What is this strange function definition syntax in C?

Question

I've seen a few function definitions like this recently while playing with GNU Bison:

static VALUE
ripper_pos(self)
    VALUE self;
{
    //code here
}

Why is the type of self outside of the parenthesis? Is this valid C?

Answer 1

Yes, it uses an older style of function definition in which the parameters, sans type, are listed in parentheses, followed by the declaration of those variables with their types before the opening brace of the function body. So self is of type VALUE.

Answer 2

Yes it is. As far as I know it's from those old days when the C language wasn't part of any standards.

And there ought to be people that like this syntax.

Fortunately, I'm not one of them. :)

Answer 3

This is the so-called "old" variant of declaring function arguments. In ye olden days, you couldn't just write argument types inside the parentheses, but you had to define it for each argument right after the closing parenthesis.

In other words, it is equivalent to ripper_pos( VALUE self )

Answer 4

That's the old-style C function declaration syntax.

Answer 5

It is really old C code, where you first specify the argument names, and then their types. See for example here.

Answer 6

That are old K&R style function parameter declarations, declaring the types of the parameters separately:

int func(a, b, c)
   int a;
   int b;
   int c;
{
  return a + b + c;
}

This is the same as the more modern way to declare function parameters:

int func(int a, int b, int c)
{
  return a + b + c;
}

The "new style" declarations are basically universally preferred.

Answer 7

This is old c. K&R C used this convention, before ANSI C enforced typed parameters.

static VALUE  // A static function that returns 'VALUE' type.
ripper_pos(self)  // Function 'ripper_pos' takes a parameter named 'self'.
    VALUE self;   // The 'self' parameter is of type 'VALUE'.