Deshmukh - Java SE 11 Programmer I Study Guide 2019

Oracle©
Certified Professional Java SE 11 Programmer I

Fundamentals
(Exam Code 1Z0-815)
18th Sep 2019 Build 1.7
©Hanumant Deshmukh
www.enthuware.com
Ещё больше книг в нашей группе:

https://vk.com/javatutorial
For online information and ordering of this book, please contact [email protected]. For more
information, please contact:
Hanumant Deshmukh
4A Agroha Nagar, A B Road,
Dewas, MP 452001
INDIA
Copyright © 2019 by Hanumant Deshmukh All Rights Reserved.
No part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by
any means, electronic, mechanical, photocopying, recording, scanning or otherwise, except as permitted
under the relevant laws of copyright, without the prior written permission of the Author. Requests for
permission should be addressed to [email protected]
Limit of Liability/Disclaimer of Warranty: The the author make no representations or warranties with
respect to the accuracy or completeness of the contents of this work and specifically disclaim all warranties,
including without limitation warranties of fitness for a particular purpose. No warranty may be created or
extended by sales or promotional materials. The advice and strategies contained herein may not be suitable
for every situation. This work is sold with the understanding that the author is not engaged in rendering
legal, accounting, or other professional services. If professional assistance is required, the services of a
competent professional person should be sought. The author shall not be liable for damages arising
herefrom.
The fact that an organization or website is referred to in this work as a citation and/or a potential source of
further information does not mean that the author endorses the information the organization or Web site
may provide or recommendations it may make. Further, readers should be aware that Internet Web sites
listed in this work may have changed or disappeared between when this work was written and when it is
read.
Cover Design: Kino A Lockhart, LOXarts Development, http://www.loxarts.com .
TRADEMARKS: Oracle and Java are registered trademarks of Oracle America, Inc. All other trademarks
are the property of their respective owners. The Author is not associated with any product or vendor
mentioned in this book.
To my alma mater,
Indian Institute of Technology, Varanasi

Acknowledgments
I would like to thank numerous individuals for their contribution to this book.
Thank you to Liu Yang for being the Technical Editor and Lisa Wright for being
the copy editor. Thank you to Carol Davis and Robert Nyquist for technical
proof reading. Thank you to Aakash Jangid and Bill Bruening for validating all
the code snippets in this book.
Thank you to Maaike Van Putten for her inputs on the book cover design and to
Kino A. Lockhart (LOXarts Development, http://www.loxarts.com ) for
designing the cover.
This book also wouldn't be possible without many people at Enthuware,
including Paul A Prem, who have been developing mock exams for the past
fifteen years. Their experience helped fine tune several topics in this book.
I would also like to thank Bruce Eckel, the author of "Thinking In Java" for
teaching Java to me and countless others through his book.
I am also thankful to countless CodeRanch.com and Enthuware.com forum

participants for highlighting the topics that readers often find difficult to grasp
and also for showing ways to make such topics easy to understand.
Thank you to Edward Dang, Rajashekar Kommu, Kaushik Das, Gopal Krishna
Gavara, Dinesh Chinalachiagari, Jignesh Malavia, Michael Tsuji, Hok Yee
Wong, Ketan Patel, Anil Malladi, Bob Silver, Jim Swiderski, Krishna
Mangallampalli, Shishiraj Kollengreth, Michael Knapp, Rajesh Velamala,
Aamer Adam, and Raghuveer Rawat for putting up with me :)
I would like to thank my family for their support throughout the time I was busy
writing this book.
Finally, I am also thankful for the following readers for their help in improving
the content of this book through suggestions and by reporting errors:
1. Zheng-Yu Wang
About the Author
Hanumant Deshmukh is a professional Java architect, author, and director of

software consultancy firm BetterCode Infosoft Pvt. Ltd. Hanumant specializes in
Java based multi-tier applications in financial domain. He has executed projects
for some of the top financial companies such as JP Morgan Chase, UBS, and
Bank Of America. He started Enthuware.com more than fifteen yrs ago through
which he offers training courses and learning material for various Java
certification exam. He has also co-authored a best selling book on Java
Servlets/JSP certification, published by Manning in 2003.
Hanumant achieved his Bachelor of Technology from Institute of Technology,

Banaras Hindu University (now, IIT - Varanasi) in Computer Science in 1997
and his Masters in Financial Analysis from ICFAI in 2010. After spending more
than a decade working with amazing people in the United States, he returned
back to India to pursue a degree in Law. He is a believer in freedom of speech
and expression and works on promoting it in his spare time.
You may reach him at [email protected]

Exam Objectives
The following are the exam objectives as of this writing. Oracle may tweak the
objectives at any time so please verify the current objectives published at OCP
Java 11 Certification Part 1 Exam page at Oracle .
Understanding Java Technology and environment
Describe Java Technology and the Java development environment

Identify key features of the Java language
Working With Java Primitive Data Types and String APIs
Declare and initialize variables (including casting and promoting primitive

data types)
Identify the scope of variables
Use local variable type inference
Create and manipulate Strings
Manipulate data using the StringBuilder class and its methods
Working with Java Arrays
Declare, instantiate, initialize and use a one-dimensional array

Declare, instantiate, initialize and use two-dimensional array
Creating and Using Methods
Create methods and constructors with arguments and return values

Create and invoke overloaded methods
Apply the static keyword to methods and fields
Reusing Implementations Through Inheritance
Create and use subclasses and superclasses

Create and extend abstract classes
Enable polymorphism by overriding methods
Utilize polymorphism to cast and call methods, differentiating object type
versus reference type
Distinguish overloading, overriding, and hiding
Handling Exceptions
Describe the advantages of Exception handling and differentiate among

checked exceptions, unchecked exceptions, and Errors
Create a try-catch block and determine how exceptions alter normal
program flow
Create and invoke a method that throws an exception
Creating a Simple Java Program
Create an executable Java program with a main class

Compile and run a Java program from the command line
Create and import packages
Using Operators and Decision Constructs
Use Java operators including the use of parenthesis to override operator

precedence
Use Java control statements including if, else, break and continue
Create and use do/while loops, while loop, and for looping statements
including nested loops
Describing Objects and Classes
Declare and instantiate Java objects, and explain objects' lifecycles

(including creation, dereferencing by reassignment, and garbage collection)
Define the structure of a Java class
Read or write to object fields
Applying Encapsulation
Apply access modifiers

Apply encapsulation principles to a class
Programming Abstractly Through Interfaces
Create and implement interfaces

Distinguish class inheritance from interface inheritance including abstract
classes
Declare and use List and ArrayList instances
Understanding lambda Expressions
Understanding Modules
Describe the Modular JDK

Declare modules and enable access between modules
Describe how a modular project is compiled and run
Table of Contents
Introduction
0.1 Who is this book for?
0.2 How is this book different from others?
0.3 How is this book organized?
0.4 Will this be asked in the exam?
0.5 General tips for the exam
0.6 Official Exam Details and Exam Objectives
0.7 Feedback and Reviews
Chapter 1 Kickstarter for Beginners

1.1 Key points in OOP
1.1.1 A matter of perspective
1.1.2 API
1.1.3 Type, class, enum, and interface
1.2 Why is something so?
1.3 Declaration and Definition
1.4 Relation between a class, an object, and a reference
1.5 Static and Instance
1.6 Stack and Heap
1.7 Conventions
1.7.1 What is a Convention?
1.7.2 Conventions in Java
1.8 Compilation and Execution
1.8.1 Compilation and Execution
1.8.2 Running a single file source code program
1.8.3 Packaging classes into Jar
1.8.4 Compilation error vs exception at run time
1.9 Commonly used terms in Java development
1.10 Java Identifiers
Chapter 2 Creating a Simple Java Program

2.1 Create an executable Java program with a main class
2.1.1 The main method
2.1.2 Command line arguments
2.1.3 The end of main
2.2 Run a Java program from the command line
2.3 Create and import packages
2.3.1 The package statement
2.3.2 Quiz
2.3.3 The import statement
2.3.4 Quiz
2.4 Exercises
Chapter 3 Working With Java Primitive Data Types

3.1 Data types and Variable types
3.2 Reference variables and primitive variables
3.3 Declare and initialize variables
3.3.1 Declare and initialize variables
3.3.2 Uninitialized variables and Default values
3.3.3 Assigning values to variables
3.3.4 final variables
3.4 Wrapper Classes
3.4.1 What are wrapper Classes?
3.4.2 Creating wrapper objects
3.4.3 Converting wrapper objects to primitives
3.5 Exercise
Chapter 4 Describing and Using Objects and Classes

4.1 Declare and instantiate Java objects
4.2 Read or write to object fields
4.2.1 Accessing object fields
4.2.2 What is "this"?
4.3 Define the structure of a Java class
4.3.1 Class disambiguated
4.3.2 Structure of a java source file
4.3.3 Members of a class
4.3.4 Relationship between Java source file name and class name
4.3.5 Quiz
4.4 Identify the scope of variables
4.4.1 Scope of variables
4.4.2 Scope and Visibility
4.4.3 Scope and Lifespan
4.4.4 Scopes Illustrated
4.4.5 Scope for the Exam
4.4.6 Quiz
4.5 Local variable type inference
4.6 Explain objects' lifecycles
4.6.1 Life cycle of an Object
4.6.2 Garbage Collection
4.6.3 Garbage Collection for the exam
4.7 Exercise
Chapter 5 Working with String APIs

5.1 Create and manipulate Strings
5.1.1 What is a "string"?
5.1.2 Creating Strings
5.1.3 String interning
5.1.4 String immutability
5.1.5 Manipulating Strings
5.1.6 Comparing strings
5.2 Manipulate data using the StringBuilder class and its methods
5.2.1 Why StringBuilder
5.2.2 StringBuilder API
5.3 Exercise
Chapter 6 Using Operators

6.1 Java Operators
6.1.1 Overview of operators available in Java
6.1.2 Expressions and Statements
6.1.3 Post and Pre Unary Increment/Decrement Operators
6.1.4 String concatenation using + and += operators
6.1.5 Numeric promotion and casting
6.1.6 Operator precedence and evaluation of expressions
6.2 Exercise
Chapter 7 Using Decision Constructs

7.1 Create if and if/else constructs
7.1.1 Basic syntax of if and if-else
7.1.2 Usage of if and if-else in the exam
7.2 The ternary conditional operator ?:
7.3 Creating a switch statement
7.4 Exercise
Chapter 8 Using Loop Constructs

8.1 What is a loop
8.2 Create and use while loops
8.2.1 The while loop
8.2.2 Using a while loop
8.3 The do-while loop
8.4 Create and use for loops
8.4.1 Going from a while loop to a for loop
8.4.2 Syntax of a for loop
8.4.3 Parts of a for loop
8.5 Create and use for each loops
8.5.1 The enhanced for loop
8.5.2 Syntax of the enhanced for loop
8.5.3 Enhanced for loop in practice
8.6 Use break and continue
8.6.1 Terminating a loop using break
8.6.2 Terminating an iteration of a loop using continue
8.7 Nested loops
8.7.1 Nested loop
8.7.2 breaking out of and continuing with nested loops
8.8 Comparison of loop constructs
8.9 Exercise
Chapter 9 Creating and Using Arrays

9.1 Declare, instantiate, initialize and use a one-dimensional array
9.1.1 Declaring array variables
9.1.2 Creating and initializing array objects
9.2 Using arrays
9.2.1 Array indexing
9.2.2 Members of an array object
9.2.3 Runtime behavior of arrays
9.2.4 Uses of arrays
9.3 Declare, instantiate, initialize and use multi-dimensional arrays
9.3.1 Multidimensional Arrays
9.3.2 Assigning arrays of primitives to Object variables
9.4 Exercise
Chapter 10 Creating and Using Methods

10.1 Create methods with arguments and return values
10.1.1 Creating a method
10.1.2 Returning a value from a method
10.1.3 Varargs
10.2 Create overloaded methods
10.2.1 Method signature
10.2.2 Method overloading
10.2.3 Method selection
10.3 Passing object references and primitive values into methods
10.3.1 Passing arguments to methods
10.3.2 Passing objects to methods
10.3.3 Returning a value from a method
10.4 Create and overload constructors
10.4.1 Creating instance initializers
10.4.2 Creating constructors
10.4.3 Overloading constructors
10.4.4 Instance initializers vs constructors
10.4.5 final variables revisited
10.5 Apply the static keyword to methods and fields
10.5.1 Apply the static keyword to methods and fields
10.5.2 Accessing static members
10.5.3 Accessing static members from the same class
10.5.4 Importing static fields
10.5.5 Accessing instance members from a static method
10.5.6 Class loading and static initializers
10.6 Exercise
Chapter 11 Encapsulation
11.1 Apply access modifiers
11.1.1 Accessibility
11.1.2 Access modifiers
11.1.3 Understanding protected access
11.1.4 Applying access modifiers to types
11.2 Apply encapsulation principles to a class
11.2.1 Encapsulation
11.2.2 JavaBeans naming convention
11.3 Exercise
Chapter 12 Reusing Implementations Through Inheritance

12.1 Create and use subclasses and superclasses
12.1.1 Understanding Inheritance
12.1.2 Inheriting features from a class
12.1.3 Inheritance and access modifiers
12.1.4 Inheritance of instance members vs static members
12.1.5 Benefits of inheritance
12.2 Using super and this to access objects and constructors
12.2.1 Object initialization revisited
12.2.2 Initializing super class using "super"
12.2.3 Using the implicit variable "super"
12.2.4 Order of initialization summarized
12.3 Create and extend abstract classes
12.3.1 Using abstract classes and abstract methods
12.3.2 Using final classes and final methods
12.3.3 Valid combinations of access modifiers, abstract, final, and static
12.4 Enable polymorphism by overriding methods
12.4.1 What is polymorphism
12.4.2 Overriding methods
12.4.3 Invalid overrides
12.5 Utilize polymorphism to cast and call methods
12.5.1 Type of reference and type of an object
12.5.2 Bridging the gap between compile time and run time
12.5.3 When is casting necessary
12.5.4 The instanceof operator
12.5.5 Invoking overridden methods
12.5.6 Impact of polymorphism on == and equals method
12.6 Overriding and Hiding
12.7 Exercise
Chapter 13 Programming Abstractly Through Interfaces

13.1 Create and implement interfaces
13.1.1 Using interfaces
13.1.2 Implementing an interface
13.1.3 Extending an interface
13.1.4 Instantiating abstract classes and interfaces
13.2 Difference between Interface and Abstract Class
13.3 Declare and use List and ArrayList instances
13.3.1 Introduction to Collections and Generics
13.3.2 Generics
13.3.3 Quiz
13.3.4 Collection and List API
13.3.5 ArrayList API
13.3.6 ArrayList vs array
13.3.7 Map and HashMap
13.3.8 Quiz
13.4 Exercise
Chapter 14 Lambda Expressions

14.1 Understanding Lambda Expressions
14.1.1 Lambda Expressions
14.1.2 Parts of a Lambda expression
14.1.3 Using Predicate
14.1.4 Functional Interfaces
14.1.5 Using Functional Interfaces with Collections API
14.1.6 Scope of variables in a lambda expression
14.1.7 Quiz
14.2 Exercise
Chapter 15 Handling Exceptions

15.1 Create try-catch blocks and determine how exceptions alter program flow
15.1.1 Java exception handling
15.1.2 Fundamentals of the try/catch approach
15.1.3 Pieces of the exception handling puzzle
15.2 Differentiate among checked, unchecked exceptions, and Errors
15.2.1 Checked and Unchecked exceptions
15.2.2 Commonly used exception classes
15.3 Create and invoke a method that throws an exception
15.3.1 Creating a method that throws an exception
15.3.2 Throwing exceptions from initializers and constructors
15.3.3 Invoking a method that throws an exception
15.3.4 Using multiple catch blocks
15.4 Exercise
Chapter 16 Understanding Modules

16.1 Module Basics
16.1.1 What are modules?
16.1.2 Declaring a module
16.1.3 Directory structure of a module
16.2 Describe how a modular project is compiled and run
16.2.1 Compiling a module
16.2.2 Running a module
16.3 Declare modules and enable access between modules
16.3.1 Enabling access between modules
16.3.2 Qualified exports
16.3.3 Transitive dependencies
16.4 Advanced module compilation and execution
16.4.1 Compiling multiple modules at once
16.4.2 Module jar vs regular jar
16.4.3 Summary of command line switches used for compilation
16.4.4 Summary of command line switches used for execution
16.5 Describe the modular JDK
16.5.1 Modular JDK
16.5.2 Organization of the modular JDK
16.5.3 Benefits of the modular JDK
16.6 Exercise
Chapter 17 Understanding Java Technology and environment

17.1 Java Technology and key features of the Java language
Introduction
I believe you have already got your feet wet with Java programming and are now
getting serious about your goal of being a professional Java programmer. First of
all, let me commend your decision to consider Java certification as a step
towards achieving that goal. I can assure you that working towards acquiring
Oracle's Java Certification will be very rewarding. Irrespective of whether you
get extra credit for being certified in your job hunt or not, you will be able to
speak with confidence in technical interviews and the concepts that this
certification will make you learn, will improve your performance on the job.
The Java SE 11 Programmer I exam (Exam code 1Z0-815 ), aka OCP JP-I exam,
is the first of the two exams that you need to pass in order to become an Oracle
Certified Professional - Java SE 11 Developer. This exam focuses on the
fundamental aspects of Java and is not particularly tough to pass. If you go
through a decent book and practice a few good mock exams, you should be able
to pass it with a couple of months of preparation. However, the topics covered in
this certification form the groundwork for the second step of professional
certification, i.e., the Java SE 11 Programmer II exam (Exam code 1Z0-816 ),
aka OCP JP-II exam. The OCPJP-II is a very tough exam. It is a lot tougher than
the OCPJP-I exam. You will have trouble passing that exam if your
fundamentals are weak. For this reason, it is very important to not think of just
passing the OCPJP-I exam with the bare minimum marks required (63%) but to
set a score of 90% as your target. My objective with this book is to help you
achieve 90% plus score on the OCPJP-1 exam.
About the mock exams
Mock exams are an essential preparation tool for achieving a good score on the
exam. However, having created mock exams for several certifications, I can tell
you that creating good quality questions is neither easy nor quick. Even after
multiple reviews and quality checks, it takes years of use by thousands of users
for the questions to shed all ambiguity, errors, and mistakes. I have seen users
come up with plausible interpretations of a problem statement that we could
never imagine. A bad quality mock exam will easily eat up your valuable time
and may also shake your confidence. For this reason, I have not created new
mock exams for this book. We have a team that specializes in developing mock
exams and I will recommend you to buy the exam simulator created by this team
from our website Enthuware.com. It is priced quite reasonably (only 9.99 USD)
and has stood the test of time.
0.1 Who is this book for? ☝
This book is for Java SE 11 Programmer - I exam (1Z0-815) takers who know
how to program and are at least aware of the basic Java terminology. Before
proceeding with this study guide, please answer the following questions.
Remember that you don't have to be an expert in the topic to answer yes. The
intention here is to check if you are at least familiar with the basic concepts. It is
okay if you don't know the details, the syntax, or the typical usage. I will go
through all that in this book, but I will not teach the basics of programming in
this book.
1. Do you know what OS, RAM, and CPU are?

2. Do you know what a command line is?
3. Do you know basic OS commands such as dir , cd , and mkdir (or if you
are a Linux/Mac user - Do you know how to use ls , cd , and md )?
4. Can you write a simple Hello World program in Java and run it from
the command line?
5. Do you know what variables are?
6. Do you know what loops (such as for loop and while loop) are and what
they are used for?
7. Are you aware of arrays?
8. Are you aware that Java has classes and interfaces?
9. Are you aware that classes and interfaces have methods?
10. Have you installed JDK 11 on your computer?
If you answered no to any of the above, this book is not for you. It would be
better if you go through a programming book or a computer book for beginners
first, and then come back to this book. Alternatively, be open to google a term if
you are not sure about it at any time before proceeding further while reading this
book.
0.2 How is this book different from others? ☝
With so many certification books around, I think this question is worth
answering at the outset. This book is fundamentally different from others in the
following respects:
1. Focus on concepts - I believe that if you get your basic concepts right,
everything else falls in place nicely. While working with Java beginners, I
noticed several misconceptions, misunderstandings, and bad short cuts that
would affect their learning of complex topics later. I have seen so many
people who manage to pass the exam but fail in technical interviews
because of this reason. In this book, I explain the important stuff from
different perspectives. This does increase the length of the book a bit but
the increase should be well worth your time.
2. No surgical cuts - Some books try to stick very close to the exam
objectives. So close that sometimes a topic remains nowhere close to reality
and the reader is left with imprecise and, at times, incorrect knowledge.
This strategy is good for answering multiple choice questions appearing on
the OCPJP-I exam but it bites the reader during technical interviews and
while writing code on the job. I believe that answering multiple choice
questions (MCQs) should not be your sole objective. Learning the concepts
correctly is equally important. For this reason, I go beyond the scope of
exam objectives as, and when, required. Of course, I mention it clearly
while doing so.
3. Exercises - "Write a lot of code" is advice that you will hear a lot. While it
seems quite an easy task for experienced programmers, I have observed that
beginners are often clueless about what exactly they should be writing.
When they are not sure about what exactly a test program should do, they
skip this important learning step altogether. In my training sessions, I give
code writing exercises with clear objectives. I have done the same in this
book. Instead of presenting MCQs or quizzes at the end of a topic or
chapter, I ask you to write code that uses the concepts taught in that topic or
chapter.
Besides, a question in the real exam generally requires knowledge of
multiple topics. The following is a typical code snippet appearing in the
exam:
int i = 10;
Long n = 20;
float f = 10.0;
String s = (String) i+n++;
To determine whether this code compiles or not, you need to learn four
topics - wrappers, operators, String class, and casting. Thus, presenting an
MCQ at the end of a topic, that focuses only on that one topic, creates a
false sense of confidence. I believe it is better to focus on realistic MCQs at
the end of your preparation.
4. Not being pedantic - If you are preparing for the OCPJP-I exam, I believe
you have already been through many academic exams in your life. You
already know what to expect in an exam. So, I won't advise you on the
amount of water you should drink before the exam to avoid a restroom
break, or on how much sleep you should get before the exam, or to check
the exam center location a day before. If you have not taken any computer-
based exam containing multiple choice questions, I strongly suggest you
use Enthuware's exam simulator to get familiar with this style. It closely
mimics the user interface of the real exam.
0.3 How is this book organized? ☝
This book consists of seventeen chapters plus this introduction at the beginning.
Other than the first chapter "Kickstarter for the Beginners", the chapters
correspond directly to the official exam objectives. The sections of a chapter also
correspond directly to the items of exam objectives in most cases. Each chapter
lists the exam objectives covered in that chapter at the beginning and includes a
set of coding exercises at the end. It would be best to read the book sequentially
because each chapter incrementally builds on the concepts discussed in the
previous chapters. I have included simple coding exercises throughout the book.
Try to do them. You will learn and remember the concept better when you
actually type the code instead of just reading it. If you have already had a few
years of Java development experience, you may go through the chapters in any
order.
Conventions used in this book
This book uses certain typographic styles in order to help you quickly identify
important information. These styles are as follows:
Code font - This font is used to differentiate between regular text and Java
code appearing within regular text. Code snippets containing multiple lines of
code are formatted as Java code within rectangular blocks.
Red code font - This font is used to show code that doesn't compile. It
could be because of incorrect syntax or some other error.
Output code font - This font is used to show the output generated by a
piece of code on the command line.
Bold font - I have highlighted important words, terms, and phrases using bold
font to help you mentally bookmark them. If you are cruising through the book,
the words in bold will keep you oriented besides making sure you are not
missing anything important.
Note -
Things that are not completely related to the topic at
hand are explained in notes. I have used notes to
provide additional information that you may find useful
on the job or for technical interviews but will most
likely not be required for the exam.
Exam Tip:
Exam Tips contain points that you should pay special attention to
during the exam. I have also used them to make you aware of the
tricks and traps that you will encounter in the exam.
Asking for clarification
If you need any clarification, have any doubt about any topic, or want to report
an error, feel free to ask on our dedicated forum for this book -
http://enthuware.com/forum. If you are reading this book on an electronic
device, you will see this icon beside every topic title. Clicking on this icon
will take you to an existing discussion on that particular topic in the same forum.
If the existing discussion addresses your question, great! You will have saved
time and effort. If it doesn't, post your question with the topic title in the subject
line. We use the same mechanism for addressing concerns about our mock exam
questions and have received tremendous appreciation from the users about this
feature.
0.4 Will this be asked in the exam? ☝

While going through this book, you will be tempted to ask this question many
times. Let me answer this question at the beginning itself. I do talk about
concepts in this book that are not explicitly listed in the official exam objectives
but wherever I deviate from the official exam objectives, I clearly specify so.
You are free to ignore that section and move on. But I suggest you do not skip
such sections because of the following reasons.
1. While discussing a rule of the language, I may have to refer to some terms
and concepts for the sake of completeness and technical accuracy. For
example, let's say we are talking about public classes in a file. If I state that
you cannot have more than one public class in a file, it is fine for the
purpose of the exam but it is technically incorrect because you can have any
number of public nested classes in a file. Thus, it would be better to state
that you cannot have more than one top-level class in a file. How about one
public top-level class and one public interface? Nope, you can't do that
either. Thus, the statement is still incorrect. The correct statement would be
that you cannot have more than one public top-level reference type in a file.
As you can see, it is imperative for me to mention the meaning of the terms
reference type, nested class, and top level class, even though you won't be
tested on them in the exam.
If you absolutely do not want to spend any time learning about anything
that is not part of the exam, then this book is not for you. I have tried to
stick to the objectives as much as possible but, if I believe you need to
know something, I talk about it even if it is beyond the scope of the exam.
2. I have noticed that many of the certification aspirants are new Java
programmers who are either in school or want to start their career with Java
programming. They want to get certified because they ultimately want to
land a job as a Java programmer. These programmers will be facing a lot of
technical interviews as well. I want these programmers to do well on
technical interviews.
Certification may get you a foot in the door but you will need to back it up
with strong knowledge of fundamentals in the interview. Therefore, if I
believe that something is important for you to know or that something will
be helpful to you in your technical interview, irrespective of whether it will
be asked in the exam or not, I discuss it.
3. Official exam objectives are neither detailed nor exhaustive. They list top
level topics that you need to study but leave out finer details. You will be
asked questions that require you to know those concepts.
4. Oracle adds new questions to the exam before formally adding a new topic
in the official exam objectives. These questions may not be included in your
final score, i.e., your answers on such questions are not counted towards
your score on the exam. However, test takers do not know if a question is
unscored and so, they must attempt it as if it will be counted towards their
final score.
Since we, at Enthuware, conduct classroom training as well, we get to

interact with a lot of test takers. We receive feedback from test takers about
getting questions on topics that are not there in the exam objectives. After
receiving such multiple reports, we may decide to add that topic to our
content. We clearly specify the reason for their inclusion.
5. Official exam objectives are not constant. Although not frequently, Oracle
does add and remove topics from the objectives from time to time. This
may render some of the content not relevant for the exam. I will update the
content as soon as possible.
If you are interested in getting your basics right, then I suggest you do not worry
too much about the exam objectives while following this book. Even if you
spend a little more time (not more than 10%, I promise) in your preparation
because of this extra content, it will be worth your while.
0.5 General tips for the exam ☝

Here is a list of things that you should keep in mind while preparing for the
exam -
1. Code Formatting - You may not find nicely formatted code in the exam.
For example, you may expect a piece of code nicely formatted like this:
if(flag){
while(b<10){
}
}else if(a>10) {
invokeM(a);
}
else{
System.out.println(10);
}
But you may get the same code formatted like this:
if(flag){
while(b<10){ }
} else
if(a>10) { invokeM(a); }
else { System.out.println(10); }
They do this most likely to save space. But it may also happen inadvertently
due to variations in display screen size and resolution.
2. Assumptions - Several questions give you partial code listings, aka "code
snippets". For example, what will the following code print?
ArrayList al = new ArrayList();

al.remove(0);
System.out.println(al);
Obviously, the code will not compile as given because it is just a code
fragment. You have to assume that this code appears in a valid context such
as within a method of a class. You also need to assume that appropriate
import statements are in place.
You should not fret over the missing stuff. Just focus on the code that is
given and assume that everything else is irrelevant and is not required to
arrive at the answer.
3. Tricky Code - You will see really weird looking code in the exam. Code
that you may never even see in real life. You will feel as if the exam is
about puzzles rather than Java programming. To some extent, that is
correct. If you have decided to go through the certification, there is no point
in questioning the relevance. If you feel frustrated, I understand. Please feel
free to vent out your anger on our forum and get back to work!
4. Number of correct options - Every question in the exam will tell you
exactly how many options you have to select to answer that question
correctly. Remember that there is no negative marking. In other words,
marks will not be deducted for answering a question incorrectly. Therefore,
do not leave a question unanswered. If you don't know the answer, select
the required number of options anyway. There is a slight chance that you
will have picked the correct answer.
5. Eliminate wrong options - Even better than not leaving a question
unanswered is make intelligent guesses by eliminating obviously incorrect
options. You may see options that are contradictory to each other. This
makes it a bit easy to narrow down the correct options.
That's about it. Hope this book helps you become a better Java programmer
besides getting you the certification.
0.6 Official Exam Details and Exam Objectives ☝

The following are the official exam details published by Oracle as of 1st July
2019. As mentioned before, Oracle may change these details at any time. They
have done it in the past. Several times. Therefore, it would be a good idea to
check the official exam page at https://education.oracle.com/java-se-11-
programmer-i/pexam_1Z0-815 during your preparation.
Exam Details
Duration : 180 Minutes
Number of Questions: 80
Passing Score: 63%
Format: Multiple Choice
Exam Price: USD 245 (varies by country)
Exam Objectives
Understanding Java Technology and environment
Describe Java Technology and the Java development environment

Working With Java Primitive Data Types and String APIs

data types)
Working with Java Arrays

Declare, instantiate, initialize and use two-dimensional array
Creating and Using Methods

Reusing Implementations Through Inheritance

Handling Exceptions

checked exceptions, unchecked exceptions, and Errors
Create a try-catch block and determine how exceptions alter normal
program flow
Creating a Simple Java Program

Using Operators and Decision Constructs

precedence
Use Java control statements including if, else, break and continue
Create and use do/while loops, while loop, and for looping statements
including nested loops
Describing Objects and Classes
Declare and instantiate Java objects, and explain objects' lifecycles

(including creation, dereferencing by reassignment, and garbage collection)
Applying Encapsulation

Programming Abstractly Through Interfaces

classes
Understanding lambda Expressions
Understanding Modules

0.7 Feedback and Reviews ☝

This is the first draft of the book and we are currently reviewing the content and
the code snippets for technical correctness. If you have any query regarding the
contents of this book or if you find any error, please do let me know on
Enthuware Forum .
I hope you enjoy reading this book. If you learn a few things and find it
interesting, I would be very grateful if you would consider leaving a review on
Amazon or Google Play with a few kind words. If you have received a review
copy of this book, please mention so, in your review.
thank you,
Hanumant Deshmukh
Chapter 1 Kickstarter for Beginners
This section is for Java beginners. It does not directly relate to any exam
objective but is meant to provide a solid grounding that will help you to easily
understand the concepts taught in later chapters. The concepts covered in this
section are important because they kind of repeat over and over throughout this
course. If we get these repetitions over with now, you will be happier later on!
1.1 Key points in OOP

1.1.1 A matter of perspective ☝
A couple of years ago, while I was visiting India, I had a tough time plugging in
my laptop charger in the 3-pin sockets. Even the international socket adapter kit,
which had adapter pins of various sizes, was not of much help. Sometimes the
receiver would be a bit too small and the pins wouldn't make steady contact or
the pins would be a bit too wide and won't go into the receiver. I had to finally
cut my cord and stick the bare copper wire ends directly into the sockets. I
wondered, why do all these sockets in the same country have slight differences.
During my stay, I observed that such minor variations were present in other
things as well. Doors that wouldn't completely close, nuts that wouldn't turn
properly, taps that wouldn't stop leaking, and other differences. Most of the time,
people there take the trial and error approach when replacing parts. They work
with the expectation that even if they get a part with the right size, it still may
not fit perfectly. In other words, minor variations are expected and well tolerated.
This was unimaginable to me in the US, where everything just fits. I could buy a
bolt from one shop and a nut from another, and it would work perfectly.
Everything, from screws, nuts, and bolts, to wood panels, electrical parts,
packing boxes, is standardized. One can easily replace a part with another built
by a totally different company. You just have to specify the right "size".
This experience led me to a potential cause of why some OOP learners find OOP
concepts confusing. Especially, beginners from non-western background find it
really tough to grasp the fundamental concepts because they do not know the
rationale behind so many rules of OOP. This is reflected in their application
design.
In the US, and I imagine in other developed countries as well, things are
extremely well defined. Products clearly specify how they should be used and in
what cases they will fail. People can and do rely on these specifications because
products work as expected and fail as defined. At the same time, people expect
products to come with detailed specifications. Ready to assemble furniture is a
prime example of how detailed these specifications can be. It's because of
detailed and clear specifications that people feel comfortable in buying
complicated ready-to-assemble furniture.
In short, people know exactly what they are getting when they acquire
something. I think of it as the society being naturally "object-oriented".
Object orientation is just a name for the same natural sense of things fitting
nicely with each other. A piece of code is not much different from the physical
things I mentioned earlier. If you code it to a specification, it will fit just as
nicely as your .16 inch nut on a .16 inch bolt, irrespective of who manufactured
the nut and who manufactured the bolt. The point here is that the source of the
concept of object-oriented programming is the physical world. If you want to
grasp OOP really well, you have to start thinking of your piece of code as a
physical component...a "thing" that has a well defined behavior and that can be
replaced with another component that adheres to the same behavior. You would
not be happy if you bought a tire that doesn't fit on your car's wheel even though
you bought same 'size', would you? You should think about the same issues
when you develop your software component. Someone somewhere is going to
use it and they won't be happy if it fails at run time with an exception that you
didn't say it would throw in a particular situation.
1.1.2 API ☝
You probably know that API stands for Application Programming Interface .
But do you understand what it really means? This goes back to my previous
observation about relating programming concepts to real life. When you operate
a switch do you really care about what exists inside the switch? Do you really
care how it works? You just connect the switch to a light bulb and press it to
switch the bulb on or off. It is the same with a car. A car provides you with a few
controls that allow you to turn, accelerate, and brake. These controls are all you
need to know how to drive a car.
You should think about developing software components in the same way. A
software component doesn't necessarily mean a bunch of classes and packages
bundled together in jar file. A software component could be as simple as a single
class with just one method. But while developing even the smallest and the
simplest of software components, you should think about how you expect the
users to use it. You should think about various ways a user can use the
component. You should also think about how the component should behave
when a user doesn't use it the way it is expected to be used. In other words, you
should specify an interface to your component, even before you start coding for
it. Here, by interface, I do not mean it in the strict sense of a Java interface but a
specification that details how to interact with your component. In a physical
world, the user's manual of any appliance is basically its interface. In the
software world, the specification of the publicly usable classes, methods, fields,
enums, et cetera of a software component is its interface. As an application
programmer , if you want to use a component developed by someone else, you
need to worry only about the interface of that component. You don't need to
worry about what else it might contain and how it works. Hence the phrase
'Application Programming Interface'.
In the Java world, a collection of classes supplied by a provider for a particular

purpose is called a library and the JavaDoc documentation for those classes
describes its API. When you install the Java Runtime Environment (JRE), it
includes a huge number of classes and packages. These are collectively called
the standard Java library and the collection of its public classes and interfaces is
called the Java API .
The Java API contains a huge amount of ready-made components for doing
basic programming activities such as file manipulation, data structures,
networking, dates, and formatting. When you write a Java program, you actually
build upon the Java API to achieve your goal. Instead of writing all the logic of
your application from scratch, you make use of the functionality already
provided to you, free of cost, by the Java library and only write code that is
specific to your needs. This saves a lot of time and effort. Therefore, a basic
understanding of the Java API is very important for a Java programmer. You
don't need to know by heart all the classes and their methods. It is practically
impossible to know them all, to be honest. But you should have a broad idea
about what the Java API provides at a high level so that when the need arises,
you know where to look for the details. For example, you should know that the
standard Java library contains a lot of classes for manipulating files. Now, when
you actually need to manipulate a file, you should be able to go through the
relevant Java packages and find a Java class that can help you do what you want
to do.
The OCPJP 11 Part 1 exam requires that you know about only a few packages
and classes of the Java API. I will go through them in detail in this book.
If you keep the above discussion in mind, I believe it will be very easy for you to
grasp the concepts that I am going to talk about throughout this book.
1.1.3 Type, class, enum, and interface ☝
A type is nothing but a name given to a behavior. For example, when you define
how a bank account behaves when you interact with it, you are defining a type
and if you give this behavior a name, say Account, then you have essentially
defined the Account type.
From this perspective, a class , an enum , and an interface help you define
certain kinds of behaviors and are thus, types of types.
A class allows you to combine the description of a behavior and the

implementation that is used to realize this behavior. The implementation
includes logic as well data. For example, an account allows you to withdraw and
deposit money, which is the description of its behavior, and uses "account
balance", which is the data that it manipulates to realize this behavior. Thus,
Account could be a class. Once you define the behavior of an account and also
provide the implementation to realize this behavior, you can have as many
accounts as you want.
An enum , which is a short form for enumeration, also allows you to combine
the description of a behavior and the implementation that is used to realize this
behavior. However, in addition, it provides a fixed number of instances of this
type and restricts you from creating any new instances of this type. For example,
there are only 7 days in a week (from Monday to Sunday). Thus, if you define
DayOfWeek, you wouldn't want to create a day other than those predefined 7
days. Thus, DayOfWeek could be an enum with 7 predefined unchangeable
instances.
An interface allows you to define just the behavior without any implementation
for it. For example, if you want to describe something that moves, you can call it
Movable. It doesn't tell you anything about how that entity moves. A cow, a car,
or even a stock price all move, but obviously, they move very differently. Thus,
Movable could be an interface.
The key point about an interface is that you cannot have an instance of an
interface because it is just a description of the way you can interact with
something and is not the description of the thing itself. For example, you cannot
really have just a Movable. You must have a cow or a car or something else that
exhibits the behavior described by Movable. In that sense, an interface is always
abstract . It cannot exist on its own. You need a class to implement the
behavior described by an interface.
Besides the above three, there is something called abstract class . An abstract
class lies somewhere in between a class and an interface. Just like a class, it
defines behavior as well as implementation but the implementation that it
provides is not complete enough for you to create instances of it. Therefore, just
like an interface, it cannot exist on its own. For example, if you define the
behavior that is common to animals along with some implementation that is
common to all animals in a class. But you know that you can't really have just an
Animal. You can have a cat, or a dog, or a cow, which are all animals, but not
just an animal.
1.2 Why is something so

1.2.1 Why is something so? ☝
Why does Java not have pointers ? Why does Java permit static fields and
methods? Why does Java not have multiple inheritance ? Why does this work
but that doesn't? While learning Java, curious minds get such questions very
often. Throughout the book, you will come across rules and conventions that will
trigger such questions. Most of the times the reason is not too complicated. I will
explain four points below that will help you answer most of such questions. I
will also refer to them throughout the book wherever warranted.
1. To help componentize the code - As discussed earlier in the API section,

while writing Java code, you should think about developing components
instead of writing just programs . The difference between the two is in the
way they allow themselves to be used interchangeably. Can you imagine a 3
pin socket that has the ground pin on the left instead of on the top? No one
makes such a thing because it won't allow any other plug to be plugged in.
In that sense, Components are like generic Lego blocks. You can mix and
match the blocks with basic functionality and build even bigger blocks. You
can take out one block and replace it with another block that has the same
connectors. It is the same with software components.
A well-developed software component is as generic as possible. It is built

to do one thing and allows other components to make use of it without
making them dependent on it. Dependency here means that you should be
able to easily replace this component with another component that does the
same thing. Indeed, you should be able to replace a 3 pin socket from one
manufacturer with another without needing to replace the entire appliance!
A program, on the other hand, is a monolithic pile of code that tries to do

everything without exposing generic and clear interfaces. Once you start
using a program, it is almost impossible to replace it with another one
without impacting all other pieces that work with that program. It is very
much like a proprietary connector that connects a device to a computer. You
have to buy the whole new PC card to support that connector. If the
connector wire goes missing, you are dependent on the maker of that
proprietary connector to provide you with a replacement, at which point,
you will wish that you had bought a device with a USB connector instead.
Only a few companies can pull this stunt off on their customers.
Java is designed with this in mind. You will see that many seemingly
confusing rules are there precisely because they promote the development
of interchangeable components. For example, an overriding method cannot
throw a more generic exception that the one declared by the overridden
method. On the other hand the constructor of a subclass cannot throw only a
more specific exception than the one thrown by the constructor of the
superclass. Think about that.
2. To eliminate the scope for bugs - Java designers have tried to limit or
eliminate features that increase the possibility of bugs in a piece of code.
Pointer arithmetic and goto are examples of that. They have also tried to
add features that help writing bug-free code. Automatic Garbage
Collection and Generics are examples of that.
3. Make life easier for the programmer - Many older languages such as
C/C++ were built with the flexibility and power to do various kinds of
things. Putting restrictions on what a programmer can do was thought of as
a bad idea. On the contrary, how to add features that will let the
programmer do more and more was the focus. Every new language added
more new features. For example, C++ has pointer arithmetic, global
functions, operator overloading, extern declarations, preprocessor
directives, unsigned data types, and so many other "features" that Java
simply does not have. These are some really powerful tools in the hands of
a C++ programmer. So, why doesn't Java have them? Java has actually
gone in reverse with respect to features. Java does not have a lot of features
that are found in languages that came before Java. The reason is simple.
Java follows the philosophy of making life simpler for the programmer .
Having more and more features is not necessarily a good thing. For
example, having pointer arithmetic and manual allocation and deallocation
of objects is powerful but it makes life hell for the programmer. Thus,
unlike C++, there is no need to allocate memory in Java because all objects
are created on the heap. Why should a programmer have to worry about
something that can be taken care of by the language? Instead of focusing on
mastering complicated features, the programmers should be spending more
time in developing business logic. Thus, unlike C++, there is no need to
deallocate memory in Java because Java performs garbage collection
automatically.
Furthermore, the cost of maintaining complicated code cascades very
quickly. A piece of code is written once but is read and is overwritten
numerous times. What is "clever" for one programmer becomes a nightmare
for the one who follows that programmer.
Java has therefore, introduced several restrictions (I consider them features,

actually) that make Java development substantially simpler overall. For
example, in C++, there is no restriction on the file name in which a class
exists. But in Java, a public class has to be in a file by that name. This is a
restriction that doesn't seem to make sense at first because after compilation
all classes are in the class files with the same names as that of the classes.
But when you think about the organization of your source code and the ease
of locating the code for a class, it makes perfect sense. Forcing every class
to be coded in its own independent file would have been impractical and
letting any class to be in a file by any name would have been too chaotic.
So, forcing a public class to be in a file by that name is a nice balance.
4. To become commercially successful - "If Java is an Object-Oriented

language, then why does it allow XYZ?" I see this question asked so many
times. The answer is simple. Java was designed by pragmatic folks rather
than idealistic ones. Java was designed at a time when C/C++ was
extremely popular. Java designers wanted to create a language that
remained faithful to OOP as much as possible but at the same time did not
alienate the huge community of C/C++ programmers. This community was
seen as potential Java developers and several compromises were made to
make it easier for them to program in Java. Furthermore, not all non-OOP
features are completely useless. Features that add value in certain
commonly occurring situations find a place in Java even if they are not
strictly OOP. Static methods is one such feature.
Then there is a matter of a "judgement call". Java designers are a bunch of

smart people. Some things may not make complete sense from a purely
logical or technical perspective but that's how they designed those things
anyway. They made the decisions based on their experience and wisdom.
For example, it is technically possible to design a compiler that can figure
out the value of a non-final local variable with 100% certainty in the
following code but the Java compiler does not flag an error for
"unreachable code" here:
int x = 0;
if(x==0){
throw new Exception();
}
x = 20;//unreachable code here but no compilation
error
Sometimes there is a logical explanation for a seemingly confusing rule but

the reason is not very well known. For example, the following code
compiles fine even though the compiler knows that the code is unreachable:
class ConditionalCompilation {
public static final boolean DEBUG = false;
public void method(){
if(DEBUG){
System.out.println("debug statement here")
;
} //works
}
}
But a similar code causes the compiler to flag "unreachable code" error:
class ConditionalCompilation{
public static final boolean DEBUG = false;
public void method(){
while(DEBUG){ //doesn't work
System.out.println("debug statement her

e");
}
}
}
The reason is that historically, developers have used the combination of a
boolean variable and an 'if' statement to include or exclude debug
statements from the compiled code. A developer has to change the value of
the flag at just one place to eliminate all debug statements. The 'if'
statement in the code above works because Java designers decided to
permitted this type of unreachable code so that conditional compilation
could occur.
In conclusion, if you ever find yourself in a situation where you have to explain
the reason behind a weird Java rule or concept, one of the above four would be
your best bet. For example, reason 3 answers the question that you asked in the
previous section, "why does Java allow fields and methods to be defined in an
interface?". Why doesn't Java allow multiple inheritance? Reason 3. Why are all
objects in Java rooted under Object class?. Reason 3.
1.3 Declaration and Definition

1.3.1 Declaration and Definition ☝
In a technical interview, you should always know what you are talking about. A
smart interviewer will catch you in no time if you talk loose. If you answer
imprecisely, your credibility will evaporate faster than water in a frying pan. The
certification exam requires the same attitude. You will lose marks for not
knowing the basics.
It is surprising how many people use the terms declaration and definition
incorrectly. So, let's just get this straight from the get-go. A declaration just
means that something exists. A definition describes exactly what it is. For
example,
class SomeClass //class declaration
//class definition starts

{
public void m1()//method declaration
//method definition starts
{
}
//method definition ends
}
//class definition ends
As you can see, a declaration provides only partial information. You can't make
use of partial information. The definition makes it complete and thus, usable.
In terms of variables, Java doesn't have a distinction between declaration and

definition because all the information required to define a variable is included in
the declaration itself. For example,
int i; //this declaration cum definition is complete in

itself
However, Java does make a distinction between variable declaration and variable
initialization. Initialization gives a value to a variable. For example, int i =
10; Here i is defined as an int and also initialized to 10. Object obj =
null; Here obj is defined as an Object and is also initialized to null . I
will discuss more about declaration and initialization later.
The above is a general idea but you should be aware that there are multiple
viewpoints with minor differences. Here are some links that elaborate more. You
should go through at least the first link below.
StackOverflow - Difference between declaration and definition in Java

JavaRanch - Declaration and Definition
Can you now answer the question what does an interface contain - method
declarations or method definitions?
Well, there was a time when interfaces contained only method declarations, but
since Java 8, interfaces contain method declarations as well as definitions.
1.4 Object and Reference

1.4.1 Relation between a class, an object, and a reference ☝
A class is a template using which objects are created. In other words, an object
is an instance of a class. A class defines what the actual object will contain once
created. You can think of a class as a cookie cutter. Just as you create cookies out
of dough using a cookie cutter, you create objects out of memory space using a
class.
To access an object, you need to know exactly where that object resides in
memory. In other words, you need to know the "address" of an object. Once you
know the address, you can call methods or access fields of that object. It is this
"address" that is stored in a reference variable.
If you have trouble understanding this concept, try to imagine the relationship
between a Television (TV) and a Remote. The TV is the object and the Remote
is the reference variable pointing to that object. Just like you operate the TV
using the remote, you operate on an object using a reference pointing to that
object. Notice that I did not say, "you operate on an object using its reference".
That's because an object doesn't have any special reference associated with it.
Just as a TV can have multiple remotes, an object can have any number of
references pointing to it. One reference is as good as any other for the purpose of
accessing that object. There is no difference between two references pointing to
the same object except that they are two different references. In other words,
they are mutually interchangeable.
Now, think about what happens when the batteries of a remote die. Does that
mean the TV stops working? No, right? Does that mean the other remote stops
working? Of course not! Similarly, if you lose one reference to an object, the
object is still there and you can use another reference, if you have it, to access
that object.
What happens when you take one remote to another room for operating another
TV? Does it mean the other remote stops controlling the other TV? No, right?
Similarly, if you change one reference to point to some other object, that doesn't
change other references pointing to the that object.The following picture
illustrates the situation:
Let me now move to an example that is closer to the programming world. Let's
say, you have the following code:
String str = "hello";
"hello" is the actual object that resides somewhere in the program's memory.
Here, str is the remote and "hello" is the TV. You can use str to invoke
methods on the "hello" object.
A program's memory can be thought of as a long array of bytes starting with 0 to
NNNN , where NNNN is the location of last byte of the array. Let's say, within this
memory, the object "hello" resides at the memory location 2222 . Therefore,
the variable str actually contains just 2222 . It doesn't contain "hello" . It is
no different from an int variable that contains 2222 in that sense.
But there is a fundamental difference in the way Java treats reference variables
and non-reference variables (aka primitive variables ).If you print an int
variable containing 2222 , you will see 2222 printed. However, if you try to
print the value str , you won't see 2222 . You will see "hello" . This is
because the JVM knows that str is defined as a reference variable and it needs
to use the value contained in this variable to go to the memory location and do
whatever you want to do with the object present at that location. In case of an
int (or any other primitive variable), the JVM just uses the value contained in
the variable as it is. Since this is an important concept, let me give you another
example to visualize it. Let us say Paul has been given 2222 dollars and Robert
has been given bank locker number 2222. Observe that both Paul and Robert
have the same number but Paul's number denotes actual money in his hands
while Robert doesn't have actual money at all. Robert has an address of the
location that has money. Thus, Paul is like a primitive variable while Robert is
like a reference variable.
Another important point is that you cannot make a reference variable point to a
memory location directly. For example, you can set the int variable to 2250
but you can't do that to str i.e. you can't do str = 2250 . It will not compile.
You can set str to another string and if that new string resides at a memory
location 2250 , str will indeed contain 2250 but you can't just store the
address of any memory location yourself in any reference variable.
As a matter of fact, there is no way in Java to see and manipulate the exact value
contained in a reference variable. You can do that in C/C++ but not in Java
because Java designers decided not to allow messing with the memory directly.
You can have as many references to an object as you want. When you assign one
reference to another, you basically just copy the value contained in one reference
into another. For example, if you do String str2 = str; you are just
copying 2222 into str2 . Understand that you are not copying "hello" into
str2 . There is only one string containing "hello" but two reference
variables referring to it. Figure 1 illustrates this more clearly.
If you later do str = "goodbye"; you will just be changing str to point to
a different string object. It does not affect str2 . str2 will still point to the
string "hello" .
The question that should pop into your head now is what would a reference
variable contain if it is not pointing at any object? In Java, such a variable is said
to be null . After all, as discussed above, a reference variable is no different
from a primitive variable in terms of what it contains. Both contain a number.
Therefore, it is entirely possible that a reference that is not pointing to any object
may actually contain the value 0. However, it would be wrong to say so, because
a reference variable is interpreted differently by the JVM. A particular
implementation of JVM may even store a value of -1 in the reference variable if
it does not point to any object. For this reason, a reference variable that does not
point to any object is just null. At the same time, a primitive variable can never
be null because the JVM knows that a primitive variable can never refer to an
object. It contains a value that is to be interpreted as it is. Therefore,
String str = null; //Okay

int n = 0; //Okay
String str = 0; //will not compile

int n = null; //will not compile.
1.5 static and instance

1.5.1 Static and Instance ☝
You will read the word "static" a lot in Java tutorials or books. So, it is better to
form a clear understanding of this word as soon as possible. In English, the word
static means something that doesn't change or move. From that perspective, it is
a misnomer. Java has a different word for something that doesn't change: final . I
will talk more about "final" later.
In Java, static means something that belongs to a class instead of belonging to an

instance of that class. As we discussed in the "Object and Reference" section, a
class is just a template. You can instantiate a class as many times as you want
and every time you instantiate a class you create an instance of that class. Now,
recall our cookie cutter analogy here. If a class is the cookie cutter, the fields
defined in the class are its patterns. Each instance of that class is then the cookie
and each field will be imprinted on the cookie - except the fields defined as
static. In that sense, a static member is kind of a tag stuck to a cookie cutter. It
doesn't apply to the instances. It stays only with the class.
Consider the following code:
class Account {
String accountNumber;
static int numberOfAccounts;
}
...
//Create a new Account instance
Account acct1 = new Account();
//This Account instance has its own accountNumber fiel

d.
acct1.accountNumber = "A1";
//But the numberOfAccounts fields does not belong to t

he instance, it belongs to the Account class
Account.numberOfAccounts = Account.numberOfAccounts +
1;
//Create another Account instance
Account acct2 = new Account();
//This instance has its own accountNumber field.
acct2.accountNumber = "A2";
//the following line accesses the same class field and
therefore, numberOfAccounts is incremented to 2
Account.numberOfAccounts = Account.numberOfAccounts +
1;
Important points about static -
1. static is considered a non object-oriented feature because as you can see in

the above code, static fields do not belong to an object. So, why does Java
have it? Check out the "Why is something so?" section.
2. Here is a zinger from Java designers - even though static fields belong to a
class and should be accessed through the name of the class, for
example, Account.numberOfAccounts , it is not an error if you
access it through a variable of that class, i.e.,
acct1.numberOfAccounts . Accessing it this way doesn't change its
behavior. It is still static and belongs to the class. Therefore,
acct2.numberOfAccounts will also refer to the same field as
acct1.numberOfAccounts . This style only causes confusion and is
therefore, strongly discouraged. Don't write such code. Ideally, they should
have disallowed this usage with a compilation error.
3. Just like fields, methods can be static as well. A static method belongs to
the class and can be accessed either using the name of the class or through a
variable of that class.
4. The opposite of static is instance. There is no keyword by that name
though. If a class member is not defined as static, it is an instance member.
1.6 Stack and Heap

1.6.1 Stack and Heap ☝
When you execute a program, the Operating System (OS) allocates and gives
memory to that program. This memory is used by the program to keep its
variables and data. For example, whenever you create a variable, its value needs
to be preserved as long as the program wants to use it. The program uses its
allocated memory to keep it. A program may ask the OS for more memory if it
requires and the OS will oblige if the OS has free memory available. A program
may also release some memory that it does not want back to the OS. Once the
OS gives out a chunk of memory to the program, it is the responsibility of the
program to manage it. Once the program ends, this memory is released and goes
back to the OS. This is basically how any executable program works.
Now, think about the following situation. Your program has a method that prints
"hello" 100 times. Something like this:
public class Test{

private String str = new String("hello"); //Using ne
w is not a good way to create strings, but bear with m
e for a moment.
public void print(){

int i = 0;
while(i++<100){
System.out.println(this.str);
}
}
public static void main(String[] args){

Test t = new Test();
t.print();
}
}
In the above class, it is the main method that calls the print method but there
could also be another class, which could make use of the same print method to
print hello a 100 times. When the print method is called, it creates the variable
i to keep track of the number of times the while loop has iterated. This variable
needs to be kept somewhere as long as the print method runs. Similarly, it uses
the variable str to print the string that you want the print method to print.
The question is, what happens when the print method ends? The variable i has
served its purpose and is not required anymore. It is not used anywhere except
within this method. Therefore, it need not be kept longer than the execution of
the print method. But the variable str still can be used whenever the print
method is called. Therefore, the value of str needs to be kept irrespective of the
execution lifetime of the print method.
It should now be clear that a program needs two kinds of memory spaces to keep
the stuff. One for temporary stuff that can be cleaned up as soon as a method call
ends and one for permanent stuff that remains in use for longer than a single
method call. The space for storing the temporary stuff is called Stack space and
the space for storing all other stuff is called Heap space . The reason why they
are called Stack and Heap will be clear soon.
In Java, each thread is given a fixed amount of stack space. In the above
example, when you execute the program, a main thread is created with a fixed
amount of stack space. All this space is initially empty. When this thread invokes
the main method, all the temporary variables created by this method are kept on
this stack. In the above example, the main method gets one reference variable
named args and inside the method, it creates another reference variable named
t . (Note that since args and t are reference variables, they contain the address
of the location where actual objects referred to by args and t respectively
reside). Therefore, the stack fills up by the amount of space required by these
reference variables. Before the main method ends, it calls the print method on
the reference t . Since print is an instance method, a variable named "this" is
automatically put on the stack for it so that the method can access the instance
fields on this object. The variable this is also a reference variable and it
contains the address of the location where the Test object actually resides. The
print method creates one more temporary variable i . This variable is also kept
on the same stack on top of this . Thus, the stack fills up a little more by
amount of space required for storing two variables. When the print method ends,
the space used for this and i is reverted back to the stack and the stack is thus,
emptied out a little. The control goes back to the main method. This method also
ends and the space used for storing args and t is cleaned up. There is nothing
left on the stack anymore at this point. Thus, the stack is completely empty
again. The following five figures illustrate this process.

As you can observe, the stack space looks like a stack of chips that are kept one
on top of the other. The temporary variables created by a method are added on
top of the stack one by one as and when they are created. As soon as the method
ends, all those variables are removed from the top. Observe that they are
removed only after the method ends. If the method calls another method, then
the variables created by the called method are pushed on to the same stack on
top of the variables stored by the caller method. When a thread dies, its stack
space is reverted back to the JVM. Since this space behaves like a stack, it is
called stack space .
The heap space , on the other hand, is, well, like a heap! Objects lie in a heap
just as they please. JVM goes to great lengths to organize the heap space.
Organization of the heap space is an advanced topic and is very important when
you analyse the performance of an application. However, it is not relevant for the
exam and so, I will not discuss it. From the program perspective, there is not
much of an organization in a heap.
Whenever any object is created anywhere in the code (i.e. whether in a method
or in a class), the JVM allocates space for that object on the heap and puts its
contents in that space. In Java, a program never releases this space explicitly. It
is managed by the JVM. Again, recall that an object can only be accessed using
its reference. For a method to access an object, it must use a reference that points
to that object. It could get that reference either from a variable kept on its stack
space (if the object was created in this method itself) or through a reference to
another object whose reference is kept on the stack space (if that object has a
reference to the required object). In either case, a method has to start with a
reference that exists on its stack space. If there is no reference on any stack space
through which an object can be accessed directly or indirectly, that object is
considered garbage. It is cleaned up automatically by the JVM using a garbage
collector.
Recall from our discussion on References and Objects

that a reference is merely a variable that stores the
address of the location where the actual object is
stored. In that sense, a reference variable is no
different than an int variable. They both store a
number. A reference variable stores a number that
indicates the memory location where you can find the
actual object, while an int variable stores a number that
is interpreted as a number. It doesn't indicate anything
else. If you create a variable in a method, whether a
reference variable or a primitive variable, it is kept on
the stack but when you create an object, that object is
stored on the heap.
Typically, an object is created using the new keyword.

