4 from __future__
import unicode_literals
37 import xml
.etree
.ElementTree
41 compat_HTMLParseError
,
46 compat_ctypes_WINFUNCTYPE
,
47 compat_etree_fromstring
,
50 compat_html_entities_html5
,
63 compat_urllib_parse_urlencode
,
64 compat_urllib_parse_urlparse
,
65 compat_urllib_parse_urlunparse
,
66 compat_urllib_parse_quote
,
67 compat_urllib_parse_quote_plus
,
68 compat_urllib_parse_unquote_plus
,
69 compat_urllib_request
,
80 def register_socks_protocols():
81 # "Register" SOCKS protocols
82 # In Python < 2.6.5, urlsplit() suffers from bug https://bugs.python.org/issue7904
83 # URLs with protocols not in urlparse.uses_netloc are not handled correctly
84 for scheme
in ('socks', 'socks4', 'socks4a', 'socks5'):
85 if scheme
not in compat_urlparse
.uses_netloc
:
86 compat_urlparse
.uses_netloc
.append(scheme
)
89 # This is not clearly defined otherwise
90 compiled_regex_type
= type(re
.compile(''))
93 def random_user_agent():
94 _USER_AGENT_TPL
= 'Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/%s Safari/537.36'
1673 return _USER_AGENT_TPL
% random
.choice(_CHROME_VERSIONS
)
1677 'User-Agent': random_user_agent(),
1678 'Accept-Charset': 'ISO-8859-1,utf-8;q=0.7,*;q=0.7',
1679 'Accept': 'text/html,application/xhtml+xml,application/xml;q=0.9,*/*;q=0.8',
1680 'Accept-Encoding': 'gzip, deflate',
1681 'Accept-Language': 'en-us,en;q=0.5',
1686 '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',
1690 NO_DEFAULT
= object()
1692 ENGLISH_MONTH_NAMES
= [
1693 'January', 'February', 'March', 'April', 'May', 'June',
1694 'July', 'August', 'September', 'October', 'November', 'December']
1697 'en': ENGLISH_MONTH_NAMES
,
1699 'janvier', 'février', 'mars', 'avril', 'mai', 'juin',
1700 'juillet', 'août', 'septembre', 'octobre', 'novembre', 'décembre'],
1703 KNOWN_EXTENSIONS
= (
1704 'mp4', 'm4a', 'm4p', 'm4b', 'm4r', 'm4v', 'aac',
1705 'flv', 'f4v', 'f4a', 'f4b',
1706 'webm', 'ogg', 'ogv', 'oga', 'ogx', 'spx', 'opus',
1707 'mkv', 'mka', 'mk3d',
1710 'asf', 'wmv', 'wma',
1716 'f4f', 'f4m', 'm3u8', 'smil')
1718 REMUX_EXTENSIONS
= ('mp4', 'mkv', 'flv', 'webm', 'mov', 'avi', 'mp3', 'mka', 'm4a', 'ogg', 'opus')
1720 # needed for sanitizing filenames in restricted mode
1721 ACCENT_CHARS
= dict(zip('ÂÃÄÀÁÅÆÇÈÉÊËÌÍÎÏÐÑÒÓÔÕÖŐØŒÙÚÛÜŰÝÞßàáâãäåæçèéêëìíîïðñòóôõöőøœùúûüűýþÿ',
1722 itertools
.chain('AAAAAA', ['AE'], 'CEEEEIIIIDNOOOOOOO', ['OE'], 'UUUUUY', ['TH', 'ss'],
1723 'aaaaaa', ['ae'], 'ceeeeiiiionooooooo', ['oe'], 'uuuuuy', ['th'], 'y')))
1746 '%Y/%m/%d %H:%M:%S',
1748 '%Y-%m-%d %H:%M:%S',
1749 '%Y-%m-%d %H:%M:%S.%f',
1752 '%Y-%m-%dT%H:%M:%SZ',
1753 '%Y-%m-%dT%H:%M:%S.%fZ',
1754 '%Y-%m-%dT%H:%M:%S.%f0Z',
1755 '%Y-%m-%dT%H:%M:%S',
1756 '%Y-%m-%dT%H:%M:%S.%f',
1758 '%b %d %Y at %H:%M',
1759 '%b %d %Y at %H:%M:%S',
1760 '%B %d %Y at %H:%M',
1761 '%B %d %Y at %H:%M:%S',
1764 DATE_FORMATS_DAY_FIRST
= list(DATE_FORMATS
)
1765 DATE_FORMATS_DAY_FIRST
.extend([
1771 '%d/%m/%Y %H:%M:%S',
1774 DATE_FORMATS_MONTH_FIRST
= list(DATE_FORMATS
)
1775 DATE_FORMATS_MONTH_FIRST
.extend([
1780 '%m/%d/%Y %H:%M:%S',
1783 PACKED_CODES_RE
= r
"}\('(.+)',(\d+),(\d+),'([^']+)'\.split\('\|'\)"
1784 JSON_LD_RE
= r
'(?is)<script[^>]+type=(["\']?
)application
/ld\
+json\
1[^
>]*>(?P
<json_ld
>.+?
)</script
>'
1787 def preferredencoding():
1788 """Get preferred encoding.
1790 Returns the best encoding scheme for the system, based on
1791 locale.getpreferredencoding() and some further tweaks.
1794 pref = locale.getpreferredencoding()
1802 def write_json_file(obj, fn):
1803 """ Encode obj as JSON and write it to fn, atomically if possible """
1805 fn = encodeFilename(fn)
1806 if sys.version_info < (3, 0) and sys.platform != 'win32
':
1807 encoding = get_filesystem_encoding()
1808 # os.path.basename returns a bytes object, but NamedTemporaryFile
1809 # will fail if the filename contains non ascii characters unless we
1810 # use a unicode object
1811 path_basename = lambda f: os.path.basename(fn).decode(encoding)
1812 # the same for os.path.dirname
1813 path_dirname = lambda f: os.path.dirname(fn).decode(encoding)
1815 path_basename = os.path.basename
1816 path_dirname = os.path.dirname
1820 'prefix
': path_basename(fn) + '.',
1821 'dir': path_dirname(fn),
1825 # In Python 2.x, json.dump expects a bytestream.
1826 # In Python 3.x, it writes to a character stream
1827 if sys.version_info < (3, 0):
1832 'encoding
': 'utf
-8',
1835 tf = tempfile.NamedTemporaryFile(**compat_kwargs(args))
1840 if sys.platform == 'win32
':
1841 # Need to remove existing file on Windows, else os.rename raises
1842 # WindowsError or FileExistsError.
1850 os.chmod(tf.name, 0o666 & ~mask)
1853 os.rename(tf.name, fn)
1862 if sys.version_info >= (2, 7):
1863 def find_xpath_attr(node, xpath, key, val=None):
1864 """ Find the xpath xpath[@key=val] """
1865 assert re.match(r'^
[a
-zA
-Z_
-]+$
', key)
1866 expr = xpath + ('[@%s]' % key if val is None else "[@%s='%s']" % (key, val))
1867 return node.find(expr)
1869 def find_xpath_attr(node, xpath, key, val=None):
1870 for f in node.findall(compat_xpath(xpath)):
1871 if key not in f.attrib:
1873 if val is None or f.attrib.get(key) == val:
1877 # On python2.6 the xml.etree.ElementTree.Element methods don't support
1878 # the namespace parameter
1881 def xpath_with_ns(path
, ns_map
):
1882 components
= [c
.split(':') for c
in path
.split('/')]
1884 for c
in components
:
1886 replaced
.append(c
[0])
1889 replaced
.append('{%s}%s' % (ns_map
[ns
], tag
))
1890 return '/'.join(replaced
)
1893 def xpath_element(node
, xpath
, name
=None, fatal
=False, default
=NO_DEFAULT
):
1894 def _find_xpath(xpath
):
1895 return node
.find(compat_xpath(xpath
))
1897 if isinstance(xpath
, (str, compat_str
)):
1898 n
= _find_xpath(xpath
)
1906 if default
is not NO_DEFAULT
:
1909 name
= xpath
if name
is None else name
1910 raise ExtractorError('Could not find XML element %s' % name
)
1916 def xpath_text(node
, xpath
, name
=None, fatal
=False, default
=NO_DEFAULT
):
1917 n
= xpath_element(node
, xpath
, name
, fatal
=fatal
, default
=default
)
1918 if n
is None or n
== default
:
1921 if default
is not NO_DEFAULT
:
1924 name
= xpath
if name
is None else name
1925 raise ExtractorError('Could not find XML element\'s text %s' % name
)
1931 def xpath_attr(node
, xpath
, key
, name
=None, fatal
=False, default
=NO_DEFAULT
):
1932 n
= find_xpath_attr(node
, xpath
, key
)
1934 if default
is not NO_DEFAULT
:
1937 name
= '%s[@%s]' % (xpath
, key
) if name
is None else name
1938 raise ExtractorError('Could not find XML attribute %s' % name
)
1941 return n
.attrib
[key
]
1944 def get_element_by_id(id, html
):
1945 """Return the content of the tag with the specified ID in the passed HTML document"""
1946 return get_element_by_attribute('id', id, html
)
1949 def get_element_by_class(class_name
, html
):
1950 """Return the content of the first tag with the specified class in the passed HTML document"""
1951 retval
= get_elements_by_class(class_name
, html
)
1952 return retval
[0] if retval
else None
1955 def get_element_by_attribute(attribute
, value
, html
, escape_value
=True):
1956 retval
= get_elements_by_attribute(attribute
, value
, html
, escape_value
)
1957 return retval
[0] if retval
else None
1960 def get_elements_by_class(class_name
, html
):
1961 """Return the content of all tags with the specified class in the passed HTML document as a list"""
1962 return get_elements_by_attribute(
1963 'class', r
'[^\'"]*\b%s\b[^\'"]*' % re.escape(class_name),
1964 html, escape_value=False)
1967 def get_elements_by_attribute(attribute, value, html, escape_value=True):
1968 """Return the content of the tag with the specified attribute in the passed HTML document"""
1970 value = re.escape(value) if escape_value else value
1973 for m in re.finditer(r'''(?xs)
1975 (?:\s+[a-zA-Z0-9:._-]+(?:=[a-zA-Z0-9:._-]*|="[^"]*"|='[^
']*'|
))*?
1977 (?:\s+[a-zA-Z0-9:._-]+(?:=[a-zA-Z0-9:._-]*|="[^
"]*"|
='[^']*'|))*?
1981 ''' % (re.escape(attribute), value), html):
1982 res = m.group('content
')
1984 if res.startswith('"') or res.startswith("'"):
1987 retlist.append(unescapeHTML(res))
1992 class HTMLAttributeParser(compat_HTMLParser):
1993 """Trivial HTML parser to gather the attributes for a single element"""
1997 compat_HTMLParser.__init__(self)
1999 def handle_starttag(self, tag, attrs):
2000 self.attrs = dict(attrs)
2003 def extract_attributes(html_element):
2004 """Given a string for an HTML element such as
2006 a="foo" B="bar" c="&98;az" d=boz
2007 empty= noval entity="&"
2010 Decode and return a dictionary of attributes.
2012 'a
': 'foo
', 'b
': 'bar
', c: 'baz
', d: 'boz
',
2013 'empty
': '', 'noval
': None, 'entity
': '&',
2014 'sq
': '"', 'dq': '\''
2016 NB HTMLParser is stricter in Python 2.6 & 3.2 than in later versions,
2017 but the cases in the unit test will work for all of 2.6, 2.7, 3.2-3.5.
2019 parser = HTMLAttributeParser()
2021 parser.feed(html_element)
2023 # Older Python may throw HTMLParseError in case of malformed HTML
2024 except compat_HTMLParseError:
2029 def clean_html(html):
2030 """Clean an HTML snippet into a readable string"""
2032 if html is None: # Convenience for sanitizing descriptions etc.
2036 html = html.replace('\n', ' ')
2037 html = re.sub(r'(?u)\s*<\s*br\s*/?\s*>\s*', '\n', html)
2038 html = re.sub(r'(?u)<\s*/\s*p\s*>\s*<\s*p[^>]*>', '\n', html)
2040 html = re.sub('<.*?>', '', html)
2041 # Replace html entities
2042 html = unescapeHTML(html)
2046 def sanitize_open(filename, open_mode):
2047 """Try to open the given filename, and slightly tweak it if this fails.
2049 Attempts to open the given filename. If this fails, it tries to change
2050 the filename slightly, step by step, until it's either able to open it
2051 or it fails and raises a final exception, like the standard open()
2054 It returns the tuple (stream, definitive_file_name).
2058 if sys.platform == 'win32':
2060 msvcrt.setmode(sys.stdout.fileno(), os.O_BINARY)
2061 return (sys.stdout.buffer if hasattr(sys.stdout, 'buffer') else sys.stdout, filename)
2062 stream = open(encodeFilename(filename), open_mode)
2063 return (stream, filename)
2064 except (IOError, OSError) as err:
2065 if err.errno in (errno.EACCES,):
2068 # In case of error, try to remove win32 forbidden chars
2069 alt_filename = sanitize_path(filename)
2070 if alt_filename == filename:
2073 # An exception here should be caught in the caller
2074 stream = open(encodeFilename(alt_filename), open_mode)
2075 return (stream, alt_filename)
2078 def timeconvert(timestr):
2079 """Convert RFC 2822 defined time string into system timestamp"""
2081 timetuple = email.utils.parsedate_tz(timestr)
2082 if timetuple is not None:
2083 timestamp = email.utils.mktime_tz(timetuple)
2087 def sanitize_filename(s, restricted=False, is_id=False):
2088 """Sanitizes a string so it could be used as part of a filename.
2089 If restricted is set, use a stricter subset of allowed characters.
2090 Set is_id if this is not an arbitrary string, but an ID that should be kept
2093 def replace_insane(char):
2094 if restricted and char in ACCENT_CHARS:
2095 return ACCENT_CHARS[char]
2096 if char == '?' or ord(char) < 32 or ord(char) == 127:
2099 return '' if restricted else '\''
2101 return '_
-' if restricted else ' -'
2102 elif char in '\\/|
*<>':
2104 if restricted and (char in '!&\'()[]{}$
;`^
,#' or char.isspace()):
2106 if restricted
and ord(char
) > 127:
2111 s
= re
.sub(r
'[0-9]+(?::[0-9]+)+', lambda m
: m
.group(0).replace(':', '_'), s
)
2112 result
= ''.join(map(replace_insane
, s
))
2114 while '__' in result
:
2115 result
= result
.replace('__', '_')
2116 result
= result
.strip('_')
2117 # Common case of "Foreign band name - English song title"
2118 if restricted
and result
.startswith('-_'):
2120 if result
.startswith('-'):
2121 result
= '_' + result
[len('-'):]
2122 result
= result
.lstrip('.')
2128 def sanitize_path(s
, force
=False):
2129 """Sanitizes and normalizes path on Windows"""
2130 if sys
.platform
== 'win32':
2132 drive_or_unc
, _
= os
.path
.splitdrive(s
)
2133 if sys
.version_info
< (2, 7) and not drive_or_unc
:
2134 drive_or_unc
, _
= os
.path
.splitunc(s
)
2140 norm_path
= os
.path
.normpath(remove_start(s
, drive_or_unc
)).split(os
.path
.sep
)
2144 path_part
if path_part
in ['.', '..'] else re
.sub(r
'(?:[/<>:"\|\\?\*]|[\s.]$)', '#', path_part
)
2145 for path_part
in norm_path
]
2147 sanitized_path
.insert(0, drive_or_unc
+ os
.path
.sep
)
2148 elif force
and s
[0] == os
.path
.sep
:
2149 sanitized_path
.insert(0, os
.path
.sep
)
2150 return os
.path
.join(*sanitized_path
)
2153 def sanitize_url(url
):
2154 # Prepend protocol-less URLs with `http:` scheme in order to mitigate
2155 # the number of unwanted failures due to missing protocol
2156 if url
.startswith('//'):
2157 return 'http:%s' % url
2158 # Fix some common typos seen so far
2160 # https://github.com/ytdl-org/youtube-dl/issues/15649
2161 (r
'^httpss://', r
'https://'),
2162 # https://bx1.be/lives/direct-tv/
2163 (r
'^rmtp([es]?)://', r
'rtmp\1://'),
2165 for mistake
, fixup
in COMMON_TYPOS
:
2166 if re
.match(mistake
, url
):
2167 return re
.sub(mistake
, fixup
, url
)
2171 def sanitized_Request(url
, *args
, **kwargs
):
2172 return compat_urllib_request
.Request(sanitize_url(url
), *args
, **kwargs
)
2176 """Expand shell variables and ~"""
2177 return os
.path
.expandvars(compat_expanduser(s
))
2180 def orderedSet(iterable
):
2181 """ Remove all duplicates from the input iterable """
2189 def _htmlentity_transform(entity_with_semicolon
):
2190 """Transforms an HTML entity to a character."""
