Object Oriented Programming (CS304)

VU

CS304-Handouts

Last Updated: 20/08/2009

Virtual University of Pakistan

Object Oriented Programming (CS304)

VU

LECTURE NO.01...................................................................................................................................8
01.1.
01.2.
01.3.
01.4.
01.5.
01.6.
01.7.

INTRODUCTION....................................................................................................................8
WHAT IS A MODEL?...........................................................................................................10
OO MODELS: .....................................................................................................................11
OBJECT-ORIENTATION - ADVANTAGES...........................................................................12
WHAT IS AN OBJECT? ........................................................................................................12
TANGIBLE AND INTANGIBLE OBJECTS .............................................................................12
SUMMARY:..........................................................................................................................13

LECTURE NO.02.................................................................................................................................15
02.1.
02.2.
02.3.
02.4.
02.5.
02.6.
02.7.

INFORMATION HIDING: ....................................................................................................15

ENCAPSULATION ................................................................................................................16
INTERFACE ..........................................................................................................................17
IMPLEMENTATION .............................................................................................................18
SEPARATION OF INTERFACE & IMPLEMENTATION .........................................................19
MESSAGES ..........................................................................................................................19
SUMMARY ...........................................................................................................................19

LECTURE NO.03.................................................................................................................................21
03.1.
03.2.
03.3.

ABSTRACTION ....................................................................................................................21
CLASSES ..............................................................................................................................23
INHERITANCE .....................................................................................................................24

LECTURE NO.04.................................................................................................................................29
04.1.
04.2.
04.3.
04.4.
04.5.
04.6.
04.7.

CONCEPTS RELATED WITH INHERITANCE .......................................................................29

GENERALIZATION ..............................................................................................................29
SUB-TYPING (EXTENSION).................................................................................................31
SPECIALIZATION (RESTRICTION) .....................................................................................32
OVERRIDING ......................................................................................................................34
ABSTRACT CLASSES ...........................................................................................................36
CONCRETE CLASSES ..........................................................................................................38

LECTURE NO.05.................................................................................................................................40
05.3.
05.4.
05.5.

SIMPLE ASSOCIATION .......................................................................................................49

COMPOSITION ....................................................................................................................52
AGGREGATION...................................................................................................................53

LECTURE NO.06.................................................................................................................................55
06.1.
06.2.
06.3.
06.4.
06.5.

CLASS COMPATIBILITY ......................................................................................................55

POLYMORPHISM .................................................................................................................56
POLYMORPHISM IN OO MODEL .......................................................................................56
POLYMORPHISM ADVANTAGES .....................................................................................57
OBJECT-ORIENTED MODELING AN EXAMPLE...................................................................58

LECTURE NO.07.................................................................................................................................65
07.1.
07.2.
07.3.
07.4.
07.5.
07.6.
07.7.

CLASS ..................................................................................................................................65
TYPE IN C++ .......................................................................................................................65
ABSTRACTION ....................................................................................................................66
DEFINING A NEW USER DEFINED TYPE ...........................................................................66
OBJECT AND CLASS: ..........................................................................................................68
ACCESSING MEMBERS........................................................................................................68
ACCESS SPECIFIERS ............................................................................................................69

LECTURE NO.08.................................................................................................................................71
08.1.

MEMBER FUNCTIONS ........................................................................................................71

Virtual University of Pakistan

Object Oriented Programming (CS304)

08.2.
08.3.
08.4.
08.5.
08.6.
08.7.
08.8.
08.9.
08.10.
08.11.

VU

DEFINING MEMBER FUNCTIONS ...................................................................................... 71

INLINE FUNCTIONS ............................................................................................................ 72
CONSTRUCTOR .................................................................................................................. 74
CONSTRUCTOR PROPERTIES ............................................................................................. 74
DEFAULT CONSTRUCTOR ................................................................................................. 75
CONSTRUCTOR OVERLOADING ....................................................................................... 75
CONSTRUCTOR OVERLOADING ....................................................................................... 77
COPY CONSTRUCTOR ........................................................................................................ 78
SHALLOW COPY ............................................................................................................ 79
DEEP COPY .................................................................................................................... 81

LECTURE NO.09 ................................................................................................................................ 83

09.1.
09.2.
09.3.
09.4.
09.5.
09.6.

SHALLOW COPY ................................................................................................................. 84

DEEP COPY ......................................................................................................................... 89
IMPORTANT POINTS ABOUT COPY CONSTRUCTOR: ........................................................ 91
DESTRUCTOR ..................................................................................................................... 92
ACCESSOR FUNCTIONS ..................................................................................................... 93
THIS POINTER ..................................................................................................................... 94

LECTURE NO.10 ................................................................................................................................ 97

10.1.
10.2.
10.3.
10.4.
10.5.

USES OF THIS POINTER ...................................................................................................... 97

SEPARATION OF INTERFACE AND IMPLEMENTATION ..................................................... 97
COMPLEX NUMBER ............................................................................................................ 97
CONST MEMBER FUNCTIONS .......................................................................................... 100
THIS POINTER AND CONST MEMBER FUNCTION .......................................................... 102

LECTURE NO.11 .............................................................................................................................. 103

11.1.
11.2.
11.3.
11.4.
11.5.

USAGE EXAMPLE OF CONSTANT MEMBER FUNCTIONS ................................................. 103

DIFFERENCE BETWEEN INITIALIZATION AND ASSIGNMENT:....................................... 104
MEMBER INITIALIZER LIST ............................................................................................. 104
CONST OBJECTS ............................................................................................................... 105
STATIC VARIABLES .......................................................................................................... 107

LECTURE NO.12 .............................................................................................................................. 110

12.1.
12.2.
12.3.
12.4.
12.5.
12.6.

ACCESSING STATIC DATA MEMBER .............................................................................. 111

LIFE OF STATIC DATA MEMBER ..................................................................................... 111
STATIC MEMBER FUNCTION ........................................................................................... 113
THIS POINTER AND STATIC MEMBER FUNCTIONS ......................................................... 114
GLOBAL VARIABLE VS. STATIC MEMBERS .................................................................... 114
ARRAY OF OBJECTS ......................................................................................................... 114

LECTURE NO.13 .............................................................................................................................. 116

13.1.
13.2.
13.3.

POINTER TO OBJECTS ...................................................................................................... 116

BREAKUP OF NEW OPERATION ....................................................................................... 117
CASE STUDY ..................................................................................................................... 117

LECTURE NO.14 .............................................................................................................................. 120

14.1.

COMPOSITION.................................................................................................................. 120

LECTURE NO.15 .............................................................................................................................. 128

15.1.
15.2.

AGGREGATION ................................................................................................................ 132

FRIEND FUNCTIONS ......................................................................................................... 135

LECTURE NO.16 .............................................................................................................................. 138

16.1.

OPERATOR OVERLOADING ............................................................................................. 138

LECTURE NO.17 .............................................................................................................................. 145

Virtual University of Pakistan

Object Oriented Programming (CS304)

17.1.

VU

OVERLOADING ASSIGNMENT OPERATOR .....................................................................147

LECTURE NO.18...............................................................................................................................151
18.1.
18.2.
18.3.

SELF ASSIGNMENT PROBLEM: .........................................................................................151

OTHER BINARY OPERATORS ...........................................................................................152
FRIEND FUNCTIONS AND OPERATOR OVERLOADING ..................................................153

LECTURE NO.19...............................................................................................................................154
19.1.
19.2.
19.3.
19.4.
19.5.

STREAM INSERTION OPERATOR......................................................................................154

STREAM EXTRACTION OPERATOR ..................................................................................154
OVERLOADING STREAM INSERTION OPERATOR ..........................................................155
OVERLOADING STREAM EXTRACTION OPERATOR:......................................................156
OTHER BINARY OPERATORS: ..........................................................................................157

LECTURE NO.20...............................................................................................................................159
20.1.
20.2.
20.3.
20.4.
20.5.

SUBSCRIPT [] OPERATOR .................................................................................................160

OVERLOADING SUBSCRIPT [] OPERATOR ......................................................................160
OVERLOADING FUNCTION () OPERATOR .......................................................................161
FUNCTION OPERATOR PERFORMING SUB STRING OPERATION,..................................161
UNARY OPERATORS.........................................................................................................162

LECTURE NO.21...............................................................................................................................164
21.1.
21.2.
21.3.
21.4.
21.5.

BEHAVIOR OF ++ AND -- FOR PRE-DEFINED TYPES: .......................................................164

POST-INCREMENT OPERATOR: ........................................................................................165
TYPE CONVERSION ..........................................................................................................166
USER DEFINED TYPES:......................................................................................................169
DRAWBACKS OF TYPE CONVERSION OPERATOR: .........................................................170

LECTURE NO.22...............................................................................................................................171
22.1.
22.2.
22.3.
22.4.
22.5.

PRACTICAL IMPLEMENTATION OF INHERITANCE IN C++ .............................................171

INHERITANCE IN CLASSES...............................................................................................171
UML NOTATION .............................................................................................................171
INHERITANCE IN C++ ......................................................................................................171
IS A RELATIONSHIP .....................................................................................................172

LECTURE NO.23...............................................................................................................................179
23.1.
23.2.
23.3.
23.4.

ACCESSING BASE CLASS MEMBER FUNCTIONS IN DERIVED CLASS: .............................179

PROTECTED ACCESS SPECIFIER: ...................................................................................181
IS A RELATIONSHIP .....................................................................................................182
STATIC TYPE .....................................................................................................................185

LECTURE NO.24...............................................................................................................................187
24.1.

MODIFIED DEFAULT CONSTRUCTOR .............................................................................195

LECTURE NO.25...............................................................................................................................203
25.1.
25.2.

OVERLOADING VS. OVERRIDING ...................................................................................203

HIERARCHY OF INHERITANCE.........................................................................................207

LECTURE NO.26...............................................................................................................................209
26.1.
26.2.
26.3.

BASE INITIALIZATION .......................................................................................................209

TYPES OF INHERITANCE...................................................................................................210
PRIVATE INHERITANCE ...................................................................................................212

LECTURE NO.27...............................................................................................................................214
27.1.
27.2.
27.3.

SPECIALIZATION (RESTRICTION) ...................................................................................214

PROTECTED INHERITANCE ..............................................................................................218
PROPERTIES OF PROTECTED INHERITANCE ...................................................................219

Virtual University of Pakistan

Object Oriented Programming (CS304)

VU

LECTURE NO.28 .............................................................................................................................. 221

28.1.
28.2.
28.3.
28.4.

VIRTUAL FUNCTIONS ...................................................................................................... 221

VIRTUAL FUNCTIONS: ..................................................................................................... 226
SHAPE HIERARCHY .......................................................................................................... 226
STATIC VS DYNAMIC BINDING ...................................................................................... 227

LECTURE NO.29 .............................................................................................................................. 229

29.1.
29.2.
29.3.
29.4.
29.5.
29.6.
29.7.
29.8.
29.9.

ABSTRACT CLASSES ........................................................................................................ 229

CONCRETE CLASSES ........................................................................................................ 230
ABSTRACT CLASSES IN C++ ........................................................................................... 230
PURE VIRTUAL FUNCTIONS ............................................................................................ 230
SHAPE HIERARCHY .......................................................................................................... 230
VIRTUAL DESTRUCTORS ................................................................................................. 232
VIRTUAL FUNCTIONS USAGE ...................................................................................... 235
V TABLE ............................................................................................................................ 236
DYNAMIC DISPATCH (DYNAMIC BINDING) ................................................................. 239

LECTURE NO.30 .............................................................................................................................. 240

30.1.
30.2.

POLYMORPHISM CASE STUDY: A SIMPLE PAYROLL APPLICATION .......................... 240

SHAPE HIERARCHY REVISITED:...................................................................................... 243

LECTURE NO.31 .............................................................................................................................. 247

31.1.
31.2.
31.3.

MULTIPLE INHERITANCE................................................................................................. 247

PROBLEMS IN MULTIPLE INHERITANCE......................................................................... 248
VIRTUAL INHERITANCE .................................................................................................. 253

LECTURE NO.32 .............................................................................................................................. 255

32.1.
32.2.
32.3.

GENERIC PROGRAMMING............................................................................................... 256

TEMPLATES....................................................................................................................... 256
FUNCTION TEMPLATES ................................................................................................... 257

LECTURE NO.33 .............................................................................................................................. 261

33.1.
33.2.
33.3.
33.4.
33.5.
33.6.
33.7.
33.8.

MULTIPLE TYPE ARGUMENTS......................................................................................... 261

USER-DEFINED TYPES ..................................................................................................... 261
OVERLOADING VS. TEMPLATES ..................................................................................... 262
TEMPLATE ARGUMENTS AS POLICY:.............................................................................. 263
FIRST SOLUTION:............................................................................................................. 263
SECOND SOLUTION: ........................................................................................................ 264
THIRD SOLUTION ............................................................................................................ 264
DEFAULT POLICY ............................................................................................................. 265

LECTURE NO.34 .............................................................................................................................. 267

34.1.
34.2.
34.3.

GENERIC ALGORITHMS .................................................................................................. 267

CLASS TEMPLATES........................................................................................................... 269
EXAMPLE CLASS TEMPLATE ......................................................................................... 270

LECTURE NO.35 .............................................................................................................................. 273

35.1.
35.2.

MEMBER TEMPLATES: ..................................................................................................... 273

CLASS TEMPLATE SPECIALIZATION ............................................................................... 275

LECTURE NO.36 .............................................................................................................................. 279

36.1.
36.2.
36.3.
36.4.
36.5.
36.6.

MEMBER TEMPLATES REVISITED ................................................................................... 279

PARTIAL SPECIALIZATION: ............................................................................................. 280
FUNCTION TEMPLATES ................................................................................................... 281
COMPLETE SPECIALIZATION........................................................................................... 282
USING DIFFERENT SPECIALIZATIONS ............................................................................ 283
NON-TYPE PARAMETERS ................................................................................................. 283

Virtual University of Pakistan

Object Oriented Programming (CS304)

36.7.
36.8.
36.9.

VU

EXAMPLE TEMPLATE CLASS ARRAY .............................................................................283

DEFAULT NON-TYPE PARAMETERS ................................................................................284
DEFAULT TYPE PARAMETERS..........................................................................................285

LECTURE NO.37...............................................................................................................................285
37.1.
37.2.
37.3.
37.4.
37.5.

RESOLUTION ORDER .......................................................................................................285

FUNCTION TEMPLATE OVERLOADING ...........................................................................287
RESOLUTION ORDER .......................................................................................................287
TEMPLATES AND INHERITANCE ......................................................................................287
DERIVATIONS IN CASE OF A GENERAL TEMPLATE CLASS ............................................288

LECTURE NO.38...............................................................................................................................292
38.1.
38.2.
38.3.
38.4.
38.5.

TEMPLATES AND FRIENDS...............................................................................................292

TEMPLATES AND FRIENDS RULE 1 ...............................................................................292
TEMPLATES AND FRIENDS RULE 2 ...............................................................................293
TEMPLATES AND FRIENDS RULE 3 ...............................................................................295
TEMPLATES AND FRIENDS RULE 4 ...............................................................................296

LECTURE NO.39...............................................................................................................................299
39.1.
39.2.
39.3.
39.4.
39.5.
39.6.

TEMPLATES & STATIC MEMBERS ...................................................................................299

TEMPLATES CONCLUSION ............................................................................................300
GENERIC ALGORITHMS REVISITED................................................................................301
GENERIC ALGORITHMS REVISITED................................................................................301
GENERIC ALGORITHM .....................................................................................................303
PROBLEMS ........................................................................................................................304

LECTURE NO.40...............................................................................................................................305
40.1.
40.2.

CURSORS ..........................................................................................................................305
ITERATORS........................................................................................................................308

LECTURE NO.41...............................................................................................................................312
41.1.
41.2.
41.3.
41.4.
41.5.

STANDARD TEMPLATE LIBRARY:....................................................................................312

STL CONTAINERS ............................................................................................................312
COMMON FUNCTIONS FOR ALL CONTAINERS ..............................................................317
FUNCTIONS FOR FIRST-CLASS CONTAINERS .................................................................318
CONTAINER REQUIREMENTS ..........................................................................................319

LECTURE NO.42...............................................................................................................................320
42.1.
42.2.
42.3.
42.4.
42.5.
42.6.
42.7.
42.8.
42.9.

ITERATORS........................................................................................................................320
ITERATOR CATEGORIES...................................................................................................320
ITERATOR SUMMARY:......................................................................................................320
CONTAINER AND ITERATOR TYPES: ...............................................................................321
SEQUENCE CONTAINERS .................................................................................................321
ASSOCIATIVE CONTAINERS ............................................................................................322
CONTAINER ADAPTERS ...................................................................................................322
ITERATOR OPERATIONS ..................................................................................................322
ALGORITHMS ...................................................................................................................326

LECTURE NO.43...............................................................................................................................329
43.1.
43.2.
43.3.
43.4.
43.5.

EXAMPLE ABNORMAL TERMINATION .........................................................................329

GRACEFUL TERMINATION...............................................................................................330
ERROR HANDLING ...........................................................................................................330
EXCEPTION HANDLING ...................................................................................................333
EXCEPTION HANDLING PROCESS ...................................................................................333

LECTURE NO.44...............................................................................................................................336
44.1.

STACK UNWINDING ........................................................................................................336

Virtual University of Pakistan

Object Oriented Programming (CS304)

VU

LECTURE NO.45 .............................................................................................................................. 343

45.1.

RESOURCE MANAGEMENT ............................................................................................. 343

Virtual University of Pakistan

Object Oriented Programming (CS304)

VU

Lecture No.01
01.1. Introduction
Course Objective:
Objective of this course is to make students familiar with the concepts of
object oriented programming. These concepts will be reinforced by their
implementation in C++.
Course Contents:
The main topics that we will study in the 45 lectures of this course are given
below,
Object Orientation
Objects and Classes
Overloading
Inheritance
Polymorphism
Generic Programming
Exception Handling
Introduction to Design Patterns
Recommended Text Book:
C++ How to Program ( Deitel & Deitel )
Reference Books:
1. Object-Oriented Software Engineering
By Jacobson, Christerson, Jonsson, Overgaard
(For object oriented programming introductory concepts)
2. The C++ Programming Language
By Bjarne Stroustrup
(For better c++ understanding)

Virtual University of Pakistan

Object Oriented Programming (CS304)

VU

Object-Orientation (OO)
What is Object-Orientation?
It is a technique in which we visualize our programming problems in the form of
objects and their interactions as happens in real life.
Examples:
We have different objects around us in our real life that interact with each other to
perform different operations for example,

A Person

A House

A Tree

A Car
Different Objects

These objects interact with each other to perform different operations,

Lives in

Ali

House

Drives

Car

Tree

Take another example of a School; the objects in a school are student, teacher, books,
pen ,school bag, classroom, parents, playground and so on ,

Objects in a School

Virtual University of Pakistan

Object Oriented Programming (CS304)

VU

Teacher

Student

School Bag

Book

Pen

Playground

Parents
Classroom
Library
Suppose we want to develop a fee collection system for a school for this we will need
to find out related objects and their interactions as happens in real life.
In this way we can say that object orientation makes it easier for us to solve our real
world problems by thinking solution of the problem in terms of real world objects.
So we can say that in our daily life everything can be taken as an object that behaves in a
certain way and has certain attributes.
In object orientation we move our concentration to objects in contrast to procedural
paradigm in which we simply write our code in functions and call them in our main
program.
01.2. What is a Model?
A model is an abstraction of something real or conceptual.
We need models to understand an aspect of reality.
Model Examples

10

Virtual University of Pakistan

Object Oriented Programming (CS304)

VU

Highway maps
Architectural models
Mechanical models
01.3. OO Models:
In the context of programming models are used to understand the problem before
starting developing it.
We make Object Oriented models showing several interacting objects to understand
a system given to us for implementation.
Example 1 Object Oriented Model

lives-in
Ali

House

drives

Car

Objects
Ali, Car, House, Tree

Tree

Interactions
Ali lives in the house
Ali drives the car

Example 2 Object Oriented Model (A School Model)

Virtual University of Pakistan

11

Object Oriented Programming (CS304)

VU

A School Model
Teacher

Teaches

Student

Plays-in

Playground

Has

Book

Objects
Teacher, Student, School Bag, Pen,
Book Playground

Pen

School Bag

Interactions
Teacher teaches Student.
Student has School Bag, Book and Pen

01.4. Object-Orientation - Advantages

As Object Oriented Models map directly to reality as we have seen in examples
above therefore,
We can easily develop an object oriented model for a problem.
Everyone can easily understand an object oriented model.
We can easily implement an object oriented model for a problem using any object
oriented language like c++ using its features1 like classes, inheritance, virtual
functions and so on
01.5. What is an Object?
An object is,
1. Something tangible (Ali, School, House, Car).
2. Something conceptual (that can be apprehended intellectually for example
time, date and so on).
An object has,
1. State (attributes)
2. Well-defined behavior (operations)
3. Unique identity
01.6. Tangible and Intangible Objects
Examples of Tangible Objects:

We will study these features in detail in this course

12

Virtual University of Pakistan

Object Oriented Programming (CS304)

VU

Ali is a tangible object, having some characteristics (attributes) and behavior as given
below,
Ali
Characteristics (attributes)
Name
Age

Behaviour (operations)
Walks
Eats

We will identify Ali using his name.

Car is also a tangible object having some characteristics (attributes) and behavior
given below,
Car
State (attributes)
Color
Model

Behavior (operations)
Accelerate
Start Car
Change Gear

We can identify Car using its registration number

Examples of Intangible Objects (also called as conceptual objects):
Time is an intangible (conceptual) object
Time
State (attributes)

Behavior (operations)

Hours
Seconds
Minutes

Set/Get Hours
Set/Get Seconds
Set/Get Minutes

We will assign our own generated unique ID in the model for Time object
Date is also an intangible (conceptual) object
State (attributes)
Year
Day
Month

Behavior (operations)
Set/Get Year
Set/Get Day
Set/Get Month

We will assign our own generated unique ID in the model for Date object.
01.7. Summary:

Virtual University of Pakistan

13

Object Oriented Programming (CS304)

14

VU

Model is the abstraction of some real word scenario. It helps us to understand

that scenario.
Object oriented model of any scenario (problem) describes that scenario
(problem) in the form of interacting objects.
We use Object Orientation because it helps us in mapping real world problem
in a programming language.
Object Orientation is achieved using objects and their relationships.
Properties of an object are described using its data members and behavior of an
object is described using its functions.
Objects may be tangible (physical) or intangible (also called conceptual or
virtual).
Generally when we have given a certain problem description, nouns in that
problem description are candidates for becoming objects of our system.
There may be more than one aspects of an object
It is not necessary that every object has a specific role in implementation of a
problem there may be some objects without any role, like school parking in
our school.
It is easier to develop programs using Object Oriented Programming because
it is closer to real life.

Virtual University of Pakistan

Object Oriented Programming (CS304)

VU

Lecture No.02
Lecture Contents
1.
2.
3.
4.
5.
6.

Information Hiding
Encapsulation
Interface
Implementation
Separation of Interface & Implementation
Messages

02.1. Information Hiding:

Information hiding is one of the most important principles of OOP inspired from real
life which says that all information should not be accessible to all persons. Private
information should only be accessible to its owner.
By Information Hiding we mean Showing only those details to the outside world which
are necessary for the outside world and hiding all other details from the outside world.
Real Life Examples of Information Hiding
1. Alis name and other personal information is stored in his brain we cant
access this information directly. For getting this information we need to ask
Ali about it and it will be up to Ali how much details he would like to share
with us.
2. An email server may have account information of millions of people but it
will share only our account information with us if we request it to send
anyone else accounts information our request will be refused.
3. A phone SIM card may store several phone numbers but we cant read the
numbers directly from the SIM card rather phone-set reads this information
for us and if the owner of this phone has not allowed others to see the
numbers saved in this phone we will not be able to see those phone numbers
using phone.
In object oriented programming approach we have objects with their attributes and
behaviors that are hidden from other classes, so we can say that object oriented
programming follows the principle of information hiding.
In the perspective of Object Oriented Programming Information Hiding is,
Hiding the object details (state and behavior) from the users
Here by users we mean an object of another class that is calling functions
of this class using the reference of this class object or it may be some other
program in which we are using this class.
Information Hiding is achieved in Object Oriented Programming using the
following principles,

Virtual University of Pakistan

15

Object Oriented Programming (CS304)

VU

All information related to an object is stored within the object

It is hidden from the outside world
It can only be manipulated by the object itself

Advantages of Information Hiding

Following are two major advantages of information hiding,
It simplifies our Object Oriented Model:
As we saw earlier that our object oriented model only had objects and their
interactions hiding implementation details so it makes it easier for everyone to
understand our object oriented model.
It is a barrier against change propagation
As implementation of functions is limited to our class and we have only given the
name of functions to user along with description of parameters so if we change
implementation of function it doesnt affect the object oriented model.
We can achieve information hiding using Encapsulation and Abstraction, so we see
these two concepts in detail now,
02.2. Encapsulation
Encapsulation means we have enclosed all the characteristics of an object in the object
itself
Encapsulation and information hiding are much related concepts (information
hiding is achieved using Encapsulation)
We have seen in previous lecture that object characteristics include data members
and behavior of the object in the form of functions.
So we can say that Data and Behavior are tightly coupled inside an object and
both the information structure and implementation details of its operations are
hidden from the outer world.
Examples of Encapsulation
Consider the same example of object Ali of previous lecture we described it as
follows,

16

Virtual University of Pakistan

Object Oriented Programming (CS304)

VU

Ali
Characteristics
(attributes)

Name

Age
Behavior
(operations)
Walks
Eats

You can see that Ali stores his personal information in itself and its behavior is
also implemented in it.
Now it is up to object Ali whether he wants to share that information with
outside world or not. Same thing stands for its behavior if some other object in
real life wants to use his behavior of walking it can not use it without the
permission of Ali.
So we say that attributes and behavior of Ali are encapsulated in it.
Any other object dont know about these things unless Ali share this information
with that object through an interface,
Same concept also applies to phone which has some data and behavior of
showing that data to user we can only access the information stored in the phone
if phone interface allow us to do so.
Advantages of Encapsulation
The following are the main advantages of Encapsulation,
a. Simplicity and clarity
As all data and functions are stored in the objects so there is no data or function
around in program that is not part of any object and is this way it becomes very
easy to understand the purpose of each data member and function in an object.
b. Low complexity
As data members and functions are hidden in objects and each object has a
specific behavior so there is less complexity in code there will be no such
situations that a functions is using some other function and that functions is
using some other function.
c. Better understanding
Everyone will be able to understand whole scenario by simple looking into object
diagrams without any issue as each object has specific role and specific relation
with other objects.
02.3. Interface

Virtual University of Pakistan

17

Object Oriented Programming (CS304)

VU

Interface is a set of functions of an object that he wants to expose to other objects.

As we discussed previously that data and behavior of each object is hidden in
that object it self so we have to use the concept of interface of the object to expose
its behavior to outer word objects.

Different objects may need different functions of an object so interface of

an object may be different for different objects.
Interfaces are necessary for object communication. Each object provides
interface/s
(operations) to other objects through these interfaces other
objects communicate with this object.

Example Interface of a Car

Steer Wheels
Accelerate
Change Gear
Apply Brakes
Turn Lights On/Off

Example Interface of a Phone

Input Number
Place Call
Disconnect Call
Add number to address book
Remove number
Update number

02.4. Implementation
It is actual implementation of the behavior of the object in any Object Oriented
language.
It has two parts,

Internal data structures to hold an object state that will be hidden from us
it will store values for an object data members.
Functionality in the form of member functions to provide required
behavior.

Examples of Implementation
a. Gear Box in car system
Consider object Gear Box in car system it has a certain structure and
functionality. When this object will be implemented it will have two things,
Physical structure of the gear box
Functionality implemented in this structure to change gear.
Both these things are part of implementation.

18

Virtual University of Pakistan

Object Oriented Programming (CS304)

VU

So it has,
Data Structure in the form of Mechanical structure of gear box
Functionality mechanism to change gear

b. Address Book in a Phone

Similarly take the example of contact details saved in the SIM of a phone,
In that case we can say physical structure of SIM card as Data Structure
And Read/write operations provided by the phone as Functionality.
02.5. Separation of Interface & Implementation
As discussed earlier we only show interface of an object to outside world and
hide actual implementation from outside world. The benefit of using this
approach is that our object interface to outside word becomes independent
from inside implementation of that interface.
This is achieved through the concepts of encapsulation and information
hiding.
Real Life example of separation of interface and implementations

Driver has a standard interface to drive a car and using that interface
he drive can drive any car regardless of its model or type whatever
engine type it has or whatever type of fuel it is using.
02.6. Messages
Objects communicate through messages they send messages (stimuli) by
invoking appropriate operations on the target object. The number and kind of
messages that can be sent to an object depends upon its interface
Examples Messages
A Person sends message (stimulus) stop to a Car by applying brakes
A Person sends message place call to a Phone by pressing appropriate button
02.7. Summary

Information hiding is achieved through encapsulation.

Encapsulation and Information Hiding are related to each other.
Interface of an object provides us the list of available functions.
An object may have more than one interface.
Interface and implementation are separated from each other to achieve
Information Hiding.
Objects communicate with each other using messages.

Virtual University of Pakistan

19

Object Oriented Programming (CS304)

VU

Useful Links:
http://www.alice.org/
A Graphical Programming Environment to teach Computer Programming.

20

Virtual University of Pakistan

Object Oriented Programming (CS304)

VU

Lecture No.03
Lecture Contents:

Abstraction
Classes
Inheritance
Major benefits of inheritance (Reuse)

03.1. Abstraction
Real life objects have a lot of attributes and many kind of behaviors but most of the
time we are interested in only that part of the objects that is related to the problem
we are currently going to solve, for example in implementing a school system we
dont need to take care of the personnel life of a student or a teacher as it will not
effect our system in any way so we will see these objects in the perspective of school
system and will ignore their other characteristics, this concept is called Abstraction.
Abstraction is a way to cope with complexity and it is used to simplify things.
Principle of abstraction:
Capture only those details about an object that are relevant to current perspective
Abstraction Example:
Suppose we want to implement abstraction for the following statement,
Ali is a PhD student and teaches BS students
Here object Ali has two perspectives one is his student perspective and second is his
teacher perspective.
We can sum up Alis attributes as follows,
Name
Age
Student Roll No
Year of Study
CGPA
Employee ID
Designation
Salary

As you can see out of all these listed attributes some belong to Alis student
perspective(Roll No, CGPA, Year of study) and some belong to Alis teacher
perspective(Employee ID, Designation, Salary).
Similarly we can sum up Alis behavior as follows,
Study
DevelopExam
GiveExam

Virtual University of Pakistan

21

Object Oriented Programming (CS304)

VU

TakeExam
PlaySports
Eat
DeliverLecture
Walk

As was the case with attributes of object Ali, its behavior can also be divided in Alis
student perspective as well as Alis teacher perspective.
Students Perspective
Attributes:
- Name
- Student Roll No
- Year of Study
- CGPA

- Employee ID
- Designation
- Salary
- Age

Behaviour:
- Study
- GiveExam
- PlaySports
- DeliverLecture

- DevelopExam
- TakeExam
- Eat
- Walk

Teachers Perspective
Attributes:
- Name
- Student Roll No
- Year of Study
- CGPA
Behaviour:
- Study
- GiveExam
- PlaySports
- DeliverLecture

- Employee ID
- Designation
- Salary
- Age

- DevelopExam
- TakeExam
- Eat
- Walk

A cat can be viewed with different perspectives

Ordinary Perspective
A pet animal with
Four Legs
A Tail
Two Ears
Sharp Teeth

22

Surgeons Perspective
A being with
A Skeleton
Heart
Kidney
Stomach

Virtual University of Pakistan

Object Oriented Programming (CS304)

VU

A car can be viewed with different perspectives

Drivers View

Engineers View

Abstraction Advantages
Abstraction has following major advantages,
1. It helps us understanding and solving a problem using object oriented
approach as it hides extra irrelevant details of objects.
2. Focusing on single perspective of an object provides us freedom to change
implementation for other aspects of for an object later.
Similar to Encapsulation Abstraction is also used for achieving information hiding as
we show only relevant details to related objects, and hide other details.
03.2. Classes
In OOP we create a general sketch for each kind of objects and then we create
different instances using this sketch we call this sketch or prototype or map as
class.
All objects of same kind exhibit identical characteristics (information structure and
behavior) however they have data of their own.
Class Example 1
Consider the objects given below,
Ali studies mathematics
Anam studies physics
Sohail studies chemistry
Each one is a Student so we say these objects are instances of the Student class.
Class Example 2
Consider the objects given below,

Ahsan teaches mathematics

Aamir teaches computer science
Virtual University of Pakistan

23

Object Oriented Programming (CS304)

VU

Atif teaches physics

Each one is a teacher so we say these objects are instances of the Teacher class
Class Representation:
we can represent a class using a rectangle as follows,
(Class Name)
(Attributes)
(Operations)
(Class Name)
Normal Form

Suppressed Form

Class Example: Circle

Circle
center
radius

draw
computeArea

Normal Form

Circle

Suppressed Form

Class Example: Person

Person

name
age
gender

eat
walk
Person
Normal Form

Suppressed Form

03.3. Inheritance

24

Virtual University of Pakistan

Object Oriented Programming (CS304)

VU

A child inherits characteristics of its parents, besides inherited characteristics, a child

may have its own unique characteristics
Inheritance in Classes
If a class B inherits from class A then it contains all the characteristics (information
structure and behaviour) of class A
The parent class is called base class and the child class is called derived class
Besides inherited characteristics, derived class may have its own unique
characteristics

Person

Doctor

Student
Teacher

Shape

Triangle

Line
Circle

Inheritance IS A or IS A KIND OF Relationship

Each derived class is a kind of its base class

Virtual University of Pakistan

25

Object Oriented Programming (CS304)

VU

Person
name
age
gender
eat
walk

Student

Teacher

Doctor

program
studyYear

designation
salary

designation
salary

study
heldExam

teach
takeExam

checkUp
prescribe

Here,
Student IS A Person
Teacher IS A Person
Doctor IS A Person

Shape
color
coord
draw
rotate
setColor

Triangle

Circle
radius
draw
computeArea

Line
length
draw

Here,
Circle IS A Shape
Line IS A Shape
Triangle IS A Shape
Inheritance Advantages
1. Reuse

26

Virtual University of Pakistan

angle
draw
computeArea

Object Oriented Programming (CS304)

VU

2. Less redundancy
3. Increased maintainability
Reuse with Inheritance
Main purpose of inheritance is reuse, we can easily add new classes by inheriting from
existing classes.
Select an existing class closer to the desired functionality, create a new class and inherit it
from the selected class, add to and/or modify the inherited functionality
Shape
color
coord
draw
rotate
setColor

Triangle

Circle
radius

Line

draw
computeArea

length

angle
draw
computeArea

draw

Person
name
age
gender
eat
walk

Student

Teacher

Doctor

program
studyYear

designation
salary

designation
salary

study
heldExam

teach
takeExam

checkUp
prescribe

Virtual University of Pakistan

27

Object Oriented Programming (CS304)

VU

Person
name
age
gender
eat
walk

Student

28

Teacher

Doctor

program
studyYear

designation
salary

designation
salary

study
heldExam

teach
takeExam

checkUp
prescribe

Virtual University of Pakistan

Object Oriented Programming (CS304)

VU

Lecture No.04
Lecture Contents
Generalization
Sub typing (extension)
Specialization (restriction)
Overriding
Abstract classes
Concrete classes
Recap Inheritance

Derived class inherits all the characteristics of the base class

Besides inherited characteristics, derived class may have its own
unique characteristics
Major benefit of inheritance is reuse

04.1. Concepts Related with Inheritance

o
o
o

Generalization
Subtyping (extension)
Specialization (restriction)

04.2. Generalization
In OO models, some classes may have common characteristics.
We extract these features into a new class and inherit original classes from this new
class. There are many objects with common characteristics in object model. The
common characteristics (attributes and behaviour) of all these objects are combined
in a single general class. Base class encapsulates the idea of commonality of derived
classes. Base class is general class representing common behaviour of all derived
classes.
This concept is known as Generalization.
It reduces the redundancy and gives us reusability, using generalization our solution
becomes less complex.
In generalization there should be Is a Kind of Relationship (also called Is A
relationship) between base and child classes.
Example: Line, Circle and Triangle

Virtual University of Pakistan

29

Object Oriented Programming (CS304)

VU

Line

Circle

Triangle

color
vertices
length

color
vertices
radius

color
vertices
angle

move
setColor
getLength

move
setColor
computeArea

move
setColor
computeArea

Line is shape

Circle is a shape

Triangle is a shape

Shape
color
vertices
move
setColor

Triangle

Circle
radius
computeArea

Line
length
getLength

Common attributes
Color vertices
Common behaviour
Set Color, Move
Example: Student Doctor and Teacher

30

Virtual University of Pakistan

angle
computeArea

Object Oriented Programming (CS304)

Student

VU

Teacher

Doctor

name
age
gender
program
studyYear

name
age
gender
designation
salary

name
age
gender
designation
salary

study
heldExam
eat
walk

teach
takeExam
eat
walk

checkUp
prescribe
eat
walk

Person
name
age
gender
eat
walk

Student

Teacher

Doctor

program
studyYear

designation
salary

designation
salary

study
heldExam

teach
takeExam

checkUp
prescribe

Common attributes,
Name, age, gender

Common behaviour
Eat, Walk

Sub-typing & Specialization

We want to add a new class to an existing model
We have developed an existing class hierarchy
Find an existing class that already implements some of the desired state and
behaviour
Inherit the new class from this class and add unique behaviour to the new
class
04.3. Sub-typing (Extension)
Sub-typing means that derived class is behaviourally compatible with the
base class
Derived class has all the characteristics of base class plus some extra
characteristics
Behaviourally compatible means that base class can be replaced by the
derived class

Virtual University of Pakistan

31

Object Oriented Programming (CS304)

VU

Sub-typing (Extension) - Example

Person
Shape
color
vertices

name
age
gender

setColor
move

eats
walks

Student

Circle

program
studyYear

radius
computeCF
computeArea

study
takeExam

Circle is extending the behaviour of

shape, it is extending attributes of shape
by adding radius similarly it is extending
behaviour of shape by adding compute
Circumference and compute Area.

Student has two extra attributes program

and studyYear
Similarly it has extended behaviour by
adding study and takeExam.

Subtyping and generalization are related concepts, Subtyping (extension) and

generalization is a way to look same thing in two ways.
Sub typing is looking at things from Top to bottom whereas in generalization we
look at things from bottom to top.
04.4. Specialization (Restriction)
We want to add a class to existing hierarchy of classes having many similarities to
already existing classes but some part of its behaviour is different or restricted. In
that case we will use the concept of specialization.
Specialization means that derived class is behaviourally incompatible with
the base class
Behaviourally incompatibility means that base class cant always be replaced
by the derived class
Derived class has some different of restricted characteristics than of base
class.
Example Specialization (Restriction)
Suppose we want to add one more class of Adult for some special requirement like
for ID card generation such that it is a person but its age is greater than 18 and
having all other behaviour of that of person class. One solution is that we write

32

Virtual University of Pakistan

Object Oriented Programming (CS304)

VU

another class from beginning and write all code of person again in it with age limit,
but better solution is that we derive adult class from person class and restrict age in
that class as shown below in diagram,

Person
age : [0..100]

setAge( a )

Adult
age : [18..100]

age = a

If age < 18 then

error
else
age = a

setAge( a )

Similarly Natural Numbers2 are also Integers3 with the restriction that natural
numbers set can NOT contain zero or negative integers it consists of only positive
integers so we can implement this relationship also as specialization,

Natural numbers: positive integers only (numbers from 1 to

.onwards)

Integers: all positive and negative numbers (..-3 , -2 , -1 , 0 , 1 , 2 ,

3)

Virtual University of Pakistan

33

Object Oriented Programming (CS304)

VU

IntegerSet

add( elem )

NaturalSet

add( elem )

add element to the

set

If elem < 1 then

error
else
add element to
the set

Add method behaviour is present in both base and derived classes but derived class
behaviour is different in derived class. Derived class will not exhibit the behaviour of
base class but it is overriding behaviour of base class with its own behaviour.
04.5. Overriding
A class may need to override the default behaviour provided by its base class
Derived class overrides the behaviour of its base class.
Reasons for overriding
Provide behaviour specific to a derived class (specialization)
Extend the default behaviour (extension)
Restrict the default behaviour (restriction)
Improve performance
It is used for the implementation of inheritance.
Example Specific Behaviour (Specialization)

34

Virtual University of Pakistan

Object Oriented Programming (CS304)

VU

Shape
color
vertices
draw
move
setColor

Line

Circle
radius

length

draw
computeArea

draw

Triangle
angle
draw
computeArea

Example Extention
Window
width
height
open
close
draw

DialogBox
controls

1- Invoke Windows
draw
2- draw the dialog
box

enable
draw

Example Restriction

Virtual University of Pakistan

35

Object Oriented Programming (CS304)

VU

IntegerSet

Add element to
the set

add( elem )

If elem < 1 then

give error
else
Add element
to the set

NaturalSet

add( elem )

Example Improve Performance

Class Circle overrides rotate operation of class Shape with a Null operation.
Shape
color
coord
draw
rotate
setColor

Circle
radius
draw
rotate

04.6. Abstract Classes

In our examples we made classes for shape and person. These are abstract concepts
and the classes we make against abstract concepts are called abstract classes. They
are present at or near the top in the class hierarchy to present most generalized
behaviour.
An abstract class implements an abstract concept
Main purpose is to be inherited by other classes
Cant be instantiated
Promotes reuse

36

Virtual University of Pakistan

Object Oriented Programming (CS304)

VU

Abstract Classes - Example I

Shape
color
vertices
draw
move
setColor

Triangle

Line

Circle
Here, Shape is an abstract class
Abstract Class
Concrete Classes

Shape
Circle

Line

Triangle

Abstract Classes - Example II

Person
name
age
gender
eat
walk

Student

Doctor
Teacher

Here, Person is an abstract class

Abstract Class

Person

Concrete Classes

Student

Teacher

Doctor

Engineer Director .

