43 import xml
.etree
.ElementTree
45 from . import traversal
47 from ..compat
import functools
# isort: split
48 from ..compat
import (
49 compat_etree_fromstring
,
51 compat_HTMLParseError
,
54 from ..dependencies
import xattr
56 __name__
= __name__
.rsplit('.', 1)[0] # Pretend to be the parent module
58 # This is not clearly defined otherwise
59 compiled_regex_type
= type(re
.compile(''))
70 ENGLISH_MONTH_NAMES
= [
71 'January', 'February', 'March', 'April', 'May', 'June',
72 'July', 'August', 'September', 'October', 'November', 'December']
75 'en': ENGLISH_MONTH_NAMES
,
77 'janvier', 'février', 'mars', 'avril', 'mai', 'juin',
78 'juillet', 'août', 'septembre', 'octobre', 'novembre', 'décembre'],
79 # these follow the genitive grammatical case (dopełniacz)
80 # some websites might be using nominative, which will require another month list
81 # https://en.wikibooks.org/wiki/Polish/Noun_cases
82 'pl': ['stycznia', 'lutego', 'marca', 'kwietnia', 'maja', 'czerwca',
83 'lipca', 'sierpnia', 'września', 'października', 'listopada', 'grudnia'],
86 # From https://github.com/python/cpython/blob/3.11/Lib/email/_parseaddr.py#L36-L42
88 'UT': 0, 'UTC': 0, 'GMT': 0, 'Z': 0,
89 'AST': -4, 'ADT': -3, # Atlantic (used in Canada)
90 'EST': -5, 'EDT': -4, # Eastern
91 'CST': -6, 'CDT': -5, # Central
92 'MST': -7, 'MDT': -6, # Mountain
93 'PST': -8, 'PDT': -7 # Pacific
96 # needed for sanitizing filenames in restricted mode
97 ACCENT_CHARS
= dict(zip('ÂÃÄÀÁÅÆÇÈÉÊËÌÍÎÏÐÑÒÓÔÕÖŐØŒÙÚÛÜŰÝÞßàáâãäåæçèéêëìíîïðñòóôõöőøœùúûüűýþÿ',
98 itertools
.chain('AAAAAA', ['AE'], 'CEEEEIIIIDNOOOOOOO', ['OE'], 'UUUUUY', ['TH', 'ss'],
99 'aaaaaa', ['ae'], 'ceeeeiiiionooooooo', ['oe'], 'uuuuuy', ['th'], 'y')))
129 '%Y-%m-%d %H:%M:%S.%f',
130 '%Y-%m-%d %H:%M:%S:%f',
133 '%Y-%m-%dT%H:%M:%SZ',
134 '%Y-%m-%dT%H:%M:%S.%fZ',
135 '%Y-%m-%dT%H:%M:%S.%f0Z',
137 '%Y-%m-%dT%H:%M:%S.%f',
140 '%b %d %Y at %H:%M:%S',
142 '%B %d %Y at %H:%M:%S',
146 DATE_FORMATS_DAY_FIRST
= list(DATE_FORMATS
)
147 DATE_FORMATS_DAY_FIRST
.extend([
158 DATE_FORMATS_MONTH_FIRST
= list(DATE_FORMATS
)
159 DATE_FORMATS_MONTH_FIRST
.extend([
167 PACKED_CODES_RE
= r
"}\('(.+)',(\d+),(\d+),'([^']+)'\.split\('\|'\)"
168 JSON_LD_RE
= r
'(?is)<script[^>]+type=(["\']?
)application
/ld\
+json\
1[^
>]*>\s
*(?P
<json_ld
>{.+?}|\
[.+?\
])\s
*</script
>'
170 NUMBER_RE = r'\d
+(?
:\
.\d
+)?
'
174 def preferredencoding():
175 """Get preferred encoding.
177 Returns the best encoding scheme for the system, based on
178 locale.getpreferredencoding() and some further tweaks.
181 pref = locale.getpreferredencoding()
189 def write_json_file(obj, fn):
190 """ Encode obj as JSON and write it to fn, atomically if possible """
192 tf = tempfile.NamedTemporaryFile(
193 prefix=f'{os.path.basename(fn)}
.', dir=os.path.dirname(fn),
194 suffix='.tmp
', delete=False, mode='w
', encoding='utf
-8')
198 json.dump(obj, tf, ensure_ascii=False)
199 if sys.platform == 'win32
':
200 # Need to remove existing file on Windows, else os.rename raises
201 # WindowsError or FileExistsError.
202 with contextlib.suppress(OSError):
204 with contextlib.suppress(OSError):
207 os.chmod(tf.name, 0o666 & ~mask)
208 os.rename(tf.name, fn)
210 with contextlib.suppress(OSError):
215 def find_xpath_attr(node, xpath, key, val=None):
216 """ Find the xpath xpath[@key=val] """
217 assert re.match(r'^
[a
-zA
-Z_
-]+$
', key)
218 expr = xpath + ('[@%s]' % key if val is None else f"[@{key}='{val}
']")
219 return node.find(expr)
221 # On python2.6 the xml.etree.ElementTree.Element methods don't support
222 # the namespace parameter
225 def xpath_with_ns(path
, ns_map
):
226 components
= [c
.split(':') for c
in path
.split('/')]
230 replaced
.append(c
[0])
233 replaced
.append('{%s}%s' % (ns_map
[ns
], tag
))
234 return '/'.join(replaced
)
237 def xpath_element(node
, xpath
, name
=None, fatal
=False, default
=NO_DEFAULT
):
238 def _find_xpath(xpath
):
239 return node
.find(xpath
)
241 if isinstance(xpath
, str):
242 n
= _find_xpath(xpath
)
250 if default
is not NO_DEFAULT
:
253 name
= xpath
if name
is None else name
254 raise ExtractorError('Could not find XML element %s' % name
)
260 def xpath_text(node
, xpath
, name
=None, fatal
=False, default
=NO_DEFAULT
):
261 n
= xpath_element(node
, xpath
, name
, fatal
=fatal
, default
=default
)
262 if n
is None or n
== default
:
265 if default
is not NO_DEFAULT
:
268 name
= xpath
if name
is None else name
269 raise ExtractorError('Could not find XML element\'s text %s' % name
)
275 def xpath_attr(node
, xpath
, key
, name
=None, fatal
=False, default
=NO_DEFAULT
):
276 n
= find_xpath_attr(node
, xpath
, key
)
278 if default
is not NO_DEFAULT
:
281 name
= f
'{xpath}[@{key}]' if name
is None else name
282 raise ExtractorError('Could not find XML attribute %s' % name
)
288 def get_element_by_id(id, html
, **kwargs
):
289 """Return the content of the tag with the specified ID in the passed HTML document"""
290 return get_element_by_attribute('id', id, html
, **kwargs
)
293 def get_element_html_by_id(id, html
, **kwargs
):
294 """Return the html of the tag with the specified ID in the passed HTML document"""
295 return get_element_html_by_attribute('id', id, html
, **kwargs
)
298 def get_element_by_class(class_name
, html
):
299 """Return the content of the first tag with the specified class in the passed HTML document"""
300 retval
= get_elements_by_class(class_name
, html
)
301 return retval
[0] if retval
else None
304 def get_element_html_by_class(class_name
, html
):
305 """Return the html of the first tag with the specified class in the passed HTML document"""
306 retval
= get_elements_html_by_class(class_name
, html
)
307 return retval
[0] if retval
else None
310 def get_element_by_attribute(attribute
, value
, html
, **kwargs
):
311 retval
= get_elements_by_attribute(attribute
, value
, html
, **kwargs
)
312 return retval
[0] if retval
else None
315 def get_element_html_by_attribute(attribute
, value
, html
, **kargs
):
316 retval
= get_elements_html_by_attribute(attribute
, value
, html
, **kargs
)
317 return retval
[0] if retval
else None
320 def get_elements_by_class(class_name
, html
, **kargs
):
321 """Return the content of all tags with the specified class in the passed HTML document as a list"""
322 return get_elements_by_attribute(
323 'class', r
'[^\'"]*(?<=[\'"\s
])%s(?
=[\'"\s])[^\'"]*' % re.escape(class_name),
324 html, escape_value=False)
327 def get_elements_html_by_class(class_name, html):
328 """Return the html of all tags with the specified class in the passed HTML document as a list"""
329 return get_elements_html_by_attribute(
330 'class', r'[^
\'"]*(?<=[\'"\s
])%s(?
=[\'"\s])[^\'"]*' % re.escape(class_name),
331 html, escape_value=False)
334 def get_elements_by_attribute(*args, **kwargs):
335 """Return the content of the tag with the specified attribute in the passed HTML document"""
336 return [content for content, _ in get_elements_text_and_html_by_attribute(*args, **kwargs)]
339 def get_elements_html_by_attribute(*args, **kwargs):
340 """Return the html of the tag with the specified attribute in the passed HTML document"""
341 return [whole for _, whole in get_elements_text_and_html_by_attribute(*args, **kwargs)]
344 def get_elements_text_and_html_by_attribute(attribute, value, html, *, tag=r'[\w
:.-]+', escape_value=True):
346 Return the text (content) and the html (whole) of the tag with the specified
347 attribute in the passed HTML document
352 quote = '' if re.match(r'''[\s"'`
=<>]''', value) else '?'
354 value = re.escape(value) if escape_value else value
356 partial_element_re = rf'''(?x
)
358 (?
:\
s(?
:[^
>"']|"[^
"]*"|
'[^']*')*)?
359 \s{re.escape(attribute)}\s*=\s*(?P<_q>['"]{quote})(?-x:{value})(?P=_q)
362 for m in re.finditer(partial_element_re, html):
363 content, whole = get_element_text_and_html_by_tag(m.group('tag'), html[m.start():])
366 unescapeHTML(re.sub(r'^(?P<q>["\'])(?P
<content
>.*)(?P
=q
)$
', r'\g
<content
>', content, flags=re.DOTALL)),
371 class HTMLBreakOnClosingTagParser(html.parser.HTMLParser):
373 HTML parser which raises HTMLBreakOnClosingTagException upon reaching the
374 closing tag for the first opening tag it has encountered, and can be used
378 class HTMLBreakOnClosingTagException(Exception):
382 self.tagstack = collections.deque()
383 html.parser.HTMLParser.__init__(self)
388 def __exit__(self, *_):
392 # handle_endtag does not return upon raising HTMLBreakOnClosingTagException,
393 # so data remains buffered; we no longer have any interest in it, thus
394 # override this method to discard it
397 def handle_starttag(self, tag, _):
398 self.tagstack.append(tag)
400 def handle_endtag(self, tag):
401 if not self.tagstack:
402 raise compat_HTMLParseError('no tags
in the stack
')
404 inner_tag = self.tagstack.pop()
408 raise compat_HTMLParseError(f'matching opening tag
for closing {tag} tag
not found
')
409 if not self.tagstack:
410 raise self.HTMLBreakOnClosingTagException()
413 # XXX: This should be far less strict
414 def get_element_text_and_html_by_tag(tag, html):
416 For the first element with the specified tag in the passed HTML document
417 return its' content (text
) and the whole
element (html
)
419 def find_or_raise(haystack, needle, exc):
421 return haystack.index(needle)
424 closing_tag = f'</{tag}>'
425 whole_start = find_or_raise(
426 html, f'<{tag}', compat_HTMLParseError(f'opening {tag} tag not found'))
427 content_start = find_or_raise(
428 html[whole_start:], '>', compat_HTMLParseError(f'malformed opening {tag} tag'))
429 content_start += whole_start + 1
430 with HTMLBreakOnClosingTagParser() as parser:
431 parser.feed(html[whole_start:content_start])
432 if not parser.tagstack or parser.tagstack[0] != tag:
433 raise compat_HTMLParseError(f'parser did not match opening {tag} tag')
434 offset = content_start
435 while offset < len(html):
436 next_closing_tag_start = find_or_raise(
437 html[offset:], closing_tag,
438 compat_HTMLParseError(f'closing {tag} tag not found'))
439 next_closing_tag_end = next_closing_tag_start + len(closing_tag)
441 parser.feed(html[offset:offset + next_closing_tag_end])
442 offset += next_closing_tag_end
443 except HTMLBreakOnClosingTagParser.HTMLBreakOnClosingTagException:
444 return html[content_start:offset + next_closing_tag_start], \
445 html[whole_start:offset + next_closing_tag_end]
446 raise compat_HTMLParseError('unexpected end of html')
449 class HTMLAttributeParser(html.parser.HTMLParser):
450 """Trivial HTML parser to gather the attributes
for a single element
"""
454 html.parser.HTMLParser.__init__(self)
456 def handle_starttag(self, tag, attrs):
457 self.attrs = dict(attrs)
458 raise compat_HTMLParseError('done')
461 class HTMLListAttrsParser(html.parser.HTMLParser):
462 """HTML parser to gather the attributes
for the elements of a
list"""
465 html.parser.HTMLParser.__init__(self)
469 def handle_starttag(self, tag, attrs):
470 if tag == 'li' and self._level == 0:
471 self.items.append(dict(attrs))
474 def handle_endtag(self, tag):
478 def extract_attributes(html_element):
479 """Given a string
for an HTML element such
as
481 a
="foo" B
="bar" c
="&98;az" d
=boz
482 empty
= noval entity
="&"
485 Decode
and return a dictionary of attributes
.
487 'a': 'foo', 'b': 'bar', c
: 'baz', d
: 'boz',
488 'empty': '', 'noval': None, 'entity': '&',
489 'sq': '"', 'dq': '\''
492 parser = HTMLAttributeParser()
493 with contextlib.suppress(compat_HTMLParseError):
494 parser.feed(html_element)
499 def parse_list(webpage):
500 """Given a string
for an series of HTML
<li
> elements
,
501 return a dictionary of their attributes
"""
502 parser = HTMLListAttrsParser()
508 def clean_html(html):
509 """Clean an HTML snippet into a readable string
"""
511 if html is None: # Convenience for sanitizing descriptions etc.
514 html = re.sub(r'\s+', ' ', html)
515 html = re.sub(r'(?u)\s?<\s?br\s?/?\s?>\s?', '\n', html)
516 html = re.sub(r'(?u)<\s?/\s?p\s?>\s?<\s?p[^>]*>', '\n', html)
518 html = re.sub('<.*?>', '', html)
519 # Replace html entities
520 html = unescapeHTML(html)
524 class LenientJSONDecoder(json.JSONDecoder):
526 def __init__(self, *args, transform_source=None, ignore_extra=False, close_objects=0, **kwargs):
527 self.transform_source, self.ignore_extra = transform_source, ignore_extra
528 self._close_attempts = 2 * close_objects
529 super().__init__(*args, **kwargs)
532 def _close_object(err):
533 doc = err.doc[:err.pos]
534 # We need to add comma first to get the correct error message
535 if err.msg.startswith('Expecting \',\''):
537 elif not doc.endswith(','):
540 if err.msg.startswith('Expecting property name'):
541 return doc[:-1] + '}'
542 elif err.msg.startswith('Expecting value'):
543 return doc[:-1] + ']'
546 if self.transform_source:
547 s = self.transform_source(s)
548 for attempt in range(self._close_attempts + 1):
550 if self.ignore_extra:
551 return self.raw_decode(s.lstrip())[0]
552 return super().decode(s)
553 except json.JSONDecodeError as e:
556 elif attempt < self._close_attempts:
557 s = self._close_object(e)
560 raise type(e)(f'{e.msg} in {s[e.pos - 10:e.pos + 10]!r}', s, e.pos)
561 assert False, 'Too many attempts to decode JSON'
564 def sanitize_open(filename, open_mode):
565 """Try to
open the given filename
, and slightly tweak it
if this fails
.
567 Attempts to
open the given filename
. If this fails
, it tries to change
568 the filename slightly
, step by step
, until it
's either able to open it
569 or it fails and raises a final exception, like the standard open()
572 It returns the tuple (stream, definitive_file_name).
575 if sys.platform == 'win32
':
578 # stdout may be any IO stream, e.g. when using contextlib.redirect_stdout
579 with contextlib.suppress(io.UnsupportedOperation):
580 msvcrt.setmode(sys.stdout.fileno(), os.O_BINARY)
581 return (sys.stdout.buffer if hasattr(sys.stdout, 'buffer') else sys.stdout, filename)
583 for attempt in range(2):
586 if sys.platform == 'win32
':
587 # FIXME: An exclusive lock also locks the file from being read.
588 # Since windows locks are mandatory, don't lock the
file on
windows (for now
).
589 # Ref: https://github.com/yt-dlp/yt-dlp/issues/3124
590 raise LockingUnsupportedError()
591 stream
= locked_file(filename
, open_mode
, block
=False).__enter
__()
593 stream
= open(filename
, open_mode
)
594 return stream
, filename
595 except OSError as err
:
596 if attempt
or err
.errno
in (errno
.EACCES
,):
598 old_filename
, filename
= filename
, sanitize_path(filename
)
599 if old_filename
== filename
:
603 def timeconvert(timestr
):
604 """Convert RFC 2822 defined time string into system timestamp"""
606 timetuple
= email
.utils
.parsedate_tz(timestr
)
607 if timetuple
is not None:
608 timestamp
= email
.utils
.mktime_tz(timetuple
)
612 def sanitize_filename(s
, restricted
=False, is_id
=NO_DEFAULT
):
613 """Sanitizes a string so it could be used as part of a filename.
614 @param restricted Use a stricter subset of allowed characters
615 @param is_id Whether this is an ID that should be kept unchanged if possible.
616 If unset, yt-dlp's new sanitization rules are in effect
621 def replace_insane(char
):
622 if restricted
and char
in ACCENT_CHARS
:
623 return ACCENT_CHARS
[char
]
624 elif not restricted
and char
== '\n':
626 elif is_id
is NO_DEFAULT
and not restricted
and char
in '"*:<>?|/\\':
627 # Replace with their full-width unicode counterparts
628 return {'/': '\u29F8', '\\': '\u29f9'}
.get(char
, chr(ord(char
) + 0xfee0))
629 elif char
== '?' or ord(char
) < 32 or ord(char
) == 127:
632 return '' if restricted
else '\''
634 return '\0_\0-' if restricted
else '\0 \0-'
635 elif char
in '\\/|*<>':
637 if restricted
and (char
in '!&\'()[]{}$;`^,#' or char
.isspace() or ord(char
) > 127):
638 return '' if unicodedata
.category(char
)[0] in 'CM' else '\0_'
641 # Replace look-alike Unicode glyphs
642 if restricted
and (is_id
is NO_DEFAULT
or not is_id
):
643 s
= unicodedata
.normalize('NFKC', s
)
644 s
= re
.sub(r
'[0-9]+(?::[0-9]+)+', lambda m
: m
.group(0).replace(':', '_'), s
) # Handle timestamps
645 result
= ''.join(map(replace_insane
, s
))
646 if is_id
is NO_DEFAULT
:
647 result
= re
.sub(r
'(\0.)(?:(?=\1)..)+', r
'\1', result
) # Remove repeated substitute chars
648 STRIP_RE
= r
'(?:\0.|[ _-])*'
649 result
= re
.sub(f
'^\0.{STRIP_RE}|{STRIP_RE}\0.$', '', result
) # Remove substitute chars from start/end
650 result
= result
.replace('\0', '') or '_'
653 while '__' in result
:
654 result
= result
.replace('__', '_')
655 result
= result
.strip('_')
656 # Common case of "Foreign band name - English song title"
657 if restricted
and result
.startswith('-_'):
659 if result
.startswith('-'):
660 result
= '_' + result
[len('-'):]
661 result
= result
.lstrip('.')
667 def sanitize_path(s
, force
=False):
668 """Sanitizes and normalizes path on Windows"""
669 # XXX: this handles drive relative paths (c:sth) incorrectly
670 if sys
.platform
== 'win32':
672 drive_or_unc
, _
= os
.path
.splitdrive(s
)
678 norm_path
= os
.path
.normpath(remove_start(s
, drive_or_unc
)).split(os
.path
.sep
)
682 path_part
if path_part
in ['.', '..'] else re
.sub(r
'(?:[/<>:"\|\\?\*]|[\s.]$)', '#', path_part
)
683 for path_part
in norm_path
]
685 sanitized_path
.insert(0, drive_or_unc
+ os
.path
.sep
)
686 elif force
and s
and s
[0] == os
.path
.sep
:
687 sanitized_path
.insert(0, os
.path
.sep
)
688 # TODO: Fix behavioral differences <3.12
689 # The workaround using `normpath` only superficially passes tests
690 # Ref: https://github.com/python/cpython/pull/100351
691 return os
.path
.normpath(os
.path
.join(*sanitized_path
))
694 def sanitize_url(url
, *, scheme
='http'):
695 # Prepend protocol-less URLs with `http:` scheme in order to mitigate
696 # the number of unwanted failures due to missing protocol
699 elif url
.startswith('//'):
700 return f
'{scheme}:{url}'
701 # Fix some common typos seen so far
703 # https://github.com/ytdl-org/youtube-dl/issues/15649
704 (r
'^httpss://', r
'https://'),
705 # https://bx1.be/lives/direct-tv/
706 (r
'^rmtp([es]?)://', r
'rtmp\1://'),
708 for mistake
, fixup
in COMMON_TYPOS
:
709 if re
.match(mistake
, url
):
710 return re
.sub(mistake
, fixup
, url
)
714 def extract_basic_auth(url
):
715 parts
= urllib
.parse
.urlsplit(url
)
716 if parts
.username
is None:
718 url
= urllib
.parse
.urlunsplit(parts
._replace
(netloc
=(
719 parts
.hostname
if parts
.port
is None
720 else '%s:%d' % (parts
.hostname
, parts
.port
))))
721 auth_payload
= base64
.b64encode(
722 ('%s:%s' % (parts
.username
, parts
.password
or '')).encode())
723 return url
, f
'Basic {auth_payload.decode()}'
727 """Expand shell variables and ~"""
728 return os
.path
.expandvars(compat_expanduser(s
))
731 def orderedSet(iterable
, *, lazy
=False):
732 """Remove all duplicates from the input iterable"""
734 seen
= [] # Do not use set since the items can be unhashable
740 return _iter() if lazy
else list(_iter())
743 def _htmlentity_transform(entity_with_semicolon
):
744 """Transforms an HTML entity to a character."""
