45 import xml
.etree
.ElementTree
47 from . import traversal
49 from ..compat
import functools
# isort: split
50 from ..compat
import (
51 compat_etree_fromstring
,
53 compat_HTMLParseError
,
57 from ..dependencies
import websockets
, xattr
59 __name__
= __name__
.rsplit('.', 1)[0] # Pretend to be the parent module
61 # This is not clearly defined otherwise
62 compiled_regex_type
= type(re
.compile(''))
66 'Safari': 'Mozilla/5.0 (X11; Linux x86_64; rv:10.0) AppleWebKit/533.20.25 (KHTML, like Gecko) Version/5.0.4 Safari/533.20.27',
78 ENGLISH_MONTH_NAMES
= [
79 'January', 'February', 'March', 'April', 'May', 'June',
80 'July', 'August', 'September', 'October', 'November', 'December']
83 'en': ENGLISH_MONTH_NAMES
,
85 'janvier', 'février', 'mars', 'avril', 'mai', 'juin',
86 'juillet', 'août', 'septembre', 'octobre', 'novembre', 'décembre'],
87 # these follow the genitive grammatical case (dopełniacz)
88 # some websites might be using nominative, which will require another month list
89 # https://en.wikibooks.org/wiki/Polish/Noun_cases
90 'pl': ['stycznia', 'lutego', 'marca', 'kwietnia', 'maja', 'czerwca',
91 'lipca', 'sierpnia', 'września', 'października', 'listopada', 'grudnia'],
94 # From https://github.com/python/cpython/blob/3.11/Lib/email/_parseaddr.py#L36-L42
96 'UT': 0, 'UTC': 0, 'GMT': 0, 'Z': 0,
97 'AST': -4, 'ADT': -3, # Atlantic (used in Canada)
98 'EST': -5, 'EDT': -4, # Eastern
99 'CST': -6, 'CDT': -5, # Central
100 'MST': -7, 'MDT': -6, # Mountain
101 'PST': -8, 'PDT': -7 # Pacific
104 # needed for sanitizing filenames in restricted mode
105 ACCENT_CHARS
= dict(zip('ÂÃÄÀÁÅÆÇÈÉÊËÌÍÎÏÐÑÒÓÔÕÖŐØŒÙÚÛÜŰÝÞßàáâãäåæçèéêëìíîïðñòóôõöőøœùúûüűýþÿ',
106 itertools
.chain('AAAAAA', ['AE'], 'CEEEEIIIIDNOOOOOOO', ['OE'], 'UUUUUY', ['TH', 'ss'],
107 'aaaaaa', ['ae'], 'ceeeeiiiionooooooo', ['oe'], 'uuuuuy', ['th'], 'y')))
137 '%Y-%m-%d %H:%M:%S.%f',
138 '%Y-%m-%d %H:%M:%S:%f',
141 '%Y-%m-%dT%H:%M:%SZ',
142 '%Y-%m-%dT%H:%M:%S.%fZ',
143 '%Y-%m-%dT%H:%M:%S.%f0Z',
145 '%Y-%m-%dT%H:%M:%S.%f',
148 '%b %d %Y at %H:%M:%S',
150 '%B %d %Y at %H:%M:%S',
154 DATE_FORMATS_DAY_FIRST
= list(DATE_FORMATS
)
155 DATE_FORMATS_DAY_FIRST
.extend([
166 DATE_FORMATS_MONTH_FIRST
= list(DATE_FORMATS
)
167 DATE_FORMATS_MONTH_FIRST
.extend([
175 PACKED_CODES_RE
= r
"}\('(.+)',(\d+),(\d+),'([^']+)'\.split\('\|'\)"
176 JSON_LD_RE
= r
'(?is)<script[^>]+type=(["\']?
)application
/ld\
+json\
1[^
>]*>\s
*(?P
<json_ld
>{.+?}|\
[.+?\
])\s
*</script
>'
178 NUMBER_RE = r'\d
+(?
:\
.\d
+)?
'
182 def preferredencoding():
183 """Get preferred encoding.
185 Returns the best encoding scheme for the system, based on
186 locale.getpreferredencoding() and some further tweaks.
189 pref = locale.getpreferredencoding()
197 def write_json_file(obj, fn):
198 """ Encode obj as JSON and write it to fn, atomically if possible """
200 tf = tempfile.NamedTemporaryFile(
201 prefix=f'{os.path.basename(fn)}
.', dir=os.path.dirname(fn),
202 suffix='.tmp
', delete=False, mode='w
', encoding='utf
-8')
206 json.dump(obj, tf, ensure_ascii=False)
207 if sys.platform == 'win32
':
208 # Need to remove existing file on Windows, else os.rename raises
209 # WindowsError or FileExistsError.
210 with contextlib.suppress(OSError):
212 with contextlib.suppress(OSError):
215 os.chmod(tf.name, 0o666 & ~mask)
216 os.rename(tf.name, fn)
218 with contextlib.suppress(OSError):
223 def find_xpath_attr(node, xpath, key, val=None):
224 """ Find the xpath xpath[@key=val] """
225 assert re.match(r'^
[a
-zA
-Z_
-]+$
', key)
226 expr = xpath + ('[@%s]' % key if val is None else f"[@{key}='{val}
']")
227 return node.find(expr)
229 # On python2.6 the xml.etree.ElementTree.Element methods don't support
230 # the namespace parameter
233 def xpath_with_ns(path
, ns_map
):
234 components
= [c
.split(':') for c
in path
.split('/')]
238 replaced
.append(c
[0])
241 replaced
.append('{%s}%s' % (ns_map
[ns
], tag
))
242 return '/'.join(replaced
)
245 def xpath_element(node
, xpath
, name
=None, fatal
=False, default
=NO_DEFAULT
):
246 def _find_xpath(xpath
):
247 return node
.find(xpath
)
249 if isinstance(xpath
, str):
250 n
= _find_xpath(xpath
)
258 if default
is not NO_DEFAULT
:
261 name
= xpath
if name
is None else name
262 raise ExtractorError('Could not find XML element %s' % name
)
268 def xpath_text(node
, xpath
, name
=None, fatal
=False, default
=NO_DEFAULT
):
269 n
= xpath_element(node
, xpath
, name
, fatal
=fatal
, default
=default
)
270 if n
is None or n
== default
:
273 if default
is not NO_DEFAULT
:
276 name
= xpath
if name
is None else name
277 raise ExtractorError('Could not find XML element\'s text %s' % name
)
283 def xpath_attr(node
, xpath
, key
, name
=None, fatal
=False, default
=NO_DEFAULT
):
284 n
= find_xpath_attr(node
, xpath
, key
)
286 if default
is not NO_DEFAULT
:
289 name
= f
'{xpath}[@{key}]' if name
is None else name
290 raise ExtractorError('Could not find XML attribute %s' % name
)
296 def get_element_by_id(id, html
, **kwargs
):
297 """Return the content of the tag with the specified ID in the passed HTML document"""
298 return get_element_by_attribute('id', id, html
, **kwargs
)
301 def get_element_html_by_id(id, html
, **kwargs
):
302 """Return the html of the tag with the specified ID in the passed HTML document"""
303 return get_element_html_by_attribute('id', id, html
, **kwargs
)
306 def get_element_by_class(class_name
, html
):
307 """Return the content of the first tag with the specified class in the passed HTML document"""
308 retval
= get_elements_by_class(class_name
, html
)
309 return retval
[0] if retval
else None
312 def get_element_html_by_class(class_name
, html
):
313 """Return the html of the first tag with the specified class in the passed HTML document"""
314 retval
= get_elements_html_by_class(class_name
, html
)
315 return retval
[0] if retval
else None
318 def get_element_by_attribute(attribute
, value
, html
, **kwargs
):
319 retval
= get_elements_by_attribute(attribute
, value
, html
, **kwargs
)
320 return retval
[0] if retval
else None
323 def get_element_html_by_attribute(attribute
, value
, html
, **kargs
):
324 retval
= get_elements_html_by_attribute(attribute
, value
, html
, **kargs
)
325 return retval
[0] if retval
else None
328 def get_elements_by_class(class_name
, html
, **kargs
):
329 """Return the content of all tags with the specified class in the passed HTML document as a list"""
330 return get_elements_by_attribute(
331 'class', r
'[^\'"]*(?<=[\'"\s
])%s(?
=[\'"\s])[^\'"]*' % re.escape(class_name),
332 html, escape_value=False)
335 def get_elements_html_by_class(class_name, html):
336 """Return the html of all tags with the specified class in the passed HTML document as a list"""
337 return get_elements_html_by_attribute(
338 'class', r'[^
\'"]*(?<=[\'"\s
])%s(?
=[\'"\s])[^\'"]*' % re.escape(class_name),
339 html, escape_value=False)
342 def get_elements_by_attribute(*args, **kwargs):
343 """Return the content of the tag with the specified attribute in the passed HTML document"""
344 return [content for content, _ in get_elements_text_and_html_by_attribute(*args, **kwargs)]
347 def get_elements_html_by_attribute(*args, **kwargs):
348 """Return the html of the tag with the specified attribute in the passed HTML document"""
349 return [whole for _, whole in get_elements_text_and_html_by_attribute(*args, **kwargs)]
352 def get_elements_text_and_html_by_attribute(attribute, value, html, *, tag=r'[\w
:.-]+', escape_value=True):
354 Return the text (content) and the html (whole) of the tag with the specified
355 attribute in the passed HTML document
360 quote = '' if re.match(r'''[\s"'`
=<>]''', value) else '?'
362 value = re.escape(value) if escape_value else value
364 partial_element_re = rf'''(?x
)
366 (?
:\
s(?
:[^
>"']|"[^
"]*"|
'[^']*')*)?
367 \s{re.escape(attribute)}\s*=\s*(?P<_q>['"]{quote})(?-x:{value})(?P=_q)
370 for m in re.finditer(partial_element_re, html):
371 content, whole = get_element_text_and_html_by_tag(m.group('tag'), html[m.start():])
374 unescapeHTML(re.sub(r'^(?P<q>["\'])(?P
<content
>.*)(?P
=q
)$
', r'\g
<content
>', content, flags=re.DOTALL)),
379 class HTMLBreakOnClosingTagParser(html.parser.HTMLParser):
381 HTML parser which raises HTMLBreakOnClosingTagException upon reaching the
382 closing tag for the first opening tag it has encountered, and can be used
386 class HTMLBreakOnClosingTagException(Exception):
390 self.tagstack = collections.deque()
391 html.parser.HTMLParser.__init__(self)
396 def __exit__(self, *_):
400 # handle_endtag does not return upon raising HTMLBreakOnClosingTagException,
401 # so data remains buffered; we no longer have any interest in it, thus
402 # override this method to discard it
405 def handle_starttag(self, tag, _):
406 self.tagstack.append(tag)
408 def handle_endtag(self, tag):
409 if not self.tagstack:
410 raise compat_HTMLParseError('no tags
in the stack
')
412 inner_tag = self.tagstack.pop()
416 raise compat_HTMLParseError(f'matching opening tag
for closing {tag} tag
not found
')
417 if not self.tagstack:
418 raise self.HTMLBreakOnClosingTagException()
421 # XXX: This should be far less strict
422 def get_element_text_and_html_by_tag(tag, html):
424 For the first element with the specified tag in the passed HTML document
425 return its' content (text
) and the whole
element (html
)
427 def find_or_raise(haystack, needle, exc):
429 return haystack.index(needle)
432 closing_tag = f'</{tag}>'
433 whole_start = find_or_raise(
434 html, f'<{tag}', compat_HTMLParseError(f'opening {tag} tag not found'))
435 content_start = find_or_raise(
436 html[whole_start:], '>', compat_HTMLParseError(f'malformed opening {tag} tag'))
437 content_start += whole_start + 1
438 with HTMLBreakOnClosingTagParser() as parser:
439 parser.feed(html[whole_start:content_start])
440 if not parser.tagstack or parser.tagstack[0] != tag:
441 raise compat_HTMLParseError(f'parser did not match opening {tag} tag')
442 offset = content_start
443 while offset < len(html):
444 next_closing_tag_start = find_or_raise(
445 html[offset:], closing_tag,
446 compat_HTMLParseError(f'closing {tag} tag not found'))
447 next_closing_tag_end = next_closing_tag_start + len(closing_tag)
449 parser.feed(html[offset:offset + next_closing_tag_end])
450 offset += next_closing_tag_end
451 except HTMLBreakOnClosingTagParser.HTMLBreakOnClosingTagException:
452 return html[content_start:offset + next_closing_tag_start], \
453 html[whole_start:offset + next_closing_tag_end]
454 raise compat_HTMLParseError('unexpected end of html')
457 class HTMLAttributeParser(html.parser.HTMLParser):
458 """Trivial HTML parser to gather the attributes
for a single element
"""
462 html.parser.HTMLParser.__init__(self)
464 def handle_starttag(self, tag, attrs):
465 self.attrs = dict(attrs)
466 raise compat_HTMLParseError('done')
469 class HTMLListAttrsParser(html.parser.HTMLParser):
470 """HTML parser to gather the attributes
for the elements of a
list"""
473 html.parser.HTMLParser.__init__(self)
477 def handle_starttag(self, tag, attrs):
478 if tag == 'li' and self._level == 0:
479 self.items.append(dict(attrs))
482 def handle_endtag(self, tag):
486 def extract_attributes(html_element):
487 """Given a string
for an HTML element such
as
489 a
="foo" B
="bar" c
="&98;az" d
=boz
490 empty
= noval entity
="&"
493 Decode
and return a dictionary of attributes
.
495 'a': 'foo', 'b': 'bar', c
: 'baz', d
: 'boz',
496 'empty': '', 'noval': None, 'entity': '&',
497 'sq': '"', 'dq': '\''
500 parser = HTMLAttributeParser()
501 with contextlib.suppress(compat_HTMLParseError):
502 parser.feed(html_element)
507 def parse_list(webpage):
508 """Given a string
for an series of HTML
<li
> elements
,
509 return a dictionary of their attributes
"""
510 parser = HTMLListAttrsParser()
516 def clean_html(html):
517 """Clean an HTML snippet into a readable string
"""
519 if html is None: # Convenience for sanitizing descriptions etc.
522 html = re.sub(r'\s+', ' ', html)
523 html = re.sub(r'(?u)\s?<\s?br\s?/?\s?>\s?', '\n', html)
524 html = re.sub(r'(?u)<\s?/\s?p\s?>\s?<\s?p[^>]*>', '\n', html)
526 html = re.sub('<.*?>', '', html)
527 # Replace html entities
528 html = unescapeHTML(html)
532 class LenientJSONDecoder(json.JSONDecoder):
534 def __init__(self, *args, transform_source=None, ignore_extra=False, close_objects=0, **kwargs):
535 self.transform_source, self.ignore_extra = transform_source, ignore_extra
536 self._close_attempts = 2 * close_objects
537 super().__init__(*args, **kwargs)
540 def _close_object(err):
541 doc = err.doc[:err.pos]
542 # We need to add comma first to get the correct error message
543 if err.msg.startswith('Expecting \',\''):
545 elif not doc.endswith(','):
548 if err.msg.startswith('Expecting property name'):
549 return doc[:-1] + '}'
550 elif err.msg.startswith('Expecting value'):
551 return doc[:-1] + ']'
554 if self.transform_source:
555 s = self.transform_source(s)
556 for attempt in range(self._close_attempts + 1):
558 if self.ignore_extra:
559 return self.raw_decode(s.lstrip())[0]
560 return super().decode(s)
561 except json.JSONDecodeError as e:
564 elif attempt < self._close_attempts:
565 s = self._close_object(e)
568 raise type(e)(f'{e.msg} in {s[e.pos-10:e.pos+10]!r}', s, e.pos)
569 assert False, 'Too many attempts to decode JSON'
572 def sanitize_open(filename, open_mode):
573 """Try to
open the given filename
, and slightly tweak it
if this fails
.
575 Attempts to
open the given filename
. If this fails
, it tries to change
576 the filename slightly
, step by step
, until it
's either able to open it
577 or it fails and raises a final exception, like the standard open()
580 It returns the tuple (stream, definitive_file_name).
583 if sys.platform == 'win32
':
586 # stdout may be any IO stream, e.g. when using contextlib.redirect_stdout
587 with contextlib.suppress(io.UnsupportedOperation):
588 msvcrt.setmode(sys.stdout.fileno(), os.O_BINARY)
589 return (sys.stdout.buffer if hasattr(sys.stdout, 'buffer') else sys.stdout, filename)
591 for attempt in range(2):
594 if sys.platform == 'win32
':
595 # FIXME: An exclusive lock also locks the file from being read.
596 # Since windows locks are mandatory, don't lock the
file on
windows (for now
).
597 # Ref: https://github.com/yt-dlp/yt-dlp/issues/3124
598 raise LockingUnsupportedError()
599 stream
= locked_file(filename
, open_mode
, block
=False).__enter
__()
601 stream
= open(filename
, open_mode
)
602 return stream
, filename
603 except OSError as err
:
604 if attempt
or err
.errno
in (errno
.EACCES
,):
606 old_filename
, filename
= filename
, sanitize_path(filename
)
607 if old_filename
== filename
:
611 def timeconvert(timestr
):
612 """Convert RFC 2822 defined time string into system timestamp"""
614 timetuple
= email
.utils
.parsedate_tz(timestr
)
615 if timetuple
is not None:
616 timestamp
= email
.utils
.mktime_tz(timetuple
)
620 def sanitize_filename(s
, restricted
=False, is_id
=NO_DEFAULT
):
621 """Sanitizes a string so it could be used as part of a filename.
622 @param restricted Use a stricter subset of allowed characters
623 @param is_id Whether this is an ID that should be kept unchanged if possible.
624 If unset, yt-dlp's new sanitization rules are in effect
629 def replace_insane(char
):
630 if restricted
and char
in ACCENT_CHARS
:
631 return ACCENT_CHARS
[char
]
632 elif not restricted
and char
== '\n':
634 elif is_id
is NO_DEFAULT
and not restricted
and char
in '"*:<>?|/\\':
635 # Replace with their full-width unicode counterparts
636 return {'/': '\u29F8', '\\': '\u29f9'}
.get(char
, chr(ord(char
) + 0xfee0))
637 elif char
== '?' or ord(char
) < 32 or ord(char
) == 127:
640 return '' if restricted
else '\''
642 return '\0_\0-' if restricted
else '\0 \0-'
643 elif char
in '\\/|*<>':
645 if restricted
and (char
in '!&\'()[]{}$;`^,#' or char
.isspace() or ord(char
) > 127):
649 # Replace look-alike Unicode glyphs
650 if restricted
and (is_id
is NO_DEFAULT
or not is_id
):
651 s
= unicodedata
.normalize('NFKC', s
)
652 s
= re
.sub(r
'[0-9]+(?::[0-9]+)+', lambda m
: m
.group(0).replace(':', '_'), s
) # Handle timestamps
653 result
= ''.join(map(replace_insane
, s
))
654 if is_id
is NO_DEFAULT
:
655 result
= re
.sub(r
'(\0.)(?:(?=\1)..)+', r
'\1', result
) # Remove repeated substitute chars
656 STRIP_RE
= r
'(?:\0.|[ _-])*'
657 result
= re
.sub(f
'^\0.{STRIP_RE}|{STRIP_RE}\0.$', '', result
) # Remove substitute chars from start/end
658 result
= result
.replace('\0', '') or '_'
661 while '__' in result
:
662 result
= result
.replace('__', '_')
663 result
= result
.strip('_')
664 # Common case of "Foreign band name - English song title"
665 if restricted
and result
.startswith('-_'):
667 if result
.startswith('-'):
668 result
= '_' + result
[len('-'):]
669 result
= result
.lstrip('.')
675 def sanitize_path(s
, force
=False):
676 """Sanitizes and normalizes path on Windows"""
677 if sys
.platform
== 'win32':
679 drive_or_unc
, _
= os
.path
.splitdrive(s
)
685 norm_path
= os
.path
.normpath(remove_start(s
, drive_or_unc
)).split(os
.path
.sep
)
689 path_part
if path_part
in ['.', '..'] else re
.sub(r
'(?:[/<>:"\|\\?\*]|[\s.]$)', '#', path_part
)
690 for path_part
in norm_path
]
692 sanitized_path
.insert(0, drive_or_unc
+ os
.path
.sep
)
693 elif force
and s
and s
[0] == os
.path
.sep
:
694 sanitized_path
.insert(0, os
.path
.sep
)
695 return os
.path
.join(*sanitized_path
)
698 def sanitize_url(url
, *, scheme
='http'):
699 # Prepend protocol-less URLs with `http:` scheme in order to mitigate
700 # the number of unwanted failures due to missing protocol
703 elif url
.startswith('//'):
704 return f
'{scheme}:{url}'
705 # Fix some common typos seen so far
707 # https://github.com/ytdl-org/youtube-dl/issues/15649
708 (r
'^httpss://', r
'https://'),
709 # https://bx1.be/lives/direct-tv/
710 (r
'^rmtp([es]?)://', r
'rtmp\1://'),
712 for mistake
, fixup
in COMMON_TYPOS
:
713 if re
.match(mistake
, url
):
714 return re
.sub(mistake
, fixup
, url
)
718 def extract_basic_auth(url
):
719 parts
= urllib
.parse
.urlsplit(url
)
720 if parts
.username
is None:
722 url
= urllib
.parse
.urlunsplit(parts
._replace
(netloc
=(
723 parts
.hostname
if parts
.port
is None
724 else '%s:%d' % (parts
.hostname
, parts
.port
))))
725 auth_payload
= base64
.b64encode(
726 ('%s:%s' % (parts
.username
, parts
.password
or '')).encode())
727 return url
, f
'Basic {auth_payload.decode()}'
730 def sanitized_Request(url
, *args
, **kwargs
):
731 url
, auth_header
= extract_basic_auth(escape_url(sanitize_url(url
)))
732 if auth_header
is not None:
733 headers
= args
[1] if len(args
) >= 2 else kwargs
.setdefault('headers', {})
734 headers
['Authorization'] = auth_header
735 return urllib
.request
.Request(url
, *args
, **kwargs
)
739 """Expand shell variables and ~"""
740 return os
.path
.expandvars(compat_expanduser(s
))
743 def orderedSet(iterable
, *, lazy
=False):
744 """Remove all duplicates from the input iterable"""
746 seen
= [] # Do not use set since the items can be unhashable
752 return _iter() if lazy
else list(_iter())
755 def _htmlentity_transform(entity_with_semicolon
):
756 """Transforms an HTML entity to a character."""
