See screenshot

According to the specification all points within a relative Bézier command (e.g. 'c','s','q','t'), are all relative to the current point.

When a relative c or s command is used, each of the relative coordinate pairs is computed as for those in an m command. For example, the final control point of the curve of both commands is (cpx + x cos cb + y sin cb, cpy + x sin cb + y cos cb). [Where cpx and cpy represent the coordinates of the current point,and cb represents the current bearing.]

Currently the code makes each point of the Bézier command relative to the last two values added including those from the current command.

When I try to set the foreground color of a CDateTime widget, I get a NullPointerException. The Exception is thrown at following point in the CDateTime class:

public void setForeground(Color color) { super.setForeground(color); picker.updateColorsAndFont(); }
The CDateTime widget is instantiated as follows:

CDateTime dateTime = new CDateTime(composite, fieldIdentifier.styles()); dateTime.setLayoutData(fieldIdentifier.gridData()); dateTime.setNullText(""); //$NON-NLS-1$
and fieldIdentifier.styles() is CDT.BORDER | CDT.DROP_DOWN | CDT.DATE_MEDIUM | CDT.TIME_SHORT

The code tries to update the color of the picker widget. This widget is null. Maybe I am doing something wrong? What if so?

The Map Cache is backed by a LinkedHashMap which is neither thread safe, or accessed in a synchronised way. I haven't seen this cause a specific problem, but it would seem likely that there is some effect of this.

I have an issue with the PrintPreview. On one specific computer, the preview is always blank. I've tried to run the Snippet7 and Snippet3 and the preview is always blank. Trying the same snippet on another computer is working perfectly fine.

Was this problem ever reported and how to debug / fix this problem.

I'm using v2.1.0.202011020719, on windows.

Take note, Snippet2 that uses PrintViewer is working fine.

If you set the map cache size to a low value (e.g. 0, 1, 5 etc), the map tiles do not load.

If you add a Thread.sleep to the paintTile method, they then work again.

This suggests that it's not the cache size that's the problem, but some kind of race condition where if something happens too quickly, it breaks. My suspicion is that it's drawing the picture before the image has actually loaded.

If one disables a TableCombo the arrow is grayed out, but the border + text background remains the same and the text stays black.

It would be better for the user if the control is more visually "disabled", eg using SWT.COLOR_WIDGET_DISABLED_FOREGROUND, a different border color and so on.

Is this really needed? Also, why do you re-export SWT?

Opal PromptSupport does not know when the user typed the same text as the initial prompt text. PromptSupport may only show the prompt if the underlying text widget is actually empty.

The shell is accessed in the first line, after that it is checked for null what will never be true as NPE will occur in the first line already.

        /**
	 * @param shell
	 * @return the bounds of the monitor on which the shell is running
	 */
	public static Rectangle getBoundsOfMonitorOnWhichShellIsDisplayed(final Shell shell) {
		Monitor monitor = shell.getDisplay().getPrimaryMonitor();
		if(shell != null && shell.getMonitor() != null) {
			monitor = shell.getMonitor();
		} 
		
		return monitor.getBounds();
	}

proposed change

	public static Rectangle getBoundsOfMonitorOnWhichShellIsDisplayed(final Shell shell) {
		Objects.requireNonNull(shell, "Shell can't be null");
		Monitor monitor = shell.getMonitor();
		if (monitor == null) {
			monitor = shell.getDisplay().getPrimaryMonitor();
		}
		return monitor.getBounds();
	}

Currently one needs to select one column that is responsible for display the text.
It would be good to have an alternative so the user has more control, I can think about the following things:

1. Support setting a Function<Object, String>, ToIntFunction to perform conversion Object>String and int>String
2. Allow to set the column to -1 that results in simply concatenate all column text with a pipe ( C1 | C2 | C3) as a result.

Trying to bind the PasswordRevealer to a data model will result in a 'not supported' error.

