cpython/Lib/http/cookies.py

608 lines
20 KiB
Python
Raw Normal View History

####
# Copyright 2000 by Timothy O'Malley <timo@alum.mit.edu>
2001-01-14 19:36:06 -04:00
#
# All Rights Reserved
2001-01-14 19:36:06 -04:00
#
# Permission to use, copy, modify, and distribute this software
# and its documentation for any purpose and without fee is hereby
# granted, provided that the above copyright notice appear in all
# copies and that both that copyright notice and this permission
# notice appear in supporting documentation, and that the name of
# Timothy O'Malley not be used in advertising or publicity
# pertaining to distribution of the software without specific, written
2001-01-14 19:36:06 -04:00
# prior permission.
#
# Timothy O'Malley DISCLAIMS ALL WARRANTIES WITH REGARD TO THIS
# SOFTWARE, INCLUDING ALL IMPLIED WARRANTIES OF MERCHANTABILITY
# AND FITNESS, IN NO EVENT SHALL Timothy O'Malley BE LIABLE FOR
# ANY SPECIAL, INDIRECT OR CONSEQUENTIAL DAMAGES OR ANY DAMAGES
# WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR PROFITS,
# WHETHER IN AN ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS
# ACTION, ARISING OUT OF OR IN CONNECTION WITH THE USE OR
2001-01-14 19:36:06 -04:00
# PERFORMANCE OF THIS SOFTWARE.
#
####
2001-01-14 19:36:06 -04:00
#
# Id: Cookie.py,v 2.29 2000/08/23 05:28:49 timo Exp
# by Timothy O'Malley <timo@alum.mit.edu>
#
# Cookie.py is a Python module for the handling of HTTP
# cookies as a Python dictionary. See RFC 2109 for more
# information on cookies.
#
# The original idea to treat Cookies as a dictionary came from
# Dave Mitchell (davem@magnet.com) in 1995, when he released the
# first version of nscookie.py.
#
####
r"""
Here's a sample session to show how to use this module.
At the moment, this is the only documentation.
The Basics
----------
Importing is easy...
2008-05-26 13:32:26 -03:00
>>> from http import cookies
Most of the time you start by creating a cookie.
2008-05-26 13:32:26 -03:00
>>> C = cookies.SimpleCookie()
Once you've created your Cookie, you can add values just as if it were
a dictionary.
>>> C = cookies.SimpleCookie()
>>> C["fig"] = "newton"
>>> C["sugar"] = "wafer"
>>> C.output()
'Set-Cookie: fig=newton\r\nSet-Cookie: sugar=wafer'
Notice that the printable representation of a Cookie is the
appropriate format for a Set-Cookie: header. This is the
default behavior. You can change the header and printed
attributes by using the .output() function
>>> C = cookies.SimpleCookie()
>>> C["rocky"] = "road"
>>> C["rocky"]["path"] = "/cookie"
2007-02-09 16:33:44 -04:00
>>> print(C.output(header="Cookie:"))
Cookie: rocky=road; Path=/cookie
2007-02-09 16:33:44 -04:00
>>> print(C.output(attrs=[], header="Cookie:"))
Cookie: rocky=road
The load() method of a Cookie extracts cookies from a string. In a
CGI script, you would use this method to extract the cookies from the
HTTP_COOKIE environment variable.
>>> C = cookies.SimpleCookie()
>>> C.load("chips=ahoy; vienna=finger")
>>> C.output()
'Set-Cookie: chips=ahoy\r\nSet-Cookie: vienna=finger'
The load() method is darn-tootin smart about identifying cookies
within a string. Escaped quotation marks, nested semicolons, and other
such trickeries do not confuse it.
>>> C = cookies.SimpleCookie()
>>> C.load('keebler="E=everybody; L=\\"Loves\\"; fudge=\\012;";')
2007-02-09 16:33:44 -04:00
>>> print(C)
Set-Cookie: keebler="E=everybody; L=\"Loves\"; fudge=\012;"
Each element of the Cookie also supports all of the RFC 2109
Cookie attributes. Here's an example which sets the Path
attribute.
