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 if sys
.platform
== 'win32':
674 drive_or_unc
, _
= os
.path
.splitdrive(s
)
680 norm_path
= os
.path
.normpath(remove_start(s
, drive_or_unc
)).split(os
.path
.sep
)
684 path_part
if path_part
in ['.', '..'] else re
.sub(r
'(?:[/<>:"\|\\?\*]|[\s.]$)', '#', path_part
)
685 for path_part
in norm_path
]
687 sanitized_path
.insert(0, drive_or_unc
+ os
.path
.sep
)
688 elif force
and s
and s
[0] == os
.path
.sep
:
689 sanitized_path
.insert(0, os
.path
.sep
)
690 return os
.path
.join(*sanitized_path
)
693 def sanitize_url(url
, *, scheme
='http'):
694 # Prepend protocol-less URLs with `http:` scheme in order to mitigate
695 # the number of unwanted failures due to missing protocol
698 elif url
.startswith('//'):
699 return f
'{scheme}:{url}'
700 # Fix some common typos seen so far
702 # https://github.com/ytdl-org/youtube-dl/issues/15649
703 (r
'^httpss://', r
'https://'),
704 # https://bx1.be/lives/direct-tv/
705 (r
'^rmtp([es]?)://', r
'rtmp\1://'),
707 for mistake
, fixup
in COMMON_TYPOS
:
708 if re
.match(mistake
, url
):
709 return re
.sub(mistake
, fixup
, url
)
713 def extract_basic_auth(url
):
714 parts
= urllib
.parse
.urlsplit(url
)
715 if parts
.username
is None:
717 url
= urllib
.parse
.urlunsplit(parts
._replace
(netloc
=(
718 parts
.hostname
if parts
.port
is None
719 else '%s:%d' % (parts
.hostname
, parts
.port
))))
720 auth_payload
= base64
.b64encode(
721 ('%s:%s' % (parts
.username
, parts
.password
or '')).encode())
722 return url
, f
'Basic {auth_payload.decode()}'
726 """Expand shell variables and ~"""
727 return os
.path
.expandvars(compat_expanduser(s
))
730 def orderedSet(iterable
, *, lazy
=False):
731 """Remove all duplicates from the input iterable"""
733 seen
= [] # Do not use set since the items can be unhashable
739 return _iter() if lazy
else list(_iter())
742 def _htmlentity_transform(entity_with_semicolon
):
743 """Transforms an HTML entity to a character."""
744 entity
= entity_with_semicolon
[:-1]
746 # Known non-numeric HTML entity
747 if entity
in html
.entities
.name2codepoint
:
748 return chr(html
.entities
.name2codepoint
[entity
])
750 # TODO: HTML5 allows entities without a semicolon.
751 # E.g. 'Éric' should be decoded as 'Éric'.
752 if entity_with_semicolon
in html
.entities
.html5
:
753 return html
.entities
.html5
[entity_with_semicolon
]
755 mobj
= re
.match(r
'#(x[0-9a-fA-F]+|[0-9]+)', entity
)
757 numstr
= mobj
.group(1)
758 if numstr
.startswith('x'):
760 numstr
= '0%s' % numstr
763 # See https://github.com/ytdl-org/youtube-dl/issues/7518
764 with contextlib
.suppress(ValueError):
765 return chr(int(numstr
, base
))
767 # Unknown entity in name, return its literal representation
768 return '&%s;' % entity
774 assert isinstance(s
, str)
777 r
'&([^&;]+;)', lambda m
: _htmlentity_transform(m
.group(1)), s
)
780 def escapeHTML(text
):
783 .replace('&', '&')
784 .replace('<', '<')
785 .replace('>', '>')
786 .replace('"', '"')
787 .replace("'", ''')
791 class netrc_from_content(netrc
.netrc
):
792 def __init__(self
, content
):
793 self
.hosts
, self
.macros
= {}, {}
794 with io
.StringIO(content
) as stream
:
795 self
._parse
('-', stream
, False)
798 class Popen(subprocess
.Popen
):
799 if sys
.platform
== 'win32':
800 _startupinfo
= subprocess
.STARTUPINFO()
801 _startupinfo
.dwFlags |
= subprocess
.STARTF_USESHOWWINDOW
806 def _fix_pyinstaller_ld_path(env
):
807 """Restore LD_LIBRARY_PATH when using PyInstaller
808 Ref: https://github.com/pyinstaller/pyinstaller/blob/develop/doc/runtime-information.rst#ld_library_path--libpath-considerations
809 https://github.com/yt-dlp/yt-dlp/issues/4573
811 if not hasattr(sys
, '_MEIPASS'):
815 orig
= env
.get(f
'{key}_ORIG')
821 _fix('LD_LIBRARY_PATH') # Linux
822 _fix('DYLD_LIBRARY_PATH') # macOS
824 def __init__(self
, *args
, env
=None, text
=False, **kwargs
):
826 env
= os
.environ
.copy()
827 self
._fix
_pyinstaller
_ld
_path
(env
)
829 self
.__text
_mode
= kwargs
.get('encoding') or kwargs
.get('errors') or text
or kwargs
.get('universal_newlines')
831 kwargs
['universal_newlines'] = True # For 3.6 compatibility
832 kwargs
.setdefault('encoding', 'utf-8')
833 kwargs
.setdefault('errors', 'replace')
834 super().__init
__(*args
, env
=env
, **kwargs
, startupinfo
=self
._startupinfo
)
836 def communicate_or_kill(self
, *args
, **kwargs
):
838 return self
.communicate(*args
, **kwargs
)
839 except BaseException
: # Including KeyboardInterrupt
840 self
.kill(timeout
=None)
843 def kill(self
, *, timeout
=0):
846 self
.wait(timeout
=timeout
)
849 def run(cls
, *args
, timeout
=None, **kwargs
):
850 with cls(*args
, **kwargs
) as proc
:
851 default
= '' if proc
.__text
_mode
else b
''
852 stdout
, stderr
= proc
.communicate_or_kill(timeout
=timeout
)
853 return stdout
or default
, stderr
or default
, proc
.returncode
856 def encodeArgument(s
):
857 # Legacy code that uses byte strings
858 # Uncomment the following line after fixing all post processors
859 # assert isinstance(s, str), 'Internal error: %r should be of type %r, is %r' % (s, str, type(s))
860 return s
if isinstance(s
, str) else s
.decode('ascii')
863 _timetuple
= collections
.namedtuple('Time', ('hours', 'minutes', 'seconds', 'milliseconds'))
866 def timetuple_from_msec(msec
):
867 secs
, msec
= divmod(msec
, 1000)
868 mins
, secs
= divmod(secs
, 60)
869 hrs
, mins
= divmod(mins
, 60)
870 return _timetuple(hrs
, mins
, secs
, msec
)
873 def formatSeconds(secs
, delim
=':', msec
=False):
874 time
= timetuple_from_msec(secs
* 1000)
876 ret
= '%d%s%02d%s%02d' % (time
.hours
, delim
, time
.minutes
, delim
, time
.seconds
)
878 ret
= '%d%s%02d' % (time
.minutes
, delim
, time
.seconds
)
880 ret
= '%d' % time
.seconds
881 return '%s.%03d' % (ret
, time
.milliseconds
) if msec
else ret
884 def bug_reports_message(before
=';'):
885 from ..update
import REPOSITORY
887 msg
= (f
'please report this issue on https://github.com/{REPOSITORY}/issues?q= , '
888 'filling out the appropriate issue template. Confirm you are on the latest version using yt-dlp -U')
890 before
= before
.rstrip()
891 if not before
or before
.endswith(('.', '!', '?')):
892 msg
= msg
[0].title() + msg
[1:]
894 return (before
+ ' ' if before
else '') + msg
897 class YoutubeDLError(Exception):
898 """Base exception for YoutubeDL errors."""
901 def __init__(self
, msg
=None):
904 elif self
.msg
is None:
905 self
.msg
= type(self
).__name
__
906 super().__init
__(self
.msg
)
909 class ExtractorError(YoutubeDLError
):
910 """Error during info extraction."""
912 def __init__(self
, msg
, tb
=None, expected
=False, cause
=None, video_id
=None, ie
=None):
913 """ tb, if given, is the original traceback (so that it can be printed out).
914 If expected is set, this is a normal error message and most likely not a bug in yt-dlp.
916 from ..networking
.exceptions
import network_exceptions
917 if sys
.exc_info()[0] in network_exceptions
:
920 self
.orig_msg
= str(msg
)
922 self
.expected
= expected
924 self
.video_id
= video_id
926 self
.exc_info
= sys
.exc_info() # preserve original exception
927 if isinstance(self
.exc_info
[1], ExtractorError
):
928 self
.exc_info
= self
.exc_info
[1].exc_info
929 super().__init
__(self
.__msg
)
934 format_field(self
.ie
, None, '[%s] '),
935 format_field(self
.video_id
, None, '%s: '),
937 format_field(self
.cause
, None, ' (caused by %r)'),
938 '' if self
.expected
else bug_reports_message()))
940 def format_traceback(self
):
941 return join_nonempty(
942 self
.traceback
and ''.join(traceback
.format_tb(self
.traceback
)),
943 self
.cause
and ''.join(traceback
.format_exception(None, self
.cause
, self
.cause
.__traceback
__)[1:]),
946 def __setattr__(self
, name
, value
):
947 super().__setattr
__(name
, value
)
948 if getattr(self
, 'msg', None) and name
not in ('msg', 'args'):
949 self
.msg
= self
.__msg
or type(self
).__name
__
950 self
.args
= (self
.msg
, ) # Cannot be property
953 class UnsupportedError(ExtractorError
):
954 def __init__(self
, url
):
956 'Unsupported URL: %s' % url
, expected
=True)
960 class RegexNotFoundError(ExtractorError
):
961 """Error when a regex didn't match"""
965 class GeoRestrictedError(ExtractorError
):
966 """Geographic restriction Error exception.
968 This exception may be thrown when a video is not available from your
969 geographic location due to geographic restrictions imposed by a website.
972 def __init__(self
, msg
, countries
=None, **kwargs
):
973 kwargs
['expected'] = True
974 super().__init
__(msg
, **kwargs
)
975 self
.countries
= countries
978 class UserNotLive(ExtractorError
):
979 """Error when a channel/user is not live"""
981 def __init__(self
, msg
=None, **kwargs
):
982 kwargs
['expected'] = True
983 super().__init
__(msg
or 'The channel is not currently live', **kwargs
)
986 class DownloadError(YoutubeDLError
):
987 """Download Error exception.
989 This exception may be thrown by FileDownloader objects if they are not
990 configured to continue on errors. They will contain the appropriate
994 def __init__(self
, msg
, exc_info
=None):
995 """ exc_info, if given, is the original exception that caused the trouble (as returned by sys.exc_info()). """
996 super().__init
__(msg
)
997 self
.exc_info
= exc_info
1000 class EntryNotInPlaylist(YoutubeDLError
):
1001 """Entry not in playlist exception.
1003 This exception will be thrown by YoutubeDL when a requested entry
1004 is not found in the playlist info_dict
1006 msg
= 'Entry not found in info'
1009 class SameFileError(YoutubeDLError
):
1010 """Same File exception.
1012 This exception will be thrown by FileDownloader objects if they detect
1013 multiple files would have to be downloaded to the same file on disk.
1015 msg
= 'Fixed output name but more than one file to download'
1017 def __init__(self
, filename
=None):
1018 if filename
is not None:
1019 self
.msg
+= f
': {filename}'
1020 super().__init
__(self
.msg
)
1023 class PostProcessingError(YoutubeDLError
):
1024 """Post Processing exception.
1026 This exception may be raised by PostProcessor's .run() method to
1027 indicate an error in the postprocessing task.
1031 class DownloadCancelled(YoutubeDLError
):
1032 """ Exception raised when the download queue should be interrupted """
1033 msg
= 'The download was cancelled'
1036 class ExistingVideoReached(DownloadCancelled
):
1037 """ --break-on-existing triggered """
1038 msg
= 'Encountered a video that is already in the archive, stopping due to --break-on-existing'
1041 class RejectedVideoReached(DownloadCancelled
):
1042 """ --break-match-filter triggered """
1043 msg
= 'Encountered a video that did not match filter, stopping due to --break-match-filter'
1046 class MaxDownloadsReached(DownloadCancelled
):
1047 """ --max-downloads limit has been reached. """
1048 msg
= 'Maximum number of downloads reached, stopping due to --max-downloads'
1051 class ReExtractInfo(YoutubeDLError
):
1052 """ Video info needs to be re-extracted. """
1054 def __init__(self
, msg
, expected
=False):
1055 super().__init
__(msg
)
1056 self
.expected
= expected
1059 class ThrottledDownload(ReExtractInfo
):
1060 """ Download speed below --throttled-rate. """
1061 msg
= 'The download speed is below throttle limit'
1064 super().__init
__(self
.msg
, expected
=False)
1067 class UnavailableVideoError(YoutubeDLError
):
1068 """Unavailable Format exception.
1070 This exception will be thrown when a video is requested
1071 in a format that is not available for that video.
1073 msg
= 'Unable to download video'
1075 def __init__(self
, err
=None):
1077 self
.msg
+= f
': {err}'
1078 super().__init
__(self
.msg
)
1081 class ContentTooShortError(YoutubeDLError
):
1082 """Content Too Short exception.
1084 This exception may be raised by FileDownloader objects when a file they
1085 download is too small for what the server announced first, indicating
1086 the connection was probably interrupted.
1089 def __init__(self
, downloaded
, expected
):
1090 super().__init
__(f
'Downloaded {downloaded} bytes, expected {expected} bytes')
1092 self
.downloaded
= downloaded
1093 self
.expected
= expected
1096 class XAttrMetadataError(YoutubeDLError
):
1097 def __init__(self
, code
=None, msg
='Unknown error'):
1098 super().__init
__(msg
)
1102 # Parsing code and msg
1103 if (self
.code
in (errno
.ENOSPC
, errno
.EDQUOT
)
1104 or 'No space left' in self
.msg
or 'Disk quota exceeded' in self
.msg
):
1105 self
.reason
= 'NO_SPACE'
1106 elif self
.code
== errno
.E2BIG
or 'Argument list too long' in self
.msg
:
1107 self
.reason
= 'VALUE_TOO_LONG'
1109 self
.reason
= 'NOT_SUPPORTED'
1112 class XAttrUnavailableError(YoutubeDLError
):
1116 def is_path_like(f
):
1117 return isinstance(f
, (str, bytes, os
.PathLike
))
1120 def extract_timezone(date_str
):
1123 ^.{8,}? # >=8 char non-TZ prefix, if present
1124 (?P<tz>Z| # just the UTC Z, or
1125 (?:(?<=.\b\d{4}|\b\d{2}:\d\d)| # preceded by 4 digits or hh:mm or
1126 (?<!.\b[a-zA-Z]{3}|[a-zA-Z]{4}|..\b\d\d)) # not preceded by 3 alpha word or >= 4 alpha or 2 digits
1127 [ ]? # optional space
1128 (?P<sign>\+|-) # +/-
1129 (?P<hours>[0-9]{2}):?(?P<minutes>[0-9]{2}) # hh[:]mm
1133 m
= re
.search(r
'\d{1,2}:\d{1,2}(?:\.\d+)?(?P<tz>\s*[A-Z]+)$', date_str
)
1134 timezone
= TIMEZONE_NAMES
.get(m
and m
.group('tz').strip())
1135 if timezone
is not None:
1136 date_str
= date_str
[:-len(m
.group('tz'))]
1137 timezone
= datetime
.timedelta(hours
=timezone
or 0)
1139 date_str
= date_str
[:-len(m
.group('tz'))]
1140 if not m
.group('sign'):
1141 timezone
= datetime
.timedelta()
1143 sign
= 1 if m
.group('sign') == '+' else -1
1144 timezone
= datetime
.timedelta(
1145 hours
=sign
* int(m
.group('hours')),
1146 minutes
=sign
* int(m
.group('minutes')))
1147 return timezone
, date_str
1150 def parse_iso8601(date_str
, delimiter
='T', timezone
=None):
1151 """ Return a UNIX timestamp from the given date """
1153 if date_str
is None:
1156 date_str
= re
.sub(r
'\.[0-9]+', '', date_str
)
1158 if timezone
is None:
1159 timezone
, date_str
= extract_timezone(date_str
)
1161 with contextlib
.suppress(ValueError):
1162 date_format
= f
'%Y-%m-%d{delimiter}%H:%M:%S'
1163 dt
= datetime
.datetime
.strptime(date_str
, date_format
) - timezone
1164 return calendar
.timegm(dt
.timetuple())
1167 def date_formats(day_first
=True):
1168 return DATE_FORMATS_DAY_FIRST
if day_first
else DATE_FORMATS_MONTH_FIRST
1171 def unified_strdate(date_str
, day_first
=True):
1172 """Return a string with the date in the format YYYYMMDD"""
1174 if date_str
is None:
1178 date_str
= date_str
.replace(',', ' ')
1179 # Remove AM/PM + timezone
1180 date_str
= re
.sub(r
'(?i)\s*(?:AM|PM)(?:\s+[A-Z]+)?', '', date_str
)
1181 _
, date_str
= extract_timezone(date_str
)
1183 for expression
in date_formats(day_first
):
1184 with contextlib
.suppress(ValueError):
1185 upload_date
= datetime
.datetime
.strptime(date_str
, expression
).strftime('%Y%m%d')
1186 if upload_date
is None:
1187 timetuple
= email
.utils
.parsedate_tz(date_str
)
1189 with contextlib
.suppress(ValueError):
1190 upload_date
= datetime
.datetime(*timetuple
[:6]).strftime('%Y%m%d')
1191 if upload_date
is not None:
1192 return str(upload_date
)
1195 def unified_timestamp(date_str
, day_first
=True):
1196 if not isinstance(date_str
, str):
1199 date_str
= re
.sub(r
'\s+', ' ', re
.sub(
1200 r
'(?i)[,|]|(mon|tues?|wed(nes)?|thu(rs)?|fri|sat(ur)?)(day)?', '', date_str
))
1202 pm_delta
= 12 if re
.search(r
'(?i)PM', date_str
) else 0
1203 timezone
, date_str
= extract_timezone(date_str
)
1205 # Remove AM/PM + timezone
1206 date_str
= re
.sub(r
'(?i)\s*(?:AM|PM)(?:\s+[A-Z]+)?', '', date_str
)
1208 # Remove unrecognized timezones from ISO 8601 alike timestamps
1209 m
= re
.search(r
'\d{1,2}:\d{1,2}(?:\.\d+)?(?P<tz>\s*[A-Z]+)$', date_str
)
1211 date_str
= date_str
[:-len(m
.group('tz'))]
1213 # Python only supports microseconds, so remove nanoseconds
1214 m
= re
.search(r
'^([0-9]{4,}-[0-9]{1,2}-[0-9]{1,2}T[0-9]{1,2}:[0-9]{1,2}:[0-9]{1,2}\.[0-9]{6})[0-9]+$', date_str
)
1216 date_str
= m
.group(1)
1218 for expression
in date_formats(day_first
):
1219 with contextlib
.suppress(ValueError):
1220 dt
= datetime
.datetime
.strptime(date_str
, expression
) - timezone
+ datetime
.timedelta(hours
=pm_delta
)
1221 return calendar
.timegm(dt
.timetuple())
1223 timetuple
= email
.utils
.parsedate_tz(date_str
)
1225 return calendar
.timegm(timetuple
) + pm_delta
* 3600 - timezone
.total_seconds()
1228 def determine_ext(url
, default_ext
='unknown_video'):
1229 if url
is None or '.' not in url
:
1231 guess
= url
.partition('?')[0].rpartition('.')[2]
1232 if re
.match(r
'^[A-Za-z0-9]+$', guess
):
1234 # Try extract ext from URLs like http://example.com/foo/bar.mp4/?download
1235 elif guess
.rstrip('/') in KNOWN_EXTENSIONS
:
1236 return guess
.rstrip('/')
1241 def subtitles_filename(filename
, sub_lang
, sub_format
, expected_real_ext
=None):
1242 return replace_extension(filename
, sub_lang
+ '.' + sub_format
, expected_real_ext
)
1245 def datetime_from_str(date_str
, precision
='auto', format
='%Y%m%d'):
1247 Return a datetime object from a string.
1249 (now|today|yesterday|DATE)([+-]\d+(microsecond|second|minute|hour|day|week|month|year)s?)?
