Question

How is this statement a definition? Isn’t it supposed to be a declaration only as it does not allocate any memory until we define an object of the type struct date?

struct Date { int d , m , y ; };

I am readng this book called “The C++ programming language” by Bjarne Stroustrup, in which it has been said (in section 4.9) that this a declaration as well as a definition.

Answer 1

It’s not a statement in either language. C99 defines statements in 6.8, and C++11 defines statements in 6.

In C, it is not a definition, it’s a declaration only: 6.7/5 of C99 says:

A definition of an identifier is a declaration for that identifier that:
—for an object, causes storage to be reserved for that object;
—for a function, includes the function body;
—for an enumeration constant or typedef name, is the (only) declaration of the identifier.

Since this is none of those three things, it’s not a definition of an identifier. In the C99 grammar, it’s a struct-or-union-specifier (followed by a semi-colon), which in turn is a type-specifier (followed by a semi-colon), which is one of the permitted forms of a declaration (6.7/1).

In C++, it is a class-specifier or class definition: 9/2 of C++11 says

A class-specifier is commonly referred to as a class definition.

In both C and C++ it’s common to say that “every definition is a declaration”, so that’s probably why Stroustrup say’s it’s a declaration as well as a definition.

In C this is strictly true, because of the definition of “definition” above. In C++ I think it’s not actually true in the grammar that a class-specifier is a declaration, but a class definition introduces a complete type, while a class declaration introduces an incomplete type. There’s nothing you can do with an incomplete type that you can’t also do with the complete type, so the class definition is “as good as” a class declaration like struct Date;, and better.