45 import xml
.etree
.ElementTree
47 from . import traversal
49 from ..compat
import functools
# isort: split
50 from ..compat
import (
51 compat_etree_fromstring
,
53 compat_HTMLParseError
,
57 from ..dependencies
import websockets
, xattr
59 __name__
= __name__
.rsplit('.', 1)[0] # Pretend to be the parent module
61 # This is not clearly defined otherwise
62 compiled_regex_type
= type(re
.compile(''))
73 ENGLISH_MONTH_NAMES
= [
74 'January', 'February', 'March', 'April', 'May', 'June',
75 'July', 'August', 'September', 'October', 'November', 'December']
78 'en': ENGLISH_MONTH_NAMES
,
80 'janvier', 'février', 'mars', 'avril', 'mai', 'juin',
81 'juillet', 'août', 'septembre', 'octobre', 'novembre', 'décembre'],
82 # these follow the genitive grammatical case (dopełniacz)
83 # some websites might be using nominative, which will require another month list
84 # https://en.wikibooks.org/wiki/Polish/Noun_cases
85 'pl': ['stycznia', 'lutego', 'marca', 'kwietnia', 'maja', 'czerwca',
86 'lipca', 'sierpnia', 'września', 'października', 'listopada', 'grudnia'],
89 # From https://github.com/python/cpython/blob/3.11/Lib/email/_parseaddr.py#L36-L42
91 'UT': 0, 'UTC': 0, 'GMT': 0, 'Z': 0,
92 'AST': -4, 'ADT': -3, # Atlantic (used in Canada)
93 'EST': -5, 'EDT': -4, # Eastern
94 'CST': -6, 'CDT': -5, # Central
95 'MST': -7, 'MDT': -6, # Mountain
96 'PST': -8, 'PDT': -7 # Pacific
99 # needed for sanitizing filenames in restricted mode
100 ACCENT_CHARS
= dict(zip('ÂÃÄÀÁÅÆÇÈÉÊËÌÍÎÏÐÑÒÓÔÕÖŐØŒÙÚÛÜŰÝÞßàáâãäåæçèéêëìíîïðñòóôõöőøœùúûüűýþÿ',
101 itertools
.chain('AAAAAA', ['AE'], 'CEEEEIIIIDNOOOOOOO', ['OE'], 'UUUUUY', ['TH', 'ss'],
102 'aaaaaa', ['ae'], 'ceeeeiiiionooooooo', ['oe'], 'uuuuuy', ['th'], 'y')))
132 '%Y-%m-%d %H:%M:%S.%f',
133 '%Y-%m-%d %H:%M:%S:%f',
136 '%Y-%m-%dT%H:%M:%SZ',
137 '%Y-%m-%dT%H:%M:%S.%fZ',
138 '%Y-%m-%dT%H:%M:%S.%f0Z',
140 '%Y-%m-%dT%H:%M:%S.%f',
143 '%b %d %Y at %H:%M:%S',
145 '%B %d %Y at %H:%M:%S',
149 DATE_FORMATS_DAY_FIRST
= list(DATE_FORMATS
)
150 DATE_FORMATS_DAY_FIRST
.extend([
161 DATE_FORMATS_MONTH_FIRST
= list(DATE_FORMATS
)
162 DATE_FORMATS_MONTH_FIRST
.extend([
170 PACKED_CODES_RE
= r
"}\('(.+)',(\d+),(\d+),'([^']+)'\.split\('\|'\)"
171 JSON_LD_RE
= r
'(?is)<script[^>]+type=(["\']?
)application
/ld\
+json\
1[^
>]*>\s
*(?P
<json_ld
>{.+?}|\
[.+?\
])\s
*</script
>'
173 NUMBER_RE = r'\d
+(?
:\
.\d
+)?
'
177 def preferredencoding():
178 """Get preferred encoding.
180 Returns the best encoding scheme for the system, based on
181 locale.getpreferredencoding() and some further tweaks.
184 pref = locale.getpreferredencoding()
192 def write_json_file(obj, fn):
193 """ Encode obj as JSON and write it to fn, atomically if possible """
195 tf = tempfile.NamedTemporaryFile(
196 prefix=f'{os.path.basename(fn)}
.', dir=os.path.dirname(fn),
197 suffix='.tmp
', delete=False, mode='w
', encoding='utf
-8')
201 json.dump(obj, tf, ensure_ascii=False)
202 if sys.platform == 'win32
':
203 # Need to remove existing file on Windows, else os.rename raises
204 # WindowsError or FileExistsError.
205 with contextlib.suppress(OSError):
207 with contextlib.suppress(OSError):
210 os.chmod(tf.name, 0o666 & ~mask)
211 os.rename(tf.name, fn)
213 with contextlib.suppress(OSError):
218 def find_xpath_attr(node, xpath, key, val=None):
219 """ Find the xpath xpath[@key=val] """
220 assert re.match(r'^
[a
-zA
-Z_
-]+$
', key)
221 expr = xpath + ('[@%s]' % key if val is None else f"[@{key}='{val}
']")
222 return node.find(expr)
224 # On python2.6 the xml.etree.ElementTree.Element methods don't support
225 # the namespace parameter
228 def xpath_with_ns(path
, ns_map
):
229 components
= [c
.split(':') for c
in path
.split('/')]
233 replaced
.append(c
[0])
236 replaced
.append('{%s}%s' % (ns_map
[ns
], tag
))
237 return '/'.join(replaced
)
240 def xpath_element(node
, xpath
, name
=None, fatal
=False, default
=NO_DEFAULT
):
241 def _find_xpath(xpath
):
242 return node
.find(xpath
)
244 if isinstance(xpath
, str):
245 n
= _find_xpath(xpath
)
253 if default
is not NO_DEFAULT
:
256 name
= xpath
if name
is None else name
257 raise ExtractorError('Could not find XML element %s' % name
)
263 def xpath_text(node
, xpath
, name
=None, fatal
=False, default
=NO_DEFAULT
):
264 n
= xpath_element(node
, xpath
, name
, fatal
=fatal
, default
=default
)
265 if n
is None or n
== default
:
268 if default
is not NO_DEFAULT
:
271 name
= xpath
if name
is None else name
272 raise ExtractorError('Could not find XML element\'s text %s' % name
)
278 def xpath_attr(node
, xpath
, key
, name
=None, fatal
=False, default
=NO_DEFAULT
):
279 n
= find_xpath_attr(node
, xpath
, key
)
281 if default
is not NO_DEFAULT
:
284 name
= f
'{xpath}[@{key}]' if name
is None else name
285 raise ExtractorError('Could not find XML attribute %s' % name
)
291 def get_element_by_id(id, html
, **kwargs
):
292 """Return the content of the tag with the specified ID in the passed HTML document"""
293 return get_element_by_attribute('id', id, html
, **kwargs
)
296 def get_element_html_by_id(id, html
, **kwargs
):
297 """Return the html of the tag with the specified ID in the passed HTML document"""
298 return get_element_html_by_attribute('id', id, html
, **kwargs
)
301 def get_element_by_class(class_name
, html
):
302 """Return the content of the first tag with the specified class in the passed HTML document"""
303 retval
= get_elements_by_class(class_name
, html
)
304 return retval
[0] if retval
else None
307 def get_element_html_by_class(class_name
, html
):
308 """Return the html of the first tag with the specified class in the passed HTML document"""
309 retval
= get_elements_html_by_class(class_name
, html
)
310 return retval
[0] if retval
else None
313 def get_element_by_attribute(attribute
, value
, html
, **kwargs
):
314 retval
= get_elements_by_attribute(attribute
, value
, html
, **kwargs
)
315 return retval
[0] if retval
else None
318 def get_element_html_by_attribute(attribute
, value
, html
, **kargs
):
319 retval
= get_elements_html_by_attribute(attribute
, value
, html
, **kargs
)
320 return retval
[0] if retval
else None
323 def get_elements_by_class(class_name
, html
, **kargs
):
324 """Return the content of all tags with the specified class in the passed HTML document as a list"""
325 return get_elements_by_attribute(
326 'class', r
'[^\'"]*(?<=[\'"\s
])%s(?
=[\'"\s])[^\'"]*' % re.escape(class_name),
327 html, escape_value=False)
330 def get_elements_html_by_class(class_name, html):
331 """Return the html of all tags with the specified class in the passed HTML document as a list"""
332 return get_elements_html_by_attribute(
333 'class', r'[^
\'"]*(?<=[\'"\s
])%s(?
=[\'"\s])[^\'"]*' % re.escape(class_name),
334 html, escape_value=False)
337 def get_elements_by_attribute(*args, **kwargs):
338 """Return the content of the tag with the specified attribute in the passed HTML document"""
339 return [content for content, _ in get_elements_text_and_html_by_attribute(*args, **kwargs)]
342 def get_elements_html_by_attribute(*args, **kwargs):
343 """Return the html of the tag with the specified attribute in the passed HTML document"""
344 return [whole for _, whole in get_elements_text_and_html_by_attribute(*args, **kwargs)]
347 def get_elements_text_and_html_by_attribute(attribute, value, html, *, tag=r'[\w
:.-]+', escape_value=True):
349 Return the text (content) and the html (whole) of the tag with the specified
350 attribute in the passed HTML document
355 quote = '' if re.match(r'''[\s"'`
=<>]''', value) else '?'
357 value = re.escape(value) if escape_value else value
359 partial_element_re = rf'''(?x
)
361 (?
:\
s(?
:[^
>"']|"[^
"]*"|
'[^']*')*)?
362 \s{re.escape(attribute)}\s*=\s*(?P<_q>['"]{quote})(?-x:{value})(?P=_q)
365 for m in re.finditer(partial_element_re, html):
366 content, whole = get_element_text_and_html_by_tag(m.group('tag'), html[m.start():])
369 unescapeHTML(re.sub(r'^(?P<q>["\'])(?P
<content
>.*)(?P
=q
)$
', r'\g
<content
>', content, flags=re.DOTALL)),
374 class HTMLBreakOnClosingTagParser(html.parser.HTMLParser):
376 HTML parser which raises HTMLBreakOnClosingTagException upon reaching the
377 closing tag for the first opening tag it has encountered, and can be used
381 class HTMLBreakOnClosingTagException(Exception):
385 self.tagstack = collections.deque()
386 html.parser.HTMLParser.__init__(self)
391 def __exit__(self, *_):
395 # handle_endtag does not return upon raising HTMLBreakOnClosingTagException,
396 # so data remains buffered; we no longer have any interest in it, thus
397 # override this method to discard it
400 def handle_starttag(self, tag, _):
401 self.tagstack.append(tag)
403 def handle_endtag(self, tag):
404 if not self.tagstack:
405 raise compat_HTMLParseError('no tags
in the stack
')
407 inner_tag = self.tagstack.pop()
411 raise compat_HTMLParseError(f'matching opening tag
for closing {tag} tag
not found
')
412 if not self.tagstack:
413 raise self.HTMLBreakOnClosingTagException()
416 # XXX: This should be far less strict
417 def get_element_text_and_html_by_tag(tag, html):
419 For the first element with the specified tag in the passed HTML document
420 return its' content (text
) and the whole
element (html
)
422 def find_or_raise(haystack, needle, exc):
424 return haystack.index(needle)
427 closing_tag = f'</{tag}>'
428 whole_start = find_or_raise(
429 html, f'<{tag}', compat_HTMLParseError(f'opening {tag} tag not found'))
430 content_start = find_or_raise(
431 html[whole_start:], '>', compat_HTMLParseError(f'malformed opening {tag} tag'))
432 content_start += whole_start + 1
433 with HTMLBreakOnClosingTagParser() as parser:
434 parser.feed(html[whole_start:content_start])
435 if not parser.tagstack or parser.tagstack[0] != tag:
436 raise compat_HTMLParseError(f'parser did not match opening {tag} tag')
437 offset = content_start
438 while offset < len(html):
439 next_closing_tag_start = find_or_raise(
440 html[offset:], closing_tag,
441 compat_HTMLParseError(f'closing {tag} tag not found'))
442 next_closing_tag_end = next_closing_tag_start + len(closing_tag)
444 parser.feed(html[offset:offset + next_closing_tag_end])
445 offset += next_closing_tag_end
446 except HTMLBreakOnClosingTagParser.HTMLBreakOnClosingTagException:
447 return html[content_start:offset + next_closing_tag_start], \
448 html[whole_start:offset + next_closing_tag_end]
449 raise compat_HTMLParseError('unexpected end of html')
452 class HTMLAttributeParser(html.parser.HTMLParser):
453 """Trivial HTML parser to gather the attributes
for a single element
"""
457 html.parser.HTMLParser.__init__(self)
459 def handle_starttag(self, tag, attrs):
460 self.attrs = dict(attrs)
461 raise compat_HTMLParseError('done')
464 class HTMLListAttrsParser(html.parser.HTMLParser):
465 """HTML parser to gather the attributes
for the elements of a
list"""
468 html.parser.HTMLParser.__init__(self)
472 def handle_starttag(self, tag, attrs):
473 if tag == 'li' and self._level == 0:
474 self.items.append(dict(attrs))
477 def handle_endtag(self, tag):
481 def extract_attributes(html_element):
482 """Given a string
for an HTML element such
as
484 a
="foo" B
="bar" c
="&98;az" d
=boz
485 empty
= noval entity
="&"
488 Decode
and return a dictionary of attributes
.
490 'a': 'foo', 'b': 'bar', c
: 'baz', d
: 'boz',
491 'empty': '', 'noval': None, 'entity': '&',
492 'sq': '"', 'dq': '\''
495 parser = HTMLAttributeParser()
496 with contextlib.suppress(compat_HTMLParseError):
497 parser.feed(html_element)
502 def parse_list(webpage):
503 """Given a string
for an series of HTML
<li
> elements
,
504 return a dictionary of their attributes
"""
505 parser = HTMLListAttrsParser()
511 def clean_html(html):
512 """Clean an HTML snippet into a readable string
"""
514 if html is None: # Convenience for sanitizing descriptions etc.
517 html = re.sub(r'\s+', ' ', html)
518 html = re.sub(r'(?u)\s?<\s?br\s?/?\s?>\s?', '\n', html)
519 html = re.sub(r'(?u)<\s?/\s?p\s?>\s?<\s?p[^>]*>', '\n', html)
521 html = re.sub('<.*?>', '', html)
522 # Replace html entities
523 html = unescapeHTML(html)
527 class LenientJSONDecoder(json.JSONDecoder):
529 def __init__(self, *args, transform_source=None, ignore_extra=False, close_objects=0, **kwargs):
530 self.transform_source, self.ignore_extra = transform_source, ignore_extra
531 self._close_attempts = 2 * close_objects
532 super().__init__(*args, **kwargs)
535 def _close_object(err):
536 doc = err.doc[:err.pos]
537 # We need to add comma first to get the correct error message
538 if err.msg.startswith('Expecting \',\''):
540 elif not doc.endswith(','):
543 if err.msg.startswith('Expecting property name'):
544 return doc[:-1] + '}'
545 elif err.msg.startswith('Expecting value'):
546 return doc[:-1] + ']'
549 if self.transform_source:
550 s = self.transform_source(s)
551 for attempt in range(self._close_attempts + 1):
553 if self.ignore_extra:
554 return self.raw_decode(s.lstrip())[0]
555 return super().decode(s)
556 except json.JSONDecodeError as e:
559 elif attempt < self._close_attempts:
560 s = self._close_object(e)
563 raise type(e)(f'{e.msg} in {s[e.pos-10:e.pos+10]!r}', s, e.pos)
564 assert False, 'Too many attempts to decode JSON'
567 def sanitize_open(filename, open_mode):
568 """Try to
open the given filename
, and slightly tweak it
if this fails
.
570 Attempts to
open the given filename
. If this fails
, it tries to change
571 the filename slightly
, step by step
, until it
's either able to open it
572 or it fails and raises a final exception, like the standard open()
575 It returns the tuple (stream, definitive_file_name).
578 if sys.platform == 'win32
':
581 # stdout may be any IO stream, e.g. when using contextlib.redirect_stdout
582 with contextlib.suppress(io.UnsupportedOperation):
583 msvcrt.setmode(sys.stdout.fileno(), os.O_BINARY)
584 return (sys.stdout.buffer if hasattr(sys.stdout, 'buffer') else sys.stdout, filename)
586 for attempt in range(2):
589 if sys.platform == 'win32
':
590 # FIXME: An exclusive lock also locks the file from being read.
591 # Since windows locks are mandatory, don't lock the
file on
windows (for now
).
592 # Ref: https://github.com/yt-dlp/yt-dlp/issues/3124
593 raise LockingUnsupportedError()
594 stream
= locked_file(filename
, open_mode
, block
=False).__enter
__()
596 stream
= open(filename
, open_mode
)
597 return stream
, filename
598 except OSError as err
:
599 if attempt
or err
.errno
in (errno
.EACCES
,):
601 old_filename
, filename
= filename
, sanitize_path(filename
)
602 if old_filename
== filename
:
606 def timeconvert(timestr
):
607 """Convert RFC 2822 defined time string into system timestamp"""
609 timetuple
= email
.utils
.parsedate_tz(timestr
)
610 if timetuple
is not None:
611 timestamp
= email
.utils
.mktime_tz(timetuple
)
615 def sanitize_filename(s
, restricted
=False, is_id
=NO_DEFAULT
):
616 """Sanitizes a string so it could be used as part of a filename.
617 @param restricted Use a stricter subset of allowed characters
618 @param is_id Whether this is an ID that should be kept unchanged if possible.
619 If unset, yt-dlp's new sanitization rules are in effect
624 def replace_insane(char
):
625 if restricted
and char
in ACCENT_CHARS
:
626 return ACCENT_CHARS
[char
]
627 elif not restricted
and char
== '\n':
629 elif is_id
is NO_DEFAULT
and not restricted
and char
in '"*:<>?|/\\':
630 # Replace with their full-width unicode counterparts
631 return {'/': '\u29F8', '\\': '\u29f9'}
.get(char
, chr(ord(char
) + 0xfee0))
632 elif char
== '?' or ord(char
) < 32 or ord(char
) == 127:
635 return '' if restricted
else '\''
637 return '\0_\0-' if restricted
else '\0 \0-'
638 elif char
in '\\/|*<>':
640 if restricted
and (char
in '!&\'()[]{}$;`^,#' or char
.isspace() or ord(char
) > 127):
644 # Replace look-alike Unicode glyphs
645 if restricted
and (is_id
is NO_DEFAULT
or not is_id
):
646 s
= unicodedata
.normalize('NFKC', s
)
647 s
= re
.sub(r
'[0-9]+(?::[0-9]+)+', lambda m
: m
.group(0).replace(':', '_'), s
) # Handle timestamps
648 result
= ''.join(map(replace_insane
, s
))
649 if is_id
is NO_DEFAULT
:
650 result
= re
.sub(r
'(\0.)(?:(?=\1)..)+', r
'\1', result
) # Remove repeated substitute chars
651 STRIP_RE
= r
'(?:\0.|[ _-])*'
652 result
= re
.sub(f
'^\0.{STRIP_RE}|{STRIP_RE}\0.$', '', result
) # Remove substitute chars from start/end
653 result
= result
.replace('\0', '') or '_'
656 while '__' in result
:
657 result
= result
.replace('__', '_')
658 result
= result
.strip('_')
659 # Common case of "Foreign band name - English song title"
660 if restricted
and result
.startswith('-_'):
662 if result
.startswith('-'):
663 result
= '_' + result
[len('-'):]
664 result
= result
.lstrip('.')
670 def sanitize_path(s
, force
=False):
671 """Sanitizes and normalizes path on Windows"""
672 # XXX: this handles drive relative paths (c:sth) incorrectly
673 if sys
.platform
== 'win32':
675 drive_or_unc
, _
= os
.path
.splitdrive(s
)
681 norm_path
= os
.path
.normpath(remove_start(s
, drive_or_unc
)).split(os
.path
.sep
)
685 path_part
if path_part
in ['.', '..'] else re
.sub(r
'(?:[/<>:"\|\\?\*]|[\s.]$)', '#', path_part
)
686 for path_part
in norm_path
]
688 sanitized_path
.insert(0, drive_or_unc
+ os
.path
.sep
)
689 elif force
and s
and s
[0] == os
.path
.sep
:
690 sanitized_path
.insert(0, os
.path
.sep
)
691 # TODO: Fix behavioral differences <3.12
692 # The workaround using `normpath` only superficially passes tests
693 # Ref: https://github.com/python/cpython/pull/100351
694 return os
.path
.normpath(os
.path
.join(*sanitized_path
))
697 def sanitize_url(url
, *, scheme
='http'):
698 # Prepend protocol-less URLs with `http:` scheme in order to mitigate
699 # the number of unwanted failures due to missing protocol
702 elif url
.startswith('//'):
703 return f
'{scheme}:{url}'
704 # Fix some common typos seen so far
706 # https://github.com/ytdl-org/youtube-dl/issues/15649
707 (r
'^httpss://', r
'https://'),
708 # https://bx1.be/lives/direct-tv/
709 (r
'^rmtp([es]?)://', r
'rtmp\1://'),
711 for mistake
, fixup
in COMMON_TYPOS
:
712 if re
.match(mistake
, url
):
713 return re
.sub(mistake
, fixup
, url
)
717 def extract_basic_auth(url
):
718 parts
= urllib
.parse
.urlsplit(url
)
719 if parts
.username
is None:
721 url
= urllib
.parse
.urlunsplit(parts
._replace
(netloc
=(
722 parts
.hostname
if parts
.port
is None
723 else '%s:%d' % (parts
.hostname
, parts
.port
))))
724 auth_payload
= base64
.b64encode(
725 ('%s:%s' % (parts
.username
, parts
.password
or '')).encode())
726 return url
, f
'Basic {auth_payload.decode()}'
730 """Expand shell variables and ~"""
731 return os
.path
.expandvars(compat_expanduser(s
))
734 def orderedSet(iterable
, *, lazy
=False):
735 """Remove all duplicates from the input iterable"""
737 seen
= [] # Do not use set since the items can be unhashable
743 return _iter() if lazy
else list(_iter())
746 def _htmlentity_transform(entity_with_semicolon
):
747 """Transforms an HTML entity to a character."""
