#!/usr/bin/env python
-# -*- coding: utf-8 -*-
+# coding: utf-8
from __future__ import unicode_literals
+import base64
+import binascii
import calendar
import codecs
import contextlib
import errno
import functools
import gzip
-import itertools
import io
+import itertools
import json
import locale
import math
import pipes
import platform
import re
-import ssl
import socket
-import struct
+import ssl
import subprocess
import sys
import tempfile
import zlib
from .compat import (
+ compat_HTMLParser,
compat_basestring,
compat_chr,
+ compat_etree_fromstring,
compat_html_entities,
+ compat_html_entities_html5,
compat_http_client,
compat_kwargs,
+ compat_os_name,
compat_parse_qs,
+ compat_shlex_quote,
compat_socket_create_connection,
compat_str,
+ compat_struct_pack,
+ compat_struct_unpack,
compat_urllib_error,
compat_urllib_parse,
+ compat_urllib_parse_urlencode,
compat_urllib_parse_urlparse,
+ compat_urllib_parse_unquote_plus,
compat_urllib_request,
compat_urlparse,
- shlex_quote,
+ compat_xpath,
+)
+
+from .socks import (
+ ProxyType,
+ sockssocket,
)
+def register_socks_protocols():
+ # "Register" SOCKS protocols
+ # In Python < 2.6.5, urlsplit() suffers from bug https://bugs.python.org/issue7904
+ # URLs with protocols not in urlparse.uses_netloc are not handled correctly
+ for scheme in ('socks', 'socks4', 'socks4a', 'socks5'):
+ if scheme not in compat_urlparse.uses_netloc:
+ compat_urlparse.uses_netloc.append(scheme)
+
+
# This is not clearly defined otherwise
compiled_regex_type = type(re.compile(''))
std_headers = {
- 'User-Agent': 'Mozilla/5.0 (X11; Linux x86_64; rv:10.0) Gecko/20150101 Firefox/20.0 (Chrome)',
+ 'User-Agent': 'Mozilla/5.0 (X11; Linux x86_64; rv:10.0) Gecko/20150101 Firefox/47.0 (Chrome)',
'Accept-Charset': 'ISO-8859-1,utf-8;q=0.7,*;q=0.7',
'Accept': 'text/html,application/xhtml+xml,application/xml;q=0.9,*/*;q=0.8',
'Accept-Encoding': 'gzip, deflate',
'January', 'February', 'March', 'April', 'May', 'June',
'July', 'August', 'September', 'October', 'November', 'December']
+MONTH_NAMES = {
+ 'en': ENGLISH_MONTH_NAMES,
+ 'fr': [
+ 'janvier', 'février', 'mars', 'avril', 'mai', 'juin',
+ 'juillet', 'août', 'septembre', 'octobre', 'novembre', 'décembre'],
+}
+
+KNOWN_EXTENSIONS = (
+ 'mp4', 'm4a', 'm4p', 'm4b', 'm4r', 'm4v', 'aac',
+ 'flv', 'f4v', 'f4a', 'f4b',
+ 'webm', 'ogg', 'ogv', 'oga', 'ogx', 'spx', 'opus',
+ 'mkv', 'mka', 'mk3d',
+ 'avi', 'divx',
+ 'mov',
+ 'asf', 'wmv', 'wma',
+ '3gp', '3g2',
+ 'mp3',
+ 'flac',
+ 'ape',
+ 'wav',
+ 'f4f', 'f4m', 'm3u8', 'smil')
+
+# needed for sanitizing filenames in restricted mode
+ACCENT_CHARS = dict(zip('ÂÃÄÀÁÅÆÇÈÉÊËÌÍÎÏÐÑÒÓÔÕÖŐØŒÙÚÛÜŰÝÞßàáâãäåæçèéêëìíîïðñòóôõöőøœùúûüűýþÿ',
+ itertools.chain('AAAAAA', ['AE'], 'CEEEEIIIIDNOOOOOOO', ['OE'], 'UUUUUYP', ['ss'],
+ 'aaaaaa', ['ae'], 'ceeeeiiiionooooooo', ['oe'], 'uuuuuypy')))
+
+DATE_FORMATS = (
+ '%d %B %Y',
+ '%d %b %Y',
+ '%B %d %Y',
+ '%b %d %Y',
+ '%b %dst %Y %I:%M',
+ '%b %dnd %Y %I:%M',
+ '%b %dth %Y %I:%M',
+ '%Y %m %d',
+ '%Y-%m-%d',
+ '%Y/%m/%d',
+ '%Y/%m/%d %H:%M',
+ '%Y/%m/%d %H:%M:%S',
+ '%Y-%m-%d %H:%M:%S',
+ '%Y-%m-%d %H:%M:%S.%f',
+ '%d.%m.%Y %H:%M',
+ '%d.%m.%Y %H.%M',
+ '%Y-%m-%dT%H:%M:%SZ',
+ '%Y-%m-%dT%H:%M:%S.%fZ',
+ '%Y-%m-%dT%H:%M:%S.%f0Z',
+ '%Y-%m-%dT%H:%M:%S',
+ '%Y-%m-%dT%H:%M:%S.%f',
+ '%Y-%m-%dT%H:%M',
+ '%b %d %Y at %H:%M',
+ '%b %d %Y at %H:%M:%S',
+)
+
+DATE_FORMATS_DAY_FIRST = list(DATE_FORMATS)
+DATE_FORMATS_DAY_FIRST.extend([
+ '%d-%m-%Y',
+ '%d.%m.%Y',
+ '%d.%m.%y',
+ '%d/%m/%Y',
+ '%d/%m/%y',
+ '%d/%m/%Y %H:%M:%S',
+])
+
+DATE_FORMATS_MONTH_FIRST = list(DATE_FORMATS)
+DATE_FORMATS_MONTH_FIRST.extend([
+ '%m-%d-%Y',
+ '%m.%d.%Y',
+ '%m/%d/%Y',
+ '%m/%d/%y',
+ '%m/%d/%Y %H:%M:%S',
+])
+
+PACKED_CODES_RE = r"}\('(.+)',(\d+),(\d+),'([^']+)'\.split\('\|'\)"
+
def preferredencoding():
"""Get preferred encoding.
if sys.version_info >= (2, 7):
- def find_xpath_attr(node, xpath, key, val):
+ def find_xpath_attr(node, xpath, key, val=None):
""" Find the xpath xpath[@key=val] """
- assert re.match(r'^[a-zA-Z-]+$', key)
- assert re.match(r'^[a-zA-Z0-9@\s:._-]*$', val)
- expr = xpath + "[@%s='%s']" % (key, val)
+ assert re.match(r'^[a-zA-Z_-]+$', key)
+ expr = xpath + ('[@%s]' % key if val is None else "[@%s='%s']" % (key, val))
return node.find(expr)
else:
- def find_xpath_attr(node, xpath, key, val):
- # Here comes the crazy part: In 2.6, if the xpath is a unicode,
- # .//node does not match if a node is a direct child of . !
- if isinstance(xpath, compat_str):
- xpath = xpath.encode('ascii')
-
- for f in node.findall(xpath):
- if f.attrib.get(key) == val:
+ def find_xpath_attr(node, xpath, key, val=None):
+ for f in node.findall(compat_xpath(xpath)):
+ if key not in f.attrib:
+ continue
+ if val is None or f.attrib.get(key) == val:
return f
return None
return '/'.join(replaced)
-def xpath_text(node, xpath, name=None, fatal=False, default=NO_DEFAULT):
- if sys.version_info < (2, 7): # Crazy 2.6
- xpath = xpath.encode('ascii')
+def xpath_element(node, xpath, name=None, fatal=False, default=NO_DEFAULT):
+ def _find_xpath(xpath):
+ return node.find(compat_xpath(xpath))
- n = node.find(xpath)
- if n is None or n.text is None:
+ if isinstance(xpath, (str, compat_str)):
+ n = _find_xpath(xpath)
+ else:
+ for xp in xpath:
+ n = _find_xpath(xp)
+ if n is not None:
+ break
+
+ if n is None:
if default is not NO_DEFAULT:
return default
elif fatal:
raise ExtractorError('Could not find XML element %s' % name)
else:
return None
+ return n
+
+
+def xpath_text(node, xpath, name=None, fatal=False, default=NO_DEFAULT):
+ n = xpath_element(node, xpath, name, fatal=fatal, default=default)
+ if n is None or n == default:
+ return n
+ if n.text is None:
+ if default is not NO_DEFAULT:
+ return default
+ elif fatal:
+ name = xpath if name is None else name
+ raise ExtractorError('Could not find XML element\'s text %s' % name)
+ else:
+ return None
return n.text
+def xpath_attr(node, xpath, key, name=None, fatal=False, default=NO_DEFAULT):
+ n = find_xpath_attr(node, xpath, key)
+ if n is None:
+ if default is not NO_DEFAULT:
+ return default
+ elif fatal:
+ name = '%s[@%s]' % (xpath, key) if name is None else name
+ raise ExtractorError('Could not find XML attribute %s' % name)
+ else:
+ return None
+ return n.attrib[key]
+
+
def get_element_by_id(id, html):
"""Return the content of the tag with the specified ID in the passed HTML document"""
- return get_element_by_attribute("id", id, html)
+ return get_element_by_attribute('id', id, html)
+
+def get_element_by_class(class_name, html):
+ return get_element_by_attribute(
+ 'class', r'[^\'"]*\b%s\b[^\'"]*' % re.escape(class_name),
+ html, escape_value=False)
-def get_element_by_attribute(attribute, value, html):
+
+def get_element_by_attribute(attribute, value, html, escape_value=True):
"""Return the content of the tag with the specified attribute in the passed HTML document"""
+ value = re.escape(value) if escape_value else value
+
m = re.search(r'''(?xs)
<([a-zA-Z0-9:._-]+)
- (?:\s+[a-zA-Z0-9:._-]+(?:=[a-zA-Z0-9:._-]+|="[^"]+"|='[^']+'))*?
