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 def extract_attributes(html_element):
2010 """Given a string for an HTML element such as
2012 a="foo" B="bar" c="&98;az" d=boz
2013 empty= noval entity="&"
2016 Decode and return a dictionary of attributes.
2018 'a
': 'foo
', 'b
': 'bar
', c: 'baz
', d: 'boz
',
2019 'empty
': '', 'noval
': None, 'entity
': '&',
2020 'sq
': '"', 'dq': '\''
2022 NB HTMLParser is stricter in Python 2.6 & 3.2 than in later versions,
2023 but the cases in the unit test will work for all of 2.6, 2.7, 3.2-3.5.
2025 parser = HTMLAttributeParser()
2027 parser.feed(html_element)
2029 # Older Python may throw HTMLParseError in case of malformed HTML
2030 except compat_HTMLParseError:
2035 def clean_html(html):
2036 """Clean an HTML snippet into a readable string"""
2038 if html is None: # Convenience for sanitizing descriptions etc.
2042 html = html.replace('\n', ' ')
2043 html = re.sub(r'(?u)\s*<\s*br\s*/?\s*>\s*', '\n', html)
2044 html = re.sub(r'(?u)<\s*/\s*p\s*>\s*<\s*p[^>]*>', '\n', html)
2046 html = re.sub('<.*?>', '', html)
2047 # Replace html entities
2048 html = unescapeHTML(html)
2052 def sanitize_open(filename, open_mode):
2053 """Try to open the given filename, and slightly tweak it if this fails.
2055 Attempts to open the given filename. If this fails, it tries to change
2056 the filename slightly, step by step, until it's either able to open it
2057 or it fails and raises a final exception, like the standard open()
2060 It returns the tuple (stream, definitive_file_name).
2064 if sys.platform == 'win32':
2066 msvcrt.setmode(sys.stdout.fileno(), os.O_BINARY)
2067 return (sys.stdout.buffer if hasattr(sys.stdout, 'buffer') else sys.stdout, filename)
2068 stream = open(encodeFilename(filename), open_mode)
2069 return (stream, filename)
2070 except (IOError, OSError) as err:
2071 if err.errno in (errno.EACCES,):
2074 # In case of error, try to remove win32 forbidden chars
2075 alt_filename = sanitize_path(filename)
2076 if alt_filename == filename:
2079 # An exception here should be caught in the caller
2080 stream = open(encodeFilename(alt_filename), open_mode)
2081 return (stream, alt_filename)
2084 def timeconvert(timestr):
2085 """Convert RFC 2822 defined time string into system timestamp"""
2087 timetuple = email.utils.parsedate_tz(timestr)
2088 if timetuple is not None:
2089 timestamp = email.utils.mktime_tz(timetuple)
2093 def sanitize_filename(s, restricted=False, is_id=False):
2094 """Sanitizes a string so it could be used as part of a filename.
2095 If restricted is set, use a stricter subset of allowed characters.
2096 Set is_id if this is not an arbitrary string, but an ID that should be kept
2099 def replace_insane(char):
2100 if restricted and char in ACCENT_CHARS:
2101 return ACCENT_CHARS[char]
2102 elif not restricted and char == '\n':
2104 elif char == '?' or ord(char) < 32 or ord(char) == 127:
2107 return '' if restricted else '\''
2109 return '_
-' if restricted else ' -'
2110 elif char in '\\/|
*<>':
2112 if restricted and (char in '!&\'()[]{}$
;`^
,#' or char.isspace()):
2114 if restricted
and ord(char
) > 127:
2121 s
= re
.sub(r
'[0-9]+(?::[0-9]+)+', lambda m
: m
.group(0).replace(':', '_'), s
)
2122 result
= ''.join(map(replace_insane
, s
))
2124 while '__' in result
:
2125 result
= result
.replace('__', '_')
2126 result
= result
.strip('_')
2127 # Common case of "Foreign band name - English song title"
2128 if restricted
and result
.startswith('-_'):
2130 if result
.startswith('-'):
2131 result
= '_' + result
[len('-'):]
2132 result
= result
.lstrip('.')
2138 def sanitize_path(s
, force
=False):
2139 """Sanitizes and normalizes path on Windows"""
2140 if sys
.platform
== 'win32':
2142 drive_or_unc
, _
= os
.path
.splitdrive(s
)
2143 if sys
.version_info
< (2, 7) and not drive_or_unc
:
2144 drive_or_unc
, _
= os
.path
.splitunc(s
)
2150 norm_path
= os
.path
.normpath(remove_start(s
, drive_or_unc
)).split(os
.path
.sep
)
2154 path_part
if path_part
in ['.', '..'] else re
.sub(r
'(?:[/<>:"\|\\?\*]|[\s.]$)', '#', path_part
)
2155 for path_part
in norm_path
]
2157 sanitized_path
.insert(0, drive_or_unc
+ os
.path
.sep
)
2158 elif force
and s
[0] == os
.path
.sep
:
2159 sanitized_path
.insert(0, os
.path
.sep
)
2160 return os
.path
.join(*sanitized_path
)
2163 def sanitize_url(url
):
2164 # Prepend protocol-less URLs with `http:` scheme in order to mitigate
2165 # the number of unwanted failures due to missing protocol
2166 if url
.startswith('//'):
2167 return 'http:%s' % url
2168 # Fix some common typos seen so far
2170 # https://github.com/ytdl-org/youtube-dl/issues/15649
2171 (r
'^httpss://', r
'https://'),
2172 # https://bx1.be/lives/direct-tv/
2173 (r
'^rmtp([es]?)://', r
'rtmp\1://'),
2175 for mistake
, fixup
in COMMON_TYPOS
:
2176 if re
.match(mistake
, url
):
2177 return re
.sub(mistake
, fixup
, url
)
2181 def extract_basic_auth(url
):
2182 parts
= compat_urlparse
.urlsplit(url
)
2183 if parts
.username
is None:
2185 url
= compat_urlparse
.urlunsplit(parts
._replace
(netloc
=(
2186 parts
.hostname
if parts
.port
is None
2187 else '%s:%d' % (parts
.hostname
, parts
.port
))))
2188 auth_payload
= base64
.b64encode(
2189 ('%s:%s' % (parts
.username
, parts
.password
or '')).encode('utf-8'))
2190 return url
, 'Basic ' + auth_payload
.decode('utf-8')
2193 def sanitized_Request(url
, *args
, **kwargs
):
2194 url
, auth_header
= extract_basic_auth(escape_url(sanitize_url(url
)))
2195 if auth_header
is not None:
2196 headers
= args
[1] if len(args
) >= 2 else kwargs
.setdefault('headers', {})
2197 headers
['Authorization'] = auth_header
2198 return compat_urllib_request
.Request(url
, *args
, **kwargs
)
2202 """Expand shell variables and ~"""
2203 return os
.path
.expandvars(compat_expanduser(s
))
2206 def orderedSet(iterable
):
2207 """ Remove all duplicates from the input iterable """
2215 def _htmlentity_transform(entity_with_semicolon
):
2216 """Transforms an HTML entity to a character."""
2217 entity
= entity_with_semicolon
[:-1]
2219 # Known non-numeric HTML entity
2220 if entity
in compat_html_entities
.name2codepoint
:
2221 return compat_chr(compat_html_entities
.name2codepoint
[entity
])
2223 # TODO: HTML5 allows entities without a semicolon. For example,
2224 # 'Éric' should be decoded as 'Éric'.
2225 if entity_with_semicolon
in compat_html_entities_html5
:
2226 return compat_html_entities_html5
[entity_with_semicolon
]
2228 mobj
= re
.match(r
'#(x[0-9a-fA-F]+|[0-9]+)', entity
)
2229 if mobj
is not None:
2230 numstr
= mobj
.group(1)
2231 if numstr
.startswith('x'):
2233 numstr
= '0%s' % numstr
2236 # See https://github.com/ytdl-org/youtube-dl/issues/7518
2238 return compat_chr(int(numstr
, base
))
2242 # Unknown entity in name, return its literal representation
2243 return '&%s;' % entity
2246 def unescapeHTML(s
):
2249 assert type(s
) == compat_str
2252 r
'&([^&;]+;)', lambda m
: _htmlentity_transform(m
.group(1)), s
)
2255 def escapeHTML(text
):
2258 .replace('&', '&')
2259 .replace('<', '<')
2260 .replace('>', '>')
2261 .replace('"', '"')
2262 .replace("'", ''')
2266 def process_communicate_or_kill(p
, *args
, **kwargs
):
2268 return p
.communicate(*args
, **kwargs
)
2269 except BaseException
: # Including KeyboardInterrupt
2275 def get_subprocess_encoding():
2276 if sys
.platform
== 'win32' and sys
.getwindowsversion()[0] >= 5:
2277 # For subprocess calls, encode with locale encoding
2278 # Refer to http://stackoverflow.com/a/9951851/35070
2279 encoding
= preferredencoding()
2281 encoding
= sys
.getfilesystemencoding()
2282 if encoding
is None:
2287 def encodeFilename(s
, for_subprocess
=False):
2289 @param s The name of the file
2292 assert type(s
) == compat_str
2294 # Python 3 has a Unicode API
2295 if sys
.version_info
>= (3, 0):
2298 # Pass '' directly to use Unicode APIs on Windows 2000 and up
2299 # (Detecting Windows NT 4 is tricky because 'major >= 4' would
2300 # match Windows 9x series as well. Besides, NT 4 is obsolete.)
2301 if not for_subprocess
and sys
.platform
== 'win32' and sys
.getwindowsversion()[0] >= 5:
2304 # Jython assumes filenames are Unicode strings though reported as Python 2.x compatible
2305 if sys
.platform
.startswith('java'):
2308 return s
.encode(get_subprocess_encoding(), 'ignore')
2311 def decodeFilename(b
, for_subprocess
=False):
2313 if sys
.version_info
>= (3, 0):
2316 if not isinstance(b
, bytes):
2319 return b
.decode(get_subprocess_encoding(), 'ignore')
2322 def encodeArgument(s
):
2323 if not isinstance(s
, compat_str
):
2324 # Legacy code that uses byte strings
2325 # Uncomment the following line after fixing all post processors
2326 # assert False, 'Internal error: %r should be of type %r, is %r' % (s, compat_str, type(s))
2327 s
= s
.decode('ascii')
2328 return encodeFilename(s
, True)
2331 def decodeArgument(b
):
2332 return decodeFilename(b
, True)
2335 def decodeOption(optval
):
2338 if isinstance(optval
, bytes):
2339 optval
= optval
.decode(preferredencoding())
2341 assert isinstance(optval
, compat_str
)
2345 def formatSeconds(secs
, delim
=':', msec
=False):
2347 ret
= '%d%s%02d%s%02d' % (secs
// 3600, delim
, (secs
% 3600) // 60, delim
, secs
% 60)
2349 ret
= '%d%s%02d' % (secs
// 60, delim
, secs
% 60)
2352 return '%s.%03d' % (ret
, secs
% 1) if msec
else ret
2355 def _ssl_load_windows_store_certs(ssl_context
, storename
):
2356 # Code adapted from _load_windows_store_certs in https://github.com/python/cpython/blob/main/Lib/ssl.py
2358 certs
= [cert
for cert
, encoding
, trust
in ssl
.enum_certificates(storename
)
2359 if encoding
== 'x509_asn' and (
2360 trust
is True or ssl
.Purpose
.SERVER_AUTH
.oid
in trust
)]
2361 except PermissionError
:
2365 ssl_context
.load_verify_locations(cadata
=cert
)
2366 except ssl
.SSLError
:
2370 def make_HTTPS_handler(params
, **kwargs
):
2371 opts_check_certificate
= not params
.get('nocheckcertificate')
2372 context
= ssl
.SSLContext(ssl
.PROTOCOL_TLS_CLIENT
)
2373 context
.check_hostname
= opts_check_certificate
2374 context
.verify_mode
= ssl
.CERT_REQUIRED
if opts_check_certificate
else ssl
.CERT_NONE
2375 if opts_check_certificate
:
2376 # Work around the issue in load_default_certs when there are bad certificates. See:
2377 # https://github.com/yt-dlp/yt-dlp/issues/1060,
2378 # https://bugs.python.org/issue35665, https://bugs.python.org/issue4531
2379 if sys
.platform
== 'win32':
2380 for storename
in ('CA', 'ROOT'):
2381 _ssl_load_windows_store_certs(context
, storename
)
2382 context
.set_default_verify_paths()
2383 return YoutubeDLHTTPSHandler(params
, context
=context
, **kwargs
)
2386 def bug_reports_message(before
=';'):
2387 if ytdl_is_updateable():
2388 update_cmd
= 'type yt-dlp -U to update'
2390 update_cmd
= 'see https://github.com/yt-dlp/yt-dlp on how to update'
2391 msg
= 'please report this issue on https://github.com/yt-dlp/yt-dlp .'
2392 msg
+= ' Make sure you are using the latest version; %s.' % update_cmd
2393 msg
+= ' Be sure to call yt-dlp with the --verbose flag and include its complete output.'
2395 before
= before
.rstrip()
2396 if not before
or before
.endswith(('.', '!', '?')):
2397 msg
= msg
[0].title() + msg
[1:]
2399 return (before
+ ' ' if before
else '') + msg
2402 class YoutubeDLError(Exception):
2403 """Base exception for YoutubeDL errors."""
2407 network_exceptions
= [compat_urllib_error
.URLError
, compat_http_client
.HTTPException
, socket
.error
]
2408 if hasattr(ssl
, 'CertificateError'):
2409 network_exceptions
.append(ssl
.CertificateError
)
2410 network_exceptions
= tuple(network_exceptions
)
2413 class ExtractorError(YoutubeDLError
):
2414 """Error during info extraction."""
2416 def __init__(self
, msg
, tb
=None, expected
=False, cause
=None, video_id
=None, ie
=None):
2417 """ tb, if given, is the original traceback (so that it can be printed out).
2418 If expected is set, this is a normal error message and most likely not a bug in yt-dlp.
2420 if sys
.exc_info()[0] in network_exceptions
:
2425 self
.expected
= expected
2427 self
.video_id
= video_id
2429 self
.exc_info
= sys
.exc_info() # preserve original exception
2431 super(ExtractorError
, self
).__init
__(''.join((
2432 format_field(ie
, template
='[%s] '),
2433 format_field(video_id
, template
='%s: '),
2435 format_field(cause
, template
=' (caused by %r)'),
2436 '' if expected
else bug_reports_message())))
2438 def format_traceback(self
):
2439 if self
.traceback
is None:
2441 return ''.join(traceback
.format_tb(self
.traceback
))
2444 class UnsupportedError(ExtractorError
):
2445 def __init__(self
, url
):
2446 super(UnsupportedError
, self
).__init
__(
2447 'Unsupported URL: %s' % url
, expected
=True)
2451 class RegexNotFoundError(ExtractorError
):
2452 """Error when a regex didn't match"""
2456 class GeoRestrictedError(ExtractorError
):
2457 """Geographic restriction Error exception.
2459 This exception may be thrown when a video is not available from your
2460 geographic location due to geographic restrictions imposed by a website.
2463 def __init__(self
, msg
, countries
=None):
2464 super(GeoRestrictedError
, self
).__init
__(msg
, expected
=True)
2466 self
.countries
= countries
2469 class DownloadError(YoutubeDLError
):
2470 """Download Error exception.
2472 This exception may be thrown by FileDownloader objects if they are not
2473 configured to continue on errors. They will contain the appropriate
2477 def __init__(self
, msg
, exc_info
=None):
2478 """ exc_info, if given, is the original exception that caused the trouble (as returned by sys.exc_info()). """
2479 super(DownloadError
, self
).__init
__(msg
)
2480 self
.exc_info
= exc_info
2483 class EntryNotInPlaylist(YoutubeDLError
):
2484 """Entry not in playlist exception.
2486 This exception will be thrown by YoutubeDL when a requested entry
2487 is not found in the playlist info_dict
2492 class SameFileError(YoutubeDLError
):
2493 """Same File exception.
2495 This exception will be thrown by FileDownloader objects if they detect
2496 multiple files would have to be downloaded to the same file on disk.
2501 class PostProcessingError(YoutubeDLError
):
2502 """Post Processing exception.
2504 This exception may be raised by PostProcessor's .run() method to
2505 indicate an error in the postprocessing task.
2508 def __init__(self
, msg
):
2509 super(PostProcessingError
, self
).__init
__(msg
)
2513 class ExistingVideoReached(YoutubeDLError
):
2514 """ --max-downloads limit has been reached. """
2518 class RejectedVideoReached(YoutubeDLError
):
2519 """ --max-downloads limit has been reached. """
2523 class ThrottledDownload(YoutubeDLError
):
2524 """ Download speed below --throttled-rate. """
2528 class MaxDownloadsReached(YoutubeDLError
):
2529 """ --max-downloads limit has been reached. """
2533 class UnavailableVideoError(YoutubeDLError
):
2534 """Unavailable Format exception.
2536 This exception will be thrown when a video is requested
2537 in a format that is not available for that video.
2542 class ContentTooShortError(YoutubeDLError
):
2543 """Content Too Short exception.
2545 This exception may be raised by FileDownloader objects when a file they
2546 download is too small for what the server announced first, indicating
2547 the connection was probably interrupted.