748 entity
= entity_with_semicolon
[:-1]
750 # Known non-numeric HTML entity
751 if entity
in html
.entities
.name2codepoint
:
752 return chr(html
.entities
.name2codepoint
[entity
])
754 # TODO: HTML5 allows entities without a semicolon.
755 # E.g. 'Éric' should be decoded as 'Éric'.
756 if entity_with_semicolon
in html
.entities
.html5
:
757 return html
.entities
.html5
[entity_with_semicolon
]
759 mobj
= re
.match(r
'#(x[0-9a-fA-F]+|[0-9]+)', entity
)
761 numstr
= mobj
.group(1)
762 if numstr
.startswith('x'):
764 numstr
= '0%s' % numstr
767 # See https://github.com/ytdl-org/youtube-dl/issues/7518
768 with contextlib
.suppress(ValueError):
769 return chr(int(numstr
, base
))
771 # Unknown entity in name, return its literal representation
772 return '&%s;' % entity
778 assert isinstance(s
, str)
781 r
'&([^&;]+;)', lambda m
: _htmlentity_transform(m
.group(1)), s
)
784 def escapeHTML(text
):
787 .replace('&', '&')
788 .replace('<', '<')
789 .replace('>', '>')
790 .replace('"', '"')
791 .replace("'", ''')
795 class netrc_from_content(netrc
.netrc
):
796 def __init__(self
, content
):
797 self
.hosts
, self
.macros
= {}, {}
798 with io
.StringIO(content
) as stream
:
799 self
._parse
('-', stream
, False)
802 class Popen(subprocess
.Popen
):
803 if sys
.platform
== 'win32':
804 _startupinfo
= subprocess
.STARTUPINFO()
805 _startupinfo
.dwFlags |
= subprocess
.STARTF_USESHOWWINDOW
810 def _fix_pyinstaller_ld_path(env
):
811 """Restore LD_LIBRARY_PATH when using PyInstaller
812 Ref: https://github.com/pyinstaller/pyinstaller/blob/develop/doc/runtime-information.rst#ld_library_path--libpath-considerations
813 https://github.com/yt-dlp/yt-dlp/issues/4573
815 if not hasattr(sys
, '_MEIPASS'):
819 orig
= env
.get(f
'{key}_ORIG')
825 _fix('LD_LIBRARY_PATH') # Linux
826 _fix('DYLD_LIBRARY_PATH') # macOS
828 def __init__(self
, *args
, env
=None, text
=False, **kwargs
):
830 env
= os
.environ
.copy()
831 self
._fix
_pyinstaller
_ld
_path
(env
)
833 self
.__text
_mode
= kwargs
.get('encoding') or kwargs
.get('errors') or text
or kwargs
.get('universal_newlines')
835 kwargs
['universal_newlines'] = True # For 3.6 compatibility
836 kwargs
.setdefault('encoding', 'utf-8')
837 kwargs
.setdefault('errors', 'replace')
838 super().__init
__(*args
, env
=env
, **kwargs
, startupinfo
=self
._startupinfo
)
840 def communicate_or_kill(self
, *args
, **kwargs
):
842 return self
.communicate(*args
, **kwargs
)
843 except BaseException
: # Including KeyboardInterrupt
844 self
.kill(timeout
=None)
847 def kill(self
, *, timeout
=0):
850 self
.wait(timeout
=timeout
)
853 def run(cls
, *args
, timeout
=None, **kwargs
):
854 with cls(*args
, **kwargs
) as proc
:
855 default
= '' if proc
.__text
_mode
else b
''
856 stdout
, stderr
= proc
.communicate_or_kill(timeout
=timeout
)
857 return stdout
or default
, stderr
or default
, proc
.returncode
860 def encodeArgument(s
):
861 # Legacy code that uses byte strings
862 # Uncomment the following line after fixing all post processors
863 # assert isinstance(s, str), 'Internal error: %r should be of type %r, is %r' % (s, str, type(s))
864 return s
if isinstance(s
, str) else s
.decode('ascii')
867 _timetuple
= collections
.namedtuple('Time', ('hours', 'minutes', 'seconds', 'milliseconds'))
870 def timetuple_from_msec(msec
):
871 secs
, msec
= divmod(msec
, 1000)
872 mins
, secs
= divmod(secs
, 60)
873 hrs
, mins
= divmod(mins
, 60)
874 return _timetuple(hrs
, mins
, secs
, msec
)
877 def formatSeconds(secs
, delim
=':', msec
=False):
878 time
= timetuple_from_msec(secs
* 1000)
880 ret
= '%d%s%02d%s%02d' % (time
.hours
, delim
, time
.minutes
, delim
, time
.seconds
)
882 ret
= '%d%s%02d' % (time
.minutes
, delim
, time
.seconds
)
884 ret
= '%d' % time
.seconds
885 return '%s.%03d' % (ret
, time
.milliseconds
) if msec
else ret
888 def bug_reports_message(before
=';'):
889 from ..update
import REPOSITORY
891 msg
= (f
'please report this issue on https://github.com/{REPOSITORY}/issues?q= , '
892 'filling out the appropriate issue template. Confirm you are on the latest version using yt-dlp -U')
894 before
= before
.rstrip()
895 if not before
or before
.endswith(('.', '!', '?')):
896 msg
= msg
[0].title() + msg
[1:]
898 return (before
+ ' ' if before
else '') + msg
901 class YoutubeDLError(Exception):
902 """Base exception for YoutubeDL errors."""
905 def __init__(self
, msg
=None):
908 elif self
.msg
is None:
909 self
.msg
= type(self
).__name
__
910 super().__init
__(self
.msg
)
913 class ExtractorError(YoutubeDLError
):
914 """Error during info extraction."""
916 def __init__(self
, msg
, tb
=None, expected
=False, cause
=None, video_id
=None, ie
=None):
917 """ tb, if given, is the original traceback (so that it can be printed out).
918 If expected is set, this is a normal error message and most likely not a bug in yt-dlp.
920 from ..networking
.exceptions
import network_exceptions
921 if sys
.exc_info()[0] in network_exceptions
:
924 self
.orig_msg
= str(msg
)
926 self
.expected
= expected
928 self
.video_id
= video_id
930 self
.exc_info
= sys
.exc_info() # preserve original exception
931 if isinstance(self
.exc_info
[1], ExtractorError
):
932 self
.exc_info
= self
.exc_info
[1].exc_info
933 super().__init
__(self
.__msg
)
938 format_field(self
.ie
, None, '[%s] '),
939 format_field(self
.video_id
, None, '%s: '),
941 format_field(self
.cause
, None, ' (caused by %r)'),
942 '' if self
.expected
else bug_reports_message()))
944 def format_traceback(self
):
945 return join_nonempty(
946 self
.traceback
and ''.join(traceback
.format_tb(self
.traceback
)),
947 self
.cause
and ''.join(traceback
.format_exception(None, self
.cause
, self
.cause
.__traceback
__)[1:]),
950 def __setattr__(self
, name
, value
):
951 super().__setattr
__(name
, value
)
952 if getattr(self
, 'msg', None) and name
not in ('msg', 'args'):
953 self
.msg
= self
.__msg
or type(self
).__name
__
954 self
.args
= (self
.msg
, ) # Cannot be property
957 class UnsupportedError(ExtractorError
):
958 def __init__(self
, url
):
960 'Unsupported URL: %s' % url
, expected
=True)
964 class RegexNotFoundError(ExtractorError
):
965 """Error when a regex didn't match"""
969 class GeoRestrictedError(ExtractorError
):
970 """Geographic restriction Error exception.
972 This exception may be thrown when a video is not available from your
973 geographic location due to geographic restrictions imposed by a website.
976 def __init__(self
, msg
, countries
=None, **kwargs
):
977 kwargs
['expected'] = True
978 super().__init
__(msg
, **kwargs
)
979 self
.countries
= countries
982 class UserNotLive(ExtractorError
):
983 """Error when a channel/user is not live"""
985 def __init__(self
, msg
=None, **kwargs
):
986 kwargs
['expected'] = True
987 super().__init
__(msg
or 'The channel is not currently live', **kwargs
)
990 class DownloadError(YoutubeDLError
):
991 """Download Error exception.
993 This exception may be thrown by FileDownloader objects if they are not
994 configured to continue on errors. They will contain the appropriate
998 def __init__(self
, msg
, exc_info
=None):
999 """ exc_info, if given, is the original exception that caused the trouble (as returned by sys.exc_info()). """
1000 super().__init
__(msg
)
1001 self
.exc_info
= exc_info
1004 class EntryNotInPlaylist(YoutubeDLError
):
1005 """Entry not in playlist exception.
1007 This exception will be thrown by YoutubeDL when a requested entry
1008 is not found in the playlist info_dict
1010 msg
= 'Entry not found in info'
1013 class SameFileError(YoutubeDLError
):
1014 """Same File exception.
1016 This exception will be thrown by FileDownloader objects if they detect
1017 multiple files would have to be downloaded to the same file on disk.
1019 msg
= 'Fixed output name but more than one file to download'
1021 def __init__(self
, filename
=None):
1022 if filename
is not None:
1023 self
.msg
+= f
': {filename}'
1024 super().__init
__(self
.msg
)
1027 class PostProcessingError(YoutubeDLError
):
1028 """Post Processing exception.
1030 This exception may be raised by PostProcessor's .run() method to
1031 indicate an error in the postprocessing task.
1035 class DownloadCancelled(YoutubeDLError
):
1036 """ Exception raised when the download queue should be interrupted """
1037 msg
= 'The download was cancelled'
1040 class ExistingVideoReached(DownloadCancelled
):
1041 """ --break-on-existing triggered """
1042 msg
= 'Encountered a video that is already in the archive, stopping due to --break-on-existing'
1045 class RejectedVideoReached(DownloadCancelled
):
1046 """ --break-match-filter triggered """
1047 msg
= 'Encountered a video that did not match filter, stopping due to --break-match-filter'
1050 class MaxDownloadsReached(DownloadCancelled
):
1051 """ --max-downloads limit has been reached. """
1052 msg
= 'Maximum number of downloads reached, stopping due to --max-downloads'
1055 class ReExtractInfo(YoutubeDLError
):
1056 """ Video info needs to be re-extracted. """
1058 def __init__(self
, msg
, expected
=False):
1059 super().__init
__(msg
)
1060 self
.expected
= expected
1063 class ThrottledDownload(ReExtractInfo
):
1064 """ Download speed below --throttled-rate. """
1065 msg
= 'The download speed is below throttle limit'
1068 super().__init
__(self
.msg
, expected
=False)
1071 class UnavailableVideoError(YoutubeDLError
):
1072 """Unavailable Format exception.
1074 This exception will be thrown when a video is requested
1075 in a format that is not available for that video.
1077 msg
= 'Unable to download video'
1079 def __init__(self
, err
=None):
1081 self
.msg
+= f
': {err}'
1082 super().__init
__(self
.msg
)
1085 class ContentTooShortError(YoutubeDLError
):
1086 """Content Too Short exception.
1088 This exception may be raised by FileDownloader objects when a file they
1089 download is too small for what the server announced first, indicating
1090 the connection was probably interrupted.
1093 def __init__(self
, downloaded
, expected
):
1094 super().__init
__(f
'Downloaded {downloaded} bytes, expected {expected} bytes')
1096 self
.downloaded
= downloaded
1097 self
.expected
= expected
1100 class XAttrMetadataError(YoutubeDLError
):
1101 def __init__(self
, code
=None, msg
='Unknown error'):
1102 super().__init
__(msg
)
1106 # Parsing code and msg
1107 if (self
.code
in (errno
.ENOSPC
, errno
.EDQUOT
)
1108 or 'No space left' in self
.msg
or 'Disk quota exceeded' in self
.msg
):
1109 self
.reason
= 'NO_SPACE'
1110 elif self
.code
== errno
.E2BIG
or 'Argument list too long' in self
.msg
:
1111 self
.reason
= 'VALUE_TOO_LONG'
1113 self
.reason
= 'NOT_SUPPORTED'
1116 class XAttrUnavailableError(YoutubeDLError
):
1120 def is_path_like(f
):
1121 return isinstance(f
, (str, bytes, os
.PathLike
))
1124 def extract_timezone(date_str
):
1127 ^.{8,}? # >=8 char non-TZ prefix, if present
1128 (?P<tz>Z| # just the UTC Z, or
1129 (?:(?<=.\b\d{4}|\b\d{2}:\d\d)| # preceded by 4 digits or hh:mm or
1130 (?<!.\b[a-zA-Z]{3}|[a-zA-Z]{4}|..\b\d\d)) # not preceded by 3 alpha word or >= 4 alpha or 2 digits
1131 [ ]? # optional space
1132 (?P<sign>\+|-) # +/-
1133 (?P<hours>[0-9]{2}):?(?P<minutes>[0-9]{2}) # hh[:]mm
1137 m
= re
.search(r
'\d{1,2}:\d{1,2}(?:\.\d+)?(?P<tz>\s*[A-Z]+)$', date_str
)
1138 timezone
= TIMEZONE_NAMES
.get(m
and m
.group('tz').strip())
1139 if timezone
is not None:
1140 date_str
= date_str
[:-len(m
.group('tz'))]
1141 timezone
= datetime
.timedelta(hours
=timezone
or 0)
1143 date_str
= date_str
[:-len(m
.group('tz'))]
1144 if not m
.group('sign'):
1145 timezone
= datetime
.timedelta()
1147 sign
= 1 if m
.group('sign') == '+' else -1
1148 timezone
= datetime
.timedelta(
1149 hours
=sign
* int(m
.group('hours')),
1150 minutes
=sign
* int(m
.group('minutes')))
1151 return timezone
, date_str
1154 def parse_iso8601(date_str
, delimiter
='T', timezone
=None):
1155 """ Return a UNIX timestamp from the given date """
1157 if date_str
is None:
1160 date_str
= re
.sub(r
'\.[0-9]+', '', date_str
)
1162 if timezone
is None:
1163 timezone
, date_str
= extract_timezone(date_str
)
1165 with contextlib
.suppress(ValueError):
1166 date_format
= f
'%Y-%m-%d{delimiter}%H:%M:%S'
1167 dt
= datetime
.datetime
.strptime(date_str
, date_format
) - timezone
1168 return calendar
.timegm(dt
.timetuple())
1171 def date_formats(day_first
=True):
1172 return DATE_FORMATS_DAY_FIRST
if day_first
else DATE_FORMATS_MONTH_FIRST
1175 def unified_strdate(date_str
, day_first
=True):
1176 """Return a string with the date in the format YYYYMMDD"""
1178 if date_str
is None:
1182 date_str
= date_str
.replace(',', ' ')
1183 # Remove AM/PM + timezone
1184 date_str
= re
.sub(r
'(?i)\s*(?:AM|PM)(?:\s+[A-Z]+)?', '', date_str
)
1185 _
, date_str
= extract_timezone(date_str
)
1187 for expression
in date_formats(day_first
):
1188 with contextlib
.suppress(ValueError):
1189 upload_date
= datetime
.datetime
.strptime(date_str
, expression
).strftime('%Y%m%d')
1190 if upload_date
is None:
1191 timetuple
= email
.utils
.parsedate_tz(date_str
)
1193 with contextlib
.suppress(ValueError):
1194 upload_date
= datetime
.datetime(*timetuple
[:6]).strftime('%Y%m%d')
1195 if upload_date
is not None:
1196 return str(upload_date
)
1199 def unified_timestamp(date_str
, day_first
=True):
1200 if not isinstance(date_str
, str):
1203 date_str
= re
.sub(r
'\s+', ' ', re
.sub(
1204 r
'(?i)[,|]|(mon|tues?|wed(nes)?|thu(rs)?|fri|sat(ur)?)(day)?', '', date_str
))
1206 pm_delta
= 12 if re
.search(r
'(?i)PM', date_str
) else 0
1207 timezone
, date_str
= extract_timezone(date_str
)
1209 # Remove AM/PM + timezone
1210 date_str
= re
.sub(r
'(?i)\s*(?:AM|PM)(?:\s+[A-Z]+)?', '', date_str
)
1212 # Remove unrecognized timezones from ISO 8601 alike timestamps
1213 m
= re
.search(r
'\d{1,2}:\d{1,2}(?:\.\d+)?(?P<tz>\s*[A-Z]+)$', date_str
)
1215 date_str
= date_str
[:-len(m
.group('tz'))]
1217 # Python only supports microseconds, so remove nanoseconds
1218 m
= re
.search(r
'^([0-9]{4,}-[0-9]{1,2}-[0-9]{1,2}T[0-9]{1,2}:[0-9]{1,2}:[0-9]{1,2}\.[0-9]{6})[0-9]+$', date_str
)
1220 date_str
= m
.group(1)
1222 for expression
in date_formats(day_first
):
1223 with contextlib
.suppress(ValueError):
1224 dt
= datetime
.datetime
.strptime(date_str
, expression
) - timezone
+ datetime
.timedelta(hours
=pm_delta
)
1225 return calendar
.timegm(dt
.timetuple())
1227 timetuple
= email
.utils
.parsedate_tz(date_str
)
1229 return calendar
.timegm(timetuple
) + pm_delta
* 3600 - timezone
.total_seconds()
1232 def determine_ext(url
, default_ext
='unknown_video'):
1233 if url
is None or '.' not in url
:
1235 guess
= url
.partition('?')[0].rpartition('.')[2]
1236 if re
.match(r
'^[A-Za-z0-9]+$', guess
):
1238 # Try extract ext from URLs like http://example.com/foo/bar.mp4/?download
1239 elif guess
.rstrip('/') in KNOWN_EXTENSIONS
:
1240 return guess
.rstrip('/')
1245 def subtitles_filename(filename
, sub_lang
, sub_format
, expected_real_ext
=None):
1246 return replace_extension(filename
, sub_lang
+ '.' + sub_format
, expected_real_ext
)
1249 def datetime_from_str(date_str
, precision
='auto', format
='%Y%m%d'):
1251 Return a datetime object from a string.
1253 (now|today|yesterday|DATE)([+-]\d+(microsecond|second|minute|hour|day|week|month|year)s?)?
1255 @param format strftime format of DATE
1256 @param precision Round the datetime object: auto|microsecond|second|minute|hour|day
1257 auto: round to the unit provided in date_str (if applicable).
1259 auto_precision
= False
1260 if precision
== 'auto':
1261 auto_precision
= True
1262 precision
= 'microsecond'
1263 today
= datetime_round(datetime
.datetime
.now(datetime
.timezone
.utc
), precision
)
1264 if date_str
in ('now', 'today'):
1266 if date_str
== 'yesterday':
1267 return today
- datetime
.timedelta(days
=1)
1269 r
'(?P<start>.+)(?P<sign>[+-])(?P<time>\d+)(?P<unit>microsecond|second|minute|hour|day|week|month|year)s?',
1271 if match
is not None:
1272 start_time
= datetime_from_str(match
.group('start'), precision
, format
)
1273 time
= int(match
.group('time')) * (-1 if match
.group('sign') == '-' else 1)
1274 unit
= match
.group('unit')
1275 if unit
== 'month' or unit
== 'year':
1276 new_date
= datetime_add_months(start_time
, time
* 12 if unit
== 'year' else time
)
1282 delta
= datetime
.timedelta(**{unit + 's': time}
)
1283 new_date
= start_time
+ delta
1285 return datetime_round(new_date
, unit
)
1288 return datetime_round(datetime
.datetime
.strptime(date_str
, format
), precision
)
1291 def date_from_str(date_str
, format
='%Y%m%d', strict
=False):
1293 Return a date object from a string using datetime_from_str
1295 @param strict Restrict allowed patterns to "YYYYMMDD" and
1296 (now|today|yesterday)(-\d+(day|week|month|year)s?)?
1298 if strict
and not re
.fullmatch(r
'\d{8}|(now|today|yesterday)(-\d+(day|week|month|year)s?)?', date_str
):
1299 raise ValueError(f
'Invalid date format "{date_str}"')
1300 return datetime_from_str(date_str
, precision
='microsecond', format
=format
).date()
1303 def datetime_add_months(dt
, months
):
1304 """Increment/Decrement a datetime object by months."""
