38 import xml
.etree
.ElementTree
41 from .compat
import asyncio
, functools
# isort: split
45 compat_etree_fromstring
,
48 compat_html_entities_html5
,
49 compat_HTMLParseError
,
60 compat_urllib_parse_unquote_plus
,
61 compat_urllib_parse_urlencode
,
62 compat_urllib_parse_urlparse
,
63 compat_urllib_request
,
66 from .dependencies
import brotli
, certifi
, websockets
67 from .socks
import ProxyType
, sockssocket
70 def register_socks_protocols():
71 # "Register" SOCKS protocols
72 # In Python < 2.6.5, urlsplit() suffers from bug https://bugs.python.org/issue7904
73 # URLs with protocols not in urlparse.uses_netloc are not handled correctly
74 for scheme
in ('socks', 'socks4', 'socks4a', 'socks5'):
75 if scheme
not in compat_urlparse
.uses_netloc
:
76 compat_urlparse
.uses_netloc
.append(scheme
)
79 # This is not clearly defined otherwise
80 compiled_regex_type
= type(re
.compile(''))
83 def random_user_agent():
84 _USER_AGENT_TPL
= 'Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/%s Safari/537.36'
125 return _USER_AGENT_TPL
% random
.choice(_CHROME_VERSIONS
)
128 SUPPORTED_ENCODINGS
= [
132 SUPPORTED_ENCODINGS
.append('br')
135 'User-Agent': random_user_agent(),
136 'Accept': 'text/html,application/xhtml+xml,application/xml;q=0.9,*/*;q=0.8',
137 'Accept-Language': 'en-us,en;q=0.5',
138 'Sec-Fetch-Mode': 'navigate',
143 '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',
147 NO_DEFAULT
= object()
149 ENGLISH_MONTH_NAMES
= [
150 'January', 'February', 'March', 'April', 'May', 'June',
151 'July', 'August', 'September', 'October', 'November', 'December']
154 'en': ENGLISH_MONTH_NAMES
,
156 'janvier', 'février', 'mars', 'avril', 'mai', 'juin',
157 'juillet', 'août', 'septembre', 'octobre', 'novembre', 'décembre'],
161 'mp4', 'm4a', 'm4p', 'm4b', 'm4r', 'm4v', 'aac',
162 'flv', 'f4v', 'f4a', 'f4b',
163 'webm', 'ogg', 'ogv', 'oga', 'ogx', 'spx', 'opus',
164 'mkv', 'mka', 'mk3d',
173 'f4f', 'f4m', 'm3u8', 'smil')
175 # needed for sanitizing filenames in restricted mode
176 ACCENT_CHARS
= dict(zip('ÂÃÄÀÁÅÆÇÈÉÊËÌÍÎÏÐÑÒÓÔÕÖŐØŒÙÚÛÜŰÝÞßàáâãäåæçèéêëìíîïðñòóôõöőøœùúûüűýþÿ',
177 itertools
.chain('AAAAAA', ['AE'], 'CEEEEIIIIDNOOOOOOO', ['OE'], 'UUUUUY', ['TH', 'ss'],
178 'aaaaaa', ['ae'], 'ceeeeiiiionooooooo', ['oe'], 'uuuuuy', ['th'], 'y')))
208 '%Y-%m-%d %H:%M:%S.%f',
209 '%Y-%m-%d %H:%M:%S:%f',
212 '%Y-%m-%dT%H:%M:%SZ',
213 '%Y-%m-%dT%H:%M:%S.%fZ',
214 '%Y-%m-%dT%H:%M:%S.%f0Z',
216 '%Y-%m-%dT%H:%M:%S.%f',
219 '%b %d %Y at %H:%M:%S',
221 '%B %d %Y at %H:%M:%S',
225 DATE_FORMATS_DAY_FIRST
= list(DATE_FORMATS
)
226 DATE_FORMATS_DAY_FIRST
.extend([
235 DATE_FORMATS_MONTH_FIRST
= list(DATE_FORMATS
)
236 DATE_FORMATS_MONTH_FIRST
.extend([
244 PACKED_CODES_RE
= r
"}\('(.+)',(\d+),(\d+),'([^']+)'\.split\('\|'\)"
245 JSON_LD_RE
= r
'(?is)<script[^>]+type=(["\']?
)application
/ld\
+json\
1[^
>]*>(?P
<json_ld
>.+?
)</script
>'
247 NUMBER_RE = r'\d
+(?
:\
.\d
+)?
'
251 def preferredencoding():
252 """Get preferred encoding.
254 Returns the best encoding scheme for the system, based on
255 locale.getpreferredencoding() and some further tweaks.
258 pref = locale.getpreferredencoding()
266 def write_json_file(obj, fn):
267 """ Encode obj as JSON and write it to fn, atomically if possible """
269 tf = tempfile.NamedTemporaryFile(
270 prefix=f'{os.path.basename(fn)}
.', dir=os.path.dirname(fn),
271 suffix='.tmp
', delete=False, mode='w
', encoding='utf
-8')
275 json.dump(obj, tf, ensure_ascii=False)
276 if sys.platform == 'win32
':
277 # Need to remove existing file on Windows, else os.rename raises
278 # WindowsError or FileExistsError.
279 with contextlib.suppress(OSError):
281 with contextlib.suppress(OSError):
284 os.chmod(tf.name, 0o666 & ~mask)
285 os.rename(tf.name, fn)
287 with contextlib.suppress(OSError):
292 def find_xpath_attr(node, xpath, key, val=None):
293 """ Find the xpath xpath[@key=val] """
294 assert re.match(r'^
[a
-zA
-Z_
-]+$
', key)
295 expr = xpath + ('[@%s]' % key if val is None else f"[@{key}='{val}
']")
296 return node.find(expr)
298 # On python2.6 the xml.etree.ElementTree.Element methods don't support
299 # the namespace parameter
302 def xpath_with_ns(path
, ns_map
):
303 components
= [c
.split(':') for c
in path
.split('/')]
307 replaced
.append(c
[0])
310 replaced
.append('{%s}%s' % (ns_map
[ns
], tag
))
311 return '/'.join(replaced
)
314 def xpath_element(node
, xpath
, name
=None, fatal
=False, default
=NO_DEFAULT
):
315 def _find_xpath(xpath
):
316 return node
.find(xpath
)
318 if isinstance(xpath
, (str, compat_str
)):
319 n
= _find_xpath(xpath
)
327 if default
is not NO_DEFAULT
:
330 name
= xpath
if name
is None else name
331 raise ExtractorError('Could not find XML element %s' % name
)
337 def xpath_text(node
, xpath
, name
=None, fatal
=False, default
=NO_DEFAULT
):
338 n
= xpath_element(node
, xpath
, name
, fatal
=fatal
, default
=default
)
339 if n
is None or n
== default
:
342 if default
is not NO_DEFAULT
:
345 name
= xpath
if name
is None else name
346 raise ExtractorError('Could not find XML element\'s text %s' % name
)
352 def xpath_attr(node
, xpath
, key
, name
=None, fatal
=False, default
=NO_DEFAULT
):
353 n
= find_xpath_attr(node
, xpath
, key
)
355 if default
is not NO_DEFAULT
:
358 name
= f
'{xpath}[@{key}]' if name
is None else name
359 raise ExtractorError('Could not find XML attribute %s' % name
)
365 def get_element_by_id(id, html
, **kwargs
):
366 """Return the content of the tag with the specified ID in the passed HTML document"""
367 return get_element_by_attribute('id', id, html
, **kwargs
)
370 def get_element_html_by_id(id, html
, **kwargs
):
371 """Return the html of the tag with the specified ID in the passed HTML document"""
372 return get_element_html_by_attribute('id', id, html
, **kwargs
)
375 def get_element_by_class(class_name
, html
):
376 """Return the content of the first tag with the specified class in the passed HTML document"""
377 retval
= get_elements_by_class(class_name
, html
)
378 return retval
[0] if retval
else None
381 def get_element_html_by_class(class_name
, html
):
382 """Return the html of the first tag with the specified class in the passed HTML document"""
383 retval
= get_elements_html_by_class(class_name
, html
)
384 return retval
[0] if retval
else None
387 def get_element_by_attribute(attribute
, value
, html
, **kwargs
):
388 retval
= get_elements_by_attribute(attribute
, value
, html
, **kwargs
)
389 return retval
[0] if retval
else None
392 def get_element_html_by_attribute(attribute
, value
, html
, **kargs
):
393 retval
= get_elements_html_by_attribute(attribute
, value
, html
, **kargs
)
394 return retval
[0] if retval
else None
397 def get_elements_by_class(class_name
, html
, **kargs
):
398 """Return the content of all tags with the specified class in the passed HTML document as a list"""
399 return get_elements_by_attribute(
400 'class', r
'[^\'"]*\b%s\b[^\'"]*' % re.escape(class_name),
401 html, escape_value=False)
404 def get_elements_html_by_class(class_name, html):
405 """Return the html of all tags with the specified class in the passed HTML document as a list"""
406 return get_elements_html_by_attribute(
407 'class', r'[^
\'"]*\b%s\b[^\'"]*' % re.escape(class_name),
408 html, escape_value=False)
411 def get_elements_by_attribute(*args, **kwargs):
412 """Return the content of the tag with the specified attribute in the passed HTML document"""
413 return [content for content, _ in get_elements_text_and_html_by_attribute(*args, **kwargs)]
416 def get_elements_html_by_attribute(*args, **kwargs):
417 """Return the html of the tag with the specified attribute in the passed HTML document"""
418 return [whole for _, whole in get_elements_text_and_html_by_attribute(*args, **kwargs)]
421 def get_elements_text_and_html_by_attribute(attribute, value, html, escape_value=True):
423 Return the text (content) and the html (whole) of the tag with the specified
424 attribute in the passed HTML document
427 quote = '' if re.match(r'''[\s"'`
=<>]''', value) else '?'
429 value = re.escape(value) if escape_value else value
431 partial_element_re = rf'''(?x
)
432 <(?P
<tag
>[a
-zA
-Z0
-9:._-]+)
433 (?
:\
s(?
:[^
>"']|"[^
"]*"|
'[^']*')*)?
434 \s{re.escape(attribute)}\s*=\s*(?P<_q>['"]{quote})(?-x:{value})(?P=_q)
437 for m in re.finditer(partial_element_re, html):
438 content, whole = get_element_text_and_html_by_tag(m.group('tag'), html[m.start():])
441 unescapeHTML(re.sub(r'^(?P<q>["\'])(?P
<content
>.*)(?P
=q
)$
', r'\g
<content
>', content, flags=re.DOTALL)),
446 class HTMLBreakOnClosingTagParser(compat_HTMLParser):
448 HTML parser which raises HTMLBreakOnClosingTagException upon reaching the
449 closing tag for the first opening tag it has encountered, and can be used
453 class HTMLBreakOnClosingTagException(Exception):
457 self.tagstack = collections.deque()
458 compat_HTMLParser.__init__(self)
463 def __exit__(self, *_):
467 # handle_endtag does not return upon raising HTMLBreakOnClosingTagException,
468 # so data remains buffered; we no longer have any interest in it, thus
469 # override this method to discard it
472 def handle_starttag(self, tag, _):
473 self.tagstack.append(tag)
475 def handle_endtag(self, tag):
476 if not self.tagstack:
477 raise compat_HTMLParseError('no tags
in the stack
')
479 inner_tag = self.tagstack.pop()
483 raise compat_HTMLParseError(f'matching opening tag
for closing {tag} tag
not found
')
484 if not self.tagstack:
485 raise self.HTMLBreakOnClosingTagException()
488 def get_element_text_and_html_by_tag(tag, html):
490 For the first element with the specified tag in the passed HTML document
491 return its' content (text
) and the whole
element (html
)
493 def find_or_raise(haystack, needle, exc):
495 return haystack.index(needle)
498 closing_tag = f'</{tag}>'
499 whole_start = find_or_raise(
500 html, f'<{tag}', compat_HTMLParseError(f'opening {tag} tag not found'))
501 content_start = find_or_raise(
502 html[whole_start:], '>', compat_HTMLParseError(f'malformed opening {tag} tag'))
503 content_start += whole_start + 1
504 with HTMLBreakOnClosingTagParser() as parser:
505 parser.feed(html[whole_start:content_start])
506 if not parser.tagstack or parser.tagstack[0] != tag:
507 raise compat_HTMLParseError(f'parser did not match opening {tag} tag')
508 offset = content_start
509 while offset < len(html):
510 next_closing_tag_start = find_or_raise(
511 html[offset:], closing_tag,
512 compat_HTMLParseError(f'closing {tag} tag not found'))
513 next_closing_tag_end = next_closing_tag_start + len(closing_tag)
515 parser.feed(html[offset:offset + next_closing_tag_end])
516 offset += next_closing_tag_end
517 except HTMLBreakOnClosingTagParser.HTMLBreakOnClosingTagException:
518 return html[content_start:offset + next_closing_tag_start], \
519 html[whole_start:offset + next_closing_tag_end]
520 raise compat_HTMLParseError('unexpected end of html')
523 class HTMLAttributeParser(compat_HTMLParser):
524 """Trivial HTML parser to gather the attributes
for a single element
"""
528 compat_HTMLParser.__init__(self)
530 def handle_starttag(self, tag, attrs):
531 self.attrs = dict(attrs)
534 class HTMLListAttrsParser(compat_HTMLParser):
535 """HTML parser to gather the attributes
for the elements of a
list"""
538 compat_HTMLParser.__init__(self)
542 def handle_starttag(self, tag, attrs):
543 if tag == 'li' and self._level == 0:
544 self.items.append(dict(attrs))
547 def handle_endtag(self, tag):
551 def extract_attributes(html_element):
552 """Given a string
for an HTML element such
as
554 a
="foo" B
="bar" c
="&98;az" d
=boz
555 empty
= noval entity
="&"
558 Decode
and return a dictionary of attributes
.
560 'a': 'foo', 'b': 'bar', c
: 'baz', d
: 'boz',
561 'empty': '', 'noval': None, 'entity': '&',
562 'sq': '"', 'dq': '\''
565 parser = HTMLAttributeParser()
566 with contextlib.suppress(compat_HTMLParseError):
567 parser.feed(html_element)
572 def parse_list(webpage):
573 """Given a string
for an series of HTML
<li
> elements
,
574 return a dictionary of their attributes
"""
575 parser = HTMLListAttrsParser()
581 def clean_html(html):
582 """Clean an HTML snippet into a readable string
"""
584 if html is None: # Convenience for sanitizing descriptions etc.
587 html = re.sub(r'\s+', ' ', html)
588 html = re.sub(r'(?u)\s?<\s?br\s?/?\s?>\s?', '\n', html)
589 html = re.sub(r'(?u)<\s?/\s?p\s?>\s?<\s?p[^>]*>', '\n', html)
591 html = re.sub('<.*?>', '', html)
592 # Replace html entities
593 html = unescapeHTML(html)
597 def sanitize_open(filename, open_mode):
598 """Try to
open the given filename
, and slightly tweak it
if this fails
.
600 Attempts to
open the given filename
. If this fails
, it tries to change
601 the filename slightly
, step by step
, until it
's either able to open it
602 or it fails and raises a final exception, like the standard open()
605 It returns the tuple (stream, definitive_file_name).
608 if sys.platform == 'win32
':
610 msvcrt.setmode(sys.stdout.fileno(), os.O_BINARY)
611 return (sys.stdout.buffer if hasattr(sys.stdout, 'buffer') else sys.stdout, filename)
613 for attempt in range(2):
616 if sys.platform == 'win32
':
617 # FIXME: An exclusive lock also locks the file from being read.
618 # Since windows locks are mandatory, don't lock the
file on
windows (for now
).
619 # Ref: https://github.com/yt-dlp/yt-dlp/issues/3124
620 raise LockingUnsupportedError()
621 stream
= locked_file(filename
, open_mode
, block
=False).__enter
__()
622 except LockingUnsupportedError
:
623 stream
= open(filename
, open_mode
)
624 return (stream
, filename
)
625 except OSError as err
:
626 if attempt
or err
.errno
in (errno
.EACCES
,):
628 old_filename
, filename
= filename
, sanitize_path(filename
)
629 if old_filename
== filename
:
633 def timeconvert(timestr
):
634 """Convert RFC 2822 defined time string into system timestamp"""
636 timetuple
= email
.utils
.parsedate_tz(timestr
)
637 if timetuple
is not None:
638 timestamp
= email
.utils
.mktime_tz(timetuple
)
642 def sanitize_filename(s
, restricted
=False, is_id
=NO_DEFAULT
):
643 """Sanitizes a string so it could be used as part of a filename.
644 @param restricted Use a stricter subset of allowed characters
645 @param is_id Whether this is an ID that should be kept unchanged if possible.
646 If unset, yt-dlp's new sanitization rules are in effect
651 def replace_insane(char
):
652 if restricted
and char
in ACCENT_CHARS
:
653 return ACCENT_CHARS
[char
]
654 elif not restricted
and char
== '\n':
656 elif char
== '?' or ord(char
) < 32 or ord(char
) == 127:
659 return '' if restricted
else '\''
661 return '\0_\0-' if restricted
else '\0 \0-'
662 elif char
in '\\/|*<>':
664 if restricted
and (char
in '!&\'()[]{}$;`^,#' or char
.isspace() or ord(char
) > 127):
668 s
= re
.sub(r
'[0-9]+(?::[0-9]+)+', lambda m
: m
.group(0).replace(':', '_'), s
) # Handle timestamps
669 result
= ''.join(map(replace_insane
, s
))
670 if is_id
is NO_DEFAULT
:
671 result
= re
.sub('(\0.)(?:(?=\\1)..)+', r
'\1', result
) # Remove repeated substitute chars
672 STRIP_RE
= '(?:\0.|[ _-])*'
673 result
= re
.sub(f
'^\0.{STRIP_RE}|{STRIP_RE}\0.$', '', result
) # Remove substitute chars from start/end
674 result
= result
.replace('\0', '') or '_'
677 while '__' in result
:
678 result
= result
.replace('__', '_')
679 result
= result
.strip('_')
680 # Common case of "Foreign band name - English song title"
681 if restricted
and result
.startswith('-_'):
683 if result
.startswith('-'):
684 result
= '_' + result
[len('-'):]
685 result
= result
.lstrip('.')
691 def sanitize_path(s
, force
=False):
692 """Sanitizes and normalizes path on Windows"""
693 if sys
.platform
== 'win32':
695 drive_or_unc
, _
= os
.path
.splitdrive(s
)
701 norm_path
= os
.path
.normpath(remove_start(s
, drive_or_unc
)).split(os
.path
.sep
)
705 path_part
if path_part
in ['.', '..'] else re
.sub(r
'(?:[/<>:"\|\\?\*]|[\s.]$)', '#', path_part
)
706 for path_part
in norm_path
]
708 sanitized_path
.insert(0, drive_or_unc
+ os
.path
.sep
)
709 elif force
and s
and s
[0] == os
.path
.sep
:
710 sanitized_path
.insert(0, os
.path
.sep
)
711 return os
.path
.join(*sanitized_path
)
714 def sanitize_url(url
):
715 # Prepend protocol-less URLs with `http:` scheme in order to mitigate
716 # the number of unwanted failures due to missing protocol
719 elif url
.startswith('//'):
720 return 'http:%s' % url
721 # Fix some common typos seen so far
723 # https://github.com/ytdl-org/youtube-dl/issues/15649
724 (r
'^httpss://', r
'https://'),
725 # https://bx1.be/lives/direct-tv/
726 (r
'^rmtp([es]?)://', r
'rtmp\1://'),
728 for mistake
, fixup
in COMMON_TYPOS
:
729 if re
.match(mistake
, url
):
730 return re
.sub(mistake
, fixup
, url
)
734 def extract_basic_auth(url
):
735 parts
= compat_urlparse
.urlsplit(url
)
736 if parts
.username
is None:
738 url
= compat_urlparse
.urlunsplit(parts
._replace
(netloc
=(
739 parts
.hostname
if parts
.port
is None
740 else '%s:%d' % (parts
.hostname
, parts
.port
))))
741 auth_payload
= base64
.b64encode(
742 ('%s:%s' % (parts
.username
, parts
.password
or '')).encode())
743 return url
, f
'Basic {auth_payload.decode()}'
746 def sanitized_Request(url
, *args
, **kwargs
):
747 url
, auth_header
= extract_basic_auth(escape_url(sanitize_url(url
)))
748 if auth_header
is not None:
749 headers
= args
[1] if len(args
) >= 2 else kwargs
.setdefault('headers', {})
750 headers
['Authorization'] = auth_header
751 return compat_urllib_request
.Request(url
, *args
, **kwargs
)
755 """Expand shell variables and ~"""
756 return os
.path
.expandvars(compat_expanduser(s
))
759 def orderedSet(iterable
):
760 """ Remove all duplicates from the input iterable """
768 def _htmlentity_transform(entity_with_semicolon
):
769 """Transforms an HTML entity to a character."""