757 entity
= entity_with_semicolon
[:-1]
759 # Known non-numeric HTML entity
760 if entity
in html
.entities
.name2codepoint
:
761 return chr(html
.entities
.name2codepoint
[entity
])
763 # TODO: HTML5 allows entities without a semicolon.
764 # E.g. 'Éric' should be decoded as 'Éric'.
765 if entity_with_semicolon
in html
.entities
.html5
:
766 return html
.entities
.html5
[entity_with_semicolon
]
768 mobj
= re
.match(r
'#(x[0-9a-fA-F]+|[0-9]+)', entity
)
770 numstr
= mobj
.group(1)
771 if numstr
.startswith('x'):
773 numstr
= '0%s' % numstr
776 # See https://github.com/ytdl-org/youtube-dl/issues/7518
777 with contextlib
.suppress(ValueError):
778 return chr(int(numstr
, base
))
780 # Unknown entity in name, return its literal representation
781 return '&%s;' % entity
787 assert isinstance(s
, str)
790 r
'&([^&;]+;)', lambda m
: _htmlentity_transform(m
.group(1)), s
)
793 def escapeHTML(text
):
796 .replace('&', '&')
797 .replace('<', '<')
798 .replace('>', '>')
799 .replace('"', '"')
800 .replace("'", ''')
804 class netrc_from_content(netrc
.netrc
):
805 def __init__(self
, content
):
806 self
.hosts
, self
.macros
= {}, {}
807 with io
.StringIO(content
) as stream
:
808 self
._parse
('-', stream
, False)
811 class Popen(subprocess
.Popen
):
812 if sys
.platform
== 'win32':
813 _startupinfo
= subprocess
.STARTUPINFO()
814 _startupinfo
.dwFlags |
= subprocess
.STARTF_USESHOWWINDOW
819 def _fix_pyinstaller_ld_path(env
):
820 """Restore LD_LIBRARY_PATH when using PyInstaller
821 Ref: https://github.com/pyinstaller/pyinstaller/blob/develop/doc/runtime-information.rst#ld_library_path--libpath-considerations
822 https://github.com/yt-dlp/yt-dlp/issues/4573
824 if not hasattr(sys
, '_MEIPASS'):
828 orig
= env
.get(f
'{key}_ORIG')
834 _fix('LD_LIBRARY_PATH') # Linux
835 _fix('DYLD_LIBRARY_PATH') # macOS
837 def __init__(self
, *args
, env
=None, text
=False, **kwargs
):
839 env
= os
.environ
.copy()
840 self
._fix
_pyinstaller
_ld
_path
(env
)
842 self
.__text
_mode
= kwargs
.get('encoding') or kwargs
.get('errors') or text
or kwargs
.get('universal_newlines')
844 kwargs
['universal_newlines'] = True # For 3.6 compatibility
845 kwargs
.setdefault('encoding', 'utf-8')
846 kwargs
.setdefault('errors', 'replace')
847 super().__init
__(*args
, env
=env
, **kwargs
, startupinfo
=self
._startupinfo
)
849 def communicate_or_kill(self
, *args
, **kwargs
):
851 return self
.communicate(*args
, **kwargs
)
852 except BaseException
: # Including KeyboardInterrupt
853 self
.kill(timeout
=None)
856 def kill(self
, *, timeout
=0):
859 self
.wait(timeout
=timeout
)
862 def run(cls
, *args
, timeout
=None, **kwargs
):
863 with cls(*args
, **kwargs
) as proc
:
864 default
= '' if proc
.__text
_mode
else b
''
865 stdout
, stderr
= proc
.communicate_or_kill(timeout
=timeout
)
866 return stdout
or default
, stderr
or default
, proc
.returncode
869 def encodeArgument(s
):
870 # Legacy code that uses byte strings
871 # Uncomment the following line after fixing all post processors
872 # assert isinstance(s, str), 'Internal error: %r should be of type %r, is %r' % (s, str, type(s))
873 return s
if isinstance(s
, str) else s
.decode('ascii')
876 _timetuple
= collections
.namedtuple('Time', ('hours', 'minutes', 'seconds', 'milliseconds'))
879 def timetuple_from_msec(msec
):
880 secs
, msec
= divmod(msec
, 1000)
881 mins
, secs
= divmod(secs
, 60)
882 hrs
, mins
= divmod(mins
, 60)
883 return _timetuple(hrs
, mins
, secs
, msec
)
886 def formatSeconds(secs
, delim
=':', msec
=False):
887 time
= timetuple_from_msec(secs
* 1000)
889 ret
= '%d%s%02d%s%02d' % (time
.hours
, delim
, time
.minutes
, delim
, time
.seconds
)
891 ret
= '%d%s%02d' % (time
.minutes
, delim
, time
.seconds
)
893 ret
= '%d' % time
.seconds
894 return '%s.%03d' % (ret
, time
.milliseconds
) if msec
else ret
897 def make_HTTPS_handler(params
, **kwargs
):
898 from ._deprecated
import YoutubeDLHTTPSHandler
899 from ..networking
._helper
import make_ssl_context
900 return YoutubeDLHTTPSHandler(params
, context
=make_ssl_context(
901 verify
=not params
.get('nocheckcertificate'),
902 client_certificate
=params
.get('client_certificate'),
903 client_certificate_key
=params
.get('client_certificate_key'),
904 client_certificate_password
=params
.get('client_certificate_password'),
905 legacy_support
=params
.get('legacyserverconnect'),
906 use_certifi
='no-certifi' not in params
.get('compat_opts', []),
910 def bug_reports_message(before
=';'):
911 from ..update
import REPOSITORY
913 msg
= (f
'please report this issue on https://github.com/{REPOSITORY}/issues?q= , '
914 'filling out the appropriate issue template. Confirm you are on the latest version using yt-dlp -U')
916 before
= before
.rstrip()
917 if not before
or before
.endswith(('.', '!', '?')):
918 msg
= msg
[0].title() + msg
[1:]
920 return (before
+ ' ' if before
else '') + msg
923 class YoutubeDLError(Exception):
924 """Base exception for YoutubeDL errors."""
927 def __init__(self
, msg
=None):
930 elif self
.msg
is None:
931 self
.msg
= type(self
).__name
__
932 super().__init
__(self
.msg
)
935 class ExtractorError(YoutubeDLError
):
936 """Error during info extraction."""
938 def __init__(self
, msg
, tb
=None, expected
=False, cause
=None, video_id
=None, ie
=None):
939 """ tb, if given, is the original traceback (so that it can be printed out).
940 If expected is set, this is a normal error message and most likely not a bug in yt-dlp.
942 from ..networking
.exceptions
import network_exceptions
943 if sys
.exc_info()[0] in network_exceptions
:
946 self
.orig_msg
= str(msg
)
948 self
.expected
= expected
950 self
.video_id
= video_id
952 self
.exc_info
= sys
.exc_info() # preserve original exception
953 if isinstance(self
.exc_info
[1], ExtractorError
):
954 self
.exc_info
= self
.exc_info
[1].exc_info
955 super().__init
__(self
.__msg
)
960 format_field(self
.ie
, None, '[%s] '),
961 format_field(self
.video_id
, None, '%s: '),
963 format_field(self
.cause
, None, ' (caused by %r)'),
964 '' if self
.expected
else bug_reports_message()))
966 def format_traceback(self
):
967 return join_nonempty(
968 self
.traceback
and ''.join(traceback
.format_tb(self
.traceback
)),
969 self
.cause
and ''.join(traceback
.format_exception(None, self
.cause
, self
.cause
.__traceback
__)[1:]),
972 def __setattr__(self
, name
, value
):
973 super().__setattr
__(name
, value
)
974 if getattr(self
, 'msg', None) and name
not in ('msg', 'args'):
975 self
.msg
= self
.__msg
or type(self
).__name
__
976 self
.args
= (self
.msg
, ) # Cannot be property
979 class UnsupportedError(ExtractorError
):
980 def __init__(self
, url
):
982 'Unsupported URL: %s' % url
, expected
=True)
986 class RegexNotFoundError(ExtractorError
):
987 """Error when a regex didn't match"""
991 class GeoRestrictedError(ExtractorError
):
992 """Geographic restriction Error exception.
994 This exception may be thrown when a video is not available from your
995 geographic location due to geographic restrictions imposed by a website.
998 def __init__(self
, msg
, countries
=None, **kwargs
):
999 kwargs
['expected'] = True
1000 super().__init
__(msg
, **kwargs
)
1001 self
.countries
= countries
1004 class UserNotLive(ExtractorError
):
1005 """Error when a channel/user is not live"""
1007 def __init__(self
, msg
=None, **kwargs
):
1008 kwargs
['expected'] = True
1009 super().__init
__(msg
or 'The channel is not currently live', **kwargs
)
1012 class DownloadError(YoutubeDLError
):
1013 """Download Error exception.
1015 This exception may be thrown by FileDownloader objects if they are not
1016 configured to continue on errors. They will contain the appropriate
1020 def __init__(self
, msg
, exc_info
=None):
1021 """ exc_info, if given, is the original exception that caused the trouble (as returned by sys.exc_info()). """
1022 super().__init
__(msg
)
1023 self
.exc_info
= exc_info
1026 class EntryNotInPlaylist(YoutubeDLError
):
1027 """Entry not in playlist exception.
1029 This exception will be thrown by YoutubeDL when a requested entry
1030 is not found in the playlist info_dict
1032 msg
= 'Entry not found in info'
1035 class SameFileError(YoutubeDLError
):
1036 """Same File exception.
1038 This exception will be thrown by FileDownloader objects if they detect
1039 multiple files would have to be downloaded to the same file on disk.
1041 msg
= 'Fixed output name but more than one file to download'
1043 def __init__(self
, filename
=None):
1044 if filename
is not None:
1045 self
.msg
+= f
': {filename}'
1046 super().__init
__(self
.msg
)
1049 class PostProcessingError(YoutubeDLError
):
1050 """Post Processing exception.
1052 This exception may be raised by PostProcessor's .run() method to
1053 indicate an error in the postprocessing task.
1057 class DownloadCancelled(YoutubeDLError
):
1058 """ Exception raised when the download queue should be interrupted """
1059 msg
= 'The download was cancelled'
1062 class ExistingVideoReached(DownloadCancelled
):
1063 """ --break-on-existing triggered """
1064 msg
= 'Encountered a video that is already in the archive, stopping due to --break-on-existing'
1067 class RejectedVideoReached(DownloadCancelled
):
1068 """ --break-match-filter triggered """
1069 msg
= 'Encountered a video that did not match filter, stopping due to --break-match-filter'
1072 class MaxDownloadsReached(DownloadCancelled
):
1073 """ --max-downloads limit has been reached. """
1074 msg
= 'Maximum number of downloads reached, stopping due to --max-downloads'
1077 class ReExtractInfo(YoutubeDLError
):
1078 """ Video info needs to be re-extracted. """
1080 def __init__(self
, msg
, expected
=False):
1081 super().__init
__(msg
)
1082 self
.expected
= expected
1085 class ThrottledDownload(ReExtractInfo
):
1086 """ Download speed below --throttled-rate. """
1087 msg
= 'The download speed is below throttle limit'
1090 super().__init
__(self
.msg
, expected
=False)
1093 class UnavailableVideoError(YoutubeDLError
):
1094 """Unavailable Format exception.
1096 This exception will be thrown when a video is requested
1097 in a format that is not available for that video.
1099 msg
= 'Unable to download video'
1101 def __init__(self
, err
=None):
1103 self
.msg
+= f
': {err}'
1104 super().__init
__(self
.msg
)
1107 class ContentTooShortError(YoutubeDLError
):
1108 """Content Too Short exception.
1110 This exception may be raised by FileDownloader objects when a file they
1111 download is too small for what the server announced first, indicating
1112 the connection was probably interrupted.
1115 def __init__(self
, downloaded
, expected
):
1116 super().__init
__(f
'Downloaded {downloaded} bytes, expected {expected} bytes')
1118 self
.downloaded
= downloaded
1119 self
.expected
= expected
1122 class XAttrMetadataError(YoutubeDLError
):
1123 def __init__(self
, code
=None, msg
='Unknown error'):
1124 super().__init
__(msg
)
1128 # Parsing code and msg
1129 if (self
.code
in (errno
.ENOSPC
, errno
.EDQUOT
)
1130 or 'No space left' in self
.msg
or 'Disk quota exceeded' in self
.msg
):
1131 self
.reason
= 'NO_SPACE'
1132 elif self
.code
== errno
.E2BIG
or 'Argument list too long' in self
.msg
:
1133 self
.reason
= 'VALUE_TOO_LONG'
1135 self
.reason
= 'NOT_SUPPORTED'
1138 class XAttrUnavailableError(YoutubeDLError
):
1142 def is_path_like(f
):
1143 return isinstance(f
, (str, bytes, os
.PathLike
))
1146 class YoutubeDLCookieProcessor(urllib
.request
.HTTPCookieProcessor
):
1147 def __init__(self
, cookiejar
=None):
1148 urllib
.request
.HTTPCookieProcessor
.__init
__(self
, cookiejar
)
1150 def http_response(self
, request
, response
):
1151 return urllib
.request
.HTTPCookieProcessor
.http_response(self
, request
, response
)
1153 https_request
= urllib
.request
.HTTPCookieProcessor
.http_request
1154 https_response
= http_response
1157 def extract_timezone(date_str
):
1160 ^.{8,}? # >=8 char non-TZ prefix, if present
1161 (?P<tz>Z| # just the UTC Z, or
1162 (?:(?<=.\b\d{4}|\b\d{2}:\d\d)| # preceded by 4 digits or hh:mm or
1163 (?<!.\b[a-zA-Z]{3}|[a-zA-Z]{4}|..\b\d\d)) # not preceded by 3 alpha word or >= 4 alpha or 2 digits
1164 [ ]? # optional space
1165 (?P<sign>\+|-) # +/-
1166 (?P<hours>[0-9]{2}):?(?P<minutes>[0-9]{2}) # hh[:]mm
1170 m
= re
.search(r
'\d{1,2}:\d{1,2}(?:\.\d+)?(?P<tz>\s*[A-Z]+)$', date_str
)
1171 timezone
= TIMEZONE_NAMES
.get(m
and m
.group('tz').strip())
1172 if timezone
is not None:
1173 date_str
= date_str
[:-len(m
.group('tz'))]
1174 timezone
= datetime
.timedelta(hours
=timezone
or 0)
1176 date_str
= date_str
[:-len(m
.group('tz'))]
1177 if not m
.group('sign'):
1178 timezone
= datetime
.timedelta()
1180 sign
= 1 if m
.group('sign') == '+' else -1
1181 timezone
= datetime
.timedelta(
1182 hours
=sign
* int(m
.group('hours')),
1183 minutes
=sign
* int(m
.group('minutes')))
1184 return timezone
, date_str
1187 def parse_iso8601(date_str
, delimiter
='T', timezone
=None):
1188 """ Return a UNIX timestamp from the given date """
1190 if date_str
is None:
1193 date_str
= re
.sub(r
'\.[0-9]+', '', date_str
)
1195 if timezone
is None:
1196 timezone
, date_str
= extract_timezone(date_str
)
1198 with contextlib
.suppress(ValueError):
1199 date_format
= f
'%Y-%m-%d{delimiter}%H:%M:%S'
1200 dt
= datetime
.datetime
.strptime(date_str
, date_format
) - timezone
1201 return calendar
.timegm(dt
.timetuple())
1204 def date_formats(day_first
=True):
1205 return DATE_FORMATS_DAY_FIRST
if day_first
else DATE_FORMATS_MONTH_FIRST
1208 def unified_strdate(date_str
, day_first
=True):
1209 """Return a string with the date in the format YYYYMMDD"""
1211 if date_str
is None:
1215 date_str
= date_str
.replace(',', ' ')
1216 # Remove AM/PM + timezone
1217 date_str
= re
.sub(r
'(?i)\s*(?:AM|PM)(?:\s+[A-Z]+)?', '', date_str
)
1218 _
, date_str
= extract_timezone(date_str
)
1220 for expression
in date_formats(day_first
):
1221 with contextlib
.suppress(ValueError):
1222 upload_date
= datetime
.datetime
.strptime(date_str
, expression
).strftime('%Y%m%d')
1223 if upload_date
is None:
1224 timetuple
= email
.utils
.parsedate_tz(date_str
)
1226 with contextlib
.suppress(ValueError):
1227 upload_date
= datetime
.datetime(*timetuple
[:6]).strftime('%Y%m%d')
1228 if upload_date
is not None:
1229 return str(upload_date
)
1232 def unified_timestamp(date_str
, day_first
=True):
1233 if not isinstance(date_str
, str):
1236 date_str
= re
.sub(r
'\s+', ' ', re
.sub(
1237 r
'(?i)[,|]|(mon|tues?|wed(nes)?|thu(rs)?|fri|sat(ur)?)(day)?', '', date_str
))
1239 pm_delta
= 12 if re
.search(r
'(?i)PM', date_str
) else 0
1240 timezone
, date_str
= extract_timezone(date_str
)
1242 # Remove AM/PM + timezone
1243 date_str
= re
.sub(r
'(?i)\s*(?:AM|PM)(?:\s+[A-Z]+)?', '', date_str
)
1245 # Remove unrecognized timezones from ISO 8601 alike timestamps
1246 m
= re
.search(r
'\d{1,2}:\d{1,2}(?:\.\d+)?(?P<tz>\s*[A-Z]+)$', date_str
)
1248 date_str
= date_str
[:-len(m
.group('tz'))]
1250 # Python only supports microseconds, so remove nanoseconds
1251 m
= re
.search(r
'^([0-9]{4,}-[0-9]{1,2}-[0-9]{1,2}T[0-9]{1,2}:[0-9]{1,2}:[0-9]{1,2}\.[0-9]{6})[0-9]+$', date_str
)
1253 date_str
= m
.group(1)
1255 for expression
in date_formats(day_first
):
1256 with contextlib
.suppress(ValueError):
1257 dt
= datetime
.datetime
.strptime(date_str
, expression
) - timezone
+ datetime
.timedelta(hours
=pm_delta
)
1258 return calendar
.timegm(dt
.timetuple())
1260 timetuple
= email
.utils
.parsedate_tz(date_str
)
1262 return calendar
.timegm(timetuple
) + pm_delta
* 3600 - timezone
.total_seconds()
1265 def determine_ext(url
, default_ext
='unknown_video'):
1266 if url
is None or '.' not in url
:
1268 guess
= url
.partition('?')[0].rpartition('.')[2]
1269 if re
.match(r
'^[A-Za-z0-9]+$', guess
):
1271 # Try extract ext from URLs like http://example.com/foo/bar.mp4/?download
1272 elif guess
.rstrip('/') in KNOWN_EXTENSIONS
:
1273 return guess
.rstrip('/')
1278 def subtitles_filename(filename
, sub_lang
, sub_format
, expected_real_ext
=None):
1279 return replace_extension(filename
, sub_lang
+ '.' + sub_format
, expected_real_ext
)
1282 def datetime_from_str(date_str
, precision
='auto', format
='%Y%m%d'):
1284 Return a datetime object from a string.
1286 (now|today|yesterday|DATE)([+-]\d+(microsecond|second|minute|hour|day|week|month|year)s?)?
1288 @param format strftime format of DATE
1289 @param precision Round the datetime object: auto|microsecond|second|minute|hour|day
1290 auto: round to the unit provided in date_str (if applicable).
1292 auto_precision
= False
1293 if precision
== 'auto':
1294 auto_precision
= True
1295 precision
= 'microsecond'
1296 today
= datetime_round(datetime
.datetime
.utcnow(), precision
)
1297 if date_str
in ('now', 'today'):
1299 if date_str
== 'yesterday':
1300 return today
- datetime
.timedelta(days
=1)
1302 r
'(?P<start>.+)(?P<sign>[+-])(?P<time>\d+)(?P<unit>microsecond|second|minute|hour|day|week|month|year)s?',
1304 if match
is not None:
1305 start_time
= datetime_from_str(match
.group('start'), precision
, format
)
1306 time
= int(match
.group('time')) * (-1 if match
.group('sign') == '-' else 1)
1307 unit
= match
.group('unit')
1308 if unit
== 'month' or unit
== 'year':
1309 new_date
= datetime_add_months(start_time
, time
* 12 if unit
== 'year' else time
)
1315 delta
= datetime
.timedelta(**{unit + 's': time}
)
1316 new_date
= start_time
+ delta
1318 return datetime_round(new_date
, unit
)
1321 return datetime_round(datetime
.datetime
.strptime(date_str
, format
), precision
)
1324 def date_from_str(date_str
, format
='%Y%m%d', strict
=False):
1326 Return a date object from a string using datetime_from_str
1328 @param strict Restrict allowed patterns to "YYYYMMDD" and
1329 (now|today|yesterday)(-\d+(day|week|month|year)s?)?
1331 if strict
and not re
.fullmatch(r
'\d{8}|(now|today|yesterday)(-\d+(day|week|month|year)s?)?', date_str
):
1332 raise ValueError(f
'Invalid date format "{date_str}"')
1333 return datetime_from_str(date_str
, precision
='microsecond', format
=format
).date()
1336 def datetime_add_months(dt
, months
):
1337 """Increment/Decrement a datetime object by months."""
