4 from __future__
import unicode_literals
41 import xml
.etree
.ElementTree
46 compat_HTMLParseError
,
53 compat_ctypes_WINFUNCTYPE
,
54 compat_etree_fromstring
,
57 compat_html_entities_html5
,
71 compat_urllib_parse_urlencode
,
72 compat_urllib_parse_urlparse
,
73 compat_urllib_parse_urlunparse
,
74 compat_urllib_parse_quote
,
75 compat_urllib_parse_quote_plus
,
76 compat_urllib_parse_unquote_plus
,
77 compat_urllib_request
,
89 def register_socks_protocols():
90 # "Register" SOCKS protocols
91 # In Python < 2.6.5, urlsplit() suffers from bug https://bugs.python.org/issue7904
92 # URLs with protocols not in urlparse.uses_netloc are not handled correctly
93 for scheme
in ('socks', 'socks4', 'socks4a', 'socks5'):
94 if scheme
not in compat_urlparse
.uses_netloc
:
95 compat_urlparse
.uses_netloc
.append(scheme
)
98 # This is not clearly defined otherwise
99 compiled_regex_type
= type(re
.compile(''))
102 def random_user_agent():
103 _USER_AGENT_TPL
= 'Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/%s Safari/537.36'
144 return _USER_AGENT_TPL
% random
.choice(_CHROME_VERSIONS
)
147 SUPPORTED_ENCODINGS
= [
151 SUPPORTED_ENCODINGS
.append('br')
154 'User-Agent': random_user_agent(),
155 'Accept': 'text/html,application/xhtml+xml,application/xml;q=0.9,*/*;q=0.8',
156 'Accept-Encoding': ', '.join(SUPPORTED_ENCODINGS
),
157 'Accept-Language': 'en-us,en;q=0.5',
158 'Sec-Fetch-Mode': 'navigate',
163 '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',
167 NO_DEFAULT
= object()
169 ENGLISH_MONTH_NAMES
= [
170 'January', 'February', 'March', 'April', 'May', 'June',
171 'July', 'August', 'September', 'October', 'November', 'December']
174 'en': ENGLISH_MONTH_NAMES
,
176 'janvier', 'février', 'mars', 'avril', 'mai', 'juin',
177 'juillet', 'août', 'septembre', 'octobre', 'novembre', 'décembre'],
181 'mp4', 'm4a', 'm4p', 'm4b', 'm4r', 'm4v', 'aac',
182 'flv', 'f4v', 'f4a', 'f4b',
183 'webm', 'ogg', 'ogv', 'oga', 'ogx', 'spx', 'opus',
184 'mkv', 'mka', 'mk3d',
193 'f4f', 'f4m', 'm3u8', 'smil')
195 # needed for sanitizing filenames in restricted mode
196 ACCENT_CHARS
= dict(zip('ÂÃÄÀÁÅÆÇÈÉÊËÌÍÎÏÐÑÒÓÔÕÖŐØŒÙÚÛÜŰÝÞßàáâãäåæçèéêëìíîïðñòóôõöőøœùúûüűýþÿ',
197 itertools
.chain('AAAAAA', ['AE'], 'CEEEEIIIIDNOOOOOOO', ['OE'], 'UUUUUY', ['TH', 'ss'],
198 'aaaaaa', ['ae'], 'ceeeeiiiionooooooo', ['oe'], 'uuuuuy', ['th'], 'y')))
228 '%Y-%m-%d %H:%M:%S.%f',
229 '%Y-%m-%d %H:%M:%S:%f',
232 '%Y-%m-%dT%H:%M:%SZ',
233 '%Y-%m-%dT%H:%M:%S.%fZ',
234 '%Y-%m-%dT%H:%M:%S.%f0Z',
236 '%Y-%m-%dT%H:%M:%S.%f',
239 '%b %d %Y at %H:%M:%S',
241 '%B %d %Y at %H:%M:%S',
245 DATE_FORMATS_DAY_FIRST
= list(DATE_FORMATS
)
246 DATE_FORMATS_DAY_FIRST
.extend([
255 DATE_FORMATS_MONTH_FIRST
= list(DATE_FORMATS
)
256 DATE_FORMATS_MONTH_FIRST
.extend([
264 PACKED_CODES_RE
= r
"}\('(.+)',(\d+),(\d+),'([^']+)'\.split\('\|'\)"
265 JSON_LD_RE
= r
'(?is)<script[^>]+type=(["\']?
)application
/ld\
+json\
1[^
>]*>(?P
<json_ld
>.+?
)</script
>'
268 def preferredencoding():
269 """Get preferred encoding.
271 Returns the best encoding scheme for the system, based on
272 locale.getpreferredencoding() and some further tweaks.
275 pref = locale.getpreferredencoding()
283 def write_json_file(obj, fn):
284 """ Encode obj as JSON and write it to fn, atomically if possible """
286 fn = encodeFilename(fn)
287 if sys.version_info < (3, 0) and sys.platform != 'win32
':
288 encoding = get_filesystem_encoding()
289 # os.path.basename returns a bytes object, but NamedTemporaryFile
290 # will fail if the filename contains non ascii characters unless we
291 # use a unicode object
292 path_basename = lambda f: os.path.basename(fn).decode(encoding)
293 # the same for os.path.dirname
294 path_dirname = lambda f: os.path.dirname(fn).decode(encoding)
296 path_basename = os.path.basename
297 path_dirname = os.path.dirname
301 'prefix
': path_basename(fn) + '.',
302 'dir': path_dirname(fn),
306 # In Python 2.x, json.dump expects a bytestream.
307 # In Python 3.x, it writes to a character stream
308 if sys.version_info < (3, 0):
316 tf = tempfile.NamedTemporaryFile(**compat_kwargs(args))
320 json.dump(obj, tf, ensure_ascii=False)
321 if sys.platform == 'win32
':
322 # Need to remove existing file on Windows, else os.rename raises
323 # WindowsError or FileExistsError.
331 os.chmod(tf.name, 0o666 & ~mask)
334 os.rename(tf.name, fn)
343 if sys.version_info >= (2, 7):
344 def find_xpath_attr(node, xpath, key, val=None):
345 """ Find the xpath xpath[@key=val] """
346 assert re.match(r'^
[a
-zA
-Z_
-]+$
', key)
347 expr = xpath + ('[@%s]' % key if val is None else "[@%s='%s']" % (key, val))
348 return node.find(expr)
350 def find_xpath_attr(node, xpath, key, val=None):
351 for f in node.findall(compat_xpath(xpath)):
352 if key not in f.attrib:
354 if val is None or f.attrib.get(key) == val:
358 # On python2.6 the xml.etree.ElementTree.Element methods don't support
359 # the namespace parameter
362 def xpath_with_ns(path
, ns_map
):
363 components
= [c
.split(':') for c
in path
.split('/')]
367 replaced
.append(c
[0])
370 replaced
.append('{%s}%s' % (ns_map
[ns
], tag
))
371 return '/'.join(replaced
)
374 def xpath_element(node
, xpath
, name
=None, fatal
=False, default
=NO_DEFAULT
):
375 def _find_xpath(xpath
):
376 return node
.find(compat_xpath(xpath
))
378 if isinstance(xpath
, (str, compat_str
)):
379 n
= _find_xpath(xpath
)
387 if default
is not NO_DEFAULT
:
390 name
= xpath
if name
is None else name
391 raise ExtractorError('Could not find XML element %s' % name
)
397 def xpath_text(node
, xpath
, name
=None, fatal
=False, default
=NO_DEFAULT
):
398 n
= xpath_element(node
, xpath
, name
, fatal
=fatal
, default
=default
)
399 if n
is None or n
== default
:
402 if default
is not NO_DEFAULT
:
405 name
= xpath
if name
is None else name
406 raise ExtractorError('Could not find XML element\'s text %s' % name
)
412 def xpath_attr(node
, xpath
, key
, name
=None, fatal
=False, default
=NO_DEFAULT
):
413 n
= find_xpath_attr(node
, xpath
, key
)
415 if default
is not NO_DEFAULT
:
418 name
= '%s[@%s]' % (xpath
, key
) if name
is None else name
419 raise ExtractorError('Could not find XML attribute %s' % name
)
425 def get_element_by_id(id, html
):
426 """Return the content of the tag with the specified ID in the passed HTML document"""
427 return get_element_by_attribute('id', id, html
)
430 def get_element_html_by_id(id, html
):
431 """Return the html of the tag with the specified ID in the passed HTML document"""
432 return get_element_html_by_attribute('id', id, html
)
435 def get_element_by_class(class_name
, html
):
436 """Return the content of the first tag with the specified class in the passed HTML document"""
437 retval
= get_elements_by_class(class_name
, html
)
438 return retval
[0] if retval
else None
441 def get_element_html_by_class(class_name
, html
):
442 """Return the html of the first tag with the specified class in the passed HTML document"""
443 retval
= get_elements_html_by_class(class_name
, html
)
444 return retval
[0] if retval
else None
447 def get_element_by_attribute(attribute
, value
, html
, escape_value
=True):
448 retval
= get_elements_by_attribute(attribute
, value
, html
, escape_value
)
449 return retval
[0] if retval
else None
452 def get_element_html_by_attribute(attribute
, value
, html
, escape_value
=True):
453 retval
= get_elements_html_by_attribute(attribute
, value
, html
, escape_value
)
454 return retval
[0] if retval
else None
457 def get_elements_by_class(class_name
, html
):
458 """Return the content of all tags with the specified class in the passed HTML document as a list"""
459 return get_elements_by_attribute(
460 'class', r
'[^\'"]*\b%s\b[^\'"]*' % re.escape(class_name),
461 html, escape_value=False)
464 def get_elements_html_by_class(class_name, html):
465 """Return the html of all tags with the specified class in the passed HTML document as a list"""
466 return get_elements_html_by_attribute(
467 'class', r'[^
\'"]*\b%s\b[^\'"]*' % re.escape(class_name),
468 html, escape_value=False)
471 def get_elements_by_attribute(*args, **kwargs):
472 """Return the content of the tag with the specified attribute in the passed HTML document"""
473 return [content for content, _ in get_elements_text_and_html_by_attribute(*args, **kwargs)]
476 def get_elements_html_by_attribute(*args, **kwargs):
477 """Return the html of the tag with the specified attribute in the passed HTML document"""
478 return [whole for _, whole in get_elements_text_and_html_by_attribute(*args, **kwargs)]
481 def get_elements_text_and_html_by_attribute(attribute, value, html, escape_value=True):
483 Return the text (content) and the html (whole) of the tag with the specified
484 attribute in the passed HTML document
487 value_quote_optional = '' if re.match(r'''[\s"'`
=<>]''', value) else '?'
489 value = re.escape(value) if escape_value else value
491 partial_element_re = r'''(?x
)
492 <(?P
<tag
>[a
-zA
-Z0
-9:._-]+)
493 (?
:\
s(?
:[^
>"']|"[^
"]*"|
'[^']*')*)?
494 \s%(attribute)s\s*=\s*(?P<_q>['"]%(vqo)s)(?-x:%(value)s)(?P=_q)
495 ''' % {'attribute': re.escape(attribute), 'value': value, 'vqo': value_quote_optional}
497 for m in re.finditer(partial_element_re, html):
498 content, whole = get_element_text_and_html_by_tag(m.group('tag'), html[m.start():])
501 unescapeHTML(re.sub(r'^(?P<q>["\'])(?P
<content
>.*)(?P
=q
)$
', r'\g
<content
>', content, flags=re.DOTALL)),
506 class HTMLBreakOnClosingTagParser(compat_HTMLParser):
508 HTML parser which raises HTMLBreakOnClosingTagException upon reaching the
509 closing tag for the first opening tag it has encountered, and can be used
513 class HTMLBreakOnClosingTagException(Exception):
517 self.tagstack = collections.deque()
518 compat_HTMLParser.__init__(self)
523 def __exit__(self, *_):
527 # handle_endtag does not return upon raising HTMLBreakOnClosingTagException,
528 # so data remains buffered; we no longer have any interest in it, thus
529 # override this method to discard it
532 def handle_starttag(self, tag, _):
533 self.tagstack.append(tag)
535 def handle_endtag(self, tag):
536 if not self.tagstack:
537 raise compat_HTMLParseError('no tags
in the stack
')
539 inner_tag = self.tagstack.pop()
543 raise compat_HTMLParseError(f'matching opening tag
for closing {tag} tag
not found
')
544 if not self.tagstack:
545 raise self.HTMLBreakOnClosingTagException()
548 def get_element_text_and_html_by_tag(tag, html):
550 For the first element with the specified tag in the passed HTML document
551 return its' content (text
) and the whole
element (html
)
553 def find_or_raise(haystack, needle, exc):
555 return haystack.index(needle)
558 closing_tag = f'</{tag}>'
559 whole_start = find_or_raise(
560 html, f'<{tag}', compat_HTMLParseError(f'opening {tag} tag not found'))
561 content_start = find_or_raise(
562 html[whole_start:], '>', compat_HTMLParseError(f'malformed opening {tag} tag'))
563 content_start += whole_start + 1
564 with HTMLBreakOnClosingTagParser() as parser:
565 parser.feed(html[whole_start:content_start])
566 if not parser.tagstack or parser.tagstack[0] != tag:
567 raise compat_HTMLParseError(f'parser did not match opening {tag} tag')
568 offset = content_start
569 while offset < len(html):
570 next_closing_tag_start = find_or_raise(
571 html[offset:], closing_tag,
572 compat_HTMLParseError(f'closing {tag} tag not found'))
573 next_closing_tag_end = next_closing_tag_start + len(closing_tag)
575 parser.feed(html[offset:offset + next_closing_tag_end])
576 offset += next_closing_tag_end
577 except HTMLBreakOnClosingTagParser.HTMLBreakOnClosingTagException:
578 return html[content_start:offset + next_closing_tag_start], \
579 html[whole_start:offset + next_closing_tag_end]
580 raise compat_HTMLParseError('unexpected end of html')
583 class HTMLAttributeParser(compat_HTMLParser):
584 """Trivial HTML parser to gather the attributes
for a single element
"""
588 compat_HTMLParser.__init__(self)
590 def handle_starttag(self, tag, attrs):
591 self.attrs = dict(attrs)
594 class HTMLListAttrsParser(compat_HTMLParser):
595 """HTML parser to gather the attributes
for the elements of a
list"""
598 compat_HTMLParser.__init__(self)
602 def handle_starttag(self, tag, attrs):
603 if tag == 'li' and self._level == 0:
604 self.items.append(dict(attrs))
607 def handle_endtag(self, tag):
611 def extract_attributes(html_element):
612 """Given a string
for an HTML element such
as
614 a
="foo" B
="bar" c
="&98;az" d
=boz
615 empty
= noval entity
="&"
618 Decode
and return a dictionary of attributes
.
620 'a': 'foo', 'b': 'bar', c
: 'baz', d
: 'boz',
621 'empty': '', 'noval': None, 'entity': '&',
622 'sq': '"', 'dq': '\''
624 NB HTMLParser
is stricter
in Python
2.6 & 3.2 than
in later versions
,
625 but the cases
in the unit test will work
for all of
2.6, 2.7, 3.2-3.5.
627 parser = HTMLAttributeParser()
629 parser.feed(html_element)
631 # Older Python may throw HTMLParseError in case of malformed HTML
632 except compat_HTMLParseError:
637 def parse_list(webpage):
638 """Given a string
for an series of HTML
<li
> elements
,
639 return a dictionary of their attributes
"""
640 parser = HTMLListAttrsParser()
646 def clean_html(html):
647 """Clean an HTML snippet into a readable string
"""
649 if html is None: # Convenience for sanitizing descriptions etc.
652 html = re.sub(r'\s+', ' ', html)
653 html = re.sub(r'(?u)\s?<\s?br\s?/?\s?>\s?', '\n', html)
654 html = re.sub(r'(?u)<\s?/\s?p\s?>\s?<\s?p[^>]*>', '\n', html)
656 html = re.sub('<.*?>', '', html)
657 # Replace html entities
658 html = unescapeHTML(html)
662 def sanitize_open(filename, open_mode):
663 """Try to
open the given filename
, and slightly tweak it
if this fails
.
665 Attempts to
open the given filename
. If this fails
, it tries to change
666 the filename slightly
, step by step
, until it
's either able to open it
667 or it fails and raises a final exception, like the standard open()
670 It returns the tuple (stream, definitive_file_name).
674 if sys.platform == 'win32
':
676 msvcrt.setmode(sys.stdout.fileno(), os.O_BINARY)
677 return (sys.stdout.buffer if hasattr(sys.stdout, 'buffer') else sys.stdout, filename)
678 stream = locked_file(filename, open_mode, block=False).open()
679 return (stream, filename)
680 except (IOError, OSError) as err:
681 if err.errno in (errno.EACCES,):
684 # In case of error, try to remove win32 forbidden chars
685 alt_filename = sanitize_path(filename)
686 if alt_filename == filename:
689 # An exception here should be caught in the caller
690 stream = locked_file(filename, open_mode, block=False).open()
691 return (stream, alt_filename)
694 def timeconvert(timestr):
695 """Convert RFC 2822 defined time string into system timestamp"""
697 timetuple = email.utils.parsedate_tz(timestr)
698 if timetuple is not None:
699 timestamp = email.utils.mktime_tz(timetuple)
703 def sanitize_filename(s, restricted=False, is_id=False):
704 """Sanitizes a string so it could be used as part of a filename.
705 If restricted is set, use a stricter subset of allowed characters.
706 Set is_id if this is not an arbitrary string, but an ID that should be kept
709 def replace_insane(char):
710 if restricted and char in ACCENT_CHARS:
711 return ACCENT_CHARS[char]
712 elif not restricted and char == '\n':
714 elif char == '?
' or ord(char) < 32 or ord(char) == 127:
717 return '' if restricted else '\''
719 return '_-' if restricted else ' -'
720 elif char in '\\/|*<>':
722 if restricted and (char in '!&\'()[]{}$;`^,#' or char.isspace()):
724 if restricted and ord(char) > 127:
731 s = re.sub(r'[0-9]+(?::[0-9]+)+', lambda m: m.group(0).replace(':', '_'), s)
732 result = ''.join(map(replace_insane, s))
734 while '__' in result:
735 result = result.replace('__', '_')
736 result = result.strip('_')
737 # Common case of "Foreign band name
- English song title
"
738 if restricted and result.startswith('-_'):
740 if result.startswith('-'):
741 result = '_' + result[len('-'):]
742 result = result.lstrip('.')
748 def sanitize_path(s, force=False):
749 """Sanitizes and normalizes path on Windows"""
750 if sys.platform == 'win32':
752 drive_or_unc, _ = os.path.splitdrive(s)
753 if sys.version_info < (2, 7) and not drive_or_unc:
754 drive_or_unc, _ = os.path.splitunc(s)
760 norm_path = os.path.normpath(remove_start(s, drive_or_unc)).split(os.path.sep)
764 path_part if path_part in ['.', '..'] else re.sub(r'(?:[/<>:"\|
\\?\
*]|
[\s
.]$
)', '#', path_part)
765 for path_part
in norm_path
]
767 sanitized_path
.insert(0, drive_or_unc
+ os
.path
.sep
)
768 elif force
and s
[0] == os
.path
.sep
:
769 sanitized_path
.insert(0, os
.path
.sep
)
770 return os
.path
.join(*sanitized_path
)
773 def sanitize_url(url
):
774 # Prepend protocol-less URLs with `http:` scheme in order to mitigate
775 # the number of unwanted failures due to missing protocol
776 if url
.startswith('//'):
777 return 'http:%s' % url
778 # Fix some common typos seen so far
780 # https://github.com/ytdl-org/youtube-dl/issues/15649
781 (r
'^httpss://', r
'https://'),
782 # https://bx1.be/lives/direct-tv/
783 (r
'^rmtp([es]?)://', r
'rtmp\1://'),
785 for mistake
, fixup
in COMMON_TYPOS
:
786 if re
.match(mistake
, url
):
787 return re
.sub(mistake
, fixup
, url
)
791 def extract_basic_auth(url
):
792 parts
= compat_urlparse
.urlsplit(url
)
793 if parts
.username
is None:
795 url
= compat_urlparse
.urlunsplit(parts
._replace
(netloc
=(
796 parts
.hostname
if parts
.port
is None
797 else '%s:%d' % (parts
.hostname
, parts
.port
))))
798 auth_payload
= base64
.b64encode(
799 ('%s:%s' % (parts
.username
, parts
.password
or '')).encode('utf-8'))
800 return url
, 'Basic ' + auth_payload
.decode('utf-8')
803 def sanitized_Request(url
, *args
, **kwargs
):
804 url
, auth_header
= extract_basic_auth(escape_url(sanitize_url(url
)))
805 if auth_header
is not None:
806 headers
= args
[1] if len(args
) >= 2 else kwargs
.setdefault('headers', {})
807 headers
['Authorization'] = auth_header
808 return compat_urllib_request
.Request(url
, *args
, **kwargs
)
812 """Expand shell variables and ~"""
813 return os
.path
.expandvars(compat_expanduser(s
))
816 def orderedSet(iterable
):
817 """ Remove all duplicates from the input iterable """
825 def _htmlentity_transform(entity_with_semicolon
):
826 """Transforms an HTML entity to a character."""
