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 # From https://github.com/python/cpython/blob/3.11/Lib/email/_parseaddr.py#L36-L42
155 'UT': 0, 'UTC': 0, 'GMT': 0, 'Z': 0,
156 'AST': -4, 'ADT': -3, # Atlantic (used in Canada)
157 'EST': -5, 'EDT': -4, # Eastern
158 'CST': -6, 'CDT': -5, # Central
159 'MST': -7, 'MDT': -6, # Mountain
160 'PST': -8, 'PDT': -7 # Pacific
163 # needed for sanitizing filenames in restricted mode
164 ACCENT_CHARS
= dict(zip('ÂÃÄÀÁÅÆÇÈÉÊËÌÍÎÏÐÑÒÓÔÕÖŐØŒÙÚÛÜŰÝÞßàáâãäåæçèéêëìíîïðñòóôõöőøœùúûüűýþÿ',
165 itertools
.chain('AAAAAA', ['AE'], 'CEEEEIIIIDNOOOOOOO', ['OE'], 'UUUUUY', ['TH', 'ss'],
166 'aaaaaa', ['ae'], 'ceeeeiiiionooooooo', ['oe'], 'uuuuuy', ['th'], 'y')))
196 '%Y-%m-%d %H:%M:%S.%f',
197 '%Y-%m-%d %H:%M:%S:%f',
200 '%Y-%m-%dT%H:%M:%SZ',
201 '%Y-%m-%dT%H:%M:%S.%fZ',
202 '%Y-%m-%dT%H:%M:%S.%f0Z',
204 '%Y-%m-%dT%H:%M:%S.%f',
207 '%b %d %Y at %H:%M:%S',
209 '%B %d %Y at %H:%M:%S',
213 DATE_FORMATS_DAY_FIRST
= list(DATE_FORMATS
)
214 DATE_FORMATS_DAY_FIRST
.extend([
224 DATE_FORMATS_MONTH_FIRST
= list(DATE_FORMATS
)
225 DATE_FORMATS_MONTH_FIRST
.extend([
233 PACKED_CODES_RE
= r
"}\('(.+)',(\d+),(\d+),'([^']+)'\.split\('\|'\)"
234 JSON_LD_RE
= r
'(?is)<script[^>]+type=(["\']?
)application
/ld\
+json\
1[^
>]*>\s
*(?P
<json_ld
>{.+?}
)\s
*</script
>'
236 NUMBER_RE = r'\d
+(?
:\
.\d
+)?
'
240 def preferredencoding():
241 """Get preferred encoding.
243 Returns the best encoding scheme for the system, based on
244 locale.getpreferredencoding() and some further tweaks.
247 pref = locale.getpreferredencoding()
255 def write_json_file(obj, fn):
256 """ Encode obj as JSON and write it to fn, atomically if possible """
258 tf = tempfile.NamedTemporaryFile(
259 prefix=f'{os.path.basename(fn)}
.', dir=os.path.dirname(fn),
260 suffix='.tmp
', delete=False, mode='w
', encoding='utf
-8')
264 json.dump(obj, tf, ensure_ascii=False)
265 if sys.platform == 'win32
':
266 # Need to remove existing file on Windows, else os.rename raises
267 # WindowsError or FileExistsError.
268 with contextlib.suppress(OSError):
270 with contextlib.suppress(OSError):
273 os.chmod(tf.name, 0o666 & ~mask)
274 os.rename(tf.name, fn)
276 with contextlib.suppress(OSError):
281 def find_xpath_attr(node, xpath, key, val=None):
282 """ Find the xpath xpath[@key=val] """
283 assert re.match(r'^
[a
-zA
-Z_
-]+$
', key)
284 expr = xpath + ('[@%s]' % key if val is None else f"[@{key}='{val}
']")
285 return node.find(expr)
287 # On python2.6 the xml.etree.ElementTree.Element methods don't support
288 # the namespace parameter
291 def xpath_with_ns(path
, ns_map
):
292 components
= [c
.split(':') for c
in path
.split('/')]
296 replaced
.append(c
[0])
299 replaced
.append('{%s}%s' % (ns_map
[ns
], tag
))
300 return '/'.join(replaced
)
303 def xpath_element(node
, xpath
, name
=None, fatal
=False, default
=NO_DEFAULT
):
304 def _find_xpath(xpath
):
305 return node
.find(xpath
)
307 if isinstance(xpath
, str):
308 n
= _find_xpath(xpath
)
316 if default
is not NO_DEFAULT
:
319 name
= xpath
if name
is None else name
320 raise ExtractorError('Could not find XML element %s' % name
)
326 def xpath_text(node
, xpath
, name
=None, fatal
=False, default
=NO_DEFAULT
):
327 n
= xpath_element(node
, xpath
, name
, fatal
=fatal
, default
=default
)
328 if n
is None or n
== default
:
331 if default
is not NO_DEFAULT
:
334 name
= xpath
if name
is None else name
335 raise ExtractorError('Could not find XML element\'s text %s' % name
)
341 def xpath_attr(node
, xpath
, key
, name
=None, fatal
=False, default
=NO_DEFAULT
):
342 n
= find_xpath_attr(node
, xpath
, key
)
344 if default
is not NO_DEFAULT
:
347 name
= f
'{xpath}[@{key}]' if name
is None else name
348 raise ExtractorError('Could not find XML attribute %s' % name
)
354 def get_element_by_id(id, html
, **kwargs
):
355 """Return the content of the tag with the specified ID in the passed HTML document"""
356 return get_element_by_attribute('id', id, html
, **kwargs
)
359 def get_element_html_by_id(id, html
, **kwargs
):
360 """Return the html of the tag with the specified ID in the passed HTML document"""
361 return get_element_html_by_attribute('id', id, html
, **kwargs
)
364 def get_element_by_class(class_name
, html
):
365 """Return the content of the first tag with the specified class in the passed HTML document"""
366 retval
= get_elements_by_class(class_name
, html
)
367 return retval
[0] if retval
else None
370 def get_element_html_by_class(class_name
, html
):
371 """Return the html of the first tag with the specified class in the passed HTML document"""
372 retval
= get_elements_html_by_class(class_name
, html
)
373 return retval
[0] if retval
else None
376 def get_element_by_attribute(attribute
, value
, html
, **kwargs
):
377 retval
= get_elements_by_attribute(attribute
, value
, html
, **kwargs
)
378 return retval
[0] if retval
else None
381 def get_element_html_by_attribute(attribute
, value
, html
, **kargs
):
382 retval
= get_elements_html_by_attribute(attribute
, value
, html
, **kargs
)
383 return retval
[0] if retval
else None
386 def get_elements_by_class(class_name
, html
, **kargs
):
387 """Return the content of all tags with the specified class in the passed HTML document as a list"""
388 return get_elements_by_attribute(
389 'class', r
'[^\'"]*(?<=[\'"\s
])%s(?
=[\'"\s])[^\'"]*' % re.escape(class_name),
390 html, escape_value=False)
393 def get_elements_html_by_class(class_name, html):
394 """Return the html of all tags with the specified class in the passed HTML document as a list"""
395 return get_elements_html_by_attribute(
396 'class', r'[^
\'"]*(?<=[\'"\s
])%s(?
=[\'"\s])[^\'"]*' % re.escape(class_name),
397 html, escape_value=False)
400 def get_elements_by_attribute(*args, **kwargs):
401 """Return the content of the tag with the specified attribute in the passed HTML document"""
402 return [content for content, _ in get_elements_text_and_html_by_attribute(*args, **kwargs)]
405 def get_elements_html_by_attribute(*args, **kwargs):
406 """Return the html of the tag with the specified attribute in the passed HTML document"""
407 return [whole for _, whole in get_elements_text_and_html_by_attribute(*args, **kwargs)]
410 def get_elements_text_and_html_by_attribute(attribute, value, html, escape_value=True):
412 Return the text (content) and the html (whole) of the tag with the specified
413 attribute in the passed HTML document
416 quote = '' if re.match(r'''[\s"'`
=<>]''', value) else '?'
418 value = re.escape(value) if escape_value else value
420 partial_element_re = rf'''(?x
)
421 <(?P
<tag
>[a
-zA
-Z0
-9:._-]+)
422 (?
:\
s(?
:[^
>"']|"[^
"]*"|
'[^']*')*)?
423 \s{re.escape(attribute)}\s*=\s*(?P<_q>['"]{quote})(?-x:{value})(?P=_q)
426 for m in re.finditer(partial_element_re, html):
427 content, whole = get_element_text_and_html_by_tag(m.group('tag'), html[m.start():])
430 unescapeHTML(re.sub(r'^(?P<q>["\'])(?P
<content
>.*)(?P
=q
)$
', r'\g
<content
>', content, flags=re.DOTALL)),
435 class HTMLBreakOnClosingTagParser(html.parser.HTMLParser):
437 HTML parser which raises HTMLBreakOnClosingTagException upon reaching the
438 closing tag for the first opening tag it has encountered, and can be used
442 class HTMLBreakOnClosingTagException(Exception):
446 self.tagstack = collections.deque()
447 html.parser.HTMLParser.__init__(self)
452 def __exit__(self, *_):
456 # handle_endtag does not return upon raising HTMLBreakOnClosingTagException,
457 # so data remains buffered; we no longer have any interest in it, thus
458 # override this method to discard it
461 def handle_starttag(self, tag, _):
462 self.tagstack.append(tag)
464 def handle_endtag(self, tag):
465 if not self.tagstack:
466 raise compat_HTMLParseError('no tags
in the stack
')
468 inner_tag = self.tagstack.pop()
472 raise compat_HTMLParseError(f'matching opening tag
for closing {tag} tag
not found
')
473 if not self.tagstack:
474 raise self.HTMLBreakOnClosingTagException()
477 def get_element_text_and_html_by_tag(tag, html):
479 For the first element with the specified tag in the passed HTML document
480 return its' content (text
) and the whole
element (html
)
482 def find_or_raise(haystack, needle, exc):
484 return haystack.index(needle)
487 closing_tag = f'</{tag}>'
488 whole_start = find_or_raise(
489 html, f'<{tag}', compat_HTMLParseError(f'opening {tag} tag not found'))
490 content_start = find_or_raise(
491 html[whole_start:], '>', compat_HTMLParseError(f'malformed opening {tag} tag'))
492 content_start += whole_start + 1
493 with HTMLBreakOnClosingTagParser() as parser:
494 parser.feed(html[whole_start:content_start])
495 if not parser.tagstack or parser.tagstack[0] != tag:
496 raise compat_HTMLParseError(f'parser did not match opening {tag} tag')
497 offset = content_start
498 while offset < len(html):
499 next_closing_tag_start = find_or_raise(
500 html[offset:], closing_tag,
501 compat_HTMLParseError(f'closing {tag} tag not found'))
502 next_closing_tag_end = next_closing_tag_start + len(closing_tag)
504 parser.feed(html[offset:offset + next_closing_tag_end])
505 offset += next_closing_tag_end
506 except HTMLBreakOnClosingTagParser.HTMLBreakOnClosingTagException:
507 return html[content_start:offset + next_closing_tag_start], \
508 html[whole_start:offset + next_closing_tag_end]
509 raise compat_HTMLParseError('unexpected end of html')
512 class HTMLAttributeParser(html.parser.HTMLParser):
513 """Trivial HTML parser to gather the attributes
for a single element
"""
517 html.parser.HTMLParser.__init__(self)
519 def handle_starttag(self, tag, attrs):
520 self.attrs = dict(attrs)
523 class HTMLListAttrsParser(html.parser.HTMLParser):
524 """HTML parser to gather the attributes
for the elements of a
list"""
527 html.parser.HTMLParser.__init__(self)
531 def handle_starttag(self, tag, attrs):
532 if tag == 'li' and self._level == 0:
533 self.items.append(dict(attrs))
536 def handle_endtag(self, tag):
540 def extract_attributes(html_element):
541 """Given a string
for an HTML element such
as
543 a
="foo" B
="bar" c
="&98;az" d
=boz
544 empty
= noval entity
="&"
547 Decode
and return a dictionary of attributes
.
549 'a': 'foo', 'b': 'bar', c
: 'baz', d
: 'boz',
550 'empty': '', 'noval': None, 'entity': '&',
551 'sq': '"', 'dq': '\''
554 parser = HTMLAttributeParser()
555 with contextlib.suppress(compat_HTMLParseError):
556 parser.feed(html_element)
561 def parse_list(webpage):
562 """Given a string
for an series of HTML
<li
> elements
,
563 return a dictionary of their attributes
"""
564 parser = HTMLListAttrsParser()
570 def clean_html(html):
571 """Clean an HTML snippet into a readable string
"""
573 if html is None: # Convenience for sanitizing descriptions etc.
576 html = re.sub(r'\s+', ' ', html)
577 html = re.sub(r'(?u)\s?<\s?br\s?/?\s?>\s?', '\n', html)
578 html = re.sub(r'(?u)<\s?/\s?p\s?>\s?<\s?p[^>]*>', '\n', html)
580 html = re.sub('<.*?>', '', html)
581 # Replace html entities
582 html = unescapeHTML(html)
586 class LenientJSONDecoder(json.JSONDecoder):
587 def __init__(self, *args, transform_source=None, ignore_extra=False, **kwargs):
588 self.transform_source, self.ignore_extra = transform_source, ignore_extra
589 super().__init__(*args, **kwargs)
592 if self.transform_source:
593 s = self.transform_source(s)
594 if self.ignore_extra:
595 return self.raw_decode(s.lstrip())[0]
596 return super().decode(s)
599 def sanitize_open(filename, open_mode):
600 """Try to
open the given filename
, and slightly tweak it
if this fails
.
602 Attempts to
open the given filename
. If this fails
, it tries to change
603 the filename slightly
, step by step
, until it
's either able to open it
604 or it fails and raises a final exception, like the standard open()
607 It returns the tuple (stream, definitive_file_name).
610 if sys.platform == 'win32
':
613 # stdout may be any IO stream, e.g. when using contextlib.redirect_stdout
614 with contextlib.suppress(io.UnsupportedOperation):
615 msvcrt.setmode(sys.stdout.fileno(), os.O_BINARY)
616 return (sys.stdout.buffer if hasattr(sys.stdout, 'buffer') else sys.stdout, filename)
618 for attempt in range(2):
621 if sys.platform == 'win32
':
622 # FIXME: An exclusive lock also locks the file from being read.
623 # Since windows locks are mandatory, don't lock the
file on
windows (for now
).
624 # Ref: https://github.com/yt-dlp/yt-dlp/issues/3124
625 raise LockingUnsupportedError()
626 stream
= locked_file(filename
, open_mode
, block
=False).__enter
__()
628 stream
= open(filename
, open_mode
)
629 return stream
, filename
630 except OSError as err
:
631 if attempt
or err
.errno
in (errno
.EACCES
,):
633 old_filename
, filename
= filename
, sanitize_path(filename
)
634 if old_filename
== filename
:
638 def timeconvert(timestr
):
639 """Convert RFC 2822 defined time string into system timestamp"""
641 timetuple
= email
.utils
.parsedate_tz(timestr
)
642 if timetuple
is not None:
643 timestamp
= email
.utils
.mktime_tz(timetuple
)
647 def sanitize_filename(s
, restricted
=False, is_id
=NO_DEFAULT
):
648 """Sanitizes a string so it could be used as part of a filename.
649 @param restricted Use a stricter subset of allowed characters
650 @param is_id Whether this is an ID that should be kept unchanged if possible.
651 If unset, yt-dlp's new sanitization rules are in effect
656 def replace_insane(char
):
657 if restricted
and char
in ACCENT_CHARS
:
658 return ACCENT_CHARS
[char
]
659 elif not restricted
and char
== '\n':
661 elif is_id
is NO_DEFAULT
and not restricted
and char
in '"*:<>?|/\\':
662 # Replace with their full-width unicode counterparts
663 return {'/': '\u29F8', '\\': '\u29f9'}
.get(char
, chr(ord(char
) + 0xfee0))
664 elif char
== '?' or ord(char
) < 32 or ord(char
) == 127:
667 return '' if restricted
else '\''
669 return '\0_\0-' if restricted
else '\0 \0-'
670 elif char
in '\\/|*<>':
672 if restricted
and (char
in '!&\'()[]{}$;`^,#' or char
.isspace() or ord(char
) > 127):
676 if restricted
and is_id
is NO_DEFAULT
:
677 s
= unicodedata
.normalize('NFKC', s
)
678 s
= re
.sub(r
'[0-9]+(?::[0-9]+)+', lambda m
: m
.group(0).replace(':', '_'), s
) # Handle timestamps
679 result
= ''.join(map(replace_insane
, s
))
680 if is_id
is NO_DEFAULT
:
681 result
= re
.sub(r
'(\0.)(?:(?=\1)..)+', r
'\1', result
) # Remove repeated substitute chars
682 STRIP_RE
= r
'(?:\0.|[ _-])*'
683 result
= re
.sub(f
'^\0.{STRIP_RE}|{STRIP_RE}\0.$', '', result
) # Remove substitute chars from start/end
684 result
= result
.replace('\0', '') or '_'
687 while '__' in result
:
688 result
= result
.replace('__', '_')
689 result
= result
.strip('_')
690 # Common case of "Foreign band name - English song title"
691 if restricted
and result
.startswith('-_'):
693 if result
.startswith('-'):
694 result
= '_' + result
[len('-'):]
695 result
= result
.lstrip('.')
701 def sanitize_path(s
, force
=False):
702 """Sanitizes and normalizes path on Windows"""
703 if sys
.platform
== 'win32':
705 drive_or_unc
, _
= os
.path
.splitdrive(s
)
711 norm_path
= os
.path
.normpath(remove_start(s
, drive_or_unc
)).split(os
.path
.sep
)
715 path_part
if path_part
in ['.', '..'] else re
.sub(r
'(?:[/<>:"\|\\?\*]|[\s.]$)', '#', path_part
)
716 for path_part
in norm_path
]
718 sanitized_path
.insert(0, drive_or_unc
+ os
.path
.sep
)
719 elif force
and s
and s
[0] == os
.path
.sep
:
720 sanitized_path
.insert(0, os
.path
.sep
)
721 return os
.path
.join(*sanitized_path
)
724 def sanitize_url(url
, *, scheme
='http'):
725 # Prepend protocol-less URLs with `http:` scheme in order to mitigate
726 # the number of unwanted failures due to missing protocol
729 elif url
.startswith('//'):
730 return f
'{scheme}:{url}'
731 # Fix some common typos seen so far
733 # https://github.com/ytdl-org/youtube-dl/issues/15649
734 (r
'^httpss://', r
'https://'),
735 # https://bx1.be/lives/direct-tv/
736 (r
'^rmtp([es]?)://', r
'rtmp\1://'),
738 for mistake
, fixup
in COMMON_TYPOS
:
739 if re
.match(mistake
, url
):
740 return re
.sub(mistake
, fixup
, url
)
744 def extract_basic_auth(url
):
745 parts
= urllib
.parse
.urlsplit(url
)
746 if parts
.username
is None:
748 url
= urllib
.parse
.urlunsplit(parts
._replace
(netloc
=(
749 parts
.hostname
if parts
.port
is None
750 else '%s:%d' % (parts
.hostname
, parts
.port
))))
751 auth_payload
= base64
.b64encode(
752 ('%s:%s' % (parts
.username
, parts
.password
or '')).encode())
753 return url
, f
'Basic {auth_payload.decode()}'
756 def sanitized_Request(url
, *args
, **kwargs
):
757 url
, auth_header
= extract_basic_auth(escape_url(sanitize_url(url
)))
758 if auth_header
is not None:
759 headers
= args
[1] if len(args
) >= 2 else kwargs
.setdefault('headers', {})
760 headers
['Authorization'] = auth_header
761 return urllib
.request
.Request(url
, *args
, **kwargs
)
765 """Expand shell variables and ~"""
766 return os
.path
.expandvars(compat_expanduser(s
))
769 def orderedSet(iterable
, *, lazy
=False):
770 """Remove all duplicates from the input iterable"""
772 seen
= [] # Do not use set since the items can be unhashable
778 return _iter() if lazy
else list(_iter())
781 def _htmlentity_transform(entity_with_semicolon
):
782 """Transforms an HTML entity to a character."""
783 entity
= entity_with_semicolon
[:-1]
785 # Known non-numeric HTML entity
786 if entity
in html
.entities
.name2codepoint
:
787 return chr(html
.entities
.name2codepoint
[entity
])
789 # TODO: HTML5 allows entities without a semicolon.
790 # E.g. 'Éric' should be decoded as 'Éric'.
