Using [[nodiscard]] with custom types

The [[nodiscard]] attribute can be applied to functions and custom types to encourage users to handle the return value. When a [[nodiscard]] function or object is called and its return value is discarded, the compiler will generate a warning.

To use [[nodiscard]] with your custom types, you can apply the attribute to the type declaration itself or to specific member functions that should not have their return values discarded.

Here's an example of using [[nodiscard]] with a custom type:

1#include <iostream>
2
3class [[nodiscard]] Result {  
4 public:
5  bool success;
6  int value;
7
8  Result(bool s, int v) : success(s), value(v) {}
9};
10
11Result divide(int a, int b) {
12  if (b == 0) {
13    return {false, 0};
14  }
15  return {true, a / b};
16}
17
18int main() {
19  // Warning: ignoring return value
20  divide(10, 2); 
21
22  Result res = divide(10, 0);
23  if (!res.success) {
24    std::cout << "Division failed\n";
25  } else {
26    std::cout << "Result: " << res.value << '\n';
27  }
28}

1Division failed

1Warning: discarding return value of function with [[nodiscard]] attribute

In this example, the Result class is marked with [[nodiscard]], indicating that instances of this type should not have their return values discarded. The divide function returns a Result object, and if the return value is not handled (as in the first call to divide in main), the compiler will generate a warning.

You can also apply [[nodiscard]] to specific member functions instead of the entire type:

1class Result {
2public:
3  [[nodiscard]] bool success() const {  
4    return m_success; }
5
6private:
7  bool m_success;
8};

By using [[nodiscard]] with your custom types, you can enforce better code practices and encourage users to handle important return values explicitly.