4 from __future__
import unicode_literals
37 import xml
.etree
.ElementTree
41 compat_HTMLParseError
,
47 compat_ctypes_WINFUNCTYPE
,
48 compat_etree_fromstring
,
51 compat_html_entities_html5
,
64 compat_urllib_parse_urlencode
,
65 compat_urllib_parse_urlparse
,
66 compat_urllib_parse_urlunparse
,
67 compat_urllib_parse_quote
,
68 compat_urllib_parse_quote_plus
,
69 compat_urllib_parse_unquote_plus
,
70 compat_urllib_request
,
81 def register_socks_protocols():
82 # "Register" SOCKS protocols
83 # In Python < 2.6.5, urlsplit() suffers from bug https://bugs.python.org/issue7904
84 # URLs with protocols not in urlparse.uses_netloc are not handled correctly
85 for scheme
in ('socks', 'socks4', 'socks4a', 'socks5'):
86 if scheme
not in compat_urlparse
.uses_netloc
:
87 compat_urlparse
.uses_netloc
.append(scheme
)
90 # This is not clearly defined otherwise
91 compiled_regex_type
= type(re
.compile(''))
94 def random_user_agent():
95 _USER_AGENT_TPL
= 'Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/%s Safari/537.36'
1674 return _USER_AGENT_TPL
% random
.choice(_CHROME_VERSIONS
)
1678 'User-Agent': random_user_agent(),
1679 'Accept-Charset': 'ISO-8859-1,utf-8;q=0.7,*;q=0.7',
1680 'Accept': 'text/html,application/xhtml+xml,application/xml;q=0.9,*/*;q=0.8',
1681 'Accept-Encoding': 'gzip, deflate',
1682 'Accept-Language': 'en-us,en;q=0.5',
1687 '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',
1691 NO_DEFAULT
= object()
1693 ENGLISH_MONTH_NAMES
= [
1694 'January', 'February', 'March', 'April', 'May', 'June',
1695 'July', 'August', 'September', 'October', 'November', 'December']
1698 'en': ENGLISH_MONTH_NAMES
,
1700 'janvier', 'février', 'mars', 'avril', 'mai', 'juin',
1701 'juillet', 'août', 'septembre', 'octobre', 'novembre', 'décembre'],
1704 KNOWN_EXTENSIONS
= (
1705 'mp4', 'm4a', 'm4p', 'm4b', 'm4r', 'm4v', 'aac',
1706 'flv', 'f4v', 'f4a', 'f4b',
1707 'webm', 'ogg', 'ogv', 'oga', 'ogx', 'spx', 'opus',
1708 'mkv', 'mka', 'mk3d',
1711 'asf', 'wmv', 'wma',
1717 'f4f', 'f4m', 'm3u8', 'smil')
1719 # needed for sanitizing filenames in restricted mode
1720 ACCENT_CHARS
= dict(zip('ÂÃÄÀÁÅÆÇÈÉÊËÌÍÎÏÐÑÒÓÔÕÖŐØŒÙÚÛÜŰÝÞßàáâãäåæçèéêëìíîïðñòóôõöőøœùúûüűýþÿ',
1721 itertools
.chain('AAAAAA', ['AE'], 'CEEEEIIIIDNOOOOOOO', ['OE'], 'UUUUUY', ['TH', 'ss'],
1722 'aaaaaa', ['ae'], 'ceeeeiiiionooooooo', ['oe'], 'uuuuuy', ['th'], 'y')))
1745 '%Y/%m/%d %H:%M:%S',
1749 '%Y-%m-%d %H:%M:%S',
1750 '%Y-%m-%d %H:%M:%S.%f',
1751 '%Y-%m-%d %H:%M:%S:%f',
1754 '%Y-%m-%dT%H:%M:%SZ',
1755 '%Y-%m-%dT%H:%M:%S.%fZ',
1756 '%Y-%m-%dT%H:%M:%S.%f0Z',
1757 '%Y-%m-%dT%H:%M:%S',
1758 '%Y-%m-%dT%H:%M:%S.%f',
1760 '%b %d %Y at %H:%M',
1761 '%b %d %Y at %H:%M:%S',
1762 '%B %d %Y at %H:%M',
1763 '%B %d %Y at %H:%M:%S',
1766 DATE_FORMATS_DAY_FIRST
= list(DATE_FORMATS
)
1767 DATE_FORMATS_DAY_FIRST
.extend([
1773 '%d/%m/%Y %H:%M:%S',
1776 DATE_FORMATS_MONTH_FIRST
= list(DATE_FORMATS
)
1777 DATE_FORMATS_MONTH_FIRST
.extend([
1782 '%m/%d/%Y %H:%M:%S',
1785 PACKED_CODES_RE
= r
"}\('(.+)',(\d+),(\d+),'([^']+)'\.split\('\|'\)"
1786 JSON_LD_RE
= r
'(?is)<script[^>]+type=(["\']?
)application
/ld\
+json\
1[^
>]*>(?P
<json_ld
>.+?
)</script
>'
1789 def preferredencoding():
1790 """Get preferred encoding.
1792 Returns the best encoding scheme for the system, based on
1793 locale.getpreferredencoding() and some further tweaks.
1796 pref = locale.getpreferredencoding()
1804 def write_json_file(obj, fn):
1805 """ Encode obj as JSON and write it to fn, atomically if possible """
1807 fn = encodeFilename(fn)
1808 if sys.version_info < (3, 0) and sys.platform != 'win32
':
1809 encoding = get_filesystem_encoding()
1810 # os.path.basename returns a bytes object, but NamedTemporaryFile
1811 # will fail if the filename contains non ascii characters unless we
1812 # use a unicode object
1813 path_basename = lambda f: os.path.basename(fn).decode(encoding)
1814 # the same for os.path.dirname
1815 path_dirname = lambda f: os.path.dirname(fn).decode(encoding)
1817 path_basename = os.path.basename
1818 path_dirname = os.path.dirname
1822 'prefix
': path_basename(fn) + '.',
1823 'dir': path_dirname(fn),
1827 # In Python 2.x, json.dump expects a bytestream.
1828 # In Python 3.x, it writes to a character stream
1829 if sys.version_info < (3, 0):
1834 'encoding
': 'utf
-8',
1837 tf = tempfile.NamedTemporaryFile(**compat_kwargs(args))
1842 if sys.platform == 'win32
':
1843 # Need to remove existing file on Windows, else os.rename raises
1844 # WindowsError or FileExistsError.
1852 os.chmod(tf.name, 0o666 & ~mask)
1855 os.rename(tf.name, fn)
1864 if sys.version_info >= (2, 7):
1865 def find_xpath_attr(node, xpath, key, val=None):
1866 """ Find the xpath xpath[@key=val] """
1867 assert re.match(r'^
[a
-zA
-Z_
-]+$
', key)
1868 expr = xpath + ('[@%s]' % key if val is None else "[@%s='%s']" % (key, val))
1869 return node.find(expr)
1871 def find_xpath_attr(node, xpath, key, val=None):
1872 for f in node.findall(compat_xpath(xpath)):
1873 if key not in f.attrib:
1875 if val is None or f.attrib.get(key) == val:
1879 # On python2.6 the xml.etree.ElementTree.Element methods don't support
1880 # the namespace parameter
1883 def xpath_with_ns(path
, ns_map
):
1884 components
= [c
.split(':') for c
in path
.split('/')]
1886 for c
in components
:
1888 replaced
.append(c
[0])
1891 replaced
.append('{%s}%s' % (ns_map
[ns
], tag
))
1892 return '/'.join(replaced
)
1895 def xpath_element(node
, xpath
, name
=None, fatal
=False, default
=NO_DEFAULT
):
1896 def _find_xpath(xpath
):
1897 return node
.find(compat_xpath(xpath
))
1899 if isinstance(xpath
, (str, compat_str
)):
1900 n
= _find_xpath(xpath
)
1908 if default
is not NO_DEFAULT
:
1911 name
= xpath
if name
is None else name
1912 raise ExtractorError('Could not find XML element %s' % name
)
1918 def xpath_text(node
, xpath
, name
=None, fatal
=False, default
=NO_DEFAULT
):
1919 n
= xpath_element(node
, xpath
, name
, fatal
=fatal
, default
=default
)
1920 if n
is None or n
== default
:
1923 if default
is not NO_DEFAULT
:
1926 name
= xpath
if name
is None else name
1927 raise ExtractorError('Could not find XML element\'s text %s' % name
)
1933 def xpath_attr(node
, xpath
, key
, name
=None, fatal
=False, default
=NO_DEFAULT
):
1934 n
= find_xpath_attr(node
, xpath
, key
)
1936 if default
is not NO_DEFAULT
:
1939 name
= '%s[@%s]' % (xpath
, key
) if name
is None else name
1940 raise ExtractorError('Could not find XML attribute %s' % name
)
1943 return n
.attrib
[key
]
1946 def get_element_by_id(id, html
):
1947 """Return the content of the tag with the specified ID in the passed HTML document"""
1948 return get_element_by_attribute('id', id, html
)
1951 def get_element_by_class(class_name
, html
):
1952 """Return the content of the first tag with the specified class in the passed HTML document"""
1953 retval
= get_elements_by_class(class_name
, html
)
1954 return retval
[0] if retval
else None
1957 def get_element_by_attribute(attribute
, value
, html
, escape_value
=True):
1958 retval
= get_elements_by_attribute(attribute
, value
, html
, escape_value
)
1959 return retval
[0] if retval
else None
1962 def get_elements_by_class(class_name
, html
):
1963 """Return the content of all tags with the specified class in the passed HTML document as a list"""
1964 return get_elements_by_attribute(
1965 'class', r
'[^\'"]*\b%s\b[^\'"]*' % re.escape(class_name),
1966 html, escape_value=False)
1969 def get_elements_by_attribute(attribute, value, html, escape_value=True):
1970 """Return the content of the tag with the specified attribute in the passed HTML document"""
1972 value = re.escape(value) if escape_value else value
1975 for m in re.finditer(r'''(?xs)
1977 (?:\s+[a-zA-Z0-9:._-]+(?:=[a-zA-Z0-9:._-]*|="[^"]*"|='[^
']*'|
))*?
1979 (?:\s+[a-zA-Z0-9:._-]+(?:=[a-zA-Z0-9:._-]*|="[^
"]*"|
='[^']*'|))*?
1983 ''' % (re.escape(attribute), value), html):
1984 res = m.group('content
')
1986 if res.startswith('"') or res.startswith("'"):
1989 retlist.append(unescapeHTML(res))
1994 class HTMLAttributeParser(compat_HTMLParser):
1995 """Trivial HTML parser to gather the attributes for a single element"""
1999 compat_HTMLParser.__init__(self)
2001 def handle_starttag(self, tag, attrs):
2002 self.attrs = dict(attrs)
2005 def extract_attributes(html_element):
2006 """Given a string for an HTML element such as
2008 a="foo" B="bar" c="&98;az" d=boz
2009 empty= noval entity="&"
2012 Decode and return a dictionary of attributes.
2014 'a
': 'foo
', 'b
': 'bar
', c: 'baz
', d: 'boz
',
2015 'empty
': '', 'noval
': None, 'entity
': '&',
2016 'sq
': '"', 'dq': '\''
2018 NB HTMLParser is stricter in Python 2.6 & 3.2 than in later versions,
2019 but the cases in the unit test will work for all of 2.6, 2.7, 3.2-3.5.
2021 parser = HTMLAttributeParser()
2023 parser.feed(html_element)
2025 # Older Python may throw HTMLParseError in case of malformed HTML
2026 except compat_HTMLParseError:
2031 def clean_html(html):
2032 """Clean an HTML snippet into a readable string"""
2034 if html is None: # Convenience for sanitizing descriptions etc.
2038 html = html.replace('\n', ' ')
2039 html = re.sub(r'(?u)\s*<\s*br\s*/?\s*>\s*', '\n', html)
2040 html = re.sub(r'(?u)<\s*/\s*p\s*>\s*<\s*p[^>]*>', '\n', html)
2042 html = re.sub('<.*?>', '', html)
2043 # Replace html entities
2044 html = unescapeHTML(html)
2048 def sanitize_open(filename, open_mode):
2049 """Try to open the given filename, and slightly tweak it if this fails.
2051 Attempts to open the given filename. If this fails, it tries to change
2052 the filename slightly, step by step, until it's either able to open it
2053 or it fails and raises a final exception, like the standard open()
2056 It returns the tuple (stream, definitive_file_name).
2060 if sys.platform == 'win32':
2062 msvcrt.setmode(sys.stdout.fileno(), os.O_BINARY)
2063 return (sys.stdout.buffer if hasattr(sys.stdout, 'buffer') else sys.stdout, filename)
2064 stream = open(encodeFilename(filename), open_mode)
2065 return (stream, filename)
2066 except (IOError, OSError) as err:
2067 if err.errno in (errno.EACCES,):
2070 # In case of error, try to remove win32 forbidden chars
2071 alt_filename = sanitize_path(filename)
2072 if alt_filename == filename:
2075 # An exception here should be caught in the caller
2076 stream = open(encodeFilename(alt_filename), open_mode)
2077 return (stream, alt_filename)
2080 def timeconvert(timestr):
2081 """Convert RFC 2822 defined time string into system timestamp"""
2083 timetuple = email.utils.parsedate_tz(timestr)
2084 if timetuple is not None:
2085 timestamp = email.utils.mktime_tz(timetuple)
2089 def sanitize_filename(s, restricted=False, is_id=False):
2090 """Sanitizes a string so it could be used as part of a filename.
2091 If restricted is set, use a stricter subset of allowed characters.
2092 Set is_id if this is not an arbitrary string, but an ID that should be kept
2095 def replace_insane(char):
2096 if restricted and char in ACCENT_CHARS:
2097 return ACCENT_CHARS[char]
2098 if char == '?' or ord(char) < 32 or ord(char) == 127:
2101 return '' if restricted else '\''
2103 return '_
-' if restricted else ' -'
2104 elif char in '\\/|
*<>':
2106 if restricted and (char in '!&\'()[]{}$
;`^
,#' or char.isspace()):
2108 if restricted
and ord(char
) > 127:
2115 s
= re
.sub(r
'[0-9]+(?::[0-9]+)+', lambda m
: m
.group(0).replace(':', '_'), s
)
2116 result
= ''.join(map(replace_insane
, s
))
2118 while '__' in result
:
2119 result
= result
.replace('__', '_')
2120 result
= result
.strip('_')
2121 # Common case of "Foreign band name - English song title"
2122 if restricted
and result
.startswith('-_'):
2124 if result
.startswith('-'):
2125 result
= '_' + result
[len('-'):]
2126 result
= result
.lstrip('.')
2132 def sanitize_path(s
, force
=False):
2133 """Sanitizes and normalizes path on Windows"""
2134 if sys
.platform
== 'win32':
2136 drive_or_unc
, _
= os
.path
.splitdrive(s
)
2137 if sys
.version_info
< (2, 7) and not drive_or_unc
:
2138 drive_or_unc
, _
= os
.path
.splitunc(s
)
2144 norm_path
= os
.path
.normpath(remove_start(s
, drive_or_unc
)).split(os
.path
.sep
)
2148 path_part
if path_part
in ['.', '..'] else re
.sub(r
'(?:[/<>:"\|\\?\*]|[\s.]$)', '#', path_part
)
2149 for path_part
in norm_path
]
2151 sanitized_path
.insert(0, drive_or_unc
+ os
.path
.sep
)
2152 elif force
and s
[0] == os
.path
.sep
:
2153 sanitized_path
.insert(0, os
.path
.sep
)
2154 return os
.path
.join(*sanitized_path
)
2157 def sanitize_url(url
):
2158 # Prepend protocol-less URLs with `http:` scheme in order to mitigate
2159 # the number of unwanted failures due to missing protocol
2160 if url
.startswith('//'):
2161 return 'http:%s' % url
2162 # Fix some common typos seen so far
2164 # https://github.com/ytdl-org/youtube-dl/issues/15649
2165 (r
'^httpss://', r
'https://'),
2166 # https://bx1.be/lives/direct-tv/
2167 (r
'^rmtp([es]?)://', r
'rtmp\1://'),
2169 for mistake
, fixup
in COMMON_TYPOS
:
2170 if re
.match(mistake
, url
):
2171 return re
.sub(mistake
, fixup
, url
)
2175 def extract_basic_auth(url
):
2176 parts
= compat_urlparse
.urlsplit(url
)
2177 if parts
.username
is None:
2179 url
= compat_urlparse
.urlunsplit(parts
._replace
(netloc
=(
2180 parts
.hostname
if parts
.port
is None
2181 else '%s:%d' % (parts
.hostname
, parts
.port
))))
2182 auth_payload
= base64
.b64encode(
2183 ('%s:%s' % (parts
.username
, parts
.password
or '')).encode('utf-8'))
2184 return url
, 'Basic ' + auth_payload
.decode('utf-8')
2187 def sanitized_Request(url
, *args
, **kwargs
):
2188 url
, auth_header
= extract_basic_auth(escape_url(sanitize_url(url
)))
2189 if auth_header
is not None:
2190 headers
= args
[1] if len(args
) >= 2 else kwargs
.setdefault('headers', {})
2191 headers
['Authorization'] = auth_header
2192 return compat_urllib_request
.Request(url
, *args
, **kwargs
)
2196 """Expand shell variables and ~"""
2197 return os
.path
.expandvars(compat_expanduser(s
))
2200 def orderedSet(iterable
):
2201 """ Remove all duplicates from the input iterable """
2209 def _htmlentity_transform(entity_with_semicolon
):
2210 """Transforms an HTML entity to a character."""
2211 entity
= entity_with_semicolon
[:-1]
2213 # Known non-numeric HTML entity
2214 if entity
in compat_html_entities
.name2codepoint
:
2215 return compat_chr(compat_html_entities
.name2codepoint
[entity
])
2217 # TODO: HTML5 allows entities without a semicolon. For example,
2218 # 'Éric' should be decoded as 'Éric'.