770 entity
= entity_with_semicolon
[:-1]
772 # Known non-numeric HTML entity
773 if entity
in compat_html_entities
.name2codepoint
:
774 return compat_chr(compat_html_entities
.name2codepoint
[entity
])
776 # TODO: HTML5 allows entities without a semicolon. For example,
777 # 'Éric' should be decoded as 'Éric'.
778 if entity_with_semicolon
in compat_html_entities_html5
:
779 return compat_html_entities_html5
[entity_with_semicolon
]
781 mobj
= re
.match(r
'#(x[0-9a-fA-F]+|[0-9]+)', entity
)
783 numstr
= mobj
.group(1)
784 if numstr
.startswith('x'):
786 numstr
= '0%s' % numstr
789 # See https://github.com/ytdl-org/youtube-dl/issues/7518
790 with contextlib
.suppress(ValueError):
791 return compat_chr(int(numstr
, base
))
793 # Unknown entity in name, return its literal representation
794 return '&%s;' % entity
800 assert isinstance(s
, str)
803 r
'&([^&;]+;)', lambda m
: _htmlentity_transform(m
.group(1)), s
)
806 def escapeHTML(text
):
809 .replace('&', '&')
810 .replace('<', '<')
811 .replace('>', '>')
812 .replace('"', '"')
813 .replace("'", ''')
817 def process_communicate_or_kill(p
, *args
, **kwargs
):
819 return p
.communicate(*args
, **kwargs
)
820 except BaseException
: # Including KeyboardInterrupt
826 class Popen(subprocess
.Popen
):
827 if sys
.platform
== 'win32':
828 _startupinfo
= subprocess
.STARTUPINFO()
829 _startupinfo
.dwFlags |
= subprocess
.STARTF_USESHOWWINDOW
833 def __init__(self
, *args
, **kwargs
):
834 super().__init
__(*args
, **kwargs
, startupinfo
=self
._startupinfo
)
836 def communicate_or_kill(self
, *args
, **kwargs
):
837 return process_communicate_or_kill(self
, *args
, **kwargs
)
840 def get_subprocess_encoding():
841 if sys
.platform
== 'win32' and sys
.getwindowsversion()[0] >= 5:
842 # For subprocess calls, encode with locale encoding
843 # Refer to http://stackoverflow.com/a/9951851/35070
844 encoding
= preferredencoding()
846 encoding
= sys
.getfilesystemencoding()
852 def encodeFilename(s
, for_subprocess
=False):
853 assert isinstance(s
, str)
857 def decodeFilename(b
, for_subprocess
=False):
861 def encodeArgument(s
):
862 # Legacy code that uses byte strings
863 # Uncomment the following line after fixing all post processors
864 # assert isinstance(s, str), 'Internal error: %r should be of type %r, is %r' % (s, compat_str, type(s))
865 return s
if isinstance(s
, str) else s
.decode('ascii')
868 def decodeArgument(b
):
872 def decodeOption(optval
):
875 if isinstance(optval
, bytes):
876 optval
= optval
.decode(preferredencoding())
878 assert isinstance(optval
, compat_str
)
882 _timetuple
= collections
.namedtuple('Time', ('hours', 'minutes', 'seconds', 'milliseconds'))
885 def timetuple_from_msec(msec
):
886 secs
, msec
= divmod(msec
, 1000)
887 mins
, secs
= divmod(secs
, 60)
888 hrs
, mins
= divmod(mins
, 60)
889 return _timetuple(hrs
, mins
, secs
, msec
)
892 def formatSeconds(secs
, delim
=':', msec
=False):
893 time
= timetuple_from_msec(secs
* 1000)
895 ret
= '%d%s%02d%s%02d' % (time
.hours
, delim
, time
.minutes
, delim
, time
.seconds
)
897 ret
= '%d%s%02d' % (time
.minutes
, delim
, time
.seconds
)
899 ret
= '%d' % time
.seconds
900 return '%s.%03d' % (ret
, time
.milliseconds
) if msec
else ret
903 def _ssl_load_windows_store_certs(ssl_context
, storename
):
904 # Code adapted from _load_windows_store_certs in https://github.com/python/cpython/blob/main/Lib/ssl.py
906 certs
= [cert
for cert
, encoding
, trust
in ssl
.enum_certificates(storename
)
907 if encoding
== 'x509_asn' and (
908 trust
is True or ssl
.Purpose
.SERVER_AUTH
.oid
in trust
)]
909 except PermissionError
:
912 with contextlib
.suppress(ssl
.SSLError
):
913 ssl_context
.load_verify_locations(cadata
=cert
)
916 def make_HTTPS_handler(params
, **kwargs
):
917 opts_check_certificate
= not params
.get('nocheckcertificate')
918 context
= ssl
.SSLContext(ssl
.PROTOCOL_TLS_CLIENT
)
919 context
.check_hostname
= opts_check_certificate
920 if params
.get('legacyserverconnect'):
921 context
.options |
= 4 # SSL_OP_LEGACY_SERVER_CONNECT
922 # Allow use of weaker ciphers in Python 3.10+. See https://bugs.python.org/issue43998
923 context
.set_ciphers('DEFAULT')
924 context
.verify_mode
= ssl
.CERT_REQUIRED
if opts_check_certificate
else ssl
.CERT_NONE
925 if opts_check_certificate
:
926 if has_certifi
and 'no-certifi' not in params
.get('compat_opts', []):
927 context
.load_verify_locations(cafile
=certifi
.where())
930 context
.load_default_certs()
931 # Work around the issue in load_default_certs when there are bad certificates. See:
932 # https://github.com/yt-dlp/yt-dlp/issues/1060,
933 # https://bugs.python.org/issue35665, https://bugs.python.org/issue45312
935 # enum_certificates is not present in mingw python. See https://github.com/yt-dlp/yt-dlp/issues/1151
936 if sys
.platform
== 'win32' and hasattr(ssl
, 'enum_certificates'):
937 for storename
in ('CA', 'ROOT'):
938 _ssl_load_windows_store_certs(context
, storename
)
939 context
.set_default_verify_paths()
940 client_certfile
= params
.get('client_certificate')
943 context
.load_cert_chain(
944 client_certfile
, keyfile
=params
.get('client_certificate_key'),
945 password
=params
.get('client_certificate_password'))
947 raise YoutubeDLError('Unable to load client certificate')
948 return YoutubeDLHTTPSHandler(params
, context
=context
, **kwargs
)
951 def bug_reports_message(before
=';'):
952 msg
= ('please report this issue on https://github.com/yt-dlp/yt-dlp/issues?q= , '
953 'filling out the appropriate issue template. '
954 'Confirm you are on the latest version using yt-dlp -U')
956 before
= before
.rstrip()
957 if not before
or before
.endswith(('.', '!', '?')):
958 msg
= msg
[0].title() + msg
[1:]
960 return (before
+ ' ' if before
else '') + msg
963 class YoutubeDLError(Exception):
964 """Base exception for YoutubeDL errors."""
967 def __init__(self
, msg
=None):
970 elif self
.msg
is None:
971 self
.msg
= type(self
).__name
__
972 super().__init
__(self
.msg
)
975 network_exceptions
= [compat_urllib_error
.URLError
, compat_http_client
.HTTPException
, socket
.error
]
976 if hasattr(ssl
, 'CertificateError'):
977 network_exceptions
.append(ssl
.CertificateError
)
978 network_exceptions
= tuple(network_exceptions
)
981 class ExtractorError(YoutubeDLError
):
982 """Error during info extraction."""
984 def __init__(self
, msg
, tb
=None, expected
=False, cause
=None, video_id
=None, ie
=None):
985 """ tb, if given, is the original traceback (so that it can be printed out).
986 If expected is set, this is a normal error message and most likely not a bug in yt-dlp.
988 if sys
.exc_info()[0] in network_exceptions
:
991 self
.orig_msg
= str(msg
)
993 self
.expected
= expected
995 self
.video_id
= video_id
997 self
.exc_info
= sys
.exc_info() # preserve original exception
999 super().__init
__(''.join((
1000 format_field(ie
, template
='[%s] '),
1001 format_field(video_id
, template
='%s: '),
1003 format_field(cause
, template
=' (caused by %r)'),
1004 '' if expected
else bug_reports_message())))
1006 def format_traceback(self
):
1007 return join_nonempty(
1008 self
.traceback
and ''.join(traceback
.format_tb(self
.traceback
)),
1009 self
.cause
and ''.join(traceback
.format_exception(None, self
.cause
, self
.cause
.__traceback
__)[1:]),
1013 class UnsupportedError(ExtractorError
):
1014 def __init__(self
, url
):
1016 'Unsupported URL: %s' % url
, expected
=True)
1020 class RegexNotFoundError(ExtractorError
):
1021 """Error when a regex didn't match"""
1025 class GeoRestrictedError(ExtractorError
):
1026 """Geographic restriction Error exception.
1028 This exception may be thrown when a video is not available from your
1029 geographic location due to geographic restrictions imposed by a website.
1032 def __init__(self
, msg
, countries
=None, **kwargs
):
1033 kwargs
['expected'] = True
1034 super().__init
__(msg
, **kwargs
)
1035 self
.countries
= countries
1038 class DownloadError(YoutubeDLError
):
1039 """Download Error exception.
1041 This exception may be thrown by FileDownloader objects if they are not
1042 configured to continue on errors. They will contain the appropriate
1046 def __init__(self
, msg
, exc_info
=None):
1047 """ exc_info, if given, is the original exception that caused the trouble (as returned by sys.exc_info()). """
1048 super().__init
__(msg
)
1049 self
.exc_info
= exc_info
1052 class EntryNotInPlaylist(YoutubeDLError
):
1053 """Entry not in playlist exception.
1055 This exception will be thrown by YoutubeDL when a requested entry
1056 is not found in the playlist info_dict
1058 msg
= 'Entry not found in info'
1061 class SameFileError(YoutubeDLError
):
1062 """Same File exception.
1064 This exception will be thrown by FileDownloader objects if they detect
1065 multiple files would have to be downloaded to the same file on disk.
1067 msg
= 'Fixed output name but more than one file to download'
1069 def __init__(self
, filename
=None):
1070 if filename
is not None:
1071 self
.msg
+= f
': {filename}'
1072 super().__init
__(self
.msg
)
1075 class PostProcessingError(YoutubeDLError
):
1076 """Post Processing exception.
1078 This exception may be raised by PostProcessor's .run() method to
1079 indicate an error in the postprocessing task.
1083 class DownloadCancelled(YoutubeDLError
):
1084 """ Exception raised when the download queue should be interrupted """
1085 msg
= 'The download was cancelled'
1088 class ExistingVideoReached(DownloadCancelled
):
1089 """ --break-on-existing triggered """
1090 msg
= 'Encountered a video that is already in the archive, stopping due to --break-on-existing'
1093 class RejectedVideoReached(DownloadCancelled
):
1094 """ --break-on-reject triggered """
1095 msg
= 'Encountered a video that did not match filter, stopping due to --break-on-reject'
1098 class MaxDownloadsReached(DownloadCancelled
):
1099 """ --max-downloads limit has been reached. """
1100 msg
= 'Maximum number of downloads reached, stopping due to --max-downloads'
1103 class ReExtractInfo(YoutubeDLError
):
1104 """ Video info needs to be re-extracted. """
1106 def __init__(self
, msg
, expected
=False):
1107 super().__init
__(msg
)
1108 self
.expected
= expected
1111 class ThrottledDownload(ReExtractInfo
):
1112 """ Download speed below --throttled-rate. """
1113 msg
= 'The download speed is below throttle limit'
1116 super().__init
__(self
.msg
, expected
=False)
1119 class UnavailableVideoError(YoutubeDLError
):
1120 """Unavailable Format exception.
1122 This exception will be thrown when a video is requested
1123 in a format that is not available for that video.
1125 msg
= 'Unable to download video'
1127 def __init__(self
, err
=None):
1129 self
.msg
+= f
': {err}'
1130 super().__init
__(self
.msg
)
1133 class ContentTooShortError(YoutubeDLError
):
1134 """Content Too Short exception.
1136 This exception may be raised by FileDownloader objects when a file they
1137 download is too small for what the server announced first, indicating
1138 the connection was probably interrupted.
1141 def __init__(self
, downloaded
, expected
):
1142 super().__init
__(f
'Downloaded {downloaded} bytes, expected {expected} bytes')
1144 self
.downloaded
= downloaded
1145 self
.expected
= expected
1148 class XAttrMetadataError(YoutubeDLError
):
1149 def __init__(self
, code
=None, msg
='Unknown error'):
1150 super().__init
__(msg
)
1154 # Parsing code and msg
1155 if (self
.code
in (errno
.ENOSPC
, errno
.EDQUOT
)
1156 or 'No space left' in self
.msg
or 'Disk quota exceeded' in self
.msg
):
1157 self
.reason
= 'NO_SPACE'
1158 elif self
.code
== errno
.E2BIG
or 'Argument list too long' in self
.msg
:
1159 self
.reason
= 'VALUE_TOO_LONG'
1161 self
.reason
= 'NOT_SUPPORTED'
1164 class XAttrUnavailableError(YoutubeDLError
):
1168 def _create_http_connection(ydl_handler
, http_class
, is_https
, *args
, **kwargs
):
1169 hc
= http_class(*args
, **kwargs
)
1170 source_address
= ydl_handler
._params
.get('source_address')
1172 if source_address
is not None:
1173 # This is to workaround _create_connection() from socket where it will try all
1174 # address data from getaddrinfo() including IPv6. This filters the result from
1175 # getaddrinfo() based on the source_address value.
1176 # This is based on the cpython socket.create_connection() function.
1177 # https://github.com/python/cpython/blob/master/Lib/socket.py#L691
1178 def _create_connection(address
, timeout
=socket
._GLOBAL
_DEFAULT
_TIMEOUT
, source_address
=None):
1179 host
, port
= address
1181 addrs
= socket
.getaddrinfo(host
, port
, 0, socket
.SOCK_STREAM
)
1182 af
= socket
.AF_INET
if '.' in source_address
[0] else socket
.AF_INET6
1183 ip_addrs
= [addr
for addr
in addrs
if addr
[0] == af
]
1184 if addrs
and not ip_addrs
:
1185 ip_version
= 'v4' if af
== socket
.AF_INET
else 'v6'
1187 "No remote IP%s addresses available for connect, can't use '%s' as source address"
1188 % (ip_version
, source_address
[0]))
1189 for res
in ip_addrs
:
1190 af
, socktype
, proto
, canonname
, sa
= res
1193 sock
= socket
.socket(af
, socktype
, proto
)
1194 if timeout
is not socket
._GLOBAL
_DEFAULT
_TIMEOUT
:
1195 sock
.settimeout(timeout
)
1196 sock
.bind(source_address
)
1198 err
= None # Explicitly break reference cycle
1200 except OSError as _
:
1202 if sock
is not None:
1207 raise OSError('getaddrinfo returns an empty list')
1208 if hasattr(hc
, '_create_connection'):
1209 hc
._create
_connection
= _create_connection
1210 hc
.source_address
= (source_address
, 0)
1215 def handle_youtubedl_headers(headers
):
1216 filtered_headers
= headers
1218 if 'Youtubedl-no-compression' in filtered_headers
:
1219 filtered_headers
= {k: v for k, v in filtered_headers.items() if k.lower() != 'accept-encoding'}
1220 del filtered_headers
['Youtubedl-no-compression']
1222 return filtered_headers
1225 class YoutubeDLHandler(compat_urllib_request
.HTTPHandler
):
1226 """Handler for HTTP requests and responses.
1228 This class, when installed with an OpenerDirector, automatically adds
1229 the standard headers to every HTTP request and handles gzipped and
1230 deflated responses from web servers. If compression is to be avoided in
1231 a particular request, the original request in the program code only has
1232 to include the HTTP header "Youtubedl-no-compression", which will be
1233 removed before making the real request.
1235 Part of this code was copied from:
1237 http://techknack.net/python-urllib2-handlers/
1239 Andrew Rowls, the author of that code, agreed to release it to the
1243 def __init__(self
, params
, *args
, **kwargs
):
1244 compat_urllib_request
.HTTPHandler
.__init
__(self
, *args
, **kwargs
)
1245 self
._params
= params
1247 def http_open(self
, req
):
1248 conn_class
= compat_http_client
.HTTPConnection
1250 socks_proxy
= req
.headers
.get('Ytdl-socks-proxy')
1252 conn_class
= make_socks_conn_class(conn_class
, socks_proxy
)
1253 del req
.headers
['Ytdl-socks-proxy']
1255 return self
.do_open(functools
.partial(
1256 _create_http_connection
, self
, conn_class
, False),
1264 return zlib
.decompress(data
, -zlib
.MAX_WBITS
)
1266 return zlib
.decompress(data
)
1272 return brotli
.decompress(data
)
1274 def http_request(self
, req
):
1275 # According to RFC 3986, URLs can not contain non-ASCII characters, however this is not
1276 # always respected by websites, some tend to give out URLs with non percent-encoded
1277 # non-ASCII characters (see telemb.py, ard.py [#3412])
1278 # urllib chokes on URLs with non-ASCII characters (see http://bugs.python.org/issue3991)
1279 # To work around aforementioned issue we will replace request's original URL with
1280 # percent-encoded one
1281 # Since redirects are also affected (e.g. http://www.southpark.de/alle-episoden/s18e09)
1282 # the code of this workaround has been moved here from YoutubeDL.urlopen()
1283 url
= req
.get_full_url()
1284 url_escaped
= escape_url(url
)
1286 # Substitute URL if any change after escaping
1287 if url
!= url_escaped
:
1288 req
= update_Request(req
, url
=url_escaped
)
1290 for h
, v
in self
._params
.get('http_headers', std_headers
).items():
1291 # Capitalize is needed because of Python bug 2275: http://bugs.python.org/issue2275
1292 # The dict keys are capitalized because of this bug by urllib
1293 if h
.capitalize() not in req
.headers
:
1294 req
.add_header(h
, v
)
1296 if 'Accept-encoding' not in req
.headers
:
1297 req
.add_header('Accept-encoding', ', '.join(SUPPORTED_ENCODINGS
))
1299 req
.headers
= handle_youtubedl_headers(req
.headers
)
1303 def http_response(self
, req
, resp
):
1306 if resp
.headers
.get('Content-encoding', '') == 'gzip':
1307 content
= resp
.read()
1308 gz
= gzip
.GzipFile(fileobj
=io
.BytesIO(content
), mode
='rb')
1310 uncompressed
= io
.BytesIO(gz
.read())
1311 except OSError as original_ioerror
:
1312 # There may be junk add the end of the file
1313 # See http://stackoverflow.com/q/4928560/35070 for details
1314 for i
in range(1, 1024):
1316 gz
= gzip
.GzipFile(fileobj
=io
.BytesIO(content
[:-i
]), mode
='rb')
1317 uncompressed
= io
.BytesIO(gz
.read())
1322 raise original_ioerror
1323 resp
= compat_urllib_request
.addinfourl(uncompressed
, old_resp
.headers
, old_resp
.url
, old_resp
.code
)
1324 resp
.msg
= old_resp
.msg
1325 del resp
.headers
['Content-encoding']
1327 if resp
.headers
.get('Content-encoding', '') == 'deflate':
1328 gz
= io
.BytesIO(self
.deflate(resp
.read()))
1329 resp
= compat_urllib_request
.addinfourl(gz
, old_resp
.headers
, old_resp
.url
, old_resp
.code
)
1330 resp
.msg
= old_resp
.msg
1331 del resp
.headers
['Content-encoding']
1333 if resp
.headers
.get('Content-encoding', '') == 'br':
1334 resp
= compat_urllib_request
.addinfourl(
1335 io
.BytesIO(self
.brotli(resp
.read())), old_resp
.headers
, old_resp
.url
, old_resp
.code
)
1336 resp
.msg
= old_resp
.msg
1337 del resp
.headers
['Content-encoding']
1338 # Percent-encode redirect URL of Location HTTP header to satisfy RFC 3986 (see
1339 # https://github.com/ytdl-org/youtube-dl/issues/6457).
