When Can the Initialization List of a Child Class Constructor Be Used to Initialize Protected Data Members?

Initialization List and Protected Member Initialization

In object-oriented programming using C , it is a common practice for classes to inherit from parent classes and extend their functionality. However, a question arises: can the initialization list of a child class' constructor be used to initialize data members declared as protected in the parent class?

In an attempt to achieve this, one may have code similar to the following:

<code class="cpp">class Parent {
protected:
    std::string something;
};

class Child : public Parent {
private:
    Child() : something("Hello, World!") {}
};</code>

However, this attempt will result in a compiler error, indicating that the class Child does not have a field named something. This is because the data member something is declared as protected in the parent class.

To initialize protected members using the initialization list of a child class, an alternative approach is required. A constructor with the required initialization can be added to the parent class, and this constructor can then be invoked from the child class' constructor. This can be achieved through constructor forwarding:

<code class="cpp">class Parent {
protected:
    Parent(const std::string& something) : something(something) {}

    std::string something;
};

class Child : public Parent {
private:
    Child() : Parent("Hello, World!") {}
};</code>

In this modified code, a protected constructor is added to the Parent class, and it takes a string parameter. The Child class' constructor then invokes the Parent class' protected constructor using constructor forwarding to pass the necessary argument and initialize the something data member.

The above is the detailed content of When Can the Initialization List of a Child Class Constructor Be Used to Initialize Protected Data Members?. For more information, please follow other related articles on the PHP Chinese website!