Null Pointer

Source: Dev.to

Why use nullptr over NULL or 0

nullptr has the distinct type std::nullptr_t.
Because it is not an integer, overload resolution can differentiate between a null‑pointer argument and an integral argument. This eliminates ambiguous calls that can arise with NULL (typically defined as 0) or a literal 0.

Function overloading example

void func(int);
void func(bool);
void func(void*);

Calls with different null representations

func(0);      // Calls func(int)
func(NULL);   // Also calls func(int) because NULL is usually defined as 0

Both 0 and NULL are integral constants, so the overload taking an int is a better match than the one taking a void*.

Using nullptr to select the pointer overload

func(nullptr);   // Calls func(void*)

Because nullptr is of type std::nullptr_t, it converts only to pointer types, allowing the pointer overload to be chosen unambiguously.

Readability with auto and nullptr

auto result = findRecord(...);

if (result == 0) {
    // …
}

if (result == nullptr) {
    // …
}

The second comparison makes it clear that result is a pointer, improving code readability.

Fun Fact

When you allocate a non‑temporary object on the heap:

int* p = new int;

the pointer returned by new is an r‑value, but it points to an l‑value (the newly created object). Dereferencing a pointer yields an l‑value reference, and the object itself is an l‑value because it has a stable address.