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': 'same-origin',
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 context
.verify_mode
= ssl
.CERT_REQUIRED
if opts_check_certificate
else ssl
.CERT_NONE
1001 if opts_check_certificate
:
1003 context
.load_default_certs()
1004 # Work around the issue in load_default_certs when there are bad certificates. See:
1005 # https://github.com/yt-dlp/yt-dlp/issues/1060,
1006 # https://bugs.python.org/issue35665, https://bugs.python.org/issue45312
1007 except ssl
.SSLError
:
1008 # enum_certificates is not present in mingw python. See https://github.com/yt-dlp/yt-dlp/issues/1151
1009 if sys
.platform
== 'win32' and hasattr(ssl
, 'enum_certificates'):
1010 # Create a new context to discard any certificates that were already loaded
1011 context
= ssl
.SSLContext(ssl
.PROTOCOL_TLS_CLIENT
)
1012 context
.check_hostname
, context
.verify_mode
= True, ssl
.CERT_REQUIRED
1013 for storename
in ('CA', 'ROOT'):
1014 _ssl_load_windows_store_certs(context
, storename
)
1015 context
.set_default_verify_paths()
1016 return YoutubeDLHTTPSHandler(params
, context
=context
, **kwargs
)
1019 def bug_reports_message(before
=';'):
1020 if ytdl_is_updateable():
1021 update_cmd
= 'type yt-dlp -U to update'
1023 update_cmd
= 'see https://github.com/yt-dlp/yt-dlp on how to update'
1024 msg
= 'please report this issue on https://github.com/yt-dlp/yt-dlp .'
1025 msg
+= ' Make sure you are using the latest version; %s.' % update_cmd
1026 msg
+= ' Be sure to call yt-dlp with the --verbose flag and include its complete output.'
1028 before
= before
.rstrip()
1029 if not before
or before
.endswith(('.', '!', '?')):
1030 msg
= msg
[0].title() + msg
[1:]
1032 return (before
+ ' ' if before
else '') + msg
1035 class YoutubeDLError(Exception):
1036 """Base exception for YoutubeDL errors."""
1039 def __init__(self
, msg
=None):
1042 elif self
.msg
is None:
1043 self
.msg
= type(self
).__name
__
1044 super().__init
__(self
.msg
)
1047 network_exceptions
= [compat_urllib_error
.URLError
, compat_http_client
.HTTPException
, socket
.error
]
1048 if hasattr(ssl
, 'CertificateError'):
1049 network_exceptions
.append(ssl
.CertificateError
)
1050 network_exceptions
= tuple(network_exceptions
)
1053 class ExtractorError(YoutubeDLError
):
1054 """Error during info extraction."""
1056 def __init__(self
, msg
, tb
=None, expected
=False, cause
=None, video_id
=None, ie
=None):
1057 """ tb, if given, is the original traceback (so that it can be printed out).
1058 If expected is set, this is a normal error message and most likely not a bug in yt-dlp.
1060 if sys
.exc_info()[0] in network_exceptions
:
1065 self
.expected
= expected
1067 self
.video_id
= video_id
1069 self
.exc_info
= sys
.exc_info() # preserve original exception
1071 super(ExtractorError
, self
).__init
__(''.join((
1072 format_field(ie
, template
='[%s] '),
1073 format_field(video_id
, template
='%s: '),
1075 format_field(cause
, template
=' (caused by %r)'),
1076 '' if expected
else bug_reports_message())))
1078 def format_traceback(self
):
1079 if self
.traceback
is None:
1081 return ''.join(traceback
.format_tb(self
.traceback
))
1084 class UnsupportedError(ExtractorError
):
1085 def __init__(self
, url
):
1086 super(UnsupportedError
, self
).__init
__(
1087 'Unsupported URL: %s' % url
, expected
=True)
1091 class RegexNotFoundError(ExtractorError
):
1092 """Error when a regex didn't match"""
1096 class GeoRestrictedError(ExtractorError
):
1097 """Geographic restriction Error exception.
1099 This exception may be thrown when a video is not available from your
1100 geographic location due to geographic restrictions imposed by a website.
1103 def __init__(self
, msg
, countries
=None, **kwargs
):
1104 kwargs
['expected'] = True
1105 super(GeoRestrictedError
, self
).__init
__(msg
, **kwargs
)
1106 self
.countries
= countries
1109 class DownloadError(YoutubeDLError
):
1110 """Download Error exception.
1112 This exception may be thrown by FileDownloader objects if they are not
1113 configured to continue on errors. They will contain the appropriate
1117 def __init__(self
, msg
, exc_info
=None):
1118 """ exc_info, if given, is the original exception that caused the trouble (as returned by sys.exc_info()). """
1119 super(DownloadError
, self
).__init
__(msg
)
1120 self
.exc_info
= exc_info
1123 class EntryNotInPlaylist(YoutubeDLError
):
1124 """Entry not in playlist exception.
1126 This exception will be thrown by YoutubeDL when a requested entry
1127 is not found in the playlist info_dict
1129 msg
= 'Entry not found in info'
1132 class SameFileError(YoutubeDLError
):
1133 """Same File exception.
1135 This exception will be thrown by FileDownloader objects if they detect
1136 multiple files would have to be downloaded to the same file on disk.
1138 msg
= 'Fixed output name but more than one file to download'
1140 def __init__(self
, filename
=None):
1141 if filename
is not None:
1142 self
.msg
+= f
': {filename}'
1143 super().__init
__(self
.msg
)
1146 class PostProcessingError(YoutubeDLError
):
1147 """Post Processing exception.
1149 This exception may be raised by PostProcessor's .run() method to
1150 indicate an error in the postprocessing task.
1154 class DownloadCancelled(YoutubeDLError
):
1155 """ Exception raised when the download queue should be interrupted """
1156 msg
= 'The download was cancelled'
1159 class ExistingVideoReached(DownloadCancelled
):
1160 """ --break-on-existing triggered """
1161 msg
= 'Encountered a video that is already in the archive, stopping due to --break-on-existing'
1164 class RejectedVideoReached(DownloadCancelled
):
1165 """ --break-on-reject triggered """
1166 msg
= 'Encountered a video that did not match filter, stopping due to --break-on-reject'
1169 class MaxDownloadsReached(DownloadCancelled
):
1170 """ --max-downloads limit has been reached. """
1171 msg
= 'Maximum number of downloads reached, stopping due to --max-downloads'
1174 class ReExtractInfo(YoutubeDLError
):
1175 """ Video info needs to be re-extracted. """
1177 def __init__(self
, msg
, expected
=False):
1178 super().__init
__(msg
)
1179 self
.expected
= expected
1182 class ThrottledDownload(ReExtractInfo
):
1183 """ Download speed below --throttled-rate. """
1184 msg
= 'The download speed is below throttle limit'
1187 super().__init
__(self
.msg
, expected
=False)
1190 class UnavailableVideoError(YoutubeDLError
):
1191 """Unavailable Format exception.
1193 This exception will be thrown when a video is requested
1194 in a format that is not available for that video.
1196 msg
= 'Unable to download video'
1198 def __init__(self
, err
=None):
1200 self
.msg
+= f
': {err}'
1201 super().__init
__(self
.msg
)
1204 class ContentTooShortError(YoutubeDLError
):
1205 """Content Too Short exception.
1207 This exception may be raised by FileDownloader objects when a file they
1208 download is too small for what the server announced first, indicating
1209 the connection was probably interrupted.
1212 def __init__(self
, downloaded
, expected
):
1213 super(ContentTooShortError
, self
).__init
__(
1214 'Downloaded {0} bytes, expected {1} bytes'.format(downloaded
, expected
)
1217 self
.downloaded
= downloaded
1218 self
.expected
= expected
1221 class XAttrMetadataError(YoutubeDLError
):
1222 def __init__(self
, code
=None, msg
='Unknown error'):
1223 super(XAttrMetadataError
, self
).__init
__(msg
)
1227 # Parsing code and msg
1228 if (self
.code
in (errno
.ENOSPC
, errno
.EDQUOT
)
1229 or 'No space left' in self
.msg
or 'Disk quota exceeded' in self
.msg
):
1230 self
.reason
= 'NO_SPACE'
1231 elif self
.code
== errno
.E2BIG
or 'Argument list too long' in self
.msg
:
1232 self
.reason
= 'VALUE_TOO_LONG'
1234 self
.reason
= 'NOT_SUPPORTED'
1237 class XAttrUnavailableError(YoutubeDLError
):
1241 def _create_http_connection(ydl_handler
, http_class
, is_https
, *args
, **kwargs
):
1242 # Working around python 2 bug (see http://bugs.python.org/issue17849) by limiting
1243 # expected HTTP responses to meet HTTP/1.0 or later (see also
1244 # https://github.com/ytdl-org/youtube-dl/issues/6727)
1245 if sys
.version_info
< (3, 0):
1246 kwargs
['strict'] = True
1247 hc
= http_class(*args
, **compat_kwargs(kwargs
))
1248 source_address
= ydl_handler
._params
.get('source_address')
1250 if source_address
is not None:
1251 # This is to workaround _create_connection() from socket where it will try all
1252 # address data from getaddrinfo() including IPv6. This filters the result from
1253 # getaddrinfo() based on the source_address value.
1254 # This is based on the cpython socket.create_connection() function.
1255 # https://github.com/python/cpython/blob/master/Lib/socket.py#L691
1256 def _create_connection(address
, timeout
=socket
._GLOBAL
_DEFAULT
_TIMEOUT
, source_address
=None):
1257 host
, port
= address
1259 addrs
= socket
.getaddrinfo(host
, port
, 0, socket
.SOCK_STREAM
)
1260 af
= socket
.AF_INET
if '.' in source_address
[0] else socket
.AF_INET6
1261 ip_addrs
= [addr
for addr
in addrs
if addr
[0] == af
]
1262 if addrs
and not ip_addrs
:
1263 ip_version
= 'v4' if af
== socket
.AF_INET
else 'v6'
1265 "No remote IP%s addresses available for connect, can't use '%s' as source address"
1266 % (ip_version
, source_address
[0]))
1267 for res
in ip_addrs
:
1268 af
, socktype
, proto
, canonname
, sa
= res
1271 sock
= socket
.socket(af
, socktype
, proto
)
1272 if timeout
is not socket
._GLOBAL
_DEFAULT
_TIMEOUT
:
1273 sock
.settimeout(timeout
)
1274 sock
.bind(source_address
)
1276 err
= None # Explicitly break reference cycle
1278 except socket
.error
as _
:
1280 if sock
is not None:
1285 raise socket
.error('getaddrinfo returns an empty list')
1286 if hasattr(hc
, '_create_connection'):
1287 hc
._create
_connection
= _create_connection
1288 sa
= (source_address
, 0)
1289 if hasattr(hc
, 'source_address'): # Python 2.7+
1290 hc
.source_address
= sa
1292 def _hc_connect(self
, *args
, **kwargs
):
1293 sock
= _create_connection(
1294 (self
.host
, self
.port
), self
.timeout
, sa
)
1296 self
.sock
= ssl
.wrap_socket(
1297 sock
, self
.key_file
, self
.cert_file
,
1298 ssl_version
=ssl
.PROTOCOL_TLSv1
)
1301 hc
.connect
= functools
.partial(_hc_connect
, hc
)
1306 def handle_youtubedl_headers(headers
):
1307 filtered_headers
= headers
1309 if 'Youtubedl-no-compression' in filtered_headers
:
1310 filtered_headers
= dict((k
, v
) for k
, v
in filtered_headers
.items() if k
.lower() != 'accept-encoding')
1311 del filtered_headers
['Youtubedl-no-compression']
1313 return filtered_headers
1316 class YoutubeDLHandler(compat_urllib_request
.HTTPHandler
):
1317 """Handler for HTTP requests and responses.
1319 This class, when installed with an OpenerDirector, automatically adds
1320 the standard headers to every HTTP request and handles gzipped and
1321 deflated responses from web servers. If compression is to be avoided in
1322 a particular request, the original request in the program code only has
1323 to include the HTTP header "Youtubedl-no-compression", which will be
1324 removed before making the real request.
1326 Part of this code was copied from:
1328 http://techknack.net/python-urllib2-handlers/
1330 Andrew Rowls, the author of that code, agreed to release it to the
1334 def __init__(self
, params
, *args
, **kwargs
):
1335 compat_urllib_request
.HTTPHandler
.__init
__(self
, *args
, **kwargs
)
1336 self
._params
= params
1338 def http_open(self
, req
):
1339 conn_class
= compat_http_client
.HTTPConnection
1341 socks_proxy
= req
.headers
.get('Ytdl-socks-proxy')
1343 conn_class
= make_socks_conn_class(conn_class
, socks_proxy
)
1344 del req
.headers
['Ytdl-socks-proxy']
1346 return self
.do_open(functools
.partial(
1347 _create_http_connection
, self
, conn_class
, False),
1355 return zlib
.decompress(data
, -zlib
.MAX_WBITS
)
1357 return zlib
.decompress(data
)
1359 def http_request(self
, req
):
1360 # According to RFC 3986, URLs can not contain non-ASCII characters, however this is not
1361 # always respected by websites, some tend to give out URLs with non percent-encoded
1362 # non-ASCII characters (see telemb.py, ard.py [#3412])
1363 # urllib chokes on URLs with non-ASCII characters (see http://bugs.python.org/issue3991)
1364 # To work around aforementioned issue we will replace request's original URL with
1365 # percent-encoded one
1366 # Since redirects are also affected (e.g. http://www.southpark.de/alle-episoden/s18e09)
1367 # the code of this workaround has been moved here from YoutubeDL.urlopen()
1368 url
= req
.get_full_url()
1369 url_escaped
= escape_url(url
)
1371 # Substitute URL if any change after escaping
1372 if url
!= url_escaped
:
1373 req
= update_Request(req
, url
=url_escaped
)
1375 for h
, v
in std_headers
.items():
1376 # Capitalize is needed because of Python bug 2275: http://bugs.python.org/issue2275
1377 # The dict keys are capitalized because of this bug by urllib
1378 if h
.capitalize() not in req
.headers
:
1379 req
.add_header(h
, v
)
1381 req
.headers
= handle_youtubedl_headers(req
.headers
)
1383 if sys
.version_info
< (2, 7) and '#' in req
.get_full_url():
1384 # Python 2.6 is brain-dead when it comes to fragments
1385 req
._Request
__original
= req
._Request
__original
.partition('#')[0]
1386 req
._Request
__r
_type
= req
._Request
__r
_type
.partition('#')[0]
1390 def http_response(self
, req
, resp
):
1393 if resp
.headers
.get('Content-encoding', '') == 'gzip':
1394 content
= resp
.read()
1395 gz
= gzip
.GzipFile(fileobj
=io
.BytesIO(content
), mode
='rb')
1397 uncompressed
= io
.BytesIO(gz
.read())
1398 except IOError as original_ioerror
:
1399 # There may be junk add the end of the file
1400 # See http://stackoverflow.com/q/4928560/35070 for details
1401 for i
in range(1, 1024):
1403 gz
= gzip
.GzipFile(fileobj
=io
.BytesIO(content
[:-i
]), mode
='rb')
1404 uncompressed
= io
.BytesIO(gz
.read())
1409 raise original_ioerror
1410 resp
= compat_urllib_request
.addinfourl(uncompressed
, old_resp
.headers
, old_resp
.url
, old_resp
.code
)
1411 resp
.msg
= old_resp
.msg
1412 del resp
.headers
['Content-encoding']
1414 if resp
.headers
.get('Content-encoding', '') == 'deflate':
1415 gz
= io
.BytesIO(self
.deflate(resp
.read()))
1416 resp
= compat_urllib_request
.addinfourl(gz
, old_resp
.headers
, old_resp
.url
, old_resp
.code
)
1417 resp
.msg
= old_resp
.msg
1418 del resp
.headers
['Content-encoding']
1419 # Percent-encode redirect URL of Location HTTP header to satisfy RFC 3986 (see
1420 # https://github.com/ytdl-org/youtube-dl/issues/6457).
