4 from __future__
import unicode_literals
41 import xml
.etree
.ElementTree
46 compat_HTMLParseError
,
53 compat_etree_fromstring
,
56 compat_html_entities_html5
,
70 compat_urllib_parse_urlencode
,
71 compat_urllib_parse_urlparse
,
72 compat_urllib_parse_urlunparse
,
73 compat_urllib_parse_quote
,
74 compat_urllib_parse_quote_plus
,
75 compat_urllib_parse_unquote_plus
,
76 compat_urllib_request
,
94 def register_socks_protocols():
95 # "Register" SOCKS protocols
96 # In Python < 2.6.5, urlsplit() suffers from bug https://bugs.python.org/issue7904
97 # URLs with protocols not in urlparse.uses_netloc are not handled correctly
98 for scheme
in ('socks', 'socks4', 'socks4a', 'socks5'):
99 if scheme
not in compat_urlparse
.uses_netloc
:
100 compat_urlparse
.uses_netloc
.append(scheme
)
103 # This is not clearly defined otherwise
104 compiled_regex_type
= type(re
.compile(''))
107 def random_user_agent():
108 _USER_AGENT_TPL
= 'Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/%s Safari/537.36'
149 return _USER_AGENT_TPL
% random
.choice(_CHROME_VERSIONS
)
152 SUPPORTED_ENCODINGS
= [
156 SUPPORTED_ENCODINGS
.append('br')
159 'User-Agent': random_user_agent(),
160 'Accept': 'text/html,application/xhtml+xml,application/xml;q=0.9,*/*;q=0.8',
161 'Accept-Language': 'en-us,en;q=0.5',
162 'Sec-Fetch-Mode': 'navigate',
167 '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',
171 NO_DEFAULT
= object()
173 ENGLISH_MONTH_NAMES
= [
174 'January', 'February', 'March', 'April', 'May', 'June',
175 'July', 'August', 'September', 'October', 'November', 'December']
178 'en': ENGLISH_MONTH_NAMES
,
180 'janvier', 'février', 'mars', 'avril', 'mai', 'juin',
181 'juillet', 'août', 'septembre', 'octobre', 'novembre', 'décembre'],
185 'mp4', 'm4a', 'm4p', 'm4b', 'm4r', 'm4v', 'aac',
186 'flv', 'f4v', 'f4a', 'f4b',
187 'webm', 'ogg', 'ogv', 'oga', 'ogx', 'spx', 'opus',
188 'mkv', 'mka', 'mk3d',
197 'f4f', 'f4m', 'm3u8', 'smil')
199 # needed for sanitizing filenames in restricted mode
200 ACCENT_CHARS
= dict(zip('ÂÃÄÀÁÅÆÇÈÉÊËÌÍÎÏÐÑÒÓÔÕÖŐØŒÙÚÛÜŰÝÞßàáâãäåæçèéêëìíîïðñòóôõöőøœùúûüűýþÿ',
201 itertools
.chain('AAAAAA', ['AE'], 'CEEEEIIIIDNOOOOOOO', ['OE'], 'UUUUUY', ['TH', 'ss'],
202 'aaaaaa', ['ae'], 'ceeeeiiiionooooooo', ['oe'], 'uuuuuy', ['th'], 'y')))
232 '%Y-%m-%d %H:%M:%S.%f',
233 '%Y-%m-%d %H:%M:%S:%f',
236 '%Y-%m-%dT%H:%M:%SZ',
237 '%Y-%m-%dT%H:%M:%S.%fZ',
238 '%Y-%m-%dT%H:%M:%S.%f0Z',
240 '%Y-%m-%dT%H:%M:%S.%f',
243 '%b %d %Y at %H:%M:%S',
245 '%B %d %Y at %H:%M:%S',
249 DATE_FORMATS_DAY_FIRST
= list(DATE_FORMATS
)
250 DATE_FORMATS_DAY_FIRST
.extend([
259 DATE_FORMATS_MONTH_FIRST
= list(DATE_FORMATS
)
260 DATE_FORMATS_MONTH_FIRST
.extend([
268 PACKED_CODES_RE
= r
"}\('(.+)',(\d+),(\d+),'([^']+)'\.split\('\|'\)"
269 JSON_LD_RE
= r
'(?is)<script[^>]+type=(["\']?
)application
/ld\
+json\
1[^
>]*>(?P
<json_ld
>.+?
)</script
>'
272 def preferredencoding():
273 """Get preferred encoding.
275 Returns the best encoding scheme for the system, based on
276 locale.getpreferredencoding() and some further tweaks.
279 pref = locale.getpreferredencoding()
287 def write_json_file(obj, fn):
288 """ Encode obj as JSON and write it to fn, atomically if possible """
290 tf = tempfile.NamedTemporaryFile(
291 prefix=f'{os.path.basename(fn)}
.', dir=os.path.dirname(fn),
292 suffix='.tmp
', delete=False, mode='w
', encoding='utf
-8')
296 json.dump(obj, tf, ensure_ascii=False)
297 if sys.platform == 'win32
':
298 # Need to remove existing file on Windows, else os.rename raises
299 # WindowsError or FileExistsError.
307 os.chmod(tf.name, 0o666 & ~mask)
310 os.rename(tf.name, fn)
319 def find_xpath_attr(node, xpath, key, val=None):
320 """ Find the xpath xpath[@key=val] """
321 assert re.match(r'^
[a
-zA
-Z_
-]+$
', key)
322 expr = xpath + ('[@%s]' % key if val is None else "[@%s='%s']" % (key, val))
323 return node.find(expr)
325 # On python2.6 the xml.etree.ElementTree.Element methods don't support
326 # the namespace parameter
329 def xpath_with_ns(path
, ns_map
):
330 components
= [c
.split(':') for c
in path
.split('/')]
334 replaced
.append(c
[0])
337 replaced
.append('{%s}%s' % (ns_map
[ns
], tag
))
338 return '/'.join(replaced
)
341 def xpath_element(node
, xpath
, name
=None, fatal
=False, default
=NO_DEFAULT
):
342 def _find_xpath(xpath
):
343 return node
.find(compat_xpath(xpath
))
345 if isinstance(xpath
, (str, compat_str
)):
346 n
= _find_xpath(xpath
)
354 if default
is not NO_DEFAULT
:
357 name
= xpath
if name
is None else name
358 raise ExtractorError('Could not find XML element %s' % name
)
364 def xpath_text(node
, xpath
, name
=None, fatal
=False, default
=NO_DEFAULT
):
365 n
= xpath_element(node
, xpath
, name
, fatal
=fatal
, default
=default
)
366 if n
is None or n
== default
:
369 if default
is not NO_DEFAULT
:
372 name
= xpath
if name
is None else name
373 raise ExtractorError('Could not find XML element\'s text %s' % name
)
379 def xpath_attr(node
, xpath
, key
, name
=None, fatal
=False, default
=NO_DEFAULT
):
380 n
= find_xpath_attr(node
, xpath
, key
)
382 if default
is not NO_DEFAULT
:
385 name
= '%s[@%s]' % (xpath
, key
) if name
is None else name
386 raise ExtractorError('Could not find XML attribute %s' % name
)
392 def get_element_by_id(id, html
):
393 """Return the content of the tag with the specified ID in the passed HTML document"""
394 return get_element_by_attribute('id', id, html
)
397 def get_element_html_by_id(id, html
):
398 """Return the html of the tag with the specified ID in the passed HTML document"""
399 return get_element_html_by_attribute('id', id, html
)
402 def get_element_by_class(class_name
, html
):
403 """Return the content of the first tag with the specified class in the passed HTML document"""
404 retval
= get_elements_by_class(class_name
, html
)
405 return retval
[0] if retval
else None
408 def get_element_html_by_class(class_name
, html
):
409 """Return the html of the first tag with the specified class in the passed HTML document"""
410 retval
= get_elements_html_by_class(class_name
, html
)
411 return retval
[0] if retval
else None
414 def get_element_by_attribute(attribute
, value
, html
, escape_value
=True):
415 retval
= get_elements_by_attribute(attribute
, value
, html
, escape_value
)
416 return retval
[0] if retval
else None
419 def get_element_html_by_attribute(attribute
, value
, html
, escape_value
=True):
420 retval
= get_elements_html_by_attribute(attribute
, value
, html
, escape_value
)
421 return retval
[0] if retval
else None
424 def get_elements_by_class(class_name
, html
):
425 """Return the content of all tags with the specified class in the passed HTML document as a list"""
426 return get_elements_by_attribute(
427 'class', r
'[^\'"]*\b%s\b[^\'"]*' % re.escape(class_name),
428 html, escape_value=False)
431 def get_elements_html_by_class(class_name, html):
432 """Return the html of all tags with the specified class in the passed HTML document as a list"""
433 return get_elements_html_by_attribute(
434 'class', r'[^
\'"]*\b%s\b[^\'"]*' % re.escape(class_name),
435 html, escape_value=False)
438 def get_elements_by_attribute(*args, **kwargs):
439 """Return the content of the tag with the specified attribute in the passed HTML document"""
440 return [content for content, _ in get_elements_text_and_html_by_attribute(*args, **kwargs)]
443 def get_elements_html_by_attribute(*args, **kwargs):
444 """Return the html of the tag with the specified attribute in the passed HTML document"""
445 return [whole for _, whole in get_elements_text_and_html_by_attribute(*args, **kwargs)]
448 def get_elements_text_and_html_by_attribute(attribute, value, html, escape_value=True):
450 Return the text (content) and the html (whole) of the tag with the specified
451 attribute in the passed HTML document
454 value_quote_optional = '' if re.match(r'''[\s"'`
=<>]''', value) else '?'
456 value = re.escape(value) if escape_value else value
458 partial_element_re = r'''(?x
)
459 <(?P
<tag
>[a
-zA
-Z0
-9:._-]+)
460 (?
:\
s(?
:[^
>"']|"[^
"]*"|
'[^']*')*)?
461 \s%(attribute)s\s*=\s*(?P<_q>['"]%(vqo)s)(?-x:%(value)s)(?P=_q)
462 ''' % {'attribute': re.escape(attribute), 'value': value, 'vqo': value_quote_optional}
464 for m in re.finditer(partial_element_re, html):
465 content, whole = get_element_text_and_html_by_tag(m.group('tag'), html[m.start():])
468 unescapeHTML(re.sub(r'^(?P<q>["\'])(?P
<content
>.*)(?P
=q
)$
', r'\g
<content
>', content, flags=re.DOTALL)),
473 class HTMLBreakOnClosingTagParser(compat_HTMLParser):
475 HTML parser which raises HTMLBreakOnClosingTagException upon reaching the
476 closing tag for the first opening tag it has encountered, and can be used
480 class HTMLBreakOnClosingTagException(Exception):
484 self.tagstack = collections.deque()
485 compat_HTMLParser.__init__(self)
490 def __exit__(self, *_):
494 # handle_endtag does not return upon raising HTMLBreakOnClosingTagException,
495 # so data remains buffered; we no longer have any interest in it, thus
496 # override this method to discard it
499 def handle_starttag(self, tag, _):
500 self.tagstack.append(tag)
502 def handle_endtag(self, tag):
503 if not self.tagstack:
504 raise compat_HTMLParseError('no tags
in the stack
')
506 inner_tag = self.tagstack.pop()
510 raise compat_HTMLParseError(f'matching opening tag
for closing {tag} tag
not found
')
511 if not self.tagstack:
512 raise self.HTMLBreakOnClosingTagException()
515 def get_element_text_and_html_by_tag(tag, html):
517 For the first element with the specified tag in the passed HTML document
518 return its' content (text
) and the whole
element (html
)
520 def find_or_raise(haystack, needle, exc):
522 return haystack.index(needle)
525 closing_tag = f'</{tag}>'
526 whole_start = find_or_raise(
527 html, f'<{tag}', compat_HTMLParseError(f'opening {tag} tag not found'))
528 content_start = find_or_raise(
529 html[whole_start:], '>', compat_HTMLParseError(f'malformed opening {tag} tag'))
530 content_start += whole_start + 1
531 with HTMLBreakOnClosingTagParser() as parser:
532 parser.feed(html[whole_start:content_start])
533 if not parser.tagstack or parser.tagstack[0] != tag:
534 raise compat_HTMLParseError(f'parser did not match opening {tag} tag')
535 offset = content_start
536 while offset < len(html):
537 next_closing_tag_start = find_or_raise(
538 html[offset:], closing_tag,
539 compat_HTMLParseError(f'closing {tag} tag not found'))
540 next_closing_tag_end = next_closing_tag_start + len(closing_tag)
542 parser.feed(html[offset:offset + next_closing_tag_end])
543 offset += next_closing_tag_end
544 except HTMLBreakOnClosingTagParser.HTMLBreakOnClosingTagException:
545 return html[content_start:offset + next_closing_tag_start], \
546 html[whole_start:offset + next_closing_tag_end]
547 raise compat_HTMLParseError('unexpected end of html')
550 class HTMLAttributeParser(compat_HTMLParser):
551 """Trivial HTML parser to gather the attributes
for a single element
"""
555 compat_HTMLParser.__init__(self)
557 def handle_starttag(self, tag, attrs):
558 self.attrs = dict(attrs)
561 class HTMLListAttrsParser(compat_HTMLParser):
562 """HTML parser to gather the attributes
for the elements of a
list"""
565 compat_HTMLParser.__init__(self)
569 def handle_starttag(self, tag, attrs):
570 if tag == 'li' and self._level == 0:
571 self.items.append(dict(attrs))
574 def handle_endtag(self, tag):
578 def extract_attributes(html_element):
579 """Given a string
for an HTML element such
as
581 a
="foo" B
="bar" c
="&98;az" d
=boz
582 empty
= noval entity
="&"
585 Decode
and return a dictionary of attributes
.
587 'a': 'foo', 'b': 'bar', c
: 'baz', d
: 'boz',
588 'empty': '', 'noval': None, 'entity': '&',
589 'sq': '"', 'dq': '\''
592 parser = HTMLAttributeParser()
594 parser.feed(html_element)
596 # Older Python may throw HTMLParseError in case of malformed HTML
597 except compat_HTMLParseError:
602 def parse_list(webpage):
603 """Given a string
for an series of HTML
<li
> elements
,
604 return a dictionary of their attributes
"""
605 parser = HTMLListAttrsParser()
611 def clean_html(html):
612 """Clean an HTML snippet into a readable string
"""
614 if html is None: # Convenience for sanitizing descriptions etc.
617 html = re.sub(r'\s+', ' ', html)
618 html = re.sub(r'(?u)\s?<\s?br\s?/?\s?>\s?', '\n', html)
619 html = re.sub(r'(?u)<\s?/\s?p\s?>\s?<\s?p[^>]*>', '\n', html)
621 html = re.sub('<.*?>', '', html)
622 # Replace html entities
623 html = unescapeHTML(html)
627 def sanitize_open(filename, open_mode):
628 """Try to
open the given filename
, and slightly tweak it
if this fails
.
630 Attempts to
open the given filename
. If this fails
, it tries to change
631 the filename slightly
, step by step
, until it
's either able to open it
632 or it fails and raises a final exception, like the standard open()
635 It returns the tuple (stream, definitive_file_name).
638 if sys.platform == 'win32
':
640 msvcrt.setmode(sys.stdout.fileno(), os.O_BINARY)
641 return (sys.stdout.buffer if hasattr(sys.stdout, 'buffer') else sys.stdout, filename)
643 for attempt in range(2):
646 if sys.platform == 'win32
':
647 # FIXME: An exclusive lock also locks the file from being read.
648 # Since windows locks are mandatory, don't lock the
file on
windows (for now
).
649 # Ref: https://github.com/yt-dlp/yt-dlp/issues/3124
650 raise LockingUnsupportedError()
651 stream
= locked_file(filename
, open_mode
, block
=False).__enter
__()
652 except LockingUnsupportedError
:
653 stream
= open(filename
, open_mode
)
654 return (stream
, filename
)
655 except (IOError, OSError) as err
:
656 if attempt
or err
.errno
in (errno
.EACCES
,):
658 old_filename
, filename
= filename
, sanitize_path(filename
)
659 if old_filename
== filename
:
663 def timeconvert(timestr
):
664 """Convert RFC 2822 defined time string into system timestamp"""
666 timetuple
= email
.utils
.parsedate_tz(timestr
)
667 if timetuple
is not None:
668 timestamp
= email
.utils
.mktime_tz(timetuple
)
672 def sanitize_filename(s
, restricted
=False, is_id
=NO_DEFAULT
):
673 """Sanitizes a string so it could be used as part of a filename.
674 @param restricted Use a stricter subset of allowed characters
675 @param is_id Whether this is an ID that should be kept unchanged if possible.
676 If unset, yt-dlp's new sanitization rules are in effect
681 def replace_insane(char
):
682 if restricted
and char
in ACCENT_CHARS
:
683 return ACCENT_CHARS
[char
]
684 elif not restricted
and char
== '\n':
686 elif char
== '?' or ord(char
) < 32 or ord(char
) == 127:
689 return '' if restricted
else '\''
691 return '\0_\0-' if restricted
else '\0 \0-'
692 elif char
in '\\/|*<>':
694 if restricted
and (char
in '!&\'()[]{}$;`^,#' or char
.isspace() or ord(char
) > 127):
698 s
= re
.sub(r
'[0-9]+(?::[0-9]+)+', lambda m
: m
.group(0).replace(':', '_'), s
) # Handle timestamps
699 result
= ''.join(map(replace_insane
, s
))
700 if is_id
is NO_DEFAULT
:
701 result
= re
.sub('(\0.)(?:(?=\\1)..)+', r
'\1', result
) # Remove repeated substitute chars
702 STRIP_RE
= '(?:\0.|[ _-])*'
703 result
= re
.sub(f
'^\0.{STRIP_RE}|{STRIP_RE}\0.$', '', result
) # Remove substitute chars from start/end
704 result
= result
.replace('\0', '') or '_'
707 while '__' in result
:
708 result
= result
.replace('__', '_')
709 result
= result
.strip('_')
710 # Common case of "Foreign band name - English song title"
711 if restricted
and result
.startswith('-_'):
713 if result
.startswith('-'):
714 result
= '_' + result
[len('-'):]
715 result
= result
.lstrip('.')
721 def sanitize_path(s
, force
=False):
722 """Sanitizes and normalizes path on Windows"""
723 if sys
.platform
== 'win32':
725 drive_or_unc
, _
= os
.path
.splitdrive(s
)
731 norm_path
= os
.path
.normpath(remove_start(s
, drive_or_unc
)).split(os
.path
.sep
)
735 path_part
if path_part
in ['.', '..'] else re
.sub(r
'(?:[/<>:"\|\\?\*]|[\s.]$)', '#', path_part
)
736 for path_part
in norm_path
]
738 sanitized_path
.insert(0, drive_or_unc
+ os
.path
.sep
)
739 elif force
and s
and s
[0] == os
.path
.sep
:
740 sanitized_path
.insert(0, os
.path
.sep
)
741 return os
.path
.join(*sanitized_path
)
744 def sanitize_url(url
):
745 # Prepend protocol-less URLs with `http:` scheme in order to mitigate
746 # the number of unwanted failures due to missing protocol
747 if url
.startswith('//'):
748 return 'http:%s' % url
749 # Fix some common typos seen so far
751 # https://github.com/ytdl-org/youtube-dl/issues/15649
752 (r
'^httpss://', r
'https://'),
753 # https://bx1.be/lives/direct-tv/
754 (r
'^rmtp([es]?)://', r
'rtmp\1://'),
756 for mistake
, fixup
in COMMON_TYPOS
:
757 if re
.match(mistake
, url
):
758 return re
.sub(mistake
, fixup
, url
)
762 def extract_basic_auth(url
):
763 parts
= compat_urlparse
.urlsplit(url
)
764 if parts
.username
is None:
766 url
= compat_urlparse
.urlunsplit(parts
._replace
(netloc
=(
767 parts
.hostname
if parts
.port
is None
768 else '%s:%d' % (parts
.hostname
, parts
.port
))))
769 auth_payload
= base64
.b64encode(
770 ('%s:%s' % (parts
.username
, parts
.password
or '')).encode('utf-8'))
771 return url
, 'Basic ' + auth_payload
.decode('utf-8')
774 def sanitized_Request(url
, *args
, **kwargs
):
775 url
, auth_header
= extract_basic_auth(escape_url(sanitize_url(url
)))
776 if auth_header
is not None:
777 headers
= args
[1] if len(args
) >= 2 else kwargs
.setdefault('headers', {})
778 headers
['Authorization'] = auth_header
779 return compat_urllib_request
.Request(url
, *args
, **kwargs
)
783 """Expand shell variables and ~"""
784 return os
.path
.expandvars(compat_expanduser(s
))
787 def orderedSet(iterable
):
788 """ Remove all duplicates from the input iterable """
796 def _htmlentity_transform(entity_with_semicolon
):
797 """Transforms an HTML entity to a character."""
