4 from __future__
import unicode_literals
41 import xml
.etree
.ElementTree
46 compat_HTMLParseError
,
52 compat_ctypes_WINFUNCTYPE
,
53 compat_etree_fromstring
,
56 compat_html_entities_html5
,
70 compat_urllib_parse_urlencode
,
71 compat_urllib_parse_urlparse
,
72 compat_urllib_parse_urlunparse
,
73 compat_urllib_parse_quote
,
74 compat_urllib_parse_quote_plus
,
75 compat_urllib_parse_unquote_plus
,
76 compat_urllib_request
,
88 def register_socks_protocols():
89 # "Register" SOCKS protocols
90 # In Python < 2.6.5, urlsplit() suffers from bug https://bugs.python.org/issue7904
91 # URLs with protocols not in urlparse.uses_netloc are not handled correctly
92 for scheme
in ('socks', 'socks4', 'socks4a', 'socks5'):
93 if scheme
not in compat_urlparse
.uses_netloc
:
94 compat_urlparse
.uses_netloc
.append(scheme
)
97 # This is not clearly defined otherwise
98 compiled_regex_type
= type(re
.compile(''))
101 def random_user_agent():
102 _USER_AGENT_TPL
= 'Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/%s Safari/537.36'
143 return _USER_AGENT_TPL
% random
.choice(_CHROME_VERSIONS
)
147 'User-Agent': random_user_agent(),
148 'Accept': 'text/html,application/xhtml+xml,application/xml;q=0.9,*/*;q=0.8',
149 'Accept-Encoding': 'gzip, deflate',
150 'Accept-Language': 'en-us,en;q=0.5',
151 'Sec-Fetch-Mode': 'navigate',
156 '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',
160 NO_DEFAULT
= object()
162 ENGLISH_MONTH_NAMES
= [
163 'January', 'February', 'March', 'April', 'May', 'June',
164 'July', 'August', 'September', 'October', 'November', 'December']
167 'en': ENGLISH_MONTH_NAMES
,
169 'janvier', 'février', 'mars', 'avril', 'mai', 'juin',
170 'juillet', 'août', 'septembre', 'octobre', 'novembre', 'décembre'],
174 'mp4', 'm4a', 'm4p', 'm4b', 'm4r', 'm4v', 'aac',
175 'flv', 'f4v', 'f4a', 'f4b',
176 'webm', 'ogg', 'ogv', 'oga', 'ogx', 'spx', 'opus',
177 'mkv', 'mka', 'mk3d',
186 'f4f', 'f4m', 'm3u8', 'smil')
188 # needed for sanitizing filenames in restricted mode
189 ACCENT_CHARS
= dict(zip('ÂÃÄÀÁÅÆÇÈÉÊËÌÍÎÏÐÑÒÓÔÕÖŐØŒÙÚÛÜŰÝÞßàáâãäåæçèéêëìíîïðñòóôõöőøœùúûüűýþÿ',
190 itertools
.chain('AAAAAA', ['AE'], 'CEEEEIIIIDNOOOOOOO', ['OE'], 'UUUUUY', ['TH', 'ss'],
191 'aaaaaa', ['ae'], 'ceeeeiiiionooooooo', ['oe'], 'uuuuuy', ['th'], 'y')))
221 '%Y-%m-%d %H:%M:%S.%f',
222 '%Y-%m-%d %H:%M:%S:%f',
225 '%Y-%m-%dT%H:%M:%SZ',
226 '%Y-%m-%dT%H:%M:%S.%fZ',
227 '%Y-%m-%dT%H:%M:%S.%f0Z',
229 '%Y-%m-%dT%H:%M:%S.%f',
232 '%b %d %Y at %H:%M:%S',
234 '%B %d %Y at %H:%M:%S',
238 DATE_FORMATS_DAY_FIRST
= list(DATE_FORMATS
)
239 DATE_FORMATS_DAY_FIRST
.extend([
248 DATE_FORMATS_MONTH_FIRST
= list(DATE_FORMATS
)
249 DATE_FORMATS_MONTH_FIRST
.extend([
257 PACKED_CODES_RE
= r
"}\('(.+)',(\d+),(\d+),'([^']+)'\.split\('\|'\)"
258 JSON_LD_RE
= r
'(?is)<script[^>]+type=(["\']?
)application
/ld\
+json\
1[^
>]*>(?P
<json_ld
>.+?
)</script
>'
261 def preferredencoding():
262 """Get preferred encoding.
264 Returns the best encoding scheme for the system, based on
265 locale.getpreferredencoding() and some further tweaks.
268 pref = locale.getpreferredencoding()
276 def write_json_file(obj, fn):
277 """ Encode obj as JSON and write it to fn, atomically if possible """
279 fn = encodeFilename(fn)
280 if sys.version_info < (3, 0) and sys.platform != 'win32
':
281 encoding = get_filesystem_encoding()
282 # os.path.basename returns a bytes object, but NamedTemporaryFile
283 # will fail if the filename contains non ascii characters unless we
284 # use a unicode object
285 path_basename = lambda f: os.path.basename(fn).decode(encoding)
286 # the same for os.path.dirname
287 path_dirname = lambda f: os.path.dirname(fn).decode(encoding)
289 path_basename = os.path.basename
290 path_dirname = os.path.dirname
294 'prefix
': path_basename(fn) + '.',
295 'dir': path_dirname(fn),
299 # In Python 2.x, json.dump expects a bytestream.
300 # In Python 3.x, it writes to a character stream
301 if sys.version_info < (3, 0):
309 tf = tempfile.NamedTemporaryFile(**compat_kwargs(args))
313 json.dump(obj, tf, ensure_ascii=False)
314 if sys.platform == 'win32
':
315 # Need to remove existing file on Windows, else os.rename raises
316 # WindowsError or FileExistsError.
324 os.chmod(tf.name, 0o666 & ~mask)
327 os.rename(tf.name, fn)
336 if sys.version_info >= (2, 7):
337 def find_xpath_attr(node, xpath, key, val=None):
338 """ Find the xpath xpath[@key=val] """
339 assert re.match(r'^
[a
-zA
-Z_
-]+$
', key)
340 expr = xpath + ('[@%s]' % key if val is None else "[@%s='%s']" % (key, val))
341 return node.find(expr)
343 def find_xpath_attr(node, xpath, key, val=None):
344 for f in node.findall(compat_xpath(xpath)):
345 if key not in f.attrib:
347 if val is None or f.attrib.get(key) == val:
351 # On python2.6 the xml.etree.ElementTree.Element methods don't support
352 # the namespace parameter
355 def xpath_with_ns(path
, ns_map
):
356 components
= [c
.split(':') for c
in path
.split('/')]
360 replaced
.append(c
[0])
363 replaced
.append('{%s}%s' % (ns_map
[ns
], tag
))
364 return '/'.join(replaced
)
367 def xpath_element(node
, xpath
, name
=None, fatal
=False, default
=NO_DEFAULT
):
368 def _find_xpath(xpath
):
369 return node
.find(compat_xpath(xpath
))
371 if isinstance(xpath
, (str, compat_str
)):
372 n
= _find_xpath(xpath
)
380 if default
is not NO_DEFAULT
:
383 name
= xpath
if name
is None else name
384 raise ExtractorError('Could not find XML element %s' % name
)
390 def xpath_text(node
, xpath
, name
=None, fatal
=False, default
=NO_DEFAULT
):
391 n
= xpath_element(node
, xpath
, name
, fatal
=fatal
, default
=default
)
392 if n
is None or n
== default
:
395 if default
is not NO_DEFAULT
:
398 name
= xpath
if name
is None else name
399 raise ExtractorError('Could not find XML element\'s text %s' % name
)
405 def xpath_attr(node
, xpath
, key
, name
=None, fatal
=False, default
=NO_DEFAULT
):
406 n
= find_xpath_attr(node
, xpath
, key
)
408 if default
is not NO_DEFAULT
:
411 name
= '%s[@%s]' % (xpath
, key
) if name
is None else name
412 raise ExtractorError('Could not find XML attribute %s' % name
)
418 def get_element_by_id(id, html
):
419 """Return the content of the tag with the specified ID in the passed HTML document"""
420 return get_element_by_attribute('id', id, html
)
423 def get_element_html_by_id(id, html
):
424 """Return the html of the tag with the specified ID in the passed HTML document"""
425 return get_element_html_by_attribute('id', id, html
)
428 def get_element_by_class(class_name
, html
):
429 """Return the content of the first tag with the specified class in the passed HTML document"""
430 retval
= get_elements_by_class(class_name
, html
)
431 return retval
[0] if retval
else None
434 def get_element_html_by_class(class_name
, html
):
435 """Return the html of the first tag with the specified class in the passed HTML document"""
436 retval
= get_elements_html_by_class(class_name
, html
)
437 return retval
[0] if retval
else None
440 def get_element_by_attribute(attribute
, value
, html
, escape_value
=True):
441 retval
= get_elements_by_attribute(attribute
, value
, html
, escape_value
)
442 return retval
[0] if retval
else None
445 def get_element_html_by_attribute(attribute
, value
, html
, escape_value
=True):
446 retval
= get_elements_html_by_attribute(attribute
, value
, html
, escape_value
)
447 return retval
[0] if retval
else None
450 def get_elements_by_class(class_name
, html
):
451 """Return the content of all tags with the specified class in the passed HTML document as a list"""
452 return get_elements_by_attribute(
453 'class', r
'[^\'"]*\b%s\b[^\'"]*' % re.escape(class_name),
454 html, escape_value=False)
457 def get_elements_html_by_class(class_name, html):
458 """Return the html of all tags with the specified class in the passed HTML document as a list"""
459 return get_elements_html_by_attribute(
460 'class', r'[^
\'"]*\b%s\b[^\'"]*' % re.escape(class_name),
461 html, escape_value=False)
464 def get_elements_by_attribute(*args, **kwargs):
465 """Return the content of the tag with the specified attribute in the passed HTML document"""
466 return [content for content, _ in get_elements_text_and_html_by_attribute(*args, **kwargs)]
469 def get_elements_html_by_attribute(*args, **kwargs):
470 """Return the html of the tag with the specified attribute in the passed HTML document"""
471 return [whole for _, whole in get_elements_text_and_html_by_attribute(*args, **kwargs)]
474 def get_elements_text_and_html_by_attribute(attribute, value, html, escape_value=True):
476 Return the text (content) and the html (whole) of the tag with the specified
477 attribute in the passed HTML document
480 value_quote_optional = '' if re.match(r'''[\s"'`
=<>]''', value) else '?'
482 value = re.escape(value) if escape_value else value
484 partial_element_re = r'''(?x
)
485 <(?P
<tag
>[a
-zA
-Z0
-9:._-]+)
486 (?
:\
s(?
:[^
>"']|"[^
"]*"|
'[^']*')*)?
487 \s%(attribute)s\s*=\s*(?P<_q>['"]%(vqo)s)(?-x:%(value)s)(?P=_q)
488 ''' % {'attribute': re.escape(attribute), 'value': value, 'vqo': value_quote_optional}
490 for m in re.finditer(partial_element_re, html):
491 content, whole = get_element_text_and_html_by_tag(m.group('tag'), html[m.start():])
494 unescapeHTML(re.sub(r'^(?P<q>["\'])(?P
<content
>.*)(?P
=q
)$
', r'\g
<content
>', content, flags=re.DOTALL)),
499 class HTMLBreakOnClosingTagParser(compat_HTMLParser):
501 HTML parser which raises HTMLBreakOnClosingTagException upon reaching the
502 closing tag for the first opening tag it has encountered, and can be used
506 class HTMLBreakOnClosingTagException(Exception):
510 self.tagstack = collections.deque()
511 compat_HTMLParser.__init__(self)
516 def __exit__(self, *_):
520 # handle_endtag does not return upon raising HTMLBreakOnClosingTagException,
521 # so data remains buffered; we no longer have any interest in it, thus
522 # override this method to discard it
525 def handle_starttag(self, tag, _):
526 self.tagstack.append(tag)
528 def handle_endtag(self, tag):
529 if not self.tagstack:
530 raise compat_HTMLParseError('no tags
in the stack
')
532 inner_tag = self.tagstack.pop()
536 raise compat_HTMLParseError(f'matching opening tag
for closing {tag} tag
not found
')
537 if not self.tagstack:
538 raise self.HTMLBreakOnClosingTagException()
541 def get_element_text_and_html_by_tag(tag, html):
543 For the first element with the specified tag in the passed HTML document
544 return its' content (text
) and the whole
element (html
)
546 def find_or_raise(haystack, needle, exc):
548 return haystack.index(needle)
551 closing_tag = f'</{tag}>'
552 whole_start = find_or_raise(
553 html, f'<{tag}', compat_HTMLParseError(f'opening {tag} tag not found'))
554 content_start = find_or_raise(
555 html[whole_start:], '>', compat_HTMLParseError(f'malformed opening {tag} tag'))
556 content_start += whole_start + 1
557 with HTMLBreakOnClosingTagParser() as parser:
558 parser.feed(html[whole_start:content_start])
559 if not parser.tagstack or parser.tagstack[0] != tag:
560 raise compat_HTMLParseError(f'parser did not match opening {tag} tag')
561 offset = content_start
562 while offset < len(html):
563 next_closing_tag_start = find_or_raise(
564 html[offset:], closing_tag,
565 compat_HTMLParseError(f'closing {tag} tag not found'))
566 next_closing_tag_end = next_closing_tag_start + len(closing_tag)
568 parser.feed(html[offset:offset + next_closing_tag_end])
569 offset += next_closing_tag_end
570 except HTMLBreakOnClosingTagParser.HTMLBreakOnClosingTagException:
571 return html[content_start:offset + next_closing_tag_start], \
572 html[whole_start:offset + next_closing_tag_end]
573 raise compat_HTMLParseError('unexpected end of html')
576 class HTMLAttributeParser(compat_HTMLParser):
577 """Trivial HTML parser to gather the attributes
for a single element
"""
581 compat_HTMLParser.__init__(self)
583 def handle_starttag(self, tag, attrs):
584 self.attrs = dict(attrs)
587 class HTMLListAttrsParser(compat_HTMLParser):
588 """HTML parser to gather the attributes
for the elements of a
list"""
591 compat_HTMLParser.__init__(self)
595 def handle_starttag(self, tag, attrs):
596 if tag == 'li' and self._level == 0:
597 self.items.append(dict(attrs))
600 def handle_endtag(self, tag):
604 def extract_attributes(html_element):
605 """Given a string
for an HTML element such
as
607 a
="foo" B
="bar" c
="&98;az" d
=boz
608 empty
= noval entity
="&"
611 Decode
and return a dictionary of attributes
.
613 'a': 'foo', 'b': 'bar', c
: 'baz', d
: 'boz',
614 'empty': '', 'noval': None, 'entity': '&',
615 'sq': '"', 'dq': '\''
617 NB HTMLParser
is stricter
in Python
2.6 & 3.2 than
in later versions
,
618 but the cases
in the unit test will work
for all of
2.6, 2.7, 3.2-3.5.
620 parser = HTMLAttributeParser()
622 parser.feed(html_element)
624 # Older Python may throw HTMLParseError in case of malformed HTML
625 except compat_HTMLParseError:
630 def parse_list(webpage):
631 """Given a string
for an series of HTML
<li
> elements
,
632 return a dictionary of their attributes
"""
633 parser = HTMLListAttrsParser()
639 def clean_html(html):
640 """Clean an HTML snippet into a readable string
"""
642 if html is None: # Convenience for sanitizing descriptions etc.
645 html = re.sub(r'\s+', ' ', html)
646 html = re.sub(r'(?u)\s?<\s?br\s?/?\s?>\s?', '\n', html)
647 html = re.sub(r'(?u)<\s?/\s?p\s?>\s?<\s?p[^>]*>', '\n', html)
649 html = re.sub('<.*?>', '', html)
650 # Replace html entities
651 html = unescapeHTML(html)
655 def sanitize_open(filename, open_mode):
656 """Try to
open the given filename
, and slightly tweak it
if this fails
.
658 Attempts to
open the given filename
. If this fails
, it tries to change
659 the filename slightly
, step by step
, until it
's either able to open it
660 or it fails and raises a final exception, like the standard open()
663 It returns the tuple (stream, definitive_file_name).
667 if sys.platform == 'win32
':
669 msvcrt.setmode(sys.stdout.fileno(), os.O_BINARY)
670 return (sys.stdout.buffer if hasattr(sys.stdout, 'buffer') else sys.stdout, filename)
671 stream = locked_file(filename, open_mode, block=False).open()
672 return (stream, filename)
673 except (IOError, OSError) as err:
674 if err.errno in (errno.EACCES,):
677 # In case of error, try to remove win32 forbidden chars
678 alt_filename = sanitize_path(filename)
679 if alt_filename == filename:
682 # An exception here should be caught in the caller
683 stream = locked_file(filename, open_mode, block=False).open()
684 return (stream, alt_filename)
687 def timeconvert(timestr):
688 """Convert RFC 2822 defined time string into system timestamp"""
690 timetuple = email.utils.parsedate_tz(timestr)
691 if timetuple is not None:
692 timestamp = email.utils.mktime_tz(timetuple)
696 def sanitize_filename(s, restricted=False, is_id=False):
697 """Sanitizes a string so it could be used as part of a filename.
698 If restricted is set, use a stricter subset of allowed characters.
699 Set is_id if this is not an arbitrary string, but an ID that should be kept
702 def replace_insane(char):
703 if restricted and char in ACCENT_CHARS:
704 return ACCENT_CHARS[char]
705 elif not restricted and char == '\n':
707 elif char == '?
' or ord(char) < 32 or ord(char) == 127:
710 return '' if restricted else '\''
712 return '_-' if restricted else ' -'
713 elif char in '\\/|*<>':
715 if restricted and (char in '!&\'()[]{}$;`^,#' or char.isspace()):
717 if restricted and ord(char) > 127:
724 s = re.sub(r'[0-9]+(?::[0-9]+)+', lambda m: m.group(0).replace(':', '_'), s)
725 result = ''.join(map(replace_insane, s))
727 while '__' in result:
728 result = result.replace('__', '_')
729 result = result.strip('_')
730 # Common case of "Foreign band name
- English song title
"
731 if restricted and result.startswith('-_'):
733 if result.startswith('-'):
734 result = '_' + result[len('-'):]
735 result = result.lstrip('.')
741 def sanitize_path(s, force=False):
742 """Sanitizes and normalizes path on Windows"""
743 if sys.platform == 'win32':
745 drive_or_unc, _ = os.path.splitdrive(s)
746 if sys.version_info < (2, 7) and not drive_or_unc:
747 drive_or_unc, _ = os.path.splitunc(s)
753 norm_path = os.path.normpath(remove_start(s, drive_or_unc)).split(os.path.sep)
757 path_part if path_part in ['.', '..'] else re.sub(r'(?:[/<>:"\|
\\?\
*]|
[\s
.]$
)', '#', path_part)
758 for path_part
in norm_path
]
760 sanitized_path
.insert(0, drive_or_unc
+ os
.path
.sep
)
761 elif force
and s
[0] == os
.path
.sep
:
762 sanitized_path
.insert(0, os
.path
.sep
)
763 return os
.path
.join(*sanitized_path
)
766 def sanitize_url(url
):
767 # Prepend protocol-less URLs with `http:` scheme in order to mitigate
768 # the number of unwanted failures due to missing protocol
769 if url
.startswith('//'):
770 return 'http:%s' % url
771 # Fix some common typos seen so far
773 # https://github.com/ytdl-org/youtube-dl/issues/15649
774 (r
'^httpss://', r
'https://'),
775 # https://bx1.be/lives/direct-tv/
776 (r
'^rmtp([es]?)://', r
'rtmp\1://'),
778 for mistake
, fixup
in COMMON_TYPOS
:
779 if re
.match(mistake
, url
):
780 return re
.sub(mistake
, fixup
, url
)
784 def extract_basic_auth(url
):
785 parts
= compat_urlparse
.urlsplit(url
)
786 if parts
.username
is None:
788 url
= compat_urlparse
.urlunsplit(parts
._replace
(netloc
=(
789 parts
.hostname
if parts
.port
is None
790 else '%s:%d' % (parts
.hostname
, parts
.port
))))
791 auth_payload
= base64
.b64encode(
792 ('%s:%s' % (parts
.username
, parts
.password
or '')).encode('utf-8'))
793 return url
, 'Basic ' + auth_payload
.decode('utf-8')
796 def sanitized_Request(url
, *args
, **kwargs
):
797 url
, auth_header
= extract_basic_auth(escape_url(sanitize_url(url
)))
798 if auth_header
is not None:
799 headers
= args
[1] if len(args
) >= 2 else kwargs
.setdefault('headers', {})
800 headers
['Authorization'] = auth_header
801 return compat_urllib_request
.Request(url
, *args
, **kwargs
)
805 """Expand shell variables and ~"""
806 return os
.path
.expandvars(compat_expanduser(s
))
809 def orderedSet(iterable
):
810 """ Remove all duplicates from the input iterable """
818 def _htmlentity_transform(entity_with_semicolon
):
819 """Transforms an HTML entity to a character."""