1305 month
= dt
.month
+ months
- 1
1306 year
= dt
.year
+ month
// 12
1307 month
= month
% 12 + 1
1308 day
= min(dt
.day
, calendar
.monthrange(year
, month
)[1])
1309 return dt
.replace(year
, month
, day
)
1312 def datetime_round(dt
, precision
='day'):
1314 Round a datetime object's time to a specific precision
1316 if precision
== 'microsecond':
1325 roundto
= lambda x
, n
: ((x
+ n
/ 2) // n
) * n
1326 timestamp
= roundto(calendar
.timegm(dt
.timetuple()), unit_seconds
[precision
])
1327 return datetime
.datetime
.fromtimestamp(timestamp
, datetime
.timezone
.utc
)
1330 def hyphenate_date(date_str
):
1332 Convert a date in 'YYYYMMDD' format to 'YYYY-MM-DD' format"""
1333 match
= re
.match(r
'^(\d\d\d\d)(\d\d)(\d\d)$', date_str
)
1334 if match
is not None:
1335 return '-'.join(match
.groups())
1341 """Represents a time interval between two dates"""
1343 def __init__(self
, start
=None, end
=None):
1344 """start and end must be strings in the format accepted by date"""
1345 if start
is not None:
1346 self
.start
= date_from_str(start
, strict
=True)
1348 self
.start
= datetime
.datetime
.min.date()
1350 self
.end
= date_from_str(end
, strict
=True)
1352 self
.end
= datetime
.datetime
.max.date()
1353 if self
.start
> self
.end
:
1354 raise ValueError('Date range: "%s" , the start date must be before the end date' % self
)
1358 """Returns a range that only contains the given day"""
1359 return cls(day
, day
)
1361 def __contains__(self
, date
):
1362 """Check if the date is in the range"""
1363 if not isinstance(date
, datetime
.date
):
1364 date
= date_from_str(date
)
1365 return self
.start
<= date
<= self
.end
1368 return f
'{__name__}.{type(self).__name__}({self.start.isoformat()!r}, {self.end.isoformat()!r})'
1370 def __eq__(self
, other
):
1371 return (isinstance(other
, DateRange
)
1372 and self
.start
== other
.start
and self
.end
== other
.end
)
1376 def system_identifier():
1377 python_implementation
= platform
.python_implementation()
1378 if python_implementation
== 'PyPy' and hasattr(sys
, 'pypy_version_info'):
1379 python_implementation
+= ' version %d.%d.%d' % sys
.pypy_version_info
[:3]
1381 with contextlib
.suppress(OSError): # We may not have access to the executable
1382 libc_ver
= platform
.libc_ver()
1384 return 'Python %s (%s %s %s) - %s (%s%s)' % (
1385 platform
.python_version(),
1386 python_implementation
,
1388 platform
.architecture()[0],
1389 platform
.platform(),
1390 ssl
.OPENSSL_VERSION
,
1391 format_field(join_nonempty(*libc_ver
, delim
=' '), None, ', %s'),
1396 def get_windows_version():
1397 ''' Get Windows version. returns () if it's not running on Windows '''
1398 if compat_os_name
== 'nt':
1399 return version_tuple(platform
.win32_ver()[1])
1404 def write_string(s
, out
=None, encoding
=None):
1405 assert isinstance(s
, str)
1406 out
= out
or sys
.stderr
1407 # `sys.stderr` might be `None` (Ref: https://github.com/pyinstaller/pyinstaller/pull/7217)
1411 if compat_os_name
== 'nt' and supports_terminal_sequences(out
):
1412 s
= re
.sub(r
'([\r\n]+)', r
' \1', s
)
1414 enc
, buffer = None, out
1415 if 'b' in getattr(out
, 'mode', ''):
1416 enc
= encoding
or preferredencoding()
1417 elif hasattr(out
, 'buffer'):
1419 enc
= encoding
or getattr(out
, 'encoding', None) or preferredencoding()
1421 buffer.write(s
.encode(enc
, 'ignore') if enc
else s
)
1425 # TODO: Use global logger
1426 def deprecation_warning(msg
, *, printer
=None, stacklevel
=0, **kwargs
):
1427 from .. import _IN_CLI
1429 if msg
in deprecation_warning
._cache
:
1431 deprecation_warning
._cache
.add(msg
)
1433 return printer(f
'{msg}{bug_reports_message()}', **kwargs
)
1434 return write_string(f
'ERROR: {msg}{bug_reports_message()}\n', **kwargs
)
1437 warnings
.warn(DeprecationWarning(msg
), stacklevel
=stacklevel
+ 3)
1440 deprecation_warning
._cache
= set()
1443 def bytes_to_intlist(bs
):
1446 if isinstance(bs
[0], int): # Python 3
1449 return [ord(c
) for c
in bs
]
1452 def intlist_to_bytes(xs
):
1455 return struct
.pack('%dB' % len(xs
), *xs
)
1458 class LockingUnsupportedError(OSError):
1459 msg
= 'File locking is not supported'
1462 super().__init
__(self
.msg
)
1465 # Cross-platform file locking
1466 if sys
.platform
== 'win32':
1468 import ctypes
.wintypes
1471 class OVERLAPPED(ctypes
.Structure
):
1473 ('Internal', ctypes
.wintypes
.LPVOID
),
1474 ('InternalHigh', ctypes
.wintypes
.LPVOID
),
1475 ('Offset', ctypes
.wintypes
.DWORD
),
1476 ('OffsetHigh', ctypes
.wintypes
.DWORD
),
1477 ('hEvent', ctypes
.wintypes
.HANDLE
),
1480 kernel32
= ctypes
.WinDLL('kernel32')
1481 LockFileEx
= kernel32
.LockFileEx
1482 LockFileEx
.argtypes
= [
1483 ctypes
.wintypes
.HANDLE
, # hFile
1484 ctypes
.wintypes
.DWORD
, # dwFlags
1485 ctypes
.wintypes
.DWORD
, # dwReserved
1486 ctypes
.wintypes
.DWORD
, # nNumberOfBytesToLockLow
1487 ctypes
.wintypes
.DWORD
, # nNumberOfBytesToLockHigh
1488 ctypes
.POINTER(OVERLAPPED
) # Overlapped
1490 LockFileEx
.restype
= ctypes
.wintypes
.BOOL
1491 UnlockFileEx
= kernel32
.UnlockFileEx
1492 UnlockFileEx
.argtypes
= [
1493 ctypes
.wintypes
.HANDLE
, # hFile
1494 ctypes
.wintypes
.DWORD
, # dwReserved
1495 ctypes
.wintypes
.DWORD
, # nNumberOfBytesToLockLow
1496 ctypes
.wintypes
.DWORD
, # nNumberOfBytesToLockHigh
1497 ctypes
.POINTER(OVERLAPPED
) # Overlapped
1499 UnlockFileEx
.restype
= ctypes
.wintypes
.BOOL
1500 whole_low
= 0xffffffff
1501 whole_high
= 0x7fffffff
1503 def _lock_file(f
, exclusive
, block
):
1504 overlapped
= OVERLAPPED()
1505 overlapped
.Offset
= 0
1506 overlapped
.OffsetHigh
= 0
1507 overlapped
.hEvent
= 0
1508 f
._lock
_file
_overlapped
_p
= ctypes
.pointer(overlapped
)
1510 if not LockFileEx(msvcrt
.get_osfhandle(f
.fileno()),
1511 (0x2 if exclusive
else 0x0) |
(0x0 if block
else 0x1),
1512 0, whole_low
, whole_high
, f
._lock
_file
_overlapped
_p
):
1513 # NB: No argument form of "ctypes.FormatError" does not work on PyPy
1514 raise BlockingIOError(f
'Locking file failed: {ctypes.FormatError(ctypes.GetLastError())!r}')
1516 def _unlock_file(f
):
1517 assert f
._lock
_file
_overlapped
_p
1518 handle
= msvcrt
.get_osfhandle(f
.fileno())
1519 if not UnlockFileEx(handle
, 0, whole_low
, whole_high
, f
._lock
_file
_overlapped
_p
):
1520 raise OSError('Unlocking file failed: %r' % ctypes
.FormatError())
1526 def _lock_file(f
, exclusive
, block
):
1527 flags
= fcntl
.LOCK_EX
if exclusive
else fcntl
.LOCK_SH
1529 flags |
= fcntl
.LOCK_NB
1531 fcntl
.flock(f
, flags
)
1532 except BlockingIOError
:
1534 except OSError: # AOSP does not have flock()
1535 fcntl
.lockf(f
, flags
)
1537 def _unlock_file(f
):
1538 with contextlib
.suppress(OSError):
1539 return fcntl
.flock(f
, fcntl
.LOCK_UN
)
1540 with contextlib
.suppress(OSError):
1541 return fcntl
.lockf(f
, fcntl
.LOCK_UN
) # AOSP does not have flock()
1542 return fcntl
.flock(f
, fcntl
.LOCK_UN | fcntl
.LOCK_NB
) # virtiofs needs LOCK_NB on unlocking
1546 def _lock_file(f
, exclusive
, block
):
1547 raise LockingUnsupportedError()
1549 def _unlock_file(f
):
1550 raise LockingUnsupportedError()
1556 def __init__(self
, filename
, mode
, block
=True, encoding
=None):
1557 if mode
not in {'r', 'rb', 'a', 'ab', 'w', 'wb'}
:
1558 raise NotImplementedError(mode
)
1559 self
.mode
, self
.block
= mode
, block
1561 writable
= any(f
in mode
for f
in 'wax+')
1562 readable
= any(f
in mode
for f
in 'r+')
1563 flags
= functools
.reduce(operator
.ior
, (
1564 getattr(os
, 'O_CLOEXEC', 0), # UNIX only
1565 getattr(os
, 'O_BINARY', 0), # Windows only
1566 getattr(os
, 'O_NOINHERIT', 0), # Windows only
1567 os
.O_CREAT
if writable
else 0, # O_TRUNC only after locking
1568 os
.O_APPEND
if 'a' in mode
else 0,
1569 os
.O_EXCL
if 'x' in mode
else 0,
1570 os
.O_RDONLY
if not writable
else os
.O_RDWR
if readable
else os
.O_WRONLY
,
1573 self
.f
= os
.fdopen(os
.open(filename
, flags
, 0o666), mode
, encoding
=encoding
)
1575 def __enter__(self
):
1576 exclusive
= 'r' not in self
.mode
1578 _lock_file(self
.f
, exclusive
, self
.block
)
1583 if 'w' in self
.mode
:
1586 except OSError as e
:
1588 errno
.ESPIPE
, # Illegal seek - expected for FIFO
1589 errno
.EINVAL
, # Invalid argument - expected for /dev/null
1598 _unlock_file(self
.f
)
1602 def __exit__(self
, *_
):
1611 def __getattr__(self
, attr
):
1612 return getattr(self
.f
, attr
)
1619 def get_filesystem_encoding():
1620 encoding
= sys
.getfilesystemencoding()
1621 return encoding
if encoding
is not None else 'utf-8'
1624 def shell_quote(args
):
1626 encoding
= get_filesystem_encoding()
1628 if isinstance(a
, bytes):
1629 # We may get a filename encoded with 'encodeFilename'
1630 a
= a
.decode(encoding
)
1631 quoted_args
.append(compat_shlex_quote(a
))
1632 return ' '.join(quoted_args
)
1635 def smuggle_url(url
, data
):
1636 """ Pass additional data in a URL for internal use. """
1638 url
, idata
= unsmuggle_url(url
, {})
1640 sdata
= urllib
.parse
.urlencode(
1641 {'__youtubedl_smuggle': json.dumps(data)}
)
1642 return url
+ '#' + sdata
1645 def unsmuggle_url(smug_url
, default
=None):
1646 if '#__youtubedl_smuggle' not in smug_url
:
1647 return smug_url
, default
1648 url
, _
, sdata
= smug_url
.rpartition('#')
1649 jsond
= urllib
.parse
.parse_qs(sdata
)['__youtubedl_smuggle'][0]
1650 data
= json
.loads(jsond
)
1654 def format_decimal_suffix(num
, fmt
='%d%s', *, factor
=1000):
1655 """ Formats numbers with decimal sufixes like K, M, etc """
1656 num
, factor
= float_or_none(num
), float(factor
)
1657 if num
is None or num
< 0:
1659 POSSIBLE_SUFFIXES
= 'kMGTPEZY'
1660 exponent
= 0 if num
== 0 else min(int(math
.log(num
, factor
)), len(POSSIBLE_SUFFIXES
))
1661 suffix
= ['', *POSSIBLE_SUFFIXES
][exponent
]
1663 suffix
= {'k': 'Ki', '': ''}
.get(suffix
, f
'{suffix}i')
1664 converted
= num
/ (factor
** exponent
)
1665 return fmt
% (converted
, suffix
)
1668 def format_bytes(bytes):
1669 return format_decimal_suffix(bytes, '%.2f%sB', factor
=1024) or 'N/A'
1672 def lookup_unit_table(unit_table
, s
, strict
=False):
1673 num_re
= NUMBER_RE
if strict
else NUMBER_RE
.replace(R
'\.', '[,.]')
1674 units_re
= '|'.join(re
.escape(u
) for u
in unit_table
)
1675 m
= (re
.fullmatch
if strict
else re
.match
)(
1676 rf
'(?P<num>{num_re})\s*(?P<unit>{units_re})\b', s
)
1680 num
= float(m
.group('num').replace(',', '.'))
1681 mult
= unit_table
[m
.group('unit')]
1682 return round(num
* mult
)
1686 """Parse a string indicating a byte quantity into an integer"""
1687 return lookup_unit_table(
1688 {u: 1024**i for i, u in enumerate(['', *'KMGTPEZY'])}
,
1689 s
.upper(), strict
=True)
1692 def parse_filesize(s
):
1696 # The lower-case forms are of course incorrect and unofficial,
1697 # but we support those too
1714 'megabytes': 1000 ** 2,
1715 'mebibytes': 1024 ** 2,
1721 'gigabytes': 1000 ** 3,
1722 'gibibytes': 1024 ** 3,
1728 'terabytes': 1000 ** 4,
1729 'tebibytes': 1024 ** 4,
1735 'petabytes': 1000 ** 5,
1736 'pebibytes': 1024 ** 5,
1742 'exabytes': 1000 ** 6,
1743 'exbibytes': 1024 ** 6,
1749 'zettabytes': 1000 ** 7,
1750 'zebibytes': 1024 ** 7,
1756 'yottabytes': 1000 ** 8,
1757 'yobibytes': 1024 ** 8,
1760 return lookup_unit_table(_UNIT_TABLE
, s
)
1767 s
= re
.sub(r
'^[^\d]+\s', '', s
).strip()
1769 if re
.match(r
'^[\d,.]+$', s
):
1770 return str_to_int(s
)
1783 ret
= lookup_unit_table(_UNIT_TABLE
, s
)
1787 mobj
= re
.match(r
'([\d,.]+)(?:$|\s)', s
)
1789 return str_to_int(mobj
.group(1))
1792 def parse_resolution(s
, *, lenient
=False):
1797 mobj
= re
.search(r
'(?P<w>\d+)\s*[xX×,]\s*(?P<h>\d+)', s
)
1799 mobj
= re
.search(r
'(?<![a-zA-Z0-9])(?P<w>\d+)\s*[xX×,]\s*(?P<h>\d+)(?![a-zA-Z0-9])', s
)
1802 'width': int(mobj
.group('w')),
1803 'height': int(mobj
.group('h')),
1806 mobj
= re
.search(r
'(?<![a-zA-Z0-9])(\d+)[pPiI](?![a-zA-Z0-9])', s
)
1808 return {'height': int(mobj.group(1))}
1810 mobj
= re
.search(r
'\b([48])[kK]\b', s
)
1812 return {'height': int(mobj.group(1)) * 540}
1817 def parse_bitrate(s
):
1818 if not isinstance(s
, str):
1820 mobj
= re
.search(r
'\b(\d+)\s*kbps', s
)
1822 return int(mobj
.group(1))
1825 def month_by_name(name
, lang
='en'):
1826 """ Return the number of a month by (locale-independently) English name """
1828 month_names
= MONTH_NAMES
.get(lang
, MONTH_NAMES
['en'])
1831 return month_names
.index(name
) + 1
1836 def month_by_abbreviation(abbrev
):
1837 """ Return the number of a month by (locale-independently) English
1841 return [s
[:3] for s
in ENGLISH_MONTH_NAMES
].index(abbrev
) + 1
1846 def fix_xml_ampersands(xml_str
):
1847 """Replace all the '&' by '&' in XML"""
1849 r
'&(?!amp;|lt;|gt;|apos;|quot;|#x[0-9a-fA-F]{,4};|#[0-9]{,4};)',
1854 def setproctitle(title
):
1855 assert isinstance(title
, str)
1857 # Workaround for https://github.com/yt-dlp/yt-dlp/issues/4541
1864 libc
= ctypes
.cdll
.LoadLibrary('libc.so.6')
1868 # LoadLibrary in Windows Python 2.7.13 only expects
1869 # a bytestring, but since unicode_literals turns
1870 # every string into a unicode string, it fails.
1872 title_bytes
= title
.encode()
1873 buf
= ctypes
.create_string_buffer(len(title_bytes
))
1874 buf
.value
= title_bytes
1876 libc
.prctl(15, buf
, 0, 0, 0)
1877 except AttributeError:
1878 return # Strange libc, just skip this
1881 def remove_start(s
, start
):
1882 return s
[len(start
):] if s
is not None and s
.startswith(start
) else s
1885 def remove_end(s
, end
):
1886 return s
[:-len(end
)] if s
is not None and s
.endswith(end
) else s
1889 def remove_quotes(s
):
1890 if s
is None or len(s
) < 2:
1892 for quote
in ('"', "'", ):
1893 if s
[0] == quote
and s
[-1] == quote
:
1898 def get_domain(url
):
1900 This implementation is inconsistent, but is kept for compatibility.
1901 Use this only for "webpage_url_domain"
1903 return remove_start(urllib
.parse
.urlparse(url
).netloc
, 'www.') or None
1906 def url_basename(url
):
1907 path
= urllib
.parse
.urlparse(url
).path
1908 return path
.strip('/').split('/')[-1]
1912 return re
.match(r
'https?://[^?#]+/', url
).group()
1915 def urljoin(base
, path
):
1916 if isinstance(path
, bytes):
1917 path
= path
.decode()
1918 if not isinstance(path
, str) or not path
:
1920 if re
.match(r
'^(?:[a-zA-Z][a-zA-Z0-9+-.]*:)?//', path
):
1922 if isinstance(base
, bytes):
1923 base
= base
.decode()
1924 if not isinstance(base
, str) or not re
.match(
1925 r
'^(?:https?:)?//', base
):
1927 return urllib
.parse
.urljoin(base
, path
)
1930 def int_or_none(v
, scale
=1, default
=None, get_attr
=None, invscale
=1):
1931 if get_attr
and v
is not None:
1932 v
= getattr(v
, get_attr
, None)
1934 return int(v
) * invscale
// scale
1935 except (ValueError, TypeError, OverflowError):
1939 def str_or_none(v
, default
=None):
1940 return default
if v
is None else str(v
)
1943 def str_to_int(int_str
):
1944 """ A more relaxed version of int_or_none """
1945 if isinstance(int_str
, int):
1947 elif isinstance(int_str
, str):
1948 int_str
= re
.sub(r
'[,\.\+]', '', int_str
)
1949 return int_or_none(int_str
)
1952 def float_or_none(v
, scale
=1, invscale
=1, default
=None):
1956 return float(v
) * invscale
/ scale
1957 except (ValueError, TypeError):
1961 def bool_or_none(v
, default
=None):
1962 return v
if isinstance(v
, bool) else default
1965 def strip_or_none(v
, default
=None):
1966 return v
.strip() if isinstance(v
, str) else default
1969 def url_or_none(url
):
1970 if not url
or not isinstance(url
, str):
1973 return url
if re
.match(r
'^(?:(?:https?|rt(?:m(?:pt?[es]?|fp)|sp[su]?)|mms|ftps?):)?//', url
) else None
1976 def strftime_or_none(timestamp
, date_format
='%Y%m%d', default
=None):
1977 datetime_object
= None
1979 if isinstance(timestamp
, (int, float)): # unix timestamp
1980 # Using naive datetime here can break timestamp() in Windows
1981 # Ref: https://github.com/yt-dlp/yt-dlp/issues/5185, https://github.com/python/cpython/issues/94414
1982 # Also, datetime.datetime.fromtimestamp breaks for negative timestamps
1983 # Ref: https://github.com/yt-dlp/yt-dlp/issues/6706#issuecomment-1496842642
1984 datetime_object
= (datetime
.datetime
.fromtimestamp(0, datetime
.timezone
.utc
)
1985 + datetime
.timedelta(seconds
=timestamp
))
1986 elif isinstance(timestamp
, str): # assume YYYYMMDD
1987 datetime_object
= datetime
.datetime
.strptime(timestamp
, '%Y%m%d')
1988 date_format
= re
.sub( # Support %s on windows
1989 r
'(?<!%)(%%)*%s', rf
'\g<1>{int(datetime_object.timestamp())}', date_format
)
1990 return datetime_object
.strftime(date_format
)
1991 except (ValueError, TypeError, AttributeError):
1995 def parse_duration(s
):
1996 if not isinstance(s
, str):
2002 days
, hours
, mins
, secs
, ms
= [None] * 5
2003 m
= re
.match(r
'''(?x)
2005 (?:(?:(?P<days>[0-9]+):)?(?P<hours>[0-9]+):)?(?P<mins>[0-9]+):)?
2006 (?P<secs>(?(before_secs)[0-9]{1,2}|[0-9]+))
2007 (?P<ms>[.:][0-9]+)?Z?$
2010 days
, hours
, mins
, secs
, ms
= m
.group('days', 'hours', 'mins', 'secs', 'ms')
2015 [0-9]+\s*y(?:ears?)?,?\s*
2018 [0-9]+\s*m(?:onths?)?,?\s*
2021 [0-9]+\s*w(?:eeks?)?,?\s*
2024 (?P<days>[0-9]+)\s*d(?:ays?)?,?\s*
2028 (?P<hours>[0-9]+)\s*h(?:(?:ou)?rs?)?,?\s*
2031 (?P<mins>[0-9]+)\s*m(?:in(?:ute)?s?)?,?\s*
2034 (?P<secs>[0-9]+)(?P<ms>\.[0-9]+)?\s*s(?:ec(?:ond)?s?)?\s*
2037 days
, hours
, mins
, secs
, ms
= m
.groups()
2039 m
= re
.match(r
'(?i)(?:(?P<hours>[0-9.]+)\s*(?:hours?)|(?P<mins>[0-9.]+)\s*(?:mins?\.?|minutes?)\s*)Z?$', s
)
2041 hours
, mins
= m
.groups()
2046 ms
= ms
.replace(':', '.')
2047 return sum(float(part
or 0) * mult
for part
, mult
in (
2048 (days
, 86400), (hours
, 3600), (mins
, 60), (secs
, 1), (ms
, 1)))
2051 def prepend_extension(filename
, ext
, expected_real_ext
=None):
2052 name
, real_ext
= os
.path
.splitext(filename
)
2054 f
'{name}.{ext}{real_ext}'
2055 if not expected_real_ext
or real_ext
[1:] == expected_real_ext
2056 else f
'{filename}.{ext}')
2059 def replace_extension(filename
, ext
, expected_real_ext
=None):
2060 name
, real_ext
= os
.path
.splitext(filename
)
2061 return '{}.{}'.format(
2062 name
if not expected_real_ext
or real_ext
[1:] == expected_real_ext
else filename
,
2066 def check_executable(exe
, args
=[]):
2067 """ Checks if the given binary is installed somewhere in PATH, and returns its name.
2068 args can be a list of arguments for a short output (like -version) """
2070 Popen
.run([exe
] + args
, stdout
=subprocess
.PIPE
, stderr
=subprocess
.PIPE
)
2076 def _get_exe_version_output(exe
, args
):
2078 # STDIN should be redirected too. On UNIX-like systems, ffmpeg triggers
2079 # SIGTTOU if yt-dlp is run in the background.