Abstract Classes - Example III

Virtual University of Pakistan

37

Object Oriented Programming (CS304)

VU

Vehicle
color
model
accelerate
applyBrakes

Truck

Car
Bus

Here, Vehicle is an abstract class

Abstract Class
Concrete Classes

Vehicle
Car

Bus

Truck

Abstract Classes can not exist standalone in an object model

While making object model we start by finding out objects in our object model and
then we find out objects having common attributes and make them in the form of
general classes at the top of class hierarchies.
04.7. Concrete Classes
The entities that actually we see in our real world are called concrete objects and
classes made against these objects are called concrete classes.
A concrete class implements a concrete concept
These are used to instantiate objects in our programs
Provides implementation details specific to the domain context
Concrete Classes - Example I
Person

Student
program
studyYear

Doctor
Teacher

study
heldExam

38

Virtual University of Pakistan

Object Oriented Programming (CS304)

VU

Here Student, Teacher and Doctor are concrete classes

Concrete Classes - Example II
Vehicle

Truck

Car
Bus

capacity
load
unload

Here Car, Bus and Truck are concrete classes

A concrete class may exist in an object model independently

Concrete classes mostly lie below the top of class hierarchy in a good object
model.

If there is an abstract class then hierarchy exists in the object model as there will
definitely be some concrete classes as well derived from this abstract class otherwise
there is no use of abstract class.
Glossary:
a. Natural numbers: numbers from 1 to .onwards
b. Integers: all positive and negative numbers ..-3,-2,-1,0,1,2,3
c. Whole numbers: numbers from 0 ,1 ,2, 3 .onwards (natural nos including
0)
Some times whole numbers are also called numbers without fractional part.

Virtual University of Pakistan

39

Object Oriented Programming (CS304)

VU

Lecture No.05
Multiple Inheritance
Inheritance:
We saw inheritance purposes in last lecture

Generalization
Extention or sub typing
Specialization or restriction

Abstract and concrete classes, former is used to represent abstract concepts later is
used to represent concrete concepts.
Overriding derived classes override inherited classes (base classes) behaviour.
Overriding is used for Specialization, Extention, Restriction, and Performance.
05.1. Multiple Inheritance
Sometimes we want to reuse characteristics of more than one parent class, in that
case we need to inherit a class from more than one classes.
Example 1 Multiple Inheritance
Consider the example of an imaginary specie Mermaid used in fairy tales that lives in
water having features both of a women as well as of a fish, In Object Oriented
programming perspective Mermaid can be derived from two classes Women and
Fish.

Woman

Fish

Mermaid

C++ Code:
/*Program to demonstrate simple multiple inheritance*/
class Fish {
};
class Woman {
};

40

Virtual University of Pakistan

Object Oriented Programming (CS304)

VU

class Mermaid : public Woman , public Fish {

};

Our Mermaid class inherits features of both woman and fish suppose our woman
class has method wald() and fish cclass has method swim then our mermaid class
can use both methods i.e can walk as well as can swim.

Woman

Fish

void walk()

void swim()

Mermaid

c++ code:
#include <iostream>
#include <stdlib.h>
using namespace std;
/*Program to demonstrate simple multiple inheritance*/
class Fish
{
public:
void swim(){
cout<<"\n In method swim";
}
};
class Woman
{
public:
void walk(){
cout<<"\n In method walk"<<endl;
}
};
class Mermaid : public Woman,public Fish
{

Virtual University of Pakistan

41

Object Oriented Programming (CS304)

VU

};
int main(int argc, char *argv[])
{
Mermaid mermaid;
/*This Mermaid object will have two implicit objects one of Fish class and one of
Woman class*/
mermaid.swim();
mermaid.walk();
system("PAUSE");
return 0;
}

Output:
In method4 swim
In method walk

Example 2 Multiple Inheritance

Take another example of amphibious vehicle (vehicle that can run on land as well as on
water) so it has properties of both land as well as of water vehicle. The general hierarchy in
this case will be,

Vehicle

Land Vehicle

Car

Water Vehicle

Amphibious Vehicle

Boat

Here we have added a general Vehicle class as well to add all common functions of Land
Vehicles and Water Vehicles in that class, and specific functions of Land and Water vehicle
in their respective classes then we have derived Amphibious Vehicle class from Land Vehicle
and Water Vehicle classes (we can do the same in first example as well concerning Woman,
Fish and Mermaid).
C++ code:

class member functions are also called class methods

42

Virtual University of Pakistan

Object Oriented Programming (CS304)

VU

class Vehicle
{
};
class WaterVehicle : public Vehicle
{
};
class LandVehicle : public Vehicle
{
};
class AmphibiousVehicle : public LandVehicle,public
WaterVehicle
{
};

Suppose we have a changeGear method in Vehicle class that is applicable to both water and
land vehicle, we also have Float and Move methods in water and land vehicles respectively
then our amphibious vehicle will have all these methods,
C++ code:
#include <iostream>
#include <stdlib.h>
using namespace std;
/*Multiple Inheritance in case of Amphibious Vehicle*/
class Vehicle
{
public:
void changeGear(){ cout<<"\nI am Vehicle changeGear()
function..\n";}
};
class WaterVehicle : public Vehicle
{
public:
void Float(){ cout<<"\nI am float function of Water Vehicle";}
};
class LandVehicle : public Vehicle

Virtual University of Pakistan

43

Object Oriented Programming (CS304)

VU

{
public:
void Move(){ cout<<"\nI am move function of Land
Vehicle"<<endl;}
};
class AmphibiousVehicle : public LandVehicle,public
WaterVehicle
{
};
int main(int argc, char *argv[])
{
AmphibiousVehicle amphibious;
amphibious.Float();
/*Calling Float function of Water Vehicle class*/
amphibious.Move();
/*Calling Move function of Land Vehicle class*/
system("PAUSE");
return 0;
}

Output:
I am float function of Water Vehicle
I am move function of Land Vehicle

Advantage of Multiple Inheritance:

As was the case with simple (single) inheritance multiple inheritance also decreases
redundant code as we can inherit a class from many classes and can use their
functions without the need to write them again.
However there are more disadvantages of multiple inheritance, than its advantages.
Problems with Multiple Inheritance
Increased complexity

44

Virtual University of Pakistan

Object Oriented Programming (CS304)

VU

Amphibious vehicle hierarchy is a complicated as this class is derived from two

classes that will make code more complex and less understandable however this is
obvious as amphibious vehicle is a complicated vehicle. It is generic problem.
Reduced understanding
Due to increased complexity of class hierarchy the object model becomes difficult it
understand especially for someone who is looking it first time.
Duplicate features
As we are deriving a single class from more than one class so there is a chance of
duplication of features (same methods in both parents), following problems may
arise due to duplicate features,
Problem 1: Ambiguity
Consider the class hierarchy of Mermaid class below,

Woman

Fish

eat

eat

Mermaid

As mermaid also need to eat and its both parents have their own methods of eating
so here question arises,
Which eat operation Mermaid should inherit as both functions are available?
Solution We can solve this problem by explicitly calling eat method from any of
the parent classes in Mermaid class according to behaviour of Mermaid (i.e. if it eats
like a Woman we can call eat method of Woman class and if eats like Fish we can call
method of Fish class), for this we will Override the Common method in multiply
inherited class and in that class overridden method we will call the appropriate base
class function.

Virtual University of Pakistan

45

Object Oriented Programming (CS304)

VU

Woman

Fish

eat

eat

Mermaid
eat

a.
b.

Override eat method in

Mermaid class
Invoke eat operation of
desired parent class

Example C++ Code

#include <iostream>
#include <stdlib.h>
using namespace std;
/*Program to demonstrate simple multiple inheritance*/
class Fish
{
public:
void eat(){
cout<<"\n In Fish eat method ";
}
};
class Woman
{
public:
void eat(){
cout<<"\n In Woman eat method \n"<<endl;
}
};
class Mermaid : public Woman,public Fish
{
public:
void eat(){
cout<<"\n In Mermaid eat method "<<endl;
cout<<"\n Explicity calling Woman eat
method...."<<endl;

46

Virtual University of Pakistan

Object Oriented Programming (CS304)

VU

Woman::eat();
}
};
int main(int argc, char *argv[])
{
Mermaid mermaid;
/*This Mermaid object will have two implicit objects one of Fish class and one of
Woman class*/
mermaid.eat();
/*Calling Mermaid eat method*/
system("PAUSE");
return 0;
}

Problem 2: Two instances for same function (Diamond Problem)

Vehicle
changeGear

Land Vehicle

Car

Water Vehicle

Amphibious Vehicle

Boat

Here Amphibious Vehicle will have two copies of changeGear function as it will
have two objects of Vehicle class one with respect to Land Vehicle and one with
respect to Water Vehicle as shown below,

Virtual University of Pakistan

47

Object Oriented Programming (CS304)

VU

Vehicle

Vehicle

changeGear

changeGear

Land Vehicle

Water Vehicle

Amphibious Vehicle

Actual Memory Layout

Compiler will not be able to decide which changeGear operation Amphibious
Vehicle should inherit so it will generate an error as shown below (two copied of
same method),

error: request for member `changeGear' is ambiguous

error: candidates are: void Vehicle::changeGear()
void Vehicle::changeGear()
Execution terminated

Solution to Diamond Problem

Some languages disallow diamond hierarchy
Others provide mechanism to ignore characteristics from one side. There are two
cases while solving diamond problem virtual inheritance and non virtual inheritance
(we will study these details in coming lectures)

Vehicle
changeGear

Land Vehicle

Car

Water Vehicle

Amphibious Vehicle

Boat

Invoke changeGear
operation with respect to
one side

48

Virtual University of Pakistan

Object Oriented Programming (CS304)

VU

05.2. Kinds of Association:

There are two main types of association which are then further subdivided i.e
1. Class Association
2. Object Association
1. Class Association
Class association is implemented in terms of Inheritance. Inheritance implements
generalization/specialization relationship between objects. Inheritance is considered
class association.
In case of public inheritance it is IS-A relationship.
In case of private inheritance it is Implemented in terms of relationship.
This relationship ensures that public members of base class are available to derived
class in case of public inheritance.
When we create objects of classes in which we have implemented inheritance
relationships we are forcing the inheritance relationship between created objects. In
this case the derived class objects will also contain base class objects attributes and
methods.
2. Object Association
It is the interaction of stand alone objects of one class with other objects of anther
class.
It can be of one of the following types,

Simple Association
Composition
Aggregation

05.3. Simple Association

The two interacting objects have no intrinsic relationship with other object. It is the
weakest link between objects. It is a reference by which one object can interact with
some other object.
Customer gets cash from cashier
Employee works for a company
Ali lives in a house
Ali drives a car

lives-in

Ali
1

House
1

drives

Ali
1

Car
*

It is generally called as association instead of simple association

Virtual University of Pakistan

49

Object Oriented Programming (CS304)

VU

Kinds of Simple Association

Simple association can be categorized in two ways,

With respect to direction (navigation)

With respect to number of objects (cardinality)
Kinds of Simple Association w.r.t Navigation

With respect to navigation association has the following types,

a. One-way Association
b. Two-way Association
a. One-way Association
In One way association we can navigate along a single direction only, it is denoted
by an arrow towards the server object.
Examples:
lives-in

Ali
1

Ali lives in a House

drives

Ali
1

House

Car
*

Ali drives his Car

b. Two-way Association
In two way association we can navigate in both directions, it is denoted by a line
between the associated objects
Examples:
works-for

Employee
*

Company
1

Employee works for company

Company employs employees
Two-way Association - Example

50

Virtual University of Pakistan

Object Oriented Programming (CS304)

VU

friend

Yasir

Ali

Yasir is a friend of Ali

Ali is a friend of Yasir
Kinds of Simple Association w.r.t Cardinality
With respect to cardinality association has the following types,
a. Binary Association
b. Ternary Association
c. N-ary Association
a. Binary Association
It associates objects of exactly two classes; it is denoted by a line, or an arrow
between the associated objects.
Example
works-for

Employee
*

Company
1

Association works-for associates objects of exactly two classes

drives

Ali
1

Car
*

Association drives associates objects of exactly two classes

b. Ternary Association
It associates objects of exactly three classes; it is denoted by a diamond with lines
connected to associated objects.
Example
Objects of exactly three classes are associated

Virtual University of Pakistan

51

Object Oriented Programming (CS304)

VU

Student

Teacher
*

1
*
Course

c. N-ary Association
An association between 3 or more classes its practical examples are very rare.
05.4. Composition
An object may be composed of other smaller objects, the relationship between the
part objects and the whole object is known as Composition, Composition is
represented by a line with a filled-diamond head towards the composer object
Example Composition of Ali

Head
1

Arm
2

Ali

Leg
2

1
Body

Example Composition of Chair

52

Virtual University of Pakistan

Object Oriented Programming (CS304)

VU

Back
1

Chair

2
Arm

Seat

4
Leg

Composition is stronger relationship:

Composition is a stronger relationship, because
Composed object becomes a part of the composer
Composed object cant exist independently
Example I
Ali is made up of different body parts
They cant exist independent of Ali
Example II
Chairs body is made up of different parts
They cant exist independently
05.5. Aggregation
An object may contain a collection (aggregate) of other objects, the relationship
between the container and the contained object is called aggregation, Aggregation is
represented by a line with unfilled-diamond head towards the container
Example Aggregation

Virtual University of Pakistan

53

Object Oriented Programming (CS304)

VU

Bed
1

Chair

Table

Room
*

1
Cupboard

Example Aggregation

Garden

Plant

Aggregation is weaker relationship

Aggregation is weaker relationship, because
Aggregate object is not a part of the container
Aggregate object can exist independently
Example I
Furniture is not an intrinsic part of room
Furniture can be shifted to another room, and so can exist independent of a
particular room
Example II
A plant is not an intrinsic part of a garden
It can be planted in some other garden, and so can exist independent of a particular
garden
http://www.codeproject.com/KB/cpp/oopuml.aspx

54

Virtual University of Pakistan

Object Oriented Programming (CS304)

VU

Lecture No.06
06.1. Class Compatibility
A class is behaviorally compatible with another if it supports all the operations of the
other class. Such a class is called subtype. A class can be replaced by its subtype.
Derived class is usually a subtype of the base class. It can handle all the legal
messages (operations) of the base class. Therefore, base class can always be replaced
by the derived class.
Examples
Child class also includes characteristics of its base class.

Shape
color
vertices
move
setColor
draw

Triangle

Circle
Line
radius
length
draw
computeArea

draw
getLength

angle
draw
computeArea

All the three derived class are behaviourally compatible with base class.

Virtual University of Pakistan

55

Object Oriented Programming (CS304)

VU

File
size

open
print

ASCII File

PS File
PDF File

print

print

print

Wherever the file class is it can be replaced by any of its child classes.
06.2. Polymorphism
It is also essential component of object oriented modeling (paradigm).
In general, polymorphism refers to existence of different forms of a single entity. For
example, both Diamond and Coal are different forms of Carbon.
06.3. Polymorphism in OO Model
In OO model, polymorphism means that different objects can behave in different
ways for the same message (stimulus). Consequently, sender of a message does not
need to know exact class of the receiver.
Sender sends message to receiver and appropriate method is called on receiver side.
Example Polymorphism

56

Virtual University of Pakistan

Object Oriented Programming (CS304)

VU

draw

Shape

View

draw

Line

Circle

Triangle

draw

draw

draw

Shape class hierarchy shape is base class and there are three child classes line circle ,
triangle. View send draw method to shape class and draw is called according to the
nature of actual object present.

print
Editor

File
print

ASCII File

PDF File

PS File

print

print

print

Editor sends message print to file class and print is called based on the actual child
object of file class message is same and appropriate execution will be done.
06.4. Polymorphism Advantages
Messages can be interpreted in different ways depending upon the receiver class
New classes can be added without changing the existing model

draw

Shape

View
draw

Square

Line

Circle

Triangle

draw

draw

draw

draw

In general, polymorphism is a powerful tool to develop flexible and reusable systems

Virtual University of Pakistan

57

Object Oriented Programming (CS304)

VU

06.5. Object-Oriented Modeling an Example

Problem Statement
Develop a graphic editor that can draw different geometric shapes such as line, circle
and triangle. User can select, move or rotate a shape. To do so, editor provides user
with a menu listing different commands. Individual shapes can be grouped together
and can behave as a single shape.
Identify Classes
Extract nouns in the problem statement
Develop a graphic editor that can draw different geometric shapes such as line,
circle and triangle. User can select, move or rotate a shape. To do so, editor provides
user with a menu listing different commands. Individual shapes can be grouped
together and can behave as a single shape.
Eliminate irrelevant classes
Editor Very broad scope. But it is the name of overall system and we are going to
model it so we will not make its object. For example if we are going to model
computer we will not make its object but its components however if it is component
of some other system then it will behave as an object. So it is marked as irrelevant.
User Out of system boundary, it is interacting with the system from outside of the
system.
Add classes by analyzing requirements
Group (of shapes) required to behave as a shape so it should behave as an object in
our system
Individual shapes can be grouped together and can behave as a single shape
View graphic editor must have a display area to show the shapes. We made this
object using domain knowledge.

Shape
Line
Circle
Triangle
Menu
Group
View

So we have the following classes,

58

Virtual University of Pakistan

Object Oriented Programming (CS304)

Shape

VU

G roup

Line

Menu

Circle

View
Triangle

Finding Associations:
Next step is to find associations between the objects.
Identify Associations
Find relationships between objects,
1. Extract verbs connecting objects,
Individual shapes can be grouped together

Group consists of lines, circles, triangles

Group can also consists of other groups (Composition)

Line, circle and triangle have composition relationship.

2. Verify access paths
a. View contains (draws) shapes

View contains lines

View contains circles
View contains triangles
View contains groups

So there is Aggregation relationship between shapes and View.

Menu sends message to View
So there is Simple One-Way Association relationship between Menu and View.
Identify Attributes of the identified objects

Virtual University of Pakistan

59

Object Oriented Programming (CS304)

Extract properties of the object,

a. From the problem statement
Properties are not mentioned
b. From the domain knowledge

Line
i. Color
ii. Vertices
iii. Length
Circle
i. Color
ii. Vertices
iii. Radius
Triangle
i. Color
ii. Vertices
iii. Angle
Shape
i. Color
ii. Vertices
Group
i. noOfObjects
View
i. noOfObjects
ii. selected
Menu
i. Name
ii. isOpen

Object Model Graphic Editor

Object model so far is shown below,

60

Virtual University of Pakistan

VU

Object Oriented Programming (CS304)

Menu

VU

View

Shape
n
n

n
Line

Circle

Triangle

Group
n

Identify Operations
Extract verbs connected with an object
Develop a graphic editor that can draw different geometric shapes such as line, circle
and triangle. User can select, move or rotate a shape. To do so, editor provides user
with a menu listing different commands. Individual shapes can be grouped together
and can behave as a single shape.
Eliminate irrelevant operations
Develop out of system boundary
Behave have broad semantics
Following are selected operations:
Line
Draw
Select
Move
Rotate
Circle
Draw
Select
Move
Rotate
Triangle
Draw
Select
Move
Rotate
Shape
Draw
Select

Virtual University of Pakistan

61

Object Oriented Programming (CS304)

VU

Move
Rotate
Group
Draw
Select
Move
Rotate
Menu
Open
Select
Move
Rotate
Extract operations using domain knowledge

View

Add
Remove
Group
Show
Select
Move
Rotate

Object Model after adding operations:

View
noOfObjects
selected

Menu
name
isOpen

Shape
color
vertices

add()
remove()
group()
show()
select()
move()
rotate()

open()
select()
move()
rotate()

draw()
select()
move()
rotate()
n

n
n
Line
length
draw()
n

Triangle
angle

n
Circle
radius

Group
noOfObjects
n

n
draw()

Identify Inheritance

62

Virtual University of Pakistan

draw()

draw()

Object Oriented Programming (CS304)

VU

a. Search lines like is a kind of by looking at keywords like such as, for
example, etc
shapes such as line, circle and triangle

Line, Circle and Triangle inherits from Shape

By analyzing requirements
Individual shapes can be grouped together and can behave as a single shape

Group inherits from Shape

Refining the Object Model
Application of inheritance demands an iteration over the whole object model
In the inheritance hierarchy,
All attributes are shared
All associations are shared
Some operations are shared
Others are overridden
Refining the Object Model

Share associations
o
o

View contains all kind of shapes

Group consists of all kind of shapes

Shape Line, Circle, Triangle and Group

 Color, vertices

Share attributes

Share operations

Shape Line, Circle, Triangle and Group

 Select
 Move
 Rotate
Share the interface and override implementation
o

Shape Line, Circle, Triangle and Group

 Draw

Virtual University of Pakistan

63

Object Oriented Programming (CS304)

VU

View
noOfObjects
selected

Menu
name
isOpen

Shape
color
vertices

add()
remove()
group()
show()
select()
move()
rotate()

open()
select()
move()
rotate()

draw() n
select()
move()
rotate()

Line
length

Circle
radius

Triangle
angle

draw()

draw()

draw()

Group
noOfObjects
draw()

View
noOfObjects
selected

Menu
name
isOpen

Shape
color
vertices

add()
remove()
group()
show()
select()
move()
rotate()

open()
select()
move()
rotate()

draw()
select()
move()
rotate()
n

n
Line n
length
draw()
n

64

Triangle
angle

n
Circle
radius

Group
noOfObjects
n

draw()

draw()

Virtual University of Pakistan

draw()

Object Oriented Programming (CS304)

VU

Lecture No.07
The basic concept Object of Object Orientation (thinking in terms of objects) is
realized using classes in programming languages.
07.1. Class
It is a way (Mechanism) given by c++ to realize objects in a program. It is concrete
implementation of objects in c++. We capture any object attributes and behaviour in
a programming language using classes.
In other words it can be defined as facility given by c++ to create new types
according to our requirement. (Class is composite data type made from basic c++
types like integers, chars and float).
Example:
Consider the examples of entity lion there are many lions but all lions will have
similar attributes and behaviour.
Similarly consider student object all students have separate existence but all students
have similar attributes and they exhibit similar behaviour.
When we hear word student or think about student a sketch comes in our mind for
student along with its attributes and behaviour. The attributes of student comes in
our mind are its name, roll no, class, degree so on. Similarly the behaviour of student
comes in our mind are study, register and many more.
We need to capture the characteristic features of any object (attributes and
behaviour) in the programming language. The concept of class is used for this
purpose.
Now consider the scenario having many interacting objects: a University System
having many objects like student, subject, classroom, and teacher so onwe will
realize all these objects in our software using classes. These all object will use the
services of each other for example student will ask teacher to teach him. This
approach is closer to real life instead of having simple functions being called from
main here these objects will call each other to get their services. This is the reason we
say that object oriented programming mimics real life.
Uses
Objects are structured in terms of class so our problem becomes easier to understand
in the terms c++ program.
We can implement interactions easily in terms of classes.
Student objects will interact with each other to take and give services to each other as
happens in real life and mapped in object oriented programming approach.
Now we see how class mechanism helps us in implementing real life concept.
07.2.

Type in C++

We implement generic concepts using types.

Virtual University of Pakistan

65

Object Oriented Programming (CS304)

VU

We have to model generic concept of Student. But there is no built in type student in
c++ like built-in c++ types int or float. Class is mechanism in c++ that will allow us
to define student as user defined type, similarly generic concept circle will also be
implemented in the same way. User define types will be,

Student in student management system

Circle in a drawing software

As objects have attributes and behaviour so corresponding classes will also have data
members and methods as shown below,
Corresponding class

Ali

/*c++ code for class Person, we can create any

object like Ali from it*/
class Person {
private:
/* attributes are generally made
private*/
char name[]; /*char array to store name*/
int age; /*int age to store age*/
public: /* methods are generally made
public*/
Person(); /*constructor used to initialize
data members*/
void walks(); /* method walk */
void eats(); /*method eats*/
}

Characteristics (attributes)

Name
Age
Behavior (operations)

Walks
Eats

a. object

b. class code

07.3. Abstraction
We only include those details in the system that are required for making a functional
system so we will leave out irrelevant attributes and behaviour from our objects.
Take the example of student,
Student
Name
Address
Relevant to our problem
Sibling
Father Business
Not relevant to our problem
07.4. Defining a New User Defined Type
There are two ways to create user defined types for objects in c++ these are,
Structure Definition:
Partially we can use Structures to define an object
Struct{
};

66

Virtual University of Pakistan

Object Oriented Programming (CS304)

VU

In c we can not define functions in a structure however in c++ we can add functions
in both structure and classes.
Class Definition:

class keyword (small case)

class identifier

class ClassName
{
Access Specifier: (public, private or protected)
DataType MemberVariable;
.
Access Specifier: (public, private or protected)
ReturnType MemberFunction();
.

Data members

Function members

};

Example
class Student
{
private:
int rollNo;
char *name;
float CGPA;
char *address;

Member Variables

public:
void setName(char *newName);
void setRollNo(int newRollNo);

Member Functions

};
Why Member Functions:
They model the behaviors of an object,
Objects can make their data invisible (in accordance with the principle of data
hiding). Setters and getters functions are provided by class to access the its members
it also minimizes the changes to move the objects in inconsistent state as we can write
checks in our setter functions for example we can check that whether the user has
entered correct age value and has not entered negative value for age.
Object remains in consistent state
Example:

Virtual University of Pakistan

67

Object Oriented Programming (CS304)

VU

We can check that the entered roll number by user is positive or negative,
Student aStudent;
aStudent.rollNo = 514;
aStudent.rollNo = -514;

//Error

07.5. Object and Class:

Object is an instantiation of a user defined type or class. Once we have defined a
class we can create as many objects for that class as we require.
Declaring class variables
Variables of classes (objects) are declared just like variables of structures and built-in
data types as follows,
TypeName VariableName;
int var; // declaring built in int data type variable
Student aStudent; // declaring user defined class Student object
07.6. Accessing members
Members of an object can be accessed using,
a. dot operator (.) to access via the variable name
Student aStudent; // declaring Student object
aStudent. rollNo = 5;
b. arrow operator (->) to access via a pointer to an object
Student * aStudent = new Student();
// declaring and initializing Student pointer
aStudent->rollNo = 5;
Note: it is against the principle of OOP to access the data members directly using
object of class as we have done above. This code is given for example only we should
write assessor functions (setters and getters) wherever we want to access the
members of the class.
Member functions are accessed in the similar way using dot or arrow operator.
Example
class Student{
int rollNo;
void setRollNo(int aNo);
};
Student aStudent;
aStudent.setRollNo(5);
Student *ptr_student = new Student();
ptr_student->setRollNo(5);

68

Virtual University of Pakistan

Object Oriented Programming (CS304)

VU

07.7. Access specifiers

These are used to enforce access restrictions to members of a class, there are three
access specifiers,
1. public is used to tell that member can be accessed whenever you have
access to the object
2. private is used to tell that member can only be accessed from a member
function
3. protected to be discussed when we cover inheritance
Example
class Student{
private:
char * name;
Can not be accessed outside the class
int rollNo;
public:
void setName(char *);
Can be accessed outside the class
void setRollNo(int);
...
};
Example Program
class Student{
char * name;
int rollNo;
public:
void setName(char *);
void setRollNo(int aNo);
};
void Student::setName(char * aName){
if (strlen(aName) > 0)
{
name = new char[strlen(aName)];
strcpy(name,aName);
}
}
void Student::setRollNo(int arollNo){
if(arollNo > 0)
rollNo = arollNo;
}
int main(){
Student aStudent;
aStudent.rollNo = 5;
/* Error: we can not access private member of the class. */
aStudent.name = Ali;
/* Error: we can not access private member of the class */
aStudent.setRollNo(1);
aStudent.setName(Ali);

Virtual University of Pakistan

69

Object Oriented Programming (CS304)

VU

/* Correct way to access the data member using public setter functions */
}

Default access specifier

When no access specifier is mentioned then default access specifier is private.
Example

class Student
{
char * name;
int RollNo;
};

Equivalent

class Student
{
private:
char * name;
int RollNo;
};

Example
We should use keyword public before the methods of the class as given below of will
not use public keyword they will also be treated as private ad will not be accessible
outside the class as shown below,
class Student
{
char * name;
int RollNo;
void SetName(char *);
};
Student aStudent;
aStudent.SetName(Ali);
Corrected code will be,
class Student
{
char * name;
int RollNo;
public:
void setName(char *);
};
Student aStudent;
aStudent.SetName(Ali);

70

Virtual University of Pakistan

Error in accessing
SetName it will be
inaccessible

Object Oriented Programming (CS304)

VU

Lecture No.08
08.1. Member Functions

Member functions are the functions that operate on the data encapsulated in
the class
Public member functions are the interface to the class

08.2. Defining Member Functions

We can define member functions in two ways,
a. We can define member functions of the class inside the class definition when
we define any class in our program.
OR
b. We declare member function inside the class definition and declare them
outside the class.
In this case class definition is added in ClassName.h file and class implementation
code is added in ClassName.cpp file.
Function definition inside the class:
General Syntax:
class ClassName {

public:
ReturnType FunctionName() {

}
};

Example:
Define a class of student that has a roll number. This class should have a function
that can be used to set the roll number
class Student{
int rollNo;
public:
void setRollNo(int aRollNo){
rollNo = aRollNo;
}
};

Function definition outside class

General Syntax:
class ClassName {

public:

Virtual University of Pakistan

71

Object Oriented Programming (CS304)

VU

ReturnType FunctionName();
};
ReturnType ClassName::FunctionName()
{

Scope resolution operator

Example
class Student{

int rollNo;
public:
void setRollNo(int aRollNo);
};
void Student::setRollNo(int aRollNo){

rollNo = aRollNo;
}

08.3. Inline Functions

Inline functions is a way used by compilers to improve efficiency of the

program, when functions are declared inline normal process of function
calling (using stack) is not followed instead function code is added by
compiler at all points where these functions have been called. Basic concept
behind inline functions is that they are functions in our code but in compiler
generated files these functions code is added by compiler at all places where
they were called in the code.
Normally small size functions that need to be called many times during
program execution are declared inline. Inline functions decrease the code
execution time because program in their case doesnt involve function call
overhead.
Keyword inline is used to request compiler to make a function inline.
However using inline keyword with function doesnt guarantee that function
will definitely in inlined, it depends on the compiler if it finds it can make
function inline it does so otherwise it ignores the keyword inline and treat the
function as normal function.

Example
inline int Area(int len, int hi)
{
return len * hi;
}
int main()
{
cout << Area(10,20);
}

72

Virtual University of Pakistan

Object Oriented Programming (CS304)

VU

Inline Functions
The functions defined inside the class are by default inline (whether we mention
keyword inline with them or not)
In case we define function outside the class then we must use the keyword inline to
make the function inline.
However compiler decides whether it will implement these functions code as inline
or not.

Example
Inline function inside the class:
class Student{
int rollNo;
public:
void setRollNo(int aRollNo){

rollNo = aRollNo;
}
};

Function setRollNo
will be automatically
inlined by compiler.

Example
Inline function outside the class:
class Student{

public:
inline void setRollNo(int aRollNo);
};
void Student::setRollNo(int aRollNo){

rollNo = aRollNo;
}
class Student{

public:
void setRollNo(int aRollNo);
};
inline void Student::setRollNo(int
aRollNo){

rollNo = aRollNo;
}
class Student{

public:
inline void setRollNo(int aRollNo);
};
inline void Student::setRollNo(int
aRollNo){

rollNo = aRollNo;
}

Virtual University of Pakistan

73

Object Oriented Programming (CS304)

VU

08.4. Constructor
Constructor is used to initialize the objects of a class. Constructor is used to ensure
that object is in well defined state at the time of creation.
The constructor of a class is automatically generated by compiler however we can
write it by our self also.
Constructor is automatically called when the object is created. Constructors are not
usually called explicitly by us.
08.5. Constructor Properties

Constructor is a special function having same name as the class name

Constructor does not have return type
Constructors are commonly public members

Example
class Student{
int rollNo;
public:
Student(){
rollNo = 0;

Student Class
constructor

}
};
int main()
{
Student aStudent;
/*constructor is implicitly called at this point*/
}
We can assure that constructor is called automatically by adding cout statement in
constructor as shown below,
#include <iostream>
using namespace std;
class Student{
Student Class
int rollNo;
constructor
public:
Student(){
rollNo = 0;
cout<<I am constructor of Student class\n;
}
};
int main()
{
Student aStudent;
/*constructor is implicitly called at this point*/
system(pause);

74

Virtual University of Pakistan

Object Oriented Programming (CS304)

VU

return 0;
}
08.6. Default Constructor

Constructor without any parameter or with all parameters with default

values is called default constructor
If we do not define a default constructor the compiler will generate a default
constructor
Compiler generated default constructor is called implicit and user written default
constructor is called explicit
This compiler generated default constructor initialize the data members to
their default values
If we have given any constructor for a class whether it is
our own explcit default constructor ( i.e parameterless or with
parameters having default values )
or
our own constructor with parameters

Then compiler will not create implicit default constructor5.

Example
Consider the class student below it has no constructor so compiler will generate one for it,

class Student
{
int rollNo;
char *name;
float GPA;
public:

};

//no constructors

Code of Compiler generated implicit default constructor

{
rollNo = 0;
GPA = 0.0;
name = NULL;
}
08.7. Constructor Overloading
We can write constructors with parameters as well. These parameters are used to
initialize the data members with user supplied data (passed as parameter). The
5

compiler generated default constructor is called implicit and user written default constructor is called
explicit

Virtual University of Pakistan

75

Object Oriented Programming (CS304)

VU

example is shown below, here example Student class has four constructors their
prototypes are,
1.
2.
3.
4.

Student(); /* explicit default parameterless constructor */

Student(char * aName); /* constructor with one parameter* /
Student(char * aName, int aRollNo); /* constructor with two parameters */
Student(int aRollNo, int aRollNo, float aGPA); /* constructor with three
parameters */

Example
class Student{
int rollNo;
char *name;
float GPA;
public:
Student(); /* explicit default constructor */
Student(char * aName); /* constructor with one parameter* /
Student(char * aName, int aRollNo); /* constructor with two parameters */
Student(int aRollNo, int aRollNo, float aGPA); /* constructor with three
parameters */
};
Student::Student(){
rollNo = 0;
name = NULL; // to indicate it is pointing to nothing at this moment
otherwise it can generate erroneous code.
GPA = 0.0;
}
Student::Student(int aRollNo){
if(aRollNo < 0){
rollNo = 0;
}
else {
rollNo = aRollNo;
}
name = NULL;
}
Student::Student(int aRollNo,
char * aName){
if(aRollNo < 0){
rollNo = 0;
}
else {
rollNo = aRollNo;

76

Virtual University of Pakistan

Object Oriented Programming (CS304)

VU

}
if (strlen(aName) > 0)
{
name = new char[strlen(aName)+1];
strcpy(name,aName);
}
else
{
name = NULL;
}
}

We can create this Student class object using any one of these constructors as follows,
int main()
{
Student student1; // default constructor will be used
Student student2(Name); // one parameter constructor will be used
Student student3(Name, 1); // two parameter constructor will be used
Student student4(Name,1,4.0); // three parameter constructor will be used
}

08.8. Constructor Overloading

We can use default parameter values to reduce the writing effort in that case we will
have to write only one constructor and it will serve the purpose of all constructors as
given below,
Example
Student::Student(

char * aName = NULL, int aRollNo= 0, float aGPA = 0.0) {

It is equivalent to all three constructors,

Student();
Student(char * aName);
Student(char * aName, int aRollNo);
Student(char * Name, int aRollNo, float aGPA);
It will use default values if values are not passed as arguments while creating
objects) it is described in code given below,
int main()

Virtual University of Pakistan

77

Object Oriented Programming (CS304)

VU

{
Student student1; /*char * aName = NULL, int aRollNo= 0, float aGPA = 0.0*/
Student student2( Name ); /*char * aName = Name, int aRollNo= 0, float
aGPA = 0.0*/
Student student3( Name, 1 ); /*char * aName = Name, int aRollNo= 1, float
aGPA = 0.0*/
Student student4( Name, 1 , 4.0); /*char * aName = Name, int aRollNo= 1,
float aGPA = 4.0*/

08.9. Copy Constructor

Copy constructors are used when:

Initializing an object at the time of creation (we want to create an object with
state of a pre existing object)
When an object is passed by value to a function (As you know temporary
copy of object is created on stack so we need copy constructor to create that
temporary object with the state of actual object being passed).

Example
void func1(Student student){

Copy constructor will

be called to create
temporary student
object

int main(){
Student studentA;
Student studentB = studentA;
func1(studentA);
}

Copy constructor will

be called as we are
creating studentB in
terms of studentA.

Copy Constructor (Syntax)

Student::Student( const Student &obj){

78

Virtual University of Pakistan

Object Oriented Programming (CS304)

VU

/*copying values to newly created object*/

rollNo = obj.rollNo;
name = obj.name;
GPA = obj.GPA;
}

As was the case with default constructor compiler also generates copy constructor by
itself however we can override that copy constructor by writing our own copy
constructor as shown in example below,
#include <iostream>
using namespace std;
class Student{
int rollNo;
Student class default
public:
constructor
Student(){
rollNo = 0;
cout<<I am default constructor of Student class\n;
}
Student(const Student &obj){
cout<<I am copy constructor of Student class\n;
rollNo = obj.rollNo;
}
};
int main()
{
Student aStudent;
/*default constructor is implicitly called at this point*/

Student class copy

constructor

Student bStudent = aStudent;

/*copy constructor is implicitly called at this point*/
system(pause);
return 0;
}

08.10. Shallow Copy

When we initialize one object with another then the compiler copies state of
one object to the other using copy constructor by assigning data member
values of previous object to newly created object.

Shallow copy using default Copy Constructor (Syntax)

Virtual University of Pakistan

79

Object Oriented Programming (CS304)

Student::Student(

VU

const Student & obj ){

rollNo = obj.rollNo;
name = obj.name;
GPA = obj.GPA;
}

This kind of copying is called shallow copying

Example
Student studentA;
Student studentB = studentA; /*Shallow copy: compiler will use copy constructor to
assign studentA values to newly created object studentB*/

Shallow copy works fine if our class doesnt include dynamic memory allocation but
in case of dynamic memory allocation it leads to dangling pointer problem as
explained below,
Problem is Shallow Copy
Student class data member name of char * type is added to store the name of student
and it is using dynamic memory according to the length of name entered by user for
student.

studentA
Name
RollNo
GPA

Memory
studentB
A
H
M
A
D

Name
RollNo
GPA

Object studentB is also pointing to memory allocated object studentA for Student
class data member name of char * type, now there are two problems with this sort of
copying,

80

Virtual University of Pakistan

Object Oriented Programming (CS304)

VU

Suppose we delete first object studentA for some reason then its destructor will also
free memory allocated by it hence memory area containing name AHMAD will
also be freed and will be given to some other application by operating system, but
studentB member name is still pointing to that area so issue of Dangling Pointer
will arose. [Pointer pointing to incorrect memory location]. If we will try to print the
name of object studentB our program will terminate abnormally as it was pointing
memory of some other applications.
Secondly if for some reason we change name of studentA the value of object
studentB will also be changed as it pointing to same memory location.
We resolve these two issues using deep copy.

08.11. Deep Copy

We write our own deep copy code in copy constructor so that when we create new
object from an existing object using copy constructor we also allocate new dynamic
memory for data members involving dynamic memory as shown below,
Student::Student( const Student & obj){
int len = strlen(obj.name);
name = new char[len+1];
// assignming new dynamic memory to data member name of char * type for
newly created object*/
strcpy(name, obj.name);

//copy rest of the data members in the same way

}

Example

Student studentA;
Student studentB = studentA; // now copy constructor will perform deep copy
(separate memory for both objects)

Virtual University of Pakistan

81

Object Oriented Programming (CS304)

Memory
A
H
M
A
D

VU

A
Name
RollNo
GPA

A
H
M
A
D

Name
RollNo
GPA

In case our class doesnt involve dynamic memory then default copy constructor that
performs shallow copy works fine.
In case our class has any data member involving dynamic memory we have to write
our own code to do deep copy.

82

Virtual University of Pakistan

Object Oriented Programming (CS304)

VU

Lecture No.09
Review
Copy Constructor
Copy constructors are used when:
Initializing an object at the time of creation (we want to create an object with state of
a pre existing object)
When an object is passed by value to a function (As you know temporary copy of
object is created on stack so we need copy constructor to create that temporary object
with the state of actual object being passed).

Example
void func1(Student student){

Copy constructor will

be called as a
temporary student
object will be created.

int main(){
Student studentA;
Student studentB = studentA;
func1(studentA);
}

Copy constructor will

be called as we are
creating studentB in
terms of studentA.

Copy Constructor (Syntax)

Student::Student( const Student &obj){
/*copying values to newly created object*/
rollNo = obj.rollNo;
name = obj.name;
GPA = obj.GPA;
}

As was the case with default constructor compiler also generates copy constructor by
itself however we can override that copy constructor by writing our own copy
constructor as shown in example below,
As was the case with default constructor compiler also generates copy constructor by
itself however we can override that copy constructor by writing our own copy
constructor as shown in example below,
#include <iostream>

Virtual University of Pakistan

83

Object Oriented Programming (CS304)

VU

using namespace std;

class Student{
int rollNo;
Student class default
public:
constructor
Student(){
rollNo = 0;
cout<<I am default constructor of Student class\n;
}
Student(const Student &obj){
cout<<I am copy constructor of Student class\n;
rollNo = obj.rollNo;
}
};

Student class copy

constructor

int main()
{
Student aStudent;
/*default constructor is implicitly called at this point*/
Student bStudent = aStudent;
/*copy constructor is implicitly called at this point*/
system(pause);
return 0;
}

09.1. Shallow Copy

When we initialize one object with another then the compiler copies state of one
object to the other using copy constructor by assigning data member values of
previous object to newly created object. This kind of copying is called shallow
copying.
Shallow copy using default Copy Constructor (Syntax)
Student::Student(

const Student & obj ){

rollNo = obj.rollNo;
name = obj.name;
GPA = obj.GPA;
}

This kind of copying is called shallow copying

Example
84

Virtual University of Pakistan

Object Oriented Programming (CS304)

VU

Student studentA;
Student studentB = studentA; /*Shallow copy: compiler will use copy constructor to
assign studentA values to newly created object studentB*/
Example
Student studentA(Ahmad);

A
H
M
A
D

studentA
RollNo
Name

Heap

GPA

Student studentB = studentA;

studentA
RollNo

A
H
M
A
D

Name
GPA

studentB
RollNn
Name

Heap

GPA

#include <iostream>
using namespace std;
class Student{
char * name;
int rollNo;
public:
Student(char * aName, int arollNo){
name = new char[strlen(aName)+1];
strcpy(name,aName);

Virtual University of Pakistan

85

Object Oriented Programming (CS304)

VU

rollNo = arollNo;
}
Student(const Student &obj){
name = obj.name;
rollNo = obj.rollNo;

}
void showName(){
cout<<name<<endl;
}
~Student(){
delete []name;
}
};
int main()
{
Student studentA("AHMAD",1);
Student studentB = studentA;
/*copy constructor is implicitly called at this point*/
studentA.showName();
studentB.showName();
system("pause");
return 0;
}
Shallow copy works fine if our class doesnt include dynamic memory allocation but
in case of dynamic memory allocation it leads to dangling pointer problem as
explained below.
Problem is Shallow Copy
Student class data member name of char * type is added to store the name of student
and it is using dynamic memory according to the length of name entered by user for
student.
Student class data member name (char *) of object studentB is also pointing to
memory allocated for datamember name of object studentA, due to this there may be
two kinds of problems.
Suppose we delete first object studentA for some reason then its destructor will also
free memory allocated by it hence memory area containing name AHMAD will
also be freed and will be given to some other application by operating system, but

86

Virtual University of Pakistan

Object Oriented Programming (CS304)

VU

studentB member name is still pointing to that area so issue of Dangling Pointer
[Pointer pointing to incorrect memory location] will arose.
Same will happen if object studentB is deleted then studentA object data member
name will become dangling pointer. This has been explained below,
Let us change code in main to make our second object studentB in a new local scope
as shown below,
int main(){
Student studentA(Ahmad,1);
{
Student studentB = studentA;
}
}

Now if we will try to print the name of object studentA our program will not show
any output as name is pointing to some irrelevant memory address,

Irrelevant
address

studentA
RollNo
Name

Heap

GPA

Complete program code is given below,

#include <iostream>
using namespace std;
class Student{
char * name;
int rollNo;
public:
Student(char * aName, int arollNo){
name = new char[strlen(aName)+1];
strcpy(name,aName);
rollNo = arollNo;
}
Student(const Student &obj){
name = obj.name;

Virtual University of Pakistan

87

Object Oriented Programming (CS304)

VU

rollNo = obj.rollNo;
}
void showName(){
cout<<name<<endl;
}
~Student(){
delete []name;
}
};
int main()
{
Student studentA("AHMAD",1);
{
Student studentB = studentA;
/*copy constructor is implicitly called at this point*/
}
studentA.showName();
system("pause");
return 0;
}

Secondly if for some reason we change name of object studentA the value of object
studentB will also be changed as it pointing to same memory location.