1251 @param format strftime format of DATE
1252 @param precision Round the datetime object: auto|microsecond|second|minute|hour|day
1253 auto: round to the unit provided in date_str (if applicable).
1255 auto_precision
= False
1256 if precision
== 'auto':
1257 auto_precision
= True
1258 precision
= 'microsecond'
1259 today
= datetime_round(datetime
.datetime
.utcnow(), precision
)
1260 if date_str
in ('now', 'today'):
1262 if date_str
== 'yesterday':
1263 return today
- datetime
.timedelta(days
=1)
1265 r
'(?P<start>.+)(?P<sign>[+-])(?P<time>\d+)(?P<unit>microsecond|second|minute|hour|day|week|month|year)s?',
1267 if match
is not None:
1268 start_time
= datetime_from_str(match
.group('start'), precision
, format
)
1269 time
= int(match
.group('time')) * (-1 if match
.group('sign') == '-' else 1)
1270 unit
= match
.group('unit')
1271 if unit
== 'month' or unit
== 'year':
1272 new_date
= datetime_add_months(start_time
, time
* 12 if unit
== 'year' else time
)
1278 delta
= datetime
.timedelta(**{unit + 's': time}
)
1279 new_date
= start_time
+ delta
1281 return datetime_round(new_date
, unit
)
1284 return datetime_round(datetime
.datetime
.strptime(date_str
, format
), precision
)
1287 def date_from_str(date_str
, format
='%Y%m%d', strict
=False):
1289 Return a date object from a string using datetime_from_str
1291 @param strict Restrict allowed patterns to "YYYYMMDD" and
1292 (now|today|yesterday)(-\d+(day|week|month|year)s?)?
1294 if strict
and not re
.fullmatch(r
'\d{8}|(now|today|yesterday)(-\d+(day|week|month|year)s?)?', date_str
):
1295 raise ValueError(f
'Invalid date format "{date_str}"')
1296 return datetime_from_str(date_str
, precision
='microsecond', format
=format
).date()
1299 def datetime_add_months(dt
, months
):
1300 """Increment/Decrement a datetime object by months."""
1301 month
= dt
.month
+ months
- 1
1302 year
= dt
.year
+ month
// 12
1303 month
= month
% 12 + 1
1304 day
= min(dt
.day
, calendar
.monthrange(year
, month
)[1])
1305 return dt
.replace(year
, month
, day
)
1308 def datetime_round(dt
, precision
='day'):
1310 Round a datetime object's time to a specific precision
1312 if precision
== 'microsecond':
1321 roundto
= lambda x
, n
: ((x
+ n
/ 2) // n
) * n
1322 timestamp
= calendar
.timegm(dt
.timetuple())
1323 return datetime
.datetime
.utcfromtimestamp(roundto(timestamp
, unit_seconds
[precision
]))
1326 def hyphenate_date(date_str
):
1328 Convert a date in 'YYYYMMDD' format to 'YYYY-MM-DD' format"""
1329 match
= re
.match(r
'^(\d\d\d\d)(\d\d)(\d\d)$', date_str
)
1330 if match
is not None:
1331 return '-'.join(match
.groups())
1337 """Represents a time interval between two dates"""
1339 def __init__(self
, start
=None, end
=None):
1340 """start and end must be strings in the format accepted by date"""
1341 if start
is not None:
1342 self
.start
= date_from_str(start
, strict
=True)
1344 self
.start
= datetime
.datetime
.min.date()
1346 self
.end
= date_from_str(end
, strict
=True)
1348 self
.end
= datetime
.datetime
.max.date()
1349 if self
.start
> self
.end
:
1350 raise ValueError('Date range: "%s" , the start date must be before the end date' % self
)
1354 """Returns a range that only contains the given day"""
1355 return cls(day
, day
)
1357 def __contains__(self
, date
):
1358 """Check if the date is in the range"""
1359 if not isinstance(date
, datetime
.date
):
1360 date
= date_from_str(date
)
1361 return self
.start
<= date
<= self
.end
1364 return f
'{__name__}.{type(self).__name__}({self.start.isoformat()!r}, {self.end.isoformat()!r})'
1366 def __eq__(self
, other
):
1367 return (isinstance(other
, DateRange
)
1368 and self
.start
== other
.start
and self
.end
== other
.end
)
1372 def system_identifier():
1373 python_implementation
= platform
.python_implementation()
1374 if python_implementation
== 'PyPy' and hasattr(sys
, 'pypy_version_info'):
1375 python_implementation
+= ' version %d.%d.%d' % sys
.pypy_version_info
[:3]
1377 with contextlib
.suppress(OSError): # We may not have access to the executable
1378 libc_ver
= platform
.libc_ver()
1380 return 'Python %s (%s %s %s) - %s (%s%s)' % (
1381 platform
.python_version(),
1382 python_implementation
,
1384 platform
.architecture()[0],
1385 platform
.platform(),
1386 ssl
.OPENSSL_VERSION
,
1387 format_field(join_nonempty(*libc_ver
, delim
=' '), None, ', %s'),
1392 def get_windows_version():
1393 ''' Get Windows version. returns () if it's not running on Windows '''
1394 if compat_os_name
== 'nt':
1395 return version_tuple(platform
.win32_ver()[1])
1400 def write_string(s
, out
=None, encoding
=None):
1401 assert isinstance(s
, str)
1402 out
= out
or sys
.stderr
1403 # `sys.stderr` might be `None` (Ref: https://github.com/pyinstaller/pyinstaller/pull/7217)
1407 if compat_os_name
== 'nt' and supports_terminal_sequences(out
):
1408 s
= re
.sub(r
'([\r\n]+)', r
' \1', s
)
1410 enc
, buffer = None, out
1411 if 'b' in getattr(out
, 'mode', ''):
1412 enc
= encoding
or preferredencoding()
1413 elif hasattr(out
, 'buffer'):
1415 enc
= encoding
or getattr(out
, 'encoding', None) or preferredencoding()
1417 buffer.write(s
.encode(enc
, 'ignore') if enc
else s
)
1421 # TODO: Use global logger
1422 def deprecation_warning(msg
, *, printer
=None, stacklevel
=0, **kwargs
):
1423 from .. import _IN_CLI
1425 if msg
in deprecation_warning
._cache
:
1427 deprecation_warning
._cache
.add(msg
)
1429 return printer(f
'{msg}{bug_reports_message()}', **kwargs
)
1430 return write_string(f
'ERROR: {msg}{bug_reports_message()}\n', **kwargs
)
1433 warnings
.warn(DeprecationWarning(msg
), stacklevel
=stacklevel
+ 3)
1436 deprecation_warning
._cache
= set()
1439 def bytes_to_intlist(bs
):
1442 if isinstance(bs
[0], int): # Python 3
1445 return [ord(c
) for c
in bs
]
1448 def intlist_to_bytes(xs
):
1451 return struct
.pack('%dB' % len(xs
), *xs
)
1454 class LockingUnsupportedError(OSError):
1455 msg
= 'File locking is not supported'
1458 super().__init
__(self
.msg
)
1461 # Cross-platform file locking
1462 if sys
.platform
== 'win32':
1464 import ctypes
.wintypes
1467 class OVERLAPPED(ctypes
.Structure
):
1469 ('Internal', ctypes
.wintypes
.LPVOID
),
1470 ('InternalHigh', ctypes
.wintypes
.LPVOID
),
1471 ('Offset', ctypes
.wintypes
.DWORD
),
1472 ('OffsetHigh', ctypes
.wintypes
.DWORD
),
1473 ('hEvent', ctypes
.wintypes
.HANDLE
),
1476 kernel32
= ctypes
.WinDLL('kernel32')
1477 LockFileEx
= kernel32
.LockFileEx
1478 LockFileEx
.argtypes
= [
1479 ctypes
.wintypes
.HANDLE
, # hFile
1480 ctypes
.wintypes
.DWORD
, # dwFlags
1481 ctypes
.wintypes
.DWORD
, # dwReserved
1482 ctypes
.wintypes
.DWORD
, # nNumberOfBytesToLockLow
1483 ctypes
.wintypes
.DWORD
, # nNumberOfBytesToLockHigh
1484 ctypes
.POINTER(OVERLAPPED
) # Overlapped
1486 LockFileEx
.restype
= ctypes
.wintypes
.BOOL
1487 UnlockFileEx
= kernel32
.UnlockFileEx
1488 UnlockFileEx
.argtypes
= [
1489 ctypes
.wintypes
.HANDLE
, # hFile
1490 ctypes
.wintypes
.DWORD
, # dwReserved
1491 ctypes
.wintypes
.DWORD
, # nNumberOfBytesToLockLow
1492 ctypes
.wintypes
.DWORD
, # nNumberOfBytesToLockHigh
1493 ctypes
.POINTER(OVERLAPPED
) # Overlapped
1495 UnlockFileEx
.restype
= ctypes
.wintypes
.BOOL
1496 whole_low
= 0xffffffff
1497 whole_high
= 0x7fffffff
1499 def _lock_file(f
, exclusive
, block
):
1500 overlapped
= OVERLAPPED()
1501 overlapped
.Offset
= 0
1502 overlapped
.OffsetHigh
= 0
1503 overlapped
.hEvent
= 0
1504 f
._lock
_file
_overlapped
_p
= ctypes
.pointer(overlapped
)
1506 if not LockFileEx(msvcrt
.get_osfhandle(f
.fileno()),
1507 (0x2 if exclusive
else 0x0) |
(0x0 if block
else 0x1),
1508 0, whole_low
, whole_high
, f
._lock
_file
_overlapped
_p
):
1509 # NB: No argument form of "ctypes.FormatError" does not work on PyPy
1510 raise BlockingIOError(f
'Locking file failed: {ctypes.FormatError(ctypes.GetLastError())!r}')
1512 def _unlock_file(f
):
1513 assert f
._lock
_file
_overlapped
_p
1514 handle
= msvcrt
.get_osfhandle(f
.fileno())
1515 if not UnlockFileEx(handle
, 0, whole_low
, whole_high
, f
._lock
_file
_overlapped
_p
):
1516 raise OSError('Unlocking file failed: %r' % ctypes
.FormatError())
1522 def _lock_file(f
, exclusive
, block
):
1523 flags
= fcntl
.LOCK_EX
if exclusive
else fcntl
.LOCK_SH
1525 flags |
= fcntl
.LOCK_NB
1527 fcntl
.flock(f
, flags
)
1528 except BlockingIOError
:
1530 except OSError: # AOSP does not have flock()
1531 fcntl
.lockf(f
, flags
)
1533 def _unlock_file(f
):
1534 with contextlib
.suppress(OSError):
1535 return fcntl
.flock(f
, fcntl
.LOCK_UN
)
1536 with contextlib
.suppress(OSError):
1537 return fcntl
.lockf(f
, fcntl
.LOCK_UN
) # AOSP does not have flock()
1538 return fcntl
.flock(f
, fcntl
.LOCK_UN | fcntl
.LOCK_NB
) # virtiofs needs LOCK_NB on unlocking
1542 def _lock_file(f
, exclusive
, block
):
1543 raise LockingUnsupportedError()
1545 def _unlock_file(f
):
1546 raise LockingUnsupportedError()
1552 def __init__(self
, filename
, mode
, block
=True, encoding
=None):
1553 if mode
not in {'r', 'rb', 'a', 'ab', 'w', 'wb'}
:
1554 raise NotImplementedError(mode
)
1555 self
.mode
, self
.block
= mode
, block
1557 writable
= any(f
in mode
for f
in 'wax+')
1558 readable
= any(f
in mode
for f
in 'r+')
1559 flags
= functools
.reduce(operator
.ior
, (
1560 getattr(os
, 'O_CLOEXEC', 0), # UNIX only
1561 getattr(os
, 'O_BINARY', 0), # Windows only
1562 getattr(os
, 'O_NOINHERIT', 0), # Windows only
1563 os
.O_CREAT
if writable
else 0, # O_TRUNC only after locking
1564 os
.O_APPEND
if 'a' in mode
else 0,
1565 os
.O_EXCL
if 'x' in mode
else 0,
1566 os
.O_RDONLY
if not writable
else os
.O_RDWR
if readable
else os
.O_WRONLY
,
1569 self
.f
= os
.fdopen(os
.open(filename
, flags
, 0o666), mode
, encoding
=encoding
)
1571 def __enter__(self
):
1572 exclusive
= 'r' not in self
.mode
1574 _lock_file(self
.f
, exclusive
, self
.block
)
1579 if 'w' in self
.mode
:
1582 except OSError as e
:
1584 errno
.ESPIPE
, # Illegal seek - expected for FIFO
1585 errno
.EINVAL
, # Invalid argument - expected for /dev/null
1594 _unlock_file(self
.f
)
1598 def __exit__(self
, *_
):
1607 def __getattr__(self
, attr
):
1608 return getattr(self
.f
, attr
)
1615 def get_filesystem_encoding():
1616 encoding
= sys
.getfilesystemencoding()
1617 return encoding
if encoding
is not None else 'utf-8'
1620 def shell_quote(args
):
1622 encoding
= get_filesystem_encoding()
1624 if isinstance(a
, bytes):
1625 # We may get a filename encoded with 'encodeFilename'
1626 a
= a
.decode(encoding
)
1627 quoted_args
.append(compat_shlex_quote(a
))
1628 return ' '.join(quoted_args
)
1631 def smuggle_url(url
, data
):
1632 """ Pass additional data in a URL for internal use. """
1634 url
, idata
= unsmuggle_url(url
, {})
1636 sdata
= urllib
.parse
.urlencode(
1637 {'__youtubedl_smuggle': json.dumps(data)}
)
1638 return url
+ '#' + sdata
1641 def unsmuggle_url(smug_url
, default
=None):
1642 if '#__youtubedl_smuggle' not in smug_url
:
1643 return smug_url
, default
1644 url
, _
, sdata
= smug_url
.rpartition('#')
1645 jsond
= urllib
.parse
.parse_qs(sdata
)['__youtubedl_smuggle'][0]
1646 data
= json
.loads(jsond
)
1650 def format_decimal_suffix(num
, fmt
='%d%s', *, factor
=1000):
1651 """ Formats numbers with decimal sufixes like K, M, etc """
1652 num
, factor
= float_or_none(num
), float(factor
)
1653 if num
is None or num
< 0:
1655 POSSIBLE_SUFFIXES
= 'kMGTPEZY'
1656 exponent
= 0 if num
== 0 else min(int(math
.log(num
, factor
)), len(POSSIBLE_SUFFIXES
))
1657 suffix
= ['', *POSSIBLE_SUFFIXES
][exponent
]
1659 suffix
= {'k': 'Ki', '': ''}
.get(suffix
, f
'{suffix}i')
1660 converted
= num
/ (factor
** exponent
)
1661 return fmt
% (converted
, suffix
)
1664 def format_bytes(bytes):
1665 return format_decimal_suffix(bytes, '%.2f%sB', factor
=1024) or 'N/A'
1668 def lookup_unit_table(unit_table
, s
, strict
=False):
1669 num_re
= NUMBER_RE
if strict
else NUMBER_RE
.replace(R
'\.', '[,.]')
1670 units_re
= '|'.join(re
.escape(u
) for u
in unit_table
)
1671 m
= (re
.fullmatch
if strict
else re
.match
)(
1672 rf
'(?P<num>{num_re})\s*(?P<unit>{units_re})\b', s
)
1676 num
= float(m
.group('num').replace(',', '.'))
1677 mult
= unit_table
[m
.group('unit')]
1678 return round(num
* mult
)
1682 """Parse a string indicating a byte quantity into an integer"""
1683 return lookup_unit_table(
1684 {u: 1024**i for i, u in enumerate(['', *'KMGTPEZY'])}
,
1685 s
.upper(), strict
=True)
1688 def parse_filesize(s
):
1692 # The lower-case forms are of course incorrect and unofficial,
1693 # but we support those too
1710 'megabytes': 1000 ** 2,
1711 'mebibytes': 1024 ** 2,
1717 'gigabytes': 1000 ** 3,
1718 'gibibytes': 1024 ** 3,
1724 'terabytes': 1000 ** 4,
1725 'tebibytes': 1024 ** 4,
1731 'petabytes': 1000 ** 5,
1732 'pebibytes': 1024 ** 5,
1738 'exabytes': 1000 ** 6,
1739 'exbibytes': 1024 ** 6,
1745 'zettabytes': 1000 ** 7,
1746 'zebibytes': 1024 ** 7,
1752 'yottabytes': 1000 ** 8,
1753 'yobibytes': 1024 ** 8,
1756 return lookup_unit_table(_UNIT_TABLE
, s
)
1763 s
= re
.sub(r
'^[^\d]+\s', '', s
).strip()
1765 if re
.match(r
'^[\d,.]+$', s
):
1766 return str_to_int(s
)
1779 ret
= lookup_unit_table(_UNIT_TABLE
, s
)
1783 mobj
= re
.match(r
'([\d,.]+)(?:$|\s)', s
)
1785 return str_to_int(mobj
.group(1))
1788 def parse_resolution(s
, *, lenient
=False):
1793 mobj
= re
.search(r
'(?P<w>\d+)\s*[xX×,]\s*(?P<h>\d+)', s
)
1795 mobj
= re
.search(r
'(?<![a-zA-Z0-9])(?P<w>\d+)\s*[xX×,]\s*(?P<h>\d+)(?![a-zA-Z0-9])', s
)
1798 'width': int(mobj
.group('w')),
1799 'height': int(mobj
.group('h')),
1802 mobj
= re
.search(r
'(?<![a-zA-Z0-9])(\d+)[pPiI](?![a-zA-Z0-9])', s
)
1804 return {'height': int(mobj.group(1))}
1806 mobj
= re
.search(r
'\b([48])[kK]\b', s
)
1808 return {'height': int(mobj.group(1)) * 540}
1813 def parse_bitrate(s
):
1814 if not isinstance(s
, str):
1816 mobj
= re
.search(r
'\b(\d+)\s*kbps', s
)
1818 return int(mobj
.group(1))
1821 def month_by_name(name
, lang
='en'):
1822 """ Return the number of a month by (locale-independently) English name """
1824 month_names
= MONTH_NAMES
.get(lang
, MONTH_NAMES
['en'])
1827 return month_names
.index(name
) + 1
1832 def month_by_abbreviation(abbrev
):
1833 """ Return the number of a month by (locale-independently) English
1837 return [s
[:3] for s
in ENGLISH_MONTH_NAMES
].index(abbrev
) + 1
1842 def fix_xml_ampersands(xml_str
):
1843 """Replace all the '&' by '&' in XML"""
1845 r
'&(?!amp;|lt;|gt;|apos;|quot;|#x[0-9a-fA-F]{,4};|#[0-9]{,4};)',
1850 def setproctitle(title
):
1851 assert isinstance(title
, str)
1853 # Workaround for https://github.com/yt-dlp/yt-dlp/issues/4541
1860 libc
= ctypes
.cdll
.LoadLibrary('libc.so.6')
1864 # LoadLibrary in Windows Python 2.7.13 only expects
1865 # a bytestring, but since unicode_literals turns
1866 # every string into a unicode string, it fails.