820 entity
= entity_with_semicolon
[:-1]
822 # Known non-numeric HTML entity
823 if entity
in compat_html_entities
.name2codepoint
:
824 return compat_chr(compat_html_entities
.name2codepoint
[entity
])
826 # TODO: HTML5 allows entities without a semicolon. For example,
827 # 'Éric' should be decoded as 'Éric'.
828 if entity_with_semicolon
in compat_html_entities_html5
:
829 return compat_html_entities_html5
[entity_with_semicolon
]
831 mobj
= re
.match(r
'#(x[0-9a-fA-F]+|[0-9]+)', entity
)
833 numstr
= mobj
.group(1)
834 if numstr
.startswith('x'):
836 numstr
= '0%s' % numstr
839 # See https://github.com/ytdl-org/youtube-dl/issues/7518
841 return compat_chr(int(numstr
, base
))
845 # Unknown entity in name, return its literal representation
846 return '&%s;' % entity
852 assert type(s
) == compat_str
855 r
'&([^&;]+;)', lambda m
: _htmlentity_transform(m
.group(1)), s
)
858 def escapeHTML(text
):
861 .replace('&', '&')
862 .replace('<', '<')
863 .replace('>', '>')
864 .replace('"', '"')
865 .replace("'", ''')
869 def process_communicate_or_kill(p
, *args
, **kwargs
):
871 return p
.communicate(*args
, **kwargs
)
872 except BaseException
: # Including KeyboardInterrupt
878 class Popen(subprocess
.Popen
):
879 if sys
.platform
== 'win32':
880 _startupinfo
= subprocess
.STARTUPINFO()
881 _startupinfo
.dwFlags |
= subprocess
.STARTF_USESHOWWINDOW
885 def __init__(self
, *args
, **kwargs
):
886 super(Popen
, self
).__init
__(*args
, **kwargs
, startupinfo
=self
._startupinfo
)
888 def communicate_or_kill(self
, *args
, **kwargs
):
889 return process_communicate_or_kill(self
, *args
, **kwargs
)
892 def get_subprocess_encoding():
893 if sys
.platform
== 'win32' and sys
.getwindowsversion()[0] >= 5:
894 # For subprocess calls, encode with locale encoding
895 # Refer to http://stackoverflow.com/a/9951851/35070
896 encoding
= preferredencoding()
898 encoding
= sys
.getfilesystemencoding()
904 def encodeFilename(s
, for_subprocess
=False):
906 @param s The name of the file
909 assert type(s
) == compat_str
911 # Python 3 has a Unicode API
912 if sys
.version_info
>= (3, 0):
915 # Pass '' directly to use Unicode APIs on Windows 2000 and up
916 # (Detecting Windows NT 4 is tricky because 'major >= 4' would
917 # match Windows 9x series as well. Besides, NT 4 is obsolete.)
918 if not for_subprocess
and sys
.platform
== 'win32' and sys
.getwindowsversion()[0] >= 5:
921 # Jython assumes filenames are Unicode strings though reported as Python 2.x compatible
922 if sys
.platform
.startswith('java'):
925 return s
.encode(get_subprocess_encoding(), 'ignore')
928 def decodeFilename(b
, for_subprocess
=False):
930 if sys
.version_info
>= (3, 0):
933 if not isinstance(b
, bytes):
936 return b
.decode(get_subprocess_encoding(), 'ignore')
939 def encodeArgument(s
):
940 if not isinstance(s
, compat_str
):
941 # Legacy code that uses byte strings
942 # Uncomment the following line after fixing all post processors
943 # assert False, 'Internal error: %r should be of type %r, is %r' % (s, compat_str, type(s))
944 s
= s
.decode('ascii')
945 return encodeFilename(s
, True)
948 def decodeArgument(b
):
949 return decodeFilename(b
, True)
952 def decodeOption(optval
):
955 if isinstance(optval
, bytes):
956 optval
= optval
.decode(preferredencoding())
958 assert isinstance(optval
, compat_str
)
962 _timetuple
= collections
.namedtuple('Time', ('hours', 'minutes', 'seconds', 'milliseconds'))
965 def timetuple_from_msec(msec
):
966 secs
, msec
= divmod(msec
, 1000)
967 mins
, secs
= divmod(secs
, 60)
968 hrs
, mins
= divmod(mins
, 60)
969 return _timetuple(hrs
, mins
, secs
, msec
)
972 def formatSeconds(secs
, delim
=':', msec
=False):
973 time
= timetuple_from_msec(secs
* 1000)
975 ret
= '%d%s%02d%s%02d' % (time
.hours
, delim
, time
.minutes
, delim
, time
.seconds
)
977 ret
= '%d%s%02d' % (time
.minutes
, delim
, time
.seconds
)
979 ret
= '%d' % time
.seconds
980 return '%s.%03d' % (ret
, time
.milliseconds
) if msec
else ret
983 def _ssl_load_windows_store_certs(ssl_context
, storename
):
984 # Code adapted from _load_windows_store_certs in https://github.com/python/cpython/blob/main/Lib/ssl.py
986 certs
= [cert
for cert
, encoding
, trust
in ssl
.enum_certificates(storename
)
987 if encoding
== 'x509_asn' and (
988 trust
is True or ssl
.Purpose
.SERVER_AUTH
.oid
in trust
)]
989 except PermissionError
:
993 ssl_context
.load_verify_locations(cadata
=cert
)
998 def make_HTTPS_handler(params
, **kwargs
):
999 opts_check_certificate
= not params
.get('nocheckcertificate')
1000 context
= ssl
.SSLContext(ssl
.PROTOCOL_TLS_CLIENT
)
1001 context
.check_hostname
= opts_check_certificate
1002 if params
.get('legacyserverconnect'):
1003 context
.options |
= 4 # SSL_OP_LEGACY_SERVER_CONNECT
1004 context
.verify_mode
= ssl
.CERT_REQUIRED
if opts_check_certificate
else ssl
.CERT_NONE
1005 if opts_check_certificate
:
1007 context
.load_default_certs()
1008 # Work around the issue in load_default_certs when there are bad certificates. See:
1009 # https://github.com/yt-dlp/yt-dlp/issues/1060,
1010 # https://bugs.python.org/issue35665, https://bugs.python.org/issue45312
1011 except ssl
.SSLError
:
1012 # enum_certificates is not present in mingw python. See https://github.com/yt-dlp/yt-dlp/issues/1151
1013 if sys
.platform
== 'win32' and hasattr(ssl
, 'enum_certificates'):
1014 # Create a new context to discard any certificates that were already loaded
1015 context
= ssl
.SSLContext(ssl
.PROTOCOL_TLS_CLIENT
)
1016 context
.check_hostname
, context
.verify_mode
= True, ssl
.CERT_REQUIRED
1017 for storename
in ('CA', 'ROOT'):
1018 _ssl_load_windows_store_certs(context
, storename
)
1019 context
.set_default_verify_paths()
1020 return YoutubeDLHTTPSHandler(params
, context
=context
, **kwargs
)
1023 def bug_reports_message(before
=';'):
1024 msg
= ('please report this issue on https://github.com/yt-dlp/yt-dlp , '
1025 'filling out the "Broken site" issue template properly. '
1026 'Confirm you are on the latest version using -U')
1028 before
= before
.rstrip()
1029 if not before
or before
.endswith(('.', '!', '?')):
1030 msg
= msg
[0].title() + msg
[1:]
1032 return (before
+ ' ' if before
else '') + msg
1035 class YoutubeDLError(Exception):
1036 """Base exception for YoutubeDL errors."""
1039 def __init__(self
, msg
=None):
1042 elif self
.msg
is None:
1043 self
.msg
= type(self
).__name
__
1044 super().__init
__(self
.msg
)
1047 network_exceptions
= [compat_urllib_error
.URLError
, compat_http_client
.HTTPException
, socket
.error
]
1048 if hasattr(ssl
, 'CertificateError'):
1049 network_exceptions
.append(ssl
.CertificateError
)
1050 network_exceptions
= tuple(network_exceptions
)
1053 class ExtractorError(YoutubeDLError
):
1054 """Error during info extraction."""
1056 def __init__(self
, msg
, tb
=None, expected
=False, cause
=None, video_id
=None, ie
=None):
1057 """ tb, if given, is the original traceback (so that it can be printed out).
1058 If expected is set, this is a normal error message and most likely not a bug in yt-dlp.
1060 if sys
.exc_info()[0] in network_exceptions
:
1065 self
.expected
= expected
1067 self
.video_id
= video_id
1069 self
.exc_info
= sys
.exc_info() # preserve original exception
1071 super(ExtractorError
, self
).__init
__(''.join((
1072 format_field(ie
, template
='[%s] '),
1073 format_field(video_id
, template
='%s: '),
1075 format_field(cause
, template
=' (caused by %r)'),
1076 '' if expected
else bug_reports_message())))
1078 def format_traceback(self
):
1079 if self
.traceback
is None:
1081 return ''.join(traceback
.format_tb(self
.traceback
))
1084 class UnsupportedError(ExtractorError
):
1085 def __init__(self
, url
):
1086 super(UnsupportedError
, self
).__init
__(
1087 'Unsupported URL: %s' % url
, expected
=True)
1091 class RegexNotFoundError(ExtractorError
):
1092 """Error when a regex didn't match"""
1096 class GeoRestrictedError(ExtractorError
):
1097 """Geographic restriction Error exception.
1099 This exception may be thrown when a video is not available from your
1100 geographic location due to geographic restrictions imposed by a website.
1103 def __init__(self
, msg
, countries
=None, **kwargs
):
1104 kwargs
['expected'] = True
1105 super(GeoRestrictedError
, self
).__init
__(msg
, **kwargs
)
1106 self
.countries
= countries
1109 class DownloadError(YoutubeDLError
):
1110 """Download Error exception.
1112 This exception may be thrown by FileDownloader objects if they are not
1113 configured to continue on errors. They will contain the appropriate
1117 def __init__(self
, msg
, exc_info
=None):
1118 """ exc_info, if given, is the original exception that caused the trouble (as returned by sys.exc_info()). """
1119 super(DownloadError
, self
).__init
__(msg
)
1120 self
.exc_info
= exc_info
1123 class EntryNotInPlaylist(YoutubeDLError
):
1124 """Entry not in playlist exception.
1126 This exception will be thrown by YoutubeDL when a requested entry
1127 is not found in the playlist info_dict
1129 msg
= 'Entry not found in info'
1132 class SameFileError(YoutubeDLError
):
1133 """Same File exception.
1135 This exception will be thrown by FileDownloader objects if they detect
1136 multiple files would have to be downloaded to the same file on disk.
1138 msg
= 'Fixed output name but more than one file to download'
1140 def __init__(self
, filename
=None):
1141 if filename
is not None:
1142 self
.msg
+= f
': {filename}'
1143 super().__init
__(self
.msg
)
1146 class PostProcessingError(YoutubeDLError
):
1147 """Post Processing exception.
1149 This exception may be raised by PostProcessor's .run() method to
1150 indicate an error in the postprocessing task.
1154 class DownloadCancelled(YoutubeDLError
):
1155 """ Exception raised when the download queue should be interrupted """
1156 msg
= 'The download was cancelled'
1159 class ExistingVideoReached(DownloadCancelled
):
1160 """ --break-on-existing triggered """
1161 msg
= 'Encountered a video that is already in the archive, stopping due to --break-on-existing'
1164 class RejectedVideoReached(DownloadCancelled
):
1165 """ --break-on-reject triggered """
1166 msg
= 'Encountered a video that did not match filter, stopping due to --break-on-reject'
1169 class MaxDownloadsReached(DownloadCancelled
):
1170 """ --max-downloads limit has been reached. """
1171 msg
= 'Maximum number of downloads reached, stopping due to --max-downloads'
1174 class ReExtractInfo(YoutubeDLError
):
1175 """ Video info needs to be re-extracted. """
1177 def __init__(self
, msg
, expected
=False):
1178 super().__init
__(msg
)
1179 self
.expected
= expected
1182 class ThrottledDownload(ReExtractInfo
):
1183 """ Download speed below --throttled-rate. """
1184 msg
= 'The download speed is below throttle limit'
1187 super().__init
__(self
.msg
, expected
=False)
1190 class UnavailableVideoError(YoutubeDLError
):
1191 """Unavailable Format exception.
1193 This exception will be thrown when a video is requested
1194 in a format that is not available for that video.
1196 msg
= 'Unable to download video'
1198 def __init__(self
, err
=None):
1200 self
.msg
+= f
': {err}'
1201 super().__init
__(self
.msg
)
1204 class ContentTooShortError(YoutubeDLError
):
1205 """Content Too Short exception.
1207 This exception may be raised by FileDownloader objects when a file they
1208 download is too small for what the server announced first, indicating
1209 the connection was probably interrupted.
1212 def __init__(self
, downloaded
, expected
):
1213 super(ContentTooShortError
, self
).__init
__(
1214 'Downloaded {0} bytes, expected {1} bytes'.format(downloaded
, expected
)
1217 self
.downloaded
= downloaded
1218 self
.expected
= expected
1221 class XAttrMetadataError(YoutubeDLError
):
1222 def __init__(self
, code
=None, msg
='Unknown error'):
1223 super(XAttrMetadataError
, self
).__init
__(msg
)
1227 # Parsing code and msg
1228 if (self
.code
in (errno
.ENOSPC
, errno
.EDQUOT
)
1229 or 'No space left' in self
.msg
or 'Disk quota exceeded' in self
.msg
):
1230 self
.reason
= 'NO_SPACE'
1231 elif self
.code
== errno
.E2BIG
or 'Argument list too long' in self
.msg
:
1232 self
.reason
= 'VALUE_TOO_LONG'
1234 self
.reason
= 'NOT_SUPPORTED'
1237 class XAttrUnavailableError(YoutubeDLError
):
1241 def _create_http_connection(ydl_handler
, http_class
, is_https
, *args
, **kwargs
):
1242 # Working around python 2 bug (see http://bugs.python.org/issue17849) by limiting
1243 # expected HTTP responses to meet HTTP/1.0 or later (see also
1244 # https://github.com/ytdl-org/youtube-dl/issues/6727)
1245 if sys
.version_info
< (3, 0):
1246 kwargs
['strict'] = True
1247 hc
= http_class(*args
, **compat_kwargs(kwargs
))
1248 source_address
= ydl_handler
._params
.get('source_address')
1250 if source_address
is not None:
1251 # This is to workaround _create_connection() from socket where it will try all
1252 # address data from getaddrinfo() including IPv6. This filters the result from
1253 # getaddrinfo() based on the source_address value.
1254 # This is based on the cpython socket.create_connection() function.
1255 # https://github.com/python/cpython/blob/master/Lib/socket.py#L691
1256 def _create_connection(address
, timeout
=socket
._GLOBAL
_DEFAULT
_TIMEOUT
, source_address
=None):
1257 host
, port
= address
1259 addrs
= socket
.getaddrinfo(host
, port
, 0, socket
.SOCK_STREAM
)
1260 af
= socket
.AF_INET
if '.' in source_address
[0] else socket
.AF_INET6
1261 ip_addrs
= [addr
for addr
in addrs
if addr
[0] == af
]
1262 if addrs
and not ip_addrs
:
1263 ip_version
= 'v4' if af
== socket
.AF_INET
else 'v6'
1265 "No remote IP%s addresses available for connect, can't use '%s' as source address"
1266 % (ip_version
, source_address
[0]))
1267 for res
in ip_addrs
:
1268 af
, socktype
, proto
, canonname
, sa
= res
1271 sock
= socket
.socket(af
, socktype
, proto
)
1272 if timeout
is not socket
._GLOBAL
_DEFAULT
_TIMEOUT
:
1273 sock
.settimeout(timeout
)
1274 sock
.bind(source_address
)
1276 err
= None # Explicitly break reference cycle
1278 except socket
.error
as _
:
1280 if sock
is not None:
1285 raise socket
.error('getaddrinfo returns an empty list')
1286 if hasattr(hc
, '_create_connection'):
1287 hc
._create
_connection
= _create_connection
1288 sa
= (source_address
, 0)
1289 if hasattr(hc
, 'source_address'): # Python 2.7+
1290 hc
.source_address
= sa
1292 def _hc_connect(self
, *args
, **kwargs
):
1293 sock
= _create_connection(
1294 (self
.host
, self
.port
), self
.timeout
, sa
)
1296 self
.sock
= ssl
.wrap_socket(
1297 sock
, self
.key_file
, self
.cert_file
,
1298 ssl_version
=ssl
.PROTOCOL_TLSv1
)
1301 hc
.connect
= functools
.partial(_hc_connect
, hc
)
1306 def handle_youtubedl_headers(headers
):
1307 filtered_headers
= headers
1309 if 'Youtubedl-no-compression' in filtered_headers
:
1310 filtered_headers
= dict((k
, v
) for k
, v
in filtered_headers
.items() if k
.lower() != 'accept-encoding')
1311 del filtered_headers
['Youtubedl-no-compression']
1313 return filtered_headers
1316 class YoutubeDLHandler(compat_urllib_request
.HTTPHandler
):
1317 """Handler for HTTP requests and responses.
1319 This class, when installed with an OpenerDirector, automatically adds
1320 the standard headers to every HTTP request and handles gzipped and
1321 deflated responses from web servers. If compression is to be avoided in
1322 a particular request, the original request in the program code only has
1323 to include the HTTP header "Youtubedl-no-compression", which will be
1324 removed before making the real request.
1326 Part of this code was copied from:
1328 http://techknack.net/python-urllib2-handlers/
1330 Andrew Rowls, the author of that code, agreed to release it to the
1334 def __init__(self
, params
, *args
, **kwargs
):
1335 compat_urllib_request
.HTTPHandler
.__init
__(self
, *args
, **kwargs
)
1336 self
._params
= params
1338 def http_open(self
, req
):
1339 conn_class
= compat_http_client
.HTTPConnection
1341 socks_proxy
= req
.headers
.get('Ytdl-socks-proxy')
1343 conn_class
= make_socks_conn_class(conn_class
, socks_proxy
)
1344 del req
.headers
['Ytdl-socks-proxy']
1346 return self
.do_open(functools
.partial(
1347 _create_http_connection
, self
, conn_class
, False),
1355 return zlib
.decompress(data
, -zlib
.MAX_WBITS
)
1357 return zlib
.decompress(data
)
1359 def http_request(self
, req
):
1360 # According to RFC 3986, URLs can not contain non-ASCII characters, however this is not
1361 # always respected by websites, some tend to give out URLs with non percent-encoded
1362 # non-ASCII characters (see telemb.py, ard.py [#3412])
1363 # urllib chokes on URLs with non-ASCII characters (see http://bugs.python.org/issue3991)
1364 # To work around aforementioned issue we will replace request's original URL with
1365 # percent-encoded one
1366 # Since redirects are also affected (e.g. http://www.southpark.de/alle-episoden/s18e09)
1367 # the code of this workaround has been moved here from YoutubeDL.urlopen()
1368 url
= req
.get_full_url()
1369 url_escaped
= escape_url(url
)
1371 # Substitute URL if any change after escaping
1372 if url
!= url_escaped
:
1373 req
= update_Request(req
, url
=url_escaped
)
1375 for h
, v
in std_headers
.items():
1376 # Capitalize is needed because of Python bug 2275: http://bugs.python.org/issue2275
1377 # The dict keys are capitalized because of this bug by urllib
1378 if h
.capitalize() not in req
.headers
:
1379 req
.add_header(h
, v
)
1381 req
.headers
= handle_youtubedl_headers(req
.headers
)
1383 if sys
.version_info
< (2, 7) and '#' in req
.get_full_url():
1384 # Python 2.6 is brain-dead when it comes to fragments
1385 req
._Request
__original
= req
._Request
__original
.partition('#')[0]
1386 req
._Request
__r
_type
= req
._Request
__r
_type
.partition('#')[0]
1390 def http_response(self
, req
, resp
):
1393 if resp
.headers
.get('Content-encoding', '') == 'gzip':
1394 content
= resp
.read()
1395 gz
= gzip
.GzipFile(fileobj
=io
.BytesIO(content
), mode
='rb')
1397 uncompressed
= io
.BytesIO(gz
.read())
1398 except IOError as original_ioerror
:
1399 # There may be junk add the end of the file
1400 # See http://stackoverflow.com/q/4928560/35070 for details
1401 for i
in range(1, 1024):
1403 gz
= gzip
.GzipFile(fileobj
=io
.BytesIO(content
[:-i
]), mode
='rb')
1404 uncompressed
= io
.BytesIO(gz
.read())
1409 raise original_ioerror
1410 resp
= compat_urllib_request
.addinfourl(uncompressed
, old_resp
.headers
, old_resp
.url
, old_resp
.code
)
1411 resp
.msg
= old_resp
.msg
1412 del resp
.headers
['Content-encoding']
1414 if resp
.headers
.get('Content-encoding', '') == 'deflate':
1415 gz
= io
.BytesIO(self
.deflate(resp
.read()))
1416 resp
= compat_urllib_request
.addinfourl(gz
, old_resp
.headers
, old_resp
.url
, old_resp
.code
)
1417 resp
.msg
= old_resp
.msg
1418 del resp
.headers
['Content-encoding']
1419 # Percent-encode redirect URL of Location HTTP header to satisfy RFC 3986 (see
1420 # https://github.com/ytdl-org/youtube-dl/issues/6457).
