freeipa/ipatests/test_ipapython/test_cookie.py

508 lines
18 KiB
Python
Raw Normal View History

Compliant client side session cookie behavior In summary this patch does: * Follow the defined rules for cookies when: - receiving a cookie (process the attributes) - storing a cookie (store cookie + attributes) - sending a cookie + validate the cookie domain against the request URL + validate the cookie path against the request URL + validate the cookie expiration + if valid then send only the cookie, no attribtues * Modifies how a request URL is stored during a XMLRPC request/response sequence. * Refactors a bit of the request/response logic to allow for making the decision whether to send a session cookie instead of full Kerberous auth easier. * The server now includes expiration information in the session cookie it sends to the client. The server always had the information available to prevent using an expired session cookie. Now that expiration timestamp is returned to the client as well and now the client will not send an expired session cookie back to the server. * Adds a new module and unit test for cookies (see below) Formerly we were always returning the session cookie no matter what the domain or path was in the URL. We were also sending the cookie attributes which are for the client only (used to determine if to return a cookie). The attributes are not meant to be sent to the server and the previous behavior was a protocol violation. We also were not checking the cookie expiration. Cookie library issues: We need a library to create, parse, manipulate and format cookies both in a client context and a server context. Core Python has two cookie libraries, Cookie.py and cookielib.py. Why did we add a new cookie module instead of using either of these two core Python libaries? Cookie.py is designed for server side generation but can be used to parse cookies on the client. It's the library we were using in the server. However when I tried to use it in the client I discovered it has some serious bugs. There are 7 defined cookie elements, it fails to correctly parse 3 of the 7 elements which makes it unusable because we depend on those elements. Since Cookie.py was designed for server side cookie processing it's not hard to understand how fails to correctly parse a cookie because that's a client side need. (Cookie.py also has an awkward baroque API and is missing some useful functionality we would have to build on top of it). cookielib.py is designed for client side. It's fully featured and obeys all the RFC's. It would be great to use however it's tightly coupled with another core library, urllib2.py. The http request and response objects must be urllib2 objects. But we don't use urllib2, rather we use httplib because xmlrpclib uses httplib. I don't see a reason why a cookie library should be so tightly coupled to a protocol library, but it is and that means we can't use it (I tried to just pick some isolated entrypoints for our use but I kept hitting interaction/dependency problems). I decided to solve the cookie library problems by writing a minimal cookie library that does what we need and no more than that. It is a new module in ipapython shared by both client and server and comes with a new unit test. The module has plenty of documentation, no need to repeat it here. Request URL issues: We also had problems in rpc.py whereby information from the request which is needed when we process the response is not available. Most important was the requesting URL. It turns out that the way the class and object relationships are structured it's impossible to get this information. Someone else must have run into the same issue because there was a routine called reconstruct_url() which attempted to recreate the request URL from other available information. Unfortunately reconstruct_url() was not callable from inside the response handler. So I decided to store the information in the thread context and when the request is received extract it from the thread context. It's perhaps not an ideal solution but we do similar things elsewhere so at least it's consistent. I removed the reconstruct_url() function because the exact information is now in the context and trying to apply heuristics to recreate the url is probably not robust. Ticket https://fedorahosted.org/freeipa/ticket/3022
2012-12-04 17:20:17 -06:00
# Authors:
# John Dennis <jdennis@redhat.com>
#
# Copyright (C) 2012 Red Hat
# see file 'COPYING' for use and warranty information
#
# This program is free software; you can redistribute it and/or modify
# it under the terms of the GNU General Public License as published by
# the Free Software Foundation, either version 3 of the License, or
# (at your option) any later version.
#
# This program is distributed in the hope that it will be useful,
# but WITHOUT ANY WARRANTY; without even the implied warranty of
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
# GNU General Public License for more details.
#
# You should have received a copy of the GNU General Public License
# along with this program. If not, see <http://www.gnu.org/licenses/>.
import datetime
import email.utils
Compliant client side session cookie behavior In summary this patch does: * Follow the defined rules for cookies when: - receiving a cookie (process the attributes) - storing a cookie (store cookie + attributes) - sending a cookie + validate the cookie domain against the request URL + validate the cookie path against the request URL + validate the cookie expiration + if valid then send only the cookie, no attribtues * Modifies how a request URL is stored during a XMLRPC request/response sequence. * Refactors a bit of the request/response logic to allow for making the decision whether to send a session cookie instead of full Kerberous auth easier. * The server now includes expiration information in the session cookie it sends to the client. The server always had the information available to prevent using an expired session cookie. Now that expiration timestamp is returned to the client as well and now the client will not send an expired session cookie back to the server. * Adds a new module and unit test for cookies (see below) Formerly we were always returning the session cookie no matter what the domain or path was in the URL. We were also sending the cookie attributes which are for the client only (used to determine if to return a cookie). The attributes are not meant to be sent to the server and the previous behavior was a protocol violation. We also were not checking the cookie expiration. Cookie library issues: We need a library to create, parse, manipulate and format cookies both in a client context and a server context. Core Python has two cookie libraries, Cookie.py and cookielib.py. Why did we add a new cookie module instead of using either of these two core Python libaries? Cookie.py is designed for server side generation but can be used to parse cookies on the client. It's the library we were using in the server. However when I tried to use it in the client I discovered it has some serious bugs. There are 7 defined cookie elements, it fails to correctly parse 3 of the 7 elements which makes it unusable because we depend on those elements. Since Cookie.py was designed for server side cookie processing it's not hard to understand how fails to correctly parse a cookie because that's a client side need. (Cookie.py also has an awkward baroque API and is missing some useful functionality we would have to build on top of it). cookielib.py is designed for client side. It's fully featured and obeys all the RFC's. It would be great to use however it's tightly coupled with another core library, urllib2.py. The http request and response objects must be urllib2 objects. But we don't use urllib2, rather we use httplib because xmlrpclib uses httplib. I don't see a reason why a cookie library should be so tightly coupled to a protocol library, but it is and that means we can't use it (I tried to just pick some isolated entrypoints for our use but I kept hitting interaction/dependency problems). I decided to solve the cookie library problems by writing a minimal cookie library that does what we need and no more than that. It is a new module in ipapython shared by both client and server and comes with a new unit test. The module has plenty of documentation, no need to repeat it here. Request URL issues: We also had problems in rpc.py whereby information from the request which is needed when we process the response is not available. Most important was the requesting URL. It turns out that the way the class and object relationships are structured it's impossible to get this information. Someone else must have run into the same issue because there was a routine called reconstruct_url() which attempted to recreate the request URL from other available information. Unfortunately reconstruct_url() was not callable from inside the response handler. So I decided to store the information in the thread context and when the request is received extract it from the thread context. It's perhaps not an ideal solution but we do similar things elsewhere so at least it's consistent. I removed the reconstruct_url() function because the exact information is now in the context and trying to apply heuristics to recreate the url is probably not robust. Ticket https://fedorahosted.org/freeipa/ticket/3022
2012-12-04 17:20:17 -06:00
from ipapython.cookie import Cookie
from ipapython.ipautil import datetime_from_utctimestamp
import pytest
pytestmark = pytest.mark.tier0
class TestParse:
Compliant client side session cookie behavior In summary this patch does: * Follow the defined rules for cookies when: - receiving a cookie (process the attributes) - storing a cookie (store cookie + attributes) - sending a cookie + validate the cookie domain against the request URL + validate the cookie path against the request URL + validate the cookie expiration + if valid then send only the cookie, no attribtues * Modifies how a request URL is stored during a XMLRPC request/response sequence. * Refactors a bit of the request/response logic to allow for making the decision whether to send a session cookie instead of full Kerberous auth easier. * The server now includes expiration information in the session cookie it sends to the client. The server always had the information available to prevent using an expired session cookie. Now that expiration timestamp is returned to the client as well and now the client will not send an expired session cookie back to the server. * Adds a new module and unit test for cookies (see below) Formerly we were always returning the session cookie no matter what the domain or path was in the URL. We were also sending the cookie attributes which are for the client only (used to determine if to return a cookie). The attributes are not meant to be sent to the server and the previous behavior was a protocol violation. We also were not checking the cookie expiration. Cookie library issues: We need a library to create, parse, manipulate and format cookies both in a client context and a server context. Core Python has two cookie libraries, Cookie.py and cookielib.py. Why did we add a new cookie module instead of using either of these two core Python libaries? Cookie.py is designed for server side generation but can be used to parse cookies on the client. It's the library we were using in the server. However when I tried to use it in the client I discovered it has some serious bugs. There are 7 defined cookie elements, it fails to correctly parse 3 of the 7 elements which makes it unusable because we depend on those elements. Since Cookie.py was designed for server side cookie processing it's not hard to understand how fails to correctly parse a cookie because that's a client side need. (Cookie.py also has an awkward baroque API and is missing some useful functionality we would have to build on top of it). cookielib.py is designed for client side. It's fully featured and obeys all the RFC's. It would be great to use however it's tightly coupled with another core library, urllib2.py. The http request and response objects must be urllib2 objects. But we don't use urllib2, rather we use httplib because xmlrpclib uses httplib. I don't see a reason why a cookie library should be so tightly coupled to a protocol library, but it is and that means we can't use it (I tried to just pick some isolated entrypoints for our use but I kept hitting interaction/dependency problems). I decided to solve the cookie library problems by writing a minimal cookie library that does what we need and no more than that. It is a new module in ipapython shared by both client and server and comes with a new unit test. The module has plenty of documentation, no need to repeat it here. Request URL issues: We also had problems in rpc.py whereby information from the request which is needed when we process the response is not available. Most important was the requesting URL. It turns out that the way the class and object relationships are structured it's impossible to get this information. Someone else must have run into the same issue because there was a routine called reconstruct_url() which attempted to recreate the request URL from other available information. Unfortunately reconstruct_url() was not callable from inside the response handler. So I decided to store the information in the thread context and when the request is received extract it from the thread context. It's perhaps not an ideal solution but we do similar things elsewhere so at least it's consistent. I removed the reconstruct_url() function because the exact information is now in the context and trying to apply heuristics to recreate the url is probably not robust. Ticket https://fedorahosted.org/freeipa/ticket/3022
2012-12-04 17:20:17 -06:00
def test_parse(self):
# Empty string
s = ''
cookies = Cookie.parse(s)
assert len(cookies) == 0
Compliant client side session cookie behavior In summary this patch does: * Follow the defined rules for cookies when: - receiving a cookie (process the attributes) - storing a cookie (store cookie + attributes) - sending a cookie + validate the cookie domain against the request URL + validate the cookie path against the request URL + validate the cookie expiration + if valid then send only the cookie, no attribtues * Modifies how a request URL is stored during a XMLRPC request/response sequence. * Refactors a bit of the request/response logic to allow for making the decision whether to send a session cookie instead of full Kerberous auth easier. * The server now includes expiration information in the session cookie it sends to the client. The server always had the information available to prevent using an expired session cookie. Now that expiration timestamp is returned to the client as well and now the client will not send an expired session cookie back to the server. * Adds a new module and unit test for cookies (see below) Formerly we were always returning the session cookie no matter what the domain or path was in the URL. We were also sending the cookie attributes which are for the client only (used to determine if to return a cookie). The attributes are not meant to be sent to the server and the previous behavior was a protocol violation. We also were not checking the cookie expiration. Cookie library issues: We need a library to create, parse, manipulate and format cookies both in a client context and a server context. Core Python has two cookie libraries, Cookie.py and cookielib.py. Why did we add a new cookie module instead of using either of these two core Python libaries? Cookie.py is designed for server side generation but can be used to parse cookies on the client. It's the library we were using in the server. However when I tried to use it in the client I discovered it has some serious bugs. There are 7 defined cookie elements, it fails to correctly parse 3 of the 7 elements which makes it unusable because we depend on those elements. Since Cookie.py was designed for server side cookie processing it's not hard to understand how fails to correctly parse a cookie because that's a client side need. (Cookie.py also has an awkward baroque API and is missing some useful functionality we would have to build on top of it). cookielib.py is designed for client side. It's fully featured and obeys all the RFC's. It would be great to use however it's tightly coupled with another core library, urllib2.py. The http request and response objects must be urllib2 objects. But we don't use urllib2, rather we use httplib because xmlrpclib uses httplib. I don't see a reason why a cookie library should be so tightly coupled to a protocol library, but it is and that means we can't use it (I tried to just pick some isolated entrypoints for our use but I kept hitting interaction/dependency problems). I decided to solve the cookie library problems by writing a minimal cookie library that does what we need and no more than that. It is a new module in ipapython shared by both client and server and comes with a new unit test. The module has plenty of documentation, no need to repeat it here. Request URL issues: We also had problems in rpc.py whereby information from the request which is needed when we process the response is not available. Most important was the requesting URL. It turns out that the way the class and object relationships are structured it's impossible to get this information. Someone else must have run into the same issue because there was a routine called reconstruct_url() which attempted to recreate the request URL from other available information. Unfortunately reconstruct_url() was not callable from inside the response handler. So I decided to store the information in the thread context and when the request is received extract it from the thread context. It's perhaps not an ideal solution but we do similar things elsewhere so at least it's consistent. I removed the reconstruct_url() function because the exact information is now in the context and trying to apply heuristics to recreate the url is probably not robust. Ticket https://fedorahosted.org/freeipa/ticket/3022
2012-12-04 17:20:17 -06:00
# Invalid single token
s = 'color'
with pytest.raises(ValueError):
Compliant client side session cookie behavior In summary this patch does: * Follow the defined rules for cookies when: - receiving a cookie (process the attributes) - storing a cookie (store cookie + attributes) - sending a cookie + validate the cookie domain against the request URL + validate the cookie path against the request URL + validate the cookie expiration + if valid then send only the cookie, no attribtues * Modifies how a request URL is stored during a XMLRPC request/response sequence. * Refactors a bit of the request/response logic to allow for making the decision whether to send a session cookie instead of full Kerberous auth easier. * The server now includes expiration information in the session cookie it sends to the client. The server always had the information available to prevent using an expired session cookie. Now that expiration timestamp is returned to the client as well and now the client will not send an expired session cookie back to the server. * Adds a new module and unit test for cookies (see below) Formerly we were always returning the session cookie no matter what the domain or path was in the URL. We were also sending the cookie attributes which are for the client only (used to determine if to return a cookie). The attributes are not meant to be sent to the server and the previous behavior was a protocol violation. We also were not checking the cookie expiration. Cookie library issues: We need a library to create, parse, manipulate and format cookies both in a client context and a server context. Core Python has two cookie libraries, Cookie.py and cookielib.py. Why did we add a new cookie module instead of using either of these two core Python libaries? Cookie.py is designed for server side generation but can be used to parse cookies on the client. It's the library we were using in the server. However when I tried to use it in the client I discovered it has some serious bugs. There are 7 defined cookie elements, it fails to correctly parse 3 of the 7 elements which makes it unusable because we depend on those elements. Since Cookie.py was designed for server side cookie processing it's not hard to understand how fails to correctly parse a cookie because that's a client side need. (Cookie.py also has an awkward baroque API and is missing some useful functionality we would have to build on top of it). cookielib.py is designed for client side. It's fully featured and obeys all the RFC's. It would be great to use however it's tightly coupled with another core library, urllib2.py. The http request and response objects must be urllib2 objects. But we don't use urllib2, rather we use httplib because xmlrpclib uses httplib. I don't see a reason why a cookie library should be so tightly coupled to a protocol library, but it is and that means we can't use it (I tried to just pick some isolated entrypoints for our use but I kept hitting interaction/dependency problems). I decided to solve the cookie library problems by writing a minimal cookie library that does what we need and no more than that. It is a new module in ipapython shared by both client and server and comes with a new unit test. The module has plenty of documentation, no need to repeat it here. Request URL issues: We also had problems in rpc.py whereby information from the request which is needed when we process the response is not available. Most important was the requesting URL. It turns out that the way the class and object relationships are structured it's impossible to get this information. Someone else must have run into the same issue because there was a routine called reconstruct_url() which attempted to recreate the request URL from other available information. Unfortunately reconstruct_url() was not callable from inside the response handler. So I decided to store the information in the thread context and when the request is received extract it from the thread context. It's perhaps not an ideal solution but we do similar things elsewhere so at least it's consistent. I removed the reconstruct_url() function because the exact information is now in the context and trying to apply heuristics to recreate the url is probably not robust. Ticket https://fedorahosted.org/freeipa/ticket/3022
2012-12-04 17:20:17 -06:00
cookies = Cookie.parse(s)
# Invalid single token that's keyword
s = 'HttpOnly'
with pytest.raises(ValueError):
Compliant client side session cookie behavior In summary this patch does: * Follow the defined rules for cookies when: - receiving a cookie (process the attributes) - storing a cookie (store cookie + attributes) - sending a cookie + validate the cookie domain against the request URL + validate the cookie path against the request URL + validate the cookie expiration + if valid then send only the cookie, no attribtues * Modifies how a request URL is stored during a XMLRPC request/response sequence. * Refactors a bit of the request/response logic to allow for making the decision whether to send a session cookie instead of full Kerberous auth easier. * The server now includes expiration information in the session cookie it sends to the client. The server always had the information available to prevent using an expired session cookie. Now that expiration timestamp is returned to the client as well and now the client will not send an expired session cookie back to the server. * Adds a new module and unit test for cookies (see below) Formerly we were always returning the session cookie no matter what the domain or path was in the URL. We were also sending the cookie attributes which are for the client only (used to determine if to return a cookie). The attributes are not meant to be sent to the server and the previous behavior was a protocol violation. We also were not checking the cookie expiration. Cookie library issues: We need a library to create, parse, manipulate and format cookies both in a client context and a server context. Core Python has two cookie libraries, Cookie.py and cookielib.py. Why did we add a new cookie module instead of using either of these two core Python libaries? Cookie.py is designed for server side generation but can be used to parse cookies on the client. It's the library we were using in the server. However when I tried to use it in the client I discovered it has some serious bugs. There are 7 defined cookie elements, it fails to correctly parse 3 of the 7 elements which makes it unusable because we depend on those elements. Since Cookie.py was designed for server side cookie processing it's not hard to understand how fails to correctly parse a cookie because that's a client side need. (Cookie.py also has an awkward baroque API and is missing some useful functionality we would have to build on top of it). cookielib.py is designed for client side. It's fully featured and obeys all the RFC's. It would be great to use however it's tightly coupled with another core library, urllib2.py. The http request and response objects must be urllib2 objects. But we don't use urllib2, rather we use httplib because xmlrpclib uses httplib. I don't see a reason why a cookie library should be so tightly coupled to a protocol library, but it is and that means we can't use it (I tried to just pick some isolated entrypoints for our use but I kept hitting interaction/dependency problems). I decided to solve the cookie library problems by writing a minimal cookie library that does what we need and no more than that. It is a new module in ipapython shared by both client and server and comes with a new unit test. The module has plenty of documentation, no need to repeat it here. Request URL issues: We also had problems in rpc.py whereby information from the request which is needed when we process the response is not available. Most important was the requesting URL. It turns out that the way the class and object relationships are structured it's impossible to get this information. Someone else must have run into the same issue because there was a routine called reconstruct_url() which attempted to recreate the request URL from other available information. Unfortunately reconstruct_url() was not callable from inside the response handler. So I decided to store the information in the thread context and when the request is received extract it from the thread context. It's perhaps not an ideal solution but we do similar things elsewhere so at least it's consistent. I removed the reconstruct_url() function because the exact information is now in the context and trying to apply heuristics to recreate the url is probably not robust. Ticket https://fedorahosted.org/freeipa/ticket/3022
2012-12-04 17:20:17 -06:00
cookies = Cookie.parse(s)
# Invalid key/value pair whose key is a keyword
s = 'domain=example.com'
with pytest.raises(ValueError):
Compliant client side session cookie behavior In summary this patch does: * Follow the defined rules for cookies when: - receiving a cookie (process the attributes) - storing a cookie (store cookie + attributes) - sending a cookie + validate the cookie domain against the request URL + validate the cookie path against the request URL + validate the cookie expiration + if valid then send only the cookie, no attribtues * Modifies how a request URL is stored during a XMLRPC request/response sequence. * Refactors a bit of the request/response logic to allow for making the decision whether to send a session cookie instead of full Kerberous auth easier. * The server now includes expiration information in the session cookie it sends to the client. The server always had the information available to prevent using an expired session cookie. Now that expiration timestamp is returned to the client as well and now the client will not send an expired session cookie back to the server. * Adds a new module and unit test for cookies (see below) Formerly we were always returning the session cookie no matter what the domain or path was in the URL. We were also sending the cookie attributes which are for the client only (used to determine if to return a cookie). The attributes are not meant to be sent to the server and the previous behavior was a protocol violation. We also were not checking the cookie expiration. Cookie library issues: We need a library to create, parse, manipulate and format cookies both in a client context and a server context. Core Python has two cookie libraries, Cookie.py and cookielib.py. Why did we add a new cookie module instead of using either of these two core Python libaries? Cookie.py is designed for server side generation but can be used to parse cookies on the client. It's the library we were using in the server. However when I tried to use it in the client I discovered it has some serious bugs. There are 7 defined cookie elements, it fails to correctly parse 3 of the 7 elements which makes it unusable because we depend on those elements. Since Cookie.py was designed for server side cookie processing it's not hard to understand how fails to correctly parse a cookie because that's a client side need. (Cookie.py also has an awkward baroque API and is missing some useful functionality we would have to build on top of it). cookielib.py is designed for client side. It's fully featured and obeys all the RFC's. It would be great to use however it's tightly coupled with another core library, urllib2.py. The http request and response objects must be urllib2 objects. But we don't use urllib2, rather we use httplib because xmlrpclib uses httplib. I don't see a reason why a cookie library should be so tightly coupled to a protocol library, but it is and that means we can't use it (I tried to just pick some isolated entrypoints for our use but I kept hitting interaction/dependency problems). I decided to solve the cookie library problems by writing a minimal cookie library that does what we need and no more than that. It is a new module in ipapython shared by both client and server and comes with a new unit test. The module has plenty of documentation, no need to repeat it here. Request URL issues: We also had problems in rpc.py whereby information from the request which is needed when we process the response is not available. Most important was the requesting URL. It turns out that the way the class and object relationships are structured it's impossible to get this information. Someone else must have run into the same issue because there was a routine called reconstruct_url() which attempted to recreate the request URL from other available information. Unfortunately reconstruct_url() was not callable from inside the response handler. So I decided to store the information in the thread context and when the request is received extract it from the thread context. It's perhaps not an ideal solution but we do similar things elsewhere so at least it's consistent. I removed the reconstruct_url() function because the exact information is now in the context and trying to apply heuristics to recreate the url is probably not robust. Ticket https://fedorahosted.org/freeipa/ticket/3022
2012-12-04 17:20:17 -06:00
cookies = Cookie.parse(s)
# 1 cookie with empty value
s = 'color='
cookies = Cookie.parse(s)
assert len(cookies) == 1
cookie = cookies[0]
assert cookie.key == 'color'
assert cookie.value == ''
assert cookie.domain is None
assert cookie.path is None
assert cookie.max_age is None
assert cookie.expires is None
assert cookie.secure is None
assert cookie.httponly is None
assert str(cookie) == "color="
assert cookie.http_cookie() == "color=;"
Compliant client side session cookie behavior In summary this patch does: * Follow the defined rules for cookies when: - receiving a cookie (process the attributes) - storing a cookie (store cookie + attributes) - sending a cookie + validate the cookie domain against the request URL + validate the cookie path against the request URL + validate the cookie expiration + if valid then send only the cookie, no attribtues * Modifies how a request URL is stored during a XMLRPC request/response sequence. * Refactors a bit of the request/response logic to allow for making the decision whether to send a session cookie instead of full Kerberous auth easier. * The server now includes expiration information in the session cookie it sends to the client. The server always had the information available to prevent using an expired session cookie. Now that expiration timestamp is returned to the client as well and now the client will not send an expired session cookie back to the server. * Adds a new module and unit test for cookies (see below) Formerly we were always returning the session cookie no matter what the domain or path was in the URL. We were also sending the cookie attributes which are for the client only (used to determine if to return a cookie). The attributes are not meant to be sent to the server and the previous behavior was a protocol violation. We also were not checking the cookie expiration. Cookie library issues: We need a library to create, parse, manipulate and format cookies both in a client context and a server context. Core Python has two cookie libraries, Cookie.py and cookielib.py. Why did we add a new cookie module instead of using either of these two core Python libaries? Cookie.py is designed for server side generation but can be used to parse cookies on the client. It's the library we were using in the server. However when I tried to use it in the client I discovered it has some serious bugs. There are 7 defined cookie elements, it fails to correctly parse 3 of the 7 elements which makes it unusable because we depend on those elements. Since Cookie.py was designed for server side cookie processing it's not hard to understand how fails to correctly parse a cookie because that's a client side need. (Cookie.py also has an awkward baroque API and is missing some useful functionality we would have to build on top of it). cookielib.py is designed for client side. It's fully featured and obeys all the RFC's. It would be great to use however it's tightly coupled with another core library, urllib2.py. The http request and response objects must be urllib2 objects. But we don't use urllib2, rather we use httplib because xmlrpclib uses httplib. I don't see a reason why a cookie library should be so tightly coupled to a protocol library, but it is and that means we can't use it (I tried to just pick some isolated entrypoints for our use but I kept hitting interaction/dependency problems). I decided to solve the cookie library problems by writing a minimal cookie library that does what we need and no more than that. It is a new module in ipapython shared by both client and server and comes with a new unit test. The module has plenty of documentation, no need to repeat it here. Request URL issues: We also had problems in rpc.py whereby information from the request which is needed when we process the response is not available. Most important was the requesting URL. It turns out that the way the class and object relationships are structured it's impossible to get this information. Someone else must have run into the same issue because there was a routine called reconstruct_url() which attempted to recreate the request URL from other available information. Unfortunately reconstruct_url() was not callable from inside the response handler. So I decided to store the information in the thread context and when the request is received extract it from the thread context. It's perhaps not an ideal solution but we do similar things elsewhere so at least it's consistent. I removed the reconstruct_url() function because the exact information is now in the context and trying to apply heuristics to recreate the url is probably not robust. Ticket https://fedorahosted.org/freeipa/ticket/3022
2012-12-04 17:20:17 -06:00
# 1 cookie with name/value
s = 'color=blue'
cookies = Cookie.parse(s)
assert len(cookies) == 1
Compliant client side session cookie behavior In summary this patch does: * Follow the defined rules for cookies when: - receiving a cookie (process the attributes) - storing a cookie (store cookie + attributes) - sending a cookie + validate the cookie domain against the request URL + validate the cookie path against the request URL + validate the cookie expiration + if valid then send only the cookie, no attribtues * Modifies how a request URL is stored during a XMLRPC request/response sequence. * Refactors a bit of the request/response logic to allow for making the decision whether to send a session cookie instead of full Kerberous auth easier. * The server now includes expiration information in the session cookie it sends to the client. The server always had the information available to prevent using an expired session cookie. Now that expiration timestamp is returned to the client as well and now the client will not send an expired session cookie back to the server. * Adds a new module and unit test for cookies (see below) Formerly we were always returning the session cookie no matter what the domain or path was in the URL. We were also sending the cookie attributes which are for the client only (used to determine if to return a cookie). The attributes are not meant to be sent to the server and the previous behavior was a protocol violation. We also were not checking the cookie expiration. Cookie library issues: We need a library to create, parse, manipulate and format cookies both in a client context and a server context. Core Python has two cookie libraries, Cookie.py and cookielib.py. Why did we add a new cookie module instead of using either of these two core Python libaries? Cookie.py is designed for server side generation but can be used to parse cookies on the client. It's the library we were using in the server. However when I tried to use it in the client I discovered it has some serious bugs. There are 7 defined cookie elements, it fails to correctly parse 3 of the 7 elements which makes it unusable because we depend on those elements. Since Cookie.py was designed for server side cookie processing it's not hard to understand how fails to correctly parse a cookie because that's a client side need. (Cookie.py also has an awkward baroque API and is missing some useful functionality we would have to build on top of it). cookielib.py is designed for client side. It's fully featured and obeys all the RFC's. It would be great to use however it's tightly coupled with another core library, urllib2.py. The http request and response objects must be urllib2 objects. But we don't use urllib2, rather we use httplib because xmlrpclib uses httplib. I don't see a reason why a cookie library should be so tightly coupled to a protocol library, but it is and that means we can't use it (I tried to just pick some isolated entrypoints for our use but I kept hitting interaction/dependency problems). I decided to solve the cookie library problems by writing a minimal cookie library that does what we need and no more than that. It is a new module in ipapython shared by both client and server and comes with a new unit test. The module has plenty of documentation, no need to repeat it here. Request URL issues: We also had problems in rpc.py whereby information from the request which is needed when we process the response is not available. Most important was the requesting URL. It turns out that the way the class and object relationships are structured it's impossible to get this information. Someone else must have run into the same issue because there was a routine called reconstruct_url() which attempted to recreate the request URL from other available information. Unfortunately reconstruct_url() was not callable from inside the response handler. So I decided to store the information in the thread context and when the request is received extract it from the thread context. It's perhaps not an ideal solution but we do similar things elsewhere so at least it's consistent. I removed the reconstruct_url() function because the exact information is now in the context and trying to apply heuristics to recreate the url is probably not robust. Ticket https://fedorahosted.org/freeipa/ticket/3022
2012-12-04 17:20:17 -06:00
cookie = cookies[0]
assert cookie.key == 'color'
assert cookie.value == 'blue'
assert cookie.domain is None
assert cookie.path is None
assert cookie.max_age is None
assert cookie.expires is None
assert cookie.secure is None
assert cookie.httponly is None
assert str(cookie) == "color=blue"
assert cookie.http_cookie() == "color=blue;"
Compliant client side session cookie behavior In summary this patch does: * Follow the defined rules for cookies when: - receiving a cookie (process the attributes) - storing a cookie (store cookie + attributes) - sending a cookie + validate the cookie domain against the request URL + validate the cookie path against the request URL + validate the cookie expiration + if valid then send only the cookie, no attribtues * Modifies how a request URL is stored during a XMLRPC request/response sequence. * Refactors a bit of the request/response logic to allow for making the decision whether to send a session cookie instead of full Kerberous auth easier. * The server now includes expiration information in the session cookie it sends to the client. The server always had the information available to prevent using an expired session cookie. Now that expiration timestamp is returned to the client as well and now the client will not send an expired session cookie back to the server. * Adds a new module and unit test for cookies (see below) Formerly we were always returning the session cookie no matter what the domain or path was in the URL. We were also sending the cookie attributes which are for the client only (used to determine if to return a cookie). The attributes are not meant to be sent to the server and the previous behavior was a protocol violation. We also were not checking the cookie expiration. Cookie library issues: We need a library to create, parse, manipulate and format cookies both in a client context and a server context. Core Python has two cookie libraries, Cookie.py and cookielib.py. Why did we add a new cookie module instead of using either of these two core Python libaries? Cookie.py is designed for server side generation but can be used to parse cookies on the client. It's the library we were using in the server. However when I tried to use it in the client I discovered it has some serious bugs. There are 7 defined cookie elements, it fails to correctly parse 3 of the 7 elements which makes it unusable because we depend on those elements. Since Cookie.py was designed for server side cookie processing it's not hard to understand how fails to correctly parse a cookie because that's a client side need. (Cookie.py also has an awkward baroque API and is missing some useful functionality we would have to build on top of it). cookielib.py is designed for client side. It's fully featured and obeys all the RFC's. It would be great to use however it's tightly coupled with another core library, urllib2.py. The http request and response objects must be urllib2 objects. But we don't use urllib2, rather we use httplib because xmlrpclib uses httplib. I don't see a reason why a cookie library should be so tightly coupled to a protocol library, but it is and that means we can't use it (I tried to just pick some isolated entrypoints for our use but I kept hitting interaction/dependency problems). I decided to solve the cookie library problems by writing a minimal cookie library that does what we need and no more than that. It is a new module in ipapython shared by both client and server and comes with a new unit test. The module has plenty of documentation, no need to repeat it here. Request URL issues: We also had problems in rpc.py whereby information from the request which is needed when we process the response is not available. Most important was the requesting URL. It turns out that the way the class and object relationships are structured it's impossible to get this information. Someone else must have run into the same issue because there was a routine called reconstruct_url() which attempted to recreate the request URL from other available information. Unfortunately reconstruct_url() was not callable from inside the response handler. So I decided to store the information in the thread context and when the request is received extract it from the thread context. It's perhaps not an ideal solution but we do similar things elsewhere so at least it's consistent. I removed the reconstruct_url() function because the exact information is now in the context and trying to apply heuristics to recreate the url is probably not robust. Ticket https://fedorahosted.org/freeipa/ticket/3022
2012-12-04 17:20:17 -06:00
# 1 cookie with whose value is quoted
# Use "get by name" utility to extract specific cookie
s = 'color="blue"'
cookie = Cookie.get_named_cookie_from_string(s, 'color')
assert cookie is not None, Cookie
assert cookie.key == 'color'
assert cookie.value == 'blue'
assert cookie.domain is None
assert cookie.path is None
assert cookie.max_age is None
assert cookie.expires is None
assert cookie.secure is None
assert cookie.httponly is None
assert str(cookie) == "color=blue"
assert cookie.http_cookie() == "color=blue;"
Compliant client side session cookie behavior In summary this patch does: * Follow the defined rules for cookies when: - receiving a cookie (process the attributes) - storing a cookie (store cookie + attributes) - sending a cookie + validate the cookie domain against the request URL + validate the cookie path against the request URL + validate the cookie expiration + if valid then send only the cookie, no attribtues * Modifies how a request URL is stored during a XMLRPC request/response sequence. * Refactors a bit of the request/response logic to allow for making the decision whether to send a session cookie instead of full Kerberous auth easier. * The server now includes expiration information in the session cookie it sends to the client. The server always had the information available to prevent using an expired session cookie. Now that expiration timestamp is returned to the client as well and now the client will not send an expired session cookie back to the server. * Adds a new module and unit test for cookies (see below) Formerly we were always returning the session cookie no matter what the domain or path was in the URL. We were also sending the cookie attributes which are for the client only (used to determine if to return a cookie). The attributes are not meant to be sent to the server and the previous behavior was a protocol violation. We also were not checking the cookie expiration. Cookie library issues: We need a library to create, parse, manipulate and format cookies both in a client context and a server context. Core Python has two cookie libraries, Cookie.py and cookielib.py. Why did we add a new cookie module instead of using either of these two core Python libaries? Cookie.py is designed for server side generation but can be used to parse cookies on the client. It's the library we were using in the server. However when I tried to use it in the client I discovered it has some serious bugs. There are 7 defined cookie elements, it fails to correctly parse 3 of the 7 elements which makes it unusable because we depend on those elements. Since Cookie.py was designed for server side cookie processing it's not hard to understand how fails to correctly parse a cookie because that's a client side need. (Cookie.py also has an awkward baroque API and is missing some useful functionality we would have to build on top of it). cookielib.py is designed for client side. It's fully featured and obeys all the RFC's. It would be great to use however it's tightly coupled with another core library, urllib2.py. The http request and response objects must be urllib2 objects. But we don't use urllib2, rather we use httplib because xmlrpclib uses httplib. I don't see a reason why a cookie library should be so tightly coupled to a protocol library, but it is and that means we can't use it (I tried to just pick some isolated entrypoints for our use but I kept hitting interaction/dependency problems). I decided to solve the cookie library problems by writing a minimal cookie library that does what we need and no more than that. It is a new module in ipapython shared by both client and server and comes with a new unit test. The module has plenty of documentation, no need to repeat it here. Request URL issues: We also had problems in rpc.py whereby information from the request which is needed when we process the response is not available. Most important was the requesting URL. It turns out that the way the class and object relationships are structured it's impossible to get this information. Someone else must have run into the same issue because there was a routine called reconstruct_url() which attempted to recreate the request URL from other available information. Unfortunately reconstruct_url() was not callable from inside the response handler. So I decided to store the information in the thread context and when the request is received extract it from the thread context. It's perhaps not an ideal solution but we do similar things elsewhere so at least it's consistent. I removed the reconstruct_url() function because the exact information is now in the context and trying to apply heuristics to recreate the url is probably not robust. Ticket https://fedorahosted.org/freeipa/ticket/3022
