了解`std :: unordered_set`的用法

Question

#include <iostream>
#include <cmath>
#include <unordered_set>
using namespace std;

struct MN
{
    MN(const int& _m, const int& _n)
        : m(_m), n(_n) {
            value = pow (m, n);
        }

    bool operator==(const MN& rhs) const {
        return m == rhs.m && n == rhs.n;
    }

    int value;      
    int m;
    int n;
};

struct MNHash
{
    size_t operator()(const MN& rhs) const {
        return hash<int>()(rhs.m) ^ hash<int>()(rhs.n);
    }
};

int main() {
    unordered_set<MN, MNHash> st;
    st.emplace(2, 3);
    st.emplace(3, 2);

    cout << st.size() << endl; // 2
    return 0;
}

Question 1> Is it true that the unordered_set uses MN::operator== to check whether the new item is duplicate or not?

Question 2> Is it true that the unordered_set uses 'MNHash' to compute the hash key of the new item?

Question 3> If the answers to both Q1 and Q2 are YES, then is it true that we will see a hash conflict in about code because both MN(2, 3) and MN(3, 2) have the same hash code?

Thank you

Answer 1

1. More or less. The default comparator for std::unordered_set will use std::equal_to , which, unless specialized, will do return lhs == rhs .

   You can change this by either specializing std::equal_to , or by adding a third template argument to the std::unordered_set .

2. It will use the second template argument to compute the hash. In this case, you've passed MNHash , so it will use that to do the work.

3. Since the hash value of MN(2,3) and MN(3,2) are the same, they will be placed in the same bucket.