But Java treats Strings as special, and so, you can
create String objects even without the new keyword.
Thus, whether you do "hello" or new String("hello"), in
both the cases, a String object containing "hello" is
created on the heap.
Points to remember:
Local variables are always kept on the stack. Objects are always stored in
the heap. (An optimizing JVM may allocate an object on the stack space,
but it is an internal detail of the JVM and you need not worry about it. For
all we care, objects are always on the heap.)
JVM may have several threads. Each thread is given a fixed amount of
stack space that is dedicated completely and exclusively to that thread. No
one but that thread can access its stack space. This is called "stack
semantics ". A thread accesses its stack space by creating and using
variables. There is no other special way of accessing the stack space.
Heap space is shared among all threads. Any thread can use space on a heap
by creating objects. Since heap space is shared, it is possible for one thread
to access objects created by another if it has a reference to that object. This
is called "heap semantics ".
Stack space is limited for a program. So, if you have a huge chain of
method calls where each method creates a lot of temporary variables
(recursion is a good example), it is possible to run out of stack space. In
Java, the default stack space size is 64KB but it can be changed at the time
of executing the program using command line option -Xss . Heap space is
unlimited from the program's perspective. It is limited only by the amount
of space available on your machine.
Only temporary variables, i.e., variables created in a method (also known as
local variables and automatic variables) are created on the stack space.
Everything else is created on the heap space. If you have any doubt, ask
yourself this question - is this a temporary variable created in a method?
Yes? Then it is created on the stack. No? Then it is on the heap. Actual
objects are ALWAYS created on the heap.
When a method is invoked by a thread, it uses the thread's stack space to
keep its temporary variables.
Variables added to the stack space by a method are removed from the stack
when that method ends. Everything else created by a method is left on the
heap even after the method ends.
Here is a good discussion on Stack and Heap and why its understanding is
important .
1.7 Conventions
1.7.1 What is a Convention? ☝
You add a 15% tip to your bill at a restaurant. There is no law about that.
Nobody is going to put you in jail if you add nothing for a tip. But you still do it
because it is a convention. A lot of things in the world are based on convention.
In India, you drive on the left side of the road. This is a convention. It has
nothing to do with being technically correct. Indeed, people are fine driving on
the right side of the road in the US. But if you drive on the right side of the road
in India, you will cause accidents because that is not what other people expect
you to do.
It is the same in the programming world. As a programmer, you are a part of the
programmer community. The code that you write will be read by others and
while developing your code, you will read and use code written by others. It
saves everyone time and effort in going through a piece of code if it follows
conventions. It may sound ridiculous to name loop variables as i, j, or k,
but that is the convention. Anyone looking at a piece of code with a variable i
will immediately assume that it is just a temporary variable meant to iterate
through some loop.
If you decide to use a variable named i for storing some important program
element, your program will work fine but it will take other people time to realize
that and they will curse you for it.
If you are still unconvinced about the importance of conventions in

programming, let me put it another way. If I ask you to write some code in an
interview and if you use a variable named hello as a loop variable, I will not
hire you. I can assure you that most interviewers will not like that either.
Conventions are that important.
1.7.2 Conventions in Java ☝
Some of the most important conventions in Java are as follows:

1. Cases - Java uses "Camel Case" everywhere with minor differences.
1. Class names start with an uppercase letter. For example,
ReadOnlyArrayList is a good name but Readonlyarraylist
is not.
2. Package names are generally in all lowercase but they also may be in
camel case starting with a lower case letter. For example,
datastructures is a good package name but DataStructures
is not.
3. variable names start with a lower case and may include underscores.
For example, current_account is a good variable name.
2. Naming - Names should be meaningful. A program with a business purpose
should not have variables with names such as foo , bar , and fubar .
Although, such nonsensical names are used for illustrating or explaining
code in sample programs where names are not important.
3. Package names use a reverse domain name combined with a group name
and/or application name. For example, if you work at Bank of America's
Fixed Income Technologies division and if you are developing an
application named FX Blotter, all your packages for this application may
start with the name com.bofa.fit.fxblotter . The full class name
for a class named ReadOnlyArrayList could be -
com.bofa.fit.fxblotter.dataStructures.ReadOnlyArrayList
.
The reason for using a reverse domain name is that it makes it really easy to
come up with globally unique package names. For example, if a developer
in another group also creates his own ReadOnlyArrayList , the full
name of his class could be
com.bofa.derivatives.dataStructures.ReadOnlyArrayList
. There would be no problem if a third developer wants to use both the
classes at the same time in his code because their full names are different.
The important thing is that the names turned out to be different without any
of the programmers ever communicating with each other about the name of
their classes. The names are unique globally as well because the domain
names of companies are unique globally.
1.8 Compilation and Execution
1.8.1 Compilation and Execution ☝
Let us go over the basics really quickly. You know that a Java source file is
compiled into a Java class file and a class file is what is executed by the JVM.
You also know that you can organize your Java classes into packages by putting
a package statement at the top of a Java source file. The package name plus the
class name is called Fully Qualified Class Name or FQCN for short, of a Java
class. For example, consider the following code:
package accounting;
public class Account{
private String accountNumber;
System.out.println("Hello 1 2 3 testing...");
In the above code, accounting.Account is the fully qualified class name of

the class. This long name is the name that you need to use to refer to this class
from a class in another package. Of course, you can "import" accounting
package and then you can refer to this class by its short name Account . The
purpose of packages is to organize your classes according to their function to
ease their maintenance. It is no different from how you organize a physical file
cabinet where you keep your tax related papers in one drawer and bills in
another.
Packaging is meant solely for ease of maintenance. The Java compiler and the
JVM don't really care about it. You can keep all your classes in one package for
all that matters.
Let us create Account.java file and put it in your work folder (for example,
c:\javatest ). Copy the above mentioned code in the file and compile it as
follows:
c:\javatest\>javac Account.java
You should see Account.class in the same folder. Now, let us try to run it
from the same folder:
c:\javatest>java Account
You will get the following error:
Exception in thread "main" java.lang.NoClassDefFoundEr

ror: Account (wrong name: accounting/Account)
Of course, you need to use the long name to refer to the class, so, let's try this:
c:\javatest>java accounting.Account
You will now get the following error:
Error: Could not find or load main class accounting.Ac

count
Okay, now delete the Account.class file and compile the Java code like this:
c:\javatest\>javac -d . Account.java
You should now have the directory structure as shown below:
Now, run it like this:
c:\javatest>java -classpath . accounting.Account
You should see the following output:
Hello 1 2 3 testing...
What is going on? Well, by default the Java compiler compiles the Java source
file and puts the class file in the same folder as the source file. But the Java
command that launches the JVM expects the class file to be in a directory path
that mimics the package name. In this case, it expects the
Accounting.class file to be in a directory named accounting . The
accounting directory itself may lie anywhere on your file system but then that
location must be on the classpath for the JVM to find it.
One of the many command line options that javac supports is the -d option. It
directs the compiler to create the directory structure as per the package name of
the class and put the class file in the right place. In our example, it creates a
directory named accounting in the current directory and puts the class file in
that directory. The dot after -d in the javac command tells the compiler that the
dot, i.e., the current directory is the target directory for the resulting output. You
can replace dot with any other directory and the compiler will create the new
package based directory structure there. For example, the command
c:\javatest\>javac -d c:\myclassfiles Account.java will
cause the accounting directory to be created in c:\myclassfiles folder.
Now, at the time of execution you have to tell the JVM where to find the class
that you are asking it to execute. The -classpath (or its short form -cp )
option is meant exactly for that purpose. You use this option to specify where
your classes are located. You can specify multiple locations here. For example, if
you have a class located in c:\myclassfiles directory and if that class
refers to another class stored in c:\someotherdirectory , you should
specify both the locations in the classpath like this:
c:\java -classpath c:\myclassfiles;c:\someotherdirecto

ry accounting.Account
Observe that when you talk about the location of a class, it is not the location of
the class file that you are interested in but the location of the directory structure
of the class file. Thus, in the above command line, c:\myclassfiles should
contain the accounting directory and not Account.class file.
Account.class should be located inside the accounting directory. The
JVM searches in all the locations specified in the -classpath option for
classes.
Note: On *nix based systems, you need to use colon (:)

instead of semi-colon (;) and forward slash (/) instead
of back slash (\).
The Java command scans the current directory for class

files (and packages) by default, so, there is usually no
need to specify "dot" in the -classpath option. I have
specified it explicitly just to illustrate the use of the -
classpath option.
Compiling multiple source files at once ☝
Let's say you have two source files A.java and B.java in c:\javatest
directory with the following contents:
Contents of A.java:
package p1;
import p2.B;
public class A{
B b = new B();
}
Contents of B.java:
package p2;
public class B{
}
Open a command prompt, cd to c:\javatest , and compile A.java . You

will get a compilation error because class A depends on class B. Obviously, the
compiler will not be able to find B.class because you haven't compiled
B.java yet! Thus, you need to compile B.java first. Of course, as explained
before, you will need to use the -d . option while compiling B.java to make
javac create the appropriate directory structure along with the class file in
c:\javatest directory. This will create B.class in c:\javatest\p2
directory. Compilation of A.java will now succeed.
The point is that if you have two classes where one class depends on the other,
you need to compile the source file for the independent class first and the source
file for the dependent class later. However, most non-trivial Java applications are
composed of multiple classes coded in multiple source files. It is impractical to
determine the sequence of compilation of the source files manually. Moreover, it
is possible for two classes to be circularly dependent on each other. Which
source file would you compile first in such a case?
Fortunately, there is a simple solution. Just let the compiler figure out the
dependencies by specifying all the source files that you want to compile at once.
Here is how:
javac -d . A.java B.java
But again, specifying the names of all the source files would also be impractical.
Well, there is a solution for this as well:
javac -d . *.java
By specifying *.java, you are telling the compiler to compile all Java files that
exist in the current directory. The compiler will inspect all source files, figure out
the dependencies, create class files for all of them, and put the class files in an
appropriate directory structure as well. Isn't that neat?
If your Java source files refer to some preexisting class files that are stored in
another directory, you can state their availability to javac using the same -
classpath (or -cp ) option that we used for executing a class file using the
java command.
I strongly advise that you become comfortable with the compilation process by
following the steps outlined above.
1.8.2 Running a single file source code program ☝
Java designers felt that the two step compilation and execution of Java programs
is too tedious when you are trying to execute simple test programs. To make it
simple, Java 11 allows you to directly execute a Java source file using the java
command. For example, consider the following contents of TestClass.java
file in c:\javatest directory:
public class TestClass {

public static void main(String[] args) {
System.out.println("Hello world!");
}
}
You can execute this file directly from the command line using the following
command:
java TestClass.java
It will print Hello World! .
The only restriction with this approach is that your Java code must not refer to
code in any other Java file. You can have as many classes in the file as you want
but the first class that appears in the file must contain the main method because
that is the main method that the JVM will pick to execute.
Although this compilation cum execution technique is not on the exam , I have
included it here because it will save you a lot of time while trying out various
concepts using single file programs. You have to be careful about distinguishing
between compilation failure and exception at run time though. If you see
error: compilation failed on the console, it is a compilation error,
otherwise, it is an exception during execution.
1.8.3 Packaging classes into Jar ☝
It is undoubtedly easier to manage one file than multiple files. An application

may be composed of hundreds or even thousands of classes and if you want to
make that application downloadable from your website, you cannot expect the
users to download each file individually. You could zip them up but then the
users would have to unzip them to be able to run the application. To avoid this
problem, Java has created its own archive format called "Java Archive", which is
very much like a zip file but with an extension of jar.
Creating a jar file that maintains the package structure of class files is quite easy.
Let us say you have the directory structure shown below:

Go to the command prompt, cd to c:\javatest directory and run the

following command:
jar -cvf accounting.jar accounting
This command tells the jar utility to create accounting.jar file and include
the entire directory named accounting in it along with its internal files and
directories. You should now have the directory structure shown below:
Assuming that you are still in c:\javatest directory on your command

prompt, you can now run the class through the jar file like this:
java -classpath .\accounting.jar accounting.Account
Note that you must maintain the package structure of the class while creating the
jar file. If you open accounting.jar in WinZip or 7zip , you will see that
this jar contains Account.class under accounting directory.
Besides the class files, the Jar file allows you to keep information about the
contents of the jar file within the jar file itself. This information is kept in a
special file is called MANIFEST.MF and is kept inside the META-INF folder
of the jar file. (This is just like airlines using a "manifest" to document the cargo
or a list of passengers on a flight.) For example, you can specify the entry point
of an application which will allow you to run a Jar file directly (from the
command line or even by just double clicking the jar file in your file explorer)
without having to specify the class name containing the main method on the
command line. T ypical contents of this file are as follows -
Manifest-Version: 1.0
Created-By: 1.7.0_09-b05 (Oracle Corporation)
Main-Class: accounting.Account
You can actually go ahead and create mymanifest.txt file with the above
mentioned contents in c:\javatest directory and use the following
command to create the jar:
jar -cvfm accounting.jar mymanifest.txt accounting
c is for create, v is for verbose (i.e. display detailed information on command

line), f is for the output file, and m is the name of the file the contents of which
have to be included in the jar's manifest. Notice that the name of the manifest file
on the command line is not important. Only the contents of the file are
important. This command will automatically add a file named MANIFEST.MF
inside the META-INF folder of the jar file.
Once you have this information inside the jar file, all you need to do to run the
program is to execute the following command on the command line:
java -jar accounting.jar
Although packaging classes into Jar files is not on the

exam as such, you will need to know about it from the
perspective of execution of modules, which is on the
exam.
1.8.4 Compilation error vs exception at run time ☝
Understanding whether something will cause a failure during compilation or will

cause an exception to be thrown at run time is important for the exam because a
good number of questions in the exam will have these two possibilities as
options. Beginners often get frustrated while trying to distinguish between the
two situations. It will get a little easier if you keep the following three points in
mind:
1. First and foremost, it is the compiler's job to check whether the code
follows the syntactical rules of the language. This means, it will generate an
error upon encountering any syntactical mistake. For example, Java
requires that the package statement, if present, must be the first statement in
the Java code file. If you try to put the package statement after an import
statement, the compiler will complain because such a code will be
syntactically incorrect. You will see several such rules throughout this book.
Yes, you will need to memorize all those. If you use an IDE such as Eclipse,
NetBeans, or IntelliJ, you should stop using it because you need to train
your brain to spot such errors instead of relying on the IDE. Using Notepad
to write and using the command line to compile and run the test programs is
very helpful in mastering this aspect of the exam.
2. Besides being syntactically correct, the compiler wants to make sure that
the code is logically correct as well. However, the compiler is limited by the
fact that it cannot execute any code and so, it can never identify all the
logical errors that the code may have. Even so, if, based on the information
present in the code, the compiler determines that something is patently
wrong with the code, it raises an error. It is this category of errors that
causes the most frustration among beginners. For example, the statement
byte b = 200; is syntactically correct but the compiler does not like it.
The compiler knows that the value 200 is too big to fit into a byte and it
believes that the programmer is making a logical mistake here. On the other
hand, the compiler okays the statement int i = 10/0; even though
you know just by looking at the code that this statement is problematic.

3. The JVM is the ultimate guard that maintains the integrity and type safety
of the Java virtual machine at all times. Unlike the compiler, the JVM
knows about everything that the code tries to do and it throws an exception
(I am using the word exception in a general sense here and not referring to
the java.lang.Exception class) as soon as it determines that the
action may damage the integrity or the type safety of the JVM. Thus, any
potentially illegal activity that escapes the compiler will be caught by the
JVM and will result in an exception to be thrown at run time. For example,
dividing a number by zero does not generate any meaningful integral value
and that is why the JVM throws an exception if the code tries to divide an
integral value by zero.
Honestly, this is not an easy topic to master. The only way to get a handle on this
is to know about all the cases where this distinction is not so straightforward to
make. If you follow this book, you will learn about all such rules, their
exceptions, and the reasons behind them, that are required for the exam.
1.9 Nomenclature
1.9.1 Commonly used terms in Java development ☝
During your programming career you will be reading a lot. It could be books,
articles, blogs, manuals, tutorials, and even discussion forums. You will also be
interacting with other Java developers in various roles such as interviewers, team
members, architects, and colleagues. To make the most out of these interactions,
it is very important to form a clear and precise understanding of commonly used
terms.
I will explain the commonly used phrases, names, and terminology in the Java
world.
1. Class - Unless stated otherwise or unless clear from the context, the term
class includes class, interface, and enum. Usually, people mean "type" when
they say "class". You should, however, always try to be precise and use the
term class only for class.
2. Type - Type refers to classes, interfaces, enums, and also primitive types
(byte, char, short, int, long, float, double, and boolean).
3. Primitive types - byte, char, short, int, long, float, double, and boolean are
called primitive types because they just hold data and have no behavior.
You can perform operations on them but you cannot call methods on them.
They do not have any property or state other than the data value that they
contain. You access them directly and never through references.
4. Reference types - Classes, Interfaces, and Enums are called reference types
because you always refer to them through references and never directly.
Unlike primitive types, reference types have behavior and/or state.
5. Top-level reference types - Classes, interfaces, or enums that are defined

directly under a package are called top-level classes, interfaces, or enums.
6. Nested reference types - Classes, interfaces, and enums that are defined
inside another class, interface, or an enum are called nested classes,
interfaces, or enums.
7. Inner reference types - Non-static nested classes, interfaces, and enums

that are called inner classes, interfaces, or enums.
8. Local reference types - Nested reference types that are defined inside a
method (or inside another code block but not directly inside a class,
interface, or enum) are called local classes, interfaces, or enums.
9. Anonymous classes - This is a special case of a nested class where just the
class definition is present in the code and the complete declaration is
automatically inferred by the compiler through the context. An anonymous
class is always a nested class and is never static.
10. Compile time vs run time (i.e. execution time) - You know that there are
two steps in executing Java code. The first step is to compile the Java code
using the Java compiler to create a class file and the second step is to
execute the JVM and pass the class file name as an argument. Anything
that happens while compiling the code such as generation of compiler
warnings or error messages is said to happen during "compile time".
Anything that happens while executing the program is said to happen
during the "run time". For example, syntax errors such as a missing bracket
or a semicolon are caught at compile time while any exception that is
generated while executing the code is thrown at run time. It is kind of
obvious but I have seen many beginners posting questions such as, "why
does this code throw the following exception when I try to compile it?",
when they really mean, "why does this code generate the following error
message while compilation?" Another common question is, "why does this
code throw an exception even after successful compilation?" Successful
compilation is not a guarantee for successful execution! Although the
compiler tries to prevent a lot of bugs by raising warnings and error
messages while compilation, successful compilation really just means that
the code is syntactically correct.
11. Compile-time constants - Normally, it is the JVM that sets the values of
variables when a program is executed. The compiler does not execute any
code and it has no knowledge of the values that a variable might take during
the execution of the program. Even so, in certain cases, it is possible for the
compiler to figure out the value of a variable. If a compiler can determine
the value that a variable will take during the execution of the program, then
that variable is actually a compile-time constant. For example, if you define
an int variable as final int x = 10; then x is a compile time
constant because the compiler knows tha x will always have a value of 10
at run time. Similarly, literals such as the numbers 1, 2, and 3, or the
characters written in code within single quotes such as 'a' , or boolean
values true and false , are all compile time constants because the
compiler knows that these values will never change.
I will refer to these terms and will also discuss the details of these terms
throughout the course so, it will be helpful if you keep the basic idea of these
terms in mind.
1.10 Java Identifiers

1.10.1 Java Identifiers ☝
Java has specific rules to name things such as variables, methods, and classes.
All these names belong to a category of names called "identifiers".
Java defines an identifier as an unlimited-length sequence of Java letters and

Java digits, the first of which must be a Java letter. An identifier cannot have the
same spelling as a Java keyword or a literal (i.e. true , false , or null ).
For example, the following variable names are invalid:
int int; //int is a keyword
String class; //class is keyword
Account 1a; //cannot start with a digit
byte true; //true is a literal
Java letters include uppercase and lowercase ASCII Latin letters A-Z ( \u0041-
\u005a ), and a-z ( \u0061-\u007a ), and, for historical reasons, the ASCII
underscore ( _ or \u005f ) and dollar sign ( $ or \u0024 ). The "Java digits"
include the ASCII digits 0-9 ( \u0030-\u0039 ).
Older versions of the exam tested candidates on

identifying valid identifiers. However, the current exam
has moved away a bit from making the candidate a
human compiler and does not include this topic. You
should still have a basic idea about what an identifier
is though because this concept applies to all kind of
names in Java.
Chapter 2 Creating a Simple Java
Program
Exam Objectives

2.1 Create an executable Java program with a main

class
2.1.1 The main method ☝
Let's get one thing out of the way first. Java classes are not executables . You
cannot "execute " Java classes. The Java Virtual Machine (JVM) is an
executable. You execute the JVM. You actually pass the FQCN of a Java class as
an argument to the JVM. When the JVM runs, it loads the given class and looks
for a specific method in that class. If it finds that method, it passes control to that
method and this method then becomes the in-charge from there onward. If the
JVM doesn't find that specific method, it errors out. In common parlance, we call
it as executing or running a Java class or a program.
The method that the JVM is hardwired to look for in the class is called the
"main" method and this method has a very specific signature - its name must
be main and it must take exactly one parameter of type String array . In
addition to this, it must return void , must be public and must also be static . It
is free to declare any exception in its throws clause. If your class has such a
method, the JVM can invoke this method and therefore, it is possible to execute
the class. The meaning of void , public , static , and throws will be
clear as you proceed through the book but for now, just assume that this is how
the main method has to be.
Examples of a valid main method:
1. public static void main(String[] args){ } - This is the

version that you will see most of the time.
2. public static void main(String... args){ } - Note that
String... is the same as String[] as far as the JVM is concerned.
The three dots syntax is called vargargs. I will talk more about it later.
3. public static void main(String args[]) throws
Exception{ throw new Exception(); } - The main method is
allowed to throw any exception.
Examples of an invalid main method:
1. static void main(String abc[]) { } - Invalid because it is

not public.
2. public void main(String[] x){ } - Invalid because it is not

static.
3. public static void main(String[] a, String b){ } -

Invalid because it doesn't take exactly one parameter of type String array.
4. static void Main(String[] args){ } - Invalid because it is

not public and the name starts with a capital M. Remember that Java is case
sensitive.
Note that all of the above methods are valid methods in their own right. It is not
a compilation error if you have these methods in your class. But they cannot be
accepted as the "main" method. JVM will complain if you try to execute a class
on the basis of these methods. JVM has gotten smarter over the years and in Java
11, it gives out a very helpful error message that explains the problem with your
main method. For example, if it is not static, you will see the following message:
Error: Main method is not static in class TestClass,

please define the main method as:
public static void main(String[] args)
Examples of really weird looking main methods:
1. public static native void main(String[] args); - Out

of scope for the exam, but good to know. This is a valid main method. A
native method means you are going to implement this method in a separate
executable library which will be linked at run time. In this case, the JVM
will look for the implementation of your main method in a dynamically
linked library. If it finds an implementation, all is good. If not, then it will
throw an error saying it is unable to find the implementation to this native
method.
2. public abstract static void main(String[] args); -

Invalid because static methods cannot be abstract.
Just like with any other method, it is possible to have overloaded main methods
in a Java class. I will talk overloading in detail later, but for now, it means
having multiple methods with same name but different parameters. The JVM
looks for a specific main method as described above. All other main methods
have no special meaning for the JVM.
Many questions in the certification exam assume the presence of the

main method. You may be given a code snippet and asked to
determine the output. If you don't see any main method in the given
code you need to assume that there is a main method somewhere
that is invoked by the JVM and the given code or a method is
invoked through that method.
2.1.2 Command line arguments ☝
It is possible to provide any number of arguments while executing a class by

specifying them right after the name of the class on the command line. The
arguments must be separated by a space character. For example, if you want to
pass three arguments to your class named TestClass, your command would be:
java TestClass a b c
The JVM passes on the arguments specified on the command line to the main
method through its String[] parameter. In other words, the String[]
parameter of the main method contains the arguments specified on the command
line. An important implication of this is that all the arguments are passed to the
main method as Strings. For example, if your command line is java
TestClass 1 , the main method will get a String array with one String
element containing 1 and not an int 1.
Let me now present to you the following program to explain how to use
command line arguments. This simple program prints the arguments that were
passed to it from the command line.
public class TestClass{

public static void main(String[] args) throws Excep
tion{
for(int i=0; i<args.length; i++){
System.out.println("args["+i+"] = \""+args[i
]+"\"");
}
}
}
The output of this program will tell you all you need to know about the
command line arguments. The following is a table containing the command line
used to execute the program and the corresponding output generated by the
program:
Command line used Output Inference

If no argument is specified, args contains a
String array of length 0. Observe that a
NullPointerException is not raised for
java TestClass
args.length . That means args is not null .
In this case args refers to a String
array of length 0.
The first argument is stored at index 0.
The first argument is NOT "java" or
java TestClass a args[0] = "a"
even the name of the class, i.e.,
"TestClass" .
args[0] = "a" Arguments can be separated by one or
java TestClass a b c
args[1] = "b" more than one white characters. All such
args[2] = "c" separator characters are stripped away.
If you put quotes around the value that you
want to pass, everything inside the quotes
is considered one parameter.
args[0] = "a " Quotes are not considered as part of the
java TestClass "a " b
args[1] = "b" argument. Observe that the first argument
is "a " , i.e., a String containing
character 'a' followed by a space
character.
To pass a quote character as an argument,
java TestClass "\"" args[0] = """ you have to escape it using a backslash.
By the way, can you guess why the name of class is not passed in as an argument
to the main method? Remember that unlike an executable program, you cannot
change the name of a Java class file. The name of a Java class file will always be
the same as the name given to the class in the Java source file. Therefore, the
main method of a class always knows the name of its containing class.
2.1.3 The end of main ☝
As discussed before, once the JVM passes control to the main method of the
class that you are trying to execute, it is the main method that decides what to do
next. As far as the JVM is concerned, your application has been "launched" upon
invocation of the main method. In that sense, the main method is just an entry
point of your application. So, what happens when the main method ends? Does
the application end as well? Well, the answer is it depends on what the main
method does.
A simple Java program such as the one we used earlier to print arguments may
have all its code in the main method. Once the main method ends, there is
nothing else to do and so, the program ends. While you can write all your
application code within the main method, Java applications are usually
composed of multiple classes. At run time, an application consists of instances of
several classes that interact with each other by calling methods on each other. A
Java application may also perform multiple activities in parallel, so, even if an
activity implemented by one method ends, another activity implemented by
some other method may still be going on. The code in main itself is just an
activity. The end of the main method implies the end of only that activity. It
doesn't mean the end of all the activities that may be going on in an application.
If it helps, you may think of your Java application as a fast food restaurant and
the main method as its manager opening the restaurant in the morning. The
restaurant need not close immediately after opening if there are no customers
lined up. After opening the restaurant, the manager kicks off a lot of activities
such as preparing the food, setting up the dining area, and waiting for customers.
Such activities may continue side by side throughout the course of the day.
When the last customer of the day is gone and when all such actives end, the
restaurant closes for the day. The same thing happens in a Java application. The
main method may kick off other activities that run side by side and the
application ends only when all such activities come to an end.
Java allows executing activities in parallel using threads. This topic is beyond
the scope of this exam so, I will not discuss it anymore in this book. But you
should know that in a nutshell, an application doesn't end until all the threads
started by the application, including the thread that executes main, end.
2.2 Compile and run a Java program from the

command line
2.2.1 Run a Java program from the command line ☝
We have already seen the basics of how to compile and execute a Java program
in Compilation and Execution section under Kickstarter for Beginners. I will just
summarize the important points that you need to know for the exam here.
1. The standard Oracle JDK comes bundled with a Java compiler. The
executable for the compiler is named javac . In other words, the program
you need to use to compile your Java code is called "javac". You may
compile a Java source file named TestClass.java using the command
- javac TestClass.java
Notice that you have to specify the full file name including the extension.
javac does support multiple options to fine tune the compilation process but
none of these options are required for the exam.
2. Compilation of a Java file results in one of more class files depending on

the contents of the Java source file.
3. The standard Oracle JDK comes bundled with a Java Virtual Machine
(JVM). The executable for the JVM is named java. In other words, the
program you need to launch the JVM and to execute your Java program is
called "java ".
4. To execute a Java class, you can use the following command - java
TestClass
Notice the absence of file extension .class while specifying the class name.
To compile a Java source file, you must specify the extension of the file,
i.e., .java though. Just like with javac, Java command can take multiple
options to fine tune the execution of a Java program. You do not need to
know any of these options for the exam.
5. Java 11 allows what is called "single file source code programs", which
means, if your program is present in a single source file, you can run that
file directly using the java command instead of compiling it first using
javac . For example, java TestClass.java
2.3 Create and import packages

2.3.1 The package statement ☝
Every Java class belongs to some or the other package. The name of this
package is specified using the package statement contained in a source file.
There can be at the most one package statement in the entire source file and, if
present, it must be the first statement (excluding comments, of course) in the file.
All top level types defined in this file belong to this package. If there is no
package statement in a Java file, then the classes defined in that file belong to an
unnamed package which is also known as the "default" package. In other
words, if you have two Java files without any package statement, classes defined
in those two files belong to the same unnamed package.
Important points about the unnamed package ☝
1. The unnamed package has no name. Duh!

2. Default is not the name of the unnamed package. There is no package
named default . You cannot even create a package named default by
specifying default as the package name for your class though because
default is a keyword.
3. Since the unnamed package has no name, it is not possible to refer to this
package. In other words, it is not possible to import classes belonging to the
unnamed package into classes belonging to another package. You can't do
import *; in your Java file. This is one reason why it is not
recommended to create classes without a package statement.
You can name your package anything but it is recommended that you use the
reverse domain name format for package. For example, if you work at Amazon ,
you should start your package name with com.amazon . You should then append
the group name and application name to your package name so, as to make your
class unique across the globe. For example, if the name of your group is sales,
and the name of the application is itemMaster, you might name your package
com.amazon.sales.itemMaster . If the name of your class is Item , your
Item.java source file will look like this:
package com.amazon.sales.itemMaster;
public class Item{
}
Although you can use non-ascii characters in your

package name, the exam will not ask you questions
about package names with such characters.
2.3.2 Quiz ☝
Q1. Which of the following code snippets are valid?
Select 1 correct option.
A.
//in Test.java
package;
public class Test{
}
B.
//in Test.java
package mypackage;
public class Test{
}
C.
//in Test.java
package x;
public class Test{
}
package y;
class AnotherTest{
}
D.
//in Test.java
package x;
package y;
public class Test{
}
Correct answer is B.
A is incorrect because you must specify the package name along with the
keyword package.
C and D are incorrect because you cannot have more than one package statement
in a Java source file. Moreover, C is incorrect also because the package
statement must be the first statement in a Java file if it exists in the file.
2.3.3 The import statement ☝
If all of your classes are in the same package, you can just use the simple class
name of a class to refer to that class in another class. But to refer to a class in one
package from another, you need to use its "fully qualified class name" or
FQCN for short. FQCN of a class is basically the package name + dot + the
class name. For example, if the package statement in your class Test is
package com.enthuware.ocp; , the FQCN of this class is
com.enthuware.ocp.Test .
If you want to refer to this class from another class in a different package, say
com.xyz.abc , you need to use the FQCN, i.e.,
com.enthuware.ocp.Test . For example,
com.enthuware.ocp.Test t = new
com.enthuware.ocp.Test();
If you try to use just the simple class name, i.e., Test t = new Test(); ,
the compiler will assume that you mean to use the Test class from the same
package, i.e., com.xyz.abc and if it doesn't find that class in com.xyz.abc
package, it will complain that it doesn't understand what you mean by "Test".
FQCN tells the compiler exactly which class you intend to use.
If you refer to this class several times in your code, you can see that it will lead
to too many repetitions of com.enthuware.ocp in the code. The import
statement solves this problem. If you add an import statement import
com.enthuware.ocp.Test; , you can use just the simple class name
Test in your class to refer to com.enthuware.ocp.Test class.
If your class refers to multiple classes of the same package, you can use either
use one import statement for each class or you can use just one import statement
with a wild card character * for the whole package. For example, import
com.enthuware.ocp.*; The compiler will try to find the simple class
names used in your code in the imported package(s). You can have as many
import statements as you need. You can also have redundant imports or imports
that are not needed.
Although importing all the classes with a wildcard looks

a like good idea but I assure you that it is not. In
practice, if a class uses several classes from different
packages, it becomes difficult to figure out the package
to which a class referred to in the code belongs. For
this reason, well written, professional code always uses
import statements for specific classes instead of using
the wildcard format. Most IDEs even have a feature to
clean up import statements of a class. In NetBeans,
you can do it with Control+Shift+i, for example.
The import static statement ☝
Sometimes you need to define values that are to be used in various classes of
your application. For example, if you are developing an application for tax
computation, you may want to define a value for tax rate that is to be used by all
other classes. Since all code in Java must be a part of a reference type (i.e. a
class or an interface or an enum), you may define a class named Values and add
this value to this class as follows:
package taxes;
public class Values{
public static double TAX_RATE = 0.15;
}
Now, if you want to use this value in some other class, you have three options.
You know the first two options already:
Option 1 - Don't import anything and just use taxes.Values.TAX_RATE;

in your class.
Option 2 - Add import taxes.Values; or import taxes.*; and

then use Values.TAX_RATE in your class.
Option 3 - Java 7 onwards, you have a third option called "import static ". To
eliminate typing the class name multiple times, you can simply import the static
members of any class using the import static statement. In this example, you can
add import static taxes.Values.TAX_RATE; or import static
taxes.Values.*; (the wild card format imports all static members of the
class) to the list of import statements of the class and then use just TAX_RATE in
your code.
You can import static fields as well as static methods through this statement. So,
for example, if you have static utility method named apply in Values class,
you could directly use the name apply instead of Values.apply if you
include import static taxes.Values.apply; or import static
taxes.values.*;
Remember that import static does not import a class. It imports only the
static member(s) of a class. Thus, you cannot use the simple name Values in
your code if you haven't imported taxes.* or taxes.Values already using
the regular import statement.
The word "import " is really a misnomer here. The

import statement doesn't import anything into your class.
It is merely a hint to the compiler to look for classes
in the imported package. You are basically telling the
compiler that the simple class names referred in this
code are actually referring to classes in the packages
mentioned in the import statements. If the compiler is
unable to resolve a simple class name (because that
class is not in the same package as this class), it will
check the import statements and see if the packages
mentioned there contain that class. If yes, then that
class will be used, if not, a compilation error will be
generated.
Important points about the import statement -
1. You can import each class individually using import

packagename.classname; statement or all the classes of a package
using import packagename.*; or any combination thereof.
2. import statements are optional. You can refer to a class from another
package even without using import statements. You will have to write
FQCN of the class in your code in that case.
3. You can import any number of packages or classes. Duplicate import

statements and redundant import statements are allowed. You can import a
class even if you are not using that class in your code. Remember, an import
statement is just a shortcut for humans. It doesn't actually import anything
in your class.
4. java.lang package is imported automatically in all the classes. You don't

need to write import java.lang.*; in your class even if you use
classes from java.lang package. But it is not wrong to import it anyway
because redundant imports are allowed.
What you cannot do:
1. There is no way to import a "subpackage" using the import statement. For

example, import com.enthuware.*; will import all the class in
package com.enthuware but it will not import any class under
com.enthuware.ocp package. Furthermore, import
com.enthuware.*.*; is illegal. This essentially means that
technically, there is no concept of "subpackage" in Java. Each package must
be imported separately.
2. You cannot import a package or a class that doesn't exist. For example, if
you try to use some random package name such as import xyz.*; the
compiler will raise an error saying,
error: package xyz does not exist

import xyz.*;
^
1 error
How does the compiler know whether a package exists or not, you ask?
Well, if the compiler doesn't find any class in its classpath that belongs to
the package that you want to import, it draws the inference that such a
package does not exist.
3. Unpackaged classes (the phrases "unpackaged classes" and "classes in the

default or unnamed package" mean the same thing, i.e., classes that do not
have any package statement) cannot be imported in any other package. You
cannot do something like import *;
4. If a class by the same name exists in multiple packages and if you import
both the packages in your code, you cannot use the simple class name in
your code because using the simple name will be ambiguous. The compiler
cannot figure out which class you really mean. Therefore, you have to use
FQCN in such a case. You may import one package or class using the
import statement and use simple name for a class in that package and use
FQCN for classes in the other package.
The requirement to use two classes with same name

but from different packages typically used to arise a lot
while using JDBC. JDBC related classes are in
java.sql package and classes in this package use
java.sql.Date class instead of java.util.Date . But
the non-JDBC related code of the application uses
java.util.Date . In such a situation, it is preferable
to use FQCN of each class in the code to avoid any
confusion to the reader even though you can import
one package and use simple name Date to refer to the
class of that package.
However, Java 8 onward, you should use the new

Date/Time classes of the java.time package anyway,
which eliminates this annoyance.
2.3.4 Quiz ☝
Q. You have downloaded two Java libraries. Their package names are
com.xyz.util and com.abc.util . Both the packages have a class named
Calculator and both the classes have a static method
named calculate() .
You are developing your class named MyClass in com.mycompany.app

package and your class code needs to invoke calculate methods belonging to
both of the Calculator classes as follows:
public class MyClass{

//call xyz's calculate()
//call abc's calculate()
}
}
Which of the following approaches will work?

A. Add import com.*; to your class. Then use
xyz.util.Calculator.calculate(); and
abc.util.Calculator.calculate();
B. Add import com.xyz.util.Calculator; Then use

Calculator.calculate(); and
com.abc.util.Calculator.calculate();
C. Do not use any import statement. In the code, use

com.xyz.util.Calculator.calculate(); and
com.abc.util.Calculator.calculate();
D. This cannot be done.
Correct answer is B and C.
Option A is incorrect because you cannot import partial package names. While
using a class, you can either use simple class name (if you have imported the
class or package using the import statement) or use Fully Qualified Class Name.
You cannot use partial package name to refer to a class.
2.4 Exercises ☝
1. Create classes in two different named packages. Define static and instance
fields in one of those classes and access those fields from the other class.
See what happens when both the classes try to access the fields of each
other.
Hint: If you have trouble compiling classes, check out "Compilation and
Execution" section in Kickstarter for Beginners".
2. Define a local variable in a method. Update this variable in a while loop
and print it out after the while loop ends. Check what happens when you
define a variable by the same name within the while loop.
3. Create a class in package foo and another class in package foo.bar with a
static method. Invoke the static method from the class in package foo using
different import statements.
4. Create a class with a main method and execute the class with a few
arguments. Print the number of arguments.
5. Which Java feature (or lack of thereof) annoys you most. Why?
Chapter 3 Working With Java
Primitive Data Types
Exam Objectives

data types)
Identify the scope of variables (*moved to the next chapter)
Use local variable type inference (*moved to the next chapter)
3.1 Data types in Java ☝

Java has support for two kinds of Data.
3.1.1 Data types and Variable types ☝
A data type is essentially a name given to a certain kind of data. For example,
integer data is given the name "int" in Java. Boolean data is given the name
"boolean" in Java. Classifying data into different data types allows you to treat
data of the same kind in the same way. It also allows you to define a set of
operations that can be performed on data of the same kind. For example, if you
are given data of type int and of type boolean, you know that you can do
addition operation on the int data but not on the boolean data. Data type also
determines the space required to store that kind of data. For example, a byte
requires only 8 bits to store while an int requires 32 bits.
Data types are important for a programming language because they allow you
tell the compiler the kind of data you want to work with. For example, when you
say int i; you are telling the compiler that i is of type int . The compiler
will then allow you to store only an integer value in this variable. Ignoring a few
exceptions, it is not possible to store data of one type into a variable of another
type because of the difference in the amount of space required by different data
types and/or because of their compatibility.
Languages where the data type of a variable is defined

at compile time and cannot change during run time are
called "statically typed" languages. Java is, therefore, a
statically typed language. Languages that allow the data
type of a variable to change at run time are called
dynamically typed languages. JavaScript is an example
of a dynamically typed language. A variable in
JavaScript code may contain an integer value in one
statement and may point to a String at the next
statement. There are advantages and disadvantages to
each approach but discussing that would be way out of
scope. You should, however, read about it online for
interview purposes.
By defining a variable of a certain type, you automatically get the right to

perform operations that are valid for that type on that variable. For example, if i
is defined to be of type the int , the compiler will allow you to perform only
mathematical and bit wise operations on this variable. If b is defined as
a boolean , the compiler will allow you to perform only logical operations on
this variable.
Java has two fundamental kinds of data types: primitive and reference .
Primitive data types are designed to be used when you are working with raw
data such integers, floating point numbers, characters, and booleans. Java (by
Java, I mean, the Java compiler and the Java Virtual Machine) inherently knows
what these data types mean, how much space they take up, and what can be done
with them. You don't need to explain anything about them to Java. Primitive data
types are the most basic building blocks of a Java program. You combine
primitive data types together in a class to build more complicated data types.
Reference data types , on the other hand, are designed to be used when you are
working with data that has a special meaning for code that Java has no
knowledge of. For example, if you are developing an application for student
management, you might define entities such as Student, Course, and Grade. Java
has no knowledge of what Student, Course, and Grade mean. It doesn't know
how much space they take, what operations they support, or what properties they
have. Java will expect you to define all these things. Once you define them, you
can use them to implement the business logic of your application. When you
write a class, interface, or enum, you are essentially defining a reference data
type. Reference data types are built by combining primitive data types and other
reference data types.
In Java, primitive data types include integral data types (byte, char,
short, int, long ), floating point data types (float, double ), and
boolean data type (there is only one - boolean ). While reference data types
include all the classes, interfaces, and, enums, irrespective of who defines
them. If something is a class, an interface, or an enum, it is a reference data type.
Yes, String too is a reference data type because all strings are instances of type
java.lang.String class :) I will talk more about Strings later.
Note that integral and floating point data types are collectively called
numeric data types .
The following table lists out the details of primitive data types:
You will not be asked the details of the sizes of data types in the
exam. However, it is important to know about them as a Java
programmer.
Primitive data Types in Java

Size
Data Sample Operations
(in Range of values
Type Values supported
bits)
mathematical,
byte 8 -2^7 to 2^7 -1, i.e., -128 to 127 -1, 0, 1
bitwise
0, 1, 2, 'a', mathematical,
char 16 0 to 2^16-1, i.e., 0 to 65,535
'\uo0061' bitwise
-2^15 to 2^15-1, i.e., -32,768 to Mathematical,
short 16 32,767 -1, 2, 3 bitwise
mathematical,
int 32 -2^31 to 2^31-1 -1, 2, 3
bitwise
mathematical,
long 64 -2^63 to 2^63-1 -1, 2, 3
bitwise
approximately Â
float 32 1.1f, 2.0f mathematical
±3.40282347E+38F
approximately Â
double 64 1.1, 2.0 mathematical
±1.79769313486231570E+308
boolean 1 true or false true, false logical

Notes:
1. byte, char, short, int, and, long are called integral data
types because they all store precise integral values.
2. char is also an integral type that stores numbers just like byte, short, int
and long. But it cannot store a negative number. The number stored in a
char variable is interpreted as a unicode character .
3. float and double store large but imprecise values. Java follows IEEE
754 standard. You may go through it to learn more but it is not required for
the exam.
4. A boolean stores only two values and therefore, requires only one bit of
memory. Officially however, its size is not defined because the size
depends on the smallest chunk of memory that can be addressed by the
operating system. On 32 bit systems, a boolean may even use 4 bytes.
A word on void ☝
void is a keyword in Java and it means "nothing". It

is used as a return type of a method to signify that the
method never returns anything. In that sense, void is a
type specification and not a data type in itself. That is
why, even though you can declare a method as
returning void but you cannot declare a variable of
type void .
Difference between null and void ☝
null is also a keyword in Java and means "nothing".

However, null is a value. It is used to signify that a
reference variable is currently not pointing to any
object. It is not a data type and so, it is not possible
to declare a variable of type null .
Note that null is a valid value for a reference variable

while void is not. When a method invocation returns
null , it means that only that particular invocation of
the method did not return a valid reference. It does not
mean that the method never returns a valid reference.
On the other hand, void means that a method never
returns anything at all. Therefore, you cannot use void
in a return statement. In other words, return null;
can be a valid return statement for a method but
return void; is never valid. A method that declares
void as its return type, can either omit the return
statement altogether from its body or have an empty
return statement, i.e., return; .
Types of variables ☝
Java has two types of variables to work with the two types of data types, namely
primitive variables and reference variables. Primitive variables let you work with
primitive data, while reference variables let you work with reference data. Thus,
when you define int i; i is a variable of the primitive data type int , but
when you define, String str; str is a variable of the reference data type
java.lang.String .
It is very important to understand the fundamental difference between the two
types of variables. A primitive variable contains primitive data within itself,
while a reference variable stores only the address to the location where the actual
data is stored. For example, if you do i = 10; , i will contain the value 10.
But if you do str = "hello"; , str will only contain the address of the
memory location where the string "hello" resides. You can now understand
why they are called "reference" variables. Because they are merely references to
the actual data! When you perform any operation on a reference, the operation is
actually performed on the object that is located somewhere else. In that sense,
you can think of a reference variable as a "remote control" of a TV. (If you have
trouble understanding this, you should go through the "Kickstarter for
Beginners" chapter before moving forward.)
Both types of variables support the assignment operation, i.e., they allow you to
assign values to them. For example, the statement i = 20; assigns the value
20 to the variable i .
In case of a reference variable, you cannot assign the address of an object

directly. You can only do so indirectly. For example, in statement the String
str = "hello"; you are assigning the address of the memory location at
which the string "hello" is stored to str variable. str , therefore, now
contains the address of a memory location. Similarly, in statement String
str2 = str; you are assigning the value stored in str to str2 . You are
not copying "hello" to str2 . You are just copying the address stored in str
to str2 . You cannot assign a memory address to a reference variable directly
because you don't know the actual address. Only the JVM knows where an
object is stored in memory and it assigns that address to the variable while
executing the assignment operation. The only "address" you can assign to a
reference variable directly is null .
Size of variables ☝
Since a primitive variable stores actual data within itself, the size of a primitive
variable depends on the size of the primitive data. Thus, a byte variable requires
1 byte while an int variable requires 4 bytes and so on.
Since a reference variable stores only the address of a memory location, the size
of a reference variable depends on the addressing mechanism of the machine. On
a system with 32 bit OS, a reference variable will be of 4 bytes, while on a 64 bit
systems, it will be of 8 bytes.
Size of reference data types

Size of a reference data type such as a class can be
easily determined at compile time by looking at the
instance variables defined in that class. Since every
instance variable will either be a primitive variable or a
reference variable, and since you know the sizes of
each of those types, the size of an instance of that
class will simply be the sum of the sizes of its instance
variables.
This size never changes for a given class. All instances
of a given class always take exactly the same amount
of space in memory, no matter what values its internal
variables hold.
Thus, there is never a need to calculate the size of
memory space taken by an instance of a class at run
time. And for this reason, there is no such operator as
"sizeof" in Java.
3.2 Difference between reference variables and

primitive variables
3.2.1 Reference variables and primitive variables ☝
In the "Object and Reference" lesson, we discussed the relationship between a

class, an object, and a reference. I explained the fundamental difference between
an object reference and a primitive. To recap, there is no difference between an
object reference and a primitive variable from the program memory perspective.
In memory, both just store a raw number. The difference is in how the JVM
interprets that raw number. In the case of a reference variable, the JVM
interprets the number as an address of another memory location where the actual
object is stored, but in the case of a primitive variable, it interprets the raw
number as a primitive data type (i.e. a byte, char, short, int, long, float, double, or
boolean). In that sense, primitives do not have references. There is nothing like a
primitive "reference" because there is no object associated with a primitive
variable.
Another crucial point to understand here is that it is the objects that support
methods and have fields, not the references. Therefore, when you invoke a
method (or access a field) using a reference, the JVM invokes that method on the
actual object referred to by that variable and not on the variable itself.
Since primitives are not objects, you cannot "invoke" any method on a primitive
variable. But you can perform mathematical (+, -, *, /, and, %), logical ( ||, &&,
!, |, and, &), and bitwise(âˆ¼, |, and, &) operations on the primitive variables
themselves.
The following image explains the above with some code. The code assumes that
there is a class named Student defined as follows:
public class Student{

int id;
}
As you can observe in the above flow diagram, whenever you assign one
variable to another, the JVM just copies the value contained in the variable on
the right-hand side of the assignment operator to the variable on the left-hand
side. It does this irrespective of whether the variable is a primitive variable or a
reference variable. In case of a primitive variable, the value happens to be the
actual primitive value and in case of a reference variable, the value happens to
be the address of an object. In both the cases, it is the value that is copied from
one variable to another. For this reason, it is also said that Java uses "pass by
value" semantics instead of "pass by reference" . We will revisit this later
when we discuss about passing variables to method calls.
This concept is very important and you will see many questions that require you
to have a clear understanding of it. The only thing that you need to remember is
that a variable, be it of any kind, contains just a simple raw number. Assigning
one variable to another simply copies that number from one variable to another.
It is the JVM's job to interpret what that number means based on the type of the
variable. Everything else just follows from this fundamental rule.
3.3 Declare and initialize variables

3.3.1 Declare and initialize variables ☝
For better or for worse, Java has several ways of declaring and initializing
variables. The exam expects that you know them all. Although Oracle have
substantially reduced the number of questions that are based solely on quirky
syntax, you may still see weird syntax used in code snippets in questions that test
you on something else.
So, let's go through them one by one starting with the most basic - declarations
without initialization.
1. int x;
String str;
Object obj;
2. int a, b, c; //a, b, and c are declared to be of t

ype int
String s1, s2; //s1 and s2 are declared to be type

String
The following are ways to declare as well as initialize at the same time:
1. int x = 10; //initializing x using an int literal

10
2. int y = x; //initializing y by assigning the value
of another variable x
3. String str = "123"; //initializing str by creating
a new String
4. SomeClass obj = new SomeClass(); //initializing obj
by creating a new instance of SomeClass
5. Object obj2 = obj; //initializing obj2 using
another reference
6. int a = 10, b = 20, c = 30; //initializing each
variable of same type with a different value
7. String s1 = "123", s2 = "hello";
8. int m = 20; int p = m = 10; //resetting m to 10 and
using the new value of m to initialize p
Mixing the two styles mentioned above:
1. int a, b = 10, c = 20; //a is declared but not

initialized. b and c are being declared as well as
initialized
2. String s1 = "123", s2; //Only s1 is being
initialized
And the following are some illegal ones:
1. int a = 10, int b; //You can have only one type

name in one statement.
2. int a, Object b; //You can have only one type name
in one statement.
3. int x = y = 10; //Invalid, y must be defined before
using it to initialize x.
Observe that there is no difference in the way you declare a primitive variables
and a reference variables. A reference variable, however, has one additional way
of initialization - you can assign null to a reference variable. You can't do that to
a primitive variable. For example, int i = null; is invalid. But String
s1 = null; is valid.
Naming rules for a variable

A variable name must be a valid Java identifier. Conventionally however, a
variable name starts with a lower case letter and names for constant variables are
in upper case. Variables created by code generation tools usually start with an
underscore or a dollar ( _ or $) sign.
3.3.2 Uninitialized variables and Default values ☝
Given just this statement - int i; - what will be the value of i?
If you are from C/C++ world, you may say that the value is indeterminate, i.e., i
may have any value. Java designers didn't like this undefined behavior of
uninitialized variables because it is a common source of bugs in applications. A
programmer may forget to initialize a variable and that may cause unintended
behavior in the application. Uninitialized variables don't serve any purpose
either. To use a variable, you have to assign it a value anyway. Then why leave
them uninitialized?
For this reason, Java designers simply outlawed the use of uninitialized variables
altogether in Java. In fact, they went even further and made sure that if a
programmer doesn't initialize a variable, the JVM initializes them with known
pre-determined values. Well, in most cases!
Try compiling the following code:

static int i;
int y;
public static void main(String[] name){
int p;
}
}
It compiles fine without any issues. It will run fine as well but will not produce
any output. Now, try the same code with a print statement that prints i and y .

static int i;
int y;
int p;
System.out.println(i+" "+new TestClass().y);
}
}
This also compiles fine. Upon running, it will print 0 0 . Now, try the following
code that tries to print p .

static int i;
int y;
int p;
System.out.println(p);
}
}
This doesn't compile. You will get an error message saying:
TestClass.java:6: error: variable p might not have bee

n initialized
System.out.println(p);
You can draw the following conclusions from this exercise:
1. Java doesn't have a problem if you have uninitialized variables as long as

you don't try to use them. That is why the first code compiles even though
the variables have not been initialized.
2. Java initializes static and instance variables to default values if you don't
initialize them explicitly. That is why the second code prints 0 0.
3. Java doesn't initialize local variables if you don't initialize them explicitly
and it will not let the code to compile if you try to use such a variable. That
is why the third code doesn't compile.
The first point is straightforward. If a variable is not used anywhere, you don't
have to initialize it. It is possible that a smart optimizing Java compiler may even
eliminate such a variable from the resulting class file.
Let us look at the second and third points now. To make sure that variables are
always initialized to specific predetermined values before they are accessed,
Java takes two different approaches .
The first approach is to let the JVM initialize the variables to predetermined
values on its own if the programmer doesn't give them any value explicitly. This
approach is taken for instance and static variables of a class. In this approach,
the JVM assigns 0 (or 0.0 ) to all numeric variables (i.e. byte, char, short, int,
long, float, and double), false to boolean variables, and null to reference
variables. These values are called the default values of variables. The following
code, therefore, prints 0 , 0.0 , false , and null .

static int i; //i is of numeric type and is therefo
re, initialized to 0
static double d; //d is a floating numeric type and

is therefore, initialized to 0.0
static boolean f; //f is of boolean type and is the

refore, initialized to false
static String s; //s is of reference type and is th

erefore, initialized to null

System.out.println(i);
System.out.println(d);
System.out.println(f);
System.out.println(s);
}
}
Observe that since s is a reference variable, it is initialized to null . You will

learn in the next chapter that an array is also an object, which means that an
array variable, irrespective of whether it refers to an array of primitives or
objects, is a reference variable, and is, therefore, treated the same way.
The above code uses only static variables. You will get the same result with
instance variables:

int i;
double d;
boolean f;
String s;
TestClass tc = new TestClass();
System.out.println(tc.i);
System.out.println(tc.d);
System.out.println(tc.f);
System.out.println(tc.s);
}
}
The second approach is to make the programmer explicitly initialize a variable

before the variable is accessed. In this approach, the compiler raises an error if it
finds that a variable may be accessed without being initialized. This approach is
used for local variables (i.e. variables defined in a method or a block).
Basically, the compiler acts as a cop that prevents you from using an
uninitialized variable. If at any point the compiler realizes that a variable may
not have been initialized before it is accessed, the compiler flags an error. This is
called the principle of "definite assignment ". It means that a local variable
must have a definitely assigned value when any access of its value occurs. For
example, the following code compiles fine because even though the variable
val is not initialized in the same line in which it is declared, it is definitely
assigned a value before it is accessed:

public static void main(String[] args) throws Exce
ption {
int val; //val not initialized here
val = 10;
System.out.println(val); //compiles fine
}
}
A compiler must perform flow analysis of the code to determine whether an

execution path exists in which a local variable is accessed without being
initialized. If such a path exists, it must refuse to compile the code. A compiler is
only allowed to consider the values of "constant expressions" in its flow
analysis. The Java language specification does formally define the phrase
"constant expression" but I will not go into it here because it is outside the scope
of the exam. The basic idea is that a compiler cannot execute code and so, it
cannot make any inferences based on the result of execution of the code. It has to
draw inferences based only on the information that is available at compile time.
It can take into account the value of a variable only if the variable is a compile
time constant . This is illustrated by the following code:

public static void main(String[] args) throws Exce
ption {
int val;
int i = 0; //LINE 4
if(i == 0){
val = 10;
}
System.out.println(val); //val may not be initia
lized
}
}
This code will not compile. Even though we know that i is 0 and so, i == 0
will always be true, the compiler doesn't know what the actual value of the
variable i will be at the time of execution because i is not a compile time
constant. Therefore, the compiler concludes that if the if condition evaluates to
false , the variable val will be left uninitialized. In other words, the compiler
notices one execution path in which the variable val will remain uninitialized
before it is accessed. That is why it refuses to accept the print statement. If you
change line 4 to final int i = 0; the compiler can take the value of i
into account in its flow analysis because i will now be a compile time constant.
The compiler can then draw the conclusion that i==0 will always be true, that
the if block will always be executed, and that val will definitely be assigned a
value before it is accessed.
Similarly, what if we add the else clause to the if statement as shown below?
int val;
int i = 0; //i is not final
if(i == 0){
val = 10;
}else{
val = 20;
}
System.out.println(val);
Now, i is still not a compile time constant but the compiler doesn't have to know
the value of i . If-else is one statement and the compiler is now sure that no
matter what the value of i is, val will definitely be assigned a value.
Therefore, it accepts the print statement.
Let us now change our if condition a bit.
if(i == 0){
val = 10;
}
if(i != 0){
val = 20;
}
It doesn't compile. It has the exact same problem that we saw in the first version.
We, by looking at the code, know that val will definitely be initialized in this
case. We know this only because we executed the code mentally. As far as the
compiler is concerned, these are two independent if statements. Since the
compiler cannot make inferences based on the results of execution of
expressions that are not compile time constants, it cannot accept the argument
that val will definitely be assigned a value before it is accessed in the print
statement.
In conclusion, Java initializes all static and instance variables of a class

automatically if you don't initialize them explicitly. You must initialize local
variables explicitly before they are used.
3.3.3 Assigning values to variables ☝
Java, like all languages, has its own rules regarding assigning values to
variables. The most basic way to assign a value to a variable is to use a "literal".
Literals ☝
A literal is a notation for representing a fixed value in source code. For example,
10 will always mean the number 10 . You cannot change its meaning or what it
represents to something else in Java. It has to be taken literally, and hence it is
called a literal. Since it represents a number, it is a numeric literal. Similarly,
true and false are literals that represent the two boolean values. 'a' is character
literal. "hello" is a string literal. The words String and name in the statement
String name; are not literals because Java does not have an inherent
understanding of these words. They are defined by a programmer.
The word int in int i; or the word for in for(int

i=0; i<5; i++); are kinda similar to literals
because they have a fixed meaning that is defined by
the Java language itself and not by a programmer.
They are actually a bit more than literals because they
tell the compiler to treat the following code in a
particular way. They form the instruction set for the
Java compiler using which you write a Java program
and are therefore, called "keywords".
Let me list a few important rules about literals:
1. To make it easy to read and comprehend large numbers, Java allows

underscores in numeric literals. For example, 1000000.0 can also be written
as 1_000_000.0 . You cannot start or end a literal with an underscore. You
can use multiple underscores consecutively. You need not worry about the
rules governing the usage of underscores in hexadecimal, binary, and octal
number formats.
2. A number without a decimal is considered an int literal , whereas a number

containing a decimal point is considered a double literal .
3. A long literal can be written by appending a lowercase or uppercase L to

the number and a float literal can be written by appending a lowercase or
uppercase f . For example, 1234L or 1234.0f .
4. A char literal can be written by enclosing a single character within single

quotes, for example, 'a' or 'A' . Since it may not always be possible to type
the character you want, Java allows you to write a char literal using the
hexadecimal Unicode character format ('\uxxxx '), where xxxx is the
hexadecimal value of the character as defined in unicode charset. For some
special characters, you can also use escape character \ . For example, a new
line character can be written as '\n'.
Note that writing character literals using a unicode or escape sequence is
not on the exam. I have presented this brief information only for the sake of
completeness.
5. There are only two boolean literals: true and false .
6. null is also a literal. It is used to set a reference variable to point to

nothing.
7. Java allows numeric values to be written in hexadecimal, octal, as well as

binary formats. In hexadecimal format (aka hex notation), the value must
start with a 0x or 0X and must follow with one or more hexadecimal
digits. For example, you could write 0xF instead of 15 . In octal format,
the number must start with a 0 and must follow with one or more ocal
digits. For example, 017 is 15 in octal. In binary format, the number must
start with a 0b or 0B and must follow with one of more binary digits (i.e.
zeros and ones).Understanding of these formats is not required for the exam
and so, I will not discuss these formats any further.
Assignment using another variable ☝
The second way to assign a value to a variable is to copy it from another

variable. For example, int i = j; or String zipCode = zip; or
Student topper = myStudent; are all examples of copying the value
that is contained in one variable to another. This works the same way for
primitive as well as reference variables. Recall from the "Object and Reference"
section that a reference variable simply contains a memory address and not the
object itself. Thus, when you assign one reference variable to another, you are
only copying the memory address stored in one variable to another. You are not
making a copy of the actual object referred to by the variable.
Assignment using return value of a method ☝
The third way to assign a value to a variable is to use the return value of a
method. For example, Student topper = findTopper(); or int
score = evaluate(); and so on.
Assigning value of one type to a variable of another type ☝
In all of the cases listed above, I showed you how to assign a value of one type
to a variable of the same type, i.e., an int value to an int variable or a Student
object to a Student variable. But it is possible to assign a value of one type to a
variable of another as well. This topic is too broad to be covered fully in this
chapter because the rules of such assignments touch upon multiple concepts. I
will cover them as and when appropriate. Let me list them here first:
1. Simple assignments involving primitive types - This includes the

assignment of compile time constants and the concept of casting for
primitive variables. I will discuss this topic next.
2. Primitive assignments involving mathematical/arithmetic operators - This

includes values generated using binary operators as well as compound
operators, and the concept of implicit widening and narrowing of
primitives. I will discuss this topic in the "Using Operators" chapter.
3. Assignments involving reference types - This expands the scope of casting

to reference types. I will discuss this in the "Reusing Implementations
Through Inheritance" chapter.
Primitive assignment ☝
If the type of the value can fit into the type of the variable, then no special
treatment is required. For example, you know that the size of a byte (8 bits) is
smaller than the size of an int (32 bits) and a byte can therefore, fit easily into
an int . Thus, you can simply assign a byte value to an int variable. Here
are a few similar examples:
byte b = 10; //b is 8 bits
char c = 'x'; //c is 16 bits
short s = 300; //c is 16 bits
int i; //i is 32 bits
long l; //l is 64 bits
float f; //f is 32 bits

double d; //d is 64 bits
//no special care is needed for any of the assignments

below
i = b;
i = s;
l = i;
f = i;
d = f;
//observe that the type of the target variable is larg
er than the type of the source variable in all of the
assignments above.
Assigning a smaller type to a larger type is known as "widening conversion" .

Since there is no cast required for such an assignment, it can also be called
"implicit widening conversion" . It is analogous to transferring water from one
bucket to another. If your source bucket is smaller in size than the target bucket,
then you can always transfer all the water from the smaller bucket to the larger
bucket without any spillage.
What if the source type is larger than the target type? Picture the bucket
analogy again, what will happen if you transfer all the water from the larger
bucket to the smaller one? Simple! There may be spillage :) Similarly, when you
assign a value of a larger type to a variable of a smaller type, there may be a loss
of information. The Java compiler does not like that. Therefore, in general, it
does not allow you to assign a value of a type that is larger than the type of the
target variable. Thus, the following lines will cause a compilation error:
//assuming variable declarations specified above
c = i;
i = l;
b = i;
f = d;
//observe that the type of the target variable on the
left is smaller than the type of the source variable(o
n the right) in all of the assignments above.
But what if the larger bucket is not really full? What if the larger bucket has only
as much water as can be held in the smaller bucket? There will be no spillage in
this case. It follows then that the compiler should allow you to assign a variable
of larger type to the variable of a smaller type if the actual value held by the
source value can fit into the target value. It does, but with a condition.
The problem here is that the compiler does not execute any code and therefore, it
cannot determine the actual value held by the source variable unless that variable
is a compile time constant. For example, recall that the number 10 is actually an
int literal . It is not a byte but an int . Thus, even though an int is larger
than a byte, byte b = 10; will compile fine because the value 10 can fit into
a byte . But byte b = 128; will not compile because a byte can only
store values from -128 to 127 . 128 is too large to be held by a byte .
Similarly, final int i = 10; byte b = i; will also compile fine

because i is now a compile time constant. Being a compile time constant, i's
value is known to the compiler and since that value is small enough to fit into a
byte, the compiler approves the assignment.
Thus, you can assign a source variable that is a compile time constant to a target
variable of different type if the value held by source variable fits into the target
variable. This is called "implicit narrowing" . The compiler automatically
narrows the value down to a smaller type if it sees that the value can fit into the
smaller type. The compiler does this only for assignments and not for method
calls. For example, if you have a method that takes a short and if you try to
pass an int to this method, then the method call will not compile even if the
value being passed is small enough to fit into a short.
What if the source variable is not a constant? Since the compiler cannot
determine the value held by the variable at run time, it forces the programmer to
make a promise that the actual value held by the source variable at run time will
fit into the target variable. This promise is in the form of a "cast ". Java allows
you to cast the value of one primitive type to another primitive type by
specifying the target type within brackets. For example, int i = (int)
11.1; Here, I am casting the floating point value 1.1 to an int . You can
use a cast to assign any primitive integral (i.e. byte , char , short , int ,
long ) or floating point type (i.e. float and double ) value to any integral or
floating point variable. You cannot cast a boolean value to any other type or
vice versa.
Here are a few more examples of assignments that can be done successfully with
casting:
int i = 10;
char c = (char) i; //explicitly casting i to char
long l = 100;
i = (int) l; //explicitly casting l to int
byte b = (byte) i; //explicitly casting i to byte
double d = 10.0;
float f = (float) d; //explicitly casting d to float
A cast tells the compiler to just assign the value and to not worry about any
spillage. This is also known as "explicit narrowing ".
But what will happen if there is spillage?, i.e., what will happen if the actual
value held by the source variable is indeed larger than the size of the target
variable? What will happen to the extra value that can't fit into the target? For
example, what will happen in this case - int i = 128; byte b =
(byte) i; ? The explicit cast should simply assign the value that can fit into
the target variable and throw away the extra. Thus, it should just assign 127 to b
and ignore the rest, right? Wrong! If you print the value of b , you will see -128
instead of 127 . There doesn't seem to be any relation between 127 and -128 !
Understanding why this happens is not required for the exam. You will not be
asked about the values assigned to variables in such cases. But I will discuss it
briefly because it is useful to know.
Casting of primitives is pretty much like shoving an object of one shape into a
mould of another shape. It may cause some parts of the original shape to be cut
off. To understand this, you need to look at the bit patterns of int i and byte b
. The size of i is 32 bits and the value that it holds is 128, therefore, its bit
pattern is: 00000000 00000000 00000000 10000000 . Since you are
now shoving it into a byte , which is of only 8 bits, the JVM will simply cut out
the extra higher order bits that can't fit into a byte and assign the lowest order 8
bits, i.e., 10000000 to b. Thus, b's bit pattern is 10000000 . Since byte is a
signed integer, the topmost bit is the sign bit (1 means, it is a negative number).
Since negative numbers are stored in two's complement form, this number is
actually -128 (and not -0 !). This process happens in all the cases where the
target is smaller than the source or has a different range than the source. As you
can see, determining the value that will actually be assigned to the target variable
is not a simple task for a human. It is, in fact, a common source of bugs. This is
exactly why Java doesn't allow you to assign just about any value to any variable
very easily. By making you explicitly cast the source value to the target type, it
tries to bring to your attention the potential problems that it might create in your
business logic. You should, therefore, be very careful with casting.
Assigning short or byte to char ☝
As you know, the sizes of short and char are same, i.e., 16 bits. The size of
int and float are also the same, i.e., 32 bits. Thus, it should be possible to
assign a short to a char and a float to an int without any problem.
However, remember that a char is unsigned while a short is not. So, even
though their sizes are the same, their ranges are different. A char can store
values from 0 to 65535 , while a short can store values from -32768 to
32767 . Thus, it is possible to lose information while making such assignments.
Similarly, you cannot assign a byte to a char either because even
though byte (8 bits) is a smaller type than char , char cannot hold negative
values while byte can.
Here are a few examples that make this clear:
char c1 = '\u0061'; //ok, unicode for 'a'

short s1 = '\u0061'; //ok, no cast needed because '\u0
061' is a compile time constant that can fit into a sh
ort.
short s2 = c1; //will not compile - c1 is not a compil

e time constant, explicit cast is required.
char c2 = '\uFEF0'; //ok, unicode for some character.
short s2 = '\uFEF0'; //will not compile, value is beyo

nd the range of short.
short s3 = (short) '\uFEF0'; //ok because explicit cas

t is present.
char c3 = 1; //ok, even though 1 is an int but it is a

compile time constant whose value can fit into a char.
char c4 = -1; //will not compile because -1 cannot fit

into a char
short s4 = -1;
char c5 = (char) s4; //ok because explicit cast is pre
sent
Assigning float to int or double to long and vice-versa ☝

The same thing happens in the case of int and float and long and double
. Even though they are of same sizes their ranges are different. int and long
store precise integral values while float and double don't. Therefore, Java
requires an explicit cast when you assign a float to a int or a double to a
long .
The reverse, however, is a different story. Although float and double also do
lose information when you assign an int or a long to them respectively, Java
allows such assignments without a cast nonetheless. In other words, Java allows
implicit widening of int and long to float and double respectively.
Here are a few examples that make this clear:
int i = 2147483647; //Integer.MAX_VALUE
float f = i; //loses precision but ok, implicit wideni

ng of int to float is allowed
long g = 9223372036854775807L; //Long.MAX_VALUE;
double d = g; //loses precision but ok, implicit widen

ing of long to double is allowed
i = f; //will not compile, implicit narrowing of float

to int is NOT allowed
g = d; //will not compile, implicit narrowing of doubl

e to long is NOT allowed
You can, of course, assign a float or a double to an int or a long using an

explicit cast.
3.3.4 final variables ☝
A final variable is a variable whose value doesn't change once it has had a
value assigned to it. In other words, the variable is a constant. Any variable can
be made final by applying the keyword final to its declaration. For example:
class TestClass{
final int x = 10;
final static int y = 20;
public static void main(final String[] args){
final TestClass tc = new TestClass();

//x = 30; //will not compile.
//y = 40; //will not compile.
//args = new String[0]; //will not compile
//tc = new TestClass(); //will not compile
System.out.println(tc.x+" "+y+" "+args+" "+tc);

}
}
Observe that in the above code, I have made an instance variable, a static
variable, a method parameter, and a local variable final. It prints 10 20
[Ljava.lang.String;@52d1fadb TestClass@35810a60 when
compiled and run.
You cannot reassign any value to a final variable, therefore, the four statements
that try to modify their values won't compile.
Remember that when you make a reference variable final, it only means that the
reference variable cannot refer to any other object. It doesn't mean that the
contents of that object can't change. For example, consider the following code:
class Data{
int x = 10;
}
final Data d = new Data();
//d = new Data(); //won't compile because d is f

inal
d.x = 20; //this is fine because we are not chan

ging d here.
}
}
In the above code, we cannot make d refer to a different Data object once it is
initialized because d is final, however, we can certainly use d to manipulate the
Data object to which it points. If you have any confusion about this point, go
through the "Object and Reference" section in "Kickstarter for Beginners"
chapter.
There are several rules about the initialization of final variables but they depend
on the knowledge of initializers and constructors. I will revisit this topic in the
"Reusing Implementations Through Inheritance" chapter.
3.4 Wrapper Classes ☝

Wrapper classes were on the OCAJP 8 exam but, surprisingly, they have been
removed from the OCP Java 11 Part 1 exam. You may, therefore, ignore the
following discussion if you are short on time. However, it is almost necessary for
a Java developer to have a basic understanding of wrapper classes because they
are widely used.
3.4.1 What are wrapper Classes? ☝
Before I start talking about wrapper classes, let me recall three topics that are
quite fundamental to understanding wrapper classes - 1. Object and Reference, 2.
Stack and Heap, and 3. Difference between reference variables and primitive
variables. I suggest you go through them first if you are not clear on those three
concepts.
As you are aware, Java is considered an object-oriented language. Pretty much

everything in Java is about objects. You also know that all objects reside on the
heap space and can only be accessed through their references, which are cached
in reference variables. This is all good but from a performance perspective, heap
space is bit heavy as compared to stack space. It is also more permanent than
stack space. Recall that JVM performs garbage collection for getting rid of
unused objects from time to time. For many simple programming activities such
as loops, decision constructs, and temporary data storage, objects seem like too
much of a hassle.
To address this concern, Java has the provision of "primitive" data types. They
are called primitives because they don't have any behavior associated with them.
They are just raw data. If you want to run a loop 10 times, all you need is a
counter that can count from 0 to 9. The counter doesn't serve any purpose after
the loop is over and so, you don't need any permanent storage to store the
counter value. A simple int variable that is forgotten as soon as the loop is done,
is sufficient for this purpose. The stack space is perfect for keeping such values
because a stack space is wiped clean as soon as the thread that owns that stack
space is finished. There is no need of any garbage collection to happen here.
Furthermore, a program can directly manipulate the value of a primitive variable
without having to go through the indirection of a reference.
The problem with having two different kinds of data types is that it creates a
dichotomy between data represented by primitives and data represented by
objects. Due to the difference between how they are stored, passed, and
accessed, a program cannot treat primitives and objects the same way. For
example, let's say you have developed some logic to process different types of
data and you've captured this logic into a method. What type of input parameters
would you use to pass the data to this method? If you decide to use Object, you
cannot pass any primitive to this method and if you choose a primitive, you
cannot pass an object to it. You will either need to write different methods for
each kind or have the caller wrap primitive data into objects and pass the objects
to your method.
A more concrete example would be a class that manages a collection of data.