Example code

import org.eclipse.core.databinding.Binding;
import org.eclipse.core.databinding.beans.typed.BeanProperties;
import org.eclipse.core.databinding.observable.value.IObservableValue;
import org.eclipse.jface.databinding.swt.typed.WidgetProperties;
import org.eclipse.nebula.widgets.passwordrevealer.PasswordRevealer;

final PasswordRevealer revealer = new PasswordRevealer(shell, SWT.NONE);

final IObservableValue<Object> passwordModel = BeanProperties.value(CredentialModel.class,CredentialModel.PROP_PASSWORD).observe(model.getUser());
final IObservableValue<String> targetPassword = WidgetProperties.text(SWT.Modify).observe(revealer);

new DataBindingContext().bindValue(targetPassword, passwordModel);

Will result in error

java.lang.NoSuchMethodException: org.eclipse.nebula.widgets.passwordrevealer.PasswordRevealer.addPropertyChangeListener(java.beans.PropertyChangeListener)
         at java.lang.Class.getMethod(Class.java:1786)
         ... 

 
Widget [org.eclipse.nebula.widgets.passwordrevealer.PasswordRevealer] is not supported.
         at org.eclipse.e4.core.commands.internal.HandlerServiceHandler.execute(HandlerServiceHandler.java:170)
         at org.eclipse.core.commands.Command.executeWithChecks(Command.java:488)
         at org.eclipse.core.commands.ParameterizedCommand.executeWithChecks(ParameterizedCommand.java:487)
         at org.eclipse.e4.core.commands.internal.HandlerServiceImpl.executeHandler(HandlerServiceImpl.java:213)
         at org.eclipse.e4.ui.workbench.renderers.swt.HandledContributionItem.executeItem(HandledContributionItem.java:438)
         at org.eclipse.e4.ui.workbench.renderers.swt.AbstractContributionItem.handleWidgetSelection(AbstractContributionItem.java:449)
         at org.eclipse.e4.ui.workbench.renderers.swt.AbstractContributionItem.lambda$2(AbstractContributionItem.java:475)
         at org.eclipse.swt.widgets.EventTable.sendEvent(EventTable.java:89)
         at org.eclipse.swt.widgets.Display.sendEvent(Display.java:4213)
         at org.eclipse.swt.widgets.Widget.sendEvent(Widget.java:1037)
         at org.eclipse.swt.widgets.Display.runDeferredEvents(Display.java:4030)
         at org.eclipse.swt.widgets.Display.readAndDispatch(Display.java:3630)
         at org.eclipse.e4.ui.internal.workbench.swt.PartRenderingEngine$5.run(PartRenderingEngine.java:1158)
         at org.eclipse.core.databinding.observable.Realm.runWithDefault(Realm.java:338)
         at org.eclipse.e4.ui.internal.workbench.swt.PartRenderingEngine.run(PartRenderingEngine.java:1047)
         at org.eclipse.e4.ui.internal.workbench.E4Workbench.createAndRunUI(E4Workbench.java:155)
         at org.eclipse.e4.ui.internal.workbench.swt.E4Application.start(E4Application.java:166)
         at org.eclipse.equinox.internal.app.EclipseAppHandle.run(EclipseAppHandle.java:203)
         at org.eclipse.core.runtime.internal.adaptor.EclipseAppLauncher.runApplication(EclipseAppLauncher.java:137)
         at org.eclipse.core.runtime.internal.adaptor.EclipseAppLauncher.start(EclipseAppLauncher.java:107)
         at org.eclipse.core.runtime.adaptor.EclipseStarter.run(EclipseStarter.java:401)
         at org.eclipse.core.runtime.adaptor.EclipseStarter.run(EclipseStarter.java:255)
         at sun.reflect.NativeMethodAccessorImpl.invoke0(Native Method)
         at sun.reflect.NativeMethodAccessorImpl.invoke(NativeMethodAccessorImpl.java:62)
         at sun.reflect.DelegatingMethodAccessorImpl.invoke(DelegatingMethodAccessorImpl.java:43)
         at java.lang.reflect.Method.invoke(Method.java:498)
         at org.eclipse.equinox.launcher.Main.invokeFramework(Main.java:657)
         at org.eclipse.equinox.launcher.Main.basicRun(Main.java:594)
         at org.eclipse.equinox.launcher.Main.run(Main.java:1447)
         at org.eclipse.equinox.launcher.Main.main(Main.java:1420)

