(field
))
4591 if not val
if ignore
is NO_DEFAULT
else val
in variadic(ignore
):
4593 return template
% func(val
)
4596 def clean_podcast_url(url
):
4597 url
= re
.sub(r
'''(?x)
4601 media\.blubrry\.com| # https://create.blubrry.com/resources/podcast-media-download-statistics/getting-started/
4606 (?:dts|www)\.podtrac\.com/(?:pts/)?redirect\.[0-9a-z]{3,4}| # http://analytics.podtrac.com/how-to-measure
4609 cn\.co| # https://podcorn.com/analytics-prefix/
4610 st\.fm # https://podsights.com/docs/
4615 return re
.sub(r
'^\w+://(\w+://)', r
'\1', url
)
4618 _HEX_TABLE
= '0123456789abcdef'
4621 def random_uuidv4():
4622 return re
.sub(r
'[xy]', lambda x
: _HEX_TABLE
[random
.randint(0, 15)], 'xxxxxxxx-xxxx-4xxx-yxxx-xxxxxxxxxxxx')
4625 def make_dir(path
, to_screen
=None):
4627 dn
= os
.path
.dirname(path
)
4629 os
.makedirs(dn
, exist_ok
=True)
4631 except OSError as err
:
4632 if callable(to_screen
) is not None:
4633 to_screen(f
'unable to create directory {err}')
4637 def get_executable_path():
4638 from ..update
import _get_variant_and_executable_path
4640 return os
.path
.dirname(os
.path
.abspath(_get_variant_and_executable_path()[1]))
4643 def get_user_config_dirs(package_name
):
4644 # .config (e.g. ~/.config/package_name)
4645 xdg_config_home
= os
.getenv('XDG_CONFIG_HOME') or compat_expanduser('~/.config')
4646 yield os
.path
.join(xdg_config_home
, package_name
)
4648 # appdata (%APPDATA%/package_name)
4649 appdata_dir
= os
.getenv('appdata')
4651 yield os
.path
.join(appdata_dir
, package_name
)
4653 # home (~/.package_name)
4654 yield os
.path
.join(compat_expanduser('~'), f
'.{package_name}')
4657 def get_system_config_dirs(package_name
):
4659 yield os
.path
.join('/etc', package_name
)
4662 def time_seconds(**kwargs
):
4664 Returns TZ-aware time in seconds since the epoch (1970-01-01T00:00:00Z)
4666 return time
.time() + datetime
.timedelta(**kwargs
).total_seconds()
4669 # create a JSON Web Signature (jws) with HS256 algorithm
4670 # the resulting format is in JWS Compact Serialization
4671 # implemented following JWT https://www.rfc-editor.org/rfc/rfc7519.html
4672 # implemented following JWS https://www.rfc-editor.org/rfc/rfc7515.html
4673 def jwt_encode_hs256(payload_data
, key
, headers
={}):
4679 header_data
.update(headers
)
4680 header_b64
= base64
.b64encode(json
.dumps(header_data
).encode())
4681 payload_b64
= base64
.b64encode(json
.dumps(payload_data
).encode())
4682 h
= hmac
.new(key
.encode(), header_b64
+ b
'.' + payload_b64
, hashlib
.sha256
)
4683 signature_b64
= base64
.b64encode(h
.digest())
4684 token
= header_b64
+ b
'.' + payload_b64
+ b
'.' + signature_b64
4688 # can be extended in future to verify the signature and parse header and return the algorithm used if it's not HS256
4689 def jwt_decode_hs256(jwt
):
4690 header_b64
, payload_b64
, signature_b64
= jwt
.split('.')