1340 if 300 <= resp
.code
< 400:
1341 location
= resp
.headers
.get('Location')
1343 # As of RFC 2616 default charset is iso-8859-1 that is respected by python 3
1344 location
= location
.encode('iso-8859-1').decode()
1345 location_escaped
= escape_url(location
)
1346 if location
!= location_escaped
:
1347 del resp
.headers
['Location']
1348 resp
.headers
['Location'] = location_escaped
1351 https_request
= http_request
1352 https_response
= http_response
1355 def make_socks_conn_class(base_class
, socks_proxy
):
1356 assert issubclass(base_class
, (
1357 compat_http_client
.HTTPConnection
, compat_http_client
.HTTPSConnection
))
1359 url_components
= compat_urlparse
.urlparse(socks_proxy
)
1360 if url_components
.scheme
.lower() == 'socks5':
1361 socks_type
= ProxyType
.SOCKS5
1362 elif url_components
.scheme
.lower() in ('socks', 'socks4'):
1363 socks_type
= ProxyType
.SOCKS4
1364 elif url_components
.scheme
.lower() == 'socks4a':
1365 socks_type
= ProxyType
.SOCKS4A
1367 def unquote_if_non_empty(s
):
1370 return compat_urllib_parse_unquote_plus(s
)
1374 url_components
.hostname
, url_components
.port
or 1080,
1376 unquote_if_non_empty(url_components
.username
),
1377 unquote_if_non_empty(url_components
.password
),
1380 class SocksConnection(base_class
):
1382 self
.sock
= sockssocket()
1383 self
.sock
.setproxy(*proxy_args
)
1384 if isinstance(self
.timeout
, (int, float)):
1385 self
.sock
.settimeout(self
.timeout
)
1386 self
.sock
.connect((self
.host
, self
.port
))
1388 if isinstance(self
, compat_http_client
.HTTPSConnection
):
1389 if hasattr(self
, '_context'): # Python > 2.6
1390 self
.sock
= self
._context
.wrap_socket(
1391 self
.sock
, server_hostname
=self
.host
)
1393 self
.sock
= ssl
.wrap_socket(self
.sock
)
1395 return SocksConnection
1398 class YoutubeDLHTTPSHandler(compat_urllib_request
.HTTPSHandler
):
1399 def __init__(self
, params
, https_conn_class
=None, *args
, **kwargs
):
1400 compat_urllib_request
.HTTPSHandler
.__init
__(self
, *args
, **kwargs
)
1401 self
._https
_conn
_class
= https_conn_class
or compat_http_client
.HTTPSConnection
1402 self
._params
= params
1404 def https_open(self
, req
):
1406 conn_class
= self
._https
_conn
_class
1408 if hasattr(self
, '_context'): # python > 2.6
1409 kwargs
['context'] = self
._context
1410 if hasattr(self
, '_check_hostname'): # python 3.x
1411 kwargs
['check_hostname'] = self
._check
_hostname
1413 socks_proxy
= req
.headers
.get('Ytdl-socks-proxy')
1415 conn_class
= make_socks_conn_class(conn_class
, socks_proxy
)
1416 del req
.headers
['Ytdl-socks-proxy']
1419 return self
.do_open(
1420 functools
.partial(_create_http_connection
, self
, conn_class
, True), req
, **kwargs
)
1421 except urllib
.error
.URLError
as e
:
1422 if (isinstance(e
.reason
, ssl
.SSLError
)
1423 and getattr(e
.reason
, 'reason', None) == 'SSLV3_ALERT_HANDSHAKE_FAILURE'):
1424 raise YoutubeDLError('SSLV3_ALERT_HANDSHAKE_FAILURE: Try using --legacy-server-connect')
1428 class YoutubeDLCookieJar(compat_cookiejar
.MozillaCookieJar
):
1430 See [1] for cookie file format.
1432 1. https://curl.haxx.se/docs/http-cookies.html
1434 _HTTPONLY_PREFIX
= '#HttpOnly_'
1436 _HEADER
= '''# Netscape HTTP Cookie File
1437 # This file is generated by yt-dlp. Do not edit.
1440 _CookieFileEntry
= collections
.namedtuple(
1442 ('domain_name', 'include_subdomains', 'path', 'https_only', 'expires_at', 'name', 'value'))
1444 def __init__(self
, filename
=None, *args
, **kwargs
):
1445 super().__init
__(None, *args
, **kwargs
)
1446 if self
.is_path(filename
):
1447 filename
= os
.fspath(filename
)
1448 self
.filename
= filename
1451 def _true_or_false(cndn
):
1452 return 'TRUE' if cndn
else 'FALSE'
1456 return isinstance(file, (str, bytes, os
.PathLike
))
1458 @contextlib.contextmanager
1459 def open(self
, file, *, write
=False):
1460 if self
.is_path(file):
1461 with open(file, 'w' if write
else 'r', encoding
='utf-8') as f
:
1468 def _really_save(self
, f
, ignore_discard
=False, ignore_expires
=False):
1471 if (not ignore_discard
and cookie
.discard
1472 or not ignore_expires
and cookie
.is_expired(now
)):
1474 name
, value
= cookie
.name
, cookie
.value
1476 # cookies.txt regards 'Set-Cookie: foo' as a cookie
1477 # with no name, whereas http.cookiejar regards it as a
1478 # cookie with no value.
1479 name
, value
= '', name
1480 f
.write('%s\n' % '\t'.join((
1482 self
._true
_or
_false
(cookie
.domain
.startswith('.')),
1484 self
._true
_or
_false
(cookie
.secure
),
1485 str_or_none(cookie
.expires
, default
=''),
1489 def save(self
, filename
=None, *args
, **kwargs
):
1491 Save cookies to a file.
1492 Code is taken from CPython 3.6
1493 https://github.com/python/cpython/blob/8d999cbf4adea053be6dbb612b9844635c4dfb8e/Lib/http/cookiejar.py#L2091-L2117 """
1495 if filename
is None:
1496 if self
.filename
is not None:
1497 filename
= self
.filename
1499 raise ValueError(compat_cookiejar
.MISSING_FILENAME_TEXT
)
1501 # Store session cookies with `expires` set to 0 instead of an empty string
1503 if cookie
.expires
is None:
1506 with self
.open(filename
, write
=True) as f
:
1507 f
.write(self
._HEADER
)
1508 self
._really
_save
(f
, *args
, **kwargs
)
1510 def load(self
, filename
=None, ignore_discard
=False, ignore_expires
=False):
1511 """Load cookies from a file."""
1512 if filename
is None:
1513 if self
.filename
is not None:
1514 filename
= self
.filename
1516 raise ValueError(compat_cookiejar
.MISSING_FILENAME_TEXT
)
1518 def prepare_line(line
):
1519 if line
.startswith(self
._HTTPONLY
_PREFIX
):
1520 line
= line
[len(self
._HTTPONLY
_PREFIX
):]
1521 # comments and empty lines are fine
1522 if line
.startswith('#') or not line
.strip():
1524 cookie_list
= line
.split('\t')
1525 if len(cookie_list
) != self
._ENTRY
_LEN
:
1526 raise compat_cookiejar
.LoadError('invalid length %d' % len(cookie_list
))
1527 cookie
= self
._CookieFileEntry
(*cookie_list
)
1528 if cookie
.expires_at
and not cookie
.expires_at
.isdigit():
1529 raise compat_cookiejar
.LoadError('invalid expires at %s' % cookie
.expires_at
)
1533 with self
.open(filename
) as f
:
1536 cf
.write(prepare_line(line
))
1537 except compat_cookiejar
.LoadError
as e
:
1538 if f
'{line.strip()} '[0] in '[{"':
1539 raise compat_cookiejar
.LoadError(
1540 'Cookies file must be Netscape formatted, not JSON. See '
1541 'https://github.com/ytdl-org/youtube-dl#how-do-i-pass-cookies-to-youtube-dl')
1542 write_string(f
'WARNING: skipping cookie file entry due to {e}: {line!r}\n')
1545 self
._really
_load
(cf
, filename
, ignore_discard
, ignore_expires
)
1546 # Session cookies are denoted by either `expires` field set to
1547 # an empty string or 0. MozillaCookieJar only recognizes the former
1548 # (see [1]). So we need force the latter to be recognized as session
1549 # cookies on our own.
1550 # Session cookies may be important for cookies-based authentication,
1551 # e.g. usually, when user does not check 'Remember me' check box while
1552 # logging in on a site, some important cookies are stored as session
1553 # cookies so that not recognizing them will result in failed login.
1554 # 1. https://bugs.python.org/issue17164
1556 # Treat `expires=0` cookies as session cookies
1557 if cookie
.expires
== 0:
1558 cookie
.expires
= None
1559 cookie
.discard
= True
1562 class YoutubeDLCookieProcessor(compat_urllib_request
.HTTPCookieProcessor
):
1563 def __init__(self
, cookiejar
=None):
1564 compat_urllib_request
.HTTPCookieProcessor
.__init
__(self
, cookiejar
)
1566 def http_response(self
, request
, response
):
1567 return compat_urllib_request
.HTTPCookieProcessor
.http_response(self
, request
, response
)
1569 https_request
= compat_urllib_request
.HTTPCookieProcessor
.http_request
1570 https_response
= http_response
1573 class YoutubeDLRedirectHandler(compat_urllib_request
.HTTPRedirectHandler
):
1574 """YoutubeDL redirect handler
1576 The code is based on HTTPRedirectHandler implementation from CPython [1].
1578 This redirect handler solves two issues:
1579 - ensures redirect URL is always unicode under python 2
1580 - introduces support for experimental HTTP response status code
1581 308 Permanent Redirect [2] used by some sites [3]
1583 1. https://github.com/python/cpython/blob/master/Lib/urllib/request.py
1584 2. https://developer.mozilla.org/en-US/docs/Web/HTTP/Status/308
1585 3. https://github.com/ytdl-org/youtube-dl/issues/28768
1588 http_error_301
= http_error_303
= http_error_307
= http_error_308
= compat_urllib_request
.HTTPRedirectHandler
.http_error_302
1590 def redirect_request(self
, req
, fp
, code
, msg
, headers
, newurl
):
1591 """Return a Request or None in response to a redirect.
1593 This is called by the http_error_30x methods when a
1594 redirection response is received. If a redirection should
1595 take place, return a new Request to allow http_error_30x to
1596 perform the redirect. Otherwise, raise HTTPError if no-one
1597 else should try to handle this url. Return None if you can't
1598 but another Handler might.
1600 m
= req
.get_method()
1601 if (not (code
in (301, 302, 303, 307, 308) and m
in ("GET", "HEAD")
1602 or code
in (301, 302, 303) and m
== "POST")):
1603 raise compat_HTTPError(req
.full_url
, code
, msg
, headers
, fp
)
1604 # Strictly (according to RFC 2616), 301 or 302 in response to
1605 # a POST MUST NOT cause a redirection without confirmation
1606 # from the user (of urllib.request, in this case). In practice,
1607 # essentially all clients do redirect in this case, so we do
1610 # Be conciliant with URIs containing a space. This is mainly
1611 # redundant with the more complete encoding done in http_error_302(),
1612 # but it is kept for compatibility with other callers.
1613 newurl
= newurl
.replace(' ', '%20')
1615 CONTENT_HEADERS
= ("content-length", "content-type")
1616 # NB: don't use dict comprehension for python 2.6 compatibility
1617 newheaders
= {k: v for k, v in req.headers.items() if k.lower() not in CONTENT_HEADERS}
1619 # A 303 must either use GET or HEAD for subsequent request
1620 # https://datatracker.ietf.org/doc/html/rfc7231#section-6.4.4
1621 if code
== 303 and m
!= 'HEAD':
1623 # 301 and 302 redirects are commonly turned into a GET from a POST
1624 # for subsequent requests by browsers, so we'll do the same.
1625 # https://datatracker.ietf.org/doc/html/rfc7231#section-6.4.2
1626 # https://datatracker.ietf.org/doc/html/rfc7231#section-6.4.3
1627 if code
in (301, 302) and m
== 'POST':
1630 return compat_urllib_request
.Request(
1631 newurl
, headers
=newheaders
, origin_req_host
=req
.origin_req_host
,
1632 unverifiable
=True, method
=m
)
1635 def extract_timezone(date_str
):
1638 ^.{8,}? # >=8 char non-TZ prefix, if present
1639 (?P<tz>Z| # just the UTC Z, or
1640 (?:(?<=.\b\d{4}|\b\d{2}:\d\d)| # preceded by 4 digits or hh:mm or
1641 (?<!.\b[a-zA-Z]{3}|[a-zA-Z]{4}|..\b\d\d)) # not preceded by 3 alpha word or >= 4 alpha or 2 digits
1642 [ ]? # optional space
1643 (?P<sign>\+|-) # +/-
1644 (?P<hours>[0-9]{2}):?(?P<minutes>[0-9]{2}) # hh[:]mm
1648 timezone
= datetime
.timedelta()
1650 date_str
= date_str
[:-len(m
.group('tz'))]
1651 if not m
.group('sign'):
1652 timezone
= datetime
.timedelta()
1654 sign
= 1 if m
.group('sign') == '+' else -1
1655 timezone
= datetime
.timedelta(
1656 hours
=sign
* int(m
.group('hours')),
1657 minutes
=sign
* int(m
.group('minutes')))
1658 return timezone
, date_str
1661 def parse_iso8601(date_str
, delimiter
='T', timezone
=None):
1662 """ Return a UNIX timestamp from the given date """
1664 if date_str
is None:
1667 date_str
= re
.sub(r
'\.[0-9]+', '', date_str
)
1669 if timezone
is None:
1670 timezone
, date_str
= extract_timezone(date_str
)
1672 with contextlib
.suppress(ValueError):
1673 date_format
= f
'%Y-%m-%d{delimiter}%H:%M:%S'
1674 dt
= datetime
.datetime
.strptime(date_str
, date_format
) - timezone
1675 return calendar
.timegm(dt
.timetuple())
1678 def date_formats(day_first
=True):
1679 return DATE_FORMATS_DAY_FIRST
if day_first
else DATE_FORMATS_MONTH_FIRST
1682 def unified_strdate(date_str
, day_first
=True):
1683 """Return a string with the date in the format YYYYMMDD"""
1685 if date_str
is None:
1689 date_str
= date_str
.replace(',', ' ')
1690 # Remove AM/PM + timezone
1691 date_str
= re
.sub(r
'(?i)\s*(?:AM|PM)(?:\s+[A-Z]+)?', '', date_str
)
1692 _
, date_str
= extract_timezone(date_str
)
1694 for expression
in date_formats(day_first
):
1695 with contextlib
.suppress(ValueError):
1696 upload_date
= datetime
.datetime
.strptime(date_str
, expression
).strftime('%Y%m%d')
1697 if upload_date
is None:
1698 timetuple
= email
.utils
.parsedate_tz(date_str
)
1700 with contextlib
.suppress(ValueError):
1701 upload_date
= datetime
.datetime(*timetuple
[:6]).strftime('%Y%m%d')
1702 if upload_date
is not None:
1703 return compat_str(upload_date
)
1706 def unified_timestamp(date_str
, day_first
=True):
1707 if date_str
is None:
1710 date_str
= re
.sub(r
'[,|]', '', date_str
)
1712 pm_delta
= 12 if re
.search(r
'(?i)PM', date_str
) else 0
1713 timezone
, date_str
= extract_timezone(date_str
)
1715 # Remove AM/PM + timezone
1716 date_str
= re
.sub(r
'(?i)\s*(?:AM|PM)(?:\s+[A-Z]+)?', '', date_str
)
1718 # Remove unrecognized timezones from ISO 8601 alike timestamps
1719 m
= re
.search(r
'\d{1,2}:\d{1,2}(?:\.\d+)?(?P<tz>\s*[A-Z]+)$', date_str
)
1721 date_str
= date_str
[:-len(m
.group('tz'))]
1723 # Python only supports microseconds, so remove nanoseconds
1724 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]+$', date_str
)
1726 date_str
= m
.group(1)
1728 for expression
in date_formats(day_first
):
1729 with contextlib
.suppress(ValueError):
1730 dt
= datetime
.datetime
.strptime(date_str
, expression
) - timezone
+ datetime
.timedelta(hours
=pm_delta
)
1731 return calendar
.timegm(dt
.timetuple())
1732 timetuple
= email
.utils
.parsedate_tz(date_str
)
1734 return calendar
.timegm(timetuple
) + pm_delta
* 3600
1737 def determine_ext(url
, default_ext
='unknown_video'):
1738 if url
is None or '.' not in url
:
1740 guess
= url
.partition('?')[0].rpartition('.')[2]
1741 if re
.match(r
'^[A-Za-z0-9]+$', guess
):
1743 # Try extract ext from URLs like http://example.com/foo/bar.mp4/?download
1744 elif guess
.rstrip('/') in KNOWN_EXTENSIONS
:
1745 return guess
.rstrip('/')
1750 def subtitles_filename(filename
, sub_lang
, sub_format
, expected_real_ext
=None):
1751 return replace_extension(filename
, sub_lang
+ '.' + sub_format
, expected_real_ext
)
1754 def datetime_from_str(date_str
, precision
='auto', format
='%Y%m%d'):
1756 Return a datetime object from a string.
1758 (now|today|yesterday|DATE)([+-]\d+(microsecond|second|minute|hour|day|week|month|year)s?)?
1760 @param format strftime format of DATE
1761 @param precision Round the datetime object: auto|microsecond|second|minute|hour|day
1762 auto: round to the unit provided in date_str (if applicable).
1764 auto_precision
= False
1765 if precision
== 'auto':
1766 auto_precision
= True
1767 precision
= 'microsecond'
1768 today
= datetime_round(datetime
.datetime
.utcnow(), precision
)
1769 if date_str
in ('now', 'today'):
1771 if date_str
== 'yesterday':
1772 return today
- datetime
.timedelta(days
=1)
1774 r
'(?P<start>.+)(?P<sign>[+-])(?P<time>\d+)(?P<unit>microsecond|second|minute|hour|day|week|month|year)s?',
1776 if match
is not None:
1777 start_time
= datetime_from_str(match
.group('start'), precision
, format
)
1778 time
= int(match
.group('time')) * (-1 if match
.group('sign') == '-' else 1)
1779 unit
= match
.group('unit')
1780 if unit
== 'month' or unit
== 'year':
1781 new_date
= datetime_add_months(start_time
, time
* 12 if unit
== 'year' else time
)
1787 delta
= datetime
.timedelta(**{unit + 's': time}
)
1788 new_date
= start_time
+ delta
1790 return datetime_round(new_date
, unit
)
1793 return datetime_round(datetime
.datetime
.strptime(date_str
, format
), precision
)
1796 def date_from_str(date_str
, format
='%Y%m%d', strict
=False):
1798 Return a date object from a string using datetime_from_str
1800 @param strict Restrict allowed patterns to "YYYYMMDD" and
1801 (now|today|yesterday)(-\d+(day|week|month|year)s?)?
1803 if strict
and not re
.fullmatch(r
'\d{8}|(now|today|yesterday)(-\d+(day|week|month|year)s?)?', date_str
):
1804 raise ValueError(f
'Invalid date format "{date_str}"')
1805 return datetime_from_str(date_str
, precision
='microsecond', format
=format
).date()
1808 def datetime_add_months(dt
, months
):
1809 """Increment/Decrement a datetime object by months."""
