Lecture

Queues

Queues

• The name ‘Queue’ derives from objects that have to queue in order to be dealt with

• A queue is a First-In-First-Out(FIFO) or a Last-In-Last-Out(LILO) abstract data type

Applications of Queues

• Input and output buffers

• Managing batch jobs

• Management of a printer queue

• OS Scheduling

Operation On a Queue: Informal Abstract Definition

• Create: i.e. initialise an empty queue

• Enqueue: add data item to the back of the queue

• Dequeue: remove data item from the front of the queue – if queue is empty throw an exception

• IsEmpty: determine if queue is empty

class QueueInheritance extends StudentList{

public QueueInheritance () { super(“Queue");}public void enqueue (String nameIn, int ageIn) {

insertAtBack(nameIn, ageIn); } public ListNode dequeue() throws EmptyListException {

return removeFromFront(); }public boolean isEmpty () { return super.isEmpty(); }public void printQueue() { printList(); }

} //end of class QueueInheritance

Linked List Queue: Inheritance Example

QueueInheritance theQueue = new QueueInheritance();

// Use the enqueue method

theQueue.enqueue("Alice", 20 ); theQueue.printQueue();

theQueue.enqueue("John", 22); theQueue.printQueue();

theQueue.enqueue("Rita", 21); theQueue.printQueue();

theQueue.enqueue("Bob", 23); theQueue. printQueue();

Linked List Queue: Inheritance Example

24. // Use the dequeue method25. ListNode removedObj = null;26. try27. {28. while ( true )29. {30. removedObj = theQueue.dequeue();31. System.out.println( removedObj.name.toString() + "

dequeued" );32. theQueue.printQueue();33. }34. }35. catch ( EmptyListException e )36. {37. System.err.println( "\n" + e.toString() );38 }

• front and back are the indices

• Avoid overflow – more than what the array can hold

• front = back = 0 is the ‘empty condition’ & prevents underflow – remove item from empty queue

[0] [1] [2] [3] [4]

front = back = 0

q

An empty queue

Array Queue Implementation

• To enqueue (insert) an item we need:

q[back ++] = item;

• To dequeue (delete) an item we would need:

item = q[ front ++];

Array Queue Implementation

• After inserting A, B, C

A B C

[0] [1] [2] [3] [4]

front = 0 back = 3

q

Array Queue Implementation

• After deleting A, B

C

[0] [1] [2] [3] [4]

front = 2 back = 3

q

Array Queue Implementation

• After inserting D

• Any problems if we try to insert E? Overflow!

• Any remedies? Shift the entire queue after every

dequeue operation, but very costly!

C D

[0] [1] [2] [3] [4]

front = 2 back = 4

q

• Is there a better solution? Circular Queue

• Allow array element to be ‘wrapped around’

• So, now we can insert E

• But enqueue and dequeue operation must also be re-defined (see example)

C D E

[0] [1] [2] [3] [4]

back = 0 front = 2

q

• Now consider inserting F, G

• Array is full – any attempt to insert another item causes overflow

• Do you the problem? How do you test for empty queue? i.e. underflow

• back = = front is no longer the condition for testing underflow!

F C D E F EDCG

[0] [1] [2] [3] [4] [0] [1] [2] [3] [4]

back = 1 front = 2 back = front = 2

• Any solutions?

• Allow queue to grow only as large as one less than it maximum size ….

• Queue becomes full when back points to the location immediately before front

F C D E

[0] [1] [2] [3] [4]

back = 1 front = 2

• Any other solution?

Keep a count of the number of items in the queue

Re-defining enqueue and dequeue operation

• Enqueue

q[back] = item;

back = ( back + 1) % n;

• Dequeue

item = q[front];

front = (front + 1) % n;

where n is the size of the array and % is the remainder operator.