4 from __future__
import unicode_literals
39 import xml
.etree
.ElementTree
43 compat_HTMLParseError
,
49 compat_ctypes_WINFUNCTYPE
,
50 compat_etree_fromstring
,
53 compat_html_entities_html5
,
66 compat_urllib_parse_urlencode
,
67 compat_urllib_parse_urlparse
,
68 compat_urllib_parse_urlunparse
,
69 compat_urllib_parse_quote
,
70 compat_urllib_parse_quote_plus
,
71 compat_urllib_parse_unquote_plus
,
72 compat_urllib_request
,
83 def register_socks_protocols():
84 # "Register" SOCKS protocols
85 # In Python < 2.6.5, urlsplit() suffers from bug https://bugs.python.org/issue7904
86 # URLs with protocols not in urlparse.uses_netloc are not handled correctly
87 for scheme
in ('socks', 'socks4', 'socks4a', 'socks5'):
88 if scheme
not in compat_urlparse
.uses_netloc
:
89 compat_urlparse
.uses_netloc
.append(scheme
)
92 # This is not clearly defined otherwise
93 compiled_regex_type
= type(re
.compile(''))
96 def random_user_agent():
97 _USER_AGENT_TPL
= 'Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/%s Safari/537.36'
1676 return _USER_AGENT_TPL
% random
.choice(_CHROME_VERSIONS
)
1680 'User-Agent': random_user_agent(),
1681 'Accept-Charset': 'ISO-8859-1,utf-8;q=0.7,*;q=0.7',
1682 'Accept': 'text/html,application/xhtml+xml,application/xml;q=0.9,*/*;q=0.8',
1683 'Accept-Encoding': 'gzip, deflate',
1684 'Accept-Language': 'en-us,en;q=0.5',
1689 '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',
1693 NO_DEFAULT
= object()
1695 ENGLISH_MONTH_NAMES
= [
1696 'January', 'February', 'March', 'April', 'May', 'June',
1697 'July', 'August', 'September', 'October', 'November', 'December']
1700 'en': ENGLISH_MONTH_NAMES
,
1702 'janvier', 'février', 'mars', 'avril', 'mai', 'juin',
1703 'juillet', 'août', 'septembre', 'octobre', 'novembre', 'décembre'],
1706 KNOWN_EXTENSIONS
= (
1707 'mp4', 'm4a', 'm4p', 'm4b', 'm4r', 'm4v', 'aac',
1708 'flv', 'f4v', 'f4a', 'f4b',
1709 'webm', 'ogg', 'ogv', 'oga', 'ogx', 'spx', 'opus',
1710 'mkv', 'mka', 'mk3d',
1713 'asf', 'wmv', 'wma',
1719 'f4f', 'f4m', 'm3u8', 'smil')
1721 # needed for sanitizing filenames in restricted mode
1722 ACCENT_CHARS
= dict(zip('ÂÃÄÀÁÅÆÇÈÉÊËÌÍÎÏÐÑÒÓÔÕÖŐØŒÙÚÛÜŰÝÞßàáâãäåæçèéêëìíîïðñòóôõöőøœùúûüűýþÿ',
1723 itertools
.chain('AAAAAA', ['AE'], 'CEEEEIIIIDNOOOOOOO', ['OE'], 'UUUUUY', ['TH', 'ss'],
1724 'aaaaaa', ['ae'], 'ceeeeiiiionooooooo', ['oe'], 'uuuuuy', ['th'], 'y')))
1748 '%Y/%m/%d %H:%M:%S',
1752 '%Y-%m-%d %H:%M:%S',
1753 '%Y-%m-%d %H:%M:%S.%f',
1754 '%Y-%m-%d %H:%M:%S:%f',
1757 '%Y-%m-%dT%H:%M:%SZ',
1758 '%Y-%m-%dT%H:%M:%S.%fZ',
1759 '%Y-%m-%dT%H:%M:%S.%f0Z',
1760 '%Y-%m-%dT%H:%M:%S',
1761 '%Y-%m-%dT%H:%M:%S.%f',
1763 '%b %d %Y at %H:%M',
1764 '%b %d %Y at %H:%M:%S',
1765 '%B %d %Y at %H:%M',
1766 '%B %d %Y at %H:%M:%S',
1770 DATE_FORMATS_DAY_FIRST
= list(DATE_FORMATS
)
1771 DATE_FORMATS_DAY_FIRST
.extend([
1777 '%d/%m/%Y %H:%M:%S',
1780 DATE_FORMATS_MONTH_FIRST
= list(DATE_FORMATS
)
1781 DATE_FORMATS_MONTH_FIRST
.extend([
1786 '%m/%d/%Y %H:%M:%S',
1789 PACKED_CODES_RE
= r
"}\('(.+)',(\d+),(\d+),'([^']+)'\.split\('\|'\)"
1790 JSON_LD_RE
= r
'(?is)<script[^>]+type=(["\']?
)application
/ld\
+json\
1[^
>]*>(?P
<json_ld
>.+?
)</script
>'
1793 def preferredencoding():
1794 """Get preferred encoding.
1796 Returns the best encoding scheme for the system, based on
1797 locale.getpreferredencoding() and some further tweaks.
1800 pref = locale.getpreferredencoding()
1808 def write_json_file(obj, fn):
1809 """ Encode obj as JSON and write it to fn, atomically if possible """
1811 fn = encodeFilename(fn)
1812 if sys.version_info < (3, 0) and sys.platform != 'win32
':
1813 encoding = get_filesystem_encoding()
1814 # os.path.basename returns a bytes object, but NamedTemporaryFile
1815 # will fail if the filename contains non ascii characters unless we
1816 # use a unicode object
1817 path_basename = lambda f: os.path.basename(fn).decode(encoding)
1818 # the same for os.path.dirname
1819 path_dirname = lambda f: os.path.dirname(fn).decode(encoding)
1821 path_basename = os.path.basename
1822 path_dirname = os.path.dirname
1826 'prefix
': path_basename(fn) + '.',
1827 'dir': path_dirname(fn),
1831 # In Python 2.x, json.dump expects a bytestream.
1832 # In Python 3.x, it writes to a character stream
1833 if sys.version_info < (3, 0):
1838 'encoding
': 'utf
-8',
1841 tf = tempfile.NamedTemporaryFile(**compat_kwargs(args))
1846 if sys.platform == 'win32
':
1847 # Need to remove existing file on Windows, else os.rename raises
1848 # WindowsError or FileExistsError.
1856 os.chmod(tf.name, 0o666 & ~mask)
1859 os.rename(tf.name, fn)
1868 if sys.version_info >= (2, 7):
1869 def find_xpath_attr(node, xpath, key, val=None):
1870 """ Find the xpath xpath[@key=val] """
1871 assert re.match(r'^
[a
-zA
-Z_
-]+$
', key)
1872 expr = xpath + ('[@%s]' % key if val is None else "[@%s='%s']" % (key, val))
1873 return node.find(expr)
1875 def find_xpath_attr(node, xpath, key, val=None):
1876 for f in node.findall(compat_xpath(xpath)):
1877 if key not in f.attrib:
1879 if val is None or f.attrib.get(key) == val:
1883 # On python2.6 the xml.etree.ElementTree.Element methods don't support
1884 # the namespace parameter
1887 def xpath_with_ns(path
, ns_map
):
1888 components
= [c
.split(':') for c
in path
.split('/')]
1890 for c
in components
:
1892 replaced
.append(c
[0])
1895 replaced
.append('{%s}%s' % (ns_map
[ns
], tag
))
1896 return '/'.join(replaced
)
1899 def xpath_element(node
, xpath
, name
=None, fatal
=False, default
=NO_DEFAULT
):
1900 def _find_xpath(xpath
):
1901 return node
.find(compat_xpath(xpath
))
1903 if isinstance(xpath
, (str, compat_str
)):
1904 n
= _find_xpath(xpath
)
1912 if default
is not NO_DEFAULT
:
1915 name
= xpath
if name
is None else name
1916 raise ExtractorError('Could not find XML element %s' % name
)
1922 def xpath_text(node
, xpath
, name
=None, fatal
=False, default
=NO_DEFAULT
):
1923 n
= xpath_element(node
, xpath
, name
, fatal
=fatal
, default
=default
)
1924 if n
is None or n
== default
:
1927 if default
is not NO_DEFAULT
:
1930 name
= xpath
if name
is None else name
1931 raise ExtractorError('Could not find XML element\'s text %s' % name
)
1937 def xpath_attr(node
, xpath
, key
, name
=None, fatal
=False, default
=NO_DEFAULT
):
1938 n
= find_xpath_attr(node
, xpath
, key
)
1940 if default
is not NO_DEFAULT
:
1943 name
= '%s[@%s]' % (xpath
, key
) if name
is None else name
1944 raise ExtractorError('Could not find XML attribute %s' % name
)
1947 return n
.attrib
[key
]
1950 def get_element_by_id(id, html
):
1951 """Return the content of the tag with the specified ID in the passed HTML document"""
1952 return get_element_by_attribute('id', id, html
)
1955 def get_element_by_class(class_name
, html
):
1956 """Return the content of the first tag with the specified class in the passed HTML document"""
1957 retval
= get_elements_by_class(class_name
, html
)
1958 return retval
[0] if retval
else None
1961 def get_element_by_attribute(attribute
, value
, html
, escape_value
=True):
1962 retval
= get_elements_by_attribute(attribute
, value
, html
, escape_value
)
1963 return retval
[0] if retval
else None
1966 def get_elements_by_class(class_name
, html
):
1967 """Return the content of all tags with the specified class in the passed HTML document as a list"""
1968 return get_elements_by_attribute(
1969 'class', r
'[^\'"]*\b%s\b[^\'"]*' % re.escape(class_name),
1970 html, escape_value=False)
1973 def get_elements_by_attribute(attribute, value, html, escape_value=True):
1974 """Return the content of the tag with the specified attribute in the passed HTML document"""
1976 value = re.escape(value) if escape_value else value
1979 for m in re.finditer(r'''(?xs)
1981 (?:\s+[a-zA-Z0-9:._-]+(?:=[a-zA-Z0-9:._-]*|="[^"]*"|='[^
']*'|
))*?
1983 (?:\s+[a-zA-Z0-9:._-]+(?:=[a-zA-Z0-9:._-]*|="[^
"]*"|
='[^']*'|))*?
1987 ''' % (re.escape(attribute), value), html):
1988 res = m.group('content
')
1990 if res.startswith('"') or res.startswith("'"):
1993 retlist.append(unescapeHTML(res))
1998 class HTMLAttributeParser(compat_HTMLParser):
1999 """Trivial HTML parser to gather the attributes for a single element"""
2003 compat_HTMLParser.__init__(self)
2005 def handle_starttag(self, tag, attrs):
2006 self.attrs = dict(attrs)
2009 class HTMLListAttrsParser(compat_HTMLParser):
2010 """HTML parser to gather the attributes for the elements of a list"""
2013 compat_HTMLParser.__init__(self)
2017 def handle_starttag(self, tag, attrs):
2018 if tag == 'li
' and self._level == 0:
2019 self.items.append(dict(attrs))
2022 def handle_endtag(self, tag):
2026 def extract_attributes(html_element):
2027 """Given a string for an HTML element such as
2029 a="foo" B="bar" c="&98;az" d=boz
2030 empty= noval entity="&"
2033 Decode and return a dictionary of attributes.
2035 'a
': 'foo
', 'b
': 'bar
', c: 'baz
', d: 'boz
',
2036 'empty
': '', 'noval
': None, 'entity
': '&',
2037 'sq
': '"', 'dq': '\''
2039 NB HTMLParser is stricter in Python 2.6 & 3.2 than in later versions,
2040 but the cases in the unit test will work for all of 2.6, 2.7, 3.2-3.5.
2042 parser = HTMLAttributeParser()
2044 parser.feed(html_element)
2046 # Older Python may throw HTMLParseError in case of malformed HTML
2047 except compat_HTMLParseError:
2052 def parse_list(webpage):
2053 """Given a string for an series of HTML <li> elements,
2054 return a dictionary of their attributes"""
2055 parser = HTMLListAttrsParser()
2056 parser.feed(webpage)
2061 def clean_html(html):
2062 """Clean an HTML snippet into a readable string"""
2064 if html is None: # Convenience for sanitizing descriptions etc.
2068 html = html.replace('\n', ' ')
2069 html = re.sub(r'(?u)\s*<\s*br\s*/?\s*>\s*', '\n', html)
2070 html = re.sub(r'(?u)<\s*/\s*p\s*>\s*<\s*p[^>]*>', '\n', html)
2072 html = re.sub('<.*?>', '', html)
2073 # Replace html entities
2074 html = unescapeHTML(html)
2078 def sanitize_open(filename, open_mode):
2079 """Try to open the given filename, and slightly tweak it if this fails.
2081 Attempts to open the given filename. If this fails, it tries to change
2082 the filename slightly, step by step, until it's either able to open it
2083 or it fails and raises a final exception, like the standard open()
2086 It returns the tuple (stream, definitive_file_name).
2090 if sys.platform == 'win32':
2092 msvcrt.setmode(sys.stdout.fileno(), os.O_BINARY)
2093 return (sys.stdout.buffer if hasattr(sys.stdout, 'buffer') else sys.stdout, filename)
2094 stream = open(encodeFilename(filename), open_mode)
2095 return (stream, filename)
2096 except (IOError, OSError) as err:
2097 if err.errno in (errno.EACCES,):
2100 # In case of error, try to remove win32 forbidden chars
2101 alt_filename = sanitize_path(filename)
2102 if alt_filename == filename:
2105 # An exception here should be caught in the caller
2106 stream = open(encodeFilename(alt_filename), open_mode)
2107 return (stream, alt_filename)
2110 def timeconvert(timestr):
2111 """Convert RFC 2822 defined time string into system timestamp"""
2113 timetuple = email.utils.parsedate_tz(timestr)
2114 if timetuple is not None:
2115 timestamp = email.utils.mktime_tz(timetuple)
2119 def sanitize_filename(s, restricted=False, is_id=False):
2120 """Sanitizes a string so it could be used as part of a filename.
2121 If restricted is set, use a stricter subset of allowed characters.
2122 Set is_id if this is not an arbitrary string, but an ID that should be kept
2125 def replace_insane(char):
2126 if restricted and char in ACCENT_CHARS:
2127 return ACCENT_CHARS[char]
2128 elif not restricted and char == '\n':
2130 elif char == '?' or ord(char) < 32 or ord(char) == 127:
2133 return '' if restricted else '\''
2135 return '_
-' if restricted else ' -'
2136 elif char in '\\/|
*<>':
2138 if restricted and (char in '!&\'()[]{}$
;`^
,#' or char.isspace()):
2140 if restricted
and ord(char
) > 127:
2147 s
= re
.sub(r
'[0-9]+(?::[0-9]+)+', lambda m
: m
.group(0).replace(':', '_'), s
)
2148 result
= ''.join(map(replace_insane
, s
))
2150 while '__' in result
:
2151 result
= result
.replace('__', '_')
2152 result
= result
.strip('_')
2153 # Common case of "Foreign band name - English song title"
2154 if restricted
and result
.startswith('-_'):
2156 if result
.startswith('-'):
2157 result
= '_' + result
[len('-'):]
2158 result
= result
.lstrip('.')
2164 def sanitize_path(s
, force
=False):
2165 """Sanitizes and normalizes path on Windows"""
2166 if sys
.platform
== 'win32':
2168 drive_or_unc
, _
= os
.path
.splitdrive(s
)
2169 if sys
.version_info
< (2, 7) and not drive_or_unc
:
2170 drive_or_unc
, _
= os
.path
.splitunc(s
)
2176 norm_path
= os
.path
.normpath(remove_start(s
, drive_or_unc
)).split(os
.path
.sep
)
2180 path_part
if path_part
in ['.', '..'] else re
.sub(r
'(?:[/<>:"\|\\?\*]|[\s.]$)', '#', path_part
)
2181 for path_part
in norm_path
]
2183 sanitized_path
.insert(0, drive_or_unc
+ os
.path
.sep
)
2184 elif force
and s
[0] == os
.path
.sep
:
2185 sanitized_path
.insert(0, os
.path
.sep
)
2186 return os
.path
.join(*sanitized_path
)
2189 def sanitize_url(url
):
2190 # Prepend protocol-less URLs with `http:` scheme in order to mitigate
2191 # the number of unwanted failures due to missing protocol
2192 if url
.startswith('//'):
2193 return 'http:%s' % url
2194 # Fix some common typos seen so far
2196 # https://github.com/ytdl-org/youtube-dl/issues/15649
2197 (r
'^httpss://', r
'https://'),
2198 # https://bx1.be/lives/direct-tv/
2199 (r
'^rmtp([es]?)://', r
'rtmp\1://'),
2201 for mistake
, fixup
in COMMON_TYPOS
:
2202 if re
.match(mistake
, url
):
2203 return re
.sub(mistake
, fixup
, url
)
2207 def extract_basic_auth(url
):
2208 parts
= compat_urlparse
.urlsplit(url
)
2209 if parts
.username
is None:
2211 url
= compat_urlparse
.urlunsplit(parts
._replace
(netloc
=(
2212 parts
.hostname
if parts
.port
is None
2213 else '%s:%d' % (parts
.hostname
, parts
.port
))))
2214 auth_payload
= base64
.b64encode(
2215 ('%s:%s' % (parts
.username
, parts
.password
or '')).encode('utf-8'))
2216 return url
, 'Basic ' + auth_payload
.decode('utf-8')
2219 def sanitized_Request(url
, *args
, **kwargs
):
2220 url
, auth_header
= extract_basic_auth(escape_url(sanitize_url(url
)))
2221 if auth_header
is not None:
2222 headers
= args
[1] if len(args
) >= 2 else kwargs
.setdefault('headers', {})
2223 headers
['Authorization'] = auth_header
2224 return compat_urllib_request
.Request(url
, *args
, **kwargs
)
2228 """Expand shell variables and ~"""
2229 return os
.path
.expandvars(compat_expanduser(s
))
2232 def orderedSet(iterable
):
2233 """ Remove all duplicates from the input iterable """
2241 def _htmlentity_transform(entity_with_semicolon
):
2242 """Transforms an HTML entity to a character."""
2243 entity
= entity_with_semicolon
[:-1]
2245 # Known non-numeric HTML entity
2246 if entity
in compat_html_entities
.name2codepoint
:
2247 return compat_chr(compat_html_entities
.name2codepoint
[entity
])
2249 # TODO: HTML5 allows entities without a semicolon. For example,
2250 # 'Éric' should be decoded as 'Éric'.