1421 if 300 <= resp
.code
< 400:
1422 location
= resp
.headers
.get('Location')
1424 # As of RFC 2616 default charset is iso-8859-1 that is respected by python 3
1425 if sys
.version_info
>= (3, 0):
1426 location
= location
.encode('iso-8859-1').decode('utf-8')
1428 location
= location
.decode('utf-8')
1429 location_escaped
= escape_url(location
)
1430 if location
!= location_escaped
:
1431 del resp
.headers
['Location']
1432 if sys
.version_info
< (3, 0):
1433 location_escaped
= location_escaped
.encode('utf-8')
1434 resp
.headers
['Location'] = location_escaped
1437 https_request
= http_request
1438 https_response
= http_response
1441 def make_socks_conn_class(base_class
, socks_proxy
):
1442 assert issubclass(base_class
, (
1443 compat_http_client
.HTTPConnection
, compat_http_client
.HTTPSConnection
))
1445 url_components
= compat_urlparse
.urlparse(socks_proxy
)
1446 if url_components
.scheme
.lower() == 'socks5':
1447 socks_type
= ProxyType
.SOCKS5
1448 elif url_components
.scheme
.lower() in ('socks', 'socks4'):
1449 socks_type
= ProxyType
.SOCKS4
1450 elif url_components
.scheme
.lower() == 'socks4a':
1451 socks_type
= ProxyType
.SOCKS4A
1453 def unquote_if_non_empty(s
):
1456 return compat_urllib_parse_unquote_plus(s
)
1460 url_components
.hostname
, url_components
.port
or 1080,
1462 unquote_if_non_empty(url_components
.username
),
1463 unquote_if_non_empty(url_components
.password
),
1466 class SocksConnection(base_class
):
1468 self
.sock
= sockssocket()
1469 self
.sock
.setproxy(*proxy_args
)
1470 if type(self
.timeout
) in (int, float):
1471 self
.sock
.settimeout(self
.timeout
)
1472 self
.sock
.connect((self
.host
, self
.port
))
1474 if isinstance(self
, compat_http_client
.HTTPSConnection
):
1475 if hasattr(self
, '_context'): # Python > 2.6
1476 self
.sock
= self
._context
.wrap_socket(
1477 self
.sock
, server_hostname
=self
.host
)
1479 self
.sock
= ssl
.wrap_socket(self
.sock
)
1481 return SocksConnection
1484 class YoutubeDLHTTPSHandler(compat_urllib_request
.HTTPSHandler
):
1485 def __init__(self
, params
, https_conn_class
=None, *args
, **kwargs
):
1486 compat_urllib_request
.HTTPSHandler
.__init
__(self
, *args
, **kwargs
)
1487 self
._https
_conn
_class
= https_conn_class
or compat_http_client
.HTTPSConnection
1488 self
._params
= params
1490 def https_open(self
, req
):
1492 conn_class
= self
._https
_conn
_class
1494 if hasattr(self
, '_context'): # python > 2.6
1495 kwargs
['context'] = self
._context
1496 if hasattr(self
, '_check_hostname'): # python 3.x
1497 kwargs
['check_hostname'] = self
._check
_hostname
1499 socks_proxy
= req
.headers
.get('Ytdl-socks-proxy')
1501 conn_class
= make_socks_conn_class(conn_class
, socks_proxy
)
1502 del req
.headers
['Ytdl-socks-proxy']
1504 return self
.do_open(functools
.partial(
1505 _create_http_connection
, self
, conn_class
, True),
1509 class YoutubeDLCookieJar(compat_cookiejar
.MozillaCookieJar
):
1511 See [1] for cookie file format.
1513 1. https://curl.haxx.se/docs/http-cookies.html
1515 _HTTPONLY_PREFIX
= '#HttpOnly_'
1517 _HEADER
= '''# Netscape HTTP Cookie File
1518 # This file is generated by yt-dlp. Do not edit.
1521 _CookieFileEntry
= collections
.namedtuple(
1523 ('domain_name', 'include_subdomains', 'path', 'https_only', 'expires_at', 'name', 'value'))
1525 def save(self
, filename
=None, ignore_discard
=False, ignore_expires
=False):
1527 Save cookies to a file.
1529 Most of the code is taken from CPython 3.8 and slightly adapted
1530 to support cookie files with UTF-8 in both python 2 and 3.
1532 if filename
is None:
1533 if self
.filename
is not None:
1534 filename
= self
.filename
1536 raise ValueError(compat_cookiejar
.MISSING_FILENAME_TEXT
)
1538 # Store session cookies with `expires` set to 0 instead of an empty
1541 if cookie
.expires
is None:
1544 with io
.open(filename
, 'w', encoding
='utf-8') as f
:
1545 f
.write(self
._HEADER
)
1548 if not ignore_discard
and cookie
.discard
:
1550 if not ignore_expires
and cookie
.is_expired(now
):
1556 if cookie
.domain
.startswith('.'):
1557 initial_dot
= 'TRUE'
1559 initial_dot
= 'FALSE'
1560 if cookie
.expires
is not None:
1561 expires
= compat_str(cookie
.expires
)
1564 if cookie
.value
is None:
1565 # cookies.txt regards 'Set-Cookie: foo' as a cookie
1566 # with no name, whereas http.cookiejar regards it as a
1567 # cookie with no value.
1572 value
= cookie
.value
1574 '\t'.join([cookie
.domain
, initial_dot
, cookie
.path
,
1575 secure
, expires
, name
, value
]) + '\n')
1577 def load(self
, filename
=None, ignore_discard
=False, ignore_expires
=False):
1578 """Load cookies from a file."""
1579 if filename
is None:
1580 if self
.filename
is not None:
1581 filename
= self
.filename
1583 raise ValueError(compat_cookiejar
.MISSING_FILENAME_TEXT
)
1585 def prepare_line(line
):
1586 if line
.startswith(self
._HTTPONLY
_PREFIX
):
1587 line
= line
[len(self
._HTTPONLY
_PREFIX
):]
1588 # comments and empty lines are fine
1589 if line
.startswith('#') or not line
.strip():
1591 cookie_list
= line
.split('\t')
1592 if len(cookie_list
) != self
._ENTRY
_LEN
:
1593 raise compat_cookiejar
.LoadError('invalid length %d' % len(cookie_list
))
1594 cookie
= self
._CookieFileEntry
(*cookie_list
)
1595 if cookie
.expires_at
and not cookie
.expires_at
.isdigit():
1596 raise compat_cookiejar
.LoadError('invalid expires at %s' % cookie
.expires_at
)
1600 with io
.open(filename
, encoding
='utf-8') as f
:
1603 cf
.write(prepare_line(line
))
1604 except compat_cookiejar
.LoadError
as e
:
1606 'WARNING: skipping cookie file entry due to %s: %r\n'
1607 % (e
, line
), sys
.stderr
)
1610 self
._really
_load
(cf
, filename
, ignore_discard
, ignore_expires
)
1611 # Session cookies are denoted by either `expires` field set to
1612 # an empty string or 0. MozillaCookieJar only recognizes the former
1613 # (see [1]). So we need force the latter to be recognized as session
1614 # cookies on our own.
1615 # Session cookies may be important for cookies-based authentication,
1616 # e.g. usually, when user does not check 'Remember me' check box while
1617 # logging in on a site, some important cookies are stored as session
1618 # cookies so that not recognizing them will result in failed login.
1619 # 1. https://bugs.python.org/issue17164
1621 # Treat `expires=0` cookies as session cookies
1622 if cookie
.expires
== 0:
1623 cookie
.expires
= None
1624 cookie
.discard
= True
1627 class YoutubeDLCookieProcessor(compat_urllib_request
.HTTPCookieProcessor
):
1628 def __init__(self
, cookiejar
=None):
1629 compat_urllib_request
.HTTPCookieProcessor
.__init
__(self
, cookiejar
)
1631 def http_response(self
, request
, response
):
1632 # Python 2 will choke on next HTTP request in row if there are non-ASCII
1633 # characters in Set-Cookie HTTP header of last response (see
1634 # https://github.com/ytdl-org/youtube-dl/issues/6769).
1635 # In order to at least prevent crashing we will percent encode Set-Cookie
1636 # header before HTTPCookieProcessor starts processing it.
1637 # if sys.version_info < (3, 0) and response.headers:
1638 # for set_cookie_header in ('Set-Cookie', 'Set-Cookie2'):
1639 # set_cookie = response.headers.get(set_cookie_header)
1641 # set_cookie_escaped = compat_urllib_parse.quote(set_cookie, b"%/;:@&=+$,!~*'()?#[] ")
1642 # if set_cookie != set_cookie_escaped:
1643 # del response.headers[set_cookie_header]
1644 # response.headers[set_cookie_header] = set_cookie_escaped
1645 return compat_urllib_request
.HTTPCookieProcessor
.http_response(self
, request
, response
)
1647 https_request
= compat_urllib_request
.HTTPCookieProcessor
.http_request
1648 https_response
= http_response
1651 class YoutubeDLRedirectHandler(compat_urllib_request
.HTTPRedirectHandler
):
1652 """YoutubeDL redirect handler
1654 The code is based on HTTPRedirectHandler implementation from CPython [1].
1656 This redirect handler solves two issues:
1657 - ensures redirect URL is always unicode under python 2
1658 - introduces support for experimental HTTP response status code
1659 308 Permanent Redirect [2] used by some sites [3]
1661 1. https://github.com/python/cpython/blob/master/Lib/urllib/request.py
1662 2. https://developer.mozilla.org/en-US/docs/Web/HTTP/Status/308
1663 3. https://github.com/ytdl-org/youtube-dl/issues/28768
1666 http_error_301
= http_error_303
= http_error_307
= http_error_308
= compat_urllib_request
.HTTPRedirectHandler
.http_error_302
1668 def redirect_request(self
, req
, fp
, code
, msg
, headers
, newurl
):
1669 """Return a Request or None in response to a redirect.
1671 This is called by the http_error_30x methods when a
1672 redirection response is received. If a redirection should
1673 take place, return a new Request to allow http_error_30x to
1674 perform the redirect. Otherwise, raise HTTPError if no-one
1675 else should try to handle this url. Return None if you can't
1676 but another Handler might.
1678 m
= req
.get_method()
1679 if (not (code
in (301, 302, 303, 307, 308) and m
in ("GET", "HEAD")
1680 or code
in (301, 302, 303) and m
== "POST")):
1681 raise compat_HTTPError(req
.full_url
, code
, msg
, headers
, fp
)
1682 # Strictly (according to RFC 2616), 301 or 302 in response to
1683 # a POST MUST NOT cause a redirection without confirmation
1684 # from the user (of urllib.request, in this case). In practice,
1685 # essentially all clients do redirect in this case, so we do
1688 # On python 2 urlh.geturl() may sometimes return redirect URL
1689 # as byte string instead of unicode. This workaround allows
1690 # to force it always return unicode.
1691 if sys
.version_info
[0] < 3:
1692 newurl
= compat_str(newurl
)
1694 # Be conciliant with URIs containing a space. This is mainly
1695 # redundant with the more complete encoding done in http_error_302(),
1696 # but it is kept for compatibility with other callers.
1697 newurl
= newurl
.replace(' ', '%20')
1699 CONTENT_HEADERS
= ("content-length", "content-type")
1700 # NB: don't use dict comprehension for python 2.6 compatibility
1701 newheaders
= dict((k
, v
) for k
, v
in req
.headers
.items()
1702 if k
.lower() not in CONTENT_HEADERS
)
1703 return compat_urllib_request
.Request(
1704 newurl
, headers
=newheaders
, origin_req_host
=req
.origin_req_host
,
1708 def extract_timezone(date_str
):
1711 ^.{8,}? # >=8 char non-TZ prefix, if present
1712 (?P<tz>Z| # just the UTC Z, or
1713 (?:(?<=.\b\d{4}|\b\d{2}:\d\d)| # preceded by 4 digits or hh:mm or
1714 (?<!.\b[a-zA-Z]{3}|[a-zA-Z]{4}|..\b\d\d)) # not preceded by 3 alpha word or >= 4 alpha or 2 digits
1715 [ ]? # optional space
1716 (?P<sign>\+|-) # +/-
1717 (?P<hours>[0-9]{2}):?(?P<minutes>[0-9]{2}) # hh[:]mm
1721 timezone
= datetime
.timedelta()
1723 date_str
= date_str
[:-len(m
.group('tz'))]
1724 if not m
.group('sign'):
1725 timezone
= datetime
.timedelta()
1727 sign
= 1 if m
.group('sign') == '+' else -1
1728 timezone
= datetime
.timedelta(
1729 hours
=sign
* int(m
.group('hours')),
1730 minutes
=sign
* int(m
.group('minutes')))
1731 return timezone
, date_str
1734 def parse_iso8601(date_str
, delimiter
='T', timezone
=None):
1735 """ Return a UNIX timestamp from the given date """
1737 if date_str
is None:
1740 date_str
= re
.sub(r
'\.[0-9]+', '', date_str
)
1742 if timezone
is None:
1743 timezone
, date_str
= extract_timezone(date_str
)
1746 date_format
= '%Y-%m-%d{0}%H:%M:%S'.format(delimiter
)
1747 dt
= datetime
.datetime
.strptime(date_str
, date_format
) - timezone
1748 return calendar
.timegm(dt
.timetuple())
1753 def date_formats(day_first
=True):
1754 return DATE_FORMATS_DAY_FIRST
if day_first
else DATE_FORMATS_MONTH_FIRST
1757 def unified_strdate(date_str
, day_first
=True):
1758 """Return a string with the date in the format YYYYMMDD"""
1760 if date_str
is None:
1764 date_str
= date_str
.replace(',', ' ')
1765 # Remove AM/PM + timezone
1766 date_str
= re
.sub(r
'(?i)\s*(?:AM|PM)(?:\s+[A-Z]+)?', '', date_str
)
1767 _
, date_str
= extract_timezone(date_str
)
1769 for expression
in date_formats(day_first
):
1771 upload_date
= datetime
.datetime
.strptime(date_str
, expression
).strftime('%Y%m%d')
1774 if upload_date
is None:
1775 timetuple
= email
.utils
.parsedate_tz(date_str
)
1778 upload_date
= datetime
.datetime(*timetuple
[:6]).strftime('%Y%m%d')
1781 if upload_date
is not None:
1782 return compat_str(upload_date
)
1785 def unified_timestamp(date_str
, day_first
=True):
1786 if date_str
is None:
1789 date_str
= re
.sub(r
'[,|]', '', date_str
)
1791 pm_delta
= 12 if re
.search(r
'(?i)PM', date_str
) else 0
1792 timezone
, date_str
= extract_timezone(date_str
)
1794 # Remove AM/PM + timezone
1795 date_str
= re
.sub(r
'(?i)\s*(?:AM|PM)(?:\s+[A-Z]+)?', '', date_str
)
1797 # Remove unrecognized timezones from ISO 8601 alike timestamps
1798 m
= re
.search(r
'\d{1,2}:\d{1,2}(?:\.\d+)?(?P<tz>\s*[A-Z]+)$', date_str
)
1800 date_str
= date_str
[:-len(m
.group('tz'))]
1802 # Python only supports microseconds, so remove nanoseconds
1803 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]+$', date_str
)
1805 date_str
= m
.group(1)
1807 for expression
in date_formats(day_first
):
1809 dt
= datetime
.datetime
.strptime(date_str
, expression
) - timezone
+ datetime
.timedelta(hours
=pm_delta
)
1810 return calendar
.timegm(dt
.timetuple())
1813 timetuple
= email
.utils
.parsedate_tz(date_str
)
1815 return calendar
.timegm(timetuple
) + pm_delta
* 3600
1818 def determine_ext(url
, default_ext
='unknown_video'):
1819 if url
is None or '.' not in url
:
1821 guess
= url
.partition('?')[0].rpartition('.')[2]
1822 if re
.match(r
'^[A-Za-z0-9]+$', guess
):
1824 # Try extract ext from URLs like http://example.com/foo/bar.mp4/?download
1825 elif guess
.rstrip('/') in KNOWN_EXTENSIONS
:
1826 return guess
.rstrip('/')
1831 def subtitles_filename(filename
, sub_lang
, sub_format
, expected_real_ext
=None):
1832 return replace_extension(filename
, sub_lang
+ '.' + sub_format
, expected_real_ext
)
1835 def datetime_from_str(date_str
, precision
='auto', format
='%Y%m%d'):
1837 Return a datetime object from a string in the format YYYYMMDD or
1838 (now|today|yesterday|date)[+-][0-9](microsecond|second|minute|hour|day|week|month|year)(s)?
1840 format: string date format used to return datetime object from
1841 precision: round the time portion of a datetime object.
1842 auto|microsecond|second|minute|hour|day.
1843 auto: round to the unit provided in date_str (if applicable).
1845 auto_precision
= False
1846 if precision
== 'auto':
1847 auto_precision
= True
1848 precision
= 'microsecond'
1849 today
= datetime_round(datetime
.datetime
.utcnow(), precision
)
1850 if date_str
in ('now', 'today'):
1852 if date_str
== 'yesterday':
1853 return today
- datetime
.timedelta(days
=1)
1855 r
'(?P<start>.+)(?P<sign>[+-])(?P<time>\d+)(?P<unit>microsecond|second|minute|hour|day|week|month|year)(s)?',
1857 if match
is not None:
1858 start_time
= datetime_from_str(match
.group('start'), precision
, format
)
1859 time
= int(match
.group('time')) * (-1 if match
.group('sign') == '-' else 1)
1860 unit
= match
.group('unit')
1861 if unit
== 'month' or unit
== 'year':
1862 new_date
= datetime_add_months(start_time
, time
* 12 if unit
== 'year' else time
)
1868 delta
= datetime
.timedelta(**{unit + 's': time}
)
1869 new_date
= start_time
+ delta
1871 return datetime_round(new_date
, unit
)
1874 return datetime_round(datetime
.datetime
.strptime(date_str
, format
), precision
)
1877 def date_from_str(date_str
, format
='%Y%m%d', strict
=False):
1879 Return a datetime object from a string in the format YYYYMMDD or
1880 (now|today|yesterday|date)[+-][0-9](microsecond|second|minute|hour|day|week|month|year)(s)?
1882 If "strict", only (now|today)[+-][0-9](day|week|month|year)(s)? is allowed
1884 format: string date format used to return datetime object from
1886 if strict
and not re
.fullmatch(r
'\d{8}|(now|today)[+-]\d+(day|week|month|year)(s)?', date_str
):
1887 raise ValueError(f
'Invalid date format {date_str}')
1888 return datetime_from_str(date_str
, precision
='microsecond', format
=format
).date()
1891 def datetime_add_months(dt
, months
):
1892 """Increment/Decrement a datetime object by months."""
