Dre4m Shell
Server IP : 85.214.239.14  /  Your IP : 3.143.237.54
Web Server : Apache/2.4.62 (Debian)
System : Linux h2886529.stratoserver.net 4.9.0 #1 SMP Tue Jan 9 19:45:01 MSK 2024 x86_64
User : www-data ( 33)
PHP Version : 7.4.18
Disable Function : pcntl_alarm,pcntl_fork,pcntl_waitpid,pcntl_wait,pcntl_wifexited,pcntl_wifstopped,pcntl_wifsignaled,pcntl_wifcontinued,pcntl_wexitstatus,pcntl_wtermsig,pcntl_wstopsig,pcntl_signal,pcntl_signal_get_handler,pcntl_signal_dispatch,pcntl_get_last_error,pcntl_strerror,pcntl_sigprocmask,pcntl_sigwaitinfo,pcntl_sigtimedwait,pcntl_exec,pcntl_getpriority,pcntl_setpriority,pcntl_async_signals,pcntl_unshare,
MySQL : OFF  |  cURL : OFF  |  WGET : ON  |  Perl : ON  |  Python : ON  |  Sudo : ON  |  Pkexec : OFF
Directory :  /proc/self/root/proc/2/cwd/proc/3/cwd/proc/2/cwd/usr/lib/python3.11/email/

Upload File :
current_dir [ Writeable ] document_root [ Writeable ]

 

Command :


[ HOME SHELL ]     

Current File : /proc/self/root/proc/2/cwd/proc/3/cwd/proc/2/cwd/usr/lib/python3.11/email//utils.py
# Copyright (C) 2001-2010 Python Software Foundation
# Author: Barry Warsaw
# Contact: email-sig@python.org

"""Miscellaneous utilities."""

__all__ = [
    'collapse_rfc2231_value',
    'decode_params',
    'decode_rfc2231',
    'encode_rfc2231',
    'formataddr',
    'formatdate',
    'format_datetime',
    'getaddresses',
    'make_msgid',
    'mktime_tz',
    'parseaddr',
    'parsedate',
    'parsedate_tz',
    'parsedate_to_datetime',
    'unquote',
    ]

import os
import re
import time
import random
import socket
import datetime
import urllib.parse

from email._parseaddr import quote
from email._parseaddr import AddressList as _AddressList
from email._parseaddr import mktime_tz

from email._parseaddr import parsedate, parsedate_tz, _parsedate_tz

# Intrapackage imports
from email.charset import Charset

COMMASPACE = ', '
EMPTYSTRING = ''
UEMPTYSTRING = ''
CRLF = '\r\n'
TICK = "'"

specialsre = re.compile(r'[][\\()<>@,:;".]')
escapesre = re.compile(r'[\\"]')

def _has_surrogates(s):
    """Return True if s contains surrogate-escaped binary data."""
    # This check is based on the fact that unless there are surrogates, utf8
    # (Python's default encoding) can encode any string.  This is the fastest
    # way to check for surrogates, see issue 11454 for timings.
    try:
        s.encode()
        return False
    except UnicodeEncodeError:
        return True

# How to deal with a string containing bytes before handing it to the
# application through the 'normal' interface.
def _sanitize(string):
    # Turn any escaped bytes into unicode 'unknown' char.  If the escaped
    # bytes happen to be utf-8 they will instead get decoded, even if they
    # were invalid in the charset the source was supposed to be in.  This
    # seems like it is not a bad thing; a defect was still registered.
    original_bytes = string.encode('utf-8', 'surrogateescape')
    return original_bytes.decode('utf-8', 'replace')



# Helpers

def formataddr(pair, charset='utf-8'):
    """The inverse of parseaddr(), this takes a 2-tuple of the form
    (realname, email_address) and returns the string value suitable
    for an RFC 2822 From, To or Cc header.

    If the first element of pair is false, then the second element is
    returned unmodified.

    The optional charset is the character set that is used to encode
    realname in case realname is not ASCII safe.  Can be an instance of str or
    a Charset-like object which has a header_encode method.  Default is
    'utf-8'.
    """
    name, address = pair
    # The address MUST (per RFC) be ascii, so raise a UnicodeError if it isn't.
    address.encode('ascii')
    if name:
        try:
            name.encode('ascii')
        except UnicodeEncodeError:
            if isinstance(charset, str):
                charset = Charset(charset)
            encoded_name = charset.header_encode(name)
            return "%s <%s>" % (encoded_name, address)
        else:
            quotes = ''
            if specialsre.search(name):
                quotes = '"'
            name = escapesre.sub(r'\\\g<0>', name)
            return '%s%s%s <%s>' % (quotes, name, quotes, address)
    return address



def getaddresses(fieldvalues):
    """Return a list of (REALNAME, EMAIL) for each fieldvalue."""
    all = COMMASPACE.join(str(v) for v in fieldvalues)
    a = _AddressList(all)
    return a.addresslist


