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 def __init__(self
, iterable
):
3976 self
.__iterable
= iter(iterable
)
3978 self
.__reversed
= False
3982 # We need to consume the entire iterable to iterate in reverse
3983 yield from self
.exhaust()
3985 yield from self
.__cache
3986 for item
in self
.__iterable
:
3987 self
.__cache
.append(item
)
3990 def __exhaust(self
):
3991 self
.__cache
.extend(self
.__iterable
)
3995 ''' Evaluate the entire iterable '''
3996 return self
.__exhaust
()[::-1 if self
.__reversed
else 1]
3999 def __reverse_index(x
):
4000 return None if x
is None else -(x
+ 1)
4002 def __getitem__(self
, idx
):
4003 if isinstance(idx
, slice):
4005 idx
= slice(self
.__reverse
_index
(idx
.start
), self
.__reverse
_index
(idx
.stop
), -(idx
.step
or 1))
4006 start
, stop
, step
= idx
.start
, idx
.stop
, idx
.step
or 1
4007 elif isinstance(idx
, int):
4009 idx
= self
.__reverse
_index
(idx
)
4010 start
, stop
, step
= idx
, idx
, 0
4012 raise TypeError('indices must be integers or slices')
4013 if ((start
or 0) < 0 or (stop
or 0) < 0
4014 or (start
is None and step
< 0)
4015 or (stop
is None and step
> 0)):
4016 # We need to consume the entire iterable to be able to slice from the end
4017 # Obviously, never use this with infinite iterables
4018 return self
.__exhaust
()[idx
]
4020 n
= max(start
or 0, stop
or 0) - len(self
.__cache
) + 1
4022 self
.__cache
.extend(itertools
.islice(self
.__iterable
, n
))
4023 return self
.__cache
[idx
]
4027 self
[-1] if self
.__reversed
else self
[0]
4034 return len(self
.__cache
)
4037 self
.__reversed
= not self
.__reversed
4041 # repr and str should mimic a list. So we exhaust the iterable
4042 return repr(self
.exhaust())
4045 return repr(self
.exhaust())
4050 # This is only useful for tests
4051 return len(self
.getslice())
4053 def __init__(self
, pagefunc
, pagesize
, use_cache
=True):
4054 self
._pagefunc
= pagefunc
4055 self
._pagesize
= pagesize
4056 self
._use
_cache
= use_cache
4059 def getpage(self
, pagenum
):
4060 page_results
= self
._cache
.get(pagenum
) or list(self
._pagefunc
(pagenum
))
4062 self
._cache
[pagenum
] = page_results
4065 def getslice(self
, start
=0, end
=None):
4066 return list(self
._getslice
(start
, end
))
4068 def _getslice(self
, start
, end
):
4069 raise NotImplementedError('This method must be implemented by subclasses')
4071 def __getitem__(self
, idx
):
4072 # NOTE: cache must be enabled if this is used
4073 if not isinstance(idx
, int) or idx
< 0:
4074 raise TypeError('indices must be non-negative integers')
4075 entries
= self
.getslice(idx
, idx
+ 1)
4076 return entries
[0] if entries
else None
4079 class OnDemandPagedList(PagedList
):
4080 def _getslice(self
, start
, end
):
4081 for pagenum
in itertools
.count(start
// self
._pagesize
):
4082 firstid
= pagenum
* self
._pagesize
4083 nextfirstid
= pagenum
* self
._pagesize
+ self
._pagesize
4084 if start
>= nextfirstid
:
4088 start
% self
._pagesize
4089 if firstid
<= start
< nextfirstid
4092 ((end
- 1) % self
._pagesize
) + 1
4093 if (end
is not None and firstid
<= end
<= nextfirstid
)
4096 page_results
= self
.getpage(pagenum
)
4097 if startv
!= 0 or endv
is not None:
4098 page_results
= page_results
[startv
:endv
]
4099 yield from page_results
4101 # A little optimization - if current page is not "full", ie. does
4102 # not contain page_size videos then we can assume that this page
4103 # is the last one - there are no more ids on further pages -
4104 # i.e. no need to query again.
4105 if len(page_results
) + startv
< self
._pagesize
:
4108 # If we got the whole page, but the next page is not interesting,
4109 # break out early as well
4110 if end
== nextfirstid
:
4114 class InAdvancePagedList(PagedList
):
4115 def __init__(self
, pagefunc
, pagecount
, pagesize
):
4116 self
._pagecount
= pagecount
4117 PagedList
.__init
__(self
, pagefunc
, pagesize
, True)
4119 def _getslice(self
, start
, end
):
4120 start_page
= start
// self
._pagesize
4122 self
._pagecount
if end
is None else (end
// self
._pagesize
+ 1))
4123 skip_elems
= start
- start_page
* self
._pagesize
4124 only_more
= None if end
is None else end
- start
4125 for pagenum
in range(start_page
, end_page
):
4126 page_results
= self
.getpage(pagenum
)
4128 page_results
= page_results
[skip_elems
:]
4130 if only_more
is not None:
4131 if len(page_results
) < only_more
:
4132 only_more
-= len(page_results
)
4134 yield from page_results
[:only_more
]
4136 yield from page_results
4139 def uppercase_escape(s
):
4140 unicode_escape
= codecs
.getdecoder('unicode_escape')
4142 r
'\\U[0-9a-fA-F]{8}',
4143 lambda m
: unicode_escape(m
.group(0))[0],
4147 def lowercase_escape(s
):
4148 unicode_escape
= codecs
.getdecoder('unicode_escape')
4150 r
'\\u[0-9a-fA-F]{4}',
4151 lambda m
: unicode_escape(m
.group(0))[0],
4155 def escape_rfc3986(s
):
4156 """Escape non-ASCII characters as suggested by RFC 3986"""
4157 if sys
.version_info
< (3, 0) and isinstance(s
, compat_str
):
4158 s
= s
.encode('utf-8')
4159 return compat_urllib_parse
.quote(s
, b
"%/;:@&=+$,!~*'()?#[]")
4162 def escape_url(url
):
4163 """Escape URL as suggested by RFC 3986"""
4164 url_parsed
= compat_urllib_parse_urlparse(url
)
4165 return url_parsed
._replace
(
4166 netloc
=url_parsed
.netloc
.encode('idna').decode('ascii'),
4167 path
=escape_rfc3986(url_parsed
.path
),
4168 params
=escape_rfc3986(url_parsed
.params
),
4169 query
=escape_rfc3986(url_parsed
.query
),
4170 fragment
=escape_rfc3986(url_parsed
.fragment
)
4175 return compat_parse_qs(compat_urllib_parse_urlparse(url
).query
)
4178 def read_batch_urls(batch_fd
):
4180 if not isinstance(url
, compat_str
):
4181 url
= url
.decode('utf-8', 'replace')
4182 BOM_UTF8
= ('\xef\xbb\xbf', '\ufeff')
4183 for bom
in BOM_UTF8
:
4184 if url
.startswith(bom
):
4185 url
= url
[len(bom
):]
4187 if not url
or url
.startswith(('#', ';', ']')):
4189 # "#" cannot be stripped out since it is part of the URI
4190 # However, it can be safely stipped out if follwing a whitespace
4191 return re
.split(r
'\s#', url
, 1)[0].rstrip()
4193 with contextlib
.closing(batch_fd
) as fd
:
4194 return [url
for url
in map(fixup
, fd
) if url
]
4197 def urlencode_postdata(*args
, **kargs
):
4198 return compat_urllib_parse_urlencode(*args
, **kargs
).encode('ascii')
4201 def update_url_query(url
, query
):
4204 parsed_url
= compat_urlparse
.urlparse(url
)
4205 qs
= compat_parse_qs(parsed_url
.query
)
4207 return compat_urlparse
.urlunparse(parsed_url
._replace
(
4208 query
=compat_urllib_parse_urlencode(qs
, True)))
4211 def update_Request(req
, url
=None, data
=None, headers
={}, query={}
):
4212 req_headers
= req
.headers
.copy()
4213 req_headers
.update(headers
)
4214 req_data
= data
or req
.data
4215 req_url
= update_url_query(url
or req
.get_full_url(), query
)
4216 req_get_method
= req
.get_method()
4217 if req_get_method
== 'HEAD':
4218 req_type
= HEADRequest
4219 elif req_get_method
== 'PUT':
4220 req_type
= PUTRequest
4222 req_type
= compat_urllib_request
.Request
4224 req_url
, data
=req_data
, headers
=req_headers
,
4225 origin_req_host
=req
.origin_req_host
, unverifiable
=req
.unverifiable
)
4226 if hasattr(req
, 'timeout'):
4227 new_req
.timeout
= req
.timeout
4231 def _multipart_encode_impl(data
, boundary
):
4232 content_type
= 'multipart/form-data; boundary=%s' % boundary
4235 for k
, v
in data
.items():
4236 out
+= b
'--' + boundary
.encode('ascii') + b
'\r\n'
4237 if isinstance(k
, compat_str
):
4238 k
= k
.encode('utf-8')
4239 if isinstance(v
, compat_str
):
4240 v
= v
.encode('utf-8')
4241 # RFC 2047 requires non-ASCII field names to be encoded, while RFC 7578
4242 # suggests sending UTF-8 directly. Firefox sends UTF-8, too
4243 content
= b
'Content-Disposition: form-data; name="' + k
+ b
'"\r\n\r\n' + v
+ b
'\r\n'
4244 if boundary
.encode('ascii') in content
:
4245 raise ValueError('Boundary overlaps with data')
4248 out
+= b
'--' + boundary
.encode('ascii') + b
'--\r\n'
4250 return out
, content_type
4253 def multipart_encode(data
, boundary
=None):
4255 Encode a dict to RFC 7578-compliant form-data
4258 A dict where keys and values can be either Unicode or bytes-like
4261 If specified a Unicode object, it's used as the boundary. Otherwise
4262 a random boundary is generated.
