Given a random class:
public class A<T> {
public T t;
public A () {} // <-- why is this constructor necessary for B?
public A (T t) {
this.setT(t);
}
public T getT () {
return this.t;
}
protected void setT (T t) {
this.t = t;
return;
}
}
And an extended class:
public class B extends A<Integer> {
public B (Integer i) {
this.setT(i);
}
}
Why does B require A to have the empty constructor? I would have assumed it would want to use the similar constructor instead of the default constructor. I tried compiling without the default constructor, but I get the following message without it...
java.lang.NoSuchMethodError: A: method <init>()V not found at B.<init>
Can anyone explain why this is?
The important point is to understand that the first line of any constructor is to call the super constructor. The compiler makes your code shorter by inserting
super()
under the covers, if you do not invoke a super constructor yourself.Also if you do not have any constructors an empty default constructor - here
A()
orB()
- would automatically be inserted.You have a situation where you do not have a
super(...)
in your B-constructor, so the compiler inserts thesuper()
call itself, but you do have an A-constructor with arguments so the the default A()-constructor is not inserted, and you have to provide the A()-constructor manually, or invoke the A(i)-constructor instead. In this case, I would suggest just havingYou may use your own constructor in
A
, but you have to call it explicitly from the B's constructor, e.g.:If you don't do that, the compiler will attempt to instantiate
A
itself, by calling its default constructor.If you don't make a call to a super constructor using
super(i)
as the first line of your constructor it will implicitly call the default super constructor