2550 def __init__(self
, downloaded
, expected
):
2551 super(ContentTooShortError
, self
).__init
__(
2552 'Downloaded {0} bytes, expected {1} bytes'.format(downloaded
, expected
)
2555 self
.downloaded
= downloaded
2556 self
.expected
= expected
2559 class XAttrMetadataError(YoutubeDLError
):
2560 def __init__(self
, code
=None, msg
='Unknown error'):
2561 super(XAttrMetadataError
, self
).__init
__(msg
)
2565 # Parsing code and msg
2566 if (self
.code
in (errno
.ENOSPC
, errno
.EDQUOT
)
2567 or 'No space left' in self
.msg
or 'Disk quota exceeded' in self
.msg
):
2568 self
.reason
= 'NO_SPACE'
2569 elif self
.code
== errno
.E2BIG
or 'Argument list too long' in self
.msg
:
2570 self
.reason
= 'VALUE_TOO_LONG'
2572 self
.reason
= 'NOT_SUPPORTED'
2575 class XAttrUnavailableError(YoutubeDLError
):
2579 def _create_http_connection(ydl_handler
, http_class
, is_https
, *args
, **kwargs
):
2580 # Working around python 2 bug (see http://bugs.python.org/issue17849) by limiting
2581 # expected HTTP responses to meet HTTP/1.0 or later (see also
2582 # https://github.com/ytdl-org/youtube-dl/issues/6727)
2583 if sys
.version_info
< (3, 0):
2584 kwargs
['strict'] = True
2585 hc
= http_class(*args
, **compat_kwargs(kwargs
))
2586 source_address
= ydl_handler
._params
.get('source_address')
2588 if source_address
is not None:
2589 # This is to workaround _create_connection() from socket where it will try all
2590 # address data from getaddrinfo() including IPv6. This filters the result from
2591 # getaddrinfo() based on the source_address value.
2592 # This is based on the cpython socket.create_connection() function.
2593 # https://github.com/python/cpython/blob/master/Lib/socket.py#L691
2594 def _create_connection(address
, timeout
=socket
._GLOBAL
_DEFAULT
_TIMEOUT
, source_address
=None):
2595 host
, port
= address
2597 addrs
= socket
.getaddrinfo(host
, port
, 0, socket
.SOCK_STREAM
)
2598 af
= socket
.AF_INET
if '.' in source_address
[0] else socket
.AF_INET6
2599 ip_addrs
= [addr
for addr
in addrs
if addr
[0] == af
]
2600 if addrs
and not ip_addrs
:
2601 ip_version
= 'v4' if af
== socket
.AF_INET
else 'v6'
2603 "No remote IP%s addresses available for connect, can't use '%s' as source address"
2604 % (ip_version
, source_address
[0]))
2605 for res
in ip_addrs
:
2606 af
, socktype
, proto
, canonname
, sa
= res
2609 sock
= socket
.socket(af
, socktype
, proto
)
2610 if timeout
is not socket
._GLOBAL
_DEFAULT
_TIMEOUT
:
2611 sock
.settimeout(timeout
)
2612 sock
.bind(source_address
)
2614 err
= None # Explicitly break reference cycle
2616 except socket
.error
as _
:
2618 if sock
is not None:
2623 raise socket
.error('getaddrinfo returns an empty list')
2624 if hasattr(hc
, '_create_connection'):
2625 hc
._create
_connection
= _create_connection
2626 sa
= (source_address
, 0)
2627 if hasattr(hc
, 'source_address'): # Python 2.7+
2628 hc
.source_address
= sa
2630 def _hc_connect(self
, *args
, **kwargs
):
2631 sock
= _create_connection(
2632 (self
.host
, self
.port
), self
.timeout
, sa
)
2634 self
.sock
= ssl
.wrap_socket(
2635 sock
, self
.key_file
, self
.cert_file
,
2636 ssl_version
=ssl
.PROTOCOL_TLSv1
)
2639 hc
.connect
= functools
.partial(_hc_connect
, hc
)
2644 def handle_youtubedl_headers(headers
):
2645 filtered_headers
= headers
2647 if 'Youtubedl-no-compression' in filtered_headers
:
2648 filtered_headers
= dict((k
, v
) for k
, v
in filtered_headers
.items() if k
.lower() != 'accept-encoding')
2649 del filtered_headers
['Youtubedl-no-compression']
2651 return filtered_headers
2654 class YoutubeDLHandler(compat_urllib_request
.HTTPHandler
):
2655 """Handler for HTTP requests and responses.
2657 This class, when installed with an OpenerDirector, automatically adds
2658 the standard headers to every HTTP request and handles gzipped and
2659 deflated responses from web servers. If compression is to be avoided in
2660 a particular request, the original request in the program code only has
2661 to include the HTTP header "Youtubedl-no-compression", which will be
2662 removed before making the real request.
2664 Part of this code was copied from:
2666 http://techknack.net/python-urllib2-handlers/
2668 Andrew Rowls, the author of that code, agreed to release it to the
2672 def __init__(self
, params
, *args
, **kwargs
):
2673 compat_urllib_request
.HTTPHandler
.__init
__(self
, *args
, **kwargs
)
2674 self
._params
= params
2676 def http_open(self
, req
):
2677 conn_class
= compat_http_client
.HTTPConnection
2679 socks_proxy
= req
.headers
.get('Ytdl-socks-proxy')
2681 conn_class
= make_socks_conn_class(conn_class
, socks_proxy
)
2682 del req
.headers
['Ytdl-socks-proxy']
2684 return self
.do_open(functools
.partial(
2685 _create_http_connection
, self
, conn_class
, False),
2693 return zlib
.decompress(data
, -zlib
.MAX_WBITS
)
2695 return zlib
.decompress(data
)
2697 def http_request(self
, req
):
2698 # According to RFC 3986, URLs can not contain non-ASCII characters, however this is not
2699 # always respected by websites, some tend to give out URLs with non percent-encoded
2700 # non-ASCII characters (see telemb.py, ard.py [#3412])
2701 # urllib chokes on URLs with non-ASCII characters (see http://bugs.python.org/issue3991)
2702 # To work around aforementioned issue we will replace request's original URL with
2703 # percent-encoded one
2704 # Since redirects are also affected (e.g. http://www.southpark.de/alle-episoden/s18e09)
2705 # the code of this workaround has been moved here from YoutubeDL.urlopen()
2706 url
= req
.get_full_url()
2707 url_escaped
= escape_url(url
)
2709 # Substitute URL if any change after escaping
2710 if url
!= url_escaped
:
2711 req
= update_Request(req
, url
=url_escaped
)
2713 for h
, v
in std_headers
.items():
2714 # Capitalize is needed because of Python bug 2275: http://bugs.python.org/issue2275
2715 # The dict keys are capitalized because of this bug by urllib
2716 if h
.capitalize() not in req
.headers
:
2717 req
.add_header(h
, v
)
2719 req
.headers
= handle_youtubedl_headers(req
.headers
)
2721 if sys
.version_info
< (2, 7) and '#' in req
.get_full_url():
2722 # Python 2.6 is brain-dead when it comes to fragments
2723 req
._Request
__original
= req
._Request
__original
.partition('#')[0]
2724 req
._Request
__r
_type
= req
._Request
__r
_type
.partition('#')[0]
2728 def http_response(self
, req
, resp
):
2731 if resp
.headers
.get('Content-encoding', '') == 'gzip':
2732 content
= resp
.read()
2733 gz
= gzip
.GzipFile(fileobj
=io
.BytesIO(content
), mode
='rb')
2735 uncompressed
= io
.BytesIO(gz
.read())
2736 except IOError as original_ioerror
:
2737 # There may be junk add the end of the file
2738 # See http://stackoverflow.com/q/4928560/35070 for details
2739 for i
in range(1, 1024):
2741 gz
= gzip
.GzipFile(fileobj
=io
.BytesIO(content
[:-i
]), mode
='rb')
2742 uncompressed
= io
.BytesIO(gz
.read())
2747 raise original_ioerror
2748 resp
= compat_urllib_request
.addinfourl(uncompressed
, old_resp
.headers
, old_resp
.url
, old_resp
.code
)
2749 resp
.msg
= old_resp
.msg
2750 del resp
.headers
['Content-encoding']
2752 if resp
.headers
.get('Content-encoding', '') == 'deflate':
2753 gz
= io
.BytesIO(self
.deflate(resp
.read()))
2754 resp
= compat_urllib_request
.addinfourl(gz
, old_resp
.headers
, old_resp
.url
, old_resp
.code
)
2755 resp
.msg
= old_resp
.msg
2756 del resp
.headers
['Content-encoding']
2757 # Percent-encode redirect URL of Location HTTP header to satisfy RFC 3986 (see
2758 # https://github.com/ytdl-org/youtube-dl/issues/6457).
2759 if 300 <= resp
.code
< 400:
2760 location
= resp
.headers
.get('Location')
2762 # As of RFC 2616 default charset is iso-8859-1 that is respected by python 3
2763 if sys
.version_info
>= (3, 0):
2764 location
= location
.encode('iso-8859-1').decode('utf-8')
2766 location
= location
.decode('utf-8')
2767 location_escaped
= escape_url(location
)
2768 if location
!= location_escaped
:
2769 del resp
.headers
['Location']
2770 if sys
.version_info
< (3, 0):
2771 location_escaped
= location_escaped
.encode('utf-8')
2772 resp
.headers
['Location'] = location_escaped
2775 https_request
= http_request
2776 https_response
= http_response
2779 def make_socks_conn_class(base_class
, socks_proxy
):
2780 assert issubclass(base_class
, (
2781 compat_http_client
.HTTPConnection
, compat_http_client
.HTTPSConnection
))
2783 url_components
= compat_urlparse
.urlparse(socks_proxy
)
2784 if url_components
.scheme
.lower() == 'socks5':
2785 socks_type
= ProxyType
.SOCKS5
2786 elif url_components
.scheme
.lower() in ('socks', 'socks4'):
2787 socks_type
= ProxyType
.SOCKS4
2788 elif url_components
.scheme
.lower() == 'socks4a':
2789 socks_type
= ProxyType
.SOCKS4A
2791 def unquote_if_non_empty(s
):
2794 return compat_urllib_parse_unquote_plus(s
)
2798 url_components
.hostname
, url_components
.port
or 1080,
2800 unquote_if_non_empty(url_components
.username
),
2801 unquote_if_non_empty(url_components
.password
),
2804 class SocksConnection(base_class
):
2806 self
.sock
= sockssocket()
2807 self
.sock
.setproxy(*proxy_args
)
2808 if type(self
.timeout
) in (int, float):
2809 self
.sock
.settimeout(self
.timeout
)
2810 self
.sock
.connect((self
.host
, self
.port
))
2812 if isinstance(self
, compat_http_client
.HTTPSConnection
):
2813 if hasattr(self
, '_context'): # Python > 2.6
2814 self
.sock
= self
._context
.wrap_socket(
2815 self
.sock
, server_hostname
=self
.host
)
2817 self
.sock
= ssl
.wrap_socket(self
.sock
)
2819 return SocksConnection
2822 class YoutubeDLHTTPSHandler(compat_urllib_request
.HTTPSHandler
):
2823 def __init__(self
, params
, https_conn_class
=None, *args
, **kwargs
):
2824 compat_urllib_request
.HTTPSHandler
.__init
__(self
, *args
, **kwargs
)
2825 self
._https
_conn
_class
= https_conn_class
or compat_http_client
.HTTPSConnection
2826 self
._params
= params
2828 def https_open(self
, req
):
2830 conn_class
= self
._https
_conn
_class
2832 if hasattr(self
, '_context'): # python > 2.6
2833 kwargs
['context'] = self
._context
2834 if hasattr(self
, '_check_hostname'): # python 3.x
2835 kwargs
['check_hostname'] = self
._check
_hostname
2837 socks_proxy
= req
.headers
.get('Ytdl-socks-proxy')
2839 conn_class
= make_socks_conn_class(conn_class
, socks_proxy
)
2840 del req
.headers
['Ytdl-socks-proxy']
2842 return self
.do_open(functools
.partial(
2843 _create_http_connection
, self
, conn_class
, True),
2847 class YoutubeDLCookieJar(compat_cookiejar
.MozillaCookieJar
):
2849 See [1] for cookie file format.
2851 1. https://curl.haxx.se/docs/http-cookies.html
2853 _HTTPONLY_PREFIX
= '#HttpOnly_'
2855 _HEADER
= '''# Netscape HTTP Cookie File
2856 # This file is generated by yt-dlp. Do not edit.
2859 _CookieFileEntry
= collections
.namedtuple(
2861 ('domain_name', 'include_subdomains', 'path', 'https_only', 'expires_at', 'name', 'value'))
2863 def save(self
, filename
=None, ignore_discard
=False, ignore_expires
=False):
2865 Save cookies to a file.
2867 Most of the code is taken from CPython 3.8 and slightly adapted
2868 to support cookie files with UTF-8 in both python 2 and 3.
2870 if filename
is None:
2871 if self
.filename
is not None:
2872 filename
= self
.filename
2874 raise ValueError(compat_cookiejar
.MISSING_FILENAME_TEXT
)
2876 # Store session cookies with `expires` set to 0 instead of an empty
2879 if cookie
.expires
is None:
2882 with io
.open(filename
, 'w', encoding
='utf-8') as f
:
2883 f
.write(self
._HEADER
)
2886 if not ignore_discard
and cookie
.discard
:
2888 if not ignore_expires
and cookie
.is_expired(now
):
2894 if cookie
.domain
.startswith('.'):
2895 initial_dot
= 'TRUE'
2897 initial_dot
= 'FALSE'
2898 if cookie
.expires
is not None:
2899 expires
= compat_str(cookie
.expires
)
2902 if cookie
.value
is None:
2903 # cookies.txt regards 'Set-Cookie: foo' as a cookie
2904 # with no name, whereas http.cookiejar regards it as a
2905 # cookie with no value.
2910 value
= cookie
.value
2912 '\t'.join([cookie
.domain
, initial_dot
, cookie
.path
,
2913 secure
, expires
, name
, value
]) + '\n')
2915 def load(self
, filename
=None, ignore_discard
=False, ignore_expires
=False):
2916 """Load cookies from a file."""
2917 if filename
is None:
2918 if self
.filename
is not None:
2919 filename
= self
.filename
2921 raise ValueError(compat_cookiejar
.MISSING_FILENAME_TEXT
)
2923 def prepare_line(line
):
2924 if line
.startswith(self
._HTTPONLY
_PREFIX
):
2925 line
= line
[len(self
._HTTPONLY
_PREFIX
):]
2926 # comments and empty lines are fine
2927 if line
.startswith('#') or not line
.strip():
2929 cookie_list
= line
.split('\t')
2930 if len(cookie_list
) != self
._ENTRY
_LEN
:
2931 raise compat_cookiejar
.LoadError('invalid length %d' % len(cookie_list
))
2932 cookie
= self
._CookieFileEntry
(*cookie_list
)
2933 if cookie
.expires_at
and not cookie
.expires_at
.isdigit():
2934 raise compat_cookiejar
.LoadError('invalid expires at %s' % cookie
.expires_at
)
2938 with io
.open(filename
, encoding
='utf-8') as f
:
2941 cf
.write(prepare_line(line
))
2942 except compat_cookiejar
.LoadError
as e
:
2944 'WARNING: skipping cookie file entry due to %s: %r\n'
2945 % (e
, line
), sys
.stderr
)
2948 self
._really
_load
(cf
, filename
, ignore_discard
, ignore_expires
)
2949 # Session cookies are denoted by either `expires` field set to
2950 # an empty string or 0. MozillaCookieJar only recognizes the former
2951 # (see [1]). So we need force the latter to be recognized as session
2952 # cookies on our own.
2953 # Session cookies may be important for cookies-based authentication,
2954 # e.g. usually, when user does not check 'Remember me' check box while
2955 # logging in on a site, some important cookies are stored as session
2956 # cookies so that not recognizing them will result in failed login.
2957 # 1. https://bugs.python.org/issue17164
2959 # Treat `expires=0` cookies as session cookies
2960 if cookie
.expires
== 0:
2961 cookie
.expires
= None
2962 cookie
.discard
= True
2965 class YoutubeDLCookieProcessor(compat_urllib_request
.HTTPCookieProcessor
):
2966 def __init__(self
, cookiejar
=None):
2967 compat_urllib_request
.HTTPCookieProcessor
.__init
__(self
, cookiejar
)
2969 def http_response(self
, request
, response
):
2970 # Python 2 will choke on next HTTP request in row if there are non-ASCII
2971 # characters in Set-Cookie HTTP header of last response (see
2972 # https://github.com/ytdl-org/youtube-dl/issues/6769).
2973 # In order to at least prevent crashing we will percent encode Set-Cookie
2974 # header before HTTPCookieProcessor starts processing it.