4264 Reference: https://tools.ietf.org/html/rfc7578
4266 has_specified_boundary
= boundary
is not None
4269 if boundary
is None:
4270 boundary
= '---------------' + str(random
.randrange(0x0fffffff, 0xffffffff))
4273 out
, content_type
= _multipart_encode_impl(data
, boundary
)
4276 if has_specified_boundary
:
4280 return out
, content_type
4283 def dict_get(d
, key_or_keys
, default
=None, skip_false_values
=True):
4284 if isinstance(key_or_keys
, (list, tuple)):
4285 for key
in key_or_keys
:
4286 if key
not in d
or d
[key
] is None or skip_false_values
and not d
[key
]:
4290 return d
.get(key_or_keys
, default
)
4293 def try_get(src
, getter
, expected_type
=None):
4294 for get
in variadic(getter
):
4297 except (AttributeError, KeyError, TypeError, IndexError):
4300 if expected_type
is None or isinstance(v
, expected_type
):
4304 def merge_dicts(*dicts
):
4306 for a_dict
in dicts
:
4307 for k
, v
in a_dict
.items():
4311 or (isinstance(v
, compat_str
) and v
4312 and isinstance(merged
[k
], compat_str
)
4313 and not merged
[k
])):
4318 def encode_compat_str(string
, encoding
=preferredencoding(), errors
='strict'):
4319 return string
if isinstance(string
, compat_str
) else compat_str(string
, encoding
, errors
)
4331 TV_PARENTAL_GUIDELINES
= {
4341 def parse_age_limit(s
):
4343 return s
if 0 <= s
<= 21 else None
4344 if not isinstance(s
, compat_basestring
):
4346 m
= re
.match(r
'^(?P<age>\d{1,2})\+?$', s
)
4348 return int(m
.group('age'))
4351 return US_RATINGS
[s
]
4352 m
= re
.match(r
'^TV[_-]?(%s)$' % '|'.join(k
[3:] for k
in TV_PARENTAL_GUIDELINES
), s
)
4354 return TV_PARENTAL_GUIDELINES
['TV-' + m
.group(1)]
4358 def strip_jsonp(code
):
4361 (?:window\.)?(?P<func_name>[a-zA-Z0-9_.$]*)
4362 (?:\s*&&\s*(?P=func_name))?
4363 \s*\(\s*(?P<callback_data>.*)\);?
4364 \s*?(?://[^\n]*)*$''',
4365 r
'\g<callback_data>', code
)
4368 def js_to_json(code
, vars={}):
4369 # vars is a dict of var, val pairs to substitute
4370 COMMENT_RE
= r
'/\*(?:(?!\*/).)*?\*/|//[^\n]*\n'
4371 SKIP_RE
= r
'\s*(?:{comment})?\s*'.format(comment
=COMMENT_RE
)
4373 (r
'(?s)^(0[xX][0-9a-fA-F]+){skip}:?$'.format(skip
=SKIP_RE
), 16),
4374 (r
'(?s)^(0+[0-7]+){skip}:?$'.format(skip
=SKIP_RE
), 8),
4379 if v
in ('true', 'false', 'null'):
4381 elif v
.startswith('/*') or v
.startswith('//') or v
.startswith('!') or v
== ',':
4384 if v
[0] in ("'", '"'):
4385 v
= re
.sub(r
'(?s)\\.|"', lambda m
: {
4390 }.get(m
.group(0), m
.group(0)), v
[1:-1])
4392 for regex
, base
in INTEGER_TABLE
:
4393 im
= re
.match(regex
, v
)
4395 i
= int(im
.group(1), base
)
4396 return '"%d":' % i
if v
.endswith(':') else '%d' % i
4403 return re
.sub(r
'''(?sx)
4404 "(?:[^"\\]*(?:\\\\|\\['"nurtbfx/\n]))*[^"\\]*"|
4405 '(?:[^'\\]*(?:\\\\|\\['"nurtbfx/\n]))*[^'\\]*'|
4406 {comment}|,(?={skip}[\]}}])|
4407 (?:(?<![0-9])[eE]|[a-df-zA-DF-Z_])[.a-zA-Z_0-9]*|
4408 \b(?:0[xX][0-9a-fA-F]+|0+[0-7]+)(?:{skip}:)?|
4411 '''.format(comment
=COMMENT_RE
, skip
=SKIP_RE
), fix_kv
, code
)
4414 def qualities(quality_ids
):
4415 """ Get a numeric quality value out of a list of possible values """
4418 return quality_ids
.index(qid
)
4425 'default': '%(title)s [%(id)s].%(ext)s',
4426 'chapter': '%(title)s - %(section_number)03d %(section_title)s [%(id)s].%(ext)s',
4432 'description': 'description',
4433 'annotation': 'annotations.xml',
4434 'infojson': 'info.json',
4435 'pl_thumbnail': None,
4436 'pl_description': 'description',
4437 'pl_infojson': 'info.json',
4440 # As of [1] format syntax is:
4441 # %[mapping_key][conversion_flags][minimum_width][.precision][length_modifier]type
4442 # 1. https://docs.python.org/2/library/stdtypes.html#string-formatting
4443 STR_FORMAT_RE_TMPL
= r
'''(?x)
4444 (?<!%)(?P<prefix>(?:%%)*)
4446 (?P<has_key>\((?P<key>{0})\))? # mapping key
4448 (?:[#0\-+ ]+)? # conversion flags (optional)
4449 (?:\d+)? # minimum field width (optional)
4450 (?:\.\d+)? # precision (optional)
4451 [hlL]? # length modifier (optional)
4452 {1} # conversion type
4457 STR_FORMAT_TYPES
= 'diouxXeEfFgGcrs'
4460 def limit_length(s
, length
):
4461 """ Add ellipses to overly long strings """
4466 return s
[:length
- len(ELLIPSES
)] + ELLIPSES
4470 def version_tuple(v
):
4471 return tuple(int(e
) for e
in re
.split(r
'[-.]', v
))
4474 def is_outdated_version(version
, limit
, assume_new
=True):
4476 return not assume_new
4478 return version_tuple(version
) < version_tuple(limit
)
4480 return not assume_new
4483 def ytdl_is_updateable():
4484 """ Returns if yt-dlp can be updated with -U """
4487 from zipimport
import zipimporter
4489 return isinstance(globals().get('__loader__'), zipimporter
) or hasattr(sys
, 'frozen')
4492 def args_to_str(args
):
4493 # Get a short string representation for a subprocess command
4494 return ' '.join(compat_shlex_quote(a
) for a
in args
)
4497 def error_to_compat_str(err
):
4499 # On python 2 error byte string must be decoded with proper
4500 # encoding rather than ascii
4501 if sys
.version_info
[0] < 3:
4502 err_str
= err_str
.decode(preferredencoding())
4506 def mimetype2ext(mt
):
4512 # Per RFC 3003, audio/mpeg can be .mp1, .mp2 or .mp3. Here use .mp3 as
4513 # it's the most popular one
4514 'audio/mpeg': 'mp3',
4515 'audio/x-wav': 'wav',
4520 _
, _
, res
= mt
.rpartition('/')
4521 res
= res
.split(';')[0].strip().lower()
4525 'smptett+xml': 'tt',
4529 'x-mp4-fragmented': 'mp4',
4530 'x-ms-sami': 'sami',
4533 'x-mpegurl': 'm3u8',
4534 'vnd.apple.mpegurl': 'm3u8',
4538 'vnd.ms-sstr+xml': 'ism',
4545 def parse_codecs(codecs_str
):
4546 # http://tools.ietf.org/html/rfc6381
4549 split_codecs
= list(filter(None, map(
4550 str.strip
, codecs_str
.strip().strip(',').split(','))))
4551 vcodec
, acodec
= None, None
4552 for full_codec
in split_codecs
:
4553 codec
= full_codec
.split('.')[0]
4554 if codec
in ('avc1', 'avc2', 'avc3', 'avc4', 'vp9', 'vp8', 'hev1', 'hev2', 'h263', 'h264', 'mp4v', 'hvc1', 'av01', 'theora'):
4557 elif codec
in ('mp4a', 'opus', 'vorbis', 'mp3', 'aac', 'ac-3', 'ec-3', 'eac3', 'dtsc', 'dtse', 'dtsh', 'dtsl'):
4561 write_string('WARNING: Unknown codec %s\n' % full_codec
, sys
.stderr
)
4562 if not vcodec
and not acodec
:
4563 if len(split_codecs
) == 2:
4565 'vcodec': split_codecs
[0],
4566 'acodec': split_codecs
[1],
