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.
2579 class DownloadCancelled(YoutubeDLError
):
2580 """ Exception raised when the download queue should be interrupted """
2581 msg
= 'The download was cancelled'
2584 class ExistingVideoReached(DownloadCancelled
):
2585 """ --break-on-existing triggered """
2586 msg
= 'Encountered a video that is already in the archive, stopping due to --break-on-existing'
2589 class RejectedVideoReached(DownloadCancelled
):
2590 """ --break-on-reject triggered """
2591 msg
= 'Encountered a video that did not match filter, stopping due to --break-on-reject'
2594 class MaxDownloadsReached(DownloadCancelled
):
2595 """ --max-downloads limit has been reached. """
2596 msg
= 'Maximum number of downloads reached, stopping due to --max-downloads'
2599 class ReExtractInfo(YoutubeDLError
):
2600 """ Video info needs to be re-extracted. """
2602 def __init__(self
, msg
, expected
=False):
2603 super().__init
__(msg
)
2604 self
.expected
= expected
2607 class ThrottledDownload(ReExtractInfo
):
2608 """ Download speed below --throttled-rate. """
2609 msg
= 'The download speed is below throttle limit'
2612 super().__init
__(self
.msg
, expected
=False)
2615 class UnavailableVideoError(YoutubeDLError
):
2616 """Unavailable Format exception.
2618 This exception will be thrown when a video is requested
2619 in a format that is not available for that video.
2621 msg
= 'Unable to download video'
2623 def __init__(self
, err
=None):
2625 self
.msg
+= f
': {err}'
2626 super().__init
__(self
.msg
)
2629 class ContentTooShortError(YoutubeDLError
):
2630 """Content Too Short exception.
2632 This exception may be raised by FileDownloader objects when a file they
2633 download is too small for what the server announced first, indicating
2634 the connection was probably interrupted.
2637 def __init__(self
, downloaded
, expected
):
2638 super(ContentTooShortError
, self
).__init
__(
2639 'Downloaded {0} bytes, expected {1} bytes'.format(downloaded
, expected
)
2642 self
.downloaded
= downloaded
2643 self
.expected
= expected
2646 class XAttrMetadataError(YoutubeDLError
):
2647 def __init__(self
, code
=None, msg
='Unknown error'):
2648 super(XAttrMetadataError
, self
).__init
__(msg
)
2652 # Parsing code and msg
2653 if (self
.code
in (errno
.ENOSPC
, errno
.EDQUOT
)
2654 or 'No space left' in self
.msg
or 'Disk quota exceeded' in self
.msg
):
2655 self
.reason
= 'NO_SPACE'
2656 elif self
.code
== errno
.E2BIG
or 'Argument list too long' in self
.msg
:
2657 self
.reason
= 'VALUE_TOO_LONG'
2659 self
.reason
= 'NOT_SUPPORTED'
2662 class XAttrUnavailableError(YoutubeDLError
):
2666 def _create_http_connection(ydl_handler
, http_class
, is_https
, *args
, **kwargs
):
2667 # Working around python 2 bug (see http://bugs.python.org/issue17849) by limiting
2668 # expected HTTP responses to meet HTTP/1.0 or later (see also
2669 # https://github.com/ytdl-org/youtube-dl/issues/6727)
2670 if sys
.version_info
< (3, 0):
2671 kwargs
['strict'] = True
2672 hc
= http_class(*args
, **compat_kwargs(kwargs
))
2673 source_address
= ydl_handler
._params
.get('source_address')
2675 if source_address
is not None:
2676 # This is to workaround _create_connection() from socket where it will try all
2677 # address data from getaddrinfo() including IPv6. This filters the result from
2678 # getaddrinfo() based on the source_address value.
2679 # This is based on the cpython socket.create_connection() function.
2680 # https://github.com/python/cpython/blob/master/Lib/socket.py#L691
2681 def _create_connection(address
, timeout
=socket
._GLOBAL
_DEFAULT
_TIMEOUT
, source_address
=None):
2682 host
, port
= address
2684 addrs
= socket
.getaddrinfo(host
, port
, 0, socket
.SOCK_STREAM
)
2685 af
= socket
.AF_INET
if '.' in source_address
[0] else socket
.AF_INET6
2686 ip_addrs
= [addr
for addr
in addrs
if addr
[0] == af
]
2687 if addrs
and not ip_addrs
:
2688 ip_version
= 'v4' if af
== socket
.AF_INET
else 'v6'
2690 "No remote IP%s addresses available for connect, can't use '%s' as source address"
2691 % (ip_version
, source_address
[0]))
2692 for res
in ip_addrs
:
2693 af
, socktype
, proto
, canonname
, sa
= res
2696 sock
= socket
.socket(af
, socktype
, proto
)
2697 if timeout
is not socket
._GLOBAL
_DEFAULT
_TIMEOUT
:
2698 sock
.settimeout(timeout
)
2699 sock
.bind(source_address
)
2701 err
= None # Explicitly break reference cycle
2703 except socket
.error
as _
:
2705 if sock
is not None:
2710 raise socket
.error('getaddrinfo returns an empty list')
2711 if hasattr(hc
, '_create_connection'):
2712 hc
._create
_connection
= _create_connection
2713 sa
= (source_address
, 0)
2714 if hasattr(hc
, 'source_address'): # Python 2.7+
2715 hc
.source_address
= sa
2717 def _hc_connect(self
, *args
, **kwargs
):
2718 sock
= _create_connection(
2719 (self
.host
, self
.port
), self
.timeout
, sa
)
2721 self
.sock
= ssl
.wrap_socket(
2722 sock
, self
.key_file
, self
.cert_file
,
2723 ssl_version
=ssl
.PROTOCOL_TLSv1
)
2726 hc
.connect
= functools
.partial(_hc_connect
, hc
)
2731 def handle_youtubedl_headers(headers
):
2732 filtered_headers
= headers
2734 if 'Youtubedl-no-compression' in filtered_headers
:
2735 filtered_headers
= dict((k
, v
) for k
, v
in filtered_headers
.items() if k
.lower() != 'accept-encoding')
2736 del filtered_headers
['Youtubedl-no-compression']
2738 return filtered_headers
2741 class YoutubeDLHandler(compat_urllib_request
.HTTPHandler
):
2742 """Handler for HTTP requests and responses.
2744 This class, when installed with an OpenerDirector, automatically adds
2745 the standard headers to every HTTP request and handles gzipped and
2746 deflated responses from web servers. If compression is to be avoided in
2747 a particular request, the original request in the program code only has
2748 to include the HTTP header "Youtubedl-no-compression", which will be
2749 removed before making the real request.
2751 Part of this code was copied from:
2753 http://techknack.net/python-urllib2-handlers/
2755 Andrew Rowls, the author of that code, agreed to release it to the
2759 def __init__(self
, params
, *args
, **kwargs
):
2760 compat_urllib_request
.HTTPHandler
.__init
__(self
, *args
, **kwargs
)
2761 self
._params
= params
2763 def http_open(self
, req
):
2764 conn_class
= compat_http_client
.HTTPConnection
2766 socks_proxy
= req
.headers
.get('Ytdl-socks-proxy')
2768 conn_class
= make_socks_conn_class(conn_class
, socks_proxy
)
2769 del req
.headers
['Ytdl-socks-proxy']
2771 return self
.do_open(functools
.partial(
2772 _create_http_connection
, self
, conn_class
, False),
2780 return zlib
.decompress(data
, -zlib
.MAX_WBITS
)
2782 return zlib
.decompress(data
)
2784 def http_request(self
, req
):
2785 # According to RFC 3986, URLs can not contain non-ASCII characters, however this is not
2786 # always respected by websites, some tend to give out URLs with non percent-encoded
2787 # non-ASCII characters (see telemb.py, ard.py [#3412])
2788 # urllib chokes on URLs with non-ASCII characters (see http://bugs.python.org/issue3991)
2789 # To work around aforementioned issue we will replace request's original URL with
2790 # percent-encoded one
2791 # Since redirects are also affected (e.g. http://www.southpark.de/alle-episoden/s18e09)
2792 # the code of this workaround has been moved here from YoutubeDL.urlopen()
2793 url
= req
.get_full_url()
2794 url_escaped
= escape_url(url
)
2796 # Substitute URL if any change after escaping
2797 if url
!= url_escaped
:
2798 req
= update_Request(req
, url
=url_escaped
)
2800 for h
, v
in std_headers
.items():
2801 # Capitalize is needed because of Python bug 2275: http://bugs.python.org/issue2275
2802 # The dict keys are capitalized because of this bug by urllib
2803 if h
.capitalize() not in req
.headers
:
2804 req
.add_header(h
, v
)
2806 req
.headers
= handle_youtubedl_headers(req
.headers
)
2808 if sys
.version_info
< (2, 7) and '#' in req
.get_full_url():
2809 # Python 2.6 is brain-dead when it comes to fragments
2810 req
._Request
__original
= req
._Request
__original
.partition('#')[0]
2811 req
._Request
__r
_type
= req
._Request
__r
_type
.partition('#')[0]
2815 def http_response(self
, req
, resp
):
2818 if resp
.headers
.get('Content-encoding', '') == 'gzip':
2819 content
= resp
.read()
2820 gz
= gzip
.GzipFile(fileobj
=io
.BytesIO(content
), mode
='rb')
2822 uncompressed
= io
.BytesIO(gz
.read())
2823 except IOError as original_ioerror
:
2824 # There may be junk add the end of the file
2825 # See http://stackoverflow.com/q/4928560/35070 for details
2826 for i
in range(1, 1024):
2828 gz
= gzip
.GzipFile(fileobj
=io
.BytesIO(content
[:-i
]), mode
='rb')
2829 uncompressed
= io
.BytesIO(gz
.read())
2834 raise original_ioerror
2835 resp
= compat_urllib_request
.addinfourl(uncompressed
, old_resp
.headers
, old_resp
.url
, old_resp
.code
)
2836 resp
.msg
= old_resp
.msg
2837 del resp
.headers
['Content-encoding']
2839 if resp
.headers
.get('Content-encoding', '') == 'deflate':
2840 gz
= io
.BytesIO(self
.deflate(resp
.read()))
2841 resp
= compat_urllib_request
.addinfourl(gz
, old_resp
.headers
, old_resp
.url
, old_resp
.code
)
2842 resp
.msg
= old_resp
.msg
2843 del resp
.headers
['Content-encoding']
2844 # Percent-encode redirect URL of Location HTTP header to satisfy RFC 3986 (see
2845 # https://github.com/ytdl-org/youtube-dl/issues/6457).
2846 if 300 <= resp
.code
< 400:
2847 location
= resp
.headers
.get('Location')
2849 # As of RFC 2616 default charset is iso-8859-1 that is respected by python 3
2850 if sys
.version_info
>= (3, 0):
2851 location
= location
.encode('iso-8859-1').decode('utf-8')
2853 location
= location
.decode('utf-8')
2854 location_escaped
= escape_url(location
)
2855 if location
!= location_escaped
:
2856 del resp
.headers
['Location']
2857 if sys
.version_info
< (3, 0):
2858 location_escaped
= location_escaped
.encode('utf-8')
2859 resp
.headers
['Location'] = location_escaped
2862 https_request
= http_request
2863 https_response
= http_response
2866 def make_socks_conn_class(base_class
, socks_proxy
):
2867 assert issubclass(base_class
, (
2868 compat_http_client
.HTTPConnection
, compat_http_client
.HTTPSConnection
))
2870 url_components
= compat_urlparse
.urlparse(socks_proxy
)
2871 if url_components
.scheme
.lower() == 'socks5':
2872 socks_type
= ProxyType
.SOCKS5
2873 elif url_components
.scheme
.lower() in ('socks', 'socks4'):
2874 socks_type
= ProxyType
.SOCKS4
2875 elif url_components
.scheme
.lower() == 'socks4a':
2876 socks_type
= ProxyType
.SOCKS4A
2878 def unquote_if_non_empty(s
):
2881 return compat_urllib_parse_unquote_plus(s
)
2885 url_components
.hostname
, url_components
.port
or 1080,
2887 unquote_if_non_empty(url_components
.username
),
2888 unquote_if_non_empty(url_components
.password
),
2891 class SocksConnection(base_class
):
2893 self
.sock
= sockssocket()
2894 self
.sock
.setproxy(*proxy_args
)
2895 if type(self
.timeout
) in (int, float):
2896 self
.sock
.settimeout(self
.timeout
)
2897 self
.sock
.connect((self
.host
, self
.port
))
2899 if isinstance(self
, compat_http_client
.HTTPSConnection
):
2900 if hasattr(self
, '_context'): # Python > 2.6
2901 self
.sock
= self
._context
.wrap_socket(
2902 self
.sock
, server_hostname
=self
.host
)
2904 self
.sock
= ssl
.wrap_socket(self
.sock
)
2906 return SocksConnection
2909 class YoutubeDLHTTPSHandler(compat_urllib_request
.HTTPSHandler
):
2910 def __init__(self
, params
, https_conn_class
=None, *args
, **kwargs
):
2911 compat_urllib_request
.HTTPSHandler
.__init
__(self
, *args
, **kwargs
)
2912 self
._https
_conn
_class
= https_conn_class
or compat_http_client
.HTTPSConnection
2913 self
._params
= params
2915 def https_open(self
, req
):
2917 conn_class
= self
._https
_conn
_class
2919 if hasattr(self
, '_context'): # python > 2.6
2920 kwargs
['context'] = self
._context
2921 if hasattr(self
, '_check_hostname'): # python 3.x
2922 kwargs
['check_hostname'] = self
._check
_hostname
2924 socks_proxy
= req
.headers
.get('Ytdl-socks-proxy')
2926 conn_class
= make_socks_conn_class(conn_class
, socks_proxy
)
2927 del req
.headers
['Ytdl-socks-proxy']
2929 return self
.do_open(functools
.partial(
2930 _create_http_connection
, self
, conn_class
, True),
2934 class YoutubeDLCookieJar(compat_cookiejar
.MozillaCookieJar
):
2936 See [1] for cookie file format.
2938 1. https://curl.haxx.se/docs/http-cookies.html
2940 _HTTPONLY_PREFIX
= '#HttpOnly_'
2942 _HEADER
= '''# Netscape HTTP Cookie File
2943 # This file is generated by yt-dlp. Do not edit.
2946 _CookieFileEntry
= collections
.namedtuple(
2948 ('domain_name', 'include_subdomains', 'path', 'https_only', 'expires_at', 'name', 'value'))
2950 def save(self
, filename
=None, ignore_discard
=False, ignore_expires
=False):
2952 Save cookies to a file.
2954 Most of the code is taken from CPython 3.8 and slightly adapted
2955 to support cookie files with UTF-8 in both python 2 and 3.
2957 if filename
is None:
2958 if self
.filename
is not None:
2959 filename
= self
.filename
2961 raise ValueError(compat_cookiejar
.MISSING_FILENAME_TEXT
)
2963 # Store session cookies with `expires` set to 0 instead of an empty
2966 if cookie
.expires
is None:
2969 with io
.open(filename
, 'w', encoding
='utf-8') as f
:
2970 f
.write(self
._HEADER
)
2973 if not ignore_discard
and cookie
.discard
:
2975 if not ignore_expires
and cookie
.is_expired(now
):
2981 if cookie
.domain
.startswith('.'):
2982 initial_dot
= 'TRUE'
2984 initial_dot
= 'FALSE'
2985 if cookie
.expires
is not None:
2986 expires
= compat_str(cookie
.expires
)
2989 if cookie
.value
is None:
2990 # cookies.txt regards 'Set-Cookie: foo' as a cookie
2991 # with no name, whereas http.cookiejar regards it as a
2992 # cookie with no value.