2191 entity
= entity_with_semicolon
[:-1]
2193 # Known non-numeric HTML entity
2194 if entity
in compat_html_entities
.name2codepoint
:
2195 return compat_chr(compat_html_entities
.name2codepoint
[entity
])
2197 # TODO: HTML5 allows entities without a semicolon. For example,
2198 # 'Éric' should be decoded as 'Éric'.
2199 if entity_with_semicolon
in compat_html_entities_html5
:
2200 return compat_html_entities_html5
[entity_with_semicolon
]
2202 mobj
= re
.match(r
'#(x[0-9a-fA-F]+|[0-9]+)', entity
)
2203 if mobj
is not None:
2204 numstr
= mobj
.group(1)
2205 if numstr
.startswith('x'):
2207 numstr
= '0%s' % numstr
2210 # See https://github.com/ytdl-org/youtube-dl/issues/7518
2212 return compat_chr(int(numstr
, base
))
2216 # Unknown entity in name, return its literal representation
2217 return '&%s;' % entity
2220 def unescapeHTML(s
):
2223 assert type(s
) == compat_str
2226 r
'&([^&;]+;)', lambda m
: _htmlentity_transform(m
.group(1)), s
)
2229 def process_communicate_or_kill(p
, *args
, **kwargs
):
2231 return p
.communicate(*args
, **kwargs
)
2232 except BaseException
: # Including KeyboardInterrupt
2238 def get_subprocess_encoding():
2239 if sys
.platform
== 'win32' and sys
.getwindowsversion()[0] >= 5:
2240 # For subprocess calls, encode with locale encoding
2241 # Refer to http://stackoverflow.com/a/9951851/35070
2242 encoding
= preferredencoding()
2244 encoding
= sys
.getfilesystemencoding()
2245 if encoding
is None:
2250 def encodeFilename(s
, for_subprocess
=False):
2252 @param s The name of the file
2255 assert type(s
) == compat_str
2257 # Python 3 has a Unicode API
2258 if sys
.version_info
>= (3, 0):
2261 # Pass '' directly to use Unicode APIs on Windows 2000 and up
2262 # (Detecting Windows NT 4 is tricky because 'major >= 4' would
2263 # match Windows 9x series as well. Besides, NT 4 is obsolete.)
2264 if not for_subprocess
and sys
.platform
== 'win32' and sys
.getwindowsversion()[0] >= 5:
2267 # Jython assumes filenames are Unicode strings though reported as Python 2.x compatible
2268 if sys
.platform
.startswith('java'):
2271 return s
.encode(get_subprocess_encoding(), 'ignore')
2274 def decodeFilename(b
, for_subprocess
=False):
2276 if sys
.version_info
>= (3, 0):
2279 if not isinstance(b
, bytes):
2282 return b
.decode(get_subprocess_encoding(), 'ignore')
2285 def encodeArgument(s
):
2286 if not isinstance(s
, compat_str
):
2287 # Legacy code that uses byte strings
2288 # Uncomment the following line after fixing all post processors
2289 # assert False, 'Internal error: %r should be of type %r, is %r' % (s, compat_str, type(s))
2290 s
= s
.decode('ascii')
2291 return encodeFilename(s
, True)
2294 def decodeArgument(b
):
2295 return decodeFilename(b
, True)
2298 def decodeOption(optval
):
2301 if isinstance(optval
, bytes):
2302 optval
= optval
.decode(preferredencoding())
2304 assert isinstance(optval
, compat_str
)
2308 def formatSeconds(secs
, delim
=':'):
2310 return '%d%s%02d%s%02d' % (secs
// 3600, delim
, (secs
% 3600) // 60, delim
, secs
% 60)
2312 return '%d%s%02d' % (secs
// 60, delim
, secs
% 60)
2317 def make_HTTPS_handler(params
, **kwargs
):
2318 opts_no_check_certificate
= params
.get('nocheckcertificate', False)
2319 if hasattr(ssl
, 'create_default_context'): # Python >= 3.4 or 2.7.9
2320 context
= ssl
.create_default_context(ssl
.Purpose
.SERVER_AUTH
)
2321 if opts_no_check_certificate
:
2322 context
.check_hostname
= False
2323 context
.verify_mode
= ssl
.CERT_NONE
2325 return YoutubeDLHTTPSHandler(params
, context
=context
, **kwargs
)
2328 # (create_default_context present but HTTPSHandler has no context=)
2331 if sys
.version_info
< (3, 2):
2332 return YoutubeDLHTTPSHandler(params
, **kwargs
)
2333 else: # Python < 3.4
2334 context
= ssl
.SSLContext(ssl
.PROTOCOL_TLSv1
)
2335 context
.verify_mode
= (ssl
.CERT_NONE
2336 if opts_no_check_certificate
2337 else ssl
.CERT_REQUIRED
)
2338 context
.set_default_verify_paths()
2339 return YoutubeDLHTTPSHandler(params
, context
=context
, **kwargs
)
2342 def bug_reports_message():
2343 if ytdl_is_updateable():
2344 update_cmd
= 'type yt-dlp -U to update'
2346 update_cmd
= 'see https://github.com/yt-dlp/yt-dlp on how to update'
2347 msg
= '; please report this issue on https://github.com/yt-dlp/yt-dlp .'
2348 msg
+= ' Make sure you are using the latest version; %s.' % update_cmd
2349 msg
+= ' Be sure to call yt-dlp with the --verbose flag and include its complete output.'
2353 class YoutubeDLError(Exception):
2354 """Base exception for YoutubeDL errors."""
2358 class ExtractorError(YoutubeDLError
):
2359 """Error during info extraction."""
2361 def __init__(self
, msg
, tb
=None, expected
=False, cause
=None, video_id
=None):
2362 """ tb, if given, is the original traceback (so that it can be printed out).
2363 If expected is set, this is a normal error message and most likely not a bug in yt-dlp.
2366 if sys
.exc_info()[0] in (compat_urllib_error
.URLError
, socket
.timeout
, UnavailableVideoError
):
2368 if video_id
is not None:
2369 msg
= video_id
+ ': ' + msg
2371 msg
+= ' (caused by %r)' % cause
2373 msg
+= bug_reports_message()
2374 super(ExtractorError
, self
).__init
__(msg
)
2377 self
.exc_info
= sys
.exc_info() # preserve original exception
2379 self
.video_id
= video_id
2381 def format_traceback(self
):
2382 if self
.traceback
is None:
2384 return ''.join(traceback
.format_tb(self
.traceback
))
2387 class UnsupportedError(ExtractorError
):
2388 def __init__(self
, url
):
2389 super(UnsupportedError
, self
).__init
__(
2390 'Unsupported URL: %s' % url
, expected
=True)
2394 class RegexNotFoundError(ExtractorError
):
2395 """Error when a regex didn't match"""
2399 class GeoRestrictedError(ExtractorError
):
2400 """Geographic restriction Error exception.
2402 This exception may be thrown when a video is not available from your
2403 geographic location due to geographic restrictions imposed by a website.
2406 def __init__(self
, msg
, countries
=None):
2407 super(GeoRestrictedError
, self
).__init
__(msg
, expected
=True)
2409 self
.countries
= countries
2412 class DownloadError(YoutubeDLError
):
2413 """Download Error exception.
2415 This exception may be thrown by FileDownloader objects if they are not
2416 configured to continue on errors. They will contain the appropriate
2420 def __init__(self
, msg
, exc_info
=None):
2421 """ exc_info, if given, is the original exception that caused the trouble (as returned by sys.exc_info()). """
2422 super(DownloadError
, self
).__init
__(msg
)
2423 self
.exc_info
= exc_info
2426 class SameFileError(YoutubeDLError
):
2427 """Same File exception.
2429 This exception will be thrown by FileDownloader objects if they detect
2430 multiple files would have to be downloaded to the same file on disk.
2435 class PostProcessingError(YoutubeDLError
):
2436 """Post Processing exception.
2438 This exception may be raised by PostProcessor's .run() method to
2439 indicate an error in the postprocessing task.
2442 def __init__(self
, msg
):
2443 super(PostProcessingError
, self
).__init
__(msg
)
2447 class ExistingVideoReached(YoutubeDLError
):
2448 """ --max-downloads limit has been reached. """
2452 class RejectedVideoReached(YoutubeDLError
):
2453 """ --max-downloads limit has been reached. """
2457 class MaxDownloadsReached(YoutubeDLError
):
2458 """ --max-downloads limit has been reached. """
2462 class UnavailableVideoError(YoutubeDLError
):
2463 """Unavailable Format exception.
2465 This exception will be thrown when a video is requested
2466 in a format that is not available for that video.
2471 class ContentTooShortError(YoutubeDLError
):
2472 """Content Too Short exception.
2474 This exception may be raised by FileDownloader objects when a file they
2475 download is too small for what the server announced first, indicating
2476 the connection was probably interrupted.
2479 def __init__(self
, downloaded
, expected
):
2480 super(ContentTooShortError
, self
).__init
__(
2481 'Downloaded {0} bytes, expected {1} bytes'.format(downloaded
, expected
)
2484 self
.downloaded
= downloaded
2485 self
.expected
= expected
2488 class XAttrMetadataError(YoutubeDLError
):
2489 def __init__(self
, code
=None, msg
='Unknown error'):
2490 super(XAttrMetadataError
, self
).__init
__(msg
)
2494 # Parsing code and msg
2495 if (self
.code
in (errno
.ENOSPC
, errno
.EDQUOT
)
2496 or 'No space left' in self
.msg
or 'Disk quota exceeded' in self
.msg
):
2497 self
.reason
= 'NO_SPACE'
2498 elif self
.code
== errno
.E2BIG
or 'Argument list too long' in self
.msg
:
2499 self
.reason
= 'VALUE_TOO_LONG'
2501 self
.reason
= 'NOT_SUPPORTED'
2504 class XAttrUnavailableError(YoutubeDLError
):
2508 def _create_http_connection(ydl_handler
, http_class
, is_https
, *args
, **kwargs
):
2509 # Working around python 2 bug (see http://bugs.python.org/issue17849) by limiting
2510 # expected HTTP responses to meet HTTP/1.0 or later (see also
2511 # https://github.com/ytdl-org/youtube-dl/issues/6727)
2512 if sys
.version_info
< (3, 0):
2513 kwargs
['strict'] = True
2514 hc
= http_class(*args
, **compat_kwargs(kwargs
))
2515 source_address
= ydl_handler
._params
.get('source_address')
2517 if source_address
is not None:
2518 # This is to workaround _create_connection() from socket where it will try all
2519 # address data from getaddrinfo() including IPv6. This filters the result from
2520 # getaddrinfo() based on the source_address value.
2521 # This is based on the cpython socket.create_connection() function.
2522 # https://github.com/python/cpython/blob/master/Lib/socket.py#L691
2523 def _create_connection(address
, timeout
=socket
._GLOBAL
_DEFAULT
_TIMEOUT
, source_address
=None):
2524 host
, port
= address
2526 addrs
= socket
.getaddrinfo(host
, port
, 0, socket
.SOCK_STREAM
)
2527 af
= socket
.AF_INET
if '.' in source_address
[0] else socket
.AF_INET6
2528 ip_addrs
= [addr
for addr
in addrs
if addr
[0] == af
]
2529 if addrs
and not ip_addrs
:
2530 ip_version
= 'v4' if af
== socket
.AF_INET
else 'v6'
2532 "No remote IP%s addresses available for connect, can't use '%s' as source address"
2533 % (ip_version
, source_address
[0]))
2534 for res
in ip_addrs
:
2535 af
, socktype
, proto
, canonname
, sa
= res
2538 sock
= socket
.socket(af
, socktype
, proto
)
2539 if timeout
is not socket
._GLOBAL
_DEFAULT
_TIMEOUT
:
2540 sock
.settimeout(timeout
)
2541 sock
.bind(source_address
)
2543 err
= None # Explicitly break reference cycle
2545 except socket
.error
as _
:
2547 if sock
is not None:
2552 raise socket
.error('getaddrinfo returns an empty list')
2553 if hasattr(hc
, '_create_connection'):
2554 hc
._create
_connection
= _create_connection
2555 sa
= (source_address
, 0)
2556 if hasattr(hc
, 'source_address'): # Python 2.7+
2557 hc
.source_address
= sa
2559 def _hc_connect(self
, *args
, **kwargs
):
2560 sock
= _create_connection(
2561 (self
.host
, self
.port
), self
.timeout
, sa
)
2563 self
.sock
= ssl
.wrap_socket(
2564 sock
, self
.key_file
, self
.cert_file
,
2565 ssl_version
=ssl
.PROTOCOL_TLSv1
)
2568 hc
.connect
= functools
.partial(_hc_connect
, hc
)
2573 def handle_youtubedl_headers(headers
):
2574 filtered_headers
= headers
2576 if 'Youtubedl-no-compression' in filtered_headers
:
2577 filtered_headers
= dict((k
, v
) for k
, v
in filtered_headers
.items() if k
.lower() != 'accept-encoding')
2578 del filtered_headers
['Youtubedl-no-compression']
2580 return filtered_headers
2583 class YoutubeDLHandler(compat_urllib_request
.HTTPHandler
):
2584 """Handler for HTTP requests and responses.
2586 This class, when installed with an OpenerDirector, automatically adds
2587 the standard headers to every HTTP request and handles gzipped and
2588 deflated responses from web servers. If compression is to be avoided in
2589 a particular request, the original request in the program code only has
2590 to include the HTTP header "Youtubedl-no-compression", which will be
2591 removed before making the real request.
2593 Part of this code was copied from:
2595 http://techknack.net/python-urllib2-handlers/
2597 Andrew Rowls, the author of that code, agreed to release it to the
2601 def __init__(self
, params
, *args
, **kwargs
):
2602 compat_urllib_request
.HTTPHandler
.__init
__(self
, *args
, **kwargs
)
2603 self
._params
= params
2605 def http_open(self
, req
):
2606 conn_class
= compat_http_client
.HTTPConnection
2608 socks_proxy
= req
.headers
.get('Ytdl-socks-proxy')
2610 conn_class
= make_socks_conn_class(conn_class
, socks_proxy
)
2611 del req
.headers
['Ytdl-socks-proxy']
2613 return self
.do_open(functools
.partial(
2614 _create_http_connection
, self
, conn_class
, False),
2622 return zlib
.decompress(data
, -zlib
.MAX_WBITS
)
2624 return zlib
.decompress(data
)
2626 def http_request(self
, req
):
2627 # According to RFC 3986, URLs can not contain non-ASCII characters, however this is not
2628 # always respected by websites, some tend to give out URLs with non percent-encoded
2629 # non-ASCII characters (see telemb.py, ard.py [#3412])
2630 # urllib chokes on URLs with non-ASCII characters (see http://bugs.python.org/issue3991)
2631 # To work around aforementioned issue we will replace request's original URL with
2632 # percent-encoded one
2633 # Since redirects are also affected (e.g. http://www.southpark.de/alle-episoden/s18e09)
2634 # the code of this workaround has been moved here from YoutubeDL.urlopen()
2635 url
= req
.get_full_url()
2636 url_escaped
= escape_url(url
)
2638 # Substitute URL if any change after escaping
2639 if url
!= url_escaped
:
2640 req
= update_Request(req
, url
=url_escaped
)
2642 for h
, v
in std_headers
.items():
2643 # Capitalize is needed because of Python bug 2275: http://bugs.python.org/issue2275
2644 # The dict keys are capitalized because of this bug by urllib
2645 if h
.capitalize() not in req
.headers
:
2646 req
.add_header(h
, v
)
2648 req
.headers
= handle_youtubedl_headers(req
.headers
)
2650 if sys
.version_info
< (2, 7) and '#' in req
.get_full_url():
2651 # Python 2.6 is brain-dead when it comes to fragments
2652 req
._Request
__original
= req
._Request
__original
.partition('#')[0]
2653 req
._Request
__r
_type
= req
._Request
__r
_type
.partition('#')[0]
2657 def http_response(self
, req
, resp
):
2660 if resp
.headers
.get('Content-encoding', '') == 'gzip':
2661 content
= resp
.read()
2662 gz
= gzip
.GzipFile(fileobj
=io
.BytesIO(content
), mode
='rb')
2664 uncompressed
= io
.BytesIO(gz
.read())
2665 except IOError as original_ioerror
:
2666 # There may be junk add the end of the file
2667 # See http://stackoverflow.com/q/4928560/35070 for details
2668 for i
in range(1, 1024):
2670 gz
= gzip
.GzipFile(fileobj
=io
.BytesIO(content
[:-i
]), mode
='rb')
2671 uncompressed
= io
.BytesIO(gz
.read())
2676 raise original_ioerror
2677 resp
= compat_urllib_request
.addinfourl(uncompressed
, old_resp
.headers
, old_resp
.url
, old_resp
.code
)
2678 resp
.msg
= old_resp
.msg
2679 del resp
.headers
['Content-encoding']
2681 if resp
.headers
.get('Content-encoding', '') == 'deflate':
2682 gz
= io
.BytesIO(self
.deflate(resp
.read()))
2683 resp
= compat_urllib_request
.addinfourl(gz
, old_resp
.headers
, old_resp
.url
, old_resp
.code
)
2684 resp
.msg
= old_resp
.msg
2685 del resp
.headers
['Content-encoding']
2686 # Percent-encode redirect URL of Location HTTP header to satisfy RFC 3986 (see
2687 # https://github.com/ytdl-org/youtube-dl/issues/6457).
