47 import xml
.etree
.ElementTree
50 from .compat
import functools
# isort: split
52 compat_etree_fromstring
,
54 compat_HTMLParseError
,
58 from .dependencies
import brotli
, certifi
, websockets
, xattr
59 from .socks
import ProxyType
, sockssocket
62 def register_socks_protocols():
63 # "Register" SOCKS protocols
64 # In Python < 2.6.5, urlsplit() suffers from bug https://bugs.python.org/issue7904
65 # URLs with protocols not in urlparse.uses_netloc are not handled correctly
66 for scheme
in ('socks', 'socks4', 'socks4a', 'socks5'):
67 if scheme
not in urllib
.parse
.uses_netloc
:
68 urllib
.parse
.uses_netloc
.append(scheme
)
71 # This is not clearly defined otherwise
72 compiled_regex_type
= type(re
.compile(''))
75 def random_user_agent():
76 _USER_AGENT_TPL
= 'Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/%s Safari/537.36'
117 return _USER_AGENT_TPL
% random
.choice(_CHROME_VERSIONS
)
120 SUPPORTED_ENCODINGS
= [
124 SUPPORTED_ENCODINGS
.append('br')
127 'User-Agent': random_user_agent(),
128 'Accept': 'text/html,application/xhtml+xml,application/xml;q=0.9,*/*;q=0.8',
129 'Accept-Language': 'en-us,en;q=0.5',
130 'Sec-Fetch-Mode': 'navigate',
135 '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',
139 NO_DEFAULT
= object()
140 IDENTITY
= lambda x
: x
142 ENGLISH_MONTH_NAMES
= [
143 'January', 'February', 'March', 'April', 'May', 'June',
144 'July', 'August', 'September', 'October', 'November', 'December']
147 'en': ENGLISH_MONTH_NAMES
,
149 'janvier', 'février', 'mars', 'avril', 'mai', 'juin',
150 'juillet', 'août', 'septembre', 'octobre', 'novembre', 'décembre'],
153 # needed for sanitizing filenames in restricted mode
154 ACCENT_CHARS
= dict(zip('ÂÃÄÀÁÅÆÇÈÉÊËÌÍÎÏÐÑÒÓÔÕÖŐØŒÙÚÛÜŰÝÞßàáâãäåæçèéêëìíîïðñòóôõöőøœùúûüűýþÿ',
155 itertools
.chain('AAAAAA', ['AE'], 'CEEEEIIIIDNOOOOOOO', ['OE'], 'UUUUUY', ['TH', 'ss'],
156 'aaaaaa', ['ae'], 'ceeeeiiiionooooooo', ['oe'], 'uuuuuy', ['th'], 'y')))
186 '%Y-%m-%d %H:%M:%S.%f',
187 '%Y-%m-%d %H:%M:%S:%f',
190 '%Y-%m-%dT%H:%M:%SZ',
191 '%Y-%m-%dT%H:%M:%S.%fZ',
192 '%Y-%m-%dT%H:%M:%S.%f0Z',
194 '%Y-%m-%dT%H:%M:%S.%f',
197 '%b %d %Y at %H:%M:%S',
199 '%B %d %Y at %H:%M:%S',
203 DATE_FORMATS_DAY_FIRST
= list(DATE_FORMATS
)
204 DATE_FORMATS_DAY_FIRST
.extend([
213 DATE_FORMATS_MONTH_FIRST
= list(DATE_FORMATS
)
214 DATE_FORMATS_MONTH_FIRST
.extend([
222 PACKED_CODES_RE
= r
"}\('(.+)',(\d+),(\d+),'([^']+)'\.split\('\|'\)"
223 JSON_LD_RE
= r
'(?is)<script[^>]+type=(["\']?
)application
/ld\
+json\
1[^
>]*>\s
*(?P
<json_ld
>{.+?}
)\s
*</script
>'
225 NUMBER_RE = r'\d
+(?
:\
.\d
+)?
'
229 def preferredencoding():
230 """Get preferred encoding.
232 Returns the best encoding scheme for the system, based on
233 locale.getpreferredencoding() and some further tweaks.
236 pref = locale.getpreferredencoding()
244 def write_json_file(obj, fn):
245 """ Encode obj as JSON and write it to fn, atomically if possible """
247 tf = tempfile.NamedTemporaryFile(
248 prefix=f'{os.path.basename(fn)}
.', dir=os.path.dirname(fn),
249 suffix='.tmp
', delete=False, mode='w
', encoding='utf
-8')
253 json.dump(obj, tf, ensure_ascii=False)
254 if sys.platform == 'win32
':
255 # Need to remove existing file on Windows, else os.rename raises
256 # WindowsError or FileExistsError.
257 with contextlib.suppress(OSError):
259 with contextlib.suppress(OSError):
262 os.chmod(tf.name, 0o666 & ~mask)
263 os.rename(tf.name, fn)
265 with contextlib.suppress(OSError):
270 def find_xpath_attr(node, xpath, key, val=None):
271 """ Find the xpath xpath[@key=val] """
272 assert re.match(r'^
[a
-zA
-Z_
-]+$
', key)
273 expr = xpath + ('[@%s]' % key if val is None else f"[@{key}='{val}
']")
274 return node.find(expr)
276 # On python2.6 the xml.etree.ElementTree.Element methods don't support
277 # the namespace parameter
280 def xpath_with_ns(path
, ns_map
):
281 components
= [c
.split(':') for c
in path
.split('/')]
285 replaced
.append(c
[0])
288 replaced
.append('{%s}%s' % (ns_map
[ns
], tag
))
289 return '/'.join(replaced
)
292 def xpath_element(node
, xpath
, name
=None, fatal
=False, default
=NO_DEFAULT
):
293 def _find_xpath(xpath
):
294 return node
.find(xpath
)
296 if isinstance(xpath
, str):
297 n
= _find_xpath(xpath
)
305 if default
is not NO_DEFAULT
:
308 name
= xpath
if name
is None else name
309 raise ExtractorError('Could not find XML element %s' % name
)
315 def xpath_text(node
, xpath
, name
=None, fatal
=False, default
=NO_DEFAULT
):
316 n
= xpath_element(node
, xpath
, name
, fatal
=fatal
, default
=default
)
317 if n
is None or n
== default
:
320 if default
is not NO_DEFAULT
:
323 name
= xpath
if name
is None else name
324 raise ExtractorError('Could not find XML element\'s text %s' % name
)
330 def xpath_attr(node
, xpath
, key
, name
=None, fatal
=False, default
=NO_DEFAULT
):
331 n
= find_xpath_attr(node
, xpath
, key
)
333 if default
is not NO_DEFAULT
:
336 name
= f
'{xpath}[@{key}]' if name
is None else name
337 raise ExtractorError('Could not find XML attribute %s' % name
)
343 def get_element_by_id(id, html
, **kwargs
):
344 """Return the content of the tag with the specified ID in the passed HTML document"""
345 return get_element_by_attribute('id', id, html
, **kwargs
)
348 def get_element_html_by_id(id, html
, **kwargs
):
349 """Return the html of the tag with the specified ID in the passed HTML document"""
350 return get_element_html_by_attribute('id', id, html
, **kwargs
)
353 def get_element_by_class(class_name
, html
):
354 """Return the content of the first tag with the specified class in the passed HTML document"""
355 retval
= get_elements_by_class(class_name
, html
)
356 return retval
[0] if retval
else None
359 def get_element_html_by_class(class_name
, html
):
360 """Return the html of the first tag with the specified class in the passed HTML document"""
361 retval
= get_elements_html_by_class(class_name
, html
)
362 return retval
[0] if retval
else None
365 def get_element_by_attribute(attribute
, value
, html
, **kwargs
):
366 retval
= get_elements_by_attribute(attribute
, value
, html
, **kwargs
)
367 return retval
[0] if retval
else None
370 def get_element_html_by_attribute(attribute
, value
, html
, **kargs
):
371 retval
= get_elements_html_by_attribute(attribute
, value
, html
, **kargs
)
372 return retval
[0] if retval
else None
375 def get_elements_by_class(class_name
, html
, **kargs
):
376 """Return the content of all tags with the specified class in the passed HTML document as a list"""
377 return get_elements_by_attribute(
378 'class', r
'[^\'"]*(?<=[\'"\s
])%s(?
=[\'"\s])[^\'"]*' % re.escape(class_name),
379 html, escape_value=False)
382 def get_elements_html_by_class(class_name, html):
383 """Return the html of all tags with the specified class in the passed HTML document as a list"""
384 return get_elements_html_by_attribute(
385 'class', r'[^
\'"]*(?<=[\'"\s
])%s(?
=[\'"\s])[^\'"]*' % re.escape(class_name),
386 html, escape_value=False)
389 def get_elements_by_attribute(*args, **kwargs):
390 """Return the content of the tag with the specified attribute in the passed HTML document"""
391 return [content for content, _ in get_elements_text_and_html_by_attribute(*args, **kwargs)]
394 def get_elements_html_by_attribute(*args, **kwargs):
395 """Return the html of the tag with the specified attribute in the passed HTML document"""
396 return [whole for _, whole in get_elements_text_and_html_by_attribute(*args, **kwargs)]
399 def get_elements_text_and_html_by_attribute(attribute, value, html, escape_value=True):
401 Return the text (content) and the html (whole) of the tag with the specified
402 attribute in the passed HTML document
405 quote = '' if re.match(r'''[\s"'`
=<>]''', value) else '?'
407 value = re.escape(value) if escape_value else value
409 partial_element_re = rf'''(?x
)
410 <(?P
<tag
>[a
-zA
-Z0
-9:._-]+)
411 (?
:\
s(?
:[^
>"']|"[^
"]*"|
'[^']*')*)?
412 \s{re.escape(attribute)}\s*=\s*(?P<_q>['"]{quote})(?-x:{value})(?P=_q)
415 for m in re.finditer(partial_element_re, html):
416 content, whole = get_element_text_and_html_by_tag(m.group('tag'), html[m.start():])
419 unescapeHTML(re.sub(r'^(?P<q>["\'])(?P
<content
>.*)(?P
=q
)$
', r'\g
<content
>', content, flags=re.DOTALL)),
424 class HTMLBreakOnClosingTagParser(html.parser.HTMLParser):
426 HTML parser which raises HTMLBreakOnClosingTagException upon reaching the
427 closing tag for the first opening tag it has encountered, and can be used
431 class HTMLBreakOnClosingTagException(Exception):
435 self.tagstack = collections.deque()
436 html.parser.HTMLParser.__init__(self)
441 def __exit__(self, *_):
445 # handle_endtag does not return upon raising HTMLBreakOnClosingTagException,
446 # so data remains buffered; we no longer have any interest in it, thus
447 # override this method to discard it
450 def handle_starttag(self, tag, _):
451 self.tagstack.append(tag)
453 def handle_endtag(self, tag):
454 if not self.tagstack:
455 raise compat_HTMLParseError('no tags
in the stack
')
457 inner_tag = self.tagstack.pop()
461 raise compat_HTMLParseError(f'matching opening tag
for closing {tag} tag
not found
')
462 if not self.tagstack:
463 raise self.HTMLBreakOnClosingTagException()
466 def get_element_text_and_html_by_tag(tag, html):
468 For the first element with the specified tag in the passed HTML document
469 return its' content (text
) and the whole
element (html
)
471 def find_or_raise(haystack, needle, exc):
473 return haystack.index(needle)
476 closing_tag = f'</{tag}>'
477 whole_start = find_or_raise(
478 html, f'<{tag}', compat_HTMLParseError(f'opening {tag} tag not found'))
479 content_start = find_or_raise(
480 html[whole_start:], '>', compat_HTMLParseError(f'malformed opening {tag} tag'))
481 content_start += whole_start + 1
482 with HTMLBreakOnClosingTagParser() as parser:
483 parser.feed(html[whole_start:content_start])
484 if not parser.tagstack or parser.tagstack[0] != tag:
485 raise compat_HTMLParseError(f'parser did not match opening {tag} tag')
486 offset = content_start
487 while offset < len(html):
488 next_closing_tag_start = find_or_raise(
489 html[offset:], closing_tag,
490 compat_HTMLParseError(f'closing {tag} tag not found'))
491 next_closing_tag_end = next_closing_tag_start + len(closing_tag)
493 parser.feed(html[offset:offset + next_closing_tag_end])
494 offset += next_closing_tag_end
495 except HTMLBreakOnClosingTagParser.HTMLBreakOnClosingTagException:
496 return html[content_start:offset + next_closing_tag_start], \
497 html[whole_start:offset + next_closing_tag_end]
498 raise compat_HTMLParseError('unexpected end of html')
501 class HTMLAttributeParser(html.parser.HTMLParser):
502 """Trivial HTML parser to gather the attributes
for a single element
"""
506 html.parser.HTMLParser.__init__(self)
508 def handle_starttag(self, tag, attrs):
509 self.attrs = dict(attrs)
512 class HTMLListAttrsParser(html.parser.HTMLParser):
513 """HTML parser to gather the attributes
for the elements of a
list"""
516 html.parser.HTMLParser.__init__(self)
520 def handle_starttag(self, tag, attrs):
521 if tag == 'li' and self._level == 0:
522 self.items.append(dict(attrs))
525 def handle_endtag(self, tag):
529 def extract_attributes(html_element):
530 """Given a string
for an HTML element such
as
532 a
="foo" B
="bar" c
="&98;az" d
=boz
533 empty
= noval entity
="&"
536 Decode
and return a dictionary of attributes
.
538 'a': 'foo', 'b': 'bar', c
: 'baz', d
: 'boz',
539 'empty': '', 'noval': None, 'entity': '&',
540 'sq': '"', 'dq': '\''
543 parser = HTMLAttributeParser()
544 with contextlib.suppress(compat_HTMLParseError):
545 parser.feed(html_element)
550 def parse_list(webpage):
551 """Given a string
for an series of HTML
<li
> elements
,
552 return a dictionary of their attributes
"""
553 parser = HTMLListAttrsParser()
559 def clean_html(html):
560 """Clean an HTML snippet into a readable string
"""
562 if html is None: # Convenience for sanitizing descriptions etc.
565 html = re.sub(r'\s+', ' ', html)
566 html = re.sub(r'(?u)\s?<\s?br\s?/?\s?>\s?', '\n', html)
567 html = re.sub(r'(?u)<\s?/\s?p\s?>\s?<\s?p[^>]*>', '\n', html)
569 html = re.sub('<.*?>', '', html)
570 # Replace html entities
571 html = unescapeHTML(html)
575 class LenientJSONDecoder(json.JSONDecoder):
576 def __init__(self, *args, transform_source=None, ignore_extra=False, **kwargs):
577 self.transform_source, self.ignore_extra = transform_source, ignore_extra
578 super().__init__(*args, **kwargs)
581 if self.transform_source:
582 s = self.transform_source(s)
583 if self.ignore_extra:
584 return self.raw_decode(s.lstrip())[0]
585 return super().decode(s)
588 def sanitize_open(filename, open_mode):
589 """Try to
open the given filename
, and slightly tweak it
if this fails
.
591 Attempts to
open the given filename
. If this fails
, it tries to change
592 the filename slightly
, step by step
, until it
's either able to open it
593 or it fails and raises a final exception, like the standard open()
596 It returns the tuple (stream, definitive_file_name).
599 if sys.platform == 'win32
':
601 msvcrt.setmode(sys.stdout.fileno(), os.O_BINARY)
602 return (sys.stdout.buffer if hasattr(sys.stdout, 'buffer') else sys.stdout, filename)
604 for attempt in range(2):
607 if sys.platform == 'win32
':
608 # FIXME: An exclusive lock also locks the file from being read.
609 # Since windows locks are mandatory, don't lock the
file on
windows (for now
).
610 # Ref: https://github.com/yt-dlp/yt-dlp/issues/3124
611 raise LockingUnsupportedError()
612 stream
= locked_file(filename
, open_mode
, block
=False).__enter
__()
614 stream
= open(filename
, open_mode
)
615 return stream
, filename
616 except OSError as err
:
617 if attempt
or err
.errno
in (errno
.EACCES
,):
619 old_filename
, filename
= filename
, sanitize_path(filename
)
620 if old_filename
== filename
:
624 def timeconvert(timestr
):
625 """Convert RFC 2822 defined time string into system timestamp"""
627 timetuple
= email
.utils
.parsedate_tz(timestr
)
628 if timetuple
is not None:
629 timestamp
= email
.utils
.mktime_tz(timetuple
)
633 def sanitize_filename(s
, restricted
=False, is_id
=NO_DEFAULT
):
634 """Sanitizes a string so it could be used as part of a filename.
635 @param restricted Use a stricter subset of allowed characters
636 @param is_id Whether this is an ID that should be kept unchanged if possible.
637 If unset, yt-dlp's new sanitization rules are in effect
642 def replace_insane(char
):
643 if restricted
and char
in ACCENT_CHARS
:
644 return ACCENT_CHARS
[char
]
645 elif not restricted
and char
== '\n':
647 elif char
== '?' or ord(char
) < 32 or ord(char
) == 127:
650 return '' if restricted
else '\''
652 return '\0_\0-' if restricted
else '\0 \0-'
653 elif char
in '\\/|*<>':
655 if restricted
and (char
in '!&\'()[]{}$;`^,#' or char
.isspace() or ord(char
) > 127):
659 s
= re
.sub(r
'[0-9]+(?::[0-9]+)+', lambda m
: m
.group(0).replace(':', '_'), s
) # Handle timestamps
660 result
= ''.join(map(replace_insane
, s
))
661 if is_id
is NO_DEFAULT
:
662 result
= re
.sub(r
'(\0.)(?:(?=\1)..)+', r
'\1', result
) # Remove repeated substitute chars
663 STRIP_RE
= r
'(?:\0.|[ _-])*'
664 result
= re
.sub(f
'^\0.{STRIP_RE}|{STRIP_RE}\0.$', '', result
) # Remove substitute chars from start/end
665 result
= result
.replace('\0', '') or '_'
668 while '__' in result
:
669 result
= result
.replace('__', '_')
670 result
= result
.strip('_')
671 # Common case of "Foreign band name - English song title"
672 if restricted
and result
.startswith('-_'):
674 if result
.startswith('-'):
675 result
= '_' + result
[len('-'):]
676 result
= result
.lstrip('.')
682 def sanitize_path(s
, force
=False):
683 """Sanitizes and normalizes path on Windows"""
684 if sys
.platform
== 'win32':
686 drive_or_unc
, _
= os
.path
.splitdrive(s
)
692 norm_path
= os
.path
.normpath(remove_start(s
, drive_or_unc
)).split(os
.path
.sep
)
696 path_part
if path_part
in ['.', '..'] else re
.sub(r
'(?:[/<>:"\|\\?\*]|[\s.]$)', '#', path_part
)
697 for path_part
in norm_path
]
699 sanitized_path
.insert(0, drive_or_unc
+ os
.path
.sep
)
700 elif force
and s
and s
[0] == os
.path
.sep
:
701 sanitized_path
.insert(0, os
.path
.sep
)
702 return os
.path
.join(*sanitized_path
)
705 def sanitize_url(url
):
706 # Prepend protocol-less URLs with `http:` scheme in order to mitigate
707 # the number of unwanted failures due to missing protocol
710 elif url
.startswith('//'):
711 return 'http:%s' % url
712 # Fix some common typos seen so far
714 # https://github.com/ytdl-org/youtube-dl/issues/15649
715 (r
'^httpss://', r
'https://'),
716 # https://bx1.be/lives/direct-tv/
717 (r
'^rmtp([es]?)://', r
'rtmp\1://'),
719 for mistake
, fixup
in COMMON_TYPOS
:
720 if re
.match(mistake
, url
):
721 return re
.sub(mistake
, fixup
, url
)
725 def extract_basic_auth(url
):
726 parts
= urllib
.parse
.urlsplit(url
)
727 if parts
.username
is None:
729 url
= urllib
.parse
.urlunsplit(parts
._replace
(netloc
=(
730 parts
.hostname
if parts
.port
is None
731 else '%s:%d' % (parts
.hostname
, parts
.port
))))
732 auth_payload
= base64
.b64encode(
733 ('%s:%s' % (parts
.username
, parts
.password
or '')).encode())
734 return url
, f
'Basic {auth_payload.decode()}'
737 def sanitized_Request(url
, *args
, **kwargs
):
738 url
, auth_header
= extract_basic_auth(escape_url(sanitize_url(url
)))
739 if auth_header
is not None:
740 headers
= args
[1] if len(args
) >= 2 else kwargs
.setdefault('headers', {})
741 headers
['Authorization'] = auth_header
742 return urllib
.request
.Request(url
, *args
, **kwargs
)
746 """Expand shell variables and ~"""
747 return os
.path
.expandvars(compat_expanduser(s
))
750 def orderedSet(iterable
, *, lazy
=False):
751 """Remove all duplicates from the input iterable"""
753 seen
= [] # Do not use set since the items can be unhashable
759 return _iter() if lazy
else list(_iter())
762 def _htmlentity_transform(entity_with_semicolon
):
763 """Transforms an HTML entity to a character."""
764 entity
= entity_with_semicolon
[:-1]
766 # Known non-numeric HTML entity
767 if entity
in html
.entities
.name2codepoint
:
768 return chr(html
.entities
.name2codepoint
[entity
])
770 # TODO: HTML5 allows entities without a semicolon. For example,
771 # 'Éric' should be decoded as 'Éric'.
