40 import xml
.etree
.ElementTree
47 compat_etree_fromstring
,
50 compat_html_entities_html5
,
51 compat_HTMLParseError
,
62 compat_urllib_parse_unquote_plus
,
63 compat_urllib_parse_urlencode
,
64 compat_urllib_parse_urlparse
,
65 compat_urllib_request
,
69 from .socks
import ProxyType
, sockssocket
78 def register_socks_protocols():
79 # "Register" SOCKS protocols
80 # In Python < 2.6.5, urlsplit() suffers from bug https://bugs.python.org/issue7904
81 # URLs with protocols not in urlparse.uses_netloc are not handled correctly
82 for scheme
in ('socks', 'socks4', 'socks4a', 'socks5'):
83 if scheme
not in compat_urlparse
.uses_netloc
:
84 compat_urlparse
.uses_netloc
.append(scheme
)
87 # This is not clearly defined otherwise
88 compiled_regex_type
= type(re
.compile(''))
91 def random_user_agent():
92 _USER_AGENT_TPL
= 'Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/%s Safari/537.36'
133 return _USER_AGENT_TPL
% random
.choice(_CHROME_VERSIONS
)
136 SUPPORTED_ENCODINGS
= [
140 SUPPORTED_ENCODINGS
.append('br')
143 'User-Agent': random_user_agent(),
144 'Accept': 'text/html,application/xhtml+xml,application/xml;q=0.9,*/*;q=0.8',
145 'Accept-Language': 'en-us,en;q=0.5',
146 'Sec-Fetch-Mode': 'navigate',
151 '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',
155 NO_DEFAULT
= object()
157 ENGLISH_MONTH_NAMES
= [
158 'January', 'February', 'March', 'April', 'May', 'June',
159 'July', 'August', 'September', 'October', 'November', 'December']
162 'en': ENGLISH_MONTH_NAMES
,
164 'janvier', 'février', 'mars', 'avril', 'mai', 'juin',
165 'juillet', 'août', 'septembre', 'octobre', 'novembre', 'décembre'],
169 'mp4', 'm4a', 'm4p', 'm4b', 'm4r', 'm4v', 'aac',
170 'flv', 'f4v', 'f4a', 'f4b',
171 'webm', 'ogg', 'ogv', 'oga', 'ogx', 'spx', 'opus',
172 'mkv', 'mka', 'mk3d',
181 'f4f', 'f4m', 'm3u8', 'smil')
183 # needed for sanitizing filenames in restricted mode
184 ACCENT_CHARS
= dict(zip('ÂÃÄÀÁÅÆÇÈÉÊËÌÍÎÏÐÑÒÓÔÕÖŐØŒÙÚÛÜŰÝÞßàáâãäåæçèéêëìíîïðñòóôõöőøœùúûüűýþÿ',
185 itertools
.chain('AAAAAA', ['AE'], 'CEEEEIIIIDNOOOOOOO', ['OE'], 'UUUUUY', ['TH', 'ss'],
186 'aaaaaa', ['ae'], 'ceeeeiiiionooooooo', ['oe'], 'uuuuuy', ['th'], 'y')))
216 '%Y-%m-%d %H:%M:%S.%f',
217 '%Y-%m-%d %H:%M:%S:%f',
220 '%Y-%m-%dT%H:%M:%SZ',
221 '%Y-%m-%dT%H:%M:%S.%fZ',
222 '%Y-%m-%dT%H:%M:%S.%f0Z',
224 '%Y-%m-%dT%H:%M:%S.%f',
227 '%b %d %Y at %H:%M:%S',
229 '%B %d %Y at %H:%M:%S',
233 DATE_FORMATS_DAY_FIRST
= list(DATE_FORMATS
)
234 DATE_FORMATS_DAY_FIRST
.extend([
243 DATE_FORMATS_MONTH_FIRST
= list(DATE_FORMATS
)
244 DATE_FORMATS_MONTH_FIRST
.extend([
252 PACKED_CODES_RE
= r
"}\('(.+)',(\d+),(\d+),'([^']+)'\.split\('\|'\)"
253 JSON_LD_RE
= r
'(?is)<script[^>]+type=(["\']?
)application
/ld\
+json\
1[^
>]*>(?P
<json_ld
>.+?
)</script
>'
256 def preferredencoding():
257 """Get preferred encoding.
259 Returns the best encoding scheme for the system, based on
260 locale.getpreferredencoding() and some further tweaks.
263 pref = locale.getpreferredencoding()
271 def write_json_file(obj, fn):
272 """ Encode obj as JSON and write it to fn, atomically if possible """
274 tf = tempfile.NamedTemporaryFile(
275 prefix=f'{os.path.basename(fn)}
.', dir=os.path.dirname(fn),
276 suffix='.tmp
', delete=False, mode='w
', encoding='utf
-8')
280 json.dump(obj, tf, ensure_ascii=False)
281 if sys.platform == 'win32
':
282 # Need to remove existing file on Windows, else os.rename raises
283 # WindowsError or FileExistsError.
291 os.chmod(tf.name, 0o666 & ~mask)
294 os.rename(tf.name, fn)
303 def find_xpath_attr(node, xpath, key, val=None):
304 """ Find the xpath xpath[@key=val] """
305 assert re.match(r'^
[a
-zA
-Z_
-]+$
', key)
306 expr = xpath + ('[@%s]' % key if val is None else f"[@{key}='{val}
']")
307 return node.find(expr)
309 # On python2.6 the xml.etree.ElementTree.Element methods don't support
310 # the namespace parameter
313 def xpath_with_ns(path
, ns_map
):
314 components
= [c
.split(':') for c
in path
.split('/')]
318 replaced
.append(c
[0])
321 replaced
.append('{%s}%s' % (ns_map
[ns
], tag
))
322 return '/'.join(replaced
)
325 def xpath_element(node
, xpath
, name
=None, fatal
=False, default
=NO_DEFAULT
):
326 def _find_xpath(xpath
):
327 return node
.find(xpath
)
329 if isinstance(xpath
, (str, compat_str
)):
330 n
= _find_xpath(xpath
)
338 if default
is not NO_DEFAULT
:
341 name
= xpath
if name
is None else name
342 raise ExtractorError('Could not find XML element %s' % name
)
348 def xpath_text(node
, xpath
, name
=None, fatal
=False, default
=NO_DEFAULT
):
349 n
= xpath_element(node
, xpath
, name
, fatal
=fatal
, default
=default
)
350 if n
is None or n
== default
:
353 if default
is not NO_DEFAULT
:
356 name
= xpath
if name
is None else name
357 raise ExtractorError('Could not find XML element\'s text %s' % name
)
363 def xpath_attr(node
, xpath
, key
, name
=None, fatal
=False, default
=NO_DEFAULT
):
364 n
= find_xpath_attr(node
, xpath
, key
)
366 if default
is not NO_DEFAULT
:
369 name
= f
'{xpath}[@{key}]' if name
is None else name
370 raise ExtractorError('Could not find XML attribute %s' % name
)
376 def get_element_by_id(id, html
):
377 """Return the content of the tag with the specified ID in the passed HTML document"""
378 return get_element_by_attribute('id', id, html
)
381 def get_element_html_by_id(id, html
):
382 """Return the html of the tag with the specified ID in the passed HTML document"""
383 return get_element_html_by_attribute('id', id, html
)
386 def get_element_by_class(class_name
, html
):
387 """Return the content of the first tag with the specified class in the passed HTML document"""
388 retval
= get_elements_by_class(class_name
, html
)
389 return retval
[0] if retval
else None
392 def get_element_html_by_class(class_name
, html
):
393 """Return the html of the first tag with the specified class in the passed HTML document"""
394 retval
= get_elements_html_by_class(class_name
, html
)
395 return retval
[0] if retval
else None
398 def get_element_by_attribute(attribute
, value
, html
, escape_value
=True):
399 retval
= get_elements_by_attribute(attribute
, value
, html
, escape_value
)
400 return retval
[0] if retval
else None
403 def get_element_html_by_attribute(attribute
, value
, html
, escape_value
=True):
404 retval
= get_elements_html_by_attribute(attribute
, value
, html
, escape_value
)
405 return retval
[0] if retval
else None
408 def get_elements_by_class(class_name
, html
):
409 """Return the content of all tags with the specified class in the passed HTML document as a list"""
410 return get_elements_by_attribute(
411 'class', r
'[^\'"]*\b%s\b[^\'"]*' % re.escape(class_name),
412 html, escape_value=False)
415 def get_elements_html_by_class(class_name, html):
416 """Return the html of all tags with the specified class in the passed HTML document as a list"""
417 return get_elements_html_by_attribute(
418 'class', r'[^
\'"]*\b%s\b[^\'"]*' % re.escape(class_name),
419 html, escape_value=False)
422 def get_elements_by_attribute(*args, **kwargs):
423 """Return the content of the tag with the specified attribute in the passed HTML document"""
424 return [content for content, _ in get_elements_text_and_html_by_attribute(*args, **kwargs)]
427 def get_elements_html_by_attribute(*args, **kwargs):
428 """Return the html of the tag with the specified attribute in the passed HTML document"""
429 return [whole for _, whole in get_elements_text_and_html_by_attribute(*args, **kwargs)]
432 def get_elements_text_and_html_by_attribute(attribute, value, html, escape_value=True):
434 Return the text (content) and the html (whole) of the tag with the specified
435 attribute in the passed HTML document
438 quote = '' if re.match(r'''[\s"'`
=<>]''', value) else '?'
440 value = re.escape(value) if escape_value else value
442 partial_element_re = rf'''(?x
)
443 <(?P
<tag
>[a
-zA
-Z0
-9:._-]+)
444 (?
:\
s(?
:[^
>"']|"[^
"]*"|
'[^']*')*)?
445 \s{re.escape(attribute)}\s*=\s*(?P<_q>['"]{quote})(?-x:{value})(?P=_q)
448 for m in re.finditer(partial_element_re, html):
449 content, whole = get_element_text_and_html_by_tag(m.group('tag'), html[m.start():])
452 unescapeHTML(re.sub(r'^(?P<q>["\'])(?P
<content
>.*)(?P
=q
)$
', r'\g
<content
>', content, flags=re.DOTALL)),
457 class HTMLBreakOnClosingTagParser(compat_HTMLParser):
459 HTML parser which raises HTMLBreakOnClosingTagException upon reaching the
460 closing tag for the first opening tag it has encountered, and can be used
464 class HTMLBreakOnClosingTagException(Exception):
468 self.tagstack = collections.deque()
469 compat_HTMLParser.__init__(self)
474 def __exit__(self, *_):
478 # handle_endtag does not return upon raising HTMLBreakOnClosingTagException,
479 # so data remains buffered; we no longer have any interest in it, thus
480 # override this method to discard it
483 def handle_starttag(self, tag, _):
484 self.tagstack.append(tag)
486 def handle_endtag(self, tag):
487 if not self.tagstack:
488 raise compat_HTMLParseError('no tags
in the stack
')
490 inner_tag = self.tagstack.pop()
494 raise compat_HTMLParseError(f'matching opening tag
for closing {tag} tag
not found
')
495 if not self.tagstack:
496 raise self.HTMLBreakOnClosingTagException()
499 def get_element_text_and_html_by_tag(tag, html):
501 For the first element with the specified tag in the passed HTML document
502 return its' content (text
) and the whole
element (html
)
504 def find_or_raise(haystack, needle, exc):
506 return haystack.index(needle)
509 closing_tag = f'</{tag}>'
510 whole_start = find_or_raise(
511 html, f'<{tag}', compat_HTMLParseError(f'opening {tag} tag not found'))
512 content_start = find_or_raise(
513 html[whole_start:], '>', compat_HTMLParseError(f'malformed opening {tag} tag'))
514 content_start += whole_start + 1
515 with HTMLBreakOnClosingTagParser() as parser:
516 parser.feed(html[whole_start:content_start])
517 if not parser.tagstack or parser.tagstack[0] != tag:
518 raise compat_HTMLParseError(f'parser did not match opening {tag} tag')
519 offset = content_start
520 while offset < len(html):
521 next_closing_tag_start = find_or_raise(
522 html[offset:], closing_tag,
523 compat_HTMLParseError(f'closing {tag} tag not found'))
524 next_closing_tag_end = next_closing_tag_start + len(closing_tag)
526 parser.feed(html[offset:offset + next_closing_tag_end])
527 offset += next_closing_tag_end
528 except HTMLBreakOnClosingTagParser.HTMLBreakOnClosingTagException:
529 return html[content_start:offset + next_closing_tag_start], \
530 html[whole_start:offset + next_closing_tag_end]
531 raise compat_HTMLParseError('unexpected end of html')
534 class HTMLAttributeParser(compat_HTMLParser):
535 """Trivial HTML parser to gather the attributes
for a single element
"""
539 compat_HTMLParser.__init__(self)
541 def handle_starttag(self, tag, attrs):
542 self.attrs = dict(attrs)
545 class HTMLListAttrsParser(compat_HTMLParser):
546 """HTML parser to gather the attributes
for the elements of a
list"""
549 compat_HTMLParser.__init__(self)
553 def handle_starttag(self, tag, attrs):
554 if tag == 'li' and self._level == 0:
555 self.items.append(dict(attrs))
558 def handle_endtag(self, tag):
562 def extract_attributes(html_element):
563 """Given a string
for an HTML element such
as
565 a
="foo" B
="bar" c
="&98;az" d
=boz
566 empty
= noval entity
="&"
569 Decode
and return a dictionary of attributes
.
571 'a': 'foo', 'b': 'bar', c
: 'baz', d
: 'boz',
572 'empty': '', 'noval': None, 'entity': '&',
573 'sq': '"', 'dq': '\''
576 parser = HTMLAttributeParser()
578 parser.feed(html_element)
580 # Older Python may throw HTMLParseError in case of malformed HTML
581 except compat_HTMLParseError:
586 def parse_list(webpage):
587 """Given a string
for an series of HTML
<li
> elements
,
588 return a dictionary of their attributes
"""
589 parser = HTMLListAttrsParser()
595 def clean_html(html):
596 """Clean an HTML snippet into a readable string
"""
598 if html is None: # Convenience for sanitizing descriptions etc.
601 html = re.sub(r'\s+', ' ', html)
602 html = re.sub(r'(?u)\s?<\s?br\s?/?\s?>\s?', '\n', html)
603 html = re.sub(r'(?u)<\s?/\s?p\s?>\s?<\s?p[^>]*>', '\n', html)
605 html = re.sub('<.*?>', '', html)
606 # Replace html entities
607 html = unescapeHTML(html)
611 def sanitize_open(filename, open_mode):
612 """Try to
open the given filename
, and slightly tweak it
if this fails
.
614 Attempts to
open the given filename
. If this fails
, it tries to change
615 the filename slightly
, step by step
, until it
's either able to open it
616 or it fails and raises a final exception, like the standard open()
619 It returns the tuple (stream, definitive_file_name).
622 if sys.platform == 'win32
':
624 msvcrt.setmode(sys.stdout.fileno(), os.O_BINARY)
625 return (sys.stdout.buffer if hasattr(sys.stdout, 'buffer') else sys.stdout, filename)
627 for attempt in range(2):
630 if sys.platform == 'win32
':
631 # FIXME: An exclusive lock also locks the file from being read.
632 # Since windows locks are mandatory, don't lock the
file on
windows (for now
).
633 # Ref: https://github.com/yt-dlp/yt-dlp/issues/3124
634 raise LockingUnsupportedError()
635 stream
= locked_file(filename
, open_mode
, block
=False).__enter
__()
636 except LockingUnsupportedError
:
637 stream
= open(filename
, open_mode
)
638 return (stream
, filename
)
639 except OSError as err
:
640 if attempt
or err
.errno
in (errno
.EACCES
,):
642 old_filename
, filename
= filename
, sanitize_path(filename
)
643 if old_filename
== filename
:
647 def timeconvert(timestr
):
648 """Convert RFC 2822 defined time string into system timestamp"""
650 timetuple
= email
.utils
.parsedate_tz(timestr
)
651 if timetuple
is not None:
652 timestamp
= email
.utils
.mktime_tz(timetuple
)
656 def sanitize_filename(s
, restricted
=False, is_id
=NO_DEFAULT
):
657 """Sanitizes a string so it could be used as part of a filename.
658 @param restricted Use a stricter subset of allowed characters
659 @param is_id Whether this is an ID that should be kept unchanged if possible.
660 If unset, yt-dlp's new sanitization rules are in effect
665 def replace_insane(char
):
666 if restricted
and char
in ACCENT_CHARS
:
667 return ACCENT_CHARS
[char
]
668 elif not restricted
and char
== '\n':
670 elif char
== '?' or ord(char
) < 32 or ord(char
) == 127:
673 return '' if restricted
else '\''
675 return '\0_\0-' if restricted
else '\0 \0-'
676 elif char
in '\\/|*<>':
678 if restricted
and (char
in '!&\'()[]{}$;`^,#' or char
.isspace() or ord(char
) > 127):
682 s
= re
.sub(r
'[0-9]+(?::[0-9]+)+', lambda m
: m
.group(0).replace(':', '_'), s
) # Handle timestamps
683 result
= ''.join(map(replace_insane
, s
))
684 if is_id
is NO_DEFAULT
:
685 result
= re
.sub('(\0.)(?:(?=\\1)..)+', r
'\1', result
) # Remove repeated substitute chars
686 STRIP_RE
= '(?:\0.|[ _-])*'
687 result
= re
.sub(f
'^\0.{STRIP_RE}|{STRIP_RE}\0.$', '', result
) # Remove substitute chars from start/end
688 result
= result
.replace('\0', '') or '_'
691 while '__' in result
:
692 result
= result
.replace('__', '_')
693 result
= result
.strip('_')
694 # Common case of "Foreign band name - English song title"
695 if restricted
and result
.startswith('-_'):
697 if result
.startswith('-'):
698 result
= '_' + result
[len('-'):]
699 result
= result
.lstrip('.')
705 def sanitize_path(s
, force
=False):
706 """Sanitizes and normalizes path on Windows"""
707 if sys
.platform
== 'win32':
709 drive_or_unc
, _
= os
.path
.splitdrive(s
)
715 norm_path
= os
.path
.normpath(remove_start(s
, drive_or_unc
)).split(os
.path
.sep
)
719 path_part
if path_part
in ['.', '..'] else re
.sub(r
'(?:[/<>:"\|\\?\*]|[\s.]$)', '#', path_part
)
720 for path_part
in norm_path
]
722 sanitized_path
.insert(0, drive_or_unc
+ os
.path
.sep
)
723 elif force
and s
and s
[0] == os
.path
.sep
:
724 sanitized_path
.insert(0, os
.path
.sep
)
725 return os
.path
.join(*sanitized_path
)
728 def sanitize_url(url
):
729 # Prepend protocol-less URLs with `http:` scheme in order to mitigate
730 # the number of unwanted failures due to missing protocol
731 if url
.startswith('//'):
732 return 'http:%s' % url
733 # Fix some common typos seen so far
735 # https://github.com/ytdl-org/youtube-dl/issues/15649
736 (r
'^httpss://', r
'https://'),
737 # https://bx1.be/lives/direct-tv/
738 (r
'^rmtp([es]?)://', r
'rtmp\1://'),
740 for mistake
, fixup
in COMMON_TYPOS
:
741 if re
.match(mistake
, url
):
742 return re
.sub(mistake
, fixup
, url
)
746 def extract_basic_auth(url
):
747 parts
= compat_urlparse
.urlsplit(url
)
748 if parts
.username
is None:
750 url
= compat_urlparse
.urlunsplit(parts
._replace
(netloc
=(
751 parts
.hostname
if parts
.port
is None
752 else '%s:%d' % (parts
.hostname
, parts
.port
))))
753 auth_payload
= base64
.b64encode(
754 ('%s:%s' % (parts
.username
, parts
.password
or '')).encode('utf-8'))
755 return url
, 'Basic ' + auth_payload
.decode('utf-8')
758 def sanitized_Request(url
, *args
, **kwargs
):
759 url
, auth_header
= extract_basic_auth(escape_url(sanitize_url(url
)))
760 if auth_header
is not None:
761 headers
= args
[1] if len(args
) >= 2 else kwargs
.setdefault('headers', {})
762 headers
['Authorization'] = auth_header
763 return compat_urllib_request
.Request(url
, *args
, **kwargs
)
767 """Expand shell variables and ~"""
768 return os
.path
.expandvars(compat_expanduser(s
))
771 def orderedSet(iterable
):
772 """ Remove all duplicates from the input iterable """
780 def _htmlentity_transform(entity_with_semicolon
):
781 """Transforms an HTML entity to a character."""