1868 title_bytes
= title
.encode()
1869 buf
= ctypes
.create_string_buffer(len(title_bytes
))
1870 buf
.value
= title_bytes
1872 libc
.prctl(15, buf
, 0, 0, 0)
1873 except AttributeError:
1874 return # Strange libc, just skip this
1877 def remove_start(s
, start
):
1878 return s
[len(start
):] if s
is not None and s
.startswith(start
) else s
1881 def remove_end(s
, end
):
1882 return s
[:-len(end
)] if s
is not None and s
.endswith(end
) else s
1885 def remove_quotes(s
):
1886 if s
is None or len(s
) < 2:
1888 for quote
in ('"', "'", ):
1889 if s
[0] == quote
and s
[-1] == quote
:
1894 def get_domain(url
):
1896 This implementation is inconsistent, but is kept for compatibility.
1897 Use this only for "webpage_url_domain"
1899 return remove_start(urllib
.parse
.urlparse(url
).netloc
, 'www.') or None
1902 def url_basename(url
):
1903 path
= urllib
.parse
.urlparse(url
).path
1904 return path
.strip('/').split('/')[-1]
1908 return re
.match(r
'https?://[^?#]+/', url
).group()
1911 def urljoin(base
, path
):
1912 if isinstance(path
, bytes):
1913 path
= path
.decode()
1914 if not isinstance(path
, str) or not path
:
1916 if re
.match(r
'^(?:[a-zA-Z][a-zA-Z0-9+-.]*:)?//', path
):
1918 if isinstance(base
, bytes):
1919 base
= base
.decode()
1920 if not isinstance(base
, str) or not re
.match(
1921 r
'^(?:https?:)?//', base
):
1923 return urllib
.parse
.urljoin(base
, path
)
1926 def int_or_none(v
, scale
=1, default
=None, get_attr
=None, invscale
=1):
1927 if get_attr
and v
is not None:
1928 v
= getattr(v
, get_attr
, None)
1930 return int(v
) * invscale
// scale
1931 except (ValueError, TypeError, OverflowError):
1935 def str_or_none(v
, default
=None):
1936 return default
if v
is None else str(v
)
1939 def str_to_int(int_str
):
1940 """ A more relaxed version of int_or_none """
1941 if isinstance(int_str
, int):
1943 elif isinstance(int_str
, str):
1944 int_str
= re
.sub(r
'[,\.\+]', '', int_str
)
1945 return int_or_none(int_str
)
1948 def float_or_none(v
, scale
=1, invscale
=1, default
=None):
1952 return float(v
) * invscale
/ scale
1953 except (ValueError, TypeError):
1957 def bool_or_none(v
, default
=None):
1958 return v
if isinstance(v
, bool) else default
1961 def strip_or_none(v
, default
=None):
1962 return v
.strip() if isinstance(v
, str) else default
1965 def url_or_none(url
):
1966 if not url
or not isinstance(url
, str):
1969 return url
if re
.match(r
'^(?:(?:https?|rt(?:m(?:pt?[es]?|fp)|sp[su]?)|mms|ftps?):)?//', url
) else None
1972 def strftime_or_none(timestamp
, date_format
='%Y%m%d', default
=None):
1973 datetime_object
= None
1975 if isinstance(timestamp
, (int, float)): # unix timestamp
1976 # Using naive datetime here can break timestamp() in Windows
1977 # Ref: https://github.com/yt-dlp/yt-dlp/issues/5185, https://github.com/python/cpython/issues/94414
1978 # Also, datetime.datetime.fromtimestamp breaks for negative timestamps
1979 # Ref: https://github.com/yt-dlp/yt-dlp/issues/6706#issuecomment-1496842642
1980 datetime_object
= (datetime
.datetime
.fromtimestamp(0, datetime
.timezone
.utc
)
1981 + datetime
.timedelta(seconds
=timestamp
))
1982 elif isinstance(timestamp
, str): # assume YYYYMMDD
1983 datetime_object
= datetime
.datetime
.strptime(timestamp
, '%Y%m%d')
1984 date_format
= re
.sub( # Support %s on windows
1985 r
'(?<!%)(%%)*%s', rf
'\g<1>{int(datetime_object.timestamp())}', date_format
)
1986 return datetime_object
.strftime(date_format
)
1987 except (ValueError, TypeError, AttributeError):
1991 def parse_duration(s
):
1992 if not isinstance(s
, str):
1998 days
, hours
, mins
, secs
, ms
= [None] * 5
1999 m
= re
.match(r
'''(?x)
2001 (?:(?:(?P<days>[0-9]+):)?(?P<hours>[0-9]+):)?(?P<mins>[0-9]+):)?
2002 (?P<secs>(?(before_secs)[0-9]{1,2}|[0-9]+))
2003 (?P<ms>[.:][0-9]+)?Z?$
2006 days
, hours
, mins
, secs
, ms
= m
.group('days', 'hours', 'mins', 'secs', 'ms')
2011 [0-9]+\s*y(?:ears?)?,?\s*
2014 [0-9]+\s*m(?:onths?)?,?\s*
2017 [0-9]+\s*w(?:eeks?)?,?\s*
2020 (?P<days>[0-9]+)\s*d(?:ays?)?,?\s*
2024 (?P<hours>[0-9]+)\s*h(?:(?:ou)?rs?)?,?\s*
2027 (?P<mins>[0-9]+)\s*m(?:in(?:ute)?s?)?,?\s*
2030 (?P<secs>[0-9]+)(?P<ms>\.[0-9]+)?\s*s(?:ec(?:ond)?s?)?\s*
2033 days
, hours
, mins
, secs
, ms
= m
.groups()
2035 m
= re
.match(r
'(?i)(?:(?P<hours>[0-9.]+)\s*(?:hours?)|(?P<mins>[0-9.]+)\s*(?:mins?\.?|minutes?)\s*)Z?$', s
)
2037 hours
, mins
= m
.groups()
2042 ms
= ms
.replace(':', '.')
2043 return sum(float(part
or 0) * mult
for part
, mult
in (
2044 (days
, 86400), (hours
, 3600), (mins
, 60), (secs
, 1), (ms
, 1)))
2047 def prepend_extension(filename
, ext
, expected_real_ext
=None):
2048 name
, real_ext
= os
.path
.splitext(filename
)
2050 f
'{name}.{ext}{real_ext}'
2051 if not expected_real_ext
or real_ext
[1:] == expected_real_ext
2052 else f
'{filename}.{ext}')
2055 def replace_extension(filename
, ext
, expected_real_ext
=None):
2056 name
, real_ext
= os
.path
.splitext(filename
)
2057 return '{}.{}'.format(
2058 name
if not expected_real_ext
or real_ext
[1:] == expected_real_ext
else filename
,
2062 def check_executable(exe
, args
=[]):
2063 """ Checks if the given binary is installed somewhere in PATH, and returns its name.
2064 args can be a list of arguments for a short output (like -version) """
2066 Popen
.run([exe
] + args
, stdout
=subprocess
.PIPE
, stderr
=subprocess
.PIPE
)
2072 def _get_exe_version_output(exe
, args
):
2074 # STDIN should be redirected too. On UNIX-like systems, ffmpeg triggers
2075 # SIGTTOU if yt-dlp is run in the background.
2076 # See https://github.com/ytdl-org/youtube-dl/issues/955#issuecomment-209789656
2077 stdout
, _
, ret
= Popen
.run([encodeArgument(exe
)] + args
, text
=True,
2078 stdin
=subprocess
.PIPE
, stdout
=subprocess
.PIPE
, stderr
=subprocess
.STDOUT
)
2086 def detect_exe_version(output
, version_re
=None, unrecognized
='present'):
2087 assert isinstance(output
, str)
2088 if version_re
is None:
2089 version_re
= r
'version\s+([-0-9._a-zA-Z]+)'
2090 m
= re
.search(version_re
, output
)
2097 def get_exe_version(exe
, args
=['--version'],
2098 version_re
=None, unrecognized
=('present', 'broken')):
2099 """ Returns the version of the specified executable,
2100 or False if the executable is not present """
2101 unrecognized
= variadic(unrecognized
)
2102 assert len(unrecognized
) in (1, 2)
2103 out
= _get_exe_version_output(exe
, args
)
2105 return unrecognized
[-1]
2106 return out
and detect_exe_version(out
, version_re
, unrecognized
[0])
2109 def frange(start
=0, stop
=None, step
=1):
2112 start
, stop
= 0, start
2113 sign
= [-1, 1][step
> 0] if step
else 0
2114 while sign
* start
< sign
* stop
:
2119 class LazyList(collections
.abc
.Sequence
):
2120 """Lazy immutable list from an iterable
2121 Note that slices of a LazyList are lists and not LazyList"""
2123 class IndexError(IndexError):
2126 def __init__(self
, iterable
, *, reverse
=False, _cache
=None):
2127 self
._iterable
= iter(iterable
)
2128 self
._cache
= [] if _cache
is None else _cache
2129 self
._reversed
= reverse
2133 # We need to consume the entire iterable to iterate in reverse
2134 yield from self
.exhaust()
2136 yield from self
._cache
2137 for item
in self
._iterable
:
2138 self
._cache
.append(item
)
2142 self
._cache
.extend(self
._iterable
)
2143 self
._iterable
= [] # Discard the emptied iterable to make it pickle-able
2147 """Evaluate the entire iterable"""
2148 return self
._exhaust
()[::-1 if self
._reversed
else 1]
2151 def _reverse_index(x
):
2152 return None if x
is None else ~x
2154 def __getitem__(self
, idx
):
2155 if isinstance(idx
, slice):
2157 idx
= slice(self
._reverse
_index
(idx
.start
), self
._reverse
_index
(idx
.stop
), -(idx
.step
or 1))
2158 start
, stop
, step
= idx
.start
, idx
.stop
, idx
.step
or 1
2159 elif isinstance(idx
, int):
2161 idx
= self
._reverse
_index
(idx
)
2162 start
, stop
, step
= idx
, idx
, 0
2164 raise TypeError('indices must be integers or slices')
2165 if ((start
or 0) < 0 or (stop
or 0) < 0
2166 or (start
is None and step
< 0)
2167 or (stop
is None and step
> 0)):
2168 # We need to consume the entire iterable to be able to slice from the end
2169 # Obviously, never use this with infinite iterables
2172 return self
._cache
[idx
]
2173 except IndexError as e
:
2174 raise self
.IndexError(e
) from e
2175 n
= max(start
or 0, stop
or 0) - len(self
._cache
) + 1
2177 self
._cache
.extend(itertools
.islice(self
._iterable
, n
))
2179 return self
._cache
[idx
]
2180 except IndexError as e
:
2181 raise self
.IndexError(e
) from e
2185 self
[-1] if self
._reversed
else self
[0]
2186 except self
.IndexError:
2192 return len(self
._cache
)
2194 def __reversed__(self
):
2195 return type(self
)(self
._iterable
, reverse
=not self
._reversed
, _cache
=self
._cache
)
2198 return type(self
)(self
._iterable
, reverse
=self
._reversed
, _cache
=self
._cache
)
2201 # repr and str should mimic a list. So we exhaust the iterable
2202 return repr(self
.exhaust())
2205 return repr(self
.exhaust())
2210 class IndexError(IndexError):
2214 # This is only useful for tests
2215 return len(self
.getslice())
2217 def __init__(self
, pagefunc
, pagesize
, use_cache
=True):
2218 self
._pagefunc
= pagefunc
2219 self
._pagesize
= pagesize
2220 self
._pagecount
= float('inf')
2221 self
._use
_cache
= use_cache
2224 def getpage(self
, pagenum
):
2225 page_results
= self
._cache
.get(pagenum
)
2226 if page_results
is None:
2227 page_results
= [] if pagenum
> self
._pagecount
else list(self
._pagefunc
(pagenum
))
2229 self
._cache
[pagenum
] = page_results
2232 def getslice(self
, start
=0, end
=None):
2233 return list(self
._getslice
(start
, end
))
2235 def _getslice(self
, start
, end
):
2236 raise NotImplementedError('This method must be implemented by subclasses')
2238 def __getitem__(self
, idx
):
2239 assert self
._use
_cache
, 'Indexing PagedList requires cache'
2240 if not isinstance(idx
, int) or idx
< 0:
2241 raise TypeError('indices must be non-negative integers')
2242 entries
= self
.getslice(idx
, idx
+ 1)
2244 raise self
.IndexError()
2248 class OnDemandPagedList(PagedList
):
2249 """Download pages until a page with less than maximum results"""
2251 def _getslice(self
, start
, end
):
2252 for pagenum
in itertools
.count(start
// self
._pagesize
):
2253 firstid
= pagenum
* self
._pagesize
2254 nextfirstid
= pagenum
* self
._pagesize
+ self
._pagesize
2255 if start
>= nextfirstid
:
2259 start
% self
._pagesize
2260 if firstid
<= start
< nextfirstid
2263 ((end
- 1) % self
._pagesize
) + 1
2264 if (end
is not None and firstid
<= end
<= nextfirstid
)
2268 page_results
= self
.getpage(pagenum
)
2270 self
._pagecount
= pagenum
- 1
2272 if startv
!= 0 or endv
is not None:
2273 page_results
= page_results
[startv
:endv
]
2274 yield from page_results
2276 # A little optimization - if current page is not "full", ie. does
2277 # not contain page_size videos then we can assume that this page
2278 # is the last one - there are no more ids on further pages -
2279 # i.e. no need to query again.
2280 if len(page_results
) + startv
< self
._pagesize
:
2283 # If we got the whole page, but the next page is not interesting,
2284 # break out early as well
2285 if end
== nextfirstid
:
2289 class InAdvancePagedList(PagedList
):
2290 """PagedList with total number of pages known in advance"""
2292 def __init__(self
, pagefunc
, pagecount
, pagesize
):
2293 PagedList
.__init
__(self
, pagefunc
, pagesize
, True)
2294 self
._pagecount
= pagecount
2296 def _getslice(self
, start
, end
):
2297 start_page
= start
// self
._pagesize
2298 end_page
= self
._pagecount
if end
is None else min(self
._pagecount
, end
// self
._pagesize
+ 1)
2299 skip_elems
= start
- start_page
* self
._pagesize
2300 only_more
= None if end
is None else end
- start
2301 for pagenum
in range(start_page
, end_page
):
2302 page_results
= self
.getpage(pagenum
)
2304 page_results
= page_results
[skip_elems
:]
2306 if only_more
is not None:
2307 if len(page_results
) < only_more
:
2308 only_more
-= len(page_results
)
2310 yield from page_results
[:only_more
]
2312 yield from page_results
2315 class PlaylistEntries
:
2316 MissingEntry
= object()
2317 is_exhausted
= False
2319 def __init__(self
, ydl
, info_dict
):
2322 # _entries must be assigned now since infodict can change during iteration
2323 entries
= info_dict
.get('entries')
2325 raise EntryNotInPlaylist('There are no entries')
2326 elif isinstance(entries
, list):
2327 self
.is_exhausted
= True
2329 requested_entries
= info_dict
.get('requested_entries')
2330 self
.is_incomplete
= requested_entries
is not None
2331 if self
.is_incomplete
:
2332 assert self
.is_exhausted
2333 self
._entries
= [self
.MissingEntry
] * max(requested_entries
or [0])
2334 for i
, entry
in zip(requested_entries
, entries
):
2335 self
._entries
[i
- 1] = entry
2336 elif isinstance(entries
, (list, PagedList
, LazyList
)):
2337 self
._entries
= entries
2339 self
._entries
= LazyList(entries
)
2341 PLAYLIST_ITEMS_RE
= re
.compile(r
'''(?x)
2342 (?P<start>[+-]?\d+)?
2344 (?P<end>[+-]?\d+|inf(?:inite)?)?
2345 (?::(?P<step>[+-]?\d+))?
2349 def parse_playlist_items(cls
, string
):
2350 for segment
in string
.split(','):
2352 raise ValueError('There is two or more consecutive commas')
2353 mobj
= cls
.PLAYLIST_ITEMS_RE
.fullmatch(segment
)
2355 raise ValueError(f
'{segment!r} is not a valid specification')
2356 start
, end
, step
, has_range
= mobj
.group('start', 'end', 'step', 'range')
2357 if int_or_none(step
) == 0:
2358 raise ValueError(f
'Step in {segment!r} cannot be zero')
2359 yield slice(int_or_none(start
), float_or_none(end
), int_or_none(step
)) if has_range
else int(start
)
2361 def get_requested_items(self
):
2362 playlist_items
= self
.ydl
.params
.get('playlist_items')
2363 playlist_start
= self
.ydl
.params
.get('playliststart', 1)
2364 playlist_end
= self
.ydl
.params
.get('playlistend')
2365 # For backwards compatibility, interpret -1 as whole list
2366 if playlist_end
in (-1, None):
2368 if not playlist_items
:
2369 playlist_items
= f
'{playlist_start}:{playlist_end}'
2370 elif playlist_start
!= 1 or playlist_end
:
2371 self
.ydl
.report_warning('Ignoring playliststart and playlistend because playlistitems was given', only_once
=True)
2373 for index
in self
.parse_playlist_items(playlist_items
):
2374 for i
, entry
in self
[index
]:
2379 # The item may have just been added to archive. Don't break due to it
2380 if not self
.ydl
.params
.get('lazy_playlist'):
2381 # TODO: Add auto-generated fields
2382 self
.ydl
._match
_entry
(entry
, incomplete
=True, silent
=True)
2383 except (ExistingVideoReached
, RejectedVideoReached
):
2386 def get_full_count(self
):
2387 if self
.is_exhausted
and not self
.is_incomplete
:
2389 elif isinstance(self
._entries
, InAdvancePagedList
):
2390 if self
._entries
._pagesize
== 1:
2391 return self
._entries
._pagecount
2393 @functools.cached_property
2395 if isinstance(self
._entries
, list):
2398 entry
= self
._entries
[i
]
2400 entry
= self
.MissingEntry
2401 if not self
.is_incomplete
:
2402 raise self
.IndexError()
2403 if entry
is self
.MissingEntry
:
2404 raise EntryNotInPlaylist(f
'Entry {i + 1} cannot be found')
2409 return type(self
.ydl
)._handle
_extraction
_exceptions
(lambda _
, i
: self
._entries
[i
])(self
.ydl
, i
)
2410 except (LazyList
.IndexError, PagedList
.IndexError):
2411 raise self
.IndexError()
2414 def __getitem__(self
, idx
):
2415 if isinstance(idx
, int):
2416 idx
= slice(idx
, idx
)
2418 # NB: PlaylistEntries[1:10] => (0, 1, ... 9)
2419 step
= 1 if idx
.step
is None else idx
.step
2420 if idx
.start
is None:
2421 start
= 0 if step
> 0 else len(self
) - 1
2423 start
= idx
.start
- 1 if idx
.start
>= 0 else len(self
) + idx
.start
2425 # NB: Do not call len(self) when idx == [:]
2426 if idx
.stop
is None:
2427 stop
= 0 if step
< 0 else float('inf')
2429 stop
= idx
.stop
- 1 if idx
.stop
>= 0 else len(self
) + idx
.stop
2430 stop
+= [-1, 1][step
> 0]
2432 for i
in frange(start
, stop
, step
):
2436 entry
= self
._getter
(i
)
2437 except self
.IndexError:
2438 self
.is_exhausted
= True
2445 return len(tuple(self
[:]))
2447 class IndexError(IndexError):
2451 def uppercase_escape(s
):
2452 unicode_escape
= codecs
.getdecoder('unicode_escape')
2454 r
'\\U[0-9a-fA-F]{8}',
2455 lambda m
: unicode_escape(m
.group(0))[0],
2459 def lowercase_escape(s
):
2460 unicode_escape
= codecs
.getdecoder('unicode_escape')
2462 r
'\\u[0-9a-fA-F]{4}',
2463 lambda m
: unicode_escape(m
.group(0))[0],
2467 def parse_qs(url
, **kwargs
):
2468 return urllib
.parse
.parse_qs(urllib
.parse
.urlparse(url
).query
, **kwargs
)
2471 def read_batch_urls(batch_fd
):
2473 if not isinstance(url
, str):
2474 url
= url
.decode('utf-8', 'replace')
2475 BOM_UTF8
= ('\xef\xbb\xbf', '\ufeff')
2476 for bom
in BOM_UTF8
:
2477 if url
.startswith(bom
):
2478 url
= url
[len(bom
):]
2480 if not url
or url
.startswith(('#', ';', ']')):
2482 # "#" cannot be stripped out since it is part of the URI
2483 # However, it can be safely stripped out if following a whitespace
2484 return re
.split(r
'\s#', url
, 1)[0].rstrip()
2486 with contextlib
.closing(batch_fd
) as fd
:
2487 return [url
for url
in map(fixup
, fd
) if url
]
2490 def urlencode_postdata(*args
, **kargs
):
2491 return urllib
.parse
.urlencode(*args
, **kargs
).encode('ascii')
2494 def update_url(url
, *, query_update
=None, **kwargs
):
2495 """Replace URL components specified by kwargs
2496 @param url str or parse url tuple
2497 @param query_update update query
2500 if isinstance(url
, str):
2501 if not kwargs
and not query_update
:
2504 url
= urllib
.parse
.urlparse(url
)
2506 assert 'query' not in kwargs
, 'query_update and query cannot be specified at the same time'
2507 kwargs
['query'] = urllib
.parse
.urlencode({
2508 **urllib
.parse
.parse_qs(url
.query
),
2511 return urllib
.parse
.urlunparse(url
._replace
(**kwargs
))
2514 def update_url_query(url
, query
):
2515 return update_url(url
, query_update
=query
)
2518 def _multipart_encode_impl(data
, boundary
):
2519 content_type
= 'multipart/form-data; boundary=%s' % boundary
2522 for k
, v
in data
.items():
2523 out
+= b
'--' + boundary
.encode('ascii') + b
'\r\n'
2524 if isinstance(k
, str):
2526 if isinstance(v
, str):
2528 # RFC 2047 requires non-ASCII field names to be encoded, while RFC 7578
2529 # suggests sending UTF-8 directly. Firefox sends UTF-8, too
2530 content
= b
'Content-Disposition: form-data; name="' + k
+ b
'"\r\n\r\n' + v
+ b
'\r\n'
2531 if boundary
.encode('ascii') in content
:
2532 raise ValueError('Boundary overlaps with data')
2535 out
+= b
'--' + boundary
.encode('ascii') + b
'--\r\n'
2537 return out
, content_type
2540 def multipart_encode(data
, boundary
=None):
2542 Encode a dict to RFC 7578-compliant form-data
2545 A dict where keys and values can be either Unicode or bytes-like
2548 If specified a Unicode object, it's used as the boundary. Otherwise
2549 a random boundary is generated.