2688 if 300 <= resp
.code
< 400:
2689 location
= resp
.headers
.get('Location')
2691 # As of RFC 2616 default charset is iso-8859-1 that is respected by python 3
2692 if sys
.version_info
>= (3, 0):
2693 location
= location
.encode('iso-8859-1').decode('utf-8')
2695 location
= location
.decode('utf-8')
2696 location_escaped
= escape_url(location
)
2697 if location
!= location_escaped
:
2698 del resp
.headers
['Location']
2699 if sys
.version_info
< (3, 0):
2700 location_escaped
= location_escaped
.encode('utf-8')
2701 resp
.headers
['Location'] = location_escaped
2704 https_request
= http_request
2705 https_response
= http_response
2708 def make_socks_conn_class(base_class
, socks_proxy
):
2709 assert issubclass(base_class
, (
2710 compat_http_client
.HTTPConnection
, compat_http_client
.HTTPSConnection
))
2712 url_components
= compat_urlparse
.urlparse(socks_proxy
)
2713 if url_components
.scheme
.lower() == 'socks5':
2714 socks_type
= ProxyType
.SOCKS5
2715 elif url_components
.scheme
.lower() in ('socks', 'socks4'):
2716 socks_type
= ProxyType
.SOCKS4
2717 elif url_components
.scheme
.lower() == 'socks4a':
2718 socks_type
= ProxyType
.SOCKS4A
2720 def unquote_if_non_empty(s
):
2723 return compat_urllib_parse_unquote_plus(s
)
2727 url_components
.hostname
, url_components
.port
or 1080,
2729 unquote_if_non_empty(url_components
.username
),
2730 unquote_if_non_empty(url_components
.password
),
2733 class SocksConnection(base_class
):
2735 self
.sock
= sockssocket()
2736 self
.sock
.setproxy(*proxy_args
)
2737 if type(self
.timeout
) in (int, float):
2738 self
.sock
.settimeout(self
.timeout
)
2739 self
.sock
.connect((self
.host
, self
.port
))
2741 if isinstance(self
, compat_http_client
.HTTPSConnection
):
2742 if hasattr(self
, '_context'): # Python > 2.6
2743 self
.sock
= self
._context
.wrap_socket(
2744 self
.sock
, server_hostname
=self
.host
)
2746 self
.sock
= ssl
.wrap_socket(self
.sock
)
2748 return SocksConnection
2751 class YoutubeDLHTTPSHandler(compat_urllib_request
.HTTPSHandler
):
2752 def __init__(self
, params
, https_conn_class
=None, *args
, **kwargs
):
2753 compat_urllib_request
.HTTPSHandler
.__init
__(self
, *args
, **kwargs
)
2754 self
._https
_conn
_class
= https_conn_class
or compat_http_client
.HTTPSConnection
2755 self
._params
= params
2757 def https_open(self
, req
):
2759 conn_class
= self
._https
_conn
_class
2761 if hasattr(self
, '_context'): # python > 2.6
2762 kwargs
['context'] = self
._context
2763 if hasattr(self
, '_check_hostname'): # python 3.x
2764 kwargs
['check_hostname'] = self
._check
_hostname
2766 socks_proxy
= req
.headers
.get('Ytdl-socks-proxy')
2768 conn_class
= make_socks_conn_class(conn_class
, socks_proxy
)
2769 del req
.headers
['Ytdl-socks-proxy']
2771 return self
.do_open(functools
.partial(
2772 _create_http_connection
, self
, conn_class
, True),
2776 class YoutubeDLCookieJar(compat_cookiejar
.MozillaCookieJar
):
2778 See [1] for cookie file format.
2780 1. https://curl.haxx.se/docs/http-cookies.html
2782 _HTTPONLY_PREFIX
= '#HttpOnly_'
2784 _HEADER
= '''# Netscape HTTP Cookie File
2785 # This file is generated by yt-dlp. Do not edit.
2788 _CookieFileEntry
= collections
.namedtuple(
2790 ('domain_name', 'include_subdomains', 'path', 'https_only', 'expires_at', 'name', 'value'))
2792 def save(self
, filename
=None, ignore_discard
=False, ignore_expires
=False):
2794 Save cookies to a file.
2796 Most of the code is taken from CPython 3.8 and slightly adapted
2797 to support cookie files with UTF-8 in both python 2 and 3.
2799 if filename
is None:
2800 if self
.filename
is not None:
2801 filename
= self
.filename
2803 raise ValueError(compat_cookiejar
.MISSING_FILENAME_TEXT
)
2805 # Store session cookies with `expires` set to 0 instead of an empty
2808 if cookie
.expires
is None:
2811 with io
.open(filename
, 'w', encoding
='utf-8') as f
:
2812 f
.write(self
._HEADER
)
2815 if not ignore_discard
and cookie
.discard
:
2817 if not ignore_expires
and cookie
.is_expired(now
):
2823 if cookie
.domain
.startswith('.'):
2824 initial_dot
= 'TRUE'
2826 initial_dot
= 'FALSE'
2827 if cookie
.expires
is not None:
2828 expires
= compat_str(cookie
.expires
)
2831 if cookie
.value
is None:
2832 # cookies.txt regards 'Set-Cookie: foo' as a cookie
2833 # with no name, whereas http.cookiejar regards it as a
2834 # cookie with no value.
2839 value
= cookie
.value
2841 '\t'.join([cookie
.domain
, initial_dot
, cookie
.path
,
2842 secure
, expires
, name
, value
]) + '\n')
2844 def load(self
, filename
=None, ignore_discard
=False, ignore_expires
=False):
2845 """Load cookies from a file."""
2846 if filename
is None:
2847 if self
.filename
is not None:
2848 filename
= self
.filename
2850 raise ValueError(compat_cookiejar
.MISSING_FILENAME_TEXT
)
2852 def prepare_line(line
):
2853 if line
.startswith(self
._HTTPONLY
_PREFIX
):
2854 line
= line
[len(self
._HTTPONLY
_PREFIX
):]
2855 # comments and empty lines are fine
2856 if line
.startswith('#') or not line
.strip():
2858 cookie_list
= line
.split('\t')
2859 if len(cookie_list
) != self
._ENTRY
_LEN
:
2860 raise compat_cookiejar
.LoadError('invalid length %d' % len(cookie_list
))
2861 cookie
= self
._CookieFileEntry
(*cookie_list
)
2862 if cookie
.expires_at
and not cookie
.expires_at
.isdigit():
2863 raise compat_cookiejar
.LoadError('invalid expires at %s' % cookie
.expires_at
)
2867 with io
.open(filename
, encoding
='utf-8') as f
:
2870 cf
.write(prepare_line(line
))
2871 except compat_cookiejar
.LoadError
as e
:
2873 'WARNING: skipping cookie file entry due to %s: %r\n'
2874 % (e
, line
), sys
.stderr
)
2877 self
._really
_load
(cf
, filename
, ignore_discard
, ignore_expires
)
2878 # Session cookies are denoted by either `expires` field set to
2879 # an empty string or 0. MozillaCookieJar only recognizes the former
2880 # (see [1]). So we need force the latter to be recognized as session
2881 # cookies on our own.
2882 # Session cookies may be important for cookies-based authentication,
2883 # e.g. usually, when user does not check 'Remember me' check box while
2884 # logging in on a site, some important cookies are stored as session
2885 # cookies so that not recognizing them will result in failed login.
2886 # 1. https://bugs.python.org/issue17164
2888 # Treat `expires=0` cookies as session cookies
2889 if cookie
.expires
== 0:
2890 cookie
.expires
= None
2891 cookie
.discard
= True
2894 class YoutubeDLCookieProcessor(compat_urllib_request
.HTTPCookieProcessor
):
2895 def __init__(self
, cookiejar
=None):
2896 compat_urllib_request
.HTTPCookieProcessor
.__init
__(self
, cookiejar
)
2898 def http_response(self
, request
, response
):
2899 # Python 2 will choke on next HTTP request in row if there are non-ASCII
2900 # characters in Set-Cookie HTTP header of last response (see
2901 # https://github.com/ytdl-org/youtube-dl/issues/6769).
2902 # In order to at least prevent crashing we will percent encode Set-Cookie
2903 # header before HTTPCookieProcessor starts processing it.
2904 # if sys.version_info < (3, 0) and response.headers:
2905 # for set_cookie_header in ('Set-Cookie', 'Set-Cookie2'):
2906 # set_cookie = response.headers.get(set_cookie_header)
2908 # set_cookie_escaped = compat_urllib_parse.quote(set_cookie, b"%/;:@&=+$,!~*'()?#[] ")
2909 # if set_cookie != set_cookie_escaped:
2910 # del response.headers[set_cookie_header]
2911 # response.headers[set_cookie_header] = set_cookie_escaped
2912 return compat_urllib_request
.HTTPCookieProcessor
.http_response(self
, request
, response
)
2914 https_request
= compat_urllib_request
.HTTPCookieProcessor
.http_request
2915 https_response
= http_response
2918 class YoutubeDLRedirectHandler(compat_urllib_request
.HTTPRedirectHandler
):
2919 if sys
.version_info
[0] < 3:
2920 def redirect_request(self
, req
, fp
, code
, msg
, headers
, newurl
):
2921 # On python 2 urlh.geturl() may sometimes return redirect URL
2922 # as byte string instead of unicode. This workaround allows
2923 # to force it always return unicode.
2924 return compat_urllib_request
.HTTPRedirectHandler
.redirect_request(self
, req
, fp
, code
, msg
, headers
, compat_str(newurl
))
2927 def extract_timezone(date_str
):
2929 r
'^.{8,}?(?P<tz>Z$| ?(?P<sign>\+|-)(?P<hours>[0-9]{2}):?(?P<minutes>[0-9]{2})$)',
2932 timezone
= datetime
.timedelta()
2934 date_str
= date_str
[:-len(m
.group('tz'))]
2935 if not m
.group('sign'):
2936 timezone
= datetime
.timedelta()
2938 sign
= 1 if m
.group('sign') == '+' else -1
2939 timezone
= datetime
.timedelta(
2940 hours
=sign
* int(m
.group('hours')),
2941 minutes
=sign
* int(m
.group('minutes')))
2942 return timezone
, date_str
2945 def parse_iso8601(date_str
, delimiter
='T', timezone
=None):
2946 """ Return a UNIX timestamp from the given date """
2948 if date_str
is None:
2951 date_str
= re
.sub(r
'\.[0-9]+', '', date_str
)
2953 if timezone
is None:
2954 timezone
, date_str
= extract_timezone(date_str
)
2957 date_format
= '%Y-%m-%d{0}%H:%M:%S'.format(delimiter
)
2958 dt
= datetime
.datetime
.strptime(date_str
, date_format
) - timezone
2959 return calendar
.timegm(dt
.timetuple())
2964 def date_formats(day_first
=True):
2965 return DATE_FORMATS_DAY_FIRST
if day_first
else DATE_FORMATS_MONTH_FIRST
2968 def unified_strdate(date_str
, day_first
=True):
2969 """Return a string with the date in the format YYYYMMDD"""
2971 if date_str
is None:
2975 date_str
= date_str
.replace(',', ' ')
2976 # Remove AM/PM + timezone
2977 date_str
= re
.sub(r
'(?i)\s*(?:AM|PM)(?:\s+[A-Z]+)?', '', date_str
)
2978 _
, date_str
= extract_timezone(date_str
)
2980 for expression
in date_formats(day_first
):
2982 upload_date
= datetime
.datetime
.strptime(date_str
, expression
).strftime('%Y%m%d')
2985 if upload_date
is None:
2986 timetuple
= email
.utils
.parsedate_tz(date_str
)
2989 upload_date
= datetime
.datetime(*timetuple
[:6]).strftime('%Y%m%d')
2992 if upload_date
is not None:
2993 return compat_str(upload_date
)
2996 def unified_timestamp(date_str
, day_first
=True):
2997 if date_str
is None:
3000 date_str
= re
.sub(r
'[,|]', '', date_str
)
3002 pm_delta
= 12 if re
.search(r
'(?i)PM', date_str
) else 0
3003 timezone
, date_str
= extract_timezone(date_str
)
3005 # Remove AM/PM + timezone
3006 date_str
= re
.sub(r
'(?i)\s*(?:AM|PM)(?:\s+[A-Z]+)?', '', date_str
)
3008 # Remove unrecognized timezones from ISO 8601 alike timestamps
3009 m
= re
.search(r
'\d{1,2}:\d{1,2}(?:\.\d+)?(?P<tz>\s*[A-Z]+)$', date_str
)
3011 date_str
= date_str
[:-len(m
.group('tz'))]
3013 # Python only supports microseconds, so remove nanoseconds
3014 m
= re
.search(r
'^([0-9]{4,}-[0-9]{1,2}-[0-9]{1,2}T[0-9]{1,2}:[0-9]{1,2}:[0-9]{1,2}\.[0-9]{6})[0-9]+$', date_str
)
3016 date_str
= m
.group(1)
3018 for expression
in date_formats(day_first
):
3020 dt
= datetime
.datetime
.strptime(date_str
, expression
) - timezone
+ datetime
.timedelta(hours
=pm_delta
)
3021 return calendar
.timegm(dt
.timetuple())
3024 timetuple
= email
.utils
.parsedate_tz(date_str
)
3026 return calendar
.timegm(timetuple
) + pm_delta
* 3600
3029 def determine_ext(url
, default_ext
='unknown_video'):
3030 if url
is None or '.' not in url
:
3032 guess
= url
.partition('?')[0].rpartition('.')[2]
3033 if re
.match(r
'^[A-Za-z0-9]+$', guess
):
3035 # Try extract ext from URLs like http://example.com/foo/bar.mp4/?download
3036 elif guess
.rstrip('/') in KNOWN_EXTENSIONS
:
3037 return guess
.rstrip('/')
3042 def subtitles_filename(filename
, sub_lang
, sub_format
, expected_real_ext
=None):
3043 return replace_extension(filename
, sub_lang
+ '.' + sub_format
, expected_real_ext
)
3046 def date_from_str(date_str
):
3048 Return a datetime object from a string in the format YYYYMMDD or
3049 (now|today)[+-][0-9](day|week|month|year)(s)?"""
3050 today
= datetime
.date
.today()
3051 if date_str
in ('now', 'today'):
3053 if date_str
== 'yesterday':
3054 return today
- datetime
.timedelta(days
=1)
3055 match
= re
.match(r
'(now|today)(?P<sign>[+-])(?P<time>\d+)(?P<unit>day|week|month|year)(s)?', date_str
)
3056 if match
is not None:
3057 sign
= match
.group('sign')
3058 time
= int(match
.group('time'))
3061 unit
= match
.group('unit')
3062 # A bad approximation?
