Webware for Python

Nick Murtagh schrieb:
> Hi,
> 
> I have two webware apps running on the same domain, one on /default/ and
> the other on /admin/. To avoid the session cookies from colliding, I
> made a small change in HTTPResponse.py:recordSession. The line
> 
> cookie.setPath('/')
> 
> is replaced by
> 
> cookie.setPath(self._transaction.application().setting('SessionCookiePath', 
> '/'))

Nick, can you check whether the following solves the problem, too?

cookie.setPath(self._transaction.request().adapterName())

That way you will not need to configure anything (we can still make it 
configurable).

-- Chris

Hi Nick,

you recently posted the following on the Webware mailing list:

> I have two webware apps running on the same domain, one on /default/ and
> the other on /admin/. To avoid the session cookies from colliding, I
> made a small change in HTTPResponse.py:recordSession. The line
> cookie.setPath('/')
> is replaced by
> cookie.setPath(self._transaction.application().setting('SessionCookiePath', 
> '/'))
> Now I can put
> SessionCookiePath = '/default/'
> in Application.config for the first app and
> SessionCookiePath = '/admin/'
> in Application.conf for the second and hey presto the sessions are
> independent :)

Since I want to make a new Webware release soon, I'd like to have that 
fixed there. My idea is to replace the above line by the following:

cookie.setPath(self._transaction.request().adapterName())

This has the advantage that you don't need to configure anything. Can 
you give me feedback whether this would solve the problem for you?

Greetings
-- Chris

> Since I want to make a new Webware release soon, I'd like to have that
> fixed there. My idea is to replace the above line by the following:
>
> cookie.setPath(self._transaction.request().adapterName())
>
> This has the advantage that you don't need to configure anything. Can
> you give me feedback whether this would solve the problem for you?

Hi Chris,

It turned out that my original idea didn't quite work in practice,
and then I found that webware already includes the feature I needed.

In

http://svn.w4py.org/Webware/trunk/WebKit/Application.py

Application.__init__ uses a setting called SessionName
to control the name of the session for each application.
Setting SessionName differently for my two applications
solved the problem.

The only improvement I can think of would be to add a commented
out SessionName directive to the default Application config. I
only found out it existed by reading the code :)

Nick

Thanks for the feedback, Nick.

> It turned out that my original idea didn't quite work in practice,
> and then I found that webware already includes the feature I needed.
> In http://svn.w4py.org/Webware/trunk/WebKit/Application.py
> Application.__init__ uses a setting called SessionName
> to control the name of the session for each application.
> Setting SessionName differently for my two applications
> solved the problem.

Yes, this changes the name of the cookie with the session id.

> The only improvement I can think of would be to add a commented
> out SessionName directive to the default Application config. I
> only found out it existed by reading the code :)

Actually, it's described in the docs as well, but I agree reading the 
code is sometimes easier :-)
http://www.webwareforpython.org/WebKit/Docs/Configuration.html#path-handling

I'll put it in the default config as well.

Anyway, I'm curious why your original approach did not work in practice 
as well. I think setting the cookie path to the adapterName() instead of 
'/' will avoid the necessity of setting different SessionNames and at 
the same time also improve security (see 
http://www.net-security.org/dl/articles/cookie_path.pdf).

Does anything speak against such a patch?

-- Chris

Bugs item #1646644, was opened at 2007-01-28 23:46
Message generated for change (Comment added) made by cito
You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=104866&aid=1646644&group_id=4866

Please note that this message will contain a full copy of the comment thread,
including the initial issue submission, for this request,
not just the latest update.
Category: WebKit
Group: None
>Status: Closed
>Resolution: Accepted
Priority: 5
Private: No
Submitted By: Ben Parker (benparker)
>Assigned to: Christoph Zwerschke (cito)
Summary: cleanStaleSessions loads all sessions into memory

Initial Comment:
SessionStore's definition of cleanStaleSessions iterates over self.items(), which loads all the sessions into memory. This effectively eliminates the purpose of SessionFileStore. 

This was repeatedly crashing a Webware process running SessionDynamicStore during periods of peak traffic, when abnormally high numbers of sessions were created.

Changing the loop in SessionStore to iterate over the keys, rather than the items, fixes the problem. A patch is attached.

----------------------------------------------------------------------

>Comment By: Christoph Zwerschke (cito)
Date: 2007-01-31 00:47

Message:
Logged In: YES 
user_id=193957
Originator: NO

You're right, the item() call resulted in all sessions being loaded from
the filesystem. Very good fix, thank's a lot. Patched in Rev 6159.

----------------------------------------------------------------------

You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=104866&aid=1646644&group_id=4866

Patches item #1626532, was opened at 2007-01-03 00:40
Message generated for change (Comment added) made by cito
You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=304866&aid=1626532&group_id=4866

Please note that this message will contain a full copy of the comment thread,
including the initial issue submission, for this request,
not just the latest update.
Category: WebUtils
Group: None
>Status: Closed
>Resolution: Accepted
Priority: 5
Private: No
Submitted By: Veros Kaplan (veros)
Assigned to: Nobody/Anonymous (nobody)
Summary: Funcs.normURI fix

Initial Comment:
fixing typo in func normURI.  