4691 # add trailing ='s that may have been stripped, superfluous ='s are ignored
4692 payload_data
= json
.loads(base64
.urlsafe_b64decode(f
'{payload_b64}==='))
4696 WINDOWS_VT_MODE
= False if compat_os_name
== 'nt' else None
4700 def supports_terminal_sequences(stream
):
4701 if compat_os_name
== 'nt':
4702 if not WINDOWS_VT_MODE
:
4704 elif not os
.getenv('TERM'):
4707 return stream
.isatty()
4708 except BaseException
:
4712 def windows_enable_vt_mode():
4713 """Ref: https://bugs.python.org/issue30075 """
4714 if get_windows_version() < (10, 0, 10586):
4718 import ctypes
.wintypes
4721 ENABLE_VIRTUAL_TERMINAL_PROCESSING
= 0x0004
4723 dll
= ctypes
.WinDLL('kernel32', use_last_error
=False)
4724 handle
= os
.open('CONOUT$', os
.O_RDWR
)
4726 h_out
= ctypes
.wintypes
.HANDLE(msvcrt
.get_osfhandle(handle
))
4727 dw_original_mode
= ctypes
.wintypes
.DWORD()
4728 success
= dll
.GetConsoleMode(h_out
, ctypes
.byref(dw_original_mode
))
4730 raise Exception('GetConsoleMode failed')
4732 success
= dll
.SetConsoleMode(h_out
, ctypes
.wintypes
.DWORD(
4733 dw_original_mode
.value | ENABLE_VIRTUAL_TERMINAL_PROCESSING
))
4735 raise Exception('SetConsoleMode failed')
4739 global WINDOWS_VT_MODE
4740 WINDOWS_VT_MODE
= True
4741 supports_terminal_sequences
.cache_clear()
4744 _terminal_sequences_re
= re
.compile('\033\\[[^m]+m')
4747 def remove_terminal_sequences(string
):
4748 return _terminal_sequences_re
.sub('', string
)
4751 def number_of_digits(number
):
4752 return len('%d' % number
)
4755 def join_nonempty(*values
, delim
='-', from_dict
=None):
4756 if from_dict
is not None:
4757 values
= (traversal
.traverse_obj(from_dict
, variadic(v
)) for v
in values
)
4758 return delim
.join(map(str, filter(None, values
)))
4761 def scale_thumbnails_to_max_format_width(formats
, thumbnails
, url_width_re
):
4763 Find the largest format dimensions in terms of video width and, for each thumbnail:
4764 * Modify the URL: Match the width with the provided regex and replace with the former width
4767 This function is useful with video services that scale the provided thumbnails on demand
4769 _keys
= ('width', 'height')
4770 max_dimensions
= max(
4771 (tuple(format
.get(k
) or 0 for k
in _keys
) for format
in formats
),
4773 if not max_dimensions
[0]:
4777 {'url': re.sub(url_width_re, str(max_dimensions[0]), thumbnail['url'])}
,
4778 dict(zip(_keys
, max_dimensions
)), thumbnail
)
4779 for thumbnail
in thumbnails
4783 def parse_http_range(range):
4784 """ Parse value of "Range" or "Content-Range" HTTP header into tuple. """
4786 return None, None, None
4787 crg
= re
.search(r
'bytes[ =](\d+)-(\d+)?(?:/(\d+))?', range)
4789 return None, None, None
4790 return int(crg
.group(1)), int_or_none(crg
.group(2)), int_or_none(crg
.group(3))
4793 def read_stdin(what
):
4794 eof
= 'Ctrl+Z' if compat_os_name
== 'nt' else 'Ctrl+D'
4795 write_string(f
'Reading {what} from STDIN - EOF ({eof}) to end:\n')
4799 def determine_file_encoding(data
):
4801 Detect the text encoding used
4802 @returns (encoding, bytes to skip)
4805 # BOM marks are given priority over declarations
4806 for bom
, enc
in BOMS
:
4807 if data
.startswith(bom
):
4808 return enc
, len(bom
)
4810 # Strip off all null bytes to match even when UTF-16 or UTF-32 is used.
4811 # We ignore the endianness to get a good enough match
4812 data
= data
.replace(b
'\0', b
'')
4813 mobj
= re
.match(rb
'(?m)^#\s*coding\s*:\s*(\S+)\s*$', data
)
4814 return mobj
.group(1).decode() if mobj
else None, 0
4821 __initialized
= False
4823 def __init__(self
, parser
, label
=None):
4824 self
.parser
, self
.label
= parser
, label
4825 self
._loaded
_paths
, self
.configs
= set(), []
4827 def init(self
, args
=None, filename
=None):
4828 assert not self
.__initialized
4829 self
.own_args
, self
.filename
= args
, filename
4830 return self
.load_configs()
4832 def load_configs(self
):
4835 location
= os
.path
.realpath(self
.filename
)
4836 directory
= os
.path
.dirname(location
)
4837 if location
in self
._loaded
_paths
:
4839 self
._loaded
_paths
.add(location
)
4841 self
.__initialized
= True
4842 opts
, _
= self
.parser
.parse_known_args(self
.own_args