>>> C = cookies.SimpleCookie()
>>> C["oreo"] = "doublestuff"
>>> C["oreo"]["path"] = "/"
2007-02-09 16:33:44 -04:00
>>> print(C)
Set-Cookie: oreo=doublestuff; Path=/
Each dictionary element has a 'value' attribute, which gives you
2001-01-14 19:36:06 -04:00
back the value associated with the key.
>>> C = cookies.SimpleCookie()
>>> C["twix"] = "none for you"
>>> C["twix"].value
'none for you'
The SimpleCookie expects that all values should be standard strings.
Just to be sure, SimpleCookie invokes the str() builtin to convert
the value to a string, when the values are set dictionary-style.
2008-05-26 13:32:26 -03:00
>>> C = cookies.SimpleCookie()
>>> C["number"] = 7
>>> C["string"] = "seven"
>>> C["number"].value
'7'
>>> C["string"].value
'seven'
>>> C.output()
'Set-Cookie: number=7\r\nSet-Cookie: string=seven'
Finis.
"""
#
# Import our required modules
2001-01-14 19:36:06 -04:00
#
import re
2001-08-02 04:15:29 -03:00
import string
__all__ = ["CookieError", "BaseCookie", "SimpleCookie"]
_nulljoin = ''.join
_semispacejoin = '; '.join
2005-08-25 04:32:42 -03:00
_spacejoin = ' '.join
#
# Define an exception visible to External modules
#
class CookieError(Exception):
pass
# These quoting routines conform to the RFC2109 specification, which in
# turn references the character definitions from RFC2068. They provide
# a two-way quoting algorithm. Any non-text character is translated
# into a 4 character sequence: a forward-slash followed by the
# three-digit octal equivalent of the character. Any '\' or '"' is
# quoted with a preceding '\' slash.
# Because of the way browsers really handle cookies (as opposed to what
# the RFC says) we also encode "," and ";".
2001-01-14 19:36:06 -04:00
#
# These are taken from RFC2068 and RFC2109.
# _LegalChars is the list of chars which don't require "'s
# _Translator hash-table for fast quoting
#
_LegalChars = string.ascii_letters + string.digits + "!#$%&'*+-.^_`|~:"
_UnescapedChars = _LegalChars + ' ()/<=>?@[]{}'
_Translator = {n: '\\%03o' % n
for n in set(range(256)) - set(map(ord, _UnescapedChars))}
_Translator.update({
ord('"'): '\\"',
ord('\\'): '\\\\',
})
_is_legal_key = re.compile('[%s]+' % re.escape(_LegalChars)).fullmatch
def _quote(str):
2009-09-04 05:28:01 -03:00
r"""Quote a string for use in a cookie header.
If the string does not need to be double-quoted, then just return the
string. Otherwise, surround the string in doublequotes and quote
(with a \) special characters.
"""
if str is None or _is_legal_key(str):
return str
else:
return '"' + str.translate(_Translator) + '"'
_OctalPatt = re.compile(r"\\[0-3][0-7][0-7]")
_QuotePatt = re.compile(r"[\\].")
def _unquote(str):
# If there aren't any doublequotes,
# then there can't be any special characters. See RFC 2109.
if str is None or len(str) < 2:
return str
if str[0] != '"' or str[-1] != '"':