#include <iostream>
using namespace std;
class Student{
char * name;
int rollNo;
public:
Student(char * aName, int arollNo){
name = new char[strlen(aName)+1];
strcpy(name,aName);
rollNo = arollNo;
}
Student(const Student &obj){
name = obj.name;
rollNo = obj.rollNo;
}

88

Virtual University of Pakistan

Object Oriented Programming (CS304)

VU

void setName(char * aName){

strcpy(name,aName);
}
void showName(){
cout<<name<<endl;
}
~Student(){
delete []name;
}
};
int main()
{
Student studentA("AHMAD",1);
Student studentB = studentA;
/*copy constructor is implicitly called at this point*/
studentA.showName();
studentB.showName();

studentA.setName("MOEEN");
studentA.showName();
studentB.showName();
system("pause");
return 0;
}
We resolve these two issues using deep copy.
09.2. Deep Copy
We write deep copy code in copy constructor so that when we create new object from
an existing object using copy constructor we also allocate new dynamic memory for
data members involving dynamic memory as shown below,
Student::Student( const Student & obj){
int len = strlen(obj.name);
name = new char[len+1]; // assignming new
/*dynamic memory to data member name of char * type for newly created object.*/
strcpy(name, obj.name);

Virtual University of Pakistan

89

Object Oriented Programming (CS304)

VU

//copy rest of the data members in the same way

}
Now we see what happens when we created objects in main as shown below,

int main(){
Student studentA(Ahmad,1);
{
Student studentB = studentA;
}
}

A
H
M
A
D

studentA
RollNo

A
H
M
A
D

Name
GPA

studentB
RollNo
Name
GPA

Now when we will execute code with object studentB in local scope our code still
works fine and shows name for object studentA as now deletion of object studentB
has no effect on object studentA as shown below,

A
H
M
A
D

studentA
RollNo
Name

Heap

GPA

Example

90

Virtual University of Pakistan

Object Oriented Programming (CS304)

VU

#include <iostream>
using namespace std;
class Student{
char * name;
int rollNo;
public:
Student(char * aName, int arollNo){
name = new char[strlen(aName)+1];
strcpy(name,aName);
rollNo = arollNo;
}
Student(const Student &obj){
name = new char[strlen(obj.name)+1];
strcpy(name,obj.name);
rollNo = obj.rollNo;

}
void showName(){
cout<<name<<endl;
}
~Student(){
delete []name;
}
};
int main()
{
Student studentA("AHMAD",1);
{
Student studentB = studentA;
/*copy constructor is implicitly called at this point*/
}
studentA.showName();
system("pause");
return 0;
}
09.3. Important points about copy constructor:

Virtual University of Pakistan

91

Object Oriented Programming (CS304)

VU

1. In case our class doesnt involve dynamic memory then default copy
constructor that performs shallow copy works fine.
2. In case our class has any data member involving dynamic memory we have
to write our own code in copy constructor to perform deep copy.
3. Copy constructor is normally used to perform deep copy
4. If we do not make a copy constructor then the compiler performs shallow
copy
5. Shallow copy performs bitwise copy.
09.4. Destructor
1. Destructor is used to free memory that is allocated through dynamic
allocation. We have to free memory allocated using new operator by over self
in destructor otherwise it remain occupied even after our program ends.
2. Destructor is used to perform house keeping operations.
3. Destructor is a function with the same name as that of class, but preceded
with a tilde ~
Example
class Student
{

public:
~Student(){
if(name){
delete []name;
}
}
}
Overloading
Destructors cannot be overloaded.
Sequence of Calls
Constructors and destructors are called automatically
Constructors are called in the sequence in which object is declared
Destructors are called in reverse order
Example
Student::Student(char * aName){

cout << aName << Constructor\n;

}
Student::~Student(){
cout << name << Destructor\n;
}
};
Example
int main()
{

92

Virtual University of Pakistan

Object Oriented Programming (CS304)

VU

Student studentB(Ali);
Student studentA(Ahmad);
return 0;
}
Example
Output:
Ali Constructor
Ahmad Constructor
Ahmad Destructor
Ali Destructor
09.5. Accessor Functions
In accordance to principle of information hiding data members of a class are declared
as private so that outside world can not access the private data of the object only an
interface is provided to outside world in the form of functions.
Accessor functions are also used to access private data of the object, we provide
accessor functions to get and set private data members of the class.
We also add error checking code in accessor functions to reduce errors so that object
doesnt move in illegal state.
Example Accessing Data Member
Example - Setter
class Student{

int rollNo;
public:
void setRollNo(int aRollNo){
rollNo = aRollNo;
}
};
Avoiding Error
void Student::setRollNo(int aRollNo){
if(aRollNo < 0){
rollNo = 0;
}
else
{
rollNo = aRollNo;
}
}
Example - Getter
class Student{

int rollNo;

Virtual University of Pakistan

93

Object Oriented Programming (CS304)

VU

public:
int getRollNo(){
return rollNo;
}
};
Good Practice:
Never return a handle to a data member from getter function because you are never
sure that function accessing the reference will not change the value of the variable.
09.6. this Pointer
Consider the code of a general class given below,
class Student{
int rollNo;
char *name;
float GPA;
public:
int getRollNo();
void setRollNo(int aRollNo);

};
The compiler reserves space for the functions defined in the class

Function Space
getRollNo(),

Space for data is not allocated (since no object is yet created)

this Pointer
Student s1, s2, s3;

94

Virtual University of Pakistan

Object Oriented Programming (CS304)

VU

s2(rollNo,)
Function Space
getRollNo(),
s3(rollNo,)

s1(rollNo,)

this Pointer

Function space is common for every variable

Whenever a new object is created:
o Memory is reserved for variables only
o Previously defined functions are used over and over again

Memory layout for objects created:

s1
rollNo,

s2
rollNo,

s3
rollNo,

s4
rollNo,

Function
Space
getRollNo(),

this Pointer

Address of each object is passed to the calling function.

This address is de-referenced by the functions and hence they act on correct
objects

Virtual University of Pakistan

95

Object Oriented Programming (CS304)

VU

s1
rollNo,

s2
rollNo,

s3
rollNo,

s4
rollNo,

address

address

address

address

Passing this Pointer

Whenever a function is called the this pointer is passed as a parameter to that

function.
Function with n parameters is actually called with n+1 parameters

Example
void Student::setName(char *)
is internally represented as,
void Student::setName(char *, const Student *)
Declaration of this
DataType * const this;
Compiler Generated Code

Student::Student(){
rollNo = 0;
}

96

Compiler generated code Student::Student(){

this->rollNo = 0;
}

Virtual University of Pakistan

Object Oriented Programming (CS304)

VU

Lecture No.10
10.1. Uses of this Pointer
o
o

There are situations where designer wants to return reference to

current object from a function
In such cases reference is taken from this pointer like (*this)

Example
Student Student::setRollNo(int aNo)
{

return *this;
}
Student Student::setName(char *aName)
{

return *this;
}

Usage:
int main()
{
Student aStudent;
Student bStudent;
bStudent = aStudent.setName(Ahmad);

bStudent = aStudent.setName(Ali).setRollNo(2);
return 0;
}

10.2. Separation of interface and implementation

Public member functions exposed by a class are called interface.
Separation of implementation from the interface is good software
engineering.
Benefits of separating interface and implementation:
o
o

Consider the example of following complex no. class, this complex no. class two
forms of implementations one is new and one is old implementation you can observe
that if you have separated interface and implementation then we can easily change
implementation without changing interface,
10.3. Complex Number

Virtual University of Pakistan

97

Object Oriented Programming (CS304)

VU

There are two representations of complex number

Euler form

z=x+iy
Phasor form
z = |z| (cos + i sin )
z is known as the complex modulus and is known as the complex argument or
phase
Example

Old implementation

New implementation

Complex

Complex
float z
float theta

float x
float y
float getX()
float getY()
void setNumber
(float i, float j)

float getX()
float getY()
void setNumber
(float i, float j)

Uml notation to show private data members

Example
class Complex{ //old
float x;
float y;
public:
void setNumber(float i, float j){
x = i;
y = j;
}

};
Example
class Complex{ //new
float z;
float theta;
public:
void setNumber(float i, float j){
theta = arctan(j/i);

98

Virtual University of Pakistan

Object Oriented Programming (CS304)

VU

};

Advantages
1. User is only concerned about ways of accessing data (interface)
2. User has no concern about the internal representation and implementation of
the class
Separation of interface and implementation
In c++ generally we can relate the concept of interface of a class to its header (.h) file
and and implementation of a class to its (.cpp) file. However it is not complete
separation of interface and implementation.

Usually functions are defined in implementation file (.cpp) while the class
definition is given in header file (.h)
Some authors also consider this as separation of interface and implementation

Student.h
class Student{
int rollNo;
public:
void setRollNo(int aRollNo);
int getRollNo();

};
Student.cpp
#include student.h
void Student::setRollNo(int aNo){

}
int Student::getRollNo(){

We only need to include header (.h) file in main.cpp to use the Student class as
shown below,

Main.cpp (main file to run the program)

#include student.h

Virtual University of Pakistan

99

Object Oriented Programming (CS304)

VU

int main(){
Student aStudent;
}
10.4. const Member Functions
Some functions in our programs are general purpose functions to show or access
data, they are supposed to do read only tasks only however there are chances that
they can change the state of data members of the class while accessing the data
members due to programming mistake, c++ provides the solution of this problem
using constant member functions.
We make those functions as constant who need only read only access (for example
such functions that will only display data or will return the value of data members).
When we make them constant compiler generates an error if these functions try to
change the value of data members of the class.
const Member Functions
Keyword const is placed at the end of the parameter list to make any function as
constant.
Declaration:
Inside class
class ClassName{
ReturnVal Function() const;
};
Definition:
Outside class
ReturnVal ClassName::Function() const{

}
Example
class Student{
public:
int getRollNo() const {
return rollNo;
}
};
const Functions

Constant member functions cannot modify the state of any object

They are just read-only
Errors due to typing are also caught at compile time

Example

100

Virtual University of Pakistan

Object Oriented Programming (CS304)

VU

Consider the function given below that is being used to check if roll no is equal to
entered value if in this function we replace comparison statemtn == with assignment
= statement it will compile correctly but whole code logic will change and we will get
un expected result,
bool Student::isRollNo(int aNo){
if(rollNo = = aNo){
return true;
}
return false;
}

Example
bool Student::isRollNo(int aNo){
/*undetected typing mistake*/
if(rollNo = aNo){
return true;
}
return false;
}

But if we have implemented as constant then compiler will catch this error and will
produce compile time error as shown below,
Example
bool Student::isRollNo
(int aNo)const{
/*compiler error*/
if(rollNo = aNo){
return true;
}
return false;
}
const Functions
Constructors and Destructors cannot be const because Constructors and
destructors are used to modify the object to a well defined state or to clean the
memory occupied by the object.
Example
class Time{
public:
Time() const {} //error
~Time() const {} //error
};

Virtual University of Pakistan

101

Object Oriented Programming (CS304)

VU

const Function

Constant member function cannot change data member

We cannot call non constant functions in constant functions because non
constant member functions may have code for changing state of the object
that is not allowed in the constant functions.

Example
class Student{
char * name;
public:
char *getName();
void setName(char * aName);
int ConstFunc() const{
name = getName(); //error
setName(Ahmad);//error
}
};
10.5. this Pointer and const Member Function
As we know that when a class function is called an implicit this pointer is passed to
tell the function about the object it has to operate same is true for constant function
with the difference that it will bbe passed as constant pointer to const data in case of
constant member functions so that this pointer can not be used now to change the
value of data members of the object,
const Student *const this;
// In case of constant member functions
instead of
Student * const this;
// In case of ordinary member functions

102

Virtual University of Pakistan

Object Oriented Programming (CS304)

VU

Lecture No.11
11.1. Usage example of Constant member functions
Problem:
Suppose we have requirement to change the class Student such that a student is
given a roll number when the object is created and cannot be changed afterwards our
existing class is given below,
Student Class

class Student{

int rollNo;
public:
Student(int aNo);
int getRollNo();

};

Solution of this problem:

We can do this by making rollNo constant so that cannot be changed once it is
defined as shown below,
Modified Student Class

class Student{

const int rollNo;

public:
Student(int aNo);
int getRollNo();

};

Now there is only one issue of initializing this roll no with initial value but the
problem is that we cannot set the value of roll no in constructor, as when code in
constructor is executed the data member roll no has already been created and when
we try to assign value to it in constructor compiler generates error,
Example

Student::Student(int aRollNo)
{
rollNo = aRollNo;
/*error: cannot modify a constant data member assignment statement not
initialization*/
}

Virtual University of Pakistan

103

Object Oriented Programming (CS304)

VU

Second solution is to write separate function but the problem remains same that we
cant assign value to constant data member,

void Student::SetRollNo(int i)
{
rollNo = i;
/*error: cannot modify a constant data member again assignment statement not
initialization */
}

We also know that we can only declare data members in structure or class but we
cannot initialize them at the time of declaration in structure or class _____ because
before executing constructor code, the class const member roll no has not got life it
will get life along with other class members when constructor will be invoked so we
can not assign any value to this constant member while declaring it. 6
Solution:
so what is the solution of this problem as we can not initialize constant members
while declaring them and we can not initialize them in constructor also because as
soon as they go life they become constant to solve this problem C++ gives us new
mechanism (syntax) for initialization of constant data members of the structure or
class to resolve above mentioned issues,
11.2. Difference between Initialization and Assignment:
Initialization is assigning value along with creation of variable.
int i = 2;
Assignment is assigning value after creation.
int i;
i = 7;

11.3. Member Initializer List

Member initialization list is used where we cannot modify the state of data members
in the member functions of the class including constructor,

A member initializer list is a mechanism to initialize data members

It is given after closing parenthesis of parameter list of constructor
In case of more than one member use comma separated list

Example
6

In c++ static const data members can be initialized in class or structure as well.

104

Virtual University of Pakistan

Object Oriented Programming (CS304)

VU

class Student{
const int rollNo;
char *name;
float GPA;
public:
Student(int aRollNo) : rollNo(aRollNo), name(Null), GPA(0.0){ //
initialization

};

Order of Initialization

Data member are initialized in order they are declared in the class
Order in member initializer list is not significant at all

Example

class ABC{
int x;
int y;
int z;
public:
ABC();
};
ABC::ABC():y(10),x(y),z(y)
{

}
/*
x = Junk value
y = 10
z = 10 */

11.4. const Objects

Objects can be declared constant with the use of const keyword

Constant objects cannot change their state

Example

int main()
{
const Student aStudent;

Virtual University of Pakistan

105

Object Oriented Programming (CS304)

VU

return 0;
}

Example

#include <cstdlib>
#include <iostream>

#include <cstdlib>
#include <iostream>

using namespace std;

using namespace std;

class Student{
int rollNo;
public:

class Student{
int rollNo;
public:
Student(int aRollNo) :
rollNo(aRollNo){

Student( ) {
}

}
int getRollNo(){
return rollNo;
}

int getRollNo(){
return rollNo;
}

};
};
int main(){
const Student aStudent;
int a = aStudent.getRollNo();

int main(){
const Student aStudent(5);
int a = aStudent.getRollNo();
//error
system("PAUSE");
return EXIT_SUCCESS;

//error
system("PAUSE");
return EXIT_SUCCESS;

const Objects
const objects can access only const member functions so chances of change of state of
const objects once they are created are eliminated.
We make getRollNo function constant so that we can access it using constant objects,
Example
class Student{

int rollNo;
public:

106

Virtual University of Pakistan

Object Oriented Programming (CS304)

VU

int getRollNo()const{
return rollNo;
}
};
int main(){
const Student aStudent;
int a = aStudent.getRollNo();
}

Constant member functions

Make all functions that dont change the state of the object constant
This will enable constant objects to access more member functions
11.5. Static Variables
Static variables of a class are such variables which are independent of class objects.
Lifetime of static variable is throughout the program life, if static variables are not
explicitly initialized then they are initialized to 0 of appropriate type.
Example
Static variable is initialized once only throughout the program, independent of how
many times the function initializing it is called,
void func1(int i){
static int staticInt = i;
//initialization statement will be executed once
//only as static variables are initialized once
cout << staticInt << endl;
}
int main(){
Output:
func1(1);
1
func1(2);
1
}

void func1(int i){

static int staticInt;
staticInt = i;
//assignment statement will be executed with each function call
cout << staticInt << endl;
}
int main(){
func1(1);
func1(2);

Output:
1
2

Virtual University of Pakistan

107

Object Oriented Programming (CS304)

VU

}
Static Data Member
Definition
A variable that is part of a class, yet is not part of any object of that class, is called
static data member
Static Data Member
They are shared by all instances (objects) of the class
They do not belong to any particular instance of a class
Class vs. Instance Variable
Suppose we created three objects of student class as shown below,
Student s1, s2, s3;

Class Variable

Instance Variable

s2(rollNo,)
Class Space

s3(rollNo,)

s1(rollNo,)

Static Data Member (Syntax)

Keyword static is used to make a data member static
class ClassName{

static DataType VariableName;

};
Defining Static Data Member
Static data member is declared inside the class
But they are defined outside the class
Defining Static Data Member

108

Virtual University of Pakistan

Object Oriented Programming (CS304)

VU

class ClassName{

static DataType VariableName;

};
DataType ClassName::VariableName;
Initializing Static Data Member
Static data members should be initialized once at file scope
They are initialized at the time of definition
Example
class Student{
private:
static int noOfStudents;
public:

};
int Student::noOfStudents = 0;
/*private static member cannot be accessed outside the class except for
initialization*/

Initializing Static Data Member

If static data members are not explicitly initialized at the time of definition then they
are initialized to 0
Example
int Student::noOfStudents;
is equivalent to
int Student::noOfStudents=0;

Virtual University of Pakistan

109

Object Oriented Programming (CS304)

VU

Lecture No.12
Review:
Static Data Member
Definition
A variable that is part of a class, yet is not part of an object of that class, is called
static data member
Static Data members are shared by all instances of the class and they do not belong to
any particular instance of a class.
Class vs. Instance Variable
Memory for static variables in allocated in class space whereas for instance variables
it is separate for each object as shown below, if we have class Student as given below,
class Student{
private:
static int noOfStudents;
public:

};
When we will create objects of Student as s1, s2, s3 then memory will be allocated as
given below,

Static Data Member (Syntax)

Keyword static is used to make a data member static,
class ClassName{

static DataType VariableName;

};
Defining Static Data Member (allocating memory for them)
Static data member is declared inside the class
But they are defined outside the class,
class ClassName{

static DataType VariableName;

};
DataType ClassName::VariableName;
Initializing Static Data Member(assigning them some initial value)
Static data members should be initialized once at file scope

110

Virtual University of Pakistan

Object Oriented Programming (CS304)

VU

They are initialized at the time of definition,

Example
class Student{
private:
static int noOfStudents;
public:

};
int Student::noOfStudents = 0;
/*private static member cannot be accessed outside the class except for
initialization*/
Initializing Static Data Member
If static data members are not explicitly initialized at the time of definition then they
are initialized to 0
Example
int Student::noOfStudents;
is equivalent to
int Student::noOfStudents=0;
12.1. Accessing Static Data Member
To access a static data member there are two ways
Access like a normal data member (using dot operator .)
Access using a scope resolution operator ::
Example
class Student{
public:
static int noOfStudents;
};
int Student::noOfStudents;
int main(){
Student aStudent;
aStudent.noOfStudents = 1;
Student::noOfStudents = 1;
}
12.2. Life of Static Data Member
They are created even when there is no object of a class
They remain in memory even when all Objects of a class are destroyed
Example
class Student{
public:

Virtual University of Pakistan

111

Object Oriented Programming (CS304)

VU

static int noOfStudents;

};
int Student::noOfStudents;
int main(){
Student::noOfStudents = 1;
}

Example
class Student{
public:
static int noOfStudents;
};
int Student::noOfStudents;
int main(){
{
Student aStudent;
aStudent.noOfStudents = 1;
}
Student::noOfStudents = 1;
}

Uses
They can be used to store information that is required by all objects, like global
variables
Example
Modify the class Student such that one can know the number of student created in a
system
class Student{

public:
static int noOfStudents;
Student();
~Student();

};
int Student::noOfStudents = 0;
Student::Student(){
noOfStudents++;
}
Student::~Student(){
noOfStudents--;
}
int Student::noOfStudents = 0;
int main(){
cout <<Student::noOfStudents <<endl;
Student studentA;

112

Virtual University of Pakistan

Object Oriented Programming (CS304)

VU

cout <<Student::noOfStudents <<endl;

Student studentB;
cout <<Student::noOfStudents <<endl;
}

Output:
0
1
2
Problem
noOfStudents is accessible outside the class
Bad design as the local data member is kept public
The solution is that we write static member function to access static members,
12.3. Static Member Function
Definition:
The function that needs access to the members of a class, yet does not need to be
invoked by a particular object, is called static member function

They are used to access static data members

Access mechanism for static member functions is same as that of static data
members
They cannot access any non-static members

Example

class Student{
static int noOfStudents;
int rollNo;
public:
static int getTotalStudent(){
return noOfStudents;
}
};
int main(){
int i = Student::getTotalStudents();
}
Accessing non static data members

int Student::getTotalStudents(){
return rollNo;
}
int main(){
int i = Student::getTotalStudents();
/*Error: There is no instance of Student, rollNo cannot be accessed*/
}

Virtual University of Pakistan

113

Object Oriented Programming (CS304)

VU

12.4. this Pointer and static member functions

this pointer is passed implicitly to member functions

this pointer is not passed to static member functions
Reason is static member functions cannot access non static data members

12.5. Global Variable vs. Static Members

Alternative to static member is to use global variable
Global variables are accessible to all entities of the program
User of Global variables is against the principle of information hiding.
12.6. Array of Objects
Array of objects can only be created if an object can be created without
supplying an explicit initializer
There must always be a default constructor if we want to create array of
objects
Example

class Test{
public:
};
int main(){
Test array[2]; // OK
}

class Test{
public:
Test();
};
int main(){
Test array[2]; // OK
}
class Test{
public:
Test(int i);
};
int main(){
Test array[2]; // Error
}
class Test{
public:
Test(int i);
}
int main(){
Test array[2] = {Test(0),Test(0)};
}

114

Virtual University of Pakistan

Object Oriented Programming (CS304)

VU

class Test{
public:
Test(int i);
}
int main(){
Test a(1),b(2);
Test array[2] = {a,b};
}

Virtual University of Pakistan

115

Object Oriented Programming (CS304)

Lecture No.13
13.1. Pointer to Objects
Pointer to objects are similar as pointer to built-in types
They can also be used to dynamically allocate objects
Example
class Student{

public:
Student();
Student(char * aName);
void setRollNo(int aNo);
};
Example
int main(){
Student obj;
Student *ptr;
ptr = &obj;
ptr->setRollNo(10);
return 0;
}
Allocation with new Operator
new operator can be used to create objects at runtime
Example
int main(){
Student *ptr;
ptr = new Student;
ptr->setRollNo(10);
return 0;
}
Example
int main(){
Student *ptr;
ptr = new Student(Ali);
ptr->setRollNo(10);
return 0;
}
Example
int main()
{
Student *ptr = new Student[100];
for(int i = 0; i < 100;i++)
{
ptr->setRollNo(10);
}
return 0;
}

116

Virtual University of Pakistan

VU

Object Oriented Programming (CS304)

VU

13.2. Breakup of new Operation

new operator is decomposed as follows
Allocating space in memory
Calling the appropriate constructor
13.3. Case Study
Design a class date through which user must be able to perform following operations
Get and set current day, month and year
Increment by x number of days, months and year
Set default date
Attributes
Attributes that can be seen in this problem statement are
Day
Month
Year
Default date
Attributes
The default date is a feature shared by all objects
This attribute must be declared a static member
Attributes in Date.h
class Date
{
int day;
int month;
int year;
static Date defaultDate;

};
Interfaces
getDay
getMonth
getYear
setDay
setMonth
setYear
addDay
addMonth
addYear
setDefaultDate
Interfaces
As the default date is a static member the interface setDefaultDate should also be
declared static
Interfaces in Date.h
class Date{

public:
void setDay(int aDay);
int getDay() const;
void addDay(int x);

Virtual University of Pakistan

117

Object Oriented Programming (CS304)

};
Interfaces in Date.h
class Date{

public:
static void setDefaultDate(
int aDay,int aMonth, int aYear);

};
Constructors and Destructors in Date.h
Date(int aDay = 0,
int aMonth= 0, int aYear= 0);
~Date(); //Destructor
};
Implementation of Date Class
The static member variables must be initialized
Date Date::defaultDate (07,3,2005);
Constructors
Date::Date(int aDay, int aMonth,
int aYear) {
if(aDay==0) {
this->day = defaultDate.day;
}
else{
setDay(aDay);
}
//similarly for other members
}
Destructor
We are not required to do any house keeping chores in destructor
Date::~Date
{
}
Getter and Setter
void Date::setMonth(int a){
if(a > 0 && a <= 12){
month = a;
}
int getMonth() const{
return month;
}
addYear
void Date::addYear(int x){
year += x;
if(day == 29 && month == 2
&&
!leapyear(year)){
day = 1;

118

Virtual University of Pakistan

VU

Object Oriented Programming (CS304)

VU

month = 3;
}
}
Helper Function
class Date{

private:
bool leapYear(int x) const;

};
Helper Function
bool Date::leapYear(int x) const{
if((x%4 == 0 && x%100 != 0)
||
(x%400==0)){
return true;
}
return false;
}
setDefaultDate
void Date::setDefaultDate(
int d, int m, int y){
if(d >= 0 && d <= 31){
day = d;
}

Virtual University of Pakistan

119

Object Oriented Programming (CS304)

VU

Lecture No.14
14.1. Composition
Consider the following implementation of the student class we discussed in previous
lectures,

Student
gpa : float
rollNo : int
name : char *
Student(char * = NULL, int = 0,
float = 0.0);
Student(const Student &)
GetName() const : const char
*
SetName(char *) : void
~Student()

Composition
If one object is part of another object (relationship of part and whole) in composition
lifetime of one object depends upon the other. The part objects are essential
components of the whole.
For example person is composed of hands, eyes, feet so on.
In student class we assigning dynamic memory for variable name using new
operator as shown,
class Student{
private:
float gpa;
char * name;
int rollNumber;
public:
Student(char * = NULL, int = 0, float = 0.0);
Student(const Student & st);
const char * GetName() const;
// never return handle of private data members or private member functions
~Student();

};
Student::Student(char * _name, int roll, float g)
{
cout << "Constructor::Student..\n";

120

Virtual University of Pakistan

Object Oriented Programming (CS304)

VU

if (!_name){
name = new char[strlen(_name)+1];
strcpy(name,_name);
}
else name = NULL;
rollNumber = roll;
gpa = g;
}

Student::Student(const Student & st){

if(str.name != NULL){
name = new char[strlen(st.name) + 1];
strcpy(name, st.name);
}
else name = NULL;
rollNumber = st.roll;
gpa = st.g;
}

const char * Student::GetName(){

return name;
}
// never return handle of private data members or private member functions const
ensures that private data members will not be changed

Student::~Student(){
delete [] name; // deleting name array
}

In C++ it is all about code reuse

Composition is Creating objects of one class inside another class

Has a relationship:
Bird has a beak
Student has a name
Composition
Now we change code slightly replacing name char * by String so that it is whole
object of class String as it qualifies to be an object because we have to apply many
operations on it like string dynamic creation and deletion, string copy using deep
copy, searching a substring and so on.
Conceptual notation:

Virtual University of Pakistan

121

Object Oriented Programming (CS304)

VU

Student
gpa : float
rollNo : int
name : String

String
string : char *

Student(char * = NULL, int = 0,

float = 0.0);
Student(const Student &)
GetName() const : String
GetNamePtr() const : const
char *
SetName(char *) : void
~Student()

String()
SetString(char *) : void
GetString() const : const
char *
~String()

Composition
Now we see string class code to see how it simplifies original Student object and how
we have used composition here,
class String{
private:
char * ptr;
public:
String(); // default constructor
String(const String &); // copy constructor
void SetString(char *); // setter function
const char * GetString() const;
// getter function returning const pointer to data member ptr
~String()

};
String::String(){
cout << "Constructor::String..\n";
ptr = NULL;
}
String::String(const String & str){
if(str.ptr != NULL){
string = new char[strlen(str.ptr)+1];
strcpy(ptr, str.ptr);
}
else ptr = NULL;
}
void String::SetString(char * str){
if(ptr != NULL){
delete [] ptr;

122

Virtual University of Pakistan

Object Oriented Programming (CS304)

VU

ptr = NULL;
}
if(str != NULL){
ptr = new char[strlen(str)+1];
strcpy(ptr, str);
}
}
issue of memory leakage (inaccessible memory)
if we simply set pointer here
memory will be outside of our object many cause problems later
user still has pointer of passed value it can itself modify it
We resolve these two issues in SetString by allocating new memory and deleting
previous memory.
const char * String::GetString()const{
return ptr;
}
String::~String(){
delete [] ptr;
cout <<"Destructor::String..\n";
}
Make pointer equal to NULL as well any where you delete dynamic memory.

Now consider the code of Student class again, now by adding composed string object
our code has been simplified very much, (we will use methods of composed object
simply by calling them where needed)
class Student{
private:
float gpa;
int rollNumber;
String name;
public:
Student(char* =NULL, int=0,float=0.0);
Student(const Student &);
void SetName(const char *);
String GetName() const;
const char * GetNamePtr const();
~Student();

};
Student Student(char * _name,
int roll, float g){
cout <<"Constructor::Student..\n";
name.SetString(_name);

Virtual University of Pakistan

123

Object Oriented Programming (CS304)

VU

rollNumber = roll;
gpa = g;
}
Student::Student(const Student & s){
name.Setname(s.name.GetString());
gpa = s.gpa;
rollNo = s.rollNo;
}
Explanation:
1. name.Setname(s.name.GetString());
// setting composed name of newly created object
2. name.Setname(s.name.GetString());
//accessing the composed object string name of object to be copied
3. name.Setname(s.name.GetString());
//accessing the value of composed object string name by calling its member
function GetString
4. name.Setname(s.name.GetString());
//overall result : the value of composed object string of object to be copied
will be copied to newly created object composed object string.

const char * Student::GetNamePtr() const{

return name.GetString();
}
void Student::SetName(const char * n){
name.SetString(n);
}
Student::~Student(){
cout <<"Destructor::Student..\n";
}
void main(){
Student aStudent("Fakhir", 899, 3.1);
cout << endl;
cout << Name: << aStudent.GetNamePtr() << \n;
}

Output:

Constructor::String..
Constructor::Student..

124

Virtual University of Pakistan

Object Oriented Programming (CS304)

VU

Name: Fakhir
Destructor::Student..
Destructor::String..

Important Points:
1. We can access methods of composed object in the same way as we can access
methods of other objects.
Name of composed object.MemberFunction
2. Member functions of a class can access its private data members like,
Student::Student(const Student & s){
name.Setname(s.name.GetString());
// accessing private member String name of student using its object s
and then accessing String name member function GetString to access
string value two methods calss in one line
gpa = s.gpa;
// accessing private members of student in student member function

rollNo = s.rollNo;
// accessing private members of student in student member function
}

Constructors & Composition

Constructors of the sub-objects are always executed before the constructors of the
master class
Example:
As you see the example output of program given above,
Output:

Constructor::String..
Constructor::Student..
Name: Fakhir
Virtual University of Pakistan

125

Object Oriented Programming (CS304)

VU

Destructor::Student..
Destructor::String..

Constructor for the sub-object name is executed before the constructor of Student
and destructor of sub-object is called after destructor of student. It is logical as
composing object has to contain composed object so composed object should be
created first and then composing object. Similarly while destructing objects we
composing object is destructed first and then composed object as shown in diagram
below,

Composed
object
String

Student
Composing
object

Composition

Constructor calling:
Constructors are called from composed objects to composing objects.

String

126

Constructor of composed object

Virtual University of Pakistan

Object Oriented Programming (CS304)

VU

String

Student
Constructor of
composing object

Destructor calling:
Destructors are called from composing objects to composed objects.

String

Student
Destructor of
composing object

String

Destructor of composed object

Virtual University of Pakistan

127

Object Oriented Programming (CS304)

VU

Lecture No.15

Composition:
We saw composition in last lecture, its Conceptual notation is given below,
Student
gpa : float
rollNo : int
name : String

String

Student(char * = NULL, int =

0, float = 0.0);
Student(const Student &)
GetName() const : String
GetNamePtr() const : const
char *
SetName(char *) : void
~Student()

string : char *
String()
SetString(char *) : void
GetString() const : const
char *
~String()

We created student object in main as by passing name, rollno and gpa and then
displayed the name of student using GetNamePtr member function of student
class,
int main(){
Student aStudent("Fakhir", 899,3.1);
cout << endl;
cout << Name: << aStudent.GetNamePtr()<< endl;
return 0;
}

Output:
The output of our code is given below,
Constructor::String..
Constructor::Student..
Name: Fakhir
Destructor::Student..
Destructor::String..
Constructor Code:
Let us see the constructor code again,
Student::Student(char * n, int roll, float g){
cout <<"Constructor::Student..\n";
name.SetString(n);
rollNumber = roll;
gpa = g;
}

128

Virtual University of Pakistan

Object Oriented Programming (CS304)

VU

In this code we are setting string data member name of Student class using SetString
but the problem in this approach is that we have to call SetString method explicitly to
set value in string class, the reason for it is that our String class doesnt support
setting its value while creating its object this is the reason we have to use the function
SetString in the constructor.
This is an overhead and also not very good way to set any object value, we want to
initialize our string sub-object name in the student class as we initialize other objects
using constructor. For achieving this functionality we add an overloaded constructor
in the String class that takes char string as parameter and initialize the String class
object with this value using the Student constructors Member initialization list as
shown below in bold text:
class String{
char *ptr;
public:
String();
String(char *); // constructor with char * as parameter
String(const String &);
void SetName(char *);
~String();

};

String::String(char * str){
if(str != NULL){
ptr = new char[strlen(str)+1];
strcpy(ptr, str);
}
else ptr = NULL;
cout << "Overloaded Constructor::String..\n";
}

Now Student class constructor code is modified as follows:

class Student{
private:
float gpa;
int rollNumber;
String name;
public:

Student(char *=NULL, int=0, float=0.0);

};

Virtual University of Pakistan

129

Object Oriented Programming (CS304)

VU

Student::Student(char * n,int roll, float g): name(n) {

cout << "Constructor::Student..\n";
rollNumber = roll;
gpa = g;
}
int main(){
Student aStudent("Fakhir", 899, 3.1);
cout << endl;
cout << Name: << aStudent.GetNamePtr()
return 0;
}

<< endl;

Output:
Overloaded Constructor::String..
Constructor::Student..
Name: Fakhir
Destructor::Student..
Destructor::String..

Now suppose we want to add date object in student class to store student Birth Date,
the conceptual diagram will be as given below,

130

Virtual University of Pakistan

Object Oriented Programming (CS304)

VU

String
name: char *

Student

String()
String(char *)
~String()

name : String
birthDate : Date
Student()
Student( char *,
const Date &, int,
float)
SetName(char *) : void
GetName() : char *
~Student()

Date
day: int
Month: int
year: int

Date()
Date(int,int,int)
Date(const Date &)

Student class is modified as follows:

class Student{
private:

Date birthDate;
String name;
public:
Student(char *, const Date &, int, float);
~Student();

};
Composition
Student::Student(char * n, const Date & d,
name(n),birthDate(d){
cout << "Constructor::Student..\n";
rollNumber = roll;
gpa = g;
}

int roll, flaot g):

Student::~Student(){
cout << "Destructor::Student..\n";
}
int main(){
Date _date(31, 12, 1982);
Student aStudent("Fakhir", _date,899,3.5);

Virtual University of Pakistan

131

Object Oriented Programming (CS304)

VU

return 0;
}
Output:
Overloaded Constructor::Date..
Copy Constructor::Date..
Overloaded Constructor::String..
Constructor::Student..
Destructor::Student..
Destructor::String..
Destructor::Date..
Destructor::Date..

15.1. Aggregation
In composition we made separate object of those concepts that we think were worthy
to be implemented as an object within other object to make our code simpler and to
make functionality modular (divided in parts) and understandable like we made
String class in Student class, but in real life most situations are such that two distinct
(different) objects and one object is using services of the other one like student and
teacher, student and librarian, room and chair, passenger and bus, book and
bookshelf, person and computer so on. In this case we cannot make one object as part
of other object because they exist independently and only provide services to each
other like in case of,
Student and Teacher: Student or Teacher object cannot be composed of other one yet
they are taking services of each other.
Passenger and Bus: Passenger and Bus are taking services of each other but exist
standalone also, bus includes passengers but passenger are not Part of Bus they can
exist independently as well.
Composition vs. Aggregation
Aggregation is a weak relationship than composition because in this relationship two
classes get services of each other but can exist independently as well, main difference
is memory organization of two objects as shown below,

132

Virtual University of Pakistan

Object Oriented Programming (CS304)

VU

Object 1
Object 2

Composition

Object 1
Object 2

Aggregation

Example:
Take the example of Room and Chair as given below,

Room

Chair

area : float
chairs[50]:Chair *

Room(char *, int)
~Room()
FoldChair(int) : bool

Chair()
DoSomething() : void
FoldChair() : bool
UnFoldChair() : bool
~Chair()

Virtual University of Pakistan

133

Object Oriented Programming (CS304)

VU

Aggregation c++ implementation:

In aggregation, a pointer or reference to an object is created inside a class. The subobject has a life that is NOT dependant on the life of its master class.
e.g

Chairs can be moved inside or outside at anytime

When Room is destroyed, the chairs may or may not be destroyed

Aggregation:
class Room{
private:
float area;
Chair * chairs[50];
Public:
Room();
void AddChair(Chair *, int chairNo);
Chair * GetChair(int chairNo);
bool FoldChair(int chairNo);

};
Room::Room(){
for(int i = 0; i < 50; i++)
chairs[i] = NULL;
}
void Room::AddChair(Chair * chair1, int chairNo){
if(chairNo >= 0 && chairNo < 50)
chairs[chairNo] = chair1;
}
Chair * Room::GetChair(int chairNo){
if(chairNo >= 0 && chairNo < 50)
return chairs[chairNo];
else
return NULL;
}
bool Room::FoldChair(int chairNo){
if(chairNo >= 0 && chairNo < 50)
return chairs[chairNo]->FoldChair();
else
return false;
}
int main(){
Chair ch1;
{
Room r1;
r1.AddChair(&ch1, 1);

134

Virtual University of Pakistan

Object Oriented Programming (CS304)

VU

r1.FoldChair(1);
}
ch1.UnFoldChair(1);
return 0;
}
15.2. Friend Functions
The functions which are not member functions of the class yet they can access all
private members of the class are called friend functions.
Why they are needed
They are needed in situations where we have written code for some function in one
class and it need to be used by other classes as well for example,
Suppose we wrote the code to compute a complex mathematical formulae in one
class but later it was required by other classes as well, in that case we will make that
function friend of all other classes.
Are friend functions against the concept of Object Oriented Programming?
It can be said that friend functions are against the principle of object oriented
programming because they violate the principle of encapsulation which clearly says
that each object methods and functions should be encapsulated in it. But there we are
making our private member accessible to other outside functions.