2219 if entity_with_semicolon
in compat_html_entities_html5
:
2220 return compat_html_entities_html5
[entity_with_semicolon
]
2222 mobj
= re
.match(r
'#(x[0-9a-fA-F]+|[0-9]+)', entity
)
2223 if mobj
is not None:
2224 numstr
= mobj
.group(1)
2225 if numstr
.startswith('x'):
2227 numstr
= '0%s' % numstr
2230 # See https://github.com/ytdl-org/youtube-dl/issues/7518
2232 return compat_chr(int(numstr
, base
))
2236 # Unknown entity in name, return its literal representation
2237 return '&%s;' % entity
2240 def unescapeHTML(s
):
2243 assert type(s
) == compat_str
2246 r
'&([^&;]+;)', lambda m
: _htmlentity_transform(m
.group(1)), s
)
2249 def escapeHTML(text
):
2252 .replace('&', '&')
2253 .replace('<', '<')
2254 .replace('>', '>')
2255 .replace('"', '"')
2256 .replace("'", ''')
2260 def process_communicate_or_kill(p
, *args
, **kwargs
):
2262 return p
.communicate(*args
, **kwargs
)
2263 except BaseException
: # Including KeyboardInterrupt
2269 def get_subprocess_encoding():
2270 if sys
.platform
== 'win32' and sys
.getwindowsversion()[0] >= 5:
2271 # For subprocess calls, encode with locale encoding
2272 # Refer to http://stackoverflow.com/a/9951851/35070
2273 encoding
= preferredencoding()
2275 encoding
= sys
.getfilesystemencoding()
2276 if encoding
is None:
2281 def encodeFilename(s
, for_subprocess
=False):
2283 @param s The name of the file
2286 assert type(s
) == compat_str
2288 # Python 3 has a Unicode API
2289 if sys
.version_info
>= (3, 0):
2292 # Pass '' directly to use Unicode APIs on Windows 2000 and up
2293 # (Detecting Windows NT 4 is tricky because 'major >= 4' would
2294 # match Windows 9x series as well. Besides, NT 4 is obsolete.)
2295 if not for_subprocess
and sys
.platform
== 'win32' and sys
.getwindowsversion()[0] >= 5:
2298 # Jython assumes filenames are Unicode strings though reported as Python 2.x compatible
2299 if sys
.platform
.startswith('java'):
2302 return s
.encode(get_subprocess_encoding(), 'ignore')
2305 def decodeFilename(b
, for_subprocess
=False):
2307 if sys
.version_info
>= (3, 0):
2310 if not isinstance(b
, bytes):
2313 return b
.decode(get_subprocess_encoding(), 'ignore')
2316 def encodeArgument(s
):
2317 if not isinstance(s
, compat_str
):
2318 # Legacy code that uses byte strings
2319 # Uncomment the following line after fixing all post processors
2320 # assert False, 'Internal error: %r should be of type %r, is %r' % (s, compat_str, type(s))
2321 s
= s
.decode('ascii')
2322 return encodeFilename(s
, True)
2325 def decodeArgument(b
):
2326 return decodeFilename(b
, True)
2329 def decodeOption(optval
):
2332 if isinstance(optval
, bytes):
2333 optval
= optval
.decode(preferredencoding())
2335 assert isinstance(optval
, compat_str
)
2339 def formatSeconds(secs
, delim
=':', msec
=False):
2341 ret
= '%d%s%02d%s%02d' % (secs
// 3600, delim
, (secs
% 3600) // 60, delim
, secs
% 60)
2343 ret
= '%d%s%02d' % (secs
// 60, delim
, secs
% 60)
2346 return '%s.%03d' % (ret
, secs
% 1) if msec
else ret
2349 def make_HTTPS_handler(params
, **kwargs
):
2350 opts_no_check_certificate
= params
.get('nocheckcertificate', False)
2351 if hasattr(ssl
, 'create_default_context'): # Python >= 3.4 or 2.7.9
2352 context
= ssl
.create_default_context(ssl
.Purpose
.SERVER_AUTH
)
2353 if opts_no_check_certificate
:
2354 context
.check_hostname
= False
2355 context
.verify_mode
= ssl
.CERT_NONE
2357 return YoutubeDLHTTPSHandler(params
, context
=context
, **kwargs
)
2360 # (create_default_context present but HTTPSHandler has no context=)
2363 if sys
.version_info
< (3, 2):
2364 return YoutubeDLHTTPSHandler(params
, **kwargs
)
2365 else: # Python < 3.4
2366 context
= ssl
.SSLContext(ssl
.PROTOCOL_TLSv1
)
2367 context
.verify_mode
= (ssl
.CERT_NONE
2368 if opts_no_check_certificate
2369 else ssl
.CERT_REQUIRED
)
2370 context
.set_default_verify_paths()
2371 return YoutubeDLHTTPSHandler(params
, context
=context
, **kwargs
)
2374 def bug_reports_message(before
=';'):
2375 if ytdl_is_updateable():
2376 update_cmd
= 'type yt-dlp -U to update'
2378 update_cmd
= 'see https://github.com/yt-dlp/yt-dlp on how to update'
2379 msg
= 'please report this issue on https://github.com/yt-dlp/yt-dlp .'
2380 msg
+= ' Make sure you are using the latest version; %s.' % update_cmd
2381 msg
+= ' Be sure to call yt-dlp with the --verbose flag and include its complete output.'
2383 before
= before
.rstrip()
2384 if not before
or before
.endswith(('.', '!', '?')):
2385 msg
= msg
[0].title() + msg
[1:]
2387 return (before
+ ' ' if before
else '') + msg
2390 class YoutubeDLError(Exception):
2391 """Base exception for YoutubeDL errors."""
2395 network_exceptions
= [compat_urllib_error
.URLError
, compat_http_client
.HTTPException
, socket
.error
]
2396 if hasattr(ssl
, 'CertificateError'):
2397 network_exceptions
.append(ssl
.CertificateError
)
2398 network_exceptions
= tuple(network_exceptions
)
2401 class ExtractorError(YoutubeDLError
):
2402 """Error during info extraction."""
2404 def __init__(self
, msg
, tb
=None, expected
=False, cause
=None, video_id
=None, ie
=None):
2405 """ tb, if given, is the original traceback (so that it can be printed out).
2406 If expected is set, this is a normal error message and most likely not a bug in yt-dlp.
2408 if sys
.exc_info()[0] in network_exceptions
:
2413 self
.expected
= expected
2415 self
.video_id
= video_id
2417 self
.exc_info
= sys
.exc_info() # preserve original exception
2419 super(ExtractorError
, self
).__init
__(''.join((
2420 format_field(ie
, template
='[%s] '),
2421 format_field(video_id
, template
='%s: '),
2423 format_field(cause
, template
=' (caused by %r)'),
2424 '' if expected
else bug_reports_message())))
2426 def format_traceback(self
):
2427 if self
.traceback
is None:
2429 return ''.join(traceback
.format_tb(self
.traceback
))
2432 class UnsupportedError(ExtractorError
):
2433 def __init__(self
, url
):
2434 super(UnsupportedError
, self
).__init
__(
2435 'Unsupported URL: %s' % url
, expected
=True)
2439 class RegexNotFoundError(ExtractorError
):
2440 """Error when a regex didn't match"""
2444 class GeoRestrictedError(ExtractorError
):
2445 """Geographic restriction Error exception.
2447 This exception may be thrown when a video is not available from your
2448 geographic location due to geographic restrictions imposed by a website.
2451 def __init__(self
, msg
, countries
=None):
2452 super(GeoRestrictedError
, self
).__init
__(msg
, expected
=True)
2454 self
.countries
= countries
2457 class DownloadError(YoutubeDLError
):
2458 """Download Error exception.
2460 This exception may be thrown by FileDownloader objects if they are not
2461 configured to continue on errors. They will contain the appropriate
2465 def __init__(self
, msg
, exc_info
=None):
2466 """ exc_info, if given, is the original exception that caused the trouble (as returned by sys.exc_info()). """
2467 super(DownloadError
, self
).__init
__(msg
)
2468 self
.exc_info
= exc_info
2471 class EntryNotInPlaylist(YoutubeDLError
):
2472 """Entry not in playlist exception.
2474 This exception will be thrown by YoutubeDL when a requested entry
2475 is not found in the playlist info_dict
2480 class SameFileError(YoutubeDLError
):
2481 """Same File exception.
2483 This exception will be thrown by FileDownloader objects if they detect
2484 multiple files would have to be downloaded to the same file on disk.
2489 class PostProcessingError(YoutubeDLError
):
2490 """Post Processing exception.
2492 This exception may be raised by PostProcessor's .run() method to
2493 indicate an error in the postprocessing task.
2496 def __init__(self
, msg
):
2497 super(PostProcessingError
, self
).__init
__(msg
)
2501 class ExistingVideoReached(YoutubeDLError
):
2502 """ --max-downloads limit has been reached. """
2506 class RejectedVideoReached(YoutubeDLError
):
2507 """ --max-downloads limit has been reached. """
2511 class ThrottledDownload(YoutubeDLError
):
2512 """ Download speed below --throttled-rate. """
2516 class MaxDownloadsReached(YoutubeDLError
):
2517 """ --max-downloads limit has been reached. """
2521 class UnavailableVideoError(YoutubeDLError
):
2522 """Unavailable Format exception.
2524 This exception will be thrown when a video is requested
2525 in a format that is not available for that video.
2530 class ContentTooShortError(YoutubeDLError
):
2531 """Content Too Short exception.
2533 This exception may be raised by FileDownloader objects when a file they
2534 download is too small for what the server announced first, indicating
2535 the connection was probably interrupted.
2538 def __init__(self
, downloaded
, expected
):
2539 super(ContentTooShortError
, self
).__init
__(
2540 'Downloaded {0} bytes, expected {1} bytes'.format(downloaded
, expected
)
2543 self
.downloaded
= downloaded
2544 self
.expected
= expected
2547 class XAttrMetadataError(YoutubeDLError
):
2548 def __init__(self
, code
=None, msg
='Unknown error'):
2549 super(XAttrMetadataError
, self
).__init
__(msg
)
2553 # Parsing code and msg
2554 if (self
.code
in (errno
.ENOSPC
, errno
.EDQUOT
)
2555 or 'No space left' in self
.msg
or 'Disk quota exceeded' in self
.msg
):
2556 self
.reason
= 'NO_SPACE'
2557 elif self
.code
== errno
.E2BIG
or 'Argument list too long' in self
.msg
:
2558 self
.reason
= 'VALUE_TOO_LONG'
2560 self
.reason
= 'NOT_SUPPORTED'
2563 class XAttrUnavailableError(YoutubeDLError
):
2567 def _create_http_connection(ydl_handler
, http_class
, is_https
, *args
, **kwargs
):
2568 # Working around python 2 bug (see http://bugs.python.org/issue17849) by limiting
2569 # expected HTTP responses to meet HTTP/1.0 or later (see also
2570 # https://github.com/ytdl-org/youtube-dl/issues/6727)
2571 if sys
.version_info
< (3, 0):
2572 kwargs
['strict'] = True
2573 hc
= http_class(*args
, **compat_kwargs(kwargs
))
2574 source_address
= ydl_handler
._params
.get('source_address')
2576 if source_address
is not None:
2577 # This is to workaround _create_connection() from socket where it will try all
2578 # address data from getaddrinfo() including IPv6. This filters the result from
2579 # getaddrinfo() based on the source_address value.
2580 # This is based on the cpython socket.create_connection() function.
2581 # https://github.com/python/cpython/blob/master/Lib/socket.py#L691
2582 def _create_connection(address
, timeout
=socket
._GLOBAL
_DEFAULT
_TIMEOUT
, source_address
=None):
2583 host
, port
= address
2585 addrs
= socket
.getaddrinfo(host
, port
, 0, socket
.SOCK_STREAM
)
2586 af
= socket
.AF_INET
if '.' in source_address
[0] else socket
.AF_INET6
2587 ip_addrs
= [addr
for addr
in addrs
if addr
[0] == af
]
2588 if addrs
and not ip_addrs
:
2589 ip_version
= 'v4' if af
== socket
.AF_INET
else 'v6'
2591 "No remote IP%s addresses available for connect, can't use '%s' as source address"
2592 % (ip_version
, source_address
[0]))
2593 for res
in ip_addrs
:
2594 af
, socktype
, proto
, canonname
, sa
= res
2597 sock
= socket
.socket(af
, socktype
, proto
)
2598 if timeout
is not socket
._GLOBAL
_DEFAULT
_TIMEOUT
:
2599 sock
.settimeout(timeout
)
2600 sock
.bind(source_address
)
2602 err
= None # Explicitly break reference cycle
2604 except socket
.error
as _
:
2606 if sock
is not None:
2611 raise socket
.error('getaddrinfo returns an empty list')
2612 if hasattr(hc
, '_create_connection'):
2613 hc
._create
_connection
= _create_connection
2614 sa
= (source_address
, 0)
2615 if hasattr(hc
, 'source_address'): # Python 2.7+
2616 hc
.source_address
= sa
2618 def _hc_connect(self
, *args
, **kwargs
):
2619 sock
= _create_connection(
2620 (self
.host
, self
.port
), self
.timeout
, sa
)
2622 self
.sock
= ssl
.wrap_socket(
2623 sock
, self
.key_file
, self
.cert_file
,
2624 ssl_version
=ssl
.PROTOCOL_TLSv1
)
2627 hc
.connect
= functools
.partial(_hc_connect
, hc
)
2632 def handle_youtubedl_headers(headers
):
2633 filtered_headers
= headers
2635 if 'Youtubedl-no-compression' in filtered_headers
:
2636 filtered_headers
= dict((k
, v
) for k
, v
in filtered_headers
.items() if k
.lower() != 'accept-encoding')
2637 del filtered_headers
['Youtubedl-no-compression']
2639 return filtered_headers
2642 class YoutubeDLHandler(compat_urllib_request
.HTTPHandler
):
2643 """Handler for HTTP requests and responses.
2645 This class, when installed with an OpenerDirector, automatically adds
2646 the standard headers to every HTTP request and handles gzipped and
2647 deflated responses from web servers. If compression is to be avoided in
2648 a particular request, the original request in the program code only has
2649 to include the HTTP header "Youtubedl-no-compression", which will be
2650 removed before making the real request.
2652 Part of this code was copied from:
2654 http://techknack.net/python-urllib2-handlers/
2656 Andrew Rowls, the author of that code, agreed to release it to the
2660 def __init__(self
, params
, *args
, **kwargs
):
2661 compat_urllib_request
.HTTPHandler
.__init
__(self
, *args
, **kwargs
)
2662 self
._params
= params
2664 def http_open(self
, req
):
2665 conn_class
= compat_http_client
.HTTPConnection
2667 socks_proxy
= req
.headers
.get('Ytdl-socks-proxy')
2669 conn_class
= make_socks_conn_class(conn_class
, socks_proxy
)
2670 del req
.headers
['Ytdl-socks-proxy']
2672 return self
.do_open(functools
.partial(
2673 _create_http_connection
, self
, conn_class
, False),
2681 return zlib
.decompress(data
, -zlib
.MAX_WBITS
)
2683 return zlib
.decompress(data
)
2685 def http_request(self
, req
):
2686 # According to RFC 3986, URLs can not contain non-ASCII characters, however this is not
2687 # always respected by websites, some tend to give out URLs with non percent-encoded
2688 # non-ASCII characters (see telemb.py, ard.py [#3412])
2689 # urllib chokes on URLs with non-ASCII characters (see http://bugs.python.org/issue3991)
2690 # To work around aforementioned issue we will replace request's original URL with
2691 # percent-encoded one
2692 # Since redirects are also affected (e.g. http://www.southpark.de/alle-episoden/s18e09)
2693 # the code of this workaround has been moved here from YoutubeDL.urlopen()
2694 url
= req
.get_full_url()
2695 url_escaped
= escape_url(url
)
2697 # Substitute URL if any change after escaping
2698 if url
!= url_escaped
:
2699 req
= update_Request(req
, url
=url_escaped
)
2701 for h
, v
in std_headers
.items():
2702 # Capitalize is needed because of Python bug 2275: http://bugs.python.org/issue2275
2703 # The dict keys are capitalized because of this bug by urllib
2704 if h
.capitalize() not in req
.headers
:
2705 req
.add_header(h
, v
)
2707 req
.headers
= handle_youtubedl_headers(req
.headers
)
2709 if sys
.version_info
< (2, 7) and '#' in req
.get_full_url():
2710 # Python 2.6 is brain-dead when it comes to fragments
2711 req
._Request
__original
= req
._Request
__original
.partition('#')[0]
2712 req
._Request
__r
_type
= req
._Request
__r
_type
.partition('#')[0]
2716 def http_response(self
, req
, resp
):
2719 if resp
.headers
.get('Content-encoding', '') == 'gzip':
2720 content
= resp
.read()
2721 gz
= gzip
.GzipFile(fileobj
=io
.BytesIO(content
), mode
='rb')
2723 uncompressed
= io
.BytesIO(gz
.read())
2724 except IOError as original_ioerror
:
2725 # There may be junk add the end of the file
2726 # See http://stackoverflow.com/q/4928560/35070 for details
2727 for i
in range(1, 1024):
2729 gz
= gzip
.GzipFile(fileobj
=io
.BytesIO(content
[:-i
]), mode
='rb')
2730 uncompressed
= io
.BytesIO(gz
.read())
2735 raise original_ioerror
2736 resp
= compat_urllib_request
.addinfourl(uncompressed
, old_resp
.headers
, old_resp
.url
, old_resp
.code
)
2737 resp
.msg
= old_resp
.msg
2738 del resp
.headers
['Content-encoding']
2740 if resp
.headers
.get('Content-encoding', '') == 'deflate':
2741 gz
= io
.BytesIO(self
.deflate(resp
.read()))
2742 resp
= compat_urllib_request
.addinfourl(gz
, old_resp
.headers
, old_resp
.url
, old_resp
.code
)
2743 resp
.msg
= old_resp
.msg
2744 del resp
.headers
['Content-encoding']
2745 # Percent-encode redirect URL of Location HTTP header to satisfy RFC 3986 (see
2746 # https://github.com/ytdl-org/youtube-dl/issues/6457).