2251 if entity_with_semicolon
in compat_html_entities_html5
:
2252 return compat_html_entities_html5
[entity_with_semicolon
]
2254 mobj
= re
.match(r
'#(x[0-9a-fA-F]+|[0-9]+)', entity
)
2255 if mobj
is not None:
2256 numstr
= mobj
.group(1)
2257 if numstr
.startswith('x'):
2259 numstr
= '0%s' % numstr
2262 # See https://github.com/ytdl-org/youtube-dl/issues/7518
2264 return compat_chr(int(numstr
, base
))
2268 # Unknown entity in name, return its literal representation
2269 return '&%s;' % entity
2272 def unescapeHTML(s
):
2275 assert type(s
) == compat_str
2278 r
'&([^&;]+;)', lambda m
: _htmlentity_transform(m
.group(1)), s
)
2281 def escapeHTML(text
):
2284 .replace('&', '&')
2285 .replace('<', '<')
2286 .replace('>', '>')
2287 .replace('"', '"')
2288 .replace("'", ''')
2292 def process_communicate_or_kill(p
, *args
, **kwargs
):
2294 return p
.communicate(*args
, **kwargs
)
2295 except BaseException
: # Including KeyboardInterrupt
2301 class Popen(subprocess
.Popen
):
2302 if sys
.platform
== 'win32':
2303 _startupinfo
= subprocess
.STARTUPINFO()
2304 _startupinfo
.dwFlags |
= subprocess
.STARTF_USESHOWWINDOW
2308 def __init__(self
, *args
, **kwargs
):
2309 super(Popen
, self
).__init
__(*args
, **kwargs
, startupinfo
=self
._startupinfo
)
2311 def communicate_or_kill(self
, *args
, **kwargs
):
2312 return process_communicate_or_kill(self
, *args
, **kwargs
)
2315 def get_subprocess_encoding():
2316 if sys
.platform
== 'win32' and sys
.getwindowsversion()[0] >= 5:
2317 # For subprocess calls, encode with locale encoding
2318 # Refer to http://stackoverflow.com/a/9951851/35070
2319 encoding
= preferredencoding()
2321 encoding
= sys
.getfilesystemencoding()
2322 if encoding
is None:
2327 def encodeFilename(s
, for_subprocess
=False):
2329 @param s The name of the file
2332 assert type(s
) == compat_str
2334 # Python 3 has a Unicode API
2335 if sys
.version_info
>= (3, 0):
2338 # Pass '' directly to use Unicode APIs on Windows 2000 and up
2339 # (Detecting Windows NT 4 is tricky because 'major >= 4' would
2340 # match Windows 9x series as well. Besides, NT 4 is obsolete.)
2341 if not for_subprocess
and sys
.platform
== 'win32' and sys
.getwindowsversion()[0] >= 5:
2344 # Jython assumes filenames are Unicode strings though reported as Python 2.x compatible
2345 if sys
.platform
.startswith('java'):
2348 return s
.encode(get_subprocess_encoding(), 'ignore')
2351 def decodeFilename(b
, for_subprocess
=False):
2353 if sys
.version_info
>= (3, 0):
2356 if not isinstance(b
, bytes):
2359 return b
.decode(get_subprocess_encoding(), 'ignore')
2362 def encodeArgument(s
):
2363 if not isinstance(s
, compat_str
):
2364 # Legacy code that uses byte strings
2365 # Uncomment the following line after fixing all post processors
2366 # assert False, 'Internal error: %r should be of type %r, is %r' % (s, compat_str, type(s))
2367 s
= s
.decode('ascii')
2368 return encodeFilename(s
, True)
2371 def decodeArgument(b
):
2372 return decodeFilename(b
, True)
2375 def decodeOption(optval
):
2378 if isinstance(optval
, bytes):
2379 optval
= optval
.decode(preferredencoding())
2381 assert isinstance(optval
, compat_str
)
2385 _timetuple
= collections
.namedtuple('Time', ('hours', 'minutes', 'seconds', 'milliseconds'))
2388 def timetuple_from_msec(msec
):
2389 secs
, msec
= divmod(msec
, 1000)
2390 mins
, secs
= divmod(secs
, 60)
2391 hrs
, mins
= divmod(mins
, 60)
2392 return _timetuple(hrs
, mins
, secs
, msec
)
2395 def formatSeconds(secs
, delim
=':', msec
=False):
2396 time
= timetuple_from_msec(secs
* 1000)
2398 ret
= '%d%s%02d%s%02d' % (time
.hours
, delim
, time
.minutes
, delim
, time
.seconds
)
2400 ret
= '%d%s%02d' % (time
.minutes
, delim
, time
.seconds
)
2402 ret
= '%d' % time
.seconds
2403 return '%s.%03d' % (ret
, time
.milliseconds
) if msec
else ret
2406 def _ssl_load_windows_store_certs(ssl_context
, storename
):
2407 # Code adapted from _load_windows_store_certs in https://github.com/python/cpython/blob/main/Lib/ssl.py
2409 certs
= [cert
for cert
, encoding
, trust
in ssl
.enum_certificates(storename
)
2410 if encoding
== 'x509_asn' and (
2411 trust
is True or ssl
.Purpose
.SERVER_AUTH
.oid
in trust
)]
2412 except PermissionError
:
2416 ssl_context
.load_verify_locations(cadata
=cert
)
2417 except ssl
.SSLError
:
2421 def make_HTTPS_handler(params
, **kwargs
):
2422 opts_check_certificate
= not params
.get('nocheckcertificate')
2423 context
= ssl
.SSLContext(ssl
.PROTOCOL_TLS_CLIENT
)
2424 context
.check_hostname
= opts_check_certificate
2425 context
.verify_mode
= ssl
.CERT_REQUIRED
if opts_check_certificate
else ssl
.CERT_NONE
2426 if opts_check_certificate
:
2428 context
.load_default_certs()
2429 # Work around the issue in load_default_certs when there are bad certificates. See:
2430 # https://github.com/yt-dlp/yt-dlp/issues/1060,
2431 # https://bugs.python.org/issue35665, https://bugs.python.org/issue45312
2432 except ssl
.SSLError
:
2433 # enum_certificates is not present in mingw python. See https://github.com/yt-dlp/yt-dlp/issues/1151
2434 if sys
.platform
== 'win32' and hasattr(ssl
, 'enum_certificates'):
2435 # Create a new context to discard any certificates that were already loaded
2436 context
= ssl
.SSLContext(ssl
.PROTOCOL_TLS_CLIENT
)
2437 context
.check_hostname
, context
.verify_mode
= True, ssl
.CERT_REQUIRED
2438 for storename
in ('CA', 'ROOT'):
2439 _ssl_load_windows_store_certs(context
, storename
)
2440 context
.set_default_verify_paths()
2441 return YoutubeDLHTTPSHandler(params
, context
=context
, **kwargs
)
2444 def bug_reports_message(before
=';'):
2445 if ytdl_is_updateable():
2446 update_cmd
= 'type yt-dlp -U to update'
2448 update_cmd
= 'see https://github.com/yt-dlp/yt-dlp on how to update'
2449 msg
= 'please report this issue on https://github.com/yt-dlp/yt-dlp .'
2450 msg
+= ' Make sure you are using the latest version; %s.' % update_cmd
2451 msg
+= ' Be sure to call yt-dlp with the --verbose flag and include its complete output.'
2453 before
= before
.rstrip()
2454 if not before
or before
.endswith(('.', '!', '?')):
2455 msg
= msg
[0].title() + msg
[1:]
2457 return (before
+ ' ' if before
else '') + msg
2460 class YoutubeDLError(Exception):
2461 """Base exception for YoutubeDL errors."""
2464 def __init__(self
, msg
=None):
2467 elif self
.msg
is None:
2468 self
.msg
= type(self
).__name
__
2469 super().__init
__(self
.msg
)
2472 network_exceptions
= [compat_urllib_error
.URLError
, compat_http_client
.HTTPException
, socket
.error
]
2473 if hasattr(ssl
, 'CertificateError'):
2474 network_exceptions
.append(ssl
.CertificateError
)
2475 network_exceptions
= tuple(network_exceptions
)
2478 class ExtractorError(YoutubeDLError
):
2479 """Error during info extraction."""
2481 def __init__(self
, msg
, tb
=None, expected
=False, cause
=None, video_id
=None, ie
=None):
2482 """ tb, if given, is the original traceback (so that it can be printed out).
2483 If expected is set, this is a normal error message and most likely not a bug in yt-dlp.
2485 if sys
.exc_info()[0] in network_exceptions
:
2490 self
.expected
= expected
2492 self
.video_id
= video_id
2494 self
.exc_info
= sys
.exc_info() # preserve original exception
2496 super(ExtractorError
, self
).__init
__(''.join((
2497 format_field(ie
, template
='[%s] '),
2498 format_field(video_id
, template
='%s: '),
2500 format_field(cause
, template
=' (caused by %r)'),
2501 '' if expected
else bug_reports_message())))
2503 def format_traceback(self
):
2504 if self
.traceback
is None:
2506 return ''.join(traceback
.format_tb(self
.traceback
))
2509 class UnsupportedError(ExtractorError
):
2510 def __init__(self
, url
):
2511 super(UnsupportedError
, self
).__init
__(
2512 'Unsupported URL: %s' % url
, expected
=True)
2516 class RegexNotFoundError(ExtractorError
):
2517 """Error when a regex didn't match"""
2521 class GeoRestrictedError(ExtractorError
):
2522 """Geographic restriction Error exception.
2524 This exception may be thrown when a video is not available from your
2525 geographic location due to geographic restrictions imposed by a website.
2528 def __init__(self
, msg
, countries
=None, **kwargs
):
2529 kwargs
['expected'] = True
2530 super(GeoRestrictedError
, self
).__init
__(msg
, **kwargs
)
2531 self
.countries
= countries
2534 class DownloadError(YoutubeDLError
):
2535 """Download Error exception.
2537 This exception may be thrown by FileDownloader objects if they are not
2538 configured to continue on errors. They will contain the appropriate
2542 def __init__(self
, msg
, exc_info
=None):
2543 """ exc_info, if given, is the original exception that caused the trouble (as returned by sys.exc_info()). """
2544 super(DownloadError
, self
).__init
__(msg
)
2545 self
.exc_info
= exc_info
2548 class EntryNotInPlaylist(YoutubeDLError
):
2549 """Entry not in playlist exception.
2551 This exception will be thrown by YoutubeDL when a requested entry
2552 is not found in the playlist info_dict
2554 msg
= 'Entry not found in info'
2557 class SameFileError(YoutubeDLError
):
2558 """Same File exception.
2560 This exception will be thrown by FileDownloader objects if they detect
2561 multiple files would have to be downloaded to the same file on disk.
2563 msg
= 'Fixed output name but more than one file to download'
2565 def __init__(self
, filename
=None):
2566 if filename
is not None:
2567 self
.msg
+= f
': {filename}'
2568 super().__init
__(self
.msg
)
2571 class PostProcessingError(YoutubeDLError
):
2572 """Post Processing exception.
2574 This exception may be raised by PostProcessor's .run() method to
2575 indicate an error in the postprocessing task.
2578 def __init__(self
, msg
):
2579 super(PostProcessingError
, self
).__init
__(msg
)
2583 class DownloadCancelled(YoutubeDLError
):
2584 """ Exception raised when the download queue should be interrupted """
2585 msg
= 'The download was cancelled'
2588 class ExistingVideoReached(DownloadCancelled
):
2589 """ --break-on-existing triggered """
2590 msg
= 'Encountered a video that is already in the archive, stopping due to --break-on-existing'
2593 class RejectedVideoReached(DownloadCancelled
):
2594 """ --break-on-reject triggered """
2595 msg
= 'Encountered a video that did not match filter, stopping due to --break-on-reject'
2598 class MaxDownloadsReached(DownloadCancelled
):
2599 """ --max-downloads limit has been reached. """
2600 msg
= 'Maximum number of downloads reached, stopping due to --max-downloads'
2603 class ReExtractInfo(YoutubeDLError
):
2604 """ Video info needs to be re-extracted. """
2606 def __init__(self
, msg
, expected
=False):
2607 super().__init
__(msg
)
2608 self
.expected
= expected
2611 class ThrottledDownload(ReExtractInfo
):
2612 """ Download speed below --throttled-rate. """
2613 msg
= 'The download speed is below throttle limit'
2615 def __init__(self
, msg
):
2616 super().__init
__(msg
, expected
=False)
2619 class UnavailableVideoError(YoutubeDLError
):
2620 """Unavailable Format exception.
2622 This exception will be thrown when a video is requested
2623 in a format that is not available for that video.
2625 msg
= 'Unable to download video'
2627 def __init__(self
, err
=None):
2629 self
.msg
+= f
': {err}'
2630 super().__init
__(self
.msg
)
2633 class ContentTooShortError(YoutubeDLError
):
2634 """Content Too Short exception.
2636 This exception may be raised by FileDownloader objects when a file they
2637 download is too small for what the server announced first, indicating
2638 the connection was probably interrupted.
2641 def __init__(self
, downloaded
, expected
):
2642 super(ContentTooShortError
, self
).__init
__(
2643 'Downloaded {0} bytes, expected {1} bytes'.format(downloaded
, expected
)
2646 self
.downloaded
= downloaded
2647 self
.expected
= expected
2650 class XAttrMetadataError(YoutubeDLError
):
2651 def __init__(self
, code
=None, msg
='Unknown error'):
2652 super(XAttrMetadataError
, self
).__init
__(msg
)
2656 # Parsing code and msg
2657 if (self
.code
in (errno
.ENOSPC
, errno
.EDQUOT
)
2658 or 'No space left' in self
.msg
or 'Disk quota exceeded' in self
.msg
):
2659 self
.reason
= 'NO_SPACE'
2660 elif self
.code
== errno
.E2BIG
or 'Argument list too long' in self
.msg
:
2661 self
.reason
= 'VALUE_TOO_LONG'
2663 self
.reason
= 'NOT_SUPPORTED'
2666 class XAttrUnavailableError(YoutubeDLError
):
2670 def _create_http_connection(ydl_handler
, http_class
, is_https
, *args
, **kwargs
):
2671 # Working around python 2 bug (see http://bugs.python.org/issue17849) by limiting
2672 # expected HTTP responses to meet HTTP/1.0 or later (see also
2673 # https://github.com/ytdl-org/youtube-dl/issues/6727)
2674 if sys
.version_info
< (3, 0):
2675 kwargs
['strict'] = True
2676 hc
= http_class(*args
, **compat_kwargs(kwargs
))
2677 source_address
= ydl_handler
._params
.get('source_address')
2679 if source_address
is not None:
2680 # This is to workaround _create_connection() from socket where it will try all
2681 # address data from getaddrinfo() including IPv6. This filters the result from
2682 # getaddrinfo() based on the source_address value.
2683 # This is based on the cpython socket.create_connection() function.
2684 # https://github.com/python/cpython/blob/master/Lib/socket.py#L691
2685 def _create_connection(address
, timeout
=socket
._GLOBAL
_DEFAULT
_TIMEOUT
, source_address
=None):
2686 host
, port
= address
2688 addrs
= socket
.getaddrinfo(host
, port
, 0, socket
.SOCK_STREAM
)
2689 af
= socket
.AF_INET
if '.' in source_address
[0] else socket
.AF_INET6
2690 ip_addrs
= [addr
for addr
in addrs
if addr
[0] == af
]
2691 if addrs
and not ip_addrs
:
2692 ip_version
= 'v4' if af
== socket
.AF_INET
else 'v6'
2694 "No remote IP%s addresses available for connect, can't use '%s' as source address"
2695 % (ip_version
, source_address
[0]))
2696 for res
in ip_addrs
:
2697 af
, socktype
, proto
, canonname
, sa
= res
2700 sock
= socket
.socket(af
, socktype
, proto
)
2701 if timeout
is not socket
._GLOBAL
_DEFAULT
_TIMEOUT
:
2702 sock
.settimeout(timeout
)
2703 sock
.bind(source_address
)
2705 err
= None # Explicitly break reference cycle
2707 except socket
.error
as _
:
2709 if sock
is not None:
2714 raise socket
.error('getaddrinfo returns an empty list')
2715 if hasattr(hc
, '_create_connection'):
2716 hc
._create
_connection
= _create_connection
2717 sa
= (source_address
, 0)
2718 if hasattr(hc
, 'source_address'): # Python 2.7+
2719 hc
.source_address
= sa
2721 def _hc_connect(self
, *args
, **kwargs
):
2722 sock
= _create_connection(
2723 (self
.host
, self
.port
), self
.timeout
, sa
)
2725 self
.sock
= ssl
.wrap_socket(
2726 sock
, self
.key_file
, self
.cert_file
,
2727 ssl_version
=ssl
.PROTOCOL_TLSv1
)
2730 hc
.connect
= functools
.partial(_hc_connect
, hc
)
2735 def handle_youtubedl_headers(headers
):
2736 filtered_headers
= headers
2738 if 'Youtubedl-no-compression' in filtered_headers
:
2739 filtered_headers
= dict((k
, v
) for k
, v
in filtered_headers
.items() if k
.lower() != 'accept-encoding')
2740 del filtered_headers
['Youtubedl-no-compression']
2742 return filtered_headers
2745 class YoutubeDLHandler(compat_urllib_request
.HTTPHandler
):
2746 """Handler for HTTP requests and responses.
2748 This class, when installed with an OpenerDirector, automatically adds
2749 the standard headers to every HTTP request and handles gzipped and
2750 deflated responses from web servers. If compression is to be avoided in
2751 a particular request, the original request in the program code only has
2752 to include the HTTP header "Youtubedl-no-compression", which will be
2753 removed before making the real request.
2755 Part of this code was copied from:
2757 http://techknack.net/python-urllib2-handlers/
2759 Andrew Rowls, the author of that code, agreed to release it to the
2763 def __init__(self
, params
, *args
, **kwargs
):
2764 compat_urllib_request
.HTTPHandler
.__init
__(self
, *args
, **kwargs
)
2765 self
._params
= params
2767 def http_open(self
, req
):
2768 conn_class
= compat_http_client
.HTTPConnection
2770 socks_proxy
= req
.headers
.get('Ytdl-socks-proxy')
2772 conn_class
= make_socks_conn_class(conn_class
, socks_proxy
)
2773 del req
.headers
['Ytdl-socks-proxy']
2775 return self
.do_open(functools
.partial(
2776 _create_http_connection
, self
, conn_class
, False),
2784 return zlib
.decompress(data
, -zlib
.MAX_WBITS
)
2786 return zlib
.decompress(data
)
2788 def http_request(self
, req
):
2789 # According to RFC 3986, URLs can not contain non-ASCII characters, however this is not
2790 # always respected by websites, some tend to give out URLs with non percent-encoded
2791 # non-ASCII characters (see telemb.py, ard.py [#3412])
2792 # urllib chokes on URLs with non-ASCII characters (see http://bugs.python.org/issue3991)
2793 # To work around aforementioned issue we will replace request's original URL with
2794 # percent-encoded one
2795 # Since redirects are also affected (e.g. http://www.southpark.de/alle-episoden/s18e09)
2796 # the code of this workaround has been moved here from YoutubeDL.urlopen()
2797 url
= req
.get_full_url()
2798 url_escaped
= escape_url(url
)
2800 # Substitute URL if any change after escaping
2801 if url
!= url_escaped
:
2802 req
= update_Request(req
, url
=url_escaped
)
2804 for h
, v
in std_headers
.items():
2805 # Capitalize is needed because of Python bug 2275: http://bugs.python.org/issue2275
2806 # The dict keys are capitalized because of this bug by urllib
2807 if h
.capitalize() not in req
.headers
:
2808 req
.add_header(h
, v
)
2810 req
.headers
= handle_youtubedl_headers(req
.headers
)
2812 if sys
.version_info
< (2, 7) and '#' in req
.get_full_url():
2813 # Python 2.6 is brain-dead when it comes to fragments
2814 req
._Request
__original
= req
._Request
__original
.partition('#')[0]
2815 req
._Request
__r
_type
= req
._Request
__r
_type
.partition('#')[0]
2819 def http_response(self
, req
, resp
):
2822 if resp
.headers
.get('Content-encoding', '') == 'gzip':
2823 content
= resp
.read()
2824 gz
= gzip
.GzipFile(fileobj
=io
.BytesIO(content
), mode
='rb')
2826 uncompressed
= io
.BytesIO(gz
.read())
2827 except IOError as original_ioerror
:
2828 # There may be junk add the end of the file
2829 # See http://stackoverflow.com/q/4928560/35070 for details
2830 for i
in range(1, 1024):
2832 gz
= gzip
.GzipFile(fileobj
=io
.BytesIO(content
[:-i
]), mode
='rb')
2833 uncompressed
= io
.BytesIO(gz
.read())
2838 raise original_ioerror
2839 resp
= compat_urllib_request
.addinfourl(uncompressed
, old_resp
.headers
, old_resp
.url
, old_resp
.code
)
2840 resp
.msg
= old_resp
.msg
2841 del resp
.headers
['Content-encoding']
2843 if resp
.headers
.get('Content-encoding', '') == 'deflate':
2844 gz
= io
.BytesIO(self
.deflate(resp
.read()))
2845 resp
= compat_urllib_request
.addinfourl(gz
, old_resp
.headers
, old_resp
.url
, old_resp
.code
)
2846 resp
.msg
= old_resp
.msg
2847 del resp
.headers
['Content-encoding']
2848 # Percent-encode redirect URL of Location HTTP header to satisfy RFC 3986 (see
2849 # https://github.com/ytdl-org/youtube-dl/issues/6457).
