4 from __future__
import unicode_literals
39 import xml
.etree
.ElementTree
44 compat_HTMLParseError
,
50 compat_ctypes_WINFUNCTYPE
,
51 compat_etree_fromstring
,
54 compat_html_entities_html5
,
68 compat_urllib_parse_urlencode
,
69 compat_urllib_parse_urlparse
,
70 compat_urllib_parse_urlunparse
,
71 compat_urllib_parse_quote
,
72 compat_urllib_parse_quote_plus
,
73 compat_urllib_parse_unquote_plus
,
74 compat_urllib_request
,
85 def register_socks_protocols():
86 # "Register" SOCKS protocols
87 # In Python < 2.6.5, urlsplit() suffers from bug https://bugs.python.org/issue7904
88 # URLs with protocols not in urlparse.uses_netloc are not handled correctly
89 for scheme
in ('socks', 'socks4', 'socks4a', 'socks5'):
90 if scheme
not in compat_urlparse
.uses_netloc
:
91 compat_urlparse
.uses_netloc
.append(scheme
)
94 # This is not clearly defined otherwise
95 compiled_regex_type
= type(re
.compile(''))
98 def random_user_agent():
99 _USER_AGENT_TPL
= 'Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/%s Safari/537.36'
140 return _USER_AGENT_TPL
% random
.choice(_CHROME_VERSIONS
)
144 'User-Agent': random_user_agent(),
145 'Accept': 'text/html,application/xhtml+xml,application/xml;q=0.9,*/*;q=0.8',
146 'Accept-Encoding': 'gzip, deflate',
147 'Accept-Language': 'en-us,en;q=0.5',
152 'Safari': 'Mozilla/5.0 (X11; Linux x86_64; rv:10.0) AppleWebKit/533.20.25 (KHTML, like Gecko) Version/5.0.4 Safari/533.20.27',
156 NO_DEFAULT
= object()
158 ENGLISH_MONTH_NAMES
= [
159 'January', 'February', 'March', 'April', 'May', 'June',
160 'July', 'August', 'September', 'October', 'November', 'December']
163 'en': ENGLISH_MONTH_NAMES
,
165 'janvier', 'février', 'mars', 'avril', 'mai', 'juin',
166 'juillet', 'août', 'septembre', 'octobre', 'novembre', 'décembre'],
170 'mp4', 'm4a', 'm4p', 'm4b', 'm4r', 'm4v', 'aac',
171 'flv', 'f4v', 'f4a', 'f4b',
172 'webm', 'ogg', 'ogv', 'oga', 'ogx', 'spx', 'opus',
173 'mkv', 'mka', 'mk3d',
182 'f4f', 'f4m', 'm3u8', 'smil')
184 # needed for sanitizing filenames in restricted mode
185 ACCENT_CHARS
= dict(zip('ÂÃÄÀÁÅÆÇÈÉÊËÌÍÎÏÐÑÒÓÔÕÖŐØŒÙÚÛÜŰÝÞßàáâãäåæçèéêëìíîïðñòóôõöőøœùúûüűýþÿ',
186 itertools
.chain('AAAAAA', ['AE'], 'CEEEEIIIIDNOOOOOOO', ['OE'], 'UUUUUY', ['TH', 'ss'],
187 'aaaaaa', ['ae'], 'ceeeeiiiionooooooo', ['oe'], 'uuuuuy', ['th'], 'y')))
217 '%Y-%m-%d %H:%M:%S.%f',
218 '%Y-%m-%d %H:%M:%S:%f',
221 '%Y-%m-%dT%H:%M:%SZ',
222 '%Y-%m-%dT%H:%M:%S.%fZ',
223 '%Y-%m-%dT%H:%M:%S.%f0Z',
225 '%Y-%m-%dT%H:%M:%S.%f',
228 '%b %d %Y at %H:%M:%S',
230 '%B %d %Y at %H:%M:%S',
234 DATE_FORMATS_DAY_FIRST
= list(DATE_FORMATS
)
235 DATE_FORMATS_DAY_FIRST
.extend([
244 DATE_FORMATS_MONTH_FIRST
= list(DATE_FORMATS
)
245 DATE_FORMATS_MONTH_FIRST
.extend([
253 PACKED_CODES_RE
= r
"}\('(.+)',(\d+),(\d+),'([^']+)'\.split\('\|'\)"
254 JSON_LD_RE
= r
'(?is)<script[^>]+type=(["\']?
)application
/ld\
+json\
1[^
>]*>(?P
<json_ld
>.+?
)</script
>'
257 def preferredencoding():
258 """Get preferred encoding.
260 Returns the best encoding scheme for the system, based on
261 locale.getpreferredencoding() and some further tweaks.
264 pref = locale.getpreferredencoding()
272 def write_json_file(obj, fn):
273 """ Encode obj as JSON and write it to fn, atomically if possible """
275 fn = encodeFilename(fn)
276 if sys.version_info < (3, 0) and sys.platform != 'win32
':
277 encoding = get_filesystem_encoding()
278 # os.path.basename returns a bytes object, but NamedTemporaryFile
279 # will fail if the filename contains non ascii characters unless we
280 # use a unicode object
281 path_basename = lambda f: os.path.basename(fn).decode(encoding)
282 # the same for os.path.dirname
283 path_dirname = lambda f: os.path.dirname(fn).decode(encoding)
285 path_basename = os.path.basename
286 path_dirname = os.path.dirname
290 'prefix
': path_basename(fn) + '.',
291 'dir': path_dirname(fn),
295 # In Python 2.x, json.dump expects a bytestream.
296 # In Python 3.x, it writes to a character stream
297 if sys.version_info < (3, 0):
305 tf = tempfile.NamedTemporaryFile(**compat_kwargs(args))
309 json.dump(obj, tf, ensure_ascii=False)
310 if sys.platform == 'win32
':
311 # Need to remove existing file on Windows, else os.rename raises
312 # WindowsError or FileExistsError.
320 os.chmod(tf.name, 0o666 & ~mask)
323 os.rename(tf.name, fn)
332 if sys.version_info >= (2, 7):
333 def find_xpath_attr(node, xpath, key, val=None):
334 """ Find the xpath xpath[@key=val] """
335 assert re.match(r'^
[a
-zA
-Z_
-]+$
', key)
336 expr = xpath + ('[@%s]' % key if val is None else "[@%s='%s']" % (key, val))
337 return node.find(expr)
339 def find_xpath_attr(node, xpath, key, val=None):
340 for f in node.findall(compat_xpath(xpath)):
341 if key not in f.attrib:
343 if val is None or f.attrib.get(key) == val:
347 # On python2.6 the xml.etree.ElementTree.Element methods don't support
348 # the namespace parameter
351 def xpath_with_ns(path
, ns_map
):
352 components
= [c
.split(':') for c
in path
.split('/')]
356 replaced
.append(c
[0])
359 replaced
.append('{%s}%s' % (ns_map
[ns
], tag
))
360 return '/'.join(replaced
)
363 def xpath_element(node
, xpath
, name
=None, fatal
=False, default
=NO_DEFAULT
):
364 def _find_xpath(xpath
):
365 return node
.find(compat_xpath(xpath
))
367 if isinstance(xpath
, (str, compat_str
)):
368 n
= _find_xpath(xpath
)
376 if default
is not NO_DEFAULT
:
379 name
= xpath
if name
is None else name
380 raise ExtractorError('Could not find XML element %s' % name
)
386 def xpath_text(node
, xpath
, name
=None, fatal
=False, default
=NO_DEFAULT
):
387 n
= xpath_element(node
, xpath
, name
, fatal
=fatal
, default
=default
)
388 if n
is None or n
== default
:
391 if default
is not NO_DEFAULT
:
394 name
= xpath
if name
is None else name
395 raise ExtractorError('Could not find XML element\'s text %s' % name
)
401 def xpath_attr(node
, xpath
, key
, name
=None, fatal
=False, default
=NO_DEFAULT
):
402 n
= find_xpath_attr(node
, xpath
, key
)
404 if default
is not NO_DEFAULT
:
407 name
= '%s[@%s]' % (xpath
, key
) if name
is None else name
408 raise ExtractorError('Could not find XML attribute %s' % name
)
414 def get_element_by_id(id, html
):
415 """Return the content of the tag with the specified ID in the passed HTML document"""
416 return get_element_by_attribute('id', id, html
)
419 def get_element_html_by_id(id, html
):
420 """Return the html of the tag with the specified ID in the passed HTML document"""
421 return get_element_html_by_attribute('id', id, html
)
424 def get_element_by_class(class_name
, html
):
425 """Return the content of the first tag with the specified class in the passed HTML document"""
426 retval
= get_elements_by_class(class_name
, html
)
427 return retval
[0] if retval
else None
430 def get_element_html_by_class(class_name
, html
):
431 """Return the html of the first tag with the specified class in the passed HTML document"""
432 retval
= get_elements_html_by_class(class_name
, html
)
433 return retval
[0] if retval
else None
436 def get_element_by_attribute(attribute
, value
, html
, escape_value
=True):
437 retval
= get_elements_by_attribute(attribute
, value
, html
, escape_value
)
438 return retval
[0] if retval
else None
441 def get_element_html_by_attribute(attribute
, value
, html
, escape_value
=True):
442 retval
= get_elements_html_by_attribute(attribute
, value
, html
, escape_value
)
443 return retval
[0] if retval
else None
446 def get_elements_by_class(class_name
, html
):
447 """Return the content of all tags with the specified class in the passed HTML document as a list"""
448 return get_elements_by_attribute(
449 'class', r
'[^\'"]*\b%s\b[^\'"]*' % re.escape(class_name),
450 html, escape_value=False)
453 def get_elements_html_by_class(class_name, html):
454 """Return the html of all tags with the specified class in the passed HTML document as a list"""
455 return get_elements_html_by_attribute(
456 'class', r'[^
\'"]*\b%s\b[^\'"]*' % re.escape(class_name),
457 html, escape_value=False)
460 def get_elements_by_attribute(*args, **kwargs):
461 """Return the content of the tag with the specified attribute in the passed HTML document"""
462 return [content for content, _ in get_elements_text_and_html_by_attribute(*args, **kwargs)]
465 def get_elements_html_by_attribute(*args, **kwargs):
466 """Return the html of the tag with the specified attribute in the passed HTML document"""
467 return [whole for _, whole in get_elements_text_and_html_by_attribute(*args, **kwargs)]
470 def get_elements_text_and_html_by_attribute(attribute, value, html, escape_value=True):
472 Return the text (content) and the html (whole) of the tag with the specified
473 attribute in the passed HTML document
476 value = re.escape(value) if escape_value else value
479 for m in re.finditer(r'''(?xs)
480 <(?P<tag>[a-zA-Z0-9:._-]+)
481 (?:\s+[a-zA-Z0-9_:.-]+(?:=\S*?|\s*=\s*(?:"[^"]*"|'[^
']*')|
))*?
482 \s
+%(attribute)s(?
:=%(value)s|\s
*=\s
*(?P
<_q
>['"]?)%(value)s(?P=_q))
483 (?:\s+[a-zA-Z0-9_:.-]+(?:=\S*?|\s*=\s*(?:"[^"]*"|'[^
']*')|
))*?
485 ''' % {'attribute': re.escape(attribute), 'value': value}, html):
486 content, whole = get_element_text_and_html_by_tag(m.group('tag'), html[m.start():])
489 unescapeHTML(re.sub(r'(?s)^(?P<q>["\'])(?P<content>.*)(?P=q)$', r'\g<content>', content)),
496 class HTMLBreakOnClosingTagParser(compat_HTMLParser):
498 HTML parser which raises HTMLBreakOnClosingTagException upon reaching the
499 closing tag for the first opening tag it has encountered, and can be used
503 class HTMLBreakOnClosingTagException(Exception):
507 self.tagstack = collections.deque()
508 compat_HTMLParser.__init__(self)
513 def __exit__(self, *_):
517 # handle_endtag does not return upon raising HTMLBreakOnClosingTagException,
518 # so data remains buffered; we no longer have any interest in it, thus
519 # override this method to discard it
522 def handle_starttag(self, tag, _):
523 self.tagstack.append(tag)
525 def handle_endtag(self, tag):
526 if not self.tagstack:
527 raise compat_HTMLParseError('no tags in the stack')
529 inner_tag = self.tagstack.pop()
533 raise compat_HTMLParseError(f'matching opening tag for closing {tag} tag not found')
534 if not self.tagstack:
535 raise self.HTMLBreakOnClosingTagException()
538 def get_element_text_and_html_by_tag(tag, html):
540 For the first element with the specified tag in the passed HTML document
541 return its' content (text) and the whole element (html)
543 def find_or_raise(haystack, needle, exc):
545 return haystack.index(needle)
548 closing_tag = f'</{tag}>'
549 whole_start = find_or_raise(
550 html, f'<{tag}', compat_HTMLParseError(f'opening {tag} tag not found'))
551 content_start = find_or_raise(
552 html[whole_start:], '>', compat_HTMLParseError(f'malformed opening {tag} tag'))
553 content_start += whole_start + 1
554 with HTMLBreakOnClosingTagParser() as parser:
555 parser.feed(html[whole_start:content_start])
556 if not parser.tagstack or parser.tagstack[0] != tag:
557 raise compat_HTMLParseError(f'parser did not match opening {tag} tag')
558 offset = content_start
559 while offset < len(html):
560 next_closing_tag_start = find_or_raise(
561 html[offset:], closing_tag,
562 compat_HTMLParseError(f'closing {tag} tag not found'))
563 next_closing_tag_end = next_closing_tag_start + len(closing_tag)
565 parser.feed(html[offset:offset + next_closing_tag_end])
566 offset += next_closing_tag_end
567 except HTMLBreakOnClosingTagParser.HTMLBreakOnClosingTagException:
568 return html[content_start:offset + next_closing_tag_start], \
569 html[whole_start:offset + next_closing_tag_end]
570 raise compat_HTMLParseError('unexpected end of html')
573 class HTMLAttributeParser(compat_HTMLParser):
574 """Trivial HTML parser to gather the attributes for a single element"""
578 compat_HTMLParser.__init__(self)
580 def handle_starttag(self, tag, attrs):
581 self.attrs = dict(attrs)
584 class HTMLListAttrsParser(compat_HTMLParser):
585 """HTML parser to gather the attributes for the elements of a list"""
588 compat_HTMLParser.__init__(self)
592 def handle_starttag(self, tag, attrs):
593 if tag == 'li' and self._level == 0:
594 self.items.append(dict(attrs))
597 def handle_endtag(self, tag):
601 def extract_attributes(html_element):
602 """Given a string for an HTML element such as
604 a="foo" B="bar" c="&98;az" d=boz
605 empty= noval entity="&"
608 Decode and return a dictionary of attributes.
610 'a': 'foo', 'b': 'bar', c: 'baz', d: 'boz',
611 'empty': '', 'noval': None, 'entity': '&',
612 'sq': '"', 'dq': '\''
614 NB HTMLParser is stricter in Python 2.6 & 3.2 than in later versions,
615 but the cases in the unit test will work for all of 2.6, 2.7, 3.2-3.5.
617 parser = HTMLAttributeParser()
619 parser.feed(html_element)
621 # Older Python may throw HTMLParseError in case of malformed HTML
622 except compat_HTMLParseError:
627 def parse_list(webpage):
628 """Given a string for an series of HTML <li> elements,
629 return a dictionary of their attributes"""
630 parser = HTMLListAttrsParser()
636 def clean_html(html):
637 """Clean an HTML snippet into a readable string"""
639 if html is None: # Convenience for sanitizing descriptions etc.
643 html = html.replace('\n', ' ')
644 html = re.sub(r'(?u)\s*<\s*br\s*/?\s*>\s*', '\n', html)
645 html = re.sub(r'(?u)<\s*/\s*p\s*>\s*<\s*p[^>]*>', '\n', html)
647 html = re.sub('<.*?>', '', html)
648 # Replace html entities
649 html = unescapeHTML(html)
653 def sanitize_open(filename, open_mode):
654 """Try to open the given filename, and slightly tweak it if this fails.
