The '.' operator on null nullable

Question

How is it that you can access null nullable's propery HasValue?
I looked to the compiled code, and it's not a syntactic sugar.

Why this doesn't throw NullReferenceException:

int? x = null;
if (x.HasValue)
{...}

Answer 1

That's because int? is short for Nullable<int> which is a value type, not a reference type - so you will never get a NullReferenceException.

The Nullable<T> struct looks something like this:

public struct Nullable<T> where T : struct
{
    private readonly T value;
    private readonly bool hasValue;
    //..
}

When you assign null there is some magic happening with support by the compiler (which knows about Nullable<T> and treats them special in this way) which just sets the hasValue field to false for this instance - which is then returned by the HasValue property.

Answer 2

Like BrokenGlass said, an int? is actually a Nullable<T>.

Structures always contain a value. Usually you cannot set a structure variable to null, but in this special case you can, essentially setting it to default(Nullable<T>). This sets its contents to null rather than the variable itself.

When you set a Nullable<T> to a value, it uses an implicit operator to set Value = value to the new value and HasValue = true.

When you set Nullable<T> to null, it nulls all of the structure's fields. For a bool field such as HasValue, null == false.

Since a Nullable<T> variable is a structure, the variable can always be referenced because its contents is null rather than the variable itself.

There's more information on structures in the Remarks section of the MSDN page struct.