4 from __future__
import unicode_literals
39 import xml
.etree
.ElementTree
43 compat_HTMLParseError
,
49 compat_ctypes_WINFUNCTYPE
,
50 compat_etree_fromstring
,
53 compat_html_entities_html5
,
66 compat_urllib_parse_urlencode
,
67 compat_urllib_parse_urlparse
,
68 compat_urllib_parse_urlunparse
,
69 compat_urllib_parse_quote
,
70 compat_urllib_parse_quote_plus
,
71 compat_urllib_parse_unquote_plus
,
72 compat_urllib_request
,
83 def register_socks_protocols():
84 # "Register" SOCKS protocols
85 # In Python < 2.6.5, urlsplit() suffers from bug https://bugs.python.org/issue7904
86 # URLs with protocols not in urlparse.uses_netloc are not handled correctly
87 for scheme
in ('socks', 'socks4', 'socks4a', 'socks5'):
88 if scheme
not in compat_urlparse
.uses_netloc
:
89 compat_urlparse
.uses_netloc
.append(scheme
)
92 # This is not clearly defined otherwise
93 compiled_regex_type
= type(re
.compile(''))
96 def random_user_agent():
97 _USER_AGENT_TPL
= 'Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/%s Safari/537.36'
1676 return _USER_AGENT_TPL
% random
.choice(_CHROME_VERSIONS
)
1680 'User-Agent': random_user_agent(),
1681 'Accept-Charset': 'ISO-8859-1,utf-8;q=0.7,*;q=0.7',
1682 'Accept': 'text/html,application/xhtml+xml,application/xml;q=0.9,*/*;q=0.8',
1683 'Accept-Encoding': 'gzip, deflate',
1684 'Accept-Language': 'en-us,en;q=0.5',
1689 'Safari': 'Mozilla/5.0 (X11; Linux x86_64; rv:10.0) AppleWebKit/533.20.25 (KHTML, like Gecko) Version/5.0.4 Safari/533.20.27',
1693 NO_DEFAULT
= object()
1695 ENGLISH_MONTH_NAMES
= [
1696 'January', 'February', 'March', 'April', 'May', 'June',
1697 'July', 'August', 'September', 'October', 'November', 'December']
1700 'en': ENGLISH_MONTH_NAMES
,
1702 'janvier', 'février', 'mars', 'avril', 'mai', 'juin',
1703 'juillet', 'août', 'septembre', 'octobre', 'novembre', 'décembre'],
1706 KNOWN_EXTENSIONS
= (
1707 'mp4', 'm4a', 'm4p', 'm4b', 'm4r', 'm4v', 'aac',
1708 'flv', 'f4v', 'f4a', 'f4b',
1709 'webm', 'ogg', 'ogv', 'oga', 'ogx', 'spx', 'opus',
1710 'mkv', 'mka', 'mk3d',
1713 'asf', 'wmv', 'wma',
1719 'f4f', 'f4m', 'm3u8', 'smil')
1721 # needed for sanitizing filenames in restricted mode
1722 ACCENT_CHARS
= dict(zip('ÂÃÄÀÁÅÆÇÈÉÊËÌÍÎÏÐÑÒÓÔÕÖŐØŒÙÚÛÜŰÝÞßàáâãäåæçèéêëìíîïðñòóôõöőøœùúûüűýþÿ',
1723 itertools
.chain('AAAAAA', ['AE'], 'CEEEEIIIIDNOOOOOOO', ['OE'], 'UUUUUY', ['TH', 'ss'],
1724 'aaaaaa', ['ae'], 'ceeeeiiiionooooooo', ['oe'], 'uuuuuy', ['th'], 'y')))
1748 '%Y/%m/%d %H:%M:%S',
1752 '%Y-%m-%d %H:%M:%S',
1753 '%Y-%m-%d %H:%M:%S.%f',
1754 '%Y-%m-%d %H:%M:%S:%f',
1757 '%Y-%m-%dT%H:%M:%SZ',
1758 '%Y-%m-%dT%H:%M:%S.%fZ',
1759 '%Y-%m-%dT%H:%M:%S.%f0Z',
1760 '%Y-%m-%dT%H:%M:%S',
1761 '%Y-%m-%dT%H:%M:%S.%f',
1763 '%b %d %Y at %H:%M',
1764 '%b %d %Y at %H:%M:%S',
1765 '%B %d %Y at %H:%M',
1766 '%B %d %Y at %H:%M:%S',
1770 DATE_FORMATS_DAY_FIRST
= list(DATE_FORMATS
)
1771 DATE_FORMATS_DAY_FIRST
.extend([
1777 '%d/%m/%Y %H:%M:%S',
1780 DATE_FORMATS_MONTH_FIRST
= list(DATE_FORMATS
)
1781 DATE_FORMATS_MONTH_FIRST
.extend([
1786 '%m/%d/%Y %H:%M:%S',
1789 PACKED_CODES_RE
= r
"}\('(.+)',(\d+),(\d+),'([^']+)'\.split\('\|'\)"
1790 JSON_LD_RE
= r
'(?is)<script[^>]+type=(["\']?
)application
/ld\
+json\
1[^
>]*>(?P
<json_ld
>.+?
)</script
>'
1793 def preferredencoding():
1794 """Get preferred encoding.
1796 Returns the best encoding scheme for the system, based on
1797 locale.getpreferredencoding() and some further tweaks.
1800 pref = locale.getpreferredencoding()
1808 def write_json_file(obj, fn):
1809 """ Encode obj as JSON and write it to fn, atomically if possible """
1811 fn = encodeFilename(fn)
1812 if sys.version_info < (3, 0) and sys.platform != 'win32
':
1813 encoding = get_filesystem_encoding()
1814 # os.path.basename returns a bytes object, but NamedTemporaryFile
1815 # will fail if the filename contains non ascii characters unless we
1816 # use a unicode object
1817 path_basename = lambda f: os.path.basename(fn).decode(encoding)
1818 # the same for os.path.dirname
1819 path_dirname = lambda f: os.path.dirname(fn).decode(encoding)
1821 path_basename = os.path.basename
1822 path_dirname = os.path.dirname
1826 'prefix
': path_basename(fn) + '.',
1827 'dir': path_dirname(fn),
1831 # In Python 2.x, json.dump expects a bytestream.
1832 # In Python 3.x, it writes to a character stream
1833 if sys.version_info < (3, 0):
1838 'encoding
': 'utf
-8',
1841 tf = tempfile.NamedTemporaryFile(**compat_kwargs(args))
1846 if sys.platform == 'win32
':
1847 # Need to remove existing file on Windows, else os.rename raises
1848 # WindowsError or FileExistsError.
1856 os.chmod(tf.name, 0o666 & ~mask)
1859 os.rename(tf.name, fn)
1868 if sys.version_info >= (2, 7):
1869 def find_xpath_attr(node, xpath, key, val=None):
1870 """ Find the xpath xpath[@key=val] """
1871 assert re.match(r'^
[a
-zA
-Z_
-]+$
', key)
1872 expr = xpath + ('[@%s]' % key if val is None else "[@%s='%s']" % (key, val))
1873 return node.find(expr)
1875 def find_xpath_attr(node, xpath, key, val=None):
1876 for f in node.findall(compat_xpath(xpath)):
1877 if key not in f.attrib:
1879 if val is None or f.attrib.get(key) == val:
1883 # On python2.6 the xml.etree.ElementTree.Element methods don't support
1884 # the namespace parameter
1887 def xpath_with_ns(path
, ns_map
):
1888 components
= [c
.split(':') for c
in path
.split('/')]
1890 for c
in components
:
1892 replaced
.append(c
[0])
1895 replaced
.append('{%s}%s' % (ns_map
[ns
], tag
))
1896 return '/'.join(replaced
)
1899 def xpath_element(node
, xpath
, name
=None, fatal
=False, default
=NO_DEFAULT
):
1900 def _find_xpath(xpath
):
1901 return node
.find(compat_xpath(xpath
))
1903 if isinstance(xpath
, (str, compat_str
)):
1904 n
= _find_xpath(xpath
)
1912 if default
is not NO_DEFAULT
:
1915 name
= xpath
if name
is None else name
1916 raise ExtractorError('Could not find XML element %s' % name
)
1922 def xpath_text(node
, xpath
, name
=None, fatal
=False, default
=NO_DEFAULT
):
1923 n
= xpath_element(node
, xpath
, name
, fatal
=fatal
, default
=default
)
1924 if n
is None or n
== default
:
1927 if default
is not NO_DEFAULT
:
1930 name
= xpath
if name
is None else name
1931 raise ExtractorError('Could not find XML element\'s text %s' % name
)
1937 def xpath_attr(node
, xpath
, key
, name
=None, fatal
=False, default
=NO_DEFAULT
):
1938 n
= find_xpath_attr(node
, xpath
, key
)
1940 if default
is not NO_DEFAULT
:
1943 name
= '%s[@%s]' % (xpath
, key
) if name
is None else name
1944 raise ExtractorError('Could not find XML attribute %s' % name
)
1947 return n
.attrib
[key
]
1950 def get_element_by_id(id, html
):
1951 """Return the content of the tag with the specified ID in the passed HTML document"""
1952 return get_element_by_attribute('id', id, html
)
1955 def get_element_by_class(class_name
, html
):
1956 """Return the content of the first tag with the specified class in the passed HTML document"""
1957 retval
= get_elements_by_class(class_name
, html
)
1958 return retval
[0] if retval
else None
1961 def get_element_by_attribute(attribute
, value
, html
, escape_value
=True):
1962 retval
= get_elements_by_attribute(attribute
, value
, html
, escape_value
)
1963 return retval
[0] if retval
else None
1966 def get_elements_by_class(class_name
, html
):
1967 """Return the content of all tags with the specified class in the passed HTML document as a list"""
1968 return get_elements_by_attribute(
1969 'class', r
'[^\'"]*\b%s\b[^\'"]*' % re.escape(class_name),
1970 html, escape_value=False)
1973 def get_elements_by_attribute(attribute, value, html, escape_value=True):
1974 """Return the content of the tag with the specified attribute in the passed HTML document"""
1976 value = re.escape(value) if escape_value else value
1979 for m in re.finditer(r'''(?xs)
1981 (?:\s+[a-zA-Z0-9:._-]+(?:=[a-zA-Z0-9:._-]*|="[^"]*"|='[^
']*'|
))*?
1983 (?:\s+[a-zA-Z0-9:._-]+(?:=[a-zA-Z0-9:._-]*|="[^
"]*"|
='[^']*'|))*?
1987 ''' % (re.escape(attribute), value), html):
1988 res = m.group('content
')
1990 if res.startswith('"') or res.startswith("'"):
1993 retlist.append(unescapeHTML(res))
1998 class HTMLAttributeParser(compat_HTMLParser):
1999 """Trivial HTML parser to gather the attributes for a single element"""
2003 compat_HTMLParser.__init__(self)
2005 def handle_starttag(self, tag, attrs):
2006 self.attrs = dict(attrs)
2009 def extract_attributes(html_element):
2010 """Given a string for an HTML element such as
2012 a="foo" B="bar" c="&98;az" d=boz
2013 empty= noval entity="&"
2016 Decode and return a dictionary of attributes.
2018 'a
': 'foo
', 'b
': 'bar
', c: 'baz
', d: 'boz
',
2019 'empty
': '', 'noval
': None, 'entity
': '&',
2020 'sq
': '"', 'dq': '\''
2022 NB HTMLParser is stricter in Python 2.6 & 3.2 than in later versions,
2023 but the cases in the unit test will work for all of 2.6, 2.7, 3.2-3.5.
2025 parser = HTMLAttributeParser()
2027 parser.feed(html_element)
2029 # Older Python may throw HTMLParseError in case of malformed HTML
2030 except compat_HTMLParseError:
2035 def clean_html(html):
2036 """Clean an HTML snippet into a readable string"""
2038 if html is None: # Convenience for sanitizing descriptions etc.
2042 html = html.replace('\n', ' ')
2043 html = re.sub(r'(?u)\s*<\s*br\s*/?\s*>\s*', '\n', html)
2044 html = re.sub(r'(?u)<\s*/\s*p\s*>\s*<\s*p[^>]*>', '\n', html)
2046 html = re.sub('<.*?>', '', html)
2047 # Replace html entities
2048 html = unescapeHTML(html)
2052 def sanitize_open(filename, open_mode):
2053 """Try to open the given filename, and slightly tweak it if this fails.
2055 Attempts to open the given filename. If this fails, it tries to change
2056 the filename slightly, step by step, until it's either able to open it
2057 or it fails and raises a final exception, like the standard open()
2060 It returns the tuple (stream, definitive_file_name).
2064 if sys.platform == 'win32':
2066 msvcrt.setmode(sys.stdout.fileno(), os.O_BINARY)
2067 return (sys.stdout.buffer if hasattr(sys.stdout, 'buffer') else sys.stdout, filename)
2068 stream = open(encodeFilename(filename), open_mode)
2069 return (stream, filename)
2070 except (IOError, OSError) as err:
2071 if err.errno in (errno.EACCES,):
2074 # In case of error, try to remove win32 forbidden chars
2075 alt_filename = sanitize_path(filename)
2076 if alt_filename == filename:
2079 # An exception here should be caught in the caller
2080 stream = open(encodeFilename(alt_filename), open_mode)
2081 return (stream, alt_filename)
2084 def timeconvert(timestr):
2085 """Convert RFC 2822 defined time string into system timestamp"""
2087 timetuple = email.utils.parsedate_tz(timestr)
2088 if timetuple is not None:
2089 timestamp = email.utils.mktime_tz(timetuple)
2093 def sanitize_filename(s, restricted=False, is_id=False):
2094 """Sanitizes a string so it could be used as part of a filename.
2095 If restricted is set, use a stricter subset of allowed characters.
2096 Set is_id if this is not an arbitrary string, but an ID that should be kept
2099 def replace_insane(char):
2100 if restricted and char in ACCENT_CHARS:
2101 return ACCENT_CHARS[char]
2102 elif not restricted and char == '\n':
2104 elif char == '?' or ord(char) < 32 or ord(char) == 127:
2107 return '' if restricted else '\''
2109 return '_
-' if restricted else ' -'
2110 elif char in '\\/|
*<>':
2112 if restricted and (char in '!&\'()[]{}$
;`^
,#' or char.isspace()):
2114 if restricted
and ord(char
) > 127:
2121 s
= re
.sub(r
'[0-9]+(?::[0-9]+)+', lambda m
: m
.group(0).replace(':', '_'), s
)
2122 result
= ''.join(map(replace_insane
, s
))
2124 while '__' in result
:
2125 result
= result
.replace('__', '_')
2126 result
= result
.strip('_')
2127 # Common case of "Foreign band name - English song title"
2128 if restricted
and result
.startswith('-_'):
2130 if result
.startswith('-'):
2131 result
= '_' + result
[len('-'):]
2132 result
= result
.lstrip('.')
2138 def sanitize_path(s
, force
=False):
2139 """Sanitizes and normalizes path on Windows"""
2140 if sys
.platform
== 'win32':
2142 drive_or_unc
, _
= os
.path
.splitdrive(s
)
2143 if sys
.version_info
< (2, 7) and not drive_or_unc
:
2144 drive_or_unc
, _
= os
.path
.splitunc(s
)
2150 norm_path
= os
.path
.normpath(remove_start(s
, drive_or_unc
)).split(os
.path
.sep
)
2154 path_part
if path_part
in ['.', '..'] else re
.sub(r
'(?:[/<>:"\|\\?\*]|[\s.]$)', '#', path_part
)
2155 for path_part
in norm_path
]
2157 sanitized_path
.insert(0, drive_or_unc
+ os
.path
.sep
)
2158 elif force
and s
[0] == os
.path
.sep
:
2159 sanitized_path
.insert(0, os
.path
.sep
)
2160 return os
.path
.join(*sanitized_path
)
2163 def sanitize_url(url
):
2164 # Prepend protocol-less URLs with `http:` scheme in order to mitigate
2165 # the number of unwanted failures due to missing protocol
2166 if url
.startswith('//'):
2167 return 'http:%s' % url
2168 # Fix some common typos seen so far
2170 # https://github.com/ytdl-org/youtube-dl/issues/15649
2171 (r
'^httpss://', r
'https://'),
2172 # https://bx1.be/lives/direct-tv/
2173 (r
'^rmtp([es]?)://', r
'rtmp\1://'),
2175 for mistake
, fixup
in COMMON_TYPOS
:
2176 if re
.match(mistake
, url
):
2177 return re
.sub(mistake
, fixup
, url
)
2181 def extract_basic_auth(url
):
2182 parts
= compat_urlparse
.urlsplit(url
)
2183 if parts
.username
is None:
2185 url
= compat_urlparse
.urlunsplit(parts
._replace
(netloc
=(
2186 parts
.hostname
if parts
.port
is None
2187 else '%s:%d' % (parts
.hostname
, parts
.port
))))
2188 auth_payload
= base64
.b64encode(
2189 ('%s:%s' % (parts
.username
, parts
.password
or '')).encode('utf-8'))
2190 return url
, 'Basic ' + auth_payload
.decode('utf-8')
2193 def sanitized_Request(url
, *args
, **kwargs
):
2194 url
, auth_header
= extract_basic_auth(escape_url(sanitize_url(url
)))
2195 if auth_header
is not None:
2196 headers
= args
[1] if len(args
) >= 2 else kwargs
.setdefault('headers', {})
2197 headers
['Authorization'] = auth_header
2198 return compat_urllib_request
.Request(url
, *args
, **kwargs
)
2202 """Expand shell variables and ~"""
2203 return os
.path
.expandvars(compat_expanduser(s
))
2206 def orderedSet(iterable
):
2207 """ Remove all duplicates from the input iterable """
2215 def _htmlentity_transform(entity_with_semicolon
):
2216 """Transforms an HTML entity to a character."""
2217 entity
= entity_with_semicolon
[:-1]
2219 # Known non-numeric HTML entity
2220 if entity
in compat_html_entities
.name2codepoint
:
2221 return compat_chr(compat_html_entities
.name2codepoint
[entity
])
2223 # TODO: HTML5 allows entities without a semicolon. For example,
2224 # 'Éric' should be decoded as 'Éric'.