+ (?:\s+[a-zA-Z0-9:._-]+(?:=[a-zA-Z0-9:._-]*|="[^"]*"|='[^']*'))*?
\s+%s=['"]?%s['"]?
- (?:\s+[a-zA-Z0-9:._-]+(?:=[a-zA-Z0-9:._-]+|="[^"]+"|='[^']+'))*?
+ (?:\s+[a-zA-Z0-9:._-]+(?:=[a-zA-Z0-9:._-]*|="[^"]*"|='[^']*'))*?
\s*>
(?P<content>.*?)
</\1>
- ''' % (re.escape(attribute), re.escape(value)), html)
+ ''' % (re.escape(attribute), value), html)
if not m:
return None
return unescapeHTML(res)
+class HTMLAttributeParser(compat_HTMLParser):
+ """Trivial HTML parser to gather the attributes for a single element"""
+ def __init__(self):
+ self.attrs = {}
+ compat_HTMLParser.__init__(self)
+
+ def handle_starttag(self, tag, attrs):
+ self.attrs = dict(attrs)
+
+
+def extract_attributes(html_element):
+ """Given a string for an HTML element such as
+ <el
+ a="foo" B="bar" c="&98;az" d=boz
+ empty= noval entity="&"
+ sq='"' dq="'"
+ >
+ Decode and return a dictionary of attributes.
+ {
+ 'a': 'foo', 'b': 'bar', c: 'baz', d: 'boz',
+ 'empty': '', 'noval': None, 'entity': '&',
+ 'sq': '"', 'dq': '\''
+ }.
+ NB HTMLParser is stricter in Python 2.6 & 3.2 than in later versions,
+ but the cases in the unit test will work for all of 2.6, 2.7, 3.2-3.5.
+ """
+ parser = HTMLAttributeParser()
+ parser.feed(html_element)
+ parser.close()
+ return parser.attrs
+
+
def clean_html(html):
"""Clean an HTML snippet into a readable string"""
Set is_id if this is not an arbitrary string, but an ID that should be kept if possible
"""
def replace_insane(char):
+ if restricted and char in ACCENT_CHARS:
+ return ACCENT_CHARS[char]
if char == '?' or ord(char) < 32 or ord(char) == 127:
return ''
elif char == '"':
if drive_or_unc:
norm_path.pop(0)
sanitized_path = [
- path_part if path_part in ['.', '..'] else re.sub('(?:[/<>:"\\|\\\\?\\*]|\.$)', '#', path_part)
+ path_part if path_part in ['.', '..'] else re.sub('(?:[/<>:"\\|\\\\?\\*]|[\s.]$)', '#', path_part)
for path_part in norm_path]
if drive_or_unc:
sanitized_path.insert(0, drive_or_unc + os.path.sep)
return os.path.join(*sanitized_path)
+# Prepend protocol-less URLs with `http:` scheme in order to mitigate the number of
+# unwanted failures due to missing protocol
+def sanitize_url(url):
+ return 'http:%s' % url if url.startswith('//') else url
+
+
+def sanitized_Request(url, *args, **kwargs):
+ return compat_urllib_request.Request(sanitize_url(url), *args, **kwargs)
+
+
def orderedSet(iterable):
""" Remove all duplicates from the input iterable """
res = []
return res
-def _htmlentity_transform(entity):
+def _htmlentity_transform(entity_with_semicolon):
"""Transforms an HTML entity to a character."""
+ entity = entity_with_semicolon[:-1]
+
# Known non-numeric HTML entity
if entity in compat_html_entities.name2codepoint:
return compat_chr(compat_html_entities.name2codepoint[entity])
+ # TODO: HTML5 allows entities without a semicolon. For example,
+ # 'Éric' should be decoded as 'Éric'.
+ if entity_with_semicolon in compat_html_entities_html5:
+ return compat_html_entities_html5[entity_with_semicolon]
+
mobj = re.match(r'#(x[0-9a-fA-F]+|[0-9]+)', entity)
if mobj is not None:
numstr = mobj.group(1)
numstr = '0%s' % numstr
else:
base = 10
- return compat_chr(int(numstr, base))
+ # See https://github.com/rg3/youtube-dl/issues/7518
+ try:
+ return compat_chr(int(numstr, base))
+ except ValueError:
+ pass
# Unknown entity in name, return its literal representation
- return ('&%s;' % entity)
+ return '&%s;' % entity
def unescapeHTML(s):
assert type(s) == compat_str
return re.sub(
- r'&([^;]+);', lambda m: _htmlentity_transform(m.group(1)), s)
+ r'&([^;]+;)', lambda m: _htmlentity_transform(m.group(1)), s)
def get_subprocess_encoding():
if not for_subprocess and sys.platform == 'win32' and sys.getwindowsversion()[0] >= 5:
return s
+ # Jython assumes filenames are Unicode strings though reported as Python 2.x compatible
+ if sys.platform.startswith('java'):
+ return s
+
return s.encode(get_subprocess_encoding(), 'ignore')
download is too small for what the server announced first, indicating
the connection was probably interrupted.
"""
- # Both in bytes
- downloaded = None
- expected = None
def __init__(self, downloaded, expected):
+ # Both in bytes
self.downloaded = downloaded
self.expected = expected
+class XAttrMetadataError(Exception):
+ def __init__(self, code=None, msg='Unknown error'):
+ super(XAttrMetadataError, self).__init__(msg)
+ self.code = code
+ self.msg = msg
+
+ # Parsing code and msg
+ if (self.code in (errno.ENOSPC, errno.EDQUOT) or
+ 'No space left' in self.msg or 'Disk quota excedded' in self.msg):
+ self.reason = 'NO_SPACE'
+ elif self.code == errno.E2BIG or 'Argument list too long' in self.msg:
+ self.reason = 'VALUE_TOO_LONG'
+ else:
+ self.reason = 'NOT_SUPPORTED'
+
+
+class XAttrUnavailableError(Exception):
+ pass
+
+
def _create_http_connection(ydl_handler, http_class, is_https, *args, **kwargs):
+ # Working around python 2 bug (see http://bugs.python.org/issue17849) by limiting
+ # expected HTTP responses to meet HTTP/1.0 or later (see also
+ # https://github.com/rg3/youtube-dl/issues/6727)
+ if sys.version_info < (3, 0):
+ kwargs[b'strict'] = True
hc = http_class(*args, **kwargs)
source_address = ydl_handler._params.get('source_address')
if source_address is not None:
return hc
+def handle_youtubedl_headers(headers):
+ filtered_headers = headers
+
+ if 'Youtubedl-no-compression' in filtered_headers:
+ filtered_headers = dict((k, v) for k, v in filtered_headers.items() if k.lower() != 'accept-encoding')
+ del filtered_headers['Youtubedl-no-compression']
+
+ return filtered_headers
+
+
class YoutubeDLHandler(compat_urllib_request.HTTPHandler):
"""Handler for HTTP requests and responses.
the standard headers to every HTTP request and handles gzipped and
deflated responses from web servers. If compression is to be avoided in
a particular request, the original request in the program code only has
- to include the HTTP header "Youtubedl-No-Compression", which will be
+ to include the HTTP header "Youtubedl-no-compression", which will be
removed before making the real request.
