43 import xml
.etree
.ElementTree
45 from . import traversal
47 from ..compat
import functools
# isort: split
48 from ..compat
import (
49 compat_etree_fromstring
,
51 compat_HTMLParseError
,
54 from ..dependencies
import xattr
56 __name__
= __name__
.rsplit('.', 1)[0] # noqa: A001: Pretend to be the parent module
58 # This is not clearly defined otherwise
59 compiled_regex_type
= type(re
.compile(''))
70 ENGLISH_MONTH_NAMES
= [
71 'January', 'February', 'March', 'April', 'May', 'June',
72 'July', 'August', 'September', 'October', 'November', 'December']
75 'en': ENGLISH_MONTH_NAMES
,
77 'janvier', 'février', 'mars', 'avril', 'mai', 'juin',
78 'juillet', 'août', 'septembre', 'octobre', 'novembre', 'décembre'],
79 # these follow the genitive grammatical case (dopełniacz)
80 # some websites might be using nominative, which will require another month list
81 # https://en.wikibooks.org/wiki/Polish/Noun_cases
82 'pl': ['stycznia', 'lutego', 'marca', 'kwietnia', 'maja', 'czerwca',
83 'lipca', 'sierpnia', 'września', 'października', 'listopada', 'grudnia'],
86 # From https://github.com/python/cpython/blob/3.11/Lib/email/_parseaddr.py#L36-L42
88 'UT': 0, 'UTC': 0, 'GMT': 0, 'Z': 0,
89 'AST': -4, 'ADT': -3, # Atlantic (used in Canada)
90 'EST': -5, 'EDT': -4, # Eastern
91 'CST': -6, 'CDT': -5, # Central
92 'MST': -7, 'MDT': -6, # Mountain
93 'PST': -8, 'PDT': -7, # Pacific
96 # needed for sanitizing filenames in restricted mode
97 ACCENT_CHARS
= dict(zip('ÂÃÄÀÁÅÆÇÈÉÊËÌÍÎÏÐÑÒÓÔÕÖŐØŒÙÚÛÜŰÝÞßàáâãäåæçèéêëìíîïðñòóôõöőøœùúûüűýþÿ',
98 itertools
.chain('AAAAAA', ['AE'], 'CEEEEIIIIDNOOOOOOO', ['OE'], 'UUUUUY', ['TH', 'ss'],
99 'aaaaaa', ['ae'], 'ceeeeiiiionooooooo', ['oe'], 'uuuuuy', ['th'], 'y')))
129 '%Y-%m-%d %H:%M:%S.%f',
130 '%Y-%m-%d %H:%M:%S:%f',
133 '%Y-%m-%dT%H:%M:%SZ',
134 '%Y-%m-%dT%H:%M:%S.%fZ',
135 '%Y-%m-%dT%H:%M:%S.%f0Z',
137 '%Y-%m-%dT%H:%M:%S.%f',
140 '%b %d %Y at %H:%M:%S',
142 '%B %d %Y at %H:%M:%S',
146 DATE_FORMATS_DAY_FIRST
= list(DATE_FORMATS
)
147 DATE_FORMATS_DAY_FIRST
.extend([
158 DATE_FORMATS_MONTH_FIRST
= list(DATE_FORMATS
)
159 DATE_FORMATS_MONTH_FIRST
.extend([
167 PACKED_CODES_RE
= r
"}\('(.+)',(\d+),(\d+),'([^']+)'\.split\('\|'\)"
168 JSON_LD_RE
= r
'(?is)<script[^>]+type=(["\']?
)application
/ld\
+json\
1[^
>]*>\s
*(?P
<json_ld
>{.+?}|\
[.+?\
])\s
*</script
>'
170 NUMBER_RE = r'\d
+(?
:\
.\d
+)?
'
174 def preferredencoding():
175 """Get preferred encoding.
177 Returns the best encoding scheme for the system, based on
178 locale.getpreferredencoding() and some further tweaks.
181 pref = locale.getpreferredencoding()
189 def write_json_file(obj, fn):
190 """ Encode obj as JSON and write it to fn, atomically if possible """
192 tf = tempfile.NamedTemporaryFile(
193 prefix=f'{os.path.basename(fn)}
.', dir=os.path.dirname(fn),
194 suffix='.tmp
', delete=False, mode='w
', encoding='utf
-8')
198 json.dump(obj, tf, ensure_ascii=False)
199 if sys.platform == 'win32
':
200 # Need to remove existing file on Windows, else os.rename raises
201 # WindowsError or FileExistsError.
202 with contextlib.suppress(OSError):
204 with contextlib.suppress(OSError):
207 os.chmod(tf.name, 0o666 & ~mask)
208 os.rename(tf.name, fn)
210 with contextlib.suppress(OSError):
215 def find_xpath_attr(node, xpath, key, val=None):
216 """ Find the xpath xpath[@key=val] """
217 assert re.match(r'^
[a
-zA
-Z_
-]+$
', key)
218 expr = xpath + (f'[@{key}
]' if val is None else f"[@{key}='{val}
']")
219 return node.find(expr)
221 # On python2.6 the xml.etree.ElementTree.Element methods don't support
222 # the namespace parameter
225 def xpath_with_ns(path
, ns_map
):
226 components
= [c
.split(':') for c
in path
.split('/')]
230 replaced
.append(c
[0])
233 replaced
.append(f
'{{{ns_map[ns]}}}{tag}')
234 return '/'.join(replaced
)
237 def xpath_element(node
, xpath
, name
=None, fatal
=False, default
=NO_DEFAULT
):
238 def _find_xpath(xpath
):
239 return node
.find(xpath
)
241 if isinstance(xpath
, str):
242 n
= _find_xpath(xpath
)
250 if default
is not NO_DEFAULT
:
253 name
= xpath
if name
is None else name
254 raise ExtractorError(f
'Could not find XML element {name}')
260 def xpath_text(node
, xpath
, name
=None, fatal
=False, default
=NO_DEFAULT
):
261 n
= xpath_element(node
, xpath
, name
, fatal
=fatal
, default
=default
)
262 if n
is None or n
== default
:
265 if default
is not NO_DEFAULT
:
268 name
= xpath
if name
is None else name
269 raise ExtractorError(f
'Could not find XML element\'s text {name}')
275 def xpath_attr(node
, xpath
, key
, name
=None, fatal
=False, default
=NO_DEFAULT
):
276 n
= find_xpath_attr(node
, xpath
, key
)
278 if default
is not NO_DEFAULT
:
281 name
= f
'{xpath}[@{key}]' if name
is None else name
282 raise ExtractorError(f
'Could not find XML attribute {name}')
288 def get_element_by_id(id, html
, **kwargs
):
289 """Return the content of the tag with the specified ID in the passed HTML document"""
290 return get_element_by_attribute('id', id, html
, **kwargs
)
293 def get_element_html_by_id(id, html
, **kwargs
):
294 """Return the html of the tag with the specified ID in the passed HTML document"""
295 return get_element_html_by_attribute('id', id, html
, **kwargs
)
298 def get_element_by_class(class_name
, html
):
299 """Return the content of the first tag with the specified class in the passed HTML document"""
300 retval
= get_elements_by_class(class_name
, html
)
301 return retval
[0] if retval
else None
304 def get_element_html_by_class(class_name
, html
):
305 """Return the html of the first tag with the specified class in the passed HTML document"""
306 retval
= get_elements_html_by_class(class_name
, html
)
307 return retval
[0] if retval
else None
310 def get_element_by_attribute(attribute
, value
, html
, **kwargs
):
311 retval
= get_elements_by_attribute(attribute
, value
, html
, **kwargs
)
312 return retval
[0] if retval
else None
315 def get_element_html_by_attribute(attribute
, value
, html
, **kargs
):
316 retval
= get_elements_html_by_attribute(attribute
, value
, html
, **kargs
)
317 return retval
[0] if retval
else None
320 def get_elements_by_class(class_name
, html
, **kargs
):
321 """Return the content of all tags with the specified class in the passed HTML document as a list"""
322 return get_elements_by_attribute(
323 'class', rf
'[^\'"]*(?<=[\'"\s]){re.escape(class_name)}(?=[\'"\s])[^\'"]*',
324 html
, escape_value
=False)
327 def get_elements_html_by_class(class_name
, html
):
328 """Return the html of all tags with the specified class in the passed HTML document as a list"""
329 return get_elements_html_by_attribute(
330 'class', rf
'[^\'"]*(?<=[\'"\s]){re.escape(class_name)}(?=[\'"\s])[^\'"]*',
331 html
, escape_value
=False)
334 def get_elements_by_attribute(*args
, **kwargs
):
335 """Return the content of the tag with the specified attribute in the passed HTML document"""
336 return [content
for content
, _
in get_elements_text_and_html_by_attribute(*args
, **kwargs
)]
339 def get_elements_html_by_attribute(*args
, **kwargs
):
340 """Return the html of the tag with the specified attribute in the passed HTML document"""
341 return [whole
for _
, whole
in get_elements_text_and_html_by_attribute(*args
, **kwargs
)]
344 def get_elements_text_and_html_by_attribute(attribute
, value
, html
, *, tag
=r
'[\w:.-]+', escape_value
=True):
346 Return the text (content) and the html (whole) of the tag with the specified
347 attribute in the passed HTML document
352 quote
= '' if re
.match(r
'''[\s"'`=<>]''', value
) else '?'
354 value
= re
.escape(value
) if escape_value
else value
356 partial_element_re
= rf
'''(?x)
358 (?:\s(?:[^>"']|"[^"]*"|'[^']*')*)?
359 \s{re.escape(attribute)}\s*=\s*(?P<_q>['"]{quote})(?-x:{value})(?P=_q)
362 for m
in re
.finditer(partial_element_re
, html
):
363 content
, whole
= get_element_text_and_html_by_tag(m
.group('tag'), html
[m
.start():])
366 unescapeHTML(re
.sub(r
'^(?P<q>["\'])(?P
<content
>.*)(?P
=q
)$
', r'\g
<content
>', content, flags=re.DOTALL)),
371 class HTMLBreakOnClosingTagParser(html.parser.HTMLParser):
373 HTML parser which raises HTMLBreakOnClosingTagException upon reaching the
374 closing tag for the first opening tag it has encountered, and can be used
378 class HTMLBreakOnClosingTagException(Exception):
382 self.tagstack = collections.deque()
383 html.parser.HTMLParser.__init__(self)
388 def __exit__(self, *_):
392 # handle_endtag does not return upon raising HTMLBreakOnClosingTagException,
393 # so data remains buffered; we no longer have any interest in it, thus
394 # override this method to discard it
397 def handle_starttag(self, tag, _):
398 self.tagstack.append(tag)
400 def handle_endtag(self, tag):
401 if not self.tagstack:
402 raise compat_HTMLParseError('no tags
in the stack
')
404 inner_tag = self.tagstack.pop()
408 raise compat_HTMLParseError(f'matching opening tag
for closing {tag} tag
not found
')
409 if not self.tagstack:
410 raise self.HTMLBreakOnClosingTagException
413 # XXX: This should be far less strict
414 def get_element_text_and_html_by_tag(tag, html):
416 For the first element with the specified tag in the passed HTML document
417 return its' content (text
) and the whole
element (html
)
419 def find_or_raise(haystack, needle, exc):
421 return haystack.index(needle)
424 closing_tag = f'</{tag}>'
425 whole_start = find_or_raise(
426 html, f'<{tag}', compat_HTMLParseError(f'opening {tag} tag not found'))
427 content_start = find_or_raise(
428 html[whole_start:], '>', compat_HTMLParseError(f'malformed opening {tag} tag'))
429 content_start += whole_start + 1
430 with HTMLBreakOnClosingTagParser() as parser:
431 parser.feed(html[whole_start:content_start])
432 if not parser.tagstack or parser.tagstack[0] != tag:
433 raise compat_HTMLParseError(f'parser did not match opening {tag} tag')
434 offset = content_start
435 while offset < len(html):
436 next_closing_tag_start = find_or_raise(
437 html[offset:], closing_tag,
438 compat_HTMLParseError(f'closing {tag} tag not found'))
439 next_closing_tag_end = next_closing_tag_start + len(closing_tag)
441 parser.feed(html[offset:offset + next_closing_tag_end])
442 offset += next_closing_tag_end
443 except HTMLBreakOnClosingTagParser.HTMLBreakOnClosingTagException:
444 return html[content_start:offset + next_closing_tag_start], \
445 html[whole_start:offset + next_closing_tag_end]
446 raise compat_HTMLParseError('unexpected end of html')
449 class HTMLAttributeParser(html.parser.HTMLParser):
450 """Trivial HTML parser to gather the attributes
for a single element
"""
454 html.parser.HTMLParser.__init__(self)
456 def handle_starttag(self, tag, attrs):
457 self.attrs = dict(attrs)
458 raise compat_HTMLParseError('done')
461 class HTMLListAttrsParser(html.parser.HTMLParser):
462 """HTML parser to gather the attributes
for the elements of a
list"""
465 html.parser.HTMLParser.__init__(self)
469 def handle_starttag(self, tag, attrs):
470 if tag == 'li' and self._level == 0:
471 self.items.append(dict(attrs))
474 def handle_endtag(self, tag):
478 def extract_attributes(html_element):
479 """Given a string
for an HTML element such
as
481 a
="foo" B
="bar" c
="&98;az" d
=boz
482 empty
= noval entity
="&"
485 Decode
and return a dictionary of attributes
.
487 'a': 'foo', 'b': 'bar', c
: 'baz', d
: 'boz',
488 'empty': '', 'noval': None, 'entity': '&',
489 'sq': '"', 'dq': '\''
492 parser = HTMLAttributeParser()
493 with contextlib.suppress(compat_HTMLParseError):
494 parser.feed(html_element)
499 def parse_list(webpage):
500 """Given a string
for an series of HTML
<li
> elements
,
501 return a dictionary of their attributes
"""
502 parser = HTMLListAttrsParser()
508 def clean_html(html):
509 """Clean an HTML snippet into a readable string
"""
511 if html is None: # Convenience for sanitizing descriptions etc.
514 html = re.sub(r'\s+', ' ', html)
515 html = re.sub(r'(?u)\s?<\s?br\s?/?\s?>\s?', '\n', html)
516 html = re.sub(r'(?u)<\s?/\s?p\s?>\s?<\s?p[^>]*>', '\n', html)
518 html = re.sub('<.*?>', '', html)
519 # Replace html entities
520 html = unescapeHTML(html)
524 class LenientJSONDecoder(json.JSONDecoder):
526 def __init__(self, *args, transform_source=None, ignore_extra=False, close_objects=0, **kwargs):
527 self.transform_source, self.ignore_extra = transform_source, ignore_extra
528 self._close_attempts = 2 * close_objects
529 super().__init__(*args, **kwargs)
532 def _close_object(err):
533 doc = err.doc[:err.pos]
534 # We need to add comma first to get the correct error message
535 if err.msg.startswith('Expecting \',\''):
537 elif not doc.endswith(','):
540 if err.msg.startswith('Expecting property name'):
541 return doc[:-1] + '}'
542 elif err.msg.startswith('Expecting value'):
543 return doc[:-1] + ']'
546 if self.transform_source:
547 s = self.transform_source(s)
548 for attempt in range(self._close_attempts + 1):
550 if self.ignore_extra:
551 return self.raw_decode(s.lstrip())[0]
552 return super().decode(s)
553 except json.JSONDecodeError as e:
556 elif attempt < self._close_attempts:
557 s = self._close_object(e)
560 raise type(e)(f'{e.msg} in {s[e.pos - 10:e.pos + 10]!r}', s, e.pos)
561 assert False, 'Too many attempts to decode JSON'
564 def sanitize_open(filename, open_mode):
565 """Try to
open the given filename
, and slightly tweak it
if this fails
.
567 Attempts to
open the given filename
. If this fails
, it tries to change
568 the filename slightly
, step by step
, until it
's either able to open it
569 or it fails and raises a final exception, like the standard open()
572 It returns the tuple (stream, definitive_file_name).
575 if sys.platform == 'win32
':
578 # stdout may be any IO stream, e.g. when using contextlib.redirect_stdout
579 with contextlib.suppress(io.UnsupportedOperation):
580 msvcrt.setmode(sys.stdout.fileno(), os.O_BINARY)
581 return (sys.stdout.buffer if hasattr(sys.stdout, 'buffer') else sys.stdout, filename)
583 for attempt in range(2):
586 if sys.platform == 'win32
':
587 # FIXME: An exclusive lock also locks the file from being read.
588 # Since windows locks are mandatory, don't lock the
file on
windows (for now
).
589 # Ref: https://github.com/yt-dlp/yt-dlp/issues/3124
590 raise LockingUnsupportedError
591 stream
= locked_file(filename
, open_mode
, block
=False).__enter
__()
593 stream
= open(filename
, open_mode
)
594 return stream
, filename
595 except OSError as err
:
596 if attempt
or err
.errno
in (errno
.EACCES
,):
598 old_filename
, filename
= filename
, sanitize_path(filename
)
599 if old_filename
== filename
:
603 def timeconvert(timestr
):
604 """Convert RFC 2822 defined time string into system timestamp"""
606 timetuple
= email
.utils
.parsedate_tz(timestr
)
607 if timetuple
is not None:
608 timestamp
= email
.utils
.mktime_tz(timetuple
)
612 def sanitize_filename(s
, restricted
=False, is_id
=NO_DEFAULT
):
613 """Sanitizes a string so it could be used as part of a filename.
614 @param restricted Use a stricter subset of allowed characters
615 @param is_id Whether this is an ID that should be kept unchanged if possible.
616 If unset, yt-dlp's new sanitization rules are in effect
621 def replace_insane(char
):
622 if restricted
and char
in ACCENT_CHARS
:
623 return ACCENT_CHARS
[char
]
624 elif not restricted
and char
== '\n':
626 elif is_id
is NO_DEFAULT
and not restricted
and char
in '"*:<>?|/\\':
627 # Replace with their full-width unicode counterparts
628 return {'/': '\u29F8', '\\': '\u29f9'}
.get(char
, chr(ord(char
) + 0xfee0))
629 elif char
== '?' or ord(char
) < 32 or ord(char
) == 127:
632 return '' if restricted
else '\''
634 return '\0_\0-' if restricted
else '\0 \0-'
635 elif char
in '\\/|*<>':
637 if restricted
and (char
in '!&\'()[]{}$;`^,#' or char
.isspace() or ord(char
) > 127):
638 return '' if unicodedata
.category(char
)[0] in 'CM' else '\0_'
641 # Replace look-alike Unicode glyphs
642 if restricted
and (is_id
is NO_DEFAULT
or not is_id
):
643 s
= unicodedata
.normalize('NFKC', s
)
644 s
= re
.sub(r
'[0-9]+(?::[0-9]+)+', lambda m
: m
.group(0).replace(':', '_'), s
) # Handle timestamps
645 result
= ''.join(map(replace_insane
, s
))
646 if is_id
is NO_DEFAULT
:
647 result
= re
.sub(r
'(\0.)(?:(?=\1)..)+', r
'\1', result
) # Remove repeated substitute chars
648 STRIP_RE
= r
'(?:\0.|[ _-])*'
649 result
= re
.sub(f
'^\0.{STRIP_RE}|{STRIP_RE}\0.$', '', result
) # Remove substitute chars from start/end
650 result
= result
.replace('\0', '') or '_'
653 while '__' in result
:
654 result
= result
.replace('__', '_')
655 result
= result
.strip('_')
656 # Common case of "Foreign band name - English song title"
657 if restricted
and result
.startswith('-_'):
659 if result
.startswith('-'):
660 result
= '_' + result
[len('-'):]
661 result
= result
.lstrip('.')
667 def sanitize_path(s
, force
=False):
668 """Sanitizes and normalizes path on Windows"""
669 # XXX: this handles drive relative paths (c:sth) incorrectly
670 if sys
.platform
== 'win32':
672 drive_or_unc
, _
= os
.path
.splitdrive(s
)
678 norm_path
= os
.path
.normpath(remove_start(s
, drive_or_unc
)).split(os
.path
.sep
)
682 path_part
if path_part
in ['.', '..'] else re
.sub(r
'(?:[/<>:"\|\\?\*]|[\s.]$)', '#', path_part
)
683 for path_part
in norm_path
]
685 sanitized_path
.insert(0, drive_or_unc
+ os
.path
.sep
)
686 elif force
and s
and s
[0] == os
.path
.sep
:
687 sanitized_path
.insert(0, os
.path
.sep
)
688 # TODO: Fix behavioral differences <3.12
689 # The workaround using `normpath` only superficially passes tests
690 # Ref: https://github.com/python/cpython/pull/100351
691 return os
.path
.normpath(os
.path
.join(*sanitized_path
))
694 def sanitize_url(url
, *, scheme
='http'):
695 # Prepend protocol-less URLs with `http:` scheme in order to mitigate
696 # the number of unwanted failures due to missing protocol
699 elif url
.startswith('//'):
700 return f
'{scheme}:{url}'
701 # Fix some common typos seen so far
703 # https://github.com/ytdl-org/youtube-dl/issues/15649
704 (r
'^httpss://', r
'https://'),
705 # https://bx1.be/lives/direct-tv/
706 (r
'^rmtp([es]?)://', r
'rtmp\1://'),
708 for mistake
, fixup
in COMMON_TYPOS
:
709 if re
.match(mistake
, url
):
710 return re
.sub(mistake
, fixup
, url
)
714 def extract_basic_auth(url
):
715 parts
= urllib
.parse
.urlsplit(url
)
716 if parts
.username
is None:
718 url
= urllib
.parse
.urlunsplit(parts
._replace
(netloc
=(
719 parts
.hostname
if parts
.port
is None
720 else f
'{parts.hostname}:{parts.port}')))
721 auth_payload
= base64
.b64encode(
722 ('{}:{}'.format(parts
.username
, parts
.password
or '')).encode())
723 return url
, f
'Basic {auth_payload.decode()}'
727 """Expand shell variables and ~"""
728 return os
.path
.expandvars(compat_expanduser(s
))
731 def orderedSet(iterable
, *, lazy
=False):
732 """Remove all duplicates from the input iterable"""
734 seen
= [] # Do not use set since the items can be unhashable
740 return _iter() if lazy
else list(_iter())
743 def _htmlentity_transform(entity_with_semicolon
):
744 """Transforms an HTML entity to a character."""