2551 Reference: https://tools.ietf.org/html/rfc7578
2553 has_specified_boundary
= boundary
is not None
2556 if boundary
is None:
2557 boundary
= '---------------' + str(random
.randrange(0x0fffffff, 0xffffffff))
2560 out
, content_type
= _multipart_encode_impl(data
, boundary
)
2563 if has_specified_boundary
:
2567 return out
, content_type
2570 def is_iterable_like(x
, allowed_types
=collections
.abc
.Iterable
, blocked_types
=NO_DEFAULT
):
2571 if blocked_types
is NO_DEFAULT
:
2572 blocked_types
= (str, bytes, collections
.abc
.Mapping
)
2573 return isinstance(x
, allowed_types
) and not isinstance(x
, blocked_types
)
2576 def variadic(x
, allowed_types
=NO_DEFAULT
):
2577 if not isinstance(allowed_types
, (tuple, type)):
2578 deprecation_warning('allowed_types should be a tuple or a type')
2579 allowed_types
= tuple(allowed_types
)
2580 return x
if is_iterable_like(x
, blocked_types
=allowed_types
) else (x
, )
2583 def try_call(*funcs
, expected_type
=None, args
=[], kwargs
={}):
2586 val
= f(*args
, **kwargs
)
2587 except (AttributeError, KeyError, TypeError, IndexError, ValueError, ZeroDivisionError):
2590 if expected_type
is None or isinstance(val
, expected_type
):
2594 def try_get(src
, getter
, expected_type
=None):
2595 return try_call(*variadic(getter
), args
=(src
,), expected_type
=expected_type
)
2598 def filter_dict(dct
, cndn
=lambda _
, v
: v
is not None):
2599 return {k: v for k, v in dct.items() if cndn(k, v)}
2602 def merge_dicts(*dicts
):
2604 for a_dict
in dicts
:
2605 for k
, v
in a_dict
.items():
2606 if (v
is not None and k
not in merged
2607 or isinstance(v
, str) and merged
[k
] == ''):
2612 def encode_compat_str(string
, encoding
=preferredencoding(), errors
='strict'):
2613 return string
if isinstance(string
, str) else str(string
, encoding
, errors
)
2625 TV_PARENTAL_GUIDELINES
= {
2635 def parse_age_limit(s
):
2636 # isinstance(False, int) is True. So type() must be used instead
2637 if type(s
) is int: # noqa: E721
2638 return s
if 0 <= s
<= 21 else None
2639 elif not isinstance(s
, str):
2641 m
= re
.match(r
'^(?P<age>\d{1,2})\+?$', s
)
2643 return int(m
.group('age'))
2646 return US_RATINGS
[s
]
2647 m
= re
.match(r
'^TV[_-]?(%s)$' % '|'.join(k
[3:] for k
in TV_PARENTAL_GUIDELINES
), s
)
2649 return TV_PARENTAL_GUIDELINES
['TV-' + m
.group(1)]
2653 def strip_jsonp(code
):
2656 (?:window\.)?(?P<func_name>[a-zA-Z0-9_.$]*)
2657 (?:\s*&&\s*(?P=func_name))?
2658 \s*\(\s*(?P<callback_data>.*)\);?
2659 \s*?(?://[^\n]*)*$''',
2660 r
'\g<callback_data>', code
)
2663 def js_to_json(code
, vars={}, *, strict
=False):
2664 # vars is a dict of var, val pairs to substitute
2665 STRING_QUOTES
= '\'"`'
2666 STRING_RE
= '|'.join(rf
'{q}(?:\\.|[^\\{q}])*{q}' for q
in STRING_QUOTES
)
2667 COMMENT_RE
= r
'/\*(?:(?!\*/).)*?\*/|//[^\n]*\n'
2668 SKIP_RE
= fr
'\s*(?:{COMMENT_RE})?\s*'
2670 (fr
'(?s)^(0[xX][0-9a-fA-F]+){SKIP_RE}:?$', 16),
2671 (fr
'(?s)^(0+[0-7]+){SKIP_RE}:?$', 8),
2674 def process_escape(match
):
2675 JSON_PASSTHROUGH_ESCAPES
= R
'"\bfnrtu'
2676 escape
= match
.group(1) or match
.group(2)
2678 return (Rf
'\{escape}' if escape
in JSON_PASSTHROUGH_ESCAPES
2679 else R
'\u00' if escape
== 'x'
2680 else '' if escape
== '\n'
2683 def template_substitute(match
):
2684 evaluated
= js_to_json(match
.group(1), vars, strict
=strict
)
2685 if evaluated
[0] == '"':
2686 return json
.loads(evaluated
)
2691 if v
in ('true', 'false', 'null'):
2693 elif v
in ('undefined', 'void 0'):
2695 elif v
.startswith('/*') or v
.startswith('//') or v
.startswith('!') or v
== ',':
2698 if v
[0] in STRING_QUOTES
:
2699 v
= re
.sub(r
'(?s)\${([^}]+)}', template_substitute
, v
[1:-1]) if v
[0] == '`' else v
[1:-1]
2700 escaped
= re
.sub(r
'(?s)(")|\\(.)', process_escape
, v
)
2701 return f
'"{escaped}"'
2703 for regex
, base
in INTEGER_TABLE
:
2704 im
= re
.match(regex
, v
)
2706 i
= int(im
.group(1), base
)
2707 return f
'"{i}":' if v
.endswith(':') else str(i
)
2713 except json
.JSONDecodeError
:
2714 return json
.dumps(vars[v
])
2721 raise ValueError(f
'Unknown value: {v}')
2723 def create_map(mobj
):
2724 return json
.dumps(dict(json
.loads(js_to_json(mobj
.group(1) or '[]', vars=vars))))
2726 code
= re
.sub(r
'new Map\((\[.*?\])?\)', create_map
, code
)
2728 code
= re
.sub(r
'new Date\((".+")\)', r
'\g<1>', code
)
2729 code
= re
.sub(r
'new \w+\((.*?)\)', lambda m
: json
.dumps(m
.group(0)), code
)
2730 code
= re
.sub(r
'parseInt\([^\d]+(\d+)[^\d]+\)', r
'\1', code
)
2731 code
= re
.sub(r
'\(function\([^)]*\)\s*\{[^}]*\}\s*\)\s*\(\s*(["\'][^
)]*["\'])\s*\)', r'\1', code)
2733 return re.sub(rf'''(?sx)
2735 {COMMENT_RE}|,(?={SKIP_RE}[\]}}])|
2736 void\s0|(?:(?<![0-9])[eE]|[a-df-zA-DF-Z_$])[.a-zA-Z_$0-9]*|
2737 \b(?:0[xX][0-9a-fA-F]+|0+[0-7]+)(?:{SKIP_RE}:)?|
2738 [0-9]+(?={SKIP_RE}:)|
2743 def qualities(quality_ids):
2744 """ Get a numeric quality value out of a list of possible values """
2747 return quality_ids.index(qid)
2753 POSTPROCESS_WHEN = ('pre_process', 'after_filter', 'video', 'before_dl', 'post_process', 'after_move', 'after_video', 'playlist')
2757 'default': '%(title)s [%(id)s].%(ext)s',
2758 'chapter': '%(title)s - %(section_number)03d %(section_title)s [%(id)s].%(ext)s',
2764 'description': 'description',
2765 'annotation': 'annotations.xml',
2766 'infojson': 'info.json',
2769 'pl_thumbnail': None,
2770 'pl_description': 'description',
2771 'pl_infojson': 'info.json',
2774 # As of [1] format syntax is:
2775 # %[mapping_key][conversion_flags][minimum_width][.precision][length_modifier]type
2776 # 1. https://docs.python.org/2/library/stdtypes.html#string-formatting
2777 STR_FORMAT_RE_TMPL = r'''(?x)
2778 (?<!%)(?P<prefix>(?:%%)*)
2780 (?P<has_key>\((?P<key>{0})\))?
2782 (?P<conversion>[#0\-+ ]+)?
2784 (?P<precision>\.\d+)?
2785 (?P<len_mod>[hlL])? # unused in python
2786 {1} # conversion type
2791 STR_FORMAT_TYPES = 'diouxXeEfFgGcrsa'
2794 def limit_length(s, length):
2795 """ Add ellipses to overly long strings """
2800 return s[:length - len(ELLIPSES)] + ELLIPSES
2804 def version_tuple(v):
2805 return tuple(int(e) for e in re.split(r'[-.]', v))
2808 def is_outdated_version(version, limit, assume_new=True):
2810 return not assume_new
2812 return version_tuple(version) < version_tuple(limit)
2814 return not assume_new
2817 def ytdl_is_updateable():
2818 """ Returns if yt-dlp can be updated with -U """
2820 from ..update import is_non_updateable
2822 return not is_non_updateable()
2825 def args_to_str(args):
2826 # Get a short string representation for a subprocess command
2827 return ' '.join(compat_shlex_quote(a) for a in args)
2830 def error_to_str(err):
2831 return f'{type(err).__name__}: {err}'
2834 def mimetype2ext(mt, default=NO_DEFAULT):
2835 if not isinstance(mt, str):
2836 if default is not NO_DEFAULT:
2853 'x-matroska': 'mkv',
2855 'x-mp4-fragmented': 'mp4',
2860 # application (streaming playlists)
2864 'vnd.apple.mpegurl': 'm3u8',
2865 'vnd.ms-sstr+xml': 'ism',
2866 'x-mpegurl': 'm3u8',
2870 # Per RFC 3003, audio/mpeg can be .mp1, .mp2 or .mp3.
2871 # Using .mp3 as it's the most popular one
2872 'audio/mpeg': 'mp3',
2873 'audio/webm': 'webm',
2874 'audio/x-matroska': 'mka',
2875 'audio/x-mpegurl': 'm3u',
2883 'x-realaudio': 'ra',
2894 'vnd.wap.wbmp': 'wbmp',
2901 'filmstrip+json': 'fs',
2902 'smptett+xml': 'tt',
2905 'x-ms-sami': 'sami',
2914 mimetype = mt.partition(';')[0].strip().lower()
2915 _, _, subtype = mimetype.rpartition('/')
2917 ext = traversal.traverse_obj(MAP, mimetype, subtype, subtype.rsplit('+')[-1])
2920 elif default is not NO_DEFAULT:
2922 return subtype.replace('+', '.')
2925 def ext2mimetype(ext_or_url):
2928 if '.' not in ext_or_url:
2929 ext_or_url = f'file.{ext_or_url}'
2930 return mimetypes.guess_type(ext_or_url)[0]
2933 def parse_codecs(codecs_str):
2934 # http://tools.ietf.org/html/rfc6381
2937 split_codecs = list(filter(None, map(
2938 str.strip, codecs_str.strip().strip(',').split(','))))
2939 vcodec, acodec, scodec, hdr = None, None, None, None
2940 for full_codec in split_codecs:
2941 parts = re.sub(r'0+(?=\d)', '', full_codec).split('.')
2942 if parts[0] in ('avc1', 'avc2', 'avc3', 'avc4', 'vp9', 'vp8', 'hev1', 'hev2',
2943 'h263', 'h264', 'mp4v', 'hvc1', 'av1', 'theora', 'dvh1', 'dvhe'):
2947 if parts[0] in ('dvh1', 'dvhe'):
2949 elif parts[0] == 'av1' and traversal.traverse_obj(parts, 3) == '10':
2951 elif parts[:2] == ['vp9', '2']:
2953 elif parts[0] in ('flac', 'mp4a', 'opus', 'vorbis', 'mp3', 'aac', 'ac-4',
2954 'ac-3', 'ec-3', 'eac3', 'dtsc', 'dtse', 'dtsh', 'dtsl'):
2955 acodec = acodec or full_codec
2956 elif parts[0] in ('stpp', 'wvtt'):
2957 scodec = scodec or full_codec
2959 write_string(f'WARNING: Unknown codec {full_codec}\n')
2960 if vcodec or acodec or scodec:
2962 'vcodec': vcodec or 'none',
2963 'acodec': acodec or 'none',
2964 'dynamic_range': hdr,
2965 **({'scodec': scodec} if scodec is not None else {}),
2967 elif len(split_codecs) == 2:
2969 'vcodec': split_codecs[0],
2970 'acodec': split_codecs[1],
2975 def get_compatible_ext(*, vcodecs, acodecs, vexts, aexts, preferences=None):
2976 assert len(vcodecs) == len(vexts) and len(acodecs) == len(aexts)
2978 allow_mkv = not preferences or 'mkv' in preferences
2980 if allow_mkv and max(len(acodecs), len(vcodecs)) > 1:
2981 return 'mkv' # TODO: any other format allows this?
2983 # TODO: All codecs supported by parse_codecs isn't handled here
2984 COMPATIBLE_CODECS = {
2986 'av1', 'hevc', 'avc1', 'mp4a', 'ac-4', # fourcc (m3u8, mpd)
2987 'h264', 'aacl', 'ec-3', # Set in ISM
2990 'av1', 'vp9', 'vp8', 'opus', 'vrbs',
2991 'vp9x', 'vp8x', # in the webm spec
2995 sanitize_codec = functools.partial(
2996 try_get, getter=lambda x: x[0].split('.')[0].replace('0', '').lower())
2997 vcodec, acodec = sanitize_codec(vcodecs), sanitize_codec(acodecs)
2999 for ext in preferences or COMPATIBLE_CODECS.keys():
3000 codec_set = COMPATIBLE_CODECS.get(ext, set())
3001 if ext == 'mkv' or codec_set.issuperset((vcodec, acodec)):
3005 {'mp3', 'mp4', 'm4a', 'm4p', 'm4b', 'm4r', 'm4v', 'ismv', 'isma', 'mov'},
3008 for ext in preferences or vexts:
3009 current_exts = {ext, *vexts, *aexts}
3010 if ext == 'mkv' or current_exts == {ext} or any(
3011 ext_sets.issuperset(current_exts) for ext_sets in COMPATIBLE_EXTS):
3013 return 'mkv' if allow_mkv else preferences[-1]
3016 def urlhandle_detect_ext(url_handle, default=NO_DEFAULT):
3017 getheader = url_handle.headers.get
3019 cd = getheader('Content-Disposition')
3021 m = re.match(r'attachment;\s*filename="(?P
<filename
>[^
"]+)"', cd)
3023 e = determine_ext(m.group('filename
'), default_ext=None)
3027 meta_ext = getheader('x
-amz
-meta
-name
')
3029 e = meta_ext.rpartition('.')[2]
3033 return mimetype2ext(getheader('Content
-Type
'), default=default)
3036 def encode_data_uri(data, mime_type):
3037 return 'data
:%s;base64
,%s' % (mime_type, base64.b64encode(data).decode('ascii
'))
3040 def age_restricted(content_limit, age_limit):
3041 """ Returns True iff the content should be blocked """
3043 if age_limit is None: # No limit set
3045 if content_limit is None:
3046 return False # Content available for everyone
3047 return age_limit < content_limit
3050 # List of known byte-order-marks (BOM)
3052 (b'\xef\xbb\xbf', 'utf
-8'),
3053 (b'\x00\x00\xfe\xff', 'utf
-32-be
'),
3054 (b'\xff\xfe\x00\x00', 'utf
-32-le
'),
3055 (b'\xff\xfe', 'utf
-16-le
'),
3056 (b'\xfe\xff', 'utf
-16-be
'),
3060 def is_html(first_bytes):
3061 """ Detect whether a file contains HTML by examining its first bytes. """
3064 for bom, enc in BOMS:
3065 while first_bytes.startswith(bom):
3066 encoding, first_bytes = enc, first_bytes[len(bom):]
3068 return re.match(r'^\s
*<', first_bytes.decode(encoding, 'replace
'))
3071 def determine_protocol(info_dict):
3072 protocol = info_dict.get('protocol
')
3073 if protocol is not None:
3076 url = sanitize_url(info_dict['url
'])
3077 if url.startswith('rtmp
'):
3079 elif url.startswith('mms
'):
3081 elif url.startswith('rtsp
'):
3084 ext = determine_ext(url)
3086 return 'm3u8
' if info_dict.get('is_live
') else 'm3u8_native
'
3090 return urllib.parse.urlparse(url).scheme
3093 def render_table(header_row, data, delim=False, extra_gap=0, hide_empty=False):
3094 """ Render a list of rows, each as a list of values.