Part of this code was copied from:
self._params = params
def http_open(self, req):
+ conn_class = compat_http_client.HTTPConnection
+
+ socks_proxy = req.headers.get('Ytdl-socks-proxy')
+ if socks_proxy:
+ conn_class = make_socks_conn_class(conn_class, socks_proxy)
+ del req.headers['Ytdl-socks-proxy']
+
return self.do_open(functools.partial(
- _create_http_connection, self, compat_http_client.HTTPConnection, False),
+ _create_http_connection, self, conn_class, False),
req)
@staticmethod
return ret
def http_request(self, req):
+ # According to RFC 3986, URLs can not contain non-ASCII characters, however this is not
+ # always respected by websites, some tend to give out URLs with non percent-encoded
+ # non-ASCII characters (see telemb.py, ard.py [#3412])
+ # urllib chokes on URLs with non-ASCII characters (see http://bugs.python.org/issue3991)
+ # To work around aforementioned issue we will replace request's original URL with
+ # percent-encoded one
+ # Since redirects are also affected (e.g. http://www.southpark.de/alle-episoden/s18e09)
+ # the code of this workaround has been moved here from YoutubeDL.urlopen()
+ url = req.get_full_url()
+ url_escaped = escape_url(url)
+
+ # Substitute URL if any change after escaping
+ if url != url_escaped:
+ req = update_Request(req, url=url_escaped)
+
for h, v in std_headers.items():
# Capitalize is needed because of Python bug 2275: http://bugs.python.org/issue2275
# The dict keys are capitalized because of this bug by urllib
if h.capitalize() not in req.headers:
req.add_header(h, v)
- if 'Youtubedl-no-compression' in req.headers:
- if 'Accept-encoding' in req.headers:
- del req.headers['Accept-encoding']
- del req.headers['Youtubedl-no-compression']
+
+ req.headers = handle_youtubedl_headers(req.headers)
if sys.version_info < (2, 7) and '#' in req.get_full_url():
# Python 2.6 is brain-dead when it comes to fragments
raise original_ioerror
resp = self.addinfourl_wrapper(uncompressed, old_resp.headers, old_resp.url, old_resp.code)
resp.msg = old_resp.msg
+ del resp.headers['Content-encoding']
# deflate
if resp.headers.get('Content-encoding', '') == 'deflate':
gz = io.BytesIO(self.deflate(resp.read()))
resp = self.addinfourl_wrapper(gz, old_resp.headers, old_resp.url, old_resp.code)
resp.msg = old_resp.msg
+ del resp.headers['Content-encoding']
+ # Percent-encode redirect URL of Location HTTP header to satisfy RFC 3986 (see
+ # https://github.com/rg3/youtube-dl/issues/6457).
+ if 300 <= resp.code < 400:
+ location = resp.headers.get('Location')
+ if location:
+ # As of RFC 2616 default charset is iso-8859-1 that is respected by python 3
+ if sys.version_info >= (3, 0):
+ location = location.encode('iso-8859-1').decode('utf-8')
+ else:
+ location = location.decode('utf-8')
+ location_escaped = escape_url(location)
+ if location != location_escaped:
+ del resp.headers['Location']
+ if sys.version_info < (3, 0):
+ location_escaped = location_escaped.encode('utf-8')
+ resp.headers['Location'] = location_escaped
return resp
https_request = http_request
https_response = http_response
+def make_socks_conn_class(base_class, socks_proxy):
+ assert issubclass(base_class, (
+ compat_http_client.HTTPConnection, compat_http_client.HTTPSConnection))
+
+ url_components = compat_urlparse.urlparse(socks_proxy)
+ if url_components.scheme.lower() == 'socks5':
+ socks_type = ProxyType.SOCKS5
+ elif url_components.scheme.lower() in ('socks', 'socks4'):
+ socks_type = ProxyType.SOCKS4
+ elif url_components.scheme.lower() == 'socks4a':
+ socks_type = ProxyType.SOCKS4A
+
+ def unquote_if_non_empty(s):
+ if not s:
+ return s
+ return compat_urllib_parse_unquote_plus(s)
+
+ proxy_args = (
+ socks_type,
+ url_components.hostname, url_components.port or 1080,
+ True, # Remote DNS
+ unquote_if_non_empty(url_components.username),
+ unquote_if_non_empty(url_components.password),
+ )
+
+ class SocksConnection(base_class):
+ def connect(self):
+ self.sock = sockssocket()
+ self.sock.setproxy(*proxy_args)
+ if type(self.timeout) in (int, float):
+ self.sock.settimeout(self.timeout)
+ self.sock.connect((self.host, self.port))
+
+ if isinstance(self, compat_http_client.HTTPSConnection):
+ if hasattr(self, '_context'): # Python > 2.6
+ self.sock = self._context.wrap_socket(
+ self.sock, server_hostname=self.host)
+ else:
+ self.sock = ssl.wrap_socket(self.sock)
+
+ return SocksConnection
+
+
class YoutubeDLHTTPSHandler(compat_urllib_request.HTTPSHandler):
def __init__(self, params, https_conn_class=None, *args, **kwargs):
compat_urllib_request.HTTPSHandler.__init__(self, *args, **kwargs)
def https_open(self, req):
kwargs = {}
+ conn_class = self._https_conn_class
+
if hasattr(self, '_context'): # python > 2.6
kwargs['context'] = self._context
if hasattr(self, '_check_hostname'): # python 3.x
kwargs['check_hostname'] = self._check_hostname
+
+ socks_proxy = req.headers.get('Ytdl-socks-proxy')
+ if socks_proxy:
+ conn_class = make_socks_conn_class(conn_class, socks_proxy)
+ del req.headers['Ytdl-socks-proxy']
+
return self.do_open(functools.partial(
- _create_http_connection, self, self._https_conn_class, True),
+ _create_http_connection, self, conn_class, True),
req, **kwargs)
+class YoutubeDLCookieProcessor(compat_urllib_request.HTTPCookieProcessor):
+ def __init__(self, cookiejar=None):
+ compat_urllib_request.HTTPCookieProcessor.__init__(self, cookiejar)
+
+ def http_response(self, request, response):
+ # Python 2 will choke on next HTTP request in row if there are non-ASCII
+ # characters in Set-Cookie HTTP header of last response (see
+ # https://github.com/rg3/youtube-dl/issues/6769).
+ # In order to at least prevent crashing we will percent encode Set-Cookie
+ # header before HTTPCookieProcessor starts processing it.
+ # if sys.version_info < (3, 0) and response.headers:
+ # for set_cookie_header in ('Set-Cookie', 'Set-Cookie2'):
+ # set_cookie = response.headers.get(set_cookie_header)
+ # if set_cookie:
+ # set_cookie_escaped = compat_urllib_parse.quote(set_cookie, b"%/;:@&=+$,!~*'()?#[] ")
+ # if set_cookie != set_cookie_escaped:
+ # del response.headers[set_cookie_header]
+ # response.headers[set_cookie_header] = set_cookie_escaped
+ return compat_urllib_request.HTTPCookieProcessor.http_response(self, request, response)
+
+ https_request = compat_urllib_request.HTTPCookieProcessor.http_request
+ https_response = http_response
+
+
+def extract_timezone(date_str):
+ m = re.search(
+ r'^.{8,}?(?P<tz>Z$| ?(?P<sign>\+|-)(?P<hours>[0-9]{2}):?(?P<minutes>[0-9]{2})$)',
+ date_str)
+ if not m:
+ timezone = datetime.timedelta()
+ else:
+ date_str = date_str[:-len(m.group('tz'))]
+ if not m.group('sign'):
+ timezone = datetime.timedelta()
+ else:
+ sign = 1 if m.group('sign') == '+' else -1
+ timezone = datetime.timedelta(
+ hours=sign * int(m.group('hours')),
+ minutes=sign * int(m.group('minutes')))
+ return timezone, date_str
+
+