2850 if 300 <= resp
.code
< 400:
2851 location
= resp
.headers
.get('Location')
2853 # As of RFC 2616 default charset is iso-8859-1 that is respected by python 3
2854 if sys
.version_info
>= (3, 0):
2855 location
= location
.encode('iso-8859-1').decode('utf-8')
2857 location
= location
.decode('utf-8')
2858 location_escaped
= escape_url(location
)
2859 if location
!= location_escaped
:
2860 del resp
.headers
['Location']
2861 if sys
.version_info
< (3, 0):
2862 location_escaped
= location_escaped
.encode('utf-8')
2863 resp
.headers
['Location'] = location_escaped
2866 https_request
= http_request
2867 https_response
= http_response
2870 def make_socks_conn_class(base_class
, socks_proxy
):
2871 assert issubclass(base_class
, (
2872 compat_http_client
.HTTPConnection
, compat_http_client
.HTTPSConnection
))
2874 url_components
= compat_urlparse
.urlparse(socks_proxy
)
2875 if url_components
.scheme
.lower() == 'socks5':
2876 socks_type
= ProxyType
.SOCKS5
2877 elif url_components
.scheme
.lower() in ('socks', 'socks4'):
2878 socks_type
= ProxyType
.SOCKS4
2879 elif url_components
.scheme
.lower() == 'socks4a':
2880 socks_type
= ProxyType
.SOCKS4A
2882 def unquote_if_non_empty(s
):
2885 return compat_urllib_parse_unquote_plus(s
)
2889 url_components
.hostname
, url_components
.port
or 1080,
2891 unquote_if_non_empty(url_components
.username
),
2892 unquote_if_non_empty(url_components
.password
),
2895 class SocksConnection(base_class
):
2897 self
.sock
= sockssocket()
2898 self
.sock
.setproxy(*proxy_args
)
2899 if type(self
.timeout
) in (int, float):
2900 self
.sock
.settimeout(self
.timeout
)
2901 self
.sock
.connect((self
.host
, self
.port
))
2903 if isinstance(self
, compat_http_client
.HTTPSConnection
):
2904 if hasattr(self
, '_context'): # Python > 2.6
2905 self
.sock
= self
._context
.wrap_socket(
2906 self
.sock
, server_hostname
=self
.host
)
2908 self
.sock
= ssl
.wrap_socket(self
.sock
)
2910 return SocksConnection
2913 class YoutubeDLHTTPSHandler(compat_urllib_request
.HTTPSHandler
):
2914 def __init__(self
, params
, https_conn_class
=None, *args
, **kwargs
):
2915 compat_urllib_request
.HTTPSHandler
.__init
__(self
, *args
, **kwargs
)
2916 self
._https
_conn
_class
= https_conn_class
or compat_http_client
.HTTPSConnection
2917 self
._params
= params
2919 def https_open(self
, req
):
2921 conn_class
= self
._https
_conn
_class
2923 if hasattr(self
, '_context'): # python > 2.6
2924 kwargs
['context'] = self
._context
2925 if hasattr(self
, '_check_hostname'): # python 3.x
2926 kwargs
['check_hostname'] = self
._check
_hostname
2928 socks_proxy
= req
.headers
.get('Ytdl-socks-proxy')
2930 conn_class
= make_socks_conn_class(conn_class
, socks_proxy
)
2931 del req
.headers
['Ytdl-socks-proxy']
2933 return self
.do_open(functools
.partial(
2934 _create_http_connection
, self
, conn_class
, True),
2938 class YoutubeDLCookieJar(compat_cookiejar
.MozillaCookieJar
):
2940 See [1] for cookie file format.
2942 1. https://curl.haxx.se/docs/http-cookies.html
2944 _HTTPONLY_PREFIX
= '#HttpOnly_'
2946 _HEADER
= '''# Netscape HTTP Cookie File
2947 # This file is generated by yt-dlp. Do not edit.
2950 _CookieFileEntry
= collections
.namedtuple(
2952 ('domain_name', 'include_subdomains', 'path', 'https_only', 'expires_at', 'name', 'value'))
2954 def save(self
, filename
=None, ignore_discard
=False, ignore_expires
=False):
2956 Save cookies to a file.
2958 Most of the code is taken from CPython 3.8 and slightly adapted
2959 to support cookie files with UTF-8 in both python 2 and 3.
2961 if filename
is None:
2962 if self
.filename
is not None:
2963 filename
= self
.filename
2965 raise ValueError(compat_cookiejar
.MISSING_FILENAME_TEXT
)
2967 # Store session cookies with `expires` set to 0 instead of an empty
2970 if cookie
.expires
is None:
2973 with io
.open(filename
, 'w', encoding
='utf-8') as f
:
2974 f
.write(self
._HEADER
)
2977 if not ignore_discard
and cookie
.discard
:
2979 if not ignore_expires
and cookie
.is_expired(now
):
2985 if cookie
.domain
.startswith('.'):
2986 initial_dot
= 'TRUE'
2988 initial_dot
= 'FALSE'
2989 if cookie
.expires
is not None:
2990 expires
= compat_str(cookie
.expires
)
2993 if cookie
.value
is None:
2994 # cookies.txt regards 'Set-Cookie: foo' as a cookie
2995 # with no name, whereas http.cookiejar regards it as a
2996 # cookie with no value.
3001 value
= cookie
.value
3003 '\t'.join([cookie
.domain
, initial_dot
, cookie
.path
,
3004 secure
, expires
, name
, value
]) + '\n')
3006 def load(self
, filename
=None, ignore_discard
=False, ignore_expires
=False):
3007 """Load cookies from a file."""
3008 if filename
is None:
3009 if self
.filename
is not None:
3010 filename
= self
.filename
3012 raise ValueError(compat_cookiejar
.MISSING_FILENAME_TEXT
)
3014 def prepare_line(line
):
3015 if line
.startswith(self
._HTTPONLY
_PREFIX
):
3016 line
= line
[len(self
._HTTPONLY
_PREFIX
):]
3017 # comments and empty lines are fine
3018 if line
.startswith('#') or not line
.strip():
3020 cookie_list
= line
.split('\t')
3021 if len(cookie_list
) != self
._ENTRY
_LEN
:
3022 raise compat_cookiejar
.LoadError('invalid length %d' % len(cookie_list
))
3023 cookie
= self
._CookieFileEntry
(*cookie_list
)
3024 if cookie
.expires_at
and not cookie
.expires_at
.isdigit():
3025 raise compat_cookiejar
.LoadError('invalid expires at %s' % cookie
.expires_at
)
3029 with io
.open(filename
, encoding
='utf-8') as f
:
3032 cf
.write(prepare_line(line
))
3033 except compat_cookiejar
.LoadError
as e
:
3035 'WARNING: skipping cookie file entry due to %s: %r\n'
3036 % (e
, line
), sys
.stderr
)
3039 self
._really
_load
(cf
, filename
, ignore_discard
, ignore_expires
)
3040 # Session cookies are denoted by either `expires` field set to
3041 # an empty string or 0. MozillaCookieJar only recognizes the former
3042 # (see [1]). So we need force the latter to be recognized as session
3043 # cookies on our own.
3044 # Session cookies may be important for cookies-based authentication,
3045 # e.g. usually, when user does not check 'Remember me' check box while
3046 # logging in on a site, some important cookies are stored as session
3047 # cookies so that not recognizing them will result in failed login.
3048 # 1. https://bugs.python.org/issue17164
3050 # Treat `expires=0` cookies as session cookies
3051 if cookie
.expires
== 0:
3052 cookie
.expires
= None
3053 cookie
.discard
= True
3056 class YoutubeDLCookieProcessor(compat_urllib_request
.HTTPCookieProcessor
):
3057 def __init__(self
, cookiejar
=None):
3058 compat_urllib_request
.HTTPCookieProcessor
.__init
__(self
, cookiejar
)
3060 def http_response(self
, request
, response
):
3061 # Python 2 will choke on next HTTP request in row if there are non-ASCII
3062 # characters in Set-Cookie HTTP header of last response (see
3063 # https://github.com/ytdl-org/youtube-dl/issues/6769).
3064 # In order to at least prevent crashing we will percent encode Set-Cookie
3065 # header before HTTPCookieProcessor starts processing it.
3066 # if sys.version_info < (3, 0) and response.headers:
3067 # for set_cookie_header in ('Set-Cookie', 'Set-Cookie2'):
3068 # set_cookie = response.headers.get(set_cookie_header)
3070 # set_cookie_escaped = compat_urllib_parse.quote(set_cookie, b"%/;:@&=+$,!~*'()?#[] ")
3071 # if set_cookie != set_cookie_escaped:
3072 # del response.headers[set_cookie_header]
3073 # response.headers[set_cookie_header] = set_cookie_escaped
3074 return compat_urllib_request
.HTTPCookieProcessor
.http_response(self
, request
, response
)
3076 https_request
= compat_urllib_request
.HTTPCookieProcessor
.http_request
3077 https_response
= http_response
3080 class YoutubeDLRedirectHandler(compat_urllib_request
.HTTPRedirectHandler
):
3081 """YoutubeDL redirect handler
3083 The code is based on HTTPRedirectHandler implementation from CPython [1].
3085 This redirect handler solves two issues:
3086 - ensures redirect URL is always unicode under python 2
3087 - introduces support for experimental HTTP response status code
3088 308 Permanent Redirect [2] used by some sites [3]
3090 1. https://github.com/python/cpython/blob/master/Lib/urllib/request.py
3091 2. https://developer.mozilla.org/en-US/docs/Web/HTTP/Status/308
3092 3. https://github.com/ytdl-org/youtube-dl/issues/28768
3095 http_error_301
= http_error_303
= http_error_307
= http_error_308
= compat_urllib_request
.HTTPRedirectHandler
.http_error_302
3097 def redirect_request(self
, req
, fp
, code
, msg
, headers
, newurl
):
3098 """Return a Request or None in response to a redirect.
3100 This is called by the http_error_30x methods when a
3101 redirection response is received. If a redirection should
3102 take place, return a new Request to allow http_error_30x to
3103 perform the redirect. Otherwise, raise HTTPError if no-one
3104 else should try to handle this url. Return None if you can't
3105 but another Handler might.
3107 m
= req
.get_method()
3108 if (not (code
in (301, 302, 303, 307, 308) and m
in ("GET", "HEAD")
3109 or code
in (301, 302, 303) and m
== "POST")):
3110 raise compat_HTTPError(req
.full_url
, code
, msg
, headers
, fp
)
3111 # Strictly (according to RFC 2616), 301 or 302 in response to
3112 # a POST MUST NOT cause a redirection without confirmation
3113 # from the user (of urllib.request, in this case). In practice,
3114 # essentially all clients do redirect in this case, so we do
3117 # On python 2 urlh.geturl() may sometimes return redirect URL
3118 # as byte string instead of unicode. This workaround allows
3119 # to force it always return unicode.
3120 if sys
.version_info
[0] < 3:
3121 newurl
= compat_str(newurl
)
3123 # Be conciliant with URIs containing a space. This is mainly
3124 # redundant with the more complete encoding done in http_error_302(),
3125 # but it is kept for compatibility with other callers.
3126 newurl
= newurl
.replace(' ', '%20')
3128 CONTENT_HEADERS
= ("content-length", "content-type")
3129 # NB: don't use dict comprehension for python 2.6 compatibility
3130 newheaders
= dict((k
, v
) for k
, v
in req
.headers
.items()
3131 if k
.lower() not in CONTENT_HEADERS
)
3132 return compat_urllib_request
.Request(
3133 newurl
, headers
=newheaders
, origin_req_host
=req
.origin_req_host
,
3137 def extract_timezone(date_str
):
3140 ^.{8,}? # >=8 char non-TZ prefix, if present
3141 (?P<tz>Z| # just the UTC Z, or
3142 (?:(?<=.\b\d{4}|\b\d{2}:\d\d)| # preceded by 4 digits or hh:mm or
3143 (?<!.\b[a-zA-Z]{3}|[a-zA-Z]{4}|..\b\d\d)) # not preceded by 3 alpha word or >= 4 alpha or 2 digits
3144 [ ]? # optional space
3145 (?P<sign>\+|-) # +/-
3146 (?P<hours>[0-9]{2}):?(?P<minutes>[0-9]{2}) # hh[:]mm
3150 timezone
= datetime
.timedelta()
3152 date_str
= date_str
[:-len(m
.group('tz'))]
3153 if not m
.group('sign'):
3154 timezone
= datetime
.timedelta()
3156 sign
= 1 if m
.group('sign') == '+' else -1
3157 timezone
= datetime
.timedelta(
3158 hours
=sign
* int(m
.group('hours')),
3159 minutes
=sign
* int(m
.group('minutes')))
3160 return timezone
, date_str
3163 def parse_iso8601(date_str
, delimiter
='T', timezone
=None):
3164 """ Return a UNIX timestamp from the given date """
3166 if date_str
is None:
3169 date_str
= re
.sub(r
'\.[0-9]+', '', date_str
)
3171 if timezone
is None:
3172 timezone
, date_str
= extract_timezone(date_str
)
3175 date_format
= '%Y-%m-%d{0}%H:%M:%S'.format(delimiter
)
3176 dt
= datetime
.datetime
.strptime(date_str
, date_format
) - timezone
3177 return calendar
.timegm(dt
.timetuple())
3182 def date_formats(day_first
=True):
3183 return DATE_FORMATS_DAY_FIRST
if day_first
else DATE_FORMATS_MONTH_FIRST
3186 def unified_strdate(date_str
, day_first
=True):
3187 """Return a string with the date in the format YYYYMMDD"""
3189 if date_str
is None:
3193 date_str
= date_str
.replace(',', ' ')
3194 # Remove AM/PM + timezone
3195 date_str
= re
.sub(r
'(?i)\s*(?:AM|PM)(?:\s+[A-Z]+)?', '', date_str
)
3196 _
, date_str
= extract_timezone(date_str
)
3198 for expression
in date_formats(day_first
):
3200 upload_date
= datetime
.datetime
.strptime(date_str
, expression
).strftime('%Y%m%d')
3203 if upload_date
is None:
3204 timetuple
= email
.utils
.parsedate_tz(date_str
)
3207 upload_date
= datetime
.datetime(*timetuple
[:6]).strftime('%Y%m%d')
3210 if upload_date
is not None:
3211 return compat_str(upload_date
)
3214 def unified_timestamp(date_str
, day_first
=True):
3215 if date_str
is None:
3218 date_str
= re
.sub(r
'[,|]', '', date_str
)
3220 pm_delta
= 12 if re
.search(r
'(?i)PM', date_str
) else 0
3221 timezone
, date_str
= extract_timezone(date_str
)
3223 # Remove AM/PM + timezone
3224 date_str
= re
.sub(r
'(?i)\s*(?:AM|PM)(?:\s+[A-Z]+)?', '', date_str
)
3226 # Remove unrecognized timezones from ISO 8601 alike timestamps
3227 m
= re
.search(r
'\d{1,2}:\d{1,2}(?:\.\d+)?(?P<tz>\s*[A-Z]+)$', date_str
)
3229 date_str
= date_str
[:-len(m
.group('tz'))]
3231 # Python only supports microseconds, so remove nanoseconds
3232 m
= re
.search(r
'^([0-9]{4,}-[0-9]{1,2}-[0-9]{1,2}T[0-9]{1,2}:[0-9]{1,2}:[0-9]{1,2}\.[0-9]{6})[0-9]+$', date_str
)
3234 date_str
= m
.group(1)
3236 for expression
in date_formats(day_first
):
3238 dt
= datetime
.datetime
.strptime(date_str
, expression
) - timezone
+ datetime
.timedelta(hours
=pm_delta
)
3239 return calendar
.timegm(dt
.timetuple())
3242 timetuple
= email
.utils
.parsedate_tz(date_str
)
3244 return calendar
.timegm(timetuple
) + pm_delta
* 3600
3247 def determine_ext(url
, default_ext
='unknown_video'):
3248 if url
is None or '.' not in url
:
3250 guess
= url
.partition('?')[0].rpartition('.')[2]
3251 if re
.match(r
'^[A-Za-z0-9]+$', guess
):
3253 # Try extract ext from URLs like http://example.com/foo/bar.mp4/?download
3254 elif guess
.rstrip('/') in KNOWN_EXTENSIONS
:
3255 return guess
.rstrip('/')
3260 def subtitles_filename(filename
, sub_lang
, sub_format
, expected_real_ext
=None):
3261 return replace_extension(filename
, sub_lang
+ '.' + sub_format
, expected_real_ext
)
3264 def datetime_from_str(date_str
, precision
='auto', format
='%Y%m%d'):
3266 Return a datetime object from a string in the format YYYYMMDD or
3267 (now|today|date)[+-][0-9](microsecond|second|minute|hour|day|week|month|year)(s)?
3269 format: string date format used to return datetime object from
3270 precision: round the time portion of a datetime object.
3271 auto|microsecond|second|minute|hour|day.
3272 auto: round to the unit provided in date_str (if applicable).
3274 auto_precision
= False
3275 if precision
== 'auto':
3276 auto_precision
= True
3277 precision
= 'microsecond'
3278 today
= datetime_round(datetime
.datetime
.now(), precision
)
3279 if date_str
in ('now', 'today'):
3281 if date_str
== 'yesterday':
3282 return today
- datetime
.timedelta(days
=1)
3284 r
'(?P<start>.+)(?P<sign>[+-])(?P<time>\d+)(?P<unit>microsecond|second|minute|hour|day|week|month|year)(s)?',
3286 if match
is not None:
3287 start_time
= datetime_from_str(match
.group('start'), precision
, format
)
3288 time
= int(match
.group('time')) * (-1 if match
.group('sign') == '-' else 1)
3289 unit
= match
.group('unit')
3290 if unit
== 'month' or unit
== 'year':
3291 new_date
= datetime_add_months(start_time
, time
* 12 if unit
== 'year' else time
)
3297 delta
= datetime
.timedelta(**{unit + 's': time}
)
3298 new_date
= start_time
+ delta
3300 return datetime_round(new_date
, unit
)
3303 return datetime_round(datetime
.datetime
.strptime(date_str
, format
), precision
)
3306 def date_from_str(date_str
, format
='%Y%m%d'):
3308 Return a datetime object from a string in the format YYYYMMDD or
3309 (now|today|date)[+-][0-9](microsecond|second|minute|hour|day|week|month|year)(s)?