Java has several standard classes for managing collections. The one most
commonly used, and which is also included in the exam objectives, is
java.util.ArrayList . I will discuss this class in detail later, but
basically, you use an ArrayList to collect a bunch of objects. If you have a
primitive value that you want to keep in the collection managed by an
ArrayList , you need to wrap that value into an object because ArrayList
only works with objects.
Java designers realized this problem and added ready-made wrapper classes for
each of the primitive data types to Java core library. These classes are - Byte,
Character, Short, Integer, Long, Float, Double, and
Boolean . All of these classes are in java.lang package.
Wrapper classes meant for the types that are used to represent numeric data, i.e.,
Byte, Short, Integer, Long, Float, and Double , extend from a
common base class named java.lang.Number .
Further, Number, Character, and Boolean , classes extend from

java.lang.Object .
3.4.2 Creating wrapper objects ☝
There are three ways to create objects of wrapper classes. Let us look at them
one by one.
1. Using constructors - Like any other object, objects of wrapper classes can
be created using their constructors. For example,
Integer i1 = new Integer(10);

Integer i2 = new Integer("10");
Character c = new Character('c');
Boolean b1 = new Boolean(true);

Boolean b2 = new Boolean("true");
float f = 10.2f;
Float f1 = new Float(f);
Float f2 = new Float(10.2); //valid, even though 1
0.2 is a double
Float f3 = new Float("10.2");
short s = 10;
Short s1 = new Short(s);
Short s2 = new Short("10");
Short s3 = new Short(10); //this will not compile.
Observe that in the above code, I have created instances of wrapper classes
using primitive values as well as String values. All wrapper classes
except java.lang.Character provide two constructors each - one
that takes the relevant primitive type and the second one that takes a
String . java.lang.Character class provides only the char
constructor and does not provide the String constructor. Furthermore,
java.lang.Float class provides a third constructor that takes a
double , which is why the line Float f2 = new Float(10.2);
works even though the literal 10.2 is a double .
Further observe that Short s3 = new Short(10); will not compile

because 10 is an int and, as explained above, Short does not have any
constructor that takes as int . It has only two constructors - one that takes
a String and one that takes a short . (Recall that "implicit narrowing"
does not happen for method or constructor arguments.)
2. Using the valueOf methods - All wrapper classes have two static
valueOf methods each. One that takes the relevant type as a parameter,
and the second one that takes a String . Character class is an
exception because it has only one valueOf method that takes a char .
Unlike constructors, Float class does not have a third valueOf method
that takes a double ! Here are a few examples:
float f = 10.2f;
Float f1 = Float.valueOf(f);
Float f2 = Float.valueOf("10.2");
Float f2 = Float.valueOf(10.2);//will not compile

because 10.2 is a double
Integer i1 = Integer.valueOf(10);
Byte b = Byte.valueOf("10");
Boolean bool = Boolean.valueOf("true");
In the case of the String versions of valueOf methods (and also of the
constructors), you have to be careful about the value that you pass to the
method because if you pass a null or a value that cannot be converted into
the required wrapper type, a NumberFormatException will be thrown.
For example, Integer i = Integer.valueOf("10.2"); will
throw a NumberFormatException because 10.2 cannot be parsed
into an integer.
It is interesting to know that you can pass null or any string to Boolean
's constructor or valueOf method without any exception. All such values
will cause it to create a Boolean object containing false . To get a
Boolean containing true , you can pass "true" any case (upper, lower,
or even mixed).
So, what's the point of having valueOf methods if you can do the same
thing using constructors?
Well, the difference between the two is that a constructor will always create
a new object, while the valueOf method may return a cached object. The
valueOf methods are therefore, more efficient than the constructors and
should be used when you don't need to use separate wrapper objects for the
same value.
3. Through auto-boxing - Up until Java 1.4, every time you wanted to use a
wrapper object, you would have had to either use a constructor or a
valueOf method to create it. Since wrapper objects are used quite often,
this was considered too much of a typing effort for such a mundane thing.
To make things a little easier and cleaner, Java 1.5 introduced the concept
of "auto-boxing". All it means is that if you assign a primitive value to a
wrapper variable, the compiler will automatically box the primitive value
into a wrapper object. So, basically, instead of writing Integer i =
Integer.valueOf(100); you can just write Integer i = 100; .
Similarly, if a method expects an object as an input parameter, you can just

pass in the primitive and the compiler will automatically box it into a
wrapper object. For example, instead of writing
myList.add(Integer.valueOf(100)); , you can write,
myList.add(100); .
Observe that I have used the valueOf method instead of a constructor to

illustrate the equivalency of the explicit creation of wrapper objects and
autoboxing. This is deliberate. The reason is that autoboxing for byte ,
short , char , int , long , and boolean uses cached objects instead
of creating new instances just like the valueOf methods. Thus, i1 and i2
in the following code will refer to the same Integer wrapper object:
Integer i1 = 100;
Integer i2 = 100; //i2 will refer to the same
object as i1.
Just like the valueOf methods, autoboxing of values from -128 to 127 ,
true and false , and '\u0000' to '\u007f' will result in cached
objects. Wrappers for other values may also be cached but that is not
guaranteed.
3.4.3 Converting wrapper objects to primitives ☝
There are two ways to get primitive values from wrapper objects:
1. Using xxxValue methods - All wrapper classes provide an instance

method that returns the value wrapped by that wrapper object as a
primitive. The name of this method follows the pattern <type>Value .
For example, Integer class has intValue , Boolean class has
booleanValue , and Character class has charValue that
returns int , boolean , and char respectively.
Recall that wrapper classes for numeric types (i.e. Byte, Short,
Integer, Long, Float, and Double ) have a common base class
called Number . This class defines byteValue and shortValue
methods and declares intValue, longValue, floatValue, and
doubleValue methods. Therefore, it is possible to get a primitive value
of any of these types from any of numeric wrapper class objects.
2. Through unboxing - This is just the opposite of autoboxing. You can

assign (or pass as an argument) any wrapper object to a variable of
primitive type directly and the compiler will automatically extract the
primitive value from it and assign it to the target. For example,
Integer i1 = 10; //autoboxing int value 10 into an

Integer object.
int i2 = i1; //unboxing Integer object and
assigning the resulting value to an int variable.
Remember that unboxing will compile only if the type of the target is wide
enough to accept the type of the wrapper. For example, byte b = i1;
will not compile because int is wider than byte , while float f =
i1; will compile fine because float is wide enough to hold any int
value.
Besides letting you convert wrapper objects to primitives, wrapper classes also
contain parseXxx methods that let you get primitive values from Strings. For
example, Integer has parseInt method that takes in a String as an
argument and returns an int . Of course, the argument must have a valid value
that can be parsed as an int otherwise a NumberFormatException will be
thrown. There are several variations of this method but you need not memorize
them for the exam.
Wrapper classes contain many useful methods and even though they are not on
the exam, I suggest you quickly browse through their API descriptions, as that
will help you on the job.
3.5 Exercise ☝
1. Identify all the primitive and reference data types as well as primitive and
reference variables used in the following code:
public class Person {

int id;
String name;
java.util.Date dob;
boolean VIP;
Person p1 = new Person();
p2 = p1;
int id = p2.id;
p1.name = args[0];
}
public String getName(){ return name; }
}
2. What are the values of the primitive variables used in the above code. What
operations can be performed on these variables?
3. Identify the methods that can be invoked using the reference variables used
in the above code and also identify the objects on which those methods will
be invoked.
4. Change the declaration of various instance members of Person class to
include initial values.
Chapter 4 Describing and Using
Objects and Classes
Declare and instantiate Java objects
Explain objects' lifecycles (including creation, dereferencing by
reassignment, and garbage collection)
4.1 Declare and instantiate Java objects

4.1.1 Declare and instantiate Java objects ☝
In the previous chapter you saw the difference between primitive variables and
reference variables and how to declare both kind of variables. The basic syntax
of declaring primitive and reference variables is the same:
<type name> <variable name>;
In case of reference variables, the type name can be a simple name or a FQCN
depending on whether the class (or its package) has been imported using an
appropriate import statement or not. For example:
String str;
Object obj;
SomeClass scRef; //assumming com.abc.SomeClass has bee
n imported
com.xyz.SomeOtherClass socRef; //assumming com.xyz.Som

eClass hasn't been imported
You can declare multiple variables of the same type in a single statement:
SomeClass scRef1, scRef2;
As with primitive variables, it is possible to assign a value to a reference variable

at the time of declaration itself:
SomeClass scRef1 = null; //scRef1 is initialized with

null.
SomeClass scRef2 = scRef1, scRef3; //scRef2 is initial

ized with the same value as scRef1 but scRef3 remains
uninitialized.
Object a = null, b = a; //a is initialized to null and

then a is assigned to b
Here are a couple of invalid declarations:
String str1, Object o1; //can't change the type within

a statement
Object a = b, b = null; //b is being assigned to a but

b hasn't been defined at that point
You will not get overly tricky questions on declaration of variables in

the exam. If you stick to the basics shown above, you will not have
trouble answering exam questions.
Instantiating objects ☝
As discussed in the previous chapter, a class defines a new data type. You define
data types because they allow you to group raw data and give that group a
special meaning in your application. For example, if you are dealing with names,
dates of birth and addresses of people in your application, you may group the
three raw data elements into a Person class and add behavior to Person class in
terms of methods. You would then have several "persons" in your application.
Each person will be modeled upon the same Person template. Every person
would actually be just an instance of the Person class. So, two persons may have
different names and other details but they will exhibit the same behavior. This
commonality of behavior allows your application to treat all person instances in
the same manner.
Once you have defined the template, it is easy to construct objects. Java has
exactly three ways of instantiating objects - using the new keyword, by
deserializing an object's data, and by cloning an existing object. Here, we will
only focus on the new keyword because deserialization and cloning are not on
the Part 1 exam.
The following is probably the simplest example of instantiating an object using

the new keyword:
new java.lang.Object();
The above statement creates an instance of java.lang.Object class. Since

a class is just a template...a cookie cutter, if you will, which is used to cut objects
out of free memory, when you use the new keyword on a class, the JVM takes a
chunk of free memory and formats that memory into the various fields defined in
the class. This formatted chunk of memory becomes an object of the class. You
can create as many instances as you want using this template.
Depending on how a class has been defined, you may be able to (or even
required to) pass argument(s) while instantiating an object. For example, the
following statement creates an instance of String class with a string parameter:
new String("hello");
The new keyword actually causes a "constructor" of the class to execute. The
constructor is responsible for initializing the data members of the instance. You
will see the exact mechanism of object creation in the "Creating and using
Methods" chapter.
As discussed earlier, a reference variable doesn't contain the actual object but
just the address (or reference) to the object. This address is available only at the
time of instantiation of an object and if you want to make use of that object later
on, you must save that address somewhere. Usually, you store it in a variable,
like this:
String str = new String("1234");
Since the address of the String object is saved in str variable, you can invoke
methods on this object using the str variable. If you don't save the address of
the object in a variable that you can access, that object will be lost after creation.
In some cases, you may want exactly that. You may want to just create an object
and forget about it after that. In such a case, there is no need to save its address
in a variable.
You may see the term "instantiate a class" being used

in articles and books instead of "instantiate an object".
Both the terms mean the same thing and are equally
acceptable. Since official exam objectives use the term
"instantiate an object", I have used the same.
Similarly, the words "create" and "instantiate" are also
used interchangeably in the context of an object. So,
"create an object" and "instantiate an object" mean the
same thing.
Assigning objects to reference variables ☝

Since Java is a strongly typed language, you cannot assign just about any object
to a reference variable of any type. The rule regarding object assignment is
actually quite simple but requires an understanding of inheritance and
polymorphism. Since I haven't discussed these concepts yet, assume that if the
type of the variable is A , then only an object of type A or a sub-type of A can be
assigned to that variable. For example, Object obj = new String(); is
valid because String is a subtype of Object but String str = new
Object(); is not valid. In later chapters, you will see how this rule applies to
interfaces, abstract classes, and variables.
4.2 Read or write to object fields

4.2.1 Accessing object fields ☝
By object fields, we mean instance variables of a class. Each instance of a class

gets its own personal copy of these variables. Thus, each instance can potentially
have different values for these variables. To access these variables, i.e., to read
the values of these variables or to set these variables to a particular value, we
must know the exact instance whose variables we want to manipulate. We must
have a reference pointing to that exact object to be able to manipulate that
object's fields.
Recall our previous discussion about how an object resides in memory and a
reference variable is just a way to address that object. To access the contents of
an object or to perform operations on that object, you need to first identify that
object to the JVM. A reference variable does just that. It tells the JVM which
object you want to deal with.
For example, consider the following code:
class Student{
String name;
}
Student s1 = new Student();
Student s2 = new Student();
s1.name = "alice";
System.out.println(s1.name); //prints alice
System.out.println(s2.name); //prints null
s2.name = "bob";
System.out.println(s1.name); //prints alice
System.out.println(s2.name); //prints bob
}
}
In the above code, we created two Student objects. We then set the name
variable of one Student object and print names of both the Student objects. As
expected, the name of the second Student object is printed as null. This is
because we never set the second Student object's name variable to anything. The
JVM gave it a default value of null .
Next, we set the name variable of the second Student object and print both the
values again. This time we are able to see the two values stored separately in two
Student instances.
This simple exercise shows how to manipulate fields of an object. We take a

reference variable and apply the dot operator and the name of the variable to
reach that field of the object pointed to by that variable. It is not possible to
access the fields of an object if you do not have a reference to that object. You
may store the reference to an object in a variable (such as s1 and s2 in the code
above) when that object is created and then pass that reference around as needed.
Sometimes you may not want to keep a reference in a variable. This typically
happens when you want to create an object to call a method on it just once. The
following piece of code illustrates this:
class Calculator{
public int calculate(int[] iArray){
int sum = 0;
for(int i : iArray){ //this is a for-each loop,
we'll cover it later
sum = sum+i;
}
return sum;
}
}
int result = new Calculator().calculate( new
int[]{1, 2, 3, 4, 5} );
System.out.println(result);
}
}
Observe that we created two objects in the main method but did not store their
references anywhere - a Calculator object and an array object. Then we
called an instance method on the Calculator object directly without having a
reference variable. Since the method call is chained directly to the object
creation, the compiler is able to create a temporary reference variable pointing to
the newly created object and invoke the method using that variable. However,
this variable is not visible to the programmer and therefore, after this line, we
have lost the reference to the Calculator object and there is no way we can
access the same Calculator object again.
Within the calculate method, the same Calculator object is available

though, through a special variable called "this ", which is the topic of the next
section.
Similarly, the compiler created a temporary reference variable for the array
object and passed it in the method call. However, we don't have any reference to
this array object after this line and so, we cannot access it anymore. Within the
calculate method, however, a reference to that array object is available through
the method parameter iArray .
4.2.2 What is "this"? ☝
Let us modify our Student class a bit:
class Student{
String name;

Student s1 = new Student(); //1
s1.name = "mitchell"; //2
s1.printName(); //3 prints mitchell
public void printName(){

System.out.println(name); //5
I mentioned earlier that it is not possible to access the fields of an object without
having a reference to that object. But at //5, we are not using any such reference.
How is that possible? How is the JVM supposed to know which Student
instance we mean here?
Observe that at //3, we are calling the printName method using the reference
variable s1 . Therefore, when the JVM invokes this method, it already knows
the instance on which it is invoking the method. It is the same instance that is
being pointed to by s1. Now, in Java, if you don't specify any reference variable
explicitly within any instance method, the JVM assumes that you mean to access
the same object for which the method has been invoked. Thus, within the
printName method, the JVM determines that it needs to access the name field
of the same Student instance for which printName method has been
invoked. You can also explicitly use this reference to the same object by using
the keyword "this" . For example, //5 can be written as:
System.out.println(this.name);
Thus, the rule about having a reference to access the instance fields of an object
still applies. Java supplies the reference on its own if you don't specify it
explicitly.
By automatically assuming the existence of the

reference variable "this" while accessing a member of
an object, Java saves you a few keystrokes. However,
it is not considered a good practice to omit it. You
should always type "this." even when you know that you
are accessing the field of the same object because it
improves code readability. The usage of "this." makes
the intention of the code very clear and easy to
understand.
When is "this" necessary? ☝
When you have more than one variable with the same name accessible in code,
you may have to remove the ambiguity by using the this reference explicitly.
This typically happens in constructors and setter methods of a class. For
example,
class Student{
String id;
String name;
public void setName(String code, String name){
id = code;
name = name;
}
}
In the setName method, four variables are accessible: two method parameters -
code and name , and two instance variables - id and name . Now, within the
method code, when you do id = code; the compiler knows that you are
assigning the value of the method parameter code to the instance field id
because these names refer to exactly one variable each. But when you do name
= name; , the compiler cannot distinguish between the two name variables. It
thinks that name refers to the method parameter and assigns the value of the
method parameter to itself, which is basically redundant and is not what you
want. There is nothing wrong with it from the compiler's perspective but from a
logical perspective, the above code has a serious bug. In technical terms, this is
called "shadowing" . A variable defined in a method (i.e. either in parameter list
or as a local variable) shadows an instance or a static field of that class. It is not
possible to access the shadowed variable directly using a simple name. The
compiler needs more information from the programmer to disambiguate the
name.
To fix this, you must tell the compiler that the name on left-hand side of =
should refer to the instance field of the Student instance. This is done using
"this" , i.e., this.name = name; .
While we are on the topic of shadowing, I may as well talk about shadowing of
static variables of a class by local variables. Here is an example:
class Student{
static int count = 0;
public void doSomething(){
int count = 10;
count = count; //technically valid but logicall
y incorrect
Student.count = count; //works fine in instance

as well as static methods
this.count = count; //works fine in an instance
method
}
}
The above code has the same problem of redundant assignment. The local
variable named count shadows the static variable by the same name. Thus,
inside doSomething() , the simple name count will always refer to the
local variable and not to the static variable. To disambiguate count , ideally,
you should use Student.count if you want to refer to the static variable but
if you are trying to use it from an instance method, you can also use
this.count . Yes, using this is a horrible way to access a static variable
but it is permissible. I will talk more about static fields and methods in the
"Creating and Using Methods" chapter.
Redundant assignment is one of the traps that you will encounter in

the exam. Most IDEs flash a warning when you try to assign the
value of a variable to the same variable. But in the exam, you won't
get an IDE and so, you must watch out for it by reading the code
carefully.
Here are a few quick facts about this :
1. this is a keyword. That means you can't use it for naming anything such
as a variable or a method.
2. The type of this is the class (or an enum) in which it is used. For
example, the type of this in the printName method of Student class
is Student .
3. this is just like any other local variable that is set to point to the instance
on which a method is being invoked. You can copy it to another variable.
For example, you can do Student s3 = this; in an instance method
of Student class.
4. You can't modify this , i.e., you can't set it to null or make it point to
some other instance. It is set by the JVM. In that sense, it is final.
5. this can only be used within the context of an instance of a class. This
means, it is available in instance initializer blocks, constructors, instance
methods, and also within a class. It is not available within a static method
and a static block because static methods (and static initializer blocks) do
not belong to an object.
4.3 Define the structure of a Java class

4.3.1 Class disambiguated ☝
The word "class" may mean multiple things. It could refer to the OOP meaning
of class, i.e., an abstraction of an entity, it could refer to the code written in a
Java source file, or it could refer to the output of the Java compiler, i.e., a file
with .class extension.
For example, let us say you are developing an application for a school. You
could model the Student entity as a class. In this case, Student is a class in the
OOP sense. When you actually start coding your application, you would write
the code for Student class in Student.java file. Finally, you would compile
Student.java using javac and produce Student.class file which contains the
bytecode for the Student class.
The exam focuses primarily on the source code aspect of a class, i.e., the
contents of Student.java file of the above example. However, you do need to
know the basics of OOP as well because, after all, a Java source file is meant to
let you write code for your OOP class model. You don't have to worry about the
bytecode version of a class.
4.3.2 Structure of a Java source file ☝

If you have written some Java code before, you are already aware of the basic
structure of a Java source file. I will do a quick recap and then move on to the
interesting situations and gotchas that you need to know for the exam.
A Java source file has the following three parts:
Part 1: zero or one package statement

Part 2: zero or more import statements
Part 3: one or more reference type (i.e. class, interface, or enum) definitions.
The ordering mentioned above is important. For example, you cannot have the
package statement after the import statements or the class declaration(s).
Similarly, you cannot have import statements after the class declaration.
A Java source file must define at least one Java reference type definition in it.
You can define multiple reference types within a single source file as well. I will
talk about the rules of that later.
4.3.3 Members of a class ☝
Within a class definition, you can have field declarations, methods,

constructors, and initializers . You can also have classes, interfaces, and enums,
but more on that later. All of these are called "members" of that class.
Members can be defined as static or non-static aka instance (a member that is

not defined as static is automatically non-static).
For example, the following code shows various members of a class:
//package com.school; //optional
import java.util.Date;
//required because we are using Date instead of java.u
til.Date in the code below
public class Student
{
private static int count = 0; //static field
private String studentId; //instance field

static{ //static initializer
System.out.println("Student class loaded");

}
{ //instance initializer
Student.count = Student.count +1;

System.out.println("Student count incremented"
);
}
public Student(String id){ //constructor
this.studentId = id;
System.out.println(
new Date() +
" Student instance created. Total student
s created = "+count);
}
public String toString(){ //instance method

return "Student[studentId = "+studentId+"]";
}
public static void main(String[] args) { //static

method
Student s = new Student("A1234");

System.out.println(s.toString());
}
The package statement at the top makes the Student class a member of
com.school package. The import statement lets you use Date class of java.util
package in the code by typing just Date instead of java.util.Date .
The class uses a static field named count to track the number of Student objects
that have been created. Instance field studentId stores an id for each Student
instance.
The static initializer is executed only once when the class is first loaded by the
JVM and the instance initializer is executed just before the constructor every
time an instance is created. Don't worry if you don't understand the purpose of
static and instance initializer blocks. We will go deep into this later.
Then there is a constructor that allows you to create Student objects with a given
id and the static main method that allows you to execute this class from
command line. (Notice that I have commented out the package statement so that
it will be easier to execute the class from command line without worrying about
the directory structure.)
The following output is produced upon executing this class:
Student class loaded

Student count incremented
Mon Jul 31 09:35:19 EST 2017 Student instance created.
Total students created = 1
Student[studentId = A1234]
Important - You cannot have any statement in a class that does not belong to
any of the categories specified above. For example, the following will not
compile:

String id = ""; //this is ok because this statement
is a declaration
id = "test"; //this is not ok because this is a simp

le statement that is not a declaration, or an initiali
zer block, or a method, or a constructor.
{ //this is ok because it is an initializer block
id = "test"; //this is ok because it is inside an

instance initializer block and not directly within the
class.
}
}
Comments ☝
Java source files can also contain comments. You will not get questions on
comments in the exam but it good to know that there are two ways to write
comments in a Java source file - a single line comment, which starts with a //
and closes automatically at the end of the line (that means you don't close it
explicitly) closing and a multi line comment, which opens with /* and closes
with */ . Multi line comments don't nest. Meaning, the comment will start with a
/* and end as soon as the first */ is encountered.
Comments are completely ignored by the compiler and have no impact on the
resulting class file.
The following are a few examples:
//this is a short comment
/*
This is a multi line
comment.
*/
/*
This is a multi line
comment.
//This is another line inside a comment
*/
JavaDoc Comments ☝
Java promotes writing well documented code. It allows you to write descriptions
for fields, methods, and constructors of a class through smart use of comments.
If you write comments in a certain format, you can produce HTML
documentation for your code using the JavaDoc tool. This format is called the
JavaDoc comment format and it looks like this:
/**
* Observe the start of the comment. It has an extra
* Each line starts with a *
* There is a space after each *
* <h3>You can write HTML tags here.</h3>
* Description of each parameter starts with @param
* Description of the return value starts with @return
* @see tag is used to add a hyperlink to description
of another class
* @param name the location of the image, relative to
the url argument
* @return the image at the specified URL
* @see SomeOtherClassName
*/
public String sayHello(String name) {

return "Hello, "+name;
}
The JavaDoc tool comes bundled with the JDK. It can extract all the information
contained in the comments written in the above format and generate nicely
formatted HTML documentation. In fact, all of the standard Java library classes
contain descriptions in the above format. It is these descriptions that are used to
generate the HTML pages of the Java API documentation automatically using
the javadoc tool.
4.3.4 Relationship between Java source file name and class name
☝
Other than the fact that Java source files have an extension .java (or .jav ), there
is only one rule about the class name and the name of its source code file - the
code for a top level public type (recall that "type" implies class, interface, or
enum) must be written inside a Java file with the same name (with extension dot
java, of course!) .
For example, if you are writing code for a public class named Student , then the
name of the source code file must be Student.java
In light of the above rule, let us take a look at a few questions that might pop
into your head:
Q. Does that mean I cannot have multiple classes in a single file?

A. No, you certainly can have multiple classes in a single file. But only one of
them can be public and the name of that public class must be the same as the
name of the file. It is okay even if there is no public class in a file.
Q. What if I don't have a public class? What should be the name of the file
in that case?
A. You can code a non-public class in a file with any name. However, it is a
good programming practice to keep even a non-public class in a file by the same
name.
Q. What about interfaces? Enums?

A. The rule applies to all types, i.e., classes, interfaces, and enums. For
example, you cannot have a public class and a public interface in the same file.
There can be only one public type in one file.
Q. What about nested types? Can I have two public classes inside a class?
A. The rule applies only to top level types. So, yes, you can have more than one
public types inside another type. For example, the following is valid:

public class TestClass

{
public interface I1{ }
public class C1{ }
public static class C2{ }
public enum E1{ }
}
I1, C1, C2, and E1 are called "nested types" aka

"nested classes" because their declaration appears
within the body of another class or an interface. Types
that are not nested inside other types are called "top
level" types. The topic of nested classes is not included
in the Part 1 exam but is included in Part 2 exam, so,
it is good to know at least the terminology at this
stage.
Remember that this restriction is imposed by the Java compiler and not the JVM.
Compiler converts the source code into class files and generates an independent
class file for each type (irrespective of whether that type is public or not) defined
in that source file. Thus, if you define three classes in Java file (one public and
two non-public), three separate class files will be generated. The JVM has no
idea about the Java source file(s) from which the class files originated.
It is a common practice, however, to define each type, whether public or not, in

its own file. Defining each type in its own independent file is a very practical
approach if you think about it. While browsing the code folder of a Java project,
you only see the file name. Since you cannot see inside the file, it will be very
hard for you to find out which class is defined in which Java file if you have
multiple definitions in a single Java file.
It is interesting to know (though not required for the exam) that imposition of
this rule is actually optional . A compiler may chose to ignore this rule
altogether. Java language specification, Section 7.6 mentions that this rule may
be imposed by a Java compiler only if the source code is stored in the file system
and the type in that source file is being referred to by other types. Thus, it is
possible for a compiler to ignore this rule if, for example, the code is stored in
the database. Or if the type defined in a file is not referred to by other types.
For the purpose of the exam , all you need to know is that Oracle's Java
compiler enforces this rule.
You may see multiple public classes in the code listing of a question.
But don't immediately jump to the conclusion that the code will not
compile. Unless the problem statement explicitly says that these
classes are written in the same file, Oracle wants you to assume that
they are written in separate files.
Directory in which source files should reside ☝
Although it is a common (and a good) practice to keep the source file in the
directory that matches the package name in the file, there is no restriction on the
directory in which the source file should reside. For example, if the package
statement in your Student.java file is com.university.admin , then you should
keep Student.java file under com/university/admin directory. IDEs usually
enforce this convention. So, if you are using an IDE, you may see errors if you
keep Student.java file anywhere else but remember that this is not required by
the Java language. You can still compile the file from the command line. Check
out the Compilation and Execution section under Kickstarter for Beginners
chapter to understand the manual compilation process. In the Modules chapter,
you will see that keeping source files organized this way has an additional
benefit.
4.3.5 Quiz ☝
The following options show complete code listings of a Java file named
Student.java. Which of these will compile without any error?
Select 1 correct option.
A.
//Start of file

}
public enum Grade{ A , B, C, D }
//End of file
B.
//Start of file
class Student{
}
enum Grade{ A , B, C, D }
enum Score{ A1 , A2, A3, A4 }
//End of file
C.
//Start of file
public interface Gradable{

}
public interface Person{
}
//End of file
D.
//Start of file
class Student{
}
public class Professor{
}
//End of file
E.
//Start of file
package com.enthuware.ocajp;
//End of file
Correct answer is B.
Option A and C are incorrect because you cannot define more than one public
top level type in a source file. Option D is incorrect because the Professor class
is public. A public class must reside in a file by the same name but here, the
name of the file is Student.java . Option E is incorrect because every
source file must have at least on Java artifact defined in it. Option B is correct
because Java allows a file to have any number of non-public types.
4.4 Identify the scope of variables

4.4.1 Scope of variables ☝
Java has three visibility scopes for variables - class, method, and block.
Java has five lifespan scopes for variables - class, instance, method, for loop,
and block.
4.4.2 Scope and Visibility ☝
Scope means where all, within a program, a variable is visible or accessible

directly without any using any referencing mechanism.
For example, the scope of a President of a country is that country. If you say
"The President", it will be interpreted as the person who is the president of the
country you are in. There cannot be two presidents of a country. If you really
want to refer to the presidents of two countries, you must also specify the name
of the country. For example, the President of US and the President of India.
At the same time, you can certainly have two presidents in a country - the
President of the country, and the President of a basketball association within that
country! If you are in your basketball association meeting, and if you talk about
the president in that meeting, it will be interpreted as person who is the president
of the association and not the person who is the president of your country. But if
you do want to mean the president of the country, you will have to clearly say
something like the "President of our country". Here, "of our country" is the
referencing mechanism that removes the ambiguity from the word "president".
In this manner, you may have several "presidents" in a country. All have their
own "visibility". Depending on the context, one president may shadow or hide
(yes, the two words have different meanings in Java) another president. But you
cannot have two presidents in the same "visibility" level.
This is exactly how "scope" in Java (or any other programming language, for
that matter) works. For example, if you declare a static variable in a class, the
visibility of that variable is the class in which you have defined it. The visibility
of an instance variable is also the class in which it is defined. Since both have
same visibility, you cannot have a static variable as well as an instance variable
with the same name in the same class. It would be like having two presidents of
a country. It would be absurd and therefore, invalid.
If you declare a variable in a method (either as a method parameter or within a

method), the visibility of that variable is within that method only. Since a method
scope is different from a class scope, you can have a variable with the same
name in a method. If you are in the method and if you try to refer to that variable
directly, it will be interpreted as the variable defined in the method and not the
class variable. Here, a method scoped variable shadows a class scoped variable.
You can, of course, refer to a class scoped variable within a method in such a
case but you would have to use the class name for a static variable or an object
reference for an instance variable as a referencing mechanism to do that.
Similarly, if you declare a variable in a loop, the visibility of that variable is only
within that loop. If you declare a variable in a block such as an if, do/while, or
switch, the visibility of that variable is only within that block.
Here, visibility is not to be confused with accessibility

(public/private/protected). Visibility refers to whether the compiler is able to see
the variable at a given location directly without any help.
For example, consider the following code:
public class Area{

public static String UNIT="sq mt"; //UNIT is visib
le all over inside the class Area
public void printUnit(){
System.out.println(UNIT); //will print "sq mt
" because UNIT is visible here
}
}
public class Volume{

//Area's UNIT is accessible in this class but not
visible to the compiler directly.
public static String UNIT="cu mt";
public void printUnit(){

System.out.println(UNIT); //will print "cu mt"
System.out.println(Area.UNIT); //will print "

sq mt"
}
}
In the above code, a public static variable named UNIT of a class Area is
accessible to all other classes but that doesn't mean another class Volume
cannot also have a static variable named UNIT . This is because within the
Volume class, Area 's UNIT is not directly visible. You would need to help the
compiler by specifying Area.UNIT if you want to refer to Area 's UNIT in
class Volume. Without this help, the compiler will assume that you are talking
about Volume 's UNIT .
Besides shadowing and hiding, there is a third category

of name conflicts called "obscuring". It happens when
the compiler is not able to determine what a simple
name refers to. For example, if a class has a field
whose name is the same as the name of a package
and if you try to use that simple name in a method,
the compiler will not know whether you are trying to
refer to the field or to a member of the package by the
same name and will generate an error. It happens
rarely and is not important for the exam.
4.4.3 Scope and Lifespan ☝
Scope and Lifespan
Besides visibility, scope is also related to the lifespan or life time of a variable.
Think of it this way - what happens to the post of the president of your local
basketball association if the association itself is dissolved? The post of the
president of the association will not exist anymore, right? In other words, the life
of the post depends on the existence of the association.
Similarly, in Java, the existence of a variable depends on the existence of the

scope to which it belongs. Once its life time ends, the variable is destroyed, i.e.,
the memory allocated for that variable is taken back by the JVM. From this
perspective, Java has five scopes: block, for loop, method, instance, and class .
When a block ends, variables defined inside that block cease to exist. For
example,

{

{
int i = 0; //i exists in this block only
System.out.println(i);//OK
}
System.out.println(i);//NOT OK because i has alre
ady gone out of scope.
Variables defined in a for loop's initialization part exist as long as the for loop
executes. Notice that this is different from variables defined inside a for block,
which cease to exist after each iteration of the loop. For example,

{
for(int i = 0; i<10; i++)
{
int k = 0; //k is block scoped. It is reset to
0 in each iteration.
System.out.println(i);// i retains its value fr

om previous iteration.
}
//i and k are both out of scope here.
}
}
When a method ends, the variables defined in that method cease to exist.
When an object ceases to exist, the instance variables of that object cease to
exist.
When a class is unloaded by the JVM (not important for the exam), the static
variables of that class cease to exist
It is important to note here that lifespan scope doesn't affect compilation. The
compiler checks for the visibility scope only. In case of blocks, loops, and,
methods, the lifespan scope of the variables coincides with the visibility scope.
But it is not so for class and instance variables. Here is an example:

{
int data = 10;
TestClass t = new TestClass();
t = null;
System.out.println(t.data); //t.data is accessib
le therefore, it will compile fine even though the obj
ect referred to by t has already ceased to exist.
Lifespan scope affects the run time execution of the program. For example, the
above program throws a NullPointerException at run time because t
doesn't exist and neither does t.data at the time we are trying to access t and
t.data .
4.4.4 Scopes Illustrated ☝
The following code shows various scopes in action.
class Scopes{
int x;//visible throughout the class
static int y;//visible throughout the class
public static void method1(int param1){ //visible

throughout the method
int local1 = 0; //visible throughout the method

{
int anonymousBlock = 0;//visible in this b
lock only
anonymousBlock = 1;//compilation error

for(int loop1=0; loop1<10; loop1++){
int loop2 = 0;
//loop1 and loop2 are visible only here.
}
loop1 = 0;//compilation error
loop2 = 0;//compilation error
if(local1==0){
int block1 = 0;//visibile only in this if
block
block1 = 7; //compilation error

switch(param1){
case 0:
int block2 = 10;//visible all over cas
e block
break;
case 1:
block2 = 5;//valid
break;
default:
System.out.println(block2);//block2 is
visible here but compilation error because block2 may
be left uninitialized before access
}
block2 = 9;//compilation error
int loop1 = 0, loop2 = 0, block1 = 0, block2 =

8;//all valid
}
}
4.4.5 Scope for the Exam ☝
The important thing about scopes that you must know for the exam is when you
can and cannot let the variable with different scopes overlap. A simple rule is
that you cannot define two variables with the same name and same visibility
scope. For example, check out the following code:
class Person{
private String name; //class scope
static String name = "rob"; //class scope. NOT OK be

cause name with class scope already exists.

for(int i = 0; i<10; i++){
String name = "john"; //OK. name is scoped o
nly within this for loop block
}
String name = "bob"; //OK. name is method scoped.
System.out.println(name);//will print bob
In the above code, the static and instance name variables have the same visibility
scope and therefore, they cannot coexist. But the name variables inside the
method and inside the for loop have different visibility scopes and can therefore,
coexist.
But there is an exception to this rule. Consider the following code:
class Person{
private String name; //name is class scoped

String name = "bob"; //method scope. OK. Overlaps
with the instance field name defined in the class
int i = -1; //method scope
for(int i = 0; i<10; i++){ //i has for loop scope

. Not OK.
String name = "john"; //block scope. Not OK.
{ //starting a new block here

int i = 2; //block scope. Not OK.
}
}
}
Observe that it is possible to overlap the instance field with a method local
variable but it is not possible to overlap a method scoped variable with a loop or
block scoped variable.
4.4.6 Quiz ☝
Q1. What will the following code print when compiled and run?

{
int x = 10;
}
System.out.println(x);
}
Select 1 correct option
A. It will not compile.
B. It will print 10
C. It will print an unknown number

The correct answer is A.
Notice that x is defined inside a block. It is not visible outside that block.
Therefore, the line System.out.println(x); will not compile.
Q.2 What will the following code print when compiled and run with the
command line:
java ScopeTest hello world
public class ScopeTest {

private String[] args = new String[0];

args = new String[args.length];
for(String arg: args){
System.out.println(arg);
}
String arg = args[0];
System.out.println(arg);
}
Select 1 correct option
A. It will not compile.
B. It will print:
null
null
null
C. It will print:
null
null
hello
Answer is B.
The line args = new String[args.length]; creates a new string
array with the same length as the length of the original string array passed to the
program and assigns it back to the same variable args . All the elements of this
new array are null .
The original string array passed to the program is lost.
The instance variable args is not touched here because it is shadowed in the
method code by the method parameter named args . Also, you need to have a
reference to an object of class ScopeTest to access an instance variable from a
static method.
4.5 Use local variable type inference

4.5.1 Local variable type inference ☝
One among a few criticisms of Java is that it is too verbose. Meaning, even
writing simple code requires too much typing. Can't really argue with that when
a simple program that prints hello world takes about a hundred characters!
The extends keyword is another example. What takes merely one character (: )
in C++, requires seven in Java. This is not an oversight by Java designers. Java is
actually designed to make the programmer state their intention very clearly, in an
unambiguous and easy to understand fashion, instead of making the reader infer
the intention of the coder based on the context.
In Java's defense, verbosity makes the code more readable and thus, easily
maintainable. It also makes the code less prone to bugs.
On the other hand, too much verbosity poses problems of its own. It would be
pretty annoying if you had to type
com.enthuware.ets.data.MultipleChoiceQuestion q = new
com.enthuware.ets.data.MultipleChoiceQuestion(); (103
characters) every time you wanted declare and create a
MultipleChoiceQuestion object. Of course, importing
com.enthuware.ets.data package reduces it to just
MultipleChoiceQuestion q = new
MultipleChoiceQuestion() . This is possible because the compiler is
able to infer from the context that by MultipleChoiceQuestion , the
developer really means
com.enthuware.ets.data.MultipleChoiceQuestion .
Well, Java 10 helps you make the above statement even shorter. Java 10 onward,
you can just write var q = new MultipleChoiceQuestion(); (38
characters) and it will mean the same thing. Since you are creating an object of
type MultipleChoiceQuestion , the compiler has no problem in inferring
that the type of the variable would also be the same, i.e.,
MultipleChoiceQuestion . This is what type inference essentially means.
The logic behind this shortcut is that if the compiler can unambiguously infer the
type of the variable from the context, why make the programmer type it
explicitly?
Now, about the "local variable" part. It may be possible for a compiler to infer
the type of a variable in several contexts. However, Java allows it to do so only
for a local variable. As you know, local variables are variables that are defined
inside a method. Therefore, you can use this feature only inside a method body.
This feature is commonly referred to as LVTI in short and, since it uses the
word var in the syntax, it is also known as var declaration .
Let us now see a few examples of valid and invalid usages of this feature. Valid
usages first:
public class LVTITest1 {

static{
var str1 = "hello1"; //valid in static as well as
instance initializers
}
public LVTITest1(){
var str2 = "hello2"; //valid in constructors
}
public static void main(String[] args)
{
var i = 10; //type of i is int because 10 is an int
var f = 1.0f; //type of f is float because 1.0f is

a float
var strA = new String[]{"a", "b" }; //type of strA

is String[]
var d = Math.random(); //type of d is double becaus

e return type of Math.random is double
Object obj = "hello"; //valid, assigning a String t

o an Object variable
var obj2 = obj; //type of obj2 is Object and not St

ring
for(var str : strA){ //type of str is String
var p = str; //type of p is String because type

of str is String
}
switch(strA[0]){
case "a":
var m = new Object();//type of m is Object
}
}
}
Observe that all of the usages of var are scoped locally. Even the variable
declared within the static block is local to that block. Another important point
illustrated in the above code is that the type of obj2 will be inferred as Object
and not String . This is because the type of the source variable obj is
Object . The compiler has no knowledge of the type of the actual object that
this variable will point to at run time. Therefore, it only goes by the declared
type of obj to infer the type of obj2 .
Here are a few examples of a invalid usages that you will be required to identify
in the exam:
public class LVTITest2 {

var static value1 = 10; //can't use LVTI for class me
mbers
var value2 = 10; //can't use LVTI for instance members
public static void main(var args)//can't use LVTI for

method parameters
{
var p;//can't use LVTI for uninitialized variable
var n = null;//invalid because type of null can't b

e determined
var doubleArray = {1, 2}; //type must be specified
in array initializer if using LVTI for variable
var[] ia = new int[]{1 , 2}; //can't apply [] to va

r because var is not a type
}
public static var getValue(){//can't use LVTI for dec
laring return type of a method
return "hello";
}
}
Observe that in the above code there is no way for the compiler to infer the type
of p because there is no source variable or source value that is being assigned to
p in the same statement. In other words, the context does not have enough
information for the compiler to determine the type of p . Similarly, the compiler
cannot figure out the type of doubleArray because the type of the value given
is ambiguous. {1, 2} could be interpreted as an int array or a byte array also.
In fact, to avoid confusion, Java prohibits type inference if the type of the array
is not specified explicitly in the array initializer.
The case of instance field value in the above code is interesting. It is possible
to infer the type of the variable from the value. However, Java prohibits using
type inference for class and instance members because such members are part of
the API of a class and are therefore, used by other classes. Inferring the type
instead of explicitly stating the type makes the type of the field dependent on the
value that is being assigned to it. If you change this value later on, the type of the
variable may change and that may adversely impact other classes.
LVTI in Lambda Expressions ☝

Java 11 has allowed the usage of LVTI in lambda expressions. I will discuss it in
the Lambda Expressions
Typically, the exam has two kinds of questions on this topic.

In the first kind, you need to determine whether a particular usage of
var declaration is valid or not. This is easy because any use of var
keyword outside of a local scope (i.e. methods, constructors, or
initializer blocks), is invalid. Even within a local scope, it is valid only
if there is a value with a known type that is being assigned to the
target variable.
In the second kind, you need to determine the type of the variable
when you declare it with the var keyword. This is also simple
because the type of the variable will be exactly the same as the
declared type of the source variable that is being assigned to the
variable. In case you are directly assigning a value to the variable,
the type of the variable will be the same as the type of the value.
When should var be used? ☝
LVTI is a powerful shortcut. But you should remember that that is exactly what
it is. A shortcut. It is not a keyword or even a reserved word (so, this is actually a
valid line of code: var var = 10; ). When the compiler sees var , it simply
converts the statement to a full blown declaration. As with all shortcuts, care
must be taken to ensure that its usage doesn't affect the readability of the code.
Consider the following line of code appearing in a method:
var x = getValue();
Compiler can figure out the type of x and will accept the code happily, but can
you tell the type of x just by looking at the above code? You can't do that unless
you look at the method declaration. In fact, every time someone goes through the
method, they will have to check the getValue() method declaration to be
sure of the type of x . Obviously, saving a few keystrokes has reduced
readability tremendously here. Now, consider the following lines of code:
var mapOfStateCapitals = new HashMap<String, String>()

;
var state_listOfTownsMap = new HashMap<String, List<St
ring>>();
The usage of var in the above code actually improves readability by reducing
clutter.
4.6 Explain objects' lifecycles

4.6.1 Life cycle of an Object ☝
You know that objects are always created in heap space . The JVM allocates
space in the heap to store the values of the instance variables of an object. Since
every class ultimately extends from java.lang.Object , even if a class
does not define any instance variable of its own, it will inherit the ones defined
in the Object class. Thus, every object will take some space in the heap. Since
heap space is not unlimited, only a limited number of objects can be stored in the
heap. Thus, it is possible to run out of heap space if a program keeps creating
objects. We must, therefore, have some way of getting rid of objects that we
don't need anymore, right? Well, that is exactly what some languages such as
C++ provide. C++ lets you create as well as delete objects. It lets you allocate
and deallocate memory as you please. In other words, it makes the programmer
"manage" the heap space.
In Java, on the other hand, the heap space is managed entirely by the JVM.
There is no way for a programmer to directly manipulate the contents of this
space. The only thing a programmer can do to indirectly affect this space is to
create an object. Java does not even let you "delete" an object. There is no way
to "deallocate" the memory consumed by objects either. The question that one
may ask here is how in the world then can a program function if the memory
runs out because of objects that are not needed?
Recall our discussion about object references. An object can be accessed only
through its reference. If you have a reference to an object, you can read from or
write to its fields, or invoke methods. You can keep this reference in a variable.
You can also keep multiple copies of the same reference. You can pass it on to
other methods as well. You can think of it like a fish hooked on at the end of a
fishing line. As long as you hold on to the fishing rod, you can get hold of the
fish. You lose the rod, you lose the fish. Similarly, if you lose all the references
to an object, there is no way to get to that object any more. The JVM uses this
fact to manage the heap space.
The JVM keeps track of all the references to an object and as soon as it realizes
that there are no references to that object, it concludes that this object is not
required anymore and is basically "garbage" . It then makes arrangements to
reclaim the space occupied by the object. This arrangement for reclaiming the
space occupied by such objects is aptly called "garbage collection" . This
essentially is what automatic memory management is, because, as you can see,
the programmer does not have to deal with managing the memory. The
programmer focuses only on creating objects and using them as and when
required. The JVM cleans up the memory held by objects automatically in the
background.
With the above discussion in mind, it should be easy to visualize the life cycle
of an object. An object comes alive when it is created.
An object can be created in three ways - using the

new keyword , through deserialization , and by cloning
. Deserialization and cloning are not on the OCP Java
11 Part 1 exam.
It remains alive as long as it is being referenced from an active part of a

program. The object is dead or inaccessible once there are no references pointing
to it. An object no longer exists, once the JVM destroys, i.e., reclaims the
memory consumed by the object during garbage collection.
4.6.2 Garbage Collection ☝
Garbage Collection ☝
Now that we have established what "garbage" is, it is easy to understand what
garbage collection entails. Garbage collection is an activity performed
periodically by the JVM to reclaim the memory occupied by objects that are no
longer in use. Let us dig deeper into each part of the previous sentence:
1. Activity performed periodically - The JVM performs garbage collection at

regular intervals so that memory is made available to the program before a
request to create an object fails for want of memory. It is not possible for a
program to control the periodicity of this activity. It not possible for a
programmer to even make the JVM perform this activity instantly. There is a
method named gc in java.lang.System class that lets a programmer
request the JVM to perform garbage collection. You can call System.gc()
any time you believe it is appropriate to clean up the memory but this is just a
request. There is no guarantee that JVM will actually perform garbage collection
after invocation of this method. A JVM may provide options to customize the
behavior of its garbage collection process through command line arguments.
Although a discussion about these arguments will be beyond the scope of the
exam, I suggest you check them out in your spare time because garbage
collection is a favorite interview topic of technical managers.
2. Reclaim the memory - Reclaiming the memory means that the memory
occupied by an object is now marked as free to store new objects. After
reclaiming the memory from multiple objects, the JVM may even reorganize the
heap space by moving the objects around and creating large chunks of free
memory. This process is pretty much like the defragmentation of a hard-drive but
within the program's RAM.
3. Objects that are no longer in use - I mentioned earlier that if an object is not
referenced from any active part of a program, the JVM concludes that the object
is no longer in use. While this statement is true, it is a bit more complicated than
it looks. Let us start with the following code:

Object o1 = new Object(); //1
o1 = new Object(); //2

}
}
An object is created at line marked //1 and its reference is assigned to variable
o1 . At this point there is only one reference variable that is pointing to this
object. On the next line, another object is created and its reference is assigned to
the same variable o1 . Observe that the value held by o1 has been overwritten
by the new value. Thus, the object that was created at //1 is not being referenced
by any variable at all after the line marked //2. There is no way we can access
this object now because we have lost the only reference that we had of this
object. The JVM is aware of this fact and will mark this object for garbage
collection.
The object created at line marked //2 is being referenced by a variable and can be
accessed through this variable. It is, therefore, not eligible to be garbage
collected. Well, at least until the main method ends, after which there will be no
reference pointing to this object either, and it will also be eligible to be garbage
collected.
The above code illustrates how an object may be left without any reference
variable pointing to it. Let us take it up a notch:

Object instanceVar;
public TestClass(Object methodParam){
instanceVar = methodParam;
}
Object tempVar = new Object(); //1
TestClass tc = new TestClass(tempVar); //2
tempVar = new Object(); //3

tc.instanceVar = null; //4
}
}
An object of class Object is created at //1 and its reference is assigned to local
variable named tempVar . An object of class TestClass is created at //2 and
value of tempVar is passed to TestClass 's constructor through method
parameter named methodParam . TestClass 's constructor copies this
reference to an instance variable named instanceVar . Thus, after execution
of line //2, the object created at //1 is referred to by two reference variables -
tempVar and instanceVar .
Now, at //3, a new object is created and its reference is assigned to the local
variable tempVar . Thus, tempVar stops pointing to the object it was
pointing to earlier and starts pointing to this new object. But observe that the
instance variable instanceVar is still pointing to the object created at //1.
At //4, we make the instanceVar lose its value by assigning it null .

Therefore, after this line, the object that was created at //1 has no reference
pointing to it. There is no way to access this object after this line and thus, this
object is eligible to be garbage collected.
In the above two examples, I showed you how an object can be deemed no
longer in use when there are no reference variables pointing to it. Indeed, if there
are no variables pointing to an object, it is not possible to access that object. But
can there be a situation where there is a variable pointing to an object but that
object is still eligible for garbage collection? Let us modify the above code a bit:

Object instanceVar;
public TestClass(Object methodParam){
instanceVar = methodParam;
}
Object tempVar = new Object(); //1
TestClass tc = new TestClass(tempVar); //2
tempVar = new Object(); //3
tc = null; //4
}
}
The only change I have made in this code is in the line marked //4. Instead of
setting tc.instanceVar to null , I have set tc to null . Thus, the
variable tc does not point to the TestClass object after this line. In fact,
there is no variable that is pointing to the TestClass object created at //2.
Thus, even though the instance variable instanceVar of this TestClass
object still points to the object created at //1, there is no way to access that object
because the only way to access that object was through the TestClass
instance, which itself is not accessible anymore. Thus, both - the TestClass
instance created at //2 and the object created at //1 - are eligible to be garbage
collected.
In other words, not only the objects that have no reference to them are eligible
for garbage collection, but the objects that are referenced only by objects that are
themselves eligible to be garbage collected, are also eligible to be garbage
collected. In the above example, there was a chain of just two objects (the
TestClass instance and the Object instance) that became eligible for garbage
collection but there could be any number of such interconnected objects that
become eligible for garbage collected if none of the objects of that chain can be
referenced from any active part of a program . This graph of interconnected
objects is known as an "island" of isolation and is considered garbage because
none of the objects of that graph are reachable from an active part of a program
even though they are reachable through each other.
Let me now explain what is meant by the cryptic looking phrase "active part of
a program ". Note that it is only when the statements written in a program are
executed that objects are created. A program code may contain several
statements that create objects using the new keyword but if those statements are
not executed, obviously, no object will be created. Indeed, the code that we write
is merely a set of instructions to the JVM. Nothing will actually happen if those
instructions are not executed. In Java, execution of the code is done through
"threads". When you run a program through the command line, a thread called
"main thread" is created automatically and this thread starts executing the
statements written in the main method. The statements may be simple statements
such as i = 10; or they could be method calls, in which case the thread will
execute the statements written inside that method first before moving on to the
next statement in the main method. Java also allows you create your own threads
and give them separate sets of instructions to execute. All such code that falls
under the scope of the main thread and the threads created by the programmer is
nothing but the active part of the program. When a thread dies, that is, when it is
done executing all the instructions that fall under its scope (for example, the
main thread dies when it reaches the end of the main method), any object that
was created by this thread will be eligible for garbage collection unless the
reference of that object is still held on to by some other live thread.
For the purpose of the OCP Part 1 exam, you do not

have to worry about threads or the impact of threads
on garbage collection. The exam merely scratches the
surface of this topic. The exam only expects that you
understand the meaning of garbage collection and that
an object can be garbage collected when there are no
references pointing to that object. You should be able
to trace the reference variables pointing to an object
and identify the point at which that object loses all its
references.
However, a deep understanding of Garbage Collection is

very critical for a Java programmer and that is why it
is a favorite topic of discussion in technical interviews.
I suggest you read about the following terms if you
want to ace a Java technical interview - finalization,
finalize method, object resurrection, types of references,
algorithms used to identify garbage, customizing the
behavior of garbage collector through command line
options, and object generations.
4.6.3 Garbage Collection for the exam ☝
Typically, there are two types of questions that you will encounter in the exam:
find out the line number after which an object becomes eligible for garbage
collection; and find out how many objects will be eligible for garbage collection
right after a particular line number. For both kinds you need to keep track of the
references and the objects these references are pointing to at each line. It is
possible to do all this in your head but the questions are designed to make you
lose track of the objects and the references. Therefore, it is best to make use of a
pen and paper and draw whatever is going on in your head. Let me show you
how. Consider the following code where you are expected to find out the line
after which the object created at line 3 is eligible for garbage collection.
1: public class TestClass {

2: public static void main(String[] args){
3: Object foo = new Object();
4: Object bar = foo;
5: foo = new Object();
6: Object baz = bar;
7: foo = null;
8: bar = null;
9: baz = new Object();
10: }
11:}
To get to the answer, I will draw a diagram to show the state of affairs after each
line of code. On the left-hand side, I will write the reference variable name and
on the right, I will draw a box to show the existence of an object on the heap. I
will also write a number in the box to distinguish one object from another.
Step 1: At line 3, a new object is created and is assigned to the reference

variable foo .
Step 2: At line 4, a new variable named bar is defined and is set to the same
value as foo . Thus, foo and bar now point to the same object.
Step 3: At line 5, a new object is created and foo is made to point to this new
object.
Step 4: At line 6, a new variable named baz is defined and is assigned the value
held by bar . In other words, baz now points to whatever bar is pointing to,
i.e., obj 1.

Step 5: At line 7, foo is set to null . In other words, foo is not pointing to the
object is was pointing to earlier. Observe that obj 2 is not being pointed to by
any reference variable after the execution of line 7. Therefore, this object will be
eligible for garbage collection after this line. However, this is not the object you
are interested in. So, let's keep executing the statements.
Step 6: At line 8, bar is also being set to null .