2975 # if sys.version_info < (3, 0) and response.headers:
2976 # for set_cookie_header in ('Set-Cookie', 'Set-Cookie2'):
2977 # set_cookie = response.headers.get(set_cookie_header)
2979 # set_cookie_escaped = compat_urllib_parse.quote(set_cookie, b"%/;:@&=+$,!~*'()?#[] ")
2980 # if set_cookie != set_cookie_escaped:
2981 # del response.headers[set_cookie_header]
2982 # response.headers[set_cookie_header] = set_cookie_escaped
2983 return compat_urllib_request
.HTTPCookieProcessor
.http_response(self
, request
, response
)
2985 https_request
= compat_urllib_request
.HTTPCookieProcessor
.http_request
2986 https_response
= http_response
2989 class YoutubeDLRedirectHandler(compat_urllib_request
.HTTPRedirectHandler
):
2990 """YoutubeDL redirect handler
2992 The code is based on HTTPRedirectHandler implementation from CPython [1].
2994 This redirect handler solves two issues:
2995 - ensures redirect URL is always unicode under python 2
2996 - introduces support for experimental HTTP response status code
2997 308 Permanent Redirect [2] used by some sites [3]
2999 1. https://github.com/python/cpython/blob/master/Lib/urllib/request.py
3000 2. https://developer.mozilla.org/en-US/docs/Web/HTTP/Status/308
3001 3. https://github.com/ytdl-org/youtube-dl/issues/28768
3004 http_error_301
= http_error_303
= http_error_307
= http_error_308
= compat_urllib_request
.HTTPRedirectHandler
.http_error_302
3006 def redirect_request(self
, req
, fp
, code
, msg
, headers
, newurl
):
3007 """Return a Request or None in response to a redirect.
3009 This is called by the http_error_30x methods when a
3010 redirection response is received. If a redirection should
3011 take place, return a new Request to allow http_error_30x to
3012 perform the redirect. Otherwise, raise HTTPError if no-one
3013 else should try to handle this url. Return None if you can't
3014 but another Handler might.
3016 m
= req
.get_method()
3017 if (not (code
in (301, 302, 303, 307, 308) and m
in ("GET", "HEAD")
3018 or code
in (301, 302, 303) and m
== "POST")):
3019 raise compat_HTTPError(req
.full_url
, code
, msg
, headers
, fp
)
3020 # Strictly (according to RFC 2616), 301 or 302 in response to
3021 # a POST MUST NOT cause a redirection without confirmation
3022 # from the user (of urllib.request, in this case). In practice,
3023 # essentially all clients do redirect in this case, so we do
3026 # On python 2 urlh.geturl() may sometimes return redirect URL
3027 # as byte string instead of unicode. This workaround allows
3028 # to force it always return unicode.
3029 if sys
.version_info
[0] < 3:
3030 newurl
= compat_str(newurl
)
3032 # Be conciliant with URIs containing a space. This is mainly
3033 # redundant with the more complete encoding done in http_error_302(),
3034 # but it is kept for compatibility with other callers.
3035 newurl
= newurl
.replace(' ', '%20')
3037 CONTENT_HEADERS
= ("content-length", "content-type")
3038 # NB: don't use dict comprehension for python 2.6 compatibility
3039 newheaders
= dict((k
, v
) for k
, v
in req
.headers
.items()
3040 if k
.lower() not in CONTENT_HEADERS
)
3041 return compat_urllib_request
.Request(
3042 newurl
, headers
=newheaders
, origin_req_host
=req
.origin_req_host
,
3046 def extract_timezone(date_str
):
3049 ^.{8,}? # >=8 char non-TZ prefix, if present
3050 (?P<tz>Z| # just the UTC Z, or
3051 (?:(?<=.\b\d{4}|\b\d{2}:\d\d)| # preceded by 4 digits or hh:mm or
3052 (?<!.\b[a-zA-Z]{3}|[a-zA-Z]{4}|..\b\d\d)) # not preceded by 3 alpha word or >= 4 alpha or 2 digits
3053 [ ]? # optional space
3054 (?P<sign>\+|-) # +/-
3055 (?P<hours>[0-9]{2}):?(?P<minutes>[0-9]{2}) # hh[:]mm
3059 timezone
= datetime
.timedelta()
3061 date_str
= date_str
[:-len(m
.group('tz'))]
3062 if not m
.group('sign'):
3063 timezone
= datetime
.timedelta()
3065 sign
= 1 if m
.group('sign') == '+' else -1
3066 timezone
= datetime
.timedelta(
3067 hours
=sign
* int(m
.group('hours')),
3068 minutes
=sign
* int(m
.group('minutes')))
3069 return timezone
, date_str
3072 def parse_iso8601(date_str
, delimiter
='T', timezone
=None):
3073 """ Return a UNIX timestamp from the given date """
3075 if date_str
is None:
3078 date_str
= re
.sub(r
'\.[0-9]+', '', date_str
)
3080 if timezone
is None:
3081 timezone
, date_str
= extract_timezone(date_str
)
3084 date_format
= '%Y-%m-%d{0}%H:%M:%S'.format(delimiter
)
3085 dt
= datetime
.datetime
.strptime(date_str
, date_format
) - timezone
3086 return calendar
.timegm(dt
.timetuple())
3091 def date_formats(day_first
=True):
3092 return DATE_FORMATS_DAY_FIRST
if day_first
else DATE_FORMATS_MONTH_FIRST
3095 def unified_strdate(date_str
, day_first
=True):
3096 """Return a string with the date in the format YYYYMMDD"""
3098 if date_str
is None:
3102 date_str
= date_str
.replace(',', ' ')
3103 # Remove AM/PM + timezone
3104 date_str
= re
.sub(r
'(?i)\s*(?:AM|PM)(?:\s+[A-Z]+)?', '', date_str
)
3105 _
, date_str
= extract_timezone(date_str
)
3107 for expression
in date_formats(day_first
):
3109 upload_date
= datetime
.datetime
.strptime(date_str
, expression
).strftime('%Y%m%d')
3112 if upload_date
is None:
3113 timetuple
= email
.utils
.parsedate_tz(date_str
)
3116 upload_date
= datetime
.datetime(*timetuple
[:6]).strftime('%Y%m%d')
3119 if upload_date
is not None:
3120 return compat_str(upload_date
)
3123 def unified_timestamp(date_str
, day_first
=True):
3124 if date_str
is None:
3127 date_str
= re
.sub(r
'[,|]', '', date_str
)
3129 pm_delta
= 12 if re
.search(r
'(?i)PM', date_str
) else 0
3130 timezone
, date_str
= extract_timezone(date_str
)
3132 # Remove AM/PM + timezone
3133 date_str
= re
.sub(r
'(?i)\s*(?:AM|PM)(?:\s+[A-Z]+)?', '', date_str
)
3135 # Remove unrecognized timezones from ISO 8601 alike timestamps
3136 m
= re
.search(r
'\d{1,2}:\d{1,2}(?:\.\d+)?(?P<tz>\s*[A-Z]+)$', date_str
)
3138 date_str
= date_str
[:-len(m
.group('tz'))]
3140 # Python only supports microseconds, so remove nanoseconds
3141 m
= re
.search(r
'^([0-9]{4,}-[0-9]{1,2}-[0-9]{1,2}T[0-9]{1,2}:[0-9]{1,2}:[0-9]{1,2}\.[0-9]{6})[0-9]+$', date_str
)
3143 date_str
= m
.group(1)
3145 for expression
in date_formats(day_first
):
3147 dt
= datetime
.datetime
.strptime(date_str
, expression
) - timezone
+ datetime
.timedelta(hours
=pm_delta
)
3148 return calendar
.timegm(dt
.timetuple())
3151 timetuple
= email
.utils
.parsedate_tz(date_str
)
3153 return calendar
.timegm(timetuple
) + pm_delta
* 3600
3156 def determine_ext(url
, default_ext
='unknown_video'):
3157 if url
is None or '.' not in url
:
3159 guess
= url
.partition('?')[0].rpartition('.')[2]
3160 if re
.match(r
'^[A-Za-z0-9]+$', guess
):
3162 # Try extract ext from URLs like http://example.com/foo/bar.mp4/?download
3163 elif guess
.rstrip('/') in KNOWN_EXTENSIONS
:
3164 return guess
.rstrip('/')
3169 def subtitles_filename(filename
, sub_lang
, sub_format
, expected_real_ext
=None):
3170 return replace_extension(filename
, sub_lang
+ '.' + sub_format
, expected_real_ext
)
3173 def datetime_from_str(date_str
, precision
='auto', format
='%Y%m%d'):
3175 Return a datetime object from a string in the format YYYYMMDD or
3176 (now|today|date)[+-][0-9](microsecond|second|minute|hour|day|week|month|year)(s)?
3178 format: string date format used to return datetime object from
3179 precision: round the time portion of a datetime object.
3180 auto|microsecond|second|minute|hour|day.
3181 auto: round to the unit provided in date_str (if applicable).
3183 auto_precision
= False
3184 if precision
== 'auto':
3185 auto_precision
= True
3186 precision
= 'microsecond'
3187 today
= datetime_round(datetime
.datetime
.now(), precision
)
3188 if date_str
in ('now', 'today'):
3190 if date_str
== 'yesterday':
3191 return today
- datetime
.timedelta(days
=1)
3193 r
'(?P<start>.+)(?P<sign>[+-])(?P<time>\d+)(?P<unit>microsecond|second|minute|hour|day|week|month|year)(s)?',
3195 if match
is not None:
3196 start_time
= datetime_from_str(match
.group('start'), precision
, format
)
3197 time
= int(match
.group('time')) * (-1 if match
.group('sign') == '-' else 1)
3198 unit
= match
.group('unit')
3199 if unit
== 'month' or unit
== 'year':
3200 new_date
= datetime_add_months(start_time
, time
* 12 if unit
== 'year' else time
)
3206 delta
= datetime
.timedelta(**{unit + 's': time}
)
3207 new_date
= start_time
+ delta
3209 return datetime_round(new_date
, unit
)
3212 return datetime_round(datetime
.datetime
.strptime(date_str
, format
), precision
)
3215 def date_from_str(date_str
, format
='%Y%m%d'):
3217 Return a datetime object from a string in the format YYYYMMDD or
3218 (now|today|date)[+-][0-9](microsecond|second|minute|hour|day|week|month|year)(s)?
3220 format: string date format used to return datetime object from
3222 return datetime_from_str(date_str
, precision
='microsecond', format
=format
).date()
3225 def datetime_add_months(dt
, months
):
3226 """Increment/Decrement a datetime object by months."""
3227 month
= dt
.month
+ months
- 1
3228 year
= dt
.year
+ month
// 12
3229 month
= month
% 12 + 1
3230 day
= min(dt
.day
, calendar
.monthrange(year
, month
)[1])
3231 return dt
.replace(year
, month
, day
)
3234 def datetime_round(dt
, precision
='day'):
3236 Round a datetime object's time to a specific precision
3238 if precision
== 'microsecond':
3247 roundto
= lambda x
, n
: ((x
+ n
/ 2) // n
) * n
3248 timestamp
= calendar
.timegm(dt
.timetuple())
3249 return datetime
.datetime
.utcfromtimestamp(roundto(timestamp
, unit_seconds
[precision
]))
3252 def hyphenate_date(date_str
):
3254 Convert a date in 'YYYYMMDD' format to 'YYYY-MM-DD' format"""
3255 match
= re
.match(r
'^(\d\d\d\d)(\d\d)(\d\d)$', date_str
)
3256 if match
is not None:
3257 return '-'.join(match
.groups())
3262 class DateRange(object):
3263 """Represents a time interval between two dates"""
3265 def __init__(self
, start
=None, end
=None):
3266 """start and end must be strings in the format accepted by date"""
3267 if start
is not None:
3268 self
.start
= date_from_str(start
)
3270 self
.start
= datetime
.datetime
.min.date()
3272 self
.end
= date_from_str(end
)
3274 self
.end
= datetime
.datetime
.max.date()
3275 if self
.start
> self
.end
:
3276 raise ValueError('Date range: "%s" , the start date must be before the end date' % self
)
3280 """Returns a range that only contains the given day"""
3281 return cls(day
, day
)
3283 def __contains__(self
, date
):
3284 """Check if the date is in the range"""
3285 if not isinstance(date
, datetime
.date
):
3286 date
= date_from_str(date
)
3287 return self
.start
<= date
<= self
.end
3290 return '%s - %s' % (self
.start
.isoformat(), self
.end
.isoformat())
3293 def platform_name():
3294 """ Returns the platform name as a compat_str """
3295 res
= platform
.platform()
3296 if isinstance(res
, bytes):
3297 res
= res
.decode(preferredencoding())
3299 assert isinstance(res
, compat_str
)
3303 def get_windows_version():
3304 ''' Get Windows version. None if it's not running on Windows '''
3305 if compat_os_name
== 'nt':
3306 return version_tuple(platform
.win32_ver()[1])
3311 def _windows_write_string(s
, out
):
3312 """ Returns True if the string was written using special methods,
3313 False if it has yet to be written out."""
3314 # Adapted from http://stackoverflow.com/a/3259271/35070
3317 import ctypes
.wintypes
3325 fileno
= out
.fileno()
3326 except AttributeError:
3327 # If the output stream doesn't have a fileno, it's virtual
3329 except io
.UnsupportedOperation
:
3330 # Some strange Windows pseudo files?