1421 if 300 <= resp
.code
< 400:
1422 location
= resp
.headers
.get('Location')
1424 # As of RFC 2616 default charset is iso-8859-1 that is respected by python 3
1425 if sys
.version_info
>= (3, 0):
1426 location
= location
.encode('iso-8859-1').decode('utf-8')
1428 location
= location
.decode('utf-8')
1429 location_escaped
= escape_url(location
)
1430 if location
!= location_escaped
:
1431 del resp
.headers
['Location']
1432 if sys
.version_info
< (3, 0):
1433 location_escaped
= location_escaped
.encode('utf-8')
1434 resp
.headers
['Location'] = location_escaped
1437 https_request
= http_request
1438 https_response
= http_response
1441 def make_socks_conn_class(base_class
, socks_proxy
):
1442 assert issubclass(base_class
, (
1443 compat_http_client
.HTTPConnection
, compat_http_client
.HTTPSConnection
))
1445 url_components
= compat_urlparse
.urlparse(socks_proxy
)
1446 if url_components
.scheme
.lower() == 'socks5':
1447 socks_type
= ProxyType
.SOCKS5
1448 elif url_components
.scheme
.lower() in ('socks', 'socks4'):
1449 socks_type
= ProxyType
.SOCKS4
1450 elif url_components
.scheme
.lower() == 'socks4a':
1451 socks_type
= ProxyType
.SOCKS4A
1453 def unquote_if_non_empty(s
):
1456 return compat_urllib_parse_unquote_plus(s
)
1460 url_components
.hostname
, url_components
.port
or 1080,
1462 unquote_if_non_empty(url_components
.username
),
1463 unquote_if_non_empty(url_components
.password
),
1466 class SocksConnection(base_class
):
1468 self
.sock
= sockssocket()
1469 self
.sock
.setproxy(*proxy_args
)
1470 if type(self
.timeout
) in (int, float):
1471 self
.sock
.settimeout(self
.timeout
)
1472 self
.sock
.connect((self
.host
, self
.port
))
1474 if isinstance(self
, compat_http_client
.HTTPSConnection
):
1475 if hasattr(self
, '_context'): # Python > 2.6
1476 self
.sock
= self
._context
.wrap_socket(
1477 self
.sock
, server_hostname
=self
.host
)
1479 self
.sock
= ssl
.wrap_socket(self
.sock
)
1481 return SocksConnection
1484 class YoutubeDLHTTPSHandler(compat_urllib_request
.HTTPSHandler
):
1485 def __init__(self
, params
, https_conn_class
=None, *args
, **kwargs
):
1486 compat_urllib_request
.HTTPSHandler
.__init
__(self
, *args
, **kwargs
)
1487 self
._https
_conn
_class
= https_conn_class
or compat_http_client
.HTTPSConnection
1488 self
._params
= params
1490 def https_open(self
, req
):
1492 conn_class
= self
._https
_conn
_class
1494 if hasattr(self
, '_context'): # python > 2.6
1495 kwargs
['context'] = self
._context
1496 if hasattr(self
, '_check_hostname'): # python 3.x
1497 kwargs
['check_hostname'] = self
._check
_hostname
1499 socks_proxy
= req
.headers
.get('Ytdl-socks-proxy')
1501 conn_class
= make_socks_conn_class(conn_class
, socks_proxy
)
1502 del req
.headers
['Ytdl-socks-proxy']
1504 return self
.do_open(functools
.partial(
1505 _create_http_connection
, self
, conn_class
, True),
1509 class YoutubeDLCookieJar(compat_cookiejar
.MozillaCookieJar
):
1511 See [1] for cookie file format.
1513 1. https://curl.haxx.se/docs/http-cookies.html
1515 _HTTPONLY_PREFIX
= '#HttpOnly_'
1517 _HEADER
= '''# Netscape HTTP Cookie File
1518 # This file is generated by yt-dlp. Do not edit.
1521 _CookieFileEntry
= collections
.namedtuple(
1523 ('domain_name', 'include_subdomains', 'path', 'https_only', 'expires_at', 'name', 'value'))
1525 def save(self
, filename
=None, ignore_discard
=False, ignore_expires
=False):
1527 Save cookies to a file.
1529 Most of the code is taken from CPython 3.8 and slightly adapted
1530 to support cookie files with UTF-8 in both python 2 and 3.
1532 if filename
is None:
1533 if self
.filename
is not None:
1534 filename
= self
.filename
1536 raise ValueError(compat_cookiejar
.MISSING_FILENAME_TEXT
)
1538 # Store session cookies with `expires` set to 0 instead of an empty
1541 if cookie
.expires
is None:
1544 with io
.open(filename
, 'w', encoding
='utf-8') as f
:
1545 f
.write(self
._HEADER
)
1548 if not ignore_discard
and cookie
.discard
:
1550 if not ignore_expires
and cookie
.is_expired(now
):
1556 if cookie
.domain
.startswith('.'):
1557 initial_dot
= 'TRUE'
1559 initial_dot
= 'FALSE'
1560 if cookie
.expires
is not None:
1561 expires
= compat_str(cookie
.expires
)
1564 if cookie
.value
is None:
1565 # cookies.txt regards 'Set-Cookie: foo' as a cookie
1566 # with no name, whereas http.cookiejar regards it as a
1567 # cookie with no value.
1572 value
= cookie
.value
1574 '\t'.join([cookie
.domain
, initial_dot
, cookie
.path
,
1575 secure
, expires
, name
, value
]) + '\n')
1577 def load(self
, filename
=None, ignore_discard
=False, ignore_expires
=False):
1578 """Load cookies from a file."""
1579 if filename
is None:
1580 if self
.filename
is not None:
1581 filename
= self
.filename
1583 raise ValueError(compat_cookiejar
.MISSING_FILENAME_TEXT
)
1585 def prepare_line(line
):
1586 if line
.startswith(self
._HTTPONLY
_PREFIX
):
1587 line
= line
[len(self
._HTTPONLY
_PREFIX
):]
1588 # comments and empty lines are fine
1589 if line
.startswith('#') or not line
.strip():
1591 cookie_list
= line
.split('\t')
1592 if len(cookie_list
) != self
._ENTRY
_LEN
:
1593 raise compat_cookiejar
.LoadError('invalid length %d' % len(cookie_list
))
1594 cookie
= self
._CookieFileEntry
(*cookie_list
)
1595 if cookie
.expires_at
and not cookie
.expires_at
.isdigit():
1596 raise compat_cookiejar
.LoadError('invalid expires at %s' % cookie
.expires_at
)
1600 with io
.open(filename
, encoding
='utf-8') as f
:
1603 cf
.write(prepare_line(line
))
1604 except compat_cookiejar
.LoadError
as e
:
1606 'WARNING: skipping cookie file entry due to %s: %r\n'
1607 % (e
, line
), sys
.stderr
)
1610 self
._really
_load
(cf
, filename
, ignore_discard
, ignore_expires
)
1611 # Session cookies are denoted by either `expires` field set to
1612 # an empty string or 0. MozillaCookieJar only recognizes the former
1613 # (see [1]). So we need force the latter to be recognized as session
1614 # cookies on our own.
1615 # Session cookies may be important for cookies-based authentication,
1616 # e.g. usually, when user does not check 'Remember me' check box while
1617 # logging in on a site, some important cookies are stored as session
1618 # cookies so that not recognizing them will result in failed login.
1619 # 1. https://bugs.python.org/issue17164
1621 # Treat `expires=0` cookies as session cookies
1622 if cookie
.expires
== 0:
1623 cookie
.expires
= None
1624 cookie
.discard
= True
1627 class YoutubeDLCookieProcessor(compat_urllib_request
.HTTPCookieProcessor
):
1628 def __init__(self
, cookiejar
=None):
1629 compat_urllib_request
.HTTPCookieProcessor
.__init
__(self
, cookiejar
)
1631 def http_response(self
, request
, response
):
1632 # Python 2 will choke on next HTTP request in row if there are non-ASCII
1633 # characters in Set-Cookie HTTP header of last response (see
1634 # https://github.com/ytdl-org/youtube-dl/issues/6769).
1635 # In order to at least prevent crashing we will percent encode Set-Cookie
1636 # header before HTTPCookieProcessor starts processing it.
1637 # if sys.version_info < (3, 0) and response.headers:
1638 # for set_cookie_header in ('Set-Cookie', 'Set-Cookie2'):
1639 # set_cookie = response.headers.get(set_cookie_header)
1641 # set_cookie_escaped = compat_urllib_parse.quote(set_cookie, b"%/;:@&=+$,!~*'()?#[] ")
1642 # if set_cookie != set_cookie_escaped:
1643 # del response.headers[set_cookie_header]
1644 # response.headers[set_cookie_header] = set_cookie_escaped
1645 return compat_urllib_request
.HTTPCookieProcessor
.http_response(self
, request
, response
)
1647 https_request
= compat_urllib_request
.HTTPCookieProcessor
.http_request
1648 https_response
= http_response
1651 class YoutubeDLRedirectHandler(compat_urllib_request
.HTTPRedirectHandler
):
1652 """YoutubeDL redirect handler
1654 The code is based on HTTPRedirectHandler implementation from CPython [1].
1656 This redirect handler solves two issues:
1657 - ensures redirect URL is always unicode under python 2
1658 - introduces support for experimental HTTP response status code
1659 308 Permanent Redirect [2] used by some sites [3]
1661 1. https://github.com/python/cpython/blob/master/Lib/urllib/request.py
1662 2. https://developer.mozilla.org/en-US/docs/Web/HTTP/Status/308
1663 3. https://github.com/ytdl-org/youtube-dl/issues/28768
1666 http_error_301
= http_error_303
= http_error_307
= http_error_308
= compat_urllib_request
.HTTPRedirectHandler
.http_error_302
1668 def redirect_request(self
, req
, fp
, code
, msg
, headers
, newurl
):
1669 """Return a Request or None in response to a redirect.
1671 This is called by the http_error_30x methods when a
1672 redirection response is received. If a redirection should
1673 take place, return a new Request to allow http_error_30x to
1674 perform the redirect. Otherwise, raise HTTPError if no-one
1675 else should try to handle this url. Return None if you can't
1676 but another Handler might.
1678 m
= req
.get_method()
1679 if (not (code
in (301, 302, 303, 307, 308) and m
in ("GET", "HEAD")
1680 or code
in (301, 302, 303) and m
== "POST")):
1681 raise compat_HTTPError(req
.full_url
, code
, msg
, headers
, fp
)
1682 # Strictly (according to RFC 2616), 301 or 302 in response to
1683 # a POST MUST NOT cause a redirection without confirmation
1684 # from the user (of urllib.request, in this case). In practice,
1685 # essentially all clients do redirect in this case, so we do
1688 # On python 2 urlh.geturl() may sometimes return redirect URL
1689 # as byte string instead of unicode. This workaround allows
1690 # to force it always return unicode.
1691 if sys
.version_info
[0] < 3:
1692 newurl
= compat_str(newurl
)
1694 # Be conciliant with URIs containing a space. This is mainly
1695 # redundant with the more complete encoding done in http_error_302(),
1696 # but it is kept for compatibility with other callers.
1697 newurl
= newurl
.replace(' ', '%20')
1699 CONTENT_HEADERS
= ("content-length", "content-type")
1700 # NB: don't use dict comprehension for python 2.6 compatibility
1701 newheaders
= dict((k
, v
) for k
, v
in req
.headers
.items()
1702 if k
.lower() not in CONTENT_HEADERS
)
1703 return compat_urllib_request
.Request(
1704 newurl
, headers
=newheaders
, origin_req_host
=req
.origin_req_host
,
1708 def extract_timezone(date_str
):
1711 ^.{8,}? # >=8 char non-TZ prefix, if present
1712 (?P<tz>Z| # just the UTC Z, or
1713 (?:(?<=.\b\d{4}|\b\d{2}:\d\d)| # preceded by 4 digits or hh:mm or
1714 (?<!.\b[a-zA-Z]{3}|[a-zA-Z]{4}|..\b\d\d)) # not preceded by 3 alpha word or >= 4 alpha or 2 digits
1715 [ ]? # optional space
1716 (?P<sign>\+|-) # +/-
1717 (?P<hours>[0-9]{2}):?(?P<minutes>[0-9]{2}) # hh[:]mm
1721 timezone
= datetime
.timedelta()
1723 date_str
= date_str
[:-len(m
.group('tz'))]
1724 if not m
.group('sign'):
1725 timezone
= datetime
.timedelta()
1727 sign
= 1 if m
.group('sign') == '+' else -1
1728 timezone
= datetime
.timedelta(
1729 hours
=sign
* int(m
.group('hours')),
1730 minutes
=sign
* int(m
.group('minutes')))
1731 return timezone
, date_str
1734 def parse_iso8601(date_str
, delimiter
='T', timezone
=None):
1735 """ Return a UNIX timestamp from the given date """
1737 if date_str
is None:
1740 date_str
= re
.sub(r
'\.[0-9]+', '', date_str
)
1742 if timezone
is None:
1743 timezone
, date_str
= extract_timezone(date_str
)
1746 date_format
= '%Y-%m-%d{0}%H:%M:%S'.format(delimiter
)
1747 dt
= datetime
.datetime
.strptime(date_str
, date_format
) - timezone
1748 return calendar
.timegm(dt
.timetuple())
1753 def date_formats(day_first
=True):
1754 return DATE_FORMATS_DAY_FIRST
if day_first
else DATE_FORMATS_MONTH_FIRST
1757 def unified_strdate(date_str
, day_first
=True):
1758 """Return a string with the date in the format YYYYMMDD"""
1760 if date_str
is None:
1764 date_str
= date_str
.replace(',', ' ')
1765 # Remove AM/PM + timezone
1766 date_str
= re
.sub(r
'(?i)\s*(?:AM|PM)(?:\s+[A-Z]+)?', '', date_str
)
1767 _
, date_str
= extract_timezone(date_str
)
1769 for expression
in date_formats(day_first
):
1771 upload_date
= datetime
.datetime
.strptime(date_str
, expression
).strftime('%Y%m%d')
1774 if upload_date
is None:
1775 timetuple
= email
.utils
.parsedate_tz(date_str
)
1778 upload_date
= datetime
.datetime(*timetuple
[:6]).strftime('%Y%m%d')
1781 if upload_date
is not None:
1782 return compat_str(upload_date
)
1785 def unified_timestamp(date_str
, day_first
=True):
1786 if date_str
is None:
1789 date_str
= re
.sub(r
'[,|]', '', date_str
)
1791 pm_delta
= 12 if re
.search(r
'(?i)PM', date_str
) else 0
1792 timezone
, date_str
= extract_timezone(date_str
)
1794 # Remove AM/PM + timezone
1795 date_str
= re
.sub(r
'(?i)\s*(?:AM|PM)(?:\s+[A-Z]+)?', '', date_str
)
1797 # Remove unrecognized timezones from ISO 8601 alike timestamps
1798 m
= re
.search(r
'\d{1,2}:\d{1,2}(?:\.\d+)?(?P<tz>\s*[A-Z]+)$', date_str
)
1800 date_str
= date_str
[:-len(m
.group('tz'))]
1802 # Python only supports microseconds, so remove nanoseconds
1803 m
= re
.search(r
'^([0-9]{4,}-[0-9]{1,2}-[0-9]{1,2}T[0-9]{1,2}:[0-9]{1,2}:[0-9]{1,2}\.[0-9]{6})[0-9]+$', date_str
)
1805 date_str
= m
.group(1)
1807 for expression
in date_formats(day_first
):
1809 dt
= datetime
.datetime
.strptime(date_str
, expression
) - timezone
+ datetime
.timedelta(hours
=pm_delta
)
1810 return calendar
.timegm(dt
.timetuple())
1813 timetuple
= email
.utils
.parsedate_tz(date_str
)
1815 return calendar
.timegm(timetuple
) + pm_delta
* 3600
1818 def determine_ext(url
, default_ext
='unknown_video'):
1819 if url
is None or '.' not in url
:
1821 guess
= url
.partition('?')[0].rpartition('.')[2]
1822 if re
.match(r
'^[A-Za-z0-9]+$', guess
):
1824 # Try extract ext from URLs like http://example.com/foo/bar.mp4/?download
1825 elif guess
.rstrip('/') in KNOWN_EXTENSIONS
:
1826 return guess
.rstrip('/')
1831 def subtitles_filename(filename
, sub_lang
, sub_format
, expected_real_ext
=None):
1832 return replace_extension(filename
, sub_lang
+ '.' + sub_format
, expected_real_ext
)
1835 def datetime_from_str(date_str
, precision
='auto', format
='%Y%m%d'):
1837 Return a datetime object from a string in the format YYYYMMDD or
1838 (now|today|date)[+-][0-9](microsecond|second|minute|hour|day|week|month|year)(s)?
1840 format: string date format used to return datetime object from
1841 precision: round the time portion of a datetime object.
1842 auto|microsecond|second|minute|hour|day.
1843 auto: round to the unit provided in date_str (if applicable).
1845 auto_precision
= False
1846 if precision
== 'auto':
1847 auto_precision
= True
1848 precision
= 'microsecond'
1849 today
= datetime_round(datetime
.datetime
.utcnow(), precision
)
1850 if date_str
in ('now', 'today'):
1852 if date_str
== 'yesterday':
1853 return today
- datetime
.timedelta(days
=1)
1855 r
'(?P<start>.+)(?P<sign>[+-])(?P<time>\d+)(?P<unit>microsecond|second|minute|hour|day|week|month|year)(s)?',
1857 if match
is not None:
1858 start_time
= datetime_from_str(match
.group('start'), precision
, format
)
1859 time
= int(match
.group('time')) * (-1 if match
.group('sign') == '-' else 1)
1860 unit
= match
.group('unit')
1861 if unit
== 'month' or unit
== 'year':
1862 new_date
= datetime_add_months(start_time
, time
* 12 if unit
== 'year' else time
)
1868 delta
= datetime
.timedelta(**{unit + 's': time}
)
1869 new_date
= start_time
+ delta
1871 return datetime_round(new_date
, unit
)
1874 return datetime_round(datetime
.datetime
.strptime(date_str
, format
), precision
)
1877 def date_from_str(date_str
, format
='%Y%m%d'):
1879 Return a datetime object from a string in the format YYYYMMDD or
1880 (now|today|date)[+-][0-9](microsecond|second|minute|hour|day|week|month|year)(s)?