Consider the following class:

class X{
private:
int a, b;
public:
void MemberFunction();

}
Suppose we have a global function DoSomething that need to access the private
members of class X, when we will try to access them compiler will generate error as
outside world can not access private members of a class except its member functions.
void DoSomething(X obj){
obj.a = 3; //Error
obj.b = 4; //Error
}
Friend Functions
In order to access the member variables of the class, we must make function friend of
that class,

Virtual University of Pakistan

135

Object Oriented Programming (CS304)

class X{
private:
int a, b;
public:

friend void DoSomething(X obj);

}
Now the function DoSomething can access data members of class X
void DoSomething(X obj){
obj.a = 3;
obj.b = 4;
}
Friend Functions
Prototypes of friend functions appear in the class definition.
But friend functions are NOT member functions.
Friend Functions
Friend functions can be placed anywhere in the class without any effect
Access specifiers dont affect friend functions or classes
class X{
...
private:
friend void DoSomething(X);
public:
friend void DoAnything(X);
...
};
Friend Functions
While the definition of the friend function is:
void DoSomething(X obj){
obj.a = 3;
obj.b = 4;

// No Error
// No Error

}
friend keyword is not given in definition.
Friend Functions
If keyword friend is used in the function definition, its a syntax error

136

Virtual University of Pakistan

VU

Object Oriented Programming (CS304)

VU

//Error
friend void DoSomething(X obj){

}
Friend Classes
Similarly, one class can also be made friend of another class:
class X{
friend class Y;

};
Member functions of class Y can access private data members of class X
class X{
friend class Y;
private:
int x_var1, x_var2;
...
};
class Y{
private:
int y_var1, y_var2;
X objX;
public:
void setX(){
objX.x_var1 = 1;
}
};
int main(){
Y objY;
objY.setX();
return 0;
}

Virtual University of Pakistan

137

Object Oriented Programming (CS304)

VU

Lecture No.16
16.1. Operator overloading
Consider the following class,
class Complex{
private:
double real, img;
public:
Complex Add(const Complex &);
Complex Subtract(const Complex &);
Complex Multiply(const Complex &);

}
We want to write function to add two complex no. objects, the Add function
implementation to add two complex numbers is shown below, this function is taking
one complex no object and its adding the current (with reference to which it will be
called) and is returning result in new object,
Complex Complex::Add(
const Complex & c1){
Complex t;
t.real = real + c1.real;
t.img = img + c1.img;
return t;
}
// adds the contents of c2 to c1
// creating new object c3 and assigning it result of c1+c2
Now we can add two complex no. objects using the following statement,
Complex c3 = c1.Add(c2);
In this statement two operations are taking place,
1. One is addition of two objects using the function call Add and returning the
result in a new object.
Complex c3 = c1.Add(c2);
2. Second is copy of that temporary object to newly created object c3 using copy
constructor.
Complex c3 = c1.Add(c2);

But there are two issues with this implementation,

1. We cant add two complex no. objects by simple writing + as we can add
basic data types like int or float as shown below,
int a = 3;

138

Virtual University of Pakistan

Object Oriented Programming (CS304)

VU

int b = 5;
c = a + b; // correct
Complex c1(2,3), c2(4,5);
Complex c3 = c1 + c2; // error
Instead we have to explicitly write,
Complex c3 = c1.Add(c2)
If we give our complex no. class code some user in compiled form for use,
user will need to know how we have written Add function (no. of parameters
, return type) to add two complex no. objects so that he can call Add function
correctly.
2. If we want to perform add operation on more than two objects in a single
mathematical statement like:
c1+c2+c3+c4
We are unable to do it.
We have to explicitly write,
c1.Add(c2.Add(c3.Add(c4)))
Alternative way is:
t1 = c3.Add(c4);
t2 = c2.Add(t1);
t3 = c1.Add(t2);
This is also overhead, especially if the mathematical expression is large,
Converting it to C++ code will involve complicated mixture of function calls
Code will become less readable
Chances of human mistakes will become very high
Code produced will be very hard to maintain

The solution to this problem is simple that we can write normal operators like +,-,*,
and so on for our user defined classes as well,
It is Operator overloading
Using operator overloading we can perform basic operations (like addition,
subtraction, multiplication, division and so on) on our own defined classes objects
in the similar way as we perform them on basic built-in types(like int, float, long,
double etc.).
C++ allows us to overload common operators like +, - or * etc
With operator overloading Mathematical statements dont have to be explicitly
converted into function calls as we had to do to add two complex no objects using
function call Add.

Virtual University of Pakistan

139

Object Oriented Programming (CS304)

VU

Operator overloading
Assume that operator + has been overloaded then actual C++ code becomes:
c1+c2+c3+c4
The resultant code is very easy to read, write and maintain
Operator overloading
C++ automatically overloads operators for pre-defined types as these have also been
implemented as classes by c++.
Example of predefined types:
int
float
double
char
long
Operator overloading
float x;
int y;
x = 102.02 + 0.09; // overloaded operator + for float type will be called by c++
Y = 50 + 47; // overloaded operator + for int type will be called by c++
The compiler probably calls the correct overloaded low level function for addition
i.e:
// for integer addition:
Add(int a, int b)
// for float addition:
Add(float a, float b)
Operator overloading
Operator functions are not usually called directly, they are automatically invoked to
evaluate the operations they implement by compiler.
List of operators that can be overloaded in C++:

List of operators that cant be overloaded:

140

Virtual University of Pakistan

Object Oriented Programming (CS304)

VU

Reason: They take actual current object name, rather than value in their argument as
you have seen previously in the use of dot (.) operator,
Student std;
int roll = std.getRollNo() // dot operator is performing on actual function
(getRollNo) of class Student that will vary from program to program.
?: is the only ternary operator in C++ and cant be overloaded.
The precedence of an operator:
The precedence of an operator is order of evaluation which operator will be
evaluated first in expression.
The precedence of an operator is NOT affected due to overloading.
Example:
c1*c2+c3
c3+c2*c1
In both lines multiplication * will be done first and then addition.
Associativity:
Associativity is NOT changed due to overloading
Following arithmetic expression always is evaluated from left to right:
c1 + c2 + c3 + c4

Unary operators and assignment operator are right associative, e.g:

a=b=c is same as a=(b=c)
All other operators are left associative:
c1+c2+c3 is same as
(c1+c2)+c3
Important things to consider:
Always write code representing the operator for example adding subtraction code
inside the + operator will create chaos.
Creating a new operator is a syntax error (whether unary, binary or ternary), you
cannot create $.
Arity of Operators
Arity (no of operands it works on) of an operator is NOT affected by overloading
Example:
Division operator will take exactly two operands in any case:
b=c/d
General syntax of Operators Overloading:

Virtual University of Pakistan

141

Object Oriented Programming (CS304)

VU

In case of member functions of a class:

return_type class_name::operator operator_symbol( parameters ){
/*code*/
}
In case of non member functions of a class (in this case we will make overloaded
operator function as friend function):
return_type operator operator_symbol( parameters ){
/*code*/
}
For example:
Complex& Complex::operator + (const Complex & c){
/*code*/
}
Complex& operator + (const Complex & c){
/*code*/
}
Binary Operators Overloading:
Binary operators act on two quantities.
Examples of binary operators:

General syntax of Binary Operators Overloading:

In case of member function of a class:
TYPE class_name::operator operator_symbol( TYPE rhs ){
/*code*/
}
In case of non-member function of a class:
TYPE class_name::operator operator_symbol( TYPE rhs1, TYPE rhs2 ){
/*code*/
}

142

Virtual University of Pakistan

Object Oriented Programming (CS304)

VU

The operator OP must have at least one formal parameter of type class (user
defined type)
Following is an error:
int operator + (int, int);
Examples:
Overloading + operator:
class Complex{
private:
double real, img;
public:

Complex operator +(const Complex & rhs);

};
Complex Complex::operator +( const Complex & rhs){
Complex t;
t.real = real + rhs.real;
t.img = img + rhs.img;
return t;
}
The return type is Complex so as to facilitate complex statements like:
Complex t = c1 + c2 + c3;
The above statement is automatically converted by the compiler into appropriate
function calls:
(c1.operator +(c2)).operator +(c3);
If the return type was void,
class Complex{
...
public:
void operator+(
const Complex & rhs);
};
void Complex::operator+(const Complex & rhs){
real = real + rhs.real;
img = img + rhs.img;
};
We have to do the same operation c1+c2+c3 as:
c1+c2
c1+c3
// final result is stored in c1
Drawbacks of void return type:

Virtual University of Pakistan

143

Object Oriented Programming (CS304)

Assignments and cascaded expressions are not possible

One of the existing objects is used to store result
Code is less readable
Debugging is tough
Code is very hard to maintain

144

Virtual University of Pakistan

VU

Object Oriented Programming (CS304)

VU

Lecture No.17
Binary operators (cont.)
The binary operator is always called with reference to the left hand argument.
Example:
In c1+c2,
c1.operator+(c2)
// c1 is calling overloaded + operator and c2 is being passed as
// reference in that function.
In c2+c1,
c2.operator+(c1)
// c2 is calling overloaded + operator and c1 is being passed as
// reference in that function.
Adding basic data type to complex number class:
The overloading code we discussed before for complex no. class can add two
complex number objects but it can not handle the following situation:
Complex c1;
c1 + 2.325 (as we are adding basic data type double to complex no.)
To do this, we have to modify the Complex class.
Modifying the complex class:
class Complex{
...
Complex operator+(const Complex & rhs);
Complex operator+(const double& rhs);
};
Complex operator + (const double& rhs){
Complex t;
t.real = real + rhs;
t.img = img;
return t;
}
Now we can write both forms of statements,
Complex c2, c3;
Complex c1 = c2 + c3;
Complex c4 = c2 + 235.01;
But problem arises if we do the following:
Complex c5 = 450.120 + c1;
The + operator is called with reference to 450.120

Virtual University of Pakistan

145

Object Oriented Programming (CS304)

VU

No predefined overloaded + operator is there that takes Complex as an argument

Now if we write the following two functions to the class, we can add a Complex to a
real or vice versa :
Class Complex{

friend Complex operator + (const Complex & lhs, const double & rhs);
friend Complex operator + (const double & lhs, const Complex & rhs);
}
We have made them as friend so that we can write them as non member functions
and they are not called with respect to complex no. class object instead we pass both
arguments (complex no. object and double value) to this function compiler invoke
them according to arguments passed. Their implementation is similar,
Complex operator +(const Complex & lhs, const double& rhs){
Complex t;
t.real = lhs.real + rhs;
t.img = lhs.img;
return t;
}
Complex operator + (const double & lhs, const Complex & rhs){
Complex t;
t.real = lhs + rhs.real;
t.img = rhs.img;
return t;
}
Binary operators
So adding three overloaded versions for + operator as shown below allow us to write
code to,
Add two complex objects
Add complex object and a double value.
Class Complex{

Complex operator + (const Complex &);

friend Complex operator + (const Complex &, const double &);
friend Complex operator + (const double &, const Complex &);
};
Non members which are not friend call also achieve this functionality but in that case
we need extra four functions two getters and two setters each for real and imag part.
Compiler searches overloaded operator code in member function first and then in
non member functions.

146

Virtual University of Pakistan

Object Oriented Programming (CS304)

VU

Binary operators
Other binary operators are overloaded in similar to the + operator as demonstrated
in the above examples
Example:
Complex operator * (const Complex & c1, const Complex & c2);
Complex operator / (const Complex & c1, const Complex & c2);
Complex operator - (const Complex & c1, const Complex & c2);
17.1. Overloading Assignment operator
As we know compiler can generate the following three functions for a class if
required on its own,

Default Constructor (in case we have not written any other constructor for a
class)
Copy Constructor
Assignment Operator

However as we discussed previously if our class has any data member using
dynamic memory then we had to write our own code for default constructor, copy
constructor and similarly assignment operator as compiler generated version of these
functions performs shallow copy that creates dangling pointer, and memory leakage
issues in case of dynamic memory allocation.
We have already seen code of default constructor and copy constructor and the code
for overloaded assignment operator is similar,
Consider the string class:
class String{
int size;
char * bufferPtr;
public:
String(); // default constructor
String(char *); // overloaded constructor
String(const String &); // copy constructor

};
String::String(){
bufferPtr = NULL;
size = 0;
}
String::String(char * ptr){
if(ptr != NULL){
size = strlen(ptr);
bufferPtr = new char[size+1];
strcpy(bufferPtr, ptr);
}

Virtual University of Pakistan

147

Object Oriented Programming (CS304)

VU

else{
bufferPtr = NULL;
size = 0;
}
}
String::String(const String & rhs){
size = rhs.size;
if(rhs.size != 0){
bufferPtr = new char[size+1];
strcpy(bufferPtr, ptr);
}
else
bufferPtr = NULL;
}
int main(){
String str1(Hello");
String str2(World);
str1 = str2;78
return 0;
}

Member wise copy

assignment

Assignment operator (Shallow Copy)

Result of str1 = str2 (memory leak)

World

Hello

str2

str1

Second issue is dangling pointer issue as was in the case of copy constructor.
Modified Assignment Operator Code:
So we add overloaded assignment operator to perform deep copy as given below,

Here by term member wise copy we mean copying values of members of class one by one blindly
also called bitwise copy.
8
In ANSI c++ standard term member wise copy has also been used to indicate logical copy (the deep
copy)

148

Virtual University of Pakistan

Object Oriented Programming (CS304)

VU

class String{

public:

void operator =(const String &);

};
void String::operator = (const String & rhs){
size = rhs.size;
if(rhs.size != 0){
delete [] bufferPtr; // resolving inaccessible memory issue
bufferPtr = new char[rhs.size+1]; // creating new dynamic memory
strcpy(bufferPtr,rhs.bufferPtr); // deep copy
}
else
bufferPtr = NULL;
}
int main(){
String str1(ABC");
String str2(DE), str3(FG);
str1 = str2;
// Valid
str1 = str2 = str3; // Error
return 0;
}

The problem in statement

str1 = str2 = str3 is,
str1=str2=str3 is resolved as:
str1.operator=(str2.operator=(str3))
Assignment operator is beiing called two times one for part str2 = str3 and then for
str1 = (str2 = str3) as assignment operator is right associate so first str2=str3 will be
executed, and str2 will become equal to str3, then first overloaded assignment
operator execution result will be assigned to s1,
str1.operator=(str2.operator=(str3))
Problem is when compiler will try to invoke second assignment operator to assign
value to str1 error will be returned becuase of void return type of oveloaded
assignment operator the reasn is explained below,
As we have studied before values are passed as parameters in operator overloadnig,
str2 = str3 means str2.operator = (str3)
// str3 is being called as parameter
and
str1 = str2 = str3 means str3.operator(str2.operator = (str3))
// str2.operator = (str3) is being passed as parameter
This issue can be resolved by introducing minor change in our code of copy
assignment operator to make it return String object instead of void as shown below,

Virtual University of Pakistan

149

Object Oriented Programming (CS304)

VU

class String{

public:

String & operator = (const String &);

};
String & String :: operator = (const String & rhs){
size = rhs.size;
delete [] bufferPtr;
if(rhs.size != 0){
bufferPtr = new char[rhs.size+1];
strcpy(bufferPtr,rhs.bufferPtr);
}
else bufferPtr = NULL;
return *this;
}
Now we are returning the value by reference of the object with respect to which this
overloaded assignment operator will be called. It will be str2 in the case of str2 = str3,
now when part str1 = (str2 = str3) will be executed, str2 will be passed as argument,
that will be assigned to str1.

150

Virtual University of Pakistan

Object Oriented Programming (CS304)

VU

Lecture No.18
18.1. Self assignment problem:
In we assign same string to itself as done in main function below our program will
produce unexpected results as source and destination operands for copying are
same,
int main(){
String str1("Fakhir");
str1 = str1; // Self Assignment problem
return 0;
}
Result of str1 = str1

// size = rhs.size;
// delete [] bufferPtr;

???
Fakhir

str1

We can resolve this issue by adding a simple if condition to ensure that both strings
are not same
String & String :: operator = (const String & rhs){
if(this != &rhs){
size = rhs.size;
delete [] bufferPtr; // deleting memory of left hand side operand
if(rhs.bufferPtr != NULL){
bufferPtr = new char[rhs.size+1];
strcpy(bufferPtr,rhs.bufferPtr);
// memory access violation or incorrect data copy
}
else bufferPtr = NULL;
}
return *this;
}
Now self-assignment is properly handled:
int main(){
String str1("Fakhir");
str1 = str1;
return 0;
}
We can make return type String & as constant to avoid assignment to sub
expressions, like (str1 = str2) = str3

Virtual University of Pakistan

151

Object Oriented Programming (CS304)

VU

class String{

public:

const String & operator=9

(const String &);
};
int main(){
String s1(ABC),
s2(DEF),
s3(GHI);
// Error
(s1 = s2) = s3;
return 0;
}
But as we can do that with primitive types so we can allow assignment to sub
expressions by making return type as String & only as we have done before.
int main(){
int a, b, c;
(a = b) = c;
return 0;
}
18.2. Other Binary operators
Overloading += operator:
class Complex{
double real, img;
public:
Complex & operator+=(const Complex & rhs);
Complex & operator+=(count double & rhs);
...
};
Complex & Complex::operator += (const Complex & rhs){
real = real + rhs.real;
img = img + rhs.img;
return * this;
}
Complex & Complex::operator +=
real = real + rhs;
return * this;

(const double & rhs){

We have seen previously that we should not return handle to any private data
member

152

Virtual University of Pakistan

Object Oriented Programming (CS304)

VU

}
int main(){
Complex c1, c2, c3;
c1 += c2;
c3 += 0.087;
return 0;
}
18.3. Friend Functions and Operator overloading
Friend functions minimize encapsulation as we can access private data of any class
using friend functions,
This can result in:
Data vulnerability
Programming bugs
Tough debugging
Hence, use of friend functions must be limited we can overload operators without
declaring them friend functions of a class, for example the + operator can be defined
as a non-member, non-friend function as shown below,
(Three versions of overloaded + operator for Complex no class to handle three kinds
of statements)
obj1 + obj2
o + obj1
obj1 + 3.78
Complex operator + (const Complex & a, const Complex & b){
Complex t = a; // creating temporary object t to store a+b
return t += b; // returning t by reference
}
Complex operator + (const double & a, const Complex & b){
Complex t = b;
return t += a;
}
Complex operator + (const Complex & a, const double & b){
Complex t = a;
return t += b;
}
Other Binary operators
The operators
-=, /=, *=, |=, %=, &=, ^=, <<=, >>=, !=
can be overloaded in a very similar fashion.

Virtual University of Pakistan

153

Object Oriented Programming (CS304)

VU

Lecture No.19
Overloading stream insertion extraction operators
19.1. Stream Insertion operator
Often we need to display the data on the screen c++ provides us insertion operator
(<<)to put data on output stream default is console but it can be any file or network
socket as well to send data on network from our program.
Example:
int i=1, j=2;
cout << i= << i << \n;
cout << j= << j << \n;
19.2. Stream Extraction operator
We also need to get data from the console or from file or network this is achieved
through c++ provided stream extraction operator (>>) that is used to get data from
input stream, again default input stream is from console.
Example:
int i,j;
cin >> i >> j; // getting value of i and j from user
Explanation:
cin and cout are objects of istream and ostream classes used for input and output
respectively, the insertion and extractions operators have been overloaded in istream
and ostream classes to do these tasks.
When we write lines like,
int i;
cin>> i;
cout << i;
Actually we are using istream and ostream class objects and using these objects we
are calling these classes overloaded (>> and <<) operators that have been overloaded
for all basic types like integer, float , long , double and char *.
We have seen previously that actual call of overloaded operators for a class takes
place by passing the objects to overloaded function as parameter like shown below,
cin>>i;
istream & operator >> (istream & in, int & i)
Here cin will be passed as istream object along with int i to be displayed and code of
this functions is returning istream object by reference & to accommodate multiple
input statement in a single line like,
int i , j;
cin>>i >> j;
as we did in our string class before.

154

Virtual University of Pakistan

Object Oriented Programming (CS304)

VU

Same goes for insertion operator <<

cout << i;
ostream & operator >> (ostream & os, const int & i)
Stream insertion and extraction operator have been overloaded for basic data types
but if we try to use them for user defined data types like our Complex no. class
compiler will generate error as it will not find any overloaded operator code for our
complex no class
Complex c1;
cout << c1; // Error
cout << c1 << 2; // Error cascaded statement
// Compiler error: binary '<<' : no operator // defined which takes a right-hand operand of
type class Complex
Same error will be for stream extraction operator so will need to overload these two
operators (<< and >>) for our Complex no. class.

19.3. Overloading Stream Insertion Operator

First we try to overload insertion << operator as member function as we did before,
class Complex{

public:

void operator << (const

Complex & rhs);
};
But when we will call this overloaded function is main compiler will generate errors
as shown below,
int main(){
Complex c1;
cout << c1;
// Error
c1 << cout;
c1 << cout << 2; // Error
return 0;
};
class Complex{

public:

void operator << (ostream &);

};
void Complex::operator << (ostream & os){

Virtual University of Pakistan

155

Object Oriented Programming (CS304)

os

VU

<< ( << real

<< , << img << );

}
Now the statement c1 << cout will work but it has two limitations,
Difficult to understand and remember statement syntax (c1 << cout ; )
Cascaded statements not possible ( cout << c1 << 2 ;)
Better syntax is given below to resolve these two issues,
class Complex{
...
friend ostream & operator << (ostream & os, const Complex & c);
};
Stream Insertion operator
// we want the output as: (real, img)
ostream & operator << (ostream & os, const Complex & c){
os << ( << c.real
<< ,
<< c.img << );
return os;
}
ostream reference can not be const as it store the data in its buffer to insert on output
stream, however Complex reference will be constant as we are only getting data from
Complex object and inserting it to output stream.
Complex c1(1.01, 20.1), c2(0.01, 12.0);
cout << c1 << endl << c2;
Stream Insertion operator
Output:
( 1.01 , 20.1 )
( 0.01 , 12.0 )
Now cascading statements are also possible as given below,
cout << c1 << c2;
is equivalent to
operator<<( operator<<(cout,c1),c2);
Because insertion operator is Left to right associative so first left part cout << c1 << c2
will be executed and then the next part as opposed to copy assignment operator that
will right associative.
Same thing can be done with stream extraction operator,
19.4. Overloading Stream Extraction Operator:

156

Virtual University of Pakistan

Object Oriented Programming (CS304)

VU

class Complex{
...
friend istream & operator >> (istream & i, Complex & c);
};
istream can not be cont and istream buffer will change as we will get data from it and
assign it to complex reference similarly Complex object can not be const for stream
extraction operator as well because we will add data to it and hence its state will
change.
Stream Extraction Operator Code:
istream & operator << (istream & in, Complex & c){
in >> c.real;
in >> c.img;
return in;
}
Main Program:
Complex c1(1.01, 20.1);
cin >> c1;
// suppose we entered // 1.0025 for c1.real and
cout << c1;

// 0.0241 for c1.img

Output:
( 1.0025 , 0.0241 )

19.5. Other Binary operators:

Overloading comparison operators (Equality and Inequality operators)
class Complex{
public:
bool operator == (const Complex & c);
//friend bool operator == (const //Complex & c1, const Complex & c2);
bool operator != (const Complex & c);
//friend bool operator != (const //Complex & c1, const Complex & c2);

};
Equality operator:
bool Complex::operator ==(const Complex & c){
if((real == c.real) &&
(img == c.img)){

Virtual University of Pakistan

157

Object Oriented Programming (CS304)

return true;
}
else
return false;
}
As non member friend function:
bool operator ==(const Complex& lhs, const Complex& rhs){
if((lhs.real == rhs.real) &&
(lhs.img == rhs.img)){
return true;
}
else
return false;
}
Inequality Operator:
bool Complex::operator !=(const Complex & c){
if((real != c.real) ||
(img != c.img)){
return true;
}
else
return false;
}

158

Virtual University of Pakistan

VU

Object Oriented Programming (CS304)

VU

Lecture No.20
Modified String Class:
We have seen the following string class till now,
class String{
private:
char * bufferPtr;
int size;
public:
String();
String(char * ptr);
void SetString(char * ptr);
const char * GetString();
...
};
int main(){
String str1(Test);
String str2;
str2.SetString(Ping);
return 0;
}
What if we want to change the string from Ping to Pong?? {ONLY 1 character to be
changed}
Possible solution:

Call: str2.SetString(Pong);
This will delete the current buffer and allocate a new one
Too much overhead if string is too big

Or, we can add a function which changes a character at nth location

class String{
...
public:
void SetChar(char c, int pos);
...
};
void SetChar(char c, int pos){
if(bufferPtr != NULL){
if(pos>0 && pos<=size)
bufferPtr[pos] = c;
}
}
Other Binary Operators
Now we can efficiently change a single character:
String str1(Ping);

Virtual University of Pakistan

159

Object Oriented Programming (CS304)

VU

str1.SetChar(o, 2);
// str1 is now changed to Pong
20.1. Subscript [] Operator
There is another elegant solution present for this problem. It is subscript operator
that is used on basic char [] data type as shown below,
char array[5] = Ping;
array[1] = o;
We want to get same functionality for our own defined String class for this we
overload the subscript [] operator.
We want function like given below for subscript operator in our String class,
int main(){
String str2;
str2.SetString(Ping);
str[2] = o;
// acting as l-value (left value so that we can assign it some value)
cout << str[2];
// acting as r-value (we are reading value using subscript operator)
return 0;
}
20.2. Overloading Subscript [] Operator
Subscript operator must be overloaded as member function of the class with one
parameter of integer type,
class String{
...
public:
char & operator[](int);
...
};
char & String::operator[]( int pos){
assert(pos>0 && pos<=size);
return stringPtr[pos-1];
}
int main() {
String s1(Ping);
cout <<str.GetString()<< endl;
s1[2] = o;
cout << str.GetString();
return 0;
}
Output:

160

Virtual University of Pakistan

Object Oriented Programming (CS304)

VU

Ping
Pong
20.3. Overloading Function () operator
Must be a member function
Any number of parameters can be specified
Any return type can be specified
Operator() can perform any generic operation
Function Operator
class String{
...
public:
char & operator()(int);
...
};
char & String::operator()
(int pos){
assert(pos>0 && pos<=size);
return bufferPtr[pos-1];
}
int main(){
String s1(Ping);
char g = s1(2); // g = i
s1(2) = o;
cout << g << \n;
cout << str.GetString();
return 0;
}
Output:

i
Pong

20.4. Function Operator performing Sub String operation,

class String{
...
public:
String operator()(int, int);
...
};
String String::operator()(int index, int subLength){
assert(index>0 && index+subLength-1<=size);
char * ptr = new char[subLength+1];

Virtual University of Pakistan

161

Object Oriented Programming (CS304)

VU

for (int i=0; i < subLength; ++i)

ptr[i] = bufferPtr[i+index-1];
ptr[subLength] = \0;
String str(ptr);
delete [] ptr;
return str;
}
int main(){
String s(Hello World);
// << is overloaded
cout << s(1, 5);
return 0;
}
Function Operator
Output:

Hello

20.5. Unary Operators

Unary operators take one operand, they act on the object with reference to which
they have been called as shown below,
& * + - ++ -- ! ~
Examples:

--x
-(x++)
!(*ptr ++)

Unary Operators
Unary operators are usually prefix, except for ++ and -++ and -- both act as prefix and postfix
Example:
h++;
g-- + ++h - --i;
General syntax for unary operators
As Member Functions:
TYPE & operator OP (); // no argument the object with respect to which it is called
is taken as one operand
As Non-member Functions:
Friend TYPE & operator OP (TYPE & t);
// one argument object with respect to which it is called.

162

Virtual University of Pakistan

Object Oriented Programming (CS304)

VU

Overloading unary -:
class Complex{
...
Complex operator - ();
// friend Complex operator
//
-(Complex &);
}
Complex Complex::operator -(){
Complex temp;
temp.real = -real;
temp.img = -img;
return temp;
}
Complex c1(1.0 , 2.0), c2;
c2 = -c1;
// c2.real = -1.0
// c2.img = -2.0
Unary + is overloaded in the same way.

Virtual University of Pakistan

163

Object Oriented Programming (CS304)

Lecture No.21
Unary Operators
21.1. Behavior of ++ and -- for pre-defined types:
Post-increment ++:
Post-increment operator ++ increments the current value and then returns the
previous value
Post-decrement --: Works exactly like post ++
Example:
int x = 1, y = 2;
cout << y++ << endl;
cout << y;
Output:
2
3
Example:
int y = 2;
y++++;// Error
y++ = x;
// Error
Behavior of ++ and -- for pre-defined types:
Pre-increment ++:
Pre-increment operator ++ increments the current value and then returns its
reference
Pre-decrement --:
Works exactly like Pre-increment ++
Example:
int y = 2;
cout << ++y << endl;
cout << y << endl;

Output:
3
3
Example:
int x = 2, y = 2;
++++y;
cout << y;
++y = x;
cout << y;
Output:
4
2

164

Virtual University of Pakistan

VU

Object Oriented Programming (CS304)

VU

Example (Pre-increment):
class Complex{
double real, img;
public:
...
Complex & operator ++ ();
// friend Complex & operator ++(Complex &);
}
Complex & Complex::operator++(){
real = real + 1;
return * this;
}

// member function

Complex & operator ++ (Complex & h){ // non member function

h.real += 1;
return h;
}
Example:
Complex h1, h2, h3;
++h1;
Function operator++() returns a reference so that the object can be used as an lvalue
++h1 = h2 + ++h3;
How does a compiler know whether it is a pre-increment or a post-increment?
A post-fix unary operator is implemented using:
Member function with 1 dummy int argument
OR
Non-member function with two arguments
In post increment, current value of the object is stored in a temporary variable
Current object is incremented
Value of the temporary variable is returned
21.2. Post-increment operator:
class Complex{
...
Complex operator ++ (int);
// friend Complex operator ++(const Complex &, int);
}
Complex Complex::operator ++ (int){
complex t = *this;

Virtual University of Pakistan

165

Object Oriented Programming (CS304)

VU

real += 1;
return t;
}
Complex operator ++ (const
Complex & h, int){
complex t = h;
h.real += 1;
return t;
}
How does a compiler know whether it is a pre-increment or a post-increment?
The dummy parameter in the operator function tells compiler that it is postincrement
Example:
Complex h1, h2, h3;
h1++;
h3++ = h2 + h3++; // Error
The pre and post decrement operator -- is implemented in exactly the same way
21.3. Type Conversion
The compiler automatically performs a type coercion of compatible types
e.g:
int f = 0.021;
double g = 34;
// type float is automatically converted into int. Compiler only issues a warning
Type Conversion
The user can also explicitly convert between types:
int g = (int)0.0210;
double h = double(35);
// type float is explicitly converted
// is issued now

(casted) into int. Not even a warning

Type Conversion
For user defined classes, there are two types of conversions
From any other type to current type
From current type to any other type
We can do the following type conversion by our self,
Conversion from any other type to current type:
Requires a constructor with a single parameter
For example,
String str = 135;
Conversion from current type to any other type:
Requires an overloaded operator

166

Virtual University of Pakistan

Object Oriented Programming (CS304)

VU

For example,
String str;
char * ptr = str;
Conversion from other type to current type (int to String):
class String{
...
public:
String(int a);
char * GetStringPtr()const;
};
String::String(int a){
cout << "String(int) called..." << endl;
char array[15];
itoa(a, array, 10);
size = strlen(array);
bufferPtr = new char [size + 1];
strcpy(bufferPtr, array);
}
char * String::GetStringPtr() const{
return bufferPtr;
}
int main(){
String s = 345;
cout << s.GetStringPtr() << endl;
return 0;
}
Output:

String(int) called
345

Automatic conversion like shown above using constructor has drawbacks conversion
takes place transparently even if the user didnt wanted the conversion for example
see the code below, in it user can write the following code to initialize the string with
a single character:
int main(){
String s = A;
// ASCII value of A that is 65 is being taken automatically it may be what the user //
wants, perhaps user intent was String s = A to store A in the string object
// but instead 65 is being stored
cout << s.GetStringPtr()<< endl << s.GetSize() << endl;
return 0;

Virtual University of Pakistan

167

Object Oriented Programming (CS304)

VU

}
Output:
String(int) called
65
2

Keyword explicit
There is a mechanism in C++ to restrict automatic conversions like this using
constructor code it is to use keyword explicit with such constructors if we have used
this keyword then casting must be explicitly performed by the user.
Keyword explicit only works with constructors.
Example:
class String{

public:

explicit String(int);
};
int main(){
String s;
s = A;
return 0;
}

// Error

int main(){
String s1, s2;
s1 = String(101);
// OR
s2 = (String)204;
return 0;
}

// valid, explicit casting

Type Conversion
There is another method for type conversion:
Operator overloading
It is used for converting from current type (user defined) to any other basic type or
user defined type.
General Syntax:
TYPE1::operator TYPE2();
Like,
String::operator char * (); // TYPE1 = String , TYPE2 = char *

168

Virtual University of Pakistan

Object Oriented Programming (CS304)

VU

It will convert our string object to char *.

We write such functions as member function of the class.
NO return type and arguments are specified
Return type is implicitly taken to be TYPE2 by compiler
Type Conversion
Overloading pre-defined types:
class String{

public:

operator int();
operator char *();
};
String::operator int(){
if(size > 0)
return atoi(bufferPtr);
else
return -1;
}
String::operator char *(){
return bufferPtr;
}

int main(){
String s("2324");
cout << (int)s << endl << (char *)s;
// int a = (int)s;
// int a = int (s);
return 0;
}
Output:
2324
2324

21.4. User Defined types:

User-defined types can be overloaded in exactly the same way
Only prototype is shown below:
class String{

operator Complex();
operator HugeInt();
operator IntVector();

Virtual University of Pakistan

169

Object Oriented Programming (CS304)

VU

};
21.5. Drawbacks of Type Conversion Operator:
class String{

public:

String(char *);
operator int();
};
int main(){
String s(Fakhir");
// << is NOT overloaded
cout << s; // compiler is automatically converting s to int
return 0;
}
Output:
Junk Returned
To avoid this problem DO NOT use type conversion operators instead use separate
member function for such type conversion as shown below,
Modifying String class:
class String{

public:

String(char *);
int AsInt();
};
int String::AsInt(){
if(size > 0)
return atoi(bufferPtr);
else
return -1;
}
int main(){
String s(434");
// << is NOT overloaded
cout << s; //error
cout << s.AsInt();
return 0;
}

170

Virtual University of Pakistan

Object Oriented Programming (CS304)

VU

Lecture No.22
22.1. Practical implementation of Inheritance in c++
Topics to be discussed in this lecture are,

Inheritance in Classes
UML Notation of Inheritance
Types of Inheritance in c++
IS A relationship
Accessing members
Allocation in memory
Constructors
Base class initializers
Initializing members
Destructors
Order of execution constructors and destructors
Examples of C++ implementation of inheritance

22.2. Inheritance in Classes

If a class B inherits from class A, then B contains all the characteristics (information
structure and behavior) of class A.
The class whose behavior is being inherited is called base class and the class who
inherits the behavior of base class is called derived class. Base class is also called
parent class and child class is called also derived class
Besides inherited characteristics, derived class may have its own unique
characteristics
22.3. UML Notation
We use arrow from derived class to the parent class to show inheritance as shown
below,

22.4. Inheritance in C++

Virtual University of Pakistan

171

Object Oriented Programming (CS304)

VU

In c++ we can inherit a class from another class in three ways,

Public
Private
Protected
22.5. IS A Relationship
Inheritance represents IS A relationship for example a student IS A person.
In general words we can say that inheritance represents,
Derived class IS A kind of Parent class
C++ Syntax of Inheritance
class ChildClass
: public BaseClass{
...
};
Example
class Person{
...
};
class Student: public Person{
...
};
Accessing Members
Public members of base class become public member of derived class.
Private members of base class are not accessible from outside of base class, even in
the derived class (Information Hiding)
Example
In the code given below Student and Teacher classes has been derived from single
Person class,

Person

Student

Teacher

class Person{
char *name;
int age;
...
public:
const char *GetName() const;
int GetAge() const;
...
};

172

Virtual University of Pakistan

Object Oriented Programming (CS304)

class Teacher: public

Person{
char * dept;
int course;
...
public:
char * GetDept() const;
int GetCourse() const;
void Print() const;
...

VU

class Student: public

Person{
int semester;
int rollNo;
...
public:
int GetSemester() const;
int GetRollNo() const;
void Print() const;
...

Example
void Student::Print()

Error
{
cout << name << is in << semester << semester;
}
corrected Code:
void Student::Print()
{
cout << GetName() << is in semester << semester;
}
int main(){
Student stdt;
stdt.semester = 0;//error
stdt.name = NULL; //error
cout << stdt.GetSemester();
cout << stdt.GetName();
return 0;
}
Explanation of above code (char * data type)
In c++ char arrays ( char []) are handled in two ways one way is statically using
statements like,
char name[30]; // static array of length 30 characters
or dynamically as shown below,
char * name;
name = new char[30];
In dynamic creation of arrays we simply store char * in class and assign it a dynamic
memory according to our need using new operator.

Virtual University of Pakistan

173

Object Oriented Programming (CS304)

VU

Allocation in Memory
The object of derived class is represented in memory as follows

base member1
base member2
...

Data members of base

class

derived member1
derived member2
...

Data members of derived

class

Derived Class Object

Every object of derived class has an anonymous object of base class

Constructors
The anonymous object of base class must be initialized using constructor of
base class
When a derived class object is created the constructor of base class is executed
before the constructor of derived class

base member1
base member2
...

Base class constructor initializes

the anonymous (base class) object

derived member1
derived member2
...

Derived class constructor

initializes the derived class object

Derived Class Object

Example
class Parent{
public:
Parent(){ cout <<
Parent Constructor...;}
};
class Child : public Parent{
public:
Child(){
cout <<
Child Constructor...;}
};

174

Virtual University of Pakistan

Object Oriented Programming (CS304)

VU

int main(){
Child cobj;
return 0;
}
Output:
Parent Constructor...
Child Constructor...

Constructor
If default constructor of base class does not exist then the compiler will try to
generate a default constructor for base class and execute it before executing
constructor of derived class
If the user has given only an overloaded constructor for base class, the
compiler will not generate default constructor for base class
Example
class Parent{
public:
Parent(int i){}
};
class Child : public Parent{
public:
Child(){}
} Child_Object; //ERROR
Definition of Some Terms:
Default constructor: Default constructor is such constructor which either has no
parameter or if it has some parameters these have default values. The benefit of
default constructor is that it can be used to create class object without passing any
argument.
Implicit Default constructor:
Compiler generates implicit default constructor for any class in case we have not
given any constructor for the class.
Explicit Default constructor:
If user has given constructor for any class without any arguments or with all
arguments with default values then it is also default constructor according to
definition but it is explicit (user defined) default constructor.
Now if a base class has only non-default constructor (constructor with parameters
without default values), then when we will create object of any class derived from
this base class compiler will not be able to call base class constructor as base class has
no default constructor ( constructor that can be called without giving any
parameters) so compiler will generate error.
We can avoid this error by calling base class non-default constructor in derived class
constructor initializer list by ourself.

Virtual University of Pakistan

175

Object Oriented Programming (CS304)

VU

Base Class Initializer

C++ has provided a mechanism to explicitly call a constructor of base class
from derived class

The syntax is similar to member initializer and is referred as base-class

initialization

Example
class Parent{
public:
Parent(int i){};
};
class Child : public Parent{
public:
Child(int i): Parent(i)
{}
};
Example
class Parent{
public:
Parent(){cout <<
Parent Constructor...;}
...
};
class Child : public Parent{
public:
Child():Parent()
{cout << Child Constructor...;}
...
};
Base Class Initializer
User can provide base class initializer and member initializer simultaneously
Example
class Parent{
public:
Parent(){}
};
class Child : public Parent{
int member;
public:
Child():member(0), Parent()
{}
};
Base Class Initializer

176

Virtual University of Pakistan

Object Oriented Programming (CS304)

VU

The base class initializer can be written after member initializer for derived
class
The base class constructor is executed before the initialization of data
members of derived class.
Initializing Members
Derived class can only initialize members of base class using overloaded
constructors
o Derived class can not initialize the public data member of base class
using member initialization list
Example
class Person{
public:
int age;
char *name;
...
public:
Person();
};

Example
class Student: public Person{
private:
int semester;
...
public:
Student(int a):age(a)
{
}
};

//error

Reason
It will be an assignment not an initialization
Destructors
Destructors are called in reverse order of constructor called
Derived class destructor is called before the base class destructor is called
Example
class Parent{
public:
Parent(){cout <<Parent Constructor;}
~Parent(){cout<<Parent Destructor;}
};
class Child : public Parent{
public:
Child(){cout << Child Constructor;}
~Child(){cout << Child Destructo;}
};
Example

Virtual University of Pakistan

177

Object Oriented Programming (CS304)

Output:
Parent Constructor
Child Constructor
Child Destructor
Parent Destructor

178

Virtual University of Pakistan

VU

Object Oriented Programming (CS304)

VU

Lecture No.23
Lecture Contents:

Protected Access Specifier in Inheritance

Implicit and explicit use of IS A relationship

Previous lecture discussion:

Definition of Some Terms:
Default constructor: Default constructor is such constructor which either has no
parameter or if it has some parameters these have default values. The benefit of
default constructor is that it can be used to create class object without passing any
argument.
Implicit Default constructor:
Compiler generates implicit default constructor for any class in case we have not
given any constructor for the class.
Explicit Default constructor:
If user has given constructor for any class without any arguments or with all
arguments with default parameters then it is also default constructor according to
definition but it is explicit default constructor.
Now if a base class has only non-default constructor (implicit or explicit), then when
we will create object of any class derived from this base class compiler will not be
able to call base class constructor as base class has no default constructor (
constructor that can be called without giving any parameters) so compiler will
generate error.
We can avoid this error by calling base class non-default constructor in derived class
constructor initializer list.
23.1. Accessing base class member functions in derived class:
Public methods of base class can directly be accessed in its derived class (derived
class interface consists of its own member functions plus member functions of its
base class).
However there are some class members functions that are written just as helper
functions for other class member functions and they need not to be called directly
using class object for example,
1. Suppose some function in our class wants to get some input from user in the
form of integers only for this we can write another function that checks
whether the entered string by user consists of integers only.
2. Suppose our class has implement encryption of data, and it encodes and
decodes data using some functions these functions will also to helper
functions and should not be accessible to outside word.

Virtual University of Pakistan

179

Object Oriented Programming (CS304)

VU

3. Similarly take the example of our Date class we studied in lecture no.13 has a
function to check whether an year is leap year or not and according to result
it allows or disallows the days of February to be 29. This bool IsLeapYear(int
) will used by other members functions of Date class like void setDay( int )
and void addDay( int ), this functions is also helper function.
These helper functions are made private as there is no need to access them using
class object directly as shown below,
class Date{
int day, month, year;
static Date defaultDate;
public:
void SetDay(int aDay);
int GetDay() const;
void AddDay(int x);

static void SetDefaultDate(int aDay,int aMonth, int aYear);

...
private:
bool IsLeapYear(int );
};
int main(){
Error:
Date aDate;
bool IsLeapYear(int ) is private method of Date
aDate.IsLeapYear( year );
class so it is not accessible here.
return 0;
}
Making such functions private works fine as long as we dont derive any child class
from this class, but when we derive some class for specialization these functions will
not be accessible in the derived class as private members functions are accessible
only in those class to which they belong but these functions are needed in derived
classes as well example is given below,
Example:
Suppose we specialize our Date class by adding a child class SpecialDate to handle
only working days in the an year, we will need to use IsLeapYear function in this
child class as well but IsLeapYear function is private member function of base Date
class (in accordance with the principles of data hiding and encapsulation), so it will
not be accessible in child class as shown below,
class SpecialDate: public Date{

bool IsLeapYear(int ) is
private method of Date class
public:
so it is not accessible in its
void AddSpecialYear(int i){
derived SpecialDate class.
...
if(day == 29 && month == 2
&&
!IsLeapyear(year+i)){
//ERROR!
...