3095 Text after a \t will be right aligned """
3097 return len(remove_terminal_sequences(string).replace('\t', ''))
3099 def get_max_lens(table):
3100 return [max(width(str(v)) for v in col) for col in zip(*table)]
3102 def filter_using_list(row, filterArray):
3103 return [col for take, col in itertools.zip_longest(filterArray, row, fillvalue=True) if take]
3105 max_lens = get_max_lens(data) if hide_empty else []
3106 header_row = filter_using_list(header_row, max_lens)
3107 data = [filter_using_list(row, max_lens) for row in data]
3109 table = [header_row] + data
3110 max_lens = get_max_lens(table)
3113 table = [header_row, [delim * (ml + extra_gap) for ml in max_lens]] + data
3114 table[1][-1] = table[1][-1][:-extra_gap * len(delim)] # Remove extra_gap from end of delimiter
3116 for pos, text in enumerate(map(str, row)):
3118 row[pos] = text.replace('\t', ' ' * (max_lens[pos] - width(text))) + ' ' * extra_gap
3120 row[pos] = text + ' ' * (max_lens[pos] - width(text) + extra_gap)
3121 ret = '\n'.join(''.join(row).rstrip() for row in table)
3125 def _match_one(filter_part, dct, incomplete):
3126 # TODO: Generalize code with YoutubeDL._build_format_filter
3127 STRING_OPERATORS = {
3128 '*=': operator.contains,
3129 '^
=': lambda attr, value: attr.startswith(value),
3130 '$
=': lambda attr, value: attr.endswith(value),
3131 '~
=': lambda attr, value: re.search(value, attr),
3133 COMPARISON_OPERATORS = {
3135 '<=': operator.le, # "<=" must be defined above "<"
3142 if isinstance(incomplete, bool):
3143 is_incomplete = lambda _: incomplete
3145 is_incomplete = lambda k: k in incomplete
3147 operator_rex = re.compile(r'''(?x)
3149 \s*(?P<negation>!\s*)?(?P<op>%s)(?P<none_inclusive>\s*\?)?\s*
3151 (?P<quote>["\'])(?P<quotedstrval>.+?)(?P=quote)|
3154 ''' % '|
'.join(map(re.escape, COMPARISON_OPERATORS.keys())))
3155 m = operator_rex.fullmatch(filter_part.strip())
3158 unnegated_op = COMPARISON_OPERATORS[m['op
']]
3160 op = lambda attr, value: not unnegated_op(attr, value)
3163 comparison_value = m['quotedstrval
'] or m['strval
'] or m['intval
']
3165 comparison_value = comparison_value.replace(r'\
%s' % m['quote
'], m['quote
'])
3166 actual_value = dct.get(m['key
'])
3167 numeric_comparison = None
3168 if isinstance(actual_value, (int, float)):
3169 # If the original field is a string and matching comparisonvalue is
3170 # a number we should respect the origin of the original field
3171 # and process comparison value as a string (see
3172 # https://github.com/ytdl-org/youtube-dl/issues/11082)
3174 numeric_comparison = int(comparison_value)
3176 numeric_comparison = parse_filesize(comparison_value)
3177 if numeric_comparison is None:
3178 numeric_comparison = parse_filesize(f'{comparison_value}B
')
3179 if numeric_comparison is None:
3180 numeric_comparison = parse_duration(comparison_value)
3181 if numeric_comparison is not None and m['op
'] in STRING_OPERATORS:
3182 raise ValueError('Operator
%s only supports string values
!' % m['op
'])
3183 if actual_value is None:
3184 return is_incomplete(m['key
']) or m['none_inclusive
']
3185 return op(actual_value, comparison_value if numeric_comparison is None else numeric_comparison)
3188 '': lambda v: (v is True) if isinstance(v, bool) else (v is not None),
3189 '!': lambda v: (v is False) if isinstance(v, bool) else (v is None),
3191 operator_rex = re.compile(r'''(?x)
3192 (?P<op>%s)\s*(?P<key>[a-z_]+)
3193 ''' % '|
'.join(map(re.escape, UNARY_OPERATORS.keys())))
3194 m = operator_rex.fullmatch(filter_part.strip())
3196 op = UNARY_OPERATORS[m.group('op
')]
3197 actual_value = dct.get(m.group('key
'))
3198 if is_incomplete(m.group('key
')) and actual_value is None:
3200 return op(actual_value)
3202 raise ValueError('Invalid
filter part
%r' % filter_part)
3205 def match_str(filter_str, dct, incomplete=False):
3206 """ Filter a dictionary with a simple string syntax.
3207 @returns Whether the filter passes
3208 @param incomplete Set of keys that is expected to be missing from dct.
3209 Can be True/False to indicate all/none of the keys may be missing.
3210 All conditions on incomplete keys pass if the key is missing
3213 _match_one(filter_part.replace(r'\
&', '&'), dct, incomplete)
3214 for filter_part in re.split(r'(?
<!\\)&', filter_str))
3217 def match_filter_func(filters, breaking_filters=None):
3218 if not filters and not breaking_filters:
3220 breaking_filters = match_filter_func(breaking_filters) or (lambda _, __: None)
3221 filters = set(variadic(filters or []))
3223 interactive = '-' in filters
3227 def _match_func(info_dict, incomplete=False):
3228 ret = breaking_filters(info_dict, incomplete)
3230 raise RejectedVideoReached(ret)
3232 if not filters or any(match_str(f, info_dict, incomplete) for f in filters):
3233 return NO_DEFAULT if interactive and not incomplete else None
3235 video_title = info_dict.get('title
') or info_dict.get('id') or 'entry
'
3236 filter_str = ') |
('.join(map(str.strip, filters))
3237 return f'{video_title} does
not pass filter ({filter_str}
), skipping
..'
3241 class download_range_func:
3242 def __init__(self, chapters, ranges, from_info=False):
3243 self.chapters, self.ranges, self.from_info = chapters, ranges, from_info
3245 def __call__(self, info_dict, ydl):
3247 warning = ('There are no chapters matching the regex
' if info_dict.get('chapters
')
3248 else 'Cannot match chapters since chapter information
is unavailable
')
3249 for regex in self.chapters or []:
3250 for i, chapter in enumerate(info_dict.get('chapters
') or []):
3251 if re.search(regex, chapter['title
']):
3253 yield {**chapter, 'index': i}
3254 if self.chapters and warning:
3255 ydl.to_screen(f'[info
] {info_dict["id"]}
: {warning}
')
3257 for start, end in self.ranges or []:
3259 'start_time
': self._handle_negative_timestamp(start, info_dict),
3260 'end_time
': self._handle_negative_timestamp(end, info_dict),
3263 if self.from_info and (info_dict.get('start_time
') or info_dict.get('end_time
')):
3265 'start_time
': info_dict.get('start_time
') or 0,
3266 'end_time
': info_dict.get('end_time
') or float('inf
'),
3268 elif not self.ranges and not self.chapters:
3272 def _handle_negative_timestamp(time, info):
3273 return max(info['duration
'] + time, 0) if info.get('duration
') and time < 0 else time
3275 def __eq__(self, other):
3276 return (isinstance(other, download_range_func)
3277 and self.chapters == other.chapters and self.ranges == other.ranges)
3280 return f'{__name__}
.{type(self).__name__}
({self.chapters}
, {self.ranges}
)'
3283 def parse_dfxp_time_expr(time_expr):
3287 mobj = re.match(rf'^
(?P
<time_offset
>{NUMBER_RE}
)s?$
', time_expr)
3289 return float(mobj.group('time_offset
'))
3291 mobj = re.match(r'^
(\d
+):(\d\d
):(\d\
d(?
:(?
:\
.|
:)\d
+)?
)$
', time_expr)
3293 return 3600 * int(mobj.group(1)) + 60 * int(mobj.group(2)) + float(mobj.group(3).replace(':', '.'))
3296 def srt_subtitles_timecode(seconds):
3297 return '%02d
:%02d
:%02d
,%03d
' % timetuple_from_msec(seconds * 1000)
3300 def ass_subtitles_timecode(seconds):
3301 time = timetuple_from_msec(seconds * 1000)
3302 return '%01d
:%02d
:%02d
.%02d
' % (*time[:-1], time.milliseconds / 10)
3305 def dfxp2srt(dfxp_data):
3307 @param dfxp_data A bytes-like object containing DFXP data
3308 @returns A unicode object containing converted SRT data
3310 LEGACY_NAMESPACES = (
3311 (b'http
://www
.w3
.org
/ns
/ttml
', [
3312 b'http
://www
.w3
.org
/2004/11/ttaf1
',
3313 b'http
://www
.w3
.org
/2006/04/ttaf1
',
3314 b'http
://www
.w3
.org
/2006/10/ttaf1
',
3316 (b'http
://www
.w3
.org
/ns
/ttml
#styling', [
3317 b
'http://www.w3.org/ns/ttml#style',
3321 SUPPORTED_STYLING
= [
3330 _x
= functools
.partial(xpath_with_ns
, ns_map
={
3331 'xml': 'http://www.w3.org/XML/1998/namespace',
3332 'ttml': 'http://www.w3.org/ns/ttml',
3333 'tts': 'http://www.w3.org/ns/ttml#styling',
3339 class TTMLPElementParser
:
3341 _unclosed_elements
= []
3342 _applied_styles
= []
3344 def start(self
, tag
, attrib
):
3345 if tag
in (_x('ttml:br'), 'br'):
3348 unclosed_elements
= []
3350 element_style_id
= attrib
.get('style')
3352 style
.update(default_style
)
3353 if element_style_id
:
3354 style
.update(styles
.get(element_style_id
, {}))
3355 for prop
in SUPPORTED_STYLING
:
3356 prop_val
= attrib
.get(_x('tts:' + prop
))
3358 style
[prop
] = prop_val
3361 for k
, v
in sorted(style
.items()):
3362 if self
._applied
_styles
and self
._applied
_styles
[-1].get(k
) == v
:
3365 font
+= ' color="%s"' % v
3366 elif k
== 'fontSize':
3367 font
+= ' size="%s"' % v
3368 elif k
== 'fontFamily':
3369 font
+= ' face="%s"' % v
3370 elif k
== 'fontWeight' and v
== 'bold':
3372 unclosed_elements
.append('b')
3373 elif k
== 'fontStyle' and v
== 'italic':
3375 unclosed_elements
.append('i')
3376 elif k
== 'textDecoration' and v
== 'underline':
3378 unclosed_elements
.append('u')
3380 self
._out
+= '<font' + font
+ '>'
3381 unclosed_elements
.append('font')
3383 if self
._applied
_styles
:
3384 applied_style
.update(self
._applied
_styles
[-1])
3385 applied_style
.update(style
)
3386 self
._applied
_styles
.append(applied_style
)
3387 self
._unclosed
_elements
.append(unclosed_elements
)
3390 if tag
not in (_x('ttml:br'), 'br'):
3391 unclosed_elements
= self
._unclosed
_elements
.pop()
3392 for element
in reversed(unclosed_elements
):
3393 self
._out
+= '</%s>' % element
3394 if unclosed_elements
and self
._applied
_styles
:
3395 self
._applied
_styles
.pop()
3397 def data(self
, data
):
3401 return self
._out
.strip()
3403 # Fix UTF-8 encoded file wrongly marked as UTF-16. See https://github.com/yt-dlp/yt-dlp/issues/6543#issuecomment-1477169870
3404 # This will not trigger false positives since only UTF-8 text is being replaced
3405 dfxp_data
= dfxp_data
.replace(b
'encoding=\'UTF-16\'', b
'encoding=\'UTF-8\'')
3407 def parse_node(node
):
3408 target
= TTMLPElementParser()
3409 parser
= xml
.etree
.ElementTree
.XMLParser(target
=target
)
3410 parser
.feed(xml
.etree
.ElementTree
.tostring(node
))
3411 return parser
.close()
3413 for k
, v
in LEGACY_NAMESPACES
:
3415 dfxp_data
= dfxp_data
.replace(ns
, k
)
3417 dfxp
= compat_etree_fromstring(dfxp_data
)
3419 paras
= dfxp
.findall(_x('.//ttml:p')) or dfxp
.findall('.//p')
3422 raise ValueError('Invalid dfxp/TTML subtitle')
3426 for style
in dfxp
.findall(_x('.//ttml:style')):
3427 style_id
= style
.get('id') or style
.get(_x('xml:id'))
3430 parent_style_id
= style
.get('style')
3432 if parent_style_id
not in styles
:
3435 styles
[style_id
] = styles
[parent_style_id
].copy()
3436 for prop
in SUPPORTED_STYLING
:
3437 prop_val
= style
.get(_x('tts:' + prop
))
3439 styles
.setdefault(style_id
, {})[prop
] = prop_val
3445 for p
in ('body', 'div'):
3446 ele
= xpath_element(dfxp
, [_x('.//ttml:' + p
), './/' + p
])
3449 style
= styles
.get(ele
.get('style'))
3452 default_style
.update(style
)
3454 for para
, index
in zip(paras
, itertools
.count(1)):
3455 begin_time
= parse_dfxp_time_expr(para
.attrib
.get('begin'))
3456 end_time
= parse_dfxp_time_expr(para
.attrib
.get('end'))
3457 dur
= parse_dfxp_time_expr(para
.attrib
.get('dur'))
3458 if begin_time
is None:
3463 end_time
= begin_time
+ dur
3464 out
.append('%d\n%s --> %s\n%s\n\n' % (
3466 srt_subtitles_timecode(begin_time
),
3467 srt_subtitles_timecode(end_time
),
3473 def cli_option(params
, command_option
, param
, separator
=None):
3474 param
= params
.get(param
)
3475 return ([] if param
is None
3476 else [command_option
, str(param
)] if separator
is None
3477 else [f
'{command_option}{separator}{param}'])
3480 def cli_bool_option(params
, command_option
, param
, true_value
='true', false_value
='false', separator
=None):
3481 param
= params
.get(param
)
3482 assert param
in (True, False, None)
3483 return cli_option({True: true_value, False: false_value}
, command_option
, param
, separator
)
3486 def cli_valueless_option(params
, command_option
, param
, expected_value
=True):
3487 return [command_option
] if params
.get(param
) == expected_value
else []
3490 def cli_configuration_args(argdict
, keys
, default
=[], use_compat
=True):
3491 if isinstance(argdict
, (list, tuple)): # for backward compatibility
3498 assert isinstance(argdict
, dict)
3500 assert isinstance(keys
, (list, tuple))
3501 for key_list
in keys
:
3502 arg_list
= list(filter(
3503 lambda x
: x
is not None,
3504 [argdict
.get(key
.lower()) for key
in variadic(key_list
)]))
3506 return [arg
for args
in arg_list
for arg
in args
]
3510 def _configuration_args(main_key
, argdict
, exe
, keys
=None, default
=[], use_compat
=True):
3511 main_key
, exe
= main_key
.lower(), exe
.lower()
3512 root_key
= exe
if main_key
== exe
else f
'{main_key}+{exe}'
3513 keys
= [f
'{root_key}{k}' for k
in (keys
or [''])]
3514 if root_key
in keys
:
3516 keys
.append((main_key
, exe
))
3517 keys
.append('default')
3520 return cli_configuration_args(argdict
, keys
, default
, use_compat
)
3524 # See http://www.loc.gov/standards/iso639-2/ISO-639-2_utf-8.txt
3583 'iw': 'heb', # Replaced by he in 1989 revision
3593 'in': 'ind', # Replaced by id in 1989 revision
3709 'ji': 'yid', # Replaced by yi in 1989 revision
3717 def short2long(cls
, code
):
3718 """Convert language code from ISO 639-1 to ISO 639-2/T"""
3719 return cls
._lang
_map
.get(code
[:2])
3722 def long2short(cls
, code
):
3723 """Convert language code from ISO 639-2/T to ISO 639-1"""
3724 for short_name
, long_name
in cls
._lang
_map
.items():
3725 if long_name
== code
:
3730 # From http://data.okfn.org/data/core/country-list
3732 'AF': 'Afghanistan',
3733 'AX': 'Åland Islands',
3736 'AS': 'American Samoa',
3741 'AG': 'Antigua and Barbuda',
3758 'BO': 'Bolivia, Plurinational State of',
3759 'BQ': 'Bonaire, Sint Eustatius and Saba',
3760 'BA': 'Bosnia and Herzegovina',
3762 'BV': 'Bouvet Island',
3764 'IO': 'British Indian Ocean Territory',
3765 'BN': 'Brunei Darussalam',
3767 'BF': 'Burkina Faso',
3773 'KY': 'Cayman Islands',
3774 'CF': 'Central African Republic',
3778 'CX': 'Christmas Island',
3779 'CC': 'Cocos (Keeling) Islands',
3783 'CD': 'Congo, the Democratic Republic of the',
3784 'CK': 'Cook Islands',
3786 'CI': 'Côte d\'Ivoire',
3791 'CZ': 'Czech Republic',
3795 'DO': 'Dominican Republic',
3798 'SV': 'El Salvador',
3799 'GQ': 'Equatorial Guinea',
3803 'FK': 'Falkland Islands (Malvinas)',
3804 'FO': 'Faroe Islands',
3808 'GF': 'French Guiana',
3809 'PF': 'French Polynesia',
3810 'TF': 'French Southern Territories',
3825 'GW': 'Guinea-Bissau',
3828 'HM': 'Heard Island and McDonald Islands',
3829 'VA': 'Holy See (Vatican City State)',
3836 'IR': 'Iran, Islamic Republic of',
3839 'IM': 'Isle of Man',
3849 'KP': 'Korea, Democratic People\'s Republic of',
3850 'KR': 'Korea, Republic of',
3853 'LA': 'Lao People\'s Democratic Republic',
3859 'LI': 'Liechtenstein',
3863 'MK': 'Macedonia, the Former Yugoslav Republic of',
3870 'MH': 'Marshall Islands',
3876 'FM': 'Micronesia, Federated States of',
3877 'MD': 'Moldova, Republic of',
3888 'NL': 'Netherlands',
3889 'NC': 'New Caledonia',
3890 'NZ': 'New Zealand',
3895 'NF': 'Norfolk Island',
3896 'MP': 'Northern Mariana Islands',
3901 'PS': 'Palestine, State of',
3903 'PG': 'Papua New Guinea',
3906 'PH': 'Philippines',
3910 'PR': 'Puerto Rico',
3914 'RU': 'Russian Federation',
3916 'BL': 'Saint Barthélemy',
3917 'SH': 'Saint Helena, Ascension and Tristan da Cunha',
3918 'KN': 'Saint Kitts and Nevis',
3919 'LC': 'Saint Lucia',
3920 'MF': 'Saint Martin (French part)',
3921 'PM': 'Saint Pierre and Miquelon',
3922 'VC': 'Saint Vincent and the Grenadines',
3925 'ST': 'Sao Tome and Principe',