791 if entity_with_semicolon
in html
.entities
.html5
:
792 return html
.entities
.html5
[entity_with_semicolon
]
794 mobj
= re
.match(r
'#(x[0-9a-fA-F]+|[0-9]+)', entity
)
796 numstr
= mobj
.group(1)
797 if numstr
.startswith('x'):
799 numstr
= '0%s' % numstr
802 # See https://github.com/ytdl-org/youtube-dl/issues/7518
803 with contextlib
.suppress(ValueError):
804 return chr(int(numstr
, base
))
806 # Unknown entity in name, return its literal representation
807 return '&%s;' % entity
813 assert isinstance(s
, str)
816 r
'&([^&;]+;)', lambda m
: _htmlentity_transform(m
.group(1)), s
)
819 def escapeHTML(text
):
822 .replace('&', '&')
823 .replace('<', '<')
824 .replace('>', '>')
825 .replace('"', '"')
826 .replace("'", ''')
830 def process_communicate_or_kill(p
, *args
, **kwargs
):
831 deprecation_warning(f
'"{__name__}.process_communicate_or_kill" is deprecated and may be removed '
832 f
'in a future version. Use "{__name__}.Popen.communicate_or_kill" instead')
833 return Popen
.communicate_or_kill(p
, *args
, **kwargs
)
836 class Popen(subprocess
.Popen
):
837 if sys
.platform
== 'win32':
838 _startupinfo
= subprocess
.STARTUPINFO()
839 _startupinfo
.dwFlags |
= subprocess
.STARTF_USESHOWWINDOW
844 def _fix_pyinstaller_ld_path(env
):
845 """Restore LD_LIBRARY_PATH when using PyInstaller
846 Ref: https://github.com/pyinstaller/pyinstaller/blob/develop/doc/runtime-information.rst#ld_library_path--libpath-considerations
847 https://github.com/yt-dlp/yt-dlp/issues/4573
849 if not hasattr(sys
, '_MEIPASS'):
853 orig
= env
.get(f
'{key}_ORIG')
859 _fix('LD_LIBRARY_PATH') # Linux
860 _fix('DYLD_LIBRARY_PATH') # macOS
862 def __init__(self
, *args
, env
=None, text
=False, **kwargs
):
864 env
= os
.environ
.copy()
865 self
._fix
_pyinstaller
_ld
_path
(env
)
868 kwargs
['universal_newlines'] = True # For 3.6 compatibility
869 kwargs
.setdefault('encoding', 'utf-8')
870 kwargs
.setdefault('errors', 'replace')
871 super().__init
__(*args
, env
=env
, **kwargs
, startupinfo
=self
._startupinfo
)
873 def communicate_or_kill(self
, *args
, **kwargs
):
875 return self
.communicate(*args
, **kwargs
)
876 except BaseException
: # Including KeyboardInterrupt
877 self
.kill(timeout
=None)
880 def kill(self
, *, timeout
=0):
883 self
.wait(timeout
=timeout
)
886 def run(cls
, *args
, timeout
=None, **kwargs
):
887 with cls(*args
, **kwargs
) as proc
:
888 stdout
, stderr
= proc
.communicate_or_kill(timeout
=timeout
)
889 return stdout
or '', stderr
or '', proc
.returncode
892 def get_subprocess_encoding():
893 if sys
.platform
== 'win32' and sys
.getwindowsversion()[0] >= 5:
894 # For subprocess calls, encode with locale encoding
895 # Refer to http://stackoverflow.com/a/9951851/35070
896 encoding
= preferredencoding()
898 encoding
= sys
.getfilesystemencoding()
904 def encodeFilename(s
, for_subprocess
=False):
905 assert isinstance(s
, str)
909 def decodeFilename(b
, for_subprocess
=False):
913 def encodeArgument(s
):
914 # Legacy code that uses byte strings
915 # Uncomment the following line after fixing all post processors
916 # assert isinstance(s, str), 'Internal error: %r should be of type %r, is %r' % (s, str, type(s))
917 return s
if isinstance(s
, str) else s
.decode('ascii')
920 def decodeArgument(b
):
924 def decodeOption(optval
):
927 if isinstance(optval
, bytes):
928 optval
= optval
.decode(preferredencoding())
930 assert isinstance(optval
, str)
934 _timetuple
= collections
.namedtuple('Time', ('hours', 'minutes', 'seconds', 'milliseconds'))
937 def timetuple_from_msec(msec
):
938 secs
, msec
= divmod(msec
, 1000)
939 mins
, secs
= divmod(secs
, 60)
940 hrs
, mins
= divmod(mins
, 60)
941 return _timetuple(hrs
, mins
, secs
, msec
)
944 def formatSeconds(secs
, delim
=':', msec
=False):
945 time
= timetuple_from_msec(secs
* 1000)
947 ret
= '%d%s%02d%s%02d' % (time
.hours
, delim
, time
.minutes
, delim
, time
.seconds
)
949 ret
= '%d%s%02d' % (time
.minutes
, delim
, time
.seconds
)
951 ret
= '%d' % time
.seconds
952 return '%s.%03d' % (ret
, time
.milliseconds
) if msec
else ret
955 def _ssl_load_windows_store_certs(ssl_context
, storename
):
956 # Code adapted from _load_windows_store_certs in https://github.com/python/cpython/blob/main/Lib/ssl.py
958 certs
= [cert
for cert
, encoding
, trust
in ssl
.enum_certificates(storename
)
959 if encoding
== 'x509_asn' and (
960 trust
is True or ssl
.Purpose
.SERVER_AUTH
.oid
in trust
)]
961 except PermissionError
:
964 with contextlib
.suppress(ssl
.SSLError
):
965 ssl_context
.load_verify_locations(cadata
=cert
)
968 def make_HTTPS_handler(params
, **kwargs
):
969 opts_check_certificate
= not params
.get('nocheckcertificate')
970 context
= ssl
.SSLContext(ssl
.PROTOCOL_TLS_CLIENT
)
971 context
.check_hostname
= opts_check_certificate
972 if params
.get('legacyserverconnect'):
973 context
.options |
= 4 # SSL_OP_LEGACY_SERVER_CONNECT
974 # Allow use of weaker ciphers in Python 3.10+. See https://bugs.python.org/issue43998
975 context
.set_ciphers('DEFAULT')
977 context
.verify_mode
= ssl
.CERT_REQUIRED
if opts_check_certificate
else ssl
.CERT_NONE
978 if opts_check_certificate
:
979 if has_certifi
and 'no-certifi' not in params
.get('compat_opts', []):
980 context
.load_verify_locations(cafile
=certifi
.where())
983 context
.load_default_certs()
984 # Work around the issue in load_default_certs when there are bad certificates. See:
985 # https://github.com/yt-dlp/yt-dlp/issues/1060,
986 # https://bugs.python.org/issue35665, https://bugs.python.org/issue45312
988 # enum_certificates is not present in mingw python. See https://github.com/yt-dlp/yt-dlp/issues/1151
989 if sys
.platform
== 'win32' and hasattr(ssl
, 'enum_certificates'):
990 for storename
in ('CA', 'ROOT'):
991 _ssl_load_windows_store_certs(context
, storename
)
992 context
.set_default_verify_paths()
994 client_certfile
= params
.get('client_certificate')
997 context
.load_cert_chain(
998 client_certfile
, keyfile
=params
.get('client_certificate_key'),
999 password
=params
.get('client_certificate_password'))
1000 except ssl
.SSLError
:
1001 raise YoutubeDLError('Unable to load client certificate')
1003 # Some servers may reject requests if ALPN extension is not sent. See:
1004 # https://github.com/python/cpython/issues/85140
1005 # https://github.com/yt-dlp/yt-dlp/issues/3878
1006 with contextlib
.suppress(NotImplementedError):
1007 context
.set_alpn_protocols(['http/1.1'])
1009 return YoutubeDLHTTPSHandler(params
, context
=context
, **kwargs
)
1012 def bug_reports_message(before
=';'):
1013 from .update
import REPOSITORY
1015 msg
= (f
'please report this issue on https://github.com/{REPOSITORY}/issues?q= , '
1016 'filling out the appropriate issue template. Confirm you are on the latest version using yt-dlp -U')
1018 before
= before
.rstrip()
1019 if not before
or before
.endswith(('.', '!', '?')):
1020 msg
= msg
[0].title() + msg
[1:]
1022 return (before
+ ' ' if before
else '') + msg
1025 class YoutubeDLError(Exception):
1026 """Base exception for YoutubeDL errors."""
1029 def __init__(self
, msg
=None):
1032 elif self
.msg
is None:
1033 self
.msg
= type(self
).__name
__
1034 super().__init
__(self
.msg
)
1037 network_exceptions
= [urllib
.error
.URLError
, http
.client
.HTTPException
, socket
.error
]
1038 if hasattr(ssl
, 'CertificateError'):
1039 network_exceptions
.append(ssl
.CertificateError
)
1040 network_exceptions
= tuple(network_exceptions
)
1043 class ExtractorError(YoutubeDLError
):
1044 """Error during info extraction."""
1046 def __init__(self
, msg
, tb
=None, expected
=False, cause
=None, video_id
=None, ie
=None):
1047 """ tb, if given, is the original traceback (so that it can be printed out).
1048 If expected is set, this is a normal error message and most likely not a bug in yt-dlp.
1050 if sys
.exc_info()[0] in network_exceptions
:
1053 self
.orig_msg
= str(msg
)
1055 self
.expected
= expected
1057 self
.video_id
= video_id
1059 self
.exc_info
= sys
.exc_info() # preserve original exception
1060 if isinstance(self
.exc_info
[1], ExtractorError
):
1061 self
.exc_info
= self
.exc_info
[1].exc_info
1063 super().__init
__(''.join((
1064 format_field(ie
, None, '[%s] '),
1065 format_field(video_id
, None, '%s: '),
1067 format_field(cause
, None, ' (caused by %r)'),
1068 '' if expected
else bug_reports_message())))
1070 def format_traceback(self
):
1071 return join_nonempty(
1072 self
.traceback
and ''.join(traceback
.format_tb(self
.traceback
)),
1073 self
.cause
and ''.join(traceback
.format_exception(None, self
.cause
, self
.cause
.__traceback
__)[1:]),
1077 class UnsupportedError(ExtractorError
):
1078 def __init__(self
, url
):
1080 'Unsupported URL: %s' % url
, expected
=True)
1084 class RegexNotFoundError(ExtractorError
):
1085 """Error when a regex didn't match"""
1089 class GeoRestrictedError(ExtractorError
):
1090 """Geographic restriction Error exception.
1092 This exception may be thrown when a video is not available from your
1093 geographic location due to geographic restrictions imposed by a website.
1096 def __init__(self
, msg
, countries
=None, **kwargs
):
1097 kwargs
['expected'] = True
1098 super().__init
__(msg
, **kwargs
)
1099 self
.countries
= countries
1102 class UserNotLive(ExtractorError
):
1103 """Error when a channel/user is not live"""
1105 def __init__(self
, msg
=None, **kwargs
):
1106 kwargs
['expected'] = True
1107 super().__init
__(msg
or 'The channel is not currently live', **kwargs
)
1110 class DownloadError(YoutubeDLError
):
1111 """Download Error exception.
1113 This exception may be thrown by FileDownloader objects if they are not
1114 configured to continue on errors. They will contain the appropriate
1118 def __init__(self
, msg
, exc_info
=None):
1119 """ exc_info, if given, is the original exception that caused the trouble (as returned by sys.exc_info()). """
1120 super().__init
__(msg
)
1121 self
.exc_info
= exc_info
1124 class EntryNotInPlaylist(YoutubeDLError
):
1125 """Entry not in playlist exception.
1127 This exception will be thrown by YoutubeDL when a requested entry
1128 is not found in the playlist info_dict
1130 msg
= 'Entry not found in info'
1133 class SameFileError(YoutubeDLError
):
1134 """Same File exception.
1136 This exception will be thrown by FileDownloader objects if they detect
1137 multiple files would have to be downloaded to the same file on disk.
1139 msg
= 'Fixed output name but more than one file to download'
1141 def __init__(self
, filename
=None):
1142 if filename
is not None:
1143 self
.msg
+= f
': {filename}'
1144 super().__init
__(self
.msg
)
1147 class PostProcessingError(YoutubeDLError
):
1148 """Post Processing exception.
1150 This exception may be raised by PostProcessor's .run() method to
1151 indicate an error in the postprocessing task.
1155 class DownloadCancelled(YoutubeDLError
):
1156 """ Exception raised when the download queue should be interrupted """
1157 msg
= 'The download was cancelled'
1160 class ExistingVideoReached(DownloadCancelled
):
1161 """ --break-on-existing triggered """
1162 msg
= 'Encountered a video that is already in the archive, stopping due to --break-on-existing'
1165 class RejectedVideoReached(DownloadCancelled
):
1166 """ --break-on-reject triggered """
1167 msg
= 'Encountered a video that did not match filter, stopping due to --break-on-reject'
1170 class MaxDownloadsReached(DownloadCancelled
):
1171 """ --max-downloads limit has been reached. """
1172 msg
= 'Maximum number of downloads reached, stopping due to --max-downloads'
1175 class ReExtractInfo(YoutubeDLError
):
1176 """ Video info needs to be re-extracted. """
1178 def __init__(self
, msg
, expected
=False):
1179 super().__init
__(msg
)
1180 self
.expected
= expected
1183 class ThrottledDownload(ReExtractInfo
):
1184 """ Download speed below --throttled-rate. """
1185 msg
= 'The download speed is below throttle limit'
1188 super().__init
__(self
.msg
, expected
=False)
1191 class UnavailableVideoError(YoutubeDLError
):
1192 """Unavailable Format exception.
1194 This exception will be thrown when a video is requested
1195 in a format that is not available for that video.
1197 msg
= 'Unable to download video'
1199 def __init__(self
, err
=None):
1201 self
.msg
+= f
': {err}'
1202 super().__init
__(self
.msg
)
1205 class ContentTooShortError(YoutubeDLError
):
1206 """Content Too Short exception.
1208 This exception may be raised by FileDownloader objects when a file they
1209 download is too small for what the server announced first, indicating
1210 the connection was probably interrupted.
1213 def __init__(self
, downloaded
, expected
):
1214 super().__init
__(f
'Downloaded {downloaded} bytes, expected {expected} bytes')
1216 self
.downloaded
= downloaded
1217 self
.expected
= expected
1220 class XAttrMetadataError(YoutubeDLError
):
1221 def __init__(self
, code
=None, msg
='Unknown error'):
1222 super().__init
__(msg
)
1226 # Parsing code and msg
1227 if (self
.code
in (errno
.ENOSPC
, errno
.EDQUOT
)
1228 or 'No space left' in self
.msg
or 'Disk quota exceeded' in self
.msg
):
1229 self
.reason
= 'NO_SPACE'
1230 elif self
.code
== errno
.E2BIG
or 'Argument list too long' in self
.msg
:
1231 self
.reason
= 'VALUE_TOO_LONG'
1233 self
.reason
= 'NOT_SUPPORTED'
1236 class XAttrUnavailableError(YoutubeDLError
):
1240 def _create_http_connection(ydl_handler
, http_class
, is_https
, *args
, **kwargs
):
1241 hc
= http_class(*args
, **kwargs
)
1242 source_address
= ydl_handler
._params
.get('source_address')
1244 if source_address
is not None:
1245 # This is to workaround _create_connection() from socket where it will try all
1246 # address data from getaddrinfo() including IPv6. This filters the result from
1247 # getaddrinfo() based on the source_address value.
1248 # This is based on the cpython socket.create_connection() function.
1249 # https://github.com/python/cpython/blob/master/Lib/socket.py#L691
1250 def _create_connection(address
, timeout
=socket
._GLOBAL
_DEFAULT
_TIMEOUT
, source_address
=None):
1251 host
, port
= address
1253 addrs
= socket
.getaddrinfo(host
, port
, 0, socket
.SOCK_STREAM
)
1254 af
= socket
.AF_INET
if '.' in source_address
[0] else socket
.AF_INET6
1255 ip_addrs
= [addr
for addr
in addrs
if addr
[0] == af
]
1256 if addrs
and not ip_addrs
:
1257 ip_version
= 'v4' if af
== socket
.AF_INET
else 'v6'
1259 "No remote IP%s addresses available for connect, can't use '%s' as source address"
1260 % (ip_version
, source_address
[0]))
1261 for res
in ip_addrs
:
1262 af
, socktype
, proto
, canonname
, sa
= res
1265 sock
= socket
.socket(af
, socktype
, proto
)
1266 if timeout
is not socket
._GLOBAL
_DEFAULT
_TIMEOUT
:
1267 sock
.settimeout(timeout
)
1268 sock
.bind(source_address
)
1270 err
= None # Explicitly break reference cycle
1272 except OSError as _
:
1274 if sock
is not None:
1279 raise OSError('getaddrinfo returns an empty list')
1280 if hasattr(hc
, '_create_connection'):
1281 hc
._create
_connection
= _create_connection
1282 hc
.source_address
= (source_address
, 0)
1287 def handle_youtubedl_headers(headers
):
1288 filtered_headers
= headers
1290 if 'Youtubedl-no-compression' in filtered_headers
:
1291 filtered_headers
= {k: v for k, v in filtered_headers.items() if k.lower() != 'accept-encoding'}
1292 del filtered_headers
['Youtubedl-no-compression']
1294 return filtered_headers
1297 class YoutubeDLHandler(urllib
.request
.HTTPHandler
):
1298 """Handler for HTTP requests and responses.
1300 This class, when installed with an OpenerDirector, automatically adds
1301 the standard headers to every HTTP request and handles gzipped and
1302 deflated responses from web servers. If compression is to be avoided in
1303 a particular request, the original request in the program code only has
1304 to include the HTTP header "Youtubedl-no-compression", which will be
1305 removed before making the real request.
1307 Part of this code was copied from:
1309 http://techknack.net/python-urllib2-handlers/
1311 Andrew Rowls, the author of that code, agreed to release it to the
1315 def __init__(self
, params
, *args
, **kwargs
):
1316 urllib
.request
.HTTPHandler
.__init
__(self
, *args
, **kwargs
)
1317 self
._params
= params
1319 def http_open(self
, req
):
1320 conn_class
= http
.client
.HTTPConnection
1322 socks_proxy
= req
.headers
.get('Ytdl-socks-proxy')
1324 conn_class
= make_socks_conn_class(conn_class
, socks_proxy
)
1325 del req
.headers
['Ytdl-socks-proxy']
1327 return self
.do_open(functools
.partial(
1328 _create_http_connection
, self
, conn_class
, False),
1336 return zlib
.decompress(data
, -zlib
.MAX_WBITS
)
1338 return zlib
.decompress(data
)
1344 return brotli
.decompress(data
)
1346 def http_request(self
, req
):
1347 # According to RFC 3986, URLs can not contain non-ASCII characters, however this is not
1348 # always respected by websites, some tend to give out URLs with non percent-encoded
1349 # non-ASCII characters (see telemb.py, ard.py [#3412])
1350 # urllib chokes on URLs with non-ASCII characters (see http://bugs.python.org/issue3991)
1351 # To work around aforementioned issue we will replace request's original URL with
1352 # percent-encoded one
1353 # Since redirects are also affected (e.g. http://www.southpark.de/alle-episoden/s18e09)
1354 # the code of this workaround has been moved here from YoutubeDL.urlopen()
1355 url
= req
.get_full_url()
1356 url_escaped
= escape_url(url
)
1358 # Substitute URL if any change after escaping
1359 if url
!= url_escaped
:
1360 req
= update_Request(req
, url
=url_escaped
)
1362 for h
, v
in self
._params
.get('http_headers', std_headers
).items():
1363 # Capitalize is needed because of Python bug 2275: http://bugs.python.org/issue2275
1364 # The dict keys are capitalized because of this bug by urllib
1365 if h
.capitalize() not in req
.headers
:
1366 req
.add_header(h
, v
)
1368 if 'Accept-encoding' not in req
.headers
:
1369 req
.add_header('Accept-encoding', ', '.join(SUPPORTED_ENCODINGS
))
1371 req
.headers
= handle_youtubedl_headers(req
.headers
)
1373 return super().do_request_(req
)
1375 def http_response(self
, req
, resp
):
1378 if resp
.headers
.get('Content-encoding', '') == 'gzip':
1379 content
= resp
.read()
1380 gz
= gzip
.GzipFile(fileobj
=io
.BytesIO(content
), mode
='rb')
1382 uncompressed
= io
.BytesIO(gz
.read())
1383 except OSError as original_ioerror
:
1384 # There may be junk add the end of the file
1385 # See http://stackoverflow.com/q/4928560/35070 for details
1386 for i
in range(1, 1024):
1388 gz
= gzip
.GzipFile(fileobj
=io
.BytesIO(content
[:-i
]), mode
='rb')
1389 uncompressed
= io
.BytesIO(gz
.read())
1394 raise original_ioerror
1395 resp
= urllib
.request
.addinfourl(uncompressed
, old_resp
.headers
, old_resp
.url
, old_resp
.code
)
1396 resp
.msg
= old_resp
.msg
1397 del resp
.headers
['Content-encoding']
1399 if resp
.headers
.get('Content-encoding', '') == 'deflate':
1400 gz
= io
.BytesIO(self
.deflate(resp
.read()))
1401 resp
= urllib
.request
.addinfourl(gz
, old_resp
.headers
, old_resp
.url
, old_resp
.code
)
1402 resp
.msg
= old_resp
.msg
1403 del resp
.headers
['Content-encoding']
1405 if resp
.headers
.get('Content-encoding', '') == 'br':
1406 resp
= urllib
.request
.addinfourl(
1407 io
.BytesIO(self
.brotli(resp
.read())), old_resp
.headers
, old_resp
.url
, old_resp
.code
)
1408 resp
.msg
= old_resp
.msg
1409 del resp
.headers
['Content-encoding']
1410 # Percent-encode redirect URL of Location HTTP header to satisfy RFC 3986 (see
1411 # https://github.com/ytdl-org/youtube-dl/issues/6457).
1412 if 300 <= resp
.code
< 400:
1413 location
= resp
.headers
.get('Location')
1415 # As of RFC 2616 default charset is iso-8859-1 that is respected by python 3
1416 location
= location
.encode('iso-8859-1').decode()
1417 location_escaped
= escape_url(location
)
1418 if location
!= location_escaped
:
1419 del resp
.headers
['Location']
1420 resp
.headers
['Location'] = location_escaped
1423 https_request
= http_request
1424 https_response
= http_response
1427 def make_socks_conn_class(base_class
, socks_proxy
):
1428 assert issubclass(base_class
, (
1429 http
.client
.HTTPConnection
, http
.client
.HTTPSConnection
))
1431 url_components
= urllib
.parse
.urlparse(socks_proxy
)
1432 if url_components
.scheme
.lower() == 'socks5':
1433 socks_type
= ProxyType
.SOCKS5
1434 elif url_components
.scheme
.lower() in ('socks', 'socks4'):
1435 socks_type
= ProxyType
.SOCKS4
1436 elif url_components
.scheme
.lower() == 'socks4a':
1437 socks_type
= ProxyType
.SOCKS4A
1439 def unquote_if_non_empty(s
):
1442 return urllib
.parse
.unquote_plus(s
)
1446 url_components
.hostname
, url_components
.port
or 1080,
1448 unquote_if_non_empty(url_components
.username
),
1449 unquote_if_non_empty(url_components
.password
),
1452 class SocksConnection(base_class
):
1454 self
.sock
= sockssocket()
1455 self
.sock
.setproxy(*proxy_args
)
1456 if isinstance(self
.timeout
, (int, float)):
1457 self
.sock
.settimeout(self
.timeout
)
1458 self
.sock
.connect((self
.host
, self
.port
))
1460 if isinstance(self
, http
.client
.HTTPSConnection
):
1461 if hasattr(self
, '_context'): # Python > 2.6
1462 self
.sock
= self
._context
.wrap_socket(
1463 self
.sock
, server_hostname
=self
.host
)
1465 self
.sock
= ssl
.wrap_socket(self
.sock
)
1467 return SocksConnection
1470 class YoutubeDLHTTPSHandler(urllib
.request
.HTTPSHandler
):
1471 def __init__(self
, params
, https_conn_class
=None, *args
, **kwargs
):
1472 urllib
.request
.HTTPSHandler
.__init
__(self
, *args
, **kwargs
)
1473 self
._https
_conn
_class
= https_conn_class
or http
.client
.HTTPSConnection
1474 self
._params
= params
1476 def https_open(self
, req
):
1478 conn_class
= self
._https
_conn
_class
1480 if hasattr(self
, '_context'): # python > 2.6
1481 kwargs
['context'] = self
._context
1482 if hasattr(self
, '_check_hostname'): # python 3.x
1483 kwargs
['check_hostname'] = self
._check
_hostname
1485 socks_proxy
= req
.headers
.get('Ytdl-socks-proxy')
1487 conn_class
= make_socks_conn_class(conn_class
, socks_proxy
)
1488 del req
.headers
['Ytdl-socks-proxy']
1491 return self
.do_open(
1492 functools
.partial(_create_http_connection
, self
, conn_class
, True), req
, **kwargs
)
1493 except urllib
.error
.URLError
as e
:
1494 if (isinstance(e
.reason
, ssl
.SSLError
)
1495 and getattr(e
.reason
, 'reason', None) == 'SSLV3_ALERT_HANDSHAKE_FAILURE'):
1496 raise YoutubeDLError('SSLV3_ALERT_HANDSHAKE_FAILURE: Try using --legacy-server-connect')
1500 class YoutubeDLCookieJar(http
.cookiejar
.MozillaCookieJar
):
1502 See [1] for cookie file format.
1504 1. https://curl.haxx.se/docs/http-cookies.html
1506 _HTTPONLY_PREFIX
= '#HttpOnly_'
1508 _HEADER
= '''# Netscape HTTP Cookie File
1509 # This file is generated by yt-dlp. Do not edit.