745 entity
= entity_with_semicolon
[:-1]
747 # Known non-numeric HTML entity
748 if entity
in html
.entities
.name2codepoint
:
749 return chr(html
.entities
.name2codepoint
[entity
])
751 # TODO: HTML5 allows entities without a semicolon.
752 # E.g. 'Éric' should be decoded as 'Éric'.
753 if entity_with_semicolon
in html
.entities
.html5
:
754 return html
.entities
.html5
[entity_with_semicolon
]
756 mobj
= re
.match(r
'#(x[0-9a-fA-F]+|[0-9]+)', entity
)
758 numstr
= mobj
.group(1)
759 if numstr
.startswith('x'):
761 numstr
= '0%s' % numstr
764 # See https://github.com/ytdl-org/youtube-dl/issues/7518
765 with contextlib
.suppress(ValueError):
766 return chr(int(numstr
, base
))
768 # Unknown entity in name, return its literal representation
769 return '&%s;' % entity
775 assert isinstance(s
, str)
778 r
'&([^&;]+;)', lambda m
: _htmlentity_transform(m
.group(1)), s
)
781 def escapeHTML(text
):
784 .replace('&', '&')
785 .replace('<', '<')
786 .replace('>', '>')
787 .replace('"', '"')
788 .replace("'", ''')
792 class netrc_from_content(netrc
.netrc
):
793 def __init__(self
, content
):
794 self
.hosts
, self
.macros
= {}, {}
795 with io
.StringIO(content
) as stream
:
796 self
._parse
('-', stream
, False)
799 class Popen(subprocess
.Popen
):
800 if sys
.platform
== 'win32':
801 _startupinfo
= subprocess
.STARTUPINFO()
802 _startupinfo
.dwFlags |
= subprocess
.STARTF_USESHOWWINDOW
807 def _fix_pyinstaller_ld_path(env
):
808 """Restore LD_LIBRARY_PATH when using PyInstaller
809 Ref: https://github.com/pyinstaller/pyinstaller/blob/develop/doc/runtime-information.rst#ld_library_path--libpath-considerations
810 https://github.com/yt-dlp/yt-dlp/issues/4573
812 if not hasattr(sys
, '_MEIPASS'):
816 orig
= env
.get(f
'{key}_ORIG')
822 _fix('LD_LIBRARY_PATH') # Linux
823 _fix('DYLD_LIBRARY_PATH') # macOS
825 def __init__(self
, args
, *remaining
, env
=None, text
=False, shell
=False, **kwargs
):
827 env
= os
.environ
.copy()
828 self
._fix
_pyinstaller
_ld
_path
(env
)
830 self
.__text
_mode
= kwargs
.get('encoding') or kwargs
.get('errors') or text
or kwargs
.get('universal_newlines')
832 kwargs
['universal_newlines'] = True # For 3.6 compatibility
833 kwargs
.setdefault('encoding', 'utf-8')
834 kwargs
.setdefault('errors', 'replace')
836 if shell
and compat_os_name
== 'nt' and kwargs
.get('executable') is None:
837 if not isinstance(args
, str):
838 args
= shell_quote(args
, shell
=True)
840 # Set variable for `cmd.exe` newline escaping (see `utils.shell_quote`)
842 args
= f
'{self.__comspec()} /Q /S /D /V:OFF /E:ON /C "{args}"'
844 super().__init
__(args
, *remaining
, env
=env
, shell
=shell
, **kwargs
, startupinfo
=self
._startupinfo
)
847 comspec
= os
.environ
.get('ComSpec') or os
.path
.join(
848 os
.environ
.get('SystemRoot', ''), 'System32', 'cmd.exe')
849 if os
.path
.isabs(comspec
):
851 raise FileNotFoundError('shell not found: neither %ComSpec% nor %SystemRoot% is set')
853 def communicate_or_kill(self
, *args
, **kwargs
):
855 return self
.communicate(*args
, **kwargs
)
856 except BaseException
: # Including KeyboardInterrupt
857 self
.kill(timeout
=None)
860 def kill(self
, *, timeout
=0):
863 self
.wait(timeout
=timeout
)
866 def run(cls
, *args
, timeout
=None, **kwargs
):
867 with cls(*args
, **kwargs
) as proc
:
868 default
= '' if proc
.__text
_mode
else b
''
869 stdout
, stderr
= proc
.communicate_or_kill(timeout
=timeout
)
870 return stdout
or default
, stderr
or default
, proc
.returncode
873 def encodeArgument(s
):
874 # Legacy code that uses byte strings
875 # Uncomment the following line after fixing all post processors
876 # assert isinstance(s, str), 'Internal error: %r should be of type %r, is %r' % (s, str, type(s))
877 return s
if isinstance(s
, str) else s
.decode('ascii')
880 _timetuple
= collections
.namedtuple('Time', ('hours', 'minutes', 'seconds', 'milliseconds'))
883 def timetuple_from_msec(msec
):
884 secs
, msec
= divmod(msec
, 1000)
885 mins
, secs
= divmod(secs
, 60)
886 hrs
, mins
= divmod(mins
, 60)
887 return _timetuple(hrs
, mins
, secs
, msec
)
890 def formatSeconds(secs
, delim
=':', msec
=False):
891 time
= timetuple_from_msec(secs
* 1000)
893 ret
= '%d%s%02d%s%02d' % (time
.hours
, delim
, time
.minutes
, delim
, time
.seconds
)
895 ret
= '%d%s%02d' % (time
.minutes
, delim
, time
.seconds
)
897 ret
= '%d' % time
.seconds
898 return '%s.%03d' % (ret
, time
.milliseconds
) if msec
else ret
901 def bug_reports_message(before
=';'):
902 from ..update
import REPOSITORY
904 msg
= (f
'please report this issue on https://github.com/{REPOSITORY}/issues?q= , '
905 'filling out the appropriate issue template. Confirm you are on the latest version using yt-dlp -U')
907 before
= before
.rstrip()
908 if not before
or before
.endswith(('.', '!', '?')):
909 msg
= msg
[0].title() + msg
[1:]
911 return (before
+ ' ' if before
else '') + msg
914 class YoutubeDLError(Exception):
915 """Base exception for YoutubeDL errors."""
918 def __init__(self
, msg
=None):
921 elif self
.msg
is None:
922 self
.msg
= type(self
).__name
__
923 super().__init
__(self
.msg
)
926 class ExtractorError(YoutubeDLError
):
927 """Error during info extraction."""
929 def __init__(self
, msg
, tb
=None, expected
=False, cause
=None, video_id
=None, ie
=None):
930 """ tb, if given, is the original traceback (so that it can be printed out).
931 If expected is set, this is a normal error message and most likely not a bug in yt-dlp.
933 from ..networking
.exceptions
import network_exceptions
934 if sys
.exc_info()[0] in network_exceptions
:
937 self
.orig_msg
= str(msg
)
939 self
.expected
= expected
941 self
.video_id
= video_id
943 self
.exc_info
= sys
.exc_info() # preserve original exception
944 if isinstance(self
.exc_info
[1], ExtractorError
):
945 self
.exc_info
= self
.exc_info
[1].exc_info
946 super().__init
__(self
.__msg
)
951 format_field(self
.ie
, None, '[%s] '),
952 format_field(self
.video_id
, None, '%s: '),
954 format_field(self
.cause
, None, ' (caused by %r)'),
955 '' if self
.expected
else bug_reports_message()))
957 def format_traceback(self
):
958 return join_nonempty(
959 self
.traceback
and ''.join(traceback
.format_tb(self
.traceback
)),
960 self
.cause
and ''.join(traceback
.format_exception(None, self
.cause
, self
.cause
.__traceback
__)[1:]),
963 def __setattr__(self
, name
, value
):
964 super().__setattr
__(name
, value
)
965 if getattr(self
, 'msg', None) and name
not in ('msg', 'args'):
966 self
.msg
= self
.__msg
or type(self
).__name
__
967 self
.args
= (self
.msg
, ) # Cannot be property
970 class UnsupportedError(ExtractorError
):
971 def __init__(self
, url
):
973 'Unsupported URL: %s' % url
, expected
=True)
977 class RegexNotFoundError(ExtractorError
):
978 """Error when a regex didn't match"""
982 class GeoRestrictedError(ExtractorError
):
983 """Geographic restriction Error exception.
985 This exception may be thrown when a video is not available from your
986 geographic location due to geographic restrictions imposed by a website.
989 def __init__(self
, msg
, countries
=None, **kwargs
):
990 kwargs
['expected'] = True
991 super().__init
__(msg
, **kwargs
)
992 self
.countries
= countries
995 class UserNotLive(ExtractorError
):
996 """Error when a channel/user is not live"""
998 def __init__(self
, msg
=None, **kwargs
):
999 kwargs
['expected'] = True
1000 super().__init
__(msg
or 'The channel is not currently live', **kwargs
)
1003 class DownloadError(YoutubeDLError
):
1004 """Download Error exception.
1006 This exception may be thrown by FileDownloader objects if they are not
1007 configured to continue on errors. They will contain the appropriate
1011 def __init__(self
, msg
, exc_info
=None):
1012 """ exc_info, if given, is the original exception that caused the trouble (as returned by sys.exc_info()). """
1013 super().__init
__(msg
)
1014 self
.exc_info
= exc_info
1017 class EntryNotInPlaylist(YoutubeDLError
):
1018 """Entry not in playlist exception.
1020 This exception will be thrown by YoutubeDL when a requested entry
1021 is not found in the playlist info_dict
1023 msg
= 'Entry not found in info'
1026 class SameFileError(YoutubeDLError
):
1027 """Same File exception.
1029 This exception will be thrown by FileDownloader objects if they detect
1030 multiple files would have to be downloaded to the same file on disk.
1032 msg
= 'Fixed output name but more than one file to download'
1034 def __init__(self
, filename
=None):
1035 if filename
is not None:
1036 self
.msg
+= f
': {filename}'
1037 super().__init
__(self
.msg
)
1040 class PostProcessingError(YoutubeDLError
):
1041 """Post Processing exception.
1043 This exception may be raised by PostProcessor's .run() method to
1044 indicate an error in the postprocessing task.
1048 class DownloadCancelled(YoutubeDLError
):
1049 """ Exception raised when the download queue should be interrupted """
1050 msg
= 'The download was cancelled'
1053 class ExistingVideoReached(DownloadCancelled
):
1054 """ --break-on-existing triggered """
1055 msg
= 'Encountered a video that is already in the archive, stopping due to --break-on-existing'
1058 class RejectedVideoReached(DownloadCancelled
):
1059 """ --break-match-filter triggered """
1060 msg
= 'Encountered a video that did not match filter, stopping due to --break-match-filter'
1063 class MaxDownloadsReached(DownloadCancelled
):
1064 """ --max-downloads limit has been reached. """
1065 msg
= 'Maximum number of downloads reached, stopping due to --max-downloads'
1068 class ReExtractInfo(YoutubeDLError
):
1069 """ Video info needs to be re-extracted. """
1071 def __init__(self
, msg
, expected
=False):
1072 super().__init
__(msg
)
1073 self
.expected
= expected
1076 class ThrottledDownload(ReExtractInfo
):
1077 """ Download speed below --throttled-rate. """
1078 msg
= 'The download speed is below throttle limit'
1081 super().__init
__(self
.msg
, expected
=False)
1084 class UnavailableVideoError(YoutubeDLError
):
1085 """Unavailable Format exception.
1087 This exception will be thrown when a video is requested
1088 in a format that is not available for that video.
1090 msg
= 'Unable to download video'
1092 def __init__(self
, err
=None):
1094 self
.msg
+= f
': {err}'
1095 super().__init
__(self
.msg
)
1098 class ContentTooShortError(YoutubeDLError
):
1099 """Content Too Short exception.
1101 This exception may be raised by FileDownloader objects when a file they
1102 download is too small for what the server announced first, indicating
1103 the connection was probably interrupted.
1106 def __init__(self
, downloaded
, expected
):
1107 super().__init
__(f
'Downloaded {downloaded} bytes, expected {expected} bytes')
1109 self
.downloaded
= downloaded
1110 self
.expected
= expected
1113 class XAttrMetadataError(YoutubeDLError
):
1114 def __init__(self
, code
=None, msg
='Unknown error'):
1115 super().__init
__(msg
)
1119 # Parsing code and msg
1120 if (self
.code
in (errno
.ENOSPC
, errno
.EDQUOT
)
1121 or 'No space left' in self
.msg
or 'Disk quota exceeded' in self
.msg
):
1122 self
.reason
= 'NO_SPACE'
1123 elif self
.code
== errno
.E2BIG
or 'Argument list too long' in self
.msg
:
1124 self
.reason
= 'VALUE_TOO_LONG'
1126 self
.reason
= 'NOT_SUPPORTED'
1129 class XAttrUnavailableError(YoutubeDLError
):
1133 def is_path_like(f
):
1134 return isinstance(f
, (str, bytes, os
.PathLike
))
1137 def extract_timezone(date_str
):
1140 ^.{8,}? # >=8 char non-TZ prefix, if present
1141 (?P<tz>Z| # just the UTC Z, or
1142 (?:(?<=.\b\d{4}|\b\d{2}:\d\d)| # preceded by 4 digits or hh:mm or
1143 (?<!.\b[a-zA-Z]{3}|[a-zA-Z]{4}|..\b\d\d)) # not preceded by 3 alpha word or >= 4 alpha or 2 digits
1144 [ ]? # optional space
1145 (?P<sign>\+|-) # +/-
1146 (?P<hours>[0-9]{2}):?(?P<minutes>[0-9]{2}) # hh[:]mm
1150 m
= re
.search(r
'\d{1,2}:\d{1,2}(?:\.\d+)?(?P<tz>\s*[A-Z]+)$', date_str
)
1151 timezone
= TIMEZONE_NAMES
.get(m
and m
.group('tz').strip())
1152 if timezone
is not None:
1153 date_str
= date_str
[:-len(m
.group('tz'))]
1154 timezone
= dt
.timedelta(hours
=timezone
or 0)
1156 date_str
= date_str
[:-len(m
.group('tz'))]
1157 if not m
.group('sign'):
1158 timezone
= dt
.timedelta()
1160 sign
= 1 if m
.group('sign') == '+' else -1
1161 timezone
= dt
.timedelta(
1162 hours
=sign
* int(m
.group('hours')),
1163 minutes
=sign
* int(m
.group('minutes')))
1164 return timezone
, date_str
1167 def parse_iso8601(date_str
, delimiter
='T', timezone
=None):
1168 """ Return a UNIX timestamp from the given date """
1170 if date_str
is None:
1173 date_str
= re
.sub(r
'\.[0-9]+', '', date_str
)
1175 if timezone
is None:
1176 timezone
, date_str
= extract_timezone(date_str
)
1178 with contextlib
.suppress(ValueError):
1179 date_format
= f
'%Y-%m-%d{delimiter}%H:%M:%S'
1180 dt_
= dt
.datetime
.strptime(date_str
, date_format
) - timezone
1181 return calendar
.timegm(dt_
.timetuple())
1184 def date_formats(day_first
=True):
1185 return DATE_FORMATS_DAY_FIRST
if day_first
else DATE_FORMATS_MONTH_FIRST
1188 def unified_strdate(date_str
, day_first
=True):
1189 """Return a string with the date in the format YYYYMMDD"""
1191 if date_str
is None:
1195 date_str
= date_str
.replace(',', ' ')
1196 # Remove AM/PM + timezone
1197 date_str
= re
.sub(r
'(?i)\s*(?:AM|PM)(?:\s+[A-Z]+)?', '', date_str
)
1198 _
, date_str
= extract_timezone(date_str
)
1200 for expression
in date_formats(day_first
):
1201 with contextlib
.suppress(ValueError):
1202 upload_date
= dt
.datetime
.strptime(date_str
, expression
).strftime('%Y%m%d')
1203 if upload_date
is None:
1204 timetuple
= email
.utils
.parsedate_tz(date_str
)
1206 with contextlib
.suppress(ValueError):
1207 upload_date
= dt
.datetime(*timetuple
[:6]).strftime('%Y%m%d')
1208 if upload_date
is not None:
1209 return str(upload_date
)
1212 def unified_timestamp(date_str
, day_first
=True):
1213 if not isinstance(date_str
, str):
1216 date_str
= re
.sub(r
'\s+', ' ', re
.sub(
1217 r
'(?i)[,|]|(mon|tues?|wed(nes)?|thu(rs)?|fri|sat(ur)?)(day)?', '', date_str
))
1219 pm_delta
= 12 if re
.search(r
'(?i)PM', date_str
) else 0
1220 timezone
, date_str
= extract_timezone(date_str
)
1222 # Remove AM/PM + timezone
1223 date_str
= re
.sub(r
'(?i)\s*(?:AM|PM)(?:\s+[A-Z]+)?', '', date_str
)
1225 # Remove unrecognized timezones from ISO 8601 alike timestamps
1226 m
= re
.search(r
'\d{1,2}:\d{1,2}(?:\.\d+)?(?P<tz>\s*[A-Z]+)$', date_str
)
1228 date_str
= date_str
[:-len(m
.group('tz'))]
1230 # Python only supports microseconds, so remove nanoseconds
1231 m
= re
.search(r
'^([0-9]{4,}-[0-9]{1,2}-[0-9]{1,2}T[0-9]{1,2}:[0-9]{1,2}:[0-9]{1,2}\.[0-9]{6})[0-9]+$', date_str
)
1233 date_str
= m
.group(1)
1235 for expression
in date_formats(day_first
):
1236 with contextlib
.suppress(ValueError):
1237 dt_
= dt
.datetime
.strptime(date_str
, expression
) - timezone
+ dt
.timedelta(hours
=pm_delta
)
1238 return calendar
.timegm(dt_
.timetuple())
1240 timetuple
= email
.utils
.parsedate_tz(date_str
)
1242 return calendar
.timegm(timetuple
) + pm_delta
* 3600 - timezone
.total_seconds()
1245 def determine_ext(url
, default_ext
='unknown_video'):
1246 if url
is None or '.' not in url
:
1248 guess
= url
.partition('?')[0].rpartition('.')[2]
1249 if re
.match(r
'^[A-Za-z0-9]+$', guess
):
1251 # Try extract ext from URLs like http://example.com/foo/bar.mp4/?download
1252 elif guess
.rstrip('/') in KNOWN_EXTENSIONS
:
1253 return guess
.rstrip('/')
1258 def subtitles_filename(filename
, sub_lang
, sub_format
, expected_real_ext
=None):
1259 return replace_extension(filename
, sub_lang
+ '.' + sub_format
, expected_real_ext
)
1262 def datetime_from_str(date_str
, precision
='auto', format
='%Y%m%d'):
1264 Return a datetime object from a string.
1266 (now|today|yesterday|DATE)([+-]\d+(microsecond|second|minute|hour|day|week|month|year)s?)?
1268 @param format strftime format of DATE
1269 @param precision Round the datetime object: auto|microsecond|second|minute|hour|day
1270 auto: round to the unit provided in date_str (if applicable).
1272 auto_precision
= False
1273 if precision
== 'auto':
1274 auto_precision
= True
1275 precision
= 'microsecond'
1276 today
= datetime_round(dt
.datetime
.now(dt
.timezone
.utc
), precision
)
1277 if date_str
in ('now', 'today'):
1279 if date_str
== 'yesterday':
1280 return today
- dt
.timedelta(days
=1)
1282 r
'(?P<start>.+)(?P<sign>[+-])(?P<time>\d+)(?P<unit>microsecond|second|minute|hour|day|week|month|year)s?',
1284 if match
is not None:
1285 start_time
= datetime_from_str(match
.group('start'), precision
, format
)
1286 time
= int(match
.group('time')) * (-1 if match
.group('sign') == '-' else 1)
1287 unit
= match
.group('unit')
1288 if unit
== 'month' or unit
== 'year':
1289 new_date
= datetime_add_months(start_time
, time
* 12 if unit
== 'year' else time
)
1295 delta
= dt
.timedelta(**{unit + 's': time}
)
1296 new_date
= start_time
+ delta
1298 return datetime_round(new_date
, unit
)
1301 return datetime_round(dt
.datetime
.strptime(date_str
, format
), precision
)
1304 def date_from_str(date_str
, format
='%Y%m%d', strict
=False):
1306 Return a date object from a string using datetime_from_str
1308 @param strict Restrict allowed patterns to "YYYYMMDD" and
1309 (now|today|yesterday)(-\d+(day|week|month|year)s?)?
1311 if strict
and not re
.fullmatch(r
'\d{8}|(now|today|yesterday)(-\d+(day|week|month|year)s?)?', date_str
):
1312 raise ValueError(f
'Invalid date format "{date_str}"')
1313 return datetime_from_str(date_str
, precision
='microsecond', format
=format
).date()
1316 def datetime_add_months(dt_
, months
):
1317 """Increment/Decrement a datetime object by months."""
