13/30/98

CSE 143

More Collection ADTs[Sections 4.6-4.8]

23/30/98

More Collection ADTsWe’ll study the following “classic” ADTs

Stack: ordered collection accessible in LIFO mannerQueue: ordered collection accessible in FIFO mannerSet: unordered collection without duplicatesTable: unordered mapping of keys to values (e.g. names -> StudentRecords)

We’ll spend much of the rest of the quarter discussing various implementation techniquesThose techniques can then be used with every more complicated data structures.

33/30/98

Stack ADTTop: Uppermost element of stack,

first to be removedBottom: Lowest element of stack,

last to be removedHeight or Depth: Number of elementsElements are always inserted andremoved from the top (LIFO)

Homogeneous collection

...

top

bottom

aStack:

43/30/98

Stack Operationspush(item): Adds an element to top of stack, increasing stack height by one

pop(): Removes topmost element from stack and returns it, decreasing stack height by one

top(): Returns a copy of topmost element of stack, leaving stack unchanged

No cursorNo iteration via reset(), advance(), etc.

53/30/98

Stack ExampleShow the changes to the stack in the following example:

Stack s;

s.push(5);

s.push(3);

s.push(9); int v1 = s.pop();

int v2 = s.top();

s.push(6);

s.push(4);

63/30/98

Common Uses of Stacks

Implementing function calls Postfix notation

4 7 + 5 * instead of (4 + 7) * 5push operands until operator found, pop 2 values, apply operator, push result

Backtracking, e.g., to explore paths through a maze

73/30/98

Stack Interface

class IntStack {

public:

IntStack();

bool isEmpty(); // is the stack empty?

bool isFull(); // is the stack full?

void push(int item); // add item to top

int pop(); // remove top item

int top(); // show the top item

private:

. . .

}

83/30/98

A Stack Client// Read numbers and print in reverse order

void ReverseNumbers() {IntStack s;int oneNumber;

while ( cin >> oneNumber ) { assert ( !s.IsFull() ); s.push(oneNumber);}

while ( !s.IsEmpty() ) cout << s.pop() << endl;

}

93/30/98

Possible ImplementationsMany possible implementations

Array-basedDynamic memory structure (later)List-based

As implementor, use other ADTs to make job easierDon’t reinvent the wheel for every problemOften simplifies job to reuse pieces when possible

103/30/98

Stack InterfaceWe’ll use a hidden list.

#include “IntList.h”

class IntStack {

public:

Stack();

bool isEmpty(); // is the stack empty?

bool isFull(); // is the stack full?

void push(int item); // add item to top

int pop(); // remove top item

int top(); // show the top item

private:

IntList items;

}

113/30/98

Stack Implementation (2)

IntStack::IntStack() { }

// don’t need to do anything, why?

bool IntStack::isFull() {

return items.isFull();

}

bool IntStack::isEmpty() {

return items.isEmpty();

}

123/30/98

Stack Implementation (3)

void IntStack::push(int item) {

items.insertBefore(item);

}

int IntStack::pop() {

int rval = items.data();

items.deleteItem();

return rval;

}

int IntStack::top() {

return items.data();

}

133/30/98

Queue ADTHomogeneous, ordered collection, accessed only at the front (removal) and rear (insertion)Front: First element in queueRear: Last element of queue FIFO: First In, First Out

front rear

...aQueue:

143/30/98

Queue Operationsenqueue(item) : Adds an element to rear of queuedequeue() : Removes and returns element at the front of queuefront() : Returns a copy of the front element of queueNo cursorNo iteration via reset(), advance(), etc.

153/30/98

Queue ExampleShow the changes to the queue in the following example:

Queue q;

q.enqueue(4);

q.enqueue(7);

q.enqueue(5);

int v1 = q.dequeue();

int v2 = q.front();

q.enqueue(9);

q.enqueue(2);

163/30/98

Common Uses of QueuesManaging a first-come, first-serve line

Processes in an operating systemPrint jobs at the printer

Buffering input/outputWhen printing to screen with cout, characters don’t appear right away: they are buffered and sent out in groups.

Mouse events

SimulationsPeople waiting in line for service

173/30/98

Queue Interface

class IntQueue {

public:

IntQueue();

bool isFull();

bool isEmpty();

void enqueue(int item);

int dequeue();

int front();

private:

IntList items;

// again, we’ll use a hidden list

}

183/30/98

Queue Implementation (1)IntQueue::IntQueue() { }

bool IntQueue::isFull() {

return items.isFull();

}

bool IntQueue::isEmpty() {

return items.isEmpty();

}

void IntQueue::enqueue(int item) {

while (!items.endOfList())

items.advance();

items.insertBefore(item);

}

193/30/98

Queue Implementation (2)

int IntQueue::dequeue() {

items.reset();

int rval = items.data();

items.deleteItem();

return rval;

}

int IntQueue::front() {

items.reset();

return items.data();

}

203/30/98

Set ADTAttributes of set type

Nonlinear collection of elementsVarying number of elementsNo duplicates

Operations on setsElement operations: add, remove, isMemberAggregate operations: union, intersection, difference

Possibly Iteration: reset, advance, end, data

213/30/98

Set TerminologyUnion: Create a new set whose elements are those found in either set

Intersection: Create a new set whose elements are those found in both sets

Difference: Remove elements found in one set from other set

423

97

1 3 10

68

aSet:

223/30/98

Set ExamplesShow the changes to the sets in the following example

Set s1, s2;s1.add(4);s1.add(8);s1.add(14);s2.add(3);s2.add(4);s2.add(5);Set s3 = s1.union(s2);Set s4 = s1.intersection(s2);Set s5 = s1.difference(s2);

233/30/98

Set Interfaceclass IntSet {

IntSet();

// standard size(), isFull(), isEmpty() methods

bool isMember(int item);

void add(int item);

void remove(int item);

Set union(Set other);

Set intersection(Set other);

Set difference(Set other);

// iteration member functions here . . .

private:

IntList items; // again, we’ll use our trusty

} // IntList

243/30/98

Table ADTAttributes of table type

Set of key/value pairsNo duplicate keys

Operations on tableslookup value from keyinsert value at keyremove key and associated value from table

Uses:Phone book, class roster, book index, databases

Sometimes also called a dictionary

253/30/98

Table TerminologyKey: Portion of pair used to look up data, like an index (aka domain value)

Value: Portion of pair that contains data (aka range value)

aTable:“4476542K”“3828122E”“24601JVJ”

23,440

“994802WE”“8675309A”

Key (employee ID) Value (salary)

27,64015,20345,21028,776

263/30/98

Example of Operations

// phone book example:

Table pb;

pb.insert(“Julie”, 5552345);

pb.insert(“Bob”, 3450011);

pb.insert(“Bart”, 6661212);

int bartsNumber = pb.lookup(“Bart”);

pb.delete(“Bob”);

273/30/98

Table Interfaceclass Table {public: Table(); // Create new table

int size(); // Return size of table bool isEmpty(); // Is table empty? bool isFull(); // Is table full?

RangeType lookup(DomainType key); void insert(DomainType key, RangeType value);

RangeType remove(DomainType key);

private: . . .

};

283/30/98

Table ImplementationMany different possibilities

Array of structsList of structsDynamic memory structure

Implementations may make use of future topicsSearching, sortingHash tables

No implementation details at this time...

293/30/98

SummaryStack

List with LIFO structureAccess via push(item), pop(), and top()

QueueList with FIFO structureAccess via enqueue(item), dequeue(), and front()

SetUnordered collection, without duplicates

TableUnordered association of key-value pairsAccess via insert(key, value), delete(key), lookup(key)