def _format_timetuple_and_zone(timetuple, zone):
    return '%s, %02d %s %04d %02d:%02d:%02d %s' % (
        ['Mon', 'Tue', 'Wed', 'Thu', 'Fri', 'Sat', 'Sun'][timetuple[6]],
        timetuple[2],
        ['Jan', 'Feb', 'Mar', 'Apr', 'May', 'Jun',
         'Jul', 'Aug', 'Sep', 'Oct', 'Nov', 'Dec'][timetuple[1] - 1],
        timetuple[0], timetuple[3], timetuple[4], timetuple[5],
        zone)

def formatdate(timeval=None, localtime=False, usegmt=False):
    """Returns a date string as specified by RFC 2822, e.g.:

    Fri, 09 Nov 2001 01:08:47 -0000

    Optional timeval if given is a floating point time value as accepted by
    gmtime() and localtime(), otherwise the current time is used.

    Optional localtime is a flag that when True, interprets timeval, and
    returns a date relative to the local timezone instead of UTC, properly
    taking daylight savings time into account.

    Optional argument usegmt means that the timezone is written out as
    an ascii string, not numeric one (so "GMT" instead of "+0000"). This
    is needed for HTTP, and is only used when localtime==False.
    """
    # Note: we cannot use strftime() because that honors the locale and RFC
    # 2822 requires that day and month names be the English abbreviations.
    if timeval is None:
        timeval = time.time()
    if localtime or usegmt:
        dt = datetime.datetime.fromtimestamp(timeval, datetime.timezone.utc)
    else:
        dt = datetime.datetime.utcfromtimestamp(timeval)
    if localtime:
        dt = dt.astimezone()
        usegmt = False
    return format_datetime(dt, usegmt)

def format_datetime(dt, usegmt=False):
    """Turn a datetime into a date string as specified in RFC 2822.

    If usegmt is True, dt must be an aware datetime with an offset of zero.  In
    this case 'GMT' will be rendered instead of the normal +0000 required by
    RFC2822.  This is to support HTTP headers involving date stamps.
    """
    now = dt.timetuple()
    if usegmt:
        if dt.tzinfo is None or dt.tzinfo != datetime.timezone.utc:
            raise ValueError("usegmt option requires a UTC datetime")
        zone = 'GMT'
    elif dt.tzinfo is None:
        zone = '-0000'
    else:
        zone = dt.strftime("%z")
    return _format_timetuple_and_zone(now, zone)


def make_msgid(idstring=None, domain=None):
    """Returns a string suitable for RFC 2822 compliant Message-ID, e.g:

    <142480216486.20800.16526388040877946887@nightshade.la.mastaler.com>

    Optional idstring if given is a string used to strengthen the
    uniqueness of the message id.  Optional domain if given provides the
    portion of the message id after the '@'.  It defaults to the locally
    defined hostname.
    """
    timeval = int(time.time()*100)
    pid = os.getpid()
    randint = random.getrandbits(64)
    if idstring is None:
        idstring = ''
    else:
        idstring = '.' + idstring
    if domain is None:
        domain = socket.getfqdn()
    msgid = '<%d.%d.%d%s@%s>' % (timeval, pid, randint, idstring, domain)
    return msgid


def parsedate_to_datetime(data):
    parsed_date_tz = _parsedate_tz(data)
    if parsed_date_tz is None:
        raise ValueError('Invalid date value or format "%s"' % str(data))
    *dtuple, tz = parsed_date_tz
    if tz is None:
        return datetime.datetime(*dtuple[:6])
    return datetime.datetime(*dtuple[:6],
            tzinfo=datetime.timezone(datetime.timedelta(seconds=tz)))


def parseaddr(addr):
    """
    Parse addr into its constituent realname and email address parts.

    Return a tuple of realname and email address, unless the parse fails, in
    which case return a 2-tuple of ('', '').
    """
    addrs = _AddressList(addr).addresslist
    if not addrs:
        return '', ''
    return addrs[0]


# rfc822.unquote() doesn't properly de-backslash-ify in Python pre-2.3.
def unquote(str):
    """Remove quotes from a string."""
    if len(str) > 1:
        if str.startswith('"') and str.endswith('"'):
            return str[1:-1].replace('\\\\', '\\').replace('\\"', '"')
        if str.startswith('<') and str.endswith('>'):
            return str[1:-1]
    return str



# RFC2231-related functions - parameter encoding and decoding
def decode_rfc2231(s):
    """Decode string according to RFC 2231"""
    parts = s.split(TICK, 2)
    if len(parts) <= 2:
        return None, None, s
    return parts


def encode_rfc2231(s, charset=None, language=None):
    """Encode string according to RFC 2231.

