Could you please submit a simple, runnable testcase to JIRA? Thanks.
OK, the first problem is fixed in CVS, along with another bug to do with pageflows. Please test it for me, thanks.
OK, I believe everything is fixed now. I've just made some really significant (hopefully non-breaking) improvements to the test harness.
Direct use of new Script() is now deprecated, tests should start using new FacesRequest() and new NonFacesRequest().
Especially, if you have a pageflow that does redirects, you need to make sure that you process the work in two requests, a FacesRequest followed by a NonFacesRequest.
The good news is that pageflows are now really properly testable.
I'll document these changes tomorrow.
I'll try to work on this today and give you some feedback.
Ok, so I finally got a chance to look at this.
SeamTesting pageflow seems to be working fine right now.
However, if navigation is not handled by a jPDL pageflow there's an issue with conversation propagation in the following scenario:
- Start a FacesRequest with a null conversationId;
- Run an action annotated with @Begin in invokeApplication;
- Set the outcome of this action in invokeApplication;
- Navigation is NOT handled by a jPDL pageflow;
Then the run method returns a null conversationId.
If you do exactly the same thing but you DO NOT set the outcome, then the run method returns the proper conversationId.
I couldn't really understand why this is happening, but I guess it is because the method getRenderedViewId() in SeamTest. Or maybe I am doing something wrong. Anyway, this is not a major problem.
I'll take a look at this.
What does your outcome look like. I'm thinking it must be of the form "/foo.jsp". Is that right?
Yep, that is correct.
OK, in that case I know why it happens :-)