return str
# We have to assume that we must decode this string.
# Down to work.
# Remove the "s
str = str[1:-1]
# Check for special sequences. Examples:
# \012 --> \n
# \" --> "
#
i = 0
n = len(str)
res = []
while 0 <= i < n:
o_match = _OctalPatt.search(str, i)
q_match = _QuotePatt.search(str, i)
if not o_match and not q_match: # Neither matched
res.append(str[i:])
break
# else:
j = k = -1
if o_match:
j = o_match.start(0)
if q_match:
k = q_match.start(0)
if q_match and (not o_match or k < j): # QuotePatt matched
res.append(str[i:k])
res.append(str[k+1])
i = k + 2
else: # OctalPatt matched
res.append(str[i:j])
res.append(chr(int(str[j+1:j+4], 8)))
i = j + 4
return _nulljoin(res)
# The _getdate() routine is used to set the expiration time in the cookie's HTTP
# header. By default, _getdate() returns the current time in the appropriate
# "expires" format for a Set-Cookie header. The one optional argument is an
# offset from now, in seconds. For example, an offset of -3600 means "one hour
# ago". The offset may be a floating point number.
#
_weekdayname = ['Mon', 'Tue', 'Wed', 'Thu', 'Fri', 'Sat', 'Sun']
_monthname = [None,
'Jan', 'Feb', 'Mar', 'Apr', 'May', 'Jun',
'Jul', 'Aug', 'Sep', 'Oct', 'Nov', 'Dec']
def _getdate(future=0, weekdayname=_weekdayname, monthname=_monthname):
from time import gmtime, time
now = time()
year, month, day, hh, mm, ss, wd, y, z = gmtime(now + future)
return "%s, %02d %3s %4d %02d:%02d:%02d GMT" % \
(weekdayname[wd], day, monthname[month], year, hh, mm, ss)
class Morsel(dict):
"""A class to hold ONE (key, value) pair.
2009-09-04 05:28:01 -03:00
In a cookie, each such pair may have several attributes, so this class is
used to keep the attributes associated with the appropriate key,value pair.
This class also includes a coded_value attribute, which is used to hold
the network representation of the value.
2009-09-04 05:28:01 -03:00
"""
# RFC 2109 lists these attributes as reserved:
# path comment domain
# max-age secure version
2001-01-14 19:36:06 -04:00
#
# For historical reasons, these attributes are also reserved:
# expires
#
# This is an extension from Microsoft:
# httponly
#
# This dictionary provides a mapping from the lowercase
# variant on the left to the appropriate traditional
# formatting on the right.
_reserved = {
"expires" : "expires",
"path" : "Path",
"comment" : "Comment",
"domain" : "Domain",
"max-age" : "Max-Age",
"secure" : "Secure",
"httponly" : "HttpOnly",
"version" : "Version",
"samesite" : "SameSite",
}
_flags = {'secure', 'httponly'}
def __init__(self):
# Set defaults
self._key = self._value = self._coded_value = None
# Set default attributes
for key in self._reserved:
dict.__setitem__(self, key, "")
@property
def key(self):
return self._key
@property
def value(self):
return self._value
@property
def coded_value(self):
return self._coded_value
def __setitem__(self, K, V):
K = K.lower()
if not K in self._reserved:
raise CookieError("Invalid attribute %r" % (K,))
dict.__setitem__(self, K, V)
def setdefault(self, key, val=None):
key = key.lower()
if key not in self._reserved:
raise CookieError("Invalid attribute %r" % (key,))
return dict.setdefault(self, key, val)
def __eq__(self, morsel):
if not isinstance(morsel, Morsel):
return NotImplemented
return (dict.__eq__(self, morsel) and
self._value == morsel._value and
self._key == morsel._key and
self._coded_value == morsel._coded_value)
__ne__ = object.__ne__
def copy(self):
morsel = Morsel()
dict.update(morsel, self)
morsel.__dict__.update(self.__dict__)
return morsel
def update(self, values):
data = {}
for key, val in dict(values).items():
key = key.lower()
if key not in self._reserved:
raise CookieError("Invalid attribute %r" % (key,))
data[key] = val
dict.update(self, data)
def isReservedKey(self, K):
return K.lower() in self._reserved
def set(self, key, val, coded_val):
if key.lower() in self._reserved:
raise CookieError('Attempt to set a reserved key %r' % (key,))
if not _is_legal_key(key):
raise CookieError('Illegal key %r' % (key,))