782 entity
= entity_with_semicolon
[:-1]
784 # Known non-numeric HTML entity
785 if entity
in compat_html_entities
.name2codepoint
:
786 return compat_chr(compat_html_entities
.name2codepoint
[entity
])
788 # TODO: HTML5 allows entities without a semicolon. For example,
789 # 'Éric' should be decoded as 'Éric'.
790 if entity_with_semicolon
in compat_html_entities_html5
:
791 return compat_html_entities_html5
[entity_with_semicolon
]
793 mobj
= re
.match(r
'#(x[0-9a-fA-F]+|[0-9]+)', entity
)
795 numstr
= mobj
.group(1)
796 if numstr
.startswith('x'):
798 numstr
= '0%s' % numstr
801 # See https://github.com/ytdl-org/youtube-dl/issues/7518
803 return compat_chr(int(numstr
, base
))
807 # Unknown entity in name, return its literal representation
808 return '&%s;' % entity
814 assert type(s
) == compat_str
817 r
'&([^&;]+;)', lambda m
: _htmlentity_transform(m
.group(1)), s
)
820 def escapeHTML(text
):
823 .replace('&', '&')
824 .replace('<', '<')
825 .replace('>', '>')
826 .replace('"', '"')
827 .replace("'", ''')
831 def process_communicate_or_kill(p
, *args
, **kwargs
):
833 return p
.communicate(*args
, **kwargs
)
834 except BaseException
: # Including KeyboardInterrupt
840 class Popen(subprocess
.Popen
):
841 if sys
.platform
== 'win32':
842 _startupinfo
= subprocess
.STARTUPINFO()
843 _startupinfo
.dwFlags |
= subprocess
.STARTF_USESHOWWINDOW
847 def __init__(self
, *args
, **kwargs
):
848 super().__init
__(*args
, **kwargs
, startupinfo
=self
._startupinfo
)
850 def communicate_or_kill(self
, *args
, **kwargs
):
851 return process_communicate_or_kill(self
, *args
, **kwargs
)
854 def get_subprocess_encoding():
855 if sys
.platform
== 'win32' and sys
.getwindowsversion()[0] >= 5:
856 # For subprocess calls, encode with locale encoding
857 # Refer to http://stackoverflow.com/a/9951851/35070
858 encoding
= preferredencoding()
860 encoding
= sys
.getfilesystemencoding()
866 def encodeFilename(s
, for_subprocess
=False):
867 assert type(s
) == str
871 def decodeFilename(b
, for_subprocess
=False):
875 def encodeArgument(s
):
876 # Legacy code that uses byte strings
877 # Uncomment the following line after fixing all post processors
878 # assert isinstance(s, str), 'Internal error: %r should be of type %r, is %r' % (s, compat_str, type(s))
879 return s
if isinstance(s
, str) else s
.decode('ascii')
882 def decodeArgument(b
):
886 def decodeOption(optval
):
889 if isinstance(optval
, bytes):
890 optval
= optval
.decode(preferredencoding())
892 assert isinstance(optval
, compat_str
)
896 _timetuple
= collections
.namedtuple('Time', ('hours', 'minutes', 'seconds', 'milliseconds'))
899 def timetuple_from_msec(msec
):
900 secs
, msec
= divmod(msec
, 1000)
901 mins
, secs
= divmod(secs
, 60)
902 hrs
, mins
= divmod(mins
, 60)
903 return _timetuple(hrs
, mins
, secs
, msec
)
906 def formatSeconds(secs
, delim
=':', msec
=False):
907 time
= timetuple_from_msec(secs
* 1000)
909 ret
= '%d%s%02d%s%02d' % (time
.hours
, delim
, time
.minutes
, delim
, time
.seconds
)
911 ret
= '%d%s%02d' % (time
.minutes
, delim
, time
.seconds
)
913 ret
= '%d' % time
.seconds
914 return '%s.%03d' % (ret
, time
.milliseconds
) if msec
else ret
917 def _ssl_load_windows_store_certs(ssl_context
, storename
):
918 # Code adapted from _load_windows_store_certs in https://github.com/python/cpython/blob/main/Lib/ssl.py
920 certs
= [cert
for cert
, encoding
, trust
in ssl
.enum_certificates(storename
)
921 if encoding
== 'x509_asn' and (
922 trust
is True or ssl
.Purpose
.SERVER_AUTH
.oid
in trust
)]
923 except PermissionError
:
927 ssl_context
.load_verify_locations(cadata
=cert
)
932 def make_HTTPS_handler(params
, **kwargs
):
933 opts_check_certificate
= not params
.get('nocheckcertificate')
934 context
= ssl
.SSLContext(ssl
.PROTOCOL_TLS_CLIENT
)
935 context
.check_hostname
= opts_check_certificate
936 if params
.get('legacyserverconnect'):
937 context
.options |
= 4 # SSL_OP_LEGACY_SERVER_CONNECT
938 context
.verify_mode
= ssl
.CERT_REQUIRED
if opts_check_certificate
else ssl
.CERT_NONE
939 if opts_check_certificate
:
940 if has_certifi
and 'no-certifi' not in params
.get('compat_opts', []):
941 context
.load_verify_locations(cafile
=certifi
.where())
944 context
.load_default_certs()
945 # Work around the issue in load_default_certs when there are bad certificates. See:
946 # https://github.com/yt-dlp/yt-dlp/issues/1060,
947 # https://bugs.python.org/issue35665, https://bugs.python.org/issue45312
949 # enum_certificates is not present in mingw python. See https://github.com/yt-dlp/yt-dlp/issues/1151
950 if sys
.platform
== 'win32' and hasattr(ssl
, 'enum_certificates'):
951 # Create a new context to discard any certificates that were already loaded
952 context
= ssl
.SSLContext(ssl
.PROTOCOL_TLS_CLIENT
)
953 context
.check_hostname
, context
.verify_mode
= True, ssl
.CERT_REQUIRED
954 for storename
in ('CA', 'ROOT'):
955 _ssl_load_windows_store_certs(context
, storename
)
956 context
.set_default_verify_paths()
957 return YoutubeDLHTTPSHandler(params
, context
=context
, **kwargs
)
960 def bug_reports_message(before
=';'):
961 msg
= ('please report this issue on https://github.com/yt-dlp/yt-dlp/issues?q= , '
962 'filling out the appropriate issue template. '
963 'Confirm you are on the latest version using yt-dlp -U')
965 before
= before
.rstrip()
966 if not before
or before
.endswith(('.', '!', '?')):
967 msg
= msg
[0].title() + msg
[1:]
969 return (before
+ ' ' if before
else '') + msg
972 class YoutubeDLError(Exception):
973 """Base exception for YoutubeDL errors."""
976 def __init__(self
, msg
=None):
979 elif self
.msg
is None:
980 self
.msg
= type(self
).__name
__
981 super().__init
__(self
.msg
)
984 network_exceptions
= [compat_urllib_error
.URLError
, compat_http_client
.HTTPException
, socket
.error
]
985 if hasattr(ssl
, 'CertificateError'):
986 network_exceptions
.append(ssl
.CertificateError
)
987 network_exceptions
= tuple(network_exceptions
)
990 class ExtractorError(YoutubeDLError
):
991 """Error during info extraction."""
993 def __init__(self
, msg
, tb
=None, expected
=False, cause
=None, video_id
=None, ie
=None):
994 """ tb, if given, is the original traceback (so that it can be printed out).
995 If expected is set, this is a normal error message and most likely not a bug in yt-dlp.
997 if sys
.exc_info()[0] in network_exceptions
:
1000 self
.orig_msg
= str(msg
)
1002 self
.expected
= expected
1004 self
.video_id
= video_id
1006 self
.exc_info
= sys
.exc_info() # preserve original exception
1008 super().__init
__(''.join((
1009 format_field(ie
, template
='[%s] '),
1010 format_field(video_id
, template
='%s: '),
1012 format_field(cause
, template
=' (caused by %r)'),
1013 '' if expected
else bug_reports_message())))
1015 def format_traceback(self
):
1016 return join_nonempty(
1017 self
.traceback
and ''.join(traceback
.format_tb(self
.traceback
)),
1018 self
.cause
and ''.join(traceback
.format_exception(None, self
.cause
, self
.cause
.__traceback
__)[1:]),
1022 class UnsupportedError(ExtractorError
):
1023 def __init__(self
, url
):
1025 'Unsupported URL: %s' % url
, expected
=True)
1029 class RegexNotFoundError(ExtractorError
):
1030 """Error when a regex didn't match"""
1034 class GeoRestrictedError(ExtractorError
):
1035 """Geographic restriction Error exception.
1037 This exception may be thrown when a video is not available from your
1038 geographic location due to geographic restrictions imposed by a website.
1041 def __init__(self
, msg
, countries
=None, **kwargs
):
1042 kwargs
['expected'] = True
1043 super().__init
__(msg
, **kwargs
)
1044 self
.countries
= countries
1047 class DownloadError(YoutubeDLError
):
1048 """Download Error exception.
1050 This exception may be thrown by FileDownloader objects if they are not
1051 configured to continue on errors. They will contain the appropriate
1055 def __init__(self
, msg
, exc_info
=None):
1056 """ exc_info, if given, is the original exception that caused the trouble (as returned by sys.exc_info()). """
1057 super().__init
__(msg
)
1058 self
.exc_info
= exc_info
1061 class EntryNotInPlaylist(YoutubeDLError
):
1062 """Entry not in playlist exception.
1064 This exception will be thrown by YoutubeDL when a requested entry
1065 is not found in the playlist info_dict
1067 msg
= 'Entry not found in info'
1070 class SameFileError(YoutubeDLError
):
1071 """Same File exception.
1073 This exception will be thrown by FileDownloader objects if they detect
1074 multiple files would have to be downloaded to the same file on disk.
1076 msg
= 'Fixed output name but more than one file to download'
1078 def __init__(self
, filename
=None):
1079 if filename
is not None:
1080 self
.msg
+= f
': {filename}'
1081 super().__init
__(self
.msg
)
1084 class PostProcessingError(YoutubeDLError
):
1085 """Post Processing exception.
1087 This exception may be raised by PostProcessor's .run() method to
1088 indicate an error in the postprocessing task.
1092 class DownloadCancelled(YoutubeDLError
):
1093 """ Exception raised when the download queue should be interrupted """
1094 msg
= 'The download was cancelled'
1097 class ExistingVideoReached(DownloadCancelled
):
1098 """ --break-on-existing triggered """
1099 msg
= 'Encountered a video that is already in the archive, stopping due to --break-on-existing'
1102 class RejectedVideoReached(DownloadCancelled
):
1103 """ --break-on-reject triggered """
1104 msg
= 'Encountered a video that did not match filter, stopping due to --break-on-reject'
1107 class MaxDownloadsReached(DownloadCancelled
):
1108 """ --max-downloads limit has been reached. """
1109 msg
= 'Maximum number of downloads reached, stopping due to --max-downloads'
1112 class ReExtractInfo(YoutubeDLError
):
1113 """ Video info needs to be re-extracted. """
1115 def __init__(self
, msg
, expected
=False):
1116 super().__init
__(msg
)
1117 self
.expected
= expected
1120 class ThrottledDownload(ReExtractInfo
):
1121 """ Download speed below --throttled-rate. """
1122 msg
= 'The download speed is below throttle limit'
1125 super().__init
__(self
.msg
, expected
=False)
1128 class UnavailableVideoError(YoutubeDLError
):
1129 """Unavailable Format exception.
1131 This exception will be thrown when a video is requested
1132 in a format that is not available for that video.
1134 msg
= 'Unable to download video'
1136 def __init__(self
, err
=None):
1138 self
.msg
+= f
': {err}'
1139 super().__init
__(self
.msg
)
1142 class ContentTooShortError(YoutubeDLError
):
1143 """Content Too Short exception.
1145 This exception may be raised by FileDownloader objects when a file they
1146 download is too small for what the server announced first, indicating
1147 the connection was probably interrupted.
1150 def __init__(self
, downloaded
, expected
):
1151 super().__init
__(f
'Downloaded {downloaded} bytes, expected {expected} bytes')
1153 self
.downloaded
= downloaded
1154 self
.expected
= expected
1157 class XAttrMetadataError(YoutubeDLError
):
1158 def __init__(self
, code
=None, msg
='Unknown error'):
1159 super().__init
__(msg
)
1163 # Parsing code and msg
1164 if (self
.code
in (errno
.ENOSPC
, errno
.EDQUOT
)
1165 or 'No space left' in self
.msg
or 'Disk quota exceeded' in self
.msg
):
1166 self
.reason
= 'NO_SPACE'
1167 elif self
.code
== errno
.E2BIG
or 'Argument list too long' in self
.msg
:
1168 self
.reason
= 'VALUE_TOO_LONG'
1170 self
.reason
= 'NOT_SUPPORTED'
1173 class XAttrUnavailableError(YoutubeDLError
):
1177 def _create_http_connection(ydl_handler
, http_class
, is_https
, *args
, **kwargs
):
1178 hc
= http_class(*args
, **kwargs
)
1179 source_address
= ydl_handler
._params
.get('source_address')
1181 if source_address
is not None:
1182 # This is to workaround _create_connection() from socket where it will try all
1183 # address data from getaddrinfo() including IPv6. This filters the result from
1184 # getaddrinfo() based on the source_address value.
1185 # This is based on the cpython socket.create_connection() function.
1186 # https://github.com/python/cpython/blob/master/Lib/socket.py#L691
1187 def _create_connection(address
, timeout
=socket
._GLOBAL
_DEFAULT
_TIMEOUT
, source_address
=None):
1188 host
, port
= address
1190 addrs
= socket
.getaddrinfo(host
, port
, 0, socket
.SOCK_STREAM
)
1191 af
= socket
.AF_INET
if '.' in source_address
[0] else socket
.AF_INET6
1192 ip_addrs
= [addr
for addr
in addrs
if addr
[0] == af
]
1193 if addrs
and not ip_addrs
:
1194 ip_version
= 'v4' if af
== socket
.AF_INET
else 'v6'
1196 "No remote IP%s addresses available for connect, can't use '%s' as source address"
1197 % (ip_version
, source_address
[0]))
1198 for res
in ip_addrs
:
1199 af
, socktype
, proto
, canonname
, sa
= res
1202 sock
= socket
.socket(af
, socktype
, proto
)
1203 if timeout
is not socket
._GLOBAL
_DEFAULT
_TIMEOUT
:
1204 sock
.settimeout(timeout
)
1205 sock
.bind(source_address
)
1207 err
= None # Explicitly break reference cycle
1209 except OSError as _
:
1211 if sock
is not None:
1216 raise OSError('getaddrinfo returns an empty list')
1217 if hasattr(hc
, '_create_connection'):
1218 hc
._create
_connection
= _create_connection
1219 hc
.source_address
= (source_address
, 0)
1224 def handle_youtubedl_headers(headers
):
1225 filtered_headers
= headers
1227 if 'Youtubedl-no-compression' in filtered_headers
:
1228 filtered_headers
= {k: v for k, v in filtered_headers.items() if k.lower() != 'accept-encoding'}
1229 del filtered_headers
['Youtubedl-no-compression']
1231 return filtered_headers
1234 class YoutubeDLHandler(compat_urllib_request
.HTTPHandler
):
1235 """Handler for HTTP requests and responses.
1237 This class, when installed with an OpenerDirector, automatically adds
1238 the standard headers to every HTTP request and handles gzipped and
1239 deflated responses from web servers. If compression is to be avoided in
1240 a particular request, the original request in the program code only has
1241 to include the HTTP header "Youtubedl-no-compression", which will be
1242 removed before making the real request.
1244 Part of this code was copied from:
1246 http://techknack.net/python-urllib2-handlers/
1248 Andrew Rowls, the author of that code, agreed to release it to the
1252 def __init__(self
, params
, *args
, **kwargs
):
1253 compat_urllib_request
.HTTPHandler
.__init
__(self
, *args
, **kwargs
)
1254 self
._params
= params
1256 def http_open(self
, req
):
1257 conn_class
= compat_http_client
.HTTPConnection
1259 socks_proxy
= req
.headers
.get('Ytdl-socks-proxy')
1261 conn_class
= make_socks_conn_class(conn_class
, socks_proxy
)
1262 del req
.headers
['Ytdl-socks-proxy']
1264 return self
.do_open(functools
.partial(
1265 _create_http_connection
, self
, conn_class
, False),
1273 return zlib
.decompress(data
, -zlib
.MAX_WBITS
)
1275 return zlib
.decompress(data
)
1281 return compat_brotli
.decompress(data
)
1283 def http_request(self
, req
):
1284 # According to RFC 3986, URLs can not contain non-ASCII characters, however this is not
1285 # always respected by websites, some tend to give out URLs with non percent-encoded
1286 # non-ASCII characters (see telemb.py, ard.py [#3412])
1287 # urllib chokes on URLs with non-ASCII characters (see http://bugs.python.org/issue3991)
1288 # To work around aforementioned issue we will replace request's original URL with
1289 # percent-encoded one
1290 # Since redirects are also affected (e.g. http://www.southpark.de/alle-episoden/s18e09)
1291 # the code of this workaround has been moved here from YoutubeDL.urlopen()
1292 url
= req
.get_full_url()
1293 url_escaped
= escape_url(url
)
1295 # Substitute URL if any change after escaping
1296 if url
!= url_escaped
:
1297 req
= update_Request(req
, url
=url_escaped
)
1299 for h
, v
in self
._params
.get('http_headers', std_headers
).items():
1300 # Capitalize is needed because of Python bug 2275: http://bugs.python.org/issue2275
1301 # The dict keys are capitalized because of this bug by urllib
1302 if h
.capitalize() not in req
.headers
:
1303 req
.add_header(h
, v
)
1305 if 'Accept-encoding' not in req
.headers
:
1306 req
.add_header('Accept-encoding', ', '.join(SUPPORTED_ENCODINGS
))
1308 req
.headers
= handle_youtubedl_headers(req
.headers
)
1312 def http_response(self
, req
, resp
):
1315 if resp
.headers
.get('Content-encoding', '') == 'gzip':
1316 content
= resp
.read()
1317 gz
= gzip
.GzipFile(fileobj
=io
.BytesIO(content
), mode
='rb')
1319 uncompressed
= io
.BytesIO(gz
.read())
1320 except OSError as original_ioerror
:
1321 # There may be junk add the end of the file
1322 # See http://stackoverflow.com/q/4928560/35070 for details
1323 for i
in range(1, 1024):
1325 gz
= gzip
.GzipFile(fileobj
=io
.BytesIO(content
[:-i
]), mode
='rb')
1326 uncompressed
= io
.BytesIO(gz
.read())
1331 raise original_ioerror
1332 resp
= compat_urllib_request
.addinfourl(uncompressed
, old_resp
.headers
, old_resp
.url
, old_resp
.code
)
1333 resp
.msg
= old_resp
.msg
1334 del resp
.headers
['Content-encoding']
1336 if resp
.headers
.get('Content-encoding', '') == 'deflate':
1337 gz
= io
.BytesIO(self
.deflate(resp
.read()))
1338 resp
= compat_urllib_request
.addinfourl(gz
, old_resp
.headers
, old_resp
.url
, old_resp
.code
)
1339 resp
.msg
= old_resp
.msg
1340 del resp
.headers
['Content-encoding']
1342 if resp
.headers
.get('Content-encoding', '') == 'br':
1343 resp
= compat_urllib_request
.addinfourl(
1344 io
.BytesIO(self
.brotli(resp
.read())), old_resp
.headers
, old_resp
.url
, old_resp
.code
)
1345 resp
.msg
= old_resp
.msg
1346 del resp
.headers
['Content-encoding']
1347 # Percent-encode redirect URL of Location HTTP header to satisfy RFC 3986 (see
1348 # https://github.com/ytdl-org/youtube-dl/issues/6457).