3926 'SA': 'Saudi Arabia',
3930 'SL': 'Sierra Leone',
3932 'SX': 'Sint Maarten (Dutch part)',
3935 'SB': 'Solomon Islands',
3937 'ZA': 'South Africa',
3938 'GS': 'South Georgia and the South Sandwich Islands',
3939 'SS': 'South Sudan',
3944 'SJ': 'Svalbard and Jan Mayen',
3947 'CH': 'Switzerland',
3948 'SY': 'Syrian Arab Republic',
3949 'TW': 'Taiwan, Province of China',
3951 'TZ': 'Tanzania, United Republic of',
3953 'TL': 'Timor-Leste',
3957 'TT': 'Trinidad and Tobago',
3960 'TM': 'Turkmenistan',
3961 'TC': 'Turks and Caicos Islands',
3965 'AE': 'United Arab Emirates',
3966 'GB': 'United Kingdom',
3967 'US': 'United States',
3968 'UM': 'United States Minor Outlying Islands',
3972 'VE': 'Venezuela, Bolivarian Republic of',
3974 'VG': 'Virgin Islands, British',
3975 'VI': 'Virgin Islands, U.S.',
3976 'WF': 'Wallis and Futuna',
3977 'EH': 'Western Sahara',
3981 # Not ISO 3166 codes, but used for IP blocks
3982 'AP': 'Asia/Pacific Region',
3987 def short2full(cls
, code
):
3988 """Convert an ISO 3166-2 country code to the corresponding full name"""
3989 return cls
._country
_map
.get(code
.upper())
3993 # Major IPv4 address blocks per country
3995 'AD': '46.172.224.0/19',
3996 'AE': '94.200.0.0/13',
3997 'AF': '149.54.0.0/17',
3998 'AG': '209.59.64.0/18',
3999 'AI': '204.14.248.0/21',
4000 'AL': '46.99.0.0/16',
4001 'AM': '46.70.0.0/15',
4002 'AO': '105.168.0.0/13',
4003 'AP': '182.50.184.0/21',
4004 'AQ': '23.154.160.0/24',
4005 'AR': '181.0.0.0/12',
4006 'AS': '202.70.112.0/20',
4007 'AT': '77.116.0.0/14',
4008 'AU': '1.128.0.0/11',
4009 'AW': '181.41.0.0/18',
4010 'AX': '185.217.4.0/22',
4011 'AZ': '5.197.0.0/16',
4012 'BA': '31.176.128.0/17',
4013 'BB': '65.48.128.0/17',
4014 'BD': '114.130.0.0/16',
4016 'BF': '102.178.0.0/15',
4017 'BG': '95.42.0.0/15',
4018 'BH': '37.131.0.0/17',
4019 'BI': '154.117.192.0/18',
4020 'BJ': '137.255.0.0/16',
4021 'BL': '185.212.72.0/23',
4022 'BM': '196.12.64.0/18',
4023 'BN': '156.31.0.0/16',
4024 'BO': '161.56.0.0/16',
4025 'BQ': '161.0.80.0/20',
4026 'BR': '191.128.0.0/12',
4027 'BS': '24.51.64.0/18',
4028 'BT': '119.2.96.0/19',
4029 'BW': '168.167.0.0/16',
4030 'BY': '178.120.0.0/13',
4031 'BZ': '179.42.192.0/18',
4032 'CA': '99.224.0.0/11',
4033 'CD': '41.243.0.0/16',
4034 'CF': '197.242.176.0/21',
4035 'CG': '160.113.0.0/16',
4036 'CH': '85.0.0.0/13',
4037 'CI': '102.136.0.0/14',
4038 'CK': '202.65.32.0/19',
4039 'CL': '152.172.0.0/14',
4040 'CM': '102.244.0.0/14',
4041 'CN': '36.128.0.0/10',
4042 'CO': '181.240.0.0/12',
4043 'CR': '201.192.0.0/12',
4044 'CU': '152.206.0.0/15',
4045 'CV': '165.90.96.0/19',
4046 'CW': '190.88.128.0/17',
4047 'CY': '31.153.0.0/16',
4048 'CZ': '88.100.0.0/14',
4050 'DJ': '197.241.0.0/17',
4051 'DK': '87.48.0.0/12',
4052 'DM': '192.243.48.0/20',
4053 'DO': '152.166.0.0/15',
4054 'DZ': '41.96.0.0/12',
4055 'EC': '186.68.0.0/15',
4056 'EE': '90.190.0.0/15',
4057 'EG': '156.160.0.0/11',
4058 'ER': '196.200.96.0/20',
4059 'ES': '88.0.0.0/11',
4060 'ET': '196.188.0.0/14',
4061 'EU': '2.16.0.0/13',
4062 'FI': '91.152.0.0/13',
4063 'FJ': '144.120.0.0/16',
4064 'FK': '80.73.208.0/21',
4065 'FM': '119.252.112.0/20',
4066 'FO': '88.85.32.0/19',
4068 'GA': '41.158.0.0/15',
4070 'GD': '74.122.88.0/21',
4071 'GE': '31.146.0.0/16',
4072 'GF': '161.22.64.0/18',
4073 'GG': '62.68.160.0/19',
4074 'GH': '154.160.0.0/12',
4075 'GI': '95.164.0.0/16',
4076 'GL': '88.83.0.0/19',
4077 'GM': '160.182.0.0/15',
4078 'GN': '197.149.192.0/18',
4079 'GP': '104.250.0.0/19',
4080 'GQ': '105.235.224.0/20',
4081 'GR': '94.64.0.0/13',
4082 'GT': '168.234.0.0/16',
4083 'GU': '168.123.0.0/16',
4084 'GW': '197.214.80.0/20',
4085 'GY': '181.41.64.0/18',
4086 'HK': '113.252.0.0/14',
4087 'HN': '181.210.0.0/16',
4088 'HR': '93.136.0.0/13',
4089 'HT': '148.102.128.0/17',
4090 'HU': '84.0.0.0/14',
4091 'ID': '39.192.0.0/10',
4092 'IE': '87.32.0.0/12',
4093 'IL': '79.176.0.0/13',
4094 'IM': '5.62.80.0/20',
4095 'IN': '117.192.0.0/10',
4096 'IO': '203.83.48.0/21',
4097 'IQ': '37.236.0.0/14',
4098 'IR': '2.176.0.0/12',
4099 'IS': '82.221.0.0/16',
4100 'IT': '79.0.0.0/10',
4101 'JE': '87.244.64.0/18',
4102 'JM': '72.27.0.0/17',
4103 'JO': '176.29.0.0/16',
4104 'JP': '133.0.0.0/8',
4105 'KE': '105.48.0.0/12',
4106 'KG': '158.181.128.0/17',
4107 'KH': '36.37.128.0/17',
4108 'KI': '103.25.140.0/22',
4109 'KM': '197.255.224.0/20',
4110 'KN': '198.167.192.0/19',
4111 'KP': '175.45.176.0/22',
4112 'KR': '175.192.0.0/10',
4113 'KW': '37.36.0.0/14',
4114 'KY': '64.96.0.0/15',
4115 'KZ': '2.72.0.0/13',
4116 'LA': '115.84.64.0/18',
4117 'LB': '178.135.0.0/16',
4118 'LC': '24.92.144.0/20',
4119 'LI': '82.117.0.0/19',
4120 'LK': '112.134.0.0/15',
4121 'LR': '102.183.0.0/16',
4122 'LS': '129.232.0.0/17',
4123 'LT': '78.56.0.0/13',
4124 'LU': '188.42.0.0/16',
4125 'LV': '46.109.0.0/16',
4126 'LY': '41.252.0.0/14',
4127 'MA': '105.128.0.0/11',
4128 'MC': '88.209.64.0/18',
4129 'MD': '37.246.0.0/16',
4130 'ME': '178.175.0.0/17',
4131 'MF': '74.112.232.0/21',
4132 'MG': '154.126.0.0/17',
4133 'MH': '117.103.88.0/21',
4134 'MK': '77.28.0.0/15',
4135 'ML': '154.118.128.0/18',
4136 'MM': '37.111.0.0/17',
4137 'MN': '49.0.128.0/17',
4138 'MO': '60.246.0.0/16',
4139 'MP': '202.88.64.0/20',
4140 'MQ': '109.203.224.0/19',
4141 'MR': '41.188.64.0/18',
4142 'MS': '208.90.112.0/22',
4143 'MT': '46.11.0.0/16',
4144 'MU': '105.16.0.0/12',
4145 'MV': '27.114.128.0/18',
4146 'MW': '102.70.0.0/15',
4147 'MX': '187.192.0.0/11',
4148 'MY': '175.136.0.0/13',
4149 'MZ': '197.218.0.0/15',
4150 'NA': '41.182.0.0/16',
4151 'NC': '101.101.0.0/18',
4152 'NE': '197.214.0.0/18',
4153 'NF': '203.17.240.0/22',
4154 'NG': '105.112.0.0/12',
4155 'NI': '186.76.0.0/15',
4156 'NL': '145.96.0.0/11',
4157 'NO': '84.208.0.0/13',
4158 'NP': '36.252.0.0/15',
4159 'NR': '203.98.224.0/19',
4160 'NU': '49.156.48.0/22',
4161 'NZ': '49.224.0.0/14',
4162 'OM': '5.36.0.0/15',
4163 'PA': '186.72.0.0/15',
4164 'PE': '186.160.0.0/14',
4165 'PF': '123.50.64.0/18',
4166 'PG': '124.240.192.0/19',
4167 'PH': '49.144.0.0/13',
4168 'PK': '39.32.0.0/11',
4169 'PL': '83.0.0.0/11',
4170 'PM': '70.36.0.0/20',
4171 'PR': '66.50.0.0/16',
4172 'PS': '188.161.0.0/16',
4173 'PT': '85.240.0.0/13',
4174 'PW': '202.124.224.0/20',
4175 'PY': '181.120.0.0/14',
4176 'QA': '37.210.0.0/15',
4177 'RE': '102.35.0.0/16',
4178 'RO': '79.112.0.0/13',
4179 'RS': '93.86.0.0/15',
4180 'RU': '5.136.0.0/13',
4181 'RW': '41.186.0.0/16',
4182 'SA': '188.48.0.0/13',
4183 'SB': '202.1.160.0/19',
4184 'SC': '154.192.0.0/11',
4185 'SD': '102.120.0.0/13',
4186 'SE': '78.64.0.0/12',
4187 'SG': '8.128.0.0/10',
4188 'SI': '188.196.0.0/14',
4189 'SK': '78.98.0.0/15',
4190 'SL': '102.143.0.0/17',
4191 'SM': '89.186.32.0/19',
4192 'SN': '41.82.0.0/15',
4193 'SO': '154.115.192.0/18',
4194 'SR': '186.179.128.0/17',
4195 'SS': '105.235.208.0/21',
4196 'ST': '197.159.160.0/19',
4197 'SV': '168.243.0.0/16',
4198 'SX': '190.102.0.0/20',
4200 'SZ': '41.84.224.0/19',
4201 'TC': '65.255.48.0/20',
4202 'TD': '154.68.128.0/19',
4203 'TG': '196.168.0.0/14',
4204 'TH': '171.96.0.0/13',
4205 'TJ': '85.9.128.0/18',
4206 'TK': '27.96.24.0/21',
4207 'TL': '180.189.160.0/20',
4208 'TM': '95.85.96.0/19',
4209 'TN': '197.0.0.0/11',
4210 'TO': '175.176.144.0/21',
4211 'TR': '78.160.0.0/11',
4212 'TT': '186.44.0.0/15',
4213 'TV': '202.2.96.0/19',
4214 'TW': '120.96.0.0/11',
4215 'TZ': '156.156.0.0/14',
4216 'UA': '37.52.0.0/14',
4217 'UG': '102.80.0.0/13',
4219 'UY': '167.56.0.0/13',
4220 'UZ': '84.54.64.0/18',
4221 'VA': '212.77.0.0/19',
4222 'VC': '207.191.240.0/21',
4223 'VE': '186.88.0.0/13',
4224 'VG': '66.81.192.0/20',
4225 'VI': '146.226.0.0/16',
4226 'VN': '14.160.0.0/11',
4227 'VU': '202.80.32.0/20',
4228 'WF': '117.20.32.0/21',
4229 'WS': '202.4.32.0/19',
4230 'YE': '134.35.0.0/16',
4231 'YT': '41.242.116.0/22',
4232 'ZA': '41.0.0.0/11',
4233 'ZM': '102.144.0.0/13',
4234 'ZW': '102.177.192.0/18',
4238 def random_ipv4(cls
, code_or_block
):
4239 if len(code_or_block
) == 2:
4240 block
= cls
._country
_ip
_map
.get(code_or_block
.upper())
4244 block
= code_or_block
4245 addr
, preflen
= block
.split('/')
4246 addr_min
= struct
.unpack('!L', socket
.inet_aton(addr
))[0]
4247 addr_max
= addr_min |
(0xffffffff >> int(preflen
))
4248 return str(socket
.inet_ntoa(
4249 struct
.pack('!L', random
.randint(addr_min
, addr_max
))))
4252 # Both long_to_bytes and bytes_to_long are adapted from PyCrypto, which is
4253 # released into Public Domain
4254 # https://github.com/dlitz/pycrypto/blob/master/lib/Crypto/Util/number.py#L387
4256 def long_to_bytes(n
, blocksize
=0):
4257 """long_to_bytes(n:long, blocksize:int) : string
4258 Convert a long integer to a byte string.
4260 If optional blocksize is given and greater than zero, pad the front of the
4261 byte string with binary zeros so that the length is a multiple of
4264 # after much testing, this algorithm was deemed to be the fastest
4268 s
= struct
.pack('>I', n
& 0xffffffff) + s
4270 # strip off leading zeros
4271 for i
in range(len(s
)):
4272 if s
[i
] != b
'\000'[0]:
4275 # only happens when n == 0
4279 # add back some pad bytes. this could be done more efficiently w.r.t. the
4280 # de-padding being done above, but sigh...
4281 if blocksize
> 0 and len(s
) % blocksize
:
4282 s
= (blocksize
- len(s
) % blocksize
) * b
'\000' + s
4286 def bytes_to_long(s
):
4287 """bytes_to_long(string) : long
4288 Convert a byte string to a long integer.
4290 This is (essentially) the inverse of long_to_bytes().
4295 extra
= (4 - length
% 4)
4296 s
= b
'\000' * extra
+ s
4297 length
= length
+ extra
4298 for i
in range(0, length
, 4):
4299 acc
= (acc
<< 32) + struct
.unpack('>I', s
[i
:i
+ 4])[0]
4303 def ohdave_rsa_encrypt(data
, exponent
, modulus
):
4305 Implement OHDave's RSA algorithm. See http://www.ohdave.com/rsa/
4308 data: data to encrypt, bytes-like object
4309 exponent, modulus: parameter e and N of RSA algorithm, both integer
4310 Output: hex string of encrypted data
4312 Limitation: supports one block encryption only
4315 payload
= int(binascii
.hexlify(data
[::-1]), 16)
4316 encrypted
= pow(payload
, exponent
, modulus
)
4317 return '%x' % encrypted
4320 def pkcs1pad(data
, length
):
4322 Padding input data with PKCS#1 scheme
4324 @param {int[]} data input data
4325 @param {int} length target length
4326 @returns {int[]} padded data
4328 if len(data
) > length
- 11:
4329 raise ValueError('Input data too long for PKCS#1 padding')
4331 pseudo_random
= [random
.randint(0, 254) for _
in range(length
- len(data
) - 3)]
4332 return [0, 2] + pseudo_random
+ [0] + data
4335 def _base_n_table(n
, table
):
4336 if not table
and not n
:
4337 raise ValueError('Either table or n must be specified')
4338 table
= (table
or '0123456789abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ')[:n
]
4340 if n
and n
!= len(table
):
4341 raise ValueError(f
'base {n} exceeds table length {len(table)}')
4345 def encode_base_n(num
, n
=None, table
=None):
4346 """Convert given int to a base-n string"""
4347 table
= _base_n_table(n
, table
)
4351 result
, base
= '', len(table
)
4353 result
= table
[num
% base
] + result
4358 def decode_base_n(string
, n
=None, table
=None):
4359 """Convert given base-n string to int"""
4360 table
= {char: index for index, char in enumerate(_base_n_table(n, table))}
4361 result
, base
= 0, len(table
)
4363 result
= result
* base
+ table
[char
]
4367 def decode_packed_codes(code
):
4368 mobj
= re
.search(PACKED_CODES_RE
, code
)
4369 obfuscated_code
, base
, count
, symbols
= mobj
.groups()
4372 symbols
= symbols
.split('|')
4377 base_n_count
= encode_base_n(count
, base
)
4378 symbol_table
[base_n_count
] = symbols
[count
] or base_n_count
4381 r
'\b(\w+)\b', lambda mobj
: symbol_table
[mobj
.group(0)],
4385 def caesar(s
, alphabet
, shift
):
4390 alphabet
[(alphabet
.index(c
) + shift
) % l
] if c
in alphabet
else c
4395 return caesar(s
, r
'''!"#$%&'()*+,-./0123456789:;<=>?@ABCDEFGHIJKLMNOPQRSTUVWXYZ[\]^_`abcdefghijklmnopqrstuvwxyz{|}~''', 47)
4398 def parse_m3u8_attributes(attrib
):
4400 for (key
, val
) in re
.findall(r
'(?P<key>[A-Z0-9-]+)=(?P<val>"[^"]+"|[^",]+)(?:,|$)', attrib
):
4401 if val
.startswith('"'):
4407 def urshift(val
, n
):
4408 return val
>> n
if val
>= 0 else (val
+ 0x100000000) >> n
4411 def write_xattr(path
, key
, value
):
4412 # Windows: Write xattrs to NTFS Alternate Data Streams:
4413 # http://en.wikipedia.org/wiki/NTFS#Alternate_data_streams_.28ADS.29
4414 if compat_os_name
== 'nt':
4415 assert ':' not in key
4416 assert os
.path
.exists(path
)
4419 with open(f
'{path}:{key}', 'wb') as f
:
4421 except OSError as e
:
4422 raise XAttrMetadataError(e
.errno
, e
.strerror
)
4425 # UNIX Method 1. Use xattrs/pyxattrs modules
4428 if getattr(xattr
, '_yt_dlp__identifier', None) == 'pyxattr':
4429 # Unicode arguments are not supported in pyxattr until version 0.5.0
4430 # See https://github.com/ytdl-org/youtube-dl/issues/5498
4431 if version_tuple(xattr
.__version
__) >= (0, 5, 0):
4432 setxattr
= xattr
.set
4434 setxattr
= xattr
.setxattr
4438 setxattr(path
, key
, value
)
4439 except OSError as e
:
4440 raise XAttrMetadataError(e
.errno
, e
.strerror
)
4443 # UNIX Method 2. Use setfattr/xattr executables
4444 exe
= ('setfattr' if check_executable('setfattr', ['--version'])
4445 else 'xattr' if check_executable('xattr', ['-h']) else None)
4447 raise XAttrUnavailableError(
4448 'Couldn\'t find a tool to set the xattrs. Install either the python "xattr" or "pyxattr" modules or the '
4449 + ('"xattr" binary' if sys
.platform
!= 'linux' else 'GNU "attr" package (which contains the "setfattr" tool)'))
4451 value
= value
.decode()
4453 _
, stderr
, returncode
= Popen
.run(
4454 [exe
, '-w', key
, value
, path
] if exe
== 'xattr' else [exe
, '-n', key
, '-v', value
, path
],
4455 text
=True, stdout
=subprocess
.PIPE
, stderr
=subprocess
.PIPE
, stdin
=subprocess
.PIPE
)
4456 except OSError as e
:
4457 raise XAttrMetadataError(e
.errno
, e
.strerror
)
4459 raise XAttrMetadataError(returncode
, stderr
)
4462 def random_birthday(year_field
, month_field
, day_field
):
4463 start_date
= datetime
.date(1950, 1, 1)
4464 end_date
= datetime
.date(1995, 12, 31)
4465 offset
= random
.randint(0, (end_date
- start_date
).days
)
4466 random_date
= start_date
+ datetime
.timedelta(offset
)
4468 year_field
: str(random_date
.year
),
4469 month_field
: str(random_date
.month
),
4470 day_field
: str(random_date
.day
),
4474 def find_available_port(interface
=''):
4476 with socket
.socket() as sock
:
4477 sock
.bind((interface
, 0))
4478 return sock
.getsockname()[1]
4483 # Templates for internet shortcut files, which are plain text files.
4484 DOT_URL_LINK_TEMPLATE
= '''\
4489 DOT_WEBLOC_LINK_TEMPLATE
= '''\
4490 <?xml version="1.0" encoding="UTF-8"?>
4491 <!DOCTYPE plist PUBLIC "-//Apple//DTD PLIST 1.0//EN" "http://www.apple.com/DTDs/PropertyList-1.0.dtd">
4492 <plist version="1.0">
4495 \t<string>%(url)s</string>
4500 DOT_DESKTOP_LINK_TEMPLATE
= '''\
4510 'url': DOT_URL_LINK_TEMPLATE
,
4511 'desktop': DOT_DESKTOP_LINK_TEMPLATE
,
4512 'webloc': DOT_WEBLOC_LINK_TEMPLATE
,
4516 def iri_to_uri(iri
):
4518 Converts an IRI (Internationalized Resource Identifier, allowing Unicode characters) to a URI (Uniform Resource Identifier, ASCII-only).
4520 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.
4523 iri_parts
= urllib
.parse
.urlparse(iri
)
4525 if '[' in iri_parts
.netloc
:
4526 raise ValueError('IPv6 URIs are not, yet, supported.')
4527 # Querying `.netloc`, when there's only one bracket, also raises a ValueError.
4529 # 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.