2997 value
= cookie
.value
2999 '\t'.join([cookie
.domain
, initial_dot
, cookie
.path
,
3000 secure
, expires
, name
, value
]) + '\n')
3002 def load(self
, filename
=None, ignore_discard
=False, ignore_expires
=False):
3003 """Load cookies from a file."""
3004 if filename
is None:
3005 if self
.filename
is not None:
3006 filename
= self
.filename
3008 raise ValueError(compat_cookiejar
.MISSING_FILENAME_TEXT
)
3010 def prepare_line(line
):
3011 if line
.startswith(self
._HTTPONLY
_PREFIX
):
3012 line
= line
[len(self
._HTTPONLY
_PREFIX
):]
3013 # comments and empty lines are fine
3014 if line
.startswith('#') or not line
.strip():
3016 cookie_list
= line
.split('\t')
3017 if len(cookie_list
) != self
._ENTRY
_LEN
:
3018 raise compat_cookiejar
.LoadError('invalid length %d' % len(cookie_list
))
3019 cookie
= self
._CookieFileEntry
(*cookie_list
)
3020 if cookie
.expires_at
and not cookie
.expires_at
.isdigit():
3021 raise compat_cookiejar
.LoadError('invalid expires at %s' % cookie
.expires_at
)
3025 with io
.open(filename
, encoding
='utf-8') as f
:
3028 cf
.write(prepare_line(line
))
3029 except compat_cookiejar
.LoadError
as e
:
3031 'WARNING: skipping cookie file entry due to %s: %r\n'
3032 % (e
, line
), sys
.stderr
)
3035 self
._really
_load
(cf
, filename
, ignore_discard
, ignore_expires
)
3036 # Session cookies are denoted by either `expires` field set to
3037 # an empty string or 0. MozillaCookieJar only recognizes the former
3038 # (see [1]). So we need force the latter to be recognized as session
3039 # cookies on our own.
3040 # Session cookies may be important for cookies-based authentication,
3041 # e.g. usually, when user does not check 'Remember me' check box while
3042 # logging in on a site, some important cookies are stored as session
3043 # cookies so that not recognizing them will result in failed login.
3044 # 1. https://bugs.python.org/issue17164
3046 # Treat `expires=0` cookies as session cookies
3047 if cookie
.expires
== 0:
3048 cookie
.expires
= None
3049 cookie
.discard
= True
3052 class YoutubeDLCookieProcessor(compat_urllib_request
.HTTPCookieProcessor
):
3053 def __init__(self
, cookiejar
=None):
3054 compat_urllib_request
.HTTPCookieProcessor
.__init
__(self
, cookiejar
)
3056 def http_response(self
, request
, response
):
3057 # Python 2 will choke on next HTTP request in row if there are non-ASCII
3058 # characters in Set-Cookie HTTP header of last response (see
3059 # https://github.com/ytdl-org/youtube-dl/issues/6769).
3060 # In order to at least prevent crashing we will percent encode Set-Cookie
3061 # header before HTTPCookieProcessor starts processing it.
3062 # if sys.version_info < (3, 0) and response.headers:
3063 # for set_cookie_header in ('Set-Cookie', 'Set-Cookie2'):
3064 # set_cookie = response.headers.get(set_cookie_header)
3066 # set_cookie_escaped = compat_urllib_parse.quote(set_cookie, b"%/;:@&=+$,!~*'()?#[] ")
3067 # if set_cookie != set_cookie_escaped:
3068 # del response.headers[set_cookie_header]
3069 # response.headers[set_cookie_header] = set_cookie_escaped
3070 return compat_urllib_request
.HTTPCookieProcessor
.http_response(self
, request
, response
)
3072 https_request
= compat_urllib_request
.HTTPCookieProcessor
.http_request
3073 https_response
= http_response
3076 class YoutubeDLRedirectHandler(compat_urllib_request
.HTTPRedirectHandler
):
3077 """YoutubeDL redirect handler
3079 The code is based on HTTPRedirectHandler implementation from CPython [1].
3081 This redirect handler solves two issues:
3082 - ensures redirect URL is always unicode under python 2
3083 - introduces support for experimental HTTP response status code
3084 308 Permanent Redirect [2] used by some sites [3]
3086 1. https://github.com/python/cpython/blob/master/Lib/urllib/request.py
3087 2. https://developer.mozilla.org/en-US/docs/Web/HTTP/Status/308
3088 3. https://github.com/ytdl-org/youtube-dl/issues/28768
3091 http_error_301
= http_error_303
= http_error_307
= http_error_308
= compat_urllib_request
.HTTPRedirectHandler
.http_error_302
3093 def redirect_request(self
, req
, fp
, code
, msg
, headers
, newurl
):
3094 """Return a Request or None in response to a redirect.
3096 This is called by the http_error_30x methods when a
3097 redirection response is received. If a redirection should
3098 take place, return a new Request to allow http_error_30x to
3099 perform the redirect. Otherwise, raise HTTPError if no-one
3100 else should try to handle this url. Return None if you can't
3101 but another Handler might.
3103 m
= req
.get_method()
3104 if (not (code
in (301, 302, 303, 307, 308) and m
in ("GET", "HEAD")
3105 or code
in (301, 302, 303) and m
== "POST")):
3106 raise compat_HTTPError(req
.full_url
, code
, msg
, headers
, fp
)
3107 # Strictly (according to RFC 2616), 301 or 302 in response to
3108 # a POST MUST NOT cause a redirection without confirmation
3109 # from the user (of urllib.request, in this case). In practice,
3110 # essentially all clients do redirect in this case, so we do
3113 # On python 2 urlh.geturl() may sometimes return redirect URL
3114 # as byte string instead of unicode. This workaround allows
3115 # to force it always return unicode.
3116 if sys
.version_info
[0] < 3:
3117 newurl
= compat_str(newurl
)
3119 # Be conciliant with URIs containing a space. This is mainly
3120 # redundant with the more complete encoding done in http_error_302(),
3121 # but it is kept for compatibility with other callers.
3122 newurl
= newurl
.replace(' ', '%20')
3124 CONTENT_HEADERS
= ("content-length", "content-type")
3125 # NB: don't use dict comprehension for python 2.6 compatibility
3126 newheaders
= dict((k
, v
) for k
, v
in req
.headers
.items()
3127 if k
.lower() not in CONTENT_HEADERS
)
3128 return compat_urllib_request
.Request(
3129 newurl
, headers
=newheaders
, origin_req_host
=req
.origin_req_host
,
3133 def extract_timezone(date_str
):
3136 ^.{8,}? # >=8 char non-TZ prefix, if present
3137 (?P<tz>Z| # just the UTC Z, or
3138 (?:(?<=.\b\d{4}|\b\d{2}:\d\d)| # preceded by 4 digits or hh:mm or
3139 (?<!.\b[a-zA-Z]{3}|[a-zA-Z]{4}|..\b\d\d)) # not preceded by 3 alpha word or >= 4 alpha or 2 digits
3140 [ ]? # optional space
3141 (?P<sign>\+|-) # +/-
3142 (?P<hours>[0-9]{2}):?(?P<minutes>[0-9]{2}) # hh[:]mm
3146 timezone
= datetime
.timedelta()
3148 date_str
= date_str
[:-len(m
.group('tz'))]
3149 if not m
.group('sign'):
3150 timezone
= datetime
.timedelta()
3152 sign
= 1 if m
.group('sign') == '+' else -1
3153 timezone
= datetime
.timedelta(
3154 hours
=sign
* int(m
.group('hours')),
3155 minutes
=sign
* int(m
.group('minutes')))
3156 return timezone
, date_str
3159 def parse_iso8601(date_str
, delimiter
='T', timezone
=None):
3160 """ Return a UNIX timestamp from the given date """
3162 if date_str
is None:
3165 date_str
= re
.sub(r
'\.[0-9]+', '', date_str
)
3167 if timezone
is None:
3168 timezone
, date_str
= extract_timezone(date_str
)
3171 date_format
= '%Y-%m-%d{0}%H:%M:%S'.format(delimiter
)
3172 dt
= datetime
.datetime
.strptime(date_str
, date_format
) - timezone
3173 return calendar
.timegm(dt
.timetuple())
3178 def date_formats(day_first
=True):
3179 return DATE_FORMATS_DAY_FIRST
if day_first
else DATE_FORMATS_MONTH_FIRST
3182 def unified_strdate(date_str
, day_first
=True):
3183 """Return a string with the date in the format YYYYMMDD"""
3185 if date_str
is None:
3189 date_str
= date_str
.replace(',', ' ')
3190 # Remove AM/PM + timezone
3191 date_str
= re
.sub(r
'(?i)\s*(?:AM|PM)(?:\s+[A-Z]+)?', '', date_str
)
3192 _
, date_str
= extract_timezone(date_str
)
3194 for expression
in date_formats(day_first
):
3196 upload_date
= datetime
.datetime
.strptime(date_str
, expression
).strftime('%Y%m%d')
3199 if upload_date
is None:
3200 timetuple
= email
.utils
.parsedate_tz(date_str
)
3203 upload_date
= datetime
.datetime(*timetuple
[:6]).strftime('%Y%m%d')
3206 if upload_date
is not None:
3207 return compat_str(upload_date
)
3210 def unified_timestamp(date_str
, day_first
=True):
3211 if date_str
is None:
3214 date_str
= re
.sub(r
'[,|]', '', date_str
)
3216 pm_delta
= 12 if re
.search(r
'(?i)PM', date_str
) else 0
3217 timezone
, date_str
= extract_timezone(date_str
)
3219 # Remove AM/PM + timezone
3220 date_str
= re
.sub(r
'(?i)\s*(?:AM|PM)(?:\s+[A-Z]+)?', '', date_str
)
3222 # Remove unrecognized timezones from ISO 8601 alike timestamps
3223 m
= re
.search(r
'\d{1,2}:\d{1,2}(?:\.\d+)?(?P<tz>\s*[A-Z]+)$', date_str
)
3225 date_str
= date_str
[:-len(m
.group('tz'))]
3227 # Python only supports microseconds, so remove nanoseconds
3228 m
= re
.search(r
'^([0-9]{4,}-[0-9]{1,2}-[0-9]{1,2}T[0-9]{1,2}:[0-9]{1,2}:[0-9]{1,2}\.[0-9]{6})[0-9]+$', date_str
)
3230 date_str
= m
.group(1)
3232 for expression
in date_formats(day_first
):
3234 dt
= datetime
.datetime
.strptime(date_str
, expression
) - timezone
+ datetime
.timedelta(hours
=pm_delta
)
3235 return calendar
.timegm(dt
.timetuple())
3238 timetuple
= email
.utils
.parsedate_tz(date_str
)
3240 return calendar
.timegm(timetuple
) + pm_delta
* 3600
3243 def determine_ext(url
, default_ext
='unknown_video'):
3244 if url
is None or '.' not in url
:
3246 guess
= url
.partition('?')[0].rpartition('.')[2]
3247 if re
.match(r
'^[A-Za-z0-9]+$', guess
):
3249 # Try extract ext from URLs like http://example.com/foo/bar.mp4/?download
3250 elif guess
.rstrip('/') in KNOWN_EXTENSIONS
:
3251 return guess
.rstrip('/')
3256 def subtitles_filename(filename
, sub_lang
, sub_format
, expected_real_ext
=None):
3257 return replace_extension(filename
, sub_lang
+ '.' + sub_format
, expected_real_ext
)
3260 def datetime_from_str(date_str
, precision
='auto', format
='%Y%m%d'):
3262 Return a datetime object from a string in the format YYYYMMDD or
3263 (now|today|date)[+-][0-9](microsecond|second|minute|hour|day|week|month|year)(s)?
3265 format: string date format used to return datetime object from
3266 precision: round the time portion of a datetime object.
3267 auto|microsecond|second|minute|hour|day.
3268 auto: round to the unit provided in date_str (if applicable).
3270 auto_precision
= False
3271 if precision
== 'auto':
3272 auto_precision
= True
3273 precision
= 'microsecond'
3274 today
= datetime_round(datetime
.datetime
.now(), precision
)
3275 if date_str
in ('now', 'today'):
3277 if date_str
== 'yesterday':
3278 return today
- datetime
.timedelta(days
=1)
3280 r
'(?P<start>.+)(?P<sign>[+-])(?P<time>\d+)(?P<unit>microsecond|second|minute|hour|day|week|month|year)(s)?',
3282 if match
is not None:
3283 start_time
= datetime_from_str(match
.group('start'), precision
, format
)
3284 time
= int(match
.group('time')) * (-1 if match
.group('sign') == '-' else 1)
3285 unit
= match
.group('unit')
3286 if unit
== 'month' or unit
== 'year':
3287 new_date
= datetime_add_months(start_time
, time
* 12 if unit
== 'year' else time
)
3293 delta
= datetime
.timedelta(**{unit + 's': time}
)
3294 new_date
= start_time
+ delta
3296 return datetime_round(new_date
, unit
)
3299 return datetime_round(datetime
.datetime
.strptime(date_str
, format
), precision
)
3302 def date_from_str(date_str
, format
='%Y%m%d'):
3304 Return a datetime object from a string in the format YYYYMMDD or
3305 (now|today|date)[+-][0-9](microsecond|second|minute|hour|day|week|month|year)(s)?
3307 format: string date format used to return datetime object from
3309 return datetime_from_str(date_str
, precision
='microsecond', format
=format
).date()
3312 def datetime_add_months(dt
, months
):
3313 """Increment/Decrement a datetime object by months."""
3314 month
= dt
.month
+ months
- 1
3315 year
= dt
.year
+ month
// 12
3316 month
= month
% 12 + 1
3317 day
= min(dt
.day
, calendar
.monthrange(year
, month
)[1])
3318 return dt
.replace(year
, month
, day
)
3321 def datetime_round(dt
, precision
='day'):
3323 Round a datetime object's time to a specific precision
3325 if precision
== 'microsecond':
3334 roundto
= lambda x
, n
: ((x
+ n
/ 2) // n
) * n
3335 timestamp
= calendar
.timegm(dt
.timetuple())
3336 return datetime
.datetime
.utcfromtimestamp(roundto(timestamp
, unit_seconds
[precision
]))
3339 def hyphenate_date(date_str
):
3341 Convert a date in 'YYYYMMDD' format to 'YYYY-MM-DD' format"""
3342 match
= re
.match(r
'^(\d\d\d\d)(\d\d)(\d\d)$', date_str
)
3343 if match
is not None:
3344 return '-'.join(match
.groups())
3349 class DateRange(object):
3350 """Represents a time interval between two dates"""
3352 def __init__(self
, start
=None, end
=None):
3353 """start and end must be strings in the format accepted by date"""
3354 if start
is not None:
3355 self
.start
= date_from_str(start
)
3357 self
.start
= datetime
.datetime
.min.date()
3359 self
.end
= date_from_str(end
)
3361 self
.end
= datetime
.datetime
.max.date()
3362 if self
.start
> self
.end
:
3363 raise ValueError('Date range: "%s" , the start date must be before the end date' % self
)
3367 """Returns a range that only contains the given day"""
3368 return cls(day
, day
)
3370 def __contains__(self
, date
):
3371 """Check if the date is in the range"""
3372 if not isinstance(date
, datetime
.date
):
3373 date
= date_from_str(date
)
3374 return self
.start
<= date
<= self
.end
3377 return '%s - %s' % (self
.start
.isoformat(), self
.end
.isoformat())
3380 def platform_name():
3381 """ Returns the platform name as a compat_str """
3382 res
= platform
.platform()
3383 if isinstance(res
, bytes):
3384 res
= res
.decode(preferredencoding())
3386 assert isinstance(res
, compat_str
)
3390 def get_windows_version():
3391 ''' Get Windows version. None if it's not running on Windows '''
3392 if compat_os_name
== 'nt':
3393 return version_tuple(platform
.win32_ver()[1])
3398 def _windows_write_string(s
, out
):
3399 """ Returns True if the string was written using special methods,
3400 False if it has yet to be written out."""
3401 # Adapted from http://stackoverflow.com/a/3259271/35070
3404 import ctypes
.wintypes
3412 fileno
= out
.fileno()
3413 except AttributeError:
3414 # If the output stream doesn't have a fileno, it's virtual
3416 except io
.UnsupportedOperation
:
3417 # Some strange Windows pseudo files?