1318 month
= dt_
.month
+ months
- 1
1319 year
= dt_
.year
+ month
// 12
1320 month
= month
% 12 + 1
1321 day
= min(dt_
.day
, calendar
.monthrange(year
, month
)[1])
1322 return dt_
.replace(year
, month
, day
)
1325 def datetime_round(dt_
, precision
='day'):
1327 Round a datetime object's time to a specific precision
1329 if precision
== 'microsecond':
1338 roundto
= lambda x
, n
: ((x
+ n
/ 2) // n
) * n
1339 timestamp
= roundto(calendar
.timegm(dt_
.timetuple()), unit_seconds
[precision
])
1340 return dt
.datetime
.fromtimestamp(timestamp
, dt
.timezone
.utc
)
1343 def hyphenate_date(date_str
):
1345 Convert a date in 'YYYYMMDD' format to 'YYYY-MM-DD' format"""
1346 match
= re
.match(r
'^(\d\d\d\d)(\d\d)(\d\d)$', date_str
)
1347 if match
is not None:
1348 return '-'.join(match
.groups())
1354 """Represents a time interval between two dates"""
1356 def __init__(self
, start
=None, end
=None):
1357 """start and end must be strings in the format accepted by date"""
1358 if start
is not None:
1359 self
.start
= date_from_str(start
, strict
=True)
1361 self
.start
= dt
.datetime
.min.date()
1363 self
.end
= date_from_str(end
, strict
=True)
1365 self
.end
= dt
.datetime
.max.date()
1366 if self
.start
> self
.end
:
1367 raise ValueError('Date range: "%s" , the start date must be before the end date' % self
)
1371 """Returns a range that only contains the given day"""
1372 return cls(day
, day
)
1374 def __contains__(self
, date
):
1375 """Check if the date is in the range"""
1376 if not isinstance(date
, dt
.date
):
1377 date
= date_from_str(date
)
1378 return self
.start
<= date
<= self
.end
1381 return f
'{__name__}.{type(self).__name__}({self.start.isoformat()!r}, {self.end.isoformat()!r})'
1384 return f
'{self.start} to {self.end}'
1386 def __eq__(self
, other
):
1387 return (isinstance(other
, DateRange
)
1388 and self
.start
== other
.start
and self
.end
== other
.end
)
1392 def system_identifier():
1393 python_implementation
= platform
.python_implementation()
1394 if python_implementation
== 'PyPy' and hasattr(sys
, 'pypy_version_info'):
1395 python_implementation
+= ' version %d.%d.%d' % sys
.pypy_version_info
[:3]
1397 with contextlib
.suppress(OSError): # We may not have access to the executable
1398 libc_ver
= platform
.libc_ver()
1400 return 'Python %s (%s %s %s) - %s (%s%s)' % (
1401 platform
.python_version(),
1402 python_implementation
,
1404 platform
.architecture()[0],
1405 platform
.platform(),
1406 ssl
.OPENSSL_VERSION
,
1407 format_field(join_nonempty(*libc_ver
, delim
=' '), None, ', %s'),
1412 def get_windows_version():
1413 ''' Get Windows version. returns () if it's not running on Windows '''
1414 if compat_os_name
== 'nt':
1415 return version_tuple(platform
.win32_ver()[1])
1420 def write_string(s
, out
=None, encoding
=None):
1421 assert isinstance(s
, str)
1422 out
= out
or sys
.stderr
1423 # `sys.stderr` might be `None` (Ref: https://github.com/pyinstaller/pyinstaller/pull/7217)
1427 if compat_os_name
== 'nt' and supports_terminal_sequences(out
):
1428 s
= re
.sub(r
'([\r\n]+)', r
' \1', s
)
1430 enc
, buffer = None, out
1431 # `mode` might be `None` (Ref: https://github.com/yt-dlp/yt-dlp/issues/8816)
1432 if 'b' in (getattr(out
, 'mode', None) or ''):
1433 enc
= encoding
or preferredencoding()
1434 elif hasattr(out
, 'buffer'):
1436 enc
= encoding
or getattr(out
, 'encoding', None) or preferredencoding()
1438 buffer.write(s
.encode(enc
, 'ignore') if enc
else s
)
1442 # TODO: Use global logger
1443 def deprecation_warning(msg
, *, printer
=None, stacklevel
=0, **kwargs
):
1444 from .. import _IN_CLI
1446 if msg
in deprecation_warning
._cache
:
1448 deprecation_warning
._cache
.add(msg
)
1450 return printer(f
'{msg}{bug_reports_message()}', **kwargs
)
1451 return write_string(f
'ERROR: {msg}{bug_reports_message()}\n', **kwargs
)
1454 warnings
.warn(DeprecationWarning(msg
), stacklevel
=stacklevel
+ 3)
1457 deprecation_warning
._cache
= set()
1460 def bytes_to_intlist(bs
):
1463 if isinstance(bs
[0], int): # Python 3
1466 return [ord(c
) for c
in bs
]
1469 def intlist_to_bytes(xs
):
1472 return struct
.pack('%dB' % len(xs
), *xs
)
1475 class LockingUnsupportedError(OSError):
1476 msg
= 'File locking is not supported'
1479 super().__init
__(self
.msg
)
1482 # Cross-platform file locking
1483 if sys
.platform
== 'win32':
1485 import ctypes
.wintypes
1488 class OVERLAPPED(ctypes
.Structure
):
1490 ('Internal', ctypes
.wintypes
.LPVOID
),
1491 ('InternalHigh', ctypes
.wintypes
.LPVOID
),
1492 ('Offset', ctypes
.wintypes
.DWORD
),
1493 ('OffsetHigh', ctypes
.wintypes
.DWORD
),
1494 ('hEvent', ctypes
.wintypes
.HANDLE
),
1497 kernel32
= ctypes
.WinDLL('kernel32')
1498 LockFileEx
= kernel32
.LockFileEx
1499 LockFileEx
.argtypes
= [
1500 ctypes
.wintypes
.HANDLE
, # hFile
1501 ctypes
.wintypes
.DWORD
, # dwFlags
1502 ctypes
.wintypes
.DWORD
, # dwReserved
1503 ctypes
.wintypes
.DWORD
, # nNumberOfBytesToLockLow
1504 ctypes
.wintypes
.DWORD
, # nNumberOfBytesToLockHigh
1505 ctypes
.POINTER(OVERLAPPED
) # Overlapped
1507 LockFileEx
.restype
= ctypes
.wintypes
.BOOL
1508 UnlockFileEx
= kernel32
.UnlockFileEx
1509 UnlockFileEx
.argtypes
= [
1510 ctypes
.wintypes
.HANDLE
, # hFile
1511 ctypes
.wintypes
.DWORD
, # dwReserved
1512 ctypes
.wintypes
.DWORD
, # nNumberOfBytesToLockLow
1513 ctypes
.wintypes
.DWORD
, # nNumberOfBytesToLockHigh
1514 ctypes
.POINTER(OVERLAPPED
) # Overlapped
1516 UnlockFileEx
.restype
= ctypes
.wintypes
.BOOL
1517 whole_low
= 0xffffffff
1518 whole_high
= 0x7fffffff
1520 def _lock_file(f
, exclusive
, block
):
1521 overlapped
= OVERLAPPED()
1522 overlapped
.Offset
= 0
1523 overlapped
.OffsetHigh
= 0
1524 overlapped
.hEvent
= 0
1525 f
._lock
_file
_overlapped
_p
= ctypes
.pointer(overlapped
)
1527 if not LockFileEx(msvcrt
.get_osfhandle(f
.fileno()),
1528 (0x2 if exclusive
else 0x0) |
(0x0 if block
else 0x1),
1529 0, whole_low
, whole_high
, f
._lock
_file
_overlapped
_p
):
1530 # NB: No argument form of "ctypes.FormatError" does not work on PyPy
1531 raise BlockingIOError(f
'Locking file failed: {ctypes.FormatError(ctypes.GetLastError())!r}')
1533 def _unlock_file(f
):
1534 assert f
._lock
_file
_overlapped
_p
1535 handle
= msvcrt
.get_osfhandle(f
.fileno())
1536 if not UnlockFileEx(handle
, 0, whole_low
, whole_high
, f
._lock
_file
_overlapped
_p
):
1537 raise OSError('Unlocking file failed: %r' % ctypes
.FormatError())
1543 def _lock_file(f
, exclusive
, block
):
1544 flags
= fcntl
.LOCK_EX
if exclusive
else fcntl
.LOCK_SH
1546 flags |
= fcntl
.LOCK_NB
1548 fcntl
.flock(f
, flags
)
1549 except BlockingIOError
:
1551 except OSError: # AOSP does not have flock()
1552 fcntl
.lockf(f
, flags
)
1554 def _unlock_file(f
):
1555 with contextlib
.suppress(OSError):
1556 return fcntl
.flock(f
, fcntl
.LOCK_UN
)
1557 with contextlib
.suppress(OSError):
1558 return fcntl
.lockf(f
, fcntl
.LOCK_UN
) # AOSP does not have flock()
1559 return fcntl
.flock(f
, fcntl
.LOCK_UN | fcntl
.LOCK_NB
) # virtiofs needs LOCK_NB on unlocking
1563 def _lock_file(f
, exclusive
, block
):
1564 raise LockingUnsupportedError()
1566 def _unlock_file(f
):
1567 raise LockingUnsupportedError()
1573 def __init__(self
, filename
, mode
, block
=True, encoding
=None):
1574 if mode
not in {'r', 'rb', 'a', 'ab', 'w', 'wb'}
:
1575 raise NotImplementedError(mode
)
1576 self
.mode
, self
.block
= mode
, block
1578 writable
= any(f
in mode
for f
in 'wax+')
1579 readable
= any(f
in mode
for f
in 'r+')
1580 flags
= functools
.reduce(operator
.ior
, (
1581 getattr(os
, 'O_CLOEXEC', 0), # UNIX only
1582 getattr(os
, 'O_BINARY', 0), # Windows only
1583 getattr(os
, 'O_NOINHERIT', 0), # Windows only
1584 os
.O_CREAT
if writable
else 0, # O_TRUNC only after locking
1585 os
.O_APPEND
if 'a' in mode
else 0,
1586 os
.O_EXCL
if 'x' in mode
else 0,
1587 os
.O_RDONLY
if not writable
else os
.O_RDWR
if readable
else os
.O_WRONLY
,
1590 self
.f
= os
.fdopen(os
.open(filename
, flags
, 0o666), mode
, encoding
=encoding
)
1592 def __enter__(self
):
1593 exclusive
= 'r' not in self
.mode
1595 _lock_file(self
.f
, exclusive
, self
.block
)
1600 if 'w' in self
.mode
:
1603 except OSError as e
:
1605 errno
.ESPIPE
, # Illegal seek - expected for FIFO
1606 errno
.EINVAL
, # Invalid argument - expected for /dev/null
1615 _unlock_file(self
.f
)
1619 def __exit__(self
, *_
):
1628 def __getattr__(self
, attr
):
1629 return getattr(self
.f
, attr
)
1636 def get_filesystem_encoding():
1637 encoding
= sys
.getfilesystemencoding()
1638 return encoding
if encoding
is not None else 'utf-8'
1641 _WINDOWS_QUOTE_TRANS
= str.maketrans({'"': '\\"', '\\': '\\\\'}
)
1642 _CMD_QUOTE_TRANS
= str.maketrans({
1643 # Keep quotes balanced by replacing them with `""` instead of `\\"`
1645 # Requires a variable `=` containing `"^\n\n"` (set in `utils.Popen`)
1646 # `=` should be unique since variables containing `=` cannot be set using cmd
1648 # While we are only required to escape backslashes immediately before quotes,
1649 # we instead escape all of 'em anyways to be consistent
1651 # Use zero length variable replacement so `%` doesn't get expanded
1652 # `cd` is always set as long as extensions are enabled (`/E:ON` in `utils.Popen`)
1657 def shell_quote(args
, *, shell
=False):
1658 args
= list(variadic(args
))
1659 if any(isinstance(item
, bytes) for item
in args
):
1660 deprecation_warning('Passing bytes to utils.shell_quote is deprecated')
1661 encoding
= get_filesystem_encoding()
1662 for index
, item
in enumerate(args
):
1663 if isinstance(item
, bytes):
1664 args
[index
] = item
.decode(encoding
)
1666 if compat_os_name
!= 'nt':
1667 return shlex
.join(args
)
1669 trans
= _CMD_QUOTE_TRANS
if shell
else _WINDOWS_QUOTE_TRANS
1671 s
if re
.fullmatch(r
'[\w#$*\-+./:?@\\]+', s
, re
.ASCII
) else s
.translate(trans
).join('""')
1675 def smuggle_url(url
, data
):
1676 """ Pass additional data in a URL for internal use. """
1678 url
, idata
= unsmuggle_url(url
, {})
1680 sdata
= urllib
.parse
.urlencode(
1681 {'__youtubedl_smuggle': json.dumps(data)}
)
1682 return url
+ '#' + sdata
1685 def unsmuggle_url(smug_url
, default
=None):
1686 if '#__youtubedl_smuggle' not in smug_url
:
1687 return smug_url
, default
1688 url
, _
, sdata
= smug_url
.rpartition('#')
1689 jsond
= urllib
.parse
.parse_qs(sdata
)['__youtubedl_smuggle'][0]
1690 data
= json
.loads(jsond
)
1694 def format_decimal_suffix(num
, fmt
='%d%s', *, factor
=1000):
1695 """ Formats numbers with decimal sufixes like K, M, etc """
1696 num
, factor
= float_or_none(num
), float(factor
)
1697 if num
is None or num
< 0:
1699 POSSIBLE_SUFFIXES
= 'kMGTPEZY'
1700 exponent
= 0 if num
== 0 else min(int(math
.log(num
, factor
)), len(POSSIBLE_SUFFIXES
))
1701 suffix
= ['', *POSSIBLE_SUFFIXES
][exponent
]
1703 suffix
= {'k': 'Ki', '': ''}
.get(suffix
, f
'{suffix}i')
1704 converted
= num
/ (factor
** exponent
)
1705 return fmt
% (converted
, suffix
)
1708 def format_bytes(bytes):
1709 return format_decimal_suffix(bytes, '%.2f%sB', factor
=1024) or 'N/A'
1712 def lookup_unit_table(unit_table
, s
, strict
=False):
1713 num_re
= NUMBER_RE
if strict
else NUMBER_RE
.replace(R
'\.', '[,.]')
1714 units_re
= '|'.join(re
.escape(u
) for u
in unit_table
)
1715 m
= (re
.fullmatch
if strict
else re
.match
)(
1716 rf
'(?P<num>{num_re})\s*(?P<unit>{units_re})\b', s
)
1720 num
= float(m
.group('num').replace(',', '.'))
1721 mult
= unit_table
[m
.group('unit')]
1722 return round(num
* mult
)
1726 """Parse a string indicating a byte quantity into an integer"""
1727 return lookup_unit_table(
1728 {u: 1024**i for i, u in enumerate(['', *'KMGTPEZY'])}
,
1729 s
.upper(), strict
=True)
1732 def parse_filesize(s
):
1736 # The lower-case forms are of course incorrect and unofficial,
1737 # but we support those too
1754 'megabytes': 1000 ** 2,
1755 'mebibytes': 1024 ** 2,
1761 'gigabytes': 1000 ** 3,
1762 'gibibytes': 1024 ** 3,
1768 'terabytes': 1000 ** 4,
1769 'tebibytes': 1024 ** 4,
1775 'petabytes': 1000 ** 5,
1776 'pebibytes': 1024 ** 5,
1782 'exabytes': 1000 ** 6,
1783 'exbibytes': 1024 ** 6,
1789 'zettabytes': 1000 ** 7,
1790 'zebibytes': 1024 ** 7,
1796 'yottabytes': 1000 ** 8,
1797 'yobibytes': 1024 ** 8,
1800 return lookup_unit_table(_UNIT_TABLE
, s
)
1807 s
= re
.sub(r
'^[^\d]+\s', '', s
).strip()
1809 if re
.match(r
'^[\d,.]+$', s
):
1810 return str_to_int(s
)
1823 ret
= lookup_unit_table(_UNIT_TABLE
, s
)
1827 mobj
= re
.match(r
'([\d,.]+)(?:$|\s)', s
)
1829 return str_to_int(mobj
.group(1))
1832 def parse_resolution(s
, *, lenient
=False):
1837 mobj
= re
.search(r
'(?P<w>\d+)\s*[xX×,]\s*(?P<h>\d+)', s
)
1839 mobj
= re
.search(r
'(?<![a-zA-Z0-9])(?P<w>\d+)\s*[xX×,]\s*(?P<h>\d+)(?![a-zA-Z0-9])', s
)
1842 'width': int(mobj
.group('w')),
1843 'height': int(mobj
.group('h')),
1846 mobj
= re
.search(r
'(?<![a-zA-Z0-9])(\d+)[pPiI](?![a-zA-Z0-9])', s
)
1848 return {'height': int(mobj.group(1))}
1850 mobj
= re
.search(r
'\b([48])[kK]\b', s
)
1852 return {'height': int(mobj.group(1)) * 540}
1857 def parse_bitrate(s
):
1858 if not isinstance(s
, str):
1860 mobj
= re
.search(r
'\b(\d+)\s*kbps', s
)
1862 return int(mobj
.group(1))
1865 def month_by_name(name
, lang
='en'):
1866 """ Return the number of a month by (locale-independently) English name """
1868 month_names
= MONTH_NAMES
.get(lang
, MONTH_NAMES
['en'])
1871 return month_names
.index(name
) + 1
1876 def month_by_abbreviation(abbrev
):
1877 """ Return the number of a month by (locale-independently) English
1881 return [s
[:3] for s
in ENGLISH_MONTH_NAMES
].index(abbrev
) + 1
1886 def fix_xml_ampersands(xml_str
):
1887 """Replace all the '&' by '&' in XML"""
1889 r
'&(?!amp;|lt;|gt;|apos;|quot;|#x[0-9a-fA-F]{,4};|#[0-9]{,4};)',
1894 def setproctitle(title
):
1895 assert isinstance(title
, str)
1897 # Workaround for https://github.com/yt-dlp/yt-dlp/issues/4541
1904 libc
= ctypes
.cdll
.LoadLibrary('libc.so.6')
1908 # LoadLibrary in Windows Python 2.7.13 only expects
1909 # a bytestring, but since unicode_literals turns
1910 # every string into a unicode string, it fails.
1912 title_bytes
= title
.encode()
1913 buf
= ctypes
.create_string_buffer(len(title_bytes
))
1914 buf
.value
= title_bytes
1916 # PR_SET_NAME = 15 Ref: /usr/include/linux/prctl.h
1917 libc
.prctl(15, buf
, 0, 0, 0)
1918 except AttributeError:
1919 return # Strange libc, just skip this
1922 def remove_start(s
, start
):
1923 return s
[len(start
):] if s
is not None and s
.startswith(start
) else s
1926 def remove_end(s
, end
):
1927 return s
[:-len(end
)] if s
is not None and s
.endswith(end
) else s
1930 def remove_quotes(s
):
1931 if s
is None or len(s
) < 2:
1933 for quote
in ('"', "'", ):
1934 if s
[0] == quote
and s
[-1] == quote
:
1939 def get_domain(url
):
1941 This implementation is inconsistent, but is kept for compatibility.
1942 Use this only for "webpage_url_domain"
1944 return remove_start(urllib
.parse
.urlparse(url
).netloc
, 'www.') or None
1947 def url_basename(url
):
1948 path
= urllib
.parse
.urlparse(url
).path
1949 return path
.strip('/').split('/')[-1]
1953 return re
.match(r
'https?://[^?#]+/', url
).group()
1956 def urljoin(base
, path
):
1957 if isinstance(path
, bytes):
1958 path
= path
.decode()
1959 if not isinstance(path
, str) or not path
:
1961 if re
.match(r
'^(?:[a-zA-Z][a-zA-Z0-9+-.]*:)?//', path
):
1963 if isinstance(base
, bytes):
1964 base
= base
.decode()
1965 if not isinstance(base
, str) or not re
.match(
1966 r
'^(?:https?:)?//', base
):
1968 return urllib
.parse
.urljoin(base
, path
)
1971 def int_or_none(v
, scale
=1, default
=None, get_attr
=None, invscale
=1):
1972 if get_attr
and v
is not None:
1973 v
= getattr(v
, get_attr
, None)
1975 return int(v
) * invscale
// scale
1976 except (ValueError, TypeError, OverflowError):
1980 def str_or_none(v
, default
=None):
1981 return default
if v
is None else str(v
)
1984 def str_to_int(int_str
):
1985 """ A more relaxed version of int_or_none """
1986 if isinstance(int_str
, int):
1988 elif isinstance(int_str
, str):
1989 int_str
= re
.sub(r
'[,\.\+]', '', int_str
)
1990 return int_or_none(int_str
)
1993 def float_or_none(v
, scale
=1, invscale
=1, default
=None):
1997 return float(v
) * invscale
/ scale
1998 except (ValueError, TypeError):
2002 def bool_or_none(v
, default
=None):
2003 return v
if isinstance(v
, bool) else default
2006 def strip_or_none(v
, default
=None):
2007 return v
.strip() if isinstance(v
, str) else default
2010 def url_or_none(url
):
2011 if not url
or not isinstance(url
, str):
2014 return url
if re
.match(r
'^(?:(?:https?|rt(?:m(?:pt?[es]?|fp)|sp[su]?)|mms|ftps?):)?//', url
) else None
2017 def strftime_or_none(timestamp
, date_format
='%Y%m%d', default
=None):
2018 datetime_object
= None
2020 if isinstance(timestamp
, (int, float)): # unix timestamp
2021 # Using naive datetime here can break timestamp() in Windows
2022 # Ref: https://github.com/yt-dlp/yt-dlp/issues/5185, https://github.com/python/cpython/issues/94414
2023 # Also, dt.datetime.fromtimestamp breaks for negative timestamps
2024 # Ref: https://github.com/yt-dlp/yt-dlp/issues/6706#issuecomment-1496842642
2025 datetime_object
= (dt
.datetime
.fromtimestamp(0, dt
.timezone
.utc
)
2026 + dt
.timedelta(seconds
=timestamp
))
2027 elif isinstance(timestamp
, str): # assume YYYYMMDD
2028 datetime_object
= dt
.datetime
.strptime(timestamp
, '%Y%m%d')
2029 date_format
= re
.sub( # Support %s on windows
2030 r
'(?<!%)(%%)*%s', rf
'\g<1>{int(datetime_object.timestamp())}', date_format
)
2031 return datetime_object
.strftime(date_format
)
2032 except (ValueError, TypeError, AttributeError):
2036 def parse_duration(s
):
2037 if not isinstance(s
, str):
2043 days
, hours
, mins
, secs
, ms
= [None] * 5
2044 m
= re
.match(r
'''(?x)
2046 (?:(?:(?P<days>[0-9]+):)?(?P<hours>[0-9]+):)?(?P<mins>[0-9]+):)?
2047 (?P<secs>(?(before_secs)[0-9]{1,2}|[0-9]+))
2048 (?P<ms>[.:][0-9]+)?Z?$
2051 days
, hours
, mins
, secs
, ms
= m
.group('days', 'hours', 'mins', 'secs', 'ms')
2056 [0-9]+\s*y(?:ears?)?,?\s*
2059 [0-9]+\s*m(?:onths?)?,?\s*
2062 [0-9]+\s*w(?:eeks?)?,?\s*
2065 (?P<days>[0-9]+)\s*d(?:ays?)?,?\s*
2069 (?P<hours>[0-9]+)\s*h(?:(?:ou)?rs?)?,?\s*
2072 (?P<mins>[0-9]+)\s*m(?:in(?:ute)?s?)?,?\s*
2075 (?P<secs>[0-9]+)(?P<ms>\.[0-9]+)?\s*s(?:ec(?:ond)?s?)?\s*
2078 days
, hours
, mins
, secs
, ms
= m
.groups()
2080 m
= re
.match(r
'(?i)(?:(?P<hours>[0-9.]+)\s*(?:hours?)|(?P<mins>[0-9.]+)\s*(?:mins?\.?|minutes?)\s*)Z?$', s
)
2082 hours
, mins
= m
.groups()
2087 ms
= ms
.replace(':', '.')