1882 format: string date format used to return datetime object from
1884 return datetime_from_str(date_str
, precision
='microsecond', format
=format
).date()
1887 def datetime_add_months(dt
, months
):
1888 """Increment/Decrement a datetime object by months."""
1889 month
= dt
.month
+ months
- 1
1890 year
= dt
.year
+ month
// 12
1891 month
= month
% 12 + 1
1892 day
= min(dt
.day
, calendar
.monthrange(year
, month
)[1])
1893 return dt
.replace(year
, month
, day
)
1896 def datetime_round(dt
, precision
='day'):
1898 Round a datetime object's time to a specific precision
1900 if precision
== 'microsecond':
1909 roundto
= lambda x
, n
: ((x
+ n
/ 2) // n
) * n
1910 timestamp
= calendar
.timegm(dt
.timetuple())
1911 return datetime
.datetime
.utcfromtimestamp(roundto(timestamp
, unit_seconds
[precision
]))
1914 def hyphenate_date(date_str
):
1916 Convert a date in 'YYYYMMDD' format to 'YYYY-MM-DD' format"""
1917 match
= re
.match(r
'^(\d\d\d\d)(\d\d)(\d\d)$', date_str
)
1918 if match
is not None:
1919 return '-'.join(match
.groups())
1924 class DateRange(object):
1925 """Represents a time interval between two dates"""
1927 def __init__(self
, start
=None, end
=None):
1928 """start and end must be strings in the format accepted by date"""
1929 if start
is not None:
1930 self
.start
= date_from_str(start
)
1932 self
.start
= datetime
.datetime
.min.date()
1934 self
.end
= date_from_str(end
)
1936 self
.end
= datetime
.datetime
.max.date()
1937 if self
.start
> self
.end
:
1938 raise ValueError('Date range: "%s" , the start date must be before the end date' % self
)
1942 """Returns a range that only contains the given day"""
1943 return cls(day
, day
)
1945 def __contains__(self
, date
):
1946 """Check if the date is in the range"""
1947 if not isinstance(date
, datetime
.date
):
1948 date
= date_from_str(date
)
1949 return self
.start
<= date
<= self
.end
1952 return '%s - %s' % (self
.start
.isoformat(), self
.end
.isoformat())
1955 def platform_name():
1956 """ Returns the platform name as a compat_str """
1957 res
= platform
.platform()
1958 if isinstance(res
, bytes):
1959 res
= res
.decode(preferredencoding())
1961 assert isinstance(res
, compat_str
)
1965 def get_windows_version():
1966 ''' Get Windows version. None if it's not running on Windows '''
1967 if compat_os_name
== 'nt':
1968 return version_tuple(platform
.win32_ver()[1])
1973 def _windows_write_string(s
, out
):
1974 """ Returns True if the string was written using special methods,
1975 False if it has yet to be written out."""
1976 # Adapted from http://stackoverflow.com/a/3259271/35070
1978 import ctypes
.wintypes
1986 fileno
= out
.fileno()
1987 except AttributeError:
1988 # If the output stream doesn't have a fileno, it's virtual
1990 except io
.UnsupportedOperation
:
1991 # Some strange Windows pseudo files?
1993 if fileno
not in WIN_OUTPUT_IDS
:
1996 GetStdHandle
= compat_ctypes_WINFUNCTYPE(
1997 ctypes
.wintypes
.HANDLE
, ctypes
.wintypes
.DWORD
)(
1998 ('GetStdHandle', ctypes
.windll
.kernel32
))
1999 h
= GetStdHandle(WIN_OUTPUT_IDS
[fileno
])
2001 WriteConsoleW
= compat_ctypes_WINFUNCTYPE(
2002 ctypes
.wintypes
.BOOL
, ctypes
.wintypes
.HANDLE
, ctypes
.wintypes
.LPWSTR
,
2003 ctypes
.wintypes
.DWORD
, ctypes
.POINTER(ctypes
.wintypes
.DWORD
),
2004 ctypes
.wintypes
.LPVOID
)(('WriteConsoleW', ctypes
.windll
.kernel32
))
2005 written
= ctypes
.wintypes
.DWORD(0)
2007 GetFileType
= compat_ctypes_WINFUNCTYPE(ctypes
.wintypes
.DWORD
, ctypes
.wintypes
.DWORD
)(('GetFileType', ctypes
.windll
.kernel32
))
2008 FILE_TYPE_CHAR
= 0x0002
2009 FILE_TYPE_REMOTE
= 0x8000
2010 GetConsoleMode
= compat_ctypes_WINFUNCTYPE(
2011 ctypes
.wintypes
.BOOL
, ctypes
.wintypes
.HANDLE
,
2012 ctypes
.POINTER(ctypes
.wintypes
.DWORD
))(
2013 ('GetConsoleMode', ctypes
.windll
.kernel32
))
2014 INVALID_HANDLE_VALUE
= ctypes
.wintypes
.DWORD(-1).value
2016 def not_a_console(handle
):
2017 if handle
== INVALID_HANDLE_VALUE
or handle
is None:
2019 return ((GetFileType(handle
) & ~FILE_TYPE_REMOTE
) != FILE_TYPE_CHAR
2020 or GetConsoleMode(handle
, ctypes
.byref(ctypes
.wintypes
.DWORD())) == 0)
2022 if not_a_console(h
):
2025 def next_nonbmp_pos(s
):
2027 return next(i
for i
, c
in enumerate(s
) if ord(c
) > 0xffff)
2028 except StopIteration:
2032 count
= min(next_nonbmp_pos(s
), 1024)
2034 ret
= WriteConsoleW(
2035 h
, s
, count
if count
else 2, ctypes
.byref(written
), None)
2037 raise OSError('Failed to write string')
2038 if not count
: # We just wrote a non-BMP character
2039 assert written
.value
== 2
2042 assert written
.value
> 0
2043 s
= s
[written
.value
:]
2047 def write_string(s
, out
=None, encoding
=None):
2050 assert type(s
) == compat_str
2052 if sys
.platform
== 'win32' and encoding
is None and hasattr(out
, 'fileno'):
2053 if _windows_write_string(s
, out
):
2056 if ('b' in getattr(out
, 'mode', '')
2057 or sys
.version_info
[0] < 3): # Python 2 lies about mode of sys.stderr
2058 byt
= s
.encode(encoding
or preferredencoding(), 'ignore')
2060 elif hasattr(out
, 'buffer'):
2061 enc
= encoding
or getattr(out
, 'encoding', None) or preferredencoding()
2062 byt
= s
.encode(enc
, 'ignore')
2063 out
.buffer.write(byt
)
2069 def bytes_to_intlist(bs
):
2072 if isinstance(bs
[0], int): # Python 3
2075 return [ord(c
) for c
in bs
]
2078 def intlist_to_bytes(xs
):
2081 return compat_struct_pack('%dB' % len(xs
), *xs
)
2084 # Cross-platform file locking
2085 if sys
.platform
== 'win32':
2086 import ctypes
.wintypes
2089 class OVERLAPPED(ctypes
.Structure
):
2091 ('Internal', ctypes
.wintypes
.LPVOID
),
2092 ('InternalHigh', ctypes
.wintypes
.LPVOID
),
2093 ('Offset', ctypes
.wintypes
.DWORD
),
2094 ('OffsetHigh', ctypes
.wintypes
.DWORD
),
2095 ('hEvent', ctypes
.wintypes
.HANDLE
),
2098 kernel32
= ctypes
.windll
.kernel32
2099 LockFileEx
= kernel32
.LockFileEx
2100 LockFileEx
.argtypes
= [
2101 ctypes
.wintypes
.HANDLE
, # hFile
2102 ctypes
.wintypes
.DWORD
, # dwFlags
2103 ctypes
.wintypes
.DWORD
, # dwReserved
2104 ctypes
.wintypes
.DWORD
, # nNumberOfBytesToLockLow
2105 ctypes
.wintypes
.DWORD
, # nNumberOfBytesToLockHigh
2106 ctypes
.POINTER(OVERLAPPED
) # Overlapped
2108 LockFileEx
.restype
= ctypes
.wintypes
.BOOL
2109 UnlockFileEx
= kernel32
.UnlockFileEx
2110 UnlockFileEx
.argtypes
= [
2111 ctypes
.wintypes
.HANDLE
, # hFile
2112 ctypes
.wintypes
.DWORD
, # dwReserved
2113 ctypes
.wintypes
.DWORD
, # nNumberOfBytesToLockLow
2114 ctypes
.wintypes
.DWORD
, # nNumberOfBytesToLockHigh
2115 ctypes
.POINTER(OVERLAPPED
) # Overlapped
2117 UnlockFileEx
.restype
= ctypes
.wintypes
.BOOL
2118 whole_low
= 0xffffffff
2119 whole_high
= 0x7fffffff
2121 def _lock_file(f
, exclusive
):
2122 overlapped
= OVERLAPPED()
2123 overlapped
.Offset
= 0
2124 overlapped
.OffsetHigh
= 0
2125 overlapped
.hEvent
= 0
2126 f
._lock
_file
_overlapped
_p
= ctypes
.pointer(overlapped
)
2127 handle
= msvcrt
.get_osfhandle(f
.fileno())
2128 if not LockFileEx(handle
, 0x2 if exclusive
else 0x0, 0,
2129 whole_low
, whole_high
, f
._lock
_file
_overlapped
_p
):
2130 raise OSError('Locking file failed: %r' % ctypes
.FormatError())
2132 def _unlock_file(f
):
2133 assert f
._lock
_file
_overlapped
_p
2134 handle
= msvcrt
.get_osfhandle(f
.fileno())
2135 if not UnlockFileEx(handle
, 0,
2136 whole_low
, whole_high
, f
._lock
_file
_overlapped
_p
):
2137 raise OSError('Unlocking file failed: %r' % ctypes
.FormatError())
2140 # Some platforms, such as Jython, is missing fcntl
2144 def _lock_file(f
, exclusive
):
2145 fcntl
.flock(f
, fcntl
.LOCK_EX
if exclusive
else fcntl
.LOCK_SH
)
2147 def _unlock_file(f
):
2148 fcntl
.flock(f
, fcntl
.LOCK_UN
)
2150 UNSUPPORTED_MSG
= 'file locking is not supported on this platform'
2152 def _lock_file(f
, exclusive
):
2153 raise IOError(UNSUPPORTED_MSG
)
2155 def _unlock_file(f
):
2156 raise IOError(UNSUPPORTED_MSG
)
2159 class locked_file(object):
2160 def __init__(self
, filename
, mode
, encoding
=None):
2161 assert mode
in ['r', 'a', 'w']
2162 self
.f
= io
.open(filename
, mode
, encoding
=encoding
)
2165 def __enter__(self
):
2166 exclusive
= self
.mode
!= 'r'
2168 _lock_file(self
.f
, exclusive
)
2174 def __exit__(self
, etype
, value
, traceback
):
2176 _unlock_file(self
.f
)
2183 def write(self
, *args
):
2184 return self
.f
.write(*args
)
2186 def read(self
, *args
):
2187 return self
.f
.read(*args
)
2190 def get_filesystem_encoding():
2191 encoding
= sys
.getfilesystemencoding()
2192 return encoding
if encoding
is not None else 'utf-8'
2195 def shell_quote(args
):
2197 encoding
= get_filesystem_encoding()
2199 if isinstance(a
, bytes):
2200 # We may get a filename encoded with 'encodeFilename'
2201 a
= a
.decode(encoding
)
2202 quoted_args
.append(compat_shlex_quote(a
))
2203 return ' '.join(quoted_args
)
2206 def smuggle_url(url
, data
):
2207 """ Pass additional data in a URL for internal use. """
2209 url
, idata
= unsmuggle_url(url
, {})
2211 sdata
= compat_urllib_parse_urlencode(
2212 {'__youtubedl_smuggle': json.dumps(data)}
)
2213 return url
+ '#' + sdata
2216 def unsmuggle_url(smug_url
, default
=None):
2217 if '#__youtubedl_smuggle' not in smug_url
:
2218 return smug_url
, default
2219 url
, _
, sdata
= smug_url
.rpartition('#')
2220 jsond
= compat_parse_qs(sdata
)['__youtubedl_smuggle'][0]
2221 data
= json
.loads(jsond
)
2225 def format_decimal_suffix(num
, fmt
='%d%s', *, factor
=1000):
2226 """ Formats numbers with decimal sufixes like K, M, etc """
2227 num
, factor
= float_or_none(num
), float(factor
)
2230 exponent
= 0 if num
== 0 else int(math
.log(num
, factor
))
2231 suffix
= ['', *'kMGTPEZY'][exponent
]
2233 suffix
= {'k': 'Ki', '': ''}
.get(suffix
, f
'{suffix}i')
2234 converted
= num
/ (factor
** exponent
)
2235 return fmt
% (converted
, suffix
)
2238 def format_bytes(bytes):
2239 return format_decimal_suffix(bytes, '%.2f%sB', factor
=1024) or 'N/A'
2242 def lookup_unit_table(unit_table
, s
):
2243 units_re
= '|'.join(re
.escape(u
) for u
in unit_table
)
2245 r
'(?P<num>[0-9]+(?:[,.][0-9]*)?)\s*(?P<unit>%s)\b' % units_re
, s
)
2248 num_str
= m
.group('num').replace(',', '.')
2249 mult
= unit_table
[m
.group('unit')]
2250 return int(float(num_str
) * mult
)
2253 def parse_filesize(s
):
2257 # The lower-case forms are of course incorrect and unofficial,
2258 # but we support those too
2275 'megabytes': 1000 ** 2,
2276 'mebibytes': 1024 ** 2,
2282 'gigabytes': 1000 ** 3,
2283 'gibibytes': 1024 ** 3,
2289 'terabytes': 1000 ** 4,
2290 'tebibytes': 1024 ** 4,
2296 'petabytes': 1000 ** 5,
2297 'pebibytes': 1024 ** 5,
2303 'exabytes': 1000 ** 6,
2304 'exbibytes': 1024 ** 6,
2310 'zettabytes': 1000 ** 7,
2311 'zebibytes': 1024 ** 7,
2317 'yottabytes': 1000 ** 8,
2318 'yobibytes': 1024 ** 8,
2321 return lookup_unit_table(_UNIT_TABLE
, s
)
2328 s
= re
.sub(r
'^[^\d]+\s', '', s
).strip()
2330 if re
.match(r
'^[\d,.]+$', s
):
2331 return str_to_int(s
)
2344 ret
= lookup_unit_table(_UNIT_TABLE
, s
)
2348 mobj
= re
.match(r
'([\d,.]+)(?:$|\s)', s
)
2350 return str_to_int(mobj
.group(1))
2353 def parse_resolution(s
):
2357 mobj
= re
.search(r
'(?<![a-zA-Z0-9])(?P<w>\d+)\s*[xX×,]\s*(?P<h>\d+)(?![a-zA-Z0-9])', s
)
2360 'width': int(mobj
.group('w')),
2361 'height': int(mobj
.group('h')),
2364 mobj
= re
.search(r
'(?<![a-zA-Z0-9])(\d+)[pPiI](?![a-zA-Z0-9])', s
)
2366 return {'height': int(mobj.group(1))}
2368 mobj
= re
.search(r
'\b([48])[kK]\b', s
)
2370 return {'height': int(mobj.group(1)) * 540}
2375 def parse_bitrate(s
):
2376 if not isinstance(s
, compat_str
):
2378 mobj
= re
.search(r
'\b(\d+)\s*kbps', s
)
2380 return int(mobj
.group(1))
2383 def month_by_name(name
, lang
='en'):
2384 """ Return the number of a month by (locale-independently) English name """
2386 month_names
= MONTH_NAMES
.get(lang
, MONTH_NAMES
['en'])
2389 return month_names
.index(name
) + 1
2394 def month_by_abbreviation(abbrev
):
2395 """ Return the number of a month by (locale-independently) English
2399 return [s
[:3] for s
in ENGLISH_MONTH_NAMES
].index(abbrev
) + 1
2404 def fix_xml_ampersands(xml_str
):
2405 """Replace all the '&' by '&' in XML"""
2407 r
'&(?!amp;|lt;|gt;|apos;|quot;|#x[0-9a-fA-F]{,4};|#[0-9]{,4};)',
2412 def setproctitle(title
):
2413 assert isinstance(title
, compat_str
)
2415 # ctypes in Jython is not complete
2416 # http://bugs.jython.org/issue2148
2417 if sys
.platform
.startswith('java'):
2421 libc
= ctypes
.cdll
.LoadLibrary('libc.so.6')
2425 # LoadLibrary in Windows Python 2.7.13 only expects
2426 # a bytestring, but since unicode_literals turns
2427 # every string into a unicode string, it fails.