3419 if fileno
not in WIN_OUTPUT_IDS
:
3422 GetStdHandle
= compat_ctypes_WINFUNCTYPE(
3423 ctypes
.wintypes
.HANDLE
, ctypes
.wintypes
.DWORD
)(
3424 ('GetStdHandle', ctypes
.windll
.kernel32
))
3425 h
= GetStdHandle(WIN_OUTPUT_IDS
[fileno
])
3427 WriteConsoleW
= compat_ctypes_WINFUNCTYPE(
3428 ctypes
.wintypes
.BOOL
, ctypes
.wintypes
.HANDLE
, ctypes
.wintypes
.LPWSTR
,
3429 ctypes
.wintypes
.DWORD
, ctypes
.POINTER(ctypes
.wintypes
.DWORD
),
3430 ctypes
.wintypes
.LPVOID
)(('WriteConsoleW', ctypes
.windll
.kernel32
))
3431 written
= ctypes
.wintypes
.DWORD(0)
3433 GetFileType
= compat_ctypes_WINFUNCTYPE(ctypes
.wintypes
.DWORD
, ctypes
.wintypes
.DWORD
)(('GetFileType', ctypes
.windll
.kernel32
))
3434 FILE_TYPE_CHAR
= 0x0002
3435 FILE_TYPE_REMOTE
= 0x8000
3436 GetConsoleMode
= compat_ctypes_WINFUNCTYPE(
3437 ctypes
.wintypes
.BOOL
, ctypes
.wintypes
.HANDLE
,
3438 ctypes
.POINTER(ctypes
.wintypes
.DWORD
))(
3439 ('GetConsoleMode', ctypes
.windll
.kernel32
))
3440 INVALID_HANDLE_VALUE
= ctypes
.wintypes
.DWORD(-1).value
3442 def not_a_console(handle
):
3443 if handle
== INVALID_HANDLE_VALUE
or handle
is None:
3445 return ((GetFileType(handle
) & ~FILE_TYPE_REMOTE
) != FILE_TYPE_CHAR
3446 or GetConsoleMode(handle
, ctypes
.byref(ctypes
.wintypes
.DWORD())) == 0)
3448 if not_a_console(h
):
3451 def next_nonbmp_pos(s
):
3453 return next(i
for i
, c
in enumerate(s
) if ord(c
) > 0xffff)
3454 except StopIteration:
3458 count
= min(next_nonbmp_pos(s
), 1024)
3460 ret
= WriteConsoleW(
3461 h
, s
, count
if count
else 2, ctypes
.byref(written
), None)
3463 raise OSError('Failed to write string')
3464 if not count
: # We just wrote a non-BMP character
3465 assert written
.value
== 2
3468 assert written
.value
> 0
3469 s
= s
[written
.value
:]
3473 def write_string(s
, out
=None, encoding
=None):
3476 assert type(s
) == compat_str
3478 if sys
.platform
== 'win32' and encoding
is None and hasattr(out
, 'fileno'):
3479 if _windows_write_string(s
, out
):
3482 if ('b' in getattr(out
, 'mode', '')
3483 or sys
.version_info
[0] < 3): # Python 2 lies about mode of sys.stderr
3484 byt
= s
.encode(encoding
or preferredencoding(), 'ignore')
3486 elif hasattr(out
, 'buffer'):
3487 enc
= encoding
or getattr(out
, 'encoding', None) or preferredencoding()
3488 byt
= s
.encode(enc
, 'ignore')
3489 out
.buffer.write(byt
)
3495 def bytes_to_intlist(bs
):
3498 if isinstance(bs
[0], int): # Python 3
3501 return [ord(c
) for c
in bs
]
3504 def intlist_to_bytes(xs
):
3507 return compat_struct_pack('%dB' % len(xs
), *xs
)
3510 # Cross-platform file locking
3511 if sys
.platform
== 'win32':
3512 import ctypes
.wintypes
3515 class OVERLAPPED(ctypes
.Structure
):
3517 ('Internal', ctypes
.wintypes
.LPVOID
),
3518 ('InternalHigh', ctypes
.wintypes
.LPVOID
),
3519 ('Offset', ctypes
.wintypes
.DWORD
),
3520 ('OffsetHigh', ctypes
.wintypes
.DWORD
),
3521 ('hEvent', ctypes
.wintypes
.HANDLE
),
3524 kernel32
= ctypes
.windll
.kernel32
3525 LockFileEx
= kernel32
.LockFileEx
3526 LockFileEx
.argtypes
= [
3527 ctypes
.wintypes
.HANDLE
, # hFile
3528 ctypes
.wintypes
.DWORD
, # dwFlags
3529 ctypes
.wintypes
.DWORD
, # dwReserved
3530 ctypes
.wintypes
.DWORD
, # nNumberOfBytesToLockLow
3531 ctypes
.wintypes
.DWORD
, # nNumberOfBytesToLockHigh
3532 ctypes
.POINTER(OVERLAPPED
) # Overlapped
3534 LockFileEx
.restype
= ctypes
.wintypes
.BOOL
3535 UnlockFileEx
= kernel32
.UnlockFileEx
3536 UnlockFileEx
.argtypes
= [
3537 ctypes
.wintypes
.HANDLE
, # hFile
3538 ctypes
.wintypes
.DWORD
, # dwReserved
3539 ctypes
.wintypes
.DWORD
, # nNumberOfBytesToLockLow
3540 ctypes
.wintypes
.DWORD
, # nNumberOfBytesToLockHigh
3541 ctypes
.POINTER(OVERLAPPED
) # Overlapped
3543 UnlockFileEx
.restype
= ctypes
.wintypes
.BOOL
3544 whole_low
= 0xffffffff
3545 whole_high
= 0x7fffffff
3547 def _lock_file(f
, exclusive
):
3548 overlapped
= OVERLAPPED()
3549 overlapped
.Offset
= 0
3550 overlapped
.OffsetHigh
= 0
3551 overlapped
.hEvent
= 0
3552 f
._lock
_file
_overlapped
_p
= ctypes
.pointer(overlapped
)
3553 handle
= msvcrt
.get_osfhandle(f
.fileno())
3554 if not LockFileEx(handle
, 0x2 if exclusive
else 0x0, 0,
3555 whole_low
, whole_high
, f
._lock
_file
_overlapped
_p
):
3556 raise OSError('Locking file failed: %r' % ctypes
.FormatError())
3558 def _unlock_file(f
):
3559 assert f
._lock
_file
_overlapped
_p
3560 handle
= msvcrt
.get_osfhandle(f
.fileno())
3561 if not UnlockFileEx(handle
, 0,
3562 whole_low
, whole_high
, f
._lock
_file
_overlapped
_p
):
3563 raise OSError('Unlocking file failed: %r' % ctypes
.FormatError())
3566 # Some platforms, such as Jython, is missing fcntl
3570 def _lock_file(f
, exclusive
):
3571 fcntl
.flock(f
, fcntl
.LOCK_EX
if exclusive
else fcntl
.LOCK_SH
)
3573 def _unlock_file(f
):
3574 fcntl
.flock(f
, fcntl
.LOCK_UN
)
3576 UNSUPPORTED_MSG
= 'file locking is not supported on this platform'
3578 def _lock_file(f
, exclusive
):
3579 raise IOError(UNSUPPORTED_MSG
)
3581 def _unlock_file(f
):
3582 raise IOError(UNSUPPORTED_MSG
)
3585 class locked_file(object):
3586 def __init__(self
, filename
, mode
, encoding
=None):
3587 assert mode
in ['r', 'a', 'w']
3588 self
.f
= io
.open(filename
, mode
, encoding
=encoding
)
3591 def __enter__(self
):
3592 exclusive
= self
.mode
!= 'r'
3594 _lock_file(self
.f
, exclusive
)
3600 def __exit__(self
, etype
, value
, traceback
):
3602 _unlock_file(self
.f
)
3609 def write(self
, *args
):
3610 return self
.f
.write(*args
)
3612 def read(self
, *args
):
3613 return self
.f
.read(*args
)
3616 def get_filesystem_encoding():
3617 encoding
= sys
.getfilesystemencoding()
3618 return encoding
if encoding
is not None else 'utf-8'
3621 def shell_quote(args
):
3623 encoding
= get_filesystem_encoding()
3625 if isinstance(a
, bytes):
3626 # We may get a filename encoded with 'encodeFilename'
3627 a
= a
.decode(encoding
)
3628 quoted_args
.append(compat_shlex_quote(a
))
3629 return ' '.join(quoted_args
)
3632 def smuggle_url(url
, data
):
3633 """ Pass additional data in a URL for internal use. """
3635 url
, idata
= unsmuggle_url(url
, {})
3637 sdata
= compat_urllib_parse_urlencode(
3638 {'__youtubedl_smuggle': json.dumps(data)}
)
3639 return url
+ '#' + sdata
3642 def unsmuggle_url(smug_url
, default
=None):
3643 if '#__youtubedl_smuggle' not in smug_url
:
3644 return smug_url
, default
3645 url
, _
, sdata
= smug_url
.rpartition('#')
3646 jsond
= compat_parse_qs(sdata
)['__youtubedl_smuggle'][0]
3647 data
= json
.loads(jsond
)
3651 def format_bytes(bytes):
3654 if type(bytes) is str:
3655 bytes = float(bytes)
3659 exponent
= int(math
.log(bytes, 1024.0))
3660 suffix
= ['B', 'KiB', 'MiB', 'GiB', 'TiB', 'PiB', 'EiB', 'ZiB', 'YiB'][exponent
]
3661 converted
= float(bytes) / float(1024 ** exponent
)
3662 return '%.2f%s' % (converted
, suffix
)
3665 def lookup_unit_table(unit_table
, s
):
3666 units_re
= '|'.join(re
.escape(u
) for u
in unit_table
)
3668 r
'(?P<num>[0-9]+(?:[,.][0-9]*)?)\s*(?P<unit>%s)\b' % units_re
, s
)
3671 num_str
= m
.group('num').replace(',', '.')
3672 mult
= unit_table
[m
.group('unit')]
3673 return int(float(num_str
) * mult
)
3676 def parse_filesize(s
):
3680 # The lower-case forms are of course incorrect and unofficial,
3681 # but we support those too
3698 'megabytes': 1000 ** 2,
3699 'mebibytes': 1024 ** 2,
3705 'gigabytes': 1000 ** 3,
3706 'gibibytes': 1024 ** 3,
3712 'terabytes': 1000 ** 4,
3713 'tebibytes': 1024 ** 4,
3719 'petabytes': 1000 ** 5,
3720 'pebibytes': 1024 ** 5,
3726 'exabytes': 1000 ** 6,
3727 'exbibytes': 1024 ** 6,
3733 'zettabytes': 1000 ** 7,
3734 'zebibytes': 1024 ** 7,
3740 'yottabytes': 1000 ** 8,
3741 'yobibytes': 1024 ** 8,
3744 return lookup_unit_table(_UNIT_TABLE
, s
)
3753 if re
.match(r
'^[\d,.]+$', s
):
3754 return str_to_int(s
)
3765 return lookup_unit_table(_UNIT_TABLE
, s
)
3768 def parse_resolution(s
):
3772 mobj
= re
.search(r
'(?<![a-zA-Z0-9])(?P<w>\d+)\s*[xX×,]\s*(?P<h>\d+)(?![a-zA-Z0-9])', s
)
3775 'width': int(mobj
.group('w')),
3776 'height': int(mobj
.group('h')),
3779 mobj
= re
.search(r
'(?<![a-zA-Z0-9])(\d+)[pPiI](?![a-zA-Z0-9])', s
)
3781 return {'height': int(mobj.group(1))}
3783 mobj
= re
.search(r
'\b([48])[kK]\b', s
)
3785 return {'height': int(mobj.group(1)) * 540}
3790 def parse_bitrate(s
):
3791 if not isinstance(s
, compat_str
):
3793 mobj
= re
.search(r
'\b(\d+)\s*kbps', s
)
3795 return int(mobj
.group(1))
3798 def month_by_name(name
, lang
='en'):
3799 """ Return the number of a month by (locale-independently) English name """
3801 month_names
= MONTH_NAMES
.get(lang
, MONTH_NAMES
['en'])
3804 return month_names
.index(name
) + 1
3809 def month_by_abbreviation(abbrev
):
3810 """ Return the number of a month by (locale-independently) English
3814 return [s
[:3] for s
in ENGLISH_MONTH_NAMES
].index(abbrev
) + 1
3819 def fix_xml_ampersands(xml_str
):
3820 """Replace all the '&' by '&' in XML"""
3822 r
'&(?!amp;|lt;|gt;|apos;|quot;|#x[0-9a-fA-F]{,4};|#[0-9]{,4};)',
3827 def setproctitle(title
):
3828 assert isinstance(title
, compat_str
)
3830 # ctypes in Jython is not complete
3831 # http://bugs.jython.org/issue2148
3832 if sys
.platform
.startswith('java'):
3836 libc
= ctypes
.cdll
.LoadLibrary('libc.so.6')
3840 # LoadLibrary in Windows Python 2.7.13 only expects
3841 # a bytestring, but since unicode_literals turns
3842 # every string into a unicode string, it fails.