3332 if fileno
not in WIN_OUTPUT_IDS
:
3335 GetStdHandle
= compat_ctypes_WINFUNCTYPE(
3336 ctypes
.wintypes
.HANDLE
, ctypes
.wintypes
.DWORD
)(
3337 ('GetStdHandle', ctypes
.windll
.kernel32
))
3338 h
= GetStdHandle(WIN_OUTPUT_IDS
[fileno
])
3340 WriteConsoleW
= compat_ctypes_WINFUNCTYPE(
3341 ctypes
.wintypes
.BOOL
, ctypes
.wintypes
.HANDLE
, ctypes
.wintypes
.LPWSTR
,
3342 ctypes
.wintypes
.DWORD
, ctypes
.POINTER(ctypes
.wintypes
.DWORD
),
3343 ctypes
.wintypes
.LPVOID
)(('WriteConsoleW', ctypes
.windll
.kernel32
))
3344 written
= ctypes
.wintypes
.DWORD(0)
3346 GetFileType
= compat_ctypes_WINFUNCTYPE(ctypes
.wintypes
.DWORD
, ctypes
.wintypes
.DWORD
)(('GetFileType', ctypes
.windll
.kernel32
))
3347 FILE_TYPE_CHAR
= 0x0002
3348 FILE_TYPE_REMOTE
= 0x8000
3349 GetConsoleMode
= compat_ctypes_WINFUNCTYPE(
3350 ctypes
.wintypes
.BOOL
, ctypes
.wintypes
.HANDLE
,
3351 ctypes
.POINTER(ctypes
.wintypes
.DWORD
))(
3352 ('GetConsoleMode', ctypes
.windll
.kernel32
))
3353 INVALID_HANDLE_VALUE
= ctypes
.wintypes
.DWORD(-1).value
3355 def not_a_console(handle
):
3356 if handle
== INVALID_HANDLE_VALUE
or handle
is None:
3358 return ((GetFileType(handle
) & ~FILE_TYPE_REMOTE
) != FILE_TYPE_CHAR
3359 or GetConsoleMode(handle
, ctypes
.byref(ctypes
.wintypes
.DWORD())) == 0)
3361 if not_a_console(h
):
3364 def next_nonbmp_pos(s
):
3366 return next(i
for i
, c
in enumerate(s
) if ord(c
) > 0xffff)
3367 except StopIteration:
3371 count
= min(next_nonbmp_pos(s
), 1024)
3373 ret
= WriteConsoleW(
3374 h
, s
, count
if count
else 2, ctypes
.byref(written
), None)
3376 raise OSError('Failed to write string')
3377 if not count
: # We just wrote a non-BMP character
3378 assert written
.value
== 2
3381 assert written
.value
> 0
3382 s
= s
[written
.value
:]
3386 def write_string(s
, out
=None, encoding
=None):
3389 assert type(s
) == compat_str
3391 if sys
.platform
== 'win32' and encoding
is None and hasattr(out
, 'fileno'):
3392 if _windows_write_string(s
, out
):
3395 if ('b' in getattr(out
, 'mode', '')
3396 or sys
.version_info
[0] < 3): # Python 2 lies about mode of sys.stderr
3397 byt
= s
.encode(encoding
or preferredencoding(), 'ignore')
3399 elif hasattr(out
, 'buffer'):
3400 enc
= encoding
or getattr(out
, 'encoding', None) or preferredencoding()
3401 byt
= s
.encode(enc
, 'ignore')
3402 out
.buffer.write(byt
)
3408 def bytes_to_intlist(bs
):
3411 if isinstance(bs
[0], int): # Python 3
3414 return [ord(c
) for c
in bs
]
3417 def intlist_to_bytes(xs
):
3420 return compat_struct_pack('%dB' % len(xs
), *xs
)
3423 # Cross-platform file locking
3424 if sys
.platform
== 'win32':
3425 import ctypes
.wintypes
3428 class OVERLAPPED(ctypes
.Structure
):
3430 ('Internal', ctypes
.wintypes
.LPVOID
),
3431 ('InternalHigh', ctypes
.wintypes
.LPVOID
),
3432 ('Offset', ctypes
.wintypes
.DWORD
),
3433 ('OffsetHigh', ctypes
.wintypes
.DWORD
),
3434 ('hEvent', ctypes
.wintypes
.HANDLE
),
3437 kernel32
= ctypes
.windll
.kernel32
3438 LockFileEx
= kernel32
.LockFileEx
3439 LockFileEx
.argtypes
= [
3440 ctypes
.wintypes
.HANDLE
, # hFile
3441 ctypes
.wintypes
.DWORD
, # dwFlags
3442 ctypes
.wintypes
.DWORD
, # dwReserved
3443 ctypes
.wintypes
.DWORD
, # nNumberOfBytesToLockLow
3444 ctypes
.wintypes
.DWORD
, # nNumberOfBytesToLockHigh
3445 ctypes
.POINTER(OVERLAPPED
) # Overlapped
3447 LockFileEx
.restype
= ctypes
.wintypes
.BOOL
3448 UnlockFileEx
= kernel32
.UnlockFileEx
3449 UnlockFileEx
.argtypes
= [
3450 ctypes
.wintypes
.HANDLE
, # hFile
3451 ctypes
.wintypes
.DWORD
, # dwReserved
3452 ctypes
.wintypes
.DWORD
, # nNumberOfBytesToLockLow
3453 ctypes
.wintypes
.DWORD
, # nNumberOfBytesToLockHigh
3454 ctypes
.POINTER(OVERLAPPED
) # Overlapped
3456 UnlockFileEx
.restype
= ctypes
.wintypes
.BOOL
3457 whole_low
= 0xffffffff
3458 whole_high
= 0x7fffffff
3460 def _lock_file(f
, exclusive
):
3461 overlapped
= OVERLAPPED()
3462 overlapped
.Offset
= 0
3463 overlapped
.OffsetHigh
= 0
3464 overlapped
.hEvent
= 0
3465 f
._lock
_file
_overlapped
_p
= ctypes
.pointer(overlapped
)
3466 handle
= msvcrt
.get_osfhandle(f
.fileno())
3467 if not LockFileEx(handle
, 0x2 if exclusive
else 0x0, 0,
3468 whole_low
, whole_high
, f
._lock
_file
_overlapped
_p
):
3469 raise OSError('Locking file failed: %r' % ctypes
.FormatError())
3471 def _unlock_file(f
):
3472 assert f
._lock
_file
_overlapped
_p
3473 handle
= msvcrt
.get_osfhandle(f
.fileno())
3474 if not UnlockFileEx(handle
, 0,
3475 whole_low
, whole_high
, f
._lock
_file
_overlapped
_p
):
3476 raise OSError('Unlocking file failed: %r' % ctypes
.FormatError())
3479 # Some platforms, such as Jython, is missing fcntl
3483 def _lock_file(f
, exclusive
):
3484 fcntl
.flock(f
, fcntl
.LOCK_EX
if exclusive
else fcntl
.LOCK_SH
)
3486 def _unlock_file(f
):
3487 fcntl
.flock(f
, fcntl
.LOCK_UN
)
3489 UNSUPPORTED_MSG
= 'file locking is not supported on this platform'
3491 def _lock_file(f
, exclusive
):
3492 raise IOError(UNSUPPORTED_MSG
)
3494 def _unlock_file(f
):
3495 raise IOError(UNSUPPORTED_MSG
)
3498 class locked_file(object):
3499 def __init__(self
, filename
, mode
, encoding
=None):
3500 assert mode
in ['r', 'a', 'w']
3501 self
.f
= io
.open(filename
, mode
, encoding
=encoding
)
3504 def __enter__(self
):
3505 exclusive
= self
.mode
!= 'r'
3507 _lock_file(self
.f
, exclusive
)
3513 def __exit__(self
, etype
, value
, traceback
):
3515 _unlock_file(self
.f
)
3522 def write(self
, *args
):
3523 return self
.f
.write(*args
)
3525 def read(self
, *args
):
3526 return self
.f
.read(*args
)
3529 def get_filesystem_encoding():
3530 encoding
= sys
.getfilesystemencoding()
3531 return encoding
if encoding
is not None else 'utf-8'
3534 def shell_quote(args
):
3536 encoding
= get_filesystem_encoding()
3538 if isinstance(a
, bytes):
3539 # We may get a filename encoded with 'encodeFilename'
3540 a
= a
.decode(encoding
)
3541 quoted_args
.append(compat_shlex_quote(a
))
3542 return ' '.join(quoted_args
)
3545 def smuggle_url(url
, data
):
3546 """ Pass additional data in a URL for internal use. """
3548 url
, idata
= unsmuggle_url(url
, {})
3550 sdata
= compat_urllib_parse_urlencode(
3551 {'__youtubedl_smuggle': json.dumps(data)}
)
3552 return url
+ '#' + sdata
3555 def unsmuggle_url(smug_url
, default
=None):
3556 if '#__youtubedl_smuggle' not in smug_url
:
3557 return smug_url
, default
3558 url
, _
, sdata
= smug_url
.rpartition('#')
3559 jsond
= compat_parse_qs(sdata
)['__youtubedl_smuggle'][0]
3560 data
= json
.loads(jsond
)
3564 def format_bytes(bytes):
3567 if type(bytes) is str:
3568 bytes = float(bytes)
3572 exponent
= int(math
.log(bytes, 1024.0))
3573 suffix
= ['B', 'KiB', 'MiB', 'GiB', 'TiB', 'PiB', 'EiB', 'ZiB', 'YiB'][exponent
]
3574 converted
= float(bytes) / float(1024 ** exponent
)
3575 return '%.2f%s' % (converted
, suffix
)
3578 def lookup_unit_table(unit_table
, s
):
3579 units_re
= '|'.join(re
.escape(u
) for u
in unit_table
)
3581 r
'(?P<num>[0-9]+(?:[,.][0-9]*)?)\s*(?P<unit>%s)\b' % units_re
, s
)
3584 num_str
= m
.group('num').replace(',', '.')
3585 mult
= unit_table
[m
.group('unit')]
3586 return int(float(num_str
) * mult
)
3589 def parse_filesize(s
):
3593 # The lower-case forms are of course incorrect and unofficial,
3594 # but we support those too
3611 'megabytes': 1000 ** 2,
3612 'mebibytes': 1024 ** 2,
3618 'gigabytes': 1000 ** 3,
3619 'gibibytes': 1024 ** 3,
3625 'terabytes': 1000 ** 4,
3626 'tebibytes': 1024 ** 4,
3632 'petabytes': 1000 ** 5,
3633 'pebibytes': 1024 ** 5,
3639 'exabytes': 1000 ** 6,
3640 'exbibytes': 1024 ** 6,
3646 'zettabytes': 1000 ** 7,
3647 'zebibytes': 1024 ** 7,
3653 'yottabytes': 1000 ** 8,
3654 'yobibytes': 1024 ** 8,
3657 return lookup_unit_table(_UNIT_TABLE
, s
)
3666 if re
.match(r
'^[\d,.]+$', s
):
3667 return str_to_int(s
)
3678 return lookup_unit_table(_UNIT_TABLE
, s
)
3681 def parse_resolution(s
):
3685 mobj
= re
.search(r
'\b(?P<w>\d+)\s*[xX×]\s*(?P<h>\d+)\b', s
)
3688 'width': int(mobj
.group('w')),
3689 'height': int(mobj
.group('h')),
3692 mobj
= re
.search(r
'\b(\d+)[pPiI]\b', s
)
3694 return {'height': int(mobj.group(1))}
3696 mobj
= re
.search(r
'\b([48])[kK]\b', s
)
3698 return {'height': int(mobj.group(1)) * 540}
3703 def parse_bitrate(s
):
3704 if not isinstance(s
, compat_str
):
3706 mobj
= re
.search(r
'\b(\d+)\s*kbps', s
)
3708 return int(mobj
.group(1))
3711 def month_by_name(name
, lang
='en'):
3712 """ Return the number of a month by (locale-independently) English name """
3714 month_names
= MONTH_NAMES
.get(lang
, MONTH_NAMES
['en'])
3717 return month_names
.index(name
) + 1
3722 def month_by_abbreviation(abbrev
):
3723 """ Return the number of a month by (locale-independently) English
3727 return [s
[:3] for s
in ENGLISH_MONTH_NAMES
].index(abbrev
) + 1
3732 def fix_xml_ampersands(xml_str
):
3733 """Replace all the '&' by '&' in XML"""
3735 r
'&(?!amp;|lt;|gt;|apos;|quot;|#x[0-9a-fA-F]{,4};|#[0-9]{,4};)',
3740 def setproctitle(title
):
3741 assert isinstance(title
, compat_str
)
3743 # ctypes in Jython is not complete
3744 # http://bugs.jython.org/issue2148
3745 if sys
.platform
.startswith('java'):
3749 libc
= ctypes
.cdll
.LoadLibrary('libc.so.6')
3753 # LoadLibrary in Windows Python 2.7.13 only expects
3754 # a bytestring, but since unicode_literals turns
3755 # every string into a unicode string, it fails.
3757 title_bytes
= title
.encode('utf-8')
3758 buf
= ctypes
.create_string_buffer(len(title_bytes
))
3759 buf
.value
= title_bytes
3761 libc
.prctl(15, buf
, 0, 0, 0)
3762 except AttributeError:
3763 return # Strange libc, just skip this
3766 def remove_start(s
, start
):
3767 return s
[len(start
):] if s
is not None and s
.startswith(start
) else s
3770 def remove_end(s
, end
):
3771 return s
[:-len(end
)] if s
is not None and s
.endswith(end
) else s
3774 def remove_quotes(s
):
3775 if s
is None or len(s
) < 2:
3777 for quote
in ('"', "'", ):
3778 if s
[0] == quote
and s
[-1] == quote
:
3783 def get_domain(url
):
3784 domain
= re
.match(r
'(?:https?:\/\/)?(?:www\.)?(?P<domain>[^\n\/]+\.[^\n\/]+)(?:\/(.*))?', url
)
3785 return domain
.group('domain') if domain
else None
3788 def url_basename(url
):
3789 path
= compat_urlparse
.urlparse(url
).path
3790 return path
.strip('/').split('/')[-1]
3794 return re
.match(r
'https?://[^?#&]+/', url
).group()
3797 def urljoin(base
, path
):
3798 if isinstance(path
, bytes):
3799 path
= path
.decode('utf-8')
3800 if not isinstance(path
, compat_str
) or not path
:
3802 if re
.match(r
'^(?:[a-zA-Z][a-zA-Z0-9+-.]*:)?//', path
):
3804 if isinstance(base
, bytes):
3805 base
= base
.decode('utf-8')
3806 if not isinstance(base
, compat_str
) or not re
.match(
3807 r
'^(?:https?:)?//', base
):
3809 return compat_urlparse
.urljoin(base
, path
)
3812 class HEADRequest(compat_urllib_request
.Request
):
3813 def get_method(self
):
3817 class PUTRequest(compat_urllib_request
.Request
):
3818 def get_method(self
):
3822 def int_or_none(v
, scale
=1, default
=None, get_attr
=None, invscale
=1):
3825 v
= getattr(v
, get_attr
, None)
3831 return int(v
) * invscale
// scale
3832 except (ValueError, TypeError):
3836 def str_or_none(v
, default
=None):
3837 return default
if v
is None else compat_str(v
)
3840 def str_to_int(int_str
):
3841 """ A more relaxed version of int_or_none """
3842 if isinstance(int_str
, compat_integer_types
):
3844 elif isinstance(int_str
, compat_str
):
3845 int_str
= re
.sub(r
'[,\.\+]', '', int_str
)
3846 return int_or_none(int_str
)
3849 def float_or_none(v
, scale
=1, invscale
=1, default
=None):
3853 return float(v
) * invscale
/ scale
3854 except (ValueError, TypeError):
3858 def bool_or_none(v
, default
=None):
3859 return v
if isinstance(v
, bool) else default
3862 def strip_or_none(v
, default
=None):
3863 return v
.strip() if isinstance(v
, compat_str
) else default
3866 def url_or_none(url
):
3867 if not url
or not isinstance(url
, compat_str
):
3870 return url
if re
.match(r
'^(?:(?:https?|rt(?:m(?:pt?[es]?|fp)|sp[su]?)|mms|ftps?):)?//', url
) else None
3873 def strftime_or_none(timestamp
, date_format
, default
=None):
3874 datetime_object
= None
3876 if isinstance(timestamp
, compat_numeric_types
): # unix timestamp
3877 datetime_object
= datetime
.datetime
.utcfromtimestamp(timestamp
)
3878 elif isinstance(timestamp
, compat_str
): # assume YYYYMMDD
3879 datetime_object
= datetime
.datetime
.strptime(timestamp
, '%Y%m%d')
3880 return datetime_object
.strftime(date_format
)
3881 except (ValueError, TypeError, AttributeError):
3885 def parse_duration(s
):
3886 if not isinstance(s
, compat_basestring
):
3891 days
, hours
, mins
, secs
, ms
= [None] * 5
3892 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
)
3894 days
, hours
, mins
, secs
, ms
= m
.groups()
3899 [0-9]+\s*y(?:ears?)?\s*
3902 [0-9]+\s*m(?:onths?)?\s*
3905 [0-9]+\s*w(?:eeks?)?\s*
3908 (?P<days>[0-9]+)\s*d(?:ays?)?\s*
3912 (?P<hours>[0-9]+)\s*h(?:ours?)?\s*
3915 (?P<mins>[0-9]+)\s*m(?:in(?:ute)?s?)?\s*
3918 (?P<secs>[0-9]+)(?P<ms>\.[0-9]+)?\s*s(?:ec(?:ond)?s?)?\s*
3921 days
, hours
, mins
, secs
, ms
= m
.groups()
3923 m
= re
.match(r
'(?i)(?:(?P<hours>[0-9.]+)\s*(?:hours?)|(?P<mins>[0-9.]+)\s*(?:mins?\.?|minutes?)\s*)Z?$', s
)
3925 hours
, mins
= m
.groups()
3931 duration
+= float(secs
)
3933 duration
+= float(mins
) * 60
3935 duration
+= float(hours
) * 60 * 60
3937 duration
+= float(days
) * 24 * 60 * 60
3939 duration
+= float(ms
)
3943 def prepend_extension(filename
, ext
, expected_real_ext
=None):
3944 name
, real_ext
= os
.path
.splitext(filename
)
3946 '{0}.{1}{2}'.format(name
, ext
, real_ext
)
3947 if not expected_real_ext
or real_ext
[1:] == expected_real_ext
3948 else '{0}.{1}'.format(filename
, ext
))
3951 def replace_extension(filename
, ext
, expected_real_ext
=None):
3952 name
, real_ext
= os
.path
.splitext(filename
)
3953 return '{0}.{1}'.format(
3954 name
if not expected_real_ext
or real_ext
[1:] == expected_real_ext
else filename
,
3958 def check_executable(exe
, args
=[]):
3959 """ Checks if the given binary is installed somewhere in PATH, and returns its name.
3960 args can be a list of arguments for a short output (like -version) """
3962 process_communicate_or_kill(subprocess
.Popen(
3963 [exe
] + args
, stdout
=subprocess
.PIPE
, stderr
=subprocess
.PIPE
))
3969 def get_exe_version(exe
, args
=['--version'],
3970 version_re
=None, unrecognized
='present'):
3971 """ Returns the version of the specified executable,
3972 or False if the executable is not present """
3974 # STDIN should be redirected too. On UNIX-like systems, ffmpeg triggers
3975 # SIGTTOU if yt-dlp is run in the background.
