4 from __future__
import unicode_literals
39 import xml
.etree
.ElementTree
44 compat_HTMLParseError
,
50 compat_ctypes_WINFUNCTYPE
,
51 compat_etree_fromstring
,
54 compat_html_entities_html5
,
68 compat_urllib_parse_urlencode
,
69 compat_urllib_parse_urlparse
,
70 compat_urllib_parse_urlunparse
,
71 compat_urllib_parse_quote
,
72 compat_urllib_parse_quote_plus
,
73 compat_urllib_parse_unquote_plus
,
74 compat_urllib_request
,
85 def register_socks_protocols():
86 # "Register" SOCKS protocols
87 # In Python < 2.6.5, urlsplit() suffers from bug https://bugs.python.org/issue7904
88 # URLs with protocols not in urlparse.uses_netloc are not handled correctly
89 for scheme
in ('socks', 'socks4', 'socks4a', 'socks5'):
90 if scheme
not in compat_urlparse
.uses_netloc
:
91 compat_urlparse
.uses_netloc
.append(scheme
)
94 # This is not clearly defined otherwise
95 compiled_regex_type
= type(re
.compile(''))
98 def random_user_agent():
99 _USER_AGENT_TPL
= 'Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/%s Safari/537.36'
140 return _USER_AGENT_TPL
% random
.choice(_CHROME_VERSIONS
)
144 'User-Agent': random_user_agent(),
145 'Accept': 'text/html,application/xhtml+xml,application/xml;q=0.9,*/*;q=0.8',
146 'Accept-Encoding': 'gzip, deflate',
147 'Accept-Language': 'en-us,en;q=0.5',
152 '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',
156 NO_DEFAULT
= object()
158 ENGLISH_MONTH_NAMES
= [
159 'January', 'February', 'March', 'April', 'May', 'June',
160 'July', 'August', 'September', 'October', 'November', 'December']
163 'en': ENGLISH_MONTH_NAMES
,
165 'janvier', 'février', 'mars', 'avril', 'mai', 'juin',
166 'juillet', 'août', 'septembre', 'octobre', 'novembre', 'décembre'],
170 'mp4', 'm4a', 'm4p', 'm4b', 'm4r', 'm4v', 'aac',
171 'flv', 'f4v', 'f4a', 'f4b',
172 'webm', 'ogg', 'ogv', 'oga', 'ogx', 'spx', 'opus',
173 'mkv', 'mka', 'mk3d',
182 'f4f', 'f4m', 'm3u8', 'smil')
184 # needed for sanitizing filenames in restricted mode
185 ACCENT_CHARS
= dict(zip('ÂÃÄÀÁÅÆÇÈÉÊËÌÍÎÏÐÑÒÓÔÕÖŐØŒÙÚÛÜŰÝÞßàáâãäåæçèéêëìíîïðñòóôõöőøœùúûüűýþÿ',
186 itertools
.chain('AAAAAA', ['AE'], 'CEEEEIIIIDNOOOOOOO', ['OE'], 'UUUUUY', ['TH', 'ss'],
187 'aaaaaa', ['ae'], 'ceeeeiiiionooooooo', ['oe'], 'uuuuuy', ['th'], 'y')))
217 '%Y-%m-%d %H:%M:%S.%f',
218 '%Y-%m-%d %H:%M:%S:%f',
221 '%Y-%m-%dT%H:%M:%SZ',
222 '%Y-%m-%dT%H:%M:%S.%fZ',
223 '%Y-%m-%dT%H:%M:%S.%f0Z',
225 '%Y-%m-%dT%H:%M:%S.%f',
228 '%b %d %Y at %H:%M:%S',
230 '%B %d %Y at %H:%M:%S',
234 DATE_FORMATS_DAY_FIRST
= list(DATE_FORMATS
)
235 DATE_FORMATS_DAY_FIRST
.extend([
244 DATE_FORMATS_MONTH_FIRST
= list(DATE_FORMATS
)
245 DATE_FORMATS_MONTH_FIRST
.extend([
253 PACKED_CODES_RE
= r
"}\('(.+)',(\d+),(\d+),'([^']+)'\.split\('\|'\)"
254 JSON_LD_RE
= r
'(?is)<script[^>]+type=(["\']?
)application
/ld\
+json\
1[^
>]*>(?P
<json_ld
>.+?
)</script
>'
257 def preferredencoding():
258 """Get preferred encoding.
260 Returns the best encoding scheme for the system, based on
261 locale.getpreferredencoding() and some further tweaks.
264 pref = locale.getpreferredencoding()
272 def write_json_file(obj, fn):
273 """ Encode obj as JSON and write it to fn, atomically if possible """
275 fn = encodeFilename(fn)
276 if sys.version_info < (3, 0) and sys.platform != 'win32
':
277 encoding = get_filesystem_encoding()
278 # os.path.basename returns a bytes object, but NamedTemporaryFile
279 # will fail if the filename contains non ascii characters unless we
280 # use a unicode object
281 path_basename = lambda f: os.path.basename(fn).decode(encoding)
282 # the same for os.path.dirname
283 path_dirname = lambda f: os.path.dirname(fn).decode(encoding)
285 path_basename = os.path.basename
286 path_dirname = os.path.dirname
290 'prefix
': path_basename(fn) + '.',
291 'dir': path_dirname(fn),
295 # In Python 2.x, json.dump expects a bytestream.
296 # In Python 3.x, it writes to a character stream
297 if sys.version_info < (3, 0):
305 tf = tempfile.NamedTemporaryFile(**compat_kwargs(args))
309 json.dump(obj, tf, ensure_ascii=False)
310 if sys.platform == 'win32
':
311 # Need to remove existing file on Windows, else os.rename raises
312 # WindowsError or FileExistsError.
320 os.chmod(tf.name, 0o666 & ~mask)
323 os.rename(tf.name, fn)
332 if sys.version_info >= (2, 7):
333 def find_xpath_attr(node, xpath, key, val=None):
334 """ Find the xpath xpath[@key=val] """
335 assert re.match(r'^
[a
-zA
-Z_
-]+$
', key)
336 expr = xpath + ('[@%s]' % key if val is None else "[@%s='%s']" % (key, val))
337 return node.find(expr)
339 def find_xpath_attr(node, xpath, key, val=None):
340 for f in node.findall(compat_xpath(xpath)):
341 if key not in f.attrib:
343 if val is None or f.attrib.get(key) == val:
347 # On python2.6 the xml.etree.ElementTree.Element methods don't support
348 # the namespace parameter
351 def xpath_with_ns(path
, ns_map
):
352 components
= [c
.split(':') for c
in path
.split('/')]
356 replaced
.append(c
[0])
359 replaced
.append('{%s}%s' % (ns_map
[ns
], tag
))
360 return '/'.join(replaced
)
363 def xpath_element(node
, xpath
, name
=None, fatal
=False, default
=NO_DEFAULT
):
364 def _find_xpath(xpath
):
365 return node
.find(compat_xpath(xpath
))
367 if isinstance(xpath
, (str, compat_str
)):
368 n
= _find_xpath(xpath
)
376 if default
is not NO_DEFAULT
:
379 name
= xpath
if name
is None else name
380 raise ExtractorError('Could not find XML element %s' % name
)
386 def xpath_text(node
, xpath
, name
=None, fatal
=False, default
=NO_DEFAULT
):
387 n
= xpath_element(node
, xpath
, name
, fatal
=fatal
, default
=default
)
388 if n
is None or n
== default
:
391 if default
is not NO_DEFAULT
:
394 name
= xpath
if name
is None else name
395 raise ExtractorError('Could not find XML element\'s text %s' % name
)
401 def xpath_attr(node
, xpath
, key
, name
=None, fatal
=False, default
=NO_DEFAULT
):
402 n
= find_xpath_attr(node
, xpath
, key
)
404 if default
is not NO_DEFAULT
:
407 name
= '%s[@%s]' % (xpath
, key
) if name
is None else name
408 raise ExtractorError('Could not find XML attribute %s' % name
)
414 def get_element_by_id(id, html
):
415 """Return the content of the tag with the specified ID in the passed HTML document"""
416 return get_element_by_attribute('id', id, html
)
419 def get_element_html_by_id(id, html
):
420 """Return the html of the tag with the specified ID in the passed HTML document"""
421 return get_element_html_by_attribute('id', id, html
)
424 def get_element_by_class(class_name
, html
):
425 """Return the content of the first tag with the specified class in the passed HTML document"""
426 retval
= get_elements_by_class(class_name
, html
)
427 return retval
[0] if retval
else None
430 def get_element_html_by_class(class_name
, html
):
431 """Return the html of the first tag with the specified class in the passed HTML document"""
432 retval
= get_elements_html_by_class(class_name
, html
)
433 return retval
[0] if retval
else None
436 def get_element_by_attribute(attribute
, value
, html
, escape_value
=True):
437 retval
= get_elements_by_attribute(attribute
, value
, html
, escape_value
)
438 return retval
[0] if retval
else None
441 def get_element_html_by_attribute(attribute
, value
, html
, escape_value
=True):
442 retval
= get_elements_html_by_attribute(attribute
, value
, html
, escape_value
)
443 return retval
[0] if retval
else None
446 def get_elements_by_class(class_name
, html
):
447 """Return the content of all tags with the specified class in the passed HTML document as a list"""
448 return get_elements_by_attribute(
449 'class', r
'[^\'"]*\b%s\b[^\'"]*' % re.escape(class_name),
450 html, escape_value=False)
453 def get_elements_html_by_class(class_name, html):
454 """Return the html of all tags with the specified class in the passed HTML document as a list"""
455 return get_elements_html_by_attribute(
456 'class', r'[^
\'"]*\b%s\b[^\'"]*' % re.escape(class_name),
457 html, escape_value=False)
460 def get_elements_by_attribute(*args, **kwargs):
461 """Return the content of the tag with the specified attribute in the passed HTML document"""
462 return [content for content, _ in get_elements_text_and_html_by_attribute(*args, **kwargs)]
465 def get_elements_html_by_attribute(*args, **kwargs):
466 """Return the html of the tag with the specified attribute in the passed HTML document"""
467 return [whole for _, whole in get_elements_text_and_html_by_attribute(*args, **kwargs)]
470 def get_elements_text_and_html_by_attribute(attribute, value, html, escape_value=True):
472 Return the text (content) and the html (whole) of the tag with the specified
473 attribute in the passed HTML document
476 value_quote_optional = '' if re.match(r'''[\s"'`
=<>]''', value) else '?'
478 value = re.escape(value) if escape_value else value
480 partial_element_re = r'''(?x
)
481 <(?P
<tag
>[a
-zA
-Z0
-9:._-]+)
482 (?
:\
s(?
:[^
>"']|"[^
"]*"|
'[^']*')*)?
483 \s%(attribute)s\s*=\s*(?P<_q>['"]%(vqo)s)(?-x:%(value)s)(?P=_q)
484 ''' % {'attribute': re.escape(attribute), 'value': value, 'vqo': value_quote_optional}
486 for m in re.finditer(partial_element_re, html):
487 content, whole = get_element_text_and_html_by_tag(m.group('tag'), html[m.start():])
490 unescapeHTML(re.sub(r'^(?P<q>["\'])(?P
<content
>.*)(?P
=q
)$
', r'\g
<content
>', content, flags=re.DOTALL)),
495 class HTMLBreakOnClosingTagParser(compat_HTMLParser):
497 HTML parser which raises HTMLBreakOnClosingTagException upon reaching the
498 closing tag for the first opening tag it has encountered, and can be used
502 class HTMLBreakOnClosingTagException(Exception):
506 self.tagstack = collections.deque()
507 compat_HTMLParser.__init__(self)
512 def __exit__(self, *_):
516 # handle_endtag does not return upon raising HTMLBreakOnClosingTagException,
517 # so data remains buffered; we no longer have any interest in it, thus
518 # override this method to discard it
521 def handle_starttag(self, tag, _):
522 self.tagstack.append(tag)
524 def handle_endtag(self, tag):
525 if not self.tagstack:
526 raise compat_HTMLParseError('no tags
in the stack
')
528 inner_tag = self.tagstack.pop()
532 raise compat_HTMLParseError(f'matching opening tag
for closing {tag} tag
not found
')
533 if not self.tagstack:
534 raise self.HTMLBreakOnClosingTagException()
537 def get_element_text_and_html_by_tag(tag, html):
539 For the first element with the specified tag in the passed HTML document
540 return its' content (text
) and the whole
element (html
)
542 def find_or_raise(haystack, needle, exc):
544 return haystack.index(needle)
547 closing_tag = f'</{tag}>'
548 whole_start = find_or_raise(
549 html, f'<{tag}', compat_HTMLParseError(f'opening {tag} tag not found'))
550 content_start = find_or_raise(
551 html[whole_start:], '>', compat_HTMLParseError(f'malformed opening {tag} tag'))
552 content_start += whole_start + 1
553 with HTMLBreakOnClosingTagParser() as parser:
554 parser.feed(html[whole_start:content_start])
555 if not parser.tagstack or parser.tagstack[0] != tag:
556 raise compat_HTMLParseError(f'parser did not match opening {tag} tag')
557 offset = content_start
558 while offset < len(html):
559 next_closing_tag_start = find_or_raise(
560 html[offset:], closing_tag,
561 compat_HTMLParseError(f'closing {tag} tag not found'))
562 next_closing_tag_end = next_closing_tag_start + len(closing_tag)
564 parser.feed(html[offset:offset + next_closing_tag_end])
565 offset += next_closing_tag_end
566 except HTMLBreakOnClosingTagParser.HTMLBreakOnClosingTagException:
567 return html[content_start:offset + next_closing_tag_start], \
568 html[whole_start:offset + next_closing_tag_end]
569 raise compat_HTMLParseError('unexpected end of html')
572 class HTMLAttributeParser(compat_HTMLParser):
573 """Trivial HTML parser to gather the attributes
for a single element
"""
577 compat_HTMLParser.__init__(self)
579 def handle_starttag(self, tag, attrs):
580 self.attrs = dict(attrs)
583 class HTMLListAttrsParser(compat_HTMLParser):
584 """HTML parser to gather the attributes
for the elements of a
list"""
587 compat_HTMLParser.__init__(self)
591 def handle_starttag(self, tag, attrs):
592 if tag == 'li' and self._level == 0:
593 self.items.append(dict(attrs))
596 def handle_endtag(self, tag):
600 def extract_attributes(html_element):
601 """Given a string
for an HTML element such
as
603 a
="foo" B
="bar" c
="&98;az" d
=boz
604 empty
= noval entity
="&"
607 Decode
and return a dictionary of attributes
.
609 'a': 'foo', 'b': 'bar', c
: 'baz', d
: 'boz',
610 'empty': '', 'noval': None, 'entity': '&',
611 'sq': '"', 'dq': '\''
613 NB HTMLParser
is stricter
in Python
2.6 & 3.2 than
in later versions
,
614 but the cases
in the unit test will work
for all of
2.6, 2.7, 3.2-3.5.
616 parser = HTMLAttributeParser()
618 parser.feed(html_element)
620 # Older Python may throw HTMLParseError in case of malformed HTML
621 except compat_HTMLParseError:
626 def parse_list(webpage):
627 """Given a string
for an series of HTML
<li
> elements
,
628 return a dictionary of their attributes
"""
629 parser = HTMLListAttrsParser()
635 def clean_html(html):
636 """Clean an HTML snippet into a readable string
"""
638 if html is None: # Convenience for sanitizing descriptions etc.
642 html = html.replace('\n', ' ')
643 html = re.sub(r'(?u)\s*<\s*br\s*/?\s*>\s*', '\n', html)
644 html = re.sub(r'(?u)<\s*/\s*p\s*>\s*<\s*p[^>]*>', '\n', html)
646 html = re.sub('<.*?>', '', html)
647 # Replace html entities
648 html = unescapeHTML(html)
652 def sanitize_open(filename, open_mode):
653 """Try to
open the given filename
, and slightly tweak it
if this fails
.
655 Attempts to
open the given filename
. If this fails
, it tries to change
656 the filename slightly
, step by step
, until it
's either able to open it
657 or it fails and raises a final exception, like the standard open()
660 It returns the tuple (stream, definitive_file_name).
664 if sys.platform == 'win32
':
666 msvcrt.setmode(sys.stdout.fileno(), os.O_BINARY)
667 return (sys.stdout.buffer if hasattr(sys.stdout, 'buffer') else sys.stdout, filename)
668 stream = open(encodeFilename(filename), open_mode)
669 return (stream, filename)
670 except (IOError, OSError) as err:
671 if err.errno in (errno.EACCES,):
674 # In case of error, try to remove win32 forbidden chars
675 alt_filename = sanitize_path(filename)
676 if alt_filename == filename:
679 # An exception here should be caught in the caller
680 stream = open(encodeFilename(alt_filename), open_mode)
681 return (stream, alt_filename)
684 def timeconvert(timestr):
685 """Convert RFC 2822 defined time string into system timestamp"""
687 timetuple = email.utils.parsedate_tz(timestr)
688 if timetuple is not None:
689 timestamp = email.utils.mktime_tz(timetuple)
693 def sanitize_filename(s, restricted=False, is_id=False):
694 """Sanitizes a string so it could be used as part of a filename.
695 If restricted is set, use a stricter subset of allowed characters.
696 Set is_id if this is not an arbitrary string, but an ID that should be kept
699 def replace_insane(char):
700 if restricted and char in ACCENT_CHARS:
701 return ACCENT_CHARS[char]
702 elif not restricted and char == '\n':
704 elif char == '?
' or ord(char) < 32 or ord(char) == 127:
707 return '' if restricted else '\''
709 return '_-' if restricted else ' -'
710 elif char in '\\/|*<>':
712 if restricted and (char in '!&\'()[]{}$;`^,#' or char.isspace()):
714 if restricted and ord(char) > 127:
721 s = re.sub(r'[0-9]+(?::[0-9]+)+', lambda m: m.group(0).replace(':', '_'), s)
722 result = ''.join(map(replace_insane, s))
724 while '__' in result:
725 result = result.replace('__', '_')
726 result = result.strip('_')
727 # Common case of "Foreign band name
- English song title
"
728 if restricted and result.startswith('-_'):
730 if result.startswith('-'):
731 result = '_' + result[len('-'):]
732 result = result.lstrip('.')
738 def sanitize_path(s, force=False):
739 """Sanitizes and normalizes path on Windows"""
740 if sys.platform == 'win32':
742 drive_or_unc, _ = os.path.splitdrive(s)
743 if sys.version_info < (2, 7) and not drive_or_unc:
744 drive_or_unc, _ = os.path.splitunc(s)
750 norm_path = os.path.normpath(remove_start(s, drive_or_unc)).split(os.path.sep)
754 path_part if path_part in ['.', '..'] else re.sub(r'(?:[/<>:"\|
\\?\
*]|
[\s
.]$
)', '#', path_part)
755 for path_part
in norm_path
]
757 sanitized_path
.insert(0, drive_or_unc
+ os
.path
.sep
)
758 elif force
and s
[0] == os
.path
.sep
:
759 sanitized_path
.insert(0, os
.path
.sep
)
760 return os
.path
.join(*sanitized_path
)
763 def sanitize_url(url
):
764 # Prepend protocol-less URLs with `http:` scheme in order to mitigate
765 # the number of unwanted failures due to missing protocol
766 if url
.startswith('//'):
767 return 'http:%s' % url
768 # Fix some common typos seen so far
770 # https://github.com/ytdl-org/youtube-dl/issues/15649
771 (r
'^httpss://', r
'https://'),
772 # https://bx1.be/lives/direct-tv/
773 (r
'^rmtp([es]?)://', r
'rtmp\1://'),
775 for mistake
, fixup
in COMMON_TYPOS
:
776 if re
.match(mistake
, url
):
777 return re
.sub(mistake
, fixup
, url
)
781 def extract_basic_auth(url
):
782 parts
= compat_urlparse
.urlsplit(url
)
783 if parts
.username
is None:
785 url
= compat_urlparse
.urlunsplit(parts
._replace
(netloc
=(
786 parts
.hostname
if parts
.port
is None
787 else '%s:%d' % (parts
.hostname
, parts
.port
))))
788 auth_payload
= base64
.b64encode(
789 ('%s:%s' % (parts
.username
, parts
.password
or '')).encode('utf-8'))
790 return url
, 'Basic ' + auth_payload
.decode('utf-8')
793 def sanitized_Request(url
, *args
, **kwargs
):
794 url
, auth_header
= extract_basic_auth(escape_url(sanitize_url(url
)))
795 if auth_header
is not None:
796 headers
= args
[1] if len(args
) >= 2 else kwargs
.setdefault('headers', {})
797 headers
['Authorization'] = auth_header
798 return compat_urllib_request
.Request(url
, *args
, **kwargs
)
802 """Expand shell variables and ~"""
803 return os
.path
.expandvars(compat_expanduser(s
))
806 def orderedSet(iterable
):
807 """ Remove all duplicates from the input iterable """
815 def _htmlentity_transform(entity_with_semicolon
):
816 """Transforms an HTML entity to a character."""