def parse_iso8601(date_str, delimiter='T', timezone=None):
""" Return a UNIX timestamp from the given date """
if date_str is None:
return None
+ date_str = re.sub(r'\.[0-9]+', '', date_str)
+
if timezone is None:
- m = re.search(
- r'(\.[0-9]+)?(?:Z$| ?(?P<sign>\+|-)(?P<hours>[0-9]{2}):?(?P<minutes>[0-9]{2})$)',
- date_str)
- if not m:
- timezone = datetime.timedelta()
- else:
- date_str = date_str[:-len(m.group(0))]
- if not m.group('sign'):
- timezone = datetime.timedelta()
- else:
- sign = 1 if m.group('sign') == '+' else -1
- timezone = datetime.timedelta(
- hours=sign * int(m.group('hours')),
- minutes=sign * int(m.group('minutes')))
- date_format = '%Y-%m-%d{0}%H:%M:%S'.format(delimiter)
- dt = datetime.datetime.strptime(date_str, date_format) - timezone
- return calendar.timegm(dt.timetuple())
+ timezone, date_str = extract_timezone(date_str)
+
+ try:
+ date_format = '%Y-%m-%d{0}%H:%M:%S'.format(delimiter)
+ dt = datetime.datetime.strptime(date_str, date_format) - timezone
+ return calendar.timegm(dt.timetuple())
+ except ValueError:
+ pass
+
+
+def date_formats(day_first=True):
+ return DATE_FORMATS_DAY_FIRST if day_first else DATE_FORMATS_MONTH_FIRST
def unified_strdate(date_str, day_first=True):
upload_date = None
# Replace commas
date_str = date_str.replace(',', ' ')
- # %z (UTC offset) is only supported in python>=3.2
- if not re.match(r'^[0-9]{1,2}-[0-9]{1,2}-[0-9]{4}$', date_str):
- date_str = re.sub(r' ?(\+|-)[0-9]{2}:?[0-9]{2}$', '', date_str)
# Remove AM/PM + timezone
date_str = re.sub(r'(?i)\s*(?:AM|PM)(?:\s+[A-Z]+)?', '', date_str)
+ _, date_str = extract_timezone(date_str)
- format_expressions = [
- '%d %B %Y',
- '%d %b %Y',
- '%B %d %Y',
- '%b %d %Y',
- '%b %dst %Y %I:%M%p',
- '%b %dnd %Y %I:%M%p',
- '%b %dth %Y %I:%M%p',
- '%Y %m %d',
- '%Y-%m-%d',
- '%Y/%m/%d',
- '%Y/%m/%d %H:%M:%S',
- '%Y-%m-%d %H:%M:%S',
- '%Y-%m-%d %H:%M:%S.%f',
- '%d.%m.%Y %H:%M',
- '%d.%m.%Y %H.%M',
- '%Y-%m-%dT%H:%M:%SZ',
- '%Y-%m-%dT%H:%M:%S.%fZ',
- '%Y-%m-%dT%H:%M:%S.%f0Z',
- '%Y-%m-%dT%H:%M:%S',
- '%Y-%m-%dT%H:%M:%S.%f',
- '%Y-%m-%dT%H:%M',
- ]
- if day_first:
- format_expressions.extend([
- '%d-%m-%Y',
- '%d.%m.%Y',
- '%d/%m/%Y',
- '%d/%m/%y',
- '%d/%m/%Y %H:%M:%S',
- ])
- else:
- format_expressions.extend([
- '%m-%d-%Y',
- '%m.%d.%Y',
- '%m/%d/%Y',
- '%m/%d/%y',
- '%m/%d/%Y %H:%M:%S',
- ])
- for expression in format_expressions:
+ for expression in date_formats(day_first):
try:
upload_date = datetime.datetime.strptime(date_str, expression).strftime('%Y%m%d')
except ValueError:
if upload_date is None:
timetuple = email.utils.parsedate_tz(date_str)
if timetuple:
- upload_date = datetime.datetime(*timetuple[:6]).strftime('%Y%m%d')
- return upload_date
+ try:
+ upload_date = datetime.datetime(*timetuple[:6]).strftime('%Y%m%d')
+ except ValueError:
+ pass
+ if upload_date is not None:
+ return compat_str(upload_date)
+
+
+def unified_timestamp(date_str, day_first=True):
+ if date_str is None:
+ return None
+
+ date_str = date_str.replace(',', ' ')
+
+ pm_delta = 12 if re.search(r'(?i)PM', date_str) else 0
+ timezone, date_str = extract_timezone(date_str)
+
+ # Remove AM/PM + timezone
+ date_str = re.sub(r'(?i)\s*(?:AM|PM)(?:\s+[A-Z]+)?', '', date_str)
+
+ for expression in date_formats(day_first):
+ try:
+ dt = datetime.datetime.strptime(date_str, expression) - timezone + datetime.timedelta(hours=pm_delta)
+ return calendar.timegm(dt.timetuple())
+ except ValueError:
+ pass
+ timetuple = email.utils.parsedate_tz(date_str)
+ if timetuple:
+ return calendar.timegm(timetuple) + pm_delta * 3600
def determine_ext(url, default_ext='unknown_video'):
guess = url.partition('?')[0].rpartition('.')[2]
if re.match(r'^[A-Za-z0-9]+$', guess):
return guess
+ # Try extract ext from URLs like http://example.com/foo/bar.mp4/?download
+ elif guess.rstrip('/') in KNOWN_EXTENSIONS:
+ return guess.rstrip('/')
else:
return default_ext
if sign == '-':
time = -time
unit = match.group('unit')
- # A bad aproximation?
+ # A bad approximation?
if unit == 'month':
unit = 'day'
time *= 30
unit += 's'
delta = datetime.timedelta(**{unit: time})
return today + delta
- return datetime.datetime.strptime(date_str, "%Y%m%d").date()
+ return datetime.datetime.strptime(date_str, '%Y%m%d').date()
def hyphenate_date(date_str):
GetStdHandle = ctypes.WINFUNCTYPE(
ctypes.wintypes.HANDLE, ctypes.wintypes.DWORD)(
- (b"GetStdHandle", ctypes.windll.kernel32))
+ (b'GetStdHandle', ctypes.windll.kernel32))
h = GetStdHandle(WIN_OUTPUT_IDS[fileno])
WriteConsoleW = ctypes.WINFUNCTYPE(
ctypes.wintypes.BOOL, ctypes.wintypes.HANDLE, ctypes.wintypes.LPWSTR,
ctypes.wintypes.DWORD, ctypes.POINTER(ctypes.wintypes.DWORD),
- ctypes.wintypes.LPVOID)((b"WriteConsoleW", ctypes.windll.kernel32))
+ ctypes.wintypes.LPVOID)((b'WriteConsoleW', ctypes.windll.kernel32))
written = ctypes.wintypes.DWORD(0)
- GetFileType = ctypes.WINFUNCTYPE(ctypes.wintypes.DWORD, ctypes.wintypes.DWORD)((b"GetFileType", ctypes.windll.kernel32))
+ GetFileType = ctypes.WINFUNCTYPE(ctypes.wintypes.DWORD, ctypes.wintypes.DWORD)((b'GetFileType', ctypes.windll.kernel32))
FILE_TYPE_CHAR = 0x0002
FILE_TYPE_REMOTE = 0x8000
GetConsoleMode = ctypes.WINFUNCTYPE(
ctypes.wintypes.BOOL, ctypes.wintypes.HANDLE,
ctypes.POINTER(ctypes.wintypes.DWORD))(
- (b"GetConsoleMode", ctypes.windll.kernel32))
+ (b'GetConsoleMode', ctypes.windll.kernel32))
INVALID_HANDLE_VALUE = ctypes.wintypes.DWORD(-1).value
def not_a_console(handle):
def intlist_to_bytes(xs):
if not xs:
return b''
- return struct_pack('%dB' % len(xs), *xs)
+ return compat_struct_pack('%dB' % len(xs), *xs)
# Cross-platform file locking
raise OSError('Unlocking file failed: %r' % ctypes.FormatError())
else:
- import fcntl
+ # Some platforms, such as Jython, is missing fcntl
+ try:
+ import fcntl
- def _lock_file(f, exclusive):
- fcntl.flock(f, fcntl.LOCK_EX if exclusive else fcntl.LOCK_SH)
+ def _lock_file(f, exclusive):
+ fcntl.flock(f, fcntl.LOCK_EX if exclusive else fcntl.LOCK_SH)
- def _unlock_file(f):
- fcntl.flock(f, fcntl.LOCK_UN)
+ def _unlock_file(f):
+ fcntl.flock(f, fcntl.LOCK_UN)
+ except ImportError:
+ UNSUPPORTED_MSG = 'file locking is not supported on this platform'
+
+ def _lock_file(f, exclusive):
+ raise IOError(UNSUPPORTED_MSG)
+
+ def _unlock_file(f):
+ raise IOError(UNSUPPORTED_MSG)
class locked_file(object):
def smuggle_url(url, data):
""" Pass additional data in a URL for internal use. """
- sdata = compat_urllib_parse.urlencode(
+ url, idata = unsmuggle_url(url, {})
+ data.update(idata)
+ sdata = compat_urllib_parse_urlencode(
{'__youtubedl_smuggle': json.dumps(data)})
return url + '#' + sdata
return '%.2f%s' % (converted, suffix)
+def lookup_unit_table(unit_table, s):
+ units_re = '|'.join(re.escape(u) for u in unit_table)
+ m = re.match(
+ r'(?P<num>[0-9]+(?:[,.][0-9]*)?)\s*(?P<unit>%s)\b' % units_re, s)
+ if not m:
+ return None
+ num_str = m.group('num').replace(',', '.')