)
4843 self
.parsed_args
= self
.own_args
4844 for location
in opts
.config_locations
or []:
4846 if location
in self
._loaded
_paths
:
4848 self
._loaded
_paths
.add(location
)
4849 self
.append_config(shlex
.split(read_stdin('options'), comments
=True), label
='stdin')
4851 location
= os
.path
.join(directory
, expand_path(location
))
4852 if os
.path
.isdir(location
):
4853 location
= os
.path
.join(location
, 'yt-dlp.conf')
4854 if not os
.path
.exists(location
):
4855 self
.parser
.error(f
'config location {location} does not exist')
4856 self
.append_config(self
.read_file(location
), location
)
4860 label
= join_nonempty(
4861 self
.label
, 'config', f
'"{self.filename}"' if self
.filename
else '',
4863 return join_nonempty(
4864 self
.own_args
is not None and f
'{label[0].upper()}{label[1:]}: {self.hide_login_info(self.own_args)}',
4865 *(f
'\n{c}'.replace('\n', '\n| ')[1:] for c
in self
.configs
),
4869 def read_file(filename
, default
=[]):
4871 optionf
= open(filename
, 'rb')
4873 return default
# silently skip if file is not present
4875 enc
, skip
= determine_file_encoding(optionf
.read(512))
4876 optionf
.seek(skip
, io
.SEEK_SET
)
4878 enc
= None # silently skip read errors
4880 # FIXME: https://github.com/ytdl-org/youtube-dl/commit/dfe5fa49aed02cf36ba9f743b11b0903554b5e56
4881 contents
= optionf
.read().decode(enc
or preferredencoding())
4882 res
= shlex
.split(contents
, comments
=True)
4883 except Exception as err
:
4884 raise ValueError(f
'Unable to parse "{filename}": {err}')
4890 def hide_login_info(opts
):
4891 PRIVATE_OPTS
= {'-p', '--password', '-u', '--username', '--video-password', '--ap-password', '--ap-username'}
4892 eqre
= re
.compile('^(?P<key>' + ('|'.join(re
.escape(po
) for po
in PRIVATE_OPTS
)) + ')=.+$')
4897 return m
.group('key') + '=PRIVATE'
4901 opts
= list(map(_scrub_eq
, opts
))
4902 for idx
, opt
in enumerate(opts
):
4903 if opt
in PRIVATE_OPTS
and idx
+ 1 < len(opts
):
4904 opts
[idx
+ 1] = 'PRIVATE'
4907 def append_config(self
, *args
, label
=None):
4908 config
= type(self
)(self
.parser
, label
)
4909 config
._loaded
_paths
= self
._loaded
_paths
4910 if config
.init(*args
):
4911 self
.configs
.append(config
)
4915 for config
in reversed(self
.configs
):
4916 yield from config
.all_args
4917 yield from self
.parsed_args
or []
4919 def parse_known_args(self
, **kwargs
):
4920 return self
.parser
.parse_known_args(self
.all_args
, **kwargs
)
4922 def parse_args(self
):
4923 return self
.parser
.parse_args(self
.all_args
)
4926 class WebSocketsWrapper
:
4927 """Wraps websockets module to use in non-async scopes"""
4930 def __init__(self
, url
, headers
=None, connect
=True):
4931 self
.loop
= asyncio
.new_event_loop()
4932 # XXX: "loop" is deprecated
4933 self
.conn
= websockets
.connect(
4934 url
, extra_headers
=headers
, ping_interval
=None,
4935 close_timeout
=float('inf'), loop
=self
.loop
, ping_timeout
=float('inf'))
4938 atexit
.register(self
.__exit
__, None, None, None)
4940 def __enter__(self
):
4942 self
.pool
= self
.run_with_loop(self
.conn
.__aenter
__(), self
.loop
)
4945 def send(self
, *args
):
4946 self
.run_with_loop(self
.pool
.send(*args
), self
.loop
)
4948 def recv(self
, *args
):
4949 return self
.run_with_loop(self
.pool
.recv(*args
), self
.loop
)
4951 def __exit__(self
, type, value
, traceback
):
4953 return self
.run_with_loop(self
.conn
.__aexit
__(type, value
, traceback
), self
.loop
)
4956 self
._cancel
_all
_tasks
(self
.loop
)
4958 # taken from https://github.com/python/cpython/blob/3.9/Lib/asyncio/runners.py with modifications
4959 # for contributors: If there's any new library using asyncio needs to be run in non-async, move these function out of this class
4961 def run_with_loop(main
, loop
):
4962 if not asyncio
.iscoroutine(main
):
4963 raise ValueError(f
'a coroutine was expected, got {main!r}')
4966 return loop
.run_until_complete(main
)
4968 loop
.run_until_complete(loop
.shutdown_asyncgens())
4969 if hasattr(loop
, 'shutdown_default_executor'):
4970 loop
.run_until_complete(loop
.shutdown_default_executor())
4973 def _cancel_all_tasks(loop
):
4974 to_cancel
= asyncio