2747 if 300 <= resp
.code
< 400:
2748 location
= resp
.headers
.get('Location')
2750 # As of RFC 2616 default charset is iso-8859-1 that is respected by python 3
2751 if sys
.version_info
>= (3, 0):
2752 location
= location
.encode('iso-8859-1').decode('utf-8')
2754 location
= location
.decode('utf-8')
2755 location_escaped
= escape_url(location
)
2756 if location
!= location_escaped
:
2757 del resp
.headers
['Location']
2758 if sys
.version_info
< (3, 0):
2759 location_escaped
= location_escaped
.encode('utf-8')
2760 resp
.headers
['Location'] = location_escaped
2763 https_request
= http_request
2764 https_response
= http_response
2767 def make_socks_conn_class(base_class
, socks_proxy
):
2768 assert issubclass(base_class
, (
2769 compat_http_client
.HTTPConnection
, compat_http_client
.HTTPSConnection
))
2771 url_components
= compat_urlparse
.urlparse(socks_proxy
)
2772 if url_components
.scheme
.lower() == 'socks5':
2773 socks_type
= ProxyType
.SOCKS5
2774 elif url_components
.scheme
.lower() in ('socks', 'socks4'):
2775 socks_type
= ProxyType
.SOCKS4
2776 elif url_components
.scheme
.lower() == 'socks4a':
2777 socks_type
= ProxyType
.SOCKS4A
2779 def unquote_if_non_empty(s
):
2782 return compat_urllib_parse_unquote_plus(s
)
2786 url_components
.hostname
, url_components
.port
or 1080,
2788 unquote_if_non_empty(url_components
.username
),
2789 unquote_if_non_empty(url_components
.password
),
2792 class SocksConnection(base_class
):
2794 self
.sock
= sockssocket()
2795 self
.sock
.setproxy(*proxy_args
)
2796 if type(self
.timeout
) in (int, float):
2797 self
.sock
.settimeout(self
.timeout
)
2798 self
.sock
.connect((self
.host
, self
.port
))
2800 if isinstance(self
, compat_http_client
.HTTPSConnection
):
2801 if hasattr(self
, '_context'): # Python > 2.6
2802 self
.sock
= self
._context
.wrap_socket(
2803 self
.sock
, server_hostname
=self
.host
)
2805 self
.sock
= ssl
.wrap_socket(self
.sock
)
2807 return SocksConnection
2810 class YoutubeDLHTTPSHandler(compat_urllib_request
.HTTPSHandler
):
2811 def __init__(self
, params
, https_conn_class
=None, *args
, **kwargs
):
2812 compat_urllib_request
.HTTPSHandler
.__init
__(self
, *args
, **kwargs
)
2813 self
._https
_conn
_class
= https_conn_class
or compat_http_client
.HTTPSConnection
2814 self
._params
= params
2816 def https_open(self
, req
):
2818 conn_class
= self
._https
_conn
_class
2820 if hasattr(self
, '_context'): # python > 2.6
2821 kwargs
['context'] = self
._context
2822 if hasattr(self
, '_check_hostname'): # python 3.x
2823 kwargs
['check_hostname'] = self
._check
_hostname
2825 socks_proxy
= req
.headers
.get('Ytdl-socks-proxy')
2827 conn_class
= make_socks_conn_class(conn_class
, socks_proxy
)
2828 del req
.headers
['Ytdl-socks-proxy']
2830 return self
.do_open(functools
.partial(
2831 _create_http_connection
, self
, conn_class
, True),
2835 class YoutubeDLCookieJar(compat_cookiejar
.MozillaCookieJar
):
2837 See [1] for cookie file format.
2839 1. https://curl.haxx.se/docs/http-cookies.html
2841 _HTTPONLY_PREFIX
= '#HttpOnly_'
2843 _HEADER
= '''# Netscape HTTP Cookie File
2844 # This file is generated by yt-dlp. Do not edit.
2847 _CookieFileEntry
= collections
.namedtuple(
2849 ('domain_name', 'include_subdomains', 'path', 'https_only', 'expires_at', 'name', 'value'))
2851 def save(self
, filename
=None, ignore_discard
=False, ignore_expires
=False):
2853 Save cookies to a file.
2855 Most of the code is taken from CPython 3.8 and slightly adapted
2856 to support cookie files with UTF-8 in both python 2 and 3.
2858 if filename
is None:
2859 if self
.filename
is not None:
2860 filename
= self
.filename
2862 raise ValueError(compat_cookiejar
.MISSING_FILENAME_TEXT
)
2864 # Store session cookies with `expires` set to 0 instead of an empty
2867 if cookie
.expires
is None:
2870 with io
.open(filename
, 'w', encoding
='utf-8') as f
:
2871 f
.write(self
._HEADER
)
2874 if not ignore_discard
and cookie
.discard
:
2876 if not ignore_expires
and cookie
.is_expired(now
):
2882 if cookie
.domain
.startswith('.'):
2883 initial_dot
= 'TRUE'
2885 initial_dot
= 'FALSE'
2886 if cookie
.expires
is not None:
2887 expires
= compat_str(cookie
.expires
)
2890 if cookie
.value
is None:
2891 # cookies.txt regards 'Set-Cookie: foo' as a cookie
2892 # with no name, whereas http.cookiejar regards it as a
2893 # cookie with no value.
2898 value
= cookie
.value
2900 '\t'.join([cookie
.domain
, initial_dot
, cookie
.path
,
2901 secure
, expires
, name
, value
]) + '\n')
2903 def load(self
, filename
=None, ignore_discard
=False, ignore_expires
=False):
2904 """Load cookies from a file."""
2905 if filename
is None:
2906 if self
.filename
is not None:
2907 filename
= self
.filename
2909 raise ValueError(compat_cookiejar
.MISSING_FILENAME_TEXT
)
2911 def prepare_line(line
):
2912 if line
.startswith(self
._HTTPONLY
_PREFIX
):
2913 line
= line
[len(self
._HTTPONLY
_PREFIX
):]
2914 # comments and empty lines are fine
2915 if line
.startswith('#') or not line
.strip():
2917 cookie_list
= line
.split('\t')
2918 if len(cookie_list
) != self
._ENTRY
_LEN
:
2919 raise compat_cookiejar
.LoadError('invalid length %d' % len(cookie_list
))
2920 cookie
= self
._CookieFileEntry
(*cookie_list
)
2921 if cookie
.expires_at
and not cookie
.expires_at
.isdigit():
2922 raise compat_cookiejar
.LoadError('invalid expires at %s' % cookie
.expires_at
)
2926 with io
.open(filename
, encoding
='utf-8') as f
:
2929 cf
.write(prepare_line(line
))
2930 except compat_cookiejar
.LoadError
as e
:
2932 'WARNING: skipping cookie file entry due to %s: %r\n'
2933 % (e
, line
), sys
.stderr
)
2936 self
._really
_load
(cf
, filename
, ignore_discard
, ignore_expires
)
2937 # Session cookies are denoted by either `expires` field set to
2938 # an empty string or 0. MozillaCookieJar only recognizes the former
2939 # (see [1]). So we need force the latter to be recognized as session
2940 # cookies on our own.
2941 # Session cookies may be important for cookies-based authentication,
2942 # e.g. usually, when user does not check 'Remember me' check box while
2943 # logging in on a site, some important cookies are stored as session
2944 # cookies so that not recognizing them will result in failed login.
2945 # 1. https://bugs.python.org/issue17164
2947 # Treat `expires=0` cookies as session cookies
2948 if cookie
.expires
== 0:
2949 cookie
.expires
= None
2950 cookie
.discard
= True
2953 class YoutubeDLCookieProcessor(compat_urllib_request
.HTTPCookieProcessor
):
2954 def __init__(self
, cookiejar
=None):
2955 compat_urllib_request
.HTTPCookieProcessor
.__init
__(self
, cookiejar
)
2957 def http_response(self
, request
, response
):
2958 # Python 2 will choke on next HTTP request in row if there are non-ASCII
2959 # characters in Set-Cookie HTTP header of last response (see
2960 # https://github.com/ytdl-org/youtube-dl/issues/6769).
2961 # In order to at least prevent crashing we will percent encode Set-Cookie
2962 # header before HTTPCookieProcessor starts processing it.
2963 # if sys.version_info < (3, 0) and response.headers:
2964 # for set_cookie_header in ('Set-Cookie', 'Set-Cookie2'):
2965 # set_cookie = response.headers.get(set_cookie_header)
2967 # set_cookie_escaped = compat_urllib_parse.quote(set_cookie, b"%/;:@&=+$,!~*'()?#[] ")
2968 # if set_cookie != set_cookie_escaped:
2969 # del response.headers[set_cookie_header]
2970 # response.headers[set_cookie_header] = set_cookie_escaped
2971 return compat_urllib_request
.HTTPCookieProcessor
.http_response(self
, request
, response
)
2973 https_request
= compat_urllib_request
.HTTPCookieProcessor
.http_request
2974 https_response
= http_response
2977 class YoutubeDLRedirectHandler(compat_urllib_request
.HTTPRedirectHandler
):
2978 """YoutubeDL redirect handler
2980 The code is based on HTTPRedirectHandler implementation from CPython [1].
2982 This redirect handler solves two issues:
2983 - ensures redirect URL is always unicode under python 2
2984 - introduces support for experimental HTTP response status code
2985 308 Permanent Redirect [2] used by some sites [3]
2987 1. https://github.com/python/cpython/blob/master/Lib/urllib/request.py
2988 2. https://developer.mozilla.org/en-US/docs/Web/HTTP/Status/308
2989 3. https://github.com/ytdl-org/youtube-dl/issues/28768
2992 http_error_301
= http_error_303
= http_error_307
= http_error_308
= compat_urllib_request
.HTTPRedirectHandler
.http_error_302
2994 def redirect_request(self
, req
, fp
, code
, msg
, headers
, newurl
):
2995 """Return a Request or None in response to a redirect.
2997 This is called by the http_error_30x methods when a
2998 redirection response is received. If a redirection should
2999 take place, return a new Request to allow http_error_30x to
3000 perform the redirect. Otherwise, raise HTTPError if no-one
3001 else should try to handle this url. Return None if you can't
3002 but another Handler might.
3004 m
= req
.get_method()
3005 if (not (code
in (301, 302, 303, 307, 308) and m
in ("GET", "HEAD")
3006 or code
in (301, 302, 303) and m
== "POST")):
3007 raise compat_HTTPError(req
.full_url
, code
, msg
, headers
, fp
)
3008 # Strictly (according to RFC 2616), 301 or 302 in response to
3009 # a POST MUST NOT cause a redirection without confirmation
3010 # from the user (of urllib.request, in this case). In practice,
3011 # essentially all clients do redirect in this case, so we do
3014 # On python 2 urlh.geturl() may sometimes return redirect URL
3015 # as byte string instead of unicode. This workaround allows
3016 # to force it always return unicode.
3017 if sys
.version_info
[0] < 3:
3018 newurl
= compat_str(newurl
)
3020 # Be conciliant with URIs containing a space. This is mainly
3021 # redundant with the more complete encoding done in http_error_302(),
3022 # but it is kept for compatibility with other callers.
3023 newurl
= newurl
.replace(' ', '%20')
3025 CONTENT_HEADERS
= ("content-length", "content-type")
3026 # NB: don't use dict comprehension for python 2.6 compatibility
3027 newheaders
= dict((k
, v
) for k
, v
in req
.headers
.items()
3028 if k
.lower() not in CONTENT_HEADERS
)
3029 return compat_urllib_request
.Request(
3030 newurl
, headers
=newheaders
, origin_req_host
=req
.origin_req_host
,
3034 def extract_timezone(date_str
):
3036 r
'^.{8,}?(?P<tz>Z$| ?(?P<sign>\+|-)(?P<hours>[0-9]{2}):?(?P<minutes>[0-9]{2})$)',
3039 timezone
= datetime
.timedelta()
3041 date_str
= date_str
[:-len(m
.group('tz'))]
3042 if not m
.group('sign'):
3043 timezone
= datetime
.timedelta()
3045 sign
= 1 if m
.group('sign') == '+' else -1
3046 timezone
= datetime
.timedelta(
3047 hours
=sign
* int(m
.group('hours')),
3048 minutes
=sign
* int(m
.group('minutes')))
3049 return timezone
, date_str
3052 def parse_iso8601(date_str
, delimiter
='T', timezone
=None):
3053 """ Return a UNIX timestamp from the given date """
3055 if date_str
is None:
3058 date_str
= re
.sub(r
'\.[0-9]+', '', date_str
)
3060 if timezone
is None:
3061 timezone
, date_str
= extract_timezone(date_str
)
3064 date_format
= '%Y-%m-%d{0}%H:%M:%S'.format(delimiter
)
3065 dt
= datetime
.datetime
.strptime(date_str
, date_format
) - timezone
3066 return calendar
.timegm(dt
.timetuple())
3071 def date_formats(day_first
=True):
3072 return DATE_FORMATS_DAY_FIRST
if day_first
else DATE_FORMATS_MONTH_FIRST
3075 def unified_strdate(date_str
, day_first
=True):
3076 """Return a string with the date in the format YYYYMMDD"""
3078 if date_str
is None:
3082 date_str
= date_str
.replace(',', ' ')
3083 # Remove AM/PM + timezone
3084 date_str
= re
.sub(r
'(?i)\s*(?:AM|PM)(?:\s+[A-Z]+)?', '', date_str
)
3085 _
, date_str
= extract_timezone(date_str
)
3087 for expression
in date_formats(day_first
):
3089 upload_date
= datetime
.datetime
.strptime(date_str
, expression
).strftime('%Y%m%d')
3092 if upload_date
is None:
3093 timetuple
= email
.utils
.parsedate_tz(date_str
)
3096 upload_date
= datetime
.datetime(*timetuple
[:6]).strftime('%Y%m%d')
3099 if upload_date
is not None:
3100 return compat_str(upload_date
)
3103 def unified_timestamp(date_str
, day_first
=True):
3104 if date_str
is None:
3107 date_str
= re
.sub(r
'[,|]', '', date_str
)
3109 pm_delta
= 12 if re
.search(r
'(?i)PM', date_str
) else 0
3110 timezone
, date_str
= extract_timezone(date_str
)
3112 # Remove AM/PM + timezone
3113 date_str
= re
.sub(r
'(?i)\s*(?:AM|PM)(?:\s+[A-Z]+)?', '', date_str
)
3115 # Remove unrecognized timezones from ISO 8601 alike timestamps
3116 m
= re
.search(r
'\d{1,2}:\d{1,2}(?:\.\d+)?(?P<tz>\s*[A-Z]+)$', date_str
)
3118 date_str
= date_str
[:-len(m
.group('tz'))]
3120 # Python only supports microseconds, so remove nanoseconds
3121 m
= re
.search(r
'^([0-9]{4,}-[0-9]{1,2}-[0-9]{1,2}T[0-9]{1,2}:[0-9]{1,2}:[0-9]{1,2}\.[0-9]{6})[0-9]+$', date_str
)
3123 date_str
= m
.group(1)
3125 for expression
in date_formats(day_first
):
3127 dt
= datetime
.datetime
.strptime(date_str
, expression
) - timezone
+ datetime
.timedelta(hours
=pm_delta
)
3128 return calendar
.timegm(dt
.timetuple())
3131 timetuple
= email
.utils
.parsedate_tz(date_str
)
3133 return calendar
.timegm(timetuple
) + pm_delta
* 3600
3136 def determine_ext(url
, default_ext
='unknown_video'):
3137 if url
is None or '.' not in url
:
3139 guess
= url
.partition('?')[0].rpartition('.')[2]
3140 if re
.match(r
'^[A-Za-z0-9]+$', guess
):
3142 # Try extract ext from URLs like http://example.com/foo/bar.mp4/?download
3143 elif guess
.rstrip('/') in KNOWN_EXTENSIONS
:
3144 return guess
.rstrip('/')
3149 def subtitles_filename(filename
, sub_lang
, sub_format
, expected_real_ext
=None):
3150 return replace_extension(filename
, sub_lang
+ '.' + sub_format
, expected_real_ext
)
3153 def datetime_from_str(date_str
, precision
='auto', format
='%Y%m%d'):
3155 Return a datetime object from a string in the format YYYYMMDD or
3156 (now|today|date)[+-][0-9](microsecond|second|minute|hour|day|week|month|year)(s)?
3158 format: string date format used to return datetime object from
3159 precision: round the time portion of a datetime object.
3160 auto|microsecond|second|minute|hour|day.
3161 auto: round to the unit provided in date_str (if applicable).
3163 auto_precision
= False
3164 if precision
== 'auto':
3165 auto_precision
= True
3166 precision
= 'microsecond'
3167 today
= datetime_round(datetime
.datetime
.now(), precision
)
3168 if date_str
in ('now', 'today'):
3170 if date_str
== 'yesterday':
3171 return today
- datetime
.timedelta(days
=1)
3173 r
'(?P<start>.+)(?P<sign>[+-])(?P<time>\d+)(?P<unit>microsecond|second|minute|hour|day|week|month|year)(s)?',
3175 if match
is not None:
3176 start_time
= datetime_from_str(match
.group('start'), precision
, format
)
3177 time
= int(match
.group('time')) * (-1 if match
.group('sign') == '-' else 1)
3178 unit
= match
.group('unit')
3179 if unit
== 'month' or unit
== 'year':
3180 new_date
= datetime_add_months(start_time
, time
* 12 if unit
== 'year' else time
)
3186 delta
= datetime
.timedelta(**{unit + 's': time}
)
3187 new_date
= start_time
+ delta
3189 return datetime_round(new_date
, unit
)
3192 return datetime_round(datetime
.datetime
.strptime(date_str
, format
), precision
)
3195 def date_from_str(date_str
, format
='%Y%m%d'):
3197 Return a datetime object from a string in the format YYYYMMDD or
3198 (now|today|date)[+-][0-9](microsecond|second|minute|hour|day|week|month|year)(s)?
3200 format: string date format used to return datetime object from
3202 return datetime_from_str(date_str
, precision
='microsecond', format
=format
).date()
3205 def datetime_add_months(dt
, months
):
3206 """Increment/Decrement a datetime object by months."""