2429 title_bytes
= title
.encode('utf-8')
2430 buf
= ctypes
.create_string_buffer(len(title_bytes
))
2431 buf
.value
= title_bytes
2433 libc
.prctl(15, buf
, 0, 0, 0)
2434 except AttributeError:
2435 return # Strange libc, just skip this
2438 def remove_start(s
, start
):
2439 return s
[len(start
):] if s
is not None and s
.startswith(start
) else s
2442 def remove_end(s
, end
):
2443 return s
[:-len(end
)] if s
is not None and s
.endswith(end
) else s
2446 def remove_quotes(s
):
2447 if s
is None or len(s
) < 2:
2449 for quote
in ('"', "'", ):
2450 if s
[0] == quote
and s
[-1] == quote
:
2455 def get_domain(url
):
2456 domain
= re
.match(r
'(?:https?:\/\/)?(?:www\.)?(?P<domain>[^\n\/]+\.[^\n\/]+)(?:\/(.*))?', url
)
2457 return domain
.group('domain') if domain
else None
2460 def url_basename(url
):
2461 path
= compat_urlparse
.urlparse(url
).path
2462 return path
.strip('/').split('/')[-1]
2466 return re
.match(r
'https?://[^?#&]+/', url
).group()
2469 def urljoin(base
, path
):
2470 if isinstance(path
, bytes):
2471 path
= path
.decode('utf-8')
2472 if not isinstance(path
, compat_str
) or not path
:
2474 if re
.match(r
'^(?:[a-zA-Z][a-zA-Z0-9+-.]*:)?//', path
):
2476 if isinstance(base
, bytes):
2477 base
= base
.decode('utf-8')
2478 if not isinstance(base
, compat_str
) or not re
.match(
2479 r
'^(?:https?:)?//', base
):
2481 return compat_urlparse
.urljoin(base
, path
)
2484 class HEADRequest(compat_urllib_request
.Request
):
2485 def get_method(self
):
2489 class PUTRequest(compat_urllib_request
.Request
):
2490 def get_method(self
):
2494 def int_or_none(v
, scale
=1, default
=None, get_attr
=None, invscale
=1):
2495 if get_attr
and v
is not None:
2496 v
= getattr(v
, get_attr
, None)
2498 return int(v
) * invscale
// scale
2499 except (ValueError, TypeError, OverflowError):
2503 def str_or_none(v
, default
=None):
2504 return default
if v
is None else compat_str(v
)
2507 def str_to_int(int_str
):
2508 """ A more relaxed version of int_or_none """
2509 if isinstance(int_str
, compat_integer_types
):
2511 elif isinstance(int_str
, compat_str
):
2512 int_str
= re
.sub(r
'[,\.\+]', '', int_str
)
2513 return int_or_none(int_str
)
2516 def float_or_none(v
, scale
=1, invscale
=1, default
=None):
2520 return float(v
) * invscale
/ scale
2521 except (ValueError, TypeError):
2525 def bool_or_none(v
, default
=None):
2526 return v
if isinstance(v
, bool) else default
2529 def strip_or_none(v
, default
=None):
2530 return v
.strip() if isinstance(v
, compat_str
) else default
2533 def url_or_none(url
):
2534 if not url
or not isinstance(url
, compat_str
):
2537 return url
if re
.match(r
'^(?:(?:https?|rt(?:m(?:pt?[es]?|fp)|sp[su]?)|mms|ftps?):)?//', url
) else None
2540 def strftime_or_none(timestamp
, date_format
, default
=None):
2541 datetime_object
= None
2543 if isinstance(timestamp
, compat_numeric_types
): # unix timestamp
2544 datetime_object
= datetime
.datetime
.utcfromtimestamp(timestamp
)
2545 elif isinstance(timestamp
, compat_str
): # assume YYYYMMDD
2546 datetime_object
= datetime
.datetime
.strptime(timestamp
, '%Y%m%d')
2547 return datetime_object
.strftime(date_format
)
2548 except (ValueError, TypeError, AttributeError):
2552 def parse_duration(s
):
2553 if not isinstance(s
, compat_basestring
):
2559 days
, hours
, mins
, secs
, ms
= [None] * 5
2560 m
= re
.match(r
'''(?x)
2562 (?:(?:(?P<days>[0-9]+):)?(?P<hours>[0-9]+):)?(?P<mins>[0-9]+):)?
2563 (?P<secs>(?(before_secs)[0-9]{1,2}|[0-9]+))
2564 (?P<ms>[.:][0-9]+)?Z?$
2567 days
, hours
, mins
, secs
, ms
= m
.group('days', 'hours', 'mins', 'secs', 'ms')
2572 [0-9]+\s*y(?:ears?)?\s*
2575 [0-9]+\s*m(?:onths?)?\s*
2578 [0-9]+\s*w(?:eeks?)?\s*
2581 (?P<days>[0-9]+)\s*d(?:ays?)?\s*
2585 (?P<hours>[0-9]+)\s*h(?:ours?)?\s*
2588 (?P<mins>[0-9]+)\s*m(?:in(?:ute)?s?)?\s*
2591 (?P<secs>[0-9]+)(?P<ms>\.[0-9]+)?\s*s(?:ec(?:ond)?s?)?\s*
2594 days
, hours
, mins
, secs
, ms
= m
.groups()
2596 m
= re
.match(r
'(?i)(?:(?P<hours>[0-9.]+)\s*(?:hours?)|(?P<mins>[0-9.]+)\s*(?:mins?\.?|minutes?)\s*)Z?$', s
)
2598 hours
, mins
= m
.groups()
2604 duration
+= float(secs
)
2606 duration
+= float(mins
) * 60
2608 duration
+= float(hours
) * 60 * 60
2610 duration
+= float(days
) * 24 * 60 * 60
2612 duration
+= float(ms
.replace(':', '.'))
2616 def prepend_extension(filename
, ext
, expected_real_ext
=None):
2617 name
, real_ext
= os
.path
.splitext(filename
)
2619 '{0}.{1}{2}'.format(name
, ext
, real_ext
)
2620 if not expected_real_ext
or real_ext
[1:] == expected_real_ext
2621 else '{0}.{1}'.format(filename
, ext
))
2624 def replace_extension(filename
, ext
, expected_real_ext
=None):
2625 name
, real_ext
= os
.path
.splitext(filename
)
2626 return '{0}.{1}'.format(
2627 name
if not expected_real_ext
or real_ext
[1:] == expected_real_ext
else filename
,
2631 def check_executable(exe
, args
=[]):
2632 """ Checks if the given binary is installed somewhere in PATH, and returns its name.
2633 args can be a list of arguments for a short output (like -version) """
2635 Popen([exe
] + args
, stdout
=subprocess
.PIPE
, stderr
=subprocess
.PIPE
).communicate_or_kill()
2641 def _get_exe_version_output(exe
, args
):
2643 # STDIN should be redirected too. On UNIX-like systems, ffmpeg triggers
2644 # SIGTTOU if yt-dlp is run in the background.
2645 # See https://github.com/ytdl-org/youtube-dl/issues/955#issuecomment-209789656
2647 [encodeArgument(exe
)] + args
, stdin
=subprocess
.PIPE
,
2648 stdout
=subprocess
.PIPE
, stderr
=subprocess
.STDOUT
).communicate_or_kill()
2651 if isinstance(out
, bytes): # Python 2.x
2652 out
= out
.decode('ascii', 'ignore')
2656 def detect_exe_version(output
, version_re
=None, unrecognized
='present'):
2657 assert isinstance(output
, compat_str
)
2658 if version_re
is None:
2659 version_re
= r
'version\s+([-0-9._a-zA-Z]+)'
2660 m
= re
.search(version_re
, output
)
2667 def get_exe_version(exe
, args
=['--version'],
2668 version_re
=None, unrecognized
='present'):
2669 """ Returns the version of the specified executable,
2670 or False if the executable is not present """
2671 out
= _get_exe_version_output(exe
, args
)
2672 return detect_exe_version(out
, version_re
, unrecognized
) if out
else False
2675 class LazyList(collections
.abc
.Sequence
):
2676 ''' Lazy immutable list from an iterable
2677 Note that slices of a LazyList are lists and not LazyList'''
2679 class IndexError(IndexError):
2682 def __init__(self
, iterable
, *, reverse
=False, _cache
=None):
2683 self
.__iterable
= iter(iterable
)
2684 self
.__cache
= [] if _cache
is None else _cache
2685 self
.__reversed
= reverse
2689 # We need to consume the entire iterable to iterate in reverse
2690 yield from self
.exhaust()
2692 yield from self
.__cache
2693 for item
in self
.__iterable
:
2694 self
.__cache
.append(item
)
2697 def __exhaust(self
):
2698 self
.__cache
.extend(self
.__iterable
)
2699 # Discard the emptied iterable to make it pickle-able
2700 self
.__iterable
= []
2704 ''' Evaluate the entire iterable '''
2705 return self
.__exhaust
()[::-1 if self
.__reversed
else 1]
2708 def __reverse_index(x
):
2709 return None if x
is None else -(x
+ 1)
2711 def __getitem__(self
, idx
):
2712 if isinstance(idx
, slice):
2714 idx
= slice(self
.__reverse
_index
(idx
.start
), self
.__reverse
_index
(idx
.stop
), -(idx
.step
or 1))
2715 start
, stop
, step
= idx
.start
, idx
.stop
, idx
.step
or 1
2716 elif isinstance(idx
, int):
2718 idx
= self
.__reverse
_index
(idx
)
2719 start
, stop
, step
= idx
, idx
, 0
2721 raise TypeError('indices must be integers or slices')
2722 if ((start
or 0) < 0 or (stop
or 0) < 0
2723 or (start
is None and step
< 0)
2724 or (stop
is None and step
> 0)):
2725 # We need to consume the entire iterable to be able to slice from the end
2726 # Obviously, never use this with infinite iterables
2729 return self
.__cache
[idx
]
2730 except IndexError as e
:
2731 raise self
.IndexError(e
) from e
2732 n
= max(start
or 0, stop
or 0) - len(self
.__cache
) + 1
2734 self
.__cache
.extend(itertools
.islice(self
.__iterable
, n
))
2736 return self
.__cache
[idx
]
2737 except IndexError as e
:
2738 raise self
.IndexError(e
) from e
2742 self
[-1] if self
.__reversed
else self
[0]
2743 except self
.IndexError:
2749 return len(self
.__cache
)
2751 def __reversed__(self
):
2752 return type(self
)(self
.__iterable
, reverse
=not self
.__reversed
, _cache
=self
.__cache
)
2755 return type(self
)(self
.__iterable
, reverse
=self
.__reversed
, _cache
=self
.__cache
)
2758 # repr and str should mimic a list. So we exhaust the iterable
2759 return repr(self
.exhaust())
2762 return repr(self
.exhaust())
2767 class IndexError(IndexError):
2771 # This is only useful for tests
2772 return len(self
.getslice())
2774 def __init__(self
, pagefunc
, pagesize
, use_cache
=True):
2775 self
._pagefunc
= pagefunc
2776 self
._pagesize
= pagesize
2777 self
._use
_cache
= use_cache
2780 def getpage(self
, pagenum
):
2781 page_results
= self
._cache
.get(pagenum
)
2782 if page_results
is None:
2783 page_results
= list(self
._pagefunc
(pagenum
))
2785 self
._cache
[pagenum
] = page_results
2788 def getslice(self
, start
=0, end
=None):
2789 return list(self
._getslice
(start
, end
))
2791 def _getslice(self
, start
, end
):
2792 raise NotImplementedError('This method must be implemented by subclasses')
2794 def __getitem__(self
, idx
):
2795 # NOTE: cache must be enabled if this is used
2796 if not isinstance(idx
, int) or idx
< 0:
2797 raise TypeError('indices must be non-negative integers')
2798 entries
= self
.getslice(idx
, idx
+ 1)
2800 raise self
.IndexError()
2804 class OnDemandPagedList(PagedList
):
2805 def _getslice(self
, start
, end
):
2806 for pagenum
in itertools
.count(start
// self
._pagesize
):
2807 firstid
= pagenum
* self
._pagesize
2808 nextfirstid
= pagenum
* self
._pagesize
+ self
._pagesize
2809 if start
>= nextfirstid
:
2813 start
% self
._pagesize
2814 if firstid
<= start
< nextfirstid
2817 ((end
- 1) % self
._pagesize
) + 1
2818 if (end
is not None and firstid
<= end
<= nextfirstid
)
2821 page_results
= self
.getpage(pagenum
)
2822 if startv
!= 0 or endv
is not None:
2823 page_results
= page_results
[startv
:endv
]
2824 yield from page_results
2826 # A little optimization - if current page is not "full", ie. does
2827 # not contain page_size videos then we can assume that this page
2828 # is the last one - there are no more ids on further pages -
2829 # i.e. no need to query again.
2830 if len(page_results
) + startv
< self
._pagesize
:
2833 # If we got the whole page, but the next page is not interesting,
2834 # break out early as well
2835 if end
== nextfirstid
:
2839 class InAdvancePagedList(PagedList
):
2840 def __init__(self
, pagefunc
, pagecount
, pagesize
):
2841 self
._pagecount
= pagecount
2842 PagedList
.__init
__(self
, pagefunc
, pagesize
, True)
2844 def _getslice(self
, start
, end
):
2845 start_page
= start
// self
._pagesize
2847 self
._pagecount
if end
is None else (end
// self
._pagesize
+ 1))
2848 skip_elems
= start
- start_page
* self
._pagesize
2849 only_more
= None if end
is None else end
- start
2850 for pagenum
in range(start_page
, end_page
):
2851 page_results
= self
.getpage(pagenum
)
2853 page_results
= page_results
[skip_elems
:]
2855 if only_more
is not None:
2856 if len(page_results
) < only_more
:
2857 only_more
-= len(page_results
)
2859 yield from page_results
[:only_more
]
2861 yield from page_results
2864 def uppercase_escape(s
):
2865 unicode_escape
= codecs
.getdecoder('unicode_escape')
2867 r
'\\U[0-9a-fA-F]{8}',
2868 lambda m
: unicode_escape(m
.group(0))[0],
2872 def lowercase_escape(s
):
2873 unicode_escape
= codecs
.getdecoder('unicode_escape')
2875 r
'\\u[0-9a-fA-F]{4}',
2876 lambda m
: unicode_escape(m
.group(0))[0],
2880 def escape_rfc3986(s
):
2881 """Escape non-ASCII characters as suggested by RFC 3986"""
2882 if sys
.version_info
< (3, 0) and isinstance(s
, compat_str
):
2883 s
= s
.encode('utf-8')
2884 return compat_urllib_parse
.quote(s
, b
"%/;:@&=+$,!~*'()?#[]")
2887 def escape_url(url
):
2888 """Escape URL as suggested by RFC 3986"""
2889 url_parsed
= compat_urllib_parse_urlparse(url
)
2890 return url_parsed
._replace
(
2891 netloc
=url_parsed
.netloc
.encode('idna').decode('ascii'),
2892 path
=escape_rfc3986(url_parsed
.path
),
2893 params
=escape_rfc3986(url_parsed
.params
),
2894 query
=escape_rfc3986(url_parsed
.query
),
2895 fragment
=escape_rfc3986(url_parsed
.fragment
)
2900 return compat_parse_qs(compat_urllib_parse_urlparse(url
).query
)
2903 def read_batch_urls(batch_fd
):
2905 if not isinstance(url
, compat_str
):
2906 url
= url
.decode('utf-8', 'replace')
2907 BOM_UTF8
= ('\xef\xbb\xbf', '\ufeff')
2908 for bom
in BOM_UTF8
:
2909 if url
.startswith(bom
):
2910 url
= url
[len(bom
):]
2912 if not url
or url
.startswith(('#', ';', ']')):
2914 # "#" cannot be stripped out since it is part of the URI
2915 # However, it can be safely stipped out if follwing a whitespace
2916 return re
.split(r
'\s#', url
, 1)[0].rstrip()
2918 with contextlib
.closing(batch_fd
) as fd
:
2919 return [url
for url
in map(fixup
, fd
) if url
]
2922 def urlencode_postdata(*args
, **kargs
):
2923 return compat_urllib_parse_urlencode(*args
, **kargs
).encode('ascii')
2926 def update_url_query(url
, query
):
2929 parsed_url
= compat_urlparse
.urlparse(url
)
2930 qs
= compat_parse_qs(parsed_url
.query
)
2932 return compat_urlparse
.urlunparse(parsed_url
._replace
(
2933 query
=compat_urllib_parse_urlencode(qs
, True)))
2936 def update_Request(req
, url
=None, data
=None, headers
={}, query={}
):
2937 req_headers
= req
.headers
.copy()
2938 req_headers
.update(headers
)
2939 req_data
= data
or req
.data
2940 req_url
= update_url_query(url
or req
.get_full_url(), query
)
2941 req_get_method
= req
.get_method()
2942 if req_get_method
== 'HEAD':
2943 req_type
= HEADRequest
2944 elif req_get_method
== 'PUT':
2945 req_type
= PUTRequest
2947 req_type
= compat_urllib_request
.Request
2949 req_url
, data
=req_data
, headers
=req_headers
,
2950 origin_req_host
=req
.origin_req_host
, unverifiable
=req
.unverifiable
)
2951 if hasattr(req
, 'timeout'):
2952 new_req
.timeout
= req
.timeout
2956 def _multipart_encode_impl(data
, boundary
):
2957 content_type
= 'multipart/form-data; boundary=%s' % boundary
2960 for k
, v
in data
.items():
2961 out
+= b
'--' + boundary
.encode('ascii') + b
'\r\n'
2962 if isinstance(k
, compat_str
):
2963 k
= k
.encode('utf-8')
2964 if isinstance(v
, compat_str
):
2965 v
= v
.encode('utf-8')
2966 # RFC 2047 requires non-ASCII field names to be encoded, while RFC 7578
2967 # suggests sending UTF-8 directly. Firefox sends UTF-8, too
2968 content
= b
'Content-Disposition: form-data; name="' + k
+ b
'"\r\n\r\n' + v
+ b
'\r\n'
2969 if boundary
.encode('ascii') in content
:
2970 raise ValueError('Boundary overlaps with data')
2973 out
+= b
'--' + boundary
.encode('ascii') + b
'--\r\n'
2975 return out
, content_type
2978 def multipart_encode(data
, boundary
=None):
2980 Encode a dict to RFC 7578-compliant form-data
2983 A dict where keys and values can be either Unicode or bytes-like
2986 If specified a Unicode object, it's used as the boundary. Otherwise
2987 a random boundary is generated.