745 entity
= entity_with_semicolon
[:-1]
747 # Known non-numeric HTML entity
748 if entity
in html
.entities
.name2codepoint
:
749 return chr(html
.entities
.name2codepoint
[entity
])
751 # TODO: HTML5 allows entities without a semicolon.
752 # E.g. 'Éric' should be decoded as 'Éric'.
753 if entity_with_semicolon
in html
.entities
.html5
:
754 return html
.entities
.html5
[entity_with_semicolon
]
756 mobj
= re
.match(r
'#(x[0-9a-fA-F]+|[0-9]+)', entity
)
758 numstr
= mobj
.group(1)
759 if numstr
.startswith('x'):
761 numstr
= f
'0{numstr}'
764 # See https://github.com/ytdl-org/youtube-dl/issues/7518
765 with contextlib
.suppress(ValueError):
766 return chr(int(numstr
, base
))
768 # Unknown entity in name, return its literal representation
775 assert isinstance(s
, str)
778 r
'&([^&;]+;)', lambda m
: _htmlentity_transform(m
.group(1)), s
)
781 def escapeHTML(text
):
784 .replace('&', '&')
785 .replace('<', '<')
786 .replace('>', '>')
787 .replace('"', '"')
788 .replace("'", ''')
792 class netrc_from_content(netrc
.netrc
):
793 def __init__(self
, content
):
794 self
.hosts
, self
.macros
= {}, {}
795 with io
.StringIO(content
) as stream
:
796 self
._parse
('-', stream
, False)
799 class Popen(subprocess
.Popen
):
800 if sys
.platform
== 'win32':
801 _startupinfo
= subprocess
.STARTUPINFO()
802 _startupinfo
.dwFlags |
= subprocess
.STARTF_USESHOWWINDOW
807 def _fix_pyinstaller_ld_path(env
):
808 """Restore LD_LIBRARY_PATH when using PyInstaller
809 Ref: https://github.com/pyinstaller/pyinstaller/blob/develop/doc/runtime-information.rst#ld_library_path--libpath-considerations
810 https://github.com/yt-dlp/yt-dlp/issues/4573
812 if not hasattr(sys
, '_MEIPASS'):
816 orig
= env
.get(f
'{key}_ORIG')
822 _fix('LD_LIBRARY_PATH') # Linux
823 _fix('DYLD_LIBRARY_PATH') # macOS
825 def __init__(self
, args
, *remaining
, env
=None, text
=False, shell
=False, **kwargs
):
827 env
= os
.environ
.copy()
828 self
._fix
_pyinstaller
_ld
_path
(env
)
830 self
.__text
_mode
= kwargs
.get('encoding') or kwargs
.get('errors') or text
or kwargs
.get('universal_newlines')
832 kwargs
['universal_newlines'] = True # For 3.6 compatibility
833 kwargs
.setdefault('encoding', 'utf-8')
834 kwargs
.setdefault('errors', 'replace')
836 if shell
and compat_os_name
== 'nt' and kwargs
.get('executable') is None:
837 if not isinstance(args
, str):
838 args
= shell_quote(args
, shell
=True)
840 # Set variable for `cmd.exe` newline escaping (see `utils.shell_quote`)
842 args
= f
'{self.__comspec()} /Q /S /D /V:OFF /E:ON /C "{args}"'
844 super().__init
__(args
, *remaining
, env
=env
, shell
=shell
, **kwargs
, startupinfo
=self
._startupinfo
)
847 comspec
= os
.environ
.get('ComSpec') or os
.path
.join(
848 os
.environ
.get('SystemRoot', ''), 'System32', 'cmd.exe')
849 if os
.path
.isabs(comspec
):
851 raise FileNotFoundError('shell not found: neither %ComSpec% nor %SystemRoot% is set')
853 def communicate_or_kill(self
, *args
, **kwargs
):
855 return self
.communicate(*args
, **kwargs
)
856 except BaseException
: # Including KeyboardInterrupt
857 self
.kill(timeout
=None)
860 def kill(self
, *, timeout
=0):
863 self
.wait(timeout
=timeout
)
866 def run(cls
, *args
, timeout
=None, **kwargs
):
867 with cls(*args
, **kwargs
) as proc
:
868 default
= '' if proc
.__text
_mode
else b
''
869 stdout
, stderr
= proc
.communicate_or_kill(timeout
=timeout
)
870 return stdout
or default
, stderr
or default
, proc
.returncode
873 def encodeArgument(s
):
874 # Legacy code that uses byte strings
875 # Uncomment the following line after fixing all post processors
876 # assert isinstance(s, str), 'Internal error: %r should be of type %r, is %r' % (s, str, type(s))
877 return s
if isinstance(s
, str) else s
.decode('ascii')
880 _timetuple
= collections
.namedtuple('Time', ('hours', 'minutes', 'seconds', 'milliseconds'))
883 def timetuple_from_msec(msec
):
884 secs
, msec
= divmod(msec
, 1000)
885 mins
, secs
= divmod(secs
, 60)
886 hrs
, mins
= divmod(mins
, 60)
887 return _timetuple(hrs
, mins
, secs
, msec
)
890 def formatSeconds(secs
, delim
=':', msec
=False):
891 time
= timetuple_from_msec(secs
* 1000)
893 ret
= '%d%s%02d%s%02d' % (time
.hours
, delim
, time
.minutes
, delim
, time
.seconds
)
895 ret
= '%d%s%02d' % (time
.minutes
, delim
, time
.seconds
)
897 ret
= '%d' % time
.seconds
898 return '%s.%03d' % (ret
, time
.milliseconds
) if msec
else ret
901 def bug_reports_message(before
=';'):
902 from ..update
import REPOSITORY
904 msg
= (f
'please report this issue on https://github.com/{REPOSITORY}/issues?q= , '
905 'filling out the appropriate issue template. Confirm you are on the latest version using yt-dlp -U')
907 before
= before
.rstrip()
908 if not before
or before
.endswith(('.', '!', '?')):
909 msg
= msg
[0].title() + msg
[1:]
911 return (before
+ ' ' if before
else '') + msg
914 class YoutubeDLError(Exception):
915 """Base exception for YoutubeDL errors."""
918 def __init__(self
, msg
=None):
921 elif self
.msg
is None:
922 self
.msg
= type(self
).__name
__
923 super().__init
__(self
.msg
)
926 class ExtractorError(YoutubeDLError
):
927 """Error during info extraction."""
929 def __init__(self
, msg
, tb
=None, expected
=False, cause
=None, video_id
=None, ie
=None):
930 """ tb, if given, is the original traceback (so that it can be printed out).
931 If expected is set, this is a normal error message and most likely not a bug in yt-dlp.
933 from ..networking
.exceptions
import network_exceptions
934 if sys
.exc_info()[0] in network_exceptions
:
937 self
.orig_msg
= str(msg
)
939 self
.expected
= expected
941 self
.video_id
= video_id
943 self
.exc_info
= sys
.exc_info() # preserve original exception
944 if isinstance(self
.exc_info
[1], ExtractorError
):
945 self
.exc_info
= self
.exc_info
[1].exc_info
946 super().__init
__(self
.__msg
)
951 format_field(self
.ie
, None, '[%s] '),
952 format_field(self
.video_id
, None, '%s: '),
954 format_field(self
.cause
, None, ' (caused by %r)'),
955 '' if self
.expected
else bug_reports_message()))
957 def format_traceback(self
):
958 return join_nonempty(
959 self
.traceback
and ''.join(traceback
.format_tb(self
.traceback
)),
960 self
.cause
and ''.join(traceback
.format_exception(None, self
.cause
, self
.cause
.__traceback
__)[1:]),
963 def __setattr__(self
, name
, value
):
964 super().__setattr
__(name
, value
)
965 if getattr(self
, 'msg', None) and name
not in ('msg', 'args'):
966 self
.msg
= self
.__msg
or type(self
).__name
__
967 self
.args
= (self
.msg
, ) # Cannot be property
970 class UnsupportedError(ExtractorError
):
971 def __init__(self
, url
):
973 f
'Unsupported URL: {url}', expected
=True)
977 class RegexNotFoundError(ExtractorError
):
978 """Error when a regex didn't match"""
982 class GeoRestrictedError(ExtractorError
):
983 """Geographic restriction Error exception.
985 This exception may be thrown when a video is not available from your
986 geographic location due to geographic restrictions imposed by a website.
989 def __init__(self
, msg
, countries
=None, **kwargs
):
990 kwargs
['expected'] = True
991 super().__init
__(msg
, **kwargs
)
992 self
.countries
= countries
995 class UserNotLive(ExtractorError
):
996 """Error when a channel/user is not live"""
998 def __init__(self
, msg
=None, **kwargs
):
999 kwargs
['expected'] = True
1000 super().__init
__(msg
or 'The channel is not currently live', **kwargs
)
1003 class DownloadError(YoutubeDLError
):
1004 """Download Error exception.
1006 This exception may be thrown by FileDownloader objects if they are not
1007 configured to continue on errors. They will contain the appropriate
1011 def __init__(self
, msg
, exc_info
=None):
1012 """ exc_info, if given, is the original exception that caused the trouble (as returned by sys.exc_info()). """
1013 super().__init
__(msg
)
1014 self
.exc_info
= exc_info
1017 class EntryNotInPlaylist(YoutubeDLError
):
1018 """Entry not in playlist exception.
1020 This exception will be thrown by YoutubeDL when a requested entry
1021 is not found in the playlist info_dict
1023 msg
= 'Entry not found in info'
1026 class SameFileError(YoutubeDLError
):
1027 """Same File exception.
1029 This exception will be thrown by FileDownloader objects if they detect
1030 multiple files would have to be downloaded to the same file on disk.
1032 msg
= 'Fixed output name but more than one file to download'
1034 def __init__(self
, filename
=None):
1035 if filename
is not None:
1036 self
.msg
+= f
': {filename}'
1037 super().__init
__(self
.msg
)
1040 class PostProcessingError(YoutubeDLError
):
1041 """Post Processing exception.
1043 This exception may be raised by PostProcessor's .run() method to
1044 indicate an error in the postprocessing task.
1048 class DownloadCancelled(YoutubeDLError
):
1049 """ Exception raised when the download queue should be interrupted """
1050 msg
= 'The download was cancelled'
1053 class ExistingVideoReached(DownloadCancelled
):
1054 """ --break-on-existing triggered """
1055 msg
= 'Encountered a video that is already in the archive, stopping due to --break-on-existing'
1058 class RejectedVideoReached(DownloadCancelled
):
1059 """ --break-match-filter triggered """
1060 msg
= 'Encountered a video that did not match filter, stopping due to --break-match-filter'
1063 class MaxDownloadsReached(DownloadCancelled
):
1064 """ --max-downloads limit has been reached. """
1065 msg
= 'Maximum number of downloads reached, stopping due to --max-downloads'
1068 class ReExtractInfo(YoutubeDLError
):
1069 """ Video info needs to be re-extracted. """
1071 def __init__(self
, msg
, expected
=False):
1072 super().__init
__(msg
)
1073 self
.expected
= expected
1076 class ThrottledDownload(ReExtractInfo
):
1077 """ Download speed below --throttled-rate. """
1078 msg
= 'The download speed is below throttle limit'
1081 super().__init
__(self
.msg
, expected
=False)
1084 class UnavailableVideoError(YoutubeDLError
):
1085 """Unavailable Format exception.
1087 This exception will be thrown when a video is requested
1088 in a format that is not available for that video.
1090 msg
= 'Unable to download video'
1092 def __init__(self
, err
=None):
1094 self
.msg
+= f
': {err}'
1095 super().__init
__(self
.msg
)
1098 class ContentTooShortError(YoutubeDLError
):
1099 """Content Too Short exception.
1101 This exception may be raised by FileDownloader objects when a file they
1102 download is too small for what the server announced first, indicating
1103 the connection was probably interrupted.
1106 def __init__(self
, downloaded
, expected
):
1107 super().__init
__(f
'Downloaded {downloaded} bytes, expected {expected} bytes')
1109 self
.downloaded
= downloaded
1110 self
.expected
= expected
1113 class XAttrMetadataError(YoutubeDLError
):
1114 def __init__(self
, code
=None, msg
='Unknown error'):
1115 super().__init
__(msg
)
1119 # Parsing code and msg
1120 if (self
.code
in (errno
.ENOSPC
, errno
.EDQUOT
)
1121 or 'No space left' in self
.msg
or 'Disk quota exceeded' in self
.msg
):
1122 self
.reason
= 'NO_SPACE'
1123 elif self
.code
== errno
.E2BIG
or 'Argument list too long' in self
.msg
:
1124 self
.reason
= 'VALUE_TOO_LONG'
1126 self
.reason
= 'NOT_SUPPORTED'
1129 class XAttrUnavailableError(YoutubeDLError
):
1133 def is_path_like(f
):
1134 return isinstance(f
, (str, bytes, os
.PathLike
))
1137 def extract_timezone(date_str
, default
=None):
1140 ^.{8,}? # >=8 char non-TZ prefix, if present
1141 (?P<tz>Z| # just the UTC Z, or
1142 (?:(?<=.\b\d{4}|\b\d{2}:\d\d)| # preceded by 4 digits or hh:mm or
1143 (?<!.\b[a-zA-Z]{3}|[a-zA-Z]{4}|..\b\d\d)) # not preceded by 3 alpha word or >= 4 alpha or 2 digits
1144 [ ]? # optional space
1145 (?P<sign>\+|-) # +/-
1146 (?P<hours>[0-9]{2}):?(?P<minutes>[0-9]{2}) # hh[:]mm
1152 m
= re
.search(r
'\d{1,2}:\d{1,2}(?:\.\d+)?(?P<tz>\s*[A-Z]+)$', date_str
)
1153 timezone
= TIMEZONE_NAMES
.get(m
and m
.group('tz').strip())
1154 if timezone
is not None:
1155 date_str
= date_str
[:-len(m
.group('tz'))]
1156 timezone
= dt
.timedelta(hours
=timezone
)
1158 date_str
= date_str
[:-len(m
.group('tz'))]
1160 sign
= 1 if m
.group('sign') == '+' else -1
1161 timezone
= dt
.timedelta(
1162 hours
=sign
* int(m
.group('hours')),
1163 minutes
=sign
* int(m
.group('minutes')))
1165 if timezone
is None and default
is not NO_DEFAULT
:
1166 timezone
= default
or dt
.timedelta()
1168 return timezone
, date_str
1171 def parse_iso8601(date_str
, delimiter
='T', timezone
=None):
1172 """ Return a UNIX timestamp from the given date """
1174 if date_str
is None:
1177 date_str
= re
.sub(r
'\.[0-9]+', '', date_str
)
1179 timezone
, date_str
= extract_timezone(date_str
, timezone
)
1181 with contextlib
.suppress(ValueError, TypeError):
1182 date_format
= f
'%Y-%m-%d{delimiter}%H:%M:%S'
1183 dt_
= dt
.datetime
.strptime(date_str
, date_format
) - timezone
1184 return calendar
.timegm(dt_
.timetuple())
1187 def date_formats(day_first
=True):
1188 return DATE_FORMATS_DAY_FIRST
if day_first
else DATE_FORMATS_MONTH_FIRST
1191 def unified_strdate(date_str
, day_first
=True):
1192 """Return a string with the date in the format YYYYMMDD"""
1194 if date_str
is None:
1198 date_str
= date_str
.replace(',', ' ')
1199 # Remove AM/PM + timezone
1200 date_str
= re
.sub(r
'(?i)\s*(?:AM|PM)(?:\s+[A-Z]+)?', '', date_str
)
1201 _
, date_str
= extract_timezone(date_str
)
1203 for expression
in date_formats(day_first
):
1204 with contextlib
.suppress(ValueError):
1205 upload_date
= dt
.datetime
.strptime(date_str
, expression
).strftime('%Y%m%d')
1206 if upload_date
is None:
1207 timetuple
= email
.utils
.parsedate_tz(date_str
)
1209 with contextlib
.suppress(ValueError):
1210 upload_date
= dt
.datetime(*timetuple
[:6]).strftime('%Y%m%d')
1211 if upload_date
is not None:
1212 return str(upload_date
)
1215 def unified_timestamp(date_str
, day_first
=True):
1216 if not isinstance(date_str
, str):
1219 date_str
= re
.sub(r
'\s+', ' ', re
.sub(
1220 r
'(?i)[,|]|(mon|tues?|wed(nes)?|thu(rs)?|fri|sat(ur)?)(day)?', '', date_str
))
1222 pm_delta
= 12 if re
.search(r
'(?i)PM', date_str
) else 0
1223 timezone
, date_str
= extract_timezone(date_str
)
1225 # Remove AM/PM + timezone
1226 date_str
= re
.sub(r
'(?i)\s*(?:AM|PM)(?:\s+[A-Z]+)?', '', date_str
)
1228 # Remove unrecognized timezones from ISO 8601 alike timestamps
1229 m
= re
.search(r
'\d{1,2}:\d{1,2}(?:\.\d+)?(?P<tz>\s*[A-Z]+)$', date_str
)
1231 date_str
= date_str
[:-len(m
.group('tz'))]
1233 # Python only supports microseconds, so remove nanoseconds
1234 m
= re
.search(r
'^([0-9]{4,}-[0-9]{1,2}-[0-9]{1,2}T[0-9]{1,2}:[0-9]{1,2}:[0-9]{1,2}\.[0-9]{6})[0-9]+$', date_str
)
1236 date_str
= m
.group(1)
1238 for expression
in date_formats(day_first
):
1239 with contextlib
.suppress(ValueError):
1240 dt_
= dt
.datetime
.strptime(date_str
, expression
) - timezone
+ dt
.timedelta(hours
=pm_delta
)
1241 return calendar
.timegm(dt_
.timetuple())
1243 timetuple
= email
.utils
.parsedate_tz(date_str
)
1245 return calendar
.timegm(timetuple
) + pm_delta
* 3600 - timezone
.total_seconds()
1248 def determine_ext(url
, default_ext
='unknown_video'):
1249 if url
is None or '.' not in url
:
1251 guess
= url
.partition('?')[0].rpartition('.')[2]
1252 if re
.match(r
'^[A-Za-z0-9]+$', guess
):
1254 # Try extract ext from URLs like http://example.com/foo/bar.mp4/?download
1255 elif guess
.rstrip('/') in KNOWN_EXTENSIONS
:
1256 return guess
.rstrip('/')
1261 def subtitles_filename(filename
, sub_lang
, sub_format
, expected_real_ext
=None):
1262 return replace_extension(filename
, sub_lang
+ '.' + sub_format
, expected_real_ext
)
1265 def datetime_from_str(date_str
, precision
='auto', format
='%Y%m%d'):
1267 Return a datetime object from a string.
1269 (now|today|yesterday|DATE)([+-]\d+(microsecond|second|minute|hour|day|week|month|year)s?)?
1271 @param format strftime format of DATE
1272 @param precision Round the datetime object: auto|microsecond|second|minute|hour|day
1273 auto: round to the unit provided in date_str (if applicable).
1275 auto_precision
= False
1276 if precision
== 'auto':
1277 auto_precision
= True
1278 precision
= 'microsecond'
1279 today
= datetime_round(dt
.datetime
.now(dt
.timezone
.utc
), precision
)
1280 if date_str
in ('now', 'today'):
1282 if date_str
== 'yesterday':
1283 return today
- dt
.timedelta(days
=1)
1285 r
'(?P<start>.+)(?P<sign>[+-])(?P<time>\d+)(?P<unit>microsecond|second|minute|hour|day|week|month|year)s?',
1287 if match
is not None:
1288 start_time
= datetime_from_str(match
.group('start'), precision
, format
)
1289 time
= int(match
.group('time')) * (-1 if match
.group('sign') == '-' else 1)
1290 unit
= match
.group('unit')
1291 if unit
== 'month' or unit
== 'year':
1292 new_date
= datetime_add_months(start_time
, time
* 12 if unit
== 'year' else time
)
1298 delta
= dt
.timedelta(**{unit + 's': time}
)
1299 new_date
= start_time
+ delta
1301 return datetime_round(new_date
, unit
)
1304 return datetime_round(dt
.datetime
.strptime(date_str
, format
), precision
)
1307 def date_from_str(date_str
, format
='%Y%m%d', strict
=False):
1309 Return a date object from a string using datetime_from_str
1311 @param strict Restrict allowed patterns to "YYYYMMDD" and
1312 (now|today|yesterday)(-\d+(day|week|month|year)s?)?
1314 if strict
and not re
.fullmatch(r
'\d{8}|(now|today|yesterday)(-\d+(day|week|month|year)s?)?', date_str
):
1315 raise ValueError(f
'Invalid date format "{date_str}"')
1316 return datetime_from_str(date_str
, precision
='microsecond', format
=format
).date()
1319 def datetime_add_months(dt_
, months
):
1320 """Increment/Decrement a datetime object by months."""
