Creating subclasses to handle "base-cases"

Question

Every now and then I see code where the programmer has extended a certain class to handle "base cases". For example, if I were creating a Tuple class to represent a tuple of integers, I would have something like

class Tuple {
    private int[] values;

    ...
}

But some people will extend Tuple to accommodate a few base-cases, such as a tuple of length 0, 1 or 2:

class Tuple0 extends Tuple {
    ...
}

class Tuple1 extends Tuple {
    private int value1;

    ...
}

class Tuple2 extends Tuple {
    private int value1;
    private int value2;

    ...
}

Of course, the methods (denoted by ...) of each TupleN class will be much simpler than those of the generic Tuple class. So, my question is: is there any benefit to this type of practice? If so, when and why; if not, why not?

Answer 1

The big drawback of this approach is that it leads to increased code size and less maintainable code. A potential advantage is possibly increased performance, but you really need to profile to make sure.

Personally, I wouldn't do this.