817 entity
= entity_with_semicolon
[:-1]
819 # Known non-numeric HTML entity
820 if entity
in compat_html_entities
.name2codepoint
:
821 return compat_chr(compat_html_entities
.name2codepoint
[entity
])
823 # TODO: HTML5 allows entities without a semicolon. For example,
824 # 'Éric' should be decoded as 'Éric'.
825 if entity_with_semicolon
in compat_html_entities_html5
:
826 return compat_html_entities_html5
[entity_with_semicolon
]
828 mobj
= re
.match(r
'#(x[0-9a-fA-F]+|[0-9]+)', entity
)
830 numstr
= mobj
.group(1)
831 if numstr
.startswith('x'):
833 numstr
= '0%s' % numstr
836 # See https://github.com/ytdl-org/youtube-dl/issues/7518
838 return compat_chr(int(numstr
, base
))
842 # Unknown entity in name, return its literal representation
843 return '&%s;' % entity
849 assert type(s
) == compat_str
852 r
'&([^&;]+;)', lambda m
: _htmlentity_transform(m
.group(1)), s
)
855 def escapeHTML(text
):
858 .replace('&', '&')
859 .replace('<', '<')
860 .replace('>', '>')
861 .replace('"', '"')
862 .replace("'", ''')
866 def process_communicate_or_kill(p
, *args
, **kwargs
):
868 return p
.communicate(*args
, **kwargs
)
869 except BaseException
: # Including KeyboardInterrupt
875 class Popen(subprocess
.Popen
):
876 if sys
.platform
== 'win32':
877 _startupinfo
= subprocess
.STARTUPINFO()
878 _startupinfo
.dwFlags |
= subprocess
.STARTF_USESHOWWINDOW
882 def __init__(self
, *args
, **kwargs
):
883 super(Popen
, self
).__init
__(*args
, **kwargs
, startupinfo
=self
._startupinfo
)
885 def communicate_or_kill(self
, *args
, **kwargs
):
886 return process_communicate_or_kill(self
, *args
, **kwargs
)
889 def get_subprocess_encoding():
890 if sys
.platform
== 'win32' and sys
.getwindowsversion()[0] >= 5:
891 # For subprocess calls, encode with locale encoding
892 # Refer to http://stackoverflow.com/a/9951851/35070
893 encoding
= preferredencoding()
895 encoding
= sys
.getfilesystemencoding()
901 def encodeFilename(s
, for_subprocess
=False):
903 @param s The name of the file
906 assert type(s
) == compat_str
908 # Python 3 has a Unicode API
909 if sys
.version_info
>= (3, 0):
912 # Pass '' directly to use Unicode APIs on Windows 2000 and up
913 # (Detecting Windows NT 4 is tricky because 'major >= 4' would
914 # match Windows 9x series as well. Besides, NT 4 is obsolete.)
915 if not for_subprocess
and sys
.platform
== 'win32' and sys
.getwindowsversion()[0] >= 5:
918 # Jython assumes filenames are Unicode strings though reported as Python 2.x compatible
919 if sys
.platform
.startswith('java'):
922 return s
.encode(get_subprocess_encoding(), 'ignore')
925 def decodeFilename(b
, for_subprocess
=False):
927 if sys
.version_info
>= (3, 0):
930 if not isinstance(b
, bytes):
933 return b
.decode(get_subprocess_encoding(), 'ignore')
936 def encodeArgument(s
):
937 if not isinstance(s
, compat_str
):
938 # Legacy code that uses byte strings
939 # Uncomment the following line after fixing all post processors
940 # assert False, 'Internal error: %r should be of type %r, is %r' % (s, compat_str, type(s))
941 s
= s
.decode('ascii')
942 return encodeFilename(s
, True)
945 def decodeArgument(b
):
946 return decodeFilename(b
, True)
949 def decodeOption(optval
):
952 if isinstance(optval
, bytes):
953 optval
= optval
.decode(preferredencoding())
955 assert isinstance(optval
, compat_str
)
959 _timetuple
= collections
.namedtuple('Time', ('hours', 'minutes', 'seconds', 'milliseconds'))
962 def timetuple_from_msec(msec
):
963 secs
, msec
= divmod(msec
, 1000)
964 mins
, secs
= divmod(secs
, 60)
965 hrs
, mins
= divmod(mins
, 60)
966 return _timetuple(hrs
, mins
, secs
, msec
)
969 def formatSeconds(secs
, delim
=':', msec
=False):
970 time
= timetuple_from_msec(secs
* 1000)
972 ret
= '%d%s%02d%s%02d' % (time
.hours
, delim
, time
.minutes
, delim
, time
.seconds
)
974 ret
= '%d%s%02d' % (time
.minutes
, delim
, time
.seconds
)
976 ret
= '%d' % time
.seconds
977 return '%s.%03d' % (ret
, time
.milliseconds
) if msec
else ret
980 def _ssl_load_windows_store_certs(ssl_context
, storename
):
981 # Code adapted from _load_windows_store_certs in https://github.com/python/cpython/blob/main/Lib/ssl.py
983 certs
= [cert
for cert
, encoding
, trust
in ssl
.enum_certificates(storename
)
984 if encoding
== 'x509_asn' and (
985 trust
is True or ssl
.Purpose
.SERVER_AUTH
.oid
in trust
)]
986 except PermissionError
:
990 ssl_context
.load_verify_locations(cadata
=cert
)
995 def make_HTTPS_handler(params
, **kwargs
):
996 opts_check_certificate
= not params
.get('nocheckcertificate')
997 context
= ssl
.SSLContext(ssl
.PROTOCOL_TLS_CLIENT
)
998 context
.check_hostname
= opts_check_certificate
999 context
.verify_mode
= ssl
.CERT_REQUIRED
if opts_check_certificate
else ssl
.CERT_NONE
1000 if opts_check_certificate
:
1002 context
.load_default_certs()
1003 # Work around the issue in load_default_certs when there are bad certificates. See:
1004 # https://github.com/yt-dlp/yt-dlp/issues/1060,
1005 # https://bugs.python.org/issue35665, https://bugs.python.org/issue45312
1006 except ssl
.SSLError
:
1007 # enum_certificates is not present in mingw python. See https://github.com/yt-dlp/yt-dlp/issues/1151
1008 if sys
.platform
== 'win32' and hasattr(ssl
, 'enum_certificates'):
1009 # Create a new context to discard any certificates that were already loaded
1010 context
= ssl
.SSLContext(ssl
.PROTOCOL_TLS_CLIENT
)
1011 context
.check_hostname
, context
.verify_mode
= True, ssl
.CERT_REQUIRED
1012 for storename
in ('CA', 'ROOT'):
1013 _ssl_load_windows_store_certs(context
, storename
)
1014 context
.set_default_verify_paths()
1015 return YoutubeDLHTTPSHandler(params
, context
=context
, **kwargs
)
1018 def bug_reports_message(before
=';'):
1019 if ytdl_is_updateable():
1020 update_cmd
= 'type yt-dlp -U to update'
1022 update_cmd
= 'see https://github.com/yt-dlp/yt-dlp on how to update'
1023 msg
= 'please report this issue on https://github.com/yt-dlp/yt-dlp .'
1024 msg
+= ' Make sure you are using the latest version; %s.' % update_cmd
1025 msg
+= ' Be sure to call yt-dlp with the --verbose flag and include its complete output.'
1027 before
= before
.rstrip()
1028 if not before
or before
.endswith(('.', '!', '?')):
1029 msg
= msg
[0].title() + msg
[1:]
1031 return (before
+ ' ' if before
else '') + msg
1034 class YoutubeDLError(Exception):
1035 """Base exception for YoutubeDL errors."""
1038 def __init__(self
, msg
=None):
1041 elif self
.msg
is None:
1042 self
.msg
= type(self
).__name
__
1043 super().__init
__(self
.msg
)
1046 network_exceptions
= [compat_urllib_error
.URLError
, compat_http_client
.HTTPException
, socket
.error
]
1047 if hasattr(ssl
, 'CertificateError'):
1048 network_exceptions
.append(ssl
.CertificateError
)
1049 network_exceptions
= tuple(network_exceptions
)
1052 class ExtractorError(YoutubeDLError
):
1053 """Error during info extraction."""
1055 def __init__(self
, msg
, tb
=None, expected
=False, cause
=None, video_id
=None, ie
=None):
1056 """ tb, if given, is the original traceback (so that it can be printed out).
1057 If expected is set, this is a normal error message and most likely not a bug in yt-dlp.
1059 if sys
.exc_info()[0] in network_exceptions
:
1064 self
.expected
= expected
1066 self
.video_id
= video_id
1068 self
.exc_info
= sys
.exc_info() # preserve original exception
1070 super(ExtractorError
, self
).__init
__(''.join((
1071 format_field(ie
, template
='[%s] '),
1072 format_field(video_id
, template
='%s: '),
1074 format_field(cause
, template
=' (caused by %r)'),
1075 '' if expected
else bug_reports_message())))
1077 def format_traceback(self
):
1078 if self
.traceback
is None:
1080 return ''.join(traceback
.format_tb(self
.traceback
))
1083 class UnsupportedError(ExtractorError
):
1084 def __init__(self
, url
):
1085 super(UnsupportedError
, self
).__init
__(
1086 'Unsupported URL: %s' % url
, expected
=True)
1090 class RegexNotFoundError(ExtractorError
):
1091 """Error when a regex didn't match"""
1095 class GeoRestrictedError(ExtractorError
):
1096 """Geographic restriction Error exception.
1098 This exception may be thrown when a video is not available from your
1099 geographic location due to geographic restrictions imposed by a website.
1102 def __init__(self
, msg
, countries
=None, **kwargs
):
1103 kwargs
['expected'] = True
1104 super(GeoRestrictedError
, self
).__init
__(msg
, **kwargs
)
1105 self
.countries
= countries
1108 class DownloadError(YoutubeDLError
):
1109 """Download Error exception.
1111 This exception may be thrown by FileDownloader objects if they are not
1112 configured to continue on errors. They will contain the appropriate
1116 def __init__(self
, msg
, exc_info
=None):
1117 """ exc_info, if given, is the original exception that caused the trouble (as returned by sys.exc_info()). """
1118 super(DownloadError
, self
).__init
__(msg
)
1119 self
.exc_info
= exc_info
1122 class EntryNotInPlaylist(YoutubeDLError
):
1123 """Entry not in playlist exception.
1125 This exception will be thrown by YoutubeDL when a requested entry
1126 is not found in the playlist info_dict
1128 msg
= 'Entry not found in info'
1131 class SameFileError(YoutubeDLError
):
1132 """Same File exception.
1134 This exception will be thrown by FileDownloader objects if they detect
1135 multiple files would have to be downloaded to the same file on disk.
1137 msg
= 'Fixed output name but more than one file to download'
1139 def __init__(self
, filename
=None):
1140 if filename
is not None:
1141 self
.msg
+= f
': {filename}'
1142 super().__init
__(self
.msg
)
1145 class PostProcessingError(YoutubeDLError
):
1146 """Post Processing exception.
1148 This exception may be raised by PostProcessor's .run() method to
1149 indicate an error in the postprocessing task.
1153 class DownloadCancelled(YoutubeDLError
):
1154 """ Exception raised when the download queue should be interrupted """
1155 msg
= 'The download was cancelled'
1158 class ExistingVideoReached(DownloadCancelled
):
1159 """ --break-on-existing triggered """
1160 msg
= 'Encountered a video that is already in the archive, stopping due to --break-on-existing'
1163 class RejectedVideoReached(DownloadCancelled
):
1164 """ --break-on-reject triggered """
1165 msg
= 'Encountered a video that did not match filter, stopping due to --break-on-reject'
1168 class MaxDownloadsReached(DownloadCancelled
):
1169 """ --max-downloads limit has been reached. """
1170 msg
= 'Maximum number of downloads reached, stopping due to --max-downloads'
1173 class ReExtractInfo(YoutubeDLError
):
1174 """ Video info needs to be re-extracted. """
1176 def __init__(self
, msg
, expected
=False):
1177 super().__init
__(msg
)
1178 self
.expected
= expected
1181 class ThrottledDownload(ReExtractInfo
):
1182 """ Download speed below --throttled-rate. """
1183 msg
= 'The download speed is below throttle limit'
1186 super().__init
__(self
.msg
, expected
=False)
1189 class UnavailableVideoError(YoutubeDLError
):
1190 """Unavailable Format exception.
1192 This exception will be thrown when a video is requested
1193 in a format that is not available for that video.
1195 msg
= 'Unable to download video'
1197 def __init__(self
, err
=None):
1199 self
.msg
+= f
': {err}'
1200 super().__init
__(self
.msg
)
1203 class ContentTooShortError(YoutubeDLError
):
1204 """Content Too Short exception.
1206 This exception may be raised by FileDownloader objects when a file they
1207 download is too small for what the server announced first, indicating
1208 the connection was probably interrupted.
1211 def __init__(self
, downloaded
, expected
):
1212 super(ContentTooShortError
, self
).__init
__(
1213 'Downloaded {0} bytes, expected {1} bytes'.format(downloaded
, expected
)
1216 self
.downloaded
= downloaded
1217 self
.expected
= expected
1220 class XAttrMetadataError(YoutubeDLError
):
1221 def __init__(self
, code
=None, msg
='Unknown error'):
1222 super(XAttrMetadataError
, self
).__init
__(msg
)
1226 # Parsing code and msg
1227 if (self
.code
in (errno
.ENOSPC
, errno
.EDQUOT
)
1228 or 'No space left' in self
.msg
or 'Disk quota exceeded' in self
.msg
):
1229 self
.reason
= 'NO_SPACE'
1230 elif self
.code
== errno
.E2BIG
or 'Argument list too long' in self
.msg
:
1231 self
.reason
= 'VALUE_TOO_LONG'
1233 self
.reason
= 'NOT_SUPPORTED'
1236 class XAttrUnavailableError(YoutubeDLError
):
1240 def _create_http_connection(ydl_handler
, http_class
, is_https
, *args
, **kwargs
):
1241 # Working around python 2 bug (see http://bugs.python.org/issue17849) by limiting
1242 # expected HTTP responses to meet HTTP/1.0 or later (see also
1243 # https://github.com/ytdl-org/youtube-dl/issues/6727)
1244 if sys
.version_info
< (3, 0):
1245 kwargs
['strict'] = True
1246 hc
= http_class(*args
, **compat_kwargs(kwargs
))
1247 source_address
= ydl_handler
._params
.get('source_address')
1249 if source_address
is not None:
1250 # This is to workaround _create_connection() from socket where it will try all
1251 # address data from getaddrinfo() including IPv6. This filters the result from
1252 # getaddrinfo() based on the source_address value.
1253 # This is based on the cpython socket.create_connection() function.
1254 # https://github.com/python/cpython/blob/master/Lib/socket.py#L691
1255 def _create_connection(address
, timeout
=socket
._GLOBAL
_DEFAULT
_TIMEOUT
, source_address
=None):
1256 host
, port
= address
1258 addrs
= socket
.getaddrinfo(host
, port
, 0, socket
.SOCK_STREAM
)
1259 af
= socket
.AF_INET
if '.' in source_address
[0] else socket
.AF_INET6
1260 ip_addrs
= [addr
for addr
in addrs
if addr
[0] == af
]
1261 if addrs
and not ip_addrs
:
1262 ip_version
= 'v4' if af
== socket
.AF_INET
else 'v6'
1264 "No remote IP%s addresses available for connect, can't use '%s' as source address"
1265 % (ip_version
, source_address
[0]))
1266 for res
in ip_addrs
:
1267 af
, socktype
, proto
, canonname
, sa
= res
1270 sock
= socket
.socket(af
, socktype
, proto
)
1271 if timeout
is not socket
._GLOBAL
_DEFAULT
_TIMEOUT
:
1272 sock
.settimeout(timeout
)
1273 sock
.bind(source_address
)
1275 err
= None # Explicitly break reference cycle
1277 except socket
.error
as _
:
1279 if sock
is not None:
1284 raise socket
.error('getaddrinfo returns an empty list')
1285 if hasattr(hc
, '_create_connection'):
1286 hc
._create
_connection
= _create_connection
1287 sa
= (source_address
, 0)
1288 if hasattr(hc
, 'source_address'): # Python 2.7+
1289 hc
.source_address
= sa
1291 def _hc_connect(self
, *args
, **kwargs
):
1292 sock
= _create_connection(
1293 (self
.host
, self
.port
), self
.timeout
, sa
)
1295 self
.sock
= ssl
.wrap_socket(
1296 sock
, self
.key_file
, self
.cert_file
,
1297 ssl_version
=ssl
.PROTOCOL_TLSv1
)
1300 hc
.connect
= functools
.partial(_hc_connect
, hc
)
1305 def handle_youtubedl_headers(headers
):
1306 filtered_headers
= headers
1308 if 'Youtubedl-no-compression' in filtered_headers
:
1309 filtered_headers
= dict((k
, v
) for k
, v
in filtered_headers
.items() if k
.lower() != 'accept-encoding')
1310 del filtered_headers
['Youtubedl-no-compression']
1312 return filtered_headers
1315 class YoutubeDLHandler(compat_urllib_request
.HTTPHandler
):
1316 """Handler for HTTP requests and responses.
1318 This class, when installed with an OpenerDirector, automatically adds
1319 the standard headers to every HTTP request and handles gzipped and
1320 deflated responses from web servers. If compression is to be avoided in
1321 a particular request, the original request in the program code only has
1322 to include the HTTP header "Youtubedl-no-compression", which will be
1323 removed before making the real request.