Caused by: org.eclipse.e4.core.di.InjectionException: java.lang.IllegalArgumentException: Widget [org.eclipse.nebula.widgets.passwordrevealer.PasswordRevealer] is not supported.
         at org.eclipse.e4.core.internal.di.MethodRequestor.execute(MethodRequestor.java:68)
         at org.eclipse.e4.core.internal.di.InjectorImpl.invokeUsingClass(InjectorImpl.java:319)
         at org.eclipse.e4.core.internal.di.InjectorImpl.invoke(InjectorImpl.java:253)
         at org.eclipse.e4.core.contexts.ContextInjectionFactory.invoke(ContextInjectionFactory.java:173)
         at org.eclipse.e4.core.commands.internal.HandlerServiceHandler.execute(HandlerServiceHandler.java:156)
         ... 29 more

Caused by: java.lang.IllegalArgumentException: Widget [org.eclipse.nebula.widgets.passwordrevealer.PasswordRevealer] is not supported.

Solution

1. Please make this widget compatible to bind like a normal Text widget OR
2. Give directions/example code how to create a data binding between this widget and a data model

Hi

The popup window appears fleetingly and is quickly closed.
I tried the provided snippet. The same result is observed.

Environment: Linux, GTK3.

Our users are trying to type in the text field but its not writable while the text color is white so its a bit confusing. It would be nice when the multichoice is read-only to make the text also gray just like when you use the setEditable method to false on a text or combo widget.

Currently, the Nebula SVG support is inside the CWT project. We would like to pull out SVG support into its own project.

The topic alread describe it....

Currently DualList is very basic and requires one to create very special objects (DLItem) to be used. It would be better to see a DualList as a StructuredViewer with

1. a ContentProvider (like ListViewer)
2. a label provider for colors/fonts/icons and so on
3. a "selection" that represent the currently selected items "on the right"
4. a "selection" that represents the currently selected items "on the left"

A user than might add selectionchangedlisteners for any of the both lists of items

In the GanttChart widget, there is a large gap between ZOOM_HOUR_MAX (value 0) and ZOOM_MINUTES_NORMAL (value -1). On my screen, it goes from a width of about 11 hours to 10 minutes. Making impossible to correctly focus on data spanning e.g. 1 hours.

Looking at the code in GanttComposite.updateZoomLevel(), we see that the "day" width is reduced, but what a "day" represent is also changed (hour -> minute), thus creating the gap.

When a Grid is placed inside of a ScrollableComposite, then the grid contents are sometimes rendered outside of the area of the scrollable. This is most noticeable on the initial creation, and when resizing the entire window. Scrolling the grid stays within the bounds, but leaves a mess outside of the scrollable.

--- should be -->

Test code that demonstrates this: ScrollableGridTest.java.txt

Discovered on Linux with GTK (Ubuntu 20.04), and yet not a problem on Windows. This used to be fine on older SWT/GTK (Eclipse 3.8) and issue has surfaced since moving to a newer platform (i.e. Eclipse 4.1x). Perhaps something changed within GTK 3?

I think this may be due to the calls to gc.setClipping() in onPaint() before calling the paint() method of the cell renderer, as these are ignoring the clipping from the paint event.

I was able to fix it by saving the original clipping at the top of the function, and then replacing gc.setClipping(new Rectangle(...)) with gc.setClipping(originalClipping.intersection(new Rectangle(...))). This ensures that the clipping is never set to be outside that of the paint event. See Grid.diff.txt. Might need to do something with insertMarkPosFound block too.

AFAICS the PasswordRevealer shows only the PW if the user continues to hold the mouse button. I think most websites react differently. If you select "Show password" it shows the PW until the user selects this option again.

Can this also be provided by PasswordRevealer? Maybe even as default as I think this is what the user expects (I might be wrong here).

I think it is better if the preference window opens on the same screen as the running application. Currently it seems to use the laptop (I guess the primary screen) and not the screen on which my application is running.