2012-12-04 17:20:17 -06:00
# 1 cookie with name/value and domain, path attributes.
# Change up the whitespace a bit.
s = 'color =blue; domain= example.com ; path = /toplevel '
cookies = Cookie.parse(s)
assert len(cookies) == 1
Compliant client side session cookie behavior In summary this patch does: * Follow the defined rules for cookies when: - receiving a cookie (process the attributes) - storing a cookie (store cookie + attributes) - sending a cookie + validate the cookie domain against the request URL + validate the cookie path against the request URL + validate the cookie expiration + if valid then send only the cookie, no attribtues * Modifies how a request URL is stored during a XMLRPC request/response sequence. * Refactors a bit of the request/response logic to allow for making the decision whether to send a session cookie instead of full Kerberous auth easier. * The server now includes expiration information in the session cookie it sends to the client. The server always had the information available to prevent using an expired session cookie. Now that expiration timestamp is returned to the client as well and now the client will not send an expired session cookie back to the server. * Adds a new module and unit test for cookies (see below) Formerly we were always returning the session cookie no matter what the domain or path was in the URL. We were also sending the cookie attributes which are for the client only (used to determine if to return a cookie). The attributes are not meant to be sent to the server and the previous behavior was a protocol violation. We also were not checking the cookie expiration. Cookie library issues: We need a library to create, parse, manipulate and format cookies both in a client context and a server context. Core Python has two cookie libraries, Cookie.py and cookielib.py. Why did we add a new cookie module instead of using either of these two core Python libaries? Cookie.py is designed for server side generation but can be used to parse cookies on the client. It's the library we were using in the server. However when I tried to use it in the client I discovered it has some serious bugs. There are 7 defined cookie elements, it fails to correctly parse 3 of the 7 elements which makes it unusable because we depend on those elements. Since Cookie.py was designed for server side cookie processing it's not hard to understand how fails to correctly parse a cookie because that's a client side need. (Cookie.py also has an awkward baroque API and is missing some useful functionality we would have to build on top of it). cookielib.py is designed for client side. It's fully featured and obeys all the RFC's. It would be great to use however it's tightly coupled with another core library, urllib2.py. The http request and response objects must be urllib2 objects. But we don't use urllib2, rather we use httplib because xmlrpclib uses httplib. I don't see a reason why a cookie library should be so tightly coupled to a protocol library, but it is and that means we can't use it (I tried to just pick some isolated entrypoints for our use but I kept hitting interaction/dependency problems). I decided to solve the cookie library problems by writing a minimal cookie library that does what we need and no more than that. It is a new module in ipapython shared by both client and server and comes with a new unit test. The module has plenty of documentation, no need to repeat it here. Request URL issues: We also had problems in rpc.py whereby information from the request which is needed when we process the response is not available. Most important was the requesting URL. It turns out that the way the class and object relationships are structured it's impossible to get this information. Someone else must have run into the same issue because there was a routine called reconstruct_url() which attempted to recreate the request URL from other available information. Unfortunately reconstruct_url() was not callable from inside the response handler. So I decided to store the information in the thread context and when the request is received extract it from the thread context. It's perhaps not an ideal solution but we do similar things elsewhere so at least it's consistent. I removed the reconstruct_url() function because the exact information is now in the context and trying to apply heuristics to recreate the url is probably not robust. Ticket https://fedorahosted.org/freeipa/ticket/3022
2012-12-04 17:20:17 -06:00
cookie = cookies[0]
assert cookie.key == 'color'
assert cookie.value == 'blue'
assert cookie.domain == 'example.com'
assert cookie.path == '/toplevel'
assert cookie.max_age is None
assert cookie.expires is None
assert cookie.secure is None
assert cookie.httponly is None
assert str(cookie) == "color=blue; Domain=example.com; Path=/toplevel"
assert cookie.http_cookie() == "color=blue;"
Compliant client side session cookie behavior In summary this patch does: * Follow the defined rules for cookies when: - receiving a cookie (process the attributes) - storing a cookie (store cookie + attributes) - sending a cookie + validate the cookie domain against the request URL + validate the cookie path against the request URL + validate the cookie expiration + if valid then send only the cookie, no attribtues * Modifies how a request URL is stored during a XMLRPC request/response sequence. * Refactors a bit of the request/response logic to allow for making the decision whether to send a session cookie instead of full Kerberous auth easier. * The server now includes expiration information in the session cookie it sends to the client. The server always had the information available to prevent using an expired session cookie. Now that expiration timestamp is returned to the client as well and now the client will not send an expired session cookie back to the server. * Adds a new module and unit test for cookies (see below) Formerly we were always returning the session cookie no matter what the domain or path was in the URL. We were also sending the cookie attributes which are for the client only (used to determine if to return a cookie). The attributes are not meant to be sent to the server and the previous behavior was a protocol violation. We also were not checking the cookie expiration. Cookie library issues: We need a library to create, parse, manipulate and format cookies both in a client context and a server context. Core Python has two cookie libraries, Cookie.py and cookielib.py. Why did we add a new cookie module instead of using either of these two core Python libaries? Cookie.py is designed for server side generation but can be used to parse cookies on the client. It's the library we were using in the server. However when I tried to use it in the client I discovered it has some serious bugs. There are 7 defined cookie elements, it fails to correctly parse 3 of the 7 elements which makes it unusable because we depend on those elements. Since Cookie.py was designed for server side cookie processing it's not hard to understand how fails to correctly parse a cookie because that's a client side need. (Cookie.py also has an awkward baroque API and is missing some useful functionality we would have to build on top of it). cookielib.py is designed for client side. It's fully featured and obeys all the RFC's. It would be great to use however it's tightly coupled with another core library, urllib2.py. The http request and response objects must be urllib2 objects. But we don't use urllib2, rather we use httplib because xmlrpclib uses httplib. I don't see a reason why a cookie library should be so tightly coupled to a protocol library, but it is and that means we can't use it (I tried to just pick some isolated entrypoints for our use but I kept hitting interaction/dependency problems). I decided to solve the cookie library problems by writing a minimal cookie library that does what we need and no more than that. It is a new module in ipapython shared by both client and server and comes with a new unit test. The module has plenty of documentation, no need to repeat it here. Request URL issues: We also had problems in rpc.py whereby information from the request which is needed when we process the response is not available. Most important was the requesting URL. It turns out that the way the class and object relationships are structured it's impossible to get this information. Someone else must have run into the same issue because there was a routine called reconstruct_url() which attempted to recreate the request URL from other available information. Unfortunately reconstruct_url() was not callable from inside the response handler. So I decided to store the information in the thread context and when the request is received extract it from the thread context. It's perhaps not an ideal solution but we do similar things elsewhere so at least it's consistent. I removed the reconstruct_url() function because the exact information is now in the context and trying to apply heuristics to recreate the url is probably not robust. Ticket https://fedorahosted.org/freeipa/ticket/3022
2012-12-04 17:20:17 -06:00
# 2 cookies, various attributes
s = 'color=blue; Max-Age=3600; temperature=hot; HttpOnly'
cookies = Cookie.parse(s)
assert len(cookies) == 2
Compliant client side session cookie behavior In summary this patch does: * Follow the defined rules for cookies when: - receiving a cookie (process the attributes) - storing a cookie (store cookie + attributes) - sending a cookie + validate the cookie domain against the request URL + validate the cookie path against the request URL + validate the cookie expiration + if valid then send only the cookie, no attribtues * Modifies how a request URL is stored during a XMLRPC request/response sequence. * Refactors a bit of the request/response logic to allow for making the decision whether to send a session cookie instead of full Kerberous auth easier. * The server now includes expiration information in the session cookie it sends to the client. The server always had the information available to prevent using an expired session cookie. Now that expiration timestamp is returned to the client as well and now the client will not send an expired session cookie back to the server. * Adds a new module and unit test for cookies (see below) Formerly we were always returning the session cookie no matter what the domain or path was in the URL. We were also sending the cookie attributes which are for the client only (used to determine if to return a cookie). The attributes are not meant to be sent to the server and the previous behavior was a protocol violation. We also were not checking the cookie expiration. Cookie library issues: We need a library to create, parse, manipulate and format cookies both in a client context and a server context. Core Python has two cookie libraries, Cookie.py and cookielib.py. Why did we add a new cookie module instead of using either of these two core Python libaries? Cookie.py is designed for server side generation but can be used to parse cookies on the client. It's the library we were using in the server. However when I tried to use it in the client I discovered it has some serious bugs. There are 7 defined cookie elements, it fails to correctly parse 3 of the 7 elements which makes it unusable because we depend on those elements. Since Cookie.py was designed for server side cookie processing it's not hard to understand how fails to correctly parse a cookie because that's a client side need. (Cookie.py also has an awkward baroque API and is missing some useful functionality we would have to build on top of it). cookielib.py is designed for client side. It's fully featured and obeys all the RFC's. It would be great to use however it's tightly coupled with another core library, urllib2.py. The http request and response objects must be urllib2 objects. But we don't use urllib2, rather we use httplib because xmlrpclib uses httplib. I don't see a reason why a cookie library should be so tightly coupled to a protocol library, but it is and that means we can't use it (I tried to just pick some isolated entrypoints for our use but I kept hitting interaction/dependency problems). I decided to solve the cookie library problems by writing a minimal cookie library that does what we need and no more than that. It is a new module in ipapython shared by both client and server and comes with a new unit test. The module has plenty of documentation, no need to repeat it here. Request URL issues: We also had problems in rpc.py whereby information from the request which is needed when we process the response is not available. Most important was the requesting URL. It turns out that the way the class and object relationships are structured it's impossible to get this information. Someone else must have run into the same issue because there was a routine called reconstruct_url() which attempted to recreate the request URL from other available information. Unfortunately reconstruct_url() was not callable from inside the response handler. So I decided to store the information in the thread context and when the request is received extract it from the thread context. It's perhaps not an ideal solution but we do similar things elsewhere so at least it's consistent. I removed the reconstruct_url() function because the exact information is now in the context and trying to apply heuristics to recreate the url is probably not robust. Ticket https://fedorahosted.org/freeipa/ticket/3022
2012-12-04 17:20:17 -06:00
cookie = cookies[0]
assert cookie.key == 'color'
assert cookie.value == 'blue'
assert cookie.domain is None
assert cookie.path is None
assert cookie.max_age == 3600
assert cookie.expires is None
assert cookie.secure is None
assert cookie.httponly is None
assert str(cookie) == "color=blue; Max-Age=3600"
assert cookie.http_cookie() == "color=blue;"
Compliant client side session cookie behavior In summary this patch does: * Follow the defined rules for cookies when: - receiving a cookie (process the attributes) - storing a cookie (store cookie + attributes) - sending a cookie + validate the cookie domain against the request URL + validate the cookie path against the request URL + validate the cookie expiration + if valid then send only the cookie, no attribtues * Modifies how a request URL is stored during a XMLRPC request/response sequence. * Refactors a bit of the request/response logic to allow for making the decision whether to send a session cookie instead of full Kerberous auth easier. * The server now includes expiration information in the session cookie it sends to the client. The server always had the information available to prevent using an expired session cookie. Now that expiration timestamp is returned to the client as well and now the client will not send an expired session cookie back to the server. * Adds a new module and unit test for cookies (see below) Formerly we were always returning the session cookie no matter what the domain or path was in the URL. We were also sending the cookie attributes which are for the client only (used to determine if to return a cookie). The attributes are not meant to be sent to the server and the previous behavior was a protocol violation. We also were not checking the cookie expiration. Cookie library issues: We need a library to create, parse, manipulate and format cookies both in a client context and a server context. Core Python has two cookie libraries, Cookie.py and cookielib.py. Why did we add a new cookie module instead of using either of these two core Python libaries? Cookie.py is designed for server side generation but can be used to parse cookies on the client. It's the library we were using in the server. However when I tried to use it in the client I discovered it has some serious bugs. There are 7 defined cookie elements, it fails to correctly parse 3 of the 7 elements which makes it unusable because we depend on those elements. Since Cookie.py was designed for server side cookie processing it's not hard to understand how fails to correctly parse a cookie because that's a client side need. (Cookie.py also has an awkward baroque API and is missing some useful functionality we would have to build on top of it). cookielib.py is designed for client side. It's fully featured and obeys all the RFC's. It would be great to use however it's tightly coupled with another core library, urllib2.py. The http request and response objects must be urllib2 objects. But we don't use urllib2, rather we use httplib because xmlrpclib uses httplib. I don't see a reason why a cookie library should be so tightly coupled to a protocol library, but it is and that means we can't use it (I tried to just pick some isolated entrypoints for our use but I kept hitting interaction/dependency problems). I decided to solve the cookie library problems by writing a minimal cookie library that does what we need and no more than that. It is a new module in ipapython shared by both client and server and comes with a new unit test. The module has plenty of documentation, no need to repeat it here. Request URL issues: We also had problems in rpc.py whereby information from the request which is needed when we process the response is not available. Most important was the requesting URL. It turns out that the way the class and object relationships are structured it's impossible to get this information. Someone else must have run into the same issue because there was a routine called reconstruct_url() which attempted to recreate the request URL from other available information. Unfortunately reconstruct_url() was not callable from inside the response handler. So I decided to store the information in the thread context and when the request is received extract it from the thread context. It's perhaps not an ideal solution but we do similar things elsewhere so at least it's consistent. I removed the reconstruct_url() function because the exact information is now in the context and trying to apply heuristics to recreate the url is probably not robust. Ticket https://fedorahosted.org/freeipa/ticket/3022
2012-12-04 17:20:17 -06:00
cookie = cookies[1]
assert cookie.key == 'temperature'
assert cookie.value == 'hot'
assert cookie.domain is None
assert cookie.path is None
assert cookie.max_age is None
assert cookie.expires is None
assert cookie.secure is None
assert cookie.httponly is True
assert str(cookie) == "temperature=hot; HttpOnly"
assert cookie.http_cookie() == "temperature=hot;"
class TestExpires:
@pytest.fixture(autouse=True)
def expires_setup(self):
Compliant client side session cookie behavior In summary this patch does: * Follow the defined rules for cookies when: - receiving a cookie (process the attributes) - storing a cookie (store cookie + attributes) - sending a cookie + validate the cookie domain against the request URL + validate the cookie path against the request URL + validate the cookie expiration + if valid then send only the cookie, no attribtues * Modifies how a request URL is stored during a XMLRPC request/response sequence. * Refactors a bit of the request/response logic to allow for making the decision whether to send a session cookie instead of full Kerberous auth easier. * The server now includes expiration information in the session cookie it sends to the client. The server always had the information available to prevent using an expired session cookie. Now that expiration timestamp is returned to the client as well and now the client will not send an expired session cookie back to the server. * Adds a new module and unit test for cookies (see below) Formerly we were always returning the session cookie no matter what the domain or path was in the URL. We were also sending the cookie attributes which are for the client only (used to determine if to return a cookie). The attributes are not meant to be sent to the server and the previous behavior was a protocol violation. We also were not checking the cookie expiration. Cookie library issues: We need a library to create, parse, manipulate and format cookies both in a client context and a server context. Core Python has two cookie libraries, Cookie.py and cookielib.py. Why did we add a new cookie module instead of using either of these two core Python libaries? Cookie.py is designed for server side generation but can be used to parse cookies on the client. It's the library we were using in the server. However when I tried to use it in the client I discovered it has some serious bugs. There are 7 defined cookie elements, it fails to correctly parse 3 of the 7 elements which makes it unusable because we depend on those elements. Since Cookie.py was designed for server side cookie processing it's not hard to understand how fails to correctly parse a cookie because that's a client side need. (Cookie.py also has an awkward baroque API and is missing some useful functionality we would have to build on top of it). cookielib.py is designed for client side. It's fully featured and obeys all the RFC's. It would be great to use however it's tightly coupled with another core library, urllib2.py. The http request and response objects must be urllib2 objects. But we don't use urllib2, rather we use httplib because xmlrpclib uses httplib. I don't see a reason why a cookie library should be so tightly coupled to a protocol library, but it is and that means we can't use it (I tried to just pick some isolated entrypoints for our use but I kept hitting interaction/dependency problems). I decided to solve the cookie library problems by writing a minimal cookie library that does what we need and no more than that. It is a new module in ipapython shared by both client and server and comes with a new unit test. The module has plenty of documentation, no need to repeat it here. Request URL issues: We also had problems in rpc.py whereby information from the request which is needed when we process the response is not available. Most important was the requesting URL. It turns out that the way the class and object relationships are structured it's impossible to get this information. Someone else must have run into the same issue because there was a routine called reconstruct_url() which attempted to recreate the request URL from other available information. Unfortunately reconstruct_url() was not callable from inside the response handler. So I decided to store the information in the thread context and when the request is received extract it from the thread context. It's perhaps not an ideal solution but we do similar things elsewhere so at least it's consistent. I removed the reconstruct_url() function because the exact information is now in the context and trying to apply heuristics to recreate the url is probably not robust. Ticket https://fedorahosted.org/freeipa/ticket/3022
2012-12-04 17:20:17 -06:00
# Force microseconds to zero because cookie timestamps only have second resolution
self.now = datetime.datetime.now(
tz=datetime.timezone.utc).replace(microsecond=0)
self.now_timestamp = datetime_from_utctimestamp(
self.now.utctimetuple(), units=1).timestamp()
self.now_string = email.utils.formatdate(self.now_timestamp, usegmt=True)
Compliant client side session cookie behavior In summary this patch does: * Follow the defined rules for cookies when: - receiving a cookie (process the attributes) - storing a cookie (store cookie + attributes) - sending a cookie + validate the cookie domain against the request URL + validate the cookie path against the request URL + validate the cookie expiration + if valid then send only the cookie, no attribtues * Modifies how a request URL is stored during a XMLRPC request/response sequence. * Refactors a bit of the request/response logic to allow for making the decision whether to send a session cookie instead of full Kerberous auth easier. * The server now includes expiration information in the session cookie it sends to the client. The server always had the information available to prevent using an expired session cookie. Now that expiration timestamp is returned to the client as well and now the client will not send an expired session cookie back to the server. * Adds a new module and unit test for cookies (see below) Formerly we were always returning the session cookie no matter what the domain or path was in the URL. We were also sending the cookie attributes which are for the client only (used to determine if to return a cookie). The attributes are not meant to be sent to the server and the previous behavior was a protocol violation. We also were not checking the cookie expiration. Cookie library issues: We need a library to create, parse, manipulate and format cookies both in a client context and a server context. Core Python has two cookie libraries, Cookie.py and cookielib.py. Why did we add a new cookie module instead of using either of these two core Python libaries? Cookie.py is designed for server side generation but can be used to parse cookies on the client. It's the library we were using in the server. However when I tried to use it in the client I discovered it has some serious bugs. There are 7 defined cookie elements, it fails to correctly parse 3 of the 7 elements which makes it unusable because we depend on those elements. Since Cookie.py was designed for server side cookie processing it's not hard to understand how fails to correctly parse a cookie because that's a client side need. (Cookie.py also has an awkward baroque API and is missing some useful functionality we would have to build on top of it). cookielib.py is designed for client side. It's fully featured and obeys all the RFC's. It would be great to use however it's tightly coupled with another core library, urllib2.py. The http request and response objects must be urllib2 objects. But we don't use urllib2, rather we use httplib because xmlrpclib uses httplib. I don't see a reason why a cookie library should be so tightly coupled to a protocol library, but it is and that means we can't use it (I tried to just pick some isolated entrypoints for our use but I kept hitting interaction/dependency problems). I decided to solve the cookie library problems by writing a minimal cookie library that does what we need and no more than that. It is a new module in ipapython shared by both client and server and comes with a new unit test. The module has plenty of documentation, no need to repeat it here. Request URL issues: We also had problems in rpc.py whereby information from the request which is needed when we process the response is not available. Most important was the requesting URL. It turns out that the way the class and object relationships are structured it's impossible to get this information. Someone else must have run into the same issue because there was a routine called reconstruct_url() which attempted to recreate the request URL from other available information. Unfortunately reconstruct_url() was not callable from inside the response handler. So I decided to store the information in the thread context and when the request is received extract it from the thread context. It's perhaps not an ideal solution but we do similar things elsewhere so at least it's consistent. I removed the reconstruct_url() function because the exact information is now in the context and trying to apply heuristics to recreate the url is probably not robust. Ticket https://fedorahosted.org/freeipa/ticket/3022
2012-12-04 17:20:17 -06:00
self.max_age = 3600 # 1 hour
self.age_expiration = self.now + datetime.timedelta(seconds=self.max_age)
self.age_timestamp = datetime_from_utctimestamp(
self.age_expiration.utctimetuple(), units=1).timestamp()
self.age_string = email.utils.formatdate(self.age_timestamp, usegmt=True)
Compliant client side session cookie behavior In summary this patch does: * Follow the defined rules for cookies when: - receiving a cookie (process the attributes) - storing a cookie (store cookie + attributes) - sending a cookie + validate the cookie domain against the request URL + validate the cookie path against the request URL + validate the cookie expiration + if valid then send only the cookie, no attribtues * Modifies how a request URL is stored during a XMLRPC request/response sequence. * Refactors a bit of the request/response logic to allow for making the decision whether to send a session cookie instead of full Kerberous auth easier. * The server now includes expiration information in the session cookie it sends to the client. The server always had the information available to prevent using an expired session cookie. Now that expiration timestamp is returned to the client as well and now the client will not send an expired session cookie back to the server. * Adds a new module and unit test for cookies (see below) Formerly we were always returning the session cookie no matter what the domain or path was in the URL. We were also sending the cookie attributes which are for the client only (used to determine if to return a cookie). The attributes are not meant to be sent to the server and the previous behavior was a protocol violation. We also were not checking the cookie expiration. Cookie library issues: We need a library to create, parse, manipulate and format cookies both in a client context and a server context. Core Python has two cookie libraries, Cookie.py and cookielib.py. Why did we add a new cookie module instead of using either of these two core Python libaries? Cookie.py is designed for server side generation but can be used to parse cookies on the client. It's the library we were using in the server. However when I tried to use it in the client I discovered it has some serious bugs. There are 7 defined cookie elements, it fails to correctly parse 3 of the 7 elements which makes it unusable because we depend on those elements. Since Cookie.py was designed for server side cookie processing it's not hard to understand how fails to correctly parse a cookie because that's a client side need. (Cookie.py also has an awkward baroque API and is missing some useful functionality we would have to build on top of it). cookielib.py is designed for client side. It's fully featured and obeys all the RFC's. It would be great to use however it's tightly coupled with another core library, urllib2.py. The http request and response objects must be urllib2 objects. But we don't use urllib2, rather we use httplib because xmlrpclib uses httplib. I don't see a reason why a cookie library should be so tightly coupled to a protocol library, but it is and that means we can't use it (I tried to just pick some isolated entrypoints for our use but I kept hitting interaction/dependency problems). I decided to solve the cookie library problems by writing a minimal cookie library that does what we need and no more than that. It is a new module in ipapython shared by both client and server and comes with a new unit test. The module has plenty of documentation, no need to repeat it here. Request URL issues: We also had problems in rpc.py whereby information from the request which is needed when we process the response is not available. Most important was the requesting URL. It turns out that the way the class and object relationships are structured it's impossible to get this information. Someone else must have run into the same issue because there was a routine called reconstruct_url() which attempted to recreate the request URL from other available information. Unfortunately reconstruct_url() was not callable from inside the response handler. So I decided to store the information in the thread context and when the request is received extract it from the thread context. It's perhaps not an ideal solution but we do similar things elsewhere so at least it's consistent. I removed the reconstruct_url() function because the exact information is now in the context and trying to apply heuristics to recreate the url is probably not robust. Ticket https://fedorahosted.org/freeipa/ticket/3022
2012-12-04 17:20:17 -06:00
self.expires = self.now + datetime.timedelta(days=1) # 1 day
self.expires_timestamp = datetime_from_utctimestamp(
self.expires.utctimetuple(), units=1).timestamp()
self.expires_string = email.utils.formatdate(self.expires_timestamp, usegmt=True)
Compliant client side session cookie behavior In summary this patch does: * Follow the defined rules for cookies when: - receiving a cookie (process the attributes) - storing a cookie (store cookie + attributes) - sending a cookie + validate the cookie domain against the request URL + validate the cookie path against the request URL + validate the cookie expiration + if valid then send only the cookie, no attribtues * Modifies how a request URL is stored during a XMLRPC request/response sequence. * Refactors a bit of the request/response logic to allow for making the decision whether to send a session cookie instead of full Kerberous auth easier. * The server now includes expiration information in the session cookie it sends to the client. The server always had the information available to prevent using an expired session cookie. Now that expiration timestamp is returned to the client as well and now the client will not send an expired session cookie back to the server. * Adds a new module and unit test for cookies (see below) Formerly we were always returning the session cookie no matter what the domain or path was in the URL. We were also sending the cookie attributes which are for the client only (used to determine if to return a cookie). The attributes are not meant to be sent to the server and the previous behavior was a protocol violation. We also were not checking the cookie expiration. Cookie library issues: We need a library to create, parse, manipulate and format cookies both in a client context and a server context. Core Python has two cookie libraries, Cookie.py and cookielib.py. Why did we add a new cookie module instead of using either of these two core Python libaries? Cookie.py is designed for server side generation but can be used to parse cookies on the client. It's the library we were using in the server. However when I tried to use it in the client I discovered it has some serious bugs. There are 7 defined cookie elements, it fails to correctly parse 3 of the 7 elements which makes it unusable because we depend on those elements. Since Cookie.py was designed for server side cookie processing it's not hard to understand how fails to correctly parse a cookie because that's a client side need. (Cookie.py also has an awkward baroque API and is missing some useful functionality we would have to build on top of it). cookielib.py is designed for client side. It's fully featured and obeys all the RFC's. It would be great to use however it's tightly coupled with another core library, urllib2.py. The http request and response objects must be urllib2 objects. But we don't use urllib2, rather we use httplib because xmlrpclib uses httplib. I don't see a reason why a cookie library should be so tightly coupled to a protocol library, but it is and that means we can't use it (I tried to just pick some isolated entrypoints for our use but I kept hitting interaction/dependency problems). I decided to solve the cookie library problems by writing a minimal cookie library that does what we need and no more than that. It is a new module in ipapython shared by both client and server and comes with a new unit test. The module has plenty of documentation, no need to repeat it here. Request URL issues: We also had problems in rpc.py whereby information from the request which is needed when we process the response is not available. Most important was the requesting URL. It turns out that the way the class and object relationships are structured it's impossible to get this information. Someone else must have run into the same issue because there was a routine called reconstruct_url() which attempted to recreate the request URL from other available information. Unfortunately reconstruct_url() was not callable from inside the response handler. So I decided to store the information in the thread context and when the request is received extract it from the thread context. It's perhaps not an ideal solution but we do similar things elsewhere so at least it's consistent. I removed the reconstruct_url() function because the exact information is now in the context and trying to apply heuristics to recreate the url is probably not robust. Ticket https://fedorahosted.org/freeipa/ticket/3022
2012-12-04 17:20:17 -06:00
def test_expires(self):
# 1 cookie with name/value and no Max-Age and no Expires
s = 'color=blue;'
cookies = Cookie.parse(s)
assert len(cookies) == 1
Compliant client side session cookie behavior In summary this patch does: * Follow the defined rules for cookies when: - receiving a cookie (process the attributes) - storing a cookie (store cookie + attributes) - sending a cookie + validate the cookie domain against the request URL + validate the cookie path against the request URL + validate the cookie expiration + if valid then send only the cookie, no attribtues * Modifies how a request URL is stored during a XMLRPC request/response sequence. * Refactors a bit of the request/response logic to allow for making the decision whether to send a session cookie instead of full Kerberous auth easier. * The server now includes expiration information in the session cookie it sends to the client. The server always had the information available to prevent using an expired session cookie. Now that expiration timestamp is returned to the client as well and now the client will not send an expired session cookie back to the server. * Adds a new module and unit test for cookies (see below) Formerly we were always returning the session cookie no matter what the domain or path was in the URL. We were also sending the cookie attributes which are for the client only (used to determine if to return a cookie). The attributes are not meant to be sent to the server and the previous behavior was a protocol violation. We also were not checking the cookie expiration. Cookie library issues: We need a library to create, parse, manipulate and format cookies both in a client context and a server context. Core Python has two cookie libraries, Cookie.py and cookielib.py. Why did we add a new cookie module instead of using either of these two core Python libaries? Cookie.py is designed for server side generation but can be used to parse cookies on the client. It's the library we were using in the server. However when I tried to use it in the client I discovered it has some serious bugs. There are 7 defined cookie elements, it fails to correctly parse 3 of the 7 elements which makes it unusable because we depend on those elements. Since Cookie.py was designed for server side cookie processing it's not hard to understand how fails to correctly parse a cookie because that's a client side need. (Cookie.py also has an awkward baroque API and is missing some useful functionality we would have to build on top of it). cookielib.py is designed for client side. It's fully featured and obeys all the RFC's. It would be great to use however it's tightly coupled with another core library, urllib2.py. The http request and response objects must be urllib2 objects. But we don't use urllib2, rather we use httplib because xmlrpclib uses httplib. I don't see a reason why a cookie library should be so tightly coupled to a protocol library, but it is and that means we can't use it (I tried to just pick some isolated entrypoints for our use but I kept hitting interaction/dependency problems). I decided to solve the cookie library problems by writing a minimal cookie library that does what we need and no more than that. It is a new module in ipapython shared by both client and server and comes with a new unit test. The module has plenty of documentation, no need to repeat it here. Request URL issues: We also had problems in rpc.py whereby information from the request which is needed when we process the response is not available. Most important was the requesting URL. It turns out that the way the class and object relationships are structured it's impossible to get this information. Someone else must have run into the same issue because there was a routine called reconstruct_url() which attempted to recreate the request URL from other available information. Unfortunately reconstruct_url() was not callable from inside the response handler. So I decided to store the information in the thread context and when the request is received extract it from the thread context. It's perhaps not an ideal solution but we do similar things elsewhere so at least it's consistent. I removed the reconstruct_url() function because the exact information is now in the context and trying to apply heuristics to recreate the url is probably not robust. Ticket https://fedorahosted.org/freeipa/ticket/3022
2012-12-04 17:20:17 -06:00
cookie = cookies[0]
# Force timestamp to known value
cookie.timestamp = self.now
assert cookie.key == 'color'
assert cookie.value == 'blue'
assert cookie.domain is None
assert cookie.path is None
assert cookie.max_age is None
assert cookie.expires is None
assert cookie.secure is None
assert cookie.httponly is None
assert str(cookie) == "color=blue"
assert cookie.get_expiration() is None
Compliant client side session cookie behavior In summary this patch does: * Follow the defined rules for cookies when: - receiving a cookie (process the attributes) - storing a cookie (store cookie + attributes) - sending a cookie + validate the cookie domain against the request URL + validate the cookie path against the request URL + validate the cookie expiration + if valid then send only the cookie, no attribtues * Modifies how a request URL is stored during a XMLRPC request/response sequence. * Refactors a bit of the request/response logic to allow for making the decision whether to send a session cookie instead of full Kerberous auth easier. * The server now includes expiration information in the session cookie it sends to the client. The server always had the information available to prevent using an expired session cookie. Now that expiration timestamp is returned to the client as well and now the client will not send an expired session cookie back to the server. * Adds a new module and unit test for cookies (see below) Formerly we were always returning the session cookie no matter what the domain or path was in the URL. We were also sending the cookie attributes which are for the client only (used to determine if to return a cookie). The attributes are not meant to be sent to the server and the previous behavior was a protocol violation. We also were not checking the cookie expiration. Cookie library issues: We need a library to create, parse, manipulate and format cookies both in a client context and a server context. Core Python has two cookie libraries, Cookie.py and cookielib.py. Why did we add a new cookie module instead of using either of these two core Python libaries? Cookie.py is designed for server side generation but can be used to parse cookies on the client. It's the library we were using in the server. However when I tried to use it in the client I discovered it has some serious bugs. There are 7 defined cookie elements, it fails to correctly parse 3 of the 7 elements which makes it unusable because we depend on those elements. Since Cookie.py was designed for server side cookie processing it's not hard to understand how fails to correctly parse a cookie because that's a client side need. (Cookie.py also has an awkward baroque API and is missing some useful functionality we would have to build on top of it). cookielib.py is designed for client side. It's fully featured and obeys all the RFC's. It would be great to use however it's tightly coupled with another core library, urllib2.py. The http request and response objects must be urllib2 objects. But we don't use urllib2, rather we use httplib because xmlrpclib uses httplib. I don't see a reason why a cookie library should be so tightly coupled to a protocol library, but it is and that means we can't use it (I tried to just pick some isolated entrypoints for our use but I kept hitting interaction/dependency problems). I decided to solve the cookie library problems by writing a minimal cookie library that does what we need and no more than that. It is a new module in ipapython shared by both client and server and comes with a new unit test. The module has plenty of documentation, no need to repeat it here. Request URL issues: We also had problems in rpc.py whereby information from the request which is needed when we process the response is not available. Most important was the requesting URL. It turns out that the way the class and object relationships are structured it's impossible to get this information. Someone else must have run into the same issue because there was a routine called reconstruct_url() which attempted to recreate the request URL from other available information. Unfortunately reconstruct_url() was not callable from inside the response handler. So I decided to store the information in the thread context and when the request is received extract it from the thread context. It's perhaps not an ideal solution but we do similar things elsewhere so at least it's consistent. I removed the reconstruct_url() function because the exact information is now in the context and trying to apply heuristics to recreate the url is probably not robust. Ticket https://fedorahosted.org/freeipa/ticket/3022
