4 from __future__
import unicode_literals
37 import xml
.etree
.ElementTree
41 compat_HTMLParseError
,
47 compat_ctypes_WINFUNCTYPE
,
48 compat_etree_fromstring
,
51 compat_html_entities_html5
,
64 compat_urllib_parse_urlencode
,
65 compat_urllib_parse_urlparse
,
66 compat_urllib_parse_urlunparse
,
67 compat_urllib_parse_quote
,
68 compat_urllib_parse_quote_plus
,
69 compat_urllib_parse_unquote_plus
,
70 compat_urllib_request
,
81 def register_socks_protocols():
82 # "Register" SOCKS protocols
83 # In Python < 2.6.5, urlsplit() suffers from bug https://bugs.python.org/issue7904
84 # URLs with protocols not in urlparse.uses_netloc are not handled correctly
85 for scheme
in ('socks', 'socks4', 'socks4a', 'socks5'):
86 if scheme
not in compat_urlparse
.uses_netloc
:
87 compat_urlparse
.uses_netloc
.append(scheme
)
90 # This is not clearly defined otherwise
91 compiled_regex_type
= type(re
.compile(''))
94 def random_user_agent():
95 _USER_AGENT_TPL
= 'Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/%s Safari/537.36'
1674 return _USER_AGENT_TPL
% random
.choice(_CHROME_VERSIONS
)
1678 'User-Agent': random_user_agent(),
1679 'Accept-Charset': 'ISO-8859-1,utf-8;q=0.7,*;q=0.7',
1680 'Accept': 'text/html,application/xhtml+xml,application/xml;q=0.9,*/*;q=0.8',
1681 'Accept-Encoding': 'gzip, deflate',
1682 'Accept-Language': 'en-us,en;q=0.5',
1687 'Safari': 'Mozilla/5.0 (X11; Linux x86_64; rv:10.0) AppleWebKit/533.20.25 (KHTML, like Gecko) Version/5.0.4 Safari/533.20.27',
1691 NO_DEFAULT
= object()
1693 ENGLISH_MONTH_NAMES
= [
1694 'January', 'February', 'March', 'April', 'May', 'June',
1695 'July', 'August', 'September', 'October', 'November', 'December']
1698 'en': ENGLISH_MONTH_NAMES
,
1700 'janvier', 'février', 'mars', 'avril', 'mai', 'juin',
1701 'juillet', 'août', 'septembre', 'octobre', 'novembre', 'décembre'],
1704 KNOWN_EXTENSIONS
= (
1705 'mp4', 'm4a', 'm4p', 'm4b', 'm4r', 'm4v', 'aac',
1706 'flv', 'f4v', 'f4a', 'f4b',
1707 'webm', 'ogg', 'ogv', 'oga', 'ogx', 'spx', 'opus',
1708 'mkv', 'mka', 'mk3d',
1711 'asf', 'wmv', 'wma',
1717 'f4f', 'f4m', 'm3u8', 'smil')
1719 # needed for sanitizing filenames in restricted mode
1720 ACCENT_CHARS
= dict(zip('ÂÃÄÀÁÅÆÇÈÉÊËÌÍÎÏÐÑÒÓÔÕÖŐØŒÙÚÛÜŰÝÞßàáâãäåæçèéêëìíîïðñòóôõöőøœùúûüűýþÿ',
1721 itertools
.chain('AAAAAA', ['AE'], 'CEEEEIIIIDNOOOOOOO', ['OE'], 'UUUUUY', ['TH', 'ss'],
1722 'aaaaaa', ['ae'], 'ceeeeiiiionooooooo', ['oe'], 'uuuuuy', ['th'], 'y')))
1745 '%Y/%m/%d %H:%M:%S',
1747 '%Y-%m-%d %H:%M:%S',
1748 '%Y-%m-%d %H:%M:%S.%f',
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))
1839 json.dump(obj, tf, default=repr)
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:
2113 s
= re
.sub(r
'[0-9]+(?::[0-9]+)+', lambda m
: m
.group(0).replace(':', '_'), s
)
2114 result
= ''.join(map(replace_insane
, s
))
2116 while '__' in result
:
2117 result
= result
.replace('__', '_')
2118 result
= result
.strip('_')
2119 # Common case of "Foreign band name - English song title"
2120 if restricted
and result
.startswith('-_'):
2122 if result
.startswith('-'):
2123 result
= '_' + result
[len('-'):]
2124 result
= result
.lstrip('.')
2130 def sanitize_path(s
, force
=False):
2131 """Sanitizes and normalizes path on Windows"""
2132 if sys
.platform
== 'win32':
2134 drive_or_unc
, _
= os
.path
.splitdrive(s
)
2135 if sys
.version_info
< (2, 7) and not drive_or_unc
:
2136 drive_or_unc
, _
= os
.path
.splitunc(s
)
2142 norm_path
= os
.path
.normpath(remove_start(s
, drive_or_unc
)).split(os
.path
.sep
)
2146 path_part
if path_part
in ['.', '..'] else re
.sub(r
'(?:[/<>:"\|\\?\*]|[\s.]$)', '#', path_part
)
2147 for path_part
in norm_path
]
2149 sanitized_path
.insert(0, drive_or_unc
+ os
.path
.sep
)
2150 elif force
and s
[0] == os
.path
.sep
:
2151 sanitized_path
.insert(0, os
.path
.sep
)
2152 return os
.path
.join(*sanitized_path
)
2155 def sanitize_url(url
):
2156 # Prepend protocol-less URLs with `http:` scheme in order to mitigate
2157 # the number of unwanted failures due to missing protocol
2158 if url
.startswith('//'):
2159 return 'http:%s' % url
2160 # Fix some common typos seen so far
2162 # https://github.com/ytdl-org/youtube-dl/issues/15649
2163 (r
'^httpss://', r
'https://'),
2164 # https://bx1.be/lives/direct-tv/
2165 (r
'^rmtp([es]?)://', r
'rtmp\1://'),
2167 for mistake
, fixup
in COMMON_TYPOS
:
2168 if re
.match(mistake
, url
):
2169 return re
.sub(mistake
, fixup
, url
)
2173 def extract_basic_auth(url
):
2174 parts
= compat_urlparse
.urlsplit(url
)
2175 if parts
.username
is None:
2177 url
= compat_urlparse
.urlunsplit(parts
._replace
(netloc
=(
2178 parts
.hostname
if parts
.port
is None
2179 else '%s:%d' % (parts
.hostname
, parts
.port
))))
2180 auth_payload
= base64
.b64encode(
2181 ('%s:%s' % (parts
.username
, parts
.password
or '')).encode('utf-8'))
2182 return url
, 'Basic ' + auth_payload
.decode('utf-8')
2185 def sanitized_Request(url
, *args
, **kwargs
):
2186 url
, auth_header
= extract_basic_auth(escape_url(sanitize_url(url
)))
2187 if auth_header
is not None:
2188 headers
= args
[1] if len(args
) >= 2 else kwargs
.setdefault('headers', {})
2189 headers
['Authorization'] = auth_header
2190 return compat_urllib_request
.Request(url
, *args
, **kwargs
)
2194 """Expand shell variables and ~"""
2195 return os
.path
.expandvars(compat_expanduser(s
))
2198 def orderedSet(iterable
):
2199 """ Remove all duplicates from the input iterable """
2207 def _htmlentity_transform(entity_with_semicolon
):
2208 """Transforms an HTML entity to a character."""
2209 entity
= entity_with_semicolon
[:-1]
2211 # Known non-numeric HTML entity
2212 if entity
in compat_html_entities
.name2codepoint
:
2213 return compat_chr(compat_html_entities
.name2codepoint
[entity
])
2215 # TODO: HTML5 allows entities without a semicolon. For example,
2216 # 'Éric' should be decoded as 'Éric'.
2217 if entity_with_semicolon
in compat_html_entities_html5
:
2218 return compat_html_entities_html5
[entity_with_semicolon
]
2220 mobj
= re
.match(r
'#(x[0-9a-fA-F]+|[0-9]+)', entity
)
2221 if mobj
is not None:
2222 numstr
= mobj
.group(1)
2223 if numstr
.startswith('x'):
2225 numstr
= '0%s' % numstr
2228 # See https://github.com/ytdl-org/youtube-dl/issues/7518
2230 return compat_chr(int(numstr
, base
))
2234 # Unknown entity in name, return its literal representation
2235 return '&%s;' % entity
2238 def unescapeHTML(s
):
2241 assert type(s
) == compat_str
2244 r
'&([^&;]+;)', lambda m
: _htmlentity_transform(m
.group(1)), s
)
2247 def process_communicate_or_kill(p
, *args
, **kwargs
):
2249 return p
.communicate(*args
, **kwargs
)
2250 except BaseException
: # Including KeyboardInterrupt
2256 def get_subprocess_encoding():
2257 if sys
.platform
== 'win32' and sys
.getwindowsversion()[0] >= 5:
2258 # For subprocess calls, encode with locale encoding
2259 # Refer to http://stackoverflow.com/a/9951851/35070
2260 encoding
= preferredencoding()
2262 encoding
= sys
.getfilesystemencoding()
2263 if encoding
is None:
2268 def encodeFilename(s
, for_subprocess
=False):
2270 @param s The name of the file
2273 assert type(s
) == compat_str
2275 # Python 3 has a Unicode API
2276 if sys
.version_info
>= (3, 0):
2279 # Pass '' directly to use Unicode APIs on Windows 2000 and up
2280 # (Detecting Windows NT 4 is tricky because 'major >= 4' would
2281 # match Windows 9x series as well. Besides, NT 4 is obsolete.)
2282 if not for_subprocess
and sys
.platform
== 'win32' and sys
.getwindowsversion()[0] >= 5:
2285 # Jython assumes filenames are Unicode strings though reported as Python 2.x compatible
2286 if sys
.platform
.startswith('java'):
2289 return s
.encode(get_subprocess_encoding(), 'ignore')
2292 def decodeFilename(b
, for_subprocess
=False):
2294 if sys
.version_info
>= (3, 0):
2297 if not isinstance(b
, bytes):
2300 return b
.decode(get_subprocess_encoding(), 'ignore')
2303 def encodeArgument(s
):
2304 if not isinstance(s
, compat_str
):
2305 # Legacy code that uses byte strings
2306 # Uncomment the following line after fixing all post processors
2307 # assert False, 'Internal error: %r should be of type %r, is %r' % (s, compat_str, type(s))
2308 s
= s
.decode('ascii')
2309 return encodeFilename(s
, True)
2312 def decodeArgument(b
):
2313 return decodeFilename(b
, True)
2316 def decodeOption(optval
):
2319 if isinstance(optval
, bytes):
2320 optval
= optval
.decode(preferredencoding())
2322 assert isinstance(optval
, compat_str
)
2326 def formatSeconds(secs
, delim
=':'):
2328 return '%d%s%02d%s%02d' % (secs
// 3600, delim
, (secs
% 3600) // 60, delim
, secs
% 60)
2330 return '%d%s%02d' % (secs
// 60, delim
, secs
% 60)
2335 def make_HTTPS_handler(params
, **kwargs
):
2336 opts_no_check_certificate
= params
.get('nocheckcertificate', False)
2337 if hasattr(ssl
, 'create_default_context'): # Python >= 3.4 or 2.7.9
2338 context
= ssl
.create_default_context(ssl
.Purpose
.SERVER_AUTH
)
2339 if opts_no_check_certificate
:
2340 context
.check_hostname
= False
2341 context
.verify_mode
= ssl
.CERT_NONE
2343 return YoutubeDLHTTPSHandler(params
, context
=context
, **kwargs
)
2346 # (create_default_context present but HTTPSHandler has no context=)
2349 if sys
.version_info
< (3, 2):
2350 return YoutubeDLHTTPSHandler(params
, **kwargs
)
2351 else: # Python < 3.4
2352 context
= ssl
.SSLContext(ssl
.PROTOCOL_TLSv1
)
2353 context
.verify_mode
= (ssl
.CERT_NONE
2354 if opts_no_check_certificate
2355 else ssl
.CERT_REQUIRED
)
2356 context
.set_default_verify_paths()
2357 return YoutubeDLHTTPSHandler(params
, context
=context
, **kwargs
)
2360 def bug_reports_message(before
=';'):
2361 if ytdl_is_updateable():
2362 update_cmd
= 'type yt-dlp -U to update'
2364 update_cmd
= 'see https://github.com/yt-dlp/yt-dlp on how to update'
2365 msg
= 'please report this issue on https://github.com/yt-dlp/yt-dlp .'
2366 msg
+= ' Make sure you are using the latest version; %s.' % update_cmd
2367 msg
+= ' Be sure to call yt-dlp with the --verbose flag and include its complete output.'
2369 before
= before
.rstrip()
2370 if not before
or before
.endswith(('.', '!', '?')):
2371 msg
= msg
[0].title() + msg
[1:]
2373 return (before
+ ' ' if before
else '') + msg
2376 class YoutubeDLError(Exception):
2377 """Base exception for YoutubeDL errors."""
2381 network_exceptions
= [compat_urllib_error
.URLError
, compat_http_client
.HTTPException
, socket
.error
]
2382 if hasattr(ssl
, 'CertificateError'):
2383 network_exceptions
.append(ssl
.CertificateError
)
2384 network_exceptions
= tuple(network_exceptions
)
2387 class ExtractorError(YoutubeDLError
):
2388 """Error during info extraction."""
2390 def __init__(self
, msg
, tb
=None, expected
=False, cause
=None, video_id
=None):
2391 """ tb, if given, is the original traceback (so that it can be printed out).
2392 If expected is set, this is a normal error message and most likely not a bug in yt-dlp.
2395 if sys
.exc_info()[0] in network_exceptions
:
2397 if video_id
is not None:
2398 msg
= video_id
+ ': ' + msg
2400 msg
+= ' (caused by %r)' % cause
2402 msg
+= bug_reports_message()
2403 super(ExtractorError
, self
).__init
__(msg
)
2406 self
.exc_info
= sys
.exc_info() # preserve original exception
2408 self
.video_id
= video_id
2410 def format_traceback(self
):
2411 if self
.traceback
is None:
2413 return ''.join(traceback
.format_tb(self
.traceback
))
2416 class UnsupportedError(ExtractorError
):
2417 def __init__(self
, url
):
2418 super(UnsupportedError
, self
).__init
__(
2419 'Unsupported URL: %s' % url
, expected
=True)
2423 class RegexNotFoundError(ExtractorError
):
2424 """Error when a regex didn't match"""
2428 class GeoRestrictedError(ExtractorError
):
2429 """Geographic restriction Error exception.
2431 This exception may be thrown when a video is not available from your
2432 geographic location due to geographic restrictions imposed by a website.
2435 def __init__(self
, msg
, countries
=None):
2436 super(GeoRestrictedError
, self
).__init
__(msg
, expected
=True)
2438 self
.countries
= countries
2441 class DownloadError(YoutubeDLError
):
2442 """Download Error exception.
2444 This exception may be thrown by FileDownloader objects if they are not
2445 configured to continue on errors. They will contain the appropriate
2449 def __init__(self
, msg
, exc_info
=None):
2450 """ exc_info, if given, is the original exception that caused the trouble (as returned by sys.exc_info()). """
2451 super(DownloadError
, self
).__init
__(msg
)
2452 self
.exc_info
= exc_info
2455 class EntryNotInPlaylist(YoutubeDLError
):
2456 """Entry not in playlist exception.
2458 This exception will be thrown by YoutubeDL when a requested entry
2459 is not found in the playlist info_dict
2464 class SameFileError(YoutubeDLError
):
2465 """Same File exception.
2467 This exception will be thrown by FileDownloader objects if they detect
2468 multiple files would have to be downloaded to the same file on disk.
2473 class PostProcessingError(YoutubeDLError
):
2474 """Post Processing exception.
2476 This exception may be raised by PostProcessor's .run() method to
2477 indicate an error in the postprocessing task.
2480 def __init__(self
, msg
):
2481 super(PostProcessingError
, self
).__init
__(msg
)
2485 class ExistingVideoReached(YoutubeDLError
):
2486 """ --max-downloads limit has been reached. """
2490 class RejectedVideoReached(YoutubeDLError
):
2491 """ --max-downloads limit has been reached. """
2495 class MaxDownloadsReached(YoutubeDLError
):
2496 """ --max-downloads limit has been reached. """
2500 class UnavailableVideoError(YoutubeDLError
):
2501 """Unavailable Format exception.
2503 This exception will be thrown when a video is requested
2504 in a format that is not available for that video.
