Ok, this reading. How is the cookie supposed to get set on the server
side?
Our test case shows that a client handler can read the cookie when a
server handler actually sets one.
We did an optimization where the server component does not create a new
HttpSession for every request, because this is an expensive operation.
Hence, when the server component does not access the HttpSession and no
other component expicitly sets a cookie there will be no cookie for the
client to read.
cheers
-thomas
On Wed, 2007-04-04 at 09:23 -0400, Chris Laprun wrote:
On Apr 4, 2007, at 3:28 AM, Thomas Diesler wrote:
> please study the testcase associated with the the jira issue. It shows
> how to work with cookies.
I did. Here's a code sample that I use in my handler:
public boolean handleResponse(MessageContext msgContext)
{
SOAPMessageContext smc = (SOAPMessageContext)msgContext;
SOAPMessage message = smc.getMessage();
MimeHeaders mimeHeaders = message.getMimeHeaders();
String[] cookieValues = mimeHeaders.getHeader("Set-Cookie");
if (cookieValues != null)
{
// deal with cookies...
}
return super.handleResponse(msgContext);
}
The problem is that with 1.0.4.GA and up, the Set-Cookie header is
not present anymore. My code works fine with 1.0.3.sp1 and below and
when I step debug it, I can see the Set-Cookie header present in the
MimeHeaders. With 1.0.4.GA and above (that includes 1.2.1 that I
compiled yesterday), the Set-Cookie header is not present anymore
when I step debug the code. I looked at the test case configuration
to see if I was missing anything but it doesn't look like it
either... so I'm not quite sure what's going on.
Best,
Chris
==
JBoss Portal Developer / WSRP Lead
JBoss, a division of Red Hat
--
xxxxxxxxxxxxxxxxxxxxxxxxxxxx
Thomas Diesler
Web Service Lead
JBoss, a division of Red Hat
xxxxxxxxxxxxxxxxxxxxxxxxxxxx