.all_tasks(loop
)
4979 for task
in to_cancel
:
4982 # XXX: "loop" is removed in python 3.10+
4983 loop
.run_until_complete(
4984 asyncio
.gather(*to_cancel
, loop
=loop
, return_exceptions
=True))
4986 for task
in to_cancel
:
4987 if task
.cancelled():
4989 if task
.exception() is not None:
4990 loop
.call_exception_handler({
4991 'message': 'unhandled exception during asyncio.run() shutdown',
4992 'exception': task
.exception(),
4997 def merge_headers(*dicts
):
4998 """Merge dicts of http headers case insensitively, prioritizing the latter ones"""
4999 return {k.title(): v for k, v in itertools.chain.from_iterable(map(dict.items, dicts))}
5002 def cached_method(f
):
5003 """Cache a method"""
5004 signature
= inspect
.signature(f
)
5007 def wrapper(self
, *args
, **kwargs
):
5008 bound_args
= signature
.bind(self
, *args
, **kwargs
)
5009 bound_args
.apply_defaults()
5010 key
= tuple(bound_args
.arguments
.values())[1:]
5012 cache
= vars(self
).setdefault('_cached_method__cache', {}).setdefault(f.__name__, {}
)
5013 if key
not in cache
:
5014 cache
[key
] = f(self
, *args
, **kwargs
)
5019 class classproperty
:
5020 """property access for class methods with optional caching"""
5021 def __new__(cls
, func
=None, *args
, **kwargs
):
5023 return functools
.partial(cls
, *args
, **kwargs
)
5024 return super().__new
__(cls
)
5026 def __init__(self
, func
, *, cache
=False):
5027 functools
.update_wrapper(self
, func
)
5029 self
._cache
= {} if cache
else None
5031 def __get__(self
, _
, cls
):
5032 if self
._cache
is None:
5033 return self
.func(cls
)
5034 elif cls
not in self
._cache
:
5035 self
._cache
[cls
] = self
.func(cls
)
5036 return self
._cache
[cls
]
5039 class function_with_repr
:
5040 def __init__(self
, func
, repr_
=None):
5041 functools
.update_wrapper(self
, func
)
5042 self
.func
, self
.__repr
= func
, repr_
5044 def __call__(self
, *args
, **kwargs
):
5045 return self
.func(*args
, **kwargs
)
5050 return f
'{self.func.__module__}.{self.func.__qualname__}'
5053 class Namespace(types
.SimpleNamespace
):
5054 """Immutable namespace"""
5057 return iter(self
.__dict
__.values())
5061 return self
.__dict
__.items()
5064 MEDIA_EXTENSIONS
= Namespace(
5065 common_video
=('avi', 'flv', 'mkv', 'mov', 'mp4', 'webm'),
5066 video
=('3g2', '3gp', 'f4v', 'mk3d', 'divx', 'mpg', 'ogv', 'm4v', 'wmv'),
5067 common_audio
=('aiff', 'alac', 'flac', 'm4a', 'mka', 'mp3', 'ogg', 'opus', 'wav'),
5068 audio
=('aac', 'ape', 'asf', 'f4a', 'f4b', 'm4b', 'm4p', 'm4r', 'oga', 'ogx', 'spx', 'vorbis', 'wma', 'weba'),
5069 thumbnails
=('jpg', 'png', 'webp'),
5070 storyboards
=('mhtml', ),
5071 subtitles
=('srt', 'vtt', 'ass', 'lrc'),
5072 manifests
=('f4f', 'f4m', 'm3u8', 'smil', 'mpd'),
5074 MEDIA_EXTENSIONS
.video
+= MEDIA_EXTENSIONS
.common_video
5075 MEDIA_EXTENSIONS
.audio
+= MEDIA_EXTENSIONS
.common_audio
5077 KNOWN_EXTENSIONS
= (*MEDIA_EXTENSIONS
.video
, *MEDIA_EXTENSIONS
.audio
, *MEDIA_EXTENSIONS
.manifests
)
5082 for retry in RetryManager(...):
5085 except SomeException as err:
5089 attempt
, _error
= 0, None
5091 def __init__(self
, _retries
, _error_callback
, **kwargs
):
5092 self
.retries
= _retries
or 0
5093 self
.error_callback
= functools
.partial(_error_callback
, **kwargs
)
5095 def _should_retry(self
):
5096 return self
._error
is not NO_DEFAULT
and self
.attempt
<= self
.retries
5100 if self
._error
is NO_DEFAULT
:
5105 def error(self
, value
):
5109 while self
._should
_retry
():
5110 self
.error
= NO_DEFAULT
5114 self
.error_callback(self
.error
, self
.attempt
, self
.retries
)
5117 def report_retry(e
, count
, retries
, *, sleep_func
, info
, warn
, error
=None, suffix
=None):
5118 """Utility function for reporting retries"""
5121 return error(f
'{e}. Giving up after {count - 1} retries') if count
> 1 else error(str(e
))
5126 elif isinstance(e
, ExtractorError
):
5127 e
= remove_end(str_or_none(e
.cause
) or e
.orig_msg
, '.')
5128 warn(f
'{e}. Retrying{format_field(suffix, None, " %s")} ({count}/{retries})...')
5130 delay
= float_or_none(sleep_func(n
=count
- 1)) if callable(sleep_func
) else sleep_func
5132 info(f
'Sleeping {delay:.2f} seconds ...')