1349 if 300 <= resp
.code
< 400:
1350 location
= resp
.headers
.get('Location')
1352 # As of RFC 2616 default charset is iso-8859-1 that is respected by python 3
1353 location
= location
.encode('iso-8859-1').decode('utf-8')
1354 location_escaped
= escape_url(location
)
1355 if location
!= location_escaped
:
1356 del resp
.headers
['Location']
1357 resp
.headers
['Location'] = location_escaped
1360 https_request
= http_request
1361 https_response
= http_response
1364 def make_socks_conn_class(base_class
, socks_proxy
):
1365 assert issubclass(base_class
, (
1366 compat_http_client
.HTTPConnection
, compat_http_client
.HTTPSConnection
))
1368 url_components
= compat_urlparse
.urlparse(socks_proxy
)
1369 if url_components
.scheme
.lower() == 'socks5':
1370 socks_type
= ProxyType
.SOCKS5
1371 elif url_components
.scheme
.lower() in ('socks', 'socks4'):
1372 socks_type
= ProxyType
.SOCKS4
1373 elif url_components
.scheme
.lower() == 'socks4a':
1374 socks_type
= ProxyType
.SOCKS4A
1376 def unquote_if_non_empty(s
):
1379 return compat_urllib_parse_unquote_plus(s
)
1383 url_components
.hostname
, url_components
.port
or 1080,
1385 unquote_if_non_empty(url_components
.username
),
1386 unquote_if_non_empty(url_components
.password
),
1389 class SocksConnection(base_class
):
1391 self
.sock
= sockssocket()
1392 self
.sock
.setproxy(*proxy_args
)
1393 if type(self
.timeout
) in (int, float):
1394 self
.sock
.settimeout(self
.timeout
)
1395 self
.sock
.connect((self
.host
, self
.port
))
1397 if isinstance(self
, compat_http_client
.HTTPSConnection
):
1398 if hasattr(self
, '_context'): # Python > 2.6
1399 self
.sock
= self
._context
.wrap_socket(
1400 self
.sock
, server_hostname
=self
.host
)
1402 self
.sock
= ssl
.wrap_socket(self
.sock
)
1404 return SocksConnection
1407 class YoutubeDLHTTPSHandler(compat_urllib_request
.HTTPSHandler
):
1408 def __init__(self
, params
, https_conn_class
=None, *args
, **kwargs
):
1409 compat_urllib_request
.HTTPSHandler
.__init
__(self
, *args
, **kwargs
)
1410 self
._https
_conn
_class
= https_conn_class
or compat_http_client
.HTTPSConnection
1411 self
._params
= params
1413 def https_open(self
, req
):
1415 conn_class
= self
._https
_conn
_class
1417 if hasattr(self
, '_context'): # python > 2.6
1418 kwargs
['context'] = self
._context
1419 if hasattr(self
, '_check_hostname'): # python 3.x
1420 kwargs
['check_hostname'] = self
._check
_hostname
1422 socks_proxy
= req
.headers
.get('Ytdl-socks-proxy')
1424 conn_class
= make_socks_conn_class(conn_class
, socks_proxy
)
1425 del req
.headers
['Ytdl-socks-proxy']
1427 return self
.do_open(functools
.partial(
1428 _create_http_connection
, self
, conn_class
, True),
1432 class YoutubeDLCookieJar(compat_cookiejar
.MozillaCookieJar
):
1434 See [1] for cookie file format.
1436 1. https://curl.haxx.se/docs/http-cookies.html
1438 _HTTPONLY_PREFIX
= '#HttpOnly_'
1440 _HEADER
= '''# Netscape HTTP Cookie File
1441 # This file is generated by yt-dlp. Do not edit.
1444 _CookieFileEntry
= collections
.namedtuple(
1446 ('domain_name', 'include_subdomains', 'path', 'https_only', 'expires_at', 'name', 'value'))
1448 def save(self
, filename
=None, ignore_discard
=False, ignore_expires
=False):
1450 Save cookies to a file.
1452 Most of the code is taken from CPython 3.8 and slightly adapted
1453 to support cookie files with UTF-8 in both python 2 and 3.
1455 if filename
is None:
1456 if self
.filename
is not None:
1457 filename
= self
.filename
1459 raise ValueError(compat_cookiejar
.MISSING_FILENAME_TEXT
)
1461 # Store session cookies with `expires` set to 0 instead of an empty
1464 if cookie
.expires
is None:
1467 with open(filename
, 'w', encoding
='utf-8') as f
:
1468 f
.write(self
._HEADER
)
1471 if not ignore_discard
and cookie
.discard
:
1473 if not ignore_expires
and cookie
.is_expired(now
):
1479 if cookie
.domain
.startswith('.'):
1480 initial_dot
= 'TRUE'
1482 initial_dot
= 'FALSE'
1483 if cookie
.expires
is not None:
1484 expires
= compat_str(cookie
.expires
)
1487 if cookie
.value
is None:
1488 # cookies.txt regards 'Set-Cookie: foo' as a cookie
1489 # with no name, whereas http.cookiejar regards it as a
1490 # cookie with no value.
1495 value
= cookie
.value
1497 '\t'.join([cookie
.domain
, initial_dot
, cookie
.path
,
1498 secure
, expires
, name
, value
]) + '\n')
1500 def load(self
, filename
=None, ignore_discard
=False, ignore_expires
=False):
1501 """Load cookies from a file."""
1502 if filename
is None:
1503 if self
.filename
is not None:
1504 filename
= self
.filename
1506 raise ValueError(compat_cookiejar
.MISSING_FILENAME_TEXT
)
1508 def prepare_line(line
):
1509 if line
.startswith(self
._HTTPONLY
_PREFIX
):
1510 line
= line
[len(self
._HTTPONLY
_PREFIX
):]
1511 # comments and empty lines are fine
1512 if line
.startswith('#') or not line
.strip():
1514 cookie_list
= line
.split('\t')
1515 if len(cookie_list
) != self
._ENTRY
_LEN
:
1516 raise compat_cookiejar
.LoadError('invalid length %d' % len(cookie_list
))
1517 cookie
= self
._CookieFileEntry
(*cookie_list
)
1518 if cookie
.expires_at
and not cookie
.expires_at
.isdigit():
1519 raise compat_cookiejar
.LoadError('invalid expires at %s' % cookie
.expires_at
)
1523 with open(filename
, encoding
='utf-8') as f
:
1526 cf
.write(prepare_line(line
))
1527 except compat_cookiejar
.LoadError
as e
:
1529 'WARNING: skipping cookie file entry due to %s: %r\n'
1530 % (e
, line
), sys
.stderr
)
1533 self
._really
_load
(cf
, filename
, ignore_discard
, ignore_expires
)
1534 # Session cookies are denoted by either `expires` field set to
1535 # an empty string or 0. MozillaCookieJar only recognizes the former
1536 # (see [1]). So we need force the latter to be recognized as session
1537 # cookies on our own.
1538 # Session cookies may be important for cookies-based authentication,
1539 # e.g. usually, when user does not check 'Remember me' check box while
1540 # logging in on a site, some important cookies are stored as session
1541 # cookies so that not recognizing them will result in failed login.
1542 # 1. https://bugs.python.org/issue17164
1544 # Treat `expires=0` cookies as session cookies
1545 if cookie
.expires
== 0:
1546 cookie
.expires
= None
1547 cookie
.discard
= True
1550 class YoutubeDLCookieProcessor(compat_urllib_request
.HTTPCookieProcessor
):
1551 def __init__(self
, cookiejar
=None):
1552 compat_urllib_request
.HTTPCookieProcessor
.__init
__(self
, cookiejar
)
1554 def http_response(self
, request
, response
):
1555 return compat_urllib_request
.HTTPCookieProcessor
.http_response(self
, request
, response
)
1557 https_request
= compat_urllib_request
.HTTPCookieProcessor
.http_request
1558 https_response
= http_response
1561 class YoutubeDLRedirectHandler(compat_urllib_request
.HTTPRedirectHandler
):
1562 """YoutubeDL redirect handler
1564 The code is based on HTTPRedirectHandler implementation from CPython [1].
1566 This redirect handler solves two issues:
1567 - ensures redirect URL is always unicode under python 2
1568 - introduces support for experimental HTTP response status code
1569 308 Permanent Redirect [2] used by some sites [3]
1571 1. https://github.com/python/cpython/blob/master/Lib/urllib/request.py
1572 2. https://developer.mozilla.org/en-US/docs/Web/HTTP/Status/308
1573 3. https://github.com/ytdl-org/youtube-dl/issues/28768
1576 http_error_301
= http_error_303
= http_error_307
= http_error_308
= compat_urllib_request
.HTTPRedirectHandler
.http_error_302
1578 def redirect_request(self
, req
, fp
, code
, msg
, headers
, newurl
):
1579 """Return a Request or None in response to a redirect.
1581 This is called by the http_error_30x methods when a
1582 redirection response is received. If a redirection should
1583 take place, return a new Request to allow http_error_30x to
1584 perform the redirect. Otherwise, raise HTTPError if no-one
1585 else should try to handle this url. Return None if you can't
1586 but another Handler might.
1588 m
= req
.get_method()
1589 if (not (code
in (301, 302, 303, 307, 308) and m
in ("GET", "HEAD")
1590 or code
in (301, 302, 303) and m
== "POST")):
1591 raise compat_HTTPError(req
.full_url
, code
, msg
, headers
, fp
)
1592 # Strictly (according to RFC 2616), 301 or 302 in response to
1593 # a POST MUST NOT cause a redirection without confirmation
1594 # from the user (of urllib.request, in this case). In practice,
1595 # essentially all clients do redirect in this case, so we do
1598 # Be conciliant with URIs containing a space. This is mainly
1599 # redundant with the more complete encoding done in http_error_302(),
1600 # but it is kept for compatibility with other callers.
1601 newurl
= newurl
.replace(' ', '%20')
1603 CONTENT_HEADERS
= ("content-length", "content-type")
1604 # NB: don't use dict comprehension for python 2.6 compatibility
1605 newheaders
= {k: v for k, v in req.headers.items() if k.lower() not in CONTENT_HEADERS}
1606 return compat_urllib_request
.Request(
1607 newurl
, headers
=newheaders
, origin_req_host
=req
.origin_req_host
,
1611 def extract_timezone(date_str
):
1614 ^.{8,}? # >=8 char non-TZ prefix, if present
1615 (?P<tz>Z| # just the UTC Z, or
1616 (?:(?<=.\b\d{4}|\b\d{2}:\d\d)| # preceded by 4 digits or hh:mm or
1617 (?<!.\b[a-zA-Z]{3}|[a-zA-Z]{4}|..\b\d\d)) # not preceded by 3 alpha word or >= 4 alpha or 2 digits
1618 [ ]? # optional space
1619 (?P<sign>\+|-) # +/-
1620 (?P<hours>[0-9]{2}):?(?P<minutes>[0-9]{2}) # hh[:]mm
1624 timezone
= datetime
.timedelta()
1626 date_str
= date_str
[:-len(m
.group('tz'))]
1627 if not m
.group('sign'):
1628 timezone
= datetime
.timedelta()
1630 sign
= 1 if m
.group('sign') == '+' else -1
1631 timezone
= datetime
.timedelta(
1632 hours
=sign
* int(m
.group('hours')),
1633 minutes
=sign
* int(m
.group('minutes')))
1634 return timezone
, date_str
1637 def parse_iso8601(date_str
, delimiter
='T', timezone
=None):
1638 """ Return a UNIX timestamp from the given date """
1640 if date_str
is None:
1643 date_str
= re
.sub(r
'\.[0-9]+', '', date_str
)
1645 if timezone
is None:
1646 timezone
, date_str
= extract_timezone(date_str
)
1649 date_format
= f
'%Y-%m-%d{delimiter}%H:%M:%S'
1650 dt
= datetime
.datetime
.strptime(date_str
, date_format
) - timezone
1651 return calendar
.timegm(dt
.timetuple())
1656 def date_formats(day_first
=True):
1657 return DATE_FORMATS_DAY_FIRST
if day_first
else DATE_FORMATS_MONTH_FIRST
1660 def unified_strdate(date_str
, day_first
=True):
1661 """Return a string with the date in the format YYYYMMDD"""
1663 if date_str
is None:
1667 date_str
= date_str
.replace(',', ' ')
1668 # Remove AM/PM + timezone
1669 date_str
= re
.sub(r
'(?i)\s*(?:AM|PM)(?:\s+[A-Z]+)?', '', date_str
)
1670 _
, date_str
= extract_timezone(date_str
)
1672 for expression
in date_formats(day_first
):
1674 upload_date
= datetime
.datetime
.strptime(date_str
, expression
).strftime('%Y%m%d')
1677 if upload_date
is None:
1678 timetuple
= email
.utils
.parsedate_tz(date_str
)
1681 upload_date
= datetime
.datetime(*timetuple
[:6]).strftime('%Y%m%d')
1684 if upload_date
is not None:
1685 return compat_str(upload_date
)
1688 def unified_timestamp(date_str
, day_first
=True):
1689 if date_str
is None:
1692 date_str
= re
.sub(r
'[,|]', '', date_str
)
1694 pm_delta
= 12 if re
.search(r
'(?i)PM', date_str
) else 0
1695 timezone
, date_str
= extract_timezone(date_str
)
1697 # Remove AM/PM + timezone
1698 date_str
= re
.sub(r
'(?i)\s*(?:AM|PM)(?:\s+[A-Z]+)?', '', date_str
)
1700 # Remove unrecognized timezones from ISO 8601 alike timestamps
1701 m
= re
.search(r
'\d{1,2}:\d{1,2}(?:\.\d+)?(?P<tz>\s*[A-Z]+)$', date_str
)
1703 date_str
= date_str
[:-len(m
.group('tz'))]
1705 # Python only supports microseconds, so remove nanoseconds
1706 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]+$', date_str
)
1708 date_str
= m
.group(1)
1710 for expression
in date_formats(day_first
):
1712 dt
= datetime
.datetime
.strptime(date_str
, expression
) - timezone
+ datetime
.timedelta(hours
=pm_delta
)
1713 return calendar
.timegm(dt
.timetuple())
1716 timetuple
= email
.utils
.parsedate_tz(date_str
)
1718 return calendar
.timegm(timetuple
) + pm_delta
* 3600
1721 def determine_ext(url
, default_ext
='unknown_video'):
1722 if url
is None or '.' not in url
:
1724 guess
= url
.partition('?')[0].rpartition('.')[2]
1725 if re
.match(r
'^[A-Za-z0-9]+$', guess
):
1727 # Try extract ext from URLs like http://example.com/foo/bar.mp4/?download
1728 elif guess
.rstrip('/') in KNOWN_EXTENSIONS
:
1729 return guess
.rstrip('/')
1734 def subtitles_filename(filename
, sub_lang
, sub_format
, expected_real_ext
=None):
1735 return replace_extension(filename
, sub_lang
+ '.' + sub_format
, expected_real_ext
)
1738 def datetime_from_str(date_str
, precision
='auto', format
='%Y%m%d'):
1740 Return a datetime object from a string in the format YYYYMMDD or
1741 (now|today|yesterday|date)[+-][0-9](microsecond|second|minute|hour|day|week|month|year)(s)?
1743 format: string date format used to return datetime object from
1744 precision: round the time portion of a datetime object.
1745 auto|microsecond|second|minute|hour|day.
1746 auto: round to the unit provided in date_str (if applicable).
1748 auto_precision
= False
1749 if precision
== 'auto':
1750 auto_precision
= True
1751 precision
= 'microsecond'
1752 today
= datetime_round(datetime
.datetime
.utcnow(), precision
)
1753 if date_str
in ('now', 'today'):
1755 if date_str
== 'yesterday':
1756 return today
- datetime
.timedelta(days
=1)
1758 r
'(?P<start>.+)(?P<sign>[+-])(?P<time>\d+)(?P<unit>microsecond|second|minute|hour|day|week|month|year)(s)?',
1760 if match
is not None:
1761 start_time
= datetime_from_str(match
.group('start'), precision
, format
)
1762 time
= int(match
.group('time')) * (-1 if match
.group('sign') == '-' else 1)
1763 unit
= match
.group('unit')
1764 if unit
== 'month' or unit
== 'year':
1765 new_date
= datetime_add_months(start_time
, time
* 12 if unit
== 'year' else time
)
1771 delta
= datetime
.timedelta(**{unit + 's': time}
)
1772 new_date
= start_time
+ delta
1774 return datetime_round(new_date
, unit
)
1777 return datetime_round(datetime
.datetime
.strptime(date_str
, format
), precision
)
1780 def date_from_str(date_str
, format
='%Y%m%d', strict
=False):
1782 Return a datetime object from a string in the format YYYYMMDD or
1783 (now|today|yesterday|date)[+-][0-9](microsecond|second|minute|hour|day|week|month|year)(s)?
1785 If "strict", only (now|today)[+-][0-9](day|week|month|year)(s)? is allowed
1787 format: string date format used to return datetime object from
1789 if strict
and not re
.fullmatch(r
'\d{8}|(now|today)[+-]\d+(day|week|month|year)(s)?', date_str
):
1790 raise ValueError(f
'Invalid date format {date_str}')
1791 return datetime_from_str(date_str
, precision
='microsecond', format
=format
).date()
1794 def datetime_add_months(dt
, months
):
1795 """Increment/Decrement a datetime object by months."""
