Type Fields and switch Statements
One way to determine the type of an object that is incorporated in a larger program is to use a switch statement. This allows us to distinguish among object types, then invoke an appropriate action for a particular object. For example, in a hierarchy of shapes in which each shape object has a shapeType attribute, a switch statement could check the object's shapeType to determine which print function to call.
However, using switch logic exposes programs to a variety of potential problems. For example, the programmer might forget to include a type test when one is warranted, or might forget to test all possible cases in a switch statement. When modifying a switch-based system by adding new types, the programmer might forget to insert the new cases in all relevant switch statements. Every addition or deletion of a class requires the modification of every switch statement in the system; tracking these statements down can be time consuming and error prone.
Software Engineering Observation 13.6
|
Polymorphic programming can eliminate the need for unnecessary switch logic. By using the C++ polymorphism mechanism to perform the equivalent logic, programmers can avoid the kinds of errors typically associated with switch logic. |
Software Engineering Observation 13.7
|
An interesting consequence of using polymorphism is that programs take on a simplified appearance. They contain less branching logic and more simple, sequential code. This simplification facilitates testing, debugging and program maintenance. |