1810 month
= dt
.month
+ months
- 1
1811 year
= dt
.year
+ month
// 12
1812 month
= month
% 12 + 1
1813 day
= min(dt
.day
, calendar
.monthrange(year
, month
)[1])
1814 return dt
.replace(year
, month
, day
)
1817 def datetime_round(dt
, precision
='day'):
1819 Round a datetime object's time to a specific precision
1821 if precision
== 'microsecond':
1830 roundto
= lambda x
, n
: ((x
+ n
/ 2) // n
) * n
1831 timestamp
= calendar
.timegm(dt
.timetuple())
1832 return datetime
.datetime
.utcfromtimestamp(roundto(timestamp
, unit_seconds
[precision
]))
1835 def hyphenate_date(date_str
):
1837 Convert a date in 'YYYYMMDD' format to 'YYYY-MM-DD' format"""
1838 match
= re
.match(r
'^(\d\d\d\d)(\d\d)(\d\d)$', date_str
)
1839 if match
is not None:
1840 return '-'.join(match
.groups())
1846 """Represents a time interval between two dates"""
1848 def __init__(self
, start
=None, end
=None):
1849 """start and end must be strings in the format accepted by date"""
1850 if start
is not None:
1851 self
.start
= date_from_str(start
, strict
=True)
1853 self
.start
= datetime
.datetime
.min.date()
1855 self
.end
= date_from_str(end
, strict
=True)
1857 self
.end
= datetime
.datetime
.max.date()
1858 if self
.start
> self
.end
:
1859 raise ValueError('Date range: "%s" , the start date must be before the end date' % self
)
1863 """Returns a range that only contains the given day"""
1864 return cls(day
, day
)
1866 def __contains__(self
, date
):
1867 """Check if the date is in the range"""
1868 if not isinstance(date
, datetime
.date
):
1869 date
= date_from_str(date
)
1870 return self
.start
<= date
<= self
.end
1873 return f
'{self.start.isoformat()} - {self.end.isoformat()}'
1876 def platform_name():
1877 """ Returns the platform name as a compat_str """
1878 res
= platform
.platform()
1879 if isinstance(res
, bytes):
1880 res
= res
.decode(preferredencoding())
1882 assert isinstance(res
, compat_str
)
1887 def get_windows_version():
1888 ''' Get Windows version. None if it's not running on Windows '''
1889 if compat_os_name
== 'nt':
1890 return version_tuple(platform
.win32_ver()[1])
1895 def write_string(s
, out
=None, encoding
=None):
1896 assert isinstance(s
, str)
1897 out
= out
or sys
.stderr
1899 if compat_os_name
== 'nt' and supports_terminal_sequences(out
):
1900 s
= re
.sub(r
'([\r\n]+)', r
' \1', s
)
1903 if 'b' in getattr(out
, 'mode', ''):
1904 enc
= encoding
or preferredencoding()
1905 elif hasattr(out
, 'buffer'):
1907 enc
= encoding
or getattr(out
, 'encoding', None) or preferredencoding()
1909 out
.write(s
.encode(enc
, 'ignore') if enc
else s
)
1913 def bytes_to_intlist(bs
):
1916 if isinstance(bs
[0], int): # Python 3
1919 return [ord(c
) for c
in bs
]
1922 def intlist_to_bytes(xs
):
1925 return compat_struct_pack('%dB' % len(xs
), *xs
)
1928 class LockingUnsupportedError(IOError):
1929 msg
= 'File locking is not supported on this platform'
1932 super().__init
__(self
.msg
)
1935 # Cross-platform file locking
1936 if sys
.platform
== 'win32':
1937 import ctypes
.wintypes
1940 class OVERLAPPED(ctypes
.Structure
):
1942 ('Internal', ctypes
.wintypes
.LPVOID
),
1943 ('InternalHigh', ctypes
.wintypes
.LPVOID
),
1944 ('Offset', ctypes
.wintypes
.DWORD
),
1945 ('OffsetHigh', ctypes
.wintypes
.DWORD
),
1946 ('hEvent', ctypes
.wintypes
.HANDLE
),
1949 kernel32
= ctypes
.windll
.kernel32
1950 LockFileEx
= kernel32
.LockFileEx
1951 LockFileEx
.argtypes
= [
1952 ctypes
.wintypes
.HANDLE
, # hFile
1953 ctypes
.wintypes
.DWORD
, # dwFlags
1954 ctypes
.wintypes
.DWORD
, # dwReserved
1955 ctypes
.wintypes
.DWORD
, # nNumberOfBytesToLockLow
1956 ctypes
.wintypes
.DWORD
, # nNumberOfBytesToLockHigh
1957 ctypes
.POINTER(OVERLAPPED
) # Overlapped
1959 LockFileEx
.restype
= ctypes
.wintypes
.BOOL
1960 UnlockFileEx
= kernel32
.UnlockFileEx
1961 UnlockFileEx
.argtypes
= [
1962 ctypes
.wintypes
.HANDLE
, # hFile
1963 ctypes
.wintypes
.DWORD
, # dwReserved
1964 ctypes
.wintypes
.DWORD
, # nNumberOfBytesToLockLow
1965 ctypes
.wintypes
.DWORD
, # nNumberOfBytesToLockHigh
1966 ctypes
.POINTER(OVERLAPPED
) # Overlapped
1968 UnlockFileEx
.restype
= ctypes
.wintypes
.BOOL
1969 whole_low
= 0xffffffff
1970 whole_high
= 0x7fffffff
1972 def _lock_file(f
, exclusive
, block
):
1973 overlapped
= OVERLAPPED()
1974 overlapped
.Offset
= 0
1975 overlapped
.OffsetHigh
= 0
1976 overlapped
.hEvent
= 0
1977 f
._lock
_file
_overlapped
_p
= ctypes
.pointer(overlapped
)
1979 if not LockFileEx(msvcrt
.get_osfhandle(f
.fileno()),
1980 (0x2 if exclusive
else 0x0) |
(0x0 if block
else 0x1),
1981 0, whole_low
, whole_high
, f
._lock
_file
_overlapped
_p
):
1982 raise BlockingIOError('Locking file failed: %r' % ctypes
.FormatError())
1984 def _unlock_file(f
):
1985 assert f
._lock
_file
_overlapped
_p
1986 handle
= msvcrt
.get_osfhandle(f
.fileno())
1987 if not UnlockFileEx(handle
, 0, whole_low
, whole_high
, f
._lock
_file
_overlapped
_p
):
1988 raise OSError('Unlocking file failed: %r' % ctypes
.FormatError())
1994 def _lock_file(f
, exclusive
, block
):
1995 flags
= fcntl
.LOCK_EX
if exclusive
else fcntl
.LOCK_SH
1997 flags |
= fcntl
.LOCK_NB
1999 fcntl
.flock(f
, flags
)
2000 except BlockingIOError
:
2002 except OSError: # AOSP does not have flock()
2003 fcntl
.lockf(f
, flags
)
2005 def _unlock_file(f
):
2007 fcntl
.flock(f
, fcntl
.LOCK_UN
)
2009 fcntl
.lockf(f
, fcntl
.LOCK_UN
)
2013 def _lock_file(f
, exclusive
, block
):
2014 raise LockingUnsupportedError()
2016 def _unlock_file(f
):
2017 raise LockingUnsupportedError()
2023 def __init__(self
, filename
, mode
, block
=True, encoding
=None):
2024 if mode
not in {'r', 'rb', 'a', 'ab', 'w', 'wb'}
:
2025 raise NotImplementedError(mode
)
2026 self
.mode
, self
.block
= mode
, block
2028 writable
= any(f
in mode
for f
in 'wax+')
2029 readable
= any(f
in mode
for f
in 'r+')
2030 flags
= functools
.reduce(operator
.ior
, (
2031 getattr(os
, 'O_CLOEXEC', 0), # UNIX only
2032 getattr(os
, 'O_BINARY', 0), # Windows only
2033 getattr(os
, 'O_NOINHERIT', 0), # Windows only
2034 os
.O_CREAT
if writable
else 0, # O_TRUNC only after locking
2035 os
.O_APPEND
if 'a' in mode
else 0,
2036 os
.O_EXCL
if 'x' in mode
else 0,
2037 os
.O_RDONLY
if not writable
else os
.O_RDWR
if readable
else os
.O_WRONLY
,
2040 self
.f
= os
.fdopen(os
.open(filename
, flags
, 0o666), mode
, encoding
=encoding
)
2042 def __enter__(self
):
2043 exclusive
= 'r' not in self
.mode
2045 _lock_file(self
.f
, exclusive
, self
.block
)
2050 if 'w' in self
.mode
:
2053 except OSError as e
:
2054 if e
.errno
!= 29: # Illegal seek, expected when self.f is a FIFO
2062 _unlock_file(self
.f
)
2066 def __exit__(self
, *_
):
2075 def __getattr__(self
, attr
):
2076 return getattr(self
.f
, attr
)
2083 def get_filesystem_encoding():
2084 encoding
= sys
.getfilesystemencoding()
2085 return encoding
if encoding
is not None else 'utf-8'
2088 def shell_quote(args
):
2090 encoding
= get_filesystem_encoding()
2092 if isinstance(a
, bytes):
2093 # We may get a filename encoded with 'encodeFilename'
2094 a
= a
.decode(encoding
)
2095 quoted_args
.append(compat_shlex_quote(a
))
2096 return ' '.join(quoted_args
)
2099 def smuggle_url(url
, data
):
2100 """ Pass additional data in a URL for internal use. """
2102 url
, idata
= unsmuggle_url(url
, {})
2104 sdata
= compat_urllib_parse_urlencode(
2105 {'__youtubedl_smuggle': json.dumps(data)}
)
2106 return url
+ '#' + sdata
2109 def unsmuggle_url(smug_url
, default
=None):
2110 if '#__youtubedl_smuggle' not in smug_url
:
2111 return smug_url
, default
2112 url
, _
, sdata
= smug_url
.rpartition('#')
2113 jsond
= compat_parse_qs(sdata
)['__youtubedl_smuggle'][0]
2114 data
= json
.loads(jsond
)
2118 def format_decimal_suffix(num
, fmt
='%d%s', *, factor
=1000):
2119 """ Formats numbers with decimal sufixes like K, M, etc """
2120 num
, factor
= float_or_none(num
), float(factor
)
2121 if num
is None or num
< 0:
2123 POSSIBLE_SUFFIXES
= 'kMGTPEZY'
2124 exponent
= 0 if num
== 0 else min(int(math
.log(num
, factor
)), len(POSSIBLE_SUFFIXES
))
2125 suffix
= ['', *POSSIBLE_SUFFIXES
][exponent
]
2127 suffix
= {'k': 'Ki', '': ''}
.get(suffix
, f
'{suffix}i')
2128 converted
= num
/ (factor
** exponent
)
2129 return fmt
% (converted
, suffix
)
2132 def format_bytes(bytes):
2133 return format_decimal_suffix(bytes, '%.2f%sB', factor
=1024) or 'N/A'
2136 def lookup_unit_table(unit_table
, s
):
2137 units_re
= '|'.join(re
.escape(u
) for u
in unit_table
)
2139 r
'(?P<num>[0-9]+(?:[,.][0-9]*)?)\s*(?P<unit>%s)\b' % units_re
, s
)
2142 num_str
= m
.group('num').replace(',', '.')
2143 mult
= unit_table
[m
.group('unit')]
2144 return int(float(num_str
) * mult
)
2147 def parse_filesize(s
):
2151 # The lower-case forms are of course incorrect and unofficial,
2152 # but we support those too
2169 'megabytes': 1000 ** 2,
2170 'mebibytes': 1024 ** 2,
2176 'gigabytes': 1000 ** 3,
2177 'gibibytes': 1024 ** 3,
2183 'terabytes': 1000 ** 4,
2184 'tebibytes': 1024 ** 4,
2190 'petabytes': 1000 ** 5,
2191 'pebibytes': 1024 ** 5,
2197 'exabytes': 1000 ** 6,
2198 'exbibytes': 1024 ** 6,
2204 'zettabytes': 1000 ** 7,
2205 'zebibytes': 1024 ** 7,
2211 'yottabytes': 1000 ** 8,
2212 'yobibytes': 1024 ** 8,
2215 return lookup_unit_table(_UNIT_TABLE
, s
)
2222 s
= re
.sub(r
'^[^\d]+\s', '', s
).strip()
2224 if re
.match(r
'^[\d,.]+$', s
):
2225 return str_to_int(s
)
2238 ret
= lookup_unit_table(_UNIT_TABLE
, s
)
2242 mobj
= re
.match(r
'([\d,.]+)(?:$|\s)', s
)
2244 return str_to_int(mobj
.group(1))
2247 def parse_resolution(s
, *, lenient
=False):
2252 mobj
= re
.search(r
'(?P<w>\d+)\s*[xX×,]\s*(?P<h>\d+)', s
)
2254 mobj
= re
.search(r
'(?<![a-zA-Z0-9])(?P<w>\d+)\s*[xX×,]\s*(?P<h>\d+)(?![a-zA-Z0-9])', s
)
2257 'width': int(mobj
.group('w')),
2258 'height': int(mobj
.group('h')),
2261 mobj
= re
.search(r
'(?<![a-zA-Z0-9])(\d+)[pPiI](?![a-zA-Z0-9])', s
)
2263 return {'height': int(mobj.group(1))}
2265 mobj
= re
.search(r
'\b([48])[kK]\b', s
)
2267 return {'height': int(mobj.group(1)) * 540}
2272 def parse_bitrate(s
):
2273 if not isinstance(s
, compat_str
):
2275 mobj
= re
.search(r
'\b(\d+)\s*kbps', s
)
2277 return int(mobj
.group(1))
2280 def month_by_name(name
, lang
='en'):
2281 """ Return the number of a month by (locale-independently) English name """
2283 month_names
= MONTH_NAMES
.get(lang
, MONTH_NAMES
['en'])
2286 return month_names
.index(name
) + 1
2291 def month_by_abbreviation(abbrev
):
2292 """ Return the number of a month by (locale-independently) English
2296 return [s
[:3] for s
in ENGLISH_MONTH_NAMES
].index(abbrev
) + 1
2301 def fix_xml_ampersands(xml_str
):
2302 """Replace all the '&' by '&' in XML"""
2304 r
'&(?!amp;|lt;|gt;|apos;|quot;|#x[0-9a-fA-F]{,4};|#[0-9]{,4};)',
2309 def setproctitle(title
):
2310 assert isinstance(title
, compat_str
)
2312 # ctypes in Jython is not complete
2313 # http://bugs.jython.org/issue2148
2314 if sys
.platform
.startswith('java'):
2318 libc
= ctypes
.cdll
.LoadLibrary('libc.so.6')
2322 # LoadLibrary in Windows Python 2.7.13 only expects
2323 # a bytestring, but since unicode_literals turns
2324 # every string into a unicode string, it fails.
2326 title_bytes
= title
.encode()
2327 buf
= ctypes
.create_string_buffer(len(title_bytes
))
2328 buf
.value
= title_bytes
2330 libc
.prctl(15, buf
, 0, 0, 0)
2331 except AttributeError:
2332 return # Strange libc, just skip this
2335 def remove_start(s
, start
):
2336 return s
[len(start
):] if s
is not None and s
.startswith(start
) else s
2339 def remove_end(s
, end
):
2340 return s
[:-len(end
)] if s
is not None and s
.endswith(end
) else s
2343 def remove_quotes(s
):
2344 if s
is None or len(s
) < 2:
2346 for quote
in ('"', "'", ):
2347 if s
[0] == quote
and s
[-1] == quote
:
2352 def get_domain(url
):
2353 domain
= re
.match(r
'(?:https?:\/\/)?(?:www\.)?(?P<domain>[^\n\/]+\.[^\n\/]+)(?:\/(.*))?', url
)
2354 return domain
.group('domain') if domain
else None
2357 def url_basename(url
):
2358 path
= compat_urlparse
.urlparse(url
).path
2359 return path
.strip('/').split('/')[-1]
2363 return re
.match(r
'https?://[^?#&]+/', url
).group()
2366 def urljoin(base
, path
):
2367 if isinstance(path
, bytes):
2368 path
= path
.decode()
2369 if not isinstance(path
, compat_str
) or not path
:
2371 if re
.match(r
'^(?:[a-zA-Z][a-zA-Z0-9+-.]*:)?//', path
):
2373 if isinstance(base
, bytes):
2374 base
= base
.decode()
2375 if not isinstance(base
, compat_str
) or not re
.match(
2376 r
'^(?:https?:)?//', base
):
2378 return compat_urlparse
.urljoin(base
, path
)
2381 class HEADRequest(compat_urllib_request
.Request
):
2382 def get_method(self
):
2386 class PUTRequest(compat_urllib_request
.Request
):
2387 def get_method(self
):
2391 def int_or_none(v
, scale
=1, default
=None, get_attr
=None, invscale
=1):
2392 if get_attr
and v
is not None:
2393 v
= getattr(v
, get_attr
, None)
2395 return int(v
) * invscale
// scale
2396 except (ValueError, TypeError, OverflowError):
2400 def str_or_none(v
, default
=None):
2401 return default
if v
is None else compat_str(v
)
2404 def str_to_int(int_str
):
2405 """ A more relaxed version of int_or_none """
2406 if isinstance(int_str
, int):
2408 elif isinstance(int_str
, compat_str
):
2409 int_str
= re
.sub(r
'[,\.\+]', '', int_str
)
2410 return int_or_none(int_str
)
2413 def float_or_none(v
, scale
=1, invscale
=1, default
=None):
2417 return float(v
) * invscale
/ scale
2418 except (ValueError, TypeError):
2422 def bool_or_none(v
, default
=None):
2423 return v
if isinstance(v
, bool) else default
2426 def strip_or_none(v
, default
=None):
2427 return v
.strip() if isinstance(v
, compat_str
) else default
2430 def url_or_none(url
):
2431 if not url
or not isinstance(url
, compat_str
):
2434 return url
if re
.match(r
'^(?:(?:https?|rt(?:m(?:pt?[es]?|fp)|sp[su]?)|mms|ftps?):)?//', url
) else None
2437 def request_to_url(req
):
2438 if isinstance(req
, compat_urllib_request
.Request
):
2439 return req
.get_full_url()
2444 def strftime_or_none(timestamp
, date_format
, default
=None):
2445 datetime_object
= None
2447 if isinstance(timestamp
, (int, float)): # unix timestamp
2448 datetime_object
= datetime
.datetime
.utcfromtimestamp(timestamp
)
2449 elif isinstance(timestamp
, compat_str
): # assume YYYYMMDD
2450 datetime_object
= datetime
.datetime
.strptime(timestamp
, '%Y%m%d')
2451 return datetime_object
.strftime(date_format
)
2452 except (ValueError, TypeError, AttributeError):
2456 def parse_duration(s
):
2457 if not isinstance(s
, str):
2463 days
, hours
, mins
, secs
, ms
= [None] * 5
2464 m
= re
.match(r
'''(?x)
2466 (?:(?:(?P<days>[0-9]+):)?(?P<hours>[0-9]+):)?(?P<mins>[0-9]+):)?
2467 (?P<secs>(?(before_secs)[0-9]{1,2}|[0-9]+))
2468 (?P<ms>[.:][0-9]+)?Z?$
2471 days
, hours
, mins
, secs
, ms
= m
.group('days', 'hours', 'mins', 'secs', 'ms')
2476 [0-9]+\s*y(?:ears?)?,?\s*
2479 [0-9]+\s*m(?:onths?)?,?\s*
2482 [0-9]+\s*w(?:eeks?)?,?\s*
2485 (?P<days>[0-9]+)\s*d(?:ays?)?,?\s*
2489 (?P<hours>[0-9]+)\s*h(?:ours?)?,?\s*
2492 (?P<mins>[0-9]+)\s*m(?:in(?:ute)?s?)?,?\s*
2495 (?P<secs>[0-9]+)(?P<ms>\.[0-9]+)?\s*s(?:ec(?:ond)?s?)?\s*
2498 days
, hours
, mins
, secs
, ms
= m
.groups()
2500 m
= re
.match(r
'(?i)(?:(?P<hours>[0-9.]+)\s*(?:hours?)|(?P<mins>[0-9.]+)\s*(?:mins?\.?|minutes?)\s*)Z?$', s
)
2502 hours
, mins
= m
.groups()
2507 ms
= ms
.replace(':', '.')
2508 return sum(float(part
or 0) * mult
for part
, mult
in (
2509 (days
, 86400), (hours
, 3600), (mins
, 60), (secs
, 1), (ms
, 1)))
2512 def prepend_extension(filename
, ext
, expected_real_ext
=None):
2513 name
, real_ext
= os
.path
.splitext(filename
)
2515 f
'{name}.{ext}{real_ext}'
2516 if not expected_real_ext
or real_ext
[1:] == expected_real_ext
2517 else f
'{filename}.{ext}')
2520 def replace_extension(filename
, ext
, expected_real_ext
=None):
2521 name
, real_ext
= os
.path
.splitext(filename
)
2522 return '{}.{}'.format(
2523 name
if not expected_real_ext
or real_ext
[1:] == expected_real_ext
else filename
,
2527 def check_executable(exe
, args
=[]):
2528 """ Checks if the given binary is installed somewhere in PATH, and returns its name.
2529 args can be a list of arguments for a short output (like -version) """
2531 Popen([exe
] + args
, stdout
=subprocess
.PIPE
, stderr
=subprocess
.PIPE
).communicate_or_kill()
2537 def _get_exe_version_output(exe
, args
, *, to_screen
=None):
2539 to_screen(f
'Checking exe version: {shell_quote([exe] + args)}')
2541 # STDIN should be redirected too. On UNIX-like systems, ffmpeg triggers
2542 # SIGTTOU if yt-dlp is run in the background.