3311 format: string date format used to return datetime object from
3313 return datetime_from_str(date_str
, precision
='microsecond', format
=format
).date()
3316 def datetime_add_months(dt
, months
):
3317 """Increment/Decrement a datetime object by months."""
3318 month
= dt
.month
+ months
- 1
3319 year
= dt
.year
+ month
// 12
3320 month
= month
% 12 + 1
3321 day
= min(dt
.day
, calendar
.monthrange(year
, month
)[1])
3322 return dt
.replace(year
, month
, day
)
3325 def datetime_round(dt
, precision
='day'):
3327 Round a datetime object's time to a specific precision
3329 if precision
== 'microsecond':
3338 roundto
= lambda x
, n
: ((x
+ n
/ 2) // n
) * n
3339 timestamp
= calendar
.timegm(dt
.timetuple())
3340 return datetime
.datetime
.utcfromtimestamp(roundto(timestamp
, unit_seconds
[precision
]))
3343 def hyphenate_date(date_str
):
3345 Convert a date in 'YYYYMMDD' format to 'YYYY-MM-DD' format"""
3346 match
= re
.match(r
'^(\d\d\d\d)(\d\d)(\d\d)$', date_str
)
3347 if match
is not None:
3348 return '-'.join(match
.groups())
3353 class DateRange(object):
3354 """Represents a time interval between two dates"""
3356 def __init__(self
, start
=None, end
=None):
3357 """start and end must be strings in the format accepted by date"""
3358 if start
is not None:
3359 self
.start
= date_from_str(start
)
3361 self
.start
= datetime
.datetime
.min.date()
3363 self
.end
= date_from_str(end
)
3365 self
.end
= datetime
.datetime
.max.date()
3366 if self
.start
> self
.end
:
3367 raise ValueError('Date range: "%s" , the start date must be before the end date' % self
)
3371 """Returns a range that only contains the given day"""
3372 return cls(day
, day
)
3374 def __contains__(self
, date
):
3375 """Check if the date is in the range"""
3376 if not isinstance(date
, datetime
.date
):
3377 date
= date_from_str(date
)
3378 return self
.start
<= date
<= self
.end
3381 return '%s - %s' % (self
.start
.isoformat(), self
.end
.isoformat())
3384 def platform_name():
3385 """ Returns the platform name as a compat_str """
3386 res
= platform
.platform()
3387 if isinstance(res
, bytes):
3388 res
= res
.decode(preferredencoding())
3390 assert isinstance(res
, compat_str
)
3394 def get_windows_version():
3395 ''' Get Windows version. None if it's not running on Windows '''
3396 if compat_os_name
== 'nt':
3397 return version_tuple(platform
.win32_ver()[1])
3402 def _windows_write_string(s
, out
):
3403 """ Returns True if the string was written using special methods,
3404 False if it has yet to be written out."""
3405 # Adapted from http://stackoverflow.com/a/3259271/35070
3408 import ctypes
.wintypes
3416 fileno
= out
.fileno()
3417 except AttributeError:
3418 # If the output stream doesn't have a fileno, it's virtual
3420 except io
.UnsupportedOperation
:
3421 # Some strange Windows pseudo files?
3423 if fileno
not in WIN_OUTPUT_IDS
:
3426 GetStdHandle
= compat_ctypes_WINFUNCTYPE(
3427 ctypes
.wintypes
.HANDLE
, ctypes
.wintypes
.DWORD
)(
3428 ('GetStdHandle', ctypes
.windll
.kernel32
))
3429 h
= GetStdHandle(WIN_OUTPUT_IDS
[fileno
])
3431 WriteConsoleW
= compat_ctypes_WINFUNCTYPE(
3432 ctypes
.wintypes
.BOOL
, ctypes
.wintypes
.HANDLE
, ctypes
.wintypes
.LPWSTR
,
3433 ctypes
.wintypes
.DWORD
, ctypes
.POINTER(ctypes
.wintypes
.DWORD
),
3434 ctypes
.wintypes
.LPVOID
)(('WriteConsoleW', ctypes
.windll
.kernel32
))
3435 written
= ctypes
.wintypes
.DWORD(0)
3437 GetFileType
= compat_ctypes_WINFUNCTYPE(ctypes
.wintypes
.DWORD
, ctypes
.wintypes
.DWORD
)(('GetFileType', ctypes
.windll
.kernel32
))
3438 FILE_TYPE_CHAR
= 0x0002
3439 FILE_TYPE_REMOTE
= 0x8000
3440 GetConsoleMode
= compat_ctypes_WINFUNCTYPE(
3441 ctypes
.wintypes
.BOOL
, ctypes
.wintypes
.HANDLE
,
3442 ctypes
.POINTER(ctypes
.wintypes
.DWORD
))(
3443 ('GetConsoleMode', ctypes
.windll
.kernel32
))
3444 INVALID_HANDLE_VALUE
= ctypes
.wintypes
.DWORD(-1).value
3446 def not_a_console(handle
):
3447 if handle
== INVALID_HANDLE_VALUE
or handle
is None:
3449 return ((GetFileType(handle
) & ~FILE_TYPE_REMOTE
) != FILE_TYPE_CHAR
3450 or GetConsoleMode(handle
, ctypes
.byref(ctypes
.wintypes
.DWORD())) == 0)
3452 if not_a_console(h
):
3455 def next_nonbmp_pos(s
):
3457 return next(i
for i
, c
in enumerate(s
) if ord(c
) > 0xffff)
3458 except StopIteration:
3462 count
= min(next_nonbmp_pos(s
), 1024)
3464 ret
= WriteConsoleW(
3465 h
, s
, count
if count
else 2, ctypes
.byref(written
), None)
3467 raise OSError('Failed to write string')
3468 if not count
: # We just wrote a non-BMP character
3469 assert written
.value
== 2
3472 assert written
.value
> 0
3473 s
= s
[written
.value
:]
3477 def write_string(s
, out
=None, encoding
=None):
3480 assert type(s
) == compat_str
3482 if sys
.platform
== 'win32' and encoding
is None and hasattr(out
, 'fileno'):
3483 if _windows_write_string(s
, out
):
3486 if ('b' in getattr(out
, 'mode', '')
3487 or sys
.version_info
[0] < 3): # Python 2 lies about mode of sys.stderr
3488 byt
= s
.encode(encoding
or preferredencoding(), 'ignore')
3490 elif hasattr(out
, 'buffer'):
3491 enc
= encoding
or getattr(out
, 'encoding', None) or preferredencoding()
3492 byt
= s
.encode(enc
, 'ignore')
3493 out
.buffer.write(byt
)
3499 def bytes_to_intlist(bs
):
3502 if isinstance(bs
[0], int): # Python 3
3505 return [ord(c
) for c
in bs
]
3508 def intlist_to_bytes(xs
):
3511 return compat_struct_pack('%dB' % len(xs
), *xs
)
3514 # Cross-platform file locking
3515 if sys
.platform
== 'win32':
3516 import ctypes
.wintypes
3519 class OVERLAPPED(ctypes
.Structure
):
3521 ('Internal', ctypes
.wintypes
.LPVOID
),
3522 ('InternalHigh', ctypes
.wintypes
.LPVOID
),
3523 ('Offset', ctypes
.wintypes
.DWORD
),
3524 ('OffsetHigh', ctypes
.wintypes
.DWORD
),
3525 ('hEvent', ctypes
.wintypes
.HANDLE
),
3528 kernel32
= ctypes
.windll
.kernel32
3529 LockFileEx
= kernel32
.LockFileEx
3530 LockFileEx
.argtypes
= [
3531 ctypes
.wintypes
.HANDLE
, # hFile
3532 ctypes
.wintypes
.DWORD
, # dwFlags
3533 ctypes
.wintypes
.DWORD
, # dwReserved
3534 ctypes
.wintypes
.DWORD
, # nNumberOfBytesToLockLow
3535 ctypes
.wintypes
.DWORD
, # nNumberOfBytesToLockHigh
3536 ctypes
.POINTER(OVERLAPPED
) # Overlapped
3538 LockFileEx
.restype
= ctypes
.wintypes
.BOOL
3539 UnlockFileEx
= kernel32
.UnlockFileEx
3540 UnlockFileEx
.argtypes
= [
3541 ctypes
.wintypes
.HANDLE
, # hFile
3542 ctypes
.wintypes
.DWORD
, # dwReserved
3543 ctypes
.wintypes
.DWORD
, # nNumberOfBytesToLockLow
3544 ctypes
.wintypes
.DWORD
, # nNumberOfBytesToLockHigh
3545 ctypes
.POINTER(OVERLAPPED
) # Overlapped
3547 UnlockFileEx
.restype
= ctypes
.wintypes
.BOOL
3548 whole_low
= 0xffffffff
3549 whole_high
= 0x7fffffff
3551 def _lock_file(f
, exclusive
):
3552 overlapped
= OVERLAPPED()
3553 overlapped
.Offset
= 0
3554 overlapped
.OffsetHigh
= 0
3555 overlapped
.hEvent
= 0
3556 f
._lock
_file
_overlapped
_p
= ctypes
.pointer(overlapped
)
3557 handle
= msvcrt
.get_osfhandle(f
.fileno())
3558 if not LockFileEx(handle
, 0x2 if exclusive
else 0x0, 0,
3559 whole_low
, whole_high
, f
._lock
_file
_overlapped
_p
):
3560 raise OSError('Locking file failed: %r' % ctypes
.FormatError())
3562 def _unlock_file(f
):
3563 assert f
._lock
_file
_overlapped
_p
3564 handle
= msvcrt
.get_osfhandle(f
.fileno())
3565 if not UnlockFileEx(handle
, 0,
3566 whole_low
, whole_high
, f
._lock
_file
_overlapped
_p
):
3567 raise OSError('Unlocking file failed: %r' % ctypes
.FormatError())
3570 # Some platforms, such as Jython, is missing fcntl
3574 def _lock_file(f
, exclusive
):
3575 fcntl
.flock(f
, fcntl
.LOCK_EX
if exclusive
else fcntl
.LOCK_SH
)
3577 def _unlock_file(f
):
3578 fcntl
.flock(f
, fcntl
.LOCK_UN
)
3580 UNSUPPORTED_MSG
= 'file locking is not supported on this platform'
3582 def _lock_file(f
, exclusive
):
3583 raise IOError(UNSUPPORTED_MSG
)
3585 def _unlock_file(f
):
3586 raise IOError(UNSUPPORTED_MSG
)
3589 class locked_file(object):
3590 def __init__(self
, filename
, mode
, encoding
=None):
3591 assert mode
in ['r', 'a', 'w']
3592 self
.f
= io
.open(filename
, mode
, encoding
=encoding
)
3595 def __enter__(self
):
3596 exclusive
= self
.mode
!= 'r'
3598 _lock_file(self
.f
, exclusive
)
3604 def __exit__(self
, etype
, value
, traceback
):
3606 _unlock_file(self
.f
)
3613 def write(self
, *args
):
3614 return self
.f
.write(*args
)
3616 def read(self
, *args
):
3617 return self
.f
.read(*args
)
3620 def get_filesystem_encoding():
3621 encoding
= sys
.getfilesystemencoding()
3622 return encoding
if encoding
is not None else 'utf-8'
3625 def shell_quote(args
):
3627 encoding
= get_filesystem_encoding()
3629 if isinstance(a
, bytes):
3630 # We may get a filename encoded with 'encodeFilename'
3631 a
= a
.decode(encoding
)
3632 quoted_args
.append(compat_shlex_quote(a
))
3633 return ' '.join(quoted_args
)
3636 def smuggle_url(url
, data
):
3637 """ Pass additional data in a URL for internal use. """
3639 url
, idata
= unsmuggle_url(url
, {})
3641 sdata
= compat_urllib_parse_urlencode(
3642 {'__youtubedl_smuggle': json.dumps(data)}
)
3643 return url
+ '#' + sdata
3646 def unsmuggle_url(smug_url
, default
=None):
3647 if '#__youtubedl_smuggle' not in smug_url
:
3648 return smug_url
, default
3649 url
, _
, sdata
= smug_url
.rpartition('#')
3650 jsond
= compat_parse_qs(sdata
)['__youtubedl_smuggle'][0]
3651 data
= json
.loads(jsond
)
3655 def format_bytes(bytes):
3658 if type(bytes) is str:
3659 bytes = float(bytes)
3663 exponent
= int(math
.log(bytes, 1024.0))
3664 suffix
= ['B', 'KiB', 'MiB', 'GiB', 'TiB', 'PiB', 'EiB', 'ZiB', 'YiB'][exponent
]
3665 converted
= float(bytes) / float(1024 ** exponent
)
3666 return '%.2f%s' % (converted
, suffix
)
3669 def lookup_unit_table(unit_table
, s
):
3670 units_re
= '|'.join(re
.escape(u
) for u
in unit_table
)
3672 r
'(?P<num>[0-9]+(?:[,.][0-9]*)?)\s*(?P<unit>%s)\b' % units_re
, s
)
3675 num_str
= m
.group('num').replace(',', '.')
3676 mult
= unit_table
[m
.group('unit')]
3677 return int(float(num_str
) * mult
)
3680 def parse_filesize(s
):
3684 # The lower-case forms are of course incorrect and unofficial,
3685 # but we support those too
3702 'megabytes': 1000 ** 2,
3703 'mebibytes': 1024 ** 2,
3709 'gigabytes': 1000 ** 3,
3710 'gibibytes': 1024 ** 3,
3716 'terabytes': 1000 ** 4,
3717 'tebibytes': 1024 ** 4,
3723 'petabytes': 1000 ** 5,
3724 'pebibytes': 1024 ** 5,
3730 'exabytes': 1000 ** 6,
3731 'exbibytes': 1024 ** 6,
3737 'zettabytes': 1000 ** 7,
3738 'zebibytes': 1024 ** 7,
3744 'yottabytes': 1000 ** 8,
3745 'yobibytes': 1024 ** 8,
3748 return lookup_unit_table(_UNIT_TABLE
, s
)
3757 if re
.match(r
'^[\d,.]+$', s
):
3758 return str_to_int(s
)
3769 return lookup_unit_table(_UNIT_TABLE
, s
)
3772 def parse_resolution(s
):
3776 mobj
= re
.search(r
'(?<![a-zA-Z0-9])(?P<w>\d+)\s*[xX×,]\s*(?P<h>\d+)(?![a-zA-Z0-9])', s
)
3779 'width': int(mobj
.group('w')),
3780 'height': int(mobj
.group('h')),
3783 mobj
= re
.search(r
'(?<![a-zA-Z0-9])(\d+)[pPiI](?![a-zA-Z0-9])', s
)
3785 return {'height': int(mobj.group(1))}
3787 mobj
= re
.search(r
'\b([48])[kK]\b', s
)
3789 return {'height': int(mobj.group(1)) * 540}
3794 def parse_bitrate(s
):
3795 if not isinstance(s
, compat_str
):
3797 mobj
= re
.search(r
'\b(\d+)\s*kbps', s
)
3799 return int(mobj
.group(1))
3802 def month_by_name(name
, lang
='en'):
3803 """ Return the number of a month by (locale-independently) English name """
3805 month_names
= MONTH_NAMES
.get(lang
, MONTH_NAMES
['en'])
3808 return month_names
.index(name
) + 1
3813 def month_by_abbreviation(abbrev
):
3814 """ Return the number of a month by (locale-independently) English
3818 return [s
[:3] for s
in ENGLISH_MONTH_NAMES
].index(abbrev
) + 1
3823 def fix_xml_ampersands(xml_str
):
3824 """Replace all the '&' by '&' in XML"""
3826 r
'&(?!amp;|lt;|gt;|apos;|quot;|#x[0-9a-fA-F]{,4};|#[0-9]{,4};)',
3831 def setproctitle(title
):
3832 assert isinstance(title
, compat_str
)
3834 # ctypes in Jython is not complete
3835 # http://bugs.jython.org/issue2148
3836 if sys
.platform
.startswith('java'):
3840 libc
= ctypes
.cdll
.LoadLibrary('libc.so.6')
3844 # LoadLibrary in Windows Python 2.7.13 only expects
3845 # a bytestring, but since unicode_literals turns
3846 # every string into a unicode string, it fails.
3848 title_bytes
= title
.encode('utf-8')
3849 buf
= ctypes
.create_string_buffer(len(title_bytes
))
3850 buf
.value
= title_bytes
3852 libc
.prctl(15, buf
, 0, 0, 0)
3853 except AttributeError:
3854 return # Strange libc, just skip this
3857 def remove_start(s
, start
):
3858 return s
[len(start
):] if s
is not None and s
.startswith(start
) else s
3861 def remove_end(s
, end
):
3862 return s
[:-len(end
)] if s
is not None and s
.endswith(end
) else s
3865 def remove_quotes(s
):
3866 if s
is None or len(s
) < 2:
3868 for quote
in ('"', "'", ):
3869 if s
[0] == quote
and s
[-1] == quote
:
3874 def get_domain(url
):
3875 domain
= re
.match(r
'(?:https?:\/\/)?(?:www\.)?(?P<domain>[^\n\/]+\.[^\n\/]+)(?:\/(.*))?', url
)
3876 return domain
.group('domain') if domain
else None
3879 def url_basename(url
):
3880 path
= compat_urlparse
.urlparse(url
).path
3881 return path
.strip('/').split('/')[-1]
3885 return re
.match(r
'https?://[^?#&]+/', url
).group()
3888 def urljoin(base
, path
):
3889 if isinstance(path
, bytes):
3890 path
= path
.decode('utf-8')
3891 if not isinstance(path
, compat_str
) or not path
:
3893 if re
.match(r
'^(?:[a-zA-Z][a-zA-Z0-9+-.]*:)?//', path
):
3895 if isinstance(base
, bytes):
3896 base
= base
.decode('utf-8')
3897 if not isinstance(base
, compat_str
) or not re
.match(
3898 r
'^(?:https?:)?//', base
):
3900 return compat_urlparse
.urljoin(base
, path
)
3903 class HEADRequest(compat_urllib_request
.Request
):
3904 def get_method(self
):
3908 class PUTRequest(compat_urllib_request
.Request
):
3909 def get_method(self
):
3913 def int_or_none(v
, scale
=1, default
=None, get_attr
=None, invscale
=1):
3916 v
= getattr(v
, get_attr
, None)
3922 return int(v
) * invscale
// scale
3923 except (ValueError, TypeError, OverflowError):
3927 def str_or_none(v
, default
=None):
3928 return default
if v
is None else compat_str(v
)
3931 def str_to_int(int_str
):
3932 """ A more relaxed version of int_or_none """
3933 if isinstance(int_str
, compat_integer_types
):
3935 elif isinstance(int_str
, compat_str
):
3936 int_str
= re
.sub(r
'[,\.\+]', '', int_str
)
3937 return int_or_none(int_str
)
3940 def float_or_none(v
, scale
=1, invscale
=1, default
=None):
3944 return float(v
) * invscale
/ scale
3945 except (ValueError, TypeError):
3949 def bool_or_none(v
, default
=None):
3950 return v
if isinstance(v
, bool) else default
3953 def strip_or_none(v
, default
=None):
3954 return v
.strip() if isinstance(v
, compat_str
) else default
3957 def url_or_none(url
):
3958 if not url
or not isinstance(url
, compat_str
):
3961 return url
if re
.match(r
'^(?:(?:https?|rt(?:m(?:pt?[es]?|fp)|sp[su]?)|mms|ftps?):)?//', url
) else None
3964 def strftime_or_none(timestamp
, date_format
, default
=None):
3965 datetime_object
= None
3967 if isinstance(timestamp
, compat_numeric_types
): # unix timestamp
3968 datetime_object
= datetime
.datetime
.utcfromtimestamp(timestamp
)
3969 elif isinstance(timestamp
, compat_str
): # assume YYYYMMDD
3970 datetime_object
= datetime
.datetime
.strptime(timestamp
, '%Y%m%d')
3971 return datetime_object
.strftime(date_format
)
3972 except (ValueError, TypeError, AttributeError):
3976 def parse_duration(s
):
3977 if not isinstance(s
, compat_basestring
):
3982 days
, hours
, mins
, secs
, ms
= [None] * 5
3983 m
= re
.match(r
'(?:(?:(?:(?P<days>[0-9]+):)?(?P<hours>[0-9]+):)?(?P<mins>[0-9]+):)?(?P<secs>[0-9]+)(?P<ms>\.[0-9]+)?Z?$', s
)
3985 days
, hours
, mins
, secs
, ms
= m
.groups()
3990 [0-9]+\s*y(?:ears?)?\s*
3993 [0-9]+\s*m(?:onths?)?\s*
3996 [0-9]+\s*w(?:eeks?)?\s*
3999 (?P<days>[0-9]+)\s*d(?:ays?)?\s*
4003 (?P<hours>[0-9]+)\s*h(?:ours?)?\s*
4006 (?P<mins>[0-9]+)\s*m(?:in(?:ute)?s?)?\s*
4009 (?P<secs>[0-9]+)(?P<ms>\.[0-9]+)?\s*s(?:ec(?:ond)?s?)?\s*
4012 days
, hours
, mins
, secs
, ms
= m
.groups()
4014 m
= re
.match(r
'(?i)(?:(?P<hours>[0-9.]+)\s*(?:hours?)|(?P<mins>[0-9.]+)\s*(?:mins?\.?|minutes?)\s*)Z?$', s
)
4016 hours
, mins
= m
.groups()
4022 duration
+= float(secs
)
4024 duration
+= float(mins
) * 60
4026 duration
+= float(hours
) * 60 * 60
4028 duration
+= float(days
) * 24 * 60 * 60
4030 duration
+= float(ms
)
4034 def prepend_extension(filename
, ext
, expected_real_ext
=None):
4035 name
, real_ext
= os
.path
.splitext(filename
)
4037 '{0}.{1}{2}'.format(name
, ext
, real_ext
)
4038 if not expected_real_ext
or real_ext
[1:] == expected_real_ext
4039 else '{0}.{1}'.format(filename
, ext
))
4042 def replace_extension(filename
, ext
, expected_real_ext
=None):
4043 name
, real_ext
= os
.path
.splitext(filename
)
4044 return '{0}.{1}'.format(
4045 name
if not expected_real_ext
or real_ext
[1:] == expected_real_ext
else filename
,
4049 def check_executable(exe
, args
=[]):
4050 """ Checks if the given binary is installed somewhere in PATH, and returns its name.