827 entity
= entity_with_semicolon
[:-1]
829 # Known non-numeric HTML entity
830 if entity
in compat_html_entities
.name2codepoint
:
831 return compat_chr(compat_html_entities
.name2codepoint
[entity
])
833 # TODO: HTML5 allows entities without a semicolon. For example,
834 # 'Éric' should be decoded as 'Éric'.
835 if entity_with_semicolon
in compat_html_entities_html5
:
836 return compat_html_entities_html5
[entity_with_semicolon
]
838 mobj
= re
.match(r
'#(x[0-9a-fA-F]+|[0-9]+)', entity
)
840 numstr
= mobj
.group(1)
841 if numstr
.startswith('x'):
843 numstr
= '0%s' % numstr
846 # See https://github.com/ytdl-org/youtube-dl/issues/7518
848 return compat_chr(int(numstr
, base
))
852 # Unknown entity in name, return its literal representation
853 return '&%s;' % entity
859 assert type(s
) == compat_str
862 r
'&([^&;]+;)', lambda m
: _htmlentity_transform(m
.group(1)), s
)
865 def escapeHTML(text
):
868 .replace('&', '&')
869 .replace('<', '<')
870 .replace('>', '>')
871 .replace('"', '"')
872 .replace("'", ''')
876 def process_communicate_or_kill(p
, *args
, **kwargs
):
878 return p
.communicate(*args
, **kwargs
)
879 except BaseException
: # Including KeyboardInterrupt
885 class Popen(subprocess
.Popen
):
886 if sys
.platform
== 'win32':
887 _startupinfo
= subprocess
.STARTUPINFO()
888 _startupinfo
.dwFlags |
= subprocess
.STARTF_USESHOWWINDOW
892 def __init__(self
, *args
, **kwargs
):
893 super(Popen
, self
).__init
__(*args
, **kwargs
, startupinfo
=self
._startupinfo
)
895 def communicate_or_kill(self
, *args
, **kwargs
):
896 return process_communicate_or_kill(self
, *args
, **kwargs
)
899 def get_subprocess_encoding():
900 if sys
.platform
== 'win32' and sys
.getwindowsversion()[0] >= 5:
901 # For subprocess calls, encode with locale encoding
902 # Refer to http://stackoverflow.com/a/9951851/35070
903 encoding
= preferredencoding()
905 encoding
= sys
.getfilesystemencoding()
911 def encodeFilename(s
, for_subprocess
=False):
913 @param s The name of the file
916 assert type(s
) == compat_str
918 # Python 3 has a Unicode API
919 if sys
.version_info
>= (3, 0):
922 # Pass '' directly to use Unicode APIs on Windows 2000 and up
923 # (Detecting Windows NT 4 is tricky because 'major >= 4' would
924 # match Windows 9x series as well. Besides, NT 4 is obsolete.)
925 if not for_subprocess
and sys
.platform
== 'win32' and sys
.getwindowsversion()[0] >= 5:
928 # Jython assumes filenames are Unicode strings though reported as Python 2.x compatible
929 if sys
.platform
.startswith('java'):
932 return s
.encode(get_subprocess_encoding(), 'ignore')
935 def decodeFilename(b
, for_subprocess
=False):
937 if sys
.version_info
>= (3, 0):
940 if not isinstance(b
, bytes):
943 return b
.decode(get_subprocess_encoding(), 'ignore')
946 def encodeArgument(s
):
947 if not isinstance(s
, compat_str
):
948 # Legacy code that uses byte strings
949 # Uncomment the following line after fixing all post processors
950 # assert False, 'Internal error: %r should be of type %r, is %r' % (s, compat_str, type(s))
951 s
= s
.decode('ascii')
952 return encodeFilename(s
, True)
955 def decodeArgument(b
):
956 return decodeFilename(b
, True)
959 def decodeOption(optval
):
962 if isinstance(optval
, bytes):
963 optval
= optval
.decode(preferredencoding())
965 assert isinstance(optval
, compat_str
)
969 _timetuple
= collections
.namedtuple('Time', ('hours', 'minutes', 'seconds', 'milliseconds'))
972 def timetuple_from_msec(msec
):
973 secs
, msec
= divmod(msec
, 1000)
974 mins
, secs
= divmod(secs
, 60)
975 hrs
, mins
= divmod(mins
, 60)
976 return _timetuple(hrs
, mins
, secs
, msec
)
979 def formatSeconds(secs
, delim
=':', msec
=False):
980 time
= timetuple_from_msec(secs
* 1000)
982 ret
= '%d%s%02d%s%02d' % (time
.hours
, delim
, time
.minutes
, delim
, time
.seconds
)
984 ret
= '%d%s%02d' % (time
.minutes
, delim
, time
.seconds
)
986 ret
= '%d' % time
.seconds
987 return '%s.%03d' % (ret
, time
.milliseconds
) if msec
else ret
990 def _ssl_load_windows_store_certs(ssl_context
, storename
):
991 # Code adapted from _load_windows_store_certs in https://github.com/python/cpython/blob/main/Lib/ssl.py
993 certs
= [cert
for cert
, encoding
, trust
in ssl
.enum_certificates(storename
)
994 if encoding
== 'x509_asn' and (
995 trust
is True or ssl
.Purpose
.SERVER_AUTH
.oid
in trust
)]
996 except PermissionError
:
1000 ssl_context
.load_verify_locations(cadata
=cert
)
1001 except ssl
.SSLError
:
1005 def make_HTTPS_handler(params
, **kwargs
):
1006 opts_check_certificate
= not params
.get('nocheckcertificate')
1007 context
= ssl
.SSLContext(ssl
.PROTOCOL_TLS_CLIENT
)
1008 context
.check_hostname
= opts_check_certificate
1009 if params
.get('legacyserverconnect'):
1010 context
.options |
= 4 # SSL_OP_LEGACY_SERVER_CONNECT
1011 context
.verify_mode
= ssl
.CERT_REQUIRED
if opts_check_certificate
else ssl
.CERT_NONE
1012 if opts_check_certificate
:
1014 context
.load_default_certs()
1015 # Work around the issue in load_default_certs when there are bad certificates. See:
1016 # https://github.com/yt-dlp/yt-dlp/issues/1060,
1017 # https://bugs.python.org/issue35665, https://bugs.python.org/issue45312
1018 except ssl
.SSLError
:
1019 # enum_certificates is not present in mingw python. See https://github.com/yt-dlp/yt-dlp/issues/1151
1020 if sys
.platform
== 'win32' and hasattr(ssl
, 'enum_certificates'):
1021 # Create a new context to discard any certificates that were already loaded
1022 context
= ssl
.SSLContext(ssl
.PROTOCOL_TLS_CLIENT
)
1023 context
.check_hostname
, context
.verify_mode
= True, ssl
.CERT_REQUIRED
1024 for storename
in ('CA', 'ROOT'):
1025 _ssl_load_windows_store_certs(context
, storename
)
1026 context
.set_default_verify_paths()
1027 return YoutubeDLHTTPSHandler(params
, context
=context
, **kwargs
)
1030 def bug_reports_message(before
=';'):
1031 msg
= ('please report this issue on https://github.com/yt-dlp/yt-dlp , '
1032 'filling out the "Broken site" issue template properly. '
1033 'Confirm you are on the latest version using -U')
1035 before
= before
.rstrip()
1036 if not before
or before
.endswith(('.', '!', '?')):
1037 msg
= msg
[0].title() + msg
[1:]
1039 return (before
+ ' ' if before
else '') + msg
1042 class YoutubeDLError(Exception):
1043 """Base exception for YoutubeDL errors."""
1046 def __init__(self
, msg
=None):
1049 elif self
.msg
is None:
1050 self
.msg
= type(self
).__name
__
1051 super().__init
__(self
.msg
)
1054 network_exceptions
= [compat_urllib_error
.URLError
, compat_http_client
.HTTPException
, socket
.error
]
1055 if hasattr(ssl
, 'CertificateError'):
1056 network_exceptions
.append(ssl
.CertificateError
)
1057 network_exceptions
= tuple(network_exceptions
)
1060 class ExtractorError(YoutubeDLError
):
1061 """Error during info extraction."""
1063 def __init__(self
, msg
, tb
=None, expected
=False, cause
=None, video_id
=None, ie
=None):
1064 """ tb, if given, is the original traceback (so that it can be printed out).
1065 If expected is set, this is a normal error message and most likely not a bug in yt-dlp.
1067 if sys
.exc_info()[0] in network_exceptions
:
1070 self
.orig_msg
= str(msg
)
1072 self
.expected
= expected
1074 self
.video_id
= video_id
1076 self
.exc_info
= sys
.exc_info() # preserve original exception
1078 super(ExtractorError
, self
).__init
__(''.join((
1079 format_field(ie
, template
='[%s] '),
1080 format_field(video_id
, template
='%s: '),
1082 format_field(cause
, template
=' (caused by %r)'),
1083 '' if expected
else bug_reports_message())))
1085 def format_traceback(self
):
1086 return join_nonempty(
1087 self
.traceback
and ''.join(traceback
.format_tb(self
.traceback
)),
1088 self
.cause
and ''.join(traceback
.format_exception(self
.cause
)[1:]),
1092 class UnsupportedError(ExtractorError
):
1093 def __init__(self
, url
):
1094 super(UnsupportedError
, self
).__init
__(
1095 'Unsupported URL: %s' % url
, expected
=True)
1099 class RegexNotFoundError(ExtractorError
):
1100 """Error when a regex didn't match"""
1104 class GeoRestrictedError(ExtractorError
):
1105 """Geographic restriction Error exception.
1107 This exception may be thrown when a video is not available from your
1108 geographic location due to geographic restrictions imposed by a website.
1111 def __init__(self
, msg
, countries
=None, **kwargs
):
1112 kwargs
['expected'] = True
1113 super(GeoRestrictedError
, self
).__init
__(msg
, **kwargs
)
1114 self
.countries
= countries
1117 class DownloadError(YoutubeDLError
):
1118 """Download Error exception.
1120 This exception may be thrown by FileDownloader objects if they are not
1121 configured to continue on errors. They will contain the appropriate
1125 def __init__(self
, msg
, exc_info
=None):
1126 """ exc_info, if given, is the original exception that caused the trouble (as returned by sys.exc_info()). """
1127 super(DownloadError
, self
).__init
__(msg
)
1128 self
.exc_info
= exc_info
1131 class EntryNotInPlaylist(YoutubeDLError
):
1132 """Entry not in playlist exception.
1134 This exception will be thrown by YoutubeDL when a requested entry
1135 is not found in the playlist info_dict
1137 msg
= 'Entry not found in info'
1140 class SameFileError(YoutubeDLError
):
1141 """Same File exception.
1143 This exception will be thrown by FileDownloader objects if they detect
1144 multiple files would have to be downloaded to the same file on disk.
1146 msg
= 'Fixed output name but more than one file to download'
1148 def __init__(self
, filename
=None):
1149 if filename
is not None:
1150 self
.msg
+= f
': {filename}'
1151 super().__init
__(self
.msg
)
1154 class PostProcessingError(YoutubeDLError
):
1155 """Post Processing exception.
1157 This exception may be raised by PostProcessor's .run() method to
1158 indicate an error in the postprocessing task.
1162 class DownloadCancelled(YoutubeDLError
):
1163 """ Exception raised when the download queue should be interrupted """
1164 msg
= 'The download was cancelled'
1167 class ExistingVideoReached(DownloadCancelled
):
1168 """ --break-on-existing triggered """
1169 msg
= 'Encountered a video that is already in the archive, stopping due to --break-on-existing'
1172 class RejectedVideoReached(DownloadCancelled
):
1173 """ --break-on-reject triggered """
1174 msg
= 'Encountered a video that did not match filter, stopping due to --break-on-reject'
1177 class MaxDownloadsReached(DownloadCancelled
):
1178 """ --max-downloads limit has been reached. """
1179 msg
= 'Maximum number of downloads reached, stopping due to --max-downloads'
1182 class ReExtractInfo(YoutubeDLError
):
1183 """ Video info needs to be re-extracted. """
1185 def __init__(self
, msg
, expected
=False):
1186 super().__init
__(msg
)
1187 self
.expected
= expected
1190 class ThrottledDownload(ReExtractInfo
):
1191 """ Download speed below --throttled-rate. """
1192 msg
= 'The download speed is below throttle limit'
1195 super().__init
__(self
.msg
, expected
=False)
1198 class UnavailableVideoError(YoutubeDLError
):
1199 """Unavailable Format exception.
1201 This exception will be thrown when a video is requested
1202 in a format that is not available for that video.
1204 msg
= 'Unable to download video'
1206 def __init__(self
, err
=None):
1208 self
.msg
+= f
': {err}'
1209 super().__init
__(self
.msg
)
1212 class ContentTooShortError(YoutubeDLError
):
1213 """Content Too Short exception.
1215 This exception may be raised by FileDownloader objects when a file they
1216 download is too small for what the server announced first, indicating
1217 the connection was probably interrupted.
1220 def __init__(self
, downloaded
, expected
):
1221 super(ContentTooShortError
, self
).__init
__(
1222 'Downloaded {0} bytes, expected {1} bytes'.format(downloaded
, expected
)
1225 self
.downloaded
= downloaded
1226 self
.expected
= expected
1229 class XAttrMetadataError(YoutubeDLError
):
1230 def __init__(self
, code
=None, msg
='Unknown error'):
1231 super(XAttrMetadataError
, self
).__init
__(msg
)
1235 # Parsing code and msg
1236 if (self
.code
in (errno
.ENOSPC
, errno
.EDQUOT
)
1237 or 'No space left' in self
.msg
or 'Disk quota exceeded' in self
.msg
):
1238 self
.reason
= 'NO_SPACE'
1239 elif self
.code
== errno
.E2BIG
or 'Argument list too long' in self
.msg
:
1240 self
.reason
= 'VALUE_TOO_LONG'
1242 self
.reason
= 'NOT_SUPPORTED'
1245 class XAttrUnavailableError(YoutubeDLError
):
1249 def _create_http_connection(ydl_handler
, http_class
, is_https
, *args
, **kwargs
):
1250 # Working around python 2 bug (see http://bugs.python.org/issue17849) by limiting
1251 # expected HTTP responses to meet HTTP/1.0 or later (see also
1252 # https://github.com/ytdl-org/youtube-dl/issues/6727)
1253 if sys
.version_info
< (3, 0):
1254 kwargs
['strict'] = True
1255 hc
= http_class(*args
, **compat_kwargs(kwargs
))
1256 source_address
= ydl_handler
._params
.get('source_address')
1258 if source_address
is not None:
1259 # This is to workaround _create_connection() from socket where it will try all
1260 # address data from getaddrinfo() including IPv6. This filters the result from
1261 # getaddrinfo() based on the source_address value.
1262 # This is based on the cpython socket.create_connection() function.
1263 # https://github.com/python/cpython/blob/master/Lib/socket.py#L691
1264 def _create_connection(address
, timeout
=socket
._GLOBAL
_DEFAULT
_TIMEOUT
, source_address
=None):
1265 host
, port
= address
1267 addrs
= socket
.getaddrinfo(host
, port
, 0, socket
.SOCK_STREAM
)
1268 af
= socket
.AF_INET
if '.' in source_address
[0] else socket
.AF_INET6
1269 ip_addrs
= [addr
for addr
in addrs
if addr
[0] == af
]
1270 if addrs
and not ip_addrs
:
1271 ip_version
= 'v4' if af
== socket
.AF_INET
else 'v6'
1273 "No remote IP%s addresses available for connect, can't use '%s' as source address"
1274 % (ip_version
, source_address
[0]))
1275 for res
in ip_addrs
:
1276 af
, socktype
, proto
, canonname
, sa
= res
1279 sock
= socket
.socket(af
, socktype
, proto
)
1280 if timeout
is not socket
._GLOBAL
_DEFAULT
_TIMEOUT
:
1281 sock
.settimeout(timeout
)
1282 sock
.bind(source_address
)
1284 err
= None # Explicitly break reference cycle
1286 except socket
.error
as _
:
1288 if sock
is not None:
1293 raise socket
.error('getaddrinfo returns an empty list')
1294 if hasattr(hc
, '_create_connection'):
1295 hc
._create
_connection
= _create_connection
1296 sa
= (source_address
, 0)
1297 if hasattr(hc
, 'source_address'): # Python 2.7+
1298 hc
.source_address
= sa
1300 def _hc_connect(self
, *args
, **kwargs
):
1301 sock
= _create_connection(
1302 (self
.host
, self
.port
), self
.timeout
, sa
)
1304 self
.sock
= ssl
.wrap_socket(
1305 sock
, self
.key_file
, self
.cert_file
,
1306 ssl_version
=ssl
.PROTOCOL_TLSv1
)
1309 hc
.connect
= functools
.partial(_hc_connect
, hc
)
1314 def handle_youtubedl_headers(headers
):
1315 filtered_headers
= headers
1317 if 'Youtubedl-no-compression' in filtered_headers
:
1318 filtered_headers
= dict((k
, v
) for k
, v
in filtered_headers
.items() if k
.lower() != 'accept-encoding')
1319 del filtered_headers
['Youtubedl-no-compression']
1321 return filtered_headers
1324 class YoutubeDLHandler(compat_urllib_request
.HTTPHandler
):
1325 """Handler for HTTP requests and responses.
1327 This class, when installed with an OpenerDirector, automatically adds
1328 the standard headers to every HTTP request and handles gzipped and
1329 deflated responses from web servers. If compression is to be avoided in
1330 a particular request, the original request in the program code only has
1331 to include the HTTP header "Youtubedl-no-compression", which will be
1332 removed before making the real request.
1334 Part of this code was copied from:
1336 http://techknack.net/python-urllib2-handlers/
1338 Andrew Rowls, the author of that code, agreed to release it to the
1342 def __init__(self
, params
, *args
, **kwargs
):
1343 compat_urllib_request
.HTTPHandler
.__init
__(self
, *args
, **kwargs
)
1344 self
._params
= params
1346 def http_open(self
, req
):
1347 conn_class
= compat_http_client
.HTTPConnection
1349 socks_proxy
= req
.headers
.get('Ytdl-socks-proxy')
1351 conn_class
= make_socks_conn_class(conn_class
, socks_proxy
)
1352 del req
.headers
['Ytdl-socks-proxy']
1354 return self
.do_open(functools
.partial(
1355 _create_http_connection
, self
, conn_class
, False),
1363 return zlib
.decompress(data
, -zlib
.MAX_WBITS
)
1365 return zlib
.decompress(data
)
1371 return compat_brotli
.decompress(data
)
1373 def http_request(self
, req
):
1374 # According to RFC 3986, URLs can not contain non-ASCII characters, however this is not
1375 # always respected by websites, some tend to give out URLs with non percent-encoded
1376 # non-ASCII characters (see telemb.py, ard.py [#3412])
1377 # urllib chokes on URLs with non-ASCII characters (see http://bugs.python.org/issue3991)
1378 # To work around aforementioned issue we will replace request's original URL with
1379 # percent-encoded one
1380 # Since redirects are also affected (e.g. http://www.southpark.de/alle-episoden/s18e09)
1381 # the code of this workaround has been moved here from YoutubeDL.urlopen()
1382 url
= req
.get_full_url()
1383 url_escaped
= escape_url(url
)
1385 # Substitute URL if any change after escaping
1386 if url
!= url_escaped
:
1387 req
= update_Request(req
, url
=url_escaped
)
1389 for h
, v
in self
._params
.get('http_headers', std_headers
).items():
1390 # Capitalize is needed because of Python bug 2275: http://bugs.python.org/issue2275
1391 # The dict keys are capitalized because of this bug by urllib
1392 if h
.capitalize() not in req
.headers
:
1393 req
.add_header(h
, v
)
1395 req
.headers
= handle_youtubedl_headers(req
.headers
)
1397 if sys
.version_info
< (2, 7) and '#' in req
.get_full_url():
1398 # Python 2.6 is brain-dead when it comes to fragments
1399 req
._Request
__original
= req
._Request
__original
.partition('#')[0]
1400 req
._Request
__r
_type
= req
._Request
__r
_type
.partition('#')[0]
1404 def http_response(self
, req
, resp
):
1407 if resp
.headers
.get('Content-encoding', '') == 'gzip':
1408 content
= resp
.read()
1409 gz
= gzip
.GzipFile(fileobj
=io
.BytesIO(content
), mode
='rb')
1411 uncompressed
= io
.BytesIO(gz
.read())
1412 except IOError as original_ioerror
:
1413 # There may be junk add the end of the file
1414 # See http://stackoverflow.com/q/4928560/35070 for details
1415 for i
in range(1, 1024):
1417 gz
= gzip
.GzipFile(fileobj
=io
.BytesIO(content
[:-i
]), mode
='rb')
1418 uncompressed
= io
.BytesIO(gz
.read())
1423 raise original_ioerror
1424 resp
= compat_urllib_request
.addinfourl(uncompressed
, old_resp
.headers
, old_resp
.url
, old_resp
.code
)
1425 resp
.msg
= old_resp
.msg
1426 del resp
.headers
['Content-encoding']
1428 if resp
.headers
.get('Content-encoding', '') == 'deflate':
1429 gz
= io
.BytesIO(self
.deflate(resp
.read()))
1430 resp
= compat_urllib_request
.addinfourl(gz
, old_resp
.headers
, old_resp
.url
, old_resp
.code
)
1431 resp
.msg
= old_resp
.msg
1432 del resp
.headers
['Content-encoding']
1434 if resp
.headers
.get('Content-encoding', '') == 'br':
1435 resp
= compat_urllib_request
.addinfourl(
1436 io
.BytesIO(self
.brotli(resp
.read())), old_resp
.headers
, old_resp
.url
, old_resp
.code
)
1437 resp
.msg
= old_resp
.msg
1438 del resp
.headers
['Content-encoding']
1439 # Percent-encode redirect URL of Location HTTP header to satisfy RFC 3986 (see
1440 # https://github.com/ytdl-org/youtube-dl/issues/6457).