656 Attempts to open the given filename. If this fails, it tries to change
657 the filename slightly, step by step, until it's either able to open it
658 or it fails and raises a final exception, like the standard open()
661 It returns the tuple (stream, definitive_file_name).
665 if sys.platform == 'win32':
667 msvcrt.setmode(sys.stdout.fileno(), os.O_BINARY)
668 return (sys.stdout.buffer if hasattr(sys.stdout, 'buffer') else sys.stdout, filename)
669 stream = open(encodeFilename(filename), open_mode)
670 return (stream, filename)
671 except (IOError, OSError) as err:
672 if err.errno in (errno.EACCES,):
675 # In case of error, try to remove win32 forbidden chars
676 alt_filename = sanitize_path(filename)
677 if alt_filename == filename:
680 # An exception here should be caught in the caller
681 stream = open(encodeFilename(alt_filename), open_mode)
682 return (stream, alt_filename)
685 def timeconvert(timestr):
686 """Convert RFC 2822 defined time string into system timestamp"""
688 timetuple = email.utils.parsedate_tz(timestr)
689 if timetuple is not None:
690 timestamp = email.utils.mktime_tz(timetuple)
694 def sanitize_filename(s, restricted=False, is_id=False):
695 """Sanitizes a string so it could be used as part of a filename.
696 If restricted is set, use a stricter subset of allowed characters.
697 Set is_id if this is not an arbitrary string, but an ID that should be kept
700 def replace_insane(char):
701 if restricted and char in ACCENT_CHARS:
702 return ACCENT_CHARS[char]
703 elif not restricted and char == '\n':
705 elif char == '?' or ord(char) < 32 or ord(char) == 127:
708 return '' if restricted else '\''
710 return '_-' if restricted else ' -'
711 elif char in '\\/|*<>':
713 if restricted and (char in '!&\'()[]{}$;`^,#' or char.isspace()):
715 if restricted and ord(char) > 127:
722 s = re.sub(r'[0-9]+(?::[0-9]+)+', lambda m: m.group(0).replace(':', '_'), s)
723 result = ''.join(map(replace_insane, s))
725 while '__' in result:
726 result = result.replace('__', '_')
727 result = result.strip('_')
728 # Common case of "Foreign band name - English song title"
729 if restricted and result.startswith('-_'):
731 if result.startswith('-'):
732 result = '_' + result[len('-'):]
733 result = result.lstrip('.')
739 def sanitize_path(s, force=False):
740 """Sanitizes and normalizes path on Windows"""
741 if sys.platform == 'win32':
743 drive_or_unc, _ = os.path.splitdrive(s)
744 if sys.version_info < (2, 7) and not drive_or_unc:
745 drive_or_unc, _ = os.path.splitunc(s)
751 norm_path = os.path.normpath(remove_start(s, drive_or_unc)).split(os.path.sep)
755 path_part if path_part in ['.', '..'] else re.sub(r'(?:[/<>:"\|\\?\*]|[\s.]$)', '#', path_part)
756 for path_part in norm_path]
758 sanitized_path.insert(0, drive_or_unc + os.path.sep)
759 elif force and s[0] == os.path.sep:
760 sanitized_path.insert(0, os.path.sep)
761 return os.path.join(*sanitized_path)
764 def sanitize_url(url):
765 # Prepend protocol-less URLs with `http:` scheme in order to mitigate
766 # the number of unwanted failures due to missing protocol
767 if url.startswith('//'):
768 return 'http:%s' % url
769 # Fix some common typos seen so far
771 # https://github.com/ytdl-org/youtube-dl/issues/15649
772 (r'^httpss://', r'https://'),
773 # https://bx1.be/lives/direct-tv/
774 (r'^rmtp([es]?)://', r'rtmp\1://'),
776 for mistake, fixup in COMMON_TYPOS:
777 if re.match(mistake, url):
778 return re.sub(mistake, fixup, url)
782 def extract_basic_auth(url):
783 parts = compat_urlparse.urlsplit(url)
784 if parts.username is None:
786 url = compat_urlparse.urlunsplit(parts._replace(netloc=(
787 parts.hostname if parts.port is None
788 else '%s:%d' % (parts.hostname, parts.port))))
789 auth_payload = base64.b64encode(
790 ('%s:%s' % (parts.username, parts.password or '')).encode('utf-8'))
791 return url, 'Basic ' + auth_payload.decode('utf-8')
794 def sanitized_Request(url, *args, **kwargs):
795 url, auth_header = extract_basic_auth(escape_url(sanitize_url(url)))
796 if auth_header is not None:
797 headers = args[1] if len(args) >= 2 else kwargs.setdefault('headers', {})
798 headers['Authorization'] = auth_header
799 return compat_urllib_request.Request(url, *args, **kwargs)
803 """Expand shell variables and ~"""
804 return os.path.expandvars(compat_expanduser(s))
807 def orderedSet(iterable):
808 """ Remove all duplicates from the input iterable """
816 def _htmlentity_transform(entity_with_semicolon):
817 """Transforms an HTML entity to a character."""
818 entity = entity_with_semicolon[:-1]
820 # Known non-numeric HTML entity
821 if entity in compat_html_entities.name2codepoint:
822 return compat_chr(compat_html_entities.name2codepoint[entity])
824 # TODO: HTML5 allows entities without a semicolon. For example,
825 # 'Éric' should be decoded as 'Éric'.
826 if entity_with_semicolon in compat_html_entities_html5:
827 return compat_html_entities_html5[entity_with_semicolon]
829 mobj = re.match(r'#(x[0-9a-fA-F]+|[0-9]+)', entity)
831 numstr = mobj.group(1)
832 if numstr.startswith('x'):
834 numstr = '0%s' % numstr
837 # See https://github.com/ytdl-org/youtube-dl/issues/7518
839 return compat_chr(int(numstr, base))
843 # Unknown entity in name, return its literal representation
844 return '&%s;' % entity
850 assert type(s) == compat_str
853 r'&([^&;]+;)', lambda m: _htmlentity_transform(m.group(1)), s)
856 def escapeHTML(text):
859 .replace('&', '&')
860 .replace('<', '<')
861 .replace('>', '>')
862 .replace('"', '"')
863 .replace("'", ''')
867 def process_communicate_or_kill(p, *args, **kwargs):
869 return p.communicate(*args, **kwargs)
870 except BaseException: # Including KeyboardInterrupt
876 class Popen(subprocess.Popen):
877 if sys.platform == 'win32':
878 _startupinfo = subprocess.STARTUPINFO()
879 _startupinfo.dwFlags |= subprocess.STARTF_USESHOWWINDOW
883 def __init__(self, *args, **kwargs):
884 super(Popen, self).__init__(*args, **kwargs, startupinfo=self._startupinfo)
886 def communicate_or_kill(self, *args, **kwargs):
887 return process_communicate_or_kill(self, *args, **kwargs)
890 def get_subprocess_encoding():
891 if sys.platform == 'win32' and sys.getwindowsversion()[0] >= 5:
892 # For subprocess calls, encode with locale encoding
893 # Refer to http://stackoverflow.com/a/9951851/35070
894 encoding = preferredencoding()
896 encoding = sys.getfilesystemencoding()
902 def encodeFilename(s, for_subprocess=False):
904 @param s The name of the file
907 assert type(s) == compat_str
909 # Python 3 has a Unicode API
910 if sys.version_info >= (3, 0):
913 # Pass '' directly to use Unicode APIs on Windows 2000 and up
914 # (Detecting Windows NT 4 is tricky because 'major >= 4' would
915 # match Windows 9x series as well. Besides, NT 4 is obsolete.)
916 if not for_subprocess and sys.platform == 'win32' and sys.getwindowsversion()[0] >= 5:
919 # Jython assumes filenames are Unicode strings though reported as Python 2.x compatible
920 if sys.platform.startswith('java'):
923 return s.encode(get_subprocess_encoding(), 'ignore')
926 def decodeFilename(b, for_subprocess=False):
928 if sys.version_info >= (3, 0):
931 if not isinstance(b, bytes):
934 return b.decode(get_subprocess_encoding(), 'ignore')
937 def encodeArgument(s):
938 if not isinstance(s, compat_str):
939 # Legacy code that uses byte strings
940 # Uncomment the following line after fixing all post processors
941 # assert False, 'Internal error: %r should be of type %r, is %r' % (s, compat_str, type(s))
942 s = s.decode('ascii')
943 return encodeFilename(s, True)
946 def decodeArgument(b):
947 return decodeFilename(b, True)
950 def decodeOption(optval):
953 if isinstance(optval, bytes):
954 optval = optval.decode(preferredencoding())
956 assert isinstance(optval, compat_str)
960 _timetuple = collections.namedtuple('Time', ('hours', 'minutes', 'seconds', 'milliseconds'))
963 def timetuple_from_msec(msec):
964 secs, msec = divmod(msec, 1000)
965 mins, secs = divmod(secs, 60)
966 hrs, mins = divmod(mins, 60)
967 return _timetuple(hrs, mins, secs, msec)
970 def formatSeconds(secs, delim=':', msec=False):
971 time = timetuple_from_msec(secs * 1000)
973 ret = '%d%s%02d%s%02d' % (time.hours, delim, time.minutes, delim, time.seconds)
975 ret = '%d%s%02d' % (time.minutes, delim, time.seconds)
977 ret = '%d' % time.seconds
978 return '%s.%03d' % (ret, time.milliseconds) if msec else ret
981 def _ssl_load_windows_store_certs(ssl_context, storename):
982 # Code adapted from _load_windows_store_certs in https://github.com/python/cpython/blob/main/Lib/ssl.py
984 certs = [cert for cert, encoding, trust in ssl.enum_certificates(storename)
985 if encoding == 'x509_asn' and (
986 trust is True or ssl.Purpose.SERVER_AUTH.oid in trust)]
987 except PermissionError:
991 ssl_context.load_verify_locations(cadata=cert)
996 def make_HTTPS_handler(params, **kwargs):
997 opts_check_certificate = not params.get('nocheckcertificate')
998 context = ssl.SSLContext(ssl.PROTOCOL_TLS_CLIENT)
999 context.check_hostname = opts_check_certificate
1000 context.verify_mode = ssl.CERT_REQUIRED if opts_check_certificate else ssl.CERT_NONE
1001 if opts_check_certificate:
1003 context.load_default_certs()
1004 # Work around the issue in load_default_certs when there are bad certificates. See:
1005 # https://github.com/yt-dlp/yt-dlp/issues/1060,
1006 # https://bugs.python.org/issue35665, https://bugs.python.org/issue45312
1007 except ssl.SSLError:
1008 # enum_certificates is not present in mingw python. See https://github.com/yt-dlp/yt-dlp/issues/1151
1009 if sys.platform == 'win32' and hasattr(ssl, 'enum_certificates'):
1010 # Create a new context to discard any certificates that were already loaded
1011 context = ssl.SSLContext(ssl.PROTOCOL_TLS_CLIENT)
1012 context.check_hostname, context.verify_mode = True, ssl.CERT_REQUIRED
1013 for storename in ('CA', 'ROOT'):
1014 _ssl_load_windows_store_certs(context, storename)
1015 context.set_default_verify_paths()
1016 return YoutubeDLHTTPSHandler(params, context=context, **kwargs)
1019 def bug_reports_message(before=';'):
1020 if ytdl_is_updateable():
1021 update_cmd = 'type yt-dlp -U to update'
1023 update_cmd = 'see https://github.com/yt-dlp/yt-dlp on how to update'
1024 msg = 'please report this issue on https://github.com/yt-dlp/yt-dlp .'
1025 msg += ' Make sure you are using the latest version; %s.' % update_cmd
1026 msg += ' Be sure to call yt-dlp with the --verbose flag and include its complete output.'
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|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.now(), 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'):
1879 Return a datetime object from a string in the format YYYYMMDD or
1880 (now|today|date)[+-][0-9](microsecond|second|minute|hour|day|week|month|year)(s)?
1882 format: string date format used to return datetime object from
1884 return datetime_from_str(date_str, precision='microsecond', format=format).date()
1887 def datetime_add_months(dt, months):
1888 """Increment/Decrement a datetime object by months."""
1889 month = dt.month + months - 1
1890 year = dt.year + month // 12
1891 month = month % 12 + 1
1892 day = min(dt.day, calendar.monthrange(year, month)[1])
1893 return dt.replace(year, month, day)
1896 def datetime_round(dt, precision='day'):
1898 Round a datetime object's time to a specific precision
1900 if precision == 'microsecond':
1909 roundto = lambda x, n: ((x + n / 2) // n) * n
1910 timestamp = calendar.timegm(dt.timetuple())
1911 return datetime.datetime.utcfromtimestamp(roundto(timestamp, unit_seconds[precision]))
1914 def hyphenate_date(date_str):
1916 Convert a date in 'YYYYMMDD' format to 'YYYY-MM-DD' format"""
1917 match = re.match(r'^(\d\d\d\d)(\d\d)(\d\d)$', date_str)
1918 if match is not None:
1919 return '-'.join(match.groups())
1924 class DateRange(object):
1925 """Represents a time interval between two dates"""
1927 def __init__(self, start=None, end=None):
1928 """start and end must be strings in the format accepted by date"""
1929 if start is not None:
1930 self.start = date_from_str(start)
1932 self.start = datetime.datetime.min.date()
1934 self.end = date_from_str(end)
1936 self.end = datetime.datetime.max.date()
1937 if self.start > self.end:
1938 raise ValueError('Date range: "%s" , the start date must be before the end date' % self)
1942 """Returns a range that only contains the given day"""
1943 return cls(day, day)
1945 def __contains__(self, date):
1946 """Check if the date is in the range"""
1947 if not isinstance(date, datetime.date):
1948 date = date_from_str(date)
1949 return self.start <= date <= self.end
1952 return '%s - %s' % (self.start.isoformat(), self.end.isoformat())
1955 def platform_name():
1956 """ Returns the platform name as a compat_str """
1957 res = platform.platform()
1958 if isinstance(res, bytes):
1959 res = res.decode(preferredencoding())
1961 assert isinstance(res, compat_str)
1965 def get_windows_version():
1966 ''' Get Windows version
. None if it
's not running on Windows '''
1967 if compat_os_name == 'nt
':
1968 return version_tuple(platform.win32_ver()[1])
1973 def _windows_write_string(s, out):
1974 """ Returns True if the string was written using special methods,
1975 False if it has yet to be written out."""
1976 # Adapted from http://stackoverflow.com/a/3259271/35070
1978 import ctypes.wintypes
1986 fileno = out.fileno()
1987 except AttributeError:
1988 # If the output stream doesn't have a fileno
, it
's virtual
1990 except io.UnsupportedOperation:
1991 # Some strange Windows pseudo files?