1338 month
= dt
.month
+ months
- 1
1339 year
= dt
.year
+ month
// 12
1340 month
= month
% 12 + 1
1341 day
= min(dt
.day
, calendar
.monthrange(year
, month
)[1])
1342 return dt
.replace(year
, month
, day
)
1345 def datetime_round(dt
, precision
='day'):
1347 Round a datetime object's time to a specific precision
1349 if precision
== 'microsecond':
1358 roundto
= lambda x
, n
: ((x
+ n
/ 2) // n
) * n
1359 timestamp
= calendar
.timegm(dt
.timetuple())
1360 return datetime
.datetime
.utcfromtimestamp(roundto(timestamp
, unit_seconds
[precision
]))
1363 def hyphenate_date(date_str
):
1365 Convert a date in 'YYYYMMDD' format to 'YYYY-MM-DD' format"""
1366 match
= re
.match(r
'^(\d\d\d\d)(\d\d)(\d\d)$', date_str
)
1367 if match
is not None:
1368 return '-'.join(match
.groups())
1374 """Represents a time interval between two dates"""
1376 def __init__(self
, start
=None, end
=None):
1377 """start and end must be strings in the format accepted by date"""
1378 if start
is not None:
1379 self
.start
= date_from_str(start
, strict
=True)
1381 self
.start
= datetime
.datetime
.min.date()
1383 self
.end
= date_from_str(end
, strict
=True)
1385 self
.end
= datetime
.datetime
.max.date()
1386 if self
.start
> self
.end
:
1387 raise ValueError('Date range: "%s" , the start date must be before the end date' % self
)
1391 """Returns a range that only contains the given day"""
1392 return cls(day
, day
)
1394 def __contains__(self
, date
):
1395 """Check if the date is in the range"""
1396 if not isinstance(date
, datetime
.date
):
1397 date
= date_from_str(date
)
1398 return self
.start
<= date
<= self
.end
1401 return f
'{__name__}.{type(self).__name__}({self.start.isoformat()!r}, {self.end.isoformat()!r})'
1403 def __eq__(self
, other
):
1404 return (isinstance(other
, DateRange
)
1405 and self
.start
== other
.start
and self
.end
== other
.end
)
1409 def system_identifier():
1410 python_implementation
= platform
.python_implementation()
1411 if python_implementation
== 'PyPy' and hasattr(sys
, 'pypy_version_info'):
1412 python_implementation
+= ' version %d.%d.%d' % sys
.pypy_version_info
[:3]
1414 with contextlib
.suppress(OSError): # We may not have access to the executable
1415 libc_ver
= platform
.libc_ver()
1417 return 'Python %s (%s %s %s) - %s (%s%s)' % (
1418 platform
.python_version(),
1419 python_implementation
,
1421 platform
.architecture()[0],
1422 platform
.platform(),
1423 ssl
.OPENSSL_VERSION
,
1424 format_field(join_nonempty(*libc_ver
, delim
=' '), None, ', %s'),
1429 def get_windows_version():
1430 ''' Get Windows version. returns () if it's not running on Windows '''
1431 if compat_os_name
== 'nt':
1432 return version_tuple(platform
.win32_ver()[1])
1437 def write_string(s
, out
=None, encoding
=None):
1438 assert isinstance(s
, str)
1439 out
= out
or sys
.stderr
1440 # `sys.stderr` might be `None` (Ref: https://github.com/pyinstaller/pyinstaller/pull/7217)
1444 if compat_os_name
== 'nt' and supports_terminal_sequences(out
):
1445 s
= re
.sub(r
'([\r\n]+)', r
' \1', s
)
1447 enc
, buffer = None, out
1448 if 'b' in getattr(out
, 'mode', ''):
1449 enc
= encoding
or preferredencoding()
1450 elif hasattr(out
, 'buffer'):
1452 enc
= encoding
or getattr(out
, 'encoding', None) or preferredencoding()
1454 buffer.write(s
.encode(enc
, 'ignore') if enc
else s
)
1458 def deprecation_warning(msg
, *, printer
=None, stacklevel
=0, **kwargs
):
1459 from .. import _IN_CLI
1461 if msg
in deprecation_warning
._cache
:
1463 deprecation_warning
._cache
.add(msg
)
1465 return printer(f
'{msg}{bug_reports_message()}', **kwargs
)
1466 return write_string(f
'ERROR: {msg}{bug_reports_message()}\n', **kwargs
)
1469 warnings
.warn(DeprecationWarning(msg
), stacklevel
=stacklevel
+ 3)
1472 deprecation_warning
._cache
= set()
1475 def bytes_to_intlist(bs
):
1478 if isinstance(bs
[0], int): # Python 3
1481 return [ord(c
) for c
in bs
]
1484 def intlist_to_bytes(xs
):
1487 return struct
.pack('%dB' % len(xs
), *xs
)
1490 class LockingUnsupportedError(OSError):
1491 msg
= 'File locking is not supported'
1494 super().__init
__(self
.msg
)
1497 # Cross-platform file locking
1498 if sys
.platform
== 'win32':
1500 import ctypes
.wintypes
1503 class OVERLAPPED(ctypes
.Structure
):
1505 ('Internal', ctypes
.wintypes
.LPVOID
),
1506 ('InternalHigh', ctypes
.wintypes
.LPVOID
),
1507 ('Offset', ctypes
.wintypes
.DWORD
),
1508 ('OffsetHigh', ctypes
.wintypes
.DWORD
),
1509 ('hEvent', ctypes
.wintypes
.HANDLE
),
1512 kernel32
= ctypes
.WinDLL('kernel32')
1513 LockFileEx
= kernel32
.LockFileEx
1514 LockFileEx
.argtypes
= [
1515 ctypes
.wintypes
.HANDLE
, # hFile
1516 ctypes
.wintypes
.DWORD
, # dwFlags
1517 ctypes
.wintypes
.DWORD
, # dwReserved
1518 ctypes
.wintypes
.DWORD
, # nNumberOfBytesToLockLow
1519 ctypes
.wintypes
.DWORD
, # nNumberOfBytesToLockHigh
1520 ctypes
.POINTER(OVERLAPPED
) # Overlapped
1522 LockFileEx
.restype
= ctypes
.wintypes
.BOOL
1523 UnlockFileEx
= kernel32
.UnlockFileEx
1524 UnlockFileEx
.argtypes
= [
1525 ctypes
.wintypes
.HANDLE
, # hFile
1526 ctypes
.wintypes
.DWORD
, # dwReserved
1527 ctypes
.wintypes
.DWORD
, # nNumberOfBytesToLockLow
1528 ctypes
.wintypes
.DWORD
, # nNumberOfBytesToLockHigh
1529 ctypes
.POINTER(OVERLAPPED
) # Overlapped
1531 UnlockFileEx
.restype
= ctypes
.wintypes
.BOOL
1532 whole_low
= 0xffffffff
1533 whole_high
= 0x7fffffff
1535 def _lock_file(f
, exclusive
, block
):
1536 overlapped
= OVERLAPPED()
1537 overlapped
.Offset
= 0
1538 overlapped
.OffsetHigh
= 0
1539 overlapped
.hEvent
= 0
1540 f
._lock
_file
_overlapped
_p
= ctypes
.pointer(overlapped
)
1542 if not LockFileEx(msvcrt
.get_osfhandle(f
.fileno()),
1543 (0x2 if exclusive
else 0x0) |
(0x0 if block
else 0x1),
1544 0, whole_low
, whole_high
, f
._lock
_file
_overlapped
_p
):
1545 # NB: No argument form of "ctypes.FormatError" does not work on PyPy
1546 raise BlockingIOError(f
'Locking file failed: {ctypes.FormatError(ctypes.GetLastError())!r}')
1548 def _unlock_file(f
):
1549 assert f
._lock
_file
_overlapped
_p
1550 handle
= msvcrt
.get_osfhandle(f
.fileno())
1551 if not UnlockFileEx(handle
, 0, whole_low
, whole_high
, f
._lock
_file
_overlapped
_p
):
1552 raise OSError('Unlocking file failed: %r' % ctypes
.FormatError())
1558 def _lock_file(f
, exclusive
, block
):
1559 flags
= fcntl
.LOCK_EX
if exclusive
else fcntl
.LOCK_SH
1561 flags |
= fcntl
.LOCK_NB
1563 fcntl
.flock(f
, flags
)
1564 except BlockingIOError
:
1566 except OSError: # AOSP does not have flock()
1567 fcntl
.lockf(f
, flags
)
1569 def _unlock_file(f
):
1570 with contextlib
.suppress(OSError):
1571 return fcntl
.flock(f
, fcntl
.LOCK_UN
)
1572 with contextlib
.suppress(OSError):
1573 return fcntl
.lockf(f
, fcntl
.LOCK_UN
) # AOSP does not have flock()
1574 return fcntl
.flock(f
, fcntl
.LOCK_UN | fcntl
.LOCK_NB
) # virtiofs needs LOCK_NB on unlocking
1578 def _lock_file(f
, exclusive
, block
):
1579 raise LockingUnsupportedError()
1581 def _unlock_file(f
):
1582 raise LockingUnsupportedError()
1588 def __init__(self
, filename
, mode
, block
=True, encoding
=None):
1589 if mode
not in {'r', 'rb', 'a', 'ab', 'w', 'wb'}
:
1590 raise NotImplementedError(mode
)
1591 self
.mode
, self
.block
= mode
, block
1593 writable
= any(f
in mode
for f
in 'wax+')
1594 readable
= any(f
in mode
for f
in 'r+')
1595 flags
= functools
.reduce(operator
.ior
, (
1596 getattr(os
, 'O_CLOEXEC', 0), # UNIX only
1597 getattr(os
, 'O_BINARY', 0), # Windows only
1598 getattr(os
, 'O_NOINHERIT', 0), # Windows only
1599 os
.O_CREAT
if writable
else 0, # O_TRUNC only after locking
1600 os
.O_APPEND
if 'a' in mode
else 0,
1601 os
.O_EXCL
if 'x' in mode
else 0,
1602 os
.O_RDONLY
if not writable
else os
.O_RDWR
if readable
else os
.O_WRONLY
,
1605 self
.f
= os
.fdopen(os
.open(filename
, flags
, 0o666), mode
, encoding
=encoding
)
1607 def __enter__(self
):
1608 exclusive
= 'r' not in self
.mode
1610 _lock_file(self
.f
, exclusive
, self
.block
)
1615 if 'w' in self
.mode
:
1618 except OSError as e
:
1620 errno
.ESPIPE
, # Illegal seek - expected for FIFO
1621 errno
.EINVAL
, # Invalid argument - expected for /dev/null
1630 _unlock_file(self
.f
)
1634 def __exit__(self
, *_
):
1643 def __getattr__(self
, attr
):
1644 return getattr(self
.f
, attr
)
1651 def get_filesystem_encoding():
1652 encoding
= sys
.getfilesystemencoding()
1653 return encoding
if encoding
is not None else 'utf-8'
1656 def shell_quote(args
):
1658 encoding
= get_filesystem_encoding()
1660 if isinstance(a
, bytes):
1661 # We may get a filename encoded with 'encodeFilename'
1662 a
= a
.decode(encoding
)
1663 quoted_args
.append(compat_shlex_quote(a
))
1664 return ' '.join(quoted_args
)
1667 def smuggle_url(url
, data
):
1668 """ Pass additional data in a URL for internal use. """
1670 url
, idata
= unsmuggle_url(url
, {})
1672 sdata
= urllib
.parse
.urlencode(
1673 {'__youtubedl_smuggle': json.dumps(data)}
)
1674 return url
+ '#' + sdata
1677 def unsmuggle_url(smug_url
, default
=None):
1678 if '#__youtubedl_smuggle' not in smug_url
:
1679 return smug_url
, default
1680 url
, _
, sdata
= smug_url
.rpartition('#')
1681 jsond
= urllib
.parse
.parse_qs(sdata
)['__youtubedl_smuggle'][0]
1682 data
= json
.loads(jsond
)
1686 def format_decimal_suffix(num
, fmt
='%d%s', *, factor
=1000):
1687 """ Formats numbers with decimal sufixes like K, M, etc """
1688 num
, factor
= float_or_none(num
), float(factor
)
1689 if num
is None or num
< 0:
1691 POSSIBLE_SUFFIXES
= 'kMGTPEZY'
1692 exponent
= 0 if num
== 0 else min(int(math
.log(num
, factor
)), len(POSSIBLE_SUFFIXES
))
1693 suffix
= ['', *POSSIBLE_SUFFIXES
][exponent
]
1695 suffix
= {'k': 'Ki', '': ''}
.get(suffix
, f
'{suffix}i')
1696 converted
= num
/ (factor
** exponent
)
1697 return fmt
% (converted
, suffix
)
1700 def format_bytes(bytes):
1701 return format_decimal_suffix(bytes, '%.2f%sB', factor
=1024) or 'N/A'
1704 def lookup_unit_table(unit_table
, s
, strict
=False):
1705 num_re
= NUMBER_RE
if strict
else NUMBER_RE
.replace(R
'\.', '[,.]')
1706 units_re
= '|'.join(re
.escape(u
) for u
in unit_table
)
1707 m
= (re
.fullmatch
if strict
else re
.match
)(
1708 rf
'(?P<num>{num_re})\s*(?P<unit>{units_re})\b', s
)
1712 num
= float(m
.group('num').replace(',', '.'))
1713 mult
= unit_table
[m
.group('unit')]
1714 return round(num
* mult
)
1718 """Parse a string indicating a byte quantity into an integer"""
1719 return lookup_unit_table(
1720 {u: 1024**i for i, u in enumerate(['', *'KMGTPEZY'])}
,
1721 s
.upper(), strict
=True)
1724 def parse_filesize(s
):
1728 # The lower-case forms are of course incorrect and unofficial,
1729 # but we support those too
1746 'megabytes': 1000 ** 2,
1747 'mebibytes': 1024 ** 2,
1753 'gigabytes': 1000 ** 3,
1754 'gibibytes': 1024 ** 3,
1760 'terabytes': 1000 ** 4,
1761 'tebibytes': 1024 ** 4,
1767 'petabytes': 1000 ** 5,
1768 'pebibytes': 1024 ** 5,
1774 'exabytes': 1000 ** 6,
1775 'exbibytes': 1024 ** 6,
1781 'zettabytes': 1000 ** 7,
1782 'zebibytes': 1024 ** 7,
1788 'yottabytes': 1000 ** 8,
1789 'yobibytes': 1024 ** 8,
1792 return lookup_unit_table(_UNIT_TABLE
, s
)
1799 s
= re
.sub(r
'^[^\d]+\s', '', s
).strip()
1801 if re
.match(r
'^[\d,.]+$', s
):
1802 return str_to_int(s
)
1815 ret
= lookup_unit_table(_UNIT_TABLE
, s
)
1819 mobj
= re
.match(r
'([\d,.]+)(?:$|\s)', s
)
1821 return str_to_int(mobj
.group(1))
1824 def parse_resolution(s
, *, lenient
=False):
1829 mobj
= re
.search(r
'(?P<w>\d+)\s*[xX×,]\s*(?P<h>\d+)', s
)
1831 mobj
= re
.search(r
'(?<![a-zA-Z0-9])(?P<w>\d+)\s*[xX×,]\s*(?P<h>\d+)(?![a-zA-Z0-9])', s
)
1834 'width': int(mobj
.group('w')),
1835 'height': int(mobj
.group('h')),
1838 mobj
= re
.search(r
'(?<![a-zA-Z0-9])(\d+)[pPiI](?![a-zA-Z0-9])', s
)
1840 return {'height': int(mobj.group(1))}
1842 mobj
= re
.search(r
'\b([48])[kK]\b', s
)
1844 return {'height': int(mobj.group(1)) * 540}
1849 def parse_bitrate(s
):
1850 if not isinstance(s
, str):
1852 mobj
= re
.search(r
'\b(\d+)\s*kbps', s
)
1854 return int(mobj
.group(1))
1857 def month_by_name(name
, lang
='en'):
1858 """ Return the number of a month by (locale-independently) English name """
1860 month_names
= MONTH_NAMES
.get(lang
, MONTH_NAMES
['en'])
1863 return month_names
.index(name
) + 1
1868 def month_by_abbreviation(abbrev
):
1869 """ Return the number of a month by (locale-independently) English
1873 return [s
[:3] for s
in ENGLISH_MONTH_NAMES
].index(abbrev
) + 1
1878 def fix_xml_ampersands(xml_str
):
1879 """Replace all the '&' by '&' in XML"""
1881 r
'&(?!amp;|lt;|gt;|apos;|quot;|#x[0-9a-fA-F]{,4};|#[0-9]{,4};)',
1886 def setproctitle(title
):
1887 assert isinstance(title
, str)
1889 # Workaround for https://github.com/yt-dlp/yt-dlp/issues/4541
1896 libc
= ctypes
.cdll
.LoadLibrary('libc.so.6')
1900 # LoadLibrary in Windows Python 2.7.13 only expects
1901 # a bytestring, but since unicode_literals turns
1902 # every string into a unicode string, it fails.
1904 title_bytes
= title
.encode()
1905 buf
= ctypes
.create_string_buffer(len(title_bytes
))
1906 buf
.value
= title_bytes
1908 libc
.prctl(15, buf
, 0, 0, 0)
1909 except AttributeError:
1910 return # Strange libc, just skip this
1913 def remove_start(s
, start
):
1914 return s
[len(start
):] if s
is not None and s
.startswith(start
) else s
1917 def remove_end(s
, end
):
1918 return s
[:-len(end
)] if s
is not None and s
.endswith(end
) else s
1921 def remove_quotes(s
):
1922 if s
is None or len(s
) < 2:
1924 for quote
in ('"', "'", ):
1925 if s
[0] == quote
and s
[-1] == quote
:
1930 def get_domain(url
):
1932 This implementation is inconsistent, but is kept for compatibility.
1933 Use this only for "webpage_url_domain"
1935 return remove_start(urllib
.parse
.urlparse(url
).netloc
, 'www.') or None
1938 def url_basename(url
):
1939 path
= urllib
.parse
.urlparse(url
).path
1940 return path
.strip('/').split('/')[-1]
1944 return re
.match(r
'https?://[^?#]+/', url
).group()
1947 def urljoin(base
, path
):
1948 if isinstance(path
, bytes):
1949 path
= path
.decode()
1950 if not isinstance(path
, str) or not path
:
1952 if re
.match(r
'^(?:[a-zA-Z][a-zA-Z0-9+-.]*:)?//', path
):
1954 if isinstance(base
, bytes):
1955 base
= base
.decode()
1956 if not isinstance(base
, str) or not re
.match(
1957 r
'^(?:https?:)?//', base
):
1959 return urllib
.parse
.urljoin(base
, path
)
1962 def int_or_none(v
, scale
=1, default
=None, get_attr
=None, invscale
=1):
1963 if get_attr
and v
is not None:
1964 v
= getattr(v
, get_attr
, None)
1966 return int(v
) * invscale
// scale
1967 except (ValueError, TypeError, OverflowError):
1971 def str_or_none(v
, default
=None):
1972 return default
if v
is None else str(v
)
1975 def str_to_int(int_str
):
1976 """ A more relaxed version of int_or_none """
1977 if isinstance(int_str
, int):
1979 elif isinstance(int_str
, str):
1980 int_str
= re
.sub(r
'[,\.\+]', '', int_str
)
1981 return int_or_none(int_str
)
1984 def float_or_none(v
, scale
=1, invscale
=1, default
=None):
1988 return float(v
) * invscale
/ scale
1989 except (ValueError, TypeError):
1993 def bool_or_none(v
, default
=None):
1994 return v
if isinstance(v
, bool) else default
1997 def strip_or_none(v
, default
=None):
1998 return v
.strip() if isinstance(v
, str) else default
2001 def url_or_none(url
):
2002 if not url
or not isinstance(url
, str):
2005 return url
if re
.match(r
'^(?:(?:https?|rt(?:m(?:pt?[es]?|fp)|sp[su]?)|mms|ftps?):)?//', url
) else None
2008 def request_to_url(req
):
2009 if isinstance(req
, urllib
.request
.Request
):
2010 return req
.get_full_url()
2015 def strftime_or_none(timestamp
, date_format
='%Y%m%d', default
=None):
2016 datetime_object
= None
2018 if isinstance(timestamp
, (int, float)): # unix timestamp
2019 # Using naive datetime here can break timestamp() in Windows
2020 # Ref: https://github.com/yt-dlp/yt-dlp/issues/5185, https://github.com/python/cpython/issues/94414
2021 # Also, datetime.datetime.fromtimestamp breaks for negative timestamps
2022 # Ref: https://github.com/yt-dlp/yt-dlp/issues/6706#issuecomment-1496842642
2023 datetime_object
= (datetime
.datetime
.fromtimestamp(0, datetime
.timezone
.utc
)
2024 + datetime
.timedelta(seconds
=timestamp
))
2025 elif isinstance(timestamp
, str): # assume YYYYMMDD
2026 datetime_object
= datetime
.datetime
.strptime(timestamp
, '%Y%m%d')
2027 date_format
= re
.sub( # Support %s on windows
2028 r
'(?<!%)(%%)*%s', rf
'\g<1>{int(datetime_object.timestamp())}', date_format
)
2029 return datetime_object
.strftime(date_format
)
2030 except (ValueError, TypeError, AttributeError):
2034 def parse_duration(s
):
2035 if not isinstance(s
, str):
2041 days
, hours
, mins
, secs
, ms
= [None] * 5
2042 m
= re
.match(r
'''(?x)
2044 (?:(?:(?P<days>[0-9]+):)?(?P<hours>[0-9]+):)?(?P<mins>[0-9]+):)?
2045 (?P<secs>(?(before_secs)[0-9]{1,2}|[0-9]+))
2046 (?P<ms>[.:][0-9]+)?Z?$
2049 days
, hours
, mins
, secs
, ms
= m
.group('days', 'hours', 'mins', 'secs', 'ms')
2054 [0-9]+\s*y(?:ears?)?,?\s*
2057 [0-9]+\s*m(?:onths?)?,?\s*
2060 [0-9]+\s*w(?:eeks?)?,?\s*
2063 (?P<days>[0-9]+)\s*d(?:ays?)?,?\s*
2067 (?P<hours>[0-9]+)\s*h(?:ours?)?,?\s*
2070 (?P<mins>[0-9]+)\s*m(?:in(?:ute)?s?)?,?\s*
2073 (?P<secs>[0-9]+)(?P<ms>\.[0-9]+)?\s*s(?:ec(?:ond)?s?)?\s*
2076 days
, hours
, mins
, secs
, ms
= m
.groups()
2078 m
= re
.match(r
'(?i)(?:(?P<hours>[0-9.]+)\s*(?:hours?)|(?P<mins>[0-9.]+)\s*(?:mins?\.?|minutes?)\s*)Z?$', s
)
2080 hours
, mins
= m
.groups()
2085 ms
= ms
.replace(':', '.')
2086 return sum(float(part
or 0) * mult
for part
, mult
in (
2087 (days
, 86400), (hours
, 3600), (mins
, 60), (secs
, 1), (ms
, 1)))
2090 def prepend_extension(filename
, ext
, expected_real_ext
=None):
2091 name
, real_ext
= os
.path
.splitext(filename
)
2093 f
'{name}.{ext}{real_ext}'
2094 if not expected_real_ext
or real_ext
[1:] == expected_real_ext
2095 else f
'{filename}.{ext}')
2098 def replace_extension(filename
, ext
, expected_real_ext
=None):
2099 name
, real_ext
= os
.path
.splitext(filename
)
2100 return '{}.{}'.format(
2101 name
if not expected_real_ext
or real_ext
[1:] == expected_real_ext
else filename
,
2105 def check_executable(exe
, args
=[]):
2106 """ Checks if the given binary is installed somewhere in PATH, and returns its name.