2509 class ContentTooShortError(YoutubeDLError
):
2510 """Content Too Short exception.
2512 This exception may be raised by FileDownloader objects when a file they
2513 download is too small for what the server announced first, indicating
2514 the connection was probably interrupted.
2517 def __init__(self
, downloaded
, expected
):
2518 super(ContentTooShortError
, self
).__init
__(
2519 'Downloaded {0} bytes, expected {1} bytes'.format(downloaded
, expected
)
2522 self
.downloaded
= downloaded
2523 self
.expected
= expected
2526 class XAttrMetadataError(YoutubeDLError
):
2527 def __init__(self
, code
=None, msg
='Unknown error'):
2528 super(XAttrMetadataError
, self
).__init
__(msg
)
2532 # Parsing code and msg
2533 if (self
.code
in (errno
.ENOSPC
, errno
.EDQUOT
)
2534 or 'No space left' in self
.msg
or 'Disk quota exceeded' in self
.msg
):
2535 self
.reason
= 'NO_SPACE'
2536 elif self
.code
== errno
.E2BIG
or 'Argument list too long' in self
.msg
:
2537 self
.reason
= 'VALUE_TOO_LONG'
2539 self
.reason
= 'NOT_SUPPORTED'
2542 class XAttrUnavailableError(YoutubeDLError
):
2546 def _create_http_connection(ydl_handler
, http_class
, is_https
, *args
, **kwargs
):
2547 # Working around python 2 bug (see http://bugs.python.org/issue17849) by limiting
2548 # expected HTTP responses to meet HTTP/1.0 or later (see also
2549 # https://github.com/ytdl-org/youtube-dl/issues/6727)
2550 if sys
.version_info
< (3, 0):
2551 kwargs
['strict'] = True
2552 hc
= http_class(*args
, **compat_kwargs(kwargs
))
2553 source_address
= ydl_handler
._params
.get('source_address')
2555 if source_address
is not None:
2556 # This is to workaround _create_connection() from socket where it will try all
2557 # address data from getaddrinfo() including IPv6. This filters the result from
2558 # getaddrinfo() based on the source_address value.
2559 # This is based on the cpython socket.create_connection() function.
2560 # https://github.com/python/cpython/blob/master/Lib/socket.py#L691
2561 def _create_connection(address
, timeout
=socket
._GLOBAL
_DEFAULT
_TIMEOUT
, source_address
=None):
2562 host
, port
= address
2564 addrs
= socket
.getaddrinfo(host
, port
, 0, socket
.SOCK_STREAM
)
2565 af
= socket
.AF_INET
if '.' in source_address
[0] else socket
.AF_INET6
2566 ip_addrs
= [addr
for addr
in addrs
if addr
[0] == af
]
2567 if addrs
and not ip_addrs
:
2568 ip_version
= 'v4' if af
== socket
.AF_INET
else 'v6'
2570 "No remote IP%s addresses available for connect, can't use '%s' as source address"
2571 % (ip_version
, source_address
[0]))
2572 for res
in ip_addrs
:
2573 af
, socktype
, proto
, canonname
, sa
= res
2576 sock
= socket
.socket(af
, socktype
, proto
)
2577 if timeout
is not socket
._GLOBAL
_DEFAULT
_TIMEOUT
:
2578 sock
.settimeout(timeout
)
2579 sock
.bind(source_address
)
2581 err
= None # Explicitly break reference cycle
2583 except socket
.error
as _
:
2585 if sock
is not None:
2590 raise socket
.error('getaddrinfo returns an empty list')
2591 if hasattr(hc
, '_create_connection'):
2592 hc
._create
_connection
= _create_connection
2593 sa
= (source_address
, 0)
2594 if hasattr(hc
, 'source_address'): # Python 2.7+
2595 hc
.source_address
= sa
2597 def _hc_connect(self
, *args
, **kwargs
):
2598 sock
= _create_connection(
2599 (self
.host
, self
.port
), self
.timeout
, sa
)
2601 self
.sock
= ssl
.wrap_socket(
2602 sock
, self
.key_file
, self
.cert_file
,
2603 ssl_version
=ssl
.PROTOCOL_TLSv1
)
2606 hc
.connect
= functools
.partial(_hc_connect
, hc
)
2611 def handle_youtubedl_headers(headers
):
2612 filtered_headers
= headers
2614 if 'Youtubedl-no-compression' in filtered_headers
:
2615 filtered_headers
= dict((k
, v
) for k
, v
in filtered_headers
.items() if k
.lower() != 'accept-encoding')
2616 del filtered_headers
['Youtubedl-no-compression']
2618 return filtered_headers
2621 class YoutubeDLHandler(compat_urllib_request
.HTTPHandler
):
2622 """Handler for HTTP requests and responses.
2624 This class, when installed with an OpenerDirector, automatically adds
2625 the standard headers to every HTTP request and handles gzipped and
2626 deflated responses from web servers. If compression is to be avoided in
2627 a particular request, the original request in the program code only has
2628 to include the HTTP header "Youtubedl-no-compression", which will be
2629 removed before making the real request.
2631 Part of this code was copied from:
2633 http://techknack.net/python-urllib2-handlers/
2635 Andrew Rowls, the author of that code, agreed to release it to the
2639 def __init__(self
, params
, *args
, **kwargs
):
2640 compat_urllib_request
.HTTPHandler
.__init
__(self
, *args
, **kwargs
)
2641 self
._params
= params
2643 def http_open(self
, req
):
2644 conn_class
= compat_http_client
.HTTPConnection
2646 socks_proxy
= req
.headers
.get('Ytdl-socks-proxy')
2648 conn_class
= make_socks_conn_class(conn_class
, socks_proxy
)
2649 del req
.headers
['Ytdl-socks-proxy']
2651 return self
.do_open(functools
.partial(
2652 _create_http_connection
, self
, conn_class
, False),
2660 return zlib
.decompress(data
, -zlib
.MAX_WBITS
)
2662 return zlib
.decompress(data
)
2664 def http_request(self
, req
):
2665 # According to RFC 3986, URLs can not contain non-ASCII characters, however this is not
2666 # always respected by websites, some tend to give out URLs with non percent-encoded
2667 # non-ASCII characters (see telemb.py, ard.py [#3412])
2668 # urllib chokes on URLs with non-ASCII characters (see http://bugs.python.org/issue3991)
2669 # To work around aforementioned issue we will replace request's original URL with
2670 # percent-encoded one
2671 # Since redirects are also affected (e.g. http://www.southpark.de/alle-episoden/s18e09)
2672 # the code of this workaround has been moved here from YoutubeDL.urlopen()
2673 url
= req
.get_full_url()
2674 url_escaped
= escape_url(url
)
2676 # Substitute URL if any change after escaping
2677 if url
!= url_escaped
:
2678 req
= update_Request(req
, url
=url_escaped
)
2680 for h
, v
in std_headers
.items():
2681 # Capitalize is needed because of Python bug 2275: http://bugs.python.org/issue2275
2682 # The dict keys are capitalized because of this bug by urllib
2683 if h
.capitalize() not in req
.headers
:
2684 req
.add_header(h
, v
)
2686 req
.headers
= handle_youtubedl_headers(req
.headers
)
2688 if sys
.version_info
< (2, 7) and '#' in req
.get_full_url():
2689 # Python 2.6 is brain-dead when it comes to fragments
2690 req
._Request
__original
= req
._Request
__original
.partition('#')[0]
2691 req
._Request
__r
_type
= req
._Request
__r
_type
.partition('#')[0]
2695 def http_response(self
, req
, resp
):
2698 if resp
.headers
.get('Content-encoding', '') == 'gzip':
2699 content
= resp
.read()
2700 gz
= gzip
.GzipFile(fileobj
=io
.BytesIO(content
), mode
='rb')
2702 uncompressed
= io
.BytesIO(gz
.read())
2703 except IOError as original_ioerror
:
2704 # There may be junk add the end of the file
2705 # See http://stackoverflow.com/q/4928560/35070 for details
2706 for i
in range(1, 1024):
2708 gz
= gzip
.GzipFile(fileobj
=io
.BytesIO(content
[:-i
]), mode
='rb')
2709 uncompressed
= io
.BytesIO(gz
.read())
2714 raise original_ioerror
2715 resp
= compat_urllib_request
.addinfourl(uncompressed
, old_resp
.headers
, old_resp
.url
, old_resp
.code
)
2716 resp
.msg
= old_resp
.msg
2717 del resp
.headers
['Content-encoding']
2719 if resp
.headers
.get('Content-encoding', '') == 'deflate':
2720 gz
= io
.BytesIO(self
.deflate(resp
.read()))
2721 resp
= compat_urllib_request
.addinfourl(gz
, old_resp
.headers
, old_resp
.url
, old_resp
.code
)
2722 resp
.msg
= old_resp
.msg
2723 del resp
.headers
['Content-encoding']
2724 # Percent-encode redirect URL of Location HTTP header to satisfy RFC 3986 (see
2725 # https://github.com/ytdl-org/youtube-dl/issues/6457).
2726 if 300 <= resp
.code
< 400:
2727 location
= resp
.headers
.get('Location')
2729 # As of RFC 2616 default charset is iso-8859-1 that is respected by python 3
2730 if sys
.version_info
>= (3, 0):
2731 location
= location
.encode('iso-8859-1').decode('utf-8')
2733 location
= location
.decode('utf-8')
2734 location_escaped
= escape_url(location
)
2735 if location
!= location_escaped
:
2736 del resp
.headers
['Location']
2737 if sys
.version_info
< (3, 0):
2738 location_escaped
= location_escaped
.encode('utf-8')
2739 resp
.headers
['Location'] = location_escaped
2742 https_request
= http_request
2743 https_response
= http_response
2746 def make_socks_conn_class(base_class
, socks_proxy
):
2747 assert issubclass(base_class
, (
2748 compat_http_client
.HTTPConnection
, compat_http_client
.HTTPSConnection
))
2750 url_components
= compat_urlparse
.urlparse(socks_proxy
)
2751 if url_components
.scheme
.lower() == 'socks5':
2752 socks_type
= ProxyType
.SOCKS5
2753 elif url_components
.scheme
.lower() in ('socks', 'socks4'):
2754 socks_type
= ProxyType
.SOCKS4
2755 elif url_components
.scheme
.lower() == 'socks4a':
2756 socks_type
= ProxyType
.SOCKS4A
2758 def unquote_if_non_empty(s
):
2761 return compat_urllib_parse_unquote_plus(s
)
2765 url_components
.hostname
, url_components
.port
or 1080,
2767 unquote_if_non_empty(url_components
.username
),
2768 unquote_if_non_empty(url_components
.password
),
2771 class SocksConnection(base_class
):
2773 self
.sock
= sockssocket()
2774 self
.sock
.setproxy(*proxy_args
)
2775 if type(self
.timeout
) in (int, float):
2776 self
.sock
.settimeout(self
.timeout
)
2777 self
.sock
.connect((self
.host
, self
.port
))
2779 if isinstance(self
, compat_http_client
.HTTPSConnection
):
2780 if hasattr(self
, '_context'): # Python > 2.6
2781 self
.sock
= self
._context
.wrap_socket(
2782 self
.sock
, server_hostname
=self
.host
)
2784 self
.sock
= ssl
.wrap_socket(self
.sock
)
2786 return SocksConnection
2789 class YoutubeDLHTTPSHandler(compat_urllib_request
.HTTPSHandler
):
2790 def __init__(self
, params
, https_conn_class
=None, *args
, **kwargs
):
2791 compat_urllib_request
.HTTPSHandler
.__init
__(self
, *args
, **kwargs
)
2792 self
._https
_conn
_class
= https_conn_class
or compat_http_client
.HTTPSConnection
2793 self
._params
= params
2795 def https_open(self
, req
):
2797 conn_class
= self
._https
_conn
_class
2799 if hasattr(self
, '_context'): # python > 2.6
2800 kwargs
['context'] = self
._context
2801 if hasattr(self
, '_check_hostname'): # python 3.x
2802 kwargs
['check_hostname'] = self
._check
_hostname
2804 socks_proxy
= req
.headers
.get('Ytdl-socks-proxy')
2806 conn_class
= make_socks_conn_class(conn_class
, socks_proxy
)
2807 del req
.headers
['Ytdl-socks-proxy']
2809 return self
.do_open(functools
.partial(
2810 _create_http_connection
, self
, conn_class
, True),
2814 class YoutubeDLCookieJar(compat_cookiejar
.MozillaCookieJar
):
2816 See [1] for cookie file format.
2818 1. https://curl.haxx.se/docs/http-cookies.html
2820 _HTTPONLY_PREFIX
= '#HttpOnly_'
2822 _HEADER
= '''# Netscape HTTP Cookie File
2823 # This file is generated by yt-dlp. Do not edit.
2826 _CookieFileEntry
= collections
.namedtuple(
2828 ('domain_name', 'include_subdomains', 'path', 'https_only', 'expires_at', 'name', 'value'))
2830 def save(self
, filename
=None, ignore_discard
=False, ignore_expires
=False):
2832 Save cookies to a file.
2834 Most of the code is taken from CPython 3.8 and slightly adapted
2835 to support cookie files with UTF-8 in both python 2 and 3.
2837 if filename
is None:
2838 if self
.filename
is not None:
2839 filename
= self
.filename
2841 raise ValueError(compat_cookiejar
.MISSING_FILENAME_TEXT
)
2843 # Store session cookies with `expires` set to 0 instead of an empty
2846 if cookie
.expires
is None:
2849 with io
.open(filename
, 'w', encoding
='utf-8') as f
:
2850 f
.write(self
._HEADER
)
2853 if not ignore_discard
and cookie
.discard
:
2855 if not ignore_expires
and cookie
.is_expired(now
):
2861 if cookie
.domain
.startswith('.'):
2862 initial_dot
= 'TRUE'
2864 initial_dot
= 'FALSE'
2865 if cookie
.expires
is not None:
2866 expires
= compat_str(cookie
.expires
)
2869 if cookie
.value
is None:
2870 # cookies.txt regards 'Set-Cookie: foo' as a cookie
2871 # with no name, whereas http.cookiejar regards it as a
2872 # cookie with no value.
2877 value
= cookie
.value
2879 '\t'.join([cookie
.domain
, initial_dot
, cookie
.path
,
2880 secure
, expires
, name
, value
]) + '\n')
2882 def load(self
, filename
=None, ignore_discard
=False, ignore_expires
=False):
2883 """Load cookies from a file."""
2884 if filename
is None:
2885 if self
.filename
is not None:
2886 filename
= self
.filename
2888 raise ValueError(compat_cookiejar
.MISSING_FILENAME_TEXT
)
2890 def prepare_line(line
):
2891 if line
.startswith(self
._HTTPONLY
_PREFIX
):
2892 line
= line
[len(self
._HTTPONLY
_PREFIX
):]
2893 # comments and empty lines are fine
2894 if line
.startswith('#') or not line
.strip():
2896 cookie_list
= line
.split('\t')
2897 if len(cookie_list
) != self
._ENTRY
_LEN
:
2898 raise compat_cookiejar
.LoadError('invalid length %d' % len(cookie_list
))
2899 cookie
= self
._CookieFileEntry
(*cookie_list
)
2900 if cookie
.expires_at
and not cookie
.expires_at
.isdigit():
2901 raise compat_cookiejar
.LoadError('invalid expires at %s' % cookie
.expires_at
)
2905 with io
.open(filename
, encoding
='utf-8') as f
:
2908 cf
.write(prepare_line(line
))
2909 except compat_cookiejar
.LoadError
as e
:
2911 'WARNING: skipping cookie file entry due to %s: %r\n'
2912 % (e
, line
), sys
.stderr
)
2915 self
._really
_load
(cf
, filename
, ignore_discard
, ignore_expires
)
2916 # Session cookies are denoted by either `expires` field set to
2917 # an empty string or 0. MozillaCookieJar only recognizes the former
2918 # (see [1]). So we need force the latter to be recognized as session
2919 # cookies on our own.
2920 # Session cookies may be important for cookies-based authentication,
2921 # e.g. usually, when user does not check 'Remember me' check box while
2922 # logging in on a site, some important cookies are stored as session
2923 # cookies so that not recognizing them will result in failed login.