3976 # See https://github.com/ytdl-org/youtube-dl/issues/955#issuecomment-209789656
3977 out
, _
= process_communicate_or_kill(subprocess
.Popen(
3978 [encodeArgument(exe
)] + args
,
3979 stdin
=subprocess
.PIPE
,
3980 stdout
=subprocess
.PIPE
, stderr
=subprocess
.STDOUT
))
3983 if isinstance(out
, bytes): # Python 2.x
3984 out
= out
.decode('ascii', 'ignore')
3985 return detect_exe_version(out
, version_re
, unrecognized
)
3988 def detect_exe_version(output
, version_re
=None, unrecognized
='present'):
3989 assert isinstance(output
, compat_str
)
3990 if version_re
is None:
3991 version_re
= r
'version\s+([-0-9._a-zA-Z]+)'
3992 m
= re
.search(version_re
, output
)
3999 class LazyList(collections
.abc
.Sequence
):
4000 ''' Lazy immutable list from an iterable
4001 Note that slices of a LazyList are lists and not LazyList'''
4003 class IndexError(IndexError):
4006 def __init__(self
, iterable
):
4007 self
.__iterable
= iter(iterable
)
4009 self
.__reversed
= False
4013 # We need to consume the entire iterable to iterate in reverse
4014 yield from self
.exhaust()
4016 yield from self
.__cache
4017 for item
in self
.__iterable
:
4018 self
.__cache
.append(item
)
4021 def __exhaust(self
):
4022 self
.__cache
.extend(self
.__iterable
)
4026 ''' Evaluate the entire iterable '''
4027 return self
.__exhaust
()[::-1 if self
.__reversed
else 1]
4030 def __reverse_index(x
):
4031 return None if x
is None else -(x
+ 1)
4033 def __getitem__(self
, idx
):
4034 if isinstance(idx
, slice):
4036 idx
= slice(self
.__reverse
_index
(idx
.start
), self
.__reverse
_index
(idx
.stop
), -(idx
.step
or 1))
4037 start
, stop
, step
= idx
.start
, idx
.stop
, idx
.step
or 1
4038 elif isinstance(idx
, int):
4040 idx
= self
.__reverse
_index
(idx
)
4041 start
, stop
, step
= idx
, idx
, 0
4043 raise TypeError('indices must be integers or slices')
4044 if ((start
or 0) < 0 or (stop
or 0) < 0
4045 or (start
is None and step
< 0)
4046 or (stop
is None and step
> 0)):
4047 # We need to consume the entire iterable to be able to slice from the end
4048 # Obviously, never use this with infinite iterables
4051 return self
.__cache
[idx
]
4052 except IndexError as e
:
4053 raise self
.IndexError(e
) from e
4054 n
= max(start
or 0, stop
or 0) - len(self
.__cache
) + 1
4056 self
.__cache
.extend(itertools
.islice(self
.__iterable
, n
))
4058 return self
.__cache
[idx
]
4059 except IndexError as e
:
4060 raise self
.IndexError(e
) from e
4064 self
[-1] if self
.__reversed
else self
[0]
4065 except self
.IndexError:
4071 return len(self
.__cache
)
4074 self
.__reversed
= not self
.__reversed
4078 # repr and str should mimic a list. So we exhaust the iterable
4079 return repr(self
.exhaust())
4082 return repr(self
.exhaust())
4087 # This is only useful for tests
4088 return len(self
.getslice())
4090 def __init__(self
, pagefunc
, pagesize
, use_cache
=True):
4091 self
._pagefunc
= pagefunc
4092 self
._pagesize
= pagesize
4093 self
._use
_cache
= use_cache
4096 def getpage(self
, pagenum
):
4097 page_results
= self
._cache
.get(pagenum
) or list(self
._pagefunc
(pagenum
))
4099 self
._cache
[pagenum
] = page_results
4102 def getslice(self
, start
=0, end
=None):
4103 return list(self
._getslice
(start
, end
))
4105 def _getslice(self
, start
, end
):
4106 raise NotImplementedError('This method must be implemented by subclasses')
4108 def __getitem__(self
, idx
):
4109 # NOTE: cache must be enabled if this is used
4110 if not isinstance(idx
, int) or idx
< 0:
4111 raise TypeError('indices must be non-negative integers')
4112 entries
= self
.getslice(idx
, idx
+ 1)
4113 return entries
[0] if entries
else None
4116 class OnDemandPagedList(PagedList
):
4117 def _getslice(self
, start
, end
):
4118 for pagenum
in itertools
.count(start
// self
._pagesize
):
4119 firstid
= pagenum
* self
._pagesize
4120 nextfirstid
= pagenum
* self
._pagesize
+ self
._pagesize
4121 if start
>= nextfirstid
:
4125 start
% self
._pagesize
4126 if firstid
<= start
< nextfirstid
4129 ((end
- 1) % self
._pagesize
) + 1
4130 if (end
is not None and firstid
<= end
<= nextfirstid
)
4133 page_results
= self
.getpage(pagenum
)
4134 if startv
!= 0 or endv
is not None:
4135 page_results
= page_results
[startv
:endv
]
4136 yield from page_results
4138 # A little optimization - if current page is not "full", ie. does
4139 # not contain page_size videos then we can assume that this page
4140 # is the last one - there are no more ids on further pages -
4141 # i.e. no need to query again.
4142 if len(page_results
) + startv
< self
._pagesize
:
4145 # If we got the whole page, but the next page is not interesting,
4146 # break out early as well
4147 if end
== nextfirstid
:
4151 class InAdvancePagedList(PagedList
):
4152 def __init__(self
, pagefunc
, pagecount
, pagesize
):
4153 self
._pagecount
= pagecount
4154 PagedList
.__init
__(self
, pagefunc
, pagesize
, True)
4156 def _getslice(self
, start
, end
):
4157 start_page
= start
// self
._pagesize
4159 self
._pagecount
if end
is None else (end
// self
._pagesize
+ 1))
4160 skip_elems
= start
- start_page
* self
._pagesize
4161 only_more
= None if end
is None else end
- start
4162 for pagenum
in range(start_page
, end_page
):
4163 page_results
= self
.getpage(pagenum
)
4165 page_results
= page_results
[skip_elems
:]
4167 if only_more
is not None:
4168 if len(page_results
) < only_more
:
4169 only_more
-= len(page_results
)
4171 yield from page_results
[:only_more
]
4173 yield from page_results
4176 def uppercase_escape(s
):
4177 unicode_escape
= codecs
.getdecoder('unicode_escape')
4179 r
'\\U[0-9a-fA-F]{8}',
4180 lambda m
: unicode_escape(m
.group(0))[0],
4184 def lowercase_escape(s
):
4185 unicode_escape
= codecs
.getdecoder('unicode_escape')
4187 r
'\\u[0-9a-fA-F]{4}',
4188 lambda m
: unicode_escape(m
.group(0))[0],
4192 def escape_rfc3986(s
):
4193 """Escape non-ASCII characters as suggested by RFC 3986"""
4194 if sys
.version_info
< (3, 0) and isinstance(s
, compat_str
):
4195 s
= s
.encode('utf-8')
4196 return compat_urllib_parse
.quote(s
, b
"%/;:@&=+$,!~*'()?#[]")
4199 def escape_url(url
):
4200 """Escape URL as suggested by RFC 3986"""
4201 url_parsed
= compat_urllib_parse_urlparse(url
)
4202 return url_parsed
._replace
(
4203 netloc
=url_parsed
.netloc
.encode('idna').decode('ascii'),
4204 path
=escape_rfc3986(url_parsed
.path
),
4205 params
=escape_rfc3986(url_parsed
.params
),
4206 query
=escape_rfc3986(url_parsed
.query
),
4207 fragment
=escape_rfc3986(url_parsed
.fragment
)
4212 return compat_parse_qs(compat_urllib_parse_urlparse(url
).query
)
4215 def read_batch_urls(batch_fd
):
4217 if not isinstance(url
, compat_str
):
4218 url
= url
.decode('utf-8', 'replace')
4219 BOM_UTF8
= ('\xef\xbb\xbf', '\ufeff')
4220 for bom
in BOM_UTF8
:
4221 if url
.startswith(bom
):
4222 url
= url
[len(bom
):]
4224 if not url
or url
.startswith(('#', ';', ']')):
4226 # "#" cannot be stripped out since it is part of the URI
4227 # However, it can be safely stipped out if follwing a whitespace
4228 return re
.split(r
'\s#', url
, 1)[0].rstrip()
4230 with contextlib
.closing(batch_fd
) as fd
:
4231 return [url
for url
in map(fixup
, fd
) if url
]
4234 def urlencode_postdata(*args
, **kargs
):
4235 return compat_urllib_parse_urlencode(*args
, **kargs
).encode('ascii')
4238 def update_url_query(url
, query
):
4241 parsed_url
= compat_urlparse
.urlparse(url
)
4242 qs
= compat_parse_qs(parsed_url
.query
)
4244 return compat_urlparse
.urlunparse(parsed_url
._replace
(
4245 query
=compat_urllib_parse_urlencode(qs
, True)))
4248 def update_Request(req
, url
=None, data
=None, headers
={}, query={}
):
4249 req_headers
= req
.headers
.copy()
4250 req_headers
.update(headers
)
4251 req_data
= data
or req
.data
4252 req_url
= update_url_query(url
or req
.get_full_url(), query
)
4253 req_get_method
= req
.get_method()
4254 if req_get_method
== 'HEAD':
4255 req_type
= HEADRequest
4256 elif req_get_method
== 'PUT':
4257 req_type
= PUTRequest
4259 req_type
= compat_urllib_request
.Request
4261 req_url
, data
=req_data
, headers
=req_headers
,
4262 origin_req_host
=req
.origin_req_host
, unverifiable
=req
.unverifiable
)
4263 if hasattr(req
, 'timeout'):
4264 new_req
.timeout
= req
.timeout
4268 def _multipart_encode_impl(data
, boundary
):
4269 content_type
= 'multipart/form-data; boundary=%s' % boundary
4272 for k
, v
in data
.items():
4273 out
+= b
'--' + boundary
.encode('ascii') + b
'\r\n'
4274 if isinstance(k
, compat_str
):
4275 k
= k
.encode('utf-8')
4276 if isinstance(v
, compat_str
):
4277 v
= v
.encode('utf-8')
4278 # RFC 2047 requires non-ASCII field names to be encoded, while RFC 7578
4279 # suggests sending UTF-8 directly. Firefox sends UTF-8, too
4280 content
= b
'Content-Disposition: form-data; name="' + k
+ b
'"\r\n\r\n' + v
+ b
'\r\n'
4281 if boundary
.encode('ascii') in content
:
4282 raise ValueError('Boundary overlaps with data')
4285 out
+= b
'--' + boundary
.encode('ascii') + b
'--\r\n'
4287 return out
, content_type
4290 def multipart_encode(data
, boundary
=None):
4292 Encode a dict to RFC 7578-compliant form-data
4295 A dict where keys and values can be either Unicode or bytes-like
4298 If specified a Unicode object, it's used as the boundary. Otherwise
4299 a random boundary is generated.
4301 Reference: https://tools.ietf.org/html/rfc7578
4303 has_specified_boundary
= boundary
is not None
4306 if boundary
is None:
4307 boundary
= '---------------' + str(random
.randrange(0x0fffffff, 0xffffffff))
4310 out
, content_type
= _multipart_encode_impl(data
, boundary
)
4313 if has_specified_boundary
:
4317 return out
, content_type
4320 def dict_get(d
, key_or_keys
, default
=None, skip_false_values
=True):
4321 if isinstance(key_or_keys
, (list, tuple)):
4322 for key
in key_or_keys
:
4323 if key
not in d
or d
[key
] is None or skip_false_values
and not d
[key
]:
4327 return d
.get(key_or_keys
, default
)
4330 def try_get(src
, getter
, expected_type
=None):
4331 for get
in variadic(getter
):
4334 except (AttributeError, KeyError, TypeError, IndexError):
4337 if expected_type
is None or isinstance(v
, expected_type
):
4341 def merge_dicts(*dicts
):
4343 for a_dict
in dicts
:
4344 for k
, v
in a_dict
.items():
4348 or (isinstance(v
, compat_str
) and v
4349 and isinstance(merged
[k
], compat_str
)
4350 and not merged
[k
])):
4355 def encode_compat_str(string
, encoding
=preferredencoding(), errors
='strict'):
4356 return string
if isinstance(string
, compat_str
) else compat_str(string
, encoding
, errors
)
4368 TV_PARENTAL_GUIDELINES
= {
4378 def parse_age_limit(s
):
4380 return s
if 0 <= s
<= 21 else None
4381 if not isinstance(s
, compat_basestring
):
4383 m
= re
.match(r
'^(?P<age>\d{1,2})\+?$', s
)
4385 return int(m
.group('age'))
4388 return US_RATINGS
[s
]
4389 m
= re
.match(r
'^TV[_-]?(%s)$' % '|'.join(k
[3:] for k
in TV_PARENTAL_GUIDELINES
), s
)
4391 return TV_PARENTAL_GUIDELINES
['TV-' + m
.group(1)]
4395 def strip_jsonp(code
):
4398 (?:window\.)?(?P<func_name>[a-zA-Z0-9_.$]*)
4399 (?:\s*&&\s*(?P=func_name))?
4400 \s*\(\s*(?P<callback_data>.*)\);?
4401 \s*?(?://[^\n]*)*$''',
4402 r
'\g<callback_data>', code
)
4405 def js_to_json(code
, vars={}):
4406 # vars is a dict of var, val pairs to substitute
4407 COMMENT_RE
= r
'/\*(?:(?!\*/).)*?\*/|//[^\n]*\n'
4408 SKIP_RE
= r
'\s*(?:{comment})?\s*'.format(comment
=COMMENT_RE
)
4410 (r
'(?s)^(0[xX][0-9a-fA-F]+){skip}:?$'.format(skip
=SKIP_RE
), 16),
4411 (r
'(?s)^(0+[0-7]+){skip}:?$'.format(skip
=SKIP_RE
), 8),
4416 if v
in ('true', 'false', 'null'):
4418 elif v
in ('undefined', 'void 0'):
4420 elif v
.startswith('/*') or v
.startswith('//') or v
.startswith('!') or v
== ',':
4423 if v
[0] in ("'", '"'):
4424 v
= re
.sub(r
'(?s)\\.|"', lambda m
: {
4429 }.get(m
.group(0), m
.group(0)), v
[1:-1])
4431 for regex
, base
in INTEGER_TABLE
:
4432 im
= re
.match(regex
, v
)
4434 i
= int(im
.group(1), base
)
4435 return '"%d":' % i
if v
.endswith(':') else '%d' % i
4442 return re
.sub(r
'''(?sx)
4443 "(?:[^"\\]*(?:\\\\|\\['"nurtbfx/\n]))*[^"\\]*"|
4444 '(?:[^'\\]*(?:\\\\|\\['"nurtbfx/\n]))*[^'\\]*'|
4445 {comment}|,(?={skip}[\]}}])|
4446 void\s0|(?:(?<![0-9])[eE]|[a-df-zA-DF-Z_$])[.a-zA-Z_$0-9]*|
4447 \b(?:0[xX][0-9a-fA-F]+|0+[0-7]+)(?:{skip}:)?|
4450 '''.format(comment
=COMMENT_RE
, skip
=SKIP_RE
), fix_kv
, code
)
4453 def qualities(quality_ids
):
4454 """ Get a numeric quality value out of a list of possible values """
4457 return quality_ids
.index(qid
)
4464 'default': '%(title)s [%(id)s].%(ext)s',
4465 'chapter': '%(title)s - %(section_number)03d %(section_title)s [%(id)s].%(ext)s',
4471 'description': 'description',
4472 'annotation': 'annotations.xml',
4473 'infojson': 'info.json',
4474 'pl_thumbnail': None,
4475 'pl_description': 'description',
4476 'pl_infojson': 'info.json',
4479 # As of [1] format syntax is:
4480 # %[mapping_key][conversion_flags][minimum_width][.precision][length_modifier]type
4481 # 1. https://docs.python.org/2/library/stdtypes.html#string-formatting
4482 STR_FORMAT_RE_TMPL
= r
'''(?x)
4483 (?<!%)(?P<prefix>(?:%%)*)
4485 (?P<has_key>\((?P<key>{0})\))?
4487 (?P<conversion>[#0\-+ ]+)?
4489 (?P<precision>\.\d+)?
4490 (?P<len_mod>[hlL])? # unused in python
4491 {1} # conversion type
4496 STR_FORMAT_TYPES
= 'diouxXeEfFgGcrs'
4499 def limit_length(s
, length
):
4500 """ Add ellipses to overly long strings """
4505 return s
[:length
- len(ELLIPSES
)] + ELLIPSES
4509 def version_tuple(v
):
4510 return tuple(int(e
) for e
in re
.split(r
'[-.]', v
))
4513 def is_outdated_version(version
, limit
, assume_new
=True):
4515 return not assume_new
4517 return version_tuple(version
) < version_tuple(limit
)
4519 return not assume_new
4522 def ytdl_is_updateable():
4523 """ Returns if yt-dlp can be updated with -U """
4525 from .update
import is_non_updateable
4527 return not is_non_updateable()
4530 def args_to_str(args
):
4531 # Get a short string representation for a subprocess command
4532 return ' '.join(compat_shlex_quote(a
) for a
in args
)
4535 def error_to_compat_str(err
):
4537 # On python 2 error byte string must be decoded with proper
4538 # encoding rather than ascii
4539 if sys
.version_info
[0] < 3:
4540 err_str
= err_str
.decode(preferredencoding())
4544 def mimetype2ext(mt
):
4548 mt
, _
, params
= mt
.partition(';')
4553 # Per RFC 3003, audio/mpeg can be .mp1, .mp2 or .mp3. Here use .mp3 as
4554 # it's the most popular one
4555 'audio/mpeg': 'mp3',
4556 'audio/x-wav': 'wav',
4558 'audio/wave': 'wav',
4561 ext
= FULL_MAP
.get(mt
)
4567 'smptett+xml': 'tt',
4571 'x-mp4-fragmented': 'mp4',
4572 'x-ms-sami': 'sami',
4575 'x-mpegurl': 'm3u8',
4576 'vnd.apple.mpegurl': 'm3u8',
4580 'vnd.ms-sstr+xml': 'ism',
4584 'filmstrip+json': 'fs',
4588 _
, _
, subtype
= mt
.rpartition('/')
4589 ext
= SUBTYPE_MAP
.get(subtype
.lower())
4600 _
, _
, suffix
= subtype
.partition('+')
4601 ext
= SUFFIX_MAP
.get(suffix
)
4605 return subtype
.replace('+', '.')
