[External] : Re: Q: Labeled.textTruncated Property (RFE)

Tue Mar 12 21:36:03 UTC 2024

Thank you for the feedback!

The property is being added to Labeled, which happens to be a base class for the cells' skins.  The truncated property therefore must work in all the subclasses, even those where computePrefWidth is hacked via `Properties.DEFER_TO_PARENT_PREF_WIDTH`.

> Second, I would rather want to see if this can be changed without adding some flags which once again will not help application developers that may want to extend something like this.

This is a much more difficult task:  First, it needs a careful discussion on which APIs to add, if any (there is a possibility that some people might say that no new APIs are needed as one can always create an instance of Text or Label and use it to get the sizing).  There might be a need to undo the hacks made by the founding father such as `Properties.DEFER_TO_PARENT_PREF_WIDTH`.  I would say all this is out of scope for this PR, as it achieves its goal without introducing the new APIs.

> IMO, the separation of concern is wrong. A Cell should always give back the prefWidth, other callers should think if they want to use the pref width or the table column width.

Why cells are controls is slightly beyond me, but who am I to ask?  I would say there should be one control in this picture, and that is TableView (or any other cell-based control) whose job is to size and lay out its various parts.  The current design is different, and it is out of scope for this PR to change the design.

> In order to move this forward, I would like to see only the necessary changes

And that 's exactly what you see here - only the necessary changes (that unfortunately include undoing the earlier hacks in cells).

As for tests - I certainly agree, but that would be a manual test.  The behavior can be verified with the examples in JDK-8205211, I just felt the value of a manual test is rather low.  What I would rather see is an automated test that actually verifies the new property behavior down to a pixel.  Any ideas how to make one?

-andy

From: Marius Hanl <mariushanl at web.de>
Date: Monday, March 11, 2024 at 15:49
To: Daniel Peintner <daniel.peintner at gmail.com>
Cc: Andy Goryachev <andy.goryachev at oracle.com>, openjfx-dev at openjdk.org <openjfx-dev at openjdk.org>
Subject: Aw: Re: [External] : Re: Q: Labeled.textTruncated Property (RFE)
I think this can still be implemented for cells without any special handling.
It is rather unexpected that a cell does not give the correct pref width, so we need to think about how to improve the current situation.
In my opinion, it should not be the concern of the Cell to give the actual width of the column it belongs to, but the pref width, as we would expect.
We need to check where this is called.

-- Marius

Gesendet: Donnerstag, 07. März 2024 um 16:07 Uhr
Von: "Daniel Peintner" <daniel.peintner at gmail.com>
An: "Andy Goryachev" <andy.goryachev at oracle.com>
Cc: "openjfx-dev at openjdk.org" <openjfx-dev at openjdk.org>
Betreff: Re: [External] : Re: Q: Labeled.textTruncated Property (RFE)
Andy, all,

Some comments below:

  1.  Even if it is possible to do the same with public means I think it is useful. Anyhow we need to promote the solutions better (e.g., on StackOverflow)

  2.  I tried to achieve what you did in JavaFX core for my use-case, TableView cells.

     *   In my real application I tried and failed since the boolean condition in https://github.com/openjdk/jfx/blob/8a5cb17ed6525b0a218595865ead3e399e3b29dd/modules/javafx.controls/src/main/java/javafx/scene/control/Labeled.java#L838-L841<https://urldefense.com/v3/__https:/github.com/openjdk/jfx/blob/8a5cb17ed6525b0a218595865ead3e399e3b29dd/modules/javafx.controls/src/main/java/javafx/scene/control/Labeled.java*L838-L841__;Iw!!ACWV5N9M2RV99hQ!JoAaduFdFfQdPcSVdkE7oZbkwg6oeCdmWsh0RTvQALlv9ti--OQHFGtHU3gRE_WPiNLBxlX6eVCGufL5rk0kXFjO$> always reports false
     *   Hence I tried to create a short self-contained TableView example to explore it further, see Gist https://gist.github.com/danielpeintner/9b561eb57261e8c5bc608f1346df74cf<https://urldefense.com/v3/__https:/gist.github.com/danielpeintner/9b561eb57261e8c5bc608f1346df74cf__;!!ACWV5N9M2RV99hQ!JoAaduFdFfQdPcSVdkE7oZbkwg6oeCdmWsh0RTvQALlv9ti--OQHFGtHU3gRE_WPiNLBxlX6eVCGufL5rgWNKS4B$>
It still fails :-(
     *   The Email field in the table ideally should report "true" once the text entered exceeds the visible view. Somehow it fails... getWidth() and  prefWidth(-1) always return the same double value. Please try the Gist above to confirm.
e.g., edit cell and extend the email text to longer than column width --> results in System.out "Table-Email-Cell !isWrapText: 182.0 < 182.0 --> false"
     *   Hence I also added a *normal* MyLabel example which does work as intended
Maybe I miss something obvious but TableCells seem to be special Labeled instances and behave differently.