1993 if fileno not in WIN_OUTPUT_IDS:
1996 GetStdHandle = compat_ctypes_WINFUNCTYPE(
1997 ctypes.wintypes.HANDLE, ctypes.wintypes.DWORD)(
1998 ('GetStdHandle
', ctypes.windll.kernel32))
1999 h = GetStdHandle(WIN_OUTPUT_IDS[fileno])
2001 WriteConsoleW = compat_ctypes_WINFUNCTYPE(
2002 ctypes.wintypes.BOOL, ctypes.wintypes.HANDLE, ctypes.wintypes.LPWSTR,
2003 ctypes.wintypes.DWORD, ctypes.POINTER(ctypes.wintypes.DWORD),
2004 ctypes.wintypes.LPVOID)(('WriteConsoleW
', ctypes.windll.kernel32))
2005 written = ctypes.wintypes.DWORD(0)
2007 GetFileType = compat_ctypes_WINFUNCTYPE(ctypes.wintypes.DWORD, ctypes.wintypes.DWORD)(('GetFileType
', ctypes.windll.kernel32))
2008 FILE_TYPE_CHAR = 0x0002
2009 FILE_TYPE_REMOTE = 0x8000
2010 GetConsoleMode = compat_ctypes_WINFUNCTYPE(
2011 ctypes.wintypes.BOOL, ctypes.wintypes.HANDLE,
2012 ctypes.POINTER(ctypes.wintypes.DWORD))(
2013 ('GetConsoleMode
', ctypes.windll.kernel32))
2014 INVALID_HANDLE_VALUE = ctypes.wintypes.DWORD(-1).value
2016 def not_a_console(handle):
2017 if handle == INVALID_HANDLE_VALUE or handle is None:
2019 return ((GetFileType(handle) & ~FILE_TYPE_REMOTE) != FILE_TYPE_CHAR
2020 or GetConsoleMode(handle, ctypes.byref(ctypes.wintypes.DWORD())) == 0)
2022 if not_a_console(h):
2025 def next_nonbmp_pos(s):
2027 return next(i for i, c in enumerate(s) if ord(c) > 0xffff)
2028 except StopIteration:
2032 count = min(next_nonbmp_pos(s), 1024)
2034 ret = WriteConsoleW(
2035 h, s, count if count else 2, ctypes.byref(written), None)
2037 raise OSError('Failed to write string
')
2038 if not count: # We just wrote a non-BMP character
2039 assert written.value == 2
2042 assert written.value > 0
2043 s = s[written.value:]
2047 def write_string(s, out=None, encoding=None):
2050 assert type(s) == compat_str
2052 if sys.platform == 'win32
' and encoding is None and hasattr(out, 'fileno
'):
2053 if _windows_write_string(s, out):
2056 if ('b
' in getattr(out, 'mode
', '')
2057 or sys.version_info[0] < 3): # Python 2 lies about mode of sys.stderr
2058 byt = s.encode(encoding or preferredencoding(), 'ignore
')
2060 elif hasattr(out, 'buffer'):
2061 enc = encoding or getattr(out, 'encoding
', None) or preferredencoding()
2062 byt = s.encode(enc, 'ignore
')
2063 out.buffer.write(byt)
2069 def bytes_to_intlist(bs):
2072 if isinstance(bs[0], int): # Python 3
2075 return [ord(c) for c in bs]
2078 def intlist_to_bytes(xs):
2081 return compat_struct_pack('%dB
' % len(xs), *xs)
2084 # Cross-platform file locking
2085 if sys.platform == 'win32
':
2086 import ctypes.wintypes
2089 class OVERLAPPED(ctypes.Structure):
2091 ('Internal
', ctypes.wintypes.LPVOID),
2092 ('InternalHigh
', ctypes.wintypes.LPVOID),
2093 ('Offset
', ctypes.wintypes.DWORD),
2094 ('OffsetHigh
', ctypes.wintypes.DWORD),
2095 ('hEvent
', ctypes.wintypes.HANDLE),
2098 kernel32 = ctypes.windll.kernel32
2099 LockFileEx = kernel32.LockFileEx
2100 LockFileEx.argtypes = [
2101 ctypes.wintypes.HANDLE, # hFile
2102 ctypes.wintypes.DWORD, # dwFlags
2103 ctypes.wintypes.DWORD, # dwReserved
2104 ctypes.wintypes.DWORD, # nNumberOfBytesToLockLow
2105 ctypes.wintypes.DWORD, # nNumberOfBytesToLockHigh
2106 ctypes.POINTER(OVERLAPPED) # Overlapped
2108 LockFileEx.restype = ctypes.wintypes.BOOL
2109 UnlockFileEx = kernel32.UnlockFileEx
2110 UnlockFileEx.argtypes = [
2111 ctypes.wintypes.HANDLE, # hFile
2112 ctypes.wintypes.DWORD, # dwReserved
2113 ctypes.wintypes.DWORD, # nNumberOfBytesToLockLow
2114 ctypes.wintypes.DWORD, # nNumberOfBytesToLockHigh
2115 ctypes.POINTER(OVERLAPPED) # Overlapped
2117 UnlockFileEx.restype = ctypes.wintypes.BOOL
2118 whole_low = 0xffffffff
2119 whole_high = 0x7fffffff
2121 def _lock_file(f, exclusive):
2122 overlapped = OVERLAPPED()
2123 overlapped.Offset = 0
2124 overlapped.OffsetHigh = 0
2125 overlapped.hEvent = 0
2126 f._lock_file_overlapped_p = ctypes.pointer(overlapped)
2127 handle = msvcrt.get_osfhandle(f.fileno())
2128 if not LockFileEx(handle, 0x2 if exclusive else 0x0, 0,
2129 whole_low, whole_high, f._lock_file_overlapped_p):
2130 raise OSError('Locking
file failed
: %r' % ctypes.FormatError())
2132 def _unlock_file(f):
2133 assert f._lock_file_overlapped_p
2134 handle = msvcrt.get_osfhandle(f.fileno())
2135 if not UnlockFileEx(handle, 0,
2136 whole_low, whole_high, f._lock_file_overlapped_p):
2137 raise OSError('Unlocking
file failed
: %r' % ctypes.FormatError())
2140 # Some platforms, such as Jython, is missing fcntl
2144 def _lock_file(f, exclusive):
2145 fcntl.flock(f, fcntl.LOCK_EX if exclusive else fcntl.LOCK_SH)
2147 def _unlock_file(f):
2148 fcntl.flock(f, fcntl.LOCK_UN)
2150 UNSUPPORTED_MSG = 'file locking
is not supported on this platform
'
2152 def _lock_file(f, exclusive):
2153 raise IOError(UNSUPPORTED_MSG)
2155 def _unlock_file(f):
2156 raise IOError(UNSUPPORTED_MSG)
2159 class locked_file(object):
2160 def __init__(self, filename, mode, encoding=None):
2161 assert mode in ['r
', 'a
', 'w
']
2162 self.f = io.open(filename, mode, encoding=encoding)
2165 def __enter__(self):
2166 exclusive = self.mode != 'r
'
2168 _lock_file(self.f, exclusive)
2174 def __exit__(self, etype, value, traceback):
2176 _unlock_file(self.f)
2183 def write(self, *args):
2184 return self.f.write(*args)
2186 def read(self, *args):
2187 return self.f.read(*args)
2190 def get_filesystem_encoding():
2191 encoding = sys.getfilesystemencoding()
2192 return encoding if encoding is not None else 'utf
-8'
2195 def shell_quote(args):
2197 encoding = get_filesystem_encoding()
2199 if isinstance(a, bytes):
2200 # We may get a filename encoded with 'encodeFilename
'
2201 a = a.decode(encoding)
2202 quoted_args.append(compat_shlex_quote(a))
2203 return ' '.join(quoted_args)
2206 def smuggle_url(url, data):
2207 """ Pass additional data in a URL for internal use. """
2209 url, idata = unsmuggle_url(url, {})
2211 sdata = compat_urllib_parse_urlencode(
2212 {'__youtubedl_smuggle': json.dumps(data)})
2213 return url + '#' + sdata
2216 def unsmuggle_url(smug_url
, default
=None):
2217 if '#__youtubedl_smuggle' not in smug_url
:
2218 return smug_url
, default
2219 url
, _
, sdata
= smug_url
.rpartition('#')
2220 jsond
= compat_parse_qs(sdata
)['__youtubedl_smuggle'][0]
2221 data
= json
.loads(jsond
)
2225 def format_decimal_suffix(num
, fmt
='%d%s', *, factor
=1000):
2226 """ Formats numbers with decimal sufixes like K, M, etc """
2227 num
, factor
= float_or_none(num
), float(factor
)
2230 exponent
= 0 if num
== 0 else int(math
.log(num
, factor
))
2231 suffix
= ['', *'kMGTPEZY'][exponent
]
2233 suffix
= {'k': 'Ki', '': ''}
.get(suffix
, f
'{suffix}i')
2234 converted
= num
/ (factor
** exponent
)
2235 return fmt
% (converted
, suffix
)
2238 def format_bytes(bytes):
2239 return format_decimal_suffix(bytes, '%.2f%sB', factor
=1024) or 'N/A'
2242 def lookup_unit_table(unit_table
, s
):
2243 units_re
= '|'.join(re
.escape(u
) for u
in unit_table
)
2245 r
'(?P<num>[0-9]+(?:[,.][0-9]*)?)\s*(?P<unit>%s)\b' % units_re
, s
)
2248 num_str
= m
.group('num').replace(',', '.')
2249 mult
= unit_table
[m
.group('unit')]
2250 return int(float(num_str
) * mult
)
2253 def parse_filesize(s
):
2257 # The lower-case forms are of course incorrect and unofficial,
2258 # but we support those too
2275 'megabytes': 1000 ** 2,
2276 'mebibytes': 1024 ** 2,
2282 'gigabytes': 1000 ** 3,
2283 'gibibytes': 1024 ** 3,
2289 'terabytes': 1000 ** 4,
2290 'tebibytes': 1024 ** 4,
2296 'petabytes': 1000 ** 5,
2297 'pebibytes': 1024 ** 5,
2303 'exabytes': 1000 ** 6,
2304 'exbibytes': 1024 ** 6,
2310 'zettabytes': 1000 ** 7,
2311 'zebibytes': 1024 ** 7,
2317 'yottabytes': 1000 ** 8,
2318 'yobibytes': 1024 ** 8,
2321 return lookup_unit_table(_UNIT_TABLE
, s
)
2328 s
= re
.sub(r
'^[^\d]+\s', '', s
).strip()
2330 if re
.match(r
'^[\d,.]+$', s
):
2331 return str_to_int(s
)
2344 ret
= lookup_unit_table(_UNIT_TABLE
, s
)
2348 mobj
= re
.match(r
'([\d,.]+)(?:$|\s)', s
)
2350 return str_to_int(mobj
.group(1))
2353 def parse_resolution(s
):
2357 mobj
= re
.search(r
'(?<![a-zA-Z0-9])(?P<w>\d+)\s*[xX×,]\s*(?P<h>\d+)(?![a-zA-Z0-9])', s
)
2360 'width': int(mobj
.group('w')),
2361 'height': int(mobj
.group('h')),
2364 mobj
= re
.search(r
'(?<![a-zA-Z0-9])(\d+)[pPiI](?![a-zA-Z0-9])', s
)
2366 return {'height': int(mobj.group(1))}
2368 mobj
= re
.search(r
'\b([48])[kK]\b', s
)
2370 return {'height': int(mobj.group(1)) * 540}
2375 def parse_bitrate(s
):
2376 if not isinstance(s
, compat_str
):
2378 mobj
= re
.search(r
'\b(\d+)\s*kbps', s
)
2380 return int(mobj
.group(1))
2383 def month_by_name(name
, lang
='en'):
2384 """ Return the number of a month by (locale-independently) English name """
2386 month_names
= MONTH_NAMES
.get(lang
, MONTH_NAMES
['en'])
2389 return month_names
.index(name
) + 1
2394 def month_by_abbreviation(abbrev
):
2395 """ Return the number of a month by (locale-independently) English
2399 return [s
[:3] for s
in ENGLISH_MONTH_NAMES
].index(abbrev
) + 1
2404 def fix_xml_ampersands(xml_str
):
2405 """Replace all the '&' by '&' in XML"""
2407 r
'&(?!amp;|lt;|gt;|apos;|quot;|#x[0-9a-fA-F]{,4};|#[0-9]{,4};)',
2412 def setproctitle(title
):
2413 assert isinstance(title
, compat_str
)
2415 # ctypes in Jython is not complete
2416 # http://bugs.jython.org/issue2148
2417 if sys
.platform
.startswith('java'):
2421 libc
= ctypes
.cdll
.LoadLibrary('libc.so.6')
2425 # LoadLibrary in Windows Python 2.7.13 only expects
2426 # a bytestring, but since unicode_literals turns
2427 # every string into a unicode string, it fails.
2429 title_bytes
= title
.encode('utf-8')
2430 buf
= ctypes
.create_string_buffer(len(title_bytes
))
2431 buf
.value
= title_bytes
2433 libc
.prctl(15, buf
, 0, 0, 0)
2434 except AttributeError:
2435 return # Strange libc, just skip this
2438 def remove_start(s
, start
):
2439 return s
[len(start
):] if s
is not None and s
.startswith(start
) else s
2442 def remove_end(s
, end
):
2443 return s
[:-len(end
)] if s
is not None and s
.endswith(end
) else s
2446 def remove_quotes(s
):
2447 if s
is None or len(s
) < 2:
2449 for quote
in ('"', "'", ):
2450 if s
[0] == quote
and s
[-1] == quote
:
2455 def get_domain(url
):
2456 domain
= re
.match(r
'(?:https?:\/\/)?(?:www\.)?(?P<domain>[^\n\/]+\.[^\n\/]+)(?:\/(.*))?', url
)
2457 return domain
.group('domain') if domain
else None
2460 def url_basename(url
):
2461 path
= compat_urlparse
.urlparse(url
).path
2462 return path
.strip('/').split('/')[-1]
2466 return re
.match(r
'https?://[^?#&]+/', url
).group()
2469 def urljoin(base
, path
):
2470 if isinstance(path
, bytes):
2471 path
= path
.decode('utf-8')
2472 if not isinstance(path
, compat_str
) or not path
:
2474 if re
.match(r
'^(?:[a-zA-Z][a-zA-Z0-9+-.]*:)?//', path
):
2476 if isinstance(base
, bytes):
2477 base
= base
.decode('utf-8')
2478 if not isinstance(base
, compat_str
) or not re
.match(
2479 r
'^(?:https?:)?//', base
):
2481 return compat_urlparse
.urljoin(base
, path
)
2484 class HEADRequest(compat_urllib_request
.Request
):
2485 def get_method(self
):
2489 class PUTRequest(compat_urllib_request
.Request
):
2490 def get_method(self
):
2494 def int_or_none(v
, scale
=1, default
=None, get_attr
=None, invscale
=1):
2495 if get_attr
and v
is not None:
2496 v
= getattr(v
, get_attr
, None)
2498 return int(v
) * invscale
// scale
2499 except (ValueError, TypeError, OverflowError):
2503 def str_or_none(v
, default
=None):
2504 return default
if v
is None else compat_str(v
)
2507 def str_to_int(int_str
):
2508 """ A more relaxed version of int_or_none """
2509 if isinstance(int_str
, compat_integer_types
):
2511 elif isinstance(int_str
, compat_str
):
2512 int_str
= re
.sub(r
'[,\.\+]', '', int_str
)
2513 return int_or_none(int_str
)
2516 def float_or_none(v
, scale
=1, invscale
=1, default
=None):
2520 return float(v
) * invscale
/ scale
2521 except (ValueError, TypeError):
2525 def bool_or_none(v
, default
=None):
2526 return v
if isinstance(v
, bool) else default
2529 def strip_or_none(v
, default
=None):
2530 return v
.strip() if isinstance(v
, compat_str
) else default
2533 def url_or_none(url
):
2534 if not url
or not isinstance(url
, compat_str
):
2537 return url
if re
.match(r
'^(?:(?:https?|rt(?:m(?:pt?[es]?|fp)|sp[su]?)|mms|ftps?):)?//', url
) else None
2540 def strftime_or_none(timestamp
, date_format
, default
=None):
2541 datetime_object
= None
2543 if isinstance(timestamp
, compat_numeric_types
): # unix timestamp
2544 datetime_object
= datetime
.datetime
.utcfromtimestamp(timestamp
)
2545 elif isinstance(timestamp
, compat_str
): # assume YYYYMMDD
2546 datetime_object
= datetime
.datetime
.strptime(timestamp
, '%Y%m%d')
2547 return datetime_object
.strftime(date_format
)
2548 except (ValueError, TypeError, AttributeError):
2552 def parse_duration(s
):
2553 if not isinstance(s
, compat_basestring
):
2559 days
, hours
, mins
, secs
, ms
= [None] * 5
2560 m
= re
.match(r
'(?:(?:(?:(?P<days>[0-9]+):)?(?P<hours>[0-9]+):)?(?P<mins>[0-9]+):)?(?P<secs>[0-9]+)(?P<ms>\.[0-9]+)?Z?$', s
)
2562 days
, hours
, mins
, secs
, ms
= m
.groups()
2567 [0-9]+\s*y(?:ears?)?\s*
2570 [0-9]+\s*m(?:onths?)?\s*
2573 [0-9]+\s*w(?:eeks?)?\s*
2576 (?P<days>[0-9]+)\s*d(?:ays?)?\s*
2580 (?P<hours>[0-9]+)\s*h(?:ours?)?\s*
2583 (?P<mins>[0-9]+)\s*m(?:in(?:ute)?s?)?\s*
2586 (?P<secs>[0-9]+)(?P<ms>\.[0-9]+)?\s*s(?:ec(?:ond)?s?)?\s*
2589 days
, hours
, mins
, secs
, ms
= m
.groups()
2591 m
= re
.match(r
'(?i)(?:(?P<hours>[0-9.]+)\s*(?:hours?)|(?P<mins>[0-9.]+)\s*(?:mins?\.?|minutes?)\s*)Z?$', s
)
2593 hours
, mins
= m
.groups()
2599 duration
+= float(secs
)
2601 duration
+= float(mins
) * 60
2603 duration
+= float(hours
) * 60 * 60
2605 duration
+= float(days
) * 24 * 60 * 60
2607 duration
+= float(ms
)
2611 def prepend_extension(filename
, ext
, expected_real_ext
=None):
2612 name
, real_ext
= os
.path
.splitext(filename
)
2614 '{0}.{1}{2}'.format(name
, ext
, real_ext
)
2615 if not expected_real_ext
or real_ext
[1:] == expected_real_ext
2616 else '{0}.{1}'.format(filename
, ext
))
2619 def replace_extension(filename
, ext
, expected_real_ext
=None):
2620 name
, real_ext
= os
.path
.splitext(filename
)
2621 return '{0}.{1}'.format(
2622 name
if not expected_real_ext
or real_ext
[1:] == expected_real_ext
else filename
,
2626 def check_executable(exe
, args
=[]):
2627 """ Checks if the given binary is installed somewhere in PATH, and returns its name.