1893 month
= dt
.month
+ months
- 1
1894 year
= dt
.year
+ month
// 12
1895 month
= month
% 12 + 1
1896 day
= min(dt
.day
, calendar
.monthrange(year
, month
)[1])
1897 return dt
.replace(year
, month
, day
)
1900 def datetime_round(dt
, precision
='day'):
1902 Round a datetime object's time to a specific precision
1904 if precision
== 'microsecond':
1913 roundto
= lambda x
, n
: ((x
+ n
/ 2) // n
) * n
1914 timestamp
= calendar
.timegm(dt
.timetuple())
1915 return datetime
.datetime
.utcfromtimestamp(roundto(timestamp
, unit_seconds
[precision
]))
1918 def hyphenate_date(date_str
):
1920 Convert a date in 'YYYYMMDD' format to 'YYYY-MM-DD' format"""
1921 match
= re
.match(r
'^(\d\d\d\d)(\d\d)(\d\d)$', date_str
)
1922 if match
is not None:
1923 return '-'.join(match
.groups())
1928 class DateRange(object):
1929 """Represents a time interval between two dates"""
1931 def __init__(self
, start
=None, end
=None):
1932 """start and end must be strings in the format accepted by date"""
1933 if start
is not None:
1934 self
.start
= date_from_str(start
, strict
=True)
1936 self
.start
= datetime
.datetime
.min.date()
1938 self
.end
= date_from_str(end
, strict
=True)
1940 self
.end
= datetime
.datetime
.max.date()
1941 if self
.start
> self
.end
:
1942 raise ValueError('Date range: "%s" , the start date must be before the end date' % self
)
1946 """Returns a range that only contains the given day"""
1947 return cls(day
, day
)
1949 def __contains__(self
, date
):
1950 """Check if the date is in the range"""
1951 if not isinstance(date
, datetime
.date
):
1952 date
= date_from_str(date
)
1953 return self
.start
<= date
<= self
.end
1956 return '%s - %s' % (self
.start
.isoformat(), self
.end
.isoformat())
1959 def platform_name():
1960 """ Returns the platform name as a compat_str """
1961 res
= platform
.platform()
1962 if isinstance(res
, bytes):
1963 res
= res
.decode(preferredencoding())
1965 assert isinstance(res
, compat_str
)
1969 def get_windows_version():
1970 ''' Get Windows version. None if it's not running on Windows '''
1971 if compat_os_name
== 'nt':
1972 return version_tuple(platform
.win32_ver()[1])
1977 def _windows_write_string(s
, out
):
1978 """ Returns True if the string was written using special methods,
1979 False if it has yet to be written out."""
1980 # Adapted from http://stackoverflow.com/a/3259271/35070
1982 import ctypes
.wintypes
1990 fileno
= out
.fileno()
1991 except AttributeError:
1992 # If the output stream doesn't have a fileno, it's virtual
1994 except io
.UnsupportedOperation
:
1995 # Some strange Windows pseudo files?
1997 if fileno
not in WIN_OUTPUT_IDS
:
2000 GetStdHandle
= compat_ctypes_WINFUNCTYPE(
2001 ctypes
.wintypes
.HANDLE
, ctypes
.wintypes
.DWORD
)(
2002 ('GetStdHandle', ctypes
.windll
.kernel32
))
2003 h
= GetStdHandle(WIN_OUTPUT_IDS
[fileno
])
2005 WriteConsoleW
= compat_ctypes_WINFUNCTYPE(
2006 ctypes
.wintypes
.BOOL
, ctypes
.wintypes
.HANDLE
, ctypes
.wintypes
.LPWSTR
,
2007 ctypes
.wintypes
.DWORD
, ctypes
.POINTER(ctypes
.wintypes
.DWORD
),
2008 ctypes
.wintypes
.LPVOID
)(('WriteConsoleW', ctypes
.windll
.kernel32
))
2009 written
= ctypes
.wintypes
.DWORD(0)
2011 GetFileType
= compat_ctypes_WINFUNCTYPE(ctypes
.wintypes
.DWORD
, ctypes
.wintypes
.DWORD
)(('GetFileType', ctypes
.windll
.kernel32
))
2012 FILE_TYPE_CHAR
= 0x0002
2013 FILE_TYPE_REMOTE
= 0x8000
2014 GetConsoleMode
= compat_ctypes_WINFUNCTYPE(
2015 ctypes
.wintypes
.BOOL
, ctypes
.wintypes
.HANDLE
,
2016 ctypes
.POINTER(ctypes
.wintypes
.DWORD
))(
2017 ('GetConsoleMode', ctypes
.windll
.kernel32
))
2018 INVALID_HANDLE_VALUE
= ctypes
.wintypes
.DWORD(-1).value
2020 def not_a_console(handle
):
2021 if handle
== INVALID_HANDLE_VALUE
or handle
is None:
2023 return ((GetFileType(handle
) & ~FILE_TYPE_REMOTE
) != FILE_TYPE_CHAR
2024 or GetConsoleMode(handle
, ctypes
.byref(ctypes
.wintypes
.DWORD())) == 0)
2026 if not_a_console(h
):
2029 def next_nonbmp_pos(s
):
2031 return next(i
for i
, c
in enumerate(s
) if ord(c
) > 0xffff)
2032 except StopIteration:
2036 count
= min(next_nonbmp_pos(s
), 1024)
2038 ret
= WriteConsoleW(
2039 h
, s
, count
if count
else 2, ctypes
.byref(written
), None)
2041 raise OSError('Failed to write string')
2042 if not count
: # We just wrote a non-BMP character
2043 assert written
.value
== 2
2046 assert written
.value
> 0
2047 s
= s
[written
.value
:]
2051 def write_string(s
, out
=None, encoding
=None):
2054 assert type(s
) == compat_str
2056 if sys
.platform
== 'win32' and encoding
is None and hasattr(out
, 'fileno'):
2057 if _windows_write_string(s
, out
):
2060 if ('b' in getattr(out
, 'mode', '')
2061 or sys
.version_info
[0] < 3): # Python 2 lies about mode of sys.stderr
2062 byt
= s
.encode(encoding
or preferredencoding(), 'ignore')
2064 elif hasattr(out
, 'buffer'):
2065 enc
= encoding
or getattr(out
, 'encoding', None) or preferredencoding()
2066 byt
= s
.encode(enc
, 'ignore')
2067 out
.buffer.write(byt
)
2073 def bytes_to_intlist(bs
):
2076 if isinstance(bs
[0], int): # Python 3
2079 return [ord(c
) for c
in bs
]
2082 def intlist_to_bytes(xs
):
2085 return compat_struct_pack('%dB' % len(xs
), *xs
)
2088 # Cross-platform file locking
2089 if sys
.platform
== 'win32':
2090 import ctypes
.wintypes
2093 class OVERLAPPED(ctypes
.Structure
):
2095 ('Internal', ctypes
.wintypes
.LPVOID
),
2096 ('InternalHigh', ctypes
.wintypes
.LPVOID
),
2097 ('Offset', ctypes
.wintypes
.DWORD
),
2098 ('OffsetHigh', ctypes
.wintypes
.DWORD
),
2099 ('hEvent', ctypes
.wintypes
.HANDLE
),
2102 kernel32
= ctypes
.windll
.kernel32
2103 LockFileEx
= kernel32
.LockFileEx
2104 LockFileEx
.argtypes
= [
2105 ctypes
.wintypes
.HANDLE
, # hFile
2106 ctypes
.wintypes
.DWORD
, # dwFlags
2107 ctypes
.wintypes
.DWORD
, # dwReserved
2108 ctypes
.wintypes
.DWORD
, # nNumberOfBytesToLockLow
2109 ctypes
.wintypes
.DWORD
, # nNumberOfBytesToLockHigh
2110 ctypes
.POINTER(OVERLAPPED
) # Overlapped
2112 LockFileEx
.restype
= ctypes
.wintypes
.BOOL
2113 UnlockFileEx
= kernel32
.UnlockFileEx
2114 UnlockFileEx
.argtypes
= [
2115 ctypes
.wintypes
.HANDLE
, # hFile
2116 ctypes
.wintypes
.DWORD
, # dwReserved
2117 ctypes
.wintypes
.DWORD
, # nNumberOfBytesToLockLow
2118 ctypes
.wintypes
.DWORD
, # nNumberOfBytesToLockHigh
2119 ctypes
.POINTER(OVERLAPPED
) # Overlapped
2121 UnlockFileEx
.restype
= ctypes
.wintypes
.BOOL
2122 whole_low
= 0xffffffff
2123 whole_high
= 0x7fffffff
2125 def _lock_file(f
, exclusive
, block
):
2126 overlapped
= OVERLAPPED()
2127 overlapped
.Offset
= 0
2128 overlapped
.OffsetHigh
= 0
2129 overlapped
.hEvent
= 0
2130 f
._lock
_file
_overlapped
_p
= ctypes
.pointer(overlapped
)
2132 if not LockFileEx(msvcrt
.get_osfhandle(f
.fileno()),
2133 (0x2 if exclusive
else 0x0) |
(0x0 if block
else 0x1),
2134 0, whole_low
, whole_high
, f
._lock
_file
_overlapped
_p
):
2135 raise BlockingIOError('Locking file failed: %r' % ctypes
.FormatError())
2137 def _unlock_file(f
):
2138 assert f
._lock
_file
_overlapped
_p
2139 handle
= msvcrt
.get_osfhandle(f
.fileno())
2140 if not UnlockFileEx(handle
, 0, whole_low
, whole_high
, f
._lock
_file
_overlapped
_p
):
2141 raise OSError('Unlocking file failed: %r' % ctypes
.FormatError())
2147 def _lock_file(f
, exclusive
, block
):
2150 fcntl
.LOCK_SH
if not exclusive
2151 else fcntl
.LOCK_EX
if block
2152 else fcntl
.LOCK_EX | fcntl
.LOCK_NB
)
2153 except BlockingIOError
:
2155 except OSError: # AOSP does not have flock()
2157 fcntl
.LOCK_SH
if not exclusive
2158 else fcntl
.LOCK_EX
if block
2159 else fcntl
.LOCK_EX | fcntl
.LOCK_NB
)
2161 def _unlock_file(f
):
2163 fcntl
.flock(f
, fcntl
.LOCK_UN
)
2165 fcntl
.lockf(f
, fcntl
.LOCK_UN
)
2168 UNSUPPORTED_MSG
= 'file locking is not supported on this platform'
2170 def _lock_file(f
, exclusive
, block
):
2171 raise IOError(UNSUPPORTED_MSG
)
2173 def _unlock_file(f
):
2174 raise IOError(UNSUPPORTED_MSG
)
2177 class locked_file(object):
2180 def __init__(self
, filename
, mode
, block
=True, encoding
=None):
2181 assert mode
in ['r', 'rb', 'a', 'ab', 'w', 'wb']
2182 self
.f
= io
.open(filename
, mode
, encoding
=encoding
)
2186 def __enter__(self
):
2187 exclusive
= 'r' not in self
.mode
2189 _lock_file(self
.f
, exclusive
, self
.block
)
2195 def __exit__(self
, etype
, value
, traceback
):
2197 if not self
._closed
:
2198 _unlock_file(self
.f
)
2206 def write(self
, *args
):
2207 return self
.f
.write(*args
)
2209 def read(self
, *args
):
2210 return self
.f
.read(*args
)
2216 return self
.__enter
__()
2218 def close(self
, *args
):
2219 self
.__exit
__(self
, *args
, value
=False, traceback
=False)
2222 def get_filesystem_encoding():
2223 encoding
= sys
.getfilesystemencoding()
2224 return encoding
if encoding
is not None else 'utf-8'
2227 def shell_quote(args
):
2229 encoding
= get_filesystem_encoding()
2231 if isinstance(a
, bytes):
2232 # We may get a filename encoded with 'encodeFilename'
2233 a
= a
.decode(encoding
)
2234 quoted_args
.append(compat_shlex_quote(a
))
2235 return ' '.join(quoted_args
)
2238 def smuggle_url(url
, data
):
2239 """ Pass additional data in a URL for internal use. """
2241 url
, idata
= unsmuggle_url(url
, {})
2243 sdata
= compat_urllib_parse_urlencode(
2244 {'__youtubedl_smuggle': json.dumps(data)}
)
2245 return url
+ '#' + sdata
2248 def unsmuggle_url(smug_url
, default
=None):
2249 if '#__youtubedl_smuggle' not in smug_url
:
2250 return smug_url
, default
2251 url
, _
, sdata
= smug_url
.rpartition('#')
2252 jsond
= compat_parse_qs(sdata
)['__youtubedl_smuggle'][0]
2253 data
= json
.loads(jsond
)
2257 def format_decimal_suffix(num
, fmt
='%d%s', *, factor
=1000):
2258 """ Formats numbers with decimal sufixes like K, M, etc """
2259 num
, factor
= float_or_none(num
), float(factor
)
2262 exponent
= 0 if num
== 0 else int(math
.log(num
, factor
))
2263 suffix
= ['', *'kMGTPEZY'][exponent
]
2265 suffix
= {'k': 'Ki', '': ''}
.get(suffix
, f
'{suffix}i')
2266 converted
= num
/ (factor
** exponent
)
2267 return fmt
% (converted
, suffix
)
2270 def format_bytes(bytes):
2271 return format_decimal_suffix(bytes, '%.2f%sB', factor
=1024) or 'N/A'
2274 def lookup_unit_table(unit_table
, s
):
2275 units_re
= '|'.join(re
.escape(u
) for u
in unit_table
)
2277 r
'(?P<num>[0-9]+(?:[,.][0-9]*)?)\s*(?P<unit>%s)\b' % units_re
, s
)
2280 num_str
= m
.group('num').replace(',', '.')
2281 mult
= unit_table
[m
.group('unit')]
2282 return int(float(num_str
) * mult
)
2285 def parse_filesize(s
):
2289 # The lower-case forms are of course incorrect and unofficial,
2290 # but we support those too
2307 'megabytes': 1000 ** 2,
2308 'mebibytes': 1024 ** 2,
2314 'gigabytes': 1000 ** 3,
2315 'gibibytes': 1024 ** 3,
2321 'terabytes': 1000 ** 4,
2322 'tebibytes': 1024 ** 4,
2328 'petabytes': 1000 ** 5,
2329 'pebibytes': 1024 ** 5,
2335 'exabytes': 1000 ** 6,
2336 'exbibytes': 1024 ** 6,
2342 'zettabytes': 1000 ** 7,
2343 'zebibytes': 1024 ** 7,
2349 'yottabytes': 1000 ** 8,
2350 'yobibytes': 1024 ** 8,
2353 return lookup_unit_table(_UNIT_TABLE
, s
)
2360 s
= re
.sub(r
'^[^\d]+\s', '', s
).strip()
2362 if re
.match(r
'^[\d,.]+$', s
):
2363 return str_to_int(s
)
2376 ret
= lookup_unit_table(_UNIT_TABLE
, s
)
2380 mobj
= re
.match(r
'([\d,.]+)(?:$|\s)', s
)
2382 return str_to_int(mobj
.group(1))
2385 def parse_resolution(s
):
2389 mobj
= re
.search(r
'(?<![a-zA-Z0-9])(?P<w>\d+)\s*[xX×,]\s*(?P<h>\d+)(?![a-zA-Z0-9])', s
)
2392 'width': int(mobj
.group('w')),
2393 'height': int(mobj
.group('h')),
2396 mobj
= re
.search(r
'(?<![a-zA-Z0-9])(\d+)[pPiI](?![a-zA-Z0-9])', s
)
2398 return {'height': int(mobj.group(1))}
2400 mobj
= re
.search(r
'\b([48])[kK]\b', s
)
2402 return {'height': int(mobj.group(1)) * 540}
2407 def parse_bitrate(s
):
2408 if not isinstance(s
, compat_str
):
2410 mobj
= re
.search(r
'\b(\d+)\s*kbps', s
)
2412 return int(mobj
.group(1))
2415 def month_by_name(name
, lang
='en'):
2416 """ Return the number of a month by (locale-independently) English name """
2418 month_names
= MONTH_NAMES
.get(lang
, MONTH_NAMES
['en'])
2421 return month_names
.index(name
) + 1
2426 def month_by_abbreviation(abbrev
):
2427 """ Return the number of a month by (locale-independently) English
2431 return [s
[:3] for s
in ENGLISH_MONTH_NAMES
].index(abbrev
) + 1
2436 def fix_xml_ampersands(xml_str
):
2437 """Replace all the '&' by '&' in XML"""
2439 r
'&(?!amp;|lt;|gt;|apos;|quot;|#x[0-9a-fA-F]{,4};|#[0-9]{,4};)',
2444 def setproctitle(title
):
2445 assert isinstance(title
, compat_str
)
2447 # ctypes in Jython is not complete
2448 # http://bugs.jython.org/issue2148
2449 if sys
.platform
.startswith('java'):
2453 libc
= ctypes
.cdll
.LoadLibrary('libc.so.6')
2457 # LoadLibrary in Windows Python 2.7.13 only expects
2458 # a bytestring, but since unicode_literals turns
2459 # every string into a unicode string, it fails.
2461 title_bytes
= title
.encode('utf-8')
2462 buf
= ctypes
.create_string_buffer(len(title_bytes
))
2463 buf
.value
= title_bytes
2465 libc
.prctl(15, buf
, 0, 0, 0)
2466 except AttributeError:
2467 return # Strange libc, just skip this
2470 def remove_start(s
, start
):
2471 return s
[len(start
):] if s
is not None and s
.startswith(start
) else s
2474 def remove_end(s
, end
):
2475 return s
[:-len(end
)] if s
is not None and s
.endswith(end
) else s
2478 def remove_quotes(s
):
2479 if s
is None or len(s
) < 2:
2481 for quote
in ('"', "'", ):
2482 if s
[0] == quote
and s
[-1] == quote
:
2487 def get_domain(url
):
2488 domain
= re
.match(r
'(?:https?:\/\/)?(?:www\.)?(?P<domain>[^\n\/]+\.[^\n\/]+)(?:\/(.*))?', url
)
2489 return domain
.group('domain') if domain
else None
2492 def url_basename(url
):
2493 path
= compat_urlparse
.urlparse(url
).path
2494 return path
.strip('/').split('/')[-1]
2498 return re
.match(r
'https?://[^?#&]+/', url
).group()
2501 def urljoin(base
, path
):
2502 if isinstance(path
, bytes):
2503 path
= path
.decode('utf-8')
2504 if not isinstance(path
, compat_str
) or not path
:
2506 if re
.match(r
'^(?:[a-zA-Z][a-zA-Z0-9+-.]*:)?//', path
):
2508 if isinstance(base
, bytes):
2509 base
= base
.decode('utf-8')
2510 if not isinstance(base
, compat_str
) or not re
.match(
2511 r
'^(?:https?:)?//', base
):
2513 return compat_urlparse
.urljoin(base
, path
)
2516 class HEADRequest(compat_urllib_request
.Request
):
2517 def get_method(self
):
2521 class PUTRequest(compat_urllib_request
.Request
):
2522 def get_method(self
):
2526 def int_or_none(v
, scale
=1, default
=None, get_attr
=None, invscale
=1):
2527 if get_attr
and v
is not None:
2528 v
= getattr(v
, get_attr
, None)
2530 return int(v
) * invscale
// scale
2531 except (ValueError, TypeError, OverflowError):
2535 def str_or_none(v
, default
=None):
2536 return default
if v
is None else compat_str(v
)
2539 def str_to_int(int_str
):
2540 """ A more relaxed version of int_or_none """
2541 if isinstance(int_str
, compat_integer_types
):
2543 elif isinstance(int_str
, compat_str
):
2544 int_str
= re
.sub(r
'[,\.\+]', '', int_str
)
2545 return int_or_none(int_str
)
2548 def float_or_none(v
, scale
=1, invscale
=1, default
=None):
2552 return float(v
) * invscale
/ scale
2553 except (ValueError, TypeError):
2557 def bool_or_none(v
, default
=None):
2558 return v
if isinstance(v
, bool) else default
2561 def strip_or_none(v
, default
=None):
2562 return v
.strip() if isinstance(v
, compat_str
) else default
2565 def url_or_none(url
):
2566 if not url
or not isinstance(url
, compat_str
):
2569 return url
if re
.match(r
'^(?:(?:https?|rt(?:m(?:pt?[es]?|fp)|sp[su]?)|mms|ftps?):)?//', url
) else None
2572 def request_to_url(req
):
2573 if isinstance(req
, compat_urllib_request
.Request
):
2574 return req
.get_full_url()
2579 def strftime_or_none(timestamp
, date_format
, default
=None):
2580 datetime_object
= None
2582 if isinstance(timestamp
, compat_numeric_types
): # unix timestamp
2583 datetime_object
= datetime
.datetime
.utcfromtimestamp(timestamp
)
2584 elif isinstance(timestamp
, compat_str
): # assume YYYYMMDD
2585 datetime_object
= datetime
.datetime
.strptime(timestamp
, '%Y%m%d')
2586 return datetime_object
.strftime(date_format
)
2587 except (ValueError, TypeError, AttributeError):
2591 def parse_duration(s
):
2592 if not isinstance(s
, compat_basestring
):
2598 days
, hours
, mins
, secs
, ms
= [None] * 5
2599 m
= re
.match(r
'''(?x)
2601 (?:(?:(?P<days>[0-9]+):)?(?P<hours>[0-9]+):)?(?P<mins>[0-9]+):)?
2602 (?P<secs>(?(before_secs)[0-9]{1,2}|[0-9]+))
2603 (?P<ms>[.:][0-9]+)?Z?$
2606 days
, hours
, mins
, secs
, ms
= m
.group('days', 'hours', 'mins', 'secs', 'ms')
2611 [0-9]+\s*y(?:ears?)?\s*
2614 [0-9]+\s*m(?:onths?)?\s*
2617 [0-9]+\s*w(?:eeks?)?\s*
2620 (?P<days>[0-9]+)\s*d(?:ays?)?\s*
2624 (?P<hours>[0-9]+)\s*h(?:ours?)?\s*
2627 (?P<mins>[0-9]+)\s*m(?:in(?:ute)?s?)?\s*
2630 (?P<secs>[0-9]+)(?P<ms>\.[0-9]+)?\s*s(?:ec(?:ond)?s?)?\s*
2633 days
, hours
, mins
, secs
, ms
= m
.groups()
2635 m
= re
.match(r
'(?i)(?:(?P<hours>[0-9.]+)\s*(?:hours?)|(?P<mins>[0-9.]+)\s*(?:mins?\.?|minutes?)\s*)Z?$', s
)
2637 hours
, mins
= m
.groups()
2643 duration
+= float(secs
)
2645 duration
+= float(mins
) * 60
2647 duration
+= float(hours
) * 60 * 60
2649 duration
+= float(days
) * 24 * 60 * 60
2651 duration
+= float(ms
.replace(':', '.'))