2225 if entity_with_semicolon
in compat_html_entities_html5
:
2226 return compat_html_entities_html5
[entity_with_semicolon
]
2228 mobj
= re
.match(r
'#(x[0-9a-fA-F]+|[0-9]+)', entity
)
2229 if mobj
is not None:
2230 numstr
= mobj
.group(1)
2231 if numstr
.startswith('x'):
2233 numstr
= '0%s' % numstr
2236 # See https://github.com/ytdl-org/youtube-dl/issues/7518
2238 return compat_chr(int(numstr
, base
))
2242 # Unknown entity in name, return its literal representation
2243 return '&%s;' % entity
2246 def unescapeHTML(s
):
2249 assert type(s
) == compat_str
2252 r
'&([^&;]+;)', lambda m
: _htmlentity_transform(m
.group(1)), s
)
2255 def escapeHTML(text
):
2258 .replace('&', '&')
2259 .replace('<', '<')
2260 .replace('>', '>')
2261 .replace('"', '"')
2262 .replace("'", ''')
2266 def process_communicate_or_kill(p
, *args
, **kwargs
):
2268 return p
.communicate(*args
, **kwargs
)
2269 except BaseException
: # Including KeyboardInterrupt
2275 def get_subprocess_encoding():
2276 if sys
.platform
== 'win32' and sys
.getwindowsversion()[0] >= 5:
2277 # For subprocess calls, encode with locale encoding
2278 # Refer to http://stackoverflow.com/a/9951851/35070
2279 encoding
= preferredencoding()
2281 encoding
= sys
.getfilesystemencoding()
2282 if encoding
is None:
2287 def encodeFilename(s
, for_subprocess
=False):
2289 @param s The name of the file
2292 assert type(s
) == compat_str
2294 # Python 3 has a Unicode API
2295 if sys
.version_info
>= (3, 0):
2298 # Pass '' directly to use Unicode APIs on Windows 2000 and up
2299 # (Detecting Windows NT 4 is tricky because 'major >= 4' would
2300 # match Windows 9x series as well. Besides, NT 4 is obsolete.)
2301 if not for_subprocess
and sys
.platform
== 'win32' and sys
.getwindowsversion()[0] >= 5:
2304 # Jython assumes filenames are Unicode strings though reported as Python 2.x compatible
2305 if sys
.platform
.startswith('java'):
2308 return s
.encode(get_subprocess_encoding(), 'ignore')
2311 def decodeFilename(b
, for_subprocess
=False):
2313 if sys
.version_info
>= (3, 0):
2316 if not isinstance(b
, bytes):
2319 return b
.decode(get_subprocess_encoding(), 'ignore')
2322 def encodeArgument(s
):
2323 if not isinstance(s
, compat_str
):
2324 # Legacy code that uses byte strings
2325 # Uncomment the following line after fixing all post processors
2326 # assert False, 'Internal error: %r should be of type %r, is %r' % (s, compat_str, type(s))
2327 s
= s
.decode('ascii')
2328 return encodeFilename(s
, True)
2331 def decodeArgument(b
):
2332 return decodeFilename(b
, True)
2335 def decodeOption(optval
):
2338 if isinstance(optval
, bytes):
2339 optval
= optval
.decode(preferredencoding())
2341 assert isinstance(optval
, compat_str
)
2345 def formatSeconds(secs
, delim
=':', msec
=False):
2347 ret
= '%d%s%02d%s%02d' % (secs
// 3600, delim
, (secs
% 3600) // 60, delim
, secs
% 60)
2349 ret
= '%d%s%02d' % (secs
// 60, delim
, secs
% 60)
2352 return '%s.%03d' % (ret
, secs
% 1) if msec
else ret
2355 def make_HTTPS_handler(params
, **kwargs
):
2356 opts_no_check_certificate
= params
.get('nocheckcertificate', False)
2357 if hasattr(ssl
, 'create_default_context'): # Python >= 3.4 or 2.7.9
2358 context
= ssl
.create_default_context(ssl
.Purpose
.SERVER_AUTH
)
2359 if opts_no_check_certificate
:
2360 context
.check_hostname
= False
2361 context
.verify_mode
= ssl
.CERT_NONE
2363 return YoutubeDLHTTPSHandler(params
, context
=context
, **kwargs
)
2366 # (create_default_context present but HTTPSHandler has no context=)
2369 if sys
.version_info
< (3, 2):
2370 return YoutubeDLHTTPSHandler(params
, **kwargs
)
2371 else: # Python < 3.4
2372 context
= ssl
.SSLContext(ssl
.PROTOCOL_TLSv1
)
2373 context
.verify_mode
= (ssl
.CERT_NONE
2374 if opts_no_check_certificate
2375 else ssl
.CERT_REQUIRED
)
2376 context
.set_default_verify_paths()
2377 return YoutubeDLHTTPSHandler(params
, context
=context
, **kwargs
)
2380 def bug_reports_message(before
=';'):
2381 if ytdl_is_updateable():
2382 update_cmd
= 'type yt-dlp -U to update'
2384 update_cmd
= 'see https://github.com/yt-dlp/yt-dlp on how to update'
2385 msg
= 'please report this issue on https://github.com/yt-dlp/yt-dlp .'
2386 msg
+= ' Make sure you are using the latest version; %s.' % update_cmd
2387 msg
+= ' Be sure to call yt-dlp with the --verbose flag and include its complete output.'
2389 before
= before
.rstrip()
2390 if not before
or before
.endswith(('.', '!', '?')):
2391 msg
= msg
[0].title() + msg
[1:]
2393 return (before
+ ' ' if before
else '') + msg
2396 class YoutubeDLError(Exception):
2397 """Base exception for YoutubeDL errors."""
2401 network_exceptions
= [compat_urllib_error
.URLError
, compat_http_client
.HTTPException
, socket
.error
]
2402 if hasattr(ssl
, 'CertificateError'):
2403 network_exceptions
.append(ssl
.CertificateError
)
2404 network_exceptions
= tuple(network_exceptions
)
2407 class ExtractorError(YoutubeDLError
):
2408 """Error during info extraction."""
2410 def __init__(self
, msg
, tb
=None, expected
=False, cause
=None, video_id
=None, ie
=None):
2411 """ tb, if given, is the original traceback (so that it can be printed out).
2412 If expected is set, this is a normal error message and most likely not a bug in yt-dlp.
2414 if sys
.exc_info()[0] in network_exceptions
:
2419 self
.expected
= expected
2421 self
.video_id
= video_id
2423 self
.exc_info
= sys
.exc_info() # preserve original exception
2425 super(ExtractorError
, self
).__init
__(''.join((
2426 format_field(ie
, template
='[%s] '),
2427 format_field(video_id
, template
='%s: '),
2429 format_field(cause
, template
=' (caused by %r)'),
2430 '' if expected
else bug_reports_message())))
2432 def format_traceback(self
):
2433 if self
.traceback
is None:
2435 return ''.join(traceback
.format_tb(self
.traceback
))
2438 class UnsupportedError(ExtractorError
):
2439 def __init__(self
, url
):
2440 super(UnsupportedError
, self
).__init
__(
2441 'Unsupported URL: %s' % url
, expected
=True)
2445 class RegexNotFoundError(ExtractorError
):
2446 """Error when a regex didn't match"""
2450 class GeoRestrictedError(ExtractorError
):
2451 """Geographic restriction Error exception.
2453 This exception may be thrown when a video is not available from your
2454 geographic location due to geographic restrictions imposed by a website.
2457 def __init__(self
, msg
, countries
=None):
2458 super(GeoRestrictedError
, self
).__init
__(msg
, expected
=True)
2460 self
.countries
= countries
2463 class DownloadError(YoutubeDLError
):
2464 """Download Error exception.
2466 This exception may be thrown by FileDownloader objects if they are not
2467 configured to continue on errors. They will contain the appropriate
2471 def __init__(self
, msg
, exc_info
=None):
2472 """ exc_info, if given, is the original exception that caused the trouble (as returned by sys.exc_info()). """
2473 super(DownloadError
, self
).__init
__(msg
)
2474 self
.exc_info
= exc_info
2477 class EntryNotInPlaylist(YoutubeDLError
):
2478 """Entry not in playlist exception.
2480 This exception will be thrown by YoutubeDL when a requested entry
2481 is not found in the playlist info_dict
2486 class SameFileError(YoutubeDLError
):
2487 """Same File exception.
2489 This exception will be thrown by FileDownloader objects if they detect
2490 multiple files would have to be downloaded to the same file on disk.
2495 class PostProcessingError(YoutubeDLError
):
2496 """Post Processing exception.
2498 This exception may be raised by PostProcessor's .run() method to
2499 indicate an error in the postprocessing task.
2502 def __init__(self
, msg
):
2503 super(PostProcessingError
, self
).__init
__(msg
)
2507 class ExistingVideoReached(YoutubeDLError
):
2508 """ --max-downloads limit has been reached. """
2512 class RejectedVideoReached(YoutubeDLError
):
2513 """ --max-downloads limit has been reached. """
2517 class ThrottledDownload(YoutubeDLError
):
2518 """ Download speed below --throttled-rate. """
2522 class MaxDownloadsReached(YoutubeDLError
):
2523 """ --max-downloads limit has been reached. """
2527 class UnavailableVideoError(YoutubeDLError
):
2528 """Unavailable Format exception.
2530 This exception will be thrown when a video is requested
2531 in a format that is not available for that video.
2536 class ContentTooShortError(YoutubeDLError
):
2537 """Content Too Short exception.
2539 This exception may be raised by FileDownloader objects when a file they
2540 download is too small for what the server announced first, indicating
2541 the connection was probably interrupted.
2544 def __init__(self
, downloaded
, expected
):
2545 super(ContentTooShortError
, self
).__init
__(
2546 'Downloaded {0} bytes, expected {1} bytes'.format(downloaded
, expected
)
2549 self
.downloaded
= downloaded
2550 self
.expected
= expected
2553 class XAttrMetadataError(YoutubeDLError
):
2554 def __init__(self
, code
=None, msg
='Unknown error'):
2555 super(XAttrMetadataError
, self
).__init
__(msg
)
2559 # Parsing code and msg
2560 if (self
.code
in (errno
.ENOSPC
, errno
.EDQUOT
)
2561 or 'No space left' in self
.msg
or 'Disk quota exceeded' in self
.msg
):
2562 self
.reason
= 'NO_SPACE'
2563 elif self
.code
== errno
.E2BIG
or 'Argument list too long' in self
.msg
:
2564 self
.reason
= 'VALUE_TOO_LONG'
2566 self
.reason
= 'NOT_SUPPORTED'
2569 class XAttrUnavailableError(YoutubeDLError
):
2573 def _create_http_connection(ydl_handler
, http_class
, is_https
, *args
, **kwargs
):
2574 # Working around python 2 bug (see http://bugs.python.org/issue17849) by limiting
2575 # expected HTTP responses to meet HTTP/1.0 or later (see also
2576 # https://github.com/ytdl-org/youtube-dl/issues/6727)
2577 if sys
.version_info
< (3, 0):
2578 kwargs
['strict'] = True
2579 hc
= http_class(*args
, **compat_kwargs(kwargs
))
2580 source_address
= ydl_handler
._params
.get('source_address')
2582 if source_address
is not None:
2583 # This is to workaround _create_connection() from socket where it will try all
2584 # address data from getaddrinfo() including IPv6. This filters the result from
2585 # getaddrinfo() based on the source_address value.
2586 # This is based on the cpython socket.create_connection() function.
2587 # https://github.com/python/cpython/blob/master/Lib/socket.py#L691
2588 def _create_connection(address
, timeout
=socket
._GLOBAL
_DEFAULT
_TIMEOUT
, source_address
=None):
2589 host
, port
= address
2591 addrs
= socket
.getaddrinfo(host
, port
, 0, socket
.SOCK_STREAM
)
2592 af
= socket
.AF_INET
if '.' in source_address
[0] else socket
.AF_INET6
2593 ip_addrs
= [addr
for addr
in addrs
if addr
[0] == af
]
2594 if addrs
and not ip_addrs
:
2595 ip_version
= 'v4' if af
== socket
.AF_INET
else 'v6'
2597 "No remote IP%s addresses available for connect, can't use '%s' as source address"
2598 % (ip_version
, source_address
[0]))
2599 for res
in ip_addrs
:
2600 af
, socktype
, proto
, canonname
, sa
= res
2603 sock
= socket
.socket(af
, socktype
, proto
)
2604 if timeout
is not socket
._GLOBAL
_DEFAULT
_TIMEOUT
:
2605 sock
.settimeout(timeout
)
2606 sock
.bind(source_address
)
2608 err
= None # Explicitly break reference cycle
2610 except socket
.error
as _
:
2612 if sock
is not None:
2617 raise socket
.error('getaddrinfo returns an empty list')
2618 if hasattr(hc
, '_create_connection'):
2619 hc
._create
_connection
= _create_connection
2620 sa
= (source_address
, 0)
2621 if hasattr(hc
, 'source_address'): # Python 2.7+
2622 hc
.source_address
= sa
2624 def _hc_connect(self
, *args
, **kwargs
):
2625 sock
= _create_connection(
2626 (self
.host
, self
.port
), self
.timeout
, sa
)
2628 self
.sock
= ssl
.wrap_socket(
2629 sock
, self
.key_file
, self
.cert_file
,
2630 ssl_version
=ssl
.PROTOCOL_TLSv1
)
2633 hc
.connect
= functools
.partial(_hc_connect
, hc
)
2638 def handle_youtubedl_headers(headers
):
2639 filtered_headers
= headers
2641 if 'Youtubedl-no-compression' in filtered_headers
:
2642 filtered_headers
= dict((k
, v
) for k
, v
in filtered_headers
.items() if k
.lower() != 'accept-encoding')
2643 del filtered_headers
['Youtubedl-no-compression']
2645 return filtered_headers
2648 class YoutubeDLHandler(compat_urllib_request
.HTTPHandler
):
2649 """Handler for HTTP requests and responses.
2651 This class, when installed with an OpenerDirector, automatically adds
2652 the standard headers to every HTTP request and handles gzipped and
2653 deflated responses from web servers. If compression is to be avoided in
2654 a particular request, the original request in the program code only has
2655 to include the HTTP header "Youtubedl-no-compression", which will be
2656 removed before making the real request.
2658 Part of this code was copied from:
2660 http://techknack.net/python-urllib2-handlers/
2662 Andrew Rowls, the author of that code, agreed to release it to the
2666 def __init__(self
, params
, *args
, **kwargs
):
2667 compat_urllib_request
.HTTPHandler
.__init
__(self
, *args
, **kwargs
)
2668 self
._params
= params
2670 def http_open(self
, req
):
2671 conn_class
= compat_http_client
.HTTPConnection
2673 socks_proxy
= req
.headers
.get('Ytdl-socks-proxy')
2675 conn_class
= make_socks_conn_class(conn_class
, socks_proxy
)
2676 del req
.headers
['Ytdl-socks-proxy']
2678 return self
.do_open(functools
.partial(
2679 _create_http_connection
, self
, conn_class
, False),
2687 return zlib
.decompress(data
, -zlib
.MAX_WBITS
)
2689 return zlib
.decompress(data
)
2691 def http_request(self
, req
):
2692 # According to RFC 3986, URLs can not contain non-ASCII characters, however this is not
2693 # always respected by websites, some tend to give out URLs with non percent-encoded
2694 # non-ASCII characters (see telemb.py, ard.py [#3412])
2695 # urllib chokes on URLs with non-ASCII characters (see http://bugs.python.org/issue3991)
2696 # To work around aforementioned issue we will replace request's original URL with
2697 # percent-encoded one
2698 # Since redirects are also affected (e.g. http://www.southpark.de/alle-episoden/s18e09)
2699 # the code of this workaround has been moved here from YoutubeDL.urlopen()
2700 url
= req
.get_full_url()
2701 url_escaped
= escape_url(url
)
2703 # Substitute URL if any change after escaping
2704 if url
!= url_escaped
:
2705 req
= update_Request(req
, url
=url_escaped
)
2707 for h
, v
in std_headers
.items():
2708 # Capitalize is needed because of Python bug 2275: http://bugs.python.org/issue2275
2709 # The dict keys are capitalized because of this bug by urllib
2710 if h
.capitalize() not in req
.headers
:
2711 req
.add_header(h
, v
)
2713 req
.headers
= handle_youtubedl_headers(req
.headers
)
2715 if sys
.version_info
< (2, 7) and '#' in req
.get_full_url():
2716 # Python 2.6 is brain-dead when it comes to fragments
2717 req
._Request
__original
= req
._Request
__original
.partition('#')[0]
2718 req
._Request
__r
_type
= req
._Request
__r
_type
.partition('#')[0]
2722 def http_response(self
, req
, resp
):
2725 if resp
.headers
.get('Content-encoding', '') == 'gzip':
2726 content
= resp
.read()
2727 gz
= gzip
.GzipFile(fileobj
=io
.BytesIO(content
), mode
='rb')
2729 uncompressed
= io
.BytesIO(gz
.read())
2730 except IOError as original_ioerror
:
2731 # There may be junk add the end of the file
2732 # See http://stackoverflow.com/q/4928560/35070 for details
2733 for i
in range(1, 1024):
2735 gz
= gzip
.GzipFile(fileobj
=io
.BytesIO(content
[:-i
]), mode
='rb')
2736 uncompressed
= io
.BytesIO(gz
.read())
2741 raise original_ioerror
2742 resp
= compat_urllib_request
.addinfourl(uncompressed
, old_resp
.headers
, old_resp
.url
, old_resp
.code
)
2743 resp
.msg
= old_resp
.msg
2744 del resp
.headers
['Content-encoding']
2746 if resp
.headers
.get('Content-encoding', '') == 'deflate':
2747 gz
= io
.BytesIO(self
.deflate(resp
.read()))
2748 resp
= compat_urllib_request
.addinfourl(gz
, old_resp
.headers
, old_resp
.url
, old_resp
.code
)
2749 resp
.msg
= old_resp
.msg
2750 del resp
.headers
['Content-encoding']
2751 # Percent-encode redirect URL of Location HTTP header to satisfy RFC 3986 (see
2752 # https://github.com/ytdl-org/youtube-dl/issues/6457).
