RFR: 6429812: NPE after calling JTable.updateUI() when using a header renderer + XP L&F [v3]

[Tejesh R]

On Tue, 31 May 2022 09:37:57 GMT, Tejesh R <duke at openjdk.java.net> wrote:

>>> Yeah, wanted to change to automatic, but wasn't sure about whether its ok for NPE check or not.
>> 
>> What can be wrong with it? It's just a regular test. If you can automate the test, you should.
>> 
>>> Actually if the current test executes then its a pass right......?
>> 
>> I don't understand your question, sorry. What surprises me is that the test fails automatically without the fix. But I can't see how the NPE is re-thrown on the main thread.
>> 
>> I thought the NPE would be just printed in the jtreg log file and the tester wouldn't even see it.
>> 
>> So your test is already automated to some extent. What we want is to make it pass automatically if NPE isn't thrown.
>> 
>>> Painting it to bufferedImage and then catching NPE is required........?
>> 
>> No, it's not, however, I think it's a more effective way than showing UI. The exception occurs when `paint` is called. Painting to `BufferedImage` gives you control when and how `paint` is called: you do it _explicitly_. So you wrap the call to `paint` in a try-catch block, if the exception is thrown you re-throw it on main thread to fail the test; if the exception isn't thrown, the test completed successfully. Perhaps, try-catch block isn't necessary, just let the exception propagate.
>
>> _Actually if the current test executes then its a pass right......?_ This means, if there is no NPE raised then its a pass case right.......? Yeah to some extent it is automatic, I actually didn't get how to handle the caught NPE, so just left it so that the Test case will be failed by NPE.

Ok, then will try with the _bufferedImage paint_ logic, If I'm able to get the exception without it been handled by internal code then it'll be really helpful to make it automatic.

-------------

PR: https://git.openjdk.java.net/jdk/pull/8830