2628 args can be a list of arguments for a short output (like -version) """
2630 Popen([exe
] + args
, stdout
=subprocess
.PIPE
, stderr
=subprocess
.PIPE
).communicate_or_kill()
2636 def _get_exe_version_output(exe
, args
):
2638 # STDIN should be redirected too. On UNIX-like systems, ffmpeg triggers
2639 # SIGTTOU if yt-dlp is run in the background.
2640 # See https://github.com/ytdl-org/youtube-dl/issues/955#issuecomment-209789656
2642 [encodeArgument(exe
)] + args
, stdin
=subprocess
.PIPE
,
2643 stdout
=subprocess
.PIPE
, stderr
=subprocess
.STDOUT
).communicate_or_kill()
2646 if isinstance(out
, bytes): # Python 2.x
2647 out
= out
.decode('ascii', 'ignore')
2651 def detect_exe_version(output
, version_re
=None, unrecognized
='present'):
2652 assert isinstance(output
, compat_str
)
2653 if version_re
is None:
2654 version_re
= r
'version\s+([-0-9._a-zA-Z]+)'
2655 m
= re
.search(version_re
, output
)
2662 def get_exe_version(exe
, args
=['--version'],
2663 version_re
=None, unrecognized
='present'):
2664 """ Returns the version of the specified executable,
2665 or False if the executable is not present """
2666 out
= _get_exe_version_output(exe
, args
)
2667 return detect_exe_version(out
, version_re
, unrecognized
) if out
else False
2670 class LazyList(collections
.abc
.Sequence
):
2671 ''' Lazy immutable list from an iterable
2672 Note that slices of a LazyList are lists and not LazyList'''
2674 class IndexError(IndexError):
2677 def __init__(self
, iterable
, *, reverse
=False, _cache
=None):
2678 self
.__iterable
= iter(iterable
)
2679 self
.__cache
= [] if _cache
is None else _cache
2680 self
.__reversed
= reverse
2684 # We need to consume the entire iterable to iterate in reverse
2685 yield from self
.exhaust()
2687 yield from self
.__cache
2688 for item
in self
.__iterable
:
2689 self
.__cache
.append(item
)
2692 def __exhaust(self
):
2693 self
.__cache
.extend(self
.__iterable
)
2694 # Discard the emptied iterable to make it pickle-able
2695 self
.__iterable
= []
2699 ''' Evaluate the entire iterable '''
2700 return self
.__exhaust
()[::-1 if self
.__reversed
else 1]
2703 def __reverse_index(x
):
2704 return None if x
is None else -(x
+ 1)
2706 def __getitem__(self
, idx
):
2707 if isinstance(idx
, slice):
2709 idx
= slice(self
.__reverse
_index
(idx
.start
), self
.__reverse
_index
(idx
.stop
), -(idx
.step
or 1))
2710 start
, stop
, step
= idx
.start
, idx
.stop
, idx
.step
or 1
2711 elif isinstance(idx
, int):
2713 idx
= self
.__reverse
_index
(idx
)
2714 start
, stop
, step
= idx
, idx
, 0
2716 raise TypeError('indices must be integers or slices')
2717 if ((start
or 0) < 0 or (stop
or 0) < 0
2718 or (start
is None and step
< 0)
2719 or (stop
is None and step
> 0)):
2720 # We need to consume the entire iterable to be able to slice from the end
2721 # Obviously, never use this with infinite iterables
2724 return self
.__cache
[idx
]
2725 except IndexError as e
:
2726 raise self
.IndexError(e
) from e
2727 n
= max(start
or 0, stop
or 0) - len(self
.__cache
) + 1
2729 self
.__cache
.extend(itertools
.islice(self
.__iterable
, n
))
2731 return self
.__cache
[idx
]
2732 except IndexError as e
:
2733 raise self
.IndexError(e
) from e
2737 self
[-1] if self
.__reversed
else self
[0]
2738 except self
.IndexError:
2744 return len(self
.__cache
)
2746 def __reversed__(self
):
2747 return type(self
)(self
.__iterable
, reverse
=not self
.__reversed
, _cache
=self
.__cache
)
2750 return type(self
)(self
.__iterable
, reverse
=self
.__reversed
, _cache
=self
.__cache
)
2753 # repr and str should mimic a list. So we exhaust the iterable
2754 return repr(self
.exhaust())
2757 return repr(self
.exhaust())
2762 class IndexError(IndexError):
2766 # This is only useful for tests
2767 return len(self
.getslice())
2769 def __init__(self
, pagefunc
, pagesize
, use_cache
=True):
2770 self
._pagefunc
= pagefunc
2771 self
._pagesize
= pagesize
2772 self
._use
_cache
= use_cache
2775 def getpage(self
, pagenum
):
2776 page_results
= self
._cache
.get(pagenum
)
2777 if page_results
is None:
2778 page_results
= list(self
._pagefunc
(pagenum
))
2780 self
._cache
[pagenum
] = page_results
2783 def getslice(self
, start
=0, end
=None):
2784 return list(self
._getslice
(start
, end
))
2786 def _getslice(self
, start
, end
):
2787 raise NotImplementedError('This method must be implemented by subclasses')
2789 def __getitem__(self
, idx
):
2790 # NOTE: cache must be enabled if this is used
2791 if not isinstance(idx
, int) or idx
< 0:
2792 raise TypeError('indices must be non-negative integers')
2793 entries
= self
.getslice(idx
, idx
+ 1)
2795 raise self
.IndexError()
2799 class OnDemandPagedList(PagedList
):
2800 def _getslice(self
, start
, end
):
2801 for pagenum
in itertools
.count(start
// self
._pagesize
):
2802 firstid
= pagenum
* self
._pagesize
2803 nextfirstid
= pagenum
* self
._pagesize
+ self
._pagesize
2804 if start
>= nextfirstid
:
2808 start
% self
._pagesize
2809 if firstid
<= start
< nextfirstid
2812 ((end
- 1) % self
._pagesize
) + 1
2813 if (end
is not None and firstid
<= end
<= nextfirstid
)
2816 page_results
= self
.getpage(pagenum
)
2817 if startv
!= 0 or endv
is not None:
2818 page_results
= page_results
[startv
:endv
]
2819 yield from page_results
2821 # A little optimization - if current page is not "full", ie. does
2822 # not contain page_size videos then we can assume that this page
2823 # is the last one - there are no more ids on further pages -
2824 # i.e. no need to query again.
2825 if len(page_results
) + startv
< self
._pagesize
:
2828 # If we got the whole page, but the next page is not interesting,
2829 # break out early as well
2830 if end
== nextfirstid
:
2834 class InAdvancePagedList(PagedList
):
2835 def __init__(self
, pagefunc
, pagecount
, pagesize
):
2836 self
._pagecount
= pagecount
2837 PagedList
.__init
__(self
, pagefunc
, pagesize
, True)
2839 def _getslice(self
, start
, end
):
2840 start_page
= start
// self
._pagesize
2842 self
._pagecount
if end
is None else (end
// self
._pagesize
+ 1))
2843 skip_elems
= start
- start_page
* self
._pagesize
2844 only_more
= None if end
is None else end
- start
2845 for pagenum
in range(start_page
, end_page
):
2846 page_results
= self
.getpage(pagenum
)
2848 page_results
= page_results
[skip_elems
:]
2850 if only_more
is not None:
2851 if len(page_results
) < only_more
:
2852 only_more
-= len(page_results
)
2854 yield from page_results
[:only_more
]
2856 yield from page_results
2859 def uppercase_escape(s
):
2860 unicode_escape
= codecs
.getdecoder('unicode_escape')
2862 r
'\\U[0-9a-fA-F]{8}',
2863 lambda m
: unicode_escape(m
.group(0))[0],
2867 def lowercase_escape(s
):
2868 unicode_escape
= codecs
.getdecoder('unicode_escape')
2870 r
'\\u[0-9a-fA-F]{4}',
2871 lambda m
: unicode_escape(m
.group(0))[0],
2875 def escape_rfc3986(s
):
2876 """Escape non-ASCII characters as suggested by RFC 3986"""
2877 if sys
.version_info
< (3, 0) and isinstance(s
, compat_str
):
2878 s
= s
.encode('utf-8')
2879 return compat_urllib_parse
.quote(s
, b
"%/;:@&=+$,!~*'()?#[]")
2882 def escape_url(url
):
2883 """Escape URL as suggested by RFC 3986"""
2884 url_parsed
= compat_urllib_parse_urlparse(url
)
2885 return url_parsed
._replace
(
2886 netloc
=url_parsed
.netloc
.encode('idna').decode('ascii'),
2887 path
=escape_rfc3986(url_parsed
.path
),
2888 params
=escape_rfc3986(url_parsed
.params
),
2889 query
=escape_rfc3986(url_parsed
.query
),
2890 fragment
=escape_rfc3986(url_parsed
.fragment
)
2895 return compat_parse_qs(compat_urllib_parse_urlparse(url
).query
)
2898 def read_batch_urls(batch_fd
):
2900 if not isinstance(url
, compat_str
):
2901 url
= url
.decode('utf-8', 'replace')
2902 BOM_UTF8
= ('\xef\xbb\xbf', '\ufeff')
2903 for bom
in BOM_UTF8
:
2904 if url
.startswith(bom
):
2905 url
= url
[len(bom
):]
2907 if not url
or url
.startswith(('#', ';', ']')):
2909 # "#" cannot be stripped out since it is part of the URI
2910 # However, it can be safely stipped out if follwing a whitespace
2911 return re
.split(r
'\s#', url
, 1)[0].rstrip()
2913 with contextlib
.closing(batch_fd
) as fd
:
2914 return [url
for url
in map(fixup
, fd
) if url
]
2917 def urlencode_postdata(*args
, **kargs
):
2918 return compat_urllib_parse_urlencode(*args
, **kargs
).encode('ascii')
2921 def update_url_query(url
, query
):
2924 parsed_url
= compat_urlparse
.urlparse(url
)
2925 qs
= compat_parse_qs(parsed_url
.query
)
2927 return compat_urlparse
.urlunparse(parsed_url
._replace
(
2928 query
=compat_urllib_parse_urlencode(qs
, True)))
2931 def update_Request(req
, url
=None, data
=None, headers
={}, query={}
):
2932 req_headers
= req
.headers
.copy()
2933 req_headers
.update(headers
)
2934 req_data
= data
or req
.data
2935 req_url
= update_url_query(url
or req
.get_full_url(), query
)
2936 req_get_method
= req
.get_method()
2937 if req_get_method
== 'HEAD':
2938 req_type
= HEADRequest
2939 elif req_get_method
== 'PUT':
2940 req_type
= PUTRequest
2942 req_type
= compat_urllib_request
.Request
2944 req_url
, data
=req_data
, headers
=req_headers
,
2945 origin_req_host
=req
.origin_req_host
, unverifiable
=req
.unverifiable
)
2946 if hasattr(req
, 'timeout'):
2947 new_req
.timeout
= req
.timeout
2951 def _multipart_encode_impl(data
, boundary
):
2952 content_type
= 'multipart/form-data; boundary=%s' % boundary
2955 for k
, v
in data
.items():
2956 out
+= b
'--' + boundary
.encode('ascii') + b
'\r\n'
2957 if isinstance(k
, compat_str
):
2958 k
= k
.encode('utf-8')
2959 if isinstance(v
, compat_str
):
2960 v
= v
.encode('utf-8')
2961 # RFC 2047 requires non-ASCII field names to be encoded, while RFC 7578
2962 # suggests sending UTF-8 directly. Firefox sends UTF-8, too
2963 content
= b
'Content-Disposition: form-data; name="' + k
+ b
'"\r\n\r\n' + v
+ b
'\r\n'
2964 if boundary
.encode('ascii') in content
:
2965 raise ValueError('Boundary overlaps with data')
2968 out
+= b
'--' + boundary
.encode('ascii') + b
'--\r\n'
2970 return out
, content_type
2973 def multipart_encode(data
, boundary
=None):
2975 Encode a dict to RFC 7578-compliant form-data
2978 A dict where keys and values can be either Unicode or bytes-like
2981 If specified a Unicode object, it's used as the boundary. Otherwise
2982 a random boundary is generated.
2984 Reference: https://tools.ietf.org/html/rfc7578
2986 has_specified_boundary
= boundary
is not None
2989 if boundary
is None:
2990 boundary
= '---------------' + str(random
.randrange(0x0fffffff, 0xffffffff))
2993 out
, content_type
= _multipart_encode_impl(data
, boundary
)
2996 if has_specified_boundary
:
3000 return out
, content_type
3003 def dict_get(d
, key_or_keys
, default
=None, skip_false_values
=True):
3004 if isinstance(key_or_keys
, (list, tuple)):
3005 for key
in key_or_keys
:
3006 if key
not in d
or d
[key
] is None or skip_false_values
and not d
[key
]:
3010 return d
.get(key_or_keys
, default
)
3013 def try_get(src
, getter
, expected_type
=None):
3014 for get
in variadic(getter
):
3017 except (AttributeError, KeyError, TypeError, IndexError):
3020 if expected_type
is None or isinstance(v
, expected_type
):
3024 def merge_dicts(*dicts
):
3026 for a_dict
in dicts
:
3027 for k
, v
in a_dict
.items():
3031 or (isinstance(v
, compat_str
) and v
3032 and isinstance(merged
[k
], compat_str
)
3033 and not merged
[k
])):
3038 def encode_compat_str(string
, encoding
=preferredencoding(), errors
='strict'):
3039 return string
if isinstance(string
, compat_str
) else compat_str(string
, encoding
, errors
)
3051 TV_PARENTAL_GUIDELINES
= {
3061 def parse_age_limit(s
):
3063 return s
if 0 <= s
<= 21 else None
3064 if not isinstance(s
, compat_basestring
):
3066 m
= re
.match(r
'^(?P<age>\d{1,2})\+?$', s
)
3068 return int(m
.group('age'))
3071 return US_RATINGS
[s
]
3072 m
= re
.match(r
'^TV[_-]?(%s)$' % '|'.join(k
[3:] for k
in TV_PARENTAL_GUIDELINES
), s
)
3074 return TV_PARENTAL_GUIDELINES
['TV-' + m
.group(1)]
3078 def strip_jsonp(code
):
3081 (?:window\.)?(?P<func_name>[a-zA-Z0-9_.$]*)
3082 (?:\s*&&\s*(?P=func_name))?
3083 \s*\(\s*(?P<callback_data>.*)\);?