180

Virtual University of Pakistan

Object Oriented Programming (CS304)

VU

}
}
};

Solution:
Modify Access Specifier
One solution of this problem is that we make IsLeapYear function public from
private as shown below,

class Date{
public:
...
bool IsLeapYear(int );
};
void SpecialDate :: AddSpecialYear (int i) {
...
if(day == 29 && month == 2
&&
!IsLeapyear(year+i)){
...
}
}

But the problem with making this function public is that now this functions will be
accessible to everyone using Date or SpecialDate class object but we dont want this,
we want that our base or derived class can access this function only. the solution of
this problem is protected access specifier.
23.2. protected access specifier:
c++ provides us with protected access specifier for these sorts of situations; protected
access specifier ensures that function in base class is accessible in derived class of this
base class and NOT outside of this class. So we can say that scope of protected access
specifier is somewhere in between private and public access specifiers it is similar to
private in the sense that it doesnt allow the external world to access protected
member and it is similar to public access specifier in the sense that it ensures that
protected member is accessible in derived classes of the protected member class.
Protected members of a class can not be accessed outside the class but only in the
derived class of that class. Protected members of base class become protected
member of derived class.10
class Date{

10
This is the case in public and protected inheritance whereas in private inheritance protected members
become private members of derived class public, protected and private inheritance has been discussed
in detail in lecture no.26.

Virtual University of Pakistan

181

Object Oriented Programming (CS304)

VU

protected:
bool IsLeapYear(int );
};
int main(){
Date aDate;
aDate.IsLeapYear(); //Error
return 0;
}
void SpecialDate :: AddSpecialYear (int i) {
...
if(day == 29 && month == 2
&&
!IsLeapyear(year+i)){
...
}
}

Now it is ok to call function bool

IsLeapYear in derived class
SpecialDate.

Disadvantages of protected Members:

Breaks encapsulation
The protected member is part of base classs implementation as well as derived
classs implementation. So we can say that protected members breaks the principle of
Encapsulation to some extent which says A class data members and functions
should be encapsulated in the class itself

23.3. IS A Relationship
We have seen previously that Public inheritance models the IS A relationship for
example see the diagram below,

Shape

Triangle

Line
Circle

Here,

182

Line IS A Shape
Virtual University of Pakistan

Object Oriented Programming (CS304)

VU

Circle IS A Shape
Triangle IS A Shape

Generally we can say that,

Derived Object IS A kind of Base Object
It means that derived class object is special kind of base object with extra properties
(attributes and behaviour ) so derived object has its own properties as well as
properties of its base object this is the reason why base class constructor is also called
while creating derived class object.
We can use derived class object where base class object is required as derived class
object has implicit base class object also but the reverse of this statement is not true,
we can not use base class object where derived class object is required because if we
create base class it will only have base part not its derived class part.

Implicit
Base Class
Object

Base Class Object

Derived Class Object

Example
See the code below in it first we are using derived class object sobj to call methods
GetName and GetRollNo belonging to classes Person and Student respectively, this is
in accordance to the principle of inheritance that says child class object can access
public and protected members of base class.
class Person {
char * name;
public: ...
const char * GetName();
};
class Student: public Person{
int rollNo;
public: ...

Virtual University of Pakistan

183

Object Oriented Programming (CS304)

VU

int GetRollNo();
};
int main()
{
Student sobj;
cout << sobj.GetName();
cout << sobj.GetRollNo();
return 0;
}

Now as we discussed previously that inheritance represents IS A relationship

meaning relationship like,
Derived class IS A Base class
Hence Derived Class Object can be used where its Base Class object is required but
Base Class Object can not be used where Derived Class Object is required example
given below explain this point,
int main(){
Person * pPtr = 0; // base class pointer
Student s; // derived class object
pPtr = &s;
/* assigning derived class object address to base class pointer derived class pointer is
converted to base class pointer */
cout << pPtr->GetName(); // calling base class methods using base class
reference
return 0;
}
So parent class object reference can hold the reference of derived class object as
shown below,
pPtr = &s;
But parent class reference will have access only to the interface of parent class, if we
try to access derived class functions from this parent class reference error will be
generated as shown below,
int main(){
Person * pPtr = 0;
Student s;
pPtr = &s;
cout << pPtr->GetRollNo();
return 0;
}

//Error

Compiler use the concept of static typing to decide the access to member functions as
described below,

184

Virtual University of Pakistan

Object Oriented Programming (CS304)

VU

23.4. Static Type

The type that is used to declare a reference or pointer is called its static type
1. In Person * pPtr = 0;
The static type of pPtr is Person * (Person pointer).
2. Student s;
The static type of s is Student.
Member Access
As the static type of pPtr is Person so following call is erroneous
pPtr->GetRollNo();
As pPtr static type is Person pointer and in Person class there is no GetRollNo()
function so compiler will generate an error.
In c++ we can also use references (called as reference identifiers or reference
variables or reference constants) to any type instead of pointers, for example see the
example below where the reference of base object is being initialized with derived
class object,
Example (Explicit use of IS A relationship)
int main(){
Person p;
Student s;
Person & refp = s;
/*Here refp is becoming reference (alias) of Student object but as its own static
type is Person so it can access member functions of Student class only. */
cout << refp.GetName();
cout << refp.GetRollNo(); //Error
return 0;
}
Example (Implicit use of IS A relationship)
Play(const Person& p){
cout << p.GetName()
<< is playing;
}
void Study(const Student& s){
cout << s.GetRollNo()
<< is Studying;
}
int main(){
Person p;

Virtual University of Pakistan

185

Object Oriented Programming (CS304)

VU

Student s;
Play(p); /* parameter of function Play is being initialized with argument p */
Play(s);
/* parameter of function Play is being is initialized with argument s as student IS
A kind of person*/
return 0;
}

186

Virtual University of Pakistan

Object Oriented Programming (CS304)

VU

Lecture No.24
Lecture Overview:
Protected Access Specifier
IS-A relationship
Copy constructor and inheritance

Protected Members:
Protected members are somewhere between public and private members. They are
used in inheritance.
From outside the class no once can access protected access specifier however any
publicly derived class can access the protected access specifiers and they behave as
protected members of derived class.
In a single standalone class protected access specifier has no significance but when
we derive any class from this class protected access specifier becomes significant.
IS-A relationship:
Inheritance is used when two classes have IS A kind of relationship between two
classes.
In c++ public inheritance is used for IS A relationship.
We can use publicly derived class pointer for its base class pointer because derived
class is a kind of base class or in other words derived class is base class with some
extra properties. In case of Student IS-A Person relationship student has all
properties of person for example student walks, eats, drinks and so on as person
does but student has some extra properties like student studies is some study
program.
(We can assign derived class pointer to its base class pointer because for compiler it
is explicit manifestation of IS-A relationship.)

Static type of an identifier:

Static type of an identifier is type used to declare it.
Static type of an identifier is the type used for declaration of that identifier for
example,
Student * student;
Person * person;
Here static type of pointer student is Student and person is Person.

Virtual University of Pakistan

187

Object Oriented Programming (CS304)

VU

Access to members of an identifier is governed by static type of that identifier,

compiler uses static type for an identifier to allow the identifier to use different
members (functions and variables).

Copy Constructor:11
Copy constructor is a member function of a class that is used to create an object of
that class by copying values from an already created object.

Copy Constructor in case of Inheritance:

Consider the code given below having two classes Person and Student in this code
Person is base and Student is derived class. Person has one attribute name to store
name of the person and Student has one attribute major to store study program of
the student,

#include <iostream>
using namespace std;

/*Listing 24.1 */
/*Program to demonstrate Copy Constructor in case of base and derived classes.
*/
/* Base Class Person */
class Person{
char * name;
public:
Person(char * = NULL);
const char * GetName() const;
~Person();
};
Person::Person(char * nm): name(NULL){
if (nm != NULL)
{
name = new char[strlen(nm)+1];
strcpy(name,nm);
}
}
const char * Person::GetName() const {
if (name != NULL )
11

See Lecture No.9 for more details about copy constructor.

188

Virtual University of Pakistan

Object Oriented Programming (CS304)

VU

return name;
else
return "NULL";
}
Person::~Person(){
delete []name;
}
/*
Child Class Student
class Student: public Person{
char* major;
public:
Student(char * , char * );
void Print() const;
~Student();
};

*/

Student::Student(char *_name, char *_maj) : Person(_name),

major(NULL)
{
if (_maj != NULL) {
major = new char [strlen(_maj)+1];
strcpy(major,_maj);
}
}
void Student::Print() const{
cout << "Name: "<< GetName() << endl;
cout << "Major: " << major << endl;
}
Student::~Student(){
delete []major;
}
int main(){
Student sobj1("Ali","Computer Science");
Student sobj2 = sobj1;
//Student sobj2(sobj1); Similar to above statement to call copy
constructor
sobj2.Print();
system("PAUSE");
return 0;
}

Code Description:

Virtual University of Pakistan

189

Object Oriented Programming (CS304)

VU

The code has two classes Person and Student this code is also manifestation of IS-A
kind of relationship (public inheritance in c++).
Person class has an attribute name to store name of any person as all persons have
specific names.
Student class has an attribute major to store the major of the student.
Person has the following in its public interface:
Constructor
GetName()
Destructor()

// one parameter name with default value NULL

// getter function
// used due to dynamic memory allocation

Student has the following in its public interface:

Constructor
Print()
Destructor

// two parameter constructor with default values NULL

// showing student attributes
// used to free memory dynamically allocated

The constructors code of student class is as shown below,

Student::Student(char *_name, char *_maj) : Person(_name), major(NULL)
{
if (_maj != NULL) {
major = new char [strlen(_maj)+1];
strcpy(major,_maj);
}
}
The constructor of student class is taking two char *s as inputs, one is name of
student and second is major subject of that student.
In initialization list we are initializing major with null value and we are also calling
base class constructor of person class as we discussed today base class part of an
objects is created first and then its derived class part is created.
Then there is code of student constructor in which are checking maj for NULL if it is
not (some value has passed for maj subject) we are creating memory equal to passed
value and assigning it to char * major.
The print function code of student class is given below,
void Student::Print() const {
cout << "Name: "<< GetName() << endl;
cout << "Major: " << major << endl;
}
In print method we are doing two things first we are printing student name using
person GetName method and then we are showing student major.

190

Virtual University of Pakistan

Object Oriented Programming (CS304)

VU

The main function is given below,

int main(){
Student sobj1("Ali", "Computer Science");
Student sobj2 = sobj1;
//Student sobj2(sobj1); Similar to above statement to call copy
constructor
sobj2.Print();
system("PAUSE");
return 0;
}
In main we are creating one object of Student class with name sobj1 with values
Ali and Study Program for name and major,
Student sobj1("Ali","Computer Science");
Then we are assigning sobj1 to a new student object sobj2 with the line given below,
Student sobj2 = sobj1;
This line will invoke copy constructor of Student class as we are creating an object of
student class in terms of another object that already exist for student class.
This line is exactly same as commented line,
//Student sobj2(sobj1);
Here one object sobj1 values have been assigned to newly created object using
default compiler generated copy constructor, this copy constructor is created by
compiler in the same way as compiler generates default constructor in case user has
not defined default constructor for that class.
Then we are calling print method of student class to ensure that we values have been
assigned to new object as well,
sobj2.Print();
The output of this code will be as follows:
C++ Out Put:
Name: Ali
Major: Computer Science
As we know compiler may generate copy constructor for any class, if needed. In this
case when we are assigning derived class object sobj1 to another newly created

Virtual University of Pakistan

191

Object Oriented Programming (CS304)

VU

object sobj2, derived class object Copy constructor will be invoked which in turn will
call the copy constructor of the Base class as base class anonymous object is created
first and then derived part is created.
Shallow Copy
Compiler by default uses shallow copy so compiler generated default copy
constructor will also use shallow copy which simply copies all values of already
existing object to newly created object as shown in the diagram below,

Sobj1
Ali

char * name

Computer Science

Char * major

Memory
Line No.1:

Student sobj1("Ali","Computer Science");

Sobj1
char * name
Char * major
Ali
Shallow Copy
Computer Science
Sobj2
Memory
char * name
char * major

Line No.2:

Student sobj2 = sobj1; [Shallow Copy]

Fig 24.1 In shallow copy both char * name and char * major of both objects sobj1 and sobj2 will
start pointing to the same memory location.

192

Virtual University of Pakistan

Object Oriented Programming (CS304)

VU

Compiler generates copy constructor for derived class, calls the copy constructor of
the base class and then performs the shallow copy of the derived classs data
members.
The problem with shallow copy remains same as was in the case of copy constructor
for a single class that it doesnt work correctly in case of dynamic memory because in
that case we are only storing pointer of the dynamic memory in class when we will
do copy only the pointer value will be copied and pointers of both objects of student
class will start pointing to same memory location as shown in the diagram now
suppose obj1 is destroyed it frees its memory allocated using new operator using its
destructor the result will be that obj2 pointer will also become invalid. The solution is
Deep Copy in deep copy we will write copy constructor code by our self to ensure
that when one object is copied from other object new dynamic memory is allocated
for it as well so that it doesnt rely on previous object memory.

Deep Copy:
Let we first write code for deep copy in our base class Person.
For this we will write our own copy constructor as we wrote our own default
constructor as shown below,

Person::Person(const Person & rhs): name(NULL){

/* Code for deep copy*/
if (rhs.name != NULL)
{
name = new char[strlen(rhs.name)+1];
strcpy(name,rhs.name);
}
}
int main(){
Student sobj1("Ali","Computer Science");
Student sobj2 = sobj1;
//Student sobj2(sobj1); Similar to above statement to call copy
constructor
sobj2.Print();
system("PAUSE");
return 0;
}

When we will run this code the output will again be same,
C++ Out Put:

Virtual University of Pakistan

193

Object Oriented Programming (CS304)

VU

Name: Ali
Major: Computer Science

In this case as we have written our self written copy constructor of base class so
Compiler will generate copy constructor for only derived class, will call our written
copy constructor of the base class so it will perform the Shallow copy for derived
classs data members and Deep copy for base class as our base class has our own
written copy constructor with Deep copy code.
Now let us write our own copy constructor for derived class as well as shown below,
Student::Student(const Student & rhs) : major(NULL)
{
if (rhs.major != NULL) {
major = new char [strlen(rhs.major)+1];
strcpy(major,rhs.major);
}
}
int main(){
Student sobj1("Ali","Computer Science");
Student sobj2 = sobj1;
//Student sobj2(sobj1); Similar to above statement to call copy
constructor
sobj2.Print();
system("PAUSE");
return 0;
}

But now when we executed this code the output is as follows:

C++ Out Put:

Name: NULL
Major: Computer Science

Name of sobj2 was not copied from sobj1 or we can say that only copy constructor of
derived class executed but base class copy constructor did not execute.
The reason for this error is that we have not explicitly called copy constructor of base class
from derived class. And compiler called base class default constructor by itself to complete
the object creation process.

194

Virtual University of Pakistan

Object Oriented Programming (CS304)

VU

We can check whether base class default constructor was called or not by adding a
cout statement in our base class default constructor as shown below,
24.1. Modified Default Constructor
Person::Person(char * nm): name(NULL){
if (nm != NULL)
{
name = new char[strlen(nm)+1];
strcpy(name,nm);
}
else {
cout<<"In Person default constructor..\n";
}
}
int main(){
Student sobj1("Ali","Computer Science");
Student sobj2 = sobj1;
//Student sobj2(sobj1); Similar to above statement to call copy
constructor
sobj2.Print();
system("PAUSE");
return 0;
}
Now when we executed this code the output was,
C++ Out Put:
In Person default constructor..
Name: NULL
Major: Computer Science

Now we modify our derived class copy constructor code such that it explicitly call
copy constructor of base class from its initialization list,

Example
Person::Person(const Person & rhs): name(NULL){
if (rhs.name != NULL)
{
name = new char[strlen(rhs.name)+1];
strcpy(name,rhs.name);
}

Virtual University of Pakistan

195

Object Oriented Programming (CS304)

VU

}
Student::Student(const Student & rhs) : Person(rhs),
major(NULL)
{
if (rhs.major != NULL) {
major = new char [strlen(rhs.major)+1];
strcpy(major,rhs.major);
}
}
Now our main function shown previously will give following output
C++ Out Put:
Name: Ali
Major: Computer Science

The output again is correct now so this is our final code with user defined copy constructors
for both base and derived class and derived class copy constructor calling base class copy
constructor.
In table below we have shown the steps number wise involved in creation of an object using
copy constructor when we executed following lines in main,
int main(){
Student sobj1("Ali","Computer Science");
Student sobj2 = sobj1;
// Here copy constructor of student class will be called
return 0;
}

3
4
5
1
6
2
7

Person::Person(const Person &rhs) :

name(NULL) {
//code for deep copy
}
Student::Student(const Student & rhs) :
major(NULL),
Person(rhs){
//code for deep copy
}
Table 24.1

Assignment Operator

196

Virtual University of Pakistan

Object Oriented Programming (CS304)

VU

Compiler also generates code for assignment operator for a class, if needed. In the case
of inheritance when we assign one derive class object to another derive class object
compiler uses assignment operator. Derived class copy assignment operator is
invoked which in turn calls the assignment operator of the base class.
There are lot of similarities between copy constructor and assignment operator these
are,
In case our class involves dynamic memory allocation we had to write assignment
operator code by our self as we had to write the user defined code for copy
constructor.
Similarly derived class assignment operator has to call base class assignment
operator to complete the assignment process as derived class also contains implicit
base class part in case of compiler generated assignment operator compiler does it by
itself but if case we programmer has to call assignment operator of base class
explicitly in case of user defined assignment operator.
The example given below explain this concept in detail,
Example
class Person{
public:
Person & operator =
(const Person & rhs){
cout << Person Assignment;
// Code for deep copy assignment
}
};
Example
class Student: Public Person{
public:
Student & operator = (const Student & rhs){
cout<< Student Assignment;
// Code for deep copy assignment
}
};
int main()
{
Student sobj1, sboj2(Ali, CS);
sobj1 = sobj2;
return 0;
}

Output
Student Assignment
The assignment operator of base class is not called
Calling Base Class member functions:

Virtual University of Pakistan

197

Object Oriented Programming (CS304)

VU

We can not call base class assignment operator from derived class using initialization
list.
There are two ways of calling base class function from derived class one is calling
base class function explicitly (mentioning base class name) and the other is calling
base class function implicitly (without mentioning base class name).
Explicit Way:
In the code below we are calling base class const char *GetName() method explicitly,

const char * Person::GetName() const {

...
...
}
void Student::Print() const{
//cout << "Name: "<< GetName() << endl;
cout << "Name: "<< Person::GetName() << endl;
/*explicit call to base class GetName method (if we even do not write base class
name even then call will work as derived class can call its base class public
methods*/
cout << "Major: " << major

<< endl;

}
Implicit Way:
We can call the same method implicitly as well as shown below,
const char * Person::GetName() const {
...
...
}
void Student::Print() const {
/* Implicit call to base class GetName method using this pointer*/
cout <<"Name: "<<static_cast<const Person
&>(*this).GetName()<<endl;
cout << "Major: " << major << endl;
}

198

Virtual University of Pakistan

Object Oriented Programming (CS304)

VU

Assignment Operator
In the same way we can also call assignment operator in base class in two ways also,
Calling assignment operator of base class explicitly
Calling assignment operator of base class implicitly
Explicitly Calling operator =
Student & Student ::operator =(Student & rhs){

Person::operator = (rhs); /*Explicit call to base class*/

/*Student object rhs will be converted to Person object in function call
above as derived class IS A is a kind of Base class*/
/*Implicit calls to base class part*/
//static_cast<Person &>(*this) = rhs;
//Person(*this) = rhs;
//(Person)*this = rhs;
/*All the above three statements have the same meaning.*/
if (major != NULL)
delete [] major; // deleting previous allocated memory for major
if(rhs.major != NULL){
major = new char[strlen(rhs.major)+1];
strcpy(major,rhs.major);
}
}
Implicitly Calling operator =
Student & Student ::operator =(Student & rhs){

//Person::operator = (rhs); /*Explicit call to base class*/

/*Student object rhs will be converted to Person object in function call
above as derived class IS A is a kind of Base class*/
/*Implicit calls to base class part*/
static_cast<Person &>(*this) = rhs;

/* c++ way of type casting

*/
//Person(*this) = rhs; /* c way of type casting */
//(Person)*this = rhs; /* c way of type casting */
/*All the above three statements have the same meaning.*/
if (major != NULL)

Virtual University of Pakistan

199

Object Oriented Programming (CS304)

VU

delete [] major; // deleting previous allocated memory for major

if(rhs.major != NULL){
major = new char[strlen(rhs.major)+1];
strcpy(major,rhs.major);
}
}

Appendix:
Type conversion:
C++ is strongly typed language mean we can not use one type instead of other type
we need to convert one type in other type before using it.
This conversion from one type to another type is called casting.
Casting:
Casting can be done in two ways ,
Implicit casting conversion :
Conversions that compiler can perform automatically using build in casting
operations without casting request from programmer.
Explicit casting conversion:
Conversion in which programmer requests the compiler to convert one type to
another.
There are four types of casting operators that programmer can use to convert one
type ot another as given below,
The C++ draft standard includes the following four casting operators,
1.
2.
3.
4.

static_cast
const_cast
dynamic_cast
reinterpret_cast

In perspective of inheritance casting can be of two kinds,

a. Up casting (casting from derived to base class)

b. Down casting (casting from base to derived class)

200

Virtual University of Pakistan

Object Oriented Programming (CS304)

Base Class

VU

Base Class

Upcasting
Derived Class

Downcasting
Derived Class

Fig b.

Fig a.

Downcasting:
Example: (Base to Derive class)

Dynamic cast: (polymorphic behavior )

#include <iostream>
#include <stdlib.h>
using namespace std;
class Base{
public:
virtual void function() { cout<<"I am in Base"<<endl; }
};
class Derived : public Base {
public:
void function() { cout<<"I am in Derived"<<endl; }
};
int main()
{
Base * base = new Derived();
base->function();
Derived * derived = dynamic_cast<Derived*>(base);
derived->function();
system("PAUSE");
return 0;
}
Reinterpret cast:
#include <iostream>
#include <stdlib.h>
using namespace std;

Virtual University of Pakistan

201

Object Oriented Programming (CS304)

VU

class Base{
public:
void function() { cout<<"I am in Base"<<endl; }
};
class Derived : public Base {
public:
void function() { cout<<"I am in Derived"<<endl; }
};
int main()
{
Base * base = new Derived();
base->function();
Derived * derived = reinterpret_cast<Derived*>(base);
derived->function();
system("PAUSE");
return 0;
}

Note:
1. char * is c++ built in data type that can be used exactly like char arrays in fact
char arrays are also treated as char * in c++ so we used char * where we
needed to use char array.
2. In video lecture at duration 24 min there should only be one memory location
to store values, Ali newly created object pointer will be NULL for name and
will not point to any memory location.
References:
http://www.acm.org/crossroads/xrds3-1/ovp3-1.html (Casting Tutorial see this link
for more details about casting)

Open Source Editor Note Pad++ has been used in this document for code
highlighting

202

Virtual University of Pakistan

Object Oriented Programming (CS304)

VU

Lecture No.25
Overriding Member Functions of Base Class in Derived Class (Function
Overriding)
Derived class can override the member functions of its base class. To override a
function the derived class simply provides a function with the same signature12
(prototype) as that of its base class
Parent
...
Func1

Child
...
Func1

Overriding
25.1. Overloading vs. Overriding
Function Overloading is done within the scope of one class whereas Function
Overriding is done in scope of parent and child classes (inheritance).

12

same name, parameters and return type

Virtual University of Pakistan

203

Object Oriented Programming (CS304)

VU

class Parent {
public:
void Func1();
void Func1(int);
};
class Child: public
Parent {
public:
void Func1();
};

Function
Overriding:
Two functions in two
base derived classes
same prototypes.

Function Overloading:
Two functions in same class
with same name but different
parameters and return type.

class Parent {
public:
void Func1();
void Func1(int);
};

class Parent {
public:
void Func1();
void Func1(int);
};

class Child: public

Parent {
public:
void Func1();
};

class Child: public

Parent {
public:
void Func1();
};

Overriding within the scope of single class is error due to duplicate declaration (two
member functions with same prototype)
class Parent {
public:
void Func1();
void Func1();
};

//Error

Overriding Member Functions of Base Class

Derive class can override member function of base class such that the working of
function is totally changed.
Example
class Person{
public:
void Walk();
};
class ParalyzedPerson: public Person{
public:
void Walk();
};

204

Virtual University of Pakistan

Object Oriented Programming (CS304)

VU

Overriding Member Functions of Base Class

Derive class can override member function of base class such that the working of
function is similar to former implementation.
Example
class Person{
char *name;
public:
Person(char *=NULL);
const char *GetName() const;
void Print(){
cout << Name: << name
<< endl;
}
};
class Student : public Person{
char * major;
public:
Student(char * aName, char* aMajor);
void Print(){
cout <<Name: << GetName()<<endl
<< Major: << major<< endl;
}
};
int main(){
Student a(Ahmad, Computer Science);
a.Print();
return 0;
}
Output:
Name: Ahmed
Major: Computer Science

Overriding Member Functions of Base Class

Derive class can call base class member function from its overridden member
function to perform the base class part and then can perform its own tasks for
example student class can call base class person method Print to show name of
student and then can show study program of student by itself, this approach is in
accordance with the Object Oriented Programming principles which says that each
class should perform its tasks by itself, this is shown in example below,
Example
class Student : public Person{
char * major;
public:
Student(char * aName, char* m);

Virtual University of Pakistan

205

Object Oriented Programming (CS304)

VU

void Print(){
Print();
//Calling Print of Person to print student name
cout<<Major: << major <<endl; // Displaying study program of
student
}
};
int main(){
Student a(Ahmad, Computer Science);
a.Print();
return 0;
}

But there will be no output on the screen after executing this code due to minor mistake it is
due to the reason that our both classes Student and Person have methods with name Print and
we are calling Print method of Person from Student Print method compiler will call to Print of
Student class again and again recursively as we are calling Print method form Student class.
In this case it is necessary to mention that we are calling base class Print method as
shown below,
class Student : public Person{
char * major;
public:
Student(char * aName, char* m);
void Print(){
Person::Print();
cout<<Major: << major <<endl;
}
...
};
Example
int main(){
Student a(Ahmad, Computer Science);
a.Print();
return 0;
}
Output
Name: Ahmed
Major: Computer Science

Overriding Member Functions of Base Class

206

Virtual University of Pakistan

Object Oriented Programming (CS304)

VU

As we see previously when we use pointers to call overridden methods they are
called according to static type of pointer as shown below,
Example
int main(){
Student a(Ahmad, Computer Science);
Student *sPtr = &a;
sPtr->Print(); // static type of sPtr is Student * so Student Print method will
be called
Person *pPtr = sPtr; // static type of pPtr is Person * so Person Print method
will be called
pPtr->Print();
return 0;
}
Output:
Name: Ahmed
Major: Computer Science
Name: Ahmed

This may be undesirable and to avoid this situation avoid using simple method overriding in
this way its modified form (virtual functions will be covered in coming lectures).
25.2. Hierarchy of Inheritance
We represent the classes involved in inheritance relation in tree like hierarchy

Example

Virtual University of Pakistan

207

Object Oriented Programming (CS304)

VU

GrandParent

Parent1

Parent2

Child1

Child2

Direct Base Class

A direct base class is explicitly listed in a derived class's header with a colon (:)
class Child1:public Parent1 {
// Here Parent1 is Direct Base Class of Child1

};
Indirect Base Class
An indirect base class is not explicitly listed in a derived class's header with a colon
(:)
It is inherited from two or more levels up the hierarchy of inheritance
class GrandParent {

};
class Parent1: public GrandParent {

};
class Child1:public Parent1 { // Here GrandParent is InDirect Base Class of
Child1

};

208

Virtual University of Pakistan

Object Oriented Programming (CS304)

VU

Lecture No.26
26.1. Base Initialization
We saw in the previous lectures, that in the case of copy constructor we have to call
base class constructor from the initialization list of child class because implicit base
class object is created first and then derived class object is created. Now we see this
concept in detail in the perspective of class hierarchy,

The child can only call constructor of its direct base class to perform its
initialization using its constructor initialization list.
The child cannot call the constructor of any of its indirect base classes to
perform their initialization using its constructor initialization list

It is explained in example code given below,

Example
class GrandParent{
int gpData;
public:
GrandParent() : gpData(0){...}
GrandParent(int i) : gpData(i){...}
void Print() const;
};
class Parent1: public GrandParent{
int pData;
public:
Parent1() : GrandParent(), pData(0) {}
};
class Child1 : public Parent1 {
public:
Child1() : Parent1() {...}
Child1(int i) : GrandParent (i) //Error: Child1 can not call its
indirect base class GrandParent Constructor from its constructor
initialization list.
{...}
void Print() const;
};

Overriding in case of class hierarchy:

In class hierarchy Child class can override the function of any of its Parent class
(direct or indirect) as shown in the diagram below (we see this concept in detail in
types of inheritance given below),

Virtual University of Pakistan

209

Object Oriented Programming (CS304)

VU

GrandParent
Print()

Parent1

Child1
Print()

Example
void GrandParent::Print() {
cout << GrandParent::Print
<< endl;
}
void Child1::Print() {
cout << Child1::Print << endl;
}
int main(){
Child1 obj;
obj.Print();
obj.Parent1::Print();
obj.GrandParent::Print();
return 0;
}

Output
Child1::Print
GrandParent::Print
GrandParent::Print

26.2. Types of Inheritance

There are three types of inheritance
Public
Protected
Private
We can use these keywords (public, private or protected) to specify the type of
inheritance

210

Virtual University of Pakistan

Object Oriented Programming (CS304)

VU

a. Public Inheritance

class Child: public Parent {};

Member access in
Base Class

Derived Class

Public

Public

Protected

Protected

Private

Hidden

b. Protected Inheritance

class Child: protected Parent {};

Member access in
Base Class

Derived Class

Public

Protected

Protected

Protected

Private

Hidden

c. Private Inheritance

class Child: private Parent {};

Member access in
Base Class

Derived Class

Public

Private

Protected

Private

Private

Hidden

If the user does not specifies the type of inheritance then the default type is private
inheritance,

Virtual University of Pakistan

211

Object Oriented Programming (CS304)

VU

class Child: private Parent {}

is equivalent to,
class Child: Parent {}
We have seen public inheritance in previous lecture, now we see private and
protected inheritance in detail,
26.3. Private Inheritance
We use private inheritance when we want to reuse code of some class. Private
Inheritance is used to model Implemented in terms of relationship
Example

Suppose we have a class collection to store element collection as shown

below,
class Collection {
...
public:
void AddElement(int);
bool SearchElement(int);
bool SearchElementAgain(int);
bool DeleteElement(int);
};

As you can see it supports the following methods,

AddElement: to add elements in collection
SearchElement: search any element in collection it will true as soon as any
element will be found
SearchElementAgain: finds second instance of any element in collection it
will return true as soon as it will find any duplicate entry of any element.
DeleteElement: to delete any entry from collection
You can see that Class collection allows duplicate elements.
Suppose now we want to implement class Set, class Set has very similar
functionality as that of collection class with the difference that Set class can
not allow duplicate elements in it, so we can use the concept of inheritance
here, we can derive class Set from class Collection.
But we can not use public inheritance here as it allows interface (all functions)
of base class to be accessed using derived class object, but we dont want to
allow all functions of class Collection to act on Set class object. We only want
to use some functions of Collection class (Base class) in derived Set class for to
implement Set class functionality so we will use private inheritance here the
two main advantages we will achieve through private inheritance in this case
are,
1. Specialization of class according to set class (removing extra features)

212

Virtual University of Pakistan

Object Oriented Programming (CS304)

VU

2. Making interface of collection class inaccessible from outside world

using class set reference.
Class Set
class Set: private Collection {
private:
...
public:
void AddMember(int);
bool IsMember(int);
bool DeleteMember(int);
};
void Set::AddMember(int i){
if (! IsMember(i) )
AddElement(i);
}
bool Set::IsMember(int i){
return SearchElement(i);
}

Virtual University of Pakistan

213

Object Oriented Programming (CS304)

VU

Lecture No.27
Previous Lecture Review:
In last lecture we saw the use of private inheritance using the example of classes
Collection and Set, we saw that class collection was allowing duplicate elements but
set class can not allow duplicate elements in it however other functionality was same,
so we privately inherited class set from collection and added only those member
functions in derived class set that were related to set class, the two main advantages
we achieved through private inheritance in this case were,
1. Specialization of class collection according to set class (removing extra
features)
2. Making interface of collection class inaccessible from outside world using
class set reference, to avoid any unintentional use of collection class restricted
features in set class.
The code of both classes in shown below,

Class Collection
class Collection {
...
public:
void AddElement(int);
bool SearchElement(int);
bool SearchElementAgain(int);
bool DeleteElement(int);
};

Class Set
class Set: private Collection {
private:
...
public:
void AddMember(int);
bool IsMember(int);
bool DeleteMember(int);
};

As we discussed that we achieve specialization using private inheritance so need to

see specialization (Specialization concept was discussed in Lecture No.4) again in
detail,
27.1. Specialization (Restriction)
In specialization derived class is behaviourally incompatible with the base class.
Behaviourally incompatible means that base class cant always be replaced by the
derived class.
Specialization is represented by Implemented in terms of relationship.
Specialization (Restriction) can be implemented using both private and protected
inheritance.

214

Virtual University of Pakistan

Object Oriented Programming (CS304)

VU

Person
age : [0..125]
age = a
setAge( a )

Adult
age : [18..125]

If age < 18 then

error
else
age = a

setAge( a )

Example Specialization (Restriction)

class Person{

protected:
int age;
public:
bool SetAge(int _age){
if (_age >=0 && _age <= 125) {
age = _age;
return true;
}
return false;
}
};
class Adult : private Person {
public:
bool SetAge(int _age){
if (_age >=18 && _age <= 125) {
age = _age;
return true;
}
return false;
}
};

Essential properties of Private Inheritance:

Virtual University of Pakistan

215

Object Oriented Programming (CS304)

VU

1. In Private Inheritance only member functions and friend classes or functions

of a derived class can convert pointer or reference of derived object to that of
parent object
Example

class Parent{
};
class Child : private Parent{
};
int main(){
Child cobj;
Parent *pptr = & cobj;
return 0;
}

//Error

void DoSomething(const Parent &);

Child::Child(){
Parent & pPtr = static_cast<Parent &>(*this); // fine
DoSomething(pPtr);
//
DoSomething(*this); // this single line is equal to two lines above.
}

2. As was in the case of public inheritance child class object has an anonymous
object of parent class.
3. As was in the case of public inheritance the default constructor and copy
constructor of parent class are called when constructing object of derived
class.
Example

class Parent{
public:
Parent(){
cout << Parent Constructor;
}
Parent(const Parent & prhs){
cout << Parent Copy Constructor;
}
};
class Child: private Parent{
public:
Child(){
cout << Child Constructor;
}

216

Virtual University of Pakistan

Object Oriented Programming (CS304)

VU

Child(const Child & crhs) : Parent(crhs){

cout << Child Copy Constructor;
}
};
int main() {
Child cobj1; // default constructor will be invoked
Child cobj2 = cobj1; // copy constructor will be invoked
//Child cobj2(cobj1); // another way of calling copy constructor
return 0;
}

Output:
Parent Constructor
Child Constructor
Parent Copy Constructor
Child Copy Constructor
4. In private inheritance the derived class that is more than one level down the
hierarchy cannot access the member functions of grand parent class as public
and protected members functions of derived class become private members
of privately derived class for all practical purposes. For example see the code
below here Child class is derived class that is more than one level down the
hierarchy and hence can not access the member functions of GrandParent
class.
class GrandParent{
public :
void DoSomething();
};
class Parent: private GrandParent{
void SomeFunction(){
DoSomething();
}
};
class Child: private Parent
{
public:
Child() {
DoSomething();
}
};

//Error

5. In private inheritance the derived class that is more than one level down the
hierarchy cannot convert the pointer or reference to that of GrandParent , for
example in code given below Child class can not convert its reference to its

Virtual University of Pakistan

217

Object Oriented Programming (CS304)

VU

GrandParent reference. The reason is same that in private inheritance we

implement specialization and we restrict all features of base class to privately
derived class only and its friend classes or functions.
void DoSomething(GrandParent&);
class GrandParent{
};
class Parent: private GrandParent{
public:
Parent() {DoSomething(*this);}
};
class Child: private Parent {
public:
Child()
{
DoSomething(*this);
}
};

//Error

27.2. Protected Inheritance

If a class D has been derived using protected inheritance from class B (If B is a
protected base and D is derived class) then public and protected members of B can be
accessed by member functions and friends of class D and classes derived from D.
Protected inheritance is used to build class hierarchy using Implemented in terms
of relationship.
If B is a protected base and D is derived class then member functions and friends of
class D and classes derived from D can access member functions of class B, (note that
in private inheritance only derived class can access the member functions of base
class). So we can say that protected inheritance lies between public and private
inheritance.

218

Virtual University of Pakistan

Object Oriented Programming (CS304)

VU

class GrandParent{
public :
void DoSomething();
};
class Parent: protected GrandParent{
void SomeFunction(){
DoSomething();
}
};
class Child: protected Parent
{
public:
Child()
{
DoSomething();
}
};

27.3. Properties of Protected Inheritance

If B is a protected base and D is derived class then only friends and members of D
and friends and members of class derived from D can convert D* to B* or D& to B&
(in private inheritance only derived class or its friends can convert pointer to base
class)
void DoSomething(GrandParent&);
class GrandParent{
};
class Parent: protected GrandParent{
};
class Child: protected Parent {
public:
Child()
{
DoSomething(*this);
}
};

Importance of Private and Protected inheritance:

As we have seen that private and protected inheritance is being used for implementing
Implemented in terms of relationship so it is important that we limit different features of
base class in child classes that is what we achieve using either private or protected inheritance
according to our requirement.

Virtual University of Pakistan

219

Object Oriented Programming (CS304)

VU

Comparison of public, protected and private inheritance:

We can show accessibility of public member functions of base class in derived classes in
these three different types of inheritance as follows,
Public Inheritance

Protected Inheritance

Private Inheritance

Base Class

Base Class

Base Class

Accessibility

Accessibility

Accessibility

Derived 1

Yes

Derived 1

Yes

Derived 1

Yes

Derived 2
Derived 3

Yes

Derived 2
Derived 3

Yes

Derived 2
Derived 3

No


Main
(Outside
world)

Yes

Main
(Outside
world)

No


Main
(Outside
world)

No


Private data members will NOT be accessible in any derived class or in main
function.
Protected data members will become private data members in case of private
inheritance and protected data members of derived class in case of protected
inheritance.
A Good Programming Exercise:
A good programming exercise would be that you write a program that shows
accessibility of all types of members functions for all types of inheritance in derived
class/es.