Step 7: At line 9, a new object is created and assigned to baz . Thus, bar stops
pointing to obj 1 and starts pointing to obj 3 after the execution of this line.
Observe that nobody is pointing to obj 1 now. This is the object that was created
at line 3, and since it is not being referenced by any variable, it is eligible for
garbage collection after line 9.
Now, let us look at the same code from another perspective. What if you are
asked about the number of objects that are eligible for garbage collection after,
say, line 8? It is easy to figure that out by looking at the above figures. Status of
objects after line 8 shows that obj 2 is the only object that is not being referred to
by any variable. Thus, it is the only object that is eligible for garbage collection
after line 8. What about after line 9? Two (obj 1 and obj 2). At the end of the
method (i.e.line 10)? All three objects.
You should solve a few mock questions using this approach. With practice, you
will be able to do it in your head and will not need to draw the diagrams on
paper.
Garbage collection of Strings ☝
While going through mock exams or through other preparation material, you
may encounter code snippets that show multiple String objects getting created.
Something like this:

for(int i=0; i<5; i++){
str = str + i;
}
The above creates one String object containing "hello" at the beginning and
then two more in each iteration of the loop - a String containing the int value of
i and the concatenated String. Thus, overall, the above code creates 1 + 2*5 =
11 Strings. However, if you are asked how many String will be eligible to be
garbage collected, the answer is not that easy. The Java Language Specification
mentions in Section 15.8.1 that the non-string operand of the + operator is
converted to a String at runtime but it does not clearly tell whether this String
goes to the String pool (in which case it will not be garbage collected) or not.
Let me show you another piece of code:
String s = "hello";
int k = 100;
s = s +"123"+k;
In this case, JLS section 15.8.1 clearly says that a compiler may avoid creating
multiple strings altogether by making use of a StringBuilder. Thus, it is not
possible to tell how many Strings will be created and how many will be eligible
to be garbage collected.
Don't worry, you will not get questions in the exam about garbage collection of
Strings. I have talked about it here only to make you aware of the issue. Strings
muddle the topic of garbage collection so much so that it is a bad idea to use
them while explaining garbage collection. You need not spend anymore time on
this topic.
Prior to Java 7, interned strings were never garbage

collected even if there were no references to a
particular string. Thus, it was possible for the string
pool to run out of memory. Typically, programs create a
lot of strings through concatenation while writing to log
files. It was not too uncommon for such programs to
start getting OutOfMemoryError after a while due to
this issue. From Java 7 onwards, strings in the string
pool can also be garbage collected. You will not be
required to answer questions on this aspect of Strings
in the exam.
4.7 Exercise ☝
1. Define a reference type named Bird . Define an instance method named
fly in Bird . Define a few instance as well as static variables of type int
, float , double , boolean , and String in Bird .
2. Create a TestClass that has a static variable of type Bird . Initialize
this variable with a valid Bird object. Print out the default values of static
and instance variables of Bird from the main method of TestClass .
Also print out the static variable of TestClass from main . Observe the
output.
3. Create and initialize one more instance variable of type Bird in
TestClass . Assign values to the members of the Bird instance pointed
to by this instance variable in TestClass 's main .Assign values to the
members of first Bird using the second Bird . Print the values of the
members of both the Bird objects.
4. Write code in fly method to print out the values of all members of Bird .
Alter main method of TestClass to invoke fly on both the instance of
Bird . Observe the values printed for static variables of Bird .
5. Add an instance variable of type Bird in Bird . Initialize this variable on
the same line using "new Bird ()" syntax. Instantiate a Bird object in
TestClass 's main and execute it. Observe the output.
6. Remove the initialization part of the variable that you added to Bird in
previous exercise. Initialize it with a new Bird object separately from
TestClass 's main . Identify how many Bird objects will be garbage
collected when the main method ends.
7. Add a parameter of type Float to Bird 's fly method. Return an int
value from fly by casting the method parameter to int . Invoke fly
multiple times from TestClass 's main by passing a float literal, a
Float object, a double literal, an int , an Int eger, and a String
containing a float value. Observe which calls compile.
8. Assign the return value of fly to an int variable, a float variable, a
String variable, and boolean variable. Observe which assignments
compile. Try the same assignments with an explicit cast. Print these
variables out and observe the output.
Chapter 5 Working with String APIs
5.1 Create and manipulate Strings

5.1.1 What is a "string"? ☝
In Java, a "string" is an object of class java.lang.String . It represents a series of

characters. Strings such as "1234" or "hello" are really just objects of this
class.
String is a final class, which means it cannot be extended. It extends Object
and implements java.lang.CharSequence . If you are not sure of what
final and implements mean, don't worry, I will cover these concepts while
talking about inheritance and interfaces respectively.
In the Java world, String objects are usually just called "strings" . Although a
string is just like any other regular object, it is such a fundamental object that
Java provides special treatment to strings in terms of how they are created, how
they are managed, and how they are used. Let's go over these aspects now.
5.1.2 Creating Strings ☝
Creating strings through constructors ☝
The String class has several constructors but for the purpose of the exam, you
only need to be aware of the following:
1. String() - The no-args constructor creates an empty String.
2. String(String str), String(StringBuilder sb) - Create a new String by
copying the sequence of characters currently contained in the passed String
or StringBuilder objects.
3. String(byte[] bytes) - Creates a new String by decoding the specified array
of bytes using the platform's default charset.
4. String(char[] value) - Creates a new String so that it represents the
sequence of characters currently contained in the character array argument.
Note that a string is composed of an array of char s. But that does not mean a
string is the same as a char array. Therefore, you cannot apply the array
indexing operator on a string. Thus, something like char c = str[0]; ,
where str is a String , will not compile.
Creating strings through concatenation ☝
The second common way of creating strings is by using the concatenation (, i.e.,
+ ) operator:
String s1 = "hello ";

String s12 = s1 + " world"; //produces "hello world"
The + operator is overloaded in such a way that if either one of its two operands
is a string, it converts the other operand to a string and produces a new string by
joining the two. There is no restriction on the type of operands as long as one of
them is a string.
The way + operator converts the non-string operand to a string is important:
1. If the non-string operand is a reference variable, the toString() method

is invoked on that reference to get a string representation of that object.
2. If the non-string operand is a primitive variable or a primitive literal value,
a wrapper object of the same type is created using the primitive value and
then a string representation is obtained by invoking toString() on the
wrapper object.
3. If the one of the operands is a null literal or a null reference variable, the
string "null" is used instead of invoking any method on it.
The following examples should make this clear:
String s1 = "hello ";

String s11 = s1 + 1; //produces "hello 1"
String s12 = 1 + " hello"; //produces "1 hello"
String s2 = "" + true; //produces "true";
double d = 0.0;
String s3 = "-"+d +"-"; //produces "-0.0-"
Object o = null;
String s4 = "hello "+o; //produces "hello null". No Nu
llPointerException here.
Just like a mathematical expression involving the + operator, string

concatenation is also evaluated from left to right. Therefore, while evaluating the
expression "1"+2+3 , "1"+2 is evaluated first to produce "12" and then
"12"+3 is evaluated to produce "123" . On the other hand, the expression 1
+ 2 +"3" produces "33" . Since neither of the operands to + in the
expression 1 + 2 is a String , it will be evaluated as a mathematical
expression and will therefore, produce integer 3 . 3 + "3" will then be
evaluated as "33" .
Remember that to elicit the overloaded behavior of the + operator, at least one of
its operands must be a String . That is why, the following statements will not
compile:
String x = true + 1;//Will not compile. First operand

is boolean and second is int.
Object obj = "string";//OK
String y = obj + obj; //Will not compile. Even though

obj points to a String at runtime, as far as the compi
ler is concerned, obj is an Object and not a String.
Since the toString method is defined in the Object class, every class in
Java inherits it. Ideally, you should override this method in your class but if you
do not, the implementation provided by the Object class is used. Here is an
example that shows the benefit of overriding toString in a class:
class Account{
String acctNo;
Account(String acctNo){
this.acctNo = acctNo;
}
//overriding toString.
//must be public because it is public in Object

public String toString(){
return "Account["+acctNo+"]";
}
}

Account a = new Account("A1234");
String s = "Printing account - "+a;
}
}
The above code produces the following output with and without overriding
toString :
Printing account - Account[A1234]
and
Printing account - Account@72bfaced
Observe that when compared to Object 's toString , Account 's

toString generates a meaningful string. Since the Object class has no idea
about what a class represents, it just returns a generic string consisting of the
name of the class of the object, the at-sign character '@', and the unsigned
hexadecimal representation of the hash code of the object. Don't worry, you will
not be asked to predict this value in the exam. Just don't get scared if you see
such a value in the exam.
On a side note, the print/println methods that we have often used also
behave just like the + operator with respect to generating a string representation
of the object that is passed to them. For example, when you call
System.out.print(acct); where acct refers to an Account object, the
print method invokes toString on that Account object and prints the
returned string.
The += operator ☝
In the chapter on Operators, we will see that += is a compound operator. It

applies the + operator on the two operands and then assigns the result back to the
variable on the left side. If the type of the variable on the left is String , it
performs a string concatenation. Here is an example:
String s = "1";
s += 2; //expanded to s = s + 2;
System.out.println(s); //prints "12"
The type of the variable on the left-hand side must be String or something that
can refer to a String (there are only two such types really - CharSequence
and Object ), otherwise the expression will not compile:
int x = 1;
x += "2";
The above code will not compile because you can't assign an object of type
String to a variable of type int . The type of the right operand can be
anything because if the operand is not a string, it will be converted to a string as
per the rules discussed above.
Now, can you tell what the following code will print?
Object m = 1;
m += "2";
System.out.println(m);
It will compile fine and print "12" . First, 1 will be boxed into an Integer
object, which will be assigned to m . This assignment is valid because an
Integer "is-a" Object . Next, the expression m += "2" will be expanded
to m = m + "2" . Since one of the operands of + in this expression is a string,
a string concatenation will be performed, which will produce the string "12" .
This string will be assigned to m . The assignment is also valid because a
String is an Object
Okay, how about this?
Object m = "Hello ";

m += 1;
System.out.println(m);
It will fail to compile because as far as the compiler is concerned, type of m is

Object and type of 1 is int . Therefore, when m+=1 is expanded to m = m
+ 1 , neither of the operands of + is a String!
5.1.3 String interning ☝
Since strings are objects and since all objects in Java are always stored only in
the heap space , all strings are stored in the heap space. However, Java keeps
strings created without using the new keyword in a special area of the heap
space, which is called "string pool" . Java keeps the strings created using the
new keyword in the regular heap space.
String interning ☝
The purpose of the string pool is to maintain a set of unique strings. Any time
you create a new string without using the new keyword, Java checks whether the
same string already exists in the string pool. If it does, Java returns a reference to
the same String object and if it does not, Java creates a new String object in the
string pool and returns its reference. So, for example, if you use the string
"hello" twice in your code as shown below, you will get a reference to the
same string. We can actually test this theory out by comparing two different
reference variables using the == operator as shown in the following code:
String str1 = "hello";

System.out.println(str1 == str2); //prints true
String str3 = new String("hello");

System.out.println(str1 == str3); //prints false
We will look at the == operator in detail later but for now, it simply checks
whether two references point to the same object or not and returns true if they
do. In the above code, str2 gets the reference to the same String object which
was created earlier. However, str3 and str4 get references to two entirely
different String objects. That is why str1 == str2 returns true but str1
== str3 and str3 == str4 return false .
In fact, when you do new String("hello"); two String objects are
created instead of just one if this is the first time the string "hello" is used in
the anywhere in program - one in the string pool because of the use of a quoted
string, and one in the regular heap space because of the use of new keyword.
String pooling is Java's way of saving program memory by avoiding creation of

multiple String objects containing the same value. It is possible to get a string
from the string pool for a string created using the new keyword by using String's
intern method. It is called "interning" of string objects. For example,

String str3 = str2.intern(); //get an interned string
object for str2

System.out.println(str1 == str3); //prints true
5.1.4 String immutability ☝
Strings are immutable. It is impossible to change the contents of a string once

you have created it. Let me show you some code that looks like it is altering a
string:
String s1 = "12";
s1 = s1+"34";
System.out.println(s1);//prints 1234
The output of the above code indicates that the string pointed to by s1 has been
changed from "12" to "1234" . However, in reality, the original string that s1
was pointing to remains as it is. A new string containing "1234" is created
instead and its address is assigned to s1 . So, after the last line of the above
code, the JVM would have created three different strings - "12" , "34" , and
"1234" .
There are several methods in the String class that may make you believe that
they change a string but just remember that a string cannot be mutated. Ever.
Here are some examples:
String s1 = "ab";
s1.concat("cd");
System.out.println(s1); //prints "ab"
s1.toUpperCase();
System.out.println(s1); //prints "ab"
In the above code, s1.concat("cd") does create a new string containing
"abcd" but this new string is not assigned to s1 . Therefore, the first println
statement prints "ab" instead of "abcd" . The same thing happens with
toUpperCase() . It does produce a new string containing "AB" but since this
string is not assigned to s1 , the second println statement prints "ab" instead of
"AB" . Note that the newly created strings "abcd" and "AB" will remain in
the string pool. The JVM will use them whenever it needs to create a string
containing the same characters. But as of now, we don't have any reference that
points to these strings.
Mutability ☝
Generally, mutability refers to the properties of an object of a class. Thus, an

immutable class implies that the instance variables of an object cannot be
changed once the instance is created. This is achieved by making instance
variables private and having only getter methods for reading their values.
Mutability is an important concept that has many aspects. However, a detailed
discussion on it is not in the scope for OCP Java 11 Part 1 exam. You should,
however, be aware that it is included in the Part 2 exam and is also a favorite
topic of discussion in technical interviews.
5.1.5 Manipulating Strings ☝
The String class contains several methods that help you manipulate strings.
To understand how these methods work, you may think of a string as an object
containing an array of chars internally. These methods simply work upon that
array. Since array indexing in Java starts with 0, any method that deals with the
locations or positions of characters in a string, also uses the same indexing logic.
Furthermore, any method that attempts to access an index that is beyond the
range of this array throws IndexOutOfBoundsException .
Here are the methods that belong to this category with their brief JavaDoc
descriptions:
1. int length() - Returns the length of this string.
For example, System.out.println("0123".length()); prints

4 . Observe that the index of the last character is always one less than the
length.
2. char charAt(int index) - Returns the char value at the specified

index. Throws IndexOutOfBoundsException if the index argument
is negative or is not less than the length of this string.
For example, System.out.println("0123".charAt(3)); prints

3 .
3. int indexOf(int ch)- Returns the index within this string of the
first occurrence of the specified character. Returns -1 if the character is not
found.
Examples:
System.out.println("0123".indexOf('2')); //prints 2
System.out.println("0123".indexOf('5')); //prints
-1
A design philosophy followed by the Java standard library regarding methods

that deal with the starting index and the ending index is that the starting index is
always inclusive while the ending index is always exclusive. String 's
substring methods works accordingly:
1. String substring(int beginIndex, int endIndex) -

Returns a new string that is a substring of this string.
Examples:
System.out.println("123456".substring(2, 4));
//prints 34 .
Observe that the character at index 4 is not included in the resulting
substring.
//prints 3456
//throws StringIndexOutOfBoundsException
2. String substring(int beginIndex) - This method works just

like the other substring method except that it returns all the characters from
beginIndex (i.e. including the character at the beginindex) to the end of the
string (including the last character).
Examples:
System.out.println("123456".substring(2));
//prints 3456 .
System.out.println("123456".substring(7)); //throws
StringIndexOutOfBoundsException .
The rule about not including the element at the ending

index is followed not just by the methods of the String
class but also by methods of other classes that have a
concept of element positions such as java.util.ArrayList.
The following methods return a new String object with the required changes:
1. String concat(String str) - Concatenates the specified string to

the end of this string.
Example - System.out.println("1234".concat("abcd"));
//prints 1 234abcd
2. String toLowerCase()/toUpperCase() - Converts all of the

characters in this String to lower/upper case.
Example - System.out.println("ab".toUpperCase());
//prints AB
3. String replace(char oldChar, char newChar) - Returns a
new string resulting from replacing all occurrences of oldChar in this string
with newChar.
Example: System.out.println("ababa".replace('a',
'c')); //prints cbcbc
4. String strip(), stripLeading(), stripTrailing(),

trim() - Returns a copy of the string, with leading and/or trailing white
space omitted.
Example:
System.out.println(" 123 ".strip()); //prints "123

" (without the quotes)
System.out.println(" 123 ".stripLeading()); //prin

ts "123 " (without the quotes)
System.out.println(" 123 ".stripTrailing()); //pri

nts " 123" (without the quotes)
System.out.println(" 123 ".trim()); //prints "123"

(without the quotes)
One interesting thing about the methods detailed above is that they return
the same string if there is no change in the string as a result of the
operation. Thus, the following code prints true:
String s1 = "aaa"; //there is no b in this string

String s2 = s1.replace('b', 'c'); //nothing is rep
laced
System.out.println(s1 == s2); //prints true becaus

e the same string was returned by replace
System.out.println("1234".strip() == "1234"); //pr

ints true for the same reason
It is very common to invoke these methods in the same line of code by chaining
them together:
String str = " hello ";

str = str.concat("world ").trim().concat("!").toUpperC
ase();
System.out.println(str);
The above code prints HELLO WORLD! . Note that such chaining is possible
only because these methods return a string. You will see a similar chaining of
methods in StringBuilder /StringBuffer classes as well.
Difference between trim() and strip() ☝
String class has had the trim method since the

beginning while the strip method was added in Java
11. Both the methods do the same thing, i.e., remove
"white spaces" from the beginning and end of the
String. The difference is in the way these methods
determine what constitutes a white space. Without
getting into too many details, you should be aware that
strip uses a better definition of a white space than
trim and identifies mores white spaces than trim . You
should use strip instead of trim in new code. You
may see the complete detail of difference in JavaDoc
API but it is not required for the exam.
Finally, here are a few methods that let you inspect the contents of a string:
1. boolean isBlank() : Returns true if the string is empty or contains

only white spaces, otherwise false.
2. boolean isEmpty() : Returns true if, and only if, length() is 0.
3. boolean startsWith(String prefix) : Returns true if this

string starts with the specified prefix.
4. boolean endsWith(String suffix) : Returns true if this string

ends with the specified suffix.
5. boolean contains(CharSequence s) : Returns true if and only if

this string contains the specified sequence of char values.
6. boolean equals(Object anObject) : Returns true if the contents

of this string and the passed string are exactly same. Observe that the type
of the parameter is Object . That's because this method is actually defined
in the Object class and the String class overrides this method. So, you can
pass any object to this method, but if that object is not a string, it will return
false.
7. boolean equalsIgnoreCase(String anotherString) :

Compares this String to another String, ignoring case considerations.
The above methods are fairly self-explanatory and work as one would expect
after looking at their names, so, I am not going to talk about them in detail here
but I suggest you take a look at their JavaDoc descriptions and write a few test
programs to try them out. You will not get any trick questions on these methods
in the exam.
5.1.6 Comparing strings ☝
There are two ways you can compare strings - using the == operator and using
the equals method of String class.
Comparison using == operator ☝
As I discussed earlier, the == operator, when applied to references, checks

whether two references point to the same object or not. You can, therefore, use
this operator on string references to check whether they point to the same String
object or not. If the two references point to the same string, they are obviously
"equal". The following code, for example, prints true for this reason:

System.out.println(str == "hello");//prints true
Since creating strings in Java is quite easy, it is very tempting to use the ==
operator for testing their equality. However, it is very dangerous to do so as
illustrated by the following code. All the checkCode method below wants to
do is to check whether the string passed to it matches with the string referenced
by the static variable named code .

static String code = "1234";
public static void checkCode(String str){
System.out.println(code == str);
}

checkCode("1234");
checkCode(new String("1234"));
}
}
This code prints true for the first comparison and false for the second.
Ideally, it should have printed true for both. The problem is that the
checkCode method has no knowledge of how the string that was passed to it
was created. As discussed earlier, when you create a string using the new
keyword, an entirely new String object is created. That is why the second check
returns false . It shows that == operator cannot guarantee you the right result
if you want to compare the character data of two strings.
Comparison using equals method ☝
String class has an equals method that compares the actual character data
of two strings to determine whether they are equal or not. This method returns
true if the data matches and false otherwise. Since it compares the actual
characters contained in two strings, it doesn't matter whether the two string
references are references to the same String object or are references to two
different String objects. Thus, new String("1234").equals("1234")
will always return true even though both are two different String objects.
The equals method, therefore, is a better way to compare two strings.
To fix the code shown above, just replace System.out.println(code ==

str); with System.out.println(code.equals(str)); It is
common to invoke equals on a String literal and pass a String variable instead of
the other way round., i.e., "1234".equals(str); is preferred to
str.equals("1234") because if str is null , the first style returns
false but the second style throws a NullPointerException .
5.2 Manipulate data using the StringBuilder class and

its methods
5.2.1 Why StringBuilder ☝
java.lang.StringBuilder is the mutable sibling of
java.lang.String . Both the classes directly extend Object and
implement CharSequence . Both are final as well.
You may be wondering why we need another class to deal with strings if
String allows us to do everything that we could possibly want to with strings!
Well, besides being mutable, StringBuilder is quite different from String
due to the fact that StringBuilder objects are treated just like objects of
other regular classes by the JVM. There is no "string pool" or "interning"
associated with StringBuilder objects. StringBuilder objects are
garbage collected just like other objects once they go out of scope, which means
they don't keep occupying memory forever. This makes StringBuilder
objects more suitable for creating temporary strings that have no use once a
method ends, such as, creating lengthy debug messages or building long xml
documents. It may be hard to believe but a program can create a large number of
String objects pretty quickly. For example, the following trivial code
generates an HTML view for displaying a list of names in a browser:
public String showPerson(List persons){

String html = "<h3>Persons</h3>";
for(Object o : persons){
Person p = (Person) o;
html = html +p.getName()+"<br class=\"mynewline\"
>";
}
return html;
}
The above code has the potential to wreak havoc on a program's memory.
Depending on the size of the list, it will create a large number of String
objects and all of them will sit in the memory for a long time, possibly for the
entire life-time of the program. The same method, written using
StringBuilder , is a lot more efficient:
public StringBuilder showPerson(List persons){

StringBuilder html = new StringBuilder("<h3>Persons
</h3>");
for(Object o : persons){
Person p = (Person) o;
html.append(p.getName()).append("<br class=\"myne
wline\" >");
}
return html;
}
It creates exactly two String objects and exactly one StringBuilder

object irrespective of the number of elements in the List. Furthermore, the
StringBuilder object will be garbage collected as soon as it goes out of
scope.
On the other hand, since String objects are interned, they are more suitable for
creating short strings that are used repeatedly in a program (For example, "<br
class=\"mynewline\" >" in the above code). Also, if you want to use
strings in a switch statement, then String is the only option.
5.2.2 StringBuilder API ☝
StringBuilder provides several constructors and methods and the exam

expects you to know most, if not all, of them. Let's go over the constructors first:
1. StringBuilder() : Constructs a StringBuilder with no characters in it

and an initial capacity of 16 characters. Here, "capacity" refers to the size of
an internal array that is used to store the characters. Initially, this array is
empty and is filled up as you start adding characters to the StringBuffer.
The StringBuilder object automatically allocates a new array with larger
size once this array is full.
The capacity of a StringBuilder is analogous to a bucket of water. It is
empty at the beginning and fills up as you add water to it. Once it is full,
you need to get a bigger bucket and transfer the water from the smaller
bucket to the bigger one.
2. StringBuilder(CharSequence seq) : Constructs a StringBuilder

that contains the same characters as the specified CharSequence. Recall
that String implements CharSequence as well. Thus, this constructor
can be used to create a new StringBuilder with the same data as an existing
String or StringBuilder.
3. StringBuilder(int capacity) : Constructs a StringBuilder with

no characters in it and an initial capacity specified by the capacity
argument. If you expect to build a large string, you can specify a big
capacity at the beginning to avoid reallocation of the internal storage array
later. For example, if you are building an HTML page in a method, you
might want to create a StringBuilder with a large initial capacity.
It is important to understand that specifying a capacity does not mean you
can store only that many characters in the StringBuilder. Once you fill up
the existing capacity, the StringBuilder will automatically allocate a new
and larger array to store more characters.
4. StringBuilder(String str) : Constructs a StringBuilder

initialized to the contents of the specified string. This constructor is actually
redundant because of the StringBuilder(CharSequence seq)
constructor. It exists only for backward compatibility with code written
before JDK 1.4, which is when CharSequence was first introduced.
Since the whole purpose of having a StringBuilder is to have mutable

strings, it is no wonder that it has a ton of overloaded append and insert methods.
But don't be overwhelmed because all of them follow the same pattern. The
append method only takes one argument. This argument can be of any type. The
insert method takes two arguments - an int to specify the position at which you
want to insert the second argument. Both the methods work as follows:
1. If you pass a CharSequence (which, again, implies String and

StringBuilder ) or a char[] , each character of the
CharSequence or the char array is appended to or inserted in the existing
StringBuilder.
2. For everything else, String.valueOf( ... ) is invoked to generate

a string representation of the argument that is passed. For example,
String.valueOf(123) returns the String "123" , which is then
appended to or inserted in the existing StringBuilder. In case of objects,
valueOf invokes toString() on that object to get its string
representation.
3. If you pass a null , the string "null" is appended to or inserted in the

existing StringBuilder. No NullPointerException is thrown.
4. All of the append and insert methods return a reference to the same
StringBuilder object. This makes it easy to chain multiple operations.
For example, instead of writing sb.append(1); sb.insert(0,
2); , you can write sb.append(1).insert(0, 2);
Here are a few examples of how the append methods work:
StringBuilder sb = new StringBuilder(100); //creati

ng an empty StringBuilder with an initial capacity of
100 characters

sb.append(true); //converts true to string "true" a
nd appends it to the existing string
System.out.println(sb); //prints true

sb.append(12.0); //converts 12.0 to string "12.0" a
nd appends it to the existing string
System.out.println(sb); //prints true12.0

sb.append(new Object()); //calls toString on the ob
ject and appends the result to the existing string
System.out.println(sb); //prints true12.0java.lang.
Object@32943380
And here are a couple of examples to illustrate the insert methods:
StringBuilder sb = new StringBuilder("01234");

sb.insert(2, 'A'); //converts 'A' to string "A" and
inserts it at index 2
System.out.println(sb); //prints 01A234
sb.insert(6, "hello"); //inserts "hello" at index 6
System.out.println(sb); //prints 01A234hello
In the above code, observe the location at which the string is being inserted. As
always, since indexing starts with 0, the first position at which you can insert a
string is 0 and the last position is the same as the length of the existing string. If
your position argument is negative or greater than the length of the existing
string, the insert method will throw
an StringIndexOutOfBoundsException .
The rest of the methods are quite straightforward and work as indicated by their
names. To make them easy to remember, I have categorized them into two
groups - the ones that return a self-reference (i.e. a reference to the same
StringBuilder object on which the method is invoked), which implies they can be
chained, and the ones that do not.
Methods that return a self-reference are - reverse() , delete(int
start, int end) , deleteCharAt(int index) , and
replace(int start, int end, String replacement) .
Remember that start index is always inclusive and end index is always
exclusive , so, the following code will print 0abcd34 and 0cd34 .
sb.replace(1, 3, "abcd"); //replaces only the chars

at index 1 and 2 with "abcd"
System.out.println(sb); //prints 0abcd34

sb.delete(1, 3); //deletes only the chars at index
1 and 2
System.out.println(sb); //print 0cd34
Methods that cannot be chained are int capacity() , char

charAt(int index) , int length() , int indexOf(String
str) , int indexOf(String str, int startIndex) , void
setLength(int len) , String substring(int start) , String
substring(int start, int end) , and String toString() .
The setLength method is interesting. It truncates the existing string contained

in the StringBuilder to the length passed in the argument. Thus,
sb.setLength(2); will truncate the contents of sb to 01 .
delete vs substring ☝
It is important to understand the difference between the delete and the substring
methods of StringBuilder. The delete methods affect the contents of the
StringBuilder while the substring methods do not. This is illustrated by the
following code:
String str = sb.substring(0, 2);

System.out.println(str+" "+sb);

StringBuider sb2 = sb.delete(0, 2);
System.out.println(sb2+" "+sb);
The above code prints 01 01234 and 234 234 .
Not important for the exam but you should be aware

that prior to Java 1.5, the Java standard library only
had the java.lang.StringBuffer classStringBuffer
class to deal with mutable strings. This class is thread
safe, which means it has a built-in protection
mechanism that prevents data corruption if multiple
threads try to modify its contents simultaneously.
However, the Java standard library designers realized
that StringBuffer is often used in situations where
this protection is not needed. Since this protection
incurs a substantial performance penalty, they added
java.lang.StringBuilder in JDK 1.5, which provides
exactly the same API as StringBuffer but without the
thread safety features.
You may see old code that uses StringBuffer but

unless you want to modify a string from multiple
threads, you don't need to use StringBuffer . Code
with StringBuffer will run a little slower than the
code that uses StringBuilder .
5.3 Exercise ☝
1. Write code to determine whether the toString and substring
methods of StringBuilder and String classes return an interned
string or not. Confirm your results by checking the JavaDoc API
descriptions of these methods.
2. Write a method that takes a String and returns a String of the same
length containing the 'X' character in all positions except the last 4
positions. The characters in the last 4 positions must be the same as in the
original string. For example, if the argument is "12345678" , the return
value should be "XXXX5678" .
3. Implement the same method as above but with a StringBuilder as the
input parameter.
4. Write a method that takes a String[] as an argument and returns a
String containing the concatenation of all the strings in the input array.
Invoke your method with different arguments. Make sure that your code
handles the cases where the argument is null, contains a few nulls, or
contains only nulls. Is this a good place to make use of a
StringBuilder ?
Chapter 6 Using Operators
precedence
6.1 Java Operators ☝

A program is nothing but an exercise in manipulating the data represented by
variables and objects. You manipulate this data by writing statements and
expressions with the help of operators. In that respect, operators are kind of a
glue that keeps your code together. You can hardly write a statement without
using any operator. Something as simple as creating an object or calling a
method on an object requires the use of an operator (the new operator and the
dot operator!). It is therefore, important to know what all operators does Java
have and to understand how they work.
6.1.1 Overview of operators available in Java ☝
Java has a large number of operators. They can be classified based on the type of
operations they perform (arithmetic, relational, logical, bitwise, assignment,
miscellaneous) or based on the number of operands they require (unary, binary,
and ternary). They may also be classified on the basis of the type of operands on
which they operate, i.e., primitives (including primitive wrappers) and objects.
While, as a Java programmer, you should be aware of all of them, for the
purpose of the exam, you can ignore a few of them. The following sections
provide a brief description of all the operators. The ones that are not required for
the exam are noted as such.
Arithmetic Operators
Arithmetic operators are used to perform standard mathematical operations on
all primitive variables except boolean. They can also be applied to wrapper
objects for numeric types (i.e. Byte, Short, Character, Integer,
Long, Float, and Double ) due to auto-unboxing.
Operator(s) Brief description and Examples

+, -, *, / Addition, subtraction, multiplication, and division .
(Binary)
Example:
int a = 10;
Integer b = 100; //using primitive wrapper h
ere
int c = a + b;
% Modulus operator - returns the remainder of the division of first

(Binary) operand by the second one.
Example:
int a = 10;
int b = 3;
int c = a % b;
c is assigned a value of 1 because when 10 is divided by 3 , the

remainder is 1 .
Here is another example:
Integer i = 10;
Character c = 'a';
System.out.println((i%c)); //prints 10
The above example illustrates that these operators work on

wrapper objects including Character . Don't worry, you will not
be required to perform mathematical calculations involving the
modulus operator in the exam. But as an exercise, you should try
to find out why the above code prints 10 .
- Unary minus - returns a negated value of a literal value or a

(Unary) variable without changing the value of the variable itself.
A unary plus may also be used on a literal or a variable but it is not

really an operator because it doesn't do anything.
Examples:
Using - on a literal :
int a = -10; //assigns -10 to a
Using - on a variable:
int b = -a;
Here, b is assigned the negated value of a , i.e., -(-10) , i.e., 10
. a remains -10 .
Using + on a variable:
int c = +a;
This is valid but will not assign 10 to c . It will assign -10 to c
because a is -10 . a remains -10 as well.
++, -- Unary increment and decrement operators - Unlike the unary

(Unary) minus operator, these operators can only be used on a variable and
they actually change the value of the variable on which they are
applied.
Also unlike the unary minus, they can be applied before (pre) as
well as after (post) the variable. I will explain the difference
between pre and post later.
Examples:
int a = 10;
int b = -10;
Post increment:
a++; //a is incremented from 10 to 11
b++; //b is incremented from -10 to -9
Pre increment:
++a; //a incremented from 11 to 12
++b; //b is incremented from -9 to -8
(-- works the same way)

Relational Operators
Relational operators are used to compare integral and floating point values. They can also be a
unboxing.
Operator(s) Brief description and Exa

<, >, <=, >= Less than, greater than, less than or equal to, and greater than or equal to
(Binary) numeric types and return a boolean value.
Example:
int a = 10;
Integer b = 100; //using primitive wrapper here
boolean flag = a < b; //flag is assigned a value of tr

s than the value of b.
==, != Equal to and Not equal to - These operators are a bit special because they work
(Binary) boolean as well) and reference types. When used on two primitive values or a pr
the two values are same or not.
Example:
int a = 10; Integer b = 20; char ch = 'a'; Double d =

System.out.println(a == b); //comparing an int with an
t equal to 20
System.out.println(a == 10.0); //comparing an int with

siders 10 and 10.0 as equal
System.out.println(a == ch); //comparing an int with a

qual to 'a'
System.out.println(97 == ch); //comparing an int with

f 'a' is indeed 97
System.out.println(a != d); //comparing an int with a

ve the same value
System.out.println(a != 10); //comparing two ints, pri
System.out.println(false != flag); //comparing a boole

flag is false
You cannot compare a numeric value and a non-numeric value such as a

the reference is to a primitive wrapper, of course) or even two references of "diff
numeric value can never be the same as a boolean value or a reference to a non-n
such nonsensical comparison, the compiler deems it to be a coding error. Therefo
System.out.println(10 == false); //can't compare a num
Object obj = new Object();

System.out.println(obj != 10); //can't compare a refer
System.out.println(obj == true); //can't compare a ref
Integer INT = 10;

Double D = 10.0;
System.out.println(INT == D);//can't compare an Intege
When used on references, == and != check whether the two references point to
Object o1 = new Object();

Object o2 = o1;
boolean e = (o1 == o2); //e is assigned a value of tru
e object in memory.
o2 = new Object();
System.out.println(o1 == o2); //prints false because o
ects
String s1 = "hello";
System.out.println(s1 == s2); //prints true because s1
Comparison of references using == looks straight forward but it is a source of tri

two parts. One that deals with their usage on String references and one that deals
first part in this chapter and the second part in the "Working with Inheritance - II
Logical Operators
Logical operators are used to form boolean expressions using boolean variables and values.

&&, || Short circuiting "and" and "or". They return a boolean value.
(Binary)
Example:
boolean iAmHungry = false;

boolean fridgeHasFood = false;
boolean eatUp = iAmHungry && fridgeHasFood; //eat if y
nd if there is food in the fridge
boolean tooMuchExcitement = true;

boolean eatAnyway = eatUp || tooMuchExcitement;
e or if there is too much excitement in the air!
They are called short circuiting operators because they avoid evaluating parts o
value of that part does not make any difference to the final value of the expressio
evaluation of the second operand is "conditional ". It is evaluated only if it is req
how it works with the example I gave above. You eating food depends upon two
hungry and there being food in the fridge. Now, if you are not hungry, would you
fridge to see if there is food in it or not? Of course not. Since you are not hungry
decide that you won't eat food irrespective of whether there is food in the fridge
second part of the expression, i.e., the check for fridgeHasFood
not evaluated) if the first part, i.e., check for iAmHungry is false.
Similarly, if you are hungry, do you still need excitement in the air to eat food? O
you are hungry, you can decide right there to eat food irrespective of whether the
air or not. Therefore, even here, the second part of the expression, i.e., the check
tooMuchExcitement can be short circuited if the first part, i.e.,
Short circuiting behavior is helpful in cases where parts of an expression are too
evaluate. Think of the above example again. Would you get up and walk up to th
it is empty or not when you are not hungry? Nah, it is too much of an effort, righ
expression such as iAmHungry && checkFridge() , where checkFrid
returns true or false depending on whether there is food in the fridge or not,
invoked if iAmHungry is false . Similarly, evaluating some conditions, such
looking up the database, may be too time consuming and it may be desirable to a
if their value doesn't make a difference to the final value of the expression.
You need to understand this behavior very clearly because it gets exploited a lot
expressions in professionally written code.
&, | Non-Short circuiting "and" and "or" (| is also known as inclusive or)
(Binary)
Example:
boolean iAmHungry = false;

boolean fridgeHasFood = false;
boolean eatUp = iAmHungry & fridgeHasFood; //eat if yo
d if there is food in the fridge
boolean tooMuchExcitement = true;

boolean eatAnyway = eatUp | tooMuchExcitement;
or if there is too much excitement in the air!
They are actually bitwise operators and are mostly used to operate on numeric ty
be used on boolean values just like && and || . The only difference between the
short circuit any part of an expression.
This behavior is useful in cases where parts of an expression has side effects that
avoid even if their value is irrelevant to the final value of the expression. For exa
following expression - boolean accessGranted = authenticateUse
logToAudit(userid); . Now, it is possible to decide that access has not to
is not authenticated irrespective of what logToAudit method returns. Howeve
to make sure every request for access is logged. Thus, you may want the
invoked irrespective of whether authenticateUser method returns
in this case is appropriate because if you use && instead of & , logToAudit
invoked if authenticateUser method returns false .
^ Xor aka Exclusive Or - Just like & and | , this is also a bitwise operator but wh
(Binary) boolean operands, it returns true if and only if exactly one of the operands is
Example:
boolean a = false;
boolean b = true;
boolean c = a^b; //c is assigned the value true
The question of short-circuiting does not arise here because both the operands ha
determine the result. In other words, it can never short-circuit an expression.
! Negation - This operator returns the compliment of given a boolean

(Unary)
Example:
boolean hungry = false;

boolean stuffed = !hungry; //assigns true to stuffed
:? Ternary - To be precise, "ternary" is not really the name of this operator. "Terna
(Ternary) since this is the only operator in Java that requires three operands, it is convenien
"ternary" operator. This operator is kind of a short form for the if-else statement
meaningful name. It evaluates either the second or the third operand depending o
operand. For example:
int a = 5;
String str = a == 5 ? "five" : "not five";
System.out.println( str ); //prints five
There are quite a few rules that govern the type of a ternary expression and the ty
will discuss them in the next chapter along with if/if-else .
Assignment O
Assignment operators are used assign the value of an expression given on the right-hand side
%= , += , -= , <<= , >>= , >>>= , &= , ^= , and, |= . The first one, i.e., = is the simple assig
Operator(s)
= Simple assignment - It simply copies the value on the left to the variable on the
(Binary) of references, it is the value of the reference (not the actual object pointed to by t
Example:
byte b1 = 1; //assign 1 to variable b1
Object o1 = "1234"; //assign the address of the locati
Object o2 = o1; //assigns the value contained in o1 to
//Thus, o2 starts pointing to the same memory address

//Note that there is only one instance of the String c
(If you are not clear about the difference between an object and a reference, I sug
Beginners" chapter.)
*=, /=, %=, Compound assignment - These operators are called compound assignment oper
+=, -=, <<=, and then assign the result of the operation to the variable on the left. These opera
>>=, >>>=, assignment operator, these operators do not apply to boolean and reference types
&=, ^=, |=
(Binary) Example:
int i1 = 2; //bit pattern of 1 is 00000001
int i2 = 3;
i2 *= i1; //assigns the value of i2*i1, i.e., 3*2 to i
byte b1 = 8;
b1 /= 2; //assigns 4 to b1
The easiest way to understand how these operators work is to expand them ment
can be expanded to i2 = (int) (i2 * i1); Similarly, b1 /= 2;
Notice the explicit cast in the expanded form. I will explain its reason later while
The += operator is overloaded to work with Strings as well. It combines String c
Example:
String s = "hello";
s += " world"; //creates a new String "hello world" an
Just like the other compound assignment statements mentioned above, the easies
s = s + " world";
While the primary function of an assignment operator is quite simple, there are a few nuances
1. They are all right associative , which means a = b = c = 10; will be evaluated as
separately in the "Operator precedence and evaluation of expressions" section.
2. The left-hand operand of these operators must be a variable. It can either be a named variab
example, an array element). Thus, you cannot do something like 10 = b; because
like aMethodThatReturnsAnObject() = 20; either because a method returns value
implication of the fact that Java does not have "pass-by-reference" .
3. The right-hand operand of these operators must be an expression whose type must be the sa
cannot assign boolean expression to an int variable. In case of primitive types, compatibil
numeric type if the type of the value fits within the range of the type of the variable. Otherwis
chapter. Compatibility in case of references is a bit complicated. I will talk about this more wh
Bitwise Operators (Not required for the exam but good to

Bitwise operators are used to apply logical operations on individual bits of given numeric valu
objects). I will not discuss them in detail because they are not required for the exam. They are
(If you want to understand these operators better, try applying them to various numeric values
Integer.toBinaryString method to print out the bit pattern of any given numeric valu

&, |, ^ Bitwise "and", "or", and "xor"
(Binary)
Example:
byte b1 = 1; //bit pattern of 1 is 00000001

byte b3 = (byte) (b1 & b2); //b3 gets 0
byte b4 = (byte) (b1 | b2); //b4 gets 3
byte b5 = (byte) (b1 ^ b2); //b5 gets 3
(I will explain the reason for explicit casting of the results to byte later while dis
~ Bitwise complement - It toggles, i.e., turns a 0 to 1 and a 1 to 0, individual bits o

(Unary) primitive numeric wrapper objects)
Example:
byte b2 = (byte) ~b1; //b2 gets 11111110, which is -2.
>>, << Bitwise signed right and left shift - They shift the given bit pattern of the left o
(Binary) of places specified by the right operand while keeping the sign of the number sam
Example:
byte b1 = -4; //bit pattern of 4 is 11111100
byte b2 = 1;
byte b3 = (byte) (b1 >> b2); //b3 gets 11111110, i.e.,
byte b4 = (byte) (b1 << b2); //b4 gets 11111000, i.e.,
Observe that when we shifted the bits of -2 to right by 1 place, the sign bit (i.e. th
>>> Bitwise unsigned right shift - This operator works the same way as
(Binary) pushes in zeros from the left irrespective of the sign of the number. Thus, a nega
Example:
int i = -4; //bit pattern of -4 is 11111111 11111111 1
int i2 = 1;
int i3 = i1 >>> i2; //i3 gets 01111111 11111111 111111
646
Observe that when we shifted the bits of -4 to right by 1 place, the sign bit (i.e. th
over. Also observe that I have used int instead of byte variables in this exam
casting from affecting the result. Again, this is not required for the exam
byte variables instead of int and compare the values of i3 using
Miscellaneous
Operator(s) Brief description and Examp
+ String concatenation - + operator can also be used to concatenate two strings to
(Binary)
Example:
String s2 = " world";
String s = s1 + s2; //creates a new String "hello worl
The above example illustrates the most straight forward use of the + operator. Ho
versatile and can be used to join any kind of object or primitive value with a Stri
. (dot) The dot operator - You have definitely seen it but most likely have not noticed
(Binary) variable or access a member of a class or an object, you use the dot operator. It d
nonetheless. It is applied to a reference to access the members of the object poin
It will always throw a NullPointerException if you apply it to a null refe

object)
instanceof instanceof - Although not required for the exam, a thorough understanding of t
(Binary) it is closely tied to polymorphism, a concept that I haven't yet talked about, I wil
instanceof is used to check whether an object pointed to by a reference varia
of this operator. It returns true if the object pointed to by the reference variable
given on the right and false otherwise.
Example:
Object obj = "hello"; //declared type of obj is Object
boolean isString = (obj instanceof String); //assigns

nt to a String object
boolean isNumber = (obj instanceof Number); //assigns

t point to a Number object
( ) The cast operator - This operator can be used on numeric values and references
(Binary) while assigning values of a larger type to variables of a smaller type. For exampl
When used on a reference, it casts the reference of one type to another. I will dis
polymorphism and instanceof operator.
-> The lambda operator - This operator is used to write lambda expressions. I wil
(Binary)
[ ] The array access operator - This operator is used access the elements of an arra
(Binary) Java Data Types" chapter.
6.1.2 Expressions and Statements ☝
Difference between an Expression and a Statement ☝
Before I move on to the intricacies of various operators, let me talk a bit about
expressions and statements . This will give you a good perspective on how
various operators work and why they work so.
Wikipedia defines an expression as a combination of one or more explicit

values, constants, variables, operators, and functions that the programming
language interprets (according to its particular rules of precedence and of
association) and computes to produce ("to return", in a stateful environment)
another value.
The point to note here is that an expression has a value . This value could be a
primitive value or a reference. You can combine these values together using
various operators to create even bigger expressions as per the rules of the
language. You can say that if something has a value that can be assigned to a
variable, then that something is an expression.
A statement , on the other hand, is a complete line of code that may or may not
have any value of its own. You cannot combine statements to produce another
statement. You can, of course, write one statement after another to create a
program .
Thus, an expression may be a statement on its own. For example, consider the
following code:

int a = 10;
int b = 20;
a + b; //this line will not compile
a = a + b;
}
}
In the above code, int a = 10; is a statement. int b = 20; is another

statement. But neither of them are expressions because they don't have a value of
their own. You cannot assign int a = 10 to a variable. On the other hand, a
+ b is an expression but is not a valid statement. If you try to compile the above
code, you will get an error saying "Not a statement" at line a + b;
However, a = a + b; is a valid statement as well a valid expression. It is, in

fact, an expression made by combining two expressions - a and a + b using
the = operator. Furthermore, a + b is also an expression made by combining
two expressions - a and b using the + operator.
The question that should pop into your head now is, if a = a + b is a valid
expression, does it have a value? Yes, it does. As a matter of fact, the value of an
expression built using the assignment operator is the same as the value that is
being assigned to the variable on the left side of the = operator. In this case, for
example, a is being assigned a value of a + b . Thus, the value of the
expression a = a + b is the value produced by the expression on the right of
= operator, i.e., a + b . You can actually test it out by assigning this whole
expression to another variable like this - int k = (a = a + b);
What values can be combined using which operators is really the subject of this
chapter.
6.1.3 Post and Pre Unary Increment/Decrement Operators ☝
The ++ and -- operators can be applied to a variable in the postfix form (i.e.
appearing after the variable) or in the prefix form (i.e. appearing before the
variable). In both the cases, the value of the variable will be incremented (or
decremented) by 1 . The difference between the two is that the postfix operator
returns the existing value of the variable while prefix operator returns the
updated value of the variable. This is illustrated in the following code:
int i = 1, post = 0, pre = 0;

post = i++;
System.out.println(i+", "+post); //prints 2, 1
i = 1; //resetting to i back to 1.
pre = ++i;
System.out.println(i+", "+pre); //prints 2, 2
To understand this, you need to look at process of evaluation of the expression

post = i++ step by step. The expression post = i++ ; is composed of a
two parts - the variable post and the expression i++ - joined using the
assignment (= ) operator. To evaluate this expression, you need to first evaluate
the expression that is on the right side of = , i.e., i++ .
Now, since a postfix operator increments the variable but returns the existing
value of the variable, and since the existing value of the variable i is 1 , the
expression i++ evaluates to 1 even though the value of i has been incremented
to 2 . Thus, the variable post is assigned a value of 1 .
Let us follow the same process for evaluating the expression pre = ++i.
Here, since the prefix operator increments the variable and returns the updated
value of the variable, and since the updated value of the variable i is 2 , the
expression ++i evaluates to 2 . Thus, the variable pre is assigned a value of 2 .
You need to appreciate the fact that the value of the variable i and the value of
the expression i++ (or ++i ) are two different things and they may or may not
be the same depending on whether you use pre or post form of the
increment/decrement operator. When you do int x = ++i; you are not
assigning the value of the variable i to x . You are assigning the value of the
expression ++i to x independently from the process of applying ++ to i . If
you truly understand this concept then you should be able to determine the
output of the following code:
int i = 1;
i = i++;
System.out.println(i); //what will this line print?
postfix and prefix for the exam ☝
In the exam, you may get questions where you are required to evaluate a
compound expression containing multiple pre/post increment operators. They do
get tricky sometimes but the key to solving such problems is to apply the
concept explained above without losing focus. I will now show you a
representative question and the steps to work out the answer.
Q. What will the following code print?
int a = 2;
int b = 5;
int c = a * (a++ - --b) * a * b;
System.out.println(a+" "+b+" "+c);
1. We start with putting the values of the variables in the expression from left to
right. Since the value of a at the beginning of the evaluation is 2 , the expression
becomes:
c = 2 * (a++ - --b) * a * b
2. Next, a is incremented using the post increment operator. Therefore, the value
of a used in the expression will be the existing value of a , i.e., 2 and then a is
incremented to 3 . Therefore, the expression now becomes:
c = 2 * (2 - --b) * a * b ; (a is 3 now)
3. b is decremented using the pre decrement operator. Therefore, b is

decremented first to from 5 to 4 and then the new value of b , i.e., 4 is used in
the expression. Therefore, the expression becomes:
c = 2 * (2 - 4) * a * b; (a is now 3 and b is now 4 )
4. There are no further operations left to be applied on a and b anymore so, their
values can be substituted in the expression. The expression now becomes:
c = 2 * (2 - 4) * 3 * 4;
5. Now you can apply the usual rules of operator precedence and brackets to
evaluate the expression. You can easily see that the value of 2 * (2 - 4) *
3 * 4 is -48 , which is the value that is assigned to the variable c .
6. Thus, the print statement will print 3 4 -48
The key point that you should observe in the evaluation process described above
is how the value of the variables change while the expression is being evaluated
and their impact on the expression. For example, the value of variable a changed
from 2 to 3 while the expression was being evaluated. This change causes part
of the expression to use old value of a while the subsequent part to use the
updated value of a . Although a similar change is not noticeable in case of b in
this expression because b is decremented using pre decrement operator instead
of post decrement, had there been a use of b in the expression before
encountering --b , the old value of b would have been used just like in the case
of a .
To test this theory, try to compute the value of this slightly modified expression:
c = b * a * (a++ - --b) * a * b
When to use postfix and when to use prefix? ☝
As shown in the code above, the difference between the two is very subtle and is
material only when you use the unary increment and decrement operators inside
of another expression. So, the obvious answer is to use postfix when you want to
use the existing value of the variable and then update it, and use prefix when you
want to update the value first and then use its value.
Ideally, however, you should avoid using these operators in a compound

expression altogether because they cause confusion and are a common source of
hard to find bugs in the code. My advice is to get into the habit of using just one
form of increment or decrement operator in all of your code.
Using unary increment/decrement operators on wrappers ☝
You can use ++ and -- operators on any numeric primitive wrapper and they
work the same way as explained above. You should, however, remember that
primitive wrapper objects are immutable and therefore, incrementing or
decrementing a wrapper object does not change that particular wrapper object.
Instead, a new wrapper object is assigned to the reference on which increment or
decrement operator is applied. The following code illustrate this point:
Integer i = 1;
Integer j = i; //now, i and j point to the same In
teger object
i++; //a different Integer object containing 2 is

assigned to i
//j still points to the same Integer object contai

ning 1
System.out.println(i+" "+j);//prints 2 1
6.1.4 String concatenation using + and += operators ☝
The + operator is quite versatile. Besides performing the mathematical addition

of numeric operands, it can also "add" two Strings together to create a new
String. In that sense, you can say that the + operator is overloaded because its
behavior changes based on the type of operands. When both of its operands are
numeric values (or their primitive wrappers), it performs the mathematical
addition but if either of its operands is of type String, it performs the String
addition. By the way, the technical term for String addition is "concatenation"
and so, I will use this term from now onward.
To trigger the String concatenating behavior of the + operator, the declared type
of at least one of its operands must be a String. If one of the operands is a String
and the other one is not, the other operand will be converted to a String first and
then both the operands will be concatenated to produce the new String. Let us
see a few examples to make it clear:
String s1 = "hello" + " world"; //both the operands ar

e Strings
System.out.println(s1); //prints "hello world"
String s2 = "hello " + 1; //first operand is a String

and second is an int
System.out.println(s2); //prints "hello 1"
Double d = 1.0;
String s = "2";
String s3 = d + s; //first operand is a Double and sec
ond is a String
System.out.println(s3); //prints "1.02"
There is no restriction on the type of the non-String operand. It can be of any

type. The interesting part is how the non-String operand is converted to a String.
The answer is simple. Recall that java.lang.Object is the root class of all objects
in Java. This class contains a toString method that returns a String
representation of that object. The + operator invokes this toString method on
the non-String operand to get a String value. If the non-String operand is a
primitive, then the primitive value is first converted to its corresponding wrapper
object and toString is invoked on the resulting wrapper object. The following
program illustrates this process.

String str = tc.toString();
System.out.println( str );
System.out.println( "hello " + tc );
}
}
This program produces the following output:
TestClass@15db9742
hello TestClass@15db9742
As you can see, the TestClass@15db9742 part of the concatenated String

on the second line of the output is the same as the String returned by the toString
method.
The + operator has one more trick up its sleeve. Check out this code:
TestClass tc = null;
System.out.println( "hello " + tc );

}
}
It prints hello null . Normally, when you invoke any method using a null
reference, the JVM throws a NullPointerException . But in this case, no
NullPointerException was thrown. How come? The reason is that the + operator
checks whether the operand is null before invoking toString on it. If it is null, it
uses the String "null" in place of that operand.
String concatenation using += operator ☝
The += operator works in the same way as the + operator but with the additional
responsibility of an assignment operator. The operand on the left of += must be a
String variable but the operand on right can be a value or variable of any type.
For example,
String str = "2";

str += 1; //this is the same as str = str + 1;
System.out.print(str); //prints "21"
Here is another example:

String str = null;
str += tc; //same as str = str + tc;
System.out.print( str ); //prints nullTestClass@15

db9742
}
}
Here, even though str is null and the second operand to += is not a String,
the String concatenation behavior of += will be triggered because the declared
type of str is String . Thus, toString will be called on the non-string
operand tc . Furthermore, since str is null , the String "null" will be used
while concatenating str and the string value returned by the call to
tc.toString() .
You need to keep in mind that compound assignment operators do not work in
declarations. For example, the statement String str += "test"; will not
compile because str is being declared in this statement.
6.1.5 Numeric promotion and casting ☝
Take a look at the following code:

byte b = 1;
short s = -b;
System.out.println(b);
}
}
Looks quite simple, right? One may believe that it will print -1. But the fact is
that it doesn't compile! It produces the following error message:
TestClass.java:4: error: incompatible types: possible
lossy conversion from int to short
short s = -b;
^
1 error
There is no use of int anywhere in the code, yet, the error message is talking
about converting something from int to short . The reason is that Java
applies the rules of "numeric promotion" while working with operators that
deal with numeric values, which are, in a nutshell, as follows:
1. Unary numeric promotion - If the type of an operand to a unary operator is

smaller than int , the operand will automatically be promoted to int before
applying the operator.
2. Binary numeric promotion - Both the operands of a binary operator are

promoted to an int but if any of the operands is larger than an int , the other
operand is promoted to the same type as the type of the larger one. Thus, for
example, if both the operands are of type byte , then both are promoted to int
but if any of the operands is a long, float, or double , the other one is
also promoted to long, float, or double , respectively. Similarly, if the
operands are of type long and float , long will be promoted to float .
A direct implication of the above two rules is that the result of applying an
operator to numeric operands is of the same type as the type of the larger
operand but it can never be smaller than an int .
If you look at the above code in light of these rules, the error message is quite
obvious. Before applying the unary minus operator on b , the compiler promotes
b to int . The result of applying the operator is, therefore, also an int . Now,
int is a larger type than short and the compiler is concerned about possible
loss of information while assigning a int value to a short variable and so, it
refuses to compile the code. The language designers could have easily allowed
such assignments without a fuss, but an inadvertent loss of information is often a
cause of bugs, and so, they decided to make the programmer aware of this issue
at the compile time itself.
You may wonder here that the value of b is 1 and the result of -b is just -1 ,
which is well within the range of short . Then what's the problem? What loss
of information is the compiler talking about? You need to realize that compiler
does not execute any code. It can't take into account the values that the variables
may take at run time, while making decisions. So, even though you know that
the result of the operation is small enough to fit into a short , the compiler
cannot draw the same inference at compile time.
Let us now take look at few more examples where numeric promotion plays a
role:
short s1 = 1;
byte b1 = 1;
byte b2 = 2;
short s2 = +s1; //won't compile because the result wil

l be an int
byte b = s1 + 2; //won't compile because the result w

ill be an int
b = b1 & b2; //won't compile because the result will

be an int
s2 = s1 << 1; //won't compile because the result will

be an int
s2 = s1 * 1; //won't compile because the result will b

e an int
float f = 1.0f; //recall that to write a float literal

you have to append it with an f or an F
double d = 1.0;
int x = f - 1; //won't compile because the result will

be a float
float f2 = f + d; //won't compile because the result w

ill be a double
All of the above operations are affected by the rule of numeric promotion and
therefore, fail to compile. To make them compile, you need to assure the
compiler that you know what you are doing, and that you are okay with the
possible loss of information if there is any. You give this assurance to the
compiler by explicitly casting the result back to the type of target variable. I have
already discussed casting of primitives in the "Working with Java Data Types"
chapter. The following code is the fixed version of the code shown above:
short s1 = 1;
byte b1 = 1;
byte b2 = 2;
short s2 = (short) +s1;

byte b = (byte) (s1 + 2);
b = (byte) (b1 & b2); //numeric promotion happens for
bit-wise operators as well
s2 = (short) (s1 << 1);

s2 = (short) (s1 * 1);
float f = 1.0f; //recall that to write a float literal

you have to append it with an f or an F
double d = 1.0;
int x = (int) (f - 1);
float f2 = (float) (f + d);
An important point to note here is these rules come into picture only when the
operands involve a variable and not when all the operands are constants and the
result of the operation lies within the range of the target variable. The need for
promotion does not arise while dealing with constants because the values are
known at compile time and therefore, there is no possibility of loss of
information at run time. If the value produced by the constants doesn't fit into the
target variable, the compiler will notice that and refuse to compile it. Thus, the
statement byte b = 200 - 100; will compile fine because 200 and 100
are compile time constants and the result of the operation fits into a byte even
though one operand of the - operator falls outside the range of byte. But byte
b = 100 + 100; will not compile because the result of 100 + 100 cannot
fit into a byte . Similarly, the following code will also compile without any
error:
final int I = 10;

byte b = I + 2;
No cast is needed because I is a compile time constant and the result of I + 2

fits into a byte .
Curious case of unary increment/decrement and compound assignment

operators ☝
To put it simply, the rules of numeric promotion do not apply to ++, -- and
the compound assignment operators such as +=, -=, and *= . They are the
exceptions to the rules of numeric promotion. Thus, the following statements
will compile fine without any explicit cast.
byte b1 = 1;
byte b2 = ++b1; //result of ++b1 will be a byte
b2 = b1--; //result of b1-- will be a byte

b1 *= b2; //result will be a byte
double d = 1.0;
float f = 2.0f;
f += d; //result of will be a float
Observe the statements b1 *= b2; and f += d; . They behave as if they are

the short hand for b1 = (byte)(b1*b2); and f = (float)(f+d); In
other words, the result of a compound assignment operation is implicitly cast
back to the target type irrespective of the type of the second operand. Similarly,
++b1; behaves like b1 = (byte) (b1 + 1);
One could certainly make the "source of bugs" argument in the case of
compound assignments as well. I guess, convenience superseded that argument
in this case :)
Numeric Promotion and Primitive Wrapper Objects ☝
Remember that to apply an operator to wrapper objects, they have to be unboxed

first into their respective primitive values. Thus, the rules of numeric promotion
and their exceptions will come into play here as well. Therefore, for example,
Byte bW = 1; bW = -bW; will not compile but --bW; will compile fine
due to the presence of an implicit cast as explained before.
The rules differ a bit in case of final wrapper variables though. For example, the
following code will not compile:
final Byte b1 = 10;

Byte b2 = -b1;//will not compile even though b1 is fin
al
This doesn't work because b1 is a reference to an object. It is this reference that
is final, not the contents of the object to which it points. The compiler doesn't
know that Byte is immutable. Thus, as far as the compiler is concerned, it is not
sure that the result of -b1 will fit into a byte and so, it refuses to compile the
code.
6.1.6 Operator precedence and evaluation of expressions ☝
I am sure you have come across simple mathematical expressions such as 2 + 6 /

2 at school. You know that this expression evaluates to 5 and not 4 because
division has higher precedence than addition and so, 6 / 2 will be grouped
together instead of 2 + 6. To change the default grouping, you use brackets (aka
parentheses), i.e., (2+6)/2. You have most likely also come across the acronym
BODMAS (or PEDMAS, in some countries), which stands for
Brackets/Parentheses, Orders/Exponents, Division, Multiplication, Addition, and
Subtration). It helps memorize the conventional precedence order in which
brackets are evaluated first, followed by powers, and then the rest in that order.
Java expressions are not much different from mathematical expressions. Their
evaluation is determined by similar conventions and rules. The only problem is
that there are a lot of operators to worry about and, as we saw earlier, the
operators are not just mathematical. They are logical, relational, assignment, and
so many other types as well. This makes Java expression evaluation a lot more
complicated than a mathematical expression. But don't worry, you will not be
required to evaluate complicated expressions in the exam. But you still need to
know a few basic principles of expression evaluation to analyze code snippets
presented in the questions correctly.
Precedence ☝
Precedence determines which operator out of two is evaluated "first", in a

conceptual sense. Another way to put it is, precedence determines how tightly an
operator binds to its operands as compared to the other applicable operator in an
expression. For example, in the case of 2 + 6 / 2 , the operand 6 can be
bound to + or to / . But the division operator, having higher precedence than
addition operator, binds to an operand more tightly than the addition operator.
The addition operator, therefore, is not able to get hold of 6 as its second operand
and has to wait until the division operator is done with it. In an expression
involving multiple operators, the operator with highest precedence gets the
operand, followed by the operator with second highest precedence, and so on.
The following table shows the precedence of all Java operators:
Operators Precedence
member and array access operators . and [ ]
cast ()
postfix expr++ expr--
unary ++expr --expr +expr -expr ~ !
multiplicative * / %
additive + -
shift << >> >>>
relational < > <= >= instanceof
equality == !=
bitwise AND &
bitwise exclusive OR ^
bitwise inclusive OR |
logical AND &&
logical OR ||
ternary ? :
assignment = += -= *= /= %= &= ^= |= <<= >>= >>>=
lambda ->
An important thing to observe from the above table is that the access operator
and the cast operators have the highest precedence among all while the
assignment operators and the lambda operator have the lowest precedence
among all.
This explains why the following code doesn't compile:
int i = 0;
byte b = (byte) i + 1;
Since the cast operator has higher precedence than the + operator, i is first cast to
byte and then the addition is performed. The end result, therefore, is an int
instead of a byte . You need to put i + 1 in parentheses like this: byte b =
(byte)(i + 1);
Associativity ☝
Associativity of operators determines the grouping of operators when an

expression has multiple operators of same precedence. For example, the value of
the expression 2 - 3 + 4 depends on whether it is grouped as (2 - 3) +
4 or as 2 - (3 + 4) . The first grouping would be used if - operator is left-
associative and the second grouping would be used if - operator is right-
associative . It turns out that operators are usually grouped in the same fashion
in which we read the expression, i.e., from left to right . In other words, almost
all of the operators in Java are defined to be left-associative . The only
exceptions are the assignment operators (simple as well as compound) and the
ternary operator. Thus, the expression 2 - 3 + 4 will be grouped as (2 -
3) + 4 and will evaluate to 3. But the expression a = b = c = 5; will be
grouped as a = ( b = (c = 5)) ; because the assignment operator is
right associative. Here is another example that shows the impact of associativity:
int i = 1;
String s2 = s1 + 1 + i;
System.out.println(s2); //prints hello11
The above code prints hello11 instead of hello2 because the + operator is
left-associative. The expression s2 = s1 + 1 + i; is grouped as s2 =
(s1 + 1) + i; . Thus, s1+1 is computed first, resulting in the string
hello1 , which is then concatenated with 1 , producing hello11 .
A programming language could easily prohibit ambiguous

expressions. There is no technical necessity for
accepting the expression 2 + 6 / 2 as valid when it
can be interpreted in two different ways. The only
reason ambiguous expressions are accepted is because
it is considered too onerous for the programmer to
resolve all ambiguity by using parenthesis when a
convention already exists to evaluate mathematical
expressions. Rules of Operator Precedence and
Associativity are basically a programming language
extension to the same convention that includes all sorts
of operators. You can, therefore, imagine that operator
precedence and evaluation order are used by the
compiler to insert parenthesis in an expression. Thus,
when a compiler see 2 + 6 / 2, it converts the
expression to 2 + ( 6 / 2 ), which is what the
programmer should have written in the first place.
You should always use parenthesis in expressions such

as 2 - 3 + 4 where the grouping of operands is not
very intuitive.
Parenthesis ☝
You can use parentheses to change how the terms of an expressions are grouped
if the default grouping based on precedence and associativity is not what you
want. For example, if you don't want 2 - 3 + 4 to be grouped as (2 - 3)
+ 4 , you could specify the parenthesis to change the grouping to 2 - (3 +
4) .
Evaluation Order ☝
Once an expression is grouped in accordance with the rules of precedence and

associativity, the process of evaluation of the expression starts. This is the step
where computation of the terms of the expression happens. In Java, expressions
are evaluated from left to right. Thus, if you have an expression getA() -
getB() + getC() , the method getA will be invoked first, followed by
getB and getC . This means, if the call to getA results in an exception,
methods getB and getC will not be invoked.
Java also makes sure that operands of an operator are evaluated fully before the
evaluation of the operator itself. Obviously, you can't compute getA() +
getB() unless you get the values for getA() and getB() first.
The important point to understand here is that evaluation order of an expression

doesn't change with grouping. Even if you use parentheses to change the
grouping of getA() - getB() + getC() to getA() - ( getB() +
getC() ) , getA() will still be invoked before getB() and getC() .
Let me show you another example of how the above rules are applied while
evaluating an expression. Consider the following code:

static boolean a ;
static boolean b ;
static boolean c ;
boolean bool = (a = true) || (b = true) && (c = t
rue) ;
System.out.println(a + ", " + b + ", "+ c );
}
}
Can you tell the output? It prints true, false, false . Surprised?
Many new programmers think that since && has higher precedence, (b =
true) && (c = true) would be evaluated first and so, it would print
true, true, true . It would be logical to think so in a Mathematics class.
However, evaluating a programming language expression is a two step process.
In the first step, you have to use the rules of precedence and associativity to
group the terms of the expression to remove ambiguity. Here, the operand (b =
true) can be applied to || as well as to && . However, since && has higher
precedence than || , this operand will be applied to && . Therefore, the
expression will be grouped as (a = true) || ( (b = true) && (c
= true) ) . After this step, there is no ambiguity left in the expression. Now,
in the second step, evaluation of the expression will start, which, in Java,
happens from from left to right. So, now, a = true will be evaluated first. The
value of this expression is true and it assigns true to a as well. Next, since
the first operand of || is true, and since || is a short circuiting operator, the
second operand will not be evaluated and so, (b = true) && (c =
true) will not be executed.
6.2 Exercise ☝
1. Work out the values of the variables after each of the following statements
on paper:
String str = "7" + 5 + 10;

str = 7 + 5 + "10";
str = "7" + (5 + 10);

int m = 50;
int n = ++m;
int o = m--;
int p = --o+m--;
int x = m<n?n<o?o<p?p:o:n:m;
int k = 4;
boolean flag = k++ == 5;
flag = !flag;
2. Which of the following lines will fail to compile and why? Write down the
value of the variables after each line.
byte b = 1;
b = b<<1;
int c = b<<1;
byte d += b;
byte e = 0;
e += b;
3. What will the following code print and why?
String s = "a";
String[] sa = { "a", s, s.substring(0, 1), new Str
ing("a"), ""+'a' };
for(int i=0; i<sa.length; i++){
System.out.println(s == sa[i]);
System.out.println(s.equals(sa[i]));
}
Chapter 7 Using Decision Constructs
Use Java control statements including:
if and if/else
switch
7.1 Create if and if/else constructs

7.1.1 Basic syntax of if and if-else ☝
The if statement is probably the most used decision construct in Java. It allows
you to execute a single statement or a block of statements if a particular
condition is true. If that condition is false, the statement (or the block of
statements) is not executed. The following are two ways you can write an if
statement:
if( booleanExpression ) single_statement;
Notice that there are no curly braces for the statement. The if statement ends
with the semi-colon. If you have multiple statements that you want to execute
instead of just one, you can put all of them within a block like this:
if ( booleanExpression ) {
zero or more statements;
}
Observe that there is no semi-colon after the closing curly brace. It is not an
error if present though.
If-else statement ☝
The if-else statement is similar to an if statement except that it also has an else
part where you can write a statement that you want to execute if the if condition
evaluates to false:
if( booleanExpression ) single statement;
//semi-colon required here
else single statement;
//semi-colon required here, if-else statement ends wi

th this semi-colon.
Again, if you have multiple statements to execute instead of just one, you can
put them within a block:
if ( booleanExpression ) single statement;
//semi-colon is required here
else {
}//no semi-colon required here, but not an error if it
exists.
or, if both the if and the else parts have multiple statements:
if ( booleanExpression ) {
} //must NOT have a semi-colon here, error if it exists
.
else {
}//no semi-colon required here, but not an error if it
exists.
If booleanExpression evaluates to true , the statement (or the block of

statements) associated with if will be executed and if the
booleanExpression evaluates to false , the else part will be executed.
Remember that an empty statement (i.e. just a semicolon) is a valid statement

and therefore, the following if and if-else constructs are valid:
boolean flag = true;
if(flag); //does nothing, but valid
if(flag); else; //does nothing, but valid
if(flag);else System.out.println(true); //does nothing

because flag is true
if(flag) System.out.println(true); else; //prints true

By the way, you may see if/if-else statements being called if-then/if-then-else
statement. This is not entirely correct. In some languages such Pascal, "then" is a
part of the syntax but it is not in Java. However, "then" is not a keyword in Java
and there is no "then" involved in the syntax of if/if-else .
7.1.2 Usage of if and if-else in the exam ☝
Let us now look at a few interesting ways if/if-else is used in the exam that
might trip you up.
Bad syntax ☝

if( flag )
else System.out.println("false"); //compilation error
In the above code, there is no statement or a block of statements for the if part. If
you don't want to have any code to be executed if the condition is true but want
to have code for the else part, you need to put an empty code block for the if part
like this:
boolean flag = false;

if( flag ) {
}
else System.out.println("false");
or

if( flag ) ; else System.out.println("false");
}
Instead of having an empty if block, it is better to negate the if condition and put
the code in the if block. Like this:

if( !
flag ) {
System.out.println("false");
}
Bad Indentation ☝
Unlike some languages such as Python, indentation (and extra white spaces, for
that matter) holds no special meaning in Java. Indentation is used solely to
improve readability of the code. Consider the following two code snippets:

if(flag)
else System.out.println("true");
{
}
and

if(flag)
else
System.out.println("true");
{
}
The above two code snippets are equivalent. However, since the second snippet
is properly indented, it is easy to understand that the last code block is not really
associated with the if/else statement. It is an independent block of code and will
be executed irrespective of the value of flag . This code will, therefore, print
true and false .
Here is another example of bad indentation:

int i = 0;
if(flag)
i = i +1;
else
i = i + 2;
The above code is trying to confuse you into thinking that there are two
statements in the if part and two statements in the else part. But, with proper
indentation, it is clear what this code is really up to:

int i = 0;
if(flag) i = i +1;
else i = i + 2;
It turns out that the last three lines of code are independent statements. The else
statement is completely out of place because it is not associated with the if
statement at all and will, therefore, cause compilation error.
Missing else ☝
As we saw earlier, the else part is not mandatory in an if statement. You can have
just the if statement. But when coupled with bad indentation, an if statement may
become hard to understand as shown in the following code:

if(flag)
{
}
The above code prints true and false because there is no else part in this
code. The second println statement is in an independent block and is not a
part of the if statement.
Dangling else ☝
"Dangling else" is a well known problem in programming languages that have if

as well as if-else statements. This is illustrated in the following piece of code
that has two if parts but only one else part:
int value = 3;
if(value == 0)
if(value == 1) System.out.println("b");
else System.out.println("c");
The question here is with which if should the else be associated? There are two
equally reasonable answers to this question as shown below:
int value = 3;
if(value == 0) {
if(value == 1) System.out.println("b");
}
else System.out.println("c");
and
int value = 3;
if(value == 0) {
if(value == 1)
System.out.println("b");
else
System.out.println("c");
}
In the first interpretation, else is associated with the first if , while in the second
interpretation, else is associated with the second if . If we go by the first
interpretation, the code will print b , and if we go by the second interpretation,
the code will not print anything. For a compiler, both are legally valid ways to
interpret the code, which makes the code ambiguous.
Since neither of the interpretations is more correct than the other, Java language
designers broke the tie by deciding to go with the second interpretation, i.e., the
else is to be associated with the nearest if . That is why the above code does not
print anything as there is no else part associated with the first if . Based on the
above discussion, you should now be able to tell the output of the following
code:
int value = 3;
if(value == 0)
if(value == 1)
else
else
System.out.println("d");
Just follow the rule that an else has to be associated with the nearest if. The
following is how the above statement will be grouped:
int value = 3;
if(value == 0){
if(value == 1)
else
}
else System.out.println("d");
It will print d .
Using assignment statement in the if condition ☝
Recall that every assignment statement itself is a valid expression with a value of
its own. Its type and value are the same as the ones of the target variable. This
fact can be used to write a very tricky if statement as shown below:

if(flag = true
){
}
else {
}
Observe that flag is not being compared with true here. It is being assigned
the value true . Thus, the value of the expression flag = true is true and
that is why the if part of the statement is executed instead of the else part. While
this type of code is not appreciated in a professional environment, a similar
construct is quite common:
String data = null;

if( (data = readData()) != null ) //assuming that read
Data() returns a String
{
//do something
The above code is fine because the assignment operation is clearly separated
from the comparison operation. The value of the expression data =
readData() is being compared with null . Remember that the value of this
expression is the same as the value that is being assigned to data . Thus, the if
body will be entered only if data is assigned a non-null value.
Using pre and post increment operations in the if condition ☝
You will see conditions that use pre and post increment (and decrement)
operators in the exam. Something like this:
int x = 0;
if(x++ > 0){ //line 2
x--; //line 3
}

if (++x == 2){ //line 6
x++; //line 7
}
System.out.println(x);
You can spot the trick easily if you have understood the difference between the
value of an expression and the value of a variable used in that expression (I
explained this in the "Using Operators" chapter). At line 2, x will be
incremented to 1 but value of the expression x++ is 0 and therefore, the
condition will be evaluated to false . Thus, line 3 will not be executed. At line
6, x is incremented to 2 and the value of the expression ++x is also 2 .
Therefore, the condition will be evaluated to true and line 7 will be executed,
thereby increasing the value of x to 3 . Thus, the above code will print 3 .
Remember that conditions are used in ternary expressions and loops as well, so,
you need to watch out for this trick there also.
7.2 Create ternary constructs