Looks like this is done in SWTGraphicUtil in PreferenceWindow#openShell

Added only this line after the main view creation:

Dialog.inform("Loading", "Test");

This is the error:

Exception in thread "main" java.lang.NoSuchMethodError: 'void org.eclipse.nebula.widgets.opal.commons.SWTGraphicUtil.addDisposer(org.eclipse.swt.widgets.Widget, org.eclipse.swt.graphics.Resource)'
	at org.eclipse.nebula.widgets.opal.dialog.DialogArea.getTitleColor(DialogArea.java:105)
	at org.eclipse.nebula.widgets.opal.dialog.MessageArea.createTitle(MessageArea.java:286)
	at org.eclipse.nebula.widgets.opal.dialog.MessageArea.render(MessageArea.java:219)
	at org.eclipse.nebula.widgets.opal.dialog.Dialog.show(Dialog.java:111)
	at org.eclipse.nebula.widgets.opal.dialog.Dialog.inform(Dialog.java:297)
	at org.eclipse.nebula.widgets.opal.dialog.Dialog.inform(Dialog.java:286)
	at net.VisualDivX.Controller.Controller.<init>(Controller.java:30)

I've only created the main Display and Shell. Afterwards, I get this error message. With java 8 was working fine, but I need java 11. 
My java version is  AdoptOpenJDK 11.0.8. Have you tested with java 8+ and above?

Sorry if I missed something, but I believe that the new RoundedSwitch lacks the code that actually broadcasts the selection event.

see

A Split Button is a Push Button that also contains a drop-down menu.

The drop-down menu typically contains modifiers for how the push button action is performed. In this example the UI allows the user to change what kind of search is performed:

Sadly in the JFace Action implementation, SWT.CHECK and SWT.DROP_DOWN are mutually exclusive:

	/**
	 * Creates a new action with the given text and style.
	 *
	 * @param text
	 *            the action's text, or <code>null</code> if there is no text
	 * @param style
	 *            one of <code>AS_PUSH_BUTTON</code>,
	 *            <code>AS_CHECK_BOX</code>, <code>AS_DROP_DOWN_MENU</code>,
	 *            <code>AS_RADIO_BUTTON</code>, and
	 *            <code>AS_UNSPECIFIED</code>.
	 */
	protected Action(String text, int style) {
		this(text);
		switch (style) {
		case AS_PUSH_BUTTON:
			value = VAL_PUSH_BTN;
			break;
		case AS_CHECK_BOX:
			value = VAL_TOGGLE_BTN_OFF;
			break;
		case AS_DROP_DOWN_MENU:
			value = VAL_DROP_DOWN_MENU;
			break;
		case AS_RADIO_BUTTON:
			value = VAL_RADIO_BTN_OFF;
			break;
		}
	}

On first impression it looks like a new version of JFAce Action is required, though maybe there's another strategy.

According to the SVG specification (https://svgwg.org/specs/paths/#PathData), the path data for all commands could have multiple values. For example, when drawing a line:

Draw a line from the current point to the given (x,y) coordinate which becomes the new current point. L (uppercase) indicates that absolute coordinates will follow; l (lowercase) indicates that relative coordinates will follow. A number of coordinates pairs may be specified to draw a polyline. At the end of the command, the new current point is set to the final set of coordinates provided.

The current method ignores any additional coordinate pairs. The multi-values are applicable to all commands except 'Z' ('z') and 'A' ('a'). 'M' ('m") is a special case and any additional coordinate pairs should be treated as 'L' ('l').

In GeoMapUtil.zoomOut, geoMap.setZoom is called, which updates the zoom level and then calls fireZoomChanged(), which causes the listeners to respond to a change in the zoom.

The problem is that the map position is only updated in GeoMapUtil.zoomOut after the geoMap.setZoom call - which means any listeners responding to the zoom change will be using an outdated map position.

I think in zoomIn & zoomOut the statements should simply be switched in order.

The description on the example page says:

A switch button is a different graphic representation of a checkbox button.

but one need to call SwitchButton.setSelection(true) to have a button that is off... that's not very intuitive and the opposite what a checkbox behaves...

1. Create a TableCombo and set it to disabled
2. set one item to it
3. set two items and enable it
4. open it -> appears to small (with scrollbars)
5. open it again --> looks fine

If a VControl only contains a SvgDocument the size is only 2x2 (plus margins) instead of using the width/height or viewBox of the SvgDocument.

Is it possible to add WMTS / WMS support to this map view?