3207 month
= dt
.month
+ months
- 1
3208 year
= dt
.year
+ month
// 12
3209 month
= month
% 12 + 1
3210 day
= min(dt
.day
, calendar
.monthrange(year
, month
)[1])
3211 return dt
.replace(year
, month
, day
)
3214 def datetime_round(dt
, precision
='day'):
3216 Round a datetime object's time to a specific precision
3218 if precision
== 'microsecond':
3227 roundto
= lambda x
, n
: ((x
+ n
/ 2) // n
) * n
3228 timestamp
= calendar
.timegm(dt
.timetuple())
3229 return datetime
.datetime
.utcfromtimestamp(roundto(timestamp
, unit_seconds
[precision
]))
3232 def hyphenate_date(date_str
):
3234 Convert a date in 'YYYYMMDD' format to 'YYYY-MM-DD' format"""
3235 match
= re
.match(r
'^(\d\d\d\d)(\d\d)(\d\d)$', date_str
)
3236 if match
is not None:
3237 return '-'.join(match
.groups())
3242 class DateRange(object):
3243 """Represents a time interval between two dates"""
3245 def __init__(self
, start
=None, end
=None):
3246 """start and end must be strings in the format accepted by date"""
3247 if start
is not None:
3248 self
.start
= date_from_str(start
)
3250 self
.start
= datetime
.datetime
.min.date()
3252 self
.end
= date_from_str(end
)
3254 self
.end
= datetime
.datetime
.max.date()
3255 if self
.start
> self
.end
:
3256 raise ValueError('Date range: "%s" , the start date must be before the end date' % self
)
3260 """Returns a range that only contains the given day"""
3261 return cls(day
, day
)
3263 def __contains__(self
, date
):
3264 """Check if the date is in the range"""
3265 if not isinstance(date
, datetime
.date
):
3266 date
= date_from_str(date
)
3267 return self
.start
<= date
<= self
.end
3270 return '%s - %s' % (self
.start
.isoformat(), self
.end
.isoformat())
3273 def platform_name():
3274 """ Returns the platform name as a compat_str """
3275 res
= platform
.platform()
3276 if isinstance(res
, bytes):
3277 res
= res
.decode(preferredencoding())
3279 assert isinstance(res
, compat_str
)
3283 def _windows_write_string(s
, out
):
3284 """ Returns True if the string was written using special methods,
3285 False if it has yet to be written out."""
3286 # Adapted from http://stackoverflow.com/a/3259271/35070
3289 import ctypes
.wintypes
3297 fileno
= out
.fileno()
3298 except AttributeError:
3299 # If the output stream doesn't have a fileno, it's virtual
3301 except io
.UnsupportedOperation
:
3302 # Some strange Windows pseudo files?
3304 if fileno
not in WIN_OUTPUT_IDS
:
3307 GetStdHandle
= compat_ctypes_WINFUNCTYPE(
3308 ctypes
.wintypes
.HANDLE
, ctypes
.wintypes
.DWORD
)(
3309 ('GetStdHandle', ctypes
.windll
.kernel32
))
3310 h
= GetStdHandle(WIN_OUTPUT_IDS
[fileno
])
3312 WriteConsoleW
= compat_ctypes_WINFUNCTYPE(
3313 ctypes
.wintypes
.BOOL
, ctypes
.wintypes
.HANDLE
, ctypes
.wintypes
.LPWSTR
,
3314 ctypes
.wintypes
.DWORD
, ctypes
.POINTER(ctypes
.wintypes
.DWORD
),
3315 ctypes
.wintypes
.LPVOID
)(('WriteConsoleW', ctypes
.windll
.kernel32
))
3316 written
= ctypes
.wintypes
.DWORD(0)
3318 GetFileType
= compat_ctypes_WINFUNCTYPE(ctypes
.wintypes
.DWORD
, ctypes
.wintypes
.DWORD
)(('GetFileType', ctypes
.windll
.kernel32
))
3319 FILE_TYPE_CHAR
= 0x0002
3320 FILE_TYPE_REMOTE
= 0x8000
3321 GetConsoleMode
= compat_ctypes_WINFUNCTYPE(
3322 ctypes
.wintypes
.BOOL
, ctypes
.wintypes
.HANDLE
,
3323 ctypes
.POINTER(ctypes
.wintypes
.DWORD
))(
3324 ('GetConsoleMode', ctypes
.windll
.kernel32
))
3325 INVALID_HANDLE_VALUE
= ctypes
.wintypes
.DWORD(-1).value
3327 def not_a_console(handle
):
3328 if handle
== INVALID_HANDLE_VALUE
or handle
is None:
3330 return ((GetFileType(handle
) & ~FILE_TYPE_REMOTE
) != FILE_TYPE_CHAR
3331 or GetConsoleMode(handle
, ctypes
.byref(ctypes
.wintypes
.DWORD())) == 0)
3333 if not_a_console(h
):
3336 def next_nonbmp_pos(s
):
3338 return next(i
for i
, c
in enumerate(s
) if ord(c
) > 0xffff)
3339 except StopIteration:
3343 count
= min(next_nonbmp_pos(s
), 1024)
3345 ret
= WriteConsoleW(
3346 h
, s
, count
if count
else 2, ctypes
.byref(written
), None)
3348 raise OSError('Failed to write string')
3349 if not count
: # We just wrote a non-BMP character
3350 assert written
.value
== 2
3353 assert written
.value
> 0
3354 s
= s
[written
.value
:]
3358 def write_string(s
, out
=None, encoding
=None):
3361 assert type(s
) == compat_str
3363 if sys
.platform
== 'win32' and encoding
is None and hasattr(out
, 'fileno'):
3364 if _windows_write_string(s
, out
):
3367 if ('b' in getattr(out
, 'mode', '')
3368 or sys
.version_info
[0] < 3): # Python 2 lies about mode of sys.stderr
3369 byt
= s
.encode(encoding
or preferredencoding(), 'ignore')
3371 elif hasattr(out
, 'buffer'):
3372 enc
= encoding
or getattr(out
, 'encoding', None) or preferredencoding()
3373 byt
= s
.encode(enc
, 'ignore')
3374 out
.buffer.write(byt
)
3380 def bytes_to_intlist(bs
):
3383 if isinstance(bs
[0], int): # Python 3
3386 return [ord(c
) for c
in bs
]
3389 def intlist_to_bytes(xs
):
3392 return compat_struct_pack('%dB' % len(xs
), *xs
)
3395 # Cross-platform file locking
3396 if sys
.platform
== 'win32':
3397 import ctypes
.wintypes
3400 class OVERLAPPED(ctypes
.Structure
):
3402 ('Internal', ctypes
.wintypes
.LPVOID
),
3403 ('InternalHigh', ctypes
.wintypes
.LPVOID
),
3404 ('Offset', ctypes
.wintypes
.DWORD
),
3405 ('OffsetHigh', ctypes
.wintypes
.DWORD
),
3406 ('hEvent', ctypes
.wintypes
.HANDLE
),
3409 kernel32
= ctypes
.windll
.kernel32
3410 LockFileEx
= kernel32
.LockFileEx
3411 LockFileEx
.argtypes
= [
3412 ctypes
.wintypes
.HANDLE
, # hFile
3413 ctypes
.wintypes
.DWORD
, # dwFlags
3414 ctypes
.wintypes
.DWORD
, # dwReserved
3415 ctypes
.wintypes
.DWORD
, # nNumberOfBytesToLockLow
3416 ctypes
.wintypes
.DWORD
, # nNumberOfBytesToLockHigh
3417 ctypes
.POINTER(OVERLAPPED
) # Overlapped
3419 LockFileEx
.restype
= ctypes
.wintypes
.BOOL
3420 UnlockFileEx
= kernel32
.UnlockFileEx
3421 UnlockFileEx
.argtypes
= [
3422 ctypes
.wintypes
.HANDLE
, # hFile
3423 ctypes
.wintypes
.DWORD
, # dwReserved
3424 ctypes
.wintypes
.DWORD
, # nNumberOfBytesToLockLow
3425 ctypes
.wintypes
.DWORD
, # nNumberOfBytesToLockHigh
3426 ctypes
.POINTER(OVERLAPPED
) # Overlapped
3428 UnlockFileEx
.restype
= ctypes
.wintypes
.BOOL
3429 whole_low
= 0xffffffff
3430 whole_high
= 0x7fffffff
3432 def _lock_file(f
, exclusive
):
3433 overlapped
= OVERLAPPED()
3434 overlapped
.Offset
= 0
3435 overlapped
.OffsetHigh
= 0
3436 overlapped
.hEvent
= 0
3437 f
._lock
_file
_overlapped
_p
= ctypes
.pointer(overlapped
)
3438 handle
= msvcrt
.get_osfhandle(f
.fileno())
3439 if not LockFileEx(handle
, 0x2 if exclusive
else 0x0, 0,
3440 whole_low
, whole_high
, f
._lock
_file
_overlapped
_p
):
3441 raise OSError('Locking file failed: %r' % ctypes
.FormatError())
3443 def _unlock_file(f
):
3444 assert f
._lock
_file
_overlapped
_p
3445 handle
= msvcrt
.get_osfhandle(f
.fileno())
3446 if not UnlockFileEx(handle
, 0,
3447 whole_low
, whole_high
, f
._lock
_file
_overlapped
_p
):
3448 raise OSError('Unlocking file failed: %r' % ctypes
.FormatError())
3451 # Some platforms, such as Jython, is missing fcntl
3455 def _lock_file(f
, exclusive
):
3456 fcntl
.flock(f
, fcntl
.LOCK_EX
if exclusive
else fcntl
.LOCK_SH
)
3458 def _unlock_file(f
):
3459 fcntl
.flock(f
, fcntl
.LOCK_UN
)
3461 UNSUPPORTED_MSG
= 'file locking is not supported on this platform'
3463 def _lock_file(f
, exclusive
):
3464 raise IOError(UNSUPPORTED_MSG
)
3466 def _unlock_file(f
):
3467 raise IOError(UNSUPPORTED_MSG
)
3470 class locked_file(object):
3471 def __init__(self
, filename
, mode
, encoding
=None):
3472 assert mode
in ['r', 'a', 'w']
3473 self
.f
= io
.open(filename
, mode
, encoding
=encoding
)
3476 def __enter__(self
):
3477 exclusive
= self
.mode
!= 'r'
3479 _lock_file(self
.f
, exclusive
)
3485 def __exit__(self
, etype
, value
, traceback
):
3487 _unlock_file(self
.f
)
3494 def write(self
, *args
):
3495 return self
.f
.write(*args
)
3497 def read(self
, *args
):
3498 return self
.f
.read(*args
)
3501 def get_filesystem_encoding():
3502 encoding
= sys
.getfilesystemencoding()
3503 return encoding
if encoding
is not None else 'utf-8'
3506 def shell_quote(args
):
3508 encoding
= get_filesystem_encoding()
3510 if isinstance(a
, bytes):
3511 # We may get a filename encoded with 'encodeFilename'
3512 a
= a
.decode(encoding
)
3513 quoted_args
.append(compat_shlex_quote(a
))
3514 return ' '.join(quoted_args
)
3517 def smuggle_url(url
, data
):
3518 """ Pass additional data in a URL for internal use. """
3520 url
, idata
= unsmuggle_url(url
, {})
3522 sdata
= compat_urllib_parse_urlencode(
3523 {'__youtubedl_smuggle': json.dumps(data)}
)
3524 return url
+ '#' + sdata
3527 def unsmuggle_url(smug_url
, default
=None):
3528 if '#__youtubedl_smuggle' not in smug_url
:
3529 return smug_url
, default
3530 url
, _
, sdata
= smug_url
.rpartition('#')
3531 jsond
= compat_parse_qs(sdata
)['__youtubedl_smuggle'][0]
3532 data
= json
.loads(jsond
)
3536 def format_bytes(bytes):
3539 if type(bytes) is str:
3540 bytes = float(bytes)
3544 exponent
= int(math
.log(bytes, 1024.0))
3545 suffix
= ['B', 'KiB', 'MiB', 'GiB', 'TiB', 'PiB', 'EiB', 'ZiB', 'YiB'][exponent
]
3546 converted
= float(bytes) / float(1024 ** exponent
)
3547 return '%.2f%s' % (converted
, suffix
)
3550 def lookup_unit_table(unit_table
, s
):
3551 units_re
= '|'.join(re
.escape(u
) for u
in unit_table
)
3553 r
'(?P<num>[0-9]+(?:[,.][0-9]*)?)\s*(?P<unit>%s)\b' % units_re
, s
)
3556 num_str
= m
.group('num').replace(',', '.')
3557 mult
= unit_table
[m
.group('unit')]
3558 return int(float(num_str
) * mult
)
3561 def parse_filesize(s
):
3565 # The lower-case forms are of course incorrect and unofficial,
3566 # but we support those too
3583 'megabytes': 1000 ** 2,
3584 'mebibytes': 1024 ** 2,
3590 'gigabytes': 1000 ** 3,
3591 'gibibytes': 1024 ** 3,
3597 'terabytes': 1000 ** 4,
3598 'tebibytes': 1024 ** 4,
3604 'petabytes': 1000 ** 5,
3605 'pebibytes': 1024 ** 5,
3611 'exabytes': 1000 ** 6,
3612 'exbibytes': 1024 ** 6,
3618 'zettabytes': 1000 ** 7,
3619 'zebibytes': 1024 ** 7,
3625 'yottabytes': 1000 ** 8,
3626 'yobibytes': 1024 ** 8,
3629 return lookup_unit_table(_UNIT_TABLE
, s
)
3638 if re
.match(r
'^[\d,.]+$', s
):
3639 return str_to_int(s
)
3650 return lookup_unit_table(_UNIT_TABLE
, s
)
3653 def parse_resolution(s
):
3657 mobj
= re
.search(r
'\b(?P<w>\d+)\s*[xX×]\s*(?P<h>\d+)\b', s
)
3660 'width': int(mobj
.group('w')),
3661 'height': int(mobj
.group('h')),
3664 mobj
= re
.search(r
'\b(\d+)[pPiI]\b', s
)
3666 return {'height': int(mobj.group(1))}
3668 mobj
= re
.search(r
'\b([48])[kK]\b', s
)
3670 return {'height': int(mobj.group(1)) * 540}
3675 def parse_bitrate(s
):
3676 if not isinstance(s
, compat_str
):
3678 mobj
= re
.search(r
'\b(\d+)\s*kbps', s
)
3680 return int(mobj
.group(1))
3683 def month_by_name(name
, lang
='en'):
3684 """ Return the number of a month by (locale-independently) English name """
3686 month_names
= MONTH_NAMES
.get(lang
, MONTH_NAMES
['en'])
3689 return month_names
.index(name
) + 1
3694 def month_by_abbreviation(abbrev
):
3695 """ Return the number of a month by (locale-independently) English
3699 return [s
[:3] for s
in ENGLISH_MONTH_NAMES
].index(abbrev
) + 1
3704 def fix_xml_ampersands(xml_str
):
3705 """Replace all the '&' by '&' in XML"""
3707 r
'&(?!amp;|lt;|gt;|apos;|quot;|#x[0-9a-fA-F]{,4};|#[0-9]{,4};)',
3712 def setproctitle(title
):
3713 assert isinstance(title
, compat_str
)
3715 # ctypes in Jython is not complete
3716 # http://bugs.jython.org/issue2148
3717 if sys
.platform
.startswith('java'):
3721 libc
= ctypes
.cdll
.LoadLibrary('libc.so.6')
3725 # LoadLibrary in Windows Python 2.7.13 only expects
3726 # a bytestring, but since unicode_literals turns
3727 # every string into a unicode string, it fails.
3729 title_bytes
= title
.encode('utf-8')
3730 buf
= ctypes
.create_string_buffer(len(title_bytes
))
3731 buf
.value
= title_bytes
3733 libc
.prctl(15, buf
, 0, 0, 0)
3734 except AttributeError:
3735 return # Strange libc, just skip this
3738 def remove_start(s
, start
):
3739 return s
[len(start
):] if s
is not None and s
.startswith(start
) else s
3742 def remove_end(s
, end
):
3743 return s
[:-len(end
)] if s
is not None and s
.endswith(end
) else s
3746 def remove_quotes(s
):
3747 if s
is None or len(s
) < 2:
3749 for quote
in ('"', "'", ):
3750 if s
[0] == quote
and s
[-1] == quote
:
3755 def get_domain(url
):
3756 domain
= re
.match(r
'(?:https?:\/\/)?(?:www\.)?(?P<domain>[^\n\/]+\.[^\n\/]+)(?:\/(.*))?', url
)
3757 return domain
.group('domain') if domain
else None
3760 def url_basename(url
):
3761 path
= compat_urlparse
.urlparse(url
).path
3762 return path
.strip('/').split('/')[-1]
3766 return re
.match(r
'https?://[^?#&]+/', url
).group()
3769 def urljoin(base
, path
):
3770 if isinstance(path
, bytes):
3771 path
= path
.decode('utf-8')
3772 if not isinstance(path
, compat_str
) or not path
:
3774 if re
.match(r
'^(?:[a-zA-Z][a-zA-Z0-9+-.]*:)?//', path
):
3776 if isinstance(base
, bytes):
3777 base
= base
.decode('utf-8')
3778 if not isinstance(base
, compat_str
) or not re
.match(
3779 r
'^(?:https?:)?//', base
):
3781 return compat_urlparse
.urljoin(base
, path
)
3784 class HEADRequest(compat_urllib_request
.Request
):
3785 def get_method(self
):
3789 class PUTRequest(compat_urllib_request
.Request
):
3790 def get_method(self
):
3794 def int_or_none(v
, scale
=1, default
=None, get_attr
=None, invscale
=1):
3797 v
= getattr(v
, get_attr
, None)
3803 return int(v
) * invscale
// scale
3804 except (ValueError, TypeError):
3808 def str_or_none(v
, default
=None):
3809 return default
if v
is None else compat_str(v
)
3812 def str_to_int(int_str
):
3813 """ A more relaxed version of int_or_none """
3814 if isinstance(int_str
, compat_integer_types
):
3816 elif isinstance(int_str
, compat_str
):
3817 int_str
= re
.sub(r
'[,\.\+]', '', int_str
)
3818 return int_or_none(int_str
)
3821 def float_or_none(v
, scale
=1, invscale
=1, default
=None):
3825 return float(v
) * invscale
/ scale
3826 except (ValueError, TypeError):
3830 def bool_or_none(v
, default
=None):
3831 return v
if isinstance(v
, bool) else default
3834 def strip_or_none(v
, default
=None):
3835 return v
.strip() if isinstance(v
, compat_str
) else default
3838 def url_or_none(url
):
3839 if not url
or not isinstance(url
, compat_str
):
3842 return url
if re
.match(r
'^(?:(?:https?|rt(?:m(?:pt?[es]?|fp)|sp[su]?)|mms|ftps?):)?//', url
) else None
3845 def strftime_or_none(timestamp
, date_format
, default
=None):
3846 datetime_object
= None
3848 if isinstance(timestamp
, compat_numeric_types
): # unix timestamp
3849 datetime_object
= datetime
.datetime
.utcfromtimestamp(timestamp
)
3850 elif isinstance(timestamp
, compat_str
): # assume YYYYMMDD
3851 datetime_object
= datetime
.datetime
.strptime(timestamp
, '%Y%m%d')
3852 return datetime_object
.strftime(date_format
)
3853 except (ValueError, TypeError, AttributeError):
3857 def parse_duration(s
):
3858 if not isinstance(s
, compat_basestring
):
3863 days
, hours
, mins
, secs
, ms
= [None] * 5
3864 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
)
3866 days
, hours
, mins
, secs
, ms
= m
.groups()
3871 [0-9]+\s*y(?:ears?)?\s*
3874 [0-9]+\s*m(?:onths?)?\s*
3877 [0-9]+\s*w(?:eeks?)?\s*
3880 (?P<days>[0-9]+)\s*d(?:ays?)?\s*
3884 (?P<hours>[0-9]+)\s*h(?:ours?)?\s*
3887 (?P<mins>[0-9]+)\s*m(?:in(?:ute)?s?)?\s*
3890 (?P<secs>[0-9]+)(?P<ms>\.[0-9]+)?\s*s(?:ec(?:ond)?s?)?\s*
3893 days
, hours
, mins
, secs
, ms
= m
.groups()
3895 m
= re
.match(r
'(?i)(?:(?P<hours>[0-9.]+)\s*(?:hours?)|(?P<mins>[0-9.]+)\s*(?:mins?\.?|minutes?)\s*)Z?$', s
)
3897 hours
, mins
= m
.groups()
3903 duration
+= float(secs
)
3905 duration
+= float(mins
) * 60
3907 duration
+= float(hours
) * 60 * 60
3909 duration
+= float(days
) * 24 * 60 * 60
3911 duration
+= float(ms
)
3915 def prepend_extension(filename
, ext
, expected_real_ext
=None):
3916 name
, real_ext
= os
.path
.splitext(filename
)
3918 '{0}.{1}{2}'.format(name
, ext
, real_ext
)
3919 if not expected_real_ext
or real_ext
[1:] == expected_real_ext
3920 else '{0}.{1}'.format(filename
, ext
))
3923 def replace_extension(filename
, ext
, expected_real_ext
=None):
3924 name
, real_ext
= os
.path
.splitext(filename
)
3925 return '{0}.{1}'.format(
3926 name
if not expected_real_ext
or real_ext
[1:] == expected_real_ext
else filename
,
3930 def check_executable(exe
, args
=[]):
3931 """ Checks if the given binary is installed somewhere in PATH, and returns its name.