798 entity
= entity_with_semicolon
[:-1]
800 # Known non-numeric HTML entity
801 if entity
in compat_html_entities
.name2codepoint
:
802 return compat_chr(compat_html_entities
.name2codepoint
[entity
])
804 # TODO: HTML5 allows entities without a semicolon. For example,
805 # 'Éric' should be decoded as 'Éric'.
806 if entity_with_semicolon
in compat_html_entities_html5
:
807 return compat_html_entities_html5
[entity_with_semicolon
]
809 mobj
= re
.match(r
'#(x[0-9a-fA-F]+|[0-9]+)', entity
)
811 numstr
= mobj
.group(1)
812 if numstr
.startswith('x'):
814 numstr
= '0%s' % numstr
817 # See https://github.com/ytdl-org/youtube-dl/issues/7518
819 return compat_chr(int(numstr
, base
))
823 # Unknown entity in name, return its literal representation
824 return '&%s;' % entity
830 assert type(s
) == compat_str
833 r
'&([^&;]+;)', lambda m
: _htmlentity_transform(m
.group(1)), s
)
836 def escapeHTML(text
):
839 .replace('&', '&')
840 .replace('<', '<')
841 .replace('>', '>')
842 .replace('"', '"')
843 .replace("'", ''')
847 def process_communicate_or_kill(p
, *args
, **kwargs
):
849 return p
.communicate(*args
, **kwargs
)
850 except BaseException
: # Including KeyboardInterrupt
856 class Popen(subprocess
.Popen
):
857 if sys
.platform
== 'win32':
858 _startupinfo
= subprocess
.STARTUPINFO()
859 _startupinfo
.dwFlags |
= subprocess
.STARTF_USESHOWWINDOW
863 def __init__(self
, *args
, **kwargs
):
864 super(Popen
, self
).__init
__(*args
, **kwargs
, startupinfo
=self
._startupinfo
)
866 def communicate_or_kill(self
, *args
, **kwargs
):
867 return process_communicate_or_kill(self
, *args
, **kwargs
)
870 def get_subprocess_encoding():
871 if sys
.platform
== 'win32' and sys
.getwindowsversion()[0] >= 5:
872 # For subprocess calls, encode with locale encoding
873 # Refer to http://stackoverflow.com/a/9951851/35070
874 encoding
= preferredencoding()
876 encoding
= sys
.getfilesystemencoding()
882 def encodeFilename(s
, for_subprocess
=False):
883 assert type(s
) == str
887 def decodeFilename(b
, for_subprocess
=False):
891 def encodeArgument(s
):
892 # Legacy code that uses byte strings
893 # Uncomment the following line after fixing all post processors
894 # assert isinstance(s, str), 'Internal error: %r should be of type %r, is %r' % (s, compat_str, type(s))
895 return s
if isinstance(s
, str) else s
.decode('ascii')
898 def decodeArgument(b
):
902 def decodeOption(optval
):
905 if isinstance(optval
, bytes):
906 optval
= optval
.decode(preferredencoding())
908 assert isinstance(optval
, compat_str
)
912 _timetuple
= collections
.namedtuple('Time', ('hours', 'minutes', 'seconds', 'milliseconds'))
915 def timetuple_from_msec(msec
):
916 secs
, msec
= divmod(msec
, 1000)
917 mins
, secs
= divmod(secs
, 60)
918 hrs
, mins
= divmod(mins
, 60)
919 return _timetuple(hrs
, mins
, secs
, msec
)
922 def formatSeconds(secs
, delim
=':', msec
=False):
923 time
= timetuple_from_msec(secs
* 1000)
925 ret
= '%d%s%02d%s%02d' % (time
.hours
, delim
, time
.minutes
, delim
, time
.seconds
)
927 ret
= '%d%s%02d' % (time
.minutes
, delim
, time
.seconds
)
929 ret
= '%d' % time
.seconds
930 return '%s.%03d' % (ret
, time
.milliseconds
) if msec
else ret
933 def _ssl_load_windows_store_certs(ssl_context
, storename
):
934 # Code adapted from _load_windows_store_certs in https://github.com/python/cpython/blob/main/Lib/ssl.py
936 certs
= [cert
for cert
, encoding
, trust
in ssl
.enum_certificates(storename
)
937 if encoding
== 'x509_asn' and (
938 trust
is True or ssl
.Purpose
.SERVER_AUTH
.oid
in trust
)]
939 except PermissionError
:
943 ssl_context
.load_verify_locations(cadata
=cert
)
948 def make_HTTPS_handler(params
, **kwargs
):
949 opts_check_certificate
= not params
.get('nocheckcertificate')
950 context
= ssl
.SSLContext(ssl
.PROTOCOL_TLS_CLIENT
)
951 context
.check_hostname
= opts_check_certificate
952 if params
.get('legacyserverconnect'):
953 context
.options |
= 4 # SSL_OP_LEGACY_SERVER_CONNECT
954 context
.verify_mode
= ssl
.CERT_REQUIRED
if opts_check_certificate
else ssl
.CERT_NONE
955 if opts_check_certificate
:
956 if has_certifi
and 'no-certifi' not in params
.get('compat_opts', []):
957 context
.load_verify_locations(cafile
=certifi
.where())
960 context
.load_default_certs()
961 # Work around the issue in load_default_certs when there are bad certificates. See:
962 # https://github.com/yt-dlp/yt-dlp/issues/1060,
963 # https://bugs.python.org/issue35665, https://bugs.python.org/issue45312
965 # enum_certificates is not present in mingw python. See https://github.com/yt-dlp/yt-dlp/issues/1151
966 if sys
.platform
== 'win32' and hasattr(ssl
, 'enum_certificates'):
967 # Create a new context to discard any certificates that were already loaded
968 context
= ssl
.SSLContext(ssl
.PROTOCOL_TLS_CLIENT
)
969 context
.check_hostname
, context
.verify_mode
= True, ssl
.CERT_REQUIRED
970 for storename
in ('CA', 'ROOT'):
971 _ssl_load_windows_store_certs(context
, storename
)
972 context
.set_default_verify_paths()
973 return YoutubeDLHTTPSHandler(params
, context
=context
, **kwargs
)
976 def bug_reports_message(before
=';'):
977 msg
= ('please report this issue on https://github.com/yt-dlp/yt-dlp/issues?q= , '
978 'filling out the appropriate issue template. '
979 'Confirm you are on the latest version using yt-dlp -U')
981 before
= before
.rstrip()
982 if not before
or before
.endswith(('.', '!', '?')):
983 msg
= msg
[0].title() + msg
[1:]
985 return (before
+ ' ' if before
else '') + msg
988 class YoutubeDLError(Exception):
989 """Base exception for YoutubeDL errors."""
992 def __init__(self
, msg
=None):
995 elif self
.msg
is None:
996 self
.msg
= type(self
).__name
__
997 super().__init
__(self
.msg
)
1000 network_exceptions
= [compat_urllib_error
.URLError
, compat_http_client
.HTTPException
, socket
.error
]
1001 if hasattr(ssl
, 'CertificateError'):
1002 network_exceptions
.append(ssl
.CertificateError
)
1003 network_exceptions
= tuple(network_exceptions
)
1006 class ExtractorError(YoutubeDLError
):
1007 """Error during info extraction."""
1009 def __init__(self
, msg
, tb
=None, expected
=False, cause
=None, video_id
=None, ie
=None):
1010 """ tb, if given, is the original traceback (so that it can be printed out).
1011 If expected is set, this is a normal error message and most likely not a bug in yt-dlp.
1013 if sys
.exc_info()[0] in network_exceptions
:
1016 self
.orig_msg
= str(msg
)
1018 self
.expected
= expected
1020 self
.video_id
= video_id
1022 self
.exc_info
= sys
.exc_info() # preserve original exception
1024 super(ExtractorError
, self
).__init
__(''.join((
1025 format_field(ie
, template
='[%s] '),
1026 format_field(video_id
, template
='%s: '),
1028 format_field(cause
, template
=' (caused by %r)'),
1029 '' if expected
else bug_reports_message())))
1031 def format_traceback(self
):
1032 return join_nonempty(
1033 self
.traceback
and ''.join(traceback
.format_tb(self
.traceback
)),
1034 self
.cause
and ''.join(traceback
.format_exception(None, self
.cause
, self
.cause
.__traceback
__)[1:]),
1038 class UnsupportedError(ExtractorError
):
1039 def __init__(self
, url
):
1040 super(UnsupportedError
, self
).__init
__(
1041 'Unsupported URL: %s' % url
, expected
=True)
1045 class RegexNotFoundError(ExtractorError
):
1046 """Error when a regex didn't match"""
1050 class GeoRestrictedError(ExtractorError
):
1051 """Geographic restriction Error exception.
1053 This exception may be thrown when a video is not available from your
1054 geographic location due to geographic restrictions imposed by a website.
1057 def __init__(self
, msg
, countries
=None, **kwargs
):
1058 kwargs
['expected'] = True
1059 super(GeoRestrictedError
, self
).__init
__(msg
, **kwargs
)
1060 self
.countries
= countries
1063 class DownloadError(YoutubeDLError
):
1064 """Download Error exception.
1066 This exception may be thrown by FileDownloader objects if they are not
1067 configured to continue on errors. They will contain the appropriate
1071 def __init__(self
, msg
, exc_info
=None):
1072 """ exc_info, if given, is the original exception that caused the trouble (as returned by sys.exc_info()). """
1073 super(DownloadError
, self
).__init
__(msg
)
1074 self
.exc_info
= exc_info
1077 class EntryNotInPlaylist(YoutubeDLError
):
1078 """Entry not in playlist exception.
1080 This exception will be thrown by YoutubeDL when a requested entry
1081 is not found in the playlist info_dict
1083 msg
= 'Entry not found in info'
1086 class SameFileError(YoutubeDLError
):
1087 """Same File exception.
1089 This exception will be thrown by FileDownloader objects if they detect
1090 multiple files would have to be downloaded to the same file on disk.
1092 msg
= 'Fixed output name but more than one file to download'
1094 def __init__(self
, filename
=None):
1095 if filename
is not None:
1096 self
.msg
+= f
': {filename}'
1097 super().__init
__(self
.msg
)
1100 class PostProcessingError(YoutubeDLError
):
1101 """Post Processing exception.
1103 This exception may be raised by PostProcessor's .run() method to
1104 indicate an error in the postprocessing task.
1108 class DownloadCancelled(YoutubeDLError
):
1109 """ Exception raised when the download queue should be interrupted """
1110 msg
= 'The download was cancelled'
1113 class ExistingVideoReached(DownloadCancelled
):
1114 """ --break-on-existing triggered """
1115 msg
= 'Encountered a video that is already in the archive, stopping due to --break-on-existing'
1118 class RejectedVideoReached(DownloadCancelled
):
1119 """ --break-on-reject triggered """
1120 msg
= 'Encountered a video that did not match filter, stopping due to --break-on-reject'
1123 class MaxDownloadsReached(DownloadCancelled
):
1124 """ --max-downloads limit has been reached. """
1125 msg
= 'Maximum number of downloads reached, stopping due to --max-downloads'
1128 class ReExtractInfo(YoutubeDLError
):
1129 """ Video info needs to be re-extracted. """
1131 def __init__(self
, msg
, expected
=False):
1132 super().__init
__(msg
)
1133 self
.expected
= expected
1136 class ThrottledDownload(ReExtractInfo
):
1137 """ Download speed below --throttled-rate. """
1138 msg
= 'The download speed is below throttle limit'
1141 super().__init
__(self
.msg
, expected
=False)
1144 class UnavailableVideoError(YoutubeDLError
):
1145 """Unavailable Format exception.
1147 This exception will be thrown when a video is requested
1148 in a format that is not available for that video.
1150 msg
= 'Unable to download video'
1152 def __init__(self
, err
=None):
1154 self
.msg
+= f
': {err}'
1155 super().__init
__(self
.msg
)
1158 class ContentTooShortError(YoutubeDLError
):
1159 """Content Too Short exception.
1161 This exception may be raised by FileDownloader objects when a file they
1162 download is too small for what the server announced first, indicating
1163 the connection was probably interrupted.
1166 def __init__(self
, downloaded
, expected
):
1167 super(ContentTooShortError
, self
).__init
__(
1168 'Downloaded {0} bytes, expected {1} bytes'.format(downloaded
, expected
)
1171 self
.downloaded
= downloaded
1172 self
.expected
= expected
1175 class XAttrMetadataError(YoutubeDLError
):
1176 def __init__(self
, code
=None, msg
='Unknown error'):
1177 super(XAttrMetadataError
, self
).__init
__(msg
)
1181 # Parsing code and msg
1182 if (self
.code
in (errno
.ENOSPC
, errno
.EDQUOT
)
1183 or 'No space left' in self
.msg
or 'Disk quota exceeded' in self
.msg
):
1184 self
.reason
= 'NO_SPACE'
1185 elif self
.code
== errno
.E2BIG
or 'Argument list too long' in self
.msg
:
1186 self
.reason
= 'VALUE_TOO_LONG'
1188 self
.reason
= 'NOT_SUPPORTED'
1191 class XAttrUnavailableError(YoutubeDLError
):
1195 def _create_http_connection(ydl_handler
, http_class
, is_https
, *args
, **kwargs
):
1196 hc
= http_class(*args
, **compat_kwargs(kwargs
))
1197 source_address
= ydl_handler
._params
.get('source_address')
1199 if source_address
is not None:
1200 # This is to workaround _create_connection() from socket where it will try all
1201 # address data from getaddrinfo() including IPv6. This filters the result from
1202 # getaddrinfo() based on the source_address value.
1203 # This is based on the cpython socket.create_connection() function.
1204 # https://github.com/python/cpython/blob/master/Lib/socket.py#L691
1205 def _create_connection(address
, timeout
=socket
._GLOBAL
_DEFAULT
_TIMEOUT
, source_address
=None):
1206 host
, port
= address
1208 addrs
= socket
.getaddrinfo(host
, port
, 0, socket
.SOCK_STREAM
)
1209 af
= socket
.AF_INET
if '.' in source_address
[0] else socket
.AF_INET6
1210 ip_addrs
= [addr
for addr
in addrs
if addr
[0] == af
]
1211 if addrs
and not ip_addrs
:
1212 ip_version
= 'v4' if af
== socket
.AF_INET
else 'v6'
1214 "No remote IP%s addresses available for connect, can't use '%s' as source address"
1215 % (ip_version
, source_address
[0]))
1216 for res
in ip_addrs
:
1217 af
, socktype
, proto
, canonname
, sa
= res
1220 sock
= socket
.socket(af
, socktype
, proto
)
1221 if timeout
is not socket
._GLOBAL
_DEFAULT
_TIMEOUT
:
1222 sock
.settimeout(timeout
)
1223 sock
.bind(source_address
)
1225 err
= None # Explicitly break reference cycle
1227 except socket
.error
as _
:
1229 if sock
is not None:
1234 raise socket
.error('getaddrinfo returns an empty list')
1235 if hasattr(hc
, '_create_connection'):
1236 hc
._create
_connection
= _create_connection
1237 hc
.source_address
= (source_address
, 0)
1242 def handle_youtubedl_headers(headers
):
1243 filtered_headers
= headers
1245 if 'Youtubedl-no-compression' in filtered_headers
:
1246 filtered_headers
= dict((k
, v
) for k
, v
in filtered_headers
.items() if k
.lower() != 'accept-encoding')
1247 del filtered_headers
['Youtubedl-no-compression']
1249 return filtered_headers
1252 class YoutubeDLHandler(compat_urllib_request
.HTTPHandler
):
1253 """Handler for HTTP requests and responses.
1255 This class, when installed with an OpenerDirector, automatically adds
1256 the standard headers to every HTTP request and handles gzipped and
1257 deflated responses from web servers. If compression is to be avoided in
1258 a particular request, the original request in the program code only has
1259 to include the HTTP header "Youtubedl-no-compression", which will be
1260 removed before making the real request.
1262 Part of this code was copied from:
1264 http://techknack.net/python-urllib2-handlers/
1266 Andrew Rowls, the author of that code, agreed to release it to the
1270 def __init__(self
, params
, *args
, **kwargs
):
1271 compat_urllib_request
.HTTPHandler
.__init
__(self
, *args
, **kwargs
)
1272 self
._params
= params
1274 def http_open(self
, req
):
1275 conn_class
= compat_http_client
.HTTPConnection
1277 socks_proxy
= req
.headers
.get('Ytdl-socks-proxy')
1279 conn_class
= make_socks_conn_class(conn_class
, socks_proxy
)
1280 del req
.headers
['Ytdl-socks-proxy']
1282 return self
.do_open(functools
.partial(
1283 _create_http_connection
, self
, conn_class
, False),
1291 return zlib
.decompress(data
, -zlib
.MAX_WBITS
)
1293 return zlib
.decompress(data
)
1299 return compat_brotli
.decompress(data
)
1301 def http_request(self
, req
):
1302 # According to RFC 3986, URLs can not contain non-ASCII characters, however this is not
1303 # always respected by websites, some tend to give out URLs with non percent-encoded
1304 # non-ASCII characters (see telemb.py, ard.py [#3412])
1305 # urllib chokes on URLs with non-ASCII characters (see http://bugs.python.org/issue3991)
1306 # To work around aforementioned issue we will replace request's original URL with
1307 # percent-encoded one
1308 # Since redirects are also affected (e.g. http://www.southpark.de/alle-episoden/s18e09)
1309 # the code of this workaround has been moved here from YoutubeDL.urlopen()
1310 url
= req
.get_full_url()
1311 url_escaped
= escape_url(url
)
1313 # Substitute URL if any change after escaping
1314 if url
!= url_escaped
:
1315 req
= update_Request(req
, url
=url_escaped
)
1317 for h
, v
in self
._params
.get('http_headers', std_headers
).items():
1318 # Capitalize is needed because of Python bug 2275: http://bugs.python.org/issue2275
1319 # The dict keys are capitalized because of this bug by urllib
1320 if h
.capitalize() not in req
.headers
:
1321 req
.add_header(h
, v
)
1323 if 'Accept-encoding' not in req
.headers
:
1324 req
.add_header('Accept-encoding', ', '.join(SUPPORTED_ENCODINGS
))
1326 req
.headers
= handle_youtubedl_headers(req
.headers
)
1330 def http_response(self
, req
, resp
):
1333 if resp
.headers
.get('Content-encoding', '') == 'gzip':
1334 content
= resp
.read()
1335 gz
= gzip
.GzipFile(fileobj
=io
.BytesIO(content
), mode
='rb')
1337 uncompressed
= io
.BytesIO(gz
.read())
1338 except IOError as original_ioerror
:
1339 # There may be junk add the end of the file
1340 # See http://stackoverflow.com/q/4928560/35070 for details
1341 for i
in range(1, 1024):
1343 gz
= gzip
.GzipFile(fileobj
=io
.BytesIO(content
[:-i
]), mode
='rb')
1344 uncompressed
= io
.BytesIO(gz
.read())
1349 raise original_ioerror
1350 resp
= compat_urllib_request
.addinfourl(uncompressed
, old_resp
.headers
, old_resp
.url
, old_resp
.code
)
1351 resp
.msg
= old_resp
.msg
1352 del resp
.headers
['Content-encoding']
1354 if resp
.headers
.get('Content-encoding', '') == 'deflate':
1355 gz
= io
.BytesIO(self
.deflate(resp
.read()))
1356 resp
= compat_urllib_request
.addinfourl(gz
, old_resp
.headers
, old_resp
.url
, old_resp
.code
)
1357 resp
.msg
= old_resp
.msg
1358 del resp
.headers
['Content-encoding']
1360 if resp
.headers
.get('Content-encoding', '') == 'br':
1361 resp
= compat_urllib_request
.addinfourl(
1362 io
.BytesIO(self
.brotli(resp
.read())), old_resp
.headers
, old_resp
.url
, old_resp
.code
)
1363 resp
.msg
= old_resp
.msg
1364 del resp
.headers
['Content-encoding']
1365 # Percent-encode redirect URL of Location HTTP header to satisfy RFC 3986 (see
1366 # https://github.com/ytdl-org/youtube-dl/issues/6457).