2080 # See https://github.com/ytdl-org/youtube-dl/issues/955#issuecomment-209789656
2081 stdout
, _
, ret
= Popen
.run([encodeArgument(exe
)] + args
, text
=True,
2082 stdin
=subprocess
.PIPE
, stdout
=subprocess
.PIPE
, stderr
=subprocess
.STDOUT
)
2090 def detect_exe_version(output
, version_re
=None, unrecognized
='present'):
2091 assert isinstance(output
, str)
2092 if version_re
is None:
2093 version_re
= r
'version\s+([-0-9._a-zA-Z]+)'
2094 m
= re
.search(version_re
, output
)
2101 def get_exe_version(exe
, args
=['--version'],
2102 version_re
=None, unrecognized
=('present', 'broken')):
2103 """ Returns the version of the specified executable,
2104 or False if the executable is not present """
2105 unrecognized
= variadic(unrecognized
)
2106 assert len(unrecognized
) in (1, 2)
2107 out
= _get_exe_version_output(exe
, args
)
2109 return unrecognized
[-1]
2110 return out
and detect_exe_version(out
, version_re
, unrecognized
[0])
2113 def frange(start
=0, stop
=None, step
=1):
2116 start
, stop
= 0, start
2117 sign
= [-1, 1][step
> 0] if step
else 0
2118 while sign
* start
< sign
* stop
:
2123 class LazyList(collections
.abc
.Sequence
):
2124 """Lazy immutable list from an iterable
2125 Note that slices of a LazyList are lists and not LazyList"""
2127 class IndexError(IndexError):
2130 def __init__(self
, iterable
, *, reverse
=False, _cache
=None):
2131 self
._iterable
= iter(iterable
)
2132 self
._cache
= [] if _cache
is None else _cache
2133 self
._reversed
= reverse
2137 # We need to consume the entire iterable to iterate in reverse
2138 yield from self
.exhaust()
2140 yield from self
._cache
2141 for item
in self
._iterable
:
2142 self
._cache
.append(item
)
2146 self
._cache
.extend(self
._iterable
)
2147 self
._iterable
= [] # Discard the emptied iterable to make it pickle-able
2151 """Evaluate the entire iterable"""
2152 return self
._exhaust
()[::-1 if self
._reversed
else 1]
2155 def _reverse_index(x
):
2156 return None if x
is None else ~x
2158 def __getitem__(self
, idx
):
2159 if isinstance(idx
, slice):
2161 idx
= slice(self
._reverse
_index
(idx
.start
), self
._reverse
_index
(idx
.stop
), -(idx
.step
or 1))
2162 start
, stop
, step
= idx
.start
, idx
.stop
, idx
.step
or 1
2163 elif isinstance(idx
, int):
2165 idx
= self
._reverse
_index
(idx
)
2166 start
, stop
, step
= idx
, idx
, 0
2168 raise TypeError('indices must be integers or slices')
2169 if ((start
or 0) < 0 or (stop
or 0) < 0
2170 or (start
is None and step
< 0)
2171 or (stop
is None and step
> 0)):
2172 # We need to consume the entire iterable to be able to slice from the end
2173 # Obviously, never use this with infinite iterables
2176 return self
._cache
[idx
]
2177 except IndexError as e
:
2178 raise self
.IndexError(e
) from e
2179 n
= max(start
or 0, stop
or 0) - len(self
._cache
) + 1
2181 self
._cache
.extend(itertools
.islice(self
._iterable
, n
))
2183 return self
._cache
[idx
]
2184 except IndexError as e
:
2185 raise self
.IndexError(e
) from e
2189 self
[-1] if self
._reversed
else self
[0]
2190 except self
.IndexError:
2196 return len(self
._cache
)
2198 def __reversed__(self
):
2199 return type(self
)(self
._iterable
, reverse
=not self
._reversed
, _cache
=self
._cache
)
2202 return type(self
)(self
._iterable
, reverse
=self
._reversed
, _cache
=self
._cache
)
2205 # repr and str should mimic a list. So we exhaust the iterable
2206 return repr(self
.exhaust())
2209 return repr(self
.exhaust())
2214 class IndexError(IndexError):
2218 # This is only useful for tests
2219 return len(self
.getslice())
2221 def __init__(self
, pagefunc
, pagesize
, use_cache
=True):
2222 self
._pagefunc
= pagefunc
2223 self
._pagesize
= pagesize
2224 self
._pagecount
= float('inf')
2225 self
._use
_cache
= use_cache
2228 def getpage(self
, pagenum
):
2229 page_results
= self
._cache
.get(pagenum
)
2230 if page_results
is None:
2231 page_results
= [] if pagenum
> self
._pagecount
else list(self
._pagefunc
(pagenum
))
2233 self
._cache
[pagenum
] = page_results
2236 def getslice(self
, start
=0, end
=None):
2237 return list(self
._getslice
(start
, end
))
2239 def _getslice(self
, start
, end
):
2240 raise NotImplementedError('This method must be implemented by subclasses')
2242 def __getitem__(self
, idx
):
2243 assert self
._use
_cache
, 'Indexing PagedList requires cache'
2244 if not isinstance(idx
, int) or idx
< 0:
2245 raise TypeError('indices must be non-negative integers')
2246 entries
= self
.getslice(idx
, idx
+ 1)
2248 raise self
.IndexError()
2252 class OnDemandPagedList(PagedList
):
2253 """Download pages until a page with less than maximum results"""
2255 def _getslice(self
, start
, end
):
2256 for pagenum
in itertools
.count(start
// self
._pagesize
):
2257 firstid
= pagenum
* self
._pagesize
2258 nextfirstid
= pagenum
* self
._pagesize
+ self
._pagesize
2259 if start
>= nextfirstid
:
2263 start
% self
._pagesize
2264 if firstid
<= start
< nextfirstid
2267 ((end
- 1) % self
._pagesize
) + 1
2268 if (end
is not None and firstid
<= end
<= nextfirstid
)
2272 page_results
= self
.getpage(pagenum
)
2274 self
._pagecount
= pagenum
- 1
2276 if startv
!= 0 or endv
is not None:
2277 page_results
= page_results
[startv
:endv
]
2278 yield from page_results
2280 # A little optimization - if current page is not "full", ie. does
2281 # not contain page_size videos then we can assume that this page
2282 # is the last one - there are no more ids on further pages -
2283 # i.e. no need to query again.
2284 if len(page_results
) + startv
< self
._pagesize
:
2287 # If we got the whole page, but the next page is not interesting,
2288 # break out early as well
2289 if end
== nextfirstid
:
2293 class InAdvancePagedList(PagedList
):
2294 """PagedList with total number of pages known in advance"""
2296 def __init__(self
, pagefunc
, pagecount
, pagesize
):
2297 PagedList
.__init
__(self
, pagefunc
, pagesize
, True)
2298 self
._pagecount
= pagecount
2300 def _getslice(self
, start
, end
):
2301 start_page
= start
// self
._pagesize
2302 end_page
= self
._pagecount
if end
is None else min(self
._pagecount
, end
// self
._pagesize
+ 1)
2303 skip_elems
= start
- start_page
* self
._pagesize
2304 only_more
= None if end
is None else end
- start
2305 for pagenum
in range(start_page
, end_page
):
2306 page_results
= self
.getpage(pagenum
)
2308 page_results
= page_results
[skip_elems
:]
2310 if only_more
is not None:
2311 if len(page_results
) < only_more
:
2312 only_more
-= len(page_results
)
2314 yield from page_results
[:only_more
]
2316 yield from page_results
2319 class PlaylistEntries
:
2320 MissingEntry
= object()
2321 is_exhausted
= False
2323 def __init__(self
, ydl
, info_dict
):
2326 # _entries must be assigned now since infodict can change during iteration
2327 entries
= info_dict
.get('entries')
2329 raise EntryNotInPlaylist('There are no entries')
2330 elif isinstance(entries
, list):
2331 self
.is_exhausted
= True
2333 requested_entries
= info_dict
.get('requested_entries')
2334 self
.is_incomplete
= requested_entries
is not None
2335 if self
.is_incomplete
:
2336 assert self
.is_exhausted
2337 self
._entries
= [self
.MissingEntry
] * max(requested_entries
or [0])
2338 for i
, entry
in zip(requested_entries
, entries
):
2339 self
._entries
[i
- 1] = entry
2340 elif isinstance(entries
, (list, PagedList
, LazyList
)):
2341 self
._entries
= entries
2343 self
._entries
= LazyList(entries
)
2345 PLAYLIST_ITEMS_RE
= re
.compile(r
'''(?x)
2346 (?P<start>[+-]?\d+)?
2348 (?P<end>[+-]?\d+|inf(?:inite)?)?
2349 (?::(?P<step>[+-]?\d+))?
2353 def parse_playlist_items(cls
, string
):
2354 for segment
in string
.split(','):
2356 raise ValueError('There is two or more consecutive commas')
2357 mobj
= cls
.PLAYLIST_ITEMS_RE
.fullmatch(segment
)
2359 raise ValueError(f
'{segment!r} is not a valid specification')
2360 start
, end
, step
, has_range
= mobj
.group('start', 'end', 'step', 'range')
2361 if int_or_none(step
) == 0:
2362 raise ValueError(f
'Step in {segment!r} cannot be zero')
2363 yield slice(int_or_none(start
), float_or_none(end
), int_or_none(step
)) if has_range
else int(start
)
2365 def get_requested_items(self
):
2366 playlist_items
= self
.ydl
.params
.get('playlist_items')
2367 playlist_start
= self
.ydl
.params
.get('playliststart', 1)
2368 playlist_end
= self
.ydl
.params
.get('playlistend')
2369 # For backwards compatibility, interpret -1 as whole list
2370 if playlist_end
in (-1, None):
2372 if not playlist_items
:
2373 playlist_items
= f
'{playlist_start}:{playlist_end}'
2374 elif playlist_start
!= 1 or playlist_end
:
2375 self
.ydl
.report_warning('Ignoring playliststart and playlistend because playlistitems was given', only_once
=True)
2377 for index
in self
.parse_playlist_items(playlist_items
):
2378 for i
, entry
in self
[index
]:
2383 # The item may have just been added to archive. Don't break due to it
2384 if not self
.ydl
.params
.get('lazy_playlist'):
2385 # TODO: Add auto-generated fields
2386 self
.ydl
._match
_entry
(entry
, incomplete
=True, silent
=True)
2387 except (ExistingVideoReached
, RejectedVideoReached
):
2390 def get_full_count(self
):
2391 if self
.is_exhausted
and not self
.is_incomplete
:
2393 elif isinstance(self
._entries
, InAdvancePagedList
):
2394 if self
._entries
._pagesize
== 1:
2395 return self
._entries
._pagecount
2397 @functools.cached_property
2399 if isinstance(self
._entries
, list):
2402 entry
= self
._entries
[i
]
2404 entry
= self
.MissingEntry
2405 if not self
.is_incomplete
:
2406 raise self
.IndexError()
2407 if entry
is self
.MissingEntry
:
2408 raise EntryNotInPlaylist(f
'Entry {i + 1} cannot be found')
2413 return type(self
.ydl
)._handle
_extraction
_exceptions
(lambda _
, i
: self
._entries
[i
])(self
.ydl
, i
)
2414 except (LazyList
.IndexError, PagedList
.IndexError):
2415 raise self
.IndexError()
2418 def __getitem__(self
, idx
):
2419 if isinstance(idx
, int):
2420 idx
= slice(idx
, idx
)
2422 # NB: PlaylistEntries[1:10] => (0, 1, ... 9)
2423 step
= 1 if idx
.step
is None else idx
.step
2424 if idx
.start
is None:
2425 start
= 0 if step
> 0 else len(self
) - 1
2427 start
= idx
.start
- 1 if idx
.start
>= 0 else len(self
) + idx
.start
2429 # NB: Do not call len(self) when idx == [:]
2430 if idx
.stop
is None:
2431 stop
= 0 if step
< 0 else float('inf')
2433 stop
= idx
.stop
- 1 if idx
.stop
>= 0 else len(self
) + idx
.stop
2434 stop
+= [-1, 1][step
> 0]
2436 for i
in frange(start
, stop
, step
):
2440 entry
= self
._getter
(i
)
2441 except self
.IndexError:
2442 self
.is_exhausted
= True
2449 return len(tuple(self
[:]))
2451 class IndexError(IndexError):
2455 def uppercase_escape(s
):
2456 unicode_escape
= codecs
.getdecoder('unicode_escape')
2458 r
'\\U[0-9a-fA-F]{8}',
2459 lambda m
: unicode_escape(m
.group(0))[0],
2463 def lowercase_escape(s
):
2464 unicode_escape
= codecs
.getdecoder('unicode_escape')
2466 r
'\\u[0-9a-fA-F]{4}',
2467 lambda m
: unicode_escape(m
.group(0))[0],
2471 def parse_qs(url
, **kwargs
):
2472 return urllib
.parse
.parse_qs(urllib
.parse
.urlparse(url
).query
, **kwargs
)
2475 def read_batch_urls(batch_fd
):
2477 if not isinstance(url
, str):
2478 url
= url
.decode('utf-8', 'replace')
2479 BOM_UTF8
= ('\xef\xbb\xbf', '\ufeff')
2480 for bom
in BOM_UTF8
:
2481 if url
.startswith(bom
):
2482 url
= url
[len(bom
):]
2484 if not url
or url
.startswith(('#', ';', ']')):
2486 # "#" cannot be stripped out since it is part of the URI
2487 # However, it can be safely stripped out if following a whitespace
2488 return re
.split(r
'\s#', url
, 1)[0].rstrip()
2490 with contextlib
.closing(batch_fd
) as fd
:
2491 return [url
for url
in map(fixup
, fd
) if url
]
2494 def urlencode_postdata(*args
, **kargs
):
2495 return urllib
.parse
.urlencode(*args
, **kargs
).encode('ascii')
2498 def update_url(url
, *, query_update
=None, **kwargs
):
2499 """Replace URL components specified by kwargs
2500 @param url str or parse url tuple
2501 @param query_update update query
2504 if isinstance(url
, str):
2505 if not kwargs
and not query_update
:
2508 url
= urllib
.parse
.urlparse(url
)
2510 assert 'query' not in kwargs
, 'query_update and query cannot be specified at the same time'
2511 kwargs
['query'] = urllib
.parse
.urlencode({
2512 **urllib
.parse
.parse_qs(url
.query
),
2515 return urllib
.parse
.urlunparse(url
._replace
(**kwargs
))
2518 def update_url_query(url
, query
):
2519 return update_url(url
, query_update
=query
)
2522 def _multipart_encode_impl(data
, boundary
):
2523 content_type
= 'multipart/form-data; boundary=%s' % boundary
2526 for k
, v
in data
.items():
2527 out
+= b
'--' + boundary
.encode('ascii') + b
'\r\n'
2528 if isinstance(k
, str):
2530 if isinstance(v
, str):
2532 # RFC 2047 requires non-ASCII field names to be encoded, while RFC 7578
2533 # suggests sending UTF-8 directly. Firefox sends UTF-8, too
2534 content
= b
'Content-Disposition: form-data; name="' + k
+ b
'"\r\n\r\n' + v
+ b
'\r\n'
2535 if boundary
.encode('ascii') in content
:
2536 raise ValueError('Boundary overlaps with data')
2539 out
+= b
'--' + boundary
.encode('ascii') + b
'--\r\n'
2541 return out
, content_type
2544 def multipart_encode(data
, boundary
=None):
2546 Encode a dict to RFC 7578-compliant form-data
2549 A dict where keys and values can be either Unicode or bytes-like
2552 If specified a Unicode object, it's used as the boundary. Otherwise
2553 a random boundary is generated.
2555 Reference: https://tools.ietf.org/html/rfc7578
2557 has_specified_boundary
= boundary
is not None
2560 if boundary
is None:
2561 boundary
= '---------------' + str(random
.randrange(0x0fffffff, 0xffffffff))
2564 out
, content_type
= _multipart_encode_impl(data
, boundary
)
2567 if has_specified_boundary
:
2571 return out
, content_type
2574 def is_iterable_like(x
, allowed_types
=collections
.abc
.Iterable
, blocked_types
=NO_DEFAULT
):
2575 if blocked_types
is NO_DEFAULT
:
2576 blocked_types
= (str, bytes, collections
.abc
.Mapping
)
2577 return isinstance(x
, allowed_types
) and not isinstance(x
, blocked_types
)
2580 def variadic(x
, allowed_types
=NO_DEFAULT
):
2581 if not isinstance(allowed_types
, (tuple, type)):
2582 deprecation_warning('allowed_types should be a tuple or a type')
2583 allowed_types
= tuple(allowed_types
)
2584 return x
if is_iterable_like(x
, blocked_types
=allowed_types
) else (x
, )
2587 def try_call(*funcs
, expected_type
=None, args
=[], kwargs
={}):
2590 val
= f(*args
, **kwargs
)
2591 except (AttributeError, KeyError, TypeError, IndexError, ValueError, ZeroDivisionError):
2594 if expected_type
is None or isinstance(val
, expected_type
):
2598 def try_get(src
, getter
, expected_type
=None):
2599 return try_call(*variadic(getter
), args
=(src
,), expected_type
=expected_type
)
2602 def filter_dict(dct
, cndn
=lambda _
, v
: v
is not None):
2603 return {k: v for k, v in dct.items() if cndn(k, v)}
2606 def merge_dicts(*dicts
):
2608 for a_dict
in dicts
:
2609 for k
, v
in a_dict
.items():
2610 if (v
is not None and k
not in merged
2611 or isinstance(v
, str) and merged
[k
] == ''):
2616 def encode_compat_str(string
, encoding
=preferredencoding(), errors
='strict'):
2617 return string
if isinstance(string
, str) else str(string
, encoding
, errors
)
2629 TV_PARENTAL_GUIDELINES
= {
2639 def parse_age_limit(s
):
2640 # isinstance(False, int) is True. So type() must be used instead
2641 if type(s
) is int: # noqa: E721
2642 return s
if 0 <= s
<= 21 else None
2643 elif not isinstance(s
, str):
2645 m
= re
.match(r
'^(?P<age>\d{1,2})\+?$', s
)
2647 return int(m
.group('age'))
2650 return US_RATINGS
[s
]
2651 m
= re
.match(r
'^TV[_-]?(%s)$' % '|'.join(k
[3:] for k
in TV_PARENTAL_GUIDELINES
), s
)
2653 return TV_PARENTAL_GUIDELINES
['TV-' + m
.group(1)]
2657 def strip_jsonp(code
):
2660 (?:window\.)?(?P<func_name>[a-zA-Z0-9_.$]*)
2661 (?:\s*&&\s*(?P=func_name))?
2662 \s*\(\s*(?P<callback_data>.*)\);?
2663 \s*?(?://[^\n]*)*$''',
2664 r
'\g<callback_data>', code
)
2667 def js_to_json(code
, vars={}, *, strict
=False):
2668 # vars is a dict of var, val pairs to substitute
2669 STRING_QUOTES
= '\'"`'
2670 STRING_RE
= '|'.join(rf
'{q}(?:\\.|[^\\{q}])*{q}' for q
in STRING_QUOTES
)
2671 COMMENT_RE
= r
'/\*(?:(?!\*/).)*?\*/|//[^\n]*\n'
2672 SKIP_RE
= fr
'\s*(?:{COMMENT_RE})?\s*'
2674 (fr
'(?s)^(0[xX][0-9a-fA-F]+){SKIP_RE}:?$', 16),
2675 (fr
'(?s)^(0+[0-7]+){SKIP_RE}:?$', 8),
2678 def process_escape(match
):
2679 JSON_PASSTHROUGH_ESCAPES
= R
'"\bfnrtu'
2680 escape
= match
.group(1) or match
.group(2)
2682 return (Rf
'\{escape}' if escape
in JSON_PASSTHROUGH_ESCAPES
2683 else R
'\u00' if escape
== 'x'
2684 else '' if escape
== '\n'
2687 def template_substitute(match
):
2688 evaluated
= js_to_json(match
.group(1), vars, strict
=strict
)
2689 if evaluated
[0] == '"':
2690 return json
.loads(evaluated
)
2695 if v
in ('true', 'false', 'null'):
2697 elif v
in ('undefined', 'void 0'):
2699 elif v
.startswith('/*') or v
.startswith('//') or v
.startswith('!') or v
== ',':
2702 if v
[0] in STRING_QUOTES
:
2703 v
= re
.sub(r
'(?s)\${([^}]+)}', template_substitute
, v
[1:-1]) if v
[0] == '`' else v
[1:-1]
2704 escaped
= re
.sub(r
'(?s)(")|\\(.)', process_escape
, v
)
2705 return f
'"{escaped}"'
2707 for regex
, base
in INTEGER_TABLE
:
2708 im
= re
.match(regex
, v
)
2710 i
= int(im
.group(1), base
)
2711 return f
'"{i}":' if v
.endswith(':') else str(i
)
2717 except json
.JSONDecodeError
:
2718 return json
.dumps(vars[v
])
2725 raise ValueError(f
'Unknown value: {v}')
2727 def create_map(mobj
):
2728 return json
.dumps(dict(json
.loads(js_to_json(mobj
.group(1) or '[]', vars=vars))))
2730 code
= re
.sub(r
'new Map\((\[.*?\])?\)', create_map
, code
)
2732 code
= re
.sub(r
'new Date\((".+")\)', r
'\g<1>', code
)
2733 code
= re
.sub(r
'new \w+\((.*?)\)', lambda m
: json
.dumps(m
.group(0)), code
)
2734 code
= re
.sub(r
'parseInt\([^\d]+(\d+)[^\d]+\)', r
'\1', code
)
2735 code
= re
.sub(r
'\(function\([^)]*\)\s*\{[^}]*\}\s*\)\s*\(\s*(["\'][^
)]*["\'])\s*\)', r'\1', code)
2737 return re.sub(rf'''(?sx)
2739 {COMMENT_RE}|,(?={SKIP_RE}[\]}}])|
2740 void\s0|(?:(?<![0-9])[eE]|[a-df-zA-DF-Z_$])[.a-zA-Z_$0-9]*|
2741 \b(?:0[xX][0-9a-fA-F]+|0+[0-7]+)(?:{SKIP_RE}:)?|
2742 [0-9]+(?={SKIP_RE}:)|
2747 def qualities(quality_ids):
2748 """ Get a numeric quality value out of a list of possible values """
2751 return quality_ids.index(qid)
2757 POSTPROCESS_WHEN = ('pre_process', 'after_filter', 'video', 'before_dl', 'post_process', 'after_move', 'after_video', 'playlist')
2761 'default': '%(title)s [%(id)s].%(ext)s',
2762 'chapter': '%(title)s - %(section_number)03d %(section_title)s [%(id)s].%(ext)s',
2768 'description': 'description',
2769 'annotation': 'annotations.xml',
2770 'infojson': 'info.json',
2773 'pl_thumbnail': None,
2774 'pl_description': 'description',
2775 'pl_infojson': 'info.json',
2778 # As of [1] format syntax is:
2779 # %[mapping_key][conversion_flags][minimum_width][.precision][length_modifier]type
2780 # 1. https://docs.python.org/2/library/stdtypes.html#string-formatting
2781 STR_FORMAT_RE_TMPL = r'''(?x)
2782 (?<!%)(?P<prefix>(?:%%)*)
2784 (?P<has_key>\((?P<key>{0})\))?