2012-12-04 17:20:17 -06:00
# Normalize
assert cookie.normalize_expiration() is None
assert cookie.max_age is None
assert cookie.expires is None
assert str(cookie) == "color=blue"
Compliant client side session cookie behavior In summary this patch does: * Follow the defined rules for cookies when: - receiving a cookie (process the attributes) - storing a cookie (store cookie + attributes) - sending a cookie + validate the cookie domain against the request URL + validate the cookie path against the request URL + validate the cookie expiration + if valid then send only the cookie, no attribtues * Modifies how a request URL is stored during a XMLRPC request/response sequence. * Refactors a bit of the request/response logic to allow for making the decision whether to send a session cookie instead of full Kerberous auth easier. * The server now includes expiration information in the session cookie it sends to the client. The server always had the information available to prevent using an expired session cookie. Now that expiration timestamp is returned to the client as well and now the client will not send an expired session cookie back to the server. * Adds a new module and unit test for cookies (see below) Formerly we were always returning the session cookie no matter what the domain or path was in the URL. We were also sending the cookie attributes which are for the client only (used to determine if to return a cookie). The attributes are not meant to be sent to the server and the previous behavior was a protocol violation. We also were not checking the cookie expiration. Cookie library issues: We need a library to create, parse, manipulate and format cookies both in a client context and a server context. Core Python has two cookie libraries, Cookie.py and cookielib.py. Why did we add a new cookie module instead of using either of these two core Python libaries? Cookie.py is designed for server side generation but can be used to parse cookies on the client. It's the library we were using in the server. However when I tried to use it in the client I discovered it has some serious bugs. There are 7 defined cookie elements, it fails to correctly parse 3 of the 7 elements which makes it unusable because we depend on those elements. Since Cookie.py was designed for server side cookie processing it's not hard to understand how fails to correctly parse a cookie because that's a client side need. (Cookie.py also has an awkward baroque API and is missing some useful functionality we would have to build on top of it). cookielib.py is designed for client side. It's fully featured and obeys all the RFC's. It would be great to use however it's tightly coupled with another core library, urllib2.py. The http request and response objects must be urllib2 objects. But we don't use urllib2, rather we use httplib because xmlrpclib uses httplib. I don't see a reason why a cookie library should be so tightly coupled to a protocol library, but it is and that means we can't use it (I tried to just pick some isolated entrypoints for our use but I kept hitting interaction/dependency problems). I decided to solve the cookie library problems by writing a minimal cookie library that does what we need and no more than that. It is a new module in ipapython shared by both client and server and comes with a new unit test. The module has plenty of documentation, no need to repeat it here. Request URL issues: We also had problems in rpc.py whereby information from the request which is needed when we process the response is not available. Most important was the requesting URL. It turns out that the way the class and object relationships are structured it's impossible to get this information. Someone else must have run into the same issue because there was a routine called reconstruct_url() which attempted to recreate the request URL from other available information. Unfortunately reconstruct_url() was not callable from inside the response handler. So I decided to store the information in the thread context and when the request is received extract it from the thread context. It's perhaps not an ideal solution but we do similar things elsewhere so at least it's consistent. I removed the reconstruct_url() function because the exact information is now in the context and trying to apply heuristics to recreate the url is probably not robust. Ticket https://fedorahosted.org/freeipa/ticket/3022
2012-12-04 17:20:17 -06:00
# 1 cookie with name/value and Max-Age
s = 'color=blue; max-age=%d' % (self.max_age)
cookies = Cookie.parse(s)
assert len(cookies) == 1
Compliant client side session cookie behavior In summary this patch does: * Follow the defined rules for cookies when: - receiving a cookie (process the attributes) - storing a cookie (store cookie + attributes) - sending a cookie + validate the cookie domain against the request URL + validate the cookie path against the request URL + validate the cookie expiration + if valid then send only the cookie, no attribtues * Modifies how a request URL is stored during a XMLRPC request/response sequence. * Refactors a bit of the request/response logic to allow for making the decision whether to send a session cookie instead of full Kerberous auth easier. * The server now includes expiration information in the session cookie it sends to the client. The server always had the information available to prevent using an expired session cookie. Now that expiration timestamp is returned to the client as well and now the client will not send an expired session cookie back to the server. * Adds a new module and unit test for cookies (see below) Formerly we were always returning the session cookie no matter what the domain or path was in the URL. We were also sending the cookie attributes which are for the client only (used to determine if to return a cookie). The attributes are not meant to be sent to the server and the previous behavior was a protocol violation. We also were not checking the cookie expiration. Cookie library issues: We need a library to create, parse, manipulate and format cookies both in a client context and a server context. Core Python has two cookie libraries, Cookie.py and cookielib.py. Why did we add a new cookie module instead of using either of these two core Python libaries? Cookie.py is designed for server side generation but can be used to parse cookies on the client. It's the library we were using in the server. However when I tried to use it in the client I discovered it has some serious bugs. There are 7 defined cookie elements, it fails to correctly parse 3 of the 7 elements which makes it unusable because we depend on those elements. Since Cookie.py was designed for server side cookie processing it's not hard to understand how fails to correctly parse a cookie because that's a client side need. (Cookie.py also has an awkward baroque API and is missing some useful functionality we would have to build on top of it). cookielib.py is designed for client side. It's fully featured and obeys all the RFC's. It would be great to use however it's tightly coupled with another core library, urllib2.py. The http request and response objects must be urllib2 objects. But we don't use urllib2, rather we use httplib because xmlrpclib uses httplib. I don't see a reason why a cookie library should be so tightly coupled to a protocol library, but it is and that means we can't use it (I tried to just pick some isolated entrypoints for our use but I kept hitting interaction/dependency problems). I decided to solve the cookie library problems by writing a minimal cookie library that does what we need and no more than that. It is a new module in ipapython shared by both client and server and comes with a new unit test. The module has plenty of documentation, no need to repeat it here. Request URL issues: We also had problems in rpc.py whereby information from the request which is needed when we process the response is not available. Most important was the requesting URL. It turns out that the way the class and object relationships are structured it's impossible to get this information. Someone else must have run into the same issue because there was a routine called reconstruct_url() which attempted to recreate the request URL from other available information. Unfortunately reconstruct_url() was not callable from inside the response handler. So I decided to store the information in the thread context and when the request is received extract it from the thread context. It's perhaps not an ideal solution but we do similar things elsewhere so at least it's consistent. I removed the reconstruct_url() function because the exact information is now in the context and trying to apply heuristics to recreate the url is probably not robust. Ticket https://fedorahosted.org/freeipa/ticket/3022
2012-12-04 17:20:17 -06:00
cookie = cookies[0]
# Force timestamp to known value
cookie.timestamp = self.now
assert cookie.key == 'color'
assert cookie.value == 'blue'
assert cookie.domain is None
assert cookie.path is None
assert cookie.max_age == self.max_age
assert cookie.expires is None
assert cookie.secure is None
assert cookie.httponly is None
assert str(cookie) == "color=blue; Max-Age=%d" % (self.max_age)
assert cookie.get_expiration() == self.age_expiration
Compliant client side session cookie behavior In summary this patch does: * Follow the defined rules for cookies when: - receiving a cookie (process the attributes) - storing a cookie (store cookie + attributes) - sending a cookie + validate the cookie domain against the request URL + validate the cookie path against the request URL + validate the cookie expiration + if valid then send only the cookie, no attribtues * Modifies how a request URL is stored during a XMLRPC request/response sequence. * Refactors a bit of the request/response logic to allow for making the decision whether to send a session cookie instead of full Kerberous auth easier. * The server now includes expiration information in the session cookie it sends to the client. The server always had the information available to prevent using an expired session cookie. Now that expiration timestamp is returned to the client as well and now the client will not send an expired session cookie back to the server. * Adds a new module and unit test for cookies (see below) Formerly we were always returning the session cookie no matter what the domain or path was in the URL. We were also sending the cookie attributes which are for the client only (used to determine if to return a cookie). The attributes are not meant to be sent to the server and the previous behavior was a protocol violation. We also were not checking the cookie expiration. Cookie library issues: We need a library to create, parse, manipulate and format cookies both in a client context and a server context. Core Python has two cookie libraries, Cookie.py and cookielib.py. Why did we add a new cookie module instead of using either of these two core Python libaries? Cookie.py is designed for server side generation but can be used to parse cookies on the client. It's the library we were using in the server. However when I tried to use it in the client I discovered it has some serious bugs. There are 7 defined cookie elements, it fails to correctly parse 3 of the 7 elements which makes it unusable because we depend on those elements. Since Cookie.py was designed for server side cookie processing it's not hard to understand how fails to correctly parse a cookie because that's a client side need. (Cookie.py also has an awkward baroque API and is missing some useful functionality we would have to build on top of it). cookielib.py is designed for client side. It's fully featured and obeys all the RFC's. It would be great to use however it's tightly coupled with another core library, urllib2.py. The http request and response objects must be urllib2 objects. But we don't use urllib2, rather we use httplib because xmlrpclib uses httplib. I don't see a reason why a cookie library should be so tightly coupled to a protocol library, but it is and that means we can't use it (I tried to just pick some isolated entrypoints for our use but I kept hitting interaction/dependency problems). I decided to solve the cookie library problems by writing a minimal cookie library that does what we need and no more than that. It is a new module in ipapython shared by both client and server and comes with a new unit test. The module has plenty of documentation, no need to repeat it here. Request URL issues: We also had problems in rpc.py whereby information from the request which is needed when we process the response is not available. Most important was the requesting URL. It turns out that the way the class and object relationships are structured it's impossible to get this information. Someone else must have run into the same issue because there was a routine called reconstruct_url() which attempted to recreate the request URL from other available information. Unfortunately reconstruct_url() was not callable from inside the response handler. So I decided to store the information in the thread context and when the request is received extract it from the thread context. It's perhaps not an ideal solution but we do similar things elsewhere so at least it's consistent. I removed the reconstruct_url() function because the exact information is now in the context and trying to apply heuristics to recreate the url is probably not robust. Ticket https://fedorahosted.org/freeipa/ticket/3022
2012-12-04 17:20:17 -06:00
# Normalize
assert cookie.normalize_expiration() == self.age_expiration
assert cookie.max_age is None
assert cookie.expires == self.age_expiration
assert str(cookie) == "color=blue; Expires=%s" % (self.age_string)
Compliant client side session cookie behavior In summary this patch does: * Follow the defined rules for cookies when: - receiving a cookie (process the attributes) - storing a cookie (store cookie + attributes) - sending a cookie + validate the cookie domain against the request URL + validate the cookie path against the request URL + validate the cookie expiration + if valid then send only the cookie, no attribtues * Modifies how a request URL is stored during a XMLRPC request/response sequence. * Refactors a bit of the request/response logic to allow for making the decision whether to send a session cookie instead of full Kerberous auth easier. * The server now includes expiration information in the session cookie it sends to the client. The server always had the information available to prevent using an expired session cookie. Now that expiration timestamp is returned to the client as well and now the client will not send an expired session cookie back to the server. * Adds a new module and unit test for cookies (see below) Formerly we were always returning the session cookie no matter what the domain or path was in the URL. We were also sending the cookie attributes which are for the client only (used to determine if to return a cookie). The attributes are not meant to be sent to the server and the previous behavior was a protocol violation. We also were not checking the cookie expiration. Cookie library issues: We need a library to create, parse, manipulate and format cookies both in a client context and a server context. Core Python has two cookie libraries, Cookie.py and cookielib.py. Why did we add a new cookie module instead of using either of these two core Python libaries? Cookie.py is designed for server side generation but can be used to parse cookies on the client. It's the library we were using in the server. However when I tried to use it in the client I discovered it has some serious bugs. There are 7 defined cookie elements, it fails to correctly parse 3 of the 7 elements which makes it unusable because we depend on those elements. Since Cookie.py was designed for server side cookie processing it's not hard to understand how fails to correctly parse a cookie because that's a client side need. (Cookie.py also has an awkward baroque API and is missing some useful functionality we would have to build on top of it). cookielib.py is designed for client side. It's fully featured and obeys all the RFC's. It would be great to use however it's tightly coupled with another core library, urllib2.py. The http request and response objects must be urllib2 objects. But we don't use urllib2, rather we use httplib because xmlrpclib uses httplib. I don't see a reason why a cookie library should be so tightly coupled to a protocol library, but it is and that means we can't use it (I tried to just pick some isolated entrypoints for our use but I kept hitting interaction/dependency problems). I decided to solve the cookie library problems by writing a minimal cookie library that does what we need and no more than that. It is a new module in ipapython shared by both client and server and comes with a new unit test. The module has plenty of documentation, no need to repeat it here. Request URL issues: We also had problems in rpc.py whereby information from the request which is needed when we process the response is not available. Most important was the requesting URL. It turns out that the way the class and object relationships are structured it's impossible to get this information. Someone else must have run into the same issue because there was a routine called reconstruct_url() which attempted to recreate the request URL from other available information. Unfortunately reconstruct_url() was not callable from inside the response handler. So I decided to store the information in the thread context and when the request is received extract it from the thread context. It's perhaps not an ideal solution but we do similar things elsewhere so at least it's consistent. I removed the reconstruct_url() function because the exact information is now in the context and trying to apply heuristics to recreate the url is probably not robust. Ticket https://fedorahosted.org/freeipa/ticket/3022
2012-12-04 17:20:17 -06:00
# 1 cookie with name/value and Expires
s = 'color=blue; Expires=%s' % (self.expires_string)
cookies = Cookie.parse(s)
assert len(cookies) == 1
Compliant client side session cookie behavior In summary this patch does: * Follow the defined rules for cookies when: - receiving a cookie (process the attributes) - storing a cookie (store cookie + attributes) - sending a cookie + validate the cookie domain against the request URL + validate the cookie path against the request URL + validate the cookie expiration + if valid then send only the cookie, no attribtues * Modifies how a request URL is stored during a XMLRPC request/response sequence. * Refactors a bit of the request/response logic to allow for making the decision whether to send a session cookie instead of full Kerberous auth easier. * The server now includes expiration information in the session cookie it sends to the client. The server always had the information available to prevent using an expired session cookie. Now that expiration timestamp is returned to the client as well and now the client will not send an expired session cookie back to the server. * Adds a new module and unit test for cookies (see below) Formerly we were always returning the session cookie no matter what the domain or path was in the URL. We were also sending the cookie attributes which are for the client only (used to determine if to return a cookie). The attributes are not meant to be sent to the server and the previous behavior was a protocol violation. We also were not checking the cookie expiration. Cookie library issues: We need a library to create, parse, manipulate and format cookies both in a client context and a server context. Core Python has two cookie libraries, Cookie.py and cookielib.py. Why did we add a new cookie module instead of using either of these two core Python libaries? Cookie.py is designed for server side generation but can be used to parse cookies on the client. It's the library we were using in the server. However when I tried to use it in the client I discovered it has some serious bugs. There are 7 defined cookie elements, it fails to correctly parse 3 of the 7 elements which makes it unusable because we depend on those elements. Since Cookie.py was designed for server side cookie processing it's not hard to understand how fails to correctly parse a cookie because that's a client side need. (Cookie.py also has an awkward baroque API and is missing some useful functionality we would have to build on top of it). cookielib.py is designed for client side. It's fully featured and obeys all the RFC's. It would be great to use however it's tightly coupled with another core library, urllib2.py. The http request and response objects must be urllib2 objects. But we don't use urllib2, rather we use httplib because xmlrpclib uses httplib. I don't see a reason why a cookie library should be so tightly coupled to a protocol library, but it is and that means we can't use it (I tried to just pick some isolated entrypoints for our use but I kept hitting interaction/dependency problems). I decided to solve the cookie library problems by writing a minimal cookie library that does what we need and no more than that. It is a new module in ipapython shared by both client and server and comes with a new unit test. The module has plenty of documentation, no need to repeat it here. Request URL issues: We also had problems in rpc.py whereby information from the request which is needed when we process the response is not available. Most important was the requesting URL. It turns out that the way the class and object relationships are structured it's impossible to get this information. Someone else must have run into the same issue because there was a routine called reconstruct_url() which attempted to recreate the request URL from other available information. Unfortunately reconstruct_url() was not callable from inside the response handler. So I decided to store the information in the thread context and when the request is received extract it from the thread context. It's perhaps not an ideal solution but we do similar things elsewhere so at least it's consistent. I removed the reconstruct_url() function because the exact information is now in the context and trying to apply heuristics to recreate the url is probably not robust. Ticket https://fedorahosted.org/freeipa/ticket/3022
2012-12-04 17:20:17 -06:00
cookie = cookies[0]
# Force timestamp to known value
cookie.timestamp = self.now
assert cookie.key == 'color'
assert cookie.value == 'blue'
assert cookie.domain is None
assert cookie.path is None
assert cookie.max_age is None
assert cookie.expires == self.expires
assert cookie.secure is None
assert cookie.httponly is None
assert str(cookie) == "color=blue; Expires=%s" % (self.expires_string)
assert cookie.get_expiration() == self.expires
Compliant client side session cookie behavior In summary this patch does: * Follow the defined rules for cookies when: - receiving a cookie (process the attributes) - storing a cookie (store cookie + attributes) - sending a cookie + validate the cookie domain against the request URL + validate the cookie path against the request URL + validate the cookie expiration + if valid then send only the cookie, no attribtues * Modifies how a request URL is stored during a XMLRPC request/response sequence. * Refactors a bit of the request/response logic to allow for making the decision whether to send a session cookie instead of full Kerberous auth easier. * The server now includes expiration information in the session cookie it sends to the client. The server always had the information available to prevent using an expired session cookie. Now that expiration timestamp is returned to the client as well and now the client will not send an expired session cookie back to the server. * Adds a new module and unit test for cookies (see below) Formerly we were always returning the session cookie no matter what the domain or path was in the URL. We were also sending the cookie attributes which are for the client only (used to determine if to return a cookie). The attributes are not meant to be sent to the server and the previous behavior was a protocol violation. We also were not checking the cookie expiration. Cookie library issues: We need a library to create, parse, manipulate and format cookies both in a client context and a server context. Core Python has two cookie libraries, Cookie.py and cookielib.py. Why did we add a new cookie module instead of using either of these two core Python libaries? Cookie.py is designed for server side generation but can be used to parse cookies on the client. It's the library we were using in the server. However when I tried to use it in the client I discovered it has some serious bugs. There are 7 defined cookie elements, it fails to correctly parse 3 of the 7 elements which makes it unusable because we depend on those elements. Since Cookie.py was designed for server side cookie processing it's not hard to understand how fails to correctly parse a cookie because that's a client side need. (Cookie.py also has an awkward baroque API and is missing some useful functionality we would have to build on top of it). cookielib.py is designed for client side. It's fully featured and obeys all the RFC's. It would be great to use however it's tightly coupled with another core library, urllib2.py. The http request and response objects must be urllib2 objects. But we don't use urllib2, rather we use httplib because xmlrpclib uses httplib. I don't see a reason why a cookie library should be so tightly coupled to a protocol library, but it is and that means we can't use it (I tried to just pick some isolated entrypoints for our use but I kept hitting interaction/dependency problems). I decided to solve the cookie library problems by writing a minimal cookie library that does what we need and no more than that. It is a new module in ipapython shared by both client and server and comes with a new unit test. The module has plenty of documentation, no need to repeat it here. Request URL issues: We also had problems in rpc.py whereby information from the request which is needed when we process the response is not available. Most important was the requesting URL. It turns out that the way the class and object relationships are structured it's impossible to get this information. Someone else must have run into the same issue because there was a routine called reconstruct_url() which attempted to recreate the request URL from other available information. Unfortunately reconstruct_url() was not callable from inside the response handler. So I decided to store the information in the thread context and when the request is received extract it from the thread context. It's perhaps not an ideal solution but we do similar things elsewhere so at least it's consistent. I removed the reconstruct_url() function because the exact information is now in the context and trying to apply heuristics to recreate the url is probably not robust. Ticket https://fedorahosted.org/freeipa/ticket/3022
2012-12-04 17:20:17 -06:00
# Normalize
assert cookie.normalize_expiration() == self.expires
assert cookie.max_age is None
assert cookie.expires == self.expires
assert str(cookie) == "color=blue; Expires=%s" % (self.expires_string)
Compliant client side session cookie behavior In summary this patch does: * Follow the defined rules for cookies when: - receiving a cookie (process the attributes) - storing a cookie (store cookie + attributes) - sending a cookie + validate the cookie domain against the request URL + validate the cookie path against the request URL + validate the cookie expiration + if valid then send only the cookie, no attribtues * Modifies how a request URL is stored during a XMLRPC request/response sequence. * Refactors a bit of the request/response logic to allow for making the decision whether to send a session cookie instead of full Kerberous auth easier. * The server now includes expiration information in the session cookie it sends to the client. The server always had the information available to prevent using an expired session cookie. Now that expiration timestamp is returned to the client as well and now the client will not send an expired session cookie back to the server. * Adds a new module and unit test for cookies (see below) Formerly we were always returning the session cookie no matter what the domain or path was in the URL. We were also sending the cookie attributes which are for the client only (used to determine if to return a cookie). The attributes are not meant to be sent to the server and the previous behavior was a protocol violation. We also were not checking the cookie expiration. Cookie library issues: We need a library to create, parse, manipulate and format cookies both in a client context and a server context. Core Python has two cookie libraries, Cookie.py and cookielib.py. Why did we add a new cookie module instead of using either of these two core Python libaries? Cookie.py is designed for server side generation but can be used to parse cookies on the client. It's the library we were using in the server. However when I tried to use it in the client I discovered it has some serious bugs. There are 7 defined cookie elements, it fails to correctly parse 3 of the 7 elements which makes it unusable because we depend on those elements. Since Cookie.py was designed for server side cookie processing it's not hard to understand how fails to correctly parse a cookie because that's a client side need. (Cookie.py also has an awkward baroque API and is missing some useful functionality we would have to build on top of it). cookielib.py is designed for client side. It's fully featured and obeys all the RFC's. It would be great to use however it's tightly coupled with another core library, urllib2.py. The http request and response objects must be urllib2 objects. But we don't use urllib2, rather we use httplib because xmlrpclib uses httplib. I don't see a reason why a cookie library should be so tightly coupled to a protocol library, but it is and that means we can't use it (I tried to just pick some isolated entrypoints for our use but I kept hitting interaction/dependency problems). I decided to solve the cookie library problems by writing a minimal cookie library that does what we need and no more than that. It is a new module in ipapython shared by both client and server and comes with a new unit test. The module has plenty of documentation, no need to repeat it here. Request URL issues: We also had problems in rpc.py whereby information from the request which is needed when we process the response is not available. Most important was the requesting URL. It turns out that the way the class and object relationships are structured it's impossible to get this information. Someone else must have run into the same issue because there was a routine called reconstruct_url() which attempted to recreate the request URL from other available information. Unfortunately reconstruct_url() was not callable from inside the response handler. So I decided to store the information in the thread context and when the request is received extract it from the thread context. It's perhaps not an ideal solution but we do similar things elsewhere so at least it's consistent. I removed the reconstruct_url() function because the exact information is now in the context and trying to apply heuristics to recreate the url is probably not robust. Ticket https://fedorahosted.org/freeipa/ticket/3022
2012-12-04 17:20:17 -06:00
# 1 cookie with name/value witht both Max-Age and Expires, Max-Age takes precedence
s = 'color=blue; Expires=%s; max-age=%d' % (self.expires_string, self.max_age)
cookies = Cookie.parse(s)
assert len(cookies) == 1
Compliant client side session cookie behavior In summary this patch does: * Follow the defined rules for cookies when: - receiving a cookie (process the attributes) - storing a cookie (store cookie + attributes) - sending a cookie + validate the cookie domain against the request URL + validate the cookie path against the request URL + validate the cookie expiration + if valid then send only the cookie, no attribtues * Modifies how a request URL is stored during a XMLRPC request/response sequence. * Refactors a bit of the request/response logic to allow for making the decision whether to send a session cookie instead of full Kerberous auth easier. * The server now includes expiration information in the session cookie it sends to the client. The server always had the information available to prevent using an expired session cookie. Now that expiration timestamp is returned to the client as well and now the client will not send an expired session cookie back to the server. * Adds a new module and unit test for cookies (see below) Formerly we were always returning the session cookie no matter what the domain or path was in the URL. We were also sending the cookie attributes which are for the client only (used to determine if to return a cookie). The attributes are not meant to be sent to the server and the previous behavior was a protocol violation. We also were not checking the cookie expiration. Cookie library issues: We need a library to create, parse, manipulate and format cookies both in a client context and a server context. Core Python has two cookie libraries, Cookie.py and cookielib.py. Why did we add a new cookie module instead of using either of these two core Python libaries? Cookie.py is designed for server side generation but can be used to parse cookies on the client. It's the library we were using in the server. However when I tried to use it in the client I discovered it has some serious bugs. There are 7 defined cookie elements, it fails to correctly parse 3 of the 7 elements which makes it unusable because we depend on those elements. Since Cookie.py was designed for server side cookie processing it's not hard to understand how fails to correctly parse a cookie because that's a client side need. (Cookie.py also has an awkward baroque API and is missing some useful functionality we would have to build on top of it). cookielib.py is designed for client side. It's fully featured and obeys all the RFC's. It would be great to use however it's tightly coupled with another core library, urllib2.py. The http request and response objects must be urllib2 objects. But we don't use urllib2, rather we use httplib because xmlrpclib uses httplib. I don't see a reason why a cookie library should be so tightly coupled to a protocol library, but it is and that means we can't use it (I tried to just pick some isolated entrypoints for our use but I kept hitting interaction/dependency problems). I decided to solve the cookie library problems by writing a minimal cookie library that does what we need and no more than that. It is a new module in ipapython shared by both client and server and comes with a new unit test. The module has plenty of documentation, no need to repeat it here. Request URL issues: We also had problems in rpc.py whereby information from the request which is needed when we process the response is not available. Most important was the requesting URL. It turns out that the way the class and object relationships are structured it's impossible to get this information. Someone else must have run into the same issue because there was a routine called reconstruct_url() which attempted to recreate the request URL from other available information. Unfortunately reconstruct_url() was not callable from inside the response handler. So I decided to store the information in the thread context and when the request is received extract it from the thread context. It's perhaps not an ideal solution but we do similar things elsewhere so at least it's consistent. I removed the reconstruct_url() function because the exact information is now in the context and trying to apply heuristics to recreate the url is probably not robust. Ticket https://fedorahosted.org/freeipa/ticket/3022
2012-12-04 17:20:17 -06:00
cookie = cookies[0]
# Force timestamp to known value
cookie.timestamp = self.now
assert cookie.key == 'color'
assert cookie.value == 'blue'
assert cookie.domain is None
assert cookie.path is None
assert cookie.max_age == self.max_age
assert cookie.expires == self.expires
assert cookie.secure is None
assert cookie.httponly is None
expected = "color=blue; Max-Age={}; Expires={}".format(
self.max_age, self.expires_string)
assert str(cookie) == expected
assert cookie.get_expiration() == self.age_expiration
Compliant client side session cookie behavior In summary this patch does: * Follow the defined rules for cookies when: - receiving a cookie (process the attributes) - storing a cookie (store cookie + attributes) - sending a cookie + validate the cookie domain against the request URL + validate the cookie path against the request URL + validate the cookie expiration + if valid then send only the cookie, no attribtues * Modifies how a request URL is stored during a XMLRPC request/response sequence. * Refactors a bit of the request/response logic to allow for making the decision whether to send a session cookie instead of full Kerberous auth easier. * The server now includes expiration information in the session cookie it sends to the client. The server always had the information available to prevent using an expired session cookie. Now that expiration timestamp is returned to the client as well and now the client will not send an expired session cookie back to the server. * Adds a new module and unit test for cookies (see below) Formerly we were always returning the session cookie no matter what the domain or path was in the URL. We were also sending the cookie attributes which are for the client only (used to determine if to return a cookie). The attributes are not meant to be sent to the server and the previous behavior was a protocol violation. We also were not checking the cookie expiration. Cookie library issues: We need a library to create, parse, manipulate and format cookies both in a client context and a server context. Core Python has two cookie libraries, Cookie.py and cookielib.py. Why did we add a new cookie module instead of using either of these two core Python libaries? Cookie.py is designed for server side generation but can be used to parse cookies on the client. It's the library we were using in the server. However when I tried to use it in the client I discovered it has some serious bugs. There are 7 defined cookie elements, it fails to correctly parse 3 of the 7 elements which makes it unusable because we depend on those elements. Since Cookie.py was designed for server side cookie processing it's not hard to understand how fails to correctly parse a cookie because that's a client side need. (Cookie.py also has an awkward baroque API and is missing some useful functionality we would have to build on top of it). cookielib.py is designed for client side. It's fully featured and obeys all the RFC's. It would be great to use however it's tightly coupled with another core library, urllib2.py. The http request and response objects must be urllib2 objects. But we don't use urllib2, rather we use httplib because xmlrpclib uses httplib. I don't see a reason why a cookie library should be so tightly coupled to a protocol library, but it is and that means we can't use it (I tried to just pick some isolated entrypoints for our use but I kept hitting interaction/dependency problems). I decided to solve the cookie library problems by writing a minimal cookie library that does what we need and no more than that. It is a new module in ipapython shared by both client and server and comes with a new unit test. The module has plenty of documentation, no need to repeat it here. Request URL issues: We also had problems in rpc.py whereby information from the request which is needed when we process the response is not available. Most important was the requesting URL. It turns out that the way the class and object relationships are structured it's impossible to get this information. Someone else must have run into the same issue because there was a routine called reconstruct_url() which attempted to recreate the request URL from other available information. Unfortunately reconstruct_url() was not callable from inside the response handler. So I decided to store the information in the thread context and when the request is received extract it from the thread context. It's perhaps not an ideal solution but we do similar things elsewhere so at least it's consistent. I removed the reconstruct_url() function because the exact information is now in the context and trying to apply heuristics to recreate the url is probably not robust. Ticket https://fedorahosted.org/freeipa/ticket/3022
2012-12-04 17:20:17 -06:00
# Normalize
assert cookie.normalize_expiration() == self.age_expiration
assert cookie.max_age is None
assert cookie.expires == self.age_expiration
assert str(cookie) == "color=blue; Expires=%s" % (self.age_string)
Compliant client side session cookie behavior In summary this patch does: * Follow the defined rules for cookies when: - receiving a cookie (process the attributes) - storing a cookie (store cookie + attributes) - sending a cookie + validate the cookie domain against the request URL + validate the cookie path against the request URL + validate the cookie expiration + if valid then send only the cookie, no attribtues * Modifies how a request URL is stored during a XMLRPC request/response sequence. * Refactors a bit of the request/response logic to allow for making the decision whether to send a session cookie instead of full Kerberous auth easier. * The server now includes expiration information in the session cookie it sends to the client. The server always had the information available to prevent using an expired session cookie. Now that expiration timestamp is returned to the client as well and now the client will not send an expired session cookie back to the server. * Adds a new module and unit test for cookies (see below) Formerly we were always returning the session cookie no matter what the domain or path was in the URL. We were also sending the cookie attributes which are for the client only (used to determine if to return a cookie). The attributes are not meant to be sent to the server and the previous behavior was a protocol violation. We also were not checking the cookie expiration. Cookie library issues: We need a library to create, parse, manipulate and format cookies both in a client context and a server context. Core Python has two cookie libraries, Cookie.py and cookielib.py. Why did we add a new cookie module instead of using either of these two core Python libaries? Cookie.py is designed for server side generation but can be used to parse cookies on the client. It's the library we were using in the server. However when I tried to use it in the client I discovered it has some serious bugs. There are 7 defined cookie elements, it fails to correctly parse 3 of the 7 elements which makes it unusable because we depend on those elements. Since Cookie.py was designed for server side cookie processing it's not hard to understand how fails to correctly parse a cookie because that's a client side need. (Cookie.py also has an awkward baroque API and is missing some useful functionality we would have to build on top of it). cookielib.py is designed for client side. It's fully featured and obeys all the RFC's. It would be great to use however it's tightly coupled with another core library, urllib2.py. The http request and response objects must be urllib2 objects. But we don't use urllib2, rather we use httplib because xmlrpclib uses httplib. I don't see a reason why a cookie library should be so tightly coupled to a protocol library, but it is and that means we can't use it (I tried to just pick some isolated entrypoints for our use but I kept hitting interaction/dependency problems). I decided to solve the cookie library problems by writing a minimal cookie library that does what we need and no more than that. It is a new module in ipapython shared by both client and server and comes with a new unit test. The module has plenty of documentation, no need to repeat it here. Request URL issues: We also had problems in rpc.py whereby information from the request which is needed when we process the response is not available. Most important was the requesting URL. It turns out that the way the class and object relationships are structured it's impossible to get this information. Someone else must have run into the same issue because there was a routine called reconstruct_url() which attempted to recreate the request URL from other available information. Unfortunately reconstruct_url() was not callable from inside the response handler. So I decided to store the information in the thread context and when the request is received extract it from the thread context. It's perhaps not an ideal solution but we do similar things elsewhere so at least it's consistent. I removed the reconstruct_url() function because the exact information is now in the context and trying to apply heuristics to recreate the url is probably not robust. Ticket https://fedorahosted.org/freeipa/ticket/3022