3084 \s*?(?://[^\n]*)*$''',
3085 r
'\g<callback_data>', code
)
3088 def js_to_json(code
, vars={}):
3089 # vars is a dict of var, val pairs to substitute
3090 COMMENT_RE
= r
'/\*(?:(?!\*/).)*?\*/|//[^\n]*\n'
3091 SKIP_RE
= r
'\s*(?:{comment})?\s*'.format(comment
=COMMENT_RE
)
3093 (r
'(?s)^(0[xX][0-9a-fA-F]+){skip}:?$'.format(skip
=SKIP_RE
), 16),
3094 (r
'(?s)^(0+[0-7]+){skip}:?$'.format(skip
=SKIP_RE
), 8),
3099 if v
in ('true', 'false', 'null'):
3101 elif v
in ('undefined', 'void 0'):
3103 elif v
.startswith('/*') or v
.startswith('//') or v
.startswith('!') or v
== ',':
3106 if v
[0] in ("'", '"'):
3107 v
= re
.sub(r
'(?s)\\.|"', lambda m
: {
3112 }.get(m
.group(0), m
.group(0)), v
[1:-1])
3114 for regex
, base
in INTEGER_TABLE
:
3115 im
= re
.match(regex
, v
)
3117 i
= int(im
.group(1), base
)
3118 return '"%d":' % i
if v
.endswith(':') else '%d' % i
3125 return re
.sub(r
'''(?sx)
3126 "(?:[^"\\]*(?:\\\\|\\['"nurtbfx/\n]))*[^"\\]*"|
3127 '(?:[^'\\]*(?:\\\\|\\['"nurtbfx/\n]))*[^'\\]*'|
3128 {comment}|,(?={skip}[\]}}])|
3129 void\s0|(?:(?<![0-9])[eE]|[a-df-zA-DF-Z_$])[.a-zA-Z_$0-9]*|
3130 \b(?:0[xX][0-9a-fA-F]+|0+[0-7]+)(?:{skip}:)?|
3133 '''.format(comment
=COMMENT_RE
, skip
=SKIP_RE
), fix_kv
, code
)
3136 def qualities(quality_ids
):
3137 """ Get a numeric quality value out of a list of possible values """
3140 return quality_ids
.index(qid
)
3146 POSTPROCESS_WHEN
= {'pre_process', 'before_dl', 'after_move', 'post_process', 'after_video', 'playlist'}
3150 'default': '%(title)s [%(id)s].%(ext)s',
3151 'chapter': '%(title)s - %(section_number)03d %(section_title)s [%(id)s].%(ext)s',
3157 'description': 'description',
3158 'annotation': 'annotations.xml',
3159 'infojson': 'info.json',
3161 'pl_thumbnail': None,
3162 'pl_description': 'description',
3163 'pl_infojson': 'info.json',
3166 # As of [1] format syntax is:
3167 # %[mapping_key][conversion_flags][minimum_width][.precision][length_modifier]type
3168 # 1. https://docs.python.org/2/library/stdtypes.html#string-formatting
3169 STR_FORMAT_RE_TMPL
= r
'''(?x)
3170 (?<!%)(?P<prefix>(?:%%)*)
3172 (?P<has_key>\((?P<key>{0})\))?
3174 (?P<conversion>[#0\-+ ]+)?
3176 (?P<precision>\.\d+)?
3177 (?P<len_mod>[hlL])? # unused in python
3178 {1} # conversion type
3183 STR_FORMAT_TYPES
= 'diouxXeEfFgGcrs'
3186 def limit_length(s
, length
):
3187 """ Add ellipses to overly long strings """
3192 return s
[:length
- len(ELLIPSES
)] + ELLIPSES
3196 def version_tuple(v
):
3197 return tuple(int(e
) for e
in re
.split(r
'[-.]', v
))
3200 def is_outdated_version(version
, limit
, assume_new
=True):
3202 return not assume_new
3204 return version_tuple(version
) < version_tuple(limit
)
3206 return not assume_new
3209 def ytdl_is_updateable():
3210 """ Returns if yt-dlp can be updated with -U """
3212 from .update
import is_non_updateable
3214 return not is_non_updateable()
3217 def args_to_str(args
):
3218 # Get a short string representation for a subprocess command
3219 return ' '.join(compat_shlex_quote(a
) for a
in args
)
3222 def error_to_compat_str(err
):
3224 # On python 2 error byte string must be decoded with proper
3225 # encoding rather than ascii
3226 if sys
.version_info
[0] < 3:
3227 err_str
= err_str
.decode(preferredencoding())
3231 def mimetype2ext(mt
):
3235 mt
, _
, params
= mt
.partition(';')
3240 # Per RFC 3003, audio/mpeg can be .mp1, .mp2 or .mp3. Here use .mp3 as
3241 # it's the most popular one
3242 'audio/mpeg': 'mp3',
3243 'audio/x-wav': 'wav',
3245 'audio/wave': 'wav',
3248 ext
= FULL_MAP
.get(mt
)
3254 'smptett+xml': 'tt',
3258 'x-mp4-fragmented': 'mp4',
3259 'x-ms-sami': 'sami',
3262 'x-mpegurl': 'm3u8',
3263 'vnd.apple.mpegurl': 'm3u8',
3267 'vnd.ms-sstr+xml': 'ism',
3271 'filmstrip+json': 'fs',
3275 _
, _
, subtype
= mt
.rpartition('/')
3276 ext
= SUBTYPE_MAP
.get(subtype
.lower())
3287 _
, _
, suffix
= subtype
.partition('+')
3288 ext
= SUFFIX_MAP
.get(suffix
)
3292 return subtype
.replace('+', '.')
3295 def ext2mimetype(ext_or_url
):
3298 if '.' not in ext_or_url
:
3299 ext_or_url
= f
'file.{ext_or_url}'
3300 return mimetypes
.guess_type(ext_or_url
)[0]
3303 def parse_codecs(codecs_str
):
3304 # http://tools.ietf.org/html/rfc6381
3307 split_codecs
= list(filter(None, map(
3308 str.strip
, codecs_str
.strip().strip(',').split(','))))
3309 vcodec
, acodec
, tcodec
, hdr
= None, None, None, None
3310 for full_codec
in split_codecs
:
3311 parts
= full_codec
.split('.')
3312 codec
= parts
[0].replace('0', '')
3313 if codec
in ('avc1', 'avc2', 'avc3', 'avc4', 'vp9', 'vp8', 'hev1', 'hev2',
3314 'h263', 'h264', 'mp4v', 'hvc1', 'av1', 'theora', 'dvh1', 'dvhe'):
3316 vcodec
= '.'.join(parts
[:4]) if codec
in ('vp9', 'av1', 'hvc1') else full_codec
3317 if codec
in ('dvh1', 'dvhe'):
3319 elif codec
== 'av1' and len(parts
) > 3 and parts
[3] == '10':
3321 elif full_codec
.replace('0', '').startswith('vp9.2'):
3323 elif codec
in ('flac', 'mp4a', 'opus', 'vorbis', 'mp3', 'aac', 'ac-3', 'ec-3', 'eac3', 'dtsc', 'dtse', 'dtsh', 'dtsl'):
3326 elif codec
in ('stpp', 'wvtt',):
3330 write_string('WARNING: Unknown codec %s\n' % full_codec
, sys
.stderr
)
3331 if vcodec
or acodec
or tcodec
:
3333 'vcodec': vcodec
or 'none',
3334 'acodec': acodec
or 'none',
3335 'dynamic_range': hdr
,
3336 **({'tcodec': tcodec}
if tcodec
is not None else {}),
3338 elif len(split_codecs
) == 2:
3340 'vcodec': split_codecs
[0],
3341 'acodec': split_codecs
[1],
3346 def urlhandle_detect_ext(url_handle
):
3347 getheader
= url_handle
.headers
.get
3349 cd
= getheader('Content-Disposition')
3351 m
= re
.match(r
'attachment;\s*filename="(?P<filename>[^"]+)"', cd
)
3353 e
= determine_ext(m
.group('filename'), default_ext
=None)
3357 return mimetype2ext(getheader('Content-Type'))
3360 def encode_data_uri(data
, mime_type
):
3361 return 'data:%s;base64,%s' % (mime_type
, base64
.b64encode(data
).decode('ascii'))
3364 def age_restricted(content_limit
, age_limit
):
3365 """ Returns True iff the content should be blocked """
3367 if age_limit
is None: # No limit set
3369 if content_limit
is None:
3370 return False # Content available for everyone
3371 return age_limit
< content_limit
3374 def is_html(first_bytes
):
3375 """ Detect whether a file contains HTML by examining its first bytes. """
3378 (b
'\xef\xbb\xbf', 'utf-8'),
3379 (b
'\x00\x00\xfe\xff', 'utf-32-be'),
3380 (b
'\xff\xfe\x00\x00', 'utf-32-le'),
3381 (b
'\xff\xfe', 'utf-16-le'),
3382 (b
'\xfe\xff', 'utf-16-be'),
3384 for bom
, enc
in BOMS
:
3385 if first_bytes
.startswith(bom
):
3386 s
= first_bytes
[len(bom
):].decode(enc
, 'replace')
3389 s
= first_bytes
.decode('utf-8', 'replace')
3391 return re
.match(r
'^\s*<', s
)
3394 def determine_protocol(info_dict
):
3395 protocol
= info_dict
.get('protocol')
3396 if protocol
is not None:
3399 url
= sanitize_url(info_dict
['url'])
3400 if url
.startswith('rtmp'):
3402 elif url
.startswith('mms'):
3404 elif url
.startswith('rtsp'):
3407 ext
= determine_ext(url
)
3413 return compat_urllib_parse_urlparse(url
).scheme
3416 def render_table(header_row
, data
, delim
=False, extra_gap
=0, hide_empty
=False):
3417 """ Render a list of rows, each as a list of values.
3418 Text after a \t will be right aligned """
3420 return len(remove_terminal_sequences(string
).replace('\t', ''))
3422 def get_max_lens(table
):
3423 return [max(width(str(v
)) for v
in col
) for col
in zip(*table
)]
3425 def filter_using_list(row
, filterArray
):
3426 return [col
for (take
, col
) in zip(filterArray
, row
) if take
]
3429 max_lens
= get_max_lens(data
)
3430 header_row
= filter_using_list(header_row
, max_lens
)
3431 data
= [filter_using_list(row
, max_lens
) for row
in data
]
3433 table
= [header_row
] + data
3434 max_lens
= get_max_lens(table
)
3437 table
= [header_row
, [delim
* (ml
+ extra_gap
) for ml
in max_lens
]] + data
3438 table
[1][-1] = table
[1][-1][:-extra_gap
] # Remove extra_gap from end of delimiter
3440 for pos
, text
in enumerate(map(str, row
)):
3442 row
[pos
] = text
.replace('\t', ' ' * (max_lens
[pos
] - width(text
))) + ' ' * extra_gap
3444 row
[pos
] = text
+ ' ' * (max_lens
[pos
] - width(text
) + extra_gap
)
3445 ret
= '\n'.join(''.join(row
).rstrip() for row
in table
)
3449 def _match_one(filter_part
, dct
, incomplete
):
3450 # TODO: Generalize code with YoutubeDL._build_format_filter
3451 STRING_OPERATORS
= {
3452 '*=': operator
.contains
,
3453 '^=': lambda attr
, value
: attr
.startswith(value
),
3454 '$=': lambda attr
, value
: attr
.endswith(value
),
3455 '~=': lambda attr
, value
: re
.search(value
, attr
),
3457 COMPARISON_OPERATORS
= {
3459 '<=': operator
.le
, # "<=" must be defined above "<"
3466 operator_rex
= re
.compile(r
'''(?x)\s*
3468 \s*(?P<negation>!\s*)?(?P<op>%s)(?P<none_inclusive>\s*\?)?\s*
3470 (?P<quote>["\'])(?P
<quotedstrval
>.+?
)(?P
=quote
)|
3474 ''' % '|'.join(map(re.escape, COMPARISON_OPERATORS.keys())))
3475 m = operator_rex.search(filter_part)
3478 unnegated_op = COMPARISON_OPERATORS[m['op']]
3480 op = lambda attr, value: not unnegated_op(attr, value)
3483 comparison_value = m['quotedstrval'] or m['strval'] or m['intval']
3485 comparison_value = comparison_value.replace(r'\%s' % m['quote'], m['quote'])
3486 actual_value = dct.get(m['key'])
3487 numeric_comparison = None
3488 if isinstance(actual_value, compat_numeric_types):
3489 # If the original field is a string and matching comparisonvalue is
3490 # a number we should respect the origin of the original field
3491 # and process comparison value as a string (see
3492 # https://github.com/ytdl-org/youtube-dl/issues/11082)
3494 numeric_comparison = int(comparison_value)
3496 numeric_comparison = parse_filesize(comparison_value)
3497 if numeric_comparison is None:
3498 numeric_comparison = parse_filesize(f'{comparison_value}B')
3499 if numeric_comparison is None:
3500 numeric_comparison = parse_duration(comparison_value)
3501 if numeric_comparison is not None and m['op'] in STRING_OPERATORS:
3502 raise ValueError('Operator %s only supports string values!' % m['op'])
3503 if actual_value is None:
3504 return incomplete or m['none_inclusive']
3505 return op(actual_value, comparison_value if numeric_comparison is None else numeric_comparison)
3508 '': lambda v: (v is True) if isinstance(v, bool) else (v is not None),
3509 '!': lambda v: (v is False) if isinstance(v, bool) else (v is None),
3511 operator_rex = re.compile(r'''(?x
)\s
*
3512 (?P
<op
>%s)\s
*(?P
<key
>[a
-z_
]+)
3514 ''' % '|'.join(map(re.escape, UNARY_OPERATORS.keys())))
3515 m = operator_rex.search(filter_part)
3517 op = UNARY_OPERATORS[m.group('op')]
3518 actual_value = dct.get(m.group('key'))
3519 if incomplete and actual_value is None:
3521 return op(actual_value)
3523 raise ValueError('Invalid filter part %r' % filter_part)
3526 def match_str(filter_str, dct, incomplete=False):
3527 """ Filter a dictionary with a simple string syntax. Returns True (=passes filter) or false
3528 When incomplete, all conditions passes on missing fields
3531 _match_one(filter_part.replace(r'\&', '&'), dct, incomplete)
3532 for filter_part in re.split(r'(?<!\\)&', filter_str))
3535 def match_filter_func(filter_str):
3536 def _match_func(info_dict, *args, **kwargs):
3537 if match_str(filter_str, info_dict, *args, **kwargs):
3540 video_title = info_dict.get('title', info_dict.get('id', 'video'))
3541 return '%s does not pass filter %s, skipping ..' % (video_title, filter_str)
3545 def parse_dfxp_time_expr(time_expr):
3549 mobj = re.match(r'^(?P<time_offset>\d+(?:\.\d+)?)s?$', time_expr)
3551 return float(mobj.group('time_offset'))
3553 mobj = re.match(r'^(\d+):(\d\d):(\d\d(?:(?:\.|:)\d+)?)$', time_expr)
3555 return 3600 * int(mobj.group(1)) + 60 * int(mobj.group(2)) + float(mobj.group(3).replace(':', '.'))