1325 Part of this code was copied from:
1327 http://techknack.net/python-urllib2-handlers/
1329 Andrew Rowls, the author of that code, agreed to release it to the
1333 def __init__(self
, params
, *args
, **kwargs
):
1334 compat_urllib_request
.HTTPHandler
.__init
__(self
, *args
, **kwargs
)
1335 self
._params
= params
1337 def http_open(self
, req
):
1338 conn_class
= compat_http_client
.HTTPConnection
1340 socks_proxy
= req
.headers
.get('Ytdl-socks-proxy')
1342 conn_class
= make_socks_conn_class(conn_class
, socks_proxy
)
1343 del req
.headers
['Ytdl-socks-proxy']
1345 return self
.do_open(functools
.partial(
1346 _create_http_connection
, self
, conn_class
, False),
1354 return zlib
.decompress(data
, -zlib
.MAX_WBITS
)
1356 return zlib
.decompress(data
)
1358 def http_request(self
, req
):
1359 # According to RFC 3986, URLs can not contain non-ASCII characters, however this is not
1360 # always respected by websites, some tend to give out URLs with non percent-encoded
1361 # non-ASCII characters (see telemb.py, ard.py [#3412])
1362 # urllib chokes on URLs with non-ASCII characters (see http://bugs.python.org/issue3991)
1363 # To work around aforementioned issue we will replace request's original URL with
1364 # percent-encoded one
1365 # Since redirects are also affected (e.g. http://www.southpark.de/alle-episoden/s18e09)
1366 # the code of this workaround has been moved here from YoutubeDL.urlopen()
1367 url
= req
.get_full_url()
1368 url_escaped
= escape_url(url
)
1370 # Substitute URL if any change after escaping
1371 if url
!= url_escaped
:
1372 req
= update_Request(req
, url
=url_escaped
)
1374 for h
, v
in std_headers
.items():
1375 # Capitalize is needed because of Python bug 2275: http://bugs.python.org/issue2275
1376 # The dict keys are capitalized because of this bug by urllib
1377 if h
.capitalize() not in req
.headers
:
1378 req
.add_header(h
, v
)
1380 req
.headers
= handle_youtubedl_headers(req
.headers
)
1382 if sys
.version_info
< (2, 7) and '#' in req
.get_full_url():
1383 # Python 2.6 is brain-dead when it comes to fragments
1384 req
._Request
__original
= req
._Request
__original
.partition('#')[0]
1385 req
._Request
__r
_type
= req
._Request
__r
_type
.partition('#')[0]
1389 def http_response(self
, req
, resp
):
1392 if resp
.headers
.get('Content-encoding', '') == 'gzip':
1393 content
= resp
.read()
1394 gz
= gzip
.GzipFile(fileobj
=io
.BytesIO(content
), mode
='rb')
1396 uncompressed
= io
.BytesIO(gz
.read())
1397 except IOError as original_ioerror
:
1398 # There may be junk add the end of the file
1399 # See http://stackoverflow.com/q/4928560/35070 for details
1400 for i
in range(1, 1024):
1402 gz
= gzip
.GzipFile(fileobj
=io
.BytesIO(content
[:-i
]), mode
='rb')
1403 uncompressed
= io
.BytesIO(gz
.read())
1408 raise original_ioerror
1409 resp
= compat_urllib_request
.addinfourl(uncompressed
, old_resp
.headers
, old_resp
.url
, old_resp
.code
)
1410 resp
.msg
= old_resp
.msg
1411 del resp
.headers
['Content-encoding']
1413 if resp
.headers
.get('Content-encoding', '') == 'deflate':
1414 gz
= io
.BytesIO(self
.deflate(resp
.read()))
1415 resp
= compat_urllib_request
.addinfourl(gz
, old_resp
.headers
, old_resp
.url
, old_resp
.code
)
1416 resp
.msg
= old_resp
.msg
1417 del resp
.headers
['Content-encoding']
1418 # Percent-encode redirect URL of Location HTTP header to satisfy RFC 3986 (see
1419 # https://github.com/ytdl-org/youtube-dl/issues/6457).
1420 if 300 <= resp
.code
< 400:
1421 location
= resp
.headers
.get('Location')
1423 # As of RFC 2616 default charset is iso-8859-1 that is respected by python 3
1424 if sys
.version_info
>= (3, 0):
1425 location
= location
.encode('iso-8859-1').decode('utf-8')
1427 location
= location
.decode('utf-8')
1428 location_escaped
= escape_url(location
)
1429 if location
!= location_escaped
:
1430 del resp
.headers
['Location']
1431 if sys
.version_info
< (3, 0):
1432 location_escaped
= location_escaped
.encode('utf-8')
1433 resp
.headers
['Location'] = location_escaped
1436 https_request
= http_request
1437 https_response
= http_response
1440 def make_socks_conn_class(base_class
, socks_proxy
):
1441 assert issubclass(base_class
, (
1442 compat_http_client
.HTTPConnection
, compat_http_client
.HTTPSConnection
))
1444 url_components
= compat_urlparse
.urlparse(socks_proxy
)
1445 if url_components
.scheme
.lower() == 'socks5':
1446 socks_type
= ProxyType
.SOCKS5
1447 elif url_components
.scheme
.lower() in ('socks', 'socks4'):
1448 socks_type
= ProxyType
.SOCKS4
1449 elif url_components
.scheme
.lower() == 'socks4a':
1450 socks_type
= ProxyType
.SOCKS4A
1452 def unquote_if_non_empty(s
):
1455 return compat_urllib_parse_unquote_plus(s
)
1459 url_components
.hostname
, url_components
.port
or 1080,
1461 unquote_if_non_empty(url_components
.username
),
1462 unquote_if_non_empty(url_components
.password
),
1465 class SocksConnection(base_class
):
1467 self
.sock
= sockssocket()
1468 self
.sock
.setproxy(*proxy_args
)
1469 if type(self
.timeout
) in (int, float):
1470 self
.sock
.settimeout(self
.timeout
)
1471 self
.sock
.connect((self
.host
, self
.port
))
1473 if isinstance(self
, compat_http_client
.HTTPSConnection
):
1474 if hasattr(self
, '_context'): # Python > 2.6
1475 self
.sock
= self
._context
.wrap_socket(
1476 self
.sock
, server_hostname
=self
.host
)
1478 self
.sock
= ssl
.wrap_socket(self
.sock
)
1480 return SocksConnection
1483 class YoutubeDLHTTPSHandler(compat_urllib_request
.HTTPSHandler
):
1484 def __init__(self
, params
, https_conn_class
=None, *args
, **kwargs
):
1485 compat_urllib_request
.HTTPSHandler
.__init
__(self
, *args
, **kwargs
)
1486 self
._https
_conn
_class
= https_conn_class
or compat_http_client
.HTTPSConnection
1487 self
._params
= params
1489 def https_open(self
, req
):
1491 conn_class
= self
._https
_conn
_class
1493 if hasattr(self
, '_context'): # python > 2.6
1494 kwargs
['context'] = self
._context
1495 if hasattr(self
, '_check_hostname'): # python 3.x
1496 kwargs
['check_hostname'] = self
._check
_hostname
1498 socks_proxy
= req
.headers
.get('Ytdl-socks-proxy')
1500 conn_class
= make_socks_conn_class(conn_class
, socks_proxy
)
1501 del req
.headers
['Ytdl-socks-proxy']
1503 return self
.do_open(functools
.partial(
1504 _create_http_connection
, self
, conn_class
, True),
1508 class YoutubeDLCookieJar(compat_cookiejar
.MozillaCookieJar
):
1510 See [1] for cookie file format.
1512 1. https://curl.haxx.se/docs/http-cookies.html
1514 _HTTPONLY_PREFIX
= '#HttpOnly_'
1516 _HEADER
= '''# Netscape HTTP Cookie File
1517 # This file is generated by yt-dlp. Do not edit.
1520 _CookieFileEntry
= collections
.namedtuple(
1522 ('domain_name', 'include_subdomains', 'path', 'https_only', 'expires_at', 'name', 'value'))
1524 def save(self
, filename
=None, ignore_discard
=False, ignore_expires
=False):
1526 Save cookies to a file.
1528 Most of the code is taken from CPython 3.8 and slightly adapted
1529 to support cookie files with UTF-8 in both python 2 and 3.
1531 if filename
is None:
1532 if self
.filename
is not None:
1533 filename
= self
.filename
1535 raise ValueError(compat_cookiejar
.MISSING_FILENAME_TEXT
)
1537 # Store session cookies with `expires` set to 0 instead of an empty
1540 if cookie
.expires
is None:
1543 with io
.open(filename
, 'w', encoding
='utf-8') as f
:
1544 f
.write(self
._HEADER
)
1547 if not ignore_discard
and cookie
.discard
:
1549 if not ignore_expires
and cookie
.is_expired(now
):
1555 if cookie
.domain
.startswith('.'):
1556 initial_dot
= 'TRUE'
1558 initial_dot
= 'FALSE'
1559 if cookie
.expires
is not None:
1560 expires
= compat_str(cookie
.expires
)
1563 if cookie
.value
is None:
1564 # cookies.txt regards 'Set-Cookie: foo' as a cookie
1565 # with no name, whereas http.cookiejar regards it as a
1566 # cookie with no value.
1571 value
= cookie
.value
1573 '\t'.join([cookie
.domain
, initial_dot
, cookie
.path
,
1574 secure
, expires
, name
, value
]) + '\n')
1576 def load(self
, filename
=None, ignore_discard
=False, ignore_expires
=False):
1577 """Load cookies from a file."""
1578 if filename
is None:
1579 if self
.filename
is not None:
1580 filename
= self
.filename
1582 raise ValueError(compat_cookiejar
.MISSING_FILENAME_TEXT
)
1584 def prepare_line(line
):
1585 if line
.startswith(self
._HTTPONLY
_PREFIX
):
1586 line
= line
[len(self
._HTTPONLY
_PREFIX
):]
1587 # comments and empty lines are fine
1588 if line
.startswith('#') or not line
.strip():
1590 cookie_list
= line
.split('\t')
1591 if len(cookie_list
) != self
._ENTRY
_LEN
:
1592 raise compat_cookiejar
.LoadError('invalid length %d' % len(cookie_list
))
1593 cookie
= self
._CookieFileEntry
(*cookie_list
)
1594 if cookie
.expires_at
and not cookie
.expires_at
.isdigit():
1595 raise compat_cookiejar
.LoadError('invalid expires at %s' % cookie
.expires_at
)
1599 with io
.open(filename
, encoding
='utf-8') as f
:
1602 cf
.write(prepare_line(line
))
1603 except compat_cookiejar
.LoadError
as e
:
1605 'WARNING: skipping cookie file entry due to %s: %r\n'
1606 % (e
, line
), sys
.stderr
)
1609 self
._really
_load
(cf
, filename
, ignore_discard
, ignore_expires
)
1610 # Session cookies are denoted by either `expires` field set to
1611 # an empty string or 0. MozillaCookieJar only recognizes the former
1612 # (see [1]). So we need force the latter to be recognized as session
1613 # cookies on our own.
1614 # Session cookies may be important for cookies-based authentication,
1615 # e.g. usually, when user does not check 'Remember me' check box while
1616 # logging in on a site, some important cookies are stored as session
1617 # cookies so that not recognizing them will result in failed login.
1618 # 1. https://bugs.python.org/issue17164
1620 # Treat `expires=0` cookies as session cookies
1621 if cookie
.expires
== 0:
1622 cookie
.expires
= None
1623 cookie
.discard
= True
1626 class YoutubeDLCookieProcessor(compat_urllib_request
.HTTPCookieProcessor
):
1627 def __init__(self
, cookiejar
=None):
1628 compat_urllib_request
.HTTPCookieProcessor
.__init
__(self
, cookiejar
)
1630 def http_response(self
, request
, response
):
1631 # Python 2 will choke on next HTTP request in row if there are non-ASCII
1632 # characters in Set-Cookie HTTP header of last response (see
1633 # https://github.com/ytdl-org/youtube-dl/issues/6769).
1634 # In order to at least prevent crashing we will percent encode Set-Cookie
1635 # header before HTTPCookieProcessor starts processing it.
1636 # if sys.version_info < (3, 0) and response.headers:
1637 # for set_cookie_header in ('Set-Cookie', 'Set-Cookie2'):
1638 # set_cookie = response.headers.get(set_cookie_header)
1640 # set_cookie_escaped = compat_urllib_parse.quote(set_cookie, b"%/;:@&=+$,!~*'()?#[] ")
1641 # if set_cookie != set_cookie_escaped:
1642 # del response.headers[set_cookie_header]
1643 # response.headers[set_cookie_header] = set_cookie_escaped
1644 return compat_urllib_request
.HTTPCookieProcessor
.http_response(self
, request
, response
)
1646 https_request
= compat_urllib_request
.HTTPCookieProcessor
.http_request
1647 https_response
= http_response
1650 class YoutubeDLRedirectHandler(compat_urllib_request
.HTTPRedirectHandler
):
1651 """YoutubeDL redirect handler
1653 The code is based on HTTPRedirectHandler implementation from CPython [1].
1655 This redirect handler solves two issues:
1656 - ensures redirect URL is always unicode under python 2
1657 - introduces support for experimental HTTP response status code
1658 308 Permanent Redirect [2] used by some sites [3]
1660 1. https://github.com/python/cpython/blob/master/Lib/urllib/request.py
1661 2. https://developer.mozilla.org/en-US/docs/Web/HTTP/Status/308
1662 3. https://github.com/ytdl-org/youtube-dl/issues/28768
1665 http_error_301
= http_error_303
= http_error_307
= http_error_308
= compat_urllib_request
.HTTPRedirectHandler
.http_error_302
1667 def redirect_request(self
, req
, fp
, code
, msg
, headers
, newurl
):
1668 """Return a Request or None in response to a redirect.
1670 This is called by the http_error_30x methods when a
1671 redirection response is received. If a redirection should
1672 take place, return a new Request to allow http_error_30x to
1673 perform the redirect. Otherwise, raise HTTPError if no-one
1674 else should try to handle this url. Return None if you can't
1675 but another Handler might.
1677 m
= req
.get_method()
1678 if (not (code
in (301, 302, 303, 307, 308) and m
in ("GET", "HEAD")
1679 or code
in (301, 302, 303) and m
== "POST")):
1680 raise compat_HTTPError(req
.full_url
, code
, msg
, headers
, fp
)
1681 # Strictly (according to RFC 2616), 301 or 302 in response to
1682 # a POST MUST NOT cause a redirection without confirmation
1683 # from the user (of urllib.request, in this case). In practice,
1684 # essentially all clients do redirect in this case, so we do
1687 # On python 2 urlh.geturl() may sometimes return redirect URL
1688 # as byte string instead of unicode. This workaround allows
1689 # to force it always return unicode.
1690 if sys
.version_info
[0] < 3:
1691 newurl
= compat_str(newurl
)
1693 # Be conciliant with URIs containing a space. This is mainly
1694 # redundant with the more complete encoding done in http_error_302(),
1695 # but it is kept for compatibility with other callers.
1696 newurl
= newurl
.replace(' ', '%20')
1698 CONTENT_HEADERS
= ("content-length", "content-type")
1699 # NB: don't use dict comprehension for python 2.6 compatibility
1700 newheaders
= dict((k
, v
) for k
, v
in req
.headers
.items()
1701 if k
.lower() not in CONTENT_HEADERS
)
1702 return compat_urllib_request
.Request(
1703 newurl
, headers
=newheaders
, origin_req_host
=req
.origin_req_host
,
1707 def extract_timezone(date_str
):
1710 ^.{8,}? # >=8 char non-TZ prefix, if present
1711 (?P<tz>Z| # just the UTC Z, or
1712 (?:(?<=.\b\d{4}|\b\d{2}:\d\d)| # preceded by 4 digits or hh:mm or
1713 (?<!.\b[a-zA-Z]{3}|[a-zA-Z]{4}|..\b\d\d)) # not preceded by 3 alpha word or >= 4 alpha or 2 digits
1714 [ ]? # optional space
1715 (?P<sign>\+|-) # +/-
1716 (?P<hours>[0-9]{2}):?(?P<minutes>[0-9]{2}) # hh[:]mm
1720 timezone
= datetime
.timedelta()
1722 date_str
= date_str
[:-len(m
.group('tz'))]
1723 if not m
.group('sign'):
1724 timezone
= datetime
.timedelta()
1726 sign
= 1 if m
.group('sign') == '+' else -1
1727 timezone
= datetime
.timedelta(
1728 hours
=sign
* int(m
.group('hours')),
1729 minutes
=sign
* int(m
.group('minutes')))
1730 return timezone
, date_str
1733 def parse_iso8601(date_str
, delimiter
='T', timezone
=None):
1734 """ Return a UNIX timestamp from the given date """
1736 if date_str
is None:
1739 date_str
= re
.sub(r
'\.[0-9]+', '', date_str
)
1741 if timezone
is None:
1742 timezone
, date_str
= extract_timezone(date_str
)
1745 date_format
= '%Y-%m-%d{0}%H:%M:%S'.format(delimiter
)
1746 dt
= datetime
.datetime
.strptime(date_str
, date_format
) - timezone
1747 return calendar
.timegm(dt
.timetuple())
1752 def date_formats(day_first
=True):
1753 return DATE_FORMATS_DAY_FIRST
if day_first
else DATE_FORMATS_MONTH_FIRST
1756 def unified_strdate(date_str
, day_first
=True):
1757 """Return a string with the date in the format YYYYMMDD"""
1759 if date_str
is None:
1763 date_str
= date_str
.replace(',', ' ')
1764 # Remove AM/PM + timezone
1765 date_str
= re
.sub(r
'(?i)\s*(?:AM|PM)(?:\s+[A-Z]+)?', '', date_str
)
1766 _
, date_str
= extract_timezone(date_str
)
1768 for expression
in date_formats(day_first
):
1770 upload_date
= datetime
.datetime
.strptime(date_str
, expression
).strftime('%Y%m%d')
1773 if upload_date
is None:
1774 timetuple
= email
.utils
.parsedate_tz(date_str
)
1777 upload_date
= datetime
.datetime(*timetuple
[:6]).strftime('%Y%m%d')
1780 if upload_date
is not None:
1781 return compat_str(upload_date
)
1784 def unified_timestamp(date_str
, day_first
=True):
1785 if date_str
is None:
1788 date_str
= re
.sub(r
'[,|]', '', date_str
)
1790 pm_delta
= 12 if re
.search(r
'(?i)PM', date_str
) else 0
1791 timezone
, date_str
= extract_timezone(date_str
)
1793 # Remove AM/PM + timezone
1794 date_str
= re
.sub(r
'(?i)\s*(?:AM|PM)(?:\s+[A-Z]+)?', '', date_str
)
1796 # Remove unrecognized timezones from ISO 8601 alike timestamps
1797 m
= re
.search(r
'\d{1,2}:\d{1,2}(?:\.\d+)?(?P<tz>\s*[A-Z]+)$', date_str
)
1799 date_str
= date_str
[:-len(m
.group('tz'))]
1801 # Python only supports microseconds, so remove nanoseconds
1802 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]+$', date_str
)
1804 date_str
= m
.group(1)
1806 for expression
in date_formats(day_first
):
1808 dt
= datetime
.datetime
.strptime(date_str
, expression
) - timezone
+ datetime
.timedelta(hours
=pm_delta
)
1809 return calendar
.timegm(dt
.timetuple())
1812 timetuple
= email
.utils
.parsedate_tz(date_str
)
1814 return calendar
.timegm(timetuple
) + pm_delta
* 3600
1817 def determine_ext(url
, default_ext
='unknown_video'):
1818 if url
is None or '.' not in url
:
1820 guess
= url
.partition('?')[0].rpartition('.')[2]
1821 if re
.match(r
'^[A-Za-z0-9]+$', guess
):
1823 # Try extract ext from URLs like http://example.com/foo/bar.mp4/?download
1824 elif guess
.rstrip('/') in KNOWN_EXTENSIONS
:
1825 return guess
.rstrip('/')
1830 def subtitles_filename(filename
, sub_lang
, sub_format
, expected_real_ext
=None):
1831 return replace_extension(filename
, sub_lang
+ '.' + sub_format
, expected_real_ext
)
1834 def datetime_from_str(date_str
, precision
='auto', format
='%Y%m%d'):
1836 Return a datetime object from a string in the format YYYYMMDD or
1837 (now|today|date)[+-][0-9](microsecond|second|minute|hour|day|week|month|year)(s)?
1839 format: string date format used to return datetime object from
1840 precision: round the time portion of a datetime object.
1841 auto|microsecond|second|minute|hour|day.
1842 auto: round to the unit provided in date_str (if applicable).
1844 auto_precision
= False
1845 if precision
== 'auto':
1846 auto_precision
= True
1847 precision
= 'microsecond'
1848 today
= datetime_round(datetime
.datetime
.now(), precision
)
1849 if date_str
in ('now', 'today'):
1851 if date_str
== 'yesterday':
1852 return today
- datetime
.timedelta(days
=1)
1854 r
'(?P<start>.+)(?P<sign>[+-])(?P<time>\d+)(?P<unit>microsecond|second|minute|hour|day|week|month|year)(s)?',
1856 if match
is not None:
1857 start_time
= datetime_from_str(match
.group('start'), precision
, format
)
1858 time
= int(match
.group('time')) * (-1 if match
.group('sign') == '-' else 1)
1859 unit
= match
.group('unit')
1860 if unit
== 'month' or unit
== 'year':
1861 new_date
= datetime_add_months(start_time
, time
* 12 if unit
== 'year' else time
)
1867 delta
= datetime
.timedelta(**{unit + 's': time}
)
1868 new_date
= start_time
+ delta
1870 return datetime_round(new_date
, unit
)
1873 return datetime_round(datetime
.datetime
.strptime(date_str
, format
), precision
)
1876 def date_from_str(date_str
, format
='%Y%m%d'):
1878 Return a datetime object from a string in the format YYYYMMDD or
1879 (now|today|date)[+-][0-9](microsecond|second|minute|hour|day|week|month|year)(s)?
1881 format: string date format used to return datetime object from
1883 return datetime_from_str(date_str
, precision
='microsecond', format
=format
).date()
1886 def datetime_add_months(dt
, months
):
1887 """Increment/Decrement a datetime object by months."""
