Why do you need to private part of a class and how to access them?

Answer 1

You don't need to make any part of a class private, but then it would behave no differently to a struct, which is public by default. The point of private access is to hide data and methods that are internal to the class; you are not supposed to be able to access them from outside the class.

There are three levels of access: private access is only accessible to the class itself, and to friends of the class. Protected access is the same as private, but is also accessible to derived classes. Public access is accessible to all.

To give an example, imagine the following structure:

struct s

{

unsigned int num;

};

Suppose the num member can be any positive value but must never be zero. That means that whenever we pass this structure to a function, we must assert or verify that num is non-zero. We are relying on those functions to perform verification that should rightly be performed by the structure. But num is public, so there's really no way to safeguard against the structure containing invalid data.

So let's replace the structure with a class:

class c

{

public:

c():num(1);

void setnum(unsigned int number){if(number !=0 ) num = number;}

unsigned int getnum()const{return( num );}

private:

unsigned int num;

};

Now we can be sure that num is never zero. The constructor automatically sets num to positive 1, and the setnum mutator only changes num if the given number parameter is non-zero. Meanwhile, the getnum accessor returns a copy of num, not a reference to num. Only the class itself has access to the private num variable, thus the data is hidden from outside influence, and the public interface guarantees that num will always be non-zero. We no longer need to verify this fact because that functionality is encapsulated within the class itself, where it belongs.