772 if entity_with_semicolon
in html
.entities
.html5
:
773 return html
.entities
.html5
[entity_with_semicolon
]
775 mobj
= re
.match(r
'#(x[0-9a-fA-F]+|[0-9]+)', entity
)
777 numstr
= mobj
.group(1)
778 if numstr
.startswith('x'):
780 numstr
= '0%s' % numstr
783 # See https://github.com/ytdl-org/youtube-dl/issues/7518
784 with contextlib
.suppress(ValueError):
785 return chr(int(numstr
, base
))
787 # Unknown entity in name, return its literal representation
788 return '&%s;' % entity
794 assert isinstance(s
, str)
797 r
'&([^&;]+;)', lambda m
: _htmlentity_transform(m
.group(1)), s
)
800 def escapeHTML(text
):
803 .replace('&', '&')
804 .replace('<', '<')
805 .replace('>', '>')
806 .replace('"', '"')
807 .replace("'", ''')
811 def process_communicate_or_kill(p
, *args
, **kwargs
):
812 write_string('DeprecationWarning: yt_dlp.utils.process_communicate_or_kill is deprecated '
813 'and may be removed in a future version. Use yt_dlp.utils.Popen.communicate_or_kill instead')
814 return Popen
.communicate_or_kill(p
, *args
, **kwargs
)
817 class Popen(subprocess
.Popen
):
818 if sys
.platform
== 'win32':
819 _startupinfo
= subprocess
.STARTUPINFO()
820 _startupinfo
.dwFlags |
= subprocess
.STARTF_USESHOWWINDOW
824 def __init__(self
, *args
, text
=False, **kwargs
):
826 kwargs
['universal_newlines'] = True # For 3.6 compatibility
827 kwargs
.setdefault('encoding', 'utf-8')
828 kwargs
.setdefault('errors', 'replace')
829 super().__init
__(*args
, **kwargs
, startupinfo
=self
._startupinfo
)
831 def communicate_or_kill(self
, *args
, **kwargs
):
833 return self
.communicate(*args
, **kwargs
)
834 except BaseException
: # Including KeyboardInterrupt
835 self
.kill(timeout
=None)
838 def kill(self
, *, timeout
=0):
841 self
.wait(timeout
=timeout
)
844 def run(cls
, *args
, **kwargs
):
845 with cls(*args
, **kwargs
) as proc
:
846 stdout
, stderr
= proc
.communicate_or_kill()
847 return stdout
or '', stderr
or '', proc
.returncode
850 def get_subprocess_encoding():
851 if sys
.platform
== 'win32' and sys
.getwindowsversion()[0] >= 5:
852 # For subprocess calls, encode with locale encoding
853 # Refer to http://stackoverflow.com/a/9951851/35070
854 encoding
= preferredencoding()
856 encoding
= sys
.getfilesystemencoding()
862 def encodeFilename(s
, for_subprocess
=False):
863 assert isinstance(s
, str)
867 def decodeFilename(b
, for_subprocess
=False):
871 def encodeArgument(s
):
872 # Legacy code that uses byte strings
873 # Uncomment the following line after fixing all post processors
874 # assert isinstance(s, str), 'Internal error: %r should be of type %r, is %r' % (s, str, type(s))
875 return s
if isinstance(s
, str) else s
.decode('ascii')
878 def decodeArgument(b
):
882 def decodeOption(optval
):
885 if isinstance(optval
, bytes):
886 optval
= optval
.decode(preferredencoding())
888 assert isinstance(optval
, str)
892 _timetuple
= collections
.namedtuple('Time', ('hours', 'minutes', 'seconds', 'milliseconds'))
895 def timetuple_from_msec(msec
):
896 secs
, msec
= divmod(msec
, 1000)
897 mins
, secs
= divmod(secs
, 60)
898 hrs
, mins
= divmod(mins
, 60)
899 return _timetuple(hrs
, mins
, secs
, msec
)
902 def formatSeconds(secs
, delim
=':', msec
=False):
903 time
= timetuple_from_msec(secs
* 1000)
905 ret
= '%d%s%02d%s%02d' % (time
.hours
, delim
, time
.minutes
, delim
, time
.seconds
)
907 ret
= '%d%s%02d' % (time
.minutes
, delim
, time
.seconds
)
909 ret
= '%d' % time
.seconds
910 return '%s.%03d' % (ret
, time
.milliseconds
) if msec
else ret
913 def _ssl_load_windows_store_certs(ssl_context
, storename
):
914 # Code adapted from _load_windows_store_certs in https://github.com/python/cpython/blob/main/Lib/ssl.py
916 certs
= [cert
for cert
, encoding
, trust
in ssl
.enum_certificates(storename
)
917 if encoding
== 'x509_asn' and (
918 trust
is True or ssl
.Purpose
.SERVER_AUTH
.oid
in trust
)]
919 except PermissionError
:
922 with contextlib
.suppress(ssl
.SSLError
):
923 ssl_context
.load_verify_locations(cadata
=cert
)
926 def make_HTTPS_handler(params
, **kwargs
):
927 opts_check_certificate
= not params
.get('nocheckcertificate')
928 context
= ssl
.SSLContext(ssl
.PROTOCOL_TLS_CLIENT
)
929 context
.check_hostname
= opts_check_certificate
930 if params
.get('legacyserverconnect'):
931 context
.options |
= 4 # SSL_OP_LEGACY_SERVER_CONNECT
932 # Allow use of weaker ciphers in Python 3.10+. See https://bugs.python.org/issue43998
933 context
.set_ciphers('DEFAULT')
935 context
.verify_mode
= ssl
.CERT_REQUIRED
if opts_check_certificate
else ssl
.CERT_NONE
936 if opts_check_certificate
:
937 if has_certifi
and 'no-certifi' not in params
.get('compat_opts', []):
938 context
.load_verify_locations(cafile
=certifi
.where())
941 context
.load_default_certs()
942 # Work around the issue in load_default_certs when there are bad certificates. See:
943 # https://github.com/yt-dlp/yt-dlp/issues/1060,
944 # https://bugs.python.org/issue35665, https://bugs.python.org/issue45312
946 # enum_certificates is not present in mingw python. See https://github.com/yt-dlp/yt-dlp/issues/1151
947 if sys
.platform
== 'win32' and hasattr(ssl
, 'enum_certificates'):
948 for storename
in ('CA', 'ROOT'):
949 _ssl_load_windows_store_certs(context
, storename
)
950 context
.set_default_verify_paths()
952 client_certfile
= params
.get('client_certificate')
955 context
.load_cert_chain(
956 client_certfile
, keyfile
=params
.get('client_certificate_key'),
957 password
=params
.get('client_certificate_password'))
959 raise YoutubeDLError('Unable to load client certificate')
961 # Some servers may reject requests if ALPN extension is not sent. See:
962 # https://github.com/python/cpython/issues/85140
963 # https://github.com/yt-dlp/yt-dlp/issues/3878
964 with contextlib
.suppress(NotImplementedError):
965 context
.set_alpn_protocols(['http/1.1'])
967 return YoutubeDLHTTPSHandler(params
, context
=context
, **kwargs
)
970 def bug_reports_message(before
=';'):
971 from .update
import REPOSITORY
973 msg
= (f
'please report this issue on https://github.com/{REPOSITORY}/issues?q= , '
974 'filling out the appropriate issue template. Confirm you are on the latest version using yt-dlp -U')
976 before
= before
.rstrip()
977 if not before
or before
.endswith(('.', '!', '?')):
978 msg
= msg
[0].title() + msg
[1:]
980 return (before
+ ' ' if before
else '') + msg
983 class YoutubeDLError(Exception):
984 """Base exception for YoutubeDL errors."""
987 def __init__(self
, msg
=None):
990 elif self
.msg
is None:
991 self
.msg
= type(self
).__name
__
992 super().__init
__(self
.msg
)
995 network_exceptions
= [urllib
.error
.URLError
, http
.client
.HTTPException
, socket
.error
]
996 if hasattr(ssl
, 'CertificateError'):
997 network_exceptions
.append(ssl
.CertificateError
)
998 network_exceptions
= tuple(network_exceptions
)
1001 class ExtractorError(YoutubeDLError
):
1002 """Error during info extraction."""
1004 def __init__(self
, msg
, tb
=None, expected
=False, cause
=None, video_id
=None, ie
=None):
1005 """ tb, if given, is the original traceback (so that it can be printed out).
1006 If expected is set, this is a normal error message and most likely not a bug in yt-dlp.
1008 if sys
.exc_info()[0] in network_exceptions
:
1011 self
.orig_msg
= str(msg
)
1013 self
.expected
= expected
1015 self
.video_id
= video_id
1017 self
.exc_info
= sys
.exc_info() # preserve original exception
1018 if isinstance(self
.exc_info
[1], ExtractorError
):
1019 self
.exc_info
= self
.exc_info
[1].exc_info
1021 super().__init
__(''.join((
1022 format_field(ie
, None, '[%s] '),
1023 format_field(video_id
, None, '%s: '),
1025 format_field(cause
, None, ' (caused by %r)'),
1026 '' if expected
else bug_reports_message())))
1028 def format_traceback(self
):
1029 return join_nonempty(
1030 self
.traceback
and ''.join(traceback
.format_tb(self
.traceback
)),
1031 self
.cause
and ''.join(traceback
.format_exception(None, self
.cause
, self
.cause
.__traceback
__)[1:]),
1035 class UnsupportedError(ExtractorError
):
1036 def __init__(self
, url
):
1038 'Unsupported URL: %s' % url
, expected
=True)
1042 class RegexNotFoundError(ExtractorError
):
1043 """Error when a regex didn't match"""
1047 class GeoRestrictedError(ExtractorError
):
1048 """Geographic restriction Error exception.
1050 This exception may be thrown when a video is not available from your
1051 geographic location due to geographic restrictions imposed by a website.
1054 def __init__(self
, msg
, countries
=None, **kwargs
):
1055 kwargs
['expected'] = True
1056 super().__init
__(msg
, **kwargs
)
1057 self
.countries
= countries
1060 class UserNotLive(ExtractorError
):
1061 """Error when a channel/user is not live"""
1063 def __init__(self
, msg
=None, **kwargs
):
1064 kwargs
['expected'] = True
1065 super().__init
__(msg
or 'The channel is not currently live', **kwargs
)
1068 class DownloadError(YoutubeDLError
):
1069 """Download Error exception.
1071 This exception may be thrown by FileDownloader objects if they are not
1072 configured to continue on errors. They will contain the appropriate
1076 def __init__(self
, msg
, exc_info
=None):
1077 """ exc_info, if given, is the original exception that caused the trouble (as returned by sys.exc_info()). """
1078 super().__init
__(msg
)
1079 self
.exc_info
= exc_info
1082 class EntryNotInPlaylist(YoutubeDLError
):
1083 """Entry not in playlist exception.
1085 This exception will be thrown by YoutubeDL when a requested entry
1086 is not found in the playlist info_dict
1088 msg
= 'Entry not found in info'
1091 class SameFileError(YoutubeDLError
):
1092 """Same File exception.
1094 This exception will be thrown by FileDownloader objects if they detect
1095 multiple files would have to be downloaded to the same file on disk.
1097 msg
= 'Fixed output name but more than one file to download'
1099 def __init__(self
, filename
=None):
1100 if filename
is not None:
1101 self
.msg
+= f
': {filename}'
1102 super().__init
__(self
.msg
)
1105 class PostProcessingError(YoutubeDLError
):
1106 """Post Processing exception.
1108 This exception may be raised by PostProcessor's .run() method to
1109 indicate an error in the postprocessing task.
1113 class DownloadCancelled(YoutubeDLError
):
1114 """ Exception raised when the download queue should be interrupted """
1115 msg
= 'The download was cancelled'
1118 class ExistingVideoReached(DownloadCancelled
):
1119 """ --break-on-existing triggered """
1120 msg
= 'Encountered a video that is already in the archive, stopping due to --break-on-existing'
1123 class RejectedVideoReached(DownloadCancelled
):
1124 """ --break-on-reject triggered """
1125 msg
= 'Encountered a video that did not match filter, stopping due to --break-on-reject'
1128 class MaxDownloadsReached(DownloadCancelled
):
1129 """ --max-downloads limit has been reached. """
1130 msg
= 'Maximum number of downloads reached, stopping due to --max-downloads'
1133 class ReExtractInfo(YoutubeDLError
):
1134 """ Video info needs to be re-extracted. """
1136 def __init__(self
, msg
, expected
=False):
1137 super().__init
__(msg
)
1138 self
.expected
= expected
1141 class ThrottledDownload(ReExtractInfo
):
1142 """ Download speed below --throttled-rate. """
1143 msg
= 'The download speed is below throttle limit'
1146 super().__init
__(self
.msg
, expected
=False)
1149 class UnavailableVideoError(YoutubeDLError
):
1150 """Unavailable Format exception.
1152 This exception will be thrown when a video is requested
1153 in a format that is not available for that video.
1155 msg
= 'Unable to download video'
1157 def __init__(self
, err
=None):
1159 self
.msg
+= f
': {err}'
1160 super().__init
__(self
.msg
)
1163 class ContentTooShortError(YoutubeDLError
):
1164 """Content Too Short exception.
1166 This exception may be raised by FileDownloader objects when a file they
1167 download is too small for what the server announced first, indicating
1168 the connection was probably interrupted.
1171 def __init__(self
, downloaded
, expected
):
1172 super().__init
__(f
'Downloaded {downloaded} bytes, expected {expected} bytes')
1174 self
.downloaded
= downloaded
1175 self
.expected
= expected
1178 class XAttrMetadataError(YoutubeDLError
):
1179 def __init__(self
, code
=None, msg
='Unknown error'):
1180 super().__init
__(msg
)
1184 # Parsing code and msg
1185 if (self
.code
in (errno
.ENOSPC
, errno
.EDQUOT
)
1186 or 'No space left' in self
.msg
or 'Disk quota exceeded' in self
.msg
):
1187 self
.reason
= 'NO_SPACE'
1188 elif self
.code
== errno
.E2BIG
or 'Argument list too long' in self
.msg
:
1189 self
.reason
= 'VALUE_TOO_LONG'
1191 self
.reason
= 'NOT_SUPPORTED'
1194 class XAttrUnavailableError(YoutubeDLError
):
1198 def _create_http_connection(ydl_handler
, http_class
, is_https
, *args
, **kwargs
):
1199 hc
= http_class(*args
, **kwargs
)
1200 source_address
= ydl_handler
._params
.get('source_address')
1202 if source_address
is not None:
1203 # This is to workaround _create_connection() from socket where it will try all
1204 # address data from getaddrinfo() including IPv6. This filters the result from
1205 # getaddrinfo() based on the source_address value.
1206 # This is based on the cpython socket.create_connection() function.
1207 # https://github.com/python/cpython/blob/master/Lib/socket.py#L691
1208 def _create_connection(address
, timeout
=socket
._GLOBAL
_DEFAULT
_TIMEOUT
, source_address
=None):
1209 host
, port
= address
1211 addrs
= socket
.getaddrinfo(host
, port
, 0, socket
.SOCK_STREAM
)
1212 af
= socket
.AF_INET
if '.' in source_address
[0] else socket
.AF_INET6
1213 ip_addrs
= [addr
for addr
in addrs
if addr
[0] == af
]
1214 if addrs
and not ip_addrs
:
1215 ip_version
= 'v4' if af
== socket
.AF_INET
else 'v6'
1217 "No remote IP%s addresses available for connect, can't use '%s' as source address"
1218 % (ip_version
, source_address
[0]))
1219 for res
in ip_addrs
:
1220 af
, socktype
, proto
, canonname
, sa
= res
1223 sock
= socket
.socket(af
, socktype
, proto
)
1224 if timeout
is not socket
._GLOBAL
_DEFAULT
_TIMEOUT
:
1225 sock
.settimeout(timeout
)
1226 sock
.bind(source_address
)
1228 err
= None # Explicitly break reference cycle
1230 except OSError as _
:
1232 if sock
is not None:
1237 raise OSError('getaddrinfo returns an empty list')
1238 if hasattr(hc
, '_create_connection'):
1239 hc
._create
_connection
= _create_connection
1240 hc
.source_address
= (source_address
, 0)
1245 def handle_youtubedl_headers(headers
):
1246 filtered_headers
= headers
1248 if 'Youtubedl-no-compression' in filtered_headers
:
1249 filtered_headers
= {k: v for k, v in filtered_headers.items() if k.lower() != 'accept-encoding'}
1250 del filtered_headers
['Youtubedl-no-compression']
1252 return filtered_headers
1255 class YoutubeDLHandler(urllib
.request
.HTTPHandler
):
1256 """Handler for HTTP requests and responses.
1258 This class, when installed with an OpenerDirector, automatically adds
1259 the standard headers to every HTTP request and handles gzipped and
1260 deflated responses from web servers. If compression is to be avoided in
1261 a particular request, the original request in the program code only has
1262 to include the HTTP header "Youtubedl-no-compression", which will be
1263 removed before making the real request.
1265 Part of this code was copied from:
1267 http://techknack.net/python-urllib2-handlers/
1269 Andrew Rowls, the author of that code, agreed to release it to the
1273 def __init__(self
, params
, *args
, **kwargs
):
1274 urllib
.request
.HTTPHandler
.__init
__(self
, *args
, **kwargs
)
1275 self
._params
= params
1277 def http_open(self
, req
):
1278 conn_class
= http
.client
.HTTPConnection
1280 socks_proxy
= req
.headers
.get('Ytdl-socks-proxy')
1282 conn_class
= make_socks_conn_class(conn_class
, socks_proxy
)
1283 del req
.headers
['Ytdl-socks-proxy']
1285 return self
.do_open(functools
.partial(
1286 _create_http_connection
, self
, conn_class
, False),
1294 return zlib
.decompress(data
, -zlib
.MAX_WBITS
)
1296 return zlib
.decompress(data
)
1302 return brotli
.decompress(data
)
1304 def http_request(self
, req
):
1305 # According to RFC 3986, URLs can not contain non-ASCII characters, however this is not
1306 # always respected by websites, some tend to give out URLs with non percent-encoded
1307 # non-ASCII characters (see telemb.py, ard.py [#3412])
1308 # urllib chokes on URLs with non-ASCII characters (see http://bugs.python.org/issue3991)
1309 # To work around aforementioned issue we will replace request's original URL with
1310 # percent-encoded one
1311 # Since redirects are also affected (e.g. http://www.southpark.de/alle-episoden/s18e09)
1312 # the code of this workaround has been moved here from YoutubeDL.urlopen()
1313 url
= req
.get_full_url()
1314 url_escaped
= escape_url(url
)
1316 # Substitute URL if any change after escaping
1317 if url
!= url_escaped
:
1318 req
= update_Request(req
, url
=url_escaped
)
1320 for h
, v
in self
._params
.get('http_headers', std_headers
).items():
1321 # Capitalize is needed because of Python bug 2275: http://bugs.python.org/issue2275
1322 # The dict keys are capitalized because of this bug by urllib
1323 if h
.capitalize() not in req
.headers
:
1324 req
.add_header(h
, v
)
1326 if 'Accept-encoding' not in req
.headers
:
1327 req
.add_header('Accept-encoding', ', '.join(SUPPORTED_ENCODINGS
))
1329 req
.headers
= handle_youtubedl_headers(req
.headers
)
1331 return super().do_request_(req
)
1333 def http_response(self
, req
, resp
):
1336 if resp
.headers
.get('Content-encoding', '') == 'gzip':
1337 content
= resp
.read()
1338 gz
= gzip
.GzipFile(fileobj
=io
.BytesIO(content
), mode
='rb')
1340 uncompressed
= io
.BytesIO(gz
.read())
1341 except OSError as original_ioerror
:
1342 # There may be junk add the end of the file
1343 # See http://stackoverflow.com/q/4928560/35070 for details
1344 for i
in range(1, 1024):
1346 gz
= gzip
.GzipFile(fileobj
=io
.BytesIO(content
[:-i
]), mode
='rb')
1347 uncompressed
= io
.BytesIO(gz
.read())
1352 raise original_ioerror
1353 resp
= urllib
.request
.addinfourl(uncompressed
, old_resp
.headers
, old_resp
.url
, old_resp
.code
)
1354 resp
.msg
= old_resp
.msg
1355 del resp
.headers
['Content-encoding']
1357 if resp
.headers
.get('Content-encoding', '') == 'deflate':
1358 gz
= io
.BytesIO(self
.deflate(resp
.read()))
1359 resp
= urllib
.request
.addinfourl(gz
, old_resp
.headers
, old_resp
.url
, old_resp
.code
)
1360 resp
.msg
= old_resp
.msg
1361 del resp
.headers
['Content-encoding']
1363 if resp
.headers
.get('Content-encoding', '') == 'br':
1364 resp
= urllib
.request
.addinfourl(
1365 io
.BytesIO(self
.brotli(resp
.read())), old_resp
.headers
, old_resp
.url
, old_resp
.code
)
1366 resp
.msg
= old_resp
.msg
1367 del resp
.headers
['Content-encoding']
1368 # Percent-encode redirect URL of Location HTTP header to satisfy RFC 3986 (see
1369 # https://github.com/ytdl-org/youtube-dl/issues/6457).