2088 return sum(float(part
or 0) * mult
for part
, mult
in (
2089 (days
, 86400), (hours
, 3600), (mins
, 60), (secs
, 1), (ms
, 1)))
2092 def prepend_extension(filename
, ext
, expected_real_ext
=None):
2093 name
, real_ext
= os
.path
.splitext(filename
)
2095 f
'{name}.{ext}{real_ext}'
2096 if not expected_real_ext
or real_ext
[1:] == expected_real_ext
2097 else f
'{filename}.{ext}')
2100 def replace_extension(filename
, ext
, expected_real_ext
=None):
2101 name
, real_ext
= os
.path
.splitext(filename
)
2102 return '{}.{}'.format(
2103 name
if not expected_real_ext
or real_ext
[1:] == expected_real_ext
else filename
,
2107 def check_executable(exe
, args
=[]):
2108 """ Checks if the given binary is installed somewhere in PATH, and returns its name.
2109 args can be a list of arguments for a short output (like -version) """
2111 Popen
.run([exe
] + args
, stdout
=subprocess
.PIPE
, stderr
=subprocess
.PIPE
)
2117 def _get_exe_version_output(exe
, args
):
2119 # STDIN should be redirected too. On UNIX-like systems, ffmpeg triggers
2120 # SIGTTOU if yt-dlp is run in the background.
2121 # See https://github.com/ytdl-org/youtube-dl/issues/955#issuecomment-209789656
2122 stdout
, _
, ret
= Popen
.run([encodeArgument(exe
)] + args
, text
=True,
2123 stdin
=subprocess
.PIPE
, stdout
=subprocess
.PIPE
, stderr
=subprocess
.STDOUT
)
2131 def detect_exe_version(output
, version_re
=None, unrecognized
='present'):
2132 assert isinstance(output
, str)
2133 if version_re
is None:
2134 version_re
= r
'version\s+([-0-9._a-zA-Z]+)'
2135 m
= re
.search(version_re
, output
)
2142 def get_exe_version(exe
, args
=['--version'],
2143 version_re
=None, unrecognized
=('present', 'broken')):
2144 """ Returns the version of the specified executable,
2145 or False if the executable is not present """
2146 unrecognized
= variadic(unrecognized
)
2147 assert len(unrecognized
) in (1, 2)
2148 out
= _get_exe_version_output(exe
, args
)
2150 return unrecognized
[-1]
2151 return out
and detect_exe_version(out
, version_re
, unrecognized
[0])
2154 def frange(start
=0, stop
=None, step
=1):
2157 start
, stop
= 0, start
2158 sign
= [-1, 1][step
> 0] if step
else 0
2159 while sign
* start
< sign
* stop
:
2164 class LazyList(collections
.abc
.Sequence
):
2165 """Lazy immutable list from an iterable
2166 Note that slices of a LazyList are lists and not LazyList"""
2168 class IndexError(IndexError):
2171 def __init__(self
, iterable
, *, reverse
=False, _cache
=None):
2172 self
._iterable
= iter(iterable
)
2173 self
._cache
= [] if _cache
is None else _cache
2174 self
._reversed
= reverse
2178 # We need to consume the entire iterable to iterate in reverse
2179 yield from self
.exhaust()
2181 yield from self
._cache
2182 for item
in self
._iterable
:
2183 self
._cache
.append(item
)
2187 self
._cache
.extend(self
._iterable
)
2188 self
._iterable
= [] # Discard the emptied iterable to make it pickle-able
2192 """Evaluate the entire iterable"""
2193 return self
._exhaust
()[::-1 if self
._reversed
else 1]
2196 def _reverse_index(x
):
2197 return None if x
is None else ~x
2199 def __getitem__(self
, idx
):
2200 if isinstance(idx
, slice):
2202 idx
= slice(self
._reverse
_index
(idx
.start
), self
._reverse
_index
(idx
.stop
), -(idx
.step
or 1))
2203 start
, stop
, step
= idx
.start
, idx
.stop
, idx
.step
or 1
2204 elif isinstance(idx
, int):
2206 idx
= self
._reverse
_index
(idx
)
2207 start
, stop
, step
= idx
, idx
, 0
2209 raise TypeError('indices must be integers or slices')
2210 if ((start
or 0) < 0 or (stop
or 0) < 0
2211 or (start
is None and step
< 0)
2212 or (stop
is None and step
> 0)):
2213 # We need to consume the entire iterable to be able to slice from the end
2214 # Obviously, never use this with infinite iterables
2217 return self
._cache
[idx
]
2218 except IndexError as e
:
2219 raise self
.IndexError(e
) from e
2220 n
= max(start
or 0, stop
or 0) - len(self
._cache
) + 1
2222 self
._cache
.extend(itertools
.islice(self
._iterable
, n
))
2224 return self
._cache
[idx
]
2225 except IndexError as e
:
2226 raise self
.IndexError(e
) from e
2230 self
[-1] if self
._reversed
else self
[0]
2231 except self
.IndexError:
2237 return len(self
._cache
)
2239 def __reversed__(self
):
2240 return type(self
)(self
._iterable
, reverse
=not self
._reversed
, _cache
=self
._cache
)
2243 return type(self
)(self
._iterable
, reverse
=self
._reversed
, _cache
=self
._cache
)
2246 # repr and str should mimic a list. So we exhaust the iterable
2247 return repr(self
.exhaust())
2250 return repr(self
.exhaust())
2255 class IndexError(IndexError):
2259 # This is only useful for tests
2260 return len(self
.getslice())
2262 def __init__(self
, pagefunc
, pagesize
, use_cache
=True):
2263 self
._pagefunc
= pagefunc
2264 self
._pagesize
= pagesize
2265 self
._pagecount
= float('inf')
2266 self
._use
_cache
= use_cache
2269 def getpage(self
, pagenum
):
2270 page_results
= self
._cache
.get(pagenum
)
2271 if page_results
is None:
2272 page_results
= [] if pagenum
> self
._pagecount
else list(self
._pagefunc
(pagenum
))
2274 self
._cache
[pagenum
] = page_results
2277 def getslice(self
, start
=0, end
=None):
2278 return list(self
._getslice
(start
, end
))
2280 def _getslice(self
, start
, end
):
2281 raise NotImplementedError('This method must be implemented by subclasses')
2283 def __getitem__(self
, idx
):
2284 assert self
._use
_cache
, 'Indexing PagedList requires cache'
2285 if not isinstance(idx
, int) or idx
< 0:
2286 raise TypeError('indices must be non-negative integers')
2287 entries
= self
.getslice(idx
, idx
+ 1)
2289 raise self
.IndexError()
2293 return bool(self
.getslice(0, 1))
2296 class OnDemandPagedList(PagedList
):
2297 """Download pages until a page with less than maximum results"""
2299 def _getslice(self
, start
, end
):
2300 for pagenum
in itertools
.count(start
// self
._pagesize
):
2301 firstid
= pagenum
* self
._pagesize
2302 nextfirstid
= pagenum
* self
._pagesize
+ self
._pagesize
2303 if start
>= nextfirstid
:
2307 start
% self
._pagesize
2308 if firstid
<= start
< nextfirstid
2311 ((end
- 1) % self
._pagesize
) + 1
2312 if (end
is not None and firstid
<= end
<= nextfirstid
)
2316 page_results
= self
.getpage(pagenum
)
2318 self
._pagecount
= pagenum
- 1
2320 if startv
!= 0 or endv
is not None:
2321 page_results
= page_results
[startv
:endv
]
2322 yield from page_results
2324 # A little optimization - if current page is not "full", ie. does
2325 # not contain page_size videos then we can assume that this page
2326 # is the last one - there are no more ids on further pages -
2327 # i.e. no need to query again.
2328 if len(page_results
) + startv
< self
._pagesize
:
2331 # If we got the whole page, but the next page is not interesting,
2332 # break out early as well
2333 if end
== nextfirstid
:
2337 class InAdvancePagedList(PagedList
):
2338 """PagedList with total number of pages known in advance"""
2340 def __init__(self
, pagefunc
, pagecount
, pagesize
):
2341 PagedList
.__init
__(self
, pagefunc
, pagesize
, True)
2342 self
._pagecount
= pagecount
2344 def _getslice(self
, start
, end
):
2345 start_page
= start
// self
._pagesize
2346 end_page
= self
._pagecount
if end
is None else min(self
._pagecount
, end
// self
._pagesize
+ 1)
2347 skip_elems
= start
- start_page
* self
._pagesize
2348 only_more
= None if end
is None else end
- start
2349 for pagenum
in range(start_page
, end_page
):
2350 page_results
= self
.getpage(pagenum
)
2352 page_results
= page_results
[skip_elems
:]
2354 if only_more
is not None:
2355 if len(page_results
) < only_more
:
2356 only_more
-= len(page_results
)
2358 yield from page_results
[:only_more
]
2360 yield from page_results
2363 class PlaylistEntries
:
2364 MissingEntry
= object()
2365 is_exhausted
= False
2367 def __init__(self
, ydl
, info_dict
):
2370 # _entries must be assigned now since infodict can change during iteration
2371 entries
= info_dict
.get('entries')
2373 raise EntryNotInPlaylist('There are no entries')
2374 elif isinstance(entries
, list):
2375 self
.is_exhausted
= True
2377 requested_entries
= info_dict
.get('requested_entries')
2378 self
.is_incomplete
= requested_entries
is not None
2379 if self
.is_incomplete
:
2380 assert self
.is_exhausted
2381 self
._entries
= [self
.MissingEntry
] * max(requested_entries
or [0])
2382 for i
, entry
in zip(requested_entries
, entries
):
2383 self
._entries
[i
- 1] = entry
2384 elif isinstance(entries
, (list, PagedList
, LazyList
)):
2385 self
._entries
= entries
2387 self
._entries
= LazyList(entries
)
2389 PLAYLIST_ITEMS_RE
= re
.compile(r
'''(?x)
2390 (?P<start>[+-]?\d+)?
2392 (?P<end>[+-]?\d+|inf(?:inite)?)?
2393 (?::(?P<step>[+-]?\d+))?
2397 def parse_playlist_items(cls
, string
):
2398 for segment
in string
.split(','):
2400 raise ValueError('There is two or more consecutive commas')
2401 mobj
= cls
.PLAYLIST_ITEMS_RE
.fullmatch(segment
)
2403 raise ValueError(f
'{segment!r} is not a valid specification')
2404 start
, end
, step
, has_range
= mobj
.group('start', 'end', 'step', 'range')
2405 if int_or_none(step
) == 0:
2406 raise ValueError(f
'Step in {segment!r} cannot be zero')
2407 yield slice(int_or_none(start
), float_or_none(end
), int_or_none(step
)) if has_range
else int(start
)
2409 def get_requested_items(self
):
2410 playlist_items
= self
.ydl
.params
.get('playlist_items')
2411 playlist_start
= self
.ydl
.params
.get('playliststart', 1)
2412 playlist_end
= self
.ydl
.params
.get('playlistend')
2413 # For backwards compatibility, interpret -1 as whole list
2414 if playlist_end
in (-1, None):
2416 if not playlist_items
:
2417 playlist_items
= f
'{playlist_start}:{playlist_end}'
2418 elif playlist_start
!= 1 or playlist_end
:
2419 self
.ydl
.report_warning('Ignoring playliststart and playlistend because playlistitems was given', only_once
=True)
2421 for index
in self
.parse_playlist_items(playlist_items
):
2422 for i
, entry
in self
[index
]:
2427 # The item may have just been added to archive. Don't break due to it
2428 if not self
.ydl
.params
.get('lazy_playlist'):
2429 # TODO: Add auto-generated fields
2430 self
.ydl
._match
_entry
(entry
, incomplete
=True, silent
=True)
2431 except (ExistingVideoReached
, RejectedVideoReached
):
2434 def get_full_count(self
):
2435 if self
.is_exhausted
and not self
.is_incomplete
:
2437 elif isinstance(self
._entries
, InAdvancePagedList
):
2438 if self
._entries
._pagesize
== 1:
2439 return self
._entries
._pagecount
2441 @functools.cached_property
2443 if isinstance(self
._entries
, list):
2446 entry
= self
._entries
[i
]
2448 entry
= self
.MissingEntry
2449 if not self
.is_incomplete
:
2450 raise self
.IndexError()
2451 if entry
is self
.MissingEntry
:
2452 raise EntryNotInPlaylist(f
'Entry {i + 1} cannot be found')
2457 return type(self
.ydl
)._handle
_extraction
_exceptions
(lambda _
, i
: self
._entries
[i
])(self
.ydl
, i
)
2458 except (LazyList
.IndexError, PagedList
.IndexError):
2459 raise self
.IndexError()
2462 def __getitem__(self
, idx
):
2463 if isinstance(idx
, int):
2464 idx
= slice(idx
, idx
)
2466 # NB: PlaylistEntries[1:10] => (0, 1, ... 9)
2467 step
= 1 if idx
.step
is None else idx
.step
2468 if idx
.start
is None:
2469 start
= 0 if step
> 0 else len(self
) - 1
2471 start
= idx
.start
- 1 if idx
.start
>= 0 else len(self
) + idx
.start
2473 # NB: Do not call len(self) when idx == [:]
2474 if idx
.stop
is None:
2475 stop
= 0 if step
< 0 else float('inf')
2477 stop
= idx
.stop
- 1 if idx
.stop
>= 0 else len(self
) + idx
.stop
2478 stop
+= [-1, 1][step
> 0]
2480 for i
in frange(start
, stop
, step
):
2484 entry
= self
._getter
(i
)
2485 except self
.IndexError:
2486 self
.is_exhausted
= True
2493 return len(tuple(self
[:]))
2495 class IndexError(IndexError):
2499 def uppercase_escape(s
):
2500 unicode_escape
= codecs
.getdecoder('unicode_escape')
2502 r
'\\U[0-9a-fA-F]{8}',
2503 lambda m
: unicode_escape(m
.group(0))[0],
2507 def lowercase_escape(s
):
2508 unicode_escape
= codecs
.getdecoder('unicode_escape')
2510 r
'\\u[0-9a-fA-F]{4}',
2511 lambda m
: unicode_escape(m
.group(0))[0],
2515 def parse_qs(url
, **kwargs
):
2516 return urllib
.parse
.parse_qs(urllib
.parse
.urlparse(url
).query
, **kwargs
)
2519 def read_batch_urls(batch_fd
):
2521 if not isinstance(url
, str):
2522 url
= url
.decode('utf-8', 'replace')
2523 BOM_UTF8
= ('\xef\xbb\xbf', '\ufeff')
2524 for bom
in BOM_UTF8
:
2525 if url
.startswith(bom
):
2526 url
= url
[len(bom
):]
2528 if not url
or url
.startswith(('#', ';', ']')):
2530 # "#" cannot be stripped out since it is part of the URI
2531 # However, it can be safely stripped out if following a whitespace
2532 return re
.split(r
'\s#', url
, 1)[0].rstrip()
2534 with contextlib
.closing(batch_fd
) as fd
:
2535 return [url
for url
in map(fixup
, fd
) if url
]
2538 def urlencode_postdata(*args
, **kargs
):
2539 return urllib
.parse
.urlencode(*args
, **kargs
).encode('ascii')
2542 def update_url(url
, *, query_update
=None, **kwargs
):
2543 """Replace URL components specified by kwargs
2544 @param url str or parse url tuple
2545 @param query_update update query
2548 if isinstance(url
, str):
2549 if not kwargs
and not query_update
:
2552 url
= urllib
.parse
.urlparse(url
)
2554 assert 'query' not in kwargs
, 'query_update and query cannot be specified at the same time'
2555 kwargs
['query'] = urllib
.parse
.urlencode({
2556 **urllib
.parse
.parse_qs(url
.query
),
2559 return urllib
.parse
.urlunparse(url
._replace
(**kwargs
))
2562 def update_url_query(url
, query
):
2563 return update_url(url
, query_update
=query
)
2566 def _multipart_encode_impl(data
, boundary
):
2567 content_type
= 'multipart/form-data; boundary=%s' % boundary
2570 for k
, v
in data
.items():
2571 out
+= b
'--' + boundary
.encode('ascii') + b
'\r\n'
2572 if isinstance(k
, str):
2574 if isinstance(v
, str):
2576 # RFC 2047 requires non-ASCII field names to be encoded, while RFC 7578
2577 # suggests sending UTF-8 directly. Firefox sends UTF-8, too
2578 content
= b
'Content-Disposition: form-data; name="' + k
+ b
'"\r\n\r\n' + v
+ b
'\r\n'
2579 if boundary
.encode('ascii') in content
:
2580 raise ValueError('Boundary overlaps with data')
2583 out
+= b
'--' + boundary
.encode('ascii') + b
'--\r\n'
2585 return out
, content_type
2588 def multipart_encode(data
, boundary
=None):
2590 Encode a dict to RFC 7578-compliant form-data
2593 A dict where keys and values can be either Unicode or bytes-like
2596 If specified a Unicode object, it's used as the boundary. Otherwise
2597 a random boundary is generated.
2599 Reference: https://tools.ietf.org/html/rfc7578
2601 has_specified_boundary
= boundary
is not None
2604 if boundary
is None:
2605 boundary
= '---------------' + str(random
.randrange(0x0fffffff, 0xffffffff))
2608 out
, content_type
= _multipart_encode_impl(data
, boundary
)
2611 if has_specified_boundary
:
2615 return out
, content_type
2618 def is_iterable_like(x
, allowed_types
=collections
.abc
.Iterable
, blocked_types
=NO_DEFAULT
):
2619 if blocked_types
is NO_DEFAULT
:
2620 blocked_types
= (str, bytes, collections
.abc
.Mapping
)
2621 return isinstance(x
, allowed_types
) and not isinstance(x
, blocked_types
)
2624 def variadic(x
, allowed_types
=NO_DEFAULT
):
2625 if not isinstance(allowed_types
, (tuple, type)):
2626 deprecation_warning('allowed_types should be a tuple or a type')
2627 allowed_types
= tuple(allowed_types
)
2628 return x
if is_iterable_like(x
, blocked_types
=allowed_types
) else (x
, )
2631 def try_call(*funcs
, expected_type
=None, args
=[], kwargs
={}):
2634 val
= f(*args
, **kwargs
)
2635 except (AttributeError, KeyError, TypeError, IndexError, ValueError, ZeroDivisionError):
2638 if expected_type
is None or isinstance(val
, expected_type
):
2642 def try_get(src
, getter
, expected_type
=None):
2643 return try_call(*variadic(getter
), args
=(src
,), expected_type
=expected_type
)
2646 def filter_dict(dct
, cndn
=lambda _
, v
: v
is not None):
2647 return {k: v for k, v in dct.items() if cndn(k, v)}
2650 def merge_dicts(*dicts
):
2652 for a_dict
in dicts
:
2653 for k
, v
in a_dict
.items():
2654 if (v
is not None and k
not in merged
2655 or isinstance(v
, str) and merged
[k
] == ''):
2660 def encode_compat_str(string
, encoding
=preferredencoding(), errors
='strict'):
2661 return string
if isinstance(string
, str) else str(string
, encoding
, errors
)
2673 TV_PARENTAL_GUIDELINES
= {
2683 def parse_age_limit(s
):
2684 # isinstance(False, int) is True. So type() must be used instead
2685 if type(s
) is int: # noqa: E721
2686 return s
if 0 <= s
<= 21 else None
2687 elif not isinstance(s
, str):
2689 m
= re
.match(r
'^(?P<age>\d{1,2})\+?$', s
)
2691 return int(m
.group('age'))
2694 return US_RATINGS
[s
]
2695 m
= re
.match(r
'^TV[_-]?(%s)$' % '|'.join(k
[3:] for k
in TV_PARENTAL_GUIDELINES
), s
)
2697 return TV_PARENTAL_GUIDELINES
['TV-' + m
.group(1)]
2701 def strip_jsonp(code
):
2704 (?:window\.)?(?P<func_name>[a-zA-Z0-9_.$]*)
2705 (?:\s*&&\s*(?P=func_name))?
2706 \s*\(\s*(?P<callback_data>.*)\);?