Create a Floating Text widget which is common in mobile applications.

Instances of this class are selectable user interface objects that allow the
user to enter and modify text with a label floating above the input area.
FloatingText controls can be either single or multi-line. In contrast to a
Text widget, when a floating text control is created with a border, a custom
drawn inset will be painted INSTEAD OF the platform specific inset.

Styles are inherited from Text and subclasses:

CENTER, ICON_CANCEL, ICON_SEARCH, LEFT, MULTI, PASSWORD, SEARCH, SINGLE,
RIGHT, READ_ONLY, WRAP, BORDER, H_SCROLL, V_SCROLL, LEFT_TO_RIGT,
RIGHT_TO_LEFT, FLIP_TEXT_DIRECTION

Can you add https://www.eclipse.org/nebula/downloads.php to your Readme.md? This would make it easier to find the update sites...

The problem has already been described here:
https://bugs.eclipse.org/bugs/show_bug.cgi?id=541345

I will copy the discussion in here anyway:

Tommy Redel

Hello,

when I use a list followed by text in a RichTextCellPainter the resulting calculated preferred row height is wrong.

The cleaned HTML in the RichTextPainter looks as follows:
<root><ul><li>Line1</li></ul>Test</root>

In the paintHTML method of the RichTextPainter this will result in just a single line. So the LinePainter will just be asked once assuming there is just one single line.

When the content is painted, everything looks fine. So the list with the single item gets its own line and the Text "Test" gets a second line.

So this is just a problem when caluclating the preferred row height.

Greetings
Tommy

Dirk Fauth

I just noticed that the issue is not in NatTable, it is in the RichTextPainter of the Nebula project.

You need to create an issue with the Nebula project. As the Nebula project has moved to GitHub, you should create an issue there and provide a pull request to fix this.

https://github.com/eclipse/nebula

IMHO the issue is that on handling the end tag of the unordered list, the currentLine variable is not set to null. Therefore the the calculation takes the following characters as part of the last line of the list.

By adding the following lines at the end of the case block (line 392) the issue should be fixed.

// set currentLine to null as any further content will start on a new line
currentLine = null;

I will close the ticket here as WONTFIX, as it is not NatTable and Nebula now uses GitHub issues for bug tracking. You can copy my answer there and provide a pull request to fix this. Hopefully this does not brake anything else. But in theory this should be the correct handling. In HTML there will be always a new line if something comes after a list end tag.

I encounter a large performance penalty when using a large Grid but only a small portion is visible.

From a quick debugging it seems (I was not able to profile it or try different strategies) that the Grid is computing a lot of values unnecessarily, for example the following code:

for (final GridColumn column : displayOrderedColumns) {

	if (column.isVisible()) {
		emptyCellRenderer.setBounds(x, y, column.getWidth(), getItemHeight());
		emptyCellRenderer.setColumn(indexOf(column));
		emptyCellRenderer.paint(e.gc, this);

		x += column.getWidth();
	}
}

if (x < getClientArea().width) {
	emptyCellRenderer.setBounds(x, y, getClientArea().width - x + 1, getItemHeight());
	emptyCellRenderer.setColumn(getColumnCount());
	emptyCellRenderer.paint(e.gc, this);
}

x = 0;

As one can see if the (final) value is higher than the bounds it is discarded so it seems suitable to break the previous loop if the limit is reached. I also noticed that column.isVisible() always calls "checkWidget" it might be good to have package private methods to bypass those checks to prevent additional efforts.

I'd like to know: Is there any "performance" tests that show on how large datasets a Grid could/should be used? Are there already any actions to make it more performant?

The Floating Text Widget was created via #11 I just clone the repo and executed FloatingTextSnippet and here is the result:

This looks wrong.

For some reason the initial look is not restored for SWT.MULTI

PreferenceWindow#buildButtons uses Ok and Cancel. Maybe these default text value coudl be "Apply" or "Close" or "Apply and Close" for the OK button?

I would be great if FloatingText also supports the "group" usage on the right hand side of the Android UX guidelines.

See https://material.io/components/text-fields#theming for the full guide.

If possible the background of the PWR should be transparent, otherwise it looks strange.

I have a TableCombo and set it to readonly, now the text is always painted "selected" when I choose a value from the drop down.