1888 month
= dt
.month
+ months
- 1
1889 year
= dt
.year
+ month
// 12
1890 month
= month
% 12 + 1
1891 day
= min(dt
.day
, calendar
.monthrange(year
, month
)[1])
1892 return dt
.replace(year
, month
, day
)
1895 def datetime_round(dt
, precision
='day'):
1897 Round a datetime object's time to a specific precision
1899 if precision
== 'microsecond':
1908 roundto
= lambda x
, n
: ((x
+ n
/ 2) // n
) * n
1909 timestamp
= calendar
.timegm(dt
.timetuple())
1910 return datetime
.datetime
.utcfromtimestamp(roundto(timestamp
, unit_seconds
[precision
]))
1913 def hyphenate_date(date_str
):
1915 Convert a date in 'YYYYMMDD' format to 'YYYY-MM-DD' format"""
1916 match
= re
.match(r
'^(\d\d\d\d)(\d\d)(\d\d)$', date_str
)
1917 if match
is not None:
1918 return '-'.join(match
.groups())
1923 class DateRange(object):
1924 """Represents a time interval between two dates"""
1926 def __init__(self
, start
=None, end
=None):
1927 """start and end must be strings in the format accepted by date"""
1928 if start
is not None:
1929 self
.start
= date_from_str(start
)
1931 self
.start
= datetime
.datetime
.min.date()
1933 self
.end
= date_from_str(end
)
1935 self
.end
= datetime
.datetime
.max.date()
1936 if self
.start
> self
.end
:
1937 raise ValueError('Date range: "%s" , the start date must be before the end date' % self
)
1941 """Returns a range that only contains the given day"""
1942 return cls(day
, day
)
1944 def __contains__(self
, date
):
1945 """Check if the date is in the range"""
1946 if not isinstance(date
, datetime
.date
):
1947 date
= date_from_str(date
)
1948 return self
.start
<= date
<= self
.end
1951 return '%s - %s' % (self
.start
.isoformat(), self
.end
.isoformat())
1954 def platform_name():
1955 """ Returns the platform name as a compat_str """
1956 res
= platform
.platform()
1957 if isinstance(res
, bytes):
1958 res
= res
.decode(preferredencoding())
1960 assert isinstance(res
, compat_str
)
1964 def get_windows_version():
1965 ''' Get Windows version. None if it's not running on Windows '''
1966 if compat_os_name
== 'nt':
1967 return version_tuple(platform
.win32_ver()[1])
1972 def _windows_write_string(s
, out
):
1973 """ Returns True if the string was written using special methods,
1974 False if it has yet to be written out."""
1975 # Adapted from http://stackoverflow.com/a/3259271/35070
1977 import ctypes
.wintypes
1985 fileno
= out
.fileno()
1986 except AttributeError:
1987 # If the output stream doesn't have a fileno, it's virtual
1989 except io
.UnsupportedOperation
:
1990 # Some strange Windows pseudo files?
1992 if fileno
not in WIN_OUTPUT_IDS
:
1995 GetStdHandle
= compat_ctypes_WINFUNCTYPE(
1996 ctypes
.wintypes
.HANDLE
, ctypes
.wintypes
.DWORD
)(
1997 ('GetStdHandle', ctypes
.windll
.kernel32
))
1998 h
= GetStdHandle(WIN_OUTPUT_IDS
[fileno
])
2000 WriteConsoleW
= compat_ctypes_WINFUNCTYPE(
2001 ctypes
.wintypes
.BOOL
, ctypes
.wintypes
.HANDLE
, ctypes
.wintypes
.LPWSTR
,
2002 ctypes
.wintypes
.DWORD
, ctypes
.POINTER(ctypes
.wintypes
.DWORD
),
2003 ctypes
.wintypes
.LPVOID
)(('WriteConsoleW', ctypes
.windll
.kernel32
))
2004 written
= ctypes
.wintypes
.DWORD(0)
2006 GetFileType
= compat_ctypes_WINFUNCTYPE(ctypes
.wintypes
.DWORD
, ctypes
.wintypes
.DWORD
)(('GetFileType', ctypes
.windll
.kernel32
))
2007 FILE_TYPE_CHAR
= 0x0002
2008 FILE_TYPE_REMOTE
= 0x8000
2009 GetConsoleMode
= compat_ctypes_WINFUNCTYPE(
2010 ctypes
.wintypes
.BOOL
, ctypes
.wintypes
.HANDLE
,
2011 ctypes
.POINTER(ctypes
.wintypes
.DWORD
))(
2012 ('GetConsoleMode', ctypes
.windll
.kernel32
))
2013 INVALID_HANDLE_VALUE
= ctypes
.wintypes
.DWORD(-1).value
2015 def not_a_console(handle
):
2016 if handle
== INVALID_HANDLE_VALUE
or handle
is None:
2018 return ((GetFileType(handle
) & ~FILE_TYPE_REMOTE
) != FILE_TYPE_CHAR
2019 or GetConsoleMode(handle
, ctypes
.byref(ctypes
.wintypes
.DWORD())) == 0)
2021 if not_a_console(h
):
2024 def next_nonbmp_pos(s
):
2026 return next(i
for i
, c
in enumerate(s
) if ord(c
) > 0xffff)
2027 except StopIteration:
2031 count
= min(next_nonbmp_pos(s
), 1024)
2033 ret
= WriteConsoleW(
2034 h
, s
, count
if count
else 2, ctypes
.byref(written
), None)
2036 raise OSError('Failed to write string')
2037 if not count
: # We just wrote a non-BMP character
2038 assert written
.value
== 2
2041 assert written
.value
> 0
2042 s
= s
[written
.value
:]
2046 def write_string(s
, out
=None, encoding
=None):
2049 assert type(s
) == compat_str
2051 if sys
.platform
== 'win32' and encoding
is None and hasattr(out
, 'fileno'):
2052 if _windows_write_string(s
, out
):
2055 if ('b' in getattr(out
, 'mode', '')
2056 or sys
.version_info
[0] < 3): # Python 2 lies about mode of sys.stderr
2057 byt
= s
.encode(encoding
or preferredencoding(), 'ignore')
2059 elif hasattr(out
, 'buffer'):
2060 enc
= encoding
or getattr(out
, 'encoding', None) or preferredencoding()
2061 byt
= s
.encode(enc
, 'ignore')
2062 out
.buffer.write(byt
)
2068 def bytes_to_intlist(bs
):
2071 if isinstance(bs
[0], int): # Python 3
2074 return [ord(c
) for c
in bs
]
2077 def intlist_to_bytes(xs
):
2080 return compat_struct_pack('%dB' % len(xs
), *xs
)
2083 # Cross-platform file locking
2084 if sys
.platform
== 'win32':
2085 import ctypes
.wintypes
2088 class OVERLAPPED(ctypes
.Structure
):
2090 ('Internal', ctypes
.wintypes
.LPVOID
),
2091 ('InternalHigh', ctypes
.wintypes
.LPVOID
),
2092 ('Offset', ctypes
.wintypes
.DWORD
),
2093 ('OffsetHigh', ctypes
.wintypes
.DWORD
),
2094 ('hEvent', ctypes
.wintypes
.HANDLE
),
2097 kernel32
= ctypes
.windll
.kernel32
2098 LockFileEx
= kernel32
.LockFileEx
2099 LockFileEx
.argtypes
= [
2100 ctypes
.wintypes
.HANDLE
, # hFile
2101 ctypes
.wintypes
.DWORD
, # dwFlags
2102 ctypes
.wintypes
.DWORD
, # dwReserved
2103 ctypes
.wintypes
.DWORD
, # nNumberOfBytesToLockLow
2104 ctypes
.wintypes
.DWORD
, # nNumberOfBytesToLockHigh
2105 ctypes
.POINTER(OVERLAPPED
) # Overlapped
2107 LockFileEx
.restype
= ctypes
.wintypes
.BOOL
2108 UnlockFileEx
= kernel32
.UnlockFileEx
2109 UnlockFileEx
.argtypes
= [
2110 ctypes
.wintypes
.HANDLE
, # hFile
2111 ctypes
.wintypes
.DWORD
, # dwReserved
2112 ctypes
.wintypes
.DWORD
, # nNumberOfBytesToLockLow
2113 ctypes
.wintypes
.DWORD
, # nNumberOfBytesToLockHigh
2114 ctypes
.POINTER(OVERLAPPED
) # Overlapped
2116 UnlockFileEx
.restype
= ctypes
.wintypes
.BOOL
2117 whole_low
= 0xffffffff
2118 whole_high
= 0x7fffffff
2120 def _lock_file(f
, exclusive
):
2121 overlapped
= OVERLAPPED()
2122 overlapped
.Offset
= 0
2123 overlapped
.OffsetHigh
= 0
2124 overlapped
.hEvent
= 0
2125 f
._lock
_file
_overlapped
_p
= ctypes
.pointer(overlapped
)
2126 handle
= msvcrt
.get_osfhandle(f
.fileno())
2127 if not LockFileEx(handle
, 0x2 if exclusive
else 0x0, 0,
2128 whole_low
, whole_high
, f
._lock
_file
_overlapped
_p
):
2129 raise OSError('Locking file failed: %r' % ctypes
.FormatError())
2131 def _unlock_file(f
):
2132 assert f
._lock
_file
_overlapped
_p
2133 handle
= msvcrt
.get_osfhandle(f
.fileno())
2134 if not UnlockFileEx(handle
, 0,
2135 whole_low
, whole_high
, f
._lock
_file
_overlapped
_p
):
2136 raise OSError('Unlocking file failed: %r' % ctypes
.FormatError())
2139 # Some platforms, such as Jython, is missing fcntl
2143 def _lock_file(f
, exclusive
):
2144 fcntl
.flock(f
, fcntl
.LOCK_EX
if exclusive
else fcntl
.LOCK_SH
)
2146 def _unlock_file(f
):
2147 fcntl
.flock(f
, fcntl
.LOCK_UN
)
2149 UNSUPPORTED_MSG
= 'file locking is not supported on this platform'
2151 def _lock_file(f
, exclusive
):
2152 raise IOError(UNSUPPORTED_MSG
)
2154 def _unlock_file(f
):
2155 raise IOError(UNSUPPORTED_MSG
)
2158 class locked_file(object):
2159 def __init__(self
, filename
, mode
, encoding
=None):
2160 assert mode
in ['r', 'a', 'w']
2161 self
.f
= io
.open(filename
, mode
, encoding
=encoding
)
2164 def __enter__(self
):
2165 exclusive
= self
.mode
!= 'r'
2167 _lock_file(self
.f
, exclusive
)
2173 def __exit__(self
, etype
, value
, traceback
):
2175 _unlock_file(self
.f
)
2182 def write(self
, *args
):
2183 return self
.f
.write(*args
)
2185 def read(self
, *args
):
2186 return self
.f
.read(*args
)
2189 def get_filesystem_encoding():
2190 encoding
= sys
.getfilesystemencoding()
2191 return encoding
if encoding
is not None else 'utf-8'
2194 def shell_quote(args
):
2196 encoding
= get_filesystem_encoding()
2198 if isinstance(a
, bytes):
2199 # We may get a filename encoded with 'encodeFilename'
2200 a
= a
.decode(encoding
)
2201 quoted_args
.append(compat_shlex_quote(a
))
2202 return ' '.join(quoted_args
)
2205 def smuggle_url(url
, data
):
2206 """ Pass additional data in a URL for internal use. """
2208 url
, idata
= unsmuggle_url(url
, {})
2210 sdata
= compat_urllib_parse_urlencode(
2211 {'__youtubedl_smuggle': json.dumps(data)}
)
2212 return url
+ '#' + sdata
2215 def unsmuggle_url(smug_url
, default
=None):
2216 if '#__youtubedl_smuggle' not in smug_url
:
2217 return smug_url
, default
2218 url
, _
, sdata
= smug_url
.rpartition('#')
2219 jsond
= compat_parse_qs(sdata
)['__youtubedl_smuggle'][0]
2220 data
= json
.loads(jsond
)
2224 def format_decimal_suffix(num
, fmt
='%d%s', *, factor
=1000):
2225 """ Formats numbers with decimal sufixes like K, M, etc """
2226 num
, factor
= float_or_none(num
), float(factor
)
2229 exponent
= 0 if num
== 0 else int(math
.log(num
, factor
))
2230 suffix
= ['', *'kMGTPEZY'][exponent
]
2232 suffix
= {'k': 'Ki', '': ''}
.get(suffix
, f
'{suffix}i')
2233 converted
= num
/ (factor
** exponent
)
2234 return fmt
% (converted
, suffix
)
2237 def format_bytes(bytes):
2238 return format_decimal_suffix(bytes, '%.2f%sB', factor
=1024) or 'N/A'
2241 def lookup_unit_table(unit_table
, s
):
2242 units_re
= '|'.join(re
.escape(u
) for u
in unit_table
)
2244 r
'(?P<num>[0-9]+(?:[,.][0-9]*)?)\s*(?P<unit>%s)\b' % units_re
, s
)
2247 num_str
= m
.group('num').replace(',', '.')
2248 mult
= unit_table
[m
.group('unit')]
2249 return int(float(num_str
) * mult
)
2252 def parse_filesize(s
):
2256 # The lower-case forms are of course incorrect and unofficial,
2257 # but we support those too
2274 'megabytes': 1000 ** 2,
2275 'mebibytes': 1024 ** 2,
2281 'gigabytes': 1000 ** 3,
2282 'gibibytes': 1024 ** 3,
2288 'terabytes': 1000 ** 4,
2289 'tebibytes': 1024 ** 4,
2295 'petabytes': 1000 ** 5,
2296 'pebibytes': 1024 ** 5,
2302 'exabytes': 1000 ** 6,
2303 'exbibytes': 1024 ** 6,
2309 'zettabytes': 1000 ** 7,
2310 'zebibytes': 1024 ** 7,
2316 'yottabytes': 1000 ** 8,
2317 'yobibytes': 1024 ** 8,
2320 return lookup_unit_table(_UNIT_TABLE
, s
)
2327 s
= re
.sub(r
'^[^\d]+\s', '', s
).strip()
2329 if re
.match(r
'^[\d,.]+$', s
):
2330 return str_to_int(s
)
2343 ret
= lookup_unit_table(_UNIT_TABLE
, s
)
2347 mobj
= re
.match(r
'([\d,.]+)(?:$|\s)', s
)
2349 return str_to_int(mobj
.group(1))
2352 def parse_resolution(s
):
2356 mobj
= re
.search(r
'(?<![a-zA-Z0-9])(?P<w>\d+)\s*[xX×,]\s*(?P<h>\d+)(?![a-zA-Z0-9])', s
)
2359 'width': int(mobj
.group('w')),
2360 'height': int(mobj
.group('h')),
2363 mobj
= re
.search(r
'(?<![a-zA-Z0-9])(\d+)[pPiI](?![a-zA-Z0-9])', s
)
2365 return {'height': int(mobj.group(1))}
2367 mobj
= re
.search(r
'\b([48])[kK]\b', s
)
2369 return {'height': int(mobj.group(1)) * 540}
2374 def parse_bitrate(s
):
2375 if not isinstance(s
, compat_str
):
2377 mobj
= re
.search(r
'\b(\d+)\s*kbps', s
)
2379 return int(mobj
.group(1))
2382 def month_by_name(name
, lang
='en'):
2383 """ Return the number of a month by (locale-independently) English name """
2385 month_names
= MONTH_NAMES
.get(lang
, MONTH_NAMES
['en'])
2388 return month_names
.index(name
) + 1
2393 def month_by_abbreviation(abbrev
):
2394 """ Return the number of a month by (locale-independently) English
2398 return [s
[:3] for s
in ENGLISH_MONTH_NAMES
].index(abbrev
) + 1
2403 def fix_xml_ampersands(xml_str
):
2404 """Replace all the '&' by '&' in XML"""
2406 r
'&(?!amp;|lt;|gt;|apos;|quot;|#x[0-9a-fA-F]{,4};|#[0-9]{,4};)',
2411 def setproctitle(title
):
2412 assert isinstance(title
, compat_str
)
2414 # ctypes in Jython is not complete
2415 # http://bugs.jython.org/issue2148
2416 if sys
.platform
.startswith('java'):
2420 libc
= ctypes
.cdll
.LoadLibrary('libc.so.6')
2424 # LoadLibrary in Windows Python 2.7.13 only expects
2425 # a bytestring, but since unicode_literals turns
2426 # every string into a unicode string, it fails.
2428 title_bytes
= title
.encode('utf-8')
2429 buf
= ctypes
.create_string_buffer(len(title_bytes
))
2430 buf
.value
= title_bytes
2432 libc
.prctl(15, buf
, 0, 0, 0)
2433 except AttributeError:
2434 return # Strange libc, just skip this
2437 def remove_start(s
, start
):
2438 return s
[len(start
):] if s
is not None and s
.startswith(start
) else s
2441 def remove_end(s
, end
):
2442 return s
[:-len(end
)] if s
is not None and s
.endswith(end
) else s
2445 def remove_quotes(s
):
2446 if s
is None or len(s
) < 2:
2448 for quote
in ('"', "'", ):
2449 if s
[0] == quote
and s
[-1] == quote
:
2454 def get_domain(url
):
2455 domain
= re
.match(r
'(?:https?:\/\/)?(?:www\.)?(?P<domain>[^\n\/]+\.[^\n\/]+)(?:\/(.*))?', url
)
2456 return domain
.group('domain') if domain
else None
2459 def url_basename(url
):
2460 path
= compat_urlparse
.urlparse(url
).path
2461 return path
.strip('/').split('/')[-1]
2465 return re
.match(r
'https?://[^?#&]+/', url
).group()
2468 def urljoin(base
, path
):
2469 if isinstance(path
, bytes):
2470 path
= path
.decode('utf-8')
2471 if not isinstance(path
, compat_str
) or not path
:
2473 if re
.match(r
'^(?:[a-zA-Z][a-zA-Z0-9+-.]*:)?//', path
):
2475 if isinstance(base
, bytes):
2476 base
= base
.decode('utf-8')
2477 if not isinstance(base
, compat_str
) or not re
.match(
2478 r
'^(?:https?:)?//', base
):
2480 return compat_urlparse
.urljoin(base
, path
)
2483 class HEADRequest(compat_urllib_request
.Request
):
2484 def get_method(self
):
2488 class PUTRequest(compat_urllib_request
.Request
):
2489 def get_method(self
):
2493 def int_or_none(v
, scale
=1, default
=None, get_attr
=None, invscale
=1):
2494 if get_attr
and v
is not None:
2495 v
= getattr(v
, get_attr
, None)
2497 return int(v
) * invscale
// scale
2498 except (ValueError, TypeError, OverflowError):
2502 def str_or_none(v
, default
=None):
2503 return default
if v
is None else compat_str(v
)
2506 def str_to_int(int_str
):
2507 """ A more relaxed version of int_or_none """
2508 if isinstance(int_str
, compat_integer_types
):
2510 elif isinstance(int_str
, compat_str
):
2511 int_str
= re
.sub(r
'[,\.\+]', '', int_str
)
2512 return int_or_none(int_str
)
2515 def float_or_none(v
, scale
=1, invscale
=1, default
=None):
2519 return float(v
) * invscale
/ scale
2520 except (ValueError, TypeError):
2524 def bool_or_none(v
, default
=None):
2525 return v
if isinstance(v
, bool) else default
2528 def strip_or_none(v
, default
=None):
2529 return v
.strip() if isinstance(v
, compat_str
) else default
2532 def url_or_none(url
):
2533 if not url
or not isinstance(url
, compat_str
):
2536 return url
if re
.match(r
'^(?:(?:https?|rt(?:m(?:pt?[es]?|fp)|sp[su]?)|mms|ftps?):)?//', url
) else None
2539 def strftime_or_none(timestamp
, date_format
, default
=None):
2540 datetime_object
= None
2542 if isinstance(timestamp
, compat_numeric_types
): # unix timestamp
2543 datetime_object
= datetime
.datetime
.utcfromtimestamp(timestamp
)
2544 elif isinstance(timestamp
, compat_str
): # assume YYYYMMDD
2545 datetime_object
= datetime
.datetime
.strptime(timestamp
, '%Y%m%d')
2546 return datetime_object
.strftime(date_format
)
2547 except (ValueError, TypeError, AttributeError):
2551 def parse_duration(s
):
2552 if not isinstance(s
, compat_basestring
):
2558 days
, hours
, mins
, secs
, ms
= [None] * 5
2559 m
= re
.match(r
'(?:(?:(?:(?P<days>[0-9]+):)?(?P<hours>[0-9]+):)?(?P<mins>[0-9]+):)?(?P<secs>[0-9]+)(?P<ms>\.[0-9]+)?Z?$', s
)
2561 days
, hours
, mins
, secs
, ms
= m
.groups()
2566 [0-9]+\s*y(?:ears?)?\s*
2569 [0-9]+\s*m(?:onths?)?\s*
2572 [0-9]+\s*w(?:eeks?)?\s*
2575 (?P<days>[0-9]+)\s*d(?:ays?)?\s*
2579 (?P<hours>[0-9]+)\s*h(?:ours?)?\s*
2582 (?P<mins>[0-9]+)\s*m(?:in(?:ute)?s?)?\s*
2585 (?P<secs>[0-9]+)(?P<ms>\.[0-9]+)?\s*s(?:ec(?:ond)?s?)?\s*
2588 days
, hours
, mins
, secs
, ms
= m
.groups()
2590 m
= re
.match(r
'(?i)(?:(?P<hours>[0-9.]+)\s*(?:hours?)|(?P<mins>[0-9.]+)\s*(?:mins?\.?|minutes?)\s*)Z?$', s
)
2592 hours
, mins
= m
.groups()
2598 duration
+= float(secs
)
2600 duration
+= float(mins
) * 60
2602 duration
+= float(hours
) * 60 * 60
2604 duration
+= float(days
) * 24 * 60 * 60
2606 duration
+= float(ms
)
2610 def prepend_extension(filename
, ext
, expected_real_ext
=None):
2611 name
, real_ext
= os
.path
.splitext(filename
)
2613 '{0}.{1}{2}'.format(name
, ext
, real_ext
)
2614 if not expected_real_ext
or real_ext
[1:] == expected_real_ext
2615 else '{0}.{1}'.format(filename
, ext
))
2618 def replace_extension(filename
, ext
, expected_real_ext
=None):
2619 name
, real_ext
= os
.path
.splitext(filename
)
2620 return '{0}.{1}'.format(
2621 name
if not expected_real_ext
or real_ext
[1:] == expected_real_ext
else filename
,
2625 def check_executable(exe
, args
=[]):
2626 """ Checks if the given binary is installed somewhere in PATH, and returns its name.