2107 args can be a list of arguments for a short output (like -version) """
2109 Popen
.run([exe
] + args
, stdout
=subprocess
.PIPE
, stderr
=subprocess
.PIPE
)
2115 def _get_exe_version_output(exe
, args
):
2117 # STDIN should be redirected too. On UNIX-like systems, ffmpeg triggers
2118 # SIGTTOU if yt-dlp is run in the background.
2119 # See https://github.com/ytdl-org/youtube-dl/issues/955#issuecomment-209789656
2120 stdout
, _
, ret
= Popen
.run([encodeArgument(exe
)] + args
, text
=True,
2121 stdin
=subprocess
.PIPE
, stdout
=subprocess
.PIPE
, stderr
=subprocess
.STDOUT
)
2129 def detect_exe_version(output
, version_re
=None, unrecognized
='present'):
2130 assert isinstance(output
, str)
2131 if version_re
is None:
2132 version_re
= r
'version\s+([-0-9._a-zA-Z]+)'
2133 m
= re
.search(version_re
, output
)
2140 def get_exe_version(exe
, args
=['--version'],
2141 version_re
=None, unrecognized
=('present', 'broken')):
2142 """ Returns the version of the specified executable,
2143 or False if the executable is not present """
2144 unrecognized
= variadic(unrecognized
)
2145 assert len(unrecognized
) in (1, 2)
2146 out
= _get_exe_version_output(exe
, args
)
2148 return unrecognized
[-1]
2149 return out
and detect_exe_version(out
, version_re
, unrecognized
[0])
2152 def frange(start
=0, stop
=None, step
=1):
2155 start
, stop
= 0, start
2156 sign
= [-1, 1][step
> 0] if step
else 0
2157 while sign
* start
< sign
* stop
:
2162 class LazyList(collections
.abc
.Sequence
):
2163 """Lazy immutable list from an iterable
2164 Note that slices of a LazyList are lists and not LazyList"""
2166 class IndexError(IndexError):
2169 def __init__(self
, iterable
, *, reverse
=False, _cache
=None):
2170 self
._iterable
= iter(iterable
)
2171 self
._cache
= [] if _cache
is None else _cache
2172 self
._reversed
= reverse
2176 # We need to consume the entire iterable to iterate in reverse
2177 yield from self
.exhaust()
2179 yield from self
._cache
2180 for item
in self
._iterable
:
2181 self
._cache
.append(item
)
2185 self
._cache
.extend(self
._iterable
)
2186 self
._iterable
= [] # Discard the emptied iterable to make it pickle-able
2190 """Evaluate the entire iterable"""
2191 return self
._exhaust
()[::-1 if self
._reversed
else 1]
2194 def _reverse_index(x
):
2195 return None if x
is None else ~x
2197 def __getitem__(self
, idx
):
2198 if isinstance(idx
, slice):
2200 idx
= slice(self
._reverse
_index
(idx
.start
), self
._reverse
_index
(idx
.stop
), -(idx
.step
or 1))
2201 start
, stop
, step
= idx
.start
, idx
.stop
, idx
.step
or 1
2202 elif isinstance(idx
, int):
2204 idx
= self
._reverse
_index
(idx
)
2205 start
, stop
, step
= idx
, idx
, 0
2207 raise TypeError('indices must be integers or slices')
2208 if ((start
or 0) < 0 or (stop
or 0) < 0
2209 or (start
is None and step
< 0)
2210 or (stop
is None and step
> 0)):
2211 # We need to consume the entire iterable to be able to slice from the end
2212 # Obviously, never use this with infinite iterables
2215 return self
._cache
[idx
]
2216 except IndexError as e
:
2217 raise self
.IndexError(e
) from e
2218 n
= max(start
or 0, stop
or 0) - len(self
._cache
) + 1
2220 self
._cache
.extend(itertools
.islice(self
._iterable
, n
))
2222 return self
._cache
[idx
]
2223 except IndexError as e
:
2224 raise self
.IndexError(e
) from e
2228 self
[-1] if self
._reversed
else self
[0]
2229 except self
.IndexError:
2235 return len(self
._cache
)
2237 def __reversed__(self
):
2238 return type(self
)(self
._iterable
, reverse
=not self
._reversed
, _cache
=self
._cache
)
2241 return type(self
)(self
._iterable
, reverse
=self
._reversed
, _cache
=self
._cache
)
2244 # repr and str should mimic a list. So we exhaust the iterable
2245 return repr(self
.exhaust())
2248 return repr(self
.exhaust())
2253 class IndexError(IndexError):
2257 # This is only useful for tests
2258 return len(self
.getslice())
2260 def __init__(self
, pagefunc
, pagesize
, use_cache
=True):
2261 self
._pagefunc
= pagefunc
2262 self
._pagesize
= pagesize
2263 self
._pagecount
= float('inf')
2264 self
._use
_cache
= use_cache
2267 def getpage(self
, pagenum
):
2268 page_results
= self
._cache
.get(pagenum
)
2269 if page_results
is None:
2270 page_results
= [] if pagenum
> self
._pagecount
else list(self
._pagefunc
(pagenum
))
2272 self
._cache
[pagenum
] = page_results
2275 def getslice(self
, start
=0, end
=None):
2276 return list(self
._getslice
(start
, end
))
2278 def _getslice(self
, start
, end
):
2279 raise NotImplementedError('This method must be implemented by subclasses')
2281 def __getitem__(self
, idx
):
2282 assert self
._use
_cache
, 'Indexing PagedList requires cache'
2283 if not isinstance(idx
, int) or idx
< 0:
2284 raise TypeError('indices must be non-negative integers')
2285 entries
= self
.getslice(idx
, idx
+ 1)
2287 raise self
.IndexError()
2291 class OnDemandPagedList(PagedList
):
2292 """Download pages until a page with less than maximum results"""
2294 def _getslice(self
, start
, end
):
2295 for pagenum
in itertools
.count(start
// self
._pagesize
):
2296 firstid
= pagenum
* self
._pagesize
2297 nextfirstid
= pagenum
* self
._pagesize
+ self
._pagesize
2298 if start
>= nextfirstid
:
2302 start
% self
._pagesize
2303 if firstid
<= start
< nextfirstid
2306 ((end
- 1) % self
._pagesize
) + 1
2307 if (end
is not None and firstid
<= end
<= nextfirstid
)
2311 page_results
= self
.getpage(pagenum
)
2313 self
._pagecount
= pagenum
- 1
2315 if startv
!= 0 or endv
is not None:
2316 page_results
= page_results
[startv
:endv
]
2317 yield from page_results
2319 # A little optimization - if current page is not "full", ie. does
2320 # not contain page_size videos then we can assume that this page
2321 # is the last one - there are no more ids on further pages -
2322 # i.e. no need to query again.
2323 if len(page_results
) + startv
< self
._pagesize
:
2326 # If we got the whole page, but the next page is not interesting,
2327 # break out early as well
2328 if end
== nextfirstid
:
2332 class InAdvancePagedList(PagedList
):
2333 """PagedList with total number of pages known in advance"""
2335 def __init__(self
, pagefunc
, pagecount
, pagesize
):
2336 PagedList
.__init
__(self
, pagefunc
, pagesize
, True)
2337 self
._pagecount
= pagecount
2339 def _getslice(self
, start
, end
):
2340 start_page
= start
// self
._pagesize
2341 end_page
= self
._pagecount
if end
is None else min(self
._pagecount
, end
// self
._pagesize
+ 1)
2342 skip_elems
= start
- start_page
* self
._pagesize
2343 only_more
= None if end
is None else end
- start
2344 for pagenum
in range(start_page
, end_page
):
2345 page_results
= self
.getpage(pagenum
)
2347 page_results
= page_results
[skip_elems
:]
2349 if only_more
is not None:
2350 if len(page_results
) < only_more
:
2351 only_more
-= len(page_results
)
2353 yield from page_results
[:only_more
]
2355 yield from page_results
2358 class PlaylistEntries
:
2359 MissingEntry
= object()
2360 is_exhausted
= False
2362 def __init__(self
, ydl
, info_dict
):
2365 # _entries must be assigned now since infodict can change during iteration
2366 entries
= info_dict
.get('entries')
2368 raise EntryNotInPlaylist('There are no entries')
2369 elif isinstance(entries
, list):
2370 self
.is_exhausted
= True
2372 requested_entries
= info_dict
.get('requested_entries')
2373 self
.is_incomplete
= requested_entries
is not None
2374 if self
.is_incomplete
:
2375 assert self
.is_exhausted
2376 self
._entries
= [self
.MissingEntry
] * max(requested_entries
or [0])
2377 for i
, entry
in zip(requested_entries
, entries
):
2378 self
._entries
[i
- 1] = entry
2379 elif isinstance(entries
, (list, PagedList
, LazyList
)):
2380 self
._entries
= entries
2382 self
._entries
= LazyList(entries
)
2384 PLAYLIST_ITEMS_RE
= re
.compile(r
'''(?x)
2385 (?P<start>[+-]?\d+)?
2387 (?P<end>[+-]?\d+|inf(?:inite)?)?
2388 (?::(?P<step>[+-]?\d+))?
2392 def parse_playlist_items(cls
, string
):
2393 for segment
in string
.split(','):
2395 raise ValueError('There is two or more consecutive commas')
2396 mobj
= cls
.PLAYLIST_ITEMS_RE
.fullmatch(segment
)
2398 raise ValueError(f
'{segment!r} is not a valid specification')
2399 start
, end
, step
, has_range
= mobj
.group('start', 'end', 'step', 'range')
2400 if int_or_none(step
) == 0:
2401 raise ValueError(f
'Step in {segment!r} cannot be zero')
2402 yield slice(int_or_none(start
), float_or_none(end
), int_or_none(step
)) if has_range
else int(start
)
2404 def get_requested_items(self
):
2405 playlist_items
= self
.ydl
.params
.get('playlist_items')
2406 playlist_start
= self
.ydl
.params
.get('playliststart', 1)
2407 playlist_end
= self
.ydl
.params
.get('playlistend')
2408 # For backwards compatibility, interpret -1 as whole list
2409 if playlist_end
in (-1, None):
2411 if not playlist_items
:
2412 playlist_items
= f
'{playlist_start}:{playlist_end}'
2413 elif playlist_start
!= 1 or playlist_end
:
2414 self
.ydl
.report_warning('Ignoring playliststart and playlistend because playlistitems was given', only_once
=True)
2416 for index
in self
.parse_playlist_items(playlist_items
):
2417 for i
, entry
in self
[index
]:
2422 # The item may have just been added to archive. Don't break due to it
2423 if not self
.ydl
.params
.get('lazy_playlist'):
2424 # TODO: Add auto-generated fields
2425 self
.ydl
._match
_entry
(entry
, incomplete
=True, silent
=True)
2426 except (ExistingVideoReached
, RejectedVideoReached
):
2429 def get_full_count(self
):
2430 if self
.is_exhausted
and not self
.is_incomplete
:
2432 elif isinstance(self
._entries
, InAdvancePagedList
):
2433 if self
._entries
._pagesize
== 1:
2434 return self
._entries
._pagecount
2436 @functools.cached_property
2438 if isinstance(self
._entries
, list):
2441 entry
= self
._entries
[i
]
2443 entry
= self
.MissingEntry
2444 if not self
.is_incomplete
:
2445 raise self
.IndexError()
2446 if entry
is self
.MissingEntry
:
2447 raise EntryNotInPlaylist(f
'Entry {i + 1} cannot be found')
2452 return type(self
.ydl
)._handle
_extraction
_exceptions
(lambda _
, i
: self
._entries
[i
])(self
.ydl
, i
)
2453 except (LazyList
.IndexError, PagedList
.IndexError):
2454 raise self
.IndexError()
2457 def __getitem__(self
, idx
):
2458 if isinstance(idx
, int):
2459 idx
= slice(idx
, idx
)
2461 # NB: PlaylistEntries[1:10] => (0, 1, ... 9)
2462 step
= 1 if idx
.step
is None else idx
.step
2463 if idx
.start
is None:
2464 start
= 0 if step
> 0 else len(self
) - 1
2466 start
= idx
.start
- 1 if idx
.start
>= 0 else len(self
) + idx
.start
2468 # NB: Do not call len(self) when idx == [:]
2469 if idx
.stop
is None:
2470 stop
= 0 if step
< 0 else float('inf')
2472 stop
= idx
.stop
- 1 if idx
.stop
>= 0 else len(self
) + idx
.stop
2473 stop
+= [-1, 1][step
> 0]
2475 for i
in frange(start
, stop
, step
):
2479 entry
= self
._getter
(i
)
2480 except self
.IndexError:
2481 self
.is_exhausted
= True
2488 return len(tuple(self
[:]))
2490 class IndexError(IndexError):
2494 def uppercase_escape(s
):
2495 unicode_escape
= codecs
.getdecoder('unicode_escape')
2497 r
'\\U[0-9a-fA-F]{8}',
2498 lambda m
: unicode_escape(m
.group(0))[0],
2502 def lowercase_escape(s
):
2503 unicode_escape
= codecs
.getdecoder('unicode_escape')
2505 r
'\\u[0-9a-fA-F]{4}',
2506 lambda m
: unicode_escape(m
.group(0))[0],
2510 def escape_rfc3986(s
):
2511 """Escape non-ASCII characters as suggested by RFC 3986"""
2512 return urllib
.parse
.quote(s
, b
"%/;:@&=+$,!~*'()?#[]")
2515 def escape_url(url
):
2516 """Escape URL as suggested by RFC 3986"""
2517 url_parsed
= urllib
.parse
.urlparse(url
)
2518 return url_parsed
._replace
(
2519 netloc
=url_parsed
.netloc
.encode('idna').decode('ascii'),
2520 path
=escape_rfc3986(url_parsed
.path
),
2521 params
=escape_rfc3986(url_parsed
.params
),
2522 query
=escape_rfc3986(url_parsed
.query
),
2523 fragment
=escape_rfc3986(url_parsed
.fragment
)
2527 def parse_qs(url
, **kwargs
):
2528 return urllib
.parse
.parse_qs(urllib
.parse
.urlparse(url
).query
, **kwargs
)
2531 def read_batch_urls(batch_fd
):
2533 if not isinstance(url
, str):
2534 url
= url
.decode('utf-8', 'replace')
2535 BOM_UTF8
= ('\xef\xbb\xbf', '\ufeff')
2536 for bom
in BOM_UTF8
:
2537 if url
.startswith(bom
):
2538 url
= url
[len(bom
):]
2540 if not url
or url
.startswith(('#', ';', ']')):
2542 # "#" cannot be stripped out since it is part of the URI
2543 # However, it can be safely stripped out if following a whitespace
2544 return re
.split(r
'\s#', url
, 1)[0].rstrip()
2546 with contextlib
.closing(batch_fd
) as fd
:
2547 return [url
for url
in map(fixup
, fd
) if url
]
2550 def urlencode_postdata(*args
, **kargs
):
2551 return urllib
.parse
.urlencode(*args
, **kargs
).encode('ascii')
2554 def update_url(url
, *, query_update
=None, **kwargs
):
2555 """Replace URL components specified by kwargs
2556 @param url str or parse url tuple
2557 @param query_update update query
2560 if isinstance(url
, str):
2561 if not kwargs
and not query_update
:
2564 url
= urllib
.parse
.urlparse(url
)
2566 assert 'query' not in kwargs
, 'query_update and query cannot be specified at the same time'
2567 kwargs
['query'] = urllib
.parse
.urlencode({
2568 **urllib
.parse
.parse_qs(url
.query
),
2571 return urllib
.parse
.urlunparse(url
._replace
(**kwargs
))
2574 def update_url_query(url
, query
):
2575 return update_url(url
, query_update
=query
)
2578 def _multipart_encode_impl(data
, boundary
):
2579 content_type
= 'multipart/form-data; boundary=%s' % boundary
2582 for k
, v
in data
.items():
2583 out
+= b
'--' + boundary
.encode('ascii') + b
'\r\n'
2584 if isinstance(k
, str):
2586 if isinstance(v
, str):
2588 # RFC 2047 requires non-ASCII field names to be encoded, while RFC 7578
2589 # suggests sending UTF-8 directly. Firefox sends UTF-8, too
2590 content
= b
'Content-Disposition: form-data; name="' + k
+ b
'"\r\n\r\n' + v
+ b
'\r\n'
2591 if boundary
.encode('ascii') in content
:
2592 raise ValueError('Boundary overlaps with data')
2595 out
+= b
'--' + boundary
.encode('ascii') + b
'--\r\n'
2597 return out
, content_type
2600 def multipart_encode(data
, boundary
=None):
2602 Encode a dict to RFC 7578-compliant form-data
2605 A dict where keys and values can be either Unicode or bytes-like
2608 If specified a Unicode object, it's used as the boundary. Otherwise
2609 a random boundary is generated.
2611 Reference: https://tools.ietf.org/html/rfc7578
2613 has_specified_boundary
= boundary
is not None
2616 if boundary
is None:
2617 boundary
= '---------------' + str(random
.randrange(0x0fffffff, 0xffffffff))
2620 out
, content_type
= _multipart_encode_impl(data
, boundary
)
2623 if has_specified_boundary
:
2627 return out
, content_type
2630 def is_iterable_like(x
, allowed_types
=collections
.abc
.Iterable
, blocked_types
=NO_DEFAULT
):
2631 if blocked_types
is NO_DEFAULT
:
2632 blocked_types
= (str, bytes, collections
.abc
.Mapping
)
2633 return isinstance(x
, allowed_types
) and not isinstance(x
, blocked_types
)
2636 def variadic(x
, allowed_types
=NO_DEFAULT
):
2637 if not isinstance(allowed_types
, (tuple, type)):
2638 deprecation_warning('allowed_types should be a tuple or a type')
2639 allowed_types
= tuple(allowed_types
)
2640 return x
if is_iterable_like(x
, blocked_types
=allowed_types
) else (x
, )
2643 def try_call(*funcs
, expected_type
=None, args
=[], kwargs
={}):
2646 val
= f(*args
, **kwargs
)
2647 except (AttributeError, KeyError, TypeError, IndexError, ValueError, ZeroDivisionError):
2650 if expected_type
is None or isinstance(val
, expected_type
):
2654 def try_get(src
, getter
, expected_type
=None):
2655 return try_call(*variadic(getter
), args
=(src
,), expected_type
=expected_type
)
2658 def filter_dict(dct
, cndn
=lambda _
, v
: v
is not None):
2659 return {k: v for k, v in dct.items() if cndn(k, v)}
2662 def merge_dicts(*dicts
):
2664 for a_dict
in dicts
:
2665 for k
, v
in a_dict
.items():
2666 if (v
is not None and k
not in merged
2667 or isinstance(v
, str) and merged
[k
] == ''):
2672 def encode_compat_str(string
, encoding
=preferredencoding(), errors
='strict'):
2673 return string
if isinstance(string
, str) else str(string
, encoding
, errors
)
2685 TV_PARENTAL_GUIDELINES
= {
2695 def parse_age_limit(s
):
2696 # isinstance(False, int) is True. So type() must be used instead
2697 if type(s
) is int: # noqa: E721
2698 return s
if 0 <= s
<= 21 else None
2699 elif not isinstance(s
, str):
2701 m
= re
.match(r
'^(?P<age>\d{1,2})\+?$', s
)
2703 return int(m
.group('age'))
2706 return US_RATINGS
[s
]
2707 m
= re
.match(r
'^TV[_-]?(%s)$' % '|'.join(k
[3:] for k
in TV_PARENTAL_GUIDELINES
), s
)
2709 return TV_PARENTAL_GUIDELINES
['TV-' + m
.group(1)]
2713 def strip_jsonp(code
):
2716 (?:window\.)?(?P<func_name>[a-zA-Z0-9_.$]*)
2717 (?:\s*&&\s*(?P=func_name))?
2718 \s*\(\s*(?P<callback_data>.*)\);?
2719 \s*?(?://[^\n]*)*$''',
2720 r
'\g<callback_data>', code
)
2723 def js_to_json(code
, vars={}, *, strict
=False):
2724 # vars is a dict of var, val pairs to substitute
2725 STRING_QUOTES
= '\'"`'
2726 STRING_RE
= '|'.join(rf
'{q}(?:\\.|[^\\{q}])*{q}' for q
in STRING_QUOTES
)
2727 COMMENT_RE
= r
'/\*(?:(?!\*/).)*?\*/|//[^\n]*\n'
2728 SKIP_RE
= fr
'\s*(?:{COMMENT_RE})?\s*'
2730 (fr
'(?s)^(0[xX][0-9a-fA-F]+){SKIP_RE}:?$', 16),
2731 (fr
'(?s)^(0+[0-7]+){SKIP_RE}:?$', 8),
2734 def process_escape(match
):
2735 JSON_PASSTHROUGH_ESCAPES
= R
'"\bfnrtu'
2736 escape
= match
.group(1) or match
.group(2)
2738 return (Rf
'\{escape}' if escape
in JSON_PASSTHROUGH_ESCAPES
2739 else R
'\u00' if escape
== 'x'
2740 else '' if escape
== '\n'
2743 def template_substitute(match
):
2744 evaluated
= js_to_json(match
.group(1), vars, strict
=strict
)
2745 if evaluated
[0] == '"':
2746 return json
.loads(evaluated
)
2751 if v
in ('true', 'false', 'null'):
2753 elif v
in ('undefined', 'void 0'):
2755 elif v
.startswith('/*') or v
.startswith('//') or v
.startswith('!') or v
== ',':
2758 if v
[0] in STRING_QUOTES
:
2759 v
= re
.sub(r
'(?s)\${([^}]+)}', template_substitute
, v
[1:-1]) if v
[0] == '`' else v
[1:-1]
2760 escaped
= re
.sub(r
'(?s)(")|\\(.)', process_escape
, v
)
2761 return f
'"{escaped}"'
2763 for regex
, base
in INTEGER_TABLE
:
2764 im
= re
.match(regex
, v
)
2766 i
= int(im
.group(1), base
)
2767 return f
'"{i}":' if v
.endswith(':') else str(i
)
2773 except json
.JSONDecodeError
:
2774 return json
.dumps(vars[v
])
2781 raise ValueError(f
'Unknown value: {v}')
2783 def create_map(mobj
):
2784 return json
.dumps(dict(json
.loads(js_to_json(mobj
.group(1) or '[]', vars=vars))))
2786 code
= re
.sub(r
'new Map\((\[.*?\])?\)', create_map
, code
)
2788 code
= re
.sub(r
'new Date\((".+")\)', r
'\g<1>', code
)
2789 code
= re
.sub(r
'new \w+\((.*?)\)', lambda m
: json
.dumps(m
.group(0)), code
)
2790 code
= re
.sub(r
'parseInt\([^\d]+(\d+)[^\d]+\)', r
'\1', code
)
2791 code
= re
.sub(r
'\(function\([^)]*\)\s*\{[^}]*\}\s*\)\s*\(\s*(["\'][^
)]*["\'])\s*\)', r'\1', code)
2793 return re.sub(rf'''(?sx)
2795 {COMMENT_RE}|,(?={SKIP_RE}[\]}}])|
2796 void\s0|(?:(?<![0-9])[eE]|[a-df-zA-DF-Z_$])[.a-zA-Z_$0-9]*|
2797 \b(?:0[xX][0-9a-fA-F]+|0+[0-7]+)(?:{SKIP_RE}:)?|
2798 [0-9]+(?={SKIP_RE}:)|
2803 def qualities(quality_ids):
2804 """ Get a numeric quality value out of a list of possible values """
2807 return quality_ids.index(qid)
2813 POSTPROCESS_WHEN = ('pre_process', 'after_filter', 'video', 'before_dl', 'post_process', 'after_move', 'after_video', 'playlist')
2817 'default': '%(title)s [%(id)s].%(ext)s',
2818 'chapter': '%(title)s - %(section_number)03d %(section_title)s [%(id)s].%(ext)s',
2824 'description': 'description',
2825 'annotation': 'annotations.xml',
2826 'infojson': 'info.json',
2829 'pl_thumbnail': None,
2830 'pl_description': 'description',
2831 'pl_infojson': 'info.json',
2834 # As of [1] format syntax is:
2835 # %[mapping_key][conversion_flags][minimum_width][.precision][length_modifier]type
2836 # 1. https://docs.python.org/2/library/stdtypes.html#string-formatting
2837 STR_FORMAT_RE_TMPL = r'''(?x)
2838 (?<!%)(?P<prefix>(?:%%)*)
2840 (?P<has_key>\((?P<key>{0})\))?