4608 def parse_codecs(codecs_str
):
4609 # http://tools.ietf.org/html/rfc6381
4612 split_codecs
= list(filter(None, map(
4613 str.strip
, codecs_str
.strip().strip(',').split(','))))
4614 vcodec
, acodec
= None, None
4615 for full_codec
in split_codecs
:
4616 codec
= full_codec
.split('.')[0]
4617 if codec
in ('avc1', 'avc2', 'avc3', 'avc4', 'vp9', 'vp8', 'hev1', 'hev2', 'h263', 'h264', 'mp4v', 'hvc1', 'av01', 'theora'):
4620 elif codec
in ('mp4a', 'opus', 'vorbis', 'mp3', 'aac', 'ac-3', 'ec-3', 'eac3', 'dtsc', 'dtse', 'dtsh', 'dtsl'):
4624 write_string('WARNING: Unknown codec %s\n' % full_codec
, sys
.stderr
)
4625 if not vcodec
and not acodec
:
4626 if len(split_codecs
) == 2:
4628 'vcodec': split_codecs
[0],
4629 'acodec': split_codecs
[1],
4633 'vcodec': vcodec
or 'none',
4634 'acodec': acodec
or 'none',
4639 def urlhandle_detect_ext(url_handle
):
4640 getheader
= url_handle
.headers
.get
4642 cd
= getheader('Content-Disposition')
4644 m
= re
.match(r
'attachment;\s*filename="(?P<filename>[^"]+)"', cd
)
4646 e
= determine_ext(m
.group('filename'), default_ext
=None)
4650 return mimetype2ext(getheader('Content-Type'))
4653 def encode_data_uri(data
, mime_type
):
4654 return 'data:%s;base64,%s' % (mime_type
, base64
.b64encode(data
).decode('ascii'))
4657 def age_restricted(content_limit
, age_limit
):
4658 """ Returns True iff the content should be blocked """
4660 if age_limit
is None: # No limit set
4662 if content_limit
is None:
4663 return False # Content available for everyone
4664 return age_limit
< content_limit
4667 def is_html(first_bytes
):
4668 """ Detect whether a file contains HTML by examining its first bytes. """
4671 (b
'\xef\xbb\xbf', 'utf-8'),
4672 (b
'\x00\x00\xfe\xff', 'utf-32-be'),
4673 (b
'\xff\xfe\x00\x00', 'utf-32-le'),
4674 (b
'\xff\xfe', 'utf-16-le'),
4675 (b
'\xfe\xff', 'utf-16-be'),
4677 for bom
, enc
in BOMS
:
4678 if first_bytes
.startswith(bom
):
4679 s
= first_bytes
[len(bom
):].decode(enc
, 'replace')
4682 s
= first_bytes
.decode('utf-8', 'replace')
4684 return re
.match(r
'^\s*<', s
)
4687 def determine_protocol(info_dict
):
4688 protocol
= info_dict
.get('protocol')
4689 if protocol
is not None:
4692 url
= info_dict
['url']
4693 if url
.startswith('rtmp'):
4695 elif url
.startswith('mms'):
4697 elif url
.startswith('rtsp'):
4700 ext
= determine_ext(url
)
4706 return compat_urllib_parse_urlparse(url
).scheme
4709 def render_table(header_row
, data
, delim
=False, extraGap
=0, hideEmpty
=False):
4710 """ Render a list of rows, each as a list of values """
4712 def get_max_lens(table
):
4713 return [max(len(compat_str(v
)) for v
in col
) for col
in zip(*table
)]
4715 def filter_using_list(row
, filterArray
):
4716 return [col
for (take
, col
) in zip(filterArray
, row
) if take
]
4719 max_lens
= get_max_lens(data
)
4720 header_row
= filter_using_list(header_row
, max_lens
)
4721 data
= [filter_using_list(row
, max_lens
) for row
in data
]
4723 table
= [header_row
] + data
4724 max_lens
= get_max_lens(table
)
4726 table
= [header_row
] + [['-' * ml
for ml
in max_lens
]] + data
4727 format_str
= ' '.join('%-' + compat_str(ml
+ extraGap
) + 's' for ml
in max_lens
[:-1]) + ' %s'
4728 return '\n'.join(format_str
% tuple(row
) for row
in table
)
4731 def _match_one(filter_part
, dct
, incomplete
):
4732 # TODO: Generalize code with YoutubeDL._build_format_filter
4733 STRING_OPERATORS
= {
4734 '*=': operator
.contains
,
4735 '^=': lambda attr
, value
: attr
.startswith(value
),
4736 '$=': lambda attr
, value
: attr
.endswith(value
),
4737 '~=': lambda attr
, value
: re
.search(value
, attr
),
4739 COMPARISON_OPERATORS
= {
4741 '<=': operator
.le
, # "<=" must be defined above "<"
4748 operator_rex
= re
.compile(r
'''(?x)\s*
4750 \s*(?P<negation>!\s*)?(?P<op>%s)(?P<none_inclusive>\s*\?)?\s*
4752 (?P<intval>[0-9.]+(?:[kKmMgGtTpPeEzZyY]i?[Bb]?)?)|
4753 (?P<quote>["\'])(?P
<quotedstrval
>.+?
)(?P
=quote
)|
4757 ''' % '|'.join(map(re.escape, COMPARISON_OPERATORS.keys())))
4758 m = operator_rex.search(filter_part)
4760 unnegated_op = COMPARISON_OPERATORS[m.group('op')]
4761 if m.group('negation'):
4762 op = lambda attr, value: not unnegated_op(attr, value)
4765 actual_value = dct.get(m.group('key'))
4766 if (m.group('quotedstrval') is not None
4767 or m.group('strval') is not None
4768 # If the original field is a string and matching comparisonvalue is
4769 # a number we should respect the origin of the original field
4770 # and process comparison value as a string (see
4771 # https://github.com/ytdl-org/youtube-dl/issues/11082).
4772 or actual_value is not None and m.group('intval') is not None
4773 and isinstance(actual_value, compat_str)):
4774 comparison_value = m.group('quotedstrval') or m.group('strval') or m.group('intval')
4775 quote = m.group('quote')
4776 if quote is not None:
4777 comparison_value = comparison_value.replace(r'\%s' % quote, quote)
4779 if m.group('op') in STRING_OPERATORS:
4780 raise ValueError('Operator %s only supports string values!' % m.group('op'))
4782 comparison_value = int(m.group('intval'))
4784 comparison_value = parse_filesize(m.group('intval'))
4785 if comparison_value is None:
4786 comparison_value = parse_filesize(m.group('intval') + 'B')
4787 if comparison_value is None:
4789 'Invalid integer value %r in filter part %r' % (
4790 m.group('intval'), filter_part))
4791 if actual_value is None:
4792 return incomplete or m.group('none_inclusive')
4793 return op(actual_value, comparison_value)
4796 '': lambda v: (v is True) if isinstance(v, bool) else (v is not None),
4797 '!': lambda v: (v is False) if isinstance(v, bool) else (v is None),
4799 operator_rex = re.compile(r'''(?x
)\s
*
4800 (?P
<op
>%s)\s
*(?P
<key
>[a
-z_
]+)
4802 ''' % '|'.join(map(re.escape, UNARY_OPERATORS.keys())))
4803 m = operator_rex.search(filter_part)
4805 op = UNARY_OPERATORS[m.group('op')]
4806 actual_value = dct.get(m.group('key'))
4807 if incomplete and actual_value is None:
4809 return op(actual_value)
4811 raise ValueError('Invalid filter part %r' % filter_part)
4814 def match_str(filter_str, dct, incomplete=False):
4815 """ Filter a dictionary with a simple string syntax. Returns True (=passes filter) or false
4816 When incomplete, all conditions passes on missing fields
4819 _match_one(filter_part.replace(r'\&', '&'), dct, incomplete)
4820 for filter_part in re.split(r'(?<!\\)&', filter_str))
4823 def match_filter_func(filter_str):
4824 def _match_func(info_dict, *args, **kwargs):
4825 if match_str(filter_str, info_dict, *args, **kwargs):
4828 video_title = info_dict.get('title', info_dict.get('id', 'video'))
4829 return '%s does not pass filter %s, skipping ..' % (video_title, filter_str)
4833 def parse_dfxp_time_expr(time_expr):
4837 mobj = re.match(r'^(?P<time_offset>\d+(?:\.\d+)?)s?$', time_expr)
4839 return float(mobj.group('time_offset'))
4841 mobj = re.match(r'^(\d+):(\d\d):(\d\d(?:(?:\.|:)\d+)?)$', time_expr)
4843 return 3600 * int(mobj.group(1)) + 60 * int(mobj.group(2)) + float(mobj.group(3).replace(':', '.'))
4846 def srt_subtitles_timecode(seconds):
4847 return '%02d:%02d:%02d,%03d' % (seconds / 3600, (seconds % 3600) / 60, seconds % 60, (seconds % 1) * 1000)
4850 def dfxp2srt(dfxp_data):
4852 @param dfxp_data A
bytes-like
object containing DFXP data
4853 @returns A
unicode object containing converted SRT data
4855 LEGACY_NAMESPACES = (
4856 (b'http://www.w3.org/ns/ttml', [
4857 b'http://www.w3.org/2004/11/ttaf1',
4858 b'http://www.w3.org/2006/04/ttaf1',
4859 b'http://www.w3.org/2006/10/ttaf1',
4861 (b'http://www.w3.org/ns/ttml#styling', [
4862 b'http://www.w3.org/ns/ttml#style',
4866 SUPPORTED_STYLING = [
4875 _x = functools.partial(xpath_with_ns, ns_map={
4876 'xml': 'http://www.w3.org/XML/1998/namespace',
4877 'ttml': 'http://www.w3.org/ns/ttml',
4878 'tts': 'http://www.w3.org/ns/ttml#styling',
4884 class TTMLPElementParser(object):
4886 _unclosed_elements = []
4887 _applied_styles = []
4889 def start(self, tag, attrib):
4890 if tag in (_x('ttml:br'), 'br'):
4893 unclosed_elements = []
4895 element_style_id = attrib.get('style')
4897 style.update(default_style)
4898 if element_style_id:
4899 style.update(styles.get(element_style_id, {}))
4900 for prop in SUPPORTED_STYLING:
4901 prop_val = attrib.get(_x('tts:' + prop))
4903 style[prop] = prop_val
4906 for k, v in sorted(style.items()):
4907 if self._applied_styles and self._applied_styles[-1].get(k) == v:
4910 font += ' color="%s"' % v
4911 elif k == 'fontSize':
4912 font += ' size="%s"' % v
4913 elif k == 'fontFamily':
4914 font += ' face="%s"' % v
4915 elif k == 'fontWeight' and v == 'bold':
4917 unclosed_elements.append('b')
4918 elif k == 'fontStyle' and v == 'italic':
4920 unclosed_elements.append('i')
4921 elif k == 'textDecoration' and v == 'underline':
4923 unclosed_elements.append('u')
4925 self._out += '<font' + font + '>'
4926 unclosed_elements.append('font')
4928 if self._applied_styles:
4929 applied_style.update(self._applied_styles[-1])
4930 applied_style.update(style)
4931 self._applied_styles.append(applied_style)
4932 self._unclosed_elements.append(unclosed_elements)
4935 if tag not in (_x('ttml:br'), 'br'):
4936 unclosed_elements = self._unclosed_elements.pop()
4937 for element in reversed(unclosed_elements):
4938 self._out += '</%s>' % element
4939 if unclosed_elements and self._applied_styles:
4940 self._applied_styles.pop()
4942 def data(self, data):
4946 return self._out.strip()
4948 def parse_node(node):
4949 target = TTMLPElementParser()
4950 parser = xml.etree.ElementTree.XMLParser(target=target)
4951 parser.feed(xml.etree.ElementTree.tostring(node))
4952 return parser.close()
4954 for k, v in LEGACY_NAMESPACES:
4956 dfxp_data = dfxp_data.replace(ns, k)
4958 dfxp = compat_etree_fromstring(dfxp_data)
4960 paras = dfxp.findall(_x('.//ttml:p')) or dfxp.findall('.//p')
4963 raise ValueError('Invalid dfxp/TTML subtitle')
4967 for style in dfxp.findall(_x('.//ttml:style')):
4968 style_id = style.get('id') or style.get(_x('xml:id'))
4971 parent_style_id = style.get('style')
4973 if parent_style_id not in styles:
4976 styles[style_id] = styles[parent_style_id].copy()
4977 for prop in SUPPORTED_STYLING:
4978 prop_val = style.get(_x('tts:' + prop))
4980 styles.setdefault(style_id, {})[prop] = prop_val
4986 for p in ('body', 'div'):
4987 ele = xpath_element(dfxp, [_x('.//ttml:' + p), './/' + p])
4990 style = styles.get(ele.get('style'))
4993 default_style.update(style)
4995 for para, index in zip(paras, itertools.count(1)):
4996 begin_time = parse_dfxp_time_expr(para.attrib.get('begin'))
4997 end_time = parse_dfxp_time_expr(para.attrib.get('end'))
4998 dur = parse_dfxp_time_expr(para.attrib.get('dur'))
4999 if begin_time is None:
5004 end_time = begin_time + dur
5005 out.append('%d\n%s --> %s\n%s\n\n' % (
5007 srt_subtitles_timecode(begin_time),
5008 srt_subtitles_timecode(end_time),
5014 def cli_option(params, command_option, param):
5015 param = params.get(param)
5017 param = compat_str(param)
5018 return [command_option, param] if param is not None else []
5021 def cli_bool_option(params, command_option, param, true_value='true', false_value='false', separator=None):
5022 param = params.get(param)
5025 assert isinstance(param, bool)
5027 return [command_option + separator + (true_value if param else false_value)]
5028 return [command_option, true_value if param else false_value]
5031 def cli_valueless_option(params, command_option, param, expected_value=True):
5032 param = params.get(param)
5033 return [command_option] if param == expected_value else []
5036 def cli_configuration_args(argdict, keys, default=[], use_compat=True):
5037 if isinstance(argdict, (list, tuple)): # for backward compatibility
5044 assert isinstance(argdict, dict)
5046 assert isinstance(keys, (list, tuple))
5047 for key_list in keys:
5048 arg_list = list(filter(
5049 lambda x: x is not None,
5050 [argdict.get(key.lower()) for key in variadic(key_list)]))
5052 return [arg for args in arg_list for arg in args]
5056 def _configuration_args(main_key, argdict, exe, keys=None, default=[], use_compat=True):
5057 main_key, exe = main_key.lower(), exe.lower()
5058 root_key = exe if main_key == exe else f'{main_key}+{exe}'
5059 keys = [f'{root_key}{k}' for k in (keys or [''])]
5060 if root_key in keys:
5062 keys.append((main_key, exe))
5063 keys.append('default')
5066 return cli_configuration_args(argdict, keys, default, use_compat)
5069 class ISO639Utils(object):
5070 # See http://www.loc.gov/standards/iso639-2/ISO-639-2_utf-8.txt
5129 'iw': 'heb', # Replaced by he in 1989 revision
5139 'in': 'ind', # Replaced by id in 1989 revision
5254 'ji': 'yid', # Replaced by yi in 1989 revision
5262 def short2long(cls, code):
5263 """Convert language code from ISO 639-1 to ISO 639-2/T"""
5264 return cls._lang_map.get(code[:2])
5267 def long2short(cls, code):
5268 """Convert language code from ISO 639-2/T to ISO 639-1"""
5269 for short_name, long_name in cls._lang_map.items():
5270 if long_name == code:
5274 class ISO3166Utils(object):
5275 # From http://data.okfn.org/data/core/country-list
5277 'AF': 'Afghanistan',
5278 'AX': 'Åland Islands',
5281 'AS': 'American Samoa',
5286 'AG': 'Antigua and Barbuda',
5303 'BO': 'Bolivia, Plurinational State of',
5304 'BQ': 'Bonaire, Sint Eustatius and Saba',
5305 'BA': 'Bosnia and Herzegovina',
5307 'BV': 'Bouvet Island',
5309 'IO': 'British Indian Ocean Territory',
5310 'BN': 'Brunei Darussalam',
5312 'BF': 'Burkina Faso',
5318 'KY': 'Cayman Islands',
5319 'CF': 'Central African Republic',
5323 'CX': 'Christmas Island',
5324 'CC': 'Cocos (Keeling) Islands',
5328 'CD': 'Congo, the Democratic Republic of the',
5329 'CK': 'Cook Islands',
5331 'CI': 'Côte d\'Ivoire',
5336 'CZ': 'Czech Republic',
5340 'DO': 'Dominican Republic',
5343 'SV': 'El Salvador',
5344 'GQ': 'Equatorial Guinea',
5348 'FK': 'Falkland Islands (Malvinas)',
5349 'FO': 'Faroe Islands',
5353 'GF': 'French Guiana',
5354 'PF': 'French Polynesia',
5355 'TF': 'French Southern Territories',
5370 'GW': 'Guinea-Bissau',