3558 def srt_subtitles_timecode(seconds):
3559 return '%02d:%02d:%02d,%03d' % timetuple_from_msec(seconds * 1000)
3562 def ass_subtitles_timecode(seconds):
3563 time = timetuple_from_msec(seconds * 1000)
3564 return '%01d:%02d:%02d.%02d' % (*time[:-1], time.milliseconds / 10)
3567 def dfxp2srt(dfxp_data):
3569 @param dfxp_data A
bytes-like
object containing DFXP data
3570 @returns A
unicode object containing converted SRT data
3572 LEGACY_NAMESPACES = (
3573 (b'http://www.w3.org/ns/ttml', [
3574 b'http://www.w3.org/2004/11/ttaf1',
3575 b'http://www.w3.org/2006/04/ttaf1',
3576 b'http://www.w3.org/2006/10/ttaf1',
3578 (b'http://www.w3.org/ns/ttml#styling', [
3579 b'http://www.w3.org/ns/ttml#style',
3583 SUPPORTED_STYLING = [
3592 _x = functools.partial(xpath_with_ns, ns_map={
3593 'xml': 'http://www.w3.org/XML/1998/namespace',
3594 'ttml': 'http://www.w3.org/ns/ttml',
3595 'tts': 'http://www.w3.org/ns/ttml#styling',
3601 class TTMLPElementParser(object):
3603 _unclosed_elements = []
3604 _applied_styles = []
3606 def start(self, tag, attrib):
3607 if tag in (_x('ttml:br'), 'br'):
3610 unclosed_elements = []
3612 element_style_id = attrib.get('style')
3614 style.update(default_style)
3615 if element_style_id:
3616 style.update(styles.get(element_style_id, {}))
3617 for prop in SUPPORTED_STYLING:
3618 prop_val = attrib.get(_x('tts:' + prop))
3620 style[prop] = prop_val
3623 for k, v in sorted(style.items()):
3624 if self._applied_styles and self._applied_styles[-1].get(k) == v:
3627 font += ' color="%s"' % v
3628 elif k == 'fontSize':
3629 font += ' size="%s"' % v
3630 elif k == 'fontFamily':
3631 font += ' face="%s"' % v
3632 elif k == 'fontWeight' and v == 'bold':
3634 unclosed_elements.append('b')
3635 elif k == 'fontStyle' and v == 'italic':
3637 unclosed_elements.append('i')
3638 elif k == 'textDecoration' and v == 'underline':
3640 unclosed_elements.append('u')
3642 self._out += '<font' + font + '>'
3643 unclosed_elements.append('font')
3645 if self._applied_styles:
3646 applied_style.update(self._applied_styles[-1])
3647 applied_style.update(style)
3648 self._applied_styles.append(applied_style)
3649 self._unclosed_elements.append(unclosed_elements)
3652 if tag not in (_x('ttml:br'), 'br'):
3653 unclosed_elements = self._unclosed_elements.pop()
3654 for element in reversed(unclosed_elements):
3655 self._out += '</%s>' % element
3656 if unclosed_elements and self._applied_styles:
3657 self._applied_styles.pop()
3659 def data(self, data):
3663 return self._out.strip()
3665 def parse_node(node):
3666 target = TTMLPElementParser()
3667 parser = xml.etree.ElementTree.XMLParser(target=target)
3668 parser.feed(xml.etree.ElementTree.tostring(node))
3669 return parser.close()
3671 for k, v in LEGACY_NAMESPACES:
3673 dfxp_data = dfxp_data.replace(ns, k)
3675 dfxp = compat_etree_fromstring(dfxp_data)
3677 paras = dfxp.findall(_x('.//ttml:p')) or dfxp.findall('.//p')
3680 raise ValueError('Invalid dfxp/TTML subtitle')
3684 for style in dfxp.findall(_x('.//ttml:style')):
3685 style_id = style.get('id') or style.get(_x('xml:id'))
3688 parent_style_id = style.get('style')
3690 if parent_style_id not in styles:
3693 styles[style_id] = styles[parent_style_id].copy()
3694 for prop in SUPPORTED_STYLING:
3695 prop_val = style.get(_x('tts:' + prop))
3697 styles.setdefault(style_id, {})[prop] = prop_val
3703 for p in ('body', 'div'):
3704 ele = xpath_element(dfxp, [_x('.//ttml:' + p), './/' + p])
3707 style = styles.get(ele.get('style'))
3710 default_style.update(style)
3712 for para, index in zip(paras, itertools.count(1)):
3713 begin_time = parse_dfxp_time_expr(para.attrib.get('begin'))
3714 end_time = parse_dfxp_time_expr(para.attrib.get('end'))
3715 dur = parse_dfxp_time_expr(para.attrib.get('dur'))
3716 if begin_time is None:
3721 end_time = begin_time + dur
3722 out.append('%d\n%s --> %s\n%s\n\n' % (
3724 srt_subtitles_timecode(begin_time),
3725 srt_subtitles_timecode(end_time),
3731 def cli_option(params, command_option, param):
3732 param = params.get(param)
3734 param = compat_str(param)
3735 return [command_option, param] if param is not None else []
3738 def cli_bool_option(params, command_option, param, true_value='true', false_value='false', separator=None):
3739 param = params.get(param)
3742 assert isinstance(param, bool)
3744 return [command_option + separator + (true_value if param else false_value)]
3745 return [command_option, true_value if param else false_value]
3748 def cli_valueless_option(params, command_option, param, expected_value=True):
3749 param = params.get(param)
3750 return [command_option] if param == expected_value else []
3753 def cli_configuration_args(argdict, keys, default=[], use_compat=True):
3754 if isinstance(argdict, (list, tuple)): # for backward compatibility
3761 assert isinstance(argdict, dict)
3763 assert isinstance(keys, (list, tuple))
3764 for key_list in keys:
3765 arg_list = list(filter(
3766 lambda x: x is not None,
3767 [argdict.get(key.lower()) for key in variadic(key_list)]))
3769 return [arg for args in arg_list for arg in args]
3773 def _configuration_args(main_key, argdict, exe, keys=None, default=[], use_compat=True):
3774 main_key, exe = main_key.lower(), exe.lower()
3775 root_key = exe if main_key == exe else f'{main_key}+{exe}'
3776 keys = [f'{root_key}{k}' for k in (keys or [''])]
3777 if root_key in keys:
3779 keys.append((main_key, exe))
3780 keys.append('default')
3783 return cli_configuration_args(argdict, keys, default, use_compat)
3786 class ISO639Utils(object):
3787 # See http://www.loc.gov/standards/iso639-2/ISO-639-2_utf-8.txt
3846 'iw': 'heb', # Replaced by he in 1989 revision
3856 'in': 'ind', # Replaced by id in 1989 revision
3971 'ji': 'yid', # Replaced by yi in 1989 revision
3979 def short2long(cls, code):
3980 """Convert language code from ISO 639-1 to ISO 639-2/T"""
3981 return cls._lang_map.get(code[:2])
3984 def long2short(cls, code):
3985 """Convert language code from ISO 639-2/T to ISO 639-1"""
3986 for short_name, long_name in cls._lang_map.items():
3987 if long_name == code:
3991 class ISO3166Utils(object):
3992 # From http://data.okfn.org/data/core/country-list
3994 'AF': 'Afghanistan',
3995 'AX': 'Åland Islands',
3998 'AS': 'American Samoa',
4003 'AG': 'Antigua and Barbuda',
4020 'BO': 'Bolivia, Plurinational State of',
4021 'BQ': 'Bonaire, Sint Eustatius and Saba',
4022 'BA': 'Bosnia and Herzegovina',
4024 'BV': 'Bouvet Island',
4026 'IO': 'British Indian Ocean Territory',
4027 'BN': 'Brunei Darussalam',
4029 'BF': 'Burkina Faso',
4035 'KY': 'Cayman Islands',
4036 'CF': 'Central African Republic',
4040 'CX': 'Christmas Island',
4041 'CC': 'Cocos (Keeling) Islands',
4045 'CD': 'Congo, the Democratic Republic of the',
4046 'CK': 'Cook Islands',
4048 'CI': 'Côte d\'Ivoire',
4053 'CZ': 'Czech Republic',
4057 'DO': 'Dominican Republic',
4060 'SV': 'El Salvador',
4061 'GQ': 'Equatorial Guinea',
4065 'FK': 'Falkland Islands (Malvinas)',
4066 'FO': 'Faroe Islands',
4070 'GF': 'French Guiana',
4071 'PF': 'French Polynesia',
4072 'TF': 'French Southern Territories',
4087 'GW': 'Guinea-Bissau',
4090 'HM': 'Heard Island and McDonald Islands',
4091 'VA': 'Holy See (Vatican City State)',
4098 'IR': 'Iran, Islamic Republic of',
4101 'IM': 'Isle of Man',
4111 'KP': 'Korea, Democratic People\'s Republic of',
4112 'KR': 'Korea, Republic of',
4115 'LA': 'Lao People\'s Democratic Republic',
4121 'LI': 'Liechtenstein',
4125 'MK': 'Macedonia, the Former Yugoslav Republic of',
4132 'MH': 'Marshall Islands',
4138 'FM': 'Micronesia, Federated States of',
4139 'MD': 'Moldova, Republic of',
4150 'NL': 'Netherlands',
4151 'NC': 'New Caledonia',
4152 'NZ': 'New Zealand',
4157 'NF': 'Norfolk Island',
4158 'MP': 'Northern Mariana Islands',
4163 'PS': 'Palestine, State of',
4165 'PG': 'Papua New Guinea',
4168 'PH': 'Philippines',
4172 'PR': 'Puerto Rico',
4176 'RU': 'Russian Federation',
4178 'BL': 'Saint Barthélemy',
4179 'SH': 'Saint Helena, Ascension and Tristan da Cunha',
4180 'KN': 'Saint Kitts and Nevis',
4181 'LC': 'Saint Lucia',
4182 'MF': 'Saint Martin (French part)',
4183 'PM': 'Saint Pierre and Miquelon',
4184 'VC': 'Saint Vincent and the Grenadines',
4187 'ST': 'Sao Tome and Principe',
4188 'SA': 'Saudi Arabia',
4192 'SL': 'Sierra Leone',
4194 'SX': 'Sint Maarten (Dutch part)',
4197 'SB': 'Solomon Islands',
4199 'ZA': 'South Africa',
4200 'GS': 'South Georgia and the South Sandwich Islands',
4201 'SS': 'South Sudan',
4206 'SJ': 'Svalbard and Jan Mayen',
4209 'CH': 'Switzerland',
4210 'SY': 'Syrian Arab Republic',
4211 'TW': 'Taiwan, Province of China',
4213 'TZ': 'Tanzania, United Republic of',
4215 'TL': 'Timor-Leste',
4219 'TT': 'Trinidad and Tobago',
4222 'TM': 'Turkmenistan',
4223 'TC': 'Turks and Caicos Islands',
4227 'AE': 'United Arab Emirates',
4228 'GB': 'United Kingdom',
4229 'US': 'United States',
4230 'UM': 'United States Minor Outlying Islands',
4234 'VE': 'Venezuela, Bolivarian Republic of',
4236 'VG': 'Virgin Islands, British',
4237 'VI': 'Virgin Islands, U.S.',
4238 'WF': 'Wallis and Futuna',
4239 'EH': 'Western Sahara',
4246 def short2full(cls, code):
4247 """Convert an ISO 3166-2 country code to the corresponding full name"""
4248 return cls._country_map.get(code.upper())
4251 class GeoUtils(object):
4252 # Major IPv4 address blocks per country
4254 'AD': '46.172.224.0/19',
4255 'AE': '94.200.0.0/13',
4256 'AF': '149.54.0.0/17',
4257 'AG': '209.59.64.0/18',
4258 'AI': '204.14.248.0/21',
4259 'AL': '46.99.0.0/16',
4260 'AM': '46.70.0.0/15',
4261 'AO': '105.168.0.0/13',
4262 'AP': '182.50.184.0/21',
4263 'AQ': '23.154.160.0/24',
4264 'AR': '181.0.0.0/12',
4265 'AS': '202.70.112.0/20',
4266 'AT': '77.116.0.0/14',
4267 'AU': '1.128.0.0/11',
4268 'AW': '181.41.0.0/18',
4269 'AX': '185.217.4.0/22',
4270 'AZ': '5.197.0.0/16',
4271 'BA': '31.176.128.0/17',
4272 'BB': '65.48.128.0/17',
4273 'BD': '114.130.0.0/16',
4275 'BF': '102.178.0.0/15',
4276 'BG': '95.42.0.0/15',
4277 'BH': '37.131.0.0/17',
4278 'BI': '154.117.192.0/18',
4279 'BJ': '137.255.0.0/16',
4280 'BL': '185.212.72.0/23',
4281 'BM': '196.12.64.0/18',
4282 'BN': '156.31.0.0/16',
4283 'BO': '161.56.0.0/16',
4284 'BQ': '161.0.80.0/20',
4285 'BR': '191.128.0.0/12',
4286 'BS': '24.51.64.0/18',
4287 'BT': '119.2.96.0/19',
4288 'BW': '168.167.0.0/16',
4289 'BY': '178.120.0.0/13',
4290 'BZ': '179.42.192.0/18',
4291 'CA': '99.224.0.0/11',
4292 'CD': '41.243.0.0/16',
4293 'CF': '197.242.176.0/21',
4294 'CG': '160.113.0.0/16',
4295 'CH': '85.0.0.0/13',
4296 'CI': '102.136.0.0/14',
4297 'CK': '202.65.32.0/19',
4298 'CL': '152.172.0.0/14',
4299 'CM': '102.244.0.0/14',
4300 'CN': '36.128.0.0/10',
4301 'CO': '181.240.0.0/12',
4302 'CR': '201.192.0.0/12',
4303 'CU': '152.206.0.0/15',
4304 'CV': '165.90.96.0/19',
4305 'CW': '190.88.128.0/17',
4306 'CY': '31.153.0.0/16',
4307 'CZ': '88.100.0.0/14',
4309 'DJ': '197.241.0.0/17',
4310 'DK': '87.48.0.0/12',
4311 'DM': '192.243.48.0/20',
4312 'DO': '152.166.0.0/15',
4313 'DZ': '41.96.0.0/12',
4314 'EC': '186.68.0.0/15',
4315 'EE': '90.190.0.0/15',
4316 'EG': '156.160.0.0/11',
4317 'ER': '196.200.96.0/20',
4318 'ES': '88.0.0.0/11',
4319 'ET': '196.188.0.0/14',
4320 'EU': '2.16.0.0/13',
4321 'FI': '91.152.0.0/13',
4322 'FJ': '144.120.0.0/16',
4323 'FK': '80.73.208.0/21',
4324 'FM': '119.252.112.0/20',
4325 'FO': '88.85.32.0/19',
4327 'GA': '41.158.0.0/15',
4329 'GD': '74.122.88.0/21',
4330 'GE': '31.146.0.0/16',
4331 'GF': '161.22.64.0/18',
4332 'GG': '62.68.160.0/19',
4333 'GH': '154.160.0.0/12',
4334 'GI': '95.164.0.0/16',
4335 'GL': '88.83.0.0/19',
4336 'GM': '160.182.0.0/15',
4337 'GN': '197.149.192.0/18',
4338 'GP': '104.250.0.0/19',
4339 'GQ': '105.235.224.0/20',
4340 'GR': '94.64.0.0/13',
4341 'GT': '168.234.0.0/16',
4342 'GU': '168.123.0.0/16',
4343 'GW': '197.214.80.0/20',
4344 'GY': '181.41.64.0/18',
4345 'HK': '113.252.0.0/14',
4346 'HN': '181.210.0.0/16',
4347 'HR': '93.136.0.0/13',
4348 'HT': '148.102.128.0/17',
4349 'HU': '84.0.0.0/14',
4350 'ID': '39.192.0.0/10',
4351 'IE': '87.32.0.0/12',
4352 'IL': '79.176.0.0/13',
4353 'IM': '5.62.80.0/20',
4354 'IN': '117.192.0.0/10',
4355 'IO': '203.83.48.0/21',
4356 'IQ': '37.236.0.0/14',
4357 'IR': '2.176.0.0/12',
4358 'IS': '82.221.0.0/16',
4359 'IT': '79.0.0.0/10',
4360 'JE': '87.244.64.0/18',
4361 'JM': '72.27.0.0/17',
4362 'JO': '176.29.0.0/16',
4363 'JP': '133.0.0.0/8',
4364 'KE': '105.48.0.0/12',
4365 'KG': '158.181.128.0/17',
