CODEWITHSUNDEEP
c++classinheritance
user13242010
user13242010

Reputation:

How to use protected type inheritance in class?

    #include<iostream>
#include<string>
#include<cmath>
using namespace std;
class Polygon {
public:
    Polygon() {}
    Polygon(int point, float length) {
        mPoint = point;
        mLength = length;
    }
    ~Polygon() {}
    virtual void calcPerimeter() {
        cout << "Perimeter: " << mLength * mPoint;
    }
    virtual void calcArea() {
        cout << "Area: " << sqrt(3) * pow(mLength, 2) / 4;
    }
protected:
    int mPoint;
    double mLength;
};

class Rectangle : public Polygon {
public:
    Rectangle() {}
    Rectangle(int point, float length):mPoint(4), mLength(4 * length){}
    ~Rectangle() {}
    void calcPerimeter() override {
        cout << "Perimeter: " << mLength;
    }
    void calcArea() override {
        cout << "Area: " << mLength * mLength;
    }

};

int main() {
    Polygon pol;
    Rectangle rec(4, 10);

    cout << "--- Polygon class---" << endl;
    pol.calcPerimeter();
    pol.calcArea();
    cout << "---Rectangle class---" << endl;
    rec.calcPerimeter();
    rec.calcArea();

    return 0;
}

I learned that if the protected part of the parent class is inherited as public, it is used like a private in the child class. By the way

Rectangle (int point, float length): mPoint (4), mLength (4 * length) {}

In this part, I get the error that mPoint and mLength are not non-static data members or the base class of Reactangle. If it's private, can't I use it like that in a class ??

If I'm misunderstanding, I hope you can tell me what's wrong. Thanks for reading.

Upvotes: 3

Views: 72

Answers (1)

Jasper Kent
Jasper Kent

Reputation: 3676

You're right about being able to use protected data in the derived class. However, that's not the problem here.

In

Rectangle (int point, float length): mPoint (4), mLength (4 * length) {}

you are initializing the fields, and that can only be done once, in the initializer list of the constructor of the class in which they are declared (Polygon).

You could assign them in the body of the constructor:

Rectangle (int point, float length)
{
    mPoint = 4;
    mLength = 4 * length;
}

but better is just to call the base class constructor and let it do its job:

Rectangle (int point, float length):Polygon (4, 4 * length) {}

All of this begs the question why you're passing point to the Rectangle constructor when you don't use it.

Also, it's widely regarded as bad practice to have protected fields (protect methods are okay) because once they are protected they are accessible all the way down the inheritance hierarchy. Better to keep them private and provide protected set methods.

Upvotes: 3

Related Questions