+ mult = unit_table[m.group('unit')]
+ return int(float(num_str) * mult)
+
+
def parse_filesize(s):
if s is None:
return None
- # The lower-case forms are of course incorrect and inofficial,
+ # The lower-case forms are of course incorrect and unofficial,
# but we support those too
_UNIT_TABLE = {
'B': 1,
'b': 1,
+ 'bytes': 1,
'KiB': 1024,
'KB': 1000,
'kB': 1024,
'Kb': 1000,
+ 'kb': 1000,
+ 'kilobytes': 1000,
+ 'kibibytes': 1024,
'MiB': 1024 ** 2,
'MB': 1000 ** 2,
'mB': 1024 ** 2,
'Mb': 1000 ** 2,
+ 'mb': 1000 ** 2,
+ 'megabytes': 1000 ** 2,
+ 'mebibytes': 1024 ** 2,
'GiB': 1024 ** 3,
'GB': 1000 ** 3,
'gB': 1024 ** 3,
'Gb': 1000 ** 3,
+ 'gb': 1000 ** 3,
+ 'gigabytes': 1000 ** 3,
+ 'gibibytes': 1024 ** 3,
'TiB': 1024 ** 4,
'TB': 1000 ** 4,
'tB': 1024 ** 4,
'Tb': 1000 ** 4,
+ 'tb': 1000 ** 4,
+ 'terabytes': 1000 ** 4,
+ 'tebibytes': 1024 ** 4,
'PiB': 1024 ** 5,
'PB': 1000 ** 5,
'pB': 1024 ** 5,
'Pb': 1000 ** 5,
+ 'pb': 1000 ** 5,
+ 'petabytes': 1000 ** 5,
+ 'pebibytes': 1024 ** 5,
'EiB': 1024 ** 6,
'EB': 1000 ** 6,
'eB': 1024 ** 6,
'Eb': 1000 ** 6,
+ 'eb': 1000 ** 6,
+ 'exabytes': 1000 ** 6,
+ 'exbibytes': 1024 ** 6,
'ZiB': 1024 ** 7,
'ZB': 1000 ** 7,
'zB': 1024 ** 7,
'Zb': 1000 ** 7,
+ 'zb': 1000 ** 7,
+ 'zettabytes': 1000 ** 7,
+ 'zebibytes': 1024 ** 7,
'YiB': 1024 ** 8,
'YB': 1000 ** 8,
'yB': 1024 ** 8,
'Yb': 1000 ** 8,
+ 'yb': 1000 ** 8,
+ 'yottabytes': 1000 ** 8,
+ 'yobibytes': 1024 ** 8,
}
- units_re = '|'.join(re.escape(u) for u in _UNIT_TABLE)
- m = re.match(
- r'(?P<num>[0-9]+(?:[,.][0-9]*)?)\s*(?P<unit>%s)' % units_re, s)
- if not m:
+ return lookup_unit_table(_UNIT_TABLE, s)
+
+
+def parse_count(s):
+ if s is None:
return None
- num_str = m.group('num').replace(',', '.')
- mult = _UNIT_TABLE[m.group('unit')]
- return int(float(num_str) * mult)
+ s = s.strip()
+
+ if re.match(r'^[\d,.]+$', s):
+ return str_to_int(s)
+
+ _UNIT_TABLE = {
+ 'k': 1000,
+ 'K': 1000,
+ 'm': 1000 ** 2,
+ 'M': 1000 ** 2,
+ 'kk': 1000 ** 2,
+ 'KK': 1000 ** 2,
+ }
+ return lookup_unit_table(_UNIT_TABLE, s)
-def month_by_name(name):
+
+def month_by_name(name, lang='en'):
""" Return the number of a month by (locale-independently) English name """
+ month_names = MONTH_NAMES.get(lang, MONTH_NAMES['en'])
+
try:
- return ENGLISH_MONTH_NAMES.index(name) + 1
+ return month_names.index(name) + 1
except ValueError:
return None
def setproctitle(title):
assert isinstance(title, compat_str)
+
+ # ctypes in Jython is not complete
+ # http://bugs.jython.org/issue2148
+ if sys.platform.startswith('java'):
+ return
+
try:
- libc = ctypes.cdll.LoadLibrary("libc.so.6")
+ libc = ctypes.cdll.LoadLibrary('libc.so.6')
except OSError:
return
title_bytes = title.encode('utf-8')
def remove_start(s, start):
- if s.startswith(start):
- return s[len(start):]
- return s
+ return s[len(start):] if s is not None and s.startswith(start) else s
def remove_end(s, end):
- if s.endswith(end):
- return s[:-len(end)]
+ return s[:-len(end)] if s is not None and s.endswith(end) else s
+
+
+def remove_quotes(s):
+ if s is None or len(s) < 2:
+ return s
+ for quote in ('"', "'", ):
+ if s[0] == quote and s[-1] == quote:
+ return s[1:-1]
return s
return path.strip('/').split('/')[-1]
+def base_url(url):
+ return re.match(r'https?://[^?#&]+/', url).group()
+
+
class HEADRequest(compat_urllib_request.Request):
def get_method(self):
- return "HEAD"
+ return 'HEAD'
+
+
+class PUTRequest(compat_urllib_request.Request):
+ def get_method(self):
+ return 'PUT'
def int_or_none(v, scale=1, default=None, get_attr=None, invscale=1):
v = getattr(v, get_attr, None)
if v == '':
v = None
- return default if v is None else (int(v) * invscale // scale)
+ if v is None:
+ return default
+ try:
+ return int(v) * invscale // scale
+ except ValueError:
+ return default
def str_or_none(v, default=None):
def float_or_none(v, scale=1, invscale=1, default=None):
- return default if v is None else (float(v) * invscale / scale)
+ if v is None:
+ return default
+ try:
+ return float(v) * invscale / scale
+ except ValueError:
+ return default
+
+
+def strip_or_none(v):
+ return None if v is None else v.strip()
def parse_duration(s):
s = s.strip()
- m = re.match(
- r'''(?ix)(?:P?T)?
- (?:
- (?P<only_mins>[0-9.]+)\s*(?:mins?|minutes?)\s*|
- (?P<only_hours>[0-9.]+)\s*(?:hours?)|
-
- \s*(?P<hours_reversed>[0-9]+)\s*(?:[:h]|hours?)\s*(?P<mins_reversed>[0-9]+)\s*(?:[:m]|mins?|minutes?)\s*|
- (?:
+ days, hours, mins, secs, ms = [None] * 5
+ m = re.match(r'(?:(?:(?:(?P<days>[0-9]+):)?(?P<hours>[0-9]+):)?(?P<mins>[0-9]+):)?(?P<secs>[0-9]+)(?P<ms>\.[0-9]+)?$', s)
+ if m:
+ days, hours, mins, secs, ms = m.groups()
+ else:
+ m = re.match(
+ r'''(?ix)(?:P?T)?
(?:
- (?:(?P<days>[0-9]+)\s*(?:[:d]|days?)\s*)?
- (?P<hours>[0-9]+)\s*(?:[:h]|hours?)\s*
+ (?P<days>[0-9]+)\s*d(?:ays?)?\s*
)?
- (?P<mins>[0-9]+)\s*(?:[:m]|mins?|minutes?)\s*
- )?
- (?P<secs>[0-9]+)(?P<ms>\.[0-9]+)?\s*(?:s|secs?|seconds?)?
- )$''', s)
- if not m:
- return None
- res = 0
- if m.group('only_mins'):
- return float_or_none(m.group('only_mins'), invscale=60)
- if m.group('only_hours'):
- return float_or_none(m.group('only_hours'), invscale=60 * 60)
- if m.group('secs'):
- res += int(m.group('secs'))
- if m.group('mins_reversed'):
- res += int(m.group('mins_reversed')) * 60
- if m.group('mins'):
- res += int(m.group('mins')) * 60
- if m.group('hours'):
- res += int(m.group('hours')) * 60 * 60
- if m.group('hours_reversed'):
- res += int(m.group('hours_reversed')) * 60 * 60
- if m.group('days'):
- res += int(m.group('days')) * 24 * 60 * 60
- if m.group('ms'):
- res += float(m.group('ms'))
- return res
+ (?:
+ (?P<hours>[0-9]+)\s*h(?:ours?)?\s*
+ )?
+ (?:
+ (?P<mins>[0-9]+)\s*m(?:in(?:ute)?s?)?\s*
+ )?
+ (?:
+ (?P<secs>[0-9]+)(?P<ms>\.[0-9]+)?\s*s(?:ec(?:ond)?s?)?\s*
+ )?$''', s)
+ if m:
+ days, hours, mins, secs, ms = m.groups()
+ else:
+ m = re.match(r'(?i)(?:(?P<hours>[0-9.]+)\s*(?:hours?)|(?P<mins>[0-9.]+)\s*(?:mins?\.?|minutes?)\s*)$', s)
+ if m:
+ hours, mins = m.groups()
+ else:
+ return None
+
+ duration = 0
+ if secs:
+ duration += float(secs)
+ if mins:
+ duration += float(mins) * 60
+ if hours:
+ duration += float(hours) * 60 * 60
+ if days:
+ duration += float(days) * 24 * 60 * 60
+ if ms:
+ duration += float(ms)
+ return duration
def prepend_extension(filename, ext, expected_real_ext=None):
""" Returns the version of the specified executable,
or False if the executable is not present """
try:
+ # STDIN should be redirected too. On UNIX-like systems, ffmpeg triggers
+ # SIGTTOU if youtube-dl is run in the background.