5136 def make_archive_id(ie
, video_id
):
5137 ie_key
= ie
if isinstance(ie
, str) else ie
.ie_key()
5138 return f
'{ie_key.lower()} {video_id}'
5141 def truncate_string(s
, left
, right
=0):
5142 assert left
> 3 and right
>= 0
5143 if s
is None or len(s
) <= left
+ right
:
5145 return f
'{s[:left-3]}...{s[-right:] if right else ""}'
5148 def orderedSet_from_options(options
, alias_dict
, *, use_regex
=False, start
=None):
5149 assert 'all' in alias_dict
, '"all" alias is required'
5150 requested
= list(start
or [])
5152 discard
= val
.startswith('-')
5156 if val
in alias_dict
:
5157 val
= alias_dict
[val
] if not discard
else [
5158 i
[1:] if i
.startswith('-') else f
'-{i}' for i
in alias_dict
[val
]]
5159 # NB: Do not allow regex in aliases for performance
5160 requested
= orderedSet_from_options(val
, alias_dict
, start
=requested
)
5163 current
= (filter(re
.compile(val
, re
.I
).fullmatch
, alias_dict
['all']) if use_regex
5164 else [val
] if val
in alias_dict
['all'] else None)
5166 raise ValueError(val
)
5169 for item
in current
:
5170 while item
in requested
:
5171 requested
.remove(item
)
5173 requested
.extend(current
)
5175 return orderedSet(requested
)
5180 regex
= r
' *((?P<reverse>\+)?(?P<field>[a-zA-Z0-9_]+)((?P<separator>[~:])(?P<limit>.*?))?)? *$'
5182 default
= ('hidden', 'aud_or_vid', 'hasvid', 'ie_pref', 'lang', 'quality',
5183 'res', 'fps', 'hdr:12', 'vcodec:vp9.2', 'channels', 'acodec',
5184 'size', 'br', 'asr', 'proto', 'ext', 'hasaud', 'source', 'id') # These must not be aliases
5185 ytdl_default
= ('hasaud', 'lang', 'quality', 'tbr', 'filesize', 'vbr',
5186 'height', 'width', 'proto', 'vext', 'abr', 'aext',
5187 'fps', 'fs_approx', 'source', 'id')
5190 'vcodec': {'type': 'ordered', 'regex': True,
5191 'order': ['av0?1', 'vp0?9.2', 'vp0?9', '[hx]265|he?vc?', '[hx]264|avc', 'vp0?8', 'mp4v|h263', 'theora', '', None, 'none']},
5192 'acodec': {'type': 'ordered', 'regex': True,
5193 'order': ['[af]lac', 'wav|aiff', 'opus', 'vorbis|ogg', 'aac', 'mp?4a?', 'mp3', 'ac-?4', 'e-?a?c-?3', 'ac-?3', 'dts', '', None, 'none']},
5194 'hdr': {'type': 'ordered', 'regex': True, 'field': 'dynamic_range',
5195 'order': ['dv', '(hdr)?12', r
'(hdr)?10\+', '(hdr)?10', 'hlg', '', 'sdr', None]},
5196 'proto': {'type': 'ordered', 'regex': True, 'field': 'protocol',
5197 'order': ['(ht|f)tps', '(ht|f)tp$', 'm3u8.*', '.*dash', 'websocket_frag', 'rtmpe?', '', 'mms|rtsp', 'ws|websocket', 'f4']},
5198 'vext': {'type': 'ordered', 'field': 'video_ext',
5199 'order': ('mp4', 'mov', 'webm', 'flv', '', 'none'),
5200 'order_free': ('webm', 'mp4', 'mov', 'flv', '', 'none')},
5201 'aext': {'type': 'ordered', 'regex': True, 'field': 'audio_ext',
5202 'order': ('m4a', 'aac', 'mp3', 'ogg', 'opus', 'web[am]', '', 'none'),
5203 'order_free': ('ogg', 'opus', 'web[am]', 'mp3', 'm4a', 'aac', '', 'none')},
5204 'hidden': {'visible': False, 'forced': True, 'type': 'extractor', 'max': -1000}
,
5205 'aud_or_vid': {'visible': False, 'forced': True, 'type': 'multiple',
5206 'field': ('vcodec', 'acodec'),
5207 'function': lambda it
: int(any(v
!= 'none' for v
in it
))},
5208 'ie_pref': {'priority': True, 'type': 'extractor'}
,
5209 'hasvid': {'priority': True, 'field': 'vcodec', 'type': 'boolean', 'not_in_list': ('none',)}
,
5210 'hasaud': {'field': 'acodec', 'type': 'boolean', 'not_in_list': ('none',)}
,
5211 'lang': {'convert': 'float', 'field': 'language_preference', 'default': -1}
,
5212 'quality': {'convert': 'float', 'default': -1}
,
5213 'filesize': {'convert': 'bytes'}
,
5214 'fs_approx': {'convert': 'bytes', 'field': 'filesize_approx'}
,
5215 'id': {'convert': 'string', 'field': 'format_id'}
,
5216 'height': {'convert': 'float_none'}
,
5217 'width': {'convert': 'float_none'}
,
5218 'fps': {'convert': 'float_none'}
,
5219 'channels': {'convert': 'float_none', 'field': 'audio_channels'}
,
5220 'tbr': {'convert': 'float_none'}
,
5221 'vbr': {'convert': 'float_none'}
,
5222 'abr': {'convert': 'float_none'}
,
5223 'asr': {'convert': 'float_none'}
,