1321 month
= dt_
.month
+ months
- 1
1322 year
= dt_
.year
+ month
// 12
1323 month
= month
% 12 + 1
1324 day
= min(dt_
.day
, calendar
.monthrange(year
, month
)[1])
1325 return dt_
.replace(year
, month
, day
)
1328 def datetime_round(dt_
, precision
='day'):
1330 Round a datetime object's time to a specific precision
1332 if precision
== 'microsecond':
1341 roundto
= lambda x
, n
: ((x
+ n
/ 2) // n
) * n
1342 timestamp
= roundto(calendar
.timegm(dt_
.timetuple()), unit_seconds
[precision
])
1343 return dt
.datetime
.fromtimestamp(timestamp
, dt
.timezone
.utc
)
1346 def hyphenate_date(date_str
):
1348 Convert a date in 'YYYYMMDD' format to 'YYYY-MM-DD' format"""
1349 match
= re
.match(r
'^(\d\d\d\d)(\d\d)(\d\d)$', date_str
)
1350 if match
is not None:
1351 return '-'.join(match
.groups())
1357 """Represents a time interval between two dates"""
1359 def __init__(self
, start
=None, end
=None):
1360 """start and end must be strings in the format accepted by date"""
1361 if start
is not None:
1362 self
.start
= date_from_str(start
, strict
=True)
1364 self
.start
= dt
.datetime
.min.date()
1366 self
.end
= date_from_str(end
, strict
=True)
1368 self
.end
= dt
.datetime
.max.date()
1369 if self
.start
> self
.end
:
1370 raise ValueError(f
'Date range: "{self}" , the start date must be before the end date')
1374 """Returns a range that only contains the given day"""
1375 return cls(day
, day
)
1377 def __contains__(self
, date
):
1378 """Check if the date is in the range"""
1379 if not isinstance(date
, dt
.date
):
1380 date
= date_from_str(date
)
1381 return self
.start
<= date
<= self
.end
1384 return f
'{__name__}.{type(self).__name__}({self.start.isoformat()!r}, {self.end.isoformat()!r})'
1387 return f
'{self.start} to {self.end}'
1389 def __eq__(self
, other
):
1390 return (isinstance(other
, DateRange
)
1391 and self
.start
== other
.start
and self
.end
== other
.end
)
1395 def system_identifier():
1396 python_implementation
= platform
.python_implementation()
1397 if python_implementation
== 'PyPy' and hasattr(sys
, 'pypy_version_info'):
1398 python_implementation
+= ' version %d.%d.%d' % sys
.pypy_version_info
[:3]
1400 with contextlib
.suppress(OSError): # We may not have access to the executable
1401 libc_ver
= platform
.libc_ver()
1403 return 'Python {} ({} {} {}) - {} ({}{})'.format(
1404 platform
.python_version(),
1405 python_implementation
,
1407 platform
.architecture()[0],
1408 platform
.platform(),
1409 ssl
.OPENSSL_VERSION
,
1410 format_field(join_nonempty(*libc_ver
, delim
=' '), None, ', %s'),
1415 def get_windows_version():
1416 """ Get Windows version. returns () if it's not running on Windows """
1417 if compat_os_name
== 'nt':
1418 return version_tuple(platform
.win32_ver()[1])
1423 def write_string(s
, out
=None, encoding
=None):
1424 assert isinstance(s
, str)
1425 out
= out
or sys
.stderr
1426 # `sys.stderr` might be `None` (Ref: https://github.com/pyinstaller/pyinstaller/pull/7217)
1430 if compat_os_name
== 'nt' and supports_terminal_sequences(out
):
1431 s
= re
.sub(r
'([\r\n]+)', r
' \1', s
)
1433 enc
, buffer = None, out
1434 # `mode` might be `None` (Ref: https://github.com/yt-dlp/yt-dlp/issues/8816)
1435 if 'b' in (getattr(out
, 'mode', None) or ''):
1436 enc
= encoding
or preferredencoding()
1437 elif hasattr(out
, 'buffer'):
1439 enc
= encoding
or getattr(out
, 'encoding', None) or preferredencoding()
1441 buffer.write(s
.encode(enc
, 'ignore') if enc
else s
)
1445 # TODO: Use global logger
1446 def deprecation_warning(msg
, *, printer
=None, stacklevel
=0, **kwargs
):
1447 from .. import _IN_CLI
1449 if msg
in deprecation_warning
._cache
:
1451 deprecation_warning
._cache
.add(msg
)
1453 return printer(f
'{msg}{bug_reports_message()}', **kwargs
)
1454 return write_string(f
'ERROR: {msg}{bug_reports_message()}\n', **kwargs
)
1457 warnings
.warn(DeprecationWarning(msg
), stacklevel
=stacklevel
+ 3)
1460 deprecation_warning
._cache
= set()
1463 def bytes_to_intlist(bs
):
1466 if isinstance(bs
[0], int): # Python 3
1469 return [ord(c
) for c
in bs
]
1472 def intlist_to_bytes(xs
):
1475 return struct
.pack('%dB' % len(xs
), *xs
)
1478 class LockingUnsupportedError(OSError):
1479 msg
= 'File locking is not supported'
1482 super().__init
__(self
.msg
)
1485 # Cross-platform file locking
1486 if sys
.platform
== 'win32':
1488 import ctypes
.wintypes
1491 class OVERLAPPED(ctypes
.Structure
):
1493 ('Internal', ctypes
.wintypes
.LPVOID
),
1494 ('InternalHigh', ctypes
.wintypes
.LPVOID
),
1495 ('Offset', ctypes
.wintypes
.DWORD
),
1496 ('OffsetHigh', ctypes
.wintypes
.DWORD
),
1497 ('hEvent', ctypes
.wintypes
.HANDLE
),
1500 kernel32
= ctypes
.WinDLL('kernel32')
1501 LockFileEx
= kernel32
.LockFileEx
1502 LockFileEx
.argtypes
= [
1503 ctypes
.wintypes
.HANDLE
, # hFile
1504 ctypes
.wintypes
.DWORD
, # dwFlags
1505 ctypes
.wintypes
.DWORD
, # dwReserved
1506 ctypes
.wintypes
.DWORD
, # nNumberOfBytesToLockLow
1507 ctypes
.wintypes
.DWORD
, # nNumberOfBytesToLockHigh
1508 ctypes
.POINTER(OVERLAPPED
), # Overlapped
1510 LockFileEx
.restype
= ctypes
.wintypes
.BOOL
1511 UnlockFileEx
= kernel32
.UnlockFileEx
1512 UnlockFileEx
.argtypes
= [
1513 ctypes
.wintypes
.HANDLE
, # hFile
1514 ctypes
.wintypes
.DWORD
, # dwReserved
1515 ctypes
.wintypes
.DWORD
, # nNumberOfBytesToLockLow
1516 ctypes
.wintypes
.DWORD
, # nNumberOfBytesToLockHigh
1517 ctypes
.POINTER(OVERLAPPED
), # Overlapped
1519 UnlockFileEx
.restype
= ctypes
.wintypes
.BOOL
1520 whole_low
= 0xffffffff
1521 whole_high
= 0x7fffffff
1523 def _lock_file(f
, exclusive
, block
):
1524 overlapped
= OVERLAPPED()
1525 overlapped
.Offset
= 0
1526 overlapped
.OffsetHigh
= 0
1527 overlapped
.hEvent
= 0
1528 f
._lock
_file
_overlapped
_p
= ctypes
.pointer(overlapped
)
1530 if not LockFileEx(msvcrt
.get_osfhandle(f
.fileno()),
1531 (0x2 if exclusive
else 0x0) |
(0x0 if block
else 0x1),
1532 0, whole_low
, whole_high
, f
._lock
_file
_overlapped
_p
):
1533 # NB: No argument form of "ctypes.FormatError" does not work on PyPy
1534 raise BlockingIOError(f
'Locking file failed: {ctypes.FormatError(ctypes.GetLastError())!r}')
1536 def _unlock_file(f
):
1537 assert f
._lock
_file
_overlapped
_p
1538 handle
= msvcrt
.get_osfhandle(f
.fileno())
1539 if not UnlockFileEx(handle
, 0, whole_low
, whole_high
, f
._lock
_file
_overlapped
_p
):
1540 raise OSError(f
'Unlocking file failed: {ctypes.FormatError()!r}')
1546 def _lock_file(f
, exclusive
, block
):
1547 flags
= fcntl
.LOCK_EX
if exclusive
else fcntl
.LOCK_SH
1549 flags |
= fcntl
.LOCK_NB
1551 fcntl
.flock(f
, flags
)
1552 except BlockingIOError
:
1554 except OSError: # AOSP does not have flock()
1555 fcntl
.lockf(f
, flags
)
1557 def _unlock_file(f
):
1558 with contextlib
.suppress(OSError):
1559 return fcntl
.flock(f
, fcntl
.LOCK_UN
)
1560 with contextlib
.suppress(OSError):
1561 return fcntl
.lockf(f
, fcntl
.LOCK_UN
) # AOSP does not have flock()
1562 return fcntl
.flock(f
, fcntl
.LOCK_UN | fcntl
.LOCK_NB
) # virtiofs needs LOCK_NB on unlocking
1566 def _lock_file(f
, exclusive
, block
):
1567 raise LockingUnsupportedError
1569 def _unlock_file(f
):
1570 raise LockingUnsupportedError
1576 def __init__(self
, filename
, mode
, block
=True, encoding
=None):
1577 if mode
not in {'r', 'rb', 'a', 'ab', 'w', 'wb'}
:
1578 raise NotImplementedError(mode
)
1579 self
.mode
, self
.block
= mode
, block
1581 writable
= any(f
in mode
for f
in 'wax+')
1582 readable
= any(f
in mode
for f
in 'r+')
1583 flags
= functools
.reduce(operator
.ior
, (
1584 getattr(os
, 'O_CLOEXEC', 0), # UNIX only
1585 getattr(os
, 'O_BINARY', 0), # Windows only
1586 getattr(os
, 'O_NOINHERIT', 0), # Windows only
1587 os
.O_CREAT
if writable
else 0, # O_TRUNC only after locking
1588 os
.O_APPEND
if 'a' in mode
else 0,
1589 os
.O_EXCL
if 'x' in mode
else 0,
1590 os
.O_RDONLY
if not writable
else os
.O_RDWR
if readable
else os
.O_WRONLY
,
1593 self
.f
= os
.fdopen(os
.open(filename
, flags
, 0o666), mode
, encoding
=encoding
)
1595 def __enter__(self
):
1596 exclusive
= 'r' not in self
.mode
1598 _lock_file(self
.f
, exclusive
, self
.block
)
1603 if 'w' in self
.mode
:
1606 except OSError as e
:
1608 errno
.ESPIPE
, # Illegal seek - expected for FIFO
1609 errno
.EINVAL
, # Invalid argument - expected for /dev/null
1618 _unlock_file(self
.f
)
1622 def __exit__(self
, *_
):
1631 def __getattr__(self
, attr
):
1632 return getattr(self
.f
, attr
)
1639 def get_filesystem_encoding():
1640 encoding
= sys
.getfilesystemencoding()
1641 return encoding
if encoding
is not None else 'utf-8'
1644 _WINDOWS_QUOTE_TRANS
= str.maketrans({'"': R'\"'}
)
1645 _CMD_QUOTE_TRANS
= str.maketrans({
1646 # Keep quotes balanced by replacing them with `""` instead of `\\"`
1648 # These require an env-variable `=` containing `"^\n\n"` (set in `utils.Popen`)
1649 # `=` should be unique since variables containing `=` cannot be set using cmd
1652 # Use zero length variable replacement so `%` doesn't get expanded
1653 # `cd` is always set as long as extensions are enabled (`/E:ON` in `utils.Popen`)
1658 def shell_quote(args
, *, shell
=False):
1659 args
= list(variadic(args
))
1661 if compat_os_name
!= 'nt':
1662 return shlex
.join(args
)
1664 trans
= _CMD_QUOTE_TRANS
if shell
else _WINDOWS_QUOTE_TRANS
1666 s
if re
.fullmatch(r
'[\w#$*\-+./:?@\\]+', s
, re
.ASCII
)
1667 else re
.sub(r
'(\\+)("|$)', r
'\1\1\2', s
).translate(trans
).join('""')
1671 def smuggle_url(url
, data
):
1672 """ Pass additional data in a URL for internal use. """
1674 url
, idata
= unsmuggle_url(url
, {})
1676 sdata
= urllib
.parse
.urlencode(
1677 {'__youtubedl_smuggle': json.dumps(data)}
)
1678 return url
+ '#' + sdata
1681 def unsmuggle_url(smug_url
, default
=None):
1682 if '#__youtubedl_smuggle' not in smug_url
:
1683 return smug_url
, default
1684 url
, _
, sdata
= smug_url
.rpartition('#')
1685 jsond
= urllib
.parse
.parse_qs(sdata
)['__youtubedl_smuggle'][0]
1686 data
= json
.loads(jsond
)
1690 def format_decimal_suffix(num
, fmt
='%d%s', *, factor
=1000):
1691 """ Formats numbers with decimal sufixes like K, M, etc """
1692 num
, factor
= float_or_none(num
), float(factor
)
1693 if num
is None or num
< 0:
1695 POSSIBLE_SUFFIXES
= 'kMGTPEZY'
1696 exponent
= 0 if num
== 0 else min(int(math
.log(num
, factor
)), len(POSSIBLE_SUFFIXES
))
1697 suffix
= ['', *POSSIBLE_SUFFIXES
][exponent
]
1699 suffix
= {'k': 'Ki', '': ''}
.get(suffix
, f
'{suffix}i')
1700 converted
= num
/ (factor
** exponent
)
1701 return fmt
% (converted
, suffix
)
1704 def format_bytes(bytes):
1705 return format_decimal_suffix(bytes, '%.2f%sB', factor
=1024) or 'N/A'
1708 def lookup_unit_table(unit_table
, s
, strict
=False):
1709 num_re
= NUMBER_RE
if strict
else NUMBER_RE
.replace(R
'\.', '[,.]')
1710 units_re
= '|'.join(re
.escape(u
) for u
in unit_table
)
1711 m
= (re
.fullmatch
if strict
else re
.match
)(
1712 rf
'(?P<num>{num_re})\s*(?P<unit>{units_re})\b', s
)
1716 num
= float(m
.group('num').replace(',', '.'))
1717 mult
= unit_table
[m
.group('unit')]
1718 return round(num
* mult
)
1722 """Parse a string indicating a byte quantity into an integer"""
1723 return lookup_unit_table(
1724 {u: 1024**i for i, u in enumerate(['', *'KMGTPEZY'])}
,
1725 s
.upper(), strict
=True)
1728 def parse_filesize(s
):
1732 # The lower-case forms are of course incorrect and unofficial,
1733 # but we support those too
1750 'megabytes': 1000 ** 2,
1751 'mebibytes': 1024 ** 2,
1757 'gigabytes': 1000 ** 3,
1758 'gibibytes': 1024 ** 3,
1764 'terabytes': 1000 ** 4,
1765 'tebibytes': 1024 ** 4,
1771 'petabytes': 1000 ** 5,
1772 'pebibytes': 1024 ** 5,
1778 'exabytes': 1000 ** 6,
1779 'exbibytes': 1024 ** 6,
1785 'zettabytes': 1000 ** 7,
1786 'zebibytes': 1024 ** 7,
1792 'yottabytes': 1000 ** 8,
1793 'yobibytes': 1024 ** 8,
1796 return lookup_unit_table(_UNIT_TABLE
, s
)
1803 s
= re
.sub(r
'^[^\d]+\s', '', s
).strip()
1805 if re
.match(r
'^[\d,.]+$', s
):
1806 return str_to_int(s
)
1819 ret
= lookup_unit_table(_UNIT_TABLE
, s
)
1823 mobj
= re
.match(r
'([\d,.]+)(?:$|\s)', s
)
1825 return str_to_int(mobj
.group(1))
1828 def parse_resolution(s
, *, lenient
=False):
1833 mobj
= re
.search(r
'(?P<w>\d+)\s*[xX×,]\s*(?P<h>\d+)', s
)
1835 mobj
= re
.search(r
'(?<![a-zA-Z0-9])(?P<w>\d+)\s*[xX×,]\s*(?P<h>\d+)(?![a-zA-Z0-9])', s
)
1838 'width': int(mobj
.group('w')),
1839 'height': int(mobj
.group('h')),
1842 mobj
= re
.search(r
'(?<![a-zA-Z0-9])(\d+)[pPiI](?![a-zA-Z0-9])', s
)
1844 return {'height': int(mobj.group(1))}
1846 mobj
= re
.search(r
'\b([48])[kK]\b', s
)
1848 return {'height': int(mobj.group(1)) * 540}
1853 def parse_bitrate(s
):
1854 if not isinstance(s
, str):
1856 mobj
= re
.search(r
'\b(\d+)\s*kbps', s
)
1858 return int(mobj
.group(1))
1861 def month_by_name(name
, lang
='en'):
1862 """ Return the number of a month by (locale-independently) English name """
1864 month_names
= MONTH_NAMES
.get(lang
, MONTH_NAMES
['en'])
1867 return month_names
.index(name
) + 1
1872 def month_by_abbreviation(abbrev
):
1873 """ Return the number of a month by (locale-independently) English
1877 return [s
[:3] for s
in ENGLISH_MONTH_NAMES
].index(abbrev
) + 1
1882 def fix_xml_ampersands(xml_str
):
1883 """Replace all the '&' by '&' in XML"""
1885 r
'&(?!amp;|lt;|gt;|apos;|quot;|#x[0-9a-fA-F]{,4};|#[0-9]{,4};)',
1890 def setproctitle(title
):
1891 assert isinstance(title
, str)
1893 # Workaround for https://github.com/yt-dlp/yt-dlp/issues/4541
1900 libc
= ctypes
.cdll
.LoadLibrary('libc.so.6')
1904 # LoadLibrary in Windows Python 2.7.13 only expects
1905 # a bytestring, but since unicode_literals turns
1906 # every string into a unicode string, it fails.
1908 title_bytes
= title
.encode()
1909 buf
= ctypes
.create_string_buffer(len(title_bytes
))
1910 buf
.value
= title_bytes
1912 # PR_SET_NAME = 15 Ref: /usr/include/linux/prctl.h
1913 libc
.prctl(15, buf
, 0, 0, 0)
1914 except AttributeError:
1915 return # Strange libc, just skip this
1918 def remove_start(s
, start
):
1919 return s
[len(start
):] if s
is not None and s
.startswith(start
) else s
1922 def remove_end(s
, end
):
1923 return s
[:-len(end
)] if s
is not None and s
.endswith(end
) else s
1926 def remove_quotes(s
):
1927 if s
is None or len(s
) < 2:
1929 for quote
in ('"', "'"):
1930 if s
[0] == quote
and s
[-1] == quote
:
1935 def get_domain(url
):
1937 This implementation is inconsistent, but is kept for compatibility.
1938 Use this only for "webpage_url_domain"
1940 return remove_start(urllib
.parse
.urlparse(url
).netloc
, 'www.') or None
1943 def url_basename(url
):
1944 path
= urllib
.parse
.urlparse(url
).path
1945 return path
.strip('/').split('/')[-1]
1949 return re
.match(r
'https?://[^?#]+/', url
).group()
1952 def urljoin(base
, path
):
1953 if isinstance(path
, bytes):
1954 path
= path
.decode()
1955 if not isinstance(path
, str) or not path
:
1957 if re
.match(r
'^(?:[a-zA-Z][a-zA-Z0-9+-.]*:)?//', path
):
1959 if isinstance(base
, bytes):
1960 base
= base
.decode()
1961 if not isinstance(base
, str) or not re
.match(
1962 r
'^(?:https?:)?//', base
):
1964 return urllib
.parse
.urljoin(base
, path
)
1967 def int_or_none(v
, scale
=1, default
=None, get_attr
=None, invscale
=1):
1968 if get_attr
and v
is not None:
1969 v
= getattr(v
, get_attr
, None)
1971 return int(v
) * invscale
// scale
1972 except (ValueError, TypeError, OverflowError):
1976 def str_or_none(v
, default
=None):
1977 return default
if v
is None else str(v
)
1980 def str_to_int(int_str
):
1981 """ A more relaxed version of int_or_none """
1982 if isinstance(int_str
, int):
1984 elif isinstance(int_str
, str):
1985 int_str
= re
.sub(r
'[,\.\+]', '', int_str
)
1986 return int_or_none(int_str
)
1989 def float_or_none(v
, scale
=1, invscale
=1, default
=None):
1993 return float(v
) * invscale
/ scale
1994 except (ValueError, TypeError):
1998 def bool_or_none(v
, default
=None):
1999 return v
if isinstance(v
, bool) else default
2002 def strip_or_none(v
, default
=None):
2003 return v
.strip() if isinstance(v
, str) else default
2006 def url_or_none(url
):
2007 if not url
or not isinstance(url
, str):
2010 return url
if re
.match(r
'^(?:(?:https?|rt(?:m(?:pt?[es]?|fp)|sp[su]?)|mms|ftps?):)?//', url
) else None
2013 def strftime_or_none(timestamp
, date_format
='%Y%m%d', default
=None):
2014 datetime_object
= None
2016 if isinstance(timestamp
, (int, float)): # unix timestamp
2017 # Using naive datetime here can break timestamp() in Windows
2018 # Ref: https://github.com/yt-dlp/yt-dlp/issues/5185, https://github.com/python/cpython/issues/94414
2019 # Also, dt.datetime.fromtimestamp breaks for negative timestamps
2020 # Ref: https://github.com/yt-dlp/yt-dlp/issues/6706#issuecomment-1496842642
2021 datetime_object
= (dt
.datetime
.fromtimestamp(0, dt
.timezone
.utc
)
2022 + dt
.timedelta(seconds
=timestamp
))
2023 elif isinstance(timestamp
, str): # assume YYYYMMDD
2024 datetime_object
= dt
.datetime
.strptime(timestamp
, '%Y%m%d')
2025 date_format
= re
.sub( # Support %s on windows
2026 r
'(?<!%)(%%)*%s', rf
'\g<1>{int(datetime_object.timestamp())}', date_format
)
2027 return datetime_object
.strftime(date_format
)
2028 except (ValueError, TypeError, AttributeError):
2032 def parse_duration(s
):
2033 if not isinstance(s
, str):
2039 days
, hours
, mins
, secs
, ms
= [None] * 5
2040 m
= re
.match(r
'''(?x)
2042 (?:(?:(?P<days>[0-9]+):)?(?P<hours>[0-9]+):)?(?P<mins>[0-9]+):)?
2043 (?P<secs>(?(before_secs)[0-9]{1,2}|[0-9]+))
2044 (?P<ms>[.:][0-9]+)?Z?$
2047 days
, hours
, mins
, secs
, ms
= m
.group('days', 'hours', 'mins', 'secs', 'ms')
2052 [0-9]+\s*y(?:ears?)?,?\s*
2055 [0-9]+\s*m(?:onths?)?,?\s*
2058 [0-9]+\s*w(?:eeks?)?,?\s*
2061 (?P<days>[0-9]+)\s*d(?:ays?)?,?\s*
2065 (?P<hours>[0-9]+)\s*h(?:(?:ou)?rs?)?,?\s*
2068 (?P<mins>[0-9]+)\s*m(?:in(?:ute)?s?)?,?\s*
2071 (?P<secs>[0-9]+)(?P<ms>\.[0-9]+)?\s*s(?:ec(?:ond)?s?)?\s*
2074 days
, hours
, mins
, secs
, ms
= m
.groups()
2076 m
= re
.match(r
'(?i)(?:(?P<hours>[0-9.]+)\s*(?:hours?)|(?P<mins>[0-9.]+)\s*(?:mins?\.?|minutes?)\s*)Z?$', s
)
2078 hours
, mins
= m
.groups()
2083 ms
= ms
.replace(':', '.')