2012-12-04 17:20:17 -06:00
# Verify different types can be assigned to the timestamp and
# expires attribute.
cookie = Cookie('color', 'blue')
cookie.timestamp = self.now
assert cookie.timestamp == self.now
Compliant client side session cookie behavior In summary this patch does: * Follow the defined rules for cookies when: - receiving a cookie (process the attributes) - storing a cookie (store cookie + attributes) - sending a cookie + validate the cookie domain against the request URL + validate the cookie path against the request URL + validate the cookie expiration + if valid then send only the cookie, no attribtues * Modifies how a request URL is stored during a XMLRPC request/response sequence. * Refactors a bit of the request/response logic to allow for making the decision whether to send a session cookie instead of full Kerberous auth easier. * The server now includes expiration information in the session cookie it sends to the client. The server always had the information available to prevent using an expired session cookie. Now that expiration timestamp is returned to the client as well and now the client will not send an expired session cookie back to the server. * Adds a new module and unit test for cookies (see below) Formerly we were always returning the session cookie no matter what the domain or path was in the URL. We were also sending the cookie attributes which are for the client only (used to determine if to return a cookie). The attributes are not meant to be sent to the server and the previous behavior was a protocol violation. We also were not checking the cookie expiration. Cookie library issues: We need a library to create, parse, manipulate and format cookies both in a client context and a server context. Core Python has two cookie libraries, Cookie.py and cookielib.py. Why did we add a new cookie module instead of using either of these two core Python libaries? Cookie.py is designed for server side generation but can be used to parse cookies on the client. It's the library we were using in the server. However when I tried to use it in the client I discovered it has some serious bugs. There are 7 defined cookie elements, it fails to correctly parse 3 of the 7 elements which makes it unusable because we depend on those elements. Since Cookie.py was designed for server side cookie processing it's not hard to understand how fails to correctly parse a cookie because that's a client side need. (Cookie.py also has an awkward baroque API and is missing some useful functionality we would have to build on top of it). cookielib.py is designed for client side. It's fully featured and obeys all the RFC's. It would be great to use however it's tightly coupled with another core library, urllib2.py. The http request and response objects must be urllib2 objects. But we don't use urllib2, rather we use httplib because xmlrpclib uses httplib. I don't see a reason why a cookie library should be so tightly coupled to a protocol library, but it is and that means we can't use it (I tried to just pick some isolated entrypoints for our use but I kept hitting interaction/dependency problems). I decided to solve the cookie library problems by writing a minimal cookie library that does what we need and no more than that. It is a new module in ipapython shared by both client and server and comes with a new unit test. The module has plenty of documentation, no need to repeat it here. Request URL issues: We also had problems in rpc.py whereby information from the request which is needed when we process the response is not available. Most important was the requesting URL. It turns out that the way the class and object relationships are structured it's impossible to get this information. Someone else must have run into the same issue because there was a routine called reconstruct_url() which attempted to recreate the request URL from other available information. Unfortunately reconstruct_url() was not callable from inside the response handler. So I decided to store the information in the thread context and when the request is received extract it from the thread context. It's perhaps not an ideal solution but we do similar things elsewhere so at least it's consistent. I removed the reconstruct_url() function because the exact information is now in the context and trying to apply heuristics to recreate the url is probably not robust. Ticket https://fedorahosted.org/freeipa/ticket/3022
2012-12-04 17:20:17 -06:00
cookie.timestamp = self.now_timestamp
assert cookie.timestamp == self.now
Compliant client side session cookie behavior In summary this patch does: * Follow the defined rules for cookies when: - receiving a cookie (process the attributes) - storing a cookie (store cookie + attributes) - sending a cookie + validate the cookie domain against the request URL + validate the cookie path against the request URL + validate the cookie expiration + if valid then send only the cookie, no attribtues * Modifies how a request URL is stored during a XMLRPC request/response sequence. * Refactors a bit of the request/response logic to allow for making the decision whether to send a session cookie instead of full Kerberous auth easier. * The server now includes expiration information in the session cookie it sends to the client. The server always had the information available to prevent using an expired session cookie. Now that expiration timestamp is returned to the client as well and now the client will not send an expired session cookie back to the server. * Adds a new module and unit test for cookies (see below) Formerly we were always returning the session cookie no matter what the domain or path was in the URL. We were also sending the cookie attributes which are for the client only (used to determine if to return a cookie). The attributes are not meant to be sent to the server and the previous behavior was a protocol violation. We also were not checking the cookie expiration. Cookie library issues: We need a library to create, parse, manipulate and format cookies both in a client context and a server context. Core Python has two cookie libraries, Cookie.py and cookielib.py. Why did we add a new cookie module instead of using either of these two core Python libaries? Cookie.py is designed for server side generation but can be used to parse cookies on the client. It's the library we were using in the server. However when I tried to use it in the client I discovered it has some serious bugs. There are 7 defined cookie elements, it fails to correctly parse 3 of the 7 elements which makes it unusable because we depend on those elements. Since Cookie.py was designed for server side cookie processing it's not hard to understand how fails to correctly parse a cookie because that's a client side need. (Cookie.py also has an awkward baroque API and is missing some useful functionality we would have to build on top of it). cookielib.py is designed for client side. It's fully featured and obeys all the RFC's. It would be great to use however it's tightly coupled with another core library, urllib2.py. The http request and response objects must be urllib2 objects. But we don't use urllib2, rather we use httplib because xmlrpclib uses httplib. I don't see a reason why a cookie library should be so tightly coupled to a protocol library, but it is and that means we can't use it (I tried to just pick some isolated entrypoints for our use but I kept hitting interaction/dependency problems). I decided to solve the cookie library problems by writing a minimal cookie library that does what we need and no more than that. It is a new module in ipapython shared by both client and server and comes with a new unit test. The module has plenty of documentation, no need to repeat it here. Request URL issues: We also had problems in rpc.py whereby information from the request which is needed when we process the response is not available. Most important was the requesting URL. It turns out that the way the class and object relationships are structured it's impossible to get this information. Someone else must have run into the same issue because there was a routine called reconstruct_url() which attempted to recreate the request URL from other available information. Unfortunately reconstruct_url() was not callable from inside the response handler. So I decided to store the information in the thread context and when the request is received extract it from the thread context. It's perhaps not an ideal solution but we do similar things elsewhere so at least it's consistent. I removed the reconstruct_url() function because the exact information is now in the context and trying to apply heuristics to recreate the url is probably not robust. Ticket https://fedorahosted.org/freeipa/ticket/3022
2012-12-04 17:20:17 -06:00
cookie.timestamp = self.now_string
assert cookie.timestamp == self.now
Compliant client side session cookie behavior In summary this patch does: * Follow the defined rules for cookies when: - receiving a cookie (process the attributes) - storing a cookie (store cookie + attributes) - sending a cookie + validate the cookie domain against the request URL + validate the cookie path against the request URL + validate the cookie expiration + if valid then send only the cookie, no attribtues * Modifies how a request URL is stored during a XMLRPC request/response sequence. * Refactors a bit of the request/response logic to allow for making the decision whether to send a session cookie instead of full Kerberous auth easier. * The server now includes expiration information in the session cookie it sends to the client. The server always had the information available to prevent using an expired session cookie. Now that expiration timestamp is returned to the client as well and now the client will not send an expired session cookie back to the server. * Adds a new module and unit test for cookies (see below) Formerly we were always returning the session cookie no matter what the domain or path was in the URL. We were also sending the cookie attributes which are for the client only (used to determine if to return a cookie). The attributes are not meant to be sent to the server and the previous behavior was a protocol violation. We also were not checking the cookie expiration. Cookie library issues: We need a library to create, parse, manipulate and format cookies both in a client context and a server context. Core Python has two cookie libraries, Cookie.py and cookielib.py. Why did we add a new cookie module instead of using either of these two core Python libaries? Cookie.py is designed for server side generation but can be used to parse cookies on the client. It's the library we were using in the server. However when I tried to use it in the client I discovered it has some serious bugs. There are 7 defined cookie elements, it fails to correctly parse 3 of the 7 elements which makes it unusable because we depend on those elements. Since Cookie.py was designed for server side cookie processing it's not hard to understand how fails to correctly parse a cookie because that's a client side need. (Cookie.py also has an awkward baroque API and is missing some useful functionality we would have to build on top of it). cookielib.py is designed for client side. It's fully featured and obeys all the RFC's. It would be great to use however it's tightly coupled with another core library, urllib2.py. The http request and response objects must be urllib2 objects. But we don't use urllib2, rather we use httplib because xmlrpclib uses httplib. I don't see a reason why a cookie library should be so tightly coupled to a protocol library, but it is and that means we can't use it (I tried to just pick some isolated entrypoints for our use but I kept hitting interaction/dependency problems). I decided to solve the cookie library problems by writing a minimal cookie library that does what we need and no more than that. It is a new module in ipapython shared by both client and server and comes with a new unit test. The module has plenty of documentation, no need to repeat it here. Request URL issues: We also had problems in rpc.py whereby information from the request which is needed when we process the response is not available. Most important was the requesting URL. It turns out that the way the class and object relationships are structured it's impossible to get this information. Someone else must have run into the same issue because there was a routine called reconstruct_url() which attempted to recreate the request URL from other available information. Unfortunately reconstruct_url() was not callable from inside the response handler. So I decided to store the information in the thread context and when the request is received extract it from the thread context. It's perhaps not an ideal solution but we do similar things elsewhere so at least it's consistent. I removed the reconstruct_url() function because the exact information is now in the context and trying to apply heuristics to recreate the url is probably not robust. Ticket https://fedorahosted.org/freeipa/ticket/3022
2012-12-04 17:20:17 -06:00
assert cookie.expires is None
Compliant client side session cookie behavior In summary this patch does: * Follow the defined rules for cookies when: - receiving a cookie (process the attributes) - storing a cookie (store cookie + attributes) - sending a cookie + validate the cookie domain against the request URL + validate the cookie path against the request URL + validate the cookie expiration + if valid then send only the cookie, no attribtues * Modifies how a request URL is stored during a XMLRPC request/response sequence. * Refactors a bit of the request/response logic to allow for making the decision whether to send a session cookie instead of full Kerberous auth easier. * The server now includes expiration information in the session cookie it sends to the client. The server always had the information available to prevent using an expired session cookie. Now that expiration timestamp is returned to the client as well and now the client will not send an expired session cookie back to the server. * Adds a new module and unit test for cookies (see below) Formerly we were always returning the session cookie no matter what the domain or path was in the URL. We were also sending the cookie attributes which are for the client only (used to determine if to return a cookie). The attributes are not meant to be sent to the server and the previous behavior was a protocol violation. We also were not checking the cookie expiration. Cookie library issues: We need a library to create, parse, manipulate and format cookies both in a client context and a server context. Core Python has two cookie libraries, Cookie.py and cookielib.py. Why did we add a new cookie module instead of using either of these two core Python libaries? Cookie.py is designed for server side generation but can be used to parse cookies on the client. It's the library we were using in the server. However when I tried to use it in the client I discovered it has some serious bugs. There are 7 defined cookie elements, it fails to correctly parse 3 of the 7 elements which makes it unusable because we depend on those elements. Since Cookie.py was designed for server side cookie processing it's not hard to understand how fails to correctly parse a cookie because that's a client side need. (Cookie.py also has an awkward baroque API and is missing some useful functionality we would have to build on top of it). cookielib.py is designed for client side. It's fully featured and obeys all the RFC's. It would be great to use however it's tightly coupled with another core library, urllib2.py. The http request and response objects must be urllib2 objects. But we don't use urllib2, rather we use httplib because xmlrpclib uses httplib. I don't see a reason why a cookie library should be so tightly coupled to a protocol library, but it is and that means we can't use it (I tried to just pick some isolated entrypoints for our use but I kept hitting interaction/dependency problems). I decided to solve the cookie library problems by writing a minimal cookie library that does what we need and no more than that. It is a new module in ipapython shared by both client and server and comes with a new unit test. The module has plenty of documentation, no need to repeat it here. Request URL issues: We also had problems in rpc.py whereby information from the request which is needed when we process the response is not available. Most important was the requesting URL. It turns out that the way the class and object relationships are structured it's impossible to get this information. Someone else must have run into the same issue because there was a routine called reconstruct_url() which attempted to recreate the request URL from other available information. Unfortunately reconstruct_url() was not callable from inside the response handler. So I decided to store the information in the thread context and when the request is received extract it from the thread context. It's perhaps not an ideal solution but we do similar things elsewhere so at least it's consistent. I removed the reconstruct_url() function because the exact information is now in the context and trying to apply heuristics to recreate the url is probably not robust. Ticket https://fedorahosted.org/freeipa/ticket/3022
2012-12-04 17:20:17 -06:00
cookie.expires = self.expires
assert cookie.expires == self.expires
Compliant client side session cookie behavior In summary this patch does: * Follow the defined rules for cookies when: - receiving a cookie (process the attributes) - storing a cookie (store cookie + attributes) - sending a cookie + validate the cookie domain against the request URL + validate the cookie path against the request URL + validate the cookie expiration + if valid then send only the cookie, no attribtues * Modifies how a request URL is stored during a XMLRPC request/response sequence. * Refactors a bit of the request/response logic to allow for making the decision whether to send a session cookie instead of full Kerberous auth easier. * The server now includes expiration information in the session cookie it sends to the client. The server always had the information available to prevent using an expired session cookie. Now that expiration timestamp is returned to the client as well and now the client will not send an expired session cookie back to the server. * Adds a new module and unit test for cookies (see below) Formerly we were always returning the session cookie no matter what the domain or path was in the URL. We were also sending the cookie attributes which are for the client only (used to determine if to return a cookie). The attributes are not meant to be sent to the server and the previous behavior was a protocol violation. We also were not checking the cookie expiration. Cookie library issues: We need a library to create, parse, manipulate and format cookies both in a client context and a server context. Core Python has two cookie libraries, Cookie.py and cookielib.py. Why did we add a new cookie module instead of using either of these two core Python libaries? Cookie.py is designed for server side generation but can be used to parse cookies on the client. It's the library we were using in the server. However when I tried to use it in the client I discovered it has some serious bugs. There are 7 defined cookie elements, it fails to correctly parse 3 of the 7 elements which makes it unusable because we depend on those elements. Since Cookie.py was designed for server side cookie processing it's not hard to understand how fails to correctly parse a cookie because that's a client side need. (Cookie.py also has an awkward baroque API and is missing some useful functionality we would have to build on top of it). cookielib.py is designed for client side. It's fully featured and obeys all the RFC's. It would be great to use however it's tightly coupled with another core library, urllib2.py. The http request and response objects must be urllib2 objects. But we don't use urllib2, rather we use httplib because xmlrpclib uses httplib. I don't see a reason why a cookie library should be so tightly coupled to a protocol library, but it is and that means we can't use it (I tried to just pick some isolated entrypoints for our use but I kept hitting interaction/dependency problems). I decided to solve the cookie library problems by writing a minimal cookie library that does what we need and no more than that. It is a new module in ipapython shared by both client and server and comes with a new unit test. The module has plenty of documentation, no need to repeat it here. Request URL issues: We also had problems in rpc.py whereby information from the request which is needed when we process the response is not available. Most important was the requesting URL. It turns out that the way the class and object relationships are structured it's impossible to get this information. Someone else must have run into the same issue because there was a routine called reconstruct_url() which attempted to recreate the request URL from other available information. Unfortunately reconstruct_url() was not callable from inside the response handler. So I decided to store the information in the thread context and when the request is received extract it from the thread context. It's perhaps not an ideal solution but we do similar things elsewhere so at least it's consistent. I removed the reconstruct_url() function because the exact information is now in the context and trying to apply heuristics to recreate the url is probably not robust. Ticket https://fedorahosted.org/freeipa/ticket/3022
2012-12-04 17:20:17 -06:00
cookie.expires = self.expires_timestamp
assert cookie.expires == self.expires
Compliant client side session cookie behavior In summary this patch does: * Follow the defined rules for cookies when: - receiving a cookie (process the attributes) - storing a cookie (store cookie + attributes) - sending a cookie + validate the cookie domain against the request URL + validate the cookie path against the request URL + validate the cookie expiration + if valid then send only the cookie, no attribtues * Modifies how a request URL is stored during a XMLRPC request/response sequence. * Refactors a bit of the request/response logic to allow for making the decision whether to send a session cookie instead of full Kerberous auth easier. * The server now includes expiration information in the session cookie it sends to the client. The server always had the information available to prevent using an expired session cookie. Now that expiration timestamp is returned to the client as well and now the client will not send an expired session cookie back to the server. * Adds a new module and unit test for cookies (see below) Formerly we were always returning the session cookie no matter what the domain or path was in the URL. We were also sending the cookie attributes which are for the client only (used to determine if to return a cookie). The attributes are not meant to be sent to the server and the previous behavior was a protocol violation. We also were not checking the cookie expiration. Cookie library issues: We need a library to create, parse, manipulate and format cookies both in a client context and a server context. Core Python has two cookie libraries, Cookie.py and cookielib.py. Why did we add a new cookie module instead of using either of these two core Python libaries? Cookie.py is designed for server side generation but can be used to parse cookies on the client. It's the library we were using in the server. However when I tried to use it in the client I discovered it has some serious bugs. There are 7 defined cookie elements, it fails to correctly parse 3 of the 7 elements which makes it unusable because we depend on those elements. Since Cookie.py was designed for server side cookie processing it's not hard to understand how fails to correctly parse a cookie because that's a client side need. (Cookie.py also has an awkward baroque API and is missing some useful functionality we would have to build on top of it). cookielib.py is designed for client side. It's fully featured and obeys all the RFC's. It would be great to use however it's tightly coupled with another core library, urllib2.py. The http request and response objects must be urllib2 objects. But we don't use urllib2, rather we use httplib because xmlrpclib uses httplib. I don't see a reason why a cookie library should be so tightly coupled to a protocol library, but it is and that means we can't use it (I tried to just pick some isolated entrypoints for our use but I kept hitting interaction/dependency problems). I decided to solve the cookie library problems by writing a minimal cookie library that does what we need and no more than that. It is a new module in ipapython shared by both client and server and comes with a new unit test. The module has plenty of documentation, no need to repeat it here. Request URL issues: We also had problems in rpc.py whereby information from the request which is needed when we process the response is not available. Most important was the requesting URL. It turns out that the way the class and object relationships are structured it's impossible to get this information. Someone else must have run into the same issue because there was a routine called reconstruct_url() which attempted to recreate the request URL from other available information. Unfortunately reconstruct_url() was not callable from inside the response handler. So I decided to store the information in the thread context and when the request is received extract it from the thread context. It's perhaps not an ideal solution but we do similar things elsewhere so at least it's consistent. I removed the reconstruct_url() function because the exact information is now in the context and trying to apply heuristics to recreate the url is probably not robust. Ticket https://fedorahosted.org/freeipa/ticket/3022
2012-12-04 17:20:17 -06:00
cookie.expires = self.expires_string
assert cookie.expires == self.expires
Compliant client side session cookie behavior In summary this patch does: * Follow the defined rules for cookies when: - receiving a cookie (process the attributes) - storing a cookie (store cookie + attributes) - sending a cookie + validate the cookie domain against the request URL + validate the cookie path against the request URL + validate the cookie expiration + if valid then send only the cookie, no attribtues * Modifies how a request URL is stored during a XMLRPC request/response sequence. * Refactors a bit of the request/response logic to allow for making the decision whether to send a session cookie instead of full Kerberous auth easier. * The server now includes expiration information in the session cookie it sends to the client. The server always had the information available to prevent using an expired session cookie. Now that expiration timestamp is returned to the client as well and now the client will not send an expired session cookie back to the server. * Adds a new module and unit test for cookies (see below) Formerly we were always returning the session cookie no matter what the domain or path was in the URL. We were also sending the cookie attributes which are for the client only (used to determine if to return a cookie). The attributes are not meant to be sent to the server and the previous behavior was a protocol violation. We also were not checking the cookie expiration. Cookie library issues: We need a library to create, parse, manipulate and format cookies both in a client context and a server context. Core Python has two cookie libraries, Cookie.py and cookielib.py. Why did we add a new cookie module instead of using either of these two core Python libaries? Cookie.py is designed for server side generation but can be used to parse cookies on the client. It's the library we were using in the server. However when I tried to use it in the client I discovered it has some serious bugs. There are 7 defined cookie elements, it fails to correctly parse 3 of the 7 elements which makes it unusable because we depend on those elements. Since Cookie.py was designed for server side cookie processing it's not hard to understand how fails to correctly parse a cookie because that's a client side need. (Cookie.py also has an awkward baroque API and is missing some useful functionality we would have to build on top of it). cookielib.py is designed for client side. It's fully featured and obeys all the RFC's. It would be great to use however it's tightly coupled with another core library, urllib2.py. The http request and response objects must be urllib2 objects. But we don't use urllib2, rather we use httplib because xmlrpclib uses httplib. I don't see a reason why a cookie library should be so tightly coupled to a protocol library, but it is and that means we can't use it (I tried to just pick some isolated entrypoints for our use but I kept hitting interaction/dependency problems). I decided to solve the cookie library problems by writing a minimal cookie library that does what we need and no more than that. It is a new module in ipapython shared by both client and server and comes with a new unit test. The module has plenty of documentation, no need to repeat it here. Request URL issues: We also had problems in rpc.py whereby information from the request which is needed when we process the response is not available. Most important was the requesting URL. It turns out that the way the class and object relationships are structured it's impossible to get this information. Someone else must have run into the same issue because there was a routine called reconstruct_url() which attempted to recreate the request URL from other available information. Unfortunately reconstruct_url() was not callable from inside the response handler. So I decided to store the information in the thread context and when the request is received extract it from the thread context. It's perhaps not an ideal solution but we do similar things elsewhere so at least it's consistent. I removed the reconstruct_url() function because the exact information is now in the context and trying to apply heuristics to recreate the url is probably not robust. Ticket https://fedorahosted.org/freeipa/ticket/3022
2012-12-04 17:20:17 -06:00
class TestInvalidAttributes:
Compliant client side session cookie behavior In summary this patch does: * Follow the defined rules for cookies when: - receiving a cookie (process the attributes) - storing a cookie (store cookie + attributes) - sending a cookie + validate the cookie domain against the request URL + validate the cookie path against the request URL + validate the cookie expiration + if valid then send only the cookie, no attribtues * Modifies how a request URL is stored during a XMLRPC request/response sequence. * Refactors a bit of the request/response logic to allow for making the decision whether to send a session cookie instead of full Kerberous auth easier. * The server now includes expiration information in the session cookie it sends to the client. The server always had the information available to prevent using an expired session cookie. Now that expiration timestamp is returned to the client as well and now the client will not send an expired session cookie back to the server. * Adds a new module and unit test for cookies (see below) Formerly we were always returning the session cookie no matter what the domain or path was in the URL. We were also sending the cookie attributes which are for the client only (used to determine if to return a cookie). The attributes are not meant to be sent to the server and the previous behavior was a protocol violation. We also were not checking the cookie expiration. Cookie library issues: We need a library to create, parse, manipulate and format cookies both in a client context and a server context. Core Python has two cookie libraries, Cookie.py and cookielib.py. Why did we add a new cookie module instead of using either of these two core Python libaries? Cookie.py is designed for server side generation but can be used to parse cookies on the client. It's the library we were using in the server. However when I tried to use it in the client I discovered it has some serious bugs. There are 7 defined cookie elements, it fails to correctly parse 3 of the 7 elements which makes it unusable because we depend on those elements. Since Cookie.py was designed for server side cookie processing it's not hard to understand how fails to correctly parse a cookie because that's a client side need. (Cookie.py also has an awkward baroque API and is missing some useful functionality we would have to build on top of it). cookielib.py is designed for client side. It's fully featured and obeys all the RFC's. It would be great to use however it's tightly coupled with another core library, urllib2.py. The http request and response objects must be urllib2 objects. But we don't use urllib2, rather we use httplib because xmlrpclib uses httplib. I don't see a reason why a cookie library should be so tightly coupled to a protocol library, but it is and that means we can't use it (I tried to just pick some isolated entrypoints for our use but I kept hitting interaction/dependency problems). I decided to solve the cookie library problems by writing a minimal cookie library that does what we need and no more than that. It is a new module in ipapython shared by both client and server and comes with a new unit test. The module has plenty of documentation, no need to repeat it here. Request URL issues: We also had problems in rpc.py whereby information from the request which is needed when we process the response is not available. Most important was the requesting URL. It turns out that the way the class and object relationships are structured it's impossible to get this information. Someone else must have run into the same issue because there was a routine called reconstruct_url() which attempted to recreate the request URL from other available information. Unfortunately reconstruct_url() was not callable from inside the response handler. So I decided to store the information in the thread context and when the request is received extract it from the thread context. It's perhaps not an ideal solution but we do similar things elsewhere so at least it's consistent. I removed the reconstruct_url() function because the exact information is now in the context and trying to apply heuristics to recreate the url is probably not robust. Ticket https://fedorahosted.org/freeipa/ticket/3022
2012-12-04 17:20:17 -06:00
def test_invalid(self):
# Invalid Max-Age
s = 'color=blue; Max-Age=over-the-hill'
with pytest.raises(ValueError):
Cookie.parse(s)
Compliant client side session cookie behavior In summary this patch does: * Follow the defined rules for cookies when: - receiving a cookie (process the attributes) - storing a cookie (store cookie + attributes) - sending a cookie + validate the cookie domain against the request URL + validate the cookie path against the request URL + validate the cookie expiration + if valid then send only the cookie, no attribtues * Modifies how a request URL is stored during a XMLRPC request/response sequence. * Refactors a bit of the request/response logic to allow for making the decision whether to send a session cookie instead of full Kerberous auth easier. * The server now includes expiration information in the session cookie it sends to the client. The server always had the information available to prevent using an expired session cookie. Now that expiration timestamp is returned to the client as well and now the client will not send an expired session cookie back to the server. * Adds a new module and unit test for cookies (see below) Formerly we were always returning the session cookie no matter what the domain or path was in the URL. We were also sending the cookie attributes which are for the client only (used to determine if to return a cookie). The attributes are not meant to be sent to the server and the previous behavior was a protocol violation. We also were not checking the cookie expiration. Cookie library issues: We need a library to create, parse, manipulate and format cookies both in a client context and a server context. Core Python has two cookie libraries, Cookie.py and cookielib.py. Why did we add a new cookie module instead of using either of these two core Python libaries? Cookie.py is designed for server side generation but can be used to parse cookies on the client. It's the library we were using in the server. However when I tried to use it in the client I discovered it has some serious bugs. There are 7 defined cookie elements, it fails to correctly parse 3 of the 7 elements which makes it unusable because we depend on those elements. Since Cookie.py was designed for server side cookie processing it's not hard to understand how fails to correctly parse a cookie because that's a client side need. (Cookie.py also has an awkward baroque API and is missing some useful functionality we would have to build on top of it). cookielib.py is designed for client side. It's fully featured and obeys all the RFC's. It would be great to use however it's tightly coupled with another core library, urllib2.py. The http request and response objects must be urllib2 objects. But we don't use urllib2, rather we use httplib because xmlrpclib uses httplib. I don't see a reason why a cookie library should be so tightly coupled to a protocol library, but it is and that means we can't use it (I tried to just pick some isolated entrypoints for our use but I kept hitting interaction/dependency problems). I decided to solve the cookie library problems by writing a minimal cookie library that does what we need and no more than that. It is a new module in ipapython shared by both client and server and comes with a new unit test. The module has plenty of documentation, no need to repeat it here. Request URL issues: We also had problems in rpc.py whereby information from the request which is needed when we process the response is not available. Most important was the requesting URL. It turns out that the way the class and object relationships are structured it's impossible to get this information. Someone else must have run into the same issue because there was a routine called reconstruct_url() which attempted to recreate the request URL from other available information. Unfortunately reconstruct_url() was not callable from inside the response handler. So I decided to store the information in the thread context and when the request is received extract it from the thread context. It's perhaps not an ideal solution but we do similar things elsewhere so at least it's consistent. I removed the reconstruct_url() function because the exact information is now in the context and trying to apply heuristics to recreate the url is probably not robust. Ticket https://fedorahosted.org/freeipa/ticket/3022
2012-12-04 17:20:17 -06:00
cookie = Cookie('color', 'blue')
with pytest.raises(ValueError):
Compliant client side session cookie behavior In summary this patch does: * Follow the defined rules for cookies when: - receiving a cookie (process the attributes) - storing a cookie (store cookie + attributes) - sending a cookie + validate the cookie domain against the request URL + validate the cookie path against the request URL + validate the cookie expiration + if valid then send only the cookie, no attribtues * Modifies how a request URL is stored during a XMLRPC request/response sequence. * Refactors a bit of the request/response logic to allow for making the decision whether to send a session cookie instead of full Kerberous auth easier. * The server now includes expiration information in the session cookie it sends to the client. The server always had the information available to prevent using an expired session cookie. Now that expiration timestamp is returned to the client as well and now the client will not send an expired session cookie back to the server. * Adds a new module and unit test for cookies (see below) Formerly we were always returning the session cookie no matter what the domain or path was in the URL. We were also sending the cookie attributes which are for the client only (used to determine if to return a cookie). The attributes are not meant to be sent to the server and the previous behavior was a protocol violation. We also were not checking the cookie expiration. Cookie library issues: We need a library to create, parse, manipulate and format cookies both in a client context and a server context. Core Python has two cookie libraries, Cookie.py and cookielib.py. Why did we add a new cookie module instead of using either of these two core Python libaries? Cookie.py is designed for server side generation but can be used to parse cookies on the client. It's the library we were using in the server. However when I tried to use it in the client I discovered it has some serious bugs. There are 7 defined cookie elements, it fails to correctly parse 3 of the 7 elements which makes it unusable because we depend on those elements. Since Cookie.py was designed for server side cookie processing it's not hard to understand how fails to correctly parse a cookie because that's a client side need. (Cookie.py also has an awkward baroque API and is missing some useful functionality we would have to build on top of it). cookielib.py is designed for client side. It's fully featured and obeys all the RFC's. It would be great to use however it's tightly coupled with another core library, urllib2.py. The http request and response objects must be urllib2 objects. But we don't use urllib2, rather we use httplib because xmlrpclib uses httplib. I don't see a reason why a cookie library should be so tightly coupled to a protocol library, but it is and that means we can't use it (I tried to just pick some isolated entrypoints for our use but I kept hitting interaction/dependency problems). I decided to solve the cookie library problems by writing a minimal cookie library that does what we need and no more than that. It is a new module in ipapython shared by both client and server and comes with a new unit test. The module has plenty of documentation, no need to repeat it here. Request URL issues: We also had problems in rpc.py whereby information from the request which is needed when we process the response is not available. Most important was the requesting URL. It turns out that the way the class and object relationships are structured it's impossible to get this information. Someone else must have run into the same issue because there was a routine called reconstruct_url() which attempted to recreate the request URL from other available information. Unfortunately reconstruct_url() was not callable from inside the response handler. So I decided to store the information in the thread context and when the request is received extract it from the thread context. It's perhaps not an ideal solution but we do similar things elsewhere so at least it's consistent. I removed the reconstruct_url() function because the exact information is now in the context and trying to apply heuristics to recreate the url is probably not robust. Ticket https://fedorahosted.org/freeipa/ticket/3022
2012-12-04 17:20:17 -06:00
cookie.max_age = 'over-the-hill'
# Invalid Expires
s = 'color=blue; Expires=Sun, 06 Xxx 1994 08:49:37 GMT'
with pytest.raises(ValueError):
Cookie.parse(s)
Compliant client side session cookie behavior In summary this patch does: * Follow the defined rules for cookies when: - receiving a cookie (process the attributes) - storing a cookie (store cookie + attributes) - sending a cookie + validate the cookie domain against the request URL + validate the cookie path against the request URL + validate the cookie expiration + if valid then send only the cookie, no attribtues * Modifies how a request URL is stored during a XMLRPC request/response sequence. * Refactors a bit of the request/response logic to allow for making the decision whether to send a session cookie instead of full Kerberous auth easier. * The server now includes expiration information in the session cookie it sends to the client. The server always had the information available to prevent using an expired session cookie. Now that expiration timestamp is returned to the client as well and now the client will not send an expired session cookie back to the server. * Adds a new module and unit test for cookies (see below) Formerly we were always returning the session cookie no matter what the domain or path was in the URL. We were also sending the cookie attributes which are for the client only (used to determine if to return a cookie). The attributes are not meant to be sent to the server and the previous behavior was a protocol violation. We also were not checking the cookie expiration. Cookie library issues: We need a library to create, parse, manipulate and format cookies both in a client context and a server context. Core Python has two cookie libraries, Cookie.py and cookielib.py. Why did we add a new cookie module instead of using either of these two core Python libaries? Cookie.py is designed for server side generation but can be used to parse cookies on the client. It's the library we were using in the server. However when I tried to use it in the client I discovered it has some serious bugs. There are 7 defined cookie elements, it fails to correctly parse 3 of the 7 elements which makes it unusable because we depend on those elements. Since Cookie.py was designed for server side cookie processing it's not hard to understand how fails to correctly parse a cookie because that's a client side need. (Cookie.py also has an awkward baroque API and is missing some useful functionality we would have to build on top of it). cookielib.py is designed for client side. It's fully featured and obeys all the RFC's. It would be great to use however it's tightly coupled with another core library, urllib2.py. The http request and response objects must be urllib2 objects. But we don't use urllib2, rather we use httplib because xmlrpclib uses httplib. I don't see a reason why a cookie library should be so tightly coupled to a protocol library, but it is and that means we can't use it (I tried to just pick some isolated entrypoints for our use but I kept hitting interaction/dependency problems). I decided to solve the cookie library problems by writing a minimal cookie library that does what we need and no more than that. It is a new module in ipapython shared by both client and server and comes with a new unit test. The module has plenty of documentation, no need to repeat it here. Request URL issues: We also had problems in rpc.py whereby information from the request which is needed when we process the response is not available. Most important was the requesting URL. It turns out that the way the class and object relationships are structured it's impossible to get this information. Someone else must have run into the same issue because there was a routine called reconstruct_url() which attempted to recreate the request URL from other available information. Unfortunately reconstruct_url() was not callable from inside the response handler. So I decided to store the information in the thread context and when the request is received extract it from the thread context. It's perhaps not an ideal solution but we do similar things elsewhere so at least it's consistent. I removed the reconstruct_url() function because the exact information is now in the context and trying to apply heuristics to recreate the url is probably not robust. Ticket https://fedorahosted.org/freeipa/ticket/3022
2012-12-04 17:20:17 -06:00
cookie = Cookie('color', 'blue')
with pytest.raises(ValueError):