5224 'source': {'convert': 'float', 'field': 'source_preference', 'default': -1}
,
5226 'codec': {'type': 'combined', 'field': ('vcodec', 'acodec')}
,
5227 'br': {'type': 'multiple', 'field': ('tbr', 'vbr', 'abr'), 'convert': 'float_none',
5228 'function': lambda it
: next(filter(None, it
), None)},
5229 'size': {'type': 'multiple', 'field': ('filesize', 'fs_approx'), 'convert': 'bytes',
5230 'function': lambda it
: next(filter(None, it
), None)},
5231 'ext': {'type': 'combined', 'field': ('vext', 'aext')}
,
5232 'res': {'type': 'multiple', 'field': ('height', 'width'),
5233 'function': lambda it
: (lambda l
: min(l
) if l
else 0)(tuple(filter(None, it
)))},
5235 # Actual field names
5236 'format_id': {'type': 'alias', 'field': 'id'}
,
5237 'preference': {'type': 'alias', 'field': 'ie_pref'}
,
5238 'language_preference': {'type': 'alias', 'field': 'lang'}
,
5239 'source_preference': {'type': 'alias', 'field': 'source'}
,
5240 'protocol': {'type': 'alias', 'field': 'proto'}
,
5241 'filesize_approx': {'type': 'alias', 'field': 'fs_approx'}
,
5242 'audio_channels': {'type': 'alias', 'field': 'channels'}
,
5245 'dimension': {'type': 'alias', 'field': 'res', 'deprecated': True}
,
5246 'resolution': {'type': 'alias', 'field': 'res', 'deprecated': True}
,
5247 'extension': {'type': 'alias', 'field': 'ext', 'deprecated': True}
,
5248 'bitrate': {'type': 'alias', 'field': 'br', 'deprecated': True}
,
5249 'total_bitrate': {'type': 'alias', 'field': 'tbr', 'deprecated': True}
,
5250 'video_bitrate': {'type': 'alias', 'field': 'vbr', 'deprecated': True}
,
5251 'audio_bitrate': {'type': 'alias', 'field': 'abr', 'deprecated': True}
,
5252 'framerate': {'type': 'alias', 'field': 'fps', 'deprecated': True}
,
5253 'filesize_estimate': {'type': 'alias', 'field': 'size', 'deprecated': True}
,
5254 'samplerate': {'type': 'alias', 'field': 'asr', 'deprecated': True}
,
5255 'video_ext': {'type': 'alias', 'field': 'vext', 'deprecated': True}
,
5256 'audio_ext': {'type': 'alias', 'field': 'aext', 'deprecated': True}
,
5257 'video_codec': {'type': 'alias', 'field': 'vcodec', 'deprecated': True}
,
5258 'audio_codec': {'type': 'alias', 'field': 'acodec', 'deprecated': True}
,
5259 'video': {'type': 'alias', 'field': 'hasvid', 'deprecated': True}
,
5260 'has_video': {'type': 'alias', 'field': 'hasvid', 'deprecated': True}
,
5261 'audio': {'type': 'alias', 'field': 'hasaud', 'deprecated': True}
,
5262 'has_audio': {'type': 'alias', 'field': 'hasaud', 'deprecated': True}
,
5263 'extractor': {'type': 'alias', 'field': 'ie_pref', 'deprecated': True}
,
5264 'extractor_preference': {'type': 'alias', 'field': 'ie_pref', 'deprecated': True}
,
5267 def __init__(self
, ydl
, field_preference
):
5270 self
.evaluate_params(self
.ydl
.params
, field_preference
)
5271 if ydl
.params
.get('verbose'):
5272 self
.print_verbose_info(self
.ydl
.write_debug
)
5274 def _get_field_setting(self
, field
, key
):
5275 if field
not in self
.settings
:
5276 if key
in ('forced', 'priority'):
5278 self
.ydl
.deprecated_feature(f
'Using arbitrary fields ({field}) for format sorting is '
5279 'deprecated and may be removed in a future version')
5280 self
.settings
[field
] = {}
5281 propObj
= self
.settings
[field
]
5282 if key
not in propObj
:
5283 type = propObj
.get('type')
5285 default
= 'preference' if type == 'extractor' else (field
,) if type in ('combined', 'multiple') else field
5286 elif key
== 'convert':
5287 default
= 'order' if type == 'ordered' else 'float_string' if field
else 'ignore'
5289 default
= {'type': 'field', 'visible': True, 'order': [], 'not_in_list': (None,)}
.get(key
, None)
5290 propObj
[key
] = default
5293 def _resolve_field_value(self
, field
, value
, convertNone
=False):
5298 value
= value
.lower()
5299 conversion
= self
._get
_field
_setting
(field
, 'convert')
5300 if conversion
== 'ignore':
5302 if conversion
== 'string':
5304 elif conversion
== 'float_none':
5305 return float_or_none(value
)
5306 elif conversion
== 'bytes':
5307 return parse_bytes(value
)
5308 elif conversion
== 'order':