2084 return sum(float(part
or 0) * mult
for part
, mult
in (
2085 (days
, 86400), (hours
, 3600), (mins
, 60), (secs
, 1), (ms
, 1)))
2088 def prepend_extension(filename
, ext
, expected_real_ext
=None):
2089 name
, real_ext
= os
.path
.splitext(filename
)
2091 f
'{name}.{ext}{real_ext}'
2092 if not expected_real_ext
or real_ext
[1:] == expected_real_ext
2093 else f
'{filename}.{ext}')
2096 def replace_extension(filename
, ext
, expected_real_ext
=None):
2097 name
, real_ext
= os
.path
.splitext(filename
)
2098 return f
'{name if not expected_real_ext or real_ext[1:] == expected_real_ext else filename}.{ext}'
2101 def check_executable(exe
, args
=[]):
2102 """ Checks if the given binary is installed somewhere in PATH, and returns its name.
2103 args can be a list of arguments for a short output (like -version) """
2105 Popen
.run([exe
, *args
], stdout
=subprocess
.PIPE
, stderr
=subprocess
.PIPE
)
2111 def _get_exe_version_output(exe
, args
):
2113 # STDIN should be redirected too. On UNIX-like systems, ffmpeg triggers
2114 # SIGTTOU if yt-dlp is run in the background.
2115 # See https://github.com/ytdl-org/youtube-dl/issues/955#issuecomment-209789656
2116 stdout
, _
, ret
= Popen
.run([encodeArgument(exe
), *args
], text
=True,
2117 stdin
=subprocess
.PIPE
, stdout
=subprocess
.PIPE
, stderr
=subprocess
.STDOUT
)
2125 def detect_exe_version(output
, version_re
=None, unrecognized
='present'):
2126 assert isinstance(output
, str)
2127 if version_re
is None:
2128 version_re
= r
'version\s+([-0-9._a-zA-Z]+)'
2129 m
= re
.search(version_re
, output
)
2136 def get_exe_version(exe
, args
=['--version'],
2137 version_re
=None, unrecognized
=('present', 'broken')):
2138 """ Returns the version of the specified executable,
2139 or False if the executable is not present """
2140 unrecognized
= variadic(unrecognized
)
2141 assert len(unrecognized
) in (1, 2)
2142 out
= _get_exe_version_output(exe
, args
)
2144 return unrecognized
[-1]
2145 return out
and detect_exe_version(out
, version_re
, unrecognized
[0])
2148 def frange(start
=0, stop
=None, step
=1):
2151 start
, stop
= 0, start
2152 sign
= [-1, 1][step
> 0] if step
else 0
2153 while sign
* start
< sign
* stop
:
2158 class LazyList(collections
.abc
.Sequence
):
2159 """Lazy immutable list from an iterable
2160 Note that slices of a LazyList are lists and not LazyList"""
2162 class IndexError(IndexError): # noqa: A001
2165 def __init__(self
, iterable
, *, reverse
=False, _cache
=None):
2166 self
._iterable
= iter(iterable
)
2167 self
._cache
= [] if _cache
is None else _cache
2168 self
._reversed
= reverse
2172 # We need to consume the entire iterable to iterate in reverse
2173 yield from self
.exhaust()
2175 yield from self
._cache
2176 for item
in self
._iterable
:
2177 self
._cache
.append(item
)
2181 self
._cache
.extend(self
._iterable
)
2182 self
._iterable
= [] # Discard the emptied iterable to make it pickle-able
2186 """Evaluate the entire iterable"""
2187 return self
._exhaust
()[::-1 if self
._reversed
else 1]
2190 def _reverse_index(x
):
2191 return None if x
is None else ~x
2193 def __getitem__(self
, idx
):
2194 if isinstance(idx
, slice):
2196 idx
= slice(self
._reverse
_index
(idx
.start
), self
._reverse
_index
(idx
.stop
), -(idx
.step
or 1))
2197 start
, stop
, step
= idx
.start
, idx
.stop
, idx
.step
or 1
2198 elif isinstance(idx
, int):
2200 idx
= self
._reverse
_index
(idx
)
2201 start
, stop
, step
= idx
, idx
, 0
2203 raise TypeError('indices must be integers or slices')
2204 if ((start
or 0) < 0 or (stop
or 0) < 0
2205 or (start
is None and step
< 0)
2206 or (stop
is None and step
> 0)):
2207 # We need to consume the entire iterable to be able to slice from the end
2208 # Obviously, never use this with infinite iterables
2211 return self
._cache
[idx
]
2212 except IndexError as e
:
2213 raise self
.IndexError(e
) from e
2214 n
= max(start
or 0, stop
or 0) - len(self
._cache
) + 1
2216 self
._cache
.extend(itertools
.islice(self
._iterable
, n
))
2218 return self
._cache
[idx
]
2219 except IndexError as e
:
2220 raise self
.IndexError(e
) from e
2224 self
[-1] if self
._reversed
else self
[0]
2225 except self
.IndexError:
2231 return len(self
._cache
)
2233 def __reversed__(self
):
2234 return type(self
)(self
._iterable
, reverse
=not self
._reversed
, _cache
=self
._cache
)
2237 return type(self
)(self
._iterable
, reverse
=self
._reversed
, _cache
=self
._cache
)
2240 # repr and str should mimic a list. So we exhaust the iterable
2241 return repr(self
.exhaust())
2244 return repr(self
.exhaust())
2249 class IndexError(IndexError): # noqa: A001
2253 # This is only useful for tests
2254 return len(self
.getslice())
2256 def __init__(self
, pagefunc
, pagesize
, use_cache
=True):
2257 self
._pagefunc
= pagefunc
2258 self
._pagesize
= pagesize
2259 self
._pagecount
= float('inf')
2260 self
._use
_cache
= use_cache
2263 def getpage(self
, pagenum
):
2264 page_results
= self
._cache
.get(pagenum
)
2265 if page_results
is None:
2266 page_results
= [] if pagenum
> self
._pagecount
else list(self
._pagefunc
(pagenum
))
2268 self
._cache
[pagenum
] = page_results
2271 def getslice(self
, start
=0, end
=None):
2272 return list(self
._getslice
(start
, end
))
2274 def _getslice(self
, start
, end
):
2275 raise NotImplementedError('This method must be implemented by subclasses')
2277 def __getitem__(self
, idx
):
2278 assert self
._use
_cache
, 'Indexing PagedList requires cache'
2279 if not isinstance(idx
, int) or idx
< 0:
2280 raise TypeError('indices must be non-negative integers')
2281 entries
= self
.getslice(idx
, idx
+ 1)
2283 raise self
.IndexError
2287 return bool(self
.getslice(0, 1))
2290 class OnDemandPagedList(PagedList
):
2291 """Download pages until a page with less than maximum results"""
2293 def _getslice(self
, start
, end
):
2294 for pagenum
in itertools
.count(start
// self
._pagesize
):
2295 firstid
= pagenum
* self
._pagesize
2296 nextfirstid
= pagenum
* self
._pagesize
+ self
._pagesize
2297 if start
>= nextfirstid
:
2301 start
% self
._pagesize
2302 if firstid
<= start
< nextfirstid
2305 ((end
- 1) % self
._pagesize
) + 1
2306 if (end
is not None and firstid
<= end
<= nextfirstid
)
2310 page_results
= self
.getpage(pagenum
)
2312 self
._pagecount
= pagenum
- 1
2314 if startv
!= 0 or endv
is not None:
2315 page_results
= page_results
[startv
:endv
]
2316 yield from page_results
2318 # A little optimization - if current page is not "full", ie. does
2319 # not contain page_size videos then we can assume that this page
2320 # is the last one - there are no more ids on further pages -
2321 # i.e. no need to query again.
2322 if len(page_results
) + startv
< self
._pagesize
:
2325 # If we got the whole page, but the next page is not interesting,
2326 # break out early as well
2327 if end
== nextfirstid
:
2331 class InAdvancePagedList(PagedList
):
2332 """PagedList with total number of pages known in advance"""
2334 def __init__(self
, pagefunc
, pagecount
, pagesize
):
2335 PagedList
.__init
__(self
, pagefunc
, pagesize
, True)
2336 self
._pagecount
= pagecount
2338 def _getslice(self
, start
, end
):
2339 start_page
= start
// self
._pagesize
2340 end_page
= self
._pagecount
if end
is None else min(self
._pagecount
, end
// self
._pagesize
+ 1)
2341 skip_elems
= start
- start_page
* self
._pagesize
2342 only_more
= None if end
is None else end
- start
2343 for pagenum
in range(start_page
, end_page
):
2344 page_results
= self
.getpage(pagenum
)
2346 page_results
= page_results
[skip_elems
:]
2348 if only_more
is not None:
2349 if len(page_results
) < only_more
:
2350 only_more
-= len(page_results
)
2352 yield from page_results
[:only_more
]
2354 yield from page_results
2357 class PlaylistEntries
:
2358 MissingEntry
= object()
2359 is_exhausted
= False
2361 def __init__(self
, ydl
, info_dict
):
2364 # _entries must be assigned now since infodict can change during iteration
2365 entries
= info_dict
.get('entries')
2367 raise EntryNotInPlaylist('There are no entries')
2368 elif isinstance(entries
, list):
2369 self
.is_exhausted
= True
2371 requested_entries
= info_dict
.get('requested_entries')
2372 self
.is_incomplete
= requested_entries
is not None
2373 if self
.is_incomplete
:
2374 assert self
.is_exhausted
2375 self
._entries
= [self
.MissingEntry
] * max(requested_entries
or [0])
2376 for i
, entry
in zip(requested_entries
, entries
):
2377 self
._entries
[i
- 1] = entry
2378 elif isinstance(entries
, (list, PagedList
, LazyList
)):
2379 self
._entries
= entries
2381 self
._entries
= LazyList(entries
)
2383 PLAYLIST_ITEMS_RE
= re
.compile(r
'''(?x)
2384 (?P<start>[+-]?\d+)?
2386 (?P<end>[+-]?\d+|inf(?:inite)?)?
2387 (?::(?P<step>[+-]?\d+))?
2391 def parse_playlist_items(cls
, string
):
2392 for segment
in string
.split(','):
2394 raise ValueError('There is two or more consecutive commas')
2395 mobj
= cls
.PLAYLIST_ITEMS_RE
.fullmatch(segment
)
2397 raise ValueError(f
'{segment!r} is not a valid specification')
2398 start
, end
, step
, has_range
= mobj
.group('start', 'end', 'step', 'range')
2399 if int_or_none(step
) == 0:
2400 raise ValueError(f
'Step in {segment!r} cannot be zero')
2401 yield slice(int_or_none(start
), float_or_none(end
), int_or_none(step
)) if has_range
else int(start
)
2403 def get_requested_items(self
):
2404 playlist_items
= self
.ydl
.params
.get('playlist_items')
2405 playlist_start
= self
.ydl
.params
.get('playliststart', 1)
2406 playlist_end
= self
.ydl
.params
.get('playlistend')
2407 # For backwards compatibility, interpret -1 as whole list
2408 if playlist_end
in (-1, None):
2410 if not playlist_items
:
2411 playlist_items
= f
'{playlist_start}:{playlist_end}'
2412 elif playlist_start
!= 1 or playlist_end
:
2413 self
.ydl
.report_warning('Ignoring playliststart and playlistend because playlistitems was given', only_once
=True)
2415 for index
in self
.parse_playlist_items(playlist_items
):
2416 for i
, entry
in self
[index
]:
2421 # The item may have just been added to archive. Don't break due to it
2422 if not self
.ydl
.params
.get('lazy_playlist'):
2423 # TODO: Add auto-generated fields
2424 self
.ydl
._match
_entry
(entry
, incomplete
=True, silent
=True)
2425 except (ExistingVideoReached
, RejectedVideoReached
):
2428 def get_full_count(self
):
2429 if self
.is_exhausted
and not self
.is_incomplete
:
2431 elif isinstance(self
._entries
, InAdvancePagedList
):
2432 if self
._entries
._pagesize
== 1:
2433 return self
._entries
._pagecount
2435 @functools.cached_property
2437 if isinstance(self
._entries
, list):
2440 entry
= self
._entries
[i
]
2442 entry
= self
.MissingEntry
2443 if not self
.is_incomplete
:
2444 raise self
.IndexError
2445 if entry
is self
.MissingEntry
:
2446 raise EntryNotInPlaylist(f
'Entry {i + 1} cannot be found')
2451 return type(self
.ydl
)._handle
_extraction
_exceptions
(lambda _
, i
: self
._entries
[i
])(self
.ydl
, i
)
2452 except (LazyList
.IndexError, PagedList
.IndexError):
2453 raise self
.IndexError
2456 def __getitem__(self
, idx
):
2457 if isinstance(idx
, int):
2458 idx
= slice(idx
, idx
)
2460 # NB: PlaylistEntries[1:10] => (0, 1, ... 9)
2461 step
= 1 if idx
.step
is None else idx
.step
2462 if idx
.start
is None:
2463 start
= 0 if step
> 0 else len(self
) - 1
2465 start
= idx
.start
- 1 if idx
.start
>= 0 else len(self
) + idx
.start
2467 # NB: Do not call len(self) when idx == [:]
2468 if idx
.stop
is None:
2469 stop
= 0 if step
< 0 else float('inf')
2471 stop
= idx
.stop
- 1 if idx
.stop
>= 0 else len(self
) + idx
.stop
2472 stop
+= [-1, 1][step
> 0]
2474 for i
in frange(start
, stop
, step
):
2478 entry
= self
._getter
(i
)
2479 except self
.IndexError:
2480 self
.is_exhausted
= True
2487 return len(tuple(self
[:]))
2489 class IndexError(IndexError): # noqa: A001
2493 def uppercase_escape(s
):
2494 unicode_escape
= codecs
.getdecoder('unicode_escape')
2496 r
'\\U[0-9a-fA-F]{8}',
2497 lambda m
: unicode_escape(m
.group(0))[0],
2501 def lowercase_escape(s
):
2502 unicode_escape
= codecs
.getdecoder('unicode_escape')
2504 r
'\\u[0-9a-fA-F]{4}',
2505 lambda m
: unicode_escape(m
.group(0))[0],
2509 def parse_qs(url
, **kwargs
):
2510 return urllib
.parse
.parse_qs(urllib
.parse
.urlparse(url
).query
, **kwargs
)
2513 def read_batch_urls(batch_fd
):
2515 if not isinstance(url
, str):
2516 url
= url
.decode('utf-8', 'replace')
2517 BOM_UTF8
= ('\xef\xbb\xbf', '\ufeff')
2518 for bom
in BOM_UTF8
:
2519 if url
.startswith(bom
):
2520 url
= url
[len(bom
):]
2522 if not url
or url
.startswith(('#', ';', ']')):
2524 # "#" cannot be stripped out since it is part of the URI
2525 # However, it can be safely stripped out if following a whitespace
2526 return re
.split(r
'\s#', url
, maxsplit
=1)[0].rstrip()
2528 with contextlib
.closing(batch_fd
) as fd
:
2529 return [url
for url
in map(fixup
, fd
) if url
]
2532 def urlencode_postdata(*args
, **kargs
):
2533 return urllib
.parse
.urlencode(*args
, **kargs
).encode('ascii')
2536 def update_url(url
, *, query_update
=None, **kwargs
):
2537 """Replace URL components specified by kwargs
2538 @param url str or parse url tuple
2539 @param query_update update query
2542 if isinstance(url
, str):
2543 if not kwargs
and not query_update
:
2546 url
= urllib
.parse
.urlparse(url
)
2548 assert 'query' not in kwargs
, 'query_update and query cannot be specified at the same time'
2549 kwargs
['query'] = urllib
.parse
.urlencode({
2550 **urllib
.parse
.parse_qs(url
.query
),
2553 return urllib
.parse
.urlunparse(url
._replace
(**kwargs
))
2556 def update_url_query(url
, query
):
2557 return update_url(url
, query_update
=query
)
2560 def _multipart_encode_impl(data
, boundary
):
2561 content_type
= f
'multipart/form-data; boundary={boundary}'
2564 for k
, v
in data
.items():
2565 out
+= b
'--' + boundary
.encode('ascii') + b
'\r\n'
2566 if isinstance(k
, str):
2568 if isinstance(v
, str):
2570 # RFC 2047 requires non-ASCII field names to be encoded, while RFC 7578
2571 # suggests sending UTF-8 directly. Firefox sends UTF-8, too
2572 content
= b
'Content-Disposition: form-data; name="' + k
+ b
'"\r\n\r\n' + v
+ b
'\r\n'
2573 if boundary
.encode('ascii') in content
:
2574 raise ValueError('Boundary overlaps with data')
2577 out
+= b
'--' + boundary
.encode('ascii') + b
'--\r\n'
2579 return out
, content_type
2582 def multipart_encode(data
, boundary
=None):
2584 Encode a dict to RFC 7578-compliant form-data
2587 A dict where keys and values can be either Unicode or bytes-like
2590 If specified a Unicode object, it's used as the boundary. Otherwise
2591 a random boundary is generated.
2593 Reference: https://tools.ietf.org/html/rfc7578
2595 has_specified_boundary
= boundary
is not None
2598 if boundary
is None:
2599 boundary
= '---------------' + str(random
.randrange(0x0fffffff, 0xffffffff))
2602 out
, content_type
= _multipart_encode_impl(data
, boundary
)
2605 if has_specified_boundary
:
2609 return out
, content_type
2612 def is_iterable_like(x
, allowed_types
=collections
.abc
.Iterable
, blocked_types
=NO_DEFAULT
):
2613 if blocked_types
is NO_DEFAULT
:
2614 blocked_types
= (str, bytes, collections
.abc
.Mapping
)
2615 return isinstance(x
, allowed_types
) and not isinstance(x
, blocked_types
)
2618 def variadic(x
, allowed_types
=NO_DEFAULT
):
2619 if not isinstance(allowed_types
, (tuple, type)):
2620 deprecation_warning('allowed_types should be a tuple or a type')
2621 allowed_types
= tuple(allowed_types
)
2622 return x
if is_iterable_like(x
, blocked_types
=allowed_types
) else (x
, )
2625 def try_call(*funcs
, expected_type
=None, args
=[], kwargs
={}):
2628 val
= f(*args
, **kwargs
)
2629 except (AttributeError, KeyError, TypeError, IndexError, ValueError, ZeroDivisionError):
2632 if expected_type
is None or isinstance(val
, expected_type
):
2636 def try_get(src
, getter
, expected_type
=None):
2637 return try_call(*variadic(getter
), args
=(src
,), expected_type
=expected_type
)
2640 def filter_dict(dct
, cndn
=lambda _
, v
: v
is not None):
2641 return {k: v for k, v in dct.items() if cndn(k, v)}
2644 def merge_dicts(*dicts
):
2646 for a_dict
in dicts
:
2647 for k
, v
in a_dict
.items():
2648 if (v
is not None and k
not in merged
2649 or isinstance(v
, str) and merged
[k
] == ''):
2654 def encode_compat_str(string
, encoding
=preferredencoding(), errors
='strict'):
2655 return string
if isinstance(string
, str) else str(string
, encoding
, errors
)
2667 TV_PARENTAL_GUIDELINES
= {
2677 def parse_age_limit(s
):
2678 # isinstance(False, int) is True. So type() must be used instead
2679 if type(s
) is int: # noqa: E721
2680 return s
if 0 <= s
<= 21 else None
2681 elif not isinstance(s
, str):
2683 m
= re
.match(r
'^(?P<age>\d{1,2})\+?$', s
)
2685 return int(m
.group('age'))
2688 return US_RATINGS
[s
]
2689 m
= re
.match(r
'^TV[_-]?({})$'.format('|'.join(k
[3:] for k
in TV_PARENTAL_GUIDELINES
)), s
)
2691 return TV_PARENTAL_GUIDELINES
['TV-' + m
.group(1)]
2695 def strip_jsonp(code
):
2698 (?:window\.)?(?P<func_name>[a-zA-Z0-9_.$]*)
2699 (?:\s*&&\s*(?P=func_name))?
2700 \s*\(\s*(?P<callback_data>.*)\);?
