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',
148 'Sec-Fetch-Mode': 'navigate',
153 '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',
157 NO_DEFAULT
= object()
159 ENGLISH_MONTH_NAMES
= [
160 'January', 'February', 'March', 'April', 'May', 'June',
161 'July', 'August', 'September', 'October', 'November', 'December']
164 'en': ENGLISH_MONTH_NAMES
,
166 'janvier', 'février', 'mars', 'avril', 'mai', 'juin',
167 'juillet', 'août', 'septembre', 'octobre', 'novembre', 'décembre'],
171 'mp4', 'm4a', 'm4p', 'm4b', 'm4r', 'm4v', 'aac',
172 'flv', 'f4v', 'f4a', 'f4b',
173 'webm', 'ogg', 'ogv', 'oga', 'ogx', 'spx', 'opus',
174 'mkv', 'mka', 'mk3d',
183 'f4f', 'f4m', 'm3u8', 'smil')
185 # needed for sanitizing filenames in restricted mode
186 ACCENT_CHARS
= dict(zip('ÂÃÄÀÁÅÆÇÈÉÊËÌÍÎÏÐÑÒÓÔÕÖŐØŒÙÚÛÜŰÝÞßàáâãäåæçèéêëìíîïðñòóôõöőøœùúûüűýþÿ',
187 itertools
.chain('AAAAAA', ['AE'], 'CEEEEIIIIDNOOOOOOO', ['OE'], 'UUUUUY', ['TH', 'ss'],
188 'aaaaaa', ['ae'], 'ceeeeiiiionooooooo', ['oe'], 'uuuuuy', ['th'], 'y')))
218 '%Y-%m-%d %H:%M:%S.%f',
219 '%Y-%m-%d %H:%M:%S:%f',
222 '%Y-%m-%dT%H:%M:%SZ',
223 '%Y-%m-%dT%H:%M:%S.%fZ',
224 '%Y-%m-%dT%H:%M:%S.%f0Z',
226 '%Y-%m-%dT%H:%M:%S.%f',
229 '%b %d %Y at %H:%M:%S',
231 '%B %d %Y at %H:%M:%S',
235 DATE_FORMATS_DAY_FIRST
= list(DATE_FORMATS
)
236 DATE_FORMATS_DAY_FIRST
.extend([
245 DATE_FORMATS_MONTH_FIRST
= list(DATE_FORMATS
)
246 DATE_FORMATS_MONTH_FIRST
.extend([
254 PACKED_CODES_RE
= r
"}\('(.+)',(\d+),(\d+),'([^']+)'\.split\('\|'\)"
255 JSON_LD_RE
= r
'(?is)<script[^>]+type=(["\']?
)application
/ld\
+json\
1[^
>]*>(?P
<json_ld
>.+?
)</script
>'
258 def preferredencoding():
259 """Get preferred encoding.
261 Returns the best encoding scheme for the system, based on
262 locale.getpreferredencoding() and some further tweaks.
265 pref = locale.getpreferredencoding()
273 def write_json_file(obj, fn):
274 """ Encode obj as JSON and write it to fn, atomically if possible """
276 fn = encodeFilename(fn)
277 if sys.version_info < (3, 0) and sys.platform != 'win32
':
278 encoding = get_filesystem_encoding()
279 # os.path.basename returns a bytes object, but NamedTemporaryFile
280 # will fail if the filename contains non ascii characters unless we
281 # use a unicode object
282 path_basename = lambda f: os.path.basename(fn).decode(encoding)
283 # the same for os.path.dirname
284 path_dirname = lambda f: os.path.dirname(fn).decode(encoding)
286 path_basename = os.path.basename
287 path_dirname = os.path.dirname
291 'prefix
': path_basename(fn) + '.',
292 'dir': path_dirname(fn),
296 # In Python 2.x, json.dump expects a bytestream.
297 # In Python 3.x, it writes to a character stream
298 if sys.version_info < (3, 0):
306 tf = tempfile.NamedTemporaryFile(**compat_kwargs(args))
310 json.dump(obj, tf, ensure_ascii=False)
311 if sys.platform == 'win32
':
312 # Need to remove existing file on Windows, else os.rename raises
313 # WindowsError or FileExistsError.
321 os.chmod(tf.name, 0o666 & ~mask)
324 os.rename(tf.name, fn)
333 if sys.version_info >= (2, 7):
334 def find_xpath_attr(node, xpath, key, val=None):
335 """ Find the xpath xpath[@key=val] """
336 assert re.match(r'^
[a
-zA
-Z_
-]+$
', key)
337 expr = xpath + ('[@%s]' % key if val is None else "[@%s='%s']" % (key, val))
338 return node.find(expr)
340 def find_xpath_attr(node, xpath, key, val=None):
341 for f in node.findall(compat_xpath(xpath)):
342 if key not in f.attrib:
344 if val is None or f.attrib.get(key) == val:
348 # On python2.6 the xml.etree.ElementTree.Element methods don't support
349 # the namespace parameter
352 def xpath_with_ns(path
, ns_map
):
353 components
= [c
.split(':') for c
in path
.split('/')]
357 replaced
.append(c
[0])
360 replaced
.append('{%s}%s' % (ns_map
[ns
], tag
))
361 return '/'.join(replaced
)
364 def xpath_element(node
, xpath
, name
=None, fatal
=False, default
=NO_DEFAULT
):
365 def _find_xpath(xpath
):
366 return node
.find(compat_xpath(xpath
))
368 if isinstance(xpath
, (str, compat_str
)):
369 n
= _find_xpath(xpath
)
377 if default
is not NO_DEFAULT
:
380 name
= xpath
if name
is None else name
381 raise ExtractorError('Could not find XML element %s' % name
)
387 def xpath_text(node
, xpath
, name
=None, fatal
=False, default
=NO_DEFAULT
):
388 n
= xpath_element(node
, xpath
, name
, fatal
=fatal
, default
=default
)
389 if n
is None or n
== default
:
392 if default
is not NO_DEFAULT
:
395 name
= xpath
if name
is None else name
396 raise ExtractorError('Could not find XML element\'s text %s' % name
)
402 def xpath_attr(node
, xpath
, key
, name
=None, fatal
=False, default
=NO_DEFAULT
):
403 n
= find_xpath_attr(node
, xpath
, key
)
405 if default
is not NO_DEFAULT
:
408 name
= '%s[@%s]' % (xpath
, key
) if name
is None else name
409 raise ExtractorError('Could not find XML attribute %s' % name
)
415 def get_element_by_id(id, html
):
416 """Return the content of the tag with the specified ID in the passed HTML document"""
417 return get_element_by_attribute('id', id, html
)
420 def get_element_html_by_id(id, html
):
421 """Return the html of the tag with the specified ID in the passed HTML document"""
422 return get_element_html_by_attribute('id', id, html
)
425 def get_element_by_class(class_name
, html
):
426 """Return the content of the first tag with the specified class in the passed HTML document"""
427 retval
= get_elements_by_class(class_name
, html
)
428 return retval
[0] if retval
else None
431 def get_element_html_by_class(class_name
, html
):
432 """Return the html of the first tag with the specified class in the passed HTML document"""
433 retval
= get_elements_html_by_class(class_name
, html
)
434 return retval
[0] if retval
else None
437 def get_element_by_attribute(attribute
, value
, html
, escape_value
=True):
438 retval
= get_elements_by_attribute(attribute
, value
, html
, escape_value
)
439 return retval
[0] if retval
else None
442 def get_element_html_by_attribute(attribute
, value
, html
, escape_value
=True):
443 retval
= get_elements_html_by_attribute(attribute
, value
, html
, escape_value
)
444 return retval
[0] if retval
else None
447 def get_elements_by_class(class_name
, html
):
448 """Return the content of all tags with the specified class in the passed HTML document as a list"""
449 return get_elements_by_attribute(
450 'class', r
'[^\'"]*\b%s\b[^\'"]*' % re.escape(class_name),
451 html, escape_value=False)
454 def get_elements_html_by_class(class_name, html):
455 """Return the html of all tags with the specified class in the passed HTML document as a list"""
456 return get_elements_html_by_attribute(
457 'class', r'[^
\'"]*\b%s\b[^\'"]*' % re.escape(class_name),
458 html, escape_value=False)
461 def get_elements_by_attribute(*args, **kwargs):
462 """Return the content of the tag with the specified attribute in the passed HTML document"""
463 return [content for content, _ in get_elements_text_and_html_by_attribute(*args, **kwargs)]
466 def get_elements_html_by_attribute(*args, **kwargs):
467 """Return the html of the tag with the specified attribute in the passed HTML document"""
468 return [whole for _, whole in get_elements_text_and_html_by_attribute(*args, **kwargs)]
471 def get_elements_text_and_html_by_attribute(attribute, value, html, escape_value=True):
473 Return the text (content) and the html (whole) of the tag with the specified
474 attribute in the passed HTML document
477 value_quote_optional = '' if re.match(r'''[\s"'`
=<>]''', value) else '?'
479 value = re.escape(value) if escape_value else value
481 partial_element_re = r'''(?x
)
482 <(?P
<tag
>[a
-zA
-Z0
-9:._-]+)
483 (?
:\
s(?
:[^
>"']|"[^
"]*"|
'[^']*')*)?
484 \s%(attribute)s\s*=\s*(?P<_q>['"]%(vqo)s)(?-x:%(value)s)(?P=_q)
485 ''' % {'attribute': re.escape(attribute), 'value': value, 'vqo': value_quote_optional}
487 for m in re.finditer(partial_element_re, html):
488 content, whole = get_element_text_and_html_by_tag(m.group('tag'), html[m.start():])
491 unescapeHTML(re.sub(r'^(?P<q>["\'])(?P
<content
>.*)(?P
=q
)$
', r'\g
<content
>', content, flags=re.DOTALL)),
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 if params
.get('legacyserverconnect'):
1001 context
.options |
= 4 # SSL_OP_LEGACY_SERVER_CONNECT
1002 context
.verify_mode
= ssl
.CERT_REQUIRED
if opts_check_certificate
else ssl
.CERT_NONE
1003 if opts_check_certificate
:
1005 context
.load_default_certs()
1006 # Work around the issue in load_default_certs when there are bad certificates. See:
1007 # https://github.com/yt-dlp/yt-dlp/issues/1060,
1008 # https://bugs.python.org/issue35665, https://bugs.python.org/issue45312
1009 except ssl
.SSLError
:
1010 # enum_certificates is not present in mingw python. See https://github.com/yt-dlp/yt-dlp/issues/1151
1011 if sys
.platform
== 'win32' and hasattr(ssl
, 'enum_certificates'):
1012 # Create a new context to discard any certificates that were already loaded
1013 context
= ssl
.SSLContext(ssl
.PROTOCOL_TLS_CLIENT
)
1014 context
.check_hostname
, context
.verify_mode
= True, ssl
.CERT_REQUIRED
1015 for storename
in ('CA', 'ROOT'):
1016 _ssl_load_windows_store_certs(context
, storename
)
1017 context
.set_default_verify_paths()
1018 return YoutubeDLHTTPSHandler(params
, context
=context
, **kwargs
)
1021 def bug_reports_message(before
=';'):
1022 if ytdl_is_updateable():
1023 update_cmd
= 'type yt-dlp -U to update'
1025 update_cmd
= 'see https://github.com/yt-dlp/yt-dlp on how to update'
1026 msg
= 'please report this issue on https://github.com/yt-dlp/yt-dlp .'
1027 msg
+= ' Make sure you are using the latest version; %s.' % update_cmd
1028 msg
+= ' Be sure to call yt-dlp with the --verbose flag and include its complete output.'
1030 before
= before
.rstrip()
1031 if not before
or before
.endswith(('.', '!', '?')):
1032 msg
= msg
[0].title() + msg
[1:]
1034 return (before
+ ' ' if before
else '') + msg
1037 class YoutubeDLError(Exception):
1038 """Base exception for YoutubeDL errors."""
1041 def __init__(self
, msg
=None):
1044 elif self
.msg
is None:
1045 self
.msg
= type(self
).__name
__
1046 super().__init
__(self
.msg
)
1049 network_exceptions
= [compat_urllib_error
.URLError
, compat_http_client
.HTTPException
, socket
.error
]
1050 if hasattr(ssl
, 'CertificateError'):
1051 network_exceptions
.append(ssl
.CertificateError
)
1052 network_exceptions
= tuple(network_exceptions
)
1055 class ExtractorError(YoutubeDLError
):
1056 """Error during info extraction."""
1058 def __init__(self
, msg
, tb
=None, expected
=False, cause
=None, video_id
=None, ie
=None):
1059 """ tb, if given, is the original traceback (so that it can be printed out).
1060 If expected is set, this is a normal error message and most likely not a bug in yt-dlp.
1062 if sys
.exc_info()[0] in network_exceptions
:
1067 self
.expected
= expected
1069 self
.video_id
= video_id
1071 self
.exc_info
= sys
.exc_info() # preserve original exception
1073 super(ExtractorError
, self
).__init
__(''.join((
1074 format_field(ie
, template
='[%s] '),
1075 format_field(video_id
, template
='%s: '),
1077 format_field(cause
, template
=' (caused by %r)'),
1078 '' if expected
else bug_reports_message())))
1080 def format_traceback(self
):
1081 if self
.traceback
is None:
1083 return ''.join(traceback
.format_tb(self
.traceback
))
1086 class UnsupportedError(ExtractorError
):
1087 def __init__(self
, url
):
1088 super(UnsupportedError
, self
).__init
__(
1089 'Unsupported URL: %s' % url
, expected
=True)
1093 class RegexNotFoundError(ExtractorError
):
1094 """Error when a regex didn't match"""
1098 class GeoRestrictedError(ExtractorError
):
1099 """Geographic restriction Error exception.
1101 This exception may be thrown when a video is not available from your
1102 geographic location due to geographic restrictions imposed by a website.
1105 def __init__(self
, msg
, countries
=None, **kwargs
):
1106 kwargs
['expected'] = True
1107 super(GeoRestrictedError
, self
).__init
__(msg
, **kwargs
)
1108 self
.countries
= countries
1111 class DownloadError(YoutubeDLError
):
1112 """Download Error exception.
1114 This exception may be thrown by FileDownloader objects if they are not
1115 configured to continue on errors. They will contain the appropriate
1119 def __init__(self
, msg
, exc_info
=None):
1120 """ exc_info, if given, is the original exception that caused the trouble (as returned by sys.exc_info()). """
1121 super(DownloadError
, self
).__init
__(msg
)
1122 self
.exc_info
= exc_info
1125 class EntryNotInPlaylist(YoutubeDLError
):
1126 """Entry not in playlist exception.
1128 This exception will be thrown by YoutubeDL when a requested entry
1129 is not found in the playlist info_dict
1131 msg
= 'Entry not found in info'
1134 class SameFileError(YoutubeDLError
):
1135 """Same File exception.
1137 This exception will be thrown by FileDownloader objects if they detect
1138 multiple files would have to be downloaded to the same file on disk.
1140 msg
= 'Fixed output name but more than one file to download'
1142 def __init__(self
, filename
=None):
1143 if filename
is not None:
1144 self
.msg
+= f
': {filename}'
1145 super().__init
__(self
.msg
)
1148 class PostProcessingError(YoutubeDLError
):
1149 """Post Processing exception.
1151 This exception may be raised by PostProcessor's .run() method to
1152 indicate an error in the postprocessing task.
1156 class DownloadCancelled(YoutubeDLError
):
1157 """ Exception raised when the download queue should be interrupted """
1158 msg
= 'The download was cancelled'
1161 class ExistingVideoReached(DownloadCancelled
):
1162 """ --break-on-existing triggered """
1163 msg
= 'Encountered a video that is already in the archive, stopping due to --break-on-existing'
1166 class RejectedVideoReached(DownloadCancelled
):
1167 """ --break-on-reject triggered """
1168 msg
= 'Encountered a video that did not match filter, stopping due to --break-on-reject'
1171 class MaxDownloadsReached(DownloadCancelled
):
1172 """ --max-downloads limit has been reached. """
1173 msg
= 'Maximum number of downloads reached, stopping due to --max-downloads'
1176 class ReExtractInfo(YoutubeDLError
):
1177 """ Video info needs to be re-extracted. """
1179 def __init__(self
, msg
, expected
=False):
1180 super().__init
__(msg
)
1181 self
.expected
= expected
1184 class ThrottledDownload(ReExtractInfo
):
1185 """ Download speed below --throttled-rate. """
1186 msg
= 'The download speed is below throttle limit'
1189 super().__init
__(self
.msg
, expected
=False)
1192 class UnavailableVideoError(YoutubeDLError
):
1193 """Unavailable Format exception.
1195 This exception will be thrown when a video is requested
1196 in a format that is not available for that video.
1198 msg
= 'Unable to download video'
1200 def __init__(self
, err
=None):
1202 self
.msg
+= f
': {err}'
1203 super().__init
__(self
.msg
)
1206 class ContentTooShortError(YoutubeDLError
):
1207 """Content Too Short exception.
1209 This exception may be raised by FileDownloader objects when a file they
1210 download is too small for what the server announced first, indicating
1211 the connection was probably interrupted.
1214 def __init__(self
, downloaded
, expected
):
1215 super(ContentTooShortError
, self
).__init
__(
1216 'Downloaded {0} bytes, expected {1} bytes'.format(downloaded
, expected
)
1219 self
.downloaded
= downloaded
1220 self
.expected
= expected
1223 class XAttrMetadataError(YoutubeDLError
):
1224 def __init__(self
, code
=None, msg
='Unknown error'):
1225 super(XAttrMetadataError
, self
).__init
__(msg
)
1229 # Parsing code and msg
1230 if (self
.code
in (errno
.ENOSPC
, errno
.EDQUOT
)
1231 or 'No space left' in self
.msg
or 'Disk quota exceeded' in self
.msg
):
1232 self
.reason
= 'NO_SPACE'
1233 elif self
.code
== errno
.E2BIG
or 'Argument list too long' in self
.msg
:
1234 self
.reason
= 'VALUE_TOO_LONG'
1236 self
.reason
= 'NOT_SUPPORTED'
1239 class XAttrUnavailableError(YoutubeDLError
):
1243 def _create_http_connection(ydl_handler
, http_class
, is_https
, *args
, **kwargs
):
1244 # Working around python 2 bug (see http://bugs.python.org/issue17849) by limiting
1245 # expected HTTP responses to meet HTTP/1.0 or later (see also
1246 # https://github.com/ytdl-org/youtube-dl/issues/6727)
1247 if sys
.version_info
< (3, 0):
1248 kwargs
['strict'] = True
1249 hc
= http_class(*args
, **compat_kwargs(kwargs
))
1250 source_address
= ydl_handler
._params
.get('source_address')
1252 if source_address
is not None:
1253 # This is to workaround _create_connection() from socket where it will try all
1254 # address data from getaddrinfo() including IPv6. This filters the result from
1255 # getaddrinfo() based on the source_address value.