2627 args can be a list of arguments for a short output (like -version) """
2629 Popen([exe
] + args
, stdout
=subprocess
.PIPE
, stderr
=subprocess
.PIPE
).communicate_or_kill()
2635 def _get_exe_version_output(exe
, args
):
2637 # STDIN should be redirected too. On UNIX-like systems, ffmpeg triggers
2638 # SIGTTOU if yt-dlp is run in the background.
2639 # See https://github.com/ytdl-org/youtube-dl/issues/955#issuecomment-209789656
2641 [encodeArgument(exe
)] + args
, stdin
=subprocess
.PIPE
,
2642 stdout
=subprocess
.PIPE
, stderr
=subprocess
.STDOUT
).communicate_or_kill()
2645 if isinstance(out
, bytes): # Python 2.x
2646 out
= out
.decode('ascii', 'ignore')
2650 def detect_exe_version(output
, version_re
=None, unrecognized
='present'):
2651 assert isinstance(output
, compat_str
)
2652 if version_re
is None:
2653 version_re
= r
'version\s+([-0-9._a-zA-Z]+)'
2654 m
= re
.search(version_re
, output
)
2661 def get_exe_version(exe
, args
=['--version'],
2662 version_re
=None, unrecognized
='present'):
2663 """ Returns the version of the specified executable,
2664 or False if the executable is not present """
2665 out
= _get_exe_version_output(exe
, args
)
2666 return detect_exe_version(out
, version_re
, unrecognized
) if out
else False
2669 class LazyList(collections
.abc
.Sequence
):
2670 ''' Lazy immutable list from an iterable
2671 Note that slices of a LazyList are lists and not LazyList'''
2673 class IndexError(IndexError):
2676 def __init__(self
, iterable
, *, reverse
=False, _cache
=None):
2677 self
.__iterable
= iter(iterable
)
2678 self
.__cache
= [] if _cache
is None else _cache
2679 self
.__reversed
= reverse
2683 # We need to consume the entire iterable to iterate in reverse
2684 yield from self
.exhaust()
2686 yield from self
.__cache
2687 for item
in self
.__iterable
:
2688 self
.__cache
.append(item
)
2691 def __exhaust(self
):
2692 self
.__cache
.extend(self
.__iterable
)
2693 # Discard the emptied iterable to make it pickle-able
2694 self
.__iterable
= []
2698 ''' Evaluate the entire iterable '''
2699 return self
.__exhaust
()[::-1 if self
.__reversed
else 1]
2702 def __reverse_index(x
):
2703 return None if x
is None else -(x
+ 1)
2705 def __getitem__(self
, idx
):
2706 if isinstance(idx
, slice):
2708 idx
= slice(self
.__reverse
_index
(idx
.start
), self
.__reverse
_index
(idx
.stop
), -(idx
.step
or 1))
2709 start
, stop
, step
= idx
.start
, idx
.stop
, idx
.step
or 1
2710 elif isinstance(idx
, int):
2712 idx
= self
.__reverse
_index
(idx
)
2713 start
, stop
, step
= idx
, idx
, 0
2715 raise TypeError('indices must be integers or slices')
2716 if ((start
or 0) < 0 or (stop
or 0) < 0
2717 or (start
is None and step
< 0)
2718 or (stop
is None and step
> 0)):
2719 # We need to consume the entire iterable to be able to slice from the end
2720 # Obviously, never use this with infinite iterables
2723 return self
.__cache
[idx
]
2724 except IndexError as e
:
2725 raise self
.IndexError(e
) from e
2726 n
= max(start
or 0, stop
or 0) - len(self
.__cache
) + 1
2728 self
.__cache
.extend(itertools
.islice(self
.__iterable
, n
))
2730 return self
.__cache
[idx
]
2731 except IndexError as e
:
2732 raise self
.IndexError(e
) from e
2736 self
[-1] if self
.__reversed
else self
[0]
2737 except self
.IndexError:
2743 return len(self
.__cache
)
2745 def __reversed__(self
):
2746 return type(self
)(self
.__iterable
, reverse
=not self
.__reversed
, _cache
=self
.__cache
)
2749 return type(self
)(self
.__iterable
, reverse
=self
.__reversed
, _cache
=self
.__cache
)
2752 # repr and str should mimic a list. So we exhaust the iterable
2753 return repr(self
.exhaust())
2756 return repr(self
.exhaust())
2761 class IndexError(IndexError):
2765 # This is only useful for tests
2766 return len(self
.getslice())
2768 def __init__(self
, pagefunc
, pagesize
, use_cache
=True):
2769 self
._pagefunc
= pagefunc
2770 self
._pagesize
= pagesize
2771 self
._use
_cache
= use_cache
2774 def getpage(self
, pagenum
):
2775 page_results
= self
._cache
.get(pagenum
)
2776 if page_results
is None:
2777 page_results
= list(self
._pagefunc
(pagenum
))
2779 self
._cache
[pagenum
] = page_results
2782 def getslice(self
, start
=0, end
=None):
2783 return list(self
._getslice
(start
, end
))
2785 def _getslice(self
, start
, end
):
2786 raise NotImplementedError('This method must be implemented by subclasses')
2788 def __getitem__(self
, idx
):
2789 # NOTE: cache must be enabled if this is used
2790 if not isinstance(idx
, int) or idx
< 0:
2791 raise TypeError('indices must be non-negative integers')
2792 entries
= self
.getslice(idx
, idx
+ 1)
2794 raise self
.IndexError()
2798 class OnDemandPagedList(PagedList
):
2799 def _getslice(self
, start
, end
):
2800 for pagenum
in itertools
.count(start
// self
._pagesize
):
2801 firstid
= pagenum
* self
._pagesize
2802 nextfirstid
= pagenum
* self
._pagesize
+ self
._pagesize
2803 if start
>= nextfirstid
:
2807 start
% self
._pagesize
2808 if firstid
<= start
< nextfirstid
2811 ((end
- 1) % self
._pagesize
) + 1
2812 if (end
is not None and firstid
<= end
<= nextfirstid
)
2815 page_results
= self
.getpage(pagenum
)
2816 if startv
!= 0 or endv
is not None:
2817 page_results
= page_results
[startv
:endv
]
2818 yield from page_results
2820 # A little optimization - if current page is not "full", ie. does
2821 # not contain page_size videos then we can assume that this page
2822 # is the last one - there are no more ids on further pages -
2823 # i.e. no need to query again.
2824 if len(page_results
) + startv
< self
._pagesize
:
2827 # If we got the whole page, but the next page is not interesting,
2828 # break out early as well
2829 if end
== nextfirstid
:
2833 class InAdvancePagedList(PagedList
):
2834 def __init__(self
, pagefunc
, pagecount
, pagesize
):
2835 self
._pagecount
= pagecount
2836 PagedList
.__init
__(self
, pagefunc
, pagesize
, True)
2838 def _getslice(self
, start
, end
):
2839 start_page
= start
// self
._pagesize
2841 self
._pagecount
if end
is None else (end
// self
._pagesize
+ 1))
2842 skip_elems
= start
- start_page
* self
._pagesize
2843 only_more
= None if end
is None else end
- start
2844 for pagenum
in range(start_page
, end_page
):
2845 page_results
= self
.getpage(pagenum
)
2847 page_results
= page_results
[skip_elems
:]
2849 if only_more
is not None:
2850 if len(page_results
) < only_more
:
2851 only_more
-= len(page_results
)
2853 yield from page_results
[:only_more
]
2855 yield from page_results
2858 def uppercase_escape(s
):
2859 unicode_escape
= codecs
.getdecoder('unicode_escape')
2861 r
'\\U[0-9a-fA-F]{8}',
2862 lambda m
: unicode_escape(m
.group(0))[0],
2866 def lowercase_escape(s
):
2867 unicode_escape
= codecs
.getdecoder('unicode_escape')
2869 r
'\\u[0-9a-fA-F]{4}',
2870 lambda m
: unicode_escape(m
.group(0))[0],
2874 def escape_rfc3986(s
):
2875 """Escape non-ASCII characters as suggested by RFC 3986"""
2876 if sys
.version_info
< (3, 0) and isinstance(s
, compat_str
):
2877 s
= s
.encode('utf-8')
2878 return compat_urllib_parse
.quote(s
, b
"%/;:@&=+$,!~*'()?#[]")
2881 def escape_url(url
):
2882 """Escape URL as suggested by RFC 3986"""
2883 url_parsed
= compat_urllib_parse_urlparse(url
)
2884 return url_parsed
._replace
(
2885 netloc
=url_parsed
.netloc
.encode('idna').decode('ascii'),
2886 path
=escape_rfc3986(url_parsed
.path
),
2887 params
=escape_rfc3986(url_parsed
.params
),
2888 query
=escape_rfc3986(url_parsed
.query
),
2889 fragment
=escape_rfc3986(url_parsed
.fragment
)
2894 return compat_parse_qs(compat_urllib_parse_urlparse(url
).query
)
2897 def read_batch_urls(batch_fd
):
2899 if not isinstance(url
, compat_str
):
2900 url
= url
.decode('utf-8', 'replace')
2901 BOM_UTF8
= ('\xef\xbb\xbf', '\ufeff')
2902 for bom
in BOM_UTF8
:
2903 if url
.startswith(bom
):
2904 url
= url
[len(bom
):]
2906 if not url
or url
.startswith(('#', ';', ']')):
2908 # "#" cannot be stripped out since it is part of the URI
2909 # However, it can be safely stipped out if follwing a whitespace
2910 return re
.split(r
'\s#', url
, 1)[0].rstrip()
2912 with contextlib
.closing(batch_fd
) as fd
:
2913 return [url
for url
in map(fixup
, fd
) if url
]
2916 def urlencode_postdata(*args
, **kargs
):
2917 return compat_urllib_parse_urlencode(*args
, **kargs
).encode('ascii')
2920 def update_url_query(url
, query
):
2923 parsed_url
= compat_urlparse
.urlparse(url
)
2924 qs
= compat_parse_qs(parsed_url
.query
)
2926 return compat_urlparse
.urlunparse(parsed_url
._replace
(
2927 query
=compat_urllib_parse_urlencode(qs
, True)))
2930 def update_Request(req
, url
=None, data
=None, headers
={}, query={}
):
2931 req_headers
= req
.headers
.copy()
2932 req_headers
.update(headers
)
2933 req_data
= data
or req
.data
2934 req_url
= update_url_query(url
or req
.get_full_url(), query
)
2935 req_get_method
= req
.get_method()
2936 if req_get_method
== 'HEAD':
2937 req_type
= HEADRequest
2938 elif req_get_method
== 'PUT':
2939 req_type
= PUTRequest
2941 req_type
= compat_urllib_request
.Request
2943 req_url
, data
=req_data
, headers
=req_headers
,
2944 origin_req_host
=req
.origin_req_host
, unverifiable
=req
.unverifiable
)
2945 if hasattr(req
, 'timeout'):
2946 new_req
.timeout
= req
.timeout
2950 def _multipart_encode_impl(data
, boundary
):
2951 content_type
= 'multipart/form-data; boundary=%s' % boundary
2954 for k
, v
in data
.items():
2955 out
+= b
'--' + boundary
.encode('ascii') + b
'\r\n'
2956 if isinstance(k
, compat_str
):
2957 k
= k
.encode('utf-8')
2958 if isinstance(v
, compat_str
):
2959 v
= v
.encode('utf-8')
2960 # RFC 2047 requires non-ASCII field names to be encoded, while RFC 7578
2961 # suggests sending UTF-8 directly. Firefox sends UTF-8, too
2962 content
= b
'Content-Disposition: form-data; name="' + k
+ b
'"\r\n\r\n' + v
+ b
'\r\n'
2963 if boundary
.encode('ascii') in content
:
2964 raise ValueError('Boundary overlaps with data')
2967 out
+= b
'--' + boundary
.encode('ascii') + b
'--\r\n'
2969 return out
, content_type
2972 def multipart_encode(data
, boundary
=None):
2974 Encode a dict to RFC 7578-compliant form-data
2977 A dict where keys and values can be either Unicode or bytes-like
2980 If specified a Unicode object, it's used as the boundary. Otherwise
2981 a random boundary is generated.
2983 Reference: https://tools.ietf.org/html/rfc7578
2985 has_specified_boundary
= boundary
is not None
2988 if boundary
is None:
2989 boundary
= '---------------' + str(random
.randrange(0x0fffffff, 0xffffffff))
2992 out
, content_type
= _multipart_encode_impl(data
, boundary
)
2995 if has_specified_boundary
:
2999 return out
, content_type
3002 def dict_get(d
, key_or_keys
, default
=None, skip_false_values
=True):
3003 if isinstance(key_or_keys
, (list, tuple)):
3004 for key
in key_or_keys
:
3005 if key
not in d
or d
[key
] is None or skip_false_values
and not d
[key
]:
3009 return d
.get(key_or_keys
, default
)
3012 def try_get(src
, getter
, expected_type
=None):
3013 for get
in variadic(getter
):
3016 except (AttributeError, KeyError, TypeError, IndexError):
3019 if expected_type
is None or isinstance(v
, expected_type
):
3023 def merge_dicts(*dicts
):
3025 for a_dict
in dicts
:
3026 for k
, v
in a_dict
.items():
3030 or (isinstance(v
, compat_str
) and v
3031 and isinstance(merged
[k
], compat_str
)
3032 and not merged
[k
])):
3037 def encode_compat_str(string
, encoding
=preferredencoding(), errors
='strict'):
3038 return string
if isinstance(string
, compat_str
) else compat_str(string
, encoding
, errors
)
3050 TV_PARENTAL_GUIDELINES
= {
3060 def parse_age_limit(s
):
3062 return s
if 0 <= s
<= 21 else None
3063 if not isinstance(s
, compat_basestring
):
3065 m
= re
.match(r
'^(?P<age>\d{1,2})\+?$', s
)
3067 return int(m
.group('age'))
3070 return US_RATINGS
[s
]
3071 m
= re
.match(r
'^TV[_-]?(%s)$' % '|'.join(k
[3:] for k
in TV_PARENTAL_GUIDELINES
), s
)
3073 return TV_PARENTAL_GUIDELINES
['TV-' + m
.group(1)]
3077 def strip_jsonp(code
):
3080 (?:window\.)?(?P<func_name>[a-zA-Z0-9_.$]*)
3081 (?:\s*&&\s*(?P=func_name))?
3082 \s*\(\s*(?P<callback_data>.*)\);?
3083 \s*?(?://[^\n]*)*$''',
3084 r
'\g<callback_data>', code
)
3087 def js_to_json(code
, vars={}):
3088 # vars is a dict of var, val pairs to substitute
3089 COMMENT_RE
= r
'/\*(?:(?!\*/).)*?\*/|//[^\n]*\n'
3090 SKIP_RE
= r
'\s*(?:{comment})?\s*'.format(comment
=COMMENT_RE
)
3092 (r
'(?s)^(0[xX][0-9a-fA-F]+){skip}:?$'.format(skip
=SKIP_RE
), 16),
3093 (r
'(?s)^(0+[0-7]+){skip}:?$'.format(skip
=SKIP_RE
), 8),
3098 if v
in ('true', 'false', 'null'):
3100 elif v
in ('undefined', 'void 0'):
3102 elif v
.startswith('/*') or v
.startswith('//') or v
.startswith('!') or v
== ',':
3105 if v
[0] in ("'", '"'):
3106 v
= re
.sub(r
'(?s)\\.|"', lambda m
: {
3111 }.get(m
.group(0), m
.group(0)), v
[1:-1])
3113 for regex
, base
in INTEGER_TABLE
:
3114 im
= re
.match(regex
, v
)
3116 i
= int(im
.group(1), base
)
3117 return '"%d":' % i
if v
.endswith(':') else '%d' % i
3124 return re
.sub(r
'''(?sx)
3125 "(?:[^"\\]*(?:\\\\|\\['"nurtbfx/\n]))*[^"\\]*"|
3126 '(?:[^'\\]*(?:\\\\|\\['"nurtbfx/\n]))*[^'\\]*'|
3127 {comment}|,(?={skip}[\]}}])|
3128 void\s0|(?:(?<![0-9])[eE]|[a-df-zA-DF-Z_$])[.a-zA-Z_$0-9]*|
3129 \b(?:0[xX][0-9a-fA-F]+|0+[0-7]+)(?:{skip}:)?|
3132 '''.format(comment
=COMMENT_RE
, skip
=SKIP_RE
), fix_kv
, code
)
3135 def qualities(quality_ids
):
3136 """ Get a numeric quality value out of a list of possible values """
3139 return quality_ids
.index(qid
)
3145 POSTPROCESS_WHEN
= {'pre_process', 'before_dl', 'after_move', 'post_process', 'after_video', 'playlist'}
3149 'default': '%(title)s [%(id)s].%(ext)s',
3150 'chapter': '%(title)s - %(section_number)03d %(section_title)s [%(id)s].%(ext)s',
3156 'description': 'description',
3157 'annotation': 'annotations.xml',
3158 'infojson': 'info.json',
3160 'pl_thumbnail': None,
3161 'pl_description': 'description',
3162 'pl_infojson': 'info.json',
3165 # As of [1] format syntax is:
3166 # %[mapping_key][conversion_flags][minimum_width][.precision][length_modifier]type
3167 # 1. https://docs.python.org/2/library/stdtypes.html#string-formatting
3168 STR_FORMAT_RE_TMPL
= r
'''(?x)
3169 (?<!%)(?P<prefix>(?:%%)*)
3171 (?P<has_key>\((?P<key>{0})\))?
3173 (?P<conversion>[#0\-+ ]+)?
3175 (?P<precision>\.\d+)?