1441 if 300 <= resp
.code
< 400:
1442 location
= resp
.headers
.get('Location')
1444 # As of RFC 2616 default charset is iso-8859-1 that is respected by python 3
1445 if sys
.version_info
>= (3, 0):
1446 location
= location
.encode('iso-8859-1').decode('utf-8')
1448 location
= location
.decode('utf-8')
1449 location_escaped
= escape_url(location
)
1450 if location
!= location_escaped
:
1451 del resp
.headers
['Location']
1452 if sys
.version_info
< (3, 0):
1453 location_escaped
= location_escaped
.encode('utf-8')
1454 resp
.headers
['Location'] = location_escaped
1457 https_request
= http_request
1458 https_response
= http_response
1461 def make_socks_conn_class(base_class
, socks_proxy
):
1462 assert issubclass(base_class
, (
1463 compat_http_client
.HTTPConnection
, compat_http_client
.HTTPSConnection
))
1465 url_components
= compat_urlparse
.urlparse(socks_proxy
)
1466 if url_components
.scheme
.lower() == 'socks5':
1467 socks_type
= ProxyType
.SOCKS5
1468 elif url_components
.scheme
.lower() in ('socks', 'socks4'):
1469 socks_type
= ProxyType
.SOCKS4
1470 elif url_components
.scheme
.lower() == 'socks4a':
1471 socks_type
= ProxyType
.SOCKS4A
1473 def unquote_if_non_empty(s
):
1476 return compat_urllib_parse_unquote_plus(s
)
1480 url_components
.hostname
, url_components
.port
or 1080,
1482 unquote_if_non_empty(url_components
.username
),
1483 unquote_if_non_empty(url_components
.password
),
1486 class SocksConnection(base_class
):
1488 self
.sock
= sockssocket()
1489 self
.sock
.setproxy(*proxy_args
)
1490 if type(self
.timeout
) in (int, float):
1491 self
.sock
.settimeout(self
.timeout
)
1492 self
.sock
.connect((self
.host
, self
.port
))
1494 if isinstance(self
, compat_http_client
.HTTPSConnection
):
1495 if hasattr(self
, '_context'): # Python > 2.6
1496 self
.sock
= self
._context
.wrap_socket(
1497 self
.sock
, server_hostname
=self
.host
)
1499 self
.sock
= ssl
.wrap_socket(self
.sock
)
1501 return SocksConnection
1504 class YoutubeDLHTTPSHandler(compat_urllib_request
.HTTPSHandler
):
1505 def __init__(self
, params
, https_conn_class
=None, *args
, **kwargs
):
1506 compat_urllib_request
.HTTPSHandler
.__init
__(self
, *args
, **kwargs
)
1507 self
._https
_conn
_class
= https_conn_class
or compat_http_client
.HTTPSConnection
1508 self
._params
= params
1510 def https_open(self
, req
):
1512 conn_class
= self
._https
_conn
_class
1514 if hasattr(self
, '_context'): # python > 2.6
1515 kwargs
['context'] = self
._context
1516 if hasattr(self
, '_check_hostname'): # python 3.x
1517 kwargs
['check_hostname'] = self
._check
_hostname
1519 socks_proxy
= req
.headers
.get('Ytdl-socks-proxy')
1521 conn_class
= make_socks_conn_class(conn_class
, socks_proxy
)
1522 del req
.headers
['Ytdl-socks-proxy']
1524 return self
.do_open(functools
.partial(
1525 _create_http_connection
, self
, conn_class
, True),
1529 class YoutubeDLCookieJar(compat_cookiejar
.MozillaCookieJar
):
1531 See [1] for cookie file format.
1533 1. https://curl.haxx.se/docs/http-cookies.html
1535 _HTTPONLY_PREFIX
= '#HttpOnly_'
1537 _HEADER
= '''# Netscape HTTP Cookie File
1538 # This file is generated by yt-dlp. Do not edit.
1541 _CookieFileEntry
= collections
.namedtuple(
1543 ('domain_name', 'include_subdomains', 'path', 'https_only', 'expires_at', 'name', 'value'))
1545 def save(self
, filename
=None, ignore_discard
=False, ignore_expires
=False):
1547 Save cookies to a file.
1549 Most of the code is taken from CPython 3.8 and slightly adapted
1550 to support cookie files with UTF-8 in both python 2 and 3.
1552 if filename
is None:
1553 if self
.filename
is not None:
1554 filename
= self
.filename
1556 raise ValueError(compat_cookiejar
.MISSING_FILENAME_TEXT
)
1558 # Store session cookies with `expires` set to 0 instead of an empty
1561 if cookie
.expires
is None:
1564 with io
.open(filename
, 'w', encoding
='utf-8') as f
:
1565 f
.write(self
._HEADER
)
1568 if not ignore_discard
and cookie
.discard
:
1570 if not ignore_expires
and cookie
.is_expired(now
):
1576 if cookie
.domain
.startswith('.'):
1577 initial_dot
= 'TRUE'
1579 initial_dot
= 'FALSE'
1580 if cookie
.expires
is not None:
1581 expires
= compat_str(cookie
.expires
)
1584 if cookie
.value
is None:
1585 # cookies.txt regards 'Set-Cookie: foo' as a cookie
1586 # with no name, whereas http.cookiejar regards it as a
1587 # cookie with no value.
1592 value
= cookie
.value
1594 '\t'.join([cookie
.domain
, initial_dot
, cookie
.path
,
1595 secure
, expires
, name
, value
]) + '\n')
1597 def load(self
, filename
=None, ignore_discard
=False, ignore_expires
=False):
1598 """Load cookies from a file."""
1599 if filename
is None:
1600 if self
.filename
is not None:
1601 filename
= self
.filename
1603 raise ValueError(compat_cookiejar
.MISSING_FILENAME_TEXT
)
1605 def prepare_line(line
):
1606 if line
.startswith(self
._HTTPONLY
_PREFIX
):
1607 line
= line
[len(self
._HTTPONLY
_PREFIX
):]
1608 # comments and empty lines are fine
1609 if line
.startswith('#') or not line
.strip():
1611 cookie_list
= line
.split('\t')
1612 if len(cookie_list
) != self
._ENTRY
_LEN
:
1613 raise compat_cookiejar
.LoadError('invalid length %d' % len(cookie_list
))
1614 cookie
= self
._CookieFileEntry
(*cookie_list
)
1615 if cookie
.expires_at
and not cookie
.expires_at
.isdigit():
1616 raise compat_cookiejar
.LoadError('invalid expires at %s' % cookie
.expires_at
)
1620 with io
.open(filename
, encoding
='utf-8') as f
:
1623 cf
.write(prepare_line(line
))
1624 except compat_cookiejar
.LoadError
as e
:
1626 'WARNING: skipping cookie file entry due to %s: %r\n'
1627 % (e
, line
), sys
.stderr
)
1630 self
._really
_load
(cf
, filename
, ignore_discard
, ignore_expires
)
1631 # Session cookies are denoted by either `expires` field set to
1632 # an empty string or 0. MozillaCookieJar only recognizes the former
1633 # (see [1]). So we need force the latter to be recognized as session
1634 # cookies on our own.
1635 # Session cookies may be important for cookies-based authentication,
1636 # e.g. usually, when user does not check 'Remember me' check box while
1637 # logging in on a site, some important cookies are stored as session
1638 # cookies so that not recognizing them will result in failed login.
1639 # 1. https://bugs.python.org/issue17164
1641 # Treat `expires=0` cookies as session cookies
1642 if cookie
.expires
== 0:
1643 cookie
.expires
= None
1644 cookie
.discard
= True
1647 class YoutubeDLCookieProcessor(compat_urllib_request
.HTTPCookieProcessor
):
1648 def __init__(self
, cookiejar
=None):
1649 compat_urllib_request
.HTTPCookieProcessor
.__init
__(self
, cookiejar
)
1651 def http_response(self
, request
, response
):
1652 # Python 2 will choke on next HTTP request in row if there are non-ASCII
1653 # characters in Set-Cookie HTTP header of last response (see
1654 # https://github.com/ytdl-org/youtube-dl/issues/6769).
1655 # In order to at least prevent crashing we will percent encode Set-Cookie
1656 # header before HTTPCookieProcessor starts processing it.
1657 # if sys.version_info < (3, 0) and response.headers:
1658 # for set_cookie_header in ('Set-Cookie', 'Set-Cookie2'):
1659 # set_cookie = response.headers.get(set_cookie_header)
1661 # set_cookie_escaped = compat_urllib_parse.quote(set_cookie, b"%/;:@&=+$,!~*'()?#[] ")
1662 # if set_cookie != set_cookie_escaped:
1663 # del response.headers[set_cookie_header]
1664 # response.headers[set_cookie_header] = set_cookie_escaped
1665 return compat_urllib_request
.HTTPCookieProcessor
.http_response(self
, request
, response
)
1667 https_request
= compat_urllib_request
.HTTPCookieProcessor
.http_request
1668 https_response
= http_response
1671 class YoutubeDLRedirectHandler(compat_urllib_request
.HTTPRedirectHandler
):
1672 """YoutubeDL redirect handler
1674 The code is based on HTTPRedirectHandler implementation from CPython [1].
1676 This redirect handler solves two issues:
1677 - ensures redirect URL is always unicode under python 2
1678 - introduces support for experimental HTTP response status code
1679 308 Permanent Redirect [2] used by some sites [3]
1681 1. https://github.com/python/cpython/blob/master/Lib/urllib/request.py
1682 2. https://developer.mozilla.org/en-US/docs/Web/HTTP/Status/308
1683 3. https://github.com/ytdl-org/youtube-dl/issues/28768
1686 http_error_301
= http_error_303
= http_error_307
= http_error_308
= compat_urllib_request
.HTTPRedirectHandler
.http_error_302
1688 def redirect_request(self
, req
, fp
, code
, msg
, headers
, newurl
):
1689 """Return a Request or None in response to a redirect.
1691 This is called by the http_error_30x methods when a
1692 redirection response is received. If a redirection should
1693 take place, return a new Request to allow http_error_30x to
1694 perform the redirect. Otherwise, raise HTTPError if no-one
1695 else should try to handle this url. Return None if you can't
1696 but another Handler might.
1698 m
= req
.get_method()
1699 if (not (code
in (301, 302, 303, 307, 308) and m
in ("GET", "HEAD")
1700 or code
in (301, 302, 303) and m
== "POST")):
1701 raise compat_HTTPError(req
.full_url
, code
, msg
, headers
, fp
)
1702 # Strictly (according to RFC 2616), 301 or 302 in response to
1703 # a POST MUST NOT cause a redirection without confirmation
1704 # from the user (of urllib.request, in this case). In practice,
1705 # essentially all clients do redirect in this case, so we do
1708 # On python 2 urlh.geturl() may sometimes return redirect URL
1709 # as byte string instead of unicode. This workaround allows
1710 # to force it always return unicode.
1711 if sys
.version_info
[0] < 3:
1712 newurl
= compat_str(newurl
)
1714 # Be conciliant with URIs containing a space. This is mainly
1715 # redundant with the more complete encoding done in http_error_302(),
1716 # but it is kept for compatibility with other callers.
1717 newurl
= newurl
.replace(' ', '%20')
1719 CONTENT_HEADERS
= ("content-length", "content-type")
1720 # NB: don't use dict comprehension for python 2.6 compatibility
1721 newheaders
= dict((k
, v
) for k
, v
in req
.headers
.items()
1722 if k
.lower() not in CONTENT_HEADERS
)
1723 return compat_urllib_request
.Request(
1724 newurl
, headers
=newheaders
, origin_req_host
=req
.origin_req_host
,
1728 def extract_timezone(date_str
):
1731 ^.{8,}? # >=8 char non-TZ prefix, if present
1732 (?P<tz>Z| # just the UTC Z, or
1733 (?:(?<=.\b\d{4}|\b\d{2}:\d\d)| # preceded by 4 digits or hh:mm or
1734 (?<!.\b[a-zA-Z]{3}|[a-zA-Z]{4}|..\b\d\d)) # not preceded by 3 alpha word or >= 4 alpha or 2 digits
1735 [ ]? # optional space
1736 (?P<sign>\+|-) # +/-
1737 (?P<hours>[0-9]{2}):?(?P<minutes>[0-9]{2}) # hh[:]mm
1741 timezone
= datetime
.timedelta()
1743 date_str
= date_str
[:-len(m
.group('tz'))]
1744 if not m
.group('sign'):
1745 timezone
= datetime
.timedelta()
1747 sign
= 1 if m
.group('sign') == '+' else -1
1748 timezone
= datetime
.timedelta(
1749 hours
=sign
* int(m
.group('hours')),
1750 minutes
=sign
* int(m
.group('minutes')))
1751 return timezone
, date_str
1754 def parse_iso8601(date_str
, delimiter
='T', timezone
=None):
1755 """ Return a UNIX timestamp from the given date """
1757 if date_str
is None:
1760 date_str
= re
.sub(r
'\.[0-9]+', '', date_str
)
1762 if timezone
is None:
1763 timezone
, date_str
= extract_timezone(date_str
)
1766 date_format
= '%Y-%m-%d{0}%H:%M:%S'.format(delimiter
)
1767 dt
= datetime
.datetime
.strptime(date_str
, date_format
) - timezone
1768 return calendar
.timegm(dt
.timetuple())
1773 def date_formats(day_first
=True):
1774 return DATE_FORMATS_DAY_FIRST
if day_first
else DATE_FORMATS_MONTH_FIRST
1777 def unified_strdate(date_str
, day_first
=True):
1778 """Return a string with the date in the format YYYYMMDD"""
1780 if date_str
is None:
1784 date_str
= date_str
.replace(',', ' ')
1785 # Remove AM/PM + timezone
1786 date_str
= re
.sub(r
'(?i)\s*(?:AM|PM)(?:\s+[A-Z]+)?', '', date_str
)
1787 _
, date_str
= extract_timezone(date_str
)
1789 for expression
in date_formats(day_first
):
1791 upload_date
= datetime
.datetime
.strptime(date_str
, expression
).strftime('%Y%m%d')
1794 if upload_date
is None:
1795 timetuple
= email
.utils
.parsedate_tz(date_str
)
1798 upload_date
= datetime
.datetime(*timetuple
[:6]).strftime('%Y%m%d')
1801 if upload_date
is not None:
1802 return compat_str(upload_date
)
1805 def unified_timestamp(date_str
, day_first
=True):
1806 if date_str
is None:
1809 date_str
= re
.sub(r
'[,|]', '', date_str
)
1811 pm_delta
= 12 if re
.search(r
'(?i)PM', date_str
) else 0
1812 timezone
, date_str
= extract_timezone(date_str
)
1814 # Remove AM/PM + timezone
1815 date_str
= re
.sub(r
'(?i)\s*(?:AM|PM)(?:\s+[A-Z]+)?', '', date_str
)
1817 # Remove unrecognized timezones from ISO 8601 alike timestamps
1818 m
= re
.search(r
'\d{1,2}:\d{1,2}(?:\.\d+)?(?P<tz>\s*[A-Z]+)$', date_str
)
1820 date_str
= date_str
[:-len(m
.group('tz'))]
1822 # Python only supports microseconds, so remove nanoseconds
1823 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]+$', date_str
)
1825 date_str
= m
.group(1)
1827 for expression
in date_formats(day_first
):
1829 dt
= datetime
.datetime
.strptime(date_str
, expression
) - timezone
+ datetime
.timedelta(hours
=pm_delta
)
1830 return calendar
.timegm(dt
.timetuple())
1833 timetuple
= email
.utils
.parsedate_tz(date_str
)
1835 return calendar
.timegm(timetuple
) + pm_delta
* 3600
1838 def determine_ext(url
, default_ext
='unknown_video'):
1839 if url
is None or '.' not in url
:
1841 guess
= url
.partition('?')[0].rpartition('.')[2]
1842 if re
.match(r
'^[A-Za-z0-9]+$', guess
):
1844 # Try extract ext from URLs like http://example.com/foo/bar.mp4/?download
1845 elif guess
.rstrip('/') in KNOWN_EXTENSIONS
:
1846 return guess
.rstrip('/')
1851 def subtitles_filename(filename
, sub_lang
, sub_format
, expected_real_ext
=None):
1852 return replace_extension(filename
, sub_lang
+ '.' + sub_format
, expected_real_ext
)
1855 def datetime_from_str(date_str
, precision
='auto', format
='%Y%m%d'):
1857 Return a datetime object from a string in the format YYYYMMDD or
1858 (now|today|yesterday|date)[+-][0-9](microsecond|second|minute|hour|day|week|month|year)(s)?
1860 format: string date format used to return datetime object from
1861 precision: round the time portion of a datetime object.
1862 auto|microsecond|second|minute|hour|day.
1863 auto: round to the unit provided in date_str (if applicable).
1865 auto_precision
= False
1866 if precision
== 'auto':
1867 auto_precision
= True
1868 precision
= 'microsecond'
1869 today
= datetime_round(datetime
.datetime
.utcnow(), precision
)
1870 if date_str
in ('now', 'today'):
1872 if date_str
== 'yesterday':
1873 return today
- datetime
.timedelta(days
=1)
1875 r
'(?P<start>.+)(?P<sign>[+-])(?P<time>\d+)(?P<unit>microsecond|second|minute|hour|day|week|month|year)(s)?',
1877 if match
is not None:
1878 start_time
= datetime_from_str(match
.group('start'), precision
, format
)
1879 time
= int(match
.group('time')) * (-1 if match
.group('sign') == '-' else 1)
1880 unit
= match
.group('unit')
1881 if unit
== 'month' or unit
== 'year':
1882 new_date
= datetime_add_months(start_time
, time
* 12 if unit
== 'year' else time
)
1888 delta
= datetime
.timedelta(**{unit + 's': time}
)
1889 new_date
= start_time
+ delta
1891 return datetime_round(new_date
, unit
)
1894 return datetime_round(datetime
.datetime
.strptime(date_str
, format
), precision
)
1897 def date_from_str(date_str
, format
='%Y%m%d', strict
=False):
1899 Return a datetime object from a string in the format YYYYMMDD or
1900 (now|today|yesterday|date)[+-][0-9](microsecond|second|minute|hour|day|week|month|year)(s)?
1902 If "strict", only (now|today)[+-][0-9](day|week|month|year)(s)? is allowed
1904 format: string date format used to return datetime object from
1906 if strict
and not re
.fullmatch(r
'\d{8}|(now|today)[+-]\d+(day|week|month|year)(s)?', date_str
):
1907 raise ValueError(f
'Invalid date format {date_str}')
1908 return datetime_from_str(date_str
, precision
='microsecond', format
=format
).date()
1911 def datetime_add_months(dt
, months
):
1912 """Increment/Decrement a datetime object by months."""