Thanks,

-- Daniel

[1] https://stackoverflow.com/a/52432862<https://urldefense.com/v3/__https:/stackoverflow.com/a/52432862__;!!ACWV5N9M2RV99hQ!JoAaduFdFfQdPcSVdkE7oZbkwg6oeCdmWsh0RTvQALlv9ti--OQHFGtHU3gRE_WPiNLBxlX6eVCGufL5riJDXXyw$>

On Wed, Mar 6, 2024 at 6:10 PM Andy Goryachev <andy.goryachev at oracle.com<mailto:andy.goryachev at oracle.com>> wrote:
Dear Daniel:

Thank you for confirming the need for this enhancement.  Your case is covered by this ticket (I included a wrong ticket in my previous message by mistake, sorry):

JDK-8327483<https://bugs.openjdk.org/browse/JDK-8327483> TreeView: Allow for tooltip when cell text is truncated

My only concern is that this functionality theoretically can be achieved by the application code just as easy, since the solution does not use any internal APIs, as can be seen here: https://github.com/openjdk/jfx/pull/1389<https://urldefense.com/v3/__https:/github.com/openjdk/jfx/pull/1389__;!!ACWV5N9M2RV99hQ!LUwCrjMEKxlGC0k5v9-zO0Axw6TqddmMJM7_6onFdHOLL1eo5MDNu1lVj7grGquhgJdnqXX_eCyA7MGEMNyIivVMcBB9XfY$>

-andy

From: Daniel Peintner <daniel.peintner at gmail.com<mailto:daniel.peintner at gmail.com>>
Date: Tuesday, March 5, 2024 at 23:53
To: Andy Goryachev <andy.goryachev at oracle.com<mailto:andy.goryachev at oracle.com>>
Cc: openjfx-dev at openjdk.org<mailto:openjfx-dev at openjdk.org> <openjfx-dev at openjdk.org<mailto:openjfx-dev at openjdk.org>>
Subject: [External] : Re: Q: Labeled.textTruncated Property (RFE)
Hi Andy, all,

Thank you for the effort which I very much appreciate.
We have a similar use-case as the one mentioned in JDK-8205211.

We would like to show a Tooltip with the corresponding text once the text is shown with ellipsis for a TableCell. The *hacky* solution we are using is sketched in [1] and requires us to calculate the width again. Moreover, once different fonts come into the game it becomes even more brittle.

Having an easy (easier?) way to detect and observe whether a label is truncated is of importance for us and definitely improves the usability of JavaFX.

Thanks,

-- Daniel

[1] https://stackoverflow.com/a/52432862<https://urldefense.com/v3/__https:/stackoverflow.com/a/52432862__;!!ACWV5N9M2RV99hQ!LUwCrjMEKxlGC0k5v9-zO0Axw6TqddmMJM7_6onFdHOLL1eo5MDNu1lVj7grGquhgJdnqXX_eCyA7MGEMNyIivVMSbAOrZg$>

On Tue, Mar 5, 2024 at 10:15 PM Andy Goryachev <andy.goryachev at oracle.com<mailto:andy.goryachev at oracle.com>> wrote:
Dear colleagues:

Do you think a new property, ReadOnlyBooleanProperty textTruncated, should be added to the Labeled control?

On one hand, we do have a couple of tickets in JBS requesting that:

  *   JDK-8091891<https://bugs.openjdk.org/browse/JDK-8091891> TreeView: There is no tooltip available on truncated node
  *   JDK-8205211<https://bugs.openjdk.org/browse/JDK-8205211> Ability to show Tooltip only when text is shown with ellipsis (...)

On the other hand, this functionality can in theory be implemented at the application level, although the number of changes required might be uncomfortably large in the case of Tree/TreeTable/TableView.

JBS:

  *   JDK-8092102<https://bugs.openjdk.org/browse/JDK-8092102> Labeled: truncated property

Pull Request: https://github.com/openjdk/jfx/pull/1389<https://urldefense.com/v3/__https:/github.com/openjdk/jfx/pull/1389__;!!ACWV5N9M2RV99hQ!LUwCrjMEKxlGC0k5v9-zO0Axw6TqddmMJM7_6onFdHOLL1eo5MDNu1lVj7grGquhgJdnqXX_eCyA7MGEMNyIivVMcBB9XfY$>

What do you think?

-andy
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://mail.openjdk.org/pipermail/openjfx-dev/attachments/20240312/3f72fc25/attachment-0001.htm>