2786 (?P<conversion>[#0\-+ ]+)?
2788 (?P<precision>\.\d+)?
2789 (?P<len_mod>[hlL])? # unused in python
2790 {1} # conversion type
2795 STR_FORMAT_TYPES = 'diouxXeEfFgGcrsa'
2798 def limit_length(s, length):
2799 """ Add ellipses to overly long strings """
2804 return s[:length - len(ELLIPSES)] + ELLIPSES
2808 def version_tuple(v):
2809 return tuple(int(e) for e in re.split(r'[-.]', v))
2812 def is_outdated_version(version, limit, assume_new=True):
2814 return not assume_new
2816 return version_tuple(version) < version_tuple(limit)
2818 return not assume_new
2821 def ytdl_is_updateable():
2822 """ Returns if yt-dlp can be updated with -U """
2824 from ..update import is_non_updateable
2826 return not is_non_updateable()
2829 def args_to_str(args):
2830 # Get a short string representation for a subprocess command
2831 return ' '.join(compat_shlex_quote(a) for a in args)
2834 def error_to_str(err):
2835 return f'{type(err).__name__}: {err}'
2838 def mimetype2ext(mt, default=NO_DEFAULT):
2839 if not isinstance(mt, str):
2840 if default is not NO_DEFAULT:
2857 'x-matroska': 'mkv',
2859 'x-mp4-fragmented': 'mp4',
2864 # application (streaming playlists)
2868 'vnd.apple.mpegurl': 'm3u8',
2869 'vnd.ms-sstr+xml': 'ism',
2870 'x-mpegurl': 'm3u8',
2874 # Per RFC 3003, audio/mpeg can be .mp1, .mp2 or .mp3.
2875 # Using .mp3 as it's the most popular one
2876 'audio/mpeg': 'mp3',
2877 'audio/webm': 'webm',
2878 'audio/x-matroska': 'mka',
2879 'audio/x-mpegurl': 'm3u',
2887 'x-realaudio': 'ra',
2898 'vnd.wap.wbmp': 'wbmp',
2905 'filmstrip+json': 'fs',
2906 'smptett+xml': 'tt',
2909 'x-ms-sami': 'sami',
2918 mimetype = mt.partition(';')[0].strip().lower()
2919 _, _, subtype = mimetype.rpartition('/')
2921 ext = traversal.traverse_obj(MAP, mimetype, subtype, subtype.rsplit('+')[-1])
2924 elif default is not NO_DEFAULT:
2926 return subtype.replace('+', '.')
2929 def ext2mimetype(ext_or_url):
2932 if '.' not in ext_or_url:
2933 ext_or_url = f'file.{ext_or_url}'
2934 return mimetypes.guess_type(ext_or_url)[0]
2937 def parse_codecs(codecs_str):
2938 # http://tools.ietf.org/html/rfc6381
2941 split_codecs = list(filter(None, map(
2942 str.strip, codecs_str.strip().strip(',').split(','))))
2943 vcodec, acodec, scodec, hdr = None, None, None, None
2944 for full_codec in split_codecs:
2945 parts = re.sub(r'0+(?=\d)', '', full_codec).split('.')
2946 if parts[0] in ('avc1', 'avc2', 'avc3', 'avc4', 'vp9', 'vp8', 'hev1', 'hev2',
2947 'h263', 'h264', 'mp4v', 'hvc1', 'av1', 'theora', 'dvh1', 'dvhe'):
2951 if parts[0] in ('dvh1', 'dvhe'):
2953 elif parts[0] == 'av1' and traversal.traverse_obj(parts, 3) == '10':
2955 elif parts[:2] == ['vp9', '2']:
2957 elif parts[0] in ('flac', 'mp4a', 'opus', 'vorbis', 'mp3', 'aac', 'ac-4',
2958 'ac-3', 'ec-3', 'eac3', 'dtsc', 'dtse', 'dtsh', 'dtsl'):
2959 acodec = acodec or full_codec
2960 elif parts[0] in ('stpp', 'wvtt'):
2961 scodec = scodec or full_codec
2963 write_string(f'WARNING: Unknown codec {full_codec}\n')
2964 if vcodec or acodec or scodec:
2966 'vcodec': vcodec or 'none',
2967 'acodec': acodec or 'none',
2968 'dynamic_range': hdr,
2969 **({'scodec': scodec} if scodec is not None else {}),
2971 elif len(split_codecs) == 2:
2973 'vcodec': split_codecs[0],
2974 'acodec': split_codecs[1],
2979 def get_compatible_ext(*, vcodecs, acodecs, vexts, aexts, preferences=None):
2980 assert len(vcodecs) == len(vexts) and len(acodecs) == len(aexts)
2982 allow_mkv = not preferences or 'mkv' in preferences
2984 if allow_mkv and max(len(acodecs), len(vcodecs)) > 1:
2985 return 'mkv' # TODO: any other format allows this?
2987 # TODO: All codecs supported by parse_codecs isn't handled here
2988 COMPATIBLE_CODECS = {
2990 'av1', 'hevc', 'avc1', 'mp4a', 'ac-4', # fourcc (m3u8, mpd)
2991 'h264', 'aacl', 'ec-3', # Set in ISM
2994 'av1', 'vp9', 'vp8', 'opus', 'vrbs',
2995 'vp9x', 'vp8x', # in the webm spec
2999 sanitize_codec = functools.partial(
3000 try_get, getter=lambda x: x[0].split('.')[0].replace('0', '').lower())
3001 vcodec, acodec = sanitize_codec(vcodecs), sanitize_codec(acodecs)
3003 for ext in preferences or COMPATIBLE_CODECS.keys():
3004 codec_set = COMPATIBLE_CODECS.get(ext, set())
3005 if ext == 'mkv' or codec_set.issuperset((vcodec, acodec)):
3009 {'mp3', 'mp4', 'm4a', 'm4p', 'm4b', 'm4r', 'm4v', 'ismv', 'isma', 'mov'},
3012 for ext in preferences or vexts:
3013 current_exts = {ext, *vexts, *aexts}
3014 if ext == 'mkv' or current_exts == {ext} or any(
3015 ext_sets.issuperset(current_exts) for ext_sets in COMPATIBLE_EXTS):
3017 return 'mkv' if allow_mkv else preferences[-1]
3020 def urlhandle_detect_ext(url_handle, default=NO_DEFAULT):
3021 getheader = url_handle.headers.get
3023 cd = getheader('Content-Disposition')
3025 m = re.match(r'attachment;\s*filename="(?P
<filename
>[^
"]+)"', cd)
3027 e = determine_ext(m.group('filename
'), default_ext=None)
3031 meta_ext = getheader('x
-amz
-meta
-name
')
3033 e = meta_ext.rpartition('.')[2]
3037 return mimetype2ext(getheader('Content
-Type
'), default=default)
3040 def encode_data_uri(data, mime_type):
3041 return 'data
:%s;base64
,%s' % (mime_type, base64.b64encode(data).decode('ascii
'))
3044 def age_restricted(content_limit, age_limit):
3045 """ Returns True iff the content should be blocked """
3047 if age_limit is None: # No limit set
3049 if content_limit is None:
3050 return False # Content available for everyone
3051 return age_limit < content_limit
3054 # List of known byte-order-marks (BOM)
3056 (b'\xef\xbb\xbf', 'utf
-8'),
3057 (b'\x00\x00\xfe\xff', 'utf
-32-be
'),
3058 (b'\xff\xfe\x00\x00', 'utf
-32-le
'),
3059 (b'\xff\xfe', 'utf
-16-le
'),
3060 (b'\xfe\xff', 'utf
-16-be
'),
3064 def is_html(first_bytes):
3065 """ Detect whether a file contains HTML by examining its first bytes. """
3068 for bom, enc in BOMS:
3069 while first_bytes.startswith(bom):
3070 encoding, first_bytes = enc, first_bytes[len(bom):]
3072 return re.match(r'^\s
*<', first_bytes.decode(encoding, 'replace
'))
3075 def determine_protocol(info_dict):
3076 protocol = info_dict.get('protocol
')
3077 if protocol is not None:
3080 url = sanitize_url(info_dict['url
'])
3081 if url.startswith('rtmp
'):
3083 elif url.startswith('mms
'):
3085 elif url.startswith('rtsp
'):
3088 ext = determine_ext(url)
3090 return 'm3u8
' if info_dict.get('is_live
') else 'm3u8_native
'
3094 return urllib.parse.urlparse(url).scheme
3097 def render_table(header_row, data, delim=False, extra_gap=0, hide_empty=False):
3098 """ Render a list of rows, each as a list of values.
3099 Text after a \t will be right aligned """
3101 return len(remove_terminal_sequences(string).replace('\t', ''))
3103 def get_max_lens(table):
3104 return [max(width(str(v)) for v in col) for col in zip(*table)]
3106 def filter_using_list(row, filterArray):
3107 return [col for take, col in itertools.zip_longest(filterArray, row, fillvalue=True) if take]
3109 max_lens = get_max_lens(data) if hide_empty else []
3110 header_row = filter_using_list(header_row, max_lens)
3111 data = [filter_using_list(row, max_lens) for row in data]
3113 table = [header_row] + data
3114 max_lens = get_max_lens(table)
3117 table = [header_row, [delim * (ml + extra_gap) for ml in max_lens]] + data
3118 table[1][-1] = table[1][-1][:-extra_gap * len(delim)] # Remove extra_gap from end of delimiter
3120 for pos, text in enumerate(map(str, row)):
3122 row[pos] = text.replace('\t', ' ' * (max_lens[pos] - width(text))) + ' ' * extra_gap
3124 row[pos] = text + ' ' * (max_lens[pos] - width(text) + extra_gap)
3125 ret = '\n'.join(''.join(row).rstrip() for row in table)
3129 def _match_one(filter_part, dct, incomplete):
3130 # TODO: Generalize code with YoutubeDL._build_format_filter
3131 STRING_OPERATORS = {
3132 '*=': operator.contains,
3133 '^
=': lambda attr, value: attr.startswith(value),
3134 '$
=': lambda attr, value: attr.endswith(value),
3135 '~
=': lambda attr, value: re.search(value, attr),
3137 COMPARISON_OPERATORS = {
3139 '<=': operator.le, # "<=" must be defined above "<"
3146 if isinstance(incomplete, bool):
3147 is_incomplete = lambda _: incomplete
3149 is_incomplete = lambda k: k in incomplete
3151 operator_rex = re.compile(r'''(?x)
3153 \s*(?P<negation>!\s*)?(?P<op>%s)(?P<none_inclusive>\s*\?)?\s*
3155 (?P<quote>["\'])(?P<quotedstrval>.+?)(?P=quote)|
3158 ''' % '|
'.join(map(re.escape, COMPARISON_OPERATORS.keys())))
3159 m = operator_rex.fullmatch(filter_part.strip())
3162 unnegated_op = COMPARISON_OPERATORS[m['op
']]
3164 op = lambda attr, value: not unnegated_op(attr, value)
3167 comparison_value = m['quotedstrval
'] or m['strval
'] or m['intval
']
3169 comparison_value = comparison_value.replace(r'\
%s' % m['quote
'], m['quote
'])
3170 actual_value = dct.get(m['key
'])
3171 numeric_comparison = None
3172 if isinstance(actual_value, (int, float)):
3173 # If the original field is a string and matching comparisonvalue is
3174 # a number we should respect the origin of the original field
3175 # and process comparison value as a string (see
3176 # https://github.com/ytdl-org/youtube-dl/issues/11082)
3178 numeric_comparison = int(comparison_value)
3180 numeric_comparison = parse_filesize(comparison_value)
3181 if numeric_comparison is None:
3182 numeric_comparison = parse_filesize(f'{comparison_value}B
')
3183 if numeric_comparison is None:
3184 numeric_comparison = parse_duration(comparison_value)
3185 if numeric_comparison is not None and m['op
'] in STRING_OPERATORS:
3186 raise ValueError('Operator
%s only supports string values
!' % m['op
'])
3187 if actual_value is None:
3188 return is_incomplete(m['key
']) or m['none_inclusive
']
3189 return op(actual_value, comparison_value if numeric_comparison is None else numeric_comparison)
3192 '': lambda v: (v is True) if isinstance(v, bool) else (v is not None),
3193 '!': lambda v: (v is False) if isinstance(v, bool) else (v is None),
3195 operator_rex = re.compile(r'''(?x)
3196 (?P<op>%s)\s*(?P<key>[a-z_]+)
3197 ''' % '|
'.join(map(re.escape, UNARY_OPERATORS.keys())))
3198 m = operator_rex.fullmatch(filter_part.strip())
3200 op = UNARY_OPERATORS[m.group('op
')]
3201 actual_value = dct.get(m.group('key
'))
3202 if is_incomplete(m.group('key
')) and actual_value is None:
3204 return op(actual_value)
3206 raise ValueError('Invalid
filter part
%r' % filter_part)
3209 def match_str(filter_str, dct, incomplete=False):
3210 """ Filter a dictionary with a simple string syntax.
3211 @returns Whether the filter passes
3212 @param incomplete Set of keys that is expected to be missing from dct.
3213 Can be True/False to indicate all/none of the keys may be missing.
3214 All conditions on incomplete keys pass if the key is missing
3217 _match_one(filter_part.replace(r'\
&', '&'), dct, incomplete)
3218 for filter_part in re.split(r'(?
<!\\)&', filter_str))
3221 def match_filter_func(filters, breaking_filters=None):
3222 if not filters and not breaking_filters:
3224 breaking_filters = match_filter_func(breaking_filters) or (lambda _, __: None)
3225 filters = set(variadic(filters or []))
3227 interactive = '-' in filters
3231 def _match_func(info_dict, incomplete=False):
3232 ret = breaking_filters(info_dict, incomplete)
3234 raise RejectedVideoReached(ret)
3236 if not filters or any(match_str(f, info_dict, incomplete) for f in filters):
3237 return NO_DEFAULT if interactive and not incomplete else None
3239 video_title = info_dict.get('title
') or info_dict.get('id') or 'entry
'
3240 filter_str = ') |
('.join(map(str.strip, filters))
3241 return f'{video_title} does
not pass filter ({filter_str}
), skipping
..'
3245 class download_range_func:
3246 def __init__(self, chapters, ranges, from_info=False):
3247 self.chapters, self.ranges, self.from_info = chapters, ranges, from_info
3249 def __call__(self, info_dict, ydl):
3251 warning = ('There are no chapters matching the regex
' if info_dict.get('chapters
')
3252 else 'Cannot match chapters since chapter information
is unavailable
')
3253 for regex in self.chapters or []:
3254 for i, chapter in enumerate(info_dict.get('chapters
') or []):
3255 if re.search(regex, chapter['title
']):
3257 yield {**chapter, 'index': i}
3258 if self.chapters and warning:
3259 ydl.to_screen(f'[info
] {info_dict["id"]}
: {warning}
')
3261 for start, end in self.ranges or []:
3263 'start_time
': self._handle_negative_timestamp(start, info_dict),
3264 'end_time
': self._handle_negative_timestamp(end, info_dict),
3267 if self.from_info and (info_dict.get('start_time
') or info_dict.get('end_time
')):
3269 'start_time
': info_dict.get('start_time
') or 0,
3270 'end_time
': info_dict.get('end_time
') or float('inf
'),
3272 elif not self.ranges and not self.chapters:
3276 def _handle_negative_timestamp(time, info):
3277 return max(info['duration
'] + time, 0) if info.get('duration
') and time < 0 else time
3279 def __eq__(self, other):
3280 return (isinstance(other, download_range_func)
3281 and self.chapters == other.chapters and self.ranges == other.ranges)
3284 return f'{__name__}
.{type(self).__name__}
({self.chapters}
, {self.ranges}
)'
3287 def parse_dfxp_time_expr(time_expr):
3291 mobj = re.match(rf'^
(?P
<time_offset
>{NUMBER_RE}
)s?$
', time_expr)
3293 return float(mobj.group('time_offset
'))
3295 mobj = re.match(r'^
(\d
+):(\d\d
):(\d\
d(?
:(?
:\
.|
:)\d
+)?
)$
', time_expr)
3297 return 3600 * int(mobj.group(1)) + 60 * int(mobj.group(2)) + float(mobj.group(3).replace(':', '.'))
