What is the correct way of declaring an iterator?

April 4, 2022

I am wondering if there is any difference between using

std::set<int,std::greater<int>>::iterator itr;

and

std::set<int>::iterator itr;

I tried the two of them in the code below and the result is the same, I would like know is there is any difference between one and the other or if there is any instance in which I should need to use one over the other (when using the STL library or another case).

CODE:

#include<iostream>
#include<iterator>
#include<set>

int main(){
    std::set<int,std::greater<int>> s1;
    s1.insert(40);
    s1.insert(30);
    s1.insert(60);
    s1.insert(20);
    s1.insert(50);
    s1.insert(10);
    // std::set<int>::iterator itr;
    std::set<int,std::greater<int>>::iterator itr;
    std::cout<<"\nThe set s1 is: \n";
    for(itr=s1.begin();itr!=s1.end();itr++){
        std::cout<<*itr<<" ";
    }
    std::cout<<"\n";
}

>Solution :

std::set<int>::iterator itr; is wrong.

It happens to work on both GCC, Clang, and MSVC by default.
But e.g. if I enable GCC’s iterator debugging (-D_GLIBCXX_DEBUG), it stops compiling.

Normally you don’t need to manually spell the iterator type. You can do for (auto itr = s1.begin(); ...). Or, if the iterator needs to be created uninitialized, decltype(s1)::iterator itr;.