1796 month
= dt
.month
+ months
- 1
1797 year
= dt
.year
+ month
// 12
1798 month
= month
% 12 + 1
1799 day
= min(dt
.day
, calendar
.monthrange(year
, month
)[1])
1800 return dt
.replace(year
, month
, day
)
1803 def datetime_round(dt
, precision
='day'):
1805 Round a datetime object's time to a specific precision
1807 if precision
== 'microsecond':
1816 roundto
= lambda x
, n
: ((x
+ n
/ 2) // n
) * n
1817 timestamp
= calendar
.timegm(dt
.timetuple())
1818 return datetime
.datetime
.utcfromtimestamp(roundto(timestamp
, unit_seconds
[precision
]))
1821 def hyphenate_date(date_str
):
1823 Convert a date in 'YYYYMMDD' format to 'YYYY-MM-DD' format"""
1824 match
= re
.match(r
'^(\d\d\d\d)(\d\d)(\d\d)$', date_str
)
1825 if match
is not None:
1826 return '-'.join(match
.groups())
1832 """Represents a time interval between two dates"""
1834 def __init__(self
, start
=None, end
=None):
1835 """start and end must be strings in the format accepted by date"""
1836 if start
is not None:
1837 self
.start
= date_from_str(start
, strict
=True)
1839 self
.start
= datetime
.datetime
.min.date()
1841 self
.end
= date_from_str(end
, strict
=True)
1843 self
.end
= datetime
.datetime
.max.date()
1844 if self
.start
> self
.end
:
1845 raise ValueError('Date range: "%s" , the start date must be before the end date' % self
)
1849 """Returns a range that only contains the given day"""
1850 return cls(day
, day
)
1852 def __contains__(self
, date
):
1853 """Check if the date is in the range"""
1854 if not isinstance(date
, datetime
.date
):
1855 date
= date_from_str(date
)
1856 return self
.start
<= date
<= self
.end
1859 return f
'{self.start.isoformat()} - {self.end.isoformat()}'
1862 def platform_name():
1863 """ Returns the platform name as a compat_str """
1864 res
= platform
.platform()
1865 if isinstance(res
, bytes):
1866 res
= res
.decode(preferredencoding())
1868 assert isinstance(res
, compat_str
)
1872 def get_windows_version():
1873 ''' Get Windows version. None if it's not running on Windows '''
1874 if compat_os_name
== 'nt':
1875 return version_tuple(platform
.win32_ver()[1])
1880 def write_string(s
, out
=None, encoding
=None):
1883 assert type(s
) == compat_str
1885 if 'b' in getattr(out
, 'mode', ''):
1886 byt
= s
.encode(encoding
or preferredencoding(), 'ignore')
1888 elif hasattr(out
, 'buffer'):
1889 enc
= encoding
or getattr(out
, 'encoding', None) or preferredencoding()
1890 byt
= s
.encode(enc
, 'ignore')
1891 out
.buffer.write(byt
)
1897 def bytes_to_intlist(bs
):
1900 if isinstance(bs
[0], int): # Python 3
1903 return [ord(c
) for c
in bs
]
1906 def intlist_to_bytes(xs
):
1909 return compat_struct_pack('%dB' % len(xs
), *xs
)
1912 class LockingUnsupportedError(IOError):
1913 msg
= 'File locking is not supported on this platform'
1916 super().__init
__(self
.msg
)
1919 # Cross-platform file locking
1920 if sys
.platform
== 'win32':
1921 import ctypes
.wintypes
1924 class OVERLAPPED(ctypes
.Structure
):
1926 ('Internal', ctypes
.wintypes
.LPVOID
),
1927 ('InternalHigh', ctypes
.wintypes
.LPVOID
),
1928 ('Offset', ctypes
.wintypes
.DWORD
),
1929 ('OffsetHigh', ctypes
.wintypes
.DWORD
),
1930 ('hEvent', ctypes
.wintypes
.HANDLE
),
1933 kernel32
= ctypes
.windll
.kernel32
1934 LockFileEx
= kernel32
.LockFileEx
1935 LockFileEx
.argtypes
= [
1936 ctypes
.wintypes
.HANDLE
, # hFile
1937 ctypes
.wintypes
.DWORD
, # dwFlags
1938 ctypes
.wintypes
.DWORD
, # dwReserved
1939 ctypes
.wintypes
.DWORD
, # nNumberOfBytesToLockLow
1940 ctypes
.wintypes
.DWORD
, # nNumberOfBytesToLockHigh
1941 ctypes
.POINTER(OVERLAPPED
) # Overlapped
1943 LockFileEx
.restype
= ctypes
.wintypes
.BOOL
1944 UnlockFileEx
= kernel32
.UnlockFileEx
1945 UnlockFileEx
.argtypes
= [
1946 ctypes
.wintypes
.HANDLE
, # hFile
1947 ctypes
.wintypes
.DWORD
, # dwReserved
1948 ctypes
.wintypes
.DWORD
, # nNumberOfBytesToLockLow
1949 ctypes
.wintypes
.DWORD
, # nNumberOfBytesToLockHigh
1950 ctypes
.POINTER(OVERLAPPED
) # Overlapped
1952 UnlockFileEx
.restype
= ctypes
.wintypes
.BOOL
1953 whole_low
= 0xffffffff
1954 whole_high
= 0x7fffffff
1956 def _lock_file(f
, exclusive
, block
):
1957 overlapped
= OVERLAPPED()
1958 overlapped
.Offset
= 0
1959 overlapped
.OffsetHigh
= 0
1960 overlapped
.hEvent
= 0
1961 f
._lock
_file
_overlapped
_p
= ctypes
.pointer(overlapped
)
1963 if not LockFileEx(msvcrt
.get_osfhandle(f
.fileno()),
1964 (0x2 if exclusive
else 0x0) |
(0x0 if block
else 0x1),
1965 0, whole_low
, whole_high
, f
._lock
_file
_overlapped
_p
):
1966 raise BlockingIOError('Locking file failed: %r' % ctypes
.FormatError())
1968 def _unlock_file(f
):
1969 assert f
._lock
_file
_overlapped
_p
1970 handle
= msvcrt
.get_osfhandle(f
.fileno())
1971 if not UnlockFileEx(handle
, 0, whole_low
, whole_high
, f
._lock
_file
_overlapped
_p
):
1972 raise OSError('Unlocking file failed: %r' % ctypes
.FormatError())
1978 def _lock_file(f
, exclusive
, block
):
1979 flags
= fcntl
.LOCK_EX
if exclusive
else fcntl
.LOCK_SH
1981 flags |
= fcntl
.LOCK_NB
1983 fcntl
.flock(f
, flags
)
1984 except BlockingIOError
:
1986 except OSError: # AOSP does not have flock()
1987 fcntl
.lockf(f
, flags
)
1989 def _unlock_file(f
):
1991 fcntl
.flock(f
, fcntl
.LOCK_UN
)
1993 fcntl
.lockf(f
, fcntl
.LOCK_UN
)
1997 def _lock_file(f
, exclusive
, block
):
1998 raise LockingUnsupportedError()
2000 def _unlock_file(f
):
2001 raise LockingUnsupportedError()
2007 def __init__(self
, filename
, mode
, block
=True, encoding
=None):
2008 if mode
not in {'r', 'rb', 'a', 'ab', 'w', 'wb'}
:
2009 raise NotImplementedError(mode
)
2010 self
.mode
, self
.block
= mode
, block
2012 writable
= any(f
in mode
for f
in 'wax+')
2013 readable
= any(f
in mode
for f
in 'r+')
2014 flags
= functools
.reduce(operator
.ior
, (
2015 getattr(os
, 'O_CLOEXEC', 0), # UNIX only
2016 getattr(os
, 'O_BINARY', 0), # Windows only
2017 getattr(os
, 'O_NOINHERIT', 0), # Windows only
2018 os
.O_CREAT
if writable
else 0, # O_TRUNC only after locking
2019 os
.O_APPEND
if 'a' in mode
else 0,
2020 os
.O_EXCL
if 'x' in mode
else 0,
2021 os
.O_RDONLY
if not writable
else os
.O_RDWR
if readable
else os
.O_WRONLY
,
2024 self
.f
= os
.fdopen(os
.open(filename
, flags
, 0o666), mode
, encoding
=encoding
)
2026 def __enter__(self
):
2027 exclusive
= 'r' not in self
.mode
2029 _lock_file(self
.f
, exclusive
, self
.block
)
2034 if 'w' in self
.mode
:
2042 _unlock_file(self
.f
)
2046 def __exit__(self
, *_
):
2055 def __getattr__(self
, attr
):
2056 return getattr(self
.f
, attr
)
2062 def get_filesystem_encoding():
2063 encoding
= sys
.getfilesystemencoding()
2064 return encoding
if encoding
is not None else 'utf-8'
2067 def shell_quote(args
):
2069 encoding
= get_filesystem_encoding()
2071 if isinstance(a
, bytes):
2072 # We may get a filename encoded with 'encodeFilename'
2073 a
= a
.decode(encoding
)
2074 quoted_args
.append(compat_shlex_quote(a
))
2075 return ' '.join(quoted_args
)
2078 def smuggle_url(url
, data
):
2079 """ Pass additional data in a URL for internal use. """
2081 url
, idata
= unsmuggle_url(url
, {})
2083 sdata
= compat_urllib_parse_urlencode(
2084 {'__youtubedl_smuggle': json.dumps(data)}
)
2085 return url
+ '#' + sdata
2088 def unsmuggle_url(smug_url
, default
=None):
2089 if '#__youtubedl_smuggle' not in smug_url
:
2090 return smug_url
, default
2091 url
, _
, sdata
= smug_url
.rpartition('#')
2092 jsond
= compat_parse_qs(sdata
)['__youtubedl_smuggle'][0]
2093 data
= json
.loads(jsond
)
2097 def format_decimal_suffix(num
, fmt
='%d%s', *, factor
=1000):
2098 """ Formats numbers with decimal sufixes like K, M, etc """
2099 num
, factor
= float_or_none(num
), float(factor
)
2100 if num
is None or num
< 0:
2102 POSSIBLE_SUFFIXES
= 'kMGTPEZY'
2103 exponent
= 0 if num
== 0 else min(int(math
.log(num
, factor
)), len(POSSIBLE_SUFFIXES
))
2104 suffix
= ['', *POSSIBLE_SUFFIXES
][exponent
]
2106 suffix
= {'k': 'Ki', '': ''}
.get(suffix
, f
'{suffix}i')
2107 converted
= num
/ (factor
** exponent
)
2108 return fmt
% (converted
, suffix
)
2111 def format_bytes(bytes):
2112 return format_decimal_suffix(bytes, '%.2f%sB', factor
=1024) or 'N/A'
2115 def lookup_unit_table(unit_table
, s
):
2116 units_re
= '|'.join(re
.escape(u
) for u
in unit_table
)
2118 r
'(?P<num>[0-9]+(?:[,.][0-9]*)?)\s*(?P<unit>%s)\b' % units_re
, s
)
2121 num_str
= m
.group('num').replace(',', '.')
2122 mult
= unit_table
[m
.group('unit')]
2123 return int(float(num_str
) * mult
)
2126 def parse_filesize(s
):
2130 # The lower-case forms are of course incorrect and unofficial,
2131 # but we support those too
2148 'megabytes': 1000 ** 2,
2149 'mebibytes': 1024 ** 2,
2155 'gigabytes': 1000 ** 3,
2156 'gibibytes': 1024 ** 3,
2162 'terabytes': 1000 ** 4,
2163 'tebibytes': 1024 ** 4,
2169 'petabytes': 1000 ** 5,
2170 'pebibytes': 1024 ** 5,
2176 'exabytes': 1000 ** 6,
2177 'exbibytes': 1024 ** 6,
2183 'zettabytes': 1000 ** 7,
2184 'zebibytes': 1024 ** 7,
2190 'yottabytes': 1000 ** 8,
2191 'yobibytes': 1024 ** 8,
2194 return lookup_unit_table(_UNIT_TABLE
, s
)
2201 s
= re
.sub(r
'^[^\d]+\s', '', s
).strip()
2203 if re
.match(r
'^[\d,.]+$', s
):
2204 return str_to_int(s
)
2217 ret
= lookup_unit_table(_UNIT_TABLE
, s
)
2221 mobj
= re
.match(r
'([\d,.]+)(?:$|\s)', s
)
2223 return str_to_int(mobj
.group(1))
2226 def parse_resolution(s
, *, lenient
=False):
2231 mobj
= re
.search(r
'(?P<w>\d+)\s*[xX×,]\s*(?P<h>\d+)', s
)
2233 mobj
= re
.search(r
'(?<![a-zA-Z0-9])(?P<w>\d+)\s*[xX×,]\s*(?P<h>\d+)(?![a-zA-Z0-9])', s
)
2236 'width': int(mobj
.group('w')),
2237 'height': int(mobj
.group('h')),
2240 mobj
= re
.search(r
'(?<![a-zA-Z0-9])(\d+)[pPiI](?![a-zA-Z0-9])', s
)
2242 return {'height': int(mobj.group(1))}
2244 mobj
= re
.search(r
'\b([48])[kK]\b', s
)
2246 return {'height': int(mobj.group(1)) * 540}
2251 def parse_bitrate(s
):
2252 if not isinstance(s
, compat_str
):
2254 mobj
= re
.search(r
'\b(\d+)\s*kbps', s
)
2256 return int(mobj
.group(1))
2259 def month_by_name(name
, lang
='en'):
2260 """ Return the number of a month by (locale-independently) English name """
2262 month_names
= MONTH_NAMES
.get(lang
, MONTH_NAMES
['en'])
2265 return month_names
.index(name
) + 1
2270 def month_by_abbreviation(abbrev
):
2271 """ Return the number of a month by (locale-independently) English
2275 return [s
[:3] for s
in ENGLISH_MONTH_NAMES
].index(abbrev
) + 1
2280 def fix_xml_ampersands(xml_str
):
2281 """Replace all the '&' by '&' in XML"""
2283 r
'&(?!amp;|lt;|gt;|apos;|quot;|#x[0-9a-fA-F]{,4};|#[0-9]{,4};)',
2288 def setproctitle(title
):
2289 assert isinstance(title
, compat_str
)
2291 # ctypes in Jython is not complete
2292 # http://bugs.jython.org/issue2148
2293 if sys
.platform
.startswith('java'):
2297 libc
= ctypes
.cdll
.LoadLibrary('libc.so.6')
2301 # LoadLibrary in Windows Python 2.7.13 only expects
2302 # a bytestring, but since unicode_literals turns
2303 # every string into a unicode string, it fails.
2305 title_bytes
= title
.encode('utf-8')
2306 buf
= ctypes
.create_string_buffer(len(title_bytes
))
2307 buf
.value
= title_bytes
2309 libc
.prctl(15, buf
, 0, 0, 0)
2310 except AttributeError:
2311 return # Strange libc, just skip this
2314 def remove_start(s
, start
):
2315 return s
[len(start
):] if s
is not None and s
.startswith(start
) else s
2318 def remove_end(s
, end
):
2319 return s
[:-len(end
)] if s
is not None and s
.endswith(end
) else s
2322 def remove_quotes(s
):
2323 if s
is None or len(s
) < 2:
2325 for quote
in ('"', "'", ):
2326 if s
[0] == quote
and s
[-1] == quote
:
2331 def get_domain(url
):
2332 domain
= re
.match(r
'(?:https?:\/\/)?(?:www\.)?(?P<domain>[^\n\/]+\.[^\n\/]+)(?:\/(.*))?', url
)
2333 return domain
.group('domain') if domain
else None
2336 def url_basename(url
):
2337 path
= compat_urlparse
.urlparse(url
).path
2338 return path
.strip('/').split('/')[-1]
2342 return re
.match(r
'https?://[^?#&]+/', url
).group()
2345 def urljoin(base
, path
):
2346 if isinstance(path
, bytes):
2347 path
= path
.decode('utf-8')
2348 if not isinstance(path
, compat_str
) or not path
:
2350 if re
.match(r
'^(?:[a-zA-Z][a-zA-Z0-9+-.]*:)?//', path
):
2352 if isinstance(base
, bytes):
2353 base
= base
.decode('utf-8')
2354 if not isinstance(base
, compat_str
) or not re
.match(
2355 r
'^(?:https?:)?//', base
):
2357 return compat_urlparse
.urljoin(base
, path
)
2360 class HEADRequest(compat_urllib_request
.Request
):
2361 def get_method(self
):
2365 class PUTRequest(compat_urllib_request
.Request
):
2366 def get_method(self
):
2370 def int_or_none(v
, scale
=1, default
=None, get_attr
=None, invscale
=1):
2371 if get_attr
and v
is not None:
2372 v
= getattr(v
, get_attr
, None)
2374 return int(v
) * invscale
// scale
2375 except (ValueError, TypeError, OverflowError):
2379 def str_or_none(v
, default
=None):
2380 return default
if v
is None else compat_str(v
)
2383 def str_to_int(int_str
):
2384 """ A more relaxed version of int_or_none """
2385 if isinstance(int_str
, int):
2387 elif isinstance(int_str
, compat_str
):
2388 int_str
= re
.sub(r
'[,\.\+]', '', int_str
)
2389 return int_or_none(int_str
)
2392 def float_or_none(v
, scale
=1, invscale
=1, default
=None):
2396 return float(v
) * invscale
/ scale
2397 except (ValueError, TypeError):
2401 def bool_or_none(v
, default
=None):
2402 return v
if isinstance(v
, bool) else default
2405 def strip_or_none(v
, default
=None):
2406 return v
.strip() if isinstance(v
, compat_str
) else default
2409 def url_or_none(url
):
2410 if not url
or not isinstance(url
, compat_str
):
2413 return url
if re
.match(r
'^(?:(?:https?|rt(?:m(?:pt?[es]?|fp)|sp[su]?)|mms|ftps?):)?//', url
) else None
2416 def request_to_url(req
):
2417 if isinstance(req
, compat_urllib_request
.Request
):
2418 return req
.get_full_url()
2423 def strftime_or_none(timestamp
, date_format
, default
=None):
2424 datetime_object
= None
2426 if isinstance(timestamp
, (int, float)): # unix timestamp
2427 datetime_object
= datetime
.datetime
.utcfromtimestamp(timestamp
)
2428 elif isinstance(timestamp
, compat_str
): # assume YYYYMMDD
2429 datetime_object
= datetime
.datetime
.strptime(timestamp
, '%Y%m%d')
2430 return datetime_object
.strftime(date_format
)
2431 except (ValueError, TypeError, AttributeError):
2435 def parse_duration(s
):
2436 if not isinstance(s
, str):
2442 days
, hours
, mins
, secs
, ms
= [None] * 5
2443 m
= re
.match(r
'''(?x)
2445 (?:(?:(?P<days>[0-9]+):)?(?P<hours>[0-9]+):)?(?P<mins>[0-9]+):)?
2446 (?P<secs>(?(before_secs)[0-9]{1,2}|[0-9]+))
2447 (?P<ms>[.:][0-9]+)?Z?$
2450 days
, hours
, mins
, secs
, ms
= m
.group('days', 'hours', 'mins', 'secs', 'ms')
2455 [0-9]+\s*y(?:ears?)?,?\s*
2458 [0-9]+\s*m(?:onths?)?,?\s*
2461 [0-9]+\s*w(?:eeks?)?,?\s*
2464 (?P<days>[0-9]+)\s*d(?:ays?)?,?\s*
2468 (?P<hours>[0-9]+)\s*h(?:ours?)?,?\s*
2471 (?P<mins>[0-9]+)\s*m(?:in(?:ute)?s?)?,?\s*
2474 (?P<secs>[0-9]+)(?P<ms>\.[0-9]+)?\s*s(?:ec(?:ond)?s?)?\s*
2477 days
, hours
, mins
, secs
, ms
= m
.groups()
2479 m
= re
.match(r
'(?i)(?:(?P<hours>[0-9.]+)\s*(?:hours?)|(?P<mins>[0-9.]+)\s*(?:mins?\.?|minutes?)\s*)Z?$', s
)
2481 hours
, mins
= m
.groups()
2487 duration
+= float(secs
)
2489 duration
+= float(mins
) * 60
2491 duration
+= float(hours
) * 60 * 60
2493 duration
+= float(days
) * 24 * 60 * 60
2495 duration
+= float(ms
.replace(':', '.'))
2499 def prepend_extension(filename
, ext
, expected_real_ext
=None):
2500 name
, real_ext
= os
.path
.splitext(filename
)
2502 f
'{name}.{ext}{real_ext}'
2503 if not expected_real_ext
or real_ext
[1:] == expected_real_ext
2504 else f
'{filename}.{ext}')
2507 def replace_extension(filename
, ext
, expected_real_ext
=None):
2508 name
, real_ext
= os
.path
.splitext(filename
)
2509 return '{}.{}'.format(
2510 name
if not expected_real_ext
or real_ext
[1:] == expected_real_ext
else filename
,
2514 def check_executable(exe
, args
=[]):
2515 """ Checks if the given binary is installed somewhere in PATH, and returns its name.
2516 args can be a list of arguments for a short output (like -version) """
2518 Popen([exe
] + args
, stdout
=subprocess
.PIPE
, stderr
=subprocess
.PIPE
).communicate_or_kill()
2524 def _get_exe_version_output(exe
, args
, *, to_screen
=None):
2526 to_screen(f
'Checking exe version: {shell_quote([exe] + args)}')
2528 # STDIN should be redirected too. On UNIX-like systems, ffmpeg triggers
2529 # SIGTTOU if yt-dlp is run in the background.