2707 \s*?(?://[^\n]*)*$''',
2708 r
'\g<callback_data>', code
)
2711 def js_to_json(code
, vars={}, *, strict
=False):
2712 # vars is a dict of var, val pairs to substitute
2713 STRING_QUOTES
= '\'"`'
2714 STRING_RE
= '|'.join(rf
'{q}(?:\\.|[^\\{q}])*{q}' for q
in STRING_QUOTES
)
2715 COMMENT_RE
= r
'/\*(?:(?!\*/).)*?\*/|//[^\n]*\n'
2716 SKIP_RE
= fr
'\s*(?:{COMMENT_RE})?\s*'
2718 (fr
'(?s)^(0[xX][0-9a-fA-F]+){SKIP_RE}:?$', 16),
2719 (fr
'(?s)^(0+[0-7]+){SKIP_RE}:?$', 8),
2722 def process_escape(match
):
2723 JSON_PASSTHROUGH_ESCAPES
= R
'"\bfnrtu'
2724 escape
= match
.group(1) or match
.group(2)
2726 return (Rf
'\{escape}' if escape
in JSON_PASSTHROUGH_ESCAPES
2727 else R
'\u00' if escape
== 'x'
2728 else '' if escape
== '\n'
2731 def template_substitute(match
):
2732 evaluated
= js_to_json(match
.group(1), vars, strict
=strict
)
2733 if evaluated
[0] == '"':
2734 return json
.loads(evaluated
)
2739 if v
in ('true', 'false', 'null'):
2741 elif v
in ('undefined', 'void 0'):
2743 elif v
.startswith('/*') or v
.startswith('//') or v
.startswith('!') or v
== ',':
2746 if v
[0] in STRING_QUOTES
:
2747 v
= re
.sub(r
'(?s)\${([^}]+)}', template_substitute
, v
[1:-1]) if v
[0] == '`' else v
[1:-1]
2748 escaped
= re
.sub(r
'(?s)(")|\\(.)', process_escape
, v
)
2749 return f
'"{escaped}"'
2751 for regex
, base
in INTEGER_TABLE
:
2752 im
= re
.match(regex
, v
)
2754 i
= int(im
.group(1), base
)
2755 return f
'"{i}":' if v
.endswith(':') else str(i
)
2761 except json
.JSONDecodeError
:
2762 return json
.dumps(vars[v
])
2769 raise ValueError(f
'Unknown value: {v}')
2771 def create_map(mobj
):
2772 return json
.dumps(dict(json
.loads(js_to_json(mobj
.group(1) or '[]', vars=vars))))
2774 code
= re
.sub(r
'(?:new\s+)?Array\((.*?)\)', r
'[\g<1>]', code
)
2775 code
= re
.sub(r
'new Map\((\[.*?\])?\)', create_map
, code
)
2777 code
= re
.sub(rf
'new Date\(({STRING_RE})\)', r
'\g<1>', code
)
2778 code
= re
.sub(r
'new \w+\((.*?)\)', lambda m
: json
.dumps(m
.group(0)), code
)
2779 code
= re
.sub(r
'parseInt\([^\d]+(\d+)[^\d]+\)', r
'\1', code
)
2780 code
= re
.sub(r
'\(function\([^)]*\)\s*\{[^}]*\}\s*\)\s*\(\s*(["\'][^
)]*["\'])\s*\)', r'\1', code)
2782 return re.sub(rf'''(?sx)
2784 {COMMENT_RE}|,(?={SKIP_RE}[\]}}])|
2785 void\s0|(?:(?<![0-9])[eE]|[a-df-zA-DF-Z_$])[.a-zA-Z_$0-9]*|
2786 \b(?:0[xX][0-9a-fA-F]+|0+[0-7]+)(?:{SKIP_RE}:)?|
2787 [0-9]+(?={SKIP_RE}:)|
2792 def qualities(quality_ids):
2793 """ Get a numeric quality value out of a list of possible values """
2796 return quality_ids.index(qid)
2802 POSTPROCESS_WHEN = ('pre_process', 'after_filter', 'video', 'before_dl', 'post_process', 'after_move', 'after_video', 'playlist')
2806 'default': '%(title)s [%(id)s].%(ext)s',
2807 'chapter': '%(title)s - %(section_number)03d %(section_title)s [%(id)s].%(ext)s',
2813 'description': 'description',
2814 'annotation': 'annotations.xml',
2815 'infojson': 'info.json',
2818 'pl_thumbnail': None,
2819 'pl_description': 'description',
2820 'pl_infojson': 'info.json',
2823 # As of [1] format syntax is:
2824 # %[mapping_key][conversion_flags][minimum_width][.precision][length_modifier]type
2825 # 1. https://docs.python.org/2/library/stdtypes.html#string-formatting
2826 STR_FORMAT_RE_TMPL = r'''(?x)
2827 (?<!%)(?P<prefix>(?:%%)*)
2829 (?P<has_key>\((?P<key>{0})\))?
2831 (?P<conversion>[#0\-+ ]+)?
2833 (?P<precision>\.\d+)?
2834 (?P<len_mod>[hlL])? # unused in python
2835 {1} # conversion type
2840 STR_FORMAT_TYPES = 'diouxXeEfFgGcrsa'
2843 def limit_length(s, length):
2844 """ Add ellipses to overly long strings """
2849 return s[:length - len(ELLIPSES)] + ELLIPSES
2853 def version_tuple(v):
2854 return tuple(int(e) for e in re.split(r'[-.]', v))
2857 def is_outdated_version(version, limit, assume_new=True):
2859 return not assume_new
2861 return version_tuple(version) < version_tuple(limit)
2863 return not assume_new
2866 def ytdl_is_updateable():
2867 """ Returns if yt-dlp can be updated with -U """
2869 from ..update import is_non_updateable
2871 return not is_non_updateable()
2874 def args_to_str(args):
2875 # Get a short string representation for a subprocess command
2876 return shell_quote(args)
2879 def error_to_str(err):
2880 return f'{type(err).__name__}: {err}'
2883 def mimetype2ext(mt, default=NO_DEFAULT):
2884 if not isinstance(mt, str):
2885 if default is not NO_DEFAULT:
2902 'x-matroska': 'mkv',
2904 'x-mp4-fragmented': 'mp4',
2909 # application (streaming playlists)
2913 'vnd.apple.mpegurl': 'm3u8',
2914 'vnd.ms-sstr+xml': 'ism',
2915 'x-mpegurl': 'm3u8',
2919 # Per RFC 3003, audio/mpeg can be .mp1, .mp2 or .mp3.
2920 # Using .mp3 as it's the most popular one
2921 'audio/mpeg': 'mp3',
2922 'audio/webm': 'webm',
2923 'audio/x-matroska': 'mka',
2924 'audio/x-mpegurl': 'm3u',
2932 'x-realaudio': 'ra',
2943 'vnd.wap.wbmp': 'wbmp',
2950 'filmstrip+json': 'fs',
2951 'smptett+xml': 'tt',
2954 'x-ms-sami': 'sami',
2963 mimetype = mt.partition(';')[0].strip().lower()
2964 _, _, subtype = mimetype.rpartition('/')
2966 ext = traversal.traverse_obj(MAP, mimetype, subtype, subtype.rsplit('+')[-1])
2969 elif default is not NO_DEFAULT:
2971 return subtype.replace('+', '.')
2974 def ext2mimetype(ext_or_url):
2977 if '.' not in ext_or_url:
2978 ext_or_url = f'file.{ext_or_url}'
2979 return mimetypes.guess_type(ext_or_url)[0]
2982 def parse_codecs(codecs_str):
2983 # http://tools.ietf.org/html/rfc6381
2986 split_codecs = list(filter(None, map(
2987 str.strip, codecs_str.strip().strip(',').split(','))))
2988 vcodec, acodec, scodec, hdr = None, None, None, None
2989 for full_codec in split_codecs:
2990 parts = re.sub(r'0+(?=\d)', '', full_codec).split('.')
2991 if parts[0] in ('avc1', 'avc2', 'avc3', 'avc4', 'vp9', 'vp8', 'hev1', 'hev2',
2992 'h263', 'h264', 'mp4v', 'hvc1', 'av1', 'theora', 'dvh1', 'dvhe'):
2996 if parts[0] in ('dvh1', 'dvhe'):
2998 elif parts[0] == 'av1' and traversal.traverse_obj(parts, 3) == '10':
3000 elif parts[:2] == ['vp9', '2']:
3002 elif parts[0] in ('flac', 'mp4a', 'opus', 'vorbis', 'mp3', 'aac', 'ac-4',
3003 'ac-3', 'ec-3', 'eac3', 'dtsc', 'dtse', 'dtsh', 'dtsl'):
3004 acodec = acodec or full_codec
3005 elif parts[0] in ('stpp', 'wvtt'):
3006 scodec = scodec or full_codec
3008 write_string(f'WARNING: Unknown codec {full_codec}\n')
3009 if vcodec or acodec or scodec:
3011 'vcodec': vcodec or 'none',
3012 'acodec': acodec or 'none',
3013 'dynamic_range': hdr,
3014 **({'scodec': scodec} if scodec is not None else {}),
3016 elif len(split_codecs) == 2:
3018 'vcodec': split_codecs[0],
3019 'acodec': split_codecs[1],
3024 def get_compatible_ext(*, vcodecs, acodecs, vexts, aexts, preferences=None):
3025 assert len(vcodecs) == len(vexts) and len(acodecs) == len(aexts)
3027 allow_mkv = not preferences or 'mkv' in preferences
3029 if allow_mkv and max(len(acodecs), len(vcodecs)) > 1:
3030 return 'mkv' # TODO: any other format allows this?
3032 # TODO: All codecs supported by parse_codecs isn't handled here
3033 COMPATIBLE_CODECS = {
3035 'av1', 'hevc', 'avc1', 'mp4a', 'ac-4', # fourcc (m3u8, mpd)
3036 'h264', 'aacl', 'ec-3', # Set in ISM
3039 'av1', 'vp9', 'vp8', 'opus', 'vrbs',
3040 'vp9x', 'vp8x', # in the webm spec
3044 sanitize_codec = functools.partial(
3045 try_get, getter=lambda x: x[0].split('.')[0].replace('0', '').lower())
3046 vcodec, acodec = sanitize_codec(vcodecs), sanitize_codec(acodecs)
3048 for ext in preferences or COMPATIBLE_CODECS.keys():
3049 codec_set = COMPATIBLE_CODECS.get(ext, set())
3050 if ext == 'mkv' or codec_set.issuperset((vcodec, acodec)):
3054 {'mp3', 'mp4', 'm4a', 'm4p', 'm4b', 'm4r', 'm4v', 'ismv', 'isma', 'mov'},
3057 for ext in preferences or vexts:
3058 current_exts = {ext, *vexts, *aexts}
3059 if ext == 'mkv' or current_exts == {ext} or any(
3060 ext_sets.issuperset(current_exts) for ext_sets in COMPATIBLE_EXTS):
3062 return 'mkv' if allow_mkv else preferences[-1]
3065 def urlhandle_detect_ext(url_handle, default=NO_DEFAULT):
3066 getheader = url_handle.headers.get
3068 cd = getheader('Content-Disposition')
3070 m = re.match(r'attachment;\s*filename="(?P
<filename
>[^
"]+)"', cd)
3072 e = determine_ext(m.group('filename
'), default_ext=None)
3076 meta_ext = getheader('x
-amz
-meta
-name
')
3078 e = meta_ext.rpartition('.')[2]
3082 return mimetype2ext(getheader('Content
-Type
'), default=default)
3085 def encode_data_uri(data, mime_type):
3086 return 'data
:%s;base64
,%s' % (mime_type, base64.b64encode(data).decode('ascii
'))
3089 def age_restricted(content_limit, age_limit):
3090 """ Returns True iff the content should be blocked """
3092 if age_limit is None: # No limit set
3094 if content_limit is None:
3095 return False # Content available for everyone
3096 return age_limit < content_limit
3099 # List of known byte-order-marks (BOM)
3101 (b'\xef\xbb\xbf', 'utf
-8'),
3102 (b'\x00\x00\xfe\xff', 'utf
-32-be
'),
3103 (b'\xff\xfe\x00\x00', 'utf
-32-le
'),
3104 (b'\xff\xfe', 'utf
-16-le
'),
3105 (b'\xfe\xff', 'utf
-16-be
'),
3109 def is_html(first_bytes):
3110 """ Detect whether a file contains HTML by examining its first bytes. """
3113 for bom, enc in BOMS:
3114 while first_bytes.startswith(bom):
3115 encoding, first_bytes = enc, first_bytes[len(bom):]
3117 return re.match(r'^\s
*<', first_bytes.decode(encoding, 'replace
'))
3120 def determine_protocol(info_dict):
3121 protocol = info_dict.get('protocol
')
3122 if protocol is not None:
3125 url = sanitize_url(info_dict['url
'])
3126 if url.startswith('rtmp
'):
3128 elif url.startswith('mms
'):
3130 elif url.startswith('rtsp
'):
3133 ext = determine_ext(url)
3135 return 'm3u8
' if info_dict.get('is_live
') else 'm3u8_native
'
3139 return urllib.parse.urlparse(url).scheme
3142 def render_table(header_row, data, delim=False, extra_gap=0, hide_empty=False):
3143 """ Render a list of rows, each as a list of values.
3144 Text after a \t will be right aligned """
3146 return len(remove_terminal_sequences(string).replace('\t', ''))
3148 def get_max_lens(table):
3149 return [max(width(str(v)) for v in col) for col in zip(*table)]
3151 def filter_using_list(row, filterArray):
3152 return [col for take, col in itertools.zip_longest(filterArray, row, fillvalue=True) if take]
3154 max_lens = get_max_lens(data) if hide_empty else []
3155 header_row = filter_using_list(header_row, max_lens)
3156 data = [filter_using_list(row, max_lens) for row in data]
3158 table = [header_row] + data
3159 max_lens = get_max_lens(table)
3162 table = [header_row, [delim * (ml + extra_gap) for ml in max_lens]] + data
3163 table[1][-1] = table[1][-1][:-extra_gap * len(delim)] # Remove extra_gap from end of delimiter
3165 for pos, text in enumerate(map(str, row)):
3167 row[pos] = text.replace('\t', ' ' * (max_lens[pos] - width(text))) + ' ' * extra_gap
3169 row[pos] = text + ' ' * (max_lens[pos] - width(text) + extra_gap)
3170 ret = '\n'.join(''.join(row).rstrip() for row in table)
3174 def _match_one(filter_part, dct, incomplete):
3175 # TODO: Generalize code with YoutubeDL._build_format_filter
3176 STRING_OPERATORS = {
3177 '*=': operator.contains,
3178 '^
=': lambda attr, value: attr.startswith(value),
3179 '$
=': lambda attr, value: attr.endswith(value),
3180 '~
=': lambda attr, value: re.search(value, attr),
3182 COMPARISON_OPERATORS = {
3184 '<=': operator.le, # "<=" must be defined above "<"
3191 if isinstance(incomplete, bool):
3192 is_incomplete = lambda _: incomplete
3194 is_incomplete = lambda k: k in incomplete
3196 operator_rex = re.compile(r'''(?x)
3198 \s*(?P<negation>!\s*)?(?P<op>%s)(?P<none_inclusive>\s*\?)?\s*
3200 (?P<quote>["\'])(?P<quotedstrval>.+?)(?P=quote)|
3203 ''' % '|
'.join(map(re.escape, COMPARISON_OPERATORS.keys())))
3204 m = operator_rex.fullmatch(filter_part.strip())
3207 unnegated_op = COMPARISON_OPERATORS[m['op
']]
3209 op = lambda attr, value: not unnegated_op(attr, value)
3212 comparison_value = m['quotedstrval
'] or m['strval
'] or m['intval
']
3214 comparison_value = comparison_value.replace(r'\
%s' % m['quote
'], m['quote
'])
3215 actual_value = dct.get(m['key
'])
3216 numeric_comparison = None
3217 if isinstance(actual_value, (int, float)):
3218 # If the original field is a string and matching comparisonvalue is
3219 # a number we should respect the origin of the original field
3220 # and process comparison value as a string (see
3221 # https://github.com/ytdl-org/youtube-dl/issues/11082)
3223 numeric_comparison = int(comparison_value)
3225 numeric_comparison = parse_filesize(comparison_value)
3226 if numeric_comparison is None:
3227 numeric_comparison = parse_filesize(f'{comparison_value}B
')
3228 if numeric_comparison is None:
3229 numeric_comparison = parse_duration(comparison_value)
3230 if numeric_comparison is not None and m['op
'] in STRING_OPERATORS:
3231 raise ValueError('Operator
%s only supports string values
!' % m['op
'])
3232 if actual_value is None:
3233 return is_incomplete(m['key
']) or m['none_inclusive
']
3234 return op(actual_value, comparison_value if numeric_comparison is None else numeric_comparison)
3237 '': lambda v: (v is True) if isinstance(v, bool) else (v is not None),
3238 '!': lambda v: (v is False) if isinstance(v, bool) else (v is None),
3240 operator_rex = re.compile(r'''(?x)
3241 (?P<op>%s)\s*(?P<key>[a-z_]+)
3242 ''' % '|
'.join(map(re.escape, UNARY_OPERATORS.keys())))
3243 m = operator_rex.fullmatch(filter_part.strip())
3245 op = UNARY_OPERATORS[m.group('op
')]
3246 actual_value = dct.get(m.group('key
'))
3247 if is_incomplete(m.group('key
')) and actual_value is None:
3249 return op(actual_value)
3251 raise ValueError('Invalid
filter part
%r' % filter_part)
3254 def match_str(filter_str, dct, incomplete=False):
3255 """ Filter a dictionary with a simple string syntax.
3256 @returns Whether the filter passes
3257 @param incomplete Set of keys that is expected to be missing from dct.
3258 Can be True/False to indicate all/none of the keys may be missing.
3259 All conditions on incomplete keys pass if the key is missing
3262 _match_one(filter_part.replace(r'\
&', '&'), dct, incomplete)
3263 for filter_part in re.split(r'(?
<!\\)&', filter_str))
3266 def match_filter_func(filters, breaking_filters=None):
3267 if not filters and not breaking_filters:
3269 repr_ = f'{match_filter_func.__module__}
.{match_filter_func.__qualname__}
({filters}
, {breaking_filters}
)'
3271 breaking_filters = match_filter_func(breaking_filters) or (lambda _, __: None)
3272 filters = set(variadic(filters or []))
3274 interactive = '-' in filters
3278 @function_with_repr.set_repr(repr_)
3279 def _match_func(info_dict, incomplete=False):
3280 ret = breaking_filters(info_dict, incomplete)
3282 raise RejectedVideoReached(ret)
3284 if not filters or any(match_str(f, info_dict, incomplete) for f in filters):
3285 return NO_DEFAULT if interactive and not incomplete else None
3287 video_title = info_dict.get('title
') or info_dict.get('id') or 'entry
'
3288 filter_str = ') |
('.join(map(str.strip, filters))
3289 return f'{video_title} does
not pass filter ({filter_str}
), skipping
..'
3293 class download_range_func:
3294 def __init__(self, chapters, ranges, from_info=False):
3295 self.chapters, self.ranges, self.from_info = chapters, ranges, from_info
3297 def __call__(self, info_dict, ydl):
3299 warning = ('There are no chapters matching the regex
' if info_dict.get('chapters
')
3300 else 'Cannot match chapters since chapter information
is unavailable
')
3301 for regex in self.chapters or []:
3302 for i, chapter in enumerate(info_dict.get('chapters
') or []):
3303 if re.search(regex, chapter['title
']):
3305 yield {**chapter, 'index': i}
3306 if self.chapters and warning:
3307 ydl.to_screen(f'[info
] {info_dict["id"]}
: {warning}
')
3309 for start, end in self.ranges or []:
3311 'start_time
': self._handle_negative_timestamp(start, info_dict),
3312 'end_time
': self._handle_negative_timestamp(end, info_dict),
3315 if self.from_info and (info_dict.get('start_time
') or info_dict.get('end_time
')):
3317 'start_time
': info_dict.get('start_time
') or 0,
3318 'end_time
': info_dict.get('end_time
') or float('inf
'),
3320 elif not self.ranges and not self.chapters:
3324 def _handle_negative_timestamp(time, info):
3325 return max(info['duration
'] + time, 0) if info.get('duration
') and time < 0 else time
3327 def __eq__(self, other):
3328 return (isinstance(other, download_range_func)
3329 and self.chapters == other.chapters and self.ranges == other.ranges)
3332 return f'{__name__}
.{type(self).__name__}
({self.chapters}
, {self.ranges}
)'
3335 def parse_dfxp_time_expr(time_expr):
3339 mobj = re.match(rf'^
(?P
<time_offset
>{NUMBER_RE}
)s?$
', time_expr)
3341 return float(mobj.group('time_offset
'))
3343 mobj = re.match(r'^
(\d
+):(\d\d
):(\d\
d(?
:(?
:\
.|
:)\d
+)?
)$
', time_expr)
3345 return 3600 * int(mobj.group(1)) + 60 * int(mobj.group(2)) + float(mobj.group(3).replace(':', '.'))