2924 # 1. https://bugs.python.org/issue17164
2926 # Treat `expires=0` cookies as session cookies
2927 if cookie
.expires
== 0:
2928 cookie
.expires
= None
2929 cookie
.discard
= True
2932 class YoutubeDLCookieProcessor(compat_urllib_request
.HTTPCookieProcessor
):
2933 def __init__(self
, cookiejar
=None):
2934 compat_urllib_request
.HTTPCookieProcessor
.__init
__(self
, cookiejar
)
2936 def http_response(self
, request
, response
):
2937 # Python 2 will choke on next HTTP request in row if there are non-ASCII
2938 # characters in Set-Cookie HTTP header of last response (see
2939 # https://github.com/ytdl-org/youtube-dl/issues/6769).
2940 # In order to at least prevent crashing we will percent encode Set-Cookie
2941 # header before HTTPCookieProcessor starts processing it.
2942 # if sys.version_info < (3, 0) and response.headers:
2943 # for set_cookie_header in ('Set-Cookie', 'Set-Cookie2'):
2944 # set_cookie = response.headers.get(set_cookie_header)
2946 # set_cookie_escaped = compat_urllib_parse.quote(set_cookie, b"%/;:@&=+$,!~*'()?#[] ")
2947 # if set_cookie != set_cookie_escaped:
2948 # del response.headers[set_cookie_header]
2949 # response.headers[set_cookie_header] = set_cookie_escaped
2950 return compat_urllib_request
.HTTPCookieProcessor
.http_response(self
, request
, response
)
2952 https_request
= compat_urllib_request
.HTTPCookieProcessor
.http_request
2953 https_response
= http_response
2956 class YoutubeDLRedirectHandler(compat_urllib_request
.HTTPRedirectHandler
):
2957 """YoutubeDL redirect handler
2959 The code is based on HTTPRedirectHandler implementation from CPython [1].
2961 This redirect handler solves two issues:
2962 - ensures redirect URL is always unicode under python 2
2963 - introduces support for experimental HTTP response status code
2964 308 Permanent Redirect [2] used by some sites [3]
2966 1. https://github.com/python/cpython/blob/master/Lib/urllib/request.py
2967 2. https://developer.mozilla.org/en-US/docs/Web/HTTP/Status/308
2968 3. https://github.com/ytdl-org/youtube-dl/issues/28768
2971 http_error_301
= http_error_303
= http_error_307
= http_error_308
= compat_urllib_request
.HTTPRedirectHandler
.http_error_302
2973 def redirect_request(self
, req
, fp
, code
, msg
, headers
, newurl
):
2974 """Return a Request or None in response to a redirect.
2976 This is called by the http_error_30x methods when a
2977 redirection response is received. If a redirection should
2978 take place, return a new Request to allow http_error_30x to
2979 perform the redirect. Otherwise, raise HTTPError if no-one
2980 else should try to handle this url. Return None if you can't
2981 but another Handler might.
2983 m
= req
.get_method()
2984 if (not (code
in (301, 302, 303, 307, 308) and m
in ("GET", "HEAD")
2985 or code
in (301, 302, 303) and m
== "POST")):
2986 raise compat_HTTPError(req
.full_url
, code
, msg
, headers
, fp
)
2987 # Strictly (according to RFC 2616), 301 or 302 in response to
2988 # a POST MUST NOT cause a redirection without confirmation
2989 # from the user (of urllib.request, in this case). In practice,
2990 # essentially all clients do redirect in this case, so we do
2993 # On python 2 urlh.geturl() may sometimes return redirect URL
2994 # as byte string instead of unicode. This workaround allows
2995 # to force it always return unicode.
2996 if sys
.version_info
[0] < 3:
2997 newurl
= compat_str(newurl
)
2999 # Be conciliant with URIs containing a space. This is mainly
3000 # redundant with the more complete encoding done in http_error_302(),
3001 # but it is kept for compatibility with other callers.
3002 newurl
= newurl
.replace(' ', '%20')
3004 CONTENT_HEADERS
= ("content-length", "content-type")
3005 # NB: don't use dict comprehension for python 2.6 compatibility
3006 newheaders
= dict((k
, v
) for k
, v
in req
.headers
.items()
3007 if k
.lower() not in CONTENT_HEADERS
)
3008 return compat_urllib_request
.Request(
3009 newurl
, headers
=newheaders
, origin_req_host
=req
.origin_req_host
,
3013 def extract_timezone(date_str
):
3015 r
'^.{8,}?(?P<tz>Z$| ?(?P<sign>\+|-)(?P<hours>[0-9]{2}):?(?P<minutes>[0-9]{2})$)',
3018 timezone
= datetime
.timedelta()
3020 date_str
= date_str
[:-len(m
.group('tz'))]
3021 if not m
.group('sign'):
3022 timezone
= datetime
.timedelta()
3024 sign
= 1 if m
.group('sign') == '+' else -1
3025 timezone
= datetime
.timedelta(
3026 hours
=sign
* int(m
.group('hours')),
3027 minutes
=sign
* int(m
.group('minutes')))
3028 return timezone
, date_str
3031 def parse_iso8601(date_str
, delimiter
='T', timezone
=None):
3032 """ Return a UNIX timestamp from the given date """
3034 if date_str
is None:
3037 date_str
= re
.sub(r
'\.[0-9]+', '', date_str
)
3039 if timezone
is None:
3040 timezone
, date_str
= extract_timezone(date_str
)
3043 date_format
= '%Y-%m-%d{0}%H:%M:%S'.format(delimiter
)
3044 dt
= datetime
.datetime
.strptime(date_str
, date_format
) - timezone
3045 return calendar
.timegm(dt
.timetuple())
3050 def date_formats(day_first
=True):
3051 return DATE_FORMATS_DAY_FIRST
if day_first
else DATE_FORMATS_MONTH_FIRST
3054 def unified_strdate(date_str
, day_first
=True):
3055 """Return a string with the date in the format YYYYMMDD"""
3057 if date_str
is None:
3061 date_str
= date_str
.replace(',', ' ')
3062 # Remove AM/PM + timezone
3063 date_str
= re
.sub(r
'(?i)\s*(?:AM|PM)(?:\s+[A-Z]+)?', '', date_str
)
3064 _
, date_str
= extract_timezone(date_str
)
3066 for expression
in date_formats(day_first
):
3068 upload_date
= datetime
.datetime
.strptime(date_str
, expression
).strftime('%Y%m%d')
3071 if upload_date
is None:
3072 timetuple
= email
.utils
.parsedate_tz(date_str
)
3075 upload_date
= datetime
.datetime(*timetuple
[:6]).strftime('%Y%m%d')
3078 if upload_date
is not None:
3079 return compat_str(upload_date
)
3082 def unified_timestamp(date_str
, day_first
=True):
3083 if date_str
is None:
3086 date_str
= re
.sub(r
'[,|]', '', date_str
)
3088 pm_delta
= 12 if re
.search(r
'(?i)PM', date_str
) else 0
3089 timezone
, date_str
= extract_timezone(date_str
)
3091 # Remove AM/PM + timezone
3092 date_str
= re
.sub(r
'(?i)\s*(?:AM|PM)(?:\s+[A-Z]+)?', '', date_str
)
3094 # Remove unrecognized timezones from ISO 8601 alike timestamps
3095 m
= re
.search(r
'\d{1,2}:\d{1,2}(?:\.\d+)?(?P<tz>\s*[A-Z]+)$', date_str
)
3097 date_str
= date_str
[:-len(m
.group('tz'))]
3099 # Python only supports microseconds, so remove nanoseconds
3100 m
= re
.search(r
'^([0-9]{4,}-[0-9]{1,2}-[0-9]{1,2}T[0-9]{1,2}:[0-9]{1,2}:[0-9]{1,2}\.[0-9]{6})[0-9]+$', date_str
)
3102 date_str
= m
.group(1)
3104 for expression
in date_formats(day_first
):
3106 dt
= datetime
.datetime
.strptime(date_str
, expression
) - timezone
+ datetime
.timedelta(hours
=pm_delta
)
3107 return calendar
.timegm(dt
.timetuple())
3110 timetuple
= email
.utils
.parsedate_tz(date_str
)
3112 return calendar
.timegm(timetuple
) + pm_delta
* 3600
3115 def determine_ext(url
, default_ext
='unknown_video'):
3116 if url
is None or '.' not in url
:
3118 guess
= url
.partition('?')[0].rpartition('.')[2]
3119 if re
.match(r
'^[A-Za-z0-9]+$', guess
):
3121 # Try extract ext from URLs like http://example.com/foo/bar.mp4/?download
3122 elif guess
.rstrip('/') in KNOWN_EXTENSIONS
:
3123 return guess
.rstrip('/')
3128 def subtitles_filename(filename
, sub_lang
, sub_format
, expected_real_ext
=None):
3129 return replace_extension(filename
, sub_lang
+ '.' + sub_format
, expected_real_ext
)
3132 def datetime_from_str(date_str
, precision
='auto', format
='%Y%m%d'):
3134 Return a datetime object from a string in the format YYYYMMDD or
3135 (now|today|date)[+-][0-9](microsecond|second|minute|hour|day|week|month|year)(s)?
3137 format: string date format used to return datetime object from
3138 precision: round the time portion of a datetime object.
3139 auto|microsecond|second|minute|hour|day.
3140 auto: round to the unit provided in date_str (if applicable).
3142 auto_precision
= False
3143 if precision
== 'auto':
3144 auto_precision
= True
3145 precision
= 'microsecond'
3146 today
= datetime_round(datetime
.datetime
.now(), precision
)
3147 if date_str
in ('now', 'today'):
3149 if date_str
== 'yesterday':
3150 return today
- datetime
.timedelta(days
=1)
3152 r
'(?P<start>.+)(?P<sign>[+-])(?P<time>\d+)(?P<unit>microsecond|second|minute|hour|day|week|month|year)(s)?',
3154 if match
is not None:
3155 start_time
= datetime_from_str(match
.group('start'), precision
, format
)
3156 time
= int(match
.group('time')) * (-1 if match
.group('sign') == '-' else 1)
3157 unit
= match
.group('unit')
3158 if unit
== 'month' or unit
== 'year':
3159 new_date
= datetime_add_months(start_time
, time
* 12 if unit
== 'year' else time
)
3165 delta
= datetime
.timedelta(**{unit + 's': time}
)
3166 new_date
= start_time
+ delta
3168 return datetime_round(new_date
, unit
)
3171 return datetime_round(datetime
.datetime
.strptime(date_str
, format
), precision
)
3174 def date_from_str(date_str
, format
='%Y%m%d'):
3176 Return a datetime object from a string in the format YYYYMMDD or
3177 (now|today|date)[+-][0-9](microsecond|second|minute|hour|day|week|month|year)(s)?
3179 format: string date format used to return datetime object from
3181 return datetime_from_str(date_str
, precision
='microsecond', format
=format
).date()
3184 def datetime_add_months(dt
, months
):
3185 """Increment/Decrement a datetime object by months."""
3186 month
= dt
.month
+ months
- 1
3187 year
= dt
.year
+ month
// 12
3188 month
= month
% 12 + 1
3189 day
= min(dt
.day
, calendar
.monthrange(year
, month
)[1])
3190 return dt
.replace(year
, month
, day
)
3193 def datetime_round(dt
, precision
='day'):
3195 Round a datetime object's time to a specific precision
3197 if precision
== 'microsecond':
3206 roundto
= lambda x
, n
: ((x
+ n
/ 2) // n
) * n
3207 timestamp
= calendar
.timegm(dt
.timetuple())
3208 return datetime
.datetime
.utcfromtimestamp(roundto(timestamp
, unit_seconds
[precision
]))
3211 def hyphenate_date(date_str
):
3213 Convert a date in 'YYYYMMDD' format to 'YYYY-MM-DD' format"""
3214 match
= re
.match(r
'^(\d\d\d\d)(\d\d)(\d\d)$', date_str
)
3215 if match
is not None:
3216 return '-'.join(match
.groups())
3221 class DateRange(object):
3222 """Represents a time interval between two dates"""
3224 def __init__(self
, start
=None, end
=None):
3225 """start and end must be strings in the format accepted by date"""
3226 if start
is not None:
3227 self
.start
= date_from_str(start
)
3229 self
.start
= datetime
.datetime
.min.date()
3231 self
.end
= date_from_str(end
)
3233 self
.end
= datetime
.datetime
.max.date()
3234 if self
.start
> self
.end
:
3235 raise ValueError('Date range: "%s" , the start date must be before the end date' % self
)
3239 """Returns a range that only contains the given day"""
3240 return cls(day
, day
)
3242 def __contains__(self
, date
):
3243 """Check if the date is in the range"""
3244 if not isinstance(date
, datetime
.date
):
3245 date
= date_from_str(date
)
3246 return self
.start
<= date
<= self
.end
3249 return '%s - %s' % (self
.start
.isoformat(), self
.end
.isoformat())
3252 def platform_name():
3253 """ Returns the platform name as a compat_str """
3254 res
= platform
.platform()
3255 if isinstance(res
, bytes):
3256 res
= res
.decode(preferredencoding())
3258 assert isinstance(res
, compat_str
)
3262 def _windows_write_string(s
, out
):
3263 """ Returns True if the string was written using special methods,
3264 False if it has yet to be written out."""
3265 # Adapted from http://stackoverflow.com/a/3259271/35070
3268 import ctypes
.wintypes
3276 fileno
= out
.fileno()
3277 except AttributeError:
3278 # If the output stream doesn't have a fileno, it's virtual
3280 except io
.UnsupportedOperation
:
3281 # Some strange Windows pseudo files?