3066 elif unit
== 'year':
3070 delta
= datetime
.timedelta(**{unit: time}
)
3071 return today
+ delta
3072 return datetime
.datetime
.strptime(date_str
, '%Y%m%d').date()
3075 def hyphenate_date(date_str
):
3077 Convert a date in 'YYYYMMDD' format to 'YYYY-MM-DD' format"""
3078 match
= re
.match(r
'^(\d\d\d\d)(\d\d)(\d\d)$', date_str
)
3079 if match
is not None:
3080 return '-'.join(match
.groups())
3085 class DateRange(object):
3086 """Represents a time interval between two dates"""
3088 def __init__(self
, start
=None, end
=None):
3089 """start and end must be strings in the format accepted by date"""
3090 if start
is not None:
3091 self
.start
= date_from_str(start
)
3093 self
.start
= datetime
.datetime
.min.date()
3095 self
.end
= date_from_str(end
)
3097 self
.end
= datetime
.datetime
.max.date()
3098 if self
.start
> self
.end
:
3099 raise ValueError('Date range: "%s" , the start date must be before the end date' % self
)
3103 """Returns a range that only contains the given day"""
3104 return cls(day
, day
)
3106 def __contains__(self
, date
):
3107 """Check if the date is in the range"""
3108 if not isinstance(date
, datetime
.date
):
3109 date
= date_from_str(date
)
3110 return self
.start
<= date
<= self
.end
3113 return '%s - %s' % (self
.start
.isoformat(), self
.end
.isoformat())
3116 def platform_name():
3117 """ Returns the platform name as a compat_str """
3118 res
= platform
.platform()
3119 if isinstance(res
, bytes):
3120 res
= res
.decode(preferredencoding())
3122 assert isinstance(res
, compat_str
)
3126 def _windows_write_string(s
, out
):
3127 """ Returns True if the string was written using special methods,
3128 False if it has yet to be written out."""
3129 # Adapted from http://stackoverflow.com/a/3259271/35070
3132 import ctypes
.wintypes
3140 fileno
= out
.fileno()
3141 except AttributeError:
3142 # If the output stream doesn't have a fileno, it's virtual
3144 except io
.UnsupportedOperation
:
3145 # Some strange Windows pseudo files?
3147 if fileno
not in WIN_OUTPUT_IDS
:
3150 GetStdHandle
= compat_ctypes_WINFUNCTYPE(
3151 ctypes
.wintypes
.HANDLE
, ctypes
.wintypes
.DWORD
)(
3152 ('GetStdHandle', ctypes
.windll
.kernel32
))
3153 h
= GetStdHandle(WIN_OUTPUT_IDS
[fileno
])
3155 WriteConsoleW
= compat_ctypes_WINFUNCTYPE(
3156 ctypes
.wintypes
.BOOL
, ctypes
.wintypes
.HANDLE
, ctypes
.wintypes
.LPWSTR
,
3157 ctypes
.wintypes
.DWORD
, ctypes
.POINTER(ctypes
.wintypes
.DWORD
),
3158 ctypes
.wintypes
.LPVOID
)(('WriteConsoleW', ctypes
.windll
.kernel32
))
3159 written
= ctypes
.wintypes
.DWORD(0)
3161 GetFileType
= compat_ctypes_WINFUNCTYPE(ctypes
.wintypes
.DWORD
, ctypes
.wintypes
.DWORD
)(('GetFileType', ctypes
.windll
.kernel32
))
3162 FILE_TYPE_CHAR
= 0x0002
3163 FILE_TYPE_REMOTE
= 0x8000
3164 GetConsoleMode
= compat_ctypes_WINFUNCTYPE(
3165 ctypes
.wintypes
.BOOL
, ctypes
.wintypes
.HANDLE
,
3166 ctypes
.POINTER(ctypes
.wintypes
.DWORD
))(
3167 ('GetConsoleMode', ctypes
.windll
.kernel32
))
3168 INVALID_HANDLE_VALUE
= ctypes
.wintypes
.DWORD(-1).value
3170 def not_a_console(handle
):
3171 if handle
== INVALID_HANDLE_VALUE
or handle
is None:
3173 return ((GetFileType(handle
) & ~FILE_TYPE_REMOTE
) != FILE_TYPE_CHAR
3174 or GetConsoleMode(handle
, ctypes
.byref(ctypes
.wintypes
.DWORD())) == 0)
3176 if not_a_console(h
):
3179 def next_nonbmp_pos(s
):
3181 return next(i
for i
, c
in enumerate(s
) if ord(c
) > 0xffff)
3182 except StopIteration:
3186 count
= min(next_nonbmp_pos(s
), 1024)
3188 ret
= WriteConsoleW(
3189 h
, s
, count
if count
else 2, ctypes
.byref(written
), None)
3191 raise OSError('Failed to write string')
3192 if not count
: # We just wrote a non-BMP character
3193 assert written
.value
== 2
3196 assert written
.value
> 0
3197 s
= s
[written
.value
:]
3201 def write_string(s
, out
=None, encoding
=None):
3204 assert type(s
) == compat_str
3206 if sys
.platform
== 'win32' and encoding
is None and hasattr(out
, 'fileno'):
3207 if _windows_write_string(s
, out
):
3210 if ('b' in getattr(out
, 'mode', '')
3211 or sys
.version_info
[0] < 3): # Python 2 lies about mode of sys.stderr
3212 byt
= s
.encode(encoding
or preferredencoding(), 'ignore')
3214 elif hasattr(out
, 'buffer'):
3215 enc
= encoding
or getattr(out
, 'encoding', None) or preferredencoding()
3216 byt
= s
.encode(enc
, 'ignore')
3217 out
.buffer.write(byt
)
3223 def bytes_to_intlist(bs
):
3226 if isinstance(bs
[0], int): # Python 3
3229 return [ord(c
) for c
in bs
]
3232 def intlist_to_bytes(xs
):
3235 return compat_struct_pack('%dB' % len(xs
), *xs
)
3238 # Cross-platform file locking
3239 if sys
.platform
== 'win32':
3240 import ctypes
.wintypes
3243 class OVERLAPPED(ctypes
.Structure
):
3245 ('Internal', ctypes
.wintypes
.LPVOID
),
3246 ('InternalHigh', ctypes
.wintypes
.LPVOID
),
3247 ('Offset', ctypes
.wintypes
.DWORD
),
3248 ('OffsetHigh', ctypes
.wintypes
.DWORD
),
3249 ('hEvent', ctypes
.wintypes
.HANDLE
),
3252 kernel32
= ctypes
.windll
.kernel32
3253 LockFileEx
= kernel32
.LockFileEx
3254 LockFileEx
.argtypes
= [
3255 ctypes
.wintypes
.HANDLE
, # hFile
3256 ctypes
.wintypes
.DWORD
, # dwFlags
3257 ctypes
.wintypes
.DWORD
, # dwReserved
3258 ctypes
.wintypes
.DWORD
, # nNumberOfBytesToLockLow
3259 ctypes
.wintypes
.DWORD
, # nNumberOfBytesToLockHigh
3260 ctypes
.POINTER(OVERLAPPED
) # Overlapped
3262 LockFileEx
.restype
= ctypes
.wintypes
.BOOL
3263 UnlockFileEx
= kernel32
.UnlockFileEx
3264 UnlockFileEx
.argtypes
= [
3265 ctypes
.wintypes
.HANDLE
, # hFile
3266 ctypes
.wintypes
.DWORD
, # dwReserved
3267 ctypes
.wintypes
.DWORD
, # nNumberOfBytesToLockLow
3268 ctypes
.wintypes
.DWORD
, # nNumberOfBytesToLockHigh
3269 ctypes
.POINTER(OVERLAPPED
) # Overlapped
3271 UnlockFileEx
.restype
= ctypes
.wintypes
.BOOL
3272 whole_low
= 0xffffffff
3273 whole_high
= 0x7fffffff
3275 def _lock_file(f
, exclusive
):
3276 overlapped
= OVERLAPPED()
3277 overlapped
.Offset
= 0
3278 overlapped
.OffsetHigh
= 0
3279 overlapped
.hEvent
= 0
3280 f
._lock
_file
_overlapped
_p
= ctypes
.pointer(overlapped
)
3281 handle
= msvcrt
.get_osfhandle(f
.fileno())
3282 if not LockFileEx(handle
, 0x2 if exclusive
else 0x0, 0,
3283 whole_low
, whole_high
, f
._lock
_file
_overlapped
_p
):
3284 raise OSError('Locking file failed: %r' % ctypes
.FormatError())
3286 def _unlock_file(f
):
3287 assert f
._lock
_file
_overlapped
_p
3288 handle
= msvcrt
.get_osfhandle(f
.fileno())
3289 if not UnlockFileEx(handle
, 0,
3290 whole_low
, whole_high
, f
._lock
_file
_overlapped
_p
):
3291 raise OSError('Unlocking file failed: %r' % ctypes
.FormatError())
3294 # Some platforms, such as Jython, is missing fcntl
3298 def _lock_file(f
, exclusive
):
3299 fcntl
.flock(f
, fcntl
.LOCK_EX
if exclusive
else fcntl
.LOCK_SH
)
3301 def _unlock_file(f
):
3302 fcntl
.flock(f
, fcntl
.LOCK_UN
)
3304 UNSUPPORTED_MSG
= 'file locking is not supported on this platform'
3306 def _lock_file(f
, exclusive
):
3307 raise IOError(UNSUPPORTED_MSG
)
3309 def _unlock_file(f
):
3310 raise IOError(UNSUPPORTED_MSG
)
3313 class locked_file(object):
3314 def __init__(self
, filename
, mode
, encoding
=None):
3315 assert mode
in ['r', 'a', 'w']
3316 self
.f
= io
.open(filename
, mode
, encoding
=encoding
)
3319 def __enter__(self
):
3320 exclusive
= self
.mode
!= 'r'
3322 _lock_file(self
.f
, exclusive
)
3328 def __exit__(self
, etype
, value
, traceback
):
3330 _unlock_file(self
.f
)
3337 def write(self
, *args
):
3338 return self
.f
.write(*args
)
3340 def read(self
, *args
):
3341 return self
.f
.read(*args
)
3344 def get_filesystem_encoding():
3345 encoding
= sys
.getfilesystemencoding()
3346 return encoding
if encoding
is not None else 'utf-8'
3349 def shell_quote(args
):
3351 encoding
= get_filesystem_encoding()
3353 if isinstance(a
, bytes):
3354 # We may get a filename encoded with 'encodeFilename'
3355 a
= a
.decode(encoding
)
3356 quoted_args
.append(compat_shlex_quote(a
))
3357 return ' '.join(quoted_args
)
3360 def smuggle_url(url
, data
):
3361 """ Pass additional data in a URL for internal use. """
3363 url
, idata
= unsmuggle_url(url
, {})
3365 sdata
= compat_urllib_parse_urlencode(
3366 {'__youtubedl_smuggle': json.dumps(data)}
)
3367 return url
+ '#' + sdata
3370 def unsmuggle_url(smug_url
, default
=None):
3371 if '#__youtubedl_smuggle' not in smug_url
:
3372 return smug_url
, default
3373 url
, _
, sdata
= smug_url
.rpartition('#')
3374 jsond
= compat_parse_qs(sdata
)['__youtubedl_smuggle'][0]
3375 data
= json
.loads(jsond
)
3379 def format_bytes(bytes):
3382 if type(bytes) is str:
3383 bytes = float(bytes)
3387 exponent
= int(math
.log(bytes, 1024.0))
3388 suffix
= ['B', 'KiB', 'MiB', 'GiB', 'TiB', 'PiB', 'EiB', 'ZiB', 'YiB'][exponent
]
3389 converted
= float(bytes) / float(1024 ** exponent
)
3390 return '%.2f%s' % (converted
, suffix
)
3393 def lookup_unit_table(unit_table
, s
):
3394 units_re
= '|'.join(re
.escape(u
) for u
in unit_table
)
3396 r
'(?P<num>[0-9]+(?:[,.][0-9]*)?)\s*(?P<unit>%s)\b' % units_re
, s
)
3399 num_str
= m
.group('num').replace(',', '.')
3400 mult
= unit_table
[m
.group('unit')]
3401 return int(float(num_str
) * mult
)
3404 def parse_filesize(s
):
3408 # The lower-case forms are of course incorrect and unofficial,
3409 # but we support those too
3426 'megabytes': 1000 ** 2,
3427 'mebibytes': 1024 ** 2,
3433 'gigabytes': 1000 ** 3,
3434 'gibibytes': 1024 ** 3,
3440 'terabytes': 1000 ** 4,
3441 'tebibytes': 1024 ** 4,
3447 'petabytes': 1000 ** 5,
3448 'pebibytes': 1024 ** 5,
3454 'exabytes': 1000 ** 6,
3455 'exbibytes': 1024 ** 6,
3461 'zettabytes': 1000 ** 7,
3462 'zebibytes': 1024 ** 7,
3468 'yottabytes': 1000 ** 8,
3469 'yobibytes': 1024 ** 8,
3472 return lookup_unit_table(_UNIT_TABLE
, s
)
3481 if re
.match(r
'^[\d,.]+$', s
):
3482 return str_to_int(s
)
3493 return lookup_unit_table(_UNIT_TABLE
, s
)
3496 def parse_resolution(s
):
3500 mobj
= re
.search(r
'\b(?P<w>\d+)\s*[xX×]\s*(?P<h>\d+)\b', s
)
3503 'width': int(mobj
.group('w')),
3504 'height': int(mobj
.group('h')),
3507 mobj
= re
.search(r
'\b(\d+)[pPiI]\b', s
)
3509 return {'height': int(mobj.group(1))}
3511 mobj
= re
.search(r
'\b([48])[kK]\b', s
)
3513 return {'height': int(mobj.group(1)) * 540}
3518 def parse_bitrate(s
):
3519 if not isinstance(s
, compat_str
):
3521 mobj
= re
.search(r
'\b(\d+)\s*kbps', s
)
3523 return int(mobj
.group(1))
3526 def month_by_name(name
, lang
='en'):
3527 """ Return the number of a month by (locale-independently) English name """
3529 month_names
= MONTH_NAMES
.get(lang
, MONTH_NAMES
['en'])
3532 return month_names
.index(name
) + 1
3537 def month_by_abbreviation(abbrev
):
3538 """ Return the number of a month by (locale-independently) English
3542 return [s
[:3] for s
in ENGLISH_MONTH_NAMES
].index(abbrev
) + 1
3547 def fix_xml_ampersands(xml_str
):
3548 """Replace all the '&' by '&' in XML"""
3550 r
'&(?!amp;|lt;|gt;|apos;|quot;|#x[0-9a-fA-F]{,4};|#[0-9]{,4};)',
3555 def setproctitle(title
):
3556 assert isinstance(title
, compat_str
)
3558 # ctypes in Jython is not complete
3559 # http://bugs.jython.org/issue2148
3560 if sys
.platform
.startswith('java'):
3564 libc
= ctypes
.cdll
.LoadLibrary('libc.so.6')
3568 # LoadLibrary in Windows Python 2.7.13 only expects
3569 # a bytestring, but since unicode_literals turns
3570 # every string into a unicode string, it fails.