4570 'vcodec': vcodec
or 'none',
4571 'acodec': acodec
or 'none',
4576 def urlhandle_detect_ext(url_handle
):
4577 getheader
= url_handle
.headers
.get
4579 cd
= getheader('Content-Disposition')
4581 m
= re
.match(r
'attachment;\s*filename="(?P<filename>[^"]+)"', cd
)
4583 e
= determine_ext(m
.group('filename'), default_ext
=None)
4587 return mimetype2ext(getheader('Content-Type'))
4590 def encode_data_uri(data
, mime_type
):
4591 return 'data:%s;base64,%s' % (mime_type
, base64
.b64encode(data
).decode('ascii'))
4594 def age_restricted(content_limit
, age_limit
):
4595 """ Returns True iff the content should be blocked """
4597 if age_limit
is None: # No limit set
4599 if content_limit
is None:
4600 return False # Content available for everyone
4601 return age_limit
< content_limit
4604 def is_html(first_bytes
):
4605 """ Detect whether a file contains HTML by examining its first bytes. """
4608 (b
'\xef\xbb\xbf', 'utf-8'),
4609 (b
'\x00\x00\xfe\xff', 'utf-32-be'),
4610 (b
'\xff\xfe\x00\x00', 'utf-32-le'),
4611 (b
'\xff\xfe', 'utf-16-le'),
4612 (b
'\xfe\xff', 'utf-16-be'),
4614 for bom
, enc
in BOMS
:
4615 if first_bytes
.startswith(bom
):
4616 s
= first_bytes
[len(bom
):].decode(enc
, 'replace')
4619 s
= first_bytes
.decode('utf-8', 'replace')
4621 return re
.match(r
'^\s*<', s
)
4624 def determine_protocol(info_dict
):
4625 protocol
= info_dict
.get('protocol')
4626 if protocol
is not None:
4629 url
= info_dict
['url']
4630 if url
.startswith('rtmp'):
4632 elif url
.startswith('mms'):
4634 elif url
.startswith('rtsp'):
4637 ext
= determine_ext(url
)
4643 return compat_urllib_parse_urlparse(url
).scheme
4646 def render_table(header_row
, data
, delim
=False, extraGap
=0, hideEmpty
=False):
4647 """ Render a list of rows, each as a list of values """
4649 def get_max_lens(table
):
4650 return [max(len(compat_str(v
)) for v
in col
) for col
in zip(*table
)]
4652 def filter_using_list(row
, filterArray
):
4653 return [col
for (take
, col
) in zip(filterArray
, row
) if take
]
4656 max_lens
= get_max_lens(data
)
4657 header_row
= filter_using_list(header_row
, max_lens
)
4658 data
= [filter_using_list(row
, max_lens
) for row
in data
]
4660 table
= [header_row
] + data
4661 max_lens
= get_max_lens(table
)
4663 table
= [header_row
] + [['-' * ml
for ml
in max_lens
]] + data
4664 format_str
= ' '.join('%-' + compat_str(ml
+ extraGap
) + 's' for ml
in max_lens
[:-1]) + ' %s'
4665 return '\n'.join(format_str
% tuple(row
) for row
in table
)
4668 def _match_one(filter_part
, dct
, incomplete
):
4669 # TODO: Generalize code with YoutubeDL._build_format_filter
4670 STRING_OPERATORS
= {
4671 '*=': operator
.contains
,
4672 '^=': lambda attr
, value
: attr
.startswith(value
),
4673 '$=': lambda attr
, value
: attr
.endswith(value
),
4674 '~=': lambda attr
, value
: re
.search(value
, attr
),
4676 COMPARISON_OPERATORS
= {
4678 '<=': operator
.le
, # "<=" must be defined above "<"
4685 operator_rex
= re
.compile(r
'''(?x)\s*
4687 \s*(?P<negation>!\s*)?(?P<op>%s)(?P<none_inclusive>\s*\?)?\s*
4689 (?P<intval>[0-9.]+(?:[kKmMgGtTpPeEzZyY]i?[Bb]?)?)|
4690 (?P<quote>["\'])(?P
<quotedstrval
>.+?
)(?P
=quote
)|
4694 ''' % '|'.join(map(re.escape, COMPARISON_OPERATORS.keys())))
4695 m = operator_rex.search(filter_part)
4697 unnegated_op = COMPARISON_OPERATORS[m.group('op')]
4698 if m.group('negation'):
4699 op = lambda attr, value: not unnegated_op(attr, value)
4702 actual_value = dct.get(m.group('key'))
4703 if (m.group('quotedstrval') is not None
4704 or m.group('strval') is not None
4705 # If the original field is a string and matching comparisonvalue is
4706 # a number we should respect the origin of the original field
4707 # and process comparison value as a string (see
4708 # https://github.com/ytdl-org/youtube-dl/issues/11082).
4709 or actual_value is not None and m.group('intval') is not None
4710 and isinstance(actual_value, compat_str)):
4711 comparison_value = m.group('quotedstrval') or m.group('strval') or m.group('intval')
4712 quote = m.group('quote')
4713 if quote is not None:
4714 comparison_value = comparison_value.replace(r'\%s' % quote, quote)
4716 if m.group('op') in STRING_OPERATORS:
4717 raise ValueError('Operator %s only supports string values!' % m.group('op'))
4719 comparison_value = int(m.group('intval'))
4721 comparison_value = parse_filesize(m.group('intval'))
4722 if comparison_value is None:
4723 comparison_value = parse_filesize(m.group('intval') + 'B')
4724 if comparison_value is None:
4726 'Invalid integer value %r in filter part %r' % (
4727 m.group('intval'), filter_part))
4728 if actual_value is None:
4729 return incomplete or m.group('none_inclusive')
4730 return op(actual_value, comparison_value)
4733 '': lambda v: (v is True) if isinstance(v, bool) else (v is not None),
4734 '!': lambda v: (v is False) if isinstance(v, bool) else (v is None),
4736 operator_rex = re.compile(r'''(?x
)\s
*
4737 (?P
<op
>%s)\s
*(?P
<key
>[a
-z_
]+)
4739 ''' % '|'.join(map(re.escape, UNARY_OPERATORS.keys())))
4740 m = operator_rex.search(filter_part)
4742 op = UNARY_OPERATORS[m.group('op')]
4743 actual_value = dct.get(m.group('key'))
4744 if incomplete and actual_value is None:
4746 return op(actual_value)
4748 raise ValueError('Invalid filter part %r' % filter_part)
4751 def match_str(filter_str, dct, incomplete=False):
4752 """ Filter a dictionary with a simple string syntax. Returns True (=passes filter) or false
4753 When incomplete, all conditions passes on missing fields
4756 _match_one(filter_part.replace(r'\&', '&'), dct, incomplete)
4757 for filter_part in re.split(r'(?<!\\)&', filter_str))
4760 def match_filter_func(filter_str):
4761 def _match_func(info_dict, *args, **kwargs):
4762 if match_str(filter_str, info_dict, *args, **kwargs):
4765 video_title = info_dict.get('title', info_dict.get('id', 'video'))
4766 return '%s does not pass filter %s, skipping ..' % (video_title, filter_str)
4770 def parse_dfxp_time_expr(time_expr):
4774 mobj = re.match(r'^(?P<time_offset>\d+(?:\.\d+)?)s?$', time_expr)
4776 return float(mobj.group('time_offset'))
4778 mobj = re.match(r'^(\d+):(\d\d):(\d\d(?:(?:\.|:)\d+)?)$', time_expr)
4780 return 3600 * int(mobj.group(1)) + 60 * int(mobj.group(2)) + float(mobj.group(3).replace(':', '.'))