3932 args can be a list of arguments for a short output (like -version) """
3934 process_communicate_or_kill(subprocess
.Popen(
3935 [exe
] + args
, stdout
=subprocess
.PIPE
, stderr
=subprocess
.PIPE
))
3941 def get_exe_version(exe
, args
=['--version'],
3942 version_re
=None, unrecognized
='present'):
3943 """ Returns the version of the specified executable,
3944 or False if the executable is not present """
3946 # STDIN should be redirected too. On UNIX-like systems, ffmpeg triggers
3947 # SIGTTOU if yt-dlp is run in the background.
3948 # See https://github.com/ytdl-org/youtube-dl/issues/955#issuecomment-209789656
3949 out
, _
= process_communicate_or_kill(subprocess
.Popen(
3950 [encodeArgument(exe
)] + args
,
3951 stdin
=subprocess
.PIPE
,
3952 stdout
=subprocess
.PIPE
, stderr
=subprocess
.STDOUT
))
3955 if isinstance(out
, bytes): # Python 2.x
3956 out
= out
.decode('ascii', 'ignore')
3957 return detect_exe_version(out
, version_re
, unrecognized
)
3960 def detect_exe_version(output
, version_re
=None, unrecognized
='present'):
3961 assert isinstance(output
, compat_str
)
3962 if version_re
is None:
3963 version_re
= r
'version\s+([-0-9._a-zA-Z]+)'
3964 m
= re
.search(version_re
, output
)
3971 class LazyList(collections
.abc
.Sequence
):
3972 ''' Lazy immutable list from an iterable
3973 Note that slices of a LazyList are lists and not LazyList'''
3975 class IndexError(IndexError):
3978 def __init__(self
, iterable
):
3979 self
.__iterable
= iter(iterable
)
3981 self
.__reversed
= False
3985 # We need to consume the entire iterable to iterate in reverse
3986 yield from self
.exhaust()
3988 yield from self
.__cache
3989 for item
in self
.__iterable
:
3990 self
.__cache
.append(item
)
3993 def __exhaust(self
):
3994 self
.__cache
.extend(self
.__iterable
)
3998 ''' Evaluate the entire iterable '''
3999 return self
.__exhaust
()[::-1 if self
.__reversed
else 1]
4002 def __reverse_index(x
):
4003 return None if x
is None else -(x
+ 1)
4005 def __getitem__(self
, idx
):
4006 if isinstance(idx
, slice):
4008 idx
= slice(self
.__reverse
_index
(idx
.start
), self
.__reverse
_index
(idx
.stop
), -(idx
.step
or 1))
4009 start
, stop
, step
= idx
.start
, idx
.stop
, idx
.step
or 1
4010 elif isinstance(idx
, int):
4012 idx
= self
.__reverse
_index
(idx
)
4013 start
, stop
, step
= idx
, idx
, 0
4015 raise TypeError('indices must be integers or slices')
4016 if ((start
or 0) < 0 or (stop
or 0) < 0
4017 or (start
is None and step
< 0)
4018 or (stop
is None and step
> 0)):
4019 # We need to consume the entire iterable to be able to slice from the end
4020 # Obviously, never use this with infinite iterables
4023 return self
.__cache
[idx
]
4024 except IndexError as e
:
4025 raise self
.IndexError(e
) from e
4026 n
= max(start
or 0, stop
or 0) - len(self
.__cache
) + 1
4028 self
.__cache
.extend(itertools
.islice(self
.__iterable
, n
))
4030 return self
.__cache
[idx
]
4031 except IndexError as e
:
4032 raise self
.IndexError(e
) from e
4036 self
[-1] if self
.__reversed
else self
[0]
4037 except self
.IndexError:
4043 return len(self
.__cache
)
4046 self
.__reversed
= not self
.__reversed
4050 # repr and str should mimic a list. So we exhaust the iterable
4051 return repr(self
.exhaust())
4054 return repr(self
.exhaust())
4059 # This is only useful for tests
4060 return len(self
.getslice())
4062 def __init__(self
, pagefunc
, pagesize
, use_cache
=True):
4063 self
._pagefunc
= pagefunc
4064 self
._pagesize
= pagesize
4065 self
._use
_cache
= use_cache
4068 def getpage(self
, pagenum
):
4069 page_results
= self
._cache
.get(pagenum
) or list(self
._pagefunc
(pagenum
))
4071 self
._cache
[pagenum
] = page_results
4074 def getslice(self
, start
=0, end
=None):
4075 return list(self
._getslice
(start
, end
))
4077 def _getslice(self
, start
, end
):
4078 raise NotImplementedError('This method must be implemented by subclasses')
4080 def __getitem__(self
, idx
):
4081 # NOTE: cache must be enabled if this is used
4082 if not isinstance(idx
, int) or idx
< 0:
4083 raise TypeError('indices must be non-negative integers')
4084 entries
= self
.getslice(idx
, idx
+ 1)
4085 return entries
[0] if entries
else None
4088 class OnDemandPagedList(PagedList
):
4089 def _getslice(self
, start
, end
):
4090 for pagenum
in itertools
.count(start
// self
._pagesize
):
4091 firstid
= pagenum
* self
._pagesize
4092 nextfirstid
= pagenum
* self
._pagesize
+ self
._pagesize
4093 if start
>= nextfirstid
:
4097 start
% self
._pagesize
4098 if firstid
<= start
< nextfirstid
4101 ((end
- 1) % self
._pagesize
) + 1
4102 if (end
is not None and firstid
<= end
<= nextfirstid
)
4105 page_results
= self
.getpage(pagenum
)
4106 if startv
!= 0 or endv
is not None:
4107 page_results
= page_results
[startv
:endv
]
4108 yield from page_results
4110 # A little optimization - if current page is not "full", ie. does
4111 # not contain page_size videos then we can assume that this page
4112 # is the last one - there are no more ids on further pages -
4113 # i.e. no need to query again.
4114 if len(page_results
) + startv
< self
._pagesize
:
4117 # If we got the whole page, but the next page is not interesting,
4118 # break out early as well
4119 if end
== nextfirstid
:
4123 class InAdvancePagedList(PagedList
):
4124 def __init__(self
, pagefunc
, pagecount
, pagesize
):
4125 self
._pagecount
= pagecount
4126 PagedList
.__init
__(self
, pagefunc
, pagesize
, True)
4128 def _getslice(self
, start
, end
):
4129 start_page
= start
// self
._pagesize
4131 self
._pagecount
if end
is None else (end
// self
._pagesize
+ 1))
4132 skip_elems
= start
- start_page
* self
._pagesize
4133 only_more
= None if end
is None else end
- start
4134 for pagenum
in range(start_page
, end_page
):
4135 page_results
= self
.getpage(pagenum
)
4137 page_results
= page_results
[skip_elems
:]
4139 if only_more
is not None:
4140 if len(page_results
) < only_more
:
4141 only_more
-= len(page_results
)
4143 yield from page_results
[:only_more
]
4145 yield from page_results
4148 def uppercase_escape(s
):
4149 unicode_escape
= codecs
.getdecoder('unicode_escape')
4151 r
'\\U[0-9a-fA-F]{8}',
4152 lambda m
: unicode_escape(m
.group(0))[0],
4156 def lowercase_escape(s
):
4157 unicode_escape
= codecs
.getdecoder('unicode_escape')
4159 r
'\\u[0-9a-fA-F]{4}',
4160 lambda m
: unicode_escape(m
.group(0))[0],
4164 def escape_rfc3986(s
):
4165 """Escape non-ASCII characters as suggested by RFC 3986"""
4166 if sys
.version_info
< (3, 0) and isinstance(s
, compat_str
):
4167 s
= s
.encode('utf-8')
4168 return compat_urllib_parse
.quote(s
, b
"%/;:@&=+$,!~*'()?#[]")
4171 def escape_url(url
):
4172 """Escape URL as suggested by RFC 3986"""
4173 url_parsed
= compat_urllib_parse_urlparse(url
)
4174 return url_parsed
._replace
(
4175 netloc
=url_parsed
.netloc
.encode('idna').decode('ascii'),
4176 path
=escape_rfc3986(url_parsed
.path
),
4177 params
=escape_rfc3986(url_parsed
.params
),
4178 query
=escape_rfc3986(url_parsed
.query
),
4179 fragment
=escape_rfc3986(url_parsed
.fragment
)
4184 return compat_parse_qs(compat_urllib_parse_urlparse(url
).query
)
4187 def read_batch_urls(batch_fd
):
4189 if not isinstance(url
, compat_str
):
4190 url
= url
.decode('utf-8', 'replace')
4191 BOM_UTF8
= ('\xef\xbb\xbf', '\ufeff')
4192 for bom
in BOM_UTF8
:
4193 if url
.startswith(bom
):
4194 url
= url
[len(bom
):]
4196 if not url
or url
.startswith(('#', ';', ']')):
4198 # "#" cannot be stripped out since it is part of the URI
4199 # However, it can be safely stipped out if follwing a whitespace
4200 return re
.split(r
'\s#', url
, 1)[0].rstrip()
4202 with contextlib
.closing(batch_fd
) as fd
:
4203 return [url
for url
in map(fixup
, fd
) if url
]
4206 def urlencode_postdata(*args
, **kargs
):
4207 return compat_urllib_parse_urlencode(*args
, **kargs
).encode('ascii')
4210 def update_url_query(url
, query
):
4213 parsed_url
= compat_urlparse
.urlparse(url
)
4214 qs
= compat_parse_qs(parsed_url
.query
)
4216 return compat_urlparse
.urlunparse(parsed_url
._replace
(
4217 query
=compat_urllib_parse_urlencode(qs
, True)))
4220 def update_Request(req
, url
=None, data
=None, headers
={}, query={}
):
4221 req_headers
= req
.headers
.copy()
4222 req_headers
.update(headers
)
4223 req_data
= data
or req
.data
4224 req_url
= update_url_query(url
or req
.get_full_url(), query
)
4225 req_get_method
= req
.get_method()
4226 if req_get_method
== 'HEAD':
4227 req_type
= HEADRequest
4228 elif req_get_method
== 'PUT':
4229 req_type
= PUTRequest
4231 req_type
= compat_urllib_request
.Request
4233 req_url
, data
=req_data
, headers
=req_headers
,
4234 origin_req_host
=req
.origin_req_host
, unverifiable
=req
.unverifiable
)
4235 if hasattr(req
, 'timeout'):
4236 new_req
.timeout
= req
.timeout
4240 def _multipart_encode_impl(data
, boundary
):
4241 content_type
= 'multipart/form-data; boundary=%s' % boundary
4244 for k
, v
in data
.items():
4245 out
+= b
'--' + boundary
.encode('ascii') + b
'\r\n'
4246 if isinstance(k
, compat_str
):
4247 k
= k
.encode('utf-8')
4248 if isinstance(v
, compat_str
):
4249 v
= v
.encode('utf-8')
4250 # RFC 2047 requires non-ASCII field names to be encoded, while RFC 7578
4251 # suggests sending UTF-8 directly. Firefox sends UTF-8, too
4252 content
= b
'Content-Disposition: form-data; name="' + k
+ b
'"\r\n\r\n' + v
+ b
'\r\n'
4253 if boundary
.encode('ascii') in content
:
4254 raise ValueError('Boundary overlaps with data')
4257 out
+= b
'--' + boundary
.encode('ascii') + b
'--\r\n'
4259 return out
, content_type
4262 def multipart_encode(data
, boundary
=None):
4264 Encode a dict to RFC 7578-compliant form-data
4267 A dict where keys and values can be either Unicode or bytes-like
4270 If specified a Unicode object, it's used as the boundary. Otherwise
4271 a random boundary is generated.
4273 Reference: https://tools.ietf.org/html/rfc7578
4275 has_specified_boundary
= boundary
is not None
4278 if boundary
is None:
4279 boundary
= '---------------' + str(random
.randrange(0x0fffffff, 0xffffffff))
4282 out
, content_type
= _multipart_encode_impl(data
, boundary
)
4285 if has_specified_boundary
:
4289 return out
, content_type
4292 def dict_get(d
, key_or_keys
, default
=None, skip_false_values
=True):
4293 if isinstance(key_or_keys
, (list, tuple)):
4294 for key
in key_or_keys
:
4295 if key
not in d
or d
[key
] is None or skip_false_values
and not d
[key
]:
4299 return d
.get(key_or_keys
, default
)
4302 def try_get(src
, getter
, expected_type
=None):
4303 for get
in variadic(getter
):
4306 except (AttributeError, KeyError, TypeError, IndexError):
4309 if expected_type
is None or isinstance(v
, expected_type
):
4313 def merge_dicts(*dicts
):
4315 for a_dict
in dicts
:
4316 for k
, v
in a_dict
.items():
4320 or (isinstance(v
, compat_str
) and v
4321 and isinstance(merged
[k
], compat_str
)
4322 and not merged
[k
])):
4327 def encode_compat_str(string
, encoding
=preferredencoding(), errors
='strict'):
4328 return string
if isinstance(string
, compat_str
) else compat_str(string
, encoding
, errors
)
4340 TV_PARENTAL_GUIDELINES
= {
4350 def parse_age_limit(s
):
4352 return s
if 0 <= s
<= 21 else None
4353 if not isinstance(s
, compat_basestring
):
4355 m
= re
.match(r
'^(?P<age>\d{1,2})\+?$', s
)
4357 return int(m
.group('age'))
4360 return US_RATINGS
[s
]
4361 m
= re
.match(r
'^TV[_-]?(%s)$' % '|'.join(k
[3:] for k
in TV_PARENTAL_GUIDELINES
), s
)
4363 return TV_PARENTAL_GUIDELINES
['TV-' + m
.group(1)]
4367 def strip_jsonp(code
):
4370 (?:window\.)?(?P<func_name>[a-zA-Z0-9_.$]*)
4371 (?:\s*&&\s*(?P=func_name))?
4372 \s*\(\s*(?P<callback_data>.*)\);?