3572 title_bytes
= title
.encode('utf-8')
3573 buf
= ctypes
.create_string_buffer(len(title_bytes
))
3574 buf
.value
= title_bytes
3576 libc
.prctl(15, buf
, 0, 0, 0)
3577 except AttributeError:
3578 return # Strange libc, just skip this
3581 def remove_start(s
, start
):
3582 return s
[len(start
):] if s
is not None and s
.startswith(start
) else s
3585 def remove_end(s
, end
):
3586 return s
[:-len(end
)] if s
is not None and s
.endswith(end
) else s
3589 def remove_quotes(s
):
3590 if s
is None or len(s
) < 2:
3592 for quote
in ('"', "'", ):
3593 if s
[0] == quote
and s
[-1] == quote
:
3598 def get_domain(url
):
3599 domain
= re
.match(r
'(?:https?:\/\/)?(?:www\.)?(?P<domain>[^\n\/]+\.[^\n\/]+)(?:\/(.*))?', url
)
3600 return domain
.group('domain') if domain
else None
3603 def url_basename(url
):
3604 path
= compat_urlparse
.urlparse(url
).path
3605 return path
.strip('/').split('/')[-1]
3609 return re
.match(r
'https?://[^?#&]+/', url
).group()
3612 def urljoin(base
, path
):
3613 if isinstance(path
, bytes):
3614 path
= path
.decode('utf-8')
3615 if not isinstance(path
, compat_str
) or not path
:
3617 if re
.match(r
'^(?:[a-zA-Z][a-zA-Z0-9+-.]*:)?//', path
):
3619 if isinstance(base
, bytes):
3620 base
= base
.decode('utf-8')
3621 if not isinstance(base
, compat_str
) or not re
.match(
3622 r
'^(?:https?:)?//', base
):
3624 return compat_urlparse
.urljoin(base
, path
)
3627 class HEADRequest(compat_urllib_request
.Request
):
3628 def get_method(self
):
3632 class PUTRequest(compat_urllib_request
.Request
):
3633 def get_method(self
):
3637 def int_or_none(v
, scale
=1, default
=None, get_attr
=None, invscale
=1):
3640 v
= getattr(v
, get_attr
, None)
3646 return int(v
) * invscale
// scale
3647 except (ValueError, TypeError):
3651 def str_or_none(v
, default
=None):
3652 return default
if v
is None else compat_str(v
)
3655 def str_to_int(int_str
):
3656 """ A more relaxed version of int_or_none """
3657 if isinstance(int_str
, compat_integer_types
):
3659 elif isinstance(int_str
, compat_str
):
3660 int_str
= re
.sub(r
'[,\.\+]', '', int_str
)
3661 return int_or_none(int_str
)
3664 def float_or_none(v
, scale
=1, invscale
=1, default
=None):
3668 return float(v
) * invscale
/ scale
3669 except (ValueError, TypeError):
3673 def bool_or_none(v
, default
=None):
3674 return v
if isinstance(v
, bool) else default
3677 def strip_or_none(v
, default
=None):
3678 return v
.strip() if isinstance(v
, compat_str
) else default
3681 def url_or_none(url
):
3682 if not url
or not isinstance(url
, compat_str
):
3685 return url
if re
.match(r
'^(?:(?:https?|rt(?:m(?:pt?[es]?|fp)|sp[su]?)|mms|ftps?):)?//', url
) else None
3688 def strftime_or_none(timestamp
, date_format
, default
=None):
3689 datetime_object
= None
3691 if isinstance(timestamp
, compat_numeric_types
): # unix timestamp
3692 datetime_object
= datetime
.datetime
.utcfromtimestamp(timestamp
)
3693 elif isinstance(timestamp
, compat_str
): # assume YYYYMMDD
3694 datetime_object
= datetime
.datetime
.strptime(timestamp
, '%Y%m%d')
3695 return datetime_object
.strftime(date_format
)
3696 except (ValueError, TypeError, AttributeError):
3700 def parse_duration(s
):
3701 if not isinstance(s
, compat_basestring
):
3706 days
, hours
, mins
, secs
, ms
= [None] * 5
3707 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
)
3709 days
, hours
, mins
, secs
, ms
= m
.groups()
3714 [0-9]+\s*y(?:ears?)?\s*
3717 [0-9]+\s*m(?:onths?)?\s*
3720 [0-9]+\s*w(?:eeks?)?\s*
3723 (?P<days>[0-9]+)\s*d(?:ays?)?\s*
3727 (?P<hours>[0-9]+)\s*h(?:ours?)?\s*
3730 (?P<mins>[0-9]+)\s*m(?:in(?:ute)?s?)?\s*
3733 (?P<secs>[0-9]+)(?P<ms>\.[0-9]+)?\s*s(?:ec(?:ond)?s?)?\s*
3736 days
, hours
, mins
, secs
, ms
= m
.groups()
3738 m
= re
.match(r
'(?i)(?:(?P<hours>[0-9.]+)\s*(?:hours?)|(?P<mins>[0-9.]+)\s*(?:mins?\.?|minutes?)\s*)Z?$', s
)
3740 hours
, mins
= m
.groups()
3746 duration
+= float(secs
)
3748 duration
+= float(mins
) * 60
3750 duration
+= float(hours
) * 60 * 60
3752 duration
+= float(days
) * 24 * 60 * 60
3754 duration
+= float(ms
)
3758 def prepend_extension(filename
, ext
, expected_real_ext
=None):
3759 name
, real_ext
= os
.path
.splitext(filename
)
3761 '{0}.{1}{2}'.format(name
, ext
, real_ext
)
3762 if not expected_real_ext
or real_ext
[1:] == expected_real_ext
3763 else '{0}.{1}'.format(filename
, ext
))
3766 def replace_extension(filename
, ext
, expected_real_ext
=None):
3767 name
, real_ext
= os
.path
.splitext(filename
)
3768 return '{0}.{1}'.format(
3769 name
if not expected_real_ext
or real_ext
[1:] == expected_real_ext
else filename
,
3773 def check_executable(exe
, args
=[]):
3774 """ Checks if the given binary is installed somewhere in PATH, and returns its name.
3775 args can be a list of arguments for a short output (like -version) """
3777 process_communicate_or_kill(subprocess
.Popen(
3778 [exe
] + args
, stdout
=subprocess
.PIPE
, stderr
=subprocess
.PIPE
))
3784 def get_exe_version(exe
, args
=['--version'],
3785 version_re
=None, unrecognized
='present'):
3786 """ Returns the version of the specified executable,
3787 or False if the executable is not present """
3789 # STDIN should be redirected too. On UNIX-like systems, ffmpeg triggers
3790 # SIGTTOU if yt-dlp is run in the background.
3791 # See https://github.com/ytdl-org/youtube-dl/issues/955#issuecomment-209789656
3792 out
, _
= process_communicate_or_kill(subprocess
.Popen(
3793 [encodeArgument(exe
)] + args
,
3794 stdin
=subprocess
.PIPE
,
3795 stdout
=subprocess
.PIPE
, stderr
=subprocess
.STDOUT
))
3798 if isinstance(out
, bytes): # Python 2.x
3799 out
= out
.decode('ascii', 'ignore')
3800 return detect_exe_version(out
, version_re
, unrecognized
)
3803 def detect_exe_version(output
, version_re
=None, unrecognized
='present'):
3804 assert isinstance(output
, compat_str
)
3805 if version_re
is None:
3806 version_re
= r
'version\s+([-0-9._a-zA-Z]+)'
3807 m
= re
.search(version_re
, output
)
3814 class PagedList(object):
3816 # This is only useful for tests
3817 return len(self
.getslice())
3820 class OnDemandPagedList(PagedList
):
3821 def __init__(self
, pagefunc
, pagesize
, use_cache
=True):
3822 self
._pagefunc
= pagefunc
3823 self
._pagesize
= pagesize
3824 self
._use
_cache
= use_cache
3828 def getslice(self
, start
=0, end
=None):
3830 for pagenum
in itertools
.count(start
// self
._pagesize
):
3831 firstid
= pagenum
* self
._pagesize
3832 nextfirstid
= pagenum
* self
._pagesize
+ self
._pagesize
3833 if start
>= nextfirstid
:
3838 page_results
= self
._cache
.get(pagenum
)
3839 if page_results
is None:
3840 page_results
= list(self
._pagefunc
(pagenum
))
3842 self
._cache
[pagenum
] = page_results
3845 start
% self
._pagesize
3846 if firstid
<= start
< nextfirstid
3850 ((end
- 1) % self
._pagesize
) + 1
3851 if (end
is not None and firstid
<= end
<= nextfirstid
)
3854 if startv
!= 0 or endv
is not None:
3855 page_results
= page_results
[startv
:endv
]
3856 res
.extend(page_results
)
3858 # A little optimization - if current page is not "full", ie. does
3859 # not contain page_size videos then we can assume that this page
3860 # is the last one - there are no more ids on further pages -
3861 # i.e. no need to query again.
3862 if len(page_results
) + startv
< self
._pagesize
:
3865 # If we got the whole page, but the next page is not interesting,
3866 # break out early as well
3867 if end
== nextfirstid
:
3872 class InAdvancePagedList(PagedList
):
3873 def __init__(self
, pagefunc
, pagecount
, pagesize
):
3874 self
._pagefunc
= pagefunc
3875 self
._pagecount
= pagecount
3876 self
._pagesize
= pagesize
3878 def getslice(self
, start
=0, end
=None):
3880 start_page
= start
// self
._pagesize
3882 self
._pagecount
if end
is None else (end
// self
._pagesize
+ 1))
3883 skip_elems
= start
- start_page
* self
._pagesize
3884 only_more
= None if end
is None else end
- start
3885 for pagenum
in range(start_page
, end_page
):
3886 page
= list(self
._pagefunc
(pagenum
))
3888 page
= page
[skip_elems
:]
3890 if only_more
is not None:
3891 if len(page
) < only_more
:
3892 only_more
-= len(page
)
3894 page
= page
[:only_more
]
3901 def uppercase_escape(s
):
3902 unicode_escape
= codecs
.getdecoder('unicode_escape')
3904 r
'\\U[0-9a-fA-F]{8}',
3905 lambda m
: unicode_escape(m
.group(0))[0],
3909 def lowercase_escape(s
):
3910 unicode_escape
= codecs
.getdecoder('unicode_escape')
3912 r
'\\u[0-9a-fA-F]{4}',
3913 lambda m
: unicode_escape(m
.group(0))[0],
3917 def escape_rfc3986(s
):
3918 """Escape non-ASCII characters as suggested by RFC 3986"""
3919 if sys
.version_info
< (3, 0) and isinstance(s
, compat_str
):
3920 s
= s
.encode('utf-8')
3921 return compat_urllib_parse
.quote(s
, b
"%/;:@&=+$,!~*'()?#[]")
3924 def escape_url(url
):
3925 """Escape URL as suggested by RFC 3986"""
3926 url_parsed
= compat_urllib_parse_urlparse(url
)
3927 return url_parsed
._replace
(
3928 netloc
=url_parsed
.netloc
.encode('idna').decode('ascii'),
3929 path
=escape_rfc3986(url_parsed
.path
),
3930 params
=escape_rfc3986(url_parsed
.params
),
3931 query
=escape_rfc3986(url_parsed
.query
),
3932 fragment
=escape_rfc3986(url_parsed
.fragment
)
3936 def read_batch_urls(batch_fd
):
3938 if not isinstance(url
, compat_str
):
3939 url
= url
.decode('utf-8', 'replace')
3940 BOM_UTF8
= ('\xef\xbb\xbf', '\ufeff')
3941 for bom
in BOM_UTF8
:
3942 if url
.startswith(bom
):
3943 url
= url
[len(bom
):]
3945 if not url
or url
.startswith(('#', ';', ']')):
3947 # "#" cannot be stripped out since it is part of the URI
3948 # However, it can be safely stipped out if follwing a whitespace
3949 return re
.split(r
'\s#', url
, 1)[0].rstrip()
3951 with contextlib
.closing(batch_fd
) as fd
:
3952 return [url
for url
in map(fixup
, fd
) if url
]
3955 def urlencode_postdata(*args
, **kargs
):
3956 return compat_urllib_parse_urlencode(*args
, **kargs
).encode('ascii')
3959 def update_url_query(url
, query
):
3962 parsed_url
= compat_urlparse
.urlparse(url
)
3963 qs
= compat_parse_qs(parsed_url
.query
)
3965 return compat_urlparse
.urlunparse(parsed_url
._replace
(
3966 query
=compat_urllib_parse_urlencode(qs
, True)))
3969 def update_Request(req
, url
=None, data
=None, headers
={}, query={}
):
3970 req_headers
= req
.headers
.copy()
3971 req_headers
.update(headers
)
3972 req_data
= data
or req
.data
3973 req_url
= update_url_query(url
or req
.get_full_url(), query
)
3974 req_get_method
= req
.get_method()
3975 if req_get_method
== 'HEAD':
3976 req_type
= HEADRequest
3977 elif req_get_method
== 'PUT':
3978 req_type
= PUTRequest
3980 req_type
= compat_urllib_request
.Request
3982 req_url
, data
=req_data
, headers
=req_headers
,
3983 origin_req_host
=req
.origin_req_host
, unverifiable
=req
.unverifiable
)
3984 if hasattr(req
, 'timeout'):
3985 new_req
.timeout
= req
.timeout
3989 def _multipart_encode_impl(data
, boundary
):
3990 content_type
= 'multipart/form-data; boundary=%s' % boundary
3993 for k
, v
in data
.items():
3994 out
+= b
'--' + boundary
.encode('ascii') + b
'\r\n'
3995 if isinstance(k
, compat_str
):
3996 k
= k
.encode('utf-8')
3997 if isinstance(v
, compat_str
):
3998 v
= v
.encode('utf-8')
3999 # RFC 2047 requires non-ASCII field names to be encoded, while RFC 7578
4000 # suggests sending UTF-8 directly. Firefox sends UTF-8, too
4001 content
= b
'Content-Disposition: form-data; name="' + k
+ b
'"\r\n\r\n' + v
+ b
'\r\n'
4002 if boundary
.encode('ascii') in content
:
4003 raise ValueError('Boundary overlaps with data')
4006 out
+= b
'--' + boundary
.encode('ascii') + b
'--\r\n'
4008 return out
, content_type
4011 def multipart_encode(data
, boundary
=None):
4013 Encode a dict to RFC 7578-compliant form-data
4016 A dict where keys and values can be either Unicode or bytes-like
4019 If specified a Unicode object, it's used as the boundary. Otherwise
4020 a random boundary is generated.
4022 Reference: https://tools.ietf.org/html/rfc7578
4024 has_specified_boundary
= boundary
is not None
4027 if boundary
is None:
4028 boundary
= '---------------' + str(random
.randrange(0x0fffffff, 0xffffffff))
4031 out
, content_type
= _multipart_encode_impl(data
, boundary
)
4034 if has_specified_boundary
:
4038 return out
, content_type
4041 def dict_get(d
, key_or_keys
, default
=None, skip_false_values
=True):
4042 if isinstance(key_or_keys
, (list, tuple)):
4043 for key
in key_or_keys
:
4044 if key
not in d
or d
[key
] is None or skip_false_values
and not d
[key
]:
4048 return d
.get(key_or_keys
, default
)
4051 def try_get(src
, getter
, expected_type
=None):
4052 if not isinstance(getter
, (list, tuple)):
4057 except (AttributeError, KeyError, TypeError, IndexError):
4060 if expected_type
is None or isinstance(v
, expected_type
):
4064 def merge_dicts(*dicts
):
4066 for a_dict
in dicts
:
4067 for k
, v
in a_dict
.items():
4071 or (isinstance(v
, compat_str
) and v
4072 and isinstance(merged
[k
], compat_str
)
4073 and not merged
[k
])):
4078 def encode_compat_str(string
, encoding
=preferredencoding(), errors
='strict'):
4079 return string
if isinstance(string
, compat_str
) else compat_str(string
, encoding
, errors
)
4091 TV_PARENTAL_GUIDELINES
= {
4101 def parse_age_limit(s
):
4103 return s
if 0 <= s
<= 21 else None
4104 if not isinstance(s
, compat_basestring
):
4106 m
= re
.match(r
'^(?P<age>\d{1,2})\+?$', s
)
4108 return int(m
.group('age'))
4110 return US_RATINGS
[s
]
4111 m
= re
.match(r
'^TV[_-]?(%s)$' % '|'.join(k
[3:] for k
in TV_PARENTAL_GUIDELINES
), s
)
4113 return TV_PARENTAL_GUIDELINES
['TV-' + m
.group(1)]
4117 def strip_jsonp(code
):
4120 (?:window\.)?(?P<func_name>[a-zA-Z0-9_.$]*)
4121 (?:\s*&&\s*(?P=func_name))?
4122 \s*\(\s*(?P<callback_data>.*)\);?
