Question

I would like to know how can I extend JEditorPane (or any other swing text editing component) to handle rectangle (column) selection mode. It is a well known feature in current text editors where you can select multiple lines (rows) starting from an offset (column) and ending by an offset (column) which look like selecting a rectangle of text, and then what you type will overwrite the selection in each line (row) concurrently.

One idea was to override the selection and create fake selection by highlighting each line in rectangle form by following the mouse events, and keeping track of such information to use it when typing. However, I am not quit sure how to override the selection and track the mouse, nor how to redirect typing to affect each line.

Any help in any form would be appreciated.

Answer 1

Found out this little code snippet, that involves a Custom Caret (to handle the fragmented selection) and Highlighter (to show the fragments):

class MyCaret extends DefaultCaret {

Point lastPoint=new Point(0,0);
public void mouseMoved(MouseEvent e) {
    super.mouseMoved(e);
    lastPoint=new Point(e.getX(),e.getY());
}
public void mouseClicked(MouseEvent e) {
    super.mouseClicked(e);
    getComponent().getHighlighter().removeAllHighlights();
}

protected void moveCaret(MouseEvent e) {
    Point pt = new Point(e.getX(), e.getY());
    Position.Bias[] biasRet = new Position.Bias[1];
    int pos = getComponent().getUI().viewToModel(getComponent(), pt, biasRet);
    if(biasRet[0] == null)
        biasRet[0] = Position.Bias.Forward;
    if (pos >= 0) {
        setDot(pos);
        Point start=new Point(Math.min(lastPoint.x,pt.x),Math.min(lastPoint.y,pt.y));
        Point end=new Point(Math.max(lastPoint.x,pt.x),Math.max(lastPoint.y,pt.y));
        customHighlight(start,end);
    }
}

protected void customHighlight(Point start, Point end) {
    getComponent().getHighlighter().removeAllHighlights();
    int y=start.y;
    int firstX=start.x;
    int lastX=end.x;

    int pos1 = getComponent().getUI().viewToModel(getComponent(), new Point(firstX,y));
    int pos2 = getComponent().getUI().viewToModel(getComponent(), new Point(lastX,y));
    try {
        getComponent().getHighlighter().addHighlight(pos1,pos2,
                 ((DefaultHighlighter)getComponent().getHighlighter()).DefaultPainter);
    }
    catch (Exception ex) {
        ex.printStackTrace();
    }
    y++;
    while (y<end.y) {
        int pos1new = getComponent().getUI().viewToModel(getComponent(), new Point(firstX,y));
        int pos2new = getComponent().getUI().viewToModel(getComponent(), new Point(lastX,y));
        if (pos1!=pos1new)  {
            pos1=pos1new;
            pos2=pos2new;
            try {
                getComponent().getHighlighter().addHighlight(pos1,pos2,
                         ((DefaultHighlighter)getComponent().getHighlighter()).DefaultPainter);
            }
            catch (Exception ex) {
                ex.printStackTrace();
            }
        }
        y++;
    }
}
}

Anyway, I’ve never run that code (it’s Stanislav’s).