3300 def srt_subtitles_timecode(seconds):
3301 return '%02d
:%02d
:%02d
,%03d
' % timetuple_from_msec(seconds * 1000)
3304 def ass_subtitles_timecode(seconds):
3305 time = timetuple_from_msec(seconds * 1000)
3306 return '%01d
:%02d
:%02d
.%02d
' % (*time[:-1], time.milliseconds / 10)
3309 def dfxp2srt(dfxp_data):
3311 @param dfxp_data A bytes-like object containing DFXP data
3312 @returns A unicode object containing converted SRT data
3314 LEGACY_NAMESPACES = (
3315 (b'http
://www
.w3
.org
/ns
/ttml
', [
3316 b'http
://www
.w3
.org
/2004/11/ttaf1
',
3317 b'http
://www
.w3
.org
/2006/04/ttaf1
',
3318 b'http
://www
.w3
.org
/2006/10/ttaf1
',
3320 (b'http
://www
.w3
.org
/ns
/ttml
#styling', [
3321 b
'http://www.w3.org/ns/ttml#style',
3325 SUPPORTED_STYLING
= [
3334 _x
= functools
.partial(xpath_with_ns
, ns_map
={
3335 'xml': 'http://www.w3.org/XML/1998/namespace',
3336 'ttml': 'http://www.w3.org/ns/ttml',
3337 'tts': 'http://www.w3.org/ns/ttml#styling',
3343 class TTMLPElementParser
:
3345 _unclosed_elements
= []
3346 _applied_styles
= []
3348 def start(self
, tag
, attrib
):
3349 if tag
in (_x('ttml:br'), 'br'):
3352 unclosed_elements
= []
3354 element_style_id
= attrib
.get('style')
3356 style
.update(default_style
)
3357 if element_style_id
:
3358 style
.update(styles
.get(element_style_id
, {}))
3359 for prop
in SUPPORTED_STYLING
:
3360 prop_val
= attrib
.get(_x('tts:' + prop
))
3362 style
[prop
] = prop_val
3365 for k
, v
in sorted(style
.items()):
3366 if self
._applied
_styles
and self
._applied
_styles
[-1].get(k
) == v
:
3369 font
+= ' color="%s"' % v
3370 elif k
== 'fontSize':
3371 font
+= ' size="%s"' % v
3372 elif k
== 'fontFamily':
3373 font
+= ' face="%s"' % v
3374 elif k
== 'fontWeight' and v
== 'bold':
3376 unclosed_elements
.append('b')
3377 elif k
== 'fontStyle' and v
== 'italic':
3379 unclosed_elements
.append('i')
3380 elif k
== 'textDecoration' and v
== 'underline':
3382 unclosed_elements
.append('u')
3384 self
._out
+= '<font' + font
+ '>'
3385 unclosed_elements
.append('font')
3387 if self
._applied
_styles
:
3388 applied_style
.update(self
._applied
_styles
[-1])
3389 applied_style
.update(style
)
3390 self
._applied
_styles
.append(applied_style
)
3391 self
._unclosed
_elements
.append(unclosed_elements
)
3394 if tag
not in (_x('ttml:br'), 'br'):
3395 unclosed_elements
= self
._unclosed
_elements
.pop()
3396 for element
in reversed(unclosed_elements
):
3397 self
._out
+= '</%s>' % element
3398 if unclosed_elements
and self
._applied
_styles
:
3399 self
._applied
_styles
.pop()
3401 def data(self
, data
):
3405 return self
._out
.strip()
3407 # Fix UTF-8 encoded file wrongly marked as UTF-16. See https://github.com/yt-dlp/yt-dlp/issues/6543#issuecomment-1477169870
3408 # This will not trigger false positives since only UTF-8 text is being replaced
3409 dfxp_data
= dfxp_data
.replace(b
'encoding=\'UTF-16\'', b
'encoding=\'UTF-8\'')
3411 def parse_node(node
):
3412 target
= TTMLPElementParser()
3413 parser
= xml
.etree
.ElementTree
.XMLParser(target
=target
)
3414 parser
.feed(xml
.etree
.ElementTree
.tostring(node
))
3415 return parser
.close()
3417 for k
, v
in LEGACY_NAMESPACES
:
3419 dfxp_data
= dfxp_data
.replace(ns
, k
)
3421 dfxp
= compat_etree_fromstring(dfxp_data
)
3423 paras
= dfxp
.findall(_x('.//ttml:p')) or dfxp
.findall('.//p')
3426 raise ValueError('Invalid dfxp/TTML subtitle')
3430 for style
in dfxp
.findall(_x('.//ttml:style')):
3431 style_id
= style
.get('id') or style
.get(_x('xml:id'))
3434 parent_style_id
= style
.get('style')
3436 if parent_style_id
not in styles
:
3439 styles
[style_id
] = styles
[parent_style_id
].copy()
3440 for prop
in SUPPORTED_STYLING
:
3441 prop_val
= style
.get(_x('tts:' + prop
))
3443 styles
.setdefault(style_id
, {})[prop
] = prop_val
3449 for p
in ('body', 'div'):
3450 ele
= xpath_element(dfxp
, [_x('.//ttml:' + p
), './/' + p
])
3453 style
= styles
.get(ele
.get('style'))
3456 default_style
.update(style
)
3458 for para
, index
in zip(paras
, itertools
.count(1)):
3459 begin_time
= parse_dfxp_time_expr(para
.attrib
.get('begin'))
3460 end_time
= parse_dfxp_time_expr(para
.attrib
.get('end'))
3461 dur
= parse_dfxp_time_expr(para
.attrib
.get('dur'))
3462 if begin_time
is None:
3467 end_time
= begin_time
+ dur
3468 out
.append('%d\n%s --> %s\n%s\n\n' % (
3470 srt_subtitles_timecode(begin_time
),
3471 srt_subtitles_timecode(end_time
),
3477 def cli_option(params
, command_option
, param
, separator
=None):
3478 param
= params
.get(param
)
3479 return ([] if param
is None
3480 else [command_option
, str(param
)] if separator
is None
3481 else [f
'{command_option}{separator}{param}'])
3484 def cli_bool_option(params
, command_option
, param
, true_value
='true', false_value
='false', separator
=None):
3485 param
= params
.get(param
)
3486 assert param
in (True, False, None)
3487 return cli_option({True: true_value, False: false_value}
, command_option
, param
, separator
)
3490 def cli_valueless_option(params
, command_option
, param
, expected_value
=True):
3491 return [command_option
] if params
.get(param
) == expected_value
else []
3494 def cli_configuration_args(argdict
, keys
, default
=[], use_compat
=True):
3495 if isinstance(argdict
, (list, tuple)): # for backward compatibility
3502 assert isinstance(argdict
, dict)
3504 assert isinstance(keys
, (list, tuple))
3505 for key_list
in keys
:
3506 arg_list
= list(filter(
3507 lambda x
: x
is not None,
3508 [argdict
.get(key
.lower()) for key
in variadic(key_list
)]))
3510 return [arg
for args
in arg_list
for arg
in args
]
3514 def _configuration_args(main_key
, argdict
, exe
, keys
=None, default
=[], use_compat
=True):
3515 main_key
, exe
= main_key
.lower(), exe
.lower()
3516 root_key
= exe
if main_key
== exe
else f
'{main_key}+{exe}'
3517 keys
= [f
'{root_key}{k}' for k
in (keys
or [''])]
3518 if root_key
in keys
:
3520 keys
.append((main_key
, exe
))
3521 keys
.append('default')
3524 return cli_configuration_args(argdict
, keys
, default
, use_compat
)
3528 # See http://www.loc.gov/standards/iso639-2/ISO-639-2_utf-8.txt
3587 'iw': 'heb', # Replaced by he in 1989 revision
3597 'in': 'ind', # Replaced by id in 1989 revision
3713 'ji': 'yid', # Replaced by yi in 1989 revision
3721 def short2long(cls
, code
):
3722 """Convert language code from ISO 639-1 to ISO 639-2/T"""
3723 return cls
._lang
_map
.get(code
[:2])
3726 def long2short(cls
, code
):
3727 """Convert language code from ISO 639-2/T to ISO 639-1"""
3728 for short_name
, long_name
in cls
._lang
_map
.items():
3729 if long_name
== code
:
3734 # From http://data.okfn.org/data/core/country-list
3736 'AF': 'Afghanistan',
3737 'AX': 'Åland Islands',
3740 'AS': 'American Samoa',
3745 'AG': 'Antigua and Barbuda',
3762 'BO': 'Bolivia, Plurinational State of',
3763 'BQ': 'Bonaire, Sint Eustatius and Saba',
3764 'BA': 'Bosnia and Herzegovina',
3766 'BV': 'Bouvet Island',
3768 'IO': 'British Indian Ocean Territory',
3769 'BN': 'Brunei Darussalam',
3771 'BF': 'Burkina Faso',
3777 'KY': 'Cayman Islands',
3778 'CF': 'Central African Republic',
3782 'CX': 'Christmas Island',
3783 'CC': 'Cocos (Keeling) Islands',
3787 'CD': 'Congo, the Democratic Republic of the',
3788 'CK': 'Cook Islands',
3790 'CI': 'Côte d\'Ivoire',
3795 'CZ': 'Czech Republic',
3799 'DO': 'Dominican Republic',
3802 'SV': 'El Salvador',
3803 'GQ': 'Equatorial Guinea',
3807 'FK': 'Falkland Islands (Malvinas)',
3808 'FO': 'Faroe Islands',
3812 'GF': 'French Guiana',
3813 'PF': 'French Polynesia',
3814 'TF': 'French Southern Territories',
3829 'GW': 'Guinea-Bissau',
3832 'HM': 'Heard Island and McDonald Islands',
3833 'VA': 'Holy See (Vatican City State)',
3840 'IR': 'Iran, Islamic Republic of',
3843 'IM': 'Isle of Man',
3853 'KP': 'Korea, Democratic People\'s Republic of',
3854 'KR': 'Korea, Republic of',
3857 'LA': 'Lao People\'s Democratic Republic',
3863 'LI': 'Liechtenstein',
3867 'MK': 'Macedonia, the Former Yugoslav Republic of',
3874 'MH': 'Marshall Islands',
3880 'FM': 'Micronesia, Federated States of',
3881 'MD': 'Moldova, Republic of',
3892 'NL': 'Netherlands',
3893 'NC': 'New Caledonia',
3894 'NZ': 'New Zealand',
3899 'NF': 'Norfolk Island',
3900 'MP': 'Northern Mariana Islands',
3905 'PS': 'Palestine, State of',
3907 'PG': 'Papua New Guinea',
3910 'PH': 'Philippines',
3914 'PR': 'Puerto Rico',
3918 'RU': 'Russian Federation',
3920 'BL': 'Saint Barthélemy',
3921 'SH': 'Saint Helena, Ascension and Tristan da Cunha',
3922 'KN': 'Saint Kitts and Nevis',
3923 'LC': 'Saint Lucia',
3924 'MF': 'Saint Martin (French part)',
3925 'PM': 'Saint Pierre and Miquelon',
3926 'VC': 'Saint Vincent and the Grenadines',
3929 'ST': 'Sao Tome and Principe',
3930 'SA': 'Saudi Arabia',
3934 'SL': 'Sierra Leone',
3936 'SX': 'Sint Maarten (Dutch part)',
3939 'SB': 'Solomon Islands',
3941 'ZA': 'South Africa',
3942 'GS': 'South Georgia and the South Sandwich Islands',
3943 'SS': 'South Sudan',
3948 'SJ': 'Svalbard and Jan Mayen',
3951 'CH': 'Switzerland',
3952 'SY': 'Syrian Arab Republic',
3953 'TW': 'Taiwan, Province of China',
3955 'TZ': 'Tanzania, United Republic of',
3957 'TL': 'Timor-Leste',
3961 'TT': 'Trinidad and Tobago',
3964 'TM': 'Turkmenistan',
3965 'TC': 'Turks and Caicos Islands',
3969 'AE': 'United Arab Emirates',
3970 'GB': 'United Kingdom',
3971 'US': 'United States',
3972 'UM': 'United States Minor Outlying Islands',
3976 'VE': 'Venezuela, Bolivarian Republic of',
3978 'VG': 'Virgin Islands, British',
3979 'VI': 'Virgin Islands, U.S.',
3980 'WF': 'Wallis and Futuna',
3981 'EH': 'Western Sahara',
3985 # Not ISO 3166 codes, but used for IP blocks
3986 'AP': 'Asia/Pacific Region',
3991 def short2full(cls
, code
):
3992 """Convert an ISO 3166-2 country code to the corresponding full name"""
3993 return cls
._country
_map
.get(code
.upper())
3997 # Major IPv4 address blocks per country
3999 'AD': '46.172.224.0/19',
4000 'AE': '94.200.0.0/13',
4001 'AF': '149.54.0.0/17',
4002 'AG': '209.59.64.0/18',
4003 'AI': '204.14.248.0/21',
4004 'AL': '46.99.0.0/16',
4005 'AM': '46.70.0.0/15',
4006 'AO': '105.168.0.0/13',
4007 'AP': '182.50.184.0/21',
4008 'AQ': '23.154.160.0/24',
4009 'AR': '181.0.0.0/12',
4010 'AS': '202.70.112.0/20',
4011 'AT': '77.116.0.0/14',
4012 'AU': '1.128.0.0/11',
4013 'AW': '181.41.0.0/18',
4014 'AX': '185.217.4.0/22',
4015 'AZ': '5.197.0.0/16',
4016 'BA': '31.176.128.0/17',
4017 'BB': '65.48.128.0/17',
4018 'BD': '114.130.0.0/16',
4020 'BF': '102.178.0.0/15',
4021 'BG': '95.42.0.0/15',
4022 'BH': '37.131.0.0/17',
4023 'BI': '154.117.192.0/18',
4024 'BJ': '137.255.0.0/16',
4025 'BL': '185.212.72.0/23',
4026 'BM': '196.12.64.0/18',
4027 'BN': '156.31.0.0/16',
4028 'BO': '161.56.0.0/16',
4029 'BQ': '161.0.80.0/20',
4030 'BR': '191.128.0.0/12',
4031 'BS': '24.51.64.0/18',
4032 'BT': '119.2.96.0/19',
4033 'BW': '168.167.0.0/16',
4034 'BY': '178.120.0.0/13',
4035 'BZ': '179.42.192.0/18',
4036 'CA': '99.224.0.0/11',
4037 'CD': '41.243.0.0/16',
4038 'CF': '197.242.176.0/21',
4039 'CG': '160.113.0.0/16',
4040 'CH': '85.0.0.0/13',
4041 'CI': '102.136.0.0/14',
4042 'CK': '202.65.32.0/19',
4043 'CL': '152.172.0.0/14',
4044 'CM': '102.244.0.0/14',
4045 'CN': '36.128.0.0/10',
4046 'CO': '181.240.0.0/12',
4047 'CR': '201.192.0.0/12',
4048 'CU': '152.206.0.0/15',
4049 'CV': '165.90.96.0/19',
4050 'CW': '190.88.128.0/17',
4051 'CY': '31.153.0.0/16',
4052 'CZ': '88.100.0.0/14',
4054 'DJ': '197.241.0.0/17',
4055 'DK': '87.48.0.0/12',
4056 'DM': '192.243.48.0/20',
4057 'DO': '152.166.0.0/15',
4058 'DZ': '41.96.0.0/12',
4059 'EC': '186.68.0.0/15',
4060 'EE': '90.190.0.0/15',
4061 'EG': '156.160.0.0/11',
4062 'ER': '196.200.96.0/20',
4063 'ES': '88.0.0.0/11',
4064 'ET': '196.188.0.0/14',
4065 'EU': '2.16.0.0/13',
4066 'FI': '91.152.0.0/13',
4067 'FJ': '144.120.0.0/16',
4068 'FK': '80.73.208.0/21',
4069 'FM': '119.252.112.0/20',
4070 'FO': '88.85.32.0/19',
4072 'GA': '41.158.0.0/15',
4074 'GD': '74.122.88.0/21',
4075 'GE': '31.146.0.0/16',
4076 'GF': '161.22.64.0/18',
4077 'GG': '62.68.160.0/19',
4078 'GH': '154.160.0.0/12',
4079 'GI': '95.164.0.0/16',
4080 'GL': '88.83.0.0/19',
4081 'GM': '160.182.0.0/15',
4082 'GN': '197.149.192.0/18',
4083 'GP': '104.250.0.0/19',
4084 'GQ': '105.235.224.0/20',
4085 'GR': '94.64.0.0/13',
4086 'GT': '168.234.0.0/16',
4087 'GU': '168.123.0.0/16',
4088 'GW': '197.214.80.0/20',
4089 'GY': '181.41.64.0/18',
4090 'HK': '113.252.0.0/14',
4091 'HN': '181.210.0.0/16',
4092 'HR': '93.136.0.0/13',
4093 'HT': '148.102.128.0/17',
4094 'HU': '84.0.0.0/14',
4095 'ID': '39.192.0.0/10',
4096 'IE': '87.32.0.0/12',
4097 'IL': '79.176.0.0/13',
4098 'IM': '5.62.80.0/20',
4099 'IN': '117.192.0.0/10',
4100 'IO': '203.83.48.0/21',
4101 'IQ': '37.236.0.0/14',
4102 'IR': '2.176.0.0/12',
4103 'IS': '82.221.0.0/16',
4104 'IT': '79.0.0.0/10',
4105 'JE': '87.244.64.0/18',
4106 'JM': '72.27.0.0/17',
4107 'JO': '176.29.0.0/16',
4108 'JP': '133.0.0.0/8',
4109 'KE': '105.48.0.0/12',
4110 'KG': '158.181.128.0/17',
4111 'KH': '36.37.128.0/17',
4112 'KI': '103.25.140.0/22',
4113 'KM': '197.255.224.0/20',
4114 'KN': '198.167.192.0/19',
4115 'KP': '175.45.176.0/22',
4116 'KR': '175.192.0.0/10',
4117 'KW': '37.36.0.0/14',
4118 'KY': '64.96.0.0/15',
4119 'KZ': '2.72.0.0/13',
4120 'LA': '115.84.64.0/18',
4121 'LB': '178.135.0.0/16',
4122 'LC': '24.92.144.0/20',
4123 'LI': '82.117.0.0/19',
4124 'LK': '112.134.0.0/15',
4125 'LR': '102.183.0.0/16',
4126 'LS': '129.232.0.0/17',
4127 'LT': '78.56.0.0/13',
4128 'LU': '188.42.0.0/16',
4129 'LV': '46.109.0.0/16',
4130 'LY': '41.252.0.0/14',
4131 'MA': '105.128.0.0/11',
4132 'MC': '88.209.64.0/18',
4133 'MD': '37.246.0.0/16',
4134 'ME': '178.175.0.0/17',
4135 'MF': '74.112.232.0/21',
4136 'MG': '154.126.0.0/17',
4137 'MH': '117.103.88.0/21',
4138 'MK': '77.28.0.0/15',
4139 'ML': '154.118.128.0/18',
4140 'MM': '37.111.0.0/17',
4141 'MN': '49.0.128.0/17',
4142 'MO': '60.246.0.0/16',
4143 'MP': '202.88.64.0/20',
4144 'MQ': '109.203.224.0/19',
4145 'MR': '41.188.64.0/18',
4146 'MS': '208.90.112.0/22',
4147 'MT': '46.11.0.0/16',
4148 'MU': '105.16.0.0/12',
4149 'MV': '27.114.128.0/18',
4150 'MW': '102.70.0.0/15',
4151 'MX': '187.192.0.0/11',
4152 'MY': '175.136.0.0/13',
4153 'MZ': '197.218.0.0/15',
4154 'NA': '41.182.0.0/16',
4155 'NC': '101.101.0.0/18',
4156 'NE': '197.214.0.0/18',
4157 'NF': '203.17.240.0/22',
4158 'NG': '105.112.0.0/12',
4159 'NI': '186.76.0.0/15',
4160 'NL': '145.96.0.0/11',
4161 'NO': '84.208.0.0/13',
4162 'NP': '36.252.0.0/15',
4163 'NR': '203.98.224.0/19',
4164 'NU': '49.156.48.0/22',
4165 'NZ': '49.224.0.0/14',
4166 'OM': '5.36.0.0/15',
4167 'PA': '186.72.0.0/15',
4168 'PE': '186.160.0.0/14',
4169 'PF': '123.50.64.0/18',
4170 'PG': '124.240.192.0/19',
4171 'PH': '49.144.0.0/13',
4172 'PK': '39.32.0.0/11',
4173 'PL': '83.0.0.0/11',
4174 'PM': '70.36.0.0/20',
4175 'PR': '66.50.0.0/16',
4176 'PS': '188.161.0.0/16',
4177 'PT': '85.240.0.0/13',
4178 'PW': '202.124.224.0/20',
4179 'PY': '181.120.0.0/14',
4180 'QA': '37.210.0.0/15',
4181 'RE': '102.35.0.0/16',
4182 'RO': '79.112.0.0/13',
4183 'RS': '93.86.0.0/15',
4184 'RU': '5.136.0.0/13',
4185 'RW': '41.186.0.0/16',
4186 'SA': '188.48.0.0/13',
4187 'SB': '202.1.160.0/19',
4188 'SC': '154.192.0.0/11',
4189 'SD': '102.120.0.0/13',
4190 'SE': '78.64.0.0/12',
4191 'SG': '8.128.0.0/10',
4192 'SI': '188.196.0.0/14',
4193 'SK': '78.98.0.0/15',
4194 'SL': '102.143.0.0/17',
4195 'SM': '89.186.32.0/19',
4196 'SN': '41.82.0.0/15',
4197 'SO': '154.115.192.0/18',
4198 'SR': '186.179.128.0/17',
4199 'SS': '105.235.208.0/21',
4200 'ST': '197.159.160.0/19',
4201 'SV': '168.243.0.0/16',
4202 'SX': '190.102.0.0/20',
4204 'SZ': '41.84.224.0/19',
4205 'TC': '65.255.48.0/20',
4206 'TD': '154.68.128.0/19',
4207 'TG': '196.168.0.0/14',
4208 'TH': '171.96.0.0/13',
4209 'TJ': '85.9.128.0/18',
4210 'TK': '27.96.24.0/21',
4211 'TL': '180.189.160.0/20',
4212 'TM': '95.85.96.0/19',
4213 'TN': '197.0.0.0/11',
4214 'TO': '175.176.144.0/21',
4215 'TR': '78.160.0.0/11',
4216 'TT': '186.44.0.0/15',
4217 'TV': '202.2.96.0/19',
4218 'TW': '120.96.0.0/11',
4219 'TZ': '156.156.0.0/14',
4220 'UA': '37.52.0.0/14',
4221 'UG': '102.80.0.0/13',
4223 'UY': '167.56.0.0/13',
4224 'UZ': '84.54.64.0/18',
4225 'VA': '212.77.0.0/19',
4226 'VC': '207.191.240.0/21',
4227 'VE': '186.88.0.0/13',
4228 'VG': '66.81.192.0/20',
4229 'VI': '146.226.0.0/16',
4230 'VN': '14.160.0.0/11',
4231 'VU': '202.80.32.0/20',
4232 'WF': '117.20.32.0/21',
4233 'WS': '202.4.32.0/19',
4234 'YE': '134.35.0.0/16',
4235 'YT': '41.242.116.0/22',
4236 'ZA': '41.0.0.0/11',
4237 'ZM': '102.144.0.0/13',
4238 'ZW': '102.177.192.0/18',
4242 def random_ipv4(cls
, code_or_block
):
4243 if len(code_or_block
) == 2:
4244 block
= cls
._country
_ip
_map
.get(code_or_block
.upper())
4248 block
= code_or_block
4249 addr
, preflen
= block
.split('/')
4250 addr_min
= struct
.unpack('!L', socket
.inet_aton(addr
))[0]
4251 addr_max
= addr_min |
(0xffffffff >> int(preflen
))
4252 return str(socket
.inet_ntoa(
4253 struct
.pack('!L', random
.randint(addr_min
, addr_max
))))
4256 # Both long_to_bytes and bytes_to_long are adapted from PyCrypto, which is
4257 # released into Public Domain
4258 # https://github.com/dlitz/pycrypto/blob/master/lib/Crypto/Util/number.py#L387
4260 def long_to_bytes(n
, blocksize
=0):
4261 """long_to_bytes(n:long, blocksize:int) : string
4262 Convert a long integer to a byte string.
4264 If optional blocksize is given and greater than zero, pad the front of the
4265 byte string with binary zeros so that the length is a multiple of
4268 # after much testing, this algorithm was deemed to be the fastest
4272 s
= struct
.pack('>I', n
& 0xffffffff) + s
4274 # strip off leading zeros
4275 for i
in range(len(s
)):
4276 if s
[i
] != b
'\000'[0]:
4279 # only happens when n == 0
4283 # add back some pad bytes. this could be done more efficiently w.r.t. the
4284 # de-padding being done above, but sigh...