2543 # See https://github.com/ytdl-org/youtube-dl/issues/955#issuecomment-209789656
2545 [encodeArgument(exe
)] + args
, stdin
=subprocess
.PIPE
,
2546 stdout
=subprocess
.PIPE
, stderr
=subprocess
.STDOUT
).communicate_or_kill()
2549 if isinstance(out
, bytes): # Python 2.x
2550 out
= out
.decode('ascii', 'ignore')
2554 def detect_exe_version(output
, version_re
=None, unrecognized
='present'):
2555 assert isinstance(output
, compat_str
)
2556 if version_re
is None:
2557 version_re
= r
'version\s+([-0-9._a-zA-Z]+)'
2558 m
= re
.search(version_re
, output
)
2565 def get_exe_version(exe
, args
=['--version'],
2566 version_re
=None, unrecognized
='present'):
2567 """ Returns the version of the specified executable,
2568 or False if the executable is not present """
2569 out
= _get_exe_version_output(exe
, args
)
2570 return detect_exe_version(out
, version_re
, unrecognized
) if out
else False
2573 class LazyList(collections
.abc
.Sequence
):
2574 """Lazy immutable list from an iterable
2575 Note that slices of a LazyList are lists and not LazyList"""
2577 class IndexError(IndexError):
2580 def __init__(self
, iterable
, *, reverse
=False, _cache
=None):
2581 self
._iterable
= iter(iterable
)
2582 self
._cache
= [] if _cache
is None else _cache
2583 self
._reversed
= reverse
2587 # We need to consume the entire iterable to iterate in reverse
2588 yield from self
.exhaust()
2590 yield from self
._cache
2591 for item
in self
._iterable
:
2592 self
._cache
.append(item
)
2596 self
._cache
.extend(self
._iterable
)
2597 self
._iterable
= [] # Discard the emptied iterable to make it pickle-able
2601 """Evaluate the entire iterable"""
2602 return self
._exhaust
()[::-1 if self
._reversed
else 1]
2605 def _reverse_index(x
):
2606 return None if x
is None else -(x
+ 1)
2608 def __getitem__(self
, idx
):
2609 if isinstance(idx
, slice):
2611 idx
= slice(self
._reverse
_index
(idx
.start
), self
._reverse
_index
(idx
.stop
), -(idx
.step
or 1))
2612 start
, stop
, step
= idx
.start
, idx
.stop
, idx
.step
or 1
2613 elif isinstance(idx
, int):
2615 idx
= self
._reverse
_index
(idx
)
2616 start
, stop
, step
= idx
, idx
, 0
2618 raise TypeError('indices must be integers or slices')
2619 if ((start
or 0) < 0 or (stop
or 0) < 0
2620 or (start
is None and step
< 0)
2621 or (stop
is None and step
> 0)):
2622 # We need to consume the entire iterable to be able to slice from the end
2623 # Obviously, never use this with infinite iterables
2626 return self
._cache
[idx
]
2627 except IndexError as e
:
2628 raise self
.IndexError(e
) from e
2629 n
= max(start
or 0, stop
or 0) - len(self
._cache
) + 1
2631 self
._cache
.extend(itertools
.islice(self
._iterable
, n
))
2633 return self
._cache
[idx
]
2634 except IndexError as e
:
2635 raise self
.IndexError(e
) from e
2639 self
[-1] if self
._reversed
else self
[0]
2640 except self
.IndexError:
2646 return len(self
._cache
)
2648 def __reversed__(self
):
2649 return type(self
)(self
._iterable
, reverse
=not self
._reversed
, _cache
=self
._cache
)
2652 return type(self
)(self
._iterable
, reverse
=self
._reversed
, _cache
=self
._cache
)
2655 # repr and str should mimic a list. So we exhaust the iterable
2656 return repr(self
.exhaust())
2659 return repr(self
.exhaust())
2664 class IndexError(IndexError):
2668 # This is only useful for tests
2669 return len(self
.getslice())
2671 def __init__(self
, pagefunc
, pagesize
, use_cache
=True):
2672 self
._pagefunc
= pagefunc
2673 self
._pagesize
= pagesize
2674 self
._pagecount
= float('inf')
2675 self
._use
_cache
= use_cache
2678 def getpage(self
, pagenum
):
2679 page_results
= self
._cache
.get(pagenum
)
2680 if page_results
is None:
2681 page_results
= [] if pagenum
> self
._pagecount
else list(self
._pagefunc
(pagenum
))
2683 self
._cache
[pagenum
] = page_results
2686 def getslice(self
, start
=0, end
=None):
2687 return list(self
._getslice
(start
, end
))
2689 def _getslice(self
, start
, end
):
2690 raise NotImplementedError('This method must be implemented by subclasses')
2692 def __getitem__(self
, idx
):
2693 assert self
._use
_cache
, 'Indexing PagedList requires cache'
2694 if not isinstance(idx
, int) or idx
< 0:
2695 raise TypeError('indices must be non-negative integers')
2696 entries
= self
.getslice(idx
, idx
+ 1)
2698 raise self
.IndexError()
2702 class OnDemandPagedList(PagedList
):
2703 """Download pages until a page with less than maximum results"""
2705 def _getslice(self
, start
, end
):
2706 for pagenum
in itertools
.count(start
// self
._pagesize
):
2707 firstid
= pagenum
* self
._pagesize
2708 nextfirstid
= pagenum
* self
._pagesize
+ self
._pagesize
2709 if start
>= nextfirstid
:
2713 start
% self
._pagesize
2714 if firstid
<= start
< nextfirstid
2717 ((end
- 1) % self
._pagesize
) + 1
2718 if (end
is not None and firstid
<= end
<= nextfirstid
)
2722 page_results
= self
.getpage(pagenum
)
2724 self
._pagecount
= pagenum
- 1
2726 if startv
!= 0 or endv
is not None:
2727 page_results
= page_results
[startv
:endv
]
2728 yield from page_results
2730 # A little optimization - if current page is not "full", ie. does
2731 # not contain page_size videos then we can assume that this page
2732 # is the last one - there are no more ids on further pages -
2733 # i.e. no need to query again.
2734 if len(page_results
) + startv
< self
._pagesize
:
2737 # If we got the whole page, but the next page is not interesting,
2738 # break out early as well
2739 if end
== nextfirstid
:
2743 class InAdvancePagedList(PagedList
):
2744 """PagedList with total number of pages known in advance"""
2746 def __init__(self
, pagefunc
, pagecount
, pagesize
):
2747 PagedList
.__init
__(self
, pagefunc
, pagesize
, True)
2748 self
._pagecount
= pagecount
2750 def _getslice(self
, start
, end
):
2751 start_page
= start
// self
._pagesize
2752 end_page
= self
._pagecount
if end
is None else min(self
._pagecount
, end
// self
._pagesize
+ 1)
2753 skip_elems
= start
- start_page
* self
._pagesize
2754 only_more
= None if end
is None else end
- start
2755 for pagenum
in range(start_page
, end_page
):
2756 page_results
= self
.getpage(pagenum
)
2758 page_results
= page_results
[skip_elems
:]
2760 if only_more
is not None:
2761 if len(page_results
) < only_more
:
2762 only_more
-= len(page_results
)
2764 yield from page_results
[:only_more
]
2766 yield from page_results
2769 def uppercase_escape(s
):
2770 unicode_escape
= codecs
.getdecoder('unicode_escape')
2772 r
'\\U[0-9a-fA-F]{8}',
2773 lambda m
: unicode_escape(m
.group(0))[0],
2777 def lowercase_escape(s
):
2778 unicode_escape
= codecs
.getdecoder('unicode_escape')
2780 r
'\\u[0-9a-fA-F]{4}',
2781 lambda m
: unicode_escape(m
.group(0))[0],
2785 def escape_rfc3986(s
):
2786 """Escape non-ASCII characters as suggested by RFC 3986"""
2787 return urllib
.parse
.quote(s
, b
"%/;:@&=+$,!~*'()?#[]")
2790 def escape_url(url
):
2791 """Escape URL as suggested by RFC 3986"""
2792 url_parsed
= compat_urllib_parse_urlparse(url
)
2793 return url_parsed
._replace
(
2794 netloc
=url_parsed
.netloc
.encode('idna').decode('ascii'),
2795 path
=escape_rfc3986(url_parsed
.path
),
2796 params
=escape_rfc3986(url_parsed
.params
),
2797 query
=escape_rfc3986(url_parsed
.query
),
2798 fragment
=escape_rfc3986(url_parsed
.fragment
)
2803 return compat_parse_qs(compat_urllib_parse_urlparse(url
).query
)
2806 def read_batch_urls(batch_fd
):
2808 if not isinstance(url
, compat_str
):
2809 url
= url
.decode('utf-8', 'replace')
2810 BOM_UTF8
= ('\xef\xbb\xbf', '\ufeff')
2811 for bom
in BOM_UTF8
:
2812 if url
.startswith(bom
):
2813 url
= url
[len(bom
):]
2815 if not url
or url
.startswith(('#', ';', ']')):
2817 # "#" cannot be stripped out since it is part of the URI
2818 # However, it can be safely stipped out if follwing a whitespace
2819 return re
.split(r
'\s#', url
, 1)[0].rstrip()
2821 with contextlib
.closing(batch_fd
) as fd
:
2822 return [url
for url
in map(fixup
, fd
) if url
]
2825 def urlencode_postdata(*args
, **kargs
):
2826 return compat_urllib_parse_urlencode(*args
, **kargs
).encode('ascii')
2829 def update_url_query(url
, query
):
2832 parsed_url
= compat_urlparse
.urlparse(url
)
2833 qs
= compat_parse_qs(parsed_url
.query
)
2835 return compat_urlparse
.urlunparse(parsed_url
._replace
(
2836 query
=compat_urllib_parse_urlencode(qs
, True)))
2839 def update_Request(req
, url
=None, data
=None, headers
={}, query={}
):
2840 req_headers
= req
.headers
.copy()
2841 req_headers
.update(headers
)
2842 req_data
= data
or req
.data
2843 req_url
= update_url_query(url
or req
.get_full_url(), query
)
2844 req_get_method
= req
.get_method()
2845 if req_get_method
== 'HEAD':
2846 req_type
= HEADRequest
2847 elif req_get_method
== 'PUT':
2848 req_type
= PUTRequest
2850 req_type
= compat_urllib_request
.Request
2852 req_url
, data
=req_data
, headers
=req_headers
,
2853 origin_req_host
=req
.origin_req_host
, unverifiable
=req
.unverifiable
)
2854 if hasattr(req
, 'timeout'):
2855 new_req
.timeout
= req
.timeout
2859 def _multipart_encode_impl(data
, boundary
):
2860 content_type
= 'multipart/form-data; boundary=%s' % boundary
2863 for k
, v
in data
.items():
2864 out
+= b
'--' + boundary
.encode('ascii') + b
'\r\n'
2865 if isinstance(k
, compat_str
):
2867 if isinstance(v
, compat_str
):
2869 # RFC 2047 requires non-ASCII field names to be encoded, while RFC 7578
2870 # suggests sending UTF-8 directly. Firefox sends UTF-8, too
2871 content
= b
'Content-Disposition: form-data; name="' + k
+ b
'"\r\n\r\n' + v
+ b
'\r\n'
2872 if boundary
.encode('ascii') in content
:
2873 raise ValueError('Boundary overlaps with data')
2876 out
+= b
'--' + boundary
.encode('ascii') + b
'--\r\n'
2878 return out
, content_type
2881 def multipart_encode(data
, boundary
=None):
2883 Encode a dict to RFC 7578-compliant form-data
2886 A dict where keys and values can be either Unicode or bytes-like
2889 If specified a Unicode object, it's used as the boundary. Otherwise
2890 a random boundary is generated.
2892 Reference: https://tools.ietf.org/html/rfc7578
2894 has_specified_boundary
= boundary
is not None
2897 if boundary
is None:
2898 boundary
= '---------------' + str(random
.randrange(0x0fffffff, 0xffffffff))
2901 out
, content_type
= _multipart_encode_impl(data
, boundary
)
2904 if has_specified_boundary
:
2908 return out
, content_type
2911 def dict_get(d
, key_or_keys
, default
=None, skip_false_values
=True):
2912 for val
in map(d
.get
, variadic(key_or_keys
)):
2913 if val
is not None and (val
or not skip_false_values
):
2918 def try_call(*funcs
, expected_type
=None, args
=[], kwargs
={}):
2921 val
= f(*args
, **kwargs
)
2922 except (AttributeError, KeyError, TypeError, IndexError, ZeroDivisionError):
2925 if expected_type
is None or isinstance(val
, expected_type
):
2929 def try_get(src
, getter
, expected_type
=None):
2930 return try_call(*variadic(getter
), args
=(src
,), expected_type
=expected_type
)
2933 def filter_dict(dct
, cndn
=lambda _
, v
: v
is not None):
2934 return {k: v for k, v in dct.items() if cndn(k, v)}
2937 def merge_dicts(*dicts
):
2939 for a_dict
in dicts
:
2940 for k
, v
in a_dict
.items():
2941 if (v
is not None and k
not in merged
2942 or isinstance(v
, str) and merged
[k
] == ''):
2947 def encode_compat_str(string
, encoding
=preferredencoding(), errors
='strict'):
2948 return string
if isinstance(string
, compat_str
) else compat_str(string
, encoding
, errors
)
2960 TV_PARENTAL_GUIDELINES
= {
2970 def parse_age_limit(s
):
2971 # isinstance(False, int) is True. So type() must be used instead
2972 if type(s
) is int: # noqa: E721
2973 return s
if 0 <= s
<= 21 else None
2974 elif not isinstance(s
, str):
2976 m
= re
.match(r
'^(?P<age>\d{1,2})\+?$', s
)
2978 return int(m
.group('age'))
2981 return US_RATINGS
[s
]
2982 m
= re
.match(r
'^TV[_-]?(%s)$' % '|'.join(k
[3:] for k
in TV_PARENTAL_GUIDELINES
), s
)
2984 return TV_PARENTAL_GUIDELINES
['TV-' + m
.group(1)]
2988 def strip_jsonp(code
):
2991 (?:window\.)?(?P<func_name>[a-zA-Z0-9_.$]*)
2992 (?:\s*&&\s*(?P=func_name))?
2993 \s*\(\s*(?P<callback_data>.*)\);?
2994 \s*?(?://[^\n]*)*$''',
2995 r
'\g<callback_data>', code
)
2998 def js_to_json(code
, vars={}):
2999 # vars is a dict of var, val pairs to substitute
3000 COMMENT_RE
= r
'/\*(?:(?!\*/).)*?\*/|//[^\n]*\n'
3001 SKIP_RE
= fr
'\s*(?:{COMMENT_RE})?\s*'
3003 (fr
'(?s)^(0[xX][0-9a-fA-F]+){SKIP_RE}:?$', 16),
3004 (fr
'(?s)^(0+[0-7]+){SKIP_RE}:?$', 8),
3009 if v
in ('true', 'false', 'null'):
3011 elif v
in ('undefined', 'void 0'):
3013 elif v
.startswith('/*') or v
.startswith('//') or v
.startswith('!') or v
== ',':
3016 if v
[0] in ("'", '"'):
3017 v
= re
.sub(r
'(?s)\\.|"', lambda m
: {
3022 }.get(m
.group(0), m
.group(0)), v
[1:-1])
3024 for regex
, base
in INTEGER_TABLE
:
3025 im
= re
.match(regex
, v
)
3027 i
= int(im
.group(1), base
)
3028 return '"%d":' % i
if v
.endswith(':') else '%d' % i
3035 code
= re
.sub(r
'new Date\((".+")\)', r
'\g<1>', code
)
3037 return re
.sub(r
'''(?sx)
3038 "(?:[^"\\]*(?:\\\\|\\['"nurtbfx/\n]))*[^"\\]*"|
3039 '(?:[^'\\]*(?:\\\\|\\['"nurtbfx/\n]))*[^'\\]*'|
3040 {comment}|,(?={skip}[\]}}])|
3041 void\s0|(?:(?<![0-9])[eE]|[a-df-zA-DF-Z_$])[.a-zA-Z_$0-9]*|
3042 \b(?:0[xX][0-9a-fA-F]+|0+[0-7]+)(?:{skip}:)?|
3045 '''.format(comment
=COMMENT_RE
, skip
=SKIP_RE
), fix_kv
, code
)
3048 def qualities(quality_ids
):
3049 """ Get a numeric quality value out of a list of possible values """
3052 return quality_ids
.index(qid
)
3058 POSTPROCESS_WHEN
= ('pre_process', 'after_filter', 'before_dl', 'after_move', 'post_process', 'after_video', 'playlist')
3062 'default': '%(title)s [%(id)s].%(ext)s',
3063 'chapter': '%(title)s - %(section_number)03d %(section_title)s [%(id)s].%(ext)s',
3069 'description': 'description',
3070 'annotation': 'annotations.xml',
3071 'infojson': 'info.json',
3074 'pl_thumbnail': None,
3075 'pl_description': 'description',
3076 'pl_infojson': 'info.json',
3079 # As of [1] format syntax is:
3080 # %[mapping_key][conversion_flags][minimum_width][.precision][length_modifier]type
3081 # 1. https://docs.python.org/2/library/stdtypes.html#string-formatting
3082 STR_FORMAT_RE_TMPL
= r
'''(?x)
3083 (?<!%)(?P<prefix>(?:%%)*)
3085 (?P<has_key>\((?P<key>{0})\))?
3087 (?P<conversion>[#0\-+ ]+)?
3089 (?P<precision>\.\d+)?
3090 (?P<len_mod>[hlL])? # unused in python
3091 {1} # conversion type
3096 STR_FORMAT_TYPES
= 'diouxXeEfFgGcrs'
3099 def limit_length(s
, length
):
3100 """ Add ellipses to overly long strings """
3105 return s
[:length
- len(ELLIPSES
)] + ELLIPSES
3109 def version_tuple(v
):
3110 return tuple(int(e
) for e
in re
.split(r
'[-.]', v
))
3113 def is_outdated_version(version
, limit
, assume_new
=True):
3115 return not assume_new
3117 return version_tuple(version
) < version_tuple(limit
)
3119 return not assume_new
3122 def ytdl_is_updateable():
3123 """ Returns if yt-dlp can be updated with -U """
3125 from .update
import is_non_updateable
3127 return not is_non_updateable()
3130 def args_to_str(args
):
3131 # Get a short string representation for a subprocess command
3132 return ' '.join(compat_shlex_quote(a
) for a
in args
)
3135 def error_to_compat_str(err
):
3139 def error_to_str(err
):
3140 return f
'{type(err).__name__}: {err}'
3143 def mimetype2ext(mt
):
3147 mt
, _
, params
= mt
.partition(';')
3152 # Per RFC 3003, audio/mpeg can be .mp1, .mp2 or .mp3. Here use .mp3 as
3153 # it's the most popular one
3154 'audio/mpeg': 'mp3',
3155 'audio/x-wav': 'wav',
3157 'audio/wave': 'wav',
3160 ext
= FULL_MAP
.get(mt
)
3166 'smptett+xml': 'tt',
3170 'x-mp4-fragmented': 'mp4',
3171 'x-ms-sami': 'sami',
3174 'x-mpegurl': 'm3u8',
3175 'vnd.apple.mpegurl': 'm3u8',
3179 'vnd.ms-sstr+xml': 'ism',
3183 'filmstrip+json': 'fs',
3187 _
, _
, subtype
= mt
.rpartition('/')
3188 ext
= SUBTYPE_MAP
.get(subtype
.lower())
3199 _
, _
, suffix
= subtype
.partition('+')
3200 ext
= SUFFIX_MAP
.get(suffix
)
3204 return subtype
.replace('+', '.')
3207 def ext2mimetype(ext_or_url
):
3210 if '.' not in ext_or_url
:
3211 ext_or_url
= f
'file.{ext_or_url}'
3212 return mimetypes
.guess_type(ext_or_url
)[0]
3215 def parse_codecs(codecs_str
):
3216 # http://tools.ietf.org/html/rfc6381
3219 split_codecs
= list(filter(None, map(
3220 str.strip
, codecs_str
.strip().strip(',').split(','))))
3221 vcodec
, acodec
, scodec
, hdr
= None, None, None, None
3222 for full_codec
in split_codecs
:
3223 parts
= full_codec
.split('.')