1367 if 300 <= resp
.code
< 400:
1368 location
= resp
.headers
.get('Location')
1370 # As of RFC 2616 default charset is iso-8859-1 that is respected by python 3
1371 location
= location
.encode('iso-8859-1').decode('utf-8')
1372 location_escaped
= escape_url(location
)
1373 if location
!= location_escaped
:
1374 del resp
.headers
['Location']
1375 resp
.headers
['Location'] = location_escaped
1378 https_request
= http_request
1379 https_response
= http_response
1382 def make_socks_conn_class(base_class
, socks_proxy
):
1383 assert issubclass(base_class
, (
1384 compat_http_client
.HTTPConnection
, compat_http_client
.HTTPSConnection
))
1386 url_components
= compat_urlparse
.urlparse(socks_proxy
)
1387 if url_components
.scheme
.lower() == 'socks5':
1388 socks_type
= ProxyType
.SOCKS5
1389 elif url_components
.scheme
.lower() in ('socks', 'socks4'):
1390 socks_type
= ProxyType
.SOCKS4
1391 elif url_components
.scheme
.lower() == 'socks4a':
1392 socks_type
= ProxyType
.SOCKS4A
1394 def unquote_if_non_empty(s
):
1397 return compat_urllib_parse_unquote_plus(s
)
1401 url_components
.hostname
, url_components
.port
or 1080,
1403 unquote_if_non_empty(url_components
.username
),
1404 unquote_if_non_empty(url_components
.password
),
1407 class SocksConnection(base_class
):
1409 self
.sock
= sockssocket()
1410 self
.sock
.setproxy(*proxy_args
)
1411 if type(self
.timeout
) in (int, float):
1412 self
.sock
.settimeout(self
.timeout
)
1413 self
.sock
.connect((self
.host
, self
.port
))
1415 if isinstance(self
, compat_http_client
.HTTPSConnection
):
1416 if hasattr(self
, '_context'): # Python > 2.6
1417 self
.sock
= self
._context
.wrap_socket(
1418 self
.sock
, server_hostname
=self
.host
)
1420 self
.sock
= ssl
.wrap_socket(self
.sock
)
1422 return SocksConnection
1425 class YoutubeDLHTTPSHandler(compat_urllib_request
.HTTPSHandler
):
1426 def __init__(self
, params
, https_conn_class
=None, *args
, **kwargs
):
1427 compat_urllib_request
.HTTPSHandler
.__init
__(self
, *args
, **kwargs
)
1428 self
._https
_conn
_class
= https_conn_class
or compat_http_client
.HTTPSConnection
1429 self
._params
= params
1431 def https_open(self
, req
):
1433 conn_class
= self
._https
_conn
_class
1435 if hasattr(self
, '_context'): # python > 2.6
1436 kwargs
['context'] = self
._context
1437 if hasattr(self
, '_check_hostname'): # python 3.x
1438 kwargs
['check_hostname'] = self
._check
_hostname
1440 socks_proxy
= req
.headers
.get('Ytdl-socks-proxy')
1442 conn_class
= make_socks_conn_class(conn_class
, socks_proxy
)
1443 del req
.headers
['Ytdl-socks-proxy']
1445 return self
.do_open(functools
.partial(
1446 _create_http_connection
, self
, conn_class
, True),
1450 class YoutubeDLCookieJar(compat_cookiejar
.MozillaCookieJar
):
1452 See [1] for cookie file format.
1454 1. https://curl.haxx.se/docs/http-cookies.html
1456 _HTTPONLY_PREFIX
= '#HttpOnly_'
1458 _HEADER
= '''# Netscape HTTP Cookie File
1459 # This file is generated by yt-dlp. Do not edit.
1462 _CookieFileEntry
= collections
.namedtuple(
1464 ('domain_name', 'include_subdomains', 'path', 'https_only', 'expires_at', 'name', 'value'))
1466 def save(self
, filename
=None, ignore_discard
=False, ignore_expires
=False):
1468 Save cookies to a file.
1470 Most of the code is taken from CPython 3.8 and slightly adapted
1471 to support cookie files with UTF-8 in both python 2 and 3.
1473 if filename
is None:
1474 if self
.filename
is not None:
1475 filename
= self
.filename
1477 raise ValueError(compat_cookiejar
.MISSING_FILENAME_TEXT
)
1479 # Store session cookies with `expires` set to 0 instead of an empty
1482 if cookie
.expires
is None:
1485 with io
.open(filename
, 'w', encoding
='utf-8') as f
:
1486 f
.write(self
._HEADER
)
1489 if not ignore_discard
and cookie
.discard
:
1491 if not ignore_expires
and cookie
.is_expired(now
):
1497 if cookie
.domain
.startswith('.'):
1498 initial_dot
= 'TRUE'
1500 initial_dot
= 'FALSE'
1501 if cookie
.expires
is not None:
1502 expires
= compat_str(cookie
.expires
)
1505 if cookie
.value
is None:
1506 # cookies.txt regards 'Set-Cookie: foo' as a cookie
1507 # with no name, whereas http.cookiejar regards it as a
1508 # cookie with no value.
1513 value
= cookie
.value
1515 '\t'.join([cookie
.domain
, initial_dot
, cookie
.path
,
1516 secure
, expires
, name
, value
]) + '\n')
1518 def load(self
, filename
=None, ignore_discard
=False, ignore_expires
=False):
1519 """Load cookies from a file."""
1520 if filename
is None:
1521 if self
.filename
is not None:
1522 filename
= self
.filename
1524 raise ValueError(compat_cookiejar
.MISSING_FILENAME_TEXT
)
1526 def prepare_line(line
):
1527 if line
.startswith(self
._HTTPONLY
_PREFIX
):
1528 line
= line
[len(self
._HTTPONLY
_PREFIX
):]
1529 # comments and empty lines are fine
1530 if line
.startswith('#') or not line
.strip():
1532 cookie_list
= line
.split('\t')
1533 if len(cookie_list
) != self
._ENTRY
_LEN
:
1534 raise compat_cookiejar
.LoadError('invalid length %d' % len(cookie_list
))
1535 cookie
= self
._CookieFileEntry
(*cookie_list
)
1536 if cookie
.expires_at
and not cookie
.expires_at
.isdigit():
1537 raise compat_cookiejar
.LoadError('invalid expires at %s' % cookie
.expires_at
)
1541 with io
.open(filename
, encoding
='utf-8') as f
:
1544 cf
.write(prepare_line(line
))
1545 except compat_cookiejar
.LoadError
as e
:
1547 'WARNING: skipping cookie file entry due to %s: %r\n'
1548 % (e
, line
), sys
.stderr
)
1551 self
._really
_load
(cf
, filename
, ignore_discard
, ignore_expires
)
1552 # Session cookies are denoted by either `expires` field set to
1553 # an empty string or 0. MozillaCookieJar only recognizes the former
1554 # (see [1]). So we need force the latter to be recognized as session
1555 # cookies on our own.
1556 # Session cookies may be important for cookies-based authentication,
1557 # e.g. usually, when user does not check 'Remember me' check box while
1558 # logging in on a site, some important cookies are stored as session
1559 # cookies so that not recognizing them will result in failed login.
1560 # 1. https://bugs.python.org/issue17164
1562 # Treat `expires=0` cookies as session cookies
1563 if cookie
.expires
== 0:
1564 cookie
.expires
= None
1565 cookie
.discard
= True
1568 class YoutubeDLCookieProcessor(compat_urllib_request
.HTTPCookieProcessor
):
1569 def __init__(self
, cookiejar
=None):
1570 compat_urllib_request
.HTTPCookieProcessor
.__init
__(self
, cookiejar
)
1572 def http_response(self
, request
, response
):
1573 return compat_urllib_request
.HTTPCookieProcessor
.http_response(self
, request
, response
)
1575 https_request
= compat_urllib_request
.HTTPCookieProcessor
.http_request
1576 https_response
= http_response
1579 class YoutubeDLRedirectHandler(compat_urllib_request
.HTTPRedirectHandler
):
1580 """YoutubeDL redirect handler
1582 The code is based on HTTPRedirectHandler implementation from CPython [1].
1584 This redirect handler solves two issues:
1585 - ensures redirect URL is always unicode under python 2
1586 - introduces support for experimental HTTP response status code
1587 308 Permanent Redirect [2] used by some sites [3]
1589 1. https://github.com/python/cpython/blob/master/Lib/urllib/request.py
1590 2. https://developer.mozilla.org/en-US/docs/Web/HTTP/Status/308
1591 3. https://github.com/ytdl-org/youtube-dl/issues/28768
1594 http_error_301
= http_error_303
= http_error_307
= http_error_308
= compat_urllib_request
.HTTPRedirectHandler
.http_error_302
1596 def redirect_request(self
, req
, fp
, code
, msg
, headers
, newurl
):
1597 """Return a Request or None in response to a redirect.
1599 This is called by the http_error_30x methods when a
1600 redirection response is received. If a redirection should
1601 take place, return a new Request to allow http_error_30x to
1602 perform the redirect. Otherwise, raise HTTPError if no-one
1603 else should try to handle this url. Return None if you can't
1604 but another Handler might.
1606 m
= req
.get_method()
1607 if (not (code
in (301, 302, 303, 307, 308) and m
in ("GET", "HEAD")
1608 or code
in (301, 302, 303) and m
== "POST")):
1609 raise compat_HTTPError(req
.full_url
, code
, msg
, headers
, fp
)
1610 # Strictly (according to RFC 2616), 301 or 302 in response to
1611 # a POST MUST NOT cause a redirection without confirmation
1612 # from the user (of urllib.request, in this case). In practice,
1613 # essentially all clients do redirect in this case, so we do
1616 # Be conciliant with URIs containing a space. This is mainly
1617 # redundant with the more complete encoding done in http_error_302(),
1618 # but it is kept for compatibility with other callers.
1619 newurl
= newurl
.replace(' ', '%20')
1621 CONTENT_HEADERS
= ("content-length", "content-type")
1622 # NB: don't use dict comprehension for python 2.6 compatibility
1623 newheaders
= dict((k
, v
) for k
, v
in req
.headers
.items()
1624 if k
.lower() not in CONTENT_HEADERS
)
1625 return compat_urllib_request
.Request(
1626 newurl
, headers
=newheaders
, origin_req_host
=req
.origin_req_host
,
1630 def extract_timezone(date_str
):
1633 ^.{8,}? # >=8 char non-TZ prefix, if present
1634 (?P<tz>Z| # just the UTC Z, or
1635 (?:(?<=.\b\d{4}|\b\d{2}:\d\d)| # preceded by 4 digits or hh:mm or
1636 (?<!.\b[a-zA-Z]{3}|[a-zA-Z]{4}|..\b\d\d)) # not preceded by 3 alpha word or >= 4 alpha or 2 digits
1637 [ ]? # optional space
1638 (?P<sign>\+|-) # +/-
1639 (?P<hours>[0-9]{2}):?(?P<minutes>[0-9]{2}) # hh[:]mm
1643 timezone
= datetime
.timedelta()
1645 date_str
= date_str
[:-len(m
.group('tz'))]
1646 if not m
.group('sign'):
1647 timezone
= datetime
.timedelta()
1649 sign
= 1 if m
.group('sign') == '+' else -1
1650 timezone
= datetime
.timedelta(
1651 hours
=sign
* int(m
.group('hours')),
1652 minutes
=sign
* int(m
.group('minutes')))
1653 return timezone
, date_str
1656 def parse_iso8601(date_str
, delimiter
='T', timezone
=None):
1657 """ Return a UNIX timestamp from the given date """
1659 if date_str
is None:
1662 date_str
= re
.sub(r
'\.[0-9]+', '', date_str
)
1664 if timezone
is None:
1665 timezone
, date_str
= extract_timezone(date_str
)
1668 date_format
= '%Y-%m-%d{0}%H:%M:%S'.format(delimiter
)
1669 dt
= datetime
.datetime
.strptime(date_str
, date_format
) - timezone
1670 return calendar
.timegm(dt
.timetuple())
1675 def date_formats(day_first
=True):
1676 return DATE_FORMATS_DAY_FIRST
if day_first
else DATE_FORMATS_MONTH_FIRST
1679 def unified_strdate(date_str
, day_first
=True):
1680 """Return a string with the date in the format YYYYMMDD"""
1682 if date_str
is None:
1686 date_str
= date_str
.replace(',', ' ')
1687 # Remove AM/PM + timezone
1688 date_str
= re
.sub(r
'(?i)\s*(?:AM|PM)(?:\s+[A-Z]+)?', '', date_str
)
1689 _
, date_str
= extract_timezone(date_str
)
1691 for expression
in date_formats(day_first
):
1693 upload_date
= datetime
.datetime
.strptime(date_str
, expression
).strftime('%Y%m%d')
1696 if upload_date
is None:
1697 timetuple
= email
.utils
.parsedate_tz(date_str
)
1700 upload_date
= datetime
.datetime(*timetuple
[:6]).strftime('%Y%m%d')
1703 if upload_date
is not None:
1704 return compat_str(upload_date
)
1707 def unified_timestamp(date_str
, day_first
=True):
1708 if date_str
is None:
1711 date_str
= re
.sub(r
'[,|]', '', date_str
)
1713 pm_delta
= 12 if re
.search(r
'(?i)PM', date_str
) else 0
1714 timezone
, date_str
= extract_timezone(date_str
)
1716 # Remove AM/PM + timezone
1717 date_str
= re
.sub(r
'(?i)\s*(?:AM|PM)(?:\s+[A-Z]+)?', '', date_str
)
1719 # Remove unrecognized timezones from ISO 8601 alike timestamps
1720 m
= re
.search(r
'\d{1,2}:\d{1,2}(?:\.\d+)?(?P<tz>\s*[A-Z]+)$', date_str
)
1722 date_str
= date_str
[:-len(m
.group('tz'))]
1724 # Python only supports microseconds, so remove nanoseconds
1725 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]+$', date_str
)
1727 date_str
= m
.group(1)
1729 for expression
in date_formats(day_first
):
1731 dt
= datetime
.datetime
.strptime(date_str
, expression
) - timezone
+ datetime
.timedelta(hours
=pm_delta
)
1732 return calendar
.timegm(dt
.timetuple())
1735 timetuple
= email
.utils
.parsedate_tz(date_str
)
1737 return calendar
.timegm(timetuple
) + pm_delta
* 3600
1740 def determine_ext(url
, default_ext
='unknown_video'):
1741 if url
is None or '.' not in url
:
1743 guess
= url
.partition('?')[0].rpartition('.')[2]
1744 if re
.match(r
'^[A-Za-z0-9]+$', guess
):
1746 # Try extract ext from URLs like http://example.com/foo/bar.mp4/?download
1747 elif guess
.rstrip('/') in KNOWN_EXTENSIONS
:
1748 return guess
.rstrip('/')
1753 def subtitles_filename(filename
, sub_lang
, sub_format
, expected_real_ext
=None):
1754 return replace_extension(filename
, sub_lang
+ '.' + sub_format
, expected_real_ext
)
1757 def datetime_from_str(date_str
, precision
='auto', format
='%Y%m%d'):
1759 Return a datetime object from a string in the format YYYYMMDD or
1760 (now|today|yesterday|date)[+-][0-9](microsecond|second|minute|hour|day|week|month|year)(s)?
1762 format: string date format used to return datetime object from
1763 precision: round the time portion of a datetime object.
1764 auto|microsecond|second|minute|hour|day.
1765 auto: round to the unit provided in date_str (if applicable).
1767 auto_precision
= False
1768 if precision
== 'auto':
1769 auto_precision
= True
1770 precision
= 'microsecond'
1771 today
= datetime_round(datetime
.datetime
.utcnow(), precision
)
1772 if date_str
in ('now', 'today'):
1774 if date_str
== 'yesterday':
1775 return today
- datetime
.timedelta(days
=1)
1777 r
'(?P<start>.+)(?P<sign>[+-])(?P<time>\d+)(?P<unit>microsecond|second|minute|hour|day|week|month|year)(s)?',
1779 if match
is not None:
1780 start_time
= datetime_from_str(match
.group('start'), precision
, format
)
1781 time
= int(match
.group('time')) * (-1 if match
.group('sign') == '-' else 1)
1782 unit
= match
.group('unit')
1783 if unit
== 'month' or unit
== 'year':
1784 new_date
= datetime_add_months(start_time
, time
* 12 if unit
== 'year' else time
)
1790 delta
= datetime
.timedelta(**{unit + 's': time}
)
1791 new_date
= start_time
+ delta
1793 return datetime_round(new_date
, unit
)
1796 return datetime_round(datetime
.datetime
.strptime(date_str
, format
), precision
)
1799 def date_from_str(date_str
, format
='%Y%m%d', strict
=False):
1801 Return a datetime object from a string in the format YYYYMMDD or
1802 (now|today|yesterday|date)[+-][0-9](microsecond|second|minute|hour|day|week|month|year)(s)?
1804 If "strict", only (now|today)[+-][0-9](day|week|month|year)(s)? is allowed
1806 format: string date format used to return datetime object from
1808 if strict
and not re
.fullmatch(r
'\d{8}|(now|today)[+-]\d+(day|week|month|year)(s)?', date_str
):
1809 raise ValueError(f
'Invalid date format {date_str}')
1810 return datetime_from_str(date_str
, precision
='microsecond', format
=format
).date()
1813 def datetime_add_months(dt
, months
):
1814 """Increment/Decrement a datetime object by months."""