# It's a good key, so save it.
self._key = key
self._value = val
self._coded_value = coded_val
def __getstate__(self):
return {
'key': self._key,
'value': self._value,
'coded_value': self._coded_value,
}
def __setstate__(self, state):
self._key = state['key']
self._value = state['value']
self._coded_value = state['coded_value']
def output(self, attrs=None, header="Set-Cookie:"):
return "%s %s" % (header, self.OutputString(attrs))
__str__ = output
def __repr__(self):
return '<%s: %s>' % (self.__class__.__name__, self.OutputString())
def js_output(self, attrs=None):
# Print javascript
return """
<script type="text/javascript">
<!-- begin hiding
document.cookie = \"%s\";
// end hiding -->
</script>
""" % (self.OutputString(attrs).replace('"', r'\"'))
def OutputString(self, attrs=None):
# Build up our result
#
result = []
append = result.append
# First, the key=value pair
append("%s=%s" % (self.key, self.coded_value))
# Now add any defined attributes
if attrs is None:
attrs = self._reserved
items = sorted(self.items())
for key, value in items:
if value == "":
continue
if key not in attrs:
continue
if key == "expires" and isinstance(value, int):
append("%s=%s" % (self._reserved[key], _getdate(value)))
elif key == "max-age" and isinstance(value, int):
append("%s=%d" % (self._reserved[key], value))
elif key in self._flags:
if value:
append(str(self._reserved[key]))
else:
append("%s=%s" % (self._reserved[key], value))
# Return the result
return _semispacejoin(result)
#
# Pattern for finding cookie
#
# This used to be strict parsing based on the RFC2109 and RFC2068
# specifications. I have since discovered that MSIE 3.0x doesn't
# follow the character rules outlined in those specs. As a
# result, the parsing rules here are less strict.
#
_LegalKeyChars = r"\w\d!#%&'~_`><@,:/\$\*\+\-\.\^\|\)\(\?\}\{\="
_LegalValueChars = _LegalKeyChars + r'\[\]'
_CookiePattern = re.compile(r"""
\s* # Optional whitespace at start of cookie
(?P<key> # Start of group 'key'
[""" + _LegalKeyChars + r"""]+? # Any word of at least one letter
) # End of group 'key'
( # Optional group: there may not be a value.
\s*=\s* # Equal Sign
(?P<val> # Start of group 'val'
"(?:[^\\"]|\\.)*" # Any doublequoted string
| # or
\w{3},\s[\w\d\s-]{9,11}\s[\d:]{8}\sGMT # Special case for "expires" attr
| # or
2015-05-23 12:38:48 -03:00
[""" + _LegalValueChars + r"""]* # Any word or empty string
) # End of group 'val'
)? # End of optional value group
\s* # Any number of spaces.
(\s+|;|$) # Ending either at space, semicolon, or EOS.
""", re.ASCII | re.VERBOSE) # re.ASCII may be removed if safe.
# At long last, here is the cookie class. Using this class is almost just like
# using a dictionary. See this module's docstring for example usage.
#
class BaseCookie(dict):
2009-09-04 05:28:01 -03:00
"""A container class for a set of Morsels."""
def value_decode(self, val):
"""real_value, coded_value = value_decode(STRING)
Called prior to setting a cookie's value from the network
representation. The VALUE is the value read from HTTP
header.
Override this function to modify the behavior of cookies.
"""
return val, val
def value_encode(self, val):
"""real_value, coded_value = value_encode(VALUE)
Called prior to setting a cookie's value from the dictionary
representation. The VALUE is the value being assigned.
Override this function to modify the behavior of cookies.