4366 'KH': '36.37.128.0/17',
4367 'KI': '103.25.140.0/22',
4368 'KM': '197.255.224.0/20',
4369 'KN': '198.167.192.0/19',
4370 'KP': '175.45.176.0/22',
4371 'KR': '175.192.0.0/10',
4372 'KW': '37.36.0.0/14',
4373 'KY': '64.96.0.0/15',
4374 'KZ': '2.72.0.0/13',
4375 'LA': '115.84.64.0/18',
4376 'LB': '178.135.0.0/16',
4377 'LC': '24.92.144.0/20',
4378 'LI': '82.117.0.0/19',
4379 'LK': '112.134.0.0/15',
4380 'LR': '102.183.0.0/16',
4381 'LS': '129.232.0.0/17',
4382 'LT': '78.56.0.0/13',
4383 'LU': '188.42.0.0/16',
4384 'LV': '46.109.0.0/16',
4385 'LY': '41.252.0.0/14',
4386 'MA': '105.128.0.0/11',
4387 'MC': '88.209.64.0/18',
4388 'MD': '37.246.0.0/16',
4389 'ME': '178.175.0.0/17',
4390 'MF': '74.112.232.0/21',
4391 'MG': '154.126.0.0/17',
4392 'MH': '117.103.88.0/21',
4393 'MK': '77.28.0.0/15',
4394 'ML': '154.118.128.0/18',
4395 'MM': '37.111.0.0/17',
4396 'MN': '49.0.128.0/17',
4397 'MO': '60.246.0.0/16',
4398 'MP': '202.88.64.0/20',
4399 'MQ': '109.203.224.0/19',
4400 'MR': '41.188.64.0/18',
4401 'MS': '208.90.112.0/22',
4402 'MT': '46.11.0.0/16',
4403 'MU': '105.16.0.0/12',
4404 'MV': '27.114.128.0/18',
4405 'MW': '102.70.0.0/15',
4406 'MX': '187.192.0.0/11',
4407 'MY': '175.136.0.0/13',
4408 'MZ': '197.218.0.0/15',
4409 'NA': '41.182.0.0/16',
4410 'NC': '101.101.0.0/18',
4411 'NE': '197.214.0.0/18',
4412 'NF': '203.17.240.0/22',
4413 'NG': '105.112.0.0/12',
4414 'NI': '186.76.0.0/15',
4415 'NL': '145.96.0.0/11',
4416 'NO': '84.208.0.0/13',
4417 'NP': '36.252.0.0/15',
4418 'NR': '203.98.224.0/19',
4419 'NU': '49.156.48.0/22',
4420 'NZ': '49.224.0.0/14',
4421 'OM': '5.36.0.0/15',
4422 'PA': '186.72.0.0/15',
4423 'PE': '186.160.0.0/14',
4424 'PF': '123.50.64.0/18',
4425 'PG': '124.240.192.0/19',
4426 'PH': '49.144.0.0/13',
4427 'PK': '39.32.0.0/11',
4428 'PL': '83.0.0.0/11',
4429 'PM': '70.36.0.0/20',
4430 'PR': '66.50.0.0/16',
4431 'PS': '188.161.0.0/16',
4432 'PT': '85.240.0.0/13',
4433 'PW': '202.124.224.0/20',
4434 'PY': '181.120.0.0/14',
4435 'QA': '37.210.0.0/15',
4436 'RE': '102.35.0.0/16',
4437 'RO': '79.112.0.0/13',
4438 'RS': '93.86.0.0/15',
4439 'RU': '5.136.0.0/13',
4440 'RW': '41.186.0.0/16',
4441 'SA': '188.48.0.0/13',
4442 'SB': '202.1.160.0/19',
4443 'SC': '154.192.0.0/11',
4444 'SD': '102.120.0.0/13',
4445 'SE': '78.64.0.0/12',
4446 'SG': '8.128.0.0/10',
4447 'SI': '188.196.0.0/14',
4448 'SK': '78.98.0.0/15',
4449 'SL': '102.143.0.0/17',
4450 'SM': '89.186.32.0/19',
4451 'SN': '41.82.0.0/15',
4452 'SO': '154.115.192.0/18',
4453 'SR': '186.179.128.0/17',
4454 'SS': '105.235.208.0/21',
4455 'ST': '197.159.160.0/19',
4456 'SV': '168.243.0.0/16',
4457 'SX': '190.102.0.0/20',
4459 'SZ': '41.84.224.0/19',
4460 'TC': '65.255.48.0/20',
4461 'TD': '154.68.128.0/19',
4462 'TG': '196.168.0.0/14',
4463 'TH': '171.96.0.0/13',
4464 'TJ': '85.9.128.0/18',
4465 'TK': '27.96.24.0/21',
4466 'TL': '180.189.160.0/20',
4467 'TM': '95.85.96.0/19',
4468 'TN': '197.0.0.0/11',
4469 'TO': '175.176.144.0/21',
4470 'TR': '78.160.0.0/11',
4471 'TT': '186.44.0.0/15',
4472 'TV': '202.2.96.0/19',
4473 'TW': '120.96.0.0/11',
4474 'TZ': '156.156.0.0/14',
4475 'UA': '37.52.0.0/14',
4476 'UG': '102.80.0.0/13',
4478 'UY': '167.56.0.0/13',
4479 'UZ': '84.54.64.0/18',
4480 'VA': '212.77.0.0/19',
4481 'VC': '207.191.240.0/21',
4482 'VE': '186.88.0.0/13',
4483 'VG': '66.81.192.0/20',
4484 'VI': '146.226.0.0/16',
4485 'VN': '14.160.0.0/11',
4486 'VU': '202.80.32.0/20',
4487 'WF': '117.20.32.0/21',
4488 'WS': '202.4.32.0/19',
4489 'YE': '134.35.0.0/16',
4490 'YT': '41.242.116.0/22',
4491 'ZA': '41.0.0.0/11',
4492 'ZM': '102.144.0.0/13',
4493 'ZW': '102.177.192.0/18',
4497 def random_ipv4(cls, code_or_block):
4498 if len(code_or_block) == 2:
4499 block = cls._country_ip_map.get(code_or_block.upper())
4503 block = code_or_block
4504 addr, preflen = block.split('/')
4505 addr_min = compat_struct_unpack('!L', socket.inet_aton(addr))[0]
4506 addr_max = addr_min | (0xffffffff >> int(preflen))
4507 return compat_str(socket.inet_ntoa(
4508 compat_struct_pack('!L', random.randint(addr_min, addr_max))))
4511 class PerRequestProxyHandler(compat_urllib_request.ProxyHandler):
4512 def __init__(self, proxies=None):
4513 # Set default handlers
4514 for type in ('http', 'https'):
4515 setattr(self, '%s_open' % type,
4516 lambda r, proxy='__noproxy__', type=type, meth=self.proxy_open:
4517 meth(r, proxy, type))
4518 compat_urllib_request.ProxyHandler.__init__(self, proxies)
4520 def proxy_open(self, req, proxy, type):
4521 req_proxy = req.headers.get('Ytdl-request-proxy')
4522 if req_proxy is not None:
4524 del req.headers['Ytdl-request-proxy']
4526 if proxy == '__noproxy__':
4527 return None # No Proxy
4528 if compat_urlparse.urlparse(proxy).scheme.lower() in ('socks', 'socks4', 'socks4a', 'socks5'):
4529 req.add_header('Ytdl-socks-proxy', proxy)
4530 # yt-dlp's http/https handlers do wrapping the socket with socks
4532 return compat_urllib_request.ProxyHandler.proxy_open(
4533 self, req, proxy, type)
4536 # Both long_to_bytes and bytes_to_long are adapted from PyCrypto, which is
4537 # released into Public Domain
4538 # https://github.com/dlitz/pycrypto/blob/master/lib/Crypto/Util/number.py#L387
4540 def long_to_bytes(n, blocksize=0):
4541 """long_to_bytes(n:long, blocksize:int) : string
4542 Convert a long integer to a byte string.
4544 If optional blocksize is given and greater than zero, pad the front of the
4545 byte string with binary zeros so that the length is a multiple of
4548 # after much testing, this algorithm was deemed to be the fastest
4552 s = compat_struct_pack('>I', n & 0xffffffff) + s
4554 # strip off leading zeros
4555 for i in range(len(s)):
4556 if s[i] != b'\000'[0]:
4559 # only happens when n == 0
4563 # add back some pad bytes. this could be done more efficiently w.r.t. the
4564 # de-padding being done above, but sigh...
4565 if blocksize > 0 and len(s) % blocksize:
4566 s = (blocksize - len(s) % blocksize) * b'\000' + s
4570 def bytes_to_long(s):
4571 """bytes_to_long(string) : long
4572 Convert a byte string to a long integer.
4574 This is (essentially) the inverse of long_to_bytes().
4579 extra = (4 - length % 4)
4580 s = b'\000' * extra + s
4581 length = length + extra
4582 for i in range(0, length, 4):
4583 acc = (acc << 32) + compat_struct_unpack('>I', s[i:i + 4])[0]
4587 def ohdave_rsa_encrypt(data, exponent, modulus):
4589 Implement OHDave
's RSA algorithm. See http://www.ohdave.com/rsa/
4592 data: data to encrypt, bytes-like object
4593 exponent, modulus: parameter e and N of RSA algorithm, both integer
4594 Output: hex string of encrypted data
4596 Limitation: supports one block encryption only
4599 payload = int(binascii.hexlify(data[::-1]), 16)
4600 encrypted = pow(payload, exponent, modulus)
4601 return '%x' % encrypted
4604 def pkcs1pad(data, length):
4606 Padding input data with PKCS#1 scheme
4608 @param {int[]} data input data
4609 @param {int} length target length
4610 @returns {int[]} padded data
4612 if len(data) > length - 11:
4613 raise ValueError('Input data too
long for PKCS
#1 padding')
4615 pseudo_random
= [random
.randint(0, 254) for _
in range(length
- len(data
) - 3)]
4616 return [0, 2] + pseudo_random
+ [0] + data
4619 def encode_base_n(num
, n
, table
=None):
4620 FULL_TABLE
= '0123456789abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ'
4622 table
= FULL_TABLE
[:n
]
4625 raise ValueError('base %d exceeds table length %d' % (n
, len(table
)))
4632 ret
= table
[num
% n
] + ret
4637 def decode_packed_codes(code
):
4638 mobj
= re
.search(PACKED_CODES_RE
, code
)
4639 obfuscated_code
, base
, count
, symbols
= mobj
.groups()
4642 symbols
= symbols
.split('|')
4647 base_n_count
= encode_base_n(count
, base
)
4648 symbol_table
[base_n_count
] = symbols
[count
] or base_n_count
4651 r
'\b(\w+)\b', lambda mobj
: symbol_table
[mobj
.group(0)],
4655 def caesar(s
, alphabet
, shift
):
4660 alphabet
[(alphabet
.index(c
) + shift
) % l
] if c
in alphabet
else c
4665 return caesar(s
, r
'''!"#$%&'()*+,-./0123456789:;<=>?@ABCDEFGHIJKLMNOPQRSTUVWXYZ[\]^_`abcdefghijklmnopqrstuvwxyz{|}~''', 47)
4668 def parse_m3u8_attributes(attrib
):
4670 for (key
, val
) in re
.findall(r
'(?P<key>[A-Z0-9-]+)=(?P<val>"[^"]+"|[^",]+)(?:,|$)', attrib
):
4671 if val
.startswith('"'):
4677 def urshift(val
, n
):
4678 return val
>> n
if val
>= 0 else (val
+ 0x100000000) >> n
4681 # Based on png2str() written by @gdkchan and improved by @yokrysty
4682 # Originally posted at https://github.com/ytdl-org/youtube-dl/issues/9706
4683 def decode_png(png_data
):
4684 # Reference: https://www.w3.org/TR/PNG/
4685 header
= png_data
[8:]
4687 if png_data
[:8] != b
'\x89PNG\x0d\x0a\x1a\x0a' or header
[4:8] != b
'IHDR':
4688 raise IOError('Not a valid PNG file.')
4690 int_map
= {1: '>B', 2: '>H', 4: '>I'}
4691 unpack_integer
= lambda x
: compat_struct_unpack(int_map
[len(x
)], x
)[0]
4696 length
= unpack_integer(header
[:4])
4699 chunk_type
= header
[:4]
4702 chunk_data
= header
[:length
]
4703 header
= header
[length
:]
4705 header
= header
[4:] # Skip CRC
4713 ihdr
= chunks
[0]['data']
4715 width
= unpack_integer(ihdr
[:4])
4716 height
= unpack_integer(ihdr
[4:8])
4720 for chunk
in chunks
:
4721 if chunk
['type'] == b
'IDAT':
4722 idat
+= chunk
['data']
4725 raise IOError('Unable to read PNG data.')
4727 decompressed_data
= bytearray(zlib
.decompress(idat
))
4732 def _get_pixel(idx
):
4737 for y
in range(height
):
4738 basePos
= y
* (1 + stride
)
4739 filter_type
= decompressed_data
[basePos
]
4743 pixels
.append(current_row
)
4745 for x
in range(stride
):
4746 color
= decompressed_data
[1 + basePos
+ x
]
4747 basex
= y
* stride
+ x
4752 left
= _get_pixel(basex
- 3)
4754 up
= _get_pixel(basex
- stride
)
4756 if filter_type
== 1: # Sub
4757 color
= (color
+ left
) & 0xff
4758 elif filter_type
== 2: # Up
4759 color
= (color
+ up
) & 0xff
4760 elif filter_type
== 3: # Average
4761 color
= (color
+ ((left
+ up
) >> 1)) & 0xff
4762 elif filter_type
== 4: # Paeth
4768 c
= _get_pixel(basex
- stride
- 3)
4776 if pa
<= pb
and pa
<= pc
:
4777 color
= (color
+ a
) & 0xff
4779 color
= (color
+ b
) & 0xff
4781 color
= (color
+ c
) & 0xff
4783 current_row
.append(color
)
4785 return width
, height
, pixels
4788 def write_xattr(path
, key
, value
):
4789 # This mess below finds the best xattr tool for the job
4791 # try the pyxattr module...
4794 if hasattr(xattr
, 'set'): # pyxattr
4795 # Unicode arguments are not supported in python-pyxattr until
4797 # See https://github.com/ytdl-org/youtube-dl/issues/5498
4798 pyxattr_required_version
= '0.5.0'
4799 if version_tuple(xattr
.__version
__) < version_tuple(pyxattr_required_version
):
4800 # TODO: fallback to CLI tools
4801 raise XAttrUnavailableError(
4802 'python-pyxattr is detected but is too old. '
4803 'yt-dlp requires %s or above while your version is %s. '
4804 'Falling back to other xattr implementations' % (
4805 pyxattr_required_version
, xattr
.__version
__))
4807 setxattr
= xattr
.set
4809 setxattr
= xattr
.setxattr
4812 setxattr(path
, key
, value
)
4813 except EnvironmentError as e
:
4814 raise XAttrMetadataError(e
.errno
, e
.strerror
)
4817 if compat_os_name
== 'nt':
4818 # Write xattrs to NTFS Alternate Data Streams:
4819 # http://en.wikipedia.org/wiki/NTFS#Alternate_data_streams_.28ADS.29
4820 assert ':' not in key
4821 assert os
.path
.exists(path
)
4823 ads_fn
= path
+ ':' + key
4825 with open(ads_fn
, 'wb') as f
:
4827 except EnvironmentError as e
:
4828 raise XAttrMetadataError(e
.errno
, e
.strerror
)
4830 user_has_setfattr
= check_executable('setfattr', ['--version'])
4831 user_has_xattr
= check_executable('xattr', ['-h'])
4833 if user_has_setfattr
or user_has_xattr
:
4835 value
= value
.decode('utf-8')
4836 if user_has_setfattr
:
4837 executable
= 'setfattr'
4838 opts
= ['-n', key
, '-v', value
]
4839 elif user_has_xattr
:
4840 executable
= 'xattr'
4841 opts
= ['-w', key
, value
]
4843 cmd
= ([encodeFilename(executable
, True)]
4844 + [encodeArgument(o
) for o
in opts
]
4845 + [encodeFilename(path
, True)])
4849 cmd
, stdout
=subprocess
.PIPE
, stderr
=subprocess
.PIPE
, stdin
=subprocess
.PIPE
)
4850 except EnvironmentError as e
:
4851 raise XAttrMetadataError(e
.errno
, e
.strerror
)
4852 stdout
, stderr
= p
.communicate_or_kill()
4853 stderr
= stderr
.decode('utf-8', 'replace')
4854 if p
.returncode
!= 0:
4855 raise XAttrMetadataError(p
.returncode
, stderr
)
4858 # On Unix, and can't find pyxattr, setfattr, or xattr.