5373 'HM': 'Heard Island and McDonald Islands',
5374 'VA': 'Holy See (Vatican City State)',
5381 'IR': 'Iran, Islamic Republic of',
5384 'IM': 'Isle of Man',
5394 'KP': 'Korea, Democratic People\'s Republic of',
5395 'KR': 'Korea, Republic of',
5398 'LA': 'Lao People\'s Democratic Republic',
5404 'LI': 'Liechtenstein',
5408 'MK': 'Macedonia, the Former Yugoslav Republic of',
5415 'MH': 'Marshall Islands',
5421 'FM': 'Micronesia, Federated States of',
5422 'MD': 'Moldova, Republic of',
5433 'NL': 'Netherlands',
5434 'NC': 'New Caledonia',
5435 'NZ': 'New Zealand',
5440 'NF': 'Norfolk Island',
5441 'MP': 'Northern Mariana Islands',
5446 'PS': 'Palestine, State of',
5448 'PG': 'Papua New Guinea',
5451 'PH': 'Philippines',
5455 'PR': 'Puerto Rico',
5459 'RU': 'Russian Federation',
5461 'BL': 'Saint Barthélemy',
5462 'SH': 'Saint Helena, Ascension and Tristan da Cunha',
5463 'KN': 'Saint Kitts and Nevis',
5464 'LC': 'Saint Lucia',
5465 'MF': 'Saint Martin (French part)',
5466 'PM': 'Saint Pierre and Miquelon',
5467 'VC': 'Saint Vincent and the Grenadines',
5470 'ST': 'Sao Tome and Principe',
5471 'SA': 'Saudi Arabia',
5475 'SL': 'Sierra Leone',
5477 'SX': 'Sint Maarten (Dutch part)',
5480 'SB': 'Solomon Islands',
5482 'ZA': 'South Africa',
5483 'GS': 'South Georgia and the South Sandwich Islands',
5484 'SS': 'South Sudan',
5489 'SJ': 'Svalbard and Jan Mayen',
5492 'CH': 'Switzerland',
5493 'SY': 'Syrian Arab Republic',
5494 'TW': 'Taiwan, Province of China',
5496 'TZ': 'Tanzania, United Republic of',
5498 'TL': 'Timor-Leste',
5502 'TT': 'Trinidad and Tobago',
5505 'TM': 'Turkmenistan',
5506 'TC': 'Turks and Caicos Islands',
5510 'AE': 'United Arab Emirates',
5511 'GB': 'United Kingdom',
5512 'US': 'United States',
5513 'UM': 'United States Minor Outlying Islands',
5517 'VE': 'Venezuela, Bolivarian Republic of',
5519 'VG': 'Virgin Islands, British',
5520 'VI': 'Virgin Islands, U.S.',
5521 'WF': 'Wallis and Futuna',
5522 'EH': 'Western Sahara',
5529 def short2full(cls, code):
5530 """Convert an ISO 3166-2 country code to the corresponding full name"""
5531 return cls._country_map.get(code.upper())
5534 class GeoUtils(object):
5535 # Major IPv4 address blocks per country
5537 'AD': '46.172.224.0/19',
5538 'AE': '94.200.0.0/13',
5539 'AF': '149.54.0.0/17',
5540 'AG': '209.59.64.0/18',
5541 'AI': '204.14.248.0/21',
5542 'AL': '46.99.0.0/16',
5543 'AM': '46.70.0.0/15',
5544 'AO': '105.168.0.0/13',
5545 'AP': '182.50.184.0/21',
5546 'AQ': '23.154.160.0/24',
5547 'AR': '181.0.0.0/12',
5548 'AS': '202.70.112.0/20',
5549 'AT': '77.116.0.0/14',
5550 'AU': '1.128.0.0/11',
5551 'AW': '181.41.0.0/18',
5552 'AX': '185.217.4.0/22',
5553 'AZ': '5.197.0.0/16',
5554 'BA': '31.176.128.0/17',
5555 'BB': '65.48.128.0/17',
5556 'BD': '114.130.0.0/16',
5558 'BF': '102.178.0.0/15',
5559 'BG': '95.42.0.0/15',
5560 'BH': '37.131.0.0/17',
5561 'BI': '154.117.192.0/18',
5562 'BJ': '137.255.0.0/16',
5563 'BL': '185.212.72.0/23',
5564 'BM': '196.12.64.0/18',
5565 'BN': '156.31.0.0/16',
5566 'BO': '161.56.0.0/16',
5567 'BQ': '161.0.80.0/20',
5568 'BR': '191.128.0.0/12',
5569 'BS': '24.51.64.0/18',
5570 'BT': '119.2.96.0/19',
5571 'BW': '168.167.0.0/16',
5572 'BY': '178.120.0.0/13',
5573 'BZ': '179.42.192.0/18',
5574 'CA': '99.224.0.0/11',
5575 'CD': '41.243.0.0/16',
5576 'CF': '197.242.176.0/21',
5577 'CG': '160.113.0.0/16',
5578 'CH': '85.0.0.0/13',
5579 'CI': '102.136.0.0/14',
5580 'CK': '202.65.32.0/19',
5581 'CL': '152.172.0.0/14',
5582 'CM': '102.244.0.0/14',
5583 'CN': '36.128.0.0/10',
5584 'CO': '181.240.0.0/12',
5585 'CR': '201.192.0.0/12',
5586 'CU': '152.206.0.0/15',
5587 'CV': '165.90.96.0/19',
5588 'CW': '190.88.128.0/17',
5589 'CY': '31.153.0.0/16',
5590 'CZ': '88.100.0.0/14',
5592 'DJ': '197.241.0.0/17',
5593 'DK': '87.48.0.0/12',
5594 'DM': '192.243.48.0/20',
5595 'DO': '152.166.0.0/15',
5596 'DZ': '41.96.0.0/12',
5597 'EC': '186.68.0.0/15',
5598 'EE': '90.190.0.0/15',
5599 'EG': '156.160.0.0/11',
5600 'ER': '196.200.96.0/20',
5601 'ES': '88.0.0.0/11',
5602 'ET': '196.188.0.0/14',
5603 'EU': '2.16.0.0/13',
5604 'FI': '91.152.0.0/13',
5605 'FJ': '144.120.0.0/16',
5606 'FK': '80.73.208.0/21',
5607 'FM': '119.252.112.0/20',
5608 'FO': '88.85.32.0/19',
5610 'GA': '41.158.0.0/15',
5612 'GD': '74.122.88.0/21',
5613 'GE': '31.146.0.0/16',
5614 'GF': '161.22.64.0/18',
5615 'GG': '62.68.160.0/19',
5616 'GH': '154.160.0.0/12',
5617 'GI': '95.164.0.0/16',
5618 'GL': '88.83.0.0/19',
5619 'GM': '160.182.0.0/15',
5620 'GN': '197.149.192.0/18',
5621 'GP': '104.250.0.0/19',
5622 'GQ': '105.235.224.0/20',
5623 'GR': '94.64.0.0/13',
5624 'GT': '168.234.0.0/16',
5625 'GU': '168.123.0.0/16',
5626 'GW': '197.214.80.0/20',
5627 'GY': '181.41.64.0/18',
5628 'HK': '113.252.0.0/14',
5629 'HN': '181.210.0.0/16',
5630 'HR': '93.136.0.0/13',
5631 'HT': '148.102.128.0/17',
5632 'HU': '84.0.0.0/14',
5633 'ID': '39.192.0.0/10',
5634 'IE': '87.32.0.0/12',
5635 'IL': '79.176.0.0/13',
5636 'IM': '5.62.80.0/20',
5637 'IN': '117.192.0.0/10',
5638 'IO': '203.83.48.0/21',
5639 'IQ': '37.236.0.0/14',
5640 'IR': '2.176.0.0/12',
5641 'IS': '82.221.0.0/16',
5642 'IT': '79.0.0.0/10',
5643 'JE': '87.244.64.0/18',
5644 'JM': '72.27.0.0/17',
5645 'JO': '176.29.0.0/16',
5646 'JP': '133.0.0.0/8',
5647 'KE': '105.48.0.0/12',
5648 'KG': '158.181.128.0/17',
5649 'KH': '36.37.128.0/17',
5650 'KI': '103.25.140.0/22',
5651 'KM': '197.255.224.0/20',
5652 'KN': '198.167.192.0/19',
5653 'KP': '175.45.176.0/22',
5654 'KR': '175.192.0.0/10',
5655 'KW': '37.36.0.0/14',
5656 'KY': '64.96.0.0/15',
5657 'KZ': '2.72.0.0/13',
5658 'LA': '115.84.64.0/18',
5659 'LB': '178.135.0.0/16',
5660 'LC': '24.92.144.0/20',
5661 'LI': '82.117.0.0/19',
5662 'LK': '112.134.0.0/15',
5663 'LR': '102.183.0.0/16',
5664 'LS': '129.232.0.0/17',
5665 'LT': '78.56.0.0/13',
5666 'LU': '188.42.0.0/16',
5667 'LV': '46.109.0.0/16',
5668 'LY': '41.252.0.0/14',
5669 'MA': '105.128.0.0/11',
5670 'MC': '88.209.64.0/18',
5671 'MD': '37.246.0.0/16',
5672 'ME': '178.175.0.0/17',
5673 'MF': '74.112.232.0/21',
5674 'MG': '154.126.0.0/17',
5675 'MH': '117.103.88.0/21',
5676 'MK': '77.28.0.0/15',
5677 'ML': '154.118.128.0/18',
5678 'MM': '37.111.0.0/17',
5679 'MN': '49.0.128.0/17',
5680 'MO': '60.246.0.0/16',
5681 'MP': '202.88.64.0/20',
5682 'MQ': '109.203.224.0/19',
5683 'MR': '41.188.64.0/18',
5684 'MS': '208.90.112.0/22',
5685 'MT': '46.11.0.0/16',
5686 'MU': '105.16.0.0/12',
5687 'MV': '27.114.128.0/18',
5688 'MW': '102.70.0.0/15',
5689 'MX': '187.192.0.0/11',
5690 'MY': '175.136.0.0/13',
5691 'MZ': '197.218.0.0/15',
5692 'NA': '41.182.0.0/16',
5693 'NC': '101.101.0.0/18',
5694 'NE': '197.214.0.0/18',
5695 'NF': '203.17.240.0/22',
5696 'NG': '105.112.0.0/12',
5697 'NI': '186.76.0.0/15',
5698 'NL': '145.96.0.0/11',
5699 'NO': '84.208.0.0/13',
5700 'NP': '36.252.0.0/15',
5701 'NR': '203.98.224.0/19',
5702 'NU': '49.156.48.0/22',
5703 'NZ': '49.224.0.0/14',
5704 'OM': '5.36.0.0/15',
5705 'PA': '186.72.0.0/15',
5706 'PE': '186.160.0.0/14',
5707 'PF': '123.50.64.0/18',
5708 'PG': '124.240.192.0/19',
5709 'PH': '49.144.0.0/13',
5710 'PK': '39.32.0.0/11',
5711 'PL': '83.0.0.0/11',
5712 'PM': '70.36.0.0/20',
5713 'PR': '66.50.0.0/16',
5714 'PS': '188.161.0.0/16',
5715 'PT': '85.240.0.0/13',
5716 'PW': '202.124.224.0/20',
5717 'PY': '181.120.0.0/14',
5718 'QA': '37.210.0.0/15',
5719 'RE': '102.35.0.0/16',
5720 'RO': '79.112.0.0/13',
5721 'RS': '93.86.0.0/15',
5722 'RU': '5.136.0.0/13',
5723 'RW': '41.186.0.0/16',
5724 'SA': '188.48.0.0/13',
5725 'SB': '202.1.160.0/19',
5726 'SC': '154.192.0.0/11',
5727 'SD': '102.120.0.0/13',
5728 'SE': '78.64.0.0/12',
5729 'SG': '8.128.0.0/10',
5730 'SI': '188.196.0.0/14',
5731 'SK': '78.98.0.0/15',
5732 'SL': '102.143.0.0/17',
5733 'SM': '89.186.32.0/19',
5734 'SN': '41.82.0.0/15',
5735 'SO': '154.115.192.0/18',
5736 'SR': '186.179.128.0/17',
5737 'SS': '105.235.208.0/21',
5738 'ST': '197.159.160.0/19',
5739 'SV': '168.243.0.0/16',
5740 'SX': '190.102.0.0/20',
5742 'SZ': '41.84.224.0/19',
5743 'TC': '65.255.48.0/20',
5744 'TD': '154.68.128.0/19',
5745 'TG': '196.168.0.0/14',
5746 'TH': '171.96.0.0/13',
5747 'TJ': '85.9.128.0/18',
5748 'TK': '27.96.24.0/21',
5749 'TL': '180.189.160.0/20',
5750 'TM': '95.85.96.0/19',
5751 'TN': '197.0.0.0/11',
5752 'TO': '175.176.144.0/21',
5753 'TR': '78.160.0.0/11',
5754 'TT': '186.44.0.0/15',
5755 'TV': '202.2.96.0/19',
5756 'TW': '120.96.0.0/11',
5757 'TZ': '156.156.0.0/14',
5758 'UA': '37.52.0.0/14',
5759 'UG': '102.80.0.0/13',
5761 'UY': '167.56.0.0/13',
5762 'UZ': '84.54.64.0/18',
5763 'VA': '212.77.0.0/19',
5764 'VC': '207.191.240.0/21',
5765 'VE': '186.88.0.0/13',
5766 'VG': '66.81.192.0/20',
5767 'VI': '146.226.0.0/16',
5768 'VN': '14.160.0.0/11',
5769 'VU': '202.80.32.0/20',
5770 'WF': '117.20.32.0/21',
5771 'WS': '202.4.32.0/19',
5772 'YE': '134.35.0.0/16',
5773 'YT': '41.242.116.0/22',
5774 'ZA': '41.0.0.0/11',
5775 'ZM': '102.144.0.0/13',
5776 'ZW': '102.177.192.0/18',
5780 def random_ipv4(cls, code_or_block):
5781 if len(code_or_block) == 2:
5782 block = cls._country_ip_map.get(code_or_block.upper())
5786 block = code_or_block
5787 addr, preflen = block.split('/')
5788 addr_min = compat_struct_unpack('!L', socket.inet_aton(addr))[0]
5789 addr_max = addr_min | (0xffffffff >> int(preflen))
5790 return compat_str(socket.inet_ntoa(
5791 compat_struct_pack('!L', random.randint(addr_min, addr_max))))
5794 class PerRequestProxyHandler(compat_urllib_request.ProxyHandler):
5795 def __init__(self, proxies=None):
5796 # Set default handlers
5797 for type in ('http', 'https'):
5798 setattr(self, '%s_open' % type,
5799 lambda r, proxy='__noproxy__', type=type, meth=self.proxy_open:
5800 meth(r, proxy, type))
5801 compat_urllib_request.ProxyHandler.__init__(self, proxies)
5803 def proxy_open(self, req, proxy, type):
5804 req_proxy = req.headers.get('Ytdl-request-proxy')
5805 if req_proxy is not None:
5807 del req.headers['Ytdl-request-proxy']
5809 if proxy == '__noproxy__':
5810 return None # No Proxy
5811 if compat_urlparse.urlparse(proxy).scheme.lower() in ('socks', 'socks4', 'socks4a', 'socks5'):
5812 req.add_header('Ytdl-socks-proxy', proxy)
5813 # yt-dlp's http/https handlers do wrapping the socket with socks
5815 return compat_urllib_request.ProxyHandler.proxy_open(
5816 self, req, proxy, type)
5819 # Both long_to_bytes and bytes_to_long are adapted from PyCrypto, which is
5820 # released into Public Domain
5821 # https://github.com/dlitz/pycrypto/blob/master/lib/Crypto/Util/number.py#L387
5823 def long_to_bytes(n, blocksize=0):
5824 """long_to_bytes(n:long, blocksize:int) : string
5825 Convert a long integer to a byte string.
5827 If optional blocksize is given and greater than zero, pad the front of the
5828 byte string with binary zeros so that the length is a multiple of
5831 # after much testing, this algorithm was deemed to be the fastest
5835 s = compat_struct_pack('>I', n & 0xffffffff) + s
5837 # strip off leading zeros
5838 for i in range(len(s)):
5839 if s[i] != b'\000'[0]:
5842 # only happens when n == 0
5846 # add back some pad bytes. this could be done more efficiently w.r.t. the
5847 # de-padding being done above, but sigh...
5848 if blocksize > 0 and len(s) % blocksize:
5849 s = (blocksize - len(s) % blocksize) * b'\000' + s
5853 def bytes_to_long(s):
5854 """bytes_to_long(string) : long
5855 Convert a byte string to a long integer.
5857 This is (essentially) the inverse of long_to_bytes().
5862 extra = (4 - length % 4)
5863 s = b'\000' * extra + s
5864 length = length + extra
5865 for i in range(0, length, 4):
5866 acc = (acc << 32) + compat_struct_unpack('>I', s[i:i + 4])[0]
5870 def ohdave_rsa_encrypt(data, exponent, modulus):
5872 Implement OHDave
's RSA algorithm. See http://www.ohdave.com/rsa/
5875 data: data to encrypt, bytes-like object
5876 exponent, modulus: parameter e and N of RSA algorithm, both integer
5877 Output: hex string of encrypted data
5879 Limitation: supports one block encryption only
5882 payload = int(binascii.hexlify(data[::-1]), 16)
5883 encrypted = pow(payload, exponent, modulus)
5884 return '%x' % encrypted
5887 def pkcs1pad(data, length):
5889 Padding input data with PKCS#1 scheme
5891 @param {int[]} data input data
5892 @param {int} length target length
5893 @returns {int[]} padded data
5895 if len(data) > length - 11:
5896 raise ValueError('Input data too
long for PKCS
#1 padding')
5898 pseudo_random
= [random
.randint(0, 254) for _
in range(length
- len(data
) - 3)]
5899 return [0, 2] + pseudo_random
+ [0] + data
5902 def encode_base_n(num
, n
, table
=None):
5903 FULL_TABLE
= '0123456789abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ'
5905 table
= FULL_TABLE
[:n
]
5908 raise ValueError('base %d exceeds table length %d' % (n
, len(table
)))
5915 ret
= table
[num
% n
] + ret
5920 def decode_packed_codes(code
):
5921 mobj
= re
.search(PACKED_CODES_RE
, code
)
5922 obfuscated_code
, base
, count
, symbols
= mobj
.groups()
5925 symbols
= symbols
.split('|')
5930 base_n_count
= encode_base_n(count
, base
)
5931 symbol_table
[base_n_count
] = symbols
[count
] or base_n_count
5934 r
'\b(\w+)\b', lambda mobj
: symbol_table
[mobj
.group(0)],
5938 def caesar(s
, alphabet
, shift
):
5943 alphabet
[(alphabet
.index(c
) + shift
) % l
] if c
in alphabet
else c
5948 return caesar(s
, r
'''!"#$%&'()*+,-./0123456789:;<=>?@ABCDEFGHIJKLMNOPQRSTUVWXYZ[\]^_`abcdefghijklmnopqrstuvwxyz{|}~''', 47)
5951 def parse_m3u8_attributes(attrib
):
5953 for (key
, val
) in re
.findall(r
'(?P<key>[A-Z0-9-]+)=(?P<val>"[^"]+"|[^",]+)(?:,|$)', attrib
):
5954 if val
.startswith('"'):
5960 def urshift(val
, n
):
5961 return val
>> n
if val
>= 0 else (val
+ 0x100000000) >> n
5964 # Based on png2str() written by @gdkchan and improved by @yokrysty
5965 # Originally posted at https://github.com/ytdl-org/youtube-dl/issues/9706
5966 def decode_png(png_data
):
5967 # Reference: https://www.w3.org/TR/PNG/
5968 header
= png_data
[8:]
5970 if png_data
[:8] != b
'\x89PNG\x0d\x0a\x1a\x0a' or header
[4:8] != b
'IHDR':
5971 raise IOError('Not a valid PNG file.')