1913 month
= dt
.month
+ months
- 1
1914 year
= dt
.year
+ month
// 12
1915 month
= month
% 12 + 1
1916 day
= min(dt
.day
, calendar
.monthrange(year
, month
)[1])
1917 return dt
.replace(year
, month
, day
)
1920 def datetime_round(dt
, precision
='day'):
1922 Round a datetime object's time to a specific precision
1924 if precision
== 'microsecond':
1933 roundto
= lambda x
, n
: ((x
+ n
/ 2) // n
) * n
1934 timestamp
= calendar
.timegm(dt
.timetuple())
1935 return datetime
.datetime
.utcfromtimestamp(roundto(timestamp
, unit_seconds
[precision
]))
1938 def hyphenate_date(date_str
):
1940 Convert a date in 'YYYYMMDD' format to 'YYYY-MM-DD' format"""
1941 match
= re
.match(r
'^(\d\d\d\d)(\d\d)(\d\d)$', date_str
)
1942 if match
is not None:
1943 return '-'.join(match
.groups())
1948 class DateRange(object):
1949 """Represents a time interval between two dates"""
1951 def __init__(self
, start
=None, end
=None):
1952 """start and end must be strings in the format accepted by date"""
1953 if start
is not None:
1954 self
.start
= date_from_str(start
, strict
=True)
1956 self
.start
= datetime
.datetime
.min.date()
1958 self
.end
= date_from_str(end
, strict
=True)
1960 self
.end
= datetime
.datetime
.max.date()
1961 if self
.start
> self
.end
:
1962 raise ValueError('Date range: "%s" , the start date must be before the end date' % self
)
1966 """Returns a range that only contains the given day"""
1967 return cls(day
, day
)
1969 def __contains__(self
, date
):
1970 """Check if the date is in the range"""
1971 if not isinstance(date
, datetime
.date
):
1972 date
= date_from_str(date
)
1973 return self
.start
<= date
<= self
.end
1976 return '%s - %s' % (self
.start
.isoformat(), self
.end
.isoformat())
1979 def platform_name():
1980 """ Returns the platform name as a compat_str """
1981 res
= platform
.platform()
1982 if isinstance(res
, bytes):
1983 res
= res
.decode(preferredencoding())
1985 assert isinstance(res
, compat_str
)
1989 def get_windows_version():
1990 ''' Get Windows version. None if it's not running on Windows '''
1991 if compat_os_name
== 'nt':
1992 return version_tuple(platform
.win32_ver()[1])
1997 def _windows_write_string(s
, out
):
1998 """ Returns True if the string was written using special methods,
1999 False if it has yet to be written out."""
2000 # Adapted from http://stackoverflow.com/a/3259271/35070
2002 import ctypes
.wintypes
2010 fileno
= out
.fileno()
2011 except AttributeError:
2012 # If the output stream doesn't have a fileno, it's virtual
2014 except io
.UnsupportedOperation
:
2015 # Some strange Windows pseudo files?
2017 if fileno
not in WIN_OUTPUT_IDS
:
2020 GetStdHandle
= compat_ctypes_WINFUNCTYPE(
2021 ctypes
.wintypes
.HANDLE
, ctypes
.wintypes
.DWORD
)(
2022 ('GetStdHandle', ctypes
.windll
.kernel32
))
2023 h
= GetStdHandle(WIN_OUTPUT_IDS
[fileno
])
2025 WriteConsoleW
= compat_ctypes_WINFUNCTYPE(
2026 ctypes
.wintypes
.BOOL
, ctypes
.wintypes
.HANDLE
, ctypes
.wintypes
.LPWSTR
,
2027 ctypes
.wintypes
.DWORD
, ctypes
.POINTER(ctypes
.wintypes
.DWORD
),
2028 ctypes
.wintypes
.LPVOID
)(('WriteConsoleW', ctypes
.windll
.kernel32
))
2029 written
= ctypes
.wintypes
.DWORD(0)
2031 GetFileType
= compat_ctypes_WINFUNCTYPE(ctypes
.wintypes
.DWORD
, ctypes
.wintypes
.DWORD
)(('GetFileType', ctypes
.windll
.kernel32
))
2032 FILE_TYPE_CHAR
= 0x0002
2033 FILE_TYPE_REMOTE
= 0x8000
2034 GetConsoleMode
= compat_ctypes_WINFUNCTYPE(
2035 ctypes
.wintypes
.BOOL
, ctypes
.wintypes
.HANDLE
,
2036 ctypes
.POINTER(ctypes
.wintypes
.DWORD
))(
2037 ('GetConsoleMode', ctypes
.windll
.kernel32
))
2038 INVALID_HANDLE_VALUE
= ctypes
.wintypes
.DWORD(-1).value
2040 def not_a_console(handle
):
2041 if handle
== INVALID_HANDLE_VALUE
or handle
is None:
2043 return ((GetFileType(handle
) & ~FILE_TYPE_REMOTE
) != FILE_TYPE_CHAR
2044 or GetConsoleMode(handle
, ctypes
.byref(ctypes
.wintypes
.DWORD())) == 0)
2046 if not_a_console(h
):
2049 def next_nonbmp_pos(s
):
2051 return next(i
for i
, c
in enumerate(s
) if ord(c
) > 0xffff)
2052 except StopIteration:
2056 count
= min(next_nonbmp_pos(s
), 1024)
2058 ret
= WriteConsoleW(
2059 h
, s
, count
if count
else 2, ctypes
.byref(written
), None)
2061 raise OSError('Failed to write string')
2062 if not count
: # We just wrote a non-BMP character
2063 assert written
.value
== 2
2066 assert written
.value
> 0
2067 s
= s
[written
.value
:]
2071 def write_string(s
, out
=None, encoding
=None):
2074 assert type(s
) == compat_str
2076 if sys
.platform
== 'win32' and encoding
is None and hasattr(out
, 'fileno'):
2077 if _windows_write_string(s
, out
):
2080 if ('b' in getattr(out
, 'mode', '')
2081 or sys
.version_info
[0] < 3): # Python 2 lies about mode of sys.stderr
2082 byt
= s
.encode(encoding
or preferredencoding(), 'ignore')
2084 elif hasattr(out
, 'buffer'):
2085 enc
= encoding
or getattr(out
, 'encoding', None) or preferredencoding()
2086 byt
= s
.encode(enc
, 'ignore')
2087 out
.buffer.write(byt
)
2093 def bytes_to_intlist(bs
):
2096 if isinstance(bs
[0], int): # Python 3
2099 return [ord(c
) for c
in bs
]
2102 def intlist_to_bytes(xs
):
2105 return compat_struct_pack('%dB' % len(xs
), *xs
)
2108 # Cross-platform file locking
2109 if sys
.platform
== 'win32':
2110 import ctypes
.wintypes
2113 class OVERLAPPED(ctypes
.Structure
):
2115 ('Internal', ctypes
.wintypes
.LPVOID
),
2116 ('InternalHigh', ctypes
.wintypes
.LPVOID
),
2117 ('Offset', ctypes
.wintypes
.DWORD
),
2118 ('OffsetHigh', ctypes
.wintypes
.DWORD
),
2119 ('hEvent', ctypes
.wintypes
.HANDLE
),
2122 kernel32
= ctypes
.windll
.kernel32
2123 LockFileEx
= kernel32
.LockFileEx
2124 LockFileEx
.argtypes
= [
2125 ctypes
.wintypes
.HANDLE
, # hFile
2126 ctypes
.wintypes
.DWORD
, # dwFlags
2127 ctypes
.wintypes
.DWORD
, # dwReserved
2128 ctypes
.wintypes
.DWORD
, # nNumberOfBytesToLockLow
2129 ctypes
.wintypes
.DWORD
, # nNumberOfBytesToLockHigh
2130 ctypes
.POINTER(OVERLAPPED
) # Overlapped
2132 LockFileEx
.restype
= ctypes
.wintypes
.BOOL
2133 UnlockFileEx
= kernel32
.UnlockFileEx
2134 UnlockFileEx
.argtypes
= [
2135 ctypes
.wintypes
.HANDLE
, # hFile
2136 ctypes
.wintypes
.DWORD
, # dwReserved
2137 ctypes
.wintypes
.DWORD
, # nNumberOfBytesToLockLow
2138 ctypes
.wintypes
.DWORD
, # nNumberOfBytesToLockHigh
2139 ctypes
.POINTER(OVERLAPPED
) # Overlapped
2141 UnlockFileEx
.restype
= ctypes
.wintypes
.BOOL
2142 whole_low
= 0xffffffff
2143 whole_high
= 0x7fffffff
2145 def _lock_file(f
, exclusive
, block
):
2146 overlapped
= OVERLAPPED()
2147 overlapped
.Offset
= 0
2148 overlapped
.OffsetHigh
= 0
2149 overlapped
.hEvent
= 0
2150 f
._lock
_file
_overlapped
_p
= ctypes
.pointer(overlapped
)
2152 if not LockFileEx(msvcrt
.get_osfhandle(f
.fileno()),
2153 (0x2 if exclusive
else 0x0) |
(0x0 if block
else 0x1),
2154 0, whole_low
, whole_high
, f
._lock
_file
_overlapped
_p
):
2155 raise BlockingIOError('Locking file failed: %r' % ctypes
.FormatError())
2157 def _unlock_file(f
):
2158 assert f
._lock
_file
_overlapped
_p
2159 handle
= msvcrt
.get_osfhandle(f
.fileno())
2160 if not UnlockFileEx(handle
, 0, whole_low
, whole_high
, f
._lock
_file
_overlapped
_p
):
2161 raise OSError('Unlocking file failed: %r' % ctypes
.FormatError())
2167 def _lock_file(f
, exclusive
, block
):
2170 fcntl
.LOCK_SH
if not exclusive
2171 else fcntl
.LOCK_EX
if block
2172 else fcntl
.LOCK_EX | fcntl
.LOCK_NB
)
2173 except BlockingIOError
:
2175 except OSError: # AOSP does not have flock()
2177 fcntl
.LOCK_SH
if not exclusive
2178 else fcntl
.LOCK_EX
if block
2179 else fcntl
.LOCK_EX | fcntl
.LOCK_NB
)
2181 def _unlock_file(f
):
2183 fcntl
.flock(f
, fcntl
.LOCK_UN
)
2185 fcntl
.lockf(f
, fcntl
.LOCK_UN
)
2188 UNSUPPORTED_MSG
= 'file locking is not supported on this platform'
2190 def _lock_file(f
, exclusive
, block
):
2191 raise IOError(UNSUPPORTED_MSG
)
2193 def _unlock_file(f
):
2194 raise IOError(UNSUPPORTED_MSG
)
2197 class locked_file(object):
2200 def __init__(self
, filename
, mode
, block
=True, encoding
=None):
2201 assert mode
in ['r', 'rb', 'a', 'ab', 'w', 'wb']
2202 self
.f
= io
.open(filename
, mode
, encoding
=encoding
)
2206 def __enter__(self
):
2207 exclusive
= 'r' not in self
.mode
2209 _lock_file(self
.f
, exclusive
, self
.block
)
2215 def __exit__(self
, etype
, value
, traceback
):
2217 if not self
._closed
:
2218 _unlock_file(self
.f
)
2226 def write(self
, *args
):
2227 return self
.f
.write(*args
)
2229 def read(self
, *args
):
2230 return self
.f
.read(*args
)
2236 return self
.__enter
__()
2238 def close(self
, *args
):
2239 self
.__exit
__(self
, *args
, value
=False, traceback
=False)
2242 def get_filesystem_encoding():
2243 encoding
= sys
.getfilesystemencoding()
2244 return encoding
if encoding
is not None else 'utf-8'
2247 def shell_quote(args
):
2249 encoding
= get_filesystem_encoding()
2251 if isinstance(a
, bytes):
2252 # We may get a filename encoded with 'encodeFilename'
2253 a
= a
.decode(encoding
)
2254 quoted_args
.append(compat_shlex_quote(a
))
2255 return ' '.join(quoted_args
)
2258 def smuggle_url(url
, data
):
2259 """ Pass additional data in a URL for internal use. """
2261 url
, idata
= unsmuggle_url(url
, {})
2263 sdata
= compat_urllib_parse_urlencode(
2264 {'__youtubedl_smuggle': json.dumps(data)}
)
2265 return url
+ '#' + sdata
2268 def unsmuggle_url(smug_url
, default
=None):
2269 if '#__youtubedl_smuggle' not in smug_url
:
2270 return smug_url
, default
2271 url
, _
, sdata
= smug_url
.rpartition('#')
2272 jsond
= compat_parse_qs(sdata
)['__youtubedl_smuggle'][0]
2273 data
= json
.loads(jsond
)
2277 def format_decimal_suffix(num
, fmt
='%d%s', *, factor
=1000):
2278 """ Formats numbers with decimal sufixes like K, M, etc """
2279 num
, factor
= float_or_none(num
), float(factor
)
2280 if num
is None or num
< 0:
2282 exponent
= 0 if num
== 0 else int(math
.log(num
, factor
))
2283 suffix
= ['', *'kMGTPEZY'][exponent
]
2285 suffix
= {'k': 'Ki', '': ''}
.get(suffix
, f
'{suffix}i')
2286 converted
= num
/ (factor
** exponent
)
2287 return fmt
% (converted
, suffix
)
2290 def format_bytes(bytes):
2291 return format_decimal_suffix(bytes, '%.2f%sB', factor
=1024) or 'N/A'
2294 def lookup_unit_table(unit_table
, s
):
2295 units_re
= '|'.join(re
.escape(u
) for u
in unit_table
)
2297 r
'(?P<num>[0-9]+(?:[,.][0-9]*)?)\s*(?P<unit>%s)\b' % units_re
, s
)
2300 num_str
= m
.group('num').replace(',', '.')
2301 mult
= unit_table
[m
.group('unit')]
2302 return int(float(num_str
) * mult
)
2305 def parse_filesize(s
):
2309 # The lower-case forms are of course incorrect and unofficial,
2310 # but we support those too
2327 'megabytes': 1000 ** 2,
2328 'mebibytes': 1024 ** 2,
2334 'gigabytes': 1000 ** 3,
2335 'gibibytes': 1024 ** 3,
2341 'terabytes': 1000 ** 4,
2342 'tebibytes': 1024 ** 4,
2348 'petabytes': 1000 ** 5,
2349 'pebibytes': 1024 ** 5,
2355 'exabytes': 1000 ** 6,
2356 'exbibytes': 1024 ** 6,
2362 'zettabytes': 1000 ** 7,
2363 'zebibytes': 1024 ** 7,
2369 'yottabytes': 1000 ** 8,
2370 'yobibytes': 1024 ** 8,
2373 return lookup_unit_table(_UNIT_TABLE
, s
)
2380 s
= re
.sub(r
'^[^\d]+\s', '', s
).strip()
2382 if re
.match(r
'^[\d,.]+$', s
):
2383 return str_to_int(s
)
2396 ret
= lookup_unit_table(_UNIT_TABLE
, s
)
2400 mobj
= re
.match(r
'([\d,.]+)(?:$|\s)', s
)
2402 return str_to_int(mobj
.group(1))
2405 def parse_resolution(s
):
2409 mobj
= re
.search(r
'(?<![a-zA-Z0-9])(?P<w>\d+)\s*[xX×,]\s*(?P<h>\d+)(?![a-zA-Z0-9])', s
)
2412 'width': int(mobj
.group('w')),
2413 'height': int(mobj
.group('h')),
2416 mobj
= re
.search(r
'(?<![a-zA-Z0-9])(\d+)[pPiI](?![a-zA-Z0-9])', s
)
2418 return {'height': int(mobj.group(1))}
2420 mobj
= re
.search(r
'\b([48])[kK]\b', s
)
2422 return {'height': int(mobj.group(1)) * 540}
2427 def parse_bitrate(s
):
2428 if not isinstance(s
, compat_str
):
2430 mobj
= re
.search(r
'\b(\d+)\s*kbps', s
)
2432 return int(mobj
.group(1))
2435 def month_by_name(name
, lang
='en'):
2436 """ Return the number of a month by (locale-independently) English name """
2438 month_names
= MONTH_NAMES
.get(lang
, MONTH_NAMES
['en'])
2441 return month_names
.index(name
) + 1
2446 def month_by_abbreviation(abbrev
):
2447 """ Return the number of a month by (locale-independently) English
2451 return [s
[:3] for s
in ENGLISH_MONTH_NAMES
].index(abbrev
) + 1
2456 def fix_xml_ampersands(xml_str
):
2457 """Replace all the '&' by '&' in XML"""
2459 r
'&(?!amp;|lt;|gt;|apos;|quot;|#x[0-9a-fA-F]{,4};|#[0-9]{,4};)',
2464 def setproctitle(title
):
2465 assert isinstance(title
, compat_str
)
2467 # ctypes in Jython is not complete
2468 # http://bugs.jython.org/issue2148
2469 if sys
.platform
.startswith('java'):
2473 libc
= ctypes
.cdll
.LoadLibrary('libc.so.6')
2477 # LoadLibrary in Windows Python 2.7.13 only expects
2478 # a bytestring, but since unicode_literals turns
2479 # every string into a unicode string, it fails.
2481 title_bytes
= title
.encode('utf-8')
2482 buf
= ctypes
.create_string_buffer(len(title_bytes
))
2483 buf
.value
= title_bytes
2485 libc
.prctl(15, buf
, 0, 0, 0)
2486 except AttributeError:
2487 return # Strange libc, just skip this
2490 def remove_start(s
, start
):
2491 return s
[len(start
):] if s
is not None and s
.startswith(start
) else s
2494 def remove_end(s
, end
):
2495 return s
[:-len(end
)] if s
is not None and s
.endswith(end
) else s
2498 def remove_quotes(s
):
2499 if s
is None or len(s
) < 2:
2501 for quote
in ('"', "'", ):
2502 if s
[0] == quote
and s
[-1] == quote
:
2507 def get_domain(url
):
2508 domain
= re
.match(r
'(?:https?:\/\/)?(?:www\.)?(?P<domain>[^\n\/]+\.[^\n\/]+)(?:\/(.*))?', url
)
2509 return domain
.group('domain') if domain
else None
2512 def url_basename(url
):
2513 path
= compat_urlparse
.urlparse(url
).path
2514 return path
.strip('/').split('/')[-1]
2518 return re
.match(r
'https?://[^?#&]+/', url
).group()
2521 def urljoin(base
, path
):
2522 if isinstance(path
, bytes):
2523 path
= path
.decode('utf-8')
2524 if not isinstance(path
, compat_str
) or not path
:
2526 if re
.match(r
'^(?:[a-zA-Z][a-zA-Z0-9+-.]*:)?//', path
):
2528 if isinstance(base
, bytes):
2529 base
= base
.decode('utf-8')
2530 if not isinstance(base
, compat_str
) or not re
.match(
2531 r
'^(?:https?:)?//', base
):
2533 return compat_urlparse
.urljoin(base
, path
)
2536 class HEADRequest(compat_urllib_request
.Request
):
2537 def get_method(self
):
2541 class PUTRequest(compat_urllib_request
.Request
):
2542 def get_method(self
):
2546 def int_or_none(v
, scale
=1, default
=None, get_attr
=None, invscale
=1):
2547 if get_attr
and v
is not None:
2548 v
= getattr(v
, get_attr
, None)
2550 return int(v
) * invscale
// scale
2551 except (ValueError, TypeError, OverflowError):
2555 def str_or_none(v
, default
=None):
2556 return default
if v
is None else compat_str(v
)
2559 def str_to_int(int_str
):
2560 """ A more relaxed version of int_or_none """
2561 if isinstance(int_str
, compat_integer_types
):
2563 elif isinstance(int_str
, compat_str
):
2564 int_str
= re
.sub(r
'[,\.\+]', '', int_str
)
2565 return int_or_none(int_str
)
2568 def float_or_none(v
, scale
=1, invscale
=1, default
=None):
2572 return float(v
) * invscale
/ scale
2573 except (ValueError, TypeError):
2577 def bool_or_none(v
, default
=None):
2578 return v
if isinstance(v
, bool) else default
2581 def strip_or_none(v
, default
=None):
2582 return v
.strip() if isinstance(v
, compat_str
) else default
2585 def url_or_none(url
):
2586 if not url
or not isinstance(url
, compat_str
):
2589 return url
if re
.match(r
'^(?:(?:https?|rt(?:m(?:pt?[es]?|fp)|sp[su]?)|mms|ftps?):)?//', url
) else None
2592 def request_to_url(req
):
2593 if isinstance(req
, compat_urllib_request
.Request
):
2594 return req
.get_full_url()
2599 def strftime_or_none(timestamp
, date_format
, default
=None):
2600 datetime_object
= None
2602 if isinstance(timestamp
, compat_numeric_types
): # unix timestamp
2603 datetime_object
= datetime
.datetime
.utcfromtimestamp(timestamp
)
2604 elif isinstance(timestamp
, compat_str
): # assume YYYYMMDD
2605 datetime_object
= datetime
.datetime
.strptime(timestamp
, '%Y%m%d')
2606 return datetime_object
.strftime(date_format
)
2607 except (ValueError, TypeError, AttributeError):
2611 def parse_duration(s
):
2612 if not isinstance(s
, compat_basestring
):
2618 days
, hours
, mins
, secs
, ms
= [None] * 5
2619 m
= re
.match(r
'''(?x)
2621 (?:(?:(?P<days>[0-9]+):)?(?P<hours>[0-9]+):)?(?P<mins>[0-9]+):)?
2622 (?P<secs>(?(before_secs)[0-9]{1,2}|[0-9]+))
2623 (?P<ms>[.:][0-9]+)?Z?$
2626 days
, hours
, mins
, secs
, ms
= m
.group('days', 'hours', 'mins', 'secs', 'ms')
2631 [0-9]+\s*y(?:ears?)?\s*
2634 [0-9]+\s*m(?:onths?)?\s*
2637 [0-9]+\s*w(?:eeks?)?\s*
2640 (?P<days>[0-9]+)\s*d(?:ays?)?\s*
2644 (?P<hours>[0-9]+)\s*h(?:ours?)?\s*
2647 (?P<mins>[0-9]+)\s*m(?:in(?:ute)?s?)?\s*
2650 (?P<secs>[0-9]+)(?P<ms>\.[0-9]+)?\s*s(?:ec(?:ond)?s?)?\s*
2653 days
, hours
, mins
, secs
, ms
= m
.groups()
2655 m
= re
.match(r
'(?i)(?:(?P<hours>[0-9.]+)\s*(?:hours?)|(?P<mins>[0-9.]+)\s*(?:mins?\.?|minutes?)\s*)Z?$', s
)
2657 hours
, mins
= m
.groups()
2663 duration
+= float(secs
)
2665 duration
+= float(mins
) * 60
2667 duration
+= float(hours
) * 60 * 60
2669 duration
+= float(days
) * 24 * 60 * 60
2671 duration
+= float(ms
.replace(':', '.'))