1815 month
= dt
.month
+ months
- 1
1816 year
= dt
.year
+ month
// 12
1817 month
= month
% 12 + 1
1818 day
= min(dt
.day
, calendar
.monthrange(year
, month
)[1])
1819 return dt
.replace(year
, month
, day
)
1822 def datetime_round(dt
, precision
='day'):
1824 Round a datetime object's time to a specific precision
1826 if precision
== 'microsecond':
1835 roundto
= lambda x
, n
: ((x
+ n
/ 2) // n
) * n
1836 timestamp
= calendar
.timegm(dt
.timetuple())
1837 return datetime
.datetime
.utcfromtimestamp(roundto(timestamp
, unit_seconds
[precision
]))
1840 def hyphenate_date(date_str
):
1842 Convert a date in 'YYYYMMDD' format to 'YYYY-MM-DD' format"""
1843 match
= re
.match(r
'^(\d\d\d\d)(\d\d)(\d\d)$', date_str
)
1844 if match
is not None:
1845 return '-'.join(match
.groups())
1850 class DateRange(object):
1851 """Represents a time interval between two dates"""
1853 def __init__(self
, start
=None, end
=None):
1854 """start and end must be strings in the format accepted by date"""
1855 if start
is not None:
1856 self
.start
= date_from_str(start
, strict
=True)
1858 self
.start
= datetime
.datetime
.min.date()
1860 self
.end
= date_from_str(end
, strict
=True)
1862 self
.end
= datetime
.datetime
.max.date()
1863 if self
.start
> self
.end
:
1864 raise ValueError('Date range: "%s" , the start date must be before the end date' % self
)
1868 """Returns a range that only contains the given day"""
1869 return cls(day
, day
)
1871 def __contains__(self
, date
):
1872 """Check if the date is in the range"""
1873 if not isinstance(date
, datetime
.date
):
1874 date
= date_from_str(date
)
1875 return self
.start
<= date
<= self
.end
1878 return '%s - %s' % (self
.start
.isoformat(), self
.end
.isoformat())
1881 def platform_name():
1882 """ Returns the platform name as a compat_str """
1883 res
= platform
.platform()
1884 if isinstance(res
, bytes):
1885 res
= res
.decode(preferredencoding())
1887 assert isinstance(res
, compat_str
)
1891 def get_windows_version():
1892 ''' Get Windows version. None if it's not running on Windows '''
1893 if compat_os_name
== 'nt':
1894 return version_tuple(platform
.win32_ver()[1])
1899 def write_string(s
, out
=None, encoding
=None):
1902 assert type(s
) == compat_str
1904 if 'b' in getattr(out
, 'mode', ''):
1905 byt
= s
.encode(encoding
or preferredencoding(), 'ignore')
1907 elif hasattr(out
, 'buffer'):
1908 enc
= encoding
or getattr(out
, 'encoding', None) or preferredencoding()
1909 byt
= s
.encode(enc
, 'ignore')
1910 out
.buffer.write(byt
)
1916 def bytes_to_intlist(bs
):
1919 if isinstance(bs
[0], int): # Python 3
1922 return [ord(c
) for c
in bs
]
1925 def intlist_to_bytes(xs
):
1928 return compat_struct_pack('%dB' % len(xs
), *xs
)
1931 class LockingUnsupportedError(IOError):
1932 msg
= 'File locking is not supported on this platform'
1935 super().__init
__(self
.msg
)
1938 # Cross-platform file locking
1939 if sys
.platform
== 'win32':
1940 import ctypes
.wintypes
1943 class OVERLAPPED(ctypes
.Structure
):
1945 ('Internal', ctypes
.wintypes
.LPVOID
),
1946 ('InternalHigh', ctypes
.wintypes
.LPVOID
),
1947 ('Offset', ctypes
.wintypes
.DWORD
),
1948 ('OffsetHigh', ctypes
.wintypes
.DWORD
),
1949 ('hEvent', ctypes
.wintypes
.HANDLE
),
1952 kernel32
= ctypes
.windll
.kernel32
1953 LockFileEx
= kernel32
.LockFileEx
1954 LockFileEx
.argtypes
= [
1955 ctypes
.wintypes
.HANDLE
, # hFile
1956 ctypes
.wintypes
.DWORD
, # dwFlags
1957 ctypes
.wintypes
.DWORD
, # dwReserved
1958 ctypes
.wintypes
.DWORD
, # nNumberOfBytesToLockLow
1959 ctypes
.wintypes
.DWORD
, # nNumberOfBytesToLockHigh
1960 ctypes
.POINTER(OVERLAPPED
) # Overlapped
1962 LockFileEx
.restype
= ctypes
.wintypes
.BOOL
1963 UnlockFileEx
= kernel32
.UnlockFileEx
1964 UnlockFileEx
.argtypes
= [
1965 ctypes
.wintypes
.HANDLE
, # hFile
1966 ctypes
.wintypes
.DWORD
, # dwReserved
1967 ctypes
.wintypes
.DWORD
, # nNumberOfBytesToLockLow
1968 ctypes
.wintypes
.DWORD
, # nNumberOfBytesToLockHigh
1969 ctypes
.POINTER(OVERLAPPED
) # Overlapped
1971 UnlockFileEx
.restype
= ctypes
.wintypes
.BOOL
1972 whole_low
= 0xffffffff
1973 whole_high
= 0x7fffffff
1975 def _lock_file(f
, exclusive
, block
):
1976 overlapped
= OVERLAPPED()
1977 overlapped
.Offset
= 0
1978 overlapped
.OffsetHigh
= 0
1979 overlapped
.hEvent
= 0
1980 f
._lock
_file
_overlapped
_p
= ctypes
.pointer(overlapped
)
1982 if not LockFileEx(msvcrt
.get_osfhandle(f
.fileno()),
1983 (0x2 if exclusive
else 0x0) |
(0x0 if block
else 0x1),
1984 0, whole_low
, whole_high
, f
._lock
_file
_overlapped
_p
):
1985 raise BlockingIOError('Locking file failed: %r' % ctypes
.FormatError())
1987 def _unlock_file(f
):
1988 assert f
._lock
_file
_overlapped
_p
1989 handle
= msvcrt
.get_osfhandle(f
.fileno())
1990 if not UnlockFileEx(handle
, 0, whole_low
, whole_high
, f
._lock
_file
_overlapped
_p
):
1991 raise OSError('Unlocking file failed: %r' % ctypes
.FormatError())
1997 def _lock_file(f
, exclusive
, block
):
1998 flags
= fcntl
.LOCK_EX
if exclusive
else fcntl
.LOCK_SH
2000 flags |
= fcntl
.LOCK_NB
2002 fcntl
.flock(f
, flags
)
2003 except BlockingIOError
:
2005 except OSError: # AOSP does not have flock()
2006 fcntl
.lockf(f
, flags
)
2008 def _unlock_file(f
):
2010 fcntl
.flock(f
, fcntl
.LOCK_UN
)
2012 fcntl
.lockf(f
, fcntl
.LOCK_UN
)
2016 def _lock_file(f
, exclusive
, block
):
2017 raise LockingUnsupportedError()
2019 def _unlock_file(f
):
2020 raise LockingUnsupportedError()
2023 class locked_file(object):
2026 def __init__(self
, filename
, mode
, block
=True, encoding
=None):
2027 if mode
not in {'r', 'rb', 'a', 'ab', 'w', 'wb'}
:
2028 raise NotImplementedError(mode
)
2029 self
.mode
, self
.block
= mode
, block
2031 writable
= any(f
in mode
for f
in 'wax+')
2032 readable
= any(f
in mode
for f
in 'r+')
2033 flags
= functools
.reduce(operator
.ior
, (
2034 getattr(os
, 'O_CLOEXEC', 0), # UNIX only
2035 getattr(os
, 'O_BINARY', 0), # Windows only
2036 getattr(os
, 'O_NOINHERIT', 0), # Windows only
2037 os
.O_CREAT
if writable
else 0, # O_TRUNC only after locking
2038 os
.O_APPEND
if 'a' in mode
else 0,
2039 os
.O_EXCL
if 'x' in mode
else 0,
2040 os
.O_RDONLY
if not writable
else os
.O_RDWR
if readable
else os
.O_WRONLY
,
2043 self
.f
= os
.fdopen(os
.open(filename
, flags
, 0o666), mode
, encoding
=encoding
)
2045 def __enter__(self
):
2046 exclusive
= 'r' not in self
.mode
2048 _lock_file(self
.f
, exclusive
, self
.block
)
2053 if 'w' in self
.mode
:
2061 _unlock_file(self
.f
)
2065 def __exit__(self
, *_
):
2074 def __getattr__(self
, attr
):
2075 return getattr(self
.f
, attr
)
2081 def get_filesystem_encoding():
2082 encoding
= sys
.getfilesystemencoding()
2083 return encoding
if encoding
is not None else 'utf-8'
2086 def shell_quote(args
):
2088 encoding
= get_filesystem_encoding()
2090 if isinstance(a
, bytes):
2091 # We may get a filename encoded with 'encodeFilename'
2092 a
= a
.decode(encoding
)
2093 quoted_args
.append(compat_shlex_quote(a
))
2094 return ' '.join(quoted_args
)
2097 def smuggle_url(url
, data
):
2098 """ Pass additional data in a URL for internal use. """
2100 url
, idata
= unsmuggle_url(url
, {})
2102 sdata
= compat_urllib_parse_urlencode(
2103 {'__youtubedl_smuggle': json.dumps(data)}
)
2104 return url
+ '#' + sdata
2107 def unsmuggle_url(smug_url
, default
=None):
2108 if '#__youtubedl_smuggle' not in smug_url
:
2109 return smug_url
, default
2110 url
, _
, sdata
= smug_url
.rpartition('#')
2111 jsond
= compat_parse_qs(sdata
)['__youtubedl_smuggle'][0]
2112 data
= json
.loads(jsond
)
2116 def format_decimal_suffix(num
, fmt
='%d%s', *, factor
=1000):
2117 """ Formats numbers with decimal sufixes like K, M, etc """
2118 num
, factor
= float_or_none(num
), float(factor
)
2119 if num
is None or num
< 0:
2121 POSSIBLE_SUFFIXES
= 'kMGTPEZY'
2122 exponent
= 0 if num
== 0 else min(int(math
.log(num
, factor
)), len(POSSIBLE_SUFFIXES
))
2123 suffix
= ['', *POSSIBLE_SUFFIXES
][exponent
]
2125 suffix
= {'k': 'Ki', '': ''}
.get(suffix
, f
'{suffix}i')
2126 converted
= num
/ (factor
** exponent
)
2127 return fmt
% (converted
, suffix
)
2130 def format_bytes(bytes):
2131 return format_decimal_suffix(bytes, '%.2f%sB', factor
=1024) or 'N/A'
2134 def lookup_unit_table(unit_table
, s
):
2135 units_re
= '|'.join(re
.escape(u
) for u
in unit_table
)
2137 r
'(?P<num>[0-9]+(?:[,.][0-9]*)?)\s*(?P<unit>%s)\b' % units_re
, s
)
2140 num_str
= m
.group('num').replace(',', '.')
2141 mult
= unit_table
[m
.group('unit')]
2142 return int(float(num_str
) * mult
)
2145 def parse_filesize(s
):
2149 # The lower-case forms are of course incorrect and unofficial,
2150 # but we support those too
2167 'megabytes': 1000 ** 2,
2168 'mebibytes': 1024 ** 2,
2174 'gigabytes': 1000 ** 3,
2175 'gibibytes': 1024 ** 3,
2181 'terabytes': 1000 ** 4,
2182 'tebibytes': 1024 ** 4,
2188 'petabytes': 1000 ** 5,
2189 'pebibytes': 1024 ** 5,
2195 'exabytes': 1000 ** 6,
2196 'exbibytes': 1024 ** 6,
2202 'zettabytes': 1000 ** 7,
2203 'zebibytes': 1024 ** 7,
2209 'yottabytes': 1000 ** 8,
2210 'yobibytes': 1024 ** 8,
2213 return lookup_unit_table(_UNIT_TABLE
, s
)
2220 s
= re
.sub(r
'^[^\d]+\s', '', s
).strip()
2222 if re
.match(r
'^[\d,.]+$', s
):
2223 return str_to_int(s
)
2236 ret
= lookup_unit_table(_UNIT_TABLE
, s
)
2240 mobj
= re
.match(r
'([\d,.]+)(?:$|\s)', s
)
2242 return str_to_int(mobj
.group(1))
2245 def parse_resolution(s
, *, lenient
=False):
2250 mobj
= re
.search(r
'(?P<w>\d+)\s*[xX×,]\s*(?P<h>\d+)', s
)
2252 mobj
= re
.search(r
'(?<![a-zA-Z0-9])(?P<w>\d+)\s*[xX×,]\s*(?P<h>\d+)(?![a-zA-Z0-9])', s
)
2255 'width': int(mobj
.group('w')),
2256 'height': int(mobj
.group('h')),
2259 mobj
= re
.search(r
'(?<![a-zA-Z0-9])(\d+)[pPiI](?![a-zA-Z0-9])', s
)
2261 return {'height': int(mobj.group(1))}
2263 mobj
= re
.search(r
'\b([48])[kK]\b', s
)
2265 return {'height': int(mobj.group(1)) * 540}
2270 def parse_bitrate(s
):
2271 if not isinstance(s
, compat_str
):
2273 mobj
= re
.search(r
'\b(\d+)\s*kbps', s
)
2275 return int(mobj
.group(1))
2278 def month_by_name(name
, lang
='en'):
2279 """ Return the number of a month by (locale-independently) English name """
2281 month_names
= MONTH_NAMES
.get(lang
, MONTH_NAMES
['en'])
2284 return month_names
.index(name
) + 1
2289 def month_by_abbreviation(abbrev
):
2290 """ Return the number of a month by (locale-independently) English
2294 return [s
[:3] for s
in ENGLISH_MONTH_NAMES
].index(abbrev
) + 1
2299 def fix_xml_ampersands(xml_str
):
2300 """Replace all the '&' by '&' in XML"""
2302 r
'&(?!amp;|lt;|gt;|apos;|quot;|#x[0-9a-fA-F]{,4};|#[0-9]{,4};)',
2307 def setproctitle(title
):
2308 assert isinstance(title
, compat_str
)
2310 # ctypes in Jython is not complete
2311 # http://bugs.jython.org/issue2148
2312 if sys
.platform
.startswith('java'):
2316 libc
= ctypes
.cdll
.LoadLibrary('libc.so.6')
2320 # LoadLibrary in Windows Python 2.7.13 only expects
2321 # a bytestring, but since unicode_literals turns
2322 # every string into a unicode string, it fails.
2324 title_bytes
= title
.encode('utf-8')
2325 buf
= ctypes
.create_string_buffer(len(title_bytes
))
2326 buf
.value
= title_bytes
2328 libc
.prctl(15, buf
, 0, 0, 0)
2329 except AttributeError:
2330 return # Strange libc, just skip this
2333 def remove_start(s
, start
):
2334 return s
[len(start
):] if s
is not None and s
.startswith(start
) else s
2337 def remove_end(s
, end
):
2338 return s
[:-len(end
)] if s
is not None and s
.endswith(end
) else s
2341 def remove_quotes(s
):
2342 if s
is None or len(s
) < 2:
2344 for quote
in ('"', "'", ):
2345 if s
[0] == quote
and s
[-1] == quote
:
2350 def get_domain(url
):
2351 domain
= re
.match(r
'(?:https?:\/\/)?(?:www\.)?(?P<domain>[^\n\/]+\.[^\n\/]+)(?:\/(.*))?', url
)
2352 return domain
.group('domain') if domain
else None
2355 def url_basename(url
):
2356 path
= compat_urlparse
.urlparse(url
).path
2357 return path
.strip('/').split('/')[-1]
2361 return re
.match(r
'https?://[^?#&]+/', url
).group()
2364 def urljoin(base
, path
):
2365 if isinstance(path
, bytes):
2366 path
= path
.decode('utf-8')
2367 if not isinstance(path
, compat_str
) or not path
:
2369 if re
.match(r
'^(?:[a-zA-Z][a-zA-Z0-9+-.]*:)?//', path
):
2371 if isinstance(base
, bytes):
2372 base
= base
.decode('utf-8')
2373 if not isinstance(base
, compat_str
) or not re
.match(
2374 r
'^(?:https?:)?//', base
):
2376 return compat_urlparse
.urljoin(base
, path
)
2379 class HEADRequest(compat_urllib_request
.Request
):
2380 def get_method(self
):
2384 class PUTRequest(compat_urllib_request
.Request
):
2385 def get_method(self
):
2389 def int_or_none(v
, scale
=1, default
=None, get_attr
=None, invscale
=1):
2390 if get_attr
and v
is not None:
2391 v
= getattr(v
, get_attr
, None)
2393 return int(v
) * invscale
// scale
2394 except (ValueError, TypeError, OverflowError):
2398 def str_or_none(v
, default
=None):
2399 return default
if v
is None else compat_str(v
)
2402 def str_to_int(int_str
):
2403 """ A more relaxed version of int_or_none """
2404 if isinstance(int_str
, compat_integer_types
):
2406 elif isinstance(int_str
, compat_str
):
2407 int_str
= re
.sub(r
'[,\.\+]', '', int_str
)
2408 return int_or_none(int_str
)
2411 def float_or_none(v
, scale
=1, invscale
=1, default
=None):
2415 return float(v
) * invscale
/ scale
2416 except (ValueError, TypeError):
2420 def bool_or_none(v
, default
=None):
2421 return v
if isinstance(v
, bool) else default
2424 def strip_or_none(v
, default
=None):
2425 return v
.strip() if isinstance(v
, compat_str
) else default
2428 def url_or_none(url
):
2429 if not url
or not isinstance(url
, compat_str
):
2432 return url
if re
.match(r
'^(?:(?:https?|rt(?:m(?:pt?[es]?|fp)|sp[su]?)|mms|ftps?):)?//', url
) else None
2435 def request_to_url(req
):
2436 if isinstance(req
, compat_urllib_request
.Request
):
2437 return req
.get_full_url()
2442 def strftime_or_none(timestamp
, date_format
, default
=None):
2443 datetime_object
= None
2445 if isinstance(timestamp
, compat_numeric_types
): # unix timestamp
2446 datetime_object
= datetime
.datetime
.utcfromtimestamp(timestamp
)
2447 elif isinstance(timestamp
, compat_str
): # assume YYYYMMDD
2448 datetime_object
= datetime
.datetime
.strptime(timestamp
, '%Y%m%d')
2449 return datetime_object
.strftime(date_format
)
2450 except (ValueError, TypeError, AttributeError):
2454 def parse_duration(s
):
2455 if not isinstance(s
, compat_basestring
):
2461 days
, hours
, mins
, secs
, ms
= [None] * 5
2462 m
= re
.match(r
'''(?x)
2464 (?:(?:(?P<days>[0-9]+):)?(?P<hours>[0-9]+):)?(?P<mins>[0-9]+):)?
2465 (?P<secs>(?(before_secs)[0-9]{1,2}|[0-9]+))
2466 (?P<ms>[.:][0-9]+)?Z?$
2469 days
, hours
, mins
, secs
, ms
= m
.group('days', 'hours', 'mins', 'secs', 'ms')
2474 [0-9]+\s*y(?:ears?)?,?\s*
2477 [0-9]+\s*m(?:onths?)?,?\s*
2480 [0-9]+\s*w(?:eeks?)?,?\s*
2483 (?P<days>[0-9]+)\s*d(?:ays?)?,?\s*
2487 (?P<hours>[0-9]+)\s*h(?:ours?)?,?\s*
2490 (?P<mins>[0-9]+)\s*m(?:in(?:ute)?s?)?,?\s*
2493 (?P<secs>[0-9]+)(?P<ms>\.[0-9]+)?\s*s(?:ec(?:ond)?s?)?\s*
2496 days
, hours
, mins
, secs
, ms
= m
.groups()
2498 m
= re
.match(r
'(?i)(?:(?P<hours>[0-9.]+)\s*(?:hours?)|(?P<mins>[0-9.]+)\s*(?:mins?\.?|minutes?)\s*)Z?$', s
)
2500 hours
, mins
= m
.groups()
2506 duration
+= float(secs
)
2508 duration
+= float(mins
) * 60
2510 duration
+= float(hours
) * 60 * 60
2512 duration
+= float(days
) * 24 * 60 * 60
2514 duration
+= float(ms
.replace(':', '.'))
2518 def prepend_extension(filename
, ext
, expected_real_ext
=None):
2519 name
, real_ext
= os
.path
.splitext(filename
)
2521 '{0}.{1}{2}'.format(name
, ext
, real_ext
)
2522 if not expected_real_ext
or real_ext
[1:] == expected_real_ext
2523 else '{0}.{1}'.format(filename
, ext
))
2526 def replace_extension(filename
, ext
, expected_real_ext
=None):
2527 name
, real_ext
= os
.path
.splitext(filename
)
2528 return '{0}.{1}'.format(
2529 name
if not expected_real_ext
or real_ext
[1:] == expected_real_ext
else filename
,
2533 def check_executable(exe
, args
=[]):
2534 """ Checks if the given binary is installed somewhere in PATH, and returns its name.
2535 args can be a list of arguments for a short output (like -version) """
2537 Popen([exe
] + args
, stdout
=subprocess
.PIPE
, stderr
=subprocess
.PIPE
).communicate_or_kill()
2543 def _get_exe_version_output(exe
, args
, *, to_screen
=None):
2545 to_screen(f
'Checking exe version: {shell_quote([exe] + args)}')
2547 # STDIN should be redirected too. On UNIX-like systems, ffmpeg triggers
2548 # SIGTTOU if yt-dlp is run in the background.