4783 def srt_subtitles_timecode(seconds):
4784 return '%02d:%02d:%02d,%03d' % (seconds / 3600, (seconds % 3600) / 60, seconds % 60, (seconds % 1) * 1000)
4787 def dfxp2srt(dfxp_data):
4789 @param dfxp_data A
bytes-like
object containing DFXP data
4790 @returns A
unicode object containing converted SRT data
4792 LEGACY_NAMESPACES = (
4793 (b'http://www.w3.org/ns/ttml', [
4794 b'http://www.w3.org/2004/11/ttaf1',
4795 b'http://www.w3.org/2006/04/ttaf1',
4796 b'http://www.w3.org/2006/10/ttaf1',
4798 (b'http://www.w3.org/ns/ttml#styling', [
4799 b'http://www.w3.org/ns/ttml#style',
4803 SUPPORTED_STYLING = [
4812 _x = functools.partial(xpath_with_ns, ns_map={
4813 'xml': 'http://www.w3.org/XML/1998/namespace',
4814 'ttml': 'http://www.w3.org/ns/ttml',
4815 'tts': 'http://www.w3.org/ns/ttml#styling',
4821 class TTMLPElementParser(object):
4823 _unclosed_elements = []
4824 _applied_styles = []
4826 def start(self, tag, attrib):
4827 if tag in (_x('ttml:br'), 'br'):
4830 unclosed_elements = []
4832 element_style_id = attrib.get('style')
4834 style.update(default_style)
4835 if element_style_id:
4836 style.update(styles.get(element_style_id, {}))
4837 for prop in SUPPORTED_STYLING:
4838 prop_val = attrib.get(_x('tts:' + prop))
4840 style[prop] = prop_val
4843 for k, v in sorted(style.items()):
4844 if self._applied_styles and self._applied_styles[-1].get(k) == v:
4847 font += ' color="%s"' % v
4848 elif k == 'fontSize':
4849 font += ' size="%s"' % v
4850 elif k == 'fontFamily':
4851 font += ' face="%s"' % v
4852 elif k == 'fontWeight' and v == 'bold':
4854 unclosed_elements.append('b')
4855 elif k == 'fontStyle' and v == 'italic':
4857 unclosed_elements.append('i')
4858 elif k == 'textDecoration' and v == 'underline':
4860 unclosed_elements.append('u')
4862 self._out += '<font' + font + '>'
4863 unclosed_elements.append('font')
4865 if self._applied_styles:
4866 applied_style.update(self._applied_styles[-1])
4867 applied_style.update(style)
4868 self._applied_styles.append(applied_style)
4869 self._unclosed_elements.append(unclosed_elements)
4872 if tag not in (_x('ttml:br'), 'br'):
4873 unclosed_elements = self._unclosed_elements.pop()
4874 for element in reversed(unclosed_elements):
4875 self._out += '</%s>' % element
4876 if unclosed_elements and self._applied_styles:
4877 self._applied_styles.pop()
4879 def data(self, data):
4883 return self._out.strip()
4885 def parse_node(node):
4886 target = TTMLPElementParser()
4887 parser = xml.etree.ElementTree.XMLParser(target=target)
4888 parser.feed(xml.etree.ElementTree.tostring(node))
4889 return parser.close()
4891 for k, v in LEGACY_NAMESPACES:
4893 dfxp_data = dfxp_data.replace(ns, k)
4895 dfxp = compat_etree_fromstring(dfxp_data)
4897 paras = dfxp.findall(_x('.//ttml:p')) or dfxp.findall('.//p')
4900 raise ValueError('Invalid dfxp/TTML subtitle')
4904 for style in dfxp.findall(_x('.//ttml:style')):
4905 style_id = style.get('id') or style.get(_x('xml:id'))
4908 parent_style_id = style.get('style')
4910 if parent_style_id not in styles:
4913 styles[style_id] = styles[parent_style_id].copy()
4914 for prop in SUPPORTED_STYLING:
4915 prop_val = style.get(_x('tts:' + prop))
4917 styles.setdefault(style_id, {})[prop] = prop_val
4923 for p in ('body', 'div'):
4924 ele = xpath_element(dfxp, [_x('.//ttml:' + p), './/' + p])
4927 style = styles.get(ele.get('style'))
4930 default_style.update(style)
4932 for para, index in zip(paras, itertools.count(1)):
4933 begin_time = parse_dfxp_time_expr(para.attrib.get('begin'))
4934 end_time = parse_dfxp_time_expr(para.attrib.get('end'))
4935 dur = parse_dfxp_time_expr(para.attrib.get('dur'))
4936 if begin_time is None:
4941 end_time = begin_time + dur
4942 out.append('%d\n%s --> %s\n%s\n\n' % (
4944 srt_subtitles_timecode(begin_time),
4945 srt_subtitles_timecode(end_time),
4951 def cli_option(params, command_option, param):
4952 param = params.get(param)
4954 param = compat_str(param)
4955 return [command_option, param] if param is not None else []
4958 def cli_bool_option(params, command_option, param, true_value='true', false_value='false', separator=None):
4959 param = params.get(param)
4962 assert isinstance(param, bool)
4964 return [command_option + separator + (true_value if param else false_value)]
4965 return [command_option, true_value if param else false_value]
4968 def cli_valueless_option(params, command_option, param, expected_value=True):
4969 param = params.get(param)
4970 return [command_option] if param == expected_value else []
4973 def cli_configuration_args(argdict, keys, default=[], use_compat=True):
4974 if isinstance(argdict, (list, tuple)): # for backward compatibility
4981 assert isinstance(argdict, dict)
4983 assert isinstance(keys, (list, tuple))
4984 for key_list in keys:
4985 arg_list = list(filter(
4986 lambda x: x is not None,
4987 [argdict.get(key.lower()) for key in variadic(key_list)]))
4989 return [arg for args in arg_list for arg in args]
4993 def _configuration_args(main_key, argdict, exe, keys=None, default=[], use_compat=True):
4994 main_key, exe = main_key.lower(), exe.lower()
4995 root_key = exe if main_key == exe else f'{main_key}+{exe}'
4996 keys = [f'{root_key}{k}' for k in (keys or [''])]
4997 if root_key in keys:
4999 keys.append((main_key, exe))
5000 keys.append('default')
5003 return cli_configuration_args(argdict, keys, default, use_compat)
5006 class ISO639Utils(object):
5007 # See http://www.loc.gov/standards/iso639-2/ISO-639-2_utf-8.txt
5066 'iw': 'heb', # Replaced by he in 1989 revision
5076 'in': 'ind', # Replaced by id in 1989 revision
5191 'ji': 'yid', # Replaced by yi in 1989 revision
5199 def short2long(cls, code):
5200 """Convert language code from ISO 639-1 to ISO 639-2/T"""
5201 return cls._lang_map.get(code[:2])
5204 def long2short(cls, code):
5205 """Convert language code from ISO 639-2/T to ISO 639-1"""
5206 for short_name, long_name in cls._lang_map.items():
5207 if long_name == code:
5211 class ISO3166Utils(object):
5212 # From http://data.okfn.org/data/core/country-list