1512 _CookieFileEntry
= collections
.namedtuple(
1514 ('domain_name', 'include_subdomains', 'path', 'https_only', 'expires_at', 'name', 'value'))
1516 def __init__(self
, filename
=None, *args
, **kwargs
):
1517 super().__init
__(None, *args
, **kwargs
)
1518 if self
.is_path(filename
):
1519 filename
= os
.fspath(filename
)
1520 self
.filename
= filename
1523 def _true_or_false(cndn
):
1524 return 'TRUE' if cndn
else 'FALSE'
1528 return isinstance(file, (str, bytes, os
.PathLike
))
1530 @contextlib.contextmanager
1531 def open(self
, file, *, write
=False):
1532 if self
.is_path(file):
1533 with open(file, 'w' if write
else 'r', encoding
='utf-8') as f
:
1540 def _really_save(self
, f
, ignore_discard
=False, ignore_expires
=False):
1543 if (not ignore_discard
and cookie
.discard
1544 or not ignore_expires
and cookie
.is_expired(now
)):
1546 name
, value
= cookie
.name
, cookie
.value
1548 # cookies.txt regards 'Set-Cookie: foo' as a cookie
1549 # with no name, whereas http.cookiejar regards it as a
1550 # cookie with no value.
1551 name
, value
= '', name
1552 f
.write('%s\n' % '\t'.join((
1554 self
._true
_or
_false
(cookie
.domain
.startswith('.')),
1556 self
._true
_or
_false
(cookie
.secure
),
1557 str_or_none(cookie
.expires
, default
=''),
1561 def save(self
, filename
=None, *args
, **kwargs
):
1563 Save cookies to a file.
1564 Code is taken from CPython 3.6
1565 https://github.com/python/cpython/blob/8d999cbf4adea053be6dbb612b9844635c4dfb8e/Lib/http/cookiejar.py#L2091-L2117 """
1567 if filename
is None:
1568 if self
.filename
is not None:
1569 filename
= self
.filename
1571 raise ValueError(http
.cookiejar
.MISSING_FILENAME_TEXT
)
1573 # Store session cookies with `expires` set to 0 instead of an empty string
1575 if cookie
.expires
is None:
1578 with self
.open(filename
, write
=True) as f
:
1579 f
.write(self
._HEADER
)
1580 self
._really
_save
(f
, *args
, **kwargs
)
1582 def load(self
, filename
=None, ignore_discard
=False, ignore_expires
=False):
1583 """Load cookies from a file."""
1584 if filename
is None:
1585 if self
.filename
is not None:
1586 filename
= self
.filename
1588 raise ValueError(http
.cookiejar
.MISSING_FILENAME_TEXT
)
1590 def prepare_line(line
):
1591 if line
.startswith(self
._HTTPONLY
_PREFIX
):
1592 line
= line
[len(self
._HTTPONLY
_PREFIX
):]
1593 # comments and empty lines are fine
1594 if line
.startswith('#') or not line
.strip():
1596 cookie_list
= line
.split('\t')
1597 if len(cookie_list
) != self
._ENTRY
_LEN
:
1598 raise http
.cookiejar
.LoadError('invalid length %d' % len(cookie_list
))
1599 cookie
= self
._CookieFileEntry
(*cookie_list
)
1600 if cookie
.expires_at
and not cookie
.expires_at
.isdigit():
1601 raise http
.cookiejar
.LoadError('invalid expires at %s' % cookie
.expires_at
)
1605 with self
.open(filename
) as f
:
1608 cf
.write(prepare_line(line
))
1609 except http
.cookiejar
.LoadError
as e
:
1610 if f
'{line.strip()} '[0] in '[{"':
1611 raise http
.cookiejar
.LoadError(
1612 'Cookies file must be Netscape formatted, not JSON. See '
1613 'https://github.com/ytdl-org/youtube-dl#how-do-i-pass-cookies-to-youtube-dl')
1614 write_string(f
'WARNING: skipping cookie file entry due to {e}: {line!r}\n')
1617 self
._really
_load
(cf
, filename
, ignore_discard
, ignore_expires
)
1618 # Session cookies are denoted by either `expires` field set to
1619 # an empty string or 0. MozillaCookieJar only recognizes the former
1620 # (see [1]). So we need force the latter to be recognized as session
1621 # cookies on our own.
1622 # Session cookies may be important for cookies-based authentication,
1623 # e.g. usually, when user does not check 'Remember me' check box while
1624 # logging in on a site, some important cookies are stored as session
1625 # cookies so that not recognizing them will result in failed login.
1626 # 1. https://bugs.python.org/issue17164
1628 # Treat `expires=0` cookies as session cookies
1629 if cookie
.expires
== 0:
1630 cookie
.expires
= None
1631 cookie
.discard
= True
1634 class YoutubeDLCookieProcessor(urllib
.request
.HTTPCookieProcessor
):
1635 def __init__(self
, cookiejar
=None):
1636 urllib
.request
.HTTPCookieProcessor
.__init
__(self
, cookiejar
)
1638 def http_response(self
, request
, response
):
1639 return urllib
.request
.HTTPCookieProcessor
.http_response(self
, request
, response
)
1641 https_request
= urllib
.request
.HTTPCookieProcessor
.http_request
1642 https_response
= http_response
1645 class YoutubeDLRedirectHandler(urllib
.request
.HTTPRedirectHandler
):
1646 """YoutubeDL redirect handler
1648 The code is based on HTTPRedirectHandler implementation from CPython [1].
1650 This redirect handler solves two issues:
1651 - ensures redirect URL is always unicode under python 2
1652 - introduces support for experimental HTTP response status code
1653 308 Permanent Redirect [2] used by some sites [3]
1655 1. https://github.com/python/cpython/blob/master/Lib/urllib/request.py
1656 2. https://developer.mozilla.org/en-US/docs/Web/HTTP/Status/308
1657 3. https://github.com/ytdl-org/youtube-dl/issues/28768
1660 http_error_301
= http_error_303
= http_error_307
= http_error_308
= urllib
.request
.HTTPRedirectHandler
.http_error_302
1662 def redirect_request(self
, req
, fp
, code
, msg
, headers
, newurl
):
1663 """Return a Request or None in response to a redirect.
1665 This is called by the http_error_30x methods when a
1666 redirection response is received. If a redirection should
1667 take place, return a new Request to allow http_error_30x to
1668 perform the redirect. Otherwise, raise HTTPError if no-one
1669 else should try to handle this url. Return None if you can't
1670 but another Handler might.
1672 m
= req
.get_method()
1673 if (not (code
in (301, 302, 303, 307, 308) and m
in ("GET", "HEAD")
1674 or code
in (301, 302, 303) and m
== "POST")):
1675 raise urllib
.error
.HTTPError(req
.full_url
, code
, msg
, headers
, fp
)
1676 # Strictly (according to RFC 2616), 301 or 302 in response to
1677 # a POST MUST NOT cause a redirection without confirmation
1678 # from the user (of urllib.request, in this case). In practice,
1679 # essentially all clients do redirect in this case, so we do
1682 # Be conciliant with URIs containing a space. This is mainly
1683 # redundant with the more complete encoding done in http_error_302(),
1684 # but it is kept for compatibility with other callers.
1685 newurl
= newurl
.replace(' ', '%20')
1687 CONTENT_HEADERS
= ("content-length", "content-type")
1688 # NB: don't use dict comprehension for python 2.6 compatibility
1689 newheaders
= {k: v for k, v in req.headers.items() if k.lower() not in CONTENT_HEADERS}
1691 # A 303 must either use GET or HEAD for subsequent request
1692 # https://datatracker.ietf.org/doc/html/rfc7231#section-6.4.4
1693 if code
== 303 and m
!= 'HEAD':
1695 # 301 and 302 redirects are commonly turned into a GET from a POST
1696 # for subsequent requests by browsers, so we'll do the same.
1697 # https://datatracker.ietf.org/doc/html/rfc7231#section-6.4.2
1698 # https://datatracker.ietf.org/doc/html/rfc7231#section-6.4.3
1699 if code
in (301, 302) and m
== 'POST':
1702 return urllib
.request
.Request(
1703 newurl
, headers
=newheaders
, origin_req_host
=req
.origin_req_host
,
1704 unverifiable
=True, method
=m
)
1707 def extract_timezone(date_str
):
1710 ^.{8,}? # >=8 char non-TZ prefix, if present
1711 (?P<tz>Z| # just the UTC Z, or
1712 (?:(?<=.\b\d{4}|\b\d{2}:\d\d)| # preceded by 4 digits or hh:mm or
1713 (?<!.\b[a-zA-Z]{3}|[a-zA-Z]{4}|..\b\d\d)) # not preceded by 3 alpha word or >= 4 alpha or 2 digits
1714 [ ]? # optional space
1715 (?P<sign>\+|-) # +/-
1716 (?P<hours>[0-9]{2}):?(?P<minutes>[0-9]{2}) # hh[:]mm
1720 m
= re
.search(r
'\d{1,2}:\d{1,2}(?:\.\d+)?(?P<tz>\s*[A-Z]+)$', date_str
)
1721 timezone
= TIMEZONE_NAMES
.get(m
and m
.group('tz').strip())
1722 if timezone
is not None:
1723 date_str
= date_str
[:-len(m
.group('tz'))]
1724 timezone
= datetime
.timedelta(hours
=timezone
or 0)
1726 date_str
= date_str
[:-len(m
.group('tz'))]
1727 if not m
.group('sign'):
1728 timezone
= datetime
.timedelta()
1730 sign
= 1 if m
.group('sign') == '+' else -1
1731 timezone
= datetime
.timedelta(
1732 hours
=sign
* int(m
.group('hours')),
1733 minutes
=sign
* int(m
.group('minutes')))
1734 return timezone
, date_str
1737 def parse_iso8601(date_str
, delimiter
='T', timezone
=None):
1738 """ Return a UNIX timestamp from the given date """
1740 if date_str
is None:
1743 date_str
= re
.sub(r
'\.[0-9]+', '', date_str
)
1745 if timezone
is None:
1746 timezone
, date_str
= extract_timezone(date_str
)
1748 with contextlib
.suppress(ValueError):
1749 date_format
= f
'%Y-%m-%d{delimiter}%H:%M:%S'
1750 dt
= datetime
.datetime
.strptime(date_str
, date_format
) - timezone
1751 return calendar
.timegm(dt
.timetuple())
1754 def date_formats(day_first
=True):
1755 return DATE_FORMATS_DAY_FIRST
if day_first
else DATE_FORMATS_MONTH_FIRST
1758 def unified_strdate(date_str
, day_first
=True):
1759 """Return a string with the date in the format YYYYMMDD"""
1761 if date_str
is None:
1765 date_str
= date_str
.replace(',', ' ')
1766 # Remove AM/PM + timezone
1767 date_str
= re
.sub(r
'(?i)\s*(?:AM|PM)(?:\s+[A-Z]+)?', '', date_str
)
1768 _
, date_str
= extract_timezone(date_str
)
1770 for expression
in date_formats(day_first
):
1771 with contextlib
.suppress(ValueError):
1772 upload_date
= datetime
.datetime
.strptime(date_str
, expression
).strftime('%Y%m%d')
1773 if upload_date
is None:
1774 timetuple
= email
.utils
.parsedate_tz(date_str
)
1776 with contextlib
.suppress(ValueError):
1777 upload_date
= datetime
.datetime(*timetuple
[:6]).strftime('%Y%m%d')
1778 if upload_date
is not None:
1779 return str(upload_date
)
1782 def unified_timestamp(date_str
, day_first
=True):
1783 if date_str
is None:
1786 date_str
= re
.sub(r
'\s+', ' ', re
.sub(
1787 r
'(?i)[,|]|(mon|tues?|wed(nes)?|thu(rs)?|fri|sat(ur)?)(day)?', '', date_str
))
1789 pm_delta
= 12 if re
.search(r
'(?i)PM', date_str
) else 0
1790 timezone
, date_str
= extract_timezone(date_str
)
1792 # Remove AM/PM + timezone
1793 date_str
= re
.sub(r
'(?i)\s*(?:AM|PM)(?:\s+[A-Z]+)?', '', date_str
)
1795 # Remove unrecognized timezones from ISO 8601 alike timestamps
1796 m
= re
.search(r
'\d{1,2}:\d{1,2}(?:\.\d+)?(?P<tz>\s*[A-Z]+)$', date_str
)
1798 date_str
= date_str
[:-len(m
.group('tz'))]
1800 # Python only supports microseconds, so remove nanoseconds
1801 m
= re
.search(r
'^([0-9]{4,}-[0-9]{1,2}-[0-9]{1,2}T[0-9]{1,2}:[0-9]{1,2}:[0-9]{1,2}\.[0-9]{6})[0-9]+$', date_str
)
1803 date_str
= m
.group(1)
1805 for expression
in date_formats(day_first
):
1806 with contextlib
.suppress(ValueError):
1807 dt
= datetime
.datetime
.strptime(date_str
, expression
) - timezone
+ datetime
.timedelta(hours
=pm_delta
)
1808 return calendar
.timegm(dt
.timetuple())
1810 timetuple
= email
.utils
.parsedate_tz(date_str
)
1812 return calendar
.timegm(timetuple
) + pm_delta
* 3600 - timezone
.total_seconds()
1815 def determine_ext(url
, default_ext
='unknown_video'):
1816 if url
is None or '.' not in url
:
1818 guess
= url
.partition('?')[0].rpartition('.')[2]
1819 if re
.match(r
'^[A-Za-z0-9]+$', guess
):
1821 # Try extract ext from URLs like http://example.com/foo/bar.mp4/?download
1822 elif guess
.rstrip('/') in KNOWN_EXTENSIONS
:
1823 return guess
.rstrip('/')
1828 def subtitles_filename(filename
, sub_lang
, sub_format
, expected_real_ext
=None):
1829 return replace_extension(filename
, sub_lang
+ '.' + sub_format
, expected_real_ext
)
1832 def datetime_from_str(date_str
, precision
='auto', format
='%Y%m%d'):
1834 Return a datetime object from a string.
1836 (now|today|yesterday|DATE)([+-]\d+(microsecond|second|minute|hour|day|week|month|year)s?)?
1838 @param format strftime format of DATE
1839 @param precision Round the datetime object: auto|microsecond|second|minute|hour|day
1840 auto: round to the unit provided in date_str (if applicable).
1842 auto_precision
= False
1843 if precision
== 'auto':
1844 auto_precision
= True
1845 precision
= 'microsecond'
1846 today
= datetime_round(datetime
.datetime
.utcnow(), precision
)
1847 if date_str
in ('now', 'today'):
1849 if date_str
== 'yesterday':
1850 return today
- datetime
.timedelta(days
=1)
1852 r
'(?P<start>.+)(?P<sign>[+-])(?P<time>\d+)(?P<unit>microsecond|second|minute|hour|day|week|month|year)s?',
1854 if match
is not None:
1855 start_time
= datetime_from_str(match
.group('start'), precision
, format
)
1856 time
= int(match
.group('time')) * (-1 if match
.group('sign') == '-' else 1)
1857 unit
= match
.group('unit')
1858 if unit
== 'month' or unit
== 'year':
1859 new_date
= datetime_add_months(start_time
, time
* 12 if unit
== 'year' else time
)
1865 delta
= datetime
.timedelta(**{unit + 's': time}
)
1866 new_date
= start_time
+ delta
1868 return datetime_round(new_date
, unit
)
1871 return datetime_round(datetime
.datetime
.strptime(date_str
, format
), precision
)
1874 def date_from_str(date_str
, format
='%Y%m%d', strict
=False):
1876 Return a date object from a string using datetime_from_str
1878 @param strict Restrict allowed patterns to "YYYYMMDD" and
1879 (now|today|yesterday)(-\d+(day|week|month|year)s?)?
1881 if strict
and not re
.fullmatch(r
'\d{8}|(now|today|yesterday)(-\d+(day|week|month|year)s?)?', date_str
):
1882 raise ValueError(f
'Invalid date format "{date_str}"')
1883 return datetime_from_str(date_str
, precision
='microsecond', format
=format
).date()
1886 def datetime_add_months(dt
, months
):
1887 """Increment/Decrement a datetime object by months."""