In org.eclipse.nebula.widgets.geomap.TileServer, a maximum zoom level is specified.

Some Tile Servers - e.g. Ordnance Survey's Maps API - specify a minimum zoom level, too - https://osdatahub.os.uk/docs/wmts/technicalSpecification

Currently the minimum zoom level is hard-coded to 0.

I think this should be dynamic.

For every Notifier instance there's one Image created (in line 211) which isn't disposed.

To verify, import the org.eclipse.swt.tools.spies project and add the following lines to the NotifierSnippet:

public static void main(final String[] args) {
  DeviceData data = new DeviceData();
  data.tracking = true;
  final Display display = new Display(data);
  Sleak sleak = new Sleak();
  Shell sleakShell = new Shell();
...

Currently one can't configure the list, eg:

• allow up/down
• if text should be centered or not ...

If you press the arrow key down without entering anything beforehand, Java throws an ArithmeticException. If you type something in and then delete it again, Java won't throw the exception.

java.lang.ArithmeticException: / by zero
at org.eclipse.nebula.widgets.opal.textassist.TextAssist$1.handleEvent(TextAssist.java:163)
at org.eclipse.swt.widgets.EventTable.sendEvent(EventTable.java:89)
at org.eclipse.swt.widgets.Display.sendEvent(Display.java:4441)
at org.eclipse.swt.widgets.Widget.sendEvent(Widget.java:1512)
at org.eclipse.swt.widgets.Widget.sendEvent(Widget.java:1535)
at org.eclipse.swt.widgets.Widget.sendEvent(Widget.java:1520)
at org.eclipse.swt.widgets.Widget.sendKeyEvent(Widget.java:1549)
at org.eclipse.swt.widgets.Text.sendKeyEvent(Text.java:1765)
at org.eclipse.swt.widgets.Widget.sendKeyEvent(Widget.java:1545)
at org.eclipse.swt.widgets.Text.sendKeyEvent(Text.java:1730)
at org.eclipse.swt.widgets.Control.doCommandBySelector(Control.java:1085)
at org.eclipse.swt.widgets.Display.windowProc(Display.java:6251)
at org.eclipse.swt.internal.cocoa.OS.objc_msgSendSuper(Native Method)
at org.eclipse.swt.widgets.Widget.callSuper(Widget.java:235)
at org.eclipse.swt.widgets.Widget.superKeyDown(Widget.java:1960)
at org.eclipse.swt.widgets.Widget.keyDown(Widget.java:1130)
at org.eclipse.swt.widgets.Control.keyDown(Control.java:2457)
at org.eclipse.swt.widgets.Display.windowProc(Display.java:6083)
at org.eclipse.swt.internal.cocoa.OS.objc_msgSendSuper(Native Method)
at org.eclipse.swt.widgets.Widget.callSuper(Widget.java:235)
at org.eclipse.swt.widgets.Widget.windowSendEvent(Widget.java:2150)
at org.eclipse.swt.widgets.Shell.windowSendEvent(Shell.java:2487)
at org.eclipse.swt.widgets.Display.windowProc(Display.java:6203)
at org.eclipse.swt.internal.cocoa.OS.objc_msgSendSuper(Native Method)
at org.eclipse.swt.widgets.Display.applicationSendEvent(Display.java:5442)
at org.eclipse.swt.widgets.Display.applicationProc(Display.java:5585)
at org.eclipse.swt.internal.cocoa.OS.objc_msgSend(Native Method)
at org.eclipse.swt.internal.cocoa.NSApplication.sendEvent(NSApplication.java:117)
at org.eclipse.swt.widgets.Display.readAndDispatch(Display.java:3834)
at org.eclipse.e4.ui.internal.workbench.swt.PartRenderingEngine$5.run(PartRenderingEngine.java:1157)
at org.eclipse.core.databinding.observable.Realm.runWithDefault(Realm.java:338)
at org.eclipse.e4.ui.internal.workbench.swt.PartRenderingEngine.run(PartRenderingEngine.java:1046)
at org.eclipse.e4.ui.internal.workbench.E4Workbench.createAndRunUI(E4Workbench.java:155)
at org.eclipse.e4.ui.internal.workbench.swt.E4Application.start(E4Application.java:168)
at org.eclipse.equinox.internal.app.EclipseAppHandle.run(EclipseAppHandle.java:203)
at org.eclipse.core.runtime.internal.adaptor.EclipseAppLauncher.runApplication(EclipseAppLauncher.java:134)
at org.eclipse.core.runtime.internal.adaptor.EclipseAppLauncher.start(EclipseAppLauncher.java:104)
at org.eclipse.core.runtime.adaptor.EclipseStarter.run(EclipseStarter.java:401)
at org.eclipse.core.runtime.adaptor.EclipseStarter.run(EclipseStarter.java:255)
at java.base/jdk.internal.reflect.NativeMethodAccessorImpl.invoke0(Native Method)
at java.base/jdk.internal.reflect.NativeMethodAccessorImpl.invoke(NativeMethodAccessorImpl.java:62)
at java.base/jdk.internal.reflect.DelegatingMethodAccessorImpl.invoke(DelegatingMethodAccessorImpl.java:43)
at java.base/java.lang.reflect.Method.invoke(Method.java:566)
at org.eclipse.equinox.launcher.Main.invokeFramework(Main.java:653)
at org.eclipse.equinox.launcher.Main.basicRun(Main.java:590)
at org.eclipse.equinox.launcher.Main.run(Main.java:1461)
at org.eclipse.equinox.launcher.Main.main(Main.java:1434)