5309 order_list
= (self
._use
_free
_order
and self
._get
_field
_setting
(field
, 'order_free')) or self
._get
_field
_setting
(field
, 'order')
5310 use_regex
= self
._get
_field
_setting
(field
, 'regex')
5311 list_length
= len(order_list
)
5312 empty_pos
= order_list
.index('') if '' in order_list
else list_length
+ 1
5313 if use_regex
and value
is not None:
5314 for i
, regex
in enumerate(order_list
):
5315 if regex
and re
.match(regex
, value
):
5316 return list_length
- i
5317 return list_length
- empty_pos
# not in list
5318 else: # not regex or value = None
5319 return list_length
- (order_list
.index(value
) if value
in order_list
else empty_pos
)
5321 if value
.isnumeric():
5324 self
.settings
[field
]['convert'] = 'string'
5327 def evaluate_params(self
, params
, sort_extractor
):
5328 self
._use
_free
_order
= params
.get('prefer_free_formats', False)
5329 self
._sort
_user
= params
.get('format_sort', [])
5330 self
._sort
_extractor
= sort_extractor
5332 def add_item(field
, reverse
, closest
, limit_text
):
5333 field
= field
.lower()
5334 if field
in self
._order
:
5336 self
._order
.append(field
)
5337 limit
= self
._resolve
_field
_value
(field
, limit_text
)
5340 'closest': False if limit
is None else closest
,
5341 'limit_text': limit_text
,
5343 if field
in self
.settings
:
5344 self
.settings
[field
].update(data
)
5346 self
.settings
[field
] = data
5349 tuple(field
for field
in self
.default
if self
._get
_field
_setting
(field
, 'forced'))
5350 + (tuple() if params
.get('format_sort_force', False)
5351 else tuple(field
for field
in self
.default
if self
._get
_field
_setting
(field
, 'priority')))
5352 + tuple(self
._sort
_user
) + tuple(sort_extractor
) + self
.default
)
5354 for item
in sort_list
:
5355 match
= re
.match(self
.regex
, item
)
5357 raise ExtractorError('Invalid format sort string "%s" given by extractor' % item
)
5358 field
= match
.group('field')
5361 if self
._get
_field
_setting
(field
, 'type') == 'alias':
5362 alias
, field
= field
, self
._get
_field
_setting
(field
, 'field')
5363 if self
._get
_field
_setting
(alias
, 'deprecated'):
5364 self
.ydl
.deprecated_feature(f
'Format sorting alias {alias} is deprecated and may '
5365 f
'be removed in a future version. Please use {field} instead')
5366 reverse
= match
.group('reverse') is not None
5367 closest
= match
.group('separator') == '~'
5368 limit_text
= match
.group('limit')
5370 has_limit
= limit_text
is not None
5371 has_multiple_fields
= self
._get
_field
_setting
(field
, 'type') == 'combined'
5372 has_multiple_limits
= has_limit
and has_multiple_fields
and not self
._get
_field
_setting
(field
, 'same_limit')
5374 fields
= self
._get
_field
_setting
(field
, 'field') if has_multiple_fields
else (field
,)
5375 limits
= limit_text
.split(':') if has_multiple_limits
else (limit_text
,) if has_limit
else tuple()
5376 limit_count
= len(limits
)
5377 for (i
, f
) in enumerate(fields
):
5378 add_item(f
, reverse
, closest
,
5379 limits
[i
] if i
< limit_count
5380 else limits
[0] if has_limit
and not has_multiple_limits
5383 def print_verbose_info(self
, write_debug
):
5385 write_debug('Sort order given by user: %s' % ', '.join(self
._sort
_user
))
5386 if self
._sort
_extractor
:
5387 write_debug('Sort order given by extractor: %s' % ', '.join(self
._sort
_extractor
))
5388 write_debug('Formats sorted by: %s' % ', '.join(['%s%s%s' % (
5389 '+' if self
._get
_field
_setting
(field
, 'reverse') else '', field
,
5390 '%s%s(%s)' % ('~' if self
._get
_field
_setting
(field
, 'closest') else ':',
5391 self
._get
_field
_setting
(field
, 'limit_text'),
5392 self
._get
_field
_setting
(field
, 'limit'))
5393 if self
._get
_field
_setting
(field
, 'limit_text') is not None else '')
5394 for field
in self
._order
if self
._get
_field
_setting
(field
, 'visible')]))
5396 def _calculate_field_preference_from_value(self
, format
, field
, type, value
):
5397 reverse
= self
._get
_field
_setting
(field
, 'reverse')
5398 closest
= self
._get
_field
_setting