220

Virtual University of Pakistan

Object Oriented Programming (CS304)

VU

Lecture No.28
28.1. Virtual Functions

Problem Statement:
Develop a function that can draw different types of geometric shapes from an array
Shape Hierarchy:
We have to implement following shape hierarchy for this,

Shape
draw
calcArea

Line
draw
calcArea

Circle
draw
calcArea

Triangle
draw
calcArea

Problem Description:
a. We want to implement this shape hierarchy in such a manner that our
function will take Shape pointers array (Shape * []) and this array size as
parameters and then it will draw the appropriate shape.
b. This Shape pointer array (Shape * []) may store pointers to all kinds of Shapes
(Line, Circle and Triangle classes) as these classes have IS-A relationship with
Shape class (We know base class pointer can store pointer of any of its
derived classes in case of public inheritance IS-A relationship).
c. The purpose of implementing this hierarchy in such a way is that we can
avoid complex code of calling draw function for each class separately after
checking the class type, instead we can use a single function call in a loop to
draw all kinds of Shapes as shown below,
void drawShapes(Shapes *array[], int size){
for (int i = 0 ; i < size ; i ++)
array[i]->draw(); // this function call will work for all types of
//Shapes (Line, Circle and Triangle)
}

Virtual University of Pakistan

221

Object Oriented Programming (CS304)

VU

Implementation:
It can be done using concept of inheritance as there exists IS-A relationship
between general Shape class and derived Line, Circle and Triangle classes.
So in code below, we wrote a general Shape class and then we publicly derived Line,
Circle and Triangle classes from it, in main we create objects of those Shapes that we
needed like Line, Circle and Triangle and stored them in single shape array, then we
called drawShapes function by passing this array to draw appropriate shapes.

class Shape {

protected:
char _type;
public:
Shape() { }
void draw(){ cout << Shape\n; }
int calcArea() { return 0; }
char getType() { return _type; }
}
class Line : public Shape {
public:
Line(Point p1, Point p2) {
}
void draw(){ cout << Line\n; }
}
class Circle : public Shape {
public:
Circle(Point center, double radius) {
}
void draw(){ cout << Circle\n; }
int calcArea() { }
}
class Triangle : public Shape {
public:
Triangle(Line l1, Line l2, double angle)
{
}
void draw(){ cout << Triangle\n; }
int calcArea() {
}
}
int main() {

222

Virtual University of Pakistan

Object Oriented Programming (CS304)

VU

Shape* _shape[ 10 ];
Point p1(0, 0), p2(10, 10);
shape[1] = new Line(p1, p2);
shape[2] = new Circle(p1, 15);
void drawShapes( shape, 10 );
return 0;
}
/*Function drawShapes()*/
void drawShapes(Shape* _shape[], int size) {
for (int i = 0; i < size; i++) {
_shape[i]->draw();
}
}

After compiling this code we will see the following output,

Sample Output
Shape
Shape
Shape
Shape

As you have seen this code is not showing correct output, it is showing text Shape
again and again instead of showing appropriate class name that was stored in Shape
array like Line, Circle or Triangle so there is some logical error in this code.
Where is problem?
Problem is that as we stored our derived classes (Line, Circle and Triangles) Objects
in Shape pointer array and then called draw method; every time draw method of
Shape class was called.
Why?
Due to the reason that static type of this array is Shape * so draw method is Shape
class will always be called whether pointer stored in it is Line, Circle or Triangle
pointer or Shape class pointer.
So we have to think about some modification,

Solution 1:
One solution is that we can modify our drawShapes function as follows,
void drawShapes( Shape* _shape[], int size) {
for (int i = 0; i < size; i++) {
// Determine object type with switch & accordingly call draw()
method

Virtual University of Pakistan

223

Object Oriented Programming (CS304)

VU

}
}
The required Switch Logic for this kind of functionality is given below,
switch ( _shape[i]->getType() )
{
case L:
static_cast<Line*>(_shape[i])->draw();
break;
case C:
static_cast<Circle*>(_shape[i])->draw();
break;

}
Equivalent If Logic is given below,
if ( _shape[i]->getType() == L )
static_cast<Line*>(_shape[i])->draw();
else if ( _shape[i]->getType() == C )
static_cast<Circle*>(_shape[i])->draw();

Sample Output:
Line
Circle
Triangle
Circle

You can see that this code is very complex for both switch or if else conditions.
But there are many Problems with this approach these are given below:
Delocalized Code
Suppose we have to write another function that prints area of each shape from an
input array. We have to write same switch or if else logic in that function
implementation also to work it correctly.
void printArea( Shape* _shape[], int size) {
for (int i = 0; i < size; i++) {
// Print shape name.
// Determine object type with
// switch & accordingly call
// calcArea() method.
}
}

224

Virtual University of Pakistan

Object Oriented Programming (CS304)

VU

Required Switch Logic

switch ( _shape[i]->getType() )
{
case L:
static_cast<Line*>(_shape[i])->calcArea(); break;
case C:
static_cast<Circle*>(_shape[i])->calcArea(); break;

}
The above switch logic is same as was in function drawArray() with difference of
name of function being called.
So this approach will result in delocalized code (same code in different places) with
following consequences,
Writing same code again and again at different places may produce errors as
programmer may forget to write switch cases.
Suppose we have added one more Shape in our program then we will have to add
one more switch case or if else condition in all functions where we have used this
logic if programmer by mistake forgets to add it in any single function whole
program will show incorrect output.
So due to above mentioned reasons this sort of code is very hard to maintain.
Solution?
To avoid switch, we need a mechanism that can select the message target (class)
automatically!
Polymorphism Revisited:
In OO model, polymorphism means that different objects can behave in different
ways for the same message (stimulus) consequently, sender of a message does not
need to know the exact class of receiver.
In other words when we have inheritance relationship and we have written basic
structure of our program correctly we need not to worry about the no. of classes in
our program. For example in case of shapes hierarchy (Line, Circle, Triangle and so
on) our program will keep working correctly if we add more shapes, draw method
of appropriate shape class should automatically be called, that is the benefit of Object
Oriented Programming,
void drawShapes(Shapes *array[], int size){
for (int i = 0 ; i < size ; i ++)
array[i]->draw(); // this function call will work for all types of
//Shapes (Line, Circle and Triangle)
}

Virtual University of Pakistan

225

Object Oriented Programming (CS304)

VU

But this kind of functionality was not being achieved in above approach in which we
had to write one separate case for each type of shape class.
To achieve this kind of functionality we have the concept of virtual functions we
make those functions virtual in base class which will be implemented by derived
classes according to their requirements.
28.2. Virtual Functions:
Virtual functions achieve exactly same kind of functionality that was achieved in
above code with complex code of switch statement.
Target class of a virtual function call is determined at run-time automatically.
In C++, we declare a function virtual by preceding the function header with
keyword virtual
Now we see how we can use virtual functions in case of our shape hierarchy,
28.3. Shape Hierarchy
Shape
draw
calcArea

Line
draw
calcArea

Circle
draw
calcArea

Triangle
draw
calcArea

We will make those functions virtual in shape that need to be overridden by derived
classes like Draw method,
class Shape {

virtual void draw();

}

class Shape {

virtual void draw();

virtual int calcArea();
}
class Line : public Shape {

virtual void draw() { cout<<Line\n;

}
}

226

Virtual University of Pakistan

Object Oriented Programming (CS304)

VU

class Circle : public Shape {

virtual void draw() { cout<<Circle.\n;

}
virtual int calcArea();
}
class Triangle : public Shape {

virtual void draw(){ cout<<Triangle\n;

}
virtual int calcArea();
}
void drawShapes(Shape* _shape[], int size) {
for (int i = 0; i < size; i++) {
_shape[i]->draw();
}
}

Sample Output:
Line
Circle
Triangle
Circle

Similarly if we have another function that will calculate and print area of different
shapes it will also be written in same way, (Now we have no need to add switch or if
else code)
void printArea(Shape* _shape[], int size) {
for (int i = 0; i < size; i++) {
// Print shape name
cout<< _shape[i]->calcArea();
cout << endl;
}
}

28.4. Static vs Dynamic Binding

Static binding means that target function for a call is selected at compile time
Dynamic binding means that target function for a call is selected at run time

Line _line;
_line.draw(); // Always Line::draw will be called

Virtual University of Pakistan

227

Object Oriented Programming (CS304)

VU

Shape* _shape = new Line();

_shape->draw(); // Shape::draw called if draw() is not virtual because of static type
of Shape *
Shape* _shape = new Line();
_shape->draw(); // Line::draw called as draw() is virtual in base Shape class

228

Virtual University of Pakistan

Object Oriented Programming (CS304)

VU

Lecture No.29
Previous Lecture Review:
The main concept of Polymorphism is that the same method can behave differently
according to the object with respect of which it has been called.
We combine all related classes having same kind of functionality in a class hierarchy
as shown below,

Shape
draw
calcArea

Line
draw
calcArea

Circle
draw
calcArea

Triangle
draw
calcArea

The draw and calcArea methods here are showing polymorphic behavior as they
exist in base as well as in derived classes, the method that will be called is decided at
runtime according to the nature of the calling object, for example in the case of above
class hierarchy if we call draw method with respect of Line class object, draw of Line
class will be called and if we call it with reference of Circle or Triangle class, draw
method of Circle or Triangle class will be called.
We achieve this type of polymorphism13 using virtual functions.
The advantage of this approach is that sender simply pass method call and
appropriate method is called automatically, as we saw in previous lecture in which
we simply used for loop to draw all kinds of shapes using shape pointers array.
We also saw the drawbacks of achieving same functionality without virtual functions
using switch statement that resulted in delocalized and complex code. Then we saw
solution of this problem using virtual functions.
29.1. Abstract Classes
In our Shape class hierarchy Shape class has no real world existence, there will be no
object in real world with name Shape there will always be an object of kind of shape
like Line, Circle or Triangle, this kind of class with no specific real world existence
are called Abstract Classes, as these classes are Abstract their behavior becomes also
abstract like Draw and calcArea methods in case of Shape class, the actual concepts
13
Polymorphism may exist in same class (function overloading) or it may exist in different classes
(function overriding) here we are referring to function overriding (same functions in base and derived
classes).

Virtual University of Pakistan

229

Object Oriented Programming (CS304)

VU

are realized in derived classes, like all kinds of shapes will have draw and calcArea
method, so we add these methods in general Shape class and all kinds of Shapes
inheriting for this general shape class given their own implementation of these
methods.
Abstract classs objects cannot be instantiated they are used for inheriting interface
and/or implementation, so that derived classes can give implementation of these
concepts.
29.2. Concrete Classes
Concrete classes Implements a concrete concept they can be instantiated they may
inherit from an abstract class or another concrete class. So far the classes we studied
were concrete classes.
29.3. Abstract Classes in C++
In C++, we can make a class abstract by making its function(s) pure virtual.
Conversely, a class with no pure virtual function is a concrete class (which object can
be instantiated)
29.4. Pure Virtual Functions
A pure virtual represents an abstract behavior and may have not implementation for
example draw method in Shape class represent abstract behavior as Shape class itself
doesnt have its existence in real world so there is no question of drawing it however
its derived concrete classes like Line, Circle and Triangle does have physical
existence and overridden draw method in these classes will have implementation . A
function is declared pure virtual by following its header with = 0.
virtual void draw() = 0;
A class having pure virtual function(s) becomes abstract
class Shape {

public:
virtual void draw() = 0;
}

Now when we will try to created object of our Shape class in our program compiler
will give an error as shown below,
Shape s;

// Error!

29.5. Shape Hierarchy

230

Virtual University of Pakistan

Object Oriented Programming (CS304)

VU

Shape
draw = 0

Line
draw

Circle

Quadrilateral

draw

Rectangle
draw

A derived class of an abstract class remains abstract until it provides implementation

for all pure virtual functions as shown below in other words we can say that at least
one class in bottom of class hierarchy should give implementation of pure virtual
function (Abstract classes are present at root or near root of the class hierarchy tree,
whereas concrete classes are near leaves of class hierarchy tree) see the code below of
above class heirarchi here Quadrilateral is also abstract class as it is derived from
Shape class but is not giving implementation of draw method we can also not create
its object and it is necessary to have one more derived class from quadrilateral that
gives implementation of draw method otherwise there will be compiler error as we
can not have all abstract classes in heirarchi there should be at least one concrete
class at leaf in class heirarchi,
class Quadrilateral : public Shape {

// No overriding draw() method

}

Quadrilateral q; // Error!
class Rectangle:public Quadrilateral{

public:
// void draw()
virtual void draw() { // once a function is declared as virtual it remains
virtual in all hierarchy
// function body
}
}

Rectangle r;
// OK

Virtual University of Pakistan

231

Object Oriented Programming (CS304)

VU

29.6. Virtual Destructors

The destructor is called according to static type of any class pointer for example if we
have saved derived class pointers in shape class pointers array as we did in previous
lecture when we will call destructor using delete operator the destructor of base class
Shape will be called as static type of array is Shape. This will destroy the base class
object only derived class object will not be destroyed this is explained in the example
code below,
class Shape {

public:
~Shape() {
cout << Shape destructor
called\n;
}
}

class Quadrilateral : public Shape {

public:
~Quadrilateral() {
cout << Quadrilateral destructor
called\n;
}
}
class Rectangle : public Quadrilateral {

public:
~Rectangle() {
cout << Rectangle destructor
called\n;
}
}
int main() {
Shape* pShape = new Rectangle();
delete pShape;
return 0;
}

232

Virtual University of Pakistan

Object Oriented Programming (CS304)

VU

When delete operator is applied to a base class pointer, base class destructor is called
regardless of the object type
Output
Shape destructor called

Result

You can see by diagram below that only base class part of object will be deleted other
parts will remain as it is this result in memory leak (wastage of memory),

pShape

pShape

Shape Part
Quad Part

Quad Part

Rect Part

Rect Part
After

Before

This issue will be solved in the same way as we solved the problem of calling
derived classes Draw methods using base class pointer by the use Virtual keyword.
Virtual Destructors
Make the base class destructor virtual as we made Draw method virtual in base class,
class Shape {

public:
virtual ~Shape() {
cout << Shape destructor called\n; }
}
class Quadrilateral : public Shape {

Virtual University of Pakistan

233

Object Oriented Programming (CS304)

VU

public:
virtual ~Quadrilateral() {
cout << Quadrilateral destructor called\n;
}
}
class Rectangle : public Quadrilateral {

public:
virtual ~Rectangle() {
cout << Rectangle destructor called\n;
}
}
Now base class destructor will run after the derived class destructor
int main() {
Shape* pShape = new Recrangle();
delete pShape;
return 0;
}

Output
Now you can see that output is correct all destructors are being called in correct
order,
Rectangle destructor called
Quadrilateral destructor called
Shape destructor called

Result

Now you can see result complete object is being deleted so there is no memory leak
(waste of memory),

234

Virtual University of Pakistan

Object Oriented Programming (CS304)

pShape

VU

pShape

Shape Part
Quad Part
Rect Part
After

Before

29.7. Virtual Functions Usage

Virtual function are used in two ways,

Virtual Functions can be used to when we want to inherit interface and

implementation (Simple virtual functions) mean base class as well as derived
class will have implementation.

virtual void draw();

Just inherit interface (Pure Virtual functions) mean only derived classes will
will have implementation base may not have implementation.

virtual void draw() = 0;

Inherit interface and implementation:

First case of simple virtual functions is useful when we have some derived classes that
will also not have implementation of virtual method for example Line is also also
Shape but it doesnt have area similarly if we had Point derived class it also doesnt
have any area in both cases we will simply not write implementation of calcArea()
method and hence calcArea of base Shape class will be called which will simply
display area as zero this is shown below,

Virtual University of Pakistan

235

Object Oriented Programming (CS304)

VU

Shape
draw = 0
calcArea

Line
draw

Circle
draw
calcArea

Triangle
draw
calcArea

class Shape {

virtual void draw() = 0; // pure virtual functions

virtual float calcArea() { // simple virtual functions
return 0;
}
}

So as calcArea() is simple virtual function so, each derived class of Shape inherits
default implementation of calcArea(), some classes may override this, such as Circle
and Triangle and Others may not, such as Point and Line.
Inherit interface only:
We have made draw method as pure virtual function because each shape will need
to draw whether it is simple point or line or any other shape so its suitable to be
decalred as pure virtual,
As draw() is Pure Virtual Function so, each derived class of Shape inherits interface
(prototype) of draw(), each concrete derived class has to provide body of draw() by
overriding it.
29.8. V Table

Now we see compiler keeps track of virtual functions and call them correctly
occording to nature of the object with respect to which they are being called,
Compiler builds a virtual function table (vTable) for each class having virtual
functions
A vTable contains a pointer for each virtual function,

236

Virtual University of Pakistan

Object Oriented Programming (CS304)

VU

Pointer to
1st virtual
function
Pointer to
2nd virtual
function
....
.
.

V Table

Consider the code below to see how virtual tables are created,
int main() {
Point p1( 10, 10 ), p2( 30, 30 );
Shape* pShape;
pShape = new Line( p1, p2 );
pShape->draw();
pShape->calcArea();
}
We are creating Line class object here and storing its refernence in Shape class
pointer and then polymorphically calling methods draw and calcArea as Shape
pointer is having Line class object reference so Line class method draw and calcArea
will be called we see how this actually happens,

Shape vTable

calcArea
Line object
0

draw

Shape

Line vTable

calcArea

point1 = p1

draw

point2 = p2

pShape

Virtual University of Pakistan

237

Object Oriented Programming (CS304)

VU

First of all when this code will be compiled v tables along with implementation code
of virtual and non virtual functions for all classes will be generated by compiler as
shown below,

Shape class implementation

Shape vTable
calcArea
0

draw

Line class implementation

Line vTable
calcArea
draw

When line object will be crated in main it will also have pointer to v table of Line
class as shown below,

238

Virtual University of Pakistan

Object Oriented Programming (CS304)

VU

Shape class implementation

Shape vTable

Line object

calcArea
0

draw

Shape
point1 = p1
point2 = p2

Line class implementation

pShape
Line vTable
calcArea

draw

29.9. Dynamic Dispatch (Dynamic Binding)

It happens in case of virtual functions, for non-virtual functions, compiler just
generates code to call the function.
In case of virtual functions, compiler generates code to
access the object
access the associated vTable
call the appropriate function
Conclusion

Virtual Functions should be added in code with care because they add,

Memory overhead due to V-Tables

Processing overhead due to extra pointer manipulation

However, this overhead is acceptable for many of the applications

Moral: Think about performance requirements before making a function virtual

Virtual University of Pakistan

239

Object Oriented Programming (CS304)

VU

Lecture No.30
30.1. Polymorphism Case Study: A Simple Payroll Application
We have studied polymorphism implementation in detail using virtual functions
now we see an example showing how it may be useful for us,
Problem Statement
Develop a simple payroll application for a company; there are three kinds of
employees in the system: salaried employee, hourly employee, and commissioned
employee. The system should take input as an array containing employee objects,
calculates salary polymorphically (according to employee object), and generates
report.
OO Model

String

Employee

pStr

name
taxRate
getName
calcSalary

String
operator =
operator <<

SalariedEmp

HourlyEmp

CommEmp

salary

hours
hourlyRate

sales
commRate

calcSalary

calcSalary

calcSalary

You can see that this model is very similar to Shape class hierarchy we saw
previously, its implementation in c++ is shown below,
Class Employee
class Employee {
private:
String name;
double taxRate;
public:
Employee( String&, double );
String getName();
virtual double calcSalary() = 0;
}
Employee::Employee( String& n, double tr ): name(n){
taxRate = tr;
}

240

Virtual University of Pakistan

Object Oriented Programming (CS304)

VU

String Employee::getName() {
return name;
}

Class SalariedEmp
class SalariedEmp : public Employee
{
private:
double salary;
public:
SalariedEmp(String&,double,double);
virtual double calcSalary();
}
SalariedEmp::SalariedEmp(String& n, double tr, double sal) : Employee( n, tr ) {
salary = sal;
}
double SalariedEmp::calcSalary() {
double tax = salary * taxRate;
return salary tax;
}

Class HourlyEmp
class HourlyEmp : public Employee {
private:
int hours;
double hourlyRate;
public:
HourlyEmp(string&,double,int,double);
virtual double calcSalary();
}
HourlyEmp ::HourlyEmp( String& n, double tr, int h, double hr ) : Employee( n, tr
){
hours = h;
hourlyRate = hr;
}
double HourlyEmp::calcSalary()
{
double grossPay, tax;
grossPay = hours * hourlyRate;
tax = grossPay * taxRate;
return grossPay tax;
}

Virtual University of Pakistan

241

Object Oriented Programming (CS304)

VU

Class CommEmp
class CommEmp : public Employee
{
private:
double sales;
double commRate;
public:
CommEmp( String&, double, double, double );
virtual double calcSalary();
}
CommEmp::CommEmp( String& n, double tr, double s, double cr ) : Employee(
n, tr ) {
sales = s;
commRate = cr;
}
double CommEmp::calcSalary()
{
double grossPay = sales * commRate;
double tax = grossPay * taxRate;
return grossPay tax;
}

A Sample Payroll
int main() {
Employee* emp[10];
emp[0] = new SalariedEmp( Aamir, 0.05, 15000 );
emp[1] = new HourlyEmp( Faakhir, 0.06, 160, 50 );
emp[2] = new CommEmp( Fuaad, 0.04, 150000, 10 );

generatePayroll( emp, 10 );
return 0;
}

Function that takes Empolyee poinsters array and calls appropriate getName
and calcSalary methods
void generatePayroll(Employee* emp[], int size) {
cout << Name\tNet Salary\n\n;
for (int i = 0; i < size; i++) {
cout << emp[i]->getName() << \t << emp[i]->calcSalary() <<
\n;
}
}

242

Virtual University of Pakistan

Object Oriented Programming (CS304)

VU

Sample Output
Name

Net Salary

Aamir
Fakhir
Fuaad

14250
7520
14400

Output is as required displaying employee names and their salary polymorphically.

Important point to note here is that Polymorphism always works with pointers of class
objects not with actual objects. In above example we used base Employee pointer array
to store pointers of derived classes of class Employee and then we polymorphically
called generatePayroll method to call getName and calcSal methods in a loop.
Never Treat Arrays Polymorphically:
If we use arrays of Objects Polymorphically then problem occurs as shown below,
30.2. Shape Hierarchy Revisited:

We have not made Shape class Abstract so that we can create its objects to show the point that
calling array of objects polymorphically results in errors,
See the shape hierarchy again,
class Shape {

public:
Shape();
virtual void draw(){
cout << Shape\n;
}
virtual int calcArea() { return 0; }
};
class Line : public Shape {

public:
Line(Point p1, Point p2);

Virtual University of Pakistan

243

Object Oriented Programming (CS304)

VU

void draw(){ cout << Line\n; }

}
drawShapes()
void drawShapes( Shape _shape[], int size ) {
for (int i = 0; i < size; i++) {
_shape[i].draw();
}
}
Polymorphism & Arrays
int main() {
Shape _shape[ 10 ];
_shape[ 0 ] = Shape();
_shape[ 1 ] = Shape();

drawShapes( _shape, 10 );
return 0;
}

In above code we created array of ten objects of Shape class and then passed it to drawShapes
function, note that in above code we passed Shapes array (Shapes []) instead of Shapes
pointers array (Shapes * []) to drawShapes function.
The code works draw method for each Object of Shape class is called and output is
displayed below,
Sample Output
Shape
Shape
Shape

Now we try to do polymorphism here by creating lines objects and passing to this
functions as we did previously in case of pointers with the difference that this time
we are passing objects array instead of pointers array.
int main() {
Point p1(10, 10), p2(20, 20),
Line _line[ 10 ];
_line[ 0 ] = Line( p1, p2 );
_line[ 1 ] = Line( p3, p4 );

drawShapes( _line, 10 );
return 0;
}

Sample Output:
Shape

244

Virtual University of Pakistan

Object Oriented Programming (CS304)

VU

// Run-time error

This time also program is compiled correctly but when we executed it only Shape
text is displayed for first object of array and then program terminates abnormally.
The reason behind this output is,
When we passed array of Line objects as parameter this array was converted
implicitly to array of Shapes objects by compiler as Line IS A kind of shape, same
thing was happened with pointers of array objects but the difference now is that this
Line objects array will be treated as Shape objects array, so draw method of first
shape object is called and it displays text Shape, for next iteration of loop when
compiler calculated address of next object it did it with respect to Shape objects but
the actual array consists of Line objects so compiler will incorrectly calculate address
of next object as size of Shape class object and Line class objects are different and
when it will call draw method with incorrect address of object program will
terminate abnormally this is give in diagram below,

0000
0010
0020
0030

0000
0015
0030
Shape Array

0045
Line Array

_shape[ i ].draw();
*(_shape + (i * sizeof(Shape))).draw();
As shown above compiler will do calculations with respect to Shape array and will calculate
next object address as 0010 because it is treating Line array as Shape array but actually next
Line object is present at address 0015 so runtime error will be generated as there is no new
object at address 0010 and program will terminate abnormally.

Original drawShapes() method that was using pointers is given below

void drawShapes(Shape* _shape[], int size) {
for (int i = 0; i < size; i++) {
_shape[i]->draw();
}
}
It shows correct output,

Virtual University of Pakistan

245

Object Oriented Programming (CS304)

VU

Sample Output:

Line
Line
Line

Because in case of pointers array, the size of each entry is same as each entry contains
a pointer and all pointers are of 4 bytes in c++ whatever they are point to Shape class
or to Line class or any other built in or user defined type. So in this case whether our
array contain Shape * or Line * next object address will be present after 4 bytes, so
our drawShapes function will execute correctly,

_line1

0000
0004
0008
0012

_line2

_line3

Shape* _shape[]

_shape[i]->draw();
(_shape + (i * sizeof(Shape*)))->draw();
So moral of the story is Never use arrays polymorphically because location of elemetns in any
array is calculated using array type and in polymorphism we have the liberty of changing
child array to parent array that will result in erraneous calcualation of location of elements in
array.

246

Virtual University of Pakistan

Object Oriented Programming (CS304)

VU

Lecture No.31
31.1. Multiple Inheritance
We have seen multiple inheritance in start of OOP now we see its implementation in
c++, A class in c++ can inherit from more than one classes like Phone class can
inherit from Transmitter or Receiver,
Examples:

Transmitter

Receiver

.......

.......

Transmit()

Receive()

Phone

class Phone: public Transmitter, public Receiver

// As phone class is publicly inherited so any class derived from Phone class will
also have access to public and protected members of Transmitter and Receiver
class
{
...
};
Derived class can inherit from public base class as well as private and protected base
classes
class Mermaid: private Woman, private Fish
// As Mermaid class is privately inherited from Woman and Fish so any class
derived from Mermaid class will Not have access to public and protected members
of Woman and Fish classes
{
...
};
Multiple Inheritance
The derived class inherits data members and functions from all the base classes
Object of derived class can perform all the tasks that an object of base class can
perform

Virtual University of Pakistan

247

Object Oriented Programming (CS304)

VU

Example
int main(){
Phone obj;
obj.Transmit(); // method of Transmitter class
obj.Receive(); // method of Receiver class
return 0;
}
Multiple Inheritance
When using public multiple inheritance, the object of derived classes can replace the
objects of all the base classes
Example
int main(){
Phone obj;
Transmitter * tPtr = &obj;
Receiver * rPtr = &obj;
return 0;
}
However note the following,
The pointer of one base class cannot be used to call the function of another base class
The functions are called based on static type
Example:
int main(){
Phone obj;
Transmitter * tPtr = &obj;
tPtr->Transmit();
tPtr->Receive();
//Error
return 0;
}

int main(){
Phone obj;
Receiver * rPtr = &obj;
rPtr->Receive();
rPtr->Transmit();
//Error
return 0;
}

31.2. Problems in Multiple Inheritance

If more than one base class have a function with same signature then the child
will have two copies of that function
Calling such function will result in ambiguity
Example:

248

Virtual University of Pakistan

Object Oriented Programming (CS304)

Land Vehicle

VU

Water Vehicle

Amphibious Vehicle

Car

Land Vehicle

Boat

Water Vehicle

Amphibious Vehicle

class LandVehicle{
public:
int GetMaxLoad();
};
class WaterVehicle{
public:
int GetMaxLoad();
};
class AmphibiousVehicle: public LandVehicle, public WaterVehicle {
};
int main(){
AmphibiousVehicle obj;
obj.GetMaxLoad();
// Error
return 0;
}

The solution of this problem is that Programmer must explicitly specify the class
name when calling ambiguous function
int main(){

Virtual University of Pakistan

249

Object Oriented Programming (CS304)

VU

AmphibiousVehicle obj;
obj.LandVehicle::GetMaxLoad();
obj.WaterVehicle::GetMaxLoad();
return 0;
}

Multiple Inheritance
The ambiguous call problem can arise when dealing with multiple level of multiple
inheritance
Example:

Vehicle

Land Vehicle

Amphibious Vehicle

Car

Boat

Vehicle

Vehicle

Land Vehicle

Water Vehicle

Car

250

Water Vehicle

Amphibious Vehicle

Virtual University of Pakistan

Boat

Object Oriented Programming (CS304)

VU

class Vehicle{
public:
int GetMaxLoad();
};
class LandVehicle : public Vehicle{
};
class WaterVehicle : public Vehicle{
};
class AmphibiousVehicle: public LandVehicle, public WaterVehicle {
};
int main(){
AmphibiousVehicle obj;
obj.GetMaxLoad();
return 0;
}

// Error

The error is due to the reason that object of AmphibiousVehicle class has two implicit
Vehicle class objects one with respect to LandVehicle and one with respect to
WaterVehicle class,
Vehicle
int GetMaxLoad();

Vehicle
int GetMaxLoad();

WaterVehicle

LandVehicle

AmphibiousVehicle

When call to GetMaxLoad method is made compiler get confused which copy of this method
should be called.

Virtual University of Pakistan

251

Object Oriented Programming (CS304)

VU

When we try to remove this error by mentioning that this method belongs to Vehicle class as
shown below error remained as it is due to above mentioned fact,
Example
int main()
{
AmphibiousVehicle obj;
obj.Vehicle::GetMaxLoad(); //Error
return 0;
}
The reason is same that Vehicle is accessible through two paths, we can avoid this error by
explicity mentioning the name of intermediate base class LandVehicle or WaterVehicle with
respect to which we want to call this GogMaxLoad method as shown below,
Example
int main(){
AmphibiousVehicle obj;
obj.LandVehicle::GetMaxLoad();
obj.WaterVehicle::GetMaxLoad();
return 0;
}
Due to the same reason as mentioned above Data member must be used with care
when dealing with more than one level on inheritance.
Example
class Vehicle{
protected:
int weight;
};
class LandVehicle : public Vehicle{
};
class WaterVehicle : public Vehicle{
};
class AmphibiousVehicle:
public LandVehicle,
public WaterVehicle{
public:
AmphibiousVehicle(){
LandVehicle::weight = 10;
WaterVehicle::weight = 10;
}
};

Here AmphibiousVehicle object has multiple copies of data member weight.

Memory View:

252

Virtual University of Pakistan

Object Oriented Programming (CS304)

VU

Data Members of Vehicle

Data Members of Vehicle

Data Members of LandVehicle

Data Members of WaterVehicle

Data Members of AmphibiousVehicle

31.3. Virtual Inheritance

The solution of avoid this problem is virtual inheritance so that in multiple
inheritance only one copy of base class is generated as shown below instead of two
separate copies.

Data Members of Vehicle

Data Members of LandVehicle

Data Members of WaterVehicle

Data Members of AmphibiousVehicle

In virtual inheritance there is exactly one copy of the anonymous base class object
Example:
class Vehicle{
protected:
int weight;
};
class LandVehicle :
public virtual Vehicle{
};
class WaterVehicle :
public virtual Vehicle{
};
Example
class AmphibiousVehicle:
public LandVehicle,
public WaterVehicle{
public:
AmphibiousVehicle(){
weight = 10;
}
};

Virtual University of Pakistan

253

Object Oriented Programming (CS304)

VU

Memory View
Virtual Inheritance
Virtual inheritance must be used when necessary
There are situation when programmer would want to use two distinct data members
inherited from base class rather than one

Example

Student
GPA

BS Student

MS Student

PhD Student

MS/PhD Student

254

Virtual University of Pakistan

Object Oriented Programming (CS304)

VU

Lecture No.32
Generic Programming

Motivation:
Following function prints an array of integer elements:
void printArray(int* array, int size)
{
for ( int i = 0; i < size; i++ )
cout << array[ i ] << , ;
}
If we want to print an array of character elements we will write similar function
again with different parameters,
void printArray(char* array, int size)
{
for ( int i = 0; i < size; i++ )
cout << array[ i ] << , ;
}
similarly for double array we can write,
void printArray(double* array, int size)
{
for ( int i = 0; i < size; i++ )
cout << array[ i ] << , ;
}
Same will be the case for array of float, short, long and so on.
You can see all these functions are doing same kind of functionality on different data types
and it looks repeated task for all data types (basically it is function overloading)
Now if we want to change the way function prints the array. e.g.
from 1, 2, 3, 4, 5
to 1 2 3 4 5 or 1 2 3 4 5 or 1 2 3 4 5
We will have to change the code in all functions for all data types.

Now consider the Array class that is developed to overcome the shortcomings of c++
built in arrays, in c++ arrays there is no check for array bounds we can insert element
at 100th position in an array of size 10 only, our program will compile correctly
however this will result in abnormal termination of the program on execution,
class Array {
int* pArray;
int size;
public:

Virtual University of Pakistan

255

Object Oriented Programming (CS304)

VU

};
This class has been written for integer arrays but we want to write the same kind of class for
all other data types also like double or boolean (true/false) as shown below,
class Array {
double* pArray;
int size;
public:

};
class Array {
bool* pArray;
int size;
public:

};
This also looks like repeated code and add effort to write code, secondly if we want to add a
function sum to Array class, we have to change all the three classes.
We want some mechanism that enables us to write single function or class that works for all
data types this technique of writing programs is called Generic Programming.
32.1. Generic Programming
Generic programming refers to programs containing generic abstractions (general
code that is same in logic for all data types like printArray function), then we
instantiate that generic program abstraction (function, class) for a particular data
type, such abstractions can work with many different types of data.
Advantages
As you can imagine major benefits of this approach are,
a. Reusability: Code can work for all data types
b. Writability: Code takes lesser time to write
c. Maintainability: Code is easy to maintain as changes are needed to be made
in a single function or class instead of many functions or classes.
32.2. Templates
In C++ generic programming is done using templates.
Templates are of two kinds,
a. Function Templates (in case we want to write general function like
printArray)
b. Class Templates (in case we want to write general class like Array class)
Compiler generates different type-specific copies from a single template
This concept is similar to concept of making prototype in the form of class for all
objects of same kind.
256

Virtual University of Pakistan

Object Oriented Programming (CS304)

VU

32.3. Function Templates

A function template can be parameterized to operate on different types of data types.
Declaration:
We write template keyword above any function make any function as template
function, they can be declared in any one of the following ways,
template< class T >
void funName( T x );
// OR
template< typename T >
void funName( T x );
// OR
template< class T, class U, >
void funName( T x, U y, );
Note here T is typename of class name and we use this T instead of data type/s for
which we want our function to work as template.
For Example Function Templates
Following function template prints an array having almost any type of elements
(note the use of T instead of int or float or char in implementation of function):
template< typename T >
void printArray( T* array, int size )
{
for ( int i = 0; i < size; i++ )
cout << array[ i ] << , ; // here data type of array is T
}
Similarly we can also write same function as,
template< class T >
void printArray( T* array, int size )
{
for ( int i = 0; i < size; i++ )
cout << array[ i ] << , ; // here data type of array is T
}
There is no difference in two implementations.
Template function will be instantiated for a particular data type according to passed
argument as shown below,
int main() {

Virtual University of Pakistan

257

Object Oriented Programming (CS304)

VU

int iArray[5] = { 1, 2, 3, 4, 5 };
void printArray( iArray, 5 ); // Instantiated for int[] as passed array is of
type int []
char cArray[3] = { a, b, c };
void printArray( cArray, 3 ); // Instantiated for char[] as argument is of type
char []
return 0;
}
Explicit Type Parameterization:
In case a function template does not have any parameter then we have to explicitly
mention the data type for which we want to create that function as shown below,
template <typename T>
T getInput() {
T x;
cin >> x;
return x;
}
Explicit Type Parameterization
int main() {
int x;
x = getInput();

// Error!

double y;
y = getInput();

// Error!

}
Explicit Type Parameterization
int main() {
int x;
x = getInput< int >();
double y;
y = getInput< double >();
}
User-defined Specializations
A template compiler generated code may not handle all the types successfully; in
that case we can give explicit specializations for a particular data type(s).
For example suppose we have written a function isEqual(, ) that compares two values of
data type and return true or false depending upon the values are equal or not,
template< typename T >

258

Virtual University of Pakistan

Object Oriented Programming (CS304)

VU

bool isEqual( T x, T y ) {
return ( x == y );
}

isEqual (6,6) should return true

isEqual (6,7) should return false
isEqual (6.6,6.6) should return true
isEqual (6.5,6.6) should return false
isEqual (A,A) should return true
isEqual (A,a) should return false
Until here the function will work correctly but consider the statement below,
isEqual (abc,xyz)
This is instantiation of isEqual function for built in type char [] or char *14, this function will
fail to give correct result simply because we have given its implementation as
return (x == y);
So here it will be translated by compiler in,
return ( char * == char *); or return ( char [] == char []);
As arrays consists of many elements, comparison of arrays in this way is not possible it will
simply compare first element of both arrays and will return result, this is given in code below,
#include <cstdlib>
#include <iostream>
using namespace std;
template< typename T >
bool isEqual( T x, T y ) {
return ( x == y );
}
int main (){
cout<<isEqual( 5, 6 ); // Ok
cout<<isEqual( 7.5, 7.5 );
// Ok
cout<<isEqual( "abc", "bca" ); // Logical Error!
system("PAUSE");
return EXIT_SUCCESS;
}

So all statements with starting same characters like below will return true,
isEqual( abc, acc );
isEqual( badac, bacc );
14

In c++ both char [] and char * are same data types

Virtual University of Pakistan

259

Object Oriented Programming (CS304)

VU

isEqual( cafaa, ccda );

This is logical error, solution of this problem is that we give our own correct implementation
of isEqual template function for char * data type, for this we can write specialization code
below general template function as follows,
Example User Specializations
#include <cstdlib>
#include <iostream>
using namespace std;
template< typename T >
bool isEqual( T x, T y ) {
return ( x == y );
}
template< >
bool isEqual< const char* >(
const char* x, const char* y ) {
return ( strcmp( x, y ) == 0 );
}

int main (){

cout<<isEqual( 5, 6 ); // OK
cout<<isEqual( 7.5, 7.5 );
// OK
cout<<isEqual( "abc", "aba" ); //OK will return False

system("PAUSE");
return EXIT_SUCCESS;
}

260

Virtual University of Pakistan

Object Oriented Programming (CS304)

VU

Lecture No.33
Recap
Templates are generic abstractions
C++ templates are of two kinds
o Function Templates
o Class Templates
A general template can be specialized to specifically handle a particular type
like we did for char []
33.1. Multiple Type Arguments
Suppose we want to write code to convert different types into one another (like char to int or
int to char or float to int or int to float), the problem is same we have to write many functions
corresponding to each type, as no of types will increase the required no. of functions will also
increase, the concept of templates can be used here as well to write general function to
convert one type into another, in this case we will need two type arguments as shown below,
template< typename T, typename U >
T my_cast( U u ) {
return (T)u; // U type will be converted to T type and will be returned
}
int main() {
double d = 10.5674;
int j = my_cast( d ); //Error
int i = my_cast< int >( d );
// need to explicity mention about type of T (int in this case) as it is used only
for
// return type not as parameter
return 0;
}

33.2. User-Defined Types

Besides primitive types, user-defined types can also be passed as type arguments to
templates, compiler performs static type checking to diagnose type errors.
Consider the String class without overloaded operator ==

class String {
char* pStr;

// Operator == not defined

};
template< typename T >
bool isEqual( T x, T y ) {
return ( x == y );
}

Virtual University of Pakistan

261

Object Oriented Programming (CS304)

VU

int main() {
String s1 = xyz, s2 = xyz;
isEqual( s1, s2 );
// Error!
return 0;
}

We can use String class objects as arguments to template function isEqual also but in this
case we should have defined overloaded == operator for our string class as friend function
because this operator is being used in isEqual function,

class String {
char* pStr;

friend bool operator ==(

};

const String&, const String& );

bool operator ==( const String& x, const String& y ) {

return strcmp(x.pStr, y.pStr) == 0;
}
template< typename T >
bool isEqual( T x, T y ) {
return ( x == y );
}
int main() {
String s1 = xyz, s2 = xyz;
isEqual( s1, s2 );
// OK
return 0;
}

33.3. Overloading vs. Templates

Function templates are used when we want to have exactly identical operations on
different data types in case of function templates we can not change implementation
from data type to data type however we can specialize implementation for a
particular data type.
In case we want to have similar operations on different data types we need function
overloading in function overloading case we can give similar but slightly different
implementation for each data type.
Example: Overloading vs. Templates

+ operation is overloaded for different operand types (different

implementation in each case)
A single function template can calculate sum of array of many types

Function Overloading

262

Virtual University of Pakistan

Object Oriented Programming (CS304)

VU

String operator +( const String& x, const String& y ) {

String tmp;
tmp.pStr = new char[strlen(x.pStr) + strlen(y.pStr) + 1 ];
strcpy( tmp.pStr, x.pStr );
strcat( tmp.pStr, y.pStr );
return tmp;
}
String operator +( const char * str1, const String& y ) {
String tmp;
tmp.pStr = new char[ strlen(str1) + strlen(y.pStr) + 1 ];
strcpy( tmp.pStr, str1 );
strcat( tmp.pStr, y.pStr );
return tmp;
}

Templates
template< class T >
T sum( T* array, int size ) {
T sum = 0;
for (int i = 0; i < size; i++)
sum = sum + array[i];
return sum;
}

33.4. Template Arguments as Policy:

We can change behaviour of a template using template parameter. We can pass a template
argument to enforce some rule (policy). For example see the problem statement below,
Write a function that compares two given character strings
This function is similar to built in string comparison function strcmp with the
difference that it can perform both case sensitive and case insensitive comparisons.
33.5. First Solution:
We can write two separate functions for both cases and call any of them as required,

int caseSencompare( char* str1, char* str2 )

{
for (int i = 0; i < strlen( str1 ) && i < strlen( str2 ); ++i)
if ( str1[i] != str2[i] )
return str1[i] - str2[i];

Virtual University of Pakistan

263

Object Oriented Programming (CS304)

VU

return strlen(str1) - strlen(str2);

}

This function will return 0 in case string length of both strings is same and they have identical
elements otherwise it will return 1.

int nonCaseSencompare( char* str1, char* str2 )

{
for (int i = 0; i < strlen( str1 ) && i < strlen( str2 ); i++)
if ( toupper( str1[i] ) != toupper( str2[i] ) )
return str1[i] - str2[i];
return strlen(str1) - strlen(str2);
}

This function will return 0 in case string length of both strings is same and they have same
alphabets (ignoring they are in lower case or in upper case) otherwise it will return 1.
33.6. Second Solution:
We write a single compare function and pass a bool type parameter to indicate type of
comparison and from this function we return result based on passed bool parameter used to
indicate case sensitive or case insensitive comparison,
int compare( char* str1, char* str2, bool caseSen )
{
for (int i = 0; i < strlen( str1 ) && i < strlen( str2 ); i++)
if ( )
return str1[i] - str2[i];
return strlen(str1) - strlen(str2);
}
// if condition:
(caseSen && str1[i] != str2[i]) || (!caseSen && toupper(str1[i]) != toupper(str2[i]))
Here bool variable caseSen will work as flag, and will activate either left sub-expression or
right sub-expression in if-statement, in case we are performaing case sensitive comparison
caseSen will be true (1) and will activate sub-expression,
( caseSen && str1[i] != str2[i] )
and in case caseSen is false it will activate second sub-expression involving !caseSen
as given below,
( !caseSen && toupper(str1[i]) != toupper(str2[i] )
other logic is same as in case 1 functions.
33.7. Third Solution

264

Virtual University of Pakistan

Object Oriented Programming (CS304)

VU

Third solution is most elegant solution out of all these, in which we write two classes one for
case sensitive and other for case insensitive comparison, and pass one of these classes as
argument when instantiating template compare function, in compare function isEqual
function of passed class is being called to perform either case sensitive or case insensitive
comparison,
class CaseSenCmp {
public:
static int isEqual( char x, char y ) {
return x == y;
}
};
class NonCaseSenCmp {
public:
static int isEqual( char x, char y ) {
return toupper(x) == toupper(y);
}
};
template< typename C >
int compare( char* str1, char* str2 )
{
for (int i = 0; i < strlen( str1 ) && i < strlen( str2 ); i++)
if ( !C::isEqual (str1[i], str2[i]) )
return str1[i] - str2[i];
return strlen(str1) - strlen(str2);
};
int main() {
int i, j;
char *x = "hello", *y = "HELLO";
i = compare< CaseSenCmp >(x, y);
j = compare< NonCaseSenCmp >(x, y);
cout << "Case Sensitive: " << i;
cout << "\nNon-Case Sensitive:
<< j << endl;
return 0;
}
Sample Output
Case Sensitive: 32
// Not Equal
Non-case Sensitive: 0 // Equal

33.8. Default Policy

We can set default class type as default comparison type as we set default parameters in case
of constructors,
template< typename C = CaseSenCmp >

Virtual University of Pakistan

265

Object Oriented Programming (CS304)

int compare( char* str1, char* str2 )

{
for (int i = 0; i < strlen( str1 ) && i < strlen( str2 ); i++)
if ( !C::isEqual
(str1[i], str2[i]) )
return str1[i] - str2[i];
return strlen(str1) - strlen(str2);
};
int main() {
int i, j;
char *x = "hello", *y = "HELLO";
i = compare(x, y);
j = compare< NonCaseSenCmp >(x, y);
cout << "Case Sensitive: " << i;
cout << "\nNon-Case Sensitive:
<< j << endl;
return 0;
}

266

Virtual University of Pakistan

VU

Object Oriented Programming (CS304)

VU

Lecture No.34
Generic Algorithms - A Case Study
Consider the template printArray function we wrote in previous lecture,
Printing an Array
template< typename T >
void printArray( T* array, int size )
{
for ( int i = 0; i < size; i++ )
cout << array[ i ] << , ;
}

Although this function will work for all data types but still it is taking array as an
argument so this function still depends upon nature of data structure used to pass
data, we want this function to be independent from data structure as well mean it
should print single values as well as arrays of basic data types.
34.1. Generic Algorithms
We want to provide such implementation that is independent of data structures also
for example in case of printing of values we want printing of both single values and
arrays, this can be achieved using Generic Programming in which a function work for
all types of containers, let us see how we can make a function generic step by step
with the help of an example, consider the find function below it is similar to
printArray function with the difference that it tries to find an element in an array and
return pointer to that element if it is found in the array otherwise it returns zero
(NULL).