3283 if fileno
not in WIN_OUTPUT_IDS
:
3286 GetStdHandle
= compat_ctypes_WINFUNCTYPE(
3287 ctypes
.wintypes
.HANDLE
, ctypes
.wintypes
.DWORD
)(
3288 ('GetStdHandle', ctypes
.windll
.kernel32
))
3289 h
= GetStdHandle(WIN_OUTPUT_IDS
[fileno
])
3291 WriteConsoleW
= compat_ctypes_WINFUNCTYPE(
3292 ctypes
.wintypes
.BOOL
, ctypes
.wintypes
.HANDLE
, ctypes
.wintypes
.LPWSTR
,
3293 ctypes
.wintypes
.DWORD
, ctypes
.POINTER(ctypes
.wintypes
.DWORD
),
3294 ctypes
.wintypes
.LPVOID
)(('WriteConsoleW', ctypes
.windll
.kernel32
))
3295 written
= ctypes
.wintypes
.DWORD(0)
3297 GetFileType
= compat_ctypes_WINFUNCTYPE(ctypes
.wintypes
.DWORD
, ctypes
.wintypes
.DWORD
)(('GetFileType', ctypes
.windll
.kernel32
))
3298 FILE_TYPE_CHAR
= 0x0002
3299 FILE_TYPE_REMOTE
= 0x8000
3300 GetConsoleMode
= compat_ctypes_WINFUNCTYPE(
3301 ctypes
.wintypes
.BOOL
, ctypes
.wintypes
.HANDLE
,
3302 ctypes
.POINTER(ctypes
.wintypes
.DWORD
))(
3303 ('GetConsoleMode', ctypes
.windll
.kernel32
))
3304 INVALID_HANDLE_VALUE
= ctypes
.wintypes
.DWORD(-1).value
3306 def not_a_console(handle
):
3307 if handle
== INVALID_HANDLE_VALUE
or handle
is None:
3309 return ((GetFileType(handle
) & ~FILE_TYPE_REMOTE
) != FILE_TYPE_CHAR
3310 or GetConsoleMode(handle
, ctypes
.byref(ctypes
.wintypes
.DWORD())) == 0)
3312 if not_a_console(h
):
3315 def next_nonbmp_pos(s
):
3317 return next(i
for i
, c
in enumerate(s
) if ord(c
) > 0xffff)
3318 except StopIteration:
3322 count
= min(next_nonbmp_pos(s
), 1024)
3324 ret
= WriteConsoleW(
3325 h
, s
, count
if count
else 2, ctypes
.byref(written
), None)
3327 raise OSError('Failed to write string')
3328 if not count
: # We just wrote a non-BMP character
3329 assert written
.value
== 2
3332 assert written
.value
> 0
3333 s
= s
[written
.value
:]
3337 def write_string(s
, out
=None, encoding
=None):
3340 assert type(s
) == compat_str
3342 if sys
.platform
== 'win32' and encoding
is None and hasattr(out
, 'fileno'):
3343 if _windows_write_string(s
, out
):
3346 if ('b' in getattr(out
, 'mode', '')
3347 or sys
.version_info
[0] < 3): # Python 2 lies about mode of sys.stderr
3348 byt
= s
.encode(encoding
or preferredencoding(), 'ignore')
3350 elif hasattr(out
, 'buffer'):
3351 enc
= encoding
or getattr(out
, 'encoding', None) or preferredencoding()
3352 byt
= s
.encode(enc
, 'ignore')
3353 out
.buffer.write(byt
)
3359 def bytes_to_intlist(bs
):
3362 if isinstance(bs
[0], int): # Python 3
3365 return [ord(c
) for c
in bs
]
3368 def intlist_to_bytes(xs
):
3371 return compat_struct_pack('%dB' % len(xs
), *xs
)
3374 # Cross-platform file locking
3375 if sys
.platform
== 'win32':
3376 import ctypes
.wintypes
3379 class OVERLAPPED(ctypes
.Structure
):
3381 ('Internal', ctypes
.wintypes
.LPVOID
),
3382 ('InternalHigh', ctypes
.wintypes
.LPVOID
),
3383 ('Offset', ctypes
.wintypes
.DWORD
),
3384 ('OffsetHigh', ctypes
.wintypes
.DWORD
),
3385 ('hEvent', ctypes
.wintypes
.HANDLE
),
3388 kernel32
= ctypes
.windll
.kernel32
3389 LockFileEx
= kernel32
.LockFileEx
3390 LockFileEx
.argtypes
= [
3391 ctypes
.wintypes
.HANDLE
, # hFile
3392 ctypes
.wintypes
.DWORD
, # dwFlags
3393 ctypes
.wintypes
.DWORD
, # dwReserved
3394 ctypes
.wintypes
.DWORD
, # nNumberOfBytesToLockLow
3395 ctypes
.wintypes
.DWORD
, # nNumberOfBytesToLockHigh
3396 ctypes
.POINTER(OVERLAPPED
) # Overlapped
3398 LockFileEx
.restype
= ctypes
.wintypes
.BOOL
3399 UnlockFileEx
= kernel32
.UnlockFileEx
3400 UnlockFileEx
.argtypes
= [
3401 ctypes
.wintypes
.HANDLE
, # hFile
3402 ctypes
.wintypes
.DWORD
, # dwReserved
3403 ctypes
.wintypes
.DWORD
, # nNumberOfBytesToLockLow
3404 ctypes
.wintypes
.DWORD
, # nNumberOfBytesToLockHigh
3405 ctypes
.POINTER(OVERLAPPED
) # Overlapped
3407 UnlockFileEx
.restype
= ctypes
.wintypes
.BOOL
3408 whole_low
= 0xffffffff
3409 whole_high
= 0x7fffffff
3411 def _lock_file(f
, exclusive
):
3412 overlapped
= OVERLAPPED()
3413 overlapped
.Offset
= 0
3414 overlapped
.OffsetHigh
= 0
3415 overlapped
.hEvent
= 0
3416 f
._lock
_file
_overlapped
_p
= ctypes
.pointer(overlapped
)
3417 handle
= msvcrt
.get_osfhandle(f
.fileno())
3418 if not LockFileEx(handle
, 0x2 if exclusive
else 0x0, 0,
3419 whole_low
, whole_high
, f
._lock
_file
_overlapped
_p
):
3420 raise OSError('Locking file failed: %r' % ctypes
.FormatError())
3422 def _unlock_file(f
):
3423 assert f
._lock
_file
_overlapped
_p
3424 handle
= msvcrt
.get_osfhandle(f
.fileno())
3425 if not UnlockFileEx(handle
, 0,
3426 whole_low
, whole_high
, f
._lock
_file
_overlapped
_p
):
3427 raise OSError('Unlocking file failed: %r' % ctypes
.FormatError())
3430 # Some platforms, such as Jython, is missing fcntl
3434 def _lock_file(f
, exclusive
):
3435 fcntl
.flock(f
, fcntl
.LOCK_EX
if exclusive
else fcntl
.LOCK_SH
)
3437 def _unlock_file(f
):
3438 fcntl
.flock(f
, fcntl
.LOCK_UN
)
3440 UNSUPPORTED_MSG
= 'file locking is not supported on this platform'
3442 def _lock_file(f
, exclusive
):
3443 raise IOError(UNSUPPORTED_MSG
)
3445 def _unlock_file(f
):
3446 raise IOError(UNSUPPORTED_MSG
)
3449 class locked_file(object):
3450 def __init__(self
, filename
, mode
, encoding
=None):
3451 assert mode
in ['r', 'a', 'w']
3452 self
.f
= io
.open(filename
, mode
, encoding
=encoding
)
3455 def __enter__(self
):
3456 exclusive
= self
.mode
!= 'r'
3458 _lock_file(self
.f
, exclusive
)
3464 def __exit__(self
, etype
, value
, traceback
):
3466 _unlock_file(self
.f
)
3473 def write(self
, *args
):
3474 return self
.f
.write(*args
)
3476 def read(self
, *args
):
3477 return self
.f
.read(*args
)
3480 def get_filesystem_encoding():
3481 encoding
= sys
.getfilesystemencoding()
3482 return encoding
if encoding
is not None else 'utf-8'
3485 def shell_quote(args
):
3487 encoding
= get_filesystem_encoding()
3489 if isinstance(a
, bytes):
3490 # We may get a filename encoded with 'encodeFilename'
3491 a
= a
.decode(encoding
)
3492 quoted_args
.append(compat_shlex_quote(a
))
3493 return ' '.join(quoted_args
)
3496 def smuggle_url(url
, data
):
3497 """ Pass additional data in a URL for internal use. """
3499 url
, idata
= unsmuggle_url(url
, {})
3501 sdata
= compat_urllib_parse_urlencode(
3502 {'__youtubedl_smuggle': json.dumps(data)}
)
3503 return url
+ '#' + sdata
3506 def unsmuggle_url(smug_url
, default
=None):
3507 if '#__youtubedl_smuggle' not in smug_url
:
3508 return smug_url
, default
3509 url
, _
, sdata
= smug_url
.rpartition('#')
3510 jsond
= compat_parse_qs(sdata
)['__youtubedl_smuggle'][0]
3511 data
= json
.loads(jsond
)
3515 def format_bytes(bytes):
3518 if type(bytes) is str:
3519 bytes = float(bytes)
3523 exponent
= int(math
.log(bytes, 1024.0))
3524 suffix
= ['B', 'KiB', 'MiB', 'GiB', 'TiB', 'PiB', 'EiB', 'ZiB', 'YiB'][exponent
]
3525 converted
= float(bytes) / float(1024 ** exponent
)
3526 return '%.2f%s' % (converted
, suffix
)
3529 def lookup_unit_table(unit_table
, s
):
3530 units_re
= '|'.join(re
.escape(u
) for u
in unit_table
)
3532 r
'(?P<num>[0-9]+(?:[,.][0-9]*)?)\s*(?P<unit>%s)\b' % units_re
, s
)
3535 num_str
= m
.group('num').replace(',', '.')
3536 mult
= unit_table
[m
.group('unit')]
3537 return int(float(num_str
) * mult
)
3540 def parse_filesize(s
):
3544 # The lower-case forms are of course incorrect and unofficial,
3545 # but we support those too
3562 'megabytes': 1000 ** 2,
3563 'mebibytes': 1024 ** 2,
3569 'gigabytes': 1000 ** 3,
3570 'gibibytes': 1024 ** 3,
3576 'terabytes': 1000 ** 4,
3577 'tebibytes': 1024 ** 4,
3583 'petabytes': 1000 ** 5,
3584 'pebibytes': 1024 ** 5,
3590 'exabytes': 1000 ** 6,
3591 'exbibytes': 1024 ** 6,
3597 'zettabytes': 1000 ** 7,
3598 'zebibytes': 1024 ** 7,
3604 'yottabytes': 1000 ** 8,
3605 'yobibytes': 1024 ** 8,
3608 return lookup_unit_table(_UNIT_TABLE
, s
)
3617 if re
.match(r
'^[\d,.]+$', s
):
3618 return str_to_int(s
)
3629 return lookup_unit_table(_UNIT_TABLE
, s
)
3632 def parse_resolution(s
):
3636 mobj
= re
.search(r
'\b(?P<w>\d+)\s*[xX×]\s*(?P<h>\d+)\b', s
)
3639 'width': int(mobj
.group('w')),
3640 'height': int(mobj
.group('h')),
3643 mobj
= re
.search(r
'\b(\d+)[pPiI]\b', s
)
3645 return {'height': int(mobj.group(1))}
3647 mobj
= re
.search(r
'\b([48])[kK]\b', s
)
3649 return {'height': int(mobj.group(1)) * 540}
3654 def parse_bitrate(s
):
3655 if not isinstance(s
, compat_str
):
3657 mobj
= re
.search(r
'\b(\d+)\s*kbps', s
)
3659 return int(mobj
.group(1))
3662 def month_by_name(name
, lang
='en'):
3663 """ Return the number of a month by (locale-independently) English name """
3665 month_names
= MONTH_NAMES
.get(lang
, MONTH_NAMES
['en'])
3668 return month_names
.index(name
) + 1
3673 def month_by_abbreviation(abbrev
):
3674 """ Return the number of a month by (locale-independently) English
3678 return [s
[:3] for s
in ENGLISH_MONTH_NAMES
].index(abbrev
) + 1
3683 def fix_xml_ampersands(xml_str
):
3684 """Replace all the '&' by '&' in XML"""
3686 r
'&(?!amp;|lt;|gt;|apos;|quot;|#x[0-9a-fA-F]{,4};|#[0-9]{,4};)',
3691 def setproctitle(title
):
3692 assert isinstance(title
, compat_str
)
3694 # ctypes in Jython is not complete
3695 # http://bugs.jython.org/issue2148
3696 if sys
.platform
.startswith('java'):
3700 libc
= ctypes
.cdll
.LoadLibrary('libc.so.6')
3704 # LoadLibrary in Windows Python 2.7.13 only expects
3705 # a bytestring, but since unicode_literals turns
3706 # every string into a unicode string, it fails.
3708 title_bytes
= title
.encode('utf-8')
3709 buf
= ctypes
.create_string_buffer(len(title_bytes
))
3710 buf
.value
= title_bytes
3712 libc
.prctl(15, buf
, 0, 0, 0)
3713 except AttributeError:
3714 return # Strange libc, just skip this
3717 def remove_start(s
, start
):
3718 return s
[len(start
):] if s
is not None and s
.startswith(start
) else s
3721 def remove_end(s
, end
):
3722 return s
[:-len(end
)] if s
is not None and s
.endswith(end
) else s
3725 def remove_quotes(s
):
3726 if s
is None or len(s
) < 2:
3728 for quote
in ('"', "'", ):
3729 if s
[0] == quote
and s
[-1] == quote
:
3734 def get_domain(url
):
3735 domain
= re
.match(r
'(?:https?:\/\/)?(?:www\.)?(?P<domain>[^\n\/]+\.[^\n\/]+)(?:\/(.*))?', url
)
3736 return domain
.group('domain') if domain
else None
3739 def url_basename(url
):
3740 path
= compat_urlparse
.urlparse(url
).path
3741 return path
.strip('/').split('/')[-1]
3745 return re
.match(r
'https?://[^?#&]+/', url
).group()
3748 def urljoin(base
, path
):
3749 if isinstance(path
, bytes):
3750 path
= path
.decode('utf-8')
3751 if not isinstance(path
, compat_str
) or not path
:
3753 if re
.match(r
'^(?:[a-zA-Z][a-zA-Z0-9+-.]*:)?//', path
):
3755 if isinstance(base
, bytes):
3756 base
= base
.decode('utf-8')
3757 if not isinstance(base
, compat_str
) or not re
.match(
3758 r
'^(?:https?:)?//', base
):
3760 return compat_urlparse
.urljoin(base
, path
)
3763 class HEADRequest(compat_urllib_request
.Request
):
3764 def get_method(self
):
3768 class PUTRequest(compat_urllib_request
.Request
):
3769 def get_method(self
):
3773 def int_or_none(v
, scale
=1, default
=None, get_attr
=None, invscale
=1):
3776 v
= getattr(v
, get_attr
, None)
3782 return int(v
) * invscale
// scale
3783 except (ValueError, TypeError):
3787 def str_or_none(v
, default
=None):
3788 return default
if v
is None else compat_str(v
)
3791 def str_to_int(int_str
):
3792 """ A more relaxed version of int_or_none """
3793 if isinstance(int_str
, compat_integer_types
):
3795 elif isinstance(int_str
, compat_str
):
3796 int_str
= re
.sub(r
'[,\.\+]', '', int_str
)
3797 return int_or_none(int_str
)
3800 def float_or_none(v
, scale
=1, invscale
=1, default
=None):
3804 return float(v
) * invscale
/ scale
3805 except (ValueError, TypeError):
3809 def bool_or_none(v
, default
=None):
3810 return v
if isinstance(v
, bool) else default
3813 def strip_or_none(v
, default
=None):
3814 return v
.strip() if isinstance(v
, compat_str
) else default
3817 def url_or_none(url
):
3818 if not url
or not isinstance(url
, compat_str
):
3821 return url
if re
.match(r
'^(?:(?:https?|rt(?:m(?:pt?[es]?|fp)|sp[su]?)|mms|ftps?):)?//', url
) else None
3824 def strftime_or_none(timestamp
, date_format
, default
=None):
3825 datetime_object
= None
3827 if isinstance(timestamp
, compat_numeric_types
): # unix timestamp
3828 datetime_object
= datetime
.datetime
.utcfromtimestamp(timestamp
)
3829 elif isinstance(timestamp
, compat_str
): # assume YYYYMMDD
3830 datetime_object
= datetime
.datetime
.strptime(timestamp
, '%Y%m%d')
3831 return datetime_object
.strftime(date_format
)
3832 except (ValueError, TypeError, AttributeError):
3836 def parse_duration(s
):
3837 if not isinstance(s
, compat_basestring
):
3842 days
, hours
, mins
, secs
, ms
= [None] * 5
3843 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
)
3845 days
, hours
, mins
, secs
, ms
= m
.groups()
3850 [0-9]+\s*y(?:ears?)?\s*
3853 [0-9]+\s*m(?:onths?)?\s*
3856 [0-9]+\s*w(?:eeks?)?\s*
3859 (?P<days>[0-9]+)\s*d(?:ays?)?\s*
3863 (?P<hours>[0-9]+)\s*h(?:ours?)?\s*
3866 (?P<mins>[0-9]+)\s*m(?:in(?:ute)?s?)?\s*
3869 (?P<secs>[0-9]+)(?P<ms>\.[0-9]+)?\s*s(?:ec(?:ond)?s?)?\s*
3872 days
, hours
, mins
, secs
, ms
= m
.groups()
3874 m
= re
.match(r
'(?i)(?:(?P<hours>[0-9.]+)\s*(?:hours?)|(?P<mins>[0-9.]+)\s*(?:mins?\.?|minutes?)\s*)Z?$', s
)
3876 hours
, mins
= m
.groups()
3882 duration
+= float(secs
)
3884 duration
+= float(mins
) * 60
3886 duration
+= float(hours
) * 60 * 60
3888 duration
+= float(days
) * 24 * 60 * 60
3890 duration
+= float(ms
)
3894 def prepend_extension(filename
, ext
, expected_real_ext
=None):
3895 name
, real_ext
= os
.path
.splitext(filename
)
3897 '{0}.{1}{2}'.format(name
, ext
, real_ext
)
3898 if not expected_real_ext
or real_ext
[1:] == expected_real_ext
3899 else '{0}.{1}'.format(filename
, ext
))
3902 def replace_extension(filename
, ext
, expected_real_ext
=None):
3903 name
, real_ext
= os
.path
.splitext(filename
)
3904 return '{0}.{1}'.format(
3905 name
if not expected_real_ext
or real_ext
[1:] == expected_real_ext
else filename
,
3909 def check_executable(exe
, args
=[]):
3910 """ Checks if the given binary is installed somewhere in PATH, and returns its name.
3911 args can be a list of arguments for a short output (like -version) """
3913 process_communicate_or_kill(subprocess
.Popen(
3914 [exe
] + args
, stdout
=subprocess
.PIPE
, stderr
=subprocess
.PIPE
))
3920 def get_exe_version(exe
, args
=['--version'],
3921 version_re
=None, unrecognized
='present'):
3922 """ Returns the version of the specified executable,
3923 or False if the executable is not present """
3925 # STDIN should be redirected too. On UNIX-like systems, ffmpeg triggers
3926 # SIGTTOU if yt-dlp is run in the background.
