Type Fields & Switch Statements

• Imagine that you have a class hierarchy of shapes:

  Shape:

  • Polygon
    • Triangle
    • Quadrilateral
      • Parallelogram
        • Rectangle
          • Square
        • Rhombuse
      • Trapezoid
  • Circle
• Imagine that you have an applet that declares an array of Shape:

  Shape arrayOfShapes[] = new Shape[100];

• Imagine that you have a method that iterates through the array, 1 element at a time, and draws each shape encountered by invoking the draw method of the appropriate subclass.

  To distinguish each subclass of Shape, you give Shape an int instance variable shapeType.

  Each subclass inherits this instance variable and statically sets it to a value representing that subclass of Shape.

  You invoke the proper draw method via a switch statement:

public void paint(Graphics g) {

for (int i = 0; i < arrayOfShapes.length; i++)

switch (arrayOfShapes[i].getShapeType() ) {

case TRIANGLE:

arrayOfShapes[i].drawTriangle(g);
break;

case SQUARE:

arrayOfShapes[i].drawSquare(g);
break;

. . .

 

case CIRCLE:

arrayOfShapes[i].drawCircle(g);
break;

}

}

• If you add a new subclass of Triangle, say Scalene (with its associated draw method), your switch statement must by modified to handle the new case.
• This approach is far more error prone and cumbersome than a polymorphic approach.