5973 int_map
= {1: '>B', 2: '>H', 4: '>I'}
5974 unpack_integer
= lambda x
: compat_struct_unpack(int_map
[len(x
)], x
)[0]
5979 length
= unpack_integer(header
[:4])
5982 chunk_type
= header
[:4]
5985 chunk_data
= header
[:length
]
5986 header
= header
[length
:]
5988 header
= header
[4:] # Skip CRC
5996 ihdr
= chunks
[0]['data']
5998 width
= unpack_integer(ihdr
[:4])
5999 height
= unpack_integer(ihdr
[4:8])
6003 for chunk
in chunks
:
6004 if chunk
['type'] == b
'IDAT':
6005 idat
+= chunk
['data']
6008 raise IOError('Unable to read PNG data.')
6010 decompressed_data
= bytearray(zlib
.decompress(idat
))
6015 def _get_pixel(idx
):
6020 for y
in range(height
):
6021 basePos
= y
* (1 + stride
)
6022 filter_type
= decompressed_data
[basePos
]
6026 pixels
.append(current_row
)
6028 for x
in range(stride
):
6029 color
= decompressed_data
[1 + basePos
+ x
]
6030 basex
= y
* stride
+ x
6035 left
= _get_pixel(basex
- 3)
6037 up
= _get_pixel(basex
- stride
)
6039 if filter_type
== 1: # Sub
6040 color
= (color
+ left
) & 0xff
6041 elif filter_type
== 2: # Up
6042 color
= (color
+ up
) & 0xff
6043 elif filter_type
== 3: # Average
6044 color
= (color
+ ((left
+ up
) >> 1)) & 0xff
6045 elif filter_type
== 4: # Paeth
6051 c
= _get_pixel(basex
- stride
- 3)
6059 if pa
<= pb
and pa
<= pc
:
6060 color
= (color
+ a
) & 0xff
6062 color
= (color
+ b
) & 0xff
6064 color
= (color
+ c
) & 0xff
6066 current_row
.append(color
)
6068 return width
, height
, pixels
6071 def write_xattr(path
, key
, value
):
6072 # This mess below finds the best xattr tool for the job
6074 # try the pyxattr module...
6077 if hasattr(xattr
, 'set'): # pyxattr
6078 # Unicode arguments are not supported in python-pyxattr until
6080 # See https://github.com/ytdl-org/youtube-dl/issues/5498
6081 pyxattr_required_version
= '0.5.0'
6082 if version_tuple(xattr
.__version
__) < version_tuple(pyxattr_required_version
):
6083 # TODO: fallback to CLI tools
6084 raise XAttrUnavailableError(
6085 'python-pyxattr is detected but is too old. '
6086 'yt-dlp requires %s or above while your version is %s. '
6087 'Falling back to other xattr implementations' % (
6088 pyxattr_required_version
, xattr
.__version
__))
6090 setxattr
= xattr
.set
6092 setxattr
= xattr
.setxattr
6095 setxattr(path
, key
, value
)
6096 except EnvironmentError as e
:
6097 raise XAttrMetadataError(e
.errno
, e
.strerror
)
6100 if compat_os_name
== 'nt':
6101 # Write xattrs to NTFS Alternate Data Streams:
6102 # http://en.wikipedia.org/wiki/NTFS#Alternate_data_streams_.28ADS.29
6103 assert ':' not in key
6104 assert os
.path
.exists(path
)
6106 ads_fn
= path
+ ':' + key
6108 with open(ads_fn
, 'wb') as f
:
6110 except EnvironmentError as e
:
6111 raise XAttrMetadataError(e
.errno
, e
.strerror
)
6113 user_has_setfattr
= check_executable('setfattr', ['--version'])
6114 user_has_xattr
= check_executable('xattr', ['-h'])
6116 if user_has_setfattr
or user_has_xattr
:
6118 value
= value
.decode('utf-8')
6119 if user_has_setfattr
:
6120 executable
= 'setfattr'
6121 opts
= ['-n', key
, '-v', value
]
6122 elif user_has_xattr
:
6123 executable
= 'xattr'
6124 opts
= ['-w', key
, value
]
6126 cmd
= ([encodeFilename(executable
, True)]
6127 + [encodeArgument(o
) for o
in opts
]
6128 + [encodeFilename(path
, True)])
6131 p
= subprocess
.Popen(
6132 cmd
, stdout
=subprocess
.PIPE
, stderr
=subprocess
.PIPE
, stdin
=subprocess
.PIPE
)
6133 except EnvironmentError as e
:
6134 raise XAttrMetadataError(e
.errno
, e
.strerror
)
6135 stdout
, stderr
= process_communicate_or_kill(p
)
6136 stderr
= stderr
.decode('utf-8', 'replace')
6137 if p
.returncode
!= 0:
6138 raise XAttrMetadataError(p
.returncode
, stderr
)
6141 # On Unix, and can't find pyxattr, setfattr, or xattr.
6142 if sys
.platform
.startswith('linux'):
6143 raise XAttrUnavailableError(
6144 "Couldn't find a tool to set the xattrs. "
6145 "Install either the python 'pyxattr' or 'xattr' "
6146 "modules, or the GNU 'attr' package "
6147 "(which contains the 'setfattr' tool).")
6149 raise XAttrUnavailableError(
6150 "Couldn't find a tool to set the xattrs. "
6151 "Install either the python 'xattr' module, "
6152 "or the 'xattr' binary.")
6155 def random_birthday(year_field
, month_field
, day_field
):
6156 start_date
= datetime
.date(1950, 1, 1)
6157 end_date
= datetime
.date(1995, 12, 31)
6158 offset
= random
.randint(0, (end_date
- start_date
).days
)
6159 random_date
= start_date
+ datetime
.timedelta(offset
)
6161 year_field
: str(random_date
.year
),
6162 month_field
: str(random_date
.month
),
6163 day_field
: str(random_date
.day
),
6167 # Templates for internet shortcut files, which are plain text files.
6168 DOT_URL_LINK_TEMPLATE
= '''
6173 DOT_WEBLOC_LINK_TEMPLATE
= '''
6174 <?xml version="1.0" encoding="UTF-8"?>
6175 <!DOCTYPE plist PUBLIC "-//Apple//DTD PLIST 1.0//EN" "http://www.apple.com/DTDs/PropertyList-1.0.dtd">
6176 <plist version="1.0">
6179 \t<string>%(url)s</string>
6184 DOT_DESKTOP_LINK_TEMPLATE
= '''
6194 def iri_to_uri(iri
):
6196 Converts an IRI (Internationalized Resource Identifier, allowing Unicode characters) to a URI (Uniform Resource Identifier, ASCII-only).
6198 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.
6201 iri_parts
= compat_urllib_parse_urlparse(iri
)
6203 if '[' in iri_parts
.netloc
:
6204 raise ValueError('IPv6 URIs are not, yet, supported.')
6205 # Querying `.netloc`, when there's only one bracket, also raises a ValueError.
6207 # 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.
6210 if iri_parts
.username
:
6211 net_location
+= compat_urllib_parse_quote(iri_parts
.username
, safe
=r
"!$%&'()*+,~")
6212 if iri_parts
.password
is not None:
6213 net_location
+= ':' + compat_urllib_parse_quote(iri_parts
.password
, safe
=r
"!$%&'()*+,~")
6216 net_location
+= iri_parts
.hostname
.encode('idna').decode('utf-8') # Punycode for Unicode hostnames.
6217 # The 'idna' encoding produces ASCII text.
6218 if iri_parts
.port
is not None and iri_parts
.port
!= 80:
6219 net_location
+= ':' + str(iri_parts
.port
)
6221 return compat_urllib_parse_urlunparse(
6225 compat_urllib_parse_quote_plus(iri_parts
.path
, safe
=r
"!$%&'()*+,/:;=@|~"),
6227 # Unsure about the `safe` argument, since this is a legacy way of handling parameters.
6228 compat_urllib_parse_quote_plus(iri_parts
.params
, safe
=r
"!$%&'()*+,/:;=@|~"),
6230 # Not totally sure about the `safe` argument, since the source does not explicitly mention the query URI component.
6231 compat_urllib_parse_quote_plus(iri_parts
.query
, safe
=r
"!$%&'()*+,/:;=?@{|}~"),
6233 compat_urllib_parse_quote_plus(iri_parts
.fragment
, safe
=r
"!#$%&'()*+,/:;=?@{|}~")))
6235 # Source for `safe` arguments: https://url.spec.whatwg.org/#percent-encoded-bytes.
6238 def to_high_limit_path(path
):
6239 if sys
.platform
in ['win32', 'cygwin']:
6240 # Work around MAX_PATH limitation on Windows. The maximum allowed length for the individual path segments may still be quite limited.
6241 return r
'\\?\ '.rstrip() + os
.path
.abspath(path
)
6246 def format_field(obj
, field
=None, template
='%s', ignore
=(None, ''), default
='', func
=None):
6248 val
= obj
if obj
is not None else default
6250 val
= obj
.get(field
, default
)
6251 if func
and val
not in ignore
:
6253 return template
% val
if val
not in ignore
else default
6256 def clean_podcast_url(url
):
6257 return re
.sub(r
'''(?x)
6261 media\.blubrry\.com| # https://create.blubrry.com/resources/podcast-media-download-statistics/getting-started/
6264 (?:dts|www)\.podtrac\.com/(?:pts/)?redirect\.[0-9a-z]{3,4}| # http://analytics.podtrac.com/how-to-measure
6267 cn\.co| # https://podcorn.com/analytics-prefix/
6268 st\.fm # https://podsights.com/docs/
6273 _HEX_TABLE
= '0123456789abcdef'
6276 def random_uuidv4():
6277 return re
.sub(r
'[xy]', lambda x
: _HEX_TABLE
[random
.randint(0, 15)], 'xxxxxxxx-xxxx-4xxx-yxxx-xxxxxxxxxxxx')
6280 def make_dir(path
, to_screen
=None):
6282 dn
= os
.path
.dirname(path
)
6283 if dn
and not os
.path
.exists(dn
):
6286 except (OSError, IOError) as err
:
6287 if callable(to_screen
) is not None:
6288 to_screen('unable to create directory ' + error_to_compat_str(err
))
6292 def get_executable_path():
6293 from zipimport
import zipimporter
6294 if hasattr(sys
, 'frozen'): # Running from PyInstaller
6295 path
= os
.path
.dirname(sys
.executable
)
6296 elif isinstance(globals().get('__loader__'), zipimporter
): # Running from ZIP
6297 path
= os
.path
.join(os
.path
.dirname(__file__
), '../..')
6299 path
= os
.path
.join(os
.path
.dirname(__file__
), '..')
6300 return os
.path
.abspath(path
)
6303 def load_plugins(name
, suffix
, namespace
):
6304 plugin_info
= [None]
6307 plugin_info
= imp
.find_module(
6308 name
, [os
.path
.join(get_executable_path(), 'ytdlp_plugins')])
6309 plugins
= imp
.load_module(name
, *plugin_info
)
6310 for name
in dir(plugins
):
6311 if name
in namespace
:
6313 if not name
.endswith(suffix
):
6315 klass
= getattr(plugins
, name
)
6316 classes
[name
] = namespace
[name
] = klass
6320 if plugin_info
[0] is not None:
6321 plugin_info
[0].close()
6326 obj
, *path_list
, default
=None, expected_type
=None, get_all
=True,
6327 casesense
=True, is_user_input
=False, traverse_string
=False):
6328 ''' Traverse nested list/dict/tuple
6329 @param path_list A list of paths which are checked one by one.
6330 Each path is a list of keys where each key is a string,
6331 a tuple of strings or "...". When a tuple is given,
6332 all the keys given in the tuple are traversed, and
6333 "..." traverses all the keys in the object
6334 @param default Default value to return
6335 @param expected_type Only accept final value of this type (Can also be any callable)
6336 @param get_all Return all the values obtained from a path or only the first one
6337 @param casesense Whether to consider dictionary keys as case sensitive
6338 @param is_user_input Whether the keys are generated from user input. If True,
6339 strings are converted to int/slice if necessary
6340 @param traverse_string Whether to traverse inside strings. If True, any
6341 non-compatible object will also be converted into a string
6345 _lower
= lambda k
: (k
.lower() if isinstance(k
, str) else k
)
6346 path_list
= (map(_lower
, variadic(path
)) for path
in path_list
)
6348 def _traverse_obj(obj
, path
, _current_depth
=0):
6352 path
= tuple(variadic(path
))
6353 for i
, key
in enumerate(path
):
6354 if isinstance(key
, (list, tuple)):
6355 obj
= [_traverse_obj(obj
, sub_key
, _current_depth
) for sub_key
in key
]
6358 obj
= (obj
.values() if isinstance(obj
, dict)
6359 else obj
if isinstance(obj
, (list, tuple, LazyList
))
6360 else str(obj
) if traverse_string
else [])
6362 depth
= max(depth
, _current_depth
)
6363 return [_traverse_obj(inner_obj
, path
[i
+ 1:], _current_depth
) for inner_obj
in obj
]
6364 elif isinstance(obj
, dict) and not (is_user_input
and key
== ':'):
6365 obj
= (obj
.get(key
) if casesense
or (key
in obj
)
6366 else next((v
for k
, v
in obj
.items() if _lower(k
) == key
), None))
6369 key
= (int_or_none(key
) if ':' not in key
6370 else slice(*map(int_or_none
, key
.split(':'))))
6371 if key
== slice(None):
6372 return _traverse_obj(obj
, (..., *path
[i
+ 1:]), _current_depth
)
6373 if not isinstance(key
, (int, slice)):
6375 if not isinstance(obj
, (list, tuple, LazyList
)):
6376 if not traverse_string
:
6385 if isinstance(expected_type
, type):
6386 type_test
= lambda val
: val
if isinstance(val
, expected_type
) else None
6387 elif expected_type
is not None:
6388 type_test
= expected_type
6390 type_test
= lambda val
: val
6392 for path
in path_list
:
6394 val
= _traverse_obj(obj
, path
)
6397 for _
in range(depth
- 1):
6398 val
= itertools
.chain
.from_iterable(v
for v
in val
if v
is not None)
6399 val
= [v
for v
in map(type_test
, val
) if v
is not None]
6401 return val
if get_all
else val
[0]
6403 val
= type_test(val
)
6409 def traverse_dict(dictn
, keys
, casesense
=True):
6410 ''' For backward compatibility. Do not use '''
6411 return traverse_obj(dictn
, keys
, casesense
=casesense
,
6412 is_user_input
=True, traverse_string
=True)
6415 def variadic(x
, allowed_types
=(str, bytes)):
6416 return x
if isinstance(x
, collections
.abc
.Iterable
) and not isinstance(x
, allowed_types
) else (x
,)
6419 # create a JSON Web Signature (jws) with HS256 algorithm
6420 # the resulting format is in JWS Compact Serialization
6421 # implemented following JWT https://www.rfc-editor.org/rfc/rfc7519.html
6422 # implemented following JWS https://www.rfc-editor.org/rfc/rfc7515.html
6423 def jwt_encode_hs256(payload_data
, key
, headers
={}):
6429 header_data
.update(headers
)
6430 header_b64
= base64
.b64encode(json
.dumps(header_data
).encode('utf-8'))
6431 payload_b64
= base64
.b64encode(json
.dumps(payload_data
).encode('utf-8'))
6432 h
= hmac
.new(key
.encode('utf-8'), header_b64
+ b
'.' + payload_b64
, hashlib
.sha256
)
6433 signature_b64
= base64
.b64encode(h
.digest())
6434 token
= header_b64
+ b
'.' + payload_b64
+ b
'.' + signature_b64
projects / yt-dlp.git / blob