7.2.1 The ternary conditional operator ?: ☝
The ternary operator is not mentioned in the exam

objectives explicitly (although it was mentioned explicitly
in OCA 8 objectives). We haven't seen any one getting
a question on it either. However, it is one of the
commonly used operators in Java and from that
perspective, it falls within the scope of the exam
objectives. You should go through the following
discussion only if you have time.
The syntax of the ternary operator is as follows:
operand 1 ? operand 2 : operand 3;
Operand 1 must be an expression that returns a boolean . The boolean value

of this expression is used to decide which one of the rest of the two other
operands should be evaluated. In other words, which of the operands 2 and 3 will
be evaluated is conditioned upon the return value of operand 1. If the boolean
expression given in operand 1 returns true, the ternary operator evaluates and
returns the value of operand 2 and if it is false, it evaluates and returns the value
of operand 3. From this perspective, it is also a "conditional operator" (as
opposed to &, |, !, and ^ , which are really just "logical" operators).
Examples:
boolean sweet = false;

int calories = sweet ? 200 : 100; //assigns 100 to cal
ories
boolean sweetflag = (calories == 100 ? false : true);/

/assigns false to sweetflag
boolean hardcoded = false;

//assuming getRateFromDB() method returns a double.
double rate = hardcoded ? 10.0 : getRateFromDB(); //in

vokes method getRateFromDB
String value = sweetflag ? "Sweetened" : "Unsweetened"

;

Object obj = sweetflag ? "Sweetened" : new Object();
The ternary conditional operator is often thought of a short form for the if/else
statement but it is similar to the if/else statement only up to conditional
evaluation of its other two operands part. Their fundamental difference lies in the
fact that the ternary expression is just an expression while an if/else statement is
a statement . As discussed earlier, every expression must have a value and so,
must the ternary expression. Since the value of a ternary expression is the value
returned by the second or the third operand, the second and third operands of the
ternary operator can comprise any expression. As the example given above
shows, they can also include non-void method invocations . There is no such
restriction with the if/else statement. The following example highlights this
difference:

if(flag) System.out.println("true");
else System.out.println("false");
The above if/else statement compiles fine and prints true but a similar code
with ternary expression does not compile.
flag ? System.out.println("true") : System.out.println

("false");
The reason for non-compilation of the above code is two fold. The first is that a
ternary expression is not a valid statement on its own, which means, you cannot
just have a free standing ternary expression. It can only be a part of a valid
statement such as an assignment. For example,
int x = flag ? System.out.println("true")

: System.out.println("false");
This brings us to the second reason. The second and third operands in this
example are invocations methods that return void . Obviously, you cannot
assign void to an int variable. In fact, you cannot assign void to any kind of
variable. Therefore, it fails to compile.
Type of a ternary conditional expression ☝

Now that we have established that a ternary conditional expression must return a
value, all that is left to discuss is the type of the value that it can return. It can
return values of three types: boolean, numeric, and reference . (There are no
other types left for that matter!)
If the second and third operands are expressions of type boolean (or Boolean),
then the return type of the ternary expression is boolean. For example:
int a = 1, b = 2;
boolean flag = a == b? true : false; //ternary express
ion that returns a boolean
If the second and third operands are expressions of a numeric types (or their
wrapper classes), then the return type of the ternary expression is the wider of
the two numeric types. For example, double d = a == b? 5 : 10.0; .
Observe here that the second operand is of type int while the third is of type
double . Since double is wider than int , the type of this ternary expression
is double . You cannot, therefore, do int d = a == b? 5 : 10.0;
because you cannot assign a double value to an int variable without a cast.
If the second and third operands are neither of the above, then the return type of
the ternary expression is reference. For example,
Object str = a == b? "good bye" : "hello"; Here operands 2
and 3 are neither numeric nor boolean. Therefore, the return type of this ternary
expression is a reference type.
The first two types are straight forward but the third type begs a little more
discussion. Consider the following line of code:
Object obj = a == b? 5 : "hello"; .
Here, the second operand is of a numeric type while the third is of type String
. Since this falls in the third category, the return type of the expression a ==
b? 5 : "hello"; must be a reference. The question before the compiler
now is what should be the type of the reference that is returned by the
expression. One operand is an int , which can be boxed into an Integer
object and another one is a String object. If the boolean condition evaluates
to true , the expression will return an Integer object and if the condition
evaluates to false , the expression will return a String object. Remember
that the compiler cannot execute any code and so, it cannot determine what the
expression will return at run time. Thus, it needs to pick a type that is compatible
with both kind of values. The compiler solves this problem by deciding to pick
the most specific common superclass of the two types as the type of the
expression. In this case, that class is java.lang.Object . By selecting the
most specific common super class, the compiler ensures that irrespective of the
result of the condition, the value returned by this expression will always be of
the same type, i.e., java.lang.Object , in this case. Here is an example,
where the most specific common super class is not Object :
Double d = 10.0;
Byte by = 1;
Number n = a == b? d : by;
Here, the most specific common superclass of Double and Byte is Number .
(Recall that all wrapper classes for integral types extend from
java.lang.Number , which in turn extends from java.lang.Object ).
You can therefore, assign the value of the expression to a variable of type
Number .
You should now be able to tell the result of the following two lines of code:
int value = a == b? 10 : "hello"; //1
System.out.println(a == b? 10 : "hello"); //2
As discussed above, the type of the expression a == b? 10 : "hello"; is

Object . Can you assign an Object to an int variable? No, right? Therefore,
the first line will not compile. Can you pass an Object to the println
method? Of course, you can. Therefore, the second line will compile and run
fine.
The short circuiting property of ?: ☝
Depending of whether the value of operand one is true or false, either operand 2
or operand 3 is evaluated. In other words, if the condition is true , operand 3 is
not evaluated and if the condition is false , operand 2 is not evaluated. In this
respect, the ternary conditional operator is similar to the other two short
circuiting operators, i.e., && and || . However, there is an important difference
between the two. While with && and || , evaluation of both the operands is
possible in certain situations, it is never the case with ?: operator. ?: evaluates
exactly one of the two operands in all situations.
The short circuiting nature of ?: provides a good opportunity for trick questions
in the exam. For example, what will the following code print?
int x = 0;
int y = 1;
System.out.println(x>y? ++x : y++);
System.out.println(x+" "+y);
This code prints:
1
0 2
Since the value of x is not greater than y , x>y evaluates to false and
therefore, the ternary expression returns the value of the third operand, which is
y++ . Since y++ uses post-increment operator, the return value of y++ will be
the current value of y , which is 1 . y will then be incremented to 2 . Observe
that evaluation of the second operand is short circuited because the first operand
evaluates to false. Therefore, ++x is never executed. Thus, the second print
statement prints 0 2 .
7.3 Use a switch statement
7.3.1 Creating a switch statement ☝
A switch statement allows you to use the value of a variable to select which code
block (or blocks) out of multiple code blocks should be executed. Here is an
example:

int i = args.length;
switch(i) { //switch block starts here
case 0 : System.out.println("No argument");

break;
case 1 : System.out.println("Only one argume
nt");
break;
case 2 : System.out.println("Two arguments")
;
break;
default : System.out.println("Too many argum
ents!");
break;
}//switch block ends here

System.out.println("All done.");
}
}
There are four blocks of code in the above switch statement. Each block of code
is associated with a case label . Depending of the value of the variable i , the
control will enter the code block associated with that particular case label and
keep on executing statements until it finds a break statement. For example, if
the value of i is 0 , the control will enter the first code block. It will print No
argument and then encounter the break statement. The break statement
causes the control to exit the switch statement and move on to the next statement
after the switch block, which prints "All done" .
If the value of i doesn't match with any of the case labels, the control looks for a
block labelled default and enters that block. If there is no default block either,
the control does not enter the switch block at all. Since this "switching" is done
based on the expression specified in the switch statement (which is just i in this
example), this expression is aptly called the "switch expression ".
Operationally, this seems quite similar to a cascaded if/else statement. Indeed,

the above code can very well be written using an if/else statement as follows:

int i = args.length;
if(i == 0) {
System.out.println("No argument");
} else if (i == 1) {
System.out.println("Only one argument");
} else if (i == 2) {
System.out.println("Two arguments");
} else {
System.out.println("Too many arguments!");
}
}
}
Well, why do you need a switch statement then, you may ask. To begin with, as
you can see, an if/else statement is a lot more verbose than a switch statement.
The switch version is also a little easier to comprehend than the if/else version.
But underneath this syntactic ease lies a complicated beast. This is evident when
we look at the moving parts involved in a switch statement more closely.
The switch expression ☝
A switch expression must evaluate to one of the following three kinds:
1. a limited set of integral types (byte, char, short, int ), and their
wrapper classes. Observe that even though long is an integral type, it
cannot be the type of a switch variable. boolean, float, and
double are not integral types anyway and therefore, cannot be the type of
a switch variable either.
2. enum type
3. java.lang.String - Generally, reference types are not allowed as
switch expressions but an exception for java.lang.String was made in Java
7. So now, you can use a String expression as a switch expression.
Compare this to an if/else statement where branching is done based on the value
of a boolean expression . This limits an if/else statement to at most two
branches. Of course, as we saw earlier, you can cascade multiple if/else
statements to achieve multiple branches.
The case labels ☝
Case labels must consist of compile time constants that are assignable to the
type of the switch expression. For example, if your switch expression is of type
byte , you cannot have a case label with a value that is larger than a byte.
Similarly, if your switch expression is of type String , the case labels must be
constant string values as illustrated in the following code:

switch(args[0]){
case "1" : System.out.println("one"); //valid becau
se "1" is a compile time constant
case "1"+"2" : System.out.println("one"); //valid b

ecause "1"+"2" is a compile time constant
case args[1] : System.out.println("same args");//wi

ll not compile because args[1] is not a compile time c
onstant
case "abc".toUpperCase() : System.out.println("ABC"

);//will not compile because "abc".toUpperCase() is no
t a compile time constant
}
}
Observe that "1" and "1"+"2" are compile time constants because the value
of these expressions is known at compile time, while args[1] and
"abc".toUpperCase() are not compile time constants because their values
can only be determined at run time when the code is executed.
The interesting thing about case labels is that they are optional . In other words,
a switch statement doesn't necessarily have to have a case label. The following
is, therefore, a superfluous yet valid switch statement.
switch(i){
default : System.out.println("This will always be p
rinted");
}
Another point worth repeating here is that although it is very common to use a
single variable as the switch expression but you can use any expression inside
the switch. And when you talk of an expression, all the baggage of numeric
promotion, casting, and operator precedence that we saw previously, comes
along with it. You need to consider all that while checking the validity of case
labels. For example, while the following code fails to compile:
byte b = 10;
switch(b){ //type of the switch expression here is byte
case 1000 : //1000 is too large to fit into a byte
System.out.println("hello!");
}
the following code compiles fine:
byte b = 10;
switch(b+1){ //type of the switch expression here is n
ow int due to numeric promotion
case 1000 : //1000 can fit into an int
System.out.println("hello!");
}
The default block ☝
There can be at most one default block in a switch statement. The purpose of the
default block is to specify a block of code that needs to be executed if the value
of the switch expression does not match with any of the case labels. Just like the
case labels, this block is also optional .
The order of case and default blocks ☝
Java does not impose any particular order for the case statements and the default
block. Thus, although it is customary to have the default block at the end of a
switch block, you can have it even at the beginning. Similarly, Java does not care
about the ordering of the case labels.
However, "does not care" does not mean "not important"! Ordering of case and
default blocks becomes very important in combination with the use of the
break statement as we will see next.
The break statement ☝
I mentioned in the beginning that the case labels determine the entry point into a
switch statement and the break statement determines the exit. That is true but the
interesting thing is that even the break statement is optional . A case block does
not necessarily have to end with a break. Let me modify the program that I
showed you in the beginning:

switch(args.length) { //switch block starts here
case 0 : System.out.println("No argument");

//break;
case 1 : System.out.println("Only one argume

nt");
//break;
case 2 : System.out.println("Two arguments")

;
//break;
default : System.out.println("Too many argum

ents!");
//break;
}//switch block ends here

}
}
I have commented out all the break statements. Now, if you run this
program without any argument , you will see the following output:
No argument
Only one argument
Two arguments
Too many arguments!
All done.
The control entered at the block labelled case 0 (because args.length is 0 ),

and executed all the statements of the switch block...even the statements
associated with other case blocks that did not match the value of
args.length . This is called "fall through" behavior of a switch statement.
In absence of a break statement, the control falls through to the next case block
and the next case block, and so on until it reaches the end of the switch
statement. This feature is used when you want to have one code block to execute
for multiple values of the switch expression. Here is an example:
char ch = 0;
int noOfVowels = 0;
while( (ch = readCharFromStream()) > 0) {
switch(ch) {
case 'a' :
case 'e' :
case 'i' :
case 'o' :
case 'u' :
noOfVowels++;
default : logCharToWhatever(ch);
}
}
System.out.println("Number of vowels received "+noOfVo
wels);
The above code logs each character it receives but increments noOfVowels
only if the character received is a lower case vowel.
You will see questions in the exam on this behavior of the switch statement. So
pay close attention to where in the switch block does the control enter and where
it exits. Always check for missing break statements that cause the control to fall
through to the next case block.
The following is a typical code snippet you may get in an exam. Try running it
with different arguments and observe the output in each case:

int i = 0;
switch(args[0]) {
default : i = i + 3;
case "2" : i = i + 2;
case "0" : break;
case "1" : i = i + 1;
}

System.out.println("i is "+i);
}
}
The fall through behavior of a switch statement does

not really get used a lot in Java but it has been used
in interesting ways to optimize code in other languages.
If you have time, you might want to google "duff's
device" to see one such usage. This is, of course, not
required for the exam :)
7.4 Exercise ☝
1. Write a method that accepts a number as input and prints whether the
number is odd or even using an if/else statement as well as a ternary
expression.
2. Accept a number between 0 to 5 as input and print the sum of numbers
from 1 to the input number using code that exploits the "fall through"
behavior of a switch statement.
3. Accept a number as input and generate output as follows using a cascaded
and/or nested if/else statement - if the number is even print "even", if it is
divisible by 3, print "three", if it is divisible by 5, print "five" and if it is not
divisible by 2, 3, or 5, print "unknown". If the number is divisible by 2 as
well as by 3, print "23", and if the number is divisible by 2, 3, and 5, print
"235".
4. Indent the following code manually such that it reflects correct if - else
associations. Use a plain text editor such as Notepad. Copy the code into a
Java editor such Netbeans or Eclipse and format it using the editor's auto
code formatting function. Compare your formatting with the editor's.
int a = 0, b = 0, c = 0, d = 0;
if (a == b)
if (c == 10)
{
if (d > a)
{
} else {
}
if (flag)
System.out.println("");
else
}
else if (flag == false)
else if (a + b < d) {
}
else
else d = b;
Chapter 8 Using Loop Constructs
Create and use while loops
Create and use do/while loops
Create and use for and for each loops including nested loops
Use break and continue statements
8.1 What is a loop

8.1.1 What is a loop ☝
A loop causes a set of instructions to execute repeatedly until a certain condition

is reached. It's like when the kids keep asking, "are we there yet?", when you are
on a long drive in a car. They ask this question in a "loop", until you are really
there :) Or until you "break" their loop by putting on a DVD.
Well, in Java, loops work similarly. They let you execute a group of statements
multiple times or even forever depending on the loop condition or until you
break out of them. Every repetition of execution of the statements is called an
iteration . So, for example, if a group of statements is executed 10 times, we can
say that the loop ran for 10 iterations.
Loops are a fundamental building block of structured programming and are

used extensively for tasks ranging from the simple such as counting the sum of a
given set of numbers to the complicated such as showing a dialog box to the user
until they select a valid file.
Java has three different ways in which you can create a loop - using a while
statement, using a do/while statement, and using a for statement. Let us take a
look at each of them one by one.
8.2 Create and use while loops
8.2.1 The while loop ☝
As the name of this loop suggests, a while loop executes a group of statements
while a condition is true. In other words, it checks a condition and if that
condition is true, it executes the given group of statements. After execution of
the statements, i.e., after finishing that iteration, it loops back to check the
condition again. If the condition is false, the group of statements is not executed
and the control goes to the next statement after the while block. The syntax of a
while loop is as follows:
while (boolean_expression) {
statement(s);
}
and here is an example of its usage:

int i = 4;
while(i>0){
i--;
}
System.out.println("Value of i after the loop is
"+i);
}
}
It produces the following output:
i is 3
i is 2
i is 1
i is 0
Value of i after the loop is 0
Observe that the condition is checked before the group of statements is executed
and that once the condition i>0 returns false , the control goes to the next
statement after the while block.
As you have seen in the past with if/else blocks, if you have only a single
statement that you need to execute in a loop, you may get rid of the curly
brackets if you want. In this case, the syntax becomes:
while(boolean_expression) statement;
The example program given above can also be written to use a single statement
like this:

int i = 4;
while(i-->0) System.out.println("i is "+i); //no
curly braces
System.out.println("Value of i after the loop is

"+i);
}
}
The above code looks the same as the previous one but produces a slightly
different output:
i is 3
i is 2
i is 1
i is 0
Value of i after the loop is -1
The difference is that I have used the post-decrement operator to decrement the
value of i within the condition expression itself. Remember that when you use
the post -decrement (or the post -increment) operator, the variable is
decremented (or incremented) after its existing value is used to evaluate the
expression. In this case, when i is 0 , the condition expression evaluates to false
and the while block is not executed, but the variable i is nevertheless
decremented to -1 . Therefore, the value of i after the loop is -1 .
8.2.2 Using a while loop ☝
Control condition and control variable ☝
When using a while loop you should be careful about the expression that you use
as the while condition. Most of the time, a while condition comprises a single
variable (which is also called the "control variable" , because it controls
whether the loop will be entered or not) that you compare against a value. For
example, i>4 or name != null or bytesRead != -1 and so on.
However, it can get arbitrarily large and complex. For example, account ==
null || (account != null && account.accountId == null)
|| (account != null && account.balance == 0) . The
expression must, however, return a boolean . Unlike some languages such as
C, Java does not allow you to use an integer value as the while condition. Thus,
while(1) System.out.println("loop forever"); will not
compile.
while body ☝
Ideally, the code in the while body should modify the control variable in such a
way that the control condition will evaluate to false when it is time to end the
loop. For example, if you are processing an array of integers, your control
variable could be set to the index of the element that you are processing, and
each iteration should increment that variable. For example,
int[] myArrayOfInts = //code to fetch the data
int i = 0; //control variable
while(i<myArrayOfInts.length){
//do something with myArrayOfInts[i]
i++;//increment i so that the control condition will e

valuate to false after the last element is processed.
There are situations where you do not want a loop to end at all. For example, a
program that listens on a socket for connections from clients may use a while
loop as follows:
Socket clientSocket = null;

while( (clientSocket = serverSocket.accept() ) != null
){
//code to hand over the clientSocket to another thre
ad and go back to the while condition to keep listenin
g for connection requests from clients.
The above is a commonly used while loop idiom.
A never ending while loop can also be as simple as this:
while(true){
System.out.println("keep printing this line forever!
");
}
On the other extreme, keep an eye for a condition that never lets the control enter
the while body:
int i = 0;
while(i>0){ // the condition is false to begin with
System.out.println("hello"); //this will never be pr
inted
i++;
}
It is possible to exit out of a while loop without checking the while condition.
This is done using the break statement. I will discuss it later.
8.3 Create and use do/while loops

8.3.1 The do-while loop ☝
A do-while loop is similar to the while loop. The only difference between the
two is that in a do-while loop the loop condition is checked after executing the
loop body. Its syntax is as follows:
do {
statement(s);
}while(boolean_expression);
Observe that a do-while statement starts with a "do" but there is no "do"
anywhere in a while statement. Another important point to understand here is
that since the loop condition is checked after the loop body is executed, the loop
body will always be executed at least once.
As with a while statement, it is ok to remove the curly brackets for the loop body
if there is only one statement in the body. Thus, the following two code snippets
are actually the same:
int i = 4;
do {
} while(i-->0);
System.out.println("Loop finished. i is "+i);
int i = 4;
do
while(i-->0);
System.out.println("Loop finished. i is "+i);
Deciding whether to use a while loop or a do while loop is easy. If you know that
the loop condition may evaluate to false at the beginning itself, i.e., if the loop
body may not execute even once, you must use a while loop because it lets you
check the condition first and then execute the body. For example, consider the
following code:
Iterator<Account> acctIterator = accounts.iterator();

while(acctIterator.hasNext()){ //no need to enter the
loop body if accounts collection is empty.
Account acct = acctIterator.next();

//do something with acct
Don't worry about the usage of Iterator or <Account> in the above code.
The point to understand here is that you want to process each account object in
the accounts collection and if there is no account object, you don't want to enter
the loop body at all. If you use a do-while loop, the code will look like this:

do {
Account acct = acctIterator.next(); //will throw a
n exception
}while(acctIterator.hasNext());
The above code will work fine in most cases but will throw an exception if the
account collection is empty. To achieve the same result with a do-while loop,
you would have to write an additional check for an empty collection at the
beginning. Something like this:

if(acctIterator.hasNext()) {//no need to enter the if
body if accounts collection is empty.
do{
Account acct = acctIterator.next();
}while(acctIterator.hasNext())
}
I think the choice is quite clear. A while loop is a natural fit in this case.
Generally, a while loop is considered more readable than a do-while loop and is
also used a lot more in practice. I have not needed to use a do-while loop in a
long while myself :)
8.4 Create and use for loops

8.4.1 Going from a while loop to a for loop ☝
The for loop is the big daddy of loops. It is the most flexible, the most
complicated, and the most used of all loop statements. It has so many different
flavors that many programmers do not get to use some of its variations despite
years of programming in Java. But don't be scared. It still follows the basic idea
of a loop, which is to let you execute a group of statements multiple times.
To ease you into it, I will morph the code for a while loop into a for loop. Here is
the code that uses a while loop:

int i = 4;
while(i>0){
i--;
}
System.out.println("Value of i after the loop is "
+i);
}
}
The output of the above code is:
i is 4
i is 3
i is 2
i is 1
Value of i after the loop is 0
Here is the same code with a for loop:

for(int i = 4; i>0; i--){
}
//System.out.println("Value of i after the loop is
"+i);
}
}
Its output is:
i is 4
i is 3
i is 2
i is 1
As you can see, it produces the same output except the last line. The following
image shows how the elements of a while loop of the first code were mapped to
the for loop of the second code.
Here are a few things that you should observe in the transformation shown
above.
1. The while statement contains just the comparison but the for statement
contains initialization, comparison, and updation. Technically, the
declaration of the loop variable i and its updation through i-- is not really
a part of the while statement. But the for loop has a provision for both.
2. The implication of declaring the loop variable i in a for statement is that it
is scoped only to the for block and is not visible outside the for block,
which is why I have commented out the last print statement in the second
code.
3. There is no semi-colon after i-- in the for statement.
This example illustrates that fundamentally there is not much difference between
the two loop constructs. Both the loops have the same three basic components -
declaration and/or initialization of a loop variable, a boolean expression that
determines whether to continue next iteration of the loop, and a statement that
updates the loop variable. But you can see that the for loop has a compact syntax
and an in-built mechanism to control the initialization and updation of the loop
variable besides the comparison as compared to the while loop.
8.4.2 Syntax of a for loop ☝
A for loop has the following syntax:
for( optional initialization section ;
optional condition section ;
optional updation section ) {

statement(s);
}
Predictably, if you have only zero or one statement in the for block, you can get
rid of the curly braces and end the statement with a semicolon like this:
optional updation section ) ;

//no body at all
or
optional updation section )

single_statement;
All three sections of a for statement are optional but the two semi-colons that
separate the three sections are not. You are allowed to omit one, two, or all of the
three sections. Thus, for( ; ; ); is actually a valid for statement but
for(); and for(;); are not.
Order of execution ☝
Various pieces of a for loop are executed in a specific order, which is as follows:
1. The initialization section is the first that is executed. It is executed exactly

once irrespective of how many times the for loop iterates. If this section is
empty , it is ignored .
2. Next, the condition section is executed. If the result of the expression in
this section is true, the next iteration, i.e., execution of the statements in the
for block is kicked off. If the result is false, the loop is terminated
immediately, i.e., neither the for block and nor the updation section is
executed. If this section is empty , it is assumed to be true .
3. The statements in the for block are executed.
4. The updation section is executed. If this section is empty , it is ignored .
5. The control goes back to the condition section .
Let us see how the above steps are performed in the context of the following for
loop.
for(int i = 3; i>0; i--){

}
1. int i = 3; is executed. So, i is now 3 .

2. Expression i>0 is evaluated and since 3 indeed greater than 0 , it evaluates
to true . Therefore, the next iteration will now commence.
3. The print statement is executed, which prints 3 .
4. i-- is executed thereby reducing i to 2 .
5. Control goes back to the condition section. i>0 evaluates to true because
2 is greater than 0 . Therefore, the next iteration will now commence.
8. Control goes back to the condition section. i>0 evaluates to true because
1 is greater than 0 . Therefore, the next iteration will now commence.
11. Control goes back to the condition section. i>0 evaluates to false
because 0 is not greater than 0 . Therefore, the loop will be terminated. The
statements in the for block and the updation section will not be executed
and the control goes to the next statement after the for block (which is the
end of the method in this case).
So far, the for statement looks quite straight forward. You will see the
complications when we turn our attention to the intricacies of the three sections
of the for statement next.
8.4.3 Parts of a for loop ☝
The initialization section ☝
The initialization section allows you to specify code that will be executed at the
beginning of the for loop. This code must belong to a category of statements that
are called "expression statements" . Expression statements are expressions that
can be used as statements on their own. These are: assignment , pre/post
increment/decrement expression , a method call , and a class instance
creation expression (e.g. using the new operator). Besides expression
statements, this section also allows you to declare local variables.
Here are a few examples of valid expression statements in a for loop:
int i = 0;
for(i = 5;
i<10; i++); //assignment
Object obj;
for(obj = "hello";
i<10; i++); //assignment
int i = 0;
int k = 0;
Object obj = "";
for(i = 0, k = 7, obj = "hello";
i<10; i++); //multiple assignments
int k = 0;
for(++k;
i<10; i++); //pre-increment
for(new ArrayList(); i<10; i++); //instance creation
int i = 0;
for(System.out.println("starting the loop now");
i<10; i++); //method call
for(k++, i--, new String()
;i<10; i++); //multiple expressions
Observe that there doesn't need to be any relationship between the variable used
in the initialization section and the variable used in other sections of a for loop.
The following are a few examples of valid local variable declarations:
for(int i = 5;
i<10; i++); //single variable declaration
for(int i = 5, k = 7;
i<10; i++); //multiple variable declaration
You can only declare variables of one type. So the following is invalid :
for(int i = 5, String str = "";
i<10; i++); //invalid
Redeclaring a variable is also invalid:

int i = 0;
for(int i = 5;
i<10; i++); //invalid because i is already declared
Another important point to note here is that a variable declared in the

initialization section is visible only in the for loop. Thus, the following will not
compile because i is not visible outside the loop.
for(int i = 5;
i<10; i++){
}
System.out.println("Final value of i is "+i); //this l

ine will not compile
The condition section ☝
This one is simple. No, really :) You can only have an expression that returns a
boolean or Boolean in this section. If there is no expression in this section,
the condition is assumed to be true .
The updation section ☝
The rules for the updation section are the same as the initialization section
except that you cannot have any declarations here. This section allows only
"expression statements", which I have already discussed above. Generally, this
section is used to update the loop variable (i++ or k = k + 2 and so on) but
as you saw in examples above, this is not the only way to use it. There doesn't
need to be any relationship between the code specified here and in other
sections. The following are a few valid examples:
int i = 0;
for(;i<10; i++
); //post-increment
for(;i<10; i = i + 2
); //increment by two
for(;i<10; i = someRef.getValue()
) //assignment
for(Object obj = new Object(); obj != null; ) { //empt

y updation section
System.out.println(obj);
obj = null;
}
for(int i = 0; i<10; callSomeMethod()
); //method call
Observe that there is no semi-colon after the expression statement. It is

terminated by the closing parenthesis of the for statement.
An infinite for loop ☝
Based on the above information, it should now be very easy to analyze the
following loop:
for( ; ; ) ;
The initialization section is empty. The condition section is empty, which means
it will assumed to be true . The updation section is empty. The loop code is
also empty. There is nothing that makes the condition section to evaluate to
false and therefore, there is nothing to stop this loop from iterating forever.
Now, as an exercise, try analyzing the following loop and determine its output:
boolean b = false;
for(int i=0 ; b = !b ; ) {
System.out.println(i++);
}
8.5 Create and use for each loops

8.5.1 The enhanced for loop ☝
Motive behind the enhanced for loop ☝
In professional Java programming, looping through a collection of objects is a

very common requirement. Here is how one might do it:
String[] values = { "a", "bb", "ccc" };

for(int i = 0; i<values.length; i++){
System.out.println(values[i]); //do something with e
ach value
The above code iterates through an array but the same pattern can be used to
iterate through any other collection such as a List. You can code this loop using a
while or do-while construct as well.
Java designers thought that this is too much code to write for performing such a
simple task. The creation of an iteration variable i and the code in the condition
section to check for collection boundary is necessary only because of the way
the for (or other) loops work. They have no purpose from a business logic point
of view. All you want is a way to do something with each object of a collection.
You don't really care about the index at which that element is inside that
collection. Furthermore, some collections such as a Set have no notion of index.
Here is the code that iterates through the elements of a Set just to give you an
idea of how you may iterate though a collection of objects that does not support
index based retrieval:
Set s = new HashSet();

s.add("a");
s.add("bb");
s.add("ccc");
Iterator it = s.iterator();
while(it.hasNext() ){
Object value = it.next();
System.out.println(value); //do something with each
value
The above code has to create an Iterator object, which has no relation to the
business logic, to iterate through the elements. Such code that is not required by
the business logic but is required only because of the way a programming
language works is called "boilerplate" code. Such code can be easily
eliminated if a programming language provides higher level constructs that
internalize the logic of this code.
In Java 5, Java designers simplified the iteration process by doing exactly that.
They introduced a new construct that internalizes the creation of iteration
variable and the boundary check and called it the "enhanced for loop" or the
"for-each loop" .for each
The enhanced for loop ☝
Let me first show how the two code snippets given above can be simplified and
then I will get into the details:
String[] values = { "a", "bb", "ccc" };

for(String value : values){
value
and
Set s = new HashSet();

s.add("a");
s.add("bb");
s.add("ccc");
for(Object value : s){

value
}
Isn't that simple? It reads easy too - "for each value in values do ..." and "for
each value in s do ...".
8.5.2 Syntax of the enhanced for loop ☝
The syntax of the enhanced for loop is as follows:
for( Type variableName : array_or_Iterable ){

statement(s);
}
or if there is only one statement in the for block:
for( Type variableName : array_or_Iterable ) statement

;
Type is the type of the elements that the array or the collection contains,
variableName is the local variable that you can use inside the block to work
with an element of the array or the collection, and array_or_iterable is the array
or an object that implements java.util.Iterable interface.
I know that I have been using the word "collection" up till now and suddenly I
have switched to Iterable . The reason for the switch is that
java.util.Iterable is a super interface of java.util.Collection
and although for-each loop is used mostly to iterate over collections, technically,
it can iterate through an object of any class that implements
java.util.Iterable . In other words, any class that wants to allow a user
to iterate through the elements that it contains must implement
java.util.Iterable interface.
The Iterable interface was introduced in Java 1.5 specifically to denote that
an object can be used as a target of the for-each loop. It has only one method
named iterate , which returns a java.util.Iterator object. Since
java.util.Collection extends java.util.Iterable , all standard
collection classes such as HashSet , and ArrayList can be iterated over
using the enhanced for loop.
For the purpose of the exam, you don't need to worry about the Iterable or the
Iterator interface but it is a good idea to develop a mental picture of the
relationship between the Iterable and the Iterator interfaces. Always remember
that for-each can only be used to "iterate" over an object that is "Iterable" .
Note that java.util.Iterator itself is not a collection of elements and

does not implement java.util.Iterable . Therefore, an Iterator object
cannot be a target of a for-each loop. Thus, the following code will not compile:
Iterator it = myList.iterator();
for(Object s : it){ //this line will not compile
For the purpose of the OCP Java 11 Part 1 exam, you only need to know that you
can use the for-each loop to iterate over any collection such as a List and an
ArrayList .
8.5.3 Enhanced for loop in practice ☝
Using enhanced for loop with typified collections ☝
In the code that I showed earlier, I used Object as the type of the variable inside
the loop. Since I was only printing the object out I didn't need to cast it to any
other type. But if you want to invoke any type specific method on the variable,
you would have to cast it explicitly like this:
List names = //get names from somewhere
for(Object obj : names){

String name = (String) obj;
System.out.println(name.toUpperCase());
}
The true power of the enhanced for loop is realized when you use generics
(which were also introduced in Java 5, by the way) to generify the collection that
you are trying to iterate through. Here is the same code with generics:
List<String>
names = //get names from somewhere
for(String name : names){

System.out.println(name.toUpperCase());
}
Although the topic of generics is not on the OCP Java 11 Part 1 exam, you will
see questions on foreach that use generics. You don't need to know much about
generics to answer the questions, but you should be aware of the basic syntax. I
will discuss it more while talking about List and ArrayList later.
Counting the number of iterations ☝
In a regular for loop, the iteration variable (usually named i) tells you which
iteration is currently going on. There is no such variable in a foreach loop. If you
do want to know about the iteration number, you will need to create and manage
another variable for it. For example:
List<String> names = //get names from somewhere
int i = 0;
for(String name : names){
i++;
System.out.println(i+" : "+name.toUpperCase());
}
System.out.println("Total number of names is "+i);
8.6 Use break and continue

8.6.1 Terminating a loop using break ☝
A loop doesn't always have to run its full course. For example, if you are
iterating through a list of names to find a particular name, there is no need to go
through the rest of the iterations after you have found that name. The break
statement does exactly that. Here is the code:
String[] names = { "ally", "bob", "charlie", "david" }

;
for(int i=0; i<names.length; i++){
System.out.println(names[i]);
if("bob".equals(names[i])){
System.out.println("Found bob at "+i);
break;
}
}
The output of the above code is:
ally
bob
Found bob at 1
Observe that charlie and david were not printed because the loop was
broken after reaching bob .
It doesn't matter which kind of loop (i.e. while, do-while, for, or enhanced for) it
is. If the code in the loop encounters a break, the loop will be broken
immediately. The condition section and the updation section (in case of a for
loop) will not be executed anymore and the control will move to the next
statement after the loop block.
8.6.2 Terminating an iteration of a loop using continue ☝
The continue statement is a little less draconian than the break statement. The
purpose of a continue statement is to just skip the rest of the statements in the
loop while executing that particular iteration. In other words, when a loop
encounters the continue statement, the remaining statements within the loop
block are skipped. The rest of the steps of executing a loop are performed as
usual, i.e., the control moves on to updation section (if it is a for loop) and then it
checks the loop condition to decide whether to execute the next iteration or not.
This is useful when you don't want to execute the rest of the statements of a loop
after encountering a particular condition. Here is an example:

;

for(String name : names){ //using a for-each loop this
time

if("bob".equals(name)){
System.out.println("Ignoring bob!");
continue;
}
System.out.println("Hi "+name+"!");
}
This code produces the following output:
Hi ally!
Ignoring bob!
Hi charlie!
Hi david!
Observe that the print statement saying Hi was not executed only when the
name was bob .
Just like the break statement, the continue statement is applicable to all
kind of loops.
Both of these statements are always used in conjunction with a conditional

statement such as an if/if-else. If a continue or a break executes unconditionally
then there would be no point in writing the code below a continue or a break. For
example, the following code will fail to compile:

;

for(int i=0; i<names.length; i++){
continue; //or break;
System.out.println("Hi "+names[i]+"!"); //This lin

e will never get to execute
The compiler will complain that the print statement is unreachable.
8.7 Nested loops

8.7.1 Nested loop ☝
A nested loop is a loop that exists inside the body of another loop. This is not a
big deal if you realize that the body of a loop is just another set of statements.
Among these set of statements, there may also be a loop statement. For example,
while looping through a list of Strings, you can also loop through the characters
of each individual String as shown in the following code:
The above code nests a regular for loop inside an enhanced for loop. The
following is the output produced by this code:
Lucky number for ally is 50

Lucky number for bob is 19
Lucky number for charlie is 56
Lucky number for david is 40
One thing that you need to be very careful about when using nested loops is
managing the loop variables correctly. Consider the following program that is
supposed to calculate the sum of all values in a given multidimensional array of
ints:
int[][] values = { {1, 2, 3} , { 2, 3}, {2 }, { 4, 5,

6, 7 } };
int sum = 0;
for(int i = 0; i<values.length; i++) {
for(int j=0; j<values[i].length; i++) {
sum = sum + values[i][j];
}
}
System.out.println("Sum is "+sum);
Read the above code carefully and try to determine the output.
It actually throws an exception: Exception in thread "main"

java.lang.ArrayIndexOutOfBoundsException: 4
Observe that the inner for loop is incrementing i instead of incrementing j .

Now, let's execute the code step by step:
Step 0 : Before the outer loop starts, values.length is 4 and sum is 0.

Step 1 : Outer loop is encountered - i is initialized to 0 , i<values.length
is true because 0 is less than 4 , so, the loop is entered.
Step 2 : Inner loop is encountered - j is initialized to 0 ,
j<values[i].length is true because i is 0, therefore, values[i] refers
to { 1, 2, 3} and values[i].length is 3 . Therefore, the loop is
entered.
Step 3 : Loop body is executed. sum is assigned 0 + values[0][0] , i.e., 1
. sum is now 1 .
Step 4 : Inner loop body is finished. Inner loop's updation section is executed,
so, i is incremented to 1 .
Step 5 : Inner loop's condition j<values[i].length is executed and it
evaluates to true because j is still 0 and i is 1 , so, values[i] refers to
{2, 3} and values[i].length is 2 . Thus, second iteration of the inner
loop will start.
Step 6 : Loop body is executed. sum is assigned 1 + values[1][0] , i.e.,
1+2 . sum is now 3 .
Step 8 : Inner loop's condition j<values[i].length is executed and it
{2} and values[i].length is 1 . Thus, third iteration of the inner loop
will start.
3+2 . sum is now 5 .
so i is incremented to 3 .
Step 11 : Inner loop's condition j<values[i].length is executed and
{4, 5, 6, 7 } and values[i].length is 4 . Thus, fourth iteration of
the inner loop will start.
5 + 4 . sum is now 9 .
Step 14 : Inner loop's condition j<values[i].length is executed. Now,
here we have a problem. i is 4 and because the length of the values array is
only 4 , values[4] exceeds the bounds of the array and therefore
values[i] throws an ArrayIndexOutOfBoundsException .
This is a very common mistake and while in this case the exception message
made it easy to find, such mistakes can actually be very difficult to debug as they
may produce plausible but incorrect results. That is why, although there is no
technical limit to how many levels deep you can nest the loops, most
professionals avoid nesting more than two levels.
In the exam you will be presented with questions containing two

levels of loops. Many students ask if there is an easy way to figure
out the output of such loops. Unfortunately, there is no short cut. It is
best if you execute the code step by step while keeping track of the
variables at each step as shown above on a piece of paper. The
code in the exam will be tricky and it may trip you up even if you are
an experienced programmer.
As an exercise, try executing the same code given above after

changing i++ to j++ .
8.7.2 breaking out of and continuing with nested loops ☝

You can use break and continue statements in nested loops exactly like I showed
you earlier with single loops. They will let you break off or continue with the
loop in which they are present. For example, check out the following code that
tries to find "bob" in multiple groups of names:

String[][] groups = { { "ally", "bob", "charlie" } , {
"bob", "alice", "boone"}, { "chad", "dave", "elliot"
} };
for(int i = 0; i<groups.length; i++){

for(String name : groups[i]){
System.out.println("Checking "+name);
if("bob".equals(name)) {
System.out.println("Found bob in Group "+i);
break;
}
}
}
The following is the output of the above code:
Checking ally
Checking bob
Found bob in Group 0
Checking bob
Checking chad
Checking dave
Checking elliot
When the inner loop encounters the String "bob" , it executes break, which
causes the inner loop to end immediately. The control goes on to execute the
next iteration of the outer loop, i.e., it starts looking for "bob" in the next group
of names. This is the reason why the above output does not contain "Checking
charlie", "Checking alice", and "Checking boone" .
Now what if you want to stop checking completely as soon as you find "bob"
in any of the groups? In other words, what if you want to break out of not just
the inner loop but also out of the outer loop as soon as you find "bob" ? Java
allows you to do that. You just have to specify which loop you want to break out
of. Here is how:
String[][] groups = { { "ally", "bob", "charlie" } ,

{ "bob", "alice", "bo
one"},
{ "chad", "dave", "el
liot" } };
MY_OUTER_LOOP:
for(int i = 0; i<groups.length; i++){

for(String name : groups[i]){
System.out.println("Checking "+name);
if("bob".equals(name)) {
System.out.println("Found bob in Group "+i);
break MY_OUTER_LOOP;
}
}
}
The following is the output of the above code:
Checking ally
Checking bob
I just "labeled" the outer for loop with MY_OUTER_LOOP and used the same
label as the target of the break statement. This is known as a "labeled break" .
A "labeled continue" works similarly.
Let me give a quick overview of labels and then I will get back to the rules of
using labeled break and continue.
What is a label? ☝
A label is nothing but a name that you generally give to a block of statements
or to statements that allow block of statements inside them (which means if, for,
while, do-while, enhanced for, and switch statements). The exam does not test
you on the precise rules for applying labels or naming of labels, but here are few
examples:
SINGLE_STMT: System.out.println("hello");
HELLO: if(a==b) callM1(); else callM2();
COME_HERE : while(i=>0) {
System.out.println("hello");
}
SOME_LABEL: { //ok, because this is a block of state

ments
System.out.println("hello1");
System.out.println("hello2");
}
Here are some examples of invalid usage of labels:
BAD1 : int x = 0; //can't apply a label to declarations
BAD2 : x++; //can't apply a label to expressions

BAD3 : public void m1() { } //can't apply a label to m
ethods
Although not necessary, the convention is to use capital letters for label names.
Also, applying a label to a statement doesn't necessarily mean that you have to
use that label as a target of any break or continue. You can label a statement just
for the sake of it :)
Rules for labeled break and continue ☝
Getting back to using labeled break and continue, you need to remember that if
you use a labeled break or a labeled continue, then that label must be present on
one of the nesting loop statements within which the labeled break or continue
exists. Thus, the following code will fail to compile:
LABEL_1 : for(String s : array) System.out.println(s)

; //usage of LABEL_1 is valid here.
for(int i = 0; i<10; i++){
if(i ==2) continue LABEL_1; //usage of continue is i

nvalid because LABEL_1 does not appear on a loop state
ment that contains this continue.
The following is invalid as well:
for(int i = 0; i<10; i++){

LABEL_1 : if(i ==2) System.out.println(2); //usage o
f LABEL_1 is valid here.
for(int j = 0; j<10; j++){
if(i ==2) continue LABEL_1; //usage of continue i

s invalid because LABEL_1 does not appear on a loop st
atement that contains this continue.
}
}
But the following is valid because the continue statement uses a label that nests
the continue statement.
LABEL_1 : for(int i = 0; i<10; i++){

if(i ==2) System.out.println(2);
for(int j = 0; j<10; j++){
if(i ==2) continue LABEL_1; //usage of continue i
s valid because it refers to an outer loop.
}
}
You can use labeled break and continue for non-nested loops as well but since
unlabeled break and continue are sufficient for breaking out of and continuing
with non-nested loops, labels are seldom used in such cases.
It is actually possible to use a labelled break (but not a labelled continue) inside
any block statement. The block doesn't necessarily have to be a loop block. For
example, the following code is valid and prints 1 3 .
public static void main (String[] args) {

MYLABEL: {
System.out.print("1 ");
if( args != null) break MYLABEL;
}
}
However, you need not spend time in learning about weird constructs because
the exam doesn't test you on obscure cases. The break and continue
statements are almost always used inside loop blocks and that is what the exam
focuses on.
To answer the questions on break and continue in the exam correctly, you need
to practice executing simple code examples on a piece of paper. The following is
one such code snippet:
public static void doIt(int h){

int x = 1;
LOOP1 : do{
LOOP2 : for(int y = 1; y < h; y++ ) {
if( y == x ) continue;
if( x*x + y*y == h*h){

System.out.println("Found "+x+" "+y);
break LOOP1;
}
}
x++;
}while(x<h);
}
Execute the above code mentally or on a piece of paper and find out what will it
print when executed with 5 as an argument and then 6 as an argument.
8.8 Comparing loop constructs

8.8.1 Comparison of loop constructs ☝
As you saw before, there are only a few technical differences between the four
kind of loops. The for, foreach, and while loops are conceptually a little
different than the do-while loop due to the fact that a do-while loop always
executes at least one iteration. The foreach loop is a little different than other
loops due to the fact that it can be used only for arrays and for classes that
implement java.util.Iterable interface.
Other than these differences, they are all interchangeable. For example, you can
always replace a while loop with a for loop as follows:
while( booleanExpression ){
for( ; booleanExpression ; ){

}
You can also replace a foreach loop with a for loop as shown below:
for( Object obj : someIterable){
for( Iterator it = someIterable.iterator() ; it.hasNex

t() ; ){
Object obj = it.next();
}
Of course, the foreach version is a lot simpler than the for version and that is the
point. While you can use any of the loops for a given problem, you should use
the one that results in code that is most understandable or intuitive.
A general rule of thumb is to use a for loop when you know the number of
iterations before hand and use a while loop when the decision to perform the
next iteration depends on the result of the operations performed in the prior
iteration. For example, if you want to print "hello" ten times, use a for loop, but
if you want to keep printing "hello" until the user enters a secret code as a
response, use a while loop!
8.9 Exercise ☝
1. Initialize a triangular matrix of ints using a for loop such the each element
contains an value equal to the sum of its row and column index. Do the
same using a while loop.
2. Write a method that determines whether a given number N is a prime
number or not by dividing that number with all the numbers from 2 to N/2
and checking the remainder.
3. Use nested for loops to print a list of prime numbers from 2 to N.
4. Use nested for loops to print out each element of the array referred to by
_3D in the format [i][j][k] = N :
int[] _1D1 = new int[]{1, 2, 3};

int[][] _2D1 = new int[][]{ _1D1 };
int[][] _2D2 = new int[][]{ _1D1, _1D1 };
int[][][] _3D = new int[][][]{ _2D1, _2D2 };
5. Write a method that accepts a String as input and count the number of
spaces in the string from start to until it finds an 'x' , or if there is no 'x'
in the string, till end.
6. The following code contains a mistake that is quite common while using
nested for loops. Identify the problem, fix it and print out all the elements of
chars array.
String[][] chars = new String[2][];

chars[0] = new String[2];
chars[1] = new String[4];
char cha = 97;
for(char c=0;c<chars.length; c++){
for(char ch=0;ch<chars.length; ch++){
chars[c][ch] = ""+cha;
cha++;
}
}
What will happen if char[0] is initialized as new String[1] instead

of new String[2] ?
7. To avoid the possibility of inadvertently introducing the mistake shown in
the above code, a programmer decided to use for-each loops instead of the
regular for loops:
for(String[] chara : chars){

for(String s : chara){
s = ""+cha;
cha++;
}
}
Is this a good idea?

8. Use an enhanced for loop to print alternate elements of an array. Can you
use an enhanced for loop to print the elements in reverse order?
9. Given two arrays of same length and type, copy the elements of the first
array into another in reverse order.
Chapter 9 Creating and Using Arrays
Declare, instantiate, initialize and use multi-dimensional arrays
9.1 Declare, instantiate, initialize and use a one-

dimensional array
9.1.1 Declaring array variables ☝
An array is an object that holds a fixed number of values of a given type. You
can think of an array as a carton of eggs. If you have a carton with 6 slots, then
that carton can hold only six eggs. Each slot of the carton can either have an egg
or can be empty. Observe that the carton itself is not an egg. Similarly, if you
have an array of size six of int values (or ints, for short), then that array can
hold six ints but the array itself is not an int .
An array of a given type cannot hold anything else except values of that type.
For example, an array of ints cannot hold long or double values. Similarly, you
can have an array of references of any given type. For example, if you have an
array of Strings, this array can only hold references to String objects. An
important point to note here is that even though we call it an "array of strings", it
does not actually contain String objects. It contains only references to String
objects. You cannot really have an array that contains actual objects.
Array declaration ☝
When you declare an array variable, you are basically specifying the type of
values that you want to deal with through that variable. The way to specify that
in Java is to apply [ ] , i.e., square brackets to the type of the values. For
example, if you want a variable through which you will deal with int values,
you will write int[] . Arrays can be multi-dimensional and there will be one
set of opening and closing brackets for each dimension. For example,
int i; //i is an int
int[] ia1, ia2; //ia1 and ia2 are one dimensional arr
ays of ints
int[][] iaa; //iaa is a two dimensional array of ints

and so on
An array declaration can never include the size of the array. Thus, the following
are declarations will not compile:
int[2] invalid1;
int[3][] invalid2;
int[][4] invalid3;
Unfortunately, the above method is not the only way to declare arrays. Java
allows you to apply square brackets to the variable name instead of type name as
well. For example,
int i, ia[]; //i is an int but ia is a one dimensiona

l array of int values
int[] ia, iaa[]; //ia is a one dimensional array of i

nts but iaa is a two dimensional array of ints and so
on
Observe that the rule of thumb of one set of square brackets per dimension still
holds. In the case of iaa , you have one set applied on the type and one set
applied on the variable, therefore iaa is a two dimensional array.
Arrays of objects are declared the same way. For example,
Object[] obja, objaa[]; //obja is a one dimensional a

rray of Objects but objaa is a two dimensional array o
f Objects
String[] strA; //strA is a one dimensional array of St

rings
Note that the statements shown above only declare array variables. They don't
actually create arrays. Array creation and initialization is a topic in its own right
and that is what I will discuss next.
9.1.2 Creating and initializing array objects ☝
Creating arrays using array creation expressions
You use the new keyword to create an array object. For example,
int[] ia = new int[10]; //an array of ints of size 10
boolean[] ba = new boolean[3]; //an array of booleans

of size 3
String[] stra = new String[5]; //an array of Strings o

f size 5
MyClass[] myca = new MyClass[5]; //an array of MyClass

of size 5
int[] invalid = new int[];//missing size. will not com
pile.
The parts on the right-hand side of = sign in the above statements are called
"array creation expressions" . These expressions merely allocate space to hold
10 ints , 3 booleans , 5 string references , and 5 MyClass references
respectively. Every element of the array is also initialized to its default value
automatically by the JVM. The default values of array elements are very
straightforward - references are initialized to null , numeric primitives to 0 , and
booleans to false . In the above lines of code, ia is set to point to an array of ten
int values and each element of the array is initialized to 0, ba is set to point to an
array of three boolean values and each element of the array is initialized to
false , stra is set to point to an array of five String references and each
element of the array is initialized to null , and finally, myca is set to point to
an array of five MyClass references and each element of the array is initialized
to null . Observe that all the elements of an array are initialized to the same
value. This shows us another important aspect of arrays - that arrays can have
duplicate values .
It is important to understand that, in the above statements, we are not creating

instances of the class of the array elements. In other words, we are not creating
instances of String or instances of MyClass . (We are not creating instances
of ints or booleans either, for that matter, but since they are primitives, and since
primitives are not objects, they don't have instances anyway.)
In Java, arrays, whether of primitives or objects, are objects of specific classes.

In other words, an array object is an instance of some class. It not an instance of
Object class but since Object is the root of every class in Java, an array object is
an Object and all methods of the Object class can be invoked on an array. Let us
now look at the following program and its output to know more about the class
of the above defined array objects -

int[] ia = new int[10];
boolean[] ba = new boolean[3];
String[] stra = new String[5];
TestClass[] ta = new TestClass[5];
System.out.println(ia.getClass().getName()+" , "+i
a.getClass().getSuperclass().getName());
System.out.println(ba.getClass().getName()+" , "+b
System.out.println(stra.getClass().getName()+" , "
+stra.getClass().getSuperclass().getName());
System.out.println(ta.getClass().getName()+" , "+t
}
}
Output:
[I , java.lang.Object
[Z , java.lang.Object
[Ljava.lang.String; , java.lang.Object
[LTestClass; , java.lang.Object
The output shows that ia , which is declared to be of type int[] , is not an

instance of int but of a class named [I . stra , which is declared to be of type
String[] , is not an instance of String but of a class named
[Ljava.lang.String and so on. These names of the classes look weird.
Actually, Java cooks up the name of the class of an array by looking at the
number of dimensions and the type of the elements. For each dimension, there is
one opening square bracket. This is followed by a letter for the class of the
elements as per the following table and, if the array is not of a primitive, the
name of the class followed by a semi-colon.
Type Letter
boolean Z
byte B
char C
short S
int I
long J
float F
double D
any Object L
Based on the above table, the name of the class for long[][] (i.e. a two
dimensional array of longs) would be [[J and the name of the class for
mypackage.SomeClass[][][] (i.e. a three dimensional array of
mypackage.SomeClass ) would be [[[Lmypackage.SomeClass; .
Although the above discussion about the class of arrays

is not included in the exam objectives, a few test
takers have reported getting a question on the exam
that requires this information.
Creating arrays using array initializers
In the previous section, we created array objects using the new keyword. It is
possible to create array objects without using the new keyword. For example, the
arrays that we created above can also be created as follows:
int[] ia = {0, 1, 2, 3, 4, 5, 6, 7, 8, 9 }; //an array

of ints of size 10
boolean[] ba = {true, false, false}; //an array of boo

leans of size 3
String[] str = {"a", "b", "c", "d", "e"}; //an array o

f Strings of size 5
MyClass[] myca = { new MyClass(), new MyClass(), new M

yClass(), new MyClass(), new MyClass()} ; //an array o
f MyClass objects of size 5
The parts on the right-hand side of the = sign in the above statements are called
"array initializers" . An array initializer is a shortcut that allows you to create
as well as initialize each element of the array with the values that you want
(instead of the default values that you get when you use array creation
expressions).
Since the compiler can find the type of the elements of the array by looking at
the declaration (i.e. the left-hand side of the statement), specifying the same on
the right-hand side is not required. Similarly, the compiler finds out the length of
the array as well by counting the number of values that are specified in the
initializer.
Array initializers can also be mixed with array creation expressions. For
example:
int[] ia = new int[]{ 1, 2, 3, 4, 5 };
Observe that size is missing from the expression on the right-hand side. Java
figures out the size of the array by counting the number of elements that are
specified in the initialization list. In fact, it is prohibited to specify the size if you
are specifying individual elements. Therefore, the following is invalid:
int[] ia = new int[2]{ 1, 2 }; //will not compile.
9.2 Using arrays

9.2.1 Array indexing ☝
In Java, array indexing starts with 0. In other words, the index of the first
element of an array is 0. For example, if you have an array variable named
accounts that refers to an array of 100 Account objects, then you can access
the first object through accounts[0] and the 100th object through
accounts[99] . accounts[0] or account[99] are like any other
variable of type Account .
Similarly, if you have an array variable ia pointing an array of 5 ints, the first
element can be accessed using ia[0] and the last element using ia[4] .
java.lang.ArrayIndexOutOfBoundsException ☝
If you try to access any array beyond its range, JVM will throw an instance of
ArrayIndexOutOfBoundsException . For example, if you have int[]
ia = new int[3]; the statements int i = ia[-1]; and int i =
ia[3]; will cause an ArrayIndexOutOfBoundsException to be
thrown.
Do not worry if you don't know much about exceptions at this point. I will
discuss exceptions in detail later.
Arrays of length zero ☝
As strange as it may sound, it is possible to have an array of length 0. For

example, int[] ia = new int[0]; Here, ia refers to an array of ints of
length 0. There are no elements in this array. It is important to understand that an
array of length 0 is not the same as null . ia = null implies that ia points to
nothing. ia = new int[0]; implies that ia points to an array of ints whose
length is 0.
A good example of an array of length 0 is the args parameter of the main

method. If you run a class with no argument, args will not be null but will
refer to an array of Strings of length 0.
Changing the size of an array ☝
Java arrays are always of fixed size(or length). Once you create an array, you
cannot change the number of elements that this array can have. So, for example,
if you have created an array of 5 ints and if you have 6 int values to store, you
will need to create a new int array. There is no way to increase or decrease the
size of an existing array. You may change the array variable to point to a
different array, and you can, of course, change the values that an array contains.
9.2.2 Members of an array object ☝
We saw earlier that arrays are actually objects of specific classes. We also saw
the names of some of these classes. But what do these classes contain? What are
the members of these classes? What functionality do these classes provide? Let
us take a look:
Array length
All array classes have one field named length , which is of type int . This field is
public and it stores the length of the array. This field is also final , which reflects
the fact that you cannot change the length of an array after its creation. Since the
length of an array can never be less than 0, the value of this field can never be
less than 0 either.
Array cloning
All array classes also have a public method named clone . This method creates a
copy of the array object. Note that it doesn't create copies of the objects referred
to by the array elements. It merely creates a new array object of the same length
and copies the contents of existing array into the new array. Which means, if the
existing array contained primitive values, those values will be copied to the
elements of the new array. If the existing array contained references to objects,
those references will be copied to the elements of the new array. Thus, the
elements of the new array will also point to the same objects. This is also known
as "shallow copy" .
For example, the following figure shows exactly what happens when you clone
an array of five ints referred to by the variable myIntArr1 using the statement
int[] myIntArr2 = (int[]) myIntArr1.clone(); Don't worry
about the explicit cast for now, but observe that a new array object with the same
number of slots is created and the contents of the slots of the existing array are
copied to the slots of the new array.

The following figure shows what happens when you clone an array of Strings
referred to by the variable myStringArr1 using the statement String[]
myStringArr2 = (String[]) myStringArr1.clone(); Observe
that the references are copied into the new array and that no new String objects
are created in the heap space.
Cloning is not required for the exam but it is an

important aspect of arrays anyway.
Members inherited from Object class
Remember that since java.lang.Object is the root class of all classes, it is

the root class of all array classes as well and therefore, array classes inherit all
the members of the Object class. This includes toString , equals , and
hashCode methods.
9.2.3 Runtime behavior of arrays ☝
Although arrays are the simplest of data structures, they are not without their
quirks. To use arrays correctly and effectively, you must be aware of the two
most important aspects of arrays in Java.
The first is that they are "reified" . Meaning, the type checking of arrays and its
elements is done at runtime by the JVM instead of at compile time by the
compiler. In other words, the type information of an array is preserved in the
compiled bytecode for use by the JVM during run time. The JVM knows about
the type of an array and enforces type checking on arrays.
For example, if you have an array of Strings, the JVM will not let you set any
element of that array to point to any object other than a String, even though the
compiler may not notice such a violation as illustrated by the following code:
String[] sa = { "1", "2", "3" };

Object[] oa = sa;
oa[0] = new Object();
The above code will compile fine. The compiler has no objection when you try
to assign an array of Strings to a variable of type array of Objects. It has no
objection when you try to set an element of this array of Strings to point to an
object that is not a String. But when you run it, the JVM will throw a
java.lang.ArrayStoreException on the third line. This is because the
JVM knows that the array pointed to by oa is actually an array of Strings. It will
not let you corrupt that array by storing random objects in it.
It is important to understand this concept because it is diametrically opposite to

how generics work. This is the reason why arrays and generics operate well with
each other. You will realize this when you learn about generics later.
The second thing about arrays is that they are "covariant" . Meaning, you can
store a subclass object in an array that is declared to be of the type of its
superclass. For example, if you have an array of type java.lang.Number ,
you can store java.lang.Integer or java.lang.Float objects into
that array because both are subclasses of java.lang.Number . Thus, the
following code will compile and run fine:
Number[] na = { 1, 2, 3 };
na[0] = new Float(1.2f);
We haven't seen anyone getting a question on array

reification and covariance in the OCP Java 11 Part 1
exam. However, any discussion about arrays is
incomplete without these two points and that is why I
have talked about them here.
9.2.4 Uses of arrays ☝
Arrays are quite powerful as a data structure but they are somewhat primitive
as a data type . As we saw earlier, arrays have only one field and merely a
couple of methods. But because of their simplicity, arrays are used as building
blocks for other data types and data structures. For example, the String data
type is built upon an array of char s. So are StringBuffer and
StringBuilder . These higher level classes really only wrap an array of
characters and provide methods for manipulating that array.
Arrays are also used extensively for building higher level data structures such as
List , Stack , and Queue . You may have come across the ArrayList
class (this class is also on the exam, by the way, and I will discuss it later).
Guess what, it is a List that manages its collection of objects using an array
inside. Since these classes provide a lot of additional features on top of arrays,
more often than not, it is these classes that get used in application programs
rather than raw arrays.
java.util.Arrays class ☝
Java standard library does include a utility class named Arrays in package
java.util that makes working with raw arrays a little easier.
java.util.Arrays class contains a large number of static utility methods
for manipulating any given array object. Although not mentioned explicitly in
the part 1 exam objectives, it is a good idea to browse through the JavaDoc API
description of Arrays class to know about these methods. We have seen
questions in the exam on the compare and mismatch methods that were
added to this class in Java 9, so, I will discuss these two methods in detail.
Arrays.compare and Arrays.mismatch methods ☝
The compare and mismatch methods are static methods and are meant to
compare two arrays of the same types. Although they can compare arrays of
primitives as well as arrays of reference types, they are most useful for
comparing primitive arrays. Efficient comparison of primitive arrays is a
common requirement in data processing. Since these methods compare data in
chunks, they are able to perform substantially better than custom code that
compares elements individually. That is probably why the OCP Java 11 exam
has questions on these methods.
If you browse through the API JavaDoc, you will notice a large number of
variants for these methods. Although these methods are heavily overloaded,
practically, there are only two flavors that you need to understand for the
purpose of the exam. The rest are mere replicas; one for each data type.
Before getting into the details of the methods, you need to understand the
meaning of the two words that are used often in the description for these
methods.
1. Lexicographical - This simply means "dictionary order", i.e., the way in

which words are ordered in a dictionary. If you were to determine which one of
the two words would come before the other in a dictionary, you would compare
the words letter by letter starting from the first position. For example, the word
"children" will appear before (and can be considered smaller) than the word
"chill" because the first four letters of the two words are the same but the
letter at the fifth position in "children" (i.e. d ) comes before (or is smaller
than) the letter at the fifth position in "chill" (i.e. l ). Observe that
lengthwise, the word "children" is bigger than "chill" but length is not
the criteria here. For the same reason, an array containing 12345 will appear
before an array containing 1235 .
2. Prefix - This refers to the common part of the two arrays. In the above
example, "chil" can be called the prefix while comparing "children" and
"chill" . Similarly, 123 is the prefix while comparing 12345 and 1235 .
Let's check out the JavaDoc descriptions of compare and mismatch methods
now.
public static int compare(int[] a, int[] b) : Compares two int arrays

lexicographically. If the two arrays share a common prefix then the
lexicographic comparison is the result of comparing two elements, as if by
Integer.compare(int, int), at an index within the respective arrays that is the
prefix length. Otherwise, one array is a proper prefix of the other and,
lexicographic comparison is the result of comparing the two array lengths.
A null array reference is considered lexicographically less than a non-null array

reference. Two null array references are considered equal.
It returns 0 if the first and second array are equal and contain the same elements
in the same order; a value less than 0 if the first array is lexicographically less
than the second array; and a value greater than 0 if the first array is
lexicographically greater than the second array.
For example, the following code will print -1 because the first array has smaller
integer at index 2.
int[] ia1 = { 0, 1, 2, 6};

int[] ia2 = { 0, 1, 5};
System.out.print(Arrays.compare(ia1, ia2)); //prints -1
Observe that the returned value (-1) does not depend on the amount of different
between the two elements (2 - 5 is -3).
public static int mismatch(int[] a, int[] b) : Finds and returns the index of the
first mismatch between two int arrays, otherwise return -1 if no mismatch is
found. The index will be in the range of 0 (inclusive) up to the length (inclusive)
of the smaller array.
If the two arrays share a common prefix then the returned index is the length of
the common prefix and it follows that there is a mismatch between the two
elements at that index within the respective arrays. If one array is a proper prefix
of the other then the returned index is the length of the smaller array and it
follows that the index is only valid for the larger array. Otherwise, there is no
mismatch.
Two non-null arrays, a and b , share a common prefix of length pl if the

following expression is true :
pl >= 0 &&
pl < Math.min(a.length, b.length) &&
Arrays.equals(a, 0, pl, b, 0, pl) &&
a[pl] != b[pl]
Note that a common prefix length of 0 indicates that the first elements from each
array mismatch.
Two non-null arrays, a and b , share a proper prefix if the following expression
is true:
a.length != b.length &&

Arrays.equals(a, 0, Math.min(a.length, b.length),
b, 0, Math.min(a.length, b.length))
For example, the following code will print 2 because the arrays differ at index 2:
int[] ia1 = { 0, 1, 2, 6};

int[] ia2 = { 0, 1, 5};
System.out.println(Arrays.mismatch(ia1, ia2)); //print
s 2
Observe that 2 is also the length of the prefix, which is {0, 1} .

