I believe so; of course, you can't express the constraint using XML schema, so it is not obvious. -mbs
On 10/3/06, Alex Boisvert <[EMAIL PROTECTED]> wrote:
You mean an initiating <receive> and a subsequent <pick> without a correlation would be considered a static error? Maciej Szefler wrote: > I think this actually qualifies as a syntax error from a standard BPEL > 2.0pointof view. > -mbs > > On 10/3/06, Alex Boisvert <[EMAIL PROTECTED]> wrote: >> >> >> It's not clear to me if you expect the unit test to succeed or fail. >> >> In my view, the test is valid but may have two different outcomes >> depending on whether you use implicit correlation (e.g. some form of >> WS-Addressing). >> >> If the initial <receive> stores correlation information into the >> partnerLink and the <pick> uses the same partnerLink, then the test may >> succeed. >> >> If the initial <receive> does not initialize any correlation in the >> partnerLink, then I think the following <pick> should fail if it doesn't >> have an explicit correlation set. >> >> alex >> >> >> Lance Waterman wrote: >> > All, >> > >> > I have created a BPEL unit test - "TestStaticPick". It contains an >> > instantiating <receive> and then a subsequent <pick>. It appears to me >> > that >> > because the <pick> does not contain a correlation key the ODE router >> > can not >> > route the subsequent message. Before I open a JIRA issue I wanted to >> > validate that the test makes sense from a BPEL standpoint ( note - the >> > process compiles and executes okay ). >> > >> > Thanks, >> > >> > Lance >> > >> >> >