2549 # See https://github.com/ytdl-org/youtube-dl/issues/955#issuecomment-209789656
2551 [encodeArgument(exe
)] + args
, stdin
=subprocess
.PIPE
,
2552 stdout
=subprocess
.PIPE
, stderr
=subprocess
.STDOUT
).communicate_or_kill()
2555 if isinstance(out
, bytes): # Python 2.x
2556 out
= out
.decode('ascii', 'ignore')
2560 def detect_exe_version(output
, version_re
=None, unrecognized
='present'):
2561 assert isinstance(output
, compat_str
)
2562 if version_re
is None:
2563 version_re
= r
'version\s+([-0-9._a-zA-Z]+)'
2564 m
= re
.search(version_re
, output
)
2571 def get_exe_version(exe
, args
=['--version'],
2572 version_re
=None, unrecognized
='present'):
2573 """ Returns the version of the specified executable,
2574 or False if the executable is not present """
2575 out
= _get_exe_version_output(exe
, args
)
2576 return detect_exe_version(out
, version_re
, unrecognized
) if out
else False
2579 class LazyList(collections
.abc
.Sequence
):
2580 ''' Lazy immutable list from an iterable
2581 Note that slices of a LazyList are lists and not LazyList'''
2583 class IndexError(IndexError):
2586 def __init__(self
, iterable
, *, reverse
=False, _cache
=None):
2587 self
.__iterable
= iter(iterable
)
2588 self
.__cache
= [] if _cache
is None else _cache
2589 self
.__reversed
= reverse
2593 # We need to consume the entire iterable to iterate in reverse
2594 yield from self
.exhaust()
2596 yield from self
.__cache
2597 for item
in self
.__iterable
:
2598 self
.__cache
.append(item
)
2601 def __exhaust(self
):
2602 self
.__cache
.extend(self
.__iterable
)
2603 # Discard the emptied iterable to make it pickle-able
2604 self
.__iterable
= []
2608 ''' Evaluate the entire iterable '''
2609 return self
.__exhaust
()[::-1 if self
.__reversed
else 1]
2612 def __reverse_index(x
):
2613 return None if x
is None else -(x
+ 1)
2615 def __getitem__(self
, idx
):
2616 if isinstance(idx
, slice):
2618 idx
= slice(self
.__reverse
_index
(idx
.start
), self
.__reverse
_index
(idx
.stop
), -(idx
.step
or 1))
2619 start
, stop
, step
= idx
.start
, idx
.stop
, idx
.step
or 1
2620 elif isinstance(idx
, int):
2622 idx
= self
.__reverse
_index
(idx
)
2623 start
, stop
, step
= idx
, idx
, 0
2625 raise TypeError('indices must be integers or slices')
2626 if ((start
or 0) < 0 or (stop
or 0) < 0
2627 or (start
is None and step
< 0)
2628 or (stop
is None and step
> 0)):
2629 # We need to consume the entire iterable to be able to slice from the end
2630 # Obviously, never use this with infinite iterables
2633 return self
.__cache
[idx
]
2634 except IndexError as e
:
2635 raise self
.IndexError(e
) from e
2636 n
= max(start
or 0, stop
or 0) - len(self
.__cache
) + 1
2638 self
.__cache
.extend(itertools
.islice(self
.__iterable
, n
))
2640 return self
.__cache
[idx
]
2641 except IndexError as e
:
2642 raise self
.IndexError(e
) from e
2646 self
[-1] if self
.__reversed
else self
[0]
2647 except self
.IndexError:
2653 return len(self
.__cache
)
2655 def __reversed__(self
):
2656 return type(self
)(self
.__iterable
, reverse
=not self
.__reversed
, _cache
=self
.__cache
)
2659 return type(self
)(self
.__iterable
, reverse
=self
.__reversed
, _cache
=self
.__cache
)
2662 # repr and str should mimic a list. So we exhaust the iterable
2663 return repr(self
.exhaust())
2666 return repr(self
.exhaust())
2671 class IndexError(IndexError):
2675 # This is only useful for tests
2676 return len(self
.getslice())
2678 def __init__(self
, pagefunc
, pagesize
, use_cache
=True):
2679 self
._pagefunc
= pagefunc
2680 self
._pagesize
= pagesize
2681 self
._pagecount
= float('inf')
2682 self
._use
_cache
= use_cache
2685 def getpage(self
, pagenum
):
2686 page_results
= self
._cache
.get(pagenum
)
2687 if page_results
is None:
2688 page_results
= [] if pagenum
> self
._pagecount
else list(self
._pagefunc
(pagenum
))
2690 self
._cache
[pagenum
] = page_results
2693 def getslice(self
, start
=0, end
=None):
2694 return list(self
._getslice
(start
, end
))
2696 def _getslice(self
, start
, end
):
2697 raise NotImplementedError('This method must be implemented by subclasses')
2699 def __getitem__(self
, idx
):
2700 assert self
._use
_cache
, 'Indexing PagedList requires cache'
2701 if not isinstance(idx
, int) or idx
< 0:
2702 raise TypeError('indices must be non-negative integers')
2703 entries
= self
.getslice(idx
, idx
+ 1)
2705 raise self
.IndexError()
2709 class OnDemandPagedList(PagedList
):
2710 """Download pages until a page with less than maximum results"""
2711 def _getslice(self
, start
, end
):
2712 for pagenum
in itertools
.count(start
// self
._pagesize
):
2713 firstid
= pagenum
* self
._pagesize
2714 nextfirstid
= pagenum
* self
._pagesize
+ self
._pagesize
2715 if start
>= nextfirstid
:
2719 start
% self
._pagesize
2720 if firstid
<= start
< nextfirstid
2723 ((end
- 1) % self
._pagesize
) + 1
2724 if (end
is not None and firstid
<= end
<= nextfirstid
)
2728 page_results
= self
.getpage(pagenum
)
2730 self
._pagecount
= pagenum
- 1
2732 if startv
!= 0 or endv
is not None:
2733 page_results
= page_results
[startv
:endv
]
2734 yield from page_results
2736 # A little optimization - if current page is not "full", ie. does
2737 # not contain page_size videos then we can assume that this page
2738 # is the last one - there are no more ids on further pages -
2739 # i.e. no need to query again.
2740 if len(page_results
) + startv
< self
._pagesize
:
2743 # If we got the whole page, but the next page is not interesting,
2744 # break out early as well
2745 if end
== nextfirstid
:
2749 class InAdvancePagedList(PagedList
):
2750 """PagedList with total number of pages known in advance"""
2751 def __init__(self
, pagefunc
, pagecount
, pagesize
):
2752 PagedList
.__init
__(self
, pagefunc
, pagesize
, True)
2753 self
._pagecount
= pagecount
2755 def _getslice(self
, start
, end
):
2756 start_page
= start
// self
._pagesize
2757 end_page
= self
._pagecount
if end
is None else min(self
._pagecount
, end
// self
._pagesize
+ 1)
2758 skip_elems
= start
- start_page
* self
._pagesize
2759 only_more
= None if end
is None else end
- start
2760 for pagenum
in range(start_page
, end_page
):
2761 page_results
= self
.getpage(pagenum
)
2763 page_results
= page_results
[skip_elems
:]
2765 if only_more
is not None:
2766 if len(page_results
) < only_more
:
2767 only_more
-= len(page_results
)
2769 yield from page_results
[:only_more
]
2771 yield from page_results
2774 def uppercase_escape(s
):
2775 unicode_escape
= codecs
.getdecoder('unicode_escape')
2777 r
'\\U[0-9a-fA-F]{8}',
2778 lambda m
: unicode_escape(m
.group(0))[0],
2782 def lowercase_escape(s
):
2783 unicode_escape
= codecs
.getdecoder('unicode_escape')
2785 r
'\\u[0-9a-fA-F]{4}',
2786 lambda m
: unicode_escape(m
.group(0))[0],
2790 def escape_rfc3986(s
):
2791 """Escape non-ASCII characters as suggested by RFC 3986"""
2792 return compat_urllib_parse
.quote(s
, b
"%/;:@&=+$,!~*'()?#[]")
2795 def escape_url(url
):
2796 """Escape URL as suggested by RFC 3986"""
2797 url_parsed
= compat_urllib_parse_urlparse(url
)
2798 return url_parsed
._replace
(
2799 netloc
=url_parsed
.netloc
.encode('idna').decode('ascii'),
2800 path
=escape_rfc3986(url_parsed
.path
),
2801 params
=escape_rfc3986(url_parsed
.params
),
2802 query
=escape_rfc3986(url_parsed
.query
),
2803 fragment
=escape_rfc3986(url_parsed
.fragment
)
2808 return compat_parse_qs(compat_urllib_parse_urlparse(url
).query
)
2811 def read_batch_urls(batch_fd
):
2813 if not isinstance(url
, compat_str
):
2814 url
= url
.decode('utf-8', 'replace')
2815 BOM_UTF8
= ('\xef\xbb\xbf', '\ufeff')
2816 for bom
in BOM_UTF8
:
2817 if url
.startswith(bom
):
2818 url
= url
[len(bom
):]
2820 if not url
or url
.startswith(('#', ';', ']')):
2822 # "#" cannot be stripped out since it is part of the URI
2823 # However, it can be safely stipped out if follwing a whitespace
2824 return re
.split(r
'\s#', url
, 1)[0].rstrip()
2826 with contextlib
.closing(batch_fd
) as fd
:
2827 return [url
for url
in map(fixup
, fd
) if url
]
2830 def urlencode_postdata(*args
, **kargs
):
2831 return compat_urllib_parse_urlencode(*args
, **kargs
).encode('ascii')
2834 def update_url_query(url
, query
):
2837 parsed_url
= compat_urlparse
.urlparse(url
)
2838 qs
= compat_parse_qs(parsed_url
.query
)
2840 return compat_urlparse
.urlunparse(parsed_url
._replace
(
2841 query
=compat_urllib_parse_urlencode(qs
, True)))
2844 def update_Request(req
, url
=None, data
=None, headers
={}, query={}
):
2845 req_headers
= req
.headers
.copy()
2846 req_headers
.update(headers
)
2847 req_data
= data
or req
.data
2848 req_url
= update_url_query(url
or req
.get_full_url(), query
)
2849 req_get_method
= req
.get_method()
2850 if req_get_method
== 'HEAD':
2851 req_type
= HEADRequest
2852 elif req_get_method
== 'PUT':
2853 req_type
= PUTRequest
2855 req_type
= compat_urllib_request
.Request
2857 req_url
, data
=req_data
, headers
=req_headers
,
2858 origin_req_host
=req
.origin_req_host
, unverifiable
=req
.unverifiable
)
2859 if hasattr(req
, 'timeout'):
2860 new_req
.timeout
= req
.timeout
2864 def _multipart_encode_impl(data
, boundary
):
2865 content_type
= 'multipart/form-data; boundary=%s' % boundary
2868 for k
, v
in data
.items():
2869 out
+= b
'--' + boundary
.encode('ascii') + b
'\r\n'
2870 if isinstance(k
, compat_str
):
2871 k
= k
.encode('utf-8')
2872 if isinstance(v
, compat_str
):
2873 v
= v
.encode('utf-8')
2874 # RFC 2047 requires non-ASCII field names to be encoded, while RFC 7578
2875 # suggests sending UTF-8 directly. Firefox sends UTF-8, too
2876 content
= b
'Content-Disposition: form-data; name="' + k
+ b
'"\r\n\r\n' + v
+ b
'\r\n'
2877 if boundary
.encode('ascii') in content
:
2878 raise ValueError('Boundary overlaps with data')
2881 out
+= b
'--' + boundary
.encode('ascii') + b
'--\r\n'
2883 return out
, content_type
2886 def multipart_encode(data
, boundary
=None):
2888 Encode a dict to RFC 7578-compliant form-data
2891 A dict where keys and values can be either Unicode or bytes-like
2894 If specified a Unicode object, it's used as the boundary. Otherwise
2895 a random boundary is generated.
2897 Reference: https://tools.ietf.org/html/rfc7578
2899 has_specified_boundary
= boundary
is not None
2902 if boundary
is None:
2903 boundary
= '---------------' + str(random
.randrange(0x0fffffff, 0xffffffff))
2906 out
, content_type
= _multipart_encode_impl(data
, boundary
)
2909 if has_specified_boundary
:
2913 return out
, content_type
2916 def dict_get(d
, key_or_keys
, default
=None, skip_false_values
=True):
2917 for val
in map(d
.get
, variadic(key_or_keys
)):
2918 if val
is not None and (val
or not skip_false_values
):
2923 def try_call(*funcs
, expected_type
=None, args
=[], kwargs
={}):
2926 val
= f(*args
, **kwargs
)
2927 except (AttributeError, KeyError, TypeError, IndexError, ZeroDivisionError):
2930 if expected_type
is None or isinstance(val
, expected_type
):
2934 def try_get(src
, getter
, expected_type
=None):
2935 return try_call(*variadic(getter
), args
=(src
,), expected_type
=expected_type
)
2938 def filter_dict(dct
, cndn
=lambda _
, v
: v
is not None):
2939 return {k: v for k, v in dct.items() if cndn(k, v)}
2942 def merge_dicts(*dicts
):
2944 for a_dict
in dicts
:
2945 for k
, v
in a_dict
.items():
2946 if (v
is not None and k
not in merged
2947 or isinstance(v
, str) and merged
[k
] == ''):
2952 def encode_compat_str(string
, encoding
=preferredencoding(), errors
='strict'):
2953 return string
if isinstance(string
, compat_str
) else compat_str(string
, encoding
, errors
)
2965 TV_PARENTAL_GUIDELINES
= {
2975 def parse_age_limit(s
):
2977 return s
if 0 <= s
<= 21 else None
2978 if not isinstance(s
, compat_basestring
):
2980 m
= re
.match(r
'^(?P<age>\d{1,2})\+?$', s
)
2982 return int(m
.group('age'))
2985 return US_RATINGS
[s
]
2986 m
= re
.match(r
'^TV[_-]?(%s)$' % '|'.join(k
[3:] for k
in TV_PARENTAL_GUIDELINES
), s
)
2988 return TV_PARENTAL_GUIDELINES
['TV-' + m
.group(1)]
2992 def strip_jsonp(code
):
2995 (?:window\.)?(?P<func_name>[a-zA-Z0-9_.$]*)
2996 (?:\s*&&\s*(?P=func_name))?
2997 \s*\(\s*(?P<callback_data>.*)\);?
2998 \s*?(?://[^\n]*)*$''',
2999 r
'\g<callback_data>', code
)
3002 def js_to_json(code
, vars={}):
3003 # vars is a dict of var, val pairs to substitute
3004 COMMENT_RE
= r
'/\*(?:(?!\*/).)*?\*/|//[^\n]*\n'
3005 SKIP_RE
= r
'\s*(?:{comment})?\s*'.format(comment
=COMMENT_RE
)
3007 (r
'(?s)^(0[xX][0-9a-fA-F]+){skip}:?$'.format(skip
=SKIP_RE
), 16),
3008 (r
'(?s)^(0+[0-7]+){skip}:?$'.format(skip
=SKIP_RE
), 8),
3013 if v
in ('true', 'false', 'null'):
3015 elif v
in ('undefined', 'void 0'):
3017 elif v
.startswith('/*') or v
.startswith('//') or v
.startswith('!') or v
== ',':
3020 if v
[0] in ("'", '"'):
3021 v
= re
.sub(r
'(?s)\\.|"', lambda m
: {
3026 }.get(m
.group(0), m
.group(0)), v
[1:-1])
3028 for regex
, base
in INTEGER_TABLE
:
3029 im
= re
.match(regex
, v
)
3031 i
= int(im
.group(1), base
)
3032 return '"%d":' % i
if v
.endswith(':') else '%d' % i
3039 code
= re
.sub(r
'new Date\((".+")\)', r
'\g<1>', code
)
3041 return re
.sub(r
'''(?sx)
3042 "(?:[^"\\]*(?:\\\\|\\['"nurtbfx/\n]))*[^"\\]*"|
3043 '(?:[^'\\]*(?:\\\\|\\['"nurtbfx/\n]))*[^'\\]*'|
3044 {comment}|,(?={skip}[\]}}])|
3045 void\s0|(?:(?<![0-9])[eE]|[a-df-zA-DF-Z_$])[.a-zA-Z_$0-9]*|
3046 \b(?:0[xX][0-9a-fA-F]+|0+[0-7]+)(?:{skip}:)?|
3049 '''.format(comment
=COMMENT_RE
, skip
=SKIP_RE
), fix_kv
, code
)
3052 def qualities(quality_ids
):
3053 """ Get a numeric quality value out of a list of possible values """
3056 return quality_ids
.index(qid
)
3062 POSTPROCESS_WHEN
= {'pre_process', 'after_filter', 'before_dl', 'after_move', 'post_process', 'after_video', 'playlist'}
3066 'default': '%(title)s [%(id)s].%(ext)s',
3067 'chapter': '%(title)s - %(section_number)03d %(section_title)s [%(id)s].%(ext)s',
3073 'description': 'description',
3074 'annotation': 'annotations.xml',
3075 'infojson': 'info.json',
3078 'pl_thumbnail': None,
3079 'pl_description': 'description',
3080 'pl_infojson': 'info.json',
3083 # As of [1] format syntax is:
3084 # %[mapping_key][conversion_flags][minimum_width][.precision][length_modifier]type
3085 # 1. https://docs.python.org/2/library/stdtypes.html#string-formatting
3086 STR_FORMAT_RE_TMPL
= r
'''(?x)
3087 (?<!%)(?P<prefix>(?:%%)*)
3089 (?P<has_key>\((?P<key>{0})\))?
3091 (?P<conversion>[#0\-+ ]+)?
3093 (?P<precision>\.\d+)?
3094 (?P<len_mod>[hlL])? # unused in python
3095 {1} # conversion type
3100 STR_FORMAT_TYPES
= 'diouxXeEfFgGcrs'
3103 def limit_length(s
, length
):
3104 """ Add ellipses to overly long strings """
3109 return s
[:length
- len(ELLIPSES
)] + ELLIPSES
3113 def version_tuple(v
):
3114 return tuple(int(e
) for e
in re
.split(r
'[-.]', v
))
3117 def is_outdated_version(version
, limit
, assume_new
=True):
3119 return not assume_new
3121 return version_tuple(version
) < version_tuple(limit
)
3123 return not assume_new
3126 def ytdl_is_updateable():
3127 """ Returns if yt-dlp can be updated with -U """
3129 from .update
import is_non_updateable
3131 return not is_non_updateable()
3134 def args_to_str(args
):
3135 # Get a short string representation for a subprocess command
3136 return ' '.join(compat_shlex_quote(a
) for a
in args
)
3139 def error_to_compat_str(err
):
3143 def error_to_str(err
):
3144 return f
'{type(err).__name__}: {err}'
3147 def mimetype2ext(mt
):
3151 mt
, _
, params
= mt
.partition(';')
3156 # Per RFC 3003, audio/mpeg can be .mp1, .mp2 or .mp3. Here use .mp3 as
3157 # it's the most popular one
3158 'audio/mpeg': 'mp3',
3159 'audio/x-wav': 'wav',
3161 'audio/wave': 'wav',
3164 ext
= FULL_MAP
.get(mt
)
3170 'smptett+xml': 'tt',
3174 'x-mp4-fragmented': 'mp4',
3175 'x-ms-sami': 'sami',
3178 'x-mpegurl': 'm3u8',
3179 'vnd.apple.mpegurl': 'm3u8',
3183 'vnd.ms-sstr+xml': 'ism',
3187 'filmstrip+json': 'fs',
3191 _
, _
, subtype
= mt
.rpartition('/')
3192 ext
= SUBTYPE_MAP
.get(subtype
.lower())
3203 _
, _
, suffix
= subtype
.partition('+')
3204 ext
= SUFFIX_MAP
.get(suffix
)
3208 return subtype
.replace('+', '.')
3211 def ext2mimetype(ext_or_url
):
3214 if '.' not in ext_or_url
:
3215 ext_or_url
= f
'file.{ext_or_url}'
3216 return mimetypes
.guess_type(ext_or_url
)[0]
3219 def parse_codecs(codecs_str
):
3220 # http://tools.ietf.org/html/rfc6381
3223 split_codecs
= list(filter(None, map(
3224 str.strip
, codecs_str
.strip().strip(',').split(','))))
3225 vcodec
, acodec
, tcodec
, hdr
= None, None, None, None
3226 for full_codec
in split_codecs
:
3227 parts
= full_codec
.split('.')