3844 title_bytes
= title
.encode('utf-8')
3845 buf
= ctypes
.create_string_buffer(len(title_bytes
))
3846 buf
.value
= title_bytes
3848 libc
.prctl(15, buf
, 0, 0, 0)
3849 except AttributeError:
3850 return # Strange libc, just skip this
3853 def remove_start(s
, start
):
3854 return s
[len(start
):] if s
is not None and s
.startswith(start
) else s
3857 def remove_end(s
, end
):
3858 return s
[:-len(end
)] if s
is not None and s
.endswith(end
) else s
3861 def remove_quotes(s
):
3862 if s
is None or len(s
) < 2:
3864 for quote
in ('"', "'", ):
3865 if s
[0] == quote
and s
[-1] == quote
:
3870 def get_domain(url
):
3871 domain
= re
.match(r
'(?:https?:\/\/)?(?:www\.)?(?P<domain>[^\n\/]+\.[^\n\/]+)(?:\/(.*))?', url
)
3872 return domain
.group('domain') if domain
else None
3875 def url_basename(url
):
3876 path
= compat_urlparse
.urlparse(url
).path
3877 return path
.strip('/').split('/')[-1]
3881 return re
.match(r
'https?://[^?#&]+/', url
).group()
3884 def urljoin(base
, path
):
3885 if isinstance(path
, bytes):
3886 path
= path
.decode('utf-8')
3887 if not isinstance(path
, compat_str
) or not path
:
3889 if re
.match(r
'^(?:[a-zA-Z][a-zA-Z0-9+-.]*:)?//', path
):
3891 if isinstance(base
, bytes):
3892 base
= base
.decode('utf-8')
3893 if not isinstance(base
, compat_str
) or not re
.match(
3894 r
'^(?:https?:)?//', base
):
3896 return compat_urlparse
.urljoin(base
, path
)
3899 class HEADRequest(compat_urllib_request
.Request
):
3900 def get_method(self
):
3904 class PUTRequest(compat_urllib_request
.Request
):
3905 def get_method(self
):
3909 def int_or_none(v
, scale
=1, default
=None, get_attr
=None, invscale
=1):
3912 v
= getattr(v
, get_attr
, None)
3918 return int(v
) * invscale
// scale
3919 except (ValueError, TypeError, OverflowError):
3923 def str_or_none(v
, default
=None):
3924 return default
if v
is None else compat_str(v
)
3927 def str_to_int(int_str
):
3928 """ A more relaxed version of int_or_none """
3929 if isinstance(int_str
, compat_integer_types
):
3931 elif isinstance(int_str
, compat_str
):
3932 int_str
= re
.sub(r
'[,\.\+]', '', int_str
)
3933 return int_or_none(int_str
)
3936 def float_or_none(v
, scale
=1, invscale
=1, default
=None):
3940 return float(v
) * invscale
/ scale
3941 except (ValueError, TypeError):
3945 def bool_or_none(v
, default
=None):
3946 return v
if isinstance(v
, bool) else default
3949 def strip_or_none(v
, default
=None):
3950 return v
.strip() if isinstance(v
, compat_str
) else default
3953 def url_or_none(url
):
3954 if not url
or not isinstance(url
, compat_str
):
3957 return url
if re
.match(r
'^(?:(?:https?|rt(?:m(?:pt?[es]?|fp)|sp[su]?)|mms|ftps?):)?//', url
) else None
3960 def strftime_or_none(timestamp
, date_format
, default
=None):
3961 datetime_object
= None
3963 if isinstance(timestamp
, compat_numeric_types
): # unix timestamp
3964 datetime_object
= datetime
.datetime
.utcfromtimestamp(timestamp
)
3965 elif isinstance(timestamp
, compat_str
): # assume YYYYMMDD
3966 datetime_object
= datetime
.datetime
.strptime(timestamp
, '%Y%m%d')
3967 return datetime_object
.strftime(date_format
)
3968 except (ValueError, TypeError, AttributeError):
3972 def parse_duration(s
):
3973 if not isinstance(s
, compat_basestring
):
3979 days
, hours
, mins
, secs
, ms
= [None] * 5
3980 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
)
3982 days
, hours
, mins
, secs
, ms
= m
.groups()
3987 [0-9]+\s*y(?:ears?)?\s*
3990 [0-9]+\s*m(?:onths?)?\s*
3993 [0-9]+\s*w(?:eeks?)?\s*
3996 (?P<days>[0-9]+)\s*d(?:ays?)?\s*
4000 (?P<hours>[0-9]+)\s*h(?:ours?)?\s*
4003 (?P<mins>[0-9]+)\s*m(?:in(?:ute)?s?)?\s*
4006 (?P<secs>[0-9]+)(?P<ms>\.[0-9]+)?\s*s(?:ec(?:ond)?s?)?\s*
4009 days
, hours
, mins
, secs
, ms
= m
.groups()
4011 m
= re
.match(r
'(?i)(?:(?P<hours>[0-9.]+)\s*(?:hours?)|(?P<mins>[0-9.]+)\s*(?:mins?\.?|minutes?)\s*)Z?$', s
)
4013 hours
, mins
= m
.groups()
4019 duration
+= float(secs
)
4021 duration
+= float(mins
) * 60
4023 duration
+= float(hours
) * 60 * 60
4025 duration
+= float(days
) * 24 * 60 * 60
4027 duration
+= float(ms
)
4031 def prepend_extension(filename
, ext
, expected_real_ext
=None):
4032 name
, real_ext
= os
.path
.splitext(filename
)
4034 '{0}.{1}{2}'.format(name
, ext
, real_ext
)
4035 if not expected_real_ext
or real_ext
[1:] == expected_real_ext
4036 else '{0}.{1}'.format(filename
, ext
))
4039 def replace_extension(filename
, ext
, expected_real_ext
=None):
4040 name
, real_ext
= os
.path
.splitext(filename
)
4041 return '{0}.{1}'.format(
4042 name
if not expected_real_ext
or real_ext
[1:] == expected_real_ext
else filename
,
4046 def check_executable(exe
, args
=[]):
4047 """ Checks if the given binary is installed somewhere in PATH, and returns its name.
4048 args can be a list of arguments for a short output (like -version) """
4050 Popen([exe
] + args
, stdout
=subprocess
.PIPE
, stderr
=subprocess
.PIPE
).communicate_or_kill()
4056 def _get_exe_version_output(exe
, args
):
4058 # STDIN should be redirected too. On UNIX-like systems, ffmpeg triggers
4059 # SIGTTOU if yt-dlp is run in the background.
4060 # See https://github.com/ytdl-org/youtube-dl/issues/955#issuecomment-209789656
4062 [encodeArgument(exe
)] + args
, stdin
=subprocess
.PIPE
,
4063 stdout
=subprocess
.PIPE
, stderr
=subprocess
.STDOUT
).communicate_or_kill()
4066 if isinstance(out
, bytes): # Python 2.x
4067 out
= out
.decode('ascii', 'ignore')
4071 def detect_exe_version(output
, version_re
=None, unrecognized
='present'):
4072 assert isinstance(output
, compat_str
)
4073 if version_re
is None:
4074 version_re
= r
'version\s+([-0-9._a-zA-Z]+)'
4075 m
= re
.search(version_re
, output
)
4082 def get_exe_version(exe
, args
=['--version'],
4083 version_re
=None, unrecognized
='present'):
4084 """ Returns the version of the specified executable,
4085 or False if the executable is not present """
4086 out
= _get_exe_version_output(exe
, args
)
4087 return detect_exe_version(out
, version_re
, unrecognized
) if out
else False
4090 class LazyList(collections
.abc
.Sequence
):
4091 ''' Lazy immutable list from an iterable
4092 Note that slices of a LazyList are lists and not LazyList'''
4094 class IndexError(IndexError):
4097 def __init__(self
, iterable
, *, reverse
=False, _cache
=None):
4098 self
.__iterable
= iter(iterable
)
4099 self
.__cache
= [] if _cache
is None else _cache
4100 self
.__reversed
= reverse
4104 # We need to consume the entire iterable to iterate in reverse
4105 yield from self
.exhaust()
4107 yield from self
.__cache
4108 for item
in self
.__iterable
:
4109 self
.__cache
.append(item
)
4112 def __exhaust(self
):
4113 self
.__cache
.extend(self
.__iterable
)
4114 # Discard the emptied iterable to make it pickle-able
4115 self
.__iterable
= []
4119 ''' Evaluate the entire iterable '''
4120 return self
.__exhaust
()[::-1 if self
.__reversed
else 1]
4123 def __reverse_index(x
):
4124 return None if x
is None else -(x
+ 1)
4126 def __getitem__(self
, idx
):
4127 if isinstance(idx
, slice):
4129 idx
= slice(self
.__reverse
_index
(idx
.start
), self
.__reverse
_index
(idx
.stop
), -(idx
.step
or 1))
4130 start
, stop
, step
= idx
.start
, idx
.stop
, idx
.step
or 1
4131 elif isinstance(idx
, int):
4133 idx
= self
.__reverse
_index
(idx
)
4134 start
, stop
, step
= idx
, idx
, 0
4136 raise TypeError('indices must be integers or slices')
4137 if ((start
or 0) < 0 or (stop
or 0) < 0
4138 or (start
is None and step
< 0)
4139 or (stop
is None and step
> 0)):
4140 # We need to consume the entire iterable to be able to slice from the end
4141 # Obviously, never use this with infinite iterables
4144 return self
.__cache
[idx
]
4145 except IndexError as e
:
4146 raise self
.IndexError(e
) from e
4147 n
= max(start
or 0, stop
or 0) - len(self
.__cache
) + 1
4149 self
.__cache
.extend(itertools
.islice(self
.__iterable
, n
))
4151 return self
.__cache
[idx
]
4152 except IndexError as e
:
4153 raise self
.IndexError(e
) from e
4157 self
[-1] if self
.__reversed
else self
[0]
4158 except self
.IndexError:
4164 return len(self
.__cache
)
4166 def __reversed__(self
):
4167 return type(self
)(self
.__iterable
, reverse
=not self
.__reversed
, _cache
=self
.__cache
)
4170 return type(self
)(self
.__iterable
, reverse
=self
.__reversed
, _cache
=self
.__cache
)
4172 def __deepcopy__(self
, memo
):
4173 # FIXME: This is actually just a shallow copy
4175 memo
[id_
] = self
.__copy
__()
4179 # repr and str should mimic a list. So we exhaust the iterable
4180 return repr(self
.exhaust())
4183 return repr(self
.exhaust())
4188 class IndexError(IndexError):
4192 # This is only useful for tests
4193 return len(self
.getslice())
4195 def __init__(self
, pagefunc
, pagesize
, use_cache
=True):
4196 self
._pagefunc
= pagefunc
4197 self
._pagesize
= pagesize
4198 self
._use
_cache
= use_cache
4201 def getpage(self
, pagenum
):
4202 page_results
= self
._cache
.get(pagenum
)
4203 if page_results
is None:
4204 page_results
= list(self
._pagefunc
(pagenum
))
4206 self
._cache
[pagenum
] = page_results
4209 def getslice(self
, start
=0, end
=None):
4210 return list(self
._getslice
(start
, end
))
4212 def _getslice(self
, start
, end
):
4213 raise NotImplementedError('This method must be implemented by subclasses')
4215 def __getitem__(self
, idx
):
4216 # NOTE: cache must be enabled if this is used
4217 if not isinstance(idx
, int) or idx
< 0:
4218 raise TypeError('indices must be non-negative integers')
4219 entries
= self
.getslice(idx
, idx
+ 1)
4221 raise self
.IndexError()
4225 class OnDemandPagedList(PagedList
):
4226 def _getslice(self
, start
, end
):
4227 for pagenum
in itertools
.count(start
// self
._pagesize
):
4228 firstid
= pagenum
* self
._pagesize
4229 nextfirstid
= pagenum
* self
._pagesize
+ self
._pagesize
4230 if start
>= nextfirstid
:
4234 start
% self
._pagesize
4235 if firstid
<= start
< nextfirstid
4238 ((end
- 1) % self
._pagesize
) + 1
4239 if (end
is not None and firstid
<= end
<= nextfirstid
)
4242 page_results
= self
.getpage(pagenum
)
4243 if startv
!= 0 or endv
is not None:
4244 page_results
= page_results
[startv
:endv
]
4245 yield from page_results
4247 # A little optimization - if current page is not "full", ie. does
4248 # not contain page_size videos then we can assume that this page
4249 # is the last one - there are no more ids on further pages -
4250 # i.e. no need to query again.
4251 if len(page_results
) + startv
< self
._pagesize
:
4254 # If we got the whole page, but the next page is not interesting,
4255 # break out early as well
4256 if end
== nextfirstid
:
4260 class InAdvancePagedList(PagedList
):
4261 def __init__(self
, pagefunc
, pagecount
, pagesize
):
4262 self
._pagecount
= pagecount
4263 PagedList
.__init
__(self
, pagefunc
, pagesize
, True)
4265 def _getslice(self
, start
, end
):
4266 start_page
= start
// self
._pagesize
4268 self
._pagecount
if end
is None else (end
// self
._pagesize
+ 1))
4269 skip_elems
= start
- start_page
* self
._pagesize
4270 only_more
= None if end
is None else end
- start
4271 for pagenum
in range(start_page
, end_page
):
4272 page_results
= self
.getpage(pagenum
)
4274 page_results
= page_results
[skip_elems
:]
4276 if only_more
is not None:
4277 if len(page_results
) < only_more
:
4278 only_more
-= len(page_results
)
4280 yield from page_results
[:only_more
]
4282 yield from page_results
4285 def uppercase_escape(s
):
4286 unicode_escape
= codecs
.getdecoder('unicode_escape')
4288 r
'\\U[0-9a-fA-F]{8}',
4289 lambda m
: unicode_escape(m
.group(0))[0],
4293 def lowercase_escape(s
):
4294 unicode_escape
= codecs
.getdecoder('unicode_escape')
4296 r
'\\u[0-9a-fA-F]{4}',
4297 lambda m
: unicode_escape(m
.group(0))[0],
4301 def escape_rfc3986(s
):
4302 """Escape non-ASCII characters as suggested by RFC 3986"""
4303 if sys
.version_info
< (3, 0) and isinstance(s
, compat_str
):
4304 s
= s
.encode('utf-8')
4305 return compat_urllib_parse
.quote(s
, b
"%/;:@&=+$,!~*'()?#[]")
4308 def escape_url(url
):
4309 """Escape URL as suggested by RFC 3986"""
4310 url_parsed
= compat_urllib_parse_urlparse(url
)
4311 return url_parsed
._replace
(
4312 netloc
=url_parsed
.netloc
.encode('idna').decode('ascii'),
4313 path
=escape_rfc3986(url_parsed
.path
),
4314 params
=escape_rfc3986(url_parsed
.params
),
4315 query
=escape_rfc3986(url_parsed
.query
),
4316 fragment
=escape_rfc3986(url_parsed
.fragment
)
4321 return compat_parse_qs(compat_urllib_parse_urlparse(url
).query
)
4324 def read_batch_urls(batch_fd
):
4326 if not isinstance(url
, compat_str
):
4327 url
= url
.decode('utf-8', 'replace')
4328 BOM_UTF8
= ('\xef\xbb\xbf', '\ufeff')
4329 for bom
in BOM_UTF8
:
4330 if url
.startswith(bom
):
4331 url
= url
[len(bom
):]
4333 if not url
or url
.startswith(('#', ';', ']')):
4335 # "#" cannot be stripped out since it is part of the URI
4336 # However, it can be safely stipped out if follwing a whitespace
4337 return re
.split(r
'\s#', url
, 1)[0].rstrip()
4339 with contextlib
.closing(batch_fd
) as fd
:
4340 return [url
for url
in map(fixup
, fd
) if url
]
4343 def urlencode_postdata(*args
, **kargs
):
4344 return compat_urllib_parse_urlencode(*args
, **kargs
).encode('ascii')
4347 def update_url_query(url
, query
):
4350 parsed_url
= compat_urlparse
.urlparse(url
)
4351 qs
= compat_parse_qs(parsed_url
.query
)
4353 return compat_urlparse
.urlunparse(parsed_url
._replace
(
4354 query
=compat_urllib_parse_urlencode(qs
, True)))
4357 def update_Request(req
, url
=None, data
=None, headers
={}, query={}
):
4358 req_headers
= req
.headers
.copy()
4359 req_headers
.update(headers
)
4360 req_data
= data
or req
.data
4361 req_url
= update_url_query(url
or req
.get_full_url(), query
)
4362 req_get_method
= req
.get_method()
4363 if req_get_method
== 'HEAD':
4364 req_type
= HEADRequest
4365 elif req_get_method
== 'PUT':
4366 req_type
= PUTRequest
4368 req_type
= compat_urllib_request
.Request
4370 req_url
, data
=req_data
, headers
=req_headers
,
4371 origin_req_host
=req
.origin_req_host
, unverifiable
=req
.unverifiable
)
4372 if hasattr(req
, 'timeout'):
4373 new_req
.timeout
= req
.timeout
4377 def _multipart_encode_impl(data
, boundary
):
4378 content_type
= 'multipart/form-data; boundary=%s' % boundary
4381 for k
, v
in data
.items():
4382 out
+= b
'--' + boundary
.encode('ascii') + b
'\r\n'
4383 if isinstance(k
, compat_str
):
4384 k
= k
.encode('utf-8')
4385 if isinstance(v
, compat_str
):
4386 v
= v
.encode('utf-8')
4387 # RFC 2047 requires non-ASCII field names to be encoded, while RFC 7578
4388 # suggests sending UTF-8 directly. Firefox sends UTF-8, too
4389 content
= b
'Content-Disposition: form-data; name="' + k
+ b
'"\r\n\r\n' + v
+ b
'\r\n'
4390 if boundary
.encode('ascii') in content
:
4391 raise ValueError('Boundary overlaps with data')
4394 out
+= b
'--' + boundary
.encode('ascii') + b
'--\r\n'
4396 return out
, content_type
4399 def multipart_encode(data
, boundary
=None):
4401 Encode a dict to RFC 7578-compliant form-data
4404 A dict where keys and values can be either Unicode or bytes-like
4407 If specified a Unicode object, it's used as the boundary. Otherwise
4408 a random boundary is generated.