3348 def srt_subtitles_timecode(seconds):
3349 return '%02d
:%02d
:%02d
,%03d
' % timetuple_from_msec(seconds * 1000)
3352 def ass_subtitles_timecode(seconds):
3353 time = timetuple_from_msec(seconds * 1000)
3354 return '%01d
:%02d
:%02d
.%02d
' % (*time[:-1], time.milliseconds / 10)
3357 def dfxp2srt(dfxp_data):
3359 @param dfxp_data A bytes-like object containing DFXP data
3360 @returns A unicode object containing converted SRT data
3362 LEGACY_NAMESPACES = (
3363 (b'http
://www
.w3
.org
/ns
/ttml
', [
3364 b'http
://www
.w3
.org
/2004/11/ttaf1
',
3365 b'http
://www
.w3
.org
/2006/04/ttaf1
',
3366 b'http
://www
.w3
.org
/2006/10/ttaf1
',
3368 (b'http
://www
.w3
.org
/ns
/ttml
#styling', [
3369 b
'http://www.w3.org/ns/ttml#style',
3373 SUPPORTED_STYLING
= [
3382 _x
= functools
.partial(xpath_with_ns
, ns_map
={
3383 'xml': 'http://www.w3.org/XML/1998/namespace',
3384 'ttml': 'http://www.w3.org/ns/ttml',
3385 'tts': 'http://www.w3.org/ns/ttml#styling',
3391 class TTMLPElementParser
:
3393 _unclosed_elements
= []
3394 _applied_styles
= []
3396 def start(self
, tag
, attrib
):
3397 if tag
in (_x('ttml:br'), 'br'):
3400 unclosed_elements
= []
3402 element_style_id
= attrib
.get('style')
3404 style
.update(default_style
)
3405 if element_style_id
:
3406 style
.update(styles
.get(element_style_id
, {}))
3407 for prop
in SUPPORTED_STYLING
:
3408 prop_val
= attrib
.get(_x('tts:' + prop
))
3410 style
[prop
] = prop_val
3413 for k
, v
in sorted(style
.items()):
3414 if self
._applied
_styles
and self
._applied
_styles
[-1].get(k
) == v
:
3417 font
+= ' color="%s"' % v
3418 elif k
== 'fontSize':
3419 font
+= ' size="%s"' % v
3420 elif k
== 'fontFamily':
3421 font
+= ' face="%s"' % v
3422 elif k
== 'fontWeight' and v
== 'bold':
3424 unclosed_elements
.append('b')
3425 elif k
== 'fontStyle' and v
== 'italic':
3427 unclosed_elements
.append('i')
3428 elif k
== 'textDecoration' and v
== 'underline':
3430 unclosed_elements
.append('u')
3432 self
._out
+= '<font' + font
+ '>'
3433 unclosed_elements
.append('font')
3435 if self
._applied
_styles
:
3436 applied_style
.update(self
._applied
_styles
[-1])
3437 applied_style
.update(style
)
3438 self
._applied
_styles
.append(applied_style
)
3439 self
._unclosed
_elements
.append(unclosed_elements
)
3442 if tag
not in (_x('ttml:br'), 'br'):
3443 unclosed_elements
= self
._unclosed
_elements
.pop()
3444 for element
in reversed(unclosed_elements
):
3445 self
._out
+= '</%s>' % element
3446 if unclosed_elements
and self
._applied
_styles
:
3447 self
._applied
_styles
.pop()
3449 def data(self
, data
):
3453 return self
._out
.strip()
3455 # Fix UTF-8 encoded file wrongly marked as UTF-16. See https://github.com/yt-dlp/yt-dlp/issues/6543#issuecomment-1477169870
3456 # This will not trigger false positives since only UTF-8 text is being replaced
3457 dfxp_data
= dfxp_data
.replace(b
'encoding=\'UTF-16\'', b
'encoding=\'UTF-8\'')
3459 def parse_node(node
):
3460 target
= TTMLPElementParser()
3461 parser
= xml
.etree
.ElementTree
.XMLParser(target
=target
)
3462 parser
.feed(xml
.etree
.ElementTree
.tostring(node
))
3463 return parser
.close()
3465 for k
, v
in LEGACY_NAMESPACES
:
3467 dfxp_data
= dfxp_data
.replace(ns
, k
)
3469 dfxp
= compat_etree_fromstring(dfxp_data
)
3471 paras
= dfxp
.findall(_x('.//ttml:p')) or dfxp
.findall('.//p')
3474 raise ValueError('Invalid dfxp/TTML subtitle')
3478 for style
in dfxp
.findall(_x('.//ttml:style')):
3479 style_id
= style
.get('id') or style
.get(_x('xml:id'))
3482 parent_style_id
= style
.get('style')
3484 if parent_style_id
not in styles
:
3487 styles
[style_id
] = styles
[parent_style_id
].copy()
3488 for prop
in SUPPORTED_STYLING
:
3489 prop_val
= style
.get(_x('tts:' + prop
))
3491 styles
.setdefault(style_id
, {})[prop
] = prop_val
3497 for p
in ('body', 'div'):
3498 ele
= xpath_element(dfxp
, [_x('.//ttml:' + p
), './/' + p
])
3501 style
= styles
.get(ele
.get('style'))
3504 default_style
.update(style
)
3506 for para
, index
in zip(paras
, itertools
.count(1)):
3507 begin_time
= parse_dfxp_time_expr(para
.attrib
.get('begin'))
3508 end_time
= parse_dfxp_time_expr(para
.attrib
.get('end'))
3509 dur
= parse_dfxp_time_expr(para
.attrib
.get('dur'))
3510 if begin_time
is None:
3515 end_time
= begin_time
+ dur
3516 out
.append('%d\n%s --> %s\n%s\n\n' % (
3518 srt_subtitles_timecode(begin_time
),
3519 srt_subtitles_timecode(end_time
),
3525 def cli_option(params
, command_option
, param
, separator
=None):
3526 param
= params
.get(param
)
3527 return ([] if param
is None
3528 else [command_option
, str(param
)] if separator
is None
3529 else [f
'{command_option}{separator}{param}'])
3532 def cli_bool_option(params
, command_option
, param
, true_value
='true', false_value
='false', separator
=None):
3533 param
= params
.get(param
)
3534 assert param
in (True, False, None)
3535 return cli_option({True: true_value, False: false_value}
, command_option
, param
, separator
)
3538 def cli_valueless_option(params
, command_option
, param
, expected_value
=True):
3539 return [command_option
] if params
.get(param
) == expected_value
else []
3542 def cli_configuration_args(argdict
, keys
, default
=[], use_compat
=True):
3543 if isinstance(argdict
, (list, tuple)): # for backward compatibility
3550 assert isinstance(argdict
, dict)
3552 assert isinstance(keys
, (list, tuple))
3553 for key_list
in keys
:
3554 arg_list
= list(filter(
3555 lambda x
: x
is not None,
3556 [argdict
.get(key
.lower()) for key
in variadic(key_list
)]))
3558 return [arg
for args
in arg_list
for arg
in args
]
3562 def _configuration_args(main_key
, argdict
, exe
, keys
=None, default
=[], use_compat
=True):
3563 main_key
, exe
= main_key
.lower(), exe
.lower()
3564 root_key
= exe
if main_key
== exe
else f
'{main_key}+{exe}'
3565 keys
= [f
'{root_key}{k}' for k
in (keys
or [''])]
3566 if root_key
in keys
:
3568 keys
.append((main_key
, exe
))
3569 keys
.append('default')
3572 return cli_configuration_args(argdict
, keys
, default
, use_compat
)
3576 # See http://www.loc.gov/standards/iso639-2/ISO-639-2_utf-8.txt
3635 'iw': 'heb', # Replaced by he in 1989 revision
3645 'in': 'ind', # Replaced by id in 1989 revision
3761 'ji': 'yid', # Replaced by yi in 1989 revision
3769 def short2long(cls
, code
):
3770 """Convert language code from ISO 639-1 to ISO 639-2/T"""
3771 return cls
._lang
_map
.get(code
[:2])
3774 def long2short(cls
, code
):
3775 """Convert language code from ISO 639-2/T to ISO 639-1"""
3776 for short_name
, long_name
in cls
._lang
_map
.items():
3777 if long_name
== code
:
3782 # From http://data.okfn.org/data/core/country-list
3784 'AF': 'Afghanistan',
3785 'AX': 'Åland Islands',
3788 'AS': 'American Samoa',
3793 'AG': 'Antigua and Barbuda',
3810 'BO': 'Bolivia, Plurinational State of',
3811 'BQ': 'Bonaire, Sint Eustatius and Saba',
3812 'BA': 'Bosnia and Herzegovina',
3814 'BV': 'Bouvet Island',
3816 'IO': 'British Indian Ocean Territory',
3817 'BN': 'Brunei Darussalam',
3819 'BF': 'Burkina Faso',
3825 'KY': 'Cayman Islands',
3826 'CF': 'Central African Republic',
3830 'CX': 'Christmas Island',
3831 'CC': 'Cocos (Keeling) Islands',
3835 'CD': 'Congo, the Democratic Republic of the',
3836 'CK': 'Cook Islands',
3838 'CI': 'Côte d\'Ivoire',
3843 'CZ': 'Czech Republic',
3847 'DO': 'Dominican Republic',
3850 'SV': 'El Salvador',
3851 'GQ': 'Equatorial Guinea',
3855 'FK': 'Falkland Islands (Malvinas)',
3856 'FO': 'Faroe Islands',
3860 'GF': 'French Guiana',
3861 'PF': 'French Polynesia',
3862 'TF': 'French Southern Territories',
3877 'GW': 'Guinea-Bissau',
3880 'HM': 'Heard Island and McDonald Islands',
3881 'VA': 'Holy See (Vatican City State)',
3888 'IR': 'Iran, Islamic Republic of',
3891 'IM': 'Isle of Man',
3901 'KP': 'Korea, Democratic People\'s Republic of',
3902 'KR': 'Korea, Republic of',
3905 'LA': 'Lao People\'s Democratic Republic',
3911 'LI': 'Liechtenstein',
3915 'MK': 'Macedonia, the Former Yugoslav Republic of',
3922 'MH': 'Marshall Islands',
3928 'FM': 'Micronesia, Federated States of',
3929 'MD': 'Moldova, Republic of',
3940 'NL': 'Netherlands',
3941 'NC': 'New Caledonia',
3942 'NZ': 'New Zealand',
3947 'NF': 'Norfolk Island',
3948 'MP': 'Northern Mariana Islands',
3953 'PS': 'Palestine, State of',
3955 'PG': 'Papua New Guinea',
3958 'PH': 'Philippines',
3962 'PR': 'Puerto Rico',
3966 'RU': 'Russian Federation',
3968 'BL': 'Saint Barthélemy',
3969 'SH': 'Saint Helena, Ascension and Tristan da Cunha',
3970 'KN': 'Saint Kitts and Nevis',
3971 'LC': 'Saint Lucia',
3972 'MF': 'Saint Martin (French part)',
3973 'PM': 'Saint Pierre and Miquelon',
3974 'VC': 'Saint Vincent and the Grenadines',
3977 'ST': 'Sao Tome and Principe',
3978 'SA': 'Saudi Arabia',
3982 'SL': 'Sierra Leone',
3984 'SX': 'Sint Maarten (Dutch part)',
3987 'SB': 'Solomon Islands',
3989 'ZA': 'South Africa',
3990 'GS': 'South Georgia and the South Sandwich Islands',
3991 'SS': 'South Sudan',
3996 'SJ': 'Svalbard and Jan Mayen',
3999 'CH': 'Switzerland',
4000 'SY': 'Syrian Arab Republic',
4001 'TW': 'Taiwan, Province of China',
4003 'TZ': 'Tanzania, United Republic of',
4005 'TL': 'Timor-Leste',
4009 'TT': 'Trinidad and Tobago',
4012 'TM': 'Turkmenistan',
4013 'TC': 'Turks and Caicos Islands',
4017 'AE': 'United Arab Emirates',
4018 'GB': 'United Kingdom',
4019 'US': 'United States',
4020 'UM': 'United States Minor Outlying Islands',
4024 'VE': 'Venezuela, Bolivarian Republic of',
4026 'VG': 'Virgin Islands, British',
4027 'VI': 'Virgin Islands, U.S.',
4028 'WF': 'Wallis and Futuna',
4029 'EH': 'Western Sahara',
4033 # Not ISO 3166 codes, but used for IP blocks
4034 'AP': 'Asia/Pacific Region',
4039 def short2full(cls
, code
):
4040 """Convert an ISO 3166-2 country code to the corresponding full name"""
4041 return cls
._country
_map
.get(code
.upper())
4045 # Major IPv4 address blocks per country
4047 'AD': '46.172.224.0/19',
4048 'AE': '94.200.0.0/13',
4049 'AF': '149.54.0.0/17',
4050 'AG': '209.59.64.0/18',
4051 'AI': '204.14.248.0/21',
4052 'AL': '46.99.0.0/16',
4053 'AM': '46.70.0.0/15',
4054 'AO': '105.168.0.0/13',
4055 'AP': '182.50.184.0/21',
4056 'AQ': '23.154.160.0/24',
4057 'AR': '181.0.0.0/12',
4058 'AS': '202.70.112.0/20',
4059 'AT': '77.116.0.0/14',
4060 'AU': '1.128.0.0/11',
4061 'AW': '181.41.0.0/18',
4062 'AX': '185.217.4.0/22',
4063 'AZ': '5.197.0.0/16',
4064 'BA': '31.176.128.0/17',
4065 'BB': '65.48.128.0/17',
4066 'BD': '114.130.0.0/16',
4068 'BF': '102.178.0.0/15',
4069 'BG': '95.42.0.0/15',
4070 'BH': '37.131.0.0/17',
4071 'BI': '154.117.192.0/18',
4072 'BJ': '137.255.0.0/16',
4073 'BL': '185.212.72.0/23',
4074 'BM': '196.12.64.0/18',
4075 'BN': '156.31.0.0/16',
4076 'BO': '161.56.0.0/16',
4077 'BQ': '161.0.80.0/20',
4078 'BR': '191.128.0.0/12',
4079 'BS': '24.51.64.0/18',
4080 'BT': '119.2.96.0/19',
4081 'BW': '168.167.0.0/16',
4082 'BY': '178.120.0.0/13',
4083 'BZ': '179.42.192.0/18',
4084 'CA': '99.224.0.0/11',
4085 'CD': '41.243.0.0/16',
4086 'CF': '197.242.176.0/21',
4087 'CG': '160.113.0.0/16',
4088 'CH': '85.0.0.0/13',
4089 'CI': '102.136.0.0/14',
4090 'CK': '202.65.32.0/19',
4091 'CL': '152.172.0.0/14',
4092 'CM': '102.244.0.0/14',
4093 'CN': '36.128.0.0/10',
4094 'CO': '181.240.0.0/12',
4095 'CR': '201.192.0.0/12',
4096 'CU': '152.206.0.0/15',
4097 'CV': '165.90.96.0/19',
4098 'CW': '190.88.128.0/17',
4099 'CY': '31.153.0.0/16',
4100 'CZ': '88.100.0.0/14',
4102 'DJ': '197.241.0.0/17',
4103 'DK': '87.48.0.0/12',
4104 'DM': '192.243.48.0/20',
4105 'DO': '152.166.0.0/15',
4106 'DZ': '41.96.0.0/12',
4107 'EC': '186.68.0.0/15',
4108 'EE': '90.190.0.0/15',
4109 'EG': '156.160.0.0/11',
4110 'ER': '196.200.96.0/20',
4111 'ES': '88.0.0.0/11',
4112 'ET': '196.188.0.0/14',
4113 'EU': '2.16.0.0/13',
4114 'FI': '91.152.0.0/13',
4115 'FJ': '144.120.0.0/16',
4116 'FK': '80.73.208.0/21',
4117 'FM': '119.252.112.0/20',
4118 'FO': '88.85.32.0/19',
4120 'GA': '41.158.0.0/15',
4122 'GD': '74.122.88.0/21',
4123 'GE': '31.146.0.0/16',
4124 'GF': '161.22.64.0/18',
4125 'GG': '62.68.160.0/19',
4126 'GH': '154.160.0.0/12',
4127 'GI': '95.164.0.0/16',
4128 'GL': '88.83.0.0/19',
4129 'GM': '160.182.0.0/15',
4130 'GN': '197.149.192.0/18',
4131 'GP': '104.250.0.0/19',
4132 'GQ': '105.235.224.0/20',
4133 'GR': '94.64.0.0/13',
4134 'GT': '168.234.0.0/16',
4135 'GU': '168.123.0.0/16',
4136 'GW': '197.214.80.0/20',
4137 'GY': '181.41.64.0/18',
4138 'HK': '113.252.0.0/14',
4139 'HN': '181.210.0.0/16',
4140 'HR': '93.136.0.0/13',
4141 'HT': '148.102.128.0/17',
4142 'HU': '84.0.0.0/14',
4143 'ID': '39.192.0.0/10',
4144 'IE': '87.32.0.0/12',
4145 'IL': '79.176.0.0/13',
4146 'IM': '5.62.80.0/20',
4147 'IN': '117.192.0.0/10',
4148 'IO': '203.83.48.0/21',
4149 'IQ': '37.236.0.0/14',
4150 'IR': '2.176.0.0/12',
4151 'IS': '82.221.0.0/16',
4152 'IT': '79.0.0.0/10',
4153 'JE': '87.244.64.0/18',
4154 'JM': '72.27.0.0/17',
4155 'JO': '176.29.0.0/16',
4156 'JP': '133.0.0.0/8',
4157 'KE': '105.48.0.0/12',
4158 'KG': '158.181.128.0/17',
4159 'KH': '36.37.128.0/17',
4160 'KI': '103.25.140.0/22',
4161 'KM': '197.255.224.0/20',
4162 'KN': '198.167.192.0/19',
4163 'KP': '175.45.176.0/22',
4164 'KR': '175.192.0.0/10',
4165 'KW': '37.36.0.0/14',
4166 'KY': '64.96.0.0/15',
4167 'KZ': '2.72.0.0/13',
4168 'LA': '115.84.64.0/18',
4169 'LB': '178.135.0.0/16',
4170 'LC': '24.92.144.0/20',
4171 'LI': '82.117.0.0/19',
4172 'LK': '112.134.0.0/15',
4173 'LR': '102.183.0.0/16',
4174 'LS': '129.232.0.0/17',
4175 'LT': '78.56.0.0/13',
4176 'LU': '188.42.0.0/16',
4177 'LV': '46.109.0.0/16',
4178 'LY': '41.252.0.0/14',
4179 'MA': '105.128.0.0/11',
4180 'MC': '88.209.64.0/18',
4181 'MD': '37.246.0.0/16',
4182 'ME': '178.175.0.0/17',
4183 'MF': '74.112.232.0/21',
4184 'MG': '154.126.0.0/17',
4185 'MH': '117.103.88.0/21',
4186 'MK': '77.28.0.0/15',
4187 'ML': '154.118.128.0/18',
4188 'MM': '37.111.0.0/17',
4189 'MN': '49.0.128.0/17',
4190 'MO': '60.246.0.0/16',
4191 'MP': '202.88.64.0/20',
4192 'MQ': '109.203.224.0/19',
4193 'MR': '41.188.64.0/18',
4194 'MS': '208.90.112.0/22',
4195 'MT': '46.11.0.0/16',
4196 'MU': '105.16.0.0/12',
4197 'MV': '27.114.128.0/18',
4198 'MW': '102.70.0.0/15',
4199 'MX': '187.192.0.0/11',
4200 'MY': '175.136.0.0/13',
4201 'MZ': '197.218.0.0/15',
4202 'NA': '41.182.0.0/16',
4203 'NC': '101.101.0.0/18',
4204 'NE': '197.214.0.0/18',
4205 'NF': '203.17.240.0/22',
4206 'NG': '105.112.0.0/12',
4207 'NI': '186.76.0.0/15',
4208 'NL': '145.96.0.0/11',
4209 'NO': '84.208.0.0/13',
4210 'NP': '36.252.0.0/15',
4211 'NR': '203.98.224.0/19',
4212 'NU': '49.156.48.0/22',
4213 'NZ': '49.224.0.0/14',
4214 'OM': '5.36.0.0/15',
4215 'PA': '186.72.0.0/15',
4216 'PE': '186.160.0.0/14',
4217 'PF': '123.50.64.0/18',
4218 'PG': '124.240.192.0/19',
4219 'PH': '49.144.0.0/13',
4220 'PK': '39.32.0.0/11',
4221 'PL': '83.0.0.0/11',
4222 'PM': '70.36.0.0/20',
4223 'PR': '66.50.0.0/16',
4224 'PS': '188.161.0.0/16',
4225 'PT': '85.240.0.0/13',
4226 'PW': '202.124.224.0/20',
4227 'PY': '181.120.0.0/14',
4228 'QA': '37.210.0.0/15',
4229 'RE': '102.35.0.0/16',
4230 'RO': '79.112.0.0/13',
4231 'RS': '93.86.0.0/15',
4232 'RU': '5.136.0.0/13',
4233 'RW': '41.186.0.0/16',
4234 'SA': '188.48.0.0/13',
4235 'SB': '202.1.160.0/19',
4236 'SC': '154.192.0.0/11',
4237 'SD': '102.120.0.0/13',
4238 'SE': '78.64.0.0/12',
4239 'SG': '8.128.0.0/10',
4240 'SI': '188.196.0.0/14',
4241 'SK': '78.98.0.0/15',
4242 'SL': '102.143.0.0/17',
4243 'SM': '89.186.32.0/19',
4244 'SN': '41.82.0.0/15',
4245 'SO': '154.115.192.0/18',
4246 'SR': '186.179.128.0/17',
4247 'SS': '105.235.208.0/21',
4248 'ST': '197.159.160.0/19',
4249 'SV': '168.243.0.0/16',
4250 'SX': '190.102.0.0/20',
4252 'SZ': '41.84.224.0/19',
4253 'TC': '65.255.48.0/20',
4254 'TD': '154.68.128.0/19',
4255 'TG': '196.168.0.0/14',
4256 'TH': '171.96.0.0/13',
4257 'TJ': '85.9.128.0/18',
4258 'TK': '27.96.24.0/21',
4259 'TL': '180.189.160.0/20',
4260 'TM': '95.85.96.0/19',
4261 'TN': '197.0.0.0/11',
4262 'TO': '175.176.144.0/21',
4263 'TR': '78.160.0.0/11',
4264 'TT': '186.44.0.0/15',
4265 'TV': '202.2.96.0/19',
4266 'TW': '120.96.0.0/11',
4267 'TZ': '156.156.0.0/14',
4268 'UA': '37.52.0.0/14',
4269 'UG': '102.80.0.0/13',
4271 'UY': '167.56.0.0/13',
4272 'UZ': '84.54.64.0/18',
4273 'VA': '212.77.0.0/19',
4274 'VC': '207.191.240.0/21',
4275 'VE': '186.88.0.0/13',
4276 'VG': '66.81.192.0/20',
4277 'VI': '146.226.0.0/16',
4278 'VN': '14.160.0.0/11',
4279 'VU': '202.80.32.0/20',
4280 'WF': '117.20.32.0/21',
4281 'WS': '202.4.32.0/19',
4282 'YE': '134.35.0.0/16',
4283 'YT': '41.242.116.0/22',
4284 'ZA': '41.0.0.0/11',
4285 'ZM': '102.144.0.0/13',
4286 'ZW': '102.177.192.0/18',
4290 def random_ipv4(cls
, code_or_block
):
4291 if len(code_or_block
) == 2:
4292 block
= cls
._country
_ip
_map
.get(code_or_block
.upper())
4296 block
= code_or_block
4297 addr
, preflen
= block
.split('/')
4298 addr_min
= struct
.unpack('!L', socket
.inet_aton(addr
))[0]
4299 addr_max
= addr_min |
(0xffffffff >> int(preflen
))
4300 return str(socket
.inet_ntoa(
4301 struct
.pack('!L', random
.randint(addr_min
, addr_max
))))
4304 # Both long_to_bytes and bytes_to_long are adapted from PyCrypto, which is
4305 # released into Public Domain
4306 # https://github.com/dlitz/pycrypto/blob/master/lib/Crypto/Util/number.py#L387
4308 def long_to_bytes(n
, blocksize
=0):
4309 """long_to_bytes(n:long, blocksize:int) : string
4310 Convert a long integer to a byte string.
4312 If optional blocksize is given and greater than zero, pad the front of the
4313 byte string with binary zeros so that the length is a multiple of
4316 # after much testing, this algorithm was deemed to be the fastest
4320 s
= struct
.pack('>I', n
& 0xffffffff) + s
4322 # strip off leading zeros
4323 for i
in range(len(s
)):
4324 if s
[i
] != b
'\000'[0]:
4327 # only happens when n == 0
4331 # add back some pad bytes. this could be done more efficiently w.r.t. the
4332 # de-padding being done above, but sigh...