2753 if 300 <= resp
.code
< 400:
2754 location
= resp
.headers
.get('Location')
2756 # As of RFC 2616 default charset is iso-8859-1 that is respected by python 3
2757 if sys
.version_info
>= (3, 0):
2758 location
= location
.encode('iso-8859-1').decode('utf-8')
2760 location
= location
.decode('utf-8')
2761 location_escaped
= escape_url(location
)
2762 if location
!= location_escaped
:
2763 del resp
.headers
['Location']
2764 if sys
.version_info
< (3, 0):
2765 location_escaped
= location_escaped
.encode('utf-8')
2766 resp
.headers
['Location'] = location_escaped
2769 https_request
= http_request
2770 https_response
= http_response
2773 def make_socks_conn_class(base_class
, socks_proxy
):
2774 assert issubclass(base_class
, (
2775 compat_http_client
.HTTPConnection
, compat_http_client
.HTTPSConnection
))
2777 url_components
= compat_urlparse
.urlparse(socks_proxy
)
2778 if url_components
.scheme
.lower() == 'socks5':
2779 socks_type
= ProxyType
.SOCKS5
2780 elif url_components
.scheme
.lower() in ('socks', 'socks4'):
2781 socks_type
= ProxyType
.SOCKS4
2782 elif url_components
.scheme
.lower() == 'socks4a':
2783 socks_type
= ProxyType
.SOCKS4A
2785 def unquote_if_non_empty(s
):
2788 return compat_urllib_parse_unquote_plus(s
)
2792 url_components
.hostname
, url_components
.port
or 1080,
2794 unquote_if_non_empty(url_components
.username
),
2795 unquote_if_non_empty(url_components
.password
),
2798 class SocksConnection(base_class
):
2800 self
.sock
= sockssocket()
2801 self
.sock
.setproxy(*proxy_args
)
2802 if type(self
.timeout
) in (int, float):
2803 self
.sock
.settimeout(self
.timeout
)
2804 self
.sock
.connect((self
.host
, self
.port
))
2806 if isinstance(self
, compat_http_client
.HTTPSConnection
):
2807 if hasattr(self
, '_context'): # Python > 2.6
2808 self
.sock
= self
._context
.wrap_socket(
2809 self
.sock
, server_hostname
=self
.host
)
2811 self
.sock
= ssl
.wrap_socket(self
.sock
)
2813 return SocksConnection
2816 class YoutubeDLHTTPSHandler(compat_urllib_request
.HTTPSHandler
):
2817 def __init__(self
, params
, https_conn_class
=None, *args
, **kwargs
):
2818 compat_urllib_request
.HTTPSHandler
.__init
__(self
, *args
, **kwargs
)
2819 self
._https
_conn
_class
= https_conn_class
or compat_http_client
.HTTPSConnection
2820 self
._params
= params
2822 def https_open(self
, req
):
2824 conn_class
= self
._https
_conn
_class
2826 if hasattr(self
, '_context'): # python > 2.6
2827 kwargs
['context'] = self
._context
2828 if hasattr(self
, '_check_hostname'): # python 3.x
2829 kwargs
['check_hostname'] = self
._check
_hostname
2831 socks_proxy
= req
.headers
.get('Ytdl-socks-proxy')
2833 conn_class
= make_socks_conn_class(conn_class
, socks_proxy
)
2834 del req
.headers
['Ytdl-socks-proxy']
2836 return self
.do_open(functools
.partial(
2837 _create_http_connection
, self
, conn_class
, True),
2841 class YoutubeDLCookieJar(compat_cookiejar
.MozillaCookieJar
):
2843 See [1] for cookie file format.
2845 1. https://curl.haxx.se/docs/http-cookies.html
2847 _HTTPONLY_PREFIX
= '#HttpOnly_'
2849 _HEADER
= '''# Netscape HTTP Cookie File
2850 # This file is generated by yt-dlp. Do not edit.
2853 _CookieFileEntry
= collections
.namedtuple(
2855 ('domain_name', 'include_subdomains', 'path', 'https_only', 'expires_at', 'name', 'value'))
2857 def save(self
, filename
=None, ignore_discard
=False, ignore_expires
=False):
2859 Save cookies to a file.
2861 Most of the code is taken from CPython 3.8 and slightly adapted
2862 to support cookie files with UTF-8 in both python 2 and 3.
2864 if filename
is None:
2865 if self
.filename
is not None:
2866 filename
= self
.filename
2868 raise ValueError(compat_cookiejar
.MISSING_FILENAME_TEXT
)
2870 # Store session cookies with `expires` set to 0 instead of an empty
2873 if cookie
.expires
is None:
2876 with io
.open(filename
, 'w', encoding
='utf-8') as f
:
2877 f
.write(self
._HEADER
)
2880 if not ignore_discard
and cookie
.discard
:
2882 if not ignore_expires
and cookie
.is_expired(now
):
2888 if cookie
.domain
.startswith('.'):
2889 initial_dot
= 'TRUE'
2891 initial_dot
= 'FALSE'
2892 if cookie
.expires
is not None:
2893 expires
= compat_str(cookie
.expires
)
2896 if cookie
.value
is None:
2897 # cookies.txt regards 'Set-Cookie: foo' as a cookie
2898 # with no name, whereas http.cookiejar regards it as a
2899 # cookie with no value.
2904 value
= cookie
.value
2906 '\t'.join([cookie
.domain
, initial_dot
, cookie
.path
,
2907 secure
, expires
, name
, value
]) + '\n')
2909 def load(self
, filename
=None, ignore_discard
=False, ignore_expires
=False):
2910 """Load cookies from a file."""
2911 if filename
is None:
2912 if self
.filename
is not None:
2913 filename
= self
.filename
2915 raise ValueError(compat_cookiejar
.MISSING_FILENAME_TEXT
)
2917 def prepare_line(line
):
2918 if line
.startswith(self
._HTTPONLY
_PREFIX
):
2919 line
= line
[len(self
._HTTPONLY
_PREFIX
):]
2920 # comments and empty lines are fine
2921 if line
.startswith('#') or not line
.strip():
2923 cookie_list
= line
.split('\t')
2924 if len(cookie_list
) != self
._ENTRY
_LEN
:
2925 raise compat_cookiejar
.LoadError('invalid length %d' % len(cookie_list
))
2926 cookie
= self
._CookieFileEntry
(*cookie_list
)
2927 if cookie
.expires_at
and not cookie
.expires_at
.isdigit():
2928 raise compat_cookiejar
.LoadError('invalid expires at %s' % cookie
.expires_at
)
2932 with io
.open(filename
, encoding
='utf-8') as f
:
2935 cf
.write(prepare_line(line
))
2936 except compat_cookiejar
.LoadError
as e
:
2938 'WARNING: skipping cookie file entry due to %s: %r\n'
2939 % (e
, line
), sys
.stderr
)
2942 self
._really
_load
(cf
, filename
, ignore_discard
, ignore_expires
)
2943 # Session cookies are denoted by either `expires` field set to
2944 # an empty string or 0. MozillaCookieJar only recognizes the former
2945 # (see [1]). So we need force the latter to be recognized as session
2946 # cookies on our own.
2947 # Session cookies may be important for cookies-based authentication,
2948 # e.g. usually, when user does not check 'Remember me' check box while
2949 # logging in on a site, some important cookies are stored as session
2950 # cookies so that not recognizing them will result in failed login.
2951 # 1. https://bugs.python.org/issue17164
2953 # Treat `expires=0` cookies as session cookies
2954 if cookie
.expires
== 0:
2955 cookie
.expires
= None
2956 cookie
.discard
= True
2959 class YoutubeDLCookieProcessor(compat_urllib_request
.HTTPCookieProcessor
):
2960 def __init__(self
, cookiejar
=None):
2961 compat_urllib_request
.HTTPCookieProcessor
.__init
__(self
, cookiejar
)
2963 def http_response(self
, request
, response
):
2964 # Python 2 will choke on next HTTP request in row if there are non-ASCII
2965 # characters in Set-Cookie HTTP header of last response (see
2966 # https://github.com/ytdl-org/youtube-dl/issues/6769).
2967 # In order to at least prevent crashing we will percent encode Set-Cookie
2968 # header before HTTPCookieProcessor starts processing it.
2969 # if sys.version_info < (3, 0) and response.headers:
2970 # for set_cookie_header in ('Set-Cookie', 'Set-Cookie2'):
2971 # set_cookie = response.headers.get(set_cookie_header)
2973 # set_cookie_escaped = compat_urllib_parse.quote(set_cookie, b"%/;:@&=+$,!~*'()?#[] ")
2974 # if set_cookie != set_cookie_escaped:
2975 # del response.headers[set_cookie_header]
2976 # response.headers[set_cookie_header] = set_cookie_escaped
2977 return compat_urllib_request
.HTTPCookieProcessor
.http_response(self
, request
, response
)
2979 https_request
= compat_urllib_request
.HTTPCookieProcessor
.http_request
2980 https_response
= http_response
2983 class YoutubeDLRedirectHandler(compat_urllib_request
.HTTPRedirectHandler
):
2984 """YoutubeDL redirect handler
2986 The code is based on HTTPRedirectHandler implementation from CPython [1].
2988 This redirect handler solves two issues:
2989 - ensures redirect URL is always unicode under python 2
2990 - introduces support for experimental HTTP response status code
2991 308 Permanent Redirect [2] used by some sites [3]
2993 1. https://github.com/python/cpython/blob/master/Lib/urllib/request.py
2994 2. https://developer.mozilla.org/en-US/docs/Web/HTTP/Status/308
2995 3. https://github.com/ytdl-org/youtube-dl/issues/28768
2998 http_error_301
= http_error_303
= http_error_307
= http_error_308
= compat_urllib_request
.HTTPRedirectHandler
.http_error_302
3000 def redirect_request(self
, req
, fp
, code
, msg
, headers
, newurl
):
3001 """Return a Request or None in response to a redirect.
3003 This is called by the http_error_30x methods when a
3004 redirection response is received. If a redirection should
3005 take place, return a new Request to allow http_error_30x to
3006 perform the redirect. Otherwise, raise HTTPError if no-one
3007 else should try to handle this url. Return None if you can't
3008 but another Handler might.
3010 m
= req
.get_method()
3011 if (not (code
in (301, 302, 303, 307, 308) and m
in ("GET", "HEAD")
3012 or code
in (301, 302, 303) and m
== "POST")):
3013 raise compat_HTTPError(req
.full_url
, code
, msg
, headers
, fp
)
3014 # Strictly (according to RFC 2616), 301 or 302 in response to
3015 # a POST MUST NOT cause a redirection without confirmation
3016 # from the user (of urllib.request, in this case). In practice,
3017 # essentially all clients do redirect in this case, so we do
3020 # On python 2 urlh.geturl() may sometimes return redirect URL
3021 # as byte string instead of unicode. This workaround allows
3022 # to force it always return unicode.
3023 if sys
.version_info
[0] < 3:
3024 newurl
= compat_str(newurl
)
3026 # Be conciliant with URIs containing a space. This is mainly
3027 # redundant with the more complete encoding done in http_error_302(),
3028 # but it is kept for compatibility with other callers.
3029 newurl
= newurl
.replace(' ', '%20')
3031 CONTENT_HEADERS
= ("content-length", "content-type")
3032 # NB: don't use dict comprehension for python 2.6 compatibility
3033 newheaders
= dict((k
, v
) for k
, v
in req
.headers
.items()
3034 if k
.lower() not in CONTENT_HEADERS
)
3035 return compat_urllib_request
.Request(
3036 newurl
, headers
=newheaders
, origin_req_host
=req
.origin_req_host
,
3040 def extract_timezone(date_str
):
3043 ^.{8,}? # >=8 char non-TZ prefix, if present
3044 (?P<tz>Z| # just the UTC Z, or
3045 (?:(?<=.\b\d{4}|\b\d{2}:\d\d)| # preceded by 4 digits or hh:mm or
3046 (?<!.\b[a-zA-Z]{3}|[a-zA-Z]{4}|..\b\d\d)) # not preceded by 3 alpha word or >= 4 alpha or 2 digits
3047 [ ]? # optional space
3048 (?P<sign>\+|-) # +/-
3049 (?P<hours>[0-9]{2}):?(?P<minutes>[0-9]{2}) # hh[:]mm
3053 timezone
= datetime
.timedelta()
3055 date_str
= date_str
[:-len(m
.group('tz'))]
3056 if not m
.group('sign'):
3057 timezone
= datetime
.timedelta()
3059 sign
= 1 if m
.group('sign') == '+' else -1
3060 timezone
= datetime
.timedelta(
3061 hours
=sign
* int(m
.group('hours')),
3062 minutes
=sign
* int(m
.group('minutes')))
3063 return timezone
, date_str
3066 def parse_iso8601(date_str
, delimiter
='T', timezone
=None):
3067 """ Return a UNIX timestamp from the given date """
3069 if date_str
is None:
3072 date_str
= re
.sub(r
'\.[0-9]+', '', date_str
)
3074 if timezone
is None:
3075 timezone
, date_str
= extract_timezone(date_str
)
3078 date_format
= '%Y-%m-%d{0}%H:%M:%S'.format(delimiter
)
3079 dt
= datetime
.datetime
.strptime(date_str
, date_format
) - timezone
3080 return calendar
.timegm(dt
.timetuple())
3085 def date_formats(day_first
=True):
3086 return DATE_FORMATS_DAY_FIRST
if day_first
else DATE_FORMATS_MONTH_FIRST
3089 def unified_strdate(date_str
, day_first
=True):
3090 """Return a string with the date in the format YYYYMMDD"""
3092 if date_str
is None:
3096 date_str
= date_str
.replace(',', ' ')
3097 # Remove AM/PM + timezone
3098 date_str
= re
.sub(r
'(?i)\s*(?:AM|PM)(?:\s+[A-Z]+)?', '', date_str
)
3099 _
, date_str
= extract_timezone(date_str
)
3101 for expression
in date_formats(day_first
):
3103 upload_date
= datetime
.datetime
.strptime(date_str
, expression
).strftime('%Y%m%d')
3106 if upload_date
is None:
3107 timetuple
= email
.utils
.parsedate_tz(date_str
)
3110 upload_date
= datetime
.datetime(*timetuple
[:6]).strftime('%Y%m%d')
3113 if upload_date
is not None:
3114 return compat_str(upload_date
)
3117 def unified_timestamp(date_str
, day_first
=True):
3118 if date_str
is None:
3121 date_str
= re
.sub(r
'[,|]', '', date_str
)
3123 pm_delta
= 12 if re
.search(r
'(?i)PM', date_str
) else 0
3124 timezone
, date_str
= extract_timezone(date_str
)
3126 # Remove AM/PM + timezone
3127 date_str
= re
.sub(r
'(?i)\s*(?:AM|PM)(?:\s+[A-Z]+)?', '', date_str
)
3129 # Remove unrecognized timezones from ISO 8601 alike timestamps
3130 m
= re
.search(r
'\d{1,2}:\d{1,2}(?:\.\d+)?(?P<tz>\s*[A-Z]+)$', date_str
)
3132 date_str
= date_str
[:-len(m
.group('tz'))]
3134 # Python only supports microseconds, so remove nanoseconds
3135 m
= re
.search(r
'^([0-9]{4,}-[0-9]{1,2}-[0-9]{1,2}T[0-9]{1,2}:[0-9]{1,2}:[0-9]{1,2}\.[0-9]{6})[0-9]+$', date_str
)
3137 date_str
= m
.group(1)
3139 for expression
in date_formats(day_first
):
3141 dt
= datetime
.datetime
.strptime(date_str
, expression
) - timezone
+ datetime
.timedelta(hours
=pm_delta
)
3142 return calendar
.timegm(dt
.timetuple())
3145 timetuple
= email
.utils
.parsedate_tz(date_str
)
3147 return calendar
.timegm(timetuple
) + pm_delta
* 3600
3150 def determine_ext(url
, default_ext
='unknown_video'):
3151 if url
is None or '.' not in url
:
3153 guess
= url
.partition('?')[0].rpartition('.')[2]
3154 if re
.match(r
'^[A-Za-z0-9]+$', guess
):
3156 # Try extract ext from URLs like http://example.com/foo/bar.mp4/?download
3157 elif guess
.rstrip('/') in KNOWN_EXTENSIONS
:
3158 return guess
.rstrip('/')
3163 def subtitles_filename(filename
, sub_lang
, sub_format
, expected_real_ext
=None):
3164 return replace_extension(filename
, sub_lang
+ '.' + sub_format
, expected_real_ext
)
3167 def datetime_from_str(date_str
, precision
='auto', format
='%Y%m%d'):
3169 Return a datetime object from a string in the format YYYYMMDD or
3170 (now|today|date)[+-][0-9](microsecond|second|minute|hour|day|week|month|year)(s)?
3172 format: string date format used to return datetime object from
3173 precision: round the time portion of a datetime object.
3174 auto|microsecond|second|minute|hour|day.
3175 auto: round to the unit provided in date_str (if applicable).
3177 auto_precision
= False
3178 if precision
== 'auto':
3179 auto_precision
= True
3180 precision
= 'microsecond'
3181 today
= datetime_round(datetime
.datetime
.now(), precision
)
3182 if date_str
in ('now', 'today'):
3184 if date_str
== 'yesterday':
3185 return today
- datetime
.timedelta(days
=1)
3187 r
'(?P<start>.+)(?P<sign>[+-])(?P<time>\d+)(?P<unit>microsecond|second|minute|hour|day|week|month|year)(s)?',
3189 if match
is not None:
3190 start_time
= datetime_from_str(match
.group('start'), precision
, format
)
3191 time
= int(match
.group('time')) * (-1 if match
.group('sign') == '-' else 1)
3192 unit
= match
.group('unit')
3193 if unit
== 'month' or unit
== 'year':
3194 new_date
= datetime_add_months(start_time
, time
* 12 if unit
== 'year' else time
)
3200 delta
= datetime
.timedelta(**{unit + 's': time}
)
3201 new_date
= start_time
+ delta
3203 return datetime_round(new_date
, unit
)
3206 return datetime_round(datetime
.datetime
.strptime(date_str
, format
), precision
)
3209 def date_from_str(date_str
, format
='%Y%m%d'):
3211 Return a datetime object from a string in the format YYYYMMDD or
3212 (now|today|date)[+-][0-9](microsecond|second|minute|hour|day|week|month|year)(s)?
3214 format: string date format used to return datetime object from
3216 return datetime_from_str(date_str
, precision
='microsecond', format
=format
).date()
3219 def datetime_add_months(dt
, months
):
3220 """Increment/Decrement a datetime object by months."""