4410 Reference: https://tools.ietf.org/html/rfc7578
4412 has_specified_boundary
= boundary
is not None
4415 if boundary
is None:
4416 boundary
= '---------------' + str(random
.randrange(0x0fffffff, 0xffffffff))
4419 out
, content_type
= _multipart_encode_impl(data
, boundary
)
4422 if has_specified_boundary
:
4426 return out
, content_type
4429 def dict_get(d
, key_or_keys
, default
=None, skip_false_values
=True):
4430 if isinstance(key_or_keys
, (list, tuple)):
4431 for key
in key_or_keys
:
4432 if key
not in d
or d
[key
] is None or skip_false_values
and not d
[key
]:
4436 return d
.get(key_or_keys
, default
)
4439 def try_get(src
, getter
, expected_type
=None):
4440 for get
in variadic(getter
):
4443 except (AttributeError, KeyError, TypeError, IndexError):
4446 if expected_type
is None or isinstance(v
, expected_type
):
4450 def merge_dicts(*dicts
):
4452 for a_dict
in dicts
:
4453 for k
, v
in a_dict
.items():
4457 or (isinstance(v
, compat_str
) and v
4458 and isinstance(merged
[k
], compat_str
)
4459 and not merged
[k
])):
4464 def encode_compat_str(string
, encoding
=preferredencoding(), errors
='strict'):
4465 return string
if isinstance(string
, compat_str
) else compat_str(string
, encoding
, errors
)
4477 TV_PARENTAL_GUIDELINES
= {
4487 def parse_age_limit(s
):
4489 return s
if 0 <= s
<= 21 else None
4490 if not isinstance(s
, compat_basestring
):
4492 m
= re
.match(r
'^(?P<age>\d{1,2})\+?$', s
)
4494 return int(m
.group('age'))
4497 return US_RATINGS
[s
]
4498 m
= re
.match(r
'^TV[_-]?(%s)$' % '|'.join(k
[3:] for k
in TV_PARENTAL_GUIDELINES
), s
)
4500 return TV_PARENTAL_GUIDELINES
['TV-' + m
.group(1)]
4504 def strip_jsonp(code
):
4507 (?:window\.)?(?P<func_name>[a-zA-Z0-9_.$]*)
4508 (?:\s*&&\s*(?P=func_name))?
4509 \s*\(\s*(?P<callback_data>.*)\);?
4510 \s*?(?://[^\n]*)*$''',
4511 r
'\g<callback_data>', code
)
4514 def js_to_json(code
, vars={}):
4515 # vars is a dict of var, val pairs to substitute
4516 COMMENT_RE
= r
'/\*(?:(?!\*/).)*?\*/|//[^\n]*\n'
4517 SKIP_RE
= r
'\s*(?:{comment})?\s*'.format(comment
=COMMENT_RE
)
4519 (r
'(?s)^(0[xX][0-9a-fA-F]+){skip}:?$'.format(skip
=SKIP_RE
), 16),
4520 (r
'(?s)^(0+[0-7]+){skip}:?$'.format(skip
=SKIP_RE
), 8),
4525 if v
in ('true', 'false', 'null'):
4527 elif v
in ('undefined', 'void 0'):
4529 elif v
.startswith('/*') or v
.startswith('//') or v
.startswith('!') or v
== ',':
4532 if v
[0] in ("'", '"'):
4533 v
= re
.sub(r
'(?s)\\.|"', lambda m
: {
4538 }.get(m
.group(0), m
.group(0)), v
[1:-1])
4540 for regex
, base
in INTEGER_TABLE
:
4541 im
= re
.match(regex
, v
)
4543 i
= int(im
.group(1), base
)
4544 return '"%d":' % i
if v
.endswith(':') else '%d' % i
4551 return re
.sub(r
'''(?sx)
4552 "(?:[^"\\]*(?:\\\\|\\['"nurtbfx/\n]))*[^"\\]*"|
4553 '(?:[^'\\]*(?:\\\\|\\['"nurtbfx/\n]))*[^'\\]*'|
4554 {comment}|,(?={skip}[\]}}])|
4555 void\s0|(?:(?<![0-9])[eE]|[a-df-zA-DF-Z_$])[.a-zA-Z_$0-9]*|
4556 \b(?:0[xX][0-9a-fA-F]+|0+[0-7]+)(?:{skip}:)?|
4559 '''.format(comment
=COMMENT_RE
, skip
=SKIP_RE
), fix_kv
, code
)
4562 def qualities(quality_ids
):
4563 """ Get a numeric quality value out of a list of possible values """
4566 return quality_ids
.index(qid
)
4573 'default': '%(title)s [%(id)s].%(ext)s',
4574 'chapter': '%(title)s - %(section_number)03d %(section_title)s [%(id)s].%(ext)s',
4580 'description': 'description',
4581 'annotation': 'annotations.xml',
4582 'infojson': 'info.json',
4584 'pl_thumbnail': None,
4585 'pl_description': 'description',
4586 'pl_infojson': 'info.json',
4589 # As of [1] format syntax is:
4590 # %[mapping_key][conversion_flags][minimum_width][.precision][length_modifier]type
4591 # 1. https://docs.python.org/2/library/stdtypes.html#string-formatting
4592 STR_FORMAT_RE_TMPL
= r
'''(?x)
4593 (?<!%)(?P<prefix>(?:%%)*)
4595 (?P<has_key>\((?P<key>{0})\))?
4597 (?P<conversion>[#0\-+ ]+)?
4599 (?P<precision>\.\d+)?
4600 (?P<len_mod>[hlL])? # unused in python
4601 {1} # conversion type
4606 STR_FORMAT_TYPES
= 'diouxXeEfFgGcrs'
4609 def limit_length(s
, length
):
4610 """ Add ellipses to overly long strings """
4615 return s
[:length
- len(ELLIPSES
)] + ELLIPSES
4619 def version_tuple(v
):
4620 return tuple(int(e
) for e
in re
.split(r
'[-.]', v
))
4623 def is_outdated_version(version
, limit
, assume_new
=True):
4625 return not assume_new
4627 return version_tuple(version
) < version_tuple(limit
)
4629 return not assume_new
4632 def ytdl_is_updateable():
4633 """ Returns if yt-dlp can be updated with -U """
4635 from .update
import is_non_updateable
4637 return not is_non_updateable()
4640 def args_to_str(args
):
4641 # Get a short string representation for a subprocess command
4642 return ' '.join(compat_shlex_quote(a
) for a
in args
)
4645 def error_to_compat_str(err
):
4647 # On python 2 error byte string must be decoded with proper
4648 # encoding rather than ascii
4649 if sys
.version_info
[0] < 3:
4650 err_str
= err_str
.decode(preferredencoding())
4654 def mimetype2ext(mt
):
4658 mt
, _
, params
= mt
.partition(';')
4663 # Per RFC 3003, audio/mpeg can be .mp1, .mp2 or .mp3. Here use .mp3 as
4664 # it's the most popular one
4665 'audio/mpeg': 'mp3',
4666 'audio/x-wav': 'wav',
4668 'audio/wave': 'wav',
4671 ext
= FULL_MAP
.get(mt
)
4677 'smptett+xml': 'tt',
4681 'x-mp4-fragmented': 'mp4',
4682 'x-ms-sami': 'sami',
4685 'x-mpegurl': 'm3u8',
4686 'vnd.apple.mpegurl': 'm3u8',
4690 'vnd.ms-sstr+xml': 'ism',
4694 'filmstrip+json': 'fs',
4698 _
, _
, subtype
= mt
.rpartition('/')
4699 ext
= SUBTYPE_MAP
.get(subtype
.lower())
4710 _
, _
, suffix
= subtype
.partition('+')
4711 ext
= SUFFIX_MAP
.get(suffix
)
4715 return subtype
.replace('+', '.')
4718 def parse_codecs(codecs_str
):
4719 # http://tools.ietf.org/html/rfc6381
4722 split_codecs
= list(filter(None, map(
4723 str.strip
, codecs_str
.strip().strip(',').split(','))))
4724 vcodec
, acodec
, hdr
= None, None, None
4725 for full_codec
in split_codecs
:
4726 parts
= full_codec
.split('.')
4727 codec
= parts
[0].replace('0', '')
4728 if codec
in ('avc1', 'avc2', 'avc3', 'avc4', 'vp9', 'vp8', 'hev1', 'hev2',
4729 'h263', 'h264', 'mp4v', 'hvc1', 'av1', 'theora', 'dvh1', 'dvhe'):
4731 vcodec
= '.'.join(parts
[:4]) if codec
in ('vp9', 'av1') else full_codec
4732 if codec
in ('dvh1', 'dvhe'):
4734 elif codec
== 'av1' and len(parts
) > 3 and parts
[3] == '10':
4736 elif full_codec
.replace('0', '').startswith('vp9.2'):
4738 elif codec
in ('mp4a', 'opus', 'vorbis', 'mp3', 'aac', 'ac-3', 'ec-3', 'eac3', 'dtsc', 'dtse', 'dtsh', 'dtsl'):
4742 write_string('WARNING: Unknown codec %s\n' % full_codec
, sys
.stderr
)
4743 if not vcodec
and not acodec
:
4744 if len(split_codecs
) == 2:
4746 'vcodec': split_codecs
[0],
4747 'acodec': split_codecs
[1],
4751 'vcodec': vcodec
or 'none',
4752 'acodec': acodec
or 'none',
4753 'dynamic_range': hdr
,
4758 def urlhandle_detect_ext(url_handle
):
4759 getheader
= url_handle
.headers
.get
4761 cd
= getheader('Content-Disposition')
4763 m
= re
.match(r
'attachment;\s*filename="(?P<filename>[^"]+)"', cd
)
4765 e
= determine_ext(m
.group('filename'), default_ext
=None)
4769 return mimetype2ext(getheader('Content-Type'))
4772 def encode_data_uri(data
, mime_type
):
4773 return 'data:%s;base64,%s' % (mime_type
, base64
.b64encode(data
).decode('ascii'))
4776 def age_restricted(content_limit
, age_limit
):
4777 """ Returns True iff the content should be blocked """
4779 if age_limit
is None: # No limit set
4781 if content_limit
is None:
4782 return False # Content available for everyone
4783 return age_limit
< content_limit
4786 def is_html(first_bytes
):
4787 """ Detect whether a file contains HTML by examining its first bytes. """
4790 (b
'\xef\xbb\xbf', 'utf-8'),
4791 (b
'\x00\x00\xfe\xff', 'utf-32-be'),
4792 (b
'\xff\xfe\x00\x00', 'utf-32-le'),
4793 (b
'\xff\xfe', 'utf-16-le'),
4794 (b
'\xfe\xff', 'utf-16-be'),
4796 for bom
, enc
in BOMS
:
4797 if first_bytes
.startswith(bom
):
4798 s
= first_bytes
[len(bom
):].decode(enc
, 'replace')
4801 s
= first_bytes
.decode('utf-8', 'replace')
4803 return re
.match(r
'^\s*<', s
)
4806 def determine_protocol(info_dict
):
4807 protocol
= info_dict
.get('protocol')
4808 if protocol
is not None:
4811 url
= sanitize_url(info_dict
['url'])
4812 if url
.startswith('rtmp'):
4814 elif url
.startswith('mms'):
4816 elif url
.startswith('rtsp'):
4819 ext
= determine_ext(url
)
4825 return compat_urllib_parse_urlparse(url
).scheme
4828 def render_table(header_row
, data
, delim
=False, extra_gap
=0, hide_empty
=False):
4829 """ Render a list of rows, each as a list of values.
4830 Text after a \t will be right aligned """
4832 return len(remove_terminal_sequences(string
).replace('\t', ''))
4834 def get_max_lens(table
):
4835 return [max(width(str(v
)) for v
in col
) for col
in zip(*table
)]
4837 def filter_using_list(row
, filterArray
):
4838 return [col
for (take
, col
) in zip(filterArray
, row
) if take
]
4841 max_lens
= get_max_lens(data
)
4842 header_row
= filter_using_list(header_row
, max_lens
)
4843 data
= [filter_using_list(row
, max_lens
) for row
in data
]
4845 table
= [header_row
] + data
4846 max_lens
= get_max_lens(table
)
4849 table
= [header_row
, [delim
* (ml
+ extra_gap
) for ml
in max_lens
]] + data
4850 table
[1][-1] = table
[1][-1][:-extra_gap
] # Remove extra_gap from end of delimiter
4852 for pos
, text
in enumerate(map(str, row
)):
4854 row
[pos
] = text
.replace('\t', ' ' * (max_lens
[pos
] - width(text
))) + ' ' * extra_gap
4856 row
[pos
] = text
+ ' ' * (max_lens
[pos
] - width(text
) + extra_gap
)
4857 ret
= '\n'.join(''.join(row
).rstrip() for row
in table
)
4861 def _match_one(filter_part
, dct
, incomplete
):
4862 # TODO: Generalize code with YoutubeDL._build_format_filter
4863 STRING_OPERATORS
= {
4864 '*=': operator
.contains
,
4865 '^=': lambda attr
, value
: attr
.startswith(value
),
4866 '$=': lambda attr
, value
: attr
.endswith(value
),
4867 '~=': lambda attr
, value
: re
.search(value
, attr
),
4869 COMPARISON_OPERATORS
= {
4871 '<=': operator
.le
, # "<=" must be defined above "<"
4878 operator_rex
= re
.compile(r
'''(?x)\s*
4880 \s*(?P<negation>!\s*)?(?P<op>%s)(?P<none_inclusive>\s*\?)?\s*
4882 (?P<quote>["\'])(?P
<quotedstrval
>.+?
)(?P
=quote
)|
4886 ''' % '|'.join(map(re.escape, COMPARISON_OPERATORS.keys())))
4887 m = operator_rex.search(filter_part)
4890 unnegated_op = COMPARISON_OPERATORS[m['op']]
4892 op = lambda attr, value: not unnegated_op(attr, value)
4895 comparison_value = m['quotedstrval'] or m['strval'] or m['intval']
4897 comparison_value = comparison_value.replace(r'\%s' % m['quote'], m['quote'])
4898 actual_value = dct.get(m['key'])
4899 numeric_comparison = None
4900 if isinstance(actual_value, compat_numeric_types):
4901 # If the original field is a string and matching comparisonvalue is
4902 # a number we should respect the origin of the original field
4903 # and process comparison value as a string (see
4904 # https://github.com/ytdl-org/youtube-dl/issues/11082)
4906 numeric_comparison = int(comparison_value)
4908 numeric_comparison = parse_filesize(comparison_value)
4909 if numeric_comparison is None:
4910 numeric_comparison = parse_filesize(f'{comparison_value}B')
4911 if numeric_comparison is None:
4912 numeric_comparison = parse_duration(comparison_value)
4913 if numeric_comparison is not None and m['op'] in STRING_OPERATORS:
4914 raise ValueError('Operator %s only supports string values!' % m['op'])
4915 if actual_value is None:
4916 return incomplete or m['none_inclusive']
4917 return op(actual_value, comparison_value if numeric_comparison is None else numeric_comparison)
4920 '': lambda v: (v is True) if isinstance(v, bool) else (v is not None),
4921 '!': lambda v: (v is False) if isinstance(v, bool) else (v is None),
4923 operator_rex = re.compile(r'''(?x
)\s
*
4924 (?P
<op
>%s)\s
*(?P
<key
>[a
-z_
]+)
4926 ''' % '|'.join(map(re.escape, UNARY_OPERATORS.keys())))
4927 m = operator_rex.search(filter_part)
4929 op = UNARY_OPERATORS[m.group('op')]
4930 actual_value = dct.get(m.group('key'))
4931 if incomplete and actual_value is None:
4933 return op(actual_value)
4935 raise ValueError('Invalid filter part %r' % filter_part)
4938 def match_str(filter_str, dct, incomplete=False):
4939 """ Filter a dictionary with a simple string syntax. Returns True (=passes filter) or false
4940 When incomplete, all conditions passes on missing fields
4943 _match_one(filter_part.replace(r'\&', '&'), dct, incomplete)
4944 for filter_part in re.split(r'(?<!\\)&', filter_str))
4947 def match_filter_func(filter_str):
4948 def _match_func(info_dict, *args, **kwargs):
4949 if match_str(filter_str, info_dict, *args, **kwargs):
4952 video_title = info_dict.get('title', info_dict.get('id', 'video'))
4953 return '%s does not pass filter %s, skipping ..' % (video_title, filter_str)
4957 def parse_dfxp_time_expr(time_expr):
4961 mobj = re.match(r'^(?P<time_offset>\d+(?:\.\d+)?)s?$', time_expr)
4963 return float(mobj.group('time_offset'))
4965 mobj = re.match(r'^(\d+):(\d\d):(\d\d(?:(?:\.|:)\d+)?)$', time_expr)
4967 return 3600 * int(mobj.group(1)) + 60 * int(mobj.group(2)) + float(mobj.group(3).replace(':', '.'))