2675 def prepend_extension(filename
, ext
, expected_real_ext
=None):
2676 name
, real_ext
= os
.path
.splitext(filename
)
2678 '{0}.{1}{2}'.format(name
, ext
, real_ext
)
2679 if not expected_real_ext
or real_ext
[1:] == expected_real_ext
2680 else '{0}.{1}'.format(filename
, ext
))
2683 def replace_extension(filename
, ext
, expected_real_ext
=None):
2684 name
, real_ext
= os
.path
.splitext(filename
)
2685 return '{0}.{1}'.format(
2686 name
if not expected_real_ext
or real_ext
[1:] == expected_real_ext
else filename
,
2690 def check_executable(exe
, args
=[]):
2691 """ Checks if the given binary is installed somewhere in PATH, and returns its name.
2692 args can be a list of arguments for a short output (like -version) """
2694 Popen([exe
] + args
, stdout
=subprocess
.PIPE
, stderr
=subprocess
.PIPE
).communicate_or_kill()
2700 def _get_exe_version_output(exe
, args
):
2702 # STDIN should be redirected too. On UNIX-like systems, ffmpeg triggers
2703 # SIGTTOU if yt-dlp is run in the background.
2704 # See https://github.com/ytdl-org/youtube-dl/issues/955#issuecomment-209789656
2706 [encodeArgument(exe
)] + args
, stdin
=subprocess
.PIPE
,
2707 stdout
=subprocess
.PIPE
, stderr
=subprocess
.STDOUT
).communicate_or_kill()
2710 if isinstance(out
, bytes): # Python 2.x
2711 out
= out
.decode('ascii', 'ignore')
2715 def detect_exe_version(output
, version_re
=None, unrecognized
='present'):
2716 assert isinstance(output
, compat_str
)
2717 if version_re
is None:
2718 version_re
= r
'version\s+([-0-9._a-zA-Z]+)'
2719 m
= re
.search(version_re
, output
)
2726 def get_exe_version(exe
, args
=['--version'],
2727 version_re
=None, unrecognized
='present'):
2728 """ Returns the version of the specified executable,
2729 or False if the executable is not present """
2730 out
= _get_exe_version_output(exe
, args
)
2731 return detect_exe_version(out
, version_re
, unrecognized
) if out
else False
2734 class LazyList(collections
.abc
.Sequence
):
2735 ''' Lazy immutable list from an iterable
2736 Note that slices of a LazyList are lists and not LazyList'''
2738 class IndexError(IndexError):
2741 def __init__(self
, iterable
, *, reverse
=False, _cache
=None):
2742 self
.__iterable
= iter(iterable
)
2743 self
.__cache
= [] if _cache
is None else _cache
2744 self
.__reversed
= reverse
2748 # We need to consume the entire iterable to iterate in reverse
2749 yield from self
.exhaust()
2751 yield from self
.__cache
2752 for item
in self
.__iterable
:
2753 self
.__cache
.append(item
)
2756 def __exhaust(self
):
2757 self
.__cache
.extend(self
.__iterable
)
2758 # Discard the emptied iterable to make it pickle-able
2759 self
.__iterable
= []
2763 ''' Evaluate the entire iterable '''
2764 return self
.__exhaust
()[::-1 if self
.__reversed
else 1]
2767 def __reverse_index(x
):
2768 return None if x
is None else -(x
+ 1)
2770 def __getitem__(self
, idx
):
2771 if isinstance(idx
, slice):
2773 idx
= slice(self
.__reverse
_index
(idx
.start
), self
.__reverse
_index
(idx
.stop
), -(idx
.step
or 1))
2774 start
, stop
, step
= idx
.start
, idx
.stop
, idx
.step
or 1
2775 elif isinstance(idx
, int):
2777 idx
= self
.__reverse
_index
(idx
)
2778 start
, stop
, step
= idx
, idx
, 0
2780 raise TypeError('indices must be integers or slices')
2781 if ((start
or 0) < 0 or (stop
or 0) < 0
2782 or (start
is None and step
< 0)
2783 or (stop
is None and step
> 0)):
2784 # We need to consume the entire iterable to be able to slice from the end
2785 # Obviously, never use this with infinite iterables
2788 return self
.__cache
[idx
]
2789 except IndexError as e
:
2790 raise self
.IndexError(e
) from e
2791 n
= max(start
or 0, stop
or 0) - len(self
.__cache
) + 1
2793 self
.__cache
.extend(itertools
.islice(self
.__iterable
, n
))
2795 return self
.__cache
[idx
]
2796 except IndexError as e
:
2797 raise self
.IndexError(e
) from e
2801 self
[-1] if self
.__reversed
else self
[0]
2802 except self
.IndexError:
2808 return len(self
.__cache
)
2810 def __reversed__(self
):
2811 return type(self
)(self
.__iterable
, reverse
=not self
.__reversed
, _cache
=self
.__cache
)
2814 return type(self
)(self
.__iterable
, reverse
=self
.__reversed
, _cache
=self
.__cache
)
2817 # repr and str should mimic a list. So we exhaust the iterable
2818 return repr(self
.exhaust())
2821 return repr(self
.exhaust())
2826 class IndexError(IndexError):
2830 # This is only useful for tests
2831 return len(self
.getslice())
2833 def __init__(self
, pagefunc
, pagesize
, use_cache
=True):
2834 self
._pagefunc
= pagefunc
2835 self
._pagesize
= pagesize
2836 self
._pagecount
= float('inf')
2837 self
._use
_cache
= use_cache
2840 def getpage(self
, pagenum
):
2841 page_results
= self
._cache
.get(pagenum
)
2842 if page_results
is None:
2843 page_results
= [] if pagenum
> self
._pagecount
else list(self
._pagefunc
(pagenum
))
2845 self
._cache
[pagenum
] = page_results
2848 def getslice(self
, start
=0, end
=None):
2849 return list(self
._getslice
(start
, end
))
2851 def _getslice(self
, start
, end
):
2852 raise NotImplementedError('This method must be implemented by subclasses')
2854 def __getitem__(self
, idx
):
2855 assert self
._use
_cache
, 'Indexing PagedList requires cache'
2856 if not isinstance(idx
, int) or idx
< 0:
2857 raise TypeError('indices must be non-negative integers')
2858 entries
= self
.getslice(idx
, idx
+ 1)
2860 raise self
.IndexError()
2864 class OnDemandPagedList(PagedList
):
2865 def _getslice(self
, start
, end
):
2866 for pagenum
in itertools
.count(start
// self
._pagesize
):
2867 firstid
= pagenum
* self
._pagesize
2868 nextfirstid
= pagenum
* self
._pagesize
+ self
._pagesize
2869 if start
>= nextfirstid
:
2873 start
% self
._pagesize
2874 if firstid
<= start
< nextfirstid
2877 ((end
- 1) % self
._pagesize
) + 1
2878 if (end
is not None and firstid
<= end
<= nextfirstid
)
2882 page_results
= self
.getpage(pagenum
)
2884 self
._pagecount
= pagenum
- 1
2886 if startv
!= 0 or endv
is not None:
2887 page_results
= page_results
[startv
:endv
]
2888 yield from page_results
2890 # A little optimization - if current page is not "full", ie. does
2891 # not contain page_size videos then we can assume that this page
2892 # is the last one - there are no more ids on further pages -
2893 # i.e. no need to query again.
2894 if len(page_results
) + startv
< self
._pagesize
:
2897 # If we got the whole page, but the next page is not interesting,
2898 # break out early as well
2899 if end
== nextfirstid
:
2903 class InAdvancePagedList(PagedList
):
2904 def __init__(self
, pagefunc
, pagecount
, pagesize
):
2905 PagedList
.__init
__(self
, pagefunc
, pagesize
, True)
2906 self
._pagecount
= pagecount
2908 def _getslice(self
, start
, end
):
2909 start_page
= start
// self
._pagesize
2910 end_page
= self
._pagecount
if end
is None else min(self
._pagecount
, end
// self
._pagesize
+ 1)
2911 skip_elems
= start
- start_page
* self
._pagesize
2912 only_more
= None if end
is None else end
- start
2913 for pagenum
in range(start_page
, end_page
):
2914 page_results
= self
.getpage(pagenum
)
2916 page_results
= page_results
[skip_elems
:]
2918 if only_more
is not None:
2919 if len(page_results
) < only_more
:
2920 only_more
-= len(page_results
)
2922 yield from page_results
[:only_more
]
2924 yield from page_results
2927 def uppercase_escape(s
):
2928 unicode_escape
= codecs
.getdecoder('unicode_escape')
2930 r
'\\U[0-9a-fA-F]{8}',
2931 lambda m
: unicode_escape(m
.group(0))[0],
2935 def lowercase_escape(s
):
2936 unicode_escape
= codecs
.getdecoder('unicode_escape')
2938 r
'\\u[0-9a-fA-F]{4}',
2939 lambda m
: unicode_escape(m
.group(0))[0],
2943 def escape_rfc3986(s
):
2944 """Escape non-ASCII characters as suggested by RFC 3986"""
2945 if sys
.version_info
< (3, 0) and isinstance(s
, compat_str
):
2946 s
= s
.encode('utf-8')
2947 return compat_urllib_parse
.quote(s
, b
"%/;:@&=+$,!~*'()?#[]")
2950 def escape_url(url
):
2951 """Escape URL as suggested by RFC 3986"""
2952 url_parsed
= compat_urllib_parse_urlparse(url
)
2953 return url_parsed
._replace
(
2954 netloc
=url_parsed
.netloc
.encode('idna').decode('ascii'),
2955 path
=escape_rfc3986(url_parsed
.path
),
2956 params
=escape_rfc3986(url_parsed
.params
),
2957 query
=escape_rfc3986(url_parsed
.query
),
2958 fragment
=escape_rfc3986(url_parsed
.fragment
)
2963 return compat_parse_qs(compat_urllib_parse_urlparse(url
).query
)
2966 def read_batch_urls(batch_fd
):
2968 if not isinstance(url
, compat_str
):
2969 url
= url
.decode('utf-8', 'replace')
2970 BOM_UTF8
= ('\xef\xbb\xbf', '\ufeff')
2971 for bom
in BOM_UTF8
:
2972 if url
.startswith(bom
):
2973 url
= url
[len(bom
):]
2975 if not url
or url
.startswith(('#', ';', ']')):
2977 # "#" cannot be stripped out since it is part of the URI
2978 # However, it can be safely stipped out if follwing a whitespace
2979 return re
.split(r
'\s#', url
, 1)[0].rstrip()
2981 with contextlib
.closing(batch_fd
) as fd
:
2982 return [url
for url
in map(fixup
, fd
) if url
]
2985 def urlencode_postdata(*args
, **kargs
):
2986 return compat_urllib_parse_urlencode(*args
, **kargs
).encode('ascii')
2989 def update_url_query(url
, query
):
2992 parsed_url
= compat_urlparse
.urlparse(url
)
2993 qs
= compat_parse_qs(parsed_url
.query
)
2995 return compat_urlparse
.urlunparse(parsed_url
._replace
(
2996 query
=compat_urllib_parse_urlencode(qs
, True)))
2999 def update_Request(req
, url
=None, data
=None, headers
={}, query={}
):
3000 req_headers
= req
.headers
.copy()
3001 req_headers
.update(headers
)
3002 req_data
= data
or req
.data
3003 req_url
= update_url_query(url
or req
.get_full_url(), query
)
3004 req_get_method
= req
.get_method()
3005 if req_get_method
== 'HEAD':
3006 req_type
= HEADRequest
3007 elif req_get_method
== 'PUT':
3008 req_type
= PUTRequest
3010 req_type
= compat_urllib_request
.Request
3012 req_url
, data
=req_data
, headers
=req_headers
,
3013 origin_req_host
=req
.origin_req_host
, unverifiable
=req
.unverifiable
)
3014 if hasattr(req
, 'timeout'):
3015 new_req
.timeout
= req
.timeout
3019 def _multipart_encode_impl(data
, boundary
):
3020 content_type
= 'multipart/form-data; boundary=%s' % boundary
3023 for k
, v
in data
.items():
3024 out
+= b
'--' + boundary
.encode('ascii') + b
'\r\n'
3025 if isinstance(k
, compat_str
):
3026 k
= k
.encode('utf-8')
3027 if isinstance(v
, compat_str
):
3028 v
= v
.encode('utf-8')
3029 # RFC 2047 requires non-ASCII field names to be encoded, while RFC 7578
3030 # suggests sending UTF-8 directly. Firefox sends UTF-8, too
3031 content
= b
'Content-Disposition: form-data; name="' + k
+ b
'"\r\n\r\n' + v
+ b
'\r\n'
3032 if boundary
.encode('ascii') in content
:
3033 raise ValueError('Boundary overlaps with data')
3036 out
+= b
'--' + boundary
.encode('ascii') + b
'--\r\n'
3038 return out
, content_type
3041 def multipart_encode(data
, boundary
=None):
3043 Encode a dict to RFC 7578-compliant form-data
3046 A dict where keys and values can be either Unicode or bytes-like
3049 If specified a Unicode object, it's used as the boundary. Otherwise
3050 a random boundary is generated.
3052 Reference: https://tools.ietf.org/html/rfc7578
3054 has_specified_boundary
= boundary
is not None
3057 if boundary
is None:
3058 boundary
= '---------------' + str(random
.randrange(0x0fffffff, 0xffffffff))
3061 out
, content_type
= _multipart_encode_impl(data
, boundary
)
3064 if has_specified_boundary
:
3068 return out
, content_type
3071 def dict_get(d
, key_or_keys
, default
=None, skip_false_values
=True):
3072 if isinstance(key_or_keys
, (list, tuple)):
3073 for key
in key_or_keys
:
3074 if key
not in d
or d
[key
] is None or skip_false_values
and not d
[key
]:
3078 return d
.get(key_or_keys
, default
)
3081 def try_get(src
, getter
, expected_type
=None):
3082 for get
in variadic(getter
):
3085 except (AttributeError, KeyError, TypeError, IndexError):
3088 if expected_type
is None or isinstance(v
, expected_type
):
3092 def merge_dicts(*dicts
):
3094 for a_dict
in dicts
:
3095 for k
, v
in a_dict
.items():
3099 or (isinstance(v
, compat_str
) and v
3100 and isinstance(merged
[k
], compat_str
)
3101 and not merged
[k
])):
3106 def encode_compat_str(string
, encoding
=preferredencoding(), errors
='strict'):
3107 return string
if isinstance(string
, compat_str
) else compat_str(string
, encoding
, errors
)
3119 TV_PARENTAL_GUIDELINES
= {
3129 def parse_age_limit(s
):
3131 return s
if 0 <= s
<= 21 else None
3132 if not isinstance(s
, compat_basestring
):
3134 m
= re
.match(r
'^(?P<age>\d{1,2})\+?$', s
)
3136 return int(m
.group('age'))
3139 return US_RATINGS
[s
]
3140 m
= re
.match(r
'^TV[_-]?(%s)$' % '|'.join(k
[3:] for k
in TV_PARENTAL_GUIDELINES
), s
)
3142 return TV_PARENTAL_GUIDELINES
['TV-' + m
.group(1)]
3146 def strip_jsonp(code
):
3149 (?:window\.)?(?P<func_name>[a-zA-Z0-9_.$]*)
3150 (?:\s*&&\s*(?P=func_name))?
3151 \s*\(\s*(?P<callback_data>.*)\);?
3152 \s*?(?://[^\n]*)*$''',
3153 r
'\g<callback_data>', code
)
3156 def js_to_json(code
, vars={}):
3157 # vars is a dict of var, val pairs to substitute
3158 COMMENT_RE
= r
'/\*(?:(?!\*/).)*?\*/|//[^\n]*\n'
3159 SKIP_RE
= r
'\s*(?:{comment})?\s*'.format(comment
=COMMENT_RE
)
3161 (r
'(?s)^(0[xX][0-9a-fA-F]+){skip}:?$'.format(skip
=SKIP_RE
), 16),
3162 (r
'(?s)^(0+[0-7]+){skip}:?$'.format(skip
=SKIP_RE
), 8),
3167 if v
in ('true', 'false', 'null'):
3169 elif v
in ('undefined', 'void 0'):
3171 elif v
.startswith('/*') or v
.startswith('//') or v
.startswith('!') or v
== ',':
3174 if v
[0] in ("'", '"'):
3175 v
= re
.sub(r
'(?s)\\.|"', lambda m
: {
3180 }.get(m
.group(0), m
.group(0)), v
[1:-1])
3182 for regex
, base
in INTEGER_TABLE
:
3183 im
= re
.match(regex
, v
)
3185 i
= int(im
.group(1), base
)
3186 return '"%d":' % i
if v
.endswith(':') else '%d' % i
3193 code
= re
.sub(r
'new Date\((".+")\)', r
'\g<1>', code
)
3195 return re
.sub(r
'''(?sx)
3196 "(?:[^"\\]*(?:\\\\|\\['"nurtbfx/\n]))*[^"\\]*"|
3197 '(?:[^'\\]*(?:\\\\|\\['"nurtbfx/\n]))*[^'\\]*'|
3198 {comment}|,(?={skip}[\]}}])|
3199 void\s0|(?:(?<![0-9])[eE]|[a-df-zA-DF-Z_$])[.a-zA-Z_$0-9]*|
3200 \b(?:0[xX][0-9a-fA-F]+|0+[0-7]+)(?:{skip}:)?|
3203 '''.format(comment
=COMMENT_RE
, skip
=SKIP_RE
), fix_kv
, code
)
3206 def qualities(quality_ids
):
3207 """ Get a numeric quality value out of a list of possible values """
3210 return quality_ids
.index(qid
)
3216 POSTPROCESS_WHEN
= {'pre_process', 'after_filter', 'before_dl', 'after_move', 'post_process', 'after_video', 'playlist'}
3220 'default': '%(title)s [%(id)s].%(ext)s',
3221 'chapter': '%(title)s - %(section_number)03d %(section_title)s [%(id)s].%(ext)s',
3227 'description': 'description',
3228 'annotation': 'annotations.xml',
3229 'infojson': 'info.json',
3232 'pl_thumbnail': None,
3233 'pl_description': 'description',
3234 'pl_infojson': 'info.json',
3237 # As of [1] format syntax is:
3238 # %[mapping_key][conversion_flags][minimum_width][.precision][length_modifier]type
3239 # 1. https://docs.python.org/2/library/stdtypes.html#string-formatting
3240 STR_FORMAT_RE_TMPL
= r
'''(?x)
3241 (?<!%)(?P<prefix>(?:%%)*)
3243 (?P<has_key>\((?P<key>{0})\))?
3245 (?P<conversion>[#0\-+ ]+)?
3247 (?P<precision>\.\d+)?
3248 (?P<len_mod>[hlL])? # unused in python
3249 {1} # conversion type
3254 STR_FORMAT_TYPES
= 'diouxXeEfFgGcrs'
3257 def limit_length(s
, length
):
3258 """ Add ellipses to overly long strings """
3263 return s
[:length
- len(ELLIPSES
)] + ELLIPSES
3267 def version_tuple(v
):
3268 return tuple(int(e
) for e
in re
.split(r
'[-.]', v
))
3271 def is_outdated_version(version
, limit
, assume_new
=True):
3273 return not assume_new
3275 return version_tuple(version
) < version_tuple(limit
)
3277 return not assume_new
3280 def ytdl_is_updateable():
3281 """ Returns if yt-dlp can be updated with -U """
3283 from .update
import is_non_updateable
3285 return not is_non_updateable()
3288 def args_to_str(args
):
3289 # Get a short string representation for a subprocess command
3290 return ' '.join(compat_shlex_quote(a
) for a
in args
)
3293 def error_to_compat_str(err
):
3295 # On python 2 error byte string must be decoded with proper
3296 # encoding rather than ascii
3297 if sys
.version_info
[0] < 3:
3298 err_str
= err_str
.decode(preferredencoding())
3302 def mimetype2ext(mt
):
3306 mt
, _
, params
= mt
.partition(';')
3311 # Per RFC 3003, audio/mpeg can be .mp1, .mp2 or .mp3. Here use .mp3 as
3312 # it's the most popular one
3313 'audio/mpeg': 'mp3',
3314 'audio/x-wav': 'wav',
3316 'audio/wave': 'wav',
3319 ext
= FULL_MAP
.get(mt
)
3325 'smptett+xml': 'tt',
3329 'x-mp4-fragmented': 'mp4',
3330 'x-ms-sami': 'sami',
3333 'x-mpegurl': 'm3u8',
3334 'vnd.apple.mpegurl': 'm3u8',
3338 'vnd.ms-sstr+xml': 'ism',
3342 'filmstrip+json': 'fs',
3346 _
, _
, subtype
= mt
.rpartition('/')
3347 ext
= SUBTYPE_MAP
.get(subtype
.lower())
3358 _
, _
, suffix
= subtype
.partition('+')
3359 ext
= SUFFIX_MAP
.get(suffix
)
3363 return subtype
.replace('+', '.')