2842 (?P<conversion>[#0\-+ ]+)?
2844 (?P<precision>\.\d+)?
2845 (?P<len_mod>[hlL])? # unused in python
2846 {1} # conversion type
2851 STR_FORMAT_TYPES = 'diouxXeEfFgGcrsa'
2854 def limit_length(s, length):
2855 """ Add ellipses to overly long strings """
2860 return s[:length - len(ELLIPSES)] + ELLIPSES
2864 def version_tuple(v):
2865 return tuple(int(e) for e in re.split(r'[-.]', v))
2868 def is_outdated_version(version, limit, assume_new=True):
2870 return not assume_new
2872 return version_tuple(version) < version_tuple(limit)
2874 return not assume_new
2877 def ytdl_is_updateable():
2878 """ Returns if yt-dlp can be updated with -U """
2880 from ..update import is_non_updateable
2882 return not is_non_updateable()
2885 def args_to_str(args):
2886 # Get a short string representation for a subprocess command
2887 return ' '.join(compat_shlex_quote(a) for a in args)
2890 def error_to_str(err):
2891 return f'{type(err).__name__}: {err}'
2894 def mimetype2ext(mt, default=NO_DEFAULT):
2895 if not isinstance(mt, str):
2896 if default is not NO_DEFAULT:
2912 'x-matroska': 'mkv',
2914 'x-mp4-fragmented': 'mp4',
2919 # application (streaming playlists)
2923 'vnd.apple.mpegurl': 'm3u8',
2924 'vnd.ms-sstr+xml': 'ism',
2925 'x-mpegurl': 'm3u8',
2929 # Per RFC 3003, audio/mpeg can be .mp1, .mp2 or .mp3.
2930 # Using .mp3 as it's the most popular one
2931 'audio/mpeg': 'mp3',
2932 'audio/webm': 'webm',
2933 'audio/x-matroska': 'mka',
2934 'audio/x-mpegurl': 'm3u',
2942 'x-realaudio': 'ra',
2953 'vnd.wap.wbmp': 'wbmp',
2960 'filmstrip+json': 'fs',
2961 'smptett+xml': 'tt',
2964 'x-ms-sami': 'sami',
2973 mimetype = mt.partition(';')[0].strip().lower()
2974 _, _, subtype = mimetype.rpartition('/')
2976 ext = traversal.traverse_obj(MAP, mimetype, subtype, subtype.rsplit('+')[-1])
2979 elif default is not NO_DEFAULT:
2981 return subtype.replace('+', '.')
2984 def ext2mimetype(ext_or_url):
2987 if '.' not in ext_or_url:
2988 ext_or_url = f'file.{ext_or_url}'
2989 return mimetypes.guess_type(ext_or_url)[0]
2992 def parse_codecs(codecs_str):
2993 # http://tools.ietf.org/html/rfc6381
2996 split_codecs = list(filter(None, map(
2997 str.strip, codecs_str.strip().strip(',').split(','))))
2998 vcodec, acodec, scodec, hdr = None, None, None, None
2999 for full_codec in split_codecs:
3000 parts = re.sub(r'0+(?=\d)', '', full_codec).split('.')
3001 if parts[0] in ('avc1', 'avc2', 'avc3', 'avc4', 'vp9', 'vp8', 'hev1', 'hev2',
3002 'h263', 'h264', 'mp4v', 'hvc1', 'av1', 'theora', 'dvh1', 'dvhe'):
3006 if parts[0] in ('dvh1', 'dvhe'):
3008 elif parts[0] == 'av1' and traversal.traverse_obj(parts, 3) == '10':
3010 elif parts[:2] == ['vp9', '2']:
3012 elif parts[0] in ('flac', 'mp4a', 'opus', 'vorbis', 'mp3', 'aac', 'ac-4',
3013 'ac-3', 'ec-3', 'eac3', 'dtsc', 'dtse', 'dtsh', 'dtsl'):
3014 acodec = acodec or full_codec
3015 elif parts[0] in ('stpp', 'wvtt'):
3016 scodec = scodec or full_codec
3018 write_string(f'WARNING: Unknown codec {full_codec}\n')
3019 if vcodec or acodec or scodec:
3021 'vcodec': vcodec or 'none',
3022 'acodec': acodec or 'none',
3023 'dynamic_range': hdr,
3024 **({'scodec': scodec} if scodec is not None else {}),
3026 elif len(split_codecs) == 2:
3028 'vcodec': split_codecs[0],
3029 'acodec': split_codecs[1],
3034 def get_compatible_ext(*, vcodecs, acodecs, vexts, aexts, preferences=None):
3035 assert len(vcodecs) == len(vexts) and len(acodecs) == len(aexts)
3037 allow_mkv = not preferences or 'mkv' in preferences
3039 if allow_mkv and max(len(acodecs), len(vcodecs)) > 1:
3040 return 'mkv' # TODO: any other format allows this?
3042 # TODO: All codecs supported by parse_codecs isn't handled here
3043 COMPATIBLE_CODECS = {
3045 'av1', 'hevc', 'avc1', 'mp4a', 'ac-4', # fourcc (m3u8, mpd)
3046 'h264', 'aacl', 'ec-3', # Set in ISM
3049 'av1', 'vp9', 'vp8', 'opus', 'vrbs',
3050 'vp9x', 'vp8x', # in the webm spec
3054 sanitize_codec = functools.partial(
3055 try_get, getter=lambda x: x[0].split('.')[0].replace('0', '').lower())
3056 vcodec, acodec = sanitize_codec(vcodecs), sanitize_codec(acodecs)
3058 for ext in preferences or COMPATIBLE_CODECS.keys():
3059 codec_set = COMPATIBLE_CODECS.get(ext, set())
3060 if ext == 'mkv' or codec_set.issuperset((vcodec, acodec)):
3064 {'mp3', 'mp4', 'm4a', 'm4p', 'm4b', 'm4r', 'm4v', 'ismv', 'isma', 'mov'},
3067 for ext in preferences or vexts:
3068 current_exts = {ext, *vexts, *aexts}
3069 if ext == 'mkv' or current_exts == {ext} or any(
3070 ext_sets.issuperset(current_exts) for ext_sets in COMPATIBLE_EXTS):
3072 return 'mkv' if allow_mkv else preferences[-1]
3075 def urlhandle_detect_ext(url_handle, default=NO_DEFAULT):
3076 getheader = url_handle.headers.get
3078 cd = getheader('Content-Disposition')
3080 m = re.match(r'attachment;\s*filename="(?P
<filename
>[^
"]+)"', cd)
3082 e = determine_ext(m.group('filename
'), default_ext=None)
3086 meta_ext = getheader('x
-amz
-meta
-name
')
3088 e = meta_ext.rpartition('.')[2]
3092 return mimetype2ext(getheader('Content
-Type
'), default=default)
3095 def encode_data_uri(data, mime_type):
3096 return 'data
:%s;base64
,%s' % (mime_type, base64.b64encode(data).decode('ascii
'))
3099 def age_restricted(content_limit, age_limit):
3100 """ Returns True iff the content should be blocked """
3102 if age_limit is None: # No limit set
3104 if content_limit is None:
3105 return False # Content available for everyone
3106 return age_limit < content_limit
3109 # List of known byte-order-marks (BOM)
3111 (b'\xef\xbb\xbf', 'utf
-8'),
3112 (b'\x00\x00\xfe\xff', 'utf
-32-be
'),
3113 (b'\xff\xfe\x00\x00', 'utf
-32-le
'),
3114 (b'\xff\xfe', 'utf
-16-le
'),
3115 (b'\xfe\xff', 'utf
-16-be
'),
3119 def is_html(first_bytes):
3120 """ Detect whether a file contains HTML by examining its first bytes. """
3123 for bom, enc in BOMS:
3124 while first_bytes.startswith(bom):
3125 encoding, first_bytes = enc, first_bytes[len(bom):]
3127 return re.match(r'^\s
*<', first_bytes.decode(encoding, 'replace
'))
3130 def determine_protocol(info_dict):
3131 protocol = info_dict.get('protocol
')
3132 if protocol is not None:
3135 url = sanitize_url(info_dict['url
'])
3136 if url.startswith('rtmp
'):
3138 elif url.startswith('mms
'):
3140 elif url.startswith('rtsp
'):
3143 ext = determine_ext(url)
3145 return 'm3u8
' if info_dict.get('is_live
') else 'm3u8_native
'
3149 return urllib.parse.urlparse(url).scheme
3152 def render_table(header_row, data, delim=False, extra_gap=0, hide_empty=False):
3153 """ Render a list of rows, each as a list of values.
3154 Text after a \t will be right aligned """
3156 return len(remove_terminal_sequences(string).replace('\t', ''))
3158 def get_max_lens(table):
3159 return [max(width(str(v)) for v in col) for col in zip(*table)]
3161 def filter_using_list(row, filterArray):
3162 return [col for take, col in itertools.zip_longest(filterArray, row, fillvalue=True) if take]
3164 max_lens = get_max_lens(data) if hide_empty else []
3165 header_row = filter_using_list(header_row, max_lens)
3166 data = [filter_using_list(row, max_lens) for row in data]
3168 table = [header_row] + data
3169 max_lens = get_max_lens(table)
3172 table = [header_row, [delim * (ml + extra_gap) for ml in max_lens]] + data
3173 table[1][-1] = table[1][-1][:-extra_gap * len(delim)] # Remove extra_gap from end of delimiter
3175 for pos, text in enumerate(map(str, row)):
3177 row[pos] = text.replace('\t', ' ' * (max_lens[pos] - width(text))) + ' ' * extra_gap
3179 row[pos] = text + ' ' * (max_lens[pos] - width(text) + extra_gap)
3180 ret = '\n'.join(''.join(row).rstrip() for row in table)
3184 def _match_one(filter_part, dct, incomplete):
3185 # TODO: Generalize code with YoutubeDL._build_format_filter
3186 STRING_OPERATORS = {
3187 '*=': operator.contains,
3188 '^
=': lambda attr, value: attr.startswith(value),
3189 '$
=': lambda attr, value: attr.endswith(value),
3190 '~
=': lambda attr, value: re.search(value, attr),
3192 COMPARISON_OPERATORS = {
3194 '<=': operator.le, # "<=" must be defined above "<"
3201 if isinstance(incomplete, bool):
3202 is_incomplete = lambda _: incomplete
3204 is_incomplete = lambda k: k in incomplete
3206 operator_rex = re.compile(r'''(?x)
3208 \s*(?P<negation>!\s*)?(?P<op>%s)(?P<none_inclusive>\s*\?)?\s*
3210 (?P<quote>["\'])(?P<quotedstrval>.+?)(?P=quote)|
3213 ''' % '|
'.join(map(re.escape, COMPARISON_OPERATORS.keys())))
3214 m = operator_rex.fullmatch(filter_part.strip())
3217 unnegated_op = COMPARISON_OPERATORS[m['op
']]
3219 op = lambda attr, value: not unnegated_op(attr, value)
3222 comparison_value = m['quotedstrval
'] or m['strval
'] or m['intval
']
3224 comparison_value = comparison_value.replace(r'\
%s' % m['quote
'], m['quote
'])
3225 actual_value = dct.get(m['key
'])
3226 numeric_comparison = None
3227 if isinstance(actual_value, (int, float)):
3228 # If the original field is a string and matching comparisonvalue is
3229 # a number we should respect the origin of the original field
3230 # and process comparison value as a string (see
3231 # https://github.com/ytdl-org/youtube-dl/issues/11082)
3233 numeric_comparison = int(comparison_value)
3235 numeric_comparison = parse_filesize(comparison_value)
3236 if numeric_comparison is None:
3237 numeric_comparison = parse_filesize(f'{comparison_value}B
')
3238 if numeric_comparison is None:
3239 numeric_comparison = parse_duration(comparison_value)
3240 if numeric_comparison is not None and m['op
'] in STRING_OPERATORS:
3241 raise ValueError('Operator
%s only supports string values
!' % m['op
'])
3242 if actual_value is None:
3243 return is_incomplete(m['key
']) or m['none_inclusive
']
3244 return op(actual_value, comparison_value if numeric_comparison is None else numeric_comparison)
3247 '': lambda v: (v is True) if isinstance(v, bool) else (v is not None),
3248 '!': lambda v: (v is False) if isinstance(v, bool) else (v is None),
3250 operator_rex = re.compile(r'''(?x)
3251 (?P<op>%s)\s*(?P<key>[a-z_]+)
3252 ''' % '|
'.join(map(re.escape, UNARY_OPERATORS.keys())))
3253 m = operator_rex.fullmatch(filter_part.strip())
3255 op = UNARY_OPERATORS[m.group('op
')]
3256 actual_value = dct.get(m.group('key
'))
3257 if is_incomplete(m.group('key
')) and actual_value is None:
3259 return op(actual_value)
3261 raise ValueError('Invalid
filter part
%r' % filter_part)
3264 def match_str(filter_str, dct, incomplete=False):
3265 """ Filter a dictionary with a simple string syntax.
3266 @returns Whether the filter passes
3267 @param incomplete Set of keys that is expected to be missing from dct.
3268 Can be True/False to indicate all/none of the keys may be missing.
3269 All conditions on incomplete keys pass if the key is missing
3272 _match_one(filter_part.replace(r'\
&', '&'), dct, incomplete)
3273 for filter_part in re.split(r'(?
<!\\)&', filter_str))
3276 def match_filter_func(filters, breaking_filters=None):
3277 if not filters and not breaking_filters:
3279 breaking_filters = match_filter_func(breaking_filters) or (lambda _, __: None)
3280 filters = set(variadic(filters or []))
3282 interactive = '-' in filters
3286 def _match_func(info_dict, incomplete=False):
3287 ret = breaking_filters(info_dict, incomplete)
3289 raise RejectedVideoReached(ret)
3291 if not filters or any(match_str(f, info_dict, incomplete) for f in filters):
3292 return NO_DEFAULT if interactive and not incomplete else None
3294 video_title = info_dict.get('title
') or info_dict.get('id') or 'entry
'
3295 filter_str = ') |
('.join(map(str.strip, filters))
3296 return f'{video_title} does
not pass filter ({filter_str}
), skipping
..'
3300 class download_range_func:
3301 def __init__(self, chapters, ranges, from_info=False):
3302 self.chapters, self.ranges, self.from_info = chapters, ranges, from_info
3304 def __call__(self, info_dict, ydl):
3306 warning = ('There are no chapters matching the regex
' if info_dict.get('chapters
')
3307 else 'Cannot match chapters since chapter information
is unavailable
')
3308 for regex in self.chapters or []:
3309 for i, chapter in enumerate(info_dict.get('chapters
') or []):
3310 if re.search(regex, chapter['title
']):
3312 yield {**chapter, 'index': i}
3313 if self.chapters and warning:
3314 ydl.to_screen(f'[info
] {info_dict["id"]}
: {warning}
')
3316 for start, end in self.ranges or []:
3318 'start_time
': self._handle_negative_timestamp(start, info_dict),
3319 'end_time
': self._handle_negative_timestamp(end, info_dict),
3322 if self.from_info and (info_dict.get('start_time
') or info_dict.get('end_time
')):
3324 'start_time
': info_dict.get('start_time
') or 0,
3325 'end_time
': info_dict.get('end_time
') or float('inf
'),
3327 elif not self.ranges and not self.chapters:
3331 def _handle_negative_timestamp(time, info):
3332 return max(info['duration
'] + time, 0) if info.get('duration
') and time < 0 else time
3334 def __eq__(self, other):
3335 return (isinstance(other, download_range_func)
3336 and self.chapters == other.chapters and self.ranges == other.ranges)
3339 return f'{__name__}
.{type(self).__name__}
({self.chapters}
, {self.ranges}
)'
3342 def parse_dfxp_time_expr(time_expr):
3346 mobj = re.match(rf'^
(?P
<time_offset
>{NUMBER_RE}
)s?$
', time_expr)
3348 return float(mobj.group('time_offset
'))
3350 mobj = re.match(r'^
(\d
+):(\d\d
):(\d\
d(?
:(?
:\
.|
:)\d
+)?
)$
', time_expr)
3352 return 3600 * int(mobj.group(1)) + 60 * int(mobj.group(2)) + float(mobj.group(3).replace(':', '.'))