4285 if blocksize
> 0 and len(s
) % blocksize
:
4286 s
= (blocksize
- len(s
) % blocksize
) * b
'\000' + s
4290 def bytes_to_long(s
):
4291 """bytes_to_long(string) : long
4292 Convert a byte string to a long integer.
4294 This is (essentially) the inverse of long_to_bytes().
4299 extra
= (4 - length
% 4)
4300 s
= b
'\000' * extra
+ s
4301 length
= length
+ extra
4302 for i
in range(0, length
, 4):
4303 acc
= (acc
<< 32) + struct
.unpack('>I', s
[i
:i
+ 4])[0]
4307 def ohdave_rsa_encrypt(data
, exponent
, modulus
):
4309 Implement OHDave's RSA algorithm. See http://www.ohdave.com/rsa/
4312 data: data to encrypt, bytes-like object
4313 exponent, modulus: parameter e and N of RSA algorithm, both integer
4314 Output: hex string of encrypted data
4316 Limitation: supports one block encryption only
4319 payload
= int(binascii
.hexlify(data
[::-1]), 16)
4320 encrypted
= pow(payload
, exponent
, modulus
)
4321 return '%x' % encrypted
4324 def pkcs1pad(data
, length
):
4326 Padding input data with PKCS#1 scheme
4328 @param {int[]} data input data
4329 @param {int} length target length
4330 @returns {int[]} padded data
4332 if len(data
) > length
- 11:
4333 raise ValueError('Input data too long for PKCS#1 padding')
4335 pseudo_random
= [random
.randint(0, 254) for _
in range(length
- len(data
) - 3)]
4336 return [0, 2] + pseudo_random
+ [0] + data
4339 def _base_n_table(n
, table
):
4340 if not table
and not n
:
4341 raise ValueError('Either table or n must be specified')
4342 table
= (table
or '0123456789abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ')[:n
]
4344 if n
and n
!= len(table
):
4345 raise ValueError(f
'base {n} exceeds table length {len(table)}')
4349 def encode_base_n(num
, n
=None, table
=None):
4350 """Convert given int to a base-n string"""
4351 table
= _base_n_table(n
, table
)
4355 result
, base
= '', len(table
)
4357 result
= table
[num
% base
] + result
4362 def decode_base_n(string
, n
=None, table
=None):
4363 """Convert given base-n string to int"""
4364 table
= {char: index for index, char in enumerate(_base_n_table(n, table))}
4365 result
, base
= 0, len(table
)
4367 result
= result
* base
+ table
[char
]
4371 def decode_packed_codes(code
):
4372 mobj
= re
.search(PACKED_CODES_RE
, code
)
4373 obfuscated_code
, base
, count
, symbols
= mobj
.groups()
4376 symbols
= symbols
.split('|')
4381 base_n_count
= encode_base_n(count
, base
)
4382 symbol_table
[base_n_count
] = symbols
[count
] or base_n_count
4385 r
'\b(\w+)\b', lambda mobj
: symbol_table
[mobj
.group(0)],
4389 def caesar(s
, alphabet
, shift
):
4394 alphabet
[(alphabet
.index(c
) + shift
) % l
] if c
in alphabet
else c
4399 return caesar(s
, r
'''!"#$%&'()*+,-./0123456789:;<=>?@ABCDEFGHIJKLMNOPQRSTUVWXYZ[\]^_`abcdefghijklmnopqrstuvwxyz{|}~''', 47)
4402 def parse_m3u8_attributes(attrib
):
4404 for (key
, val
) in re
.findall(r
'(?P<key>[A-Z0-9-]+)=(?P<val>"[^"]+"|[^",]+)(?:,|$)', attrib
):
4405 if val
.startswith('"'):
4411 def urshift(val
, n
):
4412 return val
>> n
if val
>= 0 else (val
+ 0x100000000) >> n
4415 def write_xattr(path
, key
, value
):
4416 # Windows: Write xattrs to NTFS Alternate Data Streams:
4417 # http://en.wikipedia.org/wiki/NTFS#Alternate_data_streams_.28ADS.29
4418 if compat_os_name
== 'nt':
4419 assert ':' not in key
4420 assert os
.path
.exists(path
)
4423 with open(f
'{path}:{key}', 'wb') as f
:
4425 except OSError as e
:
4426 raise XAttrMetadataError(e
.errno
, e
.strerror
)
4429 # UNIX Method 1. Use xattrs/pyxattrs modules
4432 if getattr(xattr
, '_yt_dlp__identifier', None) == 'pyxattr':
4433 # Unicode arguments are not supported in pyxattr until version 0.5.0
4434 # See https://github.com/ytdl-org/youtube-dl/issues/5498
4435 if version_tuple(xattr
.__version
__) >= (0, 5, 0):
4436 setxattr
= xattr
.set
4438 setxattr
= xattr
.setxattr
4442 setxattr(path
, key
, value
)
4443 except OSError as e
:
4444 raise XAttrMetadataError(e
.errno
, e
.strerror
)
4447 # UNIX Method 2. Use setfattr/xattr executables
4448 exe
= ('setfattr' if check_executable('setfattr', ['--version'])
4449 else 'xattr' if check_executable('xattr', ['-h']) else None)
4451 raise XAttrUnavailableError(
4452 'Couldn\'t find a tool to set the xattrs. Install either the python "xattr" or "pyxattr" modules or the '
4453 + ('"xattr" binary' if sys
.platform
!= 'linux' else 'GNU "attr" package (which contains the "setfattr" tool)'))
4455 value
= value
.decode()
4457 _
, stderr
, returncode
= Popen
.run(
4458 [exe
, '-w', key
, value
, path
] if exe
== 'xattr' else [exe
, '-n', key
, '-v', value
, path
],
4459 text
=True, stdout
=subprocess
.PIPE
, stderr
=subprocess
.PIPE
, stdin
=subprocess
.PIPE
)
4460 except OSError as e
:
4461 raise XAttrMetadataError(e
.errno
, e
.strerror
)
4463 raise XAttrMetadataError(returncode
, stderr
)
4466 def random_birthday(year_field
, month_field
, day_field
):
4467 start_date
= datetime
.date(1950, 1, 1)
4468 end_date
= datetime
.date(1995, 12, 31)
4469 offset
= random
.randint(0, (end_date
- start_date
).days
)
4470 random_date
= start_date
+ datetime
.timedelta(offset
)
4472 year_field
: str(random_date
.year
),
4473 month_field
: str(random_date
.month
),
4474 day_field
: str(random_date
.day
),
4478 def find_available_port(interface
=''):
4480 with socket
.socket() as sock
:
4481 sock
.bind((interface
, 0))
4482 return sock
.getsockname()[1]
4487 # Templates for internet shortcut files, which are plain text files.
4488 DOT_URL_LINK_TEMPLATE
= '''\
4493 DOT_WEBLOC_LINK_TEMPLATE
= '''\
4494 <?xml version="1.0" encoding="UTF-8"?>
4495 <!DOCTYPE plist PUBLIC "-//Apple//DTD PLIST 1.0//EN" "http://www.apple.com/DTDs/PropertyList-1.0.dtd">
4496 <plist version="1.0">
4499 \t<string>%(url)s</string>
4504 DOT_DESKTOP_LINK_TEMPLATE
= '''\
4514 'url': DOT_URL_LINK_TEMPLATE
,
4515 'desktop': DOT_DESKTOP_LINK_TEMPLATE
,
4516 'webloc': DOT_WEBLOC_LINK_TEMPLATE
,
4520 def iri_to_uri(iri
):
4522 Converts an IRI (Internationalized Resource Identifier, allowing Unicode characters) to a URI (Uniform Resource Identifier, ASCII-only).
4524 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.
4527 iri_parts
= urllib
.parse
.urlparse(iri
)
4529 if '[' in iri_parts
.netloc
:
4530 raise ValueError('IPv6 URIs are not, yet, supported.')
4531 # Querying `.netloc`, when there's only one bracket, also raises a ValueError.
4533 # 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.
4536 if iri_parts
.username
:
4537 net_location
+= urllib
.parse
.quote(iri_parts
.username
, safe
=r
"!$%&'()*+,~")
4538 if iri_parts
.password
is not None:
4539 net_location
+= ':' + urllib
.parse
.quote(iri_parts
.password
, safe
=r
"!$%&'()*+,~")
4542 net_location
+= iri_parts
.hostname
.encode('idna').decode() # Punycode for Unicode hostnames.
4543 # The 'idna' encoding produces ASCII text.
4544 if iri_parts
.port
is not None and iri_parts
.port
!= 80:
4545 net_location
+= ':' + str(iri_parts
.port
)
4547 return urllib
.parse
.urlunparse(
4551 urllib
.parse
.quote_plus(iri_parts
.path
, safe
=r
"!$%&'()*+,/:;=@|~"),
4553 # Unsure about the `safe` argument, since this is a legacy way of handling parameters.
4554 urllib
.parse
.quote_plus(iri_parts
.params
, safe
=r
"!$%&'()*+,/:;=@|~"),
4556 # Not totally sure about the `safe` argument, since the source does not explicitly mention the query URI component.
4557 urllib
.parse
.quote_plus(iri_parts
.query
, safe
=r
"!$%&'()*+,/:;=?@{|}~"),
4559 urllib
.parse
.quote_plus(iri_parts
.fragment
, safe
=r
"!#$%&'()*+,/:;=?@{|}~")))
4561 # Source for `safe` arguments: https://url.spec.whatwg.org/#percent-encoded-bytes.
4564 def to_high_limit_path(path
):
4565 if sys
.platform
in ['win32', 'cygwin']:
4566 # Work around MAX_PATH limitation on Windows. The maximum allowed length for the individual path segments may still be quite limited.
4567 return '\\\\?\\' + os
.path
.abspath(path
)
4572 def format_field(obj
, field
=None, template
='%s', ignore
=NO_DEFAULT
, default
='', func
=IDENTITY
):
4573 val
= traversal
.traverse_obj(obj
, *variadic(field
))
4574 if not val
if ignore
is NO_DEFAULT
else val
in variadic(ignore
):
4576 return template
% func(val
)
4579 def clean_podcast_url(url
):
4580 url
= re
.sub(r
'''(?x)
4584 media\.blubrry\.com| # https://create.blubrry.com/resources/podcast-media-download-statistics/getting-started/
4589 (?:dts|www)\.podtrac\.com/(?:pts/)?redirect\.[0-9a-z]{3,4}| # http://analytics.podtrac.com/how-to-measure
4592 cn\.co| # https://podcorn.com/analytics-prefix/
4593 st\.fm # https://podsights.com/docs/
4598 return re
.sub(r
'^\w+://(\w+://)', r
'\1', url
)
4601 _HEX_TABLE
= '0123456789abcdef'
4604 def random_uuidv4():
4605 return re
.sub(r
'[xy]', lambda x
: _HEX_TABLE
[random
.randint(0, 15)], 'xxxxxxxx-xxxx-4xxx-yxxx-xxxxxxxxxxxx')
4608 def make_dir(path
, to_screen
=None):
4610 dn
= os
.path
.dirname(path
)
4612 os
.makedirs(dn
, exist_ok
=True)
4614 except OSError as err
:
4615 if callable(to_screen
) is not None:
4616 to_screen(f
'unable to create directory {err}')
4620 def get_executable_path():
4621 from ..update
import _get_variant_and_executable_path
4623 return os
.path
.dirname(os
.path
.abspath(_get_variant_and_executable_path()[1]))
4626 def get_user_config_dirs(package_name
):
4627 # .config (e.g. ~/.config/package_name)
4628 xdg_config_home
= os
.getenv('XDG_CONFIG_HOME') or compat_expanduser('~/.config')
4629 yield os
.path
.join(xdg_config_home
, package_name
)
4631 # appdata (%APPDATA%/package_name)
4632 appdata_dir
= os
.getenv('appdata')
4634 yield os
.path
.join(appdata_dir
, package_name
)
4636 # home (~/.package_name)
4637 yield os
.path
.join(compat_expanduser('~'), f
'.{package_name}')
4640 def get_system_config_dirs(package_name
):
4642 yield os
.path
.join('/etc', package_name
)
4645 def time_seconds(**kwargs
):
4647 Returns TZ-aware time in seconds since the epoch (1970-01-01T00:00:00Z)
4649 return time
.time() + datetime
.timedelta(**kwargs
).total_seconds()
4652 # create a JSON Web Signature (jws) with HS256 algorithm
4653 # the resulting format is in JWS Compact Serialization
4654 # implemented following JWT https://www.rfc-editor.org/rfc/rfc7519.html
4655 # implemented following JWS https://www.rfc-editor.org/rfc/rfc7515.html
4656 def jwt_encode_hs256(payload_data
, key
, headers
={}):
4662 header_data
.update(headers
)
4663 header_b64
= base64
.b64encode(json
.dumps(header_data
).encode())
4664 payload_b64
= base64
.b64encode(json
.dumps(payload_data
).encode())
4665 h
= hmac
.new(key
.encode(), header_b64
+ b
'.' + payload_b64
, hashlib
.sha256
)
4666 signature_b64
= base64
.b64encode(h
.digest())
4667 token
= header_b64
+ b
'.' + payload_b64
+ b
'.' + signature_b64
4671 # can be extended in future to verify the signature and parse header and return the algorithm used if it's not HS256
4672 def jwt_decode_hs256(jwt
):
4673 header_b64
, payload_b64
, signature_b64
= jwt
.split('.')
4674 # add trailing ='s that may have been stripped, superfluous ='s are ignored
4675 payload_data
= json
.loads(base64
.urlsafe_b64decode(f
'{payload_b64}==='))
4679 WINDOWS_VT_MODE
= False if compat_os_name
== 'nt' else None
4683 def supports_terminal_sequences(stream
):
4684 if compat_os_name
== 'nt':
4685 if not WINDOWS_VT_MODE
:
4687 elif not os
.getenv('TERM'):
4690 return stream
.isatty()
4691 except BaseException
:
4695 def windows_enable_vt_mode():
4696 """Ref: https://bugs.python.org/issue30075 """
4697 if get_windows_version() < (10, 0, 10586):
4701 import ctypes
.wintypes
4704 ENABLE_VIRTUAL_TERMINAL_PROCESSING
= 0x0004
4706 dll
= ctypes
.WinDLL('kernel32', use_last_error
=False)
4707 handle
= os
.open('CONOUT$', os
.O_RDWR
)
4709 h_out
= ctypes
.wintypes
.HANDLE(msvcrt
.get_osfhandle(handle
))
4710 dw_original_mode
= ctypes
.wintypes
.DWORD()
4711 success
= dll
.GetConsoleMode(h_out
, ctypes
.byref(dw_original_mode
))
4713 raise Exception('GetConsoleMode failed')
4715 success
= dll
.SetConsoleMode(h_out
, ctypes
.wintypes
.DWORD(
4716 dw_original_mode
.value | ENABLE_VIRTUAL_TERMINAL_PROCESSING
))
4718 raise Exception('SetConsoleMode failed')
4722 global WINDOWS_VT_MODE
4723 WINDOWS_VT_MODE
= True
4724 supports_terminal_sequences
.cache_clear()
4727 _terminal_sequences_re
= re
.compile('\033\\[[^m]+m')
4730 def remove_terminal_sequences(string
):
4731 return _terminal_sequences_re
.sub('', string
)
4734 def number_of_digits(number
):
4735 return len('%d' % number
)
4738 def join_nonempty(*values
, delim
='-', from_dict
=None):
4739 if from_dict
is not None:
4740 values
= (traversal
.traverse_obj(from_dict
, variadic(v
)) for v
in values
)
4741 return delim
.join(map(str, filter(None, values
)))
4744 def scale_thumbnails_to_max_format_width(formats
, thumbnails
, url_width_re
):
4746 Find the largest format dimensions in terms of video width and, for each thumbnail:
4747 * Modify the URL: Match the width with the provided regex and replace with the former width
4750 This function is useful with video services that scale the provided thumbnails on demand
4752 _keys
= ('width', 'height')
4753 max_dimensions
= max(
4754 (tuple(format
.get(k
) or 0 for k
in _keys
) for format
in formats
),
4756 if not max_dimensions
[0]:
4760 {'url': re.sub(url_width_re, str(max_dimensions[0]), thumbnail['url'])}
,
4761 dict(zip(_keys
, max_dimensions
)), thumbnail
)
4762 for thumbnail
in thumbnails
4766 def parse_http_range(range):
4767 """ Parse value of "Range" or "Content-Range" HTTP header into tuple. """
4769 return None, None, None
4770 crg
= re
.search(r
'bytes[ =](\d+)-(\d+)?(?:/(\d+))?', range)
4772 return None, None, None
4773 return int(crg
.group(1)), int_or_none(crg
.group(2)), int_or_none(crg
.group(3))
4776 def read_stdin(what
):
4777 eof
= 'Ctrl+Z' if compat_os_name
== 'nt' else 'Ctrl+D'
4778 write_string(f
'Reading {what} from STDIN - EOF ({eof}) to end:\n')
4782 def determine_file_encoding(data
):
4784 Detect the text encoding used
4785 @returns (encoding, bytes to skip)
4788 # BOM marks are given priority over declarations
4789 for bom
, enc
in BOMS
:
4790 if data
.startswith(bom
):
4791 return enc
, len(bom
)
4793 # Strip off all null bytes to match even when UTF-16 or UTF-32 is used.
4794 # We ignore the endianness to get a good enough match
4795 data
= data
.replace(b
'\0', b
'')
4796 mobj
= re
.match(rb
'(?m)^#\s*coding\s*:\s*(\S+)\s*$', data
)
4797 return mobj
.group(1).decode() if mobj
else None, 0
4804 __initialized
= False
4806 def __init__(self
, parser
, label
=None):
4807 self
.parser
, self
.label
= parser
, label
4808 self
._loaded
_paths
, self
.configs
= set(), []
4810 def init(self
, args
=None, filename
=None):
4811 assert not self
.__initialized
4812 self
.own_args
, self
.filename
= args
, filename
4813 return self
.load_configs()
4815 def load_configs(self
):
4818 location
= os
.path
.realpath(self
.filename
)
4819 directory
= os
.path
.dirname(location
)
4820 if location
in self
._loaded
_paths
:
4822 self
._loaded
_paths
.add(location
)
4824 self
.__initialized
= True
4825 opts
, _
= self
.parser
.parse_known_args(self
.own_args
)
4826 self
.parsed_args
= self
.own_args
4827 for location
in opts
.config_locations
or []:
4829 if location
in self
._loaded
_paths
:
4831 self
._loaded
_paths
.add(location
)
4832 self
.append_config(shlex
.split(read_stdin('options'), comments
=True), label
='stdin')
4834 location
= os
.path
.join(directory
, expand_path(location
))
4835 if os
.path
.isdir(location
):
4836 location
= os
.path
.join(location
, 'yt-dlp.conf')
4837 if not os
.path
.exists(location
):
4838 self
.parser
.error(f
'config location {location} does not exist')
4839 self
.append_config(self
.read_file(location
), location
)
4843 label
= join_nonempty(
4844 self
.label
, 'config', f
'"{self.filename}"' if self
.filename
else '',
4846 return join_nonempty(
4847 self
.own_args
is not None and f
'{label[0].upper()}{label[1:]}: {self.hide_login_info(self.own_args)}',
4848 *(f
'\n{c}'.replace('\n', '\n| ')[1:] for c
in self
.configs
),
4852 def read_file(filename
, default
=[]):
4854 optionf
= open(filename
, 'rb')
4856 return default
# silently skip if file is not present
4858 enc
, skip
= determine_file_encoding(optionf
.read(512))
4859 optionf
.seek(skip
, io
.SEEK_SET
)
4861 enc
= None # silently skip read errors
4863 # FIXME: https://github.com/ytdl-org/youtube-dl/commit/dfe5fa49aed02cf36ba9f743b11b0903554b5e56
4864 contents
= optionf
.read().decode(enc
or preferredencoding())
4865 res
= shlex
.split(contents
, comments
=True)
4866 except Exception as err
:
4867 raise ValueError(f
'Unable to parse "{filename}": {err}')
4873 def hide_login_info(opts
):
4874 PRIVATE_OPTS
= {'-p', '--password', '-u', '--username', '--video-password', '--ap-password', '--ap-username'}
4875 eqre
= re
.compile('^(?P<key>' + ('|'.join(re
.escape(po
) for po
in PRIVATE_OPTS
)) + ')=.+$')
4880 return m
.group('key') + '=PRIVATE'
4884 opts
= list(map(_scrub_eq
, opts
))
4885 for idx
, opt
in enumerate(opts
):
4886 if opt
in PRIVATE_OPTS
and idx
+ 1 < len(opts
):
4887 opts
[idx
+ 1] = 'PRIVATE'
4890 def append_config(self
, *args
, label
=None):
4891 config
= type(self
)(self
.parser
, label
)
4892 config
._loaded
_paths
= self
._loaded
_paths
4893 if config
.init(*args
):
4894 self
.configs
.append(config
)
4898 for config
in reversed(self
.configs
):
4899 yield from config
.all_args
4900 yield from self
.parsed_args
or []
4902 def parse_known_args(self
, **kwargs
):
4903 return self
.parser
.parse_known_args(self
.all_args
, **kwargs
)
4905 def parse_args(self
):
4906 return self
.parser
.parse_args(self
.all_args
)
4909 class WebSocketsWrapper
:
4910 """Wraps websockets module to use in non-async scopes"""
4913 def __init__(self
, url
, headers
=None, connect
=True):
4914 self
.loop
= asyncio
.new_event_loop()
4915 # XXX: "loop" is deprecated
4916 self
.conn
= websockets
.connect(
4917 url
, extra_headers
=headers
, ping_interval
=None,
4918 close_timeout
=float('inf'), loop
=self
.loop
, ping_timeout
=float('inf'))
4921 atexit
.register(self
.__exit
__, None, None, None)
4923 def __enter__(self
):
4925 self
.pool
= self
.run_with_loop(self
.conn
.__aenter
__(), self
.loop
)
4928 def send(self
, *args
):
4929 self
.run_with_loop(self
.pool
.send(*args
), self
.loop
)
4931 def recv(self
, *args
):
4932 return self
.run_with_loop(self
.pool
.recv(*args
), self
.loop
)
4934 def __exit__(self
, type, value
, traceback
):
4936 return self
.run_with_loop(self
.conn
.__aexit
__(type, value
, traceback
), self
.loop
)
4939 self
._cancel
_all
_tasks
(self
.loop
)
4941 # taken from https://github.com/python/cpython/blob/3.9/Lib/asyncio/runners.py with modifications
4942 # for contributors: If there's any new library using asyncio needs to be run in non-async, move these function out of this class
4944 def run_with_loop(main
, loop
):
4945 if not asyncio
.iscoroutine(main
):
4946 raise ValueError(f
'a coroutine was expected, got {main!r}')
4949 return loop
.run_until_complete(main
)
4951 loop
.run_until_complete(loop
.shutdown_asyncgens())
4952 if hasattr(loop
, 'shutdown_default_executor'):
4953 loop
.run_until_complete(loop
.shutdown_default_executor())
4956 def _cancel_all_tasks(loop
):
4957 to_cancel
= asyncio
.all_tasks(loop
)
4962 for task
in to_cancel
:
4965 # XXX: "loop" is removed in python 3.10+
4966 loop
.run_until_complete(
4967 asyncio
.gather(*to_cancel
, loop
=loop
, return_exceptions
=True))
4969 for task
in to_cancel
:
4970 if task
.cancelled():
4972 if task
.exception() is not None:
4973 loop
.call_exception_handler({
4974 'message': 'unhandled exception during asyncio.run() shutdown',
4975 'exception': task
.exception(),
4980 def merge_headers(*dicts
):
4981 """Merge dicts of http headers case insensitively, prioritizing the latter ones"""
4982 return {k.title(): v for k, v in itertools.chain.from_iterable(map(dict.items, dicts))}
4985 def cached_method(f
):
4986 """Cache a method"""
4987 signature
= inspect
.signature(f
)
4990 def wrapper(self
, *args
, **kwargs
):
4991 bound_args
= signature
.bind(self
, *args
, **kwargs
)
4992 bound_args
.apply_defaults()
4993 key
= tuple(bound_args
.arguments
.values())[1:]
4995 cache
= vars(self
).setdefault('_cached_method__cache', {}).setdefault(f.__name__, {}
)
4996 if key
not in cache
:
4997 cache
[key
] = f(self
, *args
, **kwargs
)
5002 class classproperty
:
5003 """property access for class methods with optional caching"""
5004 def __new__(cls
, func
=None, *args
, **kwargs
):
5006 return functools
.partial(cls
, *args
, **kwargs
)
5007 return super().__new
__(cls
)
5009 def __init__(self
, func
, *, cache
=False):
5010 functools
.update_wrapper(self
, func
)
5012 self
._cache
= {} if cache
else None
5014 def __get__(self
, _
, cls
):
5015 if self
._cache
is None:
5016 return self
.func(cls
)
5017 elif cls
not in self
._cache
:
5018 self
._cache
[cls
] = self
.func(cls
)
5019 return self
._cache
[cls
]
5022 class function_with_repr
:
5023 def __init__(self
, func
, repr_
=None):
5024 functools
.update_wrapper(self
, func
)
5025 self
.func
, self
.__repr
= func
, repr_
5027 def __call__(self
, *args
, **kwargs
):
5028 return self
.func(*args
, **kwargs
)
5033 return f
'{self.func.__module__}.{self.func.__qualname__}'
5036 class Namespace(types
.SimpleNamespace
):
5037 """Immutable namespace"""
5040 return iter(self
.__dict
__.values())
5044 return self
.__dict
__.items()
5047 MEDIA_EXTENSIONS
= Namespace(
5048 common_video
=('avi', 'flv', 'mkv', 'mov', 'mp4', 'webm'),
5049 video
=('3g2', '3gp', 'f4v', 'mk3d', 'divx', 'mpg', 'ogv', 'm4v', 'wmv'),
5050 common_audio
=('aiff', 'alac', 'flac', 'm4a', 'mka', 'mp3', 'ogg', 'opus', 'wav'),
5051 audio
=('aac', 'ape', 'asf', 'f4a', 'f4b', 'm4b', 'm4p', 'm4r', 'oga', 'ogx', 'spx', 'vorbis', 'wma', 'weba'),
5052 thumbnails
=('jpg', 'png', 'webp'),
5053 storyboards
=('mhtml', ),
5054 subtitles
=('srt', 'vtt', 'ass', 'lrc'),
5055 manifests
=('f4f', 'f4m', 'm3u8', 'smil', 'mpd'),
5057 MEDIA_EXTENSIONS
.video
+= MEDIA_EXTENSIONS
.common_video
5058 MEDIA_EXTENSIONS
.audio
+= MEDIA_EXTENSIONS
.common_audio
5060 KNOWN_EXTENSIONS
= (*MEDIA_EXTENSIONS
.video
, *MEDIA_EXTENSIONS
.audio
, *MEDIA_EXTENSIONS
.manifests
)
5065 for retry in RetryManager(...):
5068 except SomeException as err:
5072 attempt
, _error
= 0, None
5074 def __init__(self
, _retries
, _error_callback
, **kwargs
):
5075 self
.retries
= _retries
or 0
5076 self
.error_callback
= functools
.partial(_error_callback
, **kwargs
)
5078 def _should_retry(self
):
5079 return self
._error
is not NO_DEFAULT
and self
.attempt
<= self
.retries
5083 if self
._error
is NO_DEFAULT
:
5088 def error(self
, value
):
5092 while self
._should
_retry
():
5093 self
.error
= NO_DEFAULT
5097 self
.error_callback(self
.error
, self
.attempt
, self
.retries
)
5100 def report_retry(e
, count
, retries
, *, sleep_func
, info
, warn
, error
=None, suffix
=None):
5101 """Utility function for reporting retries"""
5104 return error(f
'{e}. Giving up after {count - 1} retries') if count
> 1 else error(str(e
))
5109 elif isinstance(e
, ExtractorError
):
5110 e
= remove_end(str_or_none(e
.cause
) or e
.orig_msg
, '.')