4373 \s*?(?://[^\n]*)*$''',
4374 r
'\g<callback_data>', code
)
4377 def js_to_json(code
, vars={}):
4378 # vars is a dict of var, val pairs to substitute
4379 COMMENT_RE
= r
'/\*(?:(?!\*/).)*?\*/|//[^\n]*\n'
4380 SKIP_RE
= r
'\s*(?:{comment})?\s*'.format(comment
=COMMENT_RE
)
4382 (r
'(?s)^(0[xX][0-9a-fA-F]+){skip}:?$'.format(skip
=SKIP_RE
), 16),
4383 (r
'(?s)^(0+[0-7]+){skip}:?$'.format(skip
=SKIP_RE
), 8),
4388 if v
in ('true', 'false', 'null'):
4390 elif v
in ('undefined', 'void 0'):
4392 elif v
.startswith('/*') or v
.startswith('//') or v
.startswith('!') or v
== ',':
4395 if v
[0] in ("'", '"'):
4396 v
= re
.sub(r
'(?s)\\.|"', lambda m
: {
4401 }.get(m
.group(0), m
.group(0)), v
[1:-1])
4403 for regex
, base
in INTEGER_TABLE
:
4404 im
= re
.match(regex
, v
)
4406 i
= int(im
.group(1), base
)
4407 return '"%d":' % i
if v
.endswith(':') else '%d' % i
4414 return re
.sub(r
'''(?sx)
4415 "(?:[^"\\]*(?:\\\\|\\['"nurtbfx/\n]))*[^"\\]*"|
4416 '(?:[^'\\]*(?:\\\\|\\['"nurtbfx/\n]))*[^'\\]*'|
4417 {comment}|,(?={skip}[\]}}])|
4418 void\s0|(?:(?<![0-9])[eE]|[a-df-zA-DF-Z_$])[.a-zA-Z_$0-9]*|
4419 \b(?:0[xX][0-9a-fA-F]+|0+[0-7]+)(?:{skip}:)?|
4422 '''.format(comment
=COMMENT_RE
, skip
=SKIP_RE
), fix_kv
, code
)
4425 def qualities(quality_ids
):
4426 """ Get a numeric quality value out of a list of possible values """
4429 return quality_ids
.index(qid
)
4436 'default': '%(title)s [%(id)s].%(ext)s',
4437 'chapter': '%(title)s - %(section_number)03d %(section_title)s [%(id)s].%(ext)s',
4443 'description': 'description',
4444 'annotation': 'annotations.xml',
4445 'infojson': 'info.json',
4446 'pl_thumbnail': None,
4447 'pl_description': 'description',
4448 'pl_infojson': 'info.json',
4451 # As of [1] format syntax is:
4452 # %[mapping_key][conversion_flags][minimum_width][.precision][length_modifier]type
4453 # 1. https://docs.python.org/2/library/stdtypes.html#string-formatting
4454 STR_FORMAT_RE_TMPL
= r
'''(?x)
4455 (?<!%)(?P<prefix>(?:%%)*)
4457 (?P<has_key>\((?P<key>{0})\))? # mapping key
4459 (?:[#0\-+ ]+)? # conversion flags (optional)
4460 (?:\d+)? # minimum field width (optional)
4461 (?:\.\d+)? # precision (optional)
4462 [hlL]? # length modifier (optional)
4463 {1} # conversion type
4468 STR_FORMAT_TYPES
= 'diouxXeEfFgGcrs'
4471 def limit_length(s
, length
):
4472 """ Add ellipses to overly long strings """
4477 return s
[:length
- len(ELLIPSES
)] + ELLIPSES
4481 def version_tuple(v
):
4482 return tuple(int(e
) for e
in re
.split(r
'[-.]', v
))
4485 def is_outdated_version(version
, limit
, assume_new
=True):
4487 return not assume_new
4489 return version_tuple(version
) < version_tuple(limit
)
4491 return not assume_new
4494 def ytdl_is_updateable():
4495 """ Returns if yt-dlp can be updated with -U """
4498 from zipimport
import zipimporter
4500 return isinstance(globals().get('__loader__'), zipimporter
) or hasattr(sys
, 'frozen')
4503 def args_to_str(args
):
4504 # Get a short string representation for a subprocess command
4505 return ' '.join(compat_shlex_quote(a
) for a
in args
)
4508 def error_to_compat_str(err
):
4510 # On python 2 error byte string must be decoded with proper
4511 # encoding rather than ascii
4512 if sys
.version_info
[0] < 3:
4513 err_str
= err_str
.decode(preferredencoding())
4517 def mimetype2ext(mt
):
4523 # Per RFC 3003, audio/mpeg can be .mp1, .mp2 or .mp3. Here use .mp3 as
4524 # it's the most popular one
4525 'audio/mpeg': 'mp3',
4526 'audio/x-wav': 'wav',
4531 _
, _
, res
= mt
.rpartition('/')
4532 res
= res
.split(';')[0].strip().lower()
4536 'smptett+xml': 'tt',
4540 'x-mp4-fragmented': 'mp4',
4541 'x-ms-sami': 'sami',
4544 'x-mpegurl': 'm3u8',
4545 'vnd.apple.mpegurl': 'm3u8',
4549 'vnd.ms-sstr+xml': 'ism',
4556 def parse_codecs(codecs_str
):
4557 # http://tools.ietf.org/html/rfc6381
4560 split_codecs
= list(filter(None, map(
4561 str.strip
, codecs_str
.strip().strip(',').split(','))))
4562 vcodec
, acodec
= None, None
4563 for full_codec
in split_codecs
:
4564 codec
= full_codec
.split('.')[0]
4565 if codec
in ('avc1', 'avc2', 'avc3', 'avc4', 'vp9', 'vp8', 'hev1', 'hev2', 'h263', 'h264', 'mp4v', 'hvc1', 'av01', 'theora'):
4568 elif codec
in ('mp4a', 'opus', 'vorbis', 'mp3', 'aac', 'ac-3', 'ec-3', 'eac3', 'dtsc', 'dtse', 'dtsh', 'dtsl'):
4572 write_string('WARNING: Unknown codec %s\n' % full_codec
, sys
.stderr
)
4573 if not vcodec
and not acodec
:
4574 if len(split_codecs
) == 2:
4576 'vcodec': split_codecs
[0],
4577 'acodec': split_codecs
[1],
4581 'vcodec': vcodec
or 'none',
4582 'acodec': acodec
or 'none',
4587 def urlhandle_detect_ext(url_handle
):
4588 getheader
= url_handle
.headers
.get
4590 cd
= getheader('Content-Disposition')
4592 m
= re
.match(r
'attachment;\s*filename="(?P<filename>[^"]+)"', cd
)
4594 e
= determine_ext(m
.group('filename'), default_ext
=None)
4598 return mimetype2ext(getheader('Content-Type'))
4601 def encode_data_uri(data
, mime_type
):
4602 return 'data:%s;base64,%s' % (mime_type
, base64
.b64encode(data
).decode('ascii'))
4605 def age_restricted(content_limit
, age_limit
):
4606 """ Returns True iff the content should be blocked """
4608 if age_limit
is None: # No limit set
4610 if content_limit
is None:
4611 return False # Content available for everyone
4612 return age_limit
< content_limit
4615 def is_html(first_bytes
):
4616 """ Detect whether a file contains HTML by examining its first bytes. """
4619 (b
'\xef\xbb\xbf', 'utf-8'),
4620 (b
'\x00\x00\xfe\xff', 'utf-32-be'),
4621 (b
'\xff\xfe\x00\x00', 'utf-32-le'),
4622 (b
'\xff\xfe', 'utf-16-le'),
4623 (b
'\xfe\xff', 'utf-16-be'),
4625 for bom
, enc
in BOMS
:
4626 if first_bytes
.startswith(bom
):
4627 s
= first_bytes
[len(bom
):].decode(enc
, 'replace')
4630 s
= first_bytes
.decode('utf-8', 'replace')
4632 return re
.match(r
'^\s*<', s
)
4635 def determine_protocol(info_dict
):
4636 protocol
= info_dict
.get('protocol')
4637 if protocol
is not None:
4640 url
= info_dict
['url']
4641 if url
.startswith('rtmp'):
4643 elif url
.startswith('mms'):
4645 elif url
.startswith('rtsp'):
4648 ext
= determine_ext(url
)
4654 return compat_urllib_parse_urlparse(url
).scheme
4657 def render_table(header_row
, data
, delim
=False, extraGap
=0, hideEmpty
=False):
4658 """ Render a list of rows, each as a list of values """
4660 def get_max_lens(table
):
4661 return [max(len(compat_str(v
)) for v
in col
) for col
in zip(*table
)]
4663 def filter_using_list(row
, filterArray
):
4664 return [col
for (take
, col
) in zip(filterArray
, row
) if take
]
4667 max_lens
= get_max_lens(data
)
4668 header_row
= filter_using_list(header_row
, max_lens
)
4669 data
= [filter_using_list(row
, max_lens
) for row
in data
]
4671 table
= [header_row
] + data
4672 max_lens
= get_max_lens(table
)
4674 table
= [header_row
] + [['-' * ml
for ml
in max_lens
]] + data
4675 format_str
= ' '.join('%-' + compat_str(ml
+ extraGap
) + 's' for ml
in max_lens
[:-1]) + ' %s'
4676 return '\n'.join(format_str
% tuple(row
) for row
in table
)
4679 def _match_one(filter_part
, dct
, incomplete
):
4680 # TODO: Generalize code with YoutubeDL._build_format_filter
4681 STRING_OPERATORS
= {
4682 '*=': operator
.contains
,
4683 '^=': lambda attr
, value
: attr
.startswith(value
),
4684 '$=': lambda attr
, value
: attr
.endswith(value
),
4685 '~=': lambda attr
, value
: re
.search(value
, attr
),
4687 COMPARISON_OPERATORS
= {
4689 '<=': operator
.le
, # "<=" must be defined above "<"
4696 operator_rex
= re
.compile(r
'''(?x)\s*
4698 \s*(?P<negation>!\s*)?(?P<op>%s)(?P<none_inclusive>\s*\?)?\s*
4700 (?P<intval>[0-9.]+(?:[kKmMgGtTpPeEzZyY]i?[Bb]?)?)|
4701 (?P<quote>["\'])(?P
<quotedstrval
>.+?
)(?P
=quote
)|
4705 ''' % '|'.join(map(re.escape, COMPARISON_OPERATORS.keys())))
4706 m = operator_rex.search(filter_part)
4708 unnegated_op = COMPARISON_OPERATORS[m.group('op')]
4709 if m.group('negation'):
4710 op = lambda attr, value: not unnegated_op(attr, value)
4713 actual_value = dct.get(m.group('key'))
4714 if (m.group('quotedstrval') is not None
4715 or m.group('strval') is not None
4716 # If the original field is a string and matching comparisonvalue is
4717 # a number we should respect the origin of the original field
4718 # and process comparison value as a string (see
4719 # https://github.com/ytdl-org/youtube-dl/issues/11082).
4720 or actual_value is not None and m.group('intval') is not None
4721 and isinstance(actual_value, compat_str)):
4722 comparison_value = m.group('quotedstrval') or m.group('strval') or m.group('intval')
4723 quote = m.group('quote')
4724 if quote is not None:
4725 comparison_value = comparison_value.replace(r'\%s' % quote, quote)
4727 if m.group('op') in STRING_OPERATORS:
4728 raise ValueError('Operator %s only supports string values!' % m.group('op'))
4730 comparison_value = int(m.group('intval'))
4732 comparison_value = parse_filesize(m.group('intval'))
4733 if comparison_value is None:
4734 comparison_value = parse_filesize(m.group('intval') + 'B')
4735 if comparison_value is None:
4737 'Invalid integer value %r in filter part %r' % (
4738 m.group('intval'), filter_part))
4739 if actual_value is None:
4740 return incomplete or m.group('none_inclusive')
4741 return op(actual_value, comparison_value)
4744 '': lambda v: (v is True) if isinstance(v, bool) else (v is not None),
4745 '!': lambda v: (v is False) if isinstance(v, bool) else (v is None),
4747 operator_rex = re.compile(r'''(?x
)\s
*
4748 (?P
<op
>%s)\s
*(?P
<key
>[a
-z_
]+)
4750 ''' % '|'.join(map(re.escape, UNARY_OPERATORS.keys())))
4751 m = operator_rex.search(filter_part)
4753 op = UNARY_OPERATORS[m.group('op')]
4754 actual_value = dct.get(m.group('key'))
4755 if incomplete and actual_value is None:
4757 return op(actual_value)
4759 raise ValueError('Invalid filter part %r' % filter_part)
4762 def match_str(filter_str, dct, incomplete=False):
4763 """ Filter a dictionary with a simple string syntax. Returns True (=passes filter) or false
4764 When incomplete, all conditions passes on missing fields
4767 _match_one(filter_part.replace(r'\&', '&'), dct, incomplete)
4768 for filter_part in re.split(r'(?<!\\)&', filter_str))
4771 def match_filter_func(filter_str):
4772 def _match_func(info_dict, *args, **kwargs):
4773 if match_str(filter_str, info_dict, *args, **kwargs):
4776 video_title = info_dict.get('title', info_dict.get('id', 'video'))
4777 return '%s does not pass filter %s, skipping ..' % (video_title, filter_str)
4781 def parse_dfxp_time_expr(time_expr):
4785 mobj = re.match(r'^(?P<time_offset>\d+(?:\.\d+)?)s?$', time_expr)
4787 return float(mobj.group('time_offset'))
4789 mobj = re.match(r'^(\d+):(\d\d):(\d\d(?:(?:\.|:)\d+)?)$', time_expr)
4791 return 3600 * int(mobj.group(1)) + 60 * int(mobj.group(2)) + float(mobj.group(3).replace(':', '.'))
4794 def srt_subtitles_timecode(seconds):
4795 return '%02d:%02d:%02d,%03d' % (seconds / 3600, (seconds % 3600) / 60, seconds % 60, (seconds % 1) * 1000)
4798 def dfxp2srt(dfxp_data):
4800 @param dfxp_data A
bytes-like
object containing DFXP data
4801 @returns A
unicode object containing converted SRT data
4803 LEGACY_NAMESPACES = (
4804 (b'http://www.w3.org/ns/ttml', [
4805 b'http://www.w3.org/2004/11/ttaf1',
4806 b'http://www.w3.org/2006/04/ttaf1',
4807 b'http://www.w3.org/2006/10/ttaf1',
4809 (b'http://www.w3.org/ns/ttml#styling', [
4810 b'http://www.w3.org/ns/ttml#style',
4814 SUPPORTED_STYLING = [
4823 _x = functools.partial(xpath_with_ns, ns_map={
4824 'xml': 'http://www.w3.org/XML/1998/namespace',
4825 'ttml': 'http://www.w3.org/ns/ttml',
4826 'tts': 'http://www.w3.org/ns/ttml#styling',
4832 class TTMLPElementParser(object):
4834 _unclosed_elements = []
4835 _applied_styles = []
4837 def start(self, tag, attrib):
4838 if tag in (_x('ttml:br'), 'br'):
4841 unclosed_elements = []
4843 element_style_id = attrib.get('style')
4845 style.update(default_style)
4846 if element_style_id:
4847 style.update(styles.get(element_style_id, {}))
4848 for prop in SUPPORTED_STYLING:
4849 prop_val = attrib.get(_x('tts:' + prop))
4851 style[prop] = prop_val
4854 for k, v in sorted(style.items()):
4855 if self._applied_styles and self._applied_styles[-1].get(k) == v:
4858 font += ' color="%s"' % v
4859 elif k == 'fontSize':
4860 font += ' size="%s"' % v
4861 elif k == 'fontFamily':
4862 font += ' face="%s"' % v
4863 elif k == 'fontWeight' and v == 'bold':
4865 unclosed_elements.append('b')
4866 elif k == 'fontStyle' and v == 'italic':
4868 unclosed_elements.append('i')
4869 elif k == 'textDecoration' and v == 'underline':
4871 unclosed_elements.append('u')
4873 self._out += '<font' + font + '>'
4874 unclosed_elements.append('font')
4876 if self._applied_styles:
4877 applied_style.update(self._applied_styles[-1])
4878 applied_style.update(style)
4879 self._applied_styles.append(applied_style)
4880 self._unclosed_elements.append(unclosed_elements)
4883 if tag not in (_x('ttml:br'), 'br'):
4884 unclosed_elements = self._unclosed_elements.pop()
4885 for element in reversed(unclosed_elements):
4886 self._out += '</%s>' % element
4887 if unclosed_elements and self._applied_styles:
4888 self._applied_styles.pop()
4890 def data(self, data):
4894 return self._out.strip()
4896 def parse_node(node):
4897 target = TTMLPElementParser()
4898 parser = xml.etree.ElementTree.XMLParser(target=target)
4899 parser.feed(xml.etree.ElementTree.tostring(node))
4900 return parser.close()
4902 for k, v in LEGACY_NAMESPACES:
4904 dfxp_data = dfxp_data.replace(ns, k)
4906 dfxp = compat_etree_fromstring(dfxp_data)
4908 paras = dfxp.findall(_x('.//ttml:p')) or dfxp.findall('.//p')
4911 raise ValueError('Invalid dfxp/TTML subtitle')
4915 for style in dfxp.findall(_x('.//ttml:style')):
4916 style_id = style.get('id') or style.get(_x('xml:id'))
4919 parent_style_id = style.get('style')
4921 if parent_style_id not in styles:
4924 styles[style_id] = styles[parent_style_id].copy()
4925 for prop in SUPPORTED_STYLING:
4926 prop_val = style.get(_x('tts:' + prop))
4928 styles.setdefault(style_id, {})[prop] = prop_val
4934 for p in ('body', 'div'):
4935 ele = xpath_element(dfxp, [_x('.//ttml:' + p), './/' + p])
4938 style = styles.get(ele.get('style'))
4941 default_style.update(style)
4943 for para, index in zip(paras, itertools.count(1)):
4944 begin_time = parse_dfxp_time_expr(para.attrib.get('begin'))
4945 end_time = parse_dfxp_time_expr(para.attrib.get('end'))
4946 dur = parse_dfxp_time_expr(para.attrib.get('dur'))
4947 if begin_time is None:
4952 end_time = begin_time + dur