4333 if blocksize
> 0 and len(s
) % blocksize
:
4334 s
= (blocksize
- len(s
) % blocksize
) * b
'\000' + s
4338 def bytes_to_long(s
):
4339 """bytes_to_long(string) : long
4340 Convert a byte string to a long integer.
4342 This is (essentially) the inverse of long_to_bytes().
4347 extra
= (4 - length
% 4)
4348 s
= b
'\000' * extra
+ s
4349 length
= length
+ extra
4350 for i
in range(0, length
, 4):
4351 acc
= (acc
<< 32) + struct
.unpack('>I', s
[i
:i
+ 4])[0]
4355 def ohdave_rsa_encrypt(data
, exponent
, modulus
):
4357 Implement OHDave's RSA algorithm. See http://www.ohdave.com/rsa/
4360 data: data to encrypt, bytes-like object
4361 exponent, modulus: parameter e and N of RSA algorithm, both integer
4362 Output: hex string of encrypted data
4364 Limitation: supports one block encryption only
4367 payload
= int(binascii
.hexlify(data
[::-1]), 16)
4368 encrypted
= pow(payload
, exponent
, modulus
)
4369 return '%x' % encrypted
4372 def pkcs1pad(data
, length
):
4374 Padding input data with PKCS#1 scheme
4376 @param {int[]} data input data
4377 @param {int} length target length
4378 @returns {int[]} padded data
4380 if len(data
) > length
- 11:
4381 raise ValueError('Input data too long for PKCS#1 padding')
4383 pseudo_random
= [random
.randint(0, 254) for _
in range(length
- len(data
) - 3)]
4384 return [0, 2] + pseudo_random
+ [0] + data
4387 def _base_n_table(n
, table
):
4388 if not table
and not n
:
4389 raise ValueError('Either table or n must be specified')
4390 table
= (table
or '0123456789abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ')[:n
]
4392 if n
and n
!= len(table
):
4393 raise ValueError(f
'base {n} exceeds table length {len(table)}')
4397 def encode_base_n(num
, n
=None, table
=None):
4398 """Convert given int to a base-n string"""
4399 table
= _base_n_table(n
, table
)
4403 result
, base
= '', len(table
)
4405 result
= table
[num
% base
] + result
4410 def decode_base_n(string
, n
=None, table
=None):
4411 """Convert given base-n string to int"""
4412 table
= {char: index for index, char in enumerate(_base_n_table(n, table))}
4413 result
, base
= 0, len(table
)
4415 result
= result
* base
+ table
[char
]
4419 def decode_packed_codes(code
):
4420 mobj
= re
.search(PACKED_CODES_RE
, code
)
4421 obfuscated_code
, base
, count
, symbols
= mobj
.groups()
4424 symbols
= symbols
.split('|')
4429 base_n_count
= encode_base_n(count
, base
)
4430 symbol_table
[base_n_count
] = symbols
[count
] or base_n_count
4433 r
'\b(\w+)\b', lambda mobj
: symbol_table
[mobj
.group(0)],
4437 def caesar(s
, alphabet
, shift
):
4442 alphabet
[(alphabet
.index(c
) + shift
) % l
] if c
in alphabet
else c
4447 return caesar(s
, r
'''!"#$%&'()*+,-./0123456789:;<=>?@ABCDEFGHIJKLMNOPQRSTUVWXYZ[\]^_`abcdefghijklmnopqrstuvwxyz{|}~''', 47)
4450 def parse_m3u8_attributes(attrib
):
4452 for (key
, val
) in re
.findall(r
'(?P<key>[A-Z0-9-]+)=(?P<val>"[^"]+"|[^",]+)(?:,|$)', attrib
):
4453 if val
.startswith('"'):
4459 def urshift(val
, n
):
4460 return val
>> n
if val
>= 0 else (val
+ 0x100000000) >> n
4463 def write_xattr(path
, key
, value
):
4464 # Windows: Write xattrs to NTFS Alternate Data Streams:
4465 # http://en.wikipedia.org/wiki/NTFS#Alternate_data_streams_.28ADS.29
4466 if compat_os_name
== 'nt':
4467 assert ':' not in key
4468 assert os
.path
.exists(path
)
4471 with open(f
'{path}:{key}', 'wb') as f
:
4473 except OSError as e
:
4474 raise XAttrMetadataError(e
.errno
, e
.strerror
)
4477 # UNIX Method 1. Use os.setxattr/xattrs/pyxattrs modules
4480 if callable(getattr(os
, 'setxattr', None)):
4481 setxattr
= os
.setxattr
4482 elif getattr(xattr
, '_yt_dlp__identifier', None) == 'pyxattr':
4483 # Unicode arguments are not supported in pyxattr until version 0.5.0
4484 # See https://github.com/ytdl-org/youtube-dl/issues/5498
4485 if version_tuple(xattr
.__version
__) >= (0, 5, 0):
4486 setxattr
= xattr
.set
4488 setxattr
= xattr
.setxattr
4492 setxattr(path
, key
, value
)
4493 except OSError as e
:
4494 raise XAttrMetadataError(e
.errno
, e
.strerror
)
4497 # UNIX Method 2. Use setfattr/xattr executables
4498 exe
= ('setfattr' if check_executable('setfattr', ['--version'])
4499 else 'xattr' if check_executable('xattr', ['-h']) else None)
4501 raise XAttrUnavailableError(
4502 'Couldn\'t find a tool to set the xattrs. Install either the "xattr" or "pyxattr" Python modules or the '
4503 + ('"xattr" binary' if sys
.platform
!= 'linux' else 'GNU "attr" package (which contains the "setfattr" tool)'))
4505 value
= value
.decode()
4507 _
, stderr
, returncode
= Popen
.run(
4508 [exe
, '-w', key
, value
, path
] if exe
== 'xattr' else [exe
, '-n', key
, '-v', value
, path
],
4509 text
=True, stdout
=subprocess
.PIPE
, stderr
=subprocess
.PIPE
, stdin
=subprocess
.PIPE
)
4510 except OSError as e
:
4511 raise XAttrMetadataError(e
.errno
, e
.strerror
)
4513 raise XAttrMetadataError(returncode
, stderr
)
4516 def random_birthday(year_field
, month_field
, day_field
):
4517 start_date
= dt
.date(1950, 1, 1)
4518 end_date
= dt
.date(1995, 12, 31)
4519 offset
= random
.randint(0, (end_date
- start_date
).days
)
4520 random_date
= start_date
+ dt
.timedelta(offset
)
4522 year_field
: str(random_date
.year
),
4523 month_field
: str(random_date
.month
),
4524 day_field
: str(random_date
.day
),
4528 def find_available_port(interface
=''):
4530 with socket
.socket() as sock
:
4531 sock
.bind((interface
, 0))
4532 return sock
.getsockname()[1]
4537 # Templates for internet shortcut files, which are plain text files.
4538 DOT_URL_LINK_TEMPLATE
= '''\
4543 DOT_WEBLOC_LINK_TEMPLATE
= '''\
4544 <?xml version="1.0" encoding="UTF-8"?>
4545 <!DOCTYPE plist PUBLIC "-//Apple//DTD PLIST 1.0//EN" "http://www.apple.com/DTDs/PropertyList-1.0.dtd">
4546 <plist version="1.0">
4549 \t<string>%(url)s</string>
4554 DOT_DESKTOP_LINK_TEMPLATE
= '''\
4564 'url': DOT_URL_LINK_TEMPLATE
,
4565 'desktop': DOT_DESKTOP_LINK_TEMPLATE
,
4566 'webloc': DOT_WEBLOC_LINK_TEMPLATE
,
4570 def iri_to_uri(iri
):
4572 Converts an IRI (Internationalized Resource Identifier, allowing Unicode characters) to a URI (Uniform Resource Identifier, ASCII-only).
4574 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.
4577 iri_parts
= urllib
.parse
.urlparse(iri
)
4579 if '[' in iri_parts
.netloc
:
4580 raise ValueError('IPv6 URIs are not, yet, supported.')
4581 # Querying `.netloc`, when there's only one bracket, also raises a ValueError.
4583 # 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.
4586 if iri_parts
.username
:
4587 net_location
+= urllib
.parse
.quote(iri_parts
.username
, safe
=r
"!$%&'()*+,~")
4588 if iri_parts
.password
is not None:
4589 net_location
+= ':' + urllib
.parse
.quote(iri_parts
.password
, safe
=r
"!$%&'()*+,~")
4592 net_location
+= iri_parts
.hostname
.encode('idna').decode() # Punycode for Unicode hostnames.
4593 # The 'idna' encoding produces ASCII text.
4594 if iri_parts
.port
is not None and iri_parts
.port
!= 80:
4595 net_location
+= ':' + str(iri_parts
.port
)
4597 return urllib
.parse
.urlunparse(
4601 urllib
.parse
.quote_plus(iri_parts
.path
, safe
=r
"!$%&'()*+,/:;=@|~"),
4603 # Unsure about the `safe` argument, since this is a legacy way of handling parameters.
4604 urllib
.parse
.quote_plus(iri_parts
.params
, safe
=r
"!$%&'()*+,/:;=@|~"),
4606 # Not totally sure about the `safe` argument, since the source does not explicitly mention the query URI component.
4607 urllib
.parse
.quote_plus(iri_parts
.query
, safe
=r
"!$%&'()*+,/:;=?@{|}~"),
4609 urllib
.parse
.quote_plus(iri_parts
.fragment
, safe
=r
"!#$%&'()*+,/:;=?@{|}~")))
4611 # Source for `safe` arguments: https://url.spec.whatwg.org/#percent-encoded-bytes.
4614 def to_high_limit_path(path
):
4615 if sys
.platform
in ['win32', 'cygwin']:
4616 # Work around MAX_PATH limitation on Windows. The maximum allowed length for the individual path segments may still be quite limited.
4617 return '\\\\?\\' + os
.path
.abspath(path
)
4622 def format_field(obj
, field
=None, template
='%s', ignore
=NO_DEFAULT
, default
='', func
=IDENTITY
):
4623 val
= traversal
.traverse_obj(obj
, *variadic(field
))
4624 if not val
if ignore
is NO_DEFAULT
else val
in variadic(ignore
):
4626 return template
% func(val
)
4629 def clean_podcast_url(url
):
4630 url
= re
.sub(r
'''(?x)
4634 media\.blubrry\.com| # https://create.blubrry.com/resources/podcast-media-download-statistics/getting-started/
4639 (?:dts|www)\.podtrac\.com/(?:pts/)?redirect\.[0-9a-z]{3,4}| # http://analytics.podtrac.com/how-to-measure
4642 cn\.co| # https://podcorn.com/analytics-prefix/
4643 st\.fm # https://podsights.com/docs/
4648 return re
.sub(r
'^\w+://(\w+://)', r
'\1', url
)
4651 _HEX_TABLE
= '0123456789abcdef'
4654 def random_uuidv4():
4655 return re
.sub(r
'[xy]', lambda x
: _HEX_TABLE
[random
.randint(0, 15)], 'xxxxxxxx-xxxx-4xxx-yxxx-xxxxxxxxxxxx')
4658 def make_dir(path
, to_screen
=None):
4660 dn
= os
.path
.dirname(path
)
4662 os
.makedirs(dn
, exist_ok
=True)
4664 except OSError as err
:
4665 if callable(to_screen
) is not None:
4666 to_screen(f
'unable to create directory {err}')
4670 def get_executable_path():
4671 from ..update
import _get_variant_and_executable_path
4673 return os
.path
.dirname(os
.path
.abspath(_get_variant_and_executable_path()[1]))
4676 def get_user_config_dirs(package_name
):
4677 # .config (e.g. ~/.config/package_name)
4678 xdg_config_home
= os
.getenv('XDG_CONFIG_HOME') or compat_expanduser('~/.config')
4679 yield os
.path
.join(xdg_config_home
, package_name
)
4681 # appdata (%APPDATA%/package_name)
4682 appdata_dir
= os
.getenv('appdata')
4684 yield os
.path
.join(appdata_dir
, package_name
)
4686 # home (~/.package_name)
4687 yield os
.path
.join(compat_expanduser('~'), f
'.{package_name}')
4690 def get_system_config_dirs(package_name
):
4692 yield os
.path
.join('/etc', package_name
)
4695 def time_seconds(**kwargs
):
4697 Returns TZ-aware time in seconds since the epoch (1970-01-01T00:00:00Z)
4699 return time
.time() + dt
.timedelta(**kwargs
).total_seconds()
4702 # create a JSON Web Signature (jws) with HS256 algorithm
4703 # the resulting format is in JWS Compact Serialization
4704 # implemented following JWT https://www.rfc-editor.org/rfc/rfc7519.html
4705 # implemented following JWS https://www.rfc-editor.org/rfc/rfc7515.html
4706 def jwt_encode_hs256(payload_data
, key
, headers
={}):
4712 header_data
.update(headers
)
4713 header_b64
= base64
.b64encode(json
.dumps(header_data
).encode())
4714 payload_b64
= base64
.b64encode(json
.dumps(payload_data
).encode())
4715 h
= hmac
.new(key
.encode(), header_b64
+ b
'.' + payload_b64
, hashlib
.sha256
)
4716 signature_b64
= base64
.b64encode(h
.digest())
4717 token
= header_b64
+ b
'.' + payload_b64
+ b
'.' + signature_b64
4721 # can be extended in future to verify the signature and parse header and return the algorithm used if it's not HS256
4722 def jwt_decode_hs256(jwt
):
4723 header_b64
, payload_b64
, signature_b64
= jwt
.split('.')
4724 # add trailing ='s that may have been stripped, superfluous ='s are ignored
4725 payload_data
= json
.loads(base64
.urlsafe_b64decode(f
'{payload_b64}==='))
4729 WINDOWS_VT_MODE
= False if compat_os_name
== 'nt' else None
4733 def supports_terminal_sequences(stream
):
4734 if compat_os_name
== 'nt':
4735 if not WINDOWS_VT_MODE
:
4737 elif not os
.getenv('TERM'):
4740 return stream
.isatty()
4741 except BaseException
:
4745 def windows_enable_vt_mode():
4746 """Ref: https://bugs.python.org/issue30075 """
4747 if get_windows_version() < (10, 0, 10586):
4751 import ctypes
.wintypes
4754 ENABLE_VIRTUAL_TERMINAL_PROCESSING
= 0x0004
4756 dll
= ctypes
.WinDLL('kernel32', use_last_error
=False)
4757 handle
= os
.open('CONOUT$', os
.O_RDWR
)
4759 h_out
= ctypes
.wintypes
.HANDLE(msvcrt
.get_osfhandle(handle
))
4760 dw_original_mode
= ctypes
.wintypes
.DWORD()
4761 success
= dll
.GetConsoleMode(h_out
, ctypes
.byref(dw_original_mode
))
4763 raise Exception('GetConsoleMode failed')
4765 success
= dll
.SetConsoleMode(h_out
, ctypes
.wintypes
.DWORD(
4766 dw_original_mode
.value | ENABLE_VIRTUAL_TERMINAL_PROCESSING
))
4768 raise Exception('SetConsoleMode failed')
4772 global WINDOWS_VT_MODE
4773 WINDOWS_VT_MODE
= True
4774 supports_terminal_sequences
.cache_clear()
4777 _terminal_sequences_re
= re
.compile('\033\\[[^m]+m')
4780 def remove_terminal_sequences(string
):
4781 return _terminal_sequences_re
.sub('', string
)
4784 def number_of_digits(number
):
4785 return len('%d' % number
)
4788 def join_nonempty(*values
, delim
='-', from_dict
=None):
4789 if from_dict
is not None:
4790 values
= (traversal
.traverse_obj(from_dict
, variadic(v
)) for v
in values
)
4791 return delim
.join(map(str, filter(None, values
)))
4794 def scale_thumbnails_to_max_format_width(formats
, thumbnails
, url_width_re
):
4796 Find the largest format dimensions in terms of video width and, for each thumbnail:
4797 * Modify the URL: Match the width with the provided regex and replace with the former width
4800 This function is useful with video services that scale the provided thumbnails on demand
4802 _keys
= ('width', 'height')
4803 max_dimensions
= max(
4804 (tuple(format
.get(k
) or 0 for k
in _keys
) for format
in formats
),
4806 if not max_dimensions
[0]:
4810 {'url': re.sub(url_width_re, str(max_dimensions[0]), thumbnail['url'])}
,
4811 dict(zip(_keys
, max_dimensions
)), thumbnail
)
4812 for thumbnail
in thumbnails
4816 def parse_http_range(range):
4817 """ Parse value of "Range" or "Content-Range" HTTP header into tuple. """
4819 return None, None, None
4820 crg
= re
.search(r
'bytes[ =](\d+)-(\d+)?(?:/(\d+))?', range)
4822 return None, None, None
4823 return int(crg
.group(1)), int_or_none(crg
.group(2)), int_or_none(crg
.group(3))
4826 def read_stdin(what
):
4828 eof
= 'Ctrl+Z' if compat_os_name
== 'nt' else 'Ctrl+D'
4829 write_string(f
'Reading {what} from STDIN - EOF ({eof}) to end:\n')
4833 def determine_file_encoding(data
):
4835 Detect the text encoding used
4836 @returns (encoding, bytes to skip)
4839 # BOM marks are given priority over declarations
4840 for bom
, enc
in BOMS
:
4841 if data
.startswith(bom
):
4842 return enc
, len(bom
)
4844 # Strip off all null bytes to match even when UTF-16 or UTF-32 is used.
4845 # We ignore the endianness to get a good enough match
4846 data
= data
.replace(b
'\0', b
'')
4847 mobj
= re
.match(rb
'(?m)^#\s*coding\s*:\s*(\S+)\s*$', data
)
4848 return mobj
.group(1).decode() if mobj
else None, 0
4855 __initialized
= False
4857 def __init__(self
, parser
, label
=None):
4858 self
.parser
, self
.label
= parser
, label
4859 self
._loaded
_paths
, self
.configs
= set(), []
4861 def init(self
, args
=None, filename
=None):
4862 assert not self
.__initialized
4863 self
.own_args
, self
.filename
= args
, filename
4864 return self
.load_configs()
4866 def load_configs(self
):
4869 location
= os
.path
.realpath(self
.filename
)
4870 directory
= os
.path
.dirname(location
)
4871 if location
in self
._loaded
_paths
:
4873 self
._loaded
_paths
.add(location
)
4875 self
.__initialized
= True
4876 opts
, _
= self
.parser
.parse_known_args(self
.own_args
)
4877 self
.parsed_args
= self
.own_args
4878 for location
in opts
.config_locations
or []:
4880 if location
in self
._loaded
_paths
:
4882 self
._loaded
_paths
.add(location
)
4883 self
.append_config(shlex
.split(read_stdin('options'), comments
=True), label
='stdin')
4885 location
= os
.path
.join(directory
, expand_path(location
))
4886 if os
.path
.isdir(location
):
4887 location
= os
.path
.join(location
, 'yt-dlp.conf')
4888 if not os
.path
.exists(location
):
4889 self
.parser
.error(f
'config location {location} does not exist')
4890 self
.append_config(self
.read_file(location
), location
)
4894 label
= join_nonempty(
4895 self
.label
, 'config', f
'"{self.filename}"' if self
.filename
else '',
4897 return join_nonempty(
4898 self
.own_args
is not None and f
'{label[0].upper()}{label[1:]}: {self.hide_login_info(self.own_args)}',
4899 *(f
'\n{c}'.replace('\n', '\n| ')[1:] for c
in self
.configs
),
4903 def read_file(filename
, default
=[]):
4905 optionf
= open(filename
, 'rb')
4907 return default
# silently skip if file is not present
4909 enc
, skip
= determine_file_encoding(optionf
.read(512))
4910 optionf
.seek(skip
, io
.SEEK_SET
)
4912 enc
= None # silently skip read errors
4914 # FIXME: https://github.com/ytdl-org/youtube-dl/commit/dfe5fa49aed02cf36ba9f743b11b0903554b5e56
4915 contents
= optionf
.read().decode(enc
or preferredencoding())
4916 res
= shlex
.split(contents
, comments
=True)
4917 except Exception as err
:
4918 raise ValueError(f
'Unable to parse "{filename}": {err}')
4924 def hide_login_info(opts
):
4925 PRIVATE_OPTS
= {'-p', '--password', '-u', '--username', '--video-password', '--ap-password', '--ap-username'}
4926 eqre
= re
.compile('^(?P<key>' + ('|'.join(re
.escape(po
) for po
in PRIVATE_OPTS
)) + ')=.+$')
4931 return m
.group('key') + '=PRIVATE'
4935 opts
= list(map(_scrub_eq
, opts
))
4936 for idx
, opt
in enumerate(opts
):
4937 if opt
in PRIVATE_OPTS
and idx
+ 1 < len(opts
):
4938 opts
[idx
+ 1] = 'PRIVATE'
4941 def append_config(self
, *args
, label
=None):
4942 config
= type(self
)(self
.parser
, label
)
4943 config
._loaded
_paths
= self
._loaded
_paths
4944 if config
.init(*args
):
4945 self
.configs
.append(config
)
4949 for config
in reversed(self
.configs
):
4950 yield from config
.all_args
4951 yield from self
.parsed_args
or []
4953 def parse_known_args(self
, **kwargs
):
4954 return self
.parser
.parse_known_args(self
.all_args
, **kwargs
)
4956 def parse_args(self
):
4957 return self
.parser
.parse_args(self
.all_args
)
4960 def merge_headers(*dicts
):
4961 """Merge dicts of http headers case insensitively, prioritizing the latter ones"""
4962 return {k.title(): v for k, v in itertools.chain.from_iterable(map(dict.items, dicts))}
4965 def cached_method(f
):
4966 """Cache a method"""
4967 signature
= inspect
.signature(f
)
4970 def wrapper(self
, *args
, **kwargs
):
4971 bound_args
= signature
.bind(self
, *args
, **kwargs
)
4972 bound_args
.apply_defaults()
4973 key
= tuple(bound_args
.arguments
.values())[1:]
4975 cache
= vars(self
).setdefault('_cached_method__cache', {}).setdefault(f.__name__, {}
)
4976 if key
not in cache
:
4977 cache
[key
] = f(self
, *args
, **kwargs
)
4982 class classproperty
:
4983 """property access for class methods with optional caching"""
4984 def __new__(cls
, func
=None, *args
, **kwargs
):
4986 return functools
.partial(cls
, *args
, **kwargs
)
4987 return super().__new
__(cls
)
4989 def __init__(self
, func
, *, cache
=False):
4990 functools
.update_wrapper(self
, func
)
4992 self
._cache
= {} if cache
else None
4994 def __get__(self
, _
, cls
):
4995 if self
._cache
is None:
4996 return self
.func(cls
)
4997 elif cls
not in self
._cache
:
4998 self
._cache
[cls
] = self
.func(cls
)
4999 return self
._cache
[cls
]
5002 class function_with_repr
:
5003 def __init__(self
, func
, repr_
=None):
5004 functools
.update_wrapper(self
, func
)
5005 self
.func
, self
.__repr
= func
, repr_
5007 def __call__(self
, *args
, **kwargs
):
5008 return self
.func(*args
, **kwargs
)
5011 def set_repr(cls
, repr_
):
5012 return functools
.partial(cls
, repr_
=repr_
)
5017 return f
'{self.func.__module__}.{self.func.__qualname__}'
5020 class Namespace(types
.SimpleNamespace
):
5021 """Immutable namespace"""
5024 return iter(self
.__dict
__.values())
5028 return self
.__dict
__.items()
5031 MEDIA_EXTENSIONS
= Namespace(
5032 common_video
=('avi', 'flv', 'mkv', 'mov', 'mp4', 'webm'),
5033 video
=('3g2', '3gp', 'f4v', 'mk3d', 'divx', 'mpg', 'ogv', 'm4v', 'wmv'),
5034 common_audio
=('aiff', 'alac', 'flac', 'm4a', 'mka', 'mp3', 'ogg', 'opus', 'wav'),
5035 audio
=('aac', 'ape', 'asf', 'f4a', 'f4b', 'm4b', 'm4p', 'm4r', 'oga', 'ogx', 'spx', 'vorbis', 'wma', 'weba'),
5036 thumbnails
=('jpg', 'png', 'webp'),
5037 storyboards
=('mhtml', ),
5038 subtitles
=('srt', 'vtt', 'ass', 'lrc'),
5039 manifests
=('f4f', 'f4m', 'm3u8', 'smil', 'mpd'),
5041 MEDIA_EXTENSIONS
.video
+= MEDIA_EXTENSIONS
.common_video
5042 MEDIA_EXTENSIONS
.audio
+= MEDIA_EXTENSIONS
.common_audio
5044 KNOWN_EXTENSIONS
= (*MEDIA_EXTENSIONS
.video
, *MEDIA_EXTENSIONS
.audio
, *MEDIA_EXTENSIONS
.manifests
)
5049 for retry in RetryManager(...):
5052 except SomeException as err:
5056 attempt
, _error
= 0, None
5058 def __init__(self
, _retries
, _error_callback
, **kwargs
):
5059 self
.retries
= _retries
or 0
5060 self
.error_callback
= functools
.partial(_error_callback
, **kwargs
)
5062 def _should_retry(self
):
5063 return self
._error
is not NO_DEFAULT
and self
.attempt
<= self
.retries
5067 if self
._error
is NO_DEFAULT
:
5072 def error(self
, value
):
5076 while self
._should
_retry
():
5077 self
.error
= NO_DEFAULT
5081 self
.error_callback(self
.error
, self
.attempt
, self
.retries
)
5084 def report_retry(e
, count
, retries
, *, sleep_func
, info
, warn
, error
=None, suffix
=None):
5085 """Utility function for reporting retries"""
5088 return error(f
'{e}. Giving up after {count - 1} retries') if count
> 1 else error(str(e
))
5093 elif isinstance(e
, ExtractorError
):
5094 e
= remove_end(str_or_none(e
.cause
) or e
.orig_msg
, '.')