5214 'AF': 'Afghanistan',
5215 'AX': 'Åland Islands',
5218 'AS': 'American Samoa',
5223 'AG': 'Antigua and Barbuda',
5240 'BO': 'Bolivia, Plurinational State of',
5241 'BQ': 'Bonaire, Sint Eustatius and Saba',
5242 'BA': 'Bosnia and Herzegovina',
5244 'BV': 'Bouvet Island',
5246 'IO': 'British Indian Ocean Territory',
5247 'BN': 'Brunei Darussalam',
5249 'BF': 'Burkina Faso',
5255 'KY': 'Cayman Islands',
5256 'CF': 'Central African Republic',
5260 'CX': 'Christmas Island',
5261 'CC': 'Cocos (Keeling) Islands',
5265 'CD': 'Congo, the Democratic Republic of the',
5266 'CK': 'Cook Islands',
5268 'CI': 'Côte d\'Ivoire',
5273 'CZ': 'Czech Republic',
5277 'DO': 'Dominican Republic',
5280 'SV': 'El Salvador',
5281 'GQ': 'Equatorial Guinea',
5285 'FK': 'Falkland Islands (Malvinas)',
5286 'FO': 'Faroe Islands',
5290 'GF': 'French Guiana',
5291 'PF': 'French Polynesia',
5292 'TF': 'French Southern Territories',
5307 'GW': 'Guinea-Bissau',
5310 'HM': 'Heard Island and McDonald Islands',
5311 'VA': 'Holy See (Vatican City State)',
5318 'IR': 'Iran, Islamic Republic of',
5321 'IM': 'Isle of Man',
5331 'KP': 'Korea, Democratic People\'s Republic of',
5332 'KR': 'Korea, Republic of',
5335 'LA': 'Lao People\'s Democratic Republic',
5341 'LI': 'Liechtenstein',
5345 'MK': 'Macedonia, the Former Yugoslav Republic of',
5352 'MH': 'Marshall Islands',
5358 'FM': 'Micronesia, Federated States of',
5359 'MD': 'Moldova, Republic of',
5370 'NL': 'Netherlands',
5371 'NC': 'New Caledonia',
5372 'NZ': 'New Zealand',
5377 'NF': 'Norfolk Island',
5378 'MP': 'Northern Mariana Islands',
5383 'PS': 'Palestine, State of',
5385 'PG': 'Papua New Guinea',
5388 'PH': 'Philippines',
5392 'PR': 'Puerto Rico',
5396 'RU': 'Russian Federation',
5398 'BL': 'Saint Barthélemy',
5399 'SH': 'Saint Helena, Ascension and Tristan da Cunha',
5400 'KN': 'Saint Kitts and Nevis',
5401 'LC': 'Saint Lucia',
5402 'MF': 'Saint Martin (French part)',
5403 'PM': 'Saint Pierre and Miquelon',
5404 'VC': 'Saint Vincent and the Grenadines',
5407 'ST': 'Sao Tome and Principe',
5408 'SA': 'Saudi Arabia',
5412 'SL': 'Sierra Leone',
5414 'SX': 'Sint Maarten (Dutch part)',
5417 'SB': 'Solomon Islands',
5419 'ZA': 'South Africa',
5420 'GS': 'South Georgia and the South Sandwich Islands',
5421 'SS': 'South Sudan',
5426 'SJ': 'Svalbard and Jan Mayen',
5429 'CH': 'Switzerland',
5430 'SY': 'Syrian Arab Republic',
5431 'TW': 'Taiwan, Province of China',
5433 'TZ': 'Tanzania, United Republic of',
5435 'TL': 'Timor-Leste',
5439 'TT': 'Trinidad and Tobago',
5442 'TM': 'Turkmenistan',
5443 'TC': 'Turks and Caicos Islands',
5447 'AE': 'United Arab Emirates',
5448 'GB': 'United Kingdom',
5449 'US': 'United States',
5450 'UM': 'United States Minor Outlying Islands',
5454 'VE': 'Venezuela, Bolivarian Republic of',
5456 'VG': 'Virgin Islands, British',
5457 'VI': 'Virgin Islands, U.S.',
5458 'WF': 'Wallis and Futuna',
5459 'EH': 'Western Sahara',
5466 def short2full(cls, code):
5467 """Convert an ISO 3166-2 country code to the corresponding full name"""
5468 return cls._country_map.get(code.upper())
5471 class GeoUtils(object):
5472 # Major IPv4 address blocks per country
5474 'AD': '46.172.224.0/19',
5475 'AE': '94.200.0.0/13',
5476 'AF': '149.54.0.0/17',
5477 'AG': '209.59.64.0/18',
5478 'AI': '204.14.248.0/21',
5479 'AL': '46.99.0.0/16',
5480 'AM': '46.70.0.0/15',
5481 'AO': '105.168.0.0/13',
5482 'AP': '182.50.184.0/21',
5483 'AQ': '23.154.160.0/24',
5484 'AR': '181.0.0.0/12',
5485 'AS': '202.70.112.0/20',
5486 'AT': '77.116.0.0/14',
5487 'AU': '1.128.0.0/11',
5488 'AW': '181.41.0.0/18',
5489 'AX': '185.217.4.0/22',
5490 'AZ': '5.197.0.0/16',
5491 'BA': '31.176.128.0/17',
5492 'BB': '65.48.128.0/17',
5493 'BD': '114.130.0.0/16',
5495 'BF': '102.178.0.0/15',
5496 'BG': '95.42.0.0/15',
5497 'BH': '37.131.0.0/17',
5498 'BI': '154.117.192.0/18',
5499 'BJ': '137.255.0.0/16',
5500 'BL': '185.212.72.0/23',
5501 'BM': '196.12.64.0/18',
5502 'BN': '156.31.0.0/16',
5503 'BO': '161.56.0.0/16',
5504 'BQ': '161.0.80.0/20',
5505 'BR': '191.128.0.0/12',
5506 'BS': '24.51.64.0/18',
5507 'BT': '119.2.96.0/19',
5508 'BW': '168.167.0.0/16',
5509 'BY': '178.120.0.0/13',
5510 'BZ': '179.42.192.0/18',
5511 'CA': '99.224.0.0/11',
5512 'CD': '41.243.0.0/16',
5513 'CF': '197.242.176.0/21',
5514 'CG': '160.113.0.0/16',
5515 'CH': '85.0.0.0/13',
5516 'CI': '102.136.0.0/14',
5517 'CK': '202.65.32.0/19',
5518 'CL': '152.172.0.0/14',
5519 'CM': '102.244.0.0/14',
5520 'CN': '36.128.0.0/10',
5521 'CO': '181.240.0.0/12',
5522 'CR': '201.192.0.0/12',
5523 'CU': '152.206.0.0/15',
5524 'CV': '165.90.96.0/19',
5525 'CW': '190.88.128.0/17',
5526 'CY': '31.153.0.0/16',
5527 'CZ': '88.100.0.0/14',
5529 'DJ': '197.241.0.0/17',
5530 'DK': '87.48.0.0/12',
5531 'DM': '192.243.48.0/20',
5532 'DO': '152.166.0.0/15',
5533 'DZ': '41.96.0.0/12',
5534 'EC': '186.68.0.0/15',
5535 'EE': '90.190.0.0/15',
5536 'EG': '156.160.0.0/11',
5537 'ER': '196.200.96.0/20',
5538 'ES': '88.0.0.0/11',
5539 'ET': '196.188.0.0/14',
5540 'EU': '2.16.0.0/13',
5541 'FI': '91.152.0.0/13',
5542 'FJ': '144.120.0.0/16',
5543 'FK': '80.73.208.0/21',
5544 'FM': '119.252.112.0/20',
5545 'FO': '88.85.32.0/19',
5547 'GA': '41.158.0.0/15',
5549 'GD': '74.122.88.0/21',
5550 'GE': '31.146.0.0/16',
5551 'GF': '161.22.64.0/18',
5552 'GG': '62.68.160.0/19',
5553 'GH': '154.160.0.0/12',
5554 'GI': '95.164.0.0/16',
5555 'GL': '88.83.0.0/19',
5556 'GM': '160.182.0.0/15',
5557 'GN': '197.149.192.0/18',
5558 'GP': '104.250.0.0/19',
5559 'GQ': '105.235.224.0/20',
5560 'GR': '94.64.0.0/13',
5561 'GT': '168.234.0.0/16',
5562 'GU': '168.123.0.0/16',
5563 'GW': '197.214.80.0/20',
5564 'GY': '181.41.64.0/18',
5565 'HK': '113.252.0.0/14',
5566 'HN': '181.210.0.0/16',
5567 'HR': '93.136.0.0/13',
5568 'HT': '148.102.128.0/17',
5569 'HU': '84.0.0.0/14',
5570 'ID': '39.192.0.0/10',
5571 'IE': '87.32.0.0/12',
5572 'IL': '79.176.0.0/13',
5573 'IM': '5.62.80.0/20',
5574 'IN': '117.192.0.0/10',
5575 'IO': '203.83.48.0/21',
5576 'IQ': '37.236.0.0/14',
5577 'IR': '2.176.0.0/12',
5578 'IS': '82.221.0.0/16',