3366 def ext2mimetype(ext_or_url
):
3369 if '.' not in ext_or_url
:
3370 ext_or_url
= f
'file.{ext_or_url}'
3371 return mimetypes
.guess_type(ext_or_url
)[0]
3374 def parse_codecs(codecs_str
):
3375 # http://tools.ietf.org/html/rfc6381
3378 split_codecs
= list(filter(None, map(
3379 str.strip
, codecs_str
.strip().strip(',').split(','))))
3380 vcodec
, acodec
, tcodec
, hdr
= None, None, None, None
3381 for full_codec
in split_codecs
:
3382 parts
= full_codec
.split('.')
3383 codec
= parts
[0].replace('0', '')
3384 if codec
in ('avc1', 'avc2', 'avc3', 'avc4', 'vp9', 'vp8', 'hev1', 'hev2',
3385 'h263', 'h264', 'mp4v', 'hvc1', 'av1', 'theora', 'dvh1', 'dvhe'):
3387 vcodec
= '.'.join(parts
[:4]) if codec
in ('vp9', 'av1', 'hvc1') else full_codec
3388 if codec
in ('dvh1', 'dvhe'):
3390 elif codec
== 'av1' and len(parts
) > 3 and parts
[3] == '10':
3392 elif full_codec
.replace('0', '').startswith('vp9.2'):
3394 elif codec
in ('flac', 'mp4a', 'opus', 'vorbis', 'mp3', 'aac', 'ac-3', 'ec-3', 'eac3', 'dtsc', 'dtse', 'dtsh', 'dtsl'):
3397 elif codec
in ('stpp', 'wvtt',):
3401 write_string('WARNING: Unknown codec %s\n' % full_codec
, sys
.stderr
)
3402 if vcodec
or acodec
or tcodec
:
3404 'vcodec': vcodec
or 'none',
3405 'acodec': acodec
or 'none',
3406 'dynamic_range': hdr
,
3407 **({'tcodec': tcodec}
if tcodec
is not None else {}),
3409 elif len(split_codecs
) == 2:
3411 'vcodec': split_codecs
[0],
3412 'acodec': split_codecs
[1],
3417 def urlhandle_detect_ext(url_handle
):
3418 getheader
= url_handle
.headers
.get
3420 cd
= getheader('Content-Disposition')
3422 m
= re
.match(r
'attachment;\s*filename="(?P<filename>[^"]+)"', cd
)
3424 e
= determine_ext(m
.group('filename'), default_ext
=None)
3428 return mimetype2ext(getheader('Content-Type'))
3431 def encode_data_uri(data
, mime_type
):
3432 return 'data:%s;base64,%s' % (mime_type
, base64
.b64encode(data
).decode('ascii'))
3435 def age_restricted(content_limit
, age_limit
):
3436 """ Returns True iff the content should be blocked """
3438 if age_limit
is None: # No limit set
3440 if content_limit
is None:
3441 return False # Content available for everyone
3442 return age_limit
< content_limit
3445 def is_html(first_bytes
):
3446 """ Detect whether a file contains HTML by examining its first bytes. """
3449 (b
'\xef\xbb\xbf', 'utf-8'),
3450 (b
'\x00\x00\xfe\xff', 'utf-32-be'),
3451 (b
'\xff\xfe\x00\x00', 'utf-32-le'),
3452 (b
'\xff\xfe', 'utf-16-le'),
3453 (b
'\xfe\xff', 'utf-16-be'),
3455 for bom
, enc
in BOMS
:
3456 if first_bytes
.startswith(bom
):
3457 s
= first_bytes
[len(bom
):].decode(enc
, 'replace')
3460 s
= first_bytes
.decode('utf-8', 'replace')
3462 return re
.match(r
'^\s*<', s
)
3465 def determine_protocol(info_dict
):
3466 protocol
= info_dict
.get('protocol')
3467 if protocol
is not None:
3470 url
= sanitize_url(info_dict
['url'])
3471 if url
.startswith('rtmp'):
3473 elif url
.startswith('mms'):
3475 elif url
.startswith('rtsp'):
3478 ext
= determine_ext(url
)
3484 return compat_urllib_parse_urlparse(url
).scheme
3487 def render_table(header_row
, data
, delim
=False, extra_gap
=0, hide_empty
=False):
3488 """ Render a list of rows, each as a list of values.
3489 Text after a \t will be right aligned """
3491 return len(remove_terminal_sequences(string
).replace('\t', ''))
3493 def get_max_lens(table
):
3494 return [max(width(str(v
)) for v
in col
) for col
in zip(*table
)]
3496 def filter_using_list(row
, filterArray
):
3497 return [col
for take
, col
in itertools
.zip_longest(filterArray
, row
, fillvalue
=True) if take
]
3499 max_lens
= get_max_lens(data
) if hide_empty
else []
3500 header_row
= filter_using_list(header_row
, max_lens
)
3501 data
= [filter_using_list(row
, max_lens
) for row
in data
]
3503 table
= [header_row
] + data
3504 max_lens
= get_max_lens(table
)
3507 table
= [header_row
, [delim
* (ml
+ extra_gap
) for ml
in max_lens
]] + data
3508 table
[1][-1] = table
[1][-1][:-extra_gap
* len(delim
)] # Remove extra_gap from end of delimiter
3510 for pos
, text
in enumerate(map(str, row
)):
3512 row
[pos
] = text
.replace('\t', ' ' * (max_lens
[pos
] - width(text
))) + ' ' * extra_gap
3514 row
[pos
] = text
+ ' ' * (max_lens
[pos
] - width(text
) + extra_gap
)
3515 ret
= '\n'.join(''.join(row
).rstrip() for row
in table
)
3519 def _match_one(filter_part
, dct
, incomplete
):
3520 # TODO: Generalize code with YoutubeDL._build_format_filter
3521 STRING_OPERATORS
= {
3522 '*=': operator
.contains
,
3523 '^=': lambda attr
, value
: attr
.startswith(value
),
3524 '$=': lambda attr
, value
: attr
.endswith(value
),
3525 '~=': lambda attr
, value
: re
.search(value
, attr
),
3527 COMPARISON_OPERATORS
= {
3529 '<=': operator
.le
, # "<=" must be defined above "<"
3536 operator_rex
= re
.compile(r
'''(?x)\s*
3538 \s*(?P<negation>!\s*)?(?P<op>%s)(?P<none_inclusive>\s*\?)?\s*
3540 (?P<quote>["\'])(?P
<quotedstrval
>.+?
)(?P
=quote
)|
3544 ''' % '|'.join(map(re.escape, COMPARISON_OPERATORS.keys())))
3545 m = operator_rex.search(filter_part)
3548 unnegated_op = COMPARISON_OPERATORS[m['op']]
3550 op = lambda attr, value: not unnegated_op(attr, value)
3553 comparison_value = m['quotedstrval'] or m['strval'] or m['intval']
3555 comparison_value = comparison_value.replace(r'\%s' % m['quote'], m['quote'])
3556 actual_value = dct.get(m['key'])
3557 numeric_comparison = None
3558 if isinstance(actual_value, compat_numeric_types):
3559 # If the original field is a string and matching comparisonvalue is
3560 # a number we should respect the origin of the original field
3561 # and process comparison value as a string (see
3562 # https://github.com/ytdl-org/youtube-dl/issues/11082)
3564 numeric_comparison = int(comparison_value)
3566 numeric_comparison = parse_filesize(comparison_value)
3567 if numeric_comparison is None:
3568 numeric_comparison = parse_filesize(f'{comparison_value}B')
3569 if numeric_comparison is None:
3570 numeric_comparison = parse_duration(comparison_value)
3571 if numeric_comparison is not None and m['op'] in STRING_OPERATORS:
3572 raise ValueError('Operator %s only supports string values!' % m['op'])
3573 if actual_value is None:
3574 return incomplete or m['none_inclusive']
3575 return op(actual_value, comparison_value if numeric_comparison is None else numeric_comparison)
3578 '': lambda v: (v is True) if isinstance(v, bool) else (v is not None),
3579 '!': lambda v: (v is False) if isinstance(v, bool) else (v is None),
3581 operator_rex = re.compile(r'''(?x
)\s
*
3582 (?P
<op
>%s)\s
*(?P
<key
>[a
-z_
]+)
3584 ''' % '|'.join(map(re.escape, UNARY_OPERATORS.keys())))
3585 m = operator_rex.search(filter_part)
3587 op = UNARY_OPERATORS[m.group('op')]
3588 actual_value = dct.get(m.group('key'))
3589 if incomplete and actual_value is None:
3591 return op(actual_value)
3593 raise ValueError('Invalid filter part %r' % filter_part)
3596 def match_str(filter_str, dct, incomplete=False):
3597 """ Filter a dictionary with a simple string syntax. Returns True (=passes filter) or false
3598 When incomplete, all conditions passes on missing fields
3601 _match_one(filter_part.replace(r'\&', '&'), dct, incomplete)
3602 for filter_part in re.split(r'(?<!\\)&', filter_str))
3605 def match_filter_func(filter_str):
3606 def _match_func(info_dict, *args, **kwargs):
3607 if match_str(filter_str, info_dict, *args, **kwargs):
3610 video_title = info_dict.get('title', info_dict.get('id', 'video'))
3611 return '%s does not pass filter %s, skipping ..' % (video_title, filter_str)
3615 def parse_dfxp_time_expr(time_expr):
3619 mobj = re.match(r'^(?P<time_offset>\d+(?:\.\d+)?)s?$', time_expr)
3621 return float(mobj.group('time_offset'))
3623 mobj = re.match(r'^(\d+):(\d\d):(\d\d(?:(?:\.|:)\d+)?)$', time_expr)
3625 return 3600 * int(mobj.group(1)) + 60 * int(mobj.group(2)) + float(mobj.group(3).replace(':', '.'))
3628 def srt_subtitles_timecode(seconds):
3629 return '%02d:%02d:%02d,%03d' % timetuple_from_msec(seconds * 1000)
3632 def ass_subtitles_timecode(seconds):
3633 time = timetuple_from_msec(seconds * 1000)
3634 return '%01d:%02d:%02d.%02d' % (*time[:-1], time.milliseconds / 10)
3637 def dfxp2srt(dfxp_data):
3639 @param dfxp_data A
bytes-like
object containing DFXP data
3640 @returns A
unicode object containing converted SRT data
3642 LEGACY_NAMESPACES = (
3643 (b'http://www.w3.org/ns/ttml', [
3644 b'http://www.w3.org/2004/11/ttaf1',
3645 b'http://www.w3.org/2006/04/ttaf1',
3646 b'http://www.w3.org/2006/10/ttaf1',
3648 (b'http://www.w3.org/ns/ttml#styling', [
3649 b'http://www.w3.org/ns/ttml#style',
3653 SUPPORTED_STYLING = [
3662 _x = functools.partial(xpath_with_ns, ns_map={
3663 'xml': 'http://www.w3.org/XML/1998/namespace',
3664 'ttml': 'http://www.w3.org/ns/ttml',
3665 'tts': 'http://www.w3.org/ns/ttml#styling',
3671 class TTMLPElementParser(object):
3673 _unclosed_elements = []
3674 _applied_styles = []
3676 def start(self, tag, attrib):
3677 if tag in (_x('ttml:br'), 'br'):
3680 unclosed_elements = []
3682 element_style_id = attrib.get('style')
3684 style.update(default_style)
3685 if element_style_id:
3686 style.update(styles.get(element_style_id, {}))
3687 for prop in SUPPORTED_STYLING:
3688 prop_val = attrib.get(_x('tts:' + prop))
3690 style[prop] = prop_val
3693 for k, v in sorted(style.items()):
3694 if self._applied_styles and self._applied_styles[-1].get(k) == v:
3697 font += ' color="%s"' % v
3698 elif k == 'fontSize':
3699 font += ' size="%s"' % v
3700 elif k == 'fontFamily':
3701 font += ' face="%s"' % v
3702 elif k == 'fontWeight' and v == 'bold':
3704 unclosed_elements.append('b')
3705 elif k == 'fontStyle' and v == 'italic':
3707 unclosed_elements.append('i')
3708 elif k == 'textDecoration' and v == 'underline':
3710 unclosed_elements.append('u')
3712 self._out += '<font' + font + '>'
3713 unclosed_elements.append('font')
3715 if self._applied_styles:
3716 applied_style.update(self._applied_styles[-1])
3717 applied_style.update(style)
3718 self._applied_styles.append(applied_style)
3719 self._unclosed_elements.append(unclosed_elements)
3722 if tag not in (_x('ttml:br'), 'br'):
3723 unclosed_elements = self._unclosed_elements.pop()
3724 for element in reversed(unclosed_elements):
3725 self._out += '</%s>' % element
3726 if unclosed_elements and self._applied_styles:
3727 self._applied_styles.pop()
3729 def data(self, data):
3733 return self._out.strip()
3735 def parse_node(node):
3736 target = TTMLPElementParser()
3737 parser = xml.etree.ElementTree.XMLParser(target=target)
3738 parser.feed(xml.etree.ElementTree.tostring(node))
3739 return parser.close()
3741 for k, v in LEGACY_NAMESPACES:
3743 dfxp_data = dfxp_data.replace(ns, k)
3745 dfxp = compat_etree_fromstring(dfxp_data)
3747 paras = dfxp.findall(_x('.//ttml:p')) or dfxp.findall('.//p')
3750 raise ValueError('Invalid dfxp/TTML subtitle')
3754 for style in dfxp.findall(_x('.//ttml:style')):
3755 style_id = style.get('id') or style.get(_x('xml:id'))
3758 parent_style_id = style.get('style')
3760 if parent_style_id not in styles:
3763 styles[style_id] = styles[parent_style_id].copy()
3764 for prop in SUPPORTED_STYLING:
3765 prop_val = style.get(_x('tts:' + prop))
3767 styles.setdefault(style_id, {})[prop] = prop_val
3773 for p in ('body', 'div'):
3774 ele = xpath_element(dfxp, [_x('.//ttml:' + p), './/' + p])
3777 style = styles.get(ele.get('style'))
3780 default_style.update(style)
3782 for para, index in zip(paras, itertools.count(1)):
3783 begin_time = parse_dfxp_time_expr(para.attrib.get('begin'))
3784 end_time = parse_dfxp_time_expr(para.attrib.get('end'))
3785 dur = parse_dfxp_time_expr(para.attrib.get('dur'))
3786 if begin_time is None:
3791 end_time = begin_time + dur
3792 out.append('%d\n%s --> %s\n%s\n\n' % (
3794 srt_subtitles_timecode(begin_time),
3795 srt_subtitles_timecode(end_time),
3801 def cli_option(params, command_option, param):
3802 param = params.get(param)
3804 param = compat_str(param)
3805 return [command_option, param] if param is not None else []
3808 def cli_bool_option(params, command_option, param, true_value='true', false_value='false', separator=None):
3809 param = params.get(param)
3812 assert isinstance(param, bool)
3814 return [command_option + separator + (true_value if param else false_value)]
3815 return [command_option, true_value if param else false_value]
3818 def cli_valueless_option(params, command_option, param, expected_value=True):
3819 param = params.get(param)
3820 return [command_option] if param == expected_value else []
3823 def cli_configuration_args(argdict, keys, default=[], use_compat=True):
3824 if isinstance(argdict, (list, tuple)): # for backward compatibility
3831 assert isinstance(argdict, dict)
3833 assert isinstance(keys, (list, tuple))
3834 for key_list in keys:
3835 arg_list = list(filter(
3836 lambda x: x is not None,
3837 [argdict.get(key.lower()) for key in variadic(key_list)]))
3839 return [arg for args in arg_list for arg in args]
3843 def _configuration_args(main_key, argdict, exe, keys=None, default=[], use_compat=True):
3844 main_key, exe = main_key.lower(), exe.lower()
3845 root_key = exe if main_key == exe else f'{main_key}+{exe}'
3846 keys = [f'{root_key}{k}' for k in (keys or [''])]
3847 if root_key in keys:
3849 keys.append((main_key, exe))
3850 keys.append('default')
3853 return cli_configuration_args(argdict, keys, default, use_compat)
3856 class ISO639Utils(object):
3857 # See http://www.loc.gov/standards/iso639-2/ISO-639-2_utf-8.txt
3916 'iw': 'heb', # Replaced by he in 1989 revision
3926 'in': 'ind', # Replaced by id in 1989 revision
4041 'ji': 'yid', # Replaced by yi in 1989 revision
4049 def short2long(cls, code):
4050 """Convert language code from ISO 639-1 to ISO 639-2/T"""
4051 return cls._lang_map.get(code[:2])
4054 def long2short(cls, code):
4055 """Convert language code from ISO 639-2/T to ISO 639-1"""
4056 for short_name, long_name in cls._lang_map.items():
4057 if long_name == code:
4061 class ISO3166Utils(object):
4062 # From http://data.okfn.org/data/core/country-list
4064 'AF': 'Afghanistan',
4065 'AX': 'Åland Islands',
4068 'AS': 'American Samoa',
4073 'AG': 'Antigua and Barbuda',
4090 'BO': 'Bolivia, Plurinational State of',
4091 'BQ': 'Bonaire, Sint Eustatius and Saba',
4092 'BA': 'Bosnia and Herzegovina',
4094 'BV': 'Bouvet Island',
4096 'IO': 'British Indian Ocean Territory',
4097 'BN': 'Brunei Darussalam',
4099 'BF': 'Burkina Faso',
4105 'KY': 'Cayman Islands',
4106 'CF': 'Central African Republic',
4110 'CX': 'Christmas Island',
4111 'CC': 'Cocos (Keeling) Islands',
4115 'CD': 'Congo, the Democratic Republic of the',
4116 'CK': 'Cook Islands',
4118 'CI': 'Côte d\'Ivoire',
4123 'CZ': 'Czech Republic',
4127 'DO': 'Dominican Republic',
4130 'SV': 'El Salvador',
4131 'GQ': 'Equatorial Guinea',
4135 'FK': 'Falkland Islands (Malvinas)',
4136 'FO': 'Faroe Islands',
4140 'GF': 'French Guiana',
4141 'PF': 'French Polynesia',
4142 'TF': 'French Southern Territories',
4157 'GW': 'Guinea-Bissau',
4160 'HM': 'Heard Island and McDonald Islands',
4161 'VA': 'Holy See (Vatican City State)',
4168 'IR': 'Iran, Islamic Republic of',
4171 'IM': 'Isle of Man',
4181 'KP': 'Korea, Democratic People\'s Republic of',
4182 'KR': 'Korea, Republic of',
4185 'LA': 'Lao People\'s Democratic Republic',
4191 'LI': 'Liechtenstein',
4195 'MK': 'Macedonia, the Former Yugoslav Republic of',
4202 'MH': 'Marshall Islands',
4208 'FM': 'Micronesia, Federated States of',
4209 'MD': 'Moldova, Republic of',
4220 'NL': 'Netherlands',
4221 'NC': 'New Caledonia',
4222 'NZ': 'New Zealand',
4227 'NF': 'Norfolk Island',
4228 'MP': 'Northern Mariana Islands',
4233 'PS': 'Palestine, State of',
4235 'PG': 'Papua New Guinea',
4238 'PH': 'Philippines',
4242 'PR': 'Puerto Rico',
4246 'RU': 'Russian Federation',
4248 'BL': 'Saint Barthélemy',
4249 'SH': 'Saint Helena, Ascension and Tristan da Cunha',
4250 'KN': 'Saint Kitts and Nevis',
4251 'LC': 'Saint Lucia',
4252 'MF': 'Saint Martin (French part)',
4253 'PM': 'Saint Pierre and Miquelon',
4254 'VC': 'Saint Vincent and the Grenadines',
4257 'ST': 'Sao Tome and Principe',
4258 'SA': 'Saudi Arabia',
4262 'SL': 'Sierra Leone',
4264 'SX': 'Sint Maarten (Dutch part)',
4267 'SB': 'Solomon Islands',
4269 'ZA': 'South Africa',
4270 'GS': 'South Georgia and the South Sandwich Islands',
4271 'SS': 'South Sudan',
4276 'SJ': 'Svalbard and Jan Mayen',