1256 # This is based on the cpython socket.create_connection() function.
1257 # https://github.com/python/cpython/blob/master/Lib/socket.py#L691
1258 def _create_connection(address
, timeout
=socket
._GLOBAL
_DEFAULT
_TIMEOUT
, source_address
=None):
1259 host
, port
= address
1261 addrs
= socket
.getaddrinfo(host
, port
, 0, socket
.SOCK_STREAM
)
1262 af
= socket
.AF_INET
if '.' in source_address
[0] else socket
.AF_INET6
1263 ip_addrs
= [addr
for addr
in addrs
if addr
[0] == af
]
1264 if addrs
and not ip_addrs
:
1265 ip_version
= 'v4' if af
== socket
.AF_INET
else 'v6'
1267 "No remote IP%s addresses available for connect, can't use '%s' as source address"
1268 % (ip_version
, source_address
[0]))
1269 for res
in ip_addrs
:
1270 af
, socktype
, proto
, canonname
, sa
= res
1273 sock
= socket
.socket(af
, socktype
, proto
)
1274 if timeout
is not socket
._GLOBAL
_DEFAULT
_TIMEOUT
:
1275 sock
.settimeout(timeout
)
1276 sock
.bind(source_address
)
1278 err
= None # Explicitly break reference cycle
1280 except socket
.error
as _
:
1282 if sock
is not None:
1287 raise socket
.error('getaddrinfo returns an empty list')
1288 if hasattr(hc
, '_create_connection'):
1289 hc
._create
_connection
= _create_connection
1290 sa
= (source_address
, 0)
1291 if hasattr(hc
, 'source_address'): # Python 2.7+
1292 hc
.source_address
= sa
1294 def _hc_connect(self
, *args
, **kwargs
):
1295 sock
= _create_connection(
1296 (self
.host
, self
.port
), self
.timeout
, sa
)
1298 self
.sock
= ssl
.wrap_socket(
1299 sock
, self
.key_file
, self
.cert_file
,
1300 ssl_version
=ssl
.PROTOCOL_TLSv1
)
1303 hc
.connect
= functools
.partial(_hc_connect
, hc
)
1308 def handle_youtubedl_headers(headers
):
1309 filtered_headers
= headers
1311 if 'Youtubedl-no-compression' in filtered_headers
:
1312 filtered_headers
= dict((k
, v
) for k
, v
in filtered_headers
.items() if k
.lower() != 'accept-encoding')
1313 del filtered_headers
['Youtubedl-no-compression']
1315 return filtered_headers
1318 class YoutubeDLHandler(compat_urllib_request
.HTTPHandler
):
1319 """Handler for HTTP requests and responses.
1321 This class, when installed with an OpenerDirector, automatically adds
1322 the standard headers to every HTTP request and handles gzipped and
1323 deflated responses from web servers. If compression is to be avoided in
1324 a particular request, the original request in the program code only has
1325 to include the HTTP header "Youtubedl-no-compression", which will be
1326 removed before making the real request.
1328 Part of this code was copied from:
1330 http://techknack.net/python-urllib2-handlers/
1332 Andrew Rowls, the author of that code, agreed to release it to the
1336 def __init__(self
, params
, *args
, **kwargs
):
1337 compat_urllib_request
.HTTPHandler
.__init
__(self
, *args
, **kwargs
)
1338 self
._params
= params
1340 def http_open(self
, req
):
1341 conn_class
= compat_http_client
.HTTPConnection
1343 socks_proxy
= req
.headers
.get('Ytdl-socks-proxy')
1345 conn_class
= make_socks_conn_class(conn_class
, socks_proxy
)
1346 del req
.headers
['Ytdl-socks-proxy']
1348 return self
.do_open(functools
.partial(
1349 _create_http_connection
, self
, conn_class
, False),
1357 return zlib
.decompress(data
, -zlib
.MAX_WBITS
)
1359 return zlib
.decompress(data
)
1361 def http_request(self
, req
):
1362 # According to RFC 3986, URLs can not contain non-ASCII characters, however this is not
1363 # always respected by websites, some tend to give out URLs with non percent-encoded
1364 # non-ASCII characters (see telemb.py, ard.py [#3412])
1365 # urllib chokes on URLs with non-ASCII characters (see http://bugs.python.org/issue3991)
1366 # To work around aforementioned issue we will replace request's original URL with
1367 # percent-encoded one
1368 # Since redirects are also affected (e.g. http://www.southpark.de/alle-episoden/s18e09)
1369 # the code of this workaround has been moved here from YoutubeDL.urlopen()
1370 url
= req
.get_full_url()
1371 url_escaped
= escape_url(url
)
1373 # Substitute URL if any change after escaping
1374 if url
!= url_escaped
:
1375 req
= update_Request(req
, url
=url_escaped
)
1377 for h
, v
in std_headers
.items():
1378 # Capitalize is needed because of Python bug 2275: http://bugs.python.org/issue2275
1379 # The dict keys are capitalized because of this bug by urllib
1380 if h
.capitalize() not in req
.headers
:
1381 req
.add_header(h
, v
)
1383 req
.headers
= handle_youtubedl_headers(req
.headers
)
1385 if sys
.version_info
< (2, 7) and '#' in req
.get_full_url():
1386 # Python 2.6 is brain-dead when it comes to fragments
1387 req
._Request
__original
= req
._Request
__original
.partition('#')[0]
1388 req
._Request
__r
_type
= req
._Request
__r
_type
.partition('#')[0]
1392 def http_response(self
, req
, resp
):
1395 if resp
.headers
.get('Content-encoding', '') == 'gzip':
1396 content
= resp
.read()
1397 gz
= gzip
.GzipFile(fileobj
=io
.BytesIO(content
), mode
='rb')
1399 uncompressed
= io
.BytesIO(gz
.read())
1400 except IOError as original_ioerror
:
1401 # There may be junk add the end of the file
1402 # See http://stackoverflow.com/q/4928560/35070 for details
1403 for i
in range(1, 1024):
1405 gz
= gzip
.GzipFile(fileobj
=io
.BytesIO(content
[:-i
]), mode
='rb')
1406 uncompressed
= io
.BytesIO(gz
.read())
1411 raise original_ioerror
1412 resp
= compat_urllib_request
.addinfourl(uncompressed
, old_resp
.headers
, old_resp
.url
, old_resp
.code
)
1413 resp
.msg
= old_resp
.msg
1414 del resp
.headers
['Content-encoding']
1416 if resp
.headers
.get('Content-encoding', '') == 'deflate':
1417 gz
= io
.BytesIO(self
.deflate(resp
.read()))
1418 resp
= compat_urllib_request
.addinfourl(gz
, old_resp
.headers
, old_resp
.url
, old_resp
.code
)
1419 resp
.msg
= old_resp
.msg
1420 del resp
.headers
['Content-encoding']
1421 # Percent-encode redirect URL of Location HTTP header to satisfy RFC 3986 (see
1422 # https://github.com/ytdl-org/youtube-dl/issues/6457).
1423 if 300 <= resp
.code
< 400:
1424 location
= resp
.headers
.get('Location')
1426 # As of RFC 2616 default charset is iso-8859-1 that is respected by python 3
1427 if sys
.version_info
>= (3, 0):
1428 location
= location
.encode('iso-8859-1').decode('utf-8')
1430 location
= location
.decode('utf-8')
1431 location_escaped
= escape_url(location
)
1432 if location
!= location_escaped
:
1433 del resp
.headers
['Location']
1434 if sys
.version_info
< (3, 0):
1435 location_escaped
= location_escaped
.encode('utf-8')
1436 resp
.headers
['Location'] = location_escaped
1439 https_request
= http_request
1440 https_response
= http_response
1443 def make_socks_conn_class(base_class
, socks_proxy
):
1444 assert issubclass(base_class
, (
1445 compat_http_client
.HTTPConnection
, compat_http_client
.HTTPSConnection
))
1447 url_components
= compat_urlparse
.urlparse(socks_proxy
)
1448 if url_components
.scheme
.lower() == 'socks5':
1449 socks_type
= ProxyType
.SOCKS5
1450 elif url_components
.scheme
.lower() in ('socks', 'socks4'):
1451 socks_type
= ProxyType
.SOCKS4
1452 elif url_components
.scheme
.lower() == 'socks4a':
1453 socks_type
= ProxyType
.SOCKS4A
1455 def unquote_if_non_empty(s
):
1458 return compat_urllib_parse_unquote_plus(s
)
1462 url_components
.hostname
, url_components
.port
or 1080,
1464 unquote_if_non_empty(url_components
.username
),
1465 unquote_if_non_empty(url_components
.password
),
1468 class SocksConnection(base_class
):
1470 self
.sock
= sockssocket()
1471 self
.sock
.setproxy(*proxy_args
)
1472 if type(self
.timeout
) in (int, float):
1473 self
.sock
.settimeout(self
.timeout
)
1474 self
.sock
.connect((self
.host
, self
.port
))
1476 if isinstance(self
, compat_http_client
.HTTPSConnection
):
1477 if hasattr(self
, '_context'): # Python > 2.6
1478 self
.sock
= self
._context
.wrap_socket(
1479 self
.sock
, server_hostname
=self
.host
)
1481 self
.sock
= ssl
.wrap_socket(self
.sock
)
1483 return SocksConnection
1486 class YoutubeDLHTTPSHandler(compat_urllib_request
.HTTPSHandler
):
1487 def __init__(self
, params
, https_conn_class
=None, *args
, **kwargs
):
1488 compat_urllib_request
.HTTPSHandler
.__init
__(self
, *args
, **kwargs
)
1489 self
._https
_conn
_class
= https_conn_class
or compat_http_client
.HTTPSConnection
1490 self
._params
= params
1492 def https_open(self
, req
):
1494 conn_class
= self
._https
_conn
_class
1496 if hasattr(self
, '_context'): # python > 2.6
1497 kwargs
['context'] = self
._context
1498 if hasattr(self
, '_check_hostname'): # python 3.x
1499 kwargs
['check_hostname'] = self
._check
_hostname
1501 socks_proxy
= req
.headers
.get('Ytdl-socks-proxy')
1503 conn_class
= make_socks_conn_class(conn_class
, socks_proxy
)
1504 del req
.headers
['Ytdl-socks-proxy']
1506 return self
.do_open(functools
.partial(
1507 _create_http_connection
, self
, conn_class
, True),
1511 class YoutubeDLCookieJar(compat_cookiejar
.MozillaCookieJar
):
1513 See [1] for cookie file format.
1515 1. https://curl.haxx.se/docs/http-cookies.html
1517 _HTTPONLY_PREFIX
= '#HttpOnly_'
1519 _HEADER
= '''# Netscape HTTP Cookie File
1520 # This file is generated by yt-dlp. Do not edit.
1523 _CookieFileEntry
= collections
.namedtuple(
1525 ('domain_name', 'include_subdomains', 'path', 'https_only', 'expires_at', 'name', 'value'))
1527 def save(self
, filename
=None, ignore_discard
=False, ignore_expires
=False):
1529 Save cookies to a file.
1531 Most of the code is taken from CPython 3.8 and slightly adapted
1532 to support cookie files with UTF-8 in both python 2 and 3.
1534 if filename
is None:
1535 if self
.filename
is not None:
1536 filename
= self
.filename
1538 raise ValueError(compat_cookiejar
.MISSING_FILENAME_TEXT
)
1540 # Store session cookies with `expires` set to 0 instead of an empty
1543 if cookie
.expires
is None:
1546 with io
.open(filename
, 'w', encoding
='utf-8') as f
:
1547 f
.write(self
._HEADER
)
1550 if not ignore_discard
and cookie
.discard
:
1552 if not ignore_expires
and cookie
.is_expired(now
):
1558 if cookie
.domain
.startswith('.'):
1559 initial_dot
= 'TRUE'
1561 initial_dot
= 'FALSE'
1562 if cookie
.expires
is not None:
1563 expires
= compat_str(cookie
.expires
)
1566 if cookie
.value
is None:
1567 # cookies.txt regards 'Set-Cookie: foo' as a cookie
1568 # with no name, whereas http.cookiejar regards it as a
1569 # cookie with no value.
1574 value
= cookie
.value
1576 '\t'.join([cookie
.domain
, initial_dot
, cookie
.path
,
1577 secure
, expires
, name
, value
]) + '\n')
1579 def load(self
, filename
=None, ignore_discard
=False, ignore_expires
=False):
1580 """Load cookies from a file."""
1581 if filename
is None:
1582 if self
.filename
is not None:
1583 filename
= self
.filename
1585 raise ValueError(compat_cookiejar
.MISSING_FILENAME_TEXT
)
1587 def prepare_line(line
):
1588 if line
.startswith(self
._HTTPONLY
_PREFIX
):
1589 line
= line
[len(self
._HTTPONLY
_PREFIX
):]
1590 # comments and empty lines are fine
1591 if line
.startswith('#') or not line
.strip():
1593 cookie_list
= line
.split('\t')
1594 if len(cookie_list
) != self
._ENTRY
_LEN
:
1595 raise compat_cookiejar
.LoadError('invalid length %d' % len(cookie_list
))
1596 cookie
= self
._CookieFileEntry
(*cookie_list
)
1597 if cookie
.expires_at
and not cookie
.expires_at
.isdigit():
1598 raise compat_cookiejar
.LoadError('invalid expires at %s' % cookie
.expires_at
)
1602 with io
.open(filename
, encoding
='utf-8') as f
:
1605 cf
.write(prepare_line(line
))
1606 except compat_cookiejar
.LoadError
as e
:
1608 'WARNING: skipping cookie file entry due to %s: %r\n'
1609 % (e
, line
), sys
.stderr
)
1612 self
._really
_load
(cf
, filename
, ignore_discard
, ignore_expires
)
1613 # Session cookies are denoted by either `expires` field set to
1614 # an empty string or 0. MozillaCookieJar only recognizes the former
1615 # (see [1]). So we need force the latter to be recognized as session
1616 # cookies on our own.
1617 # Session cookies may be important for cookies-based authentication,
1618 # e.g. usually, when user does not check 'Remember me' check box while
1619 # logging in on a site, some important cookies are stored as session
1620 # cookies so that not recognizing them will result in failed login.
1621 # 1. https://bugs.python.org/issue17164
1623 # Treat `expires=0` cookies as session cookies
1624 if cookie
.expires
== 0:
1625 cookie
.expires
= None
1626 cookie
.discard
= True
1629 class YoutubeDLCookieProcessor(compat_urllib_request
.HTTPCookieProcessor
):
1630 def __init__(self
, cookiejar
=None):
1631 compat_urllib_request
.HTTPCookieProcessor
.__init
__(self
, cookiejar
)
1633 def http_response(self
, request
, response
):
1634 # Python 2 will choke on next HTTP request in row if there are non-ASCII
1635 # characters in Set-Cookie HTTP header of last response (see
1636 # https://github.com/ytdl-org/youtube-dl/issues/6769).
1637 # In order to at least prevent crashing we will percent encode Set-Cookie
1638 # header before HTTPCookieProcessor starts processing it.
1639 # if sys.version_info < (3, 0) and response.headers:
1640 # for set_cookie_header in ('Set-Cookie', 'Set-Cookie2'):
1641 # set_cookie = response.headers.get(set_cookie_header)
1643 # set_cookie_escaped = compat_urllib_parse.quote(set_cookie, b"%/;:@&=+$,!~*'()?#[] ")
1644 # if set_cookie != set_cookie_escaped:
1645 # del response.headers[set_cookie_header]
1646 # response.headers[set_cookie_header] = set_cookie_escaped
1647 return compat_urllib_request
.HTTPCookieProcessor
.http_response(self
, request
, response
)
1649 https_request
= compat_urllib_request
.HTTPCookieProcessor
.http_request
1650 https_response
= http_response
1653 class YoutubeDLRedirectHandler(compat_urllib_request
.HTTPRedirectHandler
):
1654 """YoutubeDL redirect handler
1656 The code is based on HTTPRedirectHandler implementation from CPython [1].
1658 This redirect handler solves two issues:
1659 - ensures redirect URL is always unicode under python 2
1660 - introduces support for experimental HTTP response status code
1661 308 Permanent Redirect [2] used by some sites [3]
1663 1. https://github.com/python/cpython/blob/master/Lib/urllib/request.py
1664 2. https://developer.mozilla.org/en-US/docs/Web/HTTP/Status/308
1665 3. https://github.com/ytdl-org/youtube-dl/issues/28768
1668 http_error_301
= http_error_303
= http_error_307
= http_error_308
= compat_urllib_request
.HTTPRedirectHandler
.http_error_302
1670 def redirect_request(self
, req
, fp
, code
, msg
, headers
, newurl
):
1671 """Return a Request or None in response to a redirect.
1673 This is called by the http_error_30x methods when a
1674 redirection response is received. If a redirection should
1675 take place, return a new Request to allow http_error_30x to
1676 perform the redirect. Otherwise, raise HTTPError if no-one
1677 else should try to handle this url. Return None if you can't
1678 but another Handler might.
1680 m
= req
.get_method()
1681 if (not (code
in (301, 302, 303, 307, 308) and m
in ("GET", "HEAD")
1682 or code
in (301, 302, 303) and m
== "POST")):
1683 raise compat_HTTPError(req
.full_url
, code
, msg
, headers
, fp
)
1684 # Strictly (according to RFC 2616), 301 or 302 in response to
1685 # a POST MUST NOT cause a redirection without confirmation
1686 # from the user (of urllib.request, in this case). In practice,
1687 # essentially all clients do redirect in this case, so we do
1690 # On python 2 urlh.geturl() may sometimes return redirect URL
1691 # as byte string instead of unicode. This workaround allows
1692 # to force it always return unicode.