Compliant client side session cookie behavior In summary this patch does: * Follow the defined rules for cookies when: - receiving a cookie (process the attributes) - storing a cookie (store cookie + attributes) - sending a cookie + validate the cookie domain against the request URL + validate the cookie path against the request URL + validate the cookie expiration + if valid then send only the cookie, no attribtues * Modifies how a request URL is stored during a XMLRPC request/response sequence. * Refactors a bit of the request/response logic to allow for making the decision whether to send a session cookie instead of full Kerberous auth easier. * The server now includes expiration information in the session cookie it sends to the client. The server always had the information available to prevent using an expired session cookie. Now that expiration timestamp is returned to the client as well and now the client will not send an expired session cookie back to the server. * Adds a new module and unit test for cookies (see below) Formerly we were always returning the session cookie no matter what the domain or path was in the URL. We were also sending the cookie attributes which are for the client only (used to determine if to return a cookie). The attributes are not meant to be sent to the server and the previous behavior was a protocol violation. We also were not checking the cookie expiration. Cookie library issues: We need a library to create, parse, manipulate and format cookies both in a client context and a server context. Core Python has two cookie libraries, Cookie.py and cookielib.py. Why did we add a new cookie module instead of using either of these two core Python libaries? Cookie.py is designed for server side generation but can be used to parse cookies on the client. It's the library we were using in the server. However when I tried to use it in the client I discovered it has some serious bugs. There are 7 defined cookie elements, it fails to correctly parse 3 of the 7 elements which makes it unusable because we depend on those elements. Since Cookie.py was designed for server side cookie processing it's not hard to understand how fails to correctly parse a cookie because that's a client side need. (Cookie.py also has an awkward baroque API and is missing some useful functionality we would have to build on top of it). cookielib.py is designed for client side. It's fully featured and obeys all the RFC's. It would be great to use however it's tightly coupled with another core library, urllib2.py. The http request and response objects must be urllib2 objects. But we don't use urllib2, rather we use httplib because xmlrpclib uses httplib. I don't see a reason why a cookie library should be so tightly coupled to a protocol library, but it is and that means we can't use it (I tried to just pick some isolated entrypoints for our use but I kept hitting interaction/dependency problems). I decided to solve the cookie library problems by writing a minimal cookie library that does what we need and no more than that. It is a new module in ipapython shared by both client and server and comes with a new unit test. The module has plenty of documentation, no need to repeat it here. Request URL issues: We also had problems in rpc.py whereby information from the request which is needed when we process the response is not available. Most important was the requesting URL. It turns out that the way the class and object relationships are structured it's impossible to get this information. Someone else must have run into the same issue because there was a routine called reconstruct_url() which attempted to recreate the request URL from other available information. Unfortunately reconstruct_url() was not callable from inside the response handler. So I decided to store the information in the thread context and when the request is received extract it from the thread context. It's perhaps not an ideal solution but we do similar things elsewhere so at least it's consistent. I removed the reconstruct_url() function because the exact information is now in the context and trying to apply heuristics to recreate the url is probably not robust. Ticket https://fedorahosted.org/freeipa/ticket/3022
2012-12-04 17:20:17 -06:00
cookie.expires = 'Sun, 06 Xxx 1994 08:49:37 GMT'
class TestAttributes:
Compliant client side session cookie behavior In summary this patch does: * Follow the defined rules for cookies when: - receiving a cookie (process the attributes) - storing a cookie (store cookie + attributes) - sending a cookie + validate the cookie domain against the request URL + validate the cookie path against the request URL + validate the cookie expiration + if valid then send only the cookie, no attribtues * Modifies how a request URL is stored during a XMLRPC request/response sequence. * Refactors a bit of the request/response logic to allow for making the decision whether to send a session cookie instead of full Kerberous auth easier. * The server now includes expiration information in the session cookie it sends to the client. The server always had the information available to prevent using an expired session cookie. Now that expiration timestamp is returned to the client as well and now the client will not send an expired session cookie back to the server. * Adds a new module and unit test for cookies (see below) Formerly we were always returning the session cookie no matter what the domain or path was in the URL. We were also sending the cookie attributes which are for the client only (used to determine if to return a cookie). The attributes are not meant to be sent to the server and the previous behavior was a protocol violation. We also were not checking the cookie expiration. Cookie library issues: We need a library to create, parse, manipulate and format cookies both in a client context and a server context. Core Python has two cookie libraries, Cookie.py and cookielib.py. Why did we add a new cookie module instead of using either of these two core Python libaries? Cookie.py is designed for server side generation but can be used to parse cookies on the client. It's the library we were using in the server. However when I tried to use it in the client I discovered it has some serious bugs. There are 7 defined cookie elements, it fails to correctly parse 3 of the 7 elements which makes it unusable because we depend on those elements. Since Cookie.py was designed for server side cookie processing it's not hard to understand how fails to correctly parse a cookie because that's a client side need. (Cookie.py also has an awkward baroque API and is missing some useful functionality we would have to build on top of it). cookielib.py is designed for client side. It's fully featured and obeys all the RFC's. It would be great to use however it's tightly coupled with another core library, urllib2.py. The http request and response objects must be urllib2 objects. But we don't use urllib2, rather we use httplib because xmlrpclib uses httplib. I don't see a reason why a cookie library should be so tightly coupled to a protocol library, but it is and that means we can't use it (I tried to just pick some isolated entrypoints for our use but I kept hitting interaction/dependency problems). I decided to solve the cookie library problems by writing a minimal cookie library that does what we need and no more than that. It is a new module in ipapython shared by both client and server and comes with a new unit test. The module has plenty of documentation, no need to repeat it here. Request URL issues: We also had problems in rpc.py whereby information from the request which is needed when we process the response is not available. Most important was the requesting URL. It turns out that the way the class and object relationships are structured it's impossible to get this information. Someone else must have run into the same issue because there was a routine called reconstruct_url() which attempted to recreate the request URL from other available information. Unfortunately reconstruct_url() was not callable from inside the response handler. So I decided to store the information in the thread context and when the request is received extract it from the thread context. It's perhaps not an ideal solution but we do similar things elsewhere so at least it's consistent. I removed the reconstruct_url() function because the exact information is now in the context and trying to apply heuristics to recreate the url is probably not robust. Ticket https://fedorahosted.org/freeipa/ticket/3022
2012-12-04 17:20:17 -06:00
def test_attributes(self):
cookie = Cookie('color', 'blue')
assert cookie.key == 'color'
assert cookie.value == 'blue'
assert cookie.domain is None
assert cookie.path is None
assert cookie.max_age is None
assert cookie.expires is None
assert cookie.secure is None
assert cookie.httponly is None
Compliant client side session cookie behavior In summary this patch does: * Follow the defined rules for cookies when: - receiving a cookie (process the attributes) - storing a cookie (store cookie + attributes) - sending a cookie + validate the cookie domain against the request URL + validate the cookie path against the request URL + validate the cookie expiration + if valid then send only the cookie, no attribtues * Modifies how a request URL is stored during a XMLRPC request/response sequence. * Refactors a bit of the request/response logic to allow for making the decision whether to send a session cookie instead of full Kerberous auth easier. * The server now includes expiration information in the session cookie it sends to the client. The server always had the information available to prevent using an expired session cookie. Now that expiration timestamp is returned to the client as well and now the client will not send an expired session cookie back to the server. * Adds a new module and unit test for cookies (see below) Formerly we were always returning the session cookie no matter what the domain or path was in the URL. We were also sending the cookie attributes which are for the client only (used to determine if to return a cookie). The attributes are not meant to be sent to the server and the previous behavior was a protocol violation. We also were not checking the cookie expiration. Cookie library issues: We need a library to create, parse, manipulate and format cookies both in a client context and a server context. Core Python has two cookie libraries, Cookie.py and cookielib.py. Why did we add a new cookie module instead of using either of these two core Python libaries? Cookie.py is designed for server side generation but can be used to parse cookies on the client. It's the library we were using in the server. However when I tried to use it in the client I discovered it has some serious bugs. There are 7 defined cookie elements, it fails to correctly parse 3 of the 7 elements which makes it unusable because we depend on those elements. Since Cookie.py was designed for server side cookie processing it's not hard to understand how fails to correctly parse a cookie because that's a client side need. (Cookie.py also has an awkward baroque API and is missing some useful functionality we would have to build on top of it). cookielib.py is designed for client side. It's fully featured and obeys all the RFC's. It would be great to use however it's tightly coupled with another core library, urllib2.py. The http request and response objects must be urllib2 objects. But we don't use urllib2, rather we use httplib because xmlrpclib uses httplib. I don't see a reason why a cookie library should be so tightly coupled to a protocol library, but it is and that means we can't use it (I tried to just pick some isolated entrypoints for our use but I kept hitting interaction/dependency problems). I decided to solve the cookie library problems by writing a minimal cookie library that does what we need and no more than that. It is a new module in ipapython shared by both client and server and comes with a new unit test. The module has plenty of documentation, no need to repeat it here. Request URL issues: We also had problems in rpc.py whereby information from the request which is needed when we process the response is not available. Most important was the requesting URL. It turns out that the way the class and object relationships are structured it's impossible to get this information. Someone else must have run into the same issue because there was a routine called reconstruct_url() which attempted to recreate the request URL from other available information. Unfortunately reconstruct_url() was not callable from inside the response handler. So I decided to store the information in the thread context and when the request is received extract it from the thread context. It's perhaps not an ideal solution but we do similar things elsewhere so at least it's consistent. I removed the reconstruct_url() function because the exact information is now in the context and trying to apply heuristics to recreate the url is probably not robust. Ticket https://fedorahosted.org/freeipa/ticket/3022
2012-12-04 17:20:17 -06:00
cookie.domain = 'example.com'
assert cookie.domain == 'example.com'
Compliant client side session cookie behavior In summary this patch does: * Follow the defined rules for cookies when: - receiving a cookie (process the attributes) - storing a cookie (store cookie + attributes) - sending a cookie + validate the cookie domain against the request URL + validate the cookie path against the request URL + validate the cookie expiration + if valid then send only the cookie, no attribtues * Modifies how a request URL is stored during a XMLRPC request/response sequence. * Refactors a bit of the request/response logic to allow for making the decision whether to send a session cookie instead of full Kerberous auth easier. * The server now includes expiration information in the session cookie it sends to the client. The server always had the information available to prevent using an expired session cookie. Now that expiration timestamp is returned to the client as well and now the client will not send an expired session cookie back to the server. * Adds a new module and unit test for cookies (see below) Formerly we were always returning the session cookie no matter what the domain or path was in the URL. We were also sending the cookie attributes which are for the client only (used to determine if to return a cookie). The attributes are not meant to be sent to the server and the previous behavior was a protocol violation. We also were not checking the cookie expiration. Cookie library issues: We need a library to create, parse, manipulate and format cookies both in a client context and a server context. Core Python has two cookie libraries, Cookie.py and cookielib.py. Why did we add a new cookie module instead of using either of these two core Python libaries? Cookie.py is designed for server side generation but can be used to parse cookies on the client. It's the library we were using in the server. However when I tried to use it in the client I discovered it has some serious bugs. There are 7 defined cookie elements, it fails to correctly parse 3 of the 7 elements which makes it unusable because we depend on those elements. Since Cookie.py was designed for server side cookie processing it's not hard to understand how fails to correctly parse a cookie because that's a client side need. (Cookie.py also has an awkward baroque API and is missing some useful functionality we would have to build on top of it). cookielib.py is designed for client side. It's fully featured and obeys all the RFC's. It would be great to use however it's tightly coupled with another core library, urllib2.py. The http request and response objects must be urllib2 objects. But we don't use urllib2, rather we use httplib because xmlrpclib uses httplib. I don't see a reason why a cookie library should be so tightly coupled to a protocol library, but it is and that means we can't use it (I tried to just pick some isolated entrypoints for our use but I kept hitting interaction/dependency problems). I decided to solve the cookie library problems by writing a minimal cookie library that does what we need and no more than that. It is a new module in ipapython shared by both client and server and comes with a new unit test. The module has plenty of documentation, no need to repeat it here. Request URL issues: We also had problems in rpc.py whereby information from the request which is needed when we process the response is not available. Most important was the requesting URL. It turns out that the way the class and object relationships are structured it's impossible to get this information. Someone else must have run into the same issue because there was a routine called reconstruct_url() which attempted to recreate the request URL from other available information. Unfortunately reconstruct_url() was not callable from inside the response handler. So I decided to store the information in the thread context and when the request is received extract it from the thread context. It's perhaps not an ideal solution but we do similar things elsewhere so at least it's consistent. I removed the reconstruct_url() function because the exact information is now in the context and trying to apply heuristics to recreate the url is probably not robust. Ticket https://fedorahosted.org/freeipa/ticket/3022
2012-12-04 17:20:17 -06:00
cookie.domain = None
assert cookie.domain is None
Compliant client side session cookie behavior In summary this patch does: * Follow the defined rules for cookies when: - receiving a cookie (process the attributes) - storing a cookie (store cookie + attributes) - sending a cookie + validate the cookie domain against the request URL + validate the cookie path against the request URL + validate the cookie expiration + if valid then send only the cookie, no attribtues * Modifies how a request URL is stored during a XMLRPC request/response sequence. * Refactors a bit of the request/response logic to allow for making the decision whether to send a session cookie instead of full Kerberous auth easier. * The server now includes expiration information in the session cookie it sends to the client. The server always had the information available to prevent using an expired session cookie. Now that expiration timestamp is returned to the client as well and now the client will not send an expired session cookie back to the server. * Adds a new module and unit test for cookies (see below) Formerly we were always returning the session cookie no matter what the domain or path was in the URL. We were also sending the cookie attributes which are for the client only (used to determine if to return a cookie). The attributes are not meant to be sent to the server and the previous behavior was a protocol violation. We also were not checking the cookie expiration. Cookie library issues: We need a library to create, parse, manipulate and format cookies both in a client context and a server context. Core Python has two cookie libraries, Cookie.py and cookielib.py. Why did we add a new cookie module instead of using either of these two core Python libaries? Cookie.py is designed for server side generation but can be used to parse cookies on the client. It's the library we were using in the server. However when I tried to use it in the client I discovered it has some serious bugs. There are 7 defined cookie elements, it fails to correctly parse 3 of the 7 elements which makes it unusable because we depend on those elements. Since Cookie.py was designed for server side cookie processing it's not hard to understand how fails to correctly parse a cookie because that's a client side need. (Cookie.py also has an awkward baroque API and is missing some useful functionality we would have to build on top of it). cookielib.py is designed for client side. It's fully featured and obeys all the RFC's. It would be great to use however it's tightly coupled with another core library, urllib2.py. The http request and response objects must be urllib2 objects. But we don't use urllib2, rather we use httplib because xmlrpclib uses httplib. I don't see a reason why a cookie library should be so tightly coupled to a protocol library, but it is and that means we can't use it (I tried to just pick some isolated entrypoints for our use but I kept hitting interaction/dependency problems). I decided to solve the cookie library problems by writing a minimal cookie library that does what we need and no more than that. It is a new module in ipapython shared by both client and server and comes with a new unit test. The module has plenty of documentation, no need to repeat it here. Request URL issues: We also had problems in rpc.py whereby information from the request which is needed when we process the response is not available. Most important was the requesting URL. It turns out that the way the class and object relationships are structured it's impossible to get this information. Someone else must have run into the same issue because there was a routine called reconstruct_url() which attempted to recreate the request URL from other available information. Unfortunately reconstruct_url() was not callable from inside the response handler. So I decided to store the information in the thread context and when the request is received extract it from the thread context. It's perhaps not an ideal solution but we do similar things elsewhere so at least it's consistent. I removed the reconstruct_url() function because the exact information is now in the context and trying to apply heuristics to recreate the url is probably not robust. Ticket https://fedorahosted.org/freeipa/ticket/3022
2012-12-04 17:20:17 -06:00
cookie.path = '/toplevel'
assert cookie.path == '/toplevel'
Compliant client side session cookie behavior In summary this patch does: * Follow the defined rules for cookies when: - receiving a cookie (process the attributes) - storing a cookie (store cookie + attributes) - sending a cookie + validate the cookie domain against the request URL + validate the cookie path against the request URL + validate the cookie expiration + if valid then send only the cookie, no attribtues * Modifies how a request URL is stored during a XMLRPC request/response sequence. * Refactors a bit of the request/response logic to allow for making the decision whether to send a session cookie instead of full Kerberous auth easier. * The server now includes expiration information in the session cookie it sends to the client. The server always had the information available to prevent using an expired session cookie. Now that expiration timestamp is returned to the client as well and now the client will not send an expired session cookie back to the server. * Adds a new module and unit test for cookies (see below) Formerly we were always returning the session cookie no matter what the domain or path was in the URL. We were also sending the cookie attributes which are for the client only (used to determine if to return a cookie). The attributes are not meant to be sent to the server and the previous behavior was a protocol violation. We also were not checking the cookie expiration. Cookie library issues: We need a library to create, parse, manipulate and format cookies both in a client context and a server context. Core Python has two cookie libraries, Cookie.py and cookielib.py. Why did we add a new cookie module instead of using either of these two core Python libaries? Cookie.py is designed for server side generation but can be used to parse cookies on the client. It's the library we were using in the server. However when I tried to use it in the client I discovered it has some serious bugs. There are 7 defined cookie elements, it fails to correctly parse 3 of the 7 elements which makes it unusable because we depend on those elements. Since Cookie.py was designed for server side cookie processing it's not hard to understand how fails to correctly parse a cookie because that's a client side need. (Cookie.py also has an awkward baroque API and is missing some useful functionality we would have to build on top of it). cookielib.py is designed for client side. It's fully featured and obeys all the RFC's. It would be great to use however it's tightly coupled with another core library, urllib2.py. The http request and response objects must be urllib2 objects. But we don't use urllib2, rather we use httplib because xmlrpclib uses httplib. I don't see a reason why a cookie library should be so tightly coupled to a protocol library, but it is and that means we can't use it (I tried to just pick some isolated entrypoints for our use but I kept hitting interaction/dependency problems). I decided to solve the cookie library problems by writing a minimal cookie library that does what we need and no more than that. It is a new module in ipapython shared by both client and server and comes with a new unit test. The module has plenty of documentation, no need to repeat it here. Request URL issues: We also had problems in rpc.py whereby information from the request which is needed when we process the response is not available. Most important was the requesting URL. It turns out that the way the class and object relationships are structured it's impossible to get this information. Someone else must have run into the same issue because there was a routine called reconstruct_url() which attempted to recreate the request URL from other available information. Unfortunately reconstruct_url() was not callable from inside the response handler. So I decided to store the information in the thread context and when the request is received extract it from the thread context. It's perhaps not an ideal solution but we do similar things elsewhere so at least it's consistent. I removed the reconstruct_url() function because the exact information is now in the context and trying to apply heuristics to recreate the url is probably not robust. Ticket https://fedorahosted.org/freeipa/ticket/3022
2012-12-04 17:20:17 -06:00
cookie.path = None
assert cookie.path is None
Compliant client side session cookie behavior In summary this patch does: * Follow the defined rules for cookies when: - receiving a cookie (process the attributes) - storing a cookie (store cookie + attributes) - sending a cookie + validate the cookie domain against the request URL + validate the cookie path against the request URL + validate the cookie expiration + if valid then send only the cookie, no attribtues * Modifies how a request URL is stored during a XMLRPC request/response sequence. * Refactors a bit of the request/response logic to allow for making the decision whether to send a session cookie instead of full Kerberous auth easier. * The server now includes expiration information in the session cookie it sends to the client. The server always had the information available to prevent using an expired session cookie. Now that expiration timestamp is returned to the client as well and now the client will not send an expired session cookie back to the server. * Adds a new module and unit test for cookies (see below) Formerly we were always returning the session cookie no matter what the domain or path was in the URL. We were also sending the cookie attributes which are for the client only (used to determine if to return a cookie). The attributes are not meant to be sent to the server and the previous behavior was a protocol violation. We also were not checking the cookie expiration. Cookie library issues: We need a library to create, parse, manipulate and format cookies both in a client context and a server context. Core Python has two cookie libraries, Cookie.py and cookielib.py. Why did we add a new cookie module instead of using either of these two core Python libaries? Cookie.py is designed for server side generation but can be used to parse cookies on the client. It's the library we were using in the server. However when I tried to use it in the client I discovered it has some serious bugs. There are 7 defined cookie elements, it fails to correctly parse 3 of the 7 elements which makes it unusable because we depend on those elements. Since Cookie.py was designed for server side cookie processing it's not hard to understand how fails to correctly parse a cookie because that's a client side need. (Cookie.py also has an awkward baroque API and is missing some useful functionality we would have to build on top of it). cookielib.py is designed for client side. It's fully featured and obeys all the RFC's. It would be great to use however it's tightly coupled with another core library, urllib2.py. The http request and response objects must be urllib2 objects. But we don't use urllib2, rather we use httplib because xmlrpclib uses httplib. I don't see a reason why a cookie library should be so tightly coupled to a protocol library, but it is and that means we can't use it (I tried to just pick some isolated entrypoints for our use but I kept hitting interaction/dependency problems). I decided to solve the cookie library problems by writing a minimal cookie library that does what we need and no more than that. It is a new module in ipapython shared by both client and server and comes with a new unit test. The module has plenty of documentation, no need to repeat it here. Request URL issues: We also had problems in rpc.py whereby information from the request which is needed when we process the response is not available. Most important was the requesting URL. It turns out that the way the class and object relationships are structured it's impossible to get this information. Someone else must have run into the same issue because there was a routine called reconstruct_url() which attempted to recreate the request URL from other available information. Unfortunately reconstruct_url() was not callable from inside the response handler. So I decided to store the information in the thread context and when the request is received extract it from the thread context. It's perhaps not an ideal solution but we do similar things elsewhere so at least it's consistent. I removed the reconstruct_url() function because the exact information is now in the context and trying to apply heuristics to recreate the url is probably not robust. Ticket https://fedorahosted.org/freeipa/ticket/3022
2012-12-04 17:20:17 -06:00
cookie.max_age = 400
assert cookie.max_age == 400
Compliant client side session cookie behavior In summary this patch does: * Follow the defined rules for cookies when: - receiving a cookie (process the attributes) - storing a cookie (store cookie + attributes) - sending a cookie + validate the cookie domain against the request URL + validate the cookie path against the request URL + validate the cookie expiration + if valid then send only the cookie, no attribtues * Modifies how a request URL is stored during a XMLRPC request/response sequence. * Refactors a bit of the request/response logic to allow for making the decision whether to send a session cookie instead of full Kerberous auth easier. * The server now includes expiration information in the session cookie it sends to the client. The server always had the information available to prevent using an expired session cookie. Now that expiration timestamp is returned to the client as well and now the client will not send an expired session cookie back to the server. * Adds a new module and unit test for cookies (see below) Formerly we were always returning the session cookie no matter what the domain or path was in the URL. We were also sending the cookie attributes which are for the client only (used to determine if to return a cookie). The attributes are not meant to be sent to the server and the previous behavior was a protocol violation. We also were not checking the cookie expiration. Cookie library issues: We need a library to create, parse, manipulate and format cookies both in a client context and a server context. Core Python has two cookie libraries, Cookie.py and cookielib.py. Why did we add a new cookie module instead of using either of these two core Python libaries? Cookie.py is designed for server side generation but can be used to parse cookies on the client. It's the library we were using in the server. However when I tried to use it in the client I discovered it has some serious bugs. There are 7 defined cookie elements, it fails to correctly parse 3 of the 7 elements which makes it unusable because we depend on those elements. Since Cookie.py was designed for server side cookie processing it's not hard to understand how fails to correctly parse a cookie because that's a client side need. (Cookie.py also has an awkward baroque API and is missing some useful functionality we would have to build on top of it). cookielib.py is designed for client side. It's fully featured and obeys all the RFC's. It would be great to use however it's tightly coupled with another core library, urllib2.py. The http request and response objects must be urllib2 objects. But we don't use urllib2, rather we use httplib because xmlrpclib uses httplib. I don't see a reason why a cookie library should be so tightly coupled to a protocol library, but it is and that means we can't use it (I tried to just pick some isolated entrypoints for our use but I kept hitting interaction/dependency problems). I decided to solve the cookie library problems by writing a minimal cookie library that does what we need and no more than that. It is a new module in ipapython shared by both client and server and comes with a new unit test. The module has plenty of documentation, no need to repeat it here. Request URL issues: We also had problems in rpc.py whereby information from the request which is needed when we process the response is not available. Most important was the requesting URL. It turns out that the way the class and object relationships are structured it's impossible to get this information. Someone else must have run into the same issue because there was a routine called reconstruct_url() which attempted to recreate the request URL from other available information. Unfortunately reconstruct_url() was not callable from inside the response handler. So I decided to store the information in the thread context and when the request is received extract it from the thread context. It's perhaps not an ideal solution but we do similar things elsewhere so at least it's consistent. I removed the reconstruct_url() function because the exact information is now in the context and trying to apply heuristics to recreate the url is probably not robust. Ticket https://fedorahosted.org/freeipa/ticket/3022
2012-12-04 17:20:17 -06:00
cookie.max_age = None
assert cookie.max_age is None
Compliant client side session cookie behavior In summary this patch does: * Follow the defined rules for cookies when: - receiving a cookie (process the attributes) - storing a cookie (store cookie + attributes) - sending a cookie + validate the cookie domain against the request URL + validate the cookie path against the request URL + validate the cookie expiration + if valid then send only the cookie, no attribtues * Modifies how a request URL is stored during a XMLRPC request/response sequence. * Refactors a bit of the request/response logic to allow for making the decision whether to send a session cookie instead of full Kerberous auth easier. * The server now includes expiration information in the session cookie it sends to the client. The server always had the information available to prevent using an expired session cookie. Now that expiration timestamp is returned to the client as well and now the client will not send an expired session cookie back to the server. * Adds a new module and unit test for cookies (see below) Formerly we were always returning the session cookie no matter what the domain or path was in the URL. We were also sending the cookie attributes which are for the client only (used to determine if to return a cookie). The attributes are not meant to be sent to the server and the previous behavior was a protocol violation. We also were not checking the cookie expiration. Cookie library issues: We need a library to create, parse, manipulate and format cookies both in a client context and a server context. Core Python has two cookie libraries, Cookie.py and cookielib.py. Why did we add a new cookie module instead of using either of these two core Python libaries? Cookie.py is designed for server side generation but can be used to parse cookies on the client. It's the library we were using in the server. However when I tried to use it in the client I discovered it has some serious bugs. There are 7 defined cookie elements, it fails to correctly parse 3 of the 7 elements which makes it unusable because we depend on those elements. Since Cookie.py was designed for server side cookie processing it's not hard to understand how fails to correctly parse a cookie because that's a client side need. (Cookie.py also has an awkward baroque API and is missing some useful functionality we would have to build on top of it). cookielib.py is designed for client side. It's fully featured and obeys all the RFC's. It would be great to use however it's tightly coupled with another core library, urllib2.py. The http request and response objects must be urllib2 objects. But we don't use urllib2, rather we use httplib because xmlrpclib uses httplib. I don't see a reason why a cookie library should be so tightly coupled to a protocol library, but it is and that means we can't use it (I tried to just pick some isolated entrypoints for our use but I kept hitting interaction/dependency problems). I decided to solve the cookie library problems by writing a minimal cookie library that does what we need and no more than that. It is a new module in ipapython shared by both client and server and comes with a new unit test. The module has plenty of documentation, no need to repeat it here. Request URL issues: We also had problems in rpc.py whereby information from the request which is needed when we process the response is not available. Most important was the requesting URL. It turns out that the way the class and object relationships are structured it's impossible to get this information. Someone else must have run into the same issue because there was a routine called reconstruct_url() which attempted to recreate the request URL from other available information. Unfortunately reconstruct_url() was not callable from inside the response handler. So I decided to store the information in the thread context and when the request is received extract it from the thread context. It's perhaps not an ideal solution but we do similar things elsewhere so at least it's consistent. I removed the reconstruct_url() function because the exact information is now in the context and trying to apply heuristics to recreate the url is probably not robust. Ticket https://fedorahosted.org/freeipa/ticket/3022
2012-12-04 17:20:17 -06:00
cookie.expires = 'Sun, 06 Nov 1994 08:49:37 GMT'
assert cookie.expires == datetime.datetime(
1994, 11, 6, 8, 49, 37, tzinfo=datetime.timezone.utc)
Compliant client side session cookie behavior In summary this patch does: * Follow the defined rules for cookies when: - receiving a cookie (process the attributes) - storing a cookie (store cookie + attributes) - sending a cookie + validate the cookie domain against the request URL + validate the cookie path against the request URL + validate the cookie expiration + if valid then send only the cookie, no attribtues * Modifies how a request URL is stored during a XMLRPC request/response sequence. * Refactors a bit of the request/response logic to allow for making the decision whether to send a session cookie instead of full Kerberous auth easier. * The server now includes expiration information in the session cookie it sends to the client. The server always had the information available to prevent using an expired session cookie. Now that expiration timestamp is returned to the client as well and now the client will not send an expired session cookie back to the server. * Adds a new module and unit test for cookies (see below) Formerly we were always returning the session cookie no matter what the domain or path was in the URL. We were also sending the cookie attributes which are for the client only (used to determine if to return a cookie). The attributes are not meant to be sent to the server and the previous behavior was a protocol violation. We also were not checking the cookie expiration. Cookie library issues: We need a library to create, parse, manipulate and format cookies both in a client context and a server context. Core Python has two cookie libraries, Cookie.py and cookielib.py. Why did we add a new cookie module instead of using either of these two core Python libaries? Cookie.py is designed for server side generation but can be used to parse cookies on the client. It's the library we were using in the server. However when I tried to use it in the client I discovered it has some serious bugs. There are 7 defined cookie elements, it fails to correctly parse 3 of the 7 elements which makes it unusable because we depend on those elements. Since Cookie.py was designed for server side cookie processing it's not hard to understand how fails to correctly parse a cookie because that's a client side need. (Cookie.py also has an awkward baroque API and is missing some useful functionality we would have to build on top of it). cookielib.py is designed for client side. It's fully featured and obeys all the RFC's. It would be great to use however it's tightly coupled with another core library, urllib2.py. The http request and response objects must be urllib2 objects. But we don't use urllib2, rather we use httplib because xmlrpclib uses httplib. I don't see a reason why a cookie library should be so tightly coupled to a protocol library, but it is and that means we can't use it (I tried to just pick some isolated entrypoints for our use but I kept hitting interaction/dependency problems). I decided to solve the cookie library problems by writing a minimal cookie library that does what we need and no more than that. It is a new module in ipapython shared by both client and server and comes with a new unit test. The module has plenty of documentation, no need to repeat it here. Request URL issues: We also had problems in rpc.py whereby information from the request which is needed when we process the response is not available. Most important was the requesting URL. It turns out that the way the class and object relationships are structured it's impossible to get this information. Someone else must have run into the same issue because there was a routine called reconstruct_url() which attempted to recreate the request URL from other available information. Unfortunately reconstruct_url() was not callable from inside the response handler. So I decided to store the information in the thread context and when the request is received extract it from the thread context. It's perhaps not an ideal solution but we do similar things elsewhere so at least it's consistent. I removed the reconstruct_url() function because the exact information is now in the context and trying to apply heuristics to recreate the url is probably not robust. Ticket https://fedorahosted.org/freeipa/ticket/3022
2012-12-04 17:20:17 -06:00
cookie.expires = None
assert cookie.expires is None
Compliant client side session cookie behavior In summary this patch does: * Follow the defined rules for cookies when: - receiving a cookie (process the attributes) - storing a cookie (store cookie + attributes) - sending a cookie + validate the cookie domain against the request URL + validate the cookie path against the request URL + validate the cookie expiration + if valid then send only the cookie, no attribtues * Modifies how a request URL is stored during a XMLRPC request/response sequence. * Refactors a bit of the request/response logic to allow for making the decision whether to send a session cookie instead of full Kerberous auth easier. * The server now includes expiration information in the session cookie it sends to the client. The server always had the information available to prevent using an expired session cookie. Now that expiration timestamp is returned to the client as well and now the client will not send an expired session cookie back to the server. * Adds a new module and unit test for cookies (see below) Formerly we were always returning the session cookie no matter what the domain or path was in the URL. We were also sending the cookie attributes which are for the client only (used to determine if to return a cookie). The attributes are not meant to be sent to the server and the previous behavior was a protocol violation. We also were not checking the cookie expiration. Cookie library issues: We need a library to create, parse, manipulate and format cookies both in a client context and a server context. Core Python has two cookie libraries, Cookie.py and cookielib.py. Why did we add a new cookie module instead of using either of these two core Python libaries? Cookie.py is designed for server side generation but can be used to parse cookies on the client. It's the library we were using in the server. However when I tried to use it in the client I discovered it has some serious bugs. There are 7 defined cookie elements, it fails to correctly parse 3 of the 7 elements which makes it unusable because we depend on those elements. Since Cookie.py was designed for server side cookie processing it's not hard to understand how fails to correctly parse a cookie because that's a client side need. (Cookie.py also has an awkward baroque API and is missing some useful functionality we would have to build on top of it). cookielib.py is designed for client side. It's fully featured and obeys all the RFC's. It would be great to use however it's tightly coupled with another core library, urllib2.py. The http request and response objects must be urllib2 objects. But we don't use urllib2, rather we use httplib because xmlrpclib uses httplib. I don't see a reason why a cookie library should be so tightly coupled to a protocol library, but it is and that means we can't use it (I tried to just pick some isolated entrypoints for our use but I kept hitting interaction/dependency problems). I decided to solve the cookie library problems by writing a minimal cookie library that does what we need and no more than that. It is a new module in ipapython shared by both client and server and comes with a new unit test. The module has plenty of documentation, no need to repeat it here. Request URL issues: We also had problems in rpc.py whereby information from the request which is needed when we process the response is not available. Most important was the requesting URL. It turns out that the way the class and object relationships are structured it's impossible to get this information. Someone else must have run into the same issue because there was a routine called reconstruct_url() which attempted to recreate the request URL from other available information. Unfortunately reconstruct_url() was not callable from inside the response handler. So I decided to store the information in the thread context and when the request is received extract it from the thread context. It's perhaps not an ideal solution but we do similar things elsewhere so at least it's consistent. I removed the reconstruct_url() function because the exact information is now in the context and trying to apply heuristics to recreate the url is probably not robust. Ticket https://fedorahosted.org/freeipa/ticket/3022
2012-12-04 17:20:17 -06:00
cookie.secure = True
assert cookie.secure is True
assert str(cookie) == "color=blue; Secure"
Compliant client side session cookie behavior In summary this patch does: * Follow the defined rules for cookies when: - receiving a cookie (process the attributes) - storing a cookie (store cookie + attributes) - sending a cookie + validate the cookie domain against the request URL + validate the cookie path against the request URL + validate the cookie expiration + if valid then send only the cookie, no attribtues * Modifies how a request URL is stored during a XMLRPC request/response sequence. * Refactors a bit of the request/response logic to allow for making the decision whether to send a session cookie instead of full Kerberous auth easier. * The server now includes expiration information in the session cookie it sends to the client. The server always had the information available to prevent using an expired session cookie. Now that expiration timestamp is returned to the client as well and now the client will not send an expired session cookie back to the server. * Adds a new module and unit test for cookies (see below) Formerly we were always returning the session cookie no matter what the domain or path was in the URL. We were also sending the cookie attributes which are for the client only (used to determine if to return a cookie). The attributes are not meant to be sent to the server and the previous behavior was a protocol violation. We also were not checking the cookie expiration. Cookie library issues: We need a library to create, parse, manipulate and format cookies both in a client context and a server context. Core Python has two cookie libraries, Cookie.py and cookielib.py. Why did we add a new cookie module instead of using either of these two core Python libaries? Cookie.py is designed for server side generation but can be used to parse cookies on the client. It's the library we were using in the server. However when I tried to use it in the client I discovered it has some serious bugs. There are 7 defined cookie elements, it fails to correctly parse 3 of the 7 elements which makes it unusable because we depend on those elements. Since Cookie.py was designed for server side cookie processing it's not hard to understand how fails to correctly parse a cookie because that's a client side need. (Cookie.py also has an awkward baroque API and is missing some useful functionality we would have to build on top of it). cookielib.py is designed for client side. It's fully featured and obeys all the RFC's. It would be great to use however it's tightly coupled with another core library, urllib2.py. The http request and response objects must be urllib2 objects. But we don't use urllib2, rather we use httplib because xmlrpclib uses httplib. I don't see a reason why a cookie library should be so tightly coupled to a protocol library, but it is and that means we can't use it (I tried to just pick some isolated entrypoints for our use but I kept hitting interaction/dependency problems). I decided to solve the cookie library problems by writing a minimal cookie library that does what we need and no more than that. It is a new module in ipapython shared by both client and server and comes with a new unit test. The module has plenty of documentation, no need to repeat it here. Request URL issues: We also had problems in rpc.py whereby information from the request which is needed when we process the response is not available. Most important was the requesting URL. It turns out that the way the class and object relationships are structured it's impossible to get this information. Someone else must have run into the same issue because there was a routine called reconstruct_url() which attempted to recreate the request URL from other available information. Unfortunately reconstruct_url() was not callable from inside the response handler. So I decided to store the information in the thread context and when the request is received extract it from the thread context. It's perhaps not an ideal solution but we do similar things elsewhere so at least it's consistent. I removed the reconstruct_url() function because the exact information is now in the context and trying to apply heuristics to recreate the url is probably not robust. Ticket https://fedorahosted.org/freeipa/ticket/3022