2530 # See https://github.com/ytdl-org/youtube-dl/issues/955#issuecomment-209789656
2532 [encodeArgument(exe
)] + args
, stdin
=subprocess
.PIPE
,
2533 stdout
=subprocess
.PIPE
, stderr
=subprocess
.STDOUT
).communicate_or_kill()
2536 if isinstance(out
, bytes): # Python 2.x
2537 out
= out
.decode('ascii', 'ignore')
2541 def detect_exe_version(output
, version_re
=None, unrecognized
='present'):
2542 assert isinstance(output
, compat_str
)
2543 if version_re
is None:
2544 version_re
= r
'version\s+([-0-9._a-zA-Z]+)'
2545 m
= re
.search(version_re
, output
)
2552 def get_exe_version(exe
, args
=['--version'],
2553 version_re
=None, unrecognized
='present'):
2554 """ Returns the version of the specified executable,
2555 or False if the executable is not present """
2556 out
= _get_exe_version_output(exe
, args
)
2557 return detect_exe_version(out
, version_re
, unrecognized
) if out
else False
2560 class LazyList(collections
.abc
.Sequence
):
2561 ''' Lazy immutable list from an iterable
2562 Note that slices of a LazyList are lists and not LazyList'''
2564 class IndexError(IndexError):
2567 def __init__(self
, iterable
, *, reverse
=False, _cache
=None):
2568 self
.__iterable
= iter(iterable
)
2569 self
.__cache
= [] if _cache
is None else _cache
2570 self
.__reversed
= reverse
2574 # We need to consume the entire iterable to iterate in reverse
2575 yield from self
.exhaust()
2577 yield from self
.__cache
2578 for item
in self
.__iterable
:
2579 self
.__cache
.append(item
)
2582 def __exhaust(self
):
2583 self
.__cache
.extend(self
.__iterable
)
2584 # Discard the emptied iterable to make it pickle-able
2585 self
.__iterable
= []
2589 ''' Evaluate the entire iterable '''
2590 return self
.__exhaust
()[::-1 if self
.__reversed
else 1]
2593 def __reverse_index(x
):
2594 return None if x
is None else -(x
+ 1)
2596 def __getitem__(self
, idx
):
2597 if isinstance(idx
, slice):
2599 idx
= slice(self
.__reverse
_index
(idx
.start
), self
.__reverse
_index
(idx
.stop
), -(idx
.step
or 1))
2600 start
, stop
, step
= idx
.start
, idx
.stop
, idx
.step
or 1
2601 elif isinstance(idx
, int):
2603 idx
= self
.__reverse
_index
(idx
)
2604 start
, stop
, step
= idx
, idx
, 0
2606 raise TypeError('indices must be integers or slices')
2607 if ((start
or 0) < 0 or (stop
or 0) < 0
2608 or (start
is None and step
< 0)
2609 or (stop
is None and step
> 0)):
2610 # We need to consume the entire iterable to be able to slice from the end
2611 # Obviously, never use this with infinite iterables
2614 return self
.__cache
[idx
]
2615 except IndexError as e
:
2616 raise self
.IndexError(e
) from e
2617 n
= max(start
or 0, stop
or 0) - len(self
.__cache
) + 1
2619 self
.__cache
.extend(itertools
.islice(self
.__iterable
, n
))
2621 return self
.__cache
[idx
]
2622 except IndexError as e
:
2623 raise self
.IndexError(e
) from e
2627 self
[-1] if self
.__reversed
else self
[0]
2628 except self
.IndexError:
2634 return len(self
.__cache
)
2636 def __reversed__(self
):
2637 return type(self
)(self
.__iterable
, reverse
=not self
.__reversed
, _cache
=self
.__cache
)
2640 return type(self
)(self
.__iterable
, reverse
=self
.__reversed
, _cache
=self
.__cache
)
2643 # repr and str should mimic a list. So we exhaust the iterable
2644 return repr(self
.exhaust())
2647 return repr(self
.exhaust())
2652 class IndexError(IndexError):
2656 # This is only useful for tests
2657 return len(self
.getslice())
2659 def __init__(self
, pagefunc
, pagesize
, use_cache
=True):
2660 self
._pagefunc
= pagefunc
2661 self
._pagesize
= pagesize
2662 self
._pagecount
= float('inf')
2663 self
._use
_cache
= use_cache
2666 def getpage(self
, pagenum
):
2667 page_results
= self
._cache
.get(pagenum
)
2668 if page_results
is None:
2669 page_results
= [] if pagenum
> self
._pagecount
else list(self
._pagefunc
(pagenum
))
2671 self
._cache
[pagenum
] = page_results
2674 def getslice(self
, start
=0, end
=None):
2675 return list(self
._getslice
(start
, end
))
2677 def _getslice(self
, start
, end
):
2678 raise NotImplementedError('This method must be implemented by subclasses')
2680 def __getitem__(self
, idx
):
2681 assert self
._use
_cache
, 'Indexing PagedList requires cache'
2682 if not isinstance(idx
, int) or idx
< 0:
2683 raise TypeError('indices must be non-negative integers')
2684 entries
= self
.getslice(idx
, idx
+ 1)
2686 raise self
.IndexError()
2690 class OnDemandPagedList(PagedList
):
2691 """Download pages until a page with less than maximum results"""
2693 def _getslice(self
, start
, end
):
2694 for pagenum
in itertools
.count(start
// self
._pagesize
):
2695 firstid
= pagenum
* self
._pagesize
2696 nextfirstid
= pagenum
* self
._pagesize
+ self
._pagesize
2697 if start
>= nextfirstid
:
2701 start
% self
._pagesize
2702 if firstid
<= start
< nextfirstid
2705 ((end
- 1) % self
._pagesize
) + 1
2706 if (end
is not None and firstid
<= end
<= nextfirstid
)
2710 page_results
= self
.getpage(pagenum
)
2712 self
._pagecount
= pagenum
- 1
2714 if startv
!= 0 or endv
is not None:
2715 page_results
= page_results
[startv
:endv
]
2716 yield from page_results
2718 # A little optimization - if current page is not "full", ie. does
2719 # not contain page_size videos then we can assume that this page
2720 # is the last one - there are no more ids on further pages -
2721 # i.e. no need to query again.
2722 if len(page_results
) + startv
< self
._pagesize
:
2725 # If we got the whole page, but the next page is not interesting,
2726 # break out early as well
2727 if end
== nextfirstid
:
2731 class InAdvancePagedList(PagedList
):
2732 """PagedList with total number of pages known in advance"""
2734 def __init__(self
, pagefunc
, pagecount
, pagesize
):
2735 PagedList
.__init
__(self
, pagefunc
, pagesize
, True)
2736 self
._pagecount
= pagecount
2738 def _getslice(self
, start
, end
):
2739 start_page
= start
// self
._pagesize
2740 end_page
= self
._pagecount
if end
is None else min(self
._pagecount
, end
// self
._pagesize
+ 1)
2741 skip_elems
= start
- start_page
* self
._pagesize
2742 only_more
= None if end
is None else end
- start
2743 for pagenum
in range(start_page
, end_page
):
2744 page_results
= self
.getpage(pagenum
)
2746 page_results
= page_results
[skip_elems
:]
2748 if only_more
is not None:
2749 if len(page_results
) < only_more
:
2750 only_more
-= len(page_results
)
2752 yield from page_results
[:only_more
]
2754 yield from page_results
2757 def uppercase_escape(s
):
2758 unicode_escape
= codecs
.getdecoder('unicode_escape')
2760 r
'\\U[0-9a-fA-F]{8}',
2761 lambda m
: unicode_escape(m
.group(0))[0],
2765 def lowercase_escape(s
):
2766 unicode_escape
= codecs
.getdecoder('unicode_escape')
2768 r
'\\u[0-9a-fA-F]{4}',
2769 lambda m
: unicode_escape(m
.group(0))[0],
2773 def escape_rfc3986(s
):
2774 """Escape non-ASCII characters as suggested by RFC 3986"""
2775 return urllib
.parse
.quote(s
, b
"%/;:@&=+$,!~*'()?#[]")
2778 def escape_url(url
):
2779 """Escape URL as suggested by RFC 3986"""
2780 url_parsed
= compat_urllib_parse_urlparse(url
)
2781 return url_parsed
._replace
(
2782 netloc
=url_parsed
.netloc
.encode('idna').decode('ascii'),
2783 path
=escape_rfc3986(url_parsed
.path
),
2784 params
=escape_rfc3986(url_parsed
.params
),
2785 query
=escape_rfc3986(url_parsed
.query
),
2786 fragment
=escape_rfc3986(url_parsed
.fragment
)
2791 return compat_parse_qs(compat_urllib_parse_urlparse(url
).query
)
2794 def read_batch_urls(batch_fd
):
2796 if not isinstance(url
, compat_str
):
2797 url
= url
.decode('utf-8', 'replace')
2798 BOM_UTF8
= ('\xef\xbb\xbf', '\ufeff')
2799 for bom
in BOM_UTF8
:
2800 if url
.startswith(bom
):
2801 url
= url
[len(bom
):]
2803 if not url
or url
.startswith(('#', ';', ']')):
2805 # "#" cannot be stripped out since it is part of the URI
2806 # However, it can be safely stipped out if follwing a whitespace
2807 return re
.split(r
'\s#', url
, 1)[0].rstrip()
2809 with contextlib
.closing(batch_fd
) as fd
:
2810 return [url
for url
in map(fixup
, fd
) if url
]
2813 def urlencode_postdata(*args
, **kargs
):
2814 return compat_urllib_parse_urlencode(*args
, **kargs
).encode('ascii')
2817 def update_url_query(url
, query
):
2820 parsed_url
= compat_urlparse
.urlparse(url
)
2821 qs
= compat_parse_qs(parsed_url
.query
)
2823 return compat_urlparse
.urlunparse(parsed_url
._replace
(
2824 query
=compat_urllib_parse_urlencode(qs
, True)))
2827 def update_Request(req
, url
=None, data
=None, headers
={}, query={}
):
2828 req_headers
= req
.headers
.copy()
2829 req_headers
.update(headers
)
2830 req_data
= data
or req
.data
2831 req_url
= update_url_query(url
or req
.get_full_url(), query
)
2832 req_get_method
= req
.get_method()
2833 if req_get_method
== 'HEAD':
2834 req_type
= HEADRequest
2835 elif req_get_method
== 'PUT':
2836 req_type
= PUTRequest
2838 req_type
= compat_urllib_request
.Request
2840 req_url
, data
=req_data
, headers
=req_headers
,
2841 origin_req_host
=req
.origin_req_host
, unverifiable
=req
.unverifiable
)
2842 if hasattr(req
, 'timeout'):
2843 new_req
.timeout
= req
.timeout
2847 def _multipart_encode_impl(data
, boundary
):
2848 content_type
= 'multipart/form-data; boundary=%s' % boundary
2851 for k
, v
in data
.items():
2852 out
+= b
'--' + boundary
.encode('ascii') + b
'\r\n'
2853 if isinstance(k
, compat_str
):
2854 k
= k
.encode('utf-8')
2855 if isinstance(v
, compat_str
):
2856 v
= v
.encode('utf-8')
2857 # RFC 2047 requires non-ASCII field names to be encoded, while RFC 7578
2858 # suggests sending UTF-8 directly. Firefox sends UTF-8, too
2859 content
= b
'Content-Disposition: form-data; name="' + k
+ b
'"\r\n\r\n' + v
+ b
'\r\n'
2860 if boundary
.encode('ascii') in content
:
2861 raise ValueError('Boundary overlaps with data')
2864 out
+= b
'--' + boundary
.encode('ascii') + b
'--\r\n'
2866 return out
, content_type
2869 def multipart_encode(data
, boundary
=None):
2871 Encode a dict to RFC 7578-compliant form-data
2874 A dict where keys and values can be either Unicode or bytes-like
2877 If specified a Unicode object, it's used as the boundary. Otherwise
2878 a random boundary is generated.
2880 Reference: https://tools.ietf.org/html/rfc7578
2882 has_specified_boundary
= boundary
is not None
2885 if boundary
is None:
2886 boundary
= '---------------' + str(random
.randrange(0x0fffffff, 0xffffffff))
2889 out
, content_type
= _multipart_encode_impl(data
, boundary
)
2892 if has_specified_boundary
:
2896 return out
, content_type
2899 def dict_get(d
, key_or_keys
, default
=None, skip_false_values
=True):
2900 for val
in map(d
.get
, variadic(key_or_keys
)):
2901 if val
is not None and (val
or not skip_false_values
):
2906 def try_call(*funcs
, expected_type
=None, args
=[], kwargs
={}):
2909 val
= f(*args
, **kwargs
)
2910 except (AttributeError, KeyError, TypeError, IndexError, ZeroDivisionError):
2913 if expected_type
is None or isinstance(val
, expected_type
):
2917 def try_get(src
, getter
, expected_type
=None):
2918 return try_call(*variadic(getter
), args
=(src
,), expected_type
=expected_type
)
2921 def filter_dict(dct
, cndn
=lambda _
, v
: v
is not None):
2922 return {k: v for k, v in dct.items() if cndn(k, v)}
2925 def merge_dicts(*dicts
):
2927 for a_dict
in dicts
:
2928 for k
, v
in a_dict
.items():
2929 if (v
is not None and k
not in merged
2930 or isinstance(v
, str) and merged
[k
] == ''):
2935 def encode_compat_str(string
, encoding
=preferredencoding(), errors
='strict'):
2936 return string
if isinstance(string
, compat_str
) else compat_str(string
, encoding
, errors
)
2948 TV_PARENTAL_GUIDELINES
= {
2958 def parse_age_limit(s
):
2960 return s
if 0 <= s
<= 21 else None
2961 if not isinstance(s
, str):
2963 m
= re
.match(r
'^(?P<age>\d{1,2})\+?$', s
)
2965 return int(m
.group('age'))
2968 return US_RATINGS
[s
]
2969 m
= re
.match(r
'^TV[_-]?(%s)$' % '|'.join(k
[3:] for k
in TV_PARENTAL_GUIDELINES
), s
)
2971 return TV_PARENTAL_GUIDELINES
['TV-' + m
.group(1)]
2975 def strip_jsonp(code
):
2978 (?:window\.)?(?P<func_name>[a-zA-Z0-9_.$]*)
2979 (?:\s*&&\s*(?P=func_name))?
2980 \s*\(\s*(?P<callback_data>.*)\);?
2981 \s*?(?://[^\n]*)*$''',
2982 r
'\g<callback_data>', code
)
2985 def js_to_json(code
, vars={}):
2986 # vars is a dict of var, val pairs to substitute
2987 COMMENT_RE
= r
'/\*(?:(?!\*/).)*?\*/|//[^\n]*\n'
2988 SKIP_RE
= fr
'\s*(?:{COMMENT_RE})?\s*'
2990 (fr
'(?s)^(0[xX][0-9a-fA-F]+){SKIP_RE}:?$', 16),
2991 (fr
'(?s)^(0+[0-7]+){SKIP_RE}:?$', 8),
2996 if v
in ('true', 'false', 'null'):
2998 elif v
in ('undefined', 'void 0'):
3000 elif v
.startswith('/*') or v
.startswith('//') or v
.startswith('!') or v
== ',':
3003 if v
[0] in ("'", '"'):
3004 v
= re
.sub(r
'(?s)\\.|"', lambda m
: {
3009 }.get(m
.group(0), m
.group(0)), v
[1:-1])
3011 for regex
, base
in INTEGER_TABLE
:
3012 im
= re
.match(regex
, v
)
3014 i
= int(im
.group(1), base
)
3015 return '"%d":' % i
if v
.endswith(':') else '%d' % i
3022 code
= re
.sub(r
'new Date\((".+")\)', r
'\g<1>', code
)
3024 return re
.sub(r
'''(?sx)
3025 "(?:[^"\\]*(?:\\\\|\\['"nurtbfx/\n]))*[^"\\]*"|
3026 '(?:[^'\\]*(?:\\\\|\\['"nurtbfx/\n]))*[^'\\]*'|
3027 {comment}|,(?={skip}[\]}}])|
3028 void\s0|(?:(?<![0-9])[eE]|[a-df-zA-DF-Z_$])[.a-zA-Z_$0-9]*|
3029 \b(?:0[xX][0-9a-fA-F]+|0+[0-7]+)(?:{skip}:)?|
3032 '''.format(comment
=COMMENT_RE
, skip
=SKIP_RE
), fix_kv
, code
)
3035 def qualities(quality_ids
):
3036 """ Get a numeric quality value out of a list of possible values """
3039 return quality_ids
.index(qid
)
3045 POSTPROCESS_WHEN
= {'pre_process', 'after_filter', 'before_dl', 'after_move', 'post_process', 'after_video', 'playlist'}
3049 'default': '%(title)s [%(id)s].%(ext)s',
3050 'chapter': '%(title)s - %(section_number)03d %(section_title)s [%(id)s].%(ext)s',
3056 'description': 'description',
3057 'annotation': 'annotations.xml',
3058 'infojson': 'info.json',
3061 'pl_thumbnail': None,
3062 'pl_description': 'description',
3063 'pl_infojson': 'info.json',
3066 # As of [1] format syntax is:
3067 # %[mapping_key][conversion_flags][minimum_width][.precision][length_modifier]type
3068 # 1. https://docs.python.org/2/library/stdtypes.html#string-formatting
3069 STR_FORMAT_RE_TMPL
= r
'''(?x)
3070 (?<!%)(?P<prefix>(?:%%)*)
3072 (?P<has_key>\((?P<key>{0})\))?
3074 (?P<conversion>[#0\-+ ]+)?
3076 (?P<precision>\.\d+)?
3077 (?P<len_mod>[hlL])? # unused in python
3078 {1} # conversion type
3083 STR_FORMAT_TYPES
= 'diouxXeEfFgGcrs'
3086 def limit_length(s
, length
):
3087 """ Add ellipses to overly long strings """
3092 return s
[:length
- len(ELLIPSES
)] + ELLIPSES
3096 def version_tuple(v
):
3097 return tuple(int(e
) for e
in re
.split(r
'[-.]', v
))
3100 def is_outdated_version(version
, limit
, assume_new
=True):
3102 return not assume_new
3104 return version_tuple(version
) < version_tuple(limit
)
3106 return not assume_new
3109 def ytdl_is_updateable():
3110 """ Returns if yt-dlp can be updated with -U """
3112 from .update
import is_non_updateable
3114 return not is_non_updateable()
3117 def args_to_str(args
):
3118 # Get a short string representation for a subprocess command
3119 return ' '.join(compat_shlex_quote(a
) for a
in args
)
3122 def error_to_compat_str(err
):
3126 def error_to_str(err
):
3127 return f
'{type(err).__name__}: {err}'
3130 def mimetype2ext(mt
):
3134 mt
, _
, params
= mt
.partition(';')
3139 # Per RFC 3003, audio/mpeg can be .mp1, .mp2 or .mp3. Here use .mp3 as
3140 # it's the most popular one
3141 'audio/mpeg': 'mp3',
3142 'audio/x-wav': 'wav',
3144 'audio/wave': 'wav',
3147 ext
= FULL_MAP
.get(mt
)
3153 'smptett+xml': 'tt',
3157 'x-mp4-fragmented': 'mp4',
3158 'x-ms-sami': 'sami',
3161 'x-mpegurl': 'm3u8',
3162 'vnd.apple.mpegurl': 'm3u8',
3166 'vnd.ms-sstr+xml': 'ism',
3170 'filmstrip+json': 'fs',
3174 _
, _
, subtype
= mt
.rpartition('/')
3175 ext
= SUBTYPE_MAP
.get(subtype
.lower())
3186 _
, _
, suffix
= subtype
.partition('+')
3187 ext
= SUFFIX_MAP
.get(suffix
)
3191 return subtype
.replace('+', '.')
3194 def ext2mimetype(ext_or_url
):
3197 if '.' not in ext_or_url
:
3198 ext_or_url
= f
'file.{ext_or_url}'
3199 return mimetypes
.guess_type(ext_or_url
)[0]
3202 def parse_codecs(codecs_str
):
3203 # http://tools.ietf.org/html/rfc6381
3206 split_codecs
= list(filter(None, map(
3207 str.strip
, codecs_str
.strip().strip(',').split(','))))
3208 vcodec
, acodec
, tcodec
, hdr
= None, None, None, None
3209 for full_codec
in split_codecs
:
3210 parts
= full_codec
.split('.')
