What is the type of a builtin datatype in C and C++?

Question

When we write int a;, it doesn't mean that we are creating an object of class int.

1. What does it mean?
2. What is the type of the datatype int in C and C++?
3. Which header file shows what it is?

Answer 1

When we write int a;, it doesn't mean that we are creating an object of class int.

int a; does indeed create an object in C++. It's an object of type int with indeterminate value if it has automatic storage duration; or with value 0, if it has static storage duration. But there is no "class int" because int is not a class type.

int is a:

• integral type;
• signed type;
• arithmetic type;
• fundamental type;
• scalar type;
• standard layout type;
• trivially copyable type;
• POD type;
• trivial type;

Seems like you got a bit confused in your previous question :)

In int x = 12;, you are creating an object of type int that is named x and has value 12.

The idea of object in C++ is not the same as in most other languages, and most certainly is not the same as is commonly used in object-oriented programming circles. An object in C++ is a region of storage.

If something has a type, it's either an object, a reference, or a function.

Which header file shows what it is?

The language simply requires that the type int has to exist and have certain characteristics (like being integral and having a sign). All compilers I know of simply treat all the builtin types specially and that's why you won't find a definition for them in the standard library headers. In fact, they can't provide a definition for them in any header using C++, because the language doesn't provide any means of defining fundamental types. They could only either:

• define it as a compound type (which would be wrong); or
• define it using compiler-specific extensions.

The builtin types are effectively magic.