3176 (?P<len_mod>[hlL])? # unused in python
3177 {1} # conversion type
3182 STR_FORMAT_TYPES
= 'diouxXeEfFgGcrs'
3185 def limit_length(s
, length
):
3186 """ Add ellipses to overly long strings """
3191 return s
[:length
- len(ELLIPSES
)] + ELLIPSES
3195 def version_tuple(v
):
3196 return tuple(int(e
) for e
in re
.split(r
'[-.]', v
))
3199 def is_outdated_version(version
, limit
, assume_new
=True):
3201 return not assume_new
3203 return version_tuple(version
) < version_tuple(limit
)
3205 return not assume_new
3208 def ytdl_is_updateable():
3209 """ Returns if yt-dlp can be updated with -U """
3211 from .update
import is_non_updateable
3213 return not is_non_updateable()
3216 def args_to_str(args
):
3217 # Get a short string representation for a subprocess command
3218 return ' '.join(compat_shlex_quote(a
) for a
in args
)
3221 def error_to_compat_str(err
):
3223 # On python 2 error byte string must be decoded with proper
3224 # encoding rather than ascii
3225 if sys
.version_info
[0] < 3:
3226 err_str
= err_str
.decode(preferredencoding())
3230 def mimetype2ext(mt
):
3234 mt
, _
, params
= mt
.partition(';')
3239 # Per RFC 3003, audio/mpeg can be .mp1, .mp2 or .mp3. Here use .mp3 as
3240 # it's the most popular one
3241 'audio/mpeg': 'mp3',
3242 'audio/x-wav': 'wav',
3244 'audio/wave': 'wav',
3247 ext
= FULL_MAP
.get(mt
)
3253 'smptett+xml': 'tt',
3257 'x-mp4-fragmented': 'mp4',
3258 'x-ms-sami': 'sami',
3261 'x-mpegurl': 'm3u8',
3262 'vnd.apple.mpegurl': 'm3u8',
3266 'vnd.ms-sstr+xml': 'ism',
3270 'filmstrip+json': 'fs',
3274 _
, _
, subtype
= mt
.rpartition('/')
3275 ext
= SUBTYPE_MAP
.get(subtype
.lower())
3286 _
, _
, suffix
= subtype
.partition('+')
3287 ext
= SUFFIX_MAP
.get(suffix
)
3291 return subtype
.replace('+', '.')
3294 def ext2mimetype(ext_or_url
):
3297 if '.' not in ext_or_url
:
3298 ext_or_url
= f
'file.{ext_or_url}'
3299 return mimetypes
.guess_type(ext_or_url
)[0]
3302 def parse_codecs(codecs_str
):
3303 # http://tools.ietf.org/html/rfc6381
3306 split_codecs
= list(filter(None, map(
3307 str.strip
, codecs_str
.strip().strip(',').split(','))))
3308 vcodec
, acodec
, tcodec
, hdr
= None, None, None, None
3309 for full_codec
in split_codecs
:
3310 parts
= full_codec
.split('.')
3311 codec
= parts
[0].replace('0', '')
3312 if codec
in ('avc1', 'avc2', 'avc3', 'avc4', 'vp9', 'vp8', 'hev1', 'hev2',
3313 'h263', 'h264', 'mp4v', 'hvc1', 'av1', 'theora', 'dvh1', 'dvhe'):
3315 vcodec
= '.'.join(parts
[:4]) if codec
in ('vp9', 'av1', 'hvc1') else full_codec
3316 if codec
in ('dvh1', 'dvhe'):
3318 elif codec
== 'av1' and len(parts
) > 3 and parts
[3] == '10':
3320 elif full_codec
.replace('0', '').startswith('vp9.2'):
3322 elif codec
in ('flac', 'mp4a', 'opus', 'vorbis', 'mp3', 'aac', 'ac-3', 'ec-3', 'eac3', 'dtsc', 'dtse', 'dtsh', 'dtsl'):
3325 elif codec
in ('stpp', 'wvtt',):
3329 write_string('WARNING: Unknown codec %s\n' % full_codec
, sys
.stderr
)
3330 if vcodec
or acodec
or tcodec
:
3332 'vcodec': vcodec
or 'none',
3333 'acodec': acodec
or 'none',
3334 'dynamic_range': hdr
,
3335 **({'tcodec': tcodec}
if tcodec
is not None else {}),
3337 elif len(split_codecs
) == 2:
3339 'vcodec': split_codecs
[0],
3340 'acodec': split_codecs
[1],
3345 def urlhandle_detect_ext(url_handle
):
3346 getheader
= url_handle
.headers
.get
3348 cd
= getheader('Content-Disposition')
3350 m
= re
.match(r
'attachment;\s*filename="(?P<filename>[^"]+)"', cd
)
3352 e
= determine_ext(m
.group('filename'), default_ext
=None)
3356 return mimetype2ext(getheader('Content-Type'))
3359 def encode_data_uri(data
, mime_type
):
3360 return 'data:%s;base64,%s' % (mime_type
, base64
.b64encode(data
).decode('ascii'))
3363 def age_restricted(content_limit
, age_limit
):
3364 """ Returns True iff the content should be blocked """
3366 if age_limit
is None: # No limit set
3368 if content_limit
is None:
3369 return False # Content available for everyone
3370 return age_limit
< content_limit
3373 def is_html(first_bytes
):
3374 """ Detect whether a file contains HTML by examining its first bytes. """
3377 (b
'\xef\xbb\xbf', 'utf-8'),
3378 (b
'\x00\x00\xfe\xff', 'utf-32-be'),
3379 (b
'\xff\xfe\x00\x00', 'utf-32-le'),
3380 (b
'\xff\xfe', 'utf-16-le'),
3381 (b
'\xfe\xff', 'utf-16-be'),
3383 for bom
, enc
in BOMS
:
3384 if first_bytes
.startswith(bom
):
3385 s
= first_bytes
[len(bom
):].decode(enc
, 'replace')
3388 s
= first_bytes
.decode('utf-8', 'replace')
3390 return re
.match(r
'^\s*<', s
)
3393 def determine_protocol(info_dict
):
3394 protocol
= info_dict
.get('protocol')
3395 if protocol
is not None:
3398 url
= sanitize_url(info_dict
['url'])
3399 if url
.startswith('rtmp'):
3401 elif url
.startswith('mms'):
3403 elif url
.startswith('rtsp'):
3406 ext
= determine_ext(url
)
3412 return compat_urllib_parse_urlparse(url
).scheme
3415 def render_table(header_row
, data
, delim
=False, extra_gap
=0, hide_empty
=False):
3416 """ Render a list of rows, each as a list of values.
3417 Text after a \t will be right aligned """
3419 return len(remove_terminal_sequences(string
).replace('\t', ''))
3421 def get_max_lens(table
):
3422 return [max(width(str(v
)) for v
in col
) for col
in zip(*table
)]
3424 def filter_using_list(row
, filterArray
):
3425 return [col
for (take
, col
) in zip(filterArray
, row
) if take
]
3428 max_lens
= get_max_lens(data
)
3429 header_row
= filter_using_list(header_row
, max_lens
)
3430 data
= [filter_using_list(row
, max_lens
) for row
in data
]
3432 table
= [header_row
] + data
3433 max_lens
= get_max_lens(table
)
3436 table
= [header_row
, [delim
* (ml
+ extra_gap
) for ml
in max_lens
]] + data
3437 table
[1][-1] = table
[1][-1][:-extra_gap
] # Remove extra_gap from end of delimiter
3439 for pos
, text
in enumerate(map(str, row
)):
3441 row
[pos
] = text
.replace('\t', ' ' * (max_lens
[pos
] - width(text
))) + ' ' * extra_gap
3443 row
[pos
] = text
+ ' ' * (max_lens
[pos
] - width(text
) + extra_gap
)
3444 ret
= '\n'.join(''.join(row
).rstrip() for row
in table
)
3448 def _match_one(filter_part
, dct
, incomplete
):
3449 # TODO: Generalize code with YoutubeDL._build_format_filter
3450 STRING_OPERATORS
= {
3451 '*=': operator
.contains
,
3452 '^=': lambda attr
, value
: attr
.startswith(value
),
3453 '$=': lambda attr
, value
: attr
.endswith(value
),
3454 '~=': lambda attr
, value
: re
.search(value
, attr
),
3456 COMPARISON_OPERATORS
= {
3458 '<=': operator
.le
, # "<=" must be defined above "<"
3465 operator_rex
= re
.compile(r
'''(?x)\s*
3467 \s*(?P<negation>!\s*)?(?P<op>%s)(?P<none_inclusive>\s*\?)?\s*
3469 (?P<quote>["\'])(?P
<quotedstrval
>.+?
)(?P
=quote
)|
3473 ''' % '|'.join(map(re.escape, COMPARISON_OPERATORS.keys())))
3474 m = operator_rex.search(filter_part)
3477 unnegated_op = COMPARISON_OPERATORS[m['op']]
3479 op = lambda attr, value: not unnegated_op(attr, value)
3482 comparison_value = m['quotedstrval'] or m['strval'] or m['intval']
3484 comparison_value = comparison_value.replace(r'\%s' % m['quote'], m['quote'])
3485 actual_value = dct.get(m['key'])
3486 numeric_comparison = None
3487 if isinstance(actual_value, compat_numeric_types):
3488 # If the original field is a string and matching comparisonvalue is
3489 # a number we should respect the origin of the original field
3490 # and process comparison value as a string (see
3491 # https://github.com/ytdl-org/youtube-dl/issues/11082)
3493 numeric_comparison = int(comparison_value)
3495 numeric_comparison = parse_filesize(comparison_value)
3496 if numeric_comparison is None:
3497 numeric_comparison = parse_filesize(f'{comparison_value}B')
3498 if numeric_comparison is None:
3499 numeric_comparison = parse_duration(comparison_value)
3500 if numeric_comparison is not None and m['op'] in STRING_OPERATORS:
3501 raise ValueError('Operator %s only supports string values!' % m['op'])
3502 if actual_value is None:
3503 return incomplete or m['none_inclusive']
3504 return op(actual_value, comparison_value if numeric_comparison is None else numeric_comparison)
3507 '': lambda v: (v is True) if isinstance(v, bool) else (v is not None),
3508 '!': lambda v: (v is False) if isinstance(v, bool) else (v is None),
3510 operator_rex = re.compile(r'''(?x
)\s
*
3511 (?P
<op
>%s)\s
*(?P
<key
>[a
-z_
]+)
3513 ''' % '|'.join(map(re.escape, UNARY_OPERATORS.keys())))
3514 m = operator_rex.search(filter_part)
3516 op = UNARY_OPERATORS[m.group('op')]
3517 actual_value = dct.get(m.group('key'))
3518 if incomplete and actual_value is None:
3520 return op(actual_value)
3522 raise ValueError('Invalid filter part %r' % filter_part)
3525 def match_str(filter_str, dct, incomplete=False):
3526 """ Filter a dictionary with a simple string syntax. Returns True (=passes filter) or false
3527 When incomplete, all conditions passes on missing fields
3530 _match_one(filter_part.replace(r'\&', '&'), dct, incomplete)
3531 for filter_part in re.split(r'(?<!\\)&', filter_str))
3534 def match_filter_func(filter_str):
3535 def _match_func(info_dict, *args, **kwargs):
3536 if match_str(filter_str, info_dict, *args, **kwargs):
3539 video_title = info_dict.get('title', info_dict.get('id', 'video'))
3540 return '%s does not pass filter %s, skipping ..' % (video_title, filter_str)
3544 def parse_dfxp_time_expr(time_expr):
3548 mobj = re.match(r'^(?P<time_offset>\d+(?:\.\d+)?)s?$', time_expr)
3550 return float(mobj.group('time_offset'))
3552 mobj = re.match(r'^(\d+):(\d\d):(\d\d(?:(?:\.|:)\d+)?)$', time_expr)
3554 return 3600 * int(mobj.group(1)) + 60 * int(mobj.group(2)) + float(mobj.group(3).replace(':', '.'))
3557 def srt_subtitles_timecode(seconds):
3558 return '%02d:%02d:%02d,%03d' % timetuple_from_msec(seconds * 1000)
3561 def ass_subtitles_timecode(seconds):
3562 time = timetuple_from_msec(seconds * 1000)
3563 return '%01d:%02d:%02d.%02d' % (*time[:-1], time.milliseconds / 10)
3566 def dfxp2srt(dfxp_data):
3568 @param dfxp_data A
bytes-like
object containing DFXP data
3569 @returns A
unicode object containing converted SRT data
3571 LEGACY_NAMESPACES = (
3572 (b'http://www.w3.org/ns/ttml', [
3573 b'http://www.w3.org/2004/11/ttaf1',
3574 b'http://www.w3.org/2006/04/ttaf1',
3575 b'http://www.w3.org/2006/10/ttaf1',
3577 (b'http://www.w3.org/ns/ttml#styling', [
3578 b'http://www.w3.org/ns/ttml#style',
3582 SUPPORTED_STYLING = [
3591 _x = functools.partial(xpath_with_ns, ns_map={
3592 'xml': 'http://www.w3.org/XML/1998/namespace',
3593 'ttml': 'http://www.w3.org/ns/ttml',
3594 'tts': 'http://www.w3.org/ns/ttml#styling',
3600 class TTMLPElementParser(object):
3602 _unclosed_elements = []
3603 _applied_styles = []
3605 def start(self, tag, attrib):
3606 if tag in (_x('ttml:br'), 'br'):
3609 unclosed_elements = []
3611 element_style_id = attrib.get('style')
3613 style.update(default_style)
3614 if element_style_id:
3615 style.update(styles.get(element_style_id, {}))
3616 for prop in SUPPORTED_STYLING:
3617 prop_val = attrib.get(_x('tts:' + prop))
3619 style[prop] = prop_val
3622 for k, v in sorted(style.items()):
3623 if self._applied_styles and self._applied_styles[-1].get(k) == v:
3626 font += ' color="%s"' % v
3627 elif k == 'fontSize':
3628 font += ' size="%s"' % v
3629 elif k == 'fontFamily':
3630 font += ' face="%s"' % v
3631 elif k == 'fontWeight' and v == 'bold':
3633 unclosed_elements.append('b')
3634 elif k == 'fontStyle' and v == 'italic':
3636 unclosed_elements.append('i')
3637 elif k == 'textDecoration' and v == 'underline':
3639 unclosed_elements.append('u')
3641 self._out += '<font' + font + '>'
3642 unclosed_elements.append('font')
3644 if self._applied_styles:
3645 applied_style.update(self._applied_styles[-1])
3646 applied_style.update(style)
3647 self._applied_styles.append(applied_style)
3648 self._unclosed_elements.append(unclosed_elements)
3651 if tag not in (_x('ttml:br'), 'br'):
3652 unclosed_elements = self._unclosed_elements.pop()
3653 for element in reversed(unclosed_elements):
3654 self._out += '</%s>' % element
3655 if unclosed_elements and self._applied_styles:
3656 self._applied_styles.pop()
3658 def data(self, data):
3662 return self._out.strip()
3664 def parse_node(node):
3665 target = TTMLPElementParser()
3666 parser = xml.etree.ElementTree.XMLParser(target=target)
3667 parser.feed(xml.etree.ElementTree.tostring(node))
3668 return parser.close()
3670 for k, v in LEGACY_NAMESPACES:
3672 dfxp_data = dfxp_data.replace(ns, k)
3674 dfxp = compat_etree_fromstring(dfxp_data)
3676 paras = dfxp.findall(_x('.//ttml:p')) or dfxp.findall('.//p')
3679 raise ValueError('Invalid dfxp/TTML subtitle')
3683 for style in dfxp.findall(_x('.//ttml:style')):
3684 style_id = style.get('id') or style.get(_x('xml:id'))
3687 parent_style_id = style.get('style')
3689 if parent_style_id not in styles:
3692 styles[style_id] = styles[parent_style_id].copy()
3693 for prop in SUPPORTED_STYLING:
3694 prop_val = style.get(_x('tts:' + prop))
3696 styles.setdefault(style_id, {})[prop] = prop_val
3702 for p in ('body', 'div'):
3703 ele = xpath_element(dfxp, [_x('.//ttml:' + p), './/' + p])
3706 style = styles.get(ele.get('style'))
3709 default_style.update(style)
3711 for para, index in zip(paras, itertools.count(1)):
3712 begin_time = parse_dfxp_time_expr(para.attrib.get('begin'))
3713 end_time = parse_dfxp_time_expr(para.attrib.get('end'))
3714 dur = parse_dfxp_time_expr(para.attrib.get('dur'))
3715 if begin_time is None:
3720 end_time = begin_time + dur
3721 out.append('%d\n%s --> %s\n%s\n\n' % (
3723 srt_subtitles_timecode(begin_time),
3724 srt_subtitles_timecode(end_time),
3730 def cli_option(params, command_option, param):
3731 param = params.get(param)
3733 param = compat_str(param)
3734 return [command_option, param] if param is not None else []
3737 def cli_bool_option(params, command_option, param, true_value='true', false_value='false', separator=None):
3738 param = params.get(param)
3741 assert isinstance(param, bool)
3743 return [command_option + separator + (true_value if param else false_value)]
3744 return [command_option, true_value if param else false_value]
3747 def cli_valueless_option(params, command_option, param, expected_value=True):
3748 param = params.get(param)
3749 return [command_option] if param == expected_value else []
3752 def cli_configuration_args(argdict, keys, default=[], use_compat=True):
3753 if isinstance(argdict, (list, tuple)): # for backward compatibility
3760 assert isinstance(argdict, dict)
3762 assert isinstance(keys, (list, tuple))
3763 for key_list in keys:
3764 arg_list = list(filter(
3765 lambda x: x is not None,
3766 [argdict.get(key.lower()) for key in variadic(key_list)]))
3768 return [arg for args in arg_list for arg in args]
3772 def _configuration_args(main_key, argdict, exe, keys=None, default=[], use_compat=True):
3773 main_key, exe = main_key.lower(), exe.lower()
3774 root_key = exe if main_key == exe else f'{main_key}+{exe}'
3775 keys = [f'{root_key}{k}' for k in (keys or [''])]
3776 if root_key in keys:
3778 keys.append((main_key, exe))
3779 keys.append('default')
3782 return cli_configuration_args(argdict, keys, default, use_compat)
3785 class ISO639Utils(object):
3786 # See http://www.loc.gov/standards/iso639-2/ISO-639-2_utf-8.txt
3845 'iw': 'heb', # Replaced by he in 1989 revision
3855 'in': 'ind', # Replaced by id in 1989 revision
3970 'ji': 'yid', # Replaced by yi in 1989 revision
3978 def short2long(cls, code):