4532 if iri_parts
.username
:
4533 net_location
+= urllib
.parse
.quote(iri_parts
.username
, safe
=r
"!$%&'()*+,~")
4534 if iri_parts
.password
is not None:
4535 net_location
+= ':' + urllib
.parse
.quote(iri_parts
.password
, safe
=r
"!$%&'()*+,~")
4538 net_location
+= iri_parts
.hostname
.encode('idna').decode() # Punycode for Unicode hostnames.
4539 # The 'idna' encoding produces ASCII text.
4540 if iri_parts
.port
is not None and iri_parts
.port
!= 80:
4541 net_location
+= ':' + str(iri_parts
.port
)
4543 return urllib
.parse
.urlunparse(
4547 urllib
.parse
.quote_plus(iri_parts
.path
, safe
=r
"!$%&'()*+,/:;=@|~"),
4549 # Unsure about the `safe` argument, since this is a legacy way of handling parameters.
4550 urllib
.parse
.quote_plus(iri_parts
.params
, safe
=r
"!$%&'()*+,/:;=@|~"),
4552 # Not totally sure about the `safe` argument, since the source does not explicitly mention the query URI component.
4553 urllib
.parse
.quote_plus(iri_parts
.query
, safe
=r
"!$%&'()*+,/:;=?@{|}~"),
4555 urllib
.parse
.quote_plus(iri_parts
.fragment
, safe
=r
"!#$%&'()*+,/:;=?@{|}~")))
4557 # Source for `safe` arguments: https://url.spec.whatwg.org/#percent-encoded-bytes.
4560 def to_high_limit_path(path
):
4561 if sys
.platform
in ['win32', 'cygwin']:
4562 # Work around MAX_PATH limitation on Windows. The maximum allowed length for the individual path segments may still be quite limited.
4563 return '\\\\?\\' + os
.path
.abspath(path
)
4568 def format_field(obj
, field
=None, template
='%s', ignore
=NO_DEFAULT
, default
='', func
=IDENTITY
):
4569 val
= traversal
.traverse_obj(obj
, *variadic(field
))
4570 if not val
if ignore
is NO_DEFAULT
else val
in variadic(ignore
):
4572 return template
% func(val
)
4575 def clean_podcast_url(url
):
4576 url
= re
.sub(r
'''(?x)
4580 media\.blubrry\.com| # https://create.blubrry.com/resources/podcast-media-download-statistics/getting-started/
4585 (?:dts|www)\.podtrac\.com/(?:pts/)?redirect\.[0-9a-z]{3,4}| # http://analytics.podtrac.com/how-to-measure
4588 cn\.co| # https://podcorn.com/analytics-prefix/
4589 st\.fm # https://podsights.com/docs/
4594 return re
.sub(r
'^\w+://(\w+://)', r
'\1', url
)
4597 _HEX_TABLE
= '0123456789abcdef'
4600 def random_uuidv4():
4601 return re
.sub(r
'[xy]', lambda x
: _HEX_TABLE
[random
.randint(0, 15)], 'xxxxxxxx-xxxx-4xxx-yxxx-xxxxxxxxxxxx')
4604 def make_dir(path
, to_screen
=None):
4606 dn
= os
.path
.dirname(path
)
4608 os
.makedirs(dn
, exist_ok
=True)
4610 except OSError as err
:
4611 if callable(to_screen
) is not None:
4612 to_screen(f
'unable to create directory {err}')
4616 def get_executable_path():
4617 from ..update
import _get_variant_and_executable_path
4619 return os
.path
.dirname(os
.path
.abspath(_get_variant_and_executable_path()[1]))
4622 def get_user_config_dirs(package_name
):
4623 # .config (e.g. ~/.config/package_name)
4624 xdg_config_home
= os
.getenv('XDG_CONFIG_HOME') or compat_expanduser('~/.config')
4625 yield os
.path
.join(xdg_config_home
, package_name
)
4627 # appdata (%APPDATA%/package_name)
4628 appdata_dir
= os
.getenv('appdata')
4630 yield os
.path
.join(appdata_dir
, package_name
)
4632 # home (~/.package_name)
4633 yield os
.path
.join(compat_expanduser('~'), f
'.{package_name}')
4636 def get_system_config_dirs(package_name
):
4638 yield os
.path
.join('/etc', package_name
)
4641 def time_seconds(**kwargs
):
4643 Returns TZ-aware time in seconds since the epoch (1970-01-01T00:00:00Z)
4645 return time
.time() + datetime
.timedelta(**kwargs
).total_seconds()
4648 # create a JSON Web Signature (jws) with HS256 algorithm
4649 # the resulting format is in JWS Compact Serialization
4650 # implemented following JWT https://www.rfc-editor.org/rfc/rfc7519.html
4651 # implemented following JWS https://www.rfc-editor.org/rfc/rfc7515.html
4652 def jwt_encode_hs256(payload_data
, key
, headers
={}):
4658 header_data
.update(headers
)
4659 header_b64
= base64
.b64encode(json
.dumps(header_data
).encode())
4660 payload_b64
= base64
.b64encode(json
.dumps(payload_data
).encode())
4661 h
= hmac
.new(key
.encode(), header_b64
+ b
'.' + payload_b64
, hashlib
.sha256
)
4662 signature_b64
= base64
.b64encode(h
.digest())
4663 token
= header_b64
+ b
'.' + payload_b64
+ b
'.' + signature_b64
4667 # can be extended in future to verify the signature and parse header and return the algorithm used if it's not HS256
4668 def jwt_decode_hs256(jwt
):
4669 header_b64
, payload_b64
, signature_b64
= jwt
.split('.')
4670 # add trailing ='s that may have been stripped, superfluous ='s are ignored
4671 payload_data
= json
.loads(base64
.urlsafe_b64decode(f
'{payload_b64}==='))
4675 WINDOWS_VT_MODE
= False if compat_os_name
== 'nt' else None
4679 def supports_terminal_sequences(stream
):
4680 if compat_os_name
== 'nt':
4681 if not WINDOWS_VT_MODE
:
4683 elif not os
.getenv('TERM'):
4686 return stream
.isatty()
4687 except BaseException
:
4691 def windows_enable_vt_mode():
4692 """Ref: https://bugs.python.org/issue30075 """
4693 if get_windows_version() < (10, 0, 10586):
4697 import ctypes
.wintypes
4700 ENABLE_VIRTUAL_TERMINAL_PROCESSING
= 0x0004
4702 dll
= ctypes
.WinDLL('kernel32', use_last_error
=False)
4703 handle
= os
.open('CONOUT$', os
.O_RDWR
)
4705 h_out
= ctypes
.wintypes
.HANDLE(msvcrt
.get_osfhandle(handle
))
4706 dw_original_mode
= ctypes
.wintypes
.DWORD()
4707 success
= dll
.GetConsoleMode(h_out
, ctypes
.byref(dw_original_mode
))
4709 raise Exception('GetConsoleMode failed')
4711 success
= dll
.SetConsoleMode(h_out
, ctypes
.wintypes
.DWORD(
4712 dw_original_mode
.value | ENABLE_VIRTUAL_TERMINAL_PROCESSING
))
4714 raise Exception('SetConsoleMode failed')
4718 global WINDOWS_VT_MODE
4719 WINDOWS_VT_MODE
= True
4720 supports_terminal_sequences
.cache_clear()
4723 _terminal_sequences_re
= re
.compile('\033\\[[^m]+m')
4726 def remove_terminal_sequences(string
):
4727 return _terminal_sequences_re
.sub('', string
)
4730 def number_of_digits(number
):
4731 return len('%d' % number
)
4734 def join_nonempty(*values
, delim
='-', from_dict
=None):
4735 if from_dict
is not None:
4736 values
= (traversal
.traverse_obj(from_dict
, variadic(v
)) for v
in values
)
4737 return delim
.join(map(str, filter(None, values
)))
4740 def scale_thumbnails_to_max_format_width(formats
, thumbnails
, url_width_re
):
4742 Find the largest format dimensions in terms of video width and, for each thumbnail:
4743 * Modify the URL: Match the width with the provided regex and replace with the former width
4746 This function is useful with video services that scale the provided thumbnails on demand
4748 _keys
= ('width', 'height')
4749 max_dimensions
= max(
4750 (tuple(format
.get(k
) or 0 for k
in _keys
) for format
in formats
),
4752 if not max_dimensions
[0]:
4756 {'url': re.sub(url_width_re, str(max_dimensions[0]), thumbnail['url'])}
,
4757 dict(zip(_keys
, max_dimensions
)), thumbnail
)
4758 for thumbnail
in thumbnails
4762 def parse_http_range(range):
4763 """ Parse value of "Range" or "Content-Range" HTTP header into tuple. """
4765 return None, None, None
4766 crg
= re
.search(r
'bytes[ =](\d+)-(\d+)?(?:/(\d+))?', range)
4768 return None, None, None
4769 return int(crg
.group(1)), int_or_none(crg
.group(2)), int_or_none(crg
.group(3))
4772 def read_stdin(what
):
4773 eof
= 'Ctrl+Z' if compat_os_name
== 'nt' else 'Ctrl+D'
4774 write_string(f
'Reading {what} from STDIN - EOF ({eof}) to end:\n')
4778 def determine_file_encoding(data
):
4780 Detect the text encoding used
4781 @returns (encoding, bytes to skip)
4784 # BOM marks are given priority over declarations
4785 for bom
, enc
in BOMS
:
4786 if data
.startswith(bom
):
4787 return enc
, len(bom
)
4789 # Strip off all null bytes to match even when UTF-16 or UTF-32 is used.
4790 # We ignore the endianness to get a good enough match
4791 data
= data
.replace(b
'\0', b
'')
4792 mobj
= re
.match(rb
'(?m)^#\s*coding\s*:\s*(\S+)\s*$', data
)
4793 return mobj
.group(1).decode() if mobj
else None, 0
4800 __initialized
= False
4802 def __init__(self
, parser
, label
=None):
4803 self
.parser
, self
.label
= parser
, label
4804 self
._loaded
_paths
, self
.configs
= set(), []
4806 def init(self
, args
=None, filename
=None):
4807 assert not self
.__initialized
4808 self
.own_args
, self
.filename
= args
, filename
4809 return self
.load_configs()
4811 def load_configs(self
):
4814 location
= os
.path
.realpath(self
.filename
)
4815 directory
= os
.path
.dirname(location
)
4816 if location
in self
._loaded
_paths
:
4818 self
._loaded
_paths
.add(location
)
4820 self
.__initialized
= True
4821 opts
, _
= self
.parser
.parse_known_args(self
.own_args
)
4822 self
.parsed_args
= self
.own_args
4823 for location
in opts
.config_locations
or []:
4825 if location
in self
._loaded
_paths
:
4827 self
._loaded
_paths
.add(location
)
4828 self
.append_config(shlex
.split(read_stdin('options'), comments
=True), label
='stdin')
4830 location
= os
.path
.join(directory
, expand_path(location
))
4831 if os
.path
.isdir(location
):
4832 location
= os
.path
.join(location
, 'yt-dlp.conf')
4833 if not os
.path
.exists(location
):
4834 self
.parser
.error(f
'config location {location} does not exist')
4835 self
.append_config(self
.read_file(location
), location
)
4839 label
= join_nonempty(
4840 self
.label
, 'config', f
'"{self.filename}"' if self
.filename
else '',
4842 return join_nonempty(
4843 self
.own_args
is not None and f
'{label[0].upper()}{label[1:]}: {self.hide_login_info(self.own_args)}',
4844 *(f
'\n{c}'.replace('\n', '\n| ')[1:] for c
in self
.configs
),
4848 def read_file(filename
, default
=[]):
4850 optionf
= open(filename
, 'rb')
4852 return default
# silently skip if file is not present
4854 enc
, skip
= determine_file_encoding(optionf
.read(512))
4855 optionf
.seek(skip
, io
.SEEK_SET
)
4857 enc
= None # silently skip read errors
4859 # FIXME: https://github.com/ytdl-org/youtube-dl/commit/dfe5fa49aed02cf36ba9f743b11b0903554b5e56
4860 contents
= optionf
.read().decode(enc
or preferredencoding())
4861 res
= shlex
.split(contents
, comments
=True)
4862 except Exception as err
:
4863 raise ValueError(f
'Unable to parse "{filename}": {err}')
4869 def hide_login_info(opts
):
4870 PRIVATE_OPTS
= {'-p', '--password', '-u', '--username', '--video-password', '--ap-password', '--ap-username'}
4871 eqre
= re
.compile('^(?P<key>' + ('|'.join(re
.escape(po
) for po
in PRIVATE_OPTS
)) + ')=.+$')
4876 return m
.group('key') + '=PRIVATE'
4880 opts
= list(map(_scrub_eq
, opts
))
4881 for idx
, opt
in enumerate(opts
):
4882 if opt
in PRIVATE_OPTS
and idx
+ 1 < len(opts
):
4883 opts
[idx
+ 1] = 'PRIVATE'
4886 def append_config(self
, *args
, label
=None):
4887 config
= type(self
)(self
.parser
, label
)
4888 config
._loaded
_paths
= self
._loaded
_paths
4889 if config
.init(*args
):
4890 self
.configs
.append(config
)
4894 for config
in reversed(self
.configs
):
4895 yield from config
.all_args
4896 yield from self
.parsed_args
or []
4898 def parse_known_args(self
, **kwargs
):
4899 return self
.parser
.parse_known_args(self
.all_args
, **kwargs
)
4901 def parse_args(self
):
4902 return self
.parser
.parse_args(self
.all_args
)
4905 class WebSocketsWrapper
:
4906 """Wraps websockets module to use in non-async scopes"""
4909 def __init__(self
, url
, headers
=None, connect
=True):
4910 self
.loop
= asyncio
.new_event_loop()
4911 # XXX: "loop" is deprecated
4912 self
.conn
= websockets
.connect(
4913 url
, extra_headers
=headers
, ping_interval
=None,
4914 close_timeout
=float('inf'), loop
=self
.loop
, ping_timeout
=float('inf'))
4917 atexit
.register(self
.__exit
__, None, None, None)
4919 def __enter__(self
):
4921 self
.pool
= self
.run_with_loop(self
.conn
.__aenter
__(), self
.loop
)
4924 def send(self
, *args
):
4925 self
.run_with_loop(self
.pool
.send(*args
), self
.loop
)
4927 def recv(self
, *args
):
4928 return self
.run_with_loop(self
.pool
.recv(*args
), self
.loop
)
4930 def __exit__(self
, type, value
, traceback
):
4932 return self
.run_with_loop(self
.conn
.__aexit
__(type, value
, traceback
), self
.loop
)
4935 self
._cancel
_all
_tasks
(self
.loop
)
4937 # taken from https://github.com/python/cpython/blob/3.9/Lib/asyncio/runners.py with modifications
4938 # for contributors: If there's any new library using asyncio needs to be run in non-async, move these function out of this class
4940 def run_with_loop(main
, loop
):
4941 if not asyncio
.iscoroutine(main
):
4942 raise ValueError(f
'a coroutine was expected, got {main!r}')
4945 return loop
.run_until_complete(main
)
4947 loop
.run_until_complete(loop
.shutdown_asyncgens())
4948 if hasattr(loop
, 'shutdown_default_executor'):
4949 loop
.run_until_complete(loop
.shutdown_default_executor())
4952 def _cancel_all_tasks(loop
):
4953 to_cancel
= asyncio
.all_tasks(loop
)
4958 for task
in to_cancel
:
4961 # XXX: "loop" is removed in python 3.10+
4962 loop
.run_until_complete(
4963 asyncio
.gather(*to_cancel
, loop
=loop
, return_exceptions
=True))
4965 for task
in to_cancel
:
4966 if task
.cancelled():
4968 if task
.exception() is not None:
4969 loop
.call_exception_handler({
4970 'message': 'unhandled exception during asyncio.run() shutdown',
4971 'exception': task
.exception(),
4976 def merge_headers(*dicts
):
4977 """Merge dicts of http headers case insensitively, prioritizing the latter ones"""
4978 return {k.title(): v for k, v in itertools.chain.from_iterable(map(dict.items, dicts))}
4981 def cached_method(f
):
4982 """Cache a method"""
4983 signature
= inspect
.signature(f
)
4986 def wrapper(self
, *args
, **kwargs
):
4987 bound_args
= signature
.bind(self
, *args
, **kwargs
)
4988 bound_args
.apply_defaults()
4989 key
= tuple(bound_args
.arguments
.values())[1:]
4991 cache
= vars(self
).setdefault('_cached_method__cache', {}).setdefault(f.__name__, {}
)
4992 if key
not in cache
:
4993 cache
[key
] = f(self
, *args
, **kwargs
)
4998 class classproperty
:
4999 """property access for class methods with optional caching"""
5000 def __new__(cls
, func
=None, *args
, **kwargs
):
5002 return functools
.partial(cls
, *args
, **kwargs
)
5003 return super().__new
__(cls
)
5005 def __init__(self
, func
, *, cache
=False):
5006 functools
.update_wrapper(self
, func
)
5008 self
._cache
= {} if cache
else None
5010 def __get__(self
, _
, cls
):
5011 if self
._cache
is None:
5012 return self
.func(cls
)
5013 elif cls
not in self
._cache
:
5014 self
._cache
[cls
] = self
.func(cls
)
5015 return self
._cache
[cls
]
5018 class function_with_repr
:
5019 def __init__(self
, func
, repr_
=None):
5020 functools
.update_wrapper(self
, func
)
5021 self
.func
, self
.__repr
= func
, repr_
5023 def __call__(self
, *args
, **kwargs
):
5024 return self
.func(*args
, **kwargs
)
5029 return f
'{self.func.__module__}.{self.func.__qualname__}'
5032 class Namespace(types
.SimpleNamespace
):
5033 """Immutable namespace"""
5036 return iter(self
.__dict
__.values())
5040 return self
.__dict
__.items()
5043 MEDIA_EXTENSIONS
= Namespace(
5044 common_video
=('avi', 'flv', 'mkv', 'mov', 'mp4', 'webm'),
5045 video
=('3g2', '3gp', 'f4v', 'mk3d', 'divx', 'mpg', 'ogv', 'm4v', 'wmv'),
5046 common_audio
=('aiff', 'alac', 'flac', 'm4a', 'mka', 'mp3', 'ogg', 'opus', 'wav'),
5047 audio
=('aac', 'ape', 'asf', 'f4a', 'f4b', 'm4b', 'm4p', 'm4r', 'oga', 'ogx', 'spx', 'vorbis', 'wma', 'weba'),
5048 thumbnails
=('jpg', 'png', 'webp'),
5049 storyboards
=('mhtml', ),
5050 subtitles
=('srt', 'vtt', 'ass', 'lrc'),
5051 manifests
=('f4f', 'f4m', 'm3u8', 'smil', 'mpd'),
5053 MEDIA_EXTENSIONS
.video
+= MEDIA_EXTENSIONS
.common_video
5054 MEDIA_EXTENSIONS
.audio
+= MEDIA_EXTENSIONS
.common_audio
5056 KNOWN_EXTENSIONS
= (*MEDIA_EXTENSIONS
.video
, *MEDIA_EXTENSIONS
.audio
, *MEDIA_EXTENSIONS
.manifests
)
5061 for retry in RetryManager(...):
5064 except SomeException as err:
5068 attempt
, _error
= 0, None
5070 def __init__(self
, _retries
, _error_callback
, **kwargs
):
5071 self
.retries
= _retries
or 0
5072 self
.error_callback
= functools
.partial(_error_callback
, **kwargs
)
5074 def _should_retry(self
):
5075 return self
._error
is not NO_DEFAULT
and self
.attempt
<= self
.retries
5079 if self
._error
is NO_DEFAULT
:
5084 def error(self
, value
):
5088 while self
._should
_retry
():
5089 self
.error
= NO_DEFAULT
5093 self
.error_callback(self
.error
, self
.attempt
, self
.retries
)
5096 def report_retry(e
, count
, retries
, *, sleep_func
, info
, warn
, error
=None, suffix
=None):
5097 """Utility function for reporting retries"""
5100 return error(f
'{e}. Giving up after {count - 1} retries') if count
> 1 else error(str(e
))
5105 elif isinstance(e
, ExtractorError
):
5106 e
= remove_end(str_or_none(e
.cause
) or e
.orig_msg
, '.')