Find function that tries to find an integer value in an integer array

const int* find( const int* array, int _size, int x ) {
const int* p = array;
for (int i = 0; i < _size; i++) {
if ( *p == x )
return p;
p++;
}
return 0;
}

First step obviously will be to write it as template so that it may work for all data
types instead of only integers,
We write it as template function,

Template function to find a value within an array

template< typename T >
T* find( T* array,int _size, const T& x ) {
T* p = array;

Virtual University of Pakistan

267

Object Oriented Programming (CS304)

VU

for (int i = 0; i < _size; i++) {

if ( *p == x )
return p;
p++;
}
return 0;
}

In next step we simply pass a pointer that is pointing one location next after array
instead of passing size of array as shown below, (this will make our code simplified)

template< typename T >

T* find( T* array, T* beyond, const T& x ) {
T* p = array;
while ( p != beyond ) {
if ( *p == x )
return p;
p++;
}
return 0;
}

We also change the return statement of the function by returning the beyond pointer
instead of zero in case element is not found, so now we will not check the return
value for NULL pointer we will simply check whether it points to any value or it
points to beyond pointer, (we are doing it so that we can use a single return
statement to return a single pointer as we have done below)

template< typename T >

T* find( T* array, T* beyond, const T& x ) {
T* p = array;
while ( p != beyond ) {
if ( *p == x )
return p;
p++;
}
return beyond;
}

using single return statement,

template< typename T >

T* find( T* array, T* beyond, const T& x ) {
T* p = array;
while ( p != beyond && *p != x )
p++;
268

Virtual University of Pakistan

Object Oriented Programming (CS304)

VU

return p;
}

Now our code is so generic that it will work for all types of containers (data
structures) supporting two operations,

Increment operator (++) as we are incrementing value in this container

Dereference operator (*) as we are getting value from container for
comparison by dereferencing

Although this template function will now work for all containers in same way but it
has one limitation that we will need to pass container pointers in this function this is
against the concept of generic programming so we simply remove pointer notation
from our code,

template< typename P, typename T >

P find( P start, P beyond, const T& x ) {
while ( start != beyond && *start != x )
start++;
return start;
}

Now this implementation will work for all references of containers as shown in code
below,

int main() {
int iArray[5];
iArray[0] = 15;
iArray[1] = 7;
iArray[2] = 987;

int* found;
found = find(iArray, iArray + 5, 7);
return 0;
}

We can apply the same concept of Generic Algorithms to class templates and
develop a generic Vector class that will work for all built-in types as shown below,
34.2. Class Templates
As we saw before a single class template provides functionality to operate on
different types of data and in this way facilitates reuse of classes (like we wrote our
own array class to overcome the limitations of built-in array class in c++).
We can definition a class template as follows:

Virtual University of Pakistan

269

Object Oriented Programming (CS304)

VU

template< class T > class Xyz { };

template< typename T > class Xyz { };

Now we write a template Vector class using the concept of Generic Algorithms such
that it can store everything, basically it will act as container to store anything in it, it
can store objects that are them self collections like arrays or it can store basic data
types. This will be possible because we are going to implement this Vector class
using the concept of Generic Algorithms.
34.3. Example Class Template
A Vector class template can store data elements of different types, without templates,
we need a separate Vector class for each data type.
We start with the basic definition of Vector class using templates,

Class Template
template< class T >
class Vector {
private:
int size;
T* ptr;
public:
Vector<T>( int = 10 );
Vector<T>( const Vector< T >& );
~Vector<T>();
int getSize() const;
const Vector< T >& operator =(
const Vector< T >& );
T& operator []( int );
};

Its implementation is,

template< class T >

Vector<T>::Vector<T>( int s ) {
size = s;
if ( size != 0 )
ptr = new T[size];
else
ptr = 0;
}
template< class T >
Vector<T>:: Vector<T>(
const Vector<T>& copy ) {
size = copy.getSize();
if (size != 0) {
ptr = new T[size];

270

Virtual University of Pakistan

Object Oriented Programming (CS304)

VU

for (int i = 0; i < size; i++)

ptr[i] = copy.ptr[i];
}
else ptr = 0;
}
template< class T >
Vector<T>::~Vector<T>() {
delete [] ptr;
}
template< class T >
int Vector<T>::getSize() const {
return size;
}
template< class T >
const Vector<T>& Vector<T>::operator
=( const Vector<T>& right) {
if ( this != &right ) {
delete [] ptr;
size = right.size;
if ( size != 0 ) {
ptr = new T[size];
for(int i = 0; i < size;i++)
ptr[i] = right.ptr[i];
}
else
ptr = 0;
}
return *this;
}
template< class T >
T& Vector< T >::operator []( int index ) {
if ( index < 0 || index >= size ) {
cout << Error: index out of
range\n;
exit( 1 );
}
return ptr[index];
}

We can create this vector class instances for int or char data type as given below,
Vector< int > intVector;
Vector< char > charVector;

Virtual University of Pakistan

271

Object Oriented Programming (CS304)

VU

This Vector class is parameterized class and will always be instantiated for a
particular type only. Now we can not create object of type Vector only it will be
instantiated for a particular data type like Vector<int> or Vector <float> and so on

272

Virtual University of Pakistan

Object Oriented Programming (CS304)

VU

Lecture No.35
35.1. Member Templates:
Member functions of a template class implicitly become functions templates; they work for
instantiations (int, char, float, double so on) of that class, however there are some situations
where we need explicit template functions in our class taking more template parameters other
than one implicit template parameter (parameter given to this class as parameter while
creating its object).
A class or class template can have member functions that are themselves templates

template<typename T> class Complex {

T real, imag;
public:
// Complex<T>( T r, T im )
Complex( T r, T im ) :
real(r), imag(im) {}
// Complex<T>(const Complex<T>& c)
Complex(const Complex<T>& c) :
real( c.real ), imag( c.imag ) {}

};
Note that in c++ while declaring class there is no need to mention template
parameter for class member functions as compiler implicitly understand it, however
if we are using some other template parameter like we are doing for copy constructor
then we need to give its name as well.
Now see that main function for this class in which we are assigning Complex class
float instance to double, it will result in an error,

int main() {
Complex< float > fc( 0, 0 );
Complex< double > dc = fc; // Error
return 0;
}
Because our Complex copy constructor is taking argument of same template type T
so this copy constructor will work for statements like,
Complex< float > f1c( 0, 0 );
Complex< float > f2c = f1c;
Complex< double > d1c( 0, 0 );
Complex< double > d2c = d1c;
But will not work for statements like,
Complex< float > fc( 0, 0 );
Complex< double > dc = fc;

Virtual University of Pakistan

273

Object Oriented Programming (CS304)

VU

Complex< double > d1c( 0, 0 );

Complex< float > d2c = d1c;

When we created Complex object for double compiler generated class for double like
shown below,

class Complex<double> {
double real, imag;
public:
Complex( double r, double im ) :
real(r), imag(im) {}
Complex(const Complex<double>& c) :
real( c.real ), imag( c.imag ) {}

};
So there is need to some sort of overloading of copy constructor such that we can
assign two instances of Complex class for two different data types.
Now we change copy constructor as function template explicitly, so that it may work
for copy of different types of data types as well,
template<typename T> class Complex {
T real, imag;
public:
Complex( T r, T im ) :
real(r), imag(im) {}
template <typename U>
// this copy constructor is now taking two template parameters one implicit T
// and other explicit U
Complex(const Complex<U>& c) :
real( c.real ), imag( c.imag ) {}

};
Now assignment of float Complex instance to double instance will fine,
int main() {
Complex< float > fc( 0, 0 );
Complex< double > dc = fc; // OK
return 0;
}

Because here copy constructor

Complex(const Complex<U>& c) :
Will be instantiated for implicit template type (T) double and explicit template type
(U) float because now our Copy constructor is acting as explicity template function
written for two different parameters.

274

Virtual University of Pakistan

Object Oriented Programming (CS304)

VU

Important points to Note:

1. Only that instantiation of copy constructor is generated that is required.
2. Good compilers only generate required template function instances for Complex class
instantiation for a particular data type as shown below,
For statement: Complex< double > dc = fc;
the functions that will be generated for float and double instance of complex class are shown
below,
<double> Instantiation
class Complex<double> {
double real, imag;
public:
Complex( double r, double im ) :
real(r), imag(im) {}
template <typename U>
Complex(const Complex<U>& c) :
real( c.real ), imag( c.imag ) {}

};
<float> Instantiation
class Complex<float> {
float real, imag;
public:
Complex( float r, float im ) :
real(r), imag(im) {}
// No Copy Constructor code is generated as there is no need for it

};
This approach avoid code bloat (unnecessary code generation).

35.2. Class Template Specialization

Like function templates a class template may also not handle all the types
successfully, for example for char arrays (char *) the behaviour of template class
vector may be not be as desired as shown below,
Vector class to store integers
int main() {
Vector< int > iv1( 2 );
iv1[0] = 15;
iv1[1] = 27;
Vector< int > iv2( iv1 );
Vector< int > iv3( 2 );
iv3 = iv1;

Virtual University of Pakistan

275

Object Oriented Programming (CS304)

VU

return 0;
}

Vector class to store char arrays (char *)

int main() {
Vector< char* > sv1( 2 );
sv1[0] = "Aamir";
// compiler will generate a const C String having value Aamir and will assign its
//pointer to sv1[0]
sv1[1] = "Nasir";
Vector< char* > sv2( sv1 ); // issue of shallow copy
Vector< char* > sv3( 2 );
sv3 = sv1; // issue of shallow copy
return 0;
}

We can write explicit specialization for Vector class for char arrays as we wrote for
template function isEqual,
template<>
class Vector< char* > {
private:
int size;
char** ptr;
public:
// Vector< char* >( int = 10 );
Vector( int = 10 );
Vector( const Vector< char* >& );
virtual ~Vector();
int getSize() const;
const Vector< char* >& operator =( const Vector< char* >& );
const char*& operator []( int );
void insert( char*, int );
};
template<>
Vector<char*>::Vector(int s) {
size = s;
if ( size != 0 ) {
ptr = new char*[size];
for (int i = 0; i < size; i++) ptr[i] = 0;
}
else
ptr = 0;
}
template<>
Vector< char* >::Vector( const Vector<char*>& copy ) {
size = copy.getSize();
if ( size == 0 ) {
ptr = 0;
return;

276

Virtual University of Pakistan

Object Oriented Programming (CS304)

VU

}
ptr = new char*[size];
for (int i = 0; i < size; i++)
if ( copy.ptr[i] != 0 ) {
ptr[i] = new char[ strlen( copy.ptr[i] ) + 1 ];
strcpy(ptr[i], copy.ptr[i]);
}
else
ptr[i] = 0;
}
template<>
Vector<char*>::~Vector() {
for (int i = 0; i < size; i++)
delete [] ptr[i];
delete [] ptr;
}
template<>
int Vector<char*>::getSize() const {
return size;
}
template<>
const Vector<char*>& Vector<char*>::
operator=(const Vector<char*>& right)
{
if ( this == &right )
return *this;
for (int i = 0; i < size; i++)
delete [] ptr[i];
delete [] ptr;
size = right.size;
if ( size == 0 ) {
ptr = 0;
return *this;
}
ptr = new char*[size];
for (int i = 0; i < size; i++)
if ( right.ptr[i] != 0 ) {
ptr[i] = new char[strlen( right.ptr[i] ) + 1];
strcpy( ptr[i], right.ptr[i] );
}
else
ptr[i] = 0;
}
template<>
const char*& Vector<char*>:: operator []( int index ) {
if ( index < 0 || index >= size ) {
cout << "Error: index out of range\n";

Virtual University of Pakistan

277

Object Oriented Programming (CS304)

VU

exit( 1 );
}
return ptr[index];
}
template<>
void Vector< char* >::insert( char* str, int i ) {
delete [] ptr[i];
if ( str != 0 ) {
ptr[i] = new char[strlen(str)+ 1];
strcpy( ptr[i], str );
}
else
ptr[i] = 0;
}

This code is similar to general template Vector class with the difference that now we are
allocating and taking care of dynamic memory associated with char arrays.
The main program showing usage of above main program is given below,
int main() {
Vector< char* > sv1( 2 );
sv1[0] = Aamir;
// Error as now we have to changed code of overloaded subscript operator to
return //const Vector pointer now and we can not assign new value to constant
reference //of Vector now we have to use insert function explicity written for that
purpose
sv1.insert( "Aamir", 0 );
sv1.insert( "Nasir", 1 );
Vector< char* > sv2( sv1 );
Vector< char* > sv3( 2 );
sv3 = sv1;
return 0;
}

278

Virtual University of Pakistan

Object Oriented Programming (CS304)

VU

Lecture No.36
Recap
We saw in previous lecture that how we can implement our programming problems
easily using Generic Algorithms. Then we saw how we can add explicit template
functions to our Class templates to add our desired functionality.
A class template may not handle all the types successfully explicit specializations are
required to deal such types we can implement concept of template specialization for
such classes as well as we did for function templates.
Now we move forward to see other cases for member templates,
36.1. Member Templates Revisited
We can add member templates for ordinary classes as well, for example following
code is adding any instance of Complex class to ComplexSet class, (as we know
complex class can be instantiated for int, float or double and ComplexSet class will be
collection of Complex class instantiations ) as shown below,
Complex Class objects,

real

real

5.69

img

img

8.25

Complex<int>

real

5.3286

img

5.3284

Complex<float>

Complex<double>

Virtual University of Pakistan

279

Object Oriented Programming (CS304)

VU

real

real

5.69

real

5.3286

img

img

8.25

img

5.3284

Complex<int>

Complex<float> Complex<double>

ComplexSet Object

class ComplexSet {
template< class T >
insert( Complex< T > c ) // any instance of complex class
{
// Add instance Complex class to Complex set class having many // Complex class
instances
}
};
int main() {
Complex< int > ic( 10, 5 );
Complex< float > fc( 10.5, 5.7 );
Complex< double > dc( 9.567898, 5 );
ComplexSet cs;
cs.insert( ic );
cs.insert( fc );
cs.insert( dc );
return 0;
}
36.2. Partial Specialization:
We can also perform partial specialization instead of complete specialization, a
partial specialization of a template exists between general specialization and
complete specialization for example we can specialize a class to behave in a certain
manner in case of pointers or in case of parameter of a certain type, see the example
below to understand the difference between complete and partial specialization,
template< class T, class U, class V > // general template
template< class T, class U, int > // partial specialization
template< class T, float , int > // partial specialization
template< int , float , int > // complete specialization
In partial specialization, the number of template parameters remains the same
however their nature varies (they become more specific).
Example Partial Specialization

280

Virtual University of Pakistan

Object Oriented Programming (CS304)

VU

template< class T >

class Vector { };
template< class T >
class Vector< T* > { }; // Here T can take any type pointer

Example Partial Specialization

template< class T, class U, class V >
class A {};
template< class T, class V >
class A< T, T*, V > {}; // the parameters in header of are two but in class we are
using same three parameters
template< class T, class U, int I >
class A< T, U, I > {}; // here we have changed third parameter to non type
parameter15
template< class T >
class A< int, T*, 5 > {}; // here we have changed first parameter to non type
parameter and second is T * any data type pointer and third one is constant
expression 5

Example Complete Specialization

template< class T >
class Vector { };
template< >
class Vector< char* > { };

Example Complete Specialization

template< class T, class U, class V >
class A {};
template< >
class A< int, char*, double > {};

36.3. Function Templates

Similar to class templates a function template may also have partial specializations,
Example Partial Specialization

template< class T, class U, class V >

15

non type parameters are those parameters which are not template parameters

Virtual University of Pakistan

281

Object Oriented Programming (CS304)

VU

void func( T, U, V );
template< class T, class V >
void func( T, T*, V );
template< class T, class U, int I >
void func( T, U );
template< class T >
void func( int, T, 7 );

36.4. Complete Specialization

We have already used this complete specialization in case of function templates,
template< >
bool isEqual< const char* >(
const char* x, const char* y ) {
return ( strcmp( x, y ) == 0 );
}

Example
Consider the following template
template< typename T >
bool isEqual( T x, T y ) {
return ( x == y );
}

Following partial specialization of this function deals with pointers to objects,

template< typename T >
bool isEqual( T* x, T* y ) {
return ( *x == *y );
}
So complete having all types of specializations for isEqual function is shown below,
template< typename T >
bool isEqual( T x, T y ) {
return ( x == y );
}
template< typename T >
bool isEqual( T* x, T* y ) {
return ( *x == *y );
}
template< >
bool isEqual< const char* >(

282

Virtual University of Pakistan

Object Oriented Programming (CS304)

VU

const char* x, const char* y ) {

return ( strcmp( x, y ) == 0 );
}

36.5. Using Different Specializations

The code below shows how we can use different type of specializations for isEqual function
given above,

int main() {
int i, j;
char* a, b;
Shape *s1 = new Line();
Shape *s2 = new Circle();
isEqual( i, j ); // Template
isEqual( a, b ); // Complete Sp.
isEqual( s1, s2 ); // Partial Sp.
return 0;
}
36.6. Non-type Parameters
The parameters given in template definition other than those used for mentioning templates
types are called non type parameters like,
template <class T, class U, int I>
Here int I is not type parameter.
Template parameters may include non-type parameters, the non-type parameters
may have default values like,
template <class T, class U, int I = 5>
They are treated as constants and are commonly used is static memory allocation
mean when we want to pass the length of memory we need in template at compile
time (statically) the example below shows all this,
36.7. Example template class Array
Consider the Array class again that we discussed in template classes introduction, in this
template array class we are creating a c++ array of any built in data type by passing array size
in Array object constructor as shown below,

template< class T >

class Array {
private:
T* ptr;
public:
Array( int size );
~Array();

Virtual University of Pakistan

283

Object Oriented Programming (CS304)

VU

};
template< class T >
Array<T>::Array() {
if (size > 0)
ptr = new T[size];
else
ptr = NULL;
}
int main() {
Array< char > cArray( 10 );
Array< int > iArray( 15 );
Array< double > dArray( 20 );
return 0;
}

We can do the same by passing the array size as non type parameter while creating
an Array class object itself, for this we will need to change to template class
definition by adding a non type parameter SIZE as shown below,
template< class T, int SIZE >
class Array {
private:
T ptr[SIZE];
public:
Array();

};

Now we can simply pass array size as argument while creating Array class object,
int main() {
Array< char, 10 > cArray;
Array< int, 15 > iArray;
Array< double, 20 > dArray;
return 0;
}

36.8. Default Non-type Parameters

We can set default value for this non type parameters, as we do for parameters
passed in ordinary functions,

template< class T, int SIZE = 10 >

class Array {
private:
T ptr[SIZE];
public:

284

Virtual University of Pakistan

Object Oriented Programming (CS304)

VU

void doSomething();

Now default value will be used if we will not pass any value for array size,
int main() {
Array< char > cArray; // here Array of size 10 will be created
return 0;
}

36.9. Default Type Parameters

We can also specify default type for type parameters (template parameters like T, U,
V), consider the Vector class again we can mention default type of Vector class (that
type will be used in case we have not mentioned any type while creating Vector class
object)
template< class T = int > // default type for Vector class is now int
class Vector {

}
Vector< > v; // same as Vector< int > v

Lecture No.37
37.1. Resolution Order
Consider the following specializations of template Vector class,
template< typename T >
class Vector { };
template< typename T >
class Vector< T* > { };
template< >
class Vector< char* > { };

As these all are template implementations of class Vector, when compiler have to use this
template class having many specializations from partial to complete, compiler searches all
these specializations in a particular order called resolution order, so resolution order it is
the sequence in which compiler searches for required template specialization and is given
below,
a. First of all compiler looks for complete specialization

Virtual University of Pakistan

285

Object Oriented Programming (CS304)

VU

b. If it can not find any required complete specialization then it searches for some partial
specialization
c. In the end it searches for some general template
So in other words we can say that compiler searches template specializations from
more specific to more general.
Example Resolution Order
The code below shows which particular instantiation of template Vector class will be used,
int main() {
Vector< char* > strVector;
// Vector< char* > instantiated (complete specialization used)
Vector< int* > iPtrVector;
// Vector< T* > instantiated (partial specialization used)
Vector< int > intVector;
// Vector< T > instantiated (general specialization used)
return 0;
}

Explanation of code:
In all three cases above, compiler will search for required template specialization in the order
given below,
1. Complete specialization (Vector< char* >)
2. Partial specialization (Vector< T* >)
3. General specialization (Vector< T>)
When compiler will find line,
Vector< char* > strVector;
It will find corresponding complete specialization and will stop there.
When compiler will find line,
Vector< int* > iPtrVector;
It will start its searching from complete specializations available but there is only one
complete specialization for char * (Vector< char* >) and there is no complete
specialization for int *, so compiler will move to partial specialization and will try to
match some with int *, compiler will find one (Vector <T *>) that is working for all
pointer data types and therefore it will use that.
When compiler will find line,
Vector< int > intVector;
It will start from complete specializations available but there is only one complete
specialization for char * (Vector< char* >) and there is no complete specialization for
int, so compiler will move to partial specialization and will try to match some with
int, compiler will find one partial speciation (Vector <T *>) that is working for all
pointer data types only hence not applicable here, then compiler will move to general
template class Vector <T> and will use it.

286

Virtual University of Pakistan

Object Oriented Programming (CS304)

VU

37.2. Function Template Overloading

We can specialize function templates also, this is called function template overloading,
Consider the specializations of template function sort,
template< typename T >
void sort( T );
// general template function
template< typename T >
void sort( Vector< T > & ); // specialization for template Vector class instantiated for any
data type

template< >
void sort< Vector<char*> >(Vector< char* > & );

// specialization for template

Vector instantiated for char * data type

void sort( char* ); // specialization for char * data type.

37.3. Resolution Order

Compiler searches target of a function call in the following order in case of function
of templates,

Ordinary Function
Complete Specialization
Partial Specialization
Generic Template

This is the same order as was in case of class template with the addition of ordinary functions
(non template functions with the same name),
Example Resolution Order
int main() {
char* str = Hello World!;
sort(str); // Ordinary function sort( char* )
Vector<char*> v1 = {ab, cd, };
sort(v1); //Complete specialization sort( Vector<char*> & )
Vector<int> v2 = { 5, 10, 15, 20 };
sort(v2); // partial specialization sort( Vector<T> &)
int iArray[] = { 5, 2, 6 , 70 };
sort(iArray); // general template sort( T )
return 0;
}

37.4. Templates and Inheritance

We can use inheritance comfortably with templates or their specializations, but we
must follow one rule:

Virtual University of Pakistan

287

Object Oriented Programming (CS304)

VU

If we have a template class then all classes derived from it should also be
class templates
Derived class must take at least as many template parameters as the base
class requires for an instantiation

37.5. Derivations in case of a General Template class

A class template may inherit from another class template, consider the
template class A and we are deriving another class B from it which is also
template class,
template< class T >
class A
{ };
template< class T >
class B : public A< T >

// same template parameter T in both classes declarations ensures that both classes will be instantiated for
same types

{ };
int main() {
A< int > obj1;
B< int > obj2;
return 0;
}

A partial specialization may inherit from a class template

template< class T >
class B< T* > : public A< T >// same template parameter T is used here also to

//ensure that both classes will be instantiated for same data types

{ };
int main() {
A< int > obj1;
B< int* > obj2;
return 0;
}

Complete specialization or ordinary class cannot inherit from a class template

because as you can see complete specialization is in fact instance of the class
for one data type and in this way it is similar to ordinary class as ordinary
class also can be instantiated for its members of a particular data type,

template< >
class B< char* > : public A< T >
{ };
// Error: T undefined, derived class is taking less parameters than base class

class B : public A< T >

{ };
// Error: T undefined, derived class is taking less parameters than base class

288

Virtual University of Pakistan

Object Oriented Programming (CS304)

VU

Derivations in case of a partially specialized class

A class template may inherit from a partial specialization

template< class T >

class A
{ };
template< class T >
class A< T* >
{ };
template< class T >
class B : public A< T* >
{}
int main() {
A< int* > obj1;
B< int > obj2;
return 0;
}

A partial specialization may inherit from a partial specialization

template< class T >

class B< T* > : public A< T* >
{ };
int main() {
A< int* > obj1;
B< int* > obj2;
return 0;
}

Complete specialization or ordinary class cannot inherit from a partial

specialization

template< >
class B< int* > : public A< T* >
{ } // Error: Undefined T
class B : public A< T* >
{ } // Error: Undefined T

Derivations in case of Completely Specialized class

template< class T >

class B : public A< float* >

Virtual University of Pakistan

289

Object Oriented Programming (CS304)

{ };
int main() {
A< float* > obj1;
B< int > obj2;
return 0;
}

A partial specialization may inherit from a complete specialization

template< class T >

class B< T* > : public A< float* >
{ };
int main() {
A< float* > obj1;
B< int* > obj2;
return 0;
}

A complete specialization may inherit from a complete specialization

template< >
class B< double* > : public A< float* >
{ };
int main() {
A< float* > obj1;
B< double* > obj2;
return 0;
}

An ordinary class may inherit from a complete specialization

class B : public A< float* >

{ };
int main() {
A< float* > obj1;
B obj2;
return 0;
}
Derivations in case of Ordinary Class

A class template may inherit from an ordinary class

class A
{ };

290

Virtual University of Pakistan

VU

Object Oriented Programming (CS304)

VU

template< class T >

class B : public A
{ };
int main() {
A obj1;
B< int > obj2;
return 0;
}

A partial specialization may inherit from an ordinary class

template< class T >

class B< T* > : public A
{ };
int main() {
A obj1;
B< int* > obj2;
return 0;
}

A complete specialization may inherit from an ordinary class

template< >
class B< char* > : public A
{ };
int main() {
A obj1;
B< char* > obj2;
return 0;
}

Virtual University of Pakistan

291

Object Oriented Programming (CS304)

VU

Lecture No.38
38.1. Templates and Friends
Templates or their specializations are also compatible with friendship feature of C++
as they are with inheritance.
We consider the following rules when dealing with templates and friend functions
and classes.
38.2. Templates and Friends Rule 1
When an ordinary function or class is declared as friend of a class template then it
becomes friend of each instantiation of that template class.
Consider the code below in which we are declaring a class A as friend of a class B
and in this class A, we are accessing private data of class B without any error,
In case of friend classes,

class A {

};
template< class T >
class B {
int data;
friend A; // declaring A as friend of B

};
class A {
void method() {
B< int > ib;
B< char > cb
ib.data = 5; // OK: Accessing private member data for class B
instantiation ib
cb.data = 6; // OK: Accessing private member data for class B
instantiation cb
}
};

In case of friend functions,

void doSomething( B< char >& );
template< class T >
class B {
int data;
friend void doSomething( B<char>& );
// declaring function doSomething as friend of B

};

292

Virtual University of Pakistan

Object Oriented Programming (CS304)

VU

void doSomething( B< char >& cb ) {

B< int > ib;
ib.data = 5;
// OK: Accessing private data of class B
cb.data = 6; // OK: Accessing private data of class B
}
In combined way both can be written as follows,
void doSomething( B< char >& );
class A { };
template< class T > class B {
int data;
friend void doSomething( B<char>& );
friend A;

};
void doSomething( B< char >& cb ) {
B< int > ib;
ib.data = 5;
// OK
cb.data = 6; // OK
}
class A {
void method() {
B< int > ib;
B< char > cb
ib.data = 5;
// OK
cb.data = 6;
// OK
}
};

The Rule 1 is simple now we go to Rule 2.

38.3. Templates and Friends Rule 2
Rule 2 applies when we make template function or template class friend of another
template class.
According to Rule 2 when a friend function template or friend class template is
instantiated with the type parameters of class template granting friendship then its
instantiation for a specific type is a friend of that class template instantiation for that
particular type only. The example below explains this concept,
Consider the code below it has a template friend function and a template friend
class of template class B, note that in template class B we have given the same
parameter T to both template friend function doSomething and template friend class
A, according to Rule 2 when we write implementations of friend template class A
and friend template function doSomething we create instance of class B for same
date type for which function doSomething or class A will be instantiated as shown
below otherwise compiler will generate error,

Virtual University of Pakistan

293

Object Oriented Programming (CS304)

VU

/*************************************************************
One template function and one template class
**************************************************************/
template< class U >
void doSomething( U ); // template function with template parameter U
template< class V >
class A { }; // template function with template parameter V
/*************************************************************
Making template function and template class friends of class B
Note that in code of class B below we are using same type parameter T for instantiating
template function doSomething and template class A.
**************************************************************/
template< class T >
class B {
int data;
friend void doSomething( T ); // granting friendship to template
doSomething in // class B
friend A< T >; // granting friendship to class A in class B
};
/*************************************************************
Implementation of template function doSometing and template class A
As we have used same type parameter T in class B for declaring doSomething and class A as
friend so now we can only instantiate object of class B according to the passed parameter
**************************************************************/
template< class U >
void doSomething( U u ) { // here template T will be passed as U
B< U > ib;
// it is OK as we are instantiating class B for same type that is passed from class B
i.e // T
ib.data = 78;
}
int main() {
int i = 5;
char c = x;
doSomething( i );
// OK
doSomething( c );
// OK
return 0;
}

Now suppose we change instantiation of doSomething and in it we cerate B class

instance for int,
template< class U >
void doSomething( U u ) {
B< int > ib;
ib.data = 78;

294

Virtual University of Pakistan

Object Oriented Programming (CS304)

VU

Now we can only create doSomething instance for integers as shown below, in case
we create that function instance for any other date type error will be generated by
compiler ,
int main() {
int i = 5;
char c = x;
doSomething( i );
// OK
doSomething( c );
// Error!
return 0;
}

The reason for the error is that we are creating class B instance for int data type in
function doSomething(), but function doSomething itself has been instantiated for
char in main.
Same Rule applied to template friend class A is shown below,
class B< int > {
int data;
friend void doSomething( int );
friend A< int >;
};
template< class T >
class A {
void method() {
B< char > cb; // Error!
cb.data = 8;
B< int > ib; // OK
ib.data = 9;
}
};

38.4. Templates and Friends Rule 3

When a friend function / class template takes different type parameters from the
class template granting friendship, then its each instantiation is a friend of each
instantiation of the class template granting friendship. Basically here we are
removing restriction imposed by Rule 2, due to the use of same type parameter in
class B while declaring function doSomething and class A as its friends as shown
below,
template< class U >
void doSomething( U );
template< class V >

Virtual University of Pakistan

295

Object Oriented Programming (CS304)

class A { };

template< class T >

class B {
int data;
template< class W >
friend void doSomething( W ); // type name is W
template< class S >
friend class A; // type name is S
};

template< class U >

void doSomething( U u ) {
B< int > ib; // Now it is ok to use B for int in function doSomething
instantiated for // char in main
ib.data = 78;
}

int main() {
int i = 5;
char c = x;
doSomething( i );
doSomething( c );
return 0;
}

// OK
// OK

For classes same Rule is applicable as follows,

..
/******************************************************************
class A implementation
******************************************************************/
template< class T >
class A {
void method() {
B< char > cb; // OK!
cb.data = 8;
B< int > ib;
ib.data = 9;
}
};

38.5. Templates and Friends Rule 4

296

Virtual University of Pakistan

VU

Object Oriented Programming (CS304)

VU

Rule 4 says that when we declare a template as friend of any class then all kinds
specializations of that function explicit, implicit and partial, also becomes friends of
the class granting friendship.
template< class T >
class B {
T data;
template< class U >
friend class A; // granting friendship to class A for all data types
(Rule 3)
};
template< class U >
class A {
// general template class A
A() {
B< int > ib;
ib.data = 10; // OK
}
};
template< class U >
class A< U* > { // template class A specialized for pointer data types
A() {
B< int > ib;
ib.data = 10; // OK
}
};

In case of functions demonstration of this Rule is below,

template< class T >
class B {
T data;
template< class U >
friend void doSomething( U ); // granting friendship to
doSomething for all data types (According to Rule 3)
};
template< class U >
void doSomething( U u ) {
B< int > ib;
ib.data = 56; // OK
}
template< >
void doSomething< char >( char u ) {
B< int > ib;
ib.data = 56; // OK
}

Virtual University of Pakistan

297

Object Oriented Programming (CS304)

VU

Note:
You can see that there is need to concentrate on Rule 2 as it is somewhat difficult to
understand other rules are simple.

298

Virtual University of Pakistan

Object Oriented Programming (CS304)

VU

Lecture No.39
39.1. Templates & Static Members
We have studied in case of general classes that their static members are created and
initialized at class scope, meaning that memory is reserved for them when a class is
compiled and they are initialized as well at that time, they are not dependent on the
creation of any object of a class, all these tings are true in case of static members of
template classes as well with different that they are created when a template class is
instantiated for each date type, each instantiation of template class has its own copy
of static data members, suppose we have template class with static members,
compiler will create this template class implementation for different date types as
required, and each implementation will have its own copy of static data members.16
Consider the class A with a public static data member and a static member function.
(It is not good practice to make any data member public but it is declared as public
here to demonstrate that separate copy of this data member will be created for each
data type).
#include <cstdlib>
#include <iostream>
using namespace std;
template< class T >
class A {
public:
static int data;
static void doSomething( T & );
};
template<class T>
int A<T>::data = 0;
int main() {
A< int > ia;
A< char > ca;
ia.data = 5;
ca.data = 7;
cout << "ia.data = " << ia.data << endl << "ca.data = " << ca.data;
system("PAUSE");
return 0;
}
Output

ia.data = 5
ca.data = 7
16

We can take each implementation of template class for a particular data type as a general class
written explicitly for that data type.

Virtual University of Pakistan

299

Object Oriented Programming (CS304)

VU

This output clearly shows that we have two separate copies of data for int and char types.
Also note that each date type instantiation will have single copy of static members as
demonstrated below, here objects ia,ib and ic will have single copy of data being shared by
all of them as was the case for static members of a general class,

#include <cstdlib>
#include <iostream>
using namespace std;
template< class T >
class A {
public:
static int data;
static void doSomething( T & );
};
template<class T>
int A<T>::data = 0;
int main() {
A< int > ia, ib , ic;
ia.data = 5;
ib.data = 7;
ic.data = 9;
cout << "ia.data = " << ia.data << endl << "ib.data = " << ib.data << endl <<
"ic.data = " << ic.data << endl;
system("PAUSE");
return 0;
}
Output

ia.data = 9
ib.data = 9
ic.data = 9

As same data member is being shared so same member value is being changed
39.2. Templates Conclusion
Advantages:
Templates provide
Reusability
Writability
Disadvantages:

300

Virtual University of Pakistan

Object Oriented Programming (CS304)

VU

Can consume memory if used without care.

Templates may affect reliability of a program as they give same type of
implementation for all data types with may not be correct for a particular
data type like isEqual function given below may give incorrect result for char
*.
template< typename T >
bool isEqual( T x, T y ) {
return ( x == y );
}
As we are free to give any kind of implementation for a particular data type
we may given incorrect implementation, considering the above isEqual
function, it will produce incorrect result for char * (char arrays ) as it will
compare first value of both arrays only,
int main() {
char* str1 = Hello ;
char* str2 = World!;
isEqual( str1, str2 );
// Compiler accepts!
}

39.3. Generic Algorithms Revisited

We studied the concept of Generic Algorithms before that made our code type independent as
well as independent of underlying data structure,
For this we developed step by step the following find algorithm that is so generic that it
works for all containers,

template< typename P, typename T >

P find( P start, P beyond, const T& x ) {
while ( start != beyond && *start != x )
++start;
return start;
}

int main() {
int iArray[5];
iArray[0] = 15;
iArray[1] = 7;
iArray[2] = 987;

int* found;
found = find(iArray, iArray + 5, 7);
return 0;
}

39.4. Generic Algorithms Revisited

We claimed that this algorithm is generic, because it works for any aggregate object
(container) that defines following three operations

Virtual University of Pakistan

301

Object Oriented Programming (CS304)

VU

a. Increment operator (++)

b. Dereferencing operator (*)
c. Inequality operator (!=)
Let us now try to apply this generic algorithms to our template Vector class to
examine whether our generic algorithm find works on this container (our template
Vector class can store in it data of all types), for this our Vector class need to support
the three operations given above and we will have to add one more integer data
member index to track the traversal when we will apply increment operation (++),
we will also add its setter and getter members to access its value.
So now our modified Vector class is as follows,
Example Vector
template< class T >
class Vector {
private:
T* ptr;
int size;
int index;
// initialized with zero
public:
Vector( int = 10 );
Vector( const Vector< T >& );
T& operator [](int);
int getIndex() const;
void setIndex( int i );
T& operator *();
bool operator !=( const Vector< T >& v );
Vector< T >& operator ++();
};
template< class T >
int Vector< T >::getIndex() const {
return index;
}
template< class T >
void Vector< T >::setIndex( int i ) {
if ( index >= 0 && index < size )
index = i;
}
template< class T >
Vector<T>& Vector<T>::operator ++() {
if ( index < size )
++index;
return *this;
}
template< class T >
T& Vector< T >::operator *() {
return ptr[index];
}
template< class T >

302

Virtual University of Pakistan

Object Oriented Programming (CS304)

VU

bool Vector<T>::operator !=( Vector<T>& v ) {

if ( size != v.size || index != v.index )
return true;
for ( int i = 0; i < size; i++ )
if ( ptr[i] != v.ptr[i] )
return true;
return false;
}

Now we want to apply our generic find method on this Vector container, as our generic find
algorithm is taking three parameters we will create three instances of our Vector class as
follows,
1. First Vector instance to store our data (integers in this case).
2. Second Vector instance to pass address of beyond (One position next to first Vector
instance)
3. Third Vector instance to get result

int main() {
Vector<int> iv( 3 );
// First Vector instance to store integer values from which we need to find integer
// value
iv[0] = 10;
iv[1] = 20;
iv[2] = 30;
Vector<int> beyond( iv ),found( 3 );
// declaring Second (beyond) and third (found) instance in same statement
// beyond is initialized with first Vector instance iv
// third vector instance of length 3 will be used to store the result
beyond.setIndex( iv.getSize() ); // We have set index of beyond to 3 where no
// element exists in it so it is behaving as pointing to one element next to iv.
found = find( iv, beyond, 20 );
cout<<Index: <<found.getIndex();
return 0;
}

39.5. Generic Algorithm

Now we want to apply our generic algorithm find on this Vector container,
template< typename P, typename T >
P find( P start, P beyond, const T& x ) {
while ( start != beyond && *start != x )
++start;
return start;

Virtual University of Pakistan

303

Object Oriented Programming (CS304)

VU

}
This algorithm will run as follows,
First iteration:
Iteration No
iv size beyond
#
size
3
3
1
3
3
2
3
3
3
4
3
3

iv
index
0
1
2
3

beyond
index
3
3
3
3

start !=
beyond
true
true
true
false

*start != x
true
true
true
undefined

In above scenario this algorithm will run till second iteration only as element 20 will
be found iv will be returned with index equal to 1 pointing to second element (20).
As you can see our generic algorithm is working fine with our container Vector class
instance.
However there are some other things to consider with this generic algorithm
implementation.
39.6. Problems
a. Our algorithm doesnt support multiple traversals
We can move forward in single steps that may be very inefficient approach in case of
huge Vector object with large number of values, there is no facility to make multiple
movements like one in forward direction and at the same time once in reverse
direction,
b. Our algorithm shows inconsistent behavior
We generally use pointers to mark a position in a data structure of some primitive
type and then return it but here we are using whole container as marker and return it
e.g. we returned whole found object of Vector class to tell whether we find the value
or not we could have done that using single pointer value.
c. Supports only a single traversal strategy
Also there is no way to change traversal strategy (moving beyond more than one
value in single step).