"""
strval = str(val)
return strval, strval
def __init__(self, input=None):
if input:
self.load(input)
def __set(self, key, real_value, coded_value):
"""Private method for setting a cookie's value"""
M = self.get(key, Morsel())
M.set(key, real_value, coded_value)
dict.__setitem__(self, key, M)
def __setitem__(self, key, value):
"""Dictionary style assignment."""
if isinstance(value, Morsel):
# allow assignment of constructed Morsels (e.g. for pickling)
dict.__setitem__(self, key, value)
else:
rval, cval = self.value_encode(value)
self.__set(key, rval, cval)
def output(self, attrs=None, header="Set-Cookie:", sep="\015\012"):
"""Return a string suitable for HTTP."""
result = []
items = sorted(self.items())
for key, value in items:
result.append(value.output(attrs, header))
return sep.join(result)
__str__ = output
def __repr__(self):
l = []
items = sorted(self.items())
for key, value in items:
l.append('%s=%s' % (key, repr(value.value)))
return '<%s: %s>' % (self.__class__.__name__, _spacejoin(l))
def js_output(self, attrs=None):
"""Return a string suitable for JavaScript."""
result = []
items = sorted(self.items())
for key, value in items:
result.append(value.js_output(attrs))
return _nulljoin(result)
def load(self, rawdata):
"""Load cookies from a string (presumably HTTP_COOKIE) or
from a dictionary. Loading cookies from a dictionary 'd'
is equivalent to calling:
map(Cookie.__setitem__, d.keys(), d.values())
"""
if isinstance(rawdata, str):
self.__parse_string(rawdata)
else:
Merged revisions 74277,74321,74323,74326,74355,74465,74467,74488,74492,74513,74531,74549,74553,74625,74632,74643-74644,74647,74652,74666,74671,74727,74739 via svnmerge from svn+ssh://pythondev@svn.python.org/python/trunk ........ r74277 | sean.reifschneider | 2009-08-01 18:54:55 -0500 (Sat, 01 Aug 2009) | 3 lines - Issue #6624: yArg_ParseTuple with "s" format when parsing argument with NUL: Bogus TypeError detail string. ........ r74321 | guilherme.polo | 2009-08-05 11:51:41 -0500 (Wed, 05 Aug 2009) | 1 line Easier reference to find (at least while svn continues being used). ........ r74323 | guilherme.polo | 2009-08-05 18:48:26 -0500 (Wed, 05 Aug 2009) | 1 line Typo. ........ r74326 | jesse.noller | 2009-08-05 21:05:56 -0500 (Wed, 05 Aug 2009) | 1 line Fix issue 4660: spurious task_done errors in multiprocessing, remove doc note for from_address ........ r74355 | gregory.p.smith | 2009-08-12 12:02:37 -0500 (Wed, 12 Aug 2009) | 2 lines comment typo fix ........ r74465 | vinay.sajip | 2009-08-15 18:23:12 -0500 (Sat, 15 Aug 2009) | 1 line Added section on logging to one file from multiple processes. ........ r74467 | vinay.sajip | 2009-08-15 18:34:47 -0500 (Sat, 15 Aug 2009) | 1 line Refined section on logging to one file from multiple processes. ........ r74488 | vinay.sajip | 2009-08-17 08:14:37 -0500 (Mon, 17 Aug 2009) | 1 line Further refined section on logging to one file from multiple processes. ........ r74492 | r.david.murray | 2009-08-17 14:26:49 -0500 (Mon, 17 Aug 2009) | 2 lines Issue 6685: 'toupper' -> 'upper' in cgi doc example explanation. ........ r74513 | skip.montanaro | 2009-08-18 09:37:52 -0500 (Tue, 18 Aug 2009) | 1 line missing module ref (issue6723) ........ r74531 | vinay.sajip | 2009-08-20 17:04:32 -0500 (Thu, 20 Aug 2009) | 1 line Added section on exceptions raised during logging. ........ r74549 | benjamin.peterson | 2009-08-24 12:42:36 -0500 (Mon, 24 Aug 2009) | 1 line fix pdf building by teaching latex the right encoding package ........ r74553 | r.david.murray | 2009-08-26 20:04:59 -0500 (Wed, 26 Aug 2009) | 2 lines Remove leftover text from end of sentence. ........ r74625 | benjamin.peterson | 2009-09-01 17:27:57 -0500 (Tue, 01 Sep 2009) | 1 line remove the check that classmethod's argument is a callable ........ r74632 | georg.brandl | 2009-09-03 02:27:26 -0500 (Thu, 03 Sep 2009) | 1 line #6828: fix wrongly highlighted blocks. ........ r74643 | georg.brandl | 2009-09-04 01:59:20 -0500 (Fri, 04 Sep 2009) | 2 lines Issue #2666: Handle BROWSER environment variable properly for unknown browser names in the webbrowser module. ........ r74644 | georg.brandl | 2009-09-04 02:55:14 -0500 (Fri, 04 Sep 2009) | 1 line #5047: remove Monterey support from configure. ........ r74647 | georg.brandl | 2009-09-04 03:17:04 -0500 (Fri, 04 Sep 2009) | 2 lines Issue #5275: In Cookie's Cookie.load(), properly handle non-string arguments as documented. ........ r74652 | georg.brandl | 2009-09-04 06:25:37 -0500 (Fri, 04 Sep 2009) | 1 line #6756: add some info about the "acct" parameter. ........ r74666 | georg.brandl | 2009-09-05 04:04:09 -0500 (Sat, 05 Sep 2009) | 1 line #6841: remove duplicated word. ........ r74671 | georg.brandl | 2009-09-05 11:47:17 -0500 (Sat, 05 Sep 2009) | 1 line #6843: add link from filterwarnings to where the meaning of the arguments is covered. ........ r74727 | benjamin.peterson | 2009-09-08 18:04:22 -0500 (Tue, 08 Sep 2009) | 1 line #6865 fix ref counting in initialization of pwd module ........ r74739 | georg.brandl | 2009-09-11 02:55:20 -0500 (Fri, 11 Sep 2009) | 1 line Move function back to its section. ........
2009-09-11 19:24:02 -03:00
# self.update() wouldn't call our custom __setitem__
for key, value in rawdata.items():
self[key] = value
return
def __parse_string(self, str, patt=_CookiePattern):
i = 0 # Our starting point
n = len(str) # Length of string
parsed_items = [] # Parsed (type, key, value) triples
morsel_seen = False # A key=value pair was previously encountered
TYPE_ATTRIBUTE = 1
TYPE_KEYVALUE = 2
# We first parse the whole cookie string and reject it if it's
# syntactically invalid (this helps avoid some classes of injection
# attacks).
while 0 <= i < n:
# Start looking for a cookie
match = patt.match(str, i)
if not match:
# No more cookies
break
key, value = match.group("key"), match.group("val")
i = match.end(0)
if key[0] == "$":
if not morsel_seen:
# We ignore attributes which pertain to the cookie
# mechanism as a whole, such as "$Version".
# See RFC 2965. (Does anyone care?)
continue
parsed_items.append((TYPE_ATTRIBUTE, key[1:], value))
elif key.lower() in Morsel._reserved:
if not morsel_seen:
# Invalid cookie string
return
if value is None:
if key.lower() in Morsel._flags:
parsed_items.append((TYPE_ATTRIBUTE, key, True))
else:
# Invalid cookie string
return
else:
parsed_items.append((TYPE_ATTRIBUTE, key, _unquote(value)))
elif value is not None:
parsed_items.append((TYPE_KEYVALUE, key, self.value_decode(value)))
morsel_seen = True
else:
# Invalid cookie string
return
# The cookie string is valid, apply it.
M = None # current morsel
for tp, key, value in parsed_items:
if tp == TYPE_ATTRIBUTE:
assert M is not None
M[key] = value
else:
assert tp == TYPE_KEYVALUE
rval, cval = value
self.__set(key, rval, cval)
M = self[key]
class SimpleCookie(BaseCookie):
2009-09-04 05:28:01 -03:00
"""
SimpleCookie supports strings as cookie values. When setting
the value using the dictionary assignment notation, SimpleCookie
calls the builtin str() to convert the value to a string. Values
received from HTTP are kept as strings.
"""
def value_decode(self, val):
return _unquote(val), val
def value_encode(self, val):
strval = str(val)
return strval, _quote(strval)