4970 def srt_subtitles_timecode(seconds):
4971 return '%02d:%02d:%02d,%03d' % timetuple_from_msec(seconds * 1000)
4974 def ass_subtitles_timecode(seconds):
4975 time = timetuple_from_msec(seconds * 1000)
4976 return '%01d:%02d:%02d.%02d' % (*time[:-1], time.milliseconds / 10)
4979 def dfxp2srt(dfxp_data):
4981 @param dfxp_data A
bytes-like
object containing DFXP data
4982 @returns A
unicode object containing converted SRT data
4984 LEGACY_NAMESPACES = (
4985 (b'http://www.w3.org/ns/ttml', [
4986 b'http://www.w3.org/2004/11/ttaf1',
4987 b'http://www.w3.org/2006/04/ttaf1',
4988 b'http://www.w3.org/2006/10/ttaf1',
4990 (b'http://www.w3.org/ns/ttml#styling', [
4991 b'http://www.w3.org/ns/ttml#style',
4995 SUPPORTED_STYLING = [
5004 _x = functools.partial(xpath_with_ns, ns_map={
5005 'xml': 'http://www.w3.org/XML/1998/namespace',
5006 'ttml': 'http://www.w3.org/ns/ttml',
5007 'tts': 'http://www.w3.org/ns/ttml#styling',
5013 class TTMLPElementParser(object):
5015 _unclosed_elements = []
5016 _applied_styles = []
5018 def start(self, tag, attrib):
5019 if tag in (_x('ttml:br'), 'br'):
5022 unclosed_elements = []
5024 element_style_id = attrib.get('style')
5026 style.update(default_style)
5027 if element_style_id:
5028 style.update(styles.get(element_style_id, {}))
5029 for prop in SUPPORTED_STYLING:
5030 prop_val = attrib.get(_x('tts:' + prop))
5032 style[prop] = prop_val
5035 for k, v in sorted(style.items()):
5036 if self._applied_styles and self._applied_styles[-1].get(k) == v:
5039 font += ' color="%s"' % v
5040 elif k == 'fontSize':
5041 font += ' size="%s"' % v
5042 elif k == 'fontFamily':
5043 font += ' face="%s"' % v
5044 elif k == 'fontWeight' and v == 'bold':
5046 unclosed_elements.append('b')
5047 elif k == 'fontStyle' and v == 'italic':
5049 unclosed_elements.append('i')
5050 elif k == 'textDecoration' and v == 'underline':
5052 unclosed_elements.append('u')
5054 self._out += '<font' + font + '>'
5055 unclosed_elements.append('font')
5057 if self._applied_styles:
5058 applied_style.update(self._applied_styles[-1])
5059 applied_style.update(style)
5060 self._applied_styles.append(applied_style)
5061 self._unclosed_elements.append(unclosed_elements)
5064 if tag not in (_x('ttml:br'), 'br'):
5065 unclosed_elements = self._unclosed_elements.pop()
5066 for element in reversed(unclosed_elements):
5067 self._out += '</%s>' % element
5068 if unclosed_elements and self._applied_styles:
5069 self._applied_styles.pop()
5071 def data(self, data):
5075 return self._out.strip()
5077 def parse_node(node):
5078 target = TTMLPElementParser()
5079 parser = xml.etree.ElementTree.XMLParser(target=target)
5080 parser.feed(xml.etree.ElementTree.tostring(node))
5081 return parser.close()
5083 for k, v in LEGACY_NAMESPACES:
5085 dfxp_data = dfxp_data.replace(ns, k)
5087 dfxp = compat_etree_fromstring(dfxp_data)
5089 paras = dfxp.findall(_x('.//ttml:p')) or dfxp.findall('.//p')
5092 raise ValueError('Invalid dfxp/TTML subtitle')
5096 for style in dfxp.findall(_x('.//ttml:style')):
5097 style_id = style.get('id') or style.get(_x('xml:id'))
5100 parent_style_id = style.get('style')
5102 if parent_style_id not in styles:
5105 styles[style_id] = styles[parent_style_id].copy()
5106 for prop in SUPPORTED_STYLING:
5107 prop_val = style.get(_x('tts:' + prop))
5109 styles.setdefault(style_id, {})[prop] = prop_val
5115 for p in ('body', 'div'):
5116 ele = xpath_element(dfxp, [_x('.//ttml:' + p), './/' + p])
5119 style = styles.get(ele.get('style'))
5122 default_style.update(style)
5124 for para, index in zip(paras, itertools.count(1)):
5125 begin_time = parse_dfxp_time_expr(para.attrib.get('begin'))
5126 end_time = parse_dfxp_time_expr(para.attrib.get('end'))
5127 dur = parse_dfxp_time_expr(para.attrib.get('dur'))
5128 if begin_time is None:
5133 end_time = begin_time + dur
5134 out.append('%d\n%s --> %s\n%s\n\n' % (
5136 srt_subtitles_timecode(begin_time),
5137 srt_subtitles_timecode(end_time),
5143 def cli_option(params, command_option, param):
5144 param = params.get(param)
5146 param = compat_str(param)
5147 return [command_option, param] if param is not None else []
5150 def cli_bool_option(params, command_option, param, true_value='true', false_value='false', separator=None):
5151 param = params.get(param)
5154 assert isinstance(param, bool)
5156 return [command_option + separator + (true_value if param else false_value)]
5157 return [command_option, true_value if param else false_value]
5160 def cli_valueless_option(params, command_option, param, expected_value=True):
5161 param = params.get(param)
5162 return [command_option] if param == expected_value else []
5165 def cli_configuration_args(argdict, keys, default=[], use_compat=True):
5166 if isinstance(argdict, (list, tuple)): # for backward compatibility
5173 assert isinstance(argdict, dict)
5175 assert isinstance(keys, (list, tuple))
5176 for key_list in keys:
5177 arg_list = list(filter(
5178 lambda x: x is not None,
5179 [argdict.get(key.lower()) for key in variadic(key_list)]))
5181 return [arg for args in arg_list for arg in args]
5185 def _configuration_args(main_key, argdict, exe, keys=None, default=[], use_compat=True):
5186 main_key, exe = main_key.lower(), exe.lower()
5187 root_key = exe if main_key == exe else f'{main_key}+{exe}'
5188 keys = [f'{root_key}{k}' for k in (keys or [''])]
5189 if root_key in keys:
5191 keys.append((main_key, exe))
5192 keys.append('default')
5195 return cli_configuration_args(argdict, keys, default, use_compat)
5198 class ISO639Utils(object):
5199 # See http://www.loc.gov/standards/iso639-2/ISO-639-2_utf-8.txt
5258 'iw': 'heb', # Replaced by he in 1989 revision
5268 'in': 'ind', # Replaced by id in 1989 revision
5383 'ji': 'yid', # Replaced by yi in 1989 revision
5391 def short2long(cls, code):
5392 """Convert language code from ISO 639-1 to ISO 639-2/T"""
5393 return cls._lang_map.get(code[:2])
5396 def long2short(cls, code):
5397 """Convert language code from ISO 639-2/T to ISO 639-1"""
5398 for short_name, long_name in cls._lang_map.items():
5399 if long_name == code:
5403 class ISO3166Utils(object):
5404 # From http://data.okfn.org/data/core/country-list
5406 'AF': 'Afghanistan',
5407 'AX': 'Åland Islands',
5410 'AS': 'American Samoa',
5415 'AG': 'Antigua and Barbuda',
5432 'BO': 'Bolivia, Plurinational State of',
5433 'BQ': 'Bonaire, Sint Eustatius and Saba',
5434 'BA': 'Bosnia and Herzegovina',
5436 'BV': 'Bouvet Island',
5438 'IO': 'British Indian Ocean Territory',
5439 'BN': 'Brunei Darussalam',
5441 'BF': 'Burkina Faso',
5447 'KY': 'Cayman Islands',
5448 'CF': 'Central African Republic',
5452 'CX': 'Christmas Island',
5453 'CC': 'Cocos (Keeling) Islands',
5457 'CD': 'Congo, the Democratic Republic of the',
5458 'CK': 'Cook Islands',
5460 'CI': 'Côte d\'Ivoire',
5465 'CZ': 'Czech Republic',
5469 'DO': 'Dominican Republic',
5472 'SV': 'El Salvador',
5473 'GQ': 'Equatorial Guinea',
5477 'FK': 'Falkland Islands (Malvinas)',
5478 'FO': 'Faroe Islands',
5482 'GF': 'French Guiana',
5483 'PF': 'French Polynesia',
5484 'TF': 'French Southern Territories',
5499 'GW': 'Guinea-Bissau',
5502 'HM': 'Heard Island and McDonald Islands',
5503 'VA': 'Holy See (Vatican City State)',
5510 'IR': 'Iran, Islamic Republic of',
5513 'IM': 'Isle of Man',
5523 'KP': 'Korea, Democratic People\'s Republic of',
5524 'KR': 'Korea, Republic of',
5527 'LA': 'Lao People\'s Democratic Republic',
5533 'LI': 'Liechtenstein',
5537 'MK': 'Macedonia, the Former Yugoslav Republic of',
5544 'MH': 'Marshall Islands',
5550 'FM': 'Micronesia, Federated States of',
5551 'MD': 'Moldova, Republic of',
5562 'NL': 'Netherlands',
5563 'NC': 'New Caledonia',
5564 'NZ': 'New Zealand',
5569 'NF': 'Norfolk Island',
5570 'MP': 'Northern Mariana Islands',
5575 'PS': 'Palestine, State of',
5577 'PG': 'Papua New Guinea',
5580 'PH': 'Philippines',
5584 'PR': 'Puerto Rico',
5588 'RU': 'Russian Federation',
5590 'BL': 'Saint Barthélemy',
5591 'SH': 'Saint Helena, Ascension and Tristan da Cunha',
5592 'KN': 'Saint Kitts and Nevis',
5593 'LC': 'Saint Lucia',
5594 'MF': 'Saint Martin (French part)',
5595 'PM': 'Saint Pierre and Miquelon',
5596 'VC': 'Saint Vincent and the Grenadines',
5599 'ST': 'Sao Tome and Principe',
5600 'SA': 'Saudi Arabia',
5604 'SL': 'Sierra Leone',
5606 'SX': 'Sint Maarten (Dutch part)',
5609 'SB': 'Solomon Islands',
5611 'ZA': 'South Africa',
5612 'GS': 'South Georgia and the South Sandwich Islands',
5613 'SS': 'South Sudan',
5618 'SJ': 'Svalbard and Jan Mayen',
5621 'CH': 'Switzerland',
5622 'SY': 'Syrian Arab Republic',
5623 'TW': 'Taiwan, Province of China',
5625 'TZ': 'Tanzania, United Republic of',
5627 'TL': 'Timor-Leste',
5631 'TT': 'Trinidad and Tobago',
5634 'TM': 'Turkmenistan',
5635 'TC': 'Turks and Caicos Islands',
5639 'AE': 'United Arab Emirates',
5640 'GB': 'United Kingdom',
5641 'US': 'United States',
5642 'UM': 'United States Minor Outlying Islands',
5646 'VE': 'Venezuela, Bolivarian Republic of',
5648 'VG': 'Virgin Islands, British',
5649 'VI': 'Virgin Islands, U.S.',
5650 'WF': 'Wallis and Futuna',
5651 'EH': 'Western Sahara',
5658 def short2full(cls, code):
5659 """Convert an ISO 3166-2 country code to the corresponding full name"""
5660 return cls._country_map.get(code.upper())
5663 class GeoUtils(object):
5664 # Major IPv4 address blocks per country
5666 'AD': '46.172.224.0/19',
5667 'AE': '94.200.0.0/13',
5668 'AF': '149.54.0.0/17',
5669 'AG': '209.59.64.0/18',
5670 'AI': '204.14.248.0/21',
5671 'AL': '46.99.0.0/16',
5672 'AM': '46.70.0.0/15',
5673 'AO': '105.168.0.0/13',
5674 'AP': '182.50.184.0/21',
5675 'AQ': '23.154.160.0/24',
5676 'AR': '181.0.0.0/12',
5677 'AS': '202.70.112.0/20',
5678 'AT': '77.116.0.0/14',
5679 'AU': '1.128.0.0/11',
5680 'AW': '181.41.0.0/18',
5681 'AX': '185.217.4.0/22',
5682 'AZ': '5.197.0.0/16',
5683 'BA': '31.176.128.0/17',
5684 'BB': '65.48.128.0/17',
5685 'BD': '114.130.0.0/16',
5687 'BF': '102.178.0.0/15',
5688 'BG': '95.42.0.0/15',
5689 'BH': '37.131.0.0/17',
5690 'BI': '154.117.192.0/18',
5691 'BJ': '137.255.0.0/16',
5692 'BL': '185.212.72.0/23',
5693 'BM': '196.12.64.0/18',
5694 'BN': '156.31.0.0/16',
5695 'BO': '161.56.0.0/16',
5696 'BQ': '161.0.80.0/20',
5697 'BR': '191.128.0.0/12',
5698 'BS': '24.51.64.0/18',
5699 'BT': '119.2.96.0/19',
5700 'BW': '168.167.0.0/16',
5701 'BY': '178.120.0.0/13',
5702 'BZ': '179.42.192.0/18',
5703 'CA': '99.224.0.0/11',
5704 'CD': '41.243.0.0/16',
5705 'CF': '197.242.176.0/21',
5706 'CG': '160.113.0.0/16',
5707 'CH': '85.0.0.0/13',
5708 'CI': '102.136.0.0/14',
5709 'CK': '202.65.32.0/19',
5710 'CL': '152.172.0.0/14',
5711 'CM': '102.244.0.0/14',
5712 'CN': '36.128.0.0/10',
5713 'CO': '181.240.0.0/12',
5714 'CR': '201.192.0.0/12',
5715 'CU': '152.206.0.0/15',
5716 'CV': '165.90.96.0/19',
5717 'CW': '190.88.128.0/17',
5718 'CY': '31.153.0.0/16',
5719 'CZ': '88.100.0.0/14',
5721 'DJ': '197.241.0.0/17',
5722 'DK': '87.48.0.0/12',
5723 'DM': '192.243.48.0/20',
5724 'DO': '152.166.0.0/15',
5725 'DZ': '41.96.0.0/12',
5726 'EC': '186.68.0.0/15',
5727 'EE': '90.190.0.0/15',
5728 'EG': '156.160.0.0/11',
5729 'ER': '196.200.96.0/20',
5730 'ES': '88.0.0.0/11',
5731 'ET': '196.188.0.0/14',
5732 'EU': '2.16.0.0/13',
5733 'FI': '91.152.0.0/13',
5734 'FJ': '144.120.0.0/16',
5735 'FK': '80.73.208.0/21',
5736 'FM': '119.252.112.0/20',
5737 'FO': '88.85.32.0/19',
5739 'GA': '41.158.0.0/15',