5111 warn(f
'{e}. Retrying{format_field(suffix, None, " %s")} ({count}/{retries})...')
5113 delay
= float_or_none(sleep_func(n
=count
- 1)) if callable(sleep_func
) else sleep_func
5115 info(f
'Sleeping {delay:.2f} seconds ...')
5119 def make_archive_id(ie
, video_id
):
5120 ie_key
= ie
if isinstance(ie
, str) else ie
.ie_key()
5121 return f
'{ie_key.lower()} {video_id}'
5124 def truncate_string(s
, left
, right
=0):
5125 assert left
> 3 and right
>= 0
5126 if s
is None or len(s
) <= left
+ right
:
5128 return f
'{s[:left-3]}...{s[-right:] if right else ""}'
5131 def orderedSet_from_options(options
, alias_dict
, *, use_regex
=False, start
=None):
5132 assert 'all' in alias_dict
, '"all" alias is required'
5133 requested
= list(start
or [])
5135 discard
= val
.startswith('-')
5139 if val
in alias_dict
:
5140 val
= alias_dict
[val
] if not discard
else [
5141 i
[1:] if i
.startswith('-') else f
'-{i}' for i
in alias_dict
[val
]]
5142 # NB: Do not allow regex in aliases for performance
5143 requested
= orderedSet_from_options(val
, alias_dict
, start
=requested
)
5146 current
= (filter(re
.compile(val
, re
.I
).fullmatch
, alias_dict
['all']) if use_regex
5147 else [val
] if val
in alias_dict
['all'] else None)
5149 raise ValueError(val
)
5152 for item
in current
:
5153 while item
in requested
:
5154 requested
.remove(item
)
5156 requested
.extend(current
)
5158 return orderedSet(requested
)
5163 regex
= r
' *((?P<reverse>\+)?(?P<field>[a-zA-Z0-9_]+)((?P<separator>[~:])(?P<limit>.*?))?)? *$'
5165 default
= ('hidden', 'aud_or_vid', 'hasvid', 'ie_pref', 'lang', 'quality',
5166 'res', 'fps', 'hdr:12', 'vcodec:vp9.2', 'channels', 'acodec',
5167 'size', 'br', 'asr', 'proto', 'ext', 'hasaud', 'source', 'id') # These must not be aliases
5168 ytdl_default
= ('hasaud', 'lang', 'quality', 'tbr', 'filesize', 'vbr',
5169 'height', 'width', 'proto', 'vext', 'abr', 'aext',
5170 'fps', 'fs_approx', 'source', 'id')
5173 'vcodec': {'type': 'ordered', 'regex': True,
5174 'order': ['av0?1', 'vp0?9.2', 'vp0?9', '[hx]265|he?vc?', '[hx]264|avc', 'vp0?8', 'mp4v|h263', 'theora', '', None, 'none']},
5175 'acodec': {'type': 'ordered', 'regex': True,
5176 'order': ['[af]lac', 'wav|aiff', 'opus', 'vorbis|ogg', 'aac', 'mp?4a?', 'mp3', 'ac-?4', 'e-?a?c-?3', 'ac-?3', 'dts', '', None, 'none']},
5177 'hdr': {'type': 'ordered', 'regex': True, 'field': 'dynamic_range',
5178 'order': ['dv', '(hdr)?12', r
'(hdr)?10\+', '(hdr)?10', 'hlg', '', 'sdr', None]},
5179 'proto': {'type': 'ordered', 'regex': True, 'field': 'protocol',
5180 'order': ['(ht|f)tps', '(ht|f)tp$', 'm3u8.*', '.*dash', 'websocket_frag', 'rtmpe?', '', 'mms|rtsp', 'ws|websocket', 'f4']},
5181 'vext': {'type': 'ordered', 'field': 'video_ext',
5182 'order': ('mp4', 'mov', 'webm', 'flv', '', 'none'),
5183 'order_free': ('webm', 'mp4', 'mov', 'flv', '', 'none')},
5184 'aext': {'type': 'ordered', 'regex': True, 'field': 'audio_ext',
5185 'order': ('m4a', 'aac', 'mp3', 'ogg', 'opus', 'web[am]', '', 'none'),
5186 'order_free': ('ogg', 'opus', 'web[am]', 'mp3', 'm4a', 'aac', '', 'none')},
5187 'hidden': {'visible': False, 'forced': True, 'type': 'extractor', 'max': -1000}
,
5188 'aud_or_vid': {'visible': False, 'forced': True, 'type': 'multiple',
5189 'field': ('vcodec', 'acodec'),
5190 'function': lambda it
: int(any(v
!= 'none' for v
in it
))},
5191 'ie_pref': {'priority': True, 'type': 'extractor'}
,
5192 'hasvid': {'priority': True, 'field': 'vcodec', 'type': 'boolean', 'not_in_list': ('none',)}
,
5193 'hasaud': {'field': 'acodec', 'type': 'boolean', 'not_in_list': ('none',)}
,
5194 'lang': {'convert': 'float', 'field': 'language_preference', 'default': -1}
,
5195 'quality': {'convert': 'float', 'default': -1}
,
5196 'filesize': {'convert': 'bytes'}
,
5197 'fs_approx': {'convert': 'bytes', 'field': 'filesize_approx'}
,
5198 'id': {'convert': 'string', 'field': 'format_id'}
,
5199 'height': {'convert': 'float_none'}
,
5200 'width': {'convert': 'float_none'}
,
5201 'fps': {'convert': 'float_none'}
,
5202 'channels': {'convert': 'float_none', 'field': 'audio_channels'}
,
5203 'tbr': {'convert': 'float_none'}
,
5204 'vbr': {'convert': 'float_none'}
,
5205 'abr': {'convert': 'float_none'}
,
5206 'asr': {'convert': 'float_none'}
,
5207 'source': {'convert': 'float', 'field': 'source_preference', 'default': -1}
,
5209 'codec': {'type': 'combined', 'field': ('vcodec', 'acodec')}
,
5210 'br': {'type': 'multiple', 'field': ('tbr', 'vbr', 'abr'), 'convert': 'float_none',
5211 'function': lambda it
: next(filter(None, it
), None)},
5212 'size': {'type': 'multiple', 'field': ('filesize', 'fs_approx'), 'convert': 'bytes',
5213 'function': lambda it
: next(filter(None, it
), None)},
5214 'ext': {'type': 'combined', 'field': ('vext', 'aext')}
,
5215 'res': {'type': 'multiple', 'field': ('height', 'width'),
5216 'function': lambda it
: (lambda l
: min(l
) if l
else 0)(tuple(filter(None, it
)))},
5218 # Actual field names
5219 'format_id': {'type': 'alias', 'field': 'id'}
,
5220 'preference': {'type': 'alias', 'field': 'ie_pref'}
,
5221 'language_preference': {'type': 'alias', 'field': 'lang'}
,
5222 'source_preference': {'type': 'alias', 'field': 'source'}
,
5223 'protocol': {'type': 'alias', 'field': 'proto'}
,
5224 'filesize_approx': {'type': 'alias', 'field': 'fs_approx'}
,
5225 'audio_channels': {'type': 'alias', 'field': 'channels'}
,
5228 'dimension': {'type': 'alias', 'field': 'res', 'deprecated': True}
,
5229 'resolution': {'type': 'alias', 'field': 'res', 'deprecated': True}
,
5230 'extension': {'type': 'alias', 'field': 'ext', 'deprecated': True}
,
5231 'bitrate': {'type': 'alias', 'field': 'br', 'deprecated': True}
,
5232 'total_bitrate': {'type': 'alias', 'field': 'tbr', 'deprecated': True}
,
5233 'video_bitrate': {'type': 'alias', 'field': 'vbr', 'deprecated': True}
,
5234 'audio_bitrate': {'type': 'alias', 'field': 'abr', 'deprecated': True}
,
5235 'framerate': {'type': 'alias', 'field': 'fps', 'deprecated': True}
,
5236 'filesize_estimate': {'type': 'alias', 'field': 'size', 'deprecated': True}
,
5237 'samplerate': {'type': 'alias', 'field': 'asr', 'deprecated': True}
,
5238 'video_ext': {'type': 'alias', 'field': 'vext', 'deprecated': True}
,
5239 'audio_ext': {'type': 'alias', 'field': 'aext', 'deprecated': True}
,
5240 'video_codec': {'type': 'alias', 'field': 'vcodec', 'deprecated': True}
,
5241 'audio_codec': {'type': 'alias', 'field': 'acodec', 'deprecated': True}
,
5242 'video': {'type': 'alias', 'field': 'hasvid', 'deprecated': True}
,
5243 'has_video': {'type': 'alias', 'field': 'hasvid', 'deprecated': True}
,
5244 'audio': {'type': 'alias', 'field': 'hasaud', 'deprecated': True}
,
5245 'has_audio': {'type': 'alias', 'field': 'hasaud', 'deprecated': True}
,
5246 'extractor': {'type': 'alias', 'field': 'ie_pref', 'deprecated': True}
,
5247 'extractor_preference': {'type': 'alias', 'field': 'ie_pref', 'deprecated': True}
,
5250 def __init__(self
, ydl
, field_preference
):
5253 self
.evaluate_params(self
.ydl
.params
, field_preference
)
5254 if ydl
.params
.get('verbose'):
5255 self
.print_verbose_info(self
.ydl
.write_debug
)
5257 def _get_field_setting(self
, field
, key
):
5258 if field
not in self
.settings
:
5259 if key
in ('forced', 'priority'):
5261 self
.ydl
.deprecated_feature(f
'Using arbitrary fields ({field}) for format sorting is '
5262 'deprecated and may be removed in a future version')
5263 self
.settings
[field
] = {}
5264 propObj
= self
.settings
[field
]
5265 if key
not in propObj
:
5266 type = propObj
.get('type')
5268 default
= 'preference' if type == 'extractor' else (field
,) if type in ('combined', 'multiple') else field
5269 elif key
== 'convert':
5270 default
= 'order' if type == 'ordered' else 'float_string' if field
else 'ignore'
5272 default
= {'type': 'field', 'visible': True, 'order': [], 'not_in_list': (None,)}
.get(key
, None)
5273 propObj
[key
] = default
5276 def _resolve_field_value(self
, field
, value
, convertNone
=False):
5281 value
= value
.lower()
5282 conversion
= self
._get
_field
_setting
(field
, 'convert')
5283 if conversion
== 'ignore':
5285 if conversion
== 'string':
5287 elif conversion
== 'float_none':
5288 return float_or_none(value
)
5289 elif conversion
== 'bytes':
5290 return parse_bytes(value
)
5291 elif conversion
== 'order':
5292 order_list
= (self
._use
_free
_order
and self
._get
_field
_setting
(field
, 'order_free')) or self
._get
_field
_setting
(field
, 'order')
5293 use_regex
= self
._get
_field
_setting
(field
, 'regex')
5294 list_length
= len(order_list
)
5295 empty_pos
= order_list
.index('') if '' in order_list
else list_length
+ 1
5296 if use_regex
and value
is not None:
5297 for i
, regex
in enumerate(order_list
):
5298 if regex
and re
.match(regex
, value
):
5299 return list_length
- i
5300 return list_length
- empty_pos
# not in list
5301 else: # not regex or value = None
5302 return list_length
- (order_list
.index(value
) if value
in order_list
else empty_pos
)
5304 if value
.isnumeric():
5307 self
.settings
[field
]['convert'] = 'string'
5310 def evaluate_params(self
, params
, sort_extractor
):
5311 self
._use
_free
_order
= params
.get('prefer_free_formats', False)
5312 self
._sort
_user
= params
.get('format_sort', [])
5313 self
._sort
_extractor
= sort_extractor
5315 def add_item(field
, reverse
, closest
, limit_text
):
5316 field
= field
.lower()
5317 if field
in self
._order
:
5319 self
._order
.append(field
)
5320 limit
= self
._resolve
_field
_value
(field
, limit_text
)
5323 'closest': False if limit
is None else closest
,
5324 'limit_text': limit_text
,
5326 if field
in self
.settings
:
5327 self
.settings
[field
].update(data
)
5329 self
.settings
[field
] = data
5332 tuple(field
for field
in self
.default
if self
._get
_field
_setting
(field
, 'forced'))
5333 + (tuple() if params
.get('format_sort_force', False)
5334 else tuple(field
for field
in self
.default
if self
._get
_field
_setting
(field
, 'priority')))
5335 + tuple(self
._sort
_user
) + tuple(sort_extractor
) + self
.default
)
5337 for item
in sort_list
:
5338 match
= re
.match(self
.regex
, item
)
5340 raise ExtractorError('Invalid format sort string "%s" given by extractor' % item
)
5341 field
= match
.group('field')
5344 if self
._get
_field
_setting
(field
, 'type') == 'alias':
5345 alias
, field
= field
, self
._get
_field
_setting
(field
, 'field')
5346 if self
._get
_field
_setting
(alias
, 'deprecated'):
5347 self
.ydl
.deprecated_feature(f
'Format sorting alias {alias} is deprecated and may '
5348 f
'be removed in a future version. Please use {field} instead')
5349 reverse
= match
.group('reverse') is not None
5350 closest
= match
.group('separator') == '~'
5351 limit_text
= match
.group('limit')
5353 has_limit
= limit_text
is not None
5354 has_multiple_fields
= self
._get
_field
_setting
(field
, 'type') == 'combined'
5355 has_multiple_limits
= has_limit
and has_multiple_fields
and not self
._get
_field
_setting
(field
, 'same_limit')
5357 fields
= self
._get
_field
_setting
(field
, 'field') if has_multiple_fields
else (field
,)
5358 limits
= limit_text
.split(':') if has_multiple_limits
else (limit_text
,) if has_limit
else tuple()
5359 limit_count
= len(limits
)
5360 for (i
, f
) in enumerate(fields
):
5361 add_item(f
, reverse
, closest
,
5362 limits
[i
] if i
< limit_count
5363 else limits
[0] if has_limit
and not has_multiple_limits
5366 def print_verbose_info(self
, write_debug
):
5368 write_debug('Sort order given by user: %s' % ', '.join(self
._sort
_user
))
5369 if self
._sort
_extractor
:
5370 write_debug('Sort order given by extractor: %s' % ', '.join(self
._sort
_extractor
))
5371 write_debug('Formats sorted by: %s' % ', '.join(['%s%s%s' % (
5372 '+' if self
._get
_field
_setting
(field
, 'reverse') else '', field
,
5373 '%s%s(%s)' % ('~' if self
._get
_field
_setting
(field
, 'closest') else ':',
5374 self
._get
_field
_setting
(field
, 'limit_text'),
5375 self
._get
_field
_setting
(field
, 'limit'))
5376 if self
._get
_field
_setting
(field
, 'limit_text') is not None else '')
5377 for field
in self
._order
if self
._get
_field
_setting
(field
, 'visible')]))
5379 def _calculate_field_preference_from_value(self
, format
, field
, type, value
):
5380 reverse
= self
._get
_field
_setting
(field
, 'reverse')
5381 closest
= self
._get
_field
_setting
(field
, 'closest')
5382 limit
= self
._get
_field
_setting
(field
, 'limit')
5384 if type == 'extractor':
5385 maximum
= self
._get
_field
_setting
(field
, 'max')
5386 if value
is None or (maximum
is not None and value
>= maximum
):
5388 elif type == 'boolean':
5389 in_list
= self
._get
_field
_setting
(field
, 'in_list')
5390 not_in_list
= self
._get
_field
_setting
(field
, 'not_in_list')
5391 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
5392 elif type == 'ordered':
5393 value
= self
._resolve
_field
_value
(field
, value
, True)
5395 # try to convert to number
5396 val_num
= float_or_none(value
, default
=self
._get
_field
_setting
(field
, 'default'))
5397 is_num
= self
._get
_field
_setting
(field
, 'convert') != 'string' and val_num
is not None
5401 return ((-10, 0) if value
is None
5402 else (1, value
, 0) if not is_num
# if a field has mixed strings and numbers, strings are sorted higher
5403 else (0, -abs(value
- limit
), value
- limit
if reverse
else limit
- value
) if closest
5404 else (0, value
, 0) if not reverse
and (limit
is None or value
<= limit
)
5405 else (0, -value
, 0) if limit
is None or (reverse
and value
== limit
) or value
> limit
5406 else (-1, value
, 0))
5408 def _calculate_field_preference(self
, format
, field
):
5409 type = self
._get
_field
_setting
(field
, 'type') # extractor, boolean, ordered, field, multiple
5410 get_value
= lambda f
: format
.get(self
._get
_field
_setting
(f
, 'field'))
5411 if type == 'multiple':
5412 type = 'field' # Only 'field' is allowed in multiple for now
5413 actual_fields
= self
._get
_field
_setting
(field
, 'field')
5415 value
= self
._get
_field
_setting
(field
, 'function')(get_value(f
) for f
in actual_fields
)
5417 value
= get_value(field
)
5418 return self
._calculate
_field
_preference
_from
_value
(format
, field
, type, value
)
5420 def calculate_preference(self
, format
):
5421 # Determine missing protocol
5422 if not format
.get('protocol'):
5423 format
['protocol'] = determine_protocol(format
)
5425 # Determine missing ext
5426 if not format
.get('ext') and 'url' in format
:
5427 format
['ext'] = determine_ext(format
['url'])
5428 if format
.get('vcodec') == 'none':
5429 format
['audio_ext'] = format
['ext'] if format
.get('acodec') != 'none' else 'none'
5430 format
['video_ext'] = 'none'
5432 format
['video_ext'] = format
['ext']
5433 format
['audio_ext'] = 'none'
5434 # if format.get('preference') is None and format.get('ext') in ('f4f', 'f4m'): # Not supported?
5435 # format['preference'] = -1000
5437 if format
.get('preference') is None and format
.get('ext') == 'flv' and re
.match('[hx]265|he?vc?', format
.get('vcodec') or ''):
5438 # HEVC-over-FLV is out-of-spec by FLV's original spec
5439 # ref. https://trac.ffmpeg.org/ticket/6389
5440 # ref. https://github.com/yt-dlp/yt-dlp/pull/5821
5441 format
['preference'] = -100
5443 # Determine missing bitrates
5444 if format
.get('vcodec') == 'none':
5446 if format
.get('acodec') == 'none':
5448 if not format
.get('vbr') and format
.get('vcodec') != 'none':
5449 format
['vbr'] = try_call(lambda: format
['tbr'] - format
['abr']) or None
5450 if not format
.get('abr') and format
.get('acodec') != 'none':
5451 format
['abr'] = try_call(lambda: format
['tbr'] - format
['vbr']) or None
5452 if not format
.get('tbr'):
5453 format
['tbr'] = try_call(lambda: format
['vbr'] + format
['abr']) or None
5455 return tuple(self
._calculate
_field
_preference
(format
, field
) for field
in self
._order
)
5460 def __init__(self
, ydl
=None):
5463 def debug(self
, message
):
5465 self
._ydl
.write_debug(message
)
5467 def info(self
, message
):
5469 self
._ydl
.to_screen(message
)
5471 def warning(self
, message
, *, once
=False):
5473 self
._ydl
.report_warning(message
, once
)
5475 def error(self
, message
, *, is_error
=True):
5477 self
._ydl
.report_error(message
, is_error
=is_error
)
5479 def stdout(self
, message
):
5481 self
._ydl
.to_stdout(message
)
5483 def stderr(self
, message
):
5485 self
._ydl
.to_stderr(message
)