Am I the only one in the world who is using this function? :-)

Download patch on http://veros.koren.cz/patches/webware-webutils-funcs.patch

----------------------------------------------------------------------

>Comment By: Christoph Zwerschke (cito)
Date: 2007-01-03 01:18

Message:
Logged In: YES 
user_id=193957
Originator: NO

Again my fault (tried to get rid off the string module) last year. Fixed
in the trunk now.

----------------------------------------------------------------------

You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=304866&aid=1626532&group_id=4866

Patches item #1626519, was opened at 2007-01-03 00:03
Message generated for change (Comment added) made by cito
You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=304866&aid=1626519&group_id=4866

Please note that this message will contain a full copy of the comment thread,
including the initial issue submission, for this request,
not just the latest update.
Category: WebKit
Group: None
>Status: Closed
>Resolution: Accepted
Priority: 5
Private: No
Submitted By: Veros Kaplan (veros)
Assigned to: Nobody/Anonymous (nobody)
Summary: Servet .serverSidePath fix

Initial Comment:
method Servlet.serverSidePath doesn't work as expected when using parameter with slash (eg.  "servlets/TitlePage" )

The method creates cache attribute _serverSidePath which is computed from "path" parameter. And this is the problem, because  _serverSidePath  shouldn't depend on the parameter.

The problem appeared sometimes between 0.9.1 and 0.9.2

Patch is available on http://veros.koren.cz/patches/webware-servlet.patch

Migrating from 0.8.1 to 0.9.2

----------------------------------------------------------------------

>Comment By: Christoph Zwerschke (cito)
Date: 2007-01-03 01:04

Message:
Logged In: YES 
user_id=193957
Originator: NO

Thanks. You're right. I don't know what I was thinking when I commited
revision 5444. I have fixed it in the trunk now.

----------------------------------------------------------------------

You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=304866&aid=1626519&group_id=4866

Hi Jean-Francois,

I think your work could benefit others -- I would suggest that you create a=
 ticket (in sourceforge), and post the full code there.

peace,
Jason

----- Jean-Fran=C3=A7ois Pi=C3=A9ronne <jf.pieronne@...> wrote:
> Hi,
>=20
> first, happy new year to all :-)
>=20
> I have successfully integrate pyjamas (http://pyjamas.pyworks.org/)
> with Webware.
> This mean that a page generate by pyjamas can query using JSON a
> Webware
> servlet.
> Like AjaxPage I have build a jsonPage which can be use as a root
> class.
> So a simple servlet example look like
> JSONRPC.py:
> ------------------------
> from jsonPage import jsonPage as BaseClass
>=20
> class JSONRPC(BaseClass):
>     def __init__(self):
>         BaseClass.__init__(self)
>=20
>     def echo(self, msg):
>         return msg
>=20
>     def reverse(self, msg):
>         return msg[::-1]
>=20
>     def uppercase(self, msg):
>         return msg.upper()
>=20
>     def lowercase(self, msg):
>         return msg.lower()
>=20
>     def json_methods(self):
>         return BaseClass.json_methods(self) + ['echo', 'reverse',
>=20
>                                                'uppercase',
> 'lowercase']
> ------------------------
>=20
>=20
> I have also fix a bug in HTTPRequest.py:
>=20
> Index: HTTPRequest.py
> =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D
> --- HTTPRequest.py=09(revision 6033)
> +++ HTTPRequest.py=09(working copy)
> @@ -597,7 +597,7 @@
>=20
>  =09=09"""
>  =09=09fs =3D self.fieldStorage()
> -=09=09if fs and fs.file:
> +=09=09if fs is not None and fs.file:
>  =09=09=09if rewind:
>  =09=09=09=09fs.file.seek(0)
>  =09=09=09return fs.file
>=20
>=20
>=20
> If anyone want more information I can post the complete source code
> somewhere.
>=20
>=20
> JF
>=20
> -------------------------------------------------------------------------
> Take Surveys. Earn Cash. Influence the Future of IT
> Join SourceForge.net's Techsay panel and you'll get the chance to
> share your
> opinions on IT & business topics through brief surveys - and earn
> cash
> http://www.techsay.com/default.php?page=3Djoin.php&p=3Dsourceforge&CID=3D=
DEVDEV
> _______________________________________________
> Webware-devel mailing list
> Webware-devel@...
> https://lists.sourceforge.net/lists/listinfo/webware-devel

Jean-François Piéronne schrieb:
> I have successfully integrate pyjamas (http://pyjamas.pyworks.org/)
> with Webware.
> This mean that a page generate by pyjamas can query using JSON a Webware
> servlet.

Nice. We could add that to the WebKit package. Probably we should call 
it JSONServlet instead of jsonPage since "Pages" in Webkit denote HTML 
pages (based on Page.py). Can you also add a simple demo for the 
Examples directory and/or add some comments about usage in the module 
docstring of JSONServlet?

> I have also fix a bug in HTTPRequest.py:

Thanks. Empty FieldStorage instances are indeed considered False since 
they have a __len__ method. I have already fixed that in the trunk.

-- Christoph