Understanding a Line of Quake 2 Source Code

Hi,

I have a question about a line in the following code, which is part of the released source code for Quake 2. I'm trying to understand how it works. As I understand it, this is a declaration of the gitem_t struct, which describes objects in the game:

Code:

---

typedef struct gitem_s
{
        void                (*drop)(struct edict_s *ent, struct gitem_s *item);
        void                (*weaponthink)(struct edict_s *ent);
        char                *pickup_sound;
        char                *world_model;
        int                        world_model_flags;
        char                *view_model;
} gitem_t;

---

Later in the code, the following is part of the code used to test if an item can be dropped:

Code:

---

        gitem_t                *it;
        if (!it->drop)

---

What I'm wondering about is this line:

Code:

---

void        (*drop)(struct edict_s *ent, struct gitem_s *item);

---

This doesn't look like any variable/ function declaration I've come across/ used before. It's not a function declaration because if I right click it in VS Express, and click goto definition it just stays on the same line. So presumably it's a declaration of a pointer to type "struct edict_s *ent", or "struct gitem_s *item". Does this sound right?

Sorry if this sounds dumb, it just seems a far cry from anything I've seen in my C++ book. Thanks.

This just screams C, but anyway:
It's a function pointer that points to a function that whose prototype looks like this:
void (edict_s*, gitem_s*)
The
if (!it->drop)
line checks if it's a NON-valid pointer (ie NULL pointer).

drop is a (pointer) variable that POINTS to a function with the prototype:

Code:

---

void somename(struct edict_s *ent, struct gitem_s *item);

---

As it is a pointer it can be NULL. The

Code:

---

if(!item->drop)

---

Does indeed check if the pointer is NULL or, in other words, that it doesn't point to a valid function. Apparently NULL is used here if an object can't be dropped. It could've been a pointer to a special function (drop_doNothing) as well, to name an example, but setting it to NULL is probably better and more readable.

Quote:

---

Originally Posted by EVOEx

but setting it to NULL is probably better and more readable.

---

I'd rather use item states, for obvious reasons.

Quote:

---

This just screams C, but anyway:

---

Yep, I believe its QuakeC .

Quote:

---

Originally Posted by Spidey

Yep, I believe its QuakeC .

---

I doubt it. From memory, QuakeC did (does?) not support struct types or typedefs.

Quote:

---

Originally Posted by Dae

I'd rather use item states, for obvious reasons.

---

States is the wrong word, as it is not dependent on any state, it's dependent on whether an item can be dropped (or so I assume). True, we could use a droppable flag. But personally I find this fairly readable and expect this to give less trouble than such a flag. So I say it's just a matter of style; just that my style would match this code should it be mine.
(That is, if I'd code it in C. In C++ it would be classes with virtual functions)