4051 args can be a list of arguments for a short output (like -version) """
4053 Popen([exe
] + args
, stdout
=subprocess
.PIPE
, stderr
=subprocess
.PIPE
).communicate_or_kill()
4059 def _get_exe_version_output(exe
, args
):
4061 # STDIN should be redirected too. On UNIX-like systems, ffmpeg triggers
4062 # SIGTTOU if yt-dlp is run in the background.
4063 # See https://github.com/ytdl-org/youtube-dl/issues/955#issuecomment-209789656
4065 [encodeArgument(exe
)] + args
, stdin
=subprocess
.PIPE
,
4066 stdout
=subprocess
.PIPE
, stderr
=subprocess
.STDOUT
).communicate_or_kill()
4069 if isinstance(out
, bytes): # Python 2.x
4070 out
= out
.decode('ascii', 'ignore')
4074 def detect_exe_version(output
, version_re
=None, unrecognized
='present'):
4075 assert isinstance(output
, compat_str
)
4076 if version_re
is None:
4077 version_re
= r
'version\s+([-0-9._a-zA-Z]+)'
4078 m
= re
.search(version_re
, output
)
4085 def get_exe_version(exe
, args
=['--version'],
4086 version_re
=None, unrecognized
='present'):
4087 """ Returns the version of the specified executable,
4088 or False if the executable is not present """
4089 out
= _get_exe_version_output(exe
, args
)
4090 return detect_exe_version(out
, version_re
, unrecognized
) if out
else False
4093 class LazyList(collections
.abc
.Sequence
):
4094 ''' Lazy immutable list from an iterable
4095 Note that slices of a LazyList are lists and not LazyList'''
4097 class IndexError(IndexError):
4100 def __init__(self
, iterable
, *, reverse
=False, _cache
=None):
4101 self
.__iterable
= iter(iterable
)
4102 self
.__cache
= [] if _cache
is None else _cache
4103 self
.__reversed
= reverse
4107 # We need to consume the entire iterable to iterate in reverse
4108 yield from self
.exhaust()
4110 yield from self
.__cache
4111 for item
in self
.__iterable
:
4112 self
.__cache
.append(item
)
4115 def __exhaust(self
):
4116 self
.__cache
.extend(self
.__iterable
)
4117 # Discard the emptied iterable to make it pickle-able
4118 self
.__iterable
= []
4122 ''' Evaluate the entire iterable '''
4123 return self
.__exhaust
()[::-1 if self
.__reversed
else 1]
4126 def __reverse_index(x
):
4127 return None if x
is None else -(x
+ 1)
4129 def __getitem__(self
, idx
):
4130 if isinstance(idx
, slice):
4132 idx
= slice(self
.__reverse
_index
(idx
.start
), self
.__reverse
_index
(idx
.stop
), -(idx
.step
or 1))
4133 start
, stop
, step
= idx
.start
, idx
.stop
, idx
.step
or 1
4134 elif isinstance(idx
, int):
4136 idx
= self
.__reverse
_index
(idx
)
4137 start
, stop
, step
= idx
, idx
, 0
4139 raise TypeError('indices must be integers or slices')
4140 if ((start
or 0) < 0 or (stop
or 0) < 0
4141 or (start
is None and step
< 0)
4142 or (stop
is None and step
> 0)):
4143 # We need to consume the entire iterable to be able to slice from the end
4144 # Obviously, never use this with infinite iterables
4147 return self
.__cache
[idx
]
4148 except IndexError as e
:
4149 raise self
.IndexError(e
) from e
4150 n
= max(start
or 0, stop
or 0) - len(self
.__cache
) + 1
4152 self
.__cache
.extend(itertools
.islice(self
.__iterable
, n
))
4154 return self
.__cache
[idx
]
4155 except IndexError as e
:
4156 raise self
.IndexError(e
) from e
4160 self
[-1] if self
.__reversed
else self
[0]
4161 except self
.IndexError:
4167 return len(self
.__cache
)
4169 def __reversed__(self
):
4170 return type(self
)(self
.__iterable
, reverse
=not self
.__reversed
, _cache
=self
.__cache
)
4173 return type(self
)(self
.__iterable
, reverse
=self
.__reversed
, _cache
=self
.__cache
)
4175 def __deepcopy__(self
, memo
):
4176 # FIXME: This is actually just a shallow copy
4178 memo
[id_
] = self
.__copy
__()
4182 # repr and str should mimic a list. So we exhaust the iterable
4183 return repr(self
.exhaust())
4186 return repr(self
.exhaust())
4191 class IndexError(IndexError):
4195 # This is only useful for tests
4196 return len(self
.getslice())
4198 def __init__(self
, pagefunc
, pagesize
, use_cache
=True):
4199 self
._pagefunc
= pagefunc
4200 self
._pagesize
= pagesize
4201 self
._use
_cache
= use_cache
4204 def getpage(self
, pagenum
):
4205 page_results
= self
._cache
.get(pagenum
)
4206 if page_results
is None:
4207 page_results
= list(self
._pagefunc
(pagenum
))
4209 self
._cache
[pagenum
] = page_results
4212 def getslice(self
, start
=0, end
=None):
4213 return list(self
._getslice
(start
, end
))
4215 def _getslice(self
, start
, end
):
4216 raise NotImplementedError('This method must be implemented by subclasses')
4218 def __getitem__(self
, idx
):
4219 # NOTE: cache must be enabled if this is used
4220 if not isinstance(idx
, int) or idx
< 0:
4221 raise TypeError('indices must be non-negative integers')
4222 entries
= self
.getslice(idx
, idx
+ 1)
4224 raise self
.IndexError()
4228 class OnDemandPagedList(PagedList
):
4229 def _getslice(self
, start
, end
):
4230 for pagenum
in itertools
.count(start
// self
._pagesize
):
4231 firstid
= pagenum
* self
._pagesize
4232 nextfirstid
= pagenum
* self
._pagesize
+ self
._pagesize
4233 if start
>= nextfirstid
:
4237 start
% self
._pagesize
4238 if firstid
<= start
< nextfirstid
4241 ((end
- 1) % self
._pagesize
) + 1
4242 if (end
is not None and firstid
<= end
<= nextfirstid
)
4245 page_results
= self
.getpage(pagenum
)
4246 if startv
!= 0 or endv
is not None:
4247 page_results
= page_results
[startv
:endv
]
4248 yield from page_results
4250 # A little optimization - if current page is not "full", ie. does
4251 # not contain page_size videos then we can assume that this page
4252 # is the last one - there are no more ids on further pages -
4253 # i.e. no need to query again.
4254 if len(page_results
) + startv
< self
._pagesize
:
4257 # If we got the whole page, but the next page is not interesting,
4258 # break out early as well
4259 if end
== nextfirstid
:
4263 class InAdvancePagedList(PagedList
):
4264 def __init__(self
, pagefunc
, pagecount
, pagesize
):
4265 self
._pagecount
= pagecount
4266 PagedList
.__init
__(self
, pagefunc
, pagesize
, True)
4268 def _getslice(self
, start
, end
):
4269 start_page
= start
// self
._pagesize
4271 self
._pagecount
if end
is None else (end
// self
._pagesize
+ 1))
4272 skip_elems
= start
- start_page
* self
._pagesize
4273 only_more
= None if end
is None else end
- start
4274 for pagenum
in range(start_page
, end_page
):
4275 page_results
= self
.getpage(pagenum
)
4277 page_results
= page_results
[skip_elems
:]
4279 if only_more
is not None:
4280 if len(page_results
) < only_more
:
4281 only_more
-= len(page_results
)
4283 yield from page_results
[:only_more
]
4285 yield from page_results
4288 def uppercase_escape(s
):
4289 unicode_escape
= codecs
.getdecoder('unicode_escape')
4291 r
'\\U[0-9a-fA-F]{8}',
4292 lambda m
: unicode_escape(m
.group(0))[0],
4296 def lowercase_escape(s
):
4297 unicode_escape
= codecs
.getdecoder('unicode_escape')
4299 r
'\\u[0-9a-fA-F]{4}',
4300 lambda m
: unicode_escape(m
.group(0))[0],
4304 def escape_rfc3986(s
):
4305 """Escape non-ASCII characters as suggested by RFC 3986"""
4306 if sys
.version_info
< (3, 0) and isinstance(s
, compat_str
):
4307 s
= s
.encode('utf-8')
4308 return compat_urllib_parse
.quote(s
, b
"%/;:@&=+$,!~*'()?#[]")
4311 def escape_url(url
):
4312 """Escape URL as suggested by RFC 3986"""
4313 url_parsed
= compat_urllib_parse_urlparse(url
)
4314 return url_parsed
._replace
(
4315 netloc
=url_parsed
.netloc
.encode('idna').decode('ascii'),
4316 path
=escape_rfc3986(url_parsed
.path
),
4317 params
=escape_rfc3986(url_parsed
.params
),
4318 query
=escape_rfc3986(url_parsed
.query
),
4319 fragment
=escape_rfc3986(url_parsed
.fragment
)
4324 return compat_parse_qs(compat_urllib_parse_urlparse(url
).query
)
4327 def read_batch_urls(batch_fd
):
4329 if not isinstance(url
, compat_str
):
4330 url
= url
.decode('utf-8', 'replace')
4331 BOM_UTF8
= ('\xef\xbb\xbf', '\ufeff')
4332 for bom
in BOM_UTF8
:
4333 if url
.startswith(bom
):
4334 url
= url
[len(bom
):]
4336 if not url
or url
.startswith(('#', ';', ']')):
4338 # "#" cannot be stripped out since it is part of the URI
4339 # However, it can be safely stipped out if follwing a whitespace
4340 return re
.split(r
'\s#', url
, 1)[0].rstrip()
4342 with contextlib
.closing(batch_fd
) as fd
:
4343 return [url
for url
in map(fixup
, fd
) if url
]
4346 def urlencode_postdata(*args
, **kargs
):
4347 return compat_urllib_parse_urlencode(*args
, **kargs
).encode('ascii')
4350 def update_url_query(url
, query
):
4353 parsed_url
= compat_urlparse
.urlparse(url
)
4354 qs
= compat_parse_qs(parsed_url
.query
)
4356 return compat_urlparse
.urlunparse(parsed_url
._replace
(
4357 query
=compat_urllib_parse_urlencode(qs
, True)))
4360 def update_Request(req
, url
=None, data
=None, headers
={}, query={}
):
4361 req_headers
= req
.headers
.copy()
4362 req_headers
.update(headers
)
4363 req_data
= data
or req
.data
4364 req_url
= update_url_query(url
or req
.get_full_url(), query
)
4365 req_get_method
= req
.get_method()
4366 if req_get_method
== 'HEAD':
4367 req_type
= HEADRequest
4368 elif req_get_method
== 'PUT':
4369 req_type
= PUTRequest
4371 req_type
= compat_urllib_request
.Request
4373 req_url
, data
=req_data
, headers
=req_headers
,
4374 origin_req_host
=req
.origin_req_host
, unverifiable
=req
.unverifiable
)
4375 if hasattr(req
, 'timeout'):
4376 new_req
.timeout
= req
.timeout
4380 def _multipart_encode_impl(data
, boundary
):
4381 content_type
= 'multipart/form-data; boundary=%s' % boundary
4384 for k
, v
in data
.items():
4385 out
+= b
'--' + boundary
.encode('ascii') + b
'\r\n'
4386 if isinstance(k
, compat_str
):
4387 k
= k
.encode('utf-8')
4388 if isinstance(v
, compat_str
):
4389 v
= v
.encode('utf-8')
4390 # RFC 2047 requires non-ASCII field names to be encoded, while RFC 7578
4391 # suggests sending UTF-8 directly. Firefox sends UTF-8, too
4392 content
= b
'Content-Disposition: form-data; name="' + k
+ b
'"\r\n\r\n' + v
+ b
'\r\n'
4393 if boundary
.encode('ascii') in content
:
4394 raise ValueError('Boundary overlaps with data')
4397 out
+= b
'--' + boundary
.encode('ascii') + b
'--\r\n'
4399 return out
, content_type
4402 def multipart_encode(data
, boundary
=None):
4404 Encode a dict to RFC 7578-compliant form-data
4407 A dict where keys and values can be either Unicode or bytes-like
4410 If specified a Unicode object, it's used as the boundary. Otherwise
4411 a random boundary is generated.
4413 Reference: https://tools.ietf.org/html/rfc7578
4415 has_specified_boundary
= boundary
is not None
4418 if boundary
is None:
4419 boundary
= '---------------' + str(random
.randrange(0x0fffffff, 0xffffffff))
4422 out
, content_type
= _multipart_encode_impl(data
, boundary
)
4425 if has_specified_boundary
:
4429 return out
, content_type
4432 def dict_get(d
, key_or_keys
, default
=None, skip_false_values
=True):
4433 if isinstance(key_or_keys
, (list, tuple)):
4434 for key
in key_or_keys
:
4435 if key
not in d
or d
[key
] is None or skip_false_values
and not d
[key
]:
4439 return d
.get(key_or_keys
, default
)
4442 def try_get(src
, getter
, expected_type
=None):
4443 for get
in variadic(getter
):
4446 except (AttributeError, KeyError, TypeError, IndexError):
4449 if expected_type
is None or isinstance(v
, expected_type
):
4453 def merge_dicts(*dicts
):
4455 for a_dict
in dicts
:
4456 for k
, v
in a_dict
.items():
4460 or (isinstance(v
, compat_str
) and v
4461 and isinstance(merged
[k
], compat_str
)
4462 and not merged
[k
])):
4467 def encode_compat_str(string
, encoding
=preferredencoding(), errors
='strict'):
4468 return string
if isinstance(string
, compat_str
) else compat_str(string
, encoding
, errors
)
4480 TV_PARENTAL_GUIDELINES
= {
4490 def parse_age_limit(s
):
4492 return s
if 0 <= s
<= 21 else None
4493 if not isinstance(s
, compat_basestring
):
4495 m
= re
.match(r
'^(?P<age>\d{1,2})\+?$', s
)
4497 return int(m
.group('age'))
4500 return US_RATINGS
[s
]
4501 m
= re
.match(r
'^TV[_-]?(%s)$' % '|'.join(k
[3:] for k
in TV_PARENTAL_GUIDELINES
), s
)
4503 return TV_PARENTAL_GUIDELINES
['TV-' + m
.group(1)]
4507 def strip_jsonp(code
):
4510 (?:window\.)?(?P<func_name>[a-zA-Z0-9_.$]*)
4511 (?:\s*&&\s*(?P=func_name))?
4512 \s*\(\s*(?P<callback_data>.*)\);?
4513 \s*?(?://[^\n]*)*$''',
4514 r
'\g<callback_data>', code
)
4517 def js_to_json(code
, vars={}):
4518 # vars is a dict of var, val pairs to substitute
4519 COMMENT_RE
= r
'/\*(?:(?!\*/).)*?\*/|//[^\n]*\n'
4520 SKIP_RE
= r
'\s*(?:{comment})?\s*'.format(comment
=COMMENT_RE
)
4522 (r
'(?s)^(0[xX][0-9a-fA-F]+){skip}:?$'.format(skip
=SKIP_RE
), 16),
4523 (r
'(?s)^(0+[0-7]+){skip}:?$'.format(skip
=SKIP_RE
), 8),
4528 if v
in ('true', 'false', 'null'):
4530 elif v
in ('undefined', 'void 0'):
4532 elif v
.startswith('/*') or v
.startswith('//') or v
.startswith('!') or v
== ',':
4535 if v
[0] in ("'", '"'):
4536 v
= re
.sub(r
'(?s)\\.|"', lambda m
: {
4541 }.get(m
.group(0), m
.group(0)), v
[1:-1])
4543 for regex
, base
in INTEGER_TABLE
:
4544 im
= re
.match(regex
, v
)
4546 i
= int(im
.group(1), base
)
4547 return '"%d":' % i
if v
.endswith(':') else '%d' % i
4554 return re
.sub(r
'''(?sx)
4555 "(?:[^"\\]*(?:\\\\|\\['"nurtbfx/\n]))*[^"\\]*"|
4556 '(?:[^'\\]*(?:\\\\|\\['"nurtbfx/\n]))*[^'\\]*'|
4557 {comment}|,(?={skip}[\]}}])|
4558 void\s0|(?:(?<![0-9])[eE]|[a-df-zA-DF-Z_$])[.a-zA-Z_$0-9]*|
4559 \b(?:0[xX][0-9a-fA-F]+|0+[0-7]+)(?:{skip}:)?|
4562 '''.format(comment
=COMMENT_RE
, skip
=SKIP_RE
), fix_kv
, code
)
4565 def qualities(quality_ids
):
4566 """ Get a numeric quality value out of a list of possible values """
4569 return quality_ids
.index(qid
)
4576 'default': '%(title)s [%(id)s].%(ext)s',
4577 'chapter': '%(title)s - %(section_number)03d %(section_title)s [%(id)s].%(ext)s',
4583 'description': 'description',
4584 'annotation': 'annotations.xml',
4585 'infojson': 'info.json',
4587 'pl_thumbnail': None,
4588 'pl_description': 'description',
4589 'pl_infojson': 'info.json',
4592 # As of [1] format syntax is:
4593 # %[mapping_key][conversion_flags][minimum_width][.precision][length_modifier]type
4594 # 1. https://docs.python.org/2/library/stdtypes.html#string-formatting
4595 STR_FORMAT_RE_TMPL
= r
'''(?x)
4596 (?<!%)(?P<prefix>(?:%%)*)
4598 (?P<has_key>\((?P<key>{0})\))?