2012-12-04 17:20:17 -06:00
cookie.secure = False
assert cookie.secure is False
assert str(cookie) == "color=blue"
Compliant client side session cookie behavior In summary this patch does: * Follow the defined rules for cookies when: - receiving a cookie (process the attributes) - storing a cookie (store cookie + attributes) - sending a cookie + validate the cookie domain against the request URL + validate the cookie path against the request URL + validate the cookie expiration + if valid then send only the cookie, no attribtues * Modifies how a request URL is stored during a XMLRPC request/response sequence. * Refactors a bit of the request/response logic to allow for making the decision whether to send a session cookie instead of full Kerberous auth easier. * The server now includes expiration information in the session cookie it sends to the client. The server always had the information available to prevent using an expired session cookie. Now that expiration timestamp is returned to the client as well and now the client will not send an expired session cookie back to the server. * Adds a new module and unit test for cookies (see below) Formerly we were always returning the session cookie no matter what the domain or path was in the URL. We were also sending the cookie attributes which are for the client only (used to determine if to return a cookie). The attributes are not meant to be sent to the server and the previous behavior was a protocol violation. We also were not checking the cookie expiration. Cookie library issues: We need a library to create, parse, manipulate and format cookies both in a client context and a server context. Core Python has two cookie libraries, Cookie.py and cookielib.py. Why did we add a new cookie module instead of using either of these two core Python libaries? Cookie.py is designed for server side generation but can be used to parse cookies on the client. It's the library we were using in the server. However when I tried to use it in the client I discovered it has some serious bugs. There are 7 defined cookie elements, it fails to correctly parse 3 of the 7 elements which makes it unusable because we depend on those elements. Since Cookie.py was designed for server side cookie processing it's not hard to understand how fails to correctly parse a cookie because that's a client side need. (Cookie.py also has an awkward baroque API and is missing some useful functionality we would have to build on top of it). cookielib.py is designed for client side. It's fully featured and obeys all the RFC's. It would be great to use however it's tightly coupled with another core library, urllib2.py. The http request and response objects must be urllib2 objects. But we don't use urllib2, rather we use httplib because xmlrpclib uses httplib. I don't see a reason why a cookie library should be so tightly coupled to a protocol library, but it is and that means we can't use it (I tried to just pick some isolated entrypoints for our use but I kept hitting interaction/dependency problems). I decided to solve the cookie library problems by writing a minimal cookie library that does what we need and no more than that. It is a new module in ipapython shared by both client and server and comes with a new unit test. The module has plenty of documentation, no need to repeat it here. Request URL issues: We also had problems in rpc.py whereby information from the request which is needed when we process the response is not available. Most important was the requesting URL. It turns out that the way the class and object relationships are structured it's impossible to get this information. Someone else must have run into the same issue because there was a routine called reconstruct_url() which attempted to recreate the request URL from other available information. Unfortunately reconstruct_url() was not callable from inside the response handler. So I decided to store the information in the thread context and when the request is received extract it from the thread context. It's perhaps not an ideal solution but we do similar things elsewhere so at least it's consistent. I removed the reconstruct_url() function because the exact information is now in the context and trying to apply heuristics to recreate the url is probably not robust. Ticket https://fedorahosted.org/freeipa/ticket/3022
2012-12-04 17:20:17 -06:00
cookie.secure = None
assert cookie.secure is None
assert str(cookie) == "color=blue"
Compliant client side session cookie behavior In summary this patch does: * Follow the defined rules for cookies when: - receiving a cookie (process the attributes) - storing a cookie (store cookie + attributes) - sending a cookie + validate the cookie domain against the request URL + validate the cookie path against the request URL + validate the cookie expiration + if valid then send only the cookie, no attribtues * Modifies how a request URL is stored during a XMLRPC request/response sequence. * Refactors a bit of the request/response logic to allow for making the decision whether to send a session cookie instead of full Kerberous auth easier. * The server now includes expiration information in the session cookie it sends to the client. The server always had the information available to prevent using an expired session cookie. Now that expiration timestamp is returned to the client as well and now the client will not send an expired session cookie back to the server. * Adds a new module and unit test for cookies (see below) Formerly we were always returning the session cookie no matter what the domain or path was in the URL. We were also sending the cookie attributes which are for the client only (used to determine if to return a cookie). The attributes are not meant to be sent to the server and the previous behavior was a protocol violation. We also were not checking the cookie expiration. Cookie library issues: We need a library to create, parse, manipulate and format cookies both in a client context and a server context. Core Python has two cookie libraries, Cookie.py and cookielib.py. Why did we add a new cookie module instead of using either of these two core Python libaries? Cookie.py is designed for server side generation but can be used to parse cookies on the client. It's the library we were using in the server. However when I tried to use it in the client I discovered it has some serious bugs. There are 7 defined cookie elements, it fails to correctly parse 3 of the 7 elements which makes it unusable because we depend on those elements. Since Cookie.py was designed for server side cookie processing it's not hard to understand how fails to correctly parse a cookie because that's a client side need. (Cookie.py also has an awkward baroque API and is missing some useful functionality we would have to build on top of it). cookielib.py is designed for client side. It's fully featured and obeys all the RFC's. It would be great to use however it's tightly coupled with another core library, urllib2.py. The http request and response objects must be urllib2 objects. But we don't use urllib2, rather we use httplib because xmlrpclib uses httplib. I don't see a reason why a cookie library should be so tightly coupled to a protocol library, but it is and that means we can't use it (I tried to just pick some isolated entrypoints for our use but I kept hitting interaction/dependency problems). I decided to solve the cookie library problems by writing a minimal cookie library that does what we need and no more than that. It is a new module in ipapython shared by both client and server and comes with a new unit test. The module has plenty of documentation, no need to repeat it here. Request URL issues: We also had problems in rpc.py whereby information from the request which is needed when we process the response is not available. Most important was the requesting URL. It turns out that the way the class and object relationships are structured it's impossible to get this information. Someone else must have run into the same issue because there was a routine called reconstruct_url() which attempted to recreate the request URL from other available information. Unfortunately reconstruct_url() was not callable from inside the response handler. So I decided to store the information in the thread context and when the request is received extract it from the thread context. It's perhaps not an ideal solution but we do similar things elsewhere so at least it's consistent. I removed the reconstruct_url() function because the exact information is now in the context and trying to apply heuristics to recreate the url is probably not robust. Ticket https://fedorahosted.org/freeipa/ticket/3022
2012-12-04 17:20:17 -06:00
cookie.httponly = True
assert cookie.httponly is True
assert str(cookie) == "color=blue; HttpOnly"
Compliant client side session cookie behavior In summary this patch does: * Follow the defined rules for cookies when: - receiving a cookie (process the attributes) - storing a cookie (store cookie + attributes) - sending a cookie + validate the cookie domain against the request URL + validate the cookie path against the request URL + validate the cookie expiration + if valid then send only the cookie, no attribtues * Modifies how a request URL is stored during a XMLRPC request/response sequence. * Refactors a bit of the request/response logic to allow for making the decision whether to send a session cookie instead of full Kerberous auth easier. * The server now includes expiration information in the session cookie it sends to the client. The server always had the information available to prevent using an expired session cookie. Now that expiration timestamp is returned to the client as well and now the client will not send an expired session cookie back to the server. * Adds a new module and unit test for cookies (see below) Formerly we were always returning the session cookie no matter what the domain or path was in the URL. We were also sending the cookie attributes which are for the client only (used to determine if to return a cookie). The attributes are not meant to be sent to the server and the previous behavior was a protocol violation. We also were not checking the cookie expiration. Cookie library issues: We need a library to create, parse, manipulate and format cookies both in a client context and a server context. Core Python has two cookie libraries, Cookie.py and cookielib.py. Why did we add a new cookie module instead of using either of these two core Python libaries? Cookie.py is designed for server side generation but can be used to parse cookies on the client. It's the library we were using in the server. However when I tried to use it in the client I discovered it has some serious bugs. There are 7 defined cookie elements, it fails to correctly parse 3 of the 7 elements which makes it unusable because we depend on those elements. Since Cookie.py was designed for server side cookie processing it's not hard to understand how fails to correctly parse a cookie because that's a client side need. (Cookie.py also has an awkward baroque API and is missing some useful functionality we would have to build on top of it). cookielib.py is designed for client side. It's fully featured and obeys all the RFC's. It would be great to use however it's tightly coupled with another core library, urllib2.py. The http request and response objects must be urllib2 objects. But we don't use urllib2, rather we use httplib because xmlrpclib uses httplib. I don't see a reason why a cookie library should be so tightly coupled to a protocol library, but it is and that means we can't use it (I tried to just pick some isolated entrypoints for our use but I kept hitting interaction/dependency problems). I decided to solve the cookie library problems by writing a minimal cookie library that does what we need and no more than that. It is a new module in ipapython shared by both client and server and comes with a new unit test. The module has plenty of documentation, no need to repeat it here. Request URL issues: We also had problems in rpc.py whereby information from the request which is needed when we process the response is not available. Most important was the requesting URL. It turns out that the way the class and object relationships are structured it's impossible to get this information. Someone else must have run into the same issue because there was a routine called reconstruct_url() which attempted to recreate the request URL from other available information. Unfortunately reconstruct_url() was not callable from inside the response handler. So I decided to store the information in the thread context and when the request is received extract it from the thread context. It's perhaps not an ideal solution but we do similar things elsewhere so at least it's consistent. I removed the reconstruct_url() function because the exact information is now in the context and trying to apply heuristics to recreate the url is probably not robust. Ticket https://fedorahosted.org/freeipa/ticket/3022
2012-12-04 17:20:17 -06:00
cookie.httponly = False
assert cookie.httponly is False
assert str(cookie) == "color=blue"
Compliant client side session cookie behavior In summary this patch does: * Follow the defined rules for cookies when: - receiving a cookie (process the attributes) - storing a cookie (store cookie + attributes) - sending a cookie + validate the cookie domain against the request URL + validate the cookie path against the request URL + validate the cookie expiration + if valid then send only the cookie, no attribtues * Modifies how a request URL is stored during a XMLRPC request/response sequence. * Refactors a bit of the request/response logic to allow for making the decision whether to send a session cookie instead of full Kerberous auth easier. * The server now includes expiration information in the session cookie it sends to the client. The server always had the information available to prevent using an expired session cookie. Now that expiration timestamp is returned to the client as well and now the client will not send an expired session cookie back to the server. * Adds a new module and unit test for cookies (see below) Formerly we were always returning the session cookie no matter what the domain or path was in the URL. We were also sending the cookie attributes which are for the client only (used to determine if to return a cookie). The attributes are not meant to be sent to the server and the previous behavior was a protocol violation. We also were not checking the cookie expiration. Cookie library issues: We need a library to create, parse, manipulate and format cookies both in a client context and a server context. Core Python has two cookie libraries, Cookie.py and cookielib.py. Why did we add a new cookie module instead of using either of these two core Python libaries? Cookie.py is designed for server side generation but can be used to parse cookies on the client. It's the library we were using in the server. However when I tried to use it in the client I discovered it has some serious bugs. There are 7 defined cookie elements, it fails to correctly parse 3 of the 7 elements which makes it unusable because we depend on those elements. Since Cookie.py was designed for server side cookie processing it's not hard to understand how fails to correctly parse a cookie because that's a client side need. (Cookie.py also has an awkward baroque API and is missing some useful functionality we would have to build on top of it). cookielib.py is designed for client side. It's fully featured and obeys all the RFC's. It would be great to use however it's tightly coupled with another core library, urllib2.py. The http request and response objects must be urllib2 objects. But we don't use urllib2, rather we use httplib because xmlrpclib uses httplib. I don't see a reason why a cookie library should be so tightly coupled to a protocol library, but it is and that means we can't use it (I tried to just pick some isolated entrypoints for our use but I kept hitting interaction/dependency problems). I decided to solve the cookie library problems by writing a minimal cookie library that does what we need and no more than that. It is a new module in ipapython shared by both client and server and comes with a new unit test. The module has plenty of documentation, no need to repeat it here. Request URL issues: We also had problems in rpc.py whereby information from the request which is needed when we process the response is not available. Most important was the requesting URL. It turns out that the way the class and object relationships are structured it's impossible to get this information. Someone else must have run into the same issue because there was a routine called reconstruct_url() which attempted to recreate the request URL from other available information. Unfortunately reconstruct_url() was not callable from inside the response handler. So I decided to store the information in the thread context and when the request is received extract it from the thread context. It's perhaps not an ideal solution but we do similar things elsewhere so at least it's consistent. I removed the reconstruct_url() function because the exact information is now in the context and trying to apply heuristics to recreate the url is probably not robust. Ticket https://fedorahosted.org/freeipa/ticket/3022
2012-12-04 17:20:17 -06:00
cookie.httponly = None
assert cookie.httponly is None
assert str(cookie) == "color=blue"
Compliant client side session cookie behavior In summary this patch does: * Follow the defined rules for cookies when: - receiving a cookie (process the attributes) - storing a cookie (store cookie + attributes) - sending a cookie + validate the cookie domain against the request URL + validate the cookie path against the request URL + validate the cookie expiration + if valid then send only the cookie, no attribtues * Modifies how a request URL is stored during a XMLRPC request/response sequence. * Refactors a bit of the request/response logic to allow for making the decision whether to send a session cookie instead of full Kerberous auth easier. * The server now includes expiration information in the session cookie it sends to the client. The server always had the information available to prevent using an expired session cookie. Now that expiration timestamp is returned to the client as well and now the client will not send an expired session cookie back to the server. * Adds a new module and unit test for cookies (see below) Formerly we were always returning the session cookie no matter what the domain or path was in the URL. We were also sending the cookie attributes which are for the client only (used to determine if to return a cookie). The attributes are not meant to be sent to the server and the previous behavior was a protocol violation. We also were not checking the cookie expiration. Cookie library issues: We need a library to create, parse, manipulate and format cookies both in a client context and a server context. Core Python has two cookie libraries, Cookie.py and cookielib.py. Why did we add a new cookie module instead of using either of these two core Python libaries? Cookie.py is designed for server side generation but can be used to parse cookies on the client. It's the library we were using in the server. However when I tried to use it in the client I discovered it has some serious bugs. There are 7 defined cookie elements, it fails to correctly parse 3 of the 7 elements which makes it unusable because we depend on those elements. Since Cookie.py was designed for server side cookie processing it's not hard to understand how fails to correctly parse a cookie because that's a client side need. (Cookie.py also has an awkward baroque API and is missing some useful functionality we would have to build on top of it). cookielib.py is designed for client side. It's fully featured and obeys all the RFC's. It would be great to use however it's tightly coupled with another core library, urllib2.py. The http request and response objects must be urllib2 objects. But we don't use urllib2, rather we use httplib because xmlrpclib uses httplib. I don't see a reason why a cookie library should be so tightly coupled to a protocol library, but it is and that means we can't use it (I tried to just pick some isolated entrypoints for our use but I kept hitting interaction/dependency problems). I decided to solve the cookie library problems by writing a minimal cookie library that does what we need and no more than that. It is a new module in ipapython shared by both client and server and comes with a new unit test. The module has plenty of documentation, no need to repeat it here. Request URL issues: We also had problems in rpc.py whereby information from the request which is needed when we process the response is not available. Most important was the requesting URL. It turns out that the way the class and object relationships are structured it's impossible to get this information. Someone else must have run into the same issue because there was a routine called reconstruct_url() which attempted to recreate the request URL from other available information. Unfortunately reconstruct_url() was not callable from inside the response handler. So I decided to store the information in the thread context and when the request is received extract it from the thread context. It's perhaps not an ideal solution but we do similar things elsewhere so at least it's consistent. I removed the reconstruct_url() function because the exact information is now in the context and trying to apply heuristics to recreate the url is probably not robust. Ticket https://fedorahosted.org/freeipa/ticket/3022
2012-12-04 17:20:17 -06:00
class TestHTTPReturn:
@pytest.fixture(autouse=True)
def http_return_setup(self):
Compliant client side session cookie behavior In summary this patch does: * Follow the defined rules for cookies when: - receiving a cookie (process the attributes) - storing a cookie (store cookie + attributes) - sending a cookie + validate the cookie domain against the request URL + validate the cookie path against the request URL + validate the cookie expiration + if valid then send only the cookie, no attribtues * Modifies how a request URL is stored during a XMLRPC request/response sequence. * Refactors a bit of the request/response logic to allow for making the decision whether to send a session cookie instead of full Kerberous auth easier. * The server now includes expiration information in the session cookie it sends to the client. The server always had the information available to prevent using an expired session cookie. Now that expiration timestamp is returned to the client as well and now the client will not send an expired session cookie back to the server. * Adds a new module and unit test for cookies (see below) Formerly we were always returning the session cookie no matter what the domain or path was in the URL. We were also sending the cookie attributes which are for the client only (used to determine if to return a cookie). The attributes are not meant to be sent to the server and the previous behavior was a protocol violation. We also were not checking the cookie expiration. Cookie library issues: We need a library to create, parse, manipulate and format cookies both in a client context and a server context. Core Python has two cookie libraries, Cookie.py and cookielib.py. Why did we add a new cookie module instead of using either of these two core Python libaries? Cookie.py is designed for server side generation but can be used to parse cookies on the client. It's the library we were using in the server. However when I tried to use it in the client I discovered it has some serious bugs. There are 7 defined cookie elements, it fails to correctly parse 3 of the 7 elements which makes it unusable because we depend on those elements. Since Cookie.py was designed for server side cookie processing it's not hard to understand how fails to correctly parse a cookie because that's a client side need. (Cookie.py also has an awkward baroque API and is missing some useful functionality we would have to build on top of it). cookielib.py is designed for client side. It's fully featured and obeys all the RFC's. It would be great to use however it's tightly coupled with another core library, urllib2.py. The http request and response objects must be urllib2 objects. But we don't use urllib2, rather we use httplib because xmlrpclib uses httplib. I don't see a reason why a cookie library should be so tightly coupled to a protocol library, but it is and that means we can't use it (I tried to just pick some isolated entrypoints for our use but I kept hitting interaction/dependency problems). I decided to solve the cookie library problems by writing a minimal cookie library that does what we need and no more than that. It is a new module in ipapython shared by both client and server and comes with a new unit test. The module has plenty of documentation, no need to repeat it here. Request URL issues: We also had problems in rpc.py whereby information from the request which is needed when we process the response is not available. Most important was the requesting URL. It turns out that the way the class and object relationships are structured it's impossible to get this information. Someone else must have run into the same issue because there was a routine called reconstruct_url() which attempted to recreate the request URL from other available information. Unfortunately reconstruct_url() was not callable from inside the response handler. So I decided to store the information in the thread context and when the request is received extract it from the thread context. It's perhaps not an ideal solution but we do similar things elsewhere so at least it's consistent. I removed the reconstruct_url() function because the exact information is now in the context and trying to apply heuristics to recreate the url is probably not robust. Ticket https://fedorahosted.org/freeipa/ticket/3022
2012-12-04 17:20:17 -06:00
self.url = 'http://www.foo.bar.com/one/two'
def test_no_attributes(self):
cookie = Cookie('color', 'blue')
assert cookie.http_return_ok(self.url)
Compliant client side session cookie behavior In summary this patch does: * Follow the defined rules for cookies when: - receiving a cookie (process the attributes) - storing a cookie (store cookie + attributes) - sending a cookie + validate the cookie domain against the request URL + validate the cookie path against the request URL + validate the cookie expiration + if valid then send only the cookie, no attribtues * Modifies how a request URL is stored during a XMLRPC request/response sequence. * Refactors a bit of the request/response logic to allow for making the decision whether to send a session cookie instead of full Kerberous auth easier. * The server now includes expiration information in the session cookie it sends to the client. The server always had the information available to prevent using an expired session cookie. Now that expiration timestamp is returned to the client as well and now the client will not send an expired session cookie back to the server. * Adds a new module and unit test for cookies (see below) Formerly we were always returning the session cookie no matter what the domain or path was in the URL. We were also sending the cookie attributes which are for the client only (used to determine if to return a cookie). The attributes are not meant to be sent to the server and the previous behavior was a protocol violation. We also were not checking the cookie expiration. Cookie library issues: We need a library to create, parse, manipulate and format cookies both in a client context and a server context. Core Python has two cookie libraries, Cookie.py and cookielib.py. Why did we add a new cookie module instead of using either of these two core Python libaries? Cookie.py is designed for server side generation but can be used to parse cookies on the client. It's the library we were using in the server. However when I tried to use it in the client I discovered it has some serious bugs. There are 7 defined cookie elements, it fails to correctly parse 3 of the 7 elements which makes it unusable because we depend on those elements. Since Cookie.py was designed for server side cookie processing it's not hard to understand how fails to correctly parse a cookie because that's a client side need. (Cookie.py also has an awkward baroque API and is missing some useful functionality we would have to build on top of it). cookielib.py is designed for client side. It's fully featured and obeys all the RFC's. It would be great to use however it's tightly coupled with another core library, urllib2.py. The http request and response objects must be urllib2 objects. But we don't use urllib2, rather we use httplib because xmlrpclib uses httplib. I don't see a reason why a cookie library should be so tightly coupled to a protocol library, but it is and that means we can't use it (I tried to just pick some isolated entrypoints for our use but I kept hitting interaction/dependency problems). I decided to solve the cookie library problems by writing a minimal cookie library that does what we need and no more than that. It is a new module in ipapython shared by both client and server and comes with a new unit test. The module has plenty of documentation, no need to repeat it here. Request URL issues: We also had problems in rpc.py whereby information from the request which is needed when we process the response is not available. Most important was the requesting URL. It turns out that the way the class and object relationships are structured it's impossible to get this information. Someone else must have run into the same issue because there was a routine called reconstruct_url() which attempted to recreate the request URL from other available information. Unfortunately reconstruct_url() was not callable from inside the response handler. So I decided to store the information in the thread context and when the request is received extract it from the thread context. It's perhaps not an ideal solution but we do similar things elsewhere so at least it's consistent. I removed the reconstruct_url() function because the exact information is now in the context and trying to apply heuristics to recreate the url is probably not robust. Ticket https://fedorahosted.org/freeipa/ticket/3022
2012-12-04 17:20:17 -06:00
def test_domain(self):
cookie = Cookie('color', 'blue', domain='www.foo.bar.com')
assert cookie.http_return_ok(self.url)
Compliant client side session cookie behavior In summary this patch does: * Follow the defined rules for cookies when: - receiving a cookie (process the attributes) - storing a cookie (store cookie + attributes) - sending a cookie + validate the cookie domain against the request URL + validate the cookie path against the request URL + validate the cookie expiration + if valid then send only the cookie, no attribtues * Modifies how a request URL is stored during a XMLRPC request/response sequence. * Refactors a bit of the request/response logic to allow for making the decision whether to send a session cookie instead of full Kerberous auth easier. * The server now includes expiration information in the session cookie it sends to the client. The server always had the information available to prevent using an expired session cookie. Now that expiration timestamp is returned to the client as well and now the client will not send an expired session cookie back to the server. * Adds a new module and unit test for cookies (see below) Formerly we were always returning the session cookie no matter what the domain or path was in the URL. We were also sending the cookie attributes which are for the client only (used to determine if to return a cookie). The attributes are not meant to be sent to the server and the previous behavior was a protocol violation. We also were not checking the cookie expiration. Cookie library issues: We need a library to create, parse, manipulate and format cookies both in a client context and a server context. Core Python has two cookie libraries, Cookie.py and cookielib.py. Why did we add a new cookie module instead of using either of these two core Python libaries? Cookie.py is designed for server side generation but can be used to parse cookies on the client. It's the library we were using in the server. However when I tried to use it in the client I discovered it has some serious bugs. There are 7 defined cookie elements, it fails to correctly parse 3 of the 7 elements which makes it unusable because we depend on those elements. Since Cookie.py was designed for server side cookie processing it's not hard to understand how fails to correctly parse a cookie because that's a client side need. (Cookie.py also has an awkward baroque API and is missing some useful functionality we would have to build on top of it). cookielib.py is designed for client side. It's fully featured and obeys all the RFC's. It would be great to use however it's tightly coupled with another core library, urllib2.py. The http request and response objects must be urllib2 objects. But we don't use urllib2, rather we use httplib because xmlrpclib uses httplib. I don't see a reason why a cookie library should be so tightly coupled to a protocol library, but it is and that means we can't use it (I tried to just pick some isolated entrypoints for our use but I kept hitting interaction/dependency problems). I decided to solve the cookie library problems by writing a minimal cookie library that does what we need and no more than that. It is a new module in ipapython shared by both client and server and comes with a new unit test. The module has plenty of documentation, no need to repeat it here. Request URL issues: We also had problems in rpc.py whereby information from the request which is needed when we process the response is not available. Most important was the requesting URL. It turns out that the way the class and object relationships are structured it's impossible to get this information. Someone else must have run into the same issue because there was a routine called reconstruct_url() which attempted to recreate the request URL from other available information. Unfortunately reconstruct_url() was not callable from inside the response handler. So I decided to store the information in the thread context and when the request is received extract it from the thread context. It's perhaps not an ideal solution but we do similar things elsewhere so at least it's consistent. I removed the reconstruct_url() function because the exact information is now in the context and trying to apply heuristics to recreate the url is probably not robust. Ticket https://fedorahosted.org/freeipa/ticket/3022
2012-12-04 17:20:17 -06:00
cookie = Cookie('color', 'blue', domain='.foo.bar.com')
assert cookie.http_return_ok(self.url)
Compliant client side session cookie behavior In summary this patch does: * Follow the defined rules for cookies when: - receiving a cookie (process the attributes) - storing a cookie (store cookie + attributes) - sending a cookie + validate the cookie domain against the request URL + validate the cookie path against the request URL + validate the cookie expiration + if valid then send only the cookie, no attribtues * Modifies how a request URL is stored during a XMLRPC request/response sequence. * Refactors a bit of the request/response logic to allow for making the decision whether to send a session cookie instead of full Kerberous auth easier. * The server now includes expiration information in the session cookie it sends to the client. The server always had the information available to prevent using an expired session cookie. Now that expiration timestamp is returned to the client as well and now the client will not send an expired session cookie back to the server. * Adds a new module and unit test for cookies (see below) Formerly we were always returning the session cookie no matter what the domain or path was in the URL. We were also sending the cookie attributes which are for the client only (used to determine if to return a cookie). The attributes are not meant to be sent to the server and the previous behavior was a protocol violation. We also were not checking the cookie expiration. Cookie library issues: We need a library to create, parse, manipulate and format cookies both in a client context and a server context. Core Python has two cookie libraries, Cookie.py and cookielib.py. Why did we add a new cookie module instead of using either of these two core Python libaries? Cookie.py is designed for server side generation but can be used to parse cookies on the client. It's the library we were using in the server. However when I tried to use it in the client I discovered it has some serious bugs. There are 7 defined cookie elements, it fails to correctly parse 3 of the 7 elements which makes it unusable because we depend on those elements. Since Cookie.py was designed for server side cookie processing it's not hard to understand how fails to correctly parse a cookie because that's a client side need. (Cookie.py also has an awkward baroque API and is missing some useful functionality we would have to build on top of it). cookielib.py is designed for client side. It's fully featured and obeys all the RFC's. It would be great to use however it's tightly coupled with another core library, urllib2.py. The http request and response objects must be urllib2 objects. But we don't use urllib2, rather we use httplib because xmlrpclib uses httplib. I don't see a reason why a cookie library should be so tightly coupled to a protocol library, but it is and that means we can't use it (I tried to just pick some isolated entrypoints for our use but I kept hitting interaction/dependency problems). I decided to solve the cookie library problems by writing a minimal cookie library that does what we need and no more than that. It is a new module in ipapython shared by both client and server and comes with a new unit test. The module has plenty of documentation, no need to repeat it here. Request URL issues: We also had problems in rpc.py whereby information from the request which is needed when we process the response is not available. Most important was the requesting URL. It turns out that the way the class and object relationships are structured it's impossible to get this information. Someone else must have run into the same issue because there was a routine called reconstruct_url() which attempted to recreate the request URL from other available information. Unfortunately reconstruct_url() was not callable from inside the response handler. So I decided to store the information in the thread context and when the request is received extract it from the thread context. It's perhaps not an ideal solution but we do similar things elsewhere so at least it's consistent. I removed the reconstruct_url() function because the exact information is now in the context and trying to apply heuristics to recreate the url is probably not robust. Ticket https://fedorahosted.org/freeipa/ticket/3022
2012-12-04 17:20:17 -06:00
cookie = Cookie('color', 'blue', domain='.bar.com')
assert cookie.http_return_ok(self.url)
Compliant client side session cookie behavior In summary this patch does: * Follow the defined rules for cookies when: - receiving a cookie (process the attributes) - storing a cookie (store cookie + attributes) - sending a cookie + validate the cookie domain against the request URL + validate the cookie path against the request URL + validate the cookie expiration + if valid then send only the cookie, no attribtues * Modifies how a request URL is stored during a XMLRPC request/response sequence. * Refactors a bit of the request/response logic to allow for making the decision whether to send a session cookie instead of full Kerberous auth easier. * The server now includes expiration information in the session cookie it sends to the client. The server always had the information available to prevent using an expired session cookie. Now that expiration timestamp is returned to the client as well and now the client will not send an expired session cookie back to the server. * Adds a new module and unit test for cookies (see below) Formerly we were always returning the session cookie no matter what the domain or path was in the URL. We were also sending the cookie attributes which are for the client only (used to determine if to return a cookie). The attributes are not meant to be sent to the server and the previous behavior was a protocol violation. We also were not checking the cookie expiration. Cookie library issues: We need a library to create, parse, manipulate and format cookies both in a client context and a server context. Core Python has two cookie libraries, Cookie.py and cookielib.py. Why did we add a new cookie module instead of using either of these two core Python libaries? Cookie.py is designed for server side generation but can be used to parse cookies on the client. It's the library we were using in the server. However when I tried to use it in the client I discovered it has some serious bugs. There are 7 defined cookie elements, it fails to correctly parse 3 of the 7 elements which makes it unusable because we depend on those elements. Since Cookie.py was designed for server side cookie processing it's not hard to understand how fails to correctly parse a cookie because that's a client side need. (Cookie.py also has an awkward baroque API and is missing some useful functionality we would have to build on top of it). cookielib.py is designed for client side. It's fully featured and obeys all the RFC's. It would be great to use however it's tightly coupled with another core library, urllib2.py. The http request and response objects must be urllib2 objects. But we don't use urllib2, rather we use httplib because xmlrpclib uses httplib. I don't see a reason why a cookie library should be so tightly coupled to a protocol library, but it is and that means we can't use it (I tried to just pick some isolated entrypoints for our use but I kept hitting interaction/dependency problems). I decided to solve the cookie library problems by writing a minimal cookie library that does what we need and no more than that. It is a new module in ipapython shared by both client and server and comes with a new unit test. The module has plenty of documentation, no need to repeat it here. Request URL issues: We also had problems in rpc.py whereby information from the request which is needed when we process the response is not available. Most important was the requesting URL. It turns out that the way the class and object relationships are structured it's impossible to get this information. Someone else must have run into the same issue because there was a routine called reconstruct_url() which attempted to recreate the request URL from other available information. Unfortunately reconstruct_url() was not callable from inside the response handler. So I decided to store the information in the thread context and when the request is received extract it from the thread context. It's perhaps not an ideal solution but we do similar things elsewhere so at least it's consistent. I removed the reconstruct_url() function because the exact information is now in the context and trying to apply heuristics to recreate the url is probably not robust. Ticket https://fedorahosted.org/freeipa/ticket/3022
2012-12-04 17:20:17 -06:00
cookie = Cookie('color', 'blue', domain='bar.com')
with pytest.raises(Cookie.URLMismatch):
assert cookie.http_return_ok(self.url)
Compliant client side session cookie behavior In summary this patch does: * Follow the defined rules for cookies when: - receiving a cookie (process the attributes) - storing a cookie (store cookie + attributes) - sending a cookie + validate the cookie domain against the request URL + validate the cookie path against the request URL + validate the cookie expiration + if valid then send only the cookie, no attribtues * Modifies how a request URL is stored during a XMLRPC request/response sequence. * Refactors a bit of the request/response logic to allow for making the decision whether to send a session cookie instead of full Kerberous auth easier. * The server now includes expiration information in the session cookie it sends to the client. The server always had the information available to prevent using an expired session cookie. Now that expiration timestamp is returned to the client as well and now the client will not send an expired session cookie back to the server. * Adds a new module and unit test for cookies (see below) Formerly we were always returning the session cookie no matter what the domain or path was in the URL. We were also sending the cookie attributes which are for the client only (used to determine if to return a cookie). The attributes are not meant to be sent to the server and the previous behavior was a protocol violation. We also were not checking the cookie expiration. Cookie library issues: We need a library to create, parse, manipulate and format cookies both in a client context and a server context. Core Python has two cookie libraries, Cookie.py and cookielib.py. Why did we add a new cookie module instead of using either of these two core Python libaries? Cookie.py is designed for server side generation but can be used to parse cookies on the client. It's the library we were using in the server. However when I tried to use it in the client I discovered it has some serious bugs. There are 7 defined cookie elements, it fails to correctly parse 3 of the 7 elements which makes it unusable because we depend on those elements. Since Cookie.py was designed for server side cookie processing it's not hard to understand how fails to correctly parse a cookie because that's a client side need. (Cookie.py also has an awkward baroque API and is missing some useful functionality we would have to build on top of it). cookielib.py is designed for client side. It's fully featured and obeys all the RFC's. It would be great to use however it's tightly coupled with another core library, urllib2.py. The http request and response objects must be urllib2 objects. But we don't use urllib2, rather we use httplib because xmlrpclib uses httplib. I don't see a reason why a cookie library should be so tightly coupled to a protocol library, but it is and that means we can't use it (I tried to just pick some isolated entrypoints for our use but I kept hitting interaction/dependency problems). I decided to solve the cookie library problems by writing a minimal cookie library that does what we need and no more than that. It is a new module in ipapython shared by both client and server and comes with a new unit test. The module has plenty of documentation, no need to repeat it here. Request URL issues: We also had problems in rpc.py whereby information from the request which is needed when we process the response is not available. Most important was the requesting URL. It turns out that the way the class and object relationships are structured it's impossible to get this information. Someone else must have run into the same issue because there was a routine called reconstruct_url() which attempted to recreate the request URL from other available information. Unfortunately reconstruct_url() was not callable from inside the response handler. So I decided to store the information in the thread context and when the request is received extract it from the thread context. It's perhaps not an ideal solution but we do similar things elsewhere so at least it's consistent. I removed the reconstruct_url() function because the exact information is now in the context and trying to apply heuristics to recreate the url is probably not robust. Ticket https://fedorahosted.org/freeipa/ticket/3022