2655 def prepend_extension(filename
, ext
, expected_real_ext
=None):
2656 name
, real_ext
= os
.path
.splitext(filename
)
2658 '{0}.{1}{2}'.format(name
, ext
, real_ext
)
2659 if not expected_real_ext
or real_ext
[1:] == expected_real_ext
2660 else '{0}.{1}'.format(filename
, ext
))
2663 def replace_extension(filename
, ext
, expected_real_ext
=None):
2664 name
, real_ext
= os
.path
.splitext(filename
)
2665 return '{0}.{1}'.format(
2666 name
if not expected_real_ext
or real_ext
[1:] == expected_real_ext
else filename
,
2670 def check_executable(exe
, args
=[]):
2671 """ Checks if the given binary is installed somewhere in PATH, and returns its name.
2672 args can be a list of arguments for a short output (like -version) """
2674 Popen([exe
] + args
, stdout
=subprocess
.PIPE
, stderr
=subprocess
.PIPE
).communicate_or_kill()
2680 def _get_exe_version_output(exe
, args
):
2682 # STDIN should be redirected too. On UNIX-like systems, ffmpeg triggers
2683 # SIGTTOU if yt-dlp is run in the background.
2684 # See https://github.com/ytdl-org/youtube-dl/issues/955#issuecomment-209789656
2686 [encodeArgument(exe
)] + args
, stdin
=subprocess
.PIPE
,
2687 stdout
=subprocess
.PIPE
, stderr
=subprocess
.STDOUT
).communicate_or_kill()
2690 if isinstance(out
, bytes): # Python 2.x
2691 out
= out
.decode('ascii', 'ignore')
2695 def detect_exe_version(output
, version_re
=None, unrecognized
='present'):
2696 assert isinstance(output
, compat_str
)
2697 if version_re
is None:
2698 version_re
= r
'version\s+([-0-9._a-zA-Z]+)'
2699 m
= re
.search(version_re
, output
)
2706 def get_exe_version(exe
, args
=['--version'],
2707 version_re
=None, unrecognized
='present'):
2708 """ Returns the version of the specified executable,
2709 or False if the executable is not present """
2710 out
= _get_exe_version_output(exe
, args
)
2711 return detect_exe_version(out
, version_re
, unrecognized
) if out
else False
2714 class LazyList(collections
.abc
.Sequence
):
2715 ''' Lazy immutable list from an iterable
2716 Note that slices of a LazyList are lists and not LazyList'''
2718 class IndexError(IndexError):
2721 def __init__(self
, iterable
, *, reverse
=False, _cache
=None):
2722 self
.__iterable
= iter(iterable
)
2723 self
.__cache
= [] if _cache
is None else _cache
2724 self
.__reversed
= reverse
2728 # We need to consume the entire iterable to iterate in reverse
2729 yield from self
.exhaust()
2731 yield from self
.__cache
2732 for item
in self
.__iterable
:
2733 self
.__cache
.append(item
)
2736 def __exhaust(self
):
2737 self
.__cache
.extend(self
.__iterable
)
2738 # Discard the emptied iterable to make it pickle-able
2739 self
.__iterable
= []
2743 ''' Evaluate the entire iterable '''
2744 return self
.__exhaust
()[::-1 if self
.__reversed
else 1]
2747 def __reverse_index(x
):
2748 return None if x
is None else -(x
+ 1)
2750 def __getitem__(self
, idx
):
2751 if isinstance(idx
, slice):
2753 idx
= slice(self
.__reverse
_index
(idx
.start
), self
.__reverse
_index
(idx
.stop
), -(idx
.step
or 1))
2754 start
, stop
, step
= idx
.start
, idx
.stop
, idx
.step
or 1
2755 elif isinstance(idx
, int):
2757 idx
= self
.__reverse
_index
(idx
)
2758 start
, stop
, step
= idx
, idx
, 0
2760 raise TypeError('indices must be integers or slices')
2761 if ((start
or 0) < 0 or (stop
or 0) < 0
2762 or (start
is None and step
< 0)
2763 or (stop
is None and step
> 0)):
2764 # We need to consume the entire iterable to be able to slice from the end
2765 # Obviously, never use this with infinite iterables
2768 return self
.__cache
[idx
]
2769 except IndexError as e
:
2770 raise self
.IndexError(e
) from e
2771 n
= max(start
or 0, stop
or 0) - len(self
.__cache
) + 1
2773 self
.__cache
.extend(itertools
.islice(self
.__iterable
, n
))
2775 return self
.__cache
[idx
]
2776 except IndexError as e
:
2777 raise self
.IndexError(e
) from e
2781 self
[-1] if self
.__reversed
else self
[0]
2782 except self
.IndexError:
2788 return len(self
.__cache
)
2790 def __reversed__(self
):
2791 return type(self
)(self
.__iterable
, reverse
=not self
.__reversed
, _cache
=self
.__cache
)
2794 return type(self
)(self
.__iterable
, reverse
=self
.__reversed
, _cache
=self
.__cache
)
2797 # repr and str should mimic a list. So we exhaust the iterable
2798 return repr(self
.exhaust())
2801 return repr(self
.exhaust())
2806 class IndexError(IndexError):
2810 # This is only useful for tests
2811 return len(self
.getslice())
2813 def __init__(self
, pagefunc
, pagesize
, use_cache
=True):
2814 self
._pagefunc
= pagefunc
2815 self
._pagesize
= pagesize
2816 self
._pagecount
= float('inf')
2817 self
._use
_cache
= use_cache
2820 def getpage(self
, pagenum
):
2821 page_results
= self
._cache
.get(pagenum
)
2822 if page_results
is None:
2823 page_results
= [] if pagenum
> self
._pagecount
else list(self
._pagefunc
(pagenum
))
2825 self
._cache
[pagenum
] = page_results
2828 def getslice(self
, start
=0, end
=None):
2829 return list(self
._getslice
(start
, end
))
2831 def _getslice(self
, start
, end
):
2832 raise NotImplementedError('This method must be implemented by subclasses')
2834 def __getitem__(self
, idx
):
2835 assert self
._use
_cache
, 'Indexing PagedList requires cache'
2836 if not isinstance(idx
, int) or idx
< 0:
2837 raise TypeError('indices must be non-negative integers')
2838 entries
= self
.getslice(idx
, idx
+ 1)
2840 raise self
.IndexError()
2844 class OnDemandPagedList(PagedList
):
2845 def _getslice(self
, start
, end
):
2846 for pagenum
in itertools
.count(start
// self
._pagesize
):
2847 firstid
= pagenum
* self
._pagesize
2848 nextfirstid
= pagenum
* self
._pagesize
+ self
._pagesize
2849 if start
>= nextfirstid
:
2853 start
% self
._pagesize
2854 if firstid
<= start
< nextfirstid
2857 ((end
- 1) % self
._pagesize
) + 1
2858 if (end
is not None and firstid
<= end
<= nextfirstid
)
2862 page_results
= self
.getpage(pagenum
)
2864 self
._pagecount
= pagenum
- 1
2866 if startv
!= 0 or endv
is not None:
2867 page_results
= page_results
[startv
:endv
]
2868 yield from page_results
2870 # A little optimization - if current page is not "full", ie. does
2871 # not contain page_size videos then we can assume that this page
2872 # is the last one - there are no more ids on further pages -
2873 # i.e. no need to query again.
2874 if len(page_results
) + startv
< self
._pagesize
:
2877 # If we got the whole page, but the next page is not interesting,
2878 # break out early as well
2879 if end
== nextfirstid
:
2883 class InAdvancePagedList(PagedList
):
2884 def __init__(self
, pagefunc
, pagecount
, pagesize
):
2885 PagedList
.__init
__(self
, pagefunc
, pagesize
, True)
2886 self
._pagecount
= pagecount
2888 def _getslice(self
, start
, end
):
2889 start_page
= start
// self
._pagesize
2890 end_page
= self
._pagecount
if end
is None else min(self
._pagecount
, end
// self
._pagesize
+ 1)
2891 skip_elems
= start
- start_page
* self
._pagesize
2892 only_more
= None if end
is None else end
- start
2893 for pagenum
in range(start_page
, end_page
):
2894 page_results
= self
.getpage(pagenum
)
2896 page_results
= page_results
[skip_elems
:]
2898 if only_more
is not None:
2899 if len(page_results
) < only_more
:
2900 only_more
-= len(page_results
)
2902 yield from page_results
[:only_more
]
2904 yield from page_results
2907 def uppercase_escape(s
):
2908 unicode_escape
= codecs
.getdecoder('unicode_escape')
2910 r
'\\U[0-9a-fA-F]{8}',
2911 lambda m
: unicode_escape(m
.group(0))[0],
2915 def lowercase_escape(s
):
2916 unicode_escape
= codecs
.getdecoder('unicode_escape')
2918 r
'\\u[0-9a-fA-F]{4}',
2919 lambda m
: unicode_escape(m
.group(0))[0],
2923 def escape_rfc3986(s
):
2924 """Escape non-ASCII characters as suggested by RFC 3986"""
2925 if sys
.version_info
< (3, 0) and isinstance(s
, compat_str
):
2926 s
= s
.encode('utf-8')
2927 return compat_urllib_parse
.quote(s
, b
"%/;:@&=+$,!~*'()?#[]")
2930 def escape_url(url
):
2931 """Escape URL as suggested by RFC 3986"""
2932 url_parsed
= compat_urllib_parse_urlparse(url
)
2933 return url_parsed
._replace
(
2934 netloc
=url_parsed
.netloc
.encode('idna').decode('ascii'),
2935 path
=escape_rfc3986(url_parsed
.path
),
2936 params
=escape_rfc3986(url_parsed
.params
),
2937 query
=escape_rfc3986(url_parsed
.query
),
2938 fragment
=escape_rfc3986(url_parsed
.fragment
)
2943 return compat_parse_qs(compat_urllib_parse_urlparse(url
).query
)
2946 def read_batch_urls(batch_fd
):
2948 if not isinstance(url
, compat_str
):
2949 url
= url
.decode('utf-8', 'replace')
2950 BOM_UTF8
= ('\xef\xbb\xbf', '\ufeff')
2951 for bom
in BOM_UTF8
:
2952 if url
.startswith(bom
):
2953 url
= url
[len(bom
):]
2955 if not url
or url
.startswith(('#', ';', ']')):
2957 # "#" cannot be stripped out since it is part of the URI
2958 # However, it can be safely stipped out if follwing a whitespace
2959 return re
.split(r
'\s#', url
, 1)[0].rstrip()
2961 with contextlib
.closing(batch_fd
) as fd
:
2962 return [url
for url
in map(fixup
, fd
) if url
]
2965 def urlencode_postdata(*args
, **kargs
):
2966 return compat_urllib_parse_urlencode(*args
, **kargs
).encode('ascii')
2969 def update_url_query(url
, query
):
2972 parsed_url
= compat_urlparse
.urlparse(url
)
2973 qs
= compat_parse_qs(parsed_url
.query
)
2975 return compat_urlparse
.urlunparse(parsed_url
._replace
(
2976 query
=compat_urllib_parse_urlencode(qs
, True)))
2979 def update_Request(req
, url
=None, data
=None, headers
={}, query={}
):
2980 req_headers
= req
.headers
.copy()
2981 req_headers
.update(headers
)
2982 req_data
= data
or req
.data
2983 req_url
= update_url_query(url
or req
.get_full_url(), query
)
2984 req_get_method
= req
.get_method()
2985 if req_get_method
== 'HEAD':
2986 req_type
= HEADRequest
2987 elif req_get_method
== 'PUT':
2988 req_type
= PUTRequest
2990 req_type
= compat_urllib_request
.Request
2992 req_url
, data
=req_data
, headers
=req_headers
,
2993 origin_req_host
=req
.origin_req_host
, unverifiable
=req
.unverifiable
)
2994 if hasattr(req
, 'timeout'):
2995 new_req
.timeout
= req
.timeout
2999 def _multipart_encode_impl(data
, boundary
):
3000 content_type
= 'multipart/form-data; boundary=%s' % boundary
3003 for k
, v
in data
.items():
3004 out
+= b
'--' + boundary
.encode('ascii') + b
'\r\n'
3005 if isinstance(k
, compat_str
):
3006 k
= k
.encode('utf-8')
3007 if isinstance(v
, compat_str
):
3008 v
= v
.encode('utf-8')
3009 # RFC 2047 requires non-ASCII field names to be encoded, while RFC 7578
3010 # suggests sending UTF-8 directly. Firefox sends UTF-8, too
3011 content
= b
'Content-Disposition: form-data; name="' + k
+ b
'"\r\n\r\n' + v
+ b
'\r\n'
3012 if boundary
.encode('ascii') in content
:
3013 raise ValueError('Boundary overlaps with data')
3016 out
+= b
'--' + boundary
.encode('ascii') + b
'--\r\n'
3018 return out
, content_type
3021 def multipart_encode(data
, boundary
=None):
3023 Encode a dict to RFC 7578-compliant form-data
3026 A dict where keys and values can be either Unicode or bytes-like
3029 If specified a Unicode object, it's used as the boundary. Otherwise
3030 a random boundary is generated.
3032 Reference: https://tools.ietf.org/html/rfc7578
3034 has_specified_boundary
= boundary
is not None
3037 if boundary
is None:
3038 boundary
= '---------------' + str(random
.randrange(0x0fffffff, 0xffffffff))
3041 out
, content_type
= _multipart_encode_impl(data
, boundary
)
3044 if has_specified_boundary
:
3048 return out
, content_type
3051 def dict_get(d
, key_or_keys
, default
=None, skip_false_values
=True):
3052 if isinstance(key_or_keys
, (list, tuple)):
3053 for key
in key_or_keys
:
3054 if key
not in d
or d
[key
] is None or skip_false_values
and not d
[key
]:
3058 return d
.get(key_or_keys
, default
)
3061 def try_get(src
, getter
, expected_type
=None):
3062 for get
in variadic(getter
):
3065 except (AttributeError, KeyError, TypeError, IndexError):
3068 if expected_type
is None or isinstance(v
, expected_type
):
3072 def merge_dicts(*dicts
):
3074 for a_dict
in dicts
:
3075 for k
, v
in a_dict
.items():
3079 or (isinstance(v
, compat_str
) and v
3080 and isinstance(merged
[k
], compat_str
)
3081 and not merged
[k
])):
3086 def encode_compat_str(string
, encoding
=preferredencoding(), errors
='strict'):
3087 return string
if isinstance(string
, compat_str
) else compat_str(string
, encoding
, errors
)
3099 TV_PARENTAL_GUIDELINES
= {
3109 def parse_age_limit(s
):
3111 return s
if 0 <= s
<= 21 else None
3112 if not isinstance(s
, compat_basestring
):
3114 m
= re
.match(r
'^(?P<age>\d{1,2})\+?$', s
)
3116 return int(m
.group('age'))
3119 return US_RATINGS
[s
]
3120 m
= re
.match(r
'^TV[_-]?(%s)$' % '|'.join(k
[3:] for k
in TV_PARENTAL_GUIDELINES
), s
)
3122 return TV_PARENTAL_GUIDELINES
['TV-' + m
.group(1)]
3126 def strip_jsonp(code
):
3129 (?:window\.)?(?P<func_name>[a-zA-Z0-9_.$]*)
3130 (?:\s*&&\s*(?P=func_name))?
3131 \s*\(\s*(?P<callback_data>.*)\);?
3132 \s*?(?://[^\n]*)*$''',
3133 r
'\g<callback_data>', code
)
3136 def js_to_json(code
, vars={}):
3137 # vars is a dict of var, val pairs to substitute
3138 COMMENT_RE
= r
'/\*(?:(?!\*/).)*?\*/|//[^\n]*\n'
3139 SKIP_RE
= r
'\s*(?:{comment})?\s*'.format(comment
=COMMENT_RE
)
3141 (r
'(?s)^(0[xX][0-9a-fA-F]+){skip}:?$'.format(skip
=SKIP_RE
), 16),
3142 (r
'(?s)^(0+[0-7]+){skip}:?$'.format(skip
=SKIP_RE
), 8),
3147 if v
in ('true', 'false', 'null'):
3149 elif v
in ('undefined', 'void 0'):
3151 elif v
.startswith('/*') or v
.startswith('//') or v
.startswith('!') or v
== ',':
3154 if v
[0] in ("'", '"'):
3155 v
= re
.sub(r
'(?s)\\.|"', lambda m
: {
3160 }.get(m
.group(0), m
.group(0)), v
[1:-1])
3162 for regex
, base
in INTEGER_TABLE
:
3163 im
= re
.match(regex
, v
)
3165 i
= int(im
.group(1), base
)
3166 return '"%d":' % i
if v
.endswith(':') else '%d' % i
3173 code
= re
.sub(r
'new Date\((".+")\)', r
'\g<1>', code
)
3175 return re
.sub(r
'''(?sx)
3176 "(?:[^"\\]*(?:\\\\|\\['"nurtbfx/\n]))*[^"\\]*"|
3177 '(?:[^'\\]*(?:\\\\|\\['"nurtbfx/\n]))*[^'\\]*'|
3178 {comment}|,(?={skip}[\]}}])|
3179 void\s0|(?:(?<![0-9])[eE]|[a-df-zA-DF-Z_$])[.a-zA-Z_$0-9]*|
3180 \b(?:0[xX][0-9a-fA-F]+|0+[0-7]+)(?:{skip}:)?|
3183 '''.format(comment
=COMMENT_RE
, skip
=SKIP_RE
), fix_kv
, code
)
3186 def qualities(quality_ids
):
3187 """ Get a numeric quality value out of a list of possible values """
3190 return quality_ids
.index(qid
)
3196 POSTPROCESS_WHEN
= {'pre_process', 'after_filter', 'before_dl', 'after_move', 'post_process', 'after_video', 'playlist'}
3200 'default': '%(title)s [%(id)s].%(ext)s',
3201 'chapter': '%(title)s - %(section_number)03d %(section_title)s [%(id)s].%(ext)s',
3207 'description': 'description',
3208 'annotation': 'annotations.xml',
3209 'infojson': 'info.json',
3212 'pl_thumbnail': None,
3213 'pl_description': 'description',
3214 'pl_infojson': 'info.json',
3217 # As of [1] format syntax is:
3218 # %[mapping_key][conversion_flags][minimum_width][.precision][length_modifier]type
3219 # 1. https://docs.python.org/2/library/stdtypes.html#string-formatting
3220 STR_FORMAT_RE_TMPL
= r
'''(?x)
3221 (?<!%)(?P<prefix>(?:%%)*)
3223 (?P<has_key>\((?P<key>{0})\))?
3225 (?P<conversion>[#0\-+ ]+)?
3227 (?P<precision>\.\d+)?
3228 (?P<len_mod>[hlL])? # unused in python
3229 {1} # conversion type
3234 STR_FORMAT_TYPES
= 'diouxXeEfFgGcrs'
3237 def limit_length(s
, length
):
3238 """ Add ellipses to overly long strings """
3243 return s
[:length
- len(ELLIPSES
)] + ELLIPSES
3247 def version_tuple(v
):
3248 return tuple(int(e
) for e
in re
.split(r
'[-.]', v
))
3251 def is_outdated_version(version
, limit
, assume_new
=True):
3253 return not assume_new
3255 return version_tuple(version
) < version_tuple(limit
)
3257 return not assume_new
3260 def ytdl_is_updateable():
3261 """ Returns if yt-dlp can be updated with -U """
3263 from .update
import is_non_updateable
3265 return not is_non_updateable()
3268 def args_to_str(args
):
3269 # Get a short string representation for a subprocess command
3270 return ' '.join(compat_shlex_quote(a
) for a
in args
)
3273 def error_to_compat_str(err
):
3275 # On python 2 error byte string must be decoded with proper
3276 # encoding rather than ascii
3277 if sys
.version_info
[0] < 3:
3278 err_str
= err_str
.decode(preferredencoding())
3282 def mimetype2ext(mt
):
3286 mt
, _
, params
= mt
.partition(';')
3291 # Per RFC 3003, audio/mpeg can be .mp1, .mp2 or .mp3. Here use .mp3 as
3292 # it's the most popular one
3293 'audio/mpeg': 'mp3',
3294 'audio/x-wav': 'wav',
3296 'audio/wave': 'wav',
3299 ext
= FULL_MAP
.get(mt
)
3305 'smptett+xml': 'tt',
3309 'x-mp4-fragmented': 'mp4',
3310 'x-ms-sami': 'sami',
3313 'x-mpegurl': 'm3u8',
3314 'vnd.apple.mpegurl': 'm3u8',
3318 'vnd.ms-sstr+xml': 'ism',
3322 'filmstrip+json': 'fs',
3326 _
, _
, subtype
= mt
.rpartition('/')
3327 ext
= SUBTYPE_MAP
.get(subtype
.lower())
3338 _
, _
, suffix
= subtype
.partition('+')
3339 ext
= SUFFIX_MAP
.get(suffix
)
3343 return subtype
.replace('+', '.')