4279 'CH': 'Switzerland',
4280 'SY': 'Syrian Arab Republic',
4281 'TW': 'Taiwan, Province of China',
4283 'TZ': 'Tanzania, United Republic of',
4285 'TL': 'Timor-Leste',
4289 'TT': 'Trinidad and Tobago',
4292 'TM': 'Turkmenistan',
4293 'TC': 'Turks and Caicos Islands',
4297 'AE': 'United Arab Emirates',
4298 'GB': 'United Kingdom',
4299 'US': 'United States',
4300 'UM': 'United States Minor Outlying Islands',
4304 'VE': 'Venezuela, Bolivarian Republic of',
4306 'VG': 'Virgin Islands, British',
4307 'VI': 'Virgin Islands, U.S.',
4308 'WF': 'Wallis and Futuna',
4309 'EH': 'Western Sahara',
4316 def short2full(cls, code):
4317 """Convert an ISO 3166-2 country code to the corresponding full name"""
4318 return cls._country_map.get(code.upper())
4321 class GeoUtils(object):
4322 # Major IPv4 address blocks per country
4324 'AD': '46.172.224.0/19',
4325 'AE': '94.200.0.0/13',
4326 'AF': '149.54.0.0/17',
4327 'AG': '209.59.64.0/18',
4328 'AI': '204.14.248.0/21',
4329 'AL': '46.99.0.0/16',
4330 'AM': '46.70.0.0/15',
4331 'AO': '105.168.0.0/13',
4332 'AP': '182.50.184.0/21',
4333 'AQ': '23.154.160.0/24',
4334 'AR': '181.0.0.0/12',
4335 'AS': '202.70.112.0/20',
4336 'AT': '77.116.0.0/14',
4337 'AU': '1.128.0.0/11',
4338 'AW': '181.41.0.0/18',
4339 'AX': '185.217.4.0/22',
4340 'AZ': '5.197.0.0/16',
4341 'BA': '31.176.128.0/17',
4342 'BB': '65.48.128.0/17',
4343 'BD': '114.130.0.0/16',
4345 'BF': '102.178.0.0/15',
4346 'BG': '95.42.0.0/15',
4347 'BH': '37.131.0.0/17',
4348 'BI': '154.117.192.0/18',
4349 'BJ': '137.255.0.0/16',
4350 'BL': '185.212.72.0/23',
4351 'BM': '196.12.64.0/18',
4352 'BN': '156.31.0.0/16',
4353 'BO': '161.56.0.0/16',
4354 'BQ': '161.0.80.0/20',
4355 'BR': '191.128.0.0/12',
4356 'BS': '24.51.64.0/18',
4357 'BT': '119.2.96.0/19',
4358 'BW': '168.167.0.0/16',
4359 'BY': '178.120.0.0/13',
4360 'BZ': '179.42.192.0/18',
4361 'CA': '99.224.0.0/11',
4362 'CD': '41.243.0.0/16',
4363 'CF': '197.242.176.0/21',
4364 'CG': '160.113.0.0/16',
4365 'CH': '85.0.0.0/13',
4366 'CI': '102.136.0.0/14',
4367 'CK': '202.65.32.0/19',
4368 'CL': '152.172.0.0/14',
4369 'CM': '102.244.0.0/14',
4370 'CN': '36.128.0.0/10',
4371 'CO': '181.240.0.0/12',
4372 'CR': '201.192.0.0/12',
4373 'CU': '152.206.0.0/15',
4374 'CV': '165.90.96.0/19',
4375 'CW': '190.88.128.0/17',
4376 'CY': '31.153.0.0/16',
4377 'CZ': '88.100.0.0/14',
4379 'DJ': '197.241.0.0/17',
4380 'DK': '87.48.0.0/12',
4381 'DM': '192.243.48.0/20',
4382 'DO': '152.166.0.0/15',
4383 'DZ': '41.96.0.0/12',
4384 'EC': '186.68.0.0/15',
4385 'EE': '90.190.0.0/15',
4386 'EG': '156.160.0.0/11',
4387 'ER': '196.200.96.0/20',
4388 'ES': '88.0.0.0/11',
4389 'ET': '196.188.0.0/14',
4390 'EU': '2.16.0.0/13',
4391 'FI': '91.152.0.0/13',
4392 'FJ': '144.120.0.0/16',
4393 'FK': '80.73.208.0/21',
4394 'FM': '119.252.112.0/20',
4395 'FO': '88.85.32.0/19',
4397 'GA': '41.158.0.0/15',
4399 'GD': '74.122.88.0/21',
4400 'GE': '31.146.0.0/16',
4401 'GF': '161.22.64.0/18',
4402 'GG': '62.68.160.0/19',
4403 'GH': '154.160.0.0/12',
4404 'GI': '95.164.0.0/16',
4405 'GL': '88.83.0.0/19',
4406 'GM': '160.182.0.0/15',
4407 'GN': '197.149.192.0/18',
4408 'GP': '104.250.0.0/19',
4409 'GQ': '105.235.224.0/20',
4410 'GR': '94.64.0.0/13',
4411 'GT': '168.234.0.0/16',
4412 'GU': '168.123.0.0/16',
4413 'GW': '197.214.80.0/20',
4414 'GY': '181.41.64.0/18',
4415 'HK': '113.252.0.0/14',
4416 'HN': '181.210.0.0/16',
4417 'HR': '93.136.0.0/13',
4418 'HT': '148.102.128.0/17',
4419 'HU': '84.0.0.0/14',
4420 'ID': '39.192.0.0/10',
4421 'IE': '87.32.0.0/12',
4422 'IL': '79.176.0.0/13',
4423 'IM': '5.62.80.0/20',
4424 'IN': '117.192.0.0/10',
4425 'IO': '203.83.48.0/21',
4426 'IQ': '37.236.0.0/14',
4427 'IR': '2.176.0.0/12',
4428 'IS': '82.221.0.0/16',
4429 'IT': '79.0.0.0/10',
4430 'JE': '87.244.64.0/18',
4431 'JM': '72.27.0.0/17',
4432 'JO': '176.29.0.0/16',
4433 'JP': '133.0.0.0/8',
4434 'KE': '105.48.0.0/12',
4435 'KG': '158.181.128.0/17',
4436 'KH': '36.37.128.0/17',
4437 'KI': '103.25.140.0/22',
4438 'KM': '197.255.224.0/20',
4439 'KN': '198.167.192.0/19',
4440 'KP': '175.45.176.0/22',
4441 'KR': '175.192.0.0/10',
4442 'KW': '37.36.0.0/14',
4443 'KY': '64.96.0.0/15',
4444 'KZ': '2.72.0.0/13',
4445 'LA': '115.84.64.0/18',
4446 'LB': '178.135.0.0/16',
4447 'LC': '24.92.144.0/20',
4448 'LI': '82.117.0.0/19',
4449 'LK': '112.134.0.0/15',
4450 'LR': '102.183.0.0/16',
4451 'LS': '129.232.0.0/17',
4452 'LT': '78.56.0.0/13',
4453 'LU': '188.42.0.0/16',
4454 'LV': '46.109.0.0/16',
4455 'LY': '41.252.0.0/14',
4456 'MA': '105.128.0.0/11',
4457 'MC': '88.209.64.0/18',
4458 'MD': '37.246.0.0/16',
4459 'ME': '178.175.0.0/17',
4460 'MF': '74.112.232.0/21',
4461 'MG': '154.126.0.0/17',
4462 'MH': '117.103.88.0/21',
4463 'MK': '77.28.0.0/15',
4464 'ML': '154.118.128.0/18',
4465 'MM': '37.111.0.0/17',
4466 'MN': '49.0.128.0/17',
4467 'MO': '60.246.0.0/16',
4468 'MP': '202.88.64.0/20',
4469 'MQ': '109.203.224.0/19',
4470 'MR': '41.188.64.0/18',
4471 'MS': '208.90.112.0/22',
4472 'MT': '46.11.0.0/16',
4473 'MU': '105.16.0.0/12',
4474 'MV': '27.114.128.0/18',
4475 'MW': '102.70.0.0/15',
4476 'MX': '187.192.0.0/11',
4477 'MY': '175.136.0.0/13',
4478 'MZ': '197.218.0.0/15',
4479 'NA': '41.182.0.0/16',
4480 'NC': '101.101.0.0/18',
4481 'NE': '197.214.0.0/18',
4482 'NF': '203.17.240.0/22',
4483 'NG': '105.112.0.0/12',
4484 'NI': '186.76.0.0/15',
4485 'NL': '145.96.0.0/11',
4486 'NO': '84.208.0.0/13',
4487 'NP': '36.252.0.0/15',
4488 'NR': '203.98.224.0/19',
4489 'NU': '49.156.48.0/22',
4490 'NZ': '49.224.0.0/14',
4491 'OM': '5.36.0.0/15',
4492 'PA': '186.72.0.0/15',
4493 'PE': '186.160.0.0/14',
4494 'PF': '123.50.64.0/18',
4495 'PG': '124.240.192.0/19',
4496 'PH': '49.144.0.0/13',
4497 'PK': '39.32.0.0/11',
4498 'PL': '83.0.0.0/11',
4499 'PM': '70.36.0.0/20',
4500 'PR': '66.50.0.0/16',
4501 'PS': '188.161.0.0/16',
4502 'PT': '85.240.0.0/13',
4503 'PW': '202.124.224.0/20',
4504 'PY': '181.120.0.0/14',
4505 'QA': '37.210.0.0/15',
4506 'RE': '102.35.0.0/16',
4507 'RO': '79.112.0.0/13',
4508 'RS': '93.86.0.0/15',
4509 'RU': '5.136.0.0/13',
4510 'RW': '41.186.0.0/16',
4511 'SA': '188.48.0.0/13',
4512 'SB': '202.1.160.0/19',
4513 'SC': '154.192.0.0/11',
4514 'SD': '102.120.0.0/13',
4515 'SE': '78.64.0.0/12',
4516 'SG': '8.128.0.0/10',
4517 'SI': '188.196.0.0/14',
4518 'SK': '78.98.0.0/15',
4519 'SL': '102.143.0.0/17',
4520 'SM': '89.186.32.0/19',
4521 'SN': '41.82.0.0/15',
4522 'SO': '154.115.192.0/18',
4523 'SR': '186.179.128.0/17',
4524 'SS': '105.235.208.0/21',
4525 'ST': '197.159.160.0/19',
4526 'SV': '168.243.0.0/16',
4527 'SX': '190.102.0.0/20',
4529 'SZ': '41.84.224.0/19',
4530 'TC': '65.255.48.0/20',
4531 'TD': '154.68.128.0/19',
4532 'TG': '196.168.0.0/14',
4533 'TH': '171.96.0.0/13',
4534 'TJ': '85.9.128.0/18',
4535 'TK': '27.96.24.0/21',
4536 'TL': '180.189.160.0/20',
4537 'TM': '95.85.96.0/19',
4538 'TN': '197.0.0.0/11',
4539 'TO': '175.176.144.0/21',
4540 'TR': '78.160.0.0/11',
4541 'TT': '186.44.0.0/15',
4542 'TV': '202.2.96.0/19',
4543 'TW': '120.96.0.0/11',
4544 'TZ': '156.156.0.0/14',
4545 'UA': '37.52.0.0/14',
4546 'UG': '102.80.0.0/13',
4548 'UY': '167.56.0.0/13',
4549 'UZ': '84.54.64.0/18',
4550 'VA': '212.77.0.0/19',
4551 'VC': '207.191.240.0/21',
4552 'VE': '186.88.0.0/13',
4553 'VG': '66.81.192.0/20',
4554 'VI': '146.226.0.0/16',
4555 'VN': '14.160.0.0/11',
4556 'VU': '202.80.32.0/20',
4557 'WF': '117.20.32.0/21',
4558 'WS': '202.4.32.0/19',
4559 'YE': '134.35.0.0/16',
4560 'YT': '41.242.116.0/22',
4561 'ZA': '41.0.0.0/11',
4562 'ZM': '102.144.0.0/13',
4563 'ZW': '102.177.192.0/18',
4567 def random_ipv4(cls, code_or_block):
4568 if len(code_or_block) == 2:
4569 block = cls._country_ip_map.get(code_or_block.upper())
4573 block = code_or_block
4574 addr, preflen = block.split('/')
4575 addr_min = compat_struct_unpack('!L', socket.inet_aton(addr))[0]
4576 addr_max = addr_min | (0xffffffff >> int(preflen))
4577 return compat_str(socket.inet_ntoa(
4578 compat_struct_pack('!L', random.randint(addr_min, addr_max))))
4581 class PerRequestProxyHandler(compat_urllib_request.ProxyHandler):
4582 def __init__(self, proxies=None):
4583 # Set default handlers
4584 for type in ('http', 'https'):
4585 setattr(self, '%s_open' % type,
4586 lambda r, proxy='__noproxy__', type=type, meth=self.proxy_open:
4587 meth(r, proxy, type))
4588 compat_urllib_request.ProxyHandler.__init__(self, proxies)
4590 def proxy_open(self, req, proxy, type):
4591 req_proxy = req.headers.get('Ytdl-request-proxy')
4592 if req_proxy is not None:
4594 del req.headers['Ytdl-request-proxy']
4596 if proxy == '__noproxy__':
4597 return None # No Proxy
4598 if compat_urlparse.urlparse(proxy).scheme.lower() in ('socks', 'socks4', 'socks4a', 'socks5'):
4599 req.add_header('Ytdl-socks-proxy', proxy)
4600 # yt-dlp's http/https handlers do wrapping the socket with socks
4602 return compat_urllib_request.ProxyHandler.proxy_open(
4603 self, req, proxy, type)
4606 # Both long_to_bytes and bytes_to_long are adapted from PyCrypto, which is
4607 # released into Public Domain
4608 # https://github.com/dlitz/pycrypto/blob/master/lib/Crypto/Util/number.py#L387
4610 def long_to_bytes(n, blocksize=0):
4611 """long_to_bytes(n:long, blocksize:int) : string
4612 Convert a long integer to a byte string.
4614 If optional blocksize is given and greater than zero, pad the front of the
4615 byte string with binary zeros so that the length is a multiple of
4618 # after much testing, this algorithm was deemed to be the fastest
4622 s = compat_struct_pack('>I', n & 0xffffffff) + s
4624 # strip off leading zeros
4625 for i in range(len(s)):
4626 if s[i] != b'\000'[0]:
4629 # only happens when n == 0
4633 # add back some pad bytes. this could be done more efficiently w.r.t. the
4634 # de-padding being done above, but sigh...
4635 if blocksize > 0 and len(s) % blocksize:
4636 s = (blocksize - len(s) % blocksize) * b'\000' + s
4640 def bytes_to_long(s):
4641 """bytes_to_long(string) : long
4642 Convert a byte string to a long integer.
4644 This is (essentially) the inverse of long_to_bytes().
4649 extra = (4 - length % 4)
4650 s = b'\000' * extra + s
4651 length = length + extra
4652 for i in range(0, length, 4):
4653 acc = (acc << 32) + compat_struct_unpack('>I', s[i:i + 4])[0]
4657 def ohdave_rsa_encrypt(data, exponent, modulus):
4659 Implement OHDave
's RSA algorithm. See http://www.ohdave.com/rsa/
4662 data: data to encrypt, bytes-like object
4663 exponent, modulus: parameter e and N of RSA algorithm, both integer
4664 Output: hex string of encrypted data
4666 Limitation: supports one block encryption only
4669 payload = int(binascii.hexlify(data[::-1]), 16)
4670 encrypted = pow(payload, exponent, modulus)
4671 return '%x' % encrypted
4674 def pkcs1pad(data, length):
4676 Padding input data with PKCS#1 scheme
4678 @param {int[]} data input data
4679 @param {int} length target length
4680 @returns {int[]} padded data
4682 if len(data) > length - 11:
4683 raise ValueError('Input data too
long for PKCS
#1 padding')
4685 pseudo_random
= [random
.randint(0, 254) for _
in range(length
- len(data
) - 3)]
4686 return [0, 2] + pseudo_random
+ [0] + data
4689 def encode_base_n(num
, n
, table
=None):
4690 FULL_TABLE
= '0123456789abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ'
4692 table
= FULL_TABLE
[:n
]
4695 raise ValueError('base %d exceeds table length %d' % (n
, len(table
)))
4702 ret
= table
[num
% n
] + ret
4707 def decode_packed_codes(code
):
4708 mobj
= re
.search(PACKED_CODES_RE
, code
)
4709 obfuscated_code
, base
, count
, symbols
= mobj
.groups()
4712 symbols
= symbols
.split('|')
4717 base_n_count
= encode_base_n(count
, base
)
4718 symbol_table
[base_n_count
] = symbols
[count
] or base_n_count
4721 r
'\b(\w+)\b', lambda mobj
: symbol_table
[mobj
.group(0)],
4725 def caesar(s
, alphabet
, shift
):
4730 alphabet
[(alphabet
.index(c
) + shift
) % l
] if c
in alphabet
else c
4735 return caesar(s
, r
'''!"#$%&'()*+,-./0123456789:;<=>?@ABCDEFGHIJKLMNOPQRSTUVWXYZ[\]^_`abcdefghijklmnopqrstuvwxyz{|}~''', 47)
4738 def parse_m3u8_attributes(attrib
):
4740 for (key
, val
) in re
.findall(r
'(?P<key>[A-Z0-9-]+)=(?P<val>"[^"]+"|[^",]+)(?:,|$)', attrib
):
4741 if val
.startswith('"'):
4747 def urshift(val
, n
):
4748 return val
>> n
if val
>= 0 else (val
+ 0x100000000) >> n
4751 # Based on png2str() written by @gdkchan and improved by @yokrysty
4752 # Originally posted at https://github.com/ytdl-org/youtube-dl/issues/9706
4753 def decode_png(png_data
):
4754 # Reference: https://www.w3.org/TR/PNG/
4755 header
= png_data
[8:]
4757 if png_data
[:8] != b
'\x89PNG\x0d\x0a\x1a\x0a' or header
[4:8] != b
'IHDR':
4758 raise IOError('Not a valid PNG file.')
4760 int_map
= {1: '>B', 2: '>H', 4: '>I'}
4761 unpack_integer
= lambda x
: compat_struct_unpack(int_map
[len(x
)], x
)[0]
4766 length
= unpack_integer(header
[:4])
4769 chunk_type
= header
[:4]
4772 chunk_data
= header
[:length
]
4773 header
= header
[length
:]
4775 header
= header
[4:] # Skip CRC
4783 ihdr
= chunks
[0]['data']
4785 width
= unpack_integer(ihdr
[:4])
4786 height
= unpack_integer(ihdr
[4:8])
4790 for chunk
in chunks
:
4791 if chunk
['type'] == b
'IDAT':
4792 idat
+= chunk
['data']
4795 raise IOError('Unable to read PNG data.')
4797 decompressed_data
= bytearray(zlib
.decompress(idat
))
4802 def _get_pixel(idx
):
4807 for y
in range(height
):
4808 basePos
= y
* (1 + stride
)
4809 filter_type
= decompressed_data
[basePos
]
4813 pixels
.append(current_row
)
4815 for x
in range(stride
):
4816 color
= decompressed_data
[1 + basePos
+ x
]
4817 basex
= y
* stride
+ x
4822 left
= _get_pixel(basex
- 3)
4824 up
= _get_pixel(basex
- stride
)
4826 if filter_type
== 1: # Sub
4827 color
= (color
+ left
) & 0xff
4828 elif filter_type
== 2: # Up
4829 color
= (color
+ up
) & 0xff
4830 elif filter_type
== 3: # Average
4831 color
= (color
+ ((left
+ up
) >> 1)) & 0xff
4832 elif filter_type
== 4: # Paeth
4838 c
= _get_pixel(basex
- stride
- 3)
4846 if pa
<= pb
and pa
<= pc
:
4847 color
= (color
+ a
) & 0xff
4849 color
= (color
+ b
) & 0xff
4851 color
= (color
+ c
) & 0xff
4853 current_row
.append(color
)
4855 return width
, height
, pixels
4858 def write_xattr(path
, key
, value
):
4859 # This mess below finds the best xattr tool for the job
4861 # try the pyxattr module...
4864 if hasattr(xattr
, 'set'): # pyxattr
4865 # Unicode arguments are not supported in python-pyxattr until
4867 # See https://github.com/ytdl-org/youtube-dl/issues/5498
4868 pyxattr_required_version
= '0.5.0'
4869 if version_tuple(xattr
.__version
__) < version_tuple(pyxattr_required_version
):
4870 # TODO: fallback to CLI tools
4871 raise XAttrUnavailableError(
4872 'python-pyxattr is detected but is too old. '
4873 'yt-dlp requires %s or above while your version is %s. '
4874 'Falling back to other xattr implementations' % (
4875 pyxattr_required_version
, xattr
.__version
__))
4877 setxattr
= xattr
.set
4879 setxattr
= xattr
.setxattr
4882 setxattr(path
, key
, value
)
4883 except EnvironmentError as e
:
4884 raise XAttrMetadataError(e
.errno
, e
.strerror
)
4887 if compat_os_name
== 'nt':
4888 # Write xattrs to NTFS Alternate Data Streams:
4889 # http://en.wikipedia.org/wiki/NTFS#Alternate_data_streams_.28ADS.29
4890 assert ':' not in key
4891 assert os
.path
.exists(path
)
4893 ads_fn
= path
+ ':' + key
4895 with open(ads_fn
, 'wb') as f
:
4897 except EnvironmentError as e
:
4898 raise XAttrMetadataError(e
.errno
, e
.strerror
)
4900 user_has_setfattr
= check_executable('setfattr', ['--version'])
4901 user_has_xattr
= check_executable('xattr', ['-h'])
4903 if user_has_setfattr
or user_has_xattr
:
4905 value
= value
.decode('utf-8')
4906 if user_has_setfattr
:
4907 executable
= 'setfattr'
4908 opts
= ['-n', key
, '-v', value
]
4909 elif user_has_xattr
:
4910 executable
= 'xattr'
4911 opts
= ['-w', key
, value
]
4913 cmd
= ([encodeFilename(executable
, True)]
4914 + [encodeArgument(o
) for o
in opts
]
4915 + [encodeFilename(path
, True)])
4919 cmd
, stdout
=subprocess
.PIPE
, stderr
=subprocess
.PIPE
, stdin
=subprocess
.PIPE
)
4920 except EnvironmentError as e
:
4921 raise XAttrMetadataError(e
.errno
, e
.strerror
)
4922 stdout
, stderr
= p
.communicate_or_kill()
4923 stderr
= stderr
.decode('utf-8', 'replace')
4924 if p
.returncode
!= 0:
4925 raise XAttrMetadataError(p
.returncode
, stderr
)
4928 # On Unix, and can't find pyxattr, setfattr, or xattr.