3211 codec
= parts
[0].replace('0', '')
3212 if codec
in ('avc1', 'avc2', 'avc3', 'avc4', 'vp9', 'vp8', 'hev1', 'hev2',
3213 'h263', 'h264', 'mp4v', 'hvc1', 'av1', 'theora', 'dvh1', 'dvhe'):
3215 vcodec
= '.'.join(parts
[:4]) if codec
in ('vp9', 'av1', 'hvc1') else full_codec
3216 if codec
in ('dvh1', 'dvhe'):
3218 elif codec
== 'av1' and len(parts
) > 3 and parts
[3] == '10':
3220 elif full_codec
.replace('0', '').startswith('vp9.2'):
3222 elif codec
in ('flac', 'mp4a', 'opus', 'vorbis', 'mp3', 'aac', 'ac-3', 'ec-3', 'eac3', 'dtsc', 'dtse', 'dtsh', 'dtsl'):
3225 elif codec
in ('stpp', 'wvtt',):
3229 write_string('WARNING: Unknown codec %s\n' % full_codec
, sys
.stderr
)
3230 if vcodec
or acodec
or tcodec
:
3232 'vcodec': vcodec
or 'none',
3233 'acodec': acodec
or 'none',
3234 'dynamic_range': hdr
,
3235 **({'tcodec': tcodec}
if tcodec
is not None else {}),
3237 elif len(split_codecs
) == 2:
3239 'vcodec': split_codecs
[0],
3240 'acodec': split_codecs
[1],
3245 def urlhandle_detect_ext(url_handle
):
3246 getheader
= url_handle
.headers
.get
3248 cd
= getheader('Content-Disposition')
3250 m
= re
.match(r
'attachment;\s*filename="(?P<filename>[^"]+)"', cd
)
3252 e
= determine_ext(m
.group('filename'), default_ext
=None)
3256 return mimetype2ext(getheader('Content-Type'))
3259 def encode_data_uri(data
, mime_type
):
3260 return 'data:%s;base64,%s' % (mime_type
, base64
.b64encode(data
).decode('ascii'))
3263 def age_restricted(content_limit
, age_limit
):
3264 """ Returns True iff the content should be blocked """
3266 if age_limit
is None: # No limit set
3268 if content_limit
is None:
3269 return False # Content available for everyone
3270 return age_limit
< content_limit
3273 def is_html(first_bytes
):
3274 """ Detect whether a file contains HTML by examining its first bytes. """
3277 (b
'\xef\xbb\xbf', 'utf-8'),
3278 (b
'\x00\x00\xfe\xff', 'utf-32-be'),
3279 (b
'\xff\xfe\x00\x00', 'utf-32-le'),
3280 (b
'\xff\xfe', 'utf-16-le'),
3281 (b
'\xfe\xff', 'utf-16-be'),
3283 for bom
, enc
in BOMS
:
3284 if first_bytes
.startswith(bom
):
3285 s
= first_bytes
[len(bom
):].decode(enc
, 'replace')
3288 s
= first_bytes
.decode('utf-8', 'replace')
3290 return re
.match(r
'^\s*<', s
)
3293 def determine_protocol(info_dict
):
3294 protocol
= info_dict
.get('protocol')
3295 if protocol
is not None:
3298 url
= sanitize_url(info_dict
['url'])
3299 if url
.startswith('rtmp'):
3301 elif url
.startswith('mms'):
3303 elif url
.startswith('rtsp'):
3306 ext
= determine_ext(url
)
3312 return compat_urllib_parse_urlparse(url
).scheme
3315 def render_table(header_row
, data
, delim
=False, extra_gap
=0, hide_empty
=False):
3316 """ Render a list of rows, each as a list of values.
3317 Text after a \t will be right aligned """
3319 return len(remove_terminal_sequences(string
).replace('\t', ''))
3321 def get_max_lens(table
):
3322 return [max(width(str(v
)) for v
in col
) for col
in zip(*table
)]
3324 def filter_using_list(row
, filterArray
):
3325 return [col
for take
, col
in itertools
.zip_longest(filterArray
, row
, fillvalue
=True) if take
]
3327 max_lens
= get_max_lens(data
) if hide_empty
else []
3328 header_row
= filter_using_list(header_row
, max_lens
)
3329 data
= [filter_using_list(row
, max_lens
) for row
in data
]
3331 table
= [header_row
] + data
3332 max_lens
= get_max_lens(table
)
3335 table
= [header_row
, [delim
* (ml
+ extra_gap
) for ml
in max_lens
]] + data
3336 table
[1][-1] = table
[1][-1][:-extra_gap
* len(delim
)] # Remove extra_gap from end of delimiter
3338 for pos
, text
in enumerate(map(str, row
)):
3340 row
[pos
] = text
.replace('\t', ' ' * (max_lens
[pos
] - width(text
))) + ' ' * extra_gap
3342 row
[pos
] = text
+ ' ' * (max_lens
[pos
] - width(text
) + extra_gap
)
3343 ret
= '\n'.join(''.join(row
).rstrip() for row
in table
)
3347 def _match_one(filter_part
, dct
, incomplete
):
3348 # TODO: Generalize code with YoutubeDL._build_format_filter
3349 STRING_OPERATORS
= {
3350 '*=': operator
.contains
,
3351 '^=': lambda attr
, value
: attr
.startswith(value
),
3352 '$=': lambda attr
, value
: attr
.endswith(value
),
3353 '~=': lambda attr
, value
: re
.search(value
, attr
),
3355 COMPARISON_OPERATORS
= {
3357 '<=': operator
.le
, # "<=" must be defined above "<"
3364 if isinstance(incomplete
, bool):
3365 is_incomplete
= lambda _
: incomplete
3367 is_incomplete
= lambda k
: k
in incomplete
3369 operator_rex
= re
.compile(r
'''(?x)\s*
3371 \s*(?P<negation>!\s*)?(?P<op>%s)(?P<none_inclusive>\s*\?)?\s*
3373 (?P<quote>["\'])(?P
<quotedstrval
>.+?
)(?P
=quote
)|
3377 ''' % '|'.join(map(re.escape, COMPARISON_OPERATORS.keys())))
3378 m = operator_rex.search(filter_part)
3381 unnegated_op = COMPARISON_OPERATORS[m['op']]
3383 op = lambda attr, value: not unnegated_op(attr, value)
3386 comparison_value = m['quotedstrval'] or m['strval'] or m['intval']
3388 comparison_value = comparison_value.replace(r'\%s' % m['quote'], m['quote'])
3389 actual_value = dct.get(m['key'])
3390 numeric_comparison = None
3391 if isinstance(actual_value, (int, float)):
3392 # If the original field is a string and matching comparisonvalue is
3393 # a number we should respect the origin of the original field
3394 # and process comparison value as a string (see
3395 # https://github.com/ytdl-org/youtube-dl/issues/11082)
3397 numeric_comparison = int(comparison_value)
3399 numeric_comparison = parse_filesize(comparison_value)
3400 if numeric_comparison is None:
3401 numeric_comparison = parse_filesize(f'{comparison_value}B')
3402 if numeric_comparison is None:
3403 numeric_comparison = parse_duration(comparison_value)
3404 if numeric_comparison is not None and m['op'] in STRING_OPERATORS:
3405 raise ValueError('Operator %s only supports string values!' % m['op'])
3406 if actual_value is None:
3407 return is_incomplete(m['key']) or m['none_inclusive']
3408 return op(actual_value, comparison_value if numeric_comparison is None else numeric_comparison)
3411 '': lambda v: (v is True) if isinstance(v, bool) else (v is not None),
3412 '!': lambda v: (v is False) if isinstance(v, bool) else (v is None),
3414 operator_rex = re.compile(r'''(?x
)\s
*
3415 (?P
<op
>%s)\s
*(?P
<key
>[a
-z_
]+)
3417 ''' % '|'.join(map(re.escape, UNARY_OPERATORS.keys())))
3418 m = operator_rex.search(filter_part)
3420 op = UNARY_OPERATORS[m.group('op')]
3421 actual_value = dct.get(m.group('key'))
3422 if is_incomplete(m.group('key')) and actual_value is None:
3424 return op(actual_value)
3426 raise ValueError('Invalid filter part %r' % filter_part)
3429 def match_str(filter_str, dct, incomplete=False):
3430 """ Filter a dictionary with a simple string syntax.
3431 @returns Whether the filter passes
3432 @param incomplete Set of keys that is expected to be missing from dct.
3433 Can be True/False to indicate all/none of the keys may be missing.
3434 All conditions on incomplete keys pass if the key is missing
3437 _match_one(filter_part.replace(r'\&', '&'), dct, incomplete)
3438 for filter_part in re.split(r'(?<!\\)&', filter_str))
3441 def match_filter_func(filters):
3444 filters = variadic(filters)
3446 def _match_func(info_dict, *args, **kwargs):
3447 if any(match_str(f, info_dict, *args, **kwargs) for f in filters):
3450 video_title = info_dict.get('title') or info_dict.get('id') or 'video'
3451 filter_str = ') | ('.join(map(str.strip, filters))
3452 return f'{video_title} does not pass filter ({filter_str}), skipping ..'
3456 def parse_dfxp_time_expr(time_expr):
3460 mobj = re.match(r'^(?P<time_offset>\d+(?:\.\d+)?)s?$', time_expr)
3462 return float(mobj.group('time_offset'))
3464 mobj = re.match(r'^(\d+):(\d\d):(\d\d(?:(?:\.|:)\d+)?)$', time_expr)
3466 return 3600 * int(mobj.group(1)) + 60 * int(mobj.group(2)) + float(mobj.group(3).replace(':', '.'))
3469 def srt_subtitles_timecode(seconds):
3470 return '%02d:%02d:%02d,%03d' % timetuple_from_msec(seconds * 1000)
3473 def ass_subtitles_timecode(seconds):
3474 time = timetuple_from_msec(seconds * 1000)
3475 return '%01d:%02d:%02d.%02d' % (*time[:-1], time.milliseconds / 10)
3478 def dfxp2srt(dfxp_data):
3480 @param dfxp_data A
bytes-like
object containing DFXP data
3481 @returns A
unicode object containing converted SRT data
3483 LEGACY_NAMESPACES = (
3484 (b'http://www.w3.org/ns/ttml', [
3485 b'http://www.w3.org/2004/11/ttaf1',
3486 b'http://www.w3.org/2006/04/ttaf1',
3487 b'http://www.w3.org/2006/10/ttaf1',
3489 (b'http://www.w3.org/ns/ttml#styling', [
3490 b'http://www.w3.org/ns/ttml#style',
3494 SUPPORTED_STYLING = [
3503 _x = functools.partial(xpath_with_ns, ns_map={
3504 'xml': 'http://www.w3.org/XML/1998/namespace',
3505 'ttml': 'http://www.w3.org/ns/ttml',
3506 'tts': 'http://www.w3.org/ns/ttml#styling',
3512 class TTMLPElementParser:
3514 _unclosed_elements = []
3515 _applied_styles = []
3517 def start(self, tag, attrib):
3518 if tag in (_x('ttml:br'), 'br'):
3521 unclosed_elements = []
3523 element_style_id = attrib.get('style')
3525 style.update(default_style)
3526 if element_style_id:
3527 style.update(styles.get(element_style_id, {}))
3528 for prop in SUPPORTED_STYLING:
3529 prop_val = attrib.get(_x('tts:' + prop))
3531 style[prop] = prop_val
3534 for k, v in sorted(style.items()):
3535 if self._applied_styles and self._applied_styles[-1].get(k) == v:
3538 font += ' color="%s"' % v
3539 elif k == 'fontSize':
3540 font += ' size="%s"' % v
3541 elif k == 'fontFamily':
3542 font += ' face="%s"' % v
3543 elif k == 'fontWeight' and v == 'bold':
3545 unclosed_elements.append('b')
3546 elif k == 'fontStyle' and v == 'italic':
3548 unclosed_elements.append('i')
3549 elif k == 'textDecoration' and v == 'underline':
3551 unclosed_elements.append('u')
3553 self._out += '<font' + font + '>'
3554 unclosed_elements.append('font')
3556 if self._applied_styles:
3557 applied_style.update(self._applied_styles[-1])
3558 applied_style.update(style)
3559 self._applied_styles.append(applied_style)
3560 self._unclosed_elements.append(unclosed_elements)
3563 if tag not in (_x('ttml:br'), 'br'):
3564 unclosed_elements = self._unclosed_elements.pop()
3565 for element in reversed(unclosed_elements):
3566 self._out += '</%s>' % element
3567 if unclosed_elements and self._applied_styles:
3568 self._applied_styles.pop()
3570 def data(self, data):
3574 return self._out.strip()
3576 def parse_node(node):
3577 target = TTMLPElementParser()
3578 parser = xml.etree.ElementTree.XMLParser(target=target)
3579 parser.feed(xml.etree.ElementTree.tostring(node))
3580 return parser.close()
3582 for k, v in LEGACY_NAMESPACES:
3584 dfxp_data = dfxp_data.replace(ns, k)
3586 dfxp = compat_etree_fromstring(dfxp_data)
3588 paras = dfxp.findall(_x('.//ttml:p')) or dfxp.findall('.//p')
3591 raise ValueError('Invalid dfxp/TTML subtitle')
3595 for style in dfxp.findall(_x('.//ttml:style')):
3596 style_id = style.get('id') or style.get(_x('xml:id'))
3599 parent_style_id = style.get('style')
3601 if parent_style_id not in styles:
3604 styles[style_id] = styles[parent_style_id].copy()
3605 for prop in SUPPORTED_STYLING:
3606 prop_val = style.get(_x('tts:' + prop))
3608 styles.setdefault(style_id, {})[prop] = prop_val
3614 for p in ('body', 'div'):
3615 ele = xpath_element(dfxp, [_x('.//ttml:' + p), './/' + p])
3618 style = styles.get(ele.get('style'))
3621 default_style.update(style)
3623 for para, index in zip(paras, itertools.count(1)):
3624 begin_time = parse_dfxp_time_expr(para.attrib.get('begin'))
3625 end_time = parse_dfxp_time_expr(para.attrib.get('end'))
3626 dur = parse_dfxp_time_expr(para.attrib.get('dur'))
3627 if begin_time is None:
3632 end_time = begin_time + dur
3633 out.append('%d\n%s --> %s\n%s\n\n' % (
3635 srt_subtitles_timecode(begin_time),
3636 srt_subtitles_timecode(end_time),
3642 def cli_option(params, command_option, param):
3643 param = params.get(param)
3645 param = compat_str(param)
3646 return [command_option, param] if param is not None else []
3649 def cli_bool_option(params, command_option, param, true_value='true', false_value='false', separator=None):
3650 param = params.get(param)
3653 assert isinstance(param, bool)
3655 return [command_option + separator + (true_value if param else false_value)]
3656 return [command_option, true_value if param else false_value]
3659 def cli_valueless_option(params, command_option, param, expected_value=True):
3660 param = params.get(param)
3661 return [command_option] if param == expected_value else []
3664 def cli_configuration_args(argdict, keys, default=[], use_compat=True):
3665 if isinstance(argdict, (list, tuple)): # for backward compatibility
3672 assert isinstance(argdict, dict)
3674 assert isinstance(keys, (list, tuple))
3675 for key_list in keys:
3676 arg_list = list(filter(
3677 lambda x: x is not None,
3678 [argdict.get(key.lower()) for key in variadic(key_list)]))
3680 return [arg for args in arg_list for arg in args]
3684 def _configuration_args(main_key, argdict, exe, keys=None, default=[], use_compat=True):
3685 main_key, exe = main_key.lower(), exe.lower()
3686 root_key = exe if main_key == exe else f'{main_key}+{exe}'
3687 keys = [f'{root_key}{k}' for k in (keys or [''])]
3688 if root_key in keys:
3690 keys.append((main_key, exe))
3691 keys.append('default')
3694 return cli_configuration_args(argdict, keys, default, use_compat)
3698 # See http://www.loc.gov/standards/iso639-2/ISO-639-2_utf-8.txt
3757 'iw': 'heb', # Replaced by he in 1989 revision
3767 'in': 'ind', # Replaced by id in 1989 revision
3882 'ji': 'yid', # Replaced by yi in 1989 revision
3890 def short2long(cls, code):
3891 """Convert language code from ISO 639-1 to ISO 639-2/T"""
3892 return cls._lang_map.get(code[:2])
3895 def long2short(cls, code):
3896 """Convert language code from ISO 639-2/T to ISO 639-1"""
3897 for short_name, long_name in cls._lang_map.items():
3898 if long_name == code:
3903 # From http://data.okfn.org/data/core/country-list
3905 'AF': 'Afghanistan',
3906 'AX': 'Åland Islands',
3909 'AS': 'American Samoa',
3914 'AG': 'Antigua and Barbuda',
3931 'BO': 'Bolivia, Plurinational State of',
3932 'BQ': 'Bonaire, Sint Eustatius and Saba',
3933 'BA': 'Bosnia and Herzegovina',
3935 'BV': 'Bouvet Island',
3937 'IO': 'British Indian Ocean Territory',
3938 'BN': 'Brunei Darussalam',
3940 'BF': 'Burkina Faso',
3946 'KY': 'Cayman Islands',
3947 'CF': 'Central African Republic',
3951 'CX': 'Christmas Island',
3952 'CC': 'Cocos (Keeling) Islands',
3956 'CD': 'Congo, the Democratic Republic of the',
3957 'CK': 'Cook Islands',
3959 'CI': 'Côte d\'Ivoire',
3964 'CZ': 'Czech Republic',
3968 'DO': 'Dominican Republic',
3971 'SV': 'El Salvador',
3972 'GQ': 'Equatorial Guinea',
3976 'FK': 'Falkland Islands (Malvinas)',
3977 'FO': 'Faroe Islands',
3981 'GF': 'French Guiana',
3982 'PF': 'French Polynesia',
3983 'TF': 'French Southern Territories',
3998 'GW': 'Guinea-Bissau',
4001 'HM': 'Heard Island and McDonald Islands',
4002 'VA': 'Holy See (Vatican City State)',
4009 'IR': 'Iran, Islamic Republic of',
4012 'IM': 'Isle of Man',
4022 'KP': 'Korea, Democratic People\'s Republic of',
4023 'KR': 'Korea, Republic of',
4026 'LA': 'Lao People\'s Democratic Republic',
4032 'LI': 'Liechtenstein',
4036 'MK': 'Macedonia, the Former Yugoslav Republic of',
4043 'MH': 'Marshall Islands',
4049 'FM': 'Micronesia, Federated States of',
4050 'MD': 'Moldova, Republic of',
4061 'NL': 'Netherlands',
4062 'NC': 'New Caledonia',
4063 'NZ': 'New Zealand',
4068 'NF': 'Norfolk Island',
4069 'MP': 'Northern Mariana Islands',
4074 'PS': 'Palestine, State of',
4076 'PG': 'Papua New Guinea',
4079 'PH': 'Philippines',
4083 'PR': 'Puerto Rico',
4087 'RU': 'Russian Federation',
4089 'BL': 'Saint Barthélemy',
4090 'SH': 'Saint Helena, Ascension and Tristan da Cunha',
4091 'KN': 'Saint Kitts and Nevis',
4092 'LC': 'Saint Lucia',
4093 'MF': 'Saint Martin (French part)',
4094 'PM': 'Saint Pierre and Miquelon',
4095 'VC': 'Saint Vincent and the Grenadines',
4098 'ST': 'Sao Tome and Principe',
4099 'SA': 'Saudi Arabia',
4103 'SL': 'Sierra Leone',
4105 'SX': 'Sint Maarten (Dutch part)',
4108 'SB': 'Solomon Islands',
4110 'ZA': 'South Africa',
4111 'GS': 'South Georgia and the South Sandwich Islands',
4112 'SS': 'South Sudan',
4117 'SJ': 'Svalbard and Jan Mayen',
4120 'CH': 'Switzerland',
4121 'SY': 'Syrian Arab Republic',
