Question

I am reading about FIFO queue array implementeation in Algorithms in C++ by Robert Sedwick.. The contents of the queue are all the elements in the array between “head” and “tail”, taking into account the wraparound back to 0 when the end of the array is encountered. If “head” and “tail” are equal, then we consider the queue to be empty; but if “put” would make then equal, then we consider it to be full. We are making the size of the array 1 greater than the maximum number of elements that the client expects to see in the queue so that we could augument this program to make error checks.

template <class Item>
class QUEUE
  {
    private:
      Item *q; int N, head, tail;
    public:
      QUEUE(int maxN)
        { q = new Item[maxN+1]; 
          N = maxN+1; head = N; tail = 0; }
      int empty() const
        { return head % N == tail; }
      void put(Item item)
        { q[tail++] = item; tail = tail % N; }
      Item get()
        { head = head % N; return q[head++]; }
  };

My question why author as mentioned in text allocating array 1 greater than user specified for making error checks. I am not getting how allocating 1 greater than user request will help us in error checking? Please help me with sample code.

Thanks for your time and help.

Answer 1

I think that, without the extra cell, there would be no way to distinguish between a queue with maxN elements and an empty queue, since head and tail would be in the same equivalence class modulo maxN in both cases.

So, I guess error handling could be done right after a put, to check that the empty criterion is not fulfilled (which means capacity was exceeded):

void put(Item item)
{
  q[tail++] = item;
  tail = tail % N;
  // check that capacity is not exceeded.
  if (head % N == tail)
    throw;
}

and likewise at the beginning of get, to check that the queue is not empty.

Does it make sense?