3927 # See https://github.com/ytdl-org/youtube-dl/issues/955#issuecomment-209789656
3928 out
, _
= process_communicate_or_kill(subprocess
.Popen(
3929 [encodeArgument(exe
)] + args
,
3930 stdin
=subprocess
.PIPE
,
3931 stdout
=subprocess
.PIPE
, stderr
=subprocess
.STDOUT
))
3934 if isinstance(out
, bytes): # Python 2.x
3935 out
= out
.decode('ascii', 'ignore')
3936 return detect_exe_version(out
, version_re
, unrecognized
)
3939 def detect_exe_version(output
, version_re
=None, unrecognized
='present'):
3940 assert isinstance(output
, compat_str
)
3941 if version_re
is None:
3942 version_re
= r
'version\s+([-0-9._a-zA-Z]+)'
3943 m
= re
.search(version_re
, output
)
3950 class LazyList(collections
.Sequence
):
3951 ''' Lazy immutable list from an iterable
3952 Note that slices of a LazyList are lists and not LazyList'''
3954 def __init__(self
, iterable
):
3955 self
.__iterable
= iter(iterable
)
3957 self
.__reversed
= False
3961 # We need to consume the entire iterable to iterate in reverse
3962 yield from self
.exhaust()[::-1]
3964 yield from self
.__cache
3965 for item
in self
.__iterable
:
3966 self
.__cache
.append(item
)
3970 ''' Evaluate the entire iterable '''
3971 self
.__cache
.extend(self
.__iterable
)
3975 def _reverse_index(x
):
3978 def __getitem__(self
, idx
):
3979 if isinstance(idx
, slice):
3980 step
= idx
.step
or 1
3981 start
= idx
.start
if idx
.start
is not None else 0 if step
> 0 else -1
3982 stop
= idx
.stop
if idx
.stop
is not None else -1 if step
> 0 else 0
3984 start
, stop
, step
= map(self
._reverse
_index
, (start
, stop
, step
))
3985 idx
= slice(start
, stop
, step
)
3986 elif isinstance(idx
, int):
3988 idx
= self
._reverse
_index
(idx
)
3991 raise TypeError('indices must be integers or slices')
3992 if start
< 0 or stop
< 0:
3993 # We need to consume the entire iterable to be able to slice from the end
3994 # Obviously, never use this with infinite iterables
3995 return self
.exhaust()[idx
]
3997 n
= max(start
, stop
) - len(self
.__cache
) + 1
3999 self
.__cache
.extend(itertools
.islice(self
.__iterable
, n
))
4000 return self
.__cache
[idx
]
4004 self
[-1] if self
.__reversed
else self
[0]
4011 return len(self
.__cache
)
4013 def __reversed__(self
):
4014 self
.__reversed
= not self
.__reversed
4018 # repr and str should mimic a list. So we exhaust the iterable
4019 return repr(self
.exhaust())
4022 return repr(self
.exhaust())
4025 class PagedList(object):
4027 # This is only useful for tests
4028 return len(self
.getslice())
4030 def getslice(self
, start
, end
):
4031 raise NotImplementedError('This method must be implemented by subclasses')
4033 def __getitem__(self
, idx
):
4034 if not isinstance(idx
, int) or idx
< 0:
4035 raise TypeError('indices must be non-negative integers')
4036 entries
= self
.getslice(idx
, idx
+ 1)
4037 return entries
[0] if entries
else None
4040 class OnDemandPagedList(PagedList
):
4041 def __init__(self
, pagefunc
, pagesize
, use_cache
=True):
4042 self
._pagefunc
= pagefunc
4043 self
._pagesize
= pagesize
4044 self
._use
_cache
= use_cache
4048 def getslice(self
, start
=0, end
=None):
4050 for pagenum
in itertools
.count(start
// self
._pagesize
):
4051 firstid
= pagenum
* self
._pagesize
4052 nextfirstid
= pagenum
* self
._pagesize
+ self
._pagesize
4053 if start
>= nextfirstid
:
4058 page_results
= self
._cache
.get(pagenum
)
4059 if page_results
is None:
4060 page_results
= list(self
._pagefunc
(pagenum
))
4062 self
._cache
[pagenum
] = page_results
4065 start
% self
._pagesize
4066 if firstid
<= start
< nextfirstid
4070 ((end
- 1) % self
._pagesize
) + 1
4071 if (end
is not None and firstid
<= end
<= nextfirstid
)
4074 if startv
!= 0 or endv
is not None:
4075 page_results
= page_results
[startv
:endv
]
4076 res
.extend(page_results
)
4078 # A little optimization - if current page is not "full", ie. does
4079 # not contain page_size videos then we can assume that this page
4080 # is the last one - there are no more ids on further pages -
4081 # i.e. no need to query again.
4082 if len(page_results
) + startv
< self
._pagesize
:
4085 # If we got the whole page, but the next page is not interesting,
4086 # break out early as well
4087 if end
== nextfirstid
:
4092 class InAdvancePagedList(PagedList
):
4093 def __init__(self
, pagefunc
, pagecount
, pagesize
):
4094 self
._pagefunc
= pagefunc
4095 self
._pagecount
= pagecount
4096 self
._pagesize
= pagesize
4098 def getslice(self
, start
=0, end
=None):
4100 start_page
= start
// self
._pagesize
4102 self
._pagecount
if end
is None else (end
// self
._pagesize
+ 1))
4103 skip_elems
= start
- start_page
* self
._pagesize
4104 only_more
= None if end
is None else end
- start
4105 for pagenum
in range(start_page
, end_page
):
4106 page
= list(self
._pagefunc
(pagenum
))
4108 page
= page
[skip_elems
:]
4110 if only_more
is not None:
4111 if len(page
) < only_more
:
4112 only_more
-= len(page
)
4114 page
= page
[:only_more
]
4121 def uppercase_escape(s
):
4122 unicode_escape
= codecs
.getdecoder('unicode_escape')
4124 r
'\\U[0-9a-fA-F]{8}',
4125 lambda m
: unicode_escape(m
.group(0))[0],
4129 def lowercase_escape(s
):
4130 unicode_escape
= codecs
.getdecoder('unicode_escape')
4132 r
'\\u[0-9a-fA-F]{4}',
4133 lambda m
: unicode_escape(m
.group(0))[0],
4137 def escape_rfc3986(s
):
4138 """Escape non-ASCII characters as suggested by RFC 3986"""
4139 if sys
.version_info
< (3, 0) and isinstance(s
, compat_str
):
4140 s
= s
.encode('utf-8')
4141 return compat_urllib_parse
.quote(s
, b
"%/;:@&=+$,!~*'()?#[]")
4144 def escape_url(url
):
4145 """Escape URL as suggested by RFC 3986"""
4146 url_parsed
= compat_urllib_parse_urlparse(url
)
4147 return url_parsed
._replace
(
4148 netloc
=url_parsed
.netloc
.encode('idna').decode('ascii'),
4149 path
=escape_rfc3986(url_parsed
.path
),
4150 params
=escape_rfc3986(url_parsed
.params
),
4151 query
=escape_rfc3986(url_parsed
.query
),
4152 fragment
=escape_rfc3986(url_parsed
.fragment
)
4156 def read_batch_urls(batch_fd
):
4158 if not isinstance(url
, compat_str
):
4159 url
= url
.decode('utf-8', 'replace')
4160 BOM_UTF8
= ('\xef\xbb\xbf', '\ufeff')
4161 for bom
in BOM_UTF8
:
4162 if url
.startswith(bom
):
4163 url
= url
[len(bom
):]
4165 if not url
or url
.startswith(('#', ';', ']')):
4167 # "#" cannot be stripped out since it is part of the URI
4168 # However, it can be safely stipped out if follwing a whitespace
4169 return re
.split(r
'\s#', url
, 1)[0].rstrip()
4171 with contextlib
.closing(batch_fd
) as fd
:
4172 return [url
for url
in map(fixup
, fd
) if url
]
4175 def urlencode_postdata(*args
, **kargs
):
4176 return compat_urllib_parse_urlencode(*args
, **kargs
).encode('ascii')
4179 def update_url_query(url
, query
):
4182 parsed_url
= compat_urlparse
.urlparse(url
)
4183 qs
= compat_parse_qs(parsed_url
.query
)
4185 return compat_urlparse
.urlunparse(parsed_url
._replace
(
4186 query
=compat_urllib_parse_urlencode(qs
, True)))
4189 def update_Request(req
, url
=None, data
=None, headers
={}, query={}
):
4190 req_headers
= req
.headers
.copy()
4191 req_headers
.update(headers
)
4192 req_data
= data
or req
.data
4193 req_url
= update_url_query(url
or req
.get_full_url(), query
)
4194 req_get_method
= req
.get_method()
4195 if req_get_method
== 'HEAD':
4196 req_type
= HEADRequest
4197 elif req_get_method
== 'PUT':
4198 req_type
= PUTRequest
4200 req_type
= compat_urllib_request
.Request
4202 req_url
, data
=req_data
, headers
=req_headers
,
4203 origin_req_host
=req
.origin_req_host
, unverifiable
=req
.unverifiable
)
4204 if hasattr(req
, 'timeout'):
4205 new_req
.timeout
= req
.timeout
4209 def _multipart_encode_impl(data
, boundary
):
4210 content_type
= 'multipart/form-data; boundary=%s' % boundary
4213 for k
, v
in data
.items():
4214 out
+= b
'--' + boundary
.encode('ascii') + b
'\r\n'
4215 if isinstance(k
, compat_str
):
4216 k
= k
.encode('utf-8')
4217 if isinstance(v
, compat_str
):
4218 v
= v
.encode('utf-8')
4219 # RFC 2047 requires non-ASCII field names to be encoded, while RFC 7578
4220 # suggests sending UTF-8 directly. Firefox sends UTF-8, too
4221 content
= b
'Content-Disposition: form-data; name="' + k
+ b
'"\r\n\r\n' + v
+ b
'\r\n'
4222 if boundary
.encode('ascii') in content
:
4223 raise ValueError('Boundary overlaps with data')
4226 out
+= b
'--' + boundary
.encode('ascii') + b
'--\r\n'
4228 return out
, content_type
4231 def multipart_encode(data
, boundary
=None):
4233 Encode a dict to RFC 7578-compliant form-data
4236 A dict where keys and values can be either Unicode or bytes-like
4239 If specified a Unicode object, it's used as the boundary. Otherwise
4240 a random boundary is generated.
4242 Reference: https://tools.ietf.org/html/rfc7578
4244 has_specified_boundary
= boundary
is not None
4247 if boundary
is None:
4248 boundary
= '---------------' + str(random
.randrange(0x0fffffff, 0xffffffff))
4251 out
, content_type
= _multipart_encode_impl(data
, boundary
)
4254 if has_specified_boundary
:
4258 return out
, content_type
4261 def dict_get(d
, key_or_keys
, default
=None, skip_false_values
=True):
4262 if isinstance(key_or_keys
, (list, tuple)):
4263 for key
in key_or_keys
:
4264 if key
not in d
or d
[key
] is None or skip_false_values
and not d
[key
]:
4268 return d
.get(key_or_keys
, default
)
4271 def try_get(src
, getter
, expected_type
=None):
4272 if not isinstance(getter
, (list, tuple)):
4277 except (AttributeError, KeyError, TypeError, IndexError):
4280 if expected_type
is None or isinstance(v
, expected_type
):
4284 def merge_dicts(*dicts
):
4286 for a_dict
in dicts
:
4287 for k
, v
in a_dict
.items():
4291 or (isinstance(v
, compat_str
) and v
4292 and isinstance(merged
[k
], compat_str
)
4293 and not merged
[k
])):
4298 def encode_compat_str(string
, encoding
=preferredencoding(), errors
='strict'):
4299 return string
if isinstance(string
, compat_str
) else compat_str(string
, encoding
, errors
)
4311 TV_PARENTAL_GUIDELINES
= {
4321 def parse_age_limit(s
):
4323 return s
if 0 <= s
<= 21 else None
4324 if not isinstance(s
, compat_basestring
):
4326 m
= re
.match(r
'^(?P<age>\d{1,2})\+?$', s
)
4328 return int(m
.group('age'))
4331 return US_RATINGS
[s
]
4332 m
= re
.match(r
'^TV[_-]?(%s)$' % '|'.join(k
[3:] for k
in TV_PARENTAL_GUIDELINES
), s
)
4334 return TV_PARENTAL_GUIDELINES
['TV-' + m
.group(1)]
4338 def strip_jsonp(code
):
4341 (?:window\.)?(?P<func_name>[a-zA-Z0-9_.$]*)
4342 (?:\s*&&\s*(?P=func_name))?
4343 \s*\(\s*(?P<callback_data>.*)\);?
