RFR: 8340365: Position the first window of a window list
Alexander Zvegintsev
azvegint at openjdk.org
Thu Sep 19 06:48:36 UTC 2024
On Wed, 18 Sep 2024 12:05:22 GMT, Alexey Ivanov <aivanov at openjdk.org> wrote:
> Support of multiple test UI windows in `PassFailJFrame` is still evolving. After [JDK-8340210](https://bugs.openjdk.org/browse/JDK-8340210), the `Builder` has a method `positionTestUI` to supply an implementation of the `PositionWindows` interface which handles the positioning of all test UI windows created.
>
> If `PositionWindows` is not provided, all the test UI windows are left with the default coordinates: (0, 0).
>
> If `PassFailJFrame` called `positionTestWindow` for the first window, it would allow the test developer to position other windows based on the position of the first one.
>
> This issue was raised in #21029 in [this thread of comments](https://github.com/openjdk/jdk/pull/21029#discussion_r1763744407).
>
> To make it easier to position multiple test UI windows without implementing the `PositionWindows` interface, the `PassFailJFrame` should position the first window of the list.
>
> The `LotsOfMenuItemsTest.java` test in #21029 had to hard-code the coordinates or use `PassFailJFrame.positionTestWindow` explicitly in its `ComponentListener.componentShown`.
>
> With the proposed change in this pull request, the implementation would use `ComponentListener.componentMoved`:
>
>
> @Override
> public void componentMoved(ComponentEvent e) {
> testFrame.setLocation(firstFrame.getX(),
> firstFrame.getY() + firstFrame.getHeight() + 8);
> }
>
>
> This has the advantage in that the test UI windows don't flicker for a short time in the upper left corner of the screen; and the developer has to handle only positioning the second window (frame).
>
> However, this has a funny effect: if you move the first frame with mouse or in any other way, the second frame follows. To avoid such kind of behaviour, call `removeComponentListener` in the handler.
>
> @honkar-jdk, @azvegint, could you take a look?
Marked as reviewed by azvegint (Reviewer).
-------------
PR Review: https://git.openjdk.org/jdk/pull/21057#pullrequestreview-2314554177
More information about the client-libs-dev
mailing list