1693 if sys
.version_info
[0] < 3:
1694 newurl
= compat_str(newurl
)
1696 # Be conciliant with URIs containing a space. This is mainly
1697 # redundant with the more complete encoding done in http_error_302(),
1698 # but it is kept for compatibility with other callers.
1699 newurl
= newurl
.replace(' ', '%20')
1701 CONTENT_HEADERS
= ("content-length", "content-type")
1702 # NB: don't use dict comprehension for python 2.6 compatibility
1703 newheaders
= dict((k
, v
) for k
, v
in req
.headers
.items()
1704 if k
.lower() not in CONTENT_HEADERS
)
1705 return compat_urllib_request
.Request(
1706 newurl
, headers
=newheaders
, origin_req_host
=req
.origin_req_host
,
1710 def extract_timezone(date_str
):
1713 ^.{8,}? # >=8 char non-TZ prefix, if present
1714 (?P<tz>Z| # just the UTC Z, or
1715 (?:(?<=.\b\d{4}|\b\d{2}:\d\d)| # preceded by 4 digits or hh:mm or
1716 (?<!.\b[a-zA-Z]{3}|[a-zA-Z]{4}|..\b\d\d)) # not preceded by 3 alpha word or >= 4 alpha or 2 digits
1717 [ ]? # optional space
1718 (?P<sign>\+|-) # +/-
1719 (?P<hours>[0-9]{2}):?(?P<minutes>[0-9]{2}) # hh[:]mm
1723 timezone
= datetime
.timedelta()
1725 date_str
= date_str
[:-len(m
.group('tz'))]
1726 if not m
.group('sign'):
1727 timezone
= datetime
.timedelta()
1729 sign
= 1 if m
.group('sign') == '+' else -1
1730 timezone
= datetime
.timedelta(
1731 hours
=sign
* int(m
.group('hours')),
1732 minutes
=sign
* int(m
.group('minutes')))
1733 return timezone
, date_str
1736 def parse_iso8601(date_str
, delimiter
='T', timezone
=None):
1737 """ Return a UNIX timestamp from the given date """
1739 if date_str
is None:
1742 date_str
= re
.sub(r
'\.[0-9]+', '', date_str
)
1744 if timezone
is None:
1745 timezone
, date_str
= extract_timezone(date_str
)
1748 date_format
= '%Y-%m-%d{0}%H:%M:%S'.format(delimiter
)
1749 dt
= datetime
.datetime
.strptime(date_str
, date_format
) - timezone
1750 return calendar
.timegm(dt
.timetuple())
1755 def date_formats(day_first
=True):
1756 return DATE_FORMATS_DAY_FIRST
if day_first
else DATE_FORMATS_MONTH_FIRST
1759 def unified_strdate(date_str
, day_first
=True):
1760 """Return a string with the date in the format YYYYMMDD"""
1762 if date_str
is None:
1766 date_str
= date_str
.replace(',', ' ')
1767 # Remove AM/PM + timezone
1768 date_str
= re
.sub(r
'(?i)\s*(?:AM|PM)(?:\s+[A-Z]+)?', '', date_str
)
1769 _
, date_str
= extract_timezone(date_str
)
1771 for expression
in date_formats(day_first
):
1773 upload_date
= datetime
.datetime
.strptime(date_str
, expression
).strftime('%Y%m%d')
1776 if upload_date
is None:
1777 timetuple
= email
.utils
.parsedate_tz(date_str
)
1780 upload_date
= datetime
.datetime(*timetuple
[:6]).strftime('%Y%m%d')
1783 if upload_date
is not None:
1784 return compat_str(upload_date
)
1787 def unified_timestamp(date_str
, day_first
=True):
1788 if date_str
is None:
1791 date_str
= re
.sub(r
'[,|]', '', date_str
)
1793 pm_delta
= 12 if re
.search(r
'(?i)PM', date_str
) else 0
1794 timezone
, date_str
= extract_timezone(date_str
)
1796 # Remove AM/PM + timezone
1797 date_str
= re
.sub(r
'(?i)\s*(?:AM|PM)(?:\s+[A-Z]+)?', '', date_str
)
1799 # Remove unrecognized timezones from ISO 8601 alike timestamps
1800 m
= re
.search(r
'\d{1,2}:\d{1,2}(?:\.\d+)?(?P<tz>\s*[A-Z]+)$', date_str
)
1802 date_str
= date_str
[:-len(m
.group('tz'))]
1804 # Python only supports microseconds, so remove nanoseconds
1805 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]+$', date_str
)
1807 date_str
= m
.group(1)
1809 for expression
in date_formats(day_first
):
1811 dt
= datetime
.datetime
.strptime(date_str
, expression
) - timezone
+ datetime
.timedelta(hours
=pm_delta
)
1812 return calendar
.timegm(dt
.timetuple())
1815 timetuple
= email
.utils
.parsedate_tz(date_str
)
1817 return calendar
.timegm(timetuple
) + pm_delta
* 3600
1820 def determine_ext(url
, default_ext
='unknown_video'):
1821 if url
is None or '.' not in url
:
1823 guess
= url
.partition('?')[0].rpartition('.')[2]
1824 if re
.match(r
'^[A-Za-z0-9]+$', guess
):
1826 # Try extract ext from URLs like http://example.com/foo/bar.mp4/?download
1827 elif guess
.rstrip('/') in KNOWN_EXTENSIONS
:
1828 return guess
.rstrip('/')
1833 def subtitles_filename(filename
, sub_lang
, sub_format
, expected_real_ext
=None):
1834 return replace_extension(filename
, sub_lang
+ '.' + sub_format
, expected_real_ext
)
1837 def datetime_from_str(date_str
, precision
='auto', format
='%Y%m%d'):
1839 Return a datetime object from a string in the format YYYYMMDD or
1840 (now|today|date)[+-][0-9](microsecond|second|minute|hour|day|week|month|year)(s)?
1842 format: string date format used to return datetime object from
1843 precision: round the time portion of a datetime object.
1844 auto|microsecond|second|minute|hour|day.
1845 auto: round to the unit provided in date_str (if applicable).
1847 auto_precision
= False
1848 if precision
== 'auto':
1849 auto_precision
= True
1850 precision
= 'microsecond'
1851 today
= datetime_round(datetime
.datetime
.utcnow(), precision
)
1852 if date_str
in ('now', 'today'):
1854 if date_str
== 'yesterday':
1855 return today
- datetime
.timedelta(days
=1)
1857 r
'(?P<start>.+)(?P<sign>[+-])(?P<time>\d+)(?P<unit>microsecond|second|minute|hour|day|week|month|year)(s)?',
1859 if match
is not None:
1860 start_time
= datetime_from_str(match
.group('start'), precision
, format
)
1861 time
= int(match
.group('time')) * (-1 if match
.group('sign') == '-' else 1)
1862 unit
= match
.group('unit')
1863 if unit
== 'month' or unit
== 'year':
1864 new_date
= datetime_add_months(start_time
, time
* 12 if unit
== 'year' else time
)
1870 delta
= datetime
.timedelta(**{unit + 's': time}
)
1871 new_date
= start_time
+ delta
1873 return datetime_round(new_date
, unit
)
1876 return datetime_round(datetime
.datetime
.strptime(date_str
, format
), precision
)
1879 def date_from_str(date_str
, format
='%Y%m%d'):
1881 Return a datetime object from a string in the format YYYYMMDD or
1882 (now|today|date)[+-][0-9](microsecond|second|minute|hour|day|week|month|year)(s)?
1884 format: string date format used to return datetime object from
1886 return datetime_from_str(date_str
, precision
='microsecond', format
=format
).date()
1889 def datetime_add_months(dt
, months
):
1890 """Increment/Decrement a datetime object by months."""
1891 month
= dt
.month
+ months
- 1
1892 year
= dt
.year
+ month
// 12
1893 month
= month
% 12 + 1
1894 day
= min(dt
.day
, calendar
.monthrange(year
, month
)[1])
1895 return dt
.replace(year
, month
, day
)
1898 def datetime_round(dt
, precision
='day'):
1900 Round a datetime object's time to a specific precision
1902 if precision
== 'microsecond':
1911 roundto
= lambda x
, n
: ((x
+ n
/ 2) // n
) * n
1912 timestamp
= calendar
.timegm(dt
.timetuple())
1913 return datetime
.datetime
.utcfromtimestamp(roundto(timestamp
, unit_seconds
[precision
]))
1916 def hyphenate_date(date_str
):
1918 Convert a date in 'YYYYMMDD' format to 'YYYY-MM-DD' format"""
1919 match
= re
.match(r
'^(\d\d\d\d)(\d\d)(\d\d)$', date_str
)
1920 if match
is not None:
1921 return '-'.join(match
.groups())
1926 class DateRange(object):
1927 """Represents a time interval between two dates"""
1929 def __init__(self
, start
=None, end
=None):
1930 """start and end must be strings in the format accepted by date"""
1931 if start
is not None:
1932 self
.start
= date_from_str(start
)
1934 self
.start
= datetime
.datetime
.min.date()
1936 self
.end
= date_from_str(end
)
1938 self
.end
= datetime
.datetime
.max.date()
1939 if self
.start
> self
.end
:
1940 raise ValueError('Date range: "%s" , the start date must be before the end date' % self
)
1944 """Returns a range that only contains the given day"""
1945 return cls(day
, day
)
1947 def __contains__(self
, date
):
1948 """Check if the date is in the range"""
1949 if not isinstance(date
, datetime
.date
):
1950 date
= date_from_str(date
)
1951 return self
.start
<= date
<= self
.end
1954 return '%s - %s' % (self
.start
.isoformat(), self
.end
.isoformat())
1957 def platform_name():
1958 """ Returns the platform name as a compat_str """
1959 res
= platform
.platform()
1960 if isinstance(res
, bytes):
1961 res
= res
.decode(preferredencoding())
1963 assert isinstance(res
, compat_str
)
1967 def get_windows_version():
1968 ''' Get Windows version. None if it's not running on Windows '''
1969 if compat_os_name
== 'nt':
1970 return version_tuple(platform
.win32_ver()[1])
1975 def _windows_write_string(s
, out
):
1976 """ Returns True if the string was written using special methods,
1977 False if it has yet to be written out."""
1978 # Adapted from http://stackoverflow.com/a/3259271/35070
1980 import ctypes
.wintypes
1988 fileno
= out
.fileno()
1989 except AttributeError:
1990 # If the output stream doesn't have a fileno, it's virtual
1992 except io
.UnsupportedOperation
:
1993 # Some strange Windows pseudo files?
1995 if fileno
not in WIN_OUTPUT_IDS
:
1998 GetStdHandle
= compat_ctypes_WINFUNCTYPE(
1999 ctypes
.wintypes
.HANDLE
, ctypes
.wintypes
.DWORD
)(
2000 ('GetStdHandle', ctypes
.windll
.kernel32
))
2001 h
= GetStdHandle(WIN_OUTPUT_IDS
[fileno
])
2003 WriteConsoleW
= compat_ctypes_WINFUNCTYPE(
2004 ctypes
.wintypes
.BOOL
, ctypes
.wintypes
.HANDLE
, ctypes
.wintypes
.LPWSTR
,
2005 ctypes
.wintypes
.DWORD
, ctypes
.POINTER(ctypes
.wintypes
.DWORD
),
2006 ctypes
.wintypes
.LPVOID
)(('WriteConsoleW', ctypes
.windll
.kernel32
))
2007 written
= ctypes
.wintypes
.DWORD(0)
2009 GetFileType
= compat_ctypes_WINFUNCTYPE(ctypes
.wintypes
.DWORD
, ctypes
.wintypes
.DWORD
)(('GetFileType', ctypes
.windll
.kernel32
))
2010 FILE_TYPE_CHAR
= 0x0002
2011 FILE_TYPE_REMOTE
= 0x8000
2012 GetConsoleMode
= compat_ctypes_WINFUNCTYPE(
2013 ctypes
.wintypes
.BOOL
, ctypes
.wintypes
.HANDLE
,
2014 ctypes
.POINTER(ctypes
.wintypes
.DWORD
))(
2015 ('GetConsoleMode', ctypes
.windll
.kernel32
))
2016 INVALID_HANDLE_VALUE
= ctypes
.wintypes
.DWORD(-1).value
2018 def not_a_console(handle
):
2019 if handle
== INVALID_HANDLE_VALUE
or handle
is None:
2021 return ((GetFileType(handle
) & ~FILE_TYPE_REMOTE
) != FILE_TYPE_CHAR
2022 or GetConsoleMode(handle
, ctypes
.byref(ctypes
.wintypes
.DWORD())) == 0)
2024 if not_a_console(h
):
2027 def next_nonbmp_pos(s
):
2029 return next(i
for i
, c
in enumerate(s
) if ord(c
) > 0xffff)
2030 except StopIteration:
2034 count
= min(next_nonbmp_pos(s
), 1024)
2036 ret
= WriteConsoleW(
2037 h
, s
, count
if count
else 2, ctypes
.byref(written
), None)
2039 raise OSError('Failed to write string')
2040 if not count
: # We just wrote a non-BMP character
2041 assert written
.value
== 2
2044 assert written
.value
> 0
2045 s
= s
[written
.value
:]
2049 def write_string(s
, out
=None, encoding
=None):
2052 assert type(s
) == compat_str
2054 if sys
.platform
== 'win32' and encoding
is None and hasattr(out
, 'fileno'):
2055 if _windows_write_string(s
, out
):
2058 if ('b' in getattr(out
, 'mode', '')
2059 or sys
.version_info
[0] < 3): # Python 2 lies about mode of sys.stderr
2060 byt
= s
.encode(encoding
or preferredencoding(), 'ignore')
2062 elif hasattr(out
, 'buffer'):
2063 enc
= encoding
or getattr(out
, 'encoding', None) or preferredencoding()
2064 byt
= s
.encode(enc
, 'ignore')
2065 out
.buffer.write(byt
)
2071 def bytes_to_intlist(bs
):
2074 if isinstance(bs
[0], int): # Python 3
2077 return [ord(c
) for c
in bs
]
2080 def intlist_to_bytes(xs
):
2083 return compat_struct_pack('%dB' % len(xs
), *xs
)
2086 # Cross-platform file locking
2087 if sys
.platform
== 'win32':
2088 import ctypes
.wintypes
2091 class OVERLAPPED(ctypes
.Structure
):
2093 ('Internal', ctypes
.wintypes
.LPVOID
),
2094 ('InternalHigh', ctypes
.wintypes
.LPVOID
),
2095 ('Offset', ctypes
.wintypes
.DWORD
),
2096 ('OffsetHigh', ctypes
.wintypes
.DWORD
),
2097 ('hEvent', ctypes
.wintypes
.HANDLE
),
2100 kernel32
= ctypes
.windll
.kernel32
2101 LockFileEx
= kernel32
.LockFileEx
2102 LockFileEx
.argtypes
= [
2103 ctypes
.wintypes
.HANDLE
, # hFile
2104 ctypes
.wintypes
.DWORD
, # dwFlags
2105 ctypes
.wintypes
.DWORD
, # dwReserved
2106 ctypes
.wintypes
.DWORD
, # nNumberOfBytesToLockLow
2107 ctypes
.wintypes
.DWORD
, # nNumberOfBytesToLockHigh
2108 ctypes
.POINTER(OVERLAPPED
) # Overlapped
2110 LockFileEx
.restype
= ctypes
.wintypes
.BOOL
2111 UnlockFileEx
= kernel32
.UnlockFileEx
2112 UnlockFileEx
.argtypes
= [
2113 ctypes
.wintypes
.HANDLE
, # hFile
2114 ctypes
.wintypes
.DWORD
, # dwReserved
2115 ctypes
.wintypes
.DWORD
, # nNumberOfBytesToLockLow
2116 ctypes
.wintypes
.DWORD
, # nNumberOfBytesToLockHigh
2117 ctypes
.POINTER(OVERLAPPED
) # Overlapped
2119 UnlockFileEx
.restype
= ctypes
.wintypes
.BOOL
2120 whole_low
= 0xffffffff
2121 whole_high
= 0x7fffffff
2123 def _lock_file(f
, exclusive
):
2124 overlapped
= OVERLAPPED()
2125 overlapped
.Offset
= 0
2126 overlapped
.OffsetHigh
= 0
2127 overlapped
.hEvent
= 0
2128 f
._lock
_file
_overlapped
_p
= ctypes
.pointer(overlapped
)
2129 handle
= msvcrt
.get_osfhandle(f
.fileno())
2130 if not LockFileEx(handle
, 0x2 if exclusive
else 0x0, 0,
2131 whole_low
, whole_high
, f
._lock
_file
_overlapped
_p
):
2132 raise OSError('Locking file failed: %r' % ctypes
.FormatError())
2134 def _unlock_file(f
):
2135 assert f
._lock
_file
_overlapped
_p
2136 handle
= msvcrt
.get_osfhandle(f
.fileno())
2137 if not UnlockFileEx(handle
, 0,
2138 whole_low
, whole_high
, f
._lock
_file
_overlapped
_p
):
2139 raise OSError('Unlocking file failed: %r' % ctypes
.FormatError())
2142 # Some platforms, such as Jython, is missing fcntl
2146 def _lock_file(f
, exclusive
):
2147 fcntl
.flock(f
, fcntl
.LOCK_EX
if exclusive
else fcntl
.LOCK_SH
)
2149 def _unlock_file(f
):
2150 fcntl
.flock(f
, fcntl
.LOCK_UN
)
2152 UNSUPPORTED_MSG
= 'file locking is not supported on this platform'
2154 def _lock_file(f
, exclusive
):
2155 raise IOError(UNSUPPORTED_MSG
)
2157 def _unlock_file(f
):
2158 raise IOError(UNSUPPORTED_MSG
)
2161 class locked_file(object):
2162 def __init__(self
, filename
, mode
, encoding
=None):
2163 assert mode
in ['r', 'a', 'w']
2164 self
.f
= io
.open(filename
, mode
, encoding
=encoding
)
2167 def __enter__(self
):
2168 exclusive
= self
.mode
!= 'r'
2170 _lock_file(self
.f
, exclusive
)
2176 def __exit__(self
, etype
, value
, traceback
):
2178 _unlock_file(self
.f
)
2185 def write(self
, *args
):
2186 return self
.f
.write(*args
)
2188 def read(self
, *args
):
2189 return self
.f
.read(*args
)
2192 def get_filesystem_encoding():
2193 encoding
= sys
.getfilesystemencoding()
2194 return encoding
if encoding
is not None else 'utf-8'
2197 def shell_quote(args
):
2199 encoding
= get_filesystem_encoding()
2201 if isinstance(a
, bytes):
2202 # We may get a filename encoded with 'encodeFilename'
2203 a
= a
.decode(encoding
)
2204 quoted_args
.append(compat_shlex_quote(a
))
2205 return ' '.join(quoted_args
)
2208 def smuggle_url(url
, data
):
2209 """ Pass additional data in a URL for internal use. """
2211 url
, idata
= unsmuggle_url(url
, {})
2213 sdata
= compat_urllib_parse_urlencode(
2214 {'__youtubedl_smuggle': json.dumps(data)}
)
2215 return url
+ '#' + sdata
2218 def unsmuggle_url(smug_url
, default
=None):
2219 if '#__youtubedl_smuggle' not in smug_url
:
2220 return smug_url
, default
2221 url
, _
, sdata
= smug_url
.rpartition('#')
2222 jsond
= compat_parse_qs(sdata
)['__youtubedl_smuggle'][0]
2223 data
= json
.loads(jsond
)
2227 def format_decimal_suffix(num
, fmt
='%d%s', *, factor
=1000):
2228 """ Formats numbers with decimal sufixes like K, M, etc """
2229 num
, factor
= float_or_none(num
), float(factor
)
2232 exponent
= 0 if num
== 0 else int(math
.log(num
, factor
))
2233 suffix
= ['', *'kMGTPEZY'][exponent
]
2235 suffix
= {'k': 'Ki', '': ''}
.get(suffix
, f
'{suffix}i')
2236 converted
= num
/ (factor
** exponent
)
2237 return fmt
% (converted
, suffix
)
2240 def format_bytes(bytes):
2241 return format_decimal_suffix(bytes, '%.2f%sB', factor
=1024) or 'N/A'
2244 def lookup_unit_table(unit_table
, s
):
2245 units_re
= '|'.join(re
.escape(u
) for u
in unit_table
)
2247 r
'(?P<num>[0-9]+(?:[,.][0-9]*)?)\s*(?P<unit>%s)\b' % units_re
, s
)
2250 num_str
= m
.group('num').replace(',', '.')
