- use operators for user-defined data types (classes), it is actually function overloading.
- Cannot change the usage;
- Cannot define new operators by yourself;
- Cannot change precedence rules;
- Cannot change the number of operands.
3. Assume @ is an operator overloaded as member function, please write down the declaration of the functions corresponding to the following statements.
a) x @ y; b) @x; c) y@;
- x @ y; return-type opeator@(object, object);
- @x; return-type operator@(object);
- y@; return-type operator@(object, int);
4. Assume @ is an operator overloaded as friend function, please write down the declarations of functions corresponding to the following statements.
a) x @ y; b) @x; c) y@;
- friend abcd operator @ (const abcd &x,const abcd &y);
- friend abcd operator @ (const abcd &x);
- friend abcd operator @ (const abcd &x,int);
5. When are you supposed to overload operator as member functions? And when are you supposed to overload operator as friend functions? Please try to explain your answer in detail.
- 当双目运算符的左元素是一个系统定义好的类型，不方便重载，而又需要在函数中访问右元素的私有变量，就要用friend。
- 当双目运算符的左元素是自定义类型，就可以将重载运算符作为成员函数，这时只需要一个参数就是右元素。当然也可以将其作为全局函数,参数改为两个，不需要加friend
- Because we want to use operator “<<” as “cout << obj;”, the operator occurs before object, if we overloaded it as member functions, we can only use it as “obj << obj2”, which is not expected.
7. Please write down the declaration of the function that is called for the following statement. (Note that vec3 is a well-defined object.)
Vector3 vec3_ = vec3;
- Vector3& operator = (const complex& vec_);
8. Please write down a declaration of stream operator “<<” for Vector3, making the following statements legal.
a) std::cout << vec3 << std::endl;
b) std::cout << std::left << std::setw (5) << vec3 << std::endl;
- friend ostream & operator << (ostream &os, const Vector3 & vec3);
9. Why is bit-copy dangerous, especially when there are pointer members in the class? Please give an example where bit-copy is adopted by the compiler, which potentially causes bugs.
因为默认拷贝构造函数是按位拷贝的，所以如果类中有指针变量的时候，就很可能会拷贝出一个一模一样的指针，此时如果delete掉一个类，会影响到另一个类的值。
/* example */
#include <iostream>
using namespace std;
class A
{
private:
string* m_s;
public:
A() {
m_s = NULL;
}
A(const string &s) {
m_s = new string;
m_s = s;
}
~A() {
delete m_s;
}
};
int main()
{
A a(“hello, world.”);
A b;
b = a;
return 0;
}
when a and b are destructed, the memory area m_s point will be released for twice.
- If there is pointer in the class, the assignment will first delete the original pointer to free the memory. But in self-assignment, we cannot delete it and it may cause data lost and memory leak.
11. Please state the difference between operator postfix ++ (--) and prefix ++ (--). Please write down the function body of operator postfix ++ and prefix ++ for the following class.
class Integer
{
public:
Integer (int x = 0): _x (x) {}
private:
int _x;
};
-
Using postfix, the return value is the original one while using prefix the return value is the one after operation.
const Integer& operator++ () { // Prefix _x++; return *this; } const Integer operator++ (int) { // Postfix Integer before (_x); _x++; return before; }
12. What is a function object? What is the difference between function objects and normal functions?
- An object of a class with overloaded function call operator “()”.
- Function object can store and pass data. An function object can correspond to multiple functions.
- Function objects can store and pass values.
- Normal functions can only calculate certain problems.
- Function object can store some member variables which is easy to change through interface. But function pointer can only point to a fixed function. In addition, function object can have inheritance, composition and so on, which make it easy to be reused or adapted. So function object is more valuable.
- Function objects can be safer and enables type checking. It can also utilize template to aid more powerful code reuse.
- Function objects can create lots of different kinds of objects according to the parameter passed, while function pointer can only point to existing function.
15. Please complete the missing code in the following program such that we can obtain the sum of the integers in arr.
#include <iostream>
#include <array>
#include <algorithm>
#include <random>
const size_t size = 1 << 5;
std::ostream& operator << (std::ostream& os, const Sum& sum)
{
os << "sum:" << sum._x << std::endl;
return os;
}
int main ()
{
std::random_device rd;
std::mt19937 mt (rd ());
std::uniform_int_distribution<> dis (0, size);
std::array<int, size> arr;
for (auto& element: arr)
element = dis (mt);
Sum s (0);
s = std::for_each (std::begin (arr), std::end (arr), s);
std::cout << s;
return 0;
}
#include <iostream>
#include <array>
#include <algorithm>
#include <random>
const size_t size = 1 << 5;
class Sum {
private:
int _x;
public:
Sum(int s = 0): _x(s) {}
void operator()(int x) {
this->_x += x;
}
friend std::ostream& operator << (std::ostream&, const Sum&);
};
std::ostream& operator << (std::ostream& os, const Sum& sum)
{
os << "sum:" << sum._x << std::endl;
return os;
}
int main ()
{
std::random_device rd;
std::mt19937 mt (rd ());
std::uniform_int_distribution<> dis (0, size);
std::array<int, size> arr;
for (auto& element: arr)
element = dis (mt);
Sum s (0);
s = std::for_each (std::begin (arr), std::end (arr), s);
std::cout << s;
return 0;
}