+ # See https://github.com/rg3/youtube-dl/issues/955#issuecomment-209789656
out, _ = subprocess.Popen(
[encodeArgument(exe)] + args,
+ stdin=subprocess.PIPE,
stdout=subprocess.PIPE, stderr=subprocess.STDOUT).communicate()
except OSError:
return False
class OnDemandPagedList(PagedList):
- def __init__(self, pagefunc, pagesize):
+ def __init__(self, pagefunc, pagesize, use_cache=False):
self._pagefunc = pagefunc
self._pagesize = pagesize
+ self._use_cache = use_cache
+ if use_cache:
+ self._cache = {}
def getslice(self, start=0, end=None):
res = []
if start >= nextfirstid:
continue
- page_results = list(self._pagefunc(pagenum))
+ page_results = None
+ if self._use_cache:
+ page_results = self._cache.get(pagenum)
+ if page_results is None:
+ page_results = list(self._pagefunc(pagenum))
+ if self._use_cache:
+ self._cache[pagenum] = page_results
startv = (
start % self._pagesize
"""Escape URL as suggested by RFC 3986"""
url_parsed = compat_urllib_parse_urlparse(url)
return url_parsed._replace(
+ netloc=url_parsed.netloc.encode('idna').decode('ascii'),
path=escape_rfc3986(url_parsed.path),
params=escape_rfc3986(url_parsed.params),
query=escape_rfc3986(url_parsed.query),
fragment=escape_rfc3986(url_parsed.fragment)
).geturl()
-try:
- struct.pack('!I', 0)
-except TypeError:
- # In Python 2.6 (and some 2.7 versions), struct requires a bytes argument
- def struct_pack(spec, *args):
- if isinstance(spec, compat_str):
- spec = spec.encode('ascii')
- return struct.pack(spec, *args)
-
- def struct_unpack(spec, *args):
- if isinstance(spec, compat_str):
- spec = spec.encode('ascii')
- return struct.unpack(spec, *args)
-else:
- struct_pack = struct.pack
- struct_unpack = struct.unpack
-
def read_batch_urls(batch_fd):
def fixup(url):
def urlencode_postdata(*args, **kargs):
- return compat_urllib_parse.urlencode(*args, **kargs).encode('ascii')
+ return compat_urllib_parse_urlencode(*args, **kargs).encode('ascii')
-try:
- etree_iter = xml.etree.ElementTree.Element.iter
-except AttributeError: # Python <=2.6
- etree_iter = lambda n: n.findall('.//*')
+def update_url_query(url, query):
+ if not query:
+ return url
+ parsed_url = compat_urlparse.urlparse(url)
+ qs = compat_parse_qs(parsed_url.query)
+ qs.update(query)
+ return compat_urlparse.urlunparse(parsed_url._replace(
+ query=compat_urllib_parse_urlencode(qs, True)))
+
+
+def update_Request(req, url=None, data=None, headers={}, query={}):
+ req_headers = req.headers.copy()
+ req_headers.update(headers)
+ req_data = data or req.data
+ req_url = update_url_query(url or req.get_full_url(), query)
+ req_get_method = req.get_method()
+ if req_get_method == 'HEAD':
+ req_type = HEADRequest
+ elif req_get_method == 'PUT':
+ req_type = PUTRequest
+ else:
+ req_type = compat_urllib_request.Request
+ new_req = req_type(
+ req_url, data=req_data, headers=req_headers,
+ origin_req_host=req.origin_req_host, unverifiable=req.unverifiable)
+ if hasattr(req, 'timeout'):
+ new_req.timeout = req.timeout
+ return new_req
+
+
+def dict_get(d, key_or_keys, default=None, skip_false_values=True):
+ if isinstance(key_or_keys, (list, tuple)):
+ for key in key_or_keys:
+ if key not in d or d[key] is None or skip_false_values and not d[key]:
+ continue
+ return d[key]
+ return default
+ return d.get(key_or_keys, default)
-def parse_xml(s):
- class TreeBuilder(xml.etree.ElementTree.TreeBuilder):
- def doctype(self, name, pubid, system):
- pass # Ignore doctypes
+def try_get(src, getter, expected_type=None):
+ try:
+ v = getter(src)
+ except (AttributeError, KeyError, TypeError, IndexError):
+ pass
+ else:
+ if expected_type is None or isinstance(v, expected_type):
+ return v
- parser = xml.etree.ElementTree.XMLParser(target=TreeBuilder())
- kwargs = {'parser': parser} if sys.version_info >= (2, 7) else {}
- tree = xml.etree.ElementTree.XML(s.encode('utf-8'), **kwargs)
- # Fix up XML parser in Python 2.x
- if sys.version_info < (3, 0):
- for n in etree_iter(tree):
- if n.text is not None:
- if not isinstance(n.text, compat_str):
- n.text = n.text.decode('utf-8')
- return tree
+
+def encode_compat_str(string, encoding=preferredencoding(), errors='strict'):
+ return string if isinstance(string, compat_str) else compat_str(string, encoding, errors)
US_RATINGS = {
}
+TV_PARENTAL_GUIDELINES = {
+ 'TV-Y': 0,
+ 'TV-Y7': 7,
+ 'TV-G': 0,
+ 'TV-PG': 0,
+ 'TV-14': 14,
+ 'TV-MA': 17,
+}
+
+
def parse_age_limit(s):
- if s is None:
+ if type(s) == int:
+ return s if 0 <= s <= 21 else None
+ if not isinstance(s, compat_basestring):
return None
m = re.match(r'^(?P<age>\d{1,2})\+?$', s)
- return int(m.group('age')) if m else US_RATINGS.get(s, None)
+ if m:
+ return int(m.group('age'))
+ if s in US_RATINGS:
+ return US_RATINGS[s]
+ return TV_PARENTAL_GUIDELINES.get(s)
def strip_jsonp(code):
return re.sub(
- r'(?s)^[a-zA-Z0-9_]+\s*\(\s*(.*)\);?\s*?(?://[^\n]*)*$', r'\1', code)
+ r'(?s)^[a-zA-Z0-9_.$]+\s*\(\s*(.*)\);?\s*?(?://[^\n]*)*$', r'\1', code)
def js_to_json(code):
v = m.group(0)
if v in ('true', 'false', 'null'):
return v
- if v.startswith('"'):
- return v
- if v.startswith("'"):
- v = v[1:-1]
- v = re.sub(r"\\\\|\\'|\"", lambda m: {
- '\\\\': '\\\\',
- "\\'": "'",
+ elif v.startswith('/*') or v == ',':
+ return ""
+
+ if v[0] in ("'", '"'):
+ v = re.sub(r'(?s)\\.|"', lambda m: {
'"': '\\"',
- }[m.group(0)], v)
+ "\\'": "'",
+ '\\\n': '',
+ '\\x': '\\u00',
+ }.get(m.group(0), m.group(0)), v[1:-1])
+
+ INTEGER_TABLE = (
+ (r'^(0[xX][0-9a-fA-F]+)\s*:?$', 16),
+ (r'^(0+[0-7]+)\s*:?$', 8),
+ )
+
+ for regex, base in INTEGER_TABLE:
+ im = re.match(regex, v)
+ if im:
+ i = int(im.group(1), base)
+ return '"%d":' % i if v.endswith(':') else '%d' % i
+
return '"%s"' % v
- res = re.sub(r'''(?x)
- "(?:[^"\\]*(?:\\\\|\\['"nu]))*[^"\\]*"|
- '(?:[^'\\]*(?:\\\\|\\['"nu]))*[^'\\]*'|
- [a-zA-Z_][.a-zA-Z_0-9]*
+ return re.sub(r'''(?sx)
+ "(?:[^"\\]*(?:\\\\|\\['"nurtbfx/\n]))*[^"\\]*"|
+ '(?:[^'\\]*(?:\\\\|\\['"nurtbfx/\n]))*[^'\\]*'|
+ /\*.*?\*/|,(?=\s*[\]}])|
+ [a-zA-Z_][.a-zA-Z_0-9]*|
+ \b(?:0[xX][0-9a-fA-F]+|0+[0-7]+)(?:\s*:)?|
+ [0-9]+(?=\s*:)
''', fix_kv, code)
- res = re.sub(r',(\s*[\]}])', lambda m: m.group(1), res)
- return res
def qualities(quality_ids):
def args_to_str(args):
# Get a short string representation for a subprocess command
- return ' '.join(shlex_quote(a) for a in args)
+ return ' '.join(compat_shlex_quote(a) for a in args)
+
+
+def error_to_compat_str(err):
+ err_str = str(err)
+ # On python 2 error byte string must be decoded with proper
+ # encoding rather than ascii
+ if sys.version_info[0] < 3:
+ err_str = err_str.decode(preferredencoding())
+ return err_str
def mimetype2ext(mt):
+ if mt is None:
+ return None
+
+ ext = {
+ 'audio/mp4': 'm4a',
+ # Per RFC 3003, audio/mpeg can be .mp1, .mp2 or .mp3. Here use .mp3 as
+ # it's the most popular one