2989 Reference: https://tools.ietf.org/html/rfc7578
2991 has_specified_boundary
= boundary
is not None
2994 if boundary
is None:
2995 boundary
= '---------------' + str(random
.randrange(0x0fffffff, 0xffffffff))
2998 out
, content_type
= _multipart_encode_impl(data
, boundary
)
3001 if has_specified_boundary
:
3005 return out
, content_type
3008 def dict_get(d
, key_or_keys
, default
=None, skip_false_values
=True):
3009 if isinstance(key_or_keys
, (list, tuple)):
3010 for key
in key_or_keys
:
3011 if key
not in d
or d
[key
] is None or skip_false_values
and not d
[key
]:
3015 return d
.get(key_or_keys
, default
)
3018 def try_get(src
, getter
, expected_type
=None):
3019 for get
in variadic(getter
):
3022 except (AttributeError, KeyError, TypeError, IndexError):
3025 if expected_type
is None or isinstance(v
, expected_type
):
3029 def merge_dicts(*dicts
):
3031 for a_dict
in dicts
:
3032 for k
, v
in a_dict
.items():
3036 or (isinstance(v
, compat_str
) and v
3037 and isinstance(merged
[k
], compat_str
)
3038 and not merged
[k
])):
3043 def encode_compat_str(string
, encoding
=preferredencoding(), errors
='strict'):
3044 return string
if isinstance(string
, compat_str
) else compat_str(string
, encoding
, errors
)
3056 TV_PARENTAL_GUIDELINES
= {
3066 def parse_age_limit(s
):
3068 return s
if 0 <= s
<= 21 else None
3069 if not isinstance(s
, compat_basestring
):
3071 m
= re
.match(r
'^(?P<age>\d{1,2})\+?$', s
)
3073 return int(m
.group('age'))
3076 return US_RATINGS
[s
]
3077 m
= re
.match(r
'^TV[_-]?(%s)$' % '|'.join(k
[3:] for k
in TV_PARENTAL_GUIDELINES
), s
)
3079 return TV_PARENTAL_GUIDELINES
['TV-' + m
.group(1)]
3083 def strip_jsonp(code
):
3086 (?:window\.)?(?P<func_name>[a-zA-Z0-9_.$]*)
3087 (?:\s*&&\s*(?P=func_name))?
3088 \s*\(\s*(?P<callback_data>.*)\);?
3089 \s*?(?://[^\n]*)*$''',
3090 r
'\g<callback_data>', code
)
3093 def js_to_json(code
, vars={}):
3094 # vars is a dict of var, val pairs to substitute
3095 COMMENT_RE
= r
'/\*(?:(?!\*/).)*?\*/|//[^\n]*\n'
3096 SKIP_RE
= r
'\s*(?:{comment})?\s*'.format(comment
=COMMENT_RE
)
3098 (r
'(?s)^(0[xX][0-9a-fA-F]+){skip}:?$'.format(skip
=SKIP_RE
), 16),
3099 (r
'(?s)^(0+[0-7]+){skip}:?$'.format(skip
=SKIP_RE
), 8),
3104 if v
in ('true', 'false', 'null'):
3106 elif v
in ('undefined', 'void 0'):
3108 elif v
.startswith('/*') or v
.startswith('//') or v
.startswith('!') or v
== ',':
3111 if v
[0] in ("'", '"'):
3112 v
= re
.sub(r
'(?s)\\.|"', lambda m
: {
3117 }.get(m
.group(0), m
.group(0)), v
[1:-1])
3119 for regex
, base
in INTEGER_TABLE
:
3120 im
= re
.match(regex
, v
)
3122 i
= int(im
.group(1), base
)
3123 return '"%d":' % i
if v
.endswith(':') else '%d' % i
3130 return re
.sub(r
'''(?sx)
3131 "(?:[^"\\]*(?:\\\\|\\['"nurtbfx/\n]))*[^"\\]*"|
3132 '(?:[^'\\]*(?:\\\\|\\['"nurtbfx/\n]))*[^'\\]*'|
3133 {comment}|,(?={skip}[\]}}])|
3134 void\s0|(?:(?<![0-9])[eE]|[a-df-zA-DF-Z_$])[.a-zA-Z_$0-9]*|
3135 \b(?:0[xX][0-9a-fA-F]+|0+[0-7]+)(?:{skip}:)?|
3138 '''.format(comment
=COMMENT_RE
, skip
=SKIP_RE
), fix_kv
, code
)
3141 def qualities(quality_ids
):
3142 """ Get a numeric quality value out of a list of possible values """
3145 return quality_ids
.index(qid
)
3151 POSTPROCESS_WHEN
= {'pre_process', 'before_dl', 'after_move', 'post_process', 'after_video', 'playlist'}
3155 'default': '%(title)s [%(id)s].%(ext)s',
3156 'chapter': '%(title)s - %(section_number)03d %(section_title)s [%(id)s].%(ext)s',
3162 'description': 'description',
3163 'annotation': 'annotations.xml',
3164 'infojson': 'info.json',
3167 'pl_thumbnail': None,
3168 'pl_description': 'description',
3169 'pl_infojson': 'info.json',
3172 # As of [1] format syntax is:
3173 # %[mapping_key][conversion_flags][minimum_width][.precision][length_modifier]type
3174 # 1. https://docs.python.org/2/library/stdtypes.html#string-formatting
3175 STR_FORMAT_RE_TMPL
= r
'''(?x)
3176 (?<!%)(?P<prefix>(?:%%)*)
3178 (?P<has_key>\((?P<key>{0})\))?
3180 (?P<conversion>[#0\-+ ]+)?
3182 (?P<precision>\.\d+)?
3183 (?P<len_mod>[hlL])? # unused in python
3184 {1} # conversion type
3189 STR_FORMAT_TYPES
= 'diouxXeEfFgGcrs'
3192 def limit_length(s
, length
):
3193 """ Add ellipses to overly long strings """
3198 return s
[:length
- len(ELLIPSES
)] + ELLIPSES
3202 def version_tuple(v
):
3203 return tuple(int(e
) for e
in re
.split(r
'[-.]', v
))
3206 def is_outdated_version(version
, limit
, assume_new
=True):
3208 return not assume_new
3210 return version_tuple(version
) < version_tuple(limit
)
3212 return not assume_new
3215 def ytdl_is_updateable():
3216 """ Returns if yt-dlp can be updated with -U """
3218 from .update
import is_non_updateable
3220 return not is_non_updateable()
3223 def args_to_str(args
):
3224 # Get a short string representation for a subprocess command
3225 return ' '.join(compat_shlex_quote(a
) for a
in args
)
3228 def error_to_compat_str(err
):
3230 # On python 2 error byte string must be decoded with proper
3231 # encoding rather than ascii
3232 if sys
.version_info
[0] < 3:
3233 err_str
= err_str
.decode(preferredencoding())
3237 def mimetype2ext(mt
):
3241 mt
, _
, params
= mt
.partition(';')
3246 # Per RFC 3003, audio/mpeg can be .mp1, .mp2 or .mp3. Here use .mp3 as
3247 # it's the most popular one
3248 'audio/mpeg': 'mp3',
3249 'audio/x-wav': 'wav',
3251 'audio/wave': 'wav',
3254 ext
= FULL_MAP
.get(mt
)
3260 'smptett+xml': 'tt',
3264 'x-mp4-fragmented': 'mp4',
3265 'x-ms-sami': 'sami',
3268 'x-mpegurl': 'm3u8',
3269 'vnd.apple.mpegurl': 'm3u8',
3273 'vnd.ms-sstr+xml': 'ism',
3277 'filmstrip+json': 'fs',
3281 _
, _
, subtype
= mt
.rpartition('/')
3282 ext
= SUBTYPE_MAP
.get(subtype
.lower())
3293 _
, _
, suffix
= subtype
.partition('+')
3294 ext
= SUFFIX_MAP
.get(suffix
)
3298 return subtype
.replace('+', '.')
3301 def ext2mimetype(ext_or_url
):
3304 if '.' not in ext_or_url
:
3305 ext_or_url
= f
'file.{ext_or_url}'
3306 return mimetypes
.guess_type(ext_or_url
)[0]
3309 def parse_codecs(codecs_str
):
3310 # http://tools.ietf.org/html/rfc6381
3313 split_codecs
= list(filter(None, map(
3314 str.strip
, codecs_str
.strip().strip(',').split(','))))
3315 vcodec
, acodec
, tcodec
, hdr
= None, None, None, None
3316 for full_codec
in split_codecs
:
3317 parts
= full_codec
.split('.')
3318 codec
= parts
[0].replace('0', '')
3319 if codec
in ('avc1', 'avc2', 'avc3', 'avc4', 'vp9', 'vp8', 'hev1', 'hev2',
3320 'h263', 'h264', 'mp4v', 'hvc1', 'av1', 'theora', 'dvh1', 'dvhe'):
3322 vcodec
= '.'.join(parts
[:4]) if codec
in ('vp9', 'av1', 'hvc1') else full_codec
3323 if codec
in ('dvh1', 'dvhe'):
3325 elif codec
== 'av1' and len(parts
) > 3 and parts
[3] == '10':
3327 elif full_codec
.replace('0', '').startswith('vp9.2'):
3329 elif codec
in ('flac', 'mp4a', 'opus', 'vorbis', 'mp3', 'aac', 'ac-3', 'ec-3', 'eac3', 'dtsc', 'dtse', 'dtsh', 'dtsl'):
3332 elif codec
in ('stpp', 'wvtt',):
3336 write_string('WARNING: Unknown codec %s\n' % full_codec
, sys
.stderr
)
3337 if vcodec
or acodec
or tcodec
:
3339 'vcodec': vcodec
or 'none',
3340 'acodec': acodec
or 'none',
3341 'dynamic_range': hdr
,
3342 **({'tcodec': tcodec}
if tcodec
is not None else {}),
3344 elif len(split_codecs
) == 2:
3346 'vcodec': split_codecs
[0],
3347 'acodec': split_codecs
[1],
3352 def urlhandle_detect_ext(url_handle
):
3353 getheader
= url_handle
.headers
.get
3355 cd
= getheader('Content-Disposition')
3357 m
= re
.match(r
'attachment;\s*filename="(?P<filename>[^"]+)"', cd
)
3359 e
= determine_ext(m
.group('filename'), default_ext
=None)
3363 return mimetype2ext(getheader('Content-Type'))
3366 def encode_data_uri(data
, mime_type
):
3367 return 'data:%s;base64,%s' % (mime_type
, base64
.b64encode(data
).decode('ascii'))
3370 def age_restricted(content_limit
, age_limit
):
3371 """ Returns True iff the content should be blocked """
3373 if age_limit
is None: # No limit set
3375 if content_limit
is None:
3376 return False # Content available for everyone
3377 return age_limit
< content_limit
3380 def is_html(first_bytes
):
3381 """ Detect whether a file contains HTML by examining its first bytes. """
3384 (b
'\xef\xbb\xbf', 'utf-8'),
3385 (b
'\x00\x00\xfe\xff', 'utf-32-be'),
3386 (b
'\xff\xfe\x00\x00', 'utf-32-le'),
3387 (b
'\xff\xfe', 'utf-16-le'),
3388 (b
'\xfe\xff', 'utf-16-be'),
3390 for bom
, enc
in BOMS
:
3391 if first_bytes
.startswith(bom
):
3392 s
= first_bytes
[len(bom
):].decode(enc
, 'replace')
3395 s
= first_bytes
.decode('utf-8', 'replace')
3397 return re
.match(r
'^\s*<', s
)
3400 def determine_protocol(info_dict
):
3401 protocol
= info_dict
.get('protocol')
3402 if protocol
is not None:
3405 url
= sanitize_url(info_dict
['url'])
3406 if url
.startswith('rtmp'):
3408 elif url
.startswith('mms'):
3410 elif url
.startswith('rtsp'):
3413 ext
= determine_ext(url
)
3419 return compat_urllib_parse_urlparse(url
).scheme
3422 def render_table(header_row
, data
, delim
=False, extra_gap
=0, hide_empty
=False):
3423 """ Render a list of rows, each as a list of values.
3424 Text after a \t will be right aligned """
3426 return len(remove_terminal_sequences(string
).replace('\t', ''))
3428 def get_max_lens(table
):
3429 return [max(width(str(v
)) for v
in col
) for col
in zip(*table
)]
3431 def filter_using_list(row
, filterArray
):
3432 return [col
for (take
, col
) in zip(filterArray
, row
) if take
]
3435 max_lens
= get_max_lens(data
)
3436 header_row
= filter_using_list(header_row
, max_lens
)
3437 data
= [filter_using_list(row
, max_lens
) for row
in data
]
3439 table
= [header_row
] + data
3440 max_lens
= get_max_lens(table
)
3443 table
= [header_row
, [delim
* (ml
+ extra_gap
) for ml
in max_lens
]] + data
3444 table
[1][-1] = table
[1][-1][:-extra_gap
] # Remove extra_gap from end of delimiter
3446 for pos
, text
in enumerate(map(str, row
)):
3448 row
[pos
] = text
.replace('\t', ' ' * (max_lens
[pos
] - width(text
))) + ' ' * extra_gap
3450 row
[pos
] = text
+ ' ' * (max_lens
[pos
] - width(text
) + extra_gap
)
3451 ret
= '\n'.join(''.join(row
).rstrip() for row
in table
)
3455 def _match_one(filter_part
, dct
, incomplete
):
3456 # TODO: Generalize code with YoutubeDL._build_format_filter
3457 STRING_OPERATORS
= {
3458 '*=': operator
.contains
,
3459 '^=': lambda attr
, value
: attr
.startswith(value
),
3460 '$=': lambda attr
, value
: attr
.endswith(value
),
3461 '~=': lambda attr
, value
: re
.search(value
, attr
),
3463 COMPARISON_OPERATORS
= {
3465 '<=': operator
.le
, # "<=" must be defined above "<"
3472 operator_rex
= re
.compile(r
'''(?x)\s*
3474 \s*(?P<negation>!\s*)?(?P<op>%s)(?P<none_inclusive>\s*\?)?\s*
3476 (?P<quote>["\'])(?P
<quotedstrval
>.+?
