RFR: 8150709: Mac OSX and German Keyboard Layout (Y/Z)

[Tom Schindl]

On Fri, 12 Mar 2021 21:12:41 GMT, Martin Fox <github.com+12087024+beldenfox at openjdk.org> wrote:

>> I see that the Windows pre-submit test build failed. It's clearly not related to anything in this PR, so it can be ignored.
>> 
>> I'll review this PR later (hopefully next week some time), but I have a couple general comments:
>> 
>> 1. Would it be possible to provide an automated test? Maybe not since it is sensitive to the keyboard layout.
>> 2. For the related bugs, we can either close them as duplicates of this bug or use the `/solves` command to list them here. Generally, we would do the former in the case it really is a single fix, as this seems to be. That's what I'll do once this bug is integrated unless there is a good reason not to. Normally we would use the earliest of the bugs, but in this case, I don't think it matters, so I have no problem with your using the one you chose.
>> 
>> @tomsontom Since you were the one who filed [JDK-8150709](https://bugs.openjdk.java.net/browse/JDK-8150709), and it's currently assigned to you, do you want to be the second reviewer on this?
>
> @kevinrushforth I have a manual app that can perform a simple test to verify that when a robot sends KeyCode.A through KeyCode.Z the system receives the characters 'a' to 'z'. On the Mac this sanity test was failing on German and French keyboards prior to these changes. The test is part of a key logger app I created.
> 
> I chose this bug because it has a straight-forward test attached and some recent comment activity.
> 
> @tomsontom Could you point me to the relevant code in Swing? I'm looking at the code but am getting lost in the layers.

https://github.com/openjdk/jdk/blob/8760688d213865eaf1bd675056eb809cdae67048/src/java.desktop/macosx/native/libawt_lwawt/awt/AWTEvent.m#L462

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

PR: https://git.openjdk.java.net/jfx/pull/425