2701 \s*?(?://[^\n]*)*$''',
2702 r
'\g<callback_data>', code
)
2705 def js_to_json(code
, vars={}, *, strict
=False):
2706 # vars is a dict of var, val pairs to substitute
2707 STRING_QUOTES
= '\'"`'
2708 STRING_RE
= '|'.join(rf
'{q}(?:\\.|[^\\{q}])*{q}' for q
in STRING_QUOTES
)
2709 COMMENT_RE
= r
'/\*(?:(?!\*/).)*?\*/|//[^\n]*\n'
2710 SKIP_RE
= fr
'\s*(?:{COMMENT_RE})?\s*'
2712 (fr
'(?s)^(0[xX][0-9a-fA-F]+){SKIP_RE}:?$', 16),
2713 (fr
'(?s)^(0+[0-7]+){SKIP_RE}:?$', 8),
2716 def process_escape(match
):
2717 JSON_PASSTHROUGH_ESCAPES
= R
'"\bfnrtu'
2718 escape
= match
.group(1) or match
.group(2)
2720 return (Rf
'\{escape}' if escape
in JSON_PASSTHROUGH_ESCAPES
2721 else R
'\u00' if escape
== 'x'
2722 else '' if escape
== '\n'
2725 def template_substitute(match
):
2726 evaluated
= js_to_json(match
.group(1), vars, strict
=strict
)
2727 if evaluated
[0] == '"':
2728 return json
.loads(evaluated
)
2733 if v
in ('true', 'false', 'null'):
2735 elif v
in ('undefined', 'void 0'):
2737 elif v
.startswith(('/*', '//', '!')) or v
== ',':
2740 if v
[0] in STRING_QUOTES
:
2741 v
= re
.sub(r
'(?s)\${([^}]+)}', template_substitute
, v
[1:-1]) if v
[0] == '`' else v
[1:-1]
2742 escaped
= re
.sub(r
'(?s)(")|\\(.)', process_escape
, v
)
2743 return f
'"{escaped}"'
2745 for regex
, base
in INTEGER_TABLE
:
2746 im
= re
.match(regex
, v
)
2748 i
= int(im
.group(1), base
)
2749 return f
'"{i}":' if v
.endswith(':') else str(i
)
2755 except json
.JSONDecodeError
:
2756 return json
.dumps(vars[v
])
2763 raise ValueError(f
'Unknown value: {v}')
2765 def create_map(mobj
):
2766 return json
.dumps(dict(json
.loads(js_to_json(mobj
.group(1) or '[]', vars=vars))))
2768 code
= re
.sub(r
'(?:new\s+)?Array\((.*?)\)', r
'[\g<1>]', code
)
2769 code
= re
.sub(r
'new Map\((\[.*?\])?\)', create_map
, code
)
2771 code
= re
.sub(rf
'new Date\(({STRING_RE})\)', r
'\g<1>', code
)
2772 code
= re
.sub(r
'new \w+\((.*?)\)', lambda m
: json
.dumps(m
.group(0)), code
)
2773 code
= re
.sub(r
'parseInt\([^\d]+(\d+)[^\d]+\)', r
'\1', code
)
2774 code
= re
.sub(r
'\(function\([^)]*\)\s*\{[^}]*\}\s*\)\s*\(\s*(["\'][^
)]*["\'])\s*\)', r'\1', code)
2776 return re.sub(rf'''(?sx)
2778 {COMMENT_RE}|,(?={SKIP_RE}[\]}}])|
2779 void\s0|(?:(?<![0-9])[eE]|[a-df-zA-DF-Z_$])[.a-zA-Z_$0-9]*|
2780 \b(?:0[xX][0-9a-fA-F]+|0+[0-7]+)(?:{SKIP_RE}:)?|
2781 [0-9]+(?={SKIP_RE}:)|
2786 def qualities(quality_ids):
2787 """ Get a numeric quality value out of a list of possible values """
2790 return quality_ids.index(qid)
2796 POSTPROCESS_WHEN = ('pre_process', 'after_filter', 'video', 'before_dl', 'post_process', 'after_move', 'after_video', 'playlist')
2800 'default': '%(title)s [%(id)s].%(ext)s',
2801 'chapter': '%(title)s - %(section_number)03d %(section_title)s [%(id)s].%(ext)s',
2807 'description': 'description',
2808 'annotation': 'annotations.xml',
2809 'infojson': 'info.json',
2812 'pl_thumbnail': None,
2813 'pl_description': 'description',
2814 'pl_infojson': 'info.json',
2817 # As of [1] format syntax is:
2818 # %[mapping_key][conversion_flags][minimum_width][.precision][length_modifier]type
2819 # 1. https://docs.python.org/2/library/stdtypes.html#string-formatting
2820 STR_FORMAT_RE_TMPL = r'''(?x)
2821 (?<!%)(?P<prefix>(?:%%)*)
2823 (?P<has_key>\((?P<key>{0})\))?
2825 (?P<conversion>[#0\-+ ]+)?
2827 (?P<precision>\.\d+)?
2828 (?P<len_mod>[hlL])? # unused in python
2829 {1} # conversion type
2834 STR_FORMAT_TYPES = 'diouxXeEfFgGcrsa'
2837 def limit_length(s, length):
2838 """ Add ellipses to overly long strings """
2843 return s[:length - len(ELLIPSES)] + ELLIPSES
2847 def version_tuple(v):
2848 return tuple(int(e) for e in re.split(r'[-.]', v))
2851 def is_outdated_version(version, limit, assume_new=True):
2853 return not assume_new
2855 return version_tuple(version) < version_tuple(limit)
2857 return not assume_new
2860 def ytdl_is_updateable():
2861 """ Returns if yt-dlp can be updated with -U """
2863 from ..update import is_non_updateable
2865 return not is_non_updateable()
2868 def args_to_str(args):
2869 # Get a short string representation for a subprocess command
2870 return shell_quote(args)
2873 def error_to_str(err):
2874 return f'{type(err).__name__}: {err}'
2877 def mimetype2ext(mt, default=NO_DEFAULT):
2878 if not isinstance(mt, str):
2879 if default is not NO_DEFAULT:
2896 'x-matroska': 'mkv',
2898 'x-mp4-fragmented': 'mp4',
2903 # application (streaming playlists)
2907 'vnd.apple.mpegurl': 'm3u8',
2908 'vnd.ms-sstr+xml': 'ism',
2909 'x-mpegurl': 'm3u8',
2913 # Per RFC 3003, audio/mpeg can be .mp1, .mp2 or .mp3.
2914 # Using .mp3 as it's the most popular one
2915 'audio/mpeg': 'mp3',
2916 'audio/webm': 'webm',
2917 'audio/x-matroska': 'mka',
2918 'audio/x-mpegurl': 'm3u',
2926 'x-realaudio': 'ra',
2937 'vnd.wap.wbmp': 'wbmp',
2944 'filmstrip+json': 'fs',
2945 'smptett+xml': 'tt',
2948 'x-ms-sami': 'sami',
2957 mimetype = mt.partition(';')[0].strip().lower()
2958 _, _, subtype = mimetype.rpartition('/')
2960 ext = traversal.traverse_obj(MAP, mimetype, subtype, subtype.rsplit('+')[-1])
2963 elif default is not NO_DEFAULT:
2965 return subtype.replace('+', '.')
2968 def ext2mimetype(ext_or_url):
2971 if '.' not in ext_or_url:
2972 ext_or_url = f'file.{ext_or_url}'
2973 return mimetypes.guess_type(ext_or_url)[0]
2976 def parse_codecs(codecs_str):
2977 # http://tools.ietf.org/html/rfc6381
2980 split_codecs = list(filter(None, map(
2981 str.strip, codecs_str.strip().strip(',').split(','))))
2982 vcodec, acodec, scodec, hdr = None, None, None, None
2983 for full_codec in split_codecs:
2984 parts = re.sub(r'0+(?=\d)', '', full_codec).split('.')
2985 if parts[0] in ('avc1', 'avc2', 'avc3', 'avc4', 'vp9', 'vp8', 'hev1', 'hev2',
2986 'h263', 'h264', 'mp4v', 'hvc1', 'av1', 'theora', 'dvh1', 'dvhe'):
2990 if parts[0] in ('dvh1', 'dvhe'):
2992 elif parts[0] == 'av1' and traversal.traverse_obj(parts, 3) == '10':
2994 elif parts[:2] == ['vp9', '2']:
2996 elif parts[0] in ('flac', 'mp4a', 'opus', 'vorbis', 'mp3', 'aac', 'ac-4',
2997 'ac-3', 'ec-3', 'eac3', 'dtsc', 'dtse', 'dtsh', 'dtsl'):
2998 acodec = acodec or full_codec
2999 elif parts[0] in ('stpp', 'wvtt'):
3000 scodec = scodec or full_codec
3002 write_string(f'WARNING: Unknown codec {full_codec}\n')
3003 if vcodec or acodec or scodec:
3005 'vcodec': vcodec or 'none',
3006 'acodec': acodec or 'none',
3007 'dynamic_range': hdr,
3008 **({'scodec': scodec} if scodec is not None else {}),
3010 elif len(split_codecs) == 2:
3012 'vcodec': split_codecs[0],
3013 'acodec': split_codecs[1],
3018 def get_compatible_ext(*, vcodecs, acodecs, vexts, aexts, preferences=None):
3019 assert len(vcodecs) == len(vexts) and len(acodecs) == len(aexts)
3021 allow_mkv = not preferences or 'mkv' in preferences
3023 if allow_mkv and max(len(acodecs), len(vcodecs)) > 1:
3024 return 'mkv' # TODO: any other format allows this?
3026 # TODO: All codecs supported by parse_codecs isn't handled here
3027 COMPATIBLE_CODECS = {
3029 'av1', 'hevc', 'avc1', 'mp4a', 'ac-4', # fourcc (m3u8, mpd)
3030 'h264', 'aacl', 'ec-3', # Set in ISM
3033 'av1', 'vp9', 'vp8', 'opus', 'vrbs',
3034 'vp9x', 'vp8x', # in the webm spec
3038 sanitize_codec = functools.partial(
3039 try_get, getter=lambda x: x[0].split('.')[0].replace('0', '').lower())
3040 vcodec, acodec = sanitize_codec(vcodecs), sanitize_codec(acodecs)
3042 for ext in preferences or COMPATIBLE_CODECS.keys():
3043 codec_set = COMPATIBLE_CODECS.get(ext, set())
3044 if ext == 'mkv' or codec_set.issuperset((vcodec, acodec)):
3048 {'mp3', 'mp4', 'm4a', 'm4p', 'm4b', 'm4r', 'm4v', 'ismv', 'isma', 'mov'},
3051 for ext in preferences or vexts:
3052 current_exts = {ext, *vexts, *aexts}
3053 if ext == 'mkv' or current_exts == {ext} or any(
3054 ext_sets.issuperset(current_exts) for ext_sets in COMPATIBLE_EXTS):
3056 return 'mkv' if allow_mkv else preferences[-1]
3059 def urlhandle_detect_ext(url_handle, default=NO_DEFAULT):
3060 getheader = url_handle.headers.get
3062 cd = getheader('Content-Disposition')
3064 m = re.match(r'attachment;\s*filename="(?P
<filename
>[^
"]+)"', cd)
3066 e = determine_ext(m.group('filename
'), default_ext=None)
3070 meta_ext = getheader('x
-amz
-meta
-name
')
3072 e = meta_ext.rpartition('.')[2]
3076 return mimetype2ext(getheader('Content
-Type
'), default=default)
3079 def encode_data_uri(data, mime_type):
3080 return 'data
:{};base64,{}
'.format(mime_type, base64.b64encode(data).decode('ascii
'))
3083 def age_restricted(content_limit, age_limit):
3084 """ Returns True iff the content should be blocked """
3086 if age_limit is None: # No limit set
3088 if content_limit is None:
3089 return False # Content available for everyone
3090 return age_limit < content_limit
3093 # List of known byte-order-marks (BOM)
3095 (b'\xef\xbb\xbf', 'utf
-8'),
3096 (b'\x00\x00\xfe\xff', 'utf
-32-be
'),
3097 (b'\xff\xfe\x00\x00', 'utf
-32-le
'),
3098 (b'\xff\xfe', 'utf
-16-le
'),
3099 (b'\xfe\xff', 'utf
-16-be
'),
3103 def is_html(first_bytes):
3104 """ Detect whether a file contains HTML by examining its first bytes. """
3107 for bom, enc in BOMS:
3108 while first_bytes.startswith(bom):
3109 encoding, first_bytes = enc, first_bytes[len(bom):]
3111 return re.match(r'^\s
*<', first_bytes.decode(encoding, 'replace
'))
3114 def determine_protocol(info_dict):
3115 protocol = info_dict.get('protocol
')
3116 if protocol is not None:
3119 url = sanitize_url(info_dict['url
'])
3120 if url.startswith('rtmp
'):
3122 elif url.startswith('mms
'):
3124 elif url.startswith('rtsp
'):
3127 ext = determine_ext(url)
3129 return 'm3u8
' if info_dict.get('is_live
') else 'm3u8_native
'
3133 return urllib.parse.urlparse(url).scheme
3136 def render_table(header_row, data, delim=False, extra_gap=0, hide_empty=False):
3137 """ Render a list of rows, each as a list of values.
3138 Text after a \t will be right aligned """
3140 return len(remove_terminal_sequences(string).replace('\t', ''))
3142 def get_max_lens(table):
3143 return [max(width(str(v)) for v in col) for col in zip(*table)]
3145 def filter_using_list(row, filter_array):
3146 return [col for take, col in itertools.zip_longest(filter_array, row, fillvalue=True) if take]
3148 max_lens = get_max_lens(data) if hide_empty else []
3149 header_row = filter_using_list(header_row, max_lens)
3150 data = [filter_using_list(row, max_lens) for row in data]
3152 table = [header_row, *data]
3153 max_lens = get_max_lens(table)
3156 table = [header_row, [delim * (ml + extra_gap) for ml in max_lens], *data]
3157 table[1][-1] = table[1][-1][:-extra_gap * len(delim)] # Remove extra_gap from end of delimiter
3159 for pos, text in enumerate(map(str, row)):
3161 row[pos] = text.replace('\t', ' ' * (max_lens[pos] - width(text))) + ' ' * extra_gap
3163 row[pos] = text + ' ' * (max_lens[pos] - width(text) + extra_gap)
3164 return '\n'.join(''.join(row).rstrip() for row in table)
3167 def _match_one(filter_part, dct, incomplete):
3168 # TODO: Generalize code with YoutubeDL._build_format_filter
3169 STRING_OPERATORS = {
3170 '*=': operator.contains,
3171 '^
=': lambda attr, value: attr.startswith(value),
3172 '$
=': lambda attr, value: attr.endswith(value),
3173 '~
=': lambda attr, value: re.search(value, attr),
3175 COMPARISON_OPERATORS = {
3177 '<=': operator.le, # "<=" must be defined above "<"
3184 if isinstance(incomplete, bool):
3185 is_incomplete = lambda _: incomplete
3187 is_incomplete = lambda k: k in incomplete
3189 operator_rex = re.compile(r'''(?x)
3191 \s*(?P<negation>!\s*)?(?P<op>{})(?P<none_inclusive>\s*\?)?\s*
3193 (?P<quote>["\'])(?P<quotedstrval>.+?)(?P=quote)|
3196 '''.format('|
'.join(map(re.escape, COMPARISON_OPERATORS.keys()))))
3197 m = operator_rex.fullmatch(filter_part.strip())
3200 unnegated_op = COMPARISON_OPERATORS[m['op
']]
3202 op = lambda attr, value: not unnegated_op(attr, value)
3205 comparison_value = m['quotedstrval
'] or m['strval
'] or m['intval
']
3207 comparison_value = comparison_value.replace(r'\
{}'.format(m['quote
']), m['quote
'])
3208 actual_value = dct.get(m['key
'])
3209 numeric_comparison = None
3210 if isinstance(actual_value, (int, float)):
3211 # If the original field is a string and matching comparisonvalue is
3212 # a number we should respect the origin of the original field
3213 # and process comparison value as a string (see
3214 # https://github.com/ytdl-org/youtube-dl/issues/11082)
3216 numeric_comparison = int(comparison_value)
3218 numeric_comparison = parse_filesize(comparison_value)
3219 if numeric_comparison is None:
3220 numeric_comparison = parse_filesize(f'{comparison_value}B
')
3221 if numeric_comparison is None:
3222 numeric_comparison = parse_duration(comparison_value)
3223 if numeric_comparison is not None and m['op
'] in STRING_OPERATORS:
3224 raise ValueError('Operator
{} only supports string values
!'.format(m['op
']))
3225 if actual_value is None:
3226 return is_incomplete(m['key
']) or m['none_inclusive
']
3227 return op(actual_value, comparison_value if numeric_comparison is None else numeric_comparison)
3230 '': lambda v: (v is True) if isinstance(v, bool) else (v is not None),
3231 '!': lambda v: (v is False) if isinstance(v, bool) else (v is None),
3233 operator_rex = re.compile(r'''(?x)
3234 (?P<op>{})\s*(?P<key>[a-z_]+)
3235 '''.format('|
'.join(map(re.escape, UNARY_OPERATORS.keys()))))
3236 m = operator_rex.fullmatch(filter_part.strip())
3238 op = UNARY_OPERATORS[m.group('op
')]
3239 actual_value = dct.get(m.group('key
'))
3240 if is_incomplete(m.group('key
')) and actual_value is None:
3242 return op(actual_value)
3244 raise ValueError(f'Invalid
filter part {filter_part!r}
')
3247 def match_str(filter_str, dct, incomplete=False):
3248 """ Filter a dictionary with a simple string syntax.
3249 @returns Whether the filter passes
3250 @param incomplete Set of keys that is expected to be missing from dct.
3251 Can be True/False to indicate all/none of the keys may be missing.
3252 All conditions on incomplete keys pass if the key is missing
3255 _match_one(filter_part.replace(r'\
&', '&'), dct, incomplete)
3256 for filter_part in re.split(r'(?
<!\\)&', filter_str))
3259 def match_filter_func(filters, breaking_filters=None):
3260 if not filters and not breaking_filters:
3262 repr_ = f'{match_filter_func.__module__}
.{match_filter_func.__qualname__}
({filters}
, {breaking_filters}
)'
3264 breaking_filters = match_filter_func(breaking_filters) or (lambda _, __: None)
3265 filters = set(variadic(filters or []))
3267 interactive = '-' in filters
3271 @function_with_repr.set_repr(repr_)
3272 def _match_func(info_dict, incomplete=False):
3273 ret = breaking_filters(info_dict, incomplete)
3275 raise RejectedVideoReached(ret)
3277 if not filters or any(match_str(f, info_dict, incomplete) for f in filters):
3278 return NO_DEFAULT if interactive and not incomplete else None
3280 video_title = info_dict.get('title
') or info_dict.get('id') or 'entry
'
3281 filter_str = ') |
('.join(map(str.strip, filters))
3282 return f'{video_title} does
not pass filter ({filter_str}
), skipping
..'
3286 class download_range_func:
3287 def __init__(self, chapters, ranges, from_info=False):
3288 self.chapters, self.ranges, self.from_info = chapters, ranges, from_info
3290 def __call__(self, info_dict, ydl):
3292 warning = ('There are no chapters matching the regex
' if info_dict.get('chapters
')
3293 else 'Cannot match chapters since chapter information
is unavailable
')
3294 for regex in self.chapters or []:
3295 for i, chapter in enumerate(info_dict.get('chapters
') or []):
3296 if re.search(regex, chapter['title
']):
3298 yield {**chapter, 'index': i}
3299 if self.chapters and warning:
3300 ydl.to_screen(f'[info
] {info_dict["id"]}
: {warning}
')
3302 for start, end in self.ranges or []:
3304 'start_time
': self._handle_negative_timestamp(start, info_dict),
3305 'end_time
': self._handle_negative_timestamp(end, info_dict),
3308 if self.from_info and (info_dict.get('start_time
') or info_dict.get('end_time
')):
3310 'start_time
': info_dict.get('start_time
') or 0,
3311 'end_time
': info_dict.get('end_time
') or float('inf
'),
3313 elif not self.ranges and not self.chapters:
3317 def _handle_negative_timestamp(time, info):
3318 return max(info['duration
'] + time, 0) if info.get('duration
') and time < 0 else time
3320 def __eq__(self, other):
3321 return (isinstance(other, download_range_func)
3322 and self.chapters == other.chapters and self.ranges == other.ranges)
3325 return f'{__name__}
.{type(self).__name__}
({self.chapters}
, {self.ranges}
)'
3328 def parse_dfxp_time_expr(time_expr):
3332 mobj = re.match(rf'^
(?P
<time_offset
>{NUMBER_RE}
)s?$
', time_expr)
3334 return float(mobj.group('time_offset
'))
3336 mobj = re.match(r'^
(\d
+):(\d\d
):(\d\
d(?
:(?
:\
.|
:)\d
+)?
)$
', time_expr)
3338 return 3600 * int(mobj.group(1)) + 60 * int(mobj.group(2)) + float(mobj.group(3).replace(':', '.'))
