Unofficial extensions to C, structs with functions, does that exist?

[User Name:]

I know, if I want OOP I should go to cpp, but I still prefer C for no real reason. Is there a C compiler that supports defining structure members that are functions? GCC by default, doesn't. I didn't try any other standards.

[Adak]

Turbo C - no functions allowed in a struct.

[cas]

I've never used such a compiler. You can have function pointers in your struct, though, which is a pretty good approximation--you just have to remember to point them somewhere first.

[User Name:]

From Google: Using function pointers,

#include <stdio.h>


struct t {
        int a;
        void (*fun) (int * a);
} ;

void get_a (int * a) {
        printf (" input : ");
        scanf ("%d", a);
}

int main () {
        struct t test;
        test.a = 0;

        printf ("a (before): %d\n", test.a);
        test.fun = get_a;
        test.fun(&test.a);
        printf ("a (after ): %d\n", test.a);

        return 0;
}

It even validates the params you pass. Is there any way to set the initial value of t.fun to get_a?

[Elysia]

Create an initialization function that creates a t and returns it.

I know, if I want OOP I should go to cpp, but I still prefer C for no real reason. Is there a C compiler that supports defining structure members that are functions? GCC by default, doesn't. I didn't try any other standards.

Obviously not. OOP is not the point of C. You're not going to see it anytime soon.

[laserlight]

Is there any way to set the initial value of t.fun to get_a?

There's the normal way of initialising a struct object, e.g.,

struct t test = {0, get_a};

[Elysia]

struct t test = {0, get_a};

o_O
Because for some reason that little "feature" is still in the C standard, I tend to downplay it. It's evil. Don't use it. So to speak.
Another way to do it is (and IMHO, the proper way):

struct t test = {0, &get_a};

Nevertheless, I would say it is preferable to do it in a separate function. An initializing function. The more function pointers you put, the easier it is to simply call the function instead of duplicating code.

[Bayint Naung]

In case if you are interested, there's GObject.

[Bayint Naung]

You don't need & in front of function name. It's optional.
Using C99,

struct t test = { .fun = get_a };  // If I'm not wrong,other members init to proper zero according to C99

[Elysia]

Yes, it's optional, and no, I don't recommend doing it because it's wrong. Some old compatibility crap.
To take the address of something, you use the address-of operator. No exceptions.

[laserlight]

Because for some reason that little "feature" is still in the C standard, I tend to downplay it. It's evil. Don't use it. So to speak.
Another way to do it is (and IMHO, the proper way):

Likewise, when you want to call a function through a function pointer f, remember to write (*f)(args), not f(args). The latter is evil. Don't use it. So to speak.

EDIT:

To take the address of something, you use the address-of operator. No exceptions.

Likewise, when you want to pass the address of the first element of an array x in order to simulate passing the array by reference, do not write foo(x). Rather, always write foo(&x[0]). Remember boys and girls, no exceptions. Allowing an array to be converted to a pointer to its first element is evil.

[Elysia]

Hmmm. you have a point.
(*f)(args) vs f(args): is one a compatibility syntax or is it simply two different syntax ways of doing things (kind of like (*p).me vs p->me)?

[Bayint Naung]

100% agree with laserlight.

[claudiu]

Yes, purism is evil mmkay?

[Elysia]

Extensions or backwards compatibility constructs are evil O_o And multiple syntaxes for the same thing is evil, too.