4123 \s*?(?://[^\n]*)*$''',
4124 r
'\g<callback_data>', code
)
4127 def js_to_json(code
, vars={}):
4128 # vars is a dict of var, val pairs to substitute
4129 COMMENT_RE
= r
'/\*(?:(?!\*/).)*?\*/|//[^\n]*'
4130 SKIP_RE
= r
'\s*(?:{comment})?\s*'.format(comment
=COMMENT_RE
)
4132 (r
'(?s)^(0[xX][0-9a-fA-F]+){skip}:?$'.format(skip
=SKIP_RE
), 16),
4133 (r
'(?s)^(0+[0-7]+){skip}:?$'.format(skip
=SKIP_RE
), 8),
4138 if v
in ('true', 'false', 'null'):
4140 elif v
.startswith('/*') or v
.startswith('//') or v
.startswith('!') or v
== ',':
4143 if v
[0] in ("'", '"'):
4144 v
= re
.sub(r
'(?s)\\.|"', lambda m
: {
4149 }.get(m
.group(0), m
.group(0)), v
[1:-1])
4151 for regex
, base
in INTEGER_TABLE
:
4152 im
= re
.match(regex
, v
)
4154 i
= int(im
.group(1), base
)
4155 return '"%d":' % i
if v
.endswith(':') else '%d' % i
4162 return re
.sub(r
'''(?sx)
4163 "(?:[^"\\]*(?:\\\\|\\['"nurtbfx/\n]))*[^"\\]*"|
4164 '(?:[^'\\]*(?:\\\\|\\['"nurtbfx/\n]))*[^'\\]*'|
4165 {comment}|,(?={skip}[\]}}])|
4166 (?:(?<![0-9])[eE]|[a-df-zA-DF-Z_])[.a-zA-Z_0-9]*|
4167 \b(?:0[xX][0-9a-fA-F]+|0+[0-7]+)(?:{skip}:)?|
4170 '''.format(comment
=COMMENT_RE
, skip
=SKIP_RE
), fix_kv
, code
)
4173 def qualities(quality_ids
):
4174 """ Get a numeric quality value out of a list of possible values """
4177 return quality_ids
.index(qid
)
4184 'default': '%(title)s [%(id)s].%(ext)s',
4185 'chapter': '%(title)s - %(section_number)03d %(section_title)s [%(id)s].%(ext)s',
4191 'description': 'description',
4192 'annotation': 'annotations.xml',
4193 'infojson': 'info.json',
4194 'pl_description': 'description',
4195 'pl_infojson': 'info.json',
4199 def limit_length(s
, length
):
4200 """ Add ellipses to overly long strings """
4205 return s
[:length
- len(ELLIPSES
)] + ELLIPSES
4209 def version_tuple(v
):
4210 return tuple(int(e
) for e
in re
.split(r
'[-.]', v
))
4213 def is_outdated_version(version
, limit
, assume_new
=True):
4215 return not assume_new
4217 return version_tuple(version
) < version_tuple(limit
)
4219 return not assume_new
4222 def ytdl_is_updateable():
4223 """ Returns if yt-dlp can be updated with -U """
4226 from zipimport
import zipimporter
4228 return isinstance(globals().get('__loader__'), zipimporter
) or hasattr(sys
, 'frozen')
4231 def args_to_str(args
):
4232 # Get a short string representation for a subprocess command
4233 return ' '.join(compat_shlex_quote(a
) for a
in args
)
4236 def error_to_compat_str(err
):
4238 # On python 2 error byte string must be decoded with proper
4239 # encoding rather than ascii
4240 if sys
.version_info
[0] < 3:
4241 err_str
= err_str
.decode(preferredencoding())
4245 def mimetype2ext(mt
):
4251 # Per RFC 3003, audio/mpeg can be .mp1, .mp2 or .mp3. Here use .mp3 as
4252 # it's the most popular one
4253 'audio/mpeg': 'mp3',
4254 'audio/x-wav': 'wav',
4259 _
, _
, res
= mt
.rpartition('/')
4260 res
= res
.split(';')[0].strip().lower()
4264 'smptett+xml': 'tt',
4268 'x-mp4-fragmented': 'mp4',
4269 'x-ms-sami': 'sami',
4272 'x-mpegurl': 'm3u8',
4273 'vnd.apple.mpegurl': 'm3u8',
4277 'vnd.ms-sstr+xml': 'ism',
4284 def parse_codecs(codecs_str
):
4285 # http://tools.ietf.org/html/rfc6381
4288 split_codecs
= list(filter(None, map(
4289 lambda str: str.strip(), codecs_str
.strip().strip(',').split(','))))
4290 vcodec
, acodec
= None, None
4291 for full_codec
in split_codecs
:
4292 codec
= full_codec
.split('.')[0]
4293 if codec
in ('avc1', 'avc2', 'avc3', 'avc4', 'vp9', 'vp8', 'hev1', 'hev2', 'h263', 'h264', 'mp4v', 'hvc1', 'av01', 'theora'):
4296 elif codec
in ('mp4a', 'opus', 'vorbis', 'mp3', 'aac', 'ac-3', 'ec-3', 'eac3', 'dtsc', 'dtse', 'dtsh', 'dtsl'):
4300 write_string('WARNING: Unknown codec %s\n' % full_codec
, sys
.stderr
)
4301 if not vcodec
and not acodec
:
4302 if len(split_codecs
) == 2:
4304 'vcodec': split_codecs
[0],
4305 'acodec': split_codecs
[1],
4309 'vcodec': vcodec
or 'none',
4310 'acodec': acodec
or 'none',
4315 def urlhandle_detect_ext(url_handle
):
4316 getheader
= url_handle
.headers
.get
4318 cd
= getheader('Content-Disposition')
4320 m
= re
.match(r
'attachment;\s*filename="(?P<filename>[^"]+)"', cd
)
4322 e
= determine_ext(m
.group('filename'), default_ext
=None)
4326 return mimetype2ext(getheader('Content-Type'))
4329 def encode_data_uri(data
, mime_type
):
4330 return 'data:%s;base64,%s' % (mime_type
, base64
.b64encode(data
).decode('ascii'))
4333 def age_restricted(content_limit
, age_limit
):
4334 """ Returns True iff the content should be blocked """
4336 if age_limit
is None: # No limit set
4338 if content_limit
is None:
4339 return False # Content available for everyone
4340 return age_limit
< content_limit
4343 def is_html(first_bytes
):
4344 """ Detect whether a file contains HTML by examining its first bytes. """
4347 (b
'\xef\xbb\xbf', 'utf-8'),
4348 (b
'\x00\x00\xfe\xff', 'utf-32-be'),
4349 (b
'\xff\xfe\x00\x00', 'utf-32-le'),
4350 (b
'\xff\xfe', 'utf-16-le'),
4351 (b
'\xfe\xff', 'utf-16-be'),
4353 for bom
, enc
in BOMS
:
4354 if first_bytes
.startswith(bom
):
4355 s
= first_bytes
[len(bom
):].decode(enc
, 'replace')
4358 s
= first_bytes
.decode('utf-8', 'replace')
4360 return re
.match(r
'^\s*<', s
)
4363 def determine_protocol(info_dict
):
4364 protocol
= info_dict
.get('protocol')
4365 if protocol
is not None:
4368 url
= info_dict
['url']
4369 if url
.startswith('rtmp'):
4371 elif url
.startswith('mms'):
4373 elif url
.startswith('rtsp'):
4376 ext
= determine_ext(url
)
4382 return compat_urllib_parse_urlparse(url
).scheme
4385 def render_table(header_row
, data
, delim
=False, extraGap
=0, hideEmpty
=False):
4386 """ Render a list of rows, each as a list of values """
4388 def get_max_lens(table
):
4389 return [max(len(compat_str(v
)) for v
in col
) for col
in zip(*table
)]
4391 def filter_using_list(row
, filterArray
):
4392 return [col
for (take
, col
) in zip(filterArray
, row
) if take
]
4395 max_lens
= get_max_lens(data
)
4396 header_row
= filter_using_list(header_row
, max_lens
)
4397 data
= [filter_using_list(row
, max_lens
) for row
in data
]
4399 table
= [header_row
] + data
4400 max_lens
= get_max_lens(table
)
4402 table
= [header_row
] + [['-' * ml
for ml
in max_lens
]] + data
4403 format_str
= ' '.join('%-' + compat_str(ml
+ extraGap
) + 's' for ml
in max_lens
[:-1]) + ' %s'
4404 return '\n'.join(format_str
% tuple(row
) for row
in table
)
4407 def _match_one(filter_part
, dct
):
4408 COMPARISON_OPERATORS
= {
4416 operator_rex
= re
.compile(r
'''(?x)\s*
4418 \s*(?P<op>%s)(?P<none_inclusive>\s*\?)?\s*
4420 (?P<intval>[0-9.]+(?:[kKmMgGtTpPeEzZyY]i?[Bb]?)?)|
4421 (?P<quote>["\'])(?P
<quotedstrval
>(?
:\\.|
(?
!(?P
=quote
)|
\\).)+?
)(?P
=quote
)|
4422 (?P
<strval
>(?
![0-9.])[a
-z0
-9A
-Z
]*)
4425 ''' % '|'.join(map(re.escape, COMPARISON_OPERATORS.keys())))
4426 m = operator_rex.search(filter_part)
4428 op = COMPARISON_OPERATORS[m.group('op')]
4429 actual_value = dct.get(m.group('key'))
4430 if (m.group('quotedstrval') is not None
4431 or m.group('strval') is not None
4432 # If the original field is a string and matching comparisonvalue is
4433 # a number we should respect the origin of the original field
4434 # and process comparison value as a string (see
4435 # https://github.com/ytdl-org/youtube-dl/issues/11082).
4436 or actual_value is not None and m.group('intval') is not None
4437 and isinstance(actual_value, compat_str)):
4438 if m.group('op') not in ('=', '!='):
4440 'Operator %s does not support string values!' % m.group('op'))
4441 comparison_value = m.group('quotedstrval') or m.group('strval') or m.group('intval')
4442 quote = m.group('quote')
4443 if quote is not None:
4444 comparison_value = comparison_value.replace(r'\%s' % quote, quote)
4447 comparison_value = int(m.group('intval'))
4449 comparison_value = parse_filesize(m.group('intval'))
4450 if comparison_value is None:
4451 comparison_value = parse_filesize(m.group('intval') + 'B')
4452 if comparison_value is None:
4454 'Invalid integer value %r in filter part %r' % (
4455 m.group('intval'), filter_part))
4456 if actual_value is None:
4457 return m.group('none_inclusive')
4458 return op(actual_value, comparison_value)
4461 '': lambda v: (v is True) if isinstance(v, bool) else (v is not None),
4462 '!': lambda v: (v is False) if isinstance(v, bool) else (v is None),
4464 operator_rex = re.compile(r'''(?x
)\s
*
4465 (?P
<op
>%s)\s
*(?P
<key
>[a
-z_
]+)
4467 ''' % '|'.join(map(re.escape, UNARY_OPERATORS.keys())))
4468 m = operator_rex.search(filter_part)
4470 op = UNARY_OPERATORS[m.group('op')]
4471 actual_value = dct.get(m.group('key'))
4472 return op(actual_value)
4474 raise ValueError('Invalid filter part %r' % filter_part)
4477 def match_str(filter_str, dct):
4478 """ Filter a dictionary with a simple string syntax. Returns True (=passes filter) or false """
4481 _match_one(filter_part, dct) for filter_part in filter_str.split('&'))
4484 def match_filter_func(filter_str):
4485 def _match_func(info_dict):
4486 if match_str(filter_str, info_dict):
4489 video_title = info_dict.get('title', info_dict.get('id', 'video'))
4490 return '%s does not pass filter %s, skipping ..' % (video_title, filter_str)
4494 def parse_dfxp_time_expr(time_expr):
4498 mobj = re.match(r'^(?P<time_offset>\d+(?:\.\d+)?)s?$', time_expr)
4500 return float(mobj.group('time_offset'))
4502 mobj = re.match(r'^(\d+):(\d\d):(\d\d(?:(?:\.|:)\d+)?)$', time_expr)
4504 return 3600 * int(mobj.group(1)) + 60 * int(mobj.group(2)) + float(mobj.group(3).replace(':', '.'))
4507 def srt_subtitles_timecode(seconds):
4508 return '%02d:%02d:%02d,%03d' % (seconds / 3600, (seconds % 3600) / 60, seconds % 60, (seconds % 1) * 1000)
4511 def dfxp2srt(dfxp_data):
4513 @param dfxp_data A
bytes-like
object containing DFXP data
4514 @returns A
unicode object containing converted SRT data
4516 LEGACY_NAMESPACES = (
4517 (b'http://www.w3.org/ns/ttml', [
4518 b'http://www.w3.org/2004/11/ttaf1',
4519 b'http://www.w3.org/2006/04/ttaf1',
4520 b'http://www.w3.org/2006/10/ttaf1',
4522 (b'http://www.w3.org/ns/ttml#styling', [
4523 b'http://www.w3.org/ns/ttml#style',
4527 SUPPORTED_STYLING = [
4536 _x = functools.partial(xpath_with_ns, ns_map={
4537 'xml': 'http://www.w3.org/XML/1998/namespace',
4538 'ttml': 'http://www.w3.org/ns/ttml',
4539 'tts': 'http://www.w3.org/ns/ttml#styling',
4545 class TTMLPElementParser(object):
4547 _unclosed_elements = []
4548 _applied_styles = []
4550 def start(self, tag, attrib):
4551 if tag in (_x('ttml:br'), 'br'):
4554 unclosed_elements = []
4556 element_style_id = attrib.get('style')
4558 style.update(default_style)
4559 if element_style_id:
4560 style.update(styles.get(element_style_id, {}))
4561 for prop in SUPPORTED_STYLING:
4562 prop_val = attrib.get(_x('tts:' + prop))
4564 style[prop] = prop_val
4567 for k, v in sorted(style.items()):
4568 if self._applied_styles and self._applied_styles[-1].get(k) == v:
4571 font += ' color="%s"' % v
4572 elif k == 'fontSize':
4573 font += ' size="%s"' % v
4574 elif k == 'fontFamily':
4575 font += ' face="%s"' % v
4576 elif k == 'fontWeight' and v == 'bold':
4578 unclosed_elements.append('b')
4579 elif k == 'fontStyle' and v == 'italic':
4581 unclosed_elements.append('i')
4582 elif k == 'textDecoration' and v == 'underline':
4584 unclosed_elements.append('u')
4586 self._out += '<font' + font + '>'
4587 unclosed_elements.append('font')
4589 if self._applied_styles:
4590 applied_style.update(self._applied_styles[-1])
4591 applied_style.update(style)
4592 self._applied_styles.append(applied_style)
4593 self._unclosed_elements.append(unclosed_elements)
4596 if tag not in (_x('ttml:br'), 'br'):
4597 unclosed_elements = self._unclosed_elements.pop()
4598 for element in reversed(unclosed_elements):
4599 self._out += '</%s>' % element
4600 if unclosed_elements and self._applied_styles:
4601 self._applied_styles.pop()
4603 def data(self, data):
4607 return self._out.strip()
4609 def parse_node(node):
4610 target = TTMLPElementParser()
4611 parser = xml.etree.ElementTree.XMLParser(target=target)
4612 parser.feed(xml.etree.ElementTree.tostring(node))
4613 return parser.close()
4615 for k, v in LEGACY_NAMESPACES:
4617 dfxp_data = dfxp_data.replace(ns, k)
4619 dfxp = compat_etree_fromstring(dfxp_data)
4621 paras = dfxp.findall(_x('.//ttml:p')) or dfxp.findall('.//p')
4624 raise ValueError('Invalid dfxp/TTML subtitle')
4628 for style in dfxp.findall(_x('.//ttml:style')):
4629 style_id = style.get('id') or style.get(_x('xml:id'))
4632 parent_style_id = style.get('style')
4634 if parent_style_id not in styles:
4637 styles[style_id] = styles[parent_style_id].copy()
4638 for prop in SUPPORTED_STYLING:
4639 prop_val = style.get(_x('tts:' + prop))
4641 styles.setdefault(style_id, {})[prop] = prop_val
4647 for p in ('body', 'div'):
4648 ele = xpath_element(dfxp, [_x('.//ttml:' + p), './/' + p])
4651 style = styles.get(ele.get('style'))
4654 default_style.update(style)
4656 for para, index in zip(paras, itertools.count(1)):
4657 begin_time = parse_dfxp_time_expr(para.attrib.get('begin'))
4658 end_time = parse_dfxp_time_expr(para.attrib.get('end'))
4659 dur = parse_dfxp_time_expr(para.attrib.get('dur'))
4660 if begin_time is None:
4665 end_time = begin_time + dur
4666 out.append('%d\n%s --> %s\n%s\n\n' % (
4668 srt_subtitles_timecode(begin_time),
4669 srt_subtitles_timecode(end_time),
4675 def cli_option(params, command_option, param):
4676 param = params.get(param)
4678 param = compat_str(param)
4679 return [command_option, param] if param is not None else []
4682 def cli_bool_option(params, command_option, param, true_value='true', false_value='false', separator=None):
4683 param = params.get(param)