3341 def srt_subtitles_timecode(seconds):
3342 return '%02d
:%02d
:%02d
,%03d
' % timetuple_from_msec(seconds * 1000)
3345 def ass_subtitles_timecode(seconds):
3346 time = timetuple_from_msec(seconds * 1000)
3347 return '%01d
:%02d
:%02d
.%02d
' % (*time[:-1], time.milliseconds / 10)
3350 def dfxp2srt(dfxp_data):
3352 @param dfxp_data A bytes-like object containing DFXP data
3353 @returns A unicode object containing converted SRT data
3355 LEGACY_NAMESPACES = (
3356 (b'http
://www
.w3
.org
/ns
/ttml
', [
3357 b'http
://www
.w3
.org
/2004/11/ttaf1
',
3358 b'http
://www
.w3
.org
/2006/04/ttaf1
',
3359 b'http
://www
.w3
.org
/2006/10/ttaf1
',
3361 (b'http
://www
.w3
.org
/ns
/ttml
#styling', [
3362 b
'http://www.w3.org/ns/ttml#style',
3366 SUPPORTED_STYLING
= [
3375 _x
= functools
.partial(xpath_with_ns
, ns_map
={
3376 'xml': 'http://www.w3.org/XML/1998/namespace',
3377 'ttml': 'http://www.w3.org/ns/ttml',
3378 'tts': 'http://www.w3.org/ns/ttml#styling',
3384 class TTMLPElementParser
:
3386 _unclosed_elements
= []
3387 _applied_styles
= []
3389 def start(self
, tag
, attrib
):
3390 if tag
in (_x('ttml:br'), 'br'):
3393 unclosed_elements
= []
3395 element_style_id
= attrib
.get('style')
3397 style
.update(default_style
)
3398 if element_style_id
:
3399 style
.update(styles
.get(element_style_id
, {}))
3400 for prop
in SUPPORTED_STYLING
:
3401 prop_val
= attrib
.get(_x('tts:' + prop
))
3403 style
[prop
] = prop_val
3406 for k
, v
in sorted(style
.items()):
3407 if self
._applied
_styles
and self
._applied
_styles
[-1].get(k
) == v
:
3410 font
+= f
' color="{v}"'
3411 elif k
== 'fontSize':
3412 font
+= f
' size="{v}"'
3413 elif k
== 'fontFamily':
3414 font
+= f
' face="{v}"'
3415 elif k
== 'fontWeight' and v
== 'bold':
3417 unclosed_elements
.append('b')
3418 elif k
== 'fontStyle' and v
== 'italic':
3420 unclosed_elements
.append('i')
3421 elif k
== 'textDecoration' and v
== 'underline':
3423 unclosed_elements
.append('u')
3425 self
._out
+= '<font' + font
+ '>'
3426 unclosed_elements
.append('font')
3428 if self
._applied
_styles
:
3429 applied_style
.update(self
._applied
_styles
[-1])
3430 applied_style
.update(style
)
3431 self
._applied
_styles
.append(applied_style
)
3432 self
._unclosed
_elements
.append(unclosed_elements
)
3435 if tag
not in (_x('ttml:br'), 'br'):
3436 unclosed_elements
= self
._unclosed
_elements
.pop()
3437 for element
in reversed(unclosed_elements
):
3438 self
._out
+= f
'</{element}>'
3439 if unclosed_elements
and self
._applied
_styles
:
3440 self
._applied
_styles
.pop()
3442 def data(self
, data
):
3446 return self
._out
.strip()
3448 # Fix UTF-8 encoded file wrongly marked as UTF-16. See https://github.com/yt-dlp/yt-dlp/issues/6543#issuecomment-1477169870
3449 # This will not trigger false positives since only UTF-8 text is being replaced
3450 dfxp_data
= dfxp_data
.replace(b
'encoding=\'UTF-16\'', b
'encoding=\'UTF-8\'')
3452 def parse_node(node
):
3453 target
= TTMLPElementParser()
3454 parser
= xml
.etree
.ElementTree
.XMLParser(target
=target
)
3455 parser
.feed(xml
.etree
.ElementTree
.tostring(node
))
3456 return parser
.close()
3458 for k
, v
in LEGACY_NAMESPACES
:
3460 dfxp_data
= dfxp_data
.replace(ns
, k
)
3462 dfxp
= compat_etree_fromstring(dfxp_data
)
3464 paras
= dfxp
.findall(_x('.//ttml:p')) or dfxp
.findall('.//p')
3467 raise ValueError('Invalid dfxp/TTML subtitle')
3471 for style
in dfxp
.findall(_x('.//ttml:style')):
3472 style_id
= style
.get('id') or style
.get(_x('xml:id'))
3475 parent_style_id
= style
.get('style')
3477 if parent_style_id
not in styles
:
3480 styles
[style_id
] = styles
[parent_style_id
].copy()
3481 for prop
in SUPPORTED_STYLING
:
3482 prop_val
= style
.get(_x('tts:' + prop
))
3484 styles
.setdefault(style_id
, {})[prop
] = prop_val
3490 for p
in ('body', 'div'):
3491 ele
= xpath_element(dfxp
, [_x('.//ttml:' + p
), './/' + p
])
3494 style
= styles
.get(ele
.get('style'))
3497 default_style
.update(style
)
3499 for para
, index
in zip(paras
, itertools
.count(1)):
3500 begin_time
= parse_dfxp_time_expr(para
.attrib
.get('begin'))
3501 end_time
= parse_dfxp_time_expr(para
.attrib
.get('end'))
3502 dur
= parse_dfxp_time_expr(para
.attrib
.get('dur'))
3503 if begin_time
is None:
3508 end_time
= begin_time
+ dur
3509 out
.append('%d\n%s --> %s\n%s\n\n' % (
3511 srt_subtitles_timecode(begin_time
),
3512 srt_subtitles_timecode(end_time
),
3518 def cli_option(params
, command_option
, param
, separator
=None):
3519 param
= params
.get(param
)
3520 return ([] if param
is None
3521 else [command_option
, str(param
)] if separator
is None
3522 else [f
'{command_option}{separator}{param}'])
3525 def cli_bool_option(params
, command_option
, param
, true_value
='true', false_value
='false', separator
=None):
3526 param
= params
.get(param
)
3527 assert param
in (True, False, None)
3528 return cli_option({True: true_value, False: false_value}
, command_option
, param
, separator
)
3531 def cli_valueless_option(params
, command_option
, param
, expected_value
=True):
3532 return [command_option
] if params
.get(param
) == expected_value
else []
3535 def cli_configuration_args(argdict
, keys
, default
=[], use_compat
=True):
3536 if isinstance(argdict
, (list, tuple)): # for backward compatibility
3543 assert isinstance(argdict
, dict)
3545 assert isinstance(keys
, (list, tuple))
3546 for key_list
in keys
:
3547 arg_list
= list(filter(
3548 lambda x
: x
is not None,
3549 [argdict
.get(key
.lower()) for key
in variadic(key_list
)]))
3551 return [arg
for args
in arg_list
for arg
in args
]
3555 def _configuration_args(main_key
, argdict
, exe
, keys
=None, default
=[], use_compat
=True):
3556 main_key
, exe
= main_key
.lower(), exe
.lower()
3557 root_key
= exe
if main_key
== exe
else f
'{main_key}+{exe}'
3558 keys
= [f
'{root_key}{k}' for k
in (keys
or [''])]
3559 if root_key
in keys
:
3561 keys
.append((main_key
, exe
))
3562 keys
.append('default')
3565 return cli_configuration_args(argdict
, keys
, default
, use_compat
)
3569 # See http://www.loc.gov/standards/iso639-2/ISO-639-2_utf-8.txt
3628 'iw': 'heb', # Replaced by he in 1989 revision
3638 'in': 'ind', # Replaced by id in 1989 revision
3754 'ji': 'yid', # Replaced by yi in 1989 revision
3762 def short2long(cls
, code
):
3763 """Convert language code from ISO 639-1 to ISO 639-2/T"""
3764 return cls
._lang
_map
.get(code
[:2])
3767 def long2short(cls
, code
):
3768 """Convert language code from ISO 639-2/T to ISO 639-1"""
3769 for short_name
, long_name
in cls
._lang
_map
.items():
3770 if long_name
== code
:
3775 # From http://data.okfn.org/data/core/country-list
3777 'AF': 'Afghanistan',
3778 'AX': 'Åland Islands',
3781 'AS': 'American Samoa',
3786 'AG': 'Antigua and Barbuda',
3803 'BO': 'Bolivia, Plurinational State of',
3804 'BQ': 'Bonaire, Sint Eustatius and Saba',
3805 'BA': 'Bosnia and Herzegovina',
3807 'BV': 'Bouvet Island',
3809 'IO': 'British Indian Ocean Territory',
3810 'BN': 'Brunei Darussalam',
3812 'BF': 'Burkina Faso',
3818 'KY': 'Cayman Islands',
3819 'CF': 'Central African Republic',
3823 'CX': 'Christmas Island',
3824 'CC': 'Cocos (Keeling) Islands',
3828 'CD': 'Congo, the Democratic Republic of the',
3829 'CK': 'Cook Islands',
3831 'CI': 'Côte d\'Ivoire',
3836 'CZ': 'Czech Republic',
3840 'DO': 'Dominican Republic',
3843 'SV': 'El Salvador',
3844 'GQ': 'Equatorial Guinea',
3848 'FK': 'Falkland Islands (Malvinas)',
3849 'FO': 'Faroe Islands',
3853 'GF': 'French Guiana',
3854 'PF': 'French Polynesia',
3855 'TF': 'French Southern Territories',
3870 'GW': 'Guinea-Bissau',
3873 'HM': 'Heard Island and McDonald Islands',
3874 'VA': 'Holy See (Vatican City State)',
3881 'IR': 'Iran, Islamic Republic of',
3884 'IM': 'Isle of Man',
3894 'KP': 'Korea, Democratic People\'s Republic of',
3895 'KR': 'Korea, Republic of',
3898 'LA': 'Lao People\'s Democratic Republic',
3904 'LI': 'Liechtenstein',
3908 'MK': 'Macedonia, the Former Yugoslav Republic of',
3915 'MH': 'Marshall Islands',
3921 'FM': 'Micronesia, Federated States of',
3922 'MD': 'Moldova, Republic of',
3933 'NL': 'Netherlands',
3934 'NC': 'New Caledonia',
3935 'NZ': 'New Zealand',
3940 'NF': 'Norfolk Island',
3941 'MP': 'Northern Mariana Islands',
3946 'PS': 'Palestine, State of',
3948 'PG': 'Papua New Guinea',
3951 'PH': 'Philippines',
3955 'PR': 'Puerto Rico',
3959 'RU': 'Russian Federation',
3961 'BL': 'Saint Barthélemy',
3962 'SH': 'Saint Helena, Ascension and Tristan da Cunha',
3963 'KN': 'Saint Kitts and Nevis',
3964 'LC': 'Saint Lucia',
3965 'MF': 'Saint Martin (French part)',
3966 'PM': 'Saint Pierre and Miquelon',
3967 'VC': 'Saint Vincent and the Grenadines',
3970 'ST': 'Sao Tome and Principe',
3971 'SA': 'Saudi Arabia',
3975 'SL': 'Sierra Leone',
3977 'SX': 'Sint Maarten (Dutch part)',
3980 'SB': 'Solomon Islands',
3982 'ZA': 'South Africa',
3983 'GS': 'South Georgia and the South Sandwich Islands',
3984 'SS': 'South Sudan',
3989 'SJ': 'Svalbard and Jan Mayen',
3992 'CH': 'Switzerland',
3993 'SY': 'Syrian Arab Republic',
3994 'TW': 'Taiwan, Province of China',
3996 'TZ': 'Tanzania, United Republic of',
3998 'TL': 'Timor-Leste',
4002 'TT': 'Trinidad and Tobago',
4005 'TM': 'Turkmenistan',
4006 'TC': 'Turks and Caicos Islands',
4010 'AE': 'United Arab Emirates',
4011 'GB': 'United Kingdom',
4012 'US': 'United States',
4013 'UM': 'United States Minor Outlying Islands',
4017 'VE': 'Venezuela, Bolivarian Republic of',
4019 'VG': 'Virgin Islands, British',
4020 'VI': 'Virgin Islands, U.S.',
4021 'WF': 'Wallis and Futuna',
4022 'EH': 'Western Sahara',
4026 # Not ISO 3166 codes, but used for IP blocks
4027 'AP': 'Asia/Pacific Region',
4032 def short2full(cls
, code
):
4033 """Convert an ISO 3166-2 country code to the corresponding full name"""
4034 return cls
._country
_map
.get(code
.upper())
4038 # Major IPv4 address blocks per country
4040 'AD': '46.172.224.0/19',
4041 'AE': '94.200.0.0/13',
4042 'AF': '149.54.0.0/17',
4043 'AG': '209.59.64.0/18',
4044 'AI': '204.14.248.0/21',
4045 'AL': '46.99.0.0/16',
4046 'AM': '46.70.0.0/15',
4047 'AO': '105.168.0.0/13',
4048 'AP': '182.50.184.0/21',
4049 'AQ': '23.154.160.0/24',
4050 'AR': '181.0.0.0/12',
4051 'AS': '202.70.112.0/20',
4052 'AT': '77.116.0.0/14',
4053 'AU': '1.128.0.0/11',
4054 'AW': '181.41.0.0/18',
4055 'AX': '185.217.4.0/22',
4056 'AZ': '5.197.0.0/16',
4057 'BA': '31.176.128.0/17',
4058 'BB': '65.48.128.0/17',
4059 'BD': '114.130.0.0/16',
4061 'BF': '102.178.0.0/15',
4062 'BG': '95.42.0.0/15',
4063 'BH': '37.131.0.0/17',
4064 'BI': '154.117.192.0/18',
4065 'BJ': '137.255.0.0/16',
4066 'BL': '185.212.72.0/23',
4067 'BM': '196.12.64.0/18',
4068 'BN': '156.31.0.0/16',
4069 'BO': '161.56.0.0/16',
4070 'BQ': '161.0.80.0/20',
4071 'BR': '191.128.0.0/12',
4072 'BS': '24.51.64.0/18',
4073 'BT': '119.2.96.0/19',
4074 'BW': '168.167.0.0/16',
4075 'BY': '178.120.0.0/13',
4076 'BZ': '179.42.192.0/18',
4077 'CA': '99.224.0.0/11',
4078 'CD': '41.243.0.0/16',
4079 'CF': '197.242.176.0/21',
4080 'CG': '160.113.0.0/16',
4081 'CH': '85.0.0.0/13',
4082 'CI': '102.136.0.0/14',
4083 'CK': '202.65.32.0/19',
4084 'CL': '152.172.0.0/14',
4085 'CM': '102.244.0.0/14',
4086 'CN': '36.128.0.0/10',
4087 'CO': '181.240.0.0/12',
4088 'CR': '201.192.0.0/12',
4089 'CU': '152.206.0.0/15',
4090 'CV': '165.90.96.0/19',
4091 'CW': '190.88.128.0/17',
4092 'CY': '31.153.0.0/16',
4093 'CZ': '88.100.0.0/14',
4095 'DJ': '197.241.0.0/17',
4096 'DK': '87.48.0.0/12',
4097 'DM': '192.243.48.0/20',
4098 'DO': '152.166.0.0/15',
4099 'DZ': '41.96.0.0/12',
4100 'EC': '186.68.0.0/15',
4101 'EE': '90.190.0.0/15',
4102 'EG': '156.160.0.0/11',
4103 'ER': '196.200.96.0/20',
4104 'ES': '88.0.0.0/11',
4105 'ET': '196.188.0.0/14',
4106 'EU': '2.16.0.0/13',
4107 'FI': '91.152.0.0/13',
4108 'FJ': '144.120.0.0/16',
4109 'FK': '80.73.208.0/21',
4110 'FM': '119.252.112.0/20',
4111 'FO': '88.85.32.0/19',
4113 'GA': '41.158.0.0/15',
4115 'GD': '74.122.88.0/21',
4116 'GE': '31.146.0.0/16',
4117 'GF': '161.22.64.0/18',
4118 'GG': '62.68.160.0/19',
4119 'GH': '154.160.0.0/12',
4120 'GI': '95.164.0.0/16',
4121 'GL': '88.83.0.0/19',
4122 'GM': '160.182.0.0/15',
4123 'GN': '197.149.192.0/18',
4124 'GP': '104.250.0.0/19',
4125 'GQ': '105.235.224.0/20',
4126 'GR': '94.64.0.0/13',
4127 'GT': '168.234.0.0/16',
4128 'GU': '168.123.0.0/16',
4129 'GW': '197.214.80.0/20',
4130 'GY': '181.41.64.0/18',
4131 'HK': '113.252.0.0/14',
4132 'HN': '181.210.0.0/16',
4133 'HR': '93.136.0.0/13',
4134 'HT': '148.102.128.0/17',
4135 'HU': '84.0.0.0/14',
4136 'ID': '39.192.0.0/10',
4137 'IE': '87.32.0.0/12',
4138 'IL': '79.176.0.0/13',
4139 'IM': '5.62.80.0/20',
4140 'IN': '117.192.0.0/10',
4141 'IO': '203.83.48.0/21',
4142 'IQ': '37.236.0.0/14',
4143 'IR': '2.176.0.0/12',
4144 'IS': '82.221.0.0/16',
4145 'IT': '79.0.0.0/10',
4146 'JE': '87.244.64.0/18',
4147 'JM': '72.27.0.0/17',
4148 'JO': '176.29.0.0/16',
4149 'JP': '133.0.0.0/8',
4150 'KE': '105.48.0.0/12',
4151 'KG': '158.181.128.0/17',
4152 'KH': '36.37.128.0/17',
4153 'KI': '103.25.140.0/22',
4154 'KM': '197.255.224.0/20',
4155 'KN': '198.167.192.0/19',
4156 'KP': '175.45.176.0/22',
4157 'KR': '175.192.0.0/10',
4158 'KW': '37.36.0.0/14',
4159 'KY': '64.96.0.0/15',
4160 'KZ': '2.72.0.0/13',
4161 'LA': '115.84.64.0/18',
4162 'LB': '178.135.0.0/16',
4163 'LC': '24.92.144.0/20',
4164 'LI': '82.117.0.0/19',
4165 'LK': '112.134.0.0/15',
4166 'LR': '102.183.0.0/16',
4167 'LS': '129.232.0.0/17',
4168 'LT': '78.56.0.0/13',
4169 'LU': '188.42.0.0/16',
4170 'LV': '46.109.0.0/16',
4171 'LY': '41.252.0.0/14',
4172 'MA': '105.128.0.0/11',
4173 'MC': '88.209.64.0/18',
4174 'MD': '37.246.0.0/16',
4175 'ME': '178.175.0.0/17',
4176 'MF': '74.112.232.0/21',
4177 'MG': '154.126.0.0/17',
4178 'MH': '117.103.88.0/21',
4179 'MK': '77.28.0.0/15',
4180 'ML': '154.118.128.0/18',
4181 'MM': '37.111.0.0/17',
4182 'MN': '49.0.128.0/17',
4183 'MO': '60.246.0.0/16',
4184 'MP': '202.88.64.0/20',
4185 'MQ': '109.203.224.0/19',
4186 'MR': '41.188.64.0/18',
4187 'MS': '208.90.112.0/22',
4188 'MT': '46.11.0.0/16',
4189 'MU': '105.16.0.0/12',
4190 'MV': '27.114.128.0/18',
4191 'MW': '102.70.0.0/15',
4192 'MX': '187.192.0.0/11',
4193 'MY': '175.136.0.0/13',
4194 'MZ': '197.218.0.0/15',
4195 'NA': '41.182.0.0/16',
4196 'NC': '101.101.0.0/18',
4197 'NE': '197.214.0.0/18',
4198 'NF': '203.17.240.0/22',
4199 'NG': '105.112.0.0/12',
4200 'NI': '186.76.0.0/15',
4201 'NL': '145.96.0.0/11',
4202 'NO': '84.208.0.0/13',
4203 'NP': '36.252.0.0/15',
4204 'NR': '203.98.224.0/19',
4205 'NU': '49.156.48.0/22',
4206 'NZ': '49.224.0.0/14',
4207 'OM': '5.36.0.0/15',
4208 'PA': '186.72.0.0/15',
4209 'PE': '186.160.0.0/14',
4210 'PF': '123.50.64.0/18',
4211 'PG': '124.240.192.0/19',
4212 'PH': '49.144.0.0/13',
4213 'PK': '39.32.0.0/11',
4214 'PL': '83.0.0.0/11',
4215 'PM': '70.36.0.0/20',
4216 'PR': '66.50.0.0/16',
4217 'PS': '188.161.0.0/16',
4218 'PT': '85.240.0.0/13',
4219 'PW': '202.124.224.0/20',
4220 'PY': '181.120.0.0/14',
4221 'QA': '37.210.0.0/15',
4222 'RE': '102.35.0.0/16',
4223 'RO': '79.112.0.0/13',
4224 'RS': '93.86.0.0/15',
4225 'RU': '5.136.0.0/13',
4226 'RW': '41.186.0.0/16',
4227 'SA': '188.48.0.0/13',
4228 'SB': '202.1.160.0/19',
4229 'SC': '154.192.0.0/11',
4230 'SD': '102.120.0.0/13',
4231 'SE': '78.64.0.0/12',
4232 'SG': '8.128.0.0/10',
4233 'SI': '188.196.0.0/14',
4234 'SK': '78.98.0.0/15',
4235 'SL': '102.143.0.0/17',
4236 'SM': '89.186.32.0/19',
4237 'SN': '41.82.0.0/15',
4238 'SO': '154.115.192.0/18',
4239 'SR': '186.179.128.0/17',
4240 'SS': '105.235.208.0/21',
4241 'ST': '197.159.160.0/19',
4242 'SV': '168.243.0.0/16',
4243 'SX': '190.102.0.0/20',
4245 'SZ': '41.84.224.0/19',
4246 'TC': '65.255.48.0/20',
4247 'TD': '154.68.128.0/19',
4248 'TG': '196.168.0.0/14',
4249 'TH': '171.96.0.0/13',
4250 'TJ': '85.9.128.0/18',
4251 'TK': '27.96.24.0/21',
4252 'TL': '180.189.160.0/20',
4253 'TM': '95.85.96.0/19',
4254 'TN': '197.0.0.0/11',
4255 'TO': '175.176.144.0/21',
4256 'TR': '78.160.0.0/11',
4257 'TT': '186.44.0.0/15',
4258 'TV': '202.2.96.0/19',
4259 'TW': '120.96.0.0/11',
4260 'TZ': '156.156.0.0/14',
4261 'UA': '37.52.0.0/14',
4262 'UG': '102.80.0.0/13',
4264 'UY': '167.56.0.0/13',
4265 'UZ': '84.54.64.0/18',
4266 'VA': '212.77.0.0/19',
4267 'VC': '207.191.240.0/21',
4268 'VE': '186.88.0.0/13',
4269 'VG': '66.81.192.0/20',
4270 'VI': '146.226.0.0/16',
4271 'VN': '14.160.0.0/11',
4272 'VU': '202.80.32.0/20',
4273 'WF': '117.20.32.0/21',
4274 'WS': '202.4.32.0/19',
4275 'YE': '134.35.0.0/16',
4276 'YT': '41.242.116.0/22',
4277 'ZA': '41.0.0.0/11',
4278 'ZM': '102.144.0.0/13',
4279 'ZW': '102.177.192.0/18',
4283 def random_ipv4(cls
, code_or_block
):
4284 if len(code_or_block
) == 2:
4285 block
= cls
._country
_ip
_map
.get(code_or_block
.upper())
4289 block
= code_or_block
4290 addr
, preflen
= block
.split('/')
4291 addr_min
= struct
.unpack('!L', socket
.inet_aton(addr
))[0]
4292 addr_max
= addr_min |
(0xffffffff >> int(preflen
))
4293 return str(socket
.inet_ntoa(
4294 struct
.pack('!L', random
.randint(addr_min
, addr_max
))))
4297 # Both long_to_bytes and bytes_to_long are adapted from PyCrypto, which is
4298 # released into Public Domain
4299 # https://github.com/dlitz/pycrypto/blob/master/lib/Crypto/Util/number.py#L387
4301 def long_to_bytes(n
, blocksize
=0):
4302 """long_to_bytes(n:long, blocksize:int) : string
4303 Convert a long integer to a byte string.
4305 If optional blocksize is given and greater than zero, pad the front of the
4306 byte string with binary zeros so that the length is a multiple of
4309 # after much testing, this algorithm was deemed to be the fastest
4313 s
= struct
.pack('>I', n
& 0xffffffff) + s
4315 # strip off leading zeros
4316 for i
in range(len(s
)):
4317 if s
[i
] != b
'\000'[0]:
4320 # only happens when n == 0
4324 # add back some pad bytes. this could be done more efficiently w.r.t. the
4325 # de-padding being done above, but sigh...