    If neither charset nor language is given, then s is returned as-is.  If
    charset is given but not language, the string is encoded using the empty
    string for language.
    """
    s = urllib.parse.quote(s, safe='', encoding=charset or 'ascii')
    if charset is None and language is None:
        return s
    if language is None:
        language = ''
    return "%s'%s'%s" % (charset, language, s)


rfc2231_continuation = re.compile(r'^(?P<name>\w+)\*((?P<num>[0-9]+)\*?)?$',
    re.ASCII)

def decode_params(params):
    """Decode parameters list according to RFC 2231.

    params is a sequence of 2-tuples containing (param name, string value).
    """
    new_params = [params[0]]
    # Map parameter's name to a list of continuations.  The values are a
    # 3-tuple of the continuation number, the string value, and a flag
    # specifying whether a particular segment is %-encoded.
    rfc2231_params = {}
    for name, value in params[1:]:
        encoded = name.endswith('*')
        value = unquote(value)
        mo = rfc2231_continuation.match(name)
        if mo:
            name, num = mo.group('name', 'num')
            if num is not None:
                num = int(num)
            rfc2231_params.setdefault(name, []).append((num, value, encoded))
        else:
            new_params.append((name, '"%s"' % quote(value)))
    if rfc2231_params:
        for name, continuations in rfc2231_params.items():
            value = []
            extended = False
            # Sort by number
            continuations.sort()
            # And now append all values in numerical order, converting
            # %-encodings for the encoded segments.  If any of the
            # continuation names ends in a *, then the entire string, after
            # decoding segments and concatenating, must have the charset and
            # language specifiers at the beginning of the string.
            for num, s, encoded in continuations:
                if encoded:
                    # Decode as "latin-1", so the characters in s directly
                    # represent the percent-encoded octet values.
                    # collapse_rfc2231_value treats this as an octet sequence.
                    s = urllib.parse.unquote(s, encoding="latin-1")
                    extended = True
                value.append(s)
            value = quote(EMPTYSTRING.join(value))
            if extended:
                charset, language, value = decode_rfc2231(value)
                new_params.append((name, (charset, language, '"%s"' % value)))
            else:
                new_params.append((name, '"%s"' % value))
    return new_params

def collapse_rfc2231_value(value, errors='replace',
                           fallback_charset='us-ascii'):
    if not isinstance(value, tuple) or len(value) != 3:
        return unquote(value)
    # While value comes to us as a unicode string, we need it to be a bytes
    # object.  We do not want bytes() normal utf-8 decoder, we want a straight
    # interpretation of the string as character bytes.
    charset, language, text = value
    if charset is None:
        # Issue 17369: if charset/lang is None, decode_rfc2231 couldn't parse
        # the value, so use the fallback_charset.
        charset = fallback_charset
    rawbytes = bytes(text, 'raw-unicode-escape')
    try:
        return str(rawbytes, charset, errors)
    except LookupError:
        # charset is not a known codec.
        return unquote(text)


#
# datetime doesn't provide a localtime function yet, so provide one.  Code
# adapted from the patch in issue 9527.  This may not be perfect, but it is
# better than not having it.
#

def localtime(dt=None, isdst=-1):
    """Return local time as an aware datetime object.

    If called without arguments, return current time.  Otherwise *dt*
    argument should be a datetime instance, and it is converted to the
    local time zone according to the system time zone database.  If *dt* is
    naive (that is, dt.tzinfo is None), it is assumed to be in local time.
    In this case, a positive or zero value for *isdst* causes localtime to
    presume initially that summer time (for example, Daylight Saving Time)
    is or is not (respectively) in effect for the specified time.  A
    negative value for *isdst* causes the localtime() function to attempt
    to divine whether summer time is in effect for the specified time.

    """
    if dt is None:
        return datetime.datetime.now(datetime.timezone.utc).astimezone()
    if dt.tzinfo is not None:
        return dt.astimezone()
    # We have a naive datetime.  Convert to a (localtime) timetuple and pass to
    # system mktime together with the isdst hint.  System mktime will return
    # seconds since epoch.
    tm = dt.timetuple()[:-1] + (isdst,)
    seconds = time.mktime(tm)
    localtm = time.localtime(seconds)
    try:
        delta = datetime.timedelta(seconds=localtm.tm_gmtoff)
        tz = datetime.timezone(delta, localtm.tm_zone)
    except AttributeError:
        # Compute UTC offset and compare with the value implied by tm_isdst.
        # If the values match, use the zone name implied by tm_isdst.
        delta = dt - datetime.datetime(*time.gmtime(seconds)[:6])
        dst = time.daylight and localtm.tm_isdst > 0
        gmtoff = -(time.altzone if dst else time.timezone)
        if delta == datetime.timedelta(seconds=gmtoff):
            tz = datetime.timezone(delta, time.tzname[dst])
        else:
            tz = datetime.timezone(delta)
    return dt.replace(tzinfo=tz)

Anon7 - 2022
AnonSec Team