Our TextAssistContentProvider looks like this:

        TextAssistContentProvider contentProvider = new TextAssistContentProvider() {

		@Override
		public List<String> getContent(String entry) {
			ArrayList<String> result = new ArrayList<>();
			Instant date = DateUtil.getDate(entry, locale);
			if (date == null && !entry.isEmpty()) {
				result.add("!Error converting");
			} else {
				LocalDate localDate = LocalDate.ofInstant(date, ZoneId.of("UTC"));
				result.add(localDate.format(DateTimeFormatter.ofLocalizedDate(FormatStyle.MEDIUM).withLocale(locale)));
				field.setValue(new Value(date), true);
			}
			return result;
		}

	};

The SWTGraphicUtil.centerShell(shell); centeres the shell, after opening it hence a smaller flickering is seen.
Maybe center it before opening i?

In PreferenceWindow.java:

 * Open the shell
 */
private void openShell() {
	shell.pack();
	shell.open();
	SWTGraphicUtil.centerShell(shell);

	while (!shell.isDisposed()) {
		if (!shell.getDisplay().readAndDispatch()) {
			shell.getDisplay().sleep();
		}
	}

}

/**

I try to use the FontChooser in my PreferenceWindow, but it doesn't display the value bound to the property key.

in the referenced line, the following code should be added

text.setText(buildFontInformation());

When the user adds a prompt support to a text widget, the prompt support retrieves the style (colors and font) from the widget but does not keep track of intermediate changes to the text widget regarding this style (color and font).

This might result in problems when the user changes the style after prompt support was added to the widget. e.g. the original styles might be disposed.

I tried using PasswordRevealer as drop-in replacement for a text field but it seems to have confuse the layout:

Original:
Text password = new Text(container, SWT.BORDER | SWT.PASSWORD);
Text password.setLayoutData(GridDataFactory.fillDefaults().hint(130, SWT.DEFAULT).span(2, 1).create());.

After changing the code:

PasswordRevealer password = new PasswordRevealer(parent, SWT.BORDER);
password.setLayoutData(GridDataFactory.fillDefaults().hint(130, SWT.DEFAULT).span(2, 1).create());

Maybe it modifies the layout on the parent?

org.eclipse.nebula.widgets.opal.dialog.Dialog.center() has the following code

		if (centerPolicy == CenterOption.CENTER_ON_SCREEN || shell.getParent() == null) {
			Shell activeShell = shell.getDisplay().getActiveShell();
			final Rectangle monitorBounds = SWTGraphicUtil.getBoundsOfMonitorOnWhichShellIsDisplayed(activeShell);

This will fail if there is no active shell (e.g. application does not has focus). If there is no actie shell, it would be the best to use the dialog shell...

If no image is supplied all text is centered.