2251 mult
= unit_table
[m
.group('unit')]
2252 return int(float(num_str
) * mult
)
2255 def parse_filesize(s
):
2259 # The lower-case forms are of course incorrect and unofficial,
2260 # but we support those too
2277 'megabytes': 1000 ** 2,
2278 'mebibytes': 1024 ** 2,
2284 'gigabytes': 1000 ** 3,
2285 'gibibytes': 1024 ** 3,
2291 'terabytes': 1000 ** 4,
2292 'tebibytes': 1024 ** 4,
2298 'petabytes': 1000 ** 5,
2299 'pebibytes': 1024 ** 5,
2305 'exabytes': 1000 ** 6,
2306 'exbibytes': 1024 ** 6,
2312 'zettabytes': 1000 ** 7,
2313 'zebibytes': 1024 ** 7,
2319 'yottabytes': 1000 ** 8,
2320 'yobibytes': 1024 ** 8,
2323 return lookup_unit_table(_UNIT_TABLE
, s
)
2330 s
= re
.sub(r
'^[^\d]+\s', '', s
).strip()
2332 if re
.match(r
'^[\d,.]+$', s
):
2333 return str_to_int(s
)
2346 ret
= lookup_unit_table(_UNIT_TABLE
, s
)
2350 mobj
= re
.match(r
'([\d,.]+)(?:$|\s)', s
)
2352 return str_to_int(mobj
.group(1))
2355 def parse_resolution(s
):
2359 mobj
= re
.search(r
'(?<![a-zA-Z0-9])(?P<w>\d+)\s*[xX×,]\s*(?P<h>\d+)(?![a-zA-Z0-9])', s
)
2362 'width': int(mobj
.group('w')),
2363 'height': int(mobj
.group('h')),
2366 mobj
= re
.search(r
'(?<![a-zA-Z0-9])(\d+)[pPiI](?![a-zA-Z0-9])', s
)
2368 return {'height': int(mobj.group(1))}
2370 mobj
= re
.search(r
'\b([48])[kK]\b', s
)
2372 return {'height': int(mobj.group(1)) * 540}
2377 def parse_bitrate(s
):
2378 if not isinstance(s
, compat_str
):
2380 mobj
= re
.search(r
'\b(\d+)\s*kbps', s
)
2382 return int(mobj
.group(1))
2385 def month_by_name(name
, lang
='en'):
2386 """ Return the number of a month by (locale-independently) English name """
2388 month_names
= MONTH_NAMES
.get(lang
, MONTH_NAMES
['en'])
2391 return month_names
.index(name
) + 1
2396 def month_by_abbreviation(abbrev
):
2397 """ Return the number of a month by (locale-independently) English
2401 return [s
[:3] for s
in ENGLISH_MONTH_NAMES
].index(abbrev
) + 1
2406 def fix_xml_ampersands(xml_str
):
2407 """Replace all the '&' by '&' in XML"""
2409 r
'&(?!amp;|lt;|gt;|apos;|quot;|#x[0-9a-fA-F]{,4};|#[0-9]{,4};)',
2414 def setproctitle(title
):
2415 assert isinstance(title
, compat_str
)
2417 # ctypes in Jython is not complete
2418 # http://bugs.jython.org/issue2148
2419 if sys
.platform
.startswith('java'):
2423 libc
= ctypes
.cdll
.LoadLibrary('libc.so.6')
2427 # LoadLibrary in Windows Python 2.7.13 only expects
2428 # a bytestring, but since unicode_literals turns
2429 # every string into a unicode string, it fails.
2431 title_bytes
= title
.encode('utf-8')
2432 buf
= ctypes
.create_string_buffer(len(title_bytes
))
2433 buf
.value
= title_bytes
2435 libc
.prctl(15, buf
, 0, 0, 0)
2436 except AttributeError:
2437 return # Strange libc, just skip this
2440 def remove_start(s
, start
):
2441 return s
[len(start
):] if s
is not None and s
.startswith(start
) else s
2444 def remove_end(s
, end
):
2445 return s
[:-len(end
)] if s
is not None and s
.endswith(end
) else s
2448 def remove_quotes(s
):
2449 if s
is None or len(s
) < 2:
2451 for quote
in ('"', "'", ):
2452 if s
[0] == quote
and s
[-1] == quote
:
2457 def get_domain(url
):
2458 domain
= re
.match(r
'(?:https?:\/\/)?(?:www\.)?(?P<domain>[^\n\/]+\.[^\n\/]+)(?:\/(.*))?', url
)
2459 return domain
.group('domain') if domain
else None
2462 def url_basename(url
):
2463 path
= compat_urlparse
.urlparse(url
).path
2464 return path
.strip('/').split('/')[-1]
2468 return re
.match(r
'https?://[^?#&]+/', url
).group()
2471 def urljoin(base
, path
):
2472 if isinstance(path
, bytes):
2473 path
= path
.decode('utf-8')
2474 if not isinstance(path
, compat_str
) or not path
:
2476 if re
.match(r
'^(?:[a-zA-Z][a-zA-Z0-9+-.]*:)?//', path
):
2478 if isinstance(base
, bytes):
2479 base
= base
.decode('utf-8')
2480 if not isinstance(base
, compat_str
) or not re
.match(
2481 r
'^(?:https?:)?//', base
):
2483 return compat_urlparse
.urljoin(base
, path
)
2486 class HEADRequest(compat_urllib_request
.Request
):
2487 def get_method(self
):
2491 class PUTRequest(compat_urllib_request
.Request
):
2492 def get_method(self
):
2496 def int_or_none(v
, scale
=1, default
=None, get_attr
=None, invscale
=1):
2497 if get_attr
and v
is not None:
2498 v
= getattr(v
, get_attr
, None)
2500 return int(v
) * invscale
// scale
2501 except (ValueError, TypeError, OverflowError):
2505 def str_or_none(v
, default
=None):
2506 return default
if v
is None else compat_str(v
)
2509 def str_to_int(int_str
):
2510 """ A more relaxed version of int_or_none """
2511 if isinstance(int_str
, compat_integer_types
):
2513 elif isinstance(int_str
, compat_str
):
2514 int_str
= re
.sub(r
'[,\.\+]', '', int_str
)
2515 return int_or_none(int_str
)
2518 def float_or_none(v
, scale
=1, invscale
=1, default
=None):
2522 return float(v
) * invscale
/ scale
2523 except (ValueError, TypeError):
2527 def bool_or_none(v
, default
=None):
2528 return v
if isinstance(v
, bool) else default
2531 def strip_or_none(v
, default
=None):
2532 return v
.strip() if isinstance(v
, compat_str
) else default
2535 def url_or_none(url
):
2536 if not url
or not isinstance(url
, compat_str
):
2539 return url
if re
.match(r
'^(?:(?:https?|rt(?:m(?:pt?[es]?|fp)|sp[su]?)|mms|ftps?):)?//', url
) else None
2542 def strftime_or_none(timestamp
, date_format
, default
=None):
2543 datetime_object
= None
2545 if isinstance(timestamp
, compat_numeric_types
): # unix timestamp
2546 datetime_object
= datetime
.datetime
.utcfromtimestamp(timestamp
)
2547 elif isinstance(timestamp
, compat_str
): # assume YYYYMMDD
2548 datetime_object
= datetime
.datetime
.strptime(timestamp
, '%Y%m%d')
2549 return datetime_object
.strftime(date_format
)
2550 except (ValueError, TypeError, AttributeError):
2554 def parse_duration(s
):
2555 if not isinstance(s
, compat_basestring
):
2561 days
, hours
, mins
, secs
, ms
= [None] * 5
2562 m
= re
.match(r
'''(?x)
2564 (?:(?:(?P<days>[0-9]+):)?(?P<hours>[0-9]+):)?(?P<mins>[0-9]+):)?
2565 (?P<secs>(?(before_secs)[0-9]{1,2}|[0-9]+))
2566 (?P<ms>[.:][0-9]+)?Z?$
2569 days
, hours
, mins
, secs
, ms
= m
.group('days', 'hours', 'mins', 'secs', 'ms')
2574 [0-9]+\s*y(?:ears?)?\s*
2577 [0-9]+\s*m(?:onths?)?\s*
2580 [0-9]+\s*w(?:eeks?)?\s*
2583 (?P<days>[0-9]+)\s*d(?:ays?)?\s*
2587 (?P<hours>[0-9]+)\s*h(?:ours?)?\s*
2590 (?P<mins>[0-9]+)\s*m(?:in(?:ute)?s?)?\s*
2593 (?P<secs>[0-9]+)(?P<ms>\.[0-9]+)?\s*s(?:ec(?:ond)?s?)?\s*
2596 days
, hours
, mins
, secs
, ms
= m
.groups()
2598 m
= re
.match(r
'(?i)(?:(?P<hours>[0-9.]+)\s*(?:hours?)|(?P<mins>[0-9.]+)\s*(?:mins?\.?|minutes?)\s*)Z?$', s
)
2600 hours
, mins
= m
.groups()
2606 duration
+= float(secs
)
2608 duration
+= float(mins
) * 60
2610 duration
+= float(hours
) * 60 * 60
2612 duration
+= float(days
) * 24 * 60 * 60
2614 duration
+= float(ms
.replace(':', '.'))
2618 def prepend_extension(filename
, ext
, expected_real_ext
=None):
2619 name
, real_ext
= os
.path
.splitext(filename
)
2621 '{0}.{1}{2}'.format(name
, ext
, real_ext
)
2622 if not expected_real_ext
or real_ext
[1:] == expected_real_ext
2623 else '{0}.{1}'.format(filename
, ext
))
2626 def replace_extension(filename
, ext
, expected_real_ext
=None):
2627 name
, real_ext
= os
.path
.splitext(filename
)
2628 return '{0}.{1}'.format(
2629 name
if not expected_real_ext
or real_ext
[1:] == expected_real_ext
else filename
,
2633 def check_executable(exe
, args
=[]):
2634 """ Checks if the given binary is installed somewhere in PATH, and returns its name.
2635 args can be a list of arguments for a short output (like -version) """
2637 Popen([exe
] + args
, stdout
=subprocess
.PIPE
, stderr
=subprocess
.PIPE
).communicate_or_kill()
2643 def _get_exe_version_output(exe
, args
):
2645 # STDIN should be redirected too. On UNIX-like systems, ffmpeg triggers
2646 # SIGTTOU if yt-dlp is run in the background.
2647 # See https://github.com/ytdl-org/youtube-dl/issues/955#issuecomment-209789656
2649 [encodeArgument(exe
)] + args
, stdin
=subprocess
.PIPE
,
2650 stdout
=subprocess
.PIPE
, stderr
=subprocess
.STDOUT
).communicate_or_kill()
2653 if isinstance(out
, bytes): # Python 2.x
2654 out
= out
.decode('ascii', 'ignore')
2658 def detect_exe_version(output
, version_re
=None, unrecognized
='present'):
2659 assert isinstance(output
, compat_str
)
2660 if version_re
is None:
2661 version_re
= r
'version\s+([-0-9._a-zA-Z]+)'
2662 m
= re
.search(version_re
, output
)
2669 def get_exe_version(exe
, args
=['--version'],
2670 version_re
=None, unrecognized
='present'):
2671 """ Returns the version of the specified executable,
2672 or False if the executable is not present """
2673 out
= _get_exe_version_output(exe
, args
)
2674 return detect_exe_version(out
, version_re
, unrecognized
) if out
else False
2677 class LazyList(collections
.abc
.Sequence
):
2678 ''' Lazy immutable list from an iterable
2679 Note that slices of a LazyList are lists and not LazyList'''
2681 class IndexError(IndexError):
2684 def __init__(self
, iterable
, *, reverse
=False, _cache
=None):
2685 self
.__iterable
= iter(iterable
)
2686 self
.__cache
= [] if _cache
is None else _cache
2687 self
.__reversed
= reverse
2691 # We need to consume the entire iterable to iterate in reverse
2692 yield from self
.exhaust()
2694 yield from self
.__cache
2695 for item
in self
.__iterable
:
2696 self
.__cache
.append(item
)
2699 def __exhaust(self
):
2700 self
.__cache
.extend(self
.__iterable
)
2701 # Discard the emptied iterable to make it pickle-able
2702 self
.__iterable
= []
2706 ''' Evaluate the entire iterable '''
2707 return self
.__exhaust
()[::-1 if self
.__reversed
else 1]
2710 def __reverse_index(x
):
2711 return None if x
is None else -(x
+ 1)
2713 def __getitem__(self
, idx
):
2714 if isinstance(idx
, slice):
2716 idx
= slice(self
.__reverse
_index
(idx
.start
), self
.__reverse
_index
(idx
.stop
), -(idx
.step
or 1))
2717 start
, stop
, step
= idx
.start
, idx
.stop
, idx
.step
or 1
2718 elif isinstance(idx
, int):
2720 idx
= self
.__reverse
_index
(idx
)
2721 start
, stop
, step
= idx
, idx
, 0
2723 raise TypeError('indices must be integers or slices')
2724 if ((start
or 0) < 0 or (stop
or 0) < 0
2725 or (start
is None and step
< 0)
2726 or (stop
is None and step
> 0)):
2727 # We need to consume the entire iterable to be able to slice from the end
2728 # Obviously, never use this with infinite iterables
2731 return self
.__cache
[idx
]
2732 except IndexError as e
:
2733 raise self
.IndexError(e
) from e
2734 n
= max(start
or 0, stop
or 0) - len(self
.__cache
) + 1
2736 self
.__cache
.extend(itertools
.islice(self
.__iterable
, n
))
2738 return self
.__cache
[idx
]
2739 except IndexError as e
:
2740 raise self
.IndexError(e
) from e
2744 self
[-1] if self
.__reversed
else self
[0]
2745 except self
.IndexError:
2751 return len(self
.__cache
)
2753 def __reversed__(self
):
2754 return type(self
)(self
.__iterable
, reverse
=not self
.__reversed
, _cache
=self
.__cache
)
2757 return type(self
)(self
.__iterable
, reverse
=self
.__reversed
, _cache
=self
.__cache
)
2760 # repr and str should mimic a list. So we exhaust the iterable
2761 return repr(self
.exhaust())
2764 return repr(self
.exhaust())
2769 class IndexError(IndexError):
2773 # This is only useful for tests
2774 return len(self
.getslice())
2776 def __init__(self
, pagefunc
, pagesize
, use_cache
=True):
2777 self
._pagefunc
= pagefunc
2778 self
._pagesize
= pagesize
2779 self
._use
_cache
= use_cache
2782 def getpage(self
, pagenum
):
2783 page_results
= self
._cache
.get(pagenum
)
2784 if page_results
is None:
2785 page_results
= list(self
._pagefunc
(pagenum
))
2787 self
._cache
[pagenum
] = page_results
2790 def getslice(self
, start
=0, end
=None):
2791 return list(self
._getslice
(start
, end
))
2793 def _getslice(self
, start
, end
):
2794 raise NotImplementedError('This method must be implemented by subclasses')
2796 def __getitem__(self
, idx
):
2797 # NOTE: cache must be enabled if this is used
2798 if not isinstance(idx
, int) or idx
< 0:
2799 raise TypeError('indices must be non-negative integers')
2800 entries
= self
.getslice(idx
, idx
+ 1)
2802 raise self
.IndexError()
2806 class OnDemandPagedList(PagedList
):
2807 def _getslice(self
, start
, end
):
2808 for pagenum
in itertools
.count(start
// self
._pagesize
):
2809 firstid
= pagenum
* self
._pagesize
2810 nextfirstid
= pagenum
* self
._pagesize
+ self
._pagesize
2811 if start
>= nextfirstid
:
2815 start
% self
._pagesize
2816 if firstid
<= start
< nextfirstid
2819 ((end
- 1) % self
._pagesize
) + 1
2820 if (end
is not None and firstid
<= end
<= nextfirstid
)
2823 page_results
= self
.getpage(pagenum
)
2824 if startv
!= 0 or endv
is not None:
2825 page_results
= page_results
[startv
:endv
]
2826 yield from page_results
2828 # A little optimization - if current page is not "full", ie. does
2829 # not contain page_size videos then we can assume that this page
2830 # is the last one - there are no more ids on further pages -
2831 # i.e. no need to query again.