)(?P
=quote
)|
3480 ''' % '|'.join(map(re.escape, COMPARISON_OPERATORS.keys())))
3481 m = operator_rex.search(filter_part)
3484 unnegated_op = COMPARISON_OPERATORS[m['op']]
3486 op = lambda attr, value: not unnegated_op(attr, value)
3489 comparison_value = m['quotedstrval'] or m['strval'] or m['intval']
3491 comparison_value = comparison_value.replace(r'\%s' % m['quote'], m['quote'])
3492 actual_value = dct.get(m['key'])
3493 numeric_comparison = None
3494 if isinstance(actual_value, compat_numeric_types):
3495 # If the original field is a string and matching comparisonvalue is
3496 # a number we should respect the origin of the original field
3497 # and process comparison value as a string (see
3498 # https://github.com/ytdl-org/youtube-dl/issues/11082)
3500 numeric_comparison = int(comparison_value)
3502 numeric_comparison = parse_filesize(comparison_value)
3503 if numeric_comparison is None:
3504 numeric_comparison = parse_filesize(f'{comparison_value}B')
3505 if numeric_comparison is None:
3506 numeric_comparison = parse_duration(comparison_value)
3507 if numeric_comparison is not None and m['op'] in STRING_OPERATORS:
3508 raise ValueError('Operator %s only supports string values!' % m['op'])
3509 if actual_value is None:
3510 return incomplete or m['none_inclusive']
3511 return op(actual_value, comparison_value if numeric_comparison is None else numeric_comparison)
3514 '': lambda v: (v is True) if isinstance(v, bool) else (v is not None),
3515 '!': lambda v: (v is False) if isinstance(v, bool) else (v is None),
3517 operator_rex = re.compile(r'''(?x
)\s
*
3518 (?P
<op
>%s)\s
*(?P
<key
>[a
-z_
]+)
3520 ''' % '|'.join(map(re.escape, UNARY_OPERATORS.keys())))
3521 m = operator_rex.search(filter_part)
3523 op = UNARY_OPERATORS[m.group('op')]
3524 actual_value = dct.get(m.group('key'))
3525 if incomplete and actual_value is None:
3527 return op(actual_value)
3529 raise ValueError('Invalid filter part %r' % filter_part)
3532 def match_str(filter_str, dct, incomplete=False):
3533 """ Filter a dictionary with a simple string syntax. Returns True (=passes filter) or false
3534 When incomplete, all conditions passes on missing fields
3537 _match_one(filter_part.replace(r'\&', '&'), dct, incomplete)
3538 for filter_part in re.split(r'(?<!\\)&', filter_str))
3541 def match_filter_func(filter_str):
3542 def _match_func(info_dict, *args, **kwargs):
3543 if match_str(filter_str, info_dict, *args, **kwargs):
3546 video_title = info_dict.get('title', info_dict.get('id', 'video'))
3547 return '%s does not pass filter %s, skipping ..' % (video_title, filter_str)
3551 def parse_dfxp_time_expr(time_expr):
3555 mobj = re.match(r'^(?P<time_offset>\d+(?:\.\d+)?)s?$', time_expr)
3557 return float(mobj.group('time_offset'))
3559 mobj = re.match(r'^(\d+):(\d\d):(\d\d(?:(?:\.|:)\d+)?)$', time_expr)
3561 return 3600 * int(mobj.group(1)) + 60 * int(mobj.group(2)) + float(mobj.group(3).replace(':', '.'))
3564 def srt_subtitles_timecode(seconds):
3565 return '%02d:%02d:%02d,%03d' % timetuple_from_msec(seconds * 1000)
3568 def ass_subtitles_timecode(seconds):
3569 time = timetuple_from_msec(seconds * 1000)
3570 return '%01d:%02d:%02d.%02d' % (*time[:-1], time.milliseconds / 10)
3573 def dfxp2srt(dfxp_data):
3575 @param dfxp_data A
bytes-like
object containing DFXP data
3576 @returns A
unicode object containing converted SRT data
3578 LEGACY_NAMESPACES = (
3579 (b'http://www.w3.org/ns/ttml', [
3580 b'http://www.w3.org/2004/11/ttaf1',
3581 b'http://www.w3.org/2006/04/ttaf1',
3582 b'http://www.w3.org/2006/10/ttaf1',
3584 (b'http://www.w3.org/ns/ttml#styling', [
3585 b'http://www.w3.org/ns/ttml#style',
3589 SUPPORTED_STYLING = [
3598 _x = functools.partial(xpath_with_ns, ns_map={
3599 'xml': 'http://www.w3.org/XML/1998/namespace',
3600 'ttml': 'http://www.w3.org/ns/ttml',
3601 'tts': 'http://www.w3.org/ns/ttml#styling',
3607 class TTMLPElementParser(object):
3609 _unclosed_elements = []
3610 _applied_styles = []
3612 def start(self, tag, attrib):
3613 if tag in (_x('ttml:br'), 'br'):
3616 unclosed_elements = []
3618 element_style_id = attrib.get('style')
3620 style.update(default_style)
3621 if element_style_id:
3622 style.update(styles.get(element_style_id, {}))
3623 for prop in SUPPORTED_STYLING:
3624 prop_val = attrib.get(_x('tts:' + prop))
3626 style[prop] = prop_val
3629 for k, v in sorted(style.items()):
3630 if self._applied_styles and self._applied_styles[-1].get(k) == v:
3633 font += ' color="%s"' % v
3634 elif k == 'fontSize':
3635 font += ' size="%s"' % v
3636 elif k == 'fontFamily':
3637 font += ' face="%s"' % v
3638 elif k == 'fontWeight' and v == 'bold':
3640 unclosed_elements.append('b')
3641 elif k == 'fontStyle' and v == 'italic':
3643 unclosed_elements.append('i')
3644 elif k == 'textDecoration' and v == 'underline':
3646 unclosed_elements.append('u')
3648 self._out += '<font' + font + '>'
3649 unclosed_elements.append('font')
3651 if self._applied_styles:
3652 applied_style.update(self._applied_styles[-1])
3653 applied_style.update(style)
3654 self._applied_styles.append(applied_style)
3655 self._unclosed_elements.append(unclosed_elements)
3658 if tag not in (_x('ttml:br'), 'br'):
3659 unclosed_elements = self._unclosed_elements.pop()
3660 for element in reversed(unclosed_elements):
3661 self._out += '</%s>' % element
3662 if unclosed_elements and self._applied_styles:
3663 self._applied_styles.pop()
3665 def data(self, data):
3669 return self._out.strip()
3671 def parse_node(node):
3672 target = TTMLPElementParser()
3673 parser = xml.etree.ElementTree.XMLParser(target=target)
3674 parser.feed(xml.etree.ElementTree.tostring(node))
3675 return parser.close()
3677 for k, v in LEGACY_NAMESPACES:
3679 dfxp_data = dfxp_data.replace(ns, k)
3681 dfxp = compat_etree_fromstring(dfxp_data)
3683 paras = dfxp.findall(_x('.//ttml:p')) or dfxp.findall('.//p')
3686 raise ValueError('Invalid dfxp/TTML subtitle')
3690 for style in dfxp.findall(_x('.//ttml:style')):
3691 style_id = style.get('id') or style.get(_x('xml:id'))
3694 parent_style_id = style.get('style')
3696 if parent_style_id not in styles:
3699 styles[style_id] = styles[parent_style_id].copy()
3700 for prop in SUPPORTED_STYLING:
3701 prop_val = style.get(_x('tts:' + prop))
3703 styles.setdefault(style_id, {})[prop] = prop_val
3709 for p in ('body', 'div'):
3710 ele = xpath_element(dfxp, [_x('.//ttml:' + p), './/' + p])
3713 style = styles.get(ele.get('style'))
3716 default_style.update(style)
3718 for para, index in zip(paras, itertools.count(1)):
3719 begin_time = parse_dfxp_time_expr(para.attrib.get('begin'))
3720 end_time = parse_dfxp_time_expr(para.attrib.get('end'))
3721 dur = parse_dfxp_time_expr(para.attrib.get('dur'))
3722 if begin_time is None:
3727 end_time = begin_time + dur
3728 out.append('%d\n%s --> %s\n%s\n\n' % (
3730 srt_subtitles_timecode(begin_time),
3731 srt_subtitles_timecode(end_time),
3737 def cli_option(params, command_option, param):
3738 param = params.get(param)
3740 param = compat_str(param)
3741 return [command_option, param] if param is not None else []
3744 def cli_bool_option(params, command_option, param, true_value='true', false_value='false', separator=None):
3745 param = params.get(param)
3748 assert isinstance(param, bool)
3750 return [command_option + separator + (true_value if param else false_value)]
3751 return [command_option, true_value if param else false_value]
3754 def cli_valueless_option(params, command_option, param, expected_value=True):
3755 param = params.get(param)
3756 return [command_option] if param == expected_value else []
3759 def cli_configuration_args(argdict, keys, default=[], use_compat=True):
3760 if isinstance(argdict, (list, tuple)): # for backward compatibility
3767 assert isinstance(argdict, dict)
3769 assert isinstance(keys, (list, tuple))
3770 for key_list in keys:
3771 arg_list = list(filter(
3772 lambda x: x is not None,
3773 [argdict.get(key.lower()) for key in variadic(key_list)]))
3775 return [arg for args in arg_list for arg in args]
3779 def _configuration_args(main_key, argdict, exe, keys=None, default=[], use_compat=True):
3780 main_key, exe = main_key.lower(), exe.lower()
3781 root_key = exe if main_key == exe else f'{main_key}+{exe}'
3782 keys = [f'{root_key}{k}' for k in (keys or [''])]
3783 if root_key in keys:
3785 keys.append((main_key, exe))
3786 keys.append('default')
3789 return cli_configuration_args(argdict, keys, default, use_compat)
3792 class ISO639Utils(object):
3793 # See http://www.loc.gov/standards/iso639-2/ISO-639-2_utf-8.txt
3852 'iw': 'heb', # Replaced by he in 1989 revision
3862 'in': 'ind', # Replaced by id in 1989 revision
3977 'ji': 'yid', # Replaced by yi in 1989 revision
3985 def short2long(cls, code):
3986 """Convert language code from ISO 639-1 to ISO 639-2/T"""
3987 return cls._lang_map.get(code[:2])
3990 def long2short(cls, code):
3991 """Convert language code from ISO 639-2/T to ISO 639-1"""
3992 for short_name, long_name in cls._lang_map.items():
3993 if long_name == code:
3997 class ISO3166Utils(object):
3998 # From http://data.okfn.org/data/core/country-list
4000 'AF': 'Afghanistan',
4001 'AX': 'Åland Islands',
4004 'AS': 'American Samoa',
4009 'AG': 'Antigua and Barbuda',
4026 'BO': 'Bolivia, Plurinational State of',
4027 'BQ': 'Bonaire, Sint Eustatius and Saba',
4028 'BA': 'Bosnia and Herzegovina',
4030 'BV': 'Bouvet Island',
4032 'IO': 'British Indian Ocean Territory',
4033 'BN': 'Brunei Darussalam',
4035 'BF': 'Burkina Faso',
4041 'KY': 'Cayman Islands',
4042 'CF': 'Central African Republic',
4046 'CX': 'Christmas Island',
4047 'CC': 'Cocos (Keeling) Islands',
4051 'CD': 'Congo, the Democratic Republic of the',
4052 'CK': 'Cook Islands',
4054 'CI': 'Côte d\'Ivoire',
4059 'CZ': 'Czech Republic',
4063 'DO': 'Dominican Republic',
4066 'SV': 'El Salvador',
4067 'GQ': 'Equatorial Guinea',
4071 'FK': 'Falkland Islands (Malvinas)',
4072 'FO': 'Faroe Islands',
4076 'GF': 'French Guiana',
4077 'PF': 'French Polynesia',
4078 'TF': 'French Southern Territories',
4093 'GW': 'Guinea-Bissau',
4096 'HM': 'Heard Island and McDonald Islands',
4097 'VA': 'Holy See (Vatican City State)',
4104 'IR': 'Iran, Islamic Republic of',
4107 'IM': 'Isle of Man',
4117 'KP': 'Korea, Democratic People\'s Republic of',
4118 'KR': 'Korea, Republic of',
4121 'LA': 'Lao People\'s Democratic Republic',
4127 'LI': 'Liechtenstein',
4131 'MK': 'Macedonia, the Former Yugoslav Republic of',
4138 'MH': 'Marshall Islands',
4144 'FM': 'Micronesia, Federated States of',
4145 'MD': 'Moldova, Republic of',
4156 'NL': 'Netherlands',
4157 'NC': 'New Caledonia',
4158 'NZ': 'New Zealand',
4163 'NF': 'Norfolk Island',
4164 'MP': 'Northern Mariana Islands',
4169 'PS': 'Palestine, State of',
4171 'PG': 'Papua New Guinea',
4174 'PH': 'Philippines',
4178 'PR': 'Puerto Rico',
4182 'RU': 'Russian Federation',
4184 'BL': 'Saint Barthélemy',
4185 'SH': 'Saint Helena, Ascension and Tristan da Cunha',
4186 'KN': 'Saint Kitts and Nevis',
4187 'LC': 'Saint Lucia',
4188 'MF': 'Saint Martin (French part)',
4189 'PM': 'Saint Pierre and Miquelon',
4190 'VC': 'Saint Vincent and the Grenadines',
4193 'ST': 'Sao Tome and Principe',
4194 'SA': 'Saudi Arabia',
4198 'SL': 'Sierra Leone',
4200 'SX': 'Sint Maarten (Dutch part)',
4203 'SB': 'Solomon Islands',
4205 'ZA': 'South Africa',
4206 'GS': 'South Georgia and the South Sandwich Islands',
4207 'SS': 'South Sudan',
4212 'SJ': 'Svalbard and Jan Mayen',
4215 'CH': 'Switzerland',
4216 'SY': 'Syrian Arab Republic',
4217 'TW': 'Taiwan, Province of China',
4219 'TZ': 'Tanzania, United Republic of',
4221 'TL': 'Timor-Leste',
4225 'TT': 'Trinidad and Tobago',
4228 'TM': 'Turkmenistan',
4229 'TC': 'Turks and Caicos Islands',
4233 'AE': 'United Arab Emirates',
4234 'GB': 'United Kingdom',
4235 'US': 'United States',
4236 'UM': 'United States Minor Outlying Islands',
4240 'VE': 'Venezuela, Bolivarian Republic of',
4242 'VG': 'Virgin Islands, British',
4243 'VI': 'Virgin Islands, U.S.',
4244 'WF': 'Wallis and Futuna',
4245 'EH': 'Western Sahara',
4252 def short2full(cls, code):
4253 """Convert an ISO 3166-2 country code to the corresponding full name"""
4254 return cls._country_map.get(code.upper())
4257 class GeoUtils(object):
4258 # Major IPv4 address blocks per country
4260 'AD': '46.172.224.0/19',
4261 'AE': '94.200.0.0/13',
4262 'AF': '149.54.0.0/17',
4263 'AG': '209.59.64.0/18',
4264 'AI': '204.14.248.0/21',
4265 'AL': '46.99.0.0/16',
4266 'AM': '46.70.0.0/15',
4267 'AO': '105.168.0.0/13',
4268 'AP': '182.50.184.0/21',
4269 'AQ': '23.154.160.0/24',
4270 'AR': '181.0.0.0/12',
4271 'AS': '202.70.112.0/20',
4272 'AT': '77.116.0.0/14',
4273 'AU': '1.128.0.0/11',
4274 'AW': '181.41.0.0/18',
4275 'AX': '185.217.4.0/22',
4276 'AZ': '5.197.0.0/16',
4277 'BA': '31.176.128.0/17',
4278 'BB': '65.48.128.0/17',
4279 'BD': '114.130.0.0/16',
4281 'BF': '102.178.0.0/15',
4282 'BG': '95.42.0.0/15',
4283 'BH': '37.131.0.0/17',
4284 'BI': '154.117.192.0/18',
4285 'BJ': '137.255.0.0/16',
4286 'BL': '185.212.72.0/23',
4287 'BM': '196.12.64.0/18',
4288 'BN': '156.31.0.0/16',
4289 'BO': '161.56.0.0/16',
4290 'BQ': '161.0.80.0/20',
4291 'BR': '191.128.0.0/12',
4292 'BS': '24.51.64.0/18',
4293 'BT': '119.2.96.0/19',
4294 'BW': '168.167.0.0/16',
4295 'BY': '178.120.0.0/13',
4296 'BZ': '179.42.192.0/18',
4297 'CA': '99.224.0.0/11',
4298 'CD': '41.243.0.0/16',
4299 'CF': '197.242.176.0/21',
4300 'CG': '160.113.0.0/16',
4301 'CH': '85.0.0.0/13',
4302 'CI': '102.136.0.0/14',
4303 'CK': '202.65.32.0/19',
4304 'CL': '152.172.0.0/14',
4305 'CM': '102.244.0.0/14',
4306 'CN': '36.128.0.0/10',
4307 'CO': '181.240.0.0/12',
4308 'CR': '201.192.0.0/12',
4309 'CU': '152.206.0.0/15',
4310 'CV': '165.90.96.0/19',
4311 'CW': '190.88.128.0/17',
4312 'CY': '31.153.0.0/16',
4313 'CZ': '88.100.0.0/14',
4315 'DJ': '197.241.0.0/17',
4316 'DK': '87.48.0.0/12',
4317 'DM': '192.243.48.0/20',
4318 'DO': '152.166.0.0/15',
