Re: [servlet-dev] Information required on Servlet TCK exclusion

[Stuart Douglas <sdouglas@xxxxxxxxxx>]

I think that the spec is inconsistent here, basically the test does:

startAsync();

complete();

Then it expects that isAsyncStarted() will return true.

However the Javadoc for isAsyncStarted() says:

     * This method returns <tt>false</tt> if this request was put into asynchronous mode, but has since been dispatched
     * using one of the {@link AsyncContext#dispatch} methods or released from asynchronous mode via a call to
     * {@link AsyncContext#complete}.

So according to the text isAsyncStarted should be returning false as complete has been called.

However the text for section 2.3.3.3 says that if complete is called isAsyncStarted should return true until the request returns to the container.

I think we should clarify the isAsyncStarted() javadoc to match the spec and re-enable the test.

Stuart

On Sat, 20 Nov 2021 at 03:13, Mark Thomas <markt@xxxxxxxxxx> wrote:

Hi TCK team,

I'm working on updating the Servlet TCK. Part of that work includes
reviewing the existing excludes and fixes the underlying issues so the
tests can be re-enabled. I'm stuck on the following:

com/sun/ts/tests/servlet/api/jakarta_servlet/servletrequest30/URLClient.java#asyncStartedTest3

The only clues I have found so far are:

- It may have been first excluded for Servlet 3.1
- BUG id 19793504

That ID doesn't seem to relate to any system to which I have access.

Can anyone here that has been involved in the TCKs longer than I have
shed any light?

Worst case, I'll simply re-enable the test and see what happens.

Mark
_______________________________________________
servlet-dev mailing list
servlet-dev@xxxxxxxxxxx
To unsubscribe from this list, visit https://www.eclipse.org/mailman/listinfo/servlet-dev