3979 """Convert language code from ISO 639-1 to ISO 639-2/T"""
3980 return cls._lang_map.get(code[:2])
3983 def long2short(cls, code):
3984 """Convert language code from ISO 639-2/T to ISO 639-1"""
3985 for short_name, long_name in cls._lang_map.items():
3986 if long_name == code:
3990 class ISO3166Utils(object):
3991 # From http://data.okfn.org/data/core/country-list
3993 'AF': 'Afghanistan',
3994 'AX': 'Åland Islands',
3997 'AS': 'American Samoa',
4002 'AG': 'Antigua and Barbuda',
4019 'BO': 'Bolivia, Plurinational State of',
4020 'BQ': 'Bonaire, Sint Eustatius and Saba',
4021 'BA': 'Bosnia and Herzegovina',
4023 'BV': 'Bouvet Island',
4025 'IO': 'British Indian Ocean Territory',
4026 'BN': 'Brunei Darussalam',
4028 'BF': 'Burkina Faso',
4034 'KY': 'Cayman Islands',
4035 'CF': 'Central African Republic',
4039 'CX': 'Christmas Island',
4040 'CC': 'Cocos (Keeling) Islands',
4044 'CD': 'Congo, the Democratic Republic of the',
4045 'CK': 'Cook Islands',
4047 'CI': 'Côte d\'Ivoire',
4052 'CZ': 'Czech Republic',
4056 'DO': 'Dominican Republic',
4059 'SV': 'El Salvador',
4060 'GQ': 'Equatorial Guinea',
4064 'FK': 'Falkland Islands (Malvinas)',
4065 'FO': 'Faroe Islands',
4069 'GF': 'French Guiana',
4070 'PF': 'French Polynesia',
4071 'TF': 'French Southern Territories',
4086 'GW': 'Guinea-Bissau',
4089 'HM': 'Heard Island and McDonald Islands',
4090 'VA': 'Holy See (Vatican City State)',
4097 'IR': 'Iran, Islamic Republic of',
4100 'IM': 'Isle of Man',
4110 'KP': 'Korea, Democratic People\'s Republic of',
4111 'KR': 'Korea, Republic of',
4114 'LA': 'Lao People\'s Democratic Republic',
4120 'LI': 'Liechtenstein',
4124 'MK': 'Macedonia, the Former Yugoslav Republic of',
4131 'MH': 'Marshall Islands',
4137 'FM': 'Micronesia, Federated States of',
4138 'MD': 'Moldova, Republic of',
4149 'NL': 'Netherlands',
4150 'NC': 'New Caledonia',
4151 'NZ': 'New Zealand',
4156 'NF': 'Norfolk Island',
4157 'MP': 'Northern Mariana Islands',
4162 'PS': 'Palestine, State of',
4164 'PG': 'Papua New Guinea',
4167 'PH': 'Philippines',
4171 'PR': 'Puerto Rico',
4175 'RU': 'Russian Federation',
4177 'BL': 'Saint Barthélemy',
4178 'SH': 'Saint Helena, Ascension and Tristan da Cunha',
4179 'KN': 'Saint Kitts and Nevis',
4180 'LC': 'Saint Lucia',
4181 'MF': 'Saint Martin (French part)',
4182 'PM': 'Saint Pierre and Miquelon',
4183 'VC': 'Saint Vincent and the Grenadines',
4186 'ST': 'Sao Tome and Principe',
4187 'SA': 'Saudi Arabia',
4191 'SL': 'Sierra Leone',
4193 'SX': 'Sint Maarten (Dutch part)',
4196 'SB': 'Solomon Islands',
4198 'ZA': 'South Africa',
4199 'GS': 'South Georgia and the South Sandwich Islands',
4200 'SS': 'South Sudan',
4205 'SJ': 'Svalbard and Jan Mayen',
4208 'CH': 'Switzerland',
4209 'SY': 'Syrian Arab Republic',
4210 'TW': 'Taiwan, Province of China',
4212 'TZ': 'Tanzania, United Republic of',
4214 'TL': 'Timor-Leste',
4218 'TT': 'Trinidad and Tobago',
4221 'TM': 'Turkmenistan',
4222 'TC': 'Turks and Caicos Islands',
4226 'AE': 'United Arab Emirates',
4227 'GB': 'United Kingdom',
4228 'US': 'United States',
4229 'UM': 'United States Minor Outlying Islands',
4233 'VE': 'Venezuela, Bolivarian Republic of',
4235 'VG': 'Virgin Islands, British',
4236 'VI': 'Virgin Islands, U.S.',
4237 'WF': 'Wallis and Futuna',
4238 'EH': 'Western Sahara',
4245 def short2full(cls, code):
4246 """Convert an ISO 3166-2 country code to the corresponding full name"""
4247 return cls._country_map.get(code.upper())
4250 class GeoUtils(object):
4251 # Major IPv4 address blocks per country
4253 'AD': '46.172.224.0/19',
4254 'AE': '94.200.0.0/13',
4255 'AF': '149.54.0.0/17',
4256 'AG': '209.59.64.0/18',
4257 'AI': '204.14.248.0/21',
4258 'AL': '46.99.0.0/16',
4259 'AM': '46.70.0.0/15',
4260 'AO': '105.168.0.0/13',
4261 'AP': '182.50.184.0/21',
4262 'AQ': '23.154.160.0/24',
4263 'AR': '181.0.0.0/12',
4264 'AS': '202.70.112.0/20',
4265 'AT': '77.116.0.0/14',
4266 'AU': '1.128.0.0/11',
4267 'AW': '181.41.0.0/18',
4268 'AX': '185.217.4.0/22',
4269 'AZ': '5.197.0.0/16',
4270 'BA': '31.176.128.0/17',
4271 'BB': '65.48.128.0/17',
4272 'BD': '114.130.0.0/16',
4274 'BF': '102.178.0.0/15',
4275 'BG': '95.42.0.0/15',
4276 'BH': '37.131.0.0/17',
4277 'BI': '154.117.192.0/18',
4278 'BJ': '137.255.0.0/16',
4279 'BL': '185.212.72.0/23',
4280 'BM': '196.12.64.0/18',
4281 'BN': '156.31.0.0/16',
4282 'BO': '161.56.0.0/16',
4283 'BQ': '161.0.80.0/20',
4284 'BR': '191.128.0.0/12',
4285 'BS': '24.51.64.0/18',
4286 'BT': '119.2.96.0/19',
4287 'BW': '168.167.0.0/16',
4288 'BY': '178.120.0.0/13',
4289 'BZ': '179.42.192.0/18',
4290 'CA': '99.224.0.0/11',
4291 'CD': '41.243.0.0/16',
4292 'CF': '197.242.176.0/21',
4293 'CG': '160.113.0.0/16',
4294 'CH': '85.0.0.0/13',
4295 'CI': '102.136.0.0/14',
4296 'CK': '202.65.32.0/19',
4297 'CL': '152.172.0.0/14',
4298 'CM': '102.244.0.0/14',
4299 'CN': '36.128.0.0/10',
4300 'CO': '181.240.0.0/12',
4301 'CR': '201.192.0.0/12',
4302 'CU': '152.206.0.0/15',
4303 'CV': '165.90.96.0/19',
4304 'CW': '190.88.128.0/17',
4305 'CY': '31.153.0.0/16',
4306 'CZ': '88.100.0.0/14',
4308 'DJ': '197.241.0.0/17',
4309 'DK': '87.48.0.0/12',
4310 'DM': '192.243.48.0/20',
4311 'DO': '152.166.0.0/15',
4312 'DZ': '41.96.0.0/12',
4313 'EC': '186.68.0.0/15',
4314 'EE': '90.190.0.0/15',
4315 'EG': '156.160.0.0/11',
4316 'ER': '196.200.96.0/20',
4317 'ES': '88.0.0.0/11',
4318 'ET': '196.188.0.0/14',
4319 'EU': '2.16.0.0/13',
4320 'FI': '91.152.0.0/13',
4321 'FJ': '144.120.0.0/16',
4322 'FK': '80.73.208.0/21',
4323 'FM': '119.252.112.0/20',
4324 'FO': '88.85.32.0/19',
4326 'GA': '41.158.0.0/15',
4328 'GD': '74.122.88.0/21',
4329 'GE': '31.146.0.0/16',
4330 'GF': '161.22.64.0/18',
4331 'GG': '62.68.160.0/19',
4332 'GH': '154.160.0.0/12',
4333 'GI': '95.164.0.0/16',
4334 'GL': '88.83.0.0/19',
4335 'GM': '160.182.0.0/15',
4336 'GN': '197.149.192.0/18',
4337 'GP': '104.250.0.0/19',
4338 'GQ': '105.235.224.0/20',
4339 'GR': '94.64.0.0/13',
4340 'GT': '168.234.0.0/16',
4341 'GU': '168.123.0.0/16',
4342 'GW': '197.214.80.0/20',
4343 'GY': '181.41.64.0/18',
4344 'HK': '113.252.0.0/14',
4345 'HN': '181.210.0.0/16',
4346 'HR': '93.136.0.0/13',
4347 'HT': '148.102.128.0/17',
4348 'HU': '84.0.0.0/14',
4349 'ID': '39.192.0.0/10',
4350 'IE': '87.32.0.0/12',
4351 'IL': '79.176.0.0/13',
4352 'IM': '5.62.80.0/20',
4353 'IN': '117.192.0.0/10',
4354 'IO': '203.83.48.0/21',
4355 'IQ': '37.236.0.0/14',
4356 'IR': '2.176.0.0/12',
4357 'IS': '82.221.0.0/16',
4358 'IT': '79.0.0.0/10',
4359 'JE': '87.244.64.0/18',
4360 'JM': '72.27.0.0/17',
4361 'JO': '176.29.0.0/16',
4362 'JP': '133.0.0.0/8',
4363 'KE': '105.48.0.0/12',
4364 'KG': '158.181.128.0/17',
4365 'KH': '36.37.128.0/17',
4366 'KI': '103.25.140.0/22',
4367 'KM': '197.255.224.0/20',
4368 'KN': '198.167.192.0/19',
4369 'KP': '175.45.176.0/22',
4370 'KR': '175.192.0.0/10',
4371 'KW': '37.36.0.0/14',
4372 'KY': '64.96.0.0/15',
4373 'KZ': '2.72.0.0/13',
4374 'LA': '115.84.64.0/18',
4375 'LB': '178.135.0.0/16',
4376 'LC': '24.92.144.0/20',
4377 'LI': '82.117.0.0/19',
4378 'LK': '112.134.0.0/15',
4379 'LR': '102.183.0.0/16',
4380 'LS': '129.232.0.0/17',
4381 'LT': '78.56.0.0/13',
4382 'LU': '188.42.0.0/16',
4383 'LV': '46.109.0.0/16',
4384 'LY': '41.252.0.0/14',
4385 'MA': '105.128.0.0/11',
4386 'MC': '88.209.64.0/18',
4387 'MD': '37.246.0.0/16',
4388 'ME': '178.175.0.0/17',
4389 'MF': '74.112.232.0/21',
4390 'MG': '154.126.0.0/17',
4391 'MH': '117.103.88.0/21',
4392 'MK': '77.28.0.0/15',
4393 'ML': '154.118.128.0/18',
4394 'MM': '37.111.0.0/17',
4395 'MN': '49.0.128.0/17',
4396 'MO': '60.246.0.0/16',
4397 'MP': '202.88.64.0/20',
4398 'MQ': '109.203.224.0/19',
4399 'MR': '41.188.64.0/18',
4400 'MS': '208.90.112.0/22',
4401 'MT': '46.11.0.0/16',
4402 'MU': '105.16.0.0/12',
4403 'MV': '27.114.128.0/18',
4404 'MW': '102.70.0.0/15',
4405 'MX': '187.192.0.0/11',
4406 'MY': '175.136.0.0/13',
4407 'MZ': '197.218.0.0/15',
4408 'NA': '41.182.0.0/16',
4409 'NC': '101.101.0.0/18',
4410 'NE': '197.214.0.0/18',
4411 'NF': '203.17.240.0/22',
4412 'NG': '105.112.0.0/12',
4413 'NI': '186.76.0.0/15',
4414 'NL': '145.96.0.0/11',
4415 'NO': '84.208.0.0/13',
4416 'NP': '36.252.0.0/15',
4417 'NR': '203.98.224.0/19',
4418 'NU': '49.156.48.0/22',
4419 'NZ': '49.224.0.0/14',
4420 'OM': '5.36.0.0/15',
4421 'PA': '186.72.0.0/15',
4422 'PE': '186.160.0.0/14',
4423 'PF': '123.50.64.0/18',
4424 'PG': '124.240.192.0/19',
4425 'PH': '49.144.0.0/13',
4426 'PK': '39.32.0.0/11',
4427 'PL': '83.0.0.0/11',
4428 'PM': '70.36.0.0/20',
4429 'PR': '66.50.0.0/16',
4430 'PS': '188.161.0.0/16',
4431 'PT': '85.240.0.0/13',
4432 'PW': '202.124.224.0/20',
4433 'PY': '181.120.0.0/14',
4434 'QA': '37.210.0.0/15',
4435 'RE': '102.35.0.0/16',
4436 'RO': '79.112.0.0/13',
4437 'RS': '93.86.0.0/15',
4438 'RU': '5.136.0.0/13',
4439 'RW': '41.186.0.0/16',
4440 'SA': '188.48.0.0/13',
4441 'SB': '202.1.160.0/19',
4442 'SC': '154.192.0.0/11',
4443 'SD': '102.120.0.0/13',
4444 'SE': '78.64.0.0/12',
4445 'SG': '8.128.0.0/10',
4446 'SI': '188.196.0.0/14',
4447 'SK': '78.98.0.0/15',
4448 'SL': '102.143.0.0/17',
4449 'SM': '89.186.32.0/19',
4450 'SN': '41.82.0.0/15',
4451 'SO': '154.115.192.0/18',
4452 'SR': '186.179.128.0/17',
4453 'SS': '105.235.208.0/21',
4454 'ST': '197.159.160.0/19',
4455 'SV': '168.243.0.0/16',
4456 'SX': '190.102.0.0/20',
4458 'SZ': '41.84.224.0/19',
4459 'TC': '65.255.48.0/20',
4460 'TD': '154.68.128.0/19',
4461 'TG': '196.168.0.0/14',
4462 'TH': '171.96.0.0/13',
4463 'TJ': '85.9.128.0/18',
4464 'TK': '27.96.24.0/21',
4465 'TL': '180.189.160.0/20',
4466 'TM': '95.85.96.0/19',
4467 'TN': '197.0.0.0/11',
4468 'TO': '175.176.144.0/21',
4469 'TR': '78.160.0.0/11',
4470 'TT': '186.44.0.0/15',
4471 'TV': '202.2.96.0/19',
4472 'TW': '120.96.0.0/11',
4473 'TZ': '156.156.0.0/14',
4474 'UA': '37.52.0.0/14',
4475 'UG': '102.80.0.0/13',
4477 'UY': '167.56.0.0/13',
4478 'UZ': '84.54.64.0/18',
4479 'VA': '212.77.0.0/19',
4480 'VC': '207.191.240.0/21',
4481 'VE': '186.88.0.0/13',
4482 'VG': '66.81.192.0/20',
4483 'VI': '146.226.0.0/16',
4484 'VN': '14.160.0.0/11',
4485 'VU': '202.80.32.0/20',
4486 'WF': '117.20.32.0/21',
4487 'WS': '202.4.32.0/19',
4488 'YE': '134.35.0.0/16',
4489 'YT': '41.242.116.0/22',
4490 'ZA': '41.0.0.0/11',
4491 'ZM': '102.144.0.0/13',
4492 'ZW': '102.177.192.0/18',
4496 def random_ipv4(cls, code_or_block):
4497 if len(code_or_block) == 2:
4498 block = cls._country_ip_map.get(code_or_block.upper())
4502 block = code_or_block
4503 addr, preflen = block.split('/')
4504 addr_min = compat_struct_unpack('!L', socket.inet_aton(addr))[0]
4505 addr_max = addr_min | (0xffffffff >> int(preflen))
4506 return compat_str(socket.inet_ntoa(
4507 compat_struct_pack('!L', random.randint(addr_min, addr_max))))
4510 class PerRequestProxyHandler(compat_urllib_request.ProxyHandler):
4511 def __init__(self, proxies=None):
4512 # Set default handlers
4513 for type in ('http', 'https'):
4514 setattr(self, '%s_open' % type,
4515 lambda r, proxy='__noproxy__', type=type, meth=self.proxy_open:
4516 meth(r, proxy, type))
4517 compat_urllib_request.ProxyHandler.__init__(self, proxies)
4519 def proxy_open(self, req, proxy, type):
4520 req_proxy = req.headers.get('Ytdl-request-proxy')
4521 if req_proxy is not None:
4523 del req.headers['Ytdl-request-proxy']
4525 if proxy == '__noproxy__':
4526 return None # No Proxy
4527 if compat_urlparse.urlparse(proxy).scheme.lower() in ('socks', 'socks4', 'socks4a', 'socks5'):
4528 req.add_header('Ytdl-socks-proxy', proxy)
4529 # yt-dlp's http/https handlers do wrapping the socket with socks
4531 return compat_urllib_request.ProxyHandler.proxy_open(
4532 self, req, proxy, type)
4535 # Both long_to_bytes and bytes_to_long are adapted from PyCrypto, which is
4536 # released into Public Domain
4537 # https://github.com/dlitz/pycrypto/blob/master/lib/Crypto/Util/number.py#L387
4539 def long_to_bytes(n, blocksize=0):
4540 """long_to_bytes(n:long, blocksize:int) : string
4541 Convert a long integer to a byte string.
4543 If optional blocksize is given and greater than zero, pad the front of the
4544 byte string with binary zeros so that the length is a multiple of
4547 # after much testing, this algorithm was deemed to be the fastest
4551 s = compat_struct_pack('>I', n & 0xffffffff) + s
4553 # strip off leading zeros
4554 for i in range(len(s)):
4555 if s[i] != b'\000'[0]:
4558 # only happens when n == 0
4562 # add back some pad bytes. this could be done more efficiently w.r.t. the
4563 # de-padding being done above, but sigh...
4564 if blocksize > 0 and len(s) % blocksize:
4565 s = (blocksize - len(s) % blocksize) * b'\000' + s
4569 def bytes_to_long(s):
4570 """bytes_to_long(string) : long
4571 Convert a byte string to a long integer.
4573 This is (essentially) the inverse of long_to_bytes().
4578 extra = (4 - length % 4)
4579 s = b'\000' * extra + s
4580 length = length + extra
4581 for i in range(0, length, 4):
4582 acc = (acc << 32) + compat_struct_unpack('>I', s[i:i + 4])[0]
4586 def ohdave_rsa_encrypt(data, exponent, modulus):
4588 Implement OHDave
's RSA algorithm. See http://www.ohdave.com/rsa/
4591 data: data to encrypt, bytes-like object
4592 exponent, modulus: parameter e and N of RSA algorithm, both integer
4593 Output: hex string of encrypted data
4595 Limitation: supports one block encryption only
4598 payload = int(binascii.hexlify(data[::-1]), 16)
4599 encrypted = pow(payload, exponent, modulus)
4600 return '%x' % encrypted
4603 def pkcs1pad(data, length):
4605 Padding input data with PKCS#1 scheme
4607 @param {int[]} data input data
4608 @param {int} length target length
4609 @returns {int[]} padded data
4611 if len(data) > length - 11:
4612 raise ValueError('Input data too
long for PKCS
#1 padding')
4614 pseudo_random
= [random
.randint(0, 254) for _
in range(length
- len(data
) - 3)]
4615 return [0, 2] + pseudo_random
+ [0] + data
4618 def encode_base_n(num
, n
, table
=None):
4619 FULL_TABLE
= '0123456789abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ'
4621 table
= FULL_TABLE
[:n
]
4624 raise ValueError('base %d exceeds table length %d' % (n
, len(table
)))
4631 ret
= table
[num
% n
] + ret
4636 def decode_packed_codes(code
):
4637 mobj
= re
.search(PACKED_CODES_RE
, code
)
4638 obfuscated_code
, base
, count
, symbols
= mobj
.groups()
4641 symbols
= symbols
.split('|')
4646 base_n_count
= encode_base_n(count
, base
)
4647 symbol_table
[base_n_count
] = symbols
[count
] or base_n_count
4650 r
'\b(\w+)\b', lambda mobj
: symbol_table
[mobj
.group(0)],
4654 def caesar(s
, alphabet
, shift
):
4659 alphabet
[(alphabet
.index(c
) + shift
) % l
] if c
in alphabet
else c
4664 return caesar(s
, r
'''!"#$%&'()*+,-./0123456789:;<=>?@ABCDEFGHIJKLMNOPQRSTUVWXYZ[\]^_`abcdefghijklmnopqrstuvwxyz{|}~''', 47)
4667 def parse_m3u8_attributes(attrib
):
4669 for (key
, val
) in re
.findall(r
'(?P<key>[A-Z0-9-]+)=(?P<val>"[^"]+"|[^",]+)(?:,|$)', attrib
):
4670 if val
.startswith('"'):
4676 def urshift(val
, n
):
4677 return val
>> n
if val
>= 0 else (val
+ 0x100000000) >> n
4680 # Based on png2str() written by @gdkchan and improved by @yokrysty
4681 # Originally posted at https://github.com/ytdl-org/youtube-dl/issues/9706
4682 def decode_png(png_data
):
4683 # Reference: https://www.w3.org/TR/PNG/
4684 header
= png_data
[8:]
4686 if png_data
[:8] != b
'\x89PNG\x0d\x0a\x1a\x0a' or header
[4:8] != b
'IHDR':
4687 raise IOError('Not a valid PNG file.')