5107 warn(f
'{e}. Retrying{format_field(suffix, None, " %s")} ({count}/{retries})...')
5109 delay
= float_or_none(sleep_func(n
=count
- 1)) if callable(sleep_func
) else sleep_func
5111 info(f
'Sleeping {delay:.2f} seconds ...')
5115 def make_archive_id(ie
, video_id
):
5116 ie_key
= ie
if isinstance(ie
, str) else ie
.ie_key()
5117 return f
'{ie_key.lower()} {video_id}'
5120 def truncate_string(s
, left
, right
=0):
5121 assert left
> 3 and right
>= 0
5122 if s
is None or len(s
) <= left
+ right
:
5124 return f
'{s[:left-3]}...{s[-right:] if right else ""}'
5127 def orderedSet_from_options(options
, alias_dict
, *, use_regex
=False, start
=None):
5128 assert 'all' in alias_dict
, '"all" alias is required'
5129 requested
= list(start
or [])
5131 discard
= val
.startswith('-')
5135 if val
in alias_dict
:
5136 val
= alias_dict
[val
] if not discard
else [
5137 i
[1:] if i
.startswith('-') else f
'-{i}' for i
in alias_dict
[val
]]
5138 # NB: Do not allow regex in aliases for performance
5139 requested
= orderedSet_from_options(val
, alias_dict
, start
=requested
)
5142 current
= (filter(re
.compile(val
, re
.I
).fullmatch
, alias_dict
['all']) if use_regex
5143 else [val
] if val
in alias_dict
['all'] else None)
5145 raise ValueError(val
)
5148 for item
in current
:
5149 while item
in requested
:
5150 requested
.remove(item
)
5152 requested
.extend(current
)
5154 return orderedSet(requested
)
5159 regex
= r
' *((?P<reverse>\+)?(?P<field>[a-zA-Z0-9_]+)((?P<separator>[~:])(?P<limit>.*?))?)? *$'
5161 default
= ('hidden', 'aud_or_vid', 'hasvid', 'ie_pref', 'lang', 'quality',
5162 'res', 'fps', 'hdr:12', 'vcodec:vp9.2', 'channels', 'acodec',
5163 'size', 'br', 'asr', 'proto', 'ext', 'hasaud', 'source', 'id') # These must not be aliases
5164 ytdl_default
= ('hasaud', 'lang', 'quality', 'tbr', 'filesize', 'vbr',
5165 'height', 'width', 'proto', 'vext', 'abr', 'aext',
5166 'fps', 'fs_approx', 'source', 'id')
5169 'vcodec': {'type': 'ordered', 'regex': True,
5170 'order': ['av0?1', 'vp0?9.2', 'vp0?9', '[hx]265|he?vc?', '[hx]264|avc', 'vp0?8', 'mp4v|h263', 'theora', '', None, 'none']},
5171 'acodec': {'type': 'ordered', 'regex': True,
5172 'order': ['[af]lac', 'wav|aiff', 'opus', 'vorbis|ogg', 'aac', 'mp?4a?', 'mp3', 'ac-?4', 'e-?a?c-?3', 'ac-?3', 'dts', '', None, 'none']},
5173 'hdr': {'type': 'ordered', 'regex': True, 'field': 'dynamic_range',
5174 'order': ['dv', '(hdr)?12', r
'(hdr)?10\+', '(hdr)?10', 'hlg', '', 'sdr', None]},
5175 'proto': {'type': 'ordered', 'regex': True, 'field': 'protocol',
5176 'order': ['(ht|f)tps', '(ht|f)tp$', 'm3u8.*', '.*dash', 'websocket_frag', 'rtmpe?', '', 'mms|rtsp', 'ws|websocket', 'f4']},
5177 'vext': {'type': 'ordered', 'field': 'video_ext',
5178 'order': ('mp4', 'mov', 'webm', 'flv', '', 'none'),
5179 'order_free': ('webm', 'mp4', 'mov', 'flv', '', 'none')},
5180 'aext': {'type': 'ordered', 'regex': True, 'field': 'audio_ext',
5181 'order': ('m4a', 'aac', 'mp3', 'ogg', 'opus', 'web[am]', '', 'none'),
5182 'order_free': ('ogg', 'opus', 'web[am]', 'mp3', 'm4a', 'aac', '', 'none')},
5183 'hidden': {'visible': False, 'forced': True, 'type': 'extractor', 'max': -1000}
,
5184 'aud_or_vid': {'visible': False, 'forced': True, 'type': 'multiple',
5185 'field': ('vcodec', 'acodec'),
5186 'function': lambda it
: int(any(v
!= 'none' for v
in it
))},
5187 'ie_pref': {'priority': True, 'type': 'extractor'}
,
5188 'hasvid': {'priority': True, 'field': 'vcodec', 'type': 'boolean', 'not_in_list': ('none',)}
,
5189 'hasaud': {'field': 'acodec', 'type': 'boolean', 'not_in_list': ('none',)}
,
5190 'lang': {'convert': 'float', 'field': 'language_preference', 'default': -1}
,
5191 'quality': {'convert': 'float', 'default': -1}
,
5192 'filesize': {'convert': 'bytes'}
,
5193 'fs_approx': {'convert': 'bytes', 'field': 'filesize_approx'}
,
5194 'id': {'convert': 'string', 'field': 'format_id'}
,
5195 'height': {'convert': 'float_none'}
,
5196 'width': {'convert': 'float_none'}
,
5197 'fps': {'convert': 'float_none'}
,
5198 'channels': {'convert': 'float_none', 'field': 'audio_channels'}
,
5199 'tbr': {'convert': 'float_none'}
,
5200 'vbr': {'convert': 'float_none'}
,
5201 'abr': {'convert': 'float_none'}
,
5202 'asr': {'convert': 'float_none'}
,
5203 'source': {'convert': 'float', 'field': 'source_preference', 'default': -1}
,
5205 'codec': {'type': 'combined', 'field': ('vcodec', 'acodec')}
,
5206 'br': {'type': 'multiple', 'field': ('tbr', 'vbr', 'abr'), 'convert': 'float_none',
5207 'function': lambda it
: next(filter(None, it
), None)},
5208 'size': {'type': 'multiple', 'field': ('filesize', 'fs_approx'), 'convert': 'bytes',
5209 'function': lambda it
: next(filter(None, it
), None)},
5210 'ext': {'type': 'combined', 'field': ('vext', 'aext')}
,
5211 'res': {'type': 'multiple', 'field': ('height', 'width'),
5212 'function': lambda it
: (lambda l
: min(l
) if l
else 0)(tuple(filter(None, it
)))},
5214 # Actual field names
5215 'format_id': {'type': 'alias', 'field': 'id'}
,
5216 'preference': {'type': 'alias', 'field': 'ie_pref'}
,
5217 'language_preference': {'type': 'alias', 'field': 'lang'}
,
5218 'source_preference': {'type': 'alias', 'field': 'source'}
,
5219 'protocol': {'type': 'alias', 'field': 'proto'}
,
5220 'filesize_approx': {'type': 'alias', 'field': 'fs_approx'}
,
5221 'audio_channels': {'type': 'alias', 'field': 'channels'}
,
5224 'dimension': {'type': 'alias', 'field': 'res', 'deprecated': True}
,
5225 'resolution': {'type': 'alias', 'field': 'res', 'deprecated': True}
,
5226 'extension': {'type': 'alias', 'field': 'ext', 'deprecated': True}
,
5227 'bitrate': {'type': 'alias', 'field': 'br', 'deprecated': True}
,
5228 'total_bitrate': {'type': 'alias', 'field': 'tbr', 'deprecated': True}
,
5229 'video_bitrate': {'type': 'alias', 'field': 'vbr', 'deprecated': True}
,
5230 'audio_bitrate': {'type': 'alias', 'field': 'abr', 'deprecated': True}
,
5231 'framerate': {'type': 'alias', 'field': 'fps', 'deprecated': True}
,
5232 'filesize_estimate': {'type': 'alias', 'field': 'size', 'deprecated': True}
,
5233 'samplerate': {'type': 'alias', 'field': 'asr', 'deprecated': True}
,
5234 'video_ext': {'type': 'alias', 'field': 'vext', 'deprecated': True}
,
5235 'audio_ext': {'type': 'alias', 'field': 'aext', 'deprecated': True}
,
5236 'video_codec': {'type': 'alias', 'field': 'vcodec', 'deprecated': True}
,
5237 'audio_codec': {'type': 'alias', 'field': 'acodec', 'deprecated': True}
,
5238 'video': {'type': 'alias', 'field': 'hasvid', 'deprecated': True}
,
5239 'has_video': {'type': 'alias', 'field': 'hasvid', 'deprecated': True}
,
5240 'audio': {'type': 'alias', 'field': 'hasaud', 'deprecated': True}
,
5241 'has_audio': {'type': 'alias', 'field': 'hasaud', 'deprecated': True}
,
5242 'extractor': {'type': 'alias', 'field': 'ie_pref', 'deprecated': True}
,
5243 'extractor_preference': {'type': 'alias', 'field': 'ie_pref', 'deprecated': True}
,
5246 def __init__(self
, ydl
, field_preference
):
5249 self
.evaluate_params(self
.ydl
.params
, field_preference
)
5250 if ydl
.params
.get('verbose'):
5251 self
.print_verbose_info(self
.ydl
.write_debug
)
5253 def _get_field_setting(self
, field
, key
):
5254 if field
not in self
.settings
:
5255 if key
in ('forced', 'priority'):
5257 self
.ydl
.deprecated_feature(f
'Using arbitrary fields ({field}) for format sorting is '
5258 'deprecated and may be removed in a future version')
5259 self
.settings
[field
] = {}
5260 propObj
= self
.settings
[field
]
5261 if key
not in propObj
:
5262 type = propObj
.get('type')
5264 default
= 'preference' if type == 'extractor' else (field
,) if type in ('combined', 'multiple') else field
5265 elif key
== 'convert':
5266 default
= 'order' if type == 'ordered' else 'float_string' if field
else 'ignore'
5268 default
= {'type': 'field', 'visible': True, 'order': [], 'not_in_list': (None,)}
.get(key
, None)
5269 propObj
[key
] = default
5272 def _resolve_field_value(self
, field
, value
, convertNone
=False):
5277 value
= value
.lower()
5278 conversion
= self
._get
_field
_setting
(field
, 'convert')
5279 if conversion
== 'ignore':
5281 if conversion
== 'string':
5283 elif conversion
== 'float_none':
5284 return float_or_none(value
)
5285 elif conversion
== 'bytes':
5286 return parse_bytes(value
)
5287 elif conversion
== 'order':
5288 order_list
= (self
._use
_free
_order
and self
._get
_field
_setting
(field
, 'order_free')) or self
._get
_field
_setting
(field
, 'order')
5289 use_regex
= self
._get
_field
_setting
(field
, 'regex')
5290 list_length
= len(order_list
)
5291 empty_pos
= order_list
.index('') if '' in order_list
else list_length
+ 1
5292 if use_regex
and value
is not None:
5293 for i
, regex
in enumerate(order_list
):
5294 if regex
and re
.match(regex
, value
):
5295 return list_length
- i
5296 return list_length
- empty_pos
# not in list
5297 else: # not regex or value = None
5298 return list_length
- (order_list
.index(value
) if value
in order_list
else empty_pos
)
5300 if value
.isnumeric():
5303 self
.settings
[field
]['convert'] = 'string'
5306 def evaluate_params(self
, params
, sort_extractor
):
5307 self
._use
_free
_order
= params
.get('prefer_free_formats', False)
5308 self
._sort
_user
= params
.get('format_sort', [])
5309 self
._sort
_extractor
= sort_extractor
5311 def add_item(field
, reverse
, closest
, limit_text
):
5312 field
= field
.lower()
5313 if field
in self
._order
:
5315 self
._order
.append(field
)
5316 limit
= self
._resolve
_field
_value
(field
, limit_text
)
5319 'closest': False if limit
is None else closest
,
5320 'limit_text': limit_text
,
5322 if field
in self
.settings
:
5323 self
.settings
[field
].update(data
)
5325 self
.settings
[field
] = data
5328 tuple(field
for field
in self
.default
if self
._get
_field
_setting
(field
, 'forced'))
5329 + (tuple() if params
.get('format_sort_force', False)
5330 else tuple(field
for field
in self
.default
if self
._get
_field
_setting
(field
, 'priority')))
5331 + tuple(self
._sort
_user
) + tuple(sort_extractor
) + self
.default
)
5333 for item
in sort_list
:
5334 match
= re
.match(self
.regex
, item
)
5336 raise ExtractorError('Invalid format sort string "%s" given by extractor' % item
)
5337 field
= match
.group('field')
5340 if self
._get
_field
_setting
(field
, 'type') == 'alias':
5341 alias
, field
= field
, self
._get
_field
_setting
(field
, 'field')
5342 if self
._get
_field
_setting
(alias
, 'deprecated'):
5343 self
.ydl
.deprecated_feature(f
'Format sorting alias {alias} is deprecated and may '
5344 f
'be removed in a future version. Please use {field} instead')
5345 reverse
= match
.group('reverse') is not None
5346 closest
= match
.group('separator') == '~'
5347 limit_text
= match
.group('limit')
5349 has_limit
= limit_text
is not None
5350 has_multiple_fields
= self
._get
_field
_setting
(field
, 'type') == 'combined'
5351 has_multiple_limits
= has_limit
and has_multiple_fields
and not self
._get
_field
_setting
(field
, 'same_limit')
5353 fields
= self
._get
_field
_setting
(field
, 'field') if has_multiple_fields
else (field
,)
5354 limits
= limit_text
.split(':') if has_multiple_limits
else (limit_text
,) if has_limit
else tuple()
5355 limit_count
= len(limits
)
5356 for (i
, f
) in enumerate(fields
):
5357 add_item(f
, reverse
, closest
,
5358 limits
[i
] if i
< limit_count
5359 else limits
[0] if has_limit
and not has_multiple_limits
5362 def print_verbose_info(self
, write_debug
):
5364 write_debug('Sort order given by user: %s' % ', '.join(self
._sort
_user
))
5365 if self
._sort
_extractor
:
5366 write_debug('Sort order given by extractor: %s' % ', '.join(self
._sort
_extractor
))
5367 write_debug('Formats sorted by: %s' % ', '.join(['%s%s%s' % (
5368 '+' if self
._get
_field
_setting
(field
, 'reverse') else '', field
,
5369 '%s%s(%s)' % ('~' if self
._get
_field
_setting
(field
, 'closest') else ':',
5370 self
._get
_field
_setting
(field
, 'limit_text'),
5371 self
._get
_field
_setting
(field
, 'limit'))
5372 if self
._get
_field
_setting
(field
, 'limit_text') is not None else '')
5373 for field
in self
._order
if self
._get
_field
_setting
(field
, 'visible')]))
5375 def _calculate_field_preference_from_value(self
, format
, field
, type, value
):
5376 reverse
= self
._get
_field
_setting
(field
, 'reverse')
5377 closest
= self
._get
_field
_setting
(field
, 'closest')
5378 limit
= self
._get
_field
_setting
(field
, 'limit')
5380 if type == 'extractor':
5381 maximum
= self
._get
_field
_setting
(field
, 'max')
5382 if value
is None or (maximum
is not None and value
>= maximum
):
5384 elif type == 'boolean':
5385 in_list
= self
._get
_field
_setting
(field
, 'in_list')
5386 not_in_list
= self
._get
_field
_setting
(field
, 'not_in_list')
5387 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
5388 elif type == 'ordered':
5389 value
= self
._resolve
_field
_value
(field
, value
, True)
5391 # try to convert to number
5392 val_num
= float_or_none(value
, default
=self
._get
_field
_setting
(field
, 'default'))
5393 is_num
= self
._get
_field
_setting
(field
, 'convert') != 'string' and val_num
is not None
5397 return ((-10, 0) if value
is None
5398 else (1, value
, 0) if not is_num
# if a field has mixed strings and numbers, strings are sorted higher
5399 else (0, -abs(value
- limit
), value
- limit
if reverse
else limit
- value
) if closest
5400 else (0, value
, 0) if not reverse
and (limit
is None or value
<= limit
)
5401 else (0, -value
, 0) if limit
is None or (reverse
and value
== limit
) or value
> limit
5402 else (-1, value
, 0))
5404 def _calculate_field_preference(self
, format
, field
):
5405 type = self
._get
_field
_setting
(field
, 'type') # extractor, boolean, ordered, field, multiple
5406 get_value
= lambda f
: format
.get(self
._get
_field
_setting
(f
, 'field'))
5407 if type == 'multiple':
5408 type = 'field' # Only 'field' is allowed in multiple for now
5409 actual_fields
= self
._get
_field
_setting
(field
, 'field')
5411 value
= self
._get
_field
_setting
(field
, 'function')(get_value(f
) for f
in actual_fields
)
5413 value
= get_value(field
)
5414 return self
._calculate
_field
_preference
_from
_value
(format
, field
, type, value
)
5416 def calculate_preference(self
, format
):
5417 # Determine missing protocol
5418 if not format
.get('protocol'):
5419 format
['protocol'] = determine_protocol(format
)
5421 # Determine missing ext
5422 if not format
.get('ext') and 'url' in format
:
5423 format
['ext'] = determine_ext(format
['url'])
5424 if format
.get('vcodec') == 'none':
5425 format
['audio_ext'] = format
['ext'] if format
.get('acodec') != 'none' else 'none'
5426 format
['video_ext'] = 'none'
5428 format
['video_ext'] = format
['ext']
5429 format
['audio_ext'] = 'none'
5430 # if format.get('preference') is None and format.get('ext') in ('f4f', 'f4m'): # Not supported?
5431 # format['preference'] = -1000
5433 if format
.get('preference') is None and format
.get('ext') == 'flv' and re
.match('[hx]265|he?vc?', format
.get('vcodec') or ''):
5434 # HEVC-over-FLV is out-of-spec by FLV's original spec
5435 # ref. https://trac.ffmpeg.org/ticket/6389
5436 # ref. https://github.com/yt-dlp/yt-dlp/pull/5821
5437 format
['preference'] = -100
5439 # Determine missing bitrates
5440 if format
.get('vcodec') == 'none':
5442 if format
.get('acodec') == 'none':
5444 if not format
.get('vbr') and format
.get('vcodec') != 'none':
5445 format
['vbr'] = try_call(lambda: format
['tbr'] - format
['abr']) or None
5446 if not format
.get('abr') and format
.get('acodec') != 'none':
5447 format
['abr'] = try_call(lambda: format
['tbr'] - format
['vbr']) or None
5448 if not format
.get('tbr'):
5449 format
['tbr'] = try_call(lambda: format
['vbr'] + format
['abr']) or None
5451 return tuple(self
._calculate
_field
_preference
(format
, field
) for field
in self
._order
)
5456 def __init__(self
, ydl
=None):
5459 def debug(self
, message
):
5461 self
._ydl
.write_debug(message
)
5463 def info(self
, message
):
5465 self
._ydl
.to_screen(message
)
5467 def warning(self
, message
, *, once
=False):
5469 self
._ydl
.report_warning(message
, once
)
5471 def error(self
, message
, *, is_error
=True):
5473 self
._ydl
.report_error(message
, is_error
=is_error
)
5475 def stdout(self
, message
):
5477 self
._ydl
.to_stdout(message
)
5479 def stderr(self
, message
):
5481 self
._ydl
.to_stderr(message
)