You should go through the JavaDoc API description of the various overloaded
compare and mismatch methods. You will notice that they all work in the same
basic fashion.
9.3 Declare, instantiate, initialize and use multi-

dimensional arrays
9.3.1 Multidimensional Arrays ☝
The phrase multidimensional array brings a picture of a matrix to mind. But it is

important to understand that Java doesn't have matrix kind of multidimensional
arrays. What Java has is arrays whose elements themselves can be arrays. Recall
that, in Java, every array object is an object of a specific class. For example, the
class of an array of ints is [I . Now, what if you want to have an array of
objects of this class. In other words, an array of "array of ints" . You can
declare it like this:
int[][] iaa;
Visually, the declaration looks like iaa is a two-dimensional matrix of ints. But
in reality, iaa points to a single dimensional array, where each element of the
array is an array of ints. There is a fundamental difference between the two
approaches. In a matrix, the number of elements in a given dimension are the
same at each index of the higher dimension. For example, in a two dimensional
array, the length of each row will always be the same, i.e., each row will have
same number of columns. While in an array of arrays, there is no such
restriction. Each row can refer to an array of any length. This is illustrated in the
following figure:
In this figure, iaa[2x3] , which is a made up syntax (Java does not have this),
is a matrix of size 2x3, i.e., 2 rows and 3 columns. Since the size of each
dimension is known in advance, it can be easily stored in a contiguous chunk of
2x3=6 memory cells. You can also locate the address of any element using a
simple formula.
But in case of an array of array of ints, you can only allocate 2 continuous
memory addresses to store two references - one for each array of ints. These two
references can, in turn, point to two arrays of different lengths. In the above
figure, iaa[0] points to an array stored at location 1111, and iaa[1] points to
another array stored at location 1200. iaa[0] points an array of length 2, while
iaa[1] points to an array of length 3. This is what Java has. Since the array of
arrays are not required to be symmetric, such arrays are called "jagged arrays"
. FYI, C# supports both kinds of arrays, i.e., symmetric as well as jagged.
Keeping the above discussion in mind, let us now look at the rules of declaring
and creating an array of arrays in Java:
1. The type of an array is determined by the number of pairs of square brackets
applied to the variable. For example, in case of int[] ia; and int ia[];
ia is an array of ints. In case of int[] iaa[]; and int iaa[][]; iaa is
an array of arrays of ints. int[][] iaaa[]; is an array of arrays of arrays of
ints.
2. You never specify the length of the array in the type declaration. Thus,
int[3] ia; and int[2][] iaa; are invalid declarations.
3. The rules of array creation expressions and array initializers that we talked
about in the previous lesson, are applicable here as well.
The following are a few examples:
1. int[][] iaa = new int[2][3];

iaa is created using an array creation expression. iaa refers to an array of
length 2. Each element of this array refers to an array of ints of length 3.
Each element of both the arrays of ints is initialized to 0.
2. int[][] iaa = new int[3][];

iaa is created using an array creation expression. iaa refers to an array of
length 3. Can you guess what each of the three elements of this array are
initialized to? Observe that the type of each element is "array of ints",
which means iaa is an array of objects (and not of primitives). Since every
element of array of objects is automatically initialized to null , each
element of the array pointed to by iaa is initialized to null . You can
make them point to arrays of ints like this:
iaa[0] = new int[2]; //iaa[0] points to an array of

ints of length 2
iaa[1] = new int[3]; //iaa[1] points to an array of
ints of length 3
Each element of these two arrays is now initialized to 0 but iaa[2] is still
null .
This example illustrates another important aspect of creating arrays. Notice

the difference between the specification of dimension sizes of example 1
and this one. Example 1 has [2][3] , while here, it is [3][] . We have
omitted the size of the second dimension. The size of the first dimension
tells the JVM how many references you want to store in your array.
(Remember that the type of those references that you want to store is array
of ints, i.e., int[] ). [3] implies that you want to store three references.
In other of words, the length of your array (which is of type array of ints) is
3. However, the size of the second dimension is not needed because the
arrays pointed to by those references can be created later and can be of
different lengths. You will need to specify their lengths only when you
create them.
What happens when you create a three dimensional array of ints, i.e., an
array of array of array of ints? The process is same. You only need to tell
the JVM how many number of references (whose type will be array of array
of ints, i.e., int[][] ) do you want to keep in your array. If you want only
2 such references, you can create it with new int[2][][] . The
following three figures illustrates what happens in each step as you
initialize an array of array of array of ints.
You can now set the first slot of this array to point to an array of length 2
using iaaa[0] = new int[2][];
Note that iaaa[1] is still null. Next, you can set iaaa[0][0] to point
to an array of 4 ints using iaaa[0][0] = new int[]{ 1, 2, 3,
4}; or iaaa[0][0] = { 1, 2, 3, 4}; (You can also do
iaaa[0][0] = new int[4]; and in that case the value of each
element will 0).

You cannot, however, leave out the size of a higher dimension if you want
to specify the size of a lower dimension. For example, you cannot do new
int[][2]; The reason is simple - new int[][2] tries to create an
array of int[2] objects. But it does not tell the JVM how many int[2]
objects you want to store. Without this information, the JVM has no idea
how much space it needs to allocate for this array. On the other hand, new
int[2][] is fine because now, you are telling the JVM that you want to
create an array of length 2. In this case, the JVM is clear that it needs to
allocate space to store 2 references. Remember that the size of a reference
doesn't depend on the length of the array to which it points. So, the JVM
doesn't care about the length of the arrays to which these two references
will refer. It simply allocates space to store 2 references.
3. int[][] iaa = new int[][]{ new int[]{ 1, 2 } }; This

statement uses an array creation expression coupled with array initializer.
int[][] iaa = { { 1, 2 } }; This is the same as above but with

array initializer.
In both the cases, iaa refers to an array of length 1. The first and only
element in this array refers to an array of ints of length 2.
4. Object[] obj[] = { {"string is also an object"},

{null } , { new Object() , new Integer(10)} };
obj refers to an array of array of objects. The length of the array is 3. The
first array refers to an array of objects of length 1. The only element of this
array of objects points to a String that contains "string is also an object".
The second array refers to an array of length 1. The only element of this
array points to null . The third array refers to an array of length 2. The
first element of this array points to an object of class Object and the second
element points to an object of class Integer.
Here is a simple program that prints out useful information from an array of
arrays. I suggest you play around with it by changing the arrays.

{
Object[] iaa[] = { {"string is also an object"
}, {null } , { new Object() , new Integer(10)} };
for(int i=0; i<iaa.length; i++){
System.out.println("Element["+i+"] contains
an array of length "+iaa[i].length);
for(int j = 0; j<iaa[i].length; j++){
System.out.println(" Element["+i+"]["+j
+"] contains "+iaa[i][j]);
}
}
}
}
The output is:
Element[0] contains an array of length 1

Element[0][0] contains string is also an object
Element[1][0] contains null
Element[2][0] contains java.lang.Object@2a139a55
Element[2][1] contains 10
As an exercise, modify this code to print out the contents of an array of array of
array of int s.
You will see questions in the exam that require you to iterate through
all the elements of multi-dimensional arrays using a for loop as well
as while/do-while loop.
9.3.2 Assigning arrays of primitives to Object variables ☝
I explained earlier that every array is an object. This means, you can assign any
array object to variable of type Object . Like this:
int[] intArray = new int[]{ 0, 1, 2 };

Object obj = intArray;
This is valid because an array of int s is an Object .
Then how about this - Object[] oa = intArray; ? This will not compile
because elements of the array pointed to by intArray are not Object s. They
are int s. Therefore, you cannot assign an array of int s to a variable of type
array of Object s.
You need to be very clear about this concept because it gets confusing very
quickly as the following example shows:
Object[] oa = new int[2][3]; //this is valid.
Object[][] oaa = new int[2][3]; //this will not compil

e.
Of course, each element of the array created using new int[2][3] is an

array of int s. An array of int s is an Object and therefore, an array of array
of int s is an array of Object s.
9.4 Exercise ☝
1. Create a array of booleans of length 3 inside the main method. Print the
elements of the array without initializing the array elements explicitly.
Observe the output.
2. Given int[] first = new int[3]; , int[] second = {}; ,
and int[] third = null; , print out the length of the three arrays
and print out every element of the three arrays.
3. Create an array of chars containing four values. Write assignment
statements involving the array such that the first element of the array will
contain the value of the second element, second element will have the value
that was there in the third element. and third element will contain the value
of the fourth element.
4. Declare and initialize an array of length 4 of type array of Strings without
using the new keyword such that no two arrays of Strings have the same
length. Print the length of all of the arrays one by one (including the length
of the two dimensional array).
5. Given the statement String[][] names = new String[2][3];
How many Strings will you need to fill up names completely? Initialize
each element of names with a non-null String. Print each of those values
one by one without using a loop. Do the same using nested for loops after
going through the chapter on loops.
6. Define two variables of type array of Strings. Initialize them using the
elements of the array created in the previous exercise.
7. Define a simple class named Data with a public instance field named
value of type int . Create and initialize a Data variable named d in
TestClass 's main. Create an array of Data of length 3 and initialize
each of its elements with the same Data instance. Use any of the array
elements to update the value field of the Data object. Print out the
value field of the Data object using the three elements of the array.
Finally, print the value field of the original Data using the variable d .
8. Define and initialize an array of array of ints that resembles a triangular
matrix of integers.
9. Declare another array of array of ints and initialize it using the elements of
the array created in the previous exercise in such a way that it resembles an
inverted triangular matrix of integers.
10. Declare and initialize a variable of type array of Objects of length 3.
Initialize the first element of this array with an array of ints, second with an
array of array of ints, and third with an array of Objects. See if any of the
assignments fails compilation.
11. Given the statement int[][] nums = new int[1][3]; , how
many int values can nums store? Write down how each element of nums
can be addressed.
12. Given the following code:
int[][][] nums = new int[1][4][2];

for(int i = 0; i<nums.length; i++){
for(int j = 0; j<nums[i].length; j++){
for(int k = 0; k<nums[i][j].length; k++){
nums[i][j][k] = i + j + k;
System.out.println("num["+i+"]["+j+"]["+k+"
] = "+nums[i][j][k]);
}
}
}
Execute it mentally and write down its output on paper. Run the code and
check your answer.
Chapter 10 Creating and Using
Methods
10.1 Create methods with arguments and return

values
10.1.1 Creating a method ☝
We have already seen methods in previous chapters. Indeed, we have been using
the "main" method to run our test programs all along. But we haven't discussed
them formally yet.
A method is what gives behavior to a type. Java allows classes, interfaces, and
enums to have a behavior. A behavior is nothing but a high level action
performed by a piece of code. If you expect this action to be performed as and
when required, you put this piece of code in a method and give it a name. This
lets you invoke that action whenever you need by using the name. It is like a
black box to which you give some input and get back some output in return.
Thus, the basic structure of a method is as follows:
returnType methodName(parameters) {
methodBody
}
A ReturnType specifies what the method returns as a result of its execution.

For example, a method that returns the sum of two integers may specify that it
returns an int as a result. A method that returns an Account object may
specify Account as the return type. If a method doesn't return anything, it must
specify so, using the keyword void .
It is important to know that a method can return one thing at the most.
A MethodName is the name given to this method. It must be a valid Java

identifier. Remember that an identifier is an unlimited-length sequence of Java
letters and Java digits, the first of which must be a Java letter. It cannot be a
reserved word or a keyword.
The Parameters of a method are specified by a list of comma-separated

parameter declarations. Each parameter declaration consists of a type and an
identifier that specifies the name of the parameter. If a method does not take any
parameter, the parameters list will be empty. The following are a few examples
of a valid parameter list:
void save() //no parameters
void saveAccount(Account acct) //takes an Account as a

parameter
void add(int a, int b) //takes two ints as parameters
Note that, unlike regular declarations, each parameter in a parameter list must be
specified along with its type individually. Thus, while int a, b; is a valid
statement on its own that declares two int variables, add(int a, b) is an
invalid declaration of method parameter b .
Each parameter in the parameter list may also be declared as final if the code in
the method does not change its value. For example:
void add(int a, final
int b) //b is final.
If a parameter is declared final and if the method body tries to change its value,
the code will fail to compile.
The MethodBody contains the code that is to be executed upon invocation of

that method. It must be contained within curly brackets. Observe that this is
unlike other block statements such as if/else, for, and while, where you can omit
the curly brackets if there is just one statement in their body. In the case of
methods, curly brackets are required even if the body consists of just one
statement.
There are several other bells and whistles associated with methods such as var-
args, accessibility, static, abstract, final, and exception handling. I will discuss all
of these as we go along.
10.1.2 Returning a value from a method ☝
A method must always return a value of the type that it promises to return in its
declaration after successful completion. This means that it is not possible to
return a value conditionally as shown in the following code:
public int get2X(int x){

if(x>0) return 2*x;
}
The compiler will generate an error message saying, "error: missing

return statement" because it notices that the method will not return
anything if the if condition is false. This makes sense if you consider what will
happen to the caller of the get2X method if the method doesn't return anything.
For example, what value should be assigned to y by this statement - int y =
get2X(-1); ? There is no good answer. Thus, the method must return an int
in all situations (except when it throws an exception, but you can ignore that for
now).
This rule applies even to methods that return a reference type (and not just to
method that return a primitive). But if the return type of a method is a reference
type, it is ok for the method to return null because null is a valid value for a
reference. In other words, returning null is not the same as returning nothing.
Thus, the following method is fine:
String getValue(int x){

if(x > 0) return "good day!";
else return null; //this is ok.
The only situation where a method can avoid returning a value is if it ends up
throwing an exception, which means that the method didn't really finish
successfully and therefore, it cannot be expected to return a value!
Method returning void ☝
If a method says that it doesn't return anything (i.e. its return type is specified as
void ), then it must not return anything in any situation. Obviously, it cannot
even return null because as we saw above, null is not the same as nothing.
But the interesting thing is that it cannot even return void . Thus, the following
method will not compile.
void doSomething(){
return void; //invalid
There are only two options in this case - do not have any return statement at all
or have an empty return statement, i.e., return; For example:
void doSomething(){
return; //empty return
or
void doSomething(){
//no return statement at all
Returning values of different types from a method ☝
The general rule is that Java does not allow a method to return a value that is of a
different type than the one specified in its declaration. This means that if a
method says that it returns an int , it cannot then return a boolean value.
However, there are three exceptions to this rule. Two are about primitives and
one is about references.
1. Numeric promotion - If the return type of a method is a numeric type (i.e.

byte, char, short, int, long, float, and double ), then
the return value can be of any other numeric type as long as the type of the
return value is smaller than the type of the declared return type. For
example, if a method says it returns an int , it is ok for that method to
return a byte, short, or char value. But it cannot return a long,
float, or double value. Thus, the following code will compile fine:
public int getVal(int x){

char ch = 'a';
byte b = 0;
if(x>0) return ch;
else return b;
}
This is allowed because a smaller type can easily be promoted to a larger

type without any loss of information.
2. Autoboxing/Unboxing - Java allows a return value to be a reference to a

wrapper class if the return type is of a primitive type of the same or smaller
type. Thus, the following code is ok:
public int getVal(){

return new Integer(10);//wrapper object will
be unboxed into a primitive
The reverse is also allowed:
public Integer getVal(){

return 10; //primitive will be boxed into a w
rapper object
3. The third exception is related to inheritance. Since I haven't discussed this

topic yet, I will only mention it briefly here. Don't worry if you don't
understand it completely at this point.
It is ok for a method to return a reference of a subtype of the type declared
as its return type. This is called "covariant return types." For example, if a
method declares that it returns an Object, it is ok for the method to return a
String because a String is also an Object. The following code shows some
valid possibilities:
Object getValue(){
return "hello";
//return 10; //This is ok, 10 will be boxed into
an Integer object, which is an Object.
//return null;//This is ok too.
This is the same as promising someone that you will give them a fruit, and
then give them a banana. This is ok because a banana is also a type of fruit.
But the reverse is not true. If you promise someone that you will give them
a banana, you cannot then give any other fruit. Thus, the following will fail
to compile:
String getValue(){
return new Object();// will not compile because
an Object is not a String.
It will become clearer after you learn about inheritance.
These rules ensure that only those values that are "assignable" to the declared
return type are returned by a method. In other words, if you can assign a
particular value to a variable of the declared return type directly, then you can
return that value from that method. For example, you can assign a char value to
an int variable directly, therefore, you can return a char value from a method
that declares that it returns an int .
Returning multiple values from a method ☝

Java does not allow a method to return more than one value. Period. This seems
like difficult restriction to overcome if you want to return multiple values from a
method. For example, what if your getName method wants to return first name
and last name separately? Well, the way to do that in Java is to use a class to
capture multiple values and then return a reference to an object of that class.
Here is what your getName method may look like:
Name getName(){
Name n = new Name(); //capture two values in a Name
object
n.firstName = "Ann";
n.lastName = "Rand";
return n; //return a reference to the Name object
This code assumes the existence of a separate class called Name that can capture
the two components of the name:
class Name{
String firstName, lastName;
}
You may also use arrays to overcome this restriction. For example, the getName
method can also be coded as follows:
String[] getName(){
return new String[]{"ann", "rand" };
}
Generally, classes are designed to capture values that

are related to each other. For example, you might have
a Student class that captures a student's id, name, and
address. You may, however, encounter situations where
you want to return multiple unrelated values from the
same method. You can use the same approach to
return these values. Classes that are used to capture
unrelated values are called holder classes. If you
encounter such situations too often in your code then it
is a symptom of a bad design. A detailed discussion on
this is beyond the scope of the exam.
10.1.3 Varargs ☝
If you know the type of the arguments but don't know the exact number of
arguments that the caller is going to pass to a method, you can put the arguments
in an array and pass the array to the method. For example, if you want to write a
method that computes the average of any number of integer values, you could do
something like this:
public double average(int[] values){

/* by the way, can you tell what will happen if sum
is declared as int?
Expect questions in the exam that seem to be abo
ut one topic
but are actually about something entirely differ
ent.
*/
double sum = 0;
for(int i=0; i<values.length; i++) sum += values[i]
;
return values.length==0?0 : sum/values.length;
}
The caller of this method can put all the integer values in an array and call it as
follows:
int[] values = { 1, 2, 3, 4 };
double average = average(values);
This approach works fine but is a little tedious to write. Java 5 introduced a new
syntax called "varargs" that makes passing a variable number of arguments to a
method a little easier. Instead of using an array parameter in the method
declaration, you use a varargs parameter by appending three dots to int like
this:
public double average(int...
values){
//same code as shown earlier goes here
There is absolutely no change in the method body. Within the method body,
values remains an array of integers like before. The caller, however, does not
need to create an array explicitly. It can simply pass any number of int arguments
in the method invocation. Thus, the following are all valid method invocations of
the new average method.
double average = average(); //no argument
double average = average(1); //one argument
double average = average(1, 2, 3, 4); //multiple argum

ents
An important point to understand here is that the varargs syntax is just a

syntactic sugar for the developer. It saves a few keystrokes while typing the code
but makes absolutely no difference to the resulting bytecode generated by the
compiler. When the compiler sees the invocation of a method with varargs
parameter, it simply wraps the arguments into an array and passes the array to
the method. Indeed, the old code that used an array to call the average method
still works with the new average method. If you update the method code to print
the number of elements in the values array (just add
System.out.println(values.length); ), you will see it print 0, 1,
and 4 for the above three invocations respectively.
Observe that if you don't pass any argument, the compiler will create an array
with a length of zero and pass that array. The method will not receive a null
but an array of length zero in this case. This is unlike the other ways that you
invoke the array version of the method. It is not possible to invoke the array
version without any argument. If you don't want to pass any value, you will have
to pass null . Try it out and see what happens.
You can apply the varargs approach to a parameter of any type and not just
primitives.
Restrictions on varargs ☝
Let us try to expand the usage of varargs from one parameter to two. What if we
define a method as follows:
public double test(int... p1, int... p2){

System.out.println(p1.length+" "+p2.length);
}
and call this method like as follows?
test(1, 2);
This poses a problem for the compiler. It has three equally valid possibilities for
initializing p1 and p2 . It can create two int arrays containing {1} and {2} , or
it can create {1, 2} and { } , or it can create { } and { 1, 2 } and pass
them as arguments for p1 and p2 respectively.
Resolving this ambiguity through complicated rules would make the varargs
feature too confusing to use and so, Java imposes the following two restrictions
on varargs:
1. A method cannot have more than one varargs parameter.
2. The varargs parameter, if present, must be the last parameter in the

parameter list of a method.
With the above two rules in place, it is easy to understand what will happen if
you have a method as follows:
void test(int x, int... y){

//some code here
and call it as follows:
test(1, 2); //x is assigned 1, y is assigned {2}
test(1); //x is assigned 1, y is assigned { }
test(1, 2, 3, 4); //x is assigned 1, y is assigned { 2

, 3, 4}
Note that since x is a non-varargs parameter, any invocation of the method test
must include a value for x . Therefore, a call to test() with no argument will
not compile.
10.2 Create overloaded methods
10.2.1 Method signature ☝
Before learning about method overloading, you need to know about something
called "method signature".
A method signature is kind of an "id" of a method. It uniquely identifies a

method in a class. A class may have several methods with the same name but it
cannot have more than one method with the same signature . When you call a
method of a class, you basically tell the compiler which method you mean to call
by mentioning its signature. If a class has multiple methods with the same
signature, the compiler will not be able to determine which method you mean
and will raise an error.
The question then is: what exactly constitutes a method signature? Simple.
Method signature includes just the method name and its ordered list of
parameter types . Nothing else. For example, all of the following method
declarations have the same signature:
1. void process(int a, string str);

2. public void process(int value, String name);
3. void process(int a, String str) throws Exception;
4. String process(int a, String str);
5. private int process(int a, String data);
6. static void process(int a, String str);
Observe that in all of the above cases, the method name (i.e. process) and the
ordered list of parameter types (i.e. int , String ) are exactly the same and
their access types, static/instance types, return types, parameter names, and the
throws clauses are all different. However, since these attributes are not part of
the method signature, they do not make the methods different. The compiler will
consider all of the above methods as having the same signature and will
complain if you have any two of the above methods in the same class.
10.2.2 Method overloading ☝
Now that you know about method signature, method overloading is a piece of
cake. You know that a class cannot have more than one method with the same
signature. But it can certainly have multiple methods with the same name and
different parameter types because having different parameter types would make
their signatures different. In such a situation where a class has multiple methods
with same name, it is said that the class has "overloaded" the method name.
The method name is overloaded in the sense that there are multiple possibilities
associated with that name.
From the compiler's perspective, method overloading is nothing special. Since

the compiler cares only about the method signatures, it makes no difference to
the compiler whether two methods are different because of a difference in their
method names, or their method parameters, or both. To the compiler, overloaded
methods are just different methods.
However, method overloading does hold importance for the developer and the
users of a class because it can either make the code more intuitive to use or make
it totally confusing. For example, you have been using
System.out.println methods to print out all sorts of values to the
console. You are able to use the same method named println for printing an
int as well as a String only because there are multiple println methods
that take different parameter types. Take a look at the JavaDoc API description
of java.io.PrintStream class to see how many println methods it
has. Wouldn't you be frustrated if you had to use printlnByte to print a
byte , printlnShort to print a short , printlnString to print a
String and so on? In this case, method overloading has certainly made your
life easier.
Now, consider the following code:

static void doSomething(Integer i, short s){
System.out.println("1");
}
static void doSomething(int in, Short s){

System.out.println("1");
}

int b = 10;
short x = 20;
doSomething(b, x);
}
}
Any guess on what the above code prints? Don't worry, even experienced
programmers will scratch their heads while figuring this one out. The answer is
that it will not compile. But not because of the presence of two doSomething
methods. The methods are fine. They have different signatures. The problem is
with the call to doSomething(b, x) . The compiler is not able to determine
which one of the two doSomething methods to use because both of them are
equally applicable.
You will not get questions this hard in the exam. I showed you the above code
only to illustrate how overloading can be misused to create horrible code.
10.2.3 Method selection ☝
The code that I showed you in the previous section illustrates a problem with
overloaded methods. When their parameters lists are not too different, it is very
difficult to figure out which of the overloaded methods will be picked up for a
method call. If the parameters are too similar, it may even become an impossible
task.
Java specifies precise rules that are used by the compiler to disambiguate such
method calls. While you will not get questions that are too complicated in the
exam, you still need to know a few basic rules to figure out the simple cases. So,
here they are -
1. The first rule is that having overloaded methods does not cause a
compilation error by itself. In other words, as long as their method
signatures are different, the compiler doesn't care whether they are too
similar or not. Compilation error occurs only if the compiler is not able to
successfully disambiguate a particular method call.
2. Exact match - If the compiler finds a method whose parameter list is an

exact match to the argument list of the method call, then it selects that
method. For example, consider the following two methods:
void processData(Object obj){ }

void processData(String str){ }
and the method call processData("hello"); . Since String is an

Object , the String argument matches the parameter list of both the
methods and so both the methods are capable of accepting the method call.
However, String is an exact match to the argument and so the compiler
will select the second method.
This rule applies to primitives as well. Thus, out of the following two
methods, the first method is selected when you call processData(10);
because 10 is an int, which matches exactly to the parameter type of the
first method even though the long version of the method is also perfectly
capable of accepting the value.
void processData(int value){ }

void processData(long value){ }
3. Most specific method - If more than one method is capable of accepting a

method call and none of them is an exact match, the one that is "most
specific" is chosen by the compiler. For example, consider the following
two methods:
void processData(Object obj){ }

void processData(CharSequence str){ }
and the method call processData("hello"); . Remember that

String extends CharSequence and CharSequence extends
Object . Thus, a String is a CharSequence and a String is also an
Object . So here, neither of the methods has a parameter list that is an
exact match to the type of the argument but both the methods are capable of
accepting a String . However, between Object and ChareSequence
, CharSequence is more specific and so the compiler will select the
second method.
It is important to understand what "more specific" means. It really just

means closer or more similar to the type of the parameter being passed. A
String is closer to a CharSequence than it is to an Object .
Technically, a subclass (or a subtype) is always more specific than a super
class (or supertype).
Since primitives are not classes, there is no subclass/superclass kind of

relation between them as such but Java does define the subtype relation for
them explicitly, which is as follows:
double > float > long > int > char

and
int > short > byte
What it means is, float is a subtype of double , long is a subtype of

float , int is a subtype of long , and char is subtype of int . And
also, short is a subtype of int and byte is a subtype of short .
Based on the above, you can easily determine which of the following two
methods will be picked if you call processData((byte) 10);
void processData(int value){ }

void processData(short value){ }
The short version will be picked because short is a subtype of int and
is therefore, more specific than an int .
Here is another interesting situation that can be explained easily based on

the above subtype hierarchy of primitives. What will the following code
print when compiled and run?
public static void m(char ch){
System.out.println("in char");
}
byte b = 10;
m(b);
}
}
The above code will cause a compilation error saying, "error:

incompatible types: possible lossy conversion from
byte to char" . Observe that char is not a supertype of byte .
Therefore, the method m(char ch) is not applicable for the method call
m(b) . This makes sense because even though char is a larger data type
than byte , char cannot store negative values, while byte can.
4. Consider widening before autoboxing - Since autoboxing only came into

existence when Java 5 was released, it is necessary to give higher priority to
the primitive versions if the argument can be widened to the method
parameter type so that existing code will keep working as before.
Therefore, out of the following two methods, the short version will be
picked instead of the Byte version if you call processData( (byte)
b); even though the Byte version is an exact match with byte after
autoboxing.
void processData(short value){ }
void processData(Byte value){ }
5. Consider autoboxing before varargs - This rule mandates that if an

argument can be autoboxed into a method parameter type then that method
be considered even if a method with varargs of the same type is available.
This explains why if you call processData(10); then the Integer
version (and not the int... version) is picked out of the following two.
void processData(int... values){ }
void processData(Integer value){ }
Let us now try to figure out what happens if you call processData((byte)
10); with the same two methods available.
void processData(int... values){ }
void processData(Integer value){ }
Just apply the rules one by one:
Applying rule 1 - Requires no application.
Applying rule 2 - There is no method available whose parameter type

matches exactly to byte because byte doesn't match exactly to int...
i.e.int[] or Integer .

matches to byte after widening a byte to any of the types a byte can be
widened to (i.e. short, int, long etc.). Note that a byte cannot be
widened to int[] .

matches to byte after autoboxing it to Byte . Remember that a primitive
type can only be autoboxed to the same wrapper type. So a byte can only
be autoboxed to Byte and not to anything else such as Short or
Integer .
Applying rule 5 - A byte can be widened to int and an int can be

accepted by a method that takes int... Therefore, the varargs version of
the method will be invoked
Let me show you another example. Let's determine what will happen if we call
processData(10); with the following two methods available:
void processData(Long value){ }
void processData(Long... values){ }
Observe that 10 is an int . Do we have a processData method that takes an

int ? No. So Rule 2 about exact match doesn't apply. int is not a subtype of
Long or Long[] so Rule 3 doesn't apply either. Let's see if Rule 4 is of any
help. 10 can be widened to a long , float , or double but we don't have any
method that takes any of these types. Finally, as per Rule 5, 10 can be boxed into
an Integer but that won't work either because there is no method that takes an
Integer . Remember that processData(Long ) cannot accept an
Integer because Integer is not a subtype of Long .
Well, we don't have anymore options left to try. This means the compiler can't tie
this call to any method and will therefore, raise an error saying, "Error: no
suitable method found for processData(int)" .
Method selection is not a trivial topic. The Java

language specification spends a considerable number of
pages defining the rules of method selection. Since it is
easy to get bogged down with all those rules, I have
tried to simplify them so they are easy to remember.
You will be able to answer the questions in the exam
and will also be able to figure out what a piece of
code does in real life using the four points I have
mentioned above.
Just be aware that there are several situations that

cannot be explained by these points. Most of them
involve Generics, a topic that is not on the exam. You
should go through relevant sections of JLS if you are
interested in exploring this topic further.
You will most likely get only a single question on this topic. If you
forget the above rules and find yourself taking too much time in
figuring out the answer, I suggest you leave that question and move
ahead. There is no point in wasting too much time on a question that
is very easy to answer incorrectly.
10.3 Passing object references and primitive values
into methods
10.3.1 Passing arguments to methods ☝
This section requires that you have a clear understanding of the difference
between an object and a reference. I suggest you go through the "Object and
Reference" section in "Kickstarter for Beginners" chapter and the "Difference
between reference variables and primitive variables" section in the "Working
with Java data types" chapter to refresh your memory before proceeding with
this topic.
Pass by value ☝
To understand how Java passes arguments to methods, there is just one rule that
you need to remember - Java uses pass-by-value semantics to pass arguments to
methods. There are no exceptions. I mentioned this rule right at the beginning
because if you keep this rule in your mind, this topic will feel like a piece of
cake to you. You will never get confused with any code that the exam throws at
you.
Let me start with the following simple code:

int a = doubleIt(100);
System.out.println(a);//prints 200
public static int doubleIt(int x){

return 2*x;
}
In the above code, the main method invoke the doubleIt method and passes
the value 100 to the method. Just before starting the execution of
the doubleIt method, the JVM initializes the parameter x with the value that
was passed, i.e., 100 . Thus, x now contains the value 100 . The return
statement returns the value generated by the statement 2*x back to the caller.
The JVM assigns this value to the variable a . Thus, a now contains 200 . This
is what is printed out.
Quite simple so far, right? Let me modify the code a bit now:

int a = 100;
int b = doubleIt(a);
System.out.println(a+", "+b);
}
static int doubleIt(int x){

return 2*x;
}
The only thing that I have changed is that instead of passing the value 100
directly to the doubleIt method, I am passing the variable a . The JVM
notices that the method call is using a variable as an argument. It takes the value
contained in this variable (which is 100 ) and passes it to the method. Rest is
exactly the same as before. Just before starting the execution of the doubleIt
method, the JVM initializes the parameter x with the value that was passed, i.e.,
100 . Thus, x now contains the value 100 . The return statement returns the
value generated by the statement 2*x back to the caller. The JVM assigns this
value to the variable b . Thus, b now contains 200 . Finally, 100, 200 is
printed.
Observe that doubleIt has no knowledge of the variable a . It only gets the
value contained in the variable a , i.e., 100 , which is assigned to x .
Are you with me so far? Alright, here comes the twist. Take a look at the
following code:

int a = 100;
int b = doubleIt(a);
System.out.println(a+", "+b);
}

static int doubleIt(int x){
x = 200;
return 2*x;
}
}
If you have understood the logic I explained earlier, you should be able to figure
out what the above code prints. Let's follow the same process for analyzing what
is happening here. The JVM notices that the method call is using a variable as an
argument. It takes the value contained in this variable (which is 100 ) and passes
it tp the doubleIt method. Just before starting the execution of
the doubleIt method, the JVM initializes the parameter x with the value that
was passed, i.e., 100 . Thus, x now contains the value 100 . Next, x is assigned
a new value 200 by the statement x = 200; In other words, the x is
overwritten with a new value 200 . The point to understand here is that this
assignment has no effect of the variable a that was used in the calling method
because doubleIt has absolutely no idea where the original value of 100 that
it was passed as an argument came from. It merely gets the value 100 as an
argument. In other words, the JVM passed only the value 100 to the method and
not the variable a .
Next, the return statement returns the value generated by the statement 2*x back
to the caller, which is 400 . The JVM assigns this value to the variable b . Thus,
b now contains 400 . At this time, you should realize that nothing was done to
change the value of the variable a at all. It still contains the same value as
before, i.e., 100 . Therefore, the print statement prints 100, 400 .
The above example also shows that it is not possible for a method to change the
value of the variable that was passed as an argument by the caller because that
variable is never sent to the method. Only its value is sent. That is why the term
"pass by value" is used to describe parameter passing in Java.
10.3.2 Passing objects to methods ☝
In the previous lesson, you saw how passing a primitive variable to a method
works. Passing a reference variable works exactly the same way. Check out the
following code:
class Data{
int value = 100;
}

Data d = new Data();
modifyData(d);
System.out.println(d.value);//prints 200
public static void modifyData(Data x){

x.value = 2*x.value;
}
}
In the above code, the main method creates a Data object and saves its
reference in the variable d . It then calls the modifyData method. During
execution, the JVM notices that the call to modifyData method uses a
reference variable. Recall our discussion on reference variables in "Working with
data types" chapter that a reference variable doesn't actually hold an object. It
just holds the address of the memory location where the object resides. While
invoking the modifyData method, the JVM copies the value stored in the
variable d into the method parameter x . So, for example, if the Data object is
stored at the memory address 10001 , the variable d contains 10001 and the
JVM copies this value into the variable x . Thus, the variable x also now
contains the same address as d and thus, in a manner of saying, variable x also
starts pointing to the same object as the one pointed to by the variable d . This
situation is illustrated in figure 1 below.
The point to note here is that even though it looks as if we passed the Data
object created in main to modifyData , all we actually passed was the value
stored in variable d (which was nothing but the address of the Data object) to
modifyData . The Data object itself remained exactly at the same memory
location where it was. We didn't move it, pass it, or copy it, to anywhere.
The modifyData method uses the reference variable x to modify the value
field of Data object pointed to by x to 200 . When the control goes back to the
main method, it prints d.value . , which is now 200 .
Let us make a small modification to the modifyData method of the above

code:

x = new Data();
}
All we have done is inserted the statement x = new Data(); . Can you tell
what the main method will print now?
When the JVM invokes modifyData method, it makes x point to the same
Data object as the one pointed to by variable d . This part is the same as before.
But the first statement of the updated method creates a new Data object and
makes x point to this new object. For example, if this new Data object resides
at address 20001, then x now contains 20001 (instead of 10001). The next
statement updates the value field using the variable x . Since x now points to the
new Data object, this update statement updates the value field of the new Data
object instead of the old Data object. When the control goes back to the main
method, it prints d.value . This prints 100 because we never modified the
variable d and so, d still points to the original Data object. This is illustrated in
figure 2 below.

In the two examples above, we passed a variable (the local variable d ) as an
argument to the modifyData method. Now, take a look at the following code:
class Data{
int value = 100;
}

modifyData(new Data());

}
}
Observe the statement modifyData(new Data()); . We are not passing

any variable to the modifyData method here. So, are we really passing the
Data object directly to the method? No, although we do not have any explicit
variable to store the address of the Data object, the compiler creates a
temporary reference variable implicitly. It creates the Data object in the heap
space and assigns it to this temporary variable. While making the method call, it
is the value of this variable that is passed to the modifyData method. Of
course, since we have not saved a reference to this Data object in the main
method, we won't be able to refer to this object after the modifyData method
returns.
This technique is used all the time while printing messages to the console using
the print/println method. For example, when you do
System.out.println("hello world"); the compiler passes the value
of the temporary variable that points to the String object containing "hello
world" to the println method.
10.3.3 Returning a value from a method ☝
Just like parameters, returning a value from a method also uses "pass by value"
semantics.
In the case of primitives, a return statement can return either the value directly
(e.g. return 100; return true; return 'a'; etc.) or the value of a
variable (e.g. return i; where i is an int variable).
In the case of objects, a return statement can either return the value of an explicit
variable (e.g. return str; where str is a String variable) or the value of
an implicit temporary variable that references the object (e.g. return
"hello"; return new Student(); etc.). In both cases, it is really the
address where the object is stored that is returned.
What the caller does with the return value is irrelevant to the method that returns
a value. The caller can either use the return value further in the code or ignore
the return value altogether.
10.4 Create and overload constructors
10.4.1 Creating instance initializers ☝
When you ask the JVM to create a new instance of a class, the JVM does four
things:
1. First, it checks whether that class has been initialized or not. If not, the
JVM loads and initializes the class first.
2. Second, it allocates the memory required to hold the instance variables of
the class in the heap space.
3. Third, it initializes these instance variables to their default values (i.e.
numeric and char variables to zero , boolean variables to false , and
reference variables to null ).
4. And finally, the JVM gives that instance an opportunity to set the values of
the instance variables as per the business logic of that class by executing
code written in special sections of that class. These special sections are:
instance initializers and constructors .
It is only after these four steps are complete that the instance is considered
"ready to use". Remember the use of the new operator to create instances of a
class? The JVM performs the all of the four activities mentioned above and only
then returns the reference of the newly created and initialized object to your
code.
Out of the four activities listed above, you have already seen the details of the
first one in the previous section. The second two activities are performed
transparently by the JVM. They don't require the programmer to do anything.
The fourth activity, which is the subject of this section, depends on the
programmer because it involves the code written by the programmer.
Let us take a look at instance initializers first because these are the ones that are
executed by the JVM first.
Creating instance initializers ☝
Instance initializers are blocks of code written directly within the scope of a
class. Here is an example:
class TestClass{
{
System.out.println("In instance initializer");
}
Observe that there is no method declaration or anything but just a line of code
nested inside the opening and closing curly brackets. Code inside an instance
initializer block is regular code. There is no limitation on the number of
statements or the kind of statements that an instance initialize can have. You may
have any number of such instance initializer blocks in a class. The JVM executes
them in the order that they appear in the class. The following code, for example,
prints Hello World! using two instance initializers in different places in a
class:
class TestClass{
//first instance initializer
{
System.out.print("Hello ");
}

new TestClass();
}
//second instance initializer
{
System.out.print("World!");
}
Observe that the main method does nothing except create an instance of
TestClass. As a result of this creation, the JVM executes each of the two instance
initializers and then returns the newly created instance to the main method. Of
course, the main method does not assign the reference of the newly created
instance to any variable but that is okay.
An instance initializer must appear directly within the scope of a class. The
block of code in the main method of the following code, therefore, is not an
instance initializer. It is just regular method body code. It is valid though.
{
System.out.print("Hello ");
}

//this following is not an instance initializer
{
System.out.print("World!");
}
new TestClass();
}
}
Can you tell what it will print? That's right. It will print "World!Hello "
(without the quotes, of course). While executing the main method, the JVM
encounters the print statement that prints "World!" so that is printed. Next, it
encounters an instruction to create an instance of TestClass. Note that this is the
first instance of TestClass that it is going to create. While finishing up the
creation of this instance, it executes the instance initializer that prints "Hello
" . Thus, the net output on the console is "World!Hello " .
Accessing members from instance initializers ☝
Instance initializers have access to all of the members of a class. This includes
static as well as instance fields and methods. Just like the instance methods,
instance initializers have access to the implicit variable "this" and "super" (I
will cover "super" later).
Forward referencing ☝
While accessing instance fields, the order of their appearance in relation to the
initializer block is important. Although an instance variable declared anywhere
in the class is visible to all instance initializers, an instance initializer is not
allowed to use the value of a variable if that variable is declared below the
initializer. It can assign a value to such a variable though. For example, consider
the following code:
System.out.print(i); // 1
- invalid forward reference
i = 20; //2 - v
alid forward reference

}
int i = 10;

new TestClass();
}
The line marked //1 in the above code will not compile because it is attempting
to use the value of the variable i , which is declared below the initializer, but the
line marked //2 will compile fine because it is assigning a value to i .
This rule is applicable only for the variables and not for methods. Thus, the
following code will compile fine:
{
printI(); //valid forward reference
void printI(){
System.out.print(i);
}
int i = 10;

new TestClass();
}
Although the code will compile and run fine, you may be surprised by what it
prints. JVM executes instance initializer blocks and instance variable initializer
statements in the order they appear. Therefore, in the above code, the JVM
executes the instance initializer block before it executes the initialization of the
variable in the statement int i = 10;
(Note that the JVM has already allocated space for the variable i and has
already provided a default value of 0 to i as a part of performing steps 1 and 2
that I described at the beginning of this topic.)
While executing the instance initializer, the JVM invokes the printI method.
The printI method, in turn, prints the value of i , which is 0 at this time.
10.4.2 Creating constructors ☝
Constructor of a class ☝
A constructor of a class looks very much like a method of a class but there are
two things that make a method a constructor:
1. Name - The name of a constructor is always exactly the same as the name
of the class.
2. Return type - A constructor does not have a return type. It cannot even say
that it returns void.
The following is an example of a class with a constructor.
class TestClass{
int someValue;
String someStr;
TestClass(int x)
//No return type specified
{
this.someValue = x; //initializing someValue
//return; //legal but not required.
}
}
Observe that the name of the constructor is the same as that of the class and that
there is no return type in the declaration of the constructor. Also observe that
there is no return statement because a constructor cannot return anything, not
even void. However, it is permissible to write an empty return statement as
shown in the code above.
It is interesting to note that a class can have a method with the same name as that
of the class. For example, consider the following class:
class TestClass{
int someValue;
String someStr;
void
TestClass(int x) //<-- observe the return type void
{
this.someValue = x;
}
}
In the above code, void TestClass(int x) is not a constructor but it is a

valid method nevertheless and so the code will compile fine.
Besides the above two rules there are several other rules associated with
constructors. Some of them are related to inheritance and exception handling and
I will cover them later. Here, I will talk about rules that are applicable to
constructors in general.
The default constructor ☝
This is one of the most confusing rules about constructors and it makes for a
very good trap in the exam. If I tell you that every class must have a constructor,
you might have trouble believing it because so far, I have shown you so many
classes that had no constructor at all! Well, here is the deal - it is true that every
class must have at least one constructor but the thing is that the programmer
doesn't necessarily have to provide one. If the programmer doesn't provide any
constructor, then the compiler will add a constructor to the class on its own. This
constructor, that is, the one provided by the compiler, is called the "default"
constructor. This default constructor takes no argument and has no code in its
body. In other words, it does absolutely nothing. For example, if I write and
compile this class, class Account { } , the compiler will automatically
add a constructor to this class which looks like this:
class Account{
Account(){ } //default constructor added by the co
mpiler
This sounds simple but the cause of confusion is the fact that if you write any
constructor in a class yourself, the compiler will not provide the default
constructor at all. So, for example, consider the following class:
class Account{
int id;
Account(int id){
this.id = id;
}
Account a = new Account();
}
}
I have provided a constructor explicitly in the above class. Can you can guess
what will happen if I try to compile this class? It will not compile. The compiler
will complain that Account class does not have a constructor that takes no
arguments. What happened to the default constructor, you ask? Well, since this
class provides a constructor explicitly, the compiler feels no need to add one on
its own.
The discussion on default constructors also gives me an opportunity to talk about

one misconception that I have often heard, which is that constructors must
initialize all instance members of the class. This is not true. A constructor is
provided by you, the programmer, and you can decide which instance members
you want to initialize. For example, the constructor I showed earlier for
TestClass doesn't touch the someStr variable. The default constructor
provided by the compiler also doesn't assign any values to the instance members.
Remember that the JVM always provides default values to static and instance
members anyway.
You will most certainly get a question that tests your knowledge
about the default constructor. Watch out for code that assumes the
existence of the default constructor while the class provides a
constructor explicitly.
Difference between default and user defined constructors ☝
There are only two things that you need to remember about a default constructor:
1. When is it provided - It is provided by the compiler only when a class

does not define any constructor explicitly.
2. What does it look like - The default constructor is the simplest constructor
that you will ever see. It does not take any argument, does not have any
throws clause, and does not contain any code.
But one peculiar thing about the default constructor is its accessibility. The
accessibility of the default constructor is the same as that of the class. That
means, if the class is public, the default constructor will also be public and
if the class has default accessibility, the default constructor will also have
default accessibility. If you have no idea about accessibility, don't worry, I
will talk about it in the next topic.
The default constructor is also sometimes called the "default no-args"

constructor because it does not take any arguments. But this is technically
imprecise because the default constructor is always a no-args constructor. There
is nothing like a default constructor with args!
You can, on the other hand, write a no-args constructor explicitly in a class with
a throws clause and with a different accessibility. For example, it is quite
common to have a class with private no-args constructor if you don't want
anyone to create instances of that class. I will talk more about it in the next topic.
Benefit of constructors ☝
Anything that you can do in a method, you can do in a constructor and vice-
versa. There is no limitation on what a method or a constructor can do. Why not
just have a method and name it as init() or something? Indeed, you will
encounter Java frameworks such as servlets that do exactly that. Then why do
you need a constructor? Well, a detailed discussion of the pros and cons of
constructors is beyond the scope of this book, but I will give you a couple of
pointers.
As I mentioned earlier, an instance is not considered "ready to use" until its

constructor finishes execution. Thus, a constructor helps you make sure that an
instance of your class will be initialized as per your needs before anyone can use
it. Here, "use" essentially means other code accessing that instance through its
fields or methods.
It is possible to do the same initialization in a method instead of a constructor

but in that case the users of your class would have to remember to invoke that
method explicitly after creating an instance. What if a user of your class fails to
invoke that special method after creating an instance of your class? The instance
may be in a logically inconsistent state and may produce incorrect results when
the user invokes other methods later on that instance. Therefore, a constructor is
the right place to perform all initialization activities of an object. It is a place
where you make sure that the instance is ready with all that it needs to perform
the activities it is supposed to perform in other methods.
Constructors also provide thread safety because no thread can access the object
until the constructor is finished. This protection is guaranteed by the JVM to the
constructors and is not always available to methods.
10.4.3 Overloading constructors ☝
A class can have any number of constructors as long as they have different
signatures. Since the name of a constructor is always the same as that of the
class, the only way you can have multiple constructors is if their parameter type
list is different.
Just like methods with same names, if a class has more than one constructors,
then this is called "constructor overloading" because the constructors are
different only in their list of parameter types.
There are a couple of differences between method overloading and constructor

overloading though. Recall that in case of method overloading, a method can call
another method with the same name just like it calls any other method, i.e., by
using the method name and passing the arguments in parenthesis. In case of
constructor overloading, when a constructor calls another constructor of the
same class, it is called "constructor chaining" and it works a bit differently.
Let me show you how.
Constructor chaining ☝
A constructor can invoke another constructor using the keyword this and the
arguments in parenthesis. Here is an example:
class Account{
int id;
String name;
Account(String name){
this(111, name); //invoking another constructor
here
System.out.println("returned from two args cons

tructor");
}
Account(int id, String name){
this.id = id;
this.name = name;
}
Account a = new Account("amy");
}
}
Observe that instead of using the name of the constructor, the code uses 'this ' .
That is, instead of calling Account(111, name); , the code calls
this(111, name); to invoke the other constructor. It is in fact against the
rules to try to invoke another constructor using the constructor name and it will
result in a compilation error.
Invoking another constructor from a constructor is a common technique that is

used to initialize an instance with different number of arguments. As shown in
the above example, the constructor with one argument calls the constructor with
two arguments. It passes one user supplied value and one default value to the
second constructor. This helps in keeping all initialization logic in one
constructor, while allowing the user of the class to create instances with different
parameters.
The only restriction on the call to the other constructor is that it must be the first
line of code in a constructor. This implies that a constructor can invoke another
constructor only once at the most. Thus, the following three code snippets for
constructors of Account class will not compile:
System.out.println("calling two args constructor")
;
this(111, name); //call to another constructor mus
t be the first line
Account(){
this(111);
this(111, "amy"); //call to another constructor mu
st be the first line
Account(111, name); //incorrect way to call to anot
her constructor but if the Account class had a method
named Account, this would be a valid call to that meth
od.
Invoking a constructor ☝
It is not possible to invoke the constructor of a class directly. It is invoked only

as a result of creation of a new instance using the "new" keyword. The only
exception to this rule is when a constructor invokes another constructor through
the use of "this" (and "super", which I will talk about in another chapter) as
explained above. In other words, if you are given a reference to an object, you
cannot use that reference to invoke the constructor on that object.
Now, consider the following program. Can you tell which line out of the four
lines marked LINE A, LINE B, LINE C, and LINE D should be uncommented
so that it will print 111, dummy ?
class Account{
int id;
String name;
public Account(){
id = 111;
name = "dummy";
}
public void reset(){

//this(); //<-- LINE A
//Account(); //<-- LINE B
//this = new Account(); //<-- LINE C
//new Account(); //<-- LINE D

Account a = new Account();
a.id = 2;
a.name = "amy";
a.reset();
System.out.println(a.id+", "+a.name);
}
The answer is none of these. First three are invalid attempts to invoke the
constructor - this() can only be used within a constructor, Account() is
interpreted as method call to a method named Account, but such a method
doesn't exist in the given code, and this = new Account() attempts to
change "this ", which is a final variable, to point to another object. Thus, none
of these three statements will compile.
new Account() is a valid statement but it creates an entirely new Account
object. The instance variables of this new Account object are indeed set to 111
and "dummy" , but this doesn't change the values of the current Account object.
10.4.4 Instance initializers vs constructors ☝
Instance initializers and constructors are meant for the same purpose, i.e., to give
the programmer an opportunity to initialize the instance. But there are a few
fundamental differences between them:
1. While creating an instance of a class, the JVM executes all of the instance
initializers of that class one by one but it executes only one of its
constructors (depending on the arguments passed). Of course, as you saw
earlier, it is possible for a constructor to invoke another constructor using
"this". No such explicit chaining of instance initializers is possible.
2. The restriction on the usage of variables declared below an instance

initializer is not applicable to constructors. As explained earlier, an instance
initializer can forward reference a variable but only as long as it is
assigning a value to that variable and not trying to make use of the value of
that variable.
3. The third difference is related to exceptions, a topic that I have not yet
discussed, so, don't worry if you don't get this completely. The compiler
expects an instance initializer to execute without throwing an exception. If
the compiler can figure out that an instance initializer will always end up
with an exception, it will refuse to compile the code. Here is an example:
class TestClass{
{
throw new RuntimeException(); //this line wil
l not compile
int i = 10;

new TestClass();
}
TestClass(){
throw new RuntimeException(); //this line is f
ine
}
}
The above code will fail to compile with an error message saying, "error:
initializer must be able to complete normally". The compiler notices that
there is no way this initializer will ever be able finish successfully and
therefore, there is no way an instance of this class can ever be created.
This restriction is not applicable to constructors. That is why the constructor

in the above code is fine.
Other than the above three differences, instance initializers and constructors
work the same way. So, when should you use instance initializers? Ideally,
never. Instance initializers make the code difficult to follow and are best
avoided. A well designed class should not need to use instance initializers.
However, if you have a class with multiple constructors and if you want to add
some common initialization logic to each of its constructors, you can put that
logic in an instance initializer. This will save you from repeating the same code
in the constructors. In that sense, instance initializers are a powerful hack :)
In the exam you may see a question on the basics of instance

initializers but we haven't seen anyone getting a question that
requires the details discussed in this section. I have talked about
them only because it is important for interviews and practical
programming.
10.4.5 final variables revisited ☝
Now that you know about initializers and constructors, let me go back to final
variables for a moment.
Since the value of a final variable cannot be changed, Java doesn't assign it a
default value and instead, forces the programmer to assign it a value explicitly
before it can be used by other classes. This idea of explicit initialization drives
the following rules about their initialization:
1. Explicit initialization of a static final variable - A static variable can be

used by other classes only after the class is loaded and made ready to use.
This means that the variable must be initialized by the time class loading is
complete. Thus, you can either assign a value to a final static variable at the
time of declaration or in any one of the static initializers (aka class
initializers ) as shown below:
class TestClass{
static final boolean DEBUG = true; //initializ
ing at the time of declaration
static final int value; //not initializing va

lue here

static
{
value = 10;

}

System.out.println(value);
}
}
Since a class can have multiple static initializers and all of them are
executed when a class is loaded, you can initialize a static final variable in
any one of them but not in more than one of them because the first static
initializer would have already set its value and the second one would only
be changing the existing value, which is not permitted. Thus, the following
code will not compile:
class TestClass{
final static int value;

static {
value = 10;
}
static {
value = 20; //won't compile because value i
s already initialized
}
}
2. Explicit initialization of instance final variables : Since an instance

variable can be used by other classes only after an object of this class is
created, it follows that a instance final variable can be initialized at the time
of declaration, in any of the instance initializers, or in all of the
constructors. For example:
class TestClass{
final int value; //not initializing here

{
value = 10; //initializing in an instance
initializer
}
}
Again, as with static initializers, a class can have multiple instance

initializers and since all of them are executed while creating an instance, an
instance final variable can be initialized in only one of the instance
initializers. If the variable is not initialized in any of the instance
initializers, it must be initialized in each of the constructors. For example:
class TestClass{
final int value; //not initializing here

{
//not initializing value here either
TestClass(){
value = 10;
}
TestClass(int x){
value = x;
}
TestClass(int a, int b){ //this constructor wi
ll not compile because it leaves value uninitializ
ed
Observe that initialization in constructors works differently from

initialization in instance initializers. The difference stems from the fact that
all instance initializers are executed but only one constructor is executed
while creation of an instance of that class.