4122 'TW': 'Taiwan, Province of China',
4124 'TZ': 'Tanzania, United Republic of',
4126 'TL': 'Timor-Leste',
4130 'TT': 'Trinidad and Tobago',
4133 'TM': 'Turkmenistan',
4134 'TC': 'Turks and Caicos Islands',
4138 'AE': 'United Arab Emirates',
4139 'GB': 'United Kingdom',
4140 'US': 'United States',
4141 'UM': 'United States Minor Outlying Islands',
4145 'VE': 'Venezuela, Bolivarian Republic of',
4147 'VG': 'Virgin Islands, British',
4148 'VI': 'Virgin Islands, U.S.',
4149 'WF': 'Wallis and Futuna',
4150 'EH': 'Western Sahara',
4157 def short2full(cls, code):
4158 """Convert an ISO 3166-2 country code to the corresponding full name"""
4159 return cls._country_map.get(code.upper())
4163 # Major IPv4 address blocks per country
4165 'AD': '46.172.224.0/19',
4166 'AE': '94.200.0.0/13',
4167 'AF': '149.54.0.0/17',
4168 'AG': '209.59.64.0/18',
4169 'AI': '204.14.248.0/21',
4170 'AL': '46.99.0.0/16',
4171 'AM': '46.70.0.0/15',
4172 'AO': '105.168.0.0/13',
4173 'AP': '182.50.184.0/21',
4174 'AQ': '23.154.160.0/24',
4175 'AR': '181.0.0.0/12',
4176 'AS': '202.70.112.0/20',
4177 'AT': '77.116.0.0/14',
4178 'AU': '1.128.0.0/11',
4179 'AW': '181.41.0.0/18',
4180 'AX': '185.217.4.0/22',
4181 'AZ': '5.197.0.0/16',
4182 'BA': '31.176.128.0/17',
4183 'BB': '65.48.128.0/17',
4184 'BD': '114.130.0.0/16',
4186 'BF': '102.178.0.0/15',
4187 'BG': '95.42.0.0/15',
4188 'BH': '37.131.0.0/17',
4189 'BI': '154.117.192.0/18',
4190 'BJ': '137.255.0.0/16',
4191 'BL': '185.212.72.0/23',
4192 'BM': '196.12.64.0/18',
4193 'BN': '156.31.0.0/16',
4194 'BO': '161.56.0.0/16',
4195 'BQ': '161.0.80.0/20',
4196 'BR': '191.128.0.0/12',
4197 'BS': '24.51.64.0/18',
4198 'BT': '119.2.96.0/19',
4199 'BW': '168.167.0.0/16',
4200 'BY': '178.120.0.0/13',
4201 'BZ': '179.42.192.0/18',
4202 'CA': '99.224.0.0/11',
4203 'CD': '41.243.0.0/16',
4204 'CF': '197.242.176.0/21',
4205 'CG': '160.113.0.0/16',
4206 'CH': '85.0.0.0/13',
4207 'CI': '102.136.0.0/14',
4208 'CK': '202.65.32.0/19',
4209 'CL': '152.172.0.0/14',
4210 'CM': '102.244.0.0/14',
4211 'CN': '36.128.0.0/10',
4212 'CO': '181.240.0.0/12',
4213 'CR': '201.192.0.0/12',
4214 'CU': '152.206.0.0/15',
4215 'CV': '165.90.96.0/19',
4216 'CW': '190.88.128.0/17',
4217 'CY': '31.153.0.0/16',
4218 'CZ': '88.100.0.0/14',
4220 'DJ': '197.241.0.0/17',
4221 'DK': '87.48.0.0/12',
4222 'DM': '192.243.48.0/20',
4223 'DO': '152.166.0.0/15',
4224 'DZ': '41.96.0.0/12',
4225 'EC': '186.68.0.0/15',
4226 'EE': '90.190.0.0/15',
4227 'EG': '156.160.0.0/11',
4228 'ER': '196.200.96.0/20',
4229 'ES': '88.0.0.0/11',
4230 'ET': '196.188.0.0/14',
4231 'EU': '2.16.0.0/13',
4232 'FI': '91.152.0.0/13',
4233 'FJ': '144.120.0.0/16',
4234 'FK': '80.73.208.0/21',
4235 'FM': '119.252.112.0/20',
4236 'FO': '88.85.32.0/19',
4238 'GA': '41.158.0.0/15',
4240 'GD': '74.122.88.0/21',
4241 'GE': '31.146.0.0/16',
4242 'GF': '161.22.64.0/18',
4243 'GG': '62.68.160.0/19',
4244 'GH': '154.160.0.0/12',
4245 'GI': '95.164.0.0/16',
4246 'GL': '88.83.0.0/19',
4247 'GM': '160.182.0.0/15',
4248 'GN': '197.149.192.0/18',
4249 'GP': '104.250.0.0/19',
4250 'GQ': '105.235.224.0/20',
4251 'GR': '94.64.0.0/13',
4252 'GT': '168.234.0.0/16',
4253 'GU': '168.123.0.0/16',
4254 'GW': '197.214.80.0/20',
4255 'GY': '181.41.64.0/18',
4256 'HK': '113.252.0.0/14',
4257 'HN': '181.210.0.0/16',
4258 'HR': '93.136.0.0/13',
4259 'HT': '148.102.128.0/17',
4260 'HU': '84.0.0.0/14',
4261 'ID': '39.192.0.0/10',
4262 'IE': '87.32.0.0/12',
4263 'IL': '79.176.0.0/13',
4264 'IM': '5.62.80.0/20',
4265 'IN': '117.192.0.0/10',
4266 'IO': '203.83.48.0/21',
4267 'IQ': '37.236.0.0/14',
4268 'IR': '2.176.0.0/12',
4269 'IS': '82.221.0.0/16',
4270 'IT': '79.0.0.0/10',
4271 'JE': '87.244.64.0/18',
4272 'JM': '72.27.0.0/17',
4273 'JO': '176.29.0.0/16',
4274 'JP': '133.0.0.0/8',
4275 'KE': '105.48.0.0/12',
4276 'KG': '158.181.128.0/17',
4277 'KH': '36.37.128.0/17',
4278 'KI': '103.25.140.0/22',
4279 'KM': '197.255.224.0/20',
4280 'KN': '198.167.192.0/19',
4281 'KP': '175.45.176.0/22',
4282 'KR': '175.192.0.0/10',
4283 'KW': '37.36.0.0/14',
4284 'KY': '64.96.0.0/15',
4285 'KZ': '2.72.0.0/13',
4286 'LA': '115.84.64.0/18',
4287 'LB': '178.135.0.0/16',
4288 'LC': '24.92.144.0/20',
4289 'LI': '82.117.0.0/19',
4290 'LK': '112.134.0.0/15',
4291 'LR': '102.183.0.0/16',
4292 'LS': '129.232.0.0/17',
4293 'LT': '78.56.0.0/13',
4294 'LU': '188.42.0.0/16',
4295 'LV': '46.109.0.0/16',
4296 'LY': '41.252.0.0/14',
4297 'MA': '105.128.0.0/11',
4298 'MC': '88.209.64.0/18',
4299 'MD': '37.246.0.0/16',
4300 'ME': '178.175.0.0/17',
4301 'MF': '74.112.232.0/21',
4302 'MG': '154.126.0.0/17',
4303 'MH': '117.103.88.0/21',
4304 'MK': '77.28.0.0/15',
4305 'ML': '154.118.128.0/18',
4306 'MM': '37.111.0.0/17',
4307 'MN': '49.0.128.0/17',
4308 'MO': '60.246.0.0/16',
4309 'MP': '202.88.64.0/20',
4310 'MQ': '109.203.224.0/19',
4311 'MR': '41.188.64.0/18',
4312 'MS': '208.90.112.0/22',
4313 'MT': '46.11.0.0/16',
4314 'MU': '105.16.0.0/12',
4315 'MV': '27.114.128.0/18',
4316 'MW': '102.70.0.0/15',
4317 'MX': '187.192.0.0/11',
4318 'MY': '175.136.0.0/13',
4319 'MZ': '197.218.0.0/15',
4320 'NA': '41.182.0.0/16',
4321 'NC': '101.101.0.0/18',
4322 'NE': '197.214.0.0/18',
4323 'NF': '203.17.240.0/22',
4324 'NG': '105.112.0.0/12',
4325 'NI': '186.76.0.0/15',
4326 'NL': '145.96.0.0/11',
4327 'NO': '84.208.0.0/13',
4328 'NP': '36.252.0.0/15',
4329 'NR': '203.98.224.0/19',
4330 'NU': '49.156.48.0/22',
4331 'NZ': '49.224.0.0/14',
4332 'OM': '5.36.0.0/15',
4333 'PA': '186.72.0.0/15',
4334 'PE': '186.160.0.0/14',
4335 'PF': '123.50.64.0/18',
4336 'PG': '124.240.192.0/19',
4337 'PH': '49.144.0.0/13',
4338 'PK': '39.32.0.0/11',
4339 'PL': '83.0.0.0/11',
4340 'PM': '70.36.0.0/20',
4341 'PR': '66.50.0.0/16',
4342 'PS': '188.161.0.0/16',
4343 'PT': '85.240.0.0/13',
4344 'PW': '202.124.224.0/20',
4345 'PY': '181.120.0.0/14',
4346 'QA': '37.210.0.0/15',
4347 'RE': '102.35.0.0/16',
4348 'RO': '79.112.0.0/13',
4349 'RS': '93.86.0.0/15',
4350 'RU': '5.136.0.0/13',
4351 'RW': '41.186.0.0/16',
4352 'SA': '188.48.0.0/13',
4353 'SB': '202.1.160.0/19',
4354 'SC': '154.192.0.0/11',
4355 'SD': '102.120.0.0/13',
4356 'SE': '78.64.0.0/12',
4357 'SG': '8.128.0.0/10',
4358 'SI': '188.196.0.0/14',
4359 'SK': '78.98.0.0/15',
4360 'SL': '102.143.0.0/17',
4361 'SM': '89.186.32.0/19',
4362 'SN': '41.82.0.0/15',
4363 'SO': '154.115.192.0/18',
4364 'SR': '186.179.128.0/17',
4365 'SS': '105.235.208.0/21',
4366 'ST': '197.159.160.0/19',
4367 'SV': '168.243.0.0/16',
4368 'SX': '190.102.0.0/20',
4370 'SZ': '41.84.224.0/19',
4371 'TC': '65.255.48.0/20',
4372 'TD': '154.68.128.0/19',
4373 'TG': '196.168.0.0/14',
4374 'TH': '171.96.0.0/13',
4375 'TJ': '85.9.128.0/18',
4376 'TK': '27.96.24.0/21',
4377 'TL': '180.189.160.0/20',
4378 'TM': '95.85.96.0/19',
4379 'TN': '197.0.0.0/11',
4380 'TO': '175.176.144.0/21',
4381 'TR': '78.160.0.0/11',
4382 'TT': '186.44.0.0/15',
4383 'TV': '202.2.96.0/19',
4384 'TW': '120.96.0.0/11',
4385 'TZ': '156.156.0.0/14',
4386 'UA': '37.52.0.0/14',
4387 'UG': '102.80.0.0/13',
4389 'UY': '167.56.0.0/13',
4390 'UZ': '84.54.64.0/18',
4391 'VA': '212.77.0.0/19',
4392 'VC': '207.191.240.0/21',
4393 'VE': '186.88.0.0/13',
4394 'VG': '66.81.192.0/20',
4395 'VI': '146.226.0.0/16',
4396 'VN': '14.160.0.0/11',
4397 'VU': '202.80.32.0/20',
4398 'WF': '117.20.32.0/21',
4399 'WS': '202.4.32.0/19',
4400 'YE': '134.35.0.0/16',
4401 'YT': '41.242.116.0/22',
4402 'ZA': '41.0.0.0/11',
4403 'ZM': '102.144.0.0/13',
4404 'ZW': '102.177.192.0/18',
4408 def random_ipv4(cls, code_or_block):
4409 if len(code_or_block) == 2:
4410 block = cls._country_ip_map.get(code_or_block.upper())
4414 block = code_or_block
4415 addr, preflen = block.split('/')
4416 addr_min = compat_struct_unpack('!L', socket.inet_aton(addr))[0]
4417 addr_max = addr_min | (0xffffffff >> int(preflen))
4418 return compat_str(socket.inet_ntoa(
4419 compat_struct_pack('!L', random.randint(addr_min, addr_max))))
4422 class PerRequestProxyHandler(compat_urllib_request.ProxyHandler):
4423 def __init__(self, proxies=None):
4424 # Set default handlers
4425 for type in ('http', 'https'):
4426 setattr(self, '%s_open' % type,
4427 lambda r, proxy='__noproxy__', type=type, meth=self.proxy_open:
4428 meth(r, proxy, type))
4429 compat_urllib_request.ProxyHandler.__init__(self, proxies)
4431 def proxy_open(self, req, proxy, type):
4432 req_proxy = req.headers.get('Ytdl-request-proxy')
4433 if req_proxy is not None:
4435 del req.headers['Ytdl-request-proxy']
4437 if proxy == '__noproxy__':
4438 return None # No Proxy
4439 if compat_urlparse.urlparse(proxy).scheme.lower() in ('socks', 'socks4', 'socks4a', 'socks5'):
4440 req.add_header('Ytdl-socks-proxy', proxy)
4441 # yt-dlp's http/https handlers do wrapping the socket with socks
4443 return compat_urllib_request.ProxyHandler.proxy_open(
4444 self, req, proxy, type)
4447 # Both long_to_bytes and bytes_to_long are adapted from PyCrypto, which is
4448 # released into Public Domain
4449 # https://github.com/dlitz/pycrypto/blob/master/lib/Crypto/Util/number.py#L387
4451 def long_to_bytes(n, blocksize=0):
4452 """long_to_bytes(n:long, blocksize:int) : string
4453 Convert a long integer to a byte string.
4455 If optional blocksize is given and greater than zero, pad the front of the
4456 byte string with binary zeros so that the length is a multiple of
4459 # after much testing, this algorithm was deemed to be the fastest
4463 s = compat_struct_pack('>I', n & 0xffffffff) + s
4465 # strip off leading zeros
4466 for i in range(len(s)):
4467 if s[i] != b'\000'[0]:
4470 # only happens when n == 0
4474 # add back some pad bytes. this could be done more efficiently w.r.t. the
4475 # de-padding being done above, but sigh...
4476 if blocksize > 0 and len(s) % blocksize:
4477 s = (blocksize - len(s) % blocksize) * b'\000' + s
4481 def bytes_to_long(s):
4482 """bytes_to_long(string) : long
4483 Convert a byte string to a long integer.
4485 This is (essentially) the inverse of long_to_bytes().
4490 extra = (4 - length % 4)
4491 s = b'\000' * extra + s
4492 length = length + extra
4493 for i in range(0, length, 4):
4494 acc = (acc << 32) + compat_struct_unpack('>I', s[i:i + 4])[0]
4498 def ohdave_rsa_encrypt(data, exponent, modulus):
4500 Implement OHDave
's RSA algorithm. See http://www.ohdave.com/rsa/
4503 data: data to encrypt, bytes-like object
4504 exponent, modulus: parameter e and N of RSA algorithm, both integer
4505 Output: hex string of encrypted data
4507 Limitation: supports one block encryption only
4510 payload = int(binascii.hexlify(data[::-1]), 16)
4511 encrypted = pow(payload, exponent, modulus)
4512 return '%x' % encrypted
4515 def pkcs1pad(data, length):
4517 Padding input data with PKCS#1 scheme
4519 @param {int[]} data input data
4520 @param {int} length target length
4521 @returns {int[]} padded data
4523 if len(data) > length - 11:
4524 raise ValueError('Input data too
long for PKCS
#1 padding')
4526 pseudo_random
= [random
.randint(0, 254) for _
in range(length
- len(data
) - 3)]
4527 return [0, 2] + pseudo_random
+ [0] + data
4530 def encode_base_n(num
, n
, table
=None):
4531 FULL_TABLE
= '0123456789abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ'
4533 table
= FULL_TABLE
[:n
]
4536 raise ValueError('base %d exceeds table length %d' % (n
, len(table
)))
4543 ret
= table
[num
% n
] + ret
4548 def decode_packed_codes(code
):
4549 mobj
= re
.search(PACKED_CODES_RE
, code
)
4550 obfuscated_code
, base
, count
, symbols
= mobj
.groups()
4553 symbols
= symbols
.split('|')
4558 base_n_count
= encode_base_n(count
, base
)
4559 symbol_table
[base_n_count
] = symbols
[count
] or base_n_count
4562 r
'\b(\w+)\b', lambda mobj
: symbol_table
[mobj
.group(0)],
4566 def caesar(s
, alphabet
, shift
):
4571 alphabet
[(alphabet
.index(c
) + shift
) % l
] if c
in alphabet
else c
4576 return caesar(s
, r
'''!"#$%&'()*+,-./0123456789:;<=>?@ABCDEFGHIJKLMNOPQRSTUVWXYZ[\]^_`abcdefghijklmnopqrstuvwxyz{|}~''', 47)
4579 def parse_m3u8_attributes(attrib
):
4581 for (key
, val
) in re
.findall(r
'(?P<key>[A-Z0-9-]+)=(?P<val>"[^"]+"|[^",]+)(?:,|$)', attrib
):
4582 if val
.startswith('"'):
4588 def urshift(val
, n
):
4589 return val
>> n
if val
>= 0 else (val
+ 0x100000000) >> n
4592 # Based on png2str() written by @gdkchan and improved by @yokrysty
4593 # Originally posted at https://github.com/ytdl-org/youtube-dl/issues/9706
4594 def decode_png(png_data
):
4595 # Reference: https://www.w3.org/TR/PNG/
4596 header
= png_data
[8:]
4598 if png_data
[:8] != b
'\x89PNG\x0d\x0a\x1a\x0a' or header
[4:8] != b
'IHDR':
4599 raise OSError('Not a valid PNG file.')
4601 int_map
= {1: '>B', 2: '>H', 4: '>I'}
4602 unpack_integer
= lambda x
: compat_struct_unpack(int_map
[len(x
)], x
)[0]
4607 length
= unpack_integer(header
[:4])
4610 chunk_type
= header
[:4]
4613 chunk_data
= header
[:length
]
4614 header
= header
[length
:]
4616 header
= header
[4:] # Skip CRC
4624 ihdr
= chunks
[0]['data']
4626 width
= unpack_integer(ihdr
[:4])
4627 height
= unpack_integer(ihdr
[4:8])
4631 for chunk
in chunks
:
4632 if chunk
['type'] == b
'IDAT':
4633 idat
+= chunk
['data']
4636 raise OSError('Unable to read PNG data.')
4638 decompressed_data
= bytearray(zlib
.decompress(idat
))
4643 def _get_pixel(idx
):
4648 for y
in range(height
):
4649 basePos
= y
* (1 + stride
)
4650 filter_type
= decompressed_data
[basePos
]
4654 pixels
.append(current_row
)
4656 for x
in range(stride
):
4657 color
= decompressed_data
[1 + basePos
+ x
]
4658 basex
= y
* stride
+ x
4663 left
= _get_pixel(basex
- 3)
4665 up
= _get_pixel(basex
- stride
)
4667 if filter_type
== 1: # Sub
4668 color
= (color
+ left
) & 0xff
4669 elif filter_type
== 2: # Up
4670 color
= (color
+ up
) & 0xff
4671 elif filter_type
== 3: # Average
4672 color
= (color
+ ((left
+ up
) >> 1)) & 0xff
4673 elif filter_type
== 4: # Paeth
4679 c
= _get_pixel(basex
- stride
- 3)
4687 if pa
<= pb
and pa
<= pc
:
4688 color
= (color
+ a
) & 0xff
4690 color
= (color
+ b
) & 0xff
4692 color
= (color
+ c
) & 0xff
4694 current_row
.append(color
)
4696 return width
, height
, pixels
4699 def write_xattr(path
, key
, value
):
4700 # This mess below finds the best xattr tool for the job
4702 # try the pyxattr module...
4705 if hasattr(xattr
, 'set'): # pyxattr
4706 # Unicode arguments are not supported in python-pyxattr until
4708 # See https://github.com/ytdl-org/youtube-dl/issues/5498
4709 pyxattr_required_version
= '0.5.0'
4710 if version_tuple(xattr
.__version
__) < version_tuple(pyxattr_required_version
):
4711 # TODO: fallback to CLI tools
4712 raise XAttrUnavailableError(
4713 'python-pyxattr is detected but is too old. '
4714 'yt-dlp requires %s or above while your version is %s. '
4715 'Falling back to other xattr implementations' % (
4716 pyxattr_required_version
, xattr
.__version
__))
4718 setxattr
= xattr
.set
4720 setxattr
= xattr
.setxattr
4723 setxattr(path
, key
, value
)
4724 except OSError as e
:
4725 raise XAttrMetadataError(e
.errno
, e
.strerror
)
4728 if compat_os_name
== 'nt':
4729 # Write xattrs to NTFS Alternate Data Streams:
4730 # http://en.wikipedia.org/wiki/NTFS#Alternate_data_streams_.28ADS.29
4731 assert ':' not in key
4732 assert os
.path
.exists(path
)
4734 ads_fn
= path
+ ':' + key
4736 with open(ads_fn
, 'wb') as f
:
4738 except OSError as e
:
4739 raise XAttrMetadataError(e
.errno
, e
.strerror
)
4741 user_has_setfattr
= check_executable('setfattr', ['--version'])
4742 user_has_xattr
= check_executable('xattr', ['-h'])
4744 if user_has_setfattr
or user_has_xattr
:
4746 value
= value
.decode('utf-8')
4747 if user_has_setfattr
:
4748 executable
= 'setfattr'
4749 opts
= ['-n', key
, '-v', value
]
4750 elif user_has_xattr
:
4751 executable
= 'xattr'
4752 opts
= ['-w', key
, value
]
4754 cmd
= ([encodeFilename(executable
, True)]
4755 + [encodeArgument(o
) for o
in opts
]
4756 + [encodeFilename(path
, True)])
4760 cmd
, stdout
=subprocess
.PIPE
, stderr
=subprocess
.PIPE
, stdin
=subprocess
.PIPE
)
4761 except OSError as e
:
4762 raise XAttrMetadataError(e
.errno
, e
.strerror
)
4763 stdout
, stderr
= p
.communicate_or_kill()
4764 stderr
= stderr
.decode('utf-8', 'replace')
4765 if p
.returncode
!= 0:
4766 raise XAttrMetadataError(p
.returncode
, stderr
)
4769 # On Unix, and can't find pyxattr, setfattr, or xattr.