4344 \s*?(?://[^\n]*)*$''',
4345 r
'\g<callback_data>', code
)
4348 def js_to_json(code
, vars={}):
4349 # vars is a dict of var, val pairs to substitute
4350 COMMENT_RE
= r
'/\*(?:(?!\*/).)*?\*/|//[^\n]*'
4351 SKIP_RE
= r
'\s*(?:{comment})?\s*'.format(comment
=COMMENT_RE
)
4353 (r
'(?s)^(0[xX][0-9a-fA-F]+){skip}:?$'.format(skip
=SKIP_RE
), 16),
4354 (r
'(?s)^(0+[0-7]+){skip}:?$'.format(skip
=SKIP_RE
), 8),
4359 if v
in ('true', 'false', 'null'):
4361 elif v
.startswith('/*') or v
.startswith('//') or v
.startswith('!') or v
== ',':
4364 if v
[0] in ("'", '"'):
4365 v
= re
.sub(r
'(?s)\\.|"', lambda m
: {
4370 }.get(m
.group(0), m
.group(0)), v
[1:-1])
4372 for regex
, base
in INTEGER_TABLE
:
4373 im
= re
.match(regex
, v
)
4375 i
= int(im
.group(1), base
)
4376 return '"%d":' % i
if v
.endswith(':') else '%d' % i
4383 return re
.sub(r
'''(?sx)
4384 "(?:[^"\\]*(?:\\\\|\\['"nurtbfx/\n]))*[^"\\]*"|
4385 '(?:[^'\\]*(?:\\\\|\\['"nurtbfx/\n]))*[^'\\]*'|
4386 {comment}|,(?={skip}[\]}}])|
4387 (?:(?<![0-9])[eE]|[a-df-zA-DF-Z_])[.a-zA-Z_0-9]*|
4388 \b(?:0[xX][0-9a-fA-F]+|0+[0-7]+)(?:{skip}:)?|
4391 '''.format(comment
=COMMENT_RE
, skip
=SKIP_RE
), fix_kv
, code
)
4394 def qualities(quality_ids
):
4395 """ Get a numeric quality value out of a list of possible values """
4398 return quality_ids
.index(qid
)
4405 'default': '%(title)s [%(id)s].%(ext)s',
4406 'chapter': '%(title)s - %(section_number)03d %(section_title)s [%(id)s].%(ext)s',
4412 'description': 'description',
4413 'annotation': 'annotations.xml',
4414 'infojson': 'info.json',
4415 'pl_thumbnail': None,
4416 'pl_description': 'description',
4417 'pl_infojson': 'info.json',
4420 # As of [1] format syntax is:
4421 # %[mapping_key][conversion_flags][minimum_width][.precision][length_modifier]type
4422 # 1. https://docs.python.org/2/library/stdtypes.html#string-formatting
4423 STR_FORMAT_RE
= r
'''(?x)
4426 (?P<has_key>\((?P<key>{0})\))? # mapping key
4428 (?:[#0\-+ ]+)? # conversion flags (optional)
4429 (?:\d+)? # minimum field width (optional)
4430 (?:\.\d+)? # precision (optional)
4431 [hlL]? # length modifier (optional)
4432 [diouxXeEfFgGcrs] # conversion type
4437 def limit_length(s
, length
):
4438 """ Add ellipses to overly long strings """
4443 return s
[:length
- len(ELLIPSES
)] + ELLIPSES
4447 def version_tuple(v
):
4448 return tuple(int(e
) for e
in re
.split(r
'[-.]', v
))
4451 def is_outdated_version(version
, limit
, assume_new
=True):
4453 return not assume_new
4455 return version_tuple(version
) < version_tuple(limit
)
4457 return not assume_new
4460 def ytdl_is_updateable():
4461 """ Returns if yt-dlp can be updated with -U """
4464 from zipimport
import zipimporter
4466 return isinstance(globals().get('__loader__'), zipimporter
) or hasattr(sys
, 'frozen')
4469 def args_to_str(args
):
4470 # Get a short string representation for a subprocess command
4471 return ' '.join(compat_shlex_quote(a
) for a
in args
)
4474 def error_to_compat_str(err
):
4476 # On python 2 error byte string must be decoded with proper
4477 # encoding rather than ascii
4478 if sys
.version_info
[0] < 3:
4479 err_str
= err_str
.decode(preferredencoding())
4483 def mimetype2ext(mt
):
4489 # Per RFC 3003, audio/mpeg can be .mp1, .mp2 or .mp3. Here use .mp3 as
4490 # it's the most popular one
4491 'audio/mpeg': 'mp3',
4492 'audio/x-wav': 'wav',
4497 _
, _
, res
= mt
.rpartition('/')
4498 res
= res
.split(';')[0].strip().lower()
4502 'smptett+xml': 'tt',
4506 'x-mp4-fragmented': 'mp4',
4507 'x-ms-sami': 'sami',
4510 'x-mpegurl': 'm3u8',
4511 'vnd.apple.mpegurl': 'm3u8',
4515 'vnd.ms-sstr+xml': 'ism',
4522 def parse_codecs(codecs_str
):
4523 # http://tools.ietf.org/html/rfc6381
4526 split_codecs
= list(filter(None, map(
4527 lambda str: str.strip(), codecs_str
.strip().strip(',').split(','))))
4528 vcodec
, acodec
= None, None
4529 for full_codec
in split_codecs
:
4530 codec
= full_codec
.split('.')[0]
4531 if codec
in ('avc1', 'avc2', 'avc3', 'avc4', 'vp9', 'vp8', 'hev1', 'hev2', 'h263', 'h264', 'mp4v', 'hvc1', 'av01', 'theora'):
4534 elif codec
in ('mp4a', 'opus', 'vorbis', 'mp3', 'aac', 'ac-3', 'ec-3', 'eac3', 'dtsc', 'dtse', 'dtsh', 'dtsl'):
4538 write_string('WARNING: Unknown codec %s\n' % full_codec
, sys
.stderr
)
4539 if not vcodec
and not acodec
:
4540 if len(split_codecs
) == 2:
4542 'vcodec': split_codecs
[0],
4543 'acodec': split_codecs
[1],
4547 'vcodec': vcodec
or 'none',
4548 'acodec': acodec
or 'none',
4553 def urlhandle_detect_ext(url_handle
):
4554 getheader
= url_handle
.headers
.get
4556 cd
= getheader('Content-Disposition')
4558 m
= re
.match(r
'attachment;\s*filename="(?P<filename>[^"]+)"', cd
)
4560 e
= determine_ext(m
.group('filename'), default_ext
=None)
4564 return mimetype2ext(getheader('Content-Type'))
4567 def encode_data_uri(data
, mime_type
):
4568 return 'data:%s;base64,%s' % (mime_type
, base64
.b64encode(data
).decode('ascii'))
4571 def age_restricted(content_limit
, age_limit
):
4572 """ Returns True iff the content should be blocked """
4574 if age_limit
is None: # No limit set
4576 if content_limit
is None:
4577 return False # Content available for everyone
4578 return age_limit
< content_limit
4581 def is_html(first_bytes
):
4582 """ Detect whether a file contains HTML by examining its first bytes. """
4585 (b
'\xef\xbb\xbf', 'utf-8'),
4586 (b
'\x00\x00\xfe\xff', 'utf-32-be'),
4587 (b
'\xff\xfe\x00\x00', 'utf-32-le'),
4588 (b
'\xff\xfe', 'utf-16-le'),
4589 (b
'\xfe\xff', 'utf-16-be'),
4591 for bom
, enc
in BOMS
:
4592 if first_bytes
.startswith(bom
):
4593 s
= first_bytes
[len(bom
):].decode(enc
, 'replace')
4596 s
= first_bytes
.decode('utf-8', 'replace')
4598 return re
.match(r
'^\s*<', s
)
4601 def determine_protocol(info_dict
):
4602 protocol
= info_dict
.get('protocol')
4603 if protocol
is not None:
4606 url
= info_dict
['url']
4607 if url
.startswith('rtmp'):
4609 elif url
.startswith('mms'):
4611 elif url
.startswith('rtsp'):
4614 ext
= determine_ext(url
)
4620 return compat_urllib_parse_urlparse(url
).scheme
4623 def render_table(header_row
, data
, delim
=False, extraGap
=0, hideEmpty
=False):
4624 """ Render a list of rows, each as a list of values """
4626 def get_max_lens(table
):
4627 return [max(len(compat_str(v
)) for v
in col
) for col
in zip(*table
)]
4629 def filter_using_list(row
, filterArray
):
4630 return [col
for (take
, col
) in zip(filterArray
, row
) if take
]
4633 max_lens
= get_max_lens(data
)
4634 header_row
= filter_using_list(header_row
, max_lens
)
4635 data
= [filter_using_list(row
, max_lens
) for row
in data
]
4637 table
= [header_row
] + data
4638 max_lens
= get_max_lens(table
)
4640 table
= [header_row
] + [['-' * ml
for ml
in max_lens
]] + data
4641 format_str
= ' '.join('%-' + compat_str(ml
+ extraGap
) + 's' for ml
in max_lens
[:-1]) + ' %s'
4642 return '\n'.join(format_str
% tuple(row
) for row
in table
)
4645 def _match_one(filter_part
, dct
):
4646 COMPARISON_OPERATORS
= {
4654 operator_rex
= re
.compile(r
'''(?x)\s*
4656 \s*(?P<op>%s)(?P<none_inclusive>\s*\?)?\s*
4658 (?P<intval>[0-9.]+(?:[kKmMgGtTpPeEzZyY]i?[Bb]?)?)|
4659 (?P<quote>["\'])(?P
<quotedstrval
>(?
:\\.|
(?
!(?P
=quote
)|
\\).)+?
)(?P
=quote
)|
4660 (?P
<strval
>(?
![0-9.])[a
-z0
-9A
-Z
]*)
4663 ''' % '|'.join(map(re.escape, COMPARISON_OPERATORS.keys())))
4664 m = operator_rex.search(filter_part)
4666 op = COMPARISON_OPERATORS[m.group('op')]
4667 actual_value = dct.get(m.group('key'))
4668 if (m.group('quotedstrval') is not None
4669 or m.group('strval') is not None
4670 # If the original field is a string and matching comparisonvalue is
4671 # a number we should respect the origin of the original field
4672 # and process comparison value as a string (see
4673 # https://github.com/ytdl-org/youtube-dl/issues/11082).
4674 or actual_value is not None and m.group('intval') is not None
4675 and isinstance(actual_value, compat_str)):
4676 if m.group('op') not in ('=', '!='):
4678 'Operator %s does not support string values!' % m.group('op'))
4679 comparison_value = m.group('quotedstrval') or m.group('strval') or m.group('intval')
4680 quote = m.group('quote')
4681 if quote is not None:
4682 comparison_value = comparison_value.replace(r'\%s' % quote, quote)
4685 comparison_value = int(m.group('intval'))
4687 comparison_value = parse_filesize(m.group('intval'))
4688 if comparison_value is None:
4689 comparison_value = parse_filesize(m.group('intval') + 'B')
4690 if comparison_value is None:
4692 'Invalid integer value %r in filter part %r' % (
4693 m.group('intval'), filter_part))
4694 if actual_value is None:
4695 return m.group('none_inclusive')
4696 return op(actual_value, comparison_value)
4699 '': lambda v: (v is True) if isinstance(v, bool) else (v is not None),
4700 '!': lambda v: (v is False) if isinstance(v, bool) else (v is None),
4702 operator_rex = re.compile(r'''(?x
)\s
*
4703 (?P
<op
>%s)\s
*(?P
<key
>[a
-z_
]+)
4705 ''' % '|'.join(map(re.escape, UNARY_OPERATORS.keys())))
4706 m = operator_rex.search(filter_part)
4708 op = UNARY_OPERATORS[m.group('op')]
4709 actual_value = dct.get(m.group('key'))
4710 return op(actual_value)
4712 raise ValueError('Invalid filter part %r' % filter_part)
4715 def match_str(filter_str, dct):
4716 """ Filter a dictionary with a simple string syntax. Returns True (=passes filter) or false """
4719 _match_one(filter_part, dct) for filter_part in filter_str.split('&'))
4722 def match_filter_func(filter_str):
4723 def _match_func(info_dict):
4724 if match_str(filter_str, info_dict):
4727 video_title = info_dict.get('title', info_dict.get('id', 'video'))
4728 return '%s does not pass filter %s, skipping ..' % (video_title, filter_str)
4732 def parse_dfxp_time_expr(time_expr):
4736 mobj = re.match(r'^(?P<time_offset>\d+(?:\.\d+)?)s?$', time_expr)
4738 return float(mobj.group('time_offset'))
4740 mobj = re.match(r'^(\d+):(\d\d):(\d\d(?:(?:\.|:)\d+)?)$', time_expr)
4742 return 3600 * int(mobj.group(1)) + 60 * int(mobj.group(2)) + float(mobj.group(3).replace(':', '.'))
4745 def srt_subtitles_timecode(seconds):
4746 return '%02d:%02d:%02d,%03d' % (seconds / 3600, (seconds % 3600) / 60, seconds % 60, (seconds % 1) * 1000)
4749 def dfxp2srt(dfxp_data):
4751 @param dfxp_data A
bytes-like
object containing DFXP data
4752 @returns A
unicode object containing converted SRT data
4754 LEGACY_NAMESPACES = (
4755 (b'http://www.w3.org/ns/ttml', [
4756 b'http://www.w3.org/2004/11/ttaf1',
4757 b'http://www.w3.org/2006/04/ttaf1',
4758 b'http://www.w3.org/2006/10/ttaf1',
4760 (b'http://www.w3.org/ns/ttml#styling', [
4761 b'http://www.w3.org/ns/ttml#style',
4765 SUPPORTED_STYLING = [
4774 _x = functools.partial(xpath_with_ns, ns_map={
4775 'xml': 'http://www.w3.org/XML/1998/namespace',
4776 'ttml': 'http://www.w3.org/ns/ttml',
4777 'tts': 'http://www.w3.org/ns/ttml#styling',
4783 class TTMLPElementParser(object):
4785 _unclosed_elements = []
4786 _applied_styles = []
4788 def start(self, tag, attrib):
4789 if tag in (_x('ttml:br'), 'br'):
4792 unclosed_elements = []
4794 element_style_id = attrib.get('style')
4796 style.update(default_style)
4797 if element_style_id:
4798 style.update(styles.get(element_style_id, {}))
4799 for prop in SUPPORTED_STYLING:
4800 prop_val = attrib.get(_x('tts:' + prop))
4802 style[prop] = prop_val
4805 for k, v in sorted(style.items()):
4806 if self._applied_styles and self._applied_styles[-1].get(k) == v:
4809 font += ' color="%s"' % v
4810 elif k == 'fontSize':
4811 font += ' size="%s"' % v
4812 elif k == 'fontFamily':
4813 font += ' face="%s"' % v
4814 elif k == 'fontWeight' and v == 'bold':
4816 unclosed_elements.append('b')
4817 elif k == 'fontStyle' and v == 'italic':
4819 unclosed_elements.append('i')
4820 elif k == 'textDecoration' and v == 'underline':
4822 unclosed_elements.append('u')
4824 self._out += '<font' + font + '>'
4825 unclosed_elements.append('font')
4827 if self._applied_styles:
4828 applied_style.update(self._applied_styles[-1])
4829 applied_style.update(style)
4830 self._applied_styles.append(applied_style)
4831 self._unclosed_elements.append(unclosed_elements)
4834 if tag not in (_x('ttml:br'), 'br'):
4835 unclosed_elements = self._unclosed_elements.pop()
4836 for element in reversed(unclosed_elements):
4837 self._out += '</%s>' % element
4838 if unclosed_elements and self._applied_styles:
4839 self._applied_styles.pop()
4841 def data(self, data):
4845 return self._out.strip()
4847 def parse_node(node):
4848 target = TTMLPElementParser()
4849 parser = xml.etree.ElementTree.XMLParser(target=target)
4850 parser.feed(xml.etree.ElementTree.tostring(node))
4851 return parser.close()
4853 for k, v in LEGACY_NAMESPACES:
4855 dfxp_data = dfxp_data.replace(ns, k)
4857 dfxp = compat_etree_fromstring(dfxp_data)
4859 paras = dfxp.findall(_x('.//ttml:p')) or dfxp.findall('.//p')
4862 raise ValueError('Invalid dfxp/TTML subtitle')
4866 for style in dfxp.findall(_x('.//ttml:style')):
4867 style_id = style.get('id') or style.get(_x('xml:id'))
4870 parent_style_id = style.get('style')
4872 if parent_style_id not in styles:
4875 styles[style_id] = styles[parent_style_id].copy()
4876 for prop in SUPPORTED_STYLING:
4877 prop_val = style.get(_x('tts:' + prop))
4879 styles.setdefault(style_id, {})[prop] = prop_val
4885 for p in ('body', 'div'):
4886 ele = xpath_element(dfxp, [_x('.//ttml:' + p), './/' + p])
4889 style = styles.get(ele.get('style'))
4892 default_style.update(style)
4894 for para, index in zip(paras, itertools.count(1)):
4895 begin_time = parse_dfxp_time_expr(para.attrib.get('begin'))
4896 end_time = parse_dfxp_time_expr(para.attrib.get('end'))
4897 dur = parse_dfxp_time_expr(para.attrib.get('dur'))
4898 if begin_time is None:
4903 end_time = begin_time + dur
4904 out.append('%d\n%s --> %s\n%s\n\n' % (
4906 srt_subtitles_timecode(begin_time),
4907 srt_subtitles_timecode(end_time),
4913 def cli_option(params, command_option, param):