2832 if len(page_results
) + startv
< self
._pagesize
:
2835 # If we got the whole page, but the next page is not interesting,
2836 # break out early as well
2837 if end
== nextfirstid
:
2841 class InAdvancePagedList(PagedList
):
2842 def __init__(self
, pagefunc
, pagecount
, pagesize
):
2843 self
._pagecount
= pagecount
2844 PagedList
.__init
__(self
, pagefunc
, pagesize
, True)
2846 def _getslice(self
, start
, end
):
2847 start_page
= start
// self
._pagesize
2848 end_page
= self
._pagecount
if end
is None else min(self
._pagecount
, end
// self
._pagesize
+ 1)
2849 skip_elems
= start
- start_page
* self
._pagesize
2850 only_more
= None if end
is None else end
- start
2851 for pagenum
in range(start_page
, end_page
):
2852 page_results
= self
.getpage(pagenum
)
2854 page_results
= page_results
[skip_elems
:]
2856 if only_more
is not None:
2857 if len(page_results
) < only_more
:
2858 only_more
-= len(page_results
)
2860 yield from page_results
[:only_more
]
2862 yield from page_results
2865 def uppercase_escape(s
):
2866 unicode_escape
= codecs
.getdecoder('unicode_escape')
2868 r
'\\U[0-9a-fA-F]{8}',
2869 lambda m
: unicode_escape(m
.group(0))[0],
2873 def lowercase_escape(s
):
2874 unicode_escape
= codecs
.getdecoder('unicode_escape')
2876 r
'\\u[0-9a-fA-F]{4}',
2877 lambda m
: unicode_escape(m
.group(0))[0],
2881 def escape_rfc3986(s
):
2882 """Escape non-ASCII characters as suggested by RFC 3986"""
2883 if sys
.version_info
< (3, 0) and isinstance(s
, compat_str
):
2884 s
= s
.encode('utf-8')
2885 return compat_urllib_parse
.quote(s
, b
"%/;:@&=+$,!~*'()?#[]")
2888 def escape_url(url
):
2889 """Escape URL as suggested by RFC 3986"""
2890 url_parsed
= compat_urllib_parse_urlparse(url
)
2891 return url_parsed
._replace
(
2892 netloc
=url_parsed
.netloc
.encode('idna').decode('ascii'),
2893 path
=escape_rfc3986(url_parsed
.path
),
2894 params
=escape_rfc3986(url_parsed
.params
),
2895 query
=escape_rfc3986(url_parsed
.query
),
2896 fragment
=escape_rfc3986(url_parsed
.fragment
)
2901 return compat_parse_qs(compat_urllib_parse_urlparse(url
).query
)
2904 def read_batch_urls(batch_fd
):
2906 if not isinstance(url
, compat_str
):
2907 url
= url
.decode('utf-8', 'replace')
2908 BOM_UTF8
= ('\xef\xbb\xbf', '\ufeff')
2909 for bom
in BOM_UTF8
:
2910 if url
.startswith(bom
):
2911 url
= url
[len(bom
):]
2913 if not url
or url
.startswith(('#', ';', ']')):
2915 # "#" cannot be stripped out since it is part of the URI
2916 # However, it can be safely stipped out if follwing a whitespace
2917 return re
.split(r
'\s#', url
, 1)[0].rstrip()
2919 with contextlib
.closing(batch_fd
) as fd
:
2920 return [url
for url
in map(fixup
, fd
) if url
]
2923 def urlencode_postdata(*args
, **kargs
):
2924 return compat_urllib_parse_urlencode(*args
, **kargs
).encode('ascii')
2927 def update_url_query(url
, query
):
2930 parsed_url
= compat_urlparse
.urlparse(url
)
2931 qs
= compat_parse_qs(parsed_url
.query
)
2933 return compat_urlparse
.urlunparse(parsed_url
._replace
(
2934 query
=compat_urllib_parse_urlencode(qs
, True)))
2937 def update_Request(req
, url
=None, data
=None, headers
={}, query={}
):
2938 req_headers
= req
.headers
.copy()
2939 req_headers
.update(headers
)
2940 req_data
= data
or req
.data
2941 req_url
= update_url_query(url
or req
.get_full_url(), query
)
2942 req_get_method
= req
.get_method()
2943 if req_get_method
== 'HEAD':
2944 req_type
= HEADRequest
2945 elif req_get_method
== 'PUT':
2946 req_type
= PUTRequest
2948 req_type
= compat_urllib_request
.Request
2950 req_url
, data
=req_data
, headers
=req_headers
,
2951 origin_req_host
=req
.origin_req_host
, unverifiable
=req
.unverifiable
)
2952 if hasattr(req
, 'timeout'):
2953 new_req
.timeout
= req
.timeout
2957 def _multipart_encode_impl(data
, boundary
):
2958 content_type
= 'multipart/form-data; boundary=%s' % boundary
2961 for k
, v
in data
.items():
2962 out
+= b
'--' + boundary
.encode('ascii') + b
'\r\n'
2963 if isinstance(k
, compat_str
):
2964 k
= k
.encode('utf-8')
2965 if isinstance(v
, compat_str
):
2966 v
= v
.encode('utf-8')
2967 # RFC 2047 requires non-ASCII field names to be encoded, while RFC 7578
2968 # suggests sending UTF-8 directly. Firefox sends UTF-8, too
2969 content
= b
'Content-Disposition: form-data; name="' + k
+ b
'"\r\n\r\n' + v
+ b
'\r\n'
2970 if boundary
.encode('ascii') in content
:
2971 raise ValueError('Boundary overlaps with data')
2974 out
+= b
'--' + boundary
.encode('ascii') + b
'--\r\n'
2976 return out
, content_type
2979 def multipart_encode(data
, boundary
=None):
2981 Encode a dict to RFC 7578-compliant form-data
2984 A dict where keys and values can be either Unicode or bytes-like
2987 If specified a Unicode object, it's used as the boundary. Otherwise
2988 a random boundary is generated.
2990 Reference: https://tools.ietf.org/html/rfc7578
2992 has_specified_boundary
= boundary
is not None
2995 if boundary
is None:
2996 boundary
= '---------------' + str(random
.randrange(0x0fffffff, 0xffffffff))
2999 out
, content_type
= _multipart_encode_impl(data
, boundary
)
3002 if has_specified_boundary
:
3006 return out
, content_type
3009 def dict_get(d
, key_or_keys
, default
=None, skip_false_values
=True):
3010 if isinstance(key_or_keys
, (list, tuple)):
3011 for key
in key_or_keys
:
3012 if key
not in d
or d
[key
] is None or skip_false_values
and not d
[key
]:
3016 return d
.get(key_or_keys
, default
)
3019 def try_get(src
, getter
, expected_type
=None):
3020 for get
in variadic(getter
):
3023 except (AttributeError, KeyError, TypeError, IndexError):
3026 if expected_type
is None or isinstance(v
, expected_type
):
3030 def merge_dicts(*dicts
):
3032 for a_dict
in dicts
:
3033 for k
, v
in a_dict
.items():
3037 or (isinstance(v
, compat_str
) and v
3038 and isinstance(merged
[k
], compat_str
)
3039 and not merged
[k
])):
3044 def encode_compat_str(string
, encoding
=preferredencoding(), errors
='strict'):
3045 return string
if isinstance(string
, compat_str
) else compat_str(string
, encoding
, errors
)
3057 TV_PARENTAL_GUIDELINES
= {
3067 def parse_age_limit(s
):
3069 return s
if 0 <= s
<= 21 else None
3070 if not isinstance(s
, compat_basestring
):
3072 m
= re
.match(r
'^(?P<age>\d{1,2})\+?$', s
)
3074 return int(m
.group('age'))
3077 return US_RATINGS
[s
]
3078 m
= re
.match(r
'^TV[_-]?(%s)$' % '|'.join(k
[3:] for k
in TV_PARENTAL_GUIDELINES
), s
)
3080 return TV_PARENTAL_GUIDELINES
['TV-' + m
.group(1)]
3084 def strip_jsonp(code
):
3087 (?:window\.)?(?P<func_name>[a-zA-Z0-9_.$]*)
3088 (?:\s*&&\s*(?P=func_name))?
3089 \s*\(\s*(?P<callback_data>.*)\);?
3090 \s*?(?://[^\n]*)*$''',
3091 r
'\g<callback_data>', code
)
3094 def js_to_json(code
, vars={}):
3095 # vars is a dict of var, val pairs to substitute
3096 COMMENT_RE
= r
'/\*(?:(?!\*/).)*?\*/|//[^\n]*\n'
3097 SKIP_RE
= r
'\s*(?:{comment})?\s*'.format(comment
=COMMENT_RE
)
3099 (r
'(?s)^(0[xX][0-9a-fA-F]+){skip}:?$'.format(skip
=SKIP_RE
), 16),
3100 (r
'(?s)^(0+[0-7]+){skip}:?$'.format(skip
=SKIP_RE
), 8),
3105 if v
in ('true', 'false', 'null'):
3107 elif v
in ('undefined', 'void 0'):
3109 elif v
.startswith('/*') or v
.startswith('//') or v
.startswith('!') or v
== ',':
3112 if v
[0] in ("'", '"'):
3113 v
= re
.sub(r
'(?s)\\.|"', lambda m
: {
3118 }.get(m
.group(0), m
.group(0)), v
[1:-1])
3120 for regex
, base
in INTEGER_TABLE
:
3121 im
= re
.match(regex
, v
)
3123 i
= int(im
.group(1), base
)
3124 return '"%d":' % i
if v
.endswith(':') else '%d' % i
3131 return re
.sub(r
'''(?sx)
3132 "(?:[^"\\]*(?:\\\\|\\['"nurtbfx/\n]))*[^"\\]*"|
3133 '(?:[^'\\]*(?:\\\\|\\['"nurtbfx/\n]))*[^'\\]*'|
3134 {comment}|,(?={skip}[\]}}])|
3135 void\s0|(?:(?<![0-9])[eE]|[a-df-zA-DF-Z_$])[.a-zA-Z_$0-9]*|
3136 \b(?:0[xX][0-9a-fA-F]+|0+[0-7]+)(?:{skip}:)?|
3139 '''.format(comment
=COMMENT_RE
, skip
=SKIP_RE
), fix_kv
, code
)
3142 def qualities(quality_ids
):
3143 """ Get a numeric quality value out of a list of possible values """
3146 return quality_ids
.index(qid
)
3152 POSTPROCESS_WHEN
= {'pre_process', 'before_dl', 'after_move', 'post_process', 'after_video', 'playlist'}
3156 'default': '%(title)s [%(id)s].%(ext)s',
3157 'chapter': '%(title)s - %(section_number)03d %(section_title)s [%(id)s].%(ext)s',
3163 'description': 'description',
3164 'annotation': 'annotations.xml',
3165 'infojson': 'info.json',
3168 'pl_thumbnail': None,
3169 'pl_description': 'description',
3170 'pl_infojson': 'info.json',
3173 # As of [1] format syntax is:
3174 # %[mapping_key][conversion_flags][minimum_width][.precision][length_modifier]type
3175 # 1. https://docs.python.org/2/library/stdtypes.html#string-formatting
3176 STR_FORMAT_RE_TMPL
= r
'''(?x)
3177 (?<!%)(?P<prefix>(?:%%)*)
3179 (?P<has_key>\((?P<key>{0})\))?
3181 (?P<conversion>[#0\-+ ]+)?
3183 (?P<precision>\.\d+)?
3184 (?P<len_mod>[hlL])? # unused in python
3185 {1} # conversion type
3190 STR_FORMAT_TYPES
= 'diouxXeEfFgGcrs'
3193 def limit_length(s
, length
):
3194 """ Add ellipses to overly long strings """
3199 return s
[:length
- len(ELLIPSES
)] + ELLIPSES
3203 def version_tuple(v
):
3204 return tuple(int(e
) for e
in re
.split(r
'[-.]', v
))
3207 def is_outdated_version(version
, limit
, assume_new
=True):
3209 return not assume_new
3211 return version_tuple(version
) < version_tuple(limit
)
3213 return not assume_new
3216 def ytdl_is_updateable():
3217 """ Returns if yt-dlp can be updated with -U """
3219 from .update
import is_non_updateable
3221 return not is_non_updateable()
3224 def args_to_str(args
):
3225 # Get a short string representation for a subprocess command
3226 return ' '.join(compat_shlex_quote(a
) for a
in args
)
3229 def error_to_compat_str(err
):
3231 # On python 2 error byte string must be decoded with proper
3232 # encoding rather than ascii
3233 if sys
.version_info
[0] < 3:
3234 err_str
= err_str
.decode(preferredencoding())
3238 def mimetype2ext(mt
):
3242 mt
, _
, params
= mt
.partition(';')
3247 # Per RFC 3003, audio/mpeg can be .mp1, .mp2 or .mp3. Here use .mp3 as
3248 # it's the most popular one
3249 'audio/mpeg': 'mp3',
3250 'audio/x-wav': 'wav',
3252 'audio/wave': 'wav',
3255 ext
= FULL_MAP
.get(mt
)
3261 'smptett+xml': 'tt',
3265 'x-mp4-fragmented': 'mp4',
3266 'x-ms-sami': 'sami',
3269 'x-mpegurl': 'm3u8',
3270 'vnd.apple.mpegurl': 'm3u8',
3274 'vnd.ms-sstr+xml': 'ism',
3278 'filmstrip+json': 'fs',
3282 _
, _
, subtype
= mt
.rpartition('/')
3283 ext
= SUBTYPE_MAP
.get(subtype
.lower())
3294 _
, _
, suffix
= subtype
.partition('+')
3295 ext
= SUFFIX_MAP
.get(suffix
)
3299 return subtype
.replace('+', '.')
3302 def ext2mimetype(ext_or_url
):
3305 if '.' not in ext_or_url
:
3306 ext_or_url
= f
'file.{ext_or_url}'
3307 return mimetypes
.guess_type(ext_or_url
)[0]
3310 def parse_codecs(codecs_str
):
3311 # http://tools.ietf.org/html/rfc6381
3314 split_codecs
= list(filter(None, map(
3315 str.strip
, codecs_str
.strip().strip(',').split(','))))
3316 vcodec
, acodec
, tcodec
, hdr
= None, None, None, None
3317 for full_codec
in split_codecs
:
3318 parts
= full_codec
.split('.')
3319 codec
= parts
[0].replace('0', '')
3320 if codec
in ('avc1', 'avc2', 'avc3', 'avc4', 'vp9', 'vp8', 'hev1', 'hev2',
3321 'h263', 'h264', 'mp4v', 'hvc1', 'av1', 'theora', 'dvh1', 'dvhe'):
3323 vcodec
= '.'.join(parts
[:4]) if codec
in ('vp9', 'av1', 'hvc1') else full_codec
3324 if codec
in ('dvh1', 'dvhe'):
3326 elif codec
== 'av1' and len(parts
) > 3 and parts
[3] == '10':
3328 elif full_codec
.replace('0', '').startswith('vp9.2'):
3330 elif codec
in ('flac', 'mp4a', 'opus', 'vorbis', 'mp3', 'aac', 'ac-3', 'ec-3', 'eac3', 'dtsc', 'dtse', 'dtsh', 'dtsl'):
3333 elif codec
in ('stpp', 'wvtt',):
3337 write_string('WARNING: Unknown codec %s\n' % full_codec
, sys
.stderr
)
3338 if vcodec
or acodec
or tcodec
:
3340 'vcodec': vcodec
or 'none',
3341 'acodec': acodec
or 'none',
3342 'dynamic_range': hdr
,
3343 **({'tcodec': tcodec}
if tcodec
is not None else {}),
3345 elif len(split_codecs
) == 2:
3347 'vcodec': split_codecs
[0],
3348 'acodec': split_codecs
[1],
3353 def urlhandle_detect_ext(url_handle
):
3354 getheader
= url_handle
.headers
.get
3356 cd
= getheader('Content-Disposition')
3358 m
= re
.match(r
'attachment;\s*filename="(?P<filename>[^"]+)"', cd
)
3360 e
= determine_ext(m
.group('filename'), default_ext
=None)
3364 return mimetype2ext(getheader('Content-Type'))
3367 def encode_data_uri(data
, mime_type
):
3368 return 'data:%s;base64,%s' % (mime_type
, base64
.b64encode(data
).decode('ascii'))
3371 def age_restricted(content_limit
, age_limit
):
3372 """ Returns True iff the content should be blocked """
3374 if age_limit
is None: # No limit set
3376 if content_limit
is None:
3377 return False # Content available for everyone
3378 return age_limit
< content_limit
3381 def is_html(first_bytes
):
3382 """ Detect whether a file contains HTML by examining its first bytes. """
3385 (b
'\xef\xbb\xbf', 'utf-8'),
3386 (b
'\x00\x00\xfe\xff', 'utf-32-be'),
3387 (b
'\xff\xfe\x00\x00', 'utf-32-le'),
3388 (b
'\xff\xfe', 'utf-16-le'),
3389 (b
'\xfe\xff', 'utf-16-be'),
3391 for bom
, enc
in BOMS
:
3392 if first_bytes
.startswith(bom
):
3393 s
= first_bytes
[len(bom
):].decode(enc
, 'replace')
3396 s
= first_bytes
.decode('utf-8', 'replace')
3398 return re
.match(r
'^\s*<', s
)
3401 def determine_protocol(info_dict
):
3402 protocol
= info_dict
.get('protocol')
3403 if protocol
is not None:
3406 url
= sanitize_url(info_dict
['url'])
3407 if url
.startswith('rtmp'):
3409 elif url
.startswith('mms'):
3411 elif url
.startswith('rtsp'):
3414 ext
= determine_ext(url
)
3420 return compat_urllib_parse_urlparse(url
).scheme
3423 def render_table(header_row
, data
, delim
=False, extra_gap
=0, hide_empty
=False):
3424 """ Render a list of rows, each as a list of values.