3228 codec
= parts
[0].replace('0', '')
3229 if codec
in ('avc1', 'avc2', 'avc3', 'avc4', 'vp9', 'vp8', 'hev1', 'hev2',
3230 'h263', 'h264', 'mp4v', 'hvc1', 'av1', 'theora', 'dvh1', 'dvhe'):
3232 vcodec
= '.'.join(parts
[:4]) if codec
in ('vp9', 'av1', 'hvc1') else full_codec
3233 if codec
in ('dvh1', 'dvhe'):
3235 elif codec
== 'av1' and len(parts
) > 3 and parts
[3] == '10':
3237 elif full_codec
.replace('0', '').startswith('vp9.2'):
3239 elif codec
in ('flac', 'mp4a', 'opus', 'vorbis', 'mp3', 'aac', 'ac-3', 'ec-3', 'eac3', 'dtsc', 'dtse', 'dtsh', 'dtsl'):
3242 elif codec
in ('stpp', 'wvtt',):
3246 write_string('WARNING: Unknown codec %s\n' % full_codec
, sys
.stderr
)
3247 if vcodec
or acodec
or tcodec
:
3249 'vcodec': vcodec
or 'none',
3250 'acodec': acodec
or 'none',
3251 'dynamic_range': hdr
,
3252 **({'tcodec': tcodec}
if tcodec
is not None else {}),
3254 elif len(split_codecs
) == 2:
3256 'vcodec': split_codecs
[0],
3257 'acodec': split_codecs
[1],
3262 def urlhandle_detect_ext(url_handle
):
3263 getheader
= url_handle
.headers
.get
3265 cd
= getheader('Content-Disposition')
3267 m
= re
.match(r
'attachment;\s*filename="(?P<filename>[^"]+)"', cd
)
3269 e
= determine_ext(m
.group('filename'), default_ext
=None)
3273 return mimetype2ext(getheader('Content-Type'))
3276 def encode_data_uri(data
, mime_type
):
3277 return 'data:%s;base64,%s' % (mime_type
, base64
.b64encode(data
).decode('ascii'))
3280 def age_restricted(content_limit
, age_limit
):
3281 """ Returns True iff the content should be blocked """
3283 if age_limit
is None: # No limit set
3285 if content_limit
is None:
3286 return False # Content available for everyone
3287 return age_limit
< content_limit
3290 def is_html(first_bytes
):
3291 """ Detect whether a file contains HTML by examining its first bytes. """
3294 (b
'\xef\xbb\xbf', 'utf-8'),
3295 (b
'\x00\x00\xfe\xff', 'utf-32-be'),
3296 (b
'\xff\xfe\x00\x00', 'utf-32-le'),
3297 (b
'\xff\xfe', 'utf-16-le'),
3298 (b
'\xfe\xff', 'utf-16-be'),
3300 for bom
, enc
in BOMS
:
3301 if first_bytes
.startswith(bom
):
3302 s
= first_bytes
[len(bom
):].decode(enc
, 'replace')
3305 s
= first_bytes
.decode('utf-8', 'replace')
3307 return re
.match(r
'^\s*<', s
)
3310 def determine_protocol(info_dict
):
3311 protocol
= info_dict
.get('protocol')
3312 if protocol
is not None:
3315 url
= sanitize_url(info_dict
['url'])
3316 if url
.startswith('rtmp'):
3318 elif url
.startswith('mms'):
3320 elif url
.startswith('rtsp'):
3323 ext
= determine_ext(url
)
3329 return compat_urllib_parse_urlparse(url
).scheme
3332 def render_table(header_row
, data
, delim
=False, extra_gap
=0, hide_empty
=False):
3333 """ Render a list of rows, each as a list of values.
3334 Text after a \t will be right aligned """
3336 return len(remove_terminal_sequences(string
).replace('\t', ''))
3338 def get_max_lens(table
):
3339 return [max(width(str(v
)) for v
in col
) for col
in zip(*table
)]
3341 def filter_using_list(row
, filterArray
):
3342 return [col
for take
, col
in itertools
.zip_longest(filterArray
, row
, fillvalue
=True) if take
]
3344 max_lens
= get_max_lens(data
) if hide_empty
else []
3345 header_row
= filter_using_list(header_row
, max_lens
)
3346 data
= [filter_using_list(row
, max_lens
) for row
in data
]
3348 table
= [header_row
] + data
3349 max_lens
= get_max_lens(table
)
3352 table
= [header_row
, [delim
* (ml
+ extra_gap
) for ml
in max_lens
]] + data
3353 table
[1][-1] = table
[1][-1][:-extra_gap
* len(delim
)] # Remove extra_gap from end of delimiter
3355 for pos
, text
in enumerate(map(str, row
)):
3357 row
[pos
] = text
.replace('\t', ' ' * (max_lens
[pos
] - width(text
))) + ' ' * extra_gap
3359 row
[pos
] = text
+ ' ' * (max_lens
[pos
] - width(text
) + extra_gap
)
3360 ret
= '\n'.join(''.join(row
).rstrip() for row
in table
)
3364 def _match_one(filter_part
, dct
, incomplete
):
3365 # TODO: Generalize code with YoutubeDL._build_format_filter
3366 STRING_OPERATORS
= {
3367 '*=': operator
.contains
,
3368 '^=': lambda attr
, value
: attr
.startswith(value
),
3369 '$=': lambda attr
, value
: attr
.endswith(value
),
3370 '~=': lambda attr
, value
: re
.search(value
, attr
),
3372 COMPARISON_OPERATORS
= {
3374 '<=': operator
.le
, # "<=" must be defined above "<"
3381 if isinstance(incomplete
, bool):
3382 is_incomplete
= lambda _
: incomplete
3384 is_incomplete
= lambda k
: k
in incomplete
3386 operator_rex
= re
.compile(r
'''(?x)\s*
3388 \s*(?P<negation>!\s*)?(?P<op>%s)(?P<none_inclusive>\s*\?)?\s*
3390 (?P<quote>["\'])(?P
<quotedstrval
>.+?
)(?P
=quote
)|
3394 ''' % '|'.join(map(re.escape, COMPARISON_OPERATORS.keys())))
3395 m = operator_rex.search(filter_part)
3398 unnegated_op = COMPARISON_OPERATORS[m['op']]
3400 op = lambda attr, value: not unnegated_op(attr, value)
3403 comparison_value = m['quotedstrval'] or m['strval'] or m['intval']
3405 comparison_value = comparison_value.replace(r'\%s' % m['quote'], m['quote'])
3406 actual_value = dct.get(m['key'])
3407 numeric_comparison = None
3408 if isinstance(actual_value, compat_numeric_types):
3409 # If the original field is a string and matching comparisonvalue is
3410 # a number we should respect the origin of the original field
3411 # and process comparison value as a string (see
3412 # https://github.com/ytdl-org/youtube-dl/issues/11082)
3414 numeric_comparison = int(comparison_value)
3416 numeric_comparison = parse_filesize(comparison_value)
3417 if numeric_comparison is None:
3418 numeric_comparison = parse_filesize(f'{comparison_value}B')
3419 if numeric_comparison is None:
3420 numeric_comparison = parse_duration(comparison_value)
3421 if numeric_comparison is not None and m['op'] in STRING_OPERATORS:
3422 raise ValueError('Operator %s only supports string values!' % m['op'])
3423 if actual_value is None:
3424 return is_incomplete(m['key']) or m['none_inclusive']
3425 return op(actual_value, comparison_value if numeric_comparison is None else numeric_comparison)
3428 '': lambda v: (v is True) if isinstance(v, bool) else (v is not None),
3429 '!': lambda v: (v is False) if isinstance(v, bool) else (v is None),
3431 operator_rex = re.compile(r'''(?x
)\s
*
3432 (?P
<op
>%s)\s
*(?P
<key
>[a
-z_
]+)
3434 ''' % '|'.join(map(re.escape, UNARY_OPERATORS.keys())))
3435 m = operator_rex.search(filter_part)
3437 op = UNARY_OPERATORS[m.group('op')]
3438 actual_value = dct.get(m.group('key'))
3439 if is_incomplete(m.group('key')) and actual_value is None:
3441 return op(actual_value)
3443 raise ValueError('Invalid filter part %r' % filter_part)
3446 def match_str(filter_str, dct, incomplete=False):
3447 """ Filter a dictionary with a simple string syntax.
3448 @returns Whether the filter passes
3449 @param incomplete Set of keys that is expected to be missing from dct.
3450 Can be True/False to indicate all/none of the keys may be missing.
3451 All conditions on incomplete keys pass if the key is missing
3454 _match_one(filter_part.replace(r'\&', '&'), dct, incomplete)
3455 for filter_part in re.split(r'(?<!\\)&', filter_str))
3458 def match_filter_func(filters):
3461 filters = variadic(filters)
3463 def _match_func(info_dict, *args, **kwargs):
3464 if any(match_str(f, info_dict, *args, **kwargs) for f in filters):
3467 video_title = info_dict.get('title') or info_dict.get('id') or 'video'
3468 filter_str = ') | ('.join(map(str.strip, filters))
3469 return f'{video_title} does not pass filter ({filter_str}), skipping ..'
3473 def parse_dfxp_time_expr(time_expr):
3477 mobj = re.match(r'^(?P<time_offset>\d+(?:\.\d+)?)s?$', time_expr)
3479 return float(mobj.group('time_offset'))
3481 mobj = re.match(r'^(\d+):(\d\d):(\d\d(?:(?:\.|:)\d+)?)$', time_expr)
3483 return 3600 * int(mobj.group(1)) + 60 * int(mobj.group(2)) + float(mobj.group(3).replace(':', '.'))
3486 def srt_subtitles_timecode(seconds):
3487 return '%02d:%02d:%02d,%03d' % timetuple_from_msec(seconds * 1000)
3490 def ass_subtitles_timecode(seconds):
3491 time = timetuple_from_msec(seconds * 1000)
3492 return '%01d:%02d:%02d.%02d' % (*time[:-1], time.milliseconds / 10)
3495 def dfxp2srt(dfxp_data):
3497 @param dfxp_data A
bytes-like
object containing DFXP data
3498 @returns A
unicode object containing converted SRT data
3500 LEGACY_NAMESPACES = (
3501 (b'http://www.w3.org/ns/ttml', [
3502 b'http://www.w3.org/2004/11/ttaf1',
3503 b'http://www.w3.org/2006/04/ttaf1',
3504 b'http://www.w3.org/2006/10/ttaf1',
3506 (b'http://www.w3.org/ns/ttml#styling', [
3507 b'http://www.w3.org/ns/ttml#style',
3511 SUPPORTED_STYLING = [
3520 _x = functools.partial(xpath_with_ns, ns_map={
3521 'xml': 'http://www.w3.org/XML/1998/namespace',
3522 'ttml': 'http://www.w3.org/ns/ttml',
3523 'tts': 'http://www.w3.org/ns/ttml#styling',
3529 class TTMLPElementParser(object):
3531 _unclosed_elements = []
3532 _applied_styles = []
3534 def start(self, tag, attrib):
3535 if tag in (_x('ttml:br'), 'br'):
3538 unclosed_elements = []
3540 element_style_id = attrib.get('style')
3542 style.update(default_style)
3543 if element_style_id:
3544 style.update(styles.get(element_style_id, {}))
3545 for prop in SUPPORTED_STYLING:
3546 prop_val = attrib.get(_x('tts:' + prop))
3548 style[prop] = prop_val
3551 for k, v in sorted(style.items()):
3552 if self._applied_styles and self._applied_styles[-1].get(k) == v:
3555 font += ' color="%s"' % v
3556 elif k == 'fontSize':
3557 font += ' size="%s"' % v
3558 elif k == 'fontFamily':
3559 font += ' face="%s"' % v
3560 elif k == 'fontWeight' and v == 'bold':
3562 unclosed_elements.append('b')
3563 elif k == 'fontStyle' and v == 'italic':
3565 unclosed_elements.append('i')
3566 elif k == 'textDecoration' and v == 'underline':
3568 unclosed_elements.append('u')
3570 self._out += '<font' + font + '>'
3571 unclosed_elements.append('font')
3573 if self._applied_styles:
3574 applied_style.update(self._applied_styles[-1])
3575 applied_style.update(style)
3576 self._applied_styles.append(applied_style)
3577 self._unclosed_elements.append(unclosed_elements)
3580 if tag not in (_x('ttml:br'), 'br'):
3581 unclosed_elements = self._unclosed_elements.pop()
3582 for element in reversed(unclosed_elements):
3583 self._out += '</%s>' % element
3584 if unclosed_elements and self._applied_styles:
3585 self._applied_styles.pop()
3587 def data(self, data):
3591 return self._out.strip()
3593 def parse_node(node):
3594 target = TTMLPElementParser()
3595 parser = xml.etree.ElementTree.XMLParser(target=target)
3596 parser.feed(xml.etree.ElementTree.tostring(node))
3597 return parser.close()
3599 for k, v in LEGACY_NAMESPACES:
3601 dfxp_data = dfxp_data.replace(ns, k)
3603 dfxp = compat_etree_fromstring(dfxp_data)
3605 paras = dfxp.findall(_x('.//ttml:p')) or dfxp.findall('.//p')
3608 raise ValueError('Invalid dfxp/TTML subtitle')
3612 for style in dfxp.findall(_x('.//ttml:style')):
3613 style_id = style.get('id') or style.get(_x('xml:id'))
3616 parent_style_id = style.get('style')
3618 if parent_style_id not in styles:
3621 styles[style_id] = styles[parent_style_id].copy()
3622 for prop in SUPPORTED_STYLING:
3623 prop_val = style.get(_x('tts:' + prop))
3625 styles.setdefault(style_id, {})[prop] = prop_val
3631 for p in ('body', 'div'):
3632 ele = xpath_element(dfxp, [_x('.//ttml:' + p), './/' + p])
3635 style = styles.get(ele.get('style'))
3638 default_style.update(style)
3640 for para, index in zip(paras, itertools.count(1)):
3641 begin_time = parse_dfxp_time_expr(para.attrib.get('begin'))
3642 end_time = parse_dfxp_time_expr(para.attrib.get('end'))
3643 dur = parse_dfxp_time_expr(para.attrib.get('dur'))
3644 if begin_time is None:
3649 end_time = begin_time + dur
3650 out.append('%d\n%s --> %s\n%s\n\n' % (
3652 srt_subtitles_timecode(begin_time),
3653 srt_subtitles_timecode(end_time),
3659 def cli_option(params, command_option, param):
3660 param = params.get(param)
3662 param = compat_str(param)
3663 return [command_option, param] if param is not None else []
3666 def cli_bool_option(params, command_option, param, true_value='true', false_value='false', separator=None):
3667 param = params.get(param)
3670 assert isinstance(param, bool)
3672 return [command_option + separator + (true_value if param else false_value)]
3673 return [command_option, true_value if param else false_value]
3676 def cli_valueless_option(params, command_option, param, expected_value=True):
3677 param = params.get(param)
3678 return [command_option] if param == expected_value else []
3681 def cli_configuration_args(argdict, keys, default=[], use_compat=True):
3682 if isinstance(argdict, (list, tuple)): # for backward compatibility
3689 assert isinstance(argdict, dict)
3691 assert isinstance(keys, (list, tuple))
3692 for key_list in keys:
3693 arg_list = list(filter(
3694 lambda x: x is not None,
3695 [argdict.get(key.lower()) for key in variadic(key_list)]))
3697 return [arg for args in arg_list for arg in args]
3701 def _configuration_args(main_key, argdict, exe, keys=None, default=[], use_compat=True):
3702 main_key, exe = main_key.lower(), exe.lower()
3703 root_key = exe if main_key == exe else f'{main_key}+{exe}'
3704 keys = [f'{root_key}{k}' for k in (keys or [''])]
3705 if root_key in keys:
3707 keys.append((main_key, exe))
3708 keys.append('default')
3711 return cli_configuration_args(argdict, keys, default, use_compat)
3714 class ISO639Utils(object):
3715 # See http://www.loc.gov/standards/iso639-2/ISO-639-2_utf-8.txt
3774 'iw': 'heb', # Replaced by he in 1989 revision
3784 'in': 'ind', # Replaced by id in 1989 revision
3899 'ji': 'yid', # Replaced by yi in 1989 revision
3907 def short2long(cls, code):
3908 """Convert language code from ISO 639-1 to ISO 639-2/T"""
3909 return cls._lang_map.get(code[:2])
3912 def long2short(cls, code):
3913 """Convert language code from ISO 639-2/T to ISO 639-1"""
3914 for short_name, long_name in cls._lang_map.items():
3915 if long_name == code:
3919 class ISO3166Utils(object):
3920 # From http://data.okfn.org/data/core/country-list
3922 'AF': 'Afghanistan',
3923 'AX': 'Åland Islands',
3926 'AS': 'American Samoa',
3931 'AG': 'Antigua and Barbuda',
3948 'BO': 'Bolivia, Plurinational State of',
3949 'BQ': 'Bonaire, Sint Eustatius and Saba',
3950 'BA': 'Bosnia and Herzegovina',
3952 'BV': 'Bouvet Island',
3954 'IO': 'British Indian Ocean Territory',
3955 'BN': 'Brunei Darussalam',
3957 'BF': 'Burkina Faso',
3963 'KY': 'Cayman Islands',
3964 'CF': 'Central African Republic',
3968 'CX': 'Christmas Island',
3969 'CC': 'Cocos (Keeling) Islands',
3973 'CD': 'Congo, the Democratic Republic of the',
3974 'CK': 'Cook Islands',
3976 'CI': 'Côte d\'Ivoire',
3981 'CZ': 'Czech Republic',
3985 'DO': 'Dominican Republic',
3988 'SV': 'El Salvador',
3989 'GQ': 'Equatorial Guinea',
3993 'FK': 'Falkland Islands (Malvinas)',
3994 'FO': 'Faroe Islands',
3998 'GF': 'French Guiana',
3999 'PF': 'French Polynesia',
4000 'TF': 'French Southern Territories',
4015 'GW': 'Guinea-Bissau',
4018 'HM': 'Heard Island and McDonald Islands',
4019 'VA': 'Holy See (Vatican City State)',
4026 'IR': 'Iran, Islamic Republic of',
4029 'IM': 'Isle of Man',
4039 'KP': 'Korea, Democratic People\'s Republic of',
4040 'KR': 'Korea, Republic of',
4043 'LA': 'Lao People\'s Democratic Republic',
4049 'LI': 'Liechtenstein',
4053 'MK': 'Macedonia, the Former Yugoslav Republic of',
4060 'MH': 'Marshall Islands',
4066 'FM': 'Micronesia, Federated States of',
4067 'MD': 'Moldova, Republic of',
4078 'NL': 'Netherlands',
4079 'NC': 'New Caledonia',
4080 'NZ': 'New Zealand',
4085 'NF': 'Norfolk Island',
4086 'MP': 'Northern Mariana Islands',
4091 'PS': 'Palestine, State of',
4093 'PG': 'Papua New Guinea',
4096 'PH': 'Philippines',
4100 'PR': 'Puerto Rico',
4104 'RU': 'Russian Federation',
4106 'BL': 'Saint Barthélemy',
4107 'SH': 'Saint Helena, Ascension and Tristan da Cunha',
4108 'KN': 'Saint Kitts and Nevis',
4109 'LC': 'Saint Lucia',
4110 'MF': 'Saint Martin (French part)',
4111 'PM': 'Saint Pierre and Miquelon',
4112 'VC': 'Saint Vincent and the Grenadines',
4115 'ST': 'Sao Tome and Principe',
4116 'SA': 'Saudi Arabia',
4120 'SL': 'Sierra Leone',
4122 'SX': 'Sint Maarten (Dutch part)',
4125 'SB': 'Solomon Islands',
4127 'ZA': 'South Africa',
4128 'GS': 'South Georgia and the South Sandwich Islands',
4129 'SS': 'South Sudan',
4134 'SJ': 'Svalbard and Jan Mayen',
4137 'CH': 'Switzerland',