5579 'IT': '79.0.0.0/10',
5580 'JE': '87.244.64.0/18',
5581 'JM': '72.27.0.0/17',
5582 'JO': '176.29.0.0/16',
5583 'JP': '133.0.0.0/8',
5584 'KE': '105.48.0.0/12',
5585 'KG': '158.181.128.0/17',
5586 'KH': '36.37.128.0/17',
5587 'KI': '103.25.140.0/22',
5588 'KM': '197.255.224.0/20',
5589 'KN': '198.167.192.0/19',
5590 'KP': '175.45.176.0/22',
5591 'KR': '175.192.0.0/10',
5592 'KW': '37.36.0.0/14',
5593 'KY': '64.96.0.0/15',
5594 'KZ': '2.72.0.0/13',
5595 'LA': '115.84.64.0/18',
5596 'LB': '178.135.0.0/16',
5597 'LC': '24.92.144.0/20',
5598 'LI': '82.117.0.0/19',
5599 'LK': '112.134.0.0/15',
5600 'LR': '102.183.0.0/16',
5601 'LS': '129.232.0.0/17',
5602 'LT': '78.56.0.0/13',
5603 'LU': '188.42.0.0/16',
5604 'LV': '46.109.0.0/16',
5605 'LY': '41.252.0.0/14',
5606 'MA': '105.128.0.0/11',
5607 'MC': '88.209.64.0/18',
5608 'MD': '37.246.0.0/16',
5609 'ME': '178.175.0.0/17',
5610 'MF': '74.112.232.0/21',
5611 'MG': '154.126.0.0/17',
5612 'MH': '117.103.88.0/21',
5613 'MK': '77.28.0.0/15',
5614 'ML': '154.118.128.0/18',
5615 'MM': '37.111.0.0/17',
5616 'MN': '49.0.128.0/17',
5617 'MO': '60.246.0.0/16',
5618 'MP': '202.88.64.0/20',
5619 'MQ': '109.203.224.0/19',
5620 'MR': '41.188.64.0/18',
5621 'MS': '208.90.112.0/22',
5622 'MT': '46.11.0.0/16',
5623 'MU': '105.16.0.0/12',
5624 'MV': '27.114.128.0/18',
5625 'MW': '102.70.0.0/15',
5626 'MX': '187.192.0.0/11',
5627 'MY': '175.136.0.0/13',
5628 'MZ': '197.218.0.0/15',
5629 'NA': '41.182.0.0/16',
5630 'NC': '101.101.0.0/18',
5631 'NE': '197.214.0.0/18',
5632 'NF': '203.17.240.0/22',
5633 'NG': '105.112.0.0/12',
5634 'NI': '186.76.0.0/15',
5635 'NL': '145.96.0.0/11',
5636 'NO': '84.208.0.0/13',
5637 'NP': '36.252.0.0/15',
5638 'NR': '203.98.224.0/19',
5639 'NU': '49.156.48.0/22',
5640 'NZ': '49.224.0.0/14',
5641 'OM': '5.36.0.0/15',
5642 'PA': '186.72.0.0/15',
5643 'PE': '186.160.0.0/14',
5644 'PF': '123.50.64.0/18',
5645 'PG': '124.240.192.0/19',
5646 'PH': '49.144.0.0/13',
5647 'PK': '39.32.0.0/11',
5648 'PL': '83.0.0.0/11',
5649 'PM': '70.36.0.0/20',
5650 'PR': '66.50.0.0/16',
5651 'PS': '188.161.0.0/16',
5652 'PT': '85.240.0.0/13',
5653 'PW': '202.124.224.0/20',
5654 'PY': '181.120.0.0/14',
5655 'QA': '37.210.0.0/15',
5656 'RE': '102.35.0.0/16',
5657 'RO': '79.112.0.0/13',
5658 'RS': '93.86.0.0/15',
5659 'RU': '5.136.0.0/13',
5660 'RW': '41.186.0.0/16',
5661 'SA': '188.48.0.0/13',
5662 'SB': '202.1.160.0/19',
5663 'SC': '154.192.0.0/11',
5664 'SD': '102.120.0.0/13',
5665 'SE': '78.64.0.0/12',
5666 'SG': '8.128.0.0/10',
5667 'SI': '188.196.0.0/14',
5668 'SK': '78.98.0.0/15',
5669 'SL': '102.143.0.0/17',
5670 'SM': '89.186.32.0/19',
5671 'SN': '41.82.0.0/15',
5672 'SO': '154.115.192.0/18',
5673 'SR': '186.179.128.0/17',
5674 'SS': '105.235.208.0/21',
5675 'ST': '197.159.160.0/19',
5676 'SV': '168.243.0.0/16',
5677 'SX': '190.102.0.0/20',
5679 'SZ': '41.84.224.0/19',
5680 'TC': '65.255.48.0/20',
5681 'TD': '154.68.128.0/19',
5682 'TG': '196.168.0.0/14',
5683 'TH': '171.96.0.0/13',
5684 'TJ': '85.9.128.0/18',
5685 'TK': '27.96.24.0/21',
5686 'TL': '180.189.160.0/20',
5687 'TM': '95.85.96.0/19',
5688 'TN': '197.0.0.0/11',
5689 'TO': '175.176.144.0/21',
5690 'TR': '78.160.0.0/11',
5691 'TT': '186.44.0.0/15',
5692 'TV': '202.2.96.0/19',
5693 'TW': '120.96.0.0/11',
5694 'TZ': '156.156.0.0/14',
5695 'UA': '37.52.0.0/14',
5696 'UG': '102.80.0.0/13',
5698 'UY': '167.56.0.0/13',
5699 'UZ': '84.54.64.0/18',
5700 'VA': '212.77.0.0/19',
5701 'VC': '207.191.240.0/21',
5702 'VE': '186.88.0.0/13',
5703 'VG': '66.81.192.0/20',
5704 'VI': '146.226.0.0/16',
5705 'VN': '14.160.0.0/11',
5706 'VU': '202.80.32.0/20',
5707 'WF': '117.20.32.0/21',
5708 'WS': '202.4.32.0/19',
5709 'YE': '134.35.0.0/16',
5710 'YT': '41.242.116.0/22',
5711 'ZA': '41.0.0.0/11',
5712 'ZM': '102.144.0.0/13',
5713 'ZW': '102.177.192.0/18',
5717 def random_ipv4(cls, code_or_block):
5718 if len(code_or_block) == 2:
5719 block = cls._country_ip_map.get(code_or_block.upper())
5723 block = code_or_block
5724 addr, preflen = block.split('/')
5725 addr_min = compat_struct_unpack('!L', socket.inet_aton(addr))[0]
5726 addr_max = addr_min | (0xffffffff >> int(preflen))
5727 return compat_str(socket.inet_ntoa(
5728 compat_struct_pack('!L', random.randint(addr_min, addr_max))))
5731 class PerRequestProxyHandler(compat_urllib_request.ProxyHandler):
5732 def __init__(self, proxies=None):
5733 # Set default handlers
5734 for type in ('http', 'https'):
5735 setattr(self, '%s_open' % type,
5736 lambda r, proxy='__noproxy__', type=type, meth=self.proxy_open:
5737 meth(r, proxy, type))
5738 compat_urllib_request.ProxyHandler.__init__(self, proxies)
5740 def proxy_open(self, req, proxy, type):
5741 req_proxy = req.headers.get('Ytdl-request-proxy')
5742 if req_proxy is not None:
5744 del req.headers['Ytdl-request-proxy']
5746 if proxy == '__noproxy__':
5747 return None # No Proxy
5748 if compat_urlparse.urlparse(proxy).scheme.lower() in ('socks', 'socks4', 'socks4a', 'socks5'):
5749 req.add_header('Ytdl-socks-proxy', proxy)
5750 # yt-dlp's http/https handlers do wrapping the socket with socks
5752 return compat_urllib_request.ProxyHandler.proxy_open(
5753 self, req, proxy, type)
5756 # Both long_to_bytes and bytes_to_long are adapted from PyCrypto, which is
5757 # released into Public Domain
5758 # https://github.com/dlitz/pycrypto/blob/master/lib/Crypto/Util/number.py#L387
5760 def long_to_bytes(n, blocksize=0):
5761 """long_to_bytes(n:long, blocksize:int) : string
5762 Convert a long integer to a byte string.
5764 If optional blocksize is given and greater than zero, pad the front of the
5765 byte string with binary zeros so that the length is a multiple of
5768 # after much testing, this algorithm was deemed to be the fastest
5772 s = compat_struct_pack('>I', n & 0xffffffff) + s
5774 # strip off leading zeros
5775 for i in range(len(s)):
5776 if s[i] != b'\000'[0]:
5779 # only happens when n == 0
5783 # add back some pad bytes. this could be done more efficiently w.r.t. the
5784 # de-padding being done above, but sigh...