3425 Text after a \t will be right aligned """
3427 return len(remove_terminal_sequences(string
).replace('\t', ''))
3429 def get_max_lens(table
):
3430 return [max(width(str(v
)) for v
in col
) for col
in zip(*table
)]
3432 def filter_using_list(row
, filterArray
):
3433 return [col
for (take
, col
) in zip(filterArray
, row
) if take
]
3436 max_lens
= get_max_lens(data
)
3437 header_row
= filter_using_list(header_row
, max_lens
)
3438 data
= [filter_using_list(row
, max_lens
) for row
in data
]
3440 table
= [header_row
] + data
3441 max_lens
= get_max_lens(table
)
3444 table
= [header_row
, [delim
* (ml
+ extra_gap
) for ml
in max_lens
]] + data
3445 table
[1][-1] = table
[1][-1][:-extra_gap
] # Remove extra_gap from end of delimiter
3447 for pos
, text
in enumerate(map(str, row
)):
3449 row
[pos
] = text
.replace('\t', ' ' * (max_lens
[pos
] - width(text
))) + ' ' * extra_gap
3451 row
[pos
] = text
+ ' ' * (max_lens
[pos
] - width(text
) + extra_gap
)
3452 ret
= '\n'.join(''.join(row
).rstrip() for row
in table
)
3456 def _match_one(filter_part
, dct
, incomplete
):
3457 # TODO: Generalize code with YoutubeDL._build_format_filter
3458 STRING_OPERATORS
= {
3459 '*=': operator
.contains
,
3460 '^=': lambda attr
, value
: attr
.startswith(value
),
3461 '$=': lambda attr
, value
: attr
.endswith(value
),
3462 '~=': lambda attr
, value
: re
.search(value
, attr
),
3464 COMPARISON_OPERATORS
= {
3466 '<=': operator
.le
, # "<=" must be defined above "<"
3473 operator_rex
= re
.compile(r
'''(?x)\s*
3475 \s*(?P<negation>!\s*)?(?P<op>%s)(?P<none_inclusive>\s*\?)?\s*
3477 (?P<quote>["\'])(?P
<quotedstrval
>.+?
)(?P
=quote
)|
3481 ''' % '|'.join(map(re.escape, COMPARISON_OPERATORS.keys())))
3482 m = operator_rex.search(filter_part)
3485 unnegated_op = COMPARISON_OPERATORS[m['op']]
3487 op = lambda attr, value: not unnegated_op(attr, value)
3490 comparison_value = m['quotedstrval'] or m['strval'] or m['intval']
3492 comparison_value = comparison_value.replace(r'\%s' % m['quote'], m['quote'])
3493 actual_value = dct.get(m['key'])
3494 numeric_comparison = None
3495 if isinstance(actual_value, compat_numeric_types):
3496 # If the original field is a string and matching comparisonvalue is
3497 # a number we should respect the origin of the original field
3498 # and process comparison value as a string (see
3499 # https://github.com/ytdl-org/youtube-dl/issues/11082)
3501 numeric_comparison = int(comparison_value)
3503 numeric_comparison = parse_filesize(comparison_value)
3504 if numeric_comparison is None:
3505 numeric_comparison = parse_filesize(f'{comparison_value}B')
3506 if numeric_comparison is None:
3507 numeric_comparison = parse_duration(comparison_value)
3508 if numeric_comparison is not None and m['op'] in STRING_OPERATORS:
3509 raise ValueError('Operator %s only supports string values!' % m['op'])
3510 if actual_value is None:
3511 return incomplete or m['none_inclusive']
3512 return op(actual_value, comparison_value if numeric_comparison is None else numeric_comparison)
3515 '': lambda v: (v is True) if isinstance(v, bool) else (v is not None),
3516 '!': lambda v: (v is False) if isinstance(v, bool) else (v is None),
3518 operator_rex = re.compile(r'''(?x
)\s
*
3519 (?P
<op
>%s)\s
*(?P
<key
>[a
-z_
]+)
3521 ''' % '|'.join(map(re.escape, UNARY_OPERATORS.keys())))
3522 m = operator_rex.search(filter_part)
3524 op = UNARY_OPERATORS[m.group('op')]
3525 actual_value = dct.get(m.group('key'))
3526 if incomplete and actual_value is None:
3528 return op(actual_value)
3530 raise ValueError('Invalid filter part %r' % filter_part)
3533 def match_str(filter_str, dct, incomplete=False):
3534 """ Filter a dictionary with a simple string syntax. Returns True (=passes filter) or false
3535 When incomplete, all conditions passes on missing fields
3538 _match_one(filter_part.replace(r'\&', '&'), dct, incomplete)
3539 for filter_part in re.split(r'(?<!\\)&', filter_str))
3542 def match_filter_func(filter_str):
3543 def _match_func(info_dict, *args, **kwargs):
3544 if match_str(filter_str, info_dict, *args, **kwargs):
3547 video_title = info_dict.get('title', info_dict.get('id', 'video'))
3548 return '%s does not pass filter %s, skipping ..' % (video_title, filter_str)
3552 def parse_dfxp_time_expr(time_expr):
3556 mobj = re.match(r'^(?P<time_offset>\d+(?:\.\d+)?)s?$', time_expr)
3558 return float(mobj.group('time_offset'))
3560 mobj = re.match(r'^(\d+):(\d\d):(\d\d(?:(?:\.|:)\d+)?)$', time_expr)
3562 return 3600 * int(mobj.group(1)) + 60 * int(mobj.group(2)) + float(mobj.group(3).replace(':', '.'))
3565 def srt_subtitles_timecode(seconds):
3566 return '%02d:%02d:%02d,%03d' % timetuple_from_msec(seconds * 1000)
3569 def ass_subtitles_timecode(seconds):
3570 time = timetuple_from_msec(seconds * 1000)
3571 return '%01d:%02d:%02d.%02d' % (*time[:-1], time.milliseconds / 10)
3574 def dfxp2srt(dfxp_data):
3576 @param dfxp_data A
bytes-like
object containing DFXP data
3577 @returns A
unicode object containing converted SRT data
3579 LEGACY_NAMESPACES = (
3580 (b'http://www.w3.org/ns/ttml', [
3581 b'http://www.w3.org/2004/11/ttaf1',
3582 b'http://www.w3.org/2006/04/ttaf1',
3583 b'http://www.w3.org/2006/10/ttaf1',
3585 (b'http://www.w3.org/ns/ttml#styling', [
3586 b'http://www.w3.org/ns/ttml#style',
3590 SUPPORTED_STYLING = [
3599 _x = functools.partial(xpath_with_ns, ns_map={
3600 'xml': 'http://www.w3.org/XML/1998/namespace',
3601 'ttml': 'http://www.w3.org/ns/ttml',
3602 'tts': 'http://www.w3.org/ns/ttml#styling',
3608 class TTMLPElementParser(object):
3610 _unclosed_elements = []
3611 _applied_styles = []
3613 def start(self, tag, attrib):
3614 if tag in (_x('ttml:br'), 'br'):
3617 unclosed_elements = []
3619 element_style_id = attrib.get('style')
3621 style.update(default_style)
3622 if element_style_id:
3623 style.update(styles.get(element_style_id, {}))
3624 for prop in SUPPORTED_STYLING:
3625 prop_val = attrib.get(_x('tts:' + prop))
3627 style[prop] = prop_val
3630 for k, v in sorted(style.items()):
3631 if self._applied_styles and self._applied_styles[-1].get(k) == v:
3634 font += ' color="%s"' % v
3635 elif k == 'fontSize':
3636 font += ' size="%s"' % v
3637 elif k == 'fontFamily':
3638 font += ' face="%s"' % v
3639 elif k == 'fontWeight' and v == 'bold':
3641 unclosed_elements.append('b')
3642 elif k == 'fontStyle' and v == 'italic':
3644 unclosed_elements.append('i')
3645 elif k == 'textDecoration' and v == 'underline':
3647 unclosed_elements.append('u')
3649 self._out += '<font' + font + '>'
3650 unclosed_elements.append('font')
3652 if self._applied_styles:
3653 applied_style.update(self._applied_styles[-1])
3654 applied_style.update(style)
3655 self._applied_styles.append(applied_style)
3656 self._unclosed_elements.append(unclosed_elements)
3659 if tag not in (_x('ttml:br'), 'br'):
3660 unclosed_elements = self._unclosed_elements.pop()
3661 for element in reversed(unclosed_elements):
3662 self._out += '</%s>' % element
3663 if unclosed_elements and self._applied_styles:
3664 self._applied_styles.pop()
3666 def data(self, data):
3670 return self._out.strip()
3672 def parse_node(node):
3673 target = TTMLPElementParser()
3674 parser = xml.etree.ElementTree.XMLParser(target=target)
3675 parser.feed(xml.etree.ElementTree.tostring(node))
3676 return parser.close()
3678 for k, v in LEGACY_NAMESPACES:
3680 dfxp_data = dfxp_data.replace(ns, k)
3682 dfxp = compat_etree_fromstring(dfxp_data)
3684 paras = dfxp.findall(_x('.//ttml:p')) or dfxp.findall('.//p')
3687 raise ValueError('Invalid dfxp/TTML subtitle')
3691 for style in dfxp.findall(_x('.//ttml:style')):
3692 style_id = style.get('id') or style.get(_x('xml:id'))
3695 parent_style_id = style.get('style')
3697 if parent_style_id not in styles:
3700 styles[style_id] = styles[parent_style_id].copy()
3701 for prop in SUPPORTED_STYLING:
3702 prop_val = style.get(_x('tts:' + prop))
3704 styles.setdefault(style_id, {})[prop] = prop_val
3710 for p in ('body', 'div'):
3711 ele = xpath_element(dfxp, [_x('.//ttml:' + p), './/' + p])
3714 style = styles.get(ele.get('style'))
3717 default_style.update(style)
3719 for para, index in zip(paras, itertools.count(1)):
3720 begin_time = parse_dfxp_time_expr(para.attrib.get('begin'))
3721 end_time = parse_dfxp_time_expr(para.attrib.get('end'))
3722 dur = parse_dfxp_time_expr(para.attrib.get('dur'))
3723 if begin_time is None:
3728 end_time = begin_time + dur
3729 out.append('%d\n%s --> %s\n%s\n\n' % (
3731 srt_subtitles_timecode(begin_time),
3732 srt_subtitles_timecode(end_time),
3738 def cli_option(params, command_option, param):
3739 param = params.get(param)
3741 param = compat_str(param)
3742 return [command_option, param] if param is not None else []
3745 def cli_bool_option(params, command_option, param, true_value='true', false_value='false', separator=None):
3746 param = params.get(param)
3749 assert isinstance(param, bool)
3751 return [command_option + separator + (true_value if param else false_value)]
3752 return [command_option, true_value if param else false_value]
3755 def cli_valueless_option(params, command_option, param, expected_value=True):
3756 param = params.get(param)
3757 return [command_option] if param == expected_value else []
3760 def cli_configuration_args(argdict, keys, default=[], use_compat=True):
3761 if isinstance(argdict, (list, tuple)): # for backward compatibility
3768 assert isinstance(argdict, dict)
3770 assert isinstance(keys, (list, tuple))
3771 for key_list in keys:
3772 arg_list = list(filter(
3773 lambda x: x is not None,
3774 [argdict.get(key.lower()) for key in variadic(key_list)]))
3776 return [arg for args in arg_list for arg in args]
3780 def _configuration_args(main_key, argdict, exe, keys=None, default=[], use_compat=True):
3781 main_key, exe = main_key.lower(), exe.lower()
3782 root_key = exe if main_key == exe else f'{main_key}+{exe}'
3783 keys = [f'{root_key}{k}' for k in (keys or [''])]
3784 if root_key in keys:
3786 keys.append((main_key, exe))
3787 keys.append('default')
3790 return cli_configuration_args(argdict, keys, default, use_compat)
3793 class ISO639Utils(object):
3794 # See http://www.loc.gov/standards/iso639-2/ISO-639-2_utf-8.txt
3853 'iw': 'heb', # Replaced by he in 1989 revision
3863 'in': 'ind', # Replaced by id in 1989 revision
3978 'ji': 'yid', # Replaced by yi in 1989 revision
3986 def short2long(cls, code):
3987 """Convert language code from ISO 639-1 to ISO 639-2/T"""
3988 return cls._lang_map.get(code[:2])
3991 def long2short(cls, code):
3992 """Convert language code from ISO 639-2/T to ISO 639-1"""
3993 for short_name, long_name in cls._lang_map.items():
3994 if long_name == code:
3998 class ISO3166Utils(object):
3999 # From http://data.okfn.org/data/core/country-list
4001 'AF': 'Afghanistan',
4002 'AX': 'Åland Islands',
4005 'AS': 'American Samoa',
4010 'AG': 'Antigua and Barbuda',
4027 'BO': 'Bolivia, Plurinational State of',
4028 'BQ': 'Bonaire, Sint Eustatius and Saba',
4029 'BA': 'Bosnia and Herzegovina',
4031 'BV': 'Bouvet Island',
4033 'IO': 'British Indian Ocean Territory',
4034 'BN': 'Brunei Darussalam',
4036 'BF': 'Burkina Faso',
4042 'KY': 'Cayman Islands',
4043 'CF': 'Central African Republic',
4047 'CX': 'Christmas Island',
4048 'CC': 'Cocos (Keeling) Islands',
4052 'CD': 'Congo, the Democratic Republic of the',
4053 'CK': 'Cook Islands',
4055 'CI': 'Côte d\'Ivoire',
4060 'CZ': 'Czech Republic',
4064 'DO': 'Dominican Republic',
4067 'SV': 'El Salvador',
4068 'GQ': 'Equatorial Guinea',
4072 'FK': 'Falkland Islands (Malvinas)',
4073 'FO': 'Faroe Islands',
4077 'GF': 'French Guiana',
4078 'PF': 'French Polynesia',
4079 'TF': 'French Southern Territories',
4094 'GW': 'Guinea-Bissau',
4097 'HM': 'Heard Island and McDonald Islands',
4098 'VA': 'Holy See (Vatican City State)',
4105 'IR': 'Iran, Islamic Republic of',
4108 'IM': 'Isle of Man',
4118 'KP': 'Korea, Democratic People\'s Republic of',
4119 'KR': 'Korea, Republic of',
4122 'LA': 'Lao People\'s Democratic Republic',
4128 'LI': 'Liechtenstein',
4132 'MK': 'Macedonia, the Former Yugoslav Republic of',
4139 'MH': 'Marshall Islands',
4145 'FM': 'Micronesia, Federated States of',
4146 'MD': 'Moldova, Republic of',
4157 'NL': 'Netherlands',
4158 'NC': 'New Caledonia',
4159 'NZ': 'New Zealand',
4164 'NF': 'Norfolk Island',
4165 'MP': 'Northern Mariana Islands',
4170 'PS': 'Palestine, State of',
4172 'PG': 'Papua New Guinea',
4175 'PH': 'Philippines',
4179 'PR': 'Puerto Rico',
4183 'RU': 'Russian Federation',
4185 'BL': 'Saint Barthélemy',
4186 'SH': 'Saint Helena, Ascension and Tristan da Cunha',
4187 'KN': 'Saint Kitts and Nevis',
4188 'LC': 'Saint Lucia',
4189 'MF': 'Saint Martin (French part)',
4190 'PM': 'Saint Pierre and Miquelon',
4191 'VC': 'Saint Vincent and the Grenadines',
4194 'ST': 'Sao Tome and Principe',
4195 'SA': 'Saudi Arabia',
4199 'SL': 'Sierra Leone',
4201 'SX': 'Sint Maarten (Dutch part)',
4204 'SB': 'Solomon Islands',
4206 'ZA': 'South Africa',
4207 'GS': 'South Georgia and the South Sandwich Islands',
4208 'SS': 'South Sudan',
4213 'SJ': 'Svalbard and Jan Mayen',
4216 'CH': 'Switzerland',
4217 'SY': 'Syrian Arab Republic',
4218 'TW': 'Taiwan, Province of China',
4220 'TZ': 'Tanzania, United Republic of',
4222 'TL': 'Timor-Leste',
4226 'TT': 'Trinidad and Tobago',
4229 'TM': 'Turkmenistan',
4230 'TC': 'Turks and Caicos Islands',
4234 'AE': 'United Arab Emirates',
4235 'GB': 'United Kingdom',
4236 'US': 'United States',
4237 'UM': 'United States Minor Outlying Islands',
4241 'VE': 'Venezuela, Bolivarian Republic of',
4243 'VG': 'Virgin Islands, British',
4244 'VI': 'Virgin Islands, U.S.',
4245 'WF': 'Wallis and Futuna',
4246 'EH': 'Western Sahara',