1370 if 300 <= resp
.code
< 400:
1371 location
= resp
.headers
.get('Location')
1373 # As of RFC 2616 default charset is iso-8859-1 that is respected by python 3
1374 location
= location
.encode('iso-8859-1').decode()
1375 location_escaped
= escape_url(location
)
1376 if location
!= location_escaped
:
1377 del resp
.headers
['Location']
1378 resp
.headers
['Location'] = location_escaped
1381 https_request
= http_request
1382 https_response
= http_response
1385 def make_socks_conn_class(base_class
, socks_proxy
):
1386 assert issubclass(base_class
, (
1387 http
.client
.HTTPConnection
, http
.client
.HTTPSConnection
))
1389 url_components
= urllib
.parse
.urlparse(socks_proxy
)
1390 if url_components
.scheme
.lower() == 'socks5':
1391 socks_type
= ProxyType
.SOCKS5
1392 elif url_components
.scheme
.lower() in ('socks', 'socks4'):
1393 socks_type
= ProxyType
.SOCKS4
1394 elif url_components
.scheme
.lower() == 'socks4a':
1395 socks_type
= ProxyType
.SOCKS4A
1397 def unquote_if_non_empty(s
):
1400 return urllib
.parse
.unquote_plus(s
)
1404 url_components
.hostname
, url_components
.port
or 1080,
1406 unquote_if_non_empty(url_components
.username
),
1407 unquote_if_non_empty(url_components
.password
),
1410 class SocksConnection(base_class
):
1412 self
.sock
= sockssocket()
1413 self
.sock
.setproxy(*proxy_args
)
1414 if isinstance(self
.timeout
, (int, float)):
1415 self
.sock
.settimeout(self
.timeout
)
1416 self
.sock
.connect((self
.host
, self
.port
))
1418 if isinstance(self
, http
.client
.HTTPSConnection
):
1419 if hasattr(self
, '_context'): # Python > 2.6
1420 self
.sock
= self
._context
.wrap_socket(
1421 self
.sock
, server_hostname
=self
.host
)
1423 self
.sock
= ssl
.wrap_socket(self
.sock
)
1425 return SocksConnection
1428 class YoutubeDLHTTPSHandler(urllib
.request
.HTTPSHandler
):
1429 def __init__(self
, params
, https_conn_class
=None, *args
, **kwargs
):
1430 urllib
.request
.HTTPSHandler
.__init
__(self
, *args
, **kwargs
)
1431 self
._https
_conn
_class
= https_conn_class
or http
.client
.HTTPSConnection
1432 self
._params
= params
1434 def https_open(self
, req
):
1436 conn_class
= self
._https
_conn
_class
1438 if hasattr(self
, '_context'): # python > 2.6
1439 kwargs
['context'] = self
._context
1440 if hasattr(self
, '_check_hostname'): # python 3.x
1441 kwargs
['check_hostname'] = self
._check
_hostname
1443 socks_proxy
= req
.headers
.get('Ytdl-socks-proxy')
1445 conn_class
= make_socks_conn_class(conn_class
, socks_proxy
)
1446 del req
.headers
['Ytdl-socks-proxy']
1449 return self
.do_open(
1450 functools
.partial(_create_http_connection
, self
, conn_class
, True), req
, **kwargs
)
1451 except urllib
.error
.URLError
as e
:
1452 if (isinstance(e
.reason
, ssl
.SSLError
)
1453 and getattr(e
.reason
, 'reason', None) == 'SSLV3_ALERT_HANDSHAKE_FAILURE'):
1454 raise YoutubeDLError('SSLV3_ALERT_HANDSHAKE_FAILURE: Try using --legacy-server-connect')
1458 class YoutubeDLCookieJar(http
.cookiejar
.MozillaCookieJar
):
1460 See [1] for cookie file format.
1462 1. https://curl.haxx.se/docs/http-cookies.html
1464 _HTTPONLY_PREFIX
= '#HttpOnly_'
1466 _HEADER
= '''# Netscape HTTP Cookie File
1467 # This file is generated by yt-dlp. Do not edit.
1470 _CookieFileEntry
= collections
.namedtuple(
1472 ('domain_name', 'include_subdomains', 'path', 'https_only', 'expires_at', 'name', 'value'))
1474 def __init__(self
, filename
=None, *args
, **kwargs
):
1475 super().__init
__(None, *args
, **kwargs
)
1476 if self
.is_path(filename
):
1477 filename
= os
.fspath(filename
)
1478 self
.filename
= filename
1481 def _true_or_false(cndn
):
1482 return 'TRUE' if cndn
else 'FALSE'
1486 return isinstance(file, (str, bytes, os
.PathLike
))
1488 @contextlib.contextmanager
1489 def open(self
, file, *, write
=False):
1490 if self
.is_path(file):
1491 with open(file, 'w' if write
else 'r', encoding
='utf-8') as f
:
1498 def _really_save(self
, f
, ignore_discard
=False, ignore_expires
=False):
1501 if (not ignore_discard
and cookie
.discard
1502 or not ignore_expires
and cookie
.is_expired(now
)):
1504 name
, value
= cookie
.name
, cookie
.value
1506 # cookies.txt regards 'Set-Cookie: foo' as a cookie
1507 # with no name, whereas http.cookiejar regards it as a
1508 # cookie with no value.
1509 name
, value
= '', name
1510 f
.write('%s\n' % '\t'.join((
1512 self
._true
_or
_false
(cookie
.domain
.startswith('.')),
1514 self
._true
_or
_false
(cookie
.secure
),
1515 str_or_none(cookie
.expires
, default
=''),
1519 def save(self
, filename
=None, *args
, **kwargs
):
1521 Save cookies to a file.
1522 Code is taken from CPython 3.6
1523 https://github.com/python/cpython/blob/8d999cbf4adea053be6dbb612b9844635c4dfb8e/Lib/http/cookiejar.py#L2091-L2117 """
1525 if filename
is None:
1526 if self
.filename
is not None:
1527 filename
= self
.filename
1529 raise ValueError(http
.cookiejar
.MISSING_FILENAME_TEXT
)
1531 # Store session cookies with `expires` set to 0 instead of an empty string
1533 if cookie
.expires
is None:
1536 with self
.open(filename
, write
=True) as f
:
1537 f
.write(self
._HEADER
)
1538 self
._really
_save
(f
, *args
, **kwargs
)
1540 def load(self
, filename
=None, ignore_discard
=False, ignore_expires
=False):
1541 """Load cookies from a file."""
1542 if filename
is None:
1543 if self
.filename
is not None:
1544 filename
= self
.filename
1546 raise ValueError(http
.cookiejar
.MISSING_FILENAME_TEXT
)
1548 def prepare_line(line
):
1549 if line
.startswith(self
._HTTPONLY
_PREFIX
):
1550 line
= line
[len(self
._HTTPONLY
_PREFIX
):]
1551 # comments and empty lines are fine
1552 if line
.startswith('#') or not line
.strip():
1554 cookie_list
= line
.split('\t')
1555 if len(cookie_list
) != self
._ENTRY
_LEN
:
1556 raise http
.cookiejar
.LoadError('invalid length %d' % len(cookie_list
))
1557 cookie
= self
._CookieFileEntry
(*cookie_list
)
1558 if cookie
.expires_at
and not cookie
.expires_at
.isdigit():
1559 raise http
.cookiejar
.LoadError('invalid expires at %s' % cookie
.expires_at
)
1563 with self
.open(filename
) as f
:
1566 cf
.write(prepare_line(line
))
1567 except http
.cookiejar
.LoadError
as e
:
1568 if f
'{line.strip()} '[0] in '[{"':
1569 raise http
.cookiejar
.LoadError(
1570 'Cookies file must be Netscape formatted, not JSON. See '
1571 'https://github.com/ytdl-org/youtube-dl#how-do-i-pass-cookies-to-youtube-dl')
1572 write_string(f
'WARNING: skipping cookie file entry due to {e}: {line!r}\n')
1575 self
._really
_load
(cf
, filename
, ignore_discard
, ignore_expires
)
1576 # Session cookies are denoted by either `expires` field set to
1577 # an empty string or 0. MozillaCookieJar only recognizes the former
1578 # (see [1]). So we need force the latter to be recognized as session
1579 # cookies on our own.
1580 # Session cookies may be important for cookies-based authentication,
1581 # e.g. usually, when user does not check 'Remember me' check box while
1582 # logging in on a site, some important cookies are stored as session
1583 # cookies so that not recognizing them will result in failed login.
1584 # 1. https://bugs.python.org/issue17164
1586 # Treat `expires=0` cookies as session cookies
1587 if cookie
.expires
== 0:
1588 cookie
.expires
= None
1589 cookie
.discard
= True
1592 class YoutubeDLCookieProcessor(urllib
.request
.HTTPCookieProcessor
):
1593 def __init__(self
, cookiejar
=None):
1594 urllib
.request
.HTTPCookieProcessor
.__init
__(self
, cookiejar
)
1596 def http_response(self
, request
, response
):
1597 return urllib
.request
.HTTPCookieProcessor
.http_response(self
, request
, response
)
1599 https_request
= urllib
.request
.HTTPCookieProcessor
.http_request
1600 https_response
= http_response
1603 class YoutubeDLRedirectHandler(urllib
.request
.HTTPRedirectHandler
):
1604 """YoutubeDL redirect handler
1606 The code is based on HTTPRedirectHandler implementation from CPython [1].
1608 This redirect handler solves two issues:
1609 - ensures redirect URL is always unicode under python 2
1610 - introduces support for experimental HTTP response status code
1611 308 Permanent Redirect [2] used by some sites [3]
1613 1. https://github.com/python/cpython/blob/master/Lib/urllib/request.py
1614 2. https://developer.mozilla.org/en-US/docs/Web/HTTP/Status/308
1615 3. https://github.com/ytdl-org/youtube-dl/issues/28768
1618 http_error_301
= http_error_303
= http_error_307
= http_error_308
= urllib
.request
.HTTPRedirectHandler
.http_error_302
1620 def redirect_request(self
, req
, fp
, code
, msg
, headers
, newurl
):
1621 """Return a Request or None in response to a redirect.
1623 This is called by the http_error_30x methods when a
1624 redirection response is received. If a redirection should
1625 take place, return a new Request to allow http_error_30x to
1626 perform the redirect. Otherwise, raise HTTPError if no-one
1627 else should try to handle this url. Return None if you can't
1628 but another Handler might.
1630 m
= req
.get_method()
1631 if (not (code
in (301, 302, 303, 307, 308) and m
in ("GET", "HEAD")
1632 or code
in (301, 302, 303) and m
== "POST")):
1633 raise urllib
.error
.HTTPError(req
.full_url
, code
, msg
, headers
, fp
)
1634 # Strictly (according to RFC 2616), 301 or 302 in response to
1635 # a POST MUST NOT cause a redirection without confirmation
1636 # from the user (of urllib.request, in this case). In practice,
1637 # essentially all clients do redirect in this case, so we do
1640 # Be conciliant with URIs containing a space. This is mainly
1641 # redundant with the more complete encoding done in http_error_302(),
1642 # but it is kept for compatibility with other callers.
1643 newurl
= newurl
.replace(' ', '%20')
1645 CONTENT_HEADERS
= ("content-length", "content-type")
1646 # NB: don't use dict comprehension for python 2.6 compatibility
1647 newheaders
= {k: v for k, v in req.headers.items() if k.lower() not in CONTENT_HEADERS}
1649 # A 303 must either use GET or HEAD for subsequent request
1650 # https://datatracker.ietf.org/doc/html/rfc7231#section-6.4.4
1651 if code
== 303 and m
!= 'HEAD':
1653 # 301 and 302 redirects are commonly turned into a GET from a POST
1654 # for subsequent requests by browsers, so we'll do the same.
1655 # https://datatracker.ietf.org/doc/html/rfc7231#section-6.4.2
1656 # https://datatracker.ietf.org/doc/html/rfc7231#section-6.4.3
1657 if code
in (301, 302) and m
== 'POST':
1660 return urllib
.request
.Request(
1661 newurl
, headers
=newheaders
, origin_req_host
=req
.origin_req_host
,
1662 unverifiable
=True, method
=m
)
1665 def extract_timezone(date_str
):
1668 ^.{8,}? # >=8 char non-TZ prefix, if present
1669 (?P<tz>Z| # just the UTC Z, or
1670 (?:(?<=.\b\d{4}|\b\d{2}:\d\d)| # preceded by 4 digits or hh:mm or
1671 (?<!.\b[a-zA-Z]{3}|[a-zA-Z]{4}|..\b\d\d)) # not preceded by 3 alpha word or >= 4 alpha or 2 digits
1672 [ ]? # optional space
1673 (?P<sign>\+|-) # +/-
1674 (?P<hours>[0-9]{2}):?(?P<minutes>[0-9]{2}) # hh[:]mm
1678 timezone
= datetime
.timedelta()
1680 date_str
= date_str
[:-len(m
.group('tz'))]
1681 if not m
.group('sign'):
1682 timezone
= datetime
.timedelta()
1684 sign
= 1 if m
.group('sign') == '+' else -1
1685 timezone
= datetime
.timedelta(
1686 hours
=sign
* int(m
.group('hours')),
1687 minutes
=sign
* int(m
.group('minutes')))
1688 return timezone
, date_str
1691 def parse_iso8601(date_str
, delimiter
='T', timezone
=None):
1692 """ Return a UNIX timestamp from the given date """
1694 if date_str
is None:
1697 date_str
= re
.sub(r
'\.[0-9]+', '', date_str
)
1699 if timezone
is None:
1700 timezone
, date_str
= extract_timezone(date_str
)
1702 with contextlib
.suppress(ValueError):
1703 date_format
= f
'%Y-%m-%d{delimiter}%H:%M:%S'
1704 dt
= datetime
.datetime
.strptime(date_str
, date_format
) - timezone
1705 return calendar
.timegm(dt
.timetuple())
1708 def date_formats(day_first
=True):
1709 return DATE_FORMATS_DAY_FIRST
if day_first
else DATE_FORMATS_MONTH_FIRST
1712 def unified_strdate(date_str
, day_first
=True):
1713 """Return a string with the date in the format YYYYMMDD"""
1715 if date_str
is None:
1719 date_str
= date_str
.replace(',', ' ')
1720 # Remove AM/PM + timezone
1721 date_str
= re
.sub(r
'(?i)\s*(?:AM|PM)(?:\s+[A-Z]+)?', '', date_str
)
1722 _
, date_str
= extract_timezone(date_str
)
1724 for expression
in date_formats(day_first
):
1725 with contextlib
.suppress(ValueError):
1726 upload_date
= datetime
.datetime
.strptime(date_str
, expression
).strftime('%Y%m%d')
1727 if upload_date
is None:
1728 timetuple
= email
.utils
.parsedate_tz(date_str
)
1730 with contextlib
.suppress(ValueError):
1731 upload_date
= datetime
.datetime(*timetuple
[:6]).strftime('%Y%m%d')
1732 if upload_date
is not None:
1733 return str(upload_date
)
1736 def unified_timestamp(date_str
, day_first
=True):
1737 if date_str
is None:
1740 date_str
= re
.sub(r
'[,|]', '', date_str
)
1742 pm_delta
= 12 if re
.search(r
'(?i)PM', date_str
) else 0
1743 timezone
, date_str
= extract_timezone(date_str
)
1745 # Remove AM/PM + timezone
1746 date_str
= re
.sub(r
'(?i)\s*(?:AM|PM)(?:\s+[A-Z]+)?', '', date_str
)
1748 # Remove unrecognized timezones from ISO 8601 alike timestamps
1749 m
= re
.search(r
'\d{1,2}:\d{1,2}(?:\.\d+)?(?P<tz>\s*[A-Z]+)$', date_str
)
1751 date_str
= date_str
[:-len(m
.group('tz'))]
1753 # Python only supports microseconds, so remove nanoseconds
1754 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]+$', date_str
)
1756 date_str
= m
.group(1)
1758 for expression
in date_formats(day_first
):
1759 with contextlib
.suppress(ValueError):
1760 dt
= datetime
.datetime
.strptime(date_str
, expression
) - timezone
+ datetime
.timedelta(hours
=pm_delta
)
1761 return calendar
.timegm(dt
.timetuple())
1762 timetuple
= email
.utils
.parsedate_tz(date_str
)
1764 return calendar
.timegm(timetuple
) + pm_delta
* 3600
1767 def determine_ext(url
, default_ext
='unknown_video'):
1768 if url
is None or '.' not in url
:
1770 guess
= url
.partition('?')[0].rpartition('.')[2]
1771 if re
.match(r
'^[A-Za-z0-9]+$', guess
):
1773 # Try extract ext from URLs like http://example.com/foo/bar.mp4/?download
1774 elif guess
.rstrip('/') in KNOWN_EXTENSIONS
:
1775 return guess
.rstrip('/')
1780 def subtitles_filename(filename
, sub_lang
, sub_format
, expected_real_ext
=None):
1781 return replace_extension(filename
, sub_lang
+ '.' + sub_format
, expected_real_ext
)
1784 def datetime_from_str(date_str
, precision
='auto', format
='%Y%m%d'):
1786 Return a datetime object from a string.
1788 (now|today|yesterday|DATE)([+-]\d+(microsecond|second|minute|hour|day|week|month|year)s?)?
1790 @param format strftime format of DATE
1791 @param precision Round the datetime object: auto|microsecond|second|minute|hour|day
1792 auto: round to the unit provided in date_str (if applicable).
1794 auto_precision
= False
1795 if precision
== 'auto':
1796 auto_precision
= True
1797 precision
= 'microsecond'
1798 today
= datetime_round(datetime
.datetime
.utcnow(), precision
)
1799 if date_str
in ('now', 'today'):
1801 if date_str
== 'yesterday':
1802 return today
- datetime
.timedelta(days
=1)
1804 r
'(?P<start>.+)(?P<sign>[+-])(?P<time>\d+)(?P<unit>microsecond|second|minute|hour|day|week|month|year)s?',
1806 if match
is not None:
1807 start_time
= datetime_from_str(match
.group('start'), precision
, format
)
1808 time
= int(match
.group('time')) * (-1 if match
.group('sign') == '-' else 1)
1809 unit
= match
.group('unit')
1810 if unit
== 'month' or unit
== 'year':
1811 new_date
= datetime_add_months(start_time
, time
* 12 if unit
== 'year' else time
)
1817 delta
= datetime
.timedelta(**{unit + 's': time}
)
1818 new_date
= start_time
+ delta
1820 return datetime_round(new_date
, unit
)
1823 return datetime_round(datetime
.datetime
.strptime(date_str
, format
), precision
)
1826 def date_from_str(date_str
, format
='%Y%m%d', strict
=False):
1828 Return a date object from a string using datetime_from_str
1830 @param strict Restrict allowed patterns to "YYYYMMDD" and
1831 (now|today|yesterday)(-\d+(day|week|month|year)s?)?
1833 if strict
and not re
.fullmatch(r
'\d{8}|(now|today|yesterday)(-\d+(day|week|month|year)s?)?', date_str
):
1834 raise ValueError(f
'Invalid date format "{date_str}"')
1835 return datetime_from_str(date_str
, precision
='microsecond', format
=format
).date()
1838 def datetime_add_months(dt
, months
):
1839 """Increment/Decrement a datetime object by months."""