3355 def srt_subtitles_timecode(seconds):
3356 return '%02d
:%02d
:%02d
,%03d
' % timetuple_from_msec(seconds * 1000)
3359 def ass_subtitles_timecode(seconds):
3360 time = timetuple_from_msec(seconds * 1000)
3361 return '%01d
:%02d
:%02d
.%02d
' % (*time[:-1], time.milliseconds / 10)
3364 def dfxp2srt(dfxp_data):
3366 @param dfxp_data A bytes-like object containing DFXP data
3367 @returns A unicode object containing converted SRT data
3369 LEGACY_NAMESPACES = (
3370 (b'http
://www
.w3
.org
/ns
/ttml
', [
3371 b'http
://www
.w3
.org
/2004/11/ttaf1
',
3372 b'http
://www
.w3
.org
/2006/04/ttaf1
',
3373 b'http
://www
.w3
.org
/2006/10/ttaf1
',
3375 (b'http
://www
.w3
.org
/ns
/ttml
#styling', [
3376 b
'http://www.w3.org/ns/ttml#style',
3380 SUPPORTED_STYLING
= [
3389 _x
= functools
.partial(xpath_with_ns
, ns_map
={
3390 'xml': 'http://www.w3.org/XML/1998/namespace',
3391 'ttml': 'http://www.w3.org/ns/ttml',
3392 'tts': 'http://www.w3.org/ns/ttml#styling',
3398 class TTMLPElementParser
:
3400 _unclosed_elements
= []
3401 _applied_styles
= []
3403 def start(self
, tag
, attrib
):
3404 if tag
in (_x('ttml:br'), 'br'):
3407 unclosed_elements
= []
3409 element_style_id
= attrib
.get('style')
3411 style
.update(default_style
)
3412 if element_style_id
:
3413 style
.update(styles
.get(element_style_id
, {}))
3414 for prop
in SUPPORTED_STYLING
:
3415 prop_val
= attrib
.get(_x('tts:' + prop
))
3417 style
[prop
] = prop_val
3420 for k
, v
in sorted(style
.items()):
3421 if self
._applied
_styles
and self
._applied
_styles
[-1].get(k
) == v
:
3424 font
+= ' color="%s"' % v
3425 elif k
== 'fontSize':
3426 font
+= ' size="%s"' % v
3427 elif k
== 'fontFamily':
3428 font
+= ' face="%s"' % v
3429 elif k
== 'fontWeight' and v
== 'bold':
3431 unclosed_elements
.append('b')
3432 elif k
== 'fontStyle' and v
== 'italic':
3434 unclosed_elements
.append('i')
3435 elif k
== 'textDecoration' and v
== 'underline':
3437 unclosed_elements
.append('u')
3439 self
._out
+= '<font' + font
+ '>'
3440 unclosed_elements
.append('font')
3442 if self
._applied
_styles
:
3443 applied_style
.update(self
._applied
_styles
[-1])
3444 applied_style
.update(style
)
3445 self
._applied
_styles
.append(applied_style
)
3446 self
._unclosed
_elements
.append(unclosed_elements
)
3449 if tag
not in (_x('ttml:br'), 'br'):
3450 unclosed_elements
= self
._unclosed
_elements
.pop()
3451 for element
in reversed(unclosed_elements
):
3452 self
._out
+= '</%s>' % element
3453 if unclosed_elements
and self
._applied
_styles
:
3454 self
._applied
_styles
.pop()
3456 def data(self
, data
):
3460 return self
._out
.strip()
3462 # Fix UTF-8 encoded file wrongly marked as UTF-16. See https://github.com/yt-dlp/yt-dlp/issues/6543#issuecomment-1477169870
3463 # This will not trigger false positives since only UTF-8 text is being replaced
3464 dfxp_data
= dfxp_data
.replace(b
'encoding=\'UTF-16\'', b
'encoding=\'UTF-8\'')
3466 def parse_node(node
):
3467 target
= TTMLPElementParser()
3468 parser
= xml
.etree
.ElementTree
.XMLParser(target
=target
)
3469 parser
.feed(xml
.etree
.ElementTree
.tostring(node
))
3470 return parser
.close()
3472 for k
, v
in LEGACY_NAMESPACES
:
3474 dfxp_data
= dfxp_data
.replace(ns
, k
)
3476 dfxp
= compat_etree_fromstring(dfxp_data
)
3478 paras
= dfxp
.findall(_x('.//ttml:p')) or dfxp
.findall('.//p')
3481 raise ValueError('Invalid dfxp/TTML subtitle')
3485 for style
in dfxp
.findall(_x('.//ttml:style')):
3486 style_id
= style
.get('id') or style
.get(_x('xml:id'))
3489 parent_style_id
= style
.get('style')
3491 if parent_style_id
not in styles
:
3494 styles
[style_id
] = styles
[parent_style_id
].copy()
3495 for prop
in SUPPORTED_STYLING
:
3496 prop_val
= style
.get(_x('tts:' + prop
))
3498 styles
.setdefault(style_id
, {})[prop
] = prop_val
3504 for p
in ('body', 'div'):
3505 ele
= xpath_element(dfxp
, [_x('.//ttml:' + p
), './/' + p
])
3508 style
= styles
.get(ele
.get('style'))
3511 default_style
.update(style
)
3513 for para
, index
in zip(paras
, itertools
.count(1)):
3514 begin_time
= parse_dfxp_time_expr(para
.attrib
.get('begin'))
3515 end_time
= parse_dfxp_time_expr(para
.attrib
.get('end'))
3516 dur
= parse_dfxp_time_expr(para
.attrib
.get('dur'))
3517 if begin_time
is None:
3522 end_time
= begin_time
+ dur
3523 out
.append('%d\n%s --> %s\n%s\n\n' % (
3525 srt_subtitles_timecode(begin_time
),
3526 srt_subtitles_timecode(end_time
),
3532 def cli_option(params
, command_option
, param
, separator
=None):
3533 param
= params
.get(param
)
3534 return ([] if param
is None
3535 else [command_option
, str(param
)] if separator
is None
3536 else [f
'{command_option}{separator}{param}'])
3539 def cli_bool_option(params
, command_option
, param
, true_value
='true', false_value
='false', separator
=None):
3540 param
= params
.get(param
)
3541 assert param
in (True, False, None)
3542 return cli_option({True: true_value, False: false_value}
, command_option
, param
, separator
)
3545 def cli_valueless_option(params
, command_option
, param
, expected_value
=True):
3546 return [command_option
] if params
.get(param
) == expected_value
else []
3549 def cli_configuration_args(argdict
, keys
, default
=[], use_compat
=True):
3550 if isinstance(argdict
, (list, tuple)): # for backward compatibility
3557 assert isinstance(argdict
, dict)
3559 assert isinstance(keys
, (list, tuple))
3560 for key_list
in keys
:
3561 arg_list
= list(filter(
3562 lambda x
: x
is not None,
3563 [argdict
.get(key
.lower()) for key
in variadic(key_list
)]))
3565 return [arg
for args
in arg_list
for arg
in args
]
3569 def _configuration_args(main_key
, argdict
, exe
, keys
=None, default
=[], use_compat
=True):
3570 main_key
, exe
= main_key
.lower(), exe
.lower()
3571 root_key
= exe
if main_key
== exe
else f
'{main_key}+{exe}'
3572 keys
= [f
'{root_key}{k}' for k
in (keys
or [''])]
3573 if root_key
in keys
:
3575 keys
.append((main_key
, exe
))
3576 keys
.append('default')
3579 return cli_configuration_args(argdict
, keys
, default
, use_compat
)
3583 # See http://www.loc.gov/standards/iso639-2/ISO-639-2_utf-8.txt
3642 'iw': 'heb', # Replaced by he in 1989 revision
3652 'in': 'ind', # Replaced by id in 1989 revision
3768 'ji': 'yid', # Replaced by yi in 1989 revision
3776 def short2long(cls
, code
):
3777 """Convert language code from ISO 639-1 to ISO 639-2/T"""
3778 return cls
._lang
_map
.get(code
[:2])
3781 def long2short(cls
, code
):
3782 """Convert language code from ISO 639-2/T to ISO 639-1"""
3783 for short_name
, long_name
in cls
._lang
_map
.items():
3784 if long_name
== code
:
3789 # From http://data.okfn.org/data/core/country-list
3791 'AF': 'Afghanistan',
3792 'AX': 'Åland Islands',
3795 'AS': 'American Samoa',
3800 'AG': 'Antigua and Barbuda',
3817 'BO': 'Bolivia, Plurinational State of',
3818 'BQ': 'Bonaire, Sint Eustatius and Saba',
3819 'BA': 'Bosnia and Herzegovina',
3821 'BV': 'Bouvet Island',
3823 'IO': 'British Indian Ocean Territory',
3824 'BN': 'Brunei Darussalam',
3826 'BF': 'Burkina Faso',
3832 'KY': 'Cayman Islands',
3833 'CF': 'Central African Republic',
3837 'CX': 'Christmas Island',
3838 'CC': 'Cocos (Keeling) Islands',
3842 'CD': 'Congo, the Democratic Republic of the',
3843 'CK': 'Cook Islands',
3845 'CI': 'Côte d\'Ivoire',
3850 'CZ': 'Czech Republic',
3854 'DO': 'Dominican Republic',
3857 'SV': 'El Salvador',
3858 'GQ': 'Equatorial Guinea',
3862 'FK': 'Falkland Islands (Malvinas)',
3863 'FO': 'Faroe Islands',
3867 'GF': 'French Guiana',
3868 'PF': 'French Polynesia',
3869 'TF': 'French Southern Territories',
3884 'GW': 'Guinea-Bissau',
3887 'HM': 'Heard Island and McDonald Islands',
3888 'VA': 'Holy See (Vatican City State)',
3895 'IR': 'Iran, Islamic Republic of',
3898 'IM': 'Isle of Man',
3908 'KP': 'Korea, Democratic People\'s Republic of',
3909 'KR': 'Korea, Republic of',
3912 'LA': 'Lao People\'s Democratic Republic',
3918 'LI': 'Liechtenstein',
3922 'MK': 'Macedonia, the Former Yugoslav Republic of',
3929 'MH': 'Marshall Islands',
3935 'FM': 'Micronesia, Federated States of',
3936 'MD': 'Moldova, Republic of',
3947 'NL': 'Netherlands',
3948 'NC': 'New Caledonia',
3949 'NZ': 'New Zealand',
3954 'NF': 'Norfolk Island',
3955 'MP': 'Northern Mariana Islands',
3960 'PS': 'Palestine, State of',
3962 'PG': 'Papua New Guinea',
3965 'PH': 'Philippines',
3969 'PR': 'Puerto Rico',
3973 'RU': 'Russian Federation',
3975 'BL': 'Saint Barthélemy',
3976 'SH': 'Saint Helena, Ascension and Tristan da Cunha',
3977 'KN': 'Saint Kitts and Nevis',
3978 'LC': 'Saint Lucia',
3979 'MF': 'Saint Martin (French part)',
3980 'PM': 'Saint Pierre and Miquelon',
3981 'VC': 'Saint Vincent and the Grenadines',
3984 'ST': 'Sao Tome and Principe',
3985 'SA': 'Saudi Arabia',
3989 'SL': 'Sierra Leone',
3991 'SX': 'Sint Maarten (Dutch part)',
3994 'SB': 'Solomon Islands',
3996 'ZA': 'South Africa',
3997 'GS': 'South Georgia and the South Sandwich Islands',
3998 'SS': 'South Sudan',
4003 'SJ': 'Svalbard and Jan Mayen',
4006 'CH': 'Switzerland',
4007 'SY': 'Syrian Arab Republic',
4008 'TW': 'Taiwan, Province of China',
4010 'TZ': 'Tanzania, United Republic of',
4012 'TL': 'Timor-Leste',
4016 'TT': 'Trinidad and Tobago',
4019 'TM': 'Turkmenistan',
4020 'TC': 'Turks and Caicos Islands',
4024 'AE': 'United Arab Emirates',
4025 'GB': 'United Kingdom',
4026 'US': 'United States',
4027 'UM': 'United States Minor Outlying Islands',
4031 'VE': 'Venezuela, Bolivarian Republic of',
4033 'VG': 'Virgin Islands, British',
4034 'VI': 'Virgin Islands, U.S.',
4035 'WF': 'Wallis and Futuna',
4036 'EH': 'Western Sahara',
4040 # Not ISO 3166 codes, but used for IP blocks
4041 'AP': 'Asia/Pacific Region',
4046 def short2full(cls
, code
):
4047 """Convert an ISO 3166-2 country code to the corresponding full name"""
4048 return cls
._country
_map
.get(code
.upper())
4052 # Major IPv4 address blocks per country
4054 'AD': '46.172.224.0/19',
4055 'AE': '94.200.0.0/13',
4056 'AF': '149.54.0.0/17',
4057 'AG': '209.59.64.0/18',
4058 'AI': '204.14.248.0/21',
4059 'AL': '46.99.0.0/16',
4060 'AM': '46.70.0.0/15',
4061 'AO': '105.168.0.0/13',
4062 'AP': '182.50.184.0/21',
4063 'AQ': '23.154.160.0/24',
4064 'AR': '181.0.0.0/12',
4065 'AS': '202.70.112.0/20',
4066 'AT': '77.116.0.0/14',
4067 'AU': '1.128.0.0/11',
4068 'AW': '181.41.0.0/18',
4069 'AX': '185.217.4.0/22',
4070 'AZ': '5.197.0.0/16',
4071 'BA': '31.176.128.0/17',
4072 'BB': '65.48.128.0/17',
4073 'BD': '114.130.0.0/16',
4075 'BF': '102.178.0.0/15',
4076 'BG': '95.42.0.0/15',
4077 'BH': '37.131.0.0/17',
4078 'BI': '154.117.192.0/18',
4079 'BJ': '137.255.0.0/16',
4080 'BL': '185.212.72.0/23',
4081 'BM': '196.12.64.0/18',
4082 'BN': '156.31.0.0/16',
4083 'BO': '161.56.0.0/16',
4084 'BQ': '161.0.80.0/20',
4085 'BR': '191.128.0.0/12',
4086 'BS': '24.51.64.0/18',
4087 'BT': '119.2.96.0/19',
4088 'BW': '168.167.0.0/16',
4089 'BY': '178.120.0.0/13',
4090 'BZ': '179.42.192.0/18',
4091 'CA': '99.224.0.0/11',
4092 'CD': '41.243.0.0/16',
4093 'CF': '197.242.176.0/21',
4094 'CG': '160.113.0.0/16',
4095 'CH': '85.0.0.0/13',
4096 'CI': '102.136.0.0/14',
4097 'CK': '202.65.32.0/19',
4098 'CL': '152.172.0.0/14',
4099 'CM': '102.244.0.0/14',
4100 'CN': '36.128.0.0/10',
4101 'CO': '181.240.0.0/12',
4102 'CR': '201.192.0.0/12',
4103 'CU': '152.206.0.0/15',
4104 'CV': '165.90.96.0/19',
4105 'CW': '190.88.128.0/17',
4106 'CY': '31.153.0.0/16',
4107 'CZ': '88.100.0.0/14',
4109 'DJ': '197.241.0.0/17',
4110 'DK': '87.48.0.0/12',
4111 'DM': '192.243.48.0/20',
4112 'DO': '152.166.0.0/15',
4113 'DZ': '41.96.0.0/12',
4114 'EC': '186.68.0.0/15',
4115 'EE': '90.190.0.0/15',
4116 'EG': '156.160.0.0/11',
4117 'ER': '196.200.96.0/20',
4118 'ES': '88.0.0.0/11',
4119 'ET': '196.188.0.0/14',
4120 'EU': '2.16.0.0/13',
4121 'FI': '91.152.0.0/13',
4122 'FJ': '144.120.0.0/16',
4123 'FK': '80.73.208.0/21',
4124 'FM': '119.252.112.0/20',
4125 'FO': '88.85.32.0/19',
4127 'GA': '41.158.0.0/15',
4129 'GD': '74.122.88.0/21',
4130 'GE': '31.146.0.0/16',
4131 'GF': '161.22.64.0/18',
4132 'GG': '62.68.160.0/19',
4133 'GH': '154.160.0.0/12',
4134 'GI': '95.164.0.0/16',
4135 'GL': '88.83.0.0/19',
4136 'GM': '160.182.0.0/15',
4137 'GN': '197.149.192.0/18',
4138 'GP': '104.250.0.0/19',
4139 'GQ': '105.235.224.0/20',
4140 'GR': '94.64.0.0/13',
4141 'GT': '168.234.0.0/16',
4142 'GU': '168.123.0.0/16',
4143 'GW': '197.214.80.0/20',
4144 'GY': '181.41.64.0/18',
4145 'HK': '113.252.0.0/14',
4146 'HN': '181.210.0.0/16',
4147 'HR': '93.136.0.0/13',
4148 'HT': '148.102.128.0/17',
4149 'HU': '84.0.0.0/14',
4150 'ID': '39.192.0.0/10',
4151 'IE': '87.32.0.0/12',
4152 'IL': '79.176.0.0/13',
4153 'IM': '5.62.80.0/20',
4154 'IN': '117.192.0.0/10',
4155 'IO': '203.83.48.0/21',
4156 'IQ': '37.236.0.0/14',
4157 'IR': '2.176.0.0/12',
4158 'IS': '82.221.0.0/16',
4159 'IT': '79.0.0.0/10',
4160 'JE': '87.244.64.0/18',
4161 'JM': '72.27.0.0/17',
4162 'JO': '176.29.0.0/16',
4163 'JP': '133.0.0.0/8',
4164 'KE': '105.48.0.0/12',
4165 'KG': '158.181.128.0/17',
4166 'KH': '36.37.128.0/17',
4167 'KI': '103.25.140.0/22',
4168 'KM': '197.255.224.0/20',
4169 'KN': '198.167.192.0/19',
4170 'KP': '175.45.176.0/22',
4171 'KR': '175.192.0.0/10',
4172 'KW': '37.36.0.0/14',
4173 'KY': '64.96.0.0/15',
4174 'KZ': '2.72.0.0/13',
4175 'LA': '115.84.64.0/18',
4176 'LB': '178.135.0.0/16',
4177 'LC': '24.92.144.0/20',
4178 'LI': '82.117.0.0/19',
4179 'LK': '112.134.0.0/15',
4180 'LR': '102.183.0.0/16',
4181 'LS': '129.232.0.0/17',
4182 'LT': '78.56.0.0/13',
4183 'LU': '188.42.0.0/16',
4184 'LV': '46.109.0.0/16',
4185 'LY': '41.252.0.0/14',
4186 'MA': '105.128.0.0/11',
4187 'MC': '88.209.64.0/18',
4188 'MD': '37.246.0.0/16',
4189 'ME': '178.175.0.0/17',
4190 'MF': '74.112.232.0/21',
4191 'MG': '154.126.0.0/17',
4192 'MH': '117.103.88.0/21',
4193 'MK': '77.28.0.0/15',
4194 'ML': '154.118.128.0/18',
4195 'MM': '37.111.0.0/17',
4196 'MN': '49.0.128.0/17',
4197 'MO': '60.246.0.0/16',
4198 'MP': '202.88.64.0/20',
4199 'MQ': '109.203.224.0/19',
4200 'MR': '41.188.64.0/18',
4201 'MS': '208.90.112.0/22',
4202 'MT': '46.11.0.0/16',
4203 'MU': '105.16.0.0/12',
4204 'MV': '27.114.128.0/18',
4205 'MW': '102.70.0.0/15',
4206 'MX': '187.192.0.0/11',
4207 'MY': '175.136.0.0/13',
4208 'MZ': '197.218.0.0/15',
4209 'NA': '41.182.0.0/16',
4210 'NC': '101.101.0.0/18',
4211 'NE': '197.214.0.0/18',
4212 'NF': '203.17.240.0/22',
4213 'NG': '105.112.0.0/12',
4214 'NI': '186.76.0.0/15',
4215 'NL': '145.96.0.0/11',
4216 'NO': '84.208.0.0/13',
4217 'NP': '36.252.0.0/15',
4218 'NR': '203.98.224.0/19',
4219 'NU': '49.156.48.0/22',
4220 'NZ': '49.224.0.0/14',
4221 'OM': '5.36.0.0/15',
4222 'PA': '186.72.0.0/15',
4223 'PE': '186.160.0.0/14',
4224 'PF': '123.50.64.0/18',
4225 'PG': '124.240.192.0/19',
4226 'PH': '49.144.0.0/13',
4227 'PK': '39.32.0.0/11',
4228 'PL': '83.0.0.0/11',
4229 'PM': '70.36.0.0/20',
4230 'PR': '66.50.0.0/16',
4231 'PS': '188.161.0.0/16',
4232 'PT': '85.240.0.0/13',
4233 'PW': '202.124.224.0/20',
4234 'PY': '181.120.0.0/14',
4235 'QA': '37.210.0.0/15',
4236 'RE': '102.35.0.0/16',
4237 'RO': '79.112.0.0/13',
4238 'RS': '93.86.0.0/15',
4239 'RU': '5.136.0.0/13',
4240 'RW': '41.186.0.0/16',
4241 'SA': '188.48.0.0/13',
4242 'SB': '202.1.160.0/19',
4243 'SC': '154.192.0.0/11',
4244 'SD': '102.120.0.0/13',
4245 'SE': '78.64.0.0/12',
4246 'SG': '8.128.0.0/10',
4247 'SI': '188.196.0.0/14',
4248 'SK': '78.98.0.0/15',
4249 'SL': '102.143.0.0/17',
4250 'SM': '89.186.32.0/19',
4251 'SN': '41.82.0.0/15',
4252 'SO': '154.115.192.0/18',
4253 'SR': '186.179.128.0/17',
4254 'SS': '105.235.208.0/21',
4255 'ST': '197.159.160.0/19',
4256 'SV': '168.243.0.0/16',
4257 'SX': '190.102.0.0/20',
4259 'SZ': '41.84.224.0/19',
4260 'TC': '65.255.48.0/20',
4261 'TD': '154.68.128.0/19',
4262 'TG': '196.168.0.0/14',
4263 'TH': '171.96.0.0/13',
4264 'TJ': '85.9.128.0/18',
4265 'TK': '27.96.24.0/21',
4266 'TL': '180.189.160.0/20',
4267 'TM': '95.85.96.0/19',
4268 'TN': '197.0.0.0/11',
4269 'TO': '175.176.144.0/21',
4270 'TR': '78.160.0.0/11',
4271 'TT': '186.44.0.0/15',
4272 'TV': '202.2.96.0/19',
4273 'TW': '120.96.0.0/11',
4274 'TZ': '156.156.0.0/14',
4275 'UA': '37.52.0.0/14',
4276 'UG': '102.80.0.0/13',
4278 'UY': '167.56.0.0/13',
4279 'UZ': '84.54.64.0/18',
4280 'VA': '212.77.0.0/19',
4281 'VC': '207.191.240.0/21',
4282 'VE': '186.88.0.0/13',
4283 'VG': '66.81.192.0/20',
4284 'VI': '146.226.0.0/16',
4285 'VN': '14.160.0.0/11',
4286 'VU': '202.80.32.0/20',
4287 'WF': '117.20.32.0/21',
4288 'WS': '202.4.32.0/19',
4289 'YE': '134.35.0.0/16',
4290 'YT': '41.242.116.0/22',
4291 'ZA': '41.0.0.0/11',
4292 'ZM': '102.144.0.0/13',
4293 'ZW': '102.177.192.0/18',
4297 def random_ipv4(cls
, code_or_block
):
4298 if len(code_or_block
) == 2:
4299 block
= cls
._country
_ip
_map
.get(code_or_block
.upper())
4303 block
= code_or_block
4304 addr
, preflen
= block
.split('/')
4305 addr_min
= struct
.unpack('!L', socket
.inet_aton(addr
))[0]
4306 addr_max
= addr_min |
(0xffffffff >> int(preflen
))
4307 return str(socket
.inet_ntoa(
4308 struct
.pack('!L', random
.randint(addr_min
, addr_max
))))
4311 # Both long_to_bytes and bytes_to_long are adapted from PyCrypto, which is
4312 # released into Public Domain
4313 # https://github.com/dlitz/pycrypto/blob/master/lib/Crypto/Util/number.py#L387
4315 def long_to_bytes(n
, blocksize
=0):
4316 """long_to_bytes(n:long, blocksize:int) : string
4317 Convert a long integer to a byte string.
4319 If optional blocksize is given and greater than zero, pad the front of the
4320 byte string with binary zeros so that the length is a multiple of
4323 # after much testing, this algorithm was deemed to be the fastest
4327 s
= struct
.pack('>I', n
& 0xffffffff) + s
4329 # strip off leading zeros
4330 for i
in range(len(s
)):
4331 if s
[i
] != b
'\000'[0]:
4334 # only happens when n == 0
4338 # add back some pad bytes. this could be done more efficiently w.r.t. the
4339 # de-padding being done above, but sigh...