4253 def short2full(cls, code):
4254 """Convert an ISO 3166-2 country code to the corresponding full name"""
4255 return cls._country_map.get(code.upper())
4258 class GeoUtils(object):
4259 # Major IPv4 address blocks per country
4261 'AD': '46.172.224.0/19',
4262 'AE': '94.200.0.0/13',
4263 'AF': '149.54.0.0/17',
4264 'AG': '209.59.64.0/18',
4265 'AI': '204.14.248.0/21',
4266 'AL': '46.99.0.0/16',
4267 'AM': '46.70.0.0/15',
4268 'AO': '105.168.0.0/13',
4269 'AP': '182.50.184.0/21',
4270 'AQ': '23.154.160.0/24',
4271 'AR': '181.0.0.0/12',
4272 'AS': '202.70.112.0/20',
4273 'AT': '77.116.0.0/14',
4274 'AU': '1.128.0.0/11',
4275 'AW': '181.41.0.0/18',
4276 'AX': '185.217.4.0/22',
4277 'AZ': '5.197.0.0/16',
4278 'BA': '31.176.128.0/17',
4279 'BB': '65.48.128.0/17',
4280 'BD': '114.130.0.0/16',
4282 'BF': '102.178.0.0/15',
4283 'BG': '95.42.0.0/15',
4284 'BH': '37.131.0.0/17',
4285 'BI': '154.117.192.0/18',
4286 'BJ': '137.255.0.0/16',
4287 'BL': '185.212.72.0/23',
4288 'BM': '196.12.64.0/18',
4289 'BN': '156.31.0.0/16',
4290 'BO': '161.56.0.0/16',
4291 'BQ': '161.0.80.0/20',
4292 'BR': '191.128.0.0/12',
4293 'BS': '24.51.64.0/18',
4294 'BT': '119.2.96.0/19',
4295 'BW': '168.167.0.0/16',
4296 'BY': '178.120.0.0/13',
4297 'BZ': '179.42.192.0/18',
4298 'CA': '99.224.0.0/11',
4299 'CD': '41.243.0.0/16',
4300 'CF': '197.242.176.0/21',
4301 'CG': '160.113.0.0/16',
4302 'CH': '85.0.0.0/13',
4303 'CI': '102.136.0.0/14',
4304 'CK': '202.65.32.0/19',
4305 'CL': '152.172.0.0/14',
4306 'CM': '102.244.0.0/14',
4307 'CN': '36.128.0.0/10',
4308 'CO': '181.240.0.0/12',
4309 'CR': '201.192.0.0/12',
4310 'CU': '152.206.0.0/15',
4311 'CV': '165.90.96.0/19',
4312 'CW': '190.88.128.0/17',
4313 'CY': '31.153.0.0/16',
4314 'CZ': '88.100.0.0/14',
4316 'DJ': '197.241.0.0/17',
4317 'DK': '87.48.0.0/12',
4318 'DM': '192.243.48.0/20',
4319 'DO': '152.166.0.0/15',
4320 'DZ': '41.96.0.0/12',
4321 'EC': '186.68.0.0/15',
4322 'EE': '90.190.0.0/15',
4323 'EG': '156.160.0.0/11',
4324 'ER': '196.200.96.0/20',
4325 'ES': '88.0.0.0/11',
4326 'ET': '196.188.0.0/14',
4327 'EU': '2.16.0.0/13',
4328 'FI': '91.152.0.0/13',
4329 'FJ': '144.120.0.0/16',
4330 'FK': '80.73.208.0/21',
4331 'FM': '119.252.112.0/20',
4332 'FO': '88.85.32.0/19',
4334 'GA': '41.158.0.0/15',
4336 'GD': '74.122.88.0/21',
4337 'GE': '31.146.0.0/16',
4338 'GF': '161.22.64.0/18',
4339 'GG': '62.68.160.0/19',
4340 'GH': '154.160.0.0/12',
4341 'GI': '95.164.0.0/16',
4342 'GL': '88.83.0.0/19',
4343 'GM': '160.182.0.0/15',
4344 'GN': '197.149.192.0/18',
4345 'GP': '104.250.0.0/19',
4346 'GQ': '105.235.224.0/20',
4347 'GR': '94.64.0.0/13',
4348 'GT': '168.234.0.0/16',
4349 'GU': '168.123.0.0/16',
4350 'GW': '197.214.80.0/20',
4351 'GY': '181.41.64.0/18',
4352 'HK': '113.252.0.0/14',
4353 'HN': '181.210.0.0/16',
4354 'HR': '93.136.0.0/13',
4355 'HT': '148.102.128.0/17',
4356 'HU': '84.0.0.0/14',
4357 'ID': '39.192.0.0/10',
4358 'IE': '87.32.0.0/12',
4359 'IL': '79.176.0.0/13',
4360 'IM': '5.62.80.0/20',
4361 'IN': '117.192.0.0/10',
4362 'IO': '203.83.48.0/21',
4363 'IQ': '37.236.0.0/14',
4364 'IR': '2.176.0.0/12',
4365 'IS': '82.221.0.0/16',
4366 'IT': '79.0.0.0/10',
4367 'JE': '87.244.64.0/18',
4368 'JM': '72.27.0.0/17',
4369 'JO': '176.29.0.0/16',
4370 'JP': '133.0.0.0/8',
4371 'KE': '105.48.0.0/12',
4372 'KG': '158.181.128.0/17',
4373 'KH': '36.37.128.0/17',
4374 'KI': '103.25.140.0/22',
4375 'KM': '197.255.224.0/20',
4376 'KN': '198.167.192.0/19',
4377 'KP': '175.45.176.0/22',
4378 'KR': '175.192.0.0/10',
4379 'KW': '37.36.0.0/14',
4380 'KY': '64.96.0.0/15',
4381 'KZ': '2.72.0.0/13',
4382 'LA': '115.84.64.0/18',
4383 'LB': '178.135.0.0/16',
4384 'LC': '24.92.144.0/20',
4385 'LI': '82.117.0.0/19',
4386 'LK': '112.134.0.0/15',
4387 'LR': '102.183.0.0/16',
4388 'LS': '129.232.0.0/17',
4389 'LT': '78.56.0.0/13',
4390 'LU': '188.42.0.0/16',
4391 'LV': '46.109.0.0/16',
4392 'LY': '41.252.0.0/14',
4393 'MA': '105.128.0.0/11',
4394 'MC': '88.209.64.0/18',
4395 'MD': '37.246.0.0/16',
4396 'ME': '178.175.0.0/17',
4397 'MF': '74.112.232.0/21',
4398 'MG': '154.126.0.0/17',
4399 'MH': '117.103.88.0/21',
4400 'MK': '77.28.0.0/15',
4401 'ML': '154.118.128.0/18',
4402 'MM': '37.111.0.0/17',
4403 'MN': '49.0.128.0/17',
4404 'MO': '60.246.0.0/16',
4405 'MP': '202.88.64.0/20',
4406 'MQ': '109.203.224.0/19',
4407 'MR': '41.188.64.0/18',
4408 'MS': '208.90.112.0/22',
4409 'MT': '46.11.0.0/16',
4410 'MU': '105.16.0.0/12',
4411 'MV': '27.114.128.0/18',
4412 'MW': '102.70.0.0/15',
4413 'MX': '187.192.0.0/11',
4414 'MY': '175.136.0.0/13',
4415 'MZ': '197.218.0.0/15',
4416 'NA': '41.182.0.0/16',
4417 'NC': '101.101.0.0/18',
4418 'NE': '197.214.0.0/18',
4419 'NF': '203.17.240.0/22',
4420 'NG': '105.112.0.0/12',
4421 'NI': '186.76.0.0/15',
4422 'NL': '145.96.0.0/11',
4423 'NO': '84.208.0.0/13',
4424 'NP': '36.252.0.0/15',
4425 'NR': '203.98.224.0/19',
4426 'NU': '49.156.48.0/22',
4427 'NZ': '49.224.0.0/14',
4428 'OM': '5.36.0.0/15',
4429 'PA': '186.72.0.0/15',
4430 'PE': '186.160.0.0/14',
4431 'PF': '123.50.64.0/18',
4432 'PG': '124.240.192.0/19',
4433 'PH': '49.144.0.0/13',
4434 'PK': '39.32.0.0/11',
4435 'PL': '83.0.0.0/11',
4436 'PM': '70.36.0.0/20',
4437 'PR': '66.50.0.0/16',
4438 'PS': '188.161.0.0/16',
4439 'PT': '85.240.0.0/13',
4440 'PW': '202.124.224.0/20',
4441 'PY': '181.120.0.0/14',
4442 'QA': '37.210.0.0/15',
4443 'RE': '102.35.0.0/16',
4444 'RO': '79.112.0.0/13',
4445 'RS': '93.86.0.0/15',
4446 'RU': '5.136.0.0/13',
4447 'RW': '41.186.0.0/16',
4448 'SA': '188.48.0.0/13',
4449 'SB': '202.1.160.0/19',
4450 'SC': '154.192.0.0/11',
4451 'SD': '102.120.0.0/13',
4452 'SE': '78.64.0.0/12',
4453 'SG': '8.128.0.0/10',
4454 'SI': '188.196.0.0/14',
4455 'SK': '78.98.0.0/15',
4456 'SL': '102.143.0.0/17',
4457 'SM': '89.186.32.0/19',
4458 'SN': '41.82.0.0/15',
4459 'SO': '154.115.192.0/18',
4460 'SR': '186.179.128.0/17',
4461 'SS': '105.235.208.0/21',
4462 'ST': '197.159.160.0/19',
4463 'SV': '168.243.0.0/16',
4464 'SX': '190.102.0.0/20',
4466 'SZ': '41.84.224.0/19',
4467 'TC': '65.255.48.0/20',
4468 'TD': '154.68.128.0/19',
4469 'TG': '196.168.0.0/14',
4470 'TH': '171.96.0.0/13',
4471 'TJ': '85.9.128.0/18',
4472 'TK': '27.96.24.0/21',
4473 'TL': '180.189.160.0/20',
4474 'TM': '95.85.96.0/19',
4475 'TN': '197.0.0.0/11',
4476 'TO': '175.176.144.0/21',
4477 'TR': '78.160.0.0/11',
4478 'TT': '186.44.0.0/15',
4479 'TV': '202.2.96.0/19',
4480 'TW': '120.96.0.0/11',
4481 'TZ': '156.156.0.0/14',
4482 'UA': '37.52.0.0/14',
4483 'UG': '102.80.0.0/13',
4485 'UY': '167.56.0.0/13',
4486 'UZ': '84.54.64.0/18',
4487 'VA': '212.77.0.0/19',
4488 'VC': '207.191.240.0/21',
4489 'VE': '186.88.0.0/13',
4490 'VG': '66.81.192.0/20',
4491 'VI': '146.226.0.0/16',
4492 'VN': '14.160.0.0/11',
4493 'VU': '202.80.32.0/20',
4494 'WF': '117.20.32.0/21',
4495 'WS': '202.4.32.0/19',
4496 'YE': '134.35.0.0/16',
4497 'YT': '41.242.116.0/22',
4498 'ZA': '41.0.0.0/11',
4499 'ZM': '102.144.0.0/13',
4500 'ZW': '102.177.192.0/18',
4504 def random_ipv4(cls, code_or_block):
4505 if len(code_or_block) == 2:
4506 block = cls._country_ip_map.get(code_or_block.upper())
4510 block = code_or_block
4511 addr, preflen = block.split('/')
4512 addr_min = compat_struct_unpack('!L', socket.inet_aton(addr))[0]
4513 addr_max = addr_min | (0xffffffff >> int(preflen))
4514 return compat_str(socket.inet_ntoa(
4515 compat_struct_pack('!L', random.randint(addr_min, addr_max))))
4518 class PerRequestProxyHandler(compat_urllib_request.ProxyHandler):
4519 def __init__(self, proxies=None):
4520 # Set default handlers
4521 for type in ('http', 'https'):
4522 setattr(self, '%s_open' % type,
4523 lambda r, proxy='__noproxy__', type=type, meth=self.proxy_open:
4524 meth(r, proxy, type))
4525 compat_urllib_request.ProxyHandler.__init__(self, proxies)
4527 def proxy_open(self, req, proxy, type):
4528 req_proxy = req.headers.get('Ytdl-request-proxy')
4529 if req_proxy is not None:
4531 del req.headers['Ytdl-request-proxy']
4533 if proxy == '__noproxy__':
4534 return None # No Proxy
4535 if compat_urlparse.urlparse(proxy).scheme.lower() in ('socks', 'socks4', 'socks4a', 'socks5'):
4536 req.add_header('Ytdl-socks-proxy', proxy)
4537 # yt-dlp's http/https handlers do wrapping the socket with socks
4539 return compat_urllib_request.ProxyHandler.proxy_open(
4540 self, req, proxy, type)
4543 # Both long_to_bytes and bytes_to_long are adapted from PyCrypto, which is
4544 # released into Public Domain
4545 # https://github.com/dlitz/pycrypto/blob/master/lib/Crypto/Util/number.py#L387
4547 def long_to_bytes(n, blocksize=0):
4548 """long_to_bytes(n:long, blocksize:int) : string
4549 Convert a long integer to a byte string.
4551 If optional blocksize is given and greater than zero, pad the front of the
4552 byte string with binary zeros so that the length is a multiple of
4555 # after much testing, this algorithm was deemed to be the fastest
4559 s = compat_struct_pack('>I', n & 0xffffffff) + s
4561 # strip off leading zeros
4562 for i in range(len(s)):
4563 if s[i] != b'\000'[0]:
4566 # only happens when n == 0
4570 # add back some pad bytes. this could be done more efficiently w.r.t. the
4571 # de-padding being done above, but sigh...
4572 if blocksize > 0 and len(s) % blocksize:
4573 s = (blocksize - len(s) % blocksize) * b'\000' + s
4577 def bytes_to_long(s):
4578 """bytes_to_long(string) : long
4579 Convert a byte string to a long integer.
4581 This is (essentially) the inverse of long_to_bytes().
4586 extra = (4 - length % 4)
4587 s = b'\000' * extra + s
4588 length = length + extra
4589 for i in range(0, length, 4):
4590 acc = (acc << 32) + compat_struct_unpack('>I', s[i:i + 4])[0]
4594 def ohdave_rsa_encrypt(data, exponent, modulus):
4596 Implement OHDave
's RSA algorithm. See http://www.ohdave.com/rsa/
4599 data: data to encrypt, bytes-like object
4600 exponent, modulus: parameter e and N of RSA algorithm, both integer
4601 Output: hex string of encrypted data
4603 Limitation: supports one block encryption only
4606 payload = int(binascii.hexlify(data[::-1]), 16)
4607 encrypted = pow(payload, exponent, modulus)
4608 return '%x' % encrypted
4611 def pkcs1pad(data, length):
4613 Padding input data with PKCS#1 scheme
4615 @param {int[]} data input data
4616 @param {int} length target length
4617 @returns {int[]} padded data
4619 if len(data) > length - 11:
4620 raise ValueError('Input data too
long for PKCS
#1 padding')
4622 pseudo_random
= [random
.randint(0, 254) for _
in range(length
- len(data
) - 3)]
4623 return [0, 2] + pseudo_random
+ [0] + data
4626 def encode_base_n(num
, n
, table
=None):
4627 FULL_TABLE
= '0123456789abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ'
4629 table
= FULL_TABLE
[:n
]
4632 raise ValueError('base %d exceeds table length %d' % (n
, len(table
)))
4639 ret
= table
[num
% n
] + ret
4644 def decode_packed_codes(code
):
4645 mobj
= re
.search(PACKED_CODES_RE
, code
)
4646 obfuscated_code
, base
, count
, symbols
= mobj
.groups()
4649 symbols
= symbols
.split('|')
4654 base_n_count
= encode_base_n(count
, base
)
4655 symbol_table
[base_n_count
] = symbols
[count
] or base_n_count
4658 r
'\b(\w+)\b', lambda mobj
: symbol_table
[mobj
.group(0)],
4662 def caesar(s
, alphabet
, shift
):
4667 alphabet
[(alphabet
.index(c
) + shift
) % l
] if c
in alphabet
else c
4672 return caesar(s
, r
'''!"#$%&'()*+,-./0123456789:;<=>?@ABCDEFGHIJKLMNOPQRSTUVWXYZ[\]^_`abcdefghijklmnopqrstuvwxyz{|}~''', 47)
4675 def parse_m3u8_attributes(attrib
):
4677 for (key
, val
) in re
.findall(r
'(?P<key>[A-Z0-9-]+)=(?P<val>"[^"]+"|[^",]+)(?:,|$)', attrib
):
4678 if val
.startswith('"'):
4684 def urshift(val
, n
):
4685 return val
>> n
if val
>= 0 else (val
+ 0x100000000) >> n
4688 # Based on png2str() written by @gdkchan and improved by @yokrysty
4689 # Originally posted at https://github.com/ytdl-org/youtube-dl/issues/9706
4690 def decode_png(png_data
):
4691 # Reference: https://www.w3.org/TR/PNG/
4692 header
= png_data
[8:]
4694 if png_data
[:8] != b
'\x89PNG\x0d\x0a\x1a\x0a' or header
[4:8] != b
'IHDR':
4695 raise IOError('Not a valid PNG file.')
4697 int_map
= {1: '>B', 2: '>H', 4: '>I'}
4698 unpack_integer
= lambda x
: compat_struct_unpack(int_map
[len(x
)], x
)[0]
4703 length
= unpack_integer(header
[:4])
4706 chunk_type
= header
[:4]
4709 chunk_data
= header
[:length
]
4710 header
= header
[length
:]
4712 header
= header
[4:] # Skip CRC
4720 ihdr
= chunks
[0]['data']
4722 width
= unpack_integer(ihdr
[:4])
4723 height
= unpack_integer(ihdr
[4:8])
4727 for chunk
in chunks
:
4728 if chunk
['type'] == b
'IDAT':
4729 idat
+= chunk
['data']
4732 raise IOError('Unable to read PNG data.')
4734 decompressed_data
= bytearray(zlib
.decompress(idat
))
4739 def _get_pixel(idx
):
4744 for y
in range(height
):
4745 basePos
= y
* (1 + stride
)
4746 filter_type
= decompressed_data
[basePos
]
4750 pixels
.append(current_row
)
4752 for x
in range(stride
):
4753 color
= decompressed_data
[1 + basePos
+ x
]
4754 basex
= y
* stride
+ x
4759 left
= _get_pixel(basex
- 3)
4761 up
= _get_pixel(basex
- stride
)
4763 if filter_type
== 1: # Sub
4764 color
= (color
+ left
) & 0xff
4765 elif filter_type
== 2: # Up
4766 color
= (color
+ up
) & 0xff
4767 elif filter_type
== 3: # Average
4768 color
= (color
+ ((left
+ up
) >> 1)) & 0xff
4769 elif filter_type
== 4: # Paeth
4775 c
= _get_pixel(basex
- stride
- 3)
4783 if pa
<= pb
and pa
<= pc
:
4784 color
= (color
+ a
) & 0xff
4786 color
= (color
+ b
) & 0xff
4788 color
= (color
+ c
) & 0xff
4790 current_row
.append(color
)
4792 return width
, height
, pixels
4795 def write_xattr(path
, key
, value
):
4796 # This mess below finds the best xattr tool for the job
4798 # try the pyxattr module...