1840 month
= dt
.month
+ months
- 1
1841 year
= dt
.year
+ month
// 12
1842 month
= month
% 12 + 1
1843 day
= min(dt
.day
, calendar
.monthrange(year
, month
)[1])
1844 return dt
.replace(year
, month
, day
)
1847 def datetime_round(dt
, precision
='day'):
1849 Round a datetime object's time to a specific precision
1851 if precision
== 'microsecond':
1860 roundto
= lambda x
, n
: ((x
+ n
/ 2) // n
) * n
1861 timestamp
= calendar
.timegm(dt
.timetuple())
1862 return datetime
.datetime
.utcfromtimestamp(roundto(timestamp
, unit_seconds
[precision
]))
1865 def hyphenate_date(date_str
):
1867 Convert a date in 'YYYYMMDD' format to 'YYYY-MM-DD' format"""
1868 match
= re
.match(r
'^(\d\d\d\d)(\d\d)(\d\d)$', date_str
)
1869 if match
is not None:
1870 return '-'.join(match
.groups())
1876 """Represents a time interval between two dates"""
1878 def __init__(self
, start
=None, end
=None):
1879 """start and end must be strings in the format accepted by date"""
1880 if start
is not None:
1881 self
.start
= date_from_str(start
, strict
=True)
1883 self
.start
= datetime
.datetime
.min.date()
1885 self
.end
= date_from_str(end
, strict
=True)
1887 self
.end
= datetime
.datetime
.max.date()
1888 if self
.start
> self
.end
:
1889 raise ValueError('Date range: "%s" , the start date must be before the end date' % self
)
1893 """Returns a range that only contains the given day"""
1894 return cls(day
, day
)
1896 def __contains__(self
, date
):
1897 """Check if the date is in the range"""
1898 if not isinstance(date
, datetime
.date
):
1899 date
= date_from_str(date
)
1900 return self
.start
<= date
<= self
.end
1903 return f
'{self.start.isoformat()} - {self.end.isoformat()}'
1905 def __eq__(self
, other
):
1906 return (isinstance(other
, DateRange
)
1907 and self
.start
== other
.start
and self
.end
== other
.end
)
1910 def platform_name():
1911 """ Returns the platform name as a str """
1912 write_string('DeprecationWarning: yt_dlp.utils.platform_name is deprecated, use platform.platform instead')
1913 return platform
.platform()
1917 def system_identifier():
1918 python_implementation
= platform
.python_implementation()
1919 if python_implementation
== 'PyPy' and hasattr(sys
, 'pypy_version_info'):
1920 python_implementation
+= ' version %d.%d.%d' % sys
.pypy_version_info
[:3]
1922 return 'Python %s (%s %s) - %s %s' % (
1923 platform
.python_version(),
1924 python_implementation
,
1925 platform
.architecture()[0],
1926 platform
.platform(),
1927 format_field(join_nonempty(*platform
.libc_ver(), delim
=' '), None, '(%s)'),
1932 def get_windows_version():
1933 ''' Get Windows version. returns () if it's not running on Windows '''
1934 if compat_os_name
== 'nt':
1935 return version_tuple(platform
.win32_ver()[1])
1940 def write_string(s
, out
=None, encoding
=None):
1941 assert isinstance(s
, str)
1942 out
= out
or sys
.stderr
1944 if compat_os_name
== 'nt' and supports_terminal_sequences(out
):
1945 s
= re
.sub(r
'([\r\n]+)', r
' \1', s
)
1947 enc
, buffer = None, out
1948 if 'b' in getattr(out
, 'mode', ''):
1949 enc
= encoding
or preferredencoding()
1950 elif hasattr(out
, 'buffer'):
1952 enc
= encoding
or getattr(out
, 'encoding', None) or preferredencoding()
1954 buffer.write(s
.encode(enc
, 'ignore') if enc
else s
)
1958 def bytes_to_intlist(bs
):
1961 if isinstance(bs
[0], int): # Python 3
1964 return [ord(c
) for c
in bs
]
1967 def intlist_to_bytes(xs
):
1970 return struct
.pack('%dB' % len(xs
), *xs
)
1973 class LockingUnsupportedError(OSError):
1974 msg
= 'File locking is not supported'
1977 super().__init
__(self
.msg
)
1980 # Cross-platform file locking
1981 if sys
.platform
== 'win32':
1982 import ctypes
.wintypes
1985 class OVERLAPPED(ctypes
.Structure
):
1987 ('Internal', ctypes
.wintypes
.LPVOID
),
1988 ('InternalHigh', ctypes
.wintypes
.LPVOID
),
1989 ('Offset', ctypes
.wintypes
.DWORD
),
1990 ('OffsetHigh', ctypes
.wintypes
.DWORD
),
1991 ('hEvent', ctypes
.wintypes
.HANDLE
),
1994 kernel32
= ctypes
.windll
.kernel32
1995 LockFileEx
= kernel32
.LockFileEx
1996 LockFileEx
.argtypes
= [
1997 ctypes
.wintypes
.HANDLE
, # hFile
1998 ctypes
.wintypes
.DWORD
, # dwFlags
1999 ctypes
.wintypes
.DWORD
, # dwReserved
2000 ctypes
.wintypes
.DWORD
, # nNumberOfBytesToLockLow
2001 ctypes
.wintypes
.DWORD
, # nNumberOfBytesToLockHigh
2002 ctypes
.POINTER(OVERLAPPED
) # Overlapped
2004 LockFileEx
.restype
= ctypes
.wintypes
.BOOL
2005 UnlockFileEx
= kernel32
.UnlockFileEx
2006 UnlockFileEx
.argtypes
= [
2007 ctypes
.wintypes
.HANDLE
, # hFile
2008 ctypes
.wintypes
.DWORD
, # dwReserved
2009 ctypes
.wintypes
.DWORD
, # nNumberOfBytesToLockLow
2010 ctypes
.wintypes
.DWORD
, # nNumberOfBytesToLockHigh
2011 ctypes
.POINTER(OVERLAPPED
) # Overlapped
2013 UnlockFileEx
.restype
= ctypes
.wintypes
.BOOL
2014 whole_low
= 0xffffffff
2015 whole_high
= 0x7fffffff
2017 def _lock_file(f
, exclusive
, block
):
2018 overlapped
= OVERLAPPED()
2019 overlapped
.Offset
= 0
2020 overlapped
.OffsetHigh
= 0
2021 overlapped
.hEvent
= 0
2022 f
._lock
_file
_overlapped
_p
= ctypes
.pointer(overlapped
)
2024 if not LockFileEx(msvcrt
.get_osfhandle(f
.fileno()),
2025 (0x2 if exclusive
else 0x0) |
(0x0 if block
else 0x1),
2026 0, whole_low
, whole_high
, f
._lock
_file
_overlapped
_p
):
2027 # NB: No argument form of "ctypes.FormatError" does not work on PyPy
2028 raise BlockingIOError(f
'Locking file failed: {ctypes.FormatError(ctypes.GetLastError())!r}')
2030 def _unlock_file(f
):
2031 assert f
._lock
_file
_overlapped
_p
2032 handle
= msvcrt
.get_osfhandle(f
.fileno())
2033 if not UnlockFileEx(handle
, 0, whole_low
, whole_high
, f
._lock
_file
_overlapped
_p
):
2034 raise OSError('Unlocking file failed: %r' % ctypes
.FormatError())
2040 def _lock_file(f
, exclusive
, block
):
2041 flags
= fcntl
.LOCK_EX
if exclusive
else fcntl
.LOCK_SH
2043 flags |
= fcntl
.LOCK_NB
2045 fcntl
.flock(f
, flags
)
2046 except BlockingIOError
:
2048 except OSError: # AOSP does not have flock()
2049 fcntl
.lockf(f
, flags
)
2051 def _unlock_file(f
):
2053 fcntl
.flock(f
, fcntl
.LOCK_UN
)
2055 fcntl
.lockf(f
, fcntl
.LOCK_UN
)
2059 def _lock_file(f
, exclusive
, block
):
2060 raise LockingUnsupportedError()
2062 def _unlock_file(f
):
2063 raise LockingUnsupportedError()
2069 def __init__(self
, filename
, mode
, block
=True, encoding
=None):
2070 if mode
not in {'r', 'rb', 'a', 'ab', 'w', 'wb'}
:
2071 raise NotImplementedError(mode
)
2072 self
.mode
, self
.block
= mode
, block
2074 writable
= any(f
in mode
for f
in 'wax+')
2075 readable
= any(f
in mode
for f
in 'r+')
2076 flags
= functools
.reduce(operator
.ior
, (
2077 getattr(os
, 'O_CLOEXEC', 0), # UNIX only
2078 getattr(os
, 'O_BINARY', 0), # Windows only
2079 getattr(os
, 'O_NOINHERIT', 0), # Windows only
2080 os
.O_CREAT
if writable
else 0, # O_TRUNC only after locking
2081 os
.O_APPEND
if 'a' in mode
else 0,
2082 os
.O_EXCL
if 'x' in mode
else 0,
2083 os
.O_RDONLY
if not writable
else os
.O_RDWR
if readable
else os
.O_WRONLY
,
2086 self
.f
= os
.fdopen(os
.open(filename
, flags
, 0o666), mode
, encoding
=encoding
)
2088 def __enter__(self
):
2089 exclusive
= 'r' not in self
.mode
2091 _lock_file(self
.f
, exclusive
, self
.block
)
2096 if 'w' in self
.mode
:
2099 except OSError as e
:
2101 errno
.ESPIPE
, # Illegal seek - expected for FIFO
2102 errno
.EINVAL
, # Invalid argument - expected for /dev/null
2111 _unlock_file(self
.f
)
2115 def __exit__(self
, *_
):
2124 def __getattr__(self
, attr
):
2125 return getattr(self
.f
, attr
)
2132 def get_filesystem_encoding():
2133 encoding
= sys
.getfilesystemencoding()
2134 return encoding
if encoding
is not None else 'utf-8'
2137 def shell_quote(args
):
2139 encoding
= get_filesystem_encoding()
2141 if isinstance(a
, bytes):
2142 # We may get a filename encoded with 'encodeFilename'
2143 a
= a
.decode(encoding
)
2144 quoted_args
.append(compat_shlex_quote(a
))
2145 return ' '.join(quoted_args
)
2148 def smuggle_url(url
, data
):
2149 """ Pass additional data in a URL for internal use. """
2151 url
, idata
= unsmuggle_url(url
, {})
2153 sdata
= urllib
.parse
.urlencode(
2154 {'__youtubedl_smuggle': json.dumps(data)}
)
2155 return url
+ '#' + sdata
2158 def unsmuggle_url(smug_url
, default
=None):
2159 if '#__youtubedl_smuggle' not in smug_url
:
2160 return smug_url
, default
2161 url
, _
, sdata
= smug_url
.rpartition('#')
2162 jsond
= urllib
.parse
.parse_qs(sdata
)['__youtubedl_smuggle'][0]
2163 data
= json
.loads(jsond
)
2167 def format_decimal_suffix(num
, fmt
='%d%s', *, factor
=1000):
2168 """ Formats numbers with decimal sufixes like K, M, etc """
2169 num
, factor
= float_or_none(num
), float(factor
)
2170 if num
is None or num
< 0:
2172 POSSIBLE_SUFFIXES
= 'kMGTPEZY'
2173 exponent
= 0 if num
== 0 else min(int(math
.log(num
, factor
)), len(POSSIBLE_SUFFIXES
))
2174 suffix
= ['', *POSSIBLE_SUFFIXES
][exponent
]
2176 suffix
= {'k': 'Ki', '': ''}
.get(suffix
, f
'{suffix}i')
2177 converted
= num
/ (factor
** exponent
)
2178 return fmt
% (converted
, suffix
)
2181 def format_bytes(bytes):
2182 return format_decimal_suffix(bytes, '%.2f%sB', factor
=1024) or 'N/A'
2185 def lookup_unit_table(unit_table
, s
):
2186 units_re
= '|'.join(re
.escape(u
) for u
in unit_table
)
2188 r
'(?P<num>[0-9]+(?:[,.][0-9]*)?)\s*(?P<unit>%s)\b' % units_re
, s
)
2191 num_str
= m
.group('num').replace(',', '.')
2192 mult
= unit_table
[m
.group('unit')]
2193 return int(float(num_str
) * mult
)
2196 def parse_filesize(s
):
2200 # The lower-case forms are of course incorrect and unofficial,
2201 # but we support those too
2218 'megabytes': 1000 ** 2,
2219 'mebibytes': 1024 ** 2,
2225 'gigabytes': 1000 ** 3,
2226 'gibibytes': 1024 ** 3,
2232 'terabytes': 1000 ** 4,
2233 'tebibytes': 1024 ** 4,
2239 'petabytes': 1000 ** 5,
2240 'pebibytes': 1024 ** 5,
2246 'exabytes': 1000 ** 6,
2247 'exbibytes': 1024 ** 6,
2253 'zettabytes': 1000 ** 7,
2254 'zebibytes': 1024 ** 7,
2260 'yottabytes': 1000 ** 8,
2261 'yobibytes': 1024 ** 8,
2264 return lookup_unit_table(_UNIT_TABLE
, s
)
2271 s
= re
.sub(r
'^[^\d]+\s', '', s
).strip()
2273 if re
.match(r
'^[\d,.]+$', s
):
2274 return str_to_int(s
)
2287 ret
= lookup_unit_table(_UNIT_TABLE
, s
)
2291 mobj
= re
.match(r
'([\d,.]+)(?:$|\s)', s
)
2293 return str_to_int(mobj
.group(1))
2296 def parse_resolution(s
, *, lenient
=False):
2301 mobj
= re
.search(r
'(?P<w>\d+)\s*[xX×,]\s*(?P<h>\d+)', s
)
2303 mobj
= re
.search(r
'(?<![a-zA-Z0-9])(?P<w>\d+)\s*[xX×,]\s*(?P<h>\d+)(?![a-zA-Z0-9])', s
)
2306 'width': int(mobj
.group('w')),
2307 'height': int(mobj
.group('h')),
2310 mobj
= re
.search(r
'(?<![a-zA-Z0-9])(\d+)[pPiI](?![a-zA-Z0-9])', s
)
2312 return {'height': int(mobj.group(1))}
2314 mobj
= re
.search(r
'\b([48])[kK]\b', s
)
2316 return {'height': int(mobj.group(1)) * 540}
2321 def parse_bitrate(s
):
2322 if not isinstance(s
, str):
2324 mobj
= re
.search(r
'\b(\d+)\s*kbps', s
)
2326 return int(mobj
.group(1))
2329 def month_by_name(name
, lang
='en'):
2330 """ Return the number of a month by (locale-independently) English name """
2332 month_names
= MONTH_NAMES
.get(lang
, MONTH_NAMES
['en'])
2335 return month_names
.index(name
) + 1
2340 def month_by_abbreviation(abbrev
):
2341 """ Return the number of a month by (locale-independently) English
2345 return [s
[:3] for s
in ENGLISH_MONTH_NAMES
].index(abbrev
) + 1
2350 def fix_xml_ampersands(xml_str
):
2351 """Replace all the '&' by '&' in XML"""
2353 r
'&(?!amp;|lt;|gt;|apos;|quot;|#x[0-9a-fA-F]{,4};|#[0-9]{,4};)',
2358 def setproctitle(title
):
2359 assert isinstance(title
, str)
2361 # ctypes in Jython is not complete
2362 # http://bugs.jython.org/issue2148
2363 if sys
.platform
.startswith('java'):
2367 libc
= ctypes
.cdll
.LoadLibrary('libc.so.6')
2371 # LoadLibrary in Windows Python 2.7.13 only expects
2372 # a bytestring, but since unicode_literals turns
2373 # every string into a unicode string, it fails.
2375 title_bytes
= title
.encode()
2376 buf
= ctypes
.create_string_buffer(len(title_bytes
))
2377 buf
.value
= title_bytes
2379 libc
.prctl(15, buf
, 0, 0, 0)
2380 except AttributeError:
2381 return # Strange libc, just skip this
2384 def remove_start(s
, start
):
2385 return s
[len(start
):] if s
is not None and s
.startswith(start
) else s
2388 def remove_end(s
, end
):
2389 return s
[:-len(end
)] if s
is not None and s
.endswith(end
) else s
2392 def remove_quotes(s
):
2393 if s
is None or len(s
) < 2:
2395 for quote
in ('"', "'", ):
2396 if s
[0] == quote
and s
[-1] == quote
:
2401 def get_domain(url
):
2403 This implementation is inconsistent, but is kept for compatibility.
2404 Use this only for "webpage_url_domain"
2406 return remove_start(urllib
.parse
.urlparse(url
).netloc
, 'www.') or None
2409 def url_basename(url
):
2410 path
= urllib
.parse
.urlparse(url
).path
2411 return path
.strip('/').split('/')[-1]
2415 return re
.match(r
'https?://[^?#&]+/', url
).group()
2418 def urljoin(base
, path
):
2419 if isinstance(path
, bytes):
2420 path
= path
.decode()
2421 if not isinstance(path
, str) or not path
:
2423 if re
.match(r
'^(?:[a-zA-Z][a-zA-Z0-9+-.]*:)?//', path
):
2425 if isinstance(base
, bytes):
2426 base
= base
.decode()
2427 if not isinstance(base
, str) or not re
.match(
2428 r
'^(?:https?:)?//', base
):
2430 return urllib
.parse
.urljoin(base
, path
)
2433 class HEADRequest(urllib
.request
.Request
):
2434 def get_method(self
):
2438 class PUTRequest(urllib
.request
.Request
):
2439 def get_method(self
):
2443 def int_or_none(v
, scale
=1, default
=None, get_attr
=None, invscale
=1):
2444 if get_attr
and v
is not None:
2445 v
= getattr(v
, get_attr
, None)
2447 return int(v
) * invscale
// scale
2448 except (ValueError, TypeError, OverflowError):
2452 def str_or_none(v
, default
=None):
2453 return default
if v
is None else str(v
)
2456 def str_to_int(int_str
):
2457 """ A more relaxed version of int_or_none """
2458 if isinstance(int_str
, int):
2460 elif isinstance(int_str
, str):
2461 int_str
= re
.sub(r
'[,\.\+]', '', int_str
)
2462 return int_or_none(int_str
)
2465 def float_or_none(v
, scale
=1, invscale
=1, default
=None):
2469 return float(v
) * invscale
/ scale
2470 except (ValueError, TypeError):
2474 def bool_or_none(v
, default
=None):
2475 return v
if isinstance(v
, bool) else default
2478 def strip_or_none(v
, default
=None):
2479 return v
.strip() if isinstance(v
, str) else default
2482 def url_or_none(url
):
2483 if not url
or not isinstance(url
, str):
2486 return url
if re
.match(r
'^(?:(?:https?|rt(?:m(?:pt?[es]?|fp)|sp[su]?)|mms|ftps?):)?//', url
) else None
2489 def request_to_url(req
):
2490 if isinstance(req
, urllib
.request
.Request
):
2491 return req
.get_full_url()
2496 def strftime_or_none(timestamp
, date_format
, default
=None):
2497 datetime_object
= None
2499 if isinstance(timestamp
, (int, float)): # unix timestamp
2500 datetime_object
= datetime
.datetime
.utcfromtimestamp(timestamp
)
2501 elif isinstance(timestamp
, str): # assume YYYYMMDD
2502 datetime_object
= datetime
.datetime
.strptime(timestamp
, '%Y%m%d')
2503 return datetime_object
.strftime(date_format
)
2504 except (ValueError, TypeError, AttributeError):
2508 def parse_duration(s
):
2509 if not isinstance(s
, str):
2515 days
, hours
, mins
, secs
, ms
= [None] * 5
2516 m
= re
.match(r
'''(?x)
2518 (?:(?:(?P<days>[0-9]+):)?(?P<hours>[0-9]+):)?(?P<mins>[0-9]+):)?
2519 (?P<secs>(?(before_secs)[0-9]{1,2}|[0-9]+))
2520 (?P<ms>[.:][0-9]+)?Z?$
2523 days
, hours
, mins
, secs
, ms
= m
.group('days', 'hours', 'mins', 'secs', 'ms')
2528 [0-9]+\s*y(?:ears?)?,?\s*
2531 [0-9]+\s*m(?:onths?)?,?\s*
2534 [0-9]+\s*w(?:eeks?)?,?\s*
2537 (?P<days>[0-9]+)\s*d(?:ays?)?,?\s*
2541 (?P<hours>[0-9]+)\s*h(?:ours?)?,?\s*
2544 (?P<mins>[0-9]+)\s*m(?:in(?:ute)?s?)?,?\s*
2547 (?P<secs>[0-9]+)(?P<ms>\.[0-9]+)?\s*s(?:ec(?:ond)?s?)?\s*
2550 days
, hours
, mins
, secs
, ms
= m
.groups()
2552 m
= re
.match(r
'(?i)(?:(?P<hours>[0-9.]+)\s*(?:hours?)|(?P<mins>[0-9.]+)\s*(?:mins?\.?|minutes?)\s*)Z?$', s
)
2554 hours
, mins
= m
.groups()
2559 ms
= ms
.replace(':', '.')
2560 return sum(float(part
or 0) * mult
for part
, mult
in (
2561 (days
, 86400), (hours
, 3600), (mins
, 60), (secs
, 1), (ms
, 1)))
2564 def prepend_extension(filename
, ext
, expected_real_ext
=None):
2565 name
, real_ext
= os
.path
.splitext(filename
)
2567 f
'{name}.{ext}{real_ext}'
2568 if not expected_real_ext
or real_ext
[1:] == expected_real_ext
2569 else f
'{filename}.{ext}')
2572 def replace_extension(filename
, ext
, expected_real_ext
=None):
2573 name
, real_ext
= os
.path
.splitext(filename
)
2574 return '{}.{}'.format(
2575 name
if not expected_real_ext
or real_ext
[1:] == expected_real_ext
else filename
,
2579 def check_executable(exe
, args
=[]):
2580 """ Checks if the given binary is installed somewhere in PATH, and returns its name.
2581 args can be a list of arguments for a short output (like -version) """
2583 Popen
.run([exe
] + args
, stdout
=subprocess
.PIPE
, stderr
=subprocess
.PIPE
)
2589 def _get_exe_version_output(exe
, args
, *, to_screen
=None):
2591 to_screen(f
'Checking exe version: {shell_quote([exe] + args)}')
2593 # STDIN should be redirected too. On UNIX-like systems, ffmpeg triggers
2594 # SIGTTOU if yt-dlp is run in the background.
2595 # See https://github.com/ytdl-org/youtube-dl/issues/955#issuecomment-209789656
2596 stdout
, _
, _
= Popen
.run([encodeArgument(exe
)] + args
, text
=True,
2597 stdin
=subprocess
.PIPE
, stdout
=subprocess
.PIPE
, stderr
=subprocess
.STDOUT
)
2603 def detect_exe_version(output
, version_re
=None, unrecognized
='present'):
2604 assert isinstance(output
, str)
2605 if version_re
is None:
2606 version_re
= r
'version\s+([-0-9._a-zA-Z]+)'
2607 m
= re
.search(version_re
, output
)
2614 def get_exe_version(exe
, args
=['--version'],
2615 version_re
=None, unrecognized
='present'):
2616 """ Returns the version of the specified executable,
2617 or False if the executable is not present """
2618 out
= _get_exe_version_output(exe
, args
)
2619 return detect_exe_version(out
, version_re
, unrecognized
) if out
else False
2622 def frange(start
=0, stop
=None, step
=1):
2625 start
, stop
= 0, start
2626 sign
= [-1, 1][step
> 0] if step
else 0
2627 while sign
* start
< sign
* stop
:
2632 class LazyList(collections
.abc
.Sequence
):
2633 """Lazy immutable list from an iterable
2634 Note that slices of a LazyList are lists and not LazyList"""
2636 class IndexError(IndexError):
2639 def __init__(self
, iterable
, *, reverse
=False, _cache
=None):
2640 self
._iterable
= iter(iterable
)
2641 self
._cache
= [] if _cache
is None else _cache
2642 self
._reversed
= reverse
2646 # We need to consume the entire iterable to iterate in reverse
2647 yield from self
.exhaust()
2649 yield from self
._cache
2650 for item
in self
._iterable
:
2651 self
._cache
.append(item
)
2655 self
._cache
.extend(self
._iterable
)
2656 self
._iterable
= [] # Discard the emptied iterable to make it pickle-able
2660 """Evaluate the entire iterable"""
2661 return self
._exhaust
()[::-1 if self
._reversed
else 1]
2664 def _reverse_index(x
):
2665 return None if x
is None else ~x
2667 def __getitem__(self
, idx
):
2668 if isinstance(idx
, slice):
2670 idx
= slice(self
._reverse
_index
(idx
.start
), self
._reverse
_index
(idx
.stop
), -(idx
.step
or 1))
2671 start
, stop
, step
= idx
.start
, idx
.stop
, idx
.step
or 1
2672 elif isinstance(idx
, int):
2674 idx
= self
._reverse
_index
(idx
)
2675 start
, stop
, step
= idx
, idx
, 0
2677 raise TypeError('indices must be integers or slices')
2678 if ((start
or 0) < 0 or (stop
or 0) < 0
2679 or (start
is None and step
< 0)
2680 or (stop
is None and step
> 0)):
2681 # We need to consume the entire iterable to be able to slice from the end
2682 # Obviously, never use this with infinite iterables
2685 return self
._cache
[idx
]
2686 except IndexError as e
:
2687 raise self
.IndexError(e
) from e
2688 n
= max(start
or 0, stop
or 0) - len(self
._cache
) + 1
2690 self
._cache
.extend(itertools
.islice(self
._iterable
, n
))
2692 return self
._cache
[idx
]
2693 except IndexError as e
:
2694 raise self
.IndexError(e
) from e
2698 self
[-1] if self
._reversed
else self
[0]
2699 except self
.IndexError:
2705 return len(self
._cache
)
2707 def __reversed__(self
):
2708 return type(self
)(self
._iterable
, reverse
=not self
._reversed
, _cache
=self
._cache
)
2711 return type(self
)(self
._iterable
, reverse
=self
._reversed
, _cache
=self
._cache
)
2714 # repr and str should mimic a list. So we exhaust the iterable
2715 return repr(self
.exhaust())
2718 return repr(self
.exhaust())
2723 class IndexError(IndexError):
2727 # This is only useful for tests
2728 return len(self
.getslice())
2730 def __init__(self
, pagefunc
, pagesize
, use_cache
=True):
2731 self
._pagefunc
= pagefunc
2732 self
._pagesize
= pagesize
2733 self
._pagecount
= float('inf')
2734 self
._use
_cache
= use_cache
2737 def getpage(self
, pagenum
):
2738 page_results
= self
._cache
.get(pagenum
)
2739 if page_results
is None:
2740 page_results
= [] if pagenum
> self
._pagecount
else list(self
._pagefunc
(pagenum
))
2742 self
._cache
[pagenum
] = page_results
2745 def getslice(self
, start
=0, end
=None):
2746 return list(self
._getslice
(start
, end
))
2748 def _getslice(self
, start
, end
):
2749 raise NotImplementedError('This method must be implemented by subclasses')
2751 def __getitem__(self
, idx
):
2752 assert self
._use
_cache
, 'Indexing PagedList requires cache'
2753 if not isinstance(idx
, int) or idx
< 0:
2754 raise TypeError('indices must be non-negative integers')
2755 entries
= self
.getslice(idx
, idx
+ 1)
2757 raise self
.IndexError()
2761 class OnDemandPagedList(PagedList
):
2762 """Download pages until a page with less than maximum results"""
2764 def _getslice(self
, start
, end
):
2765 for pagenum
in itertools
.count(start
// self
._pagesize
):
2766 firstid
= pagenum
* self
._pagesize
2767 nextfirstid
= pagenum
* self
._pagesize
+ self
._pagesize
2768 if start
>= nextfirstid
:
2772 start
% self
._pagesize
2773 if firstid
<= start
< nextfirstid
2776 ((end
- 1) % self
._pagesize
) + 1
2777 if (end
is not None and firstid
<= end
<= nextfirstid
)
2781 page_results
= self
.getpage(pagenum
)
2783 self
._pagecount
= pagenum
- 1
2785 if startv
!= 0 or endv
is not None:
2786 page_results
= page_results
[startv
:endv
]
2787 yield from page_results
2789 # A little optimization - if current page is not "full", ie. does
2790 # not contain page_size videos then we can assume that this page
2791 # is the last one - there are no more ids on further pages -
2792 # i.e. no need to query again.
