Question

I am very new to c and pointers. Each time I thin k I have understood it, there comes a problem that I don’t really understand (I spent some time reading c docs but pointers still remain unclear to me) :

typedef struct {
        int q[QUEUESIZE+1];
        int first;
        int last;
        int count;
} queue;

enqueue(queue *q, int x)
{
  if (q->count >= QUEUESIZE)
        printf("Warning: queue overflow enqueue x=%d\n",x);

  else {
    q->last = (q->last+1) % QUEUESIZE;
    q->q[ q->last ] = x;
    q->count = q->count + 1;
  } 
}

I hope my question will not be too opaque but could someone explain the use of pointer in enqueue function? I thought the principle of queuing was to allocate some precise successive memory addresses, but it is not that for sure….

Answer 1

enqueue takes a queue queue (queue of type queue) and add an element in it (which is made of an integer.

queue *q is a pointer, since, probably

• there could be more than one queue, and the parameter tells what queue we’re talking about
• in order to avoid a global variable, the queue is given as parameter – we want a reference to the queue so that it can be modified, and the modification will stay live even after exiting enqueue

Passing a queue by value, as

enqueue(queue q, int x) { ...

would mean

• a lot of data given as parameter (copy of the queue myqueue to the q parameter)
• when q is modified, the modification is made on q within the enqueue function.
  The initially provided queue (myqueue) as parameter would not be modified

For instance

enqueue(queue q, int x) { 
  q.count++; // only the local q.count is changed, not myqueue.count
  // ...
}

// ...

queue myqueue;
// ...
enqueue (myqueue, 3); // enqueue changes its local parameter, myqueue is not affected

Besides, the enqueue function implementation could be optimized… (see wildplasser answer below who suggests a better queue implementation)