+ 'audio/mpeg': 'mp3',
+ }.get(mt)
+ if ext is not None:
+ return ext
+
_, _, res = mt.rpartition('/')
+ res = res.split(';')[0].strip().lower()
return {
- 'x-ms-wmv': 'wmv',
- 'x-mp4-fragmented': 'mp4',
+ '3gpp': '3gp',
+ 'smptett+xml': 'tt',
+ 'srt': 'srt',
+ 'ttaf+xml': 'dfxp',
'ttml+xml': 'ttml',
+ 'vtt': 'vtt',
+ 'x-flv': 'flv',
+ 'x-mp4-fragmented': 'mp4',
+ 'x-ms-wmv': 'wmv',
+ 'mpegurl': 'm3u8',
+ 'x-mpegurl': 'm3u8',
+ 'vnd.apple.mpegurl': 'm3u8',
+ 'dash+xml': 'mpd',
+ 'f4m': 'f4m',
+ 'f4m+xml': 'f4m',
+ 'hds+xml': 'f4m',
+ 'vnd.ms-sstr+xml': 'ism',
+ 'quicktime': 'mov',
}.get(res, res)
+def parse_codecs(codecs_str):
+ # http://tools.ietf.org/html/rfc6381
+ if not codecs_str:
+ return {}
+ splited_codecs = list(filter(None, map(
+ lambda str: str.strip(), codecs_str.strip().strip(',').split(','))))
+ vcodec, acodec = None, None
+ for full_codec in splited_codecs:
+ codec = full_codec.split('.')[0]
+ if codec in ('avc1', 'avc2', 'avc3', 'avc4', 'vp9', 'vp8', 'hev1', 'hev2', 'h263', 'h264', 'mp4v'):
+ if not vcodec:
+ vcodec = full_codec
+ elif codec in ('mp4a', 'opus', 'vorbis', 'mp3', 'aac', 'ac-3'):
+ if not acodec:
+ acodec = full_codec
+ else:
+ write_string('WARNING: Unknown codec %s' % full_codec, sys.stderr)
+ if not vcodec and not acodec:
+ if len(splited_codecs) == 2:
+ return {
+ 'vcodec': vcodec,
+ 'acodec': acodec,
+ }
+ elif len(splited_codecs) == 1:
+ return {
+ 'vcodec': 'none',
+ 'acodec': vcodec,
+ }
+ else:
+ return {
+ 'vcodec': vcodec or 'none',
+ 'acodec': acodec or 'none',
+ }
+ return {}
+
+
def urlhandle_detect_ext(url_handle):
- try:
- url_handle.headers
- getheader = lambda h: url_handle.headers[h]
- except AttributeError: # Python < 3
- getheader = url_handle.info().getheader
+ getheader = url_handle.headers.get
cd = getheader('Content-Disposition')
if cd:
return mimetype2ext(getheader('Content-Type'))
+def encode_data_uri(data, mime_type):
+ return 'data:%s;base64,%s' % (mime_type, base64.b64encode(data).decode('ascii'))
+
+
def age_restricted(content_limit, age_limit):
""" Returns True iff the content should be blocked """
m = operator_rex.search(filter_part)
if m:
op = COMPARISON_OPERATORS[m.group('op')]
- if m.group('strval') is not None:
+ actual_value = dct.get(m.group('key'))
+ if (m.group('strval') is not None or
+ # If the original field is a string and matching comparisonvalue is
+ # a number we should respect the origin of the original field
+ # and process comparison value as a string (see
+ # https://github.com/rg3/youtube-dl/issues/11082).
+ actual_value is not None and m.group('intval') is not None and
+ isinstance(actual_value, compat_str)):
if m.group('op') not in ('=', '!='):
raise ValueError(
'Operator %s does not support string values!' % m.group('op'))
- comparison_value = m.group('strval')
+ comparison_value = m.group('strval') or m.group('intval')
else:
try:
comparison_value = int(m.group('intval'))
raise ValueError(
'Invalid integer value %r in filter part %r' % (
m.group('intval'), filter_part))
- actual_value = dct.get(m.group('key'))
if actual_value is None:
return m.group('none_inclusive')
return op(actual_value, comparison_value)
def parse_dfxp_time_expr(time_expr):
if not time_expr:
- return 0.0
+ return
mobj = re.match(r'^(?P<time_offset>\d+(?:\.\d+)?)s?$', time_expr)
if mobj:
return float(mobj.group('time_offset'))
- mobj = re.match(r'^(\d+):(\d\d):(\d\d(?:\.\d+)?)$', time_expr)
+ mobj = re.match(r'^(\d+):(\d\d):(\d\d(?:(?:\.|:)\d+)?)$', time_expr)
if mobj:
- return 3600 * int(mobj.group(1)) + 60 * int(mobj.group(2)) + float(mobj.group(3))
+ return 3600 * int(mobj.group(1)) + 60 * int(mobj.group(2)) + float(mobj.group(3).replace(':', '.'))
def srt_subtitles_timecode(seconds):
_x = functools.partial(xpath_with_ns, ns_map={
'ttml': 'http://www.w3.org/ns/ttml',
'ttaf1': 'http://www.w3.org/2006/10/ttaf1',
+ 'ttaf1_0604': 'http://www.w3.org/2006/04/ttaf1',
})
- def parse_node(node):
- str_or_empty = functools.partial(str_or_none, default='')
+ class TTMLPElementParser(object):
+ out = ''
- out = str_or_empty(node.text)
+ def start(self, tag, attrib):
+ if tag in (_x('ttml:br'), _x('ttaf1:br'), 'br'):
+ self.out += '\n'
- for child in node:
- if child.tag in (_x('ttml:br'), _x('ttaf1:br'), 'br'):
- out += '\n' + str_or_empty(child.tail)
- elif child.tag in (_x('ttml:span'), _x('ttaf1:span'), 'span'):
- out += str_or_empty(parse_node(child))
- else:
- out += str_or_empty(xml.etree.ElementTree.tostring(child))
+ def end(self, tag):
+ pass
- return out
+ def data(self, data):
+ self.out += data
- dfxp = xml.etree.ElementTree.fromstring(dfxp_data.encode('utf-8'))
+ def close(self):
+ return self.out.strip()
+
+ def parse_node(node):
+ target = TTMLPElementParser()
+ parser = xml.etree.ElementTree.XMLParser(target=target)
+ parser.feed(xml.etree.ElementTree.tostring(node))
+ return parser.close()
+
+ dfxp = compat_etree_fromstring(dfxp_data.encode('utf-8'))
out = []
- paras = dfxp.findall(_x('.//ttml:p')) or dfxp.findall(_x('.//ttaf1:p')) or dfxp.findall('.//p')
+ paras = dfxp.findall(_x('.//ttml:p')) or dfxp.findall(_x('.//ttaf1:p')) or dfxp.findall(_x('.//ttaf1_0604:p')) or dfxp.findall('.//p')
if not paras:
raise ValueError('Invalid dfxp/TTML subtitle')
for para, index in zip(paras, itertools.count(1)):
- begin_time = parse_dfxp_time_expr(para.attrib['begin'])
+ begin_time = parse_dfxp_time_expr(para.attrib.get('begin'))
end_time = parse_dfxp_time_expr(para.attrib.get('end'))
+ dur = parse_dfxp_time_expr(para.attrib.get('dur'))
+ if begin_time is None:
+ continue
if not end_time:
- end_time = begin_time + parse_dfxp_time_expr(para.attrib['dur'])
+ if not dur:
+ continue
+ end_time = begin_time + dur
out.append('%d\n%s --> %s\n%s\n\n' % (
index,
srt_subtitles_timecode(begin_time),
return ''.join(out)
+def cli_option(params, command_option, param):
+ param = params.get(param)
+ if param:
+ param = compat_str(param)
+ return [command_option, param] if param is not None else []
+
+
+def cli_bool_option(params, command_option, param, true_value='true', false_value='false', separator=None):
+ param = params.get(param)
+ assert isinstance(param, bool)
+ if separator:
+ return [command_option + separator + (true_value if param else false_value)]
+ return [command_option, true_value if param else false_value]
+
+
+def cli_valueless_option(params, command_option, param, expected_value=True):
+ param = params.get(param)
+ return [command_option] if param == expected_value else []
+
+
+def cli_configuration_args(params, param, default=[]):
+ ex_args = params.get(param)
+ if ex_args is None:
+ return default
+ assert isinstance(ex_args, list)
+ return ex_args
+
+
class ISO639Utils(object):
# See http://www.loc.gov/standards/iso639-2/ISO-639-2_utf-8.txt
_lang_map = {
if proxy == '__noproxy__':