1888 month
= dt
.month
+ months
- 1
1889 year
= dt
.year
+ month
// 12
1890 month
= month
% 12 + 1
1891 day
= min(dt
.day
, calendar
.monthrange(year
, month
)[1])
1892 return dt
.replace(year
, month
, day
)
1895 def datetime_round(dt
, precision
='day'):
1897 Round a datetime object's time to a specific precision
1899 if precision
== 'microsecond':
1908 roundto
= lambda x
, n
: ((x
+ n
/ 2) // n
) * n
1909 timestamp
= calendar
.timegm(dt
.timetuple())
1910 return datetime
.datetime
.utcfromtimestamp(roundto(timestamp
, unit_seconds
[precision
]))
1913 def hyphenate_date(date_str
):
1915 Convert a date in 'YYYYMMDD' format to 'YYYY-MM-DD' format"""
1916 match
= re
.match(r
'^(\d\d\d\d)(\d\d)(\d\d)$', date_str
)
1917 if match
is not None:
1918 return '-'.join(match
.groups())
1924 """Represents a time interval between two dates"""
1926 def __init__(self
, start
=None, end
=None):
1927 """start and end must be strings in the format accepted by date"""
1928 if start
is not None:
1929 self
.start
= date_from_str(start
, strict
=True)
1931 self
.start
= datetime
.datetime
.min.date()
1933 self
.end
= date_from_str(end
, strict
=True)
1935 self
.end
= datetime
.datetime
.max.date()
1936 if self
.start
> self
.end
:
1937 raise ValueError('Date range: "%s" , the start date must be before the end date' % self
)
1941 """Returns a range that only contains the given day"""
1942 return cls(day
, day
)
1944 def __contains__(self
, date
):
1945 """Check if the date is in the range"""
1946 if not isinstance(date
, datetime
.date
):
1947 date
= date_from_str(date
)
1948 return self
.start
<= date
<= self
.end
1951 return f
'{self.start.isoformat()} - {self.end.isoformat()}'
1953 def __eq__(self
, other
):
1954 return (isinstance(other
, DateRange
)
1955 and self
.start
== other
.start
and self
.end
== other
.end
)
1958 def platform_name():
1959 """ Returns the platform name as a str """
1960 deprecation_warning(f
'"{__name__}.platform_name" is deprecated, use "platform.platform" instead')
1961 return platform
.platform()
1965 def system_identifier():
1966 python_implementation
= platform
.python_implementation()
1967 if python_implementation
== 'PyPy' and hasattr(sys
, 'pypy_version_info'):
1968 python_implementation
+= ' version %d.%d.%d' % sys
.pypy_version_info
[:3]
1970 return 'Python %s (%s %s) - %s %s' % (
1971 platform
.python_version(),
1972 python_implementation
,
1973 platform
.architecture()[0],
1974 platform
.platform(),
1975 format_field(join_nonempty(*platform
.libc_ver(), delim
=' '), None, '(%s)'),
1980 def get_windows_version():
1981 ''' Get Windows version. returns () if it's not running on Windows '''
1982 if compat_os_name
== 'nt':
1983 return version_tuple(platform
.win32_ver()[1])
1988 def write_string(s
, out
=None, encoding
=None):
1989 assert isinstance(s
, str)
1990 out
= out
or sys
.stderr
1992 if compat_os_name
== 'nt' and supports_terminal_sequences(out
):
1993 s
= re
.sub(r
'([\r\n]+)', r
' \1', s
)
1995 enc
, buffer = None, out
1996 if 'b' in getattr(out
, 'mode', ''):
1997 enc
= encoding
or preferredencoding()
1998 elif hasattr(out
, 'buffer'):
2000 enc
= encoding
or getattr(out
, 'encoding', None) or preferredencoding()
2002 buffer.write(s
.encode(enc
, 'ignore') if enc
else s
)
2006 def deprecation_warning(msg
, *, printer
=None, stacklevel
=0, **kwargs
):
2007 from . import _IN_CLI
2009 if msg
in deprecation_warning
._cache
:
2011 deprecation_warning
._cache
.add(msg
)
2013 return printer(f
'{msg}{bug_reports_message()}', **kwargs
)
2014 return write_string(f
'ERROR: {msg}{bug_reports_message()}\n', **kwargs
)
2017 warnings
.warn(DeprecationWarning(msg
), stacklevel
=stacklevel
+ 3)
2020 deprecation_warning
._cache
= set()
2023 def bytes_to_intlist(bs
):
2026 if isinstance(bs
[0], int): # Python 3
2029 return [ord(c
) for c
in bs
]
2032 def intlist_to_bytes(xs
):
2035 return struct
.pack('%dB' % len(xs
), *xs
)
2038 class LockingUnsupportedError(OSError):
2039 msg
= 'File locking is not supported'
2042 super().__init
__(self
.msg
)
2045 # Cross-platform file locking
2046 if sys
.platform
== 'win32':
2048 import ctypes
.wintypes
2051 class OVERLAPPED(ctypes
.Structure
):
2053 ('Internal', ctypes
.wintypes
.LPVOID
),
2054 ('InternalHigh', ctypes
.wintypes
.LPVOID
),
2055 ('Offset', ctypes
.wintypes
.DWORD
),
2056 ('OffsetHigh', ctypes
.wintypes
.DWORD
),
2057 ('hEvent', ctypes
.wintypes
.HANDLE
),
2060 kernel32
= ctypes
.windll
.kernel32
2061 LockFileEx
= kernel32
.LockFileEx
2062 LockFileEx
.argtypes
= [
2063 ctypes
.wintypes
.HANDLE
, # hFile
2064 ctypes
.wintypes
.DWORD
, # dwFlags
2065 ctypes
.wintypes
.DWORD
, # dwReserved
2066 ctypes
.wintypes
.DWORD
, # nNumberOfBytesToLockLow
2067 ctypes
.wintypes
.DWORD
, # nNumberOfBytesToLockHigh
2068 ctypes
.POINTER(OVERLAPPED
) # Overlapped
2070 LockFileEx
.restype
= ctypes
.wintypes
.BOOL
2071 UnlockFileEx
= kernel32
.UnlockFileEx
2072 UnlockFileEx
.argtypes
= [
2073 ctypes
.wintypes
.HANDLE
, # hFile
2074 ctypes
.wintypes
.DWORD
, # dwReserved
2075 ctypes
.wintypes
.DWORD
, # nNumberOfBytesToLockLow
2076 ctypes
.wintypes
.DWORD
, # nNumberOfBytesToLockHigh
2077 ctypes
.POINTER(OVERLAPPED
) # Overlapped
2079 UnlockFileEx
.restype
= ctypes
.wintypes
.BOOL
2080 whole_low
= 0xffffffff
2081 whole_high
= 0x7fffffff
2083 def _lock_file(f
, exclusive
, block
):
2084 overlapped
= OVERLAPPED()
2085 overlapped
.Offset
= 0
2086 overlapped
.OffsetHigh
= 0
2087 overlapped
.hEvent
= 0
2088 f
._lock
_file
_overlapped
_p
= ctypes
.pointer(overlapped
)
2090 if not LockFileEx(msvcrt
.get_osfhandle(f
.fileno()),
2091 (0x2 if exclusive
else 0x0) |
(0x0 if block
else 0x1),
2092 0, whole_low
, whole_high
, f
._lock
_file
_overlapped
_p
):
2093 # NB: No argument form of "ctypes.FormatError" does not work on PyPy
2094 raise BlockingIOError(f
'Locking file failed: {ctypes.FormatError(ctypes.GetLastError())!r}')
2096 def _unlock_file(f
):
2097 assert f
._lock
_file
_overlapped
_p
2098 handle
= msvcrt
.get_osfhandle(f
.fileno())
2099 if not UnlockFileEx(handle
, 0, whole_low
, whole_high
, f
._lock
_file
_overlapped
_p
):
2100 raise OSError('Unlocking file failed: %r' % ctypes
.FormatError())
2106 def _lock_file(f
, exclusive
, block
):
2107 flags
= fcntl
.LOCK_EX
if exclusive
else fcntl
.LOCK_SH
2109 flags |
= fcntl
.LOCK_NB
2111 fcntl
.flock(f
, flags
)
2112 except BlockingIOError
:
2114 except OSError: # AOSP does not have flock()
2115 fcntl
.lockf(f
, flags
)
2117 def _unlock_file(f
):
2119 fcntl
.flock(f
, fcntl
.LOCK_UN
)
2121 fcntl
.lockf(f
, fcntl
.LOCK_UN
)
2125 def _lock_file(f
, exclusive
, block
):
2126 raise LockingUnsupportedError()
2128 def _unlock_file(f
):
2129 raise LockingUnsupportedError()
2135 def __init__(self
, filename
, mode
, block
=True, encoding
=None):
2136 if mode
not in {'r', 'rb', 'a', 'ab', 'w', 'wb'}
:
2137 raise NotImplementedError(mode
)
2138 self
.mode
, self
.block
= mode
, block
2140 writable
= any(f
in mode
for f
in 'wax+')
2141 readable
= any(f
in mode
for f
in 'r+')
2142 flags
= functools
.reduce(operator
.ior
, (
2143 getattr(os
, 'O_CLOEXEC', 0), # UNIX only
2144 getattr(os
, 'O_BINARY', 0), # Windows only
2145 getattr(os
, 'O_NOINHERIT', 0), # Windows only
2146 os
.O_CREAT
if writable
else 0, # O_TRUNC only after locking
2147 os
.O_APPEND
if 'a' in mode
else 0,
2148 os
.O_EXCL
if 'x' in mode
else 0,
2149 os
.O_RDONLY
if not writable
else os
.O_RDWR
if readable
else os
.O_WRONLY
,
2152 self
.f
= os
.fdopen(os
.open(filename
, flags
, 0o666), mode
, encoding
=encoding
)
2154 def __enter__(self
):
2155 exclusive
= 'r' not in self
.mode
2157 _lock_file(self
.f
, exclusive
, self
.block
)
2162 if 'w' in self
.mode
:
2165 except OSError as e
:
2167 errno
.ESPIPE
, # Illegal seek - expected for FIFO
2168 errno
.EINVAL
, # Invalid argument - expected for /dev/null
2177 _unlock_file(self
.f
)
2181 def __exit__(self
, *_
):
2190 def __getattr__(self
, attr
):
2191 return getattr(self
.f
, attr
)
2198 def get_filesystem_encoding():
2199 encoding
= sys
.getfilesystemencoding()
2200 return encoding
if encoding
is not None else 'utf-8'
2203 def shell_quote(args
):
2205 encoding
= get_filesystem_encoding()
2207 if isinstance(a
, bytes):
2208 # We may get a filename encoded with 'encodeFilename'
2209 a
= a
.decode(encoding
)
2210 quoted_args
.append(compat_shlex_quote(a
))
2211 return ' '.join(quoted_args
)
2214 def smuggle_url(url
, data
):
2215 """ Pass additional data in a URL for internal use. """
2217 url
, idata
= unsmuggle_url(url
, {})
2219 sdata
= urllib
.parse
.urlencode(
2220 {'__youtubedl_smuggle': json.dumps(data)}
)
2221 return url
+ '#' + sdata
2224 def unsmuggle_url(smug_url
, default
=None):
2225 if '#__youtubedl_smuggle' not in smug_url
:
2226 return smug_url
, default
2227 url
, _
, sdata
= smug_url
.rpartition('#')
2228 jsond
= urllib
.parse
.parse_qs(sdata
)['__youtubedl_smuggle'][0]
2229 data
= json
.loads(jsond
)
2233 def format_decimal_suffix(num
, fmt
='%d%s', *, factor
=1000):
2234 """ Formats numbers with decimal sufixes like K, M, etc """
2235 num
, factor
= float_or_none(num
), float(factor
)
2236 if num
is None or num
< 0:
2238 POSSIBLE_SUFFIXES
= 'kMGTPEZY'
2239 exponent
= 0 if num
== 0 else min(int(math
.log(num
, factor
)), len(POSSIBLE_SUFFIXES
))
2240 suffix
= ['', *POSSIBLE_SUFFIXES
][exponent
]
2242 suffix
= {'k': 'Ki', '': ''}
.get(suffix
, f
'{suffix}i')
2243 converted
= num
/ (factor
** exponent
)
2244 return fmt
% (converted
, suffix
)
2247 def format_bytes(bytes):
2248 return format_decimal_suffix(bytes, '%.2f%sB', factor
=1024) or 'N/A'
2251 def lookup_unit_table(unit_table
, s
):
2252 units_re
= '|'.join(re
.escape(u
) for u
in unit_table
)
2254 r
'(?P<num>[0-9]+(?:[,.][0-9]*)?)\s*(?P<unit>%s)\b' % units_re
, s
)
2257 num_str
= m
.group('num').replace(',', '.')
2258 mult
= unit_table
[m
.group('unit')]
2259 return int(float(num_str
) * mult
)
2262 def parse_filesize(s
):
2266 # The lower-case forms are of course incorrect and unofficial,
2267 # but we support those too
2284 'megabytes': 1000 ** 2,
2285 'mebibytes': 1024 ** 2,
2291 'gigabytes': 1000 ** 3,
2292 'gibibytes': 1024 ** 3,
2298 'terabytes': 1000 ** 4,
2299 'tebibytes': 1024 ** 4,
2305 'petabytes': 1000 ** 5,
2306 'pebibytes': 1024 ** 5,
2312 'exabytes': 1000 ** 6,
2313 'exbibytes': 1024 ** 6,
2319 'zettabytes': 1000 ** 7,
2320 'zebibytes': 1024 ** 7,
2326 'yottabytes': 1000 ** 8,
2327 'yobibytes': 1024 ** 8,
2330 return lookup_unit_table(_UNIT_TABLE
, s
)
2337 s
= re
.sub(r
'^[^\d]+\s', '', s
).strip()
2339 if re
.match(r
'^[\d,.]+$', s
):
2340 return str_to_int(s
)
2353 ret
= lookup_unit_table(_UNIT_TABLE
, s
)
2357 mobj
= re
.match(r
'([\d,.]+)(?:$|\s)', s
)
2359 return str_to_int(mobj
.group(1))
2362 def parse_resolution(s
, *, lenient
=False):
2367 mobj
= re
.search(r
'(?P<w>\d+)\s*[xX×,]\s*(?P<h>\d+)', s
)
2369 mobj
= re
.search(r
'(?<![a-zA-Z0-9])(?P<w>\d+)\s*[xX×,]\s*(?P<h>\d+)(?![a-zA-Z0-9])', s
)
2372 'width': int(mobj
.group('w')),
2373 'height': int(mobj
.group('h')),
2376 mobj
= re
.search(r
'(?<![a-zA-Z0-9])(\d+)[pPiI](?![a-zA-Z0-9])', s
)
2378 return {'height': int(mobj.group(1))}
2380 mobj
= re
.search(r
'\b([48])[kK]\b', s
)
2382 return {'height': int(mobj.group(1)) * 540}
2387 def parse_bitrate(s
):
2388 if not isinstance(s
, str):
2390 mobj
= re
.search(r
'\b(\d+)\s*kbps', s
)
2392 return int(mobj
.group(1))
2395 def month_by_name(name
, lang
='en'):
2396 """ Return the number of a month by (locale-independently) English name """
2398 month_names
= MONTH_NAMES
.get(lang
, MONTH_NAMES
['en'])
2401 return month_names
.index(name
) + 1
2406 def month_by_abbreviation(abbrev
):
2407 """ Return the number of a month by (locale-independently) English
2411 return [s
[:3] for s
in ENGLISH_MONTH_NAMES
].index(abbrev
) + 1
2416 def fix_xml_ampersands(xml_str
):
2417 """Replace all the '&' by '&' in XML"""
2419 r
'&(?!amp;|lt;|gt;|apos;|quot;|#x[0-9a-fA-F]{,4};|#[0-9]{,4};)',
2424 def setproctitle(title
):
2425 assert isinstance(title
, str)
2427 # Workaround for https://github.com/yt-dlp/yt-dlp/issues/4541
2434 libc
= ctypes
.cdll
.LoadLibrary('libc.so.6')
2438 # LoadLibrary in Windows Python 2.7.13 only expects
2439 # a bytestring, but since unicode_literals turns
2440 # every string into a unicode string, it fails.
2442 title_bytes
= title
.encode()
2443 buf
= ctypes
.create_string_buffer(len(title_bytes
))
2444 buf
.value
= title_bytes
2446 libc
.prctl(15, buf
, 0, 0, 0)
2447 except AttributeError:
2448 return # Strange libc, just skip this
2451 def remove_start(s
, start
):
2452 return s
[len(start
):] if s
is not None and s
.startswith(start
) else s
2455 def remove_end(s
, end
):
2456 return s
[:-len(end
)] if s
is not None and s
.endswith(end
) else s
2459 def remove_quotes(s
):
2460 if s
is None or len(s
) < 2:
2462 for quote
in ('"', "'", ):
2463 if s
[0] == quote
and s
[-1] == quote
:
2468 def get_domain(url
):
2470 This implementation is inconsistent, but is kept for compatibility.
2471 Use this only for "webpage_url_domain"
2473 return remove_start(urllib
.parse
.urlparse(url
).netloc
, 'www.') or None
2476 def url_basename(url
):
2477 path
= urllib
.parse
.urlparse(url
).path
2478 return path
.strip('/').split('/')[-1]
2482 return re
.match(r
'https?://[^?#&]+/', url
).group()
2485 def urljoin(base
, path
):
2486 if isinstance(path
, bytes):
2487 path
= path
.decode()
2488 if not isinstance(path
, str) or not path
:
2490 if re
.match(r
'^(?:[a-zA-Z][a-zA-Z0-9+-.]*:)?//', path
):
2492 if isinstance(base
, bytes):
2493 base
= base
.decode()
2494 if not isinstance(base
, str) or not re
.match(
2495 r
'^(?:https?:)?//', base
):
2497 return urllib
.parse
.urljoin(base
, path
)
2500 class HEADRequest(urllib
.request
.Request
):
2501 def get_method(self
):
2505 class PUTRequest(urllib
.request
.Request
):
2506 def get_method(self
):
2510 def int_or_none(v
, scale
=1, default
=None, get_attr
=None, invscale
=1):
2511 if get_attr
and v
is not None:
2512 v
= getattr(v
, get_attr
, None)
2514 return int(v
) * invscale
// scale
2515 except (ValueError, TypeError, OverflowError):
2519 def str_or_none(v
, default
=None):
2520 return default
if v
is None else str(v
)
2523 def str_to_int(int_str
):
2524 """ A more relaxed version of int_or_none """
2525 if isinstance(int_str
, int):
2527 elif isinstance(int_str
, str):
2528 int_str
= re
.sub(r
'[,\.\+]', '', int_str
)
2529 return int_or_none(int_str
)
2532 def float_or_none(v
, scale
=1, invscale
=1, default
=None):
2536 return float(v
) * invscale
/ scale
2537 except (ValueError, TypeError):
2541 def bool_or_none(v
, default
=None):
2542 return v
if isinstance(v
, bool) else default
2545 def strip_or_none(v
, default
=None):
2546 return v
.strip() if isinstance(v
, str) else default
2549 def url_or_none(url
):
2550 if not url
or not isinstance(url
, str):
2553 return url
if re
.match(r
'^(?:(?:https?|rt(?:m(?:pt?[es]?|fp)|sp[su]?)|mms|ftps?):)?//', url
) else None
2556 def request_to_url(req
):
2557 if isinstance(req
, urllib
.request
.Request
):
2558 return req
.get_full_url()
2563 def strftime_or_none(timestamp
, date_format
, default
=None):
2564 datetime_object
= None
2566 if isinstance(timestamp
, (int, float)): # unix timestamp
2567 datetime_object
= datetime
.datetime
.utcfromtimestamp(timestamp
)
2568 elif isinstance(timestamp
, str): # assume YYYYMMDD
2569 datetime_object
= datetime
.datetime
.strptime(timestamp
, '%Y%m%d')
2570 return datetime_object
.strftime(date_format
)
2571 except (ValueError, TypeError, AttributeError):
2575 def parse_duration(s
):
2576 if not isinstance(s
, str):
2582 days
, hours
, mins
, secs
, ms
= [None] * 5
2583 m
= re
.match(r
'''(?x)
2585 (?:(?:(?P<days>[0-9]+):)?(?P<hours>[0-9]+):)?(?P<mins>[0-9]+):)?
2586 (?P<secs>(?(before_secs)[0-9]{1,2}|[0-9]+))
2587 (?P<ms>[.:][0-9]+)?Z?$
2590 days
, hours
, mins
, secs
, ms
= m
.group('days', 'hours', 'mins', 'secs', 'ms')
2595 [0-9]+\s*y(?:ears?)?,?\s*
2598 [0-9]+\s*m(?:onths?)?,?\s*
2601 [0-9]+\s*w(?:eeks?)?,?\s*
2604 (?P<days>[0-9]+)\s*d(?:ays?)?,?\s*
2608 (?P<hours>[0-9]+)\s*h(?:ours?)?,?\s*
2611 (?P<mins>[0-9]+)\s*m(?:in(?:ute)?s?)?,?\s*
2614 (?P<secs>[0-9]+)(?P<ms>\.[0-9]+)?\s*s(?:ec(?:ond)?s?)?\s*
2617 days
, hours
, mins
, secs
, ms
= m
.groups()
2619 m
= re
.match(r
'(?i)(?:(?P<hours>[0-9.]+)\s*(?:hours?)|(?P<mins>[0-9.]+)\s*(?:mins?\.?|minutes?)\s*)Z?$', s
)
2621 hours
, mins
= m
.groups()
2626 ms
= ms
.replace(':', '.')
2627 return sum(float(part
or 0) * mult
for part
, mult
in (
2628 (days
, 86400), (hours
, 3600), (mins
, 60), (secs
, 1), (ms
, 1)))
2631 def prepend_extension(filename
, ext
, expected_real_ext
=None):
2632 name
, real_ext
= os
.path
.splitext(filename
)
2634 f
'{name}.{ext}{real_ext}'
2635 if not expected_real_ext
or real_ext
[1:] == expected_real_ext
2636 else f
'{filename}.{ext}')
2639 def replace_extension(filename
, ext
, expected_real_ext
=None):
2640 name
, real_ext
= os
.path
.splitext(filename
)
2641 return '{}.{}'.format(
2642 name
if not expected_real_ext
or real_ext
[1:] == expected_real_ext
else filename
,
2646 def check_executable(exe
, args
=[]):
2647 """ Checks if the given binary is installed somewhere in PATH, and returns its name.
2648 args can be a list of arguments for a short output (like -version) """
2650 Popen
.run([exe
] + args
, stdout
=subprocess
.PIPE
, stderr
=subprocess
.PIPE
)
2656 def _get_exe_version_output(exe
, args
, *, to_screen
=None):
2658 to_screen(f
'Checking exe version: {shell_quote([exe] + args)}')
2660 # STDIN should be redirected too. On UNIX-like systems, ffmpeg triggers
2661 # SIGTTOU if yt-dlp is run in the background.
2662 # See https://github.com/ytdl-org/youtube-dl/issues/955#issuecomment-209789656
2663 stdout
, _
, _
= Popen
.run([encodeArgument(exe
)] + args
, text
=True,
2664 stdin
=subprocess
.PIPE
, stdout
=subprocess
.PIPE
, stderr
=subprocess
.STDOUT
)
2670 def detect_exe_version(output
, version_re
=None, unrecognized
='present'):
2671 assert isinstance(output
, str)
2672 if version_re
is None:
2673 version_re
= r
'version\s+([-0-9._a-zA-Z]+)'
2674 m
= re
.search(version_re
, output
)
2681 def get_exe_version(exe
, args
=['--version'],
2682 version_re
=None, unrecognized
='present'):
2683 """ Returns the version of the specified executable,
2684 or False if the executable is not present """
2685 out
= _get_exe_version_output(exe
, args
)
2686 return detect_exe_version(out
, version_re
, unrecognized
) if out
else False
2689 def frange(start
=0, stop
=None, step
=1):
2692 start
, stop
= 0, start
2693 sign
= [-1, 1][step
> 0] if step
else 0
2694 while sign
* start
< sign
* stop
:
2699 class LazyList(collections
.abc
.Sequence
):
2700 """Lazy immutable list from an iterable
2701 Note that slices of a LazyList are lists and not LazyList"""
2703 class IndexError(IndexError):
2706 def __init__(self
, iterable
, *, reverse
=False, _cache
=None):
2707 self
._iterable
= iter(iterable
)
2708 self
._cache
= [] if _cache
is None else _cache
2709 self
._reversed
= reverse
2713 # We need to consume the entire iterable to iterate in reverse
2714 yield from self
.exhaust()
2716 yield from self
._cache
2717 for item
in self
._iterable
:
2718 self
._cache
.append(item
)
2722 self
._cache
.extend(self
._iterable
)
2723 self
._iterable
= [] # Discard the emptied iterable to make it pickle-able
2727 """Evaluate the entire iterable"""
2728 return self
._exhaust
()[::-1 if self
._reversed
else 1]
2731 def _reverse_index(x
):
2732 return None if x
is None else ~x
2734 def __getitem__(self
, idx
):
2735 if isinstance(idx
, slice):
2737 idx
= slice(self
._reverse
_index
(idx
.start
), self
._reverse
_index
(idx
.stop
), -(idx
.step
or 1))
2738 start
, stop
, step
= idx
.start
, idx
.stop
, idx
.step
or 1
2739 elif isinstance(idx
, int):
2741 idx
= self
._reverse
_index
(idx
)
2742 start
, stop
, step
= idx
, idx
, 0
2744 raise TypeError('indices must be integers or slices')
2745 if ((start
or 0) < 0 or (stop
or 0) < 0
2746 or (start
is None and step
< 0)
2747 or (stop
is None and step
> 0)):
2748 # We need to consume the entire iterable to be able to slice from the end
2749 # Obviously, never use this with infinite iterables
2752 return self
._cache
[idx
]
2753 except IndexError as e
:
2754 raise self
.IndexError(e
) from e
2755 n
= max(start
or 0, stop
or 0) - len(self
._cache
) + 1
2757 self
._cache
.extend(itertools
.islice(self
._iterable
, n
))
2759 return self
._cache
[idx
]
2760 except IndexError as e
:
2761 raise self
.IndexError(e
) from e
2765 self
[-1] if self
._reversed
else self
[0]
2766 except self
.IndexError:
2772 return len(self
._cache
)
2774 def __reversed__(self
):
2775 return type(self
)(self
._iterable
, reverse
=not self
._reversed
, _cache
=self
._cache
)
2778 return type(self
)(self
._iterable
, reverse
=self
._reversed
, _cache
=self
._cache
)
2781 # repr and str should mimic a list. So we exhaust the iterable
2782 return repr(self
.exhaust())
2785 return repr(self
.exhaust())
2790 class IndexError(IndexError):
2794 # This is only useful for tests
2795 return len(self
.getslice())
2797 def __init__(self
, pagefunc
, pagesize
, use_cache
=True):
2798 self
._pagefunc
= pagefunc
2799 self
._pagesize
= pagesize
2800 self
._pagecount
= float('inf')
2801 self
._use
_cache
= use_cache
2804 def getpage(self
, pagenum
):
2805 page_results
= self
._cache
.get(pagenum
)
2806 if page_results
is None:
2807 page_results
= [] if pagenum
> self
._pagecount
else list(self
._pagefunc
(pagenum
))
2809 self
._cache
[pagenum
] = page_results
2812 def getslice(self
, start
=0, end
=None):
2813 return list(self
._getslice
(start
, end
))
2815 def _getslice(self
, start
, end
):
2816 raise NotImplementedError('This method must be implemented by subclasses')
2818 def __getitem__(self
, idx
):
2819 assert self
._use
_cache
, 'Indexing PagedList requires cache'
2820 if not isinstance(idx
, int) or idx
< 0:
2821 raise TypeError('indices must be non-negative integers')
2822 entries
= self
.getslice(idx
, idx
+ 1)
2824 raise self
.IndexError()
2828 class OnDemandPagedList(PagedList
):
2829 """Download pages until a page with less than maximum results"""
2831 def _getslice(self
, start
, end
):
2832 for pagenum
in itertools
.count(start
// self
._pagesize
):
2833 firstid
= pagenum
* self
._pagesize
2834 nextfirstid
= pagenum
* self
._pagesize
+ self
._pagesize
2835 if start
>= nextfirstid
:
2839 start
% self
._pagesize
2840 if firstid
<= start
< nextfirstid
2843 ((end
- 1) % self
._pagesize
) + 1
2844 if (end
is not None and firstid
<= end
<= nextfirstid
)
2848 page_results
= self
.getpage(pagenum
)
2850 self
._pagecount
= pagenum
- 1
2852 if startv
!= 0 or endv
is not None:
2853 page_results
= page_results
[startv
:endv
]
2854 yield from page_results
2856 # A little optimization - if current page is not "full", ie. does
2857 # not contain page_size videos then we can assume that this page
2858 # is the last one - there are no more ids on further pages -
2859 # i.e. no need to query again.