5095 warn(f
'{e}. Retrying{format_field(suffix, None, " %s")} ({count}/{retries})...')
5097 delay
= float_or_none(sleep_func(n
=count
- 1)) if callable(sleep_func
) else sleep_func
5099 info(f
'Sleeping {delay:.2f} seconds ...')
5103 def make_archive_id(ie
, video_id
):
5104 ie_key
= ie
if isinstance(ie
, str) else ie
.ie_key()
5105 return f
'{ie_key.lower()} {video_id}'
5108 def truncate_string(s
, left
, right
=0):
5109 assert left
> 3 and right
>= 0
5110 if s
is None or len(s
) <= left
+ right
:
5112 return f
'{s[:left - 3]}...{s[-right:] if right else ""}'
5115 def orderedSet_from_options(options
, alias_dict
, *, use_regex
=False, start
=None):
5116 assert 'all' in alias_dict
, '"all" alias is required'
5117 requested
= list(start
or [])
5119 discard
= val
.startswith('-')
5123 if val
in alias_dict
:
5124 val
= alias_dict
[val
] if not discard
else [
5125 i
[1:] if i
.startswith('-') else f
'-{i}' for i
in alias_dict
[val
]]
5126 # NB: Do not allow regex in aliases for performance
5127 requested
= orderedSet_from_options(val
, alias_dict
, start
=requested
)
5130 current
= (filter(re
.compile(val
, re
.I
).fullmatch
, alias_dict
['all']) if use_regex
5131 else [val
] if val
in alias_dict
['all'] else None)
5133 raise ValueError(val
)
5136 for item
in current
:
5137 while item
in requested
:
5138 requested
.remove(item
)
5140 requested
.extend(current
)
5142 return orderedSet(requested
)
5147 regex
= r
' *((?P<reverse>\+)?(?P<field>[a-zA-Z0-9_]+)((?P<separator>[~:])(?P<limit>.*?))?)? *$'
5149 default
= ('hidden', 'aud_or_vid', 'hasvid', 'ie_pref', 'lang', 'quality',
5150 'res', 'fps', 'hdr:12', 'vcodec:vp9.2', 'channels', 'acodec',
5151 'size', 'br', 'asr', 'proto', 'ext', 'hasaud', 'source', 'id') # These must not be aliases
5152 ytdl_default
= ('hasaud', 'lang', 'quality', 'tbr', 'filesize', 'vbr',
5153 'height', 'width', 'proto', 'vext', 'abr', 'aext',
5154 'fps', 'fs_approx', 'source', 'id')
5157 'vcodec': {'type': 'ordered', 'regex': True,
5158 'order': ['av0?1', 'vp0?9.2', 'vp0?9', '[hx]265|he?vc?', '[hx]264|avc', 'vp0?8', 'mp4v|h263', 'theora', '', None, 'none']},
5159 'acodec': {'type': 'ordered', 'regex': True,
5160 'order': ['[af]lac', 'wav|aiff', 'opus', 'vorbis|ogg', 'aac', 'mp?4a?', 'mp3', 'ac-?4', 'e-?a?c-?3', 'ac-?3', 'dts', '', None, 'none']},
5161 'hdr': {'type': 'ordered', 'regex': True, 'field': 'dynamic_range',
5162 'order': ['dv', '(hdr)?12', r
'(hdr)?10\+', '(hdr)?10', 'hlg', '', 'sdr', None]},
5163 'proto': {'type': 'ordered', 'regex': True, 'field': 'protocol',
5164 'order': ['(ht|f)tps', '(ht|f)tp$', 'm3u8.*', '.*dash', 'websocket_frag', 'rtmpe?', '', 'mms|rtsp', 'ws|websocket', 'f4']},
5165 'vext': {'type': 'ordered', 'field': 'video_ext',
5166 'order': ('mp4', 'mov', 'webm', 'flv', '', 'none'),
5167 'order_free': ('webm', 'mp4', 'mov', 'flv', '', 'none')},
5168 'aext': {'type': 'ordered', 'regex': True, 'field': 'audio_ext',
5169 'order': ('m4a', 'aac', 'mp3', 'ogg', 'opus', 'web[am]', '', 'none'),
5170 'order_free': ('ogg', 'opus', 'web[am]', 'mp3', 'm4a', 'aac', '', 'none')},
5171 'hidden': {'visible': False, 'forced': True, 'type': 'extractor', 'max': -1000}
,
5172 'aud_or_vid': {'visible': False, 'forced': True, 'type': 'multiple',
5173 'field': ('vcodec', 'acodec'),
5174 'function': lambda it
: int(any(v
!= 'none' for v
in it
))},
5175 'ie_pref': {'priority': True, 'type': 'extractor'}
,
5176 'hasvid': {'priority': True, 'field': 'vcodec', 'type': 'boolean', 'not_in_list': ('none',)}
,
5177 'hasaud': {'field': 'acodec', 'type': 'boolean', 'not_in_list': ('none',)}
,
5178 'lang': {'convert': 'float', 'field': 'language_preference', 'default': -1}
,
5179 'quality': {'convert': 'float', 'default': -1}
,
5180 'filesize': {'convert': 'bytes'}
,
5181 'fs_approx': {'convert': 'bytes', 'field': 'filesize_approx'}
,
5182 'id': {'convert': 'string', 'field': 'format_id'}
,
5183 'height': {'convert': 'float_none'}
,
5184 'width': {'convert': 'float_none'}
,
5185 'fps': {'convert': 'float_none'}
,
5186 'channels': {'convert': 'float_none', 'field': 'audio_channels'}
,
5187 'tbr': {'convert': 'float_none'}
,
5188 'vbr': {'convert': 'float_none'}
,
5189 'abr': {'convert': 'float_none'}
,
5190 'asr': {'convert': 'float_none'}
,
5191 'source': {'convert': 'float', 'field': 'source_preference', 'default': -1}
,
5193 'codec': {'type': 'combined', 'field': ('vcodec', 'acodec')}
,
5194 'br': {'type': 'multiple', 'field': ('tbr', 'vbr', 'abr'), 'convert': 'float_none',
5195 'function': lambda it
: next(filter(None, it
), None)},
5196 'size': {'type': 'multiple', 'field': ('filesize', 'fs_approx'), 'convert': 'bytes',
5197 'function': lambda it
: next(filter(None, it
), None)},
5198 'ext': {'type': 'combined', 'field': ('vext', 'aext')}
,
5199 'res': {'type': 'multiple', 'field': ('height', 'width'),
5200 'function': lambda it
: (lambda l
: min(l
) if l
else 0)(tuple(filter(None, it
)))},
5202 # Actual field names
5203 'format_id': {'type': 'alias', 'field': 'id'}
,
5204 'preference': {'type': 'alias', 'field': 'ie_pref'}
,
5205 'language_preference': {'type': 'alias', 'field': 'lang'}
,
5206 'source_preference': {'type': 'alias', 'field': 'source'}
,
5207 'protocol': {'type': 'alias', 'field': 'proto'}
,
5208 'filesize_approx': {'type': 'alias', 'field': 'fs_approx'}
,
5209 'audio_channels': {'type': 'alias', 'field': 'channels'}
,
5212 'dimension': {'type': 'alias', 'field': 'res', 'deprecated': True}
,
5213 'resolution': {'type': 'alias', 'field': 'res', 'deprecated': True}
,
5214 'extension': {'type': 'alias', 'field': 'ext', 'deprecated': True}
,
5215 'bitrate': {'type': 'alias', 'field': 'br', 'deprecated': True}
,
5216 'total_bitrate': {'type': 'alias', 'field': 'tbr', 'deprecated': True}
,
5217 'video_bitrate': {'type': 'alias', 'field': 'vbr', 'deprecated': True}
,
5218 'audio_bitrate': {'type': 'alias', 'field': 'abr', 'deprecated': True}
,
5219 'framerate': {'type': 'alias', 'field': 'fps', 'deprecated': True}
,
5220 'filesize_estimate': {'type': 'alias', 'field': 'size', 'deprecated': True}
,
5221 'samplerate': {'type': 'alias', 'field': 'asr', 'deprecated': True}
,
5222 'video_ext': {'type': 'alias', 'field': 'vext', 'deprecated': True}
,
5223 'audio_ext': {'type': 'alias', 'field': 'aext', 'deprecated': True}
,
5224 'video_codec': {'type': 'alias', 'field': 'vcodec', 'deprecated': True}
,
5225 'audio_codec': {'type': 'alias', 'field': 'acodec', 'deprecated': True}
,
5226 'video': {'type': 'alias', 'field': 'hasvid', 'deprecated': True}
,
5227 'has_video': {'type': 'alias', 'field': 'hasvid', 'deprecated': True}
,
5228 'audio': {'type': 'alias', 'field': 'hasaud', 'deprecated': True}
,
5229 'has_audio': {'type': 'alias', 'field': 'hasaud', 'deprecated': True}
,
5230 'extractor': {'type': 'alias', 'field': 'ie_pref', 'deprecated': True}
,
5231 'extractor_preference': {'type': 'alias', 'field': 'ie_pref', 'deprecated': True}
,
5234 def __init__(self
, ydl
, field_preference
):
5237 self
.evaluate_params(self
.ydl
.params
, field_preference
)
5238 if ydl
.params
.get('verbose'):
5239 self
.print_verbose_info(self
.ydl
.write_debug
)
5241 def _get_field_setting(self
, field
, key
):
5242 if field
not in self
.settings
:
5243 if key
in ('forced', 'priority'):
5245 self
.ydl
.deprecated_feature(f
'Using arbitrary fields ({field}) for format sorting is '
5246 'deprecated and may be removed in a future version')
5247 self
.settings
[field
] = {}
5248 propObj
= self
.settings
[field
]
5249 if key
not in propObj
:
5250 type = propObj
.get('type')
5252 default
= 'preference' if type == 'extractor' else (field
,) if type in ('combined', 'multiple') else field
5253 elif key
== 'convert':
5254 default
= 'order' if type == 'ordered' else 'float_string' if field
else 'ignore'
5256 default
= {'type': 'field', 'visible': True, 'order': [], 'not_in_list': (None,)}
.get(key
, None)
5257 propObj
[key
] = default
5260 def _resolve_field_value(self
, field
, value
, convertNone
=False):
5265 value
= value
.lower()
5266 conversion
= self
._get
_field
_setting
(field
, 'convert')
5267 if conversion
== 'ignore':
5269 if conversion
== 'string':
5271 elif conversion
== 'float_none':
5272 return float_or_none(value
)
5273 elif conversion
== 'bytes':
5274 return parse_bytes(value
)
5275 elif conversion
== 'order':
5276 order_list
= (self
._use
_free
_order
and self
._get
_field
_setting
(field
, 'order_free')) or self
._get
_field
_setting
(field
, 'order')
5277 use_regex
= self
._get
_field
_setting
(field
, 'regex')
5278 list_length
= len(order_list
)
5279 empty_pos
= order_list
.index('') if '' in order_list
else list_length
+ 1
5280 if use_regex
and value
is not None:
5281 for i
, regex
in enumerate(order_list
):
5282 if regex
and re
.match(regex
, value
):
5283 return list_length
- i
5284 return list_length
- empty_pos
# not in list
5285 else: # not regex or value = None
5286 return list_length
- (order_list
.index(value
) if value
in order_list
else empty_pos
)
5288 if value
.isnumeric():
5291 self
.settings
[field
]['convert'] = 'string'
5294 def evaluate_params(self
, params
, sort_extractor
):
5295 self
._use
_free
_order
= params
.get('prefer_free_formats', False)
5296 self
._sort
_user
= params
.get('format_sort', [])
5297 self
._sort
_extractor
= sort_extractor
5299 def add_item(field
, reverse
, closest
, limit_text
):
5300 field
= field
.lower()
5301 if field
in self
._order
:
5303 self
._order
.append(field
)
5304 limit
= self
._resolve
_field
_value
(field
, limit_text
)
5307 'closest': False if limit
is None else closest
,
5308 'limit_text': limit_text
,
5310 if field
in self
.settings
:
5311 self
.settings
[field
].update(data
)
5313 self
.settings
[field
] = data
5316 tuple(field
for field
in self
.default
if self
._get
_field
_setting
(field
, 'forced'))
5317 + (tuple() if params
.get('format_sort_force', False)
5318 else tuple(field
for field
in self
.default
if self
._get
_field
_setting
(field
, 'priority')))
5319 + tuple(self
._sort
_user
) + tuple(sort_extractor
) + self
.default
)
5321 for item
in sort_list
:
5322 match
= re
.match(self
.regex
, item
)
5324 raise ExtractorError('Invalid format sort string "%s" given by extractor' % item
)
5325 field
= match
.group('field')
5328 if self
._get
_field
_setting
(field
, 'type') == 'alias':
5329 alias
, field
= field
, self
._get
_field
_setting
(field
, 'field')
5330 if self
._get
_field
_setting
(alias
, 'deprecated'):
5331 self
.ydl
.deprecated_feature(f
'Format sorting alias {alias} is deprecated and may '
5332 f
'be removed in a future version. Please use {field} instead')
5333 reverse
= match
.group('reverse') is not None
5334 closest
= match
.group('separator') == '~'
5335 limit_text
= match
.group('limit')
5337 has_limit
= limit_text
is not None
5338 has_multiple_fields
= self
._get
_field
_setting
(field
, 'type') == 'combined'
5339 has_multiple_limits
= has_limit
and has_multiple_fields
and not self
._get
_field
_setting
(field
, 'same_limit')
5341 fields
= self
._get
_field
_setting
(field
, 'field') if has_multiple_fields
else (field
,)
5342 limits
= limit_text
.split(':') if has_multiple_limits
else (limit_text
,) if has_limit
else tuple()
5343 limit_count
= len(limits
)
5344 for (i
, f
) in enumerate(fields
):
5345 add_item(f
, reverse
, closest
,
5346 limits
[i
] if i
< limit_count
5347 else limits
[0] if has_limit
and not has_multiple_limits
5350 def print_verbose_info(self
, write_debug
):
5352 write_debug('Sort order given by user: %s' % ', '.join(self
._sort
_user
))
5353 if self
._sort
_extractor
:
5354 write_debug('Sort order given by extractor: %s' % ', '.join(self
._sort
_extractor
))
5355 write_debug('Formats sorted by: %s' % ', '.join(['%s%s%s' % (
5356 '+' if self
._get
_field
_setting
(field
, 'reverse') else '', field
,
5357 '%s%s(%s)' % ('~' if self
._get
_field
_setting
(field
, 'closest') else ':',
5358 self
._get
_field
_setting
(field
, 'limit_text'),
5359 self
._get
_field
_setting
(field
, 'limit'))
5360 if self
._get
_field
_setting
(field
, 'limit_text') is not None else '')
5361 for field
in self
._order
if self
._get
_field
_setting
(field
, 'visible')]))
5363 def _calculate_field_preference_from_value(self
, format
, field
, type, value
):
5364 reverse
= self
._get
_field
_setting
(field
, 'reverse')
5365 closest
= self
._get
_field
_setting
(field
, 'closest')
5366 limit
= self
._get
_field
_setting
(field
, 'limit')
5368 if type == 'extractor':
5369 maximum
= self
._get
_field
_setting
(field
, 'max')
5370 if value
is None or (maximum
is not None and value
>= maximum
):
5372 elif type == 'boolean':
5373 in_list
= self
._get
_field
_setting
(field
, 'in_list')
5374 not_in_list
= self
._get
_field
_setting
(field
, 'not_in_list')
5375 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
5376 elif type == 'ordered':
5377 value
= self
._resolve
_field
_value
(field
, value
, True)
5379 # try to convert to number
5380 val_num
= float_or_none(value
, default
=self
._get
_field
_setting
(field
, 'default'))
5381 is_num
= self
._get
_field
_setting
(field
, 'convert') != 'string' and val_num
is not None
5385 return ((-10, 0) if value
is None
5386 else (1, value
, 0) if not is_num
# if a field has mixed strings and numbers, strings are sorted higher
5387 else (0, -abs(value
- limit
), value
- limit
if reverse
else limit
- value
) if closest
5388 else (0, value
, 0) if not reverse
and (limit
is None or value
<= limit
)
5389 else (0, -value
, 0) if limit
is None or (reverse
and value
== limit
) or value
> limit
5390 else (-1, value
, 0))
5392 def _calculate_field_preference(self
, format
, field
):
5393 type = self
._get
_field
_setting
(field
, 'type') # extractor, boolean, ordered, field, multiple
5394 get_value
= lambda f
: format
.get(self
._get
_field
_setting
(f
, 'field'))
5395 if type == 'multiple':
5396 type = 'field' # Only 'field' is allowed in multiple for now
5397 actual_fields
= self
._get
_field
_setting
(field
, 'field')
5399 value
= self
._get
_field
_setting
(field
, 'function')(get_value(f
) for f
in actual_fields
)
5401 value
= get_value(field
)
5402 return self
._calculate
_field
_preference
_from
_value
(format
, field
, type, value
)
5404 def calculate_preference(self
, format
):
5405 # Determine missing protocol
5406 if not format
.get('protocol'):
5407 format
['protocol'] = determine_protocol(format
)
5409 # Determine missing ext
5410 if not format
.get('ext') and 'url' in format
:
5411 format
['ext'] = determine_ext(format
['url'])
5412 if format
.get('vcodec') == 'none':
5413 format
['audio_ext'] = format
['ext'] if format
.get('acodec') != 'none' else 'none'
5414 format
['video_ext'] = 'none'
5416 format
['video_ext'] = format
['ext']
5417 format
['audio_ext'] = 'none'
5418 # if format.get('preference') is None and format.get('ext') in ('f4f', 'f4m'): # Not supported?
5419 # format['preference'] = -1000
5421 if format
.get('preference') is None and format
.get('ext') == 'flv' and re
.match('[hx]265|he?vc?', format
.get('vcodec') or ''):
5422 # HEVC-over-FLV is out-of-spec by FLV's original spec
5423 # ref. https://trac.ffmpeg.org/ticket/6389
5424 # ref. https://github.com/yt-dlp/yt-dlp/pull/5821
5425 format
['preference'] = -100
5427 # Determine missing bitrates
5428 if format
.get('vcodec') == 'none':
5430 if format
.get('acodec') == 'none':
5432 if not format
.get('vbr') and format
.get('vcodec') != 'none':
5433 format
['vbr'] = try_call(lambda: format
['tbr'] - format
['abr']) or None
5434 if not format
.get('abr') and format
.get('acodec') != 'none':
5435 format
['abr'] = try_call(lambda: format
['tbr'] - format
['vbr']) or None
5436 if not format
.get('tbr'):
5437 format
['tbr'] = try_call(lambda: format
['vbr'] + format
['abr']) or None
5439 return tuple(self
._calculate
_field
_preference
(format
, field
) for field
in self
._order
)
5442 def filesize_from_tbr(tbr
, duration
):
5444 @param tbr: Total bitrate in kbps (1000 bits/sec)
5445 @param duration: Duration in seconds
5446 @returns Filesize in bytes
5448 if tbr
is None or duration
is None:
5450 return int(duration
* tbr
* (1000 / 8))
5455 def __init__(self
, ydl
=None):
5458 def debug(self
, message
):
5460 self
._ydl
.write_debug(message
)
5462 def info(self
, message
):
5464 self
._ydl
.to_screen(message
)
5466 def warning(self
, message
, *, once
=False):
5468 self
._ydl
.report_warning(message
, once
)
5470 def error(self
, message
, *, is_error
=True):
5472 self
._ydl
.report_error(message
, is_error
=is_error
)
5474 def stdout(self
, message
):
5476 self
._ydl
.to_stdout(message
)
5478 def stderr(self
, message
):
5480 self
._ydl
.to_stderr(message
)