4340 if blocksize
> 0 and len(s
) % blocksize
:
4341 s
= (blocksize
- len(s
) % blocksize
) * b
'\000' + s
4345 def bytes_to_long(s
):
4346 """bytes_to_long(string) : long
4347 Convert a byte string to a long integer.
4349 This is (essentially) the inverse of long_to_bytes().
4354 extra
= (4 - length
% 4)
4355 s
= b
'\000' * extra
+ s
4356 length
= length
+ extra
4357 for i
in range(0, length
, 4):
4358 acc
= (acc
<< 32) + struct
.unpack('>I', s
[i
:i
+ 4])[0]
4362 def ohdave_rsa_encrypt(data
, exponent
, modulus
):
4364 Implement OHDave's RSA algorithm. See http://www.ohdave.com/rsa/
4367 data: data to encrypt, bytes-like object
4368 exponent, modulus: parameter e and N of RSA algorithm, both integer
4369 Output: hex string of encrypted data
4371 Limitation: supports one block encryption only
4374 payload
= int(binascii
.hexlify(data
[::-1]), 16)
4375 encrypted
= pow(payload
, exponent
, modulus
)
4376 return '%x' % encrypted
4379 def pkcs1pad(data
, length
):
4381 Padding input data with PKCS#1 scheme
4383 @param {int[]} data input data
4384 @param {int} length target length
4385 @returns {int[]} padded data
4387 if len(data
) > length
- 11:
4388 raise ValueError('Input data too long for PKCS#1 padding')
4390 pseudo_random
= [random
.randint(0, 254) for _
in range(length
- len(data
) - 3)]
4391 return [0, 2] + pseudo_random
+ [0] + data
4394 def _base_n_table(n
, table
):
4395 if not table
and not n
:
4396 raise ValueError('Either table or n must be specified')
4397 table
= (table
or '0123456789abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ')[:n
]
4399 if n
and n
!= len(table
):
4400 raise ValueError(f
'base {n} exceeds table length {len(table)}')
4404 def encode_base_n(num
, n
=None, table
=None):
4405 """Convert given int to a base-n string"""
4406 table
= _base_n_table(n
, table
)
4410 result
, base
= '', len(table
)
4412 result
= table
[num
% base
] + result
4417 def decode_base_n(string
, n
=None, table
=None):
4418 """Convert given base-n string to int"""
4419 table
= {char: index for index, char in enumerate(_base_n_table(n, table))}
4420 result
, base
= 0, len(table
)
4422 result
= result
* base
+ table
[char
]
4426 def decode_packed_codes(code
):
4427 mobj
= re
.search(PACKED_CODES_RE
, code
)
4428 obfuscated_code
, base
, count
, symbols
= mobj
.groups()
4431 symbols
= symbols
.split('|')
4436 base_n_count
= encode_base_n(count
, base
)
4437 symbol_table
[base_n_count
] = symbols
[count
] or base_n_count
4440 r
'\b(\w+)\b', lambda mobj
: symbol_table
[mobj
.group(0)],
4444 def caesar(s
, alphabet
, shift
):
4449 alphabet
[(alphabet
.index(c
) + shift
) % l
] if c
in alphabet
else c
4454 return caesar(s
, r
'''!"#$%&'()*+,-./0123456789:;<=>?@ABCDEFGHIJKLMNOPQRSTUVWXYZ[\]^_`abcdefghijklmnopqrstuvwxyz{|}~''', 47)
4457 def parse_m3u8_attributes(attrib
):
4459 for (key
, val
) in re
.findall(r
'(?P<key>[A-Z0-9-]+)=(?P<val>"[^"]+"|[^",]+)(?:,|$)', attrib
):
4460 if val
.startswith('"'):
4466 def urshift(val
, n
):
4467 return val
>> n
if val
>= 0 else (val
+ 0x100000000) >> n
4470 def write_xattr(path
, key
, value
):
4471 # Windows: Write xattrs to NTFS Alternate Data Streams:
4472 # http://en.wikipedia.org/wiki/NTFS#Alternate_data_streams_.28ADS.29
4473 if compat_os_name
== 'nt':
4474 assert ':' not in key
4475 assert os
.path
.exists(path
)
4478 with open(f
'{path}:{key}', 'wb') as f
:
4480 except OSError as e
:
4481 raise XAttrMetadataError(e
.errno
, e
.strerror
)
4484 # UNIX Method 1. Use xattrs/pyxattrs modules
4487 if getattr(xattr
, '_yt_dlp__identifier', None) == 'pyxattr':
4488 # Unicode arguments are not supported in pyxattr until version 0.5.0
4489 # See https://github.com/ytdl-org/youtube-dl/issues/5498
4490 if version_tuple(xattr
.__version
__) >= (0, 5, 0):
4491 setxattr
= xattr
.set
4493 setxattr
= xattr
.setxattr
4497 setxattr(path
, key
, value
)
4498 except OSError as e
:
4499 raise XAttrMetadataError(e
.errno
, e
.strerror
)
4502 # UNIX Method 2. Use setfattr/xattr executables
4503 exe
= ('setfattr' if check_executable('setfattr', ['--version'])
4504 else 'xattr' if check_executable('xattr', ['-h']) else None)
4506 raise XAttrUnavailableError(
4507 'Couldn\'t find a tool to set the xattrs. Install either the python "xattr" or "pyxattr" modules or the '
4508 + ('"xattr" binary' if sys
.platform
!= 'linux' else 'GNU "attr" package (which contains the "setfattr" tool)'))
4510 value
= value
.decode()
4512 _
, stderr
, returncode
= Popen
.run(
4513 [exe
, '-w', key
, value
, path
] if exe
== 'xattr' else [exe
, '-n', key
, '-v', value
, path
],
4514 text
=True, stdout
=subprocess
.PIPE
, stderr
=subprocess
.PIPE
, stdin
=subprocess
.PIPE
)
4515 except OSError as e
:
4516 raise XAttrMetadataError(e
.errno
, e
.strerror
)
4518 raise XAttrMetadataError(returncode
, stderr
)
4521 def random_birthday(year_field
, month_field
, day_field
):
4522 start_date
= datetime
.date(1950, 1, 1)
4523 end_date
= datetime
.date(1995, 12, 31)
4524 offset
= random
.randint(0, (end_date
- start_date
).days
)
4525 random_date
= start_date
+ datetime
.timedelta(offset
)
4527 year_field
: str(random_date
.year
),
4528 month_field
: str(random_date
.month
),
4529 day_field
: str(random_date
.day
),
4533 def find_available_port(interface
=''):
4535 with socket
.socket() as sock
:
4536 sock
.bind((interface
, 0))
4537 return sock
.getsockname()[1]
4542 # Templates for internet shortcut files, which are plain text files.
4543 DOT_URL_LINK_TEMPLATE
= '''\
4548 DOT_WEBLOC_LINK_TEMPLATE
= '''\
4549 <?xml version="1.0" encoding="UTF-8"?>
4550 <!DOCTYPE plist PUBLIC "-//Apple//DTD PLIST 1.0//EN" "http://www.apple.com/DTDs/PropertyList-1.0.dtd">
4551 <plist version="1.0">
4554 \t<string>%(url)s</string>
4559 DOT_DESKTOP_LINK_TEMPLATE
= '''\
4569 'url': DOT_URL_LINK_TEMPLATE
,
4570 'desktop': DOT_DESKTOP_LINK_TEMPLATE
,
4571 'webloc': DOT_WEBLOC_LINK_TEMPLATE
,
4575 def iri_to_uri(iri
):
4577 Converts an IRI (Internationalized Resource Identifier, allowing Unicode characters) to a URI (Uniform Resource Identifier, ASCII-only).
4579 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.
4582 iri_parts
= urllib
.parse
.urlparse(iri
)
4584 if '[' in iri_parts
.netloc
:
4585 raise ValueError('IPv6 URIs are not, yet, supported.')
4586 # Querying `.netloc`, when there's only one bracket, also raises a ValueError.
4588 # 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.
4591 if iri_parts
.username
:
4592 net_location
+= urllib
.parse
.quote(iri_parts
.username
, safe
=r
"!$%&'()*+,~")
4593 if iri_parts
.password
is not None:
4594 net_location
+= ':' + urllib
.parse
.quote(iri_parts
.password
, safe
=r
"!$%&'()*+,~")
4597 net_location
+= iri_parts
.hostname
.encode('idna').decode() # Punycode for Unicode hostnames.
4598 # The 'idna' encoding produces ASCII text.
4599 if iri_parts
.port
is not None and iri_parts
.port
!= 80:
4600 net_location
+= ':' + str(iri_parts
.port
)
4602 return urllib
.parse
.urlunparse(
4606 urllib
.parse
.quote_plus(iri_parts
.path
, safe
=r
"!$%&'()*+,/:;=@|~"),
4608 # Unsure about the `safe` argument, since this is a legacy way of handling parameters.
4609 urllib
.parse
.quote_plus(iri_parts
.params
, safe
=r
"!$%&'()*+,/:;=@|~"),
4611 # Not totally sure about the `safe` argument, since the source does not explicitly mention the query URI component.
4612 urllib
.parse
.quote_plus(iri_parts
.query
, safe
=r
"!$%&'()*+,/:;=?@{|}~"),
4614 urllib
.parse
.quote_plus(iri_parts
.fragment
, safe
=r
"!#$%&'()*+,/:;=?@{|}~")))
4616 # Source for `safe` arguments: https://url.spec.whatwg.org/#percent-encoded-bytes.
4619 def to_high_limit_path(path
):
4620 if sys
.platform
in ['win32', 'cygwin']:
4621 # Work around MAX_PATH limitation on Windows. The maximum allowed length for the individual path segments may still be quite limited.
4622 return '\\\\?\\' + os
.path
.abspath(path
)
4627 def format_field(obj
, field
=None, template
='%s', ignore
=NO_DEFAULT
, default
='', func
=IDENTITY
):
4628 val
= traversal
.traverse_obj(obj
, *variadic(field
))
4629 if not val
if ignore
is NO_DEFAULT
else val
in variadic(ignore
):
4631 return template
% func(val
)
4634 def clean_podcast_url(url
):
4635 url
= re
.sub(r
'''(?x)
4639 media\.blubrry\.com| # https://create.blubrry.com/resources/podcast-media-download-statistics/getting-started/
4644 (?:dts|www)\.podtrac\.com/(?:pts/)?redirect\.[0-9a-z]{3,4}| # http://analytics.podtrac.com/how-to-measure
4647 cn\.co| # https://podcorn.com/analytics-prefix/
4648 st\.fm # https://podsights.com/docs/
4653 return re
.sub(r
'^\w+://(\w+://)', r
'\1', url
)
4656 _HEX_TABLE
= '0123456789abcdef'
4659 def random_uuidv4():
4660 return re
.sub(r
'[xy]', lambda x
: _HEX_TABLE
[random
.randint(0, 15)], 'xxxxxxxx-xxxx-4xxx-yxxx-xxxxxxxxxxxx')
4663 def make_dir(path
, to_screen
=None):
4665 dn
= os
.path
.dirname(path
)
4667 os
.makedirs(dn
, exist_ok
=True)
4669 except OSError as err
:
4670 if callable(to_screen
) is not None:
4671 to_screen(f
'unable to create directory {err}')
4675 def get_executable_path():
4676 from ..update
import _get_variant_and_executable_path
4678 return os
.path
.dirname(os
.path
.abspath(_get_variant_and_executable_path()[1]))
4681 def get_user_config_dirs(package_name
):
4682 # .config (e.g. ~/.config/package_name)
4683 xdg_config_home
= os
.getenv('XDG_CONFIG_HOME') or compat_expanduser('~/.config')
4684 yield os
.path
.join(xdg_config_home
, package_name
)
4686 # appdata (%APPDATA%/package_name)
4687 appdata_dir
= os
.getenv('appdata')
4689 yield os
.path
.join(appdata_dir
, package_name
)
4691 # home (~/.package_name)
4692 yield os
.path
.join(compat_expanduser('~'), f
'.{package_name}')
4695 def get_system_config_dirs(package_name
):
4697 yield os
.path
.join('/etc', package_name
)
4700 def time_seconds(**kwargs
):
4702 Returns TZ-aware time in seconds since the epoch (1970-01-01T00:00:00Z)
4704 return time
.time() + datetime
.timedelta(**kwargs
).total_seconds()
4707 # create a JSON Web Signature (jws) with HS256 algorithm
4708 # the resulting format is in JWS Compact Serialization
4709 # implemented following JWT https://www.rfc-editor.org/rfc/rfc7519.html
4710 # implemented following JWS https://www.rfc-editor.org/rfc/rfc7515.html
4711 def jwt_encode_hs256(payload_data
, key
, headers
={}):
4717 header_data
.update(headers
)
4718 header_b64
= base64
.b64encode(json
.dumps(header_data
).encode())
4719 payload_b64
= base64
.b64encode(json
.dumps(payload_data
).encode())
4720 h
= hmac
.new(key
.encode(), header_b64
+ b
'.' + payload_b64
, hashlib
.sha256
)
4721 signature_b64
= base64
.b64encode(h
.digest())
4722 token
= header_b64
+ b
'.' + payload_b64
+ b
'.' + signature_b64
4726 # can be extended in future to verify the signature and parse header and return the algorithm used if it's not HS256
4727 def jwt_decode_hs256(jwt
):
4728 header_b64
, payload_b64
, signature_b64
= jwt
.split('.')
4729 # add trailing ='s that may have been stripped, superfluous ='s are ignored
4730 payload_data
= json
.loads(base64
.urlsafe_b64decode(f
'{payload_b64}==='))
4734 WINDOWS_VT_MODE
= False if compat_os_name
== 'nt' else None
4738 def supports_terminal_sequences(stream
):
4739 if compat_os_name
== 'nt':
4740 if not WINDOWS_VT_MODE
:
4742 elif not os
.getenv('TERM'):
4745 return stream
.isatty()
4746 except BaseException
:
4750 def windows_enable_vt_mode():
4751 """Ref: https://bugs.python.org/issue30075 """
4752 if get_windows_version() < (10, 0, 10586):
4756 import ctypes
.wintypes
4759 ENABLE_VIRTUAL_TERMINAL_PROCESSING
= 0x0004
4761 dll
= ctypes
.WinDLL('kernel32', use_last_error
=False)
4762 handle
= os
.open('CONOUT$', os
.O_RDWR
)
4764 h_out
= ctypes
.wintypes
.HANDLE(msvcrt
.get_osfhandle(handle
))
4765 dw_original_mode
= ctypes
.wintypes
.DWORD()
4766 success
= dll
.GetConsoleMode(h_out
, ctypes
.byref(dw_original_mode
))
4768 raise Exception('GetConsoleMode failed')
4770 success
= dll
.SetConsoleMode(h_out
, ctypes
.wintypes
.DWORD(
4771 dw_original_mode
.value | ENABLE_VIRTUAL_TERMINAL_PROCESSING
))
4773 raise Exception('SetConsoleMode failed')
4777 global WINDOWS_VT_MODE
4778 WINDOWS_VT_MODE
= True
4779 supports_terminal_sequences
.cache_clear()
4782 _terminal_sequences_re
= re
.compile('\033\\[[^m]+m')
4785 def remove_terminal_sequences(string
):
4786 return _terminal_sequences_re
.sub('', string
)
4789 def number_of_digits(number
):
4790 return len('%d' % number
)
4793 def join_nonempty(*values
, delim
='-', from_dict
=None):
4794 if from_dict
is not None:
4795 values
= (traversal
.traverse_obj(from_dict
, variadic(v
)) for v
in values
)
4796 return delim
.join(map(str, filter(None, values
)))
4799 def scale_thumbnails_to_max_format_width(formats
, thumbnails
, url_width_re
):
4801 Find the largest format dimensions in terms of video width and, for each thumbnail:
4802 * Modify the URL: Match the width with the provided regex and replace with the former width
4805 This function is useful with video services that scale the provided thumbnails on demand
4807 _keys
= ('width', 'height')
4808 max_dimensions
= max(
4809 (tuple(format
.get(k
) or 0 for k
in _keys
) for format
in formats
),
4811 if not max_dimensions
[0]:
4815 {'url': re.sub(url_width_re, str(max_dimensions[0]), thumbnail['url'])}
,
4816 dict(zip(_keys
, max_dimensions
)), thumbnail
)
4817 for thumbnail
in thumbnails
4821 def parse_http_range(range):
4822 """ Parse value of "Range" or "Content-Range" HTTP header into tuple. """
4824 return None, None, None
4825 crg
= re
.search(r
'bytes[ =](\d+)-(\d+)?(?:/(\d+))?', range)
4827 return None, None, None
4828 return int(crg
.group(1)), int_or_none(crg
.group(2)), int_or_none(crg
.group(3))
4831 def read_stdin(what
):
4832 eof
= 'Ctrl+Z' if compat_os_name
== 'nt' else 'Ctrl+D'
4833 write_string(f
'Reading {what} from STDIN - EOF ({eof}) to end:\n')
4837 def determine_file_encoding(data
):
4839 Detect the text encoding used
4840 @returns (encoding, bytes to skip)
4843 # BOM marks are given priority over declarations
4844 for bom
, enc
in BOMS
:
4845 if data
.startswith(bom
):
4846 return enc
, len(bom
)
4848 # Strip off all null bytes to match even when UTF-16 or UTF-32 is used.
4849 # We ignore the endianness to get a good enough match
4850 data
= data
.replace(b
'\0', b
'')
4851 mobj
= re
.match(rb
'(?m)^#\s*coding\s*:\s*(\S+)\s*$', data
)
4852 return mobj
.group(1).decode() if mobj
else None, 0
4859 __initialized
= False
4861 def __init__(self
, parser
, label
=None):
4862 self
.parser
, self
.label
= parser
, label
4863 self
._loaded
_paths
, self
.configs
= set(), []
4865 def init(self
, args
=None, filename
=None):
4866 assert not self
.__initialized
4867 self
.own_args
, self
.filename
= args
, filename
4868 return self
.load_configs()
4870 def load_configs(self
):
4873 location
= os
.path
.realpath(self
.filename
)
4874 directory
= os
.path
.dirname(location
)
4875 if location
in self
._loaded
_paths
:
4877 self
._loaded
_paths
.add(location
)
4879 self
.__initialized
= True
4880 opts
, _
= self
.parser
.parse_known_args(self
.own_args
)
4881 self
.parsed_args
= self
.own_args
4882 for location
in opts
.config_locations
or []:
4884 if location
in self
._loaded
_paths
:
4886 self
._loaded
_paths
.add(location
)
4887 self
.append_config(shlex
.split(read_stdin('options'), comments
=True), label
='stdin')
4889 location
= os
.path
.join(directory
, expand_path(location
))
4890 if os
.path
.isdir(location
):
4891 location
= os
.path
.join(location
, 'yt-dlp.conf')
4892 if not os
.path
.exists(location
):
4893 self
.parser
.error(f
'config location {location} does not exist')
4894 self
.append_config(self
.read_file(location
), location
)
4898 label
= join_nonempty(
4899 self
.label
, 'config', f
'"{self.filename}"' if self
.filename
else '',
4901 return join_nonempty(
4902 self
.own_args
is not None and f
'{label[0].upper()}{label[1:]}: {self.hide_login_info(self.own_args)}',
4903 *(f
'\n{c}'.replace('\n', '\n| ')[1:] for c
in self
.configs
),
4907 def read_file(filename
, default
=[]):
4909 optionf
= open(filename
, 'rb')
4911 return default
# silently skip if file is not present
4913 enc
, skip
= determine_file_encoding(optionf
.read(512))
4914 optionf
.seek(skip
, io
.SEEK_SET
)
4916 enc
= None # silently skip read errors
4918 # FIXME: https://github.com/ytdl-org/youtube-dl/commit/dfe5fa49aed02cf36ba9f743b11b0903554b5e56
4919 contents
= optionf
.read().decode(enc
or preferredencoding())
4920 res
= shlex
.split(contents
, comments
=True)
4921 except Exception as err
:
4922 raise ValueError(f
'Unable to parse "{filename}": {err}')
4928 def hide_login_info(opts
):
4929 PRIVATE_OPTS
= {'-p', '--password', '-u', '--username', '--video-password', '--ap-password', '--ap-username'}
4930 eqre
= re
.compile('^(?P<key>' + ('|'.join(re
.escape(po
) for po
in PRIVATE_OPTS
)) + ')=.+$')
4935 return m
.group('key') + '=PRIVATE'
4939 opts
= list(map(_scrub_eq
, opts
))
4940 for idx
, opt
in enumerate(opts
):
4941 if opt
in PRIVATE_OPTS
and idx
+ 1 < len(opts
):
4942 opts
[idx
+ 1] = 'PRIVATE'
4945 def append_config(self
, *args
, label
=None):
4946 config
= type(self
)(self
.parser
, label
)
4947 config
._loaded
_paths
= self
._loaded
_paths
4948 if config
.init(*args
):
4949 self
.configs
.append(config
)
4953 for config
in reversed(self
.configs
):
4954 yield from config
.all_args
4955 yield from self
.parsed_args
or []
4957 def parse_known_args(self
, **kwargs
):
4958 return self
.parser
.parse_known_args(self
.all_args
, **kwargs
)
4960 def parse_args(self
):
4961 return self
.parser
.parse_args(self
.all_args
)
4964 class WebSocketsWrapper
:
4965 """Wraps websockets module to use in non-async scopes"""
4968 def __init__(self
, url
, headers
=None, connect
=True):
4969 self
.loop
= asyncio
.new_event_loop()
4970 # XXX: "loop" is deprecated
4971 self
.conn
= websockets
.connect(
4972 url
, extra_headers
=headers
, ping_interval
=None,
4973 close_timeout
=float('inf'), loop
=self
.loop
, ping_timeout
=float('inf'))
4976 atexit
.register(self
.__exit
__, None, None, None)
4978 def __enter__(self
):
4980 self
.pool
= self
.run_with_loop(self
.conn
.__aenter
__(), self
.loop
)
4983 def send(self
, *args
):
4984 self
.run_with_loop(self
.pool
.send(*args
), self
.loop
)
4986 def recv(self
, *args
):
4987 return self
.run_with_loop(self
.pool
.recv(*args
), self
.loop
)
4989 def __exit__(self
, type, value
, traceback
):
4991 return self
.run_with_loop(self
.conn
.__aexit
__(type, value
, traceback
), self
.loop
)
4994 self
._cancel
_all
_tasks
(self
.loop
)
4996 # taken from https://github.com/python/cpython/blob/3.9/Lib/asyncio/runners.py with modifications
4997 # for contributors: If there's any new library using asyncio needs to be run in non-async, move these function out of this class
4999 def run_with_loop(main
, loop
):
5000 if not asyncio
.iscoroutine(main
):
5001 raise ValueError(f
'a coroutine was expected, got {main!r}')
5004 return loop
.run_until_complete(main
)
5006 loop
.run_until_complete(loop
.shutdown_asyncgens())
5007 if hasattr(loop
, 'shutdown_default_executor'):
5008 loop
.run_until_complete(loop
.shutdown_default_executor())
5011 def _cancel_all_tasks(loop
):
5012 to_cancel
= asyncio
.all_tasks(loop
)
5017 for task
in to_cancel
:
5020 # XXX: "loop" is removed in python 3.10+
5021 loop
.run_until_complete(
5022 asyncio
.gather(*to_cancel
, loop
=loop
, return_exceptions
=True))
5024 for task
in to_cancel
:
5025 if task
.cancelled():
5027 if task
.exception() is not None:
5028 loop
.call_exception_handler({
5029 'message': 'unhandled exception during asyncio.run() shutdown',
5030 'exception': task
.exception(),
5035 def merge_headers(*dicts
):
5036 """Merge dicts of http headers case insensitively, prioritizing the latter ones"""
5037 return {k.title(): v for k, v in itertools.chain.from_iterable(map(dict.items, dicts))}
5040 def cached_method(f
):
5041 """Cache a method"""
5042 signature
= inspect
.signature(f
)
5045 def wrapper(self
, *args
, **kwargs
):
5046 bound_args
= signature
.bind(self
, *args
, **kwargs
)
5047 bound_args
.apply_defaults()
5048 key
= tuple(bound_args
.arguments
.values())[1:]
5050 cache
= vars(self
).setdefault('_cached_method__cache', {}).setdefault(f.__name__, {}
)
5051 if key
not in cache
:
5052 cache
[key
] = f(self
, *args
, **kwargs
)
5057 class classproperty
:
5058 """property access for class methods with optional caching"""
5059 def __new__(cls
, func
=None, *args
, **kwargs
):
5061 return functools
.partial(cls
, *args
, **kwargs
)
5062 return super().__new
__(cls
)
5064 def __init__(self
, func
, *, cache
=False):
5065 functools
.update_wrapper(self
, func
)
5067 self
._cache
= {} if cache
else None
5069 def __get__(self
, _
, cls
):
5070 if self
._cache
is None:
5071 return self
.func(cls
)
5072 elif cls
not in self
._cache
:
5073 self
._cache
[cls
] = self
.func(cls
)
5074 return self
._cache
[cls
]
5077 class function_with_repr
:
5078 def __init__(self
, func
, repr_
=None):
5079 functools
.update_wrapper(self
, func
)
5080 self
.func
, self
.__repr
= func
, repr_
5082 def __call__(self
, *args
, **kwargs
):
5083 return self
.func(*args
, **kwargs
)
5088 return f
'{self.func.__module__}.{self.func.__qualname__}'
5091 class Namespace(types
.SimpleNamespace
):
5092 """Immutable namespace"""
5095 return iter(self
.__dict
__.values())
5099 return self
.__dict
__.items()
5102 MEDIA_EXTENSIONS
= Namespace(
5103 common_video
=('avi', 'flv', 'mkv', 'mov', 'mp4', 'webm'),
5104 video
=('3g2', '3gp', 'f4v', 'mk3d', 'divx', 'mpg', 'ogv', 'm4v', 'wmv'),
5105 common_audio
=('aiff', 'alac', 'flac', 'm4a', 'mka', 'mp3', 'ogg', 'opus', 'wav'),
5106 audio
=('aac', 'ape', 'asf', 'f4a', 'f4b', 'm4b', 'm4p', 'm4r', 'oga', 'ogx', 'spx', 'vorbis', 'wma', 'weba'),
5107 thumbnails
=('jpg', 'png', 'webp'),
5108 storyboards
=('mhtml', ),
5109 subtitles
=('srt', 'vtt', 'ass', 'lrc'),
5110 manifests
=('f4f', 'f4m', 'm3u8', 'smil', 'mpd'),
5112 MEDIA_EXTENSIONS
.video
+= MEDIA_EXTENSIONS
.common_video
5113 MEDIA_EXTENSIONS
.audio
+= MEDIA_EXTENSIONS
.common_audio
5115 KNOWN_EXTENSIONS
= (*MEDIA_EXTENSIONS
.video
, *MEDIA_EXTENSIONS
.audio
, *MEDIA_EXTENSIONS
.manifests
)
5120 for retry in RetryManager(...):
5123 except SomeException as err:
5127 attempt
, _error
= 0, None
5129 def __init__(self
, _retries
, _error_callback
, **kwargs
):
5130 self
.retries
= _retries
or 0
5131 self
.error_callback
= functools
.partial(_error_callback
, **kwargs
)
5133 def _should_retry(self
):
5134 return self
._error
is not NO_DEFAULT
and self
.attempt
<= self
.retries
5138 if self
._error
is NO_DEFAULT
:
5143 def error(self
, value
):
5147 while self
._should
_retry
():
5148 self
.error
= NO_DEFAULT
5152 self
.error_callback(self
.error
, self
.attempt
, self
.retries
)
5155 def report_retry(e
, count
, retries
, *, sleep_func
, info
, warn
, error
=None, suffix
=None):
5156 """Utility function for reporting retries"""
5159 return error(f
'{e}. Giving up after {count - 1} retries') if count
> 1 else error(str(e
))
5164 elif isinstance(e
, ExtractorError
):
5165 e
= remove_end(str_or_none(e
.cause
) or e
.orig_msg
, '.')