2012-12-04 17:20:17 -06:00
cookie = Cookie('color', 'blue', domain='bogus.com')
with pytest.raises(Cookie.URLMismatch):
assert cookie.http_return_ok(self.url)
Compliant client side session cookie behavior In summary this patch does: * Follow the defined rules for cookies when: - receiving a cookie (process the attributes) - storing a cookie (store cookie + attributes) - sending a cookie + validate the cookie domain against the request URL + validate the cookie path against the request URL + validate the cookie expiration + if valid then send only the cookie, no attribtues * Modifies how a request URL is stored during a XMLRPC request/response sequence. * Refactors a bit of the request/response logic to allow for making the decision whether to send a session cookie instead of full Kerberous auth easier. * The server now includes expiration information in the session cookie it sends to the client. The server always had the information available to prevent using an expired session cookie. Now that expiration timestamp is returned to the client as well and now the client will not send an expired session cookie back to the server. * Adds a new module and unit test for cookies (see below) Formerly we were always returning the session cookie no matter what the domain or path was in the URL. We were also sending the cookie attributes which are for the client only (used to determine if to return a cookie). The attributes are not meant to be sent to the server and the previous behavior was a protocol violation. We also were not checking the cookie expiration. Cookie library issues: We need a library to create, parse, manipulate and format cookies both in a client context and a server context. Core Python has two cookie libraries, Cookie.py and cookielib.py. Why did we add a new cookie module instead of using either of these two core Python libaries? Cookie.py is designed for server side generation but can be used to parse cookies on the client. It's the library we were using in the server. However when I tried to use it in the client I discovered it has some serious bugs. There are 7 defined cookie elements, it fails to correctly parse 3 of the 7 elements which makes it unusable because we depend on those elements. Since Cookie.py was designed for server side cookie processing it's not hard to understand how fails to correctly parse a cookie because that's a client side need. (Cookie.py also has an awkward baroque API and is missing some useful functionality we would have to build on top of it). cookielib.py is designed for client side. It's fully featured and obeys all the RFC's. It would be great to use however it's tightly coupled with another core library, urllib2.py. The http request and response objects must be urllib2 objects. But we don't use urllib2, rather we use httplib because xmlrpclib uses httplib. I don't see a reason why a cookie library should be so tightly coupled to a protocol library, but it is and that means we can't use it (I tried to just pick some isolated entrypoints for our use but I kept hitting interaction/dependency problems). I decided to solve the cookie library problems by writing a minimal cookie library that does what we need and no more than that. It is a new module in ipapython shared by both client and server and comes with a new unit test. The module has plenty of documentation, no need to repeat it here. Request URL issues: We also had problems in rpc.py whereby information from the request which is needed when we process the response is not available. Most important was the requesting URL. It turns out that the way the class and object relationships are structured it's impossible to get this information. Someone else must have run into the same issue because there was a routine called reconstruct_url() which attempted to recreate the request URL from other available information. Unfortunately reconstruct_url() was not callable from inside the response handler. So I decided to store the information in the thread context and when the request is received extract it from the thread context. It's perhaps not an ideal solution but we do similar things elsewhere so at least it's consistent. I removed the reconstruct_url() function because the exact information is now in the context and trying to apply heuristics to recreate the url is probably not robust. Ticket https://fedorahosted.org/freeipa/ticket/3022
2012-12-04 17:20:17 -06:00
cookie = Cookie('color', 'blue', domain='www.foo.bar.com')
with pytest.raises(Cookie.URLMismatch):
assert cookie.http_return_ok('http://192.168.1.1/one/two')
Compliant client side session cookie behavior In summary this patch does: * Follow the defined rules for cookies when: - receiving a cookie (process the attributes) - storing a cookie (store cookie + attributes) - sending a cookie + validate the cookie domain against the request URL + validate the cookie path against the request URL + validate the cookie expiration + if valid then send only the cookie, no attribtues * Modifies how a request URL is stored during a XMLRPC request/response sequence. * Refactors a bit of the request/response logic to allow for making the decision whether to send a session cookie instead of full Kerberous auth easier. * The server now includes expiration information in the session cookie it sends to the client. The server always had the information available to prevent using an expired session cookie. Now that expiration timestamp is returned to the client as well and now the client will not send an expired session cookie back to the server. * Adds a new module and unit test for cookies (see below) Formerly we were always returning the session cookie no matter what the domain or path was in the URL. We were also sending the cookie attributes which are for the client only (used to determine if to return a cookie). The attributes are not meant to be sent to the server and the previous behavior was a protocol violation. We also were not checking the cookie expiration. Cookie library issues: We need a library to create, parse, manipulate and format cookies both in a client context and a server context. Core Python has two cookie libraries, Cookie.py and cookielib.py. Why did we add a new cookie module instead of using either of these two core Python libaries? Cookie.py is designed for server side generation but can be used to parse cookies on the client. It's the library we were using in the server. However when I tried to use it in the client I discovered it has some serious bugs. There are 7 defined cookie elements, it fails to correctly parse 3 of the 7 elements which makes it unusable because we depend on those elements. Since Cookie.py was designed for server side cookie processing it's not hard to understand how fails to correctly parse a cookie because that's a client side need. (Cookie.py also has an awkward baroque API and is missing some useful functionality we would have to build on top of it). cookielib.py is designed for client side. It's fully featured and obeys all the RFC's. It would be great to use however it's tightly coupled with another core library, urllib2.py. The http request and response objects must be urllib2 objects. But we don't use urllib2, rather we use httplib because xmlrpclib uses httplib. I don't see a reason why a cookie library should be so tightly coupled to a protocol library, but it is and that means we can't use it (I tried to just pick some isolated entrypoints for our use but I kept hitting interaction/dependency problems). I decided to solve the cookie library problems by writing a minimal cookie library that does what we need and no more than that. It is a new module in ipapython shared by both client and server and comes with a new unit test. The module has plenty of documentation, no need to repeat it here. Request URL issues: We also had problems in rpc.py whereby information from the request which is needed when we process the response is not available. Most important was the requesting URL. It turns out that the way the class and object relationships are structured it's impossible to get this information. Someone else must have run into the same issue because there was a routine called reconstruct_url() which attempted to recreate the request URL from other available information. Unfortunately reconstruct_url() was not callable from inside the response handler. So I decided to store the information in the thread context and when the request is received extract it from the thread context. It's perhaps not an ideal solution but we do similar things elsewhere so at least it's consistent. I removed the reconstruct_url() function because the exact information is now in the context and trying to apply heuristics to recreate the url is probably not robust. Ticket https://fedorahosted.org/freeipa/ticket/3022
2012-12-04 17:20:17 -06:00
def test_path(self):
cookie = Cookie('color', 'blue')
assert cookie.http_return_ok(self.url)
Compliant client side session cookie behavior In summary this patch does: * Follow the defined rules for cookies when: - receiving a cookie (process the attributes) - storing a cookie (store cookie + attributes) - sending a cookie + validate the cookie domain against the request URL + validate the cookie path against the request URL + validate the cookie expiration + if valid then send only the cookie, no attribtues * Modifies how a request URL is stored during a XMLRPC request/response sequence. * Refactors a bit of the request/response logic to allow for making the decision whether to send a session cookie instead of full Kerberous auth easier. * The server now includes expiration information in the session cookie it sends to the client. The server always had the information available to prevent using an expired session cookie. Now that expiration timestamp is returned to the client as well and now the client will not send an expired session cookie back to the server. * Adds a new module and unit test for cookies (see below) Formerly we were always returning the session cookie no matter what the domain or path was in the URL. We were also sending the cookie attributes which are for the client only (used to determine if to return a cookie). The attributes are not meant to be sent to the server and the previous behavior was a protocol violation. We also were not checking the cookie expiration. Cookie library issues: We need a library to create, parse, manipulate and format cookies both in a client context and a server context. Core Python has two cookie libraries, Cookie.py and cookielib.py. Why did we add a new cookie module instead of using either of these two core Python libaries? Cookie.py is designed for server side generation but can be used to parse cookies on the client. It's the library we were using in the server. However when I tried to use it in the client I discovered it has some serious bugs. There are 7 defined cookie elements, it fails to correctly parse 3 of the 7 elements which makes it unusable because we depend on those elements. Since Cookie.py was designed for server side cookie processing it's not hard to understand how fails to correctly parse a cookie because that's a client side need. (Cookie.py also has an awkward baroque API and is missing some useful functionality we would have to build on top of it). cookielib.py is designed for client side. It's fully featured and obeys all the RFC's. It would be great to use however it's tightly coupled with another core library, urllib2.py. The http request and response objects must be urllib2 objects. But we don't use urllib2, rather we use httplib because xmlrpclib uses httplib. I don't see a reason why a cookie library should be so tightly coupled to a protocol library, but it is and that means we can't use it (I tried to just pick some isolated entrypoints for our use but I kept hitting interaction/dependency problems). I decided to solve the cookie library problems by writing a minimal cookie library that does what we need and no more than that. It is a new module in ipapython shared by both client and server and comes with a new unit test. The module has plenty of documentation, no need to repeat it here. Request URL issues: We also had problems in rpc.py whereby information from the request which is needed when we process the response is not available. Most important was the requesting URL. It turns out that the way the class and object relationships are structured it's impossible to get this information. Someone else must have run into the same issue because there was a routine called reconstruct_url() which attempted to recreate the request URL from other available information. Unfortunately reconstruct_url() was not callable from inside the response handler. So I decided to store the information in the thread context and when the request is received extract it from the thread context. It's perhaps not an ideal solution but we do similar things elsewhere so at least it's consistent. I removed the reconstruct_url() function because the exact information is now in the context and trying to apply heuristics to recreate the url is probably not robust. Ticket https://fedorahosted.org/freeipa/ticket/3022
2012-12-04 17:20:17 -06:00
cookie = Cookie('color', 'blue', path='/')
assert cookie.http_return_ok(self.url)
Compliant client side session cookie behavior In summary this patch does: * Follow the defined rules for cookies when: - receiving a cookie (process the attributes) - storing a cookie (store cookie + attributes) - sending a cookie + validate the cookie domain against the request URL + validate the cookie path against the request URL + validate the cookie expiration + if valid then send only the cookie, no attribtues * Modifies how a request URL is stored during a XMLRPC request/response sequence. * Refactors a bit of the request/response logic to allow for making the decision whether to send a session cookie instead of full Kerberous auth easier. * The server now includes expiration information in the session cookie it sends to the client. The server always had the information available to prevent using an expired session cookie. Now that expiration timestamp is returned to the client as well and now the client will not send an expired session cookie back to the server. * Adds a new module and unit test for cookies (see below) Formerly we were always returning the session cookie no matter what the domain or path was in the URL. We were also sending the cookie attributes which are for the client only (used to determine if to return a cookie). The attributes are not meant to be sent to the server and the previous behavior was a protocol violation. We also were not checking the cookie expiration. Cookie library issues: We need a library to create, parse, manipulate and format cookies both in a client context and a server context. Core Python has two cookie libraries, Cookie.py and cookielib.py. Why did we add a new cookie module instead of using either of these two core Python libaries? Cookie.py is designed for server side generation but can be used to parse cookies on the client. It's the library we were using in the server. However when I tried to use it in the client I discovered it has some serious bugs. There are 7 defined cookie elements, it fails to correctly parse 3 of the 7 elements which makes it unusable because we depend on those elements. Since Cookie.py was designed for server side cookie processing it's not hard to understand how fails to correctly parse a cookie because that's a client side need. (Cookie.py also has an awkward baroque API and is missing some useful functionality we would have to build on top of it). cookielib.py is designed for client side. It's fully featured and obeys all the RFC's. It would be great to use however it's tightly coupled with another core library, urllib2.py. The http request and response objects must be urllib2 objects. But we don't use urllib2, rather we use httplib because xmlrpclib uses httplib. I don't see a reason why a cookie library should be so tightly coupled to a protocol library, but it is and that means we can't use it (I tried to just pick some isolated entrypoints for our use but I kept hitting interaction/dependency problems). I decided to solve the cookie library problems by writing a minimal cookie library that does what we need and no more than that. It is a new module in ipapython shared by both client and server and comes with a new unit test. The module has plenty of documentation, no need to repeat it here. Request URL issues: We also had problems in rpc.py whereby information from the request which is needed when we process the response is not available. Most important was the requesting URL. It turns out that the way the class and object relationships are structured it's impossible to get this information. Someone else must have run into the same issue because there was a routine called reconstruct_url() which attempted to recreate the request URL from other available information. Unfortunately reconstruct_url() was not callable from inside the response handler. So I decided to store the information in the thread context and when the request is received extract it from the thread context. It's perhaps not an ideal solution but we do similar things elsewhere so at least it's consistent. I removed the reconstruct_url() function because the exact information is now in the context and trying to apply heuristics to recreate the url is probably not robust. Ticket https://fedorahosted.org/freeipa/ticket/3022
2012-12-04 17:20:17 -06:00
cookie = Cookie('color', 'blue', path='/one')
assert cookie.http_return_ok(self.url)
Compliant client side session cookie behavior In summary this patch does: * Follow the defined rules for cookies when: - receiving a cookie (process the attributes) - storing a cookie (store cookie + attributes) - sending a cookie + validate the cookie domain against the request URL + validate the cookie path against the request URL + validate the cookie expiration + if valid then send only the cookie, no attribtues * Modifies how a request URL is stored during a XMLRPC request/response sequence. * Refactors a bit of the request/response logic to allow for making the decision whether to send a session cookie instead of full Kerberous auth easier. * The server now includes expiration information in the session cookie it sends to the client. The server always had the information available to prevent using an expired session cookie. Now that expiration timestamp is returned to the client as well and now the client will not send an expired session cookie back to the server. * Adds a new module and unit test for cookies (see below) Formerly we were always returning the session cookie no matter what the domain or path was in the URL. We were also sending the cookie attributes which are for the client only (used to determine if to return a cookie). The attributes are not meant to be sent to the server and the previous behavior was a protocol violation. We also were not checking the cookie expiration. Cookie library issues: We need a library to create, parse, manipulate and format cookies both in a client context and a server context. Core Python has two cookie libraries, Cookie.py and cookielib.py. Why did we add a new cookie module instead of using either of these two core Python libaries? Cookie.py is designed for server side generation but can be used to parse cookies on the client. It's the library we were using in the server. However when I tried to use it in the client I discovered it has some serious bugs. There are 7 defined cookie elements, it fails to correctly parse 3 of the 7 elements which makes it unusable because we depend on those elements. Since Cookie.py was designed for server side cookie processing it's not hard to understand how fails to correctly parse a cookie because that's a client side need. (Cookie.py also has an awkward baroque API and is missing some useful functionality we would have to build on top of it). cookielib.py is designed for client side. It's fully featured and obeys all the RFC's. It would be great to use however it's tightly coupled with another core library, urllib2.py. The http request and response objects must be urllib2 objects. But we don't use urllib2, rather we use httplib because xmlrpclib uses httplib. I don't see a reason why a cookie library should be so tightly coupled to a protocol library, but it is and that means we can't use it (I tried to just pick some isolated entrypoints for our use but I kept hitting interaction/dependency problems). I decided to solve the cookie library problems by writing a minimal cookie library that does what we need and no more than that. It is a new module in ipapython shared by both client and server and comes with a new unit test. The module has plenty of documentation, no need to repeat it here. Request URL issues: We also had problems in rpc.py whereby information from the request which is needed when we process the response is not available. Most important was the requesting URL. It turns out that the way the class and object relationships are structured it's impossible to get this information. Someone else must have run into the same issue because there was a routine called reconstruct_url() which attempted to recreate the request URL from other available information. Unfortunately reconstruct_url() was not callable from inside the response handler. So I decided to store the information in the thread context and when the request is received extract it from the thread context. It's perhaps not an ideal solution but we do similar things elsewhere so at least it's consistent. I removed the reconstruct_url() function because the exact information is now in the context and trying to apply heuristics to recreate the url is probably not robust. Ticket https://fedorahosted.org/freeipa/ticket/3022
2012-12-04 17:20:17 -06:00
cookie = Cookie('color', 'blue', path='/oneX')
with pytest.raises(Cookie.URLMismatch):
assert cookie.http_return_ok(self.url)
Compliant client side session cookie behavior In summary this patch does: * Follow the defined rules for cookies when: - receiving a cookie (process the attributes) - storing a cookie (store cookie + attributes) - sending a cookie + validate the cookie domain against the request URL + validate the cookie path against the request URL + validate the cookie expiration + if valid then send only the cookie, no attribtues * Modifies how a request URL is stored during a XMLRPC request/response sequence. * Refactors a bit of the request/response logic to allow for making the decision whether to send a session cookie instead of full Kerberous auth easier. * The server now includes expiration information in the session cookie it sends to the client. The server always had the information available to prevent using an expired session cookie. Now that expiration timestamp is returned to the client as well and now the client will not send an expired session cookie back to the server. * Adds a new module and unit test for cookies (see below) Formerly we were always returning the session cookie no matter what the domain or path was in the URL. We were also sending the cookie attributes which are for the client only (used to determine if to return a cookie). The attributes are not meant to be sent to the server and the previous behavior was a protocol violation. We also were not checking the cookie expiration. Cookie library issues: We need a library to create, parse, manipulate and format cookies both in a client context and a server context. Core Python has two cookie libraries, Cookie.py and cookielib.py. Why did we add a new cookie module instead of using either of these two core Python libaries? Cookie.py is designed for server side generation but can be used to parse cookies on the client. It's the library we were using in the server. However when I tried to use it in the client I discovered it has some serious bugs. There are 7 defined cookie elements, it fails to correctly parse 3 of the 7 elements which makes it unusable because we depend on those elements. Since Cookie.py was designed for server side cookie processing it's not hard to understand how fails to correctly parse a cookie because that's a client side need. (Cookie.py also has an awkward baroque API and is missing some useful functionality we would have to build on top of it). cookielib.py is designed for client side. It's fully featured and obeys all the RFC's. It would be great to use however it's tightly coupled with another core library, urllib2.py. The http request and response objects must be urllib2 objects. But we don't use urllib2, rather we use httplib because xmlrpclib uses httplib. I don't see a reason why a cookie library should be so tightly coupled to a protocol library, but it is and that means we can't use it (I tried to just pick some isolated entrypoints for our use but I kept hitting interaction/dependency problems). I decided to solve the cookie library problems by writing a minimal cookie library that does what we need and no more than that. It is a new module in ipapython shared by both client and server and comes with a new unit test. The module has plenty of documentation, no need to repeat it here. Request URL issues: We also had problems in rpc.py whereby information from the request which is needed when we process the response is not available. Most important was the requesting URL. It turns out that the way the class and object relationships are structured it's impossible to get this information. Someone else must have run into the same issue because there was a routine called reconstruct_url() which attempted to recreate the request URL from other available information. Unfortunately reconstruct_url() was not callable from inside the response handler. So I decided to store the information in the thread context and when the request is received extract it from the thread context. It's perhaps not an ideal solution but we do similar things elsewhere so at least it's consistent. I removed the reconstruct_url() function because the exact information is now in the context and trying to apply heuristics to recreate the url is probably not robust. Ticket https://fedorahosted.org/freeipa/ticket/3022
2012-12-04 17:20:17 -06:00
def test_expires(self):
now = datetime.datetime.utcnow().replace(microsecond=0)
# expires 1 day from now
expires = now + datetime.timedelta(days=1)
cookie = Cookie('color', 'blue', expires=expires)
assert cookie.http_return_ok(self.url)
Compliant client side session cookie behavior In summary this patch does: * Follow the defined rules for cookies when: - receiving a cookie (process the attributes) - storing a cookie (store cookie + attributes) - sending a cookie + validate the cookie domain against the request URL + validate the cookie path against the request URL + validate the cookie expiration + if valid then send only the cookie, no attribtues * Modifies how a request URL is stored during a XMLRPC request/response sequence. * Refactors a bit of the request/response logic to allow for making the decision whether to send a session cookie instead of full Kerberous auth easier. * The server now includes expiration information in the session cookie it sends to the client. The server always had the information available to prevent using an expired session cookie. Now that expiration timestamp is returned to the client as well and now the client will not send an expired session cookie back to the server. * Adds a new module and unit test for cookies (see below) Formerly we were always returning the session cookie no matter what the domain or path was in the URL. We were also sending the cookie attributes which are for the client only (used to determine if to return a cookie). The attributes are not meant to be sent to the server and the previous behavior was a protocol violation. We also were not checking the cookie expiration. Cookie library issues: We need a library to create, parse, manipulate and format cookies both in a client context and a server context. Core Python has two cookie libraries, Cookie.py and cookielib.py. Why did we add a new cookie module instead of using either of these two core Python libaries? Cookie.py is designed for server side generation but can be used to parse cookies on the client. It's the library we were using in the server. However when I tried to use it in the client I discovered it has some serious bugs. There are 7 defined cookie elements, it fails to correctly parse 3 of the 7 elements which makes it unusable because we depend on those elements. Since Cookie.py was designed for server side cookie processing it's not hard to understand how fails to correctly parse a cookie because that's a client side need. (Cookie.py also has an awkward baroque API and is missing some useful functionality we would have to build on top of it). cookielib.py is designed for client side. It's fully featured and obeys all the RFC's. It would be great to use however it's tightly coupled with another core library, urllib2.py. The http request and response objects must be urllib2 objects. But we don't use urllib2, rather we use httplib because xmlrpclib uses httplib. I don't see a reason why a cookie library should be so tightly coupled to a protocol library, but it is and that means we can't use it (I tried to just pick some isolated entrypoints for our use but I kept hitting interaction/dependency problems). I decided to solve the cookie library problems by writing a minimal cookie library that does what we need and no more than that. It is a new module in ipapython shared by both client and server and comes with a new unit test. The module has plenty of documentation, no need to repeat it here. Request URL issues: We also had problems in rpc.py whereby information from the request which is needed when we process the response is not available. Most important was the requesting URL. It turns out that the way the class and object relationships are structured it's impossible to get this information. Someone else must have run into the same issue because there was a routine called reconstruct_url() which attempted to recreate the request URL from other available information. Unfortunately reconstruct_url() was not callable from inside the response handler. So I decided to store the information in the thread context and when the request is received extract it from the thread context. It's perhaps not an ideal solution but we do similar things elsewhere so at least it's consistent. I removed the reconstruct_url() function because the exact information is now in the context and trying to apply heuristics to recreate the url is probably not robust. Ticket https://fedorahosted.org/freeipa/ticket/3022
2012-12-04 17:20:17 -06:00
# expired 1 day ago
expires = now + datetime.timedelta(days=-1)
cookie = Cookie('color', 'blue', expires=expires)
with pytest.raises(Cookie.Expired):
assert cookie.http_return_ok(self.url)
Compliant client side session cookie behavior In summary this patch does: * Follow the defined rules for cookies when: - receiving a cookie (process the attributes) - storing a cookie (store cookie + attributes) - sending a cookie + validate the cookie domain against the request URL + validate the cookie path against the request URL + validate the cookie expiration + if valid then send only the cookie, no attribtues * Modifies how a request URL is stored during a XMLRPC request/response sequence. * Refactors a bit of the request/response logic to allow for making the decision whether to send a session cookie instead of full Kerberous auth easier. * The server now includes expiration information in the session cookie it sends to the client. The server always had the information available to prevent using an expired session cookie. Now that expiration timestamp is returned to the client as well and now the client will not send an expired session cookie back to the server. * Adds a new module and unit test for cookies (see below) Formerly we were always returning the session cookie no matter what the domain or path was in the URL. We were also sending the cookie attributes which are for the client only (used to determine if to return a cookie). The attributes are not meant to be sent to the server and the previous behavior was a protocol violation. We also were not checking the cookie expiration. Cookie library issues: We need a library to create, parse, manipulate and format cookies both in a client context and a server context. Core Python has two cookie libraries, Cookie.py and cookielib.py. Why did we add a new cookie module instead of using either of these two core Python libaries? Cookie.py is designed for server side generation but can be used to parse cookies on the client. It's the library we were using in the server. However when I tried to use it in the client I discovered it has some serious bugs. There are 7 defined cookie elements, it fails to correctly parse 3 of the 7 elements which makes it unusable because we depend on those elements. Since Cookie.py was designed for server side cookie processing it's not hard to understand how fails to correctly parse a cookie because that's a client side need. (Cookie.py also has an awkward baroque API and is missing some useful functionality we would have to build on top of it). cookielib.py is designed for client side. It's fully featured and obeys all the RFC's. It would be great to use however it's tightly coupled with another core library, urllib2.py. The http request and response objects must be urllib2 objects. But we don't use urllib2, rather we use httplib because xmlrpclib uses httplib. I don't see a reason why a cookie library should be so tightly coupled to a protocol library, but it is and that means we can't use it (I tried to just pick some isolated entrypoints for our use but I kept hitting interaction/dependency problems). I decided to solve the cookie library problems by writing a minimal cookie library that does what we need and no more than that. It is a new module in ipapython shared by both client and server and comes with a new unit test. The module has plenty of documentation, no need to repeat it here. Request URL issues: We also had problems in rpc.py whereby information from the request which is needed when we process the response is not available. Most important was the requesting URL. It turns out that the way the class and object relationships are structured it's impossible to get this information. Someone else must have run into the same issue because there was a routine called reconstruct_url() which attempted to recreate the request URL from other available information. Unfortunately reconstruct_url() was not callable from inside the response handler. So I decided to store the information in the thread context and when the request is received extract it from the thread context. It's perhaps not an ideal solution but we do similar things elsewhere so at least it's consistent. I removed the reconstruct_url() function because the exact information is now in the context and trying to apply heuristics to recreate the url is probably not robust. Ticket https://fedorahosted.org/freeipa/ticket/3022
2012-12-04 17:20:17 -06:00
def test_httponly(self):
cookie = Cookie('color', 'blue', httponly=True)
assert cookie.http_return_ok('http://example.com')
assert cookie.http_return_ok('https://example.com')
Compliant client side session cookie behavior In summary this patch does: * Follow the defined rules for cookies when: - receiving a cookie (process the attributes) - storing a cookie (store cookie + attributes) - sending a cookie + validate the cookie domain against the request URL + validate the cookie path against the request URL + validate the cookie expiration + if valid then send only the cookie, no attribtues * Modifies how a request URL is stored during a XMLRPC request/response sequence. * Refactors a bit of the request/response logic to allow for making the decision whether to send a session cookie instead of full Kerberous auth easier. * The server now includes expiration information in the session cookie it sends to the client. The server always had the information available to prevent using an expired session cookie. Now that expiration timestamp is returned to the client as well and now the client will not send an expired session cookie back to the server. * Adds a new module and unit test for cookies (see below) Formerly we were always returning the session cookie no matter what the domain or path was in the URL. We were also sending the cookie attributes which are for the client only (used to determine if to return a cookie). The attributes are not meant to be sent to the server and the previous behavior was a protocol violation. We also were not checking the cookie expiration. Cookie library issues: We need a library to create, parse, manipulate and format cookies both in a client context and a server context. Core Python has two cookie libraries, Cookie.py and cookielib.py. Why did we add a new cookie module instead of using either of these two core Python libaries? Cookie.py is designed for server side generation but can be used to parse cookies on the client. It's the library we were using in the server. However when I tried to use it in the client I discovered it has some serious bugs. There are 7 defined cookie elements, it fails to correctly parse 3 of the 7 elements which makes it unusable because we depend on those elements. Since Cookie.py was designed for server side cookie processing it's not hard to understand how fails to correctly parse a cookie because that's a client side need. (Cookie.py also has an awkward baroque API and is missing some useful functionality we would have to build on top of it). cookielib.py is designed for client side. It's fully featured and obeys all the RFC's. It would be great to use however it's tightly coupled with another core library, urllib2.py. The http request and response objects must be urllib2 objects. But we don't use urllib2, rather we use httplib because xmlrpclib uses httplib. I don't see a reason why a cookie library should be so tightly coupled to a protocol library, but it is and that means we can't use it (I tried to just pick some isolated entrypoints for our use but I kept hitting interaction/dependency problems). I decided to solve the cookie library problems by writing a minimal cookie library that does what we need and no more than that. It is a new module in ipapython shared by both client and server and comes with a new unit test. The module has plenty of documentation, no need to repeat it here. Request URL issues: We also had problems in rpc.py whereby information from the request which is needed when we process the response is not available. Most important was the requesting URL. It turns out that the way the class and object relationships are structured it's impossible to get this information. Someone else must have run into the same issue because there was a routine called reconstruct_url() which attempted to recreate the request URL from other available information. Unfortunately reconstruct_url() was not callable from inside the response handler. So I decided to store the information in the thread context and when the request is received extract it from the thread context. It's perhaps not an ideal solution but we do similar things elsewhere so at least it's consistent. I removed the reconstruct_url() function because the exact information is now in the context and trying to apply heuristics to recreate the url is probably not robust. Ticket https://fedorahosted.org/freeipa/ticket/3022
2012-12-04 17:20:17 -06:00
with pytest.raises(Cookie.URLMismatch):
assert cookie.http_return_ok('ftp://example.com')
Compliant client side session cookie behavior In summary this patch does: * Follow the defined rules for cookies when: - receiving a cookie (process the attributes) - storing a cookie (store cookie + attributes) - sending a cookie + validate the cookie domain against the request URL + validate the cookie path against the request URL + validate the cookie expiration + if valid then send only the cookie, no attribtues * Modifies how a request URL is stored during a XMLRPC request/response sequence. * Refactors a bit of the request/response logic to allow for making the decision whether to send a session cookie instead of full Kerberous auth easier. * The server now includes expiration information in the session cookie it sends to the client. The server always had the information available to prevent using an expired session cookie. Now that expiration timestamp is returned to the client as well and now the client will not send an expired session cookie back to the server. * Adds a new module and unit test for cookies (see below) Formerly we were always returning the session cookie no matter what the domain or path was in the URL. We were also sending the cookie attributes which are for the client only (used to determine if to return a cookie). The attributes are not meant to be sent to the server and the previous behavior was a protocol violation. We also were not checking the cookie expiration. Cookie library issues: We need a library to create, parse, manipulate and format cookies both in a client context and a server context. Core Python has two cookie libraries, Cookie.py and cookielib.py. Why did we add a new cookie module instead of using either of these two core Python libaries? Cookie.py is designed for server side generation but can be used to parse cookies on the client. It's the library we were using in the server. However when I tried to use it in the client I discovered it has some serious bugs. There are 7 defined cookie elements, it fails to correctly parse 3 of the 7 elements which makes it unusable because we depend on those elements. Since Cookie.py was designed for server side cookie processing it's not hard to understand how fails to correctly parse a cookie because that's a client side need. (Cookie.py also has an awkward baroque API and is missing some useful functionality we would have to build on top of it). cookielib.py is designed for client side. It's fully featured and obeys all the RFC's. It would be great to use however it's tightly coupled with another core library, urllib2.py. The http request and response objects must be urllib2 objects. But we don't use urllib2, rather we use httplib because xmlrpclib uses httplib. I don't see a reason why a cookie library should be so tightly coupled to a protocol library, but it is and that means we can't use it (I tried to just pick some isolated entrypoints for our use but I kept hitting interaction/dependency problems). I decided to solve the cookie library problems by writing a minimal cookie library that does what we need and no more than that. It is a new module in ipapython shared by both client and server and comes with a new unit test. The module has plenty of documentation, no need to repeat it here. Request URL issues: We also had problems in rpc.py whereby information from the request which is needed when we process the response is not available. Most important was the requesting URL. It turns out that the way the class and object relationships are structured it's impossible to get this information. Someone else must have run into the same issue because there was a routine called reconstruct_url() which attempted to recreate the request URL from other available information. Unfortunately reconstruct_url() was not callable from inside the response handler. So I decided to store the information in the thread context and when the request is received extract it from the thread context. It's perhaps not an ideal solution but we do similar things elsewhere so at least it's consistent. I removed the reconstruct_url() function because the exact information is now in the context and trying to apply heuristics to recreate the url is probably not robust. Ticket https://fedorahosted.org/freeipa/ticket/3022
2012-12-04 17:20:17 -06:00
def test_secure(self):
cookie = Cookie('color', 'blue', secure=True)
assert cookie.http_return_ok('https://Xexample.com')
with pytest.raises(Cookie.URLMismatch):
assert cookie.http_return_ok('http://Xexample.com')
Compliant client side session cookie behavior In summary this patch does: * Follow the defined rules for cookies when: - receiving a cookie (process the attributes) - storing a cookie (store cookie + attributes) - sending a cookie + validate the cookie domain against the request URL + validate the cookie path against the request URL + validate the cookie expiration + if valid then send only the cookie, no attribtues * Modifies how a request URL is stored during a XMLRPC request/response sequence. * Refactors a bit of the request/response logic to allow for making the decision whether to send a session cookie instead of full Kerberous auth easier. * The server now includes expiration information in the session cookie it sends to the client. The server always had the information available to prevent using an expired session cookie. Now that expiration timestamp is returned to the client as well and now the client will not send an expired session cookie back to the server. * Adds a new module and unit test for cookies (see below) Formerly we were always returning the session cookie no matter what the domain or path was in the URL. We were also sending the cookie attributes which are for the client only (used to determine if to return a cookie). The attributes are not meant to be sent to the server and the previous behavior was a protocol violation. We also were not checking the cookie expiration. Cookie library issues: We need a library to create, parse, manipulate and format cookies both in a client context and a server context. Core Python has two cookie libraries, Cookie.py and cookielib.py. Why did we add a new cookie module instead of using either of these two core Python libaries? Cookie.py is designed for server side generation but can be used to parse cookies on the client. It's the library we were using in the server. However when I tried to use it in the client I discovered it has some serious bugs. There are 7 defined cookie elements, it fails to correctly parse 3 of the 7 elements which makes it unusable because we depend on those elements. Since Cookie.py was designed for server side cookie processing it's not hard to understand how fails to correctly parse a cookie because that's a client side need. (Cookie.py also has an awkward baroque API and is missing some useful functionality we would have to build on top of it). cookielib.py is designed for client side. It's fully featured and obeys all the RFC's. It would be great to use however it's tightly coupled with another core library, urllib2.py. The http request and response objects must be urllib2 objects. But we don't use urllib2, rather we use httplib because xmlrpclib uses httplib. I don't see a reason why a cookie library should be so tightly coupled to a protocol library, but it is and that means we can't use it (I tried to just pick some isolated entrypoints for our use but I kept hitting interaction/dependency problems). I decided to solve the cookie library problems by writing a minimal cookie library that does what we need and no more than that. It is a new module in ipapython shared by both client and server and comes with a new unit test. The module has plenty of documentation, no need to repeat it here. Request URL issues: We also had problems in rpc.py whereby information from the request which is needed when we process the response is not available. Most important was the requesting URL. It turns out that the way the class and object relationships are structured it's impossible to get this information. Someone else must have run into the same issue because there was a routine called reconstruct_url() which attempted to recreate the request URL from other available information. Unfortunately reconstruct_url() was not callable from inside the response handler. So I decided to store the information in the thread context and when the request is received extract it from the thread context. It's perhaps not an ideal solution but we do similar things elsewhere so at least it's consistent. I removed the reconstruct_url() function because the exact information is now in the context and trying to apply heuristics to recreate the url is probably not robust. Ticket https://fedorahosted.org/freeipa/ticket/3022
2012-12-04 17:20:17 -06:00
class TestNormalization:
@pytest.fixture(autouse=True)
def normalization_setup(self):