3221 month
= dt
.month
+ months
- 1
3222 year
= dt
.year
+ month
// 12
3223 month
= month
% 12 + 1
3224 day
= min(dt
.day
, calendar
.monthrange(year
, month
)[1])
3225 return dt
.replace(year
, month
, day
)
3228 def datetime_round(dt
, precision
='day'):
3230 Round a datetime object's time to a specific precision
3232 if precision
== 'microsecond':
3241 roundto
= lambda x
, n
: ((x
+ n
/ 2) // n
) * n
3242 timestamp
= calendar
.timegm(dt
.timetuple())
3243 return datetime
.datetime
.utcfromtimestamp(roundto(timestamp
, unit_seconds
[precision
]))
3246 def hyphenate_date(date_str
):
3248 Convert a date in 'YYYYMMDD' format to 'YYYY-MM-DD' format"""
3249 match
= re
.match(r
'^(\d\d\d\d)(\d\d)(\d\d)$', date_str
)
3250 if match
is not None:
3251 return '-'.join(match
.groups())
3256 class DateRange(object):
3257 """Represents a time interval between two dates"""
3259 def __init__(self
, start
=None, end
=None):
3260 """start and end must be strings in the format accepted by date"""
3261 if start
is not None:
3262 self
.start
= date_from_str(start
)
3264 self
.start
= datetime
.datetime
.min.date()
3266 self
.end
= date_from_str(end
)
3268 self
.end
= datetime
.datetime
.max.date()
3269 if self
.start
> self
.end
:
3270 raise ValueError('Date range: "%s" , the start date must be before the end date' % self
)
3274 """Returns a range that only contains the given day"""
3275 return cls(day
, day
)
3277 def __contains__(self
, date
):
3278 """Check if the date is in the range"""
3279 if not isinstance(date
, datetime
.date
):
3280 date
= date_from_str(date
)
3281 return self
.start
<= date
<= self
.end
3284 return '%s - %s' % (self
.start
.isoformat(), self
.end
.isoformat())
3287 def platform_name():
3288 """ Returns the platform name as a compat_str """
3289 res
= platform
.platform()
3290 if isinstance(res
, bytes):
3291 res
= res
.decode(preferredencoding())
3293 assert isinstance(res
, compat_str
)
3297 def get_windows_version():
3298 ''' Get Windows version. None if it's not running on Windows '''
3299 if compat_os_name
== 'nt':
3300 return version_tuple(platform
.win32_ver()[1])
3305 def _windows_write_string(s
, out
):
3306 """ Returns True if the string was written using special methods,
3307 False if it has yet to be written out."""
3308 # Adapted from http://stackoverflow.com/a/3259271/35070
3311 import ctypes
.wintypes
3319 fileno
= out
.fileno()
3320 except AttributeError:
3321 # If the output stream doesn't have a fileno, it's virtual
3323 except io
.UnsupportedOperation
:
3324 # Some strange Windows pseudo files?
3326 if fileno
not in WIN_OUTPUT_IDS
:
3329 GetStdHandle
= compat_ctypes_WINFUNCTYPE(
3330 ctypes
.wintypes
.HANDLE
, ctypes
.wintypes
.DWORD
)(
3331 ('GetStdHandle', ctypes
.windll
.kernel32
))
3332 h
= GetStdHandle(WIN_OUTPUT_IDS
[fileno
])
3334 WriteConsoleW
= compat_ctypes_WINFUNCTYPE(
3335 ctypes
.wintypes
.BOOL
, ctypes
.wintypes
.HANDLE
, ctypes
.wintypes
.LPWSTR
,
3336 ctypes
.wintypes
.DWORD
, ctypes
.POINTER(ctypes
.wintypes
.DWORD
),
3337 ctypes
.wintypes
.LPVOID
)(('WriteConsoleW', ctypes
.windll
.kernel32
))
3338 written
= ctypes
.wintypes
.DWORD(0)
3340 GetFileType
= compat_ctypes_WINFUNCTYPE(ctypes
.wintypes
.DWORD
, ctypes
.wintypes
.DWORD
)(('GetFileType', ctypes
.windll
.kernel32
))
3341 FILE_TYPE_CHAR
= 0x0002
3342 FILE_TYPE_REMOTE
= 0x8000
3343 GetConsoleMode
= compat_ctypes_WINFUNCTYPE(
3344 ctypes
.wintypes
.BOOL
, ctypes
.wintypes
.HANDLE
,
3345 ctypes
.POINTER(ctypes
.wintypes
.DWORD
))(
3346 ('GetConsoleMode', ctypes
.windll
.kernel32
))
3347 INVALID_HANDLE_VALUE
= ctypes
.wintypes
.DWORD(-1).value
3349 def not_a_console(handle
):
3350 if handle
== INVALID_HANDLE_VALUE
or handle
is None:
3352 return ((GetFileType(handle
) & ~FILE_TYPE_REMOTE
) != FILE_TYPE_CHAR
3353 or GetConsoleMode(handle
, ctypes
.byref(ctypes
.wintypes
.DWORD())) == 0)
3355 if not_a_console(h
):
3358 def next_nonbmp_pos(s
):
3360 return next(i
for i
, c
in enumerate(s
) if ord(c
) > 0xffff)
3361 except StopIteration:
3365 count
= min(next_nonbmp_pos(s
), 1024)
3367 ret
= WriteConsoleW(
3368 h
, s
, count
if count
else 2, ctypes
.byref(written
), None)
3370 raise OSError('Failed to write string')
3371 if not count
: # We just wrote a non-BMP character
3372 assert written
.value
== 2
3375 assert written
.value
> 0
3376 s
= s
[written
.value
:]
3380 def write_string(s
, out
=None, encoding
=None):
3383 assert type(s
) == compat_str
3385 if sys
.platform
== 'win32' and encoding
is None and hasattr(out
, 'fileno'):
3386 if _windows_write_string(s
, out
):
3389 if ('b' in getattr(out
, 'mode', '')
3390 or sys
.version_info
[0] < 3): # Python 2 lies about mode of sys.stderr
3391 byt
= s
.encode(encoding
or preferredencoding(), 'ignore')
3393 elif hasattr(out
, 'buffer'):
3394 enc
= encoding
or getattr(out
, 'encoding', None) or preferredencoding()
3395 byt
= s
.encode(enc
, 'ignore')
3396 out
.buffer.write(byt
)
3402 def bytes_to_intlist(bs
):
3405 if isinstance(bs
[0], int): # Python 3
3408 return [ord(c
) for c
in bs
]
3411 def intlist_to_bytes(xs
):
3414 return compat_struct_pack('%dB' % len(xs
), *xs
)
3417 # Cross-platform file locking
3418 if sys
.platform
== 'win32':
3419 import ctypes
.wintypes
3422 class OVERLAPPED(ctypes
.Structure
):
3424 ('Internal', ctypes
.wintypes
.LPVOID
),
3425 ('InternalHigh', ctypes
.wintypes
.LPVOID
),
3426 ('Offset', ctypes
.wintypes
.DWORD
),
3427 ('OffsetHigh', ctypes
.wintypes
.DWORD
),
3428 ('hEvent', ctypes
.wintypes
.HANDLE
),
3431 kernel32
= ctypes
.windll
.kernel32
3432 LockFileEx
= kernel32
.LockFileEx
3433 LockFileEx
.argtypes
= [
3434 ctypes
.wintypes
.HANDLE
, # hFile
3435 ctypes
.wintypes
.DWORD
, # dwFlags
3436 ctypes
.wintypes
.DWORD
, # dwReserved
3437 ctypes
.wintypes
.DWORD
, # nNumberOfBytesToLockLow
3438 ctypes
.wintypes
.DWORD
, # nNumberOfBytesToLockHigh
3439 ctypes
.POINTER(OVERLAPPED
) # Overlapped
3441 LockFileEx
.restype
= ctypes
.wintypes
.BOOL
3442 UnlockFileEx
= kernel32
.UnlockFileEx
3443 UnlockFileEx
.argtypes
= [
3444 ctypes
.wintypes
.HANDLE
, # hFile
3445 ctypes
.wintypes
.DWORD
, # dwReserved
3446 ctypes
.wintypes
.DWORD
, # nNumberOfBytesToLockLow
3447 ctypes
.wintypes
.DWORD
, # nNumberOfBytesToLockHigh
3448 ctypes
.POINTER(OVERLAPPED
) # Overlapped
3450 UnlockFileEx
.restype
= ctypes
.wintypes
.BOOL
3451 whole_low
= 0xffffffff
3452 whole_high
= 0x7fffffff
3454 def _lock_file(f
, exclusive
):
3455 overlapped
= OVERLAPPED()
3456 overlapped
.Offset
= 0
3457 overlapped
.OffsetHigh
= 0
3458 overlapped
.hEvent
= 0
3459 f
._lock
_file
_overlapped
_p
= ctypes
.pointer(overlapped
)
3460 handle
= msvcrt
.get_osfhandle(f
.fileno())
3461 if not LockFileEx(handle
, 0x2 if exclusive
else 0x0, 0,
3462 whole_low
, whole_high
, f
._lock
_file
_overlapped
_p
):
3463 raise OSError('Locking file failed: %r' % ctypes
.FormatError())
3465 def _unlock_file(f
):
3466 assert f
._lock
_file
_overlapped
_p
3467 handle
= msvcrt
.get_osfhandle(f
.fileno())
3468 if not UnlockFileEx(handle
, 0,
3469 whole_low
, whole_high
, f
._lock
_file
_overlapped
_p
):
3470 raise OSError('Unlocking file failed: %r' % ctypes
.FormatError())
3473 # Some platforms, such as Jython, is missing fcntl
3477 def _lock_file(f
, exclusive
):
3478 fcntl
.flock(f
, fcntl
.LOCK_EX
if exclusive
else fcntl
.LOCK_SH
)
3480 def _unlock_file(f
):
3481 fcntl
.flock(f
, fcntl
.LOCK_UN
)
3483 UNSUPPORTED_MSG
= 'file locking is not supported on this platform'
3485 def _lock_file(f
, exclusive
):
3486 raise IOError(UNSUPPORTED_MSG
)
3488 def _unlock_file(f
):
3489 raise IOError(UNSUPPORTED_MSG
)
3492 class locked_file(object):
3493 def __init__(self
, filename
, mode
, encoding
=None):
3494 assert mode
in ['r', 'a', 'w']
3495 self
.f
= io
.open(filename
, mode
, encoding
=encoding
)
3498 def __enter__(self
):
3499 exclusive
= self
.mode
!= 'r'
3501 _lock_file(self
.f
, exclusive
)
3507 def __exit__(self
, etype
, value
, traceback
):
3509 _unlock_file(self
.f
)
3516 def write(self
, *args
):
3517 return self
.f
.write(*args
)
3519 def read(self
, *args
):
3520 return self
.f
.read(*args
)
3523 def get_filesystem_encoding():
3524 encoding
= sys
.getfilesystemencoding()
3525 return encoding
if encoding
is not None else 'utf-8'
3528 def shell_quote(args
):
3530 encoding
= get_filesystem_encoding()
3532 if isinstance(a
, bytes):
3533 # We may get a filename encoded with 'encodeFilename'
3534 a
= a
.decode(encoding
)
3535 quoted_args
.append(compat_shlex_quote(a
))
3536 return ' '.join(quoted_args
)
3539 def smuggle_url(url
, data
):
3540 """ Pass additional data in a URL for internal use. """
3542 url
, idata
= unsmuggle_url(url
, {})
3544 sdata
= compat_urllib_parse_urlencode(
3545 {'__youtubedl_smuggle': json.dumps(data)}
)
3546 return url
+ '#' + sdata
3549 def unsmuggle_url(smug_url
, default
=None):
3550 if '#__youtubedl_smuggle' not in smug_url
:
3551 return smug_url
, default
3552 url
, _
, sdata
= smug_url
.rpartition('#')
3553 jsond
= compat_parse_qs(sdata
)['__youtubedl_smuggle'][0]
3554 data
= json
.loads(jsond
)
3558 def format_bytes(bytes):
3561 if type(bytes) is str:
3562 bytes = float(bytes)
3566 exponent
= int(math
.log(bytes, 1024.0))
3567 suffix
= ['B', 'KiB', 'MiB', 'GiB', 'TiB', 'PiB', 'EiB', 'ZiB', 'YiB'][exponent
]
3568 converted
= float(bytes) / float(1024 ** exponent
)
3569 return '%.2f%s' % (converted
, suffix
)
3572 def lookup_unit_table(unit_table
, s
):
3573 units_re
= '|'.join(re
.escape(u
) for u
in unit_table
)
3575 r
'(?P<num>[0-9]+(?:[,.][0-9]*)?)\s*(?P<unit>%s)\b' % units_re
, s
)
3578 num_str
= m
.group('num').replace(',', '.')
3579 mult
= unit_table
[m
.group('unit')]
3580 return int(float(num_str
) * mult
)
3583 def parse_filesize(s
):
3587 # The lower-case forms are of course incorrect and unofficial,
3588 # but we support those too
3605 'megabytes': 1000 ** 2,
3606 'mebibytes': 1024 ** 2,
3612 'gigabytes': 1000 ** 3,
3613 'gibibytes': 1024 ** 3,
3619 'terabytes': 1000 ** 4,
3620 'tebibytes': 1024 ** 4,
3626 'petabytes': 1000 ** 5,
3627 'pebibytes': 1024 ** 5,
3633 'exabytes': 1000 ** 6,
3634 'exbibytes': 1024 ** 6,
3640 'zettabytes': 1000 ** 7,
3641 'zebibytes': 1024 ** 7,
3647 'yottabytes': 1000 ** 8,
3648 'yobibytes': 1024 ** 8,
3651 return lookup_unit_table(_UNIT_TABLE
, s
)
3660 if re
.match(r
'^[\d,.]+$', s
):
3661 return str_to_int(s
)
3672 return lookup_unit_table(_UNIT_TABLE
, s
)
3675 def parse_resolution(s
):
3679 mobj
= re
.search(r
'\b(?P<w>\d+)\s*[xX×]\s*(?P<h>\d+)\b', s
)
3682 'width': int(mobj
.group('w')),
3683 'height': int(mobj
.group('h')),
3686 mobj
= re
.search(r
'\b(\d+)[pPiI]\b', s
)
3688 return {'height': int(mobj.group(1))}
3690 mobj
= re
.search(r
'\b([48])[kK]\b', s
)
3692 return {'height': int(mobj.group(1)) * 540}
3697 def parse_bitrate(s
):
3698 if not isinstance(s
, compat_str
):
3700 mobj
= re
.search(r
'\b(\d+)\s*kbps', s
)
3702 return int(mobj
.group(1))
3705 def month_by_name(name
, lang
='en'):
3706 """ Return the number of a month by (locale-independently) English name """
3708 month_names
= MONTH_NAMES
.get(lang
, MONTH_NAMES
['en'])
3711 return month_names
.index(name
) + 1
3716 def month_by_abbreviation(abbrev
):
3717 """ Return the number of a month by (locale-independently) English
3721 return [s
[:3] for s
in ENGLISH_MONTH_NAMES
].index(abbrev
) + 1
3726 def fix_xml_ampersands(xml_str
):
3727 """Replace all the '&' by '&' in XML"""
3729 r
'&(?!amp;|lt;|gt;|apos;|quot;|#x[0-9a-fA-F]{,4};|#[0-9]{,4};)',
3734 def setproctitle(title
):
3735 assert isinstance(title
, compat_str
)
3737 # ctypes in Jython is not complete
3738 # http://bugs.jython.org/issue2148
3739 if sys
.platform
.startswith('java'):
3743 libc
= ctypes
.cdll
.LoadLibrary('libc.so.6')
3747 # LoadLibrary in Windows Python 2.7.13 only expects
3748 # a bytestring, but since unicode_literals turns
3749 # every string into a unicode string, it fails.
3751 title_bytes
= title
.encode('utf-8')
3752 buf
= ctypes
.create_string_buffer(len(title_bytes
))
3753 buf
.value
= title_bytes
3755 libc
.prctl(15, buf
, 0, 0, 0)
3756 except AttributeError:
3757 return # Strange libc, just skip this
3760 def remove_start(s
, start
):
3761 return s
[len(start
):] if s
is not None and s
.startswith(start
) else s
3764 def remove_end(s
, end
):
3765 return s
[:-len(end
)] if s
is not None and s
.endswith(end
) else s
3768 def remove_quotes(s
):
3769 if s
is None or len(s
) < 2:
3771 for quote
in ('"', "'", ):
3772 if s
[0] == quote
and s
[-1] == quote
:
3777 def get_domain(url
):
3778 domain
= re
.match(r
'(?:https?:\/\/)?(?:www\.)?(?P<domain>[^\n\/]+\.[^\n\/]+)(?:\/(.*))?', url
)
3779 return domain
.group('domain') if domain
else None
3782 def url_basename(url
):
3783 path
= compat_urlparse
.urlparse(url
).path
3784 return path
.strip('/').split('/')[-1]
3788 return re
.match(r
'https?://[^?#&]+/', url
).group()
3791 def urljoin(base
, path
):
3792 if isinstance(path
, bytes):
3793 path
= path
.decode('utf-8')
3794 if not isinstance(path
, compat_str
) or not path
:
3796 if re
.match(r
'^(?:[a-zA-Z][a-zA-Z0-9+-.]*:)?//', path
):
3798 if isinstance(base
, bytes):
3799 base
= base
.decode('utf-8')
3800 if not isinstance(base
, compat_str
) or not re
.match(
3801 r
'^(?:https?:)?//', base
):
3803 return compat_urlparse
.urljoin(base
, path
)
3806 class HEADRequest(compat_urllib_request
.Request
):
3807 def get_method(self
):
3811 class PUTRequest(compat_urllib_request
.Request
):
3812 def get_method(self
):
3816 def int_or_none(v
, scale
=1, default
=None, get_attr
=None, invscale
=1):
3819 v
= getattr(v
, get_attr
, None)
3825 return int(v
) * invscale
// scale
3826 except (ValueError, TypeError):
3830 def str_or_none(v
, default
=None):
3831 return default
if v
is None else compat_str(v
)
3834 def str_to_int(int_str
):
3835 """ A more relaxed version of int_or_none """
3836 if isinstance(int_str
, compat_integer_types
):
3838 elif isinstance(int_str
, compat_str
):
3839 int_str
= re
.sub(r
'[,\.\+]', '', int_str
)
3840 return int_or_none(int_str
)
3843 def float_or_none(v
, scale
=1, invscale
=1, default
=None):
3847 return float(v
) * invscale
/ scale
3848 except (ValueError, TypeError):
3852 def bool_or_none(v
, default
=None):
3853 return v
if isinstance(v
, bool) else default
3856 def strip_or_none(v
, default
=None):
3857 return v
.strip() if isinstance(v
, compat_str
) else default
3860 def url_or_none(url
):
3861 if not url
or not isinstance(url
, compat_str
):
3864 return url
if re
.match(r
'^(?:(?:https?|rt(?:m(?:pt?[es]?|fp)|sp[su]?)|mms|ftps?):)?//', url
) else None
3867 def strftime_or_none(timestamp
, date_format
, default
=None):
3868 datetime_object
= None
3870 if isinstance(timestamp
, compat_numeric_types
): # unix timestamp
3871 datetime_object
= datetime
.datetime
.utcfromtimestamp(timestamp
)
3872 elif isinstance(timestamp
, compat_str
): # assume YYYYMMDD
3873 datetime_object
= datetime
.datetime
.strptime(timestamp
, '%Y%m%d')
3874 return datetime_object
.strftime(date_format
)
3875 except (ValueError, TypeError, AttributeError):
3879 def parse_duration(s
):
3880 if not isinstance(s
, compat_basestring
):
3885 days
, hours
, mins
, secs
, ms
= [None] * 5
3886 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
)
3888 days
, hours
, mins
, secs
, ms
= m
.groups()
3893 [0-9]+\s*y(?:ears?)?\s*
3896 [0-9]+\s*m(?:onths?)?\s*
3899 [0-9]+\s*w(?:eeks?)?\s*
3902 (?P<days>[0-9]+)\s*d(?:ays?)?\s*
3906 (?P<hours>[0-9]+)\s*h(?:ours?)?\s*
3909 (?P<mins>[0-9]+)\s*m(?:in(?:ute)?s?)?\s*
3912 (?P<secs>[0-9]+)(?P<ms>\.[0-9]+)?\s*s(?:ec(?:ond)?s?)?\s*
3915 days
, hours
, mins
, secs
, ms
= m
.groups()
3917 m
= re
.match(r
'(?i)(?:(?P<hours>[0-9.]+)\s*(?:hours?)|(?P<mins>[0-9.]+)\s*(?:mins?\.?|minutes?)\s*)Z?$', s
)
3919 hours
, mins
= m
.groups()
3925 duration
+= float(secs
)
3927 duration
+= float(mins
) * 60
3929 duration
+= float(hours
) * 60 * 60
3931 duration
+= float(days
) * 24 * 60 * 60
3933 duration
+= float(ms
)
3937 def prepend_extension(filename
, ext
, expected_real_ext
=None):
3938 name
, real_ext
= os
.path
.splitext(filename
)
3940 '{0}.{1}{2}'.format(name
, ext
, real_ext
)
3941 if not expected_real_ext
or real_ext
[1:] == expected_real_ext
3942 else '{0}.{1}'.format(filename
, ext
))
3945 def replace_extension(filename
, ext
, expected_real_ext
=None):
3946 name
, real_ext
= os
.path
.splitext(filename
)
3947 return '{0}.{1}'.format(
3948 name
if not expected_real_ext
or real_ext
[1:] == expected_real_ext
else filename
,
3952 def check_executable(exe
, args
=[]):
3953 """ Checks if the given binary is installed somewhere in PATH, and returns its name.