4326 if blocksize
> 0 and len(s
) % blocksize
:
4327 s
= (blocksize
- len(s
) % blocksize
) * b
'\000' + s
4331 def bytes_to_long(s
):
4332 """bytes_to_long(string) : long
4333 Convert a byte string to a long integer.
4335 This is (essentially) the inverse of long_to_bytes().
4340 extra
= (4 - length
% 4)
4341 s
= b
'\000' * extra
+ s
4342 length
= length
+ extra
4343 for i
in range(0, length
, 4):
4344 acc
= (acc
<< 32) + struct
.unpack('>I', s
[i
:i
+ 4])[0]
4348 def ohdave_rsa_encrypt(data
, exponent
, modulus
):
4350 Implement OHDave's RSA algorithm. See http://www.ohdave.com/rsa/
4353 data: data to encrypt, bytes-like object
4354 exponent, modulus: parameter e and N of RSA algorithm, both integer
4355 Output: hex string of encrypted data
4357 Limitation: supports one block encryption only
4360 payload
= int(binascii
.hexlify(data
[::-1]), 16)
4361 encrypted
= pow(payload
, exponent
, modulus
)
4362 return f
'{encrypted:x}'
4365 def pkcs1pad(data
, length
):
4367 Padding input data with PKCS#1 scheme
4369 @param {int[]} data input data
4370 @param {int} length target length
4371 @returns {int[]} padded data
4373 if len(data
) > length
- 11:
4374 raise ValueError('Input data too long for PKCS#1 padding')
4376 pseudo_random
= [random
.randint(0, 254) for _
in range(length
- len(data
) - 3)]
4377 return [0, 2, *pseudo_random
, 0, *data
]
4380 def _base_n_table(n
, table
):
4381 if not table
and not n
:
4382 raise ValueError('Either table or n must be specified')
4383 table
= (table
or '0123456789abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ')[:n
]
4385 if n
and n
!= len(table
):
4386 raise ValueError(f
'base {n} exceeds table length {len(table)}')
4390 def encode_base_n(num
, n
=None, table
=None):
4391 """Convert given int to a base-n string"""
4392 table
= _base_n_table(n
, table
)
4396 result
, base
= '', len(table
)
4398 result
= table
[num
% base
] + result
4403 def decode_base_n(string
, n
=None, table
=None):
4404 """Convert given base-n string to int"""
4405 table
= {char: index for index, char in enumerate(_base_n_table(n, table))}
4406 result
, base
= 0, len(table
)
4408 result
= result
* base
+ table
[char
]
4412 def decode_packed_codes(code
):
4413 mobj
= re
.search(PACKED_CODES_RE
, code
)
4414 obfuscated_code
, base
, count
, symbols
= mobj
.groups()
4417 symbols
= symbols
.split('|')
4422 base_n_count
= encode_base_n(count
, base
)
4423 symbol_table
[base_n_count
] = symbols
[count
] or base_n_count
4426 r
'\b(\w+)\b', lambda mobj
: symbol_table
[mobj
.group(0)],
4430 def caesar(s
, alphabet
, shift
):
4435 alphabet
[(alphabet
.index(c
) + shift
) % l
] if c
in alphabet
else c
4440 return caesar(s
, r
'''!"#$%&'()*+,-./0123456789:;<=>?@ABCDEFGHIJKLMNOPQRSTUVWXYZ[\]^_`abcdefghijklmnopqrstuvwxyz{|}~''', 47)
4443 def parse_m3u8_attributes(attrib
):
4445 for (key
, val
) in re
.findall(r
'(?P<key>[A-Z0-9-]+)=(?P<val>"[^"]+"|[^",]+)(?:,|$)', attrib
):
4446 if val
.startswith('"'):
4452 def urshift(val
, n
):
4453 return val
>> n
if val
>= 0 else (val
+ 0x100000000) >> n
4456 def write_xattr(path
, key
, value
):
4457 # Windows: Write xattrs to NTFS Alternate Data Streams:
4458 # http://en.wikipedia.org/wiki/NTFS#Alternate_data_streams_.28ADS.29
4459 if compat_os_name
== 'nt':
4460 assert ':' not in key
4461 assert os
.path
.exists(path
)
4464 with open(f
'{path}:{key}', 'wb') as f
:
4466 except OSError as e
:
4467 raise XAttrMetadataError(e
.errno
, e
.strerror
)
4470 # UNIX Method 1. Use os.setxattr/xattrs/pyxattrs modules
4473 if callable(getattr(os
, 'setxattr', None)):
4474 setxattr
= os
.setxattr
4475 elif getattr(xattr
, '_yt_dlp__identifier', None) == 'pyxattr':
4476 # Unicode arguments are not supported in pyxattr until version 0.5.0
4477 # See https://github.com/ytdl-org/youtube-dl/issues/5498
4478 if version_tuple(xattr
.__version
__) >= (0, 5, 0):
4479 setxattr
= xattr
.set
4481 setxattr
= xattr
.setxattr
4485 setxattr(path
, key
, value
)
4486 except OSError as e
:
4487 raise XAttrMetadataError(e
.errno
, e
.strerror
)
4490 # UNIX Method 2. Use setfattr/xattr executables
4491 exe
= ('setfattr' if check_executable('setfattr', ['--version'])
4492 else 'xattr' if check_executable('xattr', ['-h']) else None)
4494 raise XAttrUnavailableError(
4495 'Couldn\'t find a tool to set the xattrs. Install either the "xattr" or "pyxattr" Python modules or the '
4496 + ('"xattr" binary' if sys
.platform
!= 'linux' else 'GNU "attr" package (which contains the "setfattr" tool)'))
4498 value
= value
.decode()
4500 _
, stderr
, returncode
= Popen
.run(
4501 [exe
, '-w', key
, value
, path
] if exe
== 'xattr' else [exe
, '-n', key
, '-v', value
, path
],
4502 text
=True, stdout
=subprocess
.PIPE
, stderr
=subprocess
.PIPE
, stdin
=subprocess
.PIPE
)
4503 except OSError as e
:
4504 raise XAttrMetadataError(e
.errno
, e
.strerror
)
4506 raise XAttrMetadataError(returncode
, stderr
)
4509 def random_birthday(year_field
, month_field
, day_field
):
4510 start_date
= dt
.date(1950, 1, 1)
4511 end_date
= dt
.date(1995, 12, 31)
4512 offset
= random
.randint(0, (end_date
- start_date
).days
)
4513 random_date
= start_date
+ dt
.timedelta(offset
)
4515 year_field
: str(random_date
.year
),
4516 month_field
: str(random_date
.month
),
4517 day_field
: str(random_date
.day
),
4521 def find_available_port(interface
=''):
4523 with socket
.socket() as sock
:
4524 sock
.bind((interface
, 0))
4525 return sock
.getsockname()[1]
4530 # Templates for internet shortcut files, which are plain text files.
4531 DOT_URL_LINK_TEMPLATE
= '''\
4536 DOT_WEBLOC_LINK_TEMPLATE
= '''\
4537 <?xml version="1.0" encoding="UTF-8"?>
4538 <!DOCTYPE plist PUBLIC "-//Apple//DTD PLIST 1.0//EN" "http://www.apple.com/DTDs/PropertyList-1.0.dtd">
4539 <plist version="1.0">
4542 \t<string>%(url)s</string>
4547 DOT_DESKTOP_LINK_TEMPLATE
= '''\
4557 'url': DOT_URL_LINK_TEMPLATE
,
4558 'desktop': DOT_DESKTOP_LINK_TEMPLATE
,
4559 'webloc': DOT_WEBLOC_LINK_TEMPLATE
,
4563 def iri_to_uri(iri
):
4565 Converts an IRI (Internationalized Resource Identifier, allowing Unicode characters) to a URI (Uniform Resource Identifier, ASCII-only).
4567 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.
4570 iri_parts
= urllib
.parse
.urlparse(iri
)
4572 if '[' in iri_parts
.netloc
:
4573 raise ValueError('IPv6 URIs are not, yet, supported.')
4574 # Querying `.netloc`, when there's only one bracket, also raises a ValueError.
4576 # 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.
4579 if iri_parts
.username
:
4580 net_location
+= urllib
.parse
.quote(iri_parts
.username
, safe
=r
"!$%&'()*+,~")
4581 if iri_parts
.password
is not None:
4582 net_location
+= ':' + urllib
.parse
.quote(iri_parts
.password
, safe
=r
"!$%&'()*+,~")
4585 net_location
+= iri_parts
.hostname
.encode('idna').decode() # Punycode for Unicode hostnames.
4586 # The 'idna' encoding produces ASCII text.
4587 if iri_parts
.port
is not None and iri_parts
.port
!= 80:
4588 net_location
+= ':' + str(iri_parts
.port
)
4590 return urllib
.parse
.urlunparse(
4594 urllib
.parse
.quote_plus(iri_parts
.path
, safe
=r
"!$%&'()*+,/:;=@|~"),
4596 # Unsure about the `safe` argument, since this is a legacy way of handling parameters.
4597 urllib
.parse
.quote_plus(iri_parts
.params
, safe
=r
"!$%&'()*+,/:;=@|~"),
4599 # Not totally sure about the `safe` argument, since the source does not explicitly mention the query URI component.
4600 urllib
.parse
.quote_plus(iri_parts
.query
, safe
=r
"!$%&'()*+,/:;=?@{|}~"),
4602 urllib
.parse
.quote_plus(iri_parts
.fragment
, safe
=r
"!#$%&'()*+,/:;=?@{|}~")))
4604 # Source for `safe` arguments: https://url.spec.whatwg.org/#percent-encoded-bytes.
4607 def to_high_limit_path(path
):
4608 if sys
.platform
in ['win32', 'cygwin']:
4609 # Work around MAX_PATH limitation on Windows. The maximum allowed length for the individual path segments may still be quite limited.
4610 return '\\\\?\\' + os
.path
.abspath(path
)
4615 def format_field(obj
, field
=None, template
='%s', ignore
=NO_DEFAULT
, default
='', func
=IDENTITY
):
4616 val
= traversal
.traverse_obj(obj
, *variadic(field
))
4617 if not val
if ignore
is NO_DEFAULT
else val
in variadic(ignore
):
4619 return template
% func(val
)
4622 def clean_podcast_url(url
):
4623 url
= re
.sub(r
'''(?x)
4627 media\.blubrry\.com| # https://create.blubrry.com/resources/podcast-media-download-statistics/getting-started/
4632 (?:dts|www)\.podtrac\.com/(?:pts/)?redirect\.[0-9a-z]{3,4}| # http://analytics.podtrac.com/how-to-measure
4635 cn\.co| # https://podcorn.com/analytics-prefix/
4636 st\.fm # https://podsights.com/docs/
4641 return re
.sub(r
'^\w+://(\w+://)', r
'\1', url
)
4644 _HEX_TABLE
= '0123456789abcdef'
4647 def random_uuidv4():
4648 return re
.sub(r
'[xy]', lambda x
: _HEX_TABLE
[random
.randint(0, 15)], 'xxxxxxxx-xxxx-4xxx-yxxx-xxxxxxxxxxxx')
4651 def make_dir(path
, to_screen
=None):
4653 dn
= os
.path
.dirname(path
)
4655 os
.makedirs(dn
, exist_ok
=True)
4657 except OSError as err
:
4658 if callable(to_screen
) is not None:
4659 to_screen(f
'unable to create directory {err}')
4663 def get_executable_path():
4664 from ..update
import _get_variant_and_executable_path
4666 return os
.path
.dirname(os
.path
.abspath(_get_variant_and_executable_path()[1]))
4669 def get_user_config_dirs(package_name
):
4670 # .config (e.g. ~/.config/package_name)
4671 xdg_config_home
= os
.getenv('XDG_CONFIG_HOME') or compat_expanduser('~/.config')
4672 yield os
.path
.join(xdg_config_home
, package_name
)
4674 # appdata (%APPDATA%/package_name)
4675 appdata_dir
= os
.getenv('appdata')
4677 yield os
.path
.join(appdata_dir
, package_name
)
4679 # home (~/.package_name)
4680 yield os
.path
.join(compat_expanduser('~'), f
'.{package_name}')
4683 def get_system_config_dirs(package_name
):
4685 yield os
.path
.join('/etc', package_name
)
4688 def time_seconds(**kwargs
):
4690 Returns TZ-aware time in seconds since the epoch (1970-01-01T00:00:00Z)
4692 return time
.time() + dt
.timedelta(**kwargs
).total_seconds()
4695 # create a JSON Web Signature (jws) with HS256 algorithm
4696 # the resulting format is in JWS Compact Serialization
4697 # implemented following JWT https://www.rfc-editor.org/rfc/rfc7519.html
4698 # implemented following JWS https://www.rfc-editor.org/rfc/rfc7515.html
4699 def jwt_encode_hs256(payload_data
, key
, headers
={}):
4705 header_data
.update(headers
)
4706 header_b64
= base64
.b64encode(json
.dumps(header_data
).encode())
4707 payload_b64
= base64
.b64encode(json
.dumps(payload_data
).encode())
4708 h
= hmac
.new(key
.encode(), header_b64
+ b
'.' + payload_b64
, hashlib
.sha256
)
4709 signature_b64
= base64
.b64encode(h
.digest())
4710 return header_b64
+ b
'.' + payload_b64
+ b
'.' + signature_b64
4713 # can be extended in future to verify the signature and parse header and return the algorithm used if it's not HS256
4714 def jwt_decode_hs256(jwt
):
4715 header_b64
, payload_b64
, signature_b64
= jwt
.split('.')
4716 # add trailing ='s that may have been stripped, superfluous ='s are ignored
4717 return json
.loads(base64
.urlsafe_b64decode(f
'{payload_b64}==='))
4720 WINDOWS_VT_MODE
= False if compat_os_name
== 'nt' else None
4724 def supports_terminal_sequences(stream
):
4725 if compat_os_name
== 'nt':
4726 if not WINDOWS_VT_MODE
:
4728 elif not os
.getenv('TERM'):
4731 return stream
.isatty()
4732 except BaseException
:
4736 def windows_enable_vt_mode():
4737 """Ref: https://bugs.python.org/issue30075 """
4738 if get_windows_version() < (10, 0, 10586):
4742 import ctypes
.wintypes
4745 ENABLE_VIRTUAL_TERMINAL_PROCESSING
= 0x0004
4747 dll
= ctypes
.WinDLL('kernel32', use_last_error
=False)
4748 handle
= os
.open('CONOUT$', os
.O_RDWR
)
4750 h_out
= ctypes
.wintypes
.HANDLE(msvcrt
.get_osfhandle(handle
))
4751 dw_original_mode
= ctypes
.wintypes
.DWORD()
4752 success
= dll
.GetConsoleMode(h_out
, ctypes
.byref(dw_original_mode
))
4754 raise Exception('GetConsoleMode failed')
4756 success
= dll
.SetConsoleMode(h_out
, ctypes
.wintypes
.DWORD(
4757 dw_original_mode
.value | ENABLE_VIRTUAL_TERMINAL_PROCESSING
))
4759 raise Exception('SetConsoleMode failed')
4763 global WINDOWS_VT_MODE
4764 WINDOWS_VT_MODE
= True
4765 supports_terminal_sequences
.cache_clear()
4768 _terminal_sequences_re
= re
.compile('\033\\[[^m]+m')
4771 def remove_terminal_sequences(string
):
4772 return _terminal_sequences_re
.sub('', string
)
4775 def number_of_digits(number
):
4776 return len('%d' % number
)
4779 def join_nonempty(*values
, delim
='-', from_dict
=None):
4780 if from_dict
is not None:
4781 values
= (traversal
.traverse_obj(from_dict
, variadic(v
)) for v
in values
)
4782 return delim
.join(map(str, filter(None, values
)))
4785 def scale_thumbnails_to_max_format_width(formats
, thumbnails
, url_width_re
):
4787 Find the largest format dimensions in terms of video width and, for each thumbnail:
4788 * Modify the URL: Match the width with the provided regex and replace with the former width
4791 This function is useful with video services that scale the provided thumbnails on demand
4793 _keys
= ('width', 'height')
4794 max_dimensions
= max(
4795 (tuple(fmt
.get(k
) or 0 for k
in _keys
) for fmt
in formats
),
4797 if not max_dimensions
[0]:
4801 {'url': re.sub(url_width_re, str(max_dimensions[0]), thumbnail['url'])}
,
4802 dict(zip(_keys
, max_dimensions
)), thumbnail
)
4803 for thumbnail
in thumbnails
4807 def parse_http_range(range):
4808 """ Parse value of "Range" or "Content-Range" HTTP header into tuple. """
4810 return None, None, None
4811 crg
= re
.search(r
'bytes[ =](\d+)-(\d+)?(?:/(\d+))?', range)
4813 return None, None, None
4814 return int(crg
.group(1)), int_or_none(crg
.group(2)), int_or_none(crg
.group(3))
4817 def read_stdin(what
):
4819 eof
= 'Ctrl+Z' if compat_os_name
== 'nt' else 'Ctrl+D'
4820 write_string(f
'Reading {what} from STDIN - EOF ({eof}) to end:\n')
4824 def determine_file_encoding(data
):
4826 Detect the text encoding used
4827 @returns (encoding, bytes to skip)
4830 # BOM marks are given priority over declarations
4831 for bom
, enc
in BOMS
:
4832 if data
.startswith(bom
):
4833 return enc
, len(bom
)
4835 # Strip off all null bytes to match even when UTF-16 or UTF-32 is used.
4836 # We ignore the endianness to get a good enough match
4837 data
= data
.replace(b
'\0', b
'')
4838 mobj
= re
.match(rb
'(?m)^#\s*coding\s*:\s*(\S+)\s*$', data
)
4839 return mobj
.group(1).decode() if mobj
else None, 0
4846 __initialized
= False
4848 def __init__(self
, parser
, label
=None):
4849 self
.parser
, self
.label
= parser
, label
4850 self
._loaded
_paths
, self
.configs
= set(), []
4852 def init(self
, args
=None, filename
=None):
4853 assert not self
.__initialized
4854 self
.own_args
, self
.filename
= args
, filename
4855 return self
.load_configs()
4857 def load_configs(self
):
4860 location
= os
.path
.realpath(self
.filename
)
4861 directory
= os
.path
.dirname(location
)
4862 if location
in self
._loaded
_paths
:
4864 self
._loaded
_paths
.add(location
)
4866 self
.__initialized
= True
4867 opts
, _
= self
.parser
.parse_known_args(self
.own_args
)
4868 self
.parsed_args
= self
.own_args
4869 for location
in opts
.config_locations
or []:
4871 if location
in self
._loaded
_paths
:
4873 self
._loaded
_paths
.add(location
)
4874 self
.append_config(shlex
.split(read_stdin('options'), comments
=True), label
='stdin')
4876 location
= os
.path
.join(directory
, expand_path(location
))
4877 if os
.path
.isdir(location
):
4878 location
= os
.path
.join(location
, 'yt-dlp.conf')
4879 if not os
.path
.exists(location
):
4880 self
.parser
.error(f
'config location {location} does not exist')
4881 self
.append_config(self
.read_file(location
), location
)
4885 label
= join_nonempty(
4886 self
.label
, 'config', f
'"{self.filename}"' if self
.filename
else '',
4888 return join_nonempty(
4889 self
.own_args
is not None and f
'{label[0].upper()}{label[1:]}: {self.hide_login_info(self.own_args)}',
4890 *(f
'\n{c}'.replace('\n', '\n| ')[1:] for c
in self
.configs
),
4894 def read_file(filename
, default
=[]):
4896 optionf
= open(filename
, 'rb')
4898 return default
# silently skip if file is not present
4900 enc
, skip
= determine_file_encoding(optionf
.read(512))
4901 optionf
.seek(skip
, io
.SEEK_SET
)
4903 enc
= None # silently skip read errors
4905 # FIXME: https://github.com/ytdl-org/youtube-dl/commit/dfe5fa49aed02cf36ba9f743b11b0903554b5e56
4906 contents
= optionf
.read().decode(enc
or preferredencoding())
4907 res
= shlex
.split(contents
, comments
=True)
4908 except Exception as err
:
4909 raise ValueError(f
'Unable to parse "{filename}": {err}')
4915 def hide_login_info(opts
):
4916 PRIVATE_OPTS
= {'-p', '--password', '-u', '--username', '--video-password', '--ap-password', '--ap-username'}
4917 eqre
= re
.compile('^(?P<key>' + ('|'.join(re
.escape(po
) for po
in PRIVATE_OPTS
)) + ')=.+$')
4922 return m
.group('key') + '=PRIVATE'
4926 opts
= list(map(_scrub_eq
, opts
))
4927 for idx
, opt
in enumerate(opts
):
4928 if opt
in PRIVATE_OPTS
and idx
+ 1 < len(opts
):
4929 opts
[idx
+ 1] = 'PRIVATE'
4932 def append_config(self
, *args
, label
=None):
4933 config
= type(self
)(self
.parser
, label
)
4934 config
._loaded
_paths
= self
._loaded
_paths
4935 if config
.init(*args
):
4936 self
.configs
.append(config
)
4940 for config
in reversed(self
.configs
):
4941 yield from config
.all_args
4942 yield from self
.parsed_args
or []
4944 def parse_known_args(self
, **kwargs
):
4945 return self
.parser
.parse_known_args(self
.all_args
, **kwargs
)
4947 def parse_args(self
):
4948 return self
.parser
.parse_args(self
.all_args
)
4951 def merge_headers(*dicts
):
4952 """Merge dicts of http headers case insensitively, prioritizing the latter ones"""
4953 return {k.title(): v for k, v in itertools.chain.from_iterable(map(dict.items, dicts))}
4956 def cached_method(f
):
4957 """Cache a method"""
4958 signature
= inspect
.signature(f
)
4961 def wrapper(self
, *args
, **kwargs
):
4962 bound_args
= signature
.bind(self
, *args
, **kwargs
)
4963 bound_args
.apply_defaults()
4964 key
= tuple(bound_args
.arguments
.values())[1:]
4966 cache
= vars(self
).setdefault('_cached_method__cache', {}).setdefault(f.__name__, {}
)
4967 if key
not in cache
:
4968 cache
[key
] = f(self
, *args
, **kwargs
)
4973 class classproperty
:
4974 """property access for class methods with optional caching"""
4975 def __new__(cls
, func
=None, *args
, **kwargs
):
4977 return functools
.partial(cls
, *args
, **kwargs
)
4978 return super().__new
__(cls
)
4980 def __init__(self
, func
, *, cache
=False):
4981 functools
.update_wrapper(self
, func
)
4983 self
._cache
= {} if cache
else None
4985 def __get__(self
, _
, cls
):
4986 if self
._cache
is None:
4987 return self
.func(cls
)
4988 elif cls
not in self
._cache
:
4989 self
._cache
[cls
] = self
.func(cls
)
4990 return self
._cache
[cls
]
4993 class function_with_repr
:
4994 def __init__(self
, func
, repr_
=None):
4995 functools
.update_wrapper(self
, func
)
4996 self
.func
, self
.__repr
= func
, repr_
4998 def __call__(self
, *args
, **kwargs
):
4999 return self
.func(*args
, **kwargs
)
5002 def set_repr(cls
, repr_
):
5003 return functools
.partial(cls
, repr_
=repr_
)
5008 return f
'{self.func.__module__}.{self.func.__qualname__}'
5011 class Namespace(types
.SimpleNamespace
):
5012 """Immutable namespace"""
5015 return iter(self
.__dict
__.values())
5019 return self
.__dict
__.items()
5022 MEDIA_EXTENSIONS
= Namespace(
5023 common_video
=('avi', 'flv', 'mkv', 'mov', 'mp4', 'webm'),
5024 video
=('3g2', '3gp', 'f4v', 'mk3d', 'divx', 'mpg', 'ogv', 'm4v', 'wmv'),
5025 common_audio
=('aiff', 'alac', 'flac', 'm4a', 'mka', 'mp3', 'ogg', 'opus', 'wav'),
5026 audio
=('aac', 'ape', 'asf', 'f4a', 'f4b', 'm4b', 'm4p', 'm4r', 'oga', 'ogx', 'spx', 'vorbis', 'wma', 'weba'),
5027 thumbnails
=('jpg', 'png', 'webp'),
5028 storyboards
=('mhtml', ),
5029 subtitles
=('srt', 'vtt', 'ass', 'lrc'),
5030 manifests
=('f4f', 'f4m', 'm3u8', 'smil', 'mpd'),
5032 MEDIA_EXTENSIONS
.video
+= MEDIA_EXTENSIONS
.common_video
5033 MEDIA_EXTENSIONS
.audio
+= MEDIA_EXTENSIONS
.common_audio
5035 KNOWN_EXTENSIONS
= (*MEDIA_EXTENSIONS
.video
, *MEDIA_EXTENSIONS
.audio
, *MEDIA_EXTENSIONS
.manifests
)
5040 for retry in RetryManager(...):
5043 except SomeException as err:
5047 attempt
, _error
= 0, None
5049 def __init__(self
, _retries
, _error_callback
, **kwargs
):
5050 self
.retries
= _retries
or 0
5051 self
.error_callback
= functools
.partial(_error_callback
, **kwargs
)
5053 def _should_retry(self
):
5054 return self
._error
is not NO_DEFAULT
and self
.attempt
<= self
.retries
5058 if self
._error
is NO_DEFAULT
:
5063 def error(self
, value
):
5067 while self
._should
_retry
():
5068 self
.error
= NO_DEFAULT
5072 self
.error_callback(self
.error
, self
.attempt
, self
.retries
)
5075 def report_retry(e
, count
, retries
, *, sleep_func
, info
, warn
, error
=None, suffix
=None):
5076 """Utility function for reporting retries"""
5079 return error(f
'{e}. Giving up after {count - 1} retries') if count
> 1 else error(str(e
))
5084 elif isinstance(e
, ExtractorError
):
5085 e
= remove_end(str_or_none(e
.cause
) or e
.orig_msg
, '.')
