We are currently using Eclipse Draw2D/GEF for an information page describing a process in our application. This basically consists of a matrix of large squares, each containing a matrix of smaller squares. We originally had all the squares as GEF objects, but because of the large volume of them being shown, we found that this did not scale very well and the view took a very long time to open. We then changed it so that only the large squares are Figures and we then draw the smaller squares using the graphic in paintFigure.
The problem that we are running into is that we still want the tooltip to change depending on which small square you are hovering over. I tried to do this by adding a mouseMotionListener and setting the tooltip, through setTooltip, depending on where the mouse currently is. The problem is that once the tooltip is displayed, it does not change any more when setTooltip is called.
Does any one know of an alternative way of doing this? Is there a way of getting the viewpart's PopupHelper and using that? Any help would be appreciated.
Thanks
Hmnn.. interesting problem. Since you paint your own Grid within the Figure, I would think that you have two options.
Try posting SWT events to fool Eclipse. I'd try a focus lost followed by a focused gained, to trigger tooltip machinery, at which point you could get the coordinates and display the appropriate contents.
Don't use the Figure#getTooltip strategy at all. Just show your own composite.
To dynamically change the tooltip, you can hold an instance of the tooltip Figure in your parent Figure. In the constructor of the parent Figure, create a new tooltip Figure (e.g. a Label) and use setToolTip() method to set the tooltip Figure to parent Figure.
When data model is changed, the updated tooltip text/icon can be set to the tooltip Figure. Then you just call setToolTip(tooltipFigure) method again.
You can have a method like:
protected Label toolTipLabel;
protected void updateToolTip(String text, Image icon){
toolTipLabel.setText(text);
toolTipLabel.setIcon(icon);
setToolTip(toolTipLabel);
}
The updateToolTip() method can be invoked in parent Figure's conturctor to initialize the tooltip. And this method can be invoked each time after the data model is changed.
I encountered the same problem in my code and solved it with that method. In my code, I invoked the updateToolTip() in the parentFigure.paintFigure() method.
Related
I implemented my own JFreeChart XYToolTipGenerator and, as the chart it is used on is almost full screen, sometimes the tooltip position (on screen) hides the point it is related to (e.g. in the bottom right corner, since it seems that tooltip is configured to be positioned South-East of the mouse / data point). This is a problem because the user needs to be able to click on the chart's data points (as it generates a specific action).
Is there a way to either define dynamically the position of the tooltip (e.g. for data points bottom right I would ask the tooltip to be shown North-West) or, alternatively, to define a systematic position (e.g. North-West instead of South-East as it is by default)?
This problem has given me headaches for the last few days - any help or hint is more than welcome.
Many thanks!
Thomas
Here's the answer I posted on the JFreeChart forum:
JFreeChart is using the standard Swing tool tip mechanism. In the ChartPanel class, the getToolTipText(mouseEvent) method is overridden to return the appropriate text for the tooltip, and that's it.
Swing also gives you the option to override the getToolTipLocation(mouseEvent) method, and that's probably what you need here.
I would like to display multiple texts/scores on my stage. This tutorial suggested SpriteBatch, however, the issue is that these texts won’t fade out (not actor). Certain posts suggested use table, but the issue is that I can’t set texts' position on stage. Is there any other way to display texts as an actor on stage without using table?
Thanks in advance.
Zi
You can create a subclass of actor and draw text as you like in it's overridden draw method.
Just don't forget to set alpha before drawing using setColor method. The value of alpha will be available as a parameter for draw.
You then can fade it the way you want.
Good luck.
I am trying to build a Media Player. I have all the code for music playback. And the GUI is made up of 3 main parts:-
A blue background.
A JList with JScrollPane
A JLayer Shell (with a polygon cut out to reveal the list and blue background beneath it.)
So my question is how can I create a custom list or JTable that is slightly rotated with the base of the component stretched out on both corners (using a perspective filter). Also the JScrollbar attached must be customized also. I guess at least it needs to be rotated. But would like to know how to make one by supplying my own graphics. Is this possible?
Here is a MOCKUP of the design:
http://www.splashportal.net/MOCKUP/iJuk-MOCKUP.swf
Here is a Screenshot of it:
You can transform a view component, as shown here, but mouse interaction is effectively impossible without re-writing the UI delegate. Alternatively, you may be able to adapt the approach shown here that uses an inverse transform to effect mouse control.