5741 'GD': '74.122.88.0/21',
5742 'GE': '31.146.0.0/16',
5743 'GF': '161.22.64.0/18',
5744 'GG': '62.68.160.0/19',
5745 'GH': '154.160.0.0/12',
5746 'GI': '95.164.0.0/16',
5747 'GL': '88.83.0.0/19',
5748 'GM': '160.182.0.0/15',
5749 'GN': '197.149.192.0/18',
5750 'GP': '104.250.0.0/19',
5751 'GQ': '105.235.224.0/20',
5752 'GR': '94.64.0.0/13',
5753 'GT': '168.234.0.0/16',
5754 'GU': '168.123.0.0/16',
5755 'GW': '197.214.80.0/20',
5756 'GY': '181.41.64.0/18',
5757 'HK': '113.252.0.0/14',
5758 'HN': '181.210.0.0/16',
5759 'HR': '93.136.0.0/13',
5760 'HT': '148.102.128.0/17',
5761 'HU': '84.0.0.0/14',
5762 'ID': '39.192.0.0/10',
5763 'IE': '87.32.0.0/12',
5764 'IL': '79.176.0.0/13',
5765 'IM': '5.62.80.0/20',
5766 'IN': '117.192.0.0/10',
5767 'IO': '203.83.48.0/21',
5768 'IQ': '37.236.0.0/14',
5769 'IR': '2.176.0.0/12',
5770 'IS': '82.221.0.0/16',
5771 'IT': '79.0.0.0/10',
5772 'JE': '87.244.64.0/18',
5773 'JM': '72.27.0.0/17',
5774 'JO': '176.29.0.0/16',
5775 'JP': '133.0.0.0/8',
5776 'KE': '105.48.0.0/12',
5777 'KG': '158.181.128.0/17',
5778 'KH': '36.37.128.0/17',
5779 'KI': '103.25.140.0/22',
5780 'KM': '197.255.224.0/20',
5781 'KN': '198.167.192.0/19',
5782 'KP': '175.45.176.0/22',
5783 'KR': '175.192.0.0/10',
5784 'KW': '37.36.0.0/14',
5785 'KY': '64.96.0.0/15',
5786 'KZ': '2.72.0.0/13',
5787 'LA': '115.84.64.0/18',
5788 'LB': '178.135.0.0/16',
5789 'LC': '24.92.144.0/20',
5790 'LI': '82.117.0.0/19',
5791 'LK': '112.134.0.0/15',
5792 'LR': '102.183.0.0/16',
5793 'LS': '129.232.0.0/17',
5794 'LT': '78.56.0.0/13',
5795 'LU': '188.42.0.0/16',
5796 'LV': '46.109.0.0/16',
5797 'LY': '41.252.0.0/14',
5798 'MA': '105.128.0.0/11',
5799 'MC': '88.209.64.0/18',
5800 'MD': '37.246.0.0/16',
5801 'ME': '178.175.0.0/17',
5802 'MF': '74.112.232.0/21',
5803 'MG': '154.126.0.0/17',
5804 'MH': '117.103.88.0/21',
5805 'MK': '77.28.0.0/15',
5806 'ML': '154.118.128.0/18',
5807 'MM': '37.111.0.0/17',
5808 'MN': '49.0.128.0/17',
5809 'MO': '60.246.0.0/16',
5810 'MP': '202.88.64.0/20',
5811 'MQ': '109.203.224.0/19',
5812 'MR': '41.188.64.0/18',
5813 'MS': '208.90.112.0/22',
5814 'MT': '46.11.0.0/16',
5815 'MU': '105.16.0.0/12',
5816 'MV': '27.114.128.0/18',
5817 'MW': '102.70.0.0/15',
5818 'MX': '187.192.0.0/11',
5819 'MY': '175.136.0.0/13',
5820 'MZ': '197.218.0.0/15',
5821 'NA': '41.182.0.0/16',
5822 'NC': '101.101.0.0/18',
5823 'NE': '197.214.0.0/18',
5824 'NF': '203.17.240.0/22',
5825 'NG': '105.112.0.0/12',
5826 'NI': '186.76.0.0/15',
5827 'NL': '145.96.0.0/11',
5828 'NO': '84.208.0.0/13',
5829 'NP': '36.252.0.0/15',
5830 'NR': '203.98.224.0/19',
5831 'NU': '49.156.48.0/22',
5832 'NZ': '49.224.0.0/14',
5833 'OM': '5.36.0.0/15',
5834 'PA': '186.72.0.0/15',
5835 'PE': '186.160.0.0/14',
5836 'PF': '123.50.64.0/18',
5837 'PG': '124.240.192.0/19',
5838 'PH': '49.144.0.0/13',
5839 'PK': '39.32.0.0/11',
5840 'PL': '83.0.0.0/11',
5841 'PM': '70.36.0.0/20',
5842 'PR': '66.50.0.0/16',
5843 'PS': '188.161.0.0/16',
5844 'PT': '85.240.0.0/13',
5845 'PW': '202.124.224.0/20',
5846 'PY': '181.120.0.0/14',
5847 'QA': '37.210.0.0/15',
5848 'RE': '102.35.0.0/16',
5849 'RO': '79.112.0.0/13',
5850 'RS': '93.86.0.0/15',
5851 'RU': '5.136.0.0/13',
5852 'RW': '41.186.0.0/16',
5853 'SA': '188.48.0.0/13',
5854 'SB': '202.1.160.0/19',
5855 'SC': '154.192.0.0/11',
5856 'SD': '102.120.0.0/13',
5857 'SE': '78.64.0.0/12',
5858 'SG': '8.128.0.0/10',
5859 'SI': '188.196.0.0/14',
5860 'SK': '78.98.0.0/15',
5861 'SL': '102.143.0.0/17',
5862 'SM': '89.186.32.0/19',
5863 'SN': '41.82.0.0/15',
5864 'SO': '154.115.192.0/18',
5865 'SR': '186.179.128.0/17',
5866 'SS': '105.235.208.0/21',
5867 'ST': '197.159.160.0/19',
5868 'SV': '168.243.0.0/16',
5869 'SX': '190.102.0.0/20',
5871 'SZ': '41.84.224.0/19',
5872 'TC': '65.255.48.0/20',
5873 'TD': '154.68.128.0/19',
5874 'TG': '196.168.0.0/14',
5875 'TH': '171.96.0.0/13',
5876 'TJ': '85.9.128.0/18',
5877 'TK': '27.96.24.0/21',
5878 'TL': '180.189.160.0/20',
5879 'TM': '95.85.96.0/19',
5880 'TN': '197.0.0.0/11',
5881 'TO': '175.176.144.0/21',
5882 'TR': '78.160.0.0/11',
5883 'TT': '186.44.0.0/15',
5884 'TV': '202.2.96.0/19',
5885 'TW': '120.96.0.0/11',
5886 'TZ': '156.156.0.0/14',
5887 'UA': '37.52.0.0/14',
5888 'UG': '102.80.0.0/13',
5890 'UY': '167.56.0.0/13',
5891 'UZ': '84.54.64.0/18',
5892 'VA': '212.77.0.0/19',
5893 'VC': '207.191.240.0/21',
5894 'VE': '186.88.0.0/13',
5895 'VG': '66.81.192.0/20',
5896 'VI': '146.226.0.0/16',
5897 'VN': '14.160.0.0/11',
5898 'VU': '202.80.32.0/20',
5899 'WF': '117.20.32.0/21',
5900 'WS': '202.4.32.0/19',
5901 'YE': '134.35.0.0/16',
5902 'YT': '41.242.116.0/22',
5903 'ZA': '41.0.0.0/11',
5904 'ZM': '102.144.0.0/13',
5905 'ZW': '102.177.192.0/18',
5909 def random_ipv4(cls, code_or_block):
5910 if len(code_or_block) == 2:
5911 block = cls._country_ip_map.get(code_or_block.upper())
5915 block = code_or_block
5916 addr, preflen = block.split('/')
5917 addr_min = compat_struct_unpack('!L', socket.inet_aton(addr))[0]
5918 addr_max = addr_min | (0xffffffff >> int(preflen))
5919 return compat_str(socket.inet_ntoa(
5920 compat_struct_pack('!L', random.randint(addr_min, addr_max))))
5923 class PerRequestProxyHandler(compat_urllib_request.ProxyHandler):
5924 def __init__(self, proxies=None):
5925 # Set default handlers
5926 for type in ('http', 'https'):
5927 setattr(self, '%s_open' % type,
5928 lambda r, proxy='__noproxy__', type=type, meth=self.proxy_open:
5929 meth(r, proxy, type))
5930 compat_urllib_request.ProxyHandler.__init__(self, proxies)
5932 def proxy_open(self, req, proxy, type):
5933 req_proxy = req.headers.get('Ytdl-request-proxy')
5934 if req_proxy is not None:
5936 del req.headers['Ytdl-request-proxy']
5938 if proxy == '__noproxy__':
5939 return None # No Proxy
5940 if compat_urlparse.urlparse(proxy).scheme.lower() in ('socks', 'socks4', 'socks4a', 'socks5'):
5941 req.add_header('Ytdl-socks-proxy', proxy)
5942 # yt-dlp's http/https handlers do wrapping the socket with socks
5944 return compat_urllib_request.ProxyHandler.proxy_open(
5945 self, req, proxy, type)
5948 # Both long_to_bytes and bytes_to_long are adapted from PyCrypto, which is
5949 # released into Public Domain
5950 # https://github.com/dlitz/pycrypto/blob/master/lib/Crypto/Util/number.py#L387
5952 def long_to_bytes(n, blocksize=0):
5953 """long_to_bytes(n:long, blocksize:int) : string
5954 Convert a long integer to a byte string.
5956 If optional blocksize is given and greater than zero, pad the front of the
5957 byte string with binary zeros so that the length is a multiple of
5960 # after much testing, this algorithm was deemed to be the fastest
5964 s = compat_struct_pack('>I', n & 0xffffffff) + s
5966 # strip off leading zeros
5967 for i in range(len(s)):
5968 if s[i] != b'\000'[0]:
5971 # only happens when n == 0
5975 # add back some pad bytes. this could be done more efficiently w.r.t. the
5976 # de-padding being done above, but sigh...
5977 if blocksize > 0 and len(s) % blocksize:
5978 s = (blocksize - len(s) % blocksize) * b'\000' + s
5982 def bytes_to_long(s):
5983 """bytes_to_long(string) : long
5984 Convert a byte string to a long integer.
5986 This is (essentially) the inverse of long_to_bytes().
5991 extra = (4 - length % 4)
5992 s = b'\000' * extra + s
5993 length = length + extra
5994 for i in range(0, length, 4):
5995 acc = (acc << 32) + compat_struct_unpack('>I', s[i:i + 4])[0]
5999 def ohdave_rsa_encrypt(data, exponent, modulus):
6001 Implement OHDave
's RSA algorithm. See http://www.ohdave.com/rsa/
6004 data: data to encrypt, bytes-like object
6005 exponent, modulus: parameter e and N of RSA algorithm, both integer
6006 Output: hex string of encrypted data
6008 Limitation: supports one block encryption only
6011 payload = int(binascii.hexlify(data[::-1]), 16)
6012 encrypted = pow(payload, exponent, modulus)
6013 return '%x' % encrypted
6016 def pkcs1pad(data, length):
6018 Padding input data with PKCS#1 scheme
6020 @param {int[]} data input data
6021 @param {int} length target length
6022 @returns {int[]} padded data
6024 if len(data) > length - 11:
6025 raise ValueError('Input data too
long for PKCS
#1 padding')
6027 pseudo_random
= [random
.randint(0, 254) for _
in range(length
- len(data
) - 3)]
6028 return [0, 2] + pseudo_random
+ [0] + data
6031 def encode_base_n(num
, n
, table
=None):
6032 FULL_TABLE
= '0123456789abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ'
6034 table
= FULL_TABLE
[:n
]
6037 raise ValueError('base %d exceeds table length %d' % (n
, len(table
)))
6044 ret
= table
[num
% n
] + ret
6049 def decode_packed_codes(code
):
6050 mobj
= re
.search(PACKED_CODES_RE
, code
)
6051 obfuscated_code
, base
, count
, symbols
= mobj
.groups()
6054 symbols
= symbols
.split('|')
6059 base_n_count
= encode_base_n(count
, base
)
6060 symbol_table
[base_n_count
] = symbols
[count
] or base_n_count
6063 r
'\b(\w+)\b', lambda mobj
: symbol_table
[mobj
.group(0)],
6067 def caesar(s
, alphabet
, shift
):
6072 alphabet
[(alphabet
.index(c
) + shift
) % l
] if c
in alphabet
else c
6077 return caesar(s
, r
'''!"#$%&'()*+,-./0123456789:;<=>?@ABCDEFGHIJKLMNOPQRSTUVWXYZ[\]^_`abcdefghijklmnopqrstuvwxyz{|}~''', 47)
6080 def parse_m3u8_attributes(attrib
):
6082 for (key
, val
) in re
.findall(r
'(?P<key>[A-Z0-9-]+)=(?P<val>"[^"]+"|[^",]+)(?:,|$)', attrib
):
6083 if val
.startswith('"'):
6089 def urshift(val
, n
):
6090 return val
>> n
if val
>= 0 else (val
+ 0x100000000) >> n
6093 # Based on png2str() written by @gdkchan and improved by @yokrysty
6094 # Originally posted at https://github.com/ytdl-org/youtube-dl/issues/9706
6095 def decode_png(png_data
):
6096 # Reference: https://www.w3.org/TR/PNG/
6097 header
= png_data
[8:]
6099 if png_data
[:8] != b
'\x89PNG\x0d\x0a\x1a\x0a' or header
[4:8] != b
'IHDR':
6100 raise IOError('Not a valid PNG file.')
6102 int_map
= {1: '>B', 2: '>H', 4: '>I'}
6103 unpack_integer
= lambda x
: compat_struct_unpack(int_map
[len(x
)], x
)[0]
6108 length
= unpack_integer(header
[:4])
6111 chunk_type
= header
[:4]
6114 chunk_data
= header
[:length
]
6115 header
= header
[length
:]
6117 header
= header
[4:] # Skip CRC
6125 ihdr
= chunks
[0]['data']
6127 width
= unpack_integer(ihdr
[:4])
6128 height
= unpack_integer(ihdr
[4:8])
6132 for chunk
in chunks
:
6133 if chunk
['type'] == b
'IDAT':
6134 idat
+= chunk
['data']
6137 raise IOError('Unable to read PNG data.')
6139 decompressed_data
= bytearray(zlib
.decompress(idat
))
6144 def _get_pixel(idx
):
6149 for y
in range(height
):
6150 basePos
= y
* (1 + stride
)
6151 filter_type
= decompressed_data
[basePos
]
6155 pixels
.append(current_row
)
6157 for x
in range(stride
):
6158 color
= decompressed_data
[1 + basePos
+ x
]
6159 basex
= y
* stride
+ x
6164 left
= _get_pixel(basex
- 3)
6166 up
= _get_pixel(basex
- stride
)
6168 if filter_type
== 1: # Sub
6169 color
= (color
+ left
) & 0xff
6170 elif filter_type
== 2: # Up
6171 color
= (color
+ up
) & 0xff
6172 elif filter_type
== 3: # Average
6173 color
= (color
+ ((left
+ up
) >> 1)) & 0xff
6174 elif filter_type
== 4: # Paeth
6180 c
= _get_pixel(basex
- stride
- 3)
6188 if pa
<= pb
and pa
<= pc
:
6189 color
= (color
+ a
) & 0xff
6191 color
= (color
+ b
) & 0xff
6193 color
= (color
+ c
) & 0xff
6195 current_row
.append(color
)
6197 return width
, height
, pixels
6200 def write_xattr(path
, key
, value
):
6201 # This mess below finds the best xattr tool for the job
6203 # try the pyxattr module...