3954 args can be a list of arguments for a short output (like -version) """
3956 process_communicate_or_kill(subprocess
.Popen(
3957 [exe
] + args
, stdout
=subprocess
.PIPE
, stderr
=subprocess
.PIPE
))
3963 def get_exe_version(exe
, args
=['--version'],
3964 version_re
=None, unrecognized
='present'):
3965 """ Returns the version of the specified executable,
3966 or False if the executable is not present """
3968 # STDIN should be redirected too. On UNIX-like systems, ffmpeg triggers
3969 # SIGTTOU if yt-dlp is run in the background.
3970 # See https://github.com/ytdl-org/youtube-dl/issues/955#issuecomment-209789656
3971 out
, _
= process_communicate_or_kill(subprocess
.Popen(
3972 [encodeArgument(exe
)] + args
,
3973 stdin
=subprocess
.PIPE
,
3974 stdout
=subprocess
.PIPE
, stderr
=subprocess
.STDOUT
))
3977 if isinstance(out
, bytes): # Python 2.x
3978 out
= out
.decode('ascii', 'ignore')
3979 return detect_exe_version(out
, version_re
, unrecognized
)
3982 def detect_exe_version(output
, version_re
=None, unrecognized
='present'):
3983 assert isinstance(output
, compat_str
)
3984 if version_re
is None:
3985 version_re
= r
'version\s+([-0-9._a-zA-Z]+)'
3986 m
= re
.search(version_re
, output
)
3993 class LazyList(collections
.abc
.Sequence
):
3994 ''' Lazy immutable list from an iterable
3995 Note that slices of a LazyList are lists and not LazyList'''
3997 class IndexError(IndexError):
4000 def __init__(self
, iterable
):
4001 self
.__iterable
= iter(iterable
)
4003 self
.__reversed
= False
4007 # We need to consume the entire iterable to iterate in reverse
4008 yield from self
.exhaust()
4010 yield from self
.__cache
4011 for item
in self
.__iterable
:
4012 self
.__cache
.append(item
)
4015 def __exhaust(self
):
4016 self
.__cache
.extend(self
.__iterable
)
4020 ''' Evaluate the entire iterable '''
4021 return self
.__exhaust
()[::-1 if self
.__reversed
else 1]
4024 def __reverse_index(x
):
4025 return None if x
is None else -(x
+ 1)
4027 def __getitem__(self
, idx
):
4028 if isinstance(idx
, slice):
4030 idx
= slice(self
.__reverse
_index
(idx
.start
), self
.__reverse
_index
(idx
.stop
), -(idx
.step
or 1))
4031 start
, stop
, step
= idx
.start
, idx
.stop
, idx
.step
or 1
4032 elif isinstance(idx
, int):
4034 idx
= self
.__reverse
_index
(idx
)
4035 start
, stop
, step
= idx
, idx
, 0
4037 raise TypeError('indices must be integers or slices')
4038 if ((start
or 0) < 0 or (stop
or 0) < 0
4039 or (start
is None and step
< 0)
4040 or (stop
is None and step
> 0)):
4041 # We need to consume the entire iterable to be able to slice from the end
4042 # Obviously, never use this with infinite iterables
4045 return self
.__cache
[idx
]
4046 except IndexError as e
:
4047 raise self
.IndexError(e
) from e
4048 n
= max(start
or 0, stop
or 0) - len(self
.__cache
) + 1
4050 self
.__cache
.extend(itertools
.islice(self
.__iterable
, n
))
4052 return self
.__cache
[idx
]
4053 except IndexError as e
:
4054 raise self
.IndexError(e
) from e
4058 self
[-1] if self
.__reversed
else self
[0]
4059 except self
.IndexError:
4065 return len(self
.__cache
)
4068 self
.__reversed
= not self
.__reversed
4072 # repr and str should mimic a list. So we exhaust the iterable
4073 return repr(self
.exhaust())
4076 return repr(self
.exhaust())
4081 # This is only useful for tests
4082 return len(self
.getslice())
4084 def __init__(self
, pagefunc
, pagesize
, use_cache
=True):
4085 self
._pagefunc
= pagefunc
4086 self
._pagesize
= pagesize
4087 self
._use
_cache
= use_cache
4090 def getpage(self
, pagenum
):
4091 page_results
= self
._cache
.get(pagenum
) or list(self
._pagefunc
(pagenum
))
4093 self
._cache
[pagenum
] = page_results
4096 def getslice(self
, start
=0, end
=None):
4097 return list(self
._getslice
(start
, end
))
4099 def _getslice(self
, start
, end
):
4100 raise NotImplementedError('This method must be implemented by subclasses')
4102 def __getitem__(self
, idx
):
4103 # NOTE: cache must be enabled if this is used
4104 if not isinstance(idx
, int) or idx
< 0:
4105 raise TypeError('indices must be non-negative integers')
4106 entries
= self
.getslice(idx
, idx
+ 1)
4107 return entries
[0] if entries
else None
4110 class OnDemandPagedList(PagedList
):
4111 def _getslice(self
, start
, end
):
4112 for pagenum
in itertools
.count(start
// self
._pagesize
):
4113 firstid
= pagenum
* self
._pagesize
4114 nextfirstid
= pagenum
* self
._pagesize
+ self
._pagesize
4115 if start
>= nextfirstid
:
4119 start
% self
._pagesize
4120 if firstid
<= start
< nextfirstid
4123 ((end
- 1) % self
._pagesize
) + 1
4124 if (end
is not None and firstid
<= end
<= nextfirstid
)
4127 page_results
= self
.getpage(pagenum
)
4128 if startv
!= 0 or endv
is not None:
4129 page_results
= page_results
[startv
:endv
]
4130 yield from page_results
4132 # A little optimization - if current page is not "full", ie. does
4133 # not contain page_size videos then we can assume that this page
4134 # is the last one - there are no more ids on further pages -
4135 # i.e. no need to query again.
4136 if len(page_results
) + startv
< self
._pagesize
:
4139 # If we got the whole page, but the next page is not interesting,
4140 # break out early as well
4141 if end
== nextfirstid
:
4145 class InAdvancePagedList(PagedList
):
4146 def __init__(self
, pagefunc
, pagecount
, pagesize
):
4147 self
._pagecount
= pagecount
4148 PagedList
.__init
__(self
, pagefunc
, pagesize
, True)
4150 def _getslice(self
, start
, end
):
4151 start_page
= start
// self
._pagesize
4153 self
._pagecount
if end
is None else (end
// self
._pagesize
+ 1))
4154 skip_elems
= start
- start_page
* self
._pagesize
4155 only_more
= None if end
is None else end
- start
4156 for pagenum
in range(start_page
, end_page
):
4157 page_results
= self
.getpage(pagenum
)
4159 page_results
= page_results
[skip_elems
:]
4161 if only_more
is not None:
4162 if len(page_results
) < only_more
:
4163 only_more
-= len(page_results
)
4165 yield from page_results
[:only_more
]
4167 yield from page_results
4170 def uppercase_escape(s
):
4171 unicode_escape
= codecs
.getdecoder('unicode_escape')
4173 r
'\\U[0-9a-fA-F]{8}',
4174 lambda m
: unicode_escape(m
.group(0))[0],
4178 def lowercase_escape(s
):
4179 unicode_escape
= codecs
.getdecoder('unicode_escape')
4181 r
'\\u[0-9a-fA-F]{4}',
4182 lambda m
: unicode_escape(m
.group(0))[0],
4186 def escape_rfc3986(s
):
4187 """Escape non-ASCII characters as suggested by RFC 3986"""
4188 if sys
.version_info
< (3, 0) and isinstance(s
, compat_str
):
4189 s
= s
.encode('utf-8')
4190 return compat_urllib_parse
.quote(s
, b
"%/;:@&=+$,!~*'()?#[]")
4193 def escape_url(url
):
4194 """Escape URL as suggested by RFC 3986"""
4195 url_parsed
= compat_urllib_parse_urlparse(url
)
4196 return url_parsed
._replace
(
4197 netloc
=url_parsed
.netloc
.encode('idna').decode('ascii'),
4198 path
=escape_rfc3986(url_parsed
.path
),
4199 params
=escape_rfc3986(url_parsed
.params
),
4200 query
=escape_rfc3986(url_parsed
.query
),
4201 fragment
=escape_rfc3986(url_parsed
.fragment
)
4206 return compat_parse_qs(compat_urllib_parse_urlparse(url
).query
)
4209 def read_batch_urls(batch_fd
):
4211 if not isinstance(url
, compat_str
):
4212 url
= url
.decode('utf-8', 'replace')
4213 BOM_UTF8
= ('\xef\xbb\xbf', '\ufeff')
4214 for bom
in BOM_UTF8
:
4215 if url
.startswith(bom
):
4216 url
= url
[len(bom
):]
4218 if not url
or url
.startswith(('#', ';', ']')):
4220 # "#" cannot be stripped out since it is part of the URI
4221 # However, it can be safely stipped out if follwing a whitespace
4222 return re
.split(r
'\s#', url
, 1)[0].rstrip()
4224 with contextlib
.closing(batch_fd
) as fd
:
4225 return [url
for url
in map(fixup
, fd
) if url
]
4228 def urlencode_postdata(*args
, **kargs
):
4229 return compat_urllib_parse_urlencode(*args
, **kargs
).encode('ascii')
4232 def update_url_query(url
, query
):
4235 parsed_url
= compat_urlparse
.urlparse(url
)
4236 qs
= compat_parse_qs(parsed_url
.query
)
4238 return compat_urlparse
.urlunparse(parsed_url
._replace
(
4239 query
=compat_urllib_parse_urlencode(qs
, True)))
4242 def update_Request(req
, url
=None, data
=None, headers
={}, query={}
):
4243 req_headers
= req
.headers
.copy()
4244 req_headers
.update(headers
)
4245 req_data
= data
or req
.data
4246 req_url
= update_url_query(url
or req
.get_full_url(), query
)
4247 req_get_method
= req
.get_method()
4248 if req_get_method
== 'HEAD':
4249 req_type
= HEADRequest
4250 elif req_get_method
== 'PUT':
4251 req_type
= PUTRequest
4253 req_type
= compat_urllib_request
.Request
4255 req_url
, data
=req_data
, headers
=req_headers
,
4256 origin_req_host
=req
.origin_req_host
, unverifiable
=req
.unverifiable
)
4257 if hasattr(req
, 'timeout'):
4258 new_req
.timeout
= req
.timeout
4262 def _multipart_encode_impl(data
, boundary
):
4263 content_type
= 'multipart/form-data; boundary=%s' % boundary
4266 for k
, v
in data
.items():
4267 out
+= b
'--' + boundary
.encode('ascii') + b
'\r\n'
4268 if isinstance(k
, compat_str
):
4269 k
= k
.encode('utf-8')
4270 if isinstance(v
, compat_str
):
4271 v
= v
.encode('utf-8')
4272 # RFC 2047 requires non-ASCII field names to be encoded, while RFC 7578
4273 # suggests sending UTF-8 directly. Firefox sends UTF-8, too
4274 content
= b
'Content-Disposition: form-data; name="' + k
+ b
'"\r\n\r\n' + v
+ b
'\r\n'
4275 if boundary
.encode('ascii') in content
:
4276 raise ValueError('Boundary overlaps with data')
4279 out
+= b
'--' + boundary
.encode('ascii') + b
'--\r\n'
4281 return out
, content_type
4284 def multipart_encode(data
, boundary
=None):
4286 Encode a dict to RFC 7578-compliant form-data
4289 A dict where keys and values can be either Unicode or bytes-like
4292 If specified a Unicode object, it's used as the boundary. Otherwise
4293 a random boundary is generated.
4295 Reference: https://tools.ietf.org/html/rfc7578
4297 has_specified_boundary
= boundary
is not None
4300 if boundary
is None:
4301 boundary
= '---------------' + str(random
.randrange(0x0fffffff, 0xffffffff))
4304 out
, content_type
= _multipart_encode_impl(data
, boundary
)
4307 if has_specified_boundary
:
4311 return out
, content_type
4314 def dict_get(d
, key_or_keys
, default
=None, skip_false_values
=True):
4315 if isinstance(key_or_keys
, (list, tuple)):
4316 for key
in key_or_keys
:
4317 if key
not in d
or d
[key
] is None or skip_false_values
and not d
[key
]:
4321 return d
.get(key_or_keys
, default
)
4324 def try_get(src
, getter
, expected_type
=None):
4325 for get
in variadic(getter
):
4328 except (AttributeError, KeyError, TypeError, IndexError):
4331 if expected_type
is None or isinstance(v
, expected_type
):
4335 def merge_dicts(*dicts
):
4337 for a_dict
in dicts
:
4338 for k
, v
in a_dict
.items():
4342 or (isinstance(v
, compat_str
) and v
4343 and isinstance(merged
[k
], compat_str
)
4344 and not merged
[k
])):
4349 def encode_compat_str(string
, encoding
=preferredencoding(), errors
='strict'):
4350 return string
if isinstance(string
, compat_str
) else compat_str(string
, encoding
, errors
)
4362 TV_PARENTAL_GUIDELINES
= {
4372 def parse_age_limit(s
):
4374 return s
if 0 <= s
<= 21 else None
4375 if not isinstance(s
, compat_basestring
):
4377 m
= re
.match(r
'^(?P<age>\d{1,2})\+?$', s
)
4379 return int(m
.group('age'))
4382 return US_RATINGS
[s
]
4383 m
= re
.match(r
'^TV[_-]?(%s)$' % '|'.join(k
[3:] for k
in TV_PARENTAL_GUIDELINES
), s
)
4385 return TV_PARENTAL_GUIDELINES
['TV-' + m
.group(1)]
4389 def strip_jsonp(code
):
4392 (?:window\.)?(?P<func_name>[a-zA-Z0-9_.$]*)
4393 (?:\s*&&\s*(?P=func_name))?
4394 \s*\(\s*(?P<callback_data>.*)\);?
4395 \s*?(?://[^\n]*)*$''',
4396 r
'\g<callback_data>', code
)
4399 def js_to_json(code
, vars={}):
4400 # vars is a dict of var, val pairs to substitute
4401 COMMENT_RE
= r
'/\*(?:(?!\*/).)*?\*/|//[^\n]*\n'
4402 SKIP_RE
= r
'\s*(?:{comment})?\s*'.format(comment
=COMMENT_RE
)
4404 (r
'(?s)^(0[xX][0-9a-fA-F]+){skip}:?$'.format(skip
=SKIP_RE
), 16),
4405 (r
'(?s)^(0+[0-7]+){skip}:?$'.format(skip
=SKIP_RE
), 8),
4410 if v
in ('true', 'false', 'null'):
4412 elif v
in ('undefined', 'void 0'):
4414 elif v
.startswith('/*') or v
.startswith('//') or v
.startswith('!') or v
== ',':
4417 if v
[0] in ("'", '"'):
4418 v
= re
.sub(r
'(?s)\\.|"', lambda m
: {
4423 }.get(m
.group(0), m
.group(0)), v
[1:-1])
4425 for regex
, base
in INTEGER_TABLE
:
4426 im
= re
.match(regex
, v
)
4428 i
= int(im
.group(1), base
)
4429 return '"%d":' % i
if v
.endswith(':') else '%d' % i
4436 return re
.sub(r
'''(?sx)
4437 "(?:[^"\\]*(?:\\\\|\\['"nurtbfx/\n]))*[^"\\]*"|
4438 '(?:[^'\\]*(?:\\\\|\\['"nurtbfx/\n]))*[^'\\]*'|
4439 {comment}|,(?={skip}[\]}}])|
4440 void\s0|(?:(?<![0-9])[eE]|[a-df-zA-DF-Z_$])[.a-zA-Z_$0-9]*|
4441 \b(?:0[xX][0-9a-fA-F]+|0+[0-7]+)(?:{skip}:)?|
4444 '''.format(comment
=COMMENT_RE
, skip
=SKIP_RE
), fix_kv
, code
)
4447 def qualities(quality_ids
):
4448 """ Get a numeric quality value out of a list of possible values """
4451 return quality_ids
.index(qid
)
4458 'default': '%(title)s [%(id)s].%(ext)s',
4459 'chapter': '%(title)s - %(section_number)03d %(section_title)s [%(id)s].%(ext)s',
4465 'description': 'description',
4466 'annotation': 'annotations.xml',
4467 'infojson': 'info.json',
4468 'pl_thumbnail': None,
4469 'pl_description': 'description',
4470 'pl_infojson': 'info.json',
4473 # As of [1] format syntax is:
4474 # %[mapping_key][conversion_flags][minimum_width][.precision][length_modifier]type
4475 # 1. https://docs.python.org/2/library/stdtypes.html#string-formatting
4476 STR_FORMAT_RE_TMPL
= r
'''(?x)
4477 (?<!%)(?P<prefix>(?:%%)*)
4479 (?P<has_key>\((?P<key>{0})\))?