Compliant client side session cookie behavior In summary this patch does: * Follow the defined rules for cookies when: - receiving a cookie (process the attributes) - storing a cookie (store cookie + attributes) - sending a cookie + validate the cookie domain against the request URL + validate the cookie path against the request URL + validate the cookie expiration + if valid then send only the cookie, no attribtues * Modifies how a request URL is stored during a XMLRPC request/response sequence. * Refactors a bit of the request/response logic to allow for making the decision whether to send a session cookie instead of full Kerberous auth easier. * The server now includes expiration information in the session cookie it sends to the client. The server always had the information available to prevent using an expired session cookie. Now that expiration timestamp is returned to the client as well and now the client will not send an expired session cookie back to the server. * Adds a new module and unit test for cookies (see below) Formerly we were always returning the session cookie no matter what the domain or path was in the URL. We were also sending the cookie attributes which are for the client only (used to determine if to return a cookie). The attributes are not meant to be sent to the server and the previous behavior was a protocol violation. We also were not checking the cookie expiration. Cookie library issues: We need a library to create, parse, manipulate and format cookies both in a client context and a server context. Core Python has two cookie libraries, Cookie.py and cookielib.py. Why did we add a new cookie module instead of using either of these two core Python libaries? Cookie.py is designed for server side generation but can be used to parse cookies on the client. It's the library we were using in the server. However when I tried to use it in the client I discovered it has some serious bugs. There are 7 defined cookie elements, it fails to correctly parse 3 of the 7 elements which makes it unusable because we depend on those elements. Since Cookie.py was designed for server side cookie processing it's not hard to understand how fails to correctly parse a cookie because that's a client side need. (Cookie.py also has an awkward baroque API and is missing some useful functionality we would have to build on top of it). cookielib.py is designed for client side. It's fully featured and obeys all the RFC's. It would be great to use however it's tightly coupled with another core library, urllib2.py. The http request and response objects must be urllib2 objects. But we don't use urllib2, rather we use httplib because xmlrpclib uses httplib. I don't see a reason why a cookie library should be so tightly coupled to a protocol library, but it is and that means we can't use it (I tried to just pick some isolated entrypoints for our use but I kept hitting interaction/dependency problems). I decided to solve the cookie library problems by writing a minimal cookie library that does what we need and no more than that. It is a new module in ipapython shared by both client and server and comes with a new unit test. The module has plenty of documentation, no need to repeat it here. Request URL issues: We also had problems in rpc.py whereby information from the request which is needed when we process the response is not available. Most important was the requesting URL. It turns out that the way the class and object relationships are structured it's impossible to get this information. Someone else must have run into the same issue because there was a routine called reconstruct_url() which attempted to recreate the request URL from other available information. Unfortunately reconstruct_url() was not callable from inside the response handler. So I decided to store the information in the thread context and when the request is received extract it from the thread context. It's perhaps not an ideal solution but we do similar things elsewhere so at least it's consistent. I removed the reconstruct_url() function because the exact information is now in the context and trying to apply heuristics to recreate the url is probably not robust. Ticket https://fedorahosted.org/freeipa/ticket/3022
2012-12-04 17:20:17 -06:00
# Force microseconds to zero because cookie timestamps only have second resolution
self.now = datetime.datetime.now(
tz=datetime.timezone.utc).replace(microsecond=0)
self.now_timestamp = datetime_from_utctimestamp(
self.now.utctimetuple(), units=1).timestamp()
self.now_string = email.utils.formatdate(self.now_timestamp, usegmt=True)
Compliant client side session cookie behavior In summary this patch does: * Follow the defined rules for cookies when: - receiving a cookie (process the attributes) - storing a cookie (store cookie + attributes) - sending a cookie + validate the cookie domain against the request URL + validate the cookie path against the request URL + validate the cookie expiration + if valid then send only the cookie, no attribtues * Modifies how a request URL is stored during a XMLRPC request/response sequence. * Refactors a bit of the request/response logic to allow for making the decision whether to send a session cookie instead of full Kerberous auth easier. * The server now includes expiration information in the session cookie it sends to the client. The server always had the information available to prevent using an expired session cookie. Now that expiration timestamp is returned to the client as well and now the client will not send an expired session cookie back to the server. * Adds a new module and unit test for cookies (see below) Formerly we were always returning the session cookie no matter what the domain or path was in the URL. We were also sending the cookie attributes which are for the client only (used to determine if to return a cookie). The attributes are not meant to be sent to the server and the previous behavior was a protocol violation. We also were not checking the cookie expiration. Cookie library issues: We need a library to create, parse, manipulate and format cookies both in a client context and a server context. Core Python has two cookie libraries, Cookie.py and cookielib.py. Why did we add a new cookie module instead of using either of these two core Python libaries? Cookie.py is designed for server side generation but can be used to parse cookies on the client. It's the library we were using in the server. However when I tried to use it in the client I discovered it has some serious bugs. There are 7 defined cookie elements, it fails to correctly parse 3 of the 7 elements which makes it unusable because we depend on those elements. Since Cookie.py was designed for server side cookie processing it's not hard to understand how fails to correctly parse a cookie because that's a client side need. (Cookie.py also has an awkward baroque API and is missing some useful functionality we would have to build on top of it). cookielib.py is designed for client side. It's fully featured and obeys all the RFC's. It would be great to use however it's tightly coupled with another core library, urllib2.py. The http request and response objects must be urllib2 objects. But we don't use urllib2, rather we use httplib because xmlrpclib uses httplib. I don't see a reason why a cookie library should be so tightly coupled to a protocol library, but it is and that means we can't use it (I tried to just pick some isolated entrypoints for our use but I kept hitting interaction/dependency problems). I decided to solve the cookie library problems by writing a minimal cookie library that does what we need and no more than that. It is a new module in ipapython shared by both client and server and comes with a new unit test. The module has plenty of documentation, no need to repeat it here. Request URL issues: We also had problems in rpc.py whereby information from the request which is needed when we process the response is not available. Most important was the requesting URL. It turns out that the way the class and object relationships are structured it's impossible to get this information. Someone else must have run into the same issue because there was a routine called reconstruct_url() which attempted to recreate the request URL from other available information. Unfortunately reconstruct_url() was not callable from inside the response handler. So I decided to store the information in the thread context and when the request is received extract it from the thread context. It's perhaps not an ideal solution but we do similar things elsewhere so at least it's consistent. I removed the reconstruct_url() function because the exact information is now in the context and trying to apply heuristics to recreate the url is probably not robust. Ticket https://fedorahosted.org/freeipa/ticket/3022
2012-12-04 17:20:17 -06:00
self.max_age = 3600 # 1 hour
self.age_expiration = self.now + datetime.timedelta(seconds=self.max_age)
self.age_timestamp = datetime_from_utctimestamp(
self.age_expiration.utctimetuple(), units=1).timestamp()
self.age_string = email.utils.formatdate(self.age_timestamp, usegmt=True)
Compliant client side session cookie behavior In summary this patch does: * Follow the defined rules for cookies when: - receiving a cookie (process the attributes) - storing a cookie (store cookie + attributes) - sending a cookie + validate the cookie domain against the request URL + validate the cookie path against the request URL + validate the cookie expiration + if valid then send only the cookie, no attribtues * Modifies how a request URL is stored during a XMLRPC request/response sequence. * Refactors a bit of the request/response logic to allow for making the decision whether to send a session cookie instead of full Kerberous auth easier. * The server now includes expiration information in the session cookie it sends to the client. The server always had the information available to prevent using an expired session cookie. Now that expiration timestamp is returned to the client as well and now the client will not send an expired session cookie back to the server. * Adds a new module and unit test for cookies (see below) Formerly we were always returning the session cookie no matter what the domain or path was in the URL. We were also sending the cookie attributes which are for the client only (used to determine if to return a cookie). The attributes are not meant to be sent to the server and the previous behavior was a protocol violation. We also were not checking the cookie expiration. Cookie library issues: We need a library to create, parse, manipulate and format cookies both in a client context and a server context. Core Python has two cookie libraries, Cookie.py and cookielib.py. Why did we add a new cookie module instead of using either of these two core Python libaries? Cookie.py is designed for server side generation but can be used to parse cookies on the client. It's the library we were using in the server. However when I tried to use it in the client I discovered it has some serious bugs. There are 7 defined cookie elements, it fails to correctly parse 3 of the 7 elements which makes it unusable because we depend on those elements. Since Cookie.py was designed for server side cookie processing it's not hard to understand how fails to correctly parse a cookie because that's a client side need. (Cookie.py also has an awkward baroque API and is missing some useful functionality we would have to build on top of it). cookielib.py is designed for client side. It's fully featured and obeys all the RFC's. It would be great to use however it's tightly coupled with another core library, urllib2.py. The http request and response objects must be urllib2 objects. But we don't use urllib2, rather we use httplib because xmlrpclib uses httplib. I don't see a reason why a cookie library should be so tightly coupled to a protocol library, but it is and that means we can't use it (I tried to just pick some isolated entrypoints for our use but I kept hitting interaction/dependency problems). I decided to solve the cookie library problems by writing a minimal cookie library that does what we need and no more than that. It is a new module in ipapython shared by both client and server and comes with a new unit test. The module has plenty of documentation, no need to repeat it here. Request URL issues: We also had problems in rpc.py whereby information from the request which is needed when we process the response is not available. Most important was the requesting URL. It turns out that the way the class and object relationships are structured it's impossible to get this information. Someone else must have run into the same issue because there was a routine called reconstruct_url() which attempted to recreate the request URL from other available information. Unfortunately reconstruct_url() was not callable from inside the response handler. So I decided to store the information in the thread context and when the request is received extract it from the thread context. It's perhaps not an ideal solution but we do similar things elsewhere so at least it's consistent. I removed the reconstruct_url() function because the exact information is now in the context and trying to apply heuristics to recreate the url is probably not robust. Ticket https://fedorahosted.org/freeipa/ticket/3022
2012-12-04 17:20:17 -06:00
self.expires = self.now + datetime.timedelta(days=1) # 1 day
self.expires_timestamp = datetime_from_utctimestamp(
self.expires.utctimetuple(), units=1).timestamp()
self.expires_string = email.utils.formatdate(self.expires_timestamp, usegmt=True)
Compliant client side session cookie behavior In summary this patch does: * Follow the defined rules for cookies when: - receiving a cookie (process the attributes) - storing a cookie (store cookie + attributes) - sending a cookie + validate the cookie domain against the request URL + validate the cookie path against the request URL + validate the cookie expiration + if valid then send only the cookie, no attribtues * Modifies how a request URL is stored during a XMLRPC request/response sequence. * Refactors a bit of the request/response logic to allow for making the decision whether to send a session cookie instead of full Kerberous auth easier. * The server now includes expiration information in the session cookie it sends to the client. The server always had the information available to prevent using an expired session cookie. Now that expiration timestamp is returned to the client as well and now the client will not send an expired session cookie back to the server. * Adds a new module and unit test for cookies (see below) Formerly we were always returning the session cookie no matter what the domain or path was in the URL. We were also sending the cookie attributes which are for the client only (used to determine if to return a cookie). The attributes are not meant to be sent to the server and the previous behavior was a protocol violation. We also were not checking the cookie expiration. Cookie library issues: We need a library to create, parse, manipulate and format cookies both in a client context and a server context. Core Python has two cookie libraries, Cookie.py and cookielib.py. Why did we add a new cookie module instead of using either of these two core Python libaries? Cookie.py is designed for server side generation but can be used to parse cookies on the client. It's the library we were using in the server. However when I tried to use it in the client I discovered it has some serious bugs. There are 7 defined cookie elements, it fails to correctly parse 3 of the 7 elements which makes it unusable because we depend on those elements. Since Cookie.py was designed for server side cookie processing it's not hard to understand how fails to correctly parse a cookie because that's a client side need. (Cookie.py also has an awkward baroque API and is missing some useful functionality we would have to build on top of it). cookielib.py is designed for client side. It's fully featured and obeys all the RFC's. It would be great to use however it's tightly coupled with another core library, urllib2.py. The http request and response objects must be urllib2 objects. But we don't use urllib2, rather we use httplib because xmlrpclib uses httplib. I don't see a reason why a cookie library should be so tightly coupled to a protocol library, but it is and that means we can't use it (I tried to just pick some isolated entrypoints for our use but I kept hitting interaction/dependency problems). I decided to solve the cookie library problems by writing a minimal cookie library that does what we need and no more than that. It is a new module in ipapython shared by both client and server and comes with a new unit test. The module has plenty of documentation, no need to repeat it here. Request URL issues: We also had problems in rpc.py whereby information from the request which is needed when we process the response is not available. Most important was the requesting URL. It turns out that the way the class and object relationships are structured it's impossible to get this information. Someone else must have run into the same issue because there was a routine called reconstruct_url() which attempted to recreate the request URL from other available information. Unfortunately reconstruct_url() was not callable from inside the response handler. So I decided to store the information in the thread context and when the request is received extract it from the thread context. It's perhaps not an ideal solution but we do similar things elsewhere so at least it's consistent. I removed the reconstruct_url() function because the exact information is now in the context and trying to apply heuristics to recreate the url is probably not robust. Ticket https://fedorahosted.org/freeipa/ticket/3022
2012-12-04 17:20:17 -06:00
def test_path_normalization(self):
assert Cookie.normalize_url_path('') == '/'
assert Cookie.normalize_url_path('foo') == '/'
assert Cookie.normalize_url_path('foo/') == '/'
assert Cookie.normalize_url_path('/foo') == '/'
assert Cookie.normalize_url_path('/foo/') == '/foo'
assert Cookie.normalize_url_path('/Foo/bar') == '/foo'
assert Cookie.normalize_url_path('/foo/baR/') == '/foo/bar'
Compliant client side session cookie behavior In summary this patch does: * Follow the defined rules for cookies when: - receiving a cookie (process the attributes) - storing a cookie (store cookie + attributes) - sending a cookie + validate the cookie domain against the request URL + validate the cookie path against the request URL + validate the cookie expiration + if valid then send only the cookie, no attribtues * Modifies how a request URL is stored during a XMLRPC request/response sequence. * Refactors a bit of the request/response logic to allow for making the decision whether to send a session cookie instead of full Kerberous auth easier. * The server now includes expiration information in the session cookie it sends to the client. The server always had the information available to prevent using an expired session cookie. Now that expiration timestamp is returned to the client as well and now the client will not send an expired session cookie back to the server. * Adds a new module and unit test for cookies (see below) Formerly we were always returning the session cookie no matter what the domain or path was in the URL. We were also sending the cookie attributes which are for the client only (used to determine if to return a cookie). The attributes are not meant to be sent to the server and the previous behavior was a protocol violation. We also were not checking the cookie expiration. Cookie library issues: We need a library to create, parse, manipulate and format cookies both in a client context and a server context. Core Python has two cookie libraries, Cookie.py and cookielib.py. Why did we add a new cookie module instead of using either of these two core Python libaries? Cookie.py is designed for server side generation but can be used to parse cookies on the client. It's the library we were using in the server. However when I tried to use it in the client I discovered it has some serious bugs. There are 7 defined cookie elements, it fails to correctly parse 3 of the 7 elements which makes it unusable because we depend on those elements. Since Cookie.py was designed for server side cookie processing it's not hard to understand how fails to correctly parse a cookie because that's a client side need. (Cookie.py also has an awkward baroque API and is missing some useful functionality we would have to build on top of it). cookielib.py is designed for client side. It's fully featured and obeys all the RFC's. It would be great to use however it's tightly coupled with another core library, urllib2.py. The http request and response objects must be urllib2 objects. But we don't use urllib2, rather we use httplib because xmlrpclib uses httplib. I don't see a reason why a cookie library should be so tightly coupled to a protocol library, but it is and that means we can't use it (I tried to just pick some isolated entrypoints for our use but I kept hitting interaction/dependency problems). I decided to solve the cookie library problems by writing a minimal cookie library that does what we need and no more than that. It is a new module in ipapython shared by both client and server and comes with a new unit test. The module has plenty of documentation, no need to repeat it here. Request URL issues: We also had problems in rpc.py whereby information from the request which is needed when we process the response is not available. Most important was the requesting URL. It turns out that the way the class and object relationships are structured it's impossible to get this information. Someone else must have run into the same issue because there was a routine called reconstruct_url() which attempted to recreate the request URL from other available information. Unfortunately reconstruct_url() was not callable from inside the response handler. So I decided to store the information in the thread context and when the request is received extract it from the thread context. It's perhaps not an ideal solution but we do similar things elsewhere so at least it's consistent. I removed the reconstruct_url() function because the exact information is now in the context and trying to apply heuristics to recreate the url is probably not robust. Ticket https://fedorahosted.org/freeipa/ticket/3022
2012-12-04 17:20:17 -06:00
def test_normalization(self):
cookie = Cookie('color', 'blue', expires=self.expires)
cookie.timestamp = self.now_timestamp
assert cookie.domain is None
assert cookie.path is None
Compliant client side session cookie behavior In summary this patch does: * Follow the defined rules for cookies when: - receiving a cookie (process the attributes) - storing a cookie (store cookie + attributes) - sending a cookie + validate the cookie domain against the request URL + validate the cookie path against the request URL + validate the cookie expiration + if valid then send only the cookie, no attribtues * Modifies how a request URL is stored during a XMLRPC request/response sequence. * Refactors a bit of the request/response logic to allow for making the decision whether to send a session cookie instead of full Kerberous auth easier. * The server now includes expiration information in the session cookie it sends to the client. The server always had the information available to prevent using an expired session cookie. Now that expiration timestamp is returned to the client as well and now the client will not send an expired session cookie back to the server. * Adds a new module and unit test for cookies (see below) Formerly we were always returning the session cookie no matter what the domain or path was in the URL. We were also sending the cookie attributes which are for the client only (used to determine if to return a cookie). The attributes are not meant to be sent to the server and the previous behavior was a protocol violation. We also were not checking the cookie expiration. Cookie library issues: We need a library to create, parse, manipulate and format cookies both in a client context and a server context. Core Python has two cookie libraries, Cookie.py and cookielib.py. Why did we add a new cookie module instead of using either of these two core Python libaries? Cookie.py is designed for server side generation but can be used to parse cookies on the client. It's the library we were using in the server. However when I tried to use it in the client I discovered it has some serious bugs. There are 7 defined cookie elements, it fails to correctly parse 3 of the 7 elements which makes it unusable because we depend on those elements. Since Cookie.py was designed for server side cookie processing it's not hard to understand how fails to correctly parse a cookie because that's a client side need. (Cookie.py also has an awkward baroque API and is missing some useful functionality we would have to build on top of it). cookielib.py is designed for client side. It's fully featured and obeys all the RFC's. It would be great to use however it's tightly coupled with another core library, urllib2.py. The http request and response objects must be urllib2 objects. But we don't use urllib2, rather we use httplib because xmlrpclib uses httplib. I don't see a reason why a cookie library should be so tightly coupled to a protocol library, but it is and that means we can't use it (I tried to just pick some isolated entrypoints for our use but I kept hitting interaction/dependency problems). I decided to solve the cookie library problems by writing a minimal cookie library that does what we need and no more than that. It is a new module in ipapython shared by both client and server and comes with a new unit test. The module has plenty of documentation, no need to repeat it here. Request URL issues: We also had problems in rpc.py whereby information from the request which is needed when we process the response is not available. Most important was the requesting URL. It turns out that the way the class and object relationships are structured it's impossible to get this information. Someone else must have run into the same issue because there was a routine called reconstruct_url() which attempted to recreate the request URL from other available information. Unfortunately reconstruct_url() was not callable from inside the response handler. So I decided to store the information in the thread context and when the request is received extract it from the thread context. It's perhaps not an ideal solution but we do similar things elsewhere so at least it's consistent. I removed the reconstruct_url() function because the exact information is now in the context and trying to apply heuristics to recreate the url is probably not robust. Ticket https://fedorahosted.org/freeipa/ticket/3022
2012-12-04 17:20:17 -06:00
url = 'http://example.COM/foo'
cookie.normalize(url)
assert cookie.domain == 'example.com'
assert cookie.path == '/'
assert cookie.expires == self.expires
Compliant client side session cookie behavior In summary this patch does: * Follow the defined rules for cookies when: - receiving a cookie (process the attributes) - storing a cookie (store cookie + attributes) - sending a cookie + validate the cookie domain against the request URL + validate the cookie path against the request URL + validate the cookie expiration + if valid then send only the cookie, no attribtues * Modifies how a request URL is stored during a XMLRPC request/response sequence. * Refactors a bit of the request/response logic to allow for making the decision whether to send a session cookie instead of full Kerberous auth easier. * The server now includes expiration information in the session cookie it sends to the client. The server always had the information available to prevent using an expired session cookie. Now that expiration timestamp is returned to the client as well and now the client will not send an expired session cookie back to the server. * Adds a new module and unit test for cookies (see below) Formerly we were always returning the session cookie no matter what the domain or path was in the URL. We were also sending the cookie attributes which are for the client only (used to determine if to return a cookie). The attributes are not meant to be sent to the server and the previous behavior was a protocol violation. We also were not checking the cookie expiration. Cookie library issues: We need a library to create, parse, manipulate and format cookies both in a client context and a server context. Core Python has two cookie libraries, Cookie.py and cookielib.py. Why did we add a new cookie module instead of using either of these two core Python libaries? Cookie.py is designed for server side generation but can be used to parse cookies on the client. It's the library we were using in the server. However when I tried to use it in the client I discovered it has some serious bugs. There are 7 defined cookie elements, it fails to correctly parse 3 of the 7 elements which makes it unusable because we depend on those elements. Since Cookie.py was designed for server side cookie processing it's not hard to understand how fails to correctly parse a cookie because that's a client side need. (Cookie.py also has an awkward baroque API and is missing some useful functionality we would have to build on top of it). cookielib.py is designed for client side. It's fully featured and obeys all the RFC's. It would be great to use however it's tightly coupled with another core library, urllib2.py. The http request and response objects must be urllib2 objects. But we don't use urllib2, rather we use httplib because xmlrpclib uses httplib. I don't see a reason why a cookie library should be so tightly coupled to a protocol library, but it is and that means we can't use it (I tried to just pick some isolated entrypoints for our use but I kept hitting interaction/dependency problems). I decided to solve the cookie library problems by writing a minimal cookie library that does what we need and no more than that. It is a new module in ipapython shared by both client and server and comes with a new unit test. The module has plenty of documentation, no need to repeat it here. Request URL issues: We also had problems in rpc.py whereby information from the request which is needed when we process the response is not available. Most important was the requesting URL. It turns out that the way the class and object relationships are structured it's impossible to get this information. Someone else must have run into the same issue because there was a routine called reconstruct_url() which attempted to recreate the request URL from other available information. Unfortunately reconstruct_url() was not callable from inside the response handler. So I decided to store the information in the thread context and when the request is received extract it from the thread context. It's perhaps not an ideal solution but we do similar things elsewhere so at least it's consistent. I removed the reconstruct_url() function because the exact information is now in the context and trying to apply heuristics to recreate the url is probably not robust. Ticket https://fedorahosted.org/freeipa/ticket/3022
2012-12-04 17:20:17 -06:00
cookie = Cookie('color', 'blue', max_age=self.max_age)
cookie.timestamp = self.now_timestamp
assert cookie.domain is None
assert cookie.path is None
Compliant client side session cookie behavior In summary this patch does: * Follow the defined rules for cookies when: - receiving a cookie (process the attributes) - storing a cookie (store cookie + attributes) - sending a cookie + validate the cookie domain against the request URL + validate the cookie path against the request URL + validate the cookie expiration + if valid then send only the cookie, no attribtues * Modifies how a request URL is stored during a XMLRPC request/response sequence. * Refactors a bit of the request/response logic to allow for making the decision whether to send a session cookie instead of full Kerberous auth easier. * The server now includes expiration information in the session cookie it sends to the client. The server always had the information available to prevent using an expired session cookie. Now that expiration timestamp is returned to the client as well and now the client will not send an expired session cookie back to the server. * Adds a new module and unit test for cookies (see below) Formerly we were always returning the session cookie no matter what the domain or path was in the URL. We were also sending the cookie attributes which are for the client only (used to determine if to return a cookie). The attributes are not meant to be sent to the server and the previous behavior was a protocol violation. We also were not checking the cookie expiration. Cookie library issues: We need a library to create, parse, manipulate and format cookies both in a client context and a server context. Core Python has two cookie libraries, Cookie.py and cookielib.py. Why did we add a new cookie module instead of using either of these two core Python libaries? Cookie.py is designed for server side generation but can be used to parse cookies on the client. It's the library we were using in the server. However when I tried to use it in the client I discovered it has some serious bugs. There are 7 defined cookie elements, it fails to correctly parse 3 of the 7 elements which makes it unusable because we depend on those elements. Since Cookie.py was designed for server side cookie processing it's not hard to understand how fails to correctly parse a cookie because that's a client side need. (Cookie.py also has an awkward baroque API and is missing some useful functionality we would have to build on top of it). cookielib.py is designed for client side. It's fully featured and obeys all the RFC's. It would be great to use however it's tightly coupled with another core library, urllib2.py. The http request and response objects must be urllib2 objects. But we don't use urllib2, rather we use httplib because xmlrpclib uses httplib. I don't see a reason why a cookie library should be so tightly coupled to a protocol library, but it is and that means we can't use it (I tried to just pick some isolated entrypoints for our use but I kept hitting interaction/dependency problems). I decided to solve the cookie library problems by writing a minimal cookie library that does what we need and no more than that. It is a new module in ipapython shared by both client and server and comes with a new unit test. The module has plenty of documentation, no need to repeat it here. Request URL issues: We also had problems in rpc.py whereby information from the request which is needed when we process the response is not available. Most important was the requesting URL. It turns out that the way the class and object relationships are structured it's impossible to get this information. Someone else must have run into the same issue because there was a routine called reconstruct_url() which attempted to recreate the request URL from other available information. Unfortunately reconstruct_url() was not callable from inside the response handler. So I decided to store the information in the thread context and when the request is received extract it from the thread context. It's perhaps not an ideal solution but we do similar things elsewhere so at least it's consistent. I removed the reconstruct_url() function because the exact information is now in the context and trying to apply heuristics to recreate the url is probably not robust. Ticket https://fedorahosted.org/freeipa/ticket/3022
2012-12-04 17:20:17 -06:00
url = 'http://example.com/foo/'
cookie.normalize(url)
assert cookie.domain == 'example.com'
assert cookie.path == '/foo'
assert cookie.expires == self.age_expiration
Compliant client side session cookie behavior In summary this patch does: * Follow the defined rules for cookies when: - receiving a cookie (process the attributes) - storing a cookie (store cookie + attributes) - sending a cookie + validate the cookie domain against the request URL + validate the cookie path against the request URL + validate the cookie expiration + if valid then send only the cookie, no attribtues * Modifies how a request URL is stored during a XMLRPC request/response sequence. * Refactors a bit of the request/response logic to allow for making the decision whether to send a session cookie instead of full Kerberous auth easier. * The server now includes expiration information in the session cookie it sends to the client. The server always had the information available to prevent using an expired session cookie. Now that expiration timestamp is returned to the client as well and now the client will not send an expired session cookie back to the server. * Adds a new module and unit test for cookies (see below) Formerly we were always returning the session cookie no matter what the domain or path was in the URL. We were also sending the cookie attributes which are for the client only (used to determine if to return a cookie). The attributes are not meant to be sent to the server and the previous behavior was a protocol violation. We also were not checking the cookie expiration. Cookie library issues: We need a library to create, parse, manipulate and format cookies both in a client context and a server context. Core Python has two cookie libraries, Cookie.py and cookielib.py. Why did we add a new cookie module instead of using either of these two core Python libaries? Cookie.py is designed for server side generation but can be used to parse cookies on the client. It's the library we were using in the server. However when I tried to use it in the client I discovered it has some serious bugs. There are 7 defined cookie elements, it fails to correctly parse 3 of the 7 elements which makes it unusable because we depend on those elements. Since Cookie.py was designed for server side cookie processing it's not hard to understand how fails to correctly parse a cookie because that's a client side need. (Cookie.py also has an awkward baroque API and is missing some useful functionality we would have to build on top of it). cookielib.py is designed for client side. It's fully featured and obeys all the RFC's. It would be great to use however it's tightly coupled with another core library, urllib2.py. The http request and response objects must be urllib2 objects. But we don't use urllib2, rather we use httplib because xmlrpclib uses httplib. I don't see a reason why a cookie library should be so tightly coupled to a protocol library, but it is and that means we can't use it (I tried to just pick some isolated entrypoints for our use but I kept hitting interaction/dependency problems). I decided to solve the cookie library problems by writing a minimal cookie library that does what we need and no more than that. It is a new module in ipapython shared by both client and server and comes with a new unit test. The module has plenty of documentation, no need to repeat it here. Request URL issues: We also had problems in rpc.py whereby information from the request which is needed when we process the response is not available. Most important was the requesting URL. It turns out that the way the class and object relationships are structured it's impossible to get this information. Someone else must have run into the same issue because there was a routine called reconstruct_url() which attempted to recreate the request URL from other available information. Unfortunately reconstruct_url() was not callable from inside the response handler. So I decided to store the information in the thread context and when the request is received extract it from the thread context. It's perhaps not an ideal solution but we do similar things elsewhere so at least it's consistent. I removed the reconstruct_url() function because the exact information is now in the context and trying to apply heuristics to recreate the url is probably not robust. Ticket https://fedorahosted.org/freeipa/ticket/3022
2012-12-04 17:20:17 -06:00
cookie = Cookie('color', 'blue')
url = 'http://example.com/foo'
cookie.normalize(url)
assert cookie.domain == 'example.com'
assert cookie.path == '/'
Compliant client side session cookie behavior In summary this patch does: * Follow the defined rules for cookies when: - receiving a cookie (process the attributes) - storing a cookie (store cookie + attributes) - sending a cookie + validate the cookie domain against the request URL + validate the cookie path against the request URL + validate the cookie expiration + if valid then send only the cookie, no attribtues * Modifies how a request URL is stored during a XMLRPC request/response sequence. * Refactors a bit of the request/response logic to allow for making the decision whether to send a session cookie instead of full Kerberous auth easier. * The server now includes expiration information in the session cookie it sends to the client. The server always had the information available to prevent using an expired session cookie. Now that expiration timestamp is returned to the client as well and now the client will not send an expired session cookie back to the server. * Adds a new module and unit test for cookies (see below) Formerly we were always returning the session cookie no matter what the domain or path was in the URL. We were also sending the cookie attributes which are for the client only (used to determine if to return a cookie). The attributes are not meant to be sent to the server and the previous behavior was a protocol violation. We also were not checking the cookie expiration. Cookie library issues: We need a library to create, parse, manipulate and format cookies both in a client context and a server context. Core Python has two cookie libraries, Cookie.py and cookielib.py. Why did we add a new cookie module instead of using either of these two core Python libaries? Cookie.py is designed for server side generation but can be used to parse cookies on the client. It's the library we were using in the server. However when I tried to use it in the client I discovered it has some serious bugs. There are 7 defined cookie elements, it fails to correctly parse 3 of the 7 elements which makes it unusable because we depend on those elements. Since Cookie.py was designed for server side cookie processing it's not hard to understand how fails to correctly parse a cookie because that's a client side need. (Cookie.py also has an awkward baroque API and is missing some useful functionality we would have to build on top of it). cookielib.py is designed for client side. It's fully featured and obeys all the RFC's. It would be great to use however it's tightly coupled with another core library, urllib2.py. The http request and response objects must be urllib2 objects. But we don't use urllib2, rather we use httplib because xmlrpclib uses httplib. I don't see a reason why a cookie library should be so tightly coupled to a protocol library, but it is and that means we can't use it (I tried to just pick some isolated entrypoints for our use but I kept hitting interaction/dependency problems). I decided to solve the cookie library problems by writing a minimal cookie library that does what we need and no more than that. It is a new module in ipapython shared by both client and server and comes with a new unit test. The module has plenty of documentation, no need to repeat it here. Request URL issues: We also had problems in rpc.py whereby information from the request which is needed when we process the response is not available. Most important was the requesting URL. It turns out that the way the class and object relationships are structured it's impossible to get this information. Someone else must have run into the same issue because there was a routine called reconstruct_url() which attempted to recreate the request URL from other available information. Unfortunately reconstruct_url() was not callable from inside the response handler. So I decided to store the information in the thread context and when the request is received extract it from the thread context. It's perhaps not an ideal solution but we do similar things elsewhere so at least it's consistent. I removed the reconstruct_url() function because the exact information is now in the context and trying to apply heuristics to recreate the url is probably not robust. Ticket https://fedorahosted.org/freeipa/ticket/3022
2012-12-04 17:20:17 -06:00
cookie = Cookie('color', 'blue')
url = 'http://example.com/foo/bar'
cookie.normalize(url)
assert cookie.domain == 'example.com'
assert cookie.path == '/foo'
Compliant client side session cookie behavior In summary this patch does: * Follow the defined rules for cookies when: - receiving a cookie (process the attributes) - storing a cookie (store cookie + attributes) - sending a cookie + validate the cookie domain against the request URL + validate the cookie path against the request URL + validate the cookie expiration + if valid then send only the cookie, no attribtues * Modifies how a request URL is stored during a XMLRPC request/response sequence. * Refactors a bit of the request/response logic to allow for making the decision whether to send a session cookie instead of full Kerberous auth easier. * The server now includes expiration information in the session cookie it sends to the client. The server always had the information available to prevent using an expired session cookie. Now that expiration timestamp is returned to the client as well and now the client will not send an expired session cookie back to the server. * Adds a new module and unit test for cookies (see below) Formerly we were always returning the session cookie no matter what the domain or path was in the URL. We were also sending the cookie attributes which are for the client only (used to determine if to return a cookie). The attributes are not meant to be sent to the server and the previous behavior was a protocol violation. We also were not checking the cookie expiration. Cookie library issues: We need a library to create, parse, manipulate and format cookies both in a client context and a server context. Core Python has two cookie libraries, Cookie.py and cookielib.py. Why did we add a new cookie module instead of using either of these two core Python libaries? Cookie.py is designed for server side generation but can be used to parse cookies on the client. It's the library we were using in the server. However when I tried to use it in the client I discovered it has some serious bugs. There are 7 defined cookie elements, it fails to correctly parse 3 of the 7 elements which makes it unusable because we depend on those elements. Since Cookie.py was designed for server side cookie processing it's not hard to understand how fails to correctly parse a cookie because that's a client side need. (Cookie.py also has an awkward baroque API and is missing some useful functionality we would have to build on top of it). cookielib.py is designed for client side. It's fully featured and obeys all the RFC's. It would be great to use however it's tightly coupled with another core library, urllib2.py. The http request and response objects must be urllib2 objects. But we don't use urllib2, rather we use httplib because xmlrpclib uses httplib. I don't see a reason why a cookie library should be so tightly coupled to a protocol library, but it is and that means we can't use it (I tried to just pick some isolated entrypoints for our use but I kept hitting interaction/dependency problems). I decided to solve the cookie library problems by writing a minimal cookie library that does what we need and no more than that. It is a new module in ipapython shared by both client and server and comes with a new unit test. The module has plenty of documentation, no need to repeat it here. Request URL issues: We also had problems in rpc.py whereby information from the request which is needed when we process the response is not available. Most important was the requesting URL. It turns out that the way the class and object relationships are structured it's impossible to get this information. Someone else must have run into the same issue because there was a routine called reconstruct_url() which attempted to recreate the request URL from other available information. Unfortunately reconstruct_url() was not callable from inside the response handler. So I decided to store the information in the thread context and when the request is received extract it from the thread context. It's perhaps not an ideal solution but we do similar things elsewhere so at least it's consistent. I removed the reconstruct_url() function because the exact information is now in the context and trying to apply heuristics to recreate the url is probably not robust. Ticket https://fedorahosted.org/freeipa/ticket/3022
2012-12-04 17:20:17 -06:00
cookie = Cookie('color', 'blue')
url = 'http://example.com/foo/bar/'
cookie.normalize(url)
assert cookie.domain == 'example.com'
assert cookie.path == '/foo/bar'