4914 param = params.get(param)
4916 param = compat_str(param)
4917 return [command_option, param] if param is not None else []
4920 def cli_bool_option(params, command_option, param, true_value='true', false_value='false', separator=None):
4921 param = params.get(param)
4924 assert isinstance(param, bool)
4926 return [command_option + separator + (true_value if param else false_value)]
4927 return [command_option, true_value if param else false_value]
4930 def cli_valueless_option(params, command_option, param, expected_value=True):
4931 param = params.get(param)
4932 return [command_option] if param == expected_value else []
4935 def cli_configuration_args(argdict, keys, default=[], use_compat=True):
4936 if isinstance(argdict, (list, tuple)): # for backward compatibility
4943 assert isinstance(argdict, dict)
4945 assert isinstance(keys, (list, tuple))
4946 for key_list in keys:
4947 if isinstance(key_list, compat_str):
4948 key_list = (key_list,)
4949 arg_list = list(filter(
4950 lambda x: x is not None,
4951 [argdict.get(key.lower()) for key in key_list]))
4953 return [arg for args in arg_list for arg in args]
4957 class ISO639Utils(object):
4958 # See http://www.loc.gov/standards/iso639-2/ISO-639-2_utf-8.txt
5017 'iw': 'heb', # Replaced by he in 1989 revision
5027 'in': 'ind', # Replaced by id in 1989 revision
5142 'ji': 'yid', # Replaced by yi in 1989 revision
5150 def short2long(cls, code):
5151 """Convert language code from ISO 639-1 to ISO 639-2/T"""
5152 return cls._lang_map.get(code[:2])
5155 def long2short(cls, code):
5156 """Convert language code from ISO 639-2/T to ISO 639-1"""
5157 for short_name, long_name in cls._lang_map.items():
5158 if long_name == code:
5162 class ISO3166Utils(object):
5163 # From http://data.okfn.org/data/core/country-list
5165 'AF': 'Afghanistan',
5166 'AX': 'Åland Islands',
5169 'AS': 'American Samoa',
5174 'AG': 'Antigua and Barbuda',
5191 'BO': 'Bolivia, Plurinational State of',
5192 'BQ': 'Bonaire, Sint Eustatius and Saba',
5193 'BA': 'Bosnia and Herzegovina',
5195 'BV': 'Bouvet Island',
5197 'IO': 'British Indian Ocean Territory',
5198 'BN': 'Brunei Darussalam',
5200 'BF': 'Burkina Faso',
5206 'KY': 'Cayman Islands',
5207 'CF': 'Central African Republic',
5211 'CX': 'Christmas Island',
5212 'CC': 'Cocos (Keeling) Islands',
5216 'CD': 'Congo, the Democratic Republic of the',
5217 'CK': 'Cook Islands',
5219 'CI': 'Côte d\'Ivoire',
5224 'CZ': 'Czech Republic',
5228 'DO': 'Dominican Republic',
5231 'SV': 'El Salvador',
5232 'GQ': 'Equatorial Guinea',
5236 'FK': 'Falkland Islands (Malvinas)',
5237 'FO': 'Faroe Islands',
5241 'GF': 'French Guiana',
5242 'PF': 'French Polynesia',
5243 'TF': 'French Southern Territories',
5258 'GW': 'Guinea-Bissau',
5261 'HM': 'Heard Island and McDonald Islands',
5262 'VA': 'Holy See (Vatican City State)',
5269 'IR': 'Iran, Islamic Republic of',
5272 'IM': 'Isle of Man',
5282 'KP': 'Korea, Democratic People\'s Republic of',
5283 'KR': 'Korea, Republic of',
5286 'LA': 'Lao People\'s Democratic Republic',
5292 'LI': 'Liechtenstein',
5296 'MK': 'Macedonia, the Former Yugoslav Republic of',
5303 'MH': 'Marshall Islands',
5309 'FM': 'Micronesia, Federated States of',
5310 'MD': 'Moldova, Republic of',
5321 'NL': 'Netherlands',
5322 'NC': 'New Caledonia',
5323 'NZ': 'New Zealand',
5328 'NF': 'Norfolk Island',
5329 'MP': 'Northern Mariana Islands',
5334 'PS': 'Palestine, State of',
5336 'PG': 'Papua New Guinea',
5339 'PH': 'Philippines',
5343 'PR': 'Puerto Rico',
5347 'RU': 'Russian Federation',
5349 'BL': 'Saint Barthélemy',
5350 'SH': 'Saint Helena, Ascension and Tristan da Cunha',
5351 'KN': 'Saint Kitts and Nevis',
5352 'LC': 'Saint Lucia',
5353 'MF': 'Saint Martin (French part)',
5354 'PM': 'Saint Pierre and Miquelon',
5355 'VC': 'Saint Vincent and the Grenadines',
5358 'ST': 'Sao Tome and Principe',
5359 'SA': 'Saudi Arabia',
5363 'SL': 'Sierra Leone',
5365 'SX': 'Sint Maarten (Dutch part)',
5368 'SB': 'Solomon Islands',
5370 'ZA': 'South Africa',
5371 'GS': 'South Georgia and the South Sandwich Islands',
5372 'SS': 'South Sudan',
5377 'SJ': 'Svalbard and Jan Mayen',
5380 'CH': 'Switzerland',
5381 'SY': 'Syrian Arab Republic',
5382 'TW': 'Taiwan, Province of China',
5384 'TZ': 'Tanzania, United Republic of',
5386 'TL': 'Timor-Leste',
5390 'TT': 'Trinidad and Tobago',
5393 'TM': 'Turkmenistan',
5394 'TC': 'Turks and Caicos Islands',
5398 'AE': 'United Arab Emirates',
5399 'GB': 'United Kingdom',
5400 'US': 'United States',
5401 'UM': 'United States Minor Outlying Islands',
5405 'VE': 'Venezuela, Bolivarian Republic of',
5407 'VG': 'Virgin Islands, British',
5408 'VI': 'Virgin Islands, U.S.',
5409 'WF': 'Wallis and Futuna',
5410 'EH': 'Western Sahara',
5417 def short2full(cls, code):
5418 """Convert an ISO 3166-2 country code to the corresponding full name"""
5419 return cls._country_map.get(code.upper())
5422 class GeoUtils(object):
5423 # Major IPv4 address blocks per country
5425 'AD': '46.172.224.0/19',
5426 'AE': '94.200.0.0/13',
5427 'AF': '149.54.0.0/17',
5428 'AG': '209.59.64.0/18',
5429 'AI': '204.14.248.0/21',
5430 'AL': '46.99.0.0/16',
5431 'AM': '46.70.0.0/15',
5432 'AO': '105.168.0.0/13',
5433 'AP': '182.50.184.0/21',
5434 'AQ': '23.154.160.0/24',
5435 'AR': '181.0.0.0/12',
5436 'AS': '202.70.112.0/20',
5437 'AT': '77.116.0.0/14',
5438 'AU': '1.128.0.0/11',
5439 'AW': '181.41.0.0/18',
5440 'AX': '185.217.4.0/22',
5441 'AZ': '5.197.0.0/16',
5442 'BA': '31.176.128.0/17',
5443 'BB': '65.48.128.0/17',
5444 'BD': '114.130.0.0/16',
5446 'BF': '102.178.0.0/15',
5447 'BG': '95.42.0.0/15',
5448 'BH': '37.131.0.0/17',
5449 'BI': '154.117.192.0/18',
5450 'BJ': '137.255.0.0/16',
5451 'BL': '185.212.72.0/23',
5452 'BM': '196.12.64.0/18',
5453 'BN': '156.31.0.0/16',
5454 'BO': '161.56.0.0/16',
5455 'BQ': '161.0.80.0/20',
5456 'BR': '191.128.0.0/12',
5457 'BS': '24.51.64.0/18',
5458 'BT': '119.2.96.0/19',
5459 'BW': '168.167.0.0/16',
5460 'BY': '178.120.0.0/13',
5461 'BZ': '179.42.192.0/18',
5462 'CA': '99.224.0.0/11',
5463 'CD': '41.243.0.0/16',
5464 'CF': '197.242.176.0/21',
5465 'CG': '160.113.0.0/16',
5466 'CH': '85.0.0.0/13',
5467 'CI': '102.136.0.0/14',
5468 'CK': '202.65.32.0/19',
5469 'CL': '152.172.0.0/14',
5470 'CM': '102.244.0.0/14',
5471 'CN': '36.128.0.0/10',
5472 'CO': '181.240.0.0/12',
5473 'CR': '201.192.0.0/12',
5474 'CU': '152.206.0.0/15',
5475 'CV': '165.90.96.0/19',
5476 'CW': '190.88.128.0/17',
5477 'CY': '31.153.0.0/16',
5478 'CZ': '88.100.0.0/14',
5480 'DJ': '197.241.0.0/17',
5481 'DK': '87.48.0.0/12',
5482 'DM': '192.243.48.0/20',
5483 'DO': '152.166.0.0/15',
5484 'DZ': '41.96.0.0/12',
5485 'EC': '186.68.0.0/15',
5486 'EE': '90.190.0.0/15',
5487 'EG': '156.160.0.0/11',
5488 'ER': '196.200.96.0/20',
5489 'ES': '88.0.0.0/11',
5490 'ET': '196.188.0.0/14',
5491 'EU': '2.16.0.0/13',
5492 'FI': '91.152.0.0/13',
5493 'FJ': '144.120.0.0/16',
5494 'FK': '80.73.208.0/21',
5495 'FM': '119.252.112.0/20',
5496 'FO': '88.85.32.0/19',
5498 'GA': '41.158.0.0/15',
5500 'GD': '74.122.88.0/21',
5501 'GE': '31.146.0.0/16',
5502 'GF': '161.22.64.0/18',
5503 'GG': '62.68.160.0/19',
5504 'GH': '154.160.0.0/12',
5505 'GI': '95.164.0.0/16',
5506 'GL': '88.83.0.0/19',
5507 'GM': '160.182.0.0/15',
5508 'GN': '197.149.192.0/18',
5509 'GP': '104.250.0.0/19',
5510 'GQ': '105.235.224.0/20',
5511 'GR': '94.64.0.0/13',
5512 'GT': '168.234.0.0/16',
5513 'GU': '168.123.0.0/16',
5514 'GW': '197.214.80.0/20',
5515 'GY': '181.41.64.0/18',
5516 'HK': '113.252.0.0/14',
5517 'HN': '181.210.0.0/16',
5518 'HR': '93.136.0.0/13',
5519 'HT': '148.102.128.0/17',
5520 'HU': '84.0.0.0/14',
5521 'ID': '39.192.0.0/10',
5522 'IE': '87.32.0.0/12',
5523 'IL': '79.176.0.0/13',
5524 'IM': '5.62.80.0/20',
5525 'IN': '117.192.0.0/10',
5526 'IO': '203.83.48.0/21',
5527 'IQ': '37.236.0.0/14',
5528 'IR': '2.176.0.0/12',
5529 'IS': '82.221.0.0/16',
5530 'IT': '79.0.0.0/10',
5531 'JE': '87.244.64.0/18',
5532 'JM': '72.27.0.0/17',
5533 'JO': '176.29.0.0/16',
5534 'JP': '133.0.0.0/8',
5535 'KE': '105.48.0.0/12',
5536 'KG': '158.181.128.0/17',
5537 'KH': '36.37.128.0/17',
5538 'KI': '103.25.140.0/22',
5539 'KM': '197.255.224.0/20',
5540 'KN': '198.167.192.0/19',
5541 'KP': '175.45.176.0/22',
5542 'KR': '175.192.0.0/10',
5543 'KW': '37.36.0.0/14',
5544 'KY': '64.96.0.0/15',
5545 'KZ': '2.72.0.0/13',
5546 'LA': '115.84.64.0/18',
5547 'LB': '178.135.0.0/16',
5548 'LC': '24.92.144.0/20',
5549 'LI': '82.117.0.0/19',
5550 'LK': '112.134.0.0/15',
5551 'LR': '102.183.0.0/16',
5552 'LS': '129.232.0.0/17',
5553 'LT': '78.56.0.0/13',
5554 'LU': '188.42.0.0/16',
5555 'LV': '46.109.0.0/16',
5556 'LY': '41.252.0.0/14',
5557 'MA': '105.128.0.0/11',
5558 'MC': '88.209.64.0/18',
5559 'MD': '37.246.0.0/16',
5560 'ME': '178.175.0.0/17',
5561 'MF': '74.112.232.0/21',
5562 'MG': '154.126.0.0/17',
5563 'MH': '117.103.88.0/21',
5564 'MK': '77.28.0.0/15',
5565 'ML': '154.118.128.0/18',
5566 'MM': '37.111.0.0/17',
5567 'MN': '49.0.128.0/17',
5568 'MO': '60.246.0.0/16',
5569 'MP': '202.88.64.0/20',
5570 'MQ': '109.203.224.0/19',
5571 'MR': '41.188.64.0/18',
5572 'MS': '208.90.112.0/22',
5573 'MT': '46.11.0.0/16',
5574 'MU': '105.16.0.0/12',
5575 'MV': '27.114.128.0/18',
5576 'MW': '102.70.0.0/15',
5577 'MX': '187.192.0.0/11',
5578 'MY': '175.136.0.0/13',
5579 'MZ': '197.218.0.0/15',
5580 'NA': '41.182.0.0/16',
5581 'NC': '101.101.0.0/18',
5582 'NE': '197.214.0.0/18',
5583 'NF': '203.17.240.0/22',
5584 'NG': '105.112.0.0/12',
5585 'NI': '186.76.0.0/15',
5586 'NL': '145.96.0.0/11',
5587 'NO': '84.208.0.0/13',
5588 'NP': '36.252.0.0/15',
5589 'NR': '203.98.224.0/19',
5590 'NU': '49.156.48.0/22',
5591 'NZ': '49.224.0.0/14',
5592 'OM': '5.36.0.0/15',
5593 'PA': '186.72.0.0/15',
5594 'PE': '186.160.0.0/14',
5595 'PF': '123.50.64.0/18',
5596 'PG': '124.240.192.0/19',
5597 'PH': '49.144.0.0/13',
5598 'PK': '39.32.0.0/11',
5599 'PL': '83.0.0.0/11',
5600 'PM': '70.36.0.0/20',
5601 'PR': '66.50.0.0/16',
5602 'PS': '188.161.0.0/16',
5603 'PT': '85.240.0.0/13',
5604 'PW': '202.124.224.0/20',
5605 'PY': '181.120.0.0/14',
5606 'QA': '37.210.0.0/15',
5607 'RE': '102.35.0.0/16',
5608 'RO': '79.112.0.0/13',
5609 'RS': '93.86.0.0/15',
5610 'RU': '5.136.0.0/13',
5611 'RW': '41.186.0.0/16',
5612 'SA': '188.48.0.0/13',
5613 'SB': '202.1.160.0/19',
5614 'SC': '154.192.0.0/11',
5615 'SD': '102.120.0.0/13',
5616 'SE': '78.64.0.0/12',
5617 'SG': '8.128.0.0/10',
5618 'SI': '188.196.0.0/14',
5619 'SK': '78.98.0.0/15',
5620 'SL': '102.143.0.0/17',
5621 'SM': '89.186.32.0/19',
5622 'SN': '41.82.0.0/15',
5623 'SO': '154.115.192.0/18',
5624 'SR': '186.179.128.0/17',
5625 'SS': '105.235.208.0/21',
5626 'ST': '197.159.160.0/19',
5627 'SV': '168.243.0.0/16',
5628 'SX': '190.102.0.0/20',
5630 'SZ': '41.84.224.0/19',
5631 'TC': '65.255.48.0/20',
5632 'TD': '154.68.128.0/19',
5633 'TG': '196.168.0.0/14',
5634 'TH': '171.96.0.0/13',
5635 'TJ': '85.9.128.0/18',
5636 'TK': '27.96.24.0/21',
5637 'TL': '180.189.160.0/20',
5638 'TM': '95.85.96.0/19',
5639 'TN': '197.0.0.0/11',
5640 'TO': '175.176.144.0/21',
5641 'TR': '78.160.0.0/11',
5642 'TT': '186.44.0.0/15',
5643 'TV': '202.2.96.0/19',
5644 'TW': '120.96.0.0/11',
5645 'TZ': '156.156.0.0/14',
5646 'UA': '37.52.0.0/14',
5647 'UG': '102.80.0.0/13',
5649 'UY': '167.56.0.0/13',
5650 'UZ': '84.54.64.0/18',
5651 'VA': '212.77.0.0/19',
5652 'VC': '207.191.240.0/21',
5653 'VE': '186.88.0.0/13',
5654 'VG': '66.81.192.0/20',
5655 'VI': '146.226.0.0/16',
5656 'VN': '14.160.0.0/11',
5657 'VU': '202.80.32.0/20',
5658 'WF': '117.20.32.0/21',
5659 'WS': '202.4.32.0/19',
5660 'YE': '134.35.0.0/16',
5661 'YT': '41.242.116.0/22',
5662 'ZA': '41.0.0.0/11',
5663 'ZM': '102.144.0.0/13',
5664 'ZW': '102.177.192.0/18',
5668 def random_ipv4(cls, code_or_block):
5669 if len(code_or_block) == 2:
5670 block = cls._country_ip_map.get(code_or_block.upper())
5674 block = code_or_block
5675 addr, preflen = block.split('/')
5676 addr_min = compat_struct_unpack('!L', socket.inet_aton(addr))[0]
5677 addr_max = addr_min | (0xffffffff >> int(preflen))
5678 return compat_str(socket.inet_ntoa(
5679 compat_struct_pack('!L', random.randint(addr_min, addr_max))))
5682 class PerRequestProxyHandler(compat_urllib_request.ProxyHandler):
5683 def __init__(self, proxies=None):
5684 # Set default handlers
5685 for type in ('http', 'https'):
5686 setattr(self, '%s_open' % type,
5687 lambda r, proxy='__noproxy__', type=type, meth=self.proxy_open:
5688 meth(r, proxy, type))
5689 compat_urllib_request.ProxyHandler.__init__(self, proxies)
5691 def proxy_open(self, req, proxy, type):
5692 req_proxy = req.headers.get('Ytdl-request-proxy')
5693 if req_proxy is not None:
5695 del req.headers['Ytdl-request-proxy']
5697 if proxy == '__noproxy__':
5698 return None # No Proxy
5699 if compat_urlparse.urlparse(proxy).scheme.lower() in ('socks', 'socks4', 'socks4a', 'socks5'):
5700 req.add_header('Ytdl-socks-proxy', proxy)
5701 # yt-dlp's http/https handlers do wrapping the socket with socks
5703 return compat_urllib_request.ProxyHandler.proxy_open(
5704 self, req, proxy, type)
5707 # Both long_to_bytes and bytes_to_long are adapted from PyCrypto, which is
5708 # released into Public Domain
5709 # https://github.com/dlitz/pycrypto/blob/master/lib/Crypto/Util/number.py#L387
5711 def long_to_bytes(n, blocksize=0):
5712 """long_to_bytes(n:long, blocksize:int) : string
5713 Convert a long integer to a byte string.
5715 If optional blocksize is given and greater than zero, pad the front of the
5716 byte string with binary zeros so that the length is a multiple of
5719 # after much testing, this algorithm was deemed to be the fastest
5723 s = compat_struct_pack('>I', n & 0xffffffff) + s
5725 # strip off leading zeros
5726 for i in range(len(s)):
5727 if s[i] != b'\000'[0]:
5730 # only happens when n == 0
5734 # add back some pad bytes. this could be done more efficiently w.r.t. the
5735 # de-padding being done above, but sigh...
