[
https://issues.apache.org/jira/browse/TUSCANY-1926?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12572988#action_12572988
]
Raymond Feng commented on TUSCANY-1926:
---------------------------------------
Can you attach your test case so that I can validate if the latest code handles
this wrapped/unwrapped matching?
Thanks,
Raymond
> Problem w/ JDKInvocationHandler with dynamic reference with WSDLinterface
> -------------------------------------------------------------------------
>
> Key: TUSCANY-1926
> URL: https://issues.apache.org/jira/browse/TUSCANY-1926
> Project: Tuscany
> Issue Type: Bug
> Reporter: Scott Kurz
> Assignee: Raymond Feng
> Priority: Minor
> Fix For: Java-SCA-Next
>
>
> I found another case where the JDKInvocationHandler.match() doesn't handle
> WSDL interfaces.
> Earlier problems in this area were:
> https://issues.apache.org/jira/browse/TUSCANY-1342
> https://issues.apache.org/jira/browse/TUSCANY-1473
> I'm not sure, however, if this a boundary case problem since my operation a
> zero-arg operation or if
> there's a bigger problem.
> Say MyServiceImpl is a component impl looking like:
> public class MyServiceImpl implements MyService
> {
> @Context
> public ComponentContext context;
> public MyService2 getService2()
> { return context.getService(MyService2.class, "myRef");
> }
> }
> And say the SCDL configuration of MyServiceImpl's component looks like:
> <component name="MyServiceComponent">
> <implementation.java class="MyServiceImpl"/>
> <reference name="myRef">
> <interface.wsdl interface=".....portType corresponding to
> MyService2....."/>
> <binding.ws wsdlElement="..port..."/>
> </reference>
> </component>
> Now, I think in most cases we've dealt with, the "myRef" reference of
> MyServiceComponent is going to be declared
> statically with a @Reference in MyServiceImpl. I believe, though, that
> what I pasted is legal even though the
> reference is not a part of the static Java component implementation.
> I noticed this doesn't work ... though I ran with a very special case of a
> zero-parm operation. I noticed the problem
> where the input sizes don't match up (0 in Java vs. 1 in WSDL since we have
> the single, empty input wrapper elem), like
> the one Raymond had fixed in the past in DataBindingRuntimeWireProcessor.
> But I wonder if this was only the first problem, since I noticed in my
> similar working case, my wire source operation was Java
> intf-based even with the <interface.wsdl> config in SCDL.
> I'm not sure though.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