4600 (?P<conversion>[#0\-+ ]+)?
4602 (?P<precision>\.\d+)?
4603 (?P<len_mod>[hlL])? # unused in python
4604 {1} # conversion type
4609 STR_FORMAT_TYPES
= 'diouxXeEfFgGcrs'
4612 def limit_length(s
, length
):
4613 """ Add ellipses to overly long strings """
4618 return s
[:length
- len(ELLIPSES
)] + ELLIPSES
4622 def version_tuple(v
):
4623 return tuple(int(e
) for e
in re
.split(r
'[-.]', v
))
4626 def is_outdated_version(version
, limit
, assume_new
=True):
4628 return not assume_new
4630 return version_tuple(version
) < version_tuple(limit
)
4632 return not assume_new
4635 def ytdl_is_updateable():
4636 """ Returns if yt-dlp can be updated with -U """
4638 from .update
import is_non_updateable
4640 return not is_non_updateable()
4643 def args_to_str(args
):
4644 # Get a short string representation for a subprocess command
4645 return ' '.join(compat_shlex_quote(a
) for a
in args
)
4648 def error_to_compat_str(err
):
4650 # On python 2 error byte string must be decoded with proper
4651 # encoding rather than ascii
4652 if sys
.version_info
[0] < 3:
4653 err_str
= err_str
.decode(preferredencoding())
4657 def mimetype2ext(mt
):
4661 mt
, _
, params
= mt
.partition(';')
4666 # Per RFC 3003, audio/mpeg can be .mp1, .mp2 or .mp3. Here use .mp3 as
4667 # it's the most popular one
4668 'audio/mpeg': 'mp3',
4669 'audio/x-wav': 'wav',
4671 'audio/wave': 'wav',
4674 ext
= FULL_MAP
.get(mt
)
4680 'smptett+xml': 'tt',
4684 'x-mp4-fragmented': 'mp4',
4685 'x-ms-sami': 'sami',
4688 'x-mpegurl': 'm3u8',
4689 'vnd.apple.mpegurl': 'm3u8',
4693 'vnd.ms-sstr+xml': 'ism',
4697 'filmstrip+json': 'fs',
4701 _
, _
, subtype
= mt
.rpartition('/')
4702 ext
= SUBTYPE_MAP
.get(subtype
.lower())
4713 _
, _
, suffix
= subtype
.partition('+')
4714 ext
= SUFFIX_MAP
.get(suffix
)
4718 return subtype
.replace('+', '.')
4721 def parse_codecs(codecs_str
):
4722 # http://tools.ietf.org/html/rfc6381
4725 split_codecs
= list(filter(None, map(
4726 str.strip
, codecs_str
.strip().strip(',').split(','))))
4727 vcodec
, acodec
, hdr
= None, None, None
4728 for full_codec
in split_codecs
:
4729 parts
= full_codec
.split('.')
4730 codec
= parts
[0].replace('0', '')
4731 if codec
in ('avc1', 'avc2', 'avc3', 'avc4', 'vp9', 'vp8', 'hev1', 'hev2',
4732 'h263', 'h264', 'mp4v', 'hvc1', 'av1', 'theora', 'dvh1', 'dvhe'):
4734 vcodec
= '.'.join(parts
[:4]) if codec
in ('vp9', 'av1') else full_codec
4735 if codec
in ('dvh1', 'dvhe'):
4737 elif codec
== 'av1' and len(parts
) > 3 and parts
[3] == '10':
4739 elif full_codec
.replace('0', '').startswith('vp9.2'):
4741 elif codec
in ('mp4a', 'opus', 'vorbis', 'mp3', 'aac', 'ac-3', 'ec-3', 'eac3', 'dtsc', 'dtse', 'dtsh', 'dtsl'):
4745 write_string('WARNING: Unknown codec %s\n' % full_codec
, sys
.stderr
)
4746 if not vcodec
and not acodec
:
4747 if len(split_codecs
) == 2:
4749 'vcodec': split_codecs
[0],
4750 'acodec': split_codecs
[1],
4754 'vcodec': vcodec
or 'none',
4755 'acodec': acodec
or 'none',
4756 'dynamic_range': hdr
,
4761 def urlhandle_detect_ext(url_handle
):
4762 getheader
= url_handle
.headers
.get
4764 cd
= getheader('Content-Disposition')
4766 m
= re
.match(r
'attachment;\s*filename="(?P<filename>[^"]+)"', cd
)
4768 e
= determine_ext(m
.group('filename'), default_ext
=None)
4772 return mimetype2ext(getheader('Content-Type'))
4775 def encode_data_uri(data
, mime_type
):
4776 return 'data:%s;base64,%s' % (mime_type
, base64
.b64encode(data
).decode('ascii'))
4779 def age_restricted(content_limit
, age_limit
):
4780 """ Returns True iff the content should be blocked """
4782 if age_limit
is None: # No limit set
4784 if content_limit
is None:
4785 return False # Content available for everyone
4786 return age_limit
< content_limit
4789 def is_html(first_bytes
):
4790 """ Detect whether a file contains HTML by examining its first bytes. """
4793 (b
'\xef\xbb\xbf', 'utf-8'),
4794 (b
'\x00\x00\xfe\xff', 'utf-32-be'),
4795 (b
'\xff\xfe\x00\x00', 'utf-32-le'),
4796 (b
'\xff\xfe', 'utf-16-le'),
4797 (b
'\xfe\xff', 'utf-16-be'),
4799 for bom
, enc
in BOMS
:
4800 if first_bytes
.startswith(bom
):
4801 s
= first_bytes
[len(bom
):].decode(enc
, 'replace')
4804 s
= first_bytes
.decode('utf-8', 'replace')
4806 return re
.match(r
'^\s*<', s
)
4809 def determine_protocol(info_dict
):
4810 protocol
= info_dict
.get('protocol')
4811 if protocol
is not None:
4814 url
= sanitize_url(info_dict
['url'])
4815 if url
.startswith('rtmp'):
4817 elif url
.startswith('mms'):
4819 elif url
.startswith('rtsp'):
4822 ext
= determine_ext(url
)
4828 return compat_urllib_parse_urlparse(url
).scheme
4831 def render_table(header_row
, data
, delim
=False, extra_gap
=0, hide_empty
=False):
4832 """ Render a list of rows, each as a list of values.
4833 Text after a \t will be right aligned """
4835 return len(remove_terminal_sequences(string
).replace('\t', ''))
4837 def get_max_lens(table
):
4838 return [max(width(str(v
)) for v
in col
) for col
in zip(*table
)]
4840 def filter_using_list(row
, filterArray
):
4841 return [col
for (take
, col
) in zip(filterArray
, row
) if take
]
4844 max_lens
= get_max_lens(data
)
4845 header_row
= filter_using_list(header_row
, max_lens
)
4846 data
= [filter_using_list(row
, max_lens
) for row
in data
]
4848 table
= [header_row
] + data
4849 max_lens
= get_max_lens(table
)
4852 table
= [header_row
, [delim
* (ml
+ extra_gap
) for ml
in max_lens
]] + data
4853 table
[1][-1] = table
[1][-1][:-extra_gap
] # Remove extra_gap from end of delimiter
4855 for pos
, text
in enumerate(map(str, row
)):
4857 row
[pos
] = text
.replace('\t', ' ' * (max_lens
[pos
] - width(text
))) + ' ' * extra_gap
4859 row
[pos
] = text
+ ' ' * (max_lens
[pos
] - width(text
) + extra_gap
)
4860 ret
= '\n'.join(''.join(row
).rstrip() for row
in table
)
4864 def _match_one(filter_part
, dct
, incomplete
):
4865 # TODO: Generalize code with YoutubeDL._build_format_filter
4866 STRING_OPERATORS
= {
4867 '*=': operator
.contains
,
4868 '^=': lambda attr
, value
: attr
.startswith(value
),
4869 '$=': lambda attr
, value
: attr
.endswith(value
),
4870 '~=': lambda attr
, value
: re
.search(value
, attr
),
4872 COMPARISON_OPERATORS
= {
4874 '<=': operator
.le
, # "<=" must be defined above "<"
4881 operator_rex
= re
.compile(r
'''(?x)\s*
4883 \s*(?P<negation>!\s*)?(?P<op>%s)(?P<none_inclusive>\s*\?)?\s*
4885 (?P<quote>["\'])(?P
<quotedstrval
>.+?
)(?P
=quote
)|
4889 ''' % '|'.join(map(re.escape, COMPARISON_OPERATORS.keys())))
4890 m = operator_rex.search(filter_part)
4893 unnegated_op = COMPARISON_OPERATORS[m['op']]
4895 op = lambda attr, value: not unnegated_op(attr, value)
4898 comparison_value = m['quotedstrval'] or m['strval'] or m['intval']
4900 comparison_value = comparison_value.replace(r'\%s' % m['quote'], m['quote'])
4901 actual_value = dct.get(m['key'])
4902 numeric_comparison = None
4903 if isinstance(actual_value, compat_numeric_types):
4904 # If the original field is a string and matching comparisonvalue is
4905 # a number we should respect the origin of the original field
4906 # and process comparison value as a string (see
4907 # https://github.com/ytdl-org/youtube-dl/issues/11082)
4909 numeric_comparison = int(comparison_value)
4911 numeric_comparison = parse_filesize(comparison_value)
4912 if numeric_comparison is None:
4913 numeric_comparison = parse_filesize(f'{comparison_value}B')
4914 if numeric_comparison is None:
4915 numeric_comparison = parse_duration(comparison_value)
4916 if numeric_comparison is not None and m['op'] in STRING_OPERATORS:
4917 raise ValueError('Operator %s only supports string values!' % m['op'])
4918 if actual_value is None:
4919 return incomplete or m['none_inclusive']
4920 return op(actual_value, comparison_value if numeric_comparison is None else numeric_comparison)
4923 '': lambda v: (v is True) if isinstance(v, bool) else (v is not None),
4924 '!': lambda v: (v is False) if isinstance(v, bool) else (v is None),
4926 operator_rex = re.compile(r'''(?x
)\s
*
4927 (?P
<op
>%s)\s
*(?P
<key
>[a
-z_
]+)
4929 ''' % '|'.join(map(re.escape, UNARY_OPERATORS.keys())))
4930 m = operator_rex.search(filter_part)
4932 op = UNARY_OPERATORS[m.group('op')]
4933 actual_value = dct.get(m.group('key'))
4934 if incomplete and actual_value is None:
4936 return op(actual_value)
4938 raise ValueError('Invalid filter part %r' % filter_part)
4941 def match_str(filter_str, dct, incomplete=False):
4942 """ Filter a dictionary with a simple string syntax. Returns True (=passes filter) or false
4943 When incomplete, all conditions passes on missing fields
4946 _match_one(filter_part.replace(r'\&', '&'), dct, incomplete)
4947 for filter_part in re.split(r'(?<!\\)&', filter_str))
4950 def match_filter_func(filter_str):
4951 def _match_func(info_dict, *args, **kwargs):
4952 if match_str(filter_str, info_dict, *args, **kwargs):
4955 video_title = info_dict.get('title', info_dict.get('id', 'video'))
4956 return '%s does not pass filter %s, skipping ..' % (video_title, filter_str)
4960 def parse_dfxp_time_expr(time_expr):
4964 mobj = re.match(r'^(?P<time_offset>\d+(?:\.\d+)?)s?$', time_expr)
4966 return float(mobj.group('time_offset'))
4968 mobj = re.match(r'^(\d+):(\d\d):(\d\d(?:(?:\.|:)\d+)?)$', time_expr)
4970 return 3600 * int(mobj.group(1)) + 60 * int(mobj.group(2)) + float(mobj.group(3).replace(':', '.'))
4973 def srt_subtitles_timecode(seconds):
4974 return '%02d:%02d:%02d,%03d' % timetuple_from_msec(seconds * 1000)
4977 def ass_subtitles_timecode(seconds):
4978 time = timetuple_from_msec(seconds * 1000)
4979 return '%01d:%02d:%02d.%02d' % (*time[:-1], time.milliseconds / 10)
4982 def dfxp2srt(dfxp_data):
4984 @param dfxp_data A
bytes-like
object containing DFXP data
4985 @returns A
unicode object containing converted SRT data
4987 LEGACY_NAMESPACES = (
4988 (b'http://www.w3.org/ns/ttml', [
4989 b'http://www.w3.org/2004/11/ttaf1',
4990 b'http://www.w3.org/2006/04/ttaf1',
4991 b'http://www.w3.org/2006/10/ttaf1',
4993 (b'http://www.w3.org/ns/ttml#styling', [
4994 b'http://www.w3.org/ns/ttml#style',
4998 SUPPORTED_STYLING = [
5007 _x = functools.partial(xpath_with_ns, ns_map={
5008 'xml': 'http://www.w3.org/XML/1998/namespace',
5009 'ttml': 'http://www.w3.org/ns/ttml',
5010 'tts': 'http://www.w3.org/ns/ttml#styling',
5016 class TTMLPElementParser(object):
5018 _unclosed_elements = []
5019 _applied_styles = []
5021 def start(self, tag, attrib):
5022 if tag in (_x('ttml:br'), 'br'):
5025 unclosed_elements = []
5027 element_style_id = attrib.get('style')
5029 style.update(default_style)
5030 if element_style_id:
5031 style.update(styles.get(element_style_id, {}))
5032 for prop in SUPPORTED_STYLING:
5033 prop_val = attrib.get(_x('tts:' + prop))
5035 style[prop] = prop_val
5038 for k, v in sorted(style.items()):
5039 if self._applied_styles and self._applied_styles[-1].get(k) == v:
5042 font += ' color="%s"' % v
5043 elif k == 'fontSize':
5044 font += ' size="%s"' % v
5045 elif k == 'fontFamily':
5046 font += ' face="%s"' % v
5047 elif k == 'fontWeight' and v == 'bold':
5049 unclosed_elements.append('b')
5050 elif k == 'fontStyle' and v == 'italic':
5052 unclosed_elements.append('i')
5053 elif k == 'textDecoration' and v == 'underline':
5055 unclosed_elements.append('u')
5057 self._out += '<font' + font + '>'
5058 unclosed_elements.append('font')
5060 if self._applied_styles:
5061 applied_style.update(self._applied_styles[-1])
5062 applied_style.update(style)
5063 self._applied_styles.append(applied_style)
5064 self._unclosed_elements.append(unclosed_elements)
5067 if tag not in (_x('ttml:br'), 'br'):
5068 unclosed_elements = self._unclosed_elements.pop()
5069 for element in reversed(unclosed_elements):
5070 self._out += '</%s>' % element
5071 if unclosed_elements and self._applied_styles:
5072 self._applied_styles.pop()
5074 def data(self, data):
5078 return self._out.strip()
5080 def parse_node(node):
5081 target = TTMLPElementParser()
5082 parser = xml.etree.ElementTree.XMLParser(target=target)
5083 parser.feed(xml.etree.ElementTree.tostring(node))
5084 return parser.close()
5086 for k, v in LEGACY_NAMESPACES:
5088 dfxp_data = dfxp_data.replace(ns, k)
5090 dfxp = compat_etree_fromstring(dfxp_data)
5092 paras = dfxp.findall(_x('.//ttml:p')) or dfxp.findall('.//p')
5095 raise ValueError('Invalid dfxp/TTML subtitle')
5099 for style in dfxp.findall(_x('.//ttml:style')):
5100 style_id = style.get('id') or style.get(_x('xml:id'))
5103 parent_style_id = style.get('style')
5105 if parent_style_id not in styles:
5108 styles[style_id] = styles[parent_style_id].copy()
5109 for prop in SUPPORTED_STYLING:
5110 prop_val = style.get(_x('tts:' + prop))
5112 styles.setdefault(style_id, {})[prop] = prop_val
5118 for p in ('body', 'div'):
5119 ele = xpath_element(dfxp, [_x('.//ttml:' + p), './/' + p])
5122 style = styles.get(ele.get('style'))
5125 default_style.update(style)
5127 for para, index in zip(paras, itertools.count(1)):
5128 begin_time = parse_dfxp_time_expr(para.attrib.get('begin'))
5129 end_time = parse_dfxp_time_expr(para.attrib.get('end'))
5130 dur = parse_dfxp_time_expr(para.attrib.get('dur'))
5131 if begin_time is None:
5136 end_time = begin_time + dur
5137 out.append('%d\n%s --> %s\n%s\n\n' % (
5139 srt_subtitles_timecode(begin_time),
5140 srt_subtitles_timecode(end_time),
5146 def cli_option(params, command_option, param):
5147 param = params.get(param)
5149 param = compat_str(param)
5150 return [command_option, param] if param is not None else []
5153 def cli_bool_option(params, command_option, param, true_value='true', false_value='false', separator=None):
5154 param = params.get(param)
5157 assert isinstance(param, bool)
5159 return [command_option + separator + (true_value if param else false_value)]
5160 return [command_option, true_value if param else false_value]