4953 out.append('%d\n%s --> %s\n%s\n\n' % (
4955 srt_subtitles_timecode(begin_time),
4956 srt_subtitles_timecode(end_time),
4962 def cli_option(params, command_option, param):
4963 param = params.get(param)
4965 param = compat_str(param)
4966 return [command_option, param] if param is not None else []
4969 def cli_bool_option(params, command_option, param, true_value='true', false_value='false', separator=None):
4970 param = params.get(param)
4973 assert isinstance(param, bool)
4975 return [command_option + separator + (true_value if param else false_value)]
4976 return [command_option, true_value if param else false_value]
4979 def cli_valueless_option(params, command_option, param, expected_value=True):
4980 param = params.get(param)
4981 return [command_option] if param == expected_value else []
4984 def cli_configuration_args(argdict, keys, default=[], use_compat=True):
4985 if isinstance(argdict, (list, tuple)): # for backward compatibility
4992 assert isinstance(argdict, dict)
4994 assert isinstance(keys, (list, tuple))
4995 for key_list in keys:
4996 arg_list = list(filter(
4997 lambda x: x is not None,
4998 [argdict.get(key.lower()) for key in variadic(key_list)]))
5000 return [arg for args in arg_list for arg in args]
5004 def _configuration_args(main_key, argdict, exe, keys=None, default=[], use_compat=True):
5005 main_key, exe = main_key.lower(), exe.lower()
5006 root_key = exe if main_key == exe else f'{main_key}+{exe}'
5007 keys = [f'{root_key}{k}' for k in (keys or [''])]
5008 if root_key in keys:
5010 keys.append((main_key, exe))
5011 keys.append('default')
5014 return cli_configuration_args(argdict, keys, default, use_compat)
5017 class ISO639Utils(object):
5018 # See http://www.loc.gov/standards/iso639-2/ISO-639-2_utf-8.txt
5077 'iw': 'heb', # Replaced by he in 1989 revision
5087 'in': 'ind', # Replaced by id in 1989 revision
5202 'ji': 'yid', # Replaced by yi in 1989 revision
5210 def short2long(cls, code):
5211 """Convert language code from ISO 639-1 to ISO 639-2/T"""
5212 return cls._lang_map.get(code[:2])
5215 def long2short(cls, code):
5216 """Convert language code from ISO 639-2/T to ISO 639-1"""
5217 for short_name, long_name in cls._lang_map.items():
5218 if long_name == code:
5222 class ISO3166Utils(object):
5223 # From http://data.okfn.org/data/core/country-list
5225 'AF': 'Afghanistan',
5226 'AX': 'Åland Islands',
5229 'AS': 'American Samoa',
5234 'AG': 'Antigua and Barbuda',
5251 'BO': 'Bolivia, Plurinational State of',
5252 'BQ': 'Bonaire, Sint Eustatius and Saba',
5253 'BA': 'Bosnia and Herzegovina',
5255 'BV': 'Bouvet Island',
5257 'IO': 'British Indian Ocean Territory',
5258 'BN': 'Brunei Darussalam',
5260 'BF': 'Burkina Faso',
5266 'KY': 'Cayman Islands',
5267 'CF': 'Central African Republic',
5271 'CX': 'Christmas Island',
5272 'CC': 'Cocos (Keeling) Islands',
5276 'CD': 'Congo, the Democratic Republic of the',
5277 'CK': 'Cook Islands',
5279 'CI': 'Côte d\'Ivoire',
5284 'CZ': 'Czech Republic',
5288 'DO': 'Dominican Republic',
5291 'SV': 'El Salvador',
5292 'GQ': 'Equatorial Guinea',
5296 'FK': 'Falkland Islands (Malvinas)',
5297 'FO': 'Faroe Islands',
5301 'GF': 'French Guiana',
5302 'PF': 'French Polynesia',
5303 'TF': 'French Southern Territories',
5318 'GW': 'Guinea-Bissau',
5321 'HM': 'Heard Island and McDonald Islands',
5322 'VA': 'Holy See (Vatican City State)',
5329 'IR': 'Iran, Islamic Republic of',
5332 'IM': 'Isle of Man',
5342 'KP': 'Korea, Democratic People\'s Republic of',
5343 'KR': 'Korea, Republic of',
5346 'LA': 'Lao People\'s Democratic Republic',
5352 'LI': 'Liechtenstein',
5356 'MK': 'Macedonia, the Former Yugoslav Republic of',
5363 'MH': 'Marshall Islands',
5369 'FM': 'Micronesia, Federated States of',
5370 'MD': 'Moldova, Republic of',
5381 'NL': 'Netherlands',
5382 'NC': 'New Caledonia',
5383 'NZ': 'New Zealand',
5388 'NF': 'Norfolk Island',
5389 'MP': 'Northern Mariana Islands',
5394 'PS': 'Palestine, State of',
5396 'PG': 'Papua New Guinea',
5399 'PH': 'Philippines',
5403 'PR': 'Puerto Rico',
5407 'RU': 'Russian Federation',
5409 'BL': 'Saint Barthélemy',
5410 'SH': 'Saint Helena, Ascension and Tristan da Cunha',
5411 'KN': 'Saint Kitts and Nevis',
5412 'LC': 'Saint Lucia',
5413 'MF': 'Saint Martin (French part)',
5414 'PM': 'Saint Pierre and Miquelon',
5415 'VC': 'Saint Vincent and the Grenadines',
5418 'ST': 'Sao Tome and Principe',
5419 'SA': 'Saudi Arabia',
5423 'SL': 'Sierra Leone',
5425 'SX': 'Sint Maarten (Dutch part)',
5428 'SB': 'Solomon Islands',
5430 'ZA': 'South Africa',
5431 'GS': 'South Georgia and the South Sandwich Islands',
5432 'SS': 'South Sudan',
5437 'SJ': 'Svalbard and Jan Mayen',
5440 'CH': 'Switzerland',
5441 'SY': 'Syrian Arab Republic',
5442 'TW': 'Taiwan, Province of China',
5444 'TZ': 'Tanzania, United Republic of',
5446 'TL': 'Timor-Leste',
5450 'TT': 'Trinidad and Tobago',
5453 'TM': 'Turkmenistan',
5454 'TC': 'Turks and Caicos Islands',
5458 'AE': 'United Arab Emirates',
5459 'GB': 'United Kingdom',
5460 'US': 'United States',
5461 'UM': 'United States Minor Outlying Islands',
5465 'VE': 'Venezuela, Bolivarian Republic of',
5467 'VG': 'Virgin Islands, British',
5468 'VI': 'Virgin Islands, U.S.',
5469 'WF': 'Wallis and Futuna',
5470 'EH': 'Western Sahara',
5477 def short2full(cls, code):
5478 """Convert an ISO 3166-2 country code to the corresponding full name"""
5479 return cls._country_map.get(code.upper())
5482 class GeoUtils(object):
5483 # Major IPv4 address blocks per country
5485 'AD': '46.172.224.0/19',
5486 'AE': '94.200.0.0/13',
5487 'AF': '149.54.0.0/17',
5488 'AG': '209.59.64.0/18',
5489 'AI': '204.14.248.0/21',
5490 'AL': '46.99.0.0/16',
5491 'AM': '46.70.0.0/15',
5492 'AO': '105.168.0.0/13',
5493 'AP': '182.50.184.0/21',
5494 'AQ': '23.154.160.0/24',
5495 'AR': '181.0.0.0/12',
5496 'AS': '202.70.112.0/20',
5497 'AT': '77.116.0.0/14',
5498 'AU': '1.128.0.0/11',
5499 'AW': '181.41.0.0/18',
5500 'AX': '185.217.4.0/22',
5501 'AZ': '5.197.0.0/16',
5502 'BA': '31.176.128.0/17',
5503 'BB': '65.48.128.0/17',
5504 'BD': '114.130.0.0/16',
5506 'BF': '102.178.0.0/15',
5507 'BG': '95.42.0.0/15',
5508 'BH': '37.131.0.0/17',
5509 'BI': '154.117.192.0/18',
5510 'BJ': '137.255.0.0/16',
5511 'BL': '185.212.72.0/23',
5512 'BM': '196.12.64.0/18',
5513 'BN': '156.31.0.0/16',
5514 'BO': '161.56.0.0/16',
5515 'BQ': '161.0.80.0/20',
5516 'BR': '191.128.0.0/12',
5517 'BS': '24.51.64.0/18',
5518 'BT': '119.2.96.0/19',
5519 'BW': '168.167.0.0/16',
5520 'BY': '178.120.0.0/13',
5521 'BZ': '179.42.192.0/18',
5522 'CA': '99.224.0.0/11',
5523 'CD': '41.243.0.0/16',
5524 'CF': '197.242.176.0/21',
5525 'CG': '160.113.0.0/16',
5526 'CH': '85.0.0.0/13',
5527 'CI': '102.136.0.0/14',
5528 'CK': '202.65.32.0/19',
5529 'CL': '152.172.0.0/14',
5530 'CM': '102.244.0.0/14',
5531 'CN': '36.128.0.0/10',
5532 'CO': '181.240.0.0/12',
5533 'CR': '201.192.0.0/12',
5534 'CU': '152.206.0.0/15',
5535 'CV': '165.90.96.0/19',
5536 'CW': '190.88.128.0/17',
5537 'CY': '31.153.0.0/16',
5538 'CZ': '88.100.0.0/14',
5540 'DJ': '197.241.0.0/17',
5541 'DK': '87.48.0.0/12',
5542 'DM': '192.243.48.0/20',
5543 'DO': '152.166.0.0/15',
5544 'DZ': '41.96.0.0/12',
5545 'EC': '186.68.0.0/15',
5546 'EE': '90.190.0.0/15',
5547 'EG': '156.160.0.0/11',
5548 'ER': '196.200.96.0/20',
5549 'ES': '88.0.0.0/11',
5550 'ET': '196.188.0.0/14',
5551 'EU': '2.16.0.0/13',
5552 'FI': '91.152.0.0/13',
5553 'FJ': '144.120.0.0/16',
5554 'FK': '80.73.208.0/21',
5555 'FM': '119.252.112.0/20',
5556 'FO': '88.85.32.0/19',
5558 'GA': '41.158.0.0/15',
5560 'GD': '74.122.88.0/21',
5561 'GE': '31.146.0.0/16',
5562 'GF': '161.22.64.0/18',
5563 'GG': '62.68.160.0/19',
5564 'GH': '154.160.0.0/12',
5565 'GI': '95.164.0.0/16',
5566 'GL': '88.83.0.0/19',
5567 'GM': '160.182.0.0/15',
5568 'GN': '197.149.192.0/18',
5569 'GP': '104.250.0.0/19',
5570 'GQ': '105.235.224.0/20',
5571 'GR': '94.64.0.0/13',
5572 'GT': '168.234.0.0/16',
5573 'GU': '168.123.0.0/16',
5574 'GW': '197.214.80.0/20',
5575 'GY': '181.41.64.0/18',
5576 'HK': '113.252.0.0/14',
5577 'HN': '181.210.0.0/16',
5578 'HR': '93.136.0.0/13',
5579 'HT': '148.102.128.0/17',
5580 'HU': '84.0.0.0/14',
5581 'ID': '39.192.0.0/10',
5582 'IE': '87.32.0.0/12',
5583 'IL': '79.176.0.0/13',
5584 'IM': '5.62.80.0/20',
5585 'IN': '117.192.0.0/10',
5586 'IO': '203.83.48.0/21',
5587 'IQ': '37.236.0.0/14',
5588 'IR': '2.176.0.0/12',
5589 'IS': '82.221.0.0/16',
5590 'IT': '79.0.0.0/10',
5591 'JE': '87.244.64.0/18',
5592 'JM': '72.27.0.0/17',
5593 'JO': '176.29.0.0/16',
5594 'JP': '133.0.0.0/8',
5595 'KE': '105.48.0.0/12',
5596 'KG': '158.181.128.0/17',
5597 'KH': '36.37.128.0/17',
5598 'KI': '103.25.140.0/22',
5599 'KM': '197.255.224.0/20',
5600 'KN': '198.167.192.0/19',
5601 'KP': '175.45.176.0/22',
5602 'KR': '175.192.0.0/10',
5603 'KW': '37.36.0.0/14',
5604 'KY': '64.96.0.0/15',
5605 'KZ': '2.72.0.0/13',
5606 'LA': '115.84.64.0/18',
5607 'LB': '178.135.0.0/16',
5608 'LC': '24.92.144.0/20',
5609 'LI': '82.117.0.0/19',
5610 'LK': '112.134.0.0/15',
5611 'LR': '102.183.0.0/16',
5612 'LS': '129.232.0.0/17',
5613 'LT': '78.56.0.0/13',
5614 'LU': '188.42.0.0/16',
5615 'LV': '46.109.0.0/16',
5616 'LY': '41.252.0.0/14',
5617 'MA': '105.128.0.0/11',
5618 'MC': '88.209.64.0/18',
5619 'MD': '37.246.0.0/16',
5620 'ME': '178.175.0.0/17',
5621 'MF': '74.112.232.0/21',
5622 'MG': '154.126.0.0/17',
5623 'MH': '117.103.88.0/21',
5624 'MK': '77.28.0.0/15',
5625 'ML': '154.118.128.0/18',
5626 'MM': '37.111.0.0/17',
5627 'MN': '49.0.128.0/17',
5628 'MO': '60.246.0.0/16',
5629 'MP': '202.88.64.0/20',
5630 'MQ': '109.203.224.0/19',
5631 'MR': '41.188.64.0/18',
5632 'MS': '208.90.112.0/22',
5633 'MT': '46.11.0.0/16',
5634 'MU': '105.16.0.0/12',
5635 'MV': '27.114.128.0/18',
5636 'MW': '102.70.0.0/15',
5637 'MX': '187.192.0.0/11',
5638 'MY': '175.136.0.0/13',
5639 'MZ': '197.218.0.0/15',
5640 'NA': '41.182.0.0/16',
5641 'NC': '101.101.0.0/18',
5642 'NE': '197.214.0.0/18',
5643 'NF': '203.17.240.0/22',
5644 'NG': '105.112.0.0/12',
5645 'NI': '186.76.0.0/15',
5646 'NL': '145.96.0.0/11',
5647 'NO': '84.208.0.0/13',
5648 'NP': '36.252.0.0/15',
5649 'NR': '203.98.224.0/19',
5650 'NU': '49.156.48.0/22',
5651 'NZ': '49.224.0.0/14',
5652 'OM': '5.36.0.0/15',
5653 'PA': '186.72.0.0/15',
5654 'PE': '186.160.0.0/14',
5655 'PF': '123.50.64.0/18',
5656 'PG': '124.240.192.0/19',
5657 'PH': '49.144.0.0/13',
5658 'PK': '39.32.0.0/11',
5659 'PL': '83.0.0.0/11',
5660 'PM': '70.36.0.0/20',
5661 'PR': '66.50.0.0/16',
5662 'PS': '188.161.0.0/16',
5663 'PT': '85.240.0.0/13',
5664 'PW': '202.124.224.0/20',
5665 'PY': '181.120.0.0/14',
5666 'QA': '37.210.0.0/15',
5667 'RE': '102.35.0.0/16',
5668 'RO': '79.112.0.0/13',
5669 'RS': '93.86.0.0/15',
5670 'RU': '5.136.0.0/13',
5671 'RW': '41.186.0.0/16',
5672 'SA': '188.48.0.0/13',
5673 'SB': '202.1.160.0/19',
5674 'SC': '154.192.0.0/11',
5675 'SD': '102.120.0.0/13',
5676 'SE': '78.64.0.0/12',
5677 'SG': '8.128.0.0/10',
5678 'SI': '188.196.0.0/14',
5679 'SK': '78.98.0.0/15',
5680 'SL': '102.143.0.0/17',
5681 'SM': '89.186.32.0/19',
5682 'SN': '41.82.0.0/15',
5683 'SO': '154.115.192.0/18',
5684 'SR': '186.179.128.0/17',
5685 'SS': '105.235.208.0/21',
5686 'ST': '197.159.160.0/19',
5687 'SV': '168.243.0.0/16',
5688 'SX': '190.102.0.0/20',
5690 'SZ': '41.84.224.0/19',
5691 'TC': '65.255.48.0/20',
5692 'TD': '154.68.128.0/19',
5693 'TG': '196.168.0.0/14',
5694 'TH': '171.96.0.0/13',
5695 'TJ': '85.9.128.0/18',
5696 'TK': '27.96.24.0/21',
5697 'TL': '180.189.160.0/20',
5698 'TM': '95.85.96.0/19',
5699 'TN': '197.0.0.0/11',
5700 'TO': '175.176.144.0/21',
5701 'TR': '78.160.0.0/11',
5702 'TT': '186.44.0.0/15',
5703 'TV': '202.2.96.0/19',
5704 'TW': '120.96.0.0/11',
5705 'TZ': '156.156.0.0/14',
5706 'UA': '37.52.0.0/14',
5707 'UG': '102.80.0.0/13',
5709 'UY': '167.56.0.0/13',
5710 'UZ': '84.54.64.0/18',
5711 'VA': '212.77.0.0/19',
5712 'VC': '207.191.240.0/21',
5713 'VE': '186.88.0.0/13',
5714 'VG': '66.81.192.0/20',
5715 'VI': '146.226.0.0/16',
5716 'VN': '14.160.0.0/11',
5717 'VU': '202.80.32.0/20',
5718 'WF': '117.20.32.0/21',
5719 'WS': '202.4.32.0/19',
5720 'YE': '134.35.0.0/16',
5721 'YT': '41.242.116.0/22',
5722 'ZA': '41.0.0.0/11',
5723 'ZM': '102.144.0.0/13',
5724 'ZW': '102.177.192.0/18',
5728 def random_ipv4(cls, code_or_block):
5729 if len(code_or_block) == 2:
5730 block = cls._country_ip_map.get(code_or_block.upper())
5734 block = code_or_block
5735 addr, preflen = block.split('/')
5736 addr_min = compat_struct_unpack('!L', socket.inet_aton(addr))[0]
5737 addr_max = addr_min | (0xffffffff >> int(preflen))
5738 return compat_str(socket.inet_ntoa(
5739 compat_struct_pack('!L', random.randint(addr_min, addr_max))))
5742 class PerRequestProxyHandler(compat_urllib_request.ProxyHandler):
5743 def __init__(self, proxies=None):
5744 # Set default handlers
5745 for type in ('http', 'https'):
5746 setattr(self, '%s_open' % type,
5747 lambda r, proxy='__noproxy__', type=type, meth=self.proxy_open:
5748 meth(r, proxy, type))
5749 compat_urllib_request.ProxyHandler.__init__(self, proxies)
5751 def proxy_open(self, req, proxy, type):
5752 req_proxy = req.headers.get('Ytdl-request-proxy')
5753 if req_proxy is not None:
5755 del req.headers['Ytdl-request-proxy']
5757 if proxy == '__noproxy__':
5758 return None # No Proxy
5759 if compat_urlparse.urlparse(proxy).scheme.lower() in ('socks', 'socks4', 'socks4a', 'socks5'):
5760 req.add_header('Ytdl-socks-proxy', proxy)
5761 # yt-dlp's http/https handlers do wrapping the socket with socks
5763 return compat_urllib_request.ProxyHandler.proxy_open(
5764 self, req, proxy, type)
5767 # Both long_to_bytes and bytes_to_long are adapted from PyCrypto, which is
5768 # released into Public Domain
5769 # https://github.com/dlitz/pycrypto/blob/master/lib/Crypto/Util/number.py#L387
5771 def long_to_bytes(n, blocksize=0):
5772 """long_to_bytes(n:long, blocksize:int) : string
5773 Convert a long integer to a byte string.
5775 If optional blocksize is given and greater than zero, pad the front of the
5776 byte string with binary zeros so that the length is a multiple of
5779 # after much testing, this algorithm was deemed to be the fastest
5783 s = compat_struct_pack('>I', n & 0xffffffff) + s
5785 # strip off leading zeros
5786 for i in range(len(s)):
5787 if s[i] != b'\000'[0]:
5790 # only happens when n == 0
5794 # add back some pad bytes. this could be done more efficiently w.r.t. the
5795 # de-padding being done above, but sigh...