5785 if blocksize > 0 and len(s) % blocksize:
5786 s = (blocksize - len(s) % blocksize) * b'\000' + s
5790 def bytes_to_long(s):
5791 """bytes_to_long(string) : long
5792 Convert a byte string to a long integer.
5794 This is (essentially) the inverse of long_to_bytes().
5799 extra = (4 - length % 4)
5800 s = b'\000' * extra + s
5801 length = length + extra
5802 for i in range(0, length, 4):
5803 acc = (acc << 32) + compat_struct_unpack('>I', s[i:i + 4])[0]
5807 def ohdave_rsa_encrypt(data, exponent, modulus):
5809 Implement OHDave
's RSA algorithm. See http://www.ohdave.com/rsa/
5812 data: data to encrypt, bytes-like object
5813 exponent, modulus: parameter e and N of RSA algorithm, both integer
5814 Output: hex string of encrypted data
5816 Limitation: supports one block encryption only
5819 payload = int(binascii.hexlify(data[::-1]), 16)
5820 encrypted = pow(payload, exponent, modulus)
5821 return '%x' % encrypted
5824 def pkcs1pad(data, length):
5826 Padding input data with PKCS#1 scheme
5828 @param {int[]} data input data
5829 @param {int} length target length
5830 @returns {int[]} padded data
5832 if len(data) > length - 11:
5833 raise ValueError('Input data too
long for PKCS
#1 padding')
5835 pseudo_random
= [random
.randint(0, 254) for _
in range(length
- len(data
) - 3)]
5836 return [0, 2] + pseudo_random
+ [0] + data
5839 def encode_base_n(num
, n
, table
=None):
5840 FULL_TABLE
= '0123456789abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ'
5842 table
= FULL_TABLE
[:n
]
5845 raise ValueError('base %d exceeds table length %d' % (n
, len(table
)))
5852 ret
= table
[num
% n
] + ret
5857 def decode_packed_codes(code
):
5858 mobj
= re
.search(PACKED_CODES_RE
, code
)
5859 obfuscated_code
, base
, count
, symbols
= mobj
.groups()
5862 symbols
= symbols
.split('|')
5867 base_n_count
= encode_base_n(count
, base
)
5868 symbol_table
[base_n_count
] = symbols
[count
] or base_n_count
5871 r
'\b(\w+)\b', lambda mobj
: symbol_table
[mobj
.group(0)],
5875 def caesar(s
, alphabet
, shift
):
5880 alphabet
[(alphabet
.index(c
) + shift
) % l
] if c
in alphabet
else c
5885 return caesar(s
, r
'''!"#$%&'()*+,-./0123456789:;<=>?@ABCDEFGHIJKLMNOPQRSTUVWXYZ[\]^_`abcdefghijklmnopqrstuvwxyz{|}~''', 47)
5888 def parse_m3u8_attributes(attrib
):
5890 for (key
, val
) in re
.findall(r
'(?P<key>[A-Z0-9-]+)=(?P<val>"[^"]+"|[^",]+)(?:,|$)', attrib
):
5891 if val
.startswith('"'):
5897 def urshift(val
, n
):
5898 return val
>> n
if val
>= 0 else (val
+ 0x100000000) >> n
5901 # Based on png2str() written by @gdkchan and improved by @yokrysty
5902 # Originally posted at https://github.com/ytdl-org/youtube-dl/issues/9706
5903 def decode_png(png_data
):
5904 # Reference: https://www.w3.org/TR/PNG/
5905 header
= png_data
[8:]
5907 if png_data
[:8] != b
'\x89PNG\x0d\x0a\x1a\x0a' or header
[4:8] != b
'IHDR':
5908 raise IOError('Not a valid PNG file.')
5910 int_map
= {1: '>B', 2: '>H', 4: '>I'}
5911 unpack_integer
= lambda x
: compat_struct_unpack(int_map
[len(x
)], x
)[0]
5916 length
= unpack_integer(header
[:4])
5919 chunk_type
= header
[:4]
5922 chunk_data
= header
[:length
]
5923 header
= header
[length
:]
5925 header
= header
[4:] # Skip CRC
5933 ihdr
= chunks
[0]['data']
5935 width
= unpack_integer(ihdr
[:4])
5936 height
= unpack_integer(ihdr
[4:8])
5940 for chunk
in chunks
:
5941 if chunk
['type'] == b
'IDAT':
5942 idat
+= chunk
['data']
5945 raise IOError('Unable to read PNG data.')
5947 decompressed_data
= bytearray(zlib
.decompress(idat
))
5952 def _get_pixel(idx
):
5957 for y
in range(height
):
5958 basePos
= y
* (1 + stride
)
5959 filter_type
= decompressed_data
[basePos
]
5963 pixels
.append(current_row
)
5965 for x
in range(stride
):
5966 color
= decompressed_data
[1 + basePos
+ x
]
5967 basex
= y
* stride
+ x
5972 left
= _get_pixel(basex
- 3)
5974 up
= _get_pixel(basex
- stride
)
5976 if filter_type
== 1: # Sub
5977 color
= (color
+ left
) & 0xff
5978 elif filter_type
== 2: # Up
5979 color
= (color
+ up
) & 0xff
5980 elif filter_type
== 3: # Average
5981 color
= (color
+ ((left
+ up
) >> 1)) & 0xff
5982 elif filter_type
== 4: # Paeth
5988 c
= _get_pixel(basex
- stride
- 3)
5996 if pa
<= pb
and pa
<= pc
:
5997 color
= (color
+ a
) & 0xff
5999 color
= (color
+ b
) & 0xff
6001 color
= (color
+ c
) & 0xff
6003 current_row
.append(color
)
6005 return width
, height
, pixels
6008 def write_xattr(path
, key
, value
):
6009 # This mess below finds the best xattr tool for the job
6011 # try the pyxattr module...
6014 if hasattr(xattr
, 'set'): # pyxattr
6015 # Unicode arguments are not supported in python-pyxattr until
6017 # See https://github.com/ytdl-org/youtube-dl/issues/5498
6018 pyxattr_required_version
= '0.5.0'
6019 if version_tuple(xattr
.__version
__) < version_tuple(pyxattr_required_version
):
6020 # TODO: fallback to CLI tools
6021 raise XAttrUnavailableError(
6022 'python-pyxattr is detected but is too old. '
6023 'yt-dlp requires %s or above while your version is %s. '
6024 'Falling back to other xattr implementations' % (
6025 pyxattr_required_version
, xattr
.__version
__))
6027 setxattr
= xattr
.set
6029 setxattr
= xattr
.setxattr
6032 setxattr(path
, key
, value
)
6033 except EnvironmentError as e
:
6034 raise XAttrMetadataError(e
.errno
, e
.strerror
)
6037 if compat_os_name
== 'nt':
6038 # Write xattrs to NTFS Alternate Data Streams:
6039 # http://en.wikipedia.org/wiki/NTFS#Alternate_data_streams_.28ADS.29
6040 assert ':' not in key
6041 assert os
.path
.exists(path
)
6043 ads_fn
= path
+ ':' + key
6045 with open(ads_fn
, 'wb') as f
:
6047 except EnvironmentError as e
:
6048 raise XAttrMetadataError(e
.errno
, e
.strerror
)
6050 user_has_setfattr
= check_executable('setfattr', ['--version'])
6051 user_has_xattr
= check_executable('xattr', ['-h'])
6053 if user_has_setfattr
or user_has_xattr
:
6055 value
= value
.decode('utf-8')
6056 if user_has_setfattr
:
6057 executable
= 'setfattr'
6058 opts
= ['-n', key
, '-v', value
]
6059 elif user_has_xattr
:
6060 executable
= 'xattr'
6061 opts
= ['-w', key
, value
]
6063 cmd
= ([encodeFilename(executable
, True)]
6064 + [encodeArgument(o
) for o
in opts
]
6065 + [encodeFilename(path
, True)])
6068 p
= subprocess
.Popen(
6069 cmd
, stdout
=subprocess
.PIPE
, stderr
=subprocess
.PIPE
, stdin
=subprocess
.PIPE
)
6070 except EnvironmentError as e
:
6071 raise XAttrMetadataError(e
.errno
, e
.strerror
)
6072 stdout
, stderr
= process_communicate_or_kill(p
)
6073 stderr
= stderr
.decode('utf-8', 'replace')
6074 if p
.returncode
!= 0:
6075 raise XAttrMetadataError(p
.returncode
, stderr
)
6078 # On Unix, and can't find pyxattr, setfattr, or xattr.
6079 if sys
.platform
.startswith('linux'):
6080 raise XAttrUnavailableError(
6081 "Couldn't find a tool to set the xattrs. "
6082 "Install either the python 'pyxattr' or 'xattr' "
6083 "modules, or the GNU 'attr' package "
6084 "(which contains the 'setfattr' tool).")
6086 raise XAttrUnavailableError(
6087 "Couldn't find a tool to set the xattrs. "
6088 "Install either the python 'xattr' module, "
6089 "or the 'xattr' binary.")
6092 def random_birthday(year_field
, month_field
, day_field
):
6093 start_date
= datetime
.date(1950, 1, 1)
6094 end_date
= datetime
.date(1995, 12, 31)
6095 offset
= random
.randint(0, (end_date
- start_date
).days
)
6096 random_date
= start_date
+ datetime
.timedelta(offset
)
6098 year_field
: str(random_date
.year
),
6099 month_field
: str(random_date
.month
),
6100 day_field
: str(random_date
.day
),
6104 # Templates for internet shortcut files, which are plain text files.