3224 codec
= parts
[0].replace('0', '')
3225 if codec
in ('avc1', 'avc2', 'avc3', 'avc4', 'vp9', 'vp8', 'hev1', 'hev2',
3226 'h263', 'h264', 'mp4v', 'hvc1', 'av1', 'theora', 'dvh1', 'dvhe'):
3228 vcodec
= '.'.join(parts
[:4]) if codec
in ('vp9', 'av1', 'hvc1') else full_codec
3229 if codec
in ('dvh1', 'dvhe'):
3231 elif codec
== 'av1' and len(parts
) > 3 and parts
[3] == '10':
3233 elif full_codec
.replace('0', '').startswith('vp9.2'):
3235 elif codec
in ('flac', 'mp4a', 'opus', 'vorbis', 'mp3', 'aac', 'ac-3', 'ec-3', 'eac3', 'dtsc', 'dtse', 'dtsh', 'dtsl'):
3238 elif codec
in ('stpp', 'wvtt',):
3242 write_string(f
'WARNING: Unknown codec {full_codec}\n')
3243 if vcodec
or acodec
or scodec
:
3245 'vcodec': vcodec
or 'none',
3246 'acodec': acodec
or 'none',
3247 'dynamic_range': hdr
,
3248 **({'scodec': scodec}
if scodec
is not None else {}),
3250 elif len(split_codecs
) == 2:
3252 'vcodec': split_codecs
[0],
3253 'acodec': split_codecs
[1],
3258 def urlhandle_detect_ext(url_handle
):
3259 getheader
= url_handle
.headers
.get
3261 cd
= getheader('Content-Disposition')
3263 m
= re
.match(r
'attachment;\s*filename="(?P<filename>[^"]+)"', cd
)
3265 e
= determine_ext(m
.group('filename'), default_ext
=None)
3269 return mimetype2ext(getheader('Content-Type'))
3272 def encode_data_uri(data
, mime_type
):
3273 return 'data:%s;base64,%s' % (mime_type
, base64
.b64encode(data
).decode('ascii'))
3276 def age_restricted(content_limit
, age_limit
):
3277 """ Returns True iff the content should be blocked """
3279 if age_limit
is None: # No limit set
3281 if content_limit
is None:
3282 return False # Content available for everyone
3283 return age_limit
< content_limit
3286 def is_html(first_bytes
):
3287 """ Detect whether a file contains HTML by examining its first bytes. """
3290 (b
'\xef\xbb\xbf', 'utf-8'),
3291 (b
'\x00\x00\xfe\xff', 'utf-32-be'),
3292 (b
'\xff\xfe\x00\x00', 'utf-32-le'),
3293 (b
'\xff\xfe', 'utf-16-le'),
3294 (b
'\xfe\xff', 'utf-16-be'),
3298 for bom
, enc
in BOMS
:
3299 while first_bytes
.startswith(bom
):
3300 encoding
, first_bytes
= enc
, first_bytes
[len(bom
):]
3302 return re
.match(r
'^\s*<', first_bytes
.decode(encoding
, 'replace'))
3305 def determine_protocol(info_dict
):
3306 protocol
= info_dict
.get('protocol')
3307 if protocol
is not None:
3310 url
= sanitize_url(info_dict
['url'])
3311 if url
.startswith('rtmp'):
3313 elif url
.startswith('mms'):
3315 elif url
.startswith('rtsp'):
3318 ext
= determine_ext(url
)
3324 return compat_urllib_parse_urlparse(url
).scheme
3327 def render_table(header_row
, data
, delim
=False, extra_gap
=0, hide_empty
=False):
3328 """ Render a list of rows, each as a list of values.
3329 Text after a \t will be right aligned """
3331 return len(remove_terminal_sequences(string
).replace('\t', ''))
3333 def get_max_lens(table
):
3334 return [max(width(str(v
)) for v
in col
) for col
in zip(*table
)]
3336 def filter_using_list(row
, filterArray
):
3337 return [col
for take
, col
in itertools
.zip_longest(filterArray
, row
, fillvalue
=True) if take
]
3339 max_lens
= get_max_lens(data
) if hide_empty
else []
3340 header_row
= filter_using_list(header_row
, max_lens
)
3341 data
= [filter_using_list(row
, max_lens
) for row
in data
]
3343 table
= [header_row
] + data
3344 max_lens
= get_max_lens(table
)
3347 table
= [header_row
, [delim
* (ml
+ extra_gap
) for ml
in max_lens
]] + data
3348 table
[1][-1] = table
[1][-1][:-extra_gap
* len(delim
)] # Remove extra_gap from end of delimiter
3350 for pos
, text
in enumerate(map(str, row
)):
3352 row
[pos
] = text
.replace('\t', ' ' * (max_lens
[pos
] - width(text
))) + ' ' * extra_gap
3354 row
[pos
] = text
+ ' ' * (max_lens
[pos
] - width(text
) + extra_gap
)
3355 ret
= '\n'.join(''.join(row
).rstrip() for row
in table
)
3359 def _match_one(filter_part
, dct
, incomplete
):
3360 # TODO: Generalize code with YoutubeDL._build_format_filter
3361 STRING_OPERATORS
= {
3362 '*=': operator
.contains
,
3363 '^=': lambda attr
, value
: attr
.startswith(value
),
3364 '$=': lambda attr
, value
: attr
.endswith(value
),
3365 '~=': lambda attr
, value
: re
.search(value
, attr
),
3367 COMPARISON_OPERATORS
= {
3369 '<=': operator
.le
, # "<=" must be defined above "<"
3376 if isinstance(incomplete
, bool):
3377 is_incomplete
= lambda _
: incomplete
3379 is_incomplete
= lambda k
: k
in incomplete
3381 operator_rex
= re
.compile(r
'''(?x)\s*
3383 \s*(?P<negation>!\s*)?(?P<op>%s)(?P<none_inclusive>\s*\?)?\s*
3385 (?P<quote>["\'])(?P
<quotedstrval
>.+?
)(?P
=quote
)|
3389 ''' % '|'.join(map(re.escape, COMPARISON_OPERATORS.keys())))
3390 m = operator_rex.search(filter_part)
3393 unnegated_op = COMPARISON_OPERATORS[m['op']]
3395 op = lambda attr, value: not unnegated_op(attr, value)
3398 comparison_value = m['quotedstrval'] or m['strval'] or m['intval']
3400 comparison_value = comparison_value.replace(r'\%s' % m['quote'], m['quote'])
3401 actual_value = dct.get(m['key'])
3402 numeric_comparison = None
3403 if isinstance(actual_value, (int, float)):
3404 # If the original field is a string and matching comparisonvalue is
3405 # a number we should respect the origin of the original field
3406 # and process comparison value as a string (see
3407 # https://github.com/ytdl-org/youtube-dl/issues/11082)
3409 numeric_comparison = int(comparison_value)
3411 numeric_comparison = parse_filesize(comparison_value)
3412 if numeric_comparison is None:
3413 numeric_comparison = parse_filesize(f'{comparison_value}B')
3414 if numeric_comparison is None:
3415 numeric_comparison = parse_duration(comparison_value)
3416 if numeric_comparison is not None and m['op'] in STRING_OPERATORS:
3417 raise ValueError('Operator %s only supports string values!' % m['op'])
3418 if actual_value is None:
3419 return is_incomplete(m['key']) or m['none_inclusive']
3420 return op(actual_value, comparison_value if numeric_comparison is None else numeric_comparison)
3423 '': lambda v: (v is True) if isinstance(v, bool) else (v is not None),
3424 '!': lambda v: (v is False) if isinstance(v, bool) else (v is None),
3426 operator_rex = re.compile(r'''(?x
)\s
*
3427 (?P
<op
>%s)\s
*(?P
<key
>[a
-z_
]+)
3429 ''' % '|'.join(map(re.escape, UNARY_OPERATORS.keys())))
3430 m = operator_rex.search(filter_part)
3432 op = UNARY_OPERATORS[m.group('op')]
3433 actual_value = dct.get(m.group('key'))
3434 if is_incomplete(m.group('key')) and actual_value is None:
3436 return op(actual_value)
3438 raise ValueError('Invalid filter part %r' % filter_part)
3441 def match_str(filter_str, dct, incomplete=False):
3442 """ Filter a dictionary with a simple string syntax.
3443 @returns Whether the filter passes
3444 @param incomplete Set of keys that is expected to be missing from dct.
3445 Can be True/False to indicate all/none of the keys may be missing.
3446 All conditions on incomplete keys pass if the key is missing
3449 _match_one(filter_part.replace(r'\&', '&'), dct, incomplete)
3450 for filter_part in re.split(r'(?<!\\)&', filter_str))
3453 def match_filter_func(filters):
3456 filters = set(variadic(filters))
3458 interactive = '-' in filters
3462 def _match_func(info_dict, incomplete=False):
3463 if not filters or any(match_str(f, info_dict, incomplete) for f in filters):
3464 return NO_DEFAULT if interactive and not incomplete else None
3466 video_title = info_dict.get('title') or info_dict.get('id') or 'video'
3467 filter_str = ') | ('.join(map(str.strip, filters))
3468 return f'{video_title} does not pass filter ({filter_str}), skipping ..'
3472 def parse_dfxp_time_expr(time_expr):
3476 mobj = re.match(rf'^(?P<time_offset>{NUMBER_RE})s?$', time_expr)
3478 return float(mobj.group('time_offset'))
3480 mobj = re.match(r'^(\d+):(\d\d):(\d\d(?:(?:\.|:)\d+)?)$', time_expr)
3482 return 3600 * int(mobj.group(1)) + 60 * int(mobj.group(2)) + float(mobj.group(3).replace(':', '.'))
3485 def srt_subtitles_timecode(seconds):
3486 return '%02d:%02d:%02d,%03d' % timetuple_from_msec(seconds * 1000)
3489 def ass_subtitles_timecode(seconds):
3490 time = timetuple_from_msec(seconds * 1000)
3491 return '%01d:%02d:%02d.%02d' % (*time[:-1], time.milliseconds / 10)
3494 def dfxp2srt(dfxp_data):
3496 @param dfxp_data A
bytes-like
object containing DFXP data
3497 @returns A
unicode object containing converted SRT data
3499 LEGACY_NAMESPACES = (
3500 (b'http://www.w3.org/ns/ttml', [
3501 b'http://www.w3.org/2004/11/ttaf1',
3502 b'http://www.w3.org/2006/04/ttaf1',
3503 b'http://www.w3.org/2006/10/ttaf1',
3505 (b'http://www.w3.org/ns/ttml#styling', [
3506 b'http://www.w3.org/ns/ttml#style',
3510 SUPPORTED_STYLING = [
3519 _x = functools.partial(xpath_with_ns, ns_map={
3520 'xml': 'http://www.w3.org/XML/1998/namespace',
3521 'ttml': 'http://www.w3.org/ns/ttml',
3522 'tts': 'http://www.w3.org/ns/ttml#styling',
3528 class TTMLPElementParser:
3530 _unclosed_elements = []
3531 _applied_styles = []
3533 def start(self, tag, attrib):
3534 if tag in (_x('ttml:br'), 'br'):
3537 unclosed_elements = []
3539 element_style_id = attrib.get('style')
3541 style.update(default_style)
3542 if element_style_id:
3543 style.update(styles.get(element_style_id, {}))
3544 for prop in SUPPORTED_STYLING:
3545 prop_val = attrib.get(_x('tts:' + prop))
3547 style[prop] = prop_val
3550 for k, v in sorted(style.items()):
3551 if self._applied_styles and self._applied_styles[-1].get(k) == v:
3554 font += ' color="%s"' % v
3555 elif k == 'fontSize':
3556 font += ' size="%s"' % v
3557 elif k == 'fontFamily':
3558 font += ' face="%s"' % v
3559 elif k == 'fontWeight' and v == 'bold':
3561 unclosed_elements.append('b')
3562 elif k == 'fontStyle' and v == 'italic':
3564 unclosed_elements.append('i')
3565 elif k == 'textDecoration' and v == 'underline':
3567 unclosed_elements.append('u')
3569 self._out += '<font' + font + '>'
3570 unclosed_elements.append('font')
3572 if self._applied_styles:
3573 applied_style.update(self._applied_styles[-1])
3574 applied_style.update(style)
3575 self._applied_styles.append(applied_style)
3576 self._unclosed_elements.append(unclosed_elements)
3579 if tag not in (_x('ttml:br'), 'br'):
3580 unclosed_elements = self._unclosed_elements.pop()
3581 for element in reversed(unclosed_elements):
3582 self._out += '</%s>' % element
3583 if unclosed_elements and self._applied_styles:
3584 self._applied_styles.pop()
3586 def data(self, data):
3590 return self._out.strip()
3592 def parse_node(node):
3593 target = TTMLPElementParser()
3594 parser = xml.etree.ElementTree.XMLParser(target=target)
3595 parser.feed(xml.etree.ElementTree.tostring(node))
3596 return parser.close()
3598 for k, v in LEGACY_NAMESPACES:
3600 dfxp_data = dfxp_data.replace(ns, k)
3602 dfxp = compat_etree_fromstring(dfxp_data)
3604 paras = dfxp.findall(_x('.//ttml:p')) or dfxp.findall('.//p')
3607 raise ValueError('Invalid dfxp/TTML subtitle')
3611 for style in dfxp.findall(_x('.//ttml:style')):
3612 style_id = style.get('id') or style.get(_x('xml:id'))
3615 parent_style_id = style.get('style')
3617 if parent_style_id not in styles:
3620 styles[style_id] = styles[parent_style_id].copy()
3621 for prop in SUPPORTED_STYLING:
3622 prop_val = style.get(_x('tts:' + prop))
3624 styles.setdefault(style_id, {})[prop] = prop_val
3630 for p in ('body', 'div'):
3631 ele = xpath_element(dfxp, [_x('.//ttml:' + p), './/' + p])
3634 style = styles.get(ele.get('style'))
3637 default_style.update(style)
3639 for para, index in zip(paras, itertools.count(1)):
3640 begin_time = parse_dfxp_time_expr(para.attrib.get('begin'))
3641 end_time = parse_dfxp_time_expr(para.attrib.get('end'))
3642 dur = parse_dfxp_time_expr(para.attrib.get('dur'))
3643 if begin_time is None:
3648 end_time = begin_time + dur
3649 out.append('%d\n%s --> %s\n%s\n\n' % (
3651 srt_subtitles_timecode(begin_time),
3652 srt_subtitles_timecode(end_time),
3658 def cli_option(params, command_option, param, separator=None):
3659 param = params.get(param)
3660 return ([] if param is None
3661 else [command_option, str(param)] if separator is None
3662 else [f'{command_option}{separator}{param}'])
3665 def cli_bool_option(params, command_option, param, true_value='true', false_value='false', separator=None):
3666 param = params.get(param)
3667 assert param in (True, False, None)
3668 return cli_option({True: true_value, False: false_value}, command_option, param, separator)
3671 def cli_valueless_option(params, command_option, param, expected_value=True):
3672 return [command_option] if params.get(param) == expected_value else []
3675 def cli_configuration_args(argdict, keys, default=[], use_compat=True):
3676 if isinstance(argdict, (list, tuple)): # for backward compatibility
3683 assert isinstance(argdict, dict)
3685 assert isinstance(keys, (list, tuple))
3686 for key_list in keys:
3687 arg_list = list(filter(
3688 lambda x: x is not None,
3689 [argdict.get(key.lower()) for key in variadic(key_list)]))
3691 return [arg for args in arg_list for arg in args]
3695 def _configuration_args(main_key, argdict, exe, keys=None, default=[], use_compat=True):
3696 main_key, exe = main_key.lower(), exe.lower()
3697 root_key = exe if main_key == exe else f'{main_key}+{exe}'
3698 keys = [f'{root_key}{k}' for k in (keys or [''])]
3699 if root_key in keys:
3701 keys.append((main_key, exe))
3702 keys.append('default')
3705 return cli_configuration_args(argdict, keys, default, use_compat)
3709 # See http://www.loc.gov/standards/iso639-2/ISO-639-2_utf-8.txt
3768 'iw': 'heb', # Replaced by he in 1989 revision
3778 'in': 'ind', # Replaced by id in 1989 revision
3893 'ji': 'yid', # Replaced by yi in 1989 revision
3901 def short2long(cls, code):
3902 """Convert language code from ISO 639-1 to ISO 639-2/T"""
3903 return cls._lang_map.get(code[:2])
3906 def long2short(cls, code):
3907 """Convert language code from ISO 639-2/T to ISO 639-1"""
3908 for short_name, long_name in cls._lang_map.items():
3909 if long_name == code:
3914 # From http://data.okfn.org/data/core/country-list
3916 'AF': 'Afghanistan',
3917 'AX': 'Åland Islands',
3920 'AS': 'American Samoa',
3925 'AG': 'Antigua and Barbuda',
3942 'BO': 'Bolivia, Plurinational State of',
3943 'BQ': 'Bonaire, Sint Eustatius and Saba',
3944 'BA': 'Bosnia and Herzegovina',
3946 'BV': 'Bouvet Island',
3948 'IO': 'British Indian Ocean Territory',
3949 'BN': 'Brunei Darussalam',
3951 'BF': 'Burkina Faso',
3957 'KY': 'Cayman Islands',
3958 'CF': 'Central African Republic',
3962 'CX': 'Christmas Island',
3963 'CC': 'Cocos (Keeling) Islands',
3967 'CD': 'Congo, the Democratic Republic of the',
3968 'CK': 'Cook Islands',
3970 'CI': 'Côte d\'Ivoire',
3975 'CZ': 'Czech Republic',
3979 'DO': 'Dominican Republic',
3982 'SV': 'El Salvador',
3983 'GQ': 'Equatorial Guinea',
3987 'FK': 'Falkland Islands (Malvinas)',
3988 'FO': 'Faroe Islands',
3992 'GF': 'French Guiana',
3993 'PF': 'French Polynesia',
3994 'TF': 'French Southern Territories',
4009 'GW': 'Guinea-Bissau',
4012 'HM': 'Heard Island and McDonald Islands',
4013 'VA': 'Holy See (Vatican City State)',
4020 'IR': 'Iran, Islamic Republic of',
4023 'IM': 'Isle of Man',
4033 'KP': 'Korea, Democratic People\'s Republic of',
4034 'KR': 'Korea, Republic of',
4037 'LA': 'Lao People\'s Democratic Republic',
4043 'LI': 'Liechtenstein',
4047 'MK': 'Macedonia, the Former Yugoslav Republic of',
4054 'MH': 'Marshall Islands',
4060 'FM': 'Micronesia, Federated States of',
4061 'MD': 'Moldova, Republic of',
4072 'NL': 'Netherlands',
4073 'NC': 'New Caledonia',
4074 'NZ': 'New Zealand',
4079 'NF': 'Norfolk Island',
4080 'MP': 'Northern Mariana Islands',
4085 'PS': 'Palestine, State of',
4087 'PG': 'Papua New Guinea',
4090 'PH': 'Philippines',
4094 'PR': 'Puerto Rico',
4098 'RU': 'Russian Federation',
4100 'BL': 'Saint Barthélemy',
4101 'SH': 'Saint Helena, Ascension and Tristan da Cunha',
4102 'KN': 'Saint Kitts and Nevis',
4103 'LC': 'Saint Lucia',
4104 'MF': 'Saint Martin (French part)',
4105 'PM': 'Saint Pierre and Miquelon',
4106 'VC': 'Saint Vincent and the Grenadines',
4109 'ST': 'Sao Tome and Principe',
4110 'SA': 'Saudi Arabia',
4114 'SL': 'Sierra Leone',
4116 'SX': 'Sint Maarten (Dutch part)',
4119 'SB': 'Solomon Islands',
4121 'ZA': 'South Africa',
4122 'GS': 'South Georgia and the South Sandwich Islands',
4123 'SS': 'South Sudan',
4128 'SJ': 'Svalbard and Jan Mayen',
4131 'CH': 'Switzerland',
4132 'SY': 'Syrian Arab Republic',
4133 'TW': 'Taiwan, Province of China',