2860 if len(page_results
) + startv
< self
._pagesize
:
2863 # If we got the whole page, but the next page is not interesting,
2864 # break out early as well
2865 if end
== nextfirstid
:
2869 class InAdvancePagedList(PagedList
):
2870 """PagedList with total number of pages known in advance"""
2872 def __init__(self
, pagefunc
, pagecount
, pagesize
):
2873 PagedList
.__init
__(self
, pagefunc
, pagesize
, True)
2874 self
._pagecount
= pagecount
2876 def _getslice(self
, start
, end
):
2877 start_page
= start
// self
._pagesize
2878 end_page
= self
._pagecount
if end
is None else min(self
._pagecount
, end
// self
._pagesize
+ 1)
2879 skip_elems
= start
- start_page
* self
._pagesize
2880 only_more
= None if end
is None else end
- start
2881 for pagenum
in range(start_page
, end_page
):
2882 page_results
= self
.getpage(pagenum
)
2884 page_results
= page_results
[skip_elems
:]
2886 if only_more
is not None:
2887 if len(page_results
) < only_more
:
2888 only_more
-= len(page_results
)
2890 yield from page_results
[:only_more
]
2892 yield from page_results
2895 class PlaylistEntries
:
2896 MissingEntry
= object()
2897 is_exhausted
= False
2899 def __init__(self
, ydl
, info_dict
):
2902 # _entries must be assigned now since infodict can change during iteration
2903 entries
= info_dict
.get('entries')
2905 raise EntryNotInPlaylist('There are no entries')
2906 elif isinstance(entries
, list):
2907 self
.is_exhausted
= True
2909 requested_entries
= info_dict
.get('requested_entries')
2910 self
.is_incomplete
= bool(requested_entries
)
2911 if self
.is_incomplete
:
2912 assert self
.is_exhausted
2913 self
._entries
= [self
.MissingEntry
] * max(requested_entries
)
2914 for i
, entry
in zip(requested_entries
, entries
):
2915 self
._entries
[i
- 1] = entry
2916 elif isinstance(entries
, (list, PagedList
, LazyList
)):
2917 self
._entries
= entries
2919 self
._entries
= LazyList(entries
)
2921 PLAYLIST_ITEMS_RE
= re
.compile(r
'''(?x)
2922 (?P<start>[+-]?\d+)?
2924 (?P<end>[+-]?\d+|inf(?:inite)?)?
2925 (?::(?P<step>[+-]?\d+))?
2929 def parse_playlist_items(cls
, string
):
2930 for segment
in string
.split(','):
2932 raise ValueError('There is two or more consecutive commas')
2933 mobj
= cls
.PLAYLIST_ITEMS_RE
.fullmatch(segment
)
2935 raise ValueError(f
'{segment!r} is not a valid specification')
2936 start
, end
, step
, has_range
= mobj
.group('start', 'end', 'step', 'range')
2937 if int_or_none(step
) == 0:
2938 raise ValueError(f
'Step in {segment!r} cannot be zero')
2939 yield slice(int_or_none(start
), float_or_none(end
), int_or_none(step
)) if has_range
else int(start
)
2941 def get_requested_items(self
):
2942 playlist_items
= self
.ydl
.params
.get('playlist_items')
2943 playlist_start
= self
.ydl
.params
.get('playliststart', 1)
2944 playlist_end
= self
.ydl
.params
.get('playlistend')
2945 # For backwards compatibility, interpret -1 as whole list
2946 if playlist_end
in (-1, None):
2948 if not playlist_items
:
2949 playlist_items
= f
'{playlist_start}:{playlist_end}'
2950 elif playlist_start
!= 1 or playlist_end
:
2951 self
.ydl
.report_warning('Ignoring playliststart and playlistend because playlistitems was given', only_once
=True)
2953 for index
in self
.parse_playlist_items(playlist_items
):
2954 for i
, entry
in self
[index
]:
2959 # TODO: Add auto-generated fields
2960 self
.ydl
._match
_entry
(entry
, incomplete
=True, silent
=True)
2961 except (ExistingVideoReached
, RejectedVideoReached
):
2964 def get_full_count(self
):
2965 if self
.is_exhausted
and not self
.is_incomplete
:
2967 elif isinstance(self
._entries
, InAdvancePagedList
):
2968 if self
._entries
._pagesize
== 1:
2969 return self
._entries
._pagecount
2971 @functools.cached_property
2973 if isinstance(self
._entries
, list):
2976 entry
= self
._entries
[i
]
2978 entry
= self
.MissingEntry
2979 if not self
.is_incomplete
:
2980 raise self
.IndexError()
2981 if entry
is self
.MissingEntry
:
2982 raise EntryNotInPlaylist(f
'Entry {i} cannot be found')
2987 return type(self
.ydl
)._handle
_extraction
_exceptions
(lambda _
, i
: self
._entries
[i
])(self
.ydl
, i
)
2988 except (LazyList
.IndexError, PagedList
.IndexError):
2989 raise self
.IndexError()
2992 def __getitem__(self
, idx
):
2993 if isinstance(idx
, int):
2994 idx
= slice(idx
, idx
)
2996 # NB: PlaylistEntries[1:10] => (0, 1, ... 9)
2997 step
= 1 if idx
.step
is None else idx
.step
2998 if idx
.start
is None:
2999 start
= 0 if step
> 0 else len(self
) - 1
3001 start
= idx
.start
- 1 if idx
.start
>= 0 else len(self
) + idx
.start
3003 # NB: Do not call len(self) when idx == [:]
3004 if idx
.stop
is None:
3005 stop
= 0 if step
< 0 else float('inf')
3007 stop
= idx
.stop
- 1 if idx
.stop
>= 0 else len(self
) + idx
.stop
3008 stop
+= [-1, 1][step
> 0]
3010 for i
in frange(start
, stop
, step
):
3014 entry
= self
._getter
(i
)
3015 except self
.IndexError:
3016 self
.is_exhausted
= True
3023 return len(tuple(self
[:]))
3025 class IndexError(IndexError):
3029 def uppercase_escape(s
):
3030 unicode_escape
= codecs
.getdecoder('unicode_escape')
3032 r
'\\U[0-9a-fA-F]{8}',
3033 lambda m
: unicode_escape(m
.group(0))[0],
3037 def lowercase_escape(s
):
3038 unicode_escape
= codecs
.getdecoder('unicode_escape')
3040 r
'\\u[0-9a-fA-F]{4}',
3041 lambda m
: unicode_escape(m
.group(0))[0],
3045 def escape_rfc3986(s
):
3046 """Escape non-ASCII characters as suggested by RFC 3986"""
3047 return urllib
.parse
.quote(s
, b
"%/;:@&=+$,!~*'()?#[]")
3050 def escape_url(url
):
3051 """Escape URL as suggested by RFC 3986"""
3052 url_parsed
= urllib
.parse
.urlparse(url
)
3053 return url_parsed
._replace
(
3054 netloc
=url_parsed
.netloc
.encode('idna').decode('ascii'),
3055 path
=escape_rfc3986(url_parsed
.path
),
3056 params
=escape_rfc3986(url_parsed
.params
),
3057 query
=escape_rfc3986(url_parsed
.query
),
3058 fragment
=escape_rfc3986(url_parsed
.fragment
)
3063 return urllib
.parse
.parse_qs(urllib
.parse
.urlparse(url
).query
)
3066 def read_batch_urls(batch_fd
):
3068 if not isinstance(url
, str):
3069 url
= url
.decode('utf-8', 'replace')
3070 BOM_UTF8
= ('\xef\xbb\xbf', '\ufeff')
3071 for bom
in BOM_UTF8
:
3072 if url
.startswith(bom
):
3073 url
= url
[len(bom
):]
3075 if not url
or url
.startswith(('#', ';', ']')):
3077 # "#" cannot be stripped out since it is part of the URI
3078 # However, it can be safely stripped out if following a whitespace
3079 return re
.split(r
'\s#', url
, 1)[0].rstrip()
3081 with contextlib
.closing(batch_fd
) as fd
:
3082 return [url
for url
in map(fixup
, fd
) if url
]
3085 def urlencode_postdata(*args
, **kargs
):
3086 return urllib
.parse
.urlencode(*args
, **kargs
).encode('ascii')
3089 def update_url_query(url
, query
):
3092 parsed_url
= urllib
.parse
.urlparse(url
)
3093 qs
= urllib
.parse
.parse_qs(parsed_url
.query
)
3095 return urllib
.parse
.urlunparse(parsed_url
._replace
(
3096 query
=urllib
.parse
.urlencode(qs
, True)))
3099 def update_Request(req
, url
=None, data
=None, headers
=None, query
=None):
3100 req_headers
= req
.headers
.copy()
3101 req_headers
.update(headers
or {})
3102 req_data
= data
or req
.data
3103 req_url
= update_url_query(url
or req
.get_full_url(), query
)
3104 req_get_method
= req
.get_method()
3105 if req_get_method
== 'HEAD':
3106 req_type
= HEADRequest
3107 elif req_get_method
== 'PUT':
3108 req_type
= PUTRequest
3110 req_type
= urllib
.request
.Request
3112 req_url
, data
=req_data
, headers
=req_headers
,
3113 origin_req_host
=req
.origin_req_host
, unverifiable
=req
.unverifiable
)
3114 if hasattr(req
, 'timeout'):
3115 new_req
.timeout
= req
.timeout
3119 def _multipart_encode_impl(data
, boundary
):
3120 content_type
= 'multipart/form-data; boundary=%s' % boundary
3123 for k
, v
in data
.items():
3124 out
+= b
'--' + boundary
.encode('ascii') + b
'\r\n'
3125 if isinstance(k
, str):
3127 if isinstance(v
, str):
3129 # RFC 2047 requires non-ASCII field names to be encoded, while RFC 7578
3130 # suggests sending UTF-8 directly. Firefox sends UTF-8, too
3131 content
= b
'Content-Disposition: form-data; name="' + k
+ b
'"\r\n\r\n' + v
+ b
'\r\n'
3132 if boundary
.encode('ascii') in content
:
3133 raise ValueError('Boundary overlaps with data')
3136 out
+= b
'--' + boundary
.encode('ascii') + b
'--\r\n'
3138 return out
, content_type
3141 def multipart_encode(data
, boundary
=None):
3143 Encode a dict to RFC 7578-compliant form-data
3146 A dict where keys and values can be either Unicode or bytes-like
3149 If specified a Unicode object, it's used as the boundary. Otherwise
3150 a random boundary is generated.
3152 Reference: https://tools.ietf.org/html/rfc7578
3154 has_specified_boundary
= boundary
is not None
3157 if boundary
is None:
3158 boundary
= '---------------' + str(random
.randrange(0x0fffffff, 0xffffffff))
3161 out
, content_type
= _multipart_encode_impl(data
, boundary
)
3164 if has_specified_boundary
:
3168 return out
, content_type
3171 def dict_get(d
, key_or_keys
, default
=None, skip_false_values
=True):
3172 for val
in map(d
.get
, variadic(key_or_keys
)):
3173 if val
is not None and (val
or not skip_false_values
):
3178 def try_call(*funcs
, expected_type
=None, args
=[], kwargs
={}):
3181 val
= f(*args
, **kwargs
)
3182 except (AttributeError, KeyError, TypeError, IndexError, ZeroDivisionError):
3185 if expected_type
is None or isinstance(val
, expected_type
):
3189 def try_get(src
, getter
, expected_type
=None):
3190 return try_call(*variadic(getter
), args
=(src
,), expected_type
=expected_type
)
3193 def filter_dict(dct
, cndn
=lambda _
, v
: v
is not None):
3194 return {k: v for k, v in dct.items() if cndn(k, v)}
3197 def merge_dicts(*dicts
):
3199 for a_dict
in dicts
:
3200 for k
, v
in a_dict
.items():
3201 if (v
is not None and k
not in merged
3202 or isinstance(v
, str) and merged
[k
] == ''):
3207 def encode_compat_str(string
, encoding
=preferredencoding(), errors
='strict'):
3208 return string
if isinstance(string
, str) else str(string
, encoding
, errors
)
3220 TV_PARENTAL_GUIDELINES
= {
3230 def parse_age_limit(s
):
3231 # isinstance(False, int) is True. So type() must be used instead
3232 if type(s
) is int: # noqa: E721
3233 return s
if 0 <= s
<= 21 else None
3234 elif not isinstance(s
, str):
3236 m
= re
.match(r
'^(?P<age>\d{1,2})\+?$', s
)
3238 return int(m
.group('age'))
3241 return US_RATINGS
[s
]
3242 m
= re
.match(r
'^TV[_-]?(%s)$' % '|'.join(k
[3:] for k
in TV_PARENTAL_GUIDELINES
), s
)
3244 return TV_PARENTAL_GUIDELINES
['TV-' + m
.group(1)]
3248 def strip_jsonp(code
):
3251 (?:window\.)?(?P<func_name>[a-zA-Z0-9_.$]*)
3252 (?:\s*&&\s*(?P=func_name))?
3253 \s*\(\s*(?P<callback_data>.*)\);?
3254 \s*?(?://[^\n]*)*$''',
3255 r
'\g<callback_data>', code
)
3258 def js_to_json(code
, vars={}, *, strict
=False):
3259 # vars is a dict of var, val pairs to substitute
3260 COMMENT_RE
= r
'/\*(?:(?!\*/).)*?\*/|//[^\n]*\n'
3261 SKIP_RE
= fr
'\s*(?:{COMMENT_RE})?\s*'
3263 (fr
'(?s)^(0[xX][0-9a-fA-F]+){SKIP_RE}:?$', 16),
3264 (fr
'(?s)^(0+[0-7]+){SKIP_RE}:?$', 8),
3269 if v
in ('true', 'false', 'null'):
3271 elif v
in ('undefined', 'void 0'):
3273 elif v
.startswith('/*') or v
.startswith('//') or v
.startswith('!') or v
== ',':
3276 if v
[0] in ("'", '"'):
3277 v
= re
.sub(r
'(?s)\\.|"', lambda m
: {
3282 }.get(m
.group(0), m
.group(0)), v
[1:-1])
3284 for regex
, base
in INTEGER_TABLE
:
3285 im
= re
.match(regex
, v
)
3287 i
= int(im
.group(1), base
)
3288 return '"%d":' % i
if v
.endswith(':') else '%d' % i
3293 raise ValueError(f
'Unknown value: {v}')
3297 def create_map(mobj
):
3298 return json
.dumps(dict(json
.loads(js_to_json(mobj
.group(1) or '[]', vars=vars))))
3300 code
= re
.sub(r
'new Map\((\[.*?\])?\)', create_map
, code
)
3302 code
= re
.sub(r
'new Date\((".+")\)', r
'\g<1>', code
)
3304 return re
.sub(r
'''(?sx)
3305 "(?:[^"\\]*(?:\\\\|\\['"nurtbfx/\n]))*[^"\\]*"|
3306 '(?:[^'\\]*(?:\\\\|\\['"nurtbfx/\n]))*[^'\\]*'|
3307 {comment}|,(?={skip}[\]}}])|
3308 void\s0|(?:(?<![0-9])[eE]|[a-df-zA-DF-Z_$])[.a-zA-Z_$0-9]*|
3309 \b(?:0[xX][0-9a-fA-F]+|0+[0-7]+)(?:{skip}:)?|
3312 '''.format(comment
=COMMENT_RE
, skip
=SKIP_RE
), fix_kv
, code
)
3315 def qualities(quality_ids
):
3316 """ Get a numeric quality value out of a list of possible values """
3319 return quality_ids
.index(qid
)
3325 POSTPROCESS_WHEN
= ('pre_process', 'after_filter', 'before_dl', 'post_process', 'after_move', 'after_video', 'playlist')
3329 'default': '%(title)s [%(id)s].%(ext)s',
3330 'chapter': '%(title)s - %(section_number)03d %(section_title)s [%(id)s].%(ext)s',
3336 'description': 'description',
3337 'annotation': 'annotations.xml',
3338 'infojson': 'info.json',
3341 'pl_thumbnail': None,
3342 'pl_description': 'description',
3343 'pl_infojson': 'info.json',
3346 # As of [1] format syntax is:
3347 # %[mapping_key][conversion_flags][minimum_width][.precision][length_modifier]type
3348 # 1. https://docs.python.org/2/library/stdtypes.html#string-formatting
3349 STR_FORMAT_RE_TMPL
= r
'''(?x)
3350 (?<!%)(?P<prefix>(?:%%)*)
3352 (?P<has_key>\((?P<key>{0})\))?
3354 (?P<conversion>[#0\-+ ]+)?
3356 (?P<precision>\.\d+)?
3357 (?P<len_mod>[hlL])? # unused in python
3358 {1} # conversion type
3363 STR_FORMAT_TYPES
= 'diouxXeEfFgGcrs'
3366 def limit_length(s
, length
):
3367 """ Add ellipses to overly long strings """
3372 return s
[:length
- len(ELLIPSES
)] + ELLIPSES
3376 def version_tuple(v
):
3377 return tuple(int(e
) for e
in re
.split(r
'[-.]', v
))
3380 def is_outdated_version(version
, limit
, assume_new
=True):
3382 return not assume_new
3384 return version_tuple(version
) < version_tuple(limit
)
3386 return not assume_new
3389 def ytdl_is_updateable():
3390 """ Returns if yt-dlp can be updated with -U """
3392 from .update
import is_non_updateable
3394 return not is_non_updateable()
3397 def args_to_str(args
):
3398 # Get a short string representation for a subprocess command
3399 return ' '.join(compat_shlex_quote(a
) for a
in args
)
3402 def error_to_compat_str(err
):
3406 def error_to_str(err
):
3407 return f
'{type(err).__name__}: {err}'
3410 def mimetype2ext(mt
):
3414 mt
, _
, params
= mt
.partition(';')
3419 # Per RFC 3003, audio/mpeg can be .mp1, .mp2 or .mp3. Here use .mp3 as
3420 # it's the most popular one
3421 'audio/mpeg': 'mp3',
3422 'audio/x-wav': 'wav',
3424 'audio/wave': 'wav',
3427 ext
= FULL_MAP
.get(mt
)
3433 'smptett+xml': 'tt',
3437 'x-mp4-fragmented': 'mp4',
3438 'x-ms-sami': 'sami',
3441 'x-mpegurl': 'm3u8',
3442 'vnd.apple.mpegurl': 'm3u8',
3446 'vnd.ms-sstr+xml': 'ism',
3450 'filmstrip+json': 'fs',
3454 _
, _
, subtype
= mt
.rpartition('/')
3455 ext
= SUBTYPE_MAP
.get(subtype
.lower())
3466 _
, _
, suffix
= subtype
.partition('+')
3467 ext
= SUFFIX_MAP
.get(suffix
)
3471 return subtype
.replace('+', '.')
3474 def ext2mimetype(ext_or_url
):
3477 if '.' not in ext_or_url
:
3478 ext_or_url
= f
'file.{ext_or_url}'
3479 return mimetypes
.guess_type(ext_or_url
)[0]
3482 def parse_codecs(codecs_str
):
3483 # http://tools.ietf.org/html/rfc6381
3486 split_codecs
= list(filter(None, map(
3487 str.strip
, codecs_str
.strip().strip(',').split(','))))
3488 vcodec
, acodec
, scodec
, hdr
= None, None, None, None
3489 for full_codec
in split_codecs
:
3490 parts
= re
.sub(r
'0+(?=\d)', '', full_codec
).split('.')
3491 if parts
[0] in ('avc1', 'avc2', 'avc3', 'avc4', 'vp9', 'vp8', 'hev1', 'hev2',
3492 'h263', 'h264', 'mp4v', 'hvc1', 'av1', 'theora', 'dvh1', 'dvhe'):
3496 if parts
[0] in ('dvh1', 'dvhe'):
3498 elif parts
[0] == 'av1' and traverse_obj(parts
, 3) == '10':
3500 elif parts
[:2] == ['vp9', '2']:
3502 elif parts
[0] in ('flac', 'mp4a', 'opus', 'vorbis', 'mp3', 'aac',
3503 'ac-3', 'ec-3', 'eac3', 'dtsc', 'dtse', 'dtsh', 'dtsl'):
3504 acodec
= acodec
or full_codec
3505 elif parts
[0] in ('stpp', 'wvtt'):
3506 scodec
= scodec
or full_codec
3508 write_string(f
'WARNING: Unknown codec {full_codec}\n')
3509 if vcodec
or acodec
or scodec
:
3511 'vcodec': vcodec
or 'none',
3512 'acodec': acodec
or 'none',
3513 'dynamic_range': hdr
,
3514 **({'scodec': scodec}
if scodec
is not None else {}),
3516 elif len(split_codecs
) == 2:
3518 'vcodec': split_codecs
[0],
3519 'acodec': split_codecs
[1],
3524 def get_compatible_ext(*, vcodecs
, acodecs
, vexts
, aexts
, preferences
=None):
3525 assert len(vcodecs
) == len(vexts
) and len(acodecs
) == len(aexts
)
3527 allow_mkv
= not preferences
or 'mkv' in preferences
3529 if allow_mkv
and max(len(acodecs
), len(vcodecs
)) > 1:
3530 return 'mkv' # TODO: any other format allows this?