(field
, 'closest')
5399 limit
= self
._get
_field
_setting
(field
, 'limit')
5401 if type == 'extractor':
5402 maximum
= self
._get
_field
_setting
(field
, 'max')
5403 if value
is None or (maximum
is not None and value
>= maximum
):
5405 elif type == 'boolean':
5406 in_list
= self
._get
_field
_setting
(field
, 'in_list')
5407 not_in_list
= self
._get
_field
_setting
(field
, 'not_in_list')
5408 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
5409 elif type == 'ordered':
5410 value
= self
._resolve
_field
_value
(field
, value
, True)
5412 # try to convert to number
5413 val_num
= float_or_none(value
, default
=self
._get
_field
_setting
(field
, 'default'))
5414 is_num
= self
._get
_field
_setting
(field
, 'convert') != 'string' and val_num
is not None
5418 return ((-10, 0) if value
is None
5419 else (1, value
, 0) if not is_num
# if a field has mixed strings and numbers, strings are sorted higher
5420 else (0, -abs(value
- limit
), value
- limit
if reverse
else limit
- value
) if closest
5421 else (0, value
, 0) if not reverse
and (limit
is None or value
<= limit
)
5422 else (0, -value
, 0) if limit
is None or (reverse
and value
== limit
) or value
> limit
5423 else (-1, value
, 0))
5425 def _calculate_field_preference(self
, format
, field
):
5426 type = self
._get
_field
_setting
(field
, 'type') # extractor, boolean, ordered, field, multiple
5427 get_value
= lambda f
: format
.get(self
._get
_field
_setting
(f
, 'field'))
5428 if type == 'multiple':
5429 type = 'field' # Only 'field' is allowed in multiple for now
5430 actual_fields
= self
._get
_field
_setting
(field
, 'field')
5432 value
= self
._get
_field
_setting
(field
, 'function')(get_value(f
) for f
in actual_fields
)
5434 value
= get_value(field
)
5435 return self
._calculate
_field
_preference
_from
_value
(format
, field
, type, value
)
5437 def calculate_preference(self
, format
):
5438 # Determine missing protocol
5439 if not format
.get('protocol'):
5440 format
['protocol'] = determine_protocol(format
)
5442 # Determine missing ext
5443 if not format
.get('ext') and 'url' in format
:
5444 format
['ext'] = determine_ext(format
['url'])
5445 if format
.get('vcodec') == 'none':
5446 format
['audio_ext'] = format
['ext'] if format
.get('acodec') != 'none' else 'none'
5447 format
['video_ext'] = 'none'
5449 format
['video_ext'] = format
['ext']
5450 format
['audio_ext'] = 'none'
5451 # if format.get('preference') is None and format.get('ext') in ('f4f', 'f4m'): # Not supported?
5452 # format['preference'] = -1000
5454 if format
.get('preference') is None and format
.get('ext') == 'flv' and re
.match('[hx]265|he?vc?', format
.get('vcodec') or ''):
5455 # HEVC-over-FLV is out-of-spec by FLV's original spec
5456 # ref. https://trac.ffmpeg.org/ticket/6389
5457 # ref. https://github.com/yt-dlp/yt-dlp/pull/5821
5458 format
['preference'] = -100
5460 # Determine missing bitrates
5461 if format
.get('vcodec') == 'none':
5463 if format
.get('acodec') == 'none':
5465 if not format
.get('vbr') and format
.get('vcodec') != 'none':
5466 format
['vbr'] = try_call(lambda: format
['tbr'] - format
['abr']) or None
5467 if not format
.get('abr') and format
.get('acodec') != 'none':
5468 format
['abr'] = try_call(lambda: format
['tbr'] - format
['vbr']) or None
5469 if not format
.get('tbr'):
5470 format
['tbr'] = try_call(lambda: format
['vbr'] + format
['abr']) or None
5472 return tuple(self
._calculate
_field
_preference
(format
, field
) for field
in self
._order
)
5477 def __init__(self
, ydl
=None):
5480 def debug(self
, message
):
5482 self
._ydl
.write_debug(message
)
5484 def info(self
, message
):
5486 self
._ydl
.to_screen(message
)
5488 def warning(self
, message
, *, once
=False):
5490 self
._ydl
.report_warning(message
, once
)
5492 def error(self
, message
, *, is_error
=True):
5494 self
._ydl
.report_error(message
, is_error
=is_error
)
5496 def stdout(self
, message
):
5498 self
._ydl
.to_stdout(message
)
5500 def stderr(self
, message
):
5502 self
._ydl
.to_stderr(message
)