4135 'TZ': 'Tanzania, United Republic of',
4137 'TL': 'Timor-Leste',
4141 'TT': 'Trinidad and Tobago',
4144 'TM': 'Turkmenistan',
4145 'TC': 'Turks and Caicos Islands',
4149 'AE': 'United Arab Emirates',
4150 'GB': 'United Kingdom',
4151 'US': 'United States',
4152 'UM': 'United States Minor Outlying Islands',
4156 'VE': 'Venezuela, Bolivarian Republic of',
4158 'VG': 'Virgin Islands, British',
4159 'VI': 'Virgin Islands, U.S.',
4160 'WF': 'Wallis and Futuna',
4161 'EH': 'Western Sahara',
4165 # Not ISO 3166 codes, but used for IP blocks
4166 'AP': 'Asia/Pacific Region',
4171 def short2full(cls, code):
4172 """Convert an ISO 3166-2 country code to the corresponding full name"""
4173 return cls._country_map.get(code.upper())
4177 # Major IPv4 address blocks per country
4179 'AD': '46.172.224.0/19',
4180 'AE': '94.200.0.0/13',
4181 'AF': '149.54.0.0/17',
4182 'AG': '209.59.64.0/18',
4183 'AI': '204.14.248.0/21',
4184 'AL': '46.99.0.0/16',
4185 'AM': '46.70.0.0/15',
4186 'AO': '105.168.0.0/13',
4187 'AP': '182.50.184.0/21',
4188 'AQ': '23.154.160.0/24',
4189 'AR': '181.0.0.0/12',
4190 'AS': '202.70.112.0/20',
4191 'AT': '77.116.0.0/14',
4192 'AU': '1.128.0.0/11',
4193 'AW': '181.41.0.0/18',
4194 'AX': '185.217.4.0/22',
4195 'AZ': '5.197.0.0/16',
4196 'BA': '31.176.128.0/17',
4197 'BB': '65.48.128.0/17',
4198 'BD': '114.130.0.0/16',
4200 'BF': '102.178.0.0/15',
4201 'BG': '95.42.0.0/15',
4202 'BH': '37.131.0.0/17',
4203 'BI': '154.117.192.0/18',
4204 'BJ': '137.255.0.0/16',
4205 'BL': '185.212.72.0/23',
4206 'BM': '196.12.64.0/18',
4207 'BN': '156.31.0.0/16',
4208 'BO': '161.56.0.0/16',
4209 'BQ': '161.0.80.0/20',
4210 'BR': '191.128.0.0/12',
4211 'BS': '24.51.64.0/18',
4212 'BT': '119.2.96.0/19',
4213 'BW': '168.167.0.0/16',
4214 'BY': '178.120.0.0/13',
4215 'BZ': '179.42.192.0/18',
4216 'CA': '99.224.0.0/11',
4217 'CD': '41.243.0.0/16',
4218 'CF': '197.242.176.0/21',
4219 'CG': '160.113.0.0/16',
4220 'CH': '85.0.0.0/13',
4221 'CI': '102.136.0.0/14',
4222 'CK': '202.65.32.0/19',
4223 'CL': '152.172.0.0/14',
4224 'CM': '102.244.0.0/14',
4225 'CN': '36.128.0.0/10',
4226 'CO': '181.240.0.0/12',
4227 'CR': '201.192.0.0/12',
4228 'CU': '152.206.0.0/15',
4229 'CV': '165.90.96.0/19',
4230 'CW': '190.88.128.0/17',
4231 'CY': '31.153.0.0/16',
4232 'CZ': '88.100.0.0/14',
4234 'DJ': '197.241.0.0/17',
4235 'DK': '87.48.0.0/12',
4236 'DM': '192.243.48.0/20',
4237 'DO': '152.166.0.0/15',
4238 'DZ': '41.96.0.0/12',
4239 'EC': '186.68.0.0/15',
4240 'EE': '90.190.0.0/15',
4241 'EG': '156.160.0.0/11',
4242 'ER': '196.200.96.0/20',
4243 'ES': '88.0.0.0/11',
4244 'ET': '196.188.0.0/14',
4245 'EU': '2.16.0.0/13',
4246 'FI': '91.152.0.0/13',
4247 'FJ': '144.120.0.0/16',
4248 'FK': '80.73.208.0/21',
4249 'FM': '119.252.112.0/20',
4250 'FO': '88.85.32.0/19',
4252 'GA': '41.158.0.0/15',
4254 'GD': '74.122.88.0/21',
4255 'GE': '31.146.0.0/16',
4256 'GF': '161.22.64.0/18',
4257 'GG': '62.68.160.0/19',
4258 'GH': '154.160.0.0/12',
4259 'GI': '95.164.0.0/16',
4260 'GL': '88.83.0.0/19',
4261 'GM': '160.182.0.0/15',
4262 'GN': '197.149.192.0/18',
4263 'GP': '104.250.0.0/19',
4264 'GQ': '105.235.224.0/20',
4265 'GR': '94.64.0.0/13',
4266 'GT': '168.234.0.0/16',
4267 'GU': '168.123.0.0/16',
4268 'GW': '197.214.80.0/20',
4269 'GY': '181.41.64.0/18',
4270 'HK': '113.252.0.0/14',
4271 'HN': '181.210.0.0/16',
4272 'HR': '93.136.0.0/13',
4273 'HT': '148.102.128.0/17',
4274 'HU': '84.0.0.0/14',
4275 'ID': '39.192.0.0/10',
4276 'IE': '87.32.0.0/12',
4277 'IL': '79.176.0.0/13',
4278 'IM': '5.62.80.0/20',
4279 'IN': '117.192.0.0/10',
4280 'IO': '203.83.48.0/21',
4281 'IQ': '37.236.0.0/14',
4282 'IR': '2.176.0.0/12',
4283 'IS': '82.221.0.0/16',
4284 'IT': '79.0.0.0/10',
4285 'JE': '87.244.64.0/18',
4286 'JM': '72.27.0.0/17',
4287 'JO': '176.29.0.0/16',
4288 'JP': '133.0.0.0/8',
4289 'KE': '105.48.0.0/12',
4290 'KG': '158.181.128.0/17',
4291 'KH': '36.37.128.0/17',
4292 'KI': '103.25.140.0/22',
4293 'KM': '197.255.224.0/20',
4294 'KN': '198.167.192.0/19',
4295 'KP': '175.45.176.0/22',
4296 'KR': '175.192.0.0/10',
4297 'KW': '37.36.0.0/14',
4298 'KY': '64.96.0.0/15',
4299 'KZ': '2.72.0.0/13',
4300 'LA': '115.84.64.0/18',
4301 'LB': '178.135.0.0/16',
4302 'LC': '24.92.144.0/20',
4303 'LI': '82.117.0.0/19',
4304 'LK': '112.134.0.0/15',
4305 'LR': '102.183.0.0/16',
4306 'LS': '129.232.0.0/17',
4307 'LT': '78.56.0.0/13',
4308 'LU': '188.42.0.0/16',
4309 'LV': '46.109.0.0/16',
4310 'LY': '41.252.0.0/14',
4311 'MA': '105.128.0.0/11',
4312 'MC': '88.209.64.0/18',
4313 'MD': '37.246.0.0/16',
4314 'ME': '178.175.0.0/17',
4315 'MF': '74.112.232.0/21',
4316 'MG': '154.126.0.0/17',
4317 'MH': '117.103.88.0/21',
4318 'MK': '77.28.0.0/15',
4319 'ML': '154.118.128.0/18',
4320 'MM': '37.111.0.0/17',
4321 'MN': '49.0.128.0/17',
4322 'MO': '60.246.0.0/16',
4323 'MP': '202.88.64.0/20',
4324 'MQ': '109.203.224.0/19',
4325 'MR': '41.188.64.0/18',
4326 'MS': '208.90.112.0/22',
4327 'MT': '46.11.0.0/16',
4328 'MU': '105.16.0.0/12',
4329 'MV': '27.114.128.0/18',
4330 'MW': '102.70.0.0/15',
4331 'MX': '187.192.0.0/11',
4332 'MY': '175.136.0.0/13',
4333 'MZ': '197.218.0.0/15',
4334 'NA': '41.182.0.0/16',
4335 'NC': '101.101.0.0/18',
4336 'NE': '197.214.0.0/18',
4337 'NF': '203.17.240.0/22',
4338 'NG': '105.112.0.0/12',
4339 'NI': '186.76.0.0/15',
4340 'NL': '145.96.0.0/11',
4341 'NO': '84.208.0.0/13',
4342 'NP': '36.252.0.0/15',
4343 'NR': '203.98.224.0/19',
4344 'NU': '49.156.48.0/22',
4345 'NZ': '49.224.0.0/14',
4346 'OM': '5.36.0.0/15',
4347 'PA': '186.72.0.0/15',
4348 'PE': '186.160.0.0/14',
4349 'PF': '123.50.64.0/18',
4350 'PG': '124.240.192.0/19',
4351 'PH': '49.144.0.0/13',
4352 'PK': '39.32.0.0/11',
4353 'PL': '83.0.0.0/11',
4354 'PM': '70.36.0.0/20',
4355 'PR': '66.50.0.0/16',
4356 'PS': '188.161.0.0/16',
4357 'PT': '85.240.0.0/13',
4358 'PW': '202.124.224.0/20',
4359 'PY': '181.120.0.0/14',
4360 'QA': '37.210.0.0/15',
4361 'RE': '102.35.0.0/16',
4362 'RO': '79.112.0.0/13',
4363 'RS': '93.86.0.0/15',
4364 'RU': '5.136.0.0/13',
4365 'RW': '41.186.0.0/16',
4366 'SA': '188.48.0.0/13',
4367 'SB': '202.1.160.0/19',
4368 'SC': '154.192.0.0/11',
4369 'SD': '102.120.0.0/13',
4370 'SE': '78.64.0.0/12',
4371 'SG': '8.128.0.0/10',
4372 'SI': '188.196.0.0/14',
4373 'SK': '78.98.0.0/15',
4374 'SL': '102.143.0.0/17',
4375 'SM': '89.186.32.0/19',
4376 'SN': '41.82.0.0/15',
4377 'SO': '154.115.192.0/18',
4378 'SR': '186.179.128.0/17',
4379 'SS': '105.235.208.0/21',
4380 'ST': '197.159.160.0/19',
4381 'SV': '168.243.0.0/16',
4382 'SX': '190.102.0.0/20',
4384 'SZ': '41.84.224.0/19',
4385 'TC': '65.255.48.0/20',
4386 'TD': '154.68.128.0/19',
4387 'TG': '196.168.0.0/14',
4388 'TH': '171.96.0.0/13',
4389 'TJ': '85.9.128.0/18',
4390 'TK': '27.96.24.0/21',
4391 'TL': '180.189.160.0/20',
4392 'TM': '95.85.96.0/19',
4393 'TN': '197.0.0.0/11',
4394 'TO': '175.176.144.0/21',
4395 'TR': '78.160.0.0/11',
4396 'TT': '186.44.0.0/15',
4397 'TV': '202.2.96.0/19',
4398 'TW': '120.96.0.0/11',
4399 'TZ': '156.156.0.0/14',
4400 'UA': '37.52.0.0/14',
4401 'UG': '102.80.0.0/13',
4403 'UY': '167.56.0.0/13',
4404 'UZ': '84.54.64.0/18',
4405 'VA': '212.77.0.0/19',
4406 'VC': '207.191.240.0/21',
4407 'VE': '186.88.0.0/13',
4408 'VG': '66.81.192.0/20',
4409 'VI': '146.226.0.0/16',
4410 'VN': '14.160.0.0/11',
4411 'VU': '202.80.32.0/20',
4412 'WF': '117.20.32.0/21',
4413 'WS': '202.4.32.0/19',
4414 'YE': '134.35.0.0/16',
4415 'YT': '41.242.116.0/22',
4416 'ZA': '41.0.0.0/11',
4417 'ZM': '102.144.0.0/13',
4418 'ZW': '102.177.192.0/18',
4422 def random_ipv4(cls, code_or_block):
4423 if len(code_or_block) == 2:
4424 block = cls._country_ip_map.get(code_or_block.upper())
4428 block = code_or_block
4429 addr, preflen = block.split('/')
4430 addr_min = compat_struct_unpack('!L', socket.inet_aton(addr))[0]
4431 addr_max = addr_min | (0xffffffff >> int(preflen))
4432 return compat_str(socket.inet_ntoa(
4433 compat_struct_pack('!L', random.randint(addr_min, addr_max))))
4436 class PerRequestProxyHandler(compat_urllib_request.ProxyHandler):
4437 def __init__(self, proxies=None):
4438 # Set default handlers
4439 for type in ('http', 'https'):
4440 setattr(self, '%s_open' % type,
4441 lambda r, proxy='__noproxy__', type=type, meth=self.proxy_open:
4442 meth(r, proxy, type))
4443 compat_urllib_request.ProxyHandler.__init__(self, proxies)
4445 def proxy_open(self, req, proxy, type):
4446 req_proxy = req.headers.get('Ytdl-request-proxy')
4447 if req_proxy is not None:
4449 del req.headers['Ytdl-request-proxy']
4451 if proxy == '__noproxy__':
4452 return None # No Proxy
4453 if compat_urlparse.urlparse(proxy).scheme.lower() in ('socks', 'socks4', 'socks4a', 'socks5'):
4454 req.add_header('Ytdl-socks-proxy', proxy)
4455 # yt-dlp's http/https handlers do wrapping the socket with socks
4457 return compat_urllib_request.ProxyHandler.proxy_open(
4458 self, req, proxy, type)
4461 # Both long_to_bytes and bytes_to_long are adapted from PyCrypto, which is
4462 # released into Public Domain
4463 # https://github.com/dlitz/pycrypto/blob/master/lib/Crypto/Util/number.py#L387
4465 def long_to_bytes(n, blocksize=0):
4466 """long_to_bytes(n:long, blocksize:int) : string
4467 Convert a long integer to a byte string.
4469 If optional blocksize is given and greater than zero, pad the front of the
4470 byte string with binary zeros so that the length is a multiple of
4473 # after much testing, this algorithm was deemed to be the fastest
4477 s = compat_struct_pack('>I', n & 0xffffffff) + s
4479 # strip off leading zeros
4480 for i in range(len(s)):
4481 if s[i] != b'\000'[0]:
4484 # only happens when n == 0
4488 # add back some pad bytes. this could be done more efficiently w.r.t. the
4489 # de-padding being done above, but sigh...
4490 if blocksize > 0 and len(s) % blocksize:
4491 s = (blocksize - len(s) % blocksize) * b'\000' + s
4495 def bytes_to_long(s):
4496 """bytes_to_long(string) : long
4497 Convert a byte string to a long integer.
4499 This is (essentially) the inverse of long_to_bytes().
4504 extra = (4 - length % 4)
4505 s = b'\000' * extra + s
4506 length = length + extra
4507 for i in range(0, length, 4):
4508 acc = (acc << 32) + compat_struct_unpack('>I', s[i:i + 4])[0]
4512 def ohdave_rsa_encrypt(data, exponent, modulus):
4514 Implement OHDave
's RSA algorithm. See http://www.ohdave.com/rsa/
4517 data: data to encrypt, bytes-like object
4518 exponent, modulus: parameter e and N of RSA algorithm, both integer
4519 Output: hex string of encrypted data
4521 Limitation: supports one block encryption only
4524 payload = int(binascii.hexlify(data[::-1]), 16)
4525 encrypted = pow(payload, exponent, modulus)
4526 return '%x' % encrypted
4529 def pkcs1pad(data, length):
4531 Padding input data with PKCS#1 scheme
4533 @param {int[]} data input data
4534 @param {int} length target length
4535 @returns {int[]} padded data
4537 if len(data) > length - 11:
4538 raise ValueError('Input data too
long for PKCS
#1 padding')
4540 pseudo_random
= [random
.randint(0, 254) for _
in range(length
- len(data
) - 3)]
4541 return [0, 2] + pseudo_random
+ [0] + data
4544 def encode_base_n(num
, n
, table
=None):
4545 FULL_TABLE
= '0123456789abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ'
4547 table
= FULL_TABLE
[:n
]
4550 raise ValueError('base %d exceeds table length %d' % (n
, len(table
)))
4557 ret
= table
[num
% n
] + ret
4562 def decode_packed_codes(code
):
4563 mobj
= re
.search(PACKED_CODES_RE
, code
)
4564 obfuscated_code
, base
, count
, symbols
= mobj
.groups()
4567 symbols
= symbols
.split('|')
4572 base_n_count
= encode_base_n(count
, base
)
4573 symbol_table
[base_n_count
] = symbols
[count
] or base_n_count
4576 r
'\b(\w+)\b', lambda mobj
: symbol_table
[mobj
.group(0)],
4580 def caesar(s
, alphabet
, shift
):
4585 alphabet
[(alphabet
.index(c
) + shift
) % l
] if c
in alphabet
else c
4590 return caesar(s
, r
'''!"#$%&'()*+,-./0123456789:;<=>?@ABCDEFGHIJKLMNOPQRSTUVWXYZ[\]^_`abcdefghijklmnopqrstuvwxyz{|}~''', 47)
4593 def parse_m3u8_attributes(attrib
):
4595 for (key
, val
) in re
.findall(r
'(?P<key>[A-Z0-9-]+)=(?P<val>"[^"]+"|[^",]+)(?:,|$)', attrib
):
4596 if val
.startswith('"'):
4602 def urshift(val
, n
):
4603 return val
>> n
if val
>= 0 else (val
+ 0x100000000) >> n
4606 # Based on png2str() written by @gdkchan and improved by @yokrysty
4607 # Originally posted at https://github.com/ytdl-org/youtube-dl/issues/9706
4608 def decode_png(png_data
):
4609 # Reference: https://www.w3.org/TR/PNG/
4610 header
= png_data
[8:]
4612 if png_data
[:8] != b
'\x89PNG\x0d\x0a\x1a\x0a' or header
[4:8] != b
'IHDR':
4613 raise OSError('Not a valid PNG file.')
4615 int_map
= {1: '>B', 2: '>H', 4: '>I'}
4616 unpack_integer
= lambda x
: compat_struct_unpack(int_map
[len(x
)], x
)[0]
4621 length
= unpack_integer(header
[:4])
4624 chunk_type
= header
[:4]
4627 chunk_data
= header
[:length
]
4628 header
= header
[length
:]
4630 header
= header
[4:] # Skip CRC
4638 ihdr
= chunks
[0]['data']
4640 width
= unpack_integer(ihdr
[:4])
4641 height
= unpack_integer(ihdr
[4:8])
4645 for chunk
in chunks
:
4646 if chunk
['type'] == b
'IDAT':
4647 idat
+= chunk
['data']
4650 raise OSError('Unable to read PNG data.')
4652 decompressed_data
= bytearray(zlib
.decompress(idat
))
4657 def _get_pixel(idx
):
4662 for y
in range(height
):
4663 basePos
= y
* (1 + stride
)
4664 filter_type
= decompressed_data
[basePos
]
4668 pixels
.append(current_row
)
4670 for x
in range(stride
):
4671 color
= decompressed_data
[1 + basePos
+ x
]
4672 basex
= y
* stride
+ x
4677 left
= _get_pixel(basex
- 3)
4679 up
= _get_pixel(basex
- stride
)
4681 if filter_type
== 1: # Sub
4682 color
= (color
+ left
) & 0xff
4683 elif filter_type
== 2: # Up
4684 color
= (color
+ up
) & 0xff
4685 elif filter_type
== 3: # Average
4686 color
= (color
+ ((left
+ up
) >> 1)) & 0xff
4687 elif filter_type
== 4: # Paeth
4693 c
= _get_pixel(basex
- stride
- 3)
4701 if pa
<= pb
and pa
<= pc
:
4702 color
= (color
+ a
) & 0xff
4704 color
= (color
+ b
) & 0xff
4706 color
= (color
+ c
) & 0xff
4708 current_row
.append(color
)
4710 return width
, height
, pixels
4713 def write_xattr(path
, key
, value
):
4714 # Windows: Write xattrs to NTFS Alternate Data Streams:
4715 # http://en.wikipedia.org/wiki/NTFS#Alternate_data_streams_.28ADS.29
4716 if compat_os_name
== 'nt':
4717 assert ':' not in key
4718 assert os
.path
.exists(path
)
4721 with open(f
'{path}:{key}', 'wb') as f
:
4723 except OSError as e
:
4724 raise XAttrMetadataError(e
.errno
, e
.strerror
)
4727 # UNIX Method 1. Use xattrs/pyxattrs modules
4728 from .dependencies
import xattr
4731 if getattr(xattr
, '_yt_dlp__identifier', None) == 'pyxattr':
4732 # Unicode arguments are not supported in pyxattr until version 0.5.0
4733 # See https://github.com/ytdl-org/youtube-dl/issues/5498
4734 if version_tuple(xattr
.__version
__) >= (0, 5, 0):
4735 setxattr
= xattr
.set
4737 setxattr
= xattr
.setxattr
4741 setxattr(path
, key
, value
)
4742 except OSError as e
:
4743 raise XAttrMetadataError(e
.errno
, e
.strerror
)
4746 # UNIX Method 2. Use setfattr/xattr executables
4747 exe
= ('setfattr' if check_executable('setfattr', ['--version'])
4748 else 'xattr' if check_executable('xattr', ['-h']) else None)
4750 raise XAttrUnavailableError(
4751 'Couldn\'t find a tool to set the xattrs. Install either the python "xattr" or "pyxattr" modules or the '
4752 + ('"xattr" binary' if sys
.platform
!= 'linux' else 'GNU "attr" package (which contains the "setfattr" tool)'))
4754 value
= value
.decode()
4757 [exe
, '-w', key
, value
, path
] if exe
== 'xattr' else [exe
, '-n', key
, '-v', value
, path
],
4758 stdout
=subprocess
.PIPE
, stderr
=subprocess
.PIPE
, stdin
=subprocess
.PIPE
)
4759 except OSError as e
:
4760 raise XAttrMetadataError(e
.errno
, e
.strerror
)
4761 stderr
= p
.communicate_or_kill()[1].decode('utf-8', 'replace')
4763 raise XAttrMetadataError(p
.returncode
, stderr
)
4766 def random_birthday(year_field
, month_field
, day_field
):
4767 start_date
= datetime
.date(1950, 1, 1)
4768 end_date
= datetime
.date(1995, 12, 31)
4769 offset
= random
.randint(0, (end_date
- start_date
).days
)
4770 random_date
= start_date
+ datetime
.timedelta(offset
)
4772 year_field
: str(random_date
.year
),
4773 month_field
: str(random_date
.month
),
4774 day_field
: str(random_date
.day
),
4778 # Templates for internet shortcut files, which are plain text files.