I am looking for ways to zoom in a Java Swing application. That means that I would like to resize all components in a given JPanel by a given factor as if I would take an screenshot of the UI and just applied an "Image scale" operation. The font size as well as the size of checkboxes, textboxes, cursors etc. has to be adjusted.
It is possible to scale a component by applying transforms to a graphics object:
protected Graphics getComponentGraphics(Graphics g) {
Graphics2D g2d=(Graphics2D)g;
g2d.scale(2, 2);
return super.getComponentGraphics(g2d);
}
That works as long as you don't care about self-updating components. If you have a textbox in your application this approach ceases to work since the textbox updates itself every second to show the (blinking) cursor. And since it doesn't use the modified graphics object this time the component appears at the old location. Is there a possibility to change a components graphics object permanently? There is also a problem with the mouse click event handlers.
The other possibility would be to resize all child components of the JPanel (setPreferredSize) to a new size. That doesn't work for checkboxes since the displayed picture of the checkbox doesn't change its size.
I also thought of programming my own layout manager but I don't think that this will work since layout managers only change the position (and size) of objects but are not able to zoom into checkboxes (see previous paragraph). Or am I wrong with this hypothesis?
Do you have any ideas how one could achieve a zoomable Swing GUI without programming custom components? I looked for rotatable user interfaces because the problem seems familiar but I also didn't find any satisfying solution to this problem.
Thanks for your help,
Chris
You could give a try to the JXLayer library.
There are several tools in it, which could help you to make a zoom. Check the examples shown here. I would recommend you to read more about the TransformUI, from this library. From the example, it seems like it could help solving your problem.
Scaling the view is easy enough; transforming mouse coordinates is only slightly more difficult. Here's an elementary example. I'd keep JComponents out, although it might make sense to develop an analogous ScaledComponent that knows about the geometry. That's where #Gnoupi's suggestion of using a library comes in.
hey you can try this if you want to zoom a image like any other image viewer the use a JPanel draw an image using drawImage() method now create a button and when you click the button increase the size of the panel on the frame it appears as if the image is being viewed in Zoom
You might find Piccolo2D.java API useful: http://code.google.com/p/piccolo2d/
It is very simple.
It touts in particular its smooth zooming. You essentially make a "canvas" that can contain various elements, and can then zoom by just holding right-click and panning the mouse back and forth.
I worked on a team that used it to create this: http://sourceforge.net/apps/mediawiki/guitar/index.php?title=WebGuitar#EFG.2FGUI_Visualizer
The nodes you see there are clickable links themselves.
Since Java 9, there are VM arguments (actually meant to be used for high dpi scaling) that can render a application with a higher scaling factor:
java -Dsun.java2d.uiScale=2.0 -jar MyApplication.jar
Or:
java -Dsun.java2d.win.uiScaleX=2.0 -Dsun.java2d.win.uiScaleY=2.0 -jar MyApplication.jar
I have to use a GUI Element that draws a picture at a specific screen position.
If the user selects this picture there is a border drawn around the Image.
Now we want to include another border that identifies pictures with a specific value for the user.
At the moment the Element looks at his internal state if it is selected and then decides how to draw itself.
graphic.drawImage(icon, x, y, null);
if (selected) {
drawBorder();
}
I don't like the idea of adding another if else to this drawing method.
I thought about creating a new class that inherits the behavior of the element and overwrites the draw method but that means duplicating the whole selected code in every inherited class.
Is there a nice possibility so solve this problem without creating a subclass?
Since you tagged this with design-patterns and you seem to be looking for a pattern-oriented approach, I'd suggest taking a look at the state pattern. The example on the wikipedia page even mentions keeping state while drawing a GUI. Unfortunately, this would mean you'd have to create another class with subclasses and overridden methods.
Is this going to be something that is likely to change? I.e. do you realistically think you're going to be adding new behavior to the drawing (e.g. if the user double clicks, draw a different type of border; if the user right clicks, change the color of the border), or is this it? If you see more behavior being added, I think going ahead and taking a more OO approach is wise. If it's just these two cases, I'd say just add and else if statement.
What do you have against if-else?
It makes less sense to me to create a whole new object for the selected item than to check a flag in the drawing function.
one possibility is to allow your drawBorder() method to take parameters:
private void drawBorder(boolean isSelected, boolean hasSpecialValue);
this method can determine which type of border to draw.