5736 if blocksize > 0 and len(s) % blocksize:
5737 s = (blocksize - len(s) % blocksize) * b'\000' + s
5741 def bytes_to_long(s):
5742 """bytes_to_long(string) : long
5743 Convert a byte string to a long integer.
5745 This is (essentially) the inverse of long_to_bytes().
5750 extra = (4 - length % 4)
5751 s = b'\000' * extra + s
5752 length = length + extra
5753 for i in range(0, length, 4):
5754 acc = (acc << 32) + compat_struct_unpack('>I', s[i:i + 4])[0]
5758 def ohdave_rsa_encrypt(data, exponent, modulus):
5760 Implement OHDave
's RSA algorithm. See http://www.ohdave.com/rsa/
5763 data: data to encrypt, bytes-like object
5764 exponent, modulus: parameter e and N of RSA algorithm, both integer
5765 Output: hex string of encrypted data
5767 Limitation: supports one block encryption only
5770 payload = int(binascii.hexlify(data[::-1]), 16)
5771 encrypted = pow(payload, exponent, modulus)
5772 return '%x' % encrypted
5775 def pkcs1pad(data, length):
5777 Padding input data with PKCS#1 scheme
5779 @param {int[]} data input data
5780 @param {int} length target length
5781 @returns {int[]} padded data
5783 if len(data) > length - 11:
5784 raise ValueError('Input data too
long for PKCS
#1 padding')
5786 pseudo_random
= [random
.randint(0, 254) for _
in range(length
- len(data
) - 3)]
5787 return [0, 2] + pseudo_random
+ [0] + data
5790 def encode_base_n(num
, n
, table
=None):
5791 FULL_TABLE
= '0123456789abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ'
5793 table
= FULL_TABLE
[:n
]
5796 raise ValueError('base %d exceeds table length %d' % (n
, len(table
)))
5803 ret
= table
[num
% n
] + ret
5808 def decode_packed_codes(code
):
5809 mobj
= re
.search(PACKED_CODES_RE
, code
)
5810 obfuscated_code
, base
, count
, symbols
= mobj
.groups()
5813 symbols
= symbols
.split('|')
5818 base_n_count
= encode_base_n(count
, base
)
5819 symbol_table
[base_n_count
] = symbols
[count
] or base_n_count
5822 r
'\b(\w+)\b', lambda mobj
: symbol_table
[mobj
.group(0)],
5826 def caesar(s
, alphabet
, shift
):
5831 alphabet
[(alphabet
.index(c
) + shift
) % l
] if c
in alphabet
else c
5836 return caesar(s
, r
'''!"#$%&'()*+,-./0123456789:;<=>?@ABCDEFGHIJKLMNOPQRSTUVWXYZ[\]^_`abcdefghijklmnopqrstuvwxyz{|}~''', 47)
5839 def parse_m3u8_attributes(attrib
):
5841 for (key
, val
) in re
.findall(r
'(?P<key>[A-Z0-9-]+)=(?P<val>"[^"]+"|[^",]+)(?:,|$)', attrib
):
5842 if val
.startswith('"'):
5848 def urshift(val
, n
):
5849 return val
>> n
if val
>= 0 else (val
+ 0x100000000) >> n
5852 # Based on png2str() written by @gdkchan and improved by @yokrysty
5853 # Originally posted at https://github.com/ytdl-org/youtube-dl/issues/9706
5854 def decode_png(png_data
):
5855 # Reference: https://www.w3.org/TR/PNG/
5856 header
= png_data
[8:]
5858 if png_data
[:8] != b
'\x89PNG\x0d\x0a\x1a\x0a' or header
[4:8] != b
'IHDR':
5859 raise IOError('Not a valid PNG file.')
5861 int_map
= {1: '>B', 2: '>H', 4: '>I'}
5862 unpack_integer
= lambda x
: compat_struct_unpack(int_map
[len(x
)], x
)[0]
5867 length
= unpack_integer(header
[:4])
5870 chunk_type
= header
[:4]
5873 chunk_data
= header
[:length
]
5874 header
= header
[length
:]
5876 header
= header
[4:] # Skip CRC
5884 ihdr
= chunks
[0]['data']
5886 width
= unpack_integer(ihdr
[:4])
5887 height
= unpack_integer(ihdr
[4:8])
5891 for chunk
in chunks
:
5892 if chunk
['type'] == b
'IDAT':
5893 idat
+= chunk
['data']
5896 raise IOError('Unable to read PNG data.')
5898 decompressed_data
= bytearray(zlib
.decompress(idat
))
5903 def _get_pixel(idx
):
5908 for y
in range(height
):
5909 basePos
= y
* (1 + stride
)
5910 filter_type
= decompressed_data
[basePos
]
5914 pixels
.append(current_row
)
5916 for x
in range(stride
):
5917 color
= decompressed_data
[1 + basePos
+ x
]
5918 basex
= y
* stride
+ x
5923 left
= _get_pixel(basex
- 3)
5925 up
= _get_pixel(basex
- stride
)
5927 if filter_type
== 1: # Sub
5928 color
= (color
+ left
) & 0xff
5929 elif filter_type
== 2: # Up
5930 color
= (color
+ up
) & 0xff
5931 elif filter_type
== 3: # Average
5932 color
= (color
+ ((left
+ up
) >> 1)) & 0xff
5933 elif filter_type
== 4: # Paeth
5939 c
= _get_pixel(basex
- stride
- 3)
5947 if pa
<= pb
and pa
<= pc
:
5948 color
= (color
+ a
) & 0xff
5950 color
= (color
+ b
) & 0xff
5952 color
= (color
+ c
) & 0xff
5954 current_row
.append(color
)
5956 return width
, height
, pixels
5959 def write_xattr(path
, key
, value
):
5960 # This mess below finds the best xattr tool for the job
5962 # try the pyxattr module...
5965 if hasattr(xattr
, 'set'): # pyxattr
5966 # Unicode arguments are not supported in python-pyxattr until
5968 # See https://github.com/ytdl-org/youtube-dl/issues/5498
5969 pyxattr_required_version
= '0.5.0'
5970 if version_tuple(xattr
.__version
__) < version_tuple(pyxattr_required_version
):
5971 # TODO: fallback to CLI tools
5972 raise XAttrUnavailableError(
5973 'python-pyxattr is detected but is too old. '
5974 'yt-dlp requires %s or above while your version is %s. '
5975 'Falling back to other xattr implementations' % (
5976 pyxattr_required_version
, xattr
.__version
__))
5978 setxattr
= xattr
.set
5980 setxattr
= xattr
.setxattr
5983 setxattr(path
, key
, value
)
5984 except EnvironmentError as e
:
5985 raise XAttrMetadataError(e
.errno
, e
.strerror
)
5988 if compat_os_name
== 'nt':
5989 # Write xattrs to NTFS Alternate Data Streams:
5990 # http://en.wikipedia.org/wiki/NTFS#Alternate_data_streams_.28ADS.29
5991 assert ':' not in key
5992 assert os
.path
.exists(path
)
5994 ads_fn
= path
+ ':' + key
5996 with open(ads_fn
, 'wb') as f
:
5998 except EnvironmentError as e
:
5999 raise XAttrMetadataError(e
.errno
, e
.strerror
)
6001 user_has_setfattr
= check_executable('setfattr', ['--version'])
6002 user_has_xattr
= check_executable('xattr', ['-h'])
6004 if user_has_setfattr
or user_has_xattr
:
6006 value
= value
.decode('utf-8')
6007 if user_has_setfattr
:
6008 executable
= 'setfattr'
6009 opts
= ['-n', key
, '-v', value
]
6010 elif user_has_xattr
:
6011 executable
= 'xattr'
6012 opts
= ['-w', key
, value
]
6014 cmd
= ([encodeFilename(executable
, True)]
6015 + [encodeArgument(o
) for o
in opts
]
6016 + [encodeFilename(path
, True)])
6019 p
= subprocess
.Popen(
6020 cmd
, stdout
=subprocess
.PIPE
, stderr
=subprocess
.PIPE
, stdin
=subprocess
.PIPE
)
6021 except EnvironmentError as e
:
6022 raise XAttrMetadataError(e
.errno
, e
.strerror
)
6023 stdout
, stderr
= process_communicate_or_kill(p
)
6024 stderr
= stderr
.decode('utf-8', 'replace')
6025 if p
.returncode
!= 0:
6026 raise XAttrMetadataError(p
.returncode
, stderr
)
6029 # On Unix, and can't find pyxattr, setfattr, or xattr.
6030 if sys
.platform
.startswith('linux'):
6031 raise XAttrUnavailableError(
6032 "Couldn't find a tool to set the xattrs. "
6033 "Install either the python 'pyxattr' or 'xattr' "
6034 "modules, or the GNU 'attr' package "
6035 "(which contains the 'setfattr' tool).")
6037 raise XAttrUnavailableError(
6038 "Couldn't find a tool to set the xattrs. "
6039 "Install either the python 'xattr' module, "
6040 "or the 'xattr' binary.")
6043 def random_birthday(year_field
, month_field
, day_field
):
6044 start_date
= datetime
.date(1950, 1, 1)
6045 end_date
= datetime
.date(1995, 12, 31)
6046 offset
= random
.randint(0, (end_date
- start_date
).days
)
6047 random_date
= start_date
+ datetime
.timedelta(offset
)
6049 year_field
: str(random_date
.year
),
6050 month_field
: str(random_date
.month
),
6051 day_field
: str(random_date
.day
),
6055 # Templates for internet shortcut files, which are plain text files.
6056 DOT_URL_LINK_TEMPLATE
= '''
6061 DOT_WEBLOC_LINK_TEMPLATE
= '''
6062 <?xml version="1.0" encoding="UTF-8"?>
6063 <!DOCTYPE plist PUBLIC "-//Apple//DTD PLIST 1.0//EN" "http://www.apple.com/DTDs/PropertyList-1.0.dtd">
6064 <plist version="1.0">
6067 \t<string>%(url)s</string>
6072 DOT_DESKTOP_LINK_TEMPLATE
= '''
6082 def iri_to_uri(iri
):
6084 Converts an IRI (Internationalized Resource Identifier, allowing Unicode characters) to a URI (Uniform Resource Identifier, ASCII-only).
6086 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.
6089 iri_parts
= compat_urllib_parse_urlparse(iri
)
6091 if '[' in iri_parts
.netloc
:
6092 raise ValueError('IPv6 URIs are not, yet, supported.')
6093 # Querying `.netloc`, when there's only one bracket, also raises a ValueError.
6095 # 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.
6098 if iri_parts
.username
:
6099 net_location
+= compat_urllib_parse_quote(iri_parts
.username
, safe
=r
"!$%&'()*+,~")
6100 if iri_parts
.password
is not None:
6101 net_location
+= ':' + compat_urllib_parse_quote(iri_parts
.password
, safe
=r
"!$%&'()*+,~")
6104 net_location
+= iri_parts
.hostname
.encode('idna').decode('utf-8') # Punycode for Unicode hostnames.
6105 # The 'idna' encoding produces ASCII text.
6106 if iri_parts
.port
is not None and iri_parts
.port
!= 80:
6107 net_location
+= ':' + str(iri_parts
.port
)
6109 return compat_urllib_parse_urlunparse(
6113 compat_urllib_parse_quote_plus(iri_parts
.path
, safe
=r
"!$%&'()*+,/:;=@|~"),
6115 # Unsure about the `safe` argument, since this is a legacy way of handling parameters.
6116 compat_urllib_parse_quote_plus(iri_parts
.params
, safe
=r
"!$%&'()*+,/:;=@|~"),
6118 # Not totally sure about the `safe` argument, since the source does not explicitly mention the query URI component.
6119 compat_urllib_parse_quote_plus(iri_parts
.query
, safe
=r
"!$%&'()*+,/:;=?@{|}~"),
6121 compat_urllib_parse_quote_plus(iri_parts
.fragment
, safe
=r
"!#$%&'()*+,/:;=?@{|}~")))
6123 # Source for `safe` arguments: https://url.spec.whatwg.org/#percent-encoded-bytes.
6126 def to_high_limit_path(path
):
6127 if sys
.platform
in ['win32', 'cygwin']:
6128 # Work around MAX_PATH limitation on Windows. The maximum allowed length for the individual path segments may still be quite limited.
6129 return r
'\\?\ '.rstrip() + os
.path
.abspath(path
)
6134 def format_field(obj
, field
, template
='%s', ignore
=(None, ''), default
='', func
=None):
6135 val
= obj
.get(field
, default
)
6136 if func
and val
not in ignore
:
6138 return template
% val
if val
not in ignore
else default
6141 def clean_podcast_url(url
):
6142 return re
.sub(r
'''(?x)
6146 media\.blubrry\.com| # https://create.blubrry.com/resources/podcast-media-download-statistics/getting-started/
6149 (?:dts|www)\.podtrac\.com/(?:pts/)?redirect\.[0-9a-z]{3,4}| # http://analytics.podtrac.com/how-to-measure
6152 cn\.co| # https://podcorn.com/analytics-prefix/
6153 st\.fm # https://podsights.com/docs/
6158 _HEX_TABLE
= '0123456789abcdef'
6161 def random_uuidv4():
6162 return re
.sub(r
'[xy]', lambda x
: _HEX_TABLE
[random
.randint(0, 15)], 'xxxxxxxx-xxxx-4xxx-yxxx-xxxxxxxxxxxx')
6165 def make_dir(path
, to_screen
=None):
6167 dn
= os
.path
.dirname(path
)
6168 if dn
and not os
.path
.exists(dn
):
6171 except (OSError, IOError) as err
:
6172 if callable(to_screen
) is not None:
6173 to_screen('unable to create directory ' + error_to_compat_str(err
))
6177 def get_executable_path():
6178 from zipimport
import zipimporter
6179 if hasattr(sys
, 'frozen'): # Running from PyInstaller
6180 path
= os
.path
.dirname(sys
.executable
)
6181 elif isinstance(globals().get('__loader__'), zipimporter
): # Running from ZIP
6182 path
= os
.path
.join(os
.path
.dirname(__file__
), '../..')
6184 path
= os
.path
.join(os
.path
.dirname(__file__
), '..')
6185 return os
.path
.abspath(path
)
6188 def load_plugins(name
, suffix
, namespace
):
6189 plugin_info
= [None]
6192 plugin_info
= imp
.find_module(
6193 name
, [os
.path
.join(get_executable_path(), 'ytdlp_plugins')])
6194 plugins
= imp
.load_module(name
, *plugin_info
)
6195 for name
in dir(plugins
):
6196 if name
in namespace
:
6198 if not name
.endswith(suffix
):
6200 klass
= getattr(plugins
, name
)
6201 classes
.append(klass
)
6202 namespace
[name
] = klass
6206 if plugin_info
[0] is not None:
6207 plugin_info
[0].close()
6211 def traverse_obj(obj
, keys
, *, casesense
=True, is_user_input
=False, traverse_string
=False):
6212 ''' Traverse nested list/dict/tuple
6213 @param casesense Whether to consider dictionary keys as case sensitive
6214 @param is_user_input Whether the keys are generated from user input. If True,
6215 strings are converted to int/slice if necessary
6216 @param traverse_string Whether to traverse inside strings. If True, any
6217 non-compatible object will also be converted into a string
6219 keys
= list(keys
)[::-1]
6222 if isinstance(obj
, dict):
6223 assert isinstance(key
, compat_str
)
6225 obj
= {k.lower(): v for k, v in obj.items()}
6230 key
= (int_or_none(key
) if ':' not in key
6231 else slice(*map(int_or_none
, key
.split(':'))))
6232 if not isinstance(obj
, (list, tuple)):
6234 obj
= compat_str(obj
)
6237 assert isinstance(key
, (int, slice))
6238 obj
= try_get(obj
, lambda x
: x
[key
])
6242 def traverse_dict(dictn
, keys
, casesense
=True):
6243 ''' For backward compatibility. Do not use '''
6244 return traverse_obj(dictn
, keys
, casesense
=casesense
,
6245 is_user_input
=True, traverse_string
=True)
projects / yt-dlp.git / blob