3346 def ext2mimetype(ext_or_url
):
3349 if '.' not in ext_or_url
:
3350 ext_or_url
= f
'file.{ext_or_url}'
3351 return mimetypes
.guess_type(ext_or_url
)[0]
3354 def parse_codecs(codecs_str
):
3355 # http://tools.ietf.org/html/rfc6381
3358 split_codecs
= list(filter(None, map(
3359 str.strip
, codecs_str
.strip().strip(',').split(','))))
3360 vcodec
, acodec
, tcodec
, hdr
= None, None, None, None
3361 for full_codec
in split_codecs
:
3362 parts
= full_codec
.split('.')
3363 codec
= parts
[0].replace('0', '')
3364 if codec
in ('avc1', 'avc2', 'avc3', 'avc4', 'vp9', 'vp8', 'hev1', 'hev2',
3365 'h263', 'h264', 'mp4v', 'hvc1', 'av1', 'theora', 'dvh1', 'dvhe'):
3367 vcodec
= '.'.join(parts
[:4]) if codec
in ('vp9', 'av1', 'hvc1') else full_codec
3368 if codec
in ('dvh1', 'dvhe'):
3370 elif codec
== 'av1' and len(parts
) > 3 and parts
[3] == '10':
3372 elif full_codec
.replace('0', '').startswith('vp9.2'):
3374 elif codec
in ('flac', 'mp4a', 'opus', 'vorbis', 'mp3', 'aac', 'ac-3', 'ec-3', 'eac3', 'dtsc', 'dtse', 'dtsh', 'dtsl'):
3377 elif codec
in ('stpp', 'wvtt',):
3381 write_string('WARNING: Unknown codec %s\n' % full_codec
, sys
.stderr
)
3382 if vcodec
or acodec
or tcodec
:
3384 'vcodec': vcodec
or 'none',
3385 'acodec': acodec
or 'none',
3386 'dynamic_range': hdr
,
3387 **({'tcodec': tcodec}
if tcodec
is not None else {}),
3389 elif len(split_codecs
) == 2:
3391 'vcodec': split_codecs
[0],
3392 'acodec': split_codecs
[1],
3397 def urlhandle_detect_ext(url_handle
):
3398 getheader
= url_handle
.headers
.get
3400 cd
= getheader('Content-Disposition')
3402 m
= re
.match(r
'attachment;\s*filename="(?P<filename>[^"]+)"', cd
)
3404 e
= determine_ext(m
.group('filename'), default_ext
=None)
3408 return mimetype2ext(getheader('Content-Type'))
3411 def encode_data_uri(data
, mime_type
):
3412 return 'data:%s;base64,%s' % (mime_type
, base64
.b64encode(data
).decode('ascii'))
3415 def age_restricted(content_limit
, age_limit
):
3416 """ Returns True iff the content should be blocked """
3418 if age_limit
is None: # No limit set
3420 if content_limit
is None:
3421 return False # Content available for everyone
3422 return age_limit
< content_limit
3425 def is_html(first_bytes
):
3426 """ Detect whether a file contains HTML by examining its first bytes. """
3429 (b
'\xef\xbb\xbf', 'utf-8'),
3430 (b
'\x00\x00\xfe\xff', 'utf-32-be'),
3431 (b
'\xff\xfe\x00\x00', 'utf-32-le'),
3432 (b
'\xff\xfe', 'utf-16-le'),
3433 (b
'\xfe\xff', 'utf-16-be'),
3435 for bom
, enc
in BOMS
:
3436 if first_bytes
.startswith(bom
):
3437 s
= first_bytes
[len(bom
):].decode(enc
, 'replace')
3440 s
= first_bytes
.decode('utf-8', 'replace')
3442 return re
.match(r
'^\s*<', s
)
3445 def determine_protocol(info_dict
):
3446 protocol
= info_dict
.get('protocol')
3447 if protocol
is not None:
3450 url
= sanitize_url(info_dict
['url'])
3451 if url
.startswith('rtmp'):
3453 elif url
.startswith('mms'):
3455 elif url
.startswith('rtsp'):
3458 ext
= determine_ext(url
)
3464 return compat_urllib_parse_urlparse(url
).scheme
3467 def render_table(header_row
, data
, delim
=False, extra_gap
=0, hide_empty
=False):
3468 """ Render a list of rows, each as a list of values.
3469 Text after a \t will be right aligned """
3471 return len(remove_terminal_sequences(string
).replace('\t', ''))
3473 def get_max_lens(table
):
3474 return [max(width(str(v
)) for v
in col
) for col
in zip(*table
)]
3476 def filter_using_list(row
, filterArray
):
3477 return [col
for take
, col
in itertools
.zip_longest(filterArray
, row
, fillvalue
=True) if take
]
3479 max_lens
= get_max_lens(data
) if hide_empty
else []
3480 header_row
= filter_using_list(header_row
, max_lens
)
3481 data
= [filter_using_list(row
, max_lens
) for row
in data
]
3483 table
= [header_row
] + data
3484 max_lens
= get_max_lens(table
)
3487 table
= [header_row
, [delim
* (ml
+ extra_gap
) for ml
in max_lens
]] + data
3488 table
[1][-1] = table
[1][-1][:-extra_gap
] # Remove extra_gap from end of delimiter
3490 for pos
, text
in enumerate(map(str, row
)):
3492 row
[pos
] = text
.replace('\t', ' ' * (max_lens
[pos
] - width(text
))) + ' ' * extra_gap
3494 row
[pos
] = text
+ ' ' * (max_lens
[pos
] - width(text
) + extra_gap
)
3495 ret
= '\n'.join(''.join(row
).rstrip() for row
in table
)
3499 def _match_one(filter_part
, dct
, incomplete
):
3500 # TODO: Generalize code with YoutubeDL._build_format_filter
3501 STRING_OPERATORS
= {
3502 '*=': operator
.contains
,
3503 '^=': lambda attr
, value
: attr
.startswith(value
),
3504 '$=': lambda attr
, value
: attr
.endswith(value
),
3505 '~=': lambda attr
, value
: re
.search(value
, attr
),
3507 COMPARISON_OPERATORS
= {
3509 '<=': operator
.le
, # "<=" must be defined above "<"
3516 operator_rex
= re
.compile(r
'''(?x)\s*
3518 \s*(?P<negation>!\s*)?(?P<op>%s)(?P<none_inclusive>\s*\?)?\s*
3520 (?P<quote>["\'])(?P
<quotedstrval
>.+?
)(?P
=quote
)|
3524 ''' % '|'.join(map(re.escape, COMPARISON_OPERATORS.keys())))
3525 m = operator_rex.search(filter_part)
3528 unnegated_op = COMPARISON_OPERATORS[m['op']]
3530 op = lambda attr, value: not unnegated_op(attr, value)
3533 comparison_value = m['quotedstrval'] or m['strval'] or m['intval']
3535 comparison_value = comparison_value.replace(r'\%s' % m['quote'], m['quote'])
3536 actual_value = dct.get(m['key'])
3537 numeric_comparison = None
3538 if isinstance(actual_value, compat_numeric_types):
3539 # If the original field is a string and matching comparisonvalue is
3540 # a number we should respect the origin of the original field
3541 # and process comparison value as a string (see
3542 # https://github.com/ytdl-org/youtube-dl/issues/11082)
3544 numeric_comparison = int(comparison_value)
3546 numeric_comparison = parse_filesize(comparison_value)
3547 if numeric_comparison is None:
3548 numeric_comparison = parse_filesize(f'{comparison_value}B')
3549 if numeric_comparison is None:
3550 numeric_comparison = parse_duration(comparison_value)
3551 if numeric_comparison is not None and m['op'] in STRING_OPERATORS:
3552 raise ValueError('Operator %s only supports string values!' % m['op'])
3553 if actual_value is None:
3554 return incomplete or m['none_inclusive']
3555 return op(actual_value, comparison_value if numeric_comparison is None else numeric_comparison)
3558 '': lambda v: (v is True) if isinstance(v, bool) else (v is not None),
3559 '!': lambda v: (v is False) if isinstance(v, bool) else (v is None),
3561 operator_rex = re.compile(r'''(?x
)\s
*
3562 (?P
<op
>%s)\s
*(?P
<key
>[a
-z_
]+)
3564 ''' % '|'.join(map(re.escape, UNARY_OPERATORS.keys())))
3565 m = operator_rex.search(filter_part)
3567 op = UNARY_OPERATORS[m.group('op')]
3568 actual_value = dct.get(m.group('key'))
3569 if incomplete and actual_value is None:
3571 return op(actual_value)
3573 raise ValueError('Invalid filter part %r' % filter_part)
3576 def match_str(filter_str, dct, incomplete=False):
3577 """ Filter a dictionary with a simple string syntax. Returns True (=passes filter) or false
3578 When incomplete, all conditions passes on missing fields
3581 _match_one(filter_part.replace(r'\&', '&'), dct, incomplete)
3582 for filter_part in re.split(r'(?<!\\)&', filter_str))
3585 def match_filter_func(filter_str):
3586 def _match_func(info_dict, *args, **kwargs):
3587 if match_str(filter_str, info_dict, *args, **kwargs):
3590 video_title = info_dict.get('title', info_dict.get('id', 'video'))
3591 return '%s does not pass filter %s, skipping ..' % (video_title, filter_str)
3595 def parse_dfxp_time_expr(time_expr):
3599 mobj = re.match(r'^(?P<time_offset>\d+(?:\.\d+)?)s?$', time_expr)
3601 return float(mobj.group('time_offset'))
3603 mobj = re.match(r'^(\d+):(\d\d):(\d\d(?:(?:\.|:)\d+)?)$', time_expr)
3605 return 3600 * int(mobj.group(1)) + 60 * int(mobj.group(2)) + float(mobj.group(3).replace(':', '.'))
3608 def srt_subtitles_timecode(seconds):
3609 return '%02d:%02d:%02d,%03d' % timetuple_from_msec(seconds * 1000)
3612 def ass_subtitles_timecode(seconds):
3613 time = timetuple_from_msec(seconds * 1000)
3614 return '%01d:%02d:%02d.%02d' % (*time[:-1], time.milliseconds / 10)
3617 def dfxp2srt(dfxp_data):
3619 @param dfxp_data A
bytes-like
object containing DFXP data
3620 @returns A
unicode object containing converted SRT data
3622 LEGACY_NAMESPACES = (
3623 (b'http://www.w3.org/ns/ttml', [
3624 b'http://www.w3.org/2004/11/ttaf1',
3625 b'http://www.w3.org/2006/04/ttaf1',
3626 b'http://www.w3.org/2006/10/ttaf1',
3628 (b'http://www.w3.org/ns/ttml#styling', [
3629 b'http://www.w3.org/ns/ttml#style',
3633 SUPPORTED_STYLING = [
3642 _x = functools.partial(xpath_with_ns, ns_map={
3643 'xml': 'http://www.w3.org/XML/1998/namespace',
3644 'ttml': 'http://www.w3.org/ns/ttml',
3645 'tts': 'http://www.w3.org/ns/ttml#styling',
3651 class TTMLPElementParser(object):
3653 _unclosed_elements = []
3654 _applied_styles = []
3656 def start(self, tag, attrib):
3657 if tag in (_x('ttml:br'), 'br'):
3660 unclosed_elements = []
3662 element_style_id = attrib.get('style')
3664 style.update(default_style)
3665 if element_style_id:
3666 style.update(styles.get(element_style_id, {}))
3667 for prop in SUPPORTED_STYLING:
3668 prop_val = attrib.get(_x('tts:' + prop))
3670 style[prop] = prop_val
3673 for k, v in sorted(style.items()):
3674 if self._applied_styles and self._applied_styles[-1].get(k) == v:
3677 font += ' color="%s"' % v
3678 elif k == 'fontSize':
3679 font += ' size="%s"' % v
3680 elif k == 'fontFamily':
3681 font += ' face="%s"' % v
3682 elif k == 'fontWeight' and v == 'bold':
3684 unclosed_elements.append('b')
3685 elif k == 'fontStyle' and v == 'italic':
3687 unclosed_elements.append('i')
3688 elif k == 'textDecoration' and v == 'underline':
3690 unclosed_elements.append('u')
3692 self._out += '<font' + font + '>'
3693 unclosed_elements.append('font')
3695 if self._applied_styles:
3696 applied_style.update(self._applied_styles[-1])
3697 applied_style.update(style)
3698 self._applied_styles.append(applied_style)
3699 self._unclosed_elements.append(unclosed_elements)
3702 if tag not in (_x('ttml:br'), 'br'):
3703 unclosed_elements = self._unclosed_elements.pop()
3704 for element in reversed(unclosed_elements):
3705 self._out += '</%s>' % element
3706 if unclosed_elements and self._applied_styles:
3707 self._applied_styles.pop()
3709 def data(self, data):
3713 return self._out.strip()
3715 def parse_node(node):
3716 target = TTMLPElementParser()
3717 parser = xml.etree.ElementTree.XMLParser(target=target)
3718 parser.feed(xml.etree.ElementTree.tostring(node))
3719 return parser.close()
3721 for k, v in LEGACY_NAMESPACES:
3723 dfxp_data = dfxp_data.replace(ns, k)
3725 dfxp = compat_etree_fromstring(dfxp_data)
3727 paras = dfxp.findall(_x('.//ttml:p')) or dfxp.findall('.//p')
3730 raise ValueError('Invalid dfxp/TTML subtitle')
3734 for style in dfxp.findall(_x('.//ttml:style')):
3735 style_id = style.get('id') or style.get(_x('xml:id'))
3738 parent_style_id = style.get('style')
3740 if parent_style_id not in styles:
3743 styles[style_id] = styles[parent_style_id].copy()
3744 for prop in SUPPORTED_STYLING:
3745 prop_val = style.get(_x('tts:' + prop))
3747 styles.setdefault(style_id, {})[prop] = prop_val
3753 for p in ('body', 'div'):
3754 ele = xpath_element(dfxp, [_x('.//ttml:' + p), './/' + p])
3757 style = styles.get(ele.get('style'))
3760 default_style.update(style)
3762 for para, index in zip(paras, itertools.count(1)):
3763 begin_time = parse_dfxp_time_expr(para.attrib.get('begin'))
3764 end_time = parse_dfxp_time_expr(para.attrib.get('end'))
3765 dur = parse_dfxp_time_expr(para.attrib.get('dur'))
3766 if begin_time is None:
3771 end_time = begin_time + dur
3772 out.append('%d\n%s --> %s\n%s\n\n' % (
3774 srt_subtitles_timecode(begin_time),
3775 srt_subtitles_timecode(end_time),
3781 def cli_option(params, command_option, param):
3782 param = params.get(param)
3784 param = compat_str(param)
3785 return [command_option, param] if param is not None else []
3788 def cli_bool_option(params, command_option, param, true_value='true', false_value='false', separator=None):
3789 param = params.get(param)
3792 assert isinstance(param, bool)
3794 return [command_option + separator + (true_value if param else false_value)]
3795 return [command_option, true_value if param else false_value]
3798 def cli_valueless_option(params, command_option, param, expected_value=True):
3799 param = params.get(param)
3800 return [command_option] if param == expected_value else []
3803 def cli_configuration_args(argdict, keys, default=[], use_compat=True):
3804 if isinstance(argdict, (list, tuple)): # for backward compatibility
3811 assert isinstance(argdict, dict)
3813 assert isinstance(keys, (list, tuple))
3814 for key_list in keys:
3815 arg_list = list(filter(
3816 lambda x: x is not None,
3817 [argdict.get(key.lower()) for key in variadic(key_list)]))
3819 return [arg for args in arg_list for arg in args]
3823 def _configuration_args(main_key, argdict, exe, keys=None, default=[], use_compat=True):
3824 main_key, exe = main_key.lower(), exe.lower()
3825 root_key = exe if main_key == exe else f'{main_key}+{exe}'
3826 keys = [f'{root_key}{k}' for k in (keys or [''])]
3827 if root_key in keys:
3829 keys.append((main_key, exe))
3830 keys.append('default')
3833 return cli_configuration_args(argdict, keys, default, use_compat)
3836 class ISO639Utils(object):
3837 # See http://www.loc.gov/standards/iso639-2/ISO-639-2_utf-8.txt
3896 'iw': 'heb', # Replaced by he in 1989 revision
3906 'in': 'ind', # Replaced by id in 1989 revision
4021 'ji': 'yid', # Replaced by yi in 1989 revision
4029 def short2long(cls, code):
4030 """Convert language code from ISO 639-1 to ISO 639-2/T"""
4031 return cls._lang_map.get(code[:2])
4034 def long2short(cls, code):
4035 """Convert language code from ISO 639-2/T to ISO 639-1"""
4036 for short_name, long_name in cls._lang_map.items():
4037 if long_name == code:
4041 class ISO3166Utils(object):
4042 # From http://data.okfn.org/data/core/country-list
4044 'AF': 'Afghanistan',
4045 'AX': 'Åland Islands',
4048 'AS': 'American Samoa',
4053 'AG': 'Antigua and Barbuda',
4070 'BO': 'Bolivia, Plurinational State of',
4071 'BQ': 'Bonaire, Sint Eustatius and Saba',
4072 'BA': 'Bosnia and Herzegovina',
4074 'BV': 'Bouvet Island',
4076 'IO': 'British Indian Ocean Territory',
4077 'BN': 'Brunei Darussalam',
4079 'BF': 'Burkina Faso',
4085 'KY': 'Cayman Islands',
4086 'CF': 'Central African Republic',
4090 'CX': 'Christmas Island',
4091 'CC': 'Cocos (Keeling) Islands',
4095 'CD': 'Congo, the Democratic Republic of the',
4096 'CK': 'Cook Islands',
4098 'CI': 'Côte d\'Ivoire',
4103 'CZ': 'Czech Republic',
4107 'DO': 'Dominican Republic',
4110 'SV': 'El Salvador',
4111 'GQ': 'Equatorial Guinea',
4115 'FK': 'Falkland Islands (Malvinas)',
4116 'FO': 'Faroe Islands',
4120 'GF': 'French Guiana',
4121 'PF': 'French Polynesia',
4122 'TF': 'French Southern Territories',
4137 'GW': 'Guinea-Bissau',
4140 'HM': 'Heard Island and McDonald Islands',
4141 'VA': 'Holy See (Vatican City State)',
4148 'IR': 'Iran, Islamic Republic of',
4151 'IM': 'Isle of Man',
4161 'KP': 'Korea, Democratic People\'s Republic of',
4162 'KR': 'Korea, Republic of',
4165 'LA': 'Lao People\'s Democratic Republic',
4171 'LI': 'Liechtenstein',