4689 int_map
= {1: '>B', 2: '>H', 4: '>I'}
4690 unpack_integer
= lambda x
: compat_struct_unpack(int_map
[len(x
)], x
)[0]
4695 length
= unpack_integer(header
[:4])
4698 chunk_type
= header
[:4]
4701 chunk_data
= header
[:length
]
4702 header
= header
[length
:]
4704 header
= header
[4:] # Skip CRC
4712 ihdr
= chunks
[0]['data']
4714 width
= unpack_integer(ihdr
[:4])
4715 height
= unpack_integer(ihdr
[4:8])
4719 for chunk
in chunks
:
4720 if chunk
['type'] == b
'IDAT':
4721 idat
+= chunk
['data']
4724 raise IOError('Unable to read PNG data.')
4726 decompressed_data
= bytearray(zlib
.decompress(idat
))
4731 def _get_pixel(idx
):
4736 for y
in range(height
):
4737 basePos
= y
* (1 + stride
)
4738 filter_type
= decompressed_data
[basePos
]
4742 pixels
.append(current_row
)
4744 for x
in range(stride
):
4745 color
= decompressed_data
[1 + basePos
+ x
]
4746 basex
= y
* stride
+ x
4751 left
= _get_pixel(basex
- 3)
4753 up
= _get_pixel(basex
- stride
)
4755 if filter_type
== 1: # Sub
4756 color
= (color
+ left
) & 0xff
4757 elif filter_type
== 2: # Up
4758 color
= (color
+ up
) & 0xff
4759 elif filter_type
== 3: # Average
4760 color
= (color
+ ((left
+ up
) >> 1)) & 0xff
4761 elif filter_type
== 4: # Paeth
4767 c
= _get_pixel(basex
- stride
- 3)
4775 if pa
<= pb
and pa
<= pc
:
4776 color
= (color
+ a
) & 0xff
4778 color
= (color
+ b
) & 0xff
4780 color
= (color
+ c
) & 0xff
4782 current_row
.append(color
)
4784 return width
, height
, pixels
4787 def write_xattr(path
, key
, value
):
4788 # This mess below finds the best xattr tool for the job
4790 # try the pyxattr module...
4793 if hasattr(xattr
, 'set'): # pyxattr
4794 # Unicode arguments are not supported in python-pyxattr until
4796 # See https://github.com/ytdl-org/youtube-dl/issues/5498
4797 pyxattr_required_version
= '0.5.0'
4798 if version_tuple(xattr
.__version
__) < version_tuple(pyxattr_required_version
):
4799 # TODO: fallback to CLI tools
4800 raise XAttrUnavailableError(
4801 'python-pyxattr is detected but is too old. '
4802 'yt-dlp requires %s or above while your version is %s. '
4803 'Falling back to other xattr implementations' % (
4804 pyxattr_required_version
, xattr
.__version
__))
4806 setxattr
= xattr
.set
4808 setxattr
= xattr
.setxattr
4811 setxattr(path
, key
, value
)
4812 except EnvironmentError as e
:
4813 raise XAttrMetadataError(e
.errno
, e
.strerror
)
4816 if compat_os_name
== 'nt':
4817 # Write xattrs to NTFS Alternate Data Streams:
4818 # http://en.wikipedia.org/wiki/NTFS#Alternate_data_streams_.28ADS.29
4819 assert ':' not in key
4820 assert os
.path
.exists(path
)
4822 ads_fn
= path
+ ':' + key
4824 with open(ads_fn
, 'wb') as f
:
4826 except EnvironmentError as e
:
4827 raise XAttrMetadataError(e
.errno
, e
.strerror
)
4829 user_has_setfattr
= check_executable('setfattr', ['--version'])
4830 user_has_xattr
= check_executable('xattr', ['-h'])
4832 if user_has_setfattr
or user_has_xattr
:
4834 value
= value
.decode('utf-8')
4835 if user_has_setfattr
:
4836 executable
= 'setfattr'
4837 opts
= ['-n', key
, '-v', value
]
4838 elif user_has_xattr
:
4839 executable
= 'xattr'
4840 opts
= ['-w', key
, value
]
4842 cmd
= ([encodeFilename(executable
, True)]
4843 + [encodeArgument(o
) for o
in opts
]
4844 + [encodeFilename(path
, True)])
4848 cmd
, stdout
=subprocess
.PIPE
, stderr
=subprocess
.PIPE
, stdin
=subprocess
.PIPE
)
4849 except EnvironmentError as e
:
4850 raise XAttrMetadataError(e
.errno
, e
.strerror
)
4851 stdout
, stderr
= p
.communicate_or_kill()
4852 stderr
= stderr
.decode('utf-8', 'replace')
4853 if p
.returncode
!= 0:
4854 raise XAttrMetadataError(p
.returncode
, stderr
)
4857 # On Unix, and can't find pyxattr, setfattr, or xattr.
4858 if sys
.platform
.startswith('linux'):
4859 raise XAttrUnavailableError(
4860 "Couldn't find a tool to set the xattrs. "
4861 "Install either the python 'pyxattr' or 'xattr' "
4862 "modules, or the GNU 'attr' package "
4863 "(which contains the 'setfattr' tool).")
4865 raise XAttrUnavailableError(
4866 "Couldn't find a tool to set the xattrs. "
4867 "Install either the python 'xattr' module, "
4868 "or the 'xattr' binary.")
4871 def random_birthday(year_field
, month_field
, day_field
):
4872 start_date
= datetime
.date(1950, 1, 1)
4873 end_date
= datetime
.date(1995, 12, 31)
4874 offset
= random
.randint(0, (end_date
- start_date
).days
)
4875 random_date
= start_date
+ datetime
.timedelta(offset
)
4877 year_field
: str(random_date
.year
),
4878 month_field
: str(random_date
.month
),
4879 day_field
: str(random_date
.day
),
4883 # Templates for internet shortcut files, which are plain text files.
4884 DOT_URL_LINK_TEMPLATE
= '''
4889 DOT_WEBLOC_LINK_TEMPLATE
= '''
4890 <?xml version="1.0" encoding="UTF-8"?>
4891 <!DOCTYPE plist PUBLIC "-//Apple//DTD PLIST 1.0//EN" "http://www.apple.com/DTDs/PropertyList-1.0.dtd">
4892 <plist version="1.0">
4895 \t<string>%(url)s</string>
4900 DOT_DESKTOP_LINK_TEMPLATE
= '''
4910 'url': DOT_URL_LINK_TEMPLATE
,
4911 'desktop': DOT_DESKTOP_LINK_TEMPLATE
,
4912 'webloc': DOT_WEBLOC_LINK_TEMPLATE
,
4916 def iri_to_uri(iri
):
4918 Converts an IRI (Internationalized Resource Identifier, allowing Unicode characters) to a URI (Uniform Resource Identifier, ASCII-only).
4920 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.
4923 iri_parts
= compat_urllib_parse_urlparse(iri
)
4925 if '[' in iri_parts
.netloc
:
4926 raise ValueError('IPv6 URIs are not, yet, supported.')
4927 # Querying `.netloc`, when there's only one bracket, also raises a ValueError.
4929 # 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.
4932 if iri_parts
.username
:
4933 net_location
+= compat_urllib_parse_quote(iri_parts
.username
, safe
=r
"!$%&'()*+,~")
4934 if iri_parts
.password
is not None:
4935 net_location
+= ':' + compat_urllib_parse_quote(iri_parts
.password
, safe
=r
"!$%&'()*+,~")
4938 net_location
+= iri_parts
.hostname
.encode('idna').decode('utf-8') # Punycode for Unicode hostnames.
4939 # The 'idna' encoding produces ASCII text.
4940 if iri_parts
.port
is not None and iri_parts
.port
!= 80:
4941 net_location
+= ':' + str(iri_parts
.port
)
4943 return compat_urllib_parse_urlunparse(
4947 compat_urllib_parse_quote_plus(iri_parts
.path
, safe
=r
"!$%&'()*+,/:;=@|~"),
4949 # Unsure about the `safe` argument, since this is a legacy way of handling parameters.
4950 compat_urllib_parse_quote_plus(iri_parts
.params
, safe
=r
"!$%&'()*+,/:;=@|~"),
4952 # Not totally sure about the `safe` argument, since the source does not explicitly mention the query URI component.
4953 compat_urllib_parse_quote_plus(iri_parts
.query
, safe
=r
"!$%&'()*+,/:;=?@{|}~"),
4955 compat_urllib_parse_quote_plus(iri_parts
.fragment
, safe
=r
"!#$%&'()*+,/:;=?@{|}~")))
4957 # Source for `safe` arguments: https://url.spec.whatwg.org/#percent-encoded-bytes.
4960 def to_high_limit_path(path
):
4961 if sys
.platform
in ['win32', 'cygwin']:
4962 # Work around MAX_PATH limitation on Windows. The maximum allowed length for the individual path segments may still be quite limited.
4963 return r
'\\?\ '.rstrip() + os
.path
.abspath(path
)
4968 def format_field(obj
, field
=None, template
='%s', ignore
=(None, ''), default
='', func
=None):
4970 val
= obj
if obj
is not None else default
4972 val
= obj
.get(field
, default
)
4973 if func
and val
not in ignore
:
4975 return template
% val
if val
not in ignore
else default
4978 def clean_podcast_url(url
):
4979 return re
.sub(r
'''(?x)
4983 media\.blubrry\.com| # https://create.blubrry.com/resources/podcast-media-download-statistics/getting-started/
4986 (?:dts|www)\.podtrac\.com/(?:pts/)?redirect\.[0-9a-z]{3,4}| # http://analytics.podtrac.com/how-to-measure
4989 cn\.co| # https://podcorn.com/analytics-prefix/
4990 st\.fm # https://podsights.com/docs/
4995 _HEX_TABLE
= '0123456789abcdef'
4998 def random_uuidv4():
4999 return re
.sub(r
'[xy]', lambda x
: _HEX_TABLE
[random
.randint(0, 15)], 'xxxxxxxx-xxxx-4xxx-yxxx-xxxxxxxxxxxx')
5002 def make_dir(path
, to_screen
=None):
5004 dn
= os
.path
.dirname(path
)
5005 if dn
and not os
.path
.exists(dn
):
5008 except (OSError, IOError) as err
:
5009 if callable(to_screen
) is not None:
5010 to_screen('unable to create directory ' + error_to_compat_str(err
))
5014 def get_executable_path():
5015 from zipimport
import zipimporter
5016 if hasattr(sys
, 'frozen'): # Running from PyInstaller
5017 path
= os
.path
.dirname(sys
.executable
)
5018 elif isinstance(globals().get('__loader__'), zipimporter
): # Running from ZIP
5019 path
= os
.path
.join(os
.path
.dirname(__file__
), '../..')
5021 path
= os
.path
.join(os
.path
.dirname(__file__
), '..')
5022 return os
.path
.abspath(path
)
5025 def load_plugins(name
, suffix
, namespace
):
5028 plugins_spec
= importlib
.util
.spec_from_file_location(
5029 name
, os
.path
.join(get_executable_path(), 'ytdlp_plugins', name
, '__init__.py'))
5030 plugins
= importlib
.util
.module_from_spec(plugins_spec
)
5031 sys
.modules
[plugins_spec
.name
] = plugins
5032 plugins_spec
.loader
.exec_module(plugins
)
5033 for name
in dir(plugins
):
5034 if name
in namespace
:
5036 if not name
.endswith(suffix
):
5038 klass
= getattr(plugins
, name
)
5039 classes
[name
] = namespace
[name
] = klass
5040 except FileNotFoundError
:
5046 obj
, *path_list
, default
=None, expected_type
=None, get_all
=True,
5047 casesense
=True, is_user_input
=False, traverse_string
=False):
5048 ''' Traverse nested list/dict/tuple
5049 @param path_list A list of paths which are checked one by one.
5050 Each path is a list of keys where each key is a string,
5051 a function, a tuple of strings/None or "...".
5052 When a fuction is given, it takes the key as argument and
5053 returns whether the key matches or not. When a tuple is given,
5054 all the keys given in the tuple are traversed, and
5055 "..." traverses all the keys in the object
5056 "None" returns the object without traversal
5057 @param default Default value to return
5058 @param expected_type Only accept final value of this type (Can also be any callable)
5059 @param get_all Return all the values obtained from a path or only the first one
5060 @param casesense Whether to consider dictionary keys as case sensitive
5061 @param is_user_input Whether the keys are generated from user input. If True,
5062 strings are converted to int/slice if necessary
5063 @param traverse_string Whether to traverse inside strings. If True, any
5064 non-compatible object will also be converted into a string
5068 _lower
= lambda k
: (k
.lower() if isinstance(k
, str) else k
)
5069 path_list
= (map(_lower
, variadic(path
)) for path
in path_list
)
5071 def _traverse_obj(obj
, path
, _current_depth
=0):
5073 path
= tuple(variadic(path
))
5074 for i
, key
in enumerate(path
):
5075 if None in (key
, obj
):
5077 if isinstance(key
, (list, tuple)):
5078 obj
= [_traverse_obj(obj
, sub_key
, _current_depth
) for sub_key
in key
]
5081 obj
= (obj
.values() if isinstance(obj
, dict)
5082 else obj
if isinstance(obj
, (list, tuple, LazyList
))
5083 else str(obj
) if traverse_string
else [])
5085 depth
= max(depth
, _current_depth
)
5086 return [_traverse_obj(inner_obj
, path
[i
+ 1:], _current_depth
) for inner_obj
in obj
]
5088 if isinstance(obj
, (list, tuple, LazyList
)):
5089 obj
= enumerate(obj
)
5090 elif isinstance(obj
, dict):
5093 if not traverse_string
:
5097 depth
= max(depth
, _current_depth
)
5098 return [_traverse_obj(v
, path
[i
+ 1:], _current_depth
) for k
, v
in obj
if key(k
)]
5099 elif isinstance(obj
, dict) and not (is_user_input
and key
== ':'):
5100 obj
= (obj
.get(key
) if casesense
or (key
in obj
)
5101 else next((v
for k
, v
in obj
.items() if _lower(k
) == key
), None))
5104 key
= (int_or_none(key
) if ':' not in key
5105 else slice(*map(int_or_none
, key
.split(':'))))
5106 if key
== slice(None):
5107 return _traverse_obj(obj
, (..., *path
[i
+ 1:]), _current_depth
)
5108 if not isinstance(key
, (int, slice)):
5110 if not isinstance(obj
, (list, tuple, LazyList
)):
5111 if not traverse_string
:
5120 if isinstance(expected_type
, type):
5121 type_test
= lambda val
: val
if isinstance(val
, expected_type
) else None
5122 elif expected_type
is not None:
5123 type_test
= expected_type
5125 type_test
= lambda val
: val
5127 for path
in path_list
:
5129 val
= _traverse_obj(obj
, path
)
5132 for _
in range(depth
- 1):
5133 val
= itertools
.chain
.from_iterable(v
for v
in val
if v
is not None)
5134 val
= [v
for v
in map(type_test
, val
) if v
is not None]
5136 return val
if get_all
else val
[0]
5138 val
= type_test(val
)
5144 def traverse_dict(dictn
, keys
, casesense
=True):
5145 write_string('DeprecationWarning: yt_dlp.utils.traverse_dict is deprecated '
5146 'and may be removed in a future version. Use yt_dlp.utils.traverse_obj instead')
5147 return traverse_obj(dictn
, keys
, casesense
=casesense
, is_user_input
=True, traverse_string
=True)
5150 def variadic(x
, allowed_types
=(str, bytes, dict)):
5151 return x
if isinstance(x
, collections
.abc
.Iterable
) and not isinstance(x
, allowed_types
) else (x
,)
5154 # create a JSON Web Signature (jws) with HS256 algorithm
5155 # the resulting format is in JWS Compact Serialization
5156 # implemented following JWT https://www.rfc-editor.org/rfc/rfc7519.html
5157 # implemented following JWS https://www.rfc-editor.org/rfc/rfc7515.html
5158 def jwt_encode_hs256(payload_data
, key
, headers
={}):
5164 header_data
.update(headers
)
5165 header_b64
= base64
.b64encode(json
.dumps(header_data
).encode('utf-8'))
5166 payload_b64
= base64
.b64encode(json
.dumps(payload_data
).encode('utf-8'))
5167 h
= hmac
.new(key
.encode('utf-8'), header_b64
+ b
'.' + payload_b64
, hashlib
.sha256
)
5168 signature_b64
= base64
.b64encode(h
.digest())
5169 token
= header_b64
+ b
'.' + payload_b64
+ b
'.' + signature_b64
5173 # can be extended in future to verify the signature and parse header and return the algorithm used if it's not HS256
5174 def jwt_decode_hs256(jwt
):
5175 header_b64
, payload_b64
, signature_b64
= jwt
.split('.')
5176 payload_data
= json
.loads(base64
.urlsafe_b64decode(payload_b64
))
5180 def supports_terminal_sequences(stream
):
5181 if compat_os_name
== 'nt':
5182 from .compat
import WINDOWS_VT_MODE
# Must be imported locally
5183 if not WINDOWS_VT_MODE
or get_windows_version() < (10, 0, 10586):
5185 elif not os
.getenv('TERM'):
5188 return stream
.isatty()
5189 except BaseException
:
5193 _terminal_sequences_re
= re
.compile('\033\\[[^m]+m')
5196 def remove_terminal_sequences(string
):
5197 return _terminal_sequences_re
.sub('', string
)
5200 def number_of_digits(number
):
5201 return len('%d' % number
)
5204 def join_nonempty(*values
, delim
='-', from_dict
=None):
5205 if from_dict
is not None:
5206 values
= map(from_dict
.get
, values
)
5207 return delim
.join(map(str, filter(None, values
)))
5213 __initialized
= False
5215 def __init__(self
, parser
, label
=None):
5216 self
._parser
, self
.label
= parser
, label
5217 self
._loaded
_paths
, self
.configs
= set(), []
5219 def init(self
, args
=None, filename
=None):
5220 assert not self
.__initialized
5222 location
= os
.path
.realpath(filename
)
5223 if location
in self
._loaded
_paths
:
5225 self
._loaded
_paths
.add(location
)
5227 self
.__initialized
= True
5228 self
.own_args
, self
.filename
= args
, filename
5229 for location
in self
._parser
.parse_args(args
)[0].config_locations
or []:
5230 location
= compat_expanduser(location
)
5231 if os
.path
.isdir(location
):
5232 location
= os
.path
.join(location
, 'yt-dlp.conf')
5233 if not os
.path
.exists(location
):
5234 self
._parser
.error(f
'config location {location} does not exist')
5235 self
.append_config(self
.read_file(location
), location
)
5239 label
= join_nonempty(
5240 self
.label
, 'config', f
'"{self.filename}"' if self
.filename
else '',
5242 return join_nonempty(
5243 self
.own_args
is not None and f
'{label[0].upper()}{label[1:]}: {self.hide_login_info(self.own_args)}',
5244 *(f
'\n{c}'.replace('\n', '\n| ')[1:] for c
in self
.configs
),
5248 def read_file(filename
, default
=[]):
5250 optionf
= open(filename
)
5252 return default
# silently skip if file is not present
5254 # FIXME: https://github.com/ytdl-org/youtube-dl/commit/dfe5fa49aed02cf36ba9f743b11b0903554b5e56
5255 contents
= optionf
.read()
5256 if sys
.version_info
< (3,):
5257 contents
= contents
.decode(preferredencoding())
5258 res
= compat_shlex_split(contents
, comments
=True)
5264 def hide_login_info(opts
):
5265 PRIVATE_OPTS
= set(['-p', '--password', '-u', '--username', '--video-password', '--ap-password', '--ap-username'])
5266 eqre
= re
.compile('^(?P<key>' + ('|'.join(re
.escape(po
) for po
in PRIVATE_OPTS
)) + ')=.+$')
5271 return m
.group('key') + '=PRIVATE'
5275 opts
= list(map(_scrub_eq
, opts
))
5276 for idx
, opt
in enumerate(opts
):
5277 if opt
in PRIVATE_OPTS
and idx
+ 1 < len(opts
):
5278 opts
[idx
+ 1] = 'PRIVATE'
5281 def append_config(self
, *args
, label
=None):
5282 config
= type(self
)(self
._parser
, label
)
5283 config
._loaded
_paths
= self
._loaded
_paths
5284 if config
.init(*args
):
5285 self
.configs
.append(config
)
5289 for config
in reversed(self
.configs
):
5290 yield from config
.all_args
5291 yield from self
.own_args
or []
5293 def parse_args(self
):
5294 return self
._parser
.parse_args(list(self
.all_args
))