3532 # TODO: All codecs supported by parse_codecs isn't handled here
3533 COMPATIBLE_CODECS
= {
3535 'av1', 'hevc', 'avc1', 'mp4a', # fourcc (m3u8, mpd)
3536 'h264', 'aacl', # Set in ISM
3539 'av1', 'vp9', 'vp8', 'opus', 'vrbs',
3540 'vp9x', 'vp8x', # in the webm spec
3544 sanitize_codec
= functools
.partial(try_get
, getter
=lambda x
: x
[0].split('.')[0].replace('0', ''))
3545 vcodec
, acodec
= sanitize_codec(vcodecs
), sanitize_codec(acodecs
)
3547 for ext
in preferences
or COMPATIBLE_CODECS
.keys():
3548 codec_set
= COMPATIBLE_CODECS
.get(ext
, set())
3549 if ext
== 'mkv' or codec_set
.issuperset((vcodec
, acodec
)):
3553 {'mp3', 'mp4', 'm4a', 'm4p', 'm4b', 'm4r', 'm4v', 'ismv', 'isma', 'mov'}
,
3556 for ext
in preferences
or vexts
:
3557 current_exts
= {ext, *vexts, *aexts}
3558 if ext
== 'mkv' or current_exts
== {ext}
or any(
3559 ext_sets
.issuperset(current_exts
) for ext_sets
in COMPATIBLE_EXTS
):
3561 return 'mkv' if allow_mkv
else preferences
[-1]
3564 def urlhandle_detect_ext(url_handle
):
3565 getheader
= url_handle
.headers
.get
3567 cd
= getheader('Content-Disposition')
3569 m
= re
.match(r
'attachment;\s*filename="(?P<filename>[^"]+)"', cd
)
3571 e
= determine_ext(m
.group('filename'), default_ext
=None)
3575 return mimetype2ext(getheader('Content-Type'))
3578 def encode_data_uri(data
, mime_type
):
3579 return 'data:%s;base64,%s' % (mime_type
, base64
.b64encode(data
).decode('ascii'))
3582 def age_restricted(content_limit
, age_limit
):
3583 """ Returns True iff the content should be blocked """
3585 if age_limit
is None: # No limit set
3587 if content_limit
is None:
3588 return False # Content available for everyone
3589 return age_limit
< content_limit
3592 # List of known byte-order-marks (BOM)
3594 (b
'\xef\xbb\xbf', 'utf-8'),
3595 (b
'\x00\x00\xfe\xff', 'utf-32-be'),
3596 (b
'\xff\xfe\x00\x00', 'utf-32-le'),
3597 (b
'\xff\xfe', 'utf-16-le'),
3598 (b
'\xfe\xff', 'utf-16-be'),
3602 def is_html(first_bytes
):
3603 """ Detect whether a file contains HTML by examining its first bytes. """
3606 for bom
, enc
in BOMS
:
3607 while first_bytes
.startswith(bom
):
3608 encoding
, first_bytes
= enc
, first_bytes
[len(bom
):]
3610 return re
.match(r
'^\s*<', first_bytes
.decode(encoding
, 'replace'))
3613 def determine_protocol(info_dict
):
3614 protocol
= info_dict
.get('protocol')
3615 if protocol
is not None:
3618 url
= sanitize_url(info_dict
['url'])
3619 if url
.startswith('rtmp'):
3621 elif url
.startswith('mms'):
3623 elif url
.startswith('rtsp'):
3626 ext
= determine_ext(url
)
3632 return urllib
.parse
.urlparse(url
).scheme
3635 def render_table(header_row
, data
, delim
=False, extra_gap
=0, hide_empty
=False):
3636 """ Render a list of rows, each as a list of values.
3637 Text after a \t will be right aligned """
3639 return len(remove_terminal_sequences(string
).replace('\t', ''))
3641 def get_max_lens(table
):
3642 return [max(width(str(v
)) for v
in col
) for col
in zip(*table
)]
3644 def filter_using_list(row
, filterArray
):
3645 return [col
for take
, col
in itertools
.zip_longest(filterArray
, row
, fillvalue
=True) if take
]
3647 max_lens
= get_max_lens(data
) if hide_empty
else []
3648 header_row
= filter_using_list(header_row
, max_lens
)
3649 data
= [filter_using_list(row
, max_lens
) for row
in data
]
3651 table
= [header_row
] + data
3652 max_lens
= get_max_lens(table
)
3655 table
= [header_row
, [delim
* (ml
+ extra_gap
) for ml
in max_lens
]] + data
3656 table
[1][-1] = table
[1][-1][:-extra_gap
* len(delim
)] # Remove extra_gap from end of delimiter
3658 for pos
, text
in enumerate(map(str, row
)):
3660 row
[pos
] = text
.replace('\t', ' ' * (max_lens
[pos
] - width(text
))) + ' ' * extra_gap
3662 row
[pos
] = text
+ ' ' * (max_lens
[pos
] - width(text
) + extra_gap
)
3663 ret
= '\n'.join(''.join(row
).rstrip() for row
in table
)
3667 def _match_one(filter_part
, dct
, incomplete
):
3668 # TODO: Generalize code with YoutubeDL._build_format_filter
3669 STRING_OPERATORS
= {
3670 '*=': operator
.contains
,
3671 '^=': lambda attr
, value
: attr
.startswith(value
),
3672 '$=': lambda attr
, value
: attr
.endswith(value
),
3673 '~=': lambda attr
, value
: re
.search(value
, attr
),
3675 COMPARISON_OPERATORS
= {
3677 '<=': operator
.le
, # "<=" must be defined above "<"
3684 if isinstance(incomplete
, bool):
3685 is_incomplete
= lambda _
: incomplete
3687 is_incomplete
= lambda k
: k
in incomplete
3689 operator_rex
= re
.compile(r
'''(?x)
3691 \s*(?P<negation>!\s*)?(?P<op>%s)(?P<none_inclusive>\s*\?)?\s*
3693 (?P<quote>["\'])(?P
<quotedstrval
>.+?
)(?P
=quote
)|
3696 ''' % '|'.join(map(re.escape, COMPARISON_OPERATORS.keys())))
3697 m = operator_rex.fullmatch(filter_part.strip())
3700 unnegated_op = COMPARISON_OPERATORS[m['op']]
3702 op = lambda attr, value: not unnegated_op(attr, value)
3705 comparison_value = m['quotedstrval'] or m['strval'] or m['intval']
3707 comparison_value = comparison_value.replace(r'\%s' % m['quote'], m['quote'])
3708 actual_value = dct.get(m['key'])
3709 numeric_comparison = None
3710 if isinstance(actual_value, (int, float)):
3711 # If the original field is a string and matching comparisonvalue is
3712 # a number we should respect the origin of the original field
3713 # and process comparison value as a string (see
3714 # https://github.com/ytdl-org/youtube-dl/issues/11082)
3716 numeric_comparison = int(comparison_value)
3718 numeric_comparison = parse_filesize(comparison_value)
3719 if numeric_comparison is None:
3720 numeric_comparison = parse_filesize(f'{comparison_value}B')
3721 if numeric_comparison is None:
3722 numeric_comparison = parse_duration(comparison_value)
3723 if numeric_comparison is not None and m['op'] in STRING_OPERATORS:
3724 raise ValueError('Operator %s only supports string values!' % m['op'])
3725 if actual_value is None:
3726 return is_incomplete(m['key']) or m['none_inclusive']
3727 return op(actual_value, comparison_value if numeric_comparison is None else numeric_comparison)
3730 '': lambda v: (v is True) if isinstance(v, bool) else (v is not None),
3731 '!': lambda v: (v is False) if isinstance(v, bool) else (v is None),
3733 operator_rex = re.compile(r'''(?x
)
3734 (?P
<op
>%s)\s
*(?P
<key
>[a
-z_
]+)
3735 ''' % '|'.join(map(re.escape, UNARY_OPERATORS.keys())))
3736 m = operator_rex.fullmatch(filter_part.strip())
3738 op = UNARY_OPERATORS[m.group('op')]
3739 actual_value = dct.get(m.group('key'))
3740 if is_incomplete(m.group('key')) and actual_value is None:
3742 return op(actual_value)
3744 raise ValueError('Invalid filter part %r' % filter_part)
3747 def match_str(filter_str, dct, incomplete=False):
3748 """ Filter a dictionary with a simple string syntax.
3749 @returns Whether the filter passes
3750 @param incomplete Set of keys that is expected to be missing from dct.
3751 Can be True/False to indicate all/none of the keys may be missing.
3752 All conditions on incomplete keys pass if the key is missing
3755 _match_one(filter_part.replace(r'\&', '&'), dct, incomplete)
3756 for filter_part in re.split(r'(?<!\\)&', filter_str))
3759 def match_filter_func(filters):
3762 filters = set(variadic(filters))
3764 interactive = '-' in filters
3768 def _match_func(info_dict, incomplete=False):
3769 if not filters or any(match_str(f, info_dict, incomplete) for f in filters):
3770 return NO_DEFAULT if interactive and not incomplete else None
3772 video_title = info_dict.get('title') or info_dict.get('id') or 'entry'
3773 filter_str = ') | ('.join(map(str.strip, filters))
3774 return f'{video_title} does not pass filter ({filter_str}), skipping ..'
3778 class download_range_func:
3779 def __init__(self, chapters, ranges):
3780 self.chapters, self.ranges = chapters, ranges
3782 def __call__(self, info_dict, ydl):
3783 warning = ('There are no chapters matching the regex' if info_dict.get('chapters')
3784 else 'Cannot match chapters since chapter information is unavailable')
3785 for regex in self.chapters or []:
3786 for i, chapter in enumerate(info_dict.get('chapters') or []):
3787 if re.search(regex, chapter['title']):
3789 yield {**chapter, 'index': i}
3790 if self.chapters and warning:
3791 ydl.to_screen(f'[info] {info_dict["id"]}: {warning}')
3793 yield from ({'start_time': start, 'end_time': end} for start, end in self.ranges or [])
3795 def __eq__(self, other):
3796 return (isinstance(other, download_range_func)
3797 and self.chapters == other.chapters and self.ranges == other.ranges)
3800 def parse_dfxp_time_expr(time_expr):
3804 mobj = re.match(rf'^(?P<time_offset>{NUMBER_RE})s?$', time_expr)
3806 return float(mobj.group('time_offset'))
3808 mobj = re.match(r'^(\d+):(\d\d):(\d\d(?:(?:\.|:)\d+)?)$', time_expr)
3810 return 3600 * int(mobj.group(1)) + 60 * int(mobj.group(2)) + float(mobj.group(3).replace(':', '.'))
3813 def srt_subtitles_timecode(seconds):
3814 return '%02d:%02d:%02d,%03d' % timetuple_from_msec(seconds * 1000)
3817 def ass_subtitles_timecode(seconds):
3818 time = timetuple_from_msec(seconds * 1000)
3819 return '%01d:%02d:%02d.%02d' % (*time[:-1], time.milliseconds / 10)
3822 def dfxp2srt(dfxp_data):
3824 @param dfxp_data A
bytes-like
object containing DFXP data
3825 @returns A
unicode object containing converted SRT data
3827 LEGACY_NAMESPACES = (
3828 (b'http://www.w3.org/ns/ttml', [
3829 b'http://www.w3.org/2004/11/ttaf1',
3830 b'http://www.w3.org/2006/04/ttaf1',
3831 b'http://www.w3.org/2006/10/ttaf1',
3833 (b'http://www.w3.org/ns/ttml#styling', [
3834 b'http://www.w3.org/ns/ttml#style',
3838 SUPPORTED_STYLING = [
3847 _x = functools.partial(xpath_with_ns, ns_map={
3848 'xml': 'http://www.w3.org/XML/1998/namespace',
3849 'ttml': 'http://www.w3.org/ns/ttml',
3850 'tts': 'http://www.w3.org/ns/ttml#styling',
3856 class TTMLPElementParser:
3858 _unclosed_elements = []
3859 _applied_styles = []
3861 def start(self, tag, attrib):
3862 if tag in (_x('ttml:br'), 'br'):
3865 unclosed_elements = []
3867 element_style_id = attrib.get('style')
3869 style.update(default_style)
3870 if element_style_id:
3871 style.update(styles.get(element_style_id, {}))
3872 for prop in SUPPORTED_STYLING:
3873 prop_val = attrib.get(_x('tts:' + prop))
3875 style[prop] = prop_val
3878 for k, v in sorted(style.items()):
3879 if self._applied_styles and self._applied_styles[-1].get(k) == v:
3882 font += ' color="%s"' % v
3883 elif k == 'fontSize':
3884 font += ' size="%s"' % v
3885 elif k == 'fontFamily':
3886 font += ' face="%s"' % v
3887 elif k == 'fontWeight' and v == 'bold':
3889 unclosed_elements.append('b')
3890 elif k == 'fontStyle' and v == 'italic':
3892 unclosed_elements.append('i')
3893 elif k == 'textDecoration' and v == 'underline':
3895 unclosed_elements.append('u')
3897 self._out += '<font' + font + '>'
3898 unclosed_elements.append('font')
3900 if self._applied_styles:
3901 applied_style.update(self._applied_styles[-1])
3902 applied_style.update(style)
3903 self._applied_styles.append(applied_style)
3904 self._unclosed_elements.append(unclosed_elements)
3907 if tag not in (_x('ttml:br'), 'br'):
3908 unclosed_elements = self._unclosed_elements.pop()
3909 for element in reversed(unclosed_elements):
3910 self._out += '</%s>' % element
3911 if unclosed_elements and self._applied_styles:
3912 self._applied_styles.pop()
3914 def data(self, data):
3918 return self._out.strip()
3920 def parse_node(node):
3921 target = TTMLPElementParser()
3922 parser = xml.etree.ElementTree.XMLParser(target=target)
3923 parser.feed(xml.etree.ElementTree.tostring(node))
3924 return parser.close()
3926 for k, v in LEGACY_NAMESPACES:
3928 dfxp_data = dfxp_data.replace(ns, k)
3930 dfxp = compat_etree_fromstring(dfxp_data)
3932 paras = dfxp.findall(_x('.//ttml:p')) or dfxp.findall('.//p')
3935 raise ValueError('Invalid dfxp/TTML subtitle')
3939 for style in dfxp.findall(_x('.//ttml:style')):
3940 style_id = style.get('id') or style.get(_x('xml:id'))
3943 parent_style_id = style.get('style')
3945 if parent_style_id not in styles:
3948 styles[style_id] = styles[parent_style_id].copy()
3949 for prop in SUPPORTED_STYLING:
3950 prop_val = style.get(_x('tts:' + prop))
3952 styles.setdefault(style_id, {})[prop] = prop_val
3958 for p in ('body', 'div'):
3959 ele = xpath_element(dfxp, [_x('.//ttml:' + p), './/' + p])
3962 style = styles.get(ele.get('style'))
3965 default_style.update(style)
3967 for para, index in zip(paras, itertools.count(1)):
3968 begin_time = parse_dfxp_time_expr(para.attrib.get('begin'))
3969 end_time = parse_dfxp_time_expr(para.attrib.get('end'))
3970 dur = parse_dfxp_time_expr(para.attrib.get('dur'))
3971 if begin_time is None:
3976 end_time = begin_time + dur
3977 out.append('%d\n%s --> %s\n%s\n\n' % (
3979 srt_subtitles_timecode(begin_time),
3980 srt_subtitles_timecode(end_time),
3986 def cli_option(params, command_option, param, separator=None):
3987 param = params.get(param)
3988 return ([] if param is None
3989 else [command_option, str(param)] if separator is None
3990 else [f'{command_option}{separator}{param}'])
3993 def cli_bool_option(params, command_option, param, true_value='true', false_value='false', separator=None):
3994 param = params.get(param)
3995 assert param in (True, False, None)
3996 return cli_option({True: true_value, False: false_value}, command_option, param, separator)
3999 def cli_valueless_option(params, command_option, param, expected_value=True):
4000 return [command_option] if params.get(param) == expected_value else []
4003 def cli_configuration_args(argdict, keys, default=[], use_compat=True):
4004 if isinstance(argdict, (list, tuple)): # for backward compatibility
4011 assert isinstance(argdict, dict)
4013 assert isinstance(keys, (list, tuple))
4014 for key_list in keys:
4015 arg_list = list(filter(
4016 lambda x: x is not None,
4017 [argdict.get(key.lower()) for key in variadic(key_list)]))
4019 return [arg for args in arg_list for arg in args]
4023 def _configuration_args(main_key, argdict, exe, keys=None, default=[], use_compat=True):
4024 main_key, exe = main_key.lower(), exe.lower()
4025 root_key = exe if main_key == exe else f'{main_key}+{exe}'
4026 keys = [f'{root_key}{k}' for k in (keys or [''])]
4027 if root_key in keys:
4029 keys.append((main_key, exe))
4030 keys.append('default')
4033 return cli_configuration_args(argdict, keys, default, use_compat)
4037 # See http://www.loc.gov/standards/iso639-2/ISO-639-2_utf-8.txt
4096 'iw': 'heb', # Replaced by he in 1989 revision
4106 'in': 'ind', # Replaced by id in 1989 revision
4221 'ji': 'yid', # Replaced by yi in 1989 revision
4229 def short2long(cls, code):
4230 """Convert language code from ISO 639-1 to ISO 639-2/T"""
4231 return cls._lang_map.get(code[:2])
4234 def long2short(cls, code):
4235 """Convert language code from ISO 639-2/T to ISO 639-1"""
4236 for short_name, long_name in cls._lang_map.items():
4237 if long_name == code:
4242 # From http://data.okfn.org/data/core/country-list
4244 'AF': 'Afghanistan',
4245 'AX': 'Åland Islands',
4248 'AS': 'American Samoa',
4253 'AG': 'Antigua and Barbuda',
4270 'BO': 'Bolivia, Plurinational State of',
4271 'BQ': 'Bonaire, Sint Eustatius and Saba',
4272 'BA': 'Bosnia and Herzegovina',
4274 'BV': 'Bouvet Island',
4276 'IO': 'British Indian Ocean Territory',
4277 'BN': 'Brunei Darussalam',
4279 'BF': 'Burkina Faso',
4285 'KY': 'Cayman Islands',
4286 'CF': 'Central African Republic',
4290 'CX': 'Christmas Island',
4291 'CC': 'Cocos (Keeling) Islands',
4295 'CD': 'Congo, the Democratic Republic of the',
4296 'CK': 'Cook Islands',
4298 'CI': 'Côte d\'Ivoire',
4303 'CZ': 'Czech Republic',
4307 'DO': 'Dominican Republic',
4310 'SV': 'El Salvador',
4311 'GQ': 'Equatorial Guinea',
4315 'FK': 'Falkland Islands (Malvinas)',
4316 'FO': 'Faroe Islands',
4320 'GF': 'French Guiana',
4321 'PF': 'French Polynesia',
4322 'TF': 'French Southern Territories',
4337 'GW': 'Guinea-Bissau',
4340 'HM': 'Heard Island and McDonald Islands',
4341 'VA': 'Holy See (Vatican City State)',
4348 'IR': 'Iran, Islamic Republic of',
4351 'IM': 'Isle of Man',
4361 'KP': 'Korea, Democratic People\'s Republic of',
4362 'KR': 'Korea, Republic of',
4365 'LA': 'Lao People\'s Democratic Republic',
4371 'LI': 'Liechtenstein',
4375 'MK': 'Macedonia, the Former Yugoslav Republic of',
4382 'MH': 'Marshall Islands',
4388 'FM': 'Micronesia, Federated States of',
4389 'MD': 'Moldova, Republic of',
4400 'NL': 'Netherlands',
4401 'NC': 'New Caledonia',
4402 'NZ': 'New Zealand',
4407 'NF': 'Norfolk Island',
4408 'MP': 'Northern Mariana Islands',
4413 'PS': 'Palestine, State of',
4415 'PG': 'Papua New Guinea',
4418 'PH': 'Philippines',
4422 'PR': 'Puerto Rico',
4426 'RU': 'Russian Federation',
4428 'BL': 'Saint Barthélemy',
4429 'SH': 'Saint Helena, Ascension and Tristan da Cunha',
4430 'KN': 'Saint Kitts and Nevis',
4431 'LC': 'Saint Lucia',
4432 'MF': 'Saint Martin (French part)',
4433 'PM': 'Saint Pierre and Miquelon',
4434 'VC': 'Saint Vincent and the Grenadines',
4437 'ST': 'Sao Tome and Principe',
4438 'SA': 'Saudi Arabia',
4442 'SL': 'Sierra Leone',
4444 'SX': 'Sint Maarten (Dutch part)',
4447 'SB': 'Solomon Islands',
4449 'ZA': 'South Africa',
4450 'GS': 'South Georgia and the South Sandwich Islands',
4451 'SS': 'South Sudan',
4456 'SJ': 'Svalbard and Jan Mayen',
4459 'CH': 'Switzerland',
4460 'SY': 'Syrian Arab Republic',
4461 'TW': 'Taiwan, Province of China',
4463 'TZ': 'Tanzania, United Republic of',
4465 'TL': 'Timor-Leste',
4469 'TT': 'Trinidad and Tobago',
4472 'TM': 'Turkmenistan',
4473 'TC': 'Turks and Caicos Islands',
4477 'AE': 'United Arab Emirates',
4478 'GB': 'United Kingdom',
4479 'US': 'United States',
4480 'UM': 'United States Minor Outlying Islands',
4484 'VE': 'Venezuela, Bolivarian Republic of',
4486 'VG': 'Virgin Islands, British',
4487 'VI': 'Virgin Islands, U.S.',
4488 'WF': 'Wallis and Futuna',
4489 'EH': 'Western Sahara',
4493 # Not ISO 3166 codes, but used for IP blocks
4494 'AP': 'Asia/Pacific Region',
4499 def short2full(cls, code):
4500 """Convert an ISO 3166-2 country code to the corresponding full name"""
4501 return cls._country_map.get(code.upper())
4505 # Major IPv4 address blocks per country
4507 'AD': '46.172.224.0/19',
4508 'AE': '94.200.0.0/13',
4509 'AF': '149.54.0.0/17',
4510 'AG': '209.59.64.0/18',
4511 'AI': '204.14.248.0/21',
4512 'AL': '46.99.0.0/16',
4513 'AM': '46.70.0.0/15',
4514 'AO': '105.168.0.0/13',
4515 'AP': '182.50.184.0/21',