4686 assert isinstance(param, bool)
4688 return [command_option + separator + (true_value if param else false_value)]
4689 return [command_option, true_value if param else false_value]
4692 def cli_valueless_option(params, command_option, param, expected_value=True):
4693 param = params.get(param)
4694 return [command_option] if param == expected_value else []
4697 def cli_configuration_args(argdict, keys, default=[], use_compat=True):
4698 if isinstance(argdict, (list, tuple)): # for backward compatibility
4705 assert isinstance(argdict, dict)
4707 assert isinstance(keys, (list, tuple))
4708 for key_list in keys:
4709 if isinstance(key_list, compat_str):
4710 key_list = (key_list,)
4711 arg_list = list(filter(
4712 lambda x: x is not None,
4713 [argdict.get(key.lower()) for key in key_list]))
4715 return [arg for args in arg_list for arg in args]
4719 class ISO639Utils(object):
4720 # See http://www.loc.gov/standards/iso639-2/ISO-639-2_utf-8.txt
4779 'iw': 'heb', # Replaced by he in 1989 revision
4789 'in': 'ind', # Replaced by id in 1989 revision
4904 'ji': 'yid', # Replaced by yi in 1989 revision
4912 def short2long(cls, code):
4913 """Convert language code from ISO 639-1 to ISO 639-2/T"""
4914 return cls._lang_map.get(code[:2])
4917 def long2short(cls, code):
4918 """Convert language code from ISO 639-2/T to ISO 639-1"""
4919 for short_name, long_name in cls._lang_map.items():
4920 if long_name == code:
4924 class ISO3166Utils(object):
4925 # From http://data.okfn.org/data/core/country-list
4927 'AF': 'Afghanistan',
4928 'AX': 'Åland Islands',
4931 'AS': 'American Samoa',
4936 'AG': 'Antigua and Barbuda',
4953 'BO': 'Bolivia, Plurinational State of',
4954 'BQ': 'Bonaire, Sint Eustatius and Saba',
4955 'BA': 'Bosnia and Herzegovina',
4957 'BV': 'Bouvet Island',
4959 'IO': 'British Indian Ocean Territory',
4960 'BN': 'Brunei Darussalam',
4962 'BF': 'Burkina Faso',
4968 'KY': 'Cayman Islands',
4969 'CF': 'Central African Republic',
4973 'CX': 'Christmas Island',
4974 'CC': 'Cocos (Keeling) Islands',
4978 'CD': 'Congo, the Democratic Republic of the',
4979 'CK': 'Cook Islands',
4981 'CI': 'Côte d\'Ivoire',
4986 'CZ': 'Czech Republic',
4990 'DO': 'Dominican Republic',
4993 'SV': 'El Salvador',
4994 'GQ': 'Equatorial Guinea',
4998 'FK': 'Falkland Islands (Malvinas)',
4999 'FO': 'Faroe Islands',
5003 'GF': 'French Guiana',
5004 'PF': 'French Polynesia',
5005 'TF': 'French Southern Territories',
5020 'GW': 'Guinea-Bissau',
5023 'HM': 'Heard Island and McDonald Islands',
5024 'VA': 'Holy See (Vatican City State)',
5031 'IR': 'Iran, Islamic Republic of',
5034 'IM': 'Isle of Man',
5044 'KP': 'Korea, Democratic People\'s Republic of',
5045 'KR': 'Korea, Republic of',
5048 'LA': 'Lao People\'s Democratic Republic',
5054 'LI': 'Liechtenstein',
5058 'MK': 'Macedonia, the Former Yugoslav Republic of',
5065 'MH': 'Marshall Islands',
5071 'FM': 'Micronesia, Federated States of',
5072 'MD': 'Moldova, Republic of',
5083 'NL': 'Netherlands',
5084 'NC': 'New Caledonia',
5085 'NZ': 'New Zealand',
5090 'NF': 'Norfolk Island',
5091 'MP': 'Northern Mariana Islands',
5096 'PS': 'Palestine, State of',
5098 'PG': 'Papua New Guinea',
5101 'PH': 'Philippines',
5105 'PR': 'Puerto Rico',
5109 'RU': 'Russian Federation',
5111 'BL': 'Saint Barthélemy',
5112 'SH': 'Saint Helena, Ascension and Tristan da Cunha',
5113 'KN': 'Saint Kitts and Nevis',
5114 'LC': 'Saint Lucia',
5115 'MF': 'Saint Martin (French part)',
5116 'PM': 'Saint Pierre and Miquelon',
5117 'VC': 'Saint Vincent and the Grenadines',
5120 'ST': 'Sao Tome and Principe',
5121 'SA': 'Saudi Arabia',
5125 'SL': 'Sierra Leone',
5127 'SX': 'Sint Maarten (Dutch part)',
5130 'SB': 'Solomon Islands',
5132 'ZA': 'South Africa',
5133 'GS': 'South Georgia and the South Sandwich Islands',
5134 'SS': 'South Sudan',
5139 'SJ': 'Svalbard and Jan Mayen',
5142 'CH': 'Switzerland',
5143 'SY': 'Syrian Arab Republic',
5144 'TW': 'Taiwan, Province of China',
5146 'TZ': 'Tanzania, United Republic of',
5148 'TL': 'Timor-Leste',
5152 'TT': 'Trinidad and Tobago',
5155 'TM': 'Turkmenistan',
5156 'TC': 'Turks and Caicos Islands',
5160 'AE': 'United Arab Emirates',
5161 'GB': 'United Kingdom',
5162 'US': 'United States',
5163 'UM': 'United States Minor Outlying Islands',
5167 'VE': 'Venezuela, Bolivarian Republic of',
5169 'VG': 'Virgin Islands, British',
5170 'VI': 'Virgin Islands, U.S.',
5171 'WF': 'Wallis and Futuna',
5172 'EH': 'Western Sahara',
5179 def short2full(cls, code):
5180 """Convert an ISO 3166-2 country code to the corresponding full name"""
5181 return cls._country_map.get(code.upper())
5184 class GeoUtils(object):
5185 # Major IPv4 address blocks per country
5187 'AD': '46.172.224.0/19',
5188 'AE': '94.200.0.0/13',
5189 'AF': '149.54.0.0/17',
5190 'AG': '209.59.64.0/18',
5191 'AI': '204.14.248.0/21',
5192 'AL': '46.99.0.0/16',
5193 'AM': '46.70.0.0/15',
5194 'AO': '105.168.0.0/13',
5195 'AP': '182.50.184.0/21',
5196 'AQ': '23.154.160.0/24',
5197 'AR': '181.0.0.0/12',
5198 'AS': '202.70.112.0/20',
5199 'AT': '77.116.0.0/14',
5200 'AU': '1.128.0.0/11',
5201 'AW': '181.41.0.0/18',
5202 'AX': '185.217.4.0/22',
5203 'AZ': '5.197.0.0/16',
5204 'BA': '31.176.128.0/17',
5205 'BB': '65.48.128.0/17',
5206 'BD': '114.130.0.0/16',
5208 'BF': '102.178.0.0/15',
5209 'BG': '95.42.0.0/15',
5210 'BH': '37.131.0.0/17',
5211 'BI': '154.117.192.0/18',
5212 'BJ': '137.255.0.0/16',
5213 'BL': '185.212.72.0/23',
5214 'BM': '196.12.64.0/18',
5215 'BN': '156.31.0.0/16',
5216 'BO': '161.56.0.0/16',
5217 'BQ': '161.0.80.0/20',
5218 'BR': '191.128.0.0/12',
5219 'BS': '24.51.64.0/18',
5220 'BT': '119.2.96.0/19',
5221 'BW': '168.167.0.0/16',
5222 'BY': '178.120.0.0/13',
5223 'BZ': '179.42.192.0/18',
5224 'CA': '99.224.0.0/11',
5225 'CD': '41.243.0.0/16',
5226 'CF': '197.242.176.0/21',
5227 'CG': '160.113.0.0/16',
5228 'CH': '85.0.0.0/13',
5229 'CI': '102.136.0.0/14',
5230 'CK': '202.65.32.0/19',
5231 'CL': '152.172.0.0/14',
5232 'CM': '102.244.0.0/14',
5233 'CN': '36.128.0.0/10',
5234 'CO': '181.240.0.0/12',
5235 'CR': '201.192.0.0/12',
5236 'CU': '152.206.0.0/15',
5237 'CV': '165.90.96.0/19',
5238 'CW': '190.88.128.0/17',
5239 'CY': '31.153.0.0/16',
5240 'CZ': '88.100.0.0/14',
5242 'DJ': '197.241.0.0/17',
5243 'DK': '87.48.0.0/12',
5244 'DM': '192.243.48.0/20',
5245 'DO': '152.166.0.0/15',
5246 'DZ': '41.96.0.0/12',
5247 'EC': '186.68.0.0/15',
5248 'EE': '90.190.0.0/15',
5249 'EG': '156.160.0.0/11',
5250 'ER': '196.200.96.0/20',
5251 'ES': '88.0.0.0/11',
5252 'ET': '196.188.0.0/14',
5253 'EU': '2.16.0.0/13',
5254 'FI': '91.152.0.0/13',
5255 'FJ': '144.120.0.0/16',
5256 'FK': '80.73.208.0/21',
5257 'FM': '119.252.112.0/20',
5258 'FO': '88.85.32.0/19',
5260 'GA': '41.158.0.0/15',
5262 'GD': '74.122.88.0/21',
5263 'GE': '31.146.0.0/16',
5264 'GF': '161.22.64.0/18',
5265 'GG': '62.68.160.0/19',
5266 'GH': '154.160.0.0/12',
5267 'GI': '95.164.0.0/16',
5268 'GL': '88.83.0.0/19',
5269 'GM': '160.182.0.0/15',
5270 'GN': '197.149.192.0/18',
5271 'GP': '104.250.0.0/19',
5272 'GQ': '105.235.224.0/20',
5273 'GR': '94.64.0.0/13',
5274 'GT': '168.234.0.0/16',
5275 'GU': '168.123.0.0/16',
5276 'GW': '197.214.80.0/20',
5277 'GY': '181.41.64.0/18',
5278 'HK': '113.252.0.0/14',
5279 'HN': '181.210.0.0/16',
5280 'HR': '93.136.0.0/13',
5281 'HT': '148.102.128.0/17',
5282 'HU': '84.0.0.0/14',
5283 'ID': '39.192.0.0/10',
5284 'IE': '87.32.0.0/12',
5285 'IL': '79.176.0.0/13',
5286 'IM': '5.62.80.0/20',
5287 'IN': '117.192.0.0/10',
5288 'IO': '203.83.48.0/21',
5289 'IQ': '37.236.0.0/14',
5290 'IR': '2.176.0.0/12',
5291 'IS': '82.221.0.0/16',
5292 'IT': '79.0.0.0/10',
5293 'JE': '87.244.64.0/18',
5294 'JM': '72.27.0.0/17',
5295 'JO': '176.29.0.0/16',
5296 'JP': '133.0.0.0/8',
5297 'KE': '105.48.0.0/12',
5298 'KG': '158.181.128.0/17',
5299 'KH': '36.37.128.0/17',
5300 'KI': '103.25.140.0/22',
5301 'KM': '197.255.224.0/20',
5302 'KN': '198.167.192.0/19',
5303 'KP': '175.45.176.0/22',
5304 'KR': '175.192.0.0/10',
5305 'KW': '37.36.0.0/14',
5306 'KY': '64.96.0.0/15',
5307 'KZ': '2.72.0.0/13',
5308 'LA': '115.84.64.0/18',
5309 'LB': '178.135.0.0/16',
5310 'LC': '24.92.144.0/20',
5311 'LI': '82.117.0.0/19',
5312 'LK': '112.134.0.0/15',
5313 'LR': '102.183.0.0/16',
5314 'LS': '129.232.0.0/17',
5315 'LT': '78.56.0.0/13',
5316 'LU': '188.42.0.0/16',
5317 'LV': '46.109.0.0/16',
5318 'LY': '41.252.0.0/14',
5319 'MA': '105.128.0.0/11',
5320 'MC': '88.209.64.0/18',
5321 'MD': '37.246.0.0/16',
5322 'ME': '178.175.0.0/17',
5323 'MF': '74.112.232.0/21',
5324 'MG': '154.126.0.0/17',
5325 'MH': '117.103.88.0/21',
5326 'MK': '77.28.0.0/15',
5327 'ML': '154.118.128.0/18',
5328 'MM': '37.111.0.0/17',
5329 'MN': '49.0.128.0/17',
5330 'MO': '60.246.0.0/16',
5331 'MP': '202.88.64.0/20',
5332 'MQ': '109.203.224.0/19',
5333 'MR': '41.188.64.0/18',
5334 'MS': '208.90.112.0/22',
5335 'MT': '46.11.0.0/16',
5336 'MU': '105.16.0.0/12',
5337 'MV': '27.114.128.0/18',
5338 'MW': '102.70.0.0/15',
5339 'MX': '187.192.0.0/11',
5340 'MY': '175.136.0.0/13',
5341 'MZ': '197.218.0.0/15',
5342 'NA': '41.182.0.0/16',
5343 'NC': '101.101.0.0/18',
5344 'NE': '197.214.0.0/18',
5345 'NF': '203.17.240.0/22',
5346 'NG': '105.112.0.0/12',
5347 'NI': '186.76.0.0/15',
5348 'NL': '145.96.0.0/11',
5349 'NO': '84.208.0.0/13',
5350 'NP': '36.252.0.0/15',
5351 'NR': '203.98.224.0/19',
5352 'NU': '49.156.48.0/22',
5353 'NZ': '49.224.0.0/14',
5354 'OM': '5.36.0.0/15',
5355 'PA': '186.72.0.0/15',
5356 'PE': '186.160.0.0/14',
5357 'PF': '123.50.64.0/18',
5358 'PG': '124.240.192.0/19',
5359 'PH': '49.144.0.0/13',
5360 'PK': '39.32.0.0/11',
5361 'PL': '83.0.0.0/11',
5362 'PM': '70.36.0.0/20',
5363 'PR': '66.50.0.0/16',
5364 'PS': '188.161.0.0/16',
5365 'PT': '85.240.0.0/13',
5366 'PW': '202.124.224.0/20',
5367 'PY': '181.120.0.0/14',
5368 'QA': '37.210.0.0/15',
5369 'RE': '102.35.0.0/16',
5370 'RO': '79.112.0.0/13',
5371 'RS': '93.86.0.0/15',
5372 'RU': '5.136.0.0/13',
5373 'RW': '41.186.0.0/16',
5374 'SA': '188.48.0.0/13',
5375 'SB': '202.1.160.0/19',
5376 'SC': '154.192.0.0/11',
5377 'SD': '102.120.0.0/13',
5378 'SE': '78.64.0.0/12',
5379 'SG': '8.128.0.0/10',
5380 'SI': '188.196.0.0/14',
5381 'SK': '78.98.0.0/15',
5382 'SL': '102.143.0.0/17',
5383 'SM': '89.186.32.0/19',
5384 'SN': '41.82.0.0/15',
5385 'SO': '154.115.192.0/18',
5386 'SR': '186.179.128.0/17',
5387 'SS': '105.235.208.0/21',
5388 'ST': '197.159.160.0/19',
5389 'SV': '168.243.0.0/16',
5390 'SX': '190.102.0.0/20',
5392 'SZ': '41.84.224.0/19',
5393 'TC': '65.255.48.0/20',
5394 'TD': '154.68.128.0/19',
5395 'TG': '196.168.0.0/14',
5396 'TH': '171.96.0.0/13',
5397 'TJ': '85.9.128.0/18',
5398 'TK': '27.96.24.0/21',
5399 'TL': '180.189.160.0/20',
5400 'TM': '95.85.96.0/19',
5401 'TN': '197.0.0.0/11',
5402 'TO': '175.176.144.0/21',
5403 'TR': '78.160.0.0/11',
5404 'TT': '186.44.0.0/15',
5405 'TV': '202.2.96.0/19',
5406 'TW': '120.96.0.0/11',
5407 'TZ': '156.156.0.0/14',
5408 'UA': '37.52.0.0/14',
5409 'UG': '102.80.0.0/13',
5411 'UY': '167.56.0.0/13',
5412 'UZ': '84.54.64.0/18',
5413 'VA': '212.77.0.0/19',
5414 'VC': '207.191.240.0/21',
5415 'VE': '186.88.0.0/13',
5416 'VG': '66.81.192.0/20',
5417 'VI': '146.226.0.0/16',
5418 'VN': '14.160.0.0/11',
5419 'VU': '202.80.32.0/20',
5420 'WF': '117.20.32.0/21',
5421 'WS': '202.4.32.0/19',
5422 'YE': '134.35.0.0/16',
5423 'YT': '41.242.116.0/22',
5424 'ZA': '41.0.0.0/11',
5425 'ZM': '102.144.0.0/13',
5426 'ZW': '102.177.192.0/18',
5430 def random_ipv4(cls, code_or_block):
5431 if len(code_or_block) == 2:
5432 block = cls._country_ip_map.get(code_or_block.upper())
5436 block = code_or_block
5437 addr, preflen = block.split('/')
5438 addr_min = compat_struct_unpack('!L', socket.inet_aton(addr))[0]
5439 addr_max = addr_min | (0xffffffff >> int(preflen))
5440 return compat_str(socket.inet_ntoa(
5441 compat_struct_pack('!L', random.randint(addr_min, addr_max))))
5444 class PerRequestProxyHandler(compat_urllib_request.ProxyHandler):
5445 def __init__(self, proxies=None):
5446 # Set default handlers
5447 for type in ('http', 'https'):
5448 setattr(self, '%s_open' % type,
5449 lambda r, proxy='__noproxy__', type=type, meth=self.proxy_open:
5450 meth(r, proxy, type))
5451 compat_urllib_request.ProxyHandler.__init__(self, proxies)
5453 def proxy_open(self, req, proxy, type):
5454 req_proxy = req.headers.get('Ytdl-request-proxy')
5455 if req_proxy is not None:
5457 del req.headers['Ytdl-request-proxy']
5459 if proxy == '__noproxy__':
5460 return None # No Proxy
5461 if compat_urlparse.urlparse(proxy).scheme.lower() in ('socks', 'socks4', 'socks4a', 'socks5'):
5462 req.add_header('Ytdl-socks-proxy', proxy)
5463 # yt-dlp's http/https handlers do wrapping the socket with socks
5465 return compat_urllib_request.ProxyHandler.proxy_open(
5466 self, req, proxy, type)
5469 # Both long_to_bytes and bytes_to_long are adapted from PyCrypto, which is
5470 # released into Public Domain
5471 # https://github.com/dlitz/pycrypto/blob/master/lib/Crypto/Util/number.py#L387
5473 def long_to_bytes(n, blocksize=0):
5474 """long_to_bytes(n:long, blocksize:int) : string
5475 Convert a long integer to a byte string.
5477 If optional blocksize is given and greater than zero, pad the front of the
5478 byte string with binary zeros so that the length is a multiple of
5481 # after much testing, this algorithm was deemed to be the fastest
5485 s = compat_struct_pack('>I', n & 0xffffffff) + s
5487 # strip off leading zeros
5488 for i in range(len(s)):
5489 if s[i] != b'\000'[0]:
5492 # only happens when n == 0
5496 # add back some pad bytes. this could be done more efficiently w.r.t. the
5497 # de-padding being done above, but sigh...
