Static vs. Dynamic Method Binding

A Motivating Example

Consider the following class hierarchy:

 human  robot
   \     /
sentient_being

human robot

\ /

sentient_being

And, assume each class implements its own method called print_characteristics

#include<iostream>
using namespace std;

class sentient_being {
public:
	void print_characteristics(){
		cout << "Sentient_being's print_characteristics" << endl;
	}
};

class human : public sentient_being {
public:
	void print_characteristics(){
		cout << "Human's print_characteristics" << endl;
	}
};

class robot : public sentient_being {
public:
	void print_characteristics(){
		cout << "Robot's print_characteristics" << endl;
	}
};

#include<iostream>

using namespace std;

class sentient_being {

public:

void print_characteristics(){

cout << "Sentient_being's print_characteristics" << endl;

}

};

class human : public sentient_being {

public:

void print_characteristics(){

cout << "Human's print_characteristics" << endl;

}

};

class robot : public sentient_being {

public:

void print_characteristics(){

cout << "Robot's print_characteristics" << endl;

}

};

If we define a main method which creates a sentient_being, human, and robot, then calls print_characteristics on each, it is clear what should happen.

int main() {
	sentient_being s;
	human h;
	robot r;

	s.print_characteristics();
	h.print_characteristics();
	r.print_characteristics();

	return 0;
}

int main() {

sentient_being s;

human h;

robot r;

s.print_characteristics();

h.print_characteristics();

r.print_characteristics();

return 0;

}

Output:

Sentient Being's print_characteristics
Human's print_characteristics
Robot's print_characteristics

Sentient Being's print_characteristics

Human's print_characteristics

Robot's print_characteristics

The Issue at Hand

If we add the following to main:

sentient_being *x = &h;
sentient_being *y = &r;
x->print_characteristics();
y->print_characteristics();

sentient_being *x = &h;

sentient_being *y = &r;

x->print_characteristics();

y->print_characteristics();

what should be the result of calling print_characteristics on x and y?

Two Options

1. Choose method to call based on variable type (static method binding)
2. Choose method to call based on object type (dynamic method binding)

Note: Option 2 is also known as dynamic dispatch and is a major topic in object oriented programming — subtype polymorphism.

C++ uses static method binding by default, so the output of the above additions is:

Sentient Being's print_characteristics
Sentient Being's print_characteristics

1 2	Sentient Being's print_characteristics Sentient Being's print_characteristics

Other languages like Java use dynamic method binding by default. You may use dynamic method binding in C++ through the use of the virtual keyword. The important place to put this keyword in the above program is on sentient_being’s print_characteristics method:

class sentient_being {
public:
	virtual void print_characteristics(){
		cout << "Sentient_being's print_characteristics" << endl;
	}
};

class sentient_being {

public:

virtual void print_characteristics(){

cout << "Sentient_being's print_characteristics" << endl;

}

};

With no other changes, the output becomes:

Human's print_characteristics
Robot's print_characteristics

1 2	Human's print_characteristics Robot's print_characteristics

Hi There!

Static vs. Dynamic Method Binding

A Motivating Example

The Issue at Hand

Two Options