4801 if hasattr(xattr
, 'set'): # pyxattr
4802 # Unicode arguments are not supported in python-pyxattr until
4804 # See https://github.com/ytdl-org/youtube-dl/issues/5498
4805 pyxattr_required_version
= '0.5.0'
4806 if version_tuple(xattr
.__version
__) < version_tuple(pyxattr_required_version
):
4807 # TODO: fallback to CLI tools
4808 raise XAttrUnavailableError(
4809 'python-pyxattr is detected but is too old. '
4810 'yt-dlp requires %s or above while your version is %s. '
4811 'Falling back to other xattr implementations' % (
4812 pyxattr_required_version
, xattr
.__version
__))
4814 setxattr
= xattr
.set
4816 setxattr
= xattr
.setxattr
4819 setxattr(path
, key
, value
)
4820 except EnvironmentError as e
:
4821 raise XAttrMetadataError(e
.errno
, e
.strerror
)
4824 if compat_os_name
== 'nt':
4825 # Write xattrs to NTFS Alternate Data Streams:
4826 # http://en.wikipedia.org/wiki/NTFS#Alternate_data_streams_.28ADS.29
4827 assert ':' not in key
4828 assert os
.path
.exists(path
)
4830 ads_fn
= path
+ ':' + key
4832 with open(ads_fn
, 'wb') as f
:
4834 except EnvironmentError as e
:
4835 raise XAttrMetadataError(e
.errno
, e
.strerror
)
4837 user_has_setfattr
= check_executable('setfattr', ['--version'])
4838 user_has_xattr
= check_executable('xattr', ['-h'])
4840 if user_has_setfattr
or user_has_xattr
:
4842 value
= value
.decode('utf-8')
4843 if user_has_setfattr
:
4844 executable
= 'setfattr'
4845 opts
= ['-n', key
, '-v', value
]
4846 elif user_has_xattr
:
4847 executable
= 'xattr'
4848 opts
= ['-w', key
, value
]
4850 cmd
= ([encodeFilename(executable
, True)]
4851 + [encodeArgument(o
) for o
in opts
]
4852 + [encodeFilename(path
, True)])
4856 cmd
, stdout
=subprocess
.PIPE
, stderr
=subprocess
.PIPE
, stdin
=subprocess
.PIPE
)
4857 except EnvironmentError as e
:
4858 raise XAttrMetadataError(e
.errno
, e
.strerror
)
4859 stdout
, stderr
= p
.communicate_or_kill()
4860 stderr
= stderr
.decode('utf-8', 'replace')
4861 if p
.returncode
!= 0:
4862 raise XAttrMetadataError(p
.returncode
, stderr
)
4865 # On Unix, and can't find pyxattr, setfattr, or xattr.
4866 if sys
.platform
.startswith('linux'):
4867 raise XAttrUnavailableError(
4868 "Couldn't find a tool to set the xattrs. "
4869 "Install either the python 'pyxattr' or 'xattr' "
4870 "modules, or the GNU 'attr' package "
4871 "(which contains the 'setfattr' tool).")
4873 raise XAttrUnavailableError(
4874 "Couldn't find a tool to set the xattrs. "
4875 "Install either the python 'xattr' module, "
4876 "or the 'xattr' binary.")
4879 def random_birthday(year_field
, month_field
, day_field
):
4880 start_date
= datetime
.date(1950, 1, 1)
4881 end_date
= datetime
.date(1995, 12, 31)
4882 offset
= random
.randint(0, (end_date
- start_date
).days
)
4883 random_date
= start_date
+ datetime
.timedelta(offset
)
4885 year_field
: str(random_date
.year
),
4886 month_field
: str(random_date
.month
),
4887 day_field
: str(random_date
.day
),
4891 # Templates for internet shortcut files, which are plain text files.
4892 DOT_URL_LINK_TEMPLATE
= '''
4897 DOT_WEBLOC_LINK_TEMPLATE
= '''
4898 <?xml version="1.0" encoding="UTF-8"?>
4899 <!DOCTYPE plist PUBLIC "-//Apple//DTD PLIST 1.0//EN" "http://www.apple.com/DTDs/PropertyList-1.0.dtd">
4900 <plist version="1.0">
4903 \t<string>%(url)s</string>
4908 DOT_DESKTOP_LINK_TEMPLATE
= '''
4918 'url': DOT_URL_LINK_TEMPLATE
,
4919 'desktop': DOT_DESKTOP_LINK_TEMPLATE
,
4920 'webloc': DOT_WEBLOC_LINK_TEMPLATE
,
4924 def iri_to_uri(iri
):
4926 Converts an IRI (Internationalized Resource Identifier, allowing Unicode characters) to a URI (Uniform Resource Identifier, ASCII-only).
4928 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.
4931 iri_parts
= compat_urllib_parse_urlparse(iri
)
4933 if '[' in iri_parts
.netloc
:
4934 raise ValueError('IPv6 URIs are not, yet, supported.')
4935 # Querying `.netloc`, when there's only one bracket, also raises a ValueError.
4937 # 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.
4940 if iri_parts
.username
:
4941 net_location
+= compat_urllib_parse_quote(iri_parts
.username
, safe
=r
"!$%&'()*+,~")
4942 if iri_parts
.password
is not None:
4943 net_location
+= ':' + compat_urllib_parse_quote(iri_parts
.password
, safe
=r
"!$%&'()*+,~")
4946 net_location
+= iri_parts
.hostname
.encode('idna').decode('utf-8') # Punycode for Unicode hostnames.
4947 # The 'idna' encoding produces ASCII text.
4948 if iri_parts
.port
is not None and iri_parts
.port
!= 80:
4949 net_location
+= ':' + str(iri_parts
.port
)
4951 return compat_urllib_parse_urlunparse(
4955 compat_urllib_parse_quote_plus(iri_parts
.path
, safe
=r
"!$%&'()*+,/:;=@|~"),
4957 # Unsure about the `safe` argument, since this is a legacy way of handling parameters.
4958 compat_urllib_parse_quote_plus(iri_parts
.params
, safe
=r
"!$%&'()*+,/:;=@|~"),
4960 # Not totally sure about the `safe` argument, since the source does not explicitly mention the query URI component.
4961 compat_urllib_parse_quote_plus(iri_parts
.query
, safe
=r
"!$%&'()*+,/:;=?@{|}~"),
4963 compat_urllib_parse_quote_plus(iri_parts
.fragment
, safe
=r
"!#$%&'()*+,/:;=?@{|}~")))
4965 # Source for `safe` arguments: https://url.spec.whatwg.org/#percent-encoded-bytes.
4968 def to_high_limit_path(path
):
4969 if sys
.platform
in ['win32', 'cygwin']:
4970 # Work around MAX_PATH limitation on Windows. The maximum allowed length for the individual path segments may still be quite limited.
4971 return r
'\\?\ '.rstrip() + os
.path
.abspath(path
)
4976 def format_field(obj
, field
=None, template
='%s', ignore
=(None, ''), default
='', func
=None):
4977 val
= traverse_obj(obj
, *variadic(field
))
4980 return template
% (func(val
) if func
else val
)
4983 def clean_podcast_url(url
):
4984 return re
.sub(r
'''(?x)
4988 media\.blubrry\.com| # https://create.blubrry.com/resources/podcast-media-download-statistics/getting-started/
4991 (?:dts|www)\.podtrac\.com/(?:pts/)?redirect\.[0-9a-z]{3,4}| # http://analytics.podtrac.com/how-to-measure
4994 cn\.co| # https://podcorn.com/analytics-prefix/
4995 st\.fm # https://podsights.com/docs/
5000 _HEX_TABLE
= '0123456789abcdef'
5003 def random_uuidv4():
5004 return re
.sub(r
'[xy]', lambda x
: _HEX_TABLE
[random
.randint(0, 15)], 'xxxxxxxx-xxxx-4xxx-yxxx-xxxxxxxxxxxx')
5007 def make_dir(path
, to_screen
=None):
5009 dn
= os
.path
.dirname(path
)
5010 if dn
and not os
.path
.exists(dn
):
5013 except (OSError, IOError) as err
:
5014 if callable(to_screen
) is not None:
5015 to_screen('unable to create directory ' + error_to_compat_str(err
))
5019 def get_executable_path():
5020 from zipimport
import zipimporter
5021 if hasattr(sys
, 'frozen'): # Running from PyInstaller
5022 path
= os
.path
.dirname(sys
.executable
)
5023 elif isinstance(globals().get('__loader__'), zipimporter
): # Running from ZIP
5024 path
= os
.path
.join(os
.path
.dirname(__file__
), '../..')
5026 path
= os
.path
.join(os
.path
.dirname(__file__
), '..')
5027 return os
.path
.abspath(path
)
5030 def load_plugins(name
, suffix
, namespace
):
5033 plugins_spec
= importlib
.util
.spec_from_file_location(
5034 name
, os
.path
.join(get_executable_path(), 'ytdlp_plugins', name
, '__init__.py'))
5035 plugins
= importlib
.util
.module_from_spec(plugins_spec
)
5036 sys
.modules
[plugins_spec
.name
] = plugins
5037 plugins_spec
.loader
.exec_module(plugins
)
5038 for name
in dir(plugins
):
5039 if name
in namespace
:
5041 if not name
.endswith(suffix
):
5043 klass
= getattr(plugins
, name
)
5044 classes
[name
] = namespace
[name
] = klass
5045 except FileNotFoundError
:
5051 obj
, *path_list
, default
=None, expected_type
=None, get_all
=True,
5052 casesense
=True, is_user_input
=False, traverse_string
=False):
5053 ''' Traverse nested list/dict/tuple
5054 @param path_list A list of paths which are checked one by one.
5055 Each path is a list of keys where each key is a string,
5056 a function, a tuple of strings/None or "...".
5057 When a fuction is given, it takes the key as argument and
5058 returns whether the key matches or not. When a tuple is given,
5059 all the keys given in the tuple are traversed, and
5060 "..." traverses all the keys in the object
5061 "None" returns the object without traversal
5062 @param default Default value to return
5063 @param expected_type Only accept final value of this type (Can also be any callable)
5064 @param get_all Return all the values obtained from a path or only the first one
5065 @param casesense Whether to consider dictionary keys as case sensitive
5066 @param is_user_input Whether the keys are generated from user input. If True,
5067 strings are converted to int/slice if necessary
5068 @param traverse_string Whether to traverse inside strings. If True, any
5069 non-compatible object will also be converted into a string
5073 _lower
= lambda k
: (k
.lower() if isinstance(k
, str) else k
)
5074 path_list
= (map(_lower
, variadic(path
)) for path
in path_list
)
5076 def _traverse_obj(obj
, path
, _current_depth
=0):
5078 path
= tuple(variadic(path
))
5079 for i
, key
in enumerate(path
):
5080 if None in (key
, obj
):
5082 if isinstance(key
, (list, tuple)):
5083 obj
= [_traverse_obj(obj
, sub_key
, _current_depth
) for sub_key
in key
]
5086 obj
= (obj
.values() if isinstance(obj
, dict)
5087 else obj
if isinstance(obj
, (list, tuple, LazyList
))
5088 else str(obj
) if traverse_string
else [])
5090 depth
= max(depth
, _current_depth
)
5091 return [_traverse_obj(inner_obj
, path
[i
+ 1:], _current_depth
) for inner_obj
in obj
]
5093 if isinstance(obj
, (list, tuple, LazyList
)):
5094 obj
= enumerate(obj
)
5095 elif isinstance(obj
, dict):
5098 if not traverse_string
:
5102 depth
= max(depth
, _current_depth
)
5103 return [_traverse_obj(v
, path
[i
+ 1:], _current_depth
) for k
, v
in obj
if key(k
)]
5104 elif isinstance(obj
, dict) and not (is_user_input
and key
== ':'):
5105 obj
= (obj
.get(key
) if casesense
or (key
in obj
)
5106 else next((v
for k
, v
in obj
.items() if _lower(k
) == key
), None))
5109 key
= (int_or_none(key
) if ':' not in key
5110 else slice(*map(int_or_none
, key
.split(':'))))
5111 if key
== slice(None):
5112 return _traverse_obj(obj
, (..., *path
[i
+ 1:]), _current_depth
)
5113 if not isinstance(key
, (int, slice)):
5115 if not isinstance(obj
, (list, tuple, LazyList
)):
5116 if not traverse_string
:
5125 if isinstance(expected_type
, type):
5126 type_test
= lambda val
: val
if isinstance(val
, expected_type
) else None
5127 elif expected_type
is not None:
5128 type_test
= expected_type
5130 type_test
= lambda val
: val
5132 for path
in path_list
:
5134 val
= _traverse_obj(obj
, path
)
5137 for _
in range(depth
- 1):
5138 val
= itertools
.chain
.from_iterable(v
for v
in val
if v
is not None)
5139 val
= [v
for v
in map(type_test
, val
) if v
is not None]
5141 return val
if get_all
else val
[0]
5143 val
= type_test(val
)
5149 def traverse_dict(dictn
, keys
, casesense
=True):
5150 write_string('DeprecationWarning: yt_dlp.utils.traverse_dict is deprecated '
5151 'and may be removed in a future version. Use yt_dlp.utils.traverse_obj instead')
5152 return traverse_obj(dictn
, keys
, casesense
=casesense
, is_user_input
=True, traverse_string
=True)
5155 def variadic(x
, allowed_types
=(str, bytes, dict)):
5156 return x
if isinstance(x
, collections
.abc
.Iterable
) and not isinstance(x
, allowed_types
) else (x
,)
5159 # create a JSON Web Signature (jws) with HS256 algorithm
5160 # the resulting format is in JWS Compact Serialization
5161 # implemented following JWT https://www.rfc-editor.org/rfc/rfc7519.html
5162 # implemented following JWS https://www.rfc-editor.org/rfc/rfc7515.html
5163 def jwt_encode_hs256(payload_data
, key
, headers
={}):
5169 header_data
.update(headers
)
5170 header_b64
= base64
.b64encode(json
.dumps(header_data
).encode('utf-8'))
5171 payload_b64
= base64
.b64encode(json
.dumps(payload_data
).encode('utf-8'))
5172 h
= hmac
.new(key
.encode('utf-8'), header_b64
+ b
'.' + payload_b64
, hashlib
.sha256
)
5173 signature_b64
= base64
.b64encode(h
.digest())
5174 token
= header_b64
+ b
'.' + payload_b64
+ b
'.' + signature_b64
5178 # can be extended in future to verify the signature and parse header and return the algorithm used if it's not HS256
5179 def jwt_decode_hs256(jwt
):
5180 header_b64
, payload_b64
, signature_b64
= jwt
.split('.')
5181 payload_data
= json
.loads(base64
.urlsafe_b64decode(payload_b64
))
5185 def supports_terminal_sequences(stream
):
5186 if compat_os_name
== 'nt':
5187 from .compat
import WINDOWS_VT_MODE
# Must be imported locally
5188 if not WINDOWS_VT_MODE
or get_windows_version() < (10, 0, 10586):
5190 elif not os
.getenv('TERM'):
5193 return stream
.isatty()
5194 except BaseException
:
5198 _terminal_sequences_re
= re
.compile('\033\\[[^m]+m')
5201 def remove_terminal_sequences(string
):
5202 return _terminal_sequences_re
.sub('', string
)
5205 def number_of_digits(number
):
5206 return len('%d' % number
)
5209 def join_nonempty(*values
, delim
='-', from_dict
=None):
5210 if from_dict
is not None:
5211 values
= map(from_dict
.get
, values
)
5212 return delim
.join(map(str, filter(None, values
)))
5218 __initialized
= False
5220 def __init__(self
, parser
, label
=None):
5221 self
._parser
, self
.label
= parser
, label
5222 self
._loaded
_paths
, self
.configs
= set(), []
5224 def init(self
, args
=None, filename
=None):
5225 assert not self
.__initialized
5227 location
= os
.path
.realpath(filename
)
5228 if location
in self
._loaded
_paths
:
5230 self
._loaded
_paths
.add(location
)
5232 self
.__initialized
= True
5233 self
.own_args
, self
.filename
= args
, filename
5234 for location
in self
._parser
.parse_args(args
)[0].config_locations
or []:
5235 location
= compat_expanduser(location
)
5236 if os
.path
.isdir(location
):
5237 location
= os
.path
.join(location
, 'yt-dlp.conf')
5238 if not os
.path
.exists(location
):
5239 self
._parser
.error(f
'config location {location} does not exist')
5240 self
.append_config(self
.read_file(location
), location
)
5244 label
= join_nonempty(
5245 self
.label
, 'config', f
'"{self.filename}"' if self
.filename
else '',
5247 return join_nonempty(
5248 self
.own_args
is not None and f
'{label[0].upper()}{label[1:]}: {self.hide_login_info(self.own_args)}',
5249 *(f
'\n{c}'.replace('\n', '\n| ')[1:] for c
in self
.configs
),
5253 def read_file(filename
, default
=[]):
5255 optionf
= open(filename
)
5257 return default
# silently skip if file is not present
5259 # FIXME: https://github.com/ytdl-org/youtube-dl/commit/dfe5fa49aed02cf36ba9f743b11b0903554b5e56
5260 contents
= optionf
.read()
5261 if sys
.version_info
< (3,):
5262 contents
= contents
.decode(preferredencoding())
5263 res
= compat_shlex_split(contents
, comments
=True)
5269 def hide_login_info(opts
):
5270 PRIVATE_OPTS
= set(['-p', '--password', '-u', '--username', '--video-password', '--ap-password', '--ap-username'])
5271 eqre
= re
.compile('^(?P<key>' + ('|'.join(re
.escape(po
) for po
in PRIVATE_OPTS
)) + ')=.+$')
5276 return m
.group('key') + '=PRIVATE'
5280 opts
= list(map(_scrub_eq
, opts
))
5281 for idx
, opt
in enumerate(opts
):
5282 if opt
in PRIVATE_OPTS
and idx
+ 1 < len(opts
):
5283 opts
[idx
+ 1] = 'PRIVATE'
5286 def append_config(self
, *args
, label
=None):
5287 config
= type(self
)(self
._parser
, label
)
5288 config
._loaded
_paths
= self
._loaded
_paths
5289 if config
.init(*args
):
5290 self
.configs
.append(config
)
5294 for config
in reversed(self
.configs
):
5295 yield from config
.all_args
5296 yield from self
.own_args
or []
5298 def parse_args(self
):
5299 return self
._parser
.parse_args(list(self
.all_args
))