4481 (?P<conversion>[#0\-+ ]+)?
4483 (?P<precision>\.\d+)?
4484 (?P<len_mod>[hlL])? # unused in python
4485 {1} # conversion type
4490 STR_FORMAT_TYPES
= 'diouxXeEfFgGcrs'
4493 def limit_length(s
, length
):
4494 """ Add ellipses to overly long strings """
4499 return s
[:length
- len(ELLIPSES
)] + ELLIPSES
4503 def version_tuple(v
):
4504 return tuple(int(e
) for e
in re
.split(r
'[-.]', v
))
4507 def is_outdated_version(version
, limit
, assume_new
=True):
4509 return not assume_new
4511 return version_tuple(version
) < version_tuple(limit
)
4513 return not assume_new
4516 def ytdl_is_updateable():
4517 """ Returns if yt-dlp can be updated with -U """
4520 from zipimport
import zipimporter
4522 return isinstance(globals().get('__loader__'), zipimporter
) or hasattr(sys
, 'frozen')
4525 def args_to_str(args
):
4526 # Get a short string representation for a subprocess command
4527 return ' '.join(compat_shlex_quote(a
) for a
in args
)
4530 def error_to_compat_str(err
):
4532 # On python 2 error byte string must be decoded with proper
4533 # encoding rather than ascii
4534 if sys
.version_info
[0] < 3:
4535 err_str
= err_str
.decode(preferredencoding())
4539 def mimetype2ext(mt
):
4545 # Per RFC 3003, audio/mpeg can be .mp1, .mp2 or .mp3. Here use .mp3 as
4546 # it's the most popular one
4547 'audio/mpeg': 'mp3',
4548 'audio/x-wav': 'wav',
4553 _
, _
, res
= mt
.rpartition('/')
4554 res
= res
.split(';')[0].strip().lower()
4558 'smptett+xml': 'tt',
4562 'x-mp4-fragmented': 'mp4',
4563 'x-ms-sami': 'sami',
4566 'x-mpegurl': 'm3u8',
4567 'vnd.apple.mpegurl': 'm3u8',
4571 'vnd.ms-sstr+xml': 'ism',
4578 def parse_codecs(codecs_str
):
4579 # http://tools.ietf.org/html/rfc6381
4582 split_codecs
= list(filter(None, map(
4583 str.strip
, codecs_str
.strip().strip(',').split(','))))
4584 vcodec
, acodec
= None, None
4585 for full_codec
in split_codecs
:
4586 codec
= full_codec
.split('.')[0]
4587 if codec
in ('avc1', 'avc2', 'avc3', 'avc4', 'vp9', 'vp8', 'hev1', 'hev2', 'h263', 'h264', 'mp4v', 'hvc1', 'av01', 'theora'):
4590 elif codec
in ('mp4a', 'opus', 'vorbis', 'mp3', 'aac', 'ac-3', 'ec-3', 'eac3', 'dtsc', 'dtse', 'dtsh', 'dtsl'):
4594 write_string('WARNING: Unknown codec %s\n' % full_codec
, sys
.stderr
)
4595 if not vcodec
and not acodec
:
4596 if len(split_codecs
) == 2:
4598 'vcodec': split_codecs
[0],
4599 'acodec': split_codecs
[1],
4603 'vcodec': vcodec
or 'none',
4604 'acodec': acodec
or 'none',
4609 def urlhandle_detect_ext(url_handle
):
4610 getheader
= url_handle
.headers
.get
4612 cd
= getheader('Content-Disposition')
4614 m
= re
.match(r
'attachment;\s*filename="(?P<filename>[^"]+)"', cd
)
4616 e
= determine_ext(m
.group('filename'), default_ext
=None)
4620 return mimetype2ext(getheader('Content-Type'))
4623 def encode_data_uri(data
, mime_type
):
4624 return 'data:%s;base64,%s' % (mime_type
, base64
.b64encode(data
).decode('ascii'))
4627 def age_restricted(content_limit
, age_limit
):
4628 """ Returns True iff the content should be blocked """
4630 if age_limit
is None: # No limit set
4632 if content_limit
is None:
4633 return False # Content available for everyone
4634 return age_limit
< content_limit
4637 def is_html(first_bytes
):
4638 """ Detect whether a file contains HTML by examining its first bytes. """
4641 (b
'\xef\xbb\xbf', 'utf-8'),
4642 (b
'\x00\x00\xfe\xff', 'utf-32-be'),
4643 (b
'\xff\xfe\x00\x00', 'utf-32-le'),
4644 (b
'\xff\xfe', 'utf-16-le'),
4645 (b
'\xfe\xff', 'utf-16-be'),
4647 for bom
, enc
in BOMS
:
4648 if first_bytes
.startswith(bom
):
4649 s
= first_bytes
[len(bom
):].decode(enc
, 'replace')
4652 s
= first_bytes
.decode('utf-8', 'replace')
4654 return re
.match(r
'^\s*<', s
)
4657 def determine_protocol(info_dict
):
4658 protocol
= info_dict
.get('protocol')
4659 if protocol
is not None:
4662 url
= info_dict
['url']
4663 if url
.startswith('rtmp'):
4665 elif url
.startswith('mms'):
4667 elif url
.startswith('rtsp'):
4670 ext
= determine_ext(url
)
4676 return compat_urllib_parse_urlparse(url
).scheme
4679 def render_table(header_row
, data
, delim
=False, extraGap
=0, hideEmpty
=False):
4680 """ Render a list of rows, each as a list of values """
4682 def get_max_lens(table
):
4683 return [max(len(compat_str(v
)) for v
in col
) for col
in zip(*table
)]
4685 def filter_using_list(row
, filterArray
):
4686 return [col
for (take
, col
) in zip(filterArray
, row
) if take
]
4689 max_lens
= get_max_lens(data
)
4690 header_row
= filter_using_list(header_row
, max_lens
)
4691 data
= [filter_using_list(row
, max_lens
) for row
in data
]
4693 table
= [header_row
] + data
4694 max_lens
= get_max_lens(table
)
4696 table
= [header_row
] + [['-' * ml
for ml
in max_lens
]] + data
4697 format_str
= ' '.join('%-' + compat_str(ml
+ extraGap
) + 's' for ml
in max_lens
[:-1]) + ' %s'
4698 return '\n'.join(format_str
% tuple(row
) for row
in table
)
4701 def _match_one(filter_part
, dct
, incomplete
):
4702 # TODO: Generalize code with YoutubeDL._build_format_filter
4703 STRING_OPERATORS
= {
4704 '*=': operator
.contains
,
4705 '^=': lambda attr
, value
: attr
.startswith(value
),
4706 '$=': lambda attr
, value
: attr
.endswith(value
),
4707 '~=': lambda attr
, value
: re
.search(value
, attr
),
4709 COMPARISON_OPERATORS
= {
4711 '<=': operator
.le
, # "<=" must be defined above "<"
4718 operator_rex
= re
.compile(r
'''(?x)\s*
4720 \s*(?P<negation>!\s*)?(?P<op>%s)(?P<none_inclusive>\s*\?)?\s*
4722 (?P<intval>[0-9.]+(?:[kKmMgGtTpPeEzZyY]i?[Bb]?)?)|
4723 (?P<quote>["\'])(?P
<quotedstrval
>.+?
)(?P
=quote
)|
4727 ''' % '|'.join(map(re.escape, COMPARISON_OPERATORS.keys())))
4728 m = operator_rex.search(filter_part)
4730 unnegated_op = COMPARISON_OPERATORS[m.group('op')]
4731 if m.group('negation'):
4732 op = lambda attr, value: not unnegated_op(attr, value)
4735 actual_value = dct.get(m.group('key'))
4736 if (m.group('quotedstrval') is not None
4737 or m.group('strval') is not None
4738 # If the original field is a string and matching comparisonvalue is
4739 # a number we should respect the origin of the original field
4740 # and process comparison value as a string (see
4741 # https://github.com/ytdl-org/youtube-dl/issues/11082).
4742 or actual_value is not None and m.group('intval') is not None
4743 and isinstance(actual_value, compat_str)):
4744 comparison_value = m.group('quotedstrval') or m.group('strval') or m.group('intval')
4745 quote = m.group('quote')
4746 if quote is not None:
4747 comparison_value = comparison_value.replace(r'\%s' % quote, quote)
4749 if m.group('op') in STRING_OPERATORS:
4750 raise ValueError('Operator %s only supports string values!' % m.group('op'))
4752 comparison_value = int(m.group('intval'))
4754 comparison_value = parse_filesize(m.group('intval'))
4755 if comparison_value is None:
4756 comparison_value = parse_filesize(m.group('intval') + 'B')
4757 if comparison_value is None:
4759 'Invalid integer value %r in filter part %r' % (
4760 m.group('intval'), filter_part))
4761 if actual_value is None:
4762 return incomplete or m.group('none_inclusive')
4763 return op(actual_value, comparison_value)
4766 '': lambda v: (v is True) if isinstance(v, bool) else (v is not None),
4767 '!': lambda v: (v is False) if isinstance(v, bool) else (v is None),
4769 operator_rex = re.compile(r'''(?x
)\s
*
4770 (?P
<op
>%s)\s
*(?P
<key
>[a
-z_
]+)
4772 ''' % '|'.join(map(re.escape, UNARY_OPERATORS.keys())))
4773 m = operator_rex.search(filter_part)
4775 op = UNARY_OPERATORS[m.group('op')]
4776 actual_value = dct.get(m.group('key'))
4777 if incomplete and actual_value is None:
4779 return op(actual_value)
4781 raise ValueError('Invalid filter part %r' % filter_part)
4784 def match_str(filter_str, dct, incomplete=False):
4785 """ Filter a dictionary with a simple string syntax. Returns True (=passes filter) or false
4786 When incomplete, all conditions passes on missing fields
4789 _match_one(filter_part.replace(r'\&', '&'), dct, incomplete)
4790 for filter_part in re.split(r'(?<!\\)&', filter_str))
4793 def match_filter_func(filter_str):
4794 def _match_func(info_dict, *args, **kwargs):
4795 if match_str(filter_str, info_dict, *args, **kwargs):
4798 video_title = info_dict.get('title', info_dict.get('id', 'video'))
4799 return '%s does not pass filter %s, skipping ..' % (video_title, filter_str)
4803 def parse_dfxp_time_expr(time_expr):
4807 mobj = re.match(r'^(?P<time_offset>\d+(?:\.\d+)?)s?$', time_expr)
4809 return float(mobj.group('time_offset'))
4811 mobj = re.match(r'^(\d+):(\d\d):(\d\d(?:(?:\.|:)\d+)?)$', time_expr)
4813 return 3600 * int(mobj.group(1)) + 60 * int(mobj.group(2)) + float(mobj.group(3).replace(':', '.'))
4816 def srt_subtitles_timecode(seconds):
4817 return '%02d:%02d:%02d,%03d' % (seconds / 3600, (seconds % 3600) / 60, seconds % 60, (seconds % 1) * 1000)
4820 def dfxp2srt(dfxp_data):
4822 @param dfxp_data A
bytes-like
object containing DFXP data
4823 @returns A
unicode object containing converted SRT data
4825 LEGACY_NAMESPACES = (
4826 (b'http://www.w3.org/ns/ttml', [
4827 b'http://www.w3.org/2004/11/ttaf1',
4828 b'http://www.w3.org/2006/04/ttaf1',
4829 b'http://www.w3.org/2006/10/ttaf1',
4831 (b'http://www.w3.org/ns/ttml#styling', [
4832 b'http://www.w3.org/ns/ttml#style',
4836 SUPPORTED_STYLING = [
4845 _x = functools.partial(xpath_with_ns, ns_map={
4846 'xml': 'http://www.w3.org/XML/1998/namespace',
4847 'ttml': 'http://www.w3.org/ns/ttml',
4848 'tts': 'http://www.w3.org/ns/ttml#styling',
4854 class TTMLPElementParser(object):
4856 _unclosed_elements = []
4857 _applied_styles = []
4859 def start(self, tag, attrib):
4860 if tag in (_x('ttml:br'), 'br'):
4863 unclosed_elements = []
4865 element_style_id = attrib.get('style')
4867 style.update(default_style)
4868 if element_style_id:
4869 style.update(styles.get(element_style_id, {}))
4870 for prop in SUPPORTED_STYLING:
4871 prop_val = attrib.get(_x('tts:' + prop))
4873 style[prop] = prop_val
4876 for k, v in sorted(style.items()):
4877 if self._applied_styles and self._applied_styles[-1].get(k) == v:
4880 font += ' color="%s"' % v
4881 elif k == 'fontSize':
4882 font += ' size="%s"' % v
4883 elif k == 'fontFamily':
4884 font += ' face="%s"' % v
4885 elif k == 'fontWeight' and v == 'bold':
4887 unclosed_elements.append('b')
4888 elif k == 'fontStyle' and v == 'italic':
4890 unclosed_elements.append('i')
4891 elif k == 'textDecoration' and v == 'underline':
4893 unclosed_elements.append('u')
4895 self._out += '<font' + font + '>'
4896 unclosed_elements.append('font')
4898 if self._applied_styles:
4899 applied_style.update(self._applied_styles[-1])
4900 applied_style.update(style)
4901 self._applied_styles.append(applied_style)
4902 self._unclosed_elements.append(unclosed_elements)
4905 if tag not in (_x('ttml:br'), 'br'):
4906 unclosed_elements = self._unclosed_elements.pop()
4907 for element in reversed(unclosed_elements):
4908 self._out += '</%s>' % element
4909 if unclosed_elements and self._applied_styles:
4910 self._applied_styles.pop()
4912 def data(self, data):
4916 return self._out.strip()
4918 def parse_node(node):
4919 target = TTMLPElementParser()
4920 parser = xml.etree.ElementTree.XMLParser(target=target)
4921 parser.feed(xml.etree.ElementTree.tostring(node))
4922 return parser.close()
4924 for k, v in LEGACY_NAMESPACES:
4926 dfxp_data = dfxp_data.replace(ns, k)
4928 dfxp = compat_etree_fromstring(dfxp_data)
4930 paras = dfxp.findall(_x('.//ttml:p')) or dfxp.findall('.//p')
4933 raise ValueError('Invalid dfxp/TTML subtitle')
4937 for style in dfxp.findall(_x('.//ttml:style')):
4938 style_id = style.get('id') or style.get(_x('xml:id'))
4941 parent_style_id = style.get('style')
4943 if parent_style_id not in styles:
4946 styles[style_id] = styles[parent_style_id].copy()
4947 for prop in SUPPORTED_STYLING:
4948 prop_val = style.get(_x('tts:' + prop))
4950 styles.setdefault(style_id, {})[prop] = prop_val
4956 for p in ('body', 'div'):
4957 ele = xpath_element(dfxp, [_x('.//ttml:' + p), './/' + p])
4960 style = styles.get(ele.get('style'))
4963 default_style.update(style)
4965 for para, index in zip(paras, itertools.count(1)):
4966 begin_time = parse_dfxp_time_expr(para.attrib.get('begin'))
4967 end_time = parse_dfxp_time_expr(para.attrib.get('end'))
4968 dur = parse_dfxp_time_expr(para.attrib.get('dur'))
4969 if begin_time is None:
4974 end_time = begin_time + dur
4975 out.append('%d\n%s --> %s\n%s\n\n' % (
4977 srt_subtitles_timecode(begin_time),
4978 srt_subtitles_timecode(end_time),
4984 def cli_option(params, command_option, param):
4985 param = params.get(param)
4987 param = compat_str(param)
4988 return [command_option, param] if param is not None else []
4991 def cli_bool_option(params, command_option, param, true_value='true', false_value='false', separator=None):
4992 param = params.get(param)
4995 assert isinstance(param, bool)
4997 return [command_option + separator + (true_value if param else false_value)]
4998 return [command_option, true_value if param else false_value]
5001 def cli_valueless_option(params, command_option, param, expected_value=True):
5002 param = params.get(param)
5003 return [command_option] if param == expected_value else []
5006 def cli_configuration_args(argdict, keys, default=[], use_compat=True):
5007 if isinstance(argdict, (list, tuple)): # for backward compatibility
5014 assert isinstance(argdict, dict)
5016 assert isinstance(keys, (list, tuple))