4770 if sys
.platform
.startswith('linux'):
4771 raise XAttrUnavailableError(
4772 "Couldn't find a tool to set the xattrs. "
4773 "Install either the python 'pyxattr' or 'xattr' "
4774 "modules, or the GNU 'attr' package "
4775 "(which contains the 'setfattr' tool).")
4777 raise XAttrUnavailableError(
4778 "Couldn't find a tool to set the xattrs. "
4779 "Install either the python 'xattr' module, "
4780 "or the 'xattr' binary.")
4783 def random_birthday(year_field
, month_field
, day_field
):
4784 start_date
= datetime
.date(1950, 1, 1)
4785 end_date
= datetime
.date(1995, 12, 31)
4786 offset
= random
.randint(0, (end_date
- start_date
).days
)
4787 random_date
= start_date
+ datetime
.timedelta(offset
)
4789 year_field
: str(random_date
.year
),
4790 month_field
: str(random_date
.month
),
4791 day_field
: str(random_date
.day
),
4795 # Templates for internet shortcut files, which are plain text files.
4796 DOT_URL_LINK_TEMPLATE
= '''\
4801 DOT_WEBLOC_LINK_TEMPLATE
= '''\
4802 <?xml version="1.0" encoding="UTF-8"?>
4803 <!DOCTYPE plist PUBLIC "-//Apple//DTD PLIST 1.0//EN" "http://www.apple.com/DTDs/PropertyList-1.0.dtd">
4804 <plist version="1.0">
4807 \t<string>%(url)s</string>
4812 DOT_DESKTOP_LINK_TEMPLATE
= '''\
4822 'url': DOT_URL_LINK_TEMPLATE
,
4823 'desktop': DOT_DESKTOP_LINK_TEMPLATE
,
4824 'webloc': DOT_WEBLOC_LINK_TEMPLATE
,
4828 def iri_to_uri(iri
):
4830 Converts an IRI (Internationalized Resource Identifier, allowing Unicode characters) to a URI (Uniform Resource Identifier, ASCII-only).
4832 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.
4835 iri_parts
= compat_urllib_parse_urlparse(iri
)
4837 if '[' in iri_parts
.netloc
:
4838 raise ValueError('IPv6 URIs are not, yet, supported.')
4839 # Querying `.netloc`, when there's only one bracket, also raises a ValueError.
4841 # 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.
4844 if iri_parts
.username
:
4845 net_location
+= urllib
.parse
.quote(iri_parts
.username
, safe
=r
"!$%&'()*+,~")
4846 if iri_parts
.password
is not None:
4847 net_location
+= ':' + urllib
.parse
.quote(iri_parts
.password
, safe
=r
"!$%&'()*+,~")
4850 net_location
+= iri_parts
.hostname
.encode('idna').decode('utf-8') # Punycode for Unicode hostnames.
4851 # The 'idna' encoding produces ASCII text.
4852 if iri_parts
.port
is not None and iri_parts
.port
!= 80:
4853 net_location
+= ':' + str(iri_parts
.port
)
4855 return urllib
.parse
.urlunparse(
4859 urllib
.parse
.quote_plus(iri_parts
.path
, safe
=r
"!$%&'()*+,/:;=@|~"),
4861 # Unsure about the `safe` argument, since this is a legacy way of handling parameters.
4862 urllib
.parse
.quote_plus(iri_parts
.params
, safe
=r
"!$%&'()*+,/:;=@|~"),
4864 # Not totally sure about the `safe` argument, since the source does not explicitly mention the query URI component.
4865 urllib
.parse
.quote_plus(iri_parts
.query
, safe
=r
"!$%&'()*+,/:;=?@{|}~"),
4867 urllib
.parse
.quote_plus(iri_parts
.fragment
, safe
=r
"!#$%&'()*+,/:;=?@{|}~")))
4869 # Source for `safe` arguments: https://url.spec.whatwg.org/#percent-encoded-bytes.
4872 def to_high_limit_path(path
):
4873 if sys
.platform
in ['win32', 'cygwin']:
4874 # Work around MAX_PATH limitation on Windows. The maximum allowed length for the individual path segments may still be quite limited.
4875 return '\\\\?\\' + os
.path
.abspath(path
)
4880 def format_field(obj
, field
=None, template
='%s', ignore
=(None, ''), default
='', func
=None):
4881 val
= traverse_obj(obj
, *variadic(field
))
4884 return template
% (func(val
) if func
else val
)
4887 def clean_podcast_url(url
):
4888 return re
.sub(r
'''(?x)
4892 media\.blubrry\.com| # https://create.blubrry.com/resources/podcast-media-download-statistics/getting-started/
4895 (?:dts|www)\.podtrac\.com/(?:pts/)?redirect\.[0-9a-z]{3,4}| # http://analytics.podtrac.com/how-to-measure
4898 cn\.co| # https://podcorn.com/analytics-prefix/
4899 st\.fm # https://podsights.com/docs/
4904 _HEX_TABLE
= '0123456789abcdef'
4907 def random_uuidv4():
4908 return re
.sub(r
'[xy]', lambda x
: _HEX_TABLE
[random
.randint(0, 15)], 'xxxxxxxx-xxxx-4xxx-yxxx-xxxxxxxxxxxx')
4911 def make_dir(path
, to_screen
=None):
4913 dn
= os
.path
.dirname(path
)
4914 if dn
and not os
.path
.exists(dn
):
4917 except OSError as err
:
4918 if callable(to_screen
) is not None:
4919 to_screen('unable to create directory ' + error_to_compat_str(err
))
4923 def get_executable_path():
4924 from zipimport
import zipimporter
4925 if hasattr(sys
, 'frozen'): # Running from PyInstaller
4926 path
= os
.path
.dirname(sys
.executable
)
4927 elif isinstance(__loader__
, zipimporter
): # Running from ZIP
4928 path
= os
.path
.join(os
.path
.dirname(__file__
), '../..')
4930 path
= os
.path
.join(os
.path
.dirname(__file__
), '..')
4931 return os
.path
.abspath(path
)
4934 def load_plugins(name
, suffix
, namespace
):
4937 plugins_spec
= importlib
.util
.spec_from_file_location(
4938 name
, os
.path
.join(get_executable_path(), 'ytdlp_plugins', name
, '__init__.py'))
4939 plugins
= importlib
.util
.module_from_spec(plugins_spec
)
4940 sys
.modules
[plugins_spec
.name
] = plugins
4941 plugins_spec
.loader
.exec_module(plugins
)
4942 for name
in dir(plugins
):
4943 if name
in namespace
:
4945 if not name
.endswith(suffix
):
4947 klass
= getattr(plugins
, name
)
4948 classes
[name
] = namespace
[name
] = klass
4949 except FileNotFoundError
:
4955 obj
, *path_list
, default
=None, expected_type
=None, get_all
=True,
4956 casesense
=True, is_user_input
=False, traverse_string
=False):
4957 ''' Traverse nested list/dict/tuple
4958 @param path_list A list of paths which are checked one by one.
4959 Each path is a list of keys where each key is a string,
4960 a function, a tuple of strings/None or "...".
4961 When a fuction is given, it takes the key and value as arguments
4962 and returns whether the key matches or not. When a tuple is given,
4963 all the keys given in the tuple are traversed, and
4964 "..." traverses all the keys in the object
4965 "None" returns the object without traversal
4966 @param default Default value to return
4967 @param expected_type Only accept final value of this type (Can also be any callable)
4968 @param get_all Return all the values obtained from a path or only the first one
4969 @param casesense Whether to consider dictionary keys as case sensitive
4970 @param is_user_input Whether the keys are generated from user input. If True,
4971 strings are converted to int/slice if necessary
4972 @param traverse_string Whether to traverse inside strings. If True, any
4973 non-compatible object will also be converted into a string
4977 _lower
= lambda k
: (k
.lower() if isinstance(k
, str) else k
)
4978 path_list
= (map(_lower
, variadic(path
)) for path
in path_list
)
4980 def _traverse_obj(obj
, path
, _current_depth
=0):
4982 path
= tuple(variadic(path
))
4983 for i
, key
in enumerate(path
):
4984 if None in (key
, obj
):
4986 if isinstance(key
, (list, tuple)):
4987 obj
= [_traverse_obj(obj
, sub_key
, _current_depth
) for sub_key
in key
]
4990 obj
= (obj
.values() if isinstance(obj
, dict)
4991 else obj
if isinstance(obj
, (list, tuple, LazyList
))
4992 else str(obj
) if traverse_string
else [])
4994 depth
= max(depth
, _current_depth
)
4995 return [_traverse_obj(inner_obj
, path
[i
+ 1:], _current_depth
) for inner_obj
in obj
]
4997 if isinstance(obj
, (list, tuple, LazyList
)):
4998 obj
= enumerate(obj
)
4999 elif isinstance(obj
, dict):
5002 if not traverse_string
:
5006 depth
= max(depth
, _current_depth
)
5007 return [_traverse_obj(v
, path
[i
+ 1:], _current_depth
) for k
, v
in obj
if try_call(key
, args
=(k
, v
))]
5008 elif isinstance(obj
, dict) and not (is_user_input
and key
== ':'):
5009 obj
= (obj
.get(key
) if casesense
or (key
in obj
)
5010 else next((v
for k
, v
in obj
.items() if _lower(k
) == key
), None))
5013 key
= (int_or_none(key
) if ':' not in key
5014 else slice(*map(int_or_none
, key
.split(':'))))
5015 if key
== slice(None):
5016 return _traverse_obj(obj
, (..., *path
[i
+ 1:]), _current_depth
)
5017 if not isinstance(key
, (int, slice)):
5019 if not isinstance(obj
, (list, tuple, LazyList
)):
5020 if not traverse_string
:
5029 if isinstance(expected_type
, type):
5030 type_test
= lambda val
: val
if isinstance(val
, expected_type
) else None
5031 elif expected_type
is not None:
5032 type_test
= expected_type
5034 type_test
= lambda val
: val
5036 for path
in path_list
:
5038 val
= _traverse_obj(obj
, path
)
5041 for _
in range(depth
- 1):
5042 val
= itertools
.chain
.from_iterable(v
for v
in val
if v
is not None)
5043 val
= [v
for v
in map(type_test
, val
) if v
is not None]
5045 return val
if get_all
else val
[0]
5047 val
= type_test(val
)
5053 def traverse_dict(dictn
, keys
, casesense
=True):
5054 write_string('DeprecationWarning: yt_dlp.utils.traverse_dict is deprecated '
5055 'and may be removed in a future version. Use yt_dlp.utils.traverse_obj instead')
5056 return traverse_obj(dictn
, keys
, casesense
=casesense
, is_user_input
=True, traverse_string
=True)
5059 def get_first(obj
, keys
, **kwargs
):
5060 return traverse_obj(obj
, (..., *variadic(keys
)), **kwargs
, get_all
=False)
5063 def variadic(x
, allowed_types
=(str, bytes, dict)):
5064 return x
if isinstance(x
, collections
.abc
.Iterable
) and not isinstance(x
, allowed_types
) else (x
,)
5067 def decode_base(value
, digits
):
5068 # This will convert given base-x string to scalar (long or int)
5069 table
= {char: index for index, char in enumerate(digits)}
5074 result
+= table
[chr]
5078 def time_seconds(**kwargs
):
5079 t
= datetime
.datetime
.now(datetime
.timezone(datetime
.timedelta(**kwargs
)))
5080 return t
.timestamp()
5083 # create a JSON Web Signature (jws) with HS256 algorithm
5084 # the resulting format is in JWS Compact Serialization
5085 # implemented following JWT https://www.rfc-editor.org/rfc/rfc7519.html
5086 # implemented following JWS https://www.rfc-editor.org/rfc/rfc7515.html
5087 def jwt_encode_hs256(payload_data
, key
, headers
={}):
5093 header_data
.update(headers
)
5094 header_b64
= base64
.b64encode(json
.dumps(header_data
).encode('utf-8'))
5095 payload_b64
= base64
.b64encode(json
.dumps(payload_data
).encode('utf-8'))
5096 h
= hmac
.new(key
.encode('utf-8'), header_b64
+ b
'.' + payload_b64
, hashlib
.sha256
)
5097 signature_b64
= base64
.b64encode(h
.digest())
5098 token
= header_b64
+ b
'.' + payload_b64
+ b
'.' + signature_b64
5102 # can be extended in future to verify the signature and parse header and return the algorithm used if it's not HS256
5103 def jwt_decode_hs256(jwt
):
5104 header_b64
, payload_b64
, signature_b64
= jwt
.split('.')
5105 payload_data
= json
.loads(base64
.urlsafe_b64decode(payload_b64
))
5109 def supports_terminal_sequences(stream
):
5110 if compat_os_name
== 'nt':
5111 from .compat
import WINDOWS_VT_MODE
# Must be imported locally
5112 if not WINDOWS_VT_MODE
or get_windows_version() < (10, 0, 10586):
5114 elif not os
.getenv('TERM'):
5117 return stream
.isatty()
5118 except BaseException
:
5122 _terminal_sequences_re
= re
.compile('\033\\[[^m]+m')
5125 def remove_terminal_sequences(string
):
5126 return _terminal_sequences_re
.sub('', string
)
5129 def number_of_digits(number
):
5130 return len('%d' % number
)
5133 def join_nonempty(*values
, delim
='-', from_dict
=None):
5134 if from_dict
is not None:
5135 values
= map(from_dict
.get
, values
)
5136 return delim
.join(map(str, filter(None, values
)))
5139 def scale_thumbnails_to_max_format_width(formats
, thumbnails
, url_width_re
):
5141 Find the largest format dimensions in terms of video width and, for each thumbnail:
5142 * Modify the URL: Match the width with the provided regex and replace with the former width
5145 This function is useful with video services that scale the provided thumbnails on demand
5147 _keys
= ('width', 'height')
5148 max_dimensions
= max(
5149 (tuple(format
.get(k
) or 0 for k
in _keys
) for format
in formats
),
5151 if not max_dimensions
[0]:
5155 {'url': re.sub(url_width_re, str(max_dimensions[0]), thumbnail['url'])}
,
5156 dict(zip(_keys
, max_dimensions
)), thumbnail
)
5157 for thumbnail
in thumbnails
5161 def parse_http_range(range):
5162 """ Parse value of "Range" or "Content-Range" HTTP header into tuple. """
5164 return None, None, None
5165 crg
= re
.search(r
'bytes[ =](\d+)-(\d+)?(?:/(\d+))?', range)
5167 return None, None, None
5168 return int(crg
.group(1)), int_or_none(crg
.group(2)), int_or_none(crg
.group(3))
5174 __initialized
= False
5176 def __init__(self
, parser
, label
=None):
5177 self
._parser
, self
.label
= parser
, label
5178 self
._loaded
_paths
, self
.configs
= set(), []
5180 def init(self
, args
=None, filename
=None):
5181 assert not self
.__initialized
5184 location
= os
.path
.realpath(filename
)
5185 directory
= os
.path
.dirname(location
)
5186 if location
in self
._loaded
_paths
:
5188 self
._loaded
_paths
.add(location
)
5190 self
.__initialized
= True
5191 self
.own_args
, self
.filename
= args
, filename
5192 for location
in self
._parser
.parse_args(args
)[0].config_locations
or []:
5193 location
= os
.path
.join(directory
, expand_path(location
))
5194 if os
.path
.isdir(location
):
5195 location
= os
.path
.join(location
, 'yt-dlp.conf')
5196 if not os
.path
.exists(location
):
5197 self
._parser
.error(f
'config location {location} does not exist')
5198 self
.append_config(self
.read_file(location
), location
)
5202 label
= join_nonempty(
5203 self
.label
, 'config', f
'"{self.filename}"' if self
.filename
else '',
5205 return join_nonempty(
5206 self
.own_args
is not None and f
'{label[0].upper()}{label[1:]}: {self.hide_login_info(self.own_args)}',
5207 *(f
'\n{c}'.replace('\n', '\n| ')[1:] for c
in self
.configs
),
5211 def read_file(filename
, default
=[]):
5213 optionf
= open(filename
)
5215 return default
# silently skip if file is not present
5217 # FIXME: https://github.com/ytdl-org/youtube-dl/commit/dfe5fa49aed02cf36ba9f743b11b0903554b5e56
5218 contents
= optionf
.read()
5219 res
= shlex
.split(contents
, comments
=True)
5225 def hide_login_info(opts
):
5226 PRIVATE_OPTS
= {'-p', '--password', '-u', '--username', '--video-password', '--ap-password', '--ap-username'}
5227 eqre
= re
.compile('^(?P<key>' + ('|'.join(re
.escape(po
) for po
in PRIVATE_OPTS
)) + ')=.+$')
5232 return m
.group('key') + '=PRIVATE'
5236 opts
= list(map(_scrub_eq
, opts
))
5237 for idx
, opt
in enumerate(opts
):
5238 if opt
in PRIVATE_OPTS
and idx
+ 1 < len(opts
):
5239 opts
[idx
+ 1] = 'PRIVATE'
5242 def append_config(self
, *args
, label
=None):
5243 config
= type(self
)(self
._parser
, label
)
5244 config
._loaded
_paths
= self
._loaded
_paths
5245 if config
.init(*args
):
5246 self
.configs
.append(config
)
5250 for config
in reversed(self
.configs
):
5251 yield from config
.all_args
5252 yield from self
.own_args
or []
5254 def parse_args(self
):
5255 return self
._parser
.parse_args(list(self
.all_args
))
5258 class WebSocketsWrapper():
5259 """Wraps websockets module to use in non-async scopes"""
5261 def __init__(self
, url
, headers
=None, connect
=True):
5262 self
.loop
= asyncio
.events
.new_event_loop()
5263 self
.conn
= compat_websockets
.connect(
5264 url
, extra_headers
=headers
, ping_interval
=None,
5265 close_timeout
=float('inf'), loop
=self
.loop
, ping_timeout
=float('inf'))
5268 atexit
.register(self
.__exit
__, None, None, None)
5270 def __enter__(self
):
5272 self
.pool
= self
.run_with_loop(self
.conn
.__aenter
__(), self
.loop
)
5275 def send(self
, *args
):
5276 self
.run_with_loop(self
.pool
.send(*args
), self
.loop
)
5278 def recv(self
, *args
):
5279 return self
.run_with_loop(self
.pool
.recv(*args
), self
.loop
)
5281 def __exit__(self
, type, value
, traceback
):
5283 return self
.run_with_loop(self
.conn
.__aexit
__(type, value
, traceback
), self
.loop
)
5286 self
._cancel
_all
_tasks
(self
.loop
)
5288 # taken from https://github.com/python/cpython/blob/3.9/Lib/asyncio/runners.py with modifications
5289 # for contributors: If there's any new library using asyncio needs to be run in non-async, move these function out of this class
5291 def run_with_loop(main
, loop
):
5292 if not asyncio
.coroutines
.iscoroutine(main
):
5293 raise ValueError(f
'a coroutine was expected, got {main!r}')
5296 return loop
.run_until_complete(main
)
5298 loop
.run_until_complete(loop
.shutdown_asyncgens())
5299 if hasattr(loop
, 'shutdown_default_executor'):
5300 loop
.run_until_complete(loop
.shutdown_default_executor())
5303 def _cancel_all_tasks(loop
):
5304 to_cancel
= asyncio
.tasks
.all_tasks(loop
)
5309 for task
in to_cancel
:
5312 loop
.run_until_complete(
5313 asyncio
.tasks
.gather(*to_cancel
, loop
=loop
, return_exceptions
=True))
5315 for task
in to_cancel
:
5316 if task
.cancelled():
5318 if task
.exception() is not None:
5319 loop
.call_exception_handler({
5320 'message': 'unhandled exception during asyncio.run() shutdown',
5321 'exception': task
.exception(),
5326 has_websockets
= bool(compat_websockets
)
5329 def merge_headers(*dicts
):
5330 """Merge dicts of http headers case insensitively, prioritizing the latter ones"""
5331 return {k.title(): v for k, v in itertools.chain.from_iterable(map(dict.items, dicts))}
5334 class classproperty
:
5335 def __init__(self
, f
):
5338 def __get__(self
, _
, cls
):