return None # No Proxy
+ if compat_urlparse.urlparse(proxy).scheme.lower() in ('socks', 'socks4', 'socks4a', 'socks5'):
+ req.add_header('Ytdl-socks-proxy', proxy)
+ # youtube-dl's http/https handlers do wrapping the socket with socks
+ return None
return compat_urllib_request.ProxyHandler.proxy_open(
self, req, proxy, type)
+
+
+def ohdave_rsa_encrypt(data, exponent, modulus):
+ '''
+ Implement OHDave's RSA algorithm. See http://www.ohdave.com/rsa/
+
+ Input:
+ data: data to encrypt, bytes-like object
+ exponent, modulus: parameter e and N of RSA algorithm, both integer
+ Output: hex string of encrypted data
+
+ Limitation: supports one block encryption only
+ '''
+
+ payload = int(binascii.hexlify(data[::-1]), 16)
+ encrypted = pow(payload, exponent, modulus)
+ return '%x' % encrypted
+
+
+def encode_base_n(num, n, table=None):
+ FULL_TABLE = '0123456789abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ'
+ if not table:
+ table = FULL_TABLE[:n]
+
+ if n > len(table):
+ raise ValueError('base %d exceeds table length %d' % (n, len(table)))
+
+ if num == 0:
+ return table[0]
+
+ ret = ''
+ while num:
+ ret = table[num % n] + ret
+ num = num // n
+ return ret
+
+
+def decode_packed_codes(code):
+ mobj = re.search(PACKED_CODES_RE, code)
+ obfucasted_code, base, count, symbols = mobj.groups()
+ base = int(base)
+ count = int(count)
+ symbols = symbols.split('|')
+ symbol_table = {}
+
+ while count:
+ count -= 1
+ base_n_count = encode_base_n(count, base)
+ symbol_table[base_n_count] = symbols[count] or base_n_count
+
+ return re.sub(
+ r'\b(\w+)\b', lambda mobj: symbol_table[mobj.group(0)],
+ obfucasted_code)
+
+
+def parse_m3u8_attributes(attrib):
+ info = {}
+ for (key, val) in re.findall(r'(?P<key>[A-Z0-9-]+)=(?P<val>"[^"]+"|[^",]+)(?:,|$)', attrib):
+ if val.startswith('"'):
+ val = val[1:-1]
+ info[key] = val
+ return info
+
+
+def urshift(val, n):
+ return val >> n if val >= 0 else (val + 0x100000000) >> n
+
+
+# Based on png2str() written by @gdkchan and improved by @yokrysty
+# Originally posted at https://github.com/rg3/youtube-dl/issues/9706
+def decode_png(png_data):
+ # Reference: https://www.w3.org/TR/PNG/
+ header = png_data[8:]
+
+ if png_data[:8] != b'\x89PNG\x0d\x0a\x1a\x0a' or header[4:8] != b'IHDR':
+ raise IOError('Not a valid PNG file.')
+
+ int_map = {1: '>B', 2: '>H', 4: '>I'}
+ unpack_integer = lambda x: compat_struct_unpack(int_map[len(x)], x)[0]
+
+ chunks = []
+
+ while header:
+ length = unpack_integer(header[:4])
+ header = header[4:]
+
+ chunk_type = header[:4]
+ header = header[4:]
+
+ chunk_data = header[:length]
+ header = header[length:]
+
+ header = header[4:] # Skip CRC
+
+ chunks.append({
+ 'type': chunk_type,
+ 'length': length,
+ 'data': chunk_data
+ })
+
+ ihdr = chunks[0]['data']
+
+ width = unpack_integer(ihdr[:4])
+ height = unpack_integer(ihdr[4:8])
+
+ idat = b''
+
+ for chunk in chunks:
+ if chunk['type'] == b'IDAT':
+ idat += chunk['data']
+
+ if not idat:
+ raise IOError('Unable to read PNG data.')
+
+ decompressed_data = bytearray(zlib.decompress(idat))
+
+ stride = width * 3
+ pixels = []
+
+ def _get_pixel(idx):
+ x = idx % stride
+ y = idx // stride
+ return pixels[y][x]
+
+ for y in range(height):
+ basePos = y * (1 + stride)
+ filter_type = decompressed_data[basePos]
+
+ current_row = []
+
+ pixels.append(current_row)
+
+ for x in range(stride):
+ color = decompressed_data[1 + basePos + x]
+ basex = y * stride + x
+ left = 0
+ up = 0
+
+ if x > 2:
+ left = _get_pixel(basex - 3)
+ if y > 0:
+ up = _get_pixel(basex - stride)
+
+ if filter_type == 1: # Sub
+ color = (color + left) & 0xff
+ elif filter_type == 2: # Up
+ color = (color + up) & 0xff
+ elif filter_type == 3: # Average
+ color = (color + ((left + up) >> 1)) & 0xff
+ elif filter_type == 4: # Paeth
+ a = left
+ b = up
+ c = 0
+
+ if x > 2 and y > 0:
+ c = _get_pixel(basex - stride - 3)
+
+ p = a + b - c
+
+ pa = abs(p - a)
+ pb = abs(p - b)
+ pc = abs(p - c)
+
+ if pa <= pb and pa <= pc:
+ color = (color + a) & 0xff
+ elif pb <= pc:
+ color = (color + b) & 0xff
+ else:
+ color = (color + c) & 0xff
+
+ current_row.append(color)
+
+ return width, height, pixels
+
+
+def write_xattr(path, key, value):
+ # This mess below finds the best xattr tool for the job
+ try:
+ # try the pyxattr module...
+ import xattr
+
+ if hasattr(xattr, 'set'): # pyxattr
+ # Unicode arguments are not supported in python-pyxattr until
+ # version 0.5.0
+ # See https://github.com/rg3/youtube-dl/issues/5498
+ pyxattr_required_version = '0.5.0'
+ if version_tuple(xattr.__version__) < version_tuple(pyxattr_required_version):
+ # TODO: fallback to CLI tools
+ raise XAttrUnavailableError(
+ 'python-pyxattr is detected but is too old. '
+ 'youtube-dl requires %s or above while your version is %s. '
+ 'Falling back to other xattr implementations' % (
+ pyxattr_required_version, xattr.__version__))
+
+ setxattr = xattr.set
+ else: # xattr
+ setxattr = xattr.setxattr
+
+ try:
+ setxattr(path, key, value)
+ except EnvironmentError as e:
+ raise XAttrMetadataError(e.errno, e.strerror)
+
+ except ImportError:
+ if compat_os_name == 'nt':
+ # Write xattrs to NTFS Alternate Data Streams:
+ # http://en.wikipedia.org/wiki/NTFS#Alternate_data_streams_.28ADS.29
+ assert ':' not in key
+ assert os.path.exists(path)
+
+ ads_fn = path + ':' + key
+ try:
+ with open(ads_fn, 'wb') as f:
+ f.write(value)
+ except EnvironmentError as e:
+ raise XAttrMetadataError(e.errno, e.strerror)
+ else:
+ user_has_setfattr = check_executable('setfattr', ['--version'])
+ user_has_xattr = check_executable('xattr', ['-h'])
+
+ if user_has_setfattr or user_has_xattr:
+
+ value = value.decode('utf-8')
+ if user_has_setfattr:
+ executable = 'setfattr'
+ opts = ['-n', key, '-v', value]
+ elif user_has_xattr:
+ executable = 'xattr'
+ opts = ['-w', key, value]
+
+ cmd = ([encodeFilename(executable, True)] +
+ [encodeArgument(o) for o in opts] +
+ [encodeFilename(path, True)])
+
+ try:
+ p = subprocess.Popen(
+ cmd, stdout=subprocess.PIPE, stderr=subprocess.PIPE, stdin=subprocess.PIPE)
+ except EnvironmentError as e:
+ raise XAttrMetadataError(e.errno, e.strerror)
+ stdout, stderr = p.communicate()
+ stderr = stderr.decode('utf-8', 'replace')
+ if p.returncode != 0:
+ raise XAttrMetadataError(p.returncode, stderr)
+
+ else:
+ # On Unix, and can't find pyxattr, setfattr, or xattr.
+ if sys.platform.startswith('linux'):
+ raise XAttrUnavailableError(
+ "Couldn't find a tool to set the xattrs. "
+ "Install either the python 'pyxattr' or 'xattr' "
+ "modules, or the GNU 'attr' package "
+ "(which contains the 'setfattr' tool).")
+ else:
+ raise XAttrUnavailableError(
+ "Couldn't find a tool to set the xattrs. "
+ "Install either the python 'xattr' module, "
+ "or the 'xattr' binary.")
Rapsys Git / youtubedl / blobdiff