4138 'SY': 'Syrian Arab Republic',
4139 'TW': 'Taiwan, Province of China',
4141 'TZ': 'Tanzania, United Republic of',
4143 'TL': 'Timor-Leste',
4147 'TT': 'Trinidad and Tobago',
4150 'TM': 'Turkmenistan',
4151 'TC': 'Turks and Caicos Islands',
4155 'AE': 'United Arab Emirates',
4156 'GB': 'United Kingdom',
4157 'US': 'United States',
4158 'UM': 'United States Minor Outlying Islands',
4162 'VE': 'Venezuela, Bolivarian Republic of',
4164 'VG': 'Virgin Islands, British',
4165 'VI': 'Virgin Islands, U.S.',
4166 'WF': 'Wallis and Futuna',
4167 'EH': 'Western Sahara',
4174 def short2full(cls, code):
4175 """Convert an ISO 3166-2 country code to the corresponding full name"""
4176 return cls._country_map.get(code.upper())
4179 class GeoUtils(object):
4180 # Major IPv4 address blocks per country
4182 'AD': '46.172.224.0/19',
4183 'AE': '94.200.0.0/13',
4184 'AF': '149.54.0.0/17',
4185 'AG': '209.59.64.0/18',
4186 'AI': '204.14.248.0/21',
4187 'AL': '46.99.0.0/16',
4188 'AM': '46.70.0.0/15',
4189 'AO': '105.168.0.0/13',
4190 'AP': '182.50.184.0/21',
4191 'AQ': '23.154.160.0/24',
4192 'AR': '181.0.0.0/12',
4193 'AS': '202.70.112.0/20',
4194 'AT': '77.116.0.0/14',
4195 'AU': '1.128.0.0/11',
4196 'AW': '181.41.0.0/18',
4197 'AX': '185.217.4.0/22',
4198 'AZ': '5.197.0.0/16',
4199 'BA': '31.176.128.0/17',
4200 'BB': '65.48.128.0/17',
4201 'BD': '114.130.0.0/16',
4203 'BF': '102.178.0.0/15',
4204 'BG': '95.42.0.0/15',
4205 'BH': '37.131.0.0/17',
4206 'BI': '154.117.192.0/18',
4207 'BJ': '137.255.0.0/16',
4208 'BL': '185.212.72.0/23',
4209 'BM': '196.12.64.0/18',
4210 'BN': '156.31.0.0/16',
4211 'BO': '161.56.0.0/16',
4212 'BQ': '161.0.80.0/20',
4213 'BR': '191.128.0.0/12',
4214 'BS': '24.51.64.0/18',
4215 'BT': '119.2.96.0/19',
4216 'BW': '168.167.0.0/16',
4217 'BY': '178.120.0.0/13',
4218 'BZ': '179.42.192.0/18',
4219 'CA': '99.224.0.0/11',
4220 'CD': '41.243.0.0/16',
4221 'CF': '197.242.176.0/21',
4222 'CG': '160.113.0.0/16',
4223 'CH': '85.0.0.0/13',
4224 'CI': '102.136.0.0/14',
4225 'CK': '202.65.32.0/19',
4226 'CL': '152.172.0.0/14',
4227 'CM': '102.244.0.0/14',
4228 'CN': '36.128.0.0/10',
4229 'CO': '181.240.0.0/12',
4230 'CR': '201.192.0.0/12',
4231 'CU': '152.206.0.0/15',
4232 'CV': '165.90.96.0/19',
4233 'CW': '190.88.128.0/17',
4234 'CY': '31.153.0.0/16',
4235 'CZ': '88.100.0.0/14',
4237 'DJ': '197.241.0.0/17',
4238 'DK': '87.48.0.0/12',
4239 'DM': '192.243.48.0/20',
4240 'DO': '152.166.0.0/15',
4241 'DZ': '41.96.0.0/12',
4242 'EC': '186.68.0.0/15',
4243 'EE': '90.190.0.0/15',
4244 'EG': '156.160.0.0/11',
4245 'ER': '196.200.96.0/20',
4246 'ES': '88.0.0.0/11',
4247 'ET': '196.188.0.0/14',
4248 'EU': '2.16.0.0/13',
4249 'FI': '91.152.0.0/13',
4250 'FJ': '144.120.0.0/16',
4251 'FK': '80.73.208.0/21',
4252 'FM': '119.252.112.0/20',
4253 'FO': '88.85.32.0/19',
4255 'GA': '41.158.0.0/15',
4257 'GD': '74.122.88.0/21',
4258 'GE': '31.146.0.0/16',
4259 'GF': '161.22.64.0/18',
4260 'GG': '62.68.160.0/19',
4261 'GH': '154.160.0.0/12',
4262 'GI': '95.164.0.0/16',
4263 'GL': '88.83.0.0/19',
4264 'GM': '160.182.0.0/15',
4265 'GN': '197.149.192.0/18',
4266 'GP': '104.250.0.0/19',
4267 'GQ': '105.235.224.0/20',
4268 'GR': '94.64.0.0/13',
4269 'GT': '168.234.0.0/16',
4270 'GU': '168.123.0.0/16',
4271 'GW': '197.214.80.0/20',
4272 'GY': '181.41.64.0/18',
4273 'HK': '113.252.0.0/14',
4274 'HN': '181.210.0.0/16',
4275 'HR': '93.136.0.0/13',
4276 'HT': '148.102.128.0/17',
4277 'HU': '84.0.0.0/14',
4278 'ID': '39.192.0.0/10',
4279 'IE': '87.32.0.0/12',
4280 'IL': '79.176.0.0/13',
4281 'IM': '5.62.80.0/20',
4282 'IN': '117.192.0.0/10',
4283 'IO': '203.83.48.0/21',
4284 'IQ': '37.236.0.0/14',
4285 'IR': '2.176.0.0/12',
4286 'IS': '82.221.0.0/16',
4287 'IT': '79.0.0.0/10',
4288 'JE': '87.244.64.0/18',
4289 'JM': '72.27.0.0/17',
4290 'JO': '176.29.0.0/16',
4291 'JP': '133.0.0.0/8',
4292 'KE': '105.48.0.0/12',
4293 'KG': '158.181.128.0/17',
4294 'KH': '36.37.128.0/17',
4295 'KI': '103.25.140.0/22',
4296 'KM': '197.255.224.0/20',
4297 'KN': '198.167.192.0/19',
4298 'KP': '175.45.176.0/22',
4299 'KR': '175.192.0.0/10',
4300 'KW': '37.36.0.0/14',
4301 'KY': '64.96.0.0/15',
4302 'KZ': '2.72.0.0/13',
4303 'LA': '115.84.64.0/18',
4304 'LB': '178.135.0.0/16',
4305 'LC': '24.92.144.0/20',
4306 'LI': '82.117.0.0/19',
4307 'LK': '112.134.0.0/15',
4308 'LR': '102.183.0.0/16',
4309 'LS': '129.232.0.0/17',
4310 'LT': '78.56.0.0/13',
4311 'LU': '188.42.0.0/16',
4312 'LV': '46.109.0.0/16',
4313 'LY': '41.252.0.0/14',
4314 'MA': '105.128.0.0/11',
4315 'MC': '88.209.64.0/18',
4316 'MD': '37.246.0.0/16',
4317 'ME': '178.175.0.0/17',
4318 'MF': '74.112.232.0/21',
4319 'MG': '154.126.0.0/17',
4320 'MH': '117.103.88.0/21',
4321 'MK': '77.28.0.0/15',
4322 'ML': '154.118.128.0/18',
4323 'MM': '37.111.0.0/17',
4324 'MN': '49.0.128.0/17',
4325 'MO': '60.246.0.0/16',
4326 'MP': '202.88.64.0/20',
4327 'MQ': '109.203.224.0/19',
4328 'MR': '41.188.64.0/18',
4329 'MS': '208.90.112.0/22',
4330 'MT': '46.11.0.0/16',
4331 'MU': '105.16.0.0/12',
4332 'MV': '27.114.128.0/18',
4333 'MW': '102.70.0.0/15',
4334 'MX': '187.192.0.0/11',
4335 'MY': '175.136.0.0/13',
4336 'MZ': '197.218.0.0/15',
4337 'NA': '41.182.0.0/16',
4338 'NC': '101.101.0.0/18',
4339 'NE': '197.214.0.0/18',
4340 'NF': '203.17.240.0/22',
4341 'NG': '105.112.0.0/12',
4342 'NI': '186.76.0.0/15',
4343 'NL': '145.96.0.0/11',
4344 'NO': '84.208.0.0/13',
4345 'NP': '36.252.0.0/15',
4346 'NR': '203.98.224.0/19',
4347 'NU': '49.156.48.0/22',
4348 'NZ': '49.224.0.0/14',
4349 'OM': '5.36.0.0/15',
4350 'PA': '186.72.0.0/15',
4351 'PE': '186.160.0.0/14',
4352 'PF': '123.50.64.0/18',
4353 'PG': '124.240.192.0/19',
4354 'PH': '49.144.0.0/13',
4355 'PK': '39.32.0.0/11',
4356 'PL': '83.0.0.0/11',
4357 'PM': '70.36.0.0/20',
4358 'PR': '66.50.0.0/16',
4359 'PS': '188.161.0.0/16',
4360 'PT': '85.240.0.0/13',
4361 'PW': '202.124.224.0/20',
4362 'PY': '181.120.0.0/14',
4363 'QA': '37.210.0.0/15',
4364 'RE': '102.35.0.0/16',
4365 'RO': '79.112.0.0/13',
4366 'RS': '93.86.0.0/15',
4367 'RU': '5.136.0.0/13',
4368 'RW': '41.186.0.0/16',
4369 'SA': '188.48.0.0/13',
4370 'SB': '202.1.160.0/19',
4371 'SC': '154.192.0.0/11',
4372 'SD': '102.120.0.0/13',
4373 'SE': '78.64.0.0/12',
4374 'SG': '8.128.0.0/10',
4375 'SI': '188.196.0.0/14',
4376 'SK': '78.98.0.0/15',
4377 'SL': '102.143.0.0/17',
4378 'SM': '89.186.32.0/19',
4379 'SN': '41.82.0.0/15',
4380 'SO': '154.115.192.0/18',
4381 'SR': '186.179.128.0/17',
4382 'SS': '105.235.208.0/21',
4383 'ST': '197.159.160.0/19',
4384 'SV': '168.243.0.0/16',
4385 'SX': '190.102.0.0/20',
4387 'SZ': '41.84.224.0/19',
4388 'TC': '65.255.48.0/20',
4389 'TD': '154.68.128.0/19',
4390 'TG': '196.168.0.0/14',
4391 'TH': '171.96.0.0/13',
4392 'TJ': '85.9.128.0/18',
4393 'TK': '27.96.24.0/21',
4394 'TL': '180.189.160.0/20',
4395 'TM': '95.85.96.0/19',
4396 'TN': '197.0.0.0/11',
4397 'TO': '175.176.144.0/21',
4398 'TR': '78.160.0.0/11',
4399 'TT': '186.44.0.0/15',
4400 'TV': '202.2.96.0/19',
4401 'TW': '120.96.0.0/11',
4402 'TZ': '156.156.0.0/14',
4403 'UA': '37.52.0.0/14',
4404 'UG': '102.80.0.0/13',
4406 'UY': '167.56.0.0/13',
4407 'UZ': '84.54.64.0/18',
4408 'VA': '212.77.0.0/19',
4409 'VC': '207.191.240.0/21',
4410 'VE': '186.88.0.0/13',
4411 'VG': '66.81.192.0/20',
4412 'VI': '146.226.0.0/16',
4413 'VN': '14.160.0.0/11',
4414 'VU': '202.80.32.0/20',
4415 'WF': '117.20.32.0/21',
4416 'WS': '202.4.32.0/19',
4417 'YE': '134.35.0.0/16',
4418 'YT': '41.242.116.0/22',
4419 'ZA': '41.0.0.0/11',
4420 'ZM': '102.144.0.0/13',
4421 'ZW': '102.177.192.0/18',
4425 def random_ipv4(cls, code_or_block):
4426 if len(code_or_block) == 2:
4427 block = cls._country_ip_map.get(code_or_block.upper())
4431 block = code_or_block
4432 addr, preflen = block.split('/')
4433 addr_min = compat_struct_unpack('!L', socket.inet_aton(addr))[0]
4434 addr_max = addr_min | (0xffffffff >> int(preflen))
4435 return compat_str(socket.inet_ntoa(
4436 compat_struct_pack('!L', random.randint(addr_min, addr_max))))
4439 class PerRequestProxyHandler(compat_urllib_request.ProxyHandler):
4440 def __init__(self, proxies=None):
4441 # Set default handlers
4442 for type in ('http', 'https'):
4443 setattr(self, '%s_open' % type,
4444 lambda r, proxy='__noproxy__', type=type, meth=self.proxy_open:
4445 meth(r, proxy, type))
4446 compat_urllib_request.ProxyHandler.__init__(self, proxies)
4448 def proxy_open(self, req, proxy, type):
4449 req_proxy = req.headers.get('Ytdl-request-proxy')
4450 if req_proxy is not None:
4452 del req.headers['Ytdl-request-proxy']
4454 if proxy == '__noproxy__':
4455 return None # No Proxy
4456 if compat_urlparse.urlparse(proxy).scheme.lower() in ('socks', 'socks4', 'socks4a', 'socks5'):
4457 req.add_header('Ytdl-socks-proxy', proxy)
4458 # yt-dlp's http/https handlers do wrapping the socket with socks
4460 return compat_urllib_request.ProxyHandler.proxy_open(
4461 self, req, proxy, type)
4464 # Both long_to_bytes and bytes_to_long are adapted from PyCrypto, which is
4465 # released into Public Domain
4466 # https://github.com/dlitz/pycrypto/blob/master/lib/Crypto/Util/number.py#L387
4468 def long_to_bytes(n, blocksize=0):
4469 """long_to_bytes(n:long, blocksize:int) : string
4470 Convert a long integer to a byte string.
4472 If optional blocksize is given and greater than zero, pad the front of the
4473 byte string with binary zeros so that the length is a multiple of
4476 # after much testing, this algorithm was deemed to be the fastest
4480 s = compat_struct_pack('>I', n & 0xffffffff) + s
4482 # strip off leading zeros
4483 for i in range(len(s)):
4484 if s[i] != b'\000'[0]:
4487 # only happens when n == 0
4491 # add back some pad bytes. this could be done more efficiently w.r.t. the
4492 # de-padding being done above, but sigh...
4493 if blocksize > 0 and len(s) % blocksize:
4494 s = (blocksize - len(s) % blocksize) * b'\000' + s
4498 def bytes_to_long(s):
4499 """bytes_to_long(string) : long
4500 Convert a byte string to a long integer.
4502 This is (essentially) the inverse of long_to_bytes().
4507 extra = (4 - length % 4)
4508 s = b'\000' * extra + s
4509 length = length + extra
4510 for i in range(0, length, 4):
4511 acc = (acc << 32) + compat_struct_unpack('>I', s[i:i + 4])[0]
4515 def ohdave_rsa_encrypt(data, exponent, modulus):
4517 Implement OHDave
's RSA algorithm. See http://www.ohdave.com/rsa/
4520 data: data to encrypt, bytes-like object
4521 exponent, modulus: parameter e and N of RSA algorithm, both integer
4522 Output: hex string of encrypted data
4524 Limitation: supports one block encryption only
4527 payload = int(binascii.hexlify(data[::-1]), 16)
4528 encrypted = pow(payload, exponent, modulus)
4529 return '%x' % encrypted
4532 def pkcs1pad(data, length):
4534 Padding input data with PKCS#1 scheme
4536 @param {int[]} data input data
4537 @param {int} length target length
4538 @returns {int[]} padded data
4540 if len(data) > length - 11:
4541 raise ValueError('Input data too
long for PKCS
#1 padding')
4543 pseudo_random
= [random
.randint(0, 254) for _
in range(length
- len(data
) - 3)]
4544 return [0, 2] + pseudo_random
+ [0] + data
4547 def encode_base_n(num
, n
, table
=None):
4548 FULL_TABLE
= '0123456789abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ'
4550 table
= FULL_TABLE
[:n
]
4553 raise ValueError('base %d exceeds table length %d' % (n
, len(table
)))
4560 ret
= table
[num
% n
] + ret
4565 def decode_packed_codes(code
):
4566 mobj
= re
.search(PACKED_CODES_RE
, code
)
4567 obfuscated_code
, base
, count
, symbols
= mobj
.groups()
4570 symbols
= symbols
.split('|')
4575 base_n_count
= encode_base_n(count
, base
)
4576 symbol_table
[base_n_count
] = symbols
[count
] or base_n_count
4579 r
'\b(\w+)\b', lambda mobj
: symbol_table
[mobj
.group(0)],
4583 def caesar(s
, alphabet
, shift
):
4588 alphabet
[(alphabet
.index(c
) + shift
) % l
] if c
in alphabet
else c
4593 return caesar(s
, r
'''!"#$%&'()*+,-./0123456789:;<=>?@ABCDEFGHIJKLMNOPQRSTUVWXYZ[\]^_`abcdefghijklmnopqrstuvwxyz{|}~''', 47)
4596 def parse_m3u8_attributes(attrib
):
4598 for (key
, val
) in re
.findall(r
'(?P<key>[A-Z0-9-]+)=(?P<val>"[^"]+"|[^",]+)(?:,|$)', attrib
):
4599 if val
.startswith('"'):
4605 def urshift(val
, n
):
4606 return val
>> n
if val
>= 0 else (val
+ 0x100000000) >> n
4609 # Based on png2str() written by @gdkchan and improved by @yokrysty
4610 # Originally posted at https://github.com/ytdl-org/youtube-dl/issues/9706
4611 def decode_png(png_data
):
4612 # Reference: https://www.w3.org/TR/PNG/
4613 header
= png_data
[8:]
4615 if png_data
[:8] != b
'\x89PNG\x0d\x0a\x1a\x0a' or header
[4:8] != b
'IHDR':
4616 raise IOError('Not a valid PNG file.')
4618 int_map
= {1: '>B', 2: '>H', 4: '>I'}
4619 unpack_integer
= lambda x
: compat_struct_unpack(int_map
[len(x
)], x
)[0]
4624 length
= unpack_integer(header
[:4])
4627 chunk_type
= header
[:4]
4630 chunk_data
= header
[:length
]
4631 header
= header
[length
:]
4633 header
= header
[4:] # Skip CRC
4641 ihdr
= chunks
[0]['data']
4643 width
= unpack_integer(ihdr
[:4])
4644 height
= unpack_integer(ihdr
[4:8])
4648 for chunk
in chunks
:
4649 if chunk
['type'] == b
'IDAT':
4650 idat
+= chunk
['data']
4653 raise IOError('Unable to read PNG data.')
4655 decompressed_data
= bytearray(zlib
.decompress(idat
))
4660 def _get_pixel(idx
):
4665 for y
in range(height
):
4666 basePos
= y
* (1 + stride
)
4667 filter_type
= decompressed_data
[basePos
]
4671 pixels
.append(current_row
)
4673 for x
in range(stride
):
4674 color
= decompressed_data
[1 + basePos
+ x
]
4675 basex
= y
* stride
+ x
4680 left
= _get_pixel(basex
- 3)
4682 up
= _get_pixel(basex
- stride
)
4684 if filter_type
== 1: # Sub
4685 color
= (color
+ left
) & 0xff
4686 elif filter_type
== 2: # Up
4687 color
= (color
+ up
) & 0xff
4688 elif filter_type
== 3: # Average
4689 color
= (color
+ ((left
+ up
) >> 1)) & 0xff
4690 elif filter_type
== 4: # Paeth
4696 c
= _get_pixel(basex
- stride
- 3)
4704 if pa
<= pb
and pa
<= pc
:
4705 color
= (color
+ a
) & 0xff
4707 color
= (color
+ b
) & 0xff
4709 color
= (color
+ c
) & 0xff
4711 current_row
.append(color
)
4713 return width
, height
, pixels
4716 def write_xattr(path
, key
, value
):
4717 # This mess below finds the best xattr tool for the job
4719 # try the pyxattr module...
4722 if hasattr(xattr
, 'set'): # pyxattr
4723 # Unicode arguments are not supported in python-pyxattr until
4725 # See https://github.com/ytdl-org/youtube-dl/issues/5498
4726 pyxattr_required_version
= '0.5.0'
4727 if version_tuple(xattr
.__version
__) < version_tuple(pyxattr_required_version
):
4728 # TODO: fallback to CLI tools
4729 raise XAttrUnavailableError(
4730 'python-pyxattr is detected but is too old. '
4731 'yt-dlp requires %s or above while your version is %s. '
4732 'Falling back to other xattr implementations' % (
4733 pyxattr_required_version
, xattr
.__version
__))
4735 setxattr
= xattr
.set
4737 setxattr
= xattr
.setxattr
4740 setxattr(path
, key
, value
)
4741 except EnvironmentError as e
:
4742 raise XAttrMetadataError(e
.errno
, e
.strerror
)
4745 if compat_os_name
== 'nt':
4746 # Write xattrs to NTFS Alternate Data Streams:
4747 # http://en.wikipedia.org/wiki/NTFS#Alternate_data_streams_.28ADS.29
4748 assert ':' not in key
4749 assert os
.path
.exists(path
)
4751 ads_fn
= path
+ ':' + key
4753 with open(ads_fn
, 'wb') as f
:
4755 except EnvironmentError as e
:
4756 raise XAttrMetadataError(e
.errno
, e
.strerror
)
4758 user_has_setfattr
= check_executable('setfattr', ['--version'])
4759 user_has_xattr
= check_executable('xattr', ['-h'])
4761 if user_has_setfattr
or user_has_xattr
:
4763 value
= value
.decode('utf-8')
4764 if user_has_setfattr
:
4765 executable
= 'setfattr'
4766 opts
= ['-n', key
, '-v', value
]
4767 elif user_has_xattr
:
4768 executable
= 'xattr'
4769 opts
= ['-w', key
, value
]
4771 cmd
= ([encodeFilename(executable
, True)]
4772 + [encodeArgument(o
) for o
in opts
]
4773 + [encodeFilename(path
, True)])
4777 cmd
, stdout
=subprocess
.PIPE
, stderr
=subprocess
.PIPE
, stdin
=subprocess
.PIPE
)
4778 except EnvironmentError as e
:
4779 raise XAttrMetadataError(e
.errno
, e
.strerror
)
4780 stdout
, stderr
= p
.communicate_or_kill()
4781 stderr
= stderr
.decode('utf-8', 'replace')
4782 if p
.returncode
!= 0:
4783 raise XAttrMetadataError(p
.returncode
, stderr
)
4786 # On Unix, and can't find pyxattr, setfattr, or xattr.