4175 'MK': 'Macedonia, the Former Yugoslav Republic of',
4182 'MH': 'Marshall Islands',
4188 'FM': 'Micronesia, Federated States of',
4189 'MD': 'Moldova, Republic of',
4200 'NL': 'Netherlands',
4201 'NC': 'New Caledonia',
4202 'NZ': 'New Zealand',
4207 'NF': 'Norfolk Island',
4208 'MP': 'Northern Mariana Islands',
4213 'PS': 'Palestine, State of',
4215 'PG': 'Papua New Guinea',
4218 'PH': 'Philippines',
4222 'PR': 'Puerto Rico',
4226 'RU': 'Russian Federation',
4228 'BL': 'Saint Barthélemy',
4229 'SH': 'Saint Helena, Ascension and Tristan da Cunha',
4230 'KN': 'Saint Kitts and Nevis',
4231 'LC': 'Saint Lucia',
4232 'MF': 'Saint Martin (French part)',
4233 'PM': 'Saint Pierre and Miquelon',
4234 'VC': 'Saint Vincent and the Grenadines',
4237 'ST': 'Sao Tome and Principe',
4238 'SA': 'Saudi Arabia',
4242 'SL': 'Sierra Leone',
4244 'SX': 'Sint Maarten (Dutch part)',
4247 'SB': 'Solomon Islands',
4249 'ZA': 'South Africa',
4250 'GS': 'South Georgia and the South Sandwich Islands',
4251 'SS': 'South Sudan',
4256 'SJ': 'Svalbard and Jan Mayen',
4259 'CH': 'Switzerland',
4260 'SY': 'Syrian Arab Republic',
4261 'TW': 'Taiwan, Province of China',
4263 'TZ': 'Tanzania, United Republic of',
4265 'TL': 'Timor-Leste',
4269 'TT': 'Trinidad and Tobago',
4272 'TM': 'Turkmenistan',
4273 'TC': 'Turks and Caicos Islands',
4277 'AE': 'United Arab Emirates',
4278 'GB': 'United Kingdom',
4279 'US': 'United States',
4280 'UM': 'United States Minor Outlying Islands',
4284 'VE': 'Venezuela, Bolivarian Republic of',
4286 'VG': 'Virgin Islands, British',
4287 'VI': 'Virgin Islands, U.S.',
4288 'WF': 'Wallis and Futuna',
4289 'EH': 'Western Sahara',
4296 def short2full(cls, code):
4297 """Convert an ISO 3166-2 country code to the corresponding full name"""
4298 return cls._country_map.get(code.upper())
4301 class GeoUtils(object):
4302 # Major IPv4 address blocks per country
4304 'AD': '46.172.224.0/19',
4305 'AE': '94.200.0.0/13',
4306 'AF': '149.54.0.0/17',
4307 'AG': '209.59.64.0/18',
4308 'AI': '204.14.248.0/21',
4309 'AL': '46.99.0.0/16',
4310 'AM': '46.70.0.0/15',
4311 'AO': '105.168.0.0/13',
4312 'AP': '182.50.184.0/21',
4313 'AQ': '23.154.160.0/24',
4314 'AR': '181.0.0.0/12',
4315 'AS': '202.70.112.0/20',
4316 'AT': '77.116.0.0/14',
4317 'AU': '1.128.0.0/11',
4318 'AW': '181.41.0.0/18',
4319 'AX': '185.217.4.0/22',
4320 'AZ': '5.197.0.0/16',
4321 'BA': '31.176.128.0/17',
4322 'BB': '65.48.128.0/17',
4323 'BD': '114.130.0.0/16',
4325 'BF': '102.178.0.0/15',
4326 'BG': '95.42.0.0/15',
4327 'BH': '37.131.0.0/17',
4328 'BI': '154.117.192.0/18',
4329 'BJ': '137.255.0.0/16',
4330 'BL': '185.212.72.0/23',
4331 'BM': '196.12.64.0/18',
4332 'BN': '156.31.0.0/16',
4333 'BO': '161.56.0.0/16',
4334 'BQ': '161.0.80.0/20',
4335 'BR': '191.128.0.0/12',
4336 'BS': '24.51.64.0/18',
4337 'BT': '119.2.96.0/19',
4338 'BW': '168.167.0.0/16',
4339 'BY': '178.120.0.0/13',
4340 'BZ': '179.42.192.0/18',
4341 'CA': '99.224.0.0/11',
4342 'CD': '41.243.0.0/16',
4343 'CF': '197.242.176.0/21',
4344 'CG': '160.113.0.0/16',
4345 'CH': '85.0.0.0/13',
4346 'CI': '102.136.0.0/14',
4347 'CK': '202.65.32.0/19',
4348 'CL': '152.172.0.0/14',
4349 'CM': '102.244.0.0/14',
4350 'CN': '36.128.0.0/10',
4351 'CO': '181.240.0.0/12',
4352 'CR': '201.192.0.0/12',
4353 'CU': '152.206.0.0/15',
4354 'CV': '165.90.96.0/19',
4355 'CW': '190.88.128.0/17',
4356 'CY': '31.153.0.0/16',
4357 'CZ': '88.100.0.0/14',
4359 'DJ': '197.241.0.0/17',
4360 'DK': '87.48.0.0/12',
4361 'DM': '192.243.48.0/20',
4362 'DO': '152.166.0.0/15',
4363 'DZ': '41.96.0.0/12',
4364 'EC': '186.68.0.0/15',
4365 'EE': '90.190.0.0/15',
4366 'EG': '156.160.0.0/11',
4367 'ER': '196.200.96.0/20',
4368 'ES': '88.0.0.0/11',
4369 'ET': '196.188.0.0/14',
4370 'EU': '2.16.0.0/13',
4371 'FI': '91.152.0.0/13',
4372 'FJ': '144.120.0.0/16',
4373 'FK': '80.73.208.0/21',
4374 'FM': '119.252.112.0/20',
4375 'FO': '88.85.32.0/19',
4377 'GA': '41.158.0.0/15',
4379 'GD': '74.122.88.0/21',
4380 'GE': '31.146.0.0/16',
4381 'GF': '161.22.64.0/18',
4382 'GG': '62.68.160.0/19',
4383 'GH': '154.160.0.0/12',
4384 'GI': '95.164.0.0/16',
4385 'GL': '88.83.0.0/19',
4386 'GM': '160.182.0.0/15',
4387 'GN': '197.149.192.0/18',
4388 'GP': '104.250.0.0/19',
4389 'GQ': '105.235.224.0/20',
4390 'GR': '94.64.0.0/13',
4391 'GT': '168.234.0.0/16',
4392 'GU': '168.123.0.0/16',
4393 'GW': '197.214.80.0/20',
4394 'GY': '181.41.64.0/18',
4395 'HK': '113.252.0.0/14',
4396 'HN': '181.210.0.0/16',
4397 'HR': '93.136.0.0/13',
4398 'HT': '148.102.128.0/17',
4399 'HU': '84.0.0.0/14',
4400 'ID': '39.192.0.0/10',
4401 'IE': '87.32.0.0/12',
4402 'IL': '79.176.0.0/13',
4403 'IM': '5.62.80.0/20',
4404 'IN': '117.192.0.0/10',
4405 'IO': '203.83.48.0/21',
4406 'IQ': '37.236.0.0/14',
4407 'IR': '2.176.0.0/12',
4408 'IS': '82.221.0.0/16',
4409 'IT': '79.0.0.0/10',
4410 'JE': '87.244.64.0/18',
4411 'JM': '72.27.0.0/17',
4412 'JO': '176.29.0.0/16',
4413 'JP': '133.0.0.0/8',
4414 'KE': '105.48.0.0/12',
4415 'KG': '158.181.128.0/17',
4416 'KH': '36.37.128.0/17',
4417 'KI': '103.25.140.0/22',
4418 'KM': '197.255.224.0/20',
4419 'KN': '198.167.192.0/19',
4420 'KP': '175.45.176.0/22',
4421 'KR': '175.192.0.0/10',
4422 'KW': '37.36.0.0/14',
4423 'KY': '64.96.0.0/15',
4424 'KZ': '2.72.0.0/13',
4425 'LA': '115.84.64.0/18',
4426 'LB': '178.135.0.0/16',
4427 'LC': '24.92.144.0/20',
4428 'LI': '82.117.0.0/19',
4429 'LK': '112.134.0.0/15',
4430 'LR': '102.183.0.0/16',
4431 'LS': '129.232.0.0/17',
4432 'LT': '78.56.0.0/13',
4433 'LU': '188.42.0.0/16',
4434 'LV': '46.109.0.0/16',
4435 'LY': '41.252.0.0/14',
4436 'MA': '105.128.0.0/11',
4437 'MC': '88.209.64.0/18',
4438 'MD': '37.246.0.0/16',
4439 'ME': '178.175.0.0/17',
4440 'MF': '74.112.232.0/21',
4441 'MG': '154.126.0.0/17',
4442 'MH': '117.103.88.0/21',
4443 'MK': '77.28.0.0/15',
4444 'ML': '154.118.128.0/18',
4445 'MM': '37.111.0.0/17',
4446 'MN': '49.0.128.0/17',
4447 'MO': '60.246.0.0/16',
4448 'MP': '202.88.64.0/20',
4449 'MQ': '109.203.224.0/19',
4450 'MR': '41.188.64.0/18',
4451 'MS': '208.90.112.0/22',
4452 'MT': '46.11.0.0/16',
4453 'MU': '105.16.0.0/12',
4454 'MV': '27.114.128.0/18',
4455 'MW': '102.70.0.0/15',
4456 'MX': '187.192.0.0/11',
4457 'MY': '175.136.0.0/13',
4458 'MZ': '197.218.0.0/15',
4459 'NA': '41.182.0.0/16',
4460 'NC': '101.101.0.0/18',
4461 'NE': '197.214.0.0/18',
4462 'NF': '203.17.240.0/22',
4463 'NG': '105.112.0.0/12',
4464 'NI': '186.76.0.0/15',
4465 'NL': '145.96.0.0/11',
4466 'NO': '84.208.0.0/13',
4467 'NP': '36.252.0.0/15',
4468 'NR': '203.98.224.0/19',
4469 'NU': '49.156.48.0/22',
4470 'NZ': '49.224.0.0/14',
4471 'OM': '5.36.0.0/15',
4472 'PA': '186.72.0.0/15',
4473 'PE': '186.160.0.0/14',
4474 'PF': '123.50.64.0/18',
4475 'PG': '124.240.192.0/19',
4476 'PH': '49.144.0.0/13',
4477 'PK': '39.32.0.0/11',
4478 'PL': '83.0.0.0/11',
4479 'PM': '70.36.0.0/20',
4480 'PR': '66.50.0.0/16',
4481 'PS': '188.161.0.0/16',
4482 'PT': '85.240.0.0/13',
4483 'PW': '202.124.224.0/20',
4484 'PY': '181.120.0.0/14',
4485 'QA': '37.210.0.0/15',
4486 'RE': '102.35.0.0/16',
4487 'RO': '79.112.0.0/13',
4488 'RS': '93.86.0.0/15',
4489 'RU': '5.136.0.0/13',
4490 'RW': '41.186.0.0/16',
4491 'SA': '188.48.0.0/13',
4492 'SB': '202.1.160.0/19',
4493 'SC': '154.192.0.0/11',
4494 'SD': '102.120.0.0/13',
4495 'SE': '78.64.0.0/12',
4496 'SG': '8.128.0.0/10',
4497 'SI': '188.196.0.0/14',
4498 'SK': '78.98.0.0/15',
4499 'SL': '102.143.0.0/17',
4500 'SM': '89.186.32.0/19',
4501 'SN': '41.82.0.0/15',
4502 'SO': '154.115.192.0/18',
4503 'SR': '186.179.128.0/17',
4504 'SS': '105.235.208.0/21',
4505 'ST': '197.159.160.0/19',
4506 'SV': '168.243.0.0/16',
4507 'SX': '190.102.0.0/20',
4509 'SZ': '41.84.224.0/19',
4510 'TC': '65.255.48.0/20',
4511 'TD': '154.68.128.0/19',
4512 'TG': '196.168.0.0/14',
4513 'TH': '171.96.0.0/13',
4514 'TJ': '85.9.128.0/18',
4515 'TK': '27.96.24.0/21',
4516 'TL': '180.189.160.0/20',
4517 'TM': '95.85.96.0/19',
4518 'TN': '197.0.0.0/11',
4519 'TO': '175.176.144.0/21',
4520 'TR': '78.160.0.0/11',
4521 'TT': '186.44.0.0/15',
4522 'TV': '202.2.96.0/19',
4523 'TW': '120.96.0.0/11',
4524 'TZ': '156.156.0.0/14',
4525 'UA': '37.52.0.0/14',
4526 'UG': '102.80.0.0/13',
4528 'UY': '167.56.0.0/13',
4529 'UZ': '84.54.64.0/18',
4530 'VA': '212.77.0.0/19',
4531 'VC': '207.191.240.0/21',
4532 'VE': '186.88.0.0/13',
4533 'VG': '66.81.192.0/20',
4534 'VI': '146.226.0.0/16',
4535 'VN': '14.160.0.0/11',
4536 'VU': '202.80.32.0/20',
4537 'WF': '117.20.32.0/21',
4538 'WS': '202.4.32.0/19',
4539 'YE': '134.35.0.0/16',
4540 'YT': '41.242.116.0/22',
4541 'ZA': '41.0.0.0/11',
4542 'ZM': '102.144.0.0/13',
4543 'ZW': '102.177.192.0/18',
4547 def random_ipv4(cls, code_or_block):
4548 if len(code_or_block) == 2:
4549 block = cls._country_ip_map.get(code_or_block.upper())
4553 block = code_or_block
4554 addr, preflen = block.split('/')
4555 addr_min = compat_struct_unpack('!L', socket.inet_aton(addr))[0]
4556 addr_max = addr_min | (0xffffffff >> int(preflen))
4557 return compat_str(socket.inet_ntoa(
4558 compat_struct_pack('!L', random.randint(addr_min, addr_max))))
4561 class PerRequestProxyHandler(compat_urllib_request.ProxyHandler):
4562 def __init__(self, proxies=None):
4563 # Set default handlers
4564 for type in ('http', 'https'):
4565 setattr(self, '%s_open' % type,
4566 lambda r, proxy='__noproxy__', type=type, meth=self.proxy_open:
4567 meth(r, proxy, type))
4568 compat_urllib_request.ProxyHandler.__init__(self, proxies)
4570 def proxy_open(self, req, proxy, type):
4571 req_proxy = req.headers.get('Ytdl-request-proxy')
4572 if req_proxy is not None:
4574 del req.headers['Ytdl-request-proxy']
4576 if proxy == '__noproxy__':
4577 return None # No Proxy
4578 if compat_urlparse.urlparse(proxy).scheme.lower() in ('socks', 'socks4', 'socks4a', 'socks5'):
4579 req.add_header('Ytdl-socks-proxy', proxy)
4580 # yt-dlp's http/https handlers do wrapping the socket with socks
4582 return compat_urllib_request.ProxyHandler.proxy_open(
4583 self, req, proxy, type)
4586 # Both long_to_bytes and bytes_to_long are adapted from PyCrypto, which is
4587 # released into Public Domain
4588 # https://github.com/dlitz/pycrypto/blob/master/lib/Crypto/Util/number.py#L387
4590 def long_to_bytes(n, blocksize=0):
4591 """long_to_bytes(n:long, blocksize:int) : string
4592 Convert a long integer to a byte string.
4594 If optional blocksize is given and greater than zero, pad the front of the
4595 byte string with binary zeros so that the length is a multiple of
4598 # after much testing, this algorithm was deemed to be the fastest
4602 s = compat_struct_pack('>I', n & 0xffffffff) + s
4604 # strip off leading zeros
4605 for i in range(len(s)):
4606 if s[i] != b'\000'[0]:
4609 # only happens when n == 0
4613 # add back some pad bytes. this could be done more efficiently w.r.t. the
4614 # de-padding being done above, but sigh...
4615 if blocksize > 0 and len(s) % blocksize:
4616 s = (blocksize - len(s) % blocksize) * b'\000' + s
4620 def bytes_to_long(s):
4621 """bytes_to_long(string) : long
4622 Convert a byte string to a long integer.
4624 This is (essentially) the inverse of long_to_bytes().
4629 extra = (4 - length % 4)
4630 s = b'\000' * extra + s
4631 length = length + extra
4632 for i in range(0, length, 4):
4633 acc = (acc << 32) + compat_struct_unpack('>I', s[i:i + 4])[0]
4637 def ohdave_rsa_encrypt(data, exponent, modulus):
4639 Implement OHDave
's RSA algorithm. See http://www.ohdave.com/rsa/
4642 data: data to encrypt, bytes-like object
4643 exponent, modulus: parameter e and N of RSA algorithm, both integer
4644 Output: hex string of encrypted data
4646 Limitation: supports one block encryption only
4649 payload = int(binascii.hexlify(data[::-1]), 16)
4650 encrypted = pow(payload, exponent, modulus)
4651 return '%x' % encrypted
4654 def pkcs1pad(data, length):
4656 Padding input data with PKCS#1 scheme
4658 @param {int[]} data input data
4659 @param {int} length target length
4660 @returns {int[]} padded data
4662 if len(data) > length - 11:
4663 raise ValueError('Input data too
long for PKCS
#1 padding')
4665 pseudo_random
= [random
.randint(0, 254) for _
in range(length
- len(data
) - 3)]
4666 return [0, 2] + pseudo_random
+ [0] + data
4669 def encode_base_n(num
, n
, table
=None):
4670 FULL_TABLE
= '0123456789abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ'
4672 table
= FULL_TABLE
[:n
]
4675 raise ValueError('base %d exceeds table length %d' % (n
, len(table
)))
4682 ret
= table
[num
% n
] + ret
4687 def decode_packed_codes(code
):
4688 mobj
= re
.search(PACKED_CODES_RE
, code
)
4689 obfuscated_code
, base
, count
, symbols
= mobj
.groups()
4692 symbols
= symbols
.split('|')
4697 base_n_count
= encode_base_n(count
, base
)
4698 symbol_table
[base_n_count
] = symbols
[count
] or base_n_count
4701 r
'\b(\w+)\b', lambda mobj
: symbol_table
[mobj
.group(0)],
4705 def caesar(s
, alphabet
, shift
):
4710 alphabet
[(alphabet
.index(c
) + shift
) % l
] if c
in alphabet
else c
4715 return caesar(s
, r
'''!"#$%&'()*+,-./0123456789:;<=>?@ABCDEFGHIJKLMNOPQRSTUVWXYZ[\]^_`abcdefghijklmnopqrstuvwxyz{|}~''', 47)
4718 def parse_m3u8_attributes(attrib
):
4720 for (key
, val
) in re
.findall(r
'(?P<key>[A-Z0-9-]+)=(?P<val>"[^"]+"|[^",]+)(?:,|$)', attrib
):
4721 if val
.startswith('"'):
4727 def urshift(val
, n
):
4728 return val
>> n
if val
>= 0 else (val
+ 0x100000000) >> n
4731 # Based on png2str() written by @gdkchan and improved by @yokrysty
4732 # Originally posted at https://github.com/ytdl-org/youtube-dl/issues/9706
4733 def decode_png(png_data
):
4734 # Reference: https://www.w3.org/TR/PNG/
4735 header
= png_data
[8:]
4737 if png_data
[:8] != b
'\x89PNG\x0d\x0a\x1a\x0a' or header
[4:8] != b
'IHDR':
4738 raise IOError('Not a valid PNG file.')
4740 int_map
= {1: '>B', 2: '>H', 4: '>I'}
4741 unpack_integer
= lambda x
: compat_struct_unpack(int_map
[len(x
)], x
)[0]
4746 length
= unpack_integer(header
[:4])
4749 chunk_type
= header
[:4]
4752 chunk_data
= header
[:length
]
4753 header
= header
[length
:]
4755 header
= header
[4:] # Skip CRC
4763 ihdr
= chunks
[0]['data']
4765 width
= unpack_integer(ihdr
[:4])
4766 height
= unpack_integer(ihdr
[4:8])
4770 for chunk
in chunks
:
4771 if chunk
['type'] == b
'IDAT':
4772 idat
+= chunk
['data']
4775 raise IOError('Unable to read PNG data.')
4777 decompressed_data
= bytearray(zlib
.decompress(idat
))
4782 def _get_pixel(idx
):
4787 for y
in range(height
):
4788 basePos
= y
* (1 + stride
)
4789 filter_type
= decompressed_data
[basePos
]
4793 pixels
.append(current_row
)
4795 for x
in range(stride
):
4796 color
= decompressed_data
[1 + basePos
+ x
]
4797 basex
= y
* stride
+ x
4802 left
= _get_pixel(basex
- 3)
4804 up
= _get_pixel(basex
- stride
)
4806 if filter_type
== 1: # Sub
4807 color
= (color
+ left
) & 0xff
4808 elif filter_type
== 2: # Up
4809 color
= (color
+ up
) & 0xff
4810 elif filter_type
== 3: # Average
4811 color
= (color
+ ((left
+ up
) >> 1)) & 0xff
4812 elif filter_type
== 4: # Paeth
4818 c
= _get_pixel(basex
- stride
- 3)
4826 if pa
<= pb
and pa
<= pc
:
4827 color
= (color
+ a
) & 0xff
4829 color
= (color
+ b
) & 0xff
4831 color
= (color
+ c
) & 0xff
4833 current_row
.append(color
)
4835 return width
, height
, pixels
4838 def write_xattr(path
, key
, value
):
4839 # This mess below finds the best xattr tool for the job
4841 # try the pyxattr module...