5498 if blocksize > 0 and len(s) % blocksize:
5499 s = (blocksize - len(s) % blocksize) * b'\000' + s
5503 def bytes_to_long(s):
5504 """bytes_to_long(string) : long
5505 Convert a byte string to a long integer.
5507 This is (essentially) the inverse of long_to_bytes().
5512 extra = (4 - length % 4)
5513 s = b'\000' * extra + s
5514 length = length + extra
5515 for i in range(0, length, 4):
5516 acc = (acc << 32) + compat_struct_unpack('>I', s[i:i + 4])[0]
5520 def ohdave_rsa_encrypt(data, exponent, modulus):
5522 Implement OHDave
's RSA algorithm. See http://www.ohdave.com/rsa/
5525 data: data to encrypt, bytes-like object
5526 exponent, modulus: parameter e and N of RSA algorithm, both integer
5527 Output: hex string of encrypted data
5529 Limitation: supports one block encryption only
5532 payload = int(binascii.hexlify(data[::-1]), 16)
5533 encrypted = pow(payload, exponent, modulus)
5534 return '%x' % encrypted
5537 def pkcs1pad(data, length):
5539 Padding input data with PKCS#1 scheme
5541 @param {int[]} data input data
5542 @param {int} length target length
5543 @returns {int[]} padded data
5545 if len(data) > length - 11:
5546 raise ValueError('Input data too
long for PKCS
#1 padding')
5548 pseudo_random
= [random
.randint(0, 254) for _
in range(length
- len(data
) - 3)]
5549 return [0, 2] + pseudo_random
+ [0] + data
5552 def encode_base_n(num
, n
, table
=None):
5553 FULL_TABLE
= '0123456789abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ'
5555 table
= FULL_TABLE
[:n
]
5558 raise ValueError('base %d exceeds table length %d' % (n
, len(table
)))
5565 ret
= table
[num
% n
] + ret
5570 def decode_packed_codes(code
):
5571 mobj
= re
.search(PACKED_CODES_RE
, code
)
5572 obfuscated_code
, base
, count
, symbols
= mobj
.groups()
5575 symbols
= symbols
.split('|')
5580 base_n_count
= encode_base_n(count
, base
)
5581 symbol_table
[base_n_count
] = symbols
[count
] or base_n_count
5584 r
'\b(\w+)\b', lambda mobj
: symbol_table
[mobj
.group(0)],
5588 def caesar(s
, alphabet
, shift
):
5593 alphabet
[(alphabet
.index(c
) + shift
) % l
] if c
in alphabet
else c
5598 return caesar(s
, r
'''!"#$%&'()*+,-./0123456789:;<=>?@ABCDEFGHIJKLMNOPQRSTUVWXYZ[\]^_`abcdefghijklmnopqrstuvwxyz{|}~''', 47)
5601 def parse_m3u8_attributes(attrib
):
5603 for (key
, val
) in re
.findall(r
'(?P<key>[A-Z0-9-]+)=(?P<val>"[^"]+"|[^",]+)(?:,|$)', attrib
):
5604 if val
.startswith('"'):
5610 def urshift(val
, n
):
5611 return val
>> n
if val
>= 0 else (val
+ 0x100000000) >> n
5614 # Based on png2str() written by @gdkchan and improved by @yokrysty
5615 # Originally posted at https://github.com/ytdl-org/youtube-dl/issues/9706
5616 def decode_png(png_data
):
5617 # Reference: https://www.w3.org/TR/PNG/
5618 header
= png_data
[8:]
5620 if png_data
[:8] != b
'\x89PNG\x0d\x0a\x1a\x0a' or header
[4:8] != b
'IHDR':
5621 raise IOError('Not a valid PNG file.')
5623 int_map
= {1: '>B', 2: '>H', 4: '>I'}
5624 unpack_integer
= lambda x
: compat_struct_unpack(int_map
[len(x
)], x
)[0]
5629 length
= unpack_integer(header
[:4])
5632 chunk_type
= header
[:4]
5635 chunk_data
= header
[:length
]
5636 header
= header
[length
:]
5638 header
= header
[4:] # Skip CRC
5646 ihdr
= chunks
[0]['data']
5648 width
= unpack_integer(ihdr
[:4])
5649 height
= unpack_integer(ihdr
[4:8])
5653 for chunk
in chunks
:
5654 if chunk
['type'] == b
'IDAT':
5655 idat
+= chunk
['data']
5658 raise IOError('Unable to read PNG data.')
5660 decompressed_data
= bytearray(zlib
.decompress(idat
))
5665 def _get_pixel(idx
):
5670 for y
in range(height
):
5671 basePos
= y
* (1 + stride
)
5672 filter_type
= decompressed_data
[basePos
]
5676 pixels
.append(current_row
)
5678 for x
in range(stride
):
5679 color
= decompressed_data
[1 + basePos
+ x
]
5680 basex
= y
* stride
+ x
5685 left
= _get_pixel(basex
- 3)
5687 up
= _get_pixel(basex
- stride
)
5689 if filter_type
== 1: # Sub
5690 color
= (color
+ left
) & 0xff
5691 elif filter_type
== 2: # Up
5692 color
= (color
+ up
) & 0xff
5693 elif filter_type
== 3: # Average
5694 color
= (color
+ ((left
+ up
) >> 1)) & 0xff
5695 elif filter_type
== 4: # Paeth
5701 c
= _get_pixel(basex
- stride
- 3)
5709 if pa
<= pb
and pa
<= pc
:
5710 color
= (color
+ a
) & 0xff
5712 color
= (color
+ b
) & 0xff
5714 color
= (color
+ c
) & 0xff
5716 current_row
.append(color
)
5718 return width
, height
, pixels
5721 def write_xattr(path
, key
, value
):
5722 # This mess below finds the best xattr tool for the job
5724 # try the pyxattr module...
5727 if hasattr(xattr
, 'set'): # pyxattr
5728 # Unicode arguments are not supported in python-pyxattr until
5730 # See https://github.com/ytdl-org/youtube-dl/issues/5498
5731 pyxattr_required_version
= '0.5.0'
5732 if version_tuple(xattr
.__version
__) < version_tuple(pyxattr_required_version
):
5733 # TODO: fallback to CLI tools
5734 raise XAttrUnavailableError(
5735 'python-pyxattr is detected but is too old. '
5736 'yt-dlp requires %s or above while your version is %s. '
5737 'Falling back to other xattr implementations' % (
5738 pyxattr_required_version
, xattr
.__version
__))
5740 setxattr
= xattr
.set
5742 setxattr
= xattr
.setxattr
5745 setxattr(path
, key
, value
)
5746 except EnvironmentError as e
:
5747 raise XAttrMetadataError(e
.errno
, e
.strerror
)
5750 if compat_os_name
== 'nt':
5751 # Write xattrs to NTFS Alternate Data Streams:
5752 # http://en.wikipedia.org/wiki/NTFS#Alternate_data_streams_.28ADS.29
5753 assert ':' not in key
5754 assert os
.path
.exists(path
)
5756 ads_fn
= path
+ ':' + key
5758 with open(ads_fn
, 'wb') as f
:
5760 except EnvironmentError as e
:
5761 raise XAttrMetadataError(e
.errno
, e
.strerror
)
5763 user_has_setfattr
= check_executable('setfattr', ['--version'])
5764 user_has_xattr
= check_executable('xattr', ['-h'])
5766 if user_has_setfattr
or user_has_xattr
:
5768 value
= value
.decode('utf-8')
5769 if user_has_setfattr
:
5770 executable
= 'setfattr'
5771 opts
= ['-n', key
, '-v', value
]
5772 elif user_has_xattr
:
5773 executable
= 'xattr'
5774 opts
= ['-w', key
, value
]
5776 cmd
= ([encodeFilename(executable
, True)]
5777 + [encodeArgument(o
) for o
in opts
]
5778 + [encodeFilename(path
, True)])
5781 p
= subprocess
.Popen(
5782 cmd
, stdout
=subprocess
.PIPE
, stderr
=subprocess
.PIPE
, stdin
=subprocess
.PIPE
)
5783 except EnvironmentError as e
:
5784 raise XAttrMetadataError(e
.errno
, e
.strerror
)
5785 stdout
, stderr
= process_communicate_or_kill(p
)
5786 stderr
= stderr
.decode('utf-8', 'replace')
5787 if p
.returncode
!= 0:
5788 raise XAttrMetadataError(p
.returncode
, stderr
)
5791 # On Unix, and can't find pyxattr, setfattr, or xattr.
5792 if sys
.platform
.startswith('linux'):
5793 raise XAttrUnavailableError(
5794 "Couldn't find a tool to set the xattrs. "
5795 "Install either the python 'pyxattr' or 'xattr' "
5796 "modules, or the GNU 'attr' package "
5797 "(which contains the 'setfattr' tool).")
5799 raise XAttrUnavailableError(
5800 "Couldn't find a tool to set the xattrs. "
5801 "Install either the python 'xattr' module, "
5802 "or the 'xattr' binary.")
5805 def random_birthday(year_field
, month_field
, day_field
):
5806 start_date
= datetime
.date(1950, 1, 1)
5807 end_date
= datetime
.date(1995, 12, 31)
5808 offset
= random
.randint(0, (end_date
- start_date
).days
)
5809 random_date
= start_date
+ datetime
.timedelta(offset
)
5811 year_field
: str(random_date
.year
),
5812 month_field
: str(random_date
.month
),
5813 day_field
: str(random_date
.day
),
5817 # Templates for internet shortcut files, which are plain text files.
5818 DOT_URL_LINK_TEMPLATE
= '''
5823 DOT_WEBLOC_LINK_TEMPLATE
= '''
5824 <?xml version="1.0" encoding="UTF-8"?>
5825 <!DOCTYPE plist PUBLIC "-//Apple//DTD PLIST 1.0//EN" "http://www.apple.com/DTDs/PropertyList-1.0.dtd">
5826 <plist version="1.0">
5829 \t<string>%(url)s</string>
5834 DOT_DESKTOP_LINK_TEMPLATE
= '''
5844 def iri_to_uri(iri
):
5846 Converts an IRI (Internationalized Resource Identifier, allowing Unicode characters) to a URI (Uniform Resource Identifier, ASCII-only).
5848 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.
5851 iri_parts
= compat_urllib_parse_urlparse(iri
)
5853 if '[' in iri_parts
.netloc
:
5854 raise ValueError('IPv6 URIs are not, yet, supported.')
5855 # Querying `.netloc`, when there's only one bracket, also raises a ValueError.
5857 # 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.
5860 if iri_parts
.username
:
5861 net_location
+= compat_urllib_parse_quote(iri_parts
.username
, safe
=r
"!$%&'()*+,~")
5862 if iri_parts
.password
is not None:
5863 net_location
+= ':' + compat_urllib_parse_quote(iri_parts
.password
, safe
=r
"!$%&'()*+,~")
5866 net_location
+= iri_parts
.hostname
.encode('idna').decode('utf-8') # Punycode for Unicode hostnames.
5867 # The 'idna' encoding produces ASCII text.
5868 if iri_parts
.port
is not None and iri_parts
.port
!= 80:
5869 net_location
+= ':' + str(iri_parts
.port
)
5871 return compat_urllib_parse_urlunparse(
5875 compat_urllib_parse_quote_plus(iri_parts
.path
, safe
=r
"!$%&'()*+,/:;=@|~"),
5877 # Unsure about the `safe` argument, since this is a legacy way of handling parameters.
5878 compat_urllib_parse_quote_plus(iri_parts
.params
, safe
=r
"!$%&'()*+,/:;=@|~"),
5880 # Not totally sure about the `safe` argument, since the source does not explicitly mention the query URI component.
5881 compat_urllib_parse_quote_plus(iri_parts
.query
, safe
=r
"!$%&'()*+,/:;=?@{|}~"),
5883 compat_urllib_parse_quote_plus(iri_parts
.fragment
, safe
=r
"!#$%&'()*+,/:;=?@{|}~")))
5885 # Source for `safe` arguments: https://url.spec.whatwg.org/#percent-encoded-bytes.
5888 def to_high_limit_path(path
):
5889 if sys
.platform
in ['win32', 'cygwin']:
5890 # Work around MAX_PATH limitation on Windows. The maximum allowed length for the individual path segments may still be quite limited.
5891 return r
'\\?\ '.rstrip() + os
.path
.abspath(path
)
5896 def format_field(obj
, field
, template
='%s', ignore
=(None, ''), default
='', func
=None):
5897 val
= obj
.get(field
, default
)
5898 if func
and val
not in ignore
:
5900 return template
% val
if val
not in ignore
else default
5903 def clean_podcast_url(url
):
5904 return re
.sub(r
'''(?x)
5908 media\.blubrry\.com| # https://create.blubrry.com/resources/podcast-media-download-statistics/getting-started/
5911 (?:dts|www)\.podtrac\.com/(?:pts/)?redirect\.[0-9a-z]{3,4}| # http://analytics.podtrac.com/how-to-measure
5914 cn\.co| # https://podcorn.com/analytics-prefix/
5915 st\.fm # https://podsights.com/docs/
5920 _HEX_TABLE
= '0123456789abcdef'
5923 def random_uuidv4():
5924 return re
.sub(r
'[xy]', lambda x
: _HEX_TABLE
[random
.randint(0, 15)], 'xxxxxxxx-xxxx-4xxx-yxxx-xxxxxxxxxxxx')
5927 def make_dir(path
, to_screen
=None):
5929 dn
= os
.path
.dirname(path
)
5930 if dn
and not os
.path
.exists(dn
):
5933 except (OSError, IOError) as err
:
5934 if callable(to_screen
) is not None:
5935 to_screen('unable to create directory ' + error_to_compat_str(err
))
5939 def get_executable_path():
5940 from zipimport
import zipimporter
5941 if hasattr(sys
, 'frozen'): # Running from PyInstaller
5942 path
= os
.path
.dirname(sys
.executable
)
5943 elif isinstance(globals().get('__loader__'), zipimporter
): # Running from ZIP
5944 path
= os
.path
.join(os
.path
.dirname(__file__
), '../..')
5946 path
= os
.path
.join(os
.path
.dirname(__file__
), '..')
5947 return os
.path
.abspath(path
)
5950 def load_plugins(name
, type, namespace
):
5951 plugin_info
= [None]
5954 plugin_info
= imp
.find_module(
5955 name
, [os
.path
.join(get_executable_path(), 'ytdlp_plugins')])
5956 plugins
= imp
.load_module(name
, *plugin_info
)
5957 for name
in dir(plugins
):
5958 if not name
.endswith(type):
5960 klass
= getattr(plugins
, name
)
5961 classes
.append(klass
)
5962 namespace
[name
] = klass
5966 if plugin_info
[0] is not None:
5967 plugin_info
[0].close()
5971 def traverse_dict(dictn
, keys
, casesense
=True):
5972 if not isinstance(dictn
, dict):
5976 dictn
= {key.lower(): val for key, val in dictn.items()}
5977 first_key
= first_key
.lower()
5978 value
= dictn
.get(first_key
, None)
5979 return value
if len(keys
) < 2 else traverse_dict(value
, keys
[1:], casesense
)
projects / yt-dlp.git / blob