4929 if sys
.platform
.startswith('linux'):
4930 raise XAttrUnavailableError(
4931 "Couldn't find a tool to set the xattrs. "
4932 "Install either the python 'pyxattr' or 'xattr' "
4933 "modules, or the GNU 'attr' package "
4934 "(which contains the 'setfattr' tool).")
4936 raise XAttrUnavailableError(
4937 "Couldn't find a tool to set the xattrs. "
4938 "Install either the python 'xattr' module, "
4939 "or the 'xattr' binary.")
4942 def random_birthday(year_field
, month_field
, day_field
):
4943 start_date
= datetime
.date(1950, 1, 1)
4944 end_date
= datetime
.date(1995, 12, 31)
4945 offset
= random
.randint(0, (end_date
- start_date
).days
)
4946 random_date
= start_date
+ datetime
.timedelta(offset
)
4948 year_field
: str(random_date
.year
),
4949 month_field
: str(random_date
.month
),
4950 day_field
: str(random_date
.day
),
4954 # Templates for internet shortcut files, which are plain text files.
4955 DOT_URL_LINK_TEMPLATE
= '''
4960 DOT_WEBLOC_LINK_TEMPLATE
= '''
4961 <?xml version="1.0" encoding="UTF-8"?>
4962 <!DOCTYPE plist PUBLIC "-//Apple//DTD PLIST 1.0//EN" "http://www.apple.com/DTDs/PropertyList-1.0.dtd">
4963 <plist version="1.0">
4966 \t<string>%(url)s</string>
4971 DOT_DESKTOP_LINK_TEMPLATE
= '''
4981 'url': DOT_URL_LINK_TEMPLATE
,
4982 'desktop': DOT_DESKTOP_LINK_TEMPLATE
,
4983 'webloc': DOT_WEBLOC_LINK_TEMPLATE
,
4987 def iri_to_uri(iri
):
4989 Converts an IRI (Internationalized Resource Identifier, allowing Unicode characters) to a URI (Uniform Resource Identifier, ASCII-only).
4991 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.
4994 iri_parts
= compat_urllib_parse_urlparse(iri
)
4996 if '[' in iri_parts
.netloc
:
4997 raise ValueError('IPv6 URIs are not, yet, supported.')
4998 # Querying `.netloc`, when there's only one bracket, also raises a ValueError.
5000 # 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.
5003 if iri_parts
.username
:
5004 net_location
+= compat_urllib_parse_quote(iri_parts
.username
, safe
=r
"!$%&'()*+,~")
5005 if iri_parts
.password
is not None:
5006 net_location
+= ':' + compat_urllib_parse_quote(iri_parts
.password
, safe
=r
"!$%&'()*+,~")
5009 net_location
+= iri_parts
.hostname
.encode('idna').decode('utf-8') # Punycode for Unicode hostnames.
5010 # The 'idna' encoding produces ASCII text.
5011 if iri_parts
.port
is not None and iri_parts
.port
!= 80:
5012 net_location
+= ':' + str(iri_parts
.port
)
5014 return compat_urllib_parse_urlunparse(
5018 compat_urllib_parse_quote_plus(iri_parts
.path
, safe
=r
"!$%&'()*+,/:;=@|~"),
5020 # Unsure about the `safe` argument, since this is a legacy way of handling parameters.
5021 compat_urllib_parse_quote_plus(iri_parts
.params
, safe
=r
"!$%&'()*+,/:;=@|~"),
5023 # Not totally sure about the `safe` argument, since the source does not explicitly mention the query URI component.
5024 compat_urllib_parse_quote_plus(iri_parts
.query
, safe
=r
"!$%&'()*+,/:;=?@{|}~"),
5026 compat_urllib_parse_quote_plus(iri_parts
.fragment
, safe
=r
"!#$%&'()*+,/:;=?@{|}~")))
5028 # Source for `safe` arguments: https://url.spec.whatwg.org/#percent-encoded-bytes.
5031 def to_high_limit_path(path
):
5032 if sys
.platform
in ['win32', 'cygwin']:
5033 # Work around MAX_PATH limitation on Windows. The maximum allowed length for the individual path segments may still be quite limited.
5034 return r
'\\?\ '.rstrip() + os
.path
.abspath(path
)
5039 def format_field(obj
, field
=None, template
='%s', ignore
=(None, ''), default
='', func
=None):
5040 val
= traverse_obj(obj
, *variadic(field
))
5043 return template
% (func(val
) if func
else val
)
5046 def clean_podcast_url(url
):
5047 return re
.sub(r
'''(?x)
5051 media\.blubrry\.com| # https://create.blubrry.com/resources/podcast-media-download-statistics/getting-started/
5054 (?:dts|www)\.podtrac\.com/(?:pts/)?redirect\.[0-9a-z]{3,4}| # http://analytics.podtrac.com/how-to-measure
5057 cn\.co| # https://podcorn.com/analytics-prefix/
5058 st\.fm # https://podsights.com/docs/
5063 _HEX_TABLE
= '0123456789abcdef'
5066 def random_uuidv4():
5067 return re
.sub(r
'[xy]', lambda x
: _HEX_TABLE
[random
.randint(0, 15)], 'xxxxxxxx-xxxx-4xxx-yxxx-xxxxxxxxxxxx')
5070 def make_dir(path
, to_screen
=None):
5072 dn
= os
.path
.dirname(path
)
5073 if dn
and not os
.path
.exists(dn
):
5076 except (OSError, IOError) as err
:
5077 if callable(to_screen
) is not None:
5078 to_screen('unable to create directory ' + error_to_compat_str(err
))
5082 def get_executable_path():
5083 from zipimport
import zipimporter
5084 if hasattr(sys
, 'frozen'): # Running from PyInstaller
5085 path
= os
.path
.dirname(sys
.executable
)
5086 elif isinstance(globals().get('__loader__'), zipimporter
): # Running from ZIP
5087 path
= os
.path
.join(os
.path
.dirname(__file__
), '../..')
5089 path
= os
.path
.join(os
.path
.dirname(__file__
), '..')
5090 return os
.path
.abspath(path
)
5093 def load_plugins(name
, suffix
, namespace
):
5096 plugins_spec
= importlib
.util
.spec_from_file_location(
5097 name
, os
.path
.join(get_executable_path(), 'ytdlp_plugins', name
, '__init__.py'))
5098 plugins
= importlib
.util
.module_from_spec(plugins_spec
)
5099 sys
.modules
[plugins_spec
.name
] = plugins
5100 plugins_spec
.loader
.exec_module(plugins
)
5101 for name
in dir(plugins
):
5102 if name
in namespace
:
5104 if not name
.endswith(suffix
):
5106 klass
= getattr(plugins
, name
)
5107 classes
[name
] = namespace
[name
] = klass
5108 except FileNotFoundError
:
5114 obj
, *path_list
, default
=None, expected_type
=None, get_all
=True,
5115 casesense
=True, is_user_input
=False, traverse_string
=False):
5116 ''' Traverse nested list/dict/tuple
5117 @param path_list A list of paths which are checked one by one.
5118 Each path is a list of keys where each key is a string,
5119 a function, a tuple of strings/None or "...".
5120 When a fuction is given, it takes the key as argument and
5121 returns whether the key matches or not. When a tuple is given,
5122 all the keys given in the tuple are traversed, and
5123 "..." traverses all the keys in the object
5124 "None" returns the object without traversal
5125 @param default Default value to return
5126 @param expected_type Only accept final value of this type (Can also be any callable)
5127 @param get_all Return all the values obtained from a path or only the first one
5128 @param casesense Whether to consider dictionary keys as case sensitive
5129 @param is_user_input Whether the keys are generated from user input. If True,
5130 strings are converted to int/slice if necessary
5131 @param traverse_string Whether to traverse inside strings. If True, any
5132 non-compatible object will also be converted into a string
5136 _lower
= lambda k
: (k
.lower() if isinstance(k
, str) else k
)
5137 path_list
= (map(_lower
, variadic(path
)) for path
in path_list
)
5139 def _traverse_obj(obj
, path
, _current_depth
=0):
5141 path
= tuple(variadic(path
))
5142 for i
, key
in enumerate(path
):
5143 if None in (key
, obj
):
5145 if isinstance(key
, (list, tuple)):
5146 obj
= [_traverse_obj(obj
, sub_key
, _current_depth
) for sub_key
in key
]
5149 obj
= (obj
.values() if isinstance(obj
, dict)
5150 else obj
if isinstance(obj
, (list, tuple, LazyList
))
5151 else str(obj
) if traverse_string
else [])
5153 depth
= max(depth
, _current_depth
)
5154 return [_traverse_obj(inner_obj
, path
[i
+ 1:], _current_depth
) for inner_obj
in obj
]
5156 if isinstance(obj
, (list, tuple, LazyList
)):
5157 obj
= enumerate(obj
)
5158 elif isinstance(obj
, dict):
5161 if not traverse_string
:
5165 depth
= max(depth
, _current_depth
)
5166 return [_traverse_obj(v
, path
[i
+ 1:], _current_depth
) for k
, v
in obj
if key(k
)]
5167 elif isinstance(obj
, dict) and not (is_user_input
and key
== ':'):
5168 obj
= (obj
.get(key
) if casesense
or (key
in obj
)
5169 else next((v
for k
, v
in obj
.items() if _lower(k
) == key
), None))
5172 key
= (int_or_none(key
) if ':' not in key
5173 else slice(*map(int_or_none
, key
.split(':'))))
5174 if key
== slice(None):
5175 return _traverse_obj(obj
, (..., *path
[i
+ 1:]), _current_depth
)
5176 if not isinstance(key
, (int, slice)):
5178 if not isinstance(obj
, (list, tuple, LazyList
)):
5179 if not traverse_string
:
5188 if isinstance(expected_type
, type):
5189 type_test
= lambda val
: val
if isinstance(val
, expected_type
) else None
5190 elif expected_type
is not None:
5191 type_test
= expected_type
5193 type_test
= lambda val
: val
5195 for path
in path_list
:
5197 val
= _traverse_obj(obj
, path
)
5200 for _
in range(depth
- 1):
5201 val
= itertools
.chain
.from_iterable(v
for v
in val
if v
is not None)
5202 val
= [v
for v
in map(type_test
, val
) if v
is not None]
5204 return val
if get_all
else val
[0]
5206 val
= type_test(val
)
5212 def traverse_dict(dictn
, keys
, casesense
=True):
5213 write_string('DeprecationWarning: yt_dlp.utils.traverse_dict is deprecated '
5214 'and may be removed in a future version. Use yt_dlp.utils.traverse_obj instead')
5215 return traverse_obj(dictn
, keys
, casesense
=casesense
, is_user_input
=True, traverse_string
=True)
5218 def variadic(x
, allowed_types
=(str, bytes, dict)):
5219 return x
if isinstance(x
, collections
.abc
.Iterable
) and not isinstance(x
, allowed_types
) else (x
,)
5222 def decode_base(value
, digits
):
5223 # This will convert given base-x string to scalar (long or int)
5224 table
= {char: index for index, char in enumerate(digits)}
5229 result
+= table
[chr]
5233 def time_seconds(**kwargs
):
5234 t
= datetime
.datetime
.now(datetime
.timezone(datetime
.timedelta(**kwargs
)))
5235 return t
.timestamp()
5238 # create a JSON Web Signature (jws) with HS256 algorithm
5239 # the resulting format is in JWS Compact Serialization
5240 # implemented following JWT https://www.rfc-editor.org/rfc/rfc7519.html
5241 # implemented following JWS https://www.rfc-editor.org/rfc/rfc7515.html
5242 def jwt_encode_hs256(payload_data
, key
, headers
={}):
5248 header_data
.update(headers
)
5249 header_b64
= base64
.b64encode(json
.dumps(header_data
).encode('utf-8'))
5250 payload_b64
= base64
.b64encode(json
.dumps(payload_data
).encode('utf-8'))
5251 h
= hmac
.new(key
.encode('utf-8'), header_b64
+ b
'.' + payload_b64
, hashlib
.sha256
)
5252 signature_b64
= base64
.b64encode(h
.digest())
5253 token
= header_b64
+ b
'.' + payload_b64
+ b
'.' + signature_b64
5257 # can be extended in future to verify the signature and parse header and return the algorithm used if it's not HS256
5258 def jwt_decode_hs256(jwt
):
5259 header_b64
, payload_b64
, signature_b64
= jwt
.split('.')
5260 payload_data
= json
.loads(base64
.urlsafe_b64decode(payload_b64
))
5264 def supports_terminal_sequences(stream
):
5265 if compat_os_name
== 'nt':
5266 from .compat
import WINDOWS_VT_MODE
# Must be imported locally
5267 if not WINDOWS_VT_MODE
or get_windows_version() < (10, 0, 10586):
5269 elif not os
.getenv('TERM'):
5272 return stream
.isatty()
5273 except BaseException
:
5277 _terminal_sequences_re
= re
.compile('\033\\[[^m]+m')
5280 def remove_terminal_sequences(string
):
5281 return _terminal_sequences_re
.sub('', string
)
5284 def number_of_digits(number
):
5285 return len('%d' % number
)
5288 def join_nonempty(*values
, delim
='-', from_dict
=None):
5289 if from_dict
is not None:
5290 values
= map(from_dict
.get
, values
)
5291 return delim
.join(map(str, filter(None, values
)))
5294 def scale_thumbnails_to_max_format_width(formats
, thumbnails
, url_width_re
):
5296 Find the largest format dimensions in terms of video width and, for each thumbnail:
5297 * Modify the URL: Match the width with the provided regex and replace with the former width
5300 This function is useful with video services that scale the provided thumbnails on demand
5302 _keys
= ('width', 'height')
5303 max_dimensions
= max(
5304 [tuple(format
.get(k
) or 0 for k
in _keys
) for format
in formats
],
5306 if not max_dimensions
[0]:
5310 {'url': re.sub(url_width_re, str(max_dimensions[0]), thumbnail['url'])}
,
5311 dict(zip(_keys
, max_dimensions
)), thumbnail
)
5312 for thumbnail
in thumbnails
5316 def parse_http_range(range):
5317 """ Parse value of "Range" or "Content-Range" HTTP header into tuple. """
5319 return None, None, None
5320 crg
= re
.search(r
'bytes[ =](\d+)-(\d+)?(?:/(\d+))?', range)
5322 return None, None, None
5323 return int(crg
.group(1)), int_or_none(crg
.group(2)), int_or_none(crg
.group(3))
5329 __initialized
= False
5331 def __init__(self
, parser
, label
=None):
5332 self
._parser
, self
.label
= parser
, label
5333 self
._loaded
_paths
, self
.configs
= set(), []
5335 def init(self
, args
=None, filename
=None):
5336 assert not self
.__initialized
5339 location
= os
.path
.realpath(filename
)
5340 directory
= os
.path
.dirname(location
)
5341 if location
in self
._loaded
_paths
:
5343 self
._loaded
_paths
.add(location
)
5345 self
.__initialized
= True
5346 self
.own_args
, self
.filename
= args
, filename
5347 for location
in self
._parser
.parse_args(args
)[0].config_locations
or []:
5348 location
= os
.path
.join(directory
, expand_path(location
))
5349 if os
.path
.isdir(location
):
5350 location
= os
.path
.join(location
, 'yt-dlp.conf')
5351 if not os
.path
.exists(location
):
5352 self
._parser
.error(f
'config location {location} does not exist')
5353 self
.append_config(self
.read_file(location
), location
)
5357 label
= join_nonempty(
5358 self
.label
, 'config', f
'"{self.filename}"' if self
.filename
else '',
5360 return join_nonempty(
5361 self
.own_args
is not None and f
'{label[0].upper()}{label[1:]}: {self.hide_login_info(self.own_args)}',
5362 *(f
'\n{c}'.replace('\n', '\n| ')[1:] for c
in self
.configs
),
5366 def read_file(filename
, default
=[]):
5368 optionf
= open(filename
)
5370 return default
# silently skip if file is not present
5372 # FIXME: https://github.com/ytdl-org/youtube-dl/commit/dfe5fa49aed02cf36ba9f743b11b0903554b5e56
5373 contents
= optionf
.read()
5374 if sys
.version_info
< (3,):
5375 contents
= contents
.decode(preferredencoding())
5376 res
= compat_shlex_split(contents
, comments
=True)
5382 def hide_login_info(opts
):
5383 PRIVATE_OPTS
= set(['-p', '--password', '-u', '--username', '--video-password', '--ap-password', '--ap-username'])
5384 eqre
= re
.compile('^(?P<key>' + ('|'.join(re
.escape(po
) for po
in PRIVATE_OPTS
)) + ')=.+$')
5389 return m
.group('key') + '=PRIVATE'
5393 opts
= list(map(_scrub_eq
, opts
))
5394 for idx
, opt
in enumerate(opts
):
5395 if opt
in PRIVATE_OPTS
and idx
+ 1 < len(opts
):
5396 opts
[idx
+ 1] = 'PRIVATE'
5399 def append_config(self
, *args
, label
=None):
5400 config
= type(self
)(self
._parser
, label
)
5401 config
._loaded
_paths
= self
._loaded
_paths
5402 if config
.init(*args
):
5403 self
.configs
.append(config
)
5407 for config
in reversed(self
.configs
):
5408 yield from config
.all_args
5409 yield from self
.own_args
or []
5411 def parse_args(self
):
5412 return self
._parser
.parse_args(list(self
.all_args
))
5415 class WebSocketsWrapper():
5416 """Wraps websockets module to use in non-async scopes"""
5418 def __init__(self
, url
, headers
=None):
5419 self
.loop
= asyncio
.events
.new_event_loop()
5420 self
.conn
= compat_websockets
.connect(
5421 url
, extra_headers
=headers
, ping_interval
=None,
5422 close_timeout
=float('inf'), loop
=self
.loop
, ping_timeout
=float('inf'))
5423 atexit
.register(self
.__exit
__, None, None, None)
5425 def __enter__(self
):
5426 self
.pool
= self
.run_with_loop(self
.conn
.__aenter
__(), self
.loop
)
5429 def send(self
, *args
):
5430 self
.run_with_loop(self
.pool
.send(*args
), self
.loop
)
5432 def recv(self
, *args
):
5433 return self
.run_with_loop(self
.pool
.recv(*args
), self
.loop
)
5435 def __exit__(self
, type, value
, traceback
):
5437 return self
.run_with_loop(self
.conn
.__aexit
__(type, value
, traceback
), self
.loop
)
5440 self
._cancel
_all
_tasks
(self
.loop
)
5442 # taken from https://github.com/python/cpython/blob/3.9/Lib/asyncio/runners.py with modifications
5443 # for contributors: If there's any new library using asyncio needs to be run in non-async, move these function out of this class
5445 def run_with_loop(main
, loop
):
5446 if not asyncio
.coroutines
.iscoroutine(main
):
5447 raise ValueError(f
'a coroutine was expected, got {main!r}')
5450 return loop
.run_until_complete(main
)
5452 loop
.run_until_complete(loop
.shutdown_asyncgens())
5453 if hasattr(loop
, 'shutdown_default_executor'):
5454 loop
.run_until_complete(loop
.shutdown_default_executor())
5457 def _cancel_all_tasks(loop
):
5458 to_cancel
= asyncio
.tasks
.all_tasks(loop
)
5463 for task
in to_cancel
:
5466 loop
.run_until_complete(
5467 asyncio
.tasks
.gather(*to_cancel
, loop
=loop
, return_exceptions
=True))
5469 for task
in to_cancel
:
5470 if task
.cancelled():
5472 if task
.exception() is not None:
5473 loop
.call_exception_handler({
5474 'message': 'unhandled exception during asyncio.run() shutdown',
5475 'exception': task
.exception(),
5480 has_websockets
= bool(compat_websockets
)
5483 def merge_headers(*dicts
):
5484 """Merge dicts of network headers case insensitively, prioritizing the latter ones"""
5485 return {k.capitalize(): v for k, v in itertools.chain.from_iterable(map(dict.items, dicts))}