2793 if len(page_results
) + startv
< self
._pagesize
:
2796 # If we got the whole page, but the next page is not interesting,
2797 # break out early as well
2798 if end
== nextfirstid
:
2802 class InAdvancePagedList(PagedList
):
2803 """PagedList with total number of pages known in advance"""
2805 def __init__(self
, pagefunc
, pagecount
, pagesize
):
2806 PagedList
.__init
__(self
, pagefunc
, pagesize
, True)
2807 self
._pagecount
= pagecount
2809 def _getslice(self
, start
, end
):
2810 start_page
= start
// self
._pagesize
2811 end_page
= self
._pagecount
if end
is None else min(self
._pagecount
, end
// self
._pagesize
+ 1)
2812 skip_elems
= start
- start_page
* self
._pagesize
2813 only_more
= None if end
is None else end
- start
2814 for pagenum
in range(start_page
, end_page
):
2815 page_results
= self
.getpage(pagenum
)
2817 page_results
= page_results
[skip_elems
:]
2819 if only_more
is not None:
2820 if len(page_results
) < only_more
:
2821 only_more
-= len(page_results
)
2823 yield from page_results
[:only_more
]
2825 yield from page_results
2828 class PlaylistEntries
:
2829 MissingEntry
= object()
2830 is_exhausted
= False
2832 def __init__(self
, ydl
, info_dict
):
2835 # _entries must be assigned now since infodict can change during iteration
2836 entries
= info_dict
.get('entries')
2838 raise EntryNotInPlaylist('There are no entries')
2839 elif isinstance(entries
, list):
2840 self
.is_exhausted
= True
2842 requested_entries
= info_dict
.get('requested_entries')
2843 self
.is_incomplete
= bool(requested_entries
)
2844 if self
.is_incomplete
:
2845 assert self
.is_exhausted
2846 self
._entries
= [self
.MissingEntry
] * max(requested_entries
)
2847 for i
, entry
in zip(requested_entries
, entries
):
2848 self
._entries
[i
- 1] = entry
2849 elif isinstance(entries
, (list, PagedList
, LazyList
)):
2850 self
._entries
= entries
2852 self
._entries
= LazyList(entries
)
2854 PLAYLIST_ITEMS_RE
= re
.compile(r
'''(?x)
2855 (?P<start>[+-]?\d+)?
2857 (?P<end>[+-]?\d+|inf(?:inite)?)?
2858 (?::(?P<step>[+-]?\d+))?
2862 def parse_playlist_items(cls
, string
):
2863 for segment
in string
.split(','):
2865 raise ValueError('There is two or more consecutive commas')
2866 mobj
= cls
.PLAYLIST_ITEMS_RE
.fullmatch(segment
)
2868 raise ValueError(f
'{segment!r} is not a valid specification')
2869 start
, end
, step
, has_range
= mobj
.group('start', 'end', 'step', 'range')
2870 if int_or_none(step
) == 0:
2871 raise ValueError(f
'Step in {segment!r} cannot be zero')
2872 yield slice(int_or_none(start
), float_or_none(end
), int_or_none(step
)) if has_range
else int(start
)
2874 def get_requested_items(self
):
2875 playlist_items
= self
.ydl
.params
.get('playlist_items')
2876 playlist_start
= self
.ydl
.params
.get('playliststart', 1)
2877 playlist_end
= self
.ydl
.params
.get('playlistend')
2878 # For backwards compatibility, interpret -1 as whole list
2879 if playlist_end
in (-1, None):
2881 if not playlist_items
:
2882 playlist_items
= f
'{playlist_start}:{playlist_end}'
2883 elif playlist_start
!= 1 or playlist_end
:
2884 self
.ydl
.report_warning('Ignoring playliststart and playlistend because playlistitems was given', only_once
=True)
2886 for index
in self
.parse_playlist_items(playlist_items
):
2887 for i
, entry
in self
[index
]:
2892 # TODO: Add auto-generated fields
2893 self
.ydl
._match
_entry
(entry
, incomplete
=True, silent
=True)
2894 except (ExistingVideoReached
, RejectedVideoReached
):
2897 def get_full_count(self
):
2898 if self
.is_exhausted
and not self
.is_incomplete
:
2900 elif isinstance(self
._entries
, InAdvancePagedList
):
2901 if self
._entries
._pagesize
== 1:
2902 return self
._entries
._pagecount
2904 @functools.cached_property
2906 if isinstance(self
._entries
, list):
2909 entry
= self
._entries
[i
]
2911 entry
= self
.MissingEntry
2912 if not self
.is_incomplete
:
2913 raise self
.IndexError()
2914 if entry
is self
.MissingEntry
:
2915 raise EntryNotInPlaylist(f
'Entry {i} cannot be found')
2920 return type(self
.ydl
)._handle
_extraction
_exceptions
(lambda _
, i
: self
._entries
[i
])(self
.ydl
, i
)
2921 except (LazyList
.IndexError, PagedList
.IndexError):
2922 raise self
.IndexError()
2925 def __getitem__(self
, idx
):
2926 if isinstance(idx
, int):
2927 idx
= slice(idx
, idx
)
2929 # NB: PlaylistEntries[1:10] => (0, 1, ... 9)
2930 step
= 1 if idx
.step
is None else idx
.step
2931 if idx
.start
is None:
2932 start
= 0 if step
> 0 else len(self
) - 1
2934 start
= idx
.start
- 1 if idx
.start
>= 0 else len(self
) + idx
.start
2936 # NB: Do not call len(self) when idx == [:]
2937 if idx
.stop
is None:
2938 stop
= 0 if step
< 0 else float('inf')
2940 stop
= idx
.stop
- 1 if idx
.stop
>= 0 else len(self
) + idx
.stop
2941 stop
+= [-1, 1][step
> 0]
2943 for i
in frange(start
, stop
, step
):
2947 entry
= self
._getter
(i
)
2948 except self
.IndexError:
2949 self
.is_exhausted
= True
2956 return len(tuple(self
[:]))
2958 class IndexError(IndexError):
2962 def uppercase_escape(s
):
2963 unicode_escape
= codecs
.getdecoder('unicode_escape')
2965 r
'\\U[0-9a-fA-F]{8}',
2966 lambda m
: unicode_escape(m
.group(0))[0],
2970 def lowercase_escape(s
):
2971 unicode_escape
= codecs
.getdecoder('unicode_escape')
2973 r
'\\u[0-9a-fA-F]{4}',
2974 lambda m
: unicode_escape(m
.group(0))[0],
2978 def escape_rfc3986(s
):
2979 """Escape non-ASCII characters as suggested by RFC 3986"""
2980 return urllib
.parse
.quote(s
, b
"%/;:@&=+$,!~*'()?#[]")
2983 def escape_url(url
):
2984 """Escape URL as suggested by RFC 3986"""
2985 url_parsed
= urllib
.parse
.urlparse(url
)
2986 return url_parsed
._replace
(
2987 netloc
=url_parsed
.netloc
.encode('idna').decode('ascii'),
2988 path
=escape_rfc3986(url_parsed
.path
),
2989 params
=escape_rfc3986(url_parsed
.params
),
2990 query
=escape_rfc3986(url_parsed
.query
),
2991 fragment
=escape_rfc3986(url_parsed
.fragment
)
2996 return urllib
.parse
.parse_qs(urllib
.parse
.urlparse(url
).query
)
2999 def read_batch_urls(batch_fd
):
3001 if not isinstance(url
, str):
3002 url
= url
.decode('utf-8', 'replace')
3003 BOM_UTF8
= ('\xef\xbb\xbf', '\ufeff')
3004 for bom
in BOM_UTF8
:
3005 if url
.startswith(bom
):
3006 url
= url
[len(bom
):]
3008 if not url
or url
.startswith(('#', ';', ']')):
3010 # "#" cannot be stripped out since it is part of the URI
3011 # However, it can be safely stripped out if following a whitespace
3012 return re
.split(r
'\s#', url
, 1)[0].rstrip()
3014 with contextlib
.closing(batch_fd
) as fd
:
3015 return [url
for url
in map(fixup
, fd
) if url
]
3018 def urlencode_postdata(*args
, **kargs
):
3019 return urllib
.parse
.urlencode(*args
, **kargs
).encode('ascii')
3022 def update_url_query(url
, query
):
3025 parsed_url
= urllib
.parse
.urlparse(url
)
3026 qs
= urllib
.parse
.parse_qs(parsed_url
.query
)
3028 return urllib
.parse
.urlunparse(parsed_url
._replace
(
3029 query
=urllib
.parse
.urlencode(qs
, True)))
3032 def update_Request(req
, url
=None, data
=None, headers
=None, query
=None):
3033 req_headers
= req
.headers
.copy()
3034 req_headers
.update(headers
or {})
3035 req_data
= data
or req
.data
3036 req_url
= update_url_query(url
or req
.get_full_url(), query
)
3037 req_get_method
= req
.get_method()
3038 if req_get_method
== 'HEAD':
3039 req_type
= HEADRequest
3040 elif req_get_method
== 'PUT':
3041 req_type
= PUTRequest
3043 req_type
= urllib
.request
.Request
3045 req_url
, data
=req_data
, headers
=req_headers
,
3046 origin_req_host
=req
.origin_req_host
, unverifiable
=req
.unverifiable
)
3047 if hasattr(req
, 'timeout'):
3048 new_req
.timeout
= req
.timeout
3052 def _multipart_encode_impl(data
, boundary
):
3053 content_type
= 'multipart/form-data; boundary=%s' % boundary
3056 for k
, v
in data
.items():
3057 out
+= b
'--' + boundary
.encode('ascii') + b
'\r\n'
3058 if isinstance(k
, str):
3060 if isinstance(v
, str):
3062 # RFC 2047 requires non-ASCII field names to be encoded, while RFC 7578
3063 # suggests sending UTF-8 directly. Firefox sends UTF-8, too
3064 content
= b
'Content-Disposition: form-data; name="' + k
+ b
'"\r\n\r\n' + v
+ b
'\r\n'
3065 if boundary
.encode('ascii') in content
:
3066 raise ValueError('Boundary overlaps with data')
3069 out
+= b
'--' + boundary
.encode('ascii') + b
'--\r\n'
3071 return out
, content_type
3074 def multipart_encode(data
, boundary
=None):
3076 Encode a dict to RFC 7578-compliant form-data
3079 A dict where keys and values can be either Unicode or bytes-like
3082 If specified a Unicode object, it's used as the boundary. Otherwise
3083 a random boundary is generated.
3085 Reference: https://tools.ietf.org/html/rfc7578
3087 has_specified_boundary
= boundary
is not None
3090 if boundary
is None:
3091 boundary
= '---------------' + str(random
.randrange(0x0fffffff, 0xffffffff))
3094 out
, content_type
= _multipart_encode_impl(data
, boundary
)
3097 if has_specified_boundary
:
3101 return out
, content_type
3104 def dict_get(d
, key_or_keys
, default
=None, skip_false_values
=True):
3105 for val
in map(d
.get
, variadic(key_or_keys
)):
3106 if val
is not None and (val
or not skip_false_values
):
3111 def try_call(*funcs
, expected_type
=None, args
=[], kwargs
={}):
3114 val
= f(*args
, **kwargs
)
3115 except (AttributeError, KeyError, TypeError, IndexError, ZeroDivisionError):
3118 if expected_type
is None or isinstance(val
, expected_type
):
3122 def try_get(src
, getter
, expected_type
=None):
3123 return try_call(*variadic(getter
), args
=(src
,), expected_type
=expected_type
)
3126 def filter_dict(dct
, cndn
=lambda _
, v
: v
is not None):
3127 return {k: v for k, v in dct.items() if cndn(k, v)}
3130 def merge_dicts(*dicts
):
3132 for a_dict
in dicts
:
3133 for k
, v
in a_dict
.items():
3134 if (v
is not None and k
not in merged
3135 or isinstance(v
, str) and merged
[k
] == ''):
3140 def encode_compat_str(string
, encoding
=preferredencoding(), errors
='strict'):
3141 return string
if isinstance(string
, str) else str(string
, encoding
, errors
)
3153 TV_PARENTAL_GUIDELINES
= {
3163 def parse_age_limit(s
):
3164 # isinstance(False, int) is True. So type() must be used instead
3165 if type(s
) is int: # noqa: E721
3166 return s
if 0 <= s
<= 21 else None
3167 elif not isinstance(s
, str):
3169 m
= re
.match(r
'^(?P<age>\d{1,2})\+?$', s
)
3171 return int(m
.group('age'))
3174 return US_RATINGS
[s
]
3175 m
= re
.match(r
'^TV[_-]?(%s)$' % '|'.join(k
[3:] for k
in TV_PARENTAL_GUIDELINES
), s
)
3177 return TV_PARENTAL_GUIDELINES
['TV-' + m
.group(1)]
3181 def strip_jsonp(code
):
3184 (?:window\.)?(?P<func_name>[a-zA-Z0-9_.$]*)
3185 (?:\s*&&\s*(?P=func_name))?
3186 \s*\(\s*(?P<callback_data>.*)\);?
3187 \s*?(?://[^\n]*)*$''',
3188 r
'\g<callback_data>', code
)
3191 def js_to_json(code
, vars={}):
3192 # vars is a dict of var, val pairs to substitute
3193 COMMENT_RE
= r
'/\*(?:(?!\*/).)*?\*/|//[^\n]*\n'
3194 SKIP_RE
= fr
'\s*(?:{COMMENT_RE})?\s*'
3196 (fr
'(?s)^(0[xX][0-9a-fA-F]+){SKIP_RE}:?$', 16),
3197 (fr
'(?s)^(0+[0-7]+){SKIP_RE}:?$', 8),
3202 if v
in ('true', 'false', 'null'):
3204 elif v
in ('undefined', 'void 0'):
3206 elif v
.startswith('/*') or v
.startswith('//') or v
.startswith('!') or v
== ',':
3209 if v
[0] in ("'", '"'):
3210 v
= re
.sub(r
'(?s)\\.|"', lambda m
: {
3215 }.get(m
.group(0), m
.group(0)), v
[1:-1])
3217 for regex
, base
in INTEGER_TABLE
:
3218 im
= re
.match(regex
, v
)
3220 i
= int(im
.group(1), base
)
3221 return '"%d":' % i
if v
.endswith(':') else '%d' % i
3228 def create_map(mobj
):
3229 return json
.dumps(dict(json
.loads(js_to_json(mobj
.group(1) or '[]', vars=vars))))
3231 code
= re
.sub(r
'new Date\((".+")\)', r
'\g<1>', code
)
3232 code
= re
.sub(r
'new Map\((\[.*?\])?\)', create_map
, code
)
3234 return re
.sub(r
'''(?sx)
3235 "(?:[^"\\]*(?:\\\\|\\['"nurtbfx/\n]))*[^"\\]*"|
3236 '(?:[^'\\]*(?:\\\\|\\['"nurtbfx/\n]))*[^'\\]*'|
3237 {comment}|,(?={skip}[\]}}])|
3238 void\s0|(?:(?<![0-9])[eE]|[a-df-zA-DF-Z_$])[.a-zA-Z_$0-9]*|
3239 \b(?:0[xX][0-9a-fA-F]+|0+[0-7]+)(?:{skip}:)?|
3242 '''.format(comment
=COMMENT_RE
, skip
=SKIP_RE
), fix_kv
, code
)
3245 def qualities(quality_ids
):
3246 """ Get a numeric quality value out of a list of possible values """
3249 return quality_ids
.index(qid
)
3255 POSTPROCESS_WHEN
= ('pre_process', 'after_filter', 'before_dl', 'post_process', 'after_move', 'after_video', 'playlist')
3259 'default': '%(title)s [%(id)s].%(ext)s',
3260 'chapter': '%(title)s - %(section_number)03d %(section_title)s [%(id)s].%(ext)s',
3266 'description': 'description',
3267 'annotation': 'annotations.xml',
3268 'infojson': 'info.json',
3271 'pl_thumbnail': None,
3272 'pl_description': 'description',
3273 'pl_infojson': 'info.json',
3276 # As of [1] format syntax is:
3277 # %[mapping_key][conversion_flags][minimum_width][.precision][length_modifier]type
3278 # 1. https://docs.python.org/2/library/stdtypes.html#string-formatting
3279 STR_FORMAT_RE_TMPL
= r
'''(?x)
3280 (?<!%)(?P<prefix>(?:%%)*)
3282 (?P<has_key>\((?P<key>{0})\))?
3284 (?P<conversion>[#0\-+ ]+)?
3286 (?P<precision>\.\d+)?
3287 (?P<len_mod>[hlL])? # unused in python
3288 {1} # conversion type
3293 STR_FORMAT_TYPES
= 'diouxXeEfFgGcrs'
3296 def limit_length(s
, length
):
3297 """ Add ellipses to overly long strings """
3302 return s
[:length
- len(ELLIPSES
)] + ELLIPSES
3306 def version_tuple(v
):
3307 return tuple(int(e
) for e
in re
.split(r
'[-.]', v
))
3310 def is_outdated_version(version
, limit
, assume_new
=True):
3312 return not assume_new
3314 return version_tuple(version
) < version_tuple(limit
)
3316 return not assume_new
3319 def ytdl_is_updateable():
3320 """ Returns if yt-dlp can be updated with -U """
3322 from .update
import is_non_updateable
3324 return not is_non_updateable()
3327 def args_to_str(args
):
3328 # Get a short string representation for a subprocess command
3329 return ' '.join(compat_shlex_quote(a
) for a
in args
)
3332 def error_to_compat_str(err
):
3336 def error_to_str(err
):
3337 return f
'{type(err).__name__}: {err}'
3340 def mimetype2ext(mt
):
3344 mt
, _
, params
= mt
.partition(';')
3349 # Per RFC 3003, audio/mpeg can be .mp1, .mp2 or .mp3. Here use .mp3 as
3350 # it's the most popular one
3351 'audio/mpeg': 'mp3',
3352 'audio/x-wav': 'wav',
3354 'audio/wave': 'wav',
3357 ext
= FULL_MAP
.get(mt
)
3363 'smptett+xml': 'tt',
3367 'x-mp4-fragmented': 'mp4',
3368 'x-ms-sami': 'sami',
3371 'x-mpegurl': 'm3u8',
3372 'vnd.apple.mpegurl': 'm3u8',
3376 'vnd.ms-sstr+xml': 'ism',
3380 'filmstrip+json': 'fs',
3384 _
, _
, subtype
= mt
.rpartition('/')
3385 ext
= SUBTYPE_MAP
.get(subtype
.lower())
3396 _
, _
, suffix
= subtype
.partition('+')
3397 ext
= SUFFIX_MAP
.get(suffix
)
3401 return subtype
.replace('+', '.')
3404 def ext2mimetype(ext_or_url
):
3407 if '.' not in ext_or_url
:
3408 ext_or_url
= f
'file.{ext_or_url}'
3409 return mimetypes
.guess_type(ext_or_url
)[0]
3412 def parse_codecs(codecs_str
):
3413 # http://tools.ietf.org/html/rfc6381
3416 split_codecs
= list(filter(None, map(
3417 str.strip
, codecs_str
.strip().strip(',').split(','))))
3418 vcodec
, acodec
, scodec
, hdr
= None, None, None, None
3419 for full_codec
in split_codecs
:
3420 parts
= re
.sub(r
'0+(?=\d)', '', full_codec
).split('.')