5017 for key_list in keys:
5018 arg_list = list(filter(
5019 lambda x: x is not None,
5020 [argdict.get(key.lower()) for key in variadic(key_list)]))
5022 return [arg for args in arg_list for arg in args]
5026 def _configuration_args(main_key, argdict, exe, keys=None, default=[], use_compat=True):
5027 main_key, exe = main_key.lower(), exe.lower()
5028 root_key = exe if main_key == exe else f'{main_key}+{exe}'
5029 keys = [f'{root_key}{k}' for k in (keys or [''])]
5030 if root_key in keys:
5032 keys.append((main_key, exe))
5033 keys.append('default')
5036 return cli_configuration_args(argdict, keys, default, use_compat)
5039 class ISO639Utils(object):
5040 # See http://www.loc.gov/standards/iso639-2/ISO-639-2_utf-8.txt
5099 'iw': 'heb', # Replaced by he in 1989 revision
5109 'in': 'ind', # Replaced by id in 1989 revision
5224 'ji': 'yid', # Replaced by yi in 1989 revision
5232 def short2long(cls, code):
5233 """Convert language code from ISO 639-1 to ISO 639-2/T"""
5234 return cls._lang_map.get(code[:2])
5237 def long2short(cls, code):
5238 """Convert language code from ISO 639-2/T to ISO 639-1"""
5239 for short_name, long_name in cls._lang_map.items():
5240 if long_name == code:
5244 class ISO3166Utils(object):
5245 # From http://data.okfn.org/data/core/country-list
5247 'AF': 'Afghanistan',
5248 'AX': 'Åland Islands',
5251 'AS': 'American Samoa',
5256 'AG': 'Antigua and Barbuda',
5273 'BO': 'Bolivia, Plurinational State of',
5274 'BQ': 'Bonaire, Sint Eustatius and Saba',
5275 'BA': 'Bosnia and Herzegovina',
5277 'BV': 'Bouvet Island',
5279 'IO': 'British Indian Ocean Territory',
5280 'BN': 'Brunei Darussalam',
5282 'BF': 'Burkina Faso',
5288 'KY': 'Cayman Islands',
5289 'CF': 'Central African Republic',
5293 'CX': 'Christmas Island',
5294 'CC': 'Cocos (Keeling) Islands',
5298 'CD': 'Congo, the Democratic Republic of the',
5299 'CK': 'Cook Islands',
5301 'CI': 'Côte d\'Ivoire',
5306 'CZ': 'Czech Republic',
5310 'DO': 'Dominican Republic',
5313 'SV': 'El Salvador',
5314 'GQ': 'Equatorial Guinea',
5318 'FK': 'Falkland Islands (Malvinas)',
5319 'FO': 'Faroe Islands',
5323 'GF': 'French Guiana',
5324 'PF': 'French Polynesia',
5325 'TF': 'French Southern Territories',
5340 'GW': 'Guinea-Bissau',
5343 'HM': 'Heard Island and McDonald Islands',
5344 'VA': 'Holy See (Vatican City State)',
5351 'IR': 'Iran, Islamic Republic of',
5354 'IM': 'Isle of Man',
5364 'KP': 'Korea, Democratic People\'s Republic of',
5365 'KR': 'Korea, Republic of',
5368 'LA': 'Lao People\'s Democratic Republic',
5374 'LI': 'Liechtenstein',
5378 'MK': 'Macedonia, the Former Yugoslav Republic of',
5385 'MH': 'Marshall Islands',
5391 'FM': 'Micronesia, Federated States of',
5392 'MD': 'Moldova, Republic of',
5403 'NL': 'Netherlands',
5404 'NC': 'New Caledonia',
5405 'NZ': 'New Zealand',
5410 'NF': 'Norfolk Island',
5411 'MP': 'Northern Mariana Islands',
5416 'PS': 'Palestine, State of',
5418 'PG': 'Papua New Guinea',
5421 'PH': 'Philippines',
5425 'PR': 'Puerto Rico',
5429 'RU': 'Russian Federation',
5431 'BL': 'Saint Barthélemy',
5432 'SH': 'Saint Helena, Ascension and Tristan da Cunha',
5433 'KN': 'Saint Kitts and Nevis',
5434 'LC': 'Saint Lucia',
5435 'MF': 'Saint Martin (French part)',
5436 'PM': 'Saint Pierre and Miquelon',
5437 'VC': 'Saint Vincent and the Grenadines',
5440 'ST': 'Sao Tome and Principe',
5441 'SA': 'Saudi Arabia',
5445 'SL': 'Sierra Leone',
5447 'SX': 'Sint Maarten (Dutch part)',
5450 'SB': 'Solomon Islands',
5452 'ZA': 'South Africa',
5453 'GS': 'South Georgia and the South Sandwich Islands',
5454 'SS': 'South Sudan',
5459 'SJ': 'Svalbard and Jan Mayen',
5462 'CH': 'Switzerland',
5463 'SY': 'Syrian Arab Republic',
5464 'TW': 'Taiwan, Province of China',
5466 'TZ': 'Tanzania, United Republic of',
5468 'TL': 'Timor-Leste',
5472 'TT': 'Trinidad and Tobago',
5475 'TM': 'Turkmenistan',
5476 'TC': 'Turks and Caicos Islands',
5480 'AE': 'United Arab Emirates',
5481 'GB': 'United Kingdom',
5482 'US': 'United States',
5483 'UM': 'United States Minor Outlying Islands',
5487 'VE': 'Venezuela, Bolivarian Republic of',
5489 'VG': 'Virgin Islands, British',
5490 'VI': 'Virgin Islands, U.S.',
5491 'WF': 'Wallis and Futuna',
5492 'EH': 'Western Sahara',
5499 def short2full(cls, code):
5500 """Convert an ISO 3166-2 country code to the corresponding full name"""
5501 return cls._country_map.get(code.upper())
5504 class GeoUtils(object):
5505 # Major IPv4 address blocks per country
5507 'AD': '46.172.224.0/19',
5508 'AE': '94.200.0.0/13',
5509 'AF': '149.54.0.0/17',
5510 'AG': '209.59.64.0/18',
5511 'AI': '204.14.248.0/21',
5512 'AL': '46.99.0.0/16',
5513 'AM': '46.70.0.0/15',
5514 'AO': '105.168.0.0/13',
5515 'AP': '182.50.184.0/21',
5516 'AQ': '23.154.160.0/24',
5517 'AR': '181.0.0.0/12',
5518 'AS': '202.70.112.0/20',
5519 'AT': '77.116.0.0/14',
5520 'AU': '1.128.0.0/11',
5521 'AW': '181.41.0.0/18',
5522 'AX': '185.217.4.0/22',
5523 'AZ': '5.197.0.0/16',
5524 'BA': '31.176.128.0/17',
5525 'BB': '65.48.128.0/17',
5526 'BD': '114.130.0.0/16',
5528 'BF': '102.178.0.0/15',
5529 'BG': '95.42.0.0/15',
5530 'BH': '37.131.0.0/17',
5531 'BI': '154.117.192.0/18',
5532 'BJ': '137.255.0.0/16',
5533 'BL': '185.212.72.0/23',
5534 'BM': '196.12.64.0/18',
5535 'BN': '156.31.0.0/16',
5536 'BO': '161.56.0.0/16',
5537 'BQ': '161.0.80.0/20',
5538 'BR': '191.128.0.0/12',
5539 'BS': '24.51.64.0/18',
5540 'BT': '119.2.96.0/19',
5541 'BW': '168.167.0.0/16',
5542 'BY': '178.120.0.0/13',
5543 'BZ': '179.42.192.0/18',
5544 'CA': '99.224.0.0/11',
5545 'CD': '41.243.0.0/16',
5546 'CF': '197.242.176.0/21',
5547 'CG': '160.113.0.0/16',
5548 'CH': '85.0.0.0/13',
5549 'CI': '102.136.0.0/14',
5550 'CK': '202.65.32.0/19',
5551 'CL': '152.172.0.0/14',
5552 'CM': '102.244.0.0/14',
5553 'CN': '36.128.0.0/10',
5554 'CO': '181.240.0.0/12',
5555 'CR': '201.192.0.0/12',
5556 'CU': '152.206.0.0/15',
5557 'CV': '165.90.96.0/19',
5558 'CW': '190.88.128.0/17',
5559 'CY': '31.153.0.0/16',
5560 'CZ': '88.100.0.0/14',
5562 'DJ': '197.241.0.0/17',
5563 'DK': '87.48.0.0/12',
5564 'DM': '192.243.48.0/20',
5565 'DO': '152.166.0.0/15',
5566 'DZ': '41.96.0.0/12',
5567 'EC': '186.68.0.0/15',
5568 'EE': '90.190.0.0/15',
5569 'EG': '156.160.0.0/11',
5570 'ER': '196.200.96.0/20',
5571 'ES': '88.0.0.0/11',
5572 'ET': '196.188.0.0/14',
5573 'EU': '2.16.0.0/13',
5574 'FI': '91.152.0.0/13',
5575 'FJ': '144.120.0.0/16',
5576 'FK': '80.73.208.0/21',
5577 'FM': '119.252.112.0/20',
5578 'FO': '88.85.32.0/19',
5580 'GA': '41.158.0.0/15',
5582 'GD': '74.122.88.0/21',
5583 'GE': '31.146.0.0/16',
5584 'GF': '161.22.64.0/18',
5585 'GG': '62.68.160.0/19',
5586 'GH': '154.160.0.0/12',
5587 'GI': '95.164.0.0/16',
5588 'GL': '88.83.0.0/19',
5589 'GM': '160.182.0.0/15',
5590 'GN': '197.149.192.0/18',
5591 'GP': '104.250.0.0/19',
5592 'GQ': '105.235.224.0/20',
5593 'GR': '94.64.0.0/13',
5594 'GT': '168.234.0.0/16',
5595 'GU': '168.123.0.0/16',
5596 'GW': '197.214.80.0/20',
5597 'GY': '181.41.64.0/18',
5598 'HK': '113.252.0.0/14',
5599 'HN': '181.210.0.0/16',
5600 'HR': '93.136.0.0/13',
5601 'HT': '148.102.128.0/17',
5602 'HU': '84.0.0.0/14',
5603 'ID': '39.192.0.0/10',
5604 'IE': '87.32.0.0/12',
5605 'IL': '79.176.0.0/13',
5606 'IM': '5.62.80.0/20',
5607 'IN': '117.192.0.0/10',
5608 'IO': '203.83.48.0/21',
5609 'IQ': '37.236.0.0/14',
5610 'IR': '2.176.0.0/12',
5611 'IS': '82.221.0.0/16',
5612 'IT': '79.0.0.0/10',
5613 'JE': '87.244.64.0/18',
5614 'JM': '72.27.0.0/17',
5615 'JO': '176.29.0.0/16',
5616 'JP': '133.0.0.0/8',
5617 'KE': '105.48.0.0/12',
5618 'KG': '158.181.128.0/17',
5619 'KH': '36.37.128.0/17',
5620 'KI': '103.25.140.0/22',
5621 'KM': '197.255.224.0/20',
5622 'KN': '198.167.192.0/19',
5623 'KP': '175.45.176.0/22',
5624 'KR': '175.192.0.0/10',
5625 'KW': '37.36.0.0/14',
5626 'KY': '64.96.0.0/15',
5627 'KZ': '2.72.0.0/13',
5628 'LA': '115.84.64.0/18',
5629 'LB': '178.135.0.0/16',
5630 'LC': '24.92.144.0/20',
5631 'LI': '82.117.0.0/19',
5632 'LK': '112.134.0.0/15',
5633 'LR': '102.183.0.0/16',
5634 'LS': '129.232.0.0/17',
5635 'LT': '78.56.0.0/13',
5636 'LU': '188.42.0.0/16',
5637 'LV': '46.109.0.0/16',
5638 'LY': '41.252.0.0/14',
5639 'MA': '105.128.0.0/11',
5640 'MC': '88.209.64.0/18',
5641 'MD': '37.246.0.0/16',
5642 'ME': '178.175.0.0/17',
5643 'MF': '74.112.232.0/21',
5644 'MG': '154.126.0.0/17',
5645 'MH': '117.103.88.0/21',
5646 'MK': '77.28.0.0/15',
5647 'ML': '154.118.128.0/18',
5648 'MM': '37.111.0.0/17',
5649 'MN': '49.0.128.0/17',
5650 'MO': '60.246.0.0/16',
5651 'MP': '202.88.64.0/20',
5652 'MQ': '109.203.224.0/19',
5653 'MR': '41.188.64.0/18',
5654 'MS': '208.90.112.0/22',
5655 'MT': '46.11.0.0/16',
5656 'MU': '105.16.0.0/12',
5657 'MV': '27.114.128.0/18',
5658 'MW': '102.70.0.0/15',
5659 'MX': '187.192.0.0/11',
5660 'MY': '175.136.0.0/13',
5661 'MZ': '197.218.0.0/15',
5662 'NA': '41.182.0.0/16',
5663 'NC': '101.101.0.0/18',
5664 'NE': '197.214.0.0/18',
5665 'NF': '203.17.240.0/22',
5666 'NG': '105.112.0.0/12',
5667 'NI': '186.76.0.0/15',
5668 'NL': '145.96.0.0/11',
5669 'NO': '84.208.0.0/13',
5670 'NP': '36.252.0.0/15',
5671 'NR': '203.98.224.0/19',
5672 'NU': '49.156.48.0/22',
5673 'NZ': '49.224.0.0/14',
5674 'OM': '5.36.0.0/15',
5675 'PA': '186.72.0.0/15',
5676 'PE': '186.160.0.0/14',
5677 'PF': '123.50.64.0/18',
5678 'PG': '124.240.192.0/19',
5679 'PH': '49.144.0.0/13',
5680 'PK': '39.32.0.0/11',
5681 'PL': '83.0.0.0/11',
5682 'PM': '70.36.0.0/20',
5683 'PR': '66.50.0.0/16',
5684 'PS': '188.161.0.0/16',
5685 'PT': '85.240.0.0/13',
5686 'PW': '202.124.224.0/20',
5687 'PY': '181.120.0.0/14',
5688 'QA': '37.210.0.0/15',
5689 'RE': '102.35.0.0/16',
5690 'RO': '79.112.0.0/13',
5691 'RS': '93.86.0.0/15',
5692 'RU': '5.136.0.0/13',
5693 'RW': '41.186.0.0/16',
5694 'SA': '188.48.0.0/13',
5695 'SB': '202.1.160.0/19',
5696 'SC': '154.192.0.0/11',
5697 'SD': '102.120.0.0/13',
5698 'SE': '78.64.0.0/12',
5699 'SG': '8.128.0.0/10',
5700 'SI': '188.196.0.0/14',
5701 'SK': '78.98.0.0/15',
5702 'SL': '102.143.0.0/17',
5703 'SM': '89.186.32.0/19',
5704 'SN': '41.82.0.0/15',
5705 'SO': '154.115.192.0/18',
5706 'SR': '186.179.128.0/17',
5707 'SS': '105.235.208.0/21',
5708 'ST': '197.159.160.0/19',
5709 'SV': '168.243.0.0/16',
5710 'SX': '190.102.0.0/20',
5712 'SZ': '41.84.224.0/19',
5713 'TC': '65.255.48.0/20',
5714 'TD': '154.68.128.0/19',
5715 'TG': '196.168.0.0/14',
5716 'TH': '171.96.0.0/13',
5717 'TJ': '85.9.128.0/18',
5718 'TK': '27.96.24.0/21',
5719 'TL': '180.189.160.0/20',
5720 'TM': '95.85.96.0/19',
5721 'TN': '197.0.0.0/11',
5722 'TO': '175.176.144.0/21',
5723 'TR': '78.160.0.0/11',
5724 'TT': '186.44.0.0/15',
5725 'TV': '202.2.96.0/19',
5726 'TW': '120.96.0.0/11',
5727 'TZ': '156.156.0.0/14',
5728 'UA': '37.52.0.0/14',
5729 'UG': '102.80.0.0/13',
5731 'UY': '167.56.0.0/13',
5732 'UZ': '84.54.64.0/18',
5733 'VA': '212.77.0.0/19',
5734 'VC': '207.191.240.0/21',
5735 'VE': '186.88.0.0/13',
5736 'VG': '66.81.192.0/20',
5737 'VI': '146.226.0.0/16',
5738 'VN': '14.160.0.0/11',
5739 'VU': '202.80.32.0/20',
5740 'WF': '117.20.32.0/21',
5741 'WS': '202.4.32.0/19',
5742 'YE': '134.35.0.0/16',
5743 'YT': '41.242.116.0/22',
5744 'ZA': '41.0.0.0/11',
5745 'ZM': '102.144.0.0/13',
5746 'ZW': '102.177.192.0/18',
5750 def random_ipv4(cls, code_or_block):
5751 if len(code_or_block) == 2:
5752 block = cls._country_ip_map.get(code_or_block.upper())
5756 block = code_or_block
5757 addr, preflen = block.split('/')
5758 addr_min = compat_struct_unpack('!L', socket.inet_aton(addr))[0]
5759 addr_max = addr_min | (0xffffffff >> int(preflen))
5760 return compat_str(socket.inet_ntoa(
5761 compat_struct_pack('!L', random.randint(addr_min, addr_max))))
5764 class PerRequestProxyHandler(compat_urllib_request.ProxyHandler):
5765 def __init__(self, proxies=None):
5766 # Set default handlers
5767 for type in ('http', 'https'):
5768 setattr(self, '%s_open' % type,
5769 lambda r, proxy='__noproxy__', type=type, meth=self.proxy_open:
5770 meth(r, proxy, type))
5771 compat_urllib_request.ProxyHandler.__init__(self, proxies)
5773 def proxy_open(self, req, proxy, type):
5774 req_proxy = req.headers.get('Ytdl-request-proxy')
5775 if req_proxy is not None:
5777 del req.headers['Ytdl-request-proxy']
5779 if proxy == '__noproxy__':
5780 return None # No Proxy
5781 if compat_urlparse.urlparse(proxy).scheme.lower() in ('socks', 'socks4', 'socks4a', 'socks5'):
5782 req.add_header('Ytdl-socks-proxy', proxy)
5783 # yt-dlp's http/https handlers do wrapping the socket with socks
5785 return compat_urllib_request.ProxyHandler.proxy_open(
5786 self, req, proxy, type)
5789 # Both long_to_bytes and bytes_to_long are adapted from PyCrypto, which is
5790 # released into Public Domain
5791 # https://github.com/dlitz/pycrypto/blob/master/lib/Crypto/Util/number.py#L387
5793 def long_to_bytes(n, blocksize=0):
5794 """long_to_bytes(n:long, blocksize:int) : string
5795 Convert a long integer to a byte string.
5797 If optional blocksize is given and greater than zero, pad the front of the
5798 byte string with binary zeros so that the length is a multiple of
5801 # after much testing, this algorithm was deemed to be the fastest
5805 s = compat_struct_pack('>I', n & 0xffffffff) + s
5807 # strip off leading zeros
5808 for i in range(len(s)):
5809 if s[i] != b'\000'[0]:
5812 # only happens when n == 0
5816 # add back some pad bytes. this could be done more efficiently w.r.t. the
5817 # de-padding being done above, but sigh...