5166 warn(f
'{e}. Retrying{format_field(suffix, None, " %s")} ({count}/{retries})...')
5168 delay
= float_or_none(sleep_func(n
=count
- 1)) if callable(sleep_func
) else sleep_func
5170 info(f
'Sleeping {delay:.2f} seconds ...')
5174 def make_archive_id(ie
, video_id
):
5175 ie_key
= ie
if isinstance(ie
, str) else ie
.ie_key()
5176 return f
'{ie_key.lower()} {video_id}'
5179 def truncate_string(s
, left
, right
=0):
5180 assert left
> 3 and right
>= 0
5181 if s
is None or len(s
) <= left
+ right
:
5183 return f
'{s[:left-3]}...{s[-right:] if right else ""}'
5186 def orderedSet_from_options(options
, alias_dict
, *, use_regex
=False, start
=None):
5187 assert 'all' in alias_dict
, '"all" alias is required'
5188 requested
= list(start
or [])
5190 discard
= val
.startswith('-')
5194 if val
in alias_dict
:
5195 val
= alias_dict
[val
] if not discard
else [
5196 i
[1:] if i
.startswith('-') else f
'-{i}' for i
in alias_dict
[val
]]
5197 # NB: Do not allow regex in aliases for performance
5198 requested
= orderedSet_from_options(val
, alias_dict
, start
=requested
)
5201 current
= (filter(re
.compile(val
, re
.I
).fullmatch
, alias_dict
['all']) if use_regex
5202 else [val
] if val
in alias_dict
['all'] else None)
5204 raise ValueError(val
)
5207 for item
in current
:
5208 while item
in requested
:
5209 requested
.remove(item
)
5211 requested
.extend(current
)
5213 return orderedSet(requested
)
5218 regex
= r
' *((?P<reverse>\+)?(?P<field>[a-zA-Z0-9_]+)((?P<separator>[~:])(?P<limit>.*?))?)? *$'
5220 default
= ('hidden', 'aud_or_vid', 'hasvid', 'ie_pref', 'lang', 'quality',
5221 'res', 'fps', 'hdr:12', 'vcodec:vp9.2', 'channels', 'acodec',
5222 'size', 'br', 'asr', 'proto', 'ext', 'hasaud', 'source', 'id') # These must not be aliases
5223 ytdl_default
= ('hasaud', 'lang', 'quality', 'tbr', 'filesize', 'vbr',
5224 'height', 'width', 'proto', 'vext', 'abr', 'aext',
5225 'fps', 'fs_approx', 'source', 'id')
5228 'vcodec': {'type': 'ordered', 'regex': True,
5229 'order': ['av0?1', 'vp0?9.2', 'vp0?9', '[hx]265|he?vc?', '[hx]264|avc', 'vp0?8', 'mp4v|h263', 'theora', '', None, 'none']},
5230 'acodec': {'type': 'ordered', 'regex': True,
5231 'order': ['[af]lac', 'wav|aiff', 'opus', 'vorbis|ogg', 'aac', 'mp?4a?', 'mp3', 'ac-?4', 'e-?a?c-?3', 'ac-?3', 'dts', '', None, 'none']},
5232 'hdr': {'type': 'ordered', 'regex': True, 'field': 'dynamic_range',
5233 'order': ['dv', '(hdr)?12', r
'(hdr)?10\+', '(hdr)?10', 'hlg', '', 'sdr', None]},
5234 'proto': {'type': 'ordered', 'regex': True, 'field': 'protocol',
5235 'order': ['(ht|f)tps', '(ht|f)tp$', 'm3u8.*', '.*dash', 'websocket_frag', 'rtmpe?', '', 'mms|rtsp', 'ws|websocket', 'f4']},
5236 'vext': {'type': 'ordered', 'field': 'video_ext',
5237 'order': ('mp4', 'mov', 'webm', 'flv', '', 'none'),
5238 'order_free': ('webm', 'mp4', 'mov', 'flv', '', 'none')},
5239 'aext': {'type': 'ordered', 'regex': True, 'field': 'audio_ext',
5240 'order': ('m4a', 'aac', 'mp3', 'ogg', 'opus', 'web[am]', '', 'none'),
5241 'order_free': ('ogg', 'opus', 'web[am]', 'mp3', 'm4a', 'aac', '', 'none')},
5242 'hidden': {'visible': False, 'forced': True, 'type': 'extractor', 'max': -1000}
,
5243 'aud_or_vid': {'visible': False, 'forced': True, 'type': 'multiple',
5244 'field': ('vcodec', 'acodec'),
5245 'function': lambda it
: int(any(v
!= 'none' for v
in it
))},
5246 'ie_pref': {'priority': True, 'type': 'extractor'}
,
5247 'hasvid': {'priority': True, 'field': 'vcodec', 'type': 'boolean', 'not_in_list': ('none',)}
,
5248 'hasaud': {'field': 'acodec', 'type': 'boolean', 'not_in_list': ('none',)}
,
5249 'lang': {'convert': 'float', 'field': 'language_preference', 'default': -1}
,
5250 'quality': {'convert': 'float', 'default': -1}
,
5251 'filesize': {'convert': 'bytes'}
,
5252 'fs_approx': {'convert': 'bytes', 'field': 'filesize_approx'}
,
5253 'id': {'convert': 'string', 'field': 'format_id'}
,
5254 'height': {'convert': 'float_none'}
,
5255 'width': {'convert': 'float_none'}
,
5256 'fps': {'convert': 'float_none'}
,
5257 'channels': {'convert': 'float_none', 'field': 'audio_channels'}
,
5258 'tbr': {'convert': 'float_none'}
,
5259 'vbr': {'convert': 'float_none'}
,
5260 'abr': {'convert': 'float_none'}
,
5261 'asr': {'convert': 'float_none'}
,
5262 'source': {'convert': 'float', 'field': 'source_preference', 'default': -1}
,
5264 'codec': {'type': 'combined', 'field': ('vcodec', 'acodec')}
,
5265 'br': {'type': 'multiple', 'field': ('tbr', 'vbr', 'abr'), 'convert': 'float_none',
5266 'function': lambda it
: next(filter(None, it
), None)},
5267 'size': {'type': 'multiple', 'field': ('filesize', 'fs_approx'), 'convert': 'bytes',
5268 'function': lambda it
: next(filter(None, it
), None)},
5269 'ext': {'type': 'combined', 'field': ('vext', 'aext')}
,
5270 'res': {'type': 'multiple', 'field': ('height', 'width'),
5271 'function': lambda it
: (lambda l
: min(l
) if l
else 0)(tuple(filter(None, it
)))},
5273 # Actual field names
5274 'format_id': {'type': 'alias', 'field': 'id'}
,
5275 'preference': {'type': 'alias', 'field': 'ie_pref'}
,
5276 'language_preference': {'type': 'alias', 'field': 'lang'}
,
5277 'source_preference': {'type': 'alias', 'field': 'source'}
,
5278 'protocol': {'type': 'alias', 'field': 'proto'}
,
5279 'filesize_approx': {'type': 'alias', 'field': 'fs_approx'}
,
5280 'audio_channels': {'type': 'alias', 'field': 'channels'}
,
5283 'dimension': {'type': 'alias', 'field': 'res', 'deprecated': True}
,
5284 'resolution': {'type': 'alias', 'field': 'res', 'deprecated': True}
,
5285 'extension': {'type': 'alias', 'field': 'ext', 'deprecated': True}
,
5286 'bitrate': {'type': 'alias', 'field': 'br', 'deprecated': True}
,
5287 'total_bitrate': {'type': 'alias', 'field': 'tbr', 'deprecated': True}
,
5288 'video_bitrate': {'type': 'alias', 'field': 'vbr', 'deprecated': True}
,
5289 'audio_bitrate': {'type': 'alias', 'field': 'abr', 'deprecated': True}
,
5290 'framerate': {'type': 'alias', 'field': 'fps', 'deprecated': True}
,
5291 'filesize_estimate': {'type': 'alias', 'field': 'size', 'deprecated': True}
,
5292 'samplerate': {'type': 'alias', 'field': 'asr', 'deprecated': True}
,
5293 'video_ext': {'type': 'alias', 'field': 'vext', 'deprecated': True}
,
5294 'audio_ext': {'type': 'alias', 'field': 'aext', 'deprecated': True}
,
5295 'video_codec': {'type': 'alias', 'field': 'vcodec', 'deprecated': True}
,
5296 'audio_codec': {'type': 'alias', 'field': 'acodec', 'deprecated': True}
,
5297 'video': {'type': 'alias', 'field': 'hasvid', 'deprecated': True}
,
5298 'has_video': {'type': 'alias', 'field': 'hasvid', 'deprecated': True}
,
5299 'audio': {'type': 'alias', 'field': 'hasaud', 'deprecated': True}
,
5300 'has_audio': {'type': 'alias', 'field': 'hasaud', 'deprecated': True}
,
5301 'extractor': {'type': 'alias', 'field': 'ie_pref', 'deprecated': True}
,
5302 'extractor_preference': {'type': 'alias', 'field': 'ie_pref', 'deprecated': True}
,
5305 def __init__(self
, ydl
, field_preference
):
5308 self
.evaluate_params(self
.ydl
.params
, field_preference
)
5309 if ydl
.params
.get('verbose'):
5310 self
.print_verbose_info(self
.ydl
.write_debug
)
5312 def _get_field_setting(self
, field
, key
):
5313 if field
not in self
.settings
:
5314 if key
in ('forced', 'priority'):
5316 self
.ydl
.deprecated_feature(f
'Using arbitrary fields ({field}) for format sorting is '
5317 'deprecated and may be removed in a future version')
5318 self
.settings
[field
] = {}
5319 propObj
= self
.settings
[field
]
5320 if key
not in propObj
:
5321 type = propObj
.get('type')
5323 default
= 'preference' if type == 'extractor' else (field
,) if type in ('combined', 'multiple') else field
5324 elif key
== 'convert':
5325 default
= 'order' if type == 'ordered' else 'float_string' if field
else 'ignore'
5327 default
= {'type': 'field', 'visible': True, 'order': [], 'not_in_list': (None,)}
.get(key
, None)
5328 propObj
[key
] = default
5331 def _resolve_field_value(self
, field
, value
, convertNone
=False):
5336 value
= value
.lower()
5337 conversion
= self
._get
_field
_setting
(field
, 'convert')
5338 if conversion
== 'ignore':
5340 if conversion
== 'string':
5342 elif conversion
== 'float_none':
5343 return float_or_none(value
)
5344 elif conversion
== 'bytes':
5345 return parse_bytes(value
)
5346 elif conversion
== 'order':
5347 order_list
= (self
._use
_free
_order
and self
._get
_field
_setting
(field
, 'order_free')) or self
._get
_field
_setting
(field
, 'order')
5348 use_regex
= self
._get
_field
_setting
(field
, 'regex')
5349 list_length
= len(order_list
)
5350 empty_pos
= order_list
.index('') if '' in order_list
else list_length
+ 1
5351 if use_regex
and value
is not None:
5352 for i
, regex
in enumerate(order_list
):
5353 if regex
and re
.match(regex
, value
):
5354 return list_length
- i
5355 return list_length
- empty_pos
# not in list
5356 else: # not regex or value = None
5357 return list_length
- (order_list
.index(value
) if value
in order_list
else empty_pos
)
5359 if value
.isnumeric():
5362 self
.settings
[field
]['convert'] = 'string'
5365 def evaluate_params(self
, params
, sort_extractor
):
5366 self
._use
_free
_order
= params
.get('prefer_free_formats', False)
5367 self
._sort
_user
= params
.get('format_sort', [])
5368 self
._sort
_extractor
= sort_extractor
5370 def add_item(field
, reverse
, closest
, limit_text
):
5371 field
= field
.lower()
5372 if field
in self
._order
:
5374 self
._order
.append(field
)
5375 limit
= self
._resolve
_field
_value
(field
, limit_text
)
5378 'closest': False if limit
is None else closest
,
5379 'limit_text': limit_text
,
5381 if field
in self
.settings
:
5382 self
.settings
[field
].update(data
)
5384 self
.settings
[field
] = data
5387 tuple(field
for field
in self
.default
if self
._get
_field
_setting
(field
, 'forced'))
5388 + (tuple() if params
.get('format_sort_force', False)
5389 else tuple(field
for field
in self
.default
if self
._get
_field
_setting
(field
, 'priority')))
5390 + tuple(self
._sort
_user
) + tuple(sort_extractor
) + self
.default
)
5392 for item
in sort_list
:
5393 match
= re
.match(self
.regex
, item
)
5395 raise ExtractorError('Invalid format sort string "%s" given by extractor' % item
)
5396 field
= match
.group('field')
5399 if self
._get
_field
_setting
(field
, 'type') == 'alias':
5400 alias
, field
= field
, self
._get
_field
_setting
(field
, 'field')
5401 if self
._get
_field
_setting
(alias
, 'deprecated'):
5402 self
.ydl
.deprecated_feature(f
'Format sorting alias {alias} is deprecated and may '
5403 f
'be removed in a future version. Please use {field} instead')
5404 reverse
= match
.group('reverse') is not None
5405 closest
= match
.group('separator') == '~'
5406 limit_text
= match
.group('limit')
5408 has_limit
= limit_text
is not None
5409 has_multiple_fields
= self
._get
_field
_setting
(field
, 'type') == 'combined'
5410 has_multiple_limits
= has_limit
and has_multiple_fields
and not self
._get
_field
_setting
(field
, 'same_limit')
5412 fields
= self
._get
_field
_setting
(field
, 'field') if has_multiple_fields
else (field
,)
5413 limits
= limit_text
.split(':') if has_multiple_limits
else (limit_text
,) if has_limit
else tuple()
5414 limit_count
= len(limits
)
5415 for (i
, f
) in enumerate(fields
):
5416 add_item(f
, reverse
, closest
,
5417 limits
[i
] if i
< limit_count
5418 else limits
[0] if has_limit
and not has_multiple_limits
5421 def print_verbose_info(self
, write_debug
):
5423 write_debug('Sort order given by user: %s' % ', '.join(self
._sort
_user
))
5424 if self
._sort
_extractor
:
5425 write_debug('Sort order given by extractor: %s' % ', '.join(self
._sort
_extractor
))
5426 write_debug('Formats sorted by: %s' % ', '.join(['%s%s%s' % (
5427 '+' if self
._get
_field
_setting
(field
, 'reverse') else '', field
,
5428 '%s%s(%s)' % ('~' if self
._get
_field
_setting
(field
, 'closest') else ':',
5429 self
._get
_field
_setting
(field
, 'limit_text'),
5430 self
._get
_field
_setting
(field
, 'limit'))
5431 if self
._get
_field
_setting
(field
, 'limit_text') is not None else '')
5432 for field
in self
._order
if self
._get
_field
_setting
(field
, 'visible')]))
5434 def _calculate_field_preference_from_value(self
, format
, field
, type, value
):
5435 reverse
= self
._get
_field
_setting
(field
, 'reverse')
5436 closest
= self
._get
_field
_setting
(field
, 'closest')
5437 limit
= self
._get
_field
_setting
(field
, 'limit')
5439 if type == 'extractor':
5440 maximum
= self
._get
_field
_setting
(field
, 'max')
5441 if value
is None or (maximum
is not None and value
>= maximum
):
5443 elif type == 'boolean':
5444 in_list
= self
._get
_field
_setting
(field
, 'in_list')
5445 not_in_list
= self
._get
_field
_setting
(field
, 'not_in_list')
5446 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
5447 elif type == 'ordered':
5448 value
= self
._resolve
_field
_value
(field
, value
, True)
5450 # try to convert to number
5451 val_num
= float_or_none(value
, default
=self
._get
_field
_setting
(field
, 'default'))
5452 is_num
= self
._get
_field
_setting
(field
, 'convert') != 'string' and val_num
is not None
5456 return ((-10, 0) if value
is None
5457 else (1, value
, 0) if not is_num
# if a field has mixed strings and numbers, strings are sorted higher
5458 else (0, -abs(value
- limit
), value
- limit
if reverse
else limit
- value
) if closest
5459 else (0, value
, 0) if not reverse
and (limit
is None or value
<= limit
)
5460 else (0, -value
, 0) if limit
is None or (reverse
and value
== limit
) or value
> limit
5461 else (-1, value
, 0))
5463 def _calculate_field_preference(self
, format
, field
):
5464 type = self
._get
_field
_setting
(field
, 'type') # extractor, boolean, ordered, field, multiple
5465 get_value
= lambda f
: format
.get(self
._get
_field
_setting
(f
, 'field'))
5466 if type == 'multiple':
5467 type = 'field' # Only 'field' is allowed in multiple for now
5468 actual_fields
= self
._get
_field
_setting
(field
, 'field')
5470 value
= self
._get
_field
_setting
(field
, 'function')(get_value(f
) for f
in actual_fields
)
5472 value
= get_value(field
)
5473 return self
._calculate
_field
_preference
_from
_value
(format
, field
, type, value
)
5475 def calculate_preference(self
, format
):
5476 # Determine missing protocol
5477 if not format
.get('protocol'):
5478 format
['protocol'] = determine_protocol(format
)
5480 # Determine missing ext
5481 if not format
.get('ext') and 'url' in format
:
5482 format
['ext'] = determine_ext(format
['url'])
5483 if format
.get('vcodec') == 'none':
5484 format
['audio_ext'] = format
['ext'] if format
.get('acodec') != 'none' else 'none'
5485 format
['video_ext'] = 'none'
5487 format
['video_ext'] = format
['ext']
5488 format
['audio_ext'] = 'none'
5489 # if format.get('preference') is None and format.get('ext') in ('f4f', 'f4m'): # Not supported?
5490 # format['preference'] = -1000
5492 if format
.get('preference') is None and format
.get('ext') == 'flv' and re
.match('[hx]265|he?vc?', format
.get('vcodec') or ''):
5493 # HEVC-over-FLV is out-of-spec by FLV's original spec
5494 # ref. https://trac.ffmpeg.org/ticket/6389
5495 # ref. https://github.com/yt-dlp/yt-dlp/pull/5821
5496 format
['preference'] = -100
5498 # Determine missing bitrates
5499 if format
.get('vcodec') == 'none':
5501 if format
.get('acodec') == 'none':
5503 if not format
.get('vbr') and format
.get('vcodec') != 'none':
5504 format
['vbr'] = try_call(lambda: format
['tbr'] - format
['abr']) or None
5505 if not format
.get('abr') and format
.get('acodec') != 'none':
5506 format
['abr'] = try_call(lambda: format
['tbr'] - format
['vbr']) or None
5507 if not format
.get('tbr'):
5508 format
['tbr'] = try_call(lambda: format
['vbr'] + format
['abr']) or None
5510 return tuple(self
._calculate
_field
_preference
(format
, field
) for field
in self
._order
)
5515 def __init__(self
, ydl
=None):
5518 def debug(self
, message
):
5520 self
._ydl
.write_debug(message
)
5522 def info(self
, message
):
5524 self
._ydl
.to_screen(message
)
5526 def warning(self
, message
, *, once
=False):
5528 self
._ydl
.report_warning(message
, only_once
=once
)
5530 def error(self
, message
, *, is_error
=True):
5532 self
._ydl
.report_error(message
, is_error
=is_error
)
5534 def stdout(self
, message
):
5536 self
._ydl
.to_stdout(message
)
5538 def stderr(self
, message
):
5540 self
._ydl
.to_stderr(message
)