5163 def cli_valueless_option(params, command_option, param, expected_value=True):
5164 param = params.get(param)
5165 return [command_option] if param == expected_value else []
5168 def cli_configuration_args(argdict, keys, default=[], use_compat=True):
5169 if isinstance(argdict, (list, tuple)): # for backward compatibility
5176 assert isinstance(argdict, dict)
5178 assert isinstance(keys, (list, tuple))
5179 for key_list in keys:
5180 arg_list = list(filter(
5181 lambda x: x is not None,
5182 [argdict.get(key.lower()) for key in variadic(key_list)]))
5184 return [arg for args in arg_list for arg in args]
5188 def _configuration_args(main_key, argdict, exe, keys=None, default=[], use_compat=True):
5189 main_key, exe = main_key.lower(), exe.lower()
5190 root_key = exe if main_key == exe else f'{main_key}+{exe}'
5191 keys = [f'{root_key}{k}' for k in (keys or [''])]
5192 if root_key in keys:
5194 keys.append((main_key, exe))
5195 keys.append('default')
5198 return cli_configuration_args(argdict, keys, default, use_compat)
5201 class ISO639Utils(object):
5202 # See http://www.loc.gov/standards/iso639-2/ISO-639-2_utf-8.txt
5261 'iw': 'heb', # Replaced by he in 1989 revision
5271 'in': 'ind', # Replaced by id in 1989 revision
5386 'ji': 'yid', # Replaced by yi in 1989 revision
5394 def short2long(cls, code):
5395 """Convert language code from ISO 639-1 to ISO 639-2/T"""
5396 return cls._lang_map.get(code[:2])
5399 def long2short(cls, code):
5400 """Convert language code from ISO 639-2/T to ISO 639-1"""
5401 for short_name, long_name in cls._lang_map.items():
5402 if long_name == code:
5406 class ISO3166Utils(object):
5407 # From http://data.okfn.org/data/core/country-list
5409 'AF': 'Afghanistan',
5410 'AX': 'Åland Islands',
5413 'AS': 'American Samoa',
5418 'AG': 'Antigua and Barbuda',
5435 'BO': 'Bolivia, Plurinational State of',
5436 'BQ': 'Bonaire, Sint Eustatius and Saba',
5437 'BA': 'Bosnia and Herzegovina',
5439 'BV': 'Bouvet Island',
5441 'IO': 'British Indian Ocean Territory',
5442 'BN': 'Brunei Darussalam',
5444 'BF': 'Burkina Faso',
5450 'KY': 'Cayman Islands',
5451 'CF': 'Central African Republic',
5455 'CX': 'Christmas Island',
5456 'CC': 'Cocos (Keeling) Islands',
5460 'CD': 'Congo, the Democratic Republic of the',
5461 'CK': 'Cook Islands',
5463 'CI': 'Côte d\'Ivoire',
5468 'CZ': 'Czech Republic',
5472 'DO': 'Dominican Republic',
5475 'SV': 'El Salvador',
5476 'GQ': 'Equatorial Guinea',
5480 'FK': 'Falkland Islands (Malvinas)',
5481 'FO': 'Faroe Islands',
5485 'GF': 'French Guiana',
5486 'PF': 'French Polynesia',
5487 'TF': 'French Southern Territories',
5502 'GW': 'Guinea-Bissau',
5505 'HM': 'Heard Island and McDonald Islands',
5506 'VA': 'Holy See (Vatican City State)',
5513 'IR': 'Iran, Islamic Republic of',
5516 'IM': 'Isle of Man',
5526 'KP': 'Korea, Democratic People\'s Republic of',
5527 'KR': 'Korea, Republic of',
5530 'LA': 'Lao People\'s Democratic Republic',
5536 'LI': 'Liechtenstein',
5540 'MK': 'Macedonia, the Former Yugoslav Republic of',
5547 'MH': 'Marshall Islands',
5553 'FM': 'Micronesia, Federated States of',
5554 'MD': 'Moldova, Republic of',
5565 'NL': 'Netherlands',
5566 'NC': 'New Caledonia',
5567 'NZ': 'New Zealand',
5572 'NF': 'Norfolk Island',
5573 'MP': 'Northern Mariana Islands',
5578 'PS': 'Palestine, State of',
5580 'PG': 'Papua New Guinea',
5583 'PH': 'Philippines',
5587 'PR': 'Puerto Rico',
5591 'RU': 'Russian Federation',
5593 'BL': 'Saint Barthélemy',
5594 'SH': 'Saint Helena, Ascension and Tristan da Cunha',
5595 'KN': 'Saint Kitts and Nevis',
5596 'LC': 'Saint Lucia',
5597 'MF': 'Saint Martin (French part)',
5598 'PM': 'Saint Pierre and Miquelon',
5599 'VC': 'Saint Vincent and the Grenadines',
5602 'ST': 'Sao Tome and Principe',
5603 'SA': 'Saudi Arabia',
5607 'SL': 'Sierra Leone',
5609 'SX': 'Sint Maarten (Dutch part)',
5612 'SB': 'Solomon Islands',
5614 'ZA': 'South Africa',
5615 'GS': 'South Georgia and the South Sandwich Islands',
5616 'SS': 'South Sudan',
5621 'SJ': 'Svalbard and Jan Mayen',
5624 'CH': 'Switzerland',
5625 'SY': 'Syrian Arab Republic',
5626 'TW': 'Taiwan, Province of China',
5628 'TZ': 'Tanzania, United Republic of',
5630 'TL': 'Timor-Leste',
5634 'TT': 'Trinidad and Tobago',
5637 'TM': 'Turkmenistan',
5638 'TC': 'Turks and Caicos Islands',
5642 'AE': 'United Arab Emirates',
5643 'GB': 'United Kingdom',
5644 'US': 'United States',
5645 'UM': 'United States Minor Outlying Islands',
5649 'VE': 'Venezuela, Bolivarian Republic of',
5651 'VG': 'Virgin Islands, British',
5652 'VI': 'Virgin Islands, U.S.',
5653 'WF': 'Wallis and Futuna',
5654 'EH': 'Western Sahara',
5661 def short2full(cls, code):
5662 """Convert an ISO 3166-2 country code to the corresponding full name"""
5663 return cls._country_map.get(code.upper())
5666 class GeoUtils(object):
5667 # Major IPv4 address blocks per country
5669 'AD': '46.172.224.0/19',
5670 'AE': '94.200.0.0/13',
5671 'AF': '149.54.0.0/17',
5672 'AG': '209.59.64.0/18',
5673 'AI': '204.14.248.0/21',
5674 'AL': '46.99.0.0/16',
5675 'AM': '46.70.0.0/15',
5676 'AO': '105.168.0.0/13',
5677 'AP': '182.50.184.0/21',
5678 'AQ': '23.154.160.0/24',
5679 'AR': '181.0.0.0/12',
5680 'AS': '202.70.112.0/20',
5681 'AT': '77.116.0.0/14',
5682 'AU': '1.128.0.0/11',
5683 'AW': '181.41.0.0/18',
5684 'AX': '185.217.4.0/22',
5685 'AZ': '5.197.0.0/16',
5686 'BA': '31.176.128.0/17',
5687 'BB': '65.48.128.0/17',
5688 'BD': '114.130.0.0/16',
5690 'BF': '102.178.0.0/15',
5691 'BG': '95.42.0.0/15',
5692 'BH': '37.131.0.0/17',
5693 'BI': '154.117.192.0/18',
5694 'BJ': '137.255.0.0/16',
5695 'BL': '185.212.72.0/23',
5696 'BM': '196.12.64.0/18',
5697 'BN': '156.31.0.0/16',
5698 'BO': '161.56.0.0/16',
5699 'BQ': '161.0.80.0/20',
5700 'BR': '191.128.0.0/12',
5701 'BS': '24.51.64.0/18',
5702 'BT': '119.2.96.0/19',
5703 'BW': '168.167.0.0/16',
5704 'BY': '178.120.0.0/13',
5705 'BZ': '179.42.192.0/18',
5706 'CA': '99.224.0.0/11',
5707 'CD': '41.243.0.0/16',
5708 'CF': '197.242.176.0/21',
5709 'CG': '160.113.0.0/16',
5710 'CH': '85.0.0.0/13',
5711 'CI': '102.136.0.0/14',
5712 'CK': '202.65.32.0/19',
5713 'CL': '152.172.0.0/14',
5714 'CM': '102.244.0.0/14',
5715 'CN': '36.128.0.0/10',
5716 'CO': '181.240.0.0/12',
5717 'CR': '201.192.0.0/12',
5718 'CU': '152.206.0.0/15',
5719 'CV': '165.90.96.0/19',
5720 'CW': '190.88.128.0/17',
5721 'CY': '31.153.0.0/16',
5722 'CZ': '88.100.0.0/14',
5724 'DJ': '197.241.0.0/17',
5725 'DK': '87.48.0.0/12',
5726 'DM': '192.243.48.0/20',
5727 'DO': '152.166.0.0/15',
5728 'DZ': '41.96.0.0/12',
5729 'EC': '186.68.0.0/15',
5730 'EE': '90.190.0.0/15',
5731 'EG': '156.160.0.0/11',
5732 'ER': '196.200.96.0/20',
5733 'ES': '88.0.0.0/11',
5734 'ET': '196.188.0.0/14',
5735 'EU': '2.16.0.0/13',
5736 'FI': '91.152.0.0/13',
5737 'FJ': '144.120.0.0/16',
5738 'FK': '80.73.208.0/21',
5739 'FM': '119.252.112.0/20',
5740 'FO': '88.85.32.0/19',
5742 'GA': '41.158.0.0/15',
5744 'GD': '74.122.88.0/21',
5745 'GE': '31.146.0.0/16',
5746 'GF': '161.22.64.0/18',
5747 'GG': '62.68.160.0/19',
5748 'GH': '154.160.0.0/12',
5749 'GI': '95.164.0.0/16',
5750 'GL': '88.83.0.0/19',
5751 'GM': '160.182.0.0/15',
5752 'GN': '197.149.192.0/18',
5753 'GP': '104.250.0.0/19',
5754 'GQ': '105.235.224.0/20',
5755 'GR': '94.64.0.0/13',
5756 'GT': '168.234.0.0/16',
5757 'GU': '168.123.0.0/16',
5758 'GW': '197.214.80.0/20',
5759 'GY': '181.41.64.0/18',
5760 'HK': '113.252.0.0/14',
5761 'HN': '181.210.0.0/16',
5762 'HR': '93.136.0.0/13',
5763 'HT': '148.102.128.0/17',
5764 'HU': '84.0.0.0/14',
5765 'ID': '39.192.0.0/10',
5766 'IE': '87.32.0.0/12',
5767 'IL': '79.176.0.0/13',
5768 'IM': '5.62.80.0/20',
5769 'IN': '117.192.0.0/10',
5770 'IO': '203.83.48.0/21',
5771 'IQ': '37.236.0.0/14',
5772 'IR': '2.176.0.0/12',
5773 'IS': '82.221.0.0/16',
5774 'IT': '79.0.0.0/10',
5775 'JE': '87.244.64.0/18',
5776 'JM': '72.27.0.0/17',
5777 'JO': '176.29.0.0/16',
5778 'JP': '133.0.0.0/8',
5779 'KE': '105.48.0.0/12',
5780 'KG': '158.181.128.0/17',
5781 'KH': '36.37.128.0/17',
5782 'KI': '103.25.140.0/22',
5783 'KM': '197.255.224.0/20',
5784 'KN': '198.167.192.0/19',
5785 'KP': '175.45.176.0/22',
5786 'KR': '175.192.0.0/10',
5787 'KW': '37.36.0.0/14',
5788 'KY': '64.96.0.0/15',
5789 'KZ': '2.72.0.0/13',
5790 'LA': '115.84.64.0/18',
5791 'LB': '178.135.0.0/16',
5792 'LC': '24.92.144.0/20',
5793 'LI': '82.117.0.0/19',
5794 'LK': '112.134.0.0/15',
5795 'LR': '102.183.0.0/16',
5796 'LS': '129.232.0.0/17',
5797 'LT': '78.56.0.0/13',
5798 'LU': '188.42.0.0/16',
5799 'LV': '46.109.0.0/16',
5800 'LY': '41.252.0.0/14',
5801 'MA': '105.128.0.0/11',
5802 'MC': '88.209.64.0/18',
5803 'MD': '37.246.0.0/16',
5804 'ME': '178.175.0.0/17',
5805 'MF': '74.112.232.0/21',
5806 'MG': '154.126.0.0/17',
5807 'MH': '117.103.88.0/21',
5808 'MK': '77.28.0.0/15',
5809 'ML': '154.118.128.0/18',
5810 'MM': '37.111.0.0/17',
5811 'MN': '49.0.128.0/17',
5812 'MO': '60.246.0.0/16',
5813 'MP': '202.88.64.0/20',
5814 'MQ': '109.203.224.0/19',
5815 'MR': '41.188.64.0/18',
5816 'MS': '208.90.112.0/22',
5817 'MT': '46.11.0.0/16',
5818 'MU': '105.16.0.0/12',
5819 'MV': '27.114.128.0/18',
5820 'MW': '102.70.0.0/15',
5821 'MX': '187.192.0.0/11',
5822 'MY': '175.136.0.0/13',
5823 'MZ': '197.218.0.0/15',
5824 'NA': '41.182.0.0/16',
5825 'NC': '101.101.0.0/18',
5826 'NE': '197.214.0.0/18',
5827 'NF': '203.17.240.0/22',
5828 'NG': '105.112.0.0/12',
5829 'NI': '186.76.0.0/15',
5830 'NL': '145.96.0.0/11',
5831 'NO': '84.208.0.0/13',
5832 'NP': '36.252.0.0/15',
5833 'NR': '203.98.224.0/19',
5834 'NU': '49.156.48.0/22',
5835 'NZ': '49.224.0.0/14',
5836 'OM': '5.36.0.0/15',
5837 'PA': '186.72.0.0/15',
5838 'PE': '186.160.0.0/14',
5839 'PF': '123.50.64.0/18',
5840 'PG': '124.240.192.0/19',
5841 'PH': '49.144.0.0/13',
5842 'PK': '39.32.0.0/11',
5843 'PL': '83.0.0.0/11',
5844 'PM': '70.36.0.0/20',
5845 'PR': '66.50.0.0/16',
5846 'PS': '188.161.0.0/16',
5847 'PT': '85.240.0.0/13',
5848 'PW': '202.124.224.0/20',
5849 'PY': '181.120.0.0/14',
5850 'QA': '37.210.0.0/15',
5851 'RE': '102.35.0.0/16',
5852 'RO': '79.112.0.0/13',
5853 'RS': '93.86.0.0/15',
5854 'RU': '5.136.0.0/13',
5855 'RW': '41.186.0.0/16',
5856 'SA': '188.48.0.0/13',
5857 'SB': '202.1.160.0/19',
5858 'SC': '154.192.0.0/11',
5859 'SD': '102.120.0.0/13',
5860 'SE': '78.64.0.0/12',
5861 'SG': '8.128.0.0/10',
5862 'SI': '188.196.0.0/14',
5863 'SK': '78.98.0.0/15',
5864 'SL': '102.143.0.0/17',
5865 'SM': '89.186.32.0/19',
5866 'SN': '41.82.0.0/15',
5867 'SO': '154.115.192.0/18',
5868 'SR': '186.179.128.0/17',
5869 'SS': '105.235.208.0/21',
5870 'ST': '197.159.160.0/19',
5871 'SV': '168.243.0.0/16',
5872 'SX': '190.102.0.0/20',
5874 'SZ': '41.84.224.0/19',
5875 'TC': '65.255.48.0/20',
5876 'TD': '154.68.128.0/19',
5877 'TG': '196.168.0.0/14',
5878 'TH': '171.96.0.0/13',
5879 'TJ': '85.9.128.0/18',
5880 'TK': '27.96.24.0/21',
5881 'TL': '180.189.160.0/20',
5882 'TM': '95.85.96.0/19',
5883 'TN': '197.0.0.0/11',
5884 'TO': '175.176.144.0/21',
5885 'TR': '78.160.0.0/11',
5886 'TT': '186.44.0.0/15',
5887 'TV': '202.2.96.0/19',
5888 'TW': '120.96.0.0/11',
5889 'TZ': '156.156.0.0/14',
5890 'UA': '37.52.0.0/14',
5891 'UG': '102.80.0.0/13',
5893 'UY': '167.56.0.0/13',
5894 'UZ': '84.54.64.0/18',
5895 'VA': '212.77.0.0/19',
5896 'VC': '207.191.240.0/21',
5897 'VE': '186.88.0.0/13',
5898 'VG': '66.81.192.0/20',
5899 'VI': '146.226.0.0/16',
5900 'VN': '14.160.0.0/11',
5901 'VU': '202.80.32.0/20',
5902 'WF': '117.20.32.0/21',
5903 'WS': '202.4.32.0/19',
5904 'YE': '134.35.0.0/16',
5905 'YT': '41.242.116.0/22',
5906 'ZA': '41.0.0.0/11',
5907 'ZM': '102.144.0.0/13',
5908 'ZW': '102.177.192.0/18',
5912 def random_ipv4(cls, code_or_block):
5913 if len(code_or_block) == 2:
5914 block = cls._country_ip_map.get(code_or_block.upper())
5918 block = code_or_block
5919 addr, preflen = block.split('/')
5920 addr_min = compat_struct_unpack('!L', socket.inet_aton(addr))[0]
5921 addr_max = addr_min | (0xffffffff >> int(preflen))
5922 return compat_str(socket.inet_ntoa(
5923 compat_struct_pack('!L', random.randint(addr_min, addr_max))))
5926 class PerRequestProxyHandler(compat_urllib_request.ProxyHandler):
5927 def __init__(self, proxies=None):
5928 # Set default handlers
5929 for type in ('http', 'https'):
5930 setattr(self, '%s_open' % type,
5931 lambda r, proxy='__noproxy__', type=type, meth=self.proxy_open:
5932 meth(r, proxy, type))
5933 compat_urllib_request.ProxyHandler.__init__(self, proxies)
5935 def proxy_open(self, req, proxy, type):
5936 req_proxy = req.headers.get('Ytdl-request-proxy')
5937 if req_proxy is not None:
5939 del req.headers['Ytdl-request-proxy']
5941 if proxy == '__noproxy__':
5942 return None # No Proxy
5943 if compat_urlparse.urlparse(proxy).scheme.lower() in ('socks', 'socks4', 'socks4a', 'socks5'):
5944 req.add_header('Ytdl-socks-proxy', proxy)
5945 # yt-dlp's http/https handlers do wrapping the socket with socks
5947 return compat_urllib_request.ProxyHandler.proxy_open(
5948 self, req, proxy, type)
5951 # Both long_to_bytes and bytes_to_long are adapted from PyCrypto, which is
5952 # released into Public Domain
5953 # https://github.com/dlitz/pycrypto/blob/master/lib/Crypto/Util/number.py#L387
5955 def long_to_bytes(n, blocksize=0):
5956 """long_to_bytes(n:long, blocksize:int) : string
5957 Convert a long integer to a byte string.
5959 If optional blocksize is given and greater than zero, pad the front of the
5960 byte string with binary zeros so that the length is a multiple of
5963 # after much testing, this algorithm was deemed to be the fastest
5967 s = compat_struct_pack('>I', n & 0xffffffff) + s
5969 # strip off leading zeros
5970 for i in range(len(s)):
5971 if s[i] != b'\000'[0]:
5974 # only happens when n == 0
5978 # add back some pad bytes. this could be done more efficiently w.r.t. the
5979 # de-padding being done above, but sigh...
