子类可以从与接口无关的父类继承接口实现吗

Question

This is kind of quirk or corner case. Consider this example:

interface I {
    void foo();
}

class B {
    public void foo() {}
}

class C extends B implements I {}

With my compiler (OpenJDK in Linux) it compiles and the following works:

I c = new C();
c.foo();

Why? Is it intentional and backed by the language standard? How the runtime knows to follow first "down" from I to C and then "up" from C to B ? While that would be clear in case B implements I , it seems counter intuitive given B unrelated to I .

Answer 1

Well, in the example you mentioned c is an instanceof the class C . Hence, it would require to have it's own method foo() .

Now, C has gotten all public (and default?) methods from B and foo() happens to be one that is (technically) present, thanks to B . Hence, the requirement that the interface placed on C , to have its own foo() method, is fulfilled by B s foo() .

I don't think the down from I to C, and up from C to B is relevant here, in the way you mentioned it.

EDIT:

All evidence points to the fact that the compiler has at-hand, a list of implemented functions, both that are in the given class, and 'virtual' ones that are extended from the parent class.

As a result, if the aforementioned class is implementing some interface, both extended and the classes own functions are tallied against the ones that needed to be implemented (because of the interface).

Answer 2

When you written class C extends B implements I following will happen From JLS 8.4.8 Inheritance, Overriding, and Hiding :

A class C inherits from its direct superclass all concrete methods m (both static and instance) of the superclass for which all of the following are true:

• m is a member of the direct superclass of C.
• m is public, protected, or declared with package access in the same package as C.
• No method declared in C has a signature that is a subsignature (§8.4.2) of the signature of m.

A class C inherits from its direct superclass and direct superinterfaces all abstract and default (§9.4) methods m for which all of the following are true:

• m is a member of the direct superclass or a direct superinterface, D, of C.

• m is public, protected, or declared with package access in the same package as C.

• No method declared in C has a signature that is a subsignature (§8.4.2) of the signature of m.

• No concrete method inherited by C from its direct superclass has a signature that is a subsignature of the signature of m.

• There exists no method m' that is a member of the direct superclass or a direct superinterface, D', of C (m distinct from m', D distinct from D'), such that m' from D' overrides the declaration of the method m.

So its inherits abstract method from interface I and a concrete methods with same signature from class B .

Answer 3

It doesn't really "follow" like you think. C inherits all non-private methods from B , and C can do whatever it wants with those methods; override them, use them to implement interfaces (if the method is public ), invoke them, etc.

So the method foo is a method of C , even if it came by the method by inheriting it from B . And a class can expose any of its public methods through an interface .

Answer 4

The first thing to note is the common effect of extending B and implementing I . The common effect or the obvious outcome in java is that you can have only one method with a particular name, datatype and parameters(same type and equal in number) .

Any method which has the same name and different number of arguments would be a case of method overloading. But in your case, it's pretty simple that you need to implement and define the abstarct method foo() in I in a concrete subclass which in this case is C .

It's not compulsory to redefine every method in the superclass but it becomes compulsory when you are implementing an interface. So Java considers this method foo() in the class C as the inherited method from the interface I . Otherwise you would have seen an error asking you to redefine the method foo() which is inherited from the interface.

Unknowingly this also becomes a case of method overriding when you consider the superclass B . Since the foo() method in C has the same kind of declaration as in B and I (which it should according to the standards, rules and to satisfy the rules of implementing an interface), it easily satisfies the interface as well as overrides the method in B .

In your case, you are also overriding the method foo() in class B .

So one condition as well as a complementary offer (overriding) is being satisfied.

In the end, class C has just one foo() method which overrides the method in B as well as satisfies the condition of implementing interface I .

An example:

Let the method foo() be defined (in C ) as:

void foo()
{
System.out.println("Hello");
}

Create 2 objects in C as follows:

B b=new C();
C c=new C();

Call foo() :

b.foo();
c.foo();

Ouput: Both prints Hello .

Conclusion: foo() in C satisfies the interface (no error seen) and also overrides the method foo() in B .