4319 'DZ': '41.96.0.0/12',
4320 'EC': '186.68.0.0/15',
4321 'EE': '90.190.0.0/15',
4322 'EG': '156.160.0.0/11',
4323 'ER': '196.200.96.0/20',
4324 'ES': '88.0.0.0/11',
4325 'ET': '196.188.0.0/14',
4326 'EU': '2.16.0.0/13',
4327 'FI': '91.152.0.0/13',
4328 'FJ': '144.120.0.0/16',
4329 'FK': '80.73.208.0/21',
4330 'FM': '119.252.112.0/20',
4331 'FO': '88.85.32.0/19',
4333 'GA': '41.158.0.0/15',
4335 'GD': '74.122.88.0/21',
4336 'GE': '31.146.0.0/16',
4337 'GF': '161.22.64.0/18',
4338 'GG': '62.68.160.0/19',
4339 'GH': '154.160.0.0/12',
4340 'GI': '95.164.0.0/16',
4341 'GL': '88.83.0.0/19',
4342 'GM': '160.182.0.0/15',
4343 'GN': '197.149.192.0/18',
4344 'GP': '104.250.0.0/19',
4345 'GQ': '105.235.224.0/20',
4346 'GR': '94.64.0.0/13',
4347 'GT': '168.234.0.0/16',
4348 'GU': '168.123.0.0/16',
4349 'GW': '197.214.80.0/20',
4350 'GY': '181.41.64.0/18',
4351 'HK': '113.252.0.0/14',
4352 'HN': '181.210.0.0/16',
4353 'HR': '93.136.0.0/13',
4354 'HT': '148.102.128.0/17',
4355 'HU': '84.0.0.0/14',
4356 'ID': '39.192.0.0/10',
4357 'IE': '87.32.0.0/12',
4358 'IL': '79.176.0.0/13',
4359 'IM': '5.62.80.0/20',
4360 'IN': '117.192.0.0/10',
4361 'IO': '203.83.48.0/21',
4362 'IQ': '37.236.0.0/14',
4363 'IR': '2.176.0.0/12',
4364 'IS': '82.221.0.0/16',
4365 'IT': '79.0.0.0/10',
4366 'JE': '87.244.64.0/18',
4367 'JM': '72.27.0.0/17',
4368 'JO': '176.29.0.0/16',
4369 'JP': '133.0.0.0/8',
4370 'KE': '105.48.0.0/12',
4371 'KG': '158.181.128.0/17',
4372 'KH': '36.37.128.0/17',
4373 'KI': '103.25.140.0/22',
4374 'KM': '197.255.224.0/20',
4375 'KN': '198.167.192.0/19',
4376 'KP': '175.45.176.0/22',
4377 'KR': '175.192.0.0/10',
4378 'KW': '37.36.0.0/14',
4379 'KY': '64.96.0.0/15',
4380 'KZ': '2.72.0.0/13',
4381 'LA': '115.84.64.0/18',
4382 'LB': '178.135.0.0/16',
4383 'LC': '24.92.144.0/20',
4384 'LI': '82.117.0.0/19',
4385 'LK': '112.134.0.0/15',
4386 'LR': '102.183.0.0/16',
4387 'LS': '129.232.0.0/17',
4388 'LT': '78.56.0.0/13',
4389 'LU': '188.42.0.0/16',
4390 'LV': '46.109.0.0/16',
4391 'LY': '41.252.0.0/14',
4392 'MA': '105.128.0.0/11',
4393 'MC': '88.209.64.0/18',
4394 'MD': '37.246.0.0/16',
4395 'ME': '178.175.0.0/17',
4396 'MF': '74.112.232.0/21',
4397 'MG': '154.126.0.0/17',
4398 'MH': '117.103.88.0/21',
4399 'MK': '77.28.0.0/15',
4400 'ML': '154.118.128.0/18',
4401 'MM': '37.111.0.0/17',
4402 'MN': '49.0.128.0/17',
4403 'MO': '60.246.0.0/16',
4404 'MP': '202.88.64.0/20',
4405 'MQ': '109.203.224.0/19',
4406 'MR': '41.188.64.0/18',
4407 'MS': '208.90.112.0/22',
4408 'MT': '46.11.0.0/16',
4409 'MU': '105.16.0.0/12',
4410 'MV': '27.114.128.0/18',
4411 'MW': '102.70.0.0/15',
4412 'MX': '187.192.0.0/11',
4413 'MY': '175.136.0.0/13',
4414 'MZ': '197.218.0.0/15',
4415 'NA': '41.182.0.0/16',
4416 'NC': '101.101.0.0/18',
4417 'NE': '197.214.0.0/18',
4418 'NF': '203.17.240.0/22',
4419 'NG': '105.112.0.0/12',
4420 'NI': '186.76.0.0/15',
4421 'NL': '145.96.0.0/11',
4422 'NO': '84.208.0.0/13',
4423 'NP': '36.252.0.0/15',
4424 'NR': '203.98.224.0/19',
4425 'NU': '49.156.48.0/22',
4426 'NZ': '49.224.0.0/14',
4427 'OM': '5.36.0.0/15',
4428 'PA': '186.72.0.0/15',
4429 'PE': '186.160.0.0/14',
4430 'PF': '123.50.64.0/18',
4431 'PG': '124.240.192.0/19',
4432 'PH': '49.144.0.0/13',
4433 'PK': '39.32.0.0/11',
4434 'PL': '83.0.0.0/11',
4435 'PM': '70.36.0.0/20',
4436 'PR': '66.50.0.0/16',
4437 'PS': '188.161.0.0/16',
4438 'PT': '85.240.0.0/13',
4439 'PW': '202.124.224.0/20',
4440 'PY': '181.120.0.0/14',
4441 'QA': '37.210.0.0/15',
4442 'RE': '102.35.0.0/16',
4443 'RO': '79.112.0.0/13',
4444 'RS': '93.86.0.0/15',
4445 'RU': '5.136.0.0/13',
4446 'RW': '41.186.0.0/16',
4447 'SA': '188.48.0.0/13',
4448 'SB': '202.1.160.0/19',
4449 'SC': '154.192.0.0/11',
4450 'SD': '102.120.0.0/13',
4451 'SE': '78.64.0.0/12',
4452 'SG': '8.128.0.0/10',
4453 'SI': '188.196.0.0/14',
4454 'SK': '78.98.0.0/15',
4455 'SL': '102.143.0.0/17',
4456 'SM': '89.186.32.0/19',
4457 'SN': '41.82.0.0/15',
4458 'SO': '154.115.192.0/18',
4459 'SR': '186.179.128.0/17',
4460 'SS': '105.235.208.0/21',
4461 'ST': '197.159.160.0/19',
4462 'SV': '168.243.0.0/16',
4463 'SX': '190.102.0.0/20',
4465 'SZ': '41.84.224.0/19',
4466 'TC': '65.255.48.0/20',
4467 'TD': '154.68.128.0/19',
4468 'TG': '196.168.0.0/14',
4469 'TH': '171.96.0.0/13',
4470 'TJ': '85.9.128.0/18',
4471 'TK': '27.96.24.0/21',
4472 'TL': '180.189.160.0/20',
4473 'TM': '95.85.96.0/19',
4474 'TN': '197.0.0.0/11',
4475 'TO': '175.176.144.0/21',
4476 'TR': '78.160.0.0/11',
4477 'TT': '186.44.0.0/15',
4478 'TV': '202.2.96.0/19',
4479 'TW': '120.96.0.0/11',
4480 'TZ': '156.156.0.0/14',
4481 'UA': '37.52.0.0/14',
4482 'UG': '102.80.0.0/13',
4484 'UY': '167.56.0.0/13',
4485 'UZ': '84.54.64.0/18',
4486 'VA': '212.77.0.0/19',
4487 'VC': '207.191.240.0/21',
4488 'VE': '186.88.0.0/13',
4489 'VG': '66.81.192.0/20',
4490 'VI': '146.226.0.0/16',
4491 'VN': '14.160.0.0/11',
4492 'VU': '202.80.32.0/20',
4493 'WF': '117.20.32.0/21',
4494 'WS': '202.4.32.0/19',
4495 'YE': '134.35.0.0/16',
4496 'YT': '41.242.116.0/22',
4497 'ZA': '41.0.0.0/11',
4498 'ZM': '102.144.0.0/13',
4499 'ZW': '102.177.192.0/18',
4503 def random_ipv4(cls, code_or_block):
4504 if len(code_or_block) == 2:
4505 block = cls._country_ip_map.get(code_or_block.upper())
4509 block = code_or_block
4510 addr, preflen = block.split('/')
4511 addr_min = compat_struct_unpack('!L', socket.inet_aton(addr))[0]
4512 addr_max = addr_min | (0xffffffff >> int(preflen))
4513 return compat_str(socket.inet_ntoa(
4514 compat_struct_pack('!L', random.randint(addr_min, addr_max))))
4517 class PerRequestProxyHandler(compat_urllib_request.ProxyHandler):
4518 def __init__(self, proxies=None):
4519 # Set default handlers
4520 for type in ('http', 'https'):
4521 setattr(self, '%s_open' % type,
4522 lambda r, proxy='__noproxy__', type=type, meth=self.proxy_open:
4523 meth(r, proxy, type))
4524 compat_urllib_request.ProxyHandler.__init__(self, proxies)
4526 def proxy_open(self, req, proxy, type):
4527 req_proxy = req.headers.get('Ytdl-request-proxy')
4528 if req_proxy is not None:
4530 del req.headers['Ytdl-request-proxy']
4532 if proxy == '__noproxy__':
4533 return None # No Proxy
4534 if compat_urlparse.urlparse(proxy).scheme.lower() in ('socks', 'socks4', 'socks4a', 'socks5'):
4535 req.add_header('Ytdl-socks-proxy', proxy)
4536 # yt-dlp's http/https handlers do wrapping the socket with socks
4538 return compat_urllib_request.ProxyHandler.proxy_open(
4539 self, req, proxy, type)
4542 # Both long_to_bytes and bytes_to_long are adapted from PyCrypto, which is
4543 # released into Public Domain
4544 # https://github.com/dlitz/pycrypto/blob/master/lib/Crypto/Util/number.py#L387
4546 def long_to_bytes(n, blocksize=0):
4547 """long_to_bytes(n:long, blocksize:int) : string
4548 Convert a long integer to a byte string.
4550 If optional blocksize is given and greater than zero, pad the front of the
4551 byte string with binary zeros so that the length is a multiple of
4554 # after much testing, this algorithm was deemed to be the fastest
4558 s = compat_struct_pack('>I', n & 0xffffffff) + s
4560 # strip off leading zeros
4561 for i in range(len(s)):
4562 if s[i] != b'\000'[0]:
4565 # only happens when n == 0
4569 # add back some pad bytes. this could be done more efficiently w.r.t. the
4570 # de-padding being done above, but sigh...
4571 if blocksize > 0 and len(s) % blocksize:
4572 s = (blocksize - len(s) % blocksize) * b'\000' + s
4576 def bytes_to_long(s):
4577 """bytes_to_long(string) : long
4578 Convert a byte string to a long integer.
4580 This is (essentially) the inverse of long_to_bytes().
4585 extra = (4 - length % 4)
4586 s = b'\000' * extra + s
4587 length = length + extra
4588 for i in range(0, length, 4):
4589 acc = (acc << 32) + compat_struct_unpack('>I', s[i:i + 4])[0]
4593 def ohdave_rsa_encrypt(data, exponent, modulus):
4595 Implement OHDave
's RSA algorithm. See http://www.ohdave.com/rsa/
4598 data: data to encrypt, bytes-like object
4599 exponent, modulus: parameter e and N of RSA algorithm, both integer
4600 Output: hex string of encrypted data
4602 Limitation: supports one block encryption only
4605 payload = int(binascii.hexlify(data[::-1]), 16)
4606 encrypted = pow(payload, exponent, modulus)
4607 return '%x' % encrypted
4610 def pkcs1pad(data, length):
4612 Padding input data with PKCS#1 scheme
4614 @param {int[]} data input data
4615 @param {int} length target length
4616 @returns {int[]} padded data
4618 if len(data) > length - 11:
4619 raise ValueError('Input data too
long for PKCS
#1 padding')
4621 pseudo_random
= [random
.randint(0, 254) for _
in range(length
- len(data
) - 3)]
4622 return [0, 2] + pseudo_random
+ [0] + data
4625 def encode_base_n(num
, n
, table
=None):
4626 FULL_TABLE
= '0123456789abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ'
4628 table
= FULL_TABLE
[:n
]
4631 raise ValueError('base %d exceeds table length %d' % (n
, len(table
)))
4638 ret
= table
[num
% n
] + ret
4643 def decode_packed_codes(code
):
4644 mobj
= re
.search(PACKED_CODES_RE
, code
)
4645 obfuscated_code
, base
, count
, symbols
= mobj
.groups()
4648 symbols
= symbols
.split('|')
4653 base_n_count
= encode_base_n(count
, base
)
4654 symbol_table
[base_n_count
] = symbols
[count
] or base_n_count
4657 r
'\b(\w+)\b', lambda mobj
: symbol_table
[mobj
.group(0)],
4661 def caesar(s
, alphabet
, shift
):
4666 alphabet
[(alphabet
.index(c
) + shift
) % l
] if c
in alphabet
else c
4671 return caesar(s
, r
'''!"#$%&'()*+,-./0123456789:;<=>?@ABCDEFGHIJKLMNOPQRSTUVWXYZ[\]^_`abcdefghijklmnopqrstuvwxyz{|}~''', 47)
4674 def parse_m3u8_attributes(attrib
):
4676 for (key
, val
) in re
.findall(r
'(?P<key>[A-Z0-9-]+)=(?P<val>"[^"]+"|[^",]+)(?:,|$)', attrib
):
4677 if val
.startswith('"'):
4683 def urshift(val
, n
):
4684 return val
>> n
if val
>= 0 else (val
+ 0x100000000) >> n
4687 # Based on png2str() written by @gdkchan and improved by @yokrysty
4688 # Originally posted at https://github.com/ytdl-org/youtube-dl/issues/9706
4689 def decode_png(png_data
):
4690 # Reference: https://www.w3.org/TR/PNG/
4691 header
= png_data
[8:]
4693 if png_data
[:8] != b
'\x89PNG\x0d\x0a\x1a\x0a' or header
[4:8] != b
'IHDR':
4694 raise IOError('Not a valid PNG file.')
4696 int_map
= {1: '>B', 2: '>H', 4: '>I'}
4697 unpack_integer
= lambda x
: compat_struct_unpack(int_map
[len(x
)], x
)[0]
4702 length
= unpack_integer(header
[:4])
4705 chunk_type
= header
[:4]
4708 chunk_data
= header
[:length
]
4709 header
= header
[length
:]
4711 header
= header
[4:] # Skip CRC
4719 ihdr
= chunks
[0]['data']
4721 width
= unpack_integer(ihdr
[:4])
4722 height
= unpack_integer(ihdr
[4:8])
4726 for chunk
in chunks
:
4727 if chunk
['type'] == b
'IDAT':
4728 idat
+= chunk
['data']
4731 raise IOError('Unable to read PNG data.')
4733 decompressed_data
= bytearray(zlib
.decompress(idat
))
4738 def _get_pixel(idx
):
4743 for y
in range(height
):
4744 basePos
= y
* (1 + stride
)
4745 filter_type
= decompressed_data
[basePos
]
4749 pixels
.append(current_row
)
4751 for x
in range(stride
):
4752 color
= decompressed_data
[1 + basePos
+ x
]
4753 basex
= y
* stride
+ x
4758 left
= _get_pixel(basex
- 3)
4760 up
= _get_pixel(basex
- stride
)
4762 if filter_type
== 1: # Sub
4763 color
= (color
+ left
) & 0xff
4764 elif filter_type
== 2: # Up
4765 color
= (color
+ up
) & 0xff
4766 elif filter_type
== 3: # Average
4767 color
= (color
+ ((left
+ up
) >> 1)) & 0xff
4768 elif filter_type
== 4: # Paeth
4774 c
= _get_pixel(basex
- stride
- 3)
4782 if pa
<= pb
and pa
<= pc
:
4783 color
= (color
+ a
) & 0xff
4785 color
= (color
+ b
) & 0xff
4787 color
= (color
+ c
) & 0xff
4789 current_row
.append(color
)
4791 return width
, height
, pixels
4794 def write_xattr(path
, key
, value
):
4795 # This mess below finds the best xattr tool for the job
4797 # try the pyxattr module...
4800 if hasattr(xattr
, 'set'): # pyxattr
4801 # Unicode arguments are not supported in python-pyxattr until
4803 # See https://github.com/ytdl-org/youtube-dl/issues/5498
4804 pyxattr_required_version
= '0.5.0'
4805 if version_tuple(xattr
.__version
__) < version_tuple(pyxattr_required_version
):
4806 # TODO: fallback to CLI tools
4807 raise XAttrUnavailableError(
4808 'python-pyxattr is detected but is too old. '
4809 'yt-dlp requires %s or above while your version is %s. '
4810 'Falling back to other xattr implementations' % (
4811 pyxattr_required_version
, xattr
.__version
__))
4813 setxattr
= xattr
.set
4815 setxattr
= xattr
.setxattr
4818 setxattr(path
, key
, value
)
4819 except EnvironmentError as e
:
4820 raise XAttrMetadataError(e
.errno
, e
.strerror
)
4823 if compat_os_name
== 'nt':
4824 # Write xattrs to NTFS Alternate Data Streams:
4825 # http://en.wikipedia.org/wiki/NTFS#Alternate_data_streams_.28ADS.29
4826 assert ':' not in key
4827 assert os
.path
.exists(path
)
4829 ads_fn
= path
+ ':' + key
4831 with open(ads_fn
, 'wb') as f
:
4833 except EnvironmentError as e
:
4834 raise XAttrMetadataError(e
.errno
, e
.strerror
)
4836 user_has_setfattr
= check_executable('setfattr', ['--version'])
4837 user_has_xattr
= check_executable('xattr', ['-h'])
4839 if user_has_setfattr
or user_has_xattr
:
4841 value
= value
.decode('utf-8')
4842 if user_has_setfattr
:
4843 executable
= 'setfattr'
4844 opts
= ['-n', key
, '-v', value
]
4845 elif user_has_xattr
:
4846 executable
= 'xattr'
4847 opts
= ['-w', key
, value
]
4849 cmd
= ([encodeFilename(executable
, True)]
4850 + [encodeArgument(o
) for o
in opts
]
4851 + [encodeFilename(path
, True)])
4855 cmd
, stdout
=subprocess
.PIPE
, stderr
=subprocess
.PIPE
, stdin
=subprocess
.PIPE
)
4856 except EnvironmentError as e
:
4857 raise XAttrMetadataError(e
.errno
, e
.strerror
)
4858 stdout
, stderr
= p
.communicate_or_kill()
4859 stderr
= stderr
.decode('utf-8', 'replace')
4860 if p
.returncode
!= 0:
4861 raise XAttrMetadataError(p
.returncode
, stderr
)
4864 # On Unix, and can't find pyxattr, setfattr, or xattr.