3421 if parts
[0] in ('avc1', 'avc2', 'avc3', 'avc4', 'vp9', 'vp8', 'hev1', 'hev2',
3422 'h263', 'h264', 'mp4v', 'hvc1', 'av1', 'theora', 'dvh1', 'dvhe'):
3426 if parts
[0] in ('dvh1', 'dvhe'):
3428 elif parts
[0] == 'av1' and traverse_obj(parts
, 3) == '10':
3430 elif parts
[:2] == ['vp9', '2']:
3432 elif parts
[0] in ('flac', 'mp4a', 'opus', 'vorbis', 'mp3', 'aac',
3433 'ac-3', 'ec-3', 'eac3', 'dtsc', 'dtse', 'dtsh', 'dtsl'):
3434 acodec
= acodec
or full_codec
3435 elif parts
[0] in ('stpp', 'wvtt'):
3436 scodec
= scodec
or full_codec
3438 write_string(f
'WARNING: Unknown codec {full_codec}\n')
3439 if vcodec
or acodec
or scodec
:
3441 'vcodec': vcodec
or 'none',
3442 'acodec': acodec
or 'none',
3443 'dynamic_range': hdr
,
3444 **({'scodec': scodec}
if scodec
is not None else {}),
3446 elif len(split_codecs
) == 2:
3448 'vcodec': split_codecs
[0],
3449 'acodec': split_codecs
[1],
3454 def urlhandle_detect_ext(url_handle
):
3455 getheader
= url_handle
.headers
.get
3457 cd
= getheader('Content-Disposition')
3459 m
= re
.match(r
'attachment;\s*filename="(?P<filename>[^"]+)"', cd
)
3461 e
= determine_ext(m
.group('filename'), default_ext
=None)
3465 return mimetype2ext(getheader('Content-Type'))
3468 def encode_data_uri(data
, mime_type
):
3469 return 'data:%s;base64,%s' % (mime_type
, base64
.b64encode(data
).decode('ascii'))
3472 def age_restricted(content_limit
, age_limit
):
3473 """ Returns True iff the content should be blocked """
3475 if age_limit
is None: # No limit set
3477 if content_limit
is None:
3478 return False # Content available for everyone
3479 return age_limit
< content_limit
3482 # List of known byte-order-marks (BOM)
3484 (b
'\xef\xbb\xbf', 'utf-8'),
3485 (b
'\x00\x00\xfe\xff', 'utf-32-be'),
3486 (b
'\xff\xfe\x00\x00', 'utf-32-le'),
3487 (b
'\xff\xfe', 'utf-16-le'),
3488 (b
'\xfe\xff', 'utf-16-be'),
3492 def is_html(first_bytes
):
3493 """ Detect whether a file contains HTML by examining its first bytes. """
3496 for bom
, enc
in BOMS
:
3497 while first_bytes
.startswith(bom
):
3498 encoding
, first_bytes
= enc
, first_bytes
[len(bom
):]
3500 return re
.match(r
'^\s*<', first_bytes
.decode(encoding
, 'replace'))
3503 def determine_protocol(info_dict
):
3504 protocol
= info_dict
.get('protocol')
3505 if protocol
is not None:
3508 url
= sanitize_url(info_dict
['url'])
3509 if url
.startswith('rtmp'):
3511 elif url
.startswith('mms'):
3513 elif url
.startswith('rtsp'):
3516 ext
= determine_ext(url
)
3522 return urllib
.parse
.urlparse(url
).scheme
3525 def render_table(header_row
, data
, delim
=False, extra_gap
=0, hide_empty
=False):
3526 """ Render a list of rows, each as a list of values.
3527 Text after a \t will be right aligned """
3529 return len(remove_terminal_sequences(string
).replace('\t', ''))
3531 def get_max_lens(table
):
3532 return [max(width(str(v
)) for v
in col
) for col
in zip(*table
)]
3534 def filter_using_list(row
, filterArray
):
3535 return [col
for take
, col
in itertools
.zip_longest(filterArray
, row
, fillvalue
=True) if take
]
3537 max_lens
= get_max_lens(data
) if hide_empty
else []
3538 header_row
= filter_using_list(header_row
, max_lens
)
3539 data
= [filter_using_list(row
, max_lens
) for row
in data
]
3541 table
= [header_row
] + data
3542 max_lens
= get_max_lens(table
)
3545 table
= [header_row
, [delim
* (ml
+ extra_gap
) for ml
in max_lens
]] + data
3546 table
[1][-1] = table
[1][-1][:-extra_gap
* len(delim
)] # Remove extra_gap from end of delimiter
3548 for pos
, text
in enumerate(map(str, row
)):
3550 row
[pos
] = text
.replace('\t', ' ' * (max_lens
[pos
] - width(text
))) + ' ' * extra_gap
3552 row
[pos
] = text
+ ' ' * (max_lens
[pos
] - width(text
) + extra_gap
)
3553 ret
= '\n'.join(''.join(row
).rstrip() for row
in table
)
3557 def _match_one(filter_part
, dct
, incomplete
):
3558 # TODO: Generalize code with YoutubeDL._build_format_filter
3559 STRING_OPERATORS
= {
3560 '*=': operator
.contains
,
3561 '^=': lambda attr
, value
: attr
.startswith(value
),
3562 '$=': lambda attr
, value
: attr
.endswith(value
),
3563 '~=': lambda attr
, value
: re
.search(value
, attr
),
3565 COMPARISON_OPERATORS
= {
3567 '<=': operator
.le
, # "<=" must be defined above "<"
3574 if isinstance(incomplete
, bool):
3575 is_incomplete
= lambda _
: incomplete
3577 is_incomplete
= lambda k
: k
in incomplete
3579 operator_rex
= re
.compile(r
'''(?x)
3581 \s*(?P<negation>!\s*)?(?P<op>%s)(?P<none_inclusive>\s*\?)?\s*
3583 (?P<quote>["\'])(?P
<quotedstrval
>.+?
)(?P
=quote
)|
3586 ''' % '|'.join(map(re.escape, COMPARISON_OPERATORS.keys())))
3587 m = operator_rex.fullmatch(filter_part.strip())
3590 unnegated_op = COMPARISON_OPERATORS[m['op']]
3592 op = lambda attr, value: not unnegated_op(attr, value)
3595 comparison_value = m['quotedstrval'] or m['strval'] or m['intval']
3597 comparison_value = comparison_value.replace(r'\%s' % m['quote'], m['quote'])
3598 actual_value = dct.get(m['key'])
3599 numeric_comparison = None
3600 if isinstance(actual_value, (int, float)):
3601 # If the original field is a string and matching comparisonvalue is
3602 # a number we should respect the origin of the original field
3603 # and process comparison value as a string (see
3604 # https://github.com/ytdl-org/youtube-dl/issues/11082)
3606 numeric_comparison = int(comparison_value)
3608 numeric_comparison = parse_filesize(comparison_value)
3609 if numeric_comparison is None:
3610 numeric_comparison = parse_filesize(f'{comparison_value}B')
3611 if numeric_comparison is None:
3612 numeric_comparison = parse_duration(comparison_value)
3613 if numeric_comparison is not None and m['op'] in STRING_OPERATORS:
3614 raise ValueError('Operator %s only supports string values!' % m['op'])
3615 if actual_value is None:
3616 return is_incomplete(m['key']) or m['none_inclusive']
3617 return op(actual_value, comparison_value if numeric_comparison is None else numeric_comparison)
3620 '': lambda v: (v is True) if isinstance(v, bool) else (v is not None),
3621 '!': lambda v: (v is False) if isinstance(v, bool) else (v is None),
3623 operator_rex = re.compile(r'''(?x
)
3624 (?P
<op
>%s)\s
*(?P
<key
>[a
-z_
]+)
3625 ''' % '|'.join(map(re.escape, UNARY_OPERATORS.keys())))
3626 m = operator_rex.fullmatch(filter_part.strip())
3628 op = UNARY_OPERATORS[m.group('op')]
3629 actual_value = dct.get(m.group('key'))
3630 if is_incomplete(m.group('key')) and actual_value is None:
3632 return op(actual_value)
3634 raise ValueError('Invalid filter part %r' % filter_part)
3637 def match_str(filter_str, dct, incomplete=False):
3638 """ Filter a dictionary with a simple string syntax.
3639 @returns Whether the filter passes
3640 @param incomplete Set of keys that is expected to be missing from dct.
3641 Can be True/False to indicate all/none of the keys may be missing.
3642 All conditions on incomplete keys pass if the key is missing
3645 _match_one(filter_part.replace(r'\&', '&'), dct, incomplete)
3646 for filter_part in re.split(r'(?<!\\)&', filter_str))
3649 def match_filter_func(filters):
3652 filters = set(variadic(filters))
3654 interactive = '-' in filters
3658 def _match_func(info_dict, incomplete=False):
3659 if not filters or any(match_str(f, info_dict, incomplete) for f in filters):
3660 return NO_DEFAULT if interactive and not incomplete else None
3662 video_title = info_dict.get('title') or info_dict.get('id') or 'entry'
3663 filter_str = ') | ('.join(map(str.strip, filters))
3664 return f'{video_title} does not pass filter ({filter_str}), skipping ..'
3668 class download_range_func:
3669 def __init__(self, chapters, ranges):
3670 self.chapters, self.ranges = chapters, ranges
3672 def __call__(self, info_dict, ydl):
3673 warning = ('There are no chapters matching the regex' if info_dict.get('chapters')
3674 else 'Cannot match chapters since chapter information is unavailable')
3675 for regex in self.chapters or []:
3676 for i, chapter in enumerate(info_dict.get('chapters') or []):
3677 if re.search(regex, chapter['title']):
3679 yield {**chapter, 'index': i}
3680 if self.chapters and warning:
3681 ydl.to_screen(f'[info] {info_dict["id"]}: {warning}')
3683 yield from ({'start_time': start, 'end_time': end} for start, end in self.ranges or [])
3685 def __eq__(self, other):
3686 return (isinstance(other, download_range_func)
3687 and self.chapters == other.chapters and self.ranges == other.ranges)
3690 def parse_dfxp_time_expr(time_expr):
3694 mobj = re.match(rf'^(?P<time_offset>{NUMBER_RE})s?$', time_expr)
3696 return float(mobj.group('time_offset'))
3698 mobj = re.match(r'^(\d+):(\d\d):(\d\d(?:(?:\.|:)\d+)?)$', time_expr)
3700 return 3600 * int(mobj.group(1)) + 60 * int(mobj.group(2)) + float(mobj.group(3).replace(':', '.'))
3703 def srt_subtitles_timecode(seconds):
3704 return '%02d:%02d:%02d,%03d' % timetuple_from_msec(seconds * 1000)
3707 def ass_subtitles_timecode(seconds):
3708 time = timetuple_from_msec(seconds * 1000)
3709 return '%01d:%02d:%02d.%02d' % (*time[:-1], time.milliseconds / 10)
3712 def dfxp2srt(dfxp_data):
3714 @param dfxp_data A
bytes-like
object containing DFXP data
3715 @returns A
unicode object containing converted SRT data
3717 LEGACY_NAMESPACES = (
3718 (b'http://www.w3.org/ns/ttml', [
3719 b'http://www.w3.org/2004/11/ttaf1',
3720 b'http://www.w3.org/2006/04/ttaf1',
3721 b'http://www.w3.org/2006/10/ttaf1',
3723 (b'http://www.w3.org/ns/ttml#styling', [
3724 b'http://www.w3.org/ns/ttml#style',
3728 SUPPORTED_STYLING = [
3737 _x = functools.partial(xpath_with_ns, ns_map={
3738 'xml': 'http://www.w3.org/XML/1998/namespace',
3739 'ttml': 'http://www.w3.org/ns/ttml',
3740 'tts': 'http://www.w3.org/ns/ttml#styling',
3746 class TTMLPElementParser:
3748 _unclosed_elements = []
3749 _applied_styles = []
3751 def start(self, tag, attrib):
3752 if tag in (_x('ttml:br'), 'br'):
3755 unclosed_elements = []
3757 element_style_id = attrib.get('style')
3759 style.update(default_style)
3760 if element_style_id:
3761 style.update(styles.get(element_style_id, {}))
3762 for prop in SUPPORTED_STYLING:
3763 prop_val = attrib.get(_x('tts:' + prop))
3765 style[prop] = prop_val
3768 for k, v in sorted(style.items()):
3769 if self._applied_styles and self._applied_styles[-1].get(k) == v:
3772 font += ' color="%s"' % v
3773 elif k == 'fontSize':
3774 font += ' size="%s"' % v
3775 elif k == 'fontFamily':
3776 font += ' face="%s"' % v
3777 elif k == 'fontWeight' and v == 'bold':
3779 unclosed_elements.append('b')
3780 elif k == 'fontStyle' and v == 'italic':
3782 unclosed_elements.append('i')
3783 elif k == 'textDecoration' and v == 'underline':
3785 unclosed_elements.append('u')
3787 self._out += '<font' + font + '>'
3788 unclosed_elements.append('font')
3790 if self._applied_styles:
3791 applied_style.update(self._applied_styles[-1])
3792 applied_style.update(style)
3793 self._applied_styles.append(applied_style)
3794 self._unclosed_elements.append(unclosed_elements)
3797 if tag not in (_x('ttml:br'), 'br'):
3798 unclosed_elements = self._unclosed_elements.pop()
3799 for element in reversed(unclosed_elements):
3800 self._out += '</%s>' % element
3801 if unclosed_elements and self._applied_styles:
3802 self._applied_styles.pop()
3804 def data(self, data):
3808 return self._out.strip()
3810 def parse_node(node):
3811 target = TTMLPElementParser()
3812 parser = xml.etree.ElementTree.XMLParser(target=target)
3813 parser.feed(xml.etree.ElementTree.tostring(node))
3814 return parser.close()
3816 for k, v in LEGACY_NAMESPACES:
3818 dfxp_data = dfxp_data.replace(ns, k)
3820 dfxp = compat_etree_fromstring(dfxp_data)
3822 paras = dfxp.findall(_x('.//ttml:p')) or dfxp.findall('.//p')
3825 raise ValueError('Invalid dfxp/TTML subtitle')
3829 for style in dfxp.findall(_x('.//ttml:style')):
3830 style_id = style.get('id') or style.get(_x('xml:id'))
3833 parent_style_id = style.get('style')
3835 if parent_style_id not in styles:
3838 styles[style_id] = styles[parent_style_id].copy()
3839 for prop in SUPPORTED_STYLING:
3840 prop_val = style.get(_x('tts:' + prop))
3842 styles.setdefault(style_id, {})[prop] = prop_val
3848 for p in ('body', 'div'):
3849 ele = xpath_element(dfxp, [_x('.//ttml:' + p), './/' + p])
3852 style = styles.get(ele.get('style'))
3855 default_style.update(style)
3857 for para, index in zip(paras, itertools.count(1)):
3858 begin_time = parse_dfxp_time_expr(para.attrib.get('begin'))
3859 end_time = parse_dfxp_time_expr(para.attrib.get('end'))
3860 dur = parse_dfxp_time_expr(para.attrib.get('dur'))
3861 if begin_time is None:
3866 end_time = begin_time + dur
3867 out.append('%d\n%s --> %s\n%s\n\n' % (
3869 srt_subtitles_timecode(begin_time),
3870 srt_subtitles_timecode(end_time),
3876 def cli_option(params, command_option, param, separator=None):
3877 param = params.get(param)
3878 return ([] if param is None
3879 else [command_option, str(param)] if separator is None
3880 else [f'{command_option}{separator}{param}'])
3883 def cli_bool_option(params, command_option, param, true_value='true', false_value='false', separator=None):
3884 param = params.get(param)
3885 assert param in (True, False, None)
3886 return cli_option({True: true_value, False: false_value}, command_option, param, separator)
3889 def cli_valueless_option(params, command_option, param, expected_value=True):
3890 return [command_option] if params.get(param) == expected_value else []
3893 def cli_configuration_args(argdict, keys, default=[], use_compat=True):
3894 if isinstance(argdict, (list, tuple)): # for backward compatibility
3901 assert isinstance(argdict, dict)
3903 assert isinstance(keys, (list, tuple))
3904 for key_list in keys:
3905 arg_list = list(filter(
3906 lambda x: x is not None,
3907 [argdict.get(key.lower()) for key in variadic(key_list)]))
3909 return [arg for args in arg_list for arg in args]
3913 def _configuration_args(main_key, argdict, exe, keys=None, default=[], use_compat=True):
3914 main_key, exe = main_key.lower(), exe.lower()
3915 root_key = exe if main_key == exe else f'{main_key}+{exe}'
3916 keys = [f'{root_key}{k}' for k in (keys or [''])]
3917 if root_key in keys:
3919 keys.append((main_key, exe))
3920 keys.append('default')
3923 return cli_configuration_args(argdict, keys, default, use_compat)
3927 # See http://www.loc.gov/standards/iso639-2/ISO-639-2_utf-8.txt
3986 'iw': 'heb', # Replaced by he in 1989 revision
3996 'in': 'ind', # Replaced by id in 1989 revision
4111 'ji': 'yid', # Replaced by yi in 1989 revision
4119 def short2long(cls, code):
4120 """Convert language code from ISO 639-1 to ISO 639-2/T"""
4121 return cls._lang_map.get(code[:2])
4124 def long2short(cls, code):
4125 """Convert language code from ISO 639-2/T to ISO 639-1"""
4126 for short_name, long_name in cls._lang_map.items():
4127 if long_name == code:
4132 # From http://data.okfn.org/data/core/country-list
4134 'AF': 'Afghanistan',
4135 'AX': 'Åland Islands',
4138 'AS': 'American Samoa',
4143 'AG': 'Antigua and Barbuda',
4160 'BO': 'Bolivia, Plurinational State of',
4161 'BQ': 'Bonaire, Sint Eustatius and Saba',
4162 'BA': 'Bosnia and Herzegovina',
4164 'BV': 'Bouvet Island',
4166 'IO': 'British Indian Ocean Territory',
4167 'BN': 'Brunei Darussalam',
4169 'BF': 'Burkina Faso',
4175 'KY': 'Cayman Islands',
4176 'CF': 'Central African Republic',
4180 'CX': 'Christmas Island',
4181 'CC': 'Cocos (Keeling) Islands',
4185 'CD': 'Congo, the Democratic Republic of the',
4186 'CK': 'Cook Islands',
4188 'CI': 'Côte d\'Ivoire',
4193 'CZ': 'Czech Republic',
4197 'DO': 'Dominican Republic',
4200 'SV': 'El Salvador',
4201 'GQ': 'Equatorial Guinea',
4205 'FK': 'Falkland Islands (Malvinas)',
4206 'FO': 'Faroe Islands',
4210 'GF': 'French Guiana',
4211 'PF': 'French Polynesia',
4212 'TF': 'French Southern Territories',
4227 'GW': 'Guinea-Bissau',
4230 'HM': 'Heard Island and McDonald Islands',
4231 'VA': 'Holy See (Vatican City State)',
4238 'IR': 'Iran, Islamic Republic of',
4241 'IM': 'Isle of Man',
4251 'KP': 'Korea, Democratic People\'s Republic of',
4252 'KR': 'Korea, Republic of',
4255 'LA': 'Lao People\'s Democratic Republic',
4261 'LI': 'Liechtenstein',
4265 'MK': 'Macedonia, the Former Yugoslav Republic of',
4272 'MH': 'Marshall Islands',
4278 'FM': 'Micronesia, Federated States of',
4279 'MD': 'Moldova, Republic of',
4290 'NL': 'Netherlands',
4291 'NC': 'New Caledonia',
4292 'NZ': 'New Zealand',
4297 'NF': 'Norfolk Island',
4298 'MP': 'Northern Mariana Islands',
4303 'PS': 'Palestine, State of',
4305 'PG': 'Papua New Guinea',
4308 'PH': 'Philippines',
4312 'PR': 'Puerto Rico',
4316 'RU': 'Russian Federation',
4318 'BL': 'Saint Barthélemy',
4319 'SH': 'Saint Helena, Ascension and Tristan da Cunha',
4320 'KN': 'Saint Kitts and Nevis',
4321 'LC': 'Saint Lucia',
4322 'MF': 'Saint Martin (French part)',
4323 'PM': 'Saint Pierre and Miquelon',
4324 'VC': 'Saint Vincent and the Grenadines',
4327 'ST': 'Sao Tome and Principe',
4328 'SA': 'Saudi Arabia',
4332 'SL': 'Sierra Leone',
4334 'SX': 'Sint Maarten (Dutch part)',
4337 'SB': 'Solomon Islands',
4339 'ZA': 'South Africa',
4340 'GS': 'South Georgia and the South Sandwich Islands',
4341 'SS': 'South Sudan',
4346 'SJ': 'Svalbard and Jan Mayen',
4349 'CH': 'Switzerland',
4350 'SY': 'Syrian Arab Republic',
4351 'TW': 'Taiwan, Province of China',
4353 'TZ': 'Tanzania, United Republic of',
4355 'TL': 'Timor-Leste',
4359 'TT': 'Trinidad and Tobago',
4362 'TM': 'Turkmenistan',
4363 'TC': 'Turks and Caicos Islands',
4367 'AE': 'United Arab Emirates',