4779 DOT_URL_LINK_TEMPLATE
= '''\
4784 DOT_WEBLOC_LINK_TEMPLATE
= '''\
4785 <?xml version="1.0" encoding="UTF-8"?>
4786 <!DOCTYPE plist PUBLIC "-//Apple//DTD PLIST 1.0//EN" "http://www.apple.com/DTDs/PropertyList-1.0.dtd">
4787 <plist version="1.0">
4790 \t<string>%(url)s</string>
4795 DOT_DESKTOP_LINK_TEMPLATE
= '''\
4805 'url': DOT_URL_LINK_TEMPLATE
,
4806 'desktop': DOT_DESKTOP_LINK_TEMPLATE
,
4807 'webloc': DOT_WEBLOC_LINK_TEMPLATE
,
4811 def iri_to_uri(iri
):
4813 Converts an IRI (Internationalized Resource Identifier, allowing Unicode characters) to a URI (Uniform Resource Identifier, ASCII-only).
4815 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.
4818 iri_parts
= compat_urllib_parse_urlparse(iri
)
4820 if '[' in iri_parts
.netloc
:
4821 raise ValueError('IPv6 URIs are not, yet, supported.')
4822 # Querying `.netloc`, when there's only one bracket, also raises a ValueError.
4824 # 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.
4827 if iri_parts
.username
:
4828 net_location
+= urllib
.parse
.quote(iri_parts
.username
, safe
=r
"!$%&'()*+,~")
4829 if iri_parts
.password
is not None:
4830 net_location
+= ':' + urllib
.parse
.quote(iri_parts
.password
, safe
=r
"!$%&'()*+,~")
4833 net_location
+= iri_parts
.hostname
.encode('idna').decode() # Punycode for Unicode hostnames.
4834 # The 'idna' encoding produces ASCII text.
4835 if iri_parts
.port
is not None and iri_parts
.port
!= 80:
4836 net_location
+= ':' + str(iri_parts
.port
)
4838 return urllib
.parse
.urlunparse(
4842 urllib
.parse
.quote_plus(iri_parts
.path
, safe
=r
"!$%&'()*+,/:;=@|~"),
4844 # Unsure about the `safe` argument, since this is a legacy way of handling parameters.
4845 urllib
.parse
.quote_plus(iri_parts
.params
, safe
=r
"!$%&'()*+,/:;=@|~"),
4847 # Not totally sure about the `safe` argument, since the source does not explicitly mention the query URI component.
4848 urllib
.parse
.quote_plus(iri_parts
.query
, safe
=r
"!$%&'()*+,/:;=?@{|}~"),
4850 urllib
.parse
.quote_plus(iri_parts
.fragment
, safe
=r
"!#$%&'()*+,/:;=?@{|}~")))
4852 # Source for `safe` arguments: https://url.spec.whatwg.org/#percent-encoded-bytes.
4855 def to_high_limit_path(path
):
4856 if sys
.platform
in ['win32', 'cygwin']:
4857 # Work around MAX_PATH limitation on Windows. The maximum allowed length for the individual path segments may still be quite limited.
4858 return '\\\\?\\' + os
.path
.abspath(path
)
4863 def format_field(obj
, field
=None, template
='%s', ignore
=(None, ''), default
='', func
=None):
4864 val
= traverse_obj(obj
, *variadic(field
))
4867 return template
% (func(val
) if func
else val
)
4870 def clean_podcast_url(url
):
4871 return re
.sub(r
'''(?x)
4875 media\.blubrry\.com| # https://create.blubrry.com/resources/podcast-media-download-statistics/getting-started/
4878 (?:dts|www)\.podtrac\.com/(?:pts/)?redirect\.[0-9a-z]{3,4}| # http://analytics.podtrac.com/how-to-measure
4881 cn\.co| # https://podcorn.com/analytics-prefix/
4882 st\.fm # https://podsights.com/docs/
4887 _HEX_TABLE
= '0123456789abcdef'
4890 def random_uuidv4():
4891 return re
.sub(r
'[xy]', lambda x
: _HEX_TABLE
[random
.randint(0, 15)], 'xxxxxxxx-xxxx-4xxx-yxxx-xxxxxxxxxxxx')
4894 def make_dir(path
, to_screen
=None):
4896 dn
= os
.path
.dirname(path
)
4897 if dn
and not os
.path
.exists(dn
):
4900 except OSError as err
:
4901 if callable(to_screen
) is not None:
4902 to_screen('unable to create directory ' + error_to_compat_str(err
))
4906 def get_executable_path():
4907 from .update
import _get_variant_and_executable_path
4909 return os
.path
.dirname(os
.path
.abspath(_get_variant_and_executable_path()[1]))
4912 def load_plugins(name
, suffix
, namespace
):
4914 with contextlib
.suppress(FileNotFoundError
):
4915 plugins_spec
= importlib
.util
.spec_from_file_location(
4916 name
, os
.path
.join(get_executable_path(), 'ytdlp_plugins', name
, '__init__.py'))
4917 plugins
= importlib
.util
.module_from_spec(plugins_spec
)
4918 sys
.modules
[plugins_spec
.name
] = plugins
4919 plugins_spec
.loader
.exec_module(plugins
)
4920 for name
in dir(plugins
):
4921 if name
in namespace
:
4923 if not name
.endswith(suffix
):
4925 klass
= getattr(plugins
, name
)
4926 classes
[name
] = namespace
[name
] = klass
4931 obj
, *path_list
, default
=None, expected_type
=None, get_all
=True,
4932 casesense
=True, is_user_input
=False, traverse_string
=False):
4933 ''' Traverse nested list/dict/tuple
4934 @param path_list A list of paths which are checked one by one.
4935 Each path is a list of keys where each key is a:
4937 - string: A dictionary key
4938 - int: An index into a list
4939 - tuple: A list of keys all of which will be traversed
4940 - Ellipsis: Fetch all values in the object
4941 - Function: Takes the key and value as arguments
4942 and returns whether the key matches or not
4943 @param default Default value to return
4944 @param expected_type Only accept final value of this type (Can also be any callable)
4945 @param get_all Return all the values obtained from a path or only the first one
4946 @param casesense Whether to consider dictionary keys as case sensitive
4947 @param is_user_input Whether the keys are generated from user input. If True,
4948 strings are converted to int/slice if necessary
4949 @param traverse_string Whether to traverse inside strings. If True, any
4950 non-compatible object will also be converted into a string
4954 _lower
= lambda k
: (k
.lower() if isinstance(k
, str) else k
)
4955 path_list
= (map(_lower
, variadic(path
)) for path
in path_list
)
4957 def _traverse_obj(obj
, path
, _current_depth
=0):
4959 path
= tuple(variadic(path
))
4960 for i
, key
in enumerate(path
):
4961 if None in (key
, obj
):
4963 if isinstance(key
, (list, tuple)):
4964 obj
= [_traverse_obj(obj
, sub_key
, _current_depth
) for sub_key
in key
]
4967 obj
= (obj
.values() if isinstance(obj
, dict)
4968 else obj
if isinstance(obj
, (list, tuple, LazyList
))
4969 else str(obj
) if traverse_string
else [])
4971 depth
= max(depth
, _current_depth
)
4972 return [_traverse_obj(inner_obj
, path
[i
+ 1:], _current_depth
) for inner_obj
in obj
]
4974 if isinstance(obj
, (list, tuple, LazyList
)):
4975 obj
= enumerate(obj
)
4976 elif isinstance(obj
, dict):
4979 if not traverse_string
:
4983 depth
= max(depth
, _current_depth
)
4984 return [_traverse_obj(v
, path
[i
+ 1:], _current_depth
) for k
, v
in obj
if try_call(key
, args
=(k
, v
))]
4985 elif isinstance(obj
, dict) and not (is_user_input
and key
== ':'):
4986 obj
= (obj
.get(key
) if casesense
or (key
in obj
)
4987 else next((v
for k
, v
in obj
.items() if _lower(k
) == key
), None))
4990 key
= (int_or_none(key
) if ':' not in key
4991 else slice(*map(int_or_none
, key
.split(':'))))
4992 if key
== slice(None):
4993 return _traverse_obj(obj
, (..., *path
[i
+ 1:]), _current_depth
)
4994 if not isinstance(key
, (int, slice)):
4996 if not isinstance(obj
, (list, tuple, LazyList
)):
4997 if not traverse_string
:
5006 if isinstance(expected_type
, type):
5007 type_test
= lambda val
: val
if isinstance(val
, expected_type
) else None
5008 elif expected_type
is not None:
5009 type_test
= expected_type
5011 type_test
= lambda val
: val
5013 for path
in path_list
:
5015 val
= _traverse_obj(obj
, path
)
5018 for _
in range(depth
- 1):
5019 val
= itertools
.chain
.from_iterable(v
for v
in val
if v
is not None)
5020 val
= [v
for v
in map(type_test
, val
) if v
is not None]
5022 return val
if get_all
else val
[0]
5024 val
= type_test(val
)
5030 def traverse_dict(dictn
, keys
, casesense
=True):
5031 write_string('DeprecationWarning: yt_dlp.utils.traverse_dict is deprecated '
5032 'and may be removed in a future version. Use yt_dlp.utils.traverse_obj instead')
5033 return traverse_obj(dictn
, keys
, casesense
=casesense
, is_user_input
=True, traverse_string
=True)
5036 def get_first(obj
, keys
, **kwargs
):
5037 return traverse_obj(obj
, (..., *variadic(keys
)), **kwargs
, get_all
=False)
5040 def variadic(x
, allowed_types
=(str, bytes, dict)):
5041 return x
if isinstance(x
, collections
.abc
.Iterable
) and not isinstance(x
, allowed_types
) else (x
,)
5044 def decode_base(value
, digits
):
5045 # This will convert given base-x string to scalar (long or int)
5046 table
= {char: index for index, char in enumerate(digits)}
5051 result
+= table
[chr]
5055 def time_seconds(**kwargs
):
5056 t
= datetime
.datetime
.now(datetime
.timezone(datetime
.timedelta(**kwargs
)))
5057 return t
.timestamp()
5060 # create a JSON Web Signature (jws) with HS256 algorithm
5061 # the resulting format is in JWS Compact Serialization
5062 # implemented following JWT https://www.rfc-editor.org/rfc/rfc7519.html
5063 # implemented following JWS https://www.rfc-editor.org/rfc/rfc7515.html
5064 def jwt_encode_hs256(payload_data
, key
, headers
={}):
5070 header_data
.update(headers
)
5071 header_b64
= base64
.b64encode(json
.dumps(header_data
).encode())
5072 payload_b64
= base64
.b64encode(json
.dumps(payload_data
).encode())
5073 h
= hmac
.new(key
.encode(), header_b64
+ b
'.' + payload_b64
, hashlib
.sha256
)
5074 signature_b64
= base64
.b64encode(h
.digest())
5075 token
= header_b64
+ b
'.' + payload_b64
+ b
'.' + signature_b64
5079 # can be extended in future to verify the signature and parse header and return the algorithm used if it's not HS256
5080 def jwt_decode_hs256(jwt
):
5081 header_b64
, payload_b64
, signature_b64
= jwt
.split('.')
5082 payload_data
= json
.loads(base64
.urlsafe_b64decode(payload_b64
))
5086 WINDOWS_VT_MODE
= False if compat_os_name
== 'nt' else None
5090 def supports_terminal_sequences(stream
):
5091 if compat_os_name
== 'nt':
5092 if not WINDOWS_VT_MODE
or get_windows_version() < (10, 0, 10586):
5094 elif not os
.getenv('TERM'):
5097 return stream
.isatty()
5098 except BaseException
:
5102 def windows_enable_vt_mode(): # TODO: Do this the proper way https://bugs.python.org/issue30075
5103 if compat_os_name
!= 'nt':
5105 global WINDOWS_VT_MODE
5106 startupinfo
= subprocess
.STARTUPINFO()
5107 startupinfo
.dwFlags |
= subprocess
.STARTF_USESHOWWINDOW
5109 subprocess
.Popen('', shell
=True, startupinfo
=startupinfo
).wait()
5113 WINDOWS_VT_MODE
= True
5114 supports_terminal_sequences
.cache_clear()
5117 _terminal_sequences_re
= re
.compile('\033\\[[^m]+m')
5120 def remove_terminal_sequences(string
):
5121 return _terminal_sequences_re
.sub('', string
)
5124 def number_of_digits(number
):
5125 return len('%d' % number
)
5128 def join_nonempty(*values
, delim
='-', from_dict
=None):
5129 if from_dict
is not None:
5130 values
= map(from_dict
.get
, values
)
5131 return delim
.join(map(str, filter(None, values
)))
5134 def scale_thumbnails_to_max_format_width(formats
, thumbnails
, url_width_re
):
5136 Find the largest format dimensions in terms of video width and, for each thumbnail:
5137 * Modify the URL: Match the width with the provided regex and replace with the former width
5140 This function is useful with video services that scale the provided thumbnails on demand
5142 _keys
= ('width', 'height')
5143 max_dimensions
= max(
5144 (tuple(format
.get(k
) or 0 for k
in _keys
) for format
in formats
),
5146 if not max_dimensions
[0]:
5150 {'url': re.sub(url_width_re, str(max_dimensions[0]), thumbnail['url'])}
,
5151 dict(zip(_keys
, max_dimensions
)), thumbnail
)
5152 for thumbnail
in thumbnails
5156 def parse_http_range(range):
5157 """ Parse value of "Range" or "Content-Range" HTTP header into tuple. """
5159 return None, None, None
5160 crg
= re
.search(r
'bytes[ =](\d+)-(\d+)?(?:/(\d+))?', range)
5162 return None, None, None
5163 return int(crg
.group(1)), int_or_none(crg
.group(2)), int_or_none(crg
.group(3))
5166 def read_stdin(what
):
5167 eof
= 'Ctrl+Z' if compat_os_name
== 'nt' else 'Ctrl+D'
5168 write_string(f
'Reading {what} from STDIN - EOF ({eof}) to end:\n')
5176 __initialized
= False
5178 def __init__(self
, parser
, label
=None):
5179 self
.parser
, self
.label
= parser
, label
5180 self
._loaded
_paths
, self
.configs
= set(), []
5182 def init(self
, args
=None, filename
=None):
5183 assert not self
.__initialized
5186 location
= os
.path
.realpath(filename
)
5187 directory
= os
.path
.dirname(location
)
5188 if location
in self
._loaded
_paths
:
5190 self
._loaded
_paths
.add(location
)
5192 self
.own_args
, self
.__initialized
= args
, True
5193 opts
, _
= self
.parser
.parse_known_args(args
)
5194 self
.parsed_args
, self
.filename
= args
, filename
5196 for location
in opts
.config_locations
or []:
5198 self
.append_config(shlex
.split(read_stdin('options'), comments
=True), label
='stdin')
5200 location
= os
.path
.join(directory
, expand_path(location
))
5201 if os
.path
.isdir(location
):
5202 location
= os
.path
.join(location
, 'yt-dlp.conf')
5203 if not os
.path
.exists(location
):
5204 self
.parser
.error(f
'config location {location} does not exist')
5205 self
.append_config(self
.read_file(location
), location
)
5209 label
= join_nonempty(
5210 self
.label
, 'config', f
'"{self.filename}"' if self
.filename
else '',
5212 return join_nonempty(
5213 self
.own_args
is not None and f
'{label[0].upper()}{label[1:]}: {self.hide_login_info(self.own_args)}',
5214 *(f
'\n{c}'.replace('\n', '\n| ')[1:] for c
in self
.configs
),
5218 def read_file(filename
, default
=[]):
5220 optionf
= open(filename
)
5222 return default
# silently skip if file is not present
5224 # FIXME: https://github.com/ytdl-org/youtube-dl/commit/dfe5fa49aed02cf36ba9f743b11b0903554b5e56
5225 contents
= optionf
.read()
5226 res
= shlex
.split(contents
, comments
=True)
5232 def hide_login_info(opts
):
5233 PRIVATE_OPTS
= {'-p', '--password', '-u', '--username', '--video-password', '--ap-password', '--ap-username'}
5234 eqre
= re
.compile('^(?P<key>' + ('|'.join(re
.escape(po
) for po
in PRIVATE_OPTS
)) + ')=.+$')
5239 return m
.group('key') + '=PRIVATE'
5243 opts
= list(map(_scrub_eq
, opts
))
5244 for idx
, opt
in enumerate(opts
):
5245 if opt
in PRIVATE_OPTS
and idx
+ 1 < len(opts
):
5246 opts
[idx
+ 1] = 'PRIVATE'
5249 def append_config(self
, *args
, label
=None):
5250 config
= type(self
)(self
.parser
, label
)
5251 config
._loaded
_paths
= self
._loaded
_paths
5252 if config
.init(*args
):
5253 self
.configs
.append(config
)
5257 for config
in reversed(self
.configs
):
5258 yield from config
.all_args
5259 yield from self
.parsed_args
or []
5261 def parse_known_args(self
, **kwargs
):
5262 return self
.parser
.parse_known_args(self
.all_args
, **kwargs
)
5264 def parse_args(self
):
5265 return self
.parser
.parse_args(self
.all_args
)
5268 class WebSocketsWrapper():
5269 """Wraps websockets module to use in non-async scopes"""
5272 def __init__(self
, url
, headers
=None, connect
=True):
5273 self
.loop
= asyncio
.new_event_loop()
5274 # XXX: "loop" is deprecated
5275 self
.conn
= websockets
.connect(
5276 url
, extra_headers
=headers
, ping_interval
=None,
5277 close_timeout
=float('inf'), loop
=self
.loop
, ping_timeout
=float('inf'))
5280 atexit
.register(self
.__exit
__, None, None, None)
5282 def __enter__(self
):
5284 self
.pool
= self
.run_with_loop(self
.conn
.__aenter
__(), self
.loop
)
5287 def send(self
, *args
):
5288 self
.run_with_loop(self
.pool
.send(*args
), self
.loop
)
5290 def recv(self
, *args
):
5291 return self
.run_with_loop(self
.pool
.recv(*args
), self
.loop
)
5293 def __exit__(self
, type, value
, traceback
):
5295 return self
.run_with_loop(self
.conn
.__aexit
__(type, value
, traceback
), self
.loop
)
5298 self
._cancel
_all
_tasks
(self
.loop
)
5300 # taken from https://github.com/python/cpython/blob/3.9/Lib/asyncio/runners.py with modifications
5301 # for contributors: If there's any new library using asyncio needs to be run in non-async, move these function out of this class
5303 def run_with_loop(main
, loop
):
5304 if not asyncio
.iscoroutine(main
):
5305 raise ValueError(f
'a coroutine was expected, got {main!r}')
5308 return loop
.run_until_complete(main
)
5310 loop
.run_until_complete(loop
.shutdown_asyncgens())
5311 if hasattr(loop
, 'shutdown_default_executor'):
5312 loop
.run_until_complete(loop
.shutdown_default_executor())
5315 def _cancel_all_tasks(loop
):
5316 to_cancel
= asyncio
.all_tasks(loop
)
5321 for task
in to_cancel
:
5324 # XXX: "loop" is removed in python 3.10+
5325 loop
.run_until_complete(
5326 asyncio
.gather(*to_cancel
, loop
=loop
, return_exceptions
=True))
5328 for task
in to_cancel
:
5329 if task
.cancelled():
5331 if task
.exception() is not None:
5332 loop
.call_exception_handler({
5333 'message': 'unhandled exception during asyncio.run() shutdown',
5334 'exception': task
.exception(),
5339 def merge_headers(*dicts
):
5340 """Merge dicts of http headers case insensitively, prioritizing the latter ones"""
5341 return {k.title(): v for k, v in itertools.chain.from_iterable(map(dict.items, dicts))}
5344 class classproperty
:
5345 """classmethod(property(func)) that works in py < 3.9"""
5347 def __init__(self
, func
):
5348 functools
.update_wrapper(self
, func
)
5351 def __get__(self
, _
, cls
):
5352 return self
.func(cls
)
5356 """Immutable namespace"""
5358 def __init__(self
, **kwargs
):
5361 def __getattr__(self
, attr
):
5362 return self
._dict
[attr
]
5364 def __contains__(self
, item
):
5365 return item
in self
._dict
.values()
5368 return iter(self
._dict
.items())
5371 return f
'{type(self).__name__}({", ".join(f"{k}={v}" for k, v in self)})'
5375 has_certifi
= bool(certifi
)
5376 has_websockets
= bool(websockets
)