4859 if sys
.platform
.startswith('linux'):
4860 raise XAttrUnavailableError(
4861 "Couldn't find a tool to set the xattrs. "
4862 "Install either the python 'pyxattr' or 'xattr' "
4863 "modules, or the GNU 'attr' package "
4864 "(which contains the 'setfattr' tool).")
4866 raise XAttrUnavailableError(
4867 "Couldn't find a tool to set the xattrs. "
4868 "Install either the python 'xattr' module, "
4869 "or the 'xattr' binary.")
4872 def random_birthday(year_field
, month_field
, day_field
):
4873 start_date
= datetime
.date(1950, 1, 1)
4874 end_date
= datetime
.date(1995, 12, 31)
4875 offset
= random
.randint(0, (end_date
- start_date
).days
)
4876 random_date
= start_date
+ datetime
.timedelta(offset
)
4878 year_field
: str(random_date
.year
),
4879 month_field
: str(random_date
.month
),
4880 day_field
: str(random_date
.day
),
4884 # Templates for internet shortcut files, which are plain text files.
4885 DOT_URL_LINK_TEMPLATE
= '''
4890 DOT_WEBLOC_LINK_TEMPLATE
= '''
4891 <?xml version="1.0" encoding="UTF-8"?>
4892 <!DOCTYPE plist PUBLIC "-//Apple//DTD PLIST 1.0//EN" "http://www.apple.com/DTDs/PropertyList-1.0.dtd">
4893 <plist version="1.0">
4896 \t<string>%(url)s</string>
4901 DOT_DESKTOP_LINK_TEMPLATE
= '''
4911 'url': DOT_URL_LINK_TEMPLATE
,
4912 'desktop': DOT_DESKTOP_LINK_TEMPLATE
,
4913 'webloc': DOT_WEBLOC_LINK_TEMPLATE
,
4917 def iri_to_uri(iri
):
4919 Converts an IRI (Internationalized Resource Identifier, allowing Unicode characters) to a URI (Uniform Resource Identifier, ASCII-only).
4921 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.
4924 iri_parts
= compat_urllib_parse_urlparse(iri
)
4926 if '[' in iri_parts
.netloc
:
4927 raise ValueError('IPv6 URIs are not, yet, supported.')
4928 # Querying `.netloc`, when there's only one bracket, also raises a ValueError.
4930 # 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.
4933 if iri_parts
.username
:
4934 net_location
+= compat_urllib_parse_quote(iri_parts
.username
, safe
=r
"!$%&'()*+,~")
4935 if iri_parts
.password
is not None:
4936 net_location
+= ':' + compat_urllib_parse_quote(iri_parts
.password
, safe
=r
"!$%&'()*+,~")
4939 net_location
+= iri_parts
.hostname
.encode('idna').decode('utf-8') # Punycode for Unicode hostnames.
4940 # The 'idna' encoding produces ASCII text.
4941 if iri_parts
.port
is not None and iri_parts
.port
!= 80:
4942 net_location
+= ':' + str(iri_parts
.port
)
4944 return compat_urllib_parse_urlunparse(
4948 compat_urllib_parse_quote_plus(iri_parts
.path
, safe
=r
"!$%&'()*+,/:;=@|~"),
4950 # Unsure about the `safe` argument, since this is a legacy way of handling parameters.
4951 compat_urllib_parse_quote_plus(iri_parts
.params
, safe
=r
"!$%&'()*+,/:;=@|~"),
4953 # Not totally sure about the `safe` argument, since the source does not explicitly mention the query URI component.
4954 compat_urllib_parse_quote_plus(iri_parts
.query
, safe
=r
"!$%&'()*+,/:;=?@{|}~"),
4956 compat_urllib_parse_quote_plus(iri_parts
.fragment
, safe
=r
"!#$%&'()*+,/:;=?@{|}~")))
4958 # Source for `safe` arguments: https://url.spec.whatwg.org/#percent-encoded-bytes.
4961 def to_high_limit_path(path
):
4962 if sys
.platform
in ['win32', 'cygwin']:
4963 # Work around MAX_PATH limitation on Windows. The maximum allowed length for the individual path segments may still be quite limited.
4964 return r
'\\?\ '.rstrip() + os
.path
.abspath(path
)
4969 def format_field(obj
, field
=None, template
='%s', ignore
=(None, ''), default
='', func
=None):
4971 val
= obj
if obj
is not None else default
4973 val
= obj
.get(field
, default
)
4974 if func
and val
not in ignore
:
4976 return template
% val
if val
not in ignore
else default
4979 def clean_podcast_url(url
):
4980 return re
.sub(r
'''(?x)
4984 media\.blubrry\.com| # https://create.blubrry.com/resources/podcast-media-download-statistics/getting-started/
4987 (?:dts|www)\.podtrac\.com/(?:pts/)?redirect\.[0-9a-z]{3,4}| # http://analytics.podtrac.com/how-to-measure
4990 cn\.co| # https://podcorn.com/analytics-prefix/
4991 st\.fm # https://podsights.com/docs/
4996 _HEX_TABLE
= '0123456789abcdef'
4999 def random_uuidv4():
5000 return re
.sub(r
'[xy]', lambda x
: _HEX_TABLE
[random
.randint(0, 15)], 'xxxxxxxx-xxxx-4xxx-yxxx-xxxxxxxxxxxx')
5003 def make_dir(path
, to_screen
=None):
5005 dn
= os
.path
.dirname(path
)
5006 if dn
and not os
.path
.exists(dn
):
5009 except (OSError, IOError) as err
:
5010 if callable(to_screen
) is not None:
5011 to_screen('unable to create directory ' + error_to_compat_str(err
))
5015 def get_executable_path():
5016 from zipimport
import zipimporter
5017 if hasattr(sys
, 'frozen'): # Running from PyInstaller
5018 path
= os
.path
.dirname(sys
.executable
)
5019 elif isinstance(globals().get('__loader__'), zipimporter
): # Running from ZIP
5020 path
= os
.path
.join(os
.path
.dirname(__file__
), '../..')
5022 path
= os
.path
.join(os
.path
.dirname(__file__
), '..')
5023 return os
.path
.abspath(path
)
5026 def load_plugins(name
, suffix
, namespace
):
5029 plugins_spec
= importlib
.util
.spec_from_file_location(
5030 name
, os
.path
.join(get_executable_path(), 'ytdlp_plugins', name
, '__init__.py'))
5031 plugins
= importlib
.util
.module_from_spec(plugins_spec
)
5032 sys
.modules
[plugins_spec
.name
] = plugins
5033 plugins_spec
.loader
.exec_module(plugins
)
5034 for name
in dir(plugins
):
5035 if name
in namespace
:
5037 if not name
.endswith(suffix
):
5039 klass
= getattr(plugins
, name
)
5040 classes
[name
] = namespace
[name
] = klass
5041 except FileNotFoundError
:
5047 obj
, *path_list
, default
=None, expected_type
=None, get_all
=True,
5048 casesense
=True, is_user_input
=False, traverse_string
=False):
5049 ''' Traverse nested list/dict/tuple
5050 @param path_list A list of paths which are checked one by one.
5051 Each path is a list of keys where each key is a string,
5052 a function, a tuple of strings/None or "...".
5053 When a fuction is given, it takes the key as argument and
5054 returns whether the key matches or not. When a tuple is given,
5055 all the keys given in the tuple are traversed, and
5056 "..." traverses all the keys in the object
5057 "None" returns the object without traversal
5058 @param default Default value to return
5059 @param expected_type Only accept final value of this type (Can also be any callable)
5060 @param get_all Return all the values obtained from a path or only the first one
5061 @param casesense Whether to consider dictionary keys as case sensitive
5062 @param is_user_input Whether the keys are generated from user input. If True,
5063 strings are converted to int/slice if necessary
5064 @param traverse_string Whether to traverse inside strings. If True, any
5065 non-compatible object will also be converted into a string
5069 _lower
= lambda k
: (k
.lower() if isinstance(k
, str) else k
)
5070 path_list
= (map(_lower
, variadic(path
)) for path
in path_list
)
5072 def _traverse_obj(obj
, path
, _current_depth
=0):
5074 path
= tuple(variadic(path
))
5075 for i
, key
in enumerate(path
):
5076 if None in (key
, obj
):
5078 if isinstance(key
, (list, tuple)):
5079 obj
= [_traverse_obj(obj
, sub_key
, _current_depth
) for sub_key
in key
]
5082 obj
= (obj
.values() if isinstance(obj
, dict)
5083 else obj
if isinstance(obj
, (list, tuple, LazyList
))
5084 else str(obj
) if traverse_string
else [])
5086 depth
= max(depth
, _current_depth
)
5087 return [_traverse_obj(inner_obj
, path
[i
+ 1:], _current_depth
) for inner_obj
in obj
]
5089 if isinstance(obj
, (list, tuple, LazyList
)):
5090 obj
= enumerate(obj
)
5091 elif isinstance(obj
, dict):
5094 if not traverse_string
:
5098 depth
= max(depth
, _current_depth
)
5099 return [_traverse_obj(v
, path
[i
+ 1:], _current_depth
) for k
, v
in obj
if key(k
)]
5100 elif isinstance(obj
, dict) and not (is_user_input
and key
== ':'):
5101 obj
= (obj
.get(key
) if casesense
or (key
in obj
)
5102 else next((v
for k
, v
in obj
.items() if _lower(k
) == key
), None))
5105 key
= (int_or_none(key
) if ':' not in key
5106 else slice(*map(int_or_none
, key
.split(':'))))
5107 if key
== slice(None):
5108 return _traverse_obj(obj
, (..., *path
[i
+ 1:]), _current_depth
)
5109 if not isinstance(key
, (int, slice)):
5111 if not isinstance(obj
, (list, tuple, LazyList
)):
5112 if not traverse_string
:
5121 if isinstance(expected_type
, type):
5122 type_test
= lambda val
: val
if isinstance(val
, expected_type
) else None
5123 elif expected_type
is not None:
5124 type_test
= expected_type
5126 type_test
= lambda val
: val
5128 for path
in path_list
:
5130 val
= _traverse_obj(obj
, path
)
5133 for _
in range(depth
- 1):
5134 val
= itertools
.chain
.from_iterable(v
for v
in val
if v
is not None)
5135 val
= [v
for v
in map(type_test
, val
) if v
is not None]
5137 return val
if get_all
else val
[0]
5139 val
= type_test(val
)
5145 def traverse_dict(dictn
, keys
, casesense
=True):
5146 write_string('DeprecationWarning: yt_dlp.utils.traverse_dict is deprecated '
5147 'and may be removed in a future version. Use yt_dlp.utils.traverse_obj instead')
5148 return traverse_obj(dictn
, keys
, casesense
=casesense
, is_user_input
=True, traverse_string
=True)
5151 def variadic(x
, allowed_types
=(str, bytes, dict)):
5152 return x
if isinstance(x
, collections
.abc
.Iterable
) and not isinstance(x
, allowed_types
) else (x
,)
5155 # create a JSON Web Signature (jws) with HS256 algorithm
5156 # the resulting format is in JWS Compact Serialization
5157 # implemented following JWT https://www.rfc-editor.org/rfc/rfc7519.html
5158 # implemented following JWS https://www.rfc-editor.org/rfc/rfc7515.html
5159 def jwt_encode_hs256(payload_data
, key
, headers
={}):
5165 header_data
.update(headers
)
5166 header_b64
= base64
.b64encode(json
.dumps(header_data
).encode('utf-8'))
5167 payload_b64
= base64
.b64encode(json
.dumps(payload_data
).encode('utf-8'))
5168 h
= hmac
.new(key
.encode('utf-8'), header_b64
+ b
'.' + payload_b64
, hashlib
.sha256
)
5169 signature_b64
= base64
.b64encode(h
.digest())
5170 token
= header_b64
+ b
'.' + payload_b64
+ b
'.' + signature_b64
5174 # can be extended in future to verify the signature and parse header and return the algorithm used if it's not HS256
5175 def jwt_decode_hs256(jwt
):
5176 header_b64
, payload_b64
, signature_b64
= jwt
.split('.')
5177 payload_data
= json
.loads(base64
.urlsafe_b64decode(payload_b64
))
5181 def supports_terminal_sequences(stream
):
5182 if compat_os_name
== 'nt':
5183 from .compat
import WINDOWS_VT_MODE
# Must be imported locally
5184 if not WINDOWS_VT_MODE
or get_windows_version() < (10, 0, 10586):
5186 elif not os
.getenv('TERM'):
5189 return stream
.isatty()
5190 except BaseException
:
5194 _terminal_sequences_re
= re
.compile('\033\\[[^m]+m')
5197 def remove_terminal_sequences(string
):
5198 return _terminal_sequences_re
.sub('', string
)
5201 def number_of_digits(number
):
5202 return len('%d' % number
)
5205 def join_nonempty(*values
, delim
='-', from_dict
=None):
5206 if from_dict
is not None:
5207 values
= map(from_dict
.get
, values
)
5208 return delim
.join(map(str, filter(None, values
)))
5214 __initialized
= False
5216 def __init__(self
, parser
, label
=None):
5217 self
._parser
, self
.label
= parser
, label
5218 self
._loaded
_paths
, self
.configs
= set(), []
5220 def init(self
, args
=None, filename
=None):
5221 assert not self
.__initialized
5223 location
= os
.path
.realpath(filename
)
5224 if location
in self
._loaded
_paths
:
5226 self
._loaded
_paths
.add(location
)
5228 self
.__initialized
= True
5229 self
.own_args
, self
.filename
= args
, filename
5230 for location
in self
._parser
.parse_args(args
)[0].config_locations
or []:
5231 location
= compat_expanduser(location
)
5232 if os
.path
.isdir(location
):
5233 location
= os
.path
.join(location
, 'yt-dlp.conf')
5234 if not os
.path
.exists(location
):
5235 self
._parser
.error(f
'config location {location} does not exist')
5236 self
.append_config(self
.read_file(location
), location
)
5240 label
= join_nonempty(
5241 self
.label
, 'config', f
'"{self.filename}"' if self
.filename
else '',
5243 return join_nonempty(
5244 self
.own_args
is not None and f
'{label[0].upper()}{label[1:]}: {self.hide_login_info(self.own_args)}',
5245 *(f
'\n{c}'.replace('\n', '\n| ')[1:] for c
in self
.configs
),
5249 def read_file(filename
, default
=[]):
5251 optionf
= open(filename
)
5253 return default
# silently skip if file is not present
5255 # FIXME: https://github.com/ytdl-org/youtube-dl/commit/dfe5fa49aed02cf36ba9f743b11b0903554b5e56
5256 contents
= optionf
.read()
5257 if sys
.version_info
< (3,):
5258 contents
= contents
.decode(preferredencoding())
5259 res
= compat_shlex_split(contents
, comments
=True)
5265 def hide_login_info(opts
):
5266 PRIVATE_OPTS
= set(['-p', '--password', '-u', '--username', '--video-password', '--ap-password', '--ap-username'])
5267 eqre
= re
.compile('^(?P<key>' + ('|'.join(re
.escape(po
) for po
in PRIVATE_OPTS
)) + ')=.+$')
5272 return m
.group('key') + '=PRIVATE'
5276 opts
= list(map(_scrub_eq
, opts
))
5277 for idx
, opt
in enumerate(opts
):
5278 if opt
in PRIVATE_OPTS
and idx
+ 1 < len(opts
):
5279 opts
[idx
+ 1] = 'PRIVATE'
5282 def append_config(self
, *args
, label
=None):
5283 config
= type(self
)(self
._parser
, label
)
5284 config
._loaded
_paths
= self
._loaded
_paths
5285 if config
.init(*args
):
5286 self
.configs
.append(config
)
5290 for config
in reversed(self
.configs
):
5291 yield from config
.all_args
5292 yield from self
.own_args
or []
5294 def parse_args(self
):
5295 return self
._parser
.parse_args(list(self
.all_args
))