4787 if sys
.platform
.startswith('linux'):
4788 raise XAttrUnavailableError(
4789 "Couldn't find a tool to set the xattrs. "
4790 "Install either the python 'pyxattr' or 'xattr' "
4791 "modules, or the GNU 'attr' package "
4792 "(which contains the 'setfattr' tool).")
4794 raise XAttrUnavailableError(
4795 "Couldn't find a tool to set the xattrs. "
4796 "Install either the python 'xattr' module, "
4797 "or the 'xattr' binary.")
4800 def random_birthday(year_field
, month_field
, day_field
):
4801 start_date
= datetime
.date(1950, 1, 1)
4802 end_date
= datetime
.date(1995, 12, 31)
4803 offset
= random
.randint(0, (end_date
- start_date
).days
)
4804 random_date
= start_date
+ datetime
.timedelta(offset
)
4806 year_field
: str(random_date
.year
),
4807 month_field
: str(random_date
.month
),
4808 day_field
: str(random_date
.day
),
4812 # Templates for internet shortcut files, which are plain text files.
4813 DOT_URL_LINK_TEMPLATE
= '''
4818 DOT_WEBLOC_LINK_TEMPLATE
= '''
4819 <?xml version="1.0" encoding="UTF-8"?>
4820 <!DOCTYPE plist PUBLIC "-//Apple//DTD PLIST 1.0//EN" "http://www.apple.com/DTDs/PropertyList-1.0.dtd">
4821 <plist version="1.0">
4824 \t<string>%(url)s</string>
4829 DOT_DESKTOP_LINK_TEMPLATE
= '''
4839 'url': DOT_URL_LINK_TEMPLATE
,
4840 'desktop': DOT_DESKTOP_LINK_TEMPLATE
,
4841 'webloc': DOT_WEBLOC_LINK_TEMPLATE
,
4845 def iri_to_uri(iri
):
4847 Converts an IRI (Internationalized Resource Identifier, allowing Unicode characters) to a URI (Uniform Resource Identifier, ASCII-only).
4849 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.
4852 iri_parts
= compat_urllib_parse_urlparse(iri
)
4854 if '[' in iri_parts
.netloc
:
4855 raise ValueError('IPv6 URIs are not, yet, supported.')
4856 # Querying `.netloc`, when there's only one bracket, also raises a ValueError.
4858 # 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.
4861 if iri_parts
.username
:
4862 net_location
+= compat_urllib_parse_quote(iri_parts
.username
, safe
=r
"!$%&'()*+,~")
4863 if iri_parts
.password
is not None:
4864 net_location
+= ':' + compat_urllib_parse_quote(iri_parts
.password
, safe
=r
"!$%&'()*+,~")
4867 net_location
+= iri_parts
.hostname
.encode('idna').decode('utf-8') # Punycode for Unicode hostnames.
4868 # The 'idna' encoding produces ASCII text.
4869 if iri_parts
.port
is not None and iri_parts
.port
!= 80:
4870 net_location
+= ':' + str(iri_parts
.port
)
4872 return compat_urllib_parse_urlunparse(
4876 compat_urllib_parse_quote_plus(iri_parts
.path
, safe
=r
"!$%&'()*+,/:;=@|~"),
4878 # Unsure about the `safe` argument, since this is a legacy way of handling parameters.
4879 compat_urllib_parse_quote_plus(iri_parts
.params
, safe
=r
"!$%&'()*+,/:;=@|~"),
4881 # Not totally sure about the `safe` argument, since the source does not explicitly mention the query URI component.
4882 compat_urllib_parse_quote_plus(iri_parts
.query
, safe
=r
"!$%&'()*+,/:;=?@{|}~"),
4884 compat_urllib_parse_quote_plus(iri_parts
.fragment
, safe
=r
"!#$%&'()*+,/:;=?@{|}~")))
4886 # Source for `safe` arguments: https://url.spec.whatwg.org/#percent-encoded-bytes.
4889 def to_high_limit_path(path
):
4890 if sys
.platform
in ['win32', 'cygwin']:
4891 # Work around MAX_PATH limitation on Windows. The maximum allowed length for the individual path segments may still be quite limited.
4892 return r
'\\?\ '.rstrip() + os
.path
.abspath(path
)
4897 def format_field(obj
, field
=None, template
='%s', ignore
=(None, ''), default
='', func
=None):
4898 val
= traverse_obj(obj
, *variadic(field
))
4901 return template
% (func(val
) if func
else val
)
4904 def clean_podcast_url(url
):
4905 return re
.sub(r
'''(?x)
4909 media\.blubrry\.com| # https://create.blubrry.com/resources/podcast-media-download-statistics/getting-started/
4912 (?:dts|www)\.podtrac\.com/(?:pts/)?redirect\.[0-9a-z]{3,4}| # http://analytics.podtrac.com/how-to-measure
4915 cn\.co| # https://podcorn.com/analytics-prefix/
4916 st\.fm # https://podsights.com/docs/
4921 _HEX_TABLE
= '0123456789abcdef'
4924 def random_uuidv4():
4925 return re
.sub(r
'[xy]', lambda x
: _HEX_TABLE
[random
.randint(0, 15)], 'xxxxxxxx-xxxx-4xxx-yxxx-xxxxxxxxxxxx')
4928 def make_dir(path
, to_screen
=None):
4930 dn
= os
.path
.dirname(path
)
4931 if dn
and not os
.path
.exists(dn
):
4934 except (OSError, IOError) as err
:
4935 if callable(to_screen
) is not None:
4936 to_screen('unable to create directory ' + error_to_compat_str(err
))
4940 def get_executable_path():
4941 from zipimport
import zipimporter
4942 if hasattr(sys
, 'frozen'): # Running from PyInstaller
4943 path
= os
.path
.dirname(sys
.executable
)
4944 elif isinstance(__loader__
, zipimporter
): # Running from ZIP
4945 path
= os
.path
.join(os
.path
.dirname(__file__
), '../..')
4947 path
= os
.path
.join(os
.path
.dirname(__file__
), '..')
4948 return os
.path
.abspath(path
)
4951 def load_plugins(name
, suffix
, namespace
):
4954 plugins_spec
= importlib
.util
.spec_from_file_location(
4955 name
, os
.path
.join(get_executable_path(), 'ytdlp_plugins', name
, '__init__.py'))
4956 plugins
= importlib
.util
.module_from_spec(plugins_spec
)
4957 sys
.modules
[plugins_spec
.name
] = plugins
4958 plugins_spec
.loader
.exec_module(plugins
)
4959 for name
in dir(plugins
):
4960 if name
in namespace
:
4962 if not name
.endswith(suffix
):
4964 klass
= getattr(plugins
, name
)
4965 classes
[name
] = namespace
[name
] = klass
4966 except FileNotFoundError
:
4972 obj
, *path_list
, default
=None, expected_type
=None, get_all
=True,
4973 casesense
=True, is_user_input
=False, traverse_string
=False):
4974 ''' Traverse nested list/dict/tuple
4975 @param path_list A list of paths which are checked one by one.
4976 Each path is a list of keys where each key is a string,
4977 a function, a tuple of strings/None or "...".
4978 When a fuction is given, it takes the key and value as arguments
4979 and returns whether the key matches or not. When a tuple is given,
4980 all the keys given in the tuple are traversed, and
4981 "..." traverses all the keys in the object
4982 "None" returns the object without traversal
4983 @param default Default value to return
4984 @param expected_type Only accept final value of this type (Can also be any callable)
4985 @param get_all Return all the values obtained from a path or only the first one
4986 @param casesense Whether to consider dictionary keys as case sensitive
4987 @param is_user_input Whether the keys are generated from user input. If True,
4988 strings are converted to int/slice if necessary
4989 @param traverse_string Whether to traverse inside strings. If True, any
4990 non-compatible object will also be converted into a string
4994 _lower
= lambda k
: (k
.lower() if isinstance(k
, str) else k
)
4995 path_list
= (map(_lower
, variadic(path
)) for path
in path_list
)
4997 def _traverse_obj(obj
, path
, _current_depth
=0):
4999 path
= tuple(variadic(path
))
5000 for i
, key
in enumerate(path
):
5001 if None in (key
, obj
):
5003 if isinstance(key
, (list, tuple)):
5004 obj
= [_traverse_obj(obj
, sub_key
, _current_depth
) for sub_key
in key
]
5007 obj
= (obj
.values() if isinstance(obj
, dict)
5008 else obj
if isinstance(obj
, (list, tuple, LazyList
))
5009 else str(obj
) if traverse_string
else [])
5011 depth
= max(depth
, _current_depth
)
5012 return [_traverse_obj(inner_obj
, path
[i
+ 1:], _current_depth
) for inner_obj
in obj
]
5014 if isinstance(obj
, (list, tuple, LazyList
)):
5015 obj
= enumerate(obj
)
5016 elif isinstance(obj
, dict):
5019 if not traverse_string
:
5023 depth
= max(depth
, _current_depth
)
5024 return [_traverse_obj(v
, path
[i
+ 1:], _current_depth
) for k
, v
in obj
if try_call(key
, args
=(k
, v
))]
5025 elif isinstance(obj
, dict) and not (is_user_input
and key
== ':'):
5026 obj
= (obj
.get(key
) if casesense
or (key
in obj
)
5027 else next((v
for k
, v
in obj
.items() if _lower(k
) == key
), None))
5030 key
= (int_or_none(key
) if ':' not in key
5031 else slice(*map(int_or_none
, key
.split(':'))))
5032 if key
== slice(None):
5033 return _traverse_obj(obj
, (..., *path
[i
+ 1:]), _current_depth
)
5034 if not isinstance(key
, (int, slice)):
5036 if not isinstance(obj
, (list, tuple, LazyList
)):
5037 if not traverse_string
:
5046 if isinstance(expected_type
, type):
5047 type_test
= lambda val
: val
if isinstance(val
, expected_type
) else None
5048 elif expected_type
is not None:
5049 type_test
= expected_type
5051 type_test
= lambda val
: val
5053 for path
in path_list
:
5055 val
= _traverse_obj(obj
, path
)
5058 for _
in range(depth
- 1):
5059 val
= itertools
.chain
.from_iterable(v
for v
in val
if v
is not None)
5060 val
= [v
for v
in map(type_test
, val
) if v
is not None]
5062 return val
if get_all
else val
[0]
5064 val
= type_test(val
)
5070 def traverse_dict(dictn
, keys
, casesense
=True):
5071 write_string('DeprecationWarning: yt_dlp.utils.traverse_dict is deprecated '
5072 'and may be removed in a future version. Use yt_dlp.utils.traverse_obj instead')
5073 return traverse_obj(dictn
, keys
, casesense
=casesense
, is_user_input
=True, traverse_string
=True)
5076 def get_first(obj
, keys
, **kwargs
):
5077 return traverse_obj(obj
, (..., *variadic(keys
)), **kwargs
, get_all
=False)
5080 def variadic(x
, allowed_types
=(str, bytes, dict)):
5081 return x
if isinstance(x
, collections
.abc
.Iterable
) and not isinstance(x
, allowed_types
) else (x
,)
5084 def decode_base(value
, digits
):
5085 # This will convert given base-x string to scalar (long or int)
5086 table
= {char: index for index, char in enumerate(digits)}
5091 result
+= table
[chr]
5095 def time_seconds(**kwargs
):
5096 t
= datetime
.datetime
.now(datetime
.timezone(datetime
.timedelta(**kwargs
)))
5097 return t
.timestamp()
5100 # create a JSON Web Signature (jws) with HS256 algorithm
5101 # the resulting format is in JWS Compact Serialization
5102 # implemented following JWT https://www.rfc-editor.org/rfc/rfc7519.html
5103 # implemented following JWS https://www.rfc-editor.org/rfc/rfc7515.html
5104 def jwt_encode_hs256(payload_data
, key
, headers
={}):
5110 header_data
.update(headers
)
5111 header_b64
= base64
.b64encode(json
.dumps(header_data
).encode('utf-8'))
5112 payload_b64
= base64
.b64encode(json
.dumps(payload_data
).encode('utf-8'))
5113 h
= hmac
.new(key
.encode('utf-8'), header_b64
+ b
'.' + payload_b64
, hashlib
.sha256
)
5114 signature_b64
= base64
.b64encode(h
.digest())
5115 token
= header_b64
+ b
'.' + payload_b64
+ b
'.' + signature_b64
5119 # can be extended in future to verify the signature and parse header and return the algorithm used if it's not HS256
5120 def jwt_decode_hs256(jwt
):
5121 header_b64
, payload_b64
, signature_b64
= jwt
.split('.')
5122 payload_data
= json
.loads(base64
.urlsafe_b64decode(payload_b64
))
5126 def supports_terminal_sequences(stream
):
5127 if compat_os_name
== 'nt':
5128 from .compat
import WINDOWS_VT_MODE
# Must be imported locally
5129 if not WINDOWS_VT_MODE
or get_windows_version() < (10, 0, 10586):
5131 elif not os
.getenv('TERM'):
5134 return stream
.isatty()
5135 except BaseException
:
5139 _terminal_sequences_re
= re
.compile('\033\\[[^m]+m')
5142 def remove_terminal_sequences(string
):
5143 return _terminal_sequences_re
.sub('', string
)
5146 def number_of_digits(number
):
5147 return len('%d' % number
)
5150 def join_nonempty(*values
, delim
='-', from_dict
=None):
5151 if from_dict
is not None:
5152 values
= map(from_dict
.get
, values
)
5153 return delim
.join(map(str, filter(None, values
)))
5156 def scale_thumbnails_to_max_format_width(formats
, thumbnails
, url_width_re
):
5158 Find the largest format dimensions in terms of video width and, for each thumbnail:
5159 * Modify the URL: Match the width with the provided regex and replace with the former width
5162 This function is useful with video services that scale the provided thumbnails on demand
5164 _keys
= ('width', 'height')
5165 max_dimensions
= max(
5166 [tuple(format
.get(k
) or 0 for k
in _keys
) for format
in formats
],
5168 if not max_dimensions
[0]:
5172 {'url': re.sub(url_width_re, str(max_dimensions[0]), thumbnail['url'])}
,
5173 dict(zip(_keys
, max_dimensions
)), thumbnail
)
5174 for thumbnail
in thumbnails
5178 def parse_http_range(range):
5179 """ Parse value of "Range" or "Content-Range" HTTP header into tuple. """
5181 return None, None, None
5182 crg
= re
.search(r
'bytes[ =](\d+)-(\d+)?(?:/(\d+))?', range)
5184 return None, None, None
5185 return int(crg
.group(1)), int_or_none(crg
.group(2)), int_or_none(crg
.group(3))
5191 __initialized
= False
5193 def __init__(self
, parser
, label
=None):
5194 self
._parser
, self
.label
= parser
, label
5195 self
._loaded
_paths
, self
.configs
= set(), []
5197 def init(self
, args
=None, filename
=None):
5198 assert not self
.__initialized
5201 location
= os
.path
.realpath(filename
)
5202 directory
= os
.path
.dirname(location
)
5203 if location
in self
._loaded
_paths
:
5205 self
._loaded
_paths
.add(location
)
5207 self
.__initialized
= True
5208 self
.own_args
, self
.filename
= args
, filename
5209 for location
in self
._parser
.parse_args(args
)[0].config_locations
or []:
5210 location
= os
.path
.join(directory
, expand_path(location
))
5211 if os
.path
.isdir(location
):
5212 location
= os
.path
.join(location
, 'yt-dlp.conf')
5213 if not os
.path
.exists(location
):
5214 self
._parser
.error(f
'config location {location} does not exist')
5215 self
.append_config(self
.read_file(location
), location
)
5219 label
= join_nonempty(
5220 self
.label
, 'config', f
'"{self.filename}"' if self
.filename
else '',
5222 return join_nonempty(
5223 self
.own_args
is not None and f
'{label[0].upper()}{label[1:]}: {self.hide_login_info(self.own_args)}',
5224 *(f
'\n{c}'.replace('\n', '\n| ')[1:] for c
in self
.configs
),
5228 def read_file(filename
, default
=[]):
5230 optionf
= open(filename
)
5232 return default
# silently skip if file is not present
5234 # FIXME: https://github.com/ytdl-org/youtube-dl/commit/dfe5fa49aed02cf36ba9f743b11b0903554b5e56
5235 contents
= optionf
.read()
5236 res
= compat_shlex_split(contents
, comments
=True)
5242 def hide_login_info(opts
):
5243 PRIVATE_OPTS
= set(['-p', '--password', '-u', '--username', '--video-password', '--ap-password', '--ap-username'])
5244 eqre
= re
.compile('^(?P<key>' + ('|'.join(re
.escape(po
) for po
in PRIVATE_OPTS
)) + ')=.+$')
5249 return m
.group('key') + '=PRIVATE'
5253 opts
= list(map(_scrub_eq
, opts
))
5254 for idx
, opt
in enumerate(opts
):
5255 if opt
in PRIVATE_OPTS
and idx
+ 1 < len(opts
):
5256 opts
[idx
+ 1] = 'PRIVATE'
5259 def append_config(self
, *args
, label
=None):
5260 config
= type(self
)(self
._parser
, label
)
5261 config
._loaded
_paths
= self
._loaded
_paths
5262 if config
.init(*args
):
5263 self
.configs
.append(config
)
5267 for config
in reversed(self
.configs
):
5268 yield from config
.all_args
5269 yield from self
.own_args
or []
5271 def parse_args(self
):
5272 return self
._parser
.parse_args(list(self
.all_args
))
5275 class WebSocketsWrapper():
5276 """Wraps websockets module to use in non-async scopes"""
5278 def __init__(self
, url
, headers
=None, connect
=True):
5279 self
.loop
= asyncio
.events
.new_event_loop()
5280 self
.conn
= compat_websockets
.connect(
5281 url
, extra_headers
=headers
, ping_interval
=None,
5282 close_timeout
=float('inf'), loop
=self
.loop
, ping_timeout
=float('inf'))
5285 atexit
.register(self
.__exit
__, None, None, None)
5287 def __enter__(self
):
5289 self
.pool
= self
.run_with_loop(self
.conn
.__aenter
__(), self
.loop
)
5292 def send(self
, *args
):
5293 self
.run_with_loop(self
.pool
.send(*args
), self
.loop
)
5295 def recv(self
, *args
):
5296 return self
.run_with_loop(self
.pool
.recv(*args
), self
.loop
)
5298 def __exit__(self
, type, value
, traceback
):
5300 return self
.run_with_loop(self
.conn
.__aexit
__(type, value
, traceback
), self
.loop
)
5303 self
._cancel
_all
_tasks
(self
.loop
)
5305 # taken from https://github.com/python/cpython/blob/3.9/Lib/asyncio/runners.py with modifications
5306 # for contributors: If there's any new library using asyncio needs to be run in non-async, move these function out of this class
5308 def run_with_loop(main
, loop
):
5309 if not asyncio
.coroutines
.iscoroutine(main
):
5310 raise ValueError(f
'a coroutine was expected, got {main!r}')
5313 return loop
.run_until_complete(main
)
5315 loop
.run_until_complete(loop
.shutdown_asyncgens())
5316 if hasattr(loop
, 'shutdown_default_executor'):
5317 loop
.run_until_complete(loop
.shutdown_default_executor())
5320 def _cancel_all_tasks(loop
):
5321 to_cancel
= asyncio
.tasks
.all_tasks(loop
)
5326 for task
in to_cancel
:
5329 loop
.run_until_complete(
5330 asyncio
.tasks
.gather(*to_cancel
, loop
=loop
, return_exceptions
=True))
5332 for task
in to_cancel
:
5333 if task
.cancelled():
5335 if task
.exception() is not None:
5336 loop
.call_exception_handler({
5337 'message': 'unhandled exception during asyncio.run() shutdown',
5338 'exception': task
.exception(),
5343 has_websockets
= bool(compat_websockets
)
5346 def merge_headers(*dicts
):
5347 """Merge dicts of http headers case insensitively, prioritizing the latter ones"""
5348 return {k.title(): v for k, v in itertools.chain.from_iterable(map(dict.items, dicts))}
5351 class classproperty
:
5352 def __init__(self
, f
):
5355 def __get__(self
, _
, cls
):