304

Virtual University of Pakistan

Object Oriented Programming (CS304)

VU

Lecture No.40
Recap
In previous lecture we studied that we can apply Generic Algorithms on any
container that supports three operations (++, *, !=)
Then we applied Generic Algorithms find on our Vector class by giving
implementation of these three operations in Vector class
At the end we saw that this approach has some drawbacks as given below,

No support for multiple traversals

Supports only a single traversal strategy
Inconsistent behavior

We saw that we have to made distinct three Vector objects for a simple find method
this approach may not be efficient when we have a lot of elements in Vector object.
Now we see how we can use pointers to modify this approach, we will use three
cursors (pointers) to vector class to avoid creating three Vector objects, in this
approach we will use pointers to store vector elements addresses (basic concept is
similar to link lists)
40.1. Cursors
A better way is to use cursors
A cursor is a pointer that is declared outside the container / aggregate object
Aggregate object provides methods that help a cursor to traverse the elements
T* first()
T* beyond()
T* next( T* )
In this approach we will add three methods given above in our Vector class that will
return pointer to elements of Vector elements instead of whole vector object.
Now our vector class is as follows,
Vector
template< class T >
class Vector {
private:
T* ptr;
int size;
public:
Vector( int = 10 );
Vector( const Vector< T >& );
~Vector();
int getSize() const;
const Vector< T >& operator =( const Vector< T >& );
T& operator []( int );
T* first();
T* beyond();
T* next( T* );
};
template< class T >
T* Vector< T >::first() {

Virtual University of Pakistan

305

Object Oriented Programming (CS304)

VU

return ptr;
}
template< class T >
T* Vector< T >::beyond() {
return ( ptr + size );
}
template< class T >
T* Vector< T >::next( T* current )
{
if ( current < (ptr + size) )
return ( current + 1 );
// else
return current;
}
int main() {
Vector< int > iv( 3 );
iv[0] = 10;
iv[1] = 20;
iv[2] = 30;
int* first = iv.first();
int* beyond = iv.beyond();
int* found = find(first,beyond,20);
return 0;
}

Note our generic find method remained as it as shown below, simply we used
cursors (pointers) in our main method to make our code so efficient.

template< typename P, typename T >

P find( P start, P beyond, const T& x ) {
while ( start != beyond && *start != x )
++start;
return start;
}

Cursors-Usage
This technique works fine for a contiguous sequence like c++ arrays where elements
are placed at consecutive memory locations or Vector whice we have implemented in
terms of arrays with extra functionality to overcome shortcomings of c++ arrays,
however cursors does now work with containers that use complicated data
structures and are non contiguous because in this case we can not simply use ++
operation that we used on our find method, in that case we have to rely on the that
container own traversal operations that results in inefficient approach of last lecture,

306

Virtual University of Pakistan

Object Oriented Programming (CS304)

VU

Example Works Fine

This cursor works fine when we have

Cursor
Example Problem

Cursor

Example Problem
In case of non contiguous container we have issues because when we apply find
method on non contiguous containers our find method increment operation fails as
elements are not placed at next locations in order, in code below we are calling find
method for non contiguous container Set that we are going to study in next lectures
in this case our find method will give error as indicated,

int main() {
Set< int > is( 3 );
is.add( 10 );
is.add( 20 );
is.add( 30 );
ET* first = iv.first();
ET* beyond = iv.beyond();
ET* found = find(first, beyond, 20);
return 0;
}

template< typename P, typename T >

P find( P start, P beyond, const T& x ) {
while ( start != beyond && *start != x )
++start;
// Error
return start;

Virtual University of Pakistan

307

Object Oriented Programming (CS304)

VU

In this case we need to use that container own increment operation like shown
below, then our find algorithm will work fine,
template< typename CT, typename ET >
P find( CT& cont, const ET& x ) {
ET* start = cont.first();
ET* beyond = cont.beyond();
while ( start != beyond && *start != x )
start = cont.next( start );
return start;
}

int main() {
Set< int > is( 3 );
is.add( 10 );
is.add( 20 );
is.add( 30 );
int* found = find( is, 20 );
return 0;
}

Cursors Conclusion
The main benefit is that we are using external pointer so we can do now any kind of
traversal in any way i.e. as many traversals as we need only we will need one pointer
for each traversal, however we can not use cursors in place of pointers for all
containers.
40.2. Iterators
We studied cursors previously, cursors were external pointer that we accessing
internal data of any container like vector, it is against the principle of data hiding as
we can access any container data using cursors so it is not good programming
practice to given access in container for the use of cursors (first, next, beyond
methods) we have alternate to cursors in the form of Iterators which are that traverse
a container without exposing its internal representation. Mean they are not external
pointers but internal data member of container. Iterators are for containers exactly
like pointers are for ordinary data structures (you can see this line as we have made a
mechanism to declare pointers to our containers like we declare pointers to ordinary
data types, in case of cursors we were using ordinary data type pointer but in case of
Iterators we will use container pointers without exposing their internal details)
Generic Iterators
General Iterator class can point to any container because it is implemented using
templates, basically it provides us the functionality to create any container object
using templates and operator overloading this has been explained in code below,
now we will not directly use methods first , beyond and next of container but we

308

Virtual University of Pakistan

Object Oriented Programming (CS304)

VU

will use them through Iterator class i.e we will create Iterator class objects and will
iterate container through these methods, however these objects will not be able to
directly access internal elements of container. A generic Iterator works with any kind
of container. We need the same set of operations in container class to use Iterators,
T* first()
T* beyond()
T* next( T* )
Example Generic Iterator

Iterator
operator *
operator ++
operator !=

Container
first()
beyond()
next()

Generic Iterator

template< class CT, class ET >

// Template Iterator class taking two Vector type parameters one for CT (pointer to
// container type) and ET (pointer to individual element of the container)
class Iterator {
CT* container;
ET* index;
public:
Iterator( CT* c, bool pointAtFirst = true );
Iterator( Iterator< CT, ET >& it );
Iterator& operator ++();
ET& operator *();
bool operator !=( Iterator< CT, ET >& it );
};
template< class CT, class ET >
Iterator< CT, ET >::Iterator( CT* c, bool pointAtFirst ) { // constructor
container = c;
if ( pointAtFirst )
index = container->first();
else
index = container->beyond();
}
template< class CT, class ET >
Iterator< CT, ET >::Iterator( Iterator< CT, ET >& it ) { // standard copy constructor
container = it.container;

Virtual University of Pakistan

309

Object Oriented Programming (CS304)

VU

index = it.index;
}
template< class CT, class ET >
Iterator<CT,ET>& Iterator<CT,ET>:: operator ++() { // increment operator
index = container->next( index );
return *this;
}
template< class CT, class ET >
ET& Iterator< CT, ET >::operator *() // deference operator
{
return *index;
}
template< class CT, class ET >
bool Iterator< CT, ET >::operator !=( Iterator< CT, ET >& it ) {
// inequality operator result will be false if containers are pointing
// to same memory location or at same index otherwise true
if ( container != it.container || index != it.index )
return true;
// else
return false;
}

int main() {
Vector< int > iv( 2 );
Iterator < Vector<int>, int > it( &iv ) , beyond( &iv, false );
// creating two Iterator objects of type <Vector <int >, int >
// i.e Iterator will have Vector class object for integers and element type will be
// integers so index will be integer pointer
// first object it is taking Vector class object iv as reference and making iterate
//index to point to its first element as bool pointAtFirst is true by default
// second object beyond is also taking same Vector object as reference
// so that it may also point to it by as poinAtFrist is false now so index will point
// to one position beyond the last element in iv
iv[0] = 10;
iv[1] = 20;
Iterator< Vector<int>, int > found = find( it, beyond, 20 );
// creating Iterator object found of type <Vector <int >, int > and saving in it
// result of found note that now we have saved index of our vector
// class element if element is found, without creating extra element of Vector class.
return 0;
}

310

Virtual University of Pakistan

Object Oriented Programming (CS304)

VU

Note that in above code we are NOT creating multiple objects of Vector class instead we are
creating objects of Iterator class against one Vector class object

template< typename P, typename T >

P find( P start, P beyond, const T& x ) {
while ( start != beyond && *start != x )
++start;
return start;
}

Iterators Advantages
a. With Iterators more than one traversal can be pending on a single
container
b. Iterators allow to change the traversal strategy without changing the
aggregate object
c. They contribute towards data abstraction by emulating pointers

Virtual University of Pakistan

311

Object Oriented Programming (CS304)

VU

Lecture No.41
41.1. Standard Template Library:
When we started discussing about templates and generic algorithms, our basic aim
was to make standard solutions for different problems (like searching or comparison)
that should work for all data types and for all containers.
C++ programmers started working on these concepts from very beginning and gave
many standard solutions to these problems these standard solutions after the
approval of C++ standardization committee were added to a combined library name
as Standard Template Library (STL). STL is designed to operate efficiently across
many different applications; we can use solutions from this template library in our
programs using different header files.

Standard Template Library

STL consists of three key components
Containers
Iterators
Algorithms
STL Promotes Reuse
STL promotes reuse as we dont need to rewrite the already written standard code
for different problems, it saves our development time and cost. Secondly these
solutions have been thoroughly tested so there is no change of error due to their use.
41.2. STL Containers
Container is an object that contains a collection of data elements like we have studied
before now we will study them in detail.
STL provides three kinds of containers,
1. Sequence Containers
2. Associative Containers
3. Container Adapters
Sequence Containers
A sequence organizes a finite set of objects, all of the same type, into a strictly linear
arrangement
Sequence Containers
a. vector

Rapid insertions and deletions at back end

Random access to elements

b. deque17
17

deque behaves like queue (line) such that we can add elements on both side of it.

312

Virtual University of Pakistan

Object Oriented Programming (CS304)

VU

Rapid insertions and deletions at front or back

Random access to elements
c. list18

Doubly linked list

Rapid insertions and deletions anywhere

Example STL Vector

#include <vector>
int main() {
std::vector< int > iv;
int x, y;
char ch;
do {
cout<<"Enter the first integer:";
cin >> x;
cout<<"Enter the second
integer:";
cin >> y;
iv.push_back( x );
iv.push_back( y );
cout << Current capacity of iv = << iv.capacity() << endl;
cout << Current size of iv =<< iv.size() << endl;
cout<<"Do you want to continue?";
cin >> ch;
} while ( ch == 'y' );
}

Sample Output
Enter the first integer: 1
Enter the second integer: 2
Current capacity of iv = 2
Current size of iv = 2
Do you want to continue? y

Sample Output
Enter the first integer: 3
Enter the second integer: 4
18

list is linear data structure but can not directly move to any element using index we have to move to a
certain by moving sequentially from start element.

Virtual University of Pakistan

313

Object Oriented Programming (CS304)

Current capacity of iv = 4
Current size of iv = 4
Do you want to continue? y

Sample Output
Enter the first integer: 5
Enter the second integer: 6
Current capacity of iv = 8
Current size of iv = 6
Do you want to continue? n

Example STL Deque

#include <deque>
int main() {
std::deque< int > dq;
dq.push_front( 3 );
dq.push_back( 5 );
dq.pop_front();
dq.pop_back()
return 0;
}

Example STL List

#include <list>
int main() {
std::list< float > _list;
_list.push_back( 7.8 );
_list.push_back( 8.9 );
std::list< float >::iterator it = _list.begin();
_list.insert( ++it, 5.3 );
return 0;
}

Associative Containers

314

Virtual University of Pakistan

VU

Object Oriented Programming (CS304)

VU

An associative container provide fast retrieval of data based on keys mean we add
elements in these containers using some formula and retriever the elements using the
same formula again. It ensures that we do not have to traverse the container one by
one but we directly move to the required element for example formula may be,
Value % 10 (remainder function),
Using this formula,
6 will be stored at 6th place
11 will be stored at 1st place
13 will be strored at 3rd place
and so on
Note that this remainder function will always result in a value between 0 and 9 so if
we have an array of size 10 we can use this remainder function easily.
When we need to retrieve a value we will use inverse process to find index of value
and will then retrieve it.
Associative Containers
set
No duplicates
multiset
Duplicates allowed
map
No duplicate keys
multimap
Duplicate keys allowed
Example STL Set
#include <set>
int main() {
std::set< char > cs;
cout << Size before insertions: << cs.size() << endl;
cs.insert( a );
cs.insert( b' );
cs.insert( b' );
cout << Size after
insertions:
<< cs.size();
return 0;
}

Output
Size before insertions: 0
Size after insertions: 2

Virtual University of Pakistan

315

Object Oriented Programming (CS304)

Example STL Multi-Set

#include <set>
int main() {
std::multiset< char > cms;
cout << "Size before insertions: " << cms.size() << endl;
cms.insert( 'a' );
cms.insert( 'b' );
cms.insert( 'b' );
cout << "Size after
insertions: "
<< cms.size();
return 0;
}

Output
Size before insertions: 0
Size after insertions: 3

Example STL Map

#include <map>
int main() {
typedef std::map< int, char > MyMap;
MyMap m;
m.insert(MyMap::value_type(1, 'a'));
m.insert(MyMap::value_type(2, 'b'));
m.insert(MyMap::value_type(3, 'c'));
MyMap::iterator it = m.find( 2 );
cout << "Value @ key " << it->first << " is " << it->second;
return 0;
}

Output
Value @ key 2 is b

Example STL Multi-Map

#include <map>
int main() {
typedef std::multimap< int, char > MyMap;
MyMap m;
m.insert(MyMap::value_type(1, 'a'));
m.insert(MyMap::value_type(2, 'b'));
m.insert(MyMap::value_type(3, 'b'));

316

Virtual University of Pakistan

VU

Object Oriented Programming (CS304)

VU

MyMap::iterator it1 = m.find( 2 );

MyMap::iterator it2 = m.find( 3 );
cout << "Value @ key " << it1->first << " is " << it1->second << endl;
cout << "Value @ key " << it2->first << " is " << it2->second << endl;
return 0;
}

Output
Value @ key 2 is b
Value @ key 3 is b

First-class Containers
Sequence and associative containers are collectively referred to as the first-class
containers
Container Adapters
A container adapter is a constrained version of some first-class container
Container Adapters
a. stack19
Last in first out (LIFO)
Can adapt vector, deque or list
b. queue20
First in first out ( FIFO)
Can adapt deque or list
c. priority_queue21
Always returns element with highest priority
Can adapt vector or deque

41.3. Common Functions for All Containers

Common functions that can be applied to all containers are given below (these are basically
general functions we can use them with any class),

1. Default constructor
19
Stack is basically a data structure used to store and retrieve values in speedy way because it grows
and shrinks from one end only you can consider it like pile of plates present at somewhere plates can be
added or retrieved from front end only. It is very efficient approach however it follows last in first out
(LIFO) principle meaning things are retrieved in reverse order from stack.
20
Queue is also data structure and it is similar to queue of people waiting to submit a form or pay their
utility bills queue follows first in first out principle (FIFO).
21
Priority Queue also follows FIFO principle with difference that we add elements in priority queue
according to certain priority its example may be printer queue which has priority option the person send
print having higher priority will find its data at higher level of the queue than other people having low
priority

Virtual University of Pakistan

317

Object Oriented Programming (CS304)

VU

2. Copy Constructor
3. Destructor
4. empty()
o Returns true if container contains no elements
5. max_size()
o Returns the maximum number of elements
6. size()
o Return current number of elements
7. operator = ()
o Assigns one container instance to another
8. operator < ()
o Returns true if the first container is less than the second container
9. operator <= ()
o Returns true if the first container is less than or equal to the second
container
10. operator > ()
o Returns true if the first container is greater than the second container
11. operator >= ()
o Returns true if the first container is greater than or equal to the
second container
12. operator == ()
o Returns true if the first container is equal to the second container
13. operator != ()
o Returns true if the first container is not equal to the second container
14. swap ()
o swaps the elements of the two containers
41.4. Functions for First-class Containers
As first class containers are both sequence and associative containers both allow fast retrieval
of data values so they have following additional functions,

1. begin()
o Returns an iterator object that refers to the first element of the
container
2. end()
o Returns an iterator object that refers to the next position beyond the
last element of the container
3. rbegin()
o Returns an iterator object that refers to the last element of the
container
4. rend()
o Returns an iterator object that refers to the position before the first
element
5. erase( iterator )
o Removes an element pointed to by the iterator

318

Virtual University of Pakistan

Object Oriented Programming (CS304)

VU

6. erase( iterator, iterator )

o Removes the range of elements specified by the first and the second
iterator parameters
7. clear()
o erases all elements from the container
41.5. Container Requirements
As each container need to perform certain operations on the elements added to it like
their copy while creating another instance of container or their comparison while
performaing certain operation on them like their sorting , so the elements that we are
going to add in containers should provide this kind of basic functionality.
Examples of these functionalities are given below,

When an element is inserted into a container, a copy of that element is made

using,
o Copy Constructor
o Assignment Operator

So the elements need to be added to any container should provide copy and
assignment functionality.Builtin c++ data types already provide this functionalities
and we have studied that compiler also generates copy constructor and overloaded
assignment operator for our user defined data types like structures and classes if we
have not done so.

Associative containers and many algorithms compare elements so the

elements that are added to associative containers should have this
functionality,
o Operator ==
o Operator <

c++ doesnot provide functionality of comparison operator (==) or less than operator
(<) by itself so we have to provide this functionality by ourself in element class if we
want to use it in associative containers.

[STL components Iterators and Algorithms will be discussed in next lecture.]

Virtual University of Pakistan

319

Object Oriented Programming (CS304)

VU

Lecture No.42
42.1. Iterators
We have studied about Iterators before; they are used to traverse Containers
efficiently without accessing internal details, now we see Iterators provided to us in
standard template library, Iterators are for containers like pointers are for ordinary
data structures STL Iterators provide pointer operations such as * and ++
42.2. Iterator Categories
We can divide Iterators given to us in STL in the following categories,
a.
b.
c.
d.
e.

Input Iterators
Output Iterators
Forward Iterators
Bidirectional Iterators
Random-access Iterators

Input Iterators
Using input iterators we can read an element, and we can only move in forward
direction one element at a time, these can be used in implementing those algorithms
that can be solved in one pass (moving container once in single direction from start
to end like find algorithm we studied in last lecture).
Output Iterators
Using output iterators we can read an element, and we can only move in forward
direction one element at a time, these can be used in implementing those algorithms
that can be solved in one pass (moving container once in single direction from start
to end like find algorithm we studied in last lecture).
Forward Iterators
Forward iterators have the capabilities of both input and output Iterators, in addition
they can bookmark a position in the container (we can set one position as bookmark
while traversing the container this will be more understandable when we will see
example below)
Bidirectional Iterators
They have all the capabilities of forward Iterators plus they can be moved in
backward direction, as a result they support multi-pass algorithms (algorithms that
need more that need to traverse container more than once).
Random Access Iterators
They have all the capabilities of bidirectional Iterators plus they can directly access
any element of a container.
42.3. Iterator Summary:
Following diagram shows the capabilities and scope of different iterators you can see
that Random access iterators have all the capabilities and input and output iterators
have least capabilities,

320

Virtual University of Pakistan

Object Oriented Programming (CS304)

VU

Random Access
Bidirectional
Forward
Input

Output

42.4. Container and Iterator Types:

We can use the following different types of iterators with different types of
containers, (it is according to the nature of container)
42.5. Sequence Containers

Container Type

Iterator Type

Reason

vector

random access

(as we can access any element

of vector using its index so
we can use random access
iterator)

deque

random access

(in deque we can add elements

only in front and back
however we can access any
element of deque using its
index so we can use random
access Iterator)

list

bidirectional

(in list we can move in both

directions in sequence,
however cannot access an
element at specific index
randomly so we can use
bidirectional iterator with
list)

Virtual University of Pakistan

321

Object Oriented Programming (CS304)

VU

42.6. Associative Containers

In associative containers we save values based on keys, and we cannot access
elements randomly based on indexes as elements are not stored at contiguous
memory locations, however we can traverse them in both directions so we can use
bidirectional iterators with them.

-- set
-- multiset
-- map
-- multimap

bidirectional
-- bidirectional
-- bidirectional
-- bidirectional

42.7. Container Adapters

Container adapters are made with special restrictions, most important restriction is
that they dont allow free traversal of their elements so we CANNOT use iterators
with them as given below,

-- stack
-- queue
-- priority_queue

-- (none)
-- (none)
-- (none)

42.8. Iterator Operations

Iterators support following operations,
All Iterators support,

++p
pre-increment an iterator
p++
post-increment an Iterator
Input Iterators support,
*p
Dereference operator used as rvalue (right value mean they can be used on right side
of expression) for reading only, not for assigning value lvalue (left value, assignment
only taken place when we are allowed to use them as left value)
p1 = p2
Assignment (two Iterators)
p1 == p2
Equality operator
p1 != p2
Inequality operator
p->

322

Virtual University of Pakistan

Object Oriented Programming (CS304)

VU

Access Operator

Output Iterators support,

*p
Dereference operator (here we can use it for assigning new value lvalue as iterator is
output iterator that can be used to set any value)
p1 = p2
Assignment

As Forward Iterators have combined properties of both input and output iterators so
they support operations of both input and output Iterators.
Bidirectional Iterators
As bidirectional iterators can move in backward direction also, so they support
decrementing operations also (moving pointer one element back),
--p
Pre-decrement operator
p-Post-decrement operator
Random-access Iterators
Besides the operations of bidirectional Iterators, they also support

p+i
Result is an iterator pointing at p + i
pi
Result is an iterator pointing at p i
p += i
Increment iterator p by i positions
p = i
Decrement iterator p by i positions
p[ i ]
Returns a reference of element at p + i
p1 < p2
Returns true if p1 is before p2 in the container
p1 <= p2
Returns true if p1 is before p2 in the container or p1 is equal to p2
p1 > p2
Returns true if p1 is after p2 in the container
p1 >= p2
Returns true if p1 is after p2 in the container or p1 is equal to p2

Virtual University of Pakistan

323

Object Oriented Programming (CS304)

Example Random Access Iterator

typedef std::vector< int > IntVector;
int main() {
const int SIZE = 3;
int iArray[ SIZE ] = { 1, 2, 3 };
IntVector iv(iArray, iArray + SIZE);
IntVector::iterator it = iv.begin();
cout << Vector contents: ;
for ( int i = 0; i < SIZE; ++i )
cout << it[i] << ", ";
return 0;
}

Sample Output
Vector contents: 1, 2, 3,

Example Bidirectional Iterator

typedef std::set< int > IntSet;
int main() {
const int SIZE = 3;
int iArray[ SIZE ] = { 1, 2, 3 };
IntSet is( iArray, iArray + SIZE
);
IntSet::iterator it = is.begin();
cout << Set contents: ;
for (int i = 0; i < SIZE; ++i)
cout << it[i] << ", "; // Error
return 0;
}

Example Bidirectional Iterator

typedef std::set< int > IntSet;
int main() {
const int SIZE = 3;
int iArray[ SIZE ] = { 1, 2, 3 };
IntSet is( iArray, iArray + SIZE
);
IntSet::iterator it = is.begin();
cout << Set contents: ;
for ( int i = 0; i < SIZE; ++i )
cout << *it++ << ", "; // OK
return 0;
}

324

Virtual University of Pakistan

VU

Object Oriented Programming (CS304)

VU

Sample Output
Set contents: 1, 2, 3,

Example Bidirectional Iterator

typedef std::set< int > IntSet;
int main() {
const int SIZE = 3;
int iArray[ SIZE ] = { 1, 2, 3 };
IntSet is( iArray, iArray + SIZE
IntSet::iterator it = is.end();
cout << Set contents: ;
for (int i = 0; i < SIZE; ++i)
cout << *--it << ", ";
return 0;
}

);

Sample Output
Set contents: 3, 2, 1,

Example Input Iterator

#include <iostream>
using std::cin;
using std::cout;
using std::endl;
#include <iterator>
int main() {
int x, y, z;
cout << "Enter three integers:\n";
std::istream_iterator< int > inputIt( cin );
x = *inputIt++;
y = *inputIt++;
z = *inputIt;
cout << "x = " << x << endl;
cout << "y = " << y << endl;
cout << "z = " << z << endl;
return 0;
}

int main() {
int x = 5;
std::istream_iterator< int > inputIt( cin );

Virtual University of Pakistan

325

Object Oriented Programming (CS304)

*inputIt = x;
return 0;

VU

// Error

Example Output Iterator

int main() {
int x = 1, y = 2, z = 3;
std::ostream_iterator< int > outputIt( cout, ", " );
*outputIt++ = x;
*outputIt++ = y;
*outputIt++ = z;
return 0;
}
Example Output Iterator
int main() {
int x = 1, y = 2, z = 3;
std::ostream_iterator< int > outputIt( cout, ", " );
x = *outputIt++;
// Error
return 0;
}

42.9. Algorithms
STL includes 70 standard algorithms
These algorithms may use Iterators to manipulate containers
STL algorithms also work for ordinary pointers and data structures
An algorithm works with a particular container only if that container supports a
particular Iterator category
A multi-pass algorithm for example, requires bidirectional Iterator(s) at least

Algorithm: Examples
Mutating-Sequence Algorithms (that required changing of elements position)
Copy
copy_backward
fill
fill_n
generate
generate_n

326

Virtual University of Pakistan

Object Oriented Programming (CS304)

VU

iter_swap
partition

Non-Mutating-Sequence Algorithms (that dont require changing of element

position)
adjacent_find
count
count_if
equal
find
find_each
find_end
find_first_of

Numeric Algorithms (involves mathematical calculation)

accumulate
inner_product
partial_sum
adjacent_difference

Example copy Algorithm

#include <iostream>
using std::cout;
#include <vector>
#include <algorithm>
typedef std::vector< int > IntVector;
int main() {
int iArray[] = {1, 2, 3, 4, 5, 6};
IntVector iv( iArray, iArray + 6 );
std::ostream_iterator< int >
output( cout, ", " );
std::copy( begin, end, output );
return 0;
}
Output
1, 2, 3, 4, 5, 6,

Virtual University of Pakistan

327

Object Oriented Programming (CS304)

Example fill Algorithm

#include <iostream>
using std::cout;
using std::endl;
#include <vector>
#include <algorithm>
typedef std::vector< int > IntVector;
int main() {
int iArray[] = { 1, 2, 3, 4, 5 };
IntVector iv( iArray, iArray + 5 );
std::ostream_iterator< int > output( cout, ", " );
std::copy( iv.begin(), iv.end(), output );
std::fill(iv.begin(), iv.end(), 0);
cout << endl;
std::copy( iv.begin(), iv.end(), output );
return 0;
}

328

Virtual University of Pakistan

VU

Object Oriented Programming (CS304)

VU

Lecture No.43
Techniques for Error Handling:
Sometimes our program terminates abnormally, sometimes they even crash the
system, these errors occur mostly due to incorrect memory access or due to
input/output error, sometimes it is our program fault and sometimes it is some
external resource error (like network or hard disk).
If we allow these errors to happen we may lose our work for example if a text editor
terminates abnormally without allowing us to save our work we will lost our work,
so it is important that we add some type of error handling mechanism in our
program, we use the following techniques for error handling,
a.
b.
c.
d.
e.

Abnormal termination
Graceful termination
Return the illegal value
Return error code from a function
Exception handling

43.1. Example Abnormal Termination

In abnormal termination we do nothing and our program is terminated abnormally
by operating system if case of any error without saving program data,

void GetNumbers( int &a, int &b ) {

cout << \nEnter two integers;
cin >> a >> b;
}
int Quotient( int a, int b ){
return a / b;
}
void OutputQuotient( int a, int b, int quo ) {
cout << Quotient of << a << and
<< b << is << quo << endl;
}
int main(){
int sum = 0, quot;
int a, b;
for (int i = 0; i < 10; i++){
GetNumbers(a,b);
quot = Quotient(a,b);
sum += quot;
OutputQuotient(a,b,quot);
}
cout << \nSum of ten quotients is << sum;
return 0;
}

Output

Virtual University of Pakistan

329

Object Oriented Programming (CS304)

VU

Enter two integers

10
10
Quotient of 10 and 10 is 1
Enter two integers
10
0
Program terminated abnormally

43.2. Graceful Termination

Program can be designed in such a way that instead of abnormal termination, that
causes the wastage of resources, program performs clean up tasks, mean we add
check for expected errors (using if conditions),
Example Graceful Termination
int Quotient (int a, int b ) {
if(b == 0){
cout << Denominator cant << be zero << endl;
// Do local clean up
exit(1);
}
return a / b;
}

Output
Enter two integers
10
10
Quotient of 10 and 10 is 1
Enter two integers
10
0
Denominator cant be zero

43.3. Error Handling

a. Return Illegal Value
The clean-up tasks are of local nature only; there remains the possibility of
information loss.
Example Return Illegal Value
int Quotient(int a, int b){
if(b == 0)

330

Virtual University of Pakistan

Object Oriented Programming (CS304)

VU

b = 1;
OutputQuotient(a, b, a/b);
return a / b ;
}
int main() {
int a,b,quot; GetNumbers(a,b);
quot = Quotient(a,b);
return 0;
}

Output
Enter two integers
10
0
Quotient of 10 and 1 is 10

Error Handling
b. Return Error Code
Programmer has avoided the system crash but the program is now in an inconsistent
state
Example Return Error Code
bool Quotient ( int a, int b, int & retVal ) {
if(b == 0){
return false;
}
retVal = a / b;
return true;
}

Part of main Function

for(int i = 0; i < 10; i++){
GetNumbers(a,b);
while ( ! Quotient(a, b, quot) ) {
cout << Denominator cant be << Zero. Give input again
\n;
GetNumbers(a,b);
}
sum += quot;
OutputQuotient(a, b, quot);
}

Output

Virtual University of Pakistan

331

Object Oriented Programming (CS304)

VU

Enter two integers

10
0
Denominator cant be zero. Give input again.
Enter two integers
10
10
Quotient of 10 and 10 is 1
...//there will be exactly ten quotients

Error Handling

Issues in Error Handling:

Programmer sometimes has to change the design to incorporate error
handling
Programmer has to check the return type of the function to know whether an
error has occurred
Programmer of calling function can ignore the return value
The result of the function might contain illegal value, this may cause a system
crash later
Programs Complexity Increases
The error handling code increases the complexity of the code
o Error handling code is mixed with program logic
o The code becomes less readable
o Difficult to modify

The example below shows these concepts,

Example- without error handling

int main() {
function1();
function2();
function3();
return 0;
}

Example with error handling

int main(){
if( function1() ) {
if( function2() ) {
if( function3() ) {
...
}
else cout << Error Z has occurred;

332

Virtual University of Pakistan

Object Oriented Programming (CS304)

}
else

VU

cout << Error Y has occurred;

}
else cout << Error X has occurred;
return 0;
}

43.4. Exception Handling

Exception handling is a much elegant solution as compared to other error

handling mechanisms
It enables separation of main logic and error handling code

43.5. Exception Handling Process

Programmer writes the code that is suspected to cause an exception in try

block
Code section that encounters an error throws an object that is used to
represent exception
Catch blocks follow try block to catch the object thrown

Syntax Throwing an exception

The keyword throw is used to throw an exception
Any expression can be used to represent the exception that has occurred
o throw X;
o throw (X);
Examples
int a;
Exception obj;
throw 1;
// throw with literal (const)
throw (a);
// throw with variable
throw obj;
// throw with object
throw Exception(); // anonymous object
throw 1+2*9; // mathematical expression

Primitive data types may be avoided as throw expression, as they can cause
ambiguity
Define new classes to represent the exceptions that has occurred
o This way there are less chances of ambiguity

Syntax Try and Catch

int main () {
try {
... exception with be thrown here in case of error
}

Virtual University of Pakistan

333

Object Oriented Programming (CS304)

VU

catch ( Exception1 ) { // exception with be caught here if exception

was thrown
...
}
catch ( Exception2 obj ) { // exception with be caught here if exception
was
// thrown
...
}
return 0;
}
Catch Blocks
Catch handler must be preceded by a try block or an other catch handler
Catch handlers are only executed when an exception has occurred
Catch handlers are differentiated on the basis of argument type
The catch blocks are tried in order they are written
They can be seen as switch statement that do not need break keyword
Complete Example of try catch and throw:

class DivideByZero { // class just use to indicate and throw an exception

public:
DivideByZero() {
}
};
int Quotient(int a, int b){
if(b == 0){
throw DivideByZero(); // throwing above class object as exception in
// case of error
}
return a / b;
}
Related part of main getting user input and calling Quotient method in try catch
block this method will throw exception (DividebyZero class object) if user enter no. b
(denominator, divider) equal to zero and this exception will be caught in catch block
below, (you can write it as complete program to test the actual working of program)

for(int i = 0; i < 10; i++) {

try{
GetNumbers(a,b);
quot = Quotient(a,b);
OutputQuotient(a,b,quot); sum += quot;
}
catch(DivideByZero) {
i--;
cout << \nAttempt to divide

334

Virtual University of Pakistan

Object Oriented Programming (CS304)

VU

numerator with zero;

}
}

Output

Enter two integers

10
10
Quotient of 10 and 10 is 1
Enter two integers
10
0
Attempt to divide numerator (dividend) with zero
...

Catch Handler

The catch handler catches the DivideByZero exception through anonymous

object
Program logic and error handling code are separated
We can modify this to use the object to carry information about the cause of
error

Separation of Program Logic and Error Handling

You can see now that error handling code is separated from man logic of program as shown
below,

int main() {
try {
function1();
function2();
function3();
}
catch( ErrorX) { ... }
catch( ErrorY) { ... }
catch( ErrorZ) { ... }
return 0;
}

Virtual University of Pakistan

335

Object Oriented Programming (CS304)

VU

Lecture No.44
Previous Lecture Example of Exception Handling:

class DivideByZero {
public:
DivideByZero() {
}
};
int Quotient(int a, int b){
if(b == 0){
throw DivideByZero();
}
return a / b;
}
//main Function
int main() {
try{

quot = Quotient(a,b);

}
catch(DivideByZero) {

}
return 0;
}

Now we want to see what happens actually to local variables in try block when an
exception is thrown, this concept is called stack unwinding which tells how try catch
blocks are un winded (executed ) when there are nested function calls involving try
catch blocks or nested try catch blocks themselves.
44.1. Stack Unwinding
The flow control ( the order in which code statements and function calls are made) as
a result of throw statement is referred as stack unwinding
Stack Unwinding can take place in the following two ways,
1. When we have nested try catch blocks (one try catch block into other try catch
block), for example
try {
try {

} catch( Exception e) {
}
} catch(exception e){

336

Virtual University of Pakistan

Object Oriented Programming (CS304)

VU

}
2. When exception is thrown from nested functions having try catch blocks
void function1() {
throw Exception();
}
void function2() {
function1();
}
int main() {
try{
function2();
} catch( Exception ) { }
return 0;
}
Stack unwinding is more complex than simple nested function calls (or recursive
function calls) as in case of nested try catch block, exception can be thrown from any
try block, so transfer of control to catch handler is complex, we see this is detail,
First note these points,
All the local objects of an executing block are destroyed when an exception is
thrown
Dynamically allocated memory is not destroyed automatically
If no catch handler catches the exception the function terminate is called,
which by default calls function abort
Examples
Nested Functions example:
In example below we have two functions function1 and function2, function2 is
calling function1, in function1 we have added exception throwing code so it is
necessary now to call function1 in try catch blocks otherwise compiler will generated
an error, we are calling function2 in main, now note that function2 ifself is calling
function1 that needs try catch block so we need to call function2 in try catch block,
(otherwise compiler will generate an error), in case function1 code generates an
exception stack unwinding takes place control will be returned to function2 which
will return control to main the diagram below code explain this concept,
void function1() {
throw Exception();
}
void function2() {
function1();
}
int main() {

Virtual University of Pakistan

337

Object Oriented Programming (CS304)

VU

try{
function2();
} catch( Exception ) { }
return 0;
}

function1()
function2()
main()
function2()
main()
main()

Nested Try catch blocks example:

The stack unwinding is also performed if we have nested try catch blocks,
Example
int main( ) {
try {
try {
throw 1;
}
catch( float ) { }
}
catch( int ) {
}
return 0;
}

When an exception will be thrown from a try block control will go to catch block of
that try block if it is not appropriate catch block then control will go to ctach blocks of
other try blocks above it one by one, Stack Unwinding will occur in following order
in above example,

If exception is thrown from innermost try block,

338

Virtual University of Pakistan

Object Oriented Programming (CS304)

VU

Firstly the catch handler with float parameter is tried (innermost), this
catch handler will not be executed as its parameter is of different type
no coercion (match)
Secondly the catch handler with int parameter is tried and executed

If exception is thrown from outer try block, then as there is no other

try block above it so only this block catch handler will be matched
with exception if it matches catch block will be exectuted otherwise
default terminate and abort functions (discussed in this lecture start)
will be called.

Catch Handler

We can modify the code in catch handler to use the exception object to carry
information about the cause of error
The exception object thrown is copied to the object given in the handler
We pass the exception as reference instead of by value in the catch handler to
avoid problem caused by shallow copy

Example
We have added a method Print in our exception class that to show the user cause of
error,
class DivideByZero { // exception class
int numerator;
public:
DivideByZero(int i) { // constructor taking one parameter (dividend)
numerator = i;
}
void Print() const{
cout << endl << numerator
<< was divided by zero;
}
};
int Quotient(int a, int b) {
if(b == 0){
throw DivideByZero(a);
}
return a / b;
}
for ( int i = 0; i < 10; i++ ) {
try {
GetNumbers(a, b);
quot = Quotient(a, b); ...
} catch(DivideByZero & obj) {
obj.Print();
i--;
}
}
Virtual University of Pakistan

339

Object Oriented Programming (CS304)

VU

Body of main Function

Output
Enter two integers
10
10
Quotient of 10 and 10 is 1
Enter two integers
10
0
10 was divided by zero
...

Catch Handler
The object thrown as exception is destroyed when the execution of the catch handler
completes
Avoiding too many Catch Handlers
There are two ways to catch more then one object in a single catch handler
Use inheritance
Catch every exception
Inheritance of Exceptions
In inheritance we Group all exceptions according to their categories and catch single
exception for whole category for example for code below we have divided the
exceptions as follows,
Math exceptions (Divide by Zero and IntegerOutOfRange exception)
Input Output exceptions(InputStreamError)

try{
...
}
catch(DivideByZero){
...
}
catch(IntergerOutOfRange){
...
}
catch (InputStreamError){
}

ExampleWith Inheritance

340

Virtual University of Pakistan

Object Oriented Programming (CS304)

VU

try{
...
}
catch (MathError){
}
catch (InputStreamError){
}

Catch Every Exception

C++ provides a special syntax that allows to catch every object thrown
catch ( )
{
//...
}

Re-Throw
A function can catch an exception and perform partial handling
Re-throw is a mechanism of throw the exception again after partial handling
throw; /*without any expression*/
Example
int main ( ) {
try {
Function();
}
catch(Exception&) {
...
}
return 0;
}
void Function( ) {
try {
/*Code that might throw
an Exception*/
}
catch(Exception&) {
if( can_handle_completely ) {
// handle exception
} else {
// partially handle exception
throw; //re-throw exception
}
} // end of catch
} // end of function

Virtual University of Pakistan

341

Object Oriented Programming (CS304)

VU

Order of Handlers
Order of the more than one catch handlers can cause logical errors when using
inheritance or catch all (however compiler will not generate any error in this case)
Example
try{
...
}
catch (...) {
...
}
catch ( MathError ) { ...
}
catch ( DivideByZero ) {
...
}
// last two handlers can never be invoked as general exception class will catch all
// exceptions including the next two

342

Virtual University of Pakistan

Object Oriented Programming (CS304)

VU

Lecture No.45
45.1. Resource Management

Function acquiring a resource must properly release it

Throwing an exception can cause resource wastage

Example
int function1(){
FILE *fileptr = fopen(filename.txt,w);
...
throw exception();
...
fclose(fileptr);
return 0;
}

In case of exception the call to fclose will be ignored and file will remain opened.
We can remove this issue in following ways,
First Attempt
int function1(){
try{
FILE *fileptr = fopen(filename.txt,w);
fwrite(Hello World,1,11,fileptr);
...
throw exception();
fclose(fileptr);
} catch(...) {
fclose(fileptr); // adding fclose in catch handler as well
throw;
}
return 0;
}

But it is,code duplication.

Second Attempt:
Adding a separate class file handling what will have constructor to open file and
destructor to close it, as you know objects and local variables in try block are
destroyed automatically when try block complete its execution or in case exception is
thrown, so this file object will automatically be destroyed,

class FilePtr{
FILE * f;
Virtual University of Pakistan

343

Object Oriented Programming (CS304)

VU

public:
FilePtr(const char *name, const char * mode) {
f = fopen(name, mode);
}
~FilePtr() {
fclose(f);
}
operator FILE * () {
return f;
}
};
int function1(){
FilePtr file(filename.txt,w);
fwrite(Hello World,1,11,file);
throw exception();
...
return 0;
}

The destructor of the FilePtr class will close the file

Programmer does not have to close the file explicitly in case of error as well
as in normal case

Exception in Constructors
Exception thrown in constructor cause the destructor to be called for any object built
as part of object being constructed before exception is thrown
Destructor for partially constructed object is not called
Example
class Student{
String FirstName;
String SecondName;
String EmailAddress;

}
If the constructor of the SecondName throws an exception then the destructor for the
First Name will be called.
So generally we can say that in constructor if an exception is thrown than all objects created
so far are destroyed, if EmailAddress String object had thrown exception then SecondName
and FirstName objects will be destroyed using their destructor. However destructor of Student
class itself will not be called in any case as its object was not completely constructed.
Exception in Initialization List
Exception due to constructor of any contained object or the constructor of a parent
class can be caught in the member initialization list

344

Virtual University of Pakistan

Object Oriented Programming (CS304)

VU

Example
Student::Student (String aName) : name(aName)
/*The constructor of String can throw a exception*/
{
...
}
Exception in Initialization List
The programmer may want to catch the exception and perform some action to rectify
the problem
Example
Student::Student (String aName)
try
: name(aName) {
...
}
catch() {
}
Exceptions in Destructors
Exception should not leave the destructor, because when a destructor is running it
means that we have a stack unwinding going on that has run this destructor to delete
this object if this exception will be allowed to run it will run another stack unwinding
mechanism and this will be called leaving the exception from destructor and is not
allowed. C++ allows running only one stack unwinding process at a time.
If a destructor is called due to stack unwinding, and an exception leaves the
destructor then the function std::terminate() is called, which by default calls the
std::abort()
Example
class Exception;
class Complex{

public:
~Complex(){
throw Exception();
}
};
int main(){
try{
Complex obj;
throw Exception();

Virtual University of Pakistan

345

Object Oriented Programming (CS304)

VU

catch(){
}
return 0;
}
// The program will terminate abnormally

So add catch exception in destructor itself as shown below, in this case single stack
unwinding process may handle the situation.
Example
Complex::~Complex()
{
try{
throw Exception();
}
catch(){
}
}

Exception Specification
Program can specify the list of exceptions a function is allowed to throw
This list is also called throw list
If we write empty list then the function wont be able to throw any exception
Syntax
void Function1() {}
void Function2() throw () {}
void Function3( ) throw (Exception1, ){}
Here,

Function1 can throw any exception

Function2 cannot throw any Exception
Function3 can throw any exception of type Exception1 or any class derived
from it

Exception Specification

346

If a function throws exception other then specified in the throw list then the
function unexpected is called
The function unexpected calls the function terminate and terminates the
program
If programmer wants to handle such cases also then he must provide a
handler function and tell the compiler to call handler using set_unexpected

Virtual University of Pakistan

Object Oriented Programming (CS304)

VU

Course Review
We have studied the following topics in this course,

Object Orientation
What is an object
Object-Oriented Model
o Information Hiding
o Encapsulation
o Abstraction
Classes
Object Orientation

Inheritance
o Generalization
o Sub-Typing
o Specialization
IS-A relationship
Abstract classes
Concrete classes

Object Orientation

Multiple inheritance
Types of association
o Simple association
o Composition
o Aggregation
Polymorphism

Classes C++ Constructs

Classes
o Data members
o Member functions
Access specifier
Constructors
Copy Constructors
Destructors

Classes C++ Constructs

this pointer
Constant objects
Static data member
Static member function
Dynamic allocation

Virtual University of Pakistan

347

Object Oriented Programming (CS304)

Classes C++ Constructs

Friend classes
Friend functions
Operator overloading
o Binary operator
o Unary operator
o operator []
o Type conversion

Inheritance C++ Constructs

Public inheritance
Private inheritance
Protected inheritance
Overriding
Class hierarchy

Polymorphism C++ Constructs

Static type vs. dynamic type

Virtual function
Virtual destructor
V-tables
Multiple inheritance
Virtual inheritance

Templates C++ Constructs

Generic programming
Classes template
Function templates
Generic algorithm
Templates specialization
o Partial Specialization
o Complete specialization

Templates C++ Constructs

Inheritance and templates

Friends and templates
STL
o Containers
o Iterators
o Algorithms

Writing Reliable Programs

348

Error handling techniques

o Abnormal termination

Virtual University of Pakistan

VU

Object Oriented Programming (CS304)

o
o
o
o

VU

Graceful termination
Return the illegal value
Return error code from a function
Exception handling

Please give your comments about this document at the email address
[email protected], in format
Page No.#

Issue/Error

Virtual University of Pakistan

349