3. Explicit initialization of local variables (aka method or automatic
variables) - A method variable can only be used within that method and
therefore, it may be initialized at anytime before it is accessed. It may even
be left uninitialized if it is not used anywhere in the method. For example,
the following code compiles and runs fine:
class TestClass {
final int x; //not initializing x anywhere
final int y;
y = 10; //initializing y just before it is a
ccessed
System.out.println(y);
}
}
10.5 Apply the static keyword to methods and fields
10.5.1 Apply the static keyword to methods and fields ☝
I have already discussed the meaning of the word static and what it implies when
applied to a method or a field in the "Kickstarter for Beginners" chapters.
In this section, I will dig a little deeper and explain the nuances of static from the
perspective of the exam. Let us start with the syntax.
You can declare any member of a class (i.e. a method, a field, or any nested type
definition) as static using the keyword static. In the case of a method or a field,
this keyword must appear before the return type of the method or the type of the
field respectively. For example:
class Foo{
static int bar;
static String biz;
private static void baz(){ } //order of modifiers do
esn't matter
static public final void boz(){ } //order of modifie

rs doesn't matter
In case of nested type definitions, the static keyword must appear immediately
before the type. For example:
class Boo{
static class NestedClass{
}
static interface NestedInterface{

}
static enum NestedEnum{
}
}
A top level class, interface, or an enum cannot be declared static. It wouldn't

make sense either because static is a property of the relationship between the
owner and the owned types. For example, class Boo owns NestedClass
statically. Class Boo cannot be static because it is not owned by any other type.
Local members, i.e., members defined within a method, cannot be static either.
Thus, the following code will fail to compile:
class TestClass {
static int x; //can't be static
static class Y{ //can't be static
}
}
}
Does that mean the variable x and the class Y are an instance variable and an
instance class respectively? No, such members are not members of the class at
all so, the concept of static/instance does not apply to them. Such members are
local to the method and cease to exist as soon as the method execution is
complete. They cannot be referenced from outside the method.
10.5.2 Accessing static members ☝
A static member exists as a member of the owning class and not as a member of
an instance of the owning class. In other words, a static member does not require
an instance of the owning class to exist. A static member, therefore, can be
accessed by specifying the name of the owning class. For example, the static
variables of Foo and Boo can be accessed from another class like this:
class TestClass{
System.out.println(Foo.bar); //prints 0 (why?)
System.out.println(Foo.biz); //prints null (why?)
Boo.NestedClass nc = new Boo.NestedClass();

}
}
Accessing a static member using the name of the owning class is the standard
and recommended way of accessing static members. However, Java allows a
static member to be accessed through a variable as well. For example, the static
variable bar of class Foo can also be accessed like this:
class TestClass{
Foo f = null;
System.out.println(f.bar); //prints Foo's bar
}
}
Observe that f is null . But the compiler doesn't care about f being null
because it notices that bar is a static variable and an instance of Foo is not
needed to access bar . The compiler knows what bar in the statement f.bar
implies and effectively translates it to Foo.bar .
This simple example highlights an important aspect of accessing static members:

access to static members is decided by the compiler at compile time by checking
the declared type of the variable. It is not decided by the JVM at run time. The
compiler knows that the type of the variable f is Foo and that bar is a static
member of Foo , that is why the compiler binds the call to f.bar to Foo 's
bar . The compiler doesn't care what f may point to at run time. Now, armed
with this knowledge, can you tell what the following code will do?
class TestClass{
Foo f = null;
f.boz();
}
}
That's right. It will compile and run fine (of course, without any output). But the
interesting thing is that there will be no NullPointerException . Since the
type of the reference variable f is Foo and boz() is a static method of Foo ,
the compiler binds the call to f.boz() to Foo 's boz() .
This is called static binding or compile time binding because the compiler
doesn't leave the decision of binding a call to the JVM. This concept will play an
important role when we discuss inheritance and polymorphism later.
10.5.3 Accessing static members from the same class ☝

A static member of a class can also be accessed directly from static as well as
instance members of the same class without the need to use the class name. For
example, the following class makes use of a static variable to count the number
of instances of that class:
class InstanceCounter{
static int count;
InstanceCounter(){
//directly accessing count from a constructor
count++;
}

static void printCount(){
//directly accessing count from a static method
System.out.println(count);
}
void reduceCount(){
//directly accessing count from an instance meth
od
count--;
}
}
The following class shows various ways in which you can access static members
of a class from another class:
class TestClass {
InstanceCounter ic = new InstanceCounter();
ic.printCount();//accessing static method through
a reference to an instance
new InstanceCounter().printCount(); //accessing s

tatic method through an implicit reference to an insta
nce
System.out.println(InstanceCounter.printCount()+"
"+InstanceCounter.count); //accessing static members
using the class name
}
You should be able to tell what the above code will print.
10.5.4 Importing static fields ☝
In the "Java Basics" chapter, I discussed import statements. I showed you how
to import classes as well as static members of a class. If you want to access static
members of a class several times, the import static statement provides an easy
way to save some keystrokes.
It is interesting to know that the import static statement was added in Java
7 only to avoid a kludgy work around for accessing global constants. Before
Java 7, if a developer needed to define global constants, they would define the
constants in an interface like this:
public interface Constants{

public static final double INTEREST_RATE = 0.15;
public static final boolean COMPOUNDING = true;
public static final int PERIOD = 1;
}
To use these constants in any class, they would have that class implement the
above interface like this:
public class InterestCalculator implements Constants
{
public static double compute(double principle, doub
le time){
if(COMPOUNDING){
//compute and return compound interest
}else{
//compute and return simple interest
}
}
}
The benefit of having the class implement the Constants interface is that it
saves the developer from typing the class name of the class in which the constant
is defined, i.e., instead of typing Constants.COMPOUNDING , they can just
type COMPOUNDING . But many experts believe that this makes the code
difficult to understand. It is also an abuse of "implements" keyword because
there is no functionality that is being implemented here. The usage of interface is
also objectionable because we are not defining any behavior in the interface
either.
The import static statement avoids the need for this technique by making it
simple to import static members of a class. You can now define the constants in a
class and import those constants statically like this:
package accounting; //need to put this class in a pack

age because recall that it is not possible to import a
nything from the unnamed package
public class Constants{

public static final double INTEREST_RATE = 0.15;
public static final boolean COMPOUNDING = true;
public static final int PERIOD = 1;
}
import static accounting.Constants.*;

public class InterestCalculator {
public static double compute(double principle, doub
le time){
if(COMPOUNDING){
//compute and return compound interest
}else{
//compute and return simple interest
}
}
}
Technically, you can import static members of an interface also but from a
design perspective, it is better to use a class for defining constants and reserve
the use of interfaces for defining behavior in terms of methods.
10.5.5 Accessing instance members from a static method ☝
Since a static method belongs to a class and not to an object of that class, a static
method does not execute within the context of any instance of that class. On the
other hand, an instance method is always invoked on a specific instance of a
class and so, it executes within the context of the instance upon which it is
invoked. An instance method can access this instance using the implicit variable
"this" . Since there is no instance associated with a static method, the variable
this is not available in a static method.
The reason why the "this" variable is called an implicit variable becomes
important here. This variable is not declared explicitly anywhere and the
compiler assumes its existence in an instance method. Whenever the compiler
sees an instance member being accessed from within a method directly, the
compiler uses the this variable to access that member even when you don't
type it explicitly in your code. However, when a static method tries to access an
instance member, this is not available and so, the compiler complains that a
non-static variable cannot be referenced from a static context.
The following code illustrates this point:
class Book{
int name;
static void printName1(){
System.out.println(this.name); //will not compi

le
System.out.println(name);//same as above. will

not compile
}

void printName2(){
System.out.println(this.name); //this is fine.
System.out.println(name); //same as above. this

is fine.
}
}
In the above code, the compiler realizes that name is an instance variable and so,
tries to access it through this , i.e., this.name . But since this is not
available in printName1 , it generates an error. There is no issue with
printName2 because printName2 is an instance method and this is
available in an instance method.
Remember that you don't always need to explicitly use the variable this to
access instance members. You need to use it only if there is also a local variable
with the same name declared in the method and you want to refer to the instance
variable instead of the local one. Technically, this is used to "unshadow" an
instance variable if it is shadowed by a local variable of the same name.
A common misunderstanding amongst beginners is that a static method cannot

access instance fields of a class. This misunderstanding exists because they see
or hear this statement in many places. However, it is an incomplete statement.
The correct statement is that a static method cannot access an instance member
without specifying the instance whose member it wants to access. It will be clear
when you see the following code:
class Book{
int name;
static void printName(){

Book b1 = new Book();
Book b2 = new Book();
System.out.println(b1.name); //this will compil
e fine
}
}
In the above code, we are accessing the instance variable name from within a
static method. Notice that there are two instances of Book . Each instance has its
own copy of the variable name . But because we are using the reference b1 to
access name, the compiler knows exactly whose name we intend to access. It
knows that we want to access the name field of the Book instance pointed to by
the reference variable b1 . That is why there is no problem. This shows that an
instance field can indeed be accessed from a static method as long as we specify
exactly the instance whose field we want to access.
10.5.6 Class loading and static initializers ☝

Recall that executing a Java program or a Java class essentially means executing
the "main" method of that class. But before the JVM can pass the control to this
main method, it has to first find the class file, and "load" the class into its
memory. The loading of a class is done by a class loader. Don't worry, you don't
need to worry about class loaders and the whole class loading process for the
exam. But you do need to know about one thing that happens to the class after
the class is loaded and before the main method is invoked.
While executing code, whenever the JVM encounters the usage of a class for the
first time , it allocates and initializes space for the static fields of that class.
Since static fields belong to the class and not to the objects of the class, there is
only one copy of such fields. The JVM initializes these fields to their default
values (i.e. zero for numeric and char types and null for reference types) and
executes the "static initializers" defined in the class. Static initializers are
nothing but blocks of code marked with the keyword static as shown below:
class TestClass{
//the following is a static initializer
static {
System.out.println("In static block");
}

System.out.println("In main");
}
}
The above program generates the following output:
In static block
In main
As you can observe from the above output, the main method is executed after the
static block. A static block provides an opportunity for the developer to initialize
the static fields of the class and execute any other business logic that the
developer wishes to execute before the class is put to use by any other code. The
following class shows how a developer might make use of a static initializer:
class InterestCalulator{
static double RATE;
static {
System.out.println("In static block. RATE = "+RATE
);
RATE = 10;
}

public static double computeInterest(double principl
e, int yrs){
return RATE*principle*yrs/100;
}

double interest = computeInterest(100, 1);
System.out.println(interest);
interest = computeInterest(100, 2);
The InterestCalculator class shown above computes simple interest

based on the value of RATE , which is set in the static initializer. It produces the
following output:
In static block. RATE = 0.0

10.0
20.0
The important thing to observe in the above output is that the value of RATE is
printed as 0.0 . This is because the JVM has already initialized the RATE
variable to its default value of 0.0. It is that value that is bring printed. RATE is
being set to 10 after this print statement. The second thing that you should
notice is that RATE is printed only once. This is because a static block is
executed only once when a class is loaded and initialized. The JVM never
executes the static block again no matter how many times the class is used
afterwards.
There are only a few simple rules about static blocks:
1. A class can have any number of static blocks. They are executed in the
order that they appear in the class.
2. A static block can access all static variables and static methods of the class.
However, if the declaration of a static variable appears after the static block,
then you can only set the value of that variable in the static block. Here is
an example:
class TestClass{
static int a;
static{
System.out.println(a);//valid, a is declared
before the static block
System.out.println(b);//INVALID, cannot read

b's value because b is declared after the static b
lock
b = 10; //valid because b is being assigned a

value.
m(); //valid even though m is defined later
static void m(){

System.out.println(b);//valid, a method can d
o anything with a variable that is declared later
in the code
static int b;

}
//another static block
static{
System.out.printn(b);
}
}
3. If the class has a superclass, and if the superclass hasn't been initialized
already, the JVM will initialize the superclass first and then proceed with
the initialization of this class. The following example explains this process:
class Parent{
static{
System.out.println("Initializing Parent");
}
}
class ChildA extends Parent{
static{
System.out.println("Initializing ChildA");
}
}
class ChildB extends Parent{
static{
System.out.println("Initializing ChildB");
}
}

ChildA a = new ChildA(); //will cause class
Parent and then class ChildA to be initialized
ChildB b = new ChildB(); //will cause only

ChildB to be initialized
}
}
The above code generates the following output:
Initializing Parent
Initializing ChildA
Initializing ChildB
When the JVM encounters the statement ChildA a = new ChildA()

, it realizes that ChildA is being used for the first time here. So, before
creating an object of the ChildA class, it needs to initialize the ChildA
class. Since ChildA extends Parent , the JVM must first initialize the
Parent class. This is why Initializing Parent is printed before
Initializing ChildA . Once ChildA is loaded and initialized, the
JVM creates an object of ChildA and assigns its reference to variable a .
Similarly, when the JVM encounters the statement ChildB b = new
ChildB() , it tries to initialize the ChildB class. While
initializing ChildB , the JVM realizes that its super class, i.e., Parent ,
has already been initialized and so, it does not execute Parent 's
initialization logic again. Thus, only Initializing ChildB is printed
this time. You might now wonder why we stopped at the Parent class.
What happened to the Object class? After all, Object is the super class
of Parent ! Well, Object is indeed the super class of Parent . But it is
also the super class of TestClass . The JVM initialized the Object
class while it was initializing the TestClass class.
4. There is no way to access a static block or refer to a static block. That

means, you can't "invoke" or "call" a static block explicitly from anywhere.
It can only be invoked by the JVM and that too, only once.
Technically, there is one copy of static fields per class

per class loader, but a discussion on class loaders is
way out of scope of this exam. However, be aware that
technical interviewers love to touch upon class loaders
and are impressed if the candidate knows about class
loading and its impact on static fields. I suggest you go
through a few articles on this topic such as
https://blogs.oracle.com/sundararajan/understanding-java-
class-loading if you are interviewing for a job.
10.6 Exercise ☝
1. Create a method named add that can accept any number of int values and
returns the sum of those values.
2. Create another method named add in the same class that can accept any
number of int values but returns a String containing concatenation of
all those values. What can you do to resolve compilation error due to the
presence of the two methods with same signature? Invoke these methods
from the main method of this class.
3. Create a class named Student with a few fields such as studentId, name,
and address. Should these fields be static or non-static? Add the main
method to this class and access the fields from the main method.
4. Add a static field to Student class. Access this field from another class. Use
appropriate import statement to access the field directly. Change
accessibility of the field and see its impact on the code that tries to access it.
5. Create a method named method1 in TestClass that accepts a Student object
and updates the static as well as instance fields of this object. Pass the same
Student object to another method named method2 and print the values.
Assign a new Student object to the Student variable of method2 and set its
fields to different values. After returning back to method1, print the values
again. Explain the output.
6. Add a constructor in Student class that accepts values for all of its instance
fields. Add a no-args constructor in Student class that makes use of the first
constructor to set all its instance fields to dummy values.
7. Create a class named Course in different package. Add a static method
named enroll in this class that accepts a Student. Use different access
modifiers for fields of Student class and try to access them from the enroll
method.
Chapter 11 Encapsulation
11.1 Apply access modifiers

11.1.1 Accessibility ☝
One of the objectives of object-oriented development is to encourage the users of

a component (which could be a class, interface, or an enum) to rely only on the
agreed upon contract between the user and the developer of the component and
not on any other information that the component is not willing to share.
For example, if a component provides a method to compute taxes on the items in

a shopping cart, then the user of that component should only pass the required
arguments and get the result. It should not try to access internal variables or logic
of that class because using such information will tie the user of that component
too tightly to that component. Imagine a developer writes the following code for
the TaxCalculator class:
class TaxCalculator{
double rate = 0.1;
double getTaxAmount(double price){
return rate*price;
}
}
Ideally, users of the above class should use the getTaxAmount method but let
us say they do not and access the rate variable instead, like this:
//code in some other class
double price = 95.0;

TaxCalculator tc = new TaxCalculator();
double taxAmt = price*tc.rate;
Later on, the developer realizes that "tax rate" cannot be hardcoded to 0.1. It
needs to be retrieved from the database and so, the developer makes the
following changes to the class:
class TaxCalculator{
//double rate = 0.1; //no more hardcoding
double getTaxAmount(double price){

return getRateFromDB()*price;
}
double getRateFromDB(){
//fetch rate from db using jdbc
}
}
Since there is no rate variable present in this class anymore, all other code that
is accessing this variable will now fail to compile. Had they stuck to using the
getTaxAmount method, they would not have had any problem at all. To
prevent compilation failure, the developer will now be forced to maintain the
rate variable in their new code even though this variable is not required by the
class anymore. What if the tax rate changes in the database but is not updated in
the TaxCalculator class's rate variable? What if one rogue user updates
the rate variable at an inopportune time while another user is using it to compute
taxes? Well, in both the cases the users will be computing taxes incorrectly.
There will be no error message to make the developer aware of the problem
either. This is a serious matter.
But the problem is not just with the users relying on internal details of a class.
The TaxCalculator class doesn't give any clue as to what features does it
support. How are the users supposed to know that they should be using only the
getTaxAmount method and not the rate variable? In other words, the
TaxCalculator class doesn't make its public contract clear and therefore, it
would be unfair to blame only the users of this class. This is where Java's access
modifiers of Java come into picture.
11.1.2 Access modifiers ☝
Java allows a class and members of a class to explicitly specify who can access
the class and the members using three accessibility modifiers: private,
protected, and public . Besides these three, the absence of any access modifier
is also considered an access modifier. This is known as "default" access. These
access modifiers make your intention about a member very clear not only to the
users of your class but also to the compiler. The compiler then helps you enforce
your intention by refusing to compile code that violates your intention. Here is
how these modifiers affect accessibility:
private - A private member is only accessible from within that class. It cannot
be accessed by code in any other class.
For example, the problem that I showed you with the TaxCalculator class
could have been easily avoided if the developer had simply declared the rate
variable as private . That would make the variable inaccessible from any other
class. The compiler would prevent the users of this class from using the rate
variable by refusing to compile the code that tried to use it. Since there would
have been no dependency on rate variable, the developer could have easily
removed this variable without any impact on anyone.
"default" - A member that has no access modifier applied to it is accessible to

all classes that belong to the same package. This is irrespective of whether the
class trying to access a default member of another class is a sub class or a super
class of the other class. If two classes belong to the same package, then they can
access each other's default members. This is also called "package private" or
simply "package access" .
protected - A protected member is accessible from two places - if the accessing

class belongs to the same package or if the accessing class is a subclass
irrespective of the package to which the subclass belongs. The first case is
simple because it is exactly the same as default access. All classes belonging to
the same package can access each other's protected members. The second case is
not so, simple and I will explain it separately in detail.
public - A public member is accessible from everywhere. Any code from any
class can make use of a public member of another class. For example, the
getTaxAmount method could have been made public. That would give a clear
signal to the users to use this method if they want to compute the tax amount.
If you order the four access modifiers in terms of how restrictive they are, then
private is most restrictive and public is least restrictive . The other two, i.e.,
default and protected , lie between these two. Thus, the order of access
modifiers from the most restrictive to the least would be: private > default >
protected > public.
In the exam, watch out for non-existent access modifiers such as

"friend", "private protected", and "default".
11.1.3 Understanding protected access ☝
I mentioned earlier that a protected member is accessible from any class that
belongs to the same package and from any subclass even if the subclass belongs
to another package. This sounds straightforward but the second part of that
statement is not completely true. Since it involves the concept of inheritance,
which I haven't discussed yet, I will explain this case here with an example.
Consider the following two classes that belong to two different packages.
//In file Account.java

package com.mybank.accts;
public class Account{ //observe public modifier for c
lass
protected String acctId;
//code that does something with acctId
//In file HRAccount.java
package com.mybank.hr;
import com.mybank.accts.*;
public class HRAccount extends Account{

Account simpleAcct = new Account();
simpleAcct.acctId = "111";//will not compile
HRAccount hrAcct = new HRAccount();

hrAcct.acctId = "111";//will compile fine
}
}
In the above code, the compiler will not allow acctId to be accessed using the
reference simpleAcct but it will allow acctId to be accessed through the
reference hrAcct . Ideally, since acctId is protected, and since a protected
member can be accessed from a subclass irrespective of the package of the
subclass, then simpleAcct.acctId should have been valid. Then what is
the issue?
Actually, protected allows a subclass to access its own fields that the
subclass inherits from its superclass. HRAccount class inherits the acctId
field from Account class, so HRAccount class owns the acctId field
contained in any HRAccount object but HRAccount class does not own an
Account object's acctId field. That is why HRAccount class is allowed to
access HRAccount object's acctId field but is not allowed to access
Account object's acctId field.
In technical terms, a subclass from a different package is allowed to access a

protected member of the superclass only if the subclass is involved in the
implementation of the class of the reference that it is trying to use to access that
member. In the above example, the line simpleAcct.acctId violates this
rule because simpleAcct is a reference of type Account but
class HRAccount is not involved in the implementation of class Account . It
would not compile even if simpleAcct referred to an object of type
HRAccount because the compiler checks the declared type of the reference and
not the type of the actual object to which the reference points at run time.
Let me add one more class to the mix:
//In file NewHRAccount.java
package com.mybank.newhr;
import com.mybank.hr.*;
public class NewHRAccount extends HRAccount{
protected String name;
}
The above class is a subclass of HRAccount and belongs to a third package

named com.mybank.newhr . Now, what if I modify HRAccount code to
access NewHRAccount like this:
//In file HRAccount.java
package com.mybank.hr;
import com.mybank.accts.*;
import com.mybank.newhr.*;
public class HRAccount extends Account{

NewHRAccount newHRAcct = new NewHRAccount();
newHRAcct.acctId = "111";//will this compile?
newHRAcct.name = "John";//will this compile?
}
}
In the above code, I am trying to access acctId field of NewHRAccount

from HRAccount . Does HRAccount own the acctId field of
NewHRAccount ? Yes, it does because NewHRAccount inherits it through
HRAccount . Therefore, the access to acctId in the above code is valid.
However, access to name is invalid because name is defined in
NewHRAccount and therefore HRAccount does not own this field.
It sounds confusing and it is indeed confusing. But if you think about it, it does
make sense. Imagine you develop a class. This class provides some functionality
in a way you deem appropriate. But now you want other people to use your class
and also let them implement, enhance, or tweak that functionality in their own
way. You don't want them to change the way your class works, you just want to
them to be able to reuse your class and implement that functionality the way they
deem fit in their own class. For example, Account class of
com.mybank.accts package may manage its acctId field in a certain way.
HRAccount class from com.mybank.hr package inherits acctId and is
now free to manage acctId in its own way but HRAccount should not be
able to change how the Account class manages acctId . If HRAccount is
allowed to mess with Account 's acctId field, then Account class's internal
logic may go completely haywire. This is what protected intends to achieve. It
gives a subclass full control over the fields that the subclass inherits from its
parent without compromising the integrity of the parent itself.
11.1.4 Applying access modifiers to types ☝
So far you have seen the usage of access modifiers on fields and methods of a
class. You can apply them to the reference type definitions, i.e., classes,
interfaces, and enums definitions as well. However, there are certain restrictions.
Applying access modifiers to class definitions ☝
A top level class (i.e. a class that is not defined inside another reference type)
can only have two types of access - public and default. But a nested class can use
any of the four access modifier. For example,
public
class Outer{ //cannot be private or protected
private
class Inner1 { } //valid
protected
class Inner2 { } //valid

}
Similarly, a top level interface or an enum can also only have public or default
access but a nested interface or an enum can have any of the four access
modifiers.
Applying access modifiers to members of an interface ☝
As of Java 8, members of an interface are always public. The compiler will

generate an error if you define them as private or protected. If you don't specify
any access modifier for a member of an interface, compiler will automatically
make it public (and not default, unlike in case of a class). For example:
interface Movable { //cannot be private or protected
int STEP = 10; //interface fields are always public

even if not declared public
void move(); //interface methods are always public

except when declared private (protected is not allowed
for methods in an interface)
Java 9 allows an interface to have private methods. However, fields of an

interface are still always public.
Applying access modifiers to members of an enum ☝
The enum constants are always public even when no access modifier is
specified. On the other hand, enum constructors are always private. Compiler
will generate an error if you try to make them public or protected. For example:
enum Day{ //can only be public or default
WEEKDAY, HOLIDAY; //enum constants are always publ

ic
Day(){ //enum constructor is always private
private int value; //enum field can be public, priv

ate or protected
public void count(){ //enum method can be public,

private or protected
}
}
We haven't seen candidates getting questions on details

of access modifiers as applied to enums. I have
mentioned it only for the sake of completeness. You do,
however, need to know the application of access
modifiers on classes, interfaces, and their members.
11.2 Apply encapsulation principles to a class
11.2.1 Encapsulation ☝
Encapsulation is considered as one of the three pillars of Object-Oriented

Programming. The other two being Inheritance and Polymorphism . In the
programming world, the word encapsulation may either refer to a language
mechanism for restricting direct access to an object's data fields or it may refer
to the features of a programming language that facilitate the bundling of data
with the methods operating on that data. For the purpose of the exam, we will be
focusing on the first meaning, i.e., restricting direct access to an object's fields.
As I explained in the previous section on access modifiers , letting other classes

directly access instance variables of a class causes tight coupling between
classes and that reduces the maintainability of the code. While designing a class,
your goal should be to present the functionality of this class through methods
and not through variables. In other words, a user of your class should be able to
make use of your class by invoking methods and not by accessing variables.
There are two advantages of this approach:
1. You give yourself the freedom to modify the implementation of the

functionality without affecting the users of your class. In fact, users of a
well encapsulated class do not even become aware of the internal variables
that the class uses for delivering that functionality because the
implementation details of the functionality are hidden from the users. In
that sense, encapsulation and information hiding go hand in hand.
2. You can ensure that the value of a variable is always consistent with the
business logic of the class. For example, it wouldn't make sense for the age
variable of a Person class to have a negative value. If you make this
variable public, anyone could mess with Person objects by setting their
ages to a negative value. It would therefore, be better if the Person class
has a public setter method instead, like this:
class Person{
private int age;
public void setAge(int yrs){
if(yrs<0) throw new IllegalArgumentException
();
else this.age = yrs;
}
//other code
Encapsulation in Java ☝
Since the only way to restrict access to the variables of a class from other classes
is to make them private, a well encapsulated class defines its variables as private.
The more you relax the access restrictions on the variables, the less encapsulated
a class gets. Thus, a class with public fields is not encapsulated at all. The
visibility of the methods, on other hand, can be anything ranging from private to
public, depending on the business purpose of the class.
Encapsulation of static members ☝
Encapsulation is an OOP concept and generally applies to the instance members

of a class and not to the static members. However, if you understand the spirit of
encapsulation, you will notice that all we want to do is to prevent others from
inadvertently or incorrectly accessing stuff that they should not be accessing.
This is true of static variables also. Thus, it is better to have even the static
variables as private. Ideally, the only variables that deserve to be public are the
ones that are constants.
11.2.2 JavaBeans naming convention ☝
In the previous section, I explained why a class should expose its functionality
through methods and not through variables. If you look at a well encapsulated
class from a user's perspective, you will not see any variables (because they are
private!). For example, consider the following class:
package library;
public class Book{
private String title;
public String getTitle(){ return title; }
public void setTitle(String title){ this.title = ti
tle; }
}
If you try to use this class in your code, you will not see the variable title .
You will only see methods that get and set the value for its title but no
variable named title . The following code illustrates this point:
import library.Book;
class TestClass{
Book b = new Book();
//b.title = "Java Gems"; //won't work because ti
tle is private
b.setTitle("Java Gems");
System.out.println(b.getTitle());
}
}
Thus, as far as you are concerned, a Book object doesn't seem to have any
variable named title even though it does have a notion of having a title .
In technical terms, this notion is called "property" , as in, the Book class has a
property named title .
JavaBeans ☝
JavaBeans is a set of conventions that lets anyone make use of objects by easily
recognizing the properties that these objects have. There are only two rules that
you need to remember about JavaBeans for the exam:
1. If a class has a non-private method whose name starts with a "get" followed
by an upper case letter, then that class has a property by the name that
follows the get. The type of this property is the same as the return type of
this method. For example, the Book class shown above has a public
method named getTitle and therefore, it has a property named title of
type String . This method is called the "getter" method for the title
property.
If the type of the property is boolean (or Boolean ), the name of the
getter method may also start with an "is" instead of a "get".
2. Similarly, if the same class also has a non-private method whose name
matches "set" followed by the property name with first letter capitalized and
takes a parameter of the same type as the associated getter method, this
property is considered editable and the method is called the "setter"
method for that property. For example, since Book class does have a public
setTitle method that takes a String parameter, the title property is
editable. Had it not had the setTitle method, the title property would
have been "read only" .
Observe that there is no mention of the name of the variable behind this property.
Let me show you a few examples to make this clear:
package library;
public class Book{
private String writer;

public String getAuthor(){
return writer;
}
public void setAuthor(String value){
this.writer = value;
}
public String getJunkTitle(){

return "junk";
}
private double price;

public double getprice(){ return price; }
boolean isJunk(){ return true; }
boolean getJunk(){ return true; }
Let's apply the first rule to the above class:
1. The method getAuthor is not private and starts with a get followed by
an upper case letter. Therefore, author is a valid property of Book class
and getAuthor is the getter method for this property. Observe that the
variable that the class internally uses to support this property is named
writer but that has no relevance here. Since the method is public, the
author property is publicly readable.
2. The method getJunkTitle is not private and starts with a get followed
by an upper case letter. Therefore, junkTitle is a valid property of Book
class and getJunkTitle is the getter method for this property. Observe
that this method uses no internal variable at all. Furthermore, since this
method is public, the junkTitle property is publicly readable.
3. The Book class does not have any method named getPrice (with
uppercase p) therefore Book class does not have any property named price
even though it does have a variable named price . The method
getprice (with lower case p) starts with a get but is not followed by an
upper-case letter, therefore, getprice is not a valid getter method for any
property.
4. The method isJunk is not private and starts with an is followed by an

upper-case letter and returns a boolean . Therefore, junk is a valid
boolean property of Book class and isJunk is a valid getter method for
this property. Similarly, getJunk is also a valid getter method for the junk
property. Since the method has default access, the junk property is visible
only to the classes of the same package.
Now, let us apply the second rule:
1. Since Book class also has a setAuthor method, the author property is
editable.
2. Since Book class does not have a setJunkTitle method, the junkTitle
property is uneditable, i.e., read only.
3. Since Book class does not have a setJunk method, the junk property is
uneditable, i.e., read only.
Since JavaBeans is a globally accepted convention for naming the properties,

even automated tools can see and access the properties of any object. The objects
themselves (and not the classes) are called "beans". A JavaBeans developer can
make beans provided by different providers work together by having them
access each other properties just through declarations in a visual editor without
writing any code explicitly.
11.3 Exercise ☝
1. List the access modifiers in the order of increasing restrictiveness.
2. What will be the impact if accessibility of a package protected instance
member of a class is changed to public or private?
3. Evaluate the following class code with respect to encapsulation principles:
class Book{
public String isbn;
public String title;
private String author;
public Book(){
}
getAuthor(){ return author; }
static int noOfAuthors;

}
4. What can be done to make the above class well encapsulated?

Chapter 12 Reusing Implementations
Through Inheritance
12.1 Create and use subclasses and superclasses

12.1.1 Understanding Inheritance ☝
In general terms, when a class extends another class, we say that the class
inherits from the other class. But this is a very vague view of inheritance. It is
important to understand what exactly this class inherits from the other class and
what exactly inheriting something means.
There are two things that a class contains - the instance fields defined in the
class, i.e., state , the instance methods that provide behavior to the class, i.e.,
implementation . Furthermore, a class, on its own, also defines a type . Any of
these three things can be inherited by the extending class. Thus, inheritance
could be of state, it could be of implementation, and it could be of type.
When a class inherits something, it implies that it automatically gets that thing
without you needing to explicitly define it in the class. This is very much like
real life objects. When you say that a Poodle is a Dog, or a Beagle is a Dog, you
implicitly know that a Poodle has a tail and that it barks. So, does a Beagle. You
don't have to convey this information explicitly with a Poodle or a Beagle. In
other words, both a Poodle and a Beagle inherit the tail and the barking behavior
from Dog. Note that Poodle and Beagle do not contain a Dog. Poodle is a Dog.
Beagle is a Dog. Being something is very different from containing something.
If Poodles and Beagles contained a Dog, then both of them would always bark in
exactly the same way, i.e., like a Dog. But you know that they don't bark the
same way. Both of them do bark but they bark differently.
Inheritance of state ☝
Only a class can contain state and therefore, only a class can extend another
class. Furthermore, Java restricts a class from extending more than one class and
that is why it is often said that Java does not support multiple inheritance.
However, technically, it would be more precise to say that Java does not support
multiple inheritance of state.
Inheritance of implementation ☝
Before Java 8, only classes were allowed to have implementation. Starting with
version 8, Java allows even interfaces to contain implementations in the form of
"default" methods. Thus, a class can inherit implementation by extending a class
and/or by implementing interfaces. This allows a class to inherit
implementations from more than one types. This is one form of multiple
implementation inheritance. However, due to the way default methods are
inherited in a class, it is still not possible for a class to inherit more than one
implementation of a method in Java. I will talk more about default methods later.
Inheritance of type ☝
Java allows you to define a type using an interface as well as a class (and an
enum but that is not relevant to this discussion). Thus, a class can inherit
behavior by extending a class and/or by implementing an interface. Since Java
allows a class to implement multiple interfaces, it can be said that Java supports
multiple inheritance of type.
12.1.2 Inheriting features from a class ☝
To inherit features from another class, you have to extend that class using
the extends keyword. For example, consider the following two classes:
public class Person{

public String name;
public String getName() {
return name;
}
public static
int personCount;
public static
int getPersonCount(){
return personCount;
}
}
public class Employee extends Person{

public String employeeId;
public static void main(String args[]){

Employee ee = new Employee();
ee.employeeId = "111";
ee.name = "Amy";
System.out.println(ee.getName());
In the above code, Person is the parent class (also called the super class or the
base class ) and Employee is the child class (also called sub class or derived
class ). Since Employee extends Person , it automatically "gets" the name
field as well the getName method. Thus, it is possible to access the name field
and the getName method in an Employee object as if they were defined in the
Employee class itself, just like the employeeId field.
Similarly, the Employee class also "gets" the static variable personCount
and the static method getPersonCount from Person class. Thus, it is
possible to access personCount and getPersonCount in an Employee
class as if they were defined in the Employee class itself.
Members inherited by a class from its super class are as good as the members
defined in the class itself and are therefore, passed on as inheritance to any
subclass. For example, if you have a Manager class that extends an Employee
class, the Manager class will inherit all members of the Employee class, which
includes the members defined in Employee class as well as members inherited
by Employee class from Person class.
public class Manager extends Employee {

public String projectId;

Manager m = new Manager();
m.projectId = "OCPJP-I";
m.employeeId = "111";
m.name = "Amy";
System.out.println(m.getName());
There is no limit to how deep a chain of inheritance can go.
Inheriting constructors and initializers ☝

Constructors are not considered members of a class and are therefore, never
inherited. The following code proves it:
class Person{
String name;
Person(String name){
this.name = name;
}
}
Employee ee = new Employee("Bob");
Employee does not inherit the String constructor of Person and that is why the
compiler is not happy when you try to instantiate a new Employee using a String
constructor.
Similarly, static and instance initializers of a class are also not considered to be
members of a class and they are not inherited by a subclass either.
The java.lang.Object class ☝
java.lang.Object is the root class of all classes. This means that if a class
does not explicitly extend another class, it implicitly extends the Object class.
Thus, every class inherits the members defined in the Object class either
directly or through its parent. There are several methods in this class but the only
methods that you need to be aware of for the OCP Java 11 Part 1 exam are the
equals and the toString method. It is because of the presence of these
methods in the Object class that you can call them on any kind of object. I will
talk more about them later.
Since Object is the root class of all classes, it is the only class in Java that does
not have any parent.
Extending multiple classes ☝
As I mentioned earlier, Java does not support multiple inheritance of state. Since
classes contain state, it is not possible for a class to extend more than one class.
Thus, something like the following will not compile:
public class Programmer{

}
//can't extend more than one class
public class Consultant extends Person, Programmer{
12.1.3 Inheritance and access modifiers ☝
You might have noticed that I violated an important principle of OOP in

Person and Employee classes. These classes are not well encapsulated.
Recall that in the previous chapter, I explained why it is bad to have public
fields. Here however, both the Person class, and since the Employee class
extends Person , the Employee class have a field that is exposed to the
whole world. As per the principle of encapsulation, I should make the name
field private. However, if I make name private, Employee fails to compile
with the following error message:
Error: name has private access in Person
How come, you ask? It turns out that access modifiers greatly impact
inheritance. In fact, only those members that are visible to another class as per
the rules of access modifiers are inherited in a subclass. Remember that a private
member is not visible from anywhere except from the declaring class itself.
Thus, a subclass cannot inherit a private member of the super class at all. That is
why Employee will not inherit name from Person if you make name
private.
With the above rule in mind, let us examine the impact of access modifiers on
inheritance.
private - As explained above, private members are not inherited by a subclass.
default - Since a member with default access is visible only to a class in the
same package, it can only be inherited in the subclass if the subclass belongs to
the same package.
protected - This access modifier is actually built to overcome the restriction

imposed by the default modifier. Recall that in addition to being visible to all
classes of the same package, a protected member of a class is visible to another
class that belongs to a different package if that class is a subclass of this class.
Thus, even a subclass that belongs to another package inherits protected
members of the super class.
public - public members are always inherited by a subclass.
For example, let's say you are designing an application that has three kinds of
accounts: CheckingAccount , SavingsAccount , and RetirementAccount . All
three classes need to have an account number and methods for checking balance.
Besides these common features, each of these account types also have features
that are specific only to that account type. The CheckingAccount has
an overdraftLimit , the SavingsAccount has an interestRate , and the
RetirementAccount has a maturityDate . The figure below shows the hierarchy
of the classes.
Since all accounts are to have accountNumber and balance , I have put
these members in a common super class called Account . The following is the
code for these classes:
package accts;
class Account{
int accountNumber;
protected double balance;
//public getter and setter methods for the above fi

elds
package accts;
public class CheckingAccount extends Account{
private double odLimit;
//public getter and setter methods for odLimit
}
package accts;
public class SavingsAccount extends Account{
double interestRate;
//public getter and setter methods for interestRate
package specialAccts;
import accts.SavingsAccount;
public class RetirementAccount extends SavingsAccount
{
private String maturityDate;
//getter and setter methods for maturityDate
Account a = new Account(); //will not compile
RetirementAccount ra = new RetirementAccount();

//valid
ra.balance = 100.0;//valid
ra.setAccountNumber(10);//valid
ra.setInterestRate(7.0);//valid
SavingsAccount sa = new SavingsAccount();//valid
sa.balance = 10.0; //will not compile
ra.accountNumber = 10;//will not compile
ra.interestRate = 7.0;//will not compile
}
}
To make it interesting, I have put RetirementAccount into a different

package than the rest of the classes. Let us now see the impact of the various
access modifiers on these classes:
1. Since Account has default access, it is visible only in the accts package
and thus RetirementAccount cannot access or extend the Account
class.
2. Since accountNumber has default access, it is visible only in the accts

package and thus, it will be inherited by all the subclasses of Account that
belong to the same package. Thus, accountNumber will be inherited by
CheckingAccount and SavingsAccount .
3. Since balance has protected access, it is visible in all the classes of

the accts package and all the subclasses of the Account class
irrespective of their package. Thus, it will be inherited by
CheckingAccount and SavingsAccount . Since Account is not
visible outside the accts package, making balance protected doesn't
seem to make much sense but it does . Observe what happens with
RetirementAccount .
4. Since RetirementAccount extends SavingsAccount but belongs

to a different package, it only inherits the public and protected (but not
default) members of SavingsAccount . This means, the protected field
balance is passed on to the RetirementAccount as well. Since the
interestRate field of SavingsAccount has default access, it is not
visible in RetirementAccount and therefore, is not inherited by
RetirementAccount .
5. Even though the balance field of Account is visible and inherited in

RetirementAccount , sa.balance will still not compile because
RetirementAccount class does not own SavingsAccount 's
balance as explained in the previous chapter. In technical terms, code of
RetirementAccount class is not responsible for the implementation of
SavingsAccount class and so, RetirementAccount cannot access
SavingsAccount 's balance . It can access its own balance , which
it inherits from SavingsAccount , and which is why ra.balance
compiles fine, but not SavingsAccount 's balance .
Memory impact of access modifiers on sub classes

☝
Even if a subclass does not inherit some of the

instance variables of a superclass (because of access
modifiers), a subclass object still consumes the space
required to store all of the instance variables of the
superclass. For example, when the JVM allocates
memory for a RetirementAccount object, it includes
space required to store all instance variables of
SavingsAccount and Account irrespective of whether
they are inherited in RetirementAccount or not. Thus,
from a memory perspective, access modifiers have no
impact.
So, on one hand, we are saying that a private member
is not inherited by a subclass and on the other, we are
saying that the subclass object does contain that
member in its memory. This contradiction makes an
intuitive understanding of the term "inheritance" difficult.
Unfortunately, that is how it is. For example, in Section
8.2, The Java Language Specification says, "Members of
a class that are declared private are not inherited by
subclasses of that class." This shows that the JLS links
the term inheritance with access modifiers and doesn't
give any consideration to memory.
12.1.4 Inheritance of instance members vs static members ☝
Inheritance of instance variables vs static variables ☝
In an earlier example, you saw that the Employee class inherits instance as
well as static variables of its superclass Person . However, there is a
fundamental difference between the way instance and static variables are
inherited. This difference is highlighted in the following code:
public class Person{

public String name;
public static
int personCount;
}

//inherits instance as well as static fields of Per
son
}
class TestClass{


Person p = new Person();
Employee e = new Employee();
p.name = "Amy";
e.name = "Betty";
System.out.print(p.name+" ");
System.out.println(e.name);

Employee.personCount = 2;
System.out.print(Person.personCount+" ");
System.out.println(Employee.personCount);
}
}
The above code generates the following output:
Amy Betty
2 2
Observe that it prints different values for name - one for each object, but same
value for personCount . What this means is that both the objects got their
own personal copy of name , which they were able to manipulate without
affecting each other's copy but the static variable personCount was shared by
the two classes. When the code updated personCount of Employee ,
personCount of Person was updated as well. In fact, Employee did not
get its own copy of personCount at all. It merely got access rights to
Person 's personCount .
Inheritance of instance methods vs static methods ☝

Since methods don't consume any space in an object (or a class), there is just one
copy of a method anyway, irrespective of whether it is an instance method or a
static method.
Conceptually, however, the difference that I highlighted above for variables also
exists for methods. Conceptually, a subclass inherits its own copy of an instance
method. This is proven by the fact that a subclass can completely replace the
behavior of an inherited instance method for objects of the subclass by
"overriding " it with a new implementation of its own without affecting the
behavior of the superclass's implementation for objects of the superclass. On the
other hand, a subclass merely gets access rights to a static method of its
superclass and so, the subclass cannot change the behavior of the inherited static
method. The subclass can "hide " the behavior of the superclass's static method
by providing a new implementation but it cannot replace the super class's
method.
Overriding and hiding are technical terms with precise meanings and are
closely related to polymorphism . Their understanding is crucial for the exam as
well as for being a good Java developer. I will dig deeper into these terms after
we go over the nuts and bolts of extending classes and implementing interfaces.
12.1.5 Benefits of inheritance ☝
First and foremost, Inheritance allows you to group classes by having them
extend a common parent class . Functionality that is common to all such classes
need to be defined only once in the parent class. For example, in the class
hierarchy consisting of the Account , CheckingAccount , and SavingsAccount
classes that I used at the beginning of this chapter, I put common functionality
such as account number and method for checking balance in a parent class
named Account and I put unique features of each of these account types in
separate sub classes.
There are several advantages with this approach. Let's examine them one by one.
Code reuse ☝
You can write logic for common fields only in one place and share that logic
with all classes. In the above example, there is only one copy of the code to
manage the account number. Since this logic is in the parent class, it is
automatically inherited by all the child classes.
Having a common base class also makes it easy to add new functionality or
modify the existing functionality that is common to all classes. This makes the
code more extensible and maintainable overall.
Information hiding ☝
Isolating the common features to a common parent class helps you hide the
features of individual sub classes to processes that don't need to know of those
features. As discussed before, more exposure means more risk of inadvertent
dependency, i.e., tighter coupling, which is not desirable. For example, if you
have a process that loops through all accounts and prints their account numbers
and balances, then this process doesn't need to know about overdraft limit,
interest rate, or maturity date. If you just pass an array of Accounts to this
process, that should be enough. Something like this:
public class DumpAccountInfo{

public static void printAccounts(Account[] accts){
for(Account acct : accts){
System.out.println(acct.getAccountNumber()+" "+
acct.getBalance());
}
}
}
Of course, the array will include all kinds of Account objects (i.e. objects of
sub-classes of Account class) but the process will only see them as Account
objects and not as CheckingAccount , SavingsAccount , or
RetirementAccount objects. The above code doesn't care whether an Account
object is really a SavingsAccount or a CheckingAccount. It just prints the
account number and its balance irrespective of what kind of account it is. This is
possible only due to inheritance. Without the common parent class, you would
have to have three different methods - one for each of the three account types -
to print this information.
Polymorphism ☝
Finally, and most importantly, inheritance makes polymorphism possible.

Polymorphism is a topic in its own right and I will discuss it soon in a section of
its own.
12.2 Using super and this to access objects and

constructors
12.2.1 Object initialization revisited ☝
Recall that in the "Create and overload constructors" section of the "Working
with Methods and Encapsulation" chapter, I talked about the four steps that a
JVM takes while instantiating a class. Let's see how these steps are impacted
when there is inheritance involved.
1. The first step was to load and initialize the class if it is not already loaded
and initialized.
2. The second step was to allocate the memory required to hold the instance
variables of the object in the heap space. Since the instance variables
defined in a super class are also included in object of a subclass (whether
they are accessible to the subclass or not is a different matter), the memory
allocated by the JVM for a subclass object must include space for storing
instance variables of the super class as well.
3. The third step was to initialize these variables to their default values. This
means the inherited variables also need to be initialized to their default
values.
4. The fourth step was to give that instance an opportunity to set the values of
the instance variables as per the business logic of that class by executing
code written in instance initializers and constructors. This step gets a little
more complicated when the class extends another class. Remember that the
whole purpose of inheriting features (i.e. variables as well as methods) of a
superclass is for the subclass to be able to use those features! It can use
these features even at the time of its initialization. But to be able to do that,
those features have to be initialized first. This means that subclass cannot
initialize its own features unless features of its superclass have been
initialized.
You can see where this is going. The superclass cannot be initialized before the
superclass of the superclass is initialized and so on until there is no super class
left. This chain of initialization will stretch until the Object class because Object
is the root class of all classes and it is the only class that has no super class.
You have seen that the initialization is triggered when you try to create an
instance of that class using the new keyword. When the JVM encounters the new
keyword applied to a class, it invokes the appropriate constructor of that class.
But as discussed above, the execution of this constructor cannot proceed until
the initialization of the superclass is complete.
12.2.2 Initializing super class using "super" ☝
To ensure the initialization of the fields inherited from the superclass, a

constructor must first invoke exactly one of its super class's constructors. This is
done using the super(<arguments>) syntax. For example:
class Person{
String name;
this.name = name;
}
}

public Employee(String s){
super(s);
}
}
The question that you might ask at this point is what about Person class. This
class has no extends clause and that means it implicitly extends
java.lang.Object . So where is the call to Object 's constructor in
Person 's constructor?
Good question. The call to the super class's constructor is so important that if a
class's constructor doesn't call any of the super class's constructors explicitly, the
compiler automatically inserts a call to the super class's default constructor in the
first line of that constructor. The compiler doesn't check the constructors that the
super class has. It just assumes the presence of the default constructor and inserts
a call to that constructor. Thus, basically, Person 's constructor is modified by
the compiler as follows:
class Person{
String name;
super();
//<-- inserted automatically by the compiler
this.name = name;
}
}
This means that Object 's no-args constructor will be invoked first before
proceeding with the execution of Person 's constructor. Armed with this
knowledge, let us see what happens when I modify Employee as follows:
name = s;
}
}
Any thoughts? That's right. It won't compile. Since Employee 's constructor
doesn't call super class's constructor explicitly the compiler inserts a call to
super(); on its own. This means Employee 's constructor really looks like
this:

super(); //inserted automatically by the compiler
name = s;
}
But Person does not have any no-args constructor! Recall that a default no-
args constructor is provided by the compiler to a class only if the class does not
define any constructor at all. In this case, Person does define a constructor
explicitly and so the compiler does not insert a no-args constructor in
the Person class. Thus, Employee will fail to compile.
Let me make one more change. What if I remove all constructors from
Employee class?

}
}
Again, recall the rule about the default constructor. Since Employee does not
define any constructor explicitly, the compiler inserts a no-args constructor
automatically, which looks like this:

public Employee(){ //inserted by the compiler
super(); //inserted by the compiler
}
}
Observe that this default no-args constructor also contains a call to super();
(See how important invoking a super class's constructor is?). But as discussed
above, Person doesn't have a no-args constructor and therefore, the above code
will not compile!
Since an object can be initialized only once, call to super(<arguments>);

can also be made only once. If you try to call it more than once, the code will
fail to compile.
Invoking another constructor using this(<arguments>) ☝
In the previous chapter, you saw how a constructor can invoke another
constructor of the same class using the this(<arguments>); syntax. Recall
that invocation of another constructor is only allowed if it is the first statement of
a constructor. But this poses a small problem. How can both -
this(<arguments>); and super(<arguments>); , be the first
statement of a constructor at the same time? Well, they can't be. In fact, Java
allows only one of the two statements in a constructor. In other words, if you call
this(<arguments>); then you can't call super(<arguments>); and
if you call super(<arguments>); , you can't call
this(<arguments>); .
Note that this does not violate our original premise that the super class's features
have to be initialized first before initializing subclass's features. If you call
this(<arguments>); , you don't need to call super(<arguments>);
anyway because the other constructor would have called
super(<arguments>); and initialized the super class's features. If you call
super(<arguments>); , then by prohibiting you from calling
this(<arguments>); , Java prevents the invocation of the super class's
constructor more than once.
Based on the above discussion, can you determine the output generated when
TestClass is run from the command line:
class Person{
String name;
System.out.println("In Person's constructor ");
this.name = name;
}
}
class Employee extends Person{
String empId;
Employee(){
this("dummy", "000");
System.out.println("In Employee() constructor ");
}

Employee(String name, String empId){
super(name);
System.out.println("In Employee(name, empid) const
ructor ");
}
}
class Manager extends Employee{

String grade;
Manager(String grade){
System.out.println("In Manager(grade) constructor
");
this.grade = grade;
}
}
class TestClass{
Manager m = new Manager("A");
}
}
In TestClass 's main, an object of Manager class is being created using the
only constructor it has. This constructor does not invoke superclass's constructor
explicitly, therefore, the compiler inserts a call to super(); as the first
statement of this constructor. This takes us to the no-args constructor of
Employee . Employee 's no-args constructor invokes the two argument
constructor of the same class explicitly using this("dummy", "000") .
The two arguments constructor of Employee invokes its superclass's
constructor using super(name) . This takes us to the Person 's single
argument constructor. Inside the Person(String ) constructor, there is no
explicit call to its superclass's constructor. So the compiler inserts super(); as
the first statement. This invokes java.lang.Object 's no-args constructor
because Person implicitly extends java.lang.Object . Since Object
class is the root class, this is the last constructor in the chain. Object 's
constructor doesn't produce any output.
Now, we have to start unwinding this chain of calls. After finishing the execution
of Object 's constructor, the control returns to next statement in Person 's
constructor, which is System.out.println("In Person's
constructor ") . Thus, the first line on the output is "In Person's
constructor" . Next, the name argument is assigned to the name instance
variable and the call returns back to the Employee 's constructor. The next
statement in this constructor is System.out.println("In
Employee(name, empid) constructor "); , which prints "In
Employee(name, empid) constructor " and the control goes back to
Employee 's no-args constructor, which prints "In Employee()
constructor " . The control returns to Manager 's single argument
constructor, which prints, "In Manager(grade) constructor " . This
completes the chain of constructor invocation executed during creation of a
Manager object. Thus, the output is:
In Person's constructor
In Employee(name, empid) constructor
In Employee() constructor
In Manager(grade) constructor
This chain of calls is illustrated in the following figure.

12.2.3 Using the implicit variable "super" ☝
When a method is overridden by a subclass, it is impossible for any unrelated

class to execute the super class's version of that method on a subclass object.
However, the subclass itself can access its super class's version using an implicit
variable named "super" . Every instance method of a class gets this variable. It
is very much like the other implicit variable that you saw earlier, i.e., "this" .
While this refers to the current object, super refers to the members that this
class inherits from its parent class. The following code shows how it is used:
class InterestCalculator{
public double computeInterest(double principle, int
yrs, double rate){
return principle*yrs*rate;
}
}
class CompoundInterestCalculator extends InterestCalcu

lator{
yrs, double rate){
return principle*Math.pow(1 + rate, yrs) - pr
inciple;
}
//invoke this method from TestClass's main
public void test(){

double interest = super.computeInterest(100, 2,
0.1);
System.out.println(interest); //prints 20.0
interest = computeInterest(100, 2, 0.1);

0000000014
}
CompoundInterestCalculator cic = new Compoun
dInterestCalculator();
cic.test();
}
}
The call to super. computeInterest(100, 2, 0.1); in the above

code causes InterestCalculator 's version to be invoked while the call to
computeInterest(100, 2, 0.1) causes
CompoundInterestCalculator 's version to be invoked
The variable super is available only in instance methods of a class and can be
used to access any inherited member (static as well as instance) of the super
class. Are you now thinking about forming a chain of supers? Something like
super.super.someMethod() ? No, you can't do that because there is no
member named super in the super class. In fact, you can't define a variable or a
method named super in any class. This begs the question that how can a class
access the grandparent's version of a method that is defined in its grand parent as
well as in its parent class? For example, what if you have the following
grandchild class and you want to access InterestCalculator 's version of
computeInterest in this class:
class SubCompoundInterestCalculator
extends CompoundInterestCalculator{
public void test(){
//invokes CompoundInterest's computeInterest

super.
computeInterest(100, 2, 0.1);
//can't do super.super
//super.super.computeInterest(100, 2, 0.1);
The answer is you can't. There is no way to go more than one level higher. In
other words, a subclass can only access its immediate super class's version of
methods using super.
Note that this limitation applies only to methods that are overridden by the
parent class. If the parent class does not override a method of its super class,
then the parent class inherits that method from the grandparent and the child
class can access the grandparent's version of the method as if it were the parent
class's version.
12.2.4 Order of initialization summarized ☝
You have now seen all the steps, from loading the class to executing a
constructor of the class, that a JVM takes to create an object of a class. Here is
the order of these steps for quick reference:
1. If there is a super class, initialize static fields and execute static initializers
of the super class in the order of their appearance (performed only once per
class)
2. Initialize static fields and execute static initializers of the class in the order
of their appearance (performed only once per class)
3. If there is a super class, initialize the instance fields and execute instance
initializers of the super class in the order of their appearance
4. Execute super class's constructor
5. Initialize the instance fields and execute instance initializers of the class in
the order of their appearance
6. Execute class's constructor
The following example illustrates the above steps:
class A {
static{ System.out.println("In A's static initializ
er"); }
A(){ System.out.println("In A's constructor"); }
{ System.out.println("In A's instance initializer")

; }
}
public class B extends A {

static{ System.out.println("In B's static initializ
er"); }
{ System.out.println("In B's instance initializer")

; }
B(){ System.out.println("In B's constructor"); }

System.out.println("In B.main()");
B b = new B();
B b2 = new B(); //creating B's object again
}
}
The following output is generated when class B is run:

In A's static initializer
In B's static initializer
In B.main()
In A's instance initializer
In A's constructor
In B's instance initializer
In B's constructor
In A's instance initializer
In A's constructor
In B's instance initializer
In B's constructor
You should observe the following points in the above output:
1. static initializers of A and B are invoked before B's main() is executed

because the JVM needs to load the class before it can invoke a method on
it.
2. Static initializers of A and B are invoked only once even though two objects
of B are created.
3. Instance initializer of A is executed before A's constructor even though the
instance initializer appears after the constructor in A's code.
4. While creating an object of B, A's instance initializer and constructor are
invoked before B's instance initializer and constructor.
5. Instance initializer and constructor of A are invoked each time an object of
B is created.
Order of execution of various members of a class and its superclass

can sometimes get very difficult to trace. However, the exam does
not try to trick you too much on this topic. If you can follow the above
example, you will not have any problem with the code in the real
exam.
12.3 Create and extend abstract classes

12.3.1 Using abstract classes and abstract methods ☝
I talked about abstract class as a type of type briefly in the "Kickstarter for
Beginners" chapter. An abstract class is used when you want to capture the
common features of multiple related object types while knowing that no object
that exhibits only the features captured in the abstract class can actually exist.
For example, if you are creating an application that deals with cats and dogs, you
may want to capture the common features of cats and dogs into a class called
Animal. But you know that you can't have just an Animal. You can have a cat or
a dog but not something that is just an animal without being a cat, a dog or some
other animal. In that sense, Animal is an abstract concept . Furniture is another
such concept. If you want to have a Furniture, you have to get a chair or a table
because Furniture doesn't exist on its own. It is an abstract concept.
Abstract classes are used to capture such abstract concepts. An abstract class can
be defined by using the keyword abstract in its declaration. For example, the
following is how you can define Furniture :
abstract
class Furniture{
private String material;
//public getter and setter
Observe that Furniture is pretty much like a regular class with public and
private fields and methods. However, since it is declared abstract, it cannot be
instantiated. Thus, if you try to do new Furniture() , the compiler will
generate an error saying, "Furniture is abstract; cannot be
instantiated ".
But Chair and Table are concrete concepts. They exist on their own. Since both
of them are types of Furniture, you can model them as two classes that extend
Furniture.
class Chair extends Furniture{

//fields and methods specific to Chair
}
class Table extends Furniture{
//fields and methods specific to Table
Since Chair and Table extend Furniture , they will inherit all of the
features of Furniture just like any subclass inherits all of the features of its
super class.
Note that classes that are not abstract are also called "concrete" classes because
they represent real objects.
Adding an abstract method to an abstract class ☝
Let's say every piece of furniture in a store has to provide a method for its
assembly. Since the steps to assemble a chair will be different from steps to
assemble a table, we can't include the steps for their assembly in Furniture class.
In other words, what we are saying is that the declaration of the method to
assemble is common to all furniture but their implementation is unique to each
type of furniture. Java allows you to capture declaration without implementation
in the form of an abstract method . An abstract method is declared by applying
the abstract keyword to the method declaration. It must not have a method body
either. For example:
abstract class Furniture{

private String material;
//public getter and setter
public abstract
void assemble(); //ends with a semicolon, no opening
and closing curly braces
The benefit of adding abstract methods to a class is that other components can
work with all objects uniformly. For example, a class that assembles furniture
doesn't need to worry about whether it is assembling a Chair or a Table. It can
simply call assemble() on any kind of furniture that it gets. Something like
this:
class FurnitureAssembler{
public static void assembleAllFurniture(Furniture[]
allFurniture){
for(Furniture f : allFurniture){
f.assemble();
}
}
}
But adding an abstract method to the parent class has an impact on child classes
in that the child classes are now obligated to provide an implementation for the
abstract method. For example, we will now need to update the code for our Chair
and Table classes as follows:
class Chair extends Furniture{

//fields and methods specific to Chair
public void assemble(){

System.out.println("Assembling chair!");
}
}
class Table extends Furniture{
//fields and methods specific to Table

System.out.println("Assembling table!");
}
}
This makes sense because if a piece of furniture actually exists then we must be
able to assemble it as per our definition of Furniture . The only reason why a
subclass of Furniture may be unable to have a method to assemble is if that
subclass itself is abstract! For example, what if you have a subclass of
Furniture called FoldingFurniture ? FoldingFurniture is an
abstract concept and can be modeled as an abstract class that extends
Furniture . Since it is abstract, it can get away without implementing the
assemble method. But any concrete subclass of FoldingFurniture will
have to provide an implementation for the assemble method.
If you are confused about how the code for FurnitureAssembler can
invoke the assemble methods of Chairs and Tables when it does not even
know about their existence, don't worry. I will talk about it in detail soon in the
section about polymorphism .
12.3.2 Using final classes and final methods ☝

You have seen the usage of final keyword while declaring variables. It implies
that the variable is a constant , i.e., the value of the variable does not change
throughout the execution of the program. The final keyword can also be applied
to classes and methods (static as well as instance) and implies something similar.
It means that the behavior of the class or the method cannot be changed by any
subclass.
Observe that final is diametrically opposite of abstract . You make a class or a

method abstract because you want a subclass of that class to provide different
implementation for that class or that method as per the business logic of the
subclass. On the other hand, you make a class or a method final because you
don't want the behavior of the class or of the method to be changed at all by a
subclass. In technical terms, we say that a final class cannot be subclassed and a
final method cannot be overridden . I will discuss overriding in detail soon but
you can now see why abstract and final cannot go together.
A final class doesn't necessarily have to have a final method. But practically,
since a final class cannot be extended, none of its methods can be overridden
anyway. The important thing is a final class cannot have an abstract method.
Why? Because there is absolutely no possibility of that abstract method ever
getting implemented. It follows that if a final class inherits an abstract method
from any of its ancestors, it must provide an implementation of that method.
Here is an example of a final class:
public final class Chair extends Furniture{
//since assemble is an abstract method in Furniture

, it must be implemented in this class

System.out.println("Assembling Chair");
}
}
Observe that Chair doesn't have any final method and that it must implement
the abstract method assemble that it inherits from its superclass.
A final method is defined similarly:
public abstract
class Bed extends Furniture{
//not necessary to implement assemble() because Bed

is abstract
public final
int getNumberOfLegs(){
return 4;
}
public static final void make(Bed b){

System.out.println("Making Bed: "+b);
}
}
public final class DoubleBed extends Bed{
//must implement assemble because DoubleBed is not

abstract

System.out.println("Assembling DoubleBed");
}
//Can't override getNumberOfLegs here because it is

final in Bed
final
int getHeight(){
return 18;
}
}
Observe that even though Bed has a final method, Bed itself is not only not
final, it is abstract. I made it abstract just to illustrate that any class irrespective
of whether it is final, abstract, or neither, can contain a final method.
Java standard class library includes several final classes. Most important of them
is java.lang.String class, which I will talk about at length in another
chapter.
12.3.3 Valid combinations of access modifiers, abstract, final, and

static ☝
Impact of access modifiers on abstract and final ☝
The usage of abstract makes sense only in the presence of inheritance. If

something is never inherited then there is no point in talking about whether it is
abstract or final. You also know that a private member is never inherited. Based
on this knowledge can you tell what will happen with the following code:
abstract class Sofa {

private abstract void recline();
}
That's right. The compiler will reject the above code with the message,
"illegal combination of modifiers: abstract and
private" . Because a private method is never inherited, there is no way any
subclass of Sofa can provide an implementation for this abstract method.
What if the recline method were protected or default? It would have been ok
in that case because it is possible for a subclass to inherit methods with protected
and default access.
Let us now see what happens when you make a private method final:
class Sofa {
private final void recline(){
}
}
What do you think will happen? It won't compile, you say? Wrong! If a method
cannot be inherited at all by any subclass, is it even possible to override it? No.
Thus, a private method is, practically, already final! The compiler accepts the
above code because there is no contradiction between private and final. Marking
a private method as final is, therefore, not wrong but rather redundant.
abstract and static ☝
The abstract keyword is strictly about overriding and static methods can never be
overridden . Therefore, abstract and static cannot go together. Thus, the
following code will not compile:
abstract class Bed extends Furniture{

static abstract void getWidth();
}
The compiler will generate an error message saying, "illegal

combination of modifiers: abstract and static" .
final and static ☝
Although a static method cannot be overridden, it is nevertheless inherited by a

subclass and can be hidden by the subclass. I will discuss the difference between
overriding and hiding in detail later but for now, just remember that final
prevents a subclass from hiding the super class's static method. Thus, the
following code will compile fine:
class Bed{
static final int getWidth(){
return 36;
}
}
But the following will not because getWidth is final in Bed :

class DoubleBed extends Bed
{
static int getWidth(){ //will not compile
return 60;
}
}
Summary of the application of access modifiers, final, abstract, and static

keywords ☝
Based on the previous discussion, let me summarize the rules of abstract,

concrete classes, and final classes and methods:
1. An abstract class doesn't necessarily have to have an abstract method but if

a class has an abstract method, it must be declared abstract. In other words,
a concrete class cannot have an abstract method.
2. An abstract class cannot be instantiated irrespective of whether it has an
abstract method or not.
3. A final class or a final method cannot be abstract.
4. A final class cannot contain an abstract method but an abstract class may
contain a final method .
5. A private method is always final .
6. A private method can never be abstract .
7. A static method can be final but can never be abstract .
The exam tests you on various combinations of access modifiers and

final, abstract, and static keywords as applied to classes and
methods. You should be very clear about where you can and cannot
apply these modifiers.
The exam, however, does not try to trick you on the order of access
modifiers and final and abstract keywords. For example, you will not
be asked to pick the right declaration out of, say, the following three:
public abstract void m();

final public void m(){ }
abstract protected void m();
Nevertheless, you should know that the right sequence of modifiers

in a method declaration is as follows:
<access modifier > <static > <final or abstract >

<return type > methodName(<parameter list >)
If you ever get confused about their order, just recall the signature of
the main method as:
public static final void main(String[] args) .
Of course, the main method does not have be final. I have mentioned
it above to show the position of final in a method declaration.
12.4 Enable polymorphism by overriding methods

12.4.1 What is polymorphism ☝
In simple terms, Polymorphism refers to the ability of an object to exhibit

behaviors associated with different types. In other words, if the same object
behaves differently depending on which "side" of that object you are looking at,
then that object is polymorphic . For example, if you model apples using an
Apple class that extends a Fruit class, then an apple can behave as an Apple as
well as a Fruit. Later on, if you have a RedApple class that extends the Apple
class, then a red apple will behave as an Apple and a Fruit besides behaving as a
RedApple. Similarly, an object of a StockPrice class that you saw earlier behaves
as a Readable as well as a Movable besides behaving as a StockPrice! Thus,
apples and stock prices are polymorphic objects.
It is very important to understand that the actual object doesn't change at all. The
object itself always remains the same. A red apple will always be a red apple. It
doesn't suddenly morph into an apple or a fruit. It has always been an apple and
a fruit besides being a red apple. It follows that if an object doesn't already
support a particular behavior, it won't suddenly start supporting that behavior. A
red apple will never morph into a green apple no matter what you do.
Remember that when we talk about the behavior of an object in the context of
polymorphism, we are essentially talking about its instance methods. We are not
talking about its static methods because static methods define the behavior of a
class and not the behavior of the object. (Yes, it is true that you can access a
static method using an object reference instead of the class name but that is just a
peculiarity of the Java language and has nothing to do with polymorphism). We
are not talking about instance variables either because variables merely store
data (either primitive data or references to other objects) and do not possess any
behavior. Also, the variables would not be visible to any other class anyway if
the class is well encapsulated.
Polymorphism in Java ☝
Java allows objects to be polymorphic by letting a class extend another class

and/or implement interfaces. Since every class in Java implicitly extends
java.lang.Object , you can say that every object is polymorphic because
every object exhibits the behavior of at least two classes (except an instance of
java.lang.Object , of course).
Thus, an object can be as polymorphic as the number of classes it extends

(directly as well as indirectly) and the number of interfaces it implements
(directly as well as indirectly through ancestors).
Importance of Polymorphism ☝
In the "Kickstarter for Beginners" chapter I talked about how a component can
be easily replaced with another component if they promise to honor the same
behavior. This is made possible only because of polymorphism . If an object is
allowed to behave only as a single type, then you can never replace that object
with another object of a different type.
For example, if all you want is a Fruit, then it shouldn't matter whether you are
given an Apple or an Orange. Both of them promise to honor the behavior
described by Fruit and are, therefore, equally acceptable fruits. The situation is
illustrated by the following code:
abstract class Fruit{ //must be declared abstract beca

use it has an abstract method
abstract void consume();

}
class Apple extends Fruit{

void consume(){
System.out.println("Consuming Apple...");
}
}
class Orange extends Fruit{

void consume(){
System.out.println("Consuming Orange...");
}
}
class Person{
void eatFruit(Fruit f){
f.consume();
}
}
class TestClass{
Apple a = new Apple();
Orange o = new Orange();
p.eatFruit(a);
p.eatFruit(o);
}
}
Observe that a Person object has no knowledge about what it is eating. All it
cares about is Fruit. As long as you pass it a Fruit, it is fine. You can easily pass
an Orange to a Person instead of an Apple. This is possible only because Apple
and Orange are polymorphic, they behave like a Fruit besides behaving like an
Apple or an Orange respectively.
Without polymorphism, your code would look something like this:
class Apple{
void consume(){
System.out.println("Consuming Apple...");
}
}
class Orange{
void consume(){
System.out.println("Consuming Orange...");
}
}
class Person{
void eatApple(Apple a){
a.consume();
}
void eatOrange(Orange o){
o.consume();
}
}
class TestClass{
Orange o = new Orange();
p.eatApple(a);
p.eatOrange(o);
}
}
Observe that in the absence of a common Fruit class, Person must have two
different methods - one for eating Apple and one for eating Orange. The user of
the Person class, i.e., TestClass, also has to call two different methods. What if
you wanted to feed a banana to a person? You would have to code a new method
in Person to accept a Banana and also have to change the code in TestClass. And
you know these are not the only kinds of fruit in the world, right? You get the
idea. Imagine how many eat methods you would have to write!
But that is not the only problem. Let us say you code a method for each of the
fruits that you know about in Person class and deliver this class to other people
for use. What if you or someone else comes to know about an entirely new kind
of fruit and wants to feed that fruit to a Person object? You can't change the code
of the Person class at this stage because then you would have to make everyone
using the old class get this new version of the Person class from you. Depending
on the diversity of people in different projects, this could become a nightmare
every time a new fruit is discovered.
If you think logically, it shouldn't matter to a Person if a new kind of fruit is

discovered. As long as it is a fruit, a Person should be able to accept it. This is
where Polymorphism shines. In our first design Person 's eat method accepted
just a Fruit . If you find a new fruit, say Custard Apple (amazing fruit, btw),
all you have to do is to define a CustardApple class that extends Fruit and
pass objects of this class to Person 's eatFruit(Fruit ) method. No
change is required in the Person class. This is possible only because
CustardApple honors the contract defined by Fruit by saying it extends
Fruit .
Another important thing about Polymorphism is that it allows components to be

switched not only at compile time but also at run time . You can actually
introduce completely new classes that extend existing classes (or implement
existing interfaces) into an application that is already running. An existing class
or interface serves as a contract for a specific behavior and if a new class
promises to honor this contract by extending that class or by implement that
interface, then any code can make use of instances of this new class in places
where it requires instances of the existing class or interface.
Always remember that the whole objective of polymorphism is to enable classes

to become standardized components that can be easily exchanged without any
impact on other components. The various seemingly confusing rules that you
will soon see, only exist to achieve this goal. If you ever get confused about a
particular rule, always think about how it affects the interoperability of one
component with another.
The ability to transparently replace an object with

another is called "substitutability". The substitutability
principle states that objects of a type may be replaced
with objects of its subtype without altering any of the
desirable properties of the program.
Polymorphism and Inheritance ☝
As you saw above, methods of a class or an interface basically form a contract

that a subclass or the implementing class promises to fulfill. When a class says
that it extends another class or says that it implements an interface, it agrees to
have the methods the superclass or the interface has. However, the exact things
that these methods must do is not covered by this contract. In other words, the
contract is not so detailed as to affect the internal logic of the methods. The
internal logic of a method is only governed by an informal contract that is
implicit in the name of a method. For example, if the superclass declares a
method named computeInterest , it is reasonable to expect that the
subclass' implementation would compute some kind of interest in this method.
This is not an issue with abstract methods because a subclass is required to
implement abstract methods on its own, but it may be a problem in case of non-
abstract methods. What if the superclass computes interest in one way but the
subclass wants to compute interest in another way? If you think about it, this is
an important aspect of componentizing a class. You want to componentize
classes so that you could easily replace one component with another. But why
would you replace one component with another if they behaved exactly the
same? You wouldn't.
What you really want is the ability to replace a component with another
component with which you can interact in the same way, but which behaves
differently. This is exactly what polymorphism in Java achieves. Java allows a
subclass to provide its own implementation of the method if it does not like the
behavior provided by the inherited method. The technical term for this is
"overriding" . A subclass is free to override a method that it inherits with its
own implementation.
12.4.2 Overriding methods ☝
As discussed earlier in this chapter, only non-private instance methods can be

overridden . There are several rules that you need to follow while overriding a
method. These rules govern the accessibility , return type , parameter types ,
and the throws clause of the method. The best way to understand and remember
these rules is to keep in mind that these rules are there to make sure that an
object of a class can be replaced with an object of a subclass without breaking
existing code. The objective is to be able to replace one component with another
transparently while giving flexibility to a subclass in implementing its methods.
Whenever you override a method in a subclass, think about what will happen to
the code that depends on the superclass object and if you pass it a subclass object
instead of the superclass object. If this replacement does not require the code to
be recompiled, then you have overridden the method correctly. From this
perspective, the following rules are applicable for method hiding (which happens
for non-private static methods) as well.
Let me now list all of the rules and then I will show you the reasoning behind
these rules with an example. (Remember that an "overridden method" means the
method in the superclass and an "overriding method" method means the method
in the subclass).
1. Accessibility - An overriding method must not be less accessible than the

overridden method. This means that if the overridden method is
protected , you can't make the overriding method default because
default is less accessible than protected . You can make the overriding
method more accessible. Thus, you can make it public .
2. Return type - The return type of the overriding method must either be the
same as the return type of the overridden method or it must be a sub type.
For example, if the return type of the overridden method is Fruit , then
the return type of the overriding method can either be Fruit or any
subclass of Fruit such as Apple or Mango . This is called "covariant
returns ".
In the case of primitives, the return type of the overriding method must
match exactly to the return type of the overridden method. This is because
there is no sub type relation between any two primitive types. For example,
even though a short is smaller than an int , neither one is a subtype of
the other.
3. parameters - The list of parameters that the overriding method takes must
match exactly to that of the overridden method in terms of types and order
(parameter names don't matter). Indeed, if there is a difference between the
types and/or order of parameters, it would not be an override but an
overload. I have already discussed overloading of methods.
4. throws clause - An overriding method cannot put a wider exception (i.e. a

superclass exception) in its throws clause than the ones present in the
throws clause of the overridden method. For example, if the overridden
method throws IOException , the overriding method cannot throw
Exception because Exception is a superclass of IOException .
The overriding method may throw a subclass exception such as
FileNotFoundException .
The overriding method cannot throw a new exception that is not listed in
the throws clause of the overridden method either.
The overriding method may decide to not have a throws clause altogether
though.
Note that this rule applies only to checked exceptions because these are the
only ones the compiler cares about. There is no rule regarding unchecked
exceptions.
Let's see how these rules works with respect to the following classes:
Number computeInterest(double principle, double yrs
, double rate) throws Exception {
if(yrs<0) throw new IllegalArgumentException(
"yrs should be > 0");
}
}
class Account{
double balance; double rate;
Account(double balance, double rate){

this.balance = balance;
this.rate = rate;
}
double getInterest( InterestCalculator ic, double y

rs ){
try{
Number n = ic.computeInterest(balance, yrs,
rate);
return n.doubleValue();
}catch(Exception e){
e.printStackTrace();
}
return 0.0;
}
}
class AccountManager{
Account a = new Account(100, 0.2);
InterestCalculator ic = new InterestCalculator()
;
double interest = a.getInterest(ic, 2);
}
We have an Account class that uses an InterestCalculator object to

compute interest. We also have an AccountManager class that manages
accounts and also the way accounts compute interest. AccountManager
knows the interest calculation logic that accounts must use to compute interest.
As of now, AccountManager creates an InterestCalculator object
and passes it on to an Account object to compute interest.
Our objective is to make an Account return compound interest. To achieve this,

all we need to do is to create a subclass of InterestCalculator named
CompoundInterestCalculator and pass an instance of this new class to
Account 's getInterest method from AccountManager 's main :
public class CompoundInterestCalculator extends Intere

stCalculator {
public Double computeInterest(double principle, dou
ble yrs, double rate){
inciple;
}
}
class AccountManager{
Account a = new Account(100, 0.2);
InterestCalculator ic = new CompoundInterestCalc
ulator();
double interest = a.getInterest(ic, 2);