5980 if blocksize > 0 and len(s) % blocksize:
5981 s = (blocksize - len(s) % blocksize) * b'\000' + s
5985 def bytes_to_long(s):
5986 """bytes_to_long(string) : long
5987 Convert a byte string to a long integer.
5989 This is (essentially) the inverse of long_to_bytes().
5994 extra = (4 - length % 4)
5995 s = b'\000' * extra + s
5996 length = length + extra
5997 for i in range(0, length, 4):
5998 acc = (acc << 32) + compat_struct_unpack('>I', s[i:i + 4])[0]
6002 def ohdave_rsa_encrypt(data, exponent, modulus):
6004 Implement OHDave
's RSA algorithm. See http://www.ohdave.com/rsa/
6007 data: data to encrypt, bytes-like object
6008 exponent, modulus: parameter e and N of RSA algorithm, both integer
6009 Output: hex string of encrypted data
6011 Limitation: supports one block encryption only
6014 payload = int(binascii.hexlify(data[::-1]), 16)
6015 encrypted = pow(payload, exponent, modulus)
6016 return '%x' % encrypted
6019 def pkcs1pad(data, length):
6021 Padding input data with PKCS#1 scheme
6023 @param {int[]} data input data
6024 @param {int} length target length
6025 @returns {int[]} padded data
6027 if len(data) > length - 11:
6028 raise ValueError('Input data too
long for PKCS
#1 padding')
6030 pseudo_random
= [random
.randint(0, 254) for _
in range(length
- len(data
) - 3)]
6031 return [0, 2] + pseudo_random
+ [0] + data
6034 def encode_base_n(num
, n
, table
=None):
6035 FULL_TABLE
= '0123456789abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ'
6037 table
= FULL_TABLE
[:n
]
6040 raise ValueError('base %d exceeds table length %d' % (n
, len(table
)))
6047 ret
= table
[num
% n
] + ret
6052 def decode_packed_codes(code
):
6053 mobj
= re
.search(PACKED_CODES_RE
, code
)
6054 obfuscated_code
, base
, count
, symbols
= mobj
.groups()
6057 symbols
= symbols
.split('|')
6062 base_n_count
= encode_base_n(count
, base
)
6063 symbol_table
[base_n_count
] = symbols
[count
] or base_n_count
6066 r
'\b(\w+)\b', lambda mobj
: symbol_table
[mobj
.group(0)],
6070 def caesar(s
, alphabet
, shift
):
6075 alphabet
[(alphabet
.index(c
) + shift
) % l
] if c
in alphabet
else c
6080 return caesar(s
, r
'''!"#$%&'()*+,-./0123456789:;<=>?@ABCDEFGHIJKLMNOPQRSTUVWXYZ[\]^_`abcdefghijklmnopqrstuvwxyz{|}~''', 47)
6083 def parse_m3u8_attributes(attrib
):
6085 for (key
, val
) in re
.findall(r
'(?P<key>[A-Z0-9-]+)=(?P<val>"[^"]+"|[^",]+)(?:,|$)', attrib
):
6086 if val
.startswith('"'):
6092 def urshift(val
, n
):
6093 return val
>> n
if val
>= 0 else (val
+ 0x100000000) >> n
6096 # Based on png2str() written by @gdkchan and improved by @yokrysty
6097 # Originally posted at https://github.com/ytdl-org/youtube-dl/issues/9706
6098 def decode_png(png_data
):
6099 # Reference: https://www.w3.org/TR/PNG/
6100 header
= png_data
[8:]
6102 if png_data
[:8] != b
'\x89PNG\x0d\x0a\x1a\x0a' or header
[4:8] != b
'IHDR':
6103 raise IOError('Not a valid PNG file.')
6105 int_map
= {1: '>B', 2: '>H', 4: '>I'}
6106 unpack_integer
= lambda x
: compat_struct_unpack(int_map
[len(x
)], x
)[0]
6111 length
= unpack_integer(header
[:4])
6114 chunk_type
= header
[:4]
6117 chunk_data
= header
[:length
]
6118 header
= header
[length
:]
6120 header
= header
[4:] # Skip CRC
6128 ihdr
= chunks
[0]['data']
6130 width
= unpack_integer(ihdr
[:4])
6131 height
= unpack_integer(ihdr
[4:8])
6135 for chunk
in chunks
:
6136 if chunk
['type'] == b
'IDAT':
6137 idat
+= chunk
['data']
6140 raise IOError('Unable to read PNG data.')
6142 decompressed_data
= bytearray(zlib
.decompress(idat
))
6147 def _get_pixel(idx
):
6152 for y
in range(height
):
6153 basePos
= y
* (1 + stride
)
6154 filter_type
= decompressed_data
[basePos
]
6158 pixels
.append(current_row
)
6160 for x
in range(stride
):
6161 color
= decompressed_data
[1 + basePos
+ x
]
6162 basex
= y
* stride
+ x
6167 left
= _get_pixel(basex
- 3)
6169 up
= _get_pixel(basex
- stride
)
6171 if filter_type
== 1: # Sub
6172 color
= (color
+ left
) & 0xff
6173 elif filter_type
== 2: # Up
6174 color
= (color
+ up
) & 0xff
6175 elif filter_type
== 3: # Average
6176 color
= (color
+ ((left
+ up
) >> 1)) & 0xff
6177 elif filter_type
== 4: # Paeth
6183 c
= _get_pixel(basex
- stride
- 3)
6191 if pa
<= pb
and pa
<= pc
:
6192 color
= (color
+ a
) & 0xff
6194 color
= (color
+ b
) & 0xff
6196 color
= (color
+ c
) & 0xff
6198 current_row
.append(color
)
6200 return width
, height
, pixels
6203 def write_xattr(path
, key
, value
):
6204 # This mess below finds the best xattr tool for the job
6206 # try the pyxattr module...
6209 if hasattr(xattr
, 'set'): # pyxattr
6210 # Unicode arguments are not supported in python-pyxattr until
6212 # See https://github.com/ytdl-org/youtube-dl/issues/5498
6213 pyxattr_required_version
= '0.5.0'
6214 if version_tuple(xattr
.__version
__) < version_tuple(pyxattr_required_version
):
6215 # TODO: fallback to CLI tools
6216 raise XAttrUnavailableError(
6217 'python-pyxattr is detected but is too old. '
6218 'yt-dlp requires %s or above while your version is %s. '
6219 'Falling back to other xattr implementations' % (
6220 pyxattr_required_version
, xattr
.__version
__))
6222 setxattr
= xattr
.set
6224 setxattr
= xattr
.setxattr
6227 setxattr(path
, key
, value
)
6228 except EnvironmentError as e
:
6229 raise XAttrMetadataError(e
.errno
, e
.strerror
)
6232 if compat_os_name
== 'nt':
6233 # Write xattrs to NTFS Alternate Data Streams:
6234 # http://en.wikipedia.org/wiki/NTFS#Alternate_data_streams_.28ADS.29
6235 assert ':' not in key
6236 assert os
.path
.exists(path
)
6238 ads_fn
= path
+ ':' + key
6240 with open(ads_fn
, 'wb') as f
:
6242 except EnvironmentError as e
:
6243 raise XAttrMetadataError(e
.errno
, e
.strerror
)
6245 user_has_setfattr
= check_executable('setfattr', ['--version'])
6246 user_has_xattr
= check_executable('xattr', ['-h'])
6248 if user_has_setfattr
or user_has_xattr
:
6250 value
= value
.decode('utf-8')
6251 if user_has_setfattr
:
6252 executable
= 'setfattr'
6253 opts
= ['-n', key
, '-v', value
]
6254 elif user_has_xattr
:
6255 executable
= 'xattr'
6256 opts
= ['-w', key
, value
]
6258 cmd
= ([encodeFilename(executable
, True)]
6259 + [encodeArgument(o
) for o
in opts
]
6260 + [encodeFilename(path
, True)])
6264 cmd
, stdout
=subprocess
.PIPE
, stderr
=subprocess
.PIPE
, stdin
=subprocess
.PIPE
)
6265 except EnvironmentError as e
:
6266 raise XAttrMetadataError(e
.errno
, e
.strerror
)
6267 stdout
, stderr
= p
.communicate_or_kill()
6268 stderr
= stderr
.decode('utf-8', 'replace')
6269 if p
.returncode
!= 0:
6270 raise XAttrMetadataError(p
.returncode
, stderr
)
6273 # On Unix, and can't find pyxattr, setfattr, or xattr.
6274 if sys
.platform
.startswith('linux'):
6275 raise XAttrUnavailableError(
6276 "Couldn't find a tool to set the xattrs. "
6277 "Install either the python 'pyxattr' or 'xattr' "
6278 "modules, or the GNU 'attr' package "
6279 "(which contains the 'setfattr' tool).")
6281 raise XAttrUnavailableError(
6282 "Couldn't find a tool to set the xattrs. "
6283 "Install either the python 'xattr' module, "
6284 "or the 'xattr' binary.")
6287 def random_birthday(year_field
, month_field
, day_field
):
6288 start_date
= datetime
.date(1950, 1, 1)
6289 end_date
= datetime
.date(1995, 12, 31)
6290 offset
= random
.randint(0, (end_date
- start_date
).days
)
6291 random_date
= start_date
+ datetime
.timedelta(offset
)
6293 year_field
: str(random_date
.year
),
6294 month_field
: str(random_date
.month
),
6295 day_field
: str(random_date
.day
),
6299 # Templates for internet shortcut files, which are plain text files.
6300 DOT_URL_LINK_TEMPLATE
= '''
6305 DOT_WEBLOC_LINK_TEMPLATE
= '''
6306 <?xml version="1.0" encoding="UTF-8"?>
6307 <!DOCTYPE plist PUBLIC "-//Apple//DTD PLIST 1.0//EN" "http://www.apple.com/DTDs/PropertyList-1.0.dtd">
6308 <plist version="1.0">
6311 \t<string>%(url)s</string>
6316 DOT_DESKTOP_LINK_TEMPLATE
= '''
6326 'url': DOT_URL_LINK_TEMPLATE
,
6327 'desktop': DOT_DESKTOP_LINK_TEMPLATE
,
6328 'webloc': DOT_WEBLOC_LINK_TEMPLATE
,
6332 def iri_to_uri(iri
):
6334 Converts an IRI (Internationalized Resource Identifier, allowing Unicode characters) to a URI (Uniform Resource Identifier, ASCII-only).
6336 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.
6339 iri_parts
= compat_urllib_parse_urlparse(iri
)
6341 if '[' in iri_parts
.netloc
:
6342 raise ValueError('IPv6 URIs are not, yet, supported.')
6343 # Querying `.netloc`, when there's only one bracket, also raises a ValueError.
6345 # 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.
6348 if iri_parts
.username
:
6349 net_location
+= compat_urllib_parse_quote(iri_parts
.username
, safe
=r
"!$%&'()*+,~")
6350 if iri_parts
.password
is not None:
6351 net_location
+= ':' + compat_urllib_parse_quote(iri_parts
.password
, safe
=r
"!$%&'()*+,~")
6354 net_location
+= iri_parts
.hostname
.encode('idna').decode('utf-8') # Punycode for Unicode hostnames.
6355 # The 'idna' encoding produces ASCII text.
6356 if iri_parts
.port
is not None and iri_parts
.port
!= 80:
6357 net_location
+= ':' + str(iri_parts
.port
)
6359 return compat_urllib_parse_urlunparse(
6363 compat_urllib_parse_quote_plus(iri_parts
.path
, safe
=r
"!$%&'()*+,/:;=@|~"),
6365 # Unsure about the `safe` argument, since this is a legacy way of handling parameters.
6366 compat_urllib_parse_quote_plus(iri_parts
.params
, safe
=r
"!$%&'()*+,/:;=@|~"),
6368 # Not totally sure about the `safe` argument, since the source does not explicitly mention the query URI component.
6369 compat_urllib_parse_quote_plus(iri_parts
.query
, safe
=r
"!$%&'()*+,/:;=?@{|}~"),
6371 compat_urllib_parse_quote_plus(iri_parts
.fragment
, safe
=r
"!#$%&'()*+,/:;=?@{|}~")))
6373 # Source for `safe` arguments: https://url.spec.whatwg.org/#percent-encoded-bytes.
6376 def to_high_limit_path(path
):
6377 if sys
.platform
in ['win32', 'cygwin']:
6378 # Work around MAX_PATH limitation on Windows. The maximum allowed length for the individual path segments may still be quite limited.
6379 return r
'\\?\ '.rstrip() + os
.path
.abspath(path
)
6384 def format_field(obj
, field
=None, template
='%s', ignore
=(None, ''), default
='', func
=None):
6386 val
= obj
if obj
is not None else default
6388 val
= obj
.get(field
, default
)
6389 if func
and val
not in ignore
:
6391 return template
% val
if val
not in ignore
else default
6394 def clean_podcast_url(url
):
6395 return re
.sub(r
'''(?x)
6399 media\.blubrry\.com| # https://create.blubrry.com/resources/podcast-media-download-statistics/getting-started/
6402 (?:dts|www)\.podtrac\.com/(?:pts/)?redirect\.[0-9a-z]{3,4}| # http://analytics.podtrac.com/how-to-measure
6405 cn\.co| # https://podcorn.com/analytics-prefix/
6406 st\.fm # https://podsights.com/docs/
6411 _HEX_TABLE
= '0123456789abcdef'
6414 def random_uuidv4():
6415 return re
.sub(r
'[xy]', lambda x
: _HEX_TABLE
[random
.randint(0, 15)], 'xxxxxxxx-xxxx-4xxx-yxxx-xxxxxxxxxxxx')
6418 def make_dir(path
, to_screen
=None):
6420 dn
= os
.path
.dirname(path
)
6421 if dn
and not os
.path
.exists(dn
):
6424 except (OSError, IOError) as err
:
6425 if callable(to_screen
) is not None:
6426 to_screen('unable to create directory ' + error_to_compat_str(err
))
6430 def get_executable_path():
6431 from zipimport
import zipimporter
6432 if hasattr(sys
, 'frozen'): # Running from PyInstaller
6433 path
= os
.path
.dirname(sys
.executable
)
6434 elif isinstance(globals().get('__loader__'), zipimporter
): # Running from ZIP
6435 path
= os
.path
.join(os
.path
.dirname(__file__
), '../..')
6437 path
= os
.path
.join(os
.path
.dirname(__file__
), '..')
6438 return os
.path
.abspath(path
)
6441 def load_plugins(name
, suffix
, namespace
):
6444 plugins_spec
= importlib
.util
.spec_from_file_location(
6445 name
, os
.path
.join(get_executable_path(), 'ytdlp_plugins', name
, '__init__.py'))
6446 plugins
= importlib
.util
.module_from_spec(plugins_spec
)
6447 sys
.modules
[plugins_spec
.name
] = plugins
6448 plugins_spec
.loader
.exec_module(plugins
)
6449 for name
in dir(plugins
):
6450 if name
in namespace
:
6452 if not name
.endswith(suffix
):
6454 klass
= getattr(plugins
, name
)
6455 classes
[name
] = namespace
[name
] = klass
6456 except FileNotFoundError
:
6462 obj
, *path_list
, default
=None, expected_type
=None, get_all
=True,
6463 casesense
=True, is_user_input
=False, traverse_string
=False):
6464 ''' Traverse nested list/dict/tuple
6465 @param path_list A list of paths which are checked one by one.
6466 Each path is a list of keys where each key is a string,
6467 a function, a tuple of strings or "...".
6468 When a fuction is given, it takes the key as argument and
6469 returns whether the key matches or not. When a tuple is given,
6470 all the keys given in the tuple are traversed, and
6471 "..." traverses all the keys in the object
6472 @param default Default value to return
6473 @param expected_type Only accept final value of this type (Can also be any callable)
6474 @param get_all Return all the values obtained from a path or only the first one
6475 @param casesense Whether to consider dictionary keys as case sensitive
6476 @param is_user_input Whether the keys are generated from user input. If True,
6477 strings are converted to int/slice if necessary
6478 @param traverse_string Whether to traverse inside strings. If True, any
6479 non-compatible object will also be converted into a string
6483 _lower
= lambda k
: (k
.lower() if isinstance(k
, str) else k
)
6484 path_list
= (map(_lower
, variadic(path
)) for path
in path_list
)
6486 def _traverse_obj(obj
, path
, _current_depth
=0):
6488 path
= tuple(variadic(path
))
6489 for i
, key
in enumerate(path
):
6492 if isinstance(key
, (list, tuple)):
6493 obj
= [_traverse_obj(obj
, sub_key
, _current_depth
) for sub_key
in key
]
6496 obj
= (obj
.values() if isinstance(obj
, dict)
6497 else obj
if isinstance(obj
, (list, tuple, LazyList
))
6498 else str(obj
) if traverse_string
else [])
6500 depth
= max(depth
, _current_depth
)
6501 return [_traverse_obj(inner_obj
, path
[i
+ 1:], _current_depth
) for inner_obj
in obj
]
6503 if isinstance(obj
, (list, tuple, LazyList
)):
6504 obj
= enumerate(obj
)
6505 elif isinstance(obj
, dict):
6508 if not traverse_string
:
6512 depth
= max(depth
, _current_depth
)
6513 return [_traverse_obj(v
, path
[i
+ 1:], _current_depth
) for k
, v
in obj
if key(k
)]
6514 elif isinstance(obj
, dict) and not (is_user_input
and key
== ':'):
6515 obj
= (obj
.get(key
) if casesense
or (key
in obj
)
6516 else next((v
for k
, v
in obj
.items() if _lower(k
) == key
), None))
6519 key
= (int_or_none(key
) if ':' not in key
6520 else slice(*map(int_or_none
, key
.split(':'))))
6521 if key
== slice(None):
6522 return _traverse_obj(obj
, (..., *path
[i
+ 1:]), _current_depth
)
6523 if not isinstance(key
, (int, slice)):
6525 if not isinstance(obj
, (list, tuple, LazyList
)):
6526 if not traverse_string
:
6535 if isinstance(expected_type
, type):
6536 type_test
= lambda val
: val
if isinstance(val
, expected_type
) else None
6537 elif expected_type
is not None:
6538 type_test
= expected_type
6540 type_test
= lambda val
: val
6542 for path
in path_list
:
6544 val
= _traverse_obj(obj
, path
)
6547 for _
in range(depth
- 1):
6548 val
= itertools
.chain
.from_iterable(v
for v
in val
if v
is not None)
6549 val
= [v
for v
in map(type_test
, val
) if v
is not None]
6551 return val
if get_all
else val
[0]
6553 val
= type_test(val
)
6559 def traverse_dict(dictn
, keys
, casesense
=True):
6560 ''' For backward compatibility. Do not use '''
6561 return traverse_obj(dictn
, keys
, casesense
=casesense
,
6562 is_user_input
=True, traverse_string
=True)
6565 def variadic(x
, allowed_types
=(str, bytes)):
6566 return x
if isinstance(x
, collections
.abc
.Iterable
) and not isinstance(x
, allowed_types
) else (x
,)
6569 # create a JSON Web Signature (jws) with HS256 algorithm
6570 # the resulting format is in JWS Compact Serialization
6571 # implemented following JWT https://www.rfc-editor.org/rfc/rfc7519.html
6572 # implemented following JWS https://www.rfc-editor.org/rfc/rfc7515.html
6573 def jwt_encode_hs256(payload_data
, key
, headers
={}):
6579 header_data
.update(headers
)
6580 header_b64
= base64
.b64encode(json
.dumps(header_data
).encode('utf-8'))
6581 payload_b64
= base64
.b64encode(json
.dumps(payload_data
).encode('utf-8'))
6582 h
= hmac
.new(key
.encode('utf-8'), header_b64
+ b
'.' + payload_b64
, hashlib
.sha256
)
6583 signature_b64
= base64
.b64encode(h
.digest())
6584 token
= header_b64
+ b
'.' + payload_b64
+ b
'.' + signature_b64
6588 # can be extended in future to verify the signature and parse header and return the algorithm used if it's not HS256
6589 def jwt_decode_hs256(jwt
):
6590 header_b64
, payload_b64
, signature_b64
= jwt
.split('.')
6591 payload_data
= json
.loads(base64
.urlsafe_b64decode(payload_b64
))
6595 def supports_terminal_sequences(stream
):
6596 if compat_os_name
== 'nt':
6597 if get_windows_version() < (10, 0, 10586):
6599 elif not os
.getenv('TERM'):
6602 return stream
.isatty()
6603 except BaseException
:
6607 _terminal_sequences_re
= re
.compile('\033\\[[^m]+m')
6610 def remove_terminal_sequences(string
):
6611 return _terminal_sequences_re
.sub('', string
)
6614 def number_of_digits(number
):
6615 return len('%d' % number
)
6618 def join_nonempty(*values
, delim
='-', from_dict
=None):
6619 if from_dict
is not None:
6620 values
= map(from_dict
.get
, values
)
6621 return delim
.join(map(str, filter(None, values
)))
projects / yt-dlp.git / blob