6105 DOT_URL_LINK_TEMPLATE
= '''
6110 DOT_WEBLOC_LINK_TEMPLATE
= '''
6111 <?xml version="1.0" encoding="UTF-8"?>
6112 <!DOCTYPE plist PUBLIC "-//Apple//DTD PLIST 1.0//EN" "http://www.apple.com/DTDs/PropertyList-1.0.dtd">
6113 <plist version="1.0">
6116 \t<string>%(url)s</string>
6121 DOT_DESKTOP_LINK_TEMPLATE
= '''
6131 def iri_to_uri(iri
):
6133 Converts an IRI (Internationalized Resource Identifier, allowing Unicode characters) to a URI (Uniform Resource Identifier, ASCII-only).
6135 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.
6138 iri_parts
= compat_urllib_parse_urlparse(iri
)
6140 if '[' in iri_parts
.netloc
:
6141 raise ValueError('IPv6 URIs are not, yet, supported.')
6142 # Querying `.netloc`, when there's only one bracket, also raises a ValueError.
6144 # 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.
6147 if iri_parts
.username
:
6148 net_location
+= compat_urllib_parse_quote(iri_parts
.username
, safe
=r
"!$%&'()*+,~")
6149 if iri_parts
.password
is not None:
6150 net_location
+= ':' + compat_urllib_parse_quote(iri_parts
.password
, safe
=r
"!$%&'()*+,~")
6153 net_location
+= iri_parts
.hostname
.encode('idna').decode('utf-8') # Punycode for Unicode hostnames.
6154 # The 'idna' encoding produces ASCII text.
6155 if iri_parts
.port
is not None and iri_parts
.port
!= 80:
6156 net_location
+= ':' + str(iri_parts
.port
)
6158 return compat_urllib_parse_urlunparse(
6162 compat_urllib_parse_quote_plus(iri_parts
.path
, safe
=r
"!$%&'()*+,/:;=@|~"),
6164 # Unsure about the `safe` argument, since this is a legacy way of handling parameters.
6165 compat_urllib_parse_quote_plus(iri_parts
.params
, safe
=r
"!$%&'()*+,/:;=@|~"),
6167 # Not totally sure about the `safe` argument, since the source does not explicitly mention the query URI component.
6168 compat_urllib_parse_quote_plus(iri_parts
.query
, safe
=r
"!$%&'()*+,/:;=?@{|}~"),
6170 compat_urllib_parse_quote_plus(iri_parts
.fragment
, safe
=r
"!#$%&'()*+,/:;=?@{|}~")))
6172 # Source for `safe` arguments: https://url.spec.whatwg.org/#percent-encoded-bytes.
6175 def to_high_limit_path(path
):
6176 if sys
.platform
in ['win32', 'cygwin']:
6177 # Work around MAX_PATH limitation on Windows. The maximum allowed length for the individual path segments may still be quite limited.
6178 return r
'\\?\ '.rstrip() + os
.path
.abspath(path
)
6183 def format_field(obj
, field
=None, template
='%s', ignore
=(None, ''), default
='', func
=None):
6185 val
= obj
if obj
is not None else default
6187 val
= obj
.get(field
, default
)
6188 if func
and val
not in ignore
:
6190 return template
% val
if val
not in ignore
else default
6193 def clean_podcast_url(url
):
6194 return re
.sub(r
'''(?x)
6198 media\.blubrry\.com| # https://create.blubrry.com/resources/podcast-media-download-statistics/getting-started/
6201 (?:dts|www)\.podtrac\.com/(?:pts/)?redirect\.[0-9a-z]{3,4}| # http://analytics.podtrac.com/how-to-measure
6204 cn\.co| # https://podcorn.com/analytics-prefix/
6205 st\.fm # https://podsights.com/docs/
6210 _HEX_TABLE
= '0123456789abcdef'
6213 def random_uuidv4():
6214 return re
.sub(r
'[xy]', lambda x
: _HEX_TABLE
[random
.randint(0, 15)], 'xxxxxxxx-xxxx-4xxx-yxxx-xxxxxxxxxxxx')
6217 def make_dir(path
, to_screen
=None):
6219 dn
= os
.path
.dirname(path
)
6220 if dn
and not os
.path
.exists(dn
):
6223 except (OSError, IOError) as err
:
6224 if callable(to_screen
) is not None:
6225 to_screen('unable to create directory ' + error_to_compat_str(err
))
6229 def get_executable_path():
6230 from zipimport
import zipimporter
6231 if hasattr(sys
, 'frozen'): # Running from PyInstaller
6232 path
= os
.path
.dirname(sys
.executable
)
6233 elif isinstance(globals().get('__loader__'), zipimporter
): # Running from ZIP
6234 path
= os
.path
.join(os
.path
.dirname(__file__
), '../..')
6236 path
= os
.path
.join(os
.path
.dirname(__file__
), '..')
6237 return os
.path
.abspath(path
)
6240 def load_plugins(name
, suffix
, namespace
):
6241 plugin_info
= [None]
6244 plugin_info
= imp
.find_module(
6245 name
, [os
.path
.join(get_executable_path(), 'ytdlp_plugins')])
6246 plugins
= imp
.load_module(name
, *plugin_info
)
6247 for name
in dir(plugins
):
6248 if name
in namespace
:
6250 if not name
.endswith(suffix
):
6252 klass
= getattr(plugins
, name
)
6253 classes
.append(klass
)
6254 namespace
[name
] = klass
6258 if plugin_info
[0] is not None:
6259 plugin_info
[0].close()
6264 obj
, *path_list
, default
=None, expected_type
=None, get_all
=True,
6265 casesense
=True, is_user_input
=False, traverse_string
=False):
6266 ''' Traverse nested list/dict/tuple
6267 @param path_list A list of paths which are checked one by one.
6268 Each path is a list of keys where each key is a string,
6269 a tuple of strings or "...". When a tuple is given,
6270 all the keys given in the tuple are traversed, and
6271 "..." traverses all the keys in the object
6272 @param default Default value to return
6273 @param expected_type Only accept final value of this type (Can also be any callable)
6274 @param get_all Return all the values obtained from a path or only the first one
6275 @param casesense Whether to consider dictionary keys as case sensitive
6276 @param is_user_input Whether the keys are generated from user input. If True,
6277 strings are converted to int/slice if necessary
6278 @param traverse_string Whether to traverse inside strings. If True, any
6279 non-compatible object will also be converted into a string
6283 _lower
= lambda k
: (k
.lower() if isinstance(k
, str) else k
)
6284 path_list
= (map(_lower
, variadic(path
)) for path
in path_list
)
6286 def _traverse_obj(obj
, path
, _current_depth
=0):
6290 path
= tuple(variadic(path
))
6291 for i
, key
in enumerate(path
):
6292 if isinstance(key
, (list, tuple)):
6293 obj
= [_traverse_obj(obj
, sub_key
, _current_depth
) for sub_key
in key
]
6296 obj
= (obj
.values() if isinstance(obj
, dict)
6297 else obj
if isinstance(obj
, (list, tuple, LazyList
))
6298 else str(obj
) if traverse_string
else [])
6300 depth
= max(depth
, _current_depth
)
6301 return [_traverse_obj(inner_obj
, path
[i
+ 1:], _current_depth
) for inner_obj
in obj
]
6302 elif isinstance(obj
, dict) and not (is_user_input
and key
== ':'):
6303 obj
= (obj
.get(key
) if casesense
or (key
in obj
)
6304 else next((v
for k
, v
in obj
.items() if _lower(k
) == key
), None))
6307 key
= (int_or_none(key
) if ':' not in key
6308 else slice(*map(int_or_none
, key
.split(':'))))
6309 if key
== slice(None):
6310 return _traverse_obj(obj
, (..., *path
[i
+ 1:]), _current_depth
)
6311 if not isinstance(key
, (int, slice)):
6313 if not isinstance(obj
, (list, tuple, LazyList
)):
6314 if not traverse_string
:
6323 if isinstance(expected_type
, type):
6324 type_test
= lambda val
: val
if isinstance(val
, expected_type
) else None
6325 elif expected_type
is not None:
6326 type_test
= expected_type
6328 type_test
= lambda val
: val
6330 for path
in path_list
:
6332 val
= _traverse_obj(obj
, path
)
6335 for _
in range(depth
- 1):
6336 val
= itertools
.chain
.from_iterable(v
for v
in val
if v
is not None)
6337 val
= [v
for v
in map(type_test
, val
) if v
is not None]
6339 return val
if get_all
else val
[0]
6341 val
= type_test(val
)
6347 def traverse_dict(dictn
, keys
, casesense
=True):
6348 ''' For backward compatibility. Do not use '''
6349 return traverse_obj(dictn
, keys
, casesense
=casesense
,
6350 is_user_input
=True, traverse_string
=True)
6353 def variadic(x
, allowed_types
=(str, bytes)):
6354 return x
if isinstance(x
, collections
.abc
.Iterable
) and not isinstance(x
, allowed_types
) else (x
,)
projects / yt-dlp.git / blob