Function Returns

Return-by-address

Suppose we wanted a function to return an array, which we can then use and mutate elsewhere. One way to do so is to write a function with return-by-address — a function that returns the address of stored data:

#include <iostream>
using namespace std;

int* natNumArray(int size) {
	int *p = new int[size];
	for (int i = 0; i < size; i++) {
		p[i] = i+1;
	}
	return p;
}

int main() {
	int *natNumPTR = natNumArray(5);
	return 0;
}

In the code above, we allocated memory in the heap for an int array, using the function natNumArray(). The function returns the resulting array's address, and in main(), that address is assigned to a pointer called *natNumPTR. As a result, both p in natNumArray() and natNumPTR point to the same address, effectively allowing us to modify the newly allocated array.

Return-by-reference

With return-by-address, we use an explicit pointer. We can instead use an implicit pointer with return-by-reference. By writing a function as return-by-reference, we instruct C++ that the particular reference returns a pointer to its return value, rather than the value itself. The end result: We can use functions on the left side of an assignment statement, rather than the typical right side.

#include <iostream>
using namespace std;

double& setValues(double vals[], int i) {
	return vals[i];
}

int main () {
	double arr[] = {1.1, 9.8, 3.1, 2.1, 5.7};
	cout << "Value before change" << endl;
	for ( int i = 0; i < 5; i++ ) {
		cout << "\t arr[" << i << "] = ";
		cout << arr[i] << endl;
	}

	setValues(arr, 1) = 19.2;
	setValues(arr, 3) = 58.8;

	cout << "Value after change" << endl;
	for ( int i = 0; i < 5; i++ ) {
		cout << "\t arr[" << i << "] = ";
		cout << arr[i] << endl;
	}
	return 0;
}

Value before change
	arr[0] = 1.1
	arr[1] = 9.8
	arr[2] = 3.1
	arr[3] = 2.1
	arr[4] = 5.7
Value after change
	arr[0] = 1.1
	arr[1] = 19.2
	arr[2] = 3.1
	arr[3] = 58.8
	arr[4] = 5.7