4516 'AQ': '23.154.160.0/24',
4517 'AR': '181.0.0.0/12',
4518 'AS': '202.70.112.0/20',
4519 'AT': '77.116.0.0/14',
4520 'AU': '1.128.0.0/11',
4521 'AW': '181.41.0.0/18',
4522 'AX': '185.217.4.0/22',
4523 'AZ': '5.197.0.0/16',
4524 'BA': '31.176.128.0/17',
4525 'BB': '65.48.128.0/17',
4526 'BD': '114.130.0.0/16',
4528 'BF': '102.178.0.0/15',
4529 'BG': '95.42.0.0/15',
4530 'BH': '37.131.0.0/17',
4531 'BI': '154.117.192.0/18',
4532 'BJ': '137.255.0.0/16',
4533 'BL': '185.212.72.0/23',
4534 'BM': '196.12.64.0/18',
4535 'BN': '156.31.0.0/16',
4536 'BO': '161.56.0.0/16',
4537 'BQ': '161.0.80.0/20',
4538 'BR': '191.128.0.0/12',
4539 'BS': '24.51.64.0/18',
4540 'BT': '119.2.96.0/19',
4541 'BW': '168.167.0.0/16',
4542 'BY': '178.120.0.0/13',
4543 'BZ': '179.42.192.0/18',
4544 'CA': '99.224.0.0/11',
4545 'CD': '41.243.0.0/16',
4546 'CF': '197.242.176.0/21',
4547 'CG': '160.113.0.0/16',
4548 'CH': '85.0.0.0/13',
4549 'CI': '102.136.0.0/14',
4550 'CK': '202.65.32.0/19',
4551 'CL': '152.172.0.0/14',
4552 'CM': '102.244.0.0/14',
4553 'CN': '36.128.0.0/10',
4554 'CO': '181.240.0.0/12',
4555 'CR': '201.192.0.0/12',
4556 'CU': '152.206.0.0/15',
4557 'CV': '165.90.96.0/19',
4558 'CW': '190.88.128.0/17',
4559 'CY': '31.153.0.0/16',
4560 'CZ': '88.100.0.0/14',
4562 'DJ': '197.241.0.0/17',
4563 'DK': '87.48.0.0/12',
4564 'DM': '192.243.48.0/20',
4565 'DO': '152.166.0.0/15',
4566 'DZ': '41.96.0.0/12',
4567 'EC': '186.68.0.0/15',
4568 'EE': '90.190.0.0/15',
4569 'EG': '156.160.0.0/11',
4570 'ER': '196.200.96.0/20',
4571 'ES': '88.0.0.0/11',
4572 'ET': '196.188.0.0/14',
4573 'EU': '2.16.0.0/13',
4574 'FI': '91.152.0.0/13',
4575 'FJ': '144.120.0.0/16',
4576 'FK': '80.73.208.0/21',
4577 'FM': '119.252.112.0/20',
4578 'FO': '88.85.32.0/19',
4580 'GA': '41.158.0.0/15',
4582 'GD': '74.122.88.0/21',
4583 'GE': '31.146.0.0/16',
4584 'GF': '161.22.64.0/18',
4585 'GG': '62.68.160.0/19',
4586 'GH': '154.160.0.0/12',
4587 'GI': '95.164.0.0/16',
4588 'GL': '88.83.0.0/19',
4589 'GM': '160.182.0.0/15',
4590 'GN': '197.149.192.0/18',
4591 'GP': '104.250.0.0/19',
4592 'GQ': '105.235.224.0/20',
4593 'GR': '94.64.0.0/13',
4594 'GT': '168.234.0.0/16',
4595 'GU': '168.123.0.0/16',
4596 'GW': '197.214.80.0/20',
4597 'GY': '181.41.64.0/18',
4598 'HK': '113.252.0.0/14',
4599 'HN': '181.210.0.0/16',
4600 'HR': '93.136.0.0/13',
4601 'HT': '148.102.128.0/17',
4602 'HU': '84.0.0.0/14',
4603 'ID': '39.192.0.0/10',
4604 'IE': '87.32.0.0/12',
4605 'IL': '79.176.0.0/13',
4606 'IM': '5.62.80.0/20',
4607 'IN': '117.192.0.0/10',
4608 'IO': '203.83.48.0/21',
4609 'IQ': '37.236.0.0/14',
4610 'IR': '2.176.0.0/12',
4611 'IS': '82.221.0.0/16',
4612 'IT': '79.0.0.0/10',
4613 'JE': '87.244.64.0/18',
4614 'JM': '72.27.0.0/17',
4615 'JO': '176.29.0.0/16',
4616 'JP': '133.0.0.0/8',
4617 'KE': '105.48.0.0/12',
4618 'KG': '158.181.128.0/17',
4619 'KH': '36.37.128.0/17',
4620 'KI': '103.25.140.0/22',
4621 'KM': '197.255.224.0/20',
4622 'KN': '198.167.192.0/19',
4623 'KP': '175.45.176.0/22',
4624 'KR': '175.192.0.0/10',
4625 'KW': '37.36.0.0/14',
4626 'KY': '64.96.0.0/15',
4627 'KZ': '2.72.0.0/13',
4628 'LA': '115.84.64.0/18',
4629 'LB': '178.135.0.0/16',
4630 'LC': '24.92.144.0/20',
4631 'LI': '82.117.0.0/19',
4632 'LK': '112.134.0.0/15',
4633 'LR': '102.183.0.0/16',
4634 'LS': '129.232.0.0/17',
4635 'LT': '78.56.0.0/13',
4636 'LU': '188.42.0.0/16',
4637 'LV': '46.109.0.0/16',
4638 'LY': '41.252.0.0/14',
4639 'MA': '105.128.0.0/11',
4640 'MC': '88.209.64.0/18',
4641 'MD': '37.246.0.0/16',
4642 'ME': '178.175.0.0/17',
4643 'MF': '74.112.232.0/21',
4644 'MG': '154.126.0.0/17',
4645 'MH': '117.103.88.0/21',
4646 'MK': '77.28.0.0/15',
4647 'ML': '154.118.128.0/18',
4648 'MM': '37.111.0.0/17',
4649 'MN': '49.0.128.0/17',
4650 'MO': '60.246.0.0/16',
4651 'MP': '202.88.64.0/20',
4652 'MQ': '109.203.224.0/19',
4653 'MR': '41.188.64.0/18',
4654 'MS': '208.90.112.0/22',
4655 'MT': '46.11.0.0/16',
4656 'MU': '105.16.0.0/12',
4657 'MV': '27.114.128.0/18',
4658 'MW': '102.70.0.0/15',
4659 'MX': '187.192.0.0/11',
4660 'MY': '175.136.0.0/13',
4661 'MZ': '197.218.0.0/15',
4662 'NA': '41.182.0.0/16',
4663 'NC': '101.101.0.0/18',
4664 'NE': '197.214.0.0/18',
4665 'NF': '203.17.240.0/22',
4666 'NG': '105.112.0.0/12',
4667 'NI': '186.76.0.0/15',
4668 'NL': '145.96.0.0/11',
4669 'NO': '84.208.0.0/13',
4670 'NP': '36.252.0.0/15',
4671 'NR': '203.98.224.0/19',
4672 'NU': '49.156.48.0/22',
4673 'NZ': '49.224.0.0/14',
4674 'OM': '5.36.0.0/15',
4675 'PA': '186.72.0.0/15',
4676 'PE': '186.160.0.0/14',
4677 'PF': '123.50.64.0/18',
4678 'PG': '124.240.192.0/19',
4679 'PH': '49.144.0.0/13',
4680 'PK': '39.32.0.0/11',
4681 'PL': '83.0.0.0/11',
4682 'PM': '70.36.0.0/20',
4683 'PR': '66.50.0.0/16',
4684 'PS': '188.161.0.0/16',
4685 'PT': '85.240.0.0/13',
4686 'PW': '202.124.224.0/20',
4687 'PY': '181.120.0.0/14',
4688 'QA': '37.210.0.0/15',
4689 'RE': '102.35.0.0/16',
4690 'RO': '79.112.0.0/13',
4691 'RS': '93.86.0.0/15',
4692 'RU': '5.136.0.0/13',
4693 'RW': '41.186.0.0/16',
4694 'SA': '188.48.0.0/13',
4695 'SB': '202.1.160.0/19',
4696 'SC': '154.192.0.0/11',
4697 'SD': '102.120.0.0/13',
4698 'SE': '78.64.0.0/12',
4699 'SG': '8.128.0.0/10',
4700 'SI': '188.196.0.0/14',
4701 'SK': '78.98.0.0/15',
4702 'SL': '102.143.0.0/17',
4703 'SM': '89.186.32.0/19',
4704 'SN': '41.82.0.0/15',
4705 'SO': '154.115.192.0/18',
4706 'SR': '186.179.128.0/17',
4707 'SS': '105.235.208.0/21',
4708 'ST': '197.159.160.0/19',
4709 'SV': '168.243.0.0/16',
4710 'SX': '190.102.0.0/20',
4712 'SZ': '41.84.224.0/19',
4713 'TC': '65.255.48.0/20',
4714 'TD': '154.68.128.0/19',
4715 'TG': '196.168.0.0/14',
4716 'TH': '171.96.0.0/13',
4717 'TJ': '85.9.128.0/18',
4718 'TK': '27.96.24.0/21',
4719 'TL': '180.189.160.0/20',
4720 'TM': '95.85.96.0/19',
4721 'TN': '197.0.0.0/11',
4722 'TO': '175.176.144.0/21',
4723 'TR': '78.160.0.0/11',
4724 'TT': '186.44.0.0/15',
4725 'TV': '202.2.96.0/19',
4726 'TW': '120.96.0.0/11',
4727 'TZ': '156.156.0.0/14',
4728 'UA': '37.52.0.0/14',
4729 'UG': '102.80.0.0/13',
4731 'UY': '167.56.0.0/13',
4732 'UZ': '84.54.64.0/18',
4733 'VA': '212.77.0.0/19',
4734 'VC': '207.191.240.0/21',
4735 'VE': '186.88.0.0/13',
4736 'VG': '66.81.192.0/20',
4737 'VI': '146.226.0.0/16',
4738 'VN': '14.160.0.0/11',
4739 'VU': '202.80.32.0/20',
4740 'WF': '117.20.32.0/21',
4741 'WS': '202.4.32.0/19',
4742 'YE': '134.35.0.0/16',
4743 'YT': '41.242.116.0/22',
4744 'ZA': '41.0.0.0/11',
4745 'ZM': '102.144.0.0/13',
4746 'ZW': '102.177.192.0/18',
4750 def random_ipv4(cls, code_or_block):
4751 if len(code_or_block) == 2:
4752 block = cls._country_ip_map.get(code_or_block.upper())
4756 block = code_or_block
4757 addr, preflen = block.split('/')
4758 addr_min = struct.unpack('!L', socket.inet_aton(addr))[0]
4759 addr_max = addr_min | (0xffffffff >> int(preflen))
4760 return str(socket.inet_ntoa(
4761 struct.pack('!L', random.randint(addr_min, addr_max))))
4764 class PerRequestProxyHandler(urllib.request.ProxyHandler):
4765 def __init__(self, proxies=None):
4766 # Set default handlers
4767 for type in ('http', 'https'):
4768 setattr(self, '%s_open' % type,
4769 lambda r, proxy='__noproxy__', type=type, meth=self.proxy_open:
4770 meth(r, proxy, type))
4771 urllib.request.ProxyHandler.__init__(self, proxies)
4773 def proxy_open(self, req, proxy, type):
4774 req_proxy = req.headers.get('Ytdl-request-proxy')
4775 if req_proxy is not None:
4777 del req.headers['Ytdl-request-proxy']
4779 if proxy == '__noproxy__':
4780 return None # No Proxy
4781 if urllib.parse.urlparse(proxy).scheme.lower() in ('socks', 'socks4', 'socks4a', 'socks5'):
4782 req.add_header('Ytdl-socks-proxy', proxy)
4783 # yt-dlp's http/https handlers do wrapping the socket with socks
4785 return urllib.request.ProxyHandler.proxy_open(
4786 self, req, proxy, type)
4789 # Both long_to_bytes and bytes_to_long are adapted from PyCrypto, which is
4790 # released into Public Domain
4791 # https://github.com/dlitz/pycrypto/blob/master/lib/Crypto/Util/number.py#L387
4793 def long_to_bytes(n, blocksize=0):
4794 """long_to_bytes(n:long, blocksize:int) : string
4795 Convert a long integer to a byte string.
4797 If optional blocksize is given and greater than zero, pad the front of the
4798 byte string with binary zeros so that the length is a multiple of
4801 # after much testing, this algorithm was deemed to be the fastest
4805 s = struct.pack('>I', n & 0xffffffff) + s
4807 # strip off leading zeros
4808 for i in range(len(s)):
4809 if s[i] != b'\000'[0]:
4812 # only happens when n == 0
4816 # add back some pad bytes. this could be done more efficiently w.r.t. the
4817 # de-padding being done above, but sigh...
4818 if blocksize > 0 and len(s) % blocksize:
4819 s = (blocksize - len(s) % blocksize) * b'\000' + s
4823 def bytes_to_long(s):
4824 """bytes_to_long(string) : long
4825 Convert a byte string to a long integer.
4827 This is (essentially) the inverse of long_to_bytes().
4832 extra = (4 - length % 4)
4833 s = b'\000' * extra + s
4834 length = length + extra
4835 for i in range(0, length, 4):
4836 acc = (acc << 32) + struct.unpack('>I', s[i:i + 4])[0]
4840 def ohdave_rsa_encrypt(data, exponent, modulus):
4842 Implement OHDave
's RSA algorithm. See http://www.ohdave.com/rsa/
4845 data: data to encrypt, bytes-like object
4846 exponent, modulus: parameter e and N of RSA algorithm, both integer
4847 Output: hex string of encrypted data
4849 Limitation: supports one block encryption only
4852 payload = int(binascii.hexlify(data[::-1]), 16)
4853 encrypted = pow(payload, exponent, modulus)
4854 return '%x' % encrypted
4857 def pkcs1pad(data, length):
4859 Padding input data with PKCS#1 scheme
4861 @param {int[]} data input data
4862 @param {int} length target length
4863 @returns {int[]} padded data
4865 if len(data) > length - 11:
4866 raise ValueError('Input data too
long for PKCS
#1 padding')
4868 pseudo_random
= [random
.randint(0, 254) for _
in range(length
- len(data
) - 3)]
4869 return [0, 2] + pseudo_random
+ [0] + data
4872 def _base_n_table(n
, table
):
4873 if not table
and not n
:
4874 raise ValueError('Either table or n must be specified')
4875 table
= (table
or '0123456789abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ')[:n
]
4877 if n
and n
!= len(table
):
4878 raise ValueError(f
'base {n} exceeds table length {len(table)}')
4882 def encode_base_n(num
, n
=None, table
=None):
4883 """Convert given int to a base-n string"""
4884 table
= _base_n_table(n
, table
)
4888 result
, base
= '', len(table
)
4890 result
= table
[num
% base
] + result
4895 def decode_base_n(string
, n
=None, table
=None):
4896 """Convert given base-n string to int"""
4897 table
= {char: index for index, char in enumerate(_base_n_table(n, table))}
4898 result
, base
= 0, len(table
)
4900 result
= result
* base
+ table
[char
]
4904 def decode_base(value
, digits
):
4905 deprecation_warning(f
'{__name__}.decode_base is deprecated and may be removed '
4906 f
'in a future version. Use {__name__}.decode_base_n instead')
4907 return decode_base_n(value
, table
=digits
)
4910 def decode_packed_codes(code
):
4911 mobj
= re
.search(PACKED_CODES_RE
, code
)
4912 obfuscated_code
, base
, count
, symbols
= mobj
.groups()
4915 symbols
= symbols
.split('|')
4920 base_n_count
= encode_base_n(count
, base
)
4921 symbol_table
[base_n_count
] = symbols
[count
] or base_n_count
4924 r
'\b(\w+)\b', lambda mobj
: symbol_table
[mobj
.group(0)],
4928 def caesar(s
, alphabet
, shift
):
4933 alphabet
[(alphabet
.index(c
) + shift
) % l
] if c
in alphabet
else c
4938 return caesar(s
, r
'''!"#$%&'()*+,-./0123456789:;<=>?@ABCDEFGHIJKLMNOPQRSTUVWXYZ[\]^_`abcdefghijklmnopqrstuvwxyz{|}~''', 47)
4941 def parse_m3u8_attributes(attrib
):
4943 for (key
, val
) in re
.findall(r
'(?P<key>[A-Z0-9-]+)=(?P<val>"[^"]+"|[^",]+)(?:,|$)', attrib
):
4944 if val
.startswith('"'):
4950 def urshift(val
, n
):
4951 return val
>> n
if val
>= 0 else (val
+ 0x100000000) >> n
4954 # Based on png2str() written by @gdkchan and improved by @yokrysty
4955 # Originally posted at https://github.com/ytdl-org/youtube-dl/issues/9706
4956 def decode_png(png_data
):
4957 # Reference: https://www.w3.org/TR/PNG/
4958 header
= png_data
[8:]
4960 if png_data
[:8] != b
'\x89PNG\x0d\x0a\x1a\x0a' or header
[4:8] != b
'IHDR':
4961 raise OSError('Not a valid PNG file.')
4963 int_map
= {1: '>B', 2: '>H', 4: '>I'}
4964 unpack_integer
= lambda x
: struct
.unpack(int_map
[len(x
)], x
)[0]
4969 length
= unpack_integer(header
[:4])
4972 chunk_type
= header
[:4]
4975 chunk_data
= header
[:length
]
4976 header
= header
[length
:]
4978 header
= header
[4:] # Skip CRC
4986 ihdr
= chunks
[0]['data']
4988 width
= unpack_integer(ihdr
[:4])
4989 height
= unpack_integer(ihdr
[4:8])
4993 for chunk
in chunks
:
4994 if chunk
['type'] == b
'IDAT':
4995 idat
+= chunk
['data']
4998 raise OSError('Unable to read PNG data.')
5000 decompressed_data
= bytearray(zlib
.decompress(idat
))
5005 def _get_pixel(idx
):
5010 for y
in range(height
):
5011 basePos
= y
* (1 + stride
)
5012 filter_type
= decompressed_data
[basePos
]
5016 pixels
.append(current_row
)
5018 for x
in range(stride
):
5019 color
= decompressed_data
[1 + basePos
+ x
]
5020 basex
= y
* stride
+ x
5025 left
= _get_pixel(basex
- 3)
5027 up
= _get_pixel(basex
- stride
)
5029 if filter_type
== 1: # Sub
5030 color
= (color
+ left
) & 0xff
5031 elif filter_type
== 2: # Up
5032 color
= (color
+ up
) & 0xff
5033 elif filter_type
== 3: # Average
5034 color
= (color
+ ((left
+ up
) >> 1)) & 0xff
5035 elif filter_type
== 4: # Paeth
5041 c
= _get_pixel(basex
- stride
- 3)
5049 if pa
<= pb
and pa
<= pc
:
5050 color
= (color
+ a
) & 0xff
5052 color
= (color
+ b
) & 0xff
5054 color
= (color
+ c
) & 0xff
5056 current_row
.append(color
)
5058 return width
, height
, pixels
5061 def write_xattr(path
, key
, value
):
5062 # Windows: Write xattrs to NTFS Alternate Data Streams:
5063 # http://en.wikipedia.org/wiki/NTFS#Alternate_data_streams_.28ADS.29
5064 if compat_os_name
== 'nt':
5065 assert ':' not in key
5066 assert os
.path
.exists(path
)
5069 with open(f
'{path}:{key}', 'wb') as f
:
5071 except OSError as e
:
5072 raise XAttrMetadataError(e
.errno
, e
.strerror
)
5075 # UNIX Method 1. Use xattrs/pyxattrs modules
5078 if getattr(xattr
, '_yt_dlp__identifier', None) == 'pyxattr':
5079 # Unicode arguments are not supported in pyxattr until version 0.5.0
5080 # See https://github.com/ytdl-org/youtube-dl/issues/5498
5081 if version_tuple(xattr
.__version
__) >= (0, 5, 0):
5082 setxattr
= xattr
.set
5084 setxattr
= xattr
.setxattr
5088 setxattr(path
, key
, value
)
5089 except OSError as e
:
5090 raise XAttrMetadataError(e
.errno
, e
.strerror
)
5093 # UNIX Method 2. Use setfattr/xattr executables
5094 exe
= ('setfattr' if check_executable('setfattr', ['--version'])
5095 else 'xattr' if check_executable('xattr', ['-h']) else None)
5097 raise XAttrUnavailableError(
5098 'Couldn\'t find a tool to set the xattrs. Install either the python "xattr" or "pyxattr" modules or the '
5099 + ('"xattr" binary' if sys
.platform
!= 'linux' else 'GNU "attr" package (which contains the "setfattr" tool)'))
5101 value
= value
.decode()
5103 _
, stderr
, returncode
= Popen
.run(
5104 [exe
, '-w', key
, value
, path
] if exe
== 'xattr' else [exe
, '-n', key
, '-v', value
, path
],
5105 text
=True, stdout
=subprocess
.PIPE
, stderr
=subprocess
.PIPE
, stdin
=subprocess
.PIPE
)
5106 except OSError as e
:
5107 raise XAttrMetadataError(e
.errno
, e
.strerror
)
5109 raise XAttrMetadataError(returncode
, stderr
)
5112 def random_birthday(year_field
, month_field
, day_field
):
5113 start_date
= datetime
.date(1950, 1, 1)
5114 end_date
= datetime
.date(1995, 12, 31)
5115 offset
= random
.randint(0, (end_date
- start_date
).days
)
5116 random_date
= start_date
+ datetime
.timedelta(offset
)
5118 year_field
: str(random_date
.year
),
5119 month_field
: str(random_date
.month
),
5120 day_field
: str(random_date
.day
),
5124 # Templates for internet shortcut files, which are plain text files.
5125 DOT_URL_LINK_TEMPLATE
= '''\
5130 DOT_WEBLOC_LINK_TEMPLATE
= '''\
5131 <?xml version="1.0" encoding="UTF-8"?>
5132 <!DOCTYPE plist PUBLIC "-//Apple//DTD PLIST 1.0//EN" "http://www.apple.com/DTDs/PropertyList-1.0.dtd">
5133 <plist version="1.0">
5136 \t<string>%(url)s</string>
5141 DOT_DESKTOP_LINK_TEMPLATE
= '''\
5151 'url': DOT_URL_LINK_TEMPLATE
,
5152 'desktop': DOT_DESKTOP_LINK_TEMPLATE
,
5153 'webloc': DOT_WEBLOC_LINK_TEMPLATE
,
5157 def iri_to_uri(iri
):
5159 Converts an IRI (Internationalized Resource Identifier, allowing Unicode characters) to a URI (Uniform Resource Identifier, ASCII-only).
5161 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.
5164 iri_parts
= urllib
.parse
.urlparse(iri
)
5166 if '[' in iri_parts
.netloc
:
5167 raise ValueError('IPv6 URIs are not, yet, supported.')
5168 # Querying `.netloc`, when there's only one bracket, also raises a ValueError.
5170 # 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.
5173 if iri_parts
.username
:
5174 net_location
+= urllib
.parse
.quote(iri_parts
.username
, safe
=r
"!$%&'()*+,~")
5175 if iri_parts
.password
is not None:
5176 net_location
+= ':' + urllib
.parse
.quote(iri_parts
.password
, safe
=r
"!$%&'()*+,~")
5179 net_location
+= iri_parts
.hostname
.encode('idna').decode() # Punycode for Unicode hostnames.
5180 # The 'idna' encoding produces ASCII text.
5181 if iri_parts
.port
is not None and iri_parts
.port
!= 80:
5182 net_location
+= ':' + str(iri_parts
.port
)
5184 return urllib
.parse
.urlunparse(
5188 urllib
.parse
.quote_plus(iri_parts
.path
, safe
=r
"!$%&'()*+,/:;=@|~"),
5190 # Unsure about the `safe` argument, since this is a legacy way of handling parameters.
5191 urllib
.parse
.quote_plus(iri_parts
.params
, safe
=r
"!$%&'()*+,/:;=@|~"),
5193 # Not totally sure about the `safe` argument, since the source does not explicitly mention the query URI component.
5194 urllib
.parse
.quote_plus(iri_parts
.query
, safe
=r
"!$%&'()*+,/:;=?@{|}~"),
5196 urllib
.parse
.quote_plus(iri_parts
.fragment
, safe
=r
"!#$%&'()*+,/:;=?@{|}~")))
5198 # Source for `safe` arguments: https://url.spec.whatwg.org/#percent-encoded-bytes.
5201 def to_high_limit_path(path
):
5202 if sys
.platform
in ['win32', 'cygwin']:
5203 # Work around MAX_PATH limitation on Windows. The maximum allowed length for the individual path segments may still be quite limited.