4368 'GB': 'United Kingdom',
4369 'US': 'United States',
4370 'UM': 'United States Minor Outlying Islands',
4374 'VE': 'Venezuela, Bolivarian Republic of',
4376 'VG': 'Virgin Islands, British',
4377 'VI': 'Virgin Islands, U.S.',
4378 'WF': 'Wallis and Futuna',
4379 'EH': 'Western Sahara',
4383 # Not ISO 3166 codes, but used for IP blocks
4384 'AP': 'Asia/Pacific Region',
4389 def short2full(cls, code):
4390 """Convert an ISO 3166-2 country code to the corresponding full name"""
4391 return cls._country_map.get(code.upper())
4395 # Major IPv4 address blocks per country
4397 'AD': '46.172.224.0/19',
4398 'AE': '94.200.0.0/13',
4399 'AF': '149.54.0.0/17',
4400 'AG': '209.59.64.0/18',
4401 'AI': '204.14.248.0/21',
4402 'AL': '46.99.0.0/16',
4403 'AM': '46.70.0.0/15',
4404 'AO': '105.168.0.0/13',
4405 'AP': '182.50.184.0/21',
4406 'AQ': '23.154.160.0/24',
4407 'AR': '181.0.0.0/12',
4408 'AS': '202.70.112.0/20',
4409 'AT': '77.116.0.0/14',
4410 'AU': '1.128.0.0/11',
4411 'AW': '181.41.0.0/18',
4412 'AX': '185.217.4.0/22',
4413 'AZ': '5.197.0.0/16',
4414 'BA': '31.176.128.0/17',
4415 'BB': '65.48.128.0/17',
4416 'BD': '114.130.0.0/16',
4418 'BF': '102.178.0.0/15',
4419 'BG': '95.42.0.0/15',
4420 'BH': '37.131.0.0/17',
4421 'BI': '154.117.192.0/18',
4422 'BJ': '137.255.0.0/16',
4423 'BL': '185.212.72.0/23',
4424 'BM': '196.12.64.0/18',
4425 'BN': '156.31.0.0/16',
4426 'BO': '161.56.0.0/16',
4427 'BQ': '161.0.80.0/20',
4428 'BR': '191.128.0.0/12',
4429 'BS': '24.51.64.0/18',
4430 'BT': '119.2.96.0/19',
4431 'BW': '168.167.0.0/16',
4432 'BY': '178.120.0.0/13',
4433 'BZ': '179.42.192.0/18',
4434 'CA': '99.224.0.0/11',
4435 'CD': '41.243.0.0/16',
4436 'CF': '197.242.176.0/21',
4437 'CG': '160.113.0.0/16',
4438 'CH': '85.0.0.0/13',
4439 'CI': '102.136.0.0/14',
4440 'CK': '202.65.32.0/19',
4441 'CL': '152.172.0.0/14',
4442 'CM': '102.244.0.0/14',
4443 'CN': '36.128.0.0/10',
4444 'CO': '181.240.0.0/12',
4445 'CR': '201.192.0.0/12',
4446 'CU': '152.206.0.0/15',
4447 'CV': '165.90.96.0/19',
4448 'CW': '190.88.128.0/17',
4449 'CY': '31.153.0.0/16',
4450 'CZ': '88.100.0.0/14',
4452 'DJ': '197.241.0.0/17',
4453 'DK': '87.48.0.0/12',
4454 'DM': '192.243.48.0/20',
4455 'DO': '152.166.0.0/15',
4456 'DZ': '41.96.0.0/12',
4457 'EC': '186.68.0.0/15',
4458 'EE': '90.190.0.0/15',
4459 'EG': '156.160.0.0/11',
4460 'ER': '196.200.96.0/20',
4461 'ES': '88.0.0.0/11',
4462 'ET': '196.188.0.0/14',
4463 'EU': '2.16.0.0/13',
4464 'FI': '91.152.0.0/13',
4465 'FJ': '144.120.0.0/16',
4466 'FK': '80.73.208.0/21',
4467 'FM': '119.252.112.0/20',
4468 'FO': '88.85.32.0/19',
4470 'GA': '41.158.0.0/15',
4472 'GD': '74.122.88.0/21',
4473 'GE': '31.146.0.0/16',
4474 'GF': '161.22.64.0/18',
4475 'GG': '62.68.160.0/19',
4476 'GH': '154.160.0.0/12',
4477 'GI': '95.164.0.0/16',
4478 'GL': '88.83.0.0/19',
4479 'GM': '160.182.0.0/15',
4480 'GN': '197.149.192.0/18',
4481 'GP': '104.250.0.0/19',
4482 'GQ': '105.235.224.0/20',
4483 'GR': '94.64.0.0/13',
4484 'GT': '168.234.0.0/16',
4485 'GU': '168.123.0.0/16',
4486 'GW': '197.214.80.0/20',
4487 'GY': '181.41.64.0/18',
4488 'HK': '113.252.0.0/14',
4489 'HN': '181.210.0.0/16',
4490 'HR': '93.136.0.0/13',
4491 'HT': '148.102.128.0/17',
4492 'HU': '84.0.0.0/14',
4493 'ID': '39.192.0.0/10',
4494 'IE': '87.32.0.0/12',
4495 'IL': '79.176.0.0/13',
4496 'IM': '5.62.80.0/20',
4497 'IN': '117.192.0.0/10',
4498 'IO': '203.83.48.0/21',
4499 'IQ': '37.236.0.0/14',
4500 'IR': '2.176.0.0/12',
4501 'IS': '82.221.0.0/16',
4502 'IT': '79.0.0.0/10',
4503 'JE': '87.244.64.0/18',
4504 'JM': '72.27.0.0/17',
4505 'JO': '176.29.0.0/16',
4506 'JP': '133.0.0.0/8',
4507 'KE': '105.48.0.0/12',
4508 'KG': '158.181.128.0/17',
4509 'KH': '36.37.128.0/17',
4510 'KI': '103.25.140.0/22',
4511 'KM': '197.255.224.0/20',
4512 'KN': '198.167.192.0/19',
4513 'KP': '175.45.176.0/22',
4514 'KR': '175.192.0.0/10',
4515 'KW': '37.36.0.0/14',
4516 'KY': '64.96.0.0/15',
4517 'KZ': '2.72.0.0/13',
4518 'LA': '115.84.64.0/18',
4519 'LB': '178.135.0.0/16',
4520 'LC': '24.92.144.0/20',
4521 'LI': '82.117.0.0/19',
4522 'LK': '112.134.0.0/15',
4523 'LR': '102.183.0.0/16',
4524 'LS': '129.232.0.0/17',
4525 'LT': '78.56.0.0/13',
4526 'LU': '188.42.0.0/16',
4527 'LV': '46.109.0.0/16',
4528 'LY': '41.252.0.0/14',
4529 'MA': '105.128.0.0/11',
4530 'MC': '88.209.64.0/18',
4531 'MD': '37.246.0.0/16',
4532 'ME': '178.175.0.0/17',
4533 'MF': '74.112.232.0/21',
4534 'MG': '154.126.0.0/17',
4535 'MH': '117.103.88.0/21',
4536 'MK': '77.28.0.0/15',
4537 'ML': '154.118.128.0/18',
4538 'MM': '37.111.0.0/17',
4539 'MN': '49.0.128.0/17',
4540 'MO': '60.246.0.0/16',
4541 'MP': '202.88.64.0/20',
4542 'MQ': '109.203.224.0/19',
4543 'MR': '41.188.64.0/18',
4544 'MS': '208.90.112.0/22',
4545 'MT': '46.11.0.0/16',
4546 'MU': '105.16.0.0/12',
4547 'MV': '27.114.128.0/18',
4548 'MW': '102.70.0.0/15',
4549 'MX': '187.192.0.0/11',
4550 'MY': '175.136.0.0/13',
4551 'MZ': '197.218.0.0/15',
4552 'NA': '41.182.0.0/16',
4553 'NC': '101.101.0.0/18',
4554 'NE': '197.214.0.0/18',
4555 'NF': '203.17.240.0/22',
4556 'NG': '105.112.0.0/12',
4557 'NI': '186.76.0.0/15',
4558 'NL': '145.96.0.0/11',
4559 'NO': '84.208.0.0/13',
4560 'NP': '36.252.0.0/15',
4561 'NR': '203.98.224.0/19',
4562 'NU': '49.156.48.0/22',
4563 'NZ': '49.224.0.0/14',
4564 'OM': '5.36.0.0/15',
4565 'PA': '186.72.0.0/15',
4566 'PE': '186.160.0.0/14',
4567 'PF': '123.50.64.0/18',
4568 'PG': '124.240.192.0/19',
4569 'PH': '49.144.0.0/13',
4570 'PK': '39.32.0.0/11',
4571 'PL': '83.0.0.0/11',
4572 'PM': '70.36.0.0/20',
4573 'PR': '66.50.0.0/16',
4574 'PS': '188.161.0.0/16',
4575 'PT': '85.240.0.0/13',
4576 'PW': '202.124.224.0/20',
4577 'PY': '181.120.0.0/14',
4578 'QA': '37.210.0.0/15',
4579 'RE': '102.35.0.0/16',
4580 'RO': '79.112.0.0/13',
4581 'RS': '93.86.0.0/15',
4582 'RU': '5.136.0.0/13',
4583 'RW': '41.186.0.0/16',
4584 'SA': '188.48.0.0/13',
4585 'SB': '202.1.160.0/19',
4586 'SC': '154.192.0.0/11',
4587 'SD': '102.120.0.0/13',
4588 'SE': '78.64.0.0/12',
4589 'SG': '8.128.0.0/10',
4590 'SI': '188.196.0.0/14',
4591 'SK': '78.98.0.0/15',
4592 'SL': '102.143.0.0/17',
4593 'SM': '89.186.32.0/19',
4594 'SN': '41.82.0.0/15',
4595 'SO': '154.115.192.0/18',
4596 'SR': '186.179.128.0/17',
4597 'SS': '105.235.208.0/21',
4598 'ST': '197.159.160.0/19',
4599 'SV': '168.243.0.0/16',
4600 'SX': '190.102.0.0/20',
4602 'SZ': '41.84.224.0/19',
4603 'TC': '65.255.48.0/20',
4604 'TD': '154.68.128.0/19',
4605 'TG': '196.168.0.0/14',
4606 'TH': '171.96.0.0/13',
4607 'TJ': '85.9.128.0/18',
4608 'TK': '27.96.24.0/21',
4609 'TL': '180.189.160.0/20',
4610 'TM': '95.85.96.0/19',
4611 'TN': '197.0.0.0/11',
4612 'TO': '175.176.144.0/21',
4613 'TR': '78.160.0.0/11',
4614 'TT': '186.44.0.0/15',
4615 'TV': '202.2.96.0/19',
4616 'TW': '120.96.0.0/11',
4617 'TZ': '156.156.0.0/14',
4618 'UA': '37.52.0.0/14',
4619 'UG': '102.80.0.0/13',
4621 'UY': '167.56.0.0/13',
4622 'UZ': '84.54.64.0/18',
4623 'VA': '212.77.0.0/19',
4624 'VC': '207.191.240.0/21',
4625 'VE': '186.88.0.0/13',
4626 'VG': '66.81.192.0/20',
4627 'VI': '146.226.0.0/16',
4628 'VN': '14.160.0.0/11',
4629 'VU': '202.80.32.0/20',
4630 'WF': '117.20.32.0/21',
4631 'WS': '202.4.32.0/19',
4632 'YE': '134.35.0.0/16',
4633 'YT': '41.242.116.0/22',
4634 'ZA': '41.0.0.0/11',
4635 'ZM': '102.144.0.0/13',
4636 'ZW': '102.177.192.0/18',
4640 def random_ipv4(cls, code_or_block):
4641 if len(code_or_block) == 2:
4642 block = cls._country_ip_map.get(code_or_block.upper())
4646 block = code_or_block
4647 addr, preflen = block.split('/')
4648 addr_min = struct.unpack('!L', socket.inet_aton(addr))[0]
4649 addr_max = addr_min | (0xffffffff >> int(preflen))
4650 return str(socket.inet_ntoa(
4651 struct.pack('!L', random.randint(addr_min, addr_max))))
4654 class PerRequestProxyHandler(urllib.request.ProxyHandler):
4655 def __init__(self, proxies=None):
4656 # Set default handlers
4657 for type in ('http', 'https'):
4658 setattr(self, '%s_open' % type,
4659 lambda r, proxy='__noproxy__', type=type, meth=self.proxy_open:
4660 meth(r, proxy, type))
4661 urllib.request.ProxyHandler.__init__(self, proxies)
4663 def proxy_open(self, req, proxy, type):
4664 req_proxy = req.headers.get('Ytdl-request-proxy')
4665 if req_proxy is not None:
4667 del req.headers['Ytdl-request-proxy']
4669 if proxy == '__noproxy__':
4670 return None # No Proxy
4671 if urllib.parse.urlparse(proxy).scheme.lower() in ('socks', 'socks4', 'socks4a', 'socks5'):
4672 req.add_header('Ytdl-socks-proxy', proxy)
4673 # yt-dlp's http/https handlers do wrapping the socket with socks
4675 return urllib.request.ProxyHandler.proxy_open(
4676 self, req, proxy, type)
4679 # Both long_to_bytes and bytes_to_long are adapted from PyCrypto, which is
4680 # released into Public Domain
4681 # https://github.com/dlitz/pycrypto/blob/master/lib/Crypto/Util/number.py#L387
4683 def long_to_bytes(n, blocksize=0):
4684 """long_to_bytes(n:long, blocksize:int) : string
4685 Convert a long integer to a byte string.
4687 If optional blocksize is given and greater than zero, pad the front of the
4688 byte string with binary zeros so that the length is a multiple of
4691 # after much testing, this algorithm was deemed to be the fastest
4695 s = struct.pack('>I', n & 0xffffffff) + s
4697 # strip off leading zeros
4698 for i in range(len(s)):
4699 if s[i] != b'\000'[0]:
4702 # only happens when n == 0
4706 # add back some pad bytes. this could be done more efficiently w.r.t. the
4707 # de-padding being done above, but sigh...
4708 if blocksize > 0 and len(s) % blocksize:
4709 s = (blocksize - len(s) % blocksize) * b'\000' + s
4713 def bytes_to_long(s):
4714 """bytes_to_long(string) : long
4715 Convert a byte string to a long integer.
4717 This is (essentially) the inverse of long_to_bytes().
4722 extra = (4 - length % 4)
4723 s = b'\000' * extra + s
4724 length = length + extra
4725 for i in range(0, length, 4):
4726 acc = (acc << 32) + struct.unpack('>I', s[i:i + 4])[0]
4730 def ohdave_rsa_encrypt(data, exponent, modulus):
4732 Implement OHDave
's RSA algorithm. See http://www.ohdave.com/rsa/
4735 data: data to encrypt, bytes-like object
4736 exponent, modulus: parameter e and N of RSA algorithm, both integer
4737 Output: hex string of encrypted data
4739 Limitation: supports one block encryption only
4742 payload = int(binascii.hexlify(data[::-1]), 16)
4743 encrypted = pow(payload, exponent, modulus)
4744 return '%x' % encrypted
4747 def pkcs1pad(data, length):
4749 Padding input data with PKCS#1 scheme
4751 @param {int[]} data input data
4752 @param {int} length target length
4753 @returns {int[]} padded data
4755 if len(data) > length - 11:
4756 raise ValueError('Input data too
long for PKCS
#1 padding')
4758 pseudo_random
= [random
.randint(0, 254) for _
in range(length
- len(data
) - 3)]
4759 return [0, 2] + pseudo_random
+ [0] + data
4762 def _base_n_table(n
, table
):
4763 if not table
and not n
:
4764 raise ValueError('Either table or n must be specified')
4765 table
= (table
or '0123456789abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ')[:n
]
4767 if n
and n
!= len(table
):
4768 raise ValueError(f
'base {n} exceeds table length {len(table)}')
4772 def encode_base_n(num
, n
=None, table
=None):
4773 """Convert given int to a base-n string"""
4774 table
= _base_n_table(n
, table
)
4778 result
, base
= '', len(table
)
4780 result
= table
[num
% base
] + result
4785 def decode_base_n(string
, n
=None, table
=None):
4786 """Convert given base-n string to int"""
4787 table
= {char: index for index, char in enumerate(_base_n_table(n, table))}
4788 result
, base
= 0, len(table
)
4790 result
= result
* base
+ table
[char
]
4794 def decode_base(value
, digits
):
4795 write_string('DeprecationWarning: yt_dlp.utils.decode_base is deprecated '
4796 'and may be removed in a future version. Use yt_dlp.decode_base_n instead')
4797 return decode_base_n(value
, table
=digits
)
4800 def decode_packed_codes(code
):
4801 mobj
= re
.search(PACKED_CODES_RE
, code
)
4802 obfuscated_code
, base
, count
, symbols
= mobj
.groups()
4805 symbols
= symbols
.split('|')
4810 base_n_count
= encode_base_n(count
, base
)
4811 symbol_table
[base_n_count
] = symbols
[count
] or base_n_count
4814 r
'\b(\w+)\b', lambda mobj
: symbol_table
[mobj
.group(0)],
4818 def caesar(s
, alphabet
, shift
):
4823 alphabet
[(alphabet
.index(c
) + shift
) % l
] if c
in alphabet
else c
4828 return caesar(s
, r
'''!"#$%&'()*+,-./0123456789:;<=>?@ABCDEFGHIJKLMNOPQRSTUVWXYZ[\]^_`abcdefghijklmnopqrstuvwxyz{|}~''', 47)
4831 def parse_m3u8_attributes(attrib
):
4833 for (key
, val
) in re
.findall(r
'(?P<key>[A-Z0-9-]+)=(?P<val>"[^"]+"|[^",]+)(?:,|$)', attrib
):
4834 if val
.startswith('"'):
4840 def urshift(val
, n
):
4841 return val
>> n
if val
>= 0 else (val
+ 0x100000000) >> n
4844 # Based on png2str() written by @gdkchan and improved by @yokrysty
4845 # Originally posted at https://github.com/ytdl-org/youtube-dl/issues/9706
4846 def decode_png(png_data
):
4847 # Reference: https://www.w3.org/TR/PNG/
4848 header
= png_data
[8:]
4850 if png_data
[:8] != b
'\x89PNG\x0d\x0a\x1a\x0a' or header
[4:8] != b
'IHDR':
4851 raise OSError('Not a valid PNG file.')
4853 int_map
= {1: '>B', 2: '>H', 4: '>I'}
4854 unpack_integer
= lambda x
: struct
.unpack(int_map
[len(x
)], x
)[0]
4859 length
= unpack_integer(header
[:4])
4862 chunk_type
= header
[:4]
4865 chunk_data
= header
[:length
]
4866 header
= header
[length
:]
4868 header
= header
[4:] # Skip CRC
4876 ihdr
= chunks
[0]['data']
4878 width
= unpack_integer(ihdr
[:4])
4879 height
= unpack_integer(ihdr
[4:8])
4883 for chunk
in chunks
:
4884 if chunk
['type'] == b
'IDAT':
4885 idat
+= chunk
['data']
4888 raise OSError('Unable to read PNG data.')
4890 decompressed_data
= bytearray(zlib
.decompress(idat
))
4895 def _get_pixel(idx
):
4900 for y
in range(height
):
4901 basePos
= y
* (1 + stride
)
4902 filter_type
= decompressed_data
[basePos
]
4906 pixels
.append(current_row
)
4908 for x
in range(stride
):
4909 color
= decompressed_data
[1 + basePos
+ x
]
4910 basex
= y
* stride
+ x
4915 left
= _get_pixel(basex
- 3)
4917 up
= _get_pixel(basex
- stride
)
4919 if filter_type
== 1: # Sub
4920 color
= (color
+ left
) & 0xff
4921 elif filter_type
== 2: # Up
4922 color
= (color
+ up
) & 0xff
4923 elif filter_type
== 3: # Average
4924 color
= (color
+ ((left
+ up
) >> 1)) & 0xff
4925 elif filter_type
== 4: # Paeth
4931 c
= _get_pixel(basex
- stride
- 3)
4939 if pa
<= pb
and pa
<= pc
:
4940 color
= (color
+ a
) & 0xff
4942 color
= (color
+ b
) & 0xff
4944 color
= (color
+ c
) & 0xff
4946 current_row
.append(color
)
4948 return width
, height
, pixels
4951 def write_xattr(path
, key
, value
):
4952 # Windows: Write xattrs to NTFS Alternate Data Streams:
4953 # http://en.wikipedia.org/wiki/NTFS#Alternate_data_streams_.28ADS.29
4954 if compat_os_name
== 'nt':
4955 assert ':' not in key
4956 assert os
.path
.exists(path
)
4959 with open(f
'{path}:{key}', 'wb') as f
:
4961 except OSError as e
:
4962 raise XAttrMetadataError(e
.errno
, e
.strerror
)
4965 # UNIX Method 1. Use xattrs/pyxattrs modules
4968 if getattr(xattr
, '_yt_dlp__identifier', None) == 'pyxattr':
4969 # Unicode arguments are not supported in pyxattr until version 0.5.0
4970 # See https://github.com/ytdl-org/youtube-dl/issues/5498
4971 if version_tuple(xattr
.__version
__) >= (0, 5, 0):
4972 setxattr
= xattr
.set
4974 setxattr
= xattr
.setxattr
4978 setxattr(path
, key
, value
)
4979 except OSError as e
:
4980 raise XAttrMetadataError(e
.errno
, e
.strerror
)
4983 # UNIX Method 2. Use setfattr/xattr executables
4984 exe
= ('setfattr' if check_executable('setfattr', ['--version'])
4985 else 'xattr' if check_executable('xattr', ['-h']) else None)
4987 raise XAttrUnavailableError(
4988 'Couldn\'t find a tool to set the xattrs. Install either the python "xattr" or "pyxattr" modules or the '
4989 + ('"xattr" binary' if sys
.platform
!= 'linux' else 'GNU "attr" package (which contains the "setfattr" tool)'))
4991 value
= value
.decode()
4993 _
, stderr
, returncode
= Popen
.run(
4994 [exe
, '-w', key
, value
, path
] if exe
== 'xattr' else [exe
, '-n', key
, '-v', value
, path
],
4995 text
=True, stdout
=subprocess
.PIPE
, stderr
=subprocess
.PIPE
, stdin
=subprocess
.PIPE
)
4996 except OSError as e
:
4997 raise XAttrMetadataError(e
.errno
, e
.strerror
)
4999 raise XAttrMetadataError(returncode
, stderr
)
5002 def random_birthday(year_field
, month_field
, day_field
):
5003 start_date
= datetime
.date(1950, 1, 1)
5004 end_date
= datetime
.date(1995, 12, 31)
5005 offset
= random
.randint(0, (end_date
- start_date
).days
)
5006 random_date
= start_date
+ datetime
.timedelta(offset
)
5008 year_field
: str(random_date
.year
),
5009 month_field
: str(random_date
.month
),
5010 day_field
: str(random_date
.day
),
5014 # Templates for internet shortcut files, which are plain text files.
5015 DOT_URL_LINK_TEMPLATE
= '''\
5020 DOT_WEBLOC_LINK_TEMPLATE
= '''\
5021 <?xml version="1.0" encoding="UTF-8"?>
5022 <!DOCTYPE plist PUBLIC "-//Apple//DTD PLIST 1.0//EN" "http://www.apple.com/DTDs/PropertyList-1.0.dtd">
5023 <plist version="1.0">
5026 \t<string>%(url)s</string>
5031 DOT_DESKTOP_LINK_TEMPLATE
= '''\
5041 'url': DOT_URL_LINK_TEMPLATE
,
5042 'desktop': DOT_DESKTOP_LINK_TEMPLATE
,
5043 'webloc': DOT_WEBLOC_LINK_TEMPLATE
,
5047 def iri_to_uri(iri
):
5049 Converts an IRI (Internationalized Resource Identifier, allowing Unicode characters) to a URI (Uniform Resource Identifier, ASCII-only).