5796 if blocksize > 0 and len(s) % blocksize:
5797 s = (blocksize - len(s) % blocksize) * b'\000' + s
5801 def bytes_to_long(s):
5802 """bytes_to_long(string) : long
5803 Convert a byte string to a long integer.
5805 This is (essentially) the inverse of long_to_bytes().
5810 extra = (4 - length % 4)
5811 s = b'\000' * extra + s
5812 length = length + extra
5813 for i in range(0, length, 4):
5814 acc = (acc << 32) + compat_struct_unpack('>I', s[i:i + 4])[0]
5818 def ohdave_rsa_encrypt(data, exponent, modulus):
5820 Implement OHDave
's RSA algorithm. See http://www.ohdave.com/rsa/
5823 data: data to encrypt, bytes-like object
5824 exponent, modulus: parameter e and N of RSA algorithm, both integer
5825 Output: hex string of encrypted data
5827 Limitation: supports one block encryption only
5830 payload = int(binascii.hexlify(data[::-1]), 16)
5831 encrypted = pow(payload, exponent, modulus)
5832 return '%x' % encrypted
5835 def pkcs1pad(data, length):
5837 Padding input data with PKCS#1 scheme
5839 @param {int[]} data input data
5840 @param {int} length target length
5841 @returns {int[]} padded data
5843 if len(data) > length - 11:
5844 raise ValueError('Input data too
long for PKCS
#1 padding')
5846 pseudo_random
= [random
.randint(0, 254) for _
in range(length
- len(data
) - 3)]
5847 return [0, 2] + pseudo_random
+ [0] + data
5850 def encode_base_n(num
, n
, table
=None):
5851 FULL_TABLE
= '0123456789abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ'
5853 table
= FULL_TABLE
[:n
]
5856 raise ValueError('base %d exceeds table length %d' % (n
, len(table
)))
5863 ret
= table
[num
% n
] + ret
5868 def decode_packed_codes(code
):
5869 mobj
= re
.search(PACKED_CODES_RE
, code
)
5870 obfuscated_code
, base
, count
, symbols
= mobj
.groups()
5873 symbols
= symbols
.split('|')
5878 base_n_count
= encode_base_n(count
, base
)
5879 symbol_table
[base_n_count
] = symbols
[count
] or base_n_count
5882 r
'\b(\w+)\b', lambda mobj
: symbol_table
[mobj
.group(0)],
5886 def caesar(s
, alphabet
, shift
):
5891 alphabet
[(alphabet
.index(c
) + shift
) % l
] if c
in alphabet
else c
5896 return caesar(s
, r
'''!"#$%&'()*+,-./0123456789:;<=>?@ABCDEFGHIJKLMNOPQRSTUVWXYZ[\]^_`abcdefghijklmnopqrstuvwxyz{|}~''', 47)
5899 def parse_m3u8_attributes(attrib
):
5901 for (key
, val
) in re
.findall(r
'(?P<key>[A-Z0-9-]+)=(?P<val>"[^"]+"|[^",]+)(?:,|$)', attrib
):
5902 if val
.startswith('"'):
5908 def urshift(val
, n
):
5909 return val
>> n
if val
>= 0 else (val
+ 0x100000000) >> n
5912 # Based on png2str() written by @gdkchan and improved by @yokrysty
5913 # Originally posted at https://github.com/ytdl-org/youtube-dl/issues/9706
5914 def decode_png(png_data
):
5915 # Reference: https://www.w3.org/TR/PNG/
5916 header
= png_data
[8:]
5918 if png_data
[:8] != b
'\x89PNG\x0d\x0a\x1a\x0a' or header
[4:8] != b
'IHDR':
5919 raise IOError('Not a valid PNG file.')
5921 int_map
= {1: '>B', 2: '>H', 4: '>I'}
5922 unpack_integer
= lambda x
: compat_struct_unpack(int_map
[len(x
)], x
)[0]
5927 length
= unpack_integer(header
[:4])
5930 chunk_type
= header
[:4]
5933 chunk_data
= header
[:length
]
5934 header
= header
[length
:]
5936 header
= header
[4:] # Skip CRC
5944 ihdr
= chunks
[0]['data']
5946 width
= unpack_integer(ihdr
[:4])
5947 height
= unpack_integer(ihdr
[4:8])
5951 for chunk
in chunks
:
5952 if chunk
['type'] == b
'IDAT':
5953 idat
+= chunk
['data']
5956 raise IOError('Unable to read PNG data.')
5958 decompressed_data
= bytearray(zlib
.decompress(idat
))
5963 def _get_pixel(idx
):
5968 for y
in range(height
):
5969 basePos
= y
* (1 + stride
)
5970 filter_type
= decompressed_data
[basePos
]
5974 pixels
.append(current_row
)
5976 for x
in range(stride
):
5977 color
= decompressed_data
[1 + basePos
+ x
]
5978 basex
= y
* stride
+ x
5983 left
= _get_pixel(basex
- 3)
5985 up
= _get_pixel(basex
- stride
)
5987 if filter_type
== 1: # Sub
5988 color
= (color
+ left
) & 0xff
5989 elif filter_type
== 2: # Up
5990 color
= (color
+ up
) & 0xff
5991 elif filter_type
== 3: # Average
5992 color
= (color
+ ((left
+ up
) >> 1)) & 0xff
5993 elif filter_type
== 4: # Paeth
5999 c
= _get_pixel(basex
- stride
- 3)
6007 if pa
<= pb
and pa
<= pc
:
6008 color
= (color
+ a
) & 0xff
6010 color
= (color
+ b
) & 0xff
6012 color
= (color
+ c
) & 0xff
6014 current_row
.append(color
)
6016 return width
, height
, pixels
6019 def write_xattr(path
, key
, value
):
6020 # This mess below finds the best xattr tool for the job
6022 # try the pyxattr module...
6025 if hasattr(xattr
, 'set'): # pyxattr
6026 # Unicode arguments are not supported in python-pyxattr until
6028 # See https://github.com/ytdl-org/youtube-dl/issues/5498
6029 pyxattr_required_version
= '0.5.0'
6030 if version_tuple(xattr
.__version
__) < version_tuple(pyxattr_required_version
):
6031 # TODO: fallback to CLI tools
6032 raise XAttrUnavailableError(
6033 'python-pyxattr is detected but is too old. '
6034 'yt-dlp requires %s or above while your version is %s. '
6035 'Falling back to other xattr implementations' % (
6036 pyxattr_required_version
, xattr
.__version
__))
6038 setxattr
= xattr
.set
6040 setxattr
= xattr
.setxattr
6043 setxattr(path
, key
, value
)
6044 except EnvironmentError as e
:
6045 raise XAttrMetadataError(e
.errno
, e
.strerror
)
6048 if compat_os_name
== 'nt':
6049 # Write xattrs to NTFS Alternate Data Streams:
6050 # http://en.wikipedia.org/wiki/NTFS#Alternate_data_streams_.28ADS.29
6051 assert ':' not in key
6052 assert os
.path
.exists(path
)
6054 ads_fn
= path
+ ':' + key
6056 with open(ads_fn
, 'wb') as f
:
6058 except EnvironmentError as e
:
6059 raise XAttrMetadataError(e
.errno
, e
.strerror
)
6061 user_has_setfattr
= check_executable('setfattr', ['--version'])
6062 user_has_xattr
= check_executable('xattr', ['-h'])
6064 if user_has_setfattr
or user_has_xattr
:
6066 value
= value
.decode('utf-8')
6067 if user_has_setfattr
:
6068 executable
= 'setfattr'
6069 opts
= ['-n', key
, '-v', value
]
6070 elif user_has_xattr
:
6071 executable
= 'xattr'
6072 opts
= ['-w', key
, value
]
6074 cmd
= ([encodeFilename(executable
, True)]
6075 + [encodeArgument(o
) for o
in opts
]
6076 + [encodeFilename(path
, True)])
6079 p
= subprocess
.Popen(
6080 cmd
, stdout
=subprocess
.PIPE
, stderr
=subprocess
.PIPE
, stdin
=subprocess
.PIPE
)
6081 except EnvironmentError as e
:
6082 raise XAttrMetadataError(e
.errno
, e
.strerror
)
6083 stdout
, stderr
= process_communicate_or_kill(p
)
6084 stderr
= stderr
.decode('utf-8', 'replace')
6085 if p
.returncode
!= 0:
6086 raise XAttrMetadataError(p
.returncode
, stderr
)
6089 # On Unix, and can't find pyxattr, setfattr, or xattr.
6090 if sys
.platform
.startswith('linux'):
6091 raise XAttrUnavailableError(
6092 "Couldn't find a tool to set the xattrs. "
6093 "Install either the python 'pyxattr' or 'xattr' "
6094 "modules, or the GNU 'attr' package "
6095 "(which contains the 'setfattr' tool).")
6097 raise XAttrUnavailableError(
6098 "Couldn't find a tool to set the xattrs. "
6099 "Install either the python 'xattr' module, "
6100 "or the 'xattr' binary.")
6103 def random_birthday(year_field
, month_field
, day_field
):
6104 start_date
= datetime
.date(1950, 1, 1)
6105 end_date
= datetime
.date(1995, 12, 31)
6106 offset
= random
.randint(0, (end_date
- start_date
).days
)
6107 random_date
= start_date
+ datetime
.timedelta(offset
)
6109 year_field
: str(random_date
.year
),
6110 month_field
: str(random_date
.month
),
6111 day_field
: str(random_date
.day
),
6115 # Templates for internet shortcut files, which are plain text files.
6116 DOT_URL_LINK_TEMPLATE
= '''
6121 DOT_WEBLOC_LINK_TEMPLATE
= '''
6122 <?xml version="1.0" encoding="UTF-8"?>
6123 <!DOCTYPE plist PUBLIC "-//Apple//DTD PLIST 1.0//EN" "http://www.apple.com/DTDs/PropertyList-1.0.dtd">
6124 <plist version="1.0">
6127 \t<string>%(url)s</string>
6132 DOT_DESKTOP_LINK_TEMPLATE
= '''
6142 def iri_to_uri(iri
):
6144 Converts an IRI (Internationalized Resource Identifier, allowing Unicode characters) to a URI (Uniform Resource Identifier, ASCII-only).
6146 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.
6149 iri_parts
= compat_urllib_parse_urlparse(iri
)
6151 if '[' in iri_parts
.netloc
:
6152 raise ValueError('IPv6 URIs are not, yet, supported.')
6153 # Querying `.netloc`, when there's only one bracket, also raises a ValueError.
6155 # 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.
6158 if iri_parts
.username
:
6159 net_location
+= compat_urllib_parse_quote(iri_parts
.username
, safe
=r
"!$%&'()*+,~")
6160 if iri_parts
.password
is not None:
6161 net_location
+= ':' + compat_urllib_parse_quote(iri_parts
.password
, safe
=r
"!$%&'()*+,~")
6164 net_location
+= iri_parts
.hostname
.encode('idna').decode('utf-8') # Punycode for Unicode hostnames.
6165 # The 'idna' encoding produces ASCII text.
6166 if iri_parts
.port
is not None and iri_parts
.port
!= 80:
6167 net_location
+= ':' + str(iri_parts
.port
)
6169 return compat_urllib_parse_urlunparse(
6173 compat_urllib_parse_quote_plus(iri_parts
.path
, safe
=r
"!$%&'()*+,/:;=@|~"),
6175 # Unsure about the `safe` argument, since this is a legacy way of handling parameters.
6176 compat_urllib_parse_quote_plus(iri_parts
.params
, safe
=r
"!$%&'()*+,/:;=@|~"),
6178 # Not totally sure about the `safe` argument, since the source does not explicitly mention the query URI component.
6179 compat_urllib_parse_quote_plus(iri_parts
.query
, safe
=r
"!$%&'()*+,/:;=?@{|}~"),
6181 compat_urllib_parse_quote_plus(iri_parts
.fragment
, safe
=r
"!#$%&'()*+,/:;=?@{|}~")))
6183 # Source for `safe` arguments: https://url.spec.whatwg.org/#percent-encoded-bytes.
6186 def to_high_limit_path(path
):
6187 if sys
.platform
in ['win32', 'cygwin']:
6188 # Work around MAX_PATH limitation on Windows. The maximum allowed length for the individual path segments may still be quite limited.
6189 return r
'\\?\ '.rstrip() + os
.path
.abspath(path
)
6194 def format_field(obj
, field
=None, template
='%s', ignore
=(None, ''), default
='', func
=None):
6196 val
= obj
if obj
is not None else default
6198 val
= obj
.get(field
, default
)
6199 if func
and val
not in ignore
:
6201 return template
% val
if val
not in ignore
else default
6204 def clean_podcast_url(url
):
6205 return re
.sub(r
'''(?x)
6209 media\.blubrry\.com| # https://create.blubrry.com/resources/podcast-media-download-statistics/getting-started/
6212 (?:dts|www)\.podtrac\.com/(?:pts/)?redirect\.[0-9a-z]{3,4}| # http://analytics.podtrac.com/how-to-measure
6215 cn\.co| # https://podcorn.com/analytics-prefix/
6216 st\.fm # https://podsights.com/docs/
6221 _HEX_TABLE
= '0123456789abcdef'
6224 def random_uuidv4():
6225 return re
.sub(r
'[xy]', lambda x
: _HEX_TABLE
[random
.randint(0, 15)], 'xxxxxxxx-xxxx-4xxx-yxxx-xxxxxxxxxxxx')
6228 def make_dir(path
, to_screen
=None):
6230 dn
= os
.path
.dirname(path
)
6231 if dn
and not os
.path
.exists(dn
):
6234 except (OSError, IOError) as err
:
6235 if callable(to_screen
) is not None:
6236 to_screen('unable to create directory ' + error_to_compat_str(err
))
6240 def get_executable_path():
6241 from zipimport
import zipimporter
6242 if hasattr(sys
, 'frozen'): # Running from PyInstaller
6243 path
= os
.path
.dirname(sys
.executable
)
6244 elif isinstance(globals().get('__loader__'), zipimporter
): # Running from ZIP
6245 path
= os
.path
.join(os
.path
.dirname(__file__
), '../..')
6247 path
= os
.path
.join(os
.path
.dirname(__file__
), '..')
6248 return os
.path
.abspath(path
)
6251 def load_plugins(name
, suffix
, namespace
):
6252 plugin_info
= [None]
6255 plugin_info
= imp
.find_module(
6256 name
, [os
.path
.join(get_executable_path(), 'ytdlp_plugins')])
6257 plugins
= imp
.load_module(name
, *plugin_info
)
6258 for name
in dir(plugins
):
6259 if name
in namespace
:
6261 if not name
.endswith(suffix
):
6263 klass
= getattr(plugins
, name
)
6264 classes
.append(klass
)
6265 namespace
[name
] = klass
6269 if plugin_info
[0] is not None:
6270 plugin_info
[0].close()
6275 obj
, *path_list
, default
=None, expected_type
=None, get_all
=True,
6276 casesense
=True, is_user_input
=False, traverse_string
=False):
6277 ''' Traverse nested list/dict/tuple
6278 @param path_list A list of paths which are checked one by one.
6279 Each path is a list of keys where each key is a string,
6280 a tuple of strings or "...". When a tuple is given,
6281 all the keys given in the tuple are traversed, and
6282 "..." traverses all the keys in the object
6283 @param default Default value to return
6284 @param expected_type Only accept final value of this type (Can also be any callable)
6285 @param get_all Return all the values obtained from a path or only the first one
6286 @param casesense Whether to consider dictionary keys as case sensitive
6287 @param is_user_input Whether the keys are generated from user input. If True,
6288 strings are converted to int/slice if necessary
6289 @param traverse_string Whether to traverse inside strings. If True, any
6290 non-compatible object will also be converted into a string
6294 _lower
= lambda k
: (k
.lower() if isinstance(k
, str) else k
)
6295 path_list
= (map(_lower
, variadic(path
)) for path
in path_list
)
6297 def _traverse_obj(obj
, path
, _current_depth
=0):
6301 path
= tuple(variadic(path
))
6302 for i
, key
in enumerate(path
):
6303 if isinstance(key
, (list, tuple)):
6304 obj
= [_traverse_obj(obj
, sub_key
, _current_depth
) for sub_key
in key
]
6307 obj
= (obj
.values() if isinstance(obj
, dict)
6308 else obj
if isinstance(obj
, (list, tuple, LazyList
))
6309 else str(obj
) if traverse_string
else [])
6311 depth
= max(depth
, _current_depth
)
6312 return [_traverse_obj(inner_obj
, path
[i
+ 1:], _current_depth
) for inner_obj
in obj
]
6313 elif isinstance(obj
, dict) and not (is_user_input
and key
== ':'):
6314 obj
= (obj
.get(key
) if casesense
or (key
in obj
)
6315 else next((v
for k
, v
in obj
.items() if _lower(k
) == key
), None))
6318 key
= (int_or_none(key
) if ':' not in key
6319 else slice(*map(int_or_none
, key
.split(':'))))
6320 if key
== slice(None):
6321 return _traverse_obj(obj
, (..., *path
[i
+ 1:]), _current_depth
)
6322 if not isinstance(key
, (int, slice)):
6324 if not isinstance(obj
, (list, tuple, LazyList
)):
6325 if not traverse_string
:
6334 if isinstance(expected_type
, type):
6335 type_test
= lambda val
: val
if isinstance(val
, expected_type
) else None
6336 elif expected_type
is not None:
6337 type_test
= expected_type
6339 type_test
= lambda val
: val
6341 for path
in path_list
:
6343 val
= _traverse_obj(obj
, path
)
6346 for _
in range(depth
- 1):
6347 val
= itertools
.chain
.from_iterable(v
for v
in val
if v
is not None)
6348 val
= [v
for v
in map(type_test
, val
) if v
is not None]
6350 return val
if get_all
else val
[0]
6352 val
= type_test(val
)
6358 def traverse_dict(dictn
, keys
, casesense
=True):
6359 ''' For backward compatibility. Do not use '''
6360 return traverse_obj(dictn
, keys
, casesense
=casesense
,
6361 is_user_input
=True, traverse_string
=True)
6364 def variadic(x
, allowed_types
=(str, bytes)):
6365 return x
if isinstance(x
, collections
.abc
.Iterable
) and not isinstance(x
, allowed_types
) else (x
,)
projects / yt-dlp.git / blob