4844 if hasattr(xattr
, 'set'): # pyxattr
4845 # Unicode arguments are not supported in python-pyxattr until
4847 # See https://github.com/ytdl-org/youtube-dl/issues/5498
4848 pyxattr_required_version
= '0.5.0'
4849 if version_tuple(xattr
.__version
__) < version_tuple(pyxattr_required_version
):
4850 # TODO: fallback to CLI tools
4851 raise XAttrUnavailableError(
4852 'python-pyxattr is detected but is too old. '
4853 'yt-dlp requires %s or above while your version is %s. '
4854 'Falling back to other xattr implementations' % (
4855 pyxattr_required_version
, xattr
.__version
__))
4857 setxattr
= xattr
.set
4859 setxattr
= xattr
.setxattr
4862 setxattr(path
, key
, value
)
4863 except EnvironmentError as e
:
4864 raise XAttrMetadataError(e
.errno
, e
.strerror
)
4867 if compat_os_name
== 'nt':
4868 # Write xattrs to NTFS Alternate Data Streams:
4869 # http://en.wikipedia.org/wiki/NTFS#Alternate_data_streams_.28ADS.29
4870 assert ':' not in key
4871 assert os
.path
.exists(path
)
4873 ads_fn
= path
+ ':' + key
4875 with open(ads_fn
, 'wb') as f
:
4877 except EnvironmentError as e
:
4878 raise XAttrMetadataError(e
.errno
, e
.strerror
)
4880 user_has_setfattr
= check_executable('setfattr', ['--version'])
4881 user_has_xattr
= check_executable('xattr', ['-h'])
4883 if user_has_setfattr
or user_has_xattr
:
4885 value
= value
.decode('utf-8')
4886 if user_has_setfattr
:
4887 executable
= 'setfattr'
4888 opts
= ['-n', key
, '-v', value
]
4889 elif user_has_xattr
:
4890 executable
= 'xattr'
4891 opts
= ['-w', key
, value
]
4893 cmd
= ([encodeFilename(executable
, True)]
4894 + [encodeArgument(o
) for o
in opts
]
4895 + [encodeFilename(path
, True)])
4899 cmd
, stdout
=subprocess
.PIPE
, stderr
=subprocess
.PIPE
, stdin
=subprocess
.PIPE
)
4900 except EnvironmentError as e
:
4901 raise XAttrMetadataError(e
.errno
, e
.strerror
)
4902 stdout
, stderr
= p
.communicate_or_kill()
4903 stderr
= stderr
.decode('utf-8', 'replace')
4904 if p
.returncode
!= 0:
4905 raise XAttrMetadataError(p
.returncode
, stderr
)
4908 # On Unix, and can't find pyxattr, setfattr, or xattr.
4909 if sys
.platform
.startswith('linux'):
4910 raise XAttrUnavailableError(
4911 "Couldn't find a tool to set the xattrs. "
4912 "Install either the python 'pyxattr' or 'xattr' "
4913 "modules, or the GNU 'attr' package "
4914 "(which contains the 'setfattr' tool).")
4916 raise XAttrUnavailableError(
4917 "Couldn't find a tool to set the xattrs. "
4918 "Install either the python 'xattr' module, "
4919 "or the 'xattr' binary.")
4922 def random_birthday(year_field
, month_field
, day_field
):
4923 start_date
= datetime
.date(1950, 1, 1)
4924 end_date
= datetime
.date(1995, 12, 31)
4925 offset
= random
.randint(0, (end_date
- start_date
).days
)
4926 random_date
= start_date
+ datetime
.timedelta(offset
)
4928 year_field
: str(random_date
.year
),
4929 month_field
: str(random_date
.month
),
4930 day_field
: str(random_date
.day
),
4934 # Templates for internet shortcut files, which are plain text files.
4935 DOT_URL_LINK_TEMPLATE
= '''
4940 DOT_WEBLOC_LINK_TEMPLATE
= '''
4941 <?xml version="1.0" encoding="UTF-8"?>
4942 <!DOCTYPE plist PUBLIC "-//Apple//DTD PLIST 1.0//EN" "http://www.apple.com/DTDs/PropertyList-1.0.dtd">
4943 <plist version="1.0">
4946 \t<string>%(url)s</string>
4951 DOT_DESKTOP_LINK_TEMPLATE
= '''
4961 'url': DOT_URL_LINK_TEMPLATE
,
4962 'desktop': DOT_DESKTOP_LINK_TEMPLATE
,
4963 'webloc': DOT_WEBLOC_LINK_TEMPLATE
,
4967 def iri_to_uri(iri
):
4969 Converts an IRI (Internationalized Resource Identifier, allowing Unicode characters) to a URI (Uniform Resource Identifier, ASCII-only).
4971 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.
4974 iri_parts
= compat_urllib_parse_urlparse(iri
)
4976 if '[' in iri_parts
.netloc
:
4977 raise ValueError('IPv6 URIs are not, yet, supported.')
4978 # Querying `.netloc`, when there's only one bracket, also raises a ValueError.
4980 # 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.
4983 if iri_parts
.username
:
4984 net_location
+= compat_urllib_parse_quote(iri_parts
.username
, safe
=r
"!$%&'()*+,~")
4985 if iri_parts
.password
is not None:
4986 net_location
+= ':' + compat_urllib_parse_quote(iri_parts
.password
, safe
=r
"!$%&'()*+,~")
4989 net_location
+= iri_parts
.hostname
.encode('idna').decode('utf-8') # Punycode for Unicode hostnames.
4990 # The 'idna' encoding produces ASCII text.
4991 if iri_parts
.port
is not None and iri_parts
.port
!= 80:
4992 net_location
+= ':' + str(iri_parts
.port
)
4994 return compat_urllib_parse_urlunparse(
4998 compat_urllib_parse_quote_plus(iri_parts
.path
, safe
=r
"!$%&'()*+,/:;=@|~"),
5000 # Unsure about the `safe` argument, since this is a legacy way of handling parameters.
5001 compat_urllib_parse_quote_plus(iri_parts
.params
, safe
=r
"!$%&'()*+,/:;=@|~"),
5003 # Not totally sure about the `safe` argument, since the source does not explicitly mention the query URI component.
5004 compat_urllib_parse_quote_plus(iri_parts
.query
, safe
=r
"!$%&'()*+,/:;=?@{|}~"),
5006 compat_urllib_parse_quote_plus(iri_parts
.fragment
, safe
=r
"!#$%&'()*+,/:;=?@{|}~")))
5008 # Source for `safe` arguments: https://url.spec.whatwg.org/#percent-encoded-bytes.
5011 def to_high_limit_path(path
):
5012 if sys
.platform
in ['win32', 'cygwin']:
5013 # Work around MAX_PATH limitation on Windows. The maximum allowed length for the individual path segments may still be quite limited.
5014 return r
'\\?\ '.rstrip() + os
.path
.abspath(path
)
5019 def format_field(obj
, field
=None, template
='%s', ignore
=(None, ''), default
='', func
=None):
5020 val
= traverse_obj(obj
, *variadic(field
))
5023 return template
% (func(val
) if func
else val
)
5026 def clean_podcast_url(url
):
5027 return re
.sub(r
'''(?x)
5031 media\.blubrry\.com| # https://create.blubrry.com/resources/podcast-media-download-statistics/getting-started/
5034 (?:dts|www)\.podtrac\.com/(?:pts/)?redirect\.[0-9a-z]{3,4}| # http://analytics.podtrac.com/how-to-measure
5037 cn\.co| # https://podcorn.com/analytics-prefix/
5038 st\.fm # https://podsights.com/docs/
5043 _HEX_TABLE
= '0123456789abcdef'
5046 def random_uuidv4():
5047 return re
.sub(r
'[xy]', lambda x
: _HEX_TABLE
[random
.randint(0, 15)], 'xxxxxxxx-xxxx-4xxx-yxxx-xxxxxxxxxxxx')
5050 def make_dir(path
, to_screen
=None):
5052 dn
= os
.path
.dirname(path
)
5053 if dn
and not os
.path
.exists(dn
):
5056 except (OSError, IOError) as err
:
5057 if callable(to_screen
) is not None:
5058 to_screen('unable to create directory ' + error_to_compat_str(err
))
5062 def get_executable_path():
5063 from zipimport
import zipimporter
5064 if hasattr(sys
, 'frozen'): # Running from PyInstaller
5065 path
= os
.path
.dirname(sys
.executable
)
5066 elif isinstance(globals().get('__loader__'), zipimporter
): # Running from ZIP
5067 path
= os
.path
.join(os
.path
.dirname(__file__
), '../..')
5069 path
= os
.path
.join(os
.path
.dirname(__file__
), '..')
5070 return os
.path
.abspath(path
)
5073 def load_plugins(name
, suffix
, namespace
):
5076 plugins_spec
= importlib
.util
.spec_from_file_location(
5077 name
, os
.path
.join(get_executable_path(), 'ytdlp_plugins', name
, '__init__.py'))
5078 plugins
= importlib
.util
.module_from_spec(plugins_spec
)
5079 sys
.modules
[plugins_spec
.name
] = plugins
5080 plugins_spec
.loader
.exec_module(plugins
)
5081 for name
in dir(plugins
):
5082 if name
in namespace
:
5084 if not name
.endswith(suffix
):
5086 klass
= getattr(plugins
, name
)
5087 classes
[name
] = namespace
[name
] = klass
5088 except FileNotFoundError
:
5094 obj
, *path_list
, default
=None, expected_type
=None, get_all
=True,
5095 casesense
=True, is_user_input
=False, traverse_string
=False):
5096 ''' Traverse nested list/dict/tuple
5097 @param path_list A list of paths which are checked one by one.
5098 Each path is a list of keys where each key is a string,
5099 a function, a tuple of strings/None or "...".
5100 When a fuction is given, it takes the key as argument and
5101 returns whether the key matches or not. When a tuple is given,
5102 all the keys given in the tuple are traversed, and
5103 "..." traverses all the keys in the object
5104 "None" returns the object without traversal
5105 @param default Default value to return
5106 @param expected_type Only accept final value of this type (Can also be any callable)
5107 @param get_all Return all the values obtained from a path or only the first one
5108 @param casesense Whether to consider dictionary keys as case sensitive
5109 @param is_user_input Whether the keys are generated from user input. If True,
5110 strings are converted to int/slice if necessary
5111 @param traverse_string Whether to traverse inside strings. If True, any
5112 non-compatible object will also be converted into a string
5116 _lower
= lambda k
: (k
.lower() if isinstance(k
, str) else k
)
5117 path_list
= (map(_lower
, variadic(path
)) for path
in path_list
)
5119 def _traverse_obj(obj
, path
, _current_depth
=0):
5121 path
= tuple(variadic(path
))
5122 for i
, key
in enumerate(path
):
5123 if None in (key
, obj
):
5125 if isinstance(key
, (list, tuple)):
5126 obj
= [_traverse_obj(obj
, sub_key
, _current_depth
) for sub_key
in key
]
5129 obj
= (obj
.values() if isinstance(obj
, dict)
5130 else obj
if isinstance(obj
, (list, tuple, LazyList
))
5131 else str(obj
) if traverse_string
else [])
5133 depth
= max(depth
, _current_depth
)
5134 return [_traverse_obj(inner_obj
, path
[i
+ 1:], _current_depth
) for inner_obj
in obj
]
5136 if isinstance(obj
, (list, tuple, LazyList
)):
5137 obj
= enumerate(obj
)
5138 elif isinstance(obj
, dict):
5141 if not traverse_string
:
5145 depth
= max(depth
, _current_depth
)
5146 return [_traverse_obj(v
, path
[i
+ 1:], _current_depth
) for k
, v
in obj
if key(k
)]
5147 elif isinstance(obj
, dict) and not (is_user_input
and key
== ':'):
5148 obj
= (obj
.get(key
) if casesense
or (key
in obj
)
5149 else next((v
for k
, v
in obj
.items() if _lower(k
) == key
), None))
5152 key
= (int_or_none(key
) if ':' not in key
5153 else slice(*map(int_or_none
, key
.split(':'))))
5154 if key
== slice(None):
5155 return _traverse_obj(obj
, (..., *path
[i
+ 1:]), _current_depth
)
5156 if not isinstance(key
, (int, slice)):
5158 if not isinstance(obj
, (list, tuple, LazyList
)):
5159 if not traverse_string
:
5168 if isinstance(expected_type
, type):
5169 type_test
= lambda val
: val
if isinstance(val
, expected_type
) else None
5170 elif expected_type
is not None:
5171 type_test
= expected_type
5173 type_test
= lambda val
: val
5175 for path
in path_list
:
5177 val
= _traverse_obj(obj
, path
)
5180 for _
in range(depth
- 1):
5181 val
= itertools
.chain
.from_iterable(v
for v
in val
if v
is not None)
5182 val
= [v
for v
in map(type_test
, val
) if v
is not None]
5184 return val
if get_all
else val
[0]
5186 val
= type_test(val
)
5192 def traverse_dict(dictn
, keys
, casesense
=True):
5193 write_string('DeprecationWarning: yt_dlp.utils.traverse_dict is deprecated '
5194 'and may be removed in a future version. Use yt_dlp.utils.traverse_obj instead')
5195 return traverse_obj(dictn
, keys
, casesense
=casesense
, is_user_input
=True, traverse_string
=True)
5198 def variadic(x
, allowed_types
=(str, bytes, dict)):
5199 return x
if isinstance(x
, collections
.abc
.Iterable
) and not isinstance(x
, allowed_types
) else (x
,)
5202 def decode_base(value
, digits
):
5203 # This will convert given base-x string to scalar (long or int)
5204 table
= {char: index for index, char in enumerate(digits)}
5209 result
+= table
[chr]
5213 def time_seconds(**kwargs
):
5214 t
= datetime
.datetime
.now(datetime
.timezone(datetime
.timedelta(**kwargs
)))
5215 return t
.timestamp()
5218 # create a JSON Web Signature (jws) with HS256 algorithm
5219 # the resulting format is in JWS Compact Serialization
5220 # implemented following JWT https://www.rfc-editor.org/rfc/rfc7519.html
5221 # implemented following JWS https://www.rfc-editor.org/rfc/rfc7515.html
5222 def jwt_encode_hs256(payload_data
, key
, headers
={}):
5228 header_data
.update(headers
)
5229 header_b64
= base64
.b64encode(json
.dumps(header_data
).encode('utf-8'))
5230 payload_b64
= base64
.b64encode(json
.dumps(payload_data
).encode('utf-8'))
5231 h
= hmac
.new(key
.encode('utf-8'), header_b64
+ b
'.' + payload_b64
, hashlib
.sha256
)
5232 signature_b64
= base64
.b64encode(h
.digest())
5233 token
= header_b64
+ b
'.' + payload_b64
+ b
'.' + signature_b64
5237 # can be extended in future to verify the signature and parse header and return the algorithm used if it's not HS256
5238 def jwt_decode_hs256(jwt
):
5239 header_b64
, payload_b64
, signature_b64
= jwt
.split('.')
5240 payload_data
= json
.loads(base64
.urlsafe_b64decode(payload_b64
))
5244 def supports_terminal_sequences(stream
):
5245 if compat_os_name
== 'nt':
5246 from .compat
import WINDOWS_VT_MODE
# Must be imported locally
5247 if not WINDOWS_VT_MODE
or get_windows_version() < (10, 0, 10586):
5249 elif not os
.getenv('TERM'):
5252 return stream
.isatty()
5253 except BaseException
:
5257 _terminal_sequences_re
= re
.compile('\033\\[[^m]+m')
5260 def remove_terminal_sequences(string
):
5261 return _terminal_sequences_re
.sub('', string
)
5264 def number_of_digits(number
):
5265 return len('%d' % number
)
5268 def join_nonempty(*values
, delim
='-', from_dict
=None):
5269 if from_dict
is not None:
5270 values
= map(from_dict
.get
, values
)
5271 return delim
.join(map(str, filter(None, values
)))
5274 def parse_http_range(range):
5275 """ Parse value of "Range" or "Content-Range" HTTP header into tuple. """
5277 return None, None, None
5278 crg
= re
.search(r
'bytes[ =](\d+)-(\d+)?(?:/(\d+))?', range)
5280 return None, None, None
5281 return int(crg
.group(1)), int_or_none(crg
.group(2)), int_or_none(crg
.group(3))
5287 __initialized
= False
5289 def __init__(self
, parser
, label
=None):
5290 self
._parser
, self
.label
= parser
, label
5291 self
._loaded
_paths
, self
.configs
= set(), []
5293 def init(self
, args
=None, filename
=None):
5294 assert not self
.__initialized
5297 location
= os
.path
.realpath(filename
)
5298 directory
= os
.path
.dirname(location
)
5299 if location
in self
._loaded
_paths
:
5301 self
._loaded
_paths
.add(location
)
5303 self
.__initialized
= True
5304 self
.own_args
, self
.filename
= args
, filename
5305 for location
in self
._parser
.parse_args(args
)[0].config_locations
or []:
5306 location
= os
.path
.join(directory
, expand_path(location
))
5307 if os
.path
.isdir(location
):
5308 location
= os
.path
.join(location
, 'yt-dlp.conf')
5309 if not os
.path
.exists(location
):
5310 self
._parser
.error(f
'config location {location} does not exist')
5311 self
.append_config(self
.read_file(location
), location
)
5315 label
= join_nonempty(
5316 self
.label
, 'config', f
'"{self.filename}"' if self
.filename
else '',
5318 return join_nonempty(
5319 self
.own_args
is not None and f
'{label[0].upper()}{label[1:]}: {self.hide_login_info(self.own_args)}',
5320 *(f
'\n{c}'.replace('\n', '\n| ')[1:] for c
in self
.configs
),
5324 def read_file(filename
, default
=[]):
5326 optionf
= open(filename
)
5328 return default
# silently skip if file is not present
5330 # FIXME: https://github.com/ytdl-org/youtube-dl/commit/dfe5fa49aed02cf36ba9f743b11b0903554b5e56
5331 contents
= optionf
.read()
5332 if sys
.version_info
< (3,):
5333 contents
= contents
.decode(preferredencoding())
5334 res
= compat_shlex_split(contents
, comments
=True)
5340 def hide_login_info(opts
):
5341 PRIVATE_OPTS
= set(['-p', '--password', '-u', '--username', '--video-password', '--ap-password', '--ap-username'])
5342 eqre
= re
.compile('^(?P<key>' + ('|'.join(re
.escape(po
) for po
in PRIVATE_OPTS
)) + ')=.+$')
5347 return m
.group('key') + '=PRIVATE'
5351 opts
= list(map(_scrub_eq
, opts
))
5352 for idx
, opt
in enumerate(opts
):
5353 if opt
in PRIVATE_OPTS
and idx
+ 1 < len(opts
):
5354 opts
[idx
+ 1] = 'PRIVATE'
5357 def append_config(self
, *args
, label
=None):
5358 config
= type(self
)(self
._parser
, label
)
5359 config
._loaded
_paths
= self
._loaded
_paths
5360 if config
.init(*args
):
5361 self
.configs
.append(config
)
5365 for config
in reversed(self
.configs
):
5366 yield from config
.all_args
5367 yield from self
.own_args
or []
5369 def parse_args(self
):
5370 return self
._parser
.parse_args(list(self
.all_args
))
5373 class WebSocketsWrapper():
5374 """Wraps websockets module to use in non-async scopes"""
5376 def __init__(self
, url
, headers
=None):
5377 self
.loop
= asyncio
.events
.new_event_loop()
5378 self
.conn
= compat_websockets
.connect(
5379 url
, extra_headers
=headers
, ping_interval
=None,
5380 close_timeout
=float('inf'), loop
=self
.loop
, ping_timeout
=float('inf'))
5381 atexit
.register(self
.__exit
__, None, None, None)
5383 def __enter__(self
):
5384 self
.pool
= self
.run_with_loop(self
.conn
.__aenter
__(), self
.loop
)
5387 def send(self
, *args
):
5388 self
.run_with_loop(self
.pool
.send(*args
), self
.loop
)
5390 def recv(self
, *args
):
5391 return self
.run_with_loop(self
.pool
.recv(*args
), self
.loop
)
5393 def __exit__(self
, type, value
, traceback
):
5395 return self
.run_with_loop(self
.conn
.__aexit
__(type, value
, traceback
), self
.loop
)
5398 self
._cancel
_all
_tasks
(self
.loop
)
5400 # taken from https://github.com/python/cpython/blob/3.9/Lib/asyncio/runners.py with modifications
5401 # for contributors: If there's any new library using asyncio needs to be run in non-async, move these function out of this class
5403 def run_with_loop(main
, loop
):
5404 if not asyncio
.coroutines
.iscoroutine(main
):
5405 raise ValueError(f
'a coroutine was expected, got {main!r}')
5408 return loop
.run_until_complete(main
)
5410 loop
.run_until_complete(loop
.shutdown_asyncgens())
5411 if hasattr(loop
, 'shutdown_default_executor'):
5412 loop
.run_until_complete(loop
.shutdown_default_executor())
5415 def _cancel_all_tasks(loop
):
5416 to_cancel
= asyncio
.tasks
.all_tasks(loop
)
5421 for task
in to_cancel
:
5424 loop
.run_until_complete(
5425 asyncio
.tasks
.gather(*to_cancel
, loop
=loop
, return_exceptions
=True))
5427 for task
in to_cancel
:
5428 if task
.cancelled():
5430 if task
.exception() is not None:
5431 loop
.call_exception_handler({
5432 'message': 'unhandled exception during asyncio.run() shutdown',
5433 'exception': task
.exception(),
5438 has_websockets
= bool(compat_websockets
)