5051 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.
5054 iri_parts
= urllib
.parse
.urlparse(iri
)
5056 if '[' in iri_parts
.netloc
:
5057 raise ValueError('IPv6 URIs are not, yet, supported.')
5058 # Querying `.netloc`, when there's only one bracket, also raises a ValueError.
5060 # 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.
5063 if iri_parts
.username
:
5064 net_location
+= urllib
.parse
.quote(iri_parts
.username
, safe
=r
"!$%&'()*+,~")
5065 if iri_parts
.password
is not None:
5066 net_location
+= ':' + urllib
.parse
.quote(iri_parts
.password
, safe
=r
"!$%&'()*+,~")
5069 net_location
+= iri_parts
.hostname
.encode('idna').decode() # Punycode for Unicode hostnames.
5070 # The 'idna' encoding produces ASCII text.
5071 if iri_parts
.port
is not None and iri_parts
.port
!= 80:
5072 net_location
+= ':' + str(iri_parts
.port
)
5074 return urllib
.parse
.urlunparse(
5078 urllib
.parse
.quote_plus(iri_parts
.path
, safe
=r
"!$%&'()*+,/:;=@|~"),
5080 # Unsure about the `safe` argument, since this is a legacy way of handling parameters.
5081 urllib
.parse
.quote_plus(iri_parts
.params
, safe
=r
"!$%&'()*+,/:;=@|~"),
5083 # Not totally sure about the `safe` argument, since the source does not explicitly mention the query URI component.
5084 urllib
.parse
.quote_plus(iri_parts
.query
, safe
=r
"!$%&'()*+,/:;=?@{|}~"),
5086 urllib
.parse
.quote_plus(iri_parts
.fragment
, safe
=r
"!#$%&'()*+,/:;=?@{|}~")))
5088 # Source for `safe` arguments: https://url.spec.whatwg.org/#percent-encoded-bytes.
5091 def to_high_limit_path(path
):
5092 if sys
.platform
in ['win32', 'cygwin']:
5093 # Work around MAX_PATH limitation on Windows. The maximum allowed length for the individual path segments may still be quite limited.
5094 return '\\\\?\\' + os
.path
.abspath(path
)
5099 def format_field(obj
, field
=None, template
='%s', ignore
=NO_DEFAULT
, default
='', func
=IDENTITY
):
5100 val
= traverse_obj(obj
, *variadic(field
))
5101 if (not val
and val
!= 0) if ignore
is NO_DEFAULT
else val
in variadic(ignore
):
5103 return template
% func(val
)
5106 def clean_podcast_url(url
):
5107 return re
.sub(r
'''(?x)
5111 media\.blubrry\.com| # https://create.blubrry.com/resources/podcast-media-download-statistics/getting-started/
5114 (?:dts|www)\.podtrac\.com/(?:pts/)?redirect\.[0-9a-z]{3,4}| # http://analytics.podtrac.com/how-to-measure
5117 cn\.co| # https://podcorn.com/analytics-prefix/
5118 st\.fm # https://podsights.com/docs/
5123 _HEX_TABLE
= '0123456789abcdef'
5126 def random_uuidv4():
5127 return re
.sub(r
'[xy]', lambda x
: _HEX_TABLE
[random
.randint(0, 15)], 'xxxxxxxx-xxxx-4xxx-yxxx-xxxxxxxxxxxx')
5130 def make_dir(path
, to_screen
=None):
5132 dn
= os
.path
.dirname(path
)
5133 if dn
and not os
.path
.exists(dn
):
5136 except OSError as err
:
5137 if callable(to_screen
) is not None:
5138 to_screen('unable to create directory ' + error_to_compat_str(err
))
5142 def get_executable_path():
5143 from .update
import _get_variant_and_executable_path
5145 return os
.path
.dirname(os
.path
.abspath(_get_variant_and_executable_path()[1]))
5148 def load_plugins(name
, suffix
, namespace
):
5150 with contextlib
.suppress(FileNotFoundError
):
5151 plugins_spec
= importlib
.util
.spec_from_file_location(
5152 name
, os
.path
.join(get_executable_path(), 'ytdlp_plugins', name
, '__init__.py'))
5153 plugins
= importlib
.util
.module_from_spec(plugins_spec
)
5154 sys
.modules
[plugins_spec
.name
] = plugins
5155 plugins_spec
.loader
.exec_module(plugins
)
5156 for name
in dir(plugins
):
5157 if name
in namespace
:
5159 if not name
.endswith(suffix
):
5161 klass
= getattr(plugins
, name
)
5162 classes
[name
] = namespace
[name
] = klass
5167 obj
, *path_list
, default
=None, expected_type
=None, get_all
=True,
5168 casesense
=True, is_user_input
=False, traverse_string
=False):
5169 ''' Traverse nested list/dict/tuple
5170 @param path_list A list of paths which are checked one by one.
5171 Each path is a list of keys where each key is a:
5173 - string: A dictionary key
5174 - int: An index into a list
5175 - tuple: A list of keys all of which will be traversed
5176 - Ellipsis: Fetch all values in the object
5177 - Function: Takes the key and value as arguments
5178 and returns whether the key matches or not
5179 @param default Default value to return
5180 @param expected_type Only accept final value of this type (Can also be any callable)
5181 @param get_all Return all the values obtained from a path or only the first one
5182 @param casesense Whether to consider dictionary keys as case sensitive
5183 @param is_user_input Whether the keys are generated from user input. If True,
5184 strings are converted to int/slice if necessary
5185 @param traverse_string Whether to traverse inside strings. If True, any
5186 non-compatible object will also be converted into a string
5190 _lower
= lambda k
: (k
.lower() if isinstance(k
, str) else k
)
5191 path_list
= (map(_lower
, variadic(path
)) for path
in path_list
)
5193 def _traverse_obj(obj
, path
, _current_depth
=0):
5195 path
= tuple(variadic(path
))
5196 for i
, key
in enumerate(path
):
5197 if None in (key
, obj
):
5199 if isinstance(key
, (list, tuple)):
5200 obj
= [_traverse_obj(obj
, sub_key
, _current_depth
) for sub_key
in key
]
5203 obj
= (obj
.values() if isinstance(obj
, dict)
5204 else obj
if isinstance(obj
, (list, tuple, LazyList
))
5205 else str(obj
) if traverse_string
else [])
5207 depth
= max(depth
, _current_depth
)
5208 return [_traverse_obj(inner_obj
, path
[i
+ 1:], _current_depth
) for inner_obj
in obj
]
5210 if isinstance(obj
, (list, tuple, LazyList
)):
5211 obj
= enumerate(obj
)
5212 elif isinstance(obj
, dict):
5215 if not traverse_string
:
5219 depth
= max(depth
, _current_depth
)
5220 return [_traverse_obj(v
, path
[i
+ 1:], _current_depth
) for k
, v
in obj
if try_call(key
, args
=(k
, v
))]
5221 elif isinstance(obj
, dict) and not (is_user_input
and key
== ':'):
5222 obj
= (obj
.get(key
) if casesense
or (key
in obj
)
5223 else next((v
for k
, v
in obj
.items() if _lower(k
) == key
), None))
5226 key
= (int_or_none(key
) if ':' not in key
5227 else slice(*map(int_or_none
, key
.split(':'))))
5228 if key
== slice(None):
5229 return _traverse_obj(obj
, (..., *path
[i
+ 1:]), _current_depth
)
5230 if not isinstance(key
, (int, slice)):
5232 if not isinstance(obj
, (list, tuple, LazyList
)):
5233 if not traverse_string
:
5242 if isinstance(expected_type
, type):
5243 type_test
= lambda val
: val
if isinstance(val
, expected_type
) else None
5245 type_test
= expected_type
or IDENTITY
5247 for path
in path_list
:
5249 val
= _traverse_obj(obj
, path
)
5252 for _
in range(depth
- 1):
5253 val
= itertools
.chain
.from_iterable(v
for v
in val
if v
is not None)
5254 val
= [v
for v
in map(type_test
, val
) if v
is not None]
5256 return val
if get_all
else val
[0]
5258 val
= type_test(val
)
5264 def traverse_dict(dictn
, keys
, casesense
=True):
5265 write_string('DeprecationWarning: yt_dlp.utils.traverse_dict is deprecated '
5266 'and may be removed in a future version. Use yt_dlp.utils.traverse_obj instead')
5267 return traverse_obj(dictn
, keys
, casesense
=casesense
, is_user_input
=True, traverse_string
=True)
5270 def get_first(obj
, keys
, **kwargs
):
5271 return traverse_obj(obj
, (..., *variadic(keys
)), **kwargs
, get_all
=False)
5274 def variadic(x
, allowed_types
=(str, bytes, dict)):
5275 return x
if isinstance(x
, collections
.abc
.Iterable
) and not isinstance(x
, allowed_types
) else (x
,)
5278 def time_seconds(**kwargs
):
5279 t
= datetime
.datetime
.now(datetime
.timezone(datetime
.timedelta(**kwargs
)))
5280 return t
.timestamp()
5283 # create a JSON Web Signature (jws) with HS256 algorithm
5284 # the resulting format is in JWS Compact Serialization
5285 # implemented following JWT https://www.rfc-editor.org/rfc/rfc7519.html
5286 # implemented following JWS https://www.rfc-editor.org/rfc/rfc7515.html
5287 def jwt_encode_hs256(payload_data
, key
, headers
={}):
5293 header_data
.update(headers
)
5294 header_b64
= base64
.b64encode(json
.dumps(header_data
).encode())
5295 payload_b64
= base64
.b64encode(json
.dumps(payload_data
).encode())
5296 h
= hmac
.new(key
.encode(), header_b64
+ b
'.' + payload_b64
, hashlib
.sha256
)
5297 signature_b64
= base64
.b64encode(h
.digest())
5298 token
= header_b64
+ b
'.' + payload_b64
+ b
'.' + signature_b64
5302 # can be extended in future to verify the signature and parse header and return the algorithm used if it's not HS256
5303 def jwt_decode_hs256(jwt
):
5304 header_b64
, payload_b64
, signature_b64
= jwt
.split('.')
5305 payload_data
= json
.loads(base64
.urlsafe_b64decode(payload_b64
))
5309 WINDOWS_VT_MODE
= False if compat_os_name
== 'nt' else None
5313 def supports_terminal_sequences(stream
):
5314 if compat_os_name
== 'nt':
5315 if not WINDOWS_VT_MODE
:
5317 elif not os
.getenv('TERM'):
5320 return stream
.isatty()
5321 except BaseException
:
5325 def windows_enable_vt_mode(): # TODO: Do this the proper way https://bugs.python.org/issue30075
5326 if get_windows_version() < (10, 0, 10586):
5328 global WINDOWS_VT_MODE
5330 Popen
.run('', shell
=True)
5334 WINDOWS_VT_MODE
= True
5335 supports_terminal_sequences
.cache_clear()
5338 _terminal_sequences_re
= re
.compile('\033\\[[^m]+m')
5341 def remove_terminal_sequences(string
):
5342 return _terminal_sequences_re
.sub('', string
)
5345 def number_of_digits(number
):
5346 return len('%d' % number
)
5349 def join_nonempty(*values
, delim
='-', from_dict
=None):
5350 if from_dict
is not None:
5351 values
= (traverse_obj(from_dict
, variadic(v
)) for v
in values
)
5352 return delim
.join(map(str, filter(None, values
)))
5355 def scale_thumbnails_to_max_format_width(formats
, thumbnails
, url_width_re
):
5357 Find the largest format dimensions in terms of video width and, for each thumbnail:
5358 * Modify the URL: Match the width with the provided regex and replace with the former width
5361 This function is useful with video services that scale the provided thumbnails on demand
5363 _keys
= ('width', 'height')
5364 max_dimensions
= max(
5365 (tuple(format
.get(k
) or 0 for k
in _keys
) for format
in formats
),
5367 if not max_dimensions
[0]:
5371 {'url': re.sub(url_width_re, str(max_dimensions[0]), thumbnail['url'])}
,
5372 dict(zip(_keys
, max_dimensions
)), thumbnail
)
5373 for thumbnail
in thumbnails
5377 def parse_http_range(range):
5378 """ Parse value of "Range" or "Content-Range" HTTP header into tuple. """
5380 return None, None, None
5381 crg
= re
.search(r
'bytes[ =](\d+)-(\d+)?(?:/(\d+))?', range)
5383 return None, None, None
5384 return int(crg
.group(1)), int_or_none(crg
.group(2)), int_or_none(crg
.group(3))
5387 def read_stdin(what
):
5388 eof
= 'Ctrl+Z' if compat_os_name
== 'nt' else 'Ctrl+D'
5389 write_string(f
'Reading {what} from STDIN - EOF ({eof}) to end:\n')
5393 def determine_file_encoding(data
):
5395 Detect the text encoding used
5396 @returns (encoding, bytes to skip)
5399 # BOM marks are given priority over declarations
5400 for bom
, enc
in BOMS
:
5401 if data
.startswith(bom
):
5402 return enc
, len(bom
)
5404 # Strip off all null bytes to match even when UTF-16 or UTF-32 is used.
5405 # We ignore the endianness to get a good enough match
5406 data
= data
.replace(b
'\0', b
'')
5407 mobj
= re
.match(rb
'(?m)^#\s*coding\s*:\s*(\S+)\s*$', data
)
5408 return mobj
.group(1).decode() if mobj
else None, 0
5415 __initialized
= False
5417 def __init__(self
, parser
, label
=None):
5418 self
.parser
, self
.label
= parser
, label
5419 self
._loaded
_paths
, self
.configs
= set(), []
5421 def init(self
, args
=None, filename
=None):
5422 assert not self
.__initialized
5423 self
.own_args
, self
.filename
= args
, filename
5424 return self
.load_configs()
5426 def load_configs(self
):
5429 location
= os
.path
.realpath(self
.filename
)
5430 directory
= os
.path
.dirname(location
)
5431 if location
in self
._loaded
_paths
:
5433 self
._loaded
_paths
.add(location
)
5435 self
.__initialized
= True
5436 opts
, _
= self
.parser
.parse_known_args(self
.own_args
)
5437 self
.parsed_args
= self
.own_args
5438 for location
in opts
.config_locations
or []:
5440 self
.append_config(shlex
.split(read_stdin('options'), comments
=True), label
='stdin')
5442 location
= os
.path
.join(directory
, expand_path(location
))
5443 if os
.path
.isdir(location
):
5444 location
= os
.path
.join(location
, 'yt-dlp.conf')
5445 if not os
.path
.exists(location
):
5446 self
.parser
.error(f
'config location {location} does not exist')
5447 self
.append_config(self
.read_file(location
), location
)
5451 label
= join_nonempty(
5452 self
.label
, 'config', f
'"{self.filename}"' if self
.filename
else '',
5454 return join_nonempty(
5455 self
.own_args
is not None and f
'{label[0].upper()}{label[1:]}: {self.hide_login_info(self.own_args)}',
5456 *(f
'\n{c}'.replace('\n', '\n| ')[1:] for c
in self
.configs
),
5460 def read_file(filename
, default
=[]):
5462 optionf
= open(filename
, 'rb')
5464 return default
# silently skip if file is not present
5466 enc
, skip
= determine_file_encoding(optionf
.read(512))
5467 optionf
.seek(skip
, io
.SEEK_SET
)
5469 enc
= None # silently skip read errors
5471 # FIXME: https://github.com/ytdl-org/youtube-dl/commit/dfe5fa49aed02cf36ba9f743b11b0903554b5e56
5472 contents
= optionf
.read().decode(enc
or preferredencoding())
5473 res
= shlex
.split(contents
, comments
=True)
5474 except Exception as err
:
5475 raise ValueError(f
'Unable to parse "{filename}": {err}')
5481 def hide_login_info(opts
):
5482 PRIVATE_OPTS
= {'-p', '--password', '-u', '--username', '--video-password', '--ap-password', '--ap-username'}
5483 eqre
= re
.compile('^(?P<key>' + ('|'.join(re
.escape(po
) for po
in PRIVATE_OPTS
)) + ')=.+$')
5488 return m
.group('key') + '=PRIVATE'
5492 opts
= list(map(_scrub_eq
, opts
))
5493 for idx
, opt
in enumerate(opts
):
5494 if opt
in PRIVATE_OPTS
and idx
+ 1 < len(opts
):
5495 opts
[idx
+ 1] = 'PRIVATE'
5498 def append_config(self
, *args
, label
=None):
5499 config
= type(self
)(self
.parser
, label
)
5500 config
._loaded
_paths
= self
._loaded
_paths
5501 if config
.init(*args
):
5502 self
.configs
.append(config
)
5506 for config
in reversed(self
.configs
):
5507 yield from config
.all_args
5508 yield from self
.parsed_args
or []
5510 def parse_known_args(self
, **kwargs
):
5511 return self
.parser
.parse_known_args(self
.all_args
, **kwargs
)
5513 def parse_args(self
):
5514 return self
.parser
.parse_args(self
.all_args
)
5517 class WebSocketsWrapper():
5518 """Wraps websockets module to use in non-async scopes"""
5521 def __init__(self
, url
, headers
=None, connect
=True):
5522 self
.loop
= asyncio
.new_event_loop()
5523 # XXX: "loop" is deprecated
5524 self
.conn
= websockets
.connect(
5525 url
, extra_headers
=headers
, ping_interval
=None,
5526 close_timeout
=float('inf'), loop
=self
.loop
, ping_timeout
=float('inf'))
5529 atexit
.register(self
.__exit
__, None, None, None)
5531 def __enter__(self
):
5533 self
.pool
= self
.run_with_loop(self
.conn
.__aenter
__(), self
.loop
)
5536 def send(self
, *args
):
5537 self
.run_with_loop(self
.pool
.send(*args
), self
.loop
)
5539 def recv(self
, *args
):
5540 return self
.run_with_loop(self
.pool
.recv(*args
), self
.loop
)
5542 def __exit__(self
, type, value
, traceback
):
5544 return self
.run_with_loop(self
.conn
.__aexit
__(type, value
, traceback
), self
.loop
)
5547 self
._cancel
_all
_tasks
(self
.loop
)
5549 # taken from https://github.com/python/cpython/blob/3.9/Lib/asyncio/runners.py with modifications
5550 # for contributors: If there's any new library using asyncio needs to be run in non-async, move these function out of this class
5552 def run_with_loop(main
, loop
):
5553 if not asyncio
.iscoroutine(main
):
5554 raise ValueError(f
'a coroutine was expected, got {main!r}')
5557 return loop
.run_until_complete(main
)
5559 loop
.run_until_complete(loop
.shutdown_asyncgens())
5560 if hasattr(loop
, 'shutdown_default_executor'):
5561 loop
.run_until_complete(loop
.shutdown_default_executor())
5564 def _cancel_all_tasks(loop
):
5565 to_cancel
= asyncio
.all_tasks(loop
)
5570 for task
in to_cancel
:
5573 # XXX: "loop" is removed in python 3.10+
5574 loop
.run_until_complete(
5575 asyncio
.gather(*to_cancel
, loop
=loop
, return_exceptions
=True))
5577 for task
in to_cancel
:
5578 if task
.cancelled():
5580 if task
.exception() is not None:
5581 loop
.call_exception_handler({
5582 'message': 'unhandled exception during asyncio.run() shutdown',
5583 'exception': task
.exception(),
5588 def merge_headers(*dicts
):
5589 """Merge dicts of http headers case insensitively, prioritizing the latter ones"""
5590 return {k.title(): v for k, v in itertools.chain.from_iterable(map(dict.items, dicts))}
5593 def cached_method(f
):
5594 """Cache a method"""
5595 signature
= inspect
.signature(f
)
5598 def wrapper(self
, *args
, **kwargs
):
5599 bound_args
= signature
.bind(self
, *args
, **kwargs
)
5600 bound_args
.apply_defaults()
5601 key
= tuple(bound_args
.arguments
.values())
5603 if not hasattr(self
, '__cached_method__cache'):
5604 self
.__cached
_method
__cache
= {}
5605 cache
= self
.__cached
_method
__cache
.setdefault(f
.__name
__, {})
5606 if key
not in cache
:
5607 cache
[key
] = f(self
, *args
, **kwargs
)
5612 class classproperty
:
5613 """property access for class methods"""
5615 def __init__(self
, func
):
5616 functools
.update_wrapper(self
, func
)
5619 def __get__(self
, _
, cls
):
5620 return self
.func(cls
)
5623 class Namespace(types
.SimpleNamespace
):
5624 """Immutable namespace"""
5627 return iter(self
.__dict
__.values())
5631 return self
.__dict
__.items()
5634 MEDIA_EXTENSIONS
= Namespace(
5635 common_video
=('avi', 'flv', 'mkv', 'mov', 'mp4', 'webm'),
5636 video
=('3g2', '3gp', 'f4v', 'mk3d', 'divx', 'mpg', 'ogv', 'm4v', 'wmv'),
5637 common_audio
=('aiff', 'alac', 'flac', 'm4a', 'mka', 'mp3', 'ogg', 'opus', 'wav'),
5638 audio
=('aac', 'ape', 'asf', 'f4a', 'f4b', 'm4b', 'm4p', 'm4r', 'oga', 'ogx', 'spx', 'vorbis', 'wma'),
5639 thumbnails
=('jpg', 'png', 'webp'),
5640 storyboards
=('mhtml', ),
5641 subtitles
=('srt', 'vtt', 'ass', 'lrc'),
5642 manifests
=('f4f', 'f4m', 'm3u8', 'smil', 'mpd'),
5644 MEDIA_EXTENSIONS
.video
+= MEDIA_EXTENSIONS
.common_video
5645 MEDIA_EXTENSIONS
.audio
+= MEDIA_EXTENSIONS
.common_audio
5647 KNOWN_EXTENSIONS
= (*MEDIA_EXTENSIONS
.video
, *MEDIA_EXTENSIONS
.audio
, *MEDIA_EXTENSIONS
.manifests
)
5651 has_certifi
= bool(certifi
)
5652 has_websockets
= bool(websockets
)