5204 return '\\\\?\\' + os
.path
.abspath(path
)
5209 def format_field(obj
, field
=None, template
='%s', ignore
=NO_DEFAULT
, default
='', func
=IDENTITY
):
5210 val
= traverse_obj(obj
, *variadic(field
))
5211 if (not val
and val
!= 0) if ignore
is NO_DEFAULT
else val
in variadic(ignore
):
5213 return template
% func(val
)
5216 def clean_podcast_url(url
):
5217 return re
.sub(r
'''(?x)
5221 media\.blubrry\.com| # https://create.blubrry.com/resources/podcast-media-download-statistics/getting-started/
5224 (?:dts|www)\.podtrac\.com/(?:pts/)?redirect\.[0-9a-z]{3,4}| # http://analytics.podtrac.com/how-to-measure
5227 cn\.co| # https://podcorn.com/analytics-prefix/
5228 st\.fm # https://podsights.com/docs/
5233 _HEX_TABLE
= '0123456789abcdef'
5236 def random_uuidv4():
5237 return re
.sub(r
'[xy]', lambda x
: _HEX_TABLE
[random
.randint(0, 15)], 'xxxxxxxx-xxxx-4xxx-yxxx-xxxxxxxxxxxx')
5240 def make_dir(path
, to_screen
=None):
5242 dn
= os
.path
.dirname(path
)
5243 if dn
and not os
.path
.exists(dn
):
5246 except OSError as err
:
5247 if callable(to_screen
) is not None:
5248 to_screen('unable to create directory ' + error_to_compat_str(err
))
5252 def get_executable_path():
5253 from .update
import _get_variant_and_executable_path
5255 return os
.path
.dirname(os
.path
.abspath(_get_variant_and_executable_path()[1]))
5258 def load_plugins(name
, suffix
, namespace
):
5260 with contextlib
.suppress(FileNotFoundError
):
5261 plugins_spec
= importlib
.util
.spec_from_file_location(
5262 name
, os
.path
.join(get_executable_path(), 'ytdlp_plugins', name
, '__init__.py'))
5263 plugins
= importlib
.util
.module_from_spec(plugins_spec
)
5264 sys
.modules
[plugins_spec
.name
] = plugins
5265 plugins_spec
.loader
.exec_module(plugins
)
5266 for name
in dir(plugins
):
5267 if name
in namespace
:
5269 if not name
.endswith(suffix
):
5271 klass
= getattr(plugins
, name
)
5272 classes
[name
] = namespace
[name
] = klass
5277 obj
, *path_list
, default
=None, expected_type
=None, get_all
=True,
5278 casesense
=True, is_user_input
=False, traverse_string
=False):
5279 ''' Traverse nested list/dict/tuple
5280 @param path_list A list of paths which are checked one by one.
5281 Each path is a list of keys where each key is a:
5283 - string: A dictionary key / regex group
5284 - int: An index into a list
5285 - tuple: A list of keys all of which will be traversed
5286 - Ellipsis: Fetch all values in the object
5287 - Function: Takes the key and value as arguments
5288 and returns whether the key matches or not
5289 @param default Default value to return
5290 @param expected_type Only accept final value of this type (Can also be any callable)
5291 @param get_all Return all the values obtained from a path or only the first one
5292 @param casesense Whether to consider dictionary keys as case sensitive
5294 The following are only meant to be used by YoutubeDL.prepare_outtmpl and is not part of the API
5296 @param path_list In addition to the above,
5297 - dict: Given {k:v, ...}; return {k: traverse_obj(obj, v), ...}
5298 @param is_user_input Whether the keys are generated from user input. If True,
5299 strings are converted to int/slice if necessary
5300 @param traverse_string Whether to traverse inside strings. If True, any
5301 non-compatible object will also be converted into a string
5302 ''' # TODO: Write tests
5304 _lower
= lambda k
: (k
.lower() if isinstance(k
, str) else k
)
5305 path_list
= (map(_lower
, variadic(path
)) for path
in path_list
)
5307 def _traverse_obj(obj
, path
, _current_depth
=0):
5309 path
= tuple(variadic(path
))
5310 for i
, key
in enumerate(path
):
5311 if None in (key
, obj
):
5313 if isinstance(key
, (list, tuple)):
5314 obj
= [_traverse_obj(obj
, sub_key
, _current_depth
) for sub_key
in key
]
5318 obj
= (obj
.values() if isinstance(obj
, dict)
5319 else obj
if isinstance(obj
, (list, tuple, LazyList
))
5320 else str(obj
) if traverse_string
else [])
5322 depth
= max(depth
, _current_depth
)
5323 return [_traverse_obj(inner_obj
, path
[i
+ 1:], _current_depth
) for inner_obj
in obj
]
5324 elif isinstance(key
, dict):
5325 obj
= filter_dict({k: _traverse_obj(obj, v, _current_depth) for k, v in key.items()}
)
5327 if isinstance(obj
, (list, tuple, LazyList
)):
5328 obj
= enumerate(obj
)
5329 elif isinstance(obj
, dict):
5332 if not traverse_string
:
5336 depth
= max(depth
, _current_depth
)
5337 return [_traverse_obj(v
, path
[i
+ 1:], _current_depth
) for k
, v
in obj
if try_call(key
, args
=(k
, v
))]
5338 elif isinstance(obj
, dict) and not (is_user_input
and key
== ':'):
5339 obj
= (obj
.get(key
) if casesense
or (key
in obj
)
5340 else next((v
for k
, v
in obj
.items() if _lower(k
) == key
), None))
5343 key
= (int_or_none(key
) if ':' not in key
5344 else slice(*map(int_or_none
, key
.split(':'))))
5345 if key
== slice(None):
5346 return _traverse_obj(obj
, (..., *path
[i
+ 1:]), _current_depth
)
5347 if not isinstance(key
, (int, slice)):
5349 if not isinstance(obj
, (list, tuple, LazyList
)):
5350 if not traverse_string
:
5359 if isinstance(expected_type
, type):
5360 type_test
= lambda val
: val
if isinstance(val
, expected_type
) else None
5362 type_test
= expected_type
or IDENTITY
5364 for path
in path_list
:
5366 val
= _traverse_obj(obj
, path
)
5369 for _
in range(depth
- 1):
5370 val
= itertools
.chain
.from_iterable(v
for v
in val
if v
is not None)
5371 val
= [v
for v
in map(type_test
, val
) if v
is not None]
5373 return val
if get_all
else val
[0]
5375 val
= type_test(val
)
5381 def traverse_dict(dictn
, keys
, casesense
=True):
5382 deprecation_warning(f
'"{__name__}.traverse_dict" is deprecated and may be removed '
5383 f
'in a future version. Use "{__name__}.traverse_obj" instead')
5384 return traverse_obj(dictn
, keys
, casesense
=casesense
, is_user_input
=True, traverse_string
=True)
5387 def get_first(obj
, keys
, **kwargs
):
5388 return traverse_obj(obj
, (..., *variadic(keys
)), **kwargs
, get_all
=False)
5391 def variadic(x
, allowed_types
=(str, bytes, dict)):
5392 return x
if isinstance(x
, collections
.abc
.Iterable
) and not isinstance(x
, allowed_types
) else (x
,)
5395 def time_seconds(**kwargs
):
5396 t
= datetime
.datetime
.now(datetime
.timezone(datetime
.timedelta(**kwargs
)))
5397 return t
.timestamp()
5400 # create a JSON Web Signature (jws) with HS256 algorithm
5401 # the resulting format is in JWS Compact Serialization
5402 # implemented following JWT https://www.rfc-editor.org/rfc/rfc7519.html
5403 # implemented following JWS https://www.rfc-editor.org/rfc/rfc7515.html
5404 def jwt_encode_hs256(payload_data
, key
, headers
={}):
5410 header_data
.update(headers
)
5411 header_b64
= base64
.b64encode(json
.dumps(header_data
).encode())
5412 payload_b64
= base64
.b64encode(json
.dumps(payload_data
).encode())
5413 h
= hmac
.new(key
.encode(), header_b64
+ b
'.' + payload_b64
, hashlib
.sha256
)
5414 signature_b64
= base64
.b64encode(h
.digest())
5415 token
= header_b64
+ b
'.' + payload_b64
+ b
'.' + signature_b64
5419 # can be extended in future to verify the signature and parse header and return the algorithm used if it's not HS256
5420 def jwt_decode_hs256(jwt
):
5421 header_b64
, payload_b64
, signature_b64
= jwt
.split('.')
5422 payload_data
= json
.loads(base64
.urlsafe_b64decode(payload_b64
))
5426 WINDOWS_VT_MODE
= False if compat_os_name
== 'nt' else None
5430 def supports_terminal_sequences(stream
):
5431 if compat_os_name
== 'nt':
5432 if not WINDOWS_VT_MODE
:
5434 elif not os
.getenv('TERM'):
5437 return stream
.isatty()
5438 except BaseException
:
5442 def windows_enable_vt_mode(): # TODO: Do this the proper way https://bugs.python.org/issue30075
5443 if get_windows_version() < (10, 0, 10586):
5445 global WINDOWS_VT_MODE
5447 Popen
.run('', shell
=True)
5451 WINDOWS_VT_MODE
= True
5452 supports_terminal_sequences
.cache_clear()
5455 _terminal_sequences_re
= re
.compile('\033\\[[^m]+m')
5458 def remove_terminal_sequences(string
):
5459 return _terminal_sequences_re
.sub('', string
)
5462 def number_of_digits(number
):
5463 return len('%d' % number
)
5466 def join_nonempty(*values
, delim
='-', from_dict
=None):
5467 if from_dict
is not None:
5468 values
= (traverse_obj(from_dict
, variadic(v
)) for v
in values
)
5469 return delim
.join(map(str, filter(None, values
)))
5472 def scale_thumbnails_to_max_format_width(formats
, thumbnails
, url_width_re
):
5474 Find the largest format dimensions in terms of video width and, for each thumbnail:
5475 * Modify the URL: Match the width with the provided regex and replace with the former width
5478 This function is useful with video services that scale the provided thumbnails on demand
5480 _keys
= ('width', 'height')
5481 max_dimensions
= max(
5482 (tuple(format
.get(k
) or 0 for k
in _keys
) for format
in formats
),
5484 if not max_dimensions
[0]:
5488 {'url': re.sub(url_width_re, str(max_dimensions[0]), thumbnail['url'])}
,
5489 dict(zip(_keys
, max_dimensions
)), thumbnail
)
5490 for thumbnail
in thumbnails
5494 def parse_http_range(range):
5495 """ Parse value of "Range" or "Content-Range" HTTP header into tuple. """
5497 return None, None, None
5498 crg
= re
.search(r
'bytes[ =](\d+)-(\d+)?(?:/(\d+))?', range)
5500 return None, None, None
5501 return int(crg
.group(1)), int_or_none(crg
.group(2)), int_or_none(crg
.group(3))
5504 def read_stdin(what
):
5505 eof
= 'Ctrl+Z' if compat_os_name
== 'nt' else 'Ctrl+D'
5506 write_string(f
'Reading {what} from STDIN - EOF ({eof}) to end:\n')
5510 def determine_file_encoding(data
):
5512 Detect the text encoding used
5513 @returns (encoding, bytes to skip)
5516 # BOM marks are given priority over declarations
5517 for bom
, enc
in BOMS
:
5518 if data
.startswith(bom
):
5519 return enc
, len(bom
)
5521 # Strip off all null bytes to match even when UTF-16 or UTF-32 is used.
5522 # We ignore the endianness to get a good enough match
5523 data
= data
.replace(b
'\0', b
'')
5524 mobj
= re
.match(rb
'(?m)^#\s*coding\s*:\s*(\S+)\s*$', data
)
5525 return mobj
.group(1).decode() if mobj
else None, 0
5532 __initialized
= False
5534 def __init__(self
, parser
, label
=None):
5535 self
.parser
, self
.label
= parser
, label
5536 self
._loaded
_paths
, self
.configs
= set(), []
5538 def init(self
, args
=None, filename
=None):
5539 assert not self
.__initialized
5540 self
.own_args
, self
.filename
= args
, filename
5541 return self
.load_configs()
5543 def load_configs(self
):
5546 location
= os
.path
.realpath(self
.filename
)
5547 directory
= os
.path
.dirname(location
)
5548 if location
in self
._loaded
_paths
:
5550 self
._loaded
_paths
.add(location
)
5552 self
.__initialized
= True
5553 opts
, _
= self
.parser
.parse_known_args(self
.own_args
)
5554 self
.parsed_args
= self
.own_args
5555 for location
in opts
.config_locations
or []:
5557 self
.append_config(shlex
.split(read_stdin('options'), comments
=True), label
='stdin')
5559 location
= os
.path
.join(directory
, expand_path(location
))
5560 if os
.path
.isdir(location
):
5561 location
= os
.path
.join(location
, 'yt-dlp.conf')
5562 if not os
.path
.exists(location
):
5563 self
.parser
.error(f
'config location {location} does not exist')
5564 self
.append_config(self
.read_file(location
), location
)
5568 label
= join_nonempty(
5569 self
.label
, 'config', f
'"{self.filename}"' if self
.filename
else '',
5571 return join_nonempty(
5572 self
.own_args
is not None and f
'{label[0].upper()}{label[1:]}: {self.hide_login_info(self.own_args)}',
5573 *(f
'\n{c}'.replace('\n', '\n| ')[1:] for c
in self
.configs
),
5577 def read_file(filename
, default
=[]):
5579 optionf
= open(filename
, 'rb')
5581 return default
# silently skip if file is not present
5583 enc
, skip
= determine_file_encoding(optionf
.read(512))
5584 optionf
.seek(skip
, io
.SEEK_SET
)
5586 enc
= None # silently skip read errors
5588 # FIXME: https://github.com/ytdl-org/youtube-dl/commit/dfe5fa49aed02cf36ba9f743b11b0903554b5e56
5589 contents
= optionf
.read().decode(enc
or preferredencoding())
5590 res
= shlex
.split(contents
, comments
=True)
5591 except Exception as err
:
5592 raise ValueError(f
'Unable to parse "{filename}": {err}')
5598 def hide_login_info(opts
):
5599 PRIVATE_OPTS
= {'-p', '--password', '-u', '--username', '--video-password', '--ap-password', '--ap-username'}
5600 eqre
= re
.compile('^(?P<key>' + ('|'.join(re
.escape(po
) for po
in PRIVATE_OPTS
)) + ')=.+$')
5605 return m
.group('key') + '=PRIVATE'
5609 opts
= list(map(_scrub_eq
, opts
))
5610 for idx
, opt
in enumerate(opts
):
5611 if opt
in PRIVATE_OPTS
and idx
+ 1 < len(opts
):
5612 opts
[idx
+ 1] = 'PRIVATE'
5615 def append_config(self
, *args
, label
=None):
5616 config
= type(self
)(self
.parser
, label
)
5617 config
._loaded
_paths
= self
._loaded
_paths
5618 if config
.init(*args
):
5619 self
.configs
.append(config
)
5623 for config
in reversed(self
.configs
):
5624 yield from config
.all_args
5625 yield from self
.parsed_args
or []
5627 def parse_known_args(self
, **kwargs
):
5628 return self
.parser
.parse_known_args(self
.all_args
, **kwargs
)
5630 def parse_args(self
):
5631 return self
.parser
.parse_args(self
.all_args
)
5634 class WebSocketsWrapper():
5635 """Wraps websockets module to use in non-async scopes"""
5638 def __init__(self
, url
, headers
=None, connect
=True):
5639 self
.loop
= asyncio
.new_event_loop()
5640 # XXX: "loop" is deprecated
5641 self
.conn
= websockets
.connect(
5642 url
, extra_headers
=headers
, ping_interval
=None,
5643 close_timeout
=float('inf'), loop
=self
.loop
, ping_timeout
=float('inf'))
5646 atexit
.register(self
.__exit
__, None, None, None)
5648 def __enter__(self
):
5650 self
.pool
= self
.run_with_loop(self
.conn
.__aenter
__(), self
.loop
)
5653 def send(self
, *args
):
5654 self
.run_with_loop(self
.pool
.send(*args
), self
.loop
)
5656 def recv(self
, *args
):
5657 return self
.run_with_loop(self
.pool
.recv(*args
), self
.loop
)
5659 def __exit__(self
, type, value
, traceback
):
5661 return self
.run_with_loop(self
.conn
.__aexit
__(type, value
, traceback
), self
.loop
)
5664 self
._cancel
_all
_tasks
(self
.loop
)
5666 # taken from https://github.com/python/cpython/blob/3.9/Lib/asyncio/runners.py with modifications
5667 # for contributors: If there's any new library using asyncio needs to be run in non-async, move these function out of this class
5669 def run_with_loop(main
, loop
):
5670 if not asyncio
.iscoroutine(main
):
5671 raise ValueError(f
'a coroutine was expected, got {main!r}')
5674 return loop
.run_until_complete(main
)
5676 loop
.run_until_complete(loop
.shutdown_asyncgens())
5677 if hasattr(loop
, 'shutdown_default_executor'):
5678 loop
.run_until_complete(loop
.shutdown_default_executor())
5681 def _cancel_all_tasks(loop
):
5682 to_cancel
= asyncio
.all_tasks(loop
)
5687 for task
in to_cancel
:
5690 # XXX: "loop" is removed in python 3.10+
5691 loop
.run_until_complete(
5692 asyncio
.gather(*to_cancel
, loop
=loop
, return_exceptions
=True))
5694 for task
in to_cancel
:
5695 if task
.cancelled():
5697 if task
.exception() is not None:
5698 loop
.call_exception_handler({
5699 'message': 'unhandled exception during asyncio.run() shutdown',
5700 'exception': task
.exception(),
5705 def merge_headers(*dicts
):
5706 """Merge dicts of http headers case insensitively, prioritizing the latter ones"""
5707 return {k.title(): v for k, v in itertools.chain.from_iterable(map(dict.items, dicts))}
5710 def cached_method(f
):
5711 """Cache a method"""
5712 signature
= inspect
.signature(f
)
5715 def wrapper(self
, *args
, **kwargs
):
5716 bound_args
= signature
.bind(self
, *args
, **kwargs
)
5717 bound_args
.apply_defaults()
5718 key
= tuple(bound_args
.arguments
.values())
5720 if not hasattr(self
, '__cached_method__cache'):
5721 self
.__cached
_method
__cache
= {}
5722 cache
= self
.__cached
_method
__cache
.setdefault(f
.__name
__, {})
5723 if key
not in cache
:
5724 cache
[key
] = f(self
, *args
, **kwargs
)
5729 class classproperty
:
5730 """property access for class methods"""
5732 def __init__(self
, func
):
5733 functools
.update_wrapper(self
, func
)
5736 def __get__(self
, _
, cls
):
5737 return self
.func(cls
)
5740 class Namespace(types
.SimpleNamespace
):
5741 """Immutable namespace"""
5744 return iter(self
.__dict
__.values())
5748 return self
.__dict
__.items()
5751 MEDIA_EXTENSIONS
= Namespace(
5752 common_video
=('avi', 'flv', 'mkv', 'mov', 'mp4', 'webm'),
5753 video
=('3g2', '3gp', 'f4v', 'mk3d', 'divx', 'mpg', 'ogv', 'm4v', 'wmv'),
5754 common_audio
=('aiff', 'alac', 'flac', 'm4a', 'mka', 'mp3', 'ogg', 'opus', 'wav'),
5755 audio
=('aac', 'ape', 'asf', 'f4a', 'f4b', 'm4b', 'm4p', 'm4r', 'oga', 'ogx', 'spx', 'vorbis', 'wma'),
5756 thumbnails
=('jpg', 'png', 'webp'),
5757 storyboards
=('mhtml', ),
5758 subtitles
=('srt', 'vtt', 'ass', 'lrc'),
5759 manifests
=('f4f', 'f4m', 'm3u8', 'smil', 'mpd'),
5761 MEDIA_EXTENSIONS
.video
+= MEDIA_EXTENSIONS
.common_video
5762 MEDIA_EXTENSIONS
.audio
+= MEDIA_EXTENSIONS
.common_audio
5764 KNOWN_EXTENSIONS
= (*MEDIA_EXTENSIONS
.video
, *MEDIA_EXTENSIONS
.audio
, *MEDIA_EXTENSIONS
.manifests
)
5769 for retry in RetryManager(...):
5772 except SomeException as err:
5776 attempt
, _error
= 0, None
5778 def __init__(self
, _retries
, _error_callback
, **kwargs
):
5779 self
.retries
= _retries
or 0
5780 self
.error_callback
= functools
.partial(_error_callback
, **kwargs
)
5782 def _should_retry(self
):
5783 return self
._error
is not NO_DEFAULT
and self
.attempt
<= self
.retries
5787 if self
._error
is NO_DEFAULT
:
5792 def error(self
, value
):
5796 while self
._should
_retry
():
5797 self
.error
= NO_DEFAULT
5801 self
.error_callback(self
.error
, self
.attempt
, self
.retries
)
5804 def report_retry(e
, count
, retries
, *, sleep_func
, info
, warn
, error
=None, suffix
=None):
5805 """Utility function for reporting retries"""
5808 return error(f
'{e}. Giving up after {count - 1} retries') if count
> 1 else error(str(e
))
5813 elif isinstance(e
, ExtractorError
):
5814 e
= remove_end(str_or_none(e
.cause
) or e
.orig_msg
, '.')
5815 warn(f
'{e}. Retrying{format_field(suffix, None, " %s")} ({count}/{retries})...')
5817 delay
= float_or_none(sleep_func(n
=count
- 1)) if callable(sleep_func
) else sleep_func
5819 info(f
'Sleeping {delay:.2f} seconds ...')
5823 def make_archive_id(ie
, video_id
):
5824 ie_key
= ie
if isinstance(ie
, str) else ie
.ie_key()
5825 return f
'{ie_key.lower()} {video_id}'
5828 def truncate_string(s
, left
, right
=0):
5829 assert left
> 3 and right
>= 0
5830 if s
is None or len(s
) <= left
+ right
:
5832 return f
'{s[:left-3]}...{s[-right:]}'
5835 def orderedSet_from_options(options
, alias_dict
, *, use_regex
=False, start
=None):
5836 assert 'all' in alias_dict
, '"all" alias is required'
5837 requested
= list(start
or [])
5839 discard
= val
.startswith('-')
5843 if val
in alias_dict
:
5844 val
= alias_dict
[val
] if not discard
else [
5845 i
[1:] if i
.startswith('-') else f
'-{i}' for i
in alias_dict
[val
]]
5846 # NB: Do not allow regex in aliases for performance
5847 requested
= orderedSet_from_options(val
, alias_dict
, start
=requested
)
5850 current
= (filter(re
.compile(val
, re
.I
).fullmatch
, alias_dict
['all']) if use_regex
5851 else [val
] if val
in alias_dict
['all'] else None)
5853 raise ValueError(val
)
5856 for item
in current
:
5857 while item
in requested
:
5858 requested
.remove(item
)
5860 requested
.extend(current
)
5862 return orderedSet(requested
)
5866 has_certifi
= bool(certifi
)
5867 has_websockets
= bool(websockets
)