5818 if blocksize > 0 and len(s) % blocksize:
5819 s = (blocksize - len(s) % blocksize) * b'\000' + s
5823 def bytes_to_long(s):
5824 """bytes_to_long(string) : long
5825 Convert a byte string to a long integer.
5827 This is (essentially) the inverse of long_to_bytes().
5832 extra = (4 - length % 4)
5833 s = b'\000' * extra + s
5834 length = length + extra
5835 for i in range(0, length, 4):
5836 acc = (acc << 32) + compat_struct_unpack('>I', s[i:i + 4])[0]
5840 def ohdave_rsa_encrypt(data, exponent, modulus):
5842 Implement OHDave
's RSA algorithm. See http://www.ohdave.com/rsa/
5845 data: data to encrypt, bytes-like object
5846 exponent, modulus: parameter e and N of RSA algorithm, both integer
5847 Output: hex string of encrypted data
5849 Limitation: supports one block encryption only
5852 payload = int(binascii.hexlify(data[::-1]), 16)
5853 encrypted = pow(payload, exponent, modulus)
5854 return '%x' % encrypted
5857 def pkcs1pad(data, length):
5859 Padding input data with PKCS#1 scheme
5861 @param {int[]} data input data
5862 @param {int} length target length
5863 @returns {int[]} padded data
5865 if len(data) > length - 11:
5866 raise ValueError('Input data too
long for PKCS
#1 padding')
5868 pseudo_random
= [random
.randint(0, 254) for _
in range(length
- len(data
) - 3)]
5869 return [0, 2] + pseudo_random
+ [0] + data
5872 def encode_base_n(num
, n
, table
=None):
5873 FULL_TABLE
= '0123456789abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ'
5875 table
= FULL_TABLE
[:n
]
5878 raise ValueError('base %d exceeds table length %d' % (n
, len(table
)))
5885 ret
= table
[num
% n
] + ret
5890 def decode_packed_codes(code
):
5891 mobj
= re
.search(PACKED_CODES_RE
, code
)
5892 obfuscated_code
, base
, count
, symbols
= mobj
.groups()
5895 symbols
= symbols
.split('|')
5900 base_n_count
= encode_base_n(count
, base
)
5901 symbol_table
[base_n_count
] = symbols
[count
] or base_n_count
5904 r
'\b(\w+)\b', lambda mobj
: symbol_table
[mobj
.group(0)],
5908 def caesar(s
, alphabet
, shift
):
5913 alphabet
[(alphabet
.index(c
) + shift
) % l
] if c
in alphabet
else c
5918 return caesar(s
, r
'''!"#$%&'()*+,-./0123456789:;<=>?@ABCDEFGHIJKLMNOPQRSTUVWXYZ[\]^_`abcdefghijklmnopqrstuvwxyz{|}~''', 47)
5921 def parse_m3u8_attributes(attrib
):
5923 for (key
, val
) in re
.findall(r
'(?P<key>[A-Z0-9-]+)=(?P<val>"[^"]+"|[^",]+)(?:,|$)', attrib
):
5924 if val
.startswith('"'):
5930 def urshift(val
, n
):
5931 return val
>> n
if val
>= 0 else (val
+ 0x100000000) >> n
5934 # Based on png2str() written by @gdkchan and improved by @yokrysty
5935 # Originally posted at https://github.com/ytdl-org/youtube-dl/issues/9706
5936 def decode_png(png_data
):
5937 # Reference: https://www.w3.org/TR/PNG/
5938 header
= png_data
[8:]
5940 if png_data
[:8] != b
'\x89PNG\x0d\x0a\x1a\x0a' or header
[4:8] != b
'IHDR':
5941 raise IOError('Not a valid PNG file.')
5943 int_map
= {1: '>B', 2: '>H', 4: '>I'}
5944 unpack_integer
= lambda x
: compat_struct_unpack(int_map
[len(x
)], x
)[0]
5949 length
= unpack_integer(header
[:4])
5952 chunk_type
= header
[:4]
5955 chunk_data
= header
[:length
]
5956 header
= header
[length
:]
5958 header
= header
[4:] # Skip CRC
5966 ihdr
= chunks
[0]['data']
5968 width
= unpack_integer(ihdr
[:4])
5969 height
= unpack_integer(ihdr
[4:8])
5973 for chunk
in chunks
:
5974 if chunk
['type'] == b
'IDAT':
5975 idat
+= chunk
['data']
5978 raise IOError('Unable to read PNG data.')
5980 decompressed_data
= bytearray(zlib
.decompress(idat
))
5985 def _get_pixel(idx
):
5990 for y
in range(height
):
5991 basePos
= y
* (1 + stride
)
5992 filter_type
= decompressed_data
[basePos
]
5996 pixels
.append(current_row
)
5998 for x
in range(stride
):
5999 color
= decompressed_data
[1 + basePos
+ x
]
6000 basex
= y
* stride
+ x
6005 left
= _get_pixel(basex
- 3)
6007 up
= _get_pixel(basex
- stride
)
6009 if filter_type
== 1: # Sub
6010 color
= (color
+ left
) & 0xff
6011 elif filter_type
== 2: # Up
6012 color
= (color
+ up
) & 0xff
6013 elif filter_type
== 3: # Average
6014 color
= (color
+ ((left
+ up
) >> 1)) & 0xff
6015 elif filter_type
== 4: # Paeth
6021 c
= _get_pixel(basex
- stride
- 3)
6029 if pa
<= pb
and pa
<= pc
:
6030 color
= (color
+ a
) & 0xff
6032 color
= (color
+ b
) & 0xff
6034 color
= (color
+ c
) & 0xff
6036 current_row
.append(color
)
6038 return width
, height
, pixels
6041 def write_xattr(path
, key
, value
):
6042 # This mess below finds the best xattr tool for the job
6044 # try the pyxattr module...
6047 if hasattr(xattr
, 'set'): # pyxattr
6048 # Unicode arguments are not supported in python-pyxattr until
6050 # See https://github.com/ytdl-org/youtube-dl/issues/5498
6051 pyxattr_required_version
= '0.5.0'
6052 if version_tuple(xattr
.__version
__) < version_tuple(pyxattr_required_version
):
6053 # TODO: fallback to CLI tools
6054 raise XAttrUnavailableError(
6055 'python-pyxattr is detected but is too old. '
6056 'yt-dlp requires %s or above while your version is %s. '
6057 'Falling back to other xattr implementations' % (
6058 pyxattr_required_version
, xattr
.__version
__))
6060 setxattr
= xattr
.set
6062 setxattr
= xattr
.setxattr
6065 setxattr(path
, key
, value
)
6066 except EnvironmentError as e
:
6067 raise XAttrMetadataError(e
.errno
, e
.strerror
)
6070 if compat_os_name
== 'nt':
6071 # Write xattrs to NTFS Alternate Data Streams:
6072 # http://en.wikipedia.org/wiki/NTFS#Alternate_data_streams_.28ADS.29
6073 assert ':' not in key
6074 assert os
.path
.exists(path
)
6076 ads_fn
= path
+ ':' + key
6078 with open(ads_fn
, 'wb') as f
:
6080 except EnvironmentError as e
:
6081 raise XAttrMetadataError(e
.errno
, e
.strerror
)
6083 user_has_setfattr
= check_executable('setfattr', ['--version'])
6084 user_has_xattr
= check_executable('xattr', ['-h'])
6086 if user_has_setfattr
or user_has_xattr
:
6088 value
= value
.decode('utf-8')
6089 if user_has_setfattr
:
6090 executable
= 'setfattr'
6091 opts
= ['-n', key
, '-v', value
]
6092 elif user_has_xattr
:
6093 executable
= 'xattr'
6094 opts
= ['-w', key
, value
]
6096 cmd
= ([encodeFilename(executable
, True)]
6097 + [encodeArgument(o
) for o
in opts
]
6098 + [encodeFilename(path
, True)])
6101 p
= subprocess
.Popen(
6102 cmd
, stdout
=subprocess
.PIPE
, stderr
=subprocess
.PIPE
, stdin
=subprocess
.PIPE
)
6103 except EnvironmentError as e
:
6104 raise XAttrMetadataError(e
.errno
, e
.strerror
)
6105 stdout
, stderr
= process_communicate_or_kill(p
)
6106 stderr
= stderr
.decode('utf-8', 'replace')
6107 if p
.returncode
!= 0:
6108 raise XAttrMetadataError(p
.returncode
, stderr
)
6111 # On Unix, and can't find pyxattr, setfattr, or xattr.
6112 if sys
.platform
.startswith('linux'):
6113 raise XAttrUnavailableError(
6114 "Couldn't find a tool to set the xattrs. "
6115 "Install either the python 'pyxattr' or 'xattr' "
6116 "modules, or the GNU 'attr' package "
6117 "(which contains the 'setfattr' tool).")
6119 raise XAttrUnavailableError(
6120 "Couldn't find a tool to set the xattrs. "
6121 "Install either the python 'xattr' module, "
6122 "or the 'xattr' binary.")
6125 def random_birthday(year_field
, month_field
, day_field
):
6126 start_date
= datetime
.date(1950, 1, 1)
6127 end_date
= datetime
.date(1995, 12, 31)
6128 offset
= random
.randint(0, (end_date
- start_date
).days
)
6129 random_date
= start_date
+ datetime
.timedelta(offset
)
6131 year_field
: str(random_date
.year
),
6132 month_field
: str(random_date
.month
),
6133 day_field
: str(random_date
.day
),
6137 # Templates for internet shortcut files, which are plain text files.
6138 DOT_URL_LINK_TEMPLATE
= '''
6143 DOT_WEBLOC_LINK_TEMPLATE
= '''
6144 <?xml version="1.0" encoding="UTF-8"?>
6145 <!DOCTYPE plist PUBLIC "-//Apple//DTD PLIST 1.0//EN" "http://www.apple.com/DTDs/PropertyList-1.0.dtd">
6146 <plist version="1.0">
6149 \t<string>%(url)s</string>
6154 DOT_DESKTOP_LINK_TEMPLATE
= '''
6164 def iri_to_uri(iri
):
6166 Converts an IRI (Internationalized Resource Identifier, allowing Unicode characters) to a URI (Uniform Resource Identifier, ASCII-only).
6168 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.
6171 iri_parts
= compat_urllib_parse_urlparse(iri
)
6173 if '[' in iri_parts
.netloc
:
6174 raise ValueError('IPv6 URIs are not, yet, supported.')
6175 # Querying `.netloc`, when there's only one bracket, also raises a ValueError.
6177 # 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.
6180 if iri_parts
.username
:
6181 net_location
+= compat_urllib_parse_quote(iri_parts
.username
, safe
=r
"!$%&'()*+,~")
6182 if iri_parts
.password
is not None:
6183 net_location
+= ':' + compat_urllib_parse_quote(iri_parts
.password
, safe
=r
"!$%&'()*+,~")
6186 net_location
+= iri_parts
.hostname
.encode('idna').decode('utf-8') # Punycode for Unicode hostnames.
6187 # The 'idna' encoding produces ASCII text.
6188 if iri_parts
.port
is not None and iri_parts
.port
!= 80:
6189 net_location
+= ':' + str(iri_parts
.port
)
6191 return compat_urllib_parse_urlunparse(
6195 compat_urllib_parse_quote_plus(iri_parts
.path
, safe
=r
"!$%&'()*+,/:;=@|~"),
6197 # Unsure about the `safe` argument, since this is a legacy way of handling parameters.
6198 compat_urllib_parse_quote_plus(iri_parts
.params
, safe
=r
"!$%&'()*+,/:;=@|~"),
6200 # Not totally sure about the `safe` argument, since the source does not explicitly mention the query URI component.
6201 compat_urllib_parse_quote_plus(iri_parts
.query
, safe
=r
"!$%&'()*+,/:;=?@{|}~"),
6203 compat_urllib_parse_quote_plus(iri_parts
.fragment
, safe
=r
"!#$%&'()*+,/:;=?@{|}~")))
6205 # Source for `safe` arguments: https://url.spec.whatwg.org/#percent-encoded-bytes.
6208 def to_high_limit_path(path
):
6209 if sys
.platform
in ['win32', 'cygwin']:
6210 # Work around MAX_PATH limitation on Windows. The maximum allowed length for the individual path segments may still be quite limited.
6211 return r
'\\?\ '.rstrip() + os
.path
.abspath(path
)
6216 def format_field(obj
, field
=None, template
='%s', ignore
=(None, ''), default
='', func
=None):
6218 val
= obj
if obj
is not None else default
6220 val
= obj
.get(field
, default
)
6221 if func
and val
not in ignore
:
6223 return template
% val
if val
not in ignore
else default
6226 def clean_podcast_url(url
):
6227 return re
.sub(r
'''(?x)
6231 media\.blubrry\.com| # https://create.blubrry.com/resources/podcast-media-download-statistics/getting-started/
6234 (?:dts|www)\.podtrac\.com/(?:pts/)?redirect\.[0-9a-z]{3,4}| # http://analytics.podtrac.com/how-to-measure
6237 cn\.co| # https://podcorn.com/analytics-prefix/
6238 st\.fm # https://podsights.com/docs/
6243 _HEX_TABLE
= '0123456789abcdef'
6246 def random_uuidv4():
6247 return re
.sub(r
'[xy]', lambda x
: _HEX_TABLE
[random
.randint(0, 15)], 'xxxxxxxx-xxxx-4xxx-yxxx-xxxxxxxxxxxx')
6250 def make_dir(path
, to_screen
=None):
6252 dn
= os
.path
.dirname(path
)
6253 if dn
and not os
.path
.exists(dn
):
6256 except (OSError, IOError) as err
:
6257 if callable(to_screen
) is not None:
6258 to_screen('unable to create directory ' + error_to_compat_str(err
))
6262 def get_executable_path():
6263 from zipimport
import zipimporter
6264 if hasattr(sys
, 'frozen'): # Running from PyInstaller
6265 path
= os
.path
.dirname(sys
.executable
)
6266 elif isinstance(globals().get('__loader__'), zipimporter
): # Running from ZIP
6267 path
= os
.path
.join(os
.path
.dirname(__file__
), '../..')
6269 path
= os
.path
.join(os
.path
.dirname(__file__
), '..')
6270 return os
.path
.abspath(path
)
6273 def load_plugins(name
, suffix
, namespace
):
6274 plugin_info
= [None]
6277 plugin_info
= imp
.find_module(
6278 name
, [os
.path
.join(get_executable_path(), 'ytdlp_plugins')])
6279 plugins
= imp
.load_module(name
, *plugin_info
)
6280 for name
in dir(plugins
):
6281 if name
in namespace
:
6283 if not name
.endswith(suffix
):
6285 klass
= getattr(plugins
, name
)
6286 classes
.append(klass
)
6287 namespace
[name
] = klass
6291 if plugin_info
[0] is not None:
6292 plugin_info
[0].close()
6297 obj
, *path_list
, default
=None, expected_type
=None, get_all
=True,
6298 casesense
=True, is_user_input
=False, traverse_string
=False):
6299 ''' Traverse nested list/dict/tuple
6300 @param path_list A list of paths which are checked one by one.
6301 Each path is a list of keys where each key is a string,
6302 a tuple of strings or "...". When a tuple is given,
6303 all the keys given in the tuple are traversed, and
6304 "..." traverses all the keys in the object
6305 @param default Default value to return
6306 @param expected_type Only accept final value of this type (Can also be any callable)
6307 @param get_all Return all the values obtained from a path or only the first one
6308 @param casesense Whether to consider dictionary keys as case sensitive
6309 @param is_user_input Whether the keys are generated from user input. If True,
6310 strings are converted to int/slice if necessary
6311 @param traverse_string Whether to traverse inside strings. If True, any
6312 non-compatible object will also be converted into a string
6316 _lower
= lambda k
: (k
.lower() if isinstance(k
, str) else k
)
6317 path_list
= (map(_lower
, variadic(path
)) for path
in path_list
)
6319 def _traverse_obj(obj
, path
, _current_depth
=0):
6323 path
= tuple(variadic(path
))
6324 for i
, key
in enumerate(path
):
6325 if isinstance(key
, (list, tuple)):
6326 obj
= [_traverse_obj(obj
, sub_key
, _current_depth
) for sub_key
in key
]
6329 obj
= (obj
.values() if isinstance(obj
, dict)
6330 else obj
if isinstance(obj
, (list, tuple, LazyList
))
6331 else str(obj
) if traverse_string
else [])
6333 depth
= max(depth
, _current_depth
)
6334 return [_traverse_obj(inner_obj
, path
[i
+ 1:], _current_depth
) for inner_obj
in obj
]
6335 elif isinstance(obj
, dict) and not (is_user_input
and key
== ':'):
6336 obj
= (obj
.get(key
) if casesense
or (key
in obj
)
6337 else next((v
for k
, v
in obj
.items() if _lower(k
) == key
), None))
6340 key
= (int_or_none(key
) if ':' not in key
6341 else slice(*map(int_or_none
, key
.split(':'))))
6342 if key
== slice(None):
6343 return _traverse_obj(obj
, (..., *path
[i
+ 1:]), _current_depth
)
6344 if not isinstance(key
, (int, slice)):
6346 if not isinstance(obj
, (list, tuple, LazyList
)):
6347 if not traverse_string
:
6356 if isinstance(expected_type
, type):
6357 type_test
= lambda val
: val
if isinstance(val
, expected_type
) else None
6358 elif expected_type
is not None:
6359 type_test
= expected_type
6361 type_test
= lambda val
: val
6363 for path
in path_list
:
6365 val
= _traverse_obj(obj
, path
)
6368 for _
in range(depth
- 1):
6369 val
= itertools
.chain
.from_iterable(v
for v
in val
if v
is not None)
6370 val
= [v
for v
in map(type_test
, val
) if v
is not None]
6372 return val
if get_all
else val
[0]
6374 val
= type_test(val
)
6380 def traverse_dict(dictn
, keys
, casesense
=True):
6381 ''' For backward compatibility. Do not use '''
6382 return traverse_obj(dictn
, keys
, casesense
=casesense
,
6383 is_user_input
=True, traverse_string
=True)
6386 def variadic(x
, allowed_types
=(str, bytes)):
6387 return x
if isinstance(x
, collections
.abc
.Iterable
) and not isinstance(x
, allowed_types
) else (x
,)
6390 # create a JSON Web Signature (jws) with HS256 algorithm
6391 # the resulting format is in JWS Compact Serialization
6392 # implemented following JWT https://www.rfc-editor.org/rfc/rfc7519.html
6393 # implemented following JWS https://www.rfc-editor.org/rfc/rfc7515.html
6394 def jwt_encode_hs256(payload_data
, key
, headers
={}):
6400 header_data
.update(headers
)
6401 header_b64
= base64
.b64encode(json
.dumps(header_data
).encode('utf-8'))
6402 payload_b64
= base64
.b64encode(json
.dumps(payload_data
).encode('utf-8'))
6403 h
= hmac
.new(key
.encode('utf-8'), header_b64
+ b
'.' + payload_b64
, hashlib
.sha256
)
6404 signature_b64
= base64
.b64encode(h
.digest())
6405 token
= header_b64
+ b
'.' + payload_b64
+ b
'.' + signature_b64
projects / yt-dlp.git / blob