C++虚函数机制解析

1． 无继承的情况

#include <iostream>
using namespace std;

class Base {
public:
    virtual void f() { cout << "Base::f()" << endl; }
    virtual void g() { cout << "Base::g()" << endl; }
    virtual void h() { cout << "Base::h()" << endl; }
};

int main()
{
    typedef void (*Fun)();

    Base *b = new Base;
    cout << *(int*)(&b) << endl; //虚函数表的地址存放在对象最开始的位置

    Fun funf = (Fun)(*(int*)*(int*)b);
    Fun fung = (Fun)(*((int*)*(int*)b + 1));
    Fun funh = (Fun)(*((int*)*(int*)b + 2));


    funf();
    fung();
    funh();


    cout << (Fun)(*((int*)*(int*)b + 3)); // 最后一个位置为0，表明虚函数表的结束

    return 0;
}


注意：在上面这个图中，虚函数表中最后一个节点相当于字符串的结束符，其标志了虚函数表的结束，在Codeblocks下打印为0。 

#include <iostream>
using namespace std;

class Base {
public:
    virtual void f() { cout << "Base::f()" << endl; }
    virtual void g() { cout << "Base::g()" << endl; }
    virtual void h() { cout << "Base::h()" << endl; }
};

class Derive: public Base {
    virtual void f1() { cout << "Derive::f1()" << endl; }
    virtual void g1() { cout << "Derive::g1()" << endl; }
    virtual void h1() { cout << "Derive::h1()" << endl; }
};

int main()
{
    typedef void (*Fun)();

    Base *b = new Derive;
    cout << *(int*)b << endl;
    Fun funf = (Fun)(*(int*)*(int*)b);
    Fun fung = (Fun)(*((int*)*(int*)b + 1));
    Fun funh = (Fun)(*((int*)*(int*)b + 2));
    Fun funf1 = (Fun)(*((int*)*(int*)b + 3));
    Fun fung1 = (Fun)(*((int*)*(int*)b + 4));
    Fun funh1 = (Fun)(*((int*)*(int*)b + 5));


    funf(); // Base::f()
    fung(); // Base::g()
    funh(); // Base::h()
    funf1(); // Derive::f1()
    fung1(); // Derive::g1()
    funh1(); // Derive::h1()

    cout << (Fun)(*((int*)*(int*)b + 6));
    return 0;
}


从上表可以发现：

1．  虚函数按照其声明顺序放于表中。

2．  父类的虚函数在子类的虚函数前面。 

3． 继承，虚函数覆盖的情形

#include <iostream>
using namespace std;

class Base {
public:
    virtual void f() { cout << "Base::f()" << endl; }
    virtual void g() { cout << "Base::g()" << endl; }
    virtual void h() { cout << "Base::h()" << endl; }
};

class Derive: public Base {
    virtual void f() { cout << "Derive::f()" << endl; }
    virtual void g1() { cout << "Derive::g1()" << endl; }
    virtual void h1() { cout << "Derive::h1()" << endl; }
};

int main()
{
    typedef void (*Fun)();

    Base *b = new Derive;
    cout << *(int*)b << endl;
    Fun funf = (Fun)(*(int*)*(int*)b);
    Fun fung = (Fun)(*((int*)*(int*)b + 1));
    Fun funh = (Fun)(*((int*)*(int*)b + 2));
    Fun fung1 = (Fun)(*((int*)*(int*)b + 3));
    Fun funh1 = (Fun)(*((int*)*(int*)b + 4));


    funf(); // Derive::f()
    fung(); // Base::g()
    funh(); // Base::h()
    fung1(); // Derive::g1()
    funh1(); // Derive::h1()

    cout << (Fun)(*((int*)*(int*)b + 5));
    return 0;
}


从上表可以看出：

1．  覆盖的f()函数被放到了虚表中原来父类虚函数的位置。

2．  没有被覆盖的函数依旧。

3．  可通过获取获取成员函数指针来调用成员函数（即使是private类型的），带 来一定安全性的影响。

4． 多继承的情形

#include <iostream>
using namespace std;

class Base1 {
public:
    virtual void f() { cout << "Base1::f()" << endl; }
    virtual void g() { cout << "Base1::g()" << endl; }
    virtual void h() { cout << "Base1::h()" << endl; }
};

class Base2 {
public:
    virtual void f() { cout << "Base2::f()" << endl; }
    virtual void g() { cout << "Base2::g()" << endl; }
    virtual void h() { cout << "Base2::h()" << endl; }
};


class Base3 {
public:
    virtual void f() { cout << "Base3::f()" << endl; }
    virtual void g() { cout << "Base3::g()" << endl; }
    virtual void h() { cout << "Base3::h()" << endl; }
};


class Derive: public Base1,public Base2, public Base3 {
    virtual void f() { cout << "Derive::f()" << endl; }
    virtual void g1() { cout << "Derive::g1()" << endl; }
};

int main()
{
    typedef void (*Fun)();

    Derive d;
    Base1 *b1 = &d;
    Base2 *b2 = &d;
    Base3 *b3 = &d;


    b1->f(); //Derive::f()
    b2->f(); //Derive::f()
    b3->f(); //Derive::f()
    b1->g(); //Base1::g()
    b2->g(); //Base2::g()
    b3->g(); //Base3::g()


    Fun b1fun = (Fun)(*(int*)*(int*)b1);
    Fun b2fun = (Fun)(*(int*)*((int*)b1+1));
    Fun b3fun = (Fun)(*(int*)*((int*)b1+2));

    b1fun(); // Derive::f()
    b2fun(); // Derive::f()
    b3fun(); // Derive::f()

    return 0;
}


从上表可以看出：

1． 每个父类都有自己的虚表。

2． 子类的成员函数被放到了第一个父类的表中。（所谓的第一个父类是按照声明顺序来判断的）

3． 对于多继承无虚函数覆盖的情况，布局与上图类似(Derive的位置对应Base)。