5086 warn(f
'{e}. Retrying{format_field(suffix, None, " %s")} ({count}/{retries})...')
5088 delay
= float_or_none(sleep_func(n
=count
- 1)) if callable(sleep_func
) else sleep_func
5090 info(f
'Sleeping {delay:.2f} seconds ...')
5094 def make_archive_id(ie
, video_id
):
5095 ie_key
= ie
if isinstance(ie
, str) else ie
.ie_key()
5096 return f
'{ie_key.lower()} {video_id}'
5099 def truncate_string(s
, left
, right
=0):
5100 assert left
> 3 and right
>= 0
5101 if s
is None or len(s
) <= left
+ right
:
5103 return f
'{s[:left - 3]}...{s[-right:] if right else ""}'
5106 def orderedSet_from_options(options
, alias_dict
, *, use_regex
=False, start
=None):
5107 assert 'all' in alias_dict
, '"all" alias is required'
5108 requested
= list(start
or [])
5110 discard
= val
.startswith('-')
5114 if val
in alias_dict
:
5115 val
= alias_dict
[val
] if not discard
else [
5116 i
[1:] if i
.startswith('-') else f
'-{i}' for i
in alias_dict
[val
]]
5117 # NB: Do not allow regex in aliases for performance
5118 requested
= orderedSet_from_options(val
, alias_dict
, start
=requested
)
5121 current
= (filter(re
.compile(val
, re
.I
).fullmatch
, alias_dict
['all']) if use_regex
5122 else [val
] if val
in alias_dict
['all'] else None)
5124 raise ValueError(val
)
5127 for item
in current
:
5128 while item
in requested
:
5129 requested
.remove(item
)
5131 requested
.extend(current
)
5133 return orderedSet(requested
)
5138 regex
= r
' *((?P<reverse>\+)?(?P<field>[a-zA-Z0-9_]+)((?P<separator>[~:])(?P<limit>.*?))?)? *$'
5140 default
= ('hidden', 'aud_or_vid', 'hasvid', 'ie_pref', 'lang', 'quality',
5141 'res', 'fps', 'hdr:12', 'vcodec:vp9.2', 'channels', 'acodec',
5142 'size', 'br', 'asr', 'proto', 'ext', 'hasaud', 'source', 'id') # These must not be aliases
5143 ytdl_default
= ('hasaud', 'lang', 'quality', 'tbr', 'filesize', 'vbr',
5144 'height', 'width', 'proto', 'vext', 'abr', 'aext',
5145 'fps', 'fs_approx', 'source', 'id')
5148 'vcodec': {'type': 'ordered', 'regex': True,
5149 'order': ['av0?1', 'vp0?9.2', 'vp0?9', '[hx]265|he?vc?', '[hx]264|avc', 'vp0?8', 'mp4v|h263', 'theora', '', None, 'none']},
5150 'acodec': {'type': 'ordered', 'regex': True,
5151 'order': ['[af]lac', 'wav|aiff', 'opus', 'vorbis|ogg', 'aac', 'mp?4a?', 'mp3', 'ac-?4', 'e-?a?c-?3', 'ac-?3', 'dts', '', None, 'none']},
5152 'hdr': {'type': 'ordered', 'regex': True, 'field': 'dynamic_range',
5153 'order': ['dv', '(hdr)?12', r
'(hdr)?10\+', '(hdr)?10', 'hlg', '', 'sdr', None]},
5154 'proto': {'type': 'ordered', 'regex': True, 'field': 'protocol',
5155 'order': ['(ht|f)tps', '(ht|f)tp$', 'm3u8.*', '.*dash', 'websocket_frag', 'rtmpe?', '', 'mms|rtsp', 'ws|websocket', 'f4']},
5156 'vext': {'type': 'ordered', 'field': 'video_ext',
5157 'order': ('mp4', 'mov', 'webm', 'flv', '', 'none'),
5158 'order_free': ('webm', 'mp4', 'mov', 'flv', '', 'none')},
5159 'aext': {'type': 'ordered', 'regex': True, 'field': 'audio_ext',
5160 'order': ('m4a', 'aac', 'mp3', 'ogg', 'opus', 'web[am]', '', 'none'),
5161 'order_free': ('ogg', 'opus', 'web[am]', 'mp3', 'm4a', 'aac', '', 'none')},
5162 'hidden': {'visible': False, 'forced': True, 'type': 'extractor', 'max': -1000}
,
5163 'aud_or_vid': {'visible': False, 'forced': True, 'type': 'multiple',
5164 'field': ('vcodec', 'acodec'),
5165 'function': lambda it
: int(any(v
!= 'none' for v
in it
))},
5166 'ie_pref': {'priority': True, 'type': 'extractor'}
,
5167 'hasvid': {'priority': True, 'field': 'vcodec', 'type': 'boolean', 'not_in_list': ('none',)}
,
5168 'hasaud': {'field': 'acodec', 'type': 'boolean', 'not_in_list': ('none',)}
,
5169 'lang': {'convert': 'float', 'field': 'language_preference', 'default': -1}
,
5170 'quality': {'convert': 'float', 'default': -1}
,
5171 'filesize': {'convert': 'bytes'}
,
5172 'fs_approx': {'convert': 'bytes', 'field': 'filesize_approx'}
,
5173 'id': {'convert': 'string', 'field': 'format_id'}
,
5174 'height': {'convert': 'float_none'}
,
5175 'width': {'convert': 'float_none'}
,
5176 'fps': {'convert': 'float_none'}
,
5177 'channels': {'convert': 'float_none', 'field': 'audio_channels'}
,
5178 'tbr': {'convert': 'float_none'}
,
5179 'vbr': {'convert': 'float_none'}
,
5180 'abr': {'convert': 'float_none'}
,
5181 'asr': {'convert': 'float_none'}
,
5182 'source': {'convert': 'float', 'field': 'source_preference', 'default': -1}
,
5184 'codec': {'type': 'combined', 'field': ('vcodec', 'acodec')}
,
5185 'br': {'type': 'multiple', 'field': ('tbr', 'vbr', 'abr'), 'convert': 'float_none',
5186 'function': lambda it
: next(filter(None, it
), None)},
5187 'size': {'type': 'multiple', 'field': ('filesize', 'fs_approx'), 'convert': 'bytes',
5188 'function': lambda it
: next(filter(None, it
), None)},
5189 'ext': {'type': 'combined', 'field': ('vext', 'aext')}
,
5190 'res': {'type': 'multiple', 'field': ('height', 'width'),
5191 'function': lambda it
: min(filter(None, it
), default
=0)},
5193 # Actual field names
5194 'format_id': {'type': 'alias', 'field': 'id'}
,
5195 'preference': {'type': 'alias', 'field': 'ie_pref'}
,
5196 'language_preference': {'type': 'alias', 'field': 'lang'}
,
5197 'source_preference': {'type': 'alias', 'field': 'source'}
,
5198 'protocol': {'type': 'alias', 'field': 'proto'}
,
5199 'filesize_approx': {'type': 'alias', 'field': 'fs_approx'}
,
5200 'audio_channels': {'type': 'alias', 'field': 'channels'}
,
5203 'dimension': {'type': 'alias', 'field': 'res', 'deprecated': True}
,
5204 'resolution': {'type': 'alias', 'field': 'res', 'deprecated': True}
,
5205 'extension': {'type': 'alias', 'field': 'ext', 'deprecated': True}
,
5206 'bitrate': {'type': 'alias', 'field': 'br', 'deprecated': True}
,
5207 'total_bitrate': {'type': 'alias', 'field': 'tbr', 'deprecated': True}
,
5208 'video_bitrate': {'type': 'alias', 'field': 'vbr', 'deprecated': True}
,
5209 'audio_bitrate': {'type': 'alias', 'field': 'abr', 'deprecated': True}
,
5210 'framerate': {'type': 'alias', 'field': 'fps', 'deprecated': True}
,
5211 'filesize_estimate': {'type': 'alias', 'field': 'size', 'deprecated': True}
,
5212 'samplerate': {'type': 'alias', 'field': 'asr', 'deprecated': True}
,
5213 'video_ext': {'type': 'alias', 'field': 'vext', 'deprecated': True}
,
5214 'audio_ext': {'type': 'alias', 'field': 'aext', 'deprecated': True}
,
5215 'video_codec': {'type': 'alias', 'field': 'vcodec', 'deprecated': True}
,
5216 'audio_codec': {'type': 'alias', 'field': 'acodec', 'deprecated': True}
,
5217 'video': {'type': 'alias', 'field': 'hasvid', 'deprecated': True}
,
5218 'has_video': {'type': 'alias', 'field': 'hasvid', 'deprecated': True}
,
5219 'audio': {'type': 'alias', 'field': 'hasaud', 'deprecated': True}
,
5220 'has_audio': {'type': 'alias', 'field': 'hasaud', 'deprecated': True}
,
5221 'extractor': {'type': 'alias', 'field': 'ie_pref', 'deprecated': True}
,
5222 'extractor_preference': {'type': 'alias', 'field': 'ie_pref', 'deprecated': True}
,
5225 def __init__(self
, ydl
, field_preference
):
5228 self
.evaluate_params(self
.ydl
.params
, field_preference
)
5229 if ydl
.params
.get('verbose'):
5230 self
.print_verbose_info(self
.ydl
.write_debug
)
5232 def _get_field_setting(self
, field
, key
):
5233 if field
not in self
.settings
:
5234 if key
in ('forced', 'priority'):
5236 self
.ydl
.deprecated_feature(f
'Using arbitrary fields ({field}) for format sorting is '
5237 'deprecated and may be removed in a future version')
5238 self
.settings
[field
] = {}
5239 prop_obj
= self
.settings
[field
]
5240 if key
not in prop_obj
:
5241 type_
= prop_obj
.get('type')
5243 default
= 'preference' if type_
== 'extractor' else (field
,) if type_
in ('combined', 'multiple') else field
5244 elif key
== 'convert':
5245 default
= 'order' if type_
== 'ordered' else 'float_string' if field
else 'ignore'
5247 default
= {'type': 'field', 'visible': True, 'order': [], 'not_in_list': (None,)}
.get(key
)
5248 prop_obj
[key
] = default
5249 return prop_obj
[key
]
5251 def _resolve_field_value(self
, field
, value
, convert_none
=False):
5253 if not convert_none
:
5256 value
= value
.lower()
5257 conversion
= self
._get
_field
_setting
(field
, 'convert')
5258 if conversion
== 'ignore':
5260 if conversion
== 'string':
5262 elif conversion
== 'float_none':
5263 return float_or_none(value
)
5264 elif conversion
== 'bytes':
5265 return parse_bytes(value
)
5266 elif conversion
== 'order':
5267 order_list
= (self
._use
_free
_order
and self
._get
_field
_setting
(field
, 'order_free')) or self
._get
_field
_setting
(field
, 'order')
5268 use_regex
= self
._get
_field
_setting
(field
, 'regex')
5269 list_length
= len(order_list
)
5270 empty_pos
= order_list
.index('') if '' in order_list
else list_length
+ 1
5271 if use_regex
and value
is not None:
5272 for i
, regex
in enumerate(order_list
):
5273 if regex
and re
.match(regex
, value
):
5274 return list_length
- i
5275 return list_length
- empty_pos
# not in list
5276 else: # not regex or value = None
5277 return list_length
- (order_list
.index(value
) if value
in order_list
else empty_pos
)
5279 if value
.isnumeric():
5282 self
.settings
[field
]['convert'] = 'string'
5285 def evaluate_params(self
, params
, sort_extractor
):
5286 self
._use
_free
_order
= params
.get('prefer_free_formats', False)
5287 self
._sort
_user
= params
.get('format_sort', [])
5288 self
._sort
_extractor
= sort_extractor
5290 def add_item(field
, reverse
, closest
, limit_text
):
5291 field
= field
.lower()
5292 if field
in self
._order
:
5294 self
._order
.append(field
)
5295 limit
= self
._resolve
_field
_value
(field
, limit_text
)
5298 'closest': False if limit
is None else closest
,
5299 'limit_text': limit_text
,
5301 if field
in self
.settings
:
5302 self
.settings
[field
].update(data
)
5304 self
.settings
[field
] = data
5307 tuple(field
for field
in self
.default
if self
._get
_field
_setting
(field
, 'forced'))
5308 + (tuple() if params
.get('format_sort_force', False)
5309 else tuple(field
for field
in self
.default
if self
._get
_field
_setting
(field
, 'priority')))
5310 + tuple(self
._sort
_user
) + tuple(sort_extractor
) + self
.default
)
5312 for item
in sort_list
:
5313 match
= re
.match(self
.regex
, item
)
5315 raise ExtractorError(f
'Invalid format sort string "{item}" given by extractor')
5316 field
= match
.group('field')
5319 if self
._get
_field
_setting
(field
, 'type') == 'alias':
5320 alias
, field
= field
, self
._get
_field
_setting
(field
, 'field')
5321 if self
._get
_field
_setting
(alias
, 'deprecated'):
5322 self
.ydl
.deprecated_feature(f
'Format sorting alias {alias} is deprecated and may '
5323 f
'be removed in a future version. Please use {field} instead')
5324 reverse
= match
.group('reverse') is not None
5325 closest
= match
.group('separator') == '~'
5326 limit_text
= match
.group('limit')
5328 has_limit
= limit_text
is not None
5329 has_multiple_fields
= self
._get
_field
_setting
(field
, 'type') == 'combined'
5330 has_multiple_limits
= has_limit
and has_multiple_fields
and not self
._get
_field
_setting
(field
, 'same_limit')
5332 fields
= self
._get
_field
_setting
(field
, 'field') if has_multiple_fields
else (field
,)
5333 limits
= limit_text
.split(':') if has_multiple_limits
else (limit_text
,) if has_limit
else tuple()
5334 limit_count
= len(limits
)
5335 for (i
, f
) in enumerate(fields
):
5336 add_item(f
, reverse
, closest
,
5337 limits
[i
] if i
< limit_count
5338 else limits
[0] if has_limit
and not has_multiple_limits
5341 def print_verbose_info(self
, write_debug
):
5343 write_debug('Sort order given by user: {}'.format(', '.join(self
._sort
_user
)))
5344 if self
._sort
_extractor
:
5345 write_debug('Sort order given by extractor: {}'.format(', '.join(self
._sort
_extractor
)))
5346 write_debug('Formats sorted by: {}'.format(', '.join(['{}{}{}'.format(
5347 '+' if self
._get
_field
_setting
(field
, 'reverse') else '', field
,
5348 '{}{}({})'.format('~' if self
._get
_field
_setting
(field
, 'closest') else ':',
5349 self
._get
_field
_setting
(field
, 'limit_text'),
5350 self
._get
_field
_setting
(field
, 'limit'))
5351 if self
._get
_field
_setting
(field
, 'limit_text') is not None else '')
5352 for field
in self
._order
if self
._get
_field
_setting
(field
, 'visible')])))
5354 def _calculate_field_preference_from_value(self
, format_
, field
, type_
, value
):
5355 reverse
= self
._get
_field
_setting
(field
, 'reverse')
5356 closest
= self
._get
_field
_setting
(field
, 'closest')
5357 limit
= self
._get
_field
_setting
(field
, 'limit')
5359 if type_
== 'extractor':
5360 maximum
= self
._get
_field
_setting
(field
, 'max')
5361 if value
is None or (maximum
is not None and value
>= maximum
):
5363 elif type_
== 'boolean':
5364 in_list
= self
._get
_field
_setting
(field
, 'in_list')
5365 not_in_list
= self
._get
_field
_setting
(field
, 'not_in_list')
5366 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
5367 elif type_
== 'ordered':
5368 value
= self
._resolve
_field
_value
(field
, value
, True)
5370 # try to convert to number
5371 val_num
= float_or_none(value
, default
=self
._get
_field
_setting
(field
, 'default'))
5372 is_num
= self
._get
_field
_setting
(field
, 'convert') != 'string' and val_num
is not None
5376 return ((-10, 0) if value
is None
5377 else (1, value
, 0) if not is_num
# if a field has mixed strings and numbers, strings are sorted higher
5378 else (0, -abs(value
- limit
), value
- limit
if reverse
else limit
- value
) if closest
5379 else (0, value
, 0) if not reverse
and (limit
is None or value
<= limit
)
5380 else (0, -value
, 0) if limit
is None or (reverse
and value
== limit
) or value
> limit
5381 else (-1, value
, 0))
5383 def _calculate_field_preference(self
, format_
, field
):
5384 type_
= self
._get
_field
_setting
(field
, 'type') # extractor, boolean, ordered, field, multiple
5385 get_value
= lambda f
: format_
.get(self
._get
_field
_setting
(f
, 'field'))
5386 if type_
== 'multiple':
5387 type_
= 'field' # Only 'field' is allowed in multiple for now
5388 actual_fields
= self
._get
_field
_setting
(field
, 'field')
5390 value
= self
._get
_field
_setting
(field
, 'function')(get_value(f
) for f
in actual_fields
)
5392 value
= get_value(field
)
5393 return self
._calculate
_field
_preference
_from
_value
(format_
, field
, type_
, value
)
5395 def calculate_preference(self
, format
):
5396 # Determine missing protocol
5397 if not format
.get('protocol'):
5398 format
['protocol'] = determine_protocol(format
)
5400 # Determine missing ext
5401 if not format
.get('ext') and 'url' in format
:
5402 format
['ext'] = determine_ext(format
['url'])
5403 if format
.get('vcodec') == 'none':
5404 format
['audio_ext'] = format
['ext'] if format
.get('acodec') != 'none' else 'none'
5405 format
['video_ext'] = 'none'
5407 format
['video_ext'] = format
['ext']
5408 format
['audio_ext'] = 'none'
5409 # if format.get('preference') is None and format.get('ext') in ('f4f', 'f4m'): # Not supported?
5410 # format['preference'] = -1000
5412 if format
.get('preference') is None and format
.get('ext') == 'flv' and re
.match('[hx]265|he?vc?', format
.get('vcodec') or ''):
5413 # HEVC-over-FLV is out-of-spec by FLV's original spec
5414 # ref. https://trac.ffmpeg.org/ticket/6389
5415 # ref. https://github.com/yt-dlp/yt-dlp/pull/5821
5416 format
['preference'] = -100
5418 # Determine missing bitrates
5419 if format
.get('vcodec') == 'none':
5421 if format
.get('acodec') == 'none':
5423 if not format
.get('vbr') and format
.get('vcodec') != 'none':
5424 format
['vbr'] = try_call(lambda: format
['tbr'] - format
['abr']) or None
5425 if not format
.get('abr') and format
.get('acodec') != 'none':
5426 format
['abr'] = try_call(lambda: format
['tbr'] - format
['vbr']) or None
5427 if not format
.get('tbr'):
5428 format
['tbr'] = try_call(lambda: format
['vbr'] + format
['abr']) or None
5430 return tuple(self
._calculate
_field
_preference
(format
, field
) for field
in self
._order
)
5433 def filesize_from_tbr(tbr
, duration
):
5435 @param tbr: Total bitrate in kbps (1000 bits/sec)
5436 @param duration: Duration in seconds
5437 @returns Filesize in bytes
5439 if tbr
is None or duration
is None:
5441 return int(duration
* tbr
* (1000 / 8))
5446 def __init__(self
, ydl
=None):
5449 def debug(self
, message
):
5451 self
._ydl
.write_debug(message
)
5453 def info(self
, message
):
5455 self
._ydl
.to_screen(message
)
5457 def warning(self
, message
, *, once
=False):
5459 self
._ydl
.report_warning(message
, once
)
5461 def error(self
, message
, *, is_error
=True):
5463 self
._ydl
.report_error(message
, is_error
=is_error
)
5465 def stdout(self
, message
):
5467 self
._ydl
.to_stdout(message
)
5469 def stderr(self
, message
):
5471 self
._ydl
.to_stderr(message
)