6206 if hasattr(xattr
, 'set'): # pyxattr
6207 # Unicode arguments are not supported in python-pyxattr until
6209 # See https://github.com/ytdl-org/youtube-dl/issues/5498
6210 pyxattr_required_version
= '0.5.0'
6211 if version_tuple(xattr
.__version
__) < version_tuple(pyxattr_required_version
):
6212 # TODO: fallback to CLI tools
6213 raise XAttrUnavailableError(
6214 'python-pyxattr is detected but is too old. '
6215 'yt-dlp requires %s or above while your version is %s. '
6216 'Falling back to other xattr implementations' % (
6217 pyxattr_required_version
, xattr
.__version
__))
6219 setxattr
= xattr
.set
6221 setxattr
= xattr
.setxattr
6224 setxattr(path
, key
, value
)
6225 except EnvironmentError as e
:
6226 raise XAttrMetadataError(e
.errno
, e
.strerror
)
6229 if compat_os_name
== 'nt':
6230 # Write xattrs to NTFS Alternate Data Streams:
6231 # http://en.wikipedia.org/wiki/NTFS#Alternate_data_streams_.28ADS.29
6232 assert ':' not in key
6233 assert os
.path
.exists(path
)
6235 ads_fn
= path
+ ':' + key
6237 with open(ads_fn
, 'wb') as f
:
6239 except EnvironmentError as e
:
6240 raise XAttrMetadataError(e
.errno
, e
.strerror
)
6242 user_has_setfattr
= check_executable('setfattr', ['--version'])
6243 user_has_xattr
= check_executable('xattr', ['-h'])
6245 if user_has_setfattr
or user_has_xattr
:
6247 value
= value
.decode('utf-8')
6248 if user_has_setfattr
:
6249 executable
= 'setfattr'
6250 opts
= ['-n', key
, '-v', value
]
6251 elif user_has_xattr
:
6252 executable
= 'xattr'
6253 opts
= ['-w', key
, value
]
6255 cmd
= ([encodeFilename(executable
, True)]
6256 + [encodeArgument(o
) for o
in opts
]
6257 + [encodeFilename(path
, True)])
6261 cmd
, stdout
=subprocess
.PIPE
, stderr
=subprocess
.PIPE
, stdin
=subprocess
.PIPE
)
6262 except EnvironmentError as e
:
6263 raise XAttrMetadataError(e
.errno
, e
.strerror
)
6264 stdout
, stderr
= p
.communicate_or_kill()
6265 stderr
= stderr
.decode('utf-8', 'replace')
6266 if p
.returncode
!= 0:
6267 raise XAttrMetadataError(p
.returncode
, stderr
)
6270 # On Unix, and can't find pyxattr, setfattr, or xattr.
6271 if sys
.platform
.startswith('linux'):
6272 raise XAttrUnavailableError(
6273 "Couldn't find a tool to set the xattrs. "
6274 "Install either the python 'pyxattr' or 'xattr' "
6275 "modules, or the GNU 'attr' package "
6276 "(which contains the 'setfattr' tool).")
6278 raise XAttrUnavailableError(
6279 "Couldn't find a tool to set the xattrs. "
6280 "Install either the python 'xattr' module, "
6281 "or the 'xattr' binary.")
6284 def random_birthday(year_field
, month_field
, day_field
):
6285 start_date
= datetime
.date(1950, 1, 1)
6286 end_date
= datetime
.date(1995, 12, 31)
6287 offset
= random
.randint(0, (end_date
- start_date
).days
)
6288 random_date
= start_date
+ datetime
.timedelta(offset
)
6290 year_field
: str(random_date
.year
),
6291 month_field
: str(random_date
.month
),
6292 day_field
: str(random_date
.day
),
6296 # Templates for internet shortcut files, which are plain text files.
6297 DOT_URL_LINK_TEMPLATE
= '''
6302 DOT_WEBLOC_LINK_TEMPLATE
= '''
6303 <?xml version="1.0" encoding="UTF-8"?>
6304 <!DOCTYPE plist PUBLIC "-//Apple//DTD PLIST 1.0//EN" "http://www.apple.com/DTDs/PropertyList-1.0.dtd">
6305 <plist version="1.0">
6308 \t<string>%(url)s</string>
6313 DOT_DESKTOP_LINK_TEMPLATE
= '''
6323 'url': DOT_URL_LINK_TEMPLATE
,
6324 'desktop': DOT_DESKTOP_LINK_TEMPLATE
,
6325 'webloc': DOT_WEBLOC_LINK_TEMPLATE
,
6329 def iri_to_uri(iri
):
6331 Converts an IRI (Internationalized Resource Identifier, allowing Unicode characters) to a URI (Uniform Resource Identifier, ASCII-only).
6333 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.
6336 iri_parts
= compat_urllib_parse_urlparse(iri
)
6338 if '[' in iri_parts
.netloc
:
6339 raise ValueError('IPv6 URIs are not, yet, supported.')
6340 # Querying `.netloc`, when there's only one bracket, also raises a ValueError.
6342 # 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.
6345 if iri_parts
.username
:
6346 net_location
+= compat_urllib_parse_quote(iri_parts
.username
, safe
=r
"!$%&'()*+,~")
6347 if iri_parts
.password
is not None:
6348 net_location
+= ':' + compat_urllib_parse_quote(iri_parts
.password
, safe
=r
"!$%&'()*+,~")
6351 net_location
+= iri_parts
.hostname
.encode('idna').decode('utf-8') # Punycode for Unicode hostnames.
6352 # The 'idna' encoding produces ASCII text.
6353 if iri_parts
.port
is not None and iri_parts
.port
!= 80:
6354 net_location
+= ':' + str(iri_parts
.port
)
6356 return compat_urllib_parse_urlunparse(
6360 compat_urllib_parse_quote_plus(iri_parts
.path
, safe
=r
"!$%&'()*+,/:;=@|~"),
6362 # Unsure about the `safe` argument, since this is a legacy way of handling parameters.
6363 compat_urllib_parse_quote_plus(iri_parts
.params
, safe
=r
"!$%&'()*+,/:;=@|~"),
6365 # Not totally sure about the `safe` argument, since the source does not explicitly mention the query URI component.
6366 compat_urllib_parse_quote_plus(iri_parts
.query
, safe
=r
"!$%&'()*+,/:;=?@{|}~"),
6368 compat_urllib_parse_quote_plus(iri_parts
.fragment
, safe
=r
"!#$%&'()*+,/:;=?@{|}~")))
6370 # Source for `safe` arguments: https://url.spec.whatwg.org/#percent-encoded-bytes.
6373 def to_high_limit_path(path
):
6374 if sys
.platform
in ['win32', 'cygwin']:
6375 # Work around MAX_PATH limitation on Windows. The maximum allowed length for the individual path segments may still be quite limited.
6376 return r
'\\?\ '.rstrip() + os
.path
.abspath(path
)
6381 def format_field(obj
, field
=None, template
='%s', ignore
=(None, ''), default
='', func
=None):
6383 val
= obj
if obj
is not None else default
6385 val
= obj
.get(field
, default
)
6386 if func
and val
not in ignore
:
6388 return template
% val
if val
not in ignore
else default
6391 def clean_podcast_url(url
):
6392 return re
.sub(r
'''(?x)
6396 media\.blubrry\.com| # https://create.blubrry.com/resources/podcast-media-download-statistics/getting-started/
6399 (?:dts|www)\.podtrac\.com/(?:pts/)?redirect\.[0-9a-z]{3,4}| # http://analytics.podtrac.com/how-to-measure
6402 cn\.co| # https://podcorn.com/analytics-prefix/
6403 st\.fm # https://podsights.com/docs/
6408 _HEX_TABLE
= '0123456789abcdef'
6411 def random_uuidv4():
6412 return re
.sub(r
'[xy]', lambda x
: _HEX_TABLE
[random
.randint(0, 15)], 'xxxxxxxx-xxxx-4xxx-yxxx-xxxxxxxxxxxx')
6415 def make_dir(path
, to_screen
=None):
6417 dn
= os
.path
.dirname(path
)
6418 if dn
and not os
.path
.exists(dn
):
6421 except (OSError, IOError) as err
:
6422 if callable(to_screen
) is not None:
6423 to_screen('unable to create directory ' + error_to_compat_str(err
))
6427 def get_executable_path():
6428 from zipimport
import zipimporter
6429 if hasattr(sys
, 'frozen'): # Running from PyInstaller
6430 path
= os
.path
.dirname(sys
.executable
)
6431 elif isinstance(globals().get('__loader__'), zipimporter
): # Running from ZIP
6432 path
= os
.path
.join(os
.path
.dirname(__file__
), '../..')
6434 path
= os
.path
.join(os
.path
.dirname(__file__
), '..')
6435 return os
.path
.abspath(path
)
6438 def load_plugins(name
, suffix
, namespace
):
6441 plugins_spec
= importlib
.util
.spec_from_file_location(
6442 name
, os
.path
.join(get_executable_path(), 'ytdlp_plugins', name
, '__init__.py'))
6443 plugins
= importlib
.util
.module_from_spec(plugins_spec
)
6444 sys
.modules
[plugins_spec
.name
] = plugins
6445 plugins_spec
.loader
.exec_module(plugins
)
6446 for name
in dir(plugins
):
6447 if name
in namespace
:
6449 if not name
.endswith(suffix
):
6451 klass
= getattr(plugins
, name
)
6452 classes
[name
] = namespace
[name
] = klass
6453 except FileNotFoundError
:
6459 obj
, *path_list
, default
=None, expected_type
=None, get_all
=True,
6460 casesense
=True, is_user_input
=False, traverse_string
=False):
6461 ''' Traverse nested list/dict/tuple
6462 @param path_list A list of paths which are checked one by one.
6463 Each path is a list of keys where each key is a string,
6464 a function, a tuple of strings or "...".
6465 When a fuction is given, it takes the key as argument and
6466 returns whether the key matches or not. When a tuple is given,
6467 all the keys given in the tuple are traversed, and
6468 "..." traverses all the keys in the object
6469 @param default Default value to return
6470 @param expected_type Only accept final value of this type (Can also be any callable)
6471 @param get_all Return all the values obtained from a path or only the first one
6472 @param casesense Whether to consider dictionary keys as case sensitive
6473 @param is_user_input Whether the keys are generated from user input. If True,
6474 strings are converted to int/slice if necessary
6475 @param traverse_string Whether to traverse inside strings. If True, any
6476 non-compatible object will also be converted into a string
6480 _lower
= lambda k
: (k
.lower() if isinstance(k
, str) else k
)
6481 path_list
= (map(_lower
, variadic(path
)) for path
in path_list
)
6483 def _traverse_obj(obj
, path
, _current_depth
=0):
6485 path
= tuple(variadic(path
))
6486 for i
, key
in enumerate(path
):
6489 if isinstance(key
, (list, tuple)):
6490 obj
= [_traverse_obj(obj
, sub_key
, _current_depth
) for sub_key
in key
]
6493 obj
= (obj
.values() if isinstance(obj
, dict)
6494 else obj
if isinstance(obj
, (list, tuple, LazyList
))
6495 else str(obj
) if traverse_string
else [])
6497 depth
= max(depth
, _current_depth
)
6498 return [_traverse_obj(inner_obj
, path
[i
+ 1:], _current_depth
) for inner_obj
in obj
]
6500 if isinstance(obj
, (list, tuple, LazyList
)):
6501 obj
= enumerate(obj
)
6502 elif isinstance(obj
, dict):
6505 if not traverse_string
:
6509 depth
= max(depth
, _current_depth
)
6510 return [_traverse_obj(v
, path
[i
+ 1:], _current_depth
) for k
, v
in obj
if key(k
)]
6511 elif isinstance(obj
, dict) and not (is_user_input
and key
== ':'):
6512 obj
= (obj
.get(key
) if casesense
or (key
in obj
)
6513 else next((v
for k
, v
in obj
.items() if _lower(k
) == key
), None))
6516 key
= (int_or_none(key
) if ':' not in key
6517 else slice(*map(int_or_none
, key
.split(':'))))
6518 if key
== slice(None):
6519 return _traverse_obj(obj
, (..., *path
[i
+ 1:]), _current_depth
)
6520 if not isinstance(key
, (int, slice)):
6522 if not isinstance(obj
, (list, tuple, LazyList
)):
6523 if not traverse_string
:
6532 if isinstance(expected_type
, type):
6533 type_test
= lambda val
: val
if isinstance(val
, expected_type
) else None
6534 elif expected_type
is not None:
6535 type_test
= expected_type
6537 type_test
= lambda val
: val
6539 for path
in path_list
:
6541 val
= _traverse_obj(obj
, path
)
6544 for _
in range(depth
- 1):
6545 val
= itertools
.chain
.from_iterable(v
for v
in val
if v
is not None)
6546 val
= [v
for v
in map(type_test
, val
) if v
is not None]
6548 return val
if get_all
else val
[0]
6550 val
= type_test(val
)
6557 def traverse_dict(dictn
, keys
, casesense
=True):
6558 write_string('DeprecationWarning: yt_dlp.utils.traverse_dict is deprecated '
6559 'and may be removed in a future version. Use yt_dlp.utils.traverse_obj instead')
6560 return traverse_obj(dictn
, keys
, casesense
=casesense
, is_user_input
=True, traverse_string
=True)
6563 def variadic(x
, allowed_types
=(str, bytes)):
6564 return x
if isinstance(x
, collections
.abc
.Iterable
) and not isinstance(x
, allowed_types
) else (x
,)
6567 # create a JSON Web Signature (jws) with HS256 algorithm
6568 # the resulting format is in JWS Compact Serialization
6569 # implemented following JWT https://www.rfc-editor.org/rfc/rfc7519.html
6570 # implemented following JWS https://www.rfc-editor.org/rfc/rfc7515.html
6571 def jwt_encode_hs256(payload_data
, key
, headers
={}):
6577 header_data
.update(headers
)
6578 header_b64
= base64
.b64encode(json
.dumps(header_data
).encode('utf-8'))
6579 payload_b64
= base64
.b64encode(json
.dumps(payload_data
).encode('utf-8'))
6580 h
= hmac
.new(key
.encode('utf-8'), header_b64
+ b
'.' + payload_b64
, hashlib
.sha256
)
6581 signature_b64
= base64
.b64encode(h
.digest())
6582 token
= header_b64
+ b
'.' + payload_b64
+ b
'.' + signature_b64
6586 # can be extended in future to verify the signature and parse header and return the algorithm used if it's not HS256
6587 def jwt_decode_hs256(jwt
):
6588 header_b64
, payload_b64
, signature_b64
= jwt
.split('.')
6589 payload_data
= json
.loads(base64
.urlsafe_b64decode(payload_b64
))
6593 def supports_terminal_sequences(stream
):
6594 if compat_os_name
== 'nt':
6595 from .compat
import WINDOWS_VT_MODE
# Must be imported locally
6596 if not WINDOWS_VT_MODE
or get_windows_version() < (10, 0, 10586):
6598 elif not os
.getenv('TERM'):
6601 return stream
.isatty()
6602 except BaseException
:
6606 _terminal_sequences_re
= re
.compile('\033\\[[^m]+m')
6609 def remove_terminal_sequences(string
):
6610 return _terminal_sequences_re
.sub('', string
)
6613 def number_of_digits(number
):
6614 return len('%d' % number
)
6617 def join_nonempty(*values
, delim
='-', from_dict
=None):
6618 if from_dict
is not None:
6619 values
= map(from_dict
.get
, values
)
6620 return delim
.join(map(str, filter(None, values
)))
projects / yt-dlp.git / blob