4865 if sys
.platform
.startswith('linux'):
4866 raise XAttrUnavailableError(
4867 "Couldn't find a tool to set the xattrs. "
4868 "Install either the python 'pyxattr' or 'xattr' "
4869 "modules, or the GNU 'attr' package "
4870 "(which contains the 'setfattr' tool).")
4872 raise XAttrUnavailableError(
4873 "Couldn't find a tool to set the xattrs. "
4874 "Install either the python 'xattr' module, "
4875 "or the 'xattr' binary.")
4878 def random_birthday(year_field
, month_field
, day_field
):
4879 start_date
= datetime
.date(1950, 1, 1)
4880 end_date
= datetime
.date(1995, 12, 31)
4881 offset
= random
.randint(0, (end_date
- start_date
).days
)
4882 random_date
= start_date
+ datetime
.timedelta(offset
)
4884 year_field
: str(random_date
.year
),
4885 month_field
: str(random_date
.month
),
4886 day_field
: str(random_date
.day
),
4890 # Templates for internet shortcut files, which are plain text files.
4891 DOT_URL_LINK_TEMPLATE
= '''
4896 DOT_WEBLOC_LINK_TEMPLATE
= '''
4897 <?xml version="1.0" encoding="UTF-8"?>
4898 <!DOCTYPE plist PUBLIC "-//Apple//DTD PLIST 1.0//EN" "http://www.apple.com/DTDs/PropertyList-1.0.dtd">
4899 <plist version="1.0">
4902 \t<string>%(url)s</string>
4907 DOT_DESKTOP_LINK_TEMPLATE
= '''
4917 'url': DOT_URL_LINK_TEMPLATE
,
4918 'desktop': DOT_DESKTOP_LINK_TEMPLATE
,
4919 'webloc': DOT_WEBLOC_LINK_TEMPLATE
,
4923 def iri_to_uri(iri
):
4925 Converts an IRI (Internationalized Resource Identifier, allowing Unicode characters) to a URI (Uniform Resource Identifier, ASCII-only).
4927 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.
4930 iri_parts
= compat_urllib_parse_urlparse(iri
)
4932 if '[' in iri_parts
.netloc
:
4933 raise ValueError('IPv6 URIs are not, yet, supported.')
4934 # Querying `.netloc`, when there's only one bracket, also raises a ValueError.
4936 # 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.
4939 if iri_parts
.username
:
4940 net_location
+= compat_urllib_parse_quote(iri_parts
.username
, safe
=r
"!$%&'()*+,~")
4941 if iri_parts
.password
is not None:
4942 net_location
+= ':' + compat_urllib_parse_quote(iri_parts
.password
, safe
=r
"!$%&'()*+,~")
4945 net_location
+= iri_parts
.hostname
.encode('idna').decode('utf-8') # Punycode for Unicode hostnames.
4946 # The 'idna' encoding produces ASCII text.
4947 if iri_parts
.port
is not None and iri_parts
.port
!= 80:
4948 net_location
+= ':' + str(iri_parts
.port
)
4950 return compat_urllib_parse_urlunparse(
4954 compat_urllib_parse_quote_plus(iri_parts
.path
, safe
=r
"!$%&'()*+,/:;=@|~"),
4956 # Unsure about the `safe` argument, since this is a legacy way of handling parameters.
4957 compat_urllib_parse_quote_plus(iri_parts
.params
, safe
=r
"!$%&'()*+,/:;=@|~"),
4959 # Not totally sure about the `safe` argument, since the source does not explicitly mention the query URI component.
4960 compat_urllib_parse_quote_plus(iri_parts
.query
, safe
=r
"!$%&'()*+,/:;=?@{|}~"),
4962 compat_urllib_parse_quote_plus(iri_parts
.fragment
, safe
=r
"!#$%&'()*+,/:;=?@{|}~")))
4964 # Source for `safe` arguments: https://url.spec.whatwg.org/#percent-encoded-bytes.
4967 def to_high_limit_path(path
):
4968 if sys
.platform
in ['win32', 'cygwin']:
4969 # Work around MAX_PATH limitation on Windows. The maximum allowed length for the individual path segments may still be quite limited.
4970 return r
'\\?\ '.rstrip() + os
.path
.abspath(path
)
4975 def format_field(obj
, field
=None, template
='%s', ignore
=(None, ''), default
='', func
=None):
4977 val
= obj
if obj
is not None else default
4979 val
= obj
.get(field
, default
)
4980 if func
and val
not in ignore
:
4982 return template
% val
if val
not in ignore
else default
4985 def clean_podcast_url(url
):
4986 return re
.sub(r
'''(?x)
4990 media\.blubrry\.com| # https://create.blubrry.com/resources/podcast-media-download-statistics/getting-started/
4993 (?:dts|www)\.podtrac\.com/(?:pts/)?redirect\.[0-9a-z]{3,4}| # http://analytics.podtrac.com/how-to-measure
4996 cn\.co| # https://podcorn.com/analytics-prefix/
4997 st\.fm # https://podsights.com/docs/
5002 _HEX_TABLE
= '0123456789abcdef'
5005 def random_uuidv4():
5006 return re
.sub(r
'[xy]', lambda x
: _HEX_TABLE
[random
.randint(0, 15)], 'xxxxxxxx-xxxx-4xxx-yxxx-xxxxxxxxxxxx')
5009 def make_dir(path
, to_screen
=None):
5011 dn
= os
.path
.dirname(path
)
5012 if dn
and not os
.path
.exists(dn
):
5015 except (OSError, IOError) as err
:
5016 if callable(to_screen
) is not None:
5017 to_screen('unable to create directory ' + error_to_compat_str(err
))
5021 def get_executable_path():
5022 from zipimport
import zipimporter
5023 if hasattr(sys
, 'frozen'): # Running from PyInstaller
5024 path
= os
.path
.dirname(sys
.executable
)
5025 elif isinstance(globals().get('__loader__'), zipimporter
): # Running from ZIP
5026 path
= os
.path
.join(os
.path
.dirname(__file__
), '../..')
5028 path
= os
.path
.join(os
.path
.dirname(__file__
), '..')
5029 return os
.path
.abspath(path
)
5032 def load_plugins(name
, suffix
, namespace
):
5035 plugins_spec
= importlib
.util
.spec_from_file_location(
5036 name
, os
.path
.join(get_executable_path(), 'ytdlp_plugins', name
, '__init__.py'))
5037 plugins
= importlib
.util
.module_from_spec(plugins_spec
)
5038 sys
.modules
[plugins_spec
.name
] = plugins
5039 plugins_spec
.loader
.exec_module(plugins
)
5040 for name
in dir(plugins
):
5041 if name
in namespace
:
5043 if not name
.endswith(suffix
):
5045 klass
= getattr(plugins
, name
)
5046 classes
[name
] = namespace
[name
] = klass
5047 except FileNotFoundError
:
5053 obj
, *path_list
, default
=None, expected_type
=None, get_all
=True,
5054 casesense
=True, is_user_input
=False, traverse_string
=False):
5055 ''' Traverse nested list/dict/tuple
5056 @param path_list A list of paths which are checked one by one.
5057 Each path is a list of keys where each key is a string,
5058 a function, a tuple of strings/None or "...".
5059 When a fuction is given, it takes the key as argument and
5060 returns whether the key matches or not. When a tuple is given,
5061 all the keys given in the tuple are traversed, and
5062 "..." traverses all the keys in the object
5063 "None" returns the object without traversal
5064 @param default Default value to return
5065 @param expected_type Only accept final value of this type (Can also be any callable)
5066 @param get_all Return all the values obtained from a path or only the first one
5067 @param casesense Whether to consider dictionary keys as case sensitive
5068 @param is_user_input Whether the keys are generated from user input. If True,
5069 strings are converted to int/slice if necessary
5070 @param traverse_string Whether to traverse inside strings. If True, any
5071 non-compatible object will also be converted into a string
5075 _lower
= lambda k
: (k
.lower() if isinstance(k
, str) else k
)
5076 path_list
= (map(_lower
, variadic(path
)) for path
in path_list
)
5078 def _traverse_obj(obj
, path
, _current_depth
=0):
5080 path
= tuple(variadic(path
))
5081 for i
, key
in enumerate(path
):
5082 if None in (key
, obj
):
5084 if isinstance(key
, (list, tuple)):
5085 obj
= [_traverse_obj(obj
, sub_key
, _current_depth
) for sub_key
in key
]
5088 obj
= (obj
.values() if isinstance(obj
, dict)
5089 else obj
if isinstance(obj
, (list, tuple, LazyList
))
5090 else str(obj
) if traverse_string
else [])
5092 depth
= max(depth
, _current_depth
)
5093 return [_traverse_obj(inner_obj
, path
[i
+ 1:], _current_depth
) for inner_obj
in obj
]
5095 if isinstance(obj
, (list, tuple, LazyList
)):
5096 obj
= enumerate(obj
)
5097 elif isinstance(obj
, dict):
5100 if not traverse_string
:
5104 depth
= max(depth
, _current_depth
)
5105 return [_traverse_obj(v
, path
[i
+ 1:], _current_depth
) for k
, v
in obj
if key(k
)]
5106 elif isinstance(obj
, dict) and not (is_user_input
and key
== ':'):
5107 obj
= (obj
.get(key
) if casesense
or (key
in obj
)
5108 else next((v
for k
, v
in obj
.items() if _lower(k
) == key
), None))
5111 key
= (int_or_none(key
) if ':' not in key
5112 else slice(*map(int_or_none
, key
.split(':'))))
5113 if key
== slice(None):
5114 return _traverse_obj(obj
, (..., *path
[i
+ 1:]), _current_depth
)
5115 if not isinstance(key
, (int, slice)):
5117 if not isinstance(obj
, (list, tuple, LazyList
)):
5118 if not traverse_string
:
5127 if isinstance(expected_type
, type):
5128 type_test
= lambda val
: val
if isinstance(val
, expected_type
) else None
5129 elif expected_type
is not None:
5130 type_test
= expected_type
5132 type_test
= lambda val
: val
5134 for path
in path_list
:
5136 val
= _traverse_obj(obj
, path
)
5139 for _
in range(depth
- 1):
5140 val
= itertools
.chain
.from_iterable(v
for v
in val
if v
is not None)
5141 val
= [v
for v
in map(type_test
, val
) if v
is not None]
5143 return val
if get_all
else val
[0]
5145 val
= type_test(val
)
5151 def traverse_dict(dictn
, keys
, casesense
=True):
5152 write_string('DeprecationWarning: yt_dlp.utils.traverse_dict is deprecated '
5153 'and may be removed in a future version. Use yt_dlp.utils.traverse_obj instead')
5154 return traverse_obj(dictn
, keys
, casesense
=casesense
, is_user_input
=True, traverse_string
=True)
5157 def variadic(x
, allowed_types
=(str, bytes, dict)):
5158 return x
if isinstance(x
, collections
.abc
.Iterable
) and not isinstance(x
, allowed_types
) else (x
,)
5161 # create a JSON Web Signature (jws) with HS256 algorithm
5162 # the resulting format is in JWS Compact Serialization
5163 # implemented following JWT https://www.rfc-editor.org/rfc/rfc7519.html
5164 # implemented following JWS https://www.rfc-editor.org/rfc/rfc7515.html
5165 def jwt_encode_hs256(payload_data
, key
, headers
={}):
5171 header_data
.update(headers
)
5172 header_b64
= base64
.b64encode(json
.dumps(header_data
).encode('utf-8'))
5173 payload_b64
= base64
.b64encode(json
.dumps(payload_data
).encode('utf-8'))
5174 h
= hmac
.new(key
.encode('utf-8'), header_b64
+ b
'.' + payload_b64
, hashlib
.sha256
)
5175 signature_b64
= base64
.b64encode(h
.digest())
5176 token
= header_b64
+ b
'.' + payload_b64
+ b
'.' + signature_b64
5180 # can be extended in future to verify the signature and parse header and return the algorithm used if it's not HS256
5181 def jwt_decode_hs256(jwt
):
5182 header_b64
, payload_b64
, signature_b64
= jwt
.split('.')
5183 payload_data
= json
.loads(base64
.urlsafe_b64decode(payload_b64
))
5187 def supports_terminal_sequences(stream
):
5188 if compat_os_name
== 'nt':
5189 from .compat
import WINDOWS_VT_MODE
# Must be imported locally
5190 if not WINDOWS_VT_MODE
or get_windows_version() < (10, 0, 10586):
5192 elif not os
.getenv('TERM'):
5195 return stream
.isatty()
5196 except BaseException
:
5200 _terminal_sequences_re
= re
.compile('\033\\[[^m]+m')
5203 def remove_terminal_sequences(string
):
5204 return _terminal_sequences_re
.sub('', string
)
5207 def number_of_digits(number
):
5208 return len('%d' % number
)
5211 def join_nonempty(*values
, delim
='-', from_dict
=None):
5212 if from_dict
is not None:
5213 values
= map(from_dict
.get
, values
)
5214 return delim
.join(map(str, filter(None, values
)))
5220 __initialized
= False
5222 def __init__(self
, parser
, label
=None):
5223 self
._parser
, self
.label
= parser
, label
5224 self
._loaded
_paths
, self
.configs
= set(), []
5226 def init(self
, args
=None, filename
=None):
5227 assert not self
.__initialized
5229 location
= os
.path
.realpath(filename
)
5230 if location
in self
._loaded
_paths
:
5232 self
._loaded
_paths
.add(location
)
5234 self
.__initialized
= True
5235 self
.own_args
, self
.filename
= args
, filename
5236 for location
in self
._parser
.parse_args(args
)[0].config_locations
or []:
5237 location
= compat_expanduser(location
)
5238 if os
.path
.isdir(location
):
5239 location
= os
.path
.join(location
, 'yt-dlp.conf')
5240 if not os
.path
.exists(location
):
5241 self
._parser
.error(f
'config location {location} does not exist')
5242 self
.append_config(self
.read_file(location
), location
)
5246 label
= join_nonempty(
5247 self
.label
, 'config', f
'"{self.filename}"' if self
.filename
else '',
5249 return join_nonempty(
5250 self
.own_args
is not None and f
'{label[0].upper()}{label[1:]}: {self.hide_login_info(self.own_args)}',
5251 *(f
'\n{c}'.replace('\n', '\n| ')[1:] for c
in self
.configs
),
5255 def read_file(filename
, default
=[]):
5257 optionf
= open(filename
)
5259 return default
# silently skip if file is not present
5261 # FIXME: https://github.com/ytdl-org/youtube-dl/commit/dfe5fa49aed02cf36ba9f743b11b0903554b5e56
5262 contents
= optionf
.read()
5263 if sys
.version_info
< (3,):
5264 contents
= contents
.decode(preferredencoding())
5265 res
= compat_shlex_split(contents
, comments
=True)
5271 def hide_login_info(opts
):
5272 PRIVATE_OPTS
= set(['-p', '--password', '-u', '--username', '--video-password', '--ap-password', '--ap-username'])
5273 eqre
= re
.compile('^(?P<key>' + ('|'.join(re
.escape(po
) for po
in PRIVATE_OPTS
)) + ')=.+$')
5278 return m
.group('key') + '=PRIVATE'
5282 opts
= list(map(_scrub_eq
, opts
))
5283 for idx
, opt
in enumerate(opts
):
5284 if opt
in PRIVATE_OPTS
and idx
+ 1 < len(opts
):
5285 opts
[idx
+ 1] = 'PRIVATE'
5288 def append_config(self
, *args
, label
=None):
5289 config
= type(self
)(self
._parser
, label
)
5290 config
._loaded
_paths
= self
._loaded
_paths
5291 if config
.init(*args
):
5292 self
.configs
.append(config
)
5296 for config
in reversed(self
.configs
):
5297 yield from config
.all_args
5298 yield from self
.own_args
or []
5300 def parse_args(self
):
5301 return self
._parser
.parse_args(list(self
.all_args
))