}
}
Observe that the Account class has no idea about the change that we did.
The Account class does not need recompilation because it does not even know
about the existence of CompountInterestCalculator . It interacts with
the CompountInterestCalculator object as if it were just another
InterestCalculator object.
Now, let's see how violating the rules of overriding affects Account class.
1. Accessibility - If you make computeInterest in

CompountInterestCalculator private, the JVM will be facing two
contradictory directions. The call to ic.computeInterest(...) in
Account expects the JVM to invoke the method on the actual object
referred to by ic . But the type of the actual object is
CompoundInterestCalculator and
CompoundInterestCalculator expects the JVM to reject any
attempt to invoke this method from outside the
CompoundInterestCalculator class. There is no way to obey both
the directions. Therefore, the compiler rejects the restriction on making this
method private.
Making the method public in CompoundInterestCalculator is ok

because the Account class doesn't care if classes from other packages are
able to access CompoundInterestCalculator 's
computeInterest .
2. Return type - When Account 's getInterest calls

ic.computeInterest , it expects to get back a Number . So, when
CompoundInterestCalculator 's computeInterest returns a
Double , Account has no problem because a Double is-a Number .
But if you change the return type of CompoundInterestCalculator
's computeInterest to, say, Object , code in Account 's
getInterest will fail because an Object is not necessarily a Number
and it won't be able to call doubleValue on an Object .
3. parameter types - This is kind of obvious. If you change the parameter
types, the compiler will consider this a completely different method and not
a replacement for the method in InterestCalculator . The JVM will
never invoke this new method and Account will still be computing simple
interest instead of compound interest.
4. throws clause - Since the computeInterest method of

InterestCalculator says that it may throw an Exception , the
caller of this method, i.e., Account 's getInterest is prepared to deal
with this exception. It has a catch block that is meant to catch
an Exception . The code in getInterest will work even if the
getInterest method throws IllegalArgumentException
because IllegalArgumentException is an Exception and it will
be caught by catch(Exception ) clause. It doesn't have any problem
if computeInterest doesn't throw any exception at all either. But if the
overriding method decides to throw a super class exception such as
Throwable , then the code in Account 's getInterest will fail
because it is not prepared to handle a Throwable . You will have to
update the catch clause to catch(Throwable ) and recompile
Account .
I suggest that you write a similar example to validate the above rules in case of
hiding of a static method.
12.4.3 Invalid overrides ☝
Overriding a static method with an instance method and vice versa ☝
Java does not allow a subclass to change the static/instance type of a non-private
method defined in the superclass. Thus, the following code will not compile:
class Foo{
static void foo(){
}

void moo(){
}
}
class Bar extends Foo {

void foo(){ // will not compile because overridden
method is static
static void moo(){ // will not compile because over

riding method is static
Overriding private methods ☝
Since private methods are not inherited by a subclass, it is not possible to

override them. But a subclass is allowed to have a method with the same
signature as a private method of a superclass.
Overriding final methods ☝
By marking a method as final , you prohibit subclasses from overriding that

method. Therefore, if you try to have a method with the same signature as a final
method of a superclass, it will not compile. The final keyword works
similarly for static methods as well. It prevents you from hiding a static method
in the subclass.
Note that final works very differently in case of fields. A subclass is allowed to
have a field by the same name even if that field is declared final in the
superclass. As discussed earlier, final just prevents you from the changing the
value of that field.
12.5 Utilize polymorphism to cast and call methods
12.5.1 Type of reference and type of an object ☝
In the 'Kickstarter for Beginners' chapter I discussed the difference between a

reference and an object in detail. I used the analogy of a remote and a TV to
explain that a reference is a handle with which you access an object. Just as the
remote and the TV it controls are two different things, the reference and the
object to which the reference refers are also two different things.
The type of the variable is specified at the time of defining the variable. For
example, String str; defines a variable str of type String . Specifying
the type of the variable is important because that is how the compiler is able to
determine the variables and methods that you are allowed to access through that
variable. This is what makes Java a strongly typed language. A Java compiler
will not allow you to invoke methods that are not available in the type of the
reference that you are using.
An object, on the other hand, is created at run time. Although the compiler may
sometimes be able to guess the type of an object that will be created at run time
(for example, when you instantiate a class using the new keyword), it is not
always the case. For example, while compiling the following method, the
compiler has no way of knowing the type of actual object that is referred to by
the method parameter obj :
public class Test{

public void foo(Object obj){
//do something with obj
}
}
The compiler does not know who may be calling foo and what they may be
passing to it as an argument. All it knows is that obj will refer to an Object
but it has no idea about the exact type of that object. For example, it is possible
for another developer to write a method in another class that calls Test 's foo
with a String or an Integer as an argument:
public class Other{

public void bar(){
new Test().foo("hello");
new Test().foo(new Integer(1));
}
}
The above example actually shows polymorphism in action. The variable obj in
foo could potentially behave like a String , an Integer , or like any other
subclass of Object depending on the actual class of the object that is passed to
foo .
12.5.2 Bridging the gap between compile time and run time ☝
To write polymorphic code, it is critical to understand that polymorphism

happens at run time i.e. at the time of execution of the code, but the code that
makes it happen is written at compile time . There is a difference between the
amount of information that the compiler has about the program and the amount
of information that the JVM has while it is executing that program. For example,
when you define a variable, the compiler only knows the type of that variable,
but it cannot know the exact object to which it refers because objects are created
only when the program is run, i.e., at run time. Now, Java is a strongly typed
language, which means that the type of the variable is defined at compile time
itself and it cannot be changed once defined. This means, the compiler must
make sure that the code does not assign an object of one type to a variable of
another type. It must not allow the code to invoke methods that are not supported
by the type of the variable. However, relying only on the type information
available at compile time will impose too many restrictions on the code. For
example, if you define a reference variable of type java.util.List , the
compiler must not allow the code to invoke methods defined in some other class
using that variable. However, this rigidity will make polymorphism impossible
because the whole premise of polymorphism is that an object may behave
differently at run time than what was promised at compile time.
You can see that there are two contradictory requirements that the compiler
needs to fulfill. On one hand, you want the compiler to type check the code to
make sure the code doesn't invoke random methods using a reference and on the
other hand you want the compiler to let code violate type checking in some
cases. Java tries to meet both the objectives by letting the compiler make certain
assumptions and by letting the programmer give guarantees to the compiler that
it is not trying to do anything fishy.
Let us now see how these two things are done in practice.
The "is a" test ☝
The is-a test is an intuitive test to determine the relationship between two
reference types. It allows you the check whether there is a parent child
relationship between two reference types and if there is, which one of the two
types is the parent class, and which one is the child class. For example, it makes
sense if you say that a Dog is a Pet or a Cat is a Pet but not if you say that a Dog
is a Cat. A similar relationship exists between an Apple and a Fruit or a Mango
and a Fruit. An Apple is a Fruit, A Mango is a Fruit, but an Apple is not a
Mango. If you were to model Apple, Mango, and Fruit as classes, you would put
all common features of fruits in a Fruit class and you would make Apple and
Mango classes extend Fruit . The importance of establishing this relationship
is that it allows you to use a subclass object anywhere in place of a superclass
object. Indeed, if you promised to give someone a fruit, you can certainly give
them an apple because an apple is a fruit. The reverse is obviously not true. If
you promised to give someone an apple, you cannot just give them any kind
fruit. It has to be an apple. But if you had a Macintosh apple, you could give
them that because a Macintosh apple is an apple.
The Java compiler recognizes if an is-a relationship exists between two types
and allows you to assign a subclass object to a reference of a superclass type. For
example:
class Fruit{ }
class Apple extends Fruit{ }
class Mango extends Fruit{ }
...
Mango m = new Mango();
Fruit f1 = a; //ok, because Apple is a Fruit
m = a; //will NOT compile because an apple is not a ma

ngo
Fruit f2 = m; //ok, because Mango is a Fruit
m = f1; //will NOT compile because all fruits are not

mangoes
m = f2; //will NOT compile because all fruits are not

mangoes
Note that the actual objects are created only at run time but the compiler knows
that the variable a will always refer to an Apple (i.e. to an object of a class that
satisfies the is-a test with Apple ) at run time because it would not compile the
code that tries to assign anything else to a . This is proven by the fact that the
line Mango m = a; does not compile. Similarly, the compiler knows that f1
and f2 will always point to a Fruit at run time because it will not allow you to
write code that tries to assign anything else to f1 and f2 .
The last line, i.e., m = f2; is important. Even though we know that f2 really
does point to a Mango , the compiler refuses to compile this line. Remember
that the compiler does not execute any code and therefore, the compiler doesn't
know that f2 really does point to a Mango . It only knows that f2 may refer to
any kind of Fruit . That fruit could be a Mango or an Apple . Since the
compiler cannot know for sure what f2 will point to at run time, it does not let
you assign f2 to m . If it allowed this line to compile, there would be a violation
of type safety during execution if f2 pointed to an Apple instead of a Mango
.
The same logic discussed above applies to interfaces as well. If a class

implements an interface, then an object of that class is-a that interface. For
example, if Fruit implements an interface named Edible , then any fruit will
satisfy the is-a test for Edible . In other words, a Fruit is-a Edible .
The cast Operator ( ) ☝
You saw above that the compiler does not let the lines Mango m = f1; and
Mango m = f2; compile because the compiler cannot know for sure that f1
and f2 will point to Mango objects at run time. Instead of letting an Apple get
assigned to m and thereby violating type safety, it rejects the code altogether.
However, the programmer knows what she expects f2 to refer to at run time
(because she wrote the code after all!). Java language allows a programmer to
use their knowledge about their program to assure the compiler that a reference
will point to an object of the correct type at run time using the cast operator. This
assurance convinces the compiler to let them do the assignment. Here is how this
works:
m = (Mango) f1;
m = (Mango) f2;
By casting f1 and f2 to Mango , the programmer basically guarantees to the

compiler that f1 and f2 will point to Mango objects at run time. The compiler
accepts the guarantee because it knows that Mango is a Fruit and it is possible
for a variable of type Fruit to refer to a Mango . The compiler rejected the
code earlier because the variables could point to Apple objects as well but with
the explicit guarantee given by the programmer that they will point to a Mango
objects at run time, the compiler accepts the code.
The java.lang.ClassCastException ☝
If you observe the above code carefully, you will know that f1 will not really
point to a Mango at run time. The line Fruit f1 = a; actually makes f1
point to the same object as a and a points to an Apple . Therefore, f1 will
actually point to an Apple and not a Mango at run time. So basically, we just
fooled the compiler into accepting the code by making a false guarantee. Well,
we did fool the compiler but we can't fool the JVM. The JVM knows what kind
of object f1 actually points to and it will not let the cast of f1 to Mango
succeed. The JVM will throw a ClassCastException when it executes this
line.
The JVM is essentially the second line of defence against violation of Java's type
safety. It checks at run time what the compiler is unable to check at compile
time. If the JVM sees that a reference variable is being cast to a type that does
not satisfy the is-a test for the class of the object to which the variable is
referring, it will throw a ClassCastException . Thus, the type safety of the
program is never compromised.
Fooling the compiler in this case doesn't mean that you can fool the compiler by
giving a false guarantee every time. Take a look at this code:

Apple a = (Apple) m;
The compiler rejects the above code in spite of you guaranteeing that m will
point to an Apple at run time. Well, it turns out that compiler is not completely
clueless. It knows that there is absolutely no way m can point to an Apple
because the declared class of m is Mango and there is no is-a relationship
between Mango and Apple . It knows that anything that is a Mango can never
be an Apple and so it calls your bluff. Thus, for a cast to pass compilation, the
cast must at least be plausible!
Casting a reference to an interface ☝
It is easy for a compiler to determine whether a reference of one class can ever
point to an object of another class because a class can only extend one class at
most. Therefore, if the class mentioned in the cast is a subclass of the declared
class of the reference, the compiler knows that it the cast will probably succeed
at run time.
It is not so easy with interfaces. Since a class can extend one class and can also
implement any number of interfaces at the same time, the compiler cannot rule
out cases where a reference can never point to an instance of a class that
implements the interface mentioned in the cast. Here is what I mean:
interface Poisonous{ }
class TestClass{
Fruit f = new Mango();//ok, because Mango is-a
Fruit
Poisonous p = (Poisonous) f; //compiles fine bu

t throws a ClassCastException at run time
}
}
The compiler accepts the casting of f to Poisonous even though it knows that
Fruit does not implement Poisonous . The reason is that the compiler also
knows that even though Fruit does not implement Poisonous , there could
be a subclass of Fruit that implements Poisonous and since that subclass
would be a Fruit , f could potentially point to an object of a class that
implements Poisonous . Something like this:
class StarFruit extends Fruit implements Poisonous{ }

...
Fruit f = new StarFruit();
Poisonous p = (Poisonous) f; //compiles and runs fine
The above reasoning implies that you can cast any type of reference to any
interface. That's true except in one case. If the declared class of the variable is
final and if that class does not implement the interface given in the cast, the
compiler knows that this class can never have any subclass and therefore, it
knows that there is no way the reference can point to an object of a class that
extends this class and also implements the given interface. For example,
String is a final class and the following code will not compile for the same
reason:
String s = new String();

Poisonous p = (Poisonous) s; //will not compile
Downcast vs Upcast ☝
Casting a variable of a type to a subtype (e.g. casting the variable f of type Fruit
to Apple) is called "downcasting". It is called downcasting because in a UML
diagram a subclass is always drawn below the superclass. Also, a variable of
type the Fruit may refer to any kind of fruit but when you cast it to Apple, you
have essentially reduced the possibilities for the kind of fruits that the variable
could be referring to. In other words, when you cast a variable to a subtype, you
are narrowing the type of the object to which this variable is pointing, down to a
more specific type. Hence, the term "narrowing". It means the same as
downcasting. The opposite of downcasting is "upcasting" or "widening".
As explained above, downcasting always requires a check by the JVM to make

sure that the variable is really pointing to an object of type that the programmer
has claimed it is pointing to. Upcasting, on the other hand, requires no such
check and is, in fact, almost always redundant (I will show you the one situation
where it is not redundant soon). Since an Apple is-a Fruit, you can always assign
a variable of type Apple to a variable of type Fruit without any explicit cast.
12.5.3 When is casting necessary ☝
In the previous section I used the analogy of a remote and a TV to show you that
the type of a variable and the type of the actual object referred to by that variable
are two different things and may not necessarily be the same.
Let me continue with the same analogy to show you how you can deal with the
difference between the type of a variable and type of the object referred to by the
variable. Imagine you have the remote for an old model that has a limited
number of buttons and you are using this remote to control a TV from a new
model that has a lot of functions. Obviously, you will only be able to use those
functions for which there are buttons in the remote. Even though the TV
supports many more functions, you cannot use them with the old remote because
the old remote has no knowledge of the new functions.
The reference and the object behave the same way. The type of the reference is
like the model of the remote and type of the object is like the model of the TV.
The following code makes this clear:
Object obj = "hello";

int h = obj.hashCode(); //ok because hashCode is defin
ed in Object
int i = obj.length(); //will not compile

int j = str.length(); //OK
In the above code, the declared type of the reference variable obj is Object
and the type of the object to which it refers to at run time is String . Since
String is a subclass of Object , String is-a Object and therefore, it is ok
to assign a String object to obj .
You can say that String is a kind of new model of Object and it has several
new features in addition to all the features of Object . However, since the
declared type of obj is Object , the compiler will only let you use the
functionality that is supported by Object because the compiler does not know
that obj will actually point to a string at run time.
If you want to use the new features of String class using obj , you will need to
cast it to String using the cast operator:
Object obj = "hello";

String str = (String) obj; //cast obj to String
int i = str.length();
int j = ((String) obj).length(); //casting and access

ing at the same time
Thus, the answer to the question when is casting necessary is simple. You need
casting when you want to use the features (i.e. instance variables and methods)
defined in a subclass using a reference whose declared type is of a superclass.
Remember that casting doesn't change the actual object. The purpose of casting
is to provide the compiler with the type information of the actual object to which
a variable will be pointing to at run time. Thus, casting just changes the
perspective from which the compiler views the object.
Impact of casting on static members ☝
Casting is essentially an aspect of object-oriented programming while "static" is

not. Nevertheless, due to the peculiarity of the Java language, it is possible to
access static members of a type through a variable of that type. Recall from the
"Working with Data types" chapter that static members of a type "shadow" the
static members of the same name in the super type. Casting a variable to the
super type is how you "unshadow" those members. Here is an example:
class Fruit{
static int count = 5;
static int getCount(){ return count; }
}
static int count = 10; //shadows Fruit's count
static int getCount(){ //shadows Fruit's getCount

return count;
}
}
class TestClass{
System.out.println(a.count); //prints 10;
System.out.println(a.getCount()); //prints 10;
System.out.println( ((Fruit) a).count ); //prin

ts 5;
System.out.println( ((Fruit) a).getCount() ); /

/prints 5;
Fruit f = a; //observe that no cast is needed he

re because we are widening
System.out.println( f.count); //prints 5;
System.out.println( f.getCount() ); //prints 5;
}
}
As promised, the above code shows that an upcast is not always redundant. But I
cannot stress enough that the above code is completely unprofessional. You
should never access static members through a variable.
While understanding Casting is important, it is used

sparingly in professionally designed code because it is,
after all, a means of avoiding the type safety
mechanism of the compiler. When you cast a reference
to another type, you are basically saying that the
program does something that is not evident from the
code itself and you are browbeating the compiler into
accepting that code. This reflects a bad design. Ideally,
you should almost never need to use casting.
12.5.4 The instanceof operator ☝
The instanceof operator is not on the OCP Part 1 exam

but it is important to know about this operator to
develop a good understanding of polymorphism. In fact,
the cast operator, which is on the exam, and the
instanceof operator often go together and that is why I
am going to talk about it in detail.
Since the actual type of the object which a variable is pointing to is not always
evident from the code, Java has the instanceof operator for the programmer
to check whether the actual object is an instance of a particular reference type
(i.e. class, interface, or enum) that the programmer is interested in. It takes two
arguments - a reference on the left-hand side and a type name on the right-hand
side. It returns a boolean value of true if the object pointed to by the variable
satisfies the is-a test with the type name given and false otherwise. For
example, if you are passed a Fruit as an argument to a method, here is how
you can use instanceof to know whether you have actually been given a
Mango :
class Fruit{ }
class Mango extends Fruit{ }
class Apple extends Fruit{ }
public class Juicer{

public void crush(Fruit f){
if(f instanceof Mango
){
System.out.println("crushing mango...");
} else {
System.out.println("crushing some other fruit
...");
}
}

new Juicer().crush(m);
}
}
In the above code, the method crush doesn't know the type of the actual object
that is referred to by f , but using the instanceof operator you can check if
that object is-a Mango .
The instanceof operator is handy when you want to treat an object of a

particular kind differently. For example, if Mango had a method
named removeSeed and if you want to invoke that method before crushing it,
you would want to know whether you have actually been given a Mango or not
before you cast your Fruit reference f to Mango , otherwise, you will get a
ClassCastException at run time if f does not refer to a Mango . Here is
how this can be done:
class Mango extends Fruit{

public void removeSeed(){ }
}
class Juicer{
public void crush(Fruit f){
if(f instanceof Mango){

Mango m = (Mango) f;
m.removeSeed();
}
System.out.println("crushing fruit...");
}

...
}
Remember that you can't invoke removeSeed on f because the type of f is

Fruit and Fruit doesn't have removeSeed method. Mango does.
Therefore, as discussed before, you must cast f to Mango before you can invoke
a method that is specific to a Mango .
There are a couple of things that you should understand clearly about
instanceof :
1. The instanceof operator cannot tell you the exact type of the object
being pointed to by a variable. It can only tell you whether that object is-a
something. For example, if you do f instanceof Fruit , it will
return true if the object referred to by f is-a Fruit , which means it will
return true even if f points to a Mango because a Mango is-a Fruit . In
the case of interfaces, it will return true if the class of the object pointed
to by the reference implements the given interface (directly or indirectly).
2. The compiler will let you use instanceof operator only if it is possible
for the variable to refer to an object of the type given on the right-hand side.
For example, f instanceof Mango is valid because the compiler
knows that the declared type of f is Fruit and since Mango is a Fruit ,
it is possible for f to point to a Mango . But f instanceof String
will not compile because the compiler knows that there is no way f can
ever point to a String . The instanceof operator behaves the same
way as the cast operator in this respect.
Just like the cast operator, the instanceof operator is

also used only sparingly in professionally written code.
Usage of instanceof reflects a bad design and if you
feel the need to use instanceof operator in your code
too often, you should think about redesigning your
application.
12.5.5 Invoking overridden methods ☝
If I haven't hammered it in enough yet, let me say this again - polymorphism is

all about the ability to replace one object with another without the need to
recompile existing code as long as the objects stick to an agreed upon contract.
For example, if a method requires an object of a class as an argument, then it
should work well with an object of its subclass. If a method expects an interface
as an argument, the class of the object that you pass to it shouldn't matter as long
as that class implements the interface. This flexibility makes it easier, and
thereby cheaper, to develop and maintain an application.
In technical terms, you must remember that polymorphism works only because
of dynamic binding of methods calls. When you invoke an instance method
using a reference variable, it is not the compiler but the JVM that determines
which code to execute based on the class of the actual object referenced by the
variable.
Some languages let the programmer decide whether they want to let the compiler
bind a method call to the version provided by the declared class of the variable
or to let the JVM bind the call at run time based on the class of the object
referred to by the variable. In such languages, methods that are not bound by the
compiler are called "virtual methods" . Java does not give the programmer the
ability to customize this behavior. In Java, calls to non-private and non-final
instance methods are bound dynamically by the JVM and are therefore, always
"virtual" . Everything else is bound statically at compile time by the compiler.
To take advantage of polymorphism, it is advisable to use interfaces and non-

final classes as method parameter types. For example, the interest calculator
example that I showed earlier could be redesigned as follows:
interface InterestCalculator{
//interface methods are public by default
double computeInterest(double p, double r, double

t);
}
class Account{
double balance, rate;
double getInterest(InterestCalculator ic, double t

ime){
return ic.computeInterest(balance, rate, time)
;
}
}
class SimpleInterestCalculator implements InterestCalc

ulator{
//must be public because an overriding method must

not reduce accessibility
public double computeInterest(double principle, dou

ble yrs, double rate) {
}
}
class CompoundInterestCalculator implements InterestCa

lculator{
public double computeInterest(double principle, dou
ble yrs, double rate) {
return principle*Math.pow(1 + rate, yrs) - prin
ciple;
}
}
Observe that the getInterest method now takes an

InterestCalculator as a parameter. This makes it very easy for any other
class to compute any kind of interest on an account object. As of now, there are
two classes that compute interest - SimpleInterestCalculator and
CompoundInterestCalculator , but in future if you want to compute
interest with different compounding, all you have to do is create a new class that
implements InterestCalculator interface and pass an object of this class
to the getInterest method. The Account class will not know the difference.
The JVM will bind the call to computeInterest to the code provided by
your new class automatically.
Dynamic binding of method calls may cause unexpected results if you are not
careful. Consider the following code:
class Account{
double balance = 0.0;

Account(double balance){
this.balance = balance;
this.printBalance();
}

void printBalance(){
System.out.println(balance);
}
class DummyAccount extends Account{

DummyAccount(double b ){
super(b);
}

public void printBalance(){
System.out.println("No balance in dummy account")
;
}
}

Account a = new DummyAccount(100.0);
}
}
Can you guess what it prints?

You may expect the call to this.printBalance(); in Account 's
constructor to be bound to Account 's printBalance method but observe
that this method is overridden by DummyAccount . Since the class of the actual
object is DummyAccount , the JVM will bind the call to DummyAccount 's
printBalance instead of Account 's printBalance . Therefore, it will
print "No balance in DummyAccount" .
This example shows that you need to be very careful about calling non-private
methods from a constructor. And by careful, I mean, don't do it :) A constructor
is meant to initialize the object's state variables to their appropriate values but if
you invoke a non-private method from a constructor, a subclass can easily mess
with your constructor's logic by overriding that method.
12.5.6 Impact of polymorphism on == and equals method ☝
You know that when used on references the == operator checks whether the two
operands point to the same object in memory or not. So, can you guess what the
following code will print?
String s = "hello";
Integer n = 10;
System.out.println(n == s);
false , you say? Well, the code won't even compile. The compiler applies the
same logic that it applies to the cast and the instanceof operators to check
whether it is even possible for the two reference variables to point to the same
object. It rejects the comparison if it is not. In the above code, there is no way s
and n can point to the same object because s and n are variables of two different
unrelated types. Thus, the compiler knows that this comparison is pointless and
is most probably a mistake by the programmer. Here is the same code but with a
small change:
Object s = "hello";
Integer n = 10;
System.out.println(n == s);
The above code indeed prints false . The compiler cannot reject the
comparison now because the type of s is Object and therefore, it is possible
for s to point to an Integer object (because Integer is-a Object).
The equals method, on the other hand, behaves differently. Remember that
equals is defined in the Object class and its signature is equals(Object
) . The type of the input parameter is Object and therefore, it must accept a
reference of any type. Thus, the compiler has no option but to accept even the
illogical invocations of the equals method such as "1234".equals(n); . It
is for the same reason that when a class overrides the equals method, the first
line of code in the method is usually an instanceof check:
class X{
int val;
public boolean equals(Object x){
if( ! (x instanceof X) ) return false;
//now compare the values of instance fields of thi

s and x and return true/false accordingly
return this.val == ((X) x).val

;
}
}
If you remember the rules of overriding, an overriding method is not allowed to

change the type of the input parameter to a narrower type. Thus, if a class tries to
override the equals method but changes the type of the input parameter from
Object to a more specific type, it will not be a valid "override". It will be a valid
"overload" though. For example, the following code will compile fine, but the
equals method does not override the equals method that X inherits from Object:
class X{
int val;
public boolean equals(X x){ //does not override bu
t overloads the equals method
return this.val == x.val;

}
X x1 = new X(); x1.val = 1;
X x2 = new X(); x2.val = 1;
System.out.println(x1.equals(x2)); //prints true
}
}
On the face of it, the equals method written above doesn't seem to make much
of a difference when you try to compare two X objects. But let's change the code
inside the main method as follows and see what happens:
X x1 = new X(); x1.val = 1;

Object x2 = new X(); ((X) x2).val = 1;
System.out.println(x1.equals(x2)); //what will i
t print?

If you have understood the rules of method selection that we discussed in the
"Working with Methods and Encapsulation" chapter, you should be able to
figure out that the above code will print false . The compiler sees two versions
of the equals method in class X to choose from when it tries to bind the
x1.equals(x2) method call - the version that class X inherits from
java.lang.Object , which takes Object as an argument and the version
that class X implements itself, which takes X as an argument. Since the declared
type of the variable x2 is Object , the compiler binds the call to the Object
version instead of the X version. The Object version returns false because
x1 and x2 are pointing to two different objects.
12.6 Distinguish overloading, overriding, and hiding

12.6.1 Overriding and Hiding ☝
You have seen how a class can inherit features by extending another class. These
features include static and instance fields as well as static and instance methods.
In the previous section, I explained how inheriting features is usually a good
thing because it allows a class to get functionality without writing any code. But
it could pose a problem if the subclass were not able to provide suitable behavior
to a method that it inherited.
For example, what if there is a SpecialInterestCalculator that extends

InterestCalculator and inherits a computeInterest method, but it
wants to change how the interest is computed by this method? Or what if it
wants to define a variable interestRate , but that variable is already defined
in InterestCalculator ?
Java has specific rules about what features can be tweaked and how they can be
tweaked by a subclass. These rules are categorized into two categories:
Overriding and Hiding . The rules of overriding are about polymorphism and
therefore, only apply to instance methods and the rules of hiding apply to
everything else, i.e., static methods as well as static and instance variables .
Remember that in both the cases, a member of a class has to be inherited in the
subclass first. Since the private members of a class are not inherited by a
subclass, the concepts of overriding and hiding are not applicable to them.
Similarly, constructors of a class are not inherited either and are, therefore, out of
the purview of overriding and hiding.
Overriding ☝
A class is allowed to completely replace the behavior of an instance method that

it inherited by providing its own implementation of that method. What this
means is that the behavior provided by the subclass is what will be exhibited by
any object of the subclass instead of the behavior provided by the super class.
For example:
yrs, double rate){
}
}

lator{

yrs, double rate){
inciple;//don't worry about Math.pow()!
In the above code, CompoundInterestCalculator has replaced the

implementation of computeInterest method provided by its super class
with its own implementation. If you call the computeInterest method on a
CompoundInterestCalculator object,
CompoundInterestCalculator 's version of the method will be called.
The following code proves it.
class TestClass{
InterestCalculator ic = new InterestCalculator(
);
double interest = ic.computeInterest(100, 2, 0
.1);
CompoundInterestCalculator cic = new CompoundIn

terestCalculator();
interest = cic.computeInterest(100, 2, 0.1);
}
}
Note that I am using the word replace to highlight the fact that it is not possible
for any other class to see the behavior of computeInterest method as
implemented by InterestCalculator class in a
CompoundInterestCalculator object because the behavior of the super
class has been replaced by the behavior provided by the subclass. Technically,
we say that CompoundInterestCalculator has "overridden"
computeInterest method of InterestCalculator . Let me make a
small change to the above code to make this point clear:
class TestClass{
InterestCalculator ic = new CompoundInterestCal
culator();
double interest = ic.computeInterest(100, 2, 0
.1);
}
}
In the above code, the declared type of the variable ic is

InterestCalculator but the actual object that it points to is of type
CompoundInterestCalculator . Therefore, when you call
computeInterest , CompoundInterestCalculator 's version of this
method is executed instead of InterestCalculator 's version.
The point to understand here is that in the case of instance methods, it is always
the method implemented by the class of the object that is invoked.
Hiding ☝
Hiding is a less drastic version of overriding. Like overriding, hiding lets a class
define its own version of the features implemented by its superclass, but unlike
overriding, hiding does not completely replace them with the subclass' version.
Thus, the subclass now has two versions of the same features and any unrelated
class can access both the versions. The following code illustrates this point:
public int yrs = 10;
public static double rate = 0.1;
public static String getClassName(){
return "InterestCalculator";
}
}

lator{
public int yrs = 20;
public static double rate = 0.2;
public static String getClassName(){
return "CompoundInterestCalculator";
}
}
In the above code, CompoundInterestCalculator inherits the instance

variable yrs , the static variable rate and the static method getClassName
from InterestCalculator . At the same time,
CompoundInterestCalculator defines all of these on it own as well. So,
now, CompoundInterestCalculator has two yrs variables, two rate
variables, and two versions of the getClassName method. Both the versions
can be accessed by an unrelated class as shown below:
class TestClass{
CompoundInterestCalculator cic = new CompoundIn
terestCalculator();

System.out.println(cic.yrs); //prints 20
System.out.println( ((InterestCalculator) cic).

yrs); //prints 10
System.out.println(cic.rate); //prints 0.2

rate); //prints 0.1
System.out.println(cic.getClassName()); //print
s CompoundInterestCalculator
getClassName()); //prints InterestCalculator
}
}
Observe that I used a special syntax to access the version provided by the
superclass. This special syntax is called a "cast" . The declared type of the
variable cic is CompoundInterestCalculator but I cast it to
InterestCalculator to go behind CompoundInterestCalculator
and access the versions provided by InterestCalculator . The cast tells
the compiler to treat a variable as if its declared type is the type mentioned in the
cast.
I will discuss the rules of casting later but the point to understand here is that
static methods, static variables, and instance variables are accessed as per the
declared type of the variable through which they are accessed and not according
to the actual type of the object to which the variable refers.
Contrast this with overriding, where type of the variable makes no difference to
the version of the instance method that is invoked.
12.7 Exercise ☝
1. You are developing an application that allows a user to compare
automobiles. Use abstract classes, classes, and interfaces to model Car ,
Truck , Vehicle , and Drivable entities. Declare and define a method
named drive() in appropriate places.
2. Every vehicle must have a make and model. What can you do to ensure that
a method named getMakeAndModel() can be invoked on every vehicle.
3. You need to be able to get the Vehicle Identification Number (VIN) of a
vehicle by calling getVIN() on any vehicle. Furthermore, you don't want
any subclass to change the behavior of the getVIN method. Where and how
will you code the getVIN method?
4. Create an interface named Drivable with a default method start() .
Invoke start() on instances of classes that implement Drivable .
Override start() so that it prints a different message in each class.
5. Ensure that every vehicle is created with a VIN.
6. Create a class named ToyCar that extends Car but doesn't require any
argument while creation.
7. You have a list of features such as height, width, length, power, and boot
capacity, on which you want to compare any two vehicles. New feature
names will be added to this list in future. Create a getFeature(String
featureName) method such that it will return "N.A" for any feature
that is not supported by a particular vehicle.
8. Create an interface named VehicleHelper with a static method
register(Vehicle v) that prints the VIN of the vehicle. Ensure that
VehicleHelper's register method is invoked whenever an instance of a
vehicle is created.
Chapter 13 Programming Abstractly
Through Interfaces
classes
13.1 Create and implement interfaces

13.1.1 Using interfaces ☝
An interface is used to describe behavior as captured by a group of methods. It

doesn't tell you anything about the object behind the behavior except that the
methods declared in the interface can be invoked on that object. This group of
methods therefore, is a way to interact with the object.
An interface is defined using the keyword "interface" . For example:
interface Movable{
void move(int x);
}
From a purely OOP perspective, an interface should not contain any

implementation. It should only contain method declarations. However, Java has
always permitted interfaces to contain static fields. Furthermore, from Java 8 ,
Java has permitted an interface to contain default methods and static methods
and Java 9 has allowed an interface to contain private methods (static as well as
non-static) as well.
There are several rules about interfaces that you must learn by heart. So, let me
go over them one by one.
Everything is public in an interface by default ☝
Everything declared inside an interface is implicitly public except methods that

are explicitly declared private . This means that members of an interface are
always public irrespective of whether you define them as public or not. In fact,
you are prohibited from defining them as private or protected. The following
example illustrates this point:
interface Movable{
void move1(int x); //OK, move1 is implicitly public
public void move2(int x); //OK, move2 is explicitly

public
private void move3(int x); //NOT OK, will not compil

e.
protected void move4(int x); //NOT OK, will not comp

ile.
int VALUE = 10; //OK, VALUE is implicitly public
private int PVT_VALUE = 10; //NOT OK, fields cannot

be protected or private
private void pvtMethod(){ }; //OK, methods can be pr

ivate (but not protected)
}
This makes sense because the whole purpose of an interface is to put out a way
to interact with an object in front of the world. Remember that an interface is not
about how an object is implemented. It is about how an object can be interacted
with. In that sense, an interface is a contract between an object and the rest of
the application components and since everything in a contract must be public,
everything in an interface must be public.
What if you want to disclose members of an interface only to the members of the
same package? Well, make the interface "default" . As discussed earlier, default
access allows something to be visible only to members of the same package. If
the interface itself is not visible outside the package, its members certainly won't
be.
Private methods do not really fit conceptually in an interface but the need for
private methods was felt after default methods were introduced in Java 8. If a
method gets too big or if there multiple methods with a lot of common code,
there was no way to refactor them into smaller methods without exposing all of
them to the world because everything in an interface had to be public. Java 9
fixes this problem by allowing private methods in an interface.
An interface is always abstract ☝
As mentioned above, an interface is merely a contractual description of some

behavior. It is not an entity that can actually fulfill that contract. For example,
you may describe the behavior of movement using an interface named Movable
but you need an object such as an animal or vehicle to exhibit that behavior.
Therefore, it does not make sense to instantiate an interface. For this reason, an
interface is implicitly abstract . Although legally valid, it would be redundant to
declare an interface as abstract.
Variable definitions in an interface ☝
All variables defined in an interface are implicitly public , static , and final .
The following example illustrates this point:
interface Movable{
int UNIT1 = 1;
static int UNIT2 = 1;
static final int UNIT3 = 1;
public static final int UNIT4 = 1;
}
All four of the variables declarations above are valid. All of the variables are
public , static , and final even though the first three have not been declared as
such. This implies that you cannot have instance variables in an interface.
Remember that instance variables are meant to store state , which means they
are really a part of implementation . Therefore, instance variables have no place
in an interface. To be honest, it is not a good idea to have static variables either
in an interface but for reasons best known to the language designers, Java allows
static constants in an interface. Many people use this feature to define "global
constants" but it is simply a bad design choice. Avoid them in your code.
Methods in an interface ☝
An interface can have four kinds of methods :
1. abstract methods - You have seen abstract methods in the previous section
about abstract classes. They contain just the declaration and no body. It is
the same thing here except that the keyword abstract is optional. For
example:
interface Movable{
void move1(int x); //implicitly abstract
abstract void move2(int x); //explicitly abstract

}
Both of the methods declared above are abstract.
2. default methods - Default methods are a way for an interface to include a

default implementation for a method. They are defined using the keyword
default as follows:
interface Movable{
default
void move(int x){

System.out.println("Dummy implementation. Mov
ing by "+x+" points");
}
}
Observe that default methods are just the opposite of abstract methods -
abstract methods cannot have an implementation while the whole purpose
of default methods is to provide an implementation. Therefore, a method
cannot be default as well as abstract at the same time. Thus, the following
code will not compile:
interface Movable{
//must be marked default or private because it

has a body
void move1(int x){

System.out.println("Dummy implementation. Mov
ing by "+x+" points");
}

void move2(int x); //ok, no body

}
3. static methods - As the name implies, static methods belong to the

interface itself and not to the object implementing that interface. They are
defined using the keyword static as follows:
interface Movable{
static
void sayHello(){
System.out.println("Hello!");
}
}
Since sayHello is a static method, you don't need an instance of any

object to invoke it. You can invoke it directly on the interface just like you
invoke static methods on a class. For example:

Movable.sayHello();
}
}
4. private methods - Private methods (static as well as no-static) have been

allowed since Java 9 and, as explained above, they are helpful when a
default static methods gets too big and needs to be refactored into smaller
internal methods without exposing the internal methods to the outside
world. Here is an example:
interface Movable{
private
void moveInternal(){ //don't want to make it acces
sible to others
System.out.println("in moveInternal");
}
default void move(int n){
while(n-- > 0) moveInternal();
}
}
Note that, since private methods are not visible outside the interface, they
cannot be marked default.
Since methods of an interface are meant to be implemented by the classes that

implement the interface, methods of an interface cannot be declared final.
Observe that this is opposite to the rule that is applied to variables. Variables of
an interface are always final.
Although not important for the exam, it is possible to

define classes, interfaces, and enums within an
interface. They cannot be anything other than public.
Marker interface ☝
You may encounter an interface that does not contain anything at all. Such
interfaces are called "marker interfaces". For example:
interface SpecialThing{
}
In the above code, SpecialThing is a marker interface.
The purpose of a marker interface is to tag a class with an extra piece of

information about that class itself. For example, if any class implements
SpecialThing , it implies that that class is a SpecialThing . This
information, also called "metadata", could be used by some code or some tool
that treats all SpecialThing s in a certain way.
The most common marker interface used in Java is

java.io.Serializable interface. It signifies to the JVM that objects of
classes implementing this interface can be serialized and deserialized.
You don't need to know about the term marker interface or the Serializable
interface for the Part 1 exam but it is often asked about in technical interviews.
13.1.2 Implementing an interface ☝
A class can implement any number of interfaces by specifying their names in its
implements clause. For example, in the following code the Price class
implements two interfaces:
interface Movable{
void move();
}
interface Readable{
void read();
}
class Price implements Movable, Readable{

public void move() { System.out.println("Moving...")
; }
public void read() { System.out.println("Reading..."

); }
}
Of course, once a class declares that it implements an interface, it must then have
the implementation for all of the abstract methods declared in that interface. It
could either implement the methods itself or inherit them from its ancestor class.
If an interface provides a default implementation for a method in the form of a
default method , the implementing class does not necessarily have to provide
implementation for that method.
If the class does not have implementation for even one of the abstract methods
declared in the interface that it says it implements, the class must be declared
abstract . Otherwise, the compiler will refuse to compile the class.
So, for example, if the StockPrice class does not provide an implementation
for the read method declared in Readable , it would have to be declared
abstract like so:
abstract
class StockPrice implements Movable, Readable{

public void move() { System.out.println("Moving...")
; }
}
A class is allowed to extend another class as well as implement any number of

interfaces at the same time. It this case, the class declaration will have an
extends clause as well as an implements clause. For example:
interface Printable{
void print();
}
public class StockPrice extends
Price implements
Printable{
public void print(){
System.out.println("Printing StockPrice...");
}
}
In the above code, StockPrice implements Printable explicitly and since

it extends Price , it implements Movable and Readable implicitly.
Furthermore, StockPrice inherits the move and read methods from Price
and therefore, it does not need to define them again.
Note that the order of extends clause and implements clause is important. The
extends clause must appear before the implements clause.
Inheritance of static methods of an interface ☝
Unlike the static methods of a class, the static methods of an interface cannot be
inherited. This difference is illustrated by the following code:
class Price{
static void m(){
System.out.println("In Price.m()");
}
}
interface Printable{
static void p(){
System.out.println("In Printable.p()");
}
}
class StockPrice extends Price implements Printable{
}
class TestClass{
StockPrice.m(); //works fine
StockPrice.p(); //will not compile
}
}
In the above code, since StockPrice extends Price , StockPrice inherits

the static method m() defined in Price . However, even though StockPrice
implements Printable , it does not inherit the static method p() defined in
Printable . That is why the call StockPrice.p() fails to compile.
Inheritance of multiple versions of a default method ☝
Since it is possible for a class to implement multiple interfaces, it is possible for

a class to inherit multiple implementations of a default method from more than
one interface. For example:
interface Task{
public default void doIt(){
System.out.println("Doing Task");
}
}
interface Activity{
public default void doIt(){
System.out.println("Doing Activity");
}
//will not compile
class Process implements Task, Activity{
In the above code, Process implements two interfaces and since each of them
contains a default method named doIt with the same signature, Process now
has two implementations for the same method. This is a problem because when
you call doIt method on a Process object, the JVM will not be able to
determine which implementation of doIt to invoke.
Java resolves this problem by forcing the class to provide an implementation of

the method of its own to remove the ambiguity in invocation.
public void doIt(){
System.out.println("Doing Process");
}
}
The JVM now has only one implementation of doIt method to invoke and so,
the above code compiles fine.
Let us look at a case that is even more interesting. Can you tell what happens
when you try to compile the following code:
interface Task{
int SIZE = 10;
default void doIt(){
System.out.println("Doing Task");
}
}
interface Activity{
long SIZE = 20;
void doIt();
}

Process p = new Process();
p.doIt();
}
}
Observe that doIt method is declared by both the interfaces but only one of
them provides an implementation. Thus, Process inherits only one
implementation for doIt method. There is no ambiguity for the JVM in
determining which implementation of doIt to invoke. Therefore, it should
compile, right? Wrong!
Java designers thought that even though there is only one implementation for the
doIt method, there is no guarantee that this implementation is appropriate for
the doIt method declared by another interface. The contract of the other
interface could possibly be different from the one that provides the
implementation for the doIt method. Therefore, it is better for the class to
explicitly provide an implementation of the doIt method.
Note that this issue does not arise when there is no default implementation
available at all because in that case the class will have to provide its own
implementation anyway.
Inheritance of multiple versions of a variable ☝
The situation with the SIZE variables in the above code is a bit special. Fields of
an interface are inherited by a sub class and therefore Process does get two
versions of SIZE variable. Java allows a class to inherit multiple fields with the
same name as long as you don't try to use those fields ambiguously. This means
that you will get a compilation error only if you try to use SIZE within
Process directly without specifying which SIZE field are you trying to refer
to. This is illustrated below:
interface Activity{
long SIZE = 20;
//void doIt(); //let's get rid of it to avoid the m
ethod ambiguity issue for now
}

System.out.println(SIZE);//will not compile
}
}
The above code will fail to compile with an error message, "reference to
SIZE is ambiguous " because the compiler is not able to figure out which
SIZE you are trying to use. But the following code will compile fine because
the compiler has no confusion about the SIZE that are you referring to here:

System.out.println(Activity.SIZE);
System.out.println(Task.SIZE);
}
}
13.1.3 Extending an interface ☝
It is possible for an interface to extend any number of interfaces. For example:
interface Readable{
int SIZE = 0;
void read();
}
interface Writable{
void write();
}
interface ReadWritable extends Readable, Writable{

//inherits SIZE and read() from Readable
//inherits write() from Writable
void delete();
}
Remember that a class cannot extend an interface, it can only implement an
interface. Whereas, an interface cannot implement any interface it can only
extend an interface. Thus, the following definitions of Writer will not compile:
interface Writer implements Writable{ //interface cann

ot "implement" any interface
public void write(){ }

}
class Writer extends Writable{ //class cannot "extend"

any interface
public void write(){ }

}
The extending interface inherits all the members except static methods of each
of the other extended interfaces.
Inheriting multiple versions of a default method ☝
It is possible for an interface to inherit a field or an abstract method with the

same signature from two of its super interface. But inheriting multiple default
methods or one default and one or more abstract methods with the same
signature has the same problem that you saw earlier in a class that implements
multiple interfaces. Java does not allow it. Here is an example:
interface Readable{
int SIZE = 10;
void read();
static void staticMethod(){

System.out.println("In Readable.staticMethod");
};
default void defaultMethod(){
System.out.println("In Readable.defaultMethod");

};
}
interface Writable{
int SIZE = 20;

void write();
static void staticMethod(){

System.out.println("In Writable.staticMethod");
};
/* commenting the following two methods out
default void defaultMethod(){

System.out.println("In Writable.defaultMethod");

};
void defaultMethod();
*/

//inherits SIZE, read(), and defaultMethod() from Re
adable
//inherits SIZE and write() from Writable

}
The above code compiles fine. But if you uncomment either of the
defaultMethod s in Writable , ReadWritable will fail to compile
because it would be inheriting two different implementations (or one
implementation and one declaration) of defaultMethod . The solution is the
same as what we did with the class. ReadWritable must provide its own
implementation of the defaultMethod to resolve the ambiguity:

default
void defaultMethod(){
System.out.println("In ReadWritable.defaultMetho
d");
};
}
Observe that staticMethod is also defined in both the super interfaces but it
does not cause any problem because static methods of an interface are never
inherited, which means ReadWritable does not get even a single
implementation of staticMethod from either of its superinterfaces, let alone
two, which would cause ambiguity.
Inheriting multiple variables with same name ☝
The situation with the SIZE variable (remember that it is implicitly static and
final) is, again, the same as what you saw in a class that inherits multiple
versions of a field from multiple interfaces. Static fields of an interface are
inherited by a sub interface and therefore ReadWritable does get two
versions of SIZE variable. Java allows an interface to inherit multiple fields
with the same name as long as you don't try to use those fields ambiguously. For
example, the following code will not compile :
int NEWSIZE = SIZE; //will not compile because SIZE
is being used ambiguously
The above code will fail to compile with an error message, "reference to
SIZE is ambiguous " because the compiler is not able to figure out which
SIZE are you trying to use. But the following code will compile fine because
the compiler has no confusion about the SIZE that you are referring to here:

int NEWSIZE = Readable.SIZE; //fine
13.1.4 Instantiating abstract classes and interfaces ☝

Abstract classes and interfaces are abstract. Objects of their kind do not exist,
which is why they are called abstract in the first place. Therefore, they cannot be
instantiated, period.
I have seen many students getting confused when they see code that looks like it
is instantiating an abstract class. For example:
abstract class Animal{

Animal a = new Animal(){ };
}
}
The above code seems to be instantiating Animal even though Animal is

abstract! The following is an example that seems to instantiate an interface:
interface Dummy{
}
Dummy d = new Dummy(){ };
}
}
Well, first thing that you should observe is the presence of { } between new
Animal() and ; . This is not the syntax for instantiating a class. For
instantiating Animal , you would have to write new Animal(); and if you
try to do that you will get a compilation error that says, "Animal is
abstract; cannot be instantiated ".
In fact, the above code uses the syntax for declaring as well as instantiating a
concrete inner class that extends Animal without giving this class a name.
Since no name is given to this class, it is called an anonymous class . When the
compiler sees this code, it actually creates a class, gives this class a weird
looking name, and generates a separate a class file for this class. It is this class
that is instantiated at run time. Since Animal class doesn't have any abstract
method, this anonymous class doesn't need to implement any method. Had
Animal had an abstract method, you would have to implement this method
inside the curly brackets.
The same thing is happening in the case of the interface example above. The
code is defining and instantiating in a single statement an anonymous class that
implements the Dummy interface.
Since the topic of inner classes (including anonymous inner class) is not on the
OCP Java 11 Part 1 exam, I will not discuss this any further in this book.
13.2 Distinguish class inheritance from interface
inheritance including abstract classes
13.2.1 Difference between Interface and Abstract Class ☝
"What is the difference between an interface and an abstract class " is usually
the first question that is asked in a Java "tech-check". While being a simple ice
breaker, it also an excellent question to judge a candidate's understanding of
OOP.
Candidates usually start parroting the technical differences such as an interface

cannot have method implementations (which it can, since Java 8), while abstract
class can. An interface cannot have static methods (which it can, since Java 8) or
instance fields while an abstract class can and so on.
All that is correct, but is not really impressive. The fundamental difference
between an interface and an abstract class is that an interface defines just the
behavior. An interface tells you nothing about the actual object other than how it
behaves. An abstract class, on the other hand, defines an object, which in turn,
drives the behavior.
If you understand this concept everything else about them will fall in place. For
example, "movement" is a behavior that is displayed by various kinds of objects
such as a Car, a Cat, or a StockPrice. These objects have no commonality except
that they move. Saying it the other way round, if you get an object that "moves",
you don't get any idea about what kind of an object are you going to deal with. It
could be a Car, a Cat, or a StockPrice. If you were to capture this behavior in
Java, you would use an interface named Movable with a method named move()
.
On the other hand, if you talk about Automobile, a picture of an object starts
forming in your head immediately. You can sense that an Automobile would be
something that would have an engine, would have wheels, and would move. You
intuitively know that a StockPrice or a Cat cannot be an Automobile even
though they both do move. An abstract class is meant exactly for this purpose,
when, once you identify a conceptual object, you do not need to worry about its
behavior. The behavior kind of flows automatically. If you create an abstract
class named Automobile, you are almost certain that it would have methods such
a move, turn, accelerate, or brake. It would have fields for capturing inner details
such Engine, Wheels, and Gears. You get all that just by saying the word
Automobile.
From the above discussion, it should be clear that interfaces and abstract classes
are not interchangeable. Even though an abstract class with no non-abstract
method looks functionally similar to an interface, both are fundamentally
different. If you are capturing a behavior, you must use an interface. If you are
capturing a conceptual object, you must use an abstract class.
13.3 Declare and use List and ArrayList instances

13.3.1 Introduction to Collections and Generics ☝
Processing multiple objects of the same kind in the same way is often a
requirement in applications. Printing the names of all the Employees in a list of
Employees, computing interest for a list of Accounts, or something as simple as
computing the average of a list of numbers, require you do deal with a list of
objects instead of one single object. You have already seen a data structure that
is capable of holding multiple objects of the same kind - array. An array lets you
hold references to multiple objects of a type and pass them around as a bunch.
However, arrays have a couple of limitations. First, an array cannot grow or
shrink in length after it is created. If you create an array of 10 Employees and
later find out you have 11 Employees to hold, you will need to create a new
array with 11 slots. You can't just add one more slot in an existing array. Second,
inserting a value in the middle of an array is a bit difficult. If you want to insert
an element just before the last element, you will have to first shift the last
element to the next position and then put the value. Imagine if you want to insert
an element in the middle of the list!
Although both of the limitations can be overcome by writing some extra code,
these requirements are so common that writing the same code over and over
again is just not a good idea. java.util.ArrayList is a class that
incorporates these, and several other, features out of the box. The following is an
example that shows how easy it is to manage a list of objects using an ArrayList:
import java.util.ArrayList;
ArrayList al = new ArrayList();
al.add("alice");//[alice]
al.add("bob");//[alice, bob]
al.add("charlie");//[alice, bob, charlie]
al.add(2, "david");//[alice, bob, david, charlie]
al.remove(0);//[bod, david, charlie]

for(Object o : al){ //process objects in the li
st
String name = (String) o;

System.out.println(name+" "+name.length());
}
//dump contents of the list
System.out.println("All names: "+al);

}
}
You will get a few warning messages when you compile the above program.
Ignore them for now. It prints the following output when run:
bob 3
david 5
charlie 7
All names: [bob, david, charlie]
The above program illustrates the basics of using an ArrayList , i.e., adding
objects, removing, iterating over them, and printing the contents of the
ArrayList . I suggest you write a program that does the same thing using an
array of strings instead of an ArrayList . This will give you an appreciation
for the value that ArrayList adds over a raw array.
I showed you the above code just to give you a glimpse of one of the readymade
classes in the Java library. There is a lot more that you can do with it but before
we get to that, you need to understand the bigger picture, the grand scheme of
things, into which classes such as ArrayList fit.
Collections and Collections API ☝
ArrayList belongs to a category of classes called "collections" . A

"collection" is nothing but a group of objects held up together in some form.
You can visualize a collection as a bag in which you put several things. Just like
you use a bag to take various grocery items from a store to your home, you put
the objects in a collection so that you can pass them around to other objects and
methods. A bag knows nothing about the items themselves. It merely holds the
items. You can add items to it and take items out of it. It doesn't care about the
order in which you put the items or whether there are any duplicates or how they
are stored inside the bag. The behavior of a bag is captured by an interface called
Collection , which exists in the java.util package.
Now, what if you want a collection that ensures objects can be retrieved from it
in the same order in which they were added to it? Or what if you want a
collection that does not allow duplicate objects? You can think of several
features that can be added on top of a basic collection. As you have already
learnt previously, subclassing/subinterfacing is the way you extend the behavior
of an existing class or interface. The Java standard library takes the same route
here and extends java.util.Collection interface to provide several
interfaces with various features. In fact, the Java standard library provides a huge
tree of interfaces along with classes that implement those interfaces. These
classes and interfaces are collectively known as the "Collections API" .
So what has ArrayList got to do with this, you ask? Well, ArrayList is a
class that implements one of the specialized collection interfaces called List .
java.util.List extends java.util.Collection to add the behavior
of ordering on top of a simple collection. A List is supposed to remember the
order in which objects are added to the collection. Okay, so that takes care of the
"List" part of ArrayList, what about the "Array" part? Again, as you have learnt
in previous chapters, an interface merely describes a behavior. It doesn't really
implement that behavior. You need a class to implement the behavior. In this
case, ArrayList is the class that implements the behavior described by
java.util.List and to implement this behavior, it uses an array. Hence the
name ArrayList . Remember I talked about writing code to overcome the
limitation of an array that it is not flexible enough to increase in size
automatically? ArrayList contains that code. When you add an object to an
ArrayList , it checks whether there is space left in the array. If not, it
allocates a new array with a bigger length, copies the elements from the old
array to this new array, and then adds the passed object to the new array.
Similarly, it contains code for inserting an object in the middle of the array. All
this is transparent to the user of an ArrayList . A user simply invokes
methods such as add and remove on an ArrayList without having any
knowledge of the array that is being used inside to store the object.
Generics ☝
Now, regarding the warning messages that I asked you to ignore. Observe the
for loop in the above code. The type of the loop variable o is Object (and not
String ). To invoke the length method of String on o , we need to cast o
to String . This is because, in the same way that a bag is unaware of what is
inside of it, so too does the ArrayList object have no knowledge about the
type of the objects that have been added to it. It is the responsibility of the
programmer to cast the objects that they retrieve from the list to appropriate
types. In this program, we know that we have added String s to this list and so
we know that we can cast the object that we retrieve from this list to String .
But what if we were simply given a pre-populated list as an argument? Since the
list would have been populated by another class written by someone else, what
guarantee would we have about the type of objects we find in the list? None,
actually. And as you are aware, if we try to cast an object to an inappropriate
type, we get a ClassCastException at run time. Getting a
ClassCastException at run time would be a coding mistake that will be
discovered only at run time. Discovering coding mistakes at run time is not a
good thing and the compiler is only trying to warn us about this potential issue
by generating warning messages while compiling our code.
The first warning message that it prints out is, "Warning: unchecked
call to add(E) as a member of the raw type
java.util.ArrayList " for the code al.add("alice"); at line 8. It
prints the same warning every time an object is added to the list. The compiler is
trying to tell us that it does not know what kind of objects this ArrayList is
supposed to hold and that it has no way of checking what we are adding to this
list. By printing out the warning, the compiler is telling us that it will not be held
responsible for the objects that are being put in this list. Whoever wants to
retrieve objects from this list must already know what type of objects are present
inside the list and must cast those objects appropriately upon retrieval at their
own risk. In other words, this list is basically "type unsafe" . It is unsafe
because it depends on assumptions made by humans about the contents of the
list. This assumption is not checked or validated by the compiler for its
truthfulness. One can put any kind of object in the list and others will not realize
until they are hit with a ClassCastException at run time when they try to
cast the retrieved object to its expected type.
Java solves this problem by allowing us to specify the type of objects that we are
going to store in a list while declaring the type of the list. If, instead of writing
ArrayList al = new ArrayList(); , we write
ArrayList<String> al = new ArrayList<String> (); all the
warning messages go away. The compiler now knows that the ArrayList
pointed to by al is supposed to hold only String s. Therefore, it will be able
to keep a watch on the type of objects the code tries to add this list. If it sees any
code that attempts to add anything that is not a String to the list, it will refuse
to compile the code. The error message generated by the following code
illustrates this point:
ArrayList<String> al = new ArrayList<String>();

al.add(new Object());
The error message is:
Error: no suitable method found for add(java.lang.Obje

ct)
method java.util.Collection.add(java.lang.String)
is not applicable
The compiler will not let you corrupt the list by adding objects that this list is not
supposed to keep. Similarly, the compiler will not let you make incorrect
assignments either. Here is an example:
List<String> al = new ArrayList<String>();//al points

to a List of Strings.
al.add("hello");//valid
String s = al.get(0); //valid, no cast required
Object obj = al.get(0); //valid because a String is-a

Object
Integer in = al.get(0); //Invalid. Will not compile.

The error message is:
Error: incompatible types: java.lang.String cannot be

converted to java.lang.Integer
Observe that we supplied information about the type of object the list is
supposed to keep by appending <String> to the variable declaration and the
ArrayList creation. This way of specifying the type information is a part of a
feature introduced in Java 5 known as "generics ".
The above discussion should tell you that the topics of List/ArrayList,
Collections, and Generics are interrelated. All three of them combine together to
provide you a powerful mechanism to write type safe and bug free code.
Although the exam objectives mention only List and ArrayList explicitly, we
have seen questions that require you to have a basic understanding of the
Collections API , the Collection and Map interfaces, the HashMap class and
generics as well in the exam. So, before I move on to the API details of the
important classes, I will take a diversion to discuss Generics in a little more
detail.
13.3.2 Generics ☝
Although Generics is a huge and complicated topic that deserves a chapter or

two of its own, I will try to introduce it to you in an easy to understand manner.
My objective here is not to overwhelm you with too much complexity but to
provide you with the foundation required by the OCP Java 11 Part 1 exam.
The whole purpose of introducing generics in Java is to enhance type safety of

the code. Type safely is important because it helps prevent a category of bugs
that occur due to wrong assumption about the type of objects at run time. For
example, if the developer assumes that an object pointed to by a variable will be
of type CheckingAccount and writes the code accordingly, and if at run time it
turns out that the object is actually of type SavingsAccount, the code will fail.
Had the compiler warned the developer that their assumption will turn out to be
wrong at run time, the developer would have fixed the code right then and there.
Here, "making an assumption" essentially means casting a variable that was
declared to be of one type to another type. You have seen cases where the
compile will not allow you to make such a cast. For example, the compiler will
simply refuse to compile the cast in this code - Integer i = 10; String
s = (String) i; . The compiler knows that the declared type of i is
Integer and since String and Integer classes do not have a parent-child
relationship, there is no way i can ever point to an object of type String . The
compiler correctly points out that this cast is a programming mistake. But, as
you saw with the code snippet involving an ArrayList in the previous
section, there are cases where the compiler does not have enough information to
prevent the developer from making such incorrect assumptions. In that code, the
compiler had no knowledge about the type of the objects contained in the list and
so, it had no option but to allow the compilation to succeed. It simply generated
a warning to let the programmer know that the code may run into issues at run
time.
Parameterized classes ☝
The first challenge, therefore, in enhancing the type safety in a piece of code is
to add type information to a class, which the compiler can use to prevent
inappropriate casts. One could achieve that by creating a class with more
specific types instead of Object . For example, one could copy the code from
ArrayList , make some changes to it and create a StringList class that
would use String , instead of Object , as the type of input parameters and
return types in its methods. This solution has an obvious shortcoming. You will
need to define a new class every time you need to create a list of a new type. Just
imagine the horror if you find out a bug in the code. You would have to fix it in
all the cloned classes. It is an impractical solution. The second challenge,
therefore, is to prevent duplication of the code. So, what we really need is a way
to write a class in a generic fashion, without hardcoding it to any particular type,
and also allow that class to be typed to any type as per the requirement, at the
time of using it.
This is exactly what "parameterized" classes in Java allow you to do. Instead of
using an actual type while coding a class, you use a placeholder. Whoever uses
the class later must specify the value for that placeholder so that the class can be
typed to that type. I know it sounds complicated but the following example will
make this clear:
public class DataHolder<E> {

E data;
public E getData(){ return data; }
public void setData(E e){ data = e; }
}
In the above code, instead of using any specific class as the type of data
variable, I have used a placeholder named E . The name of the placeholder
doesn't have to be E . You could name it anything but conventionally, only a
single capital letter is used for this purpose. Observe the part where it says <E>
in the class declaration. This tells the compiler that E is just a placeholder and
that E will be replaced by the actual type later on when this class is actually used
by another piece of code. The code in DataHolder is written as if the type of
the data variable is E . The code in this class is generic enough to not care
about the type with which E is replaced. In technical parlance, DataHolder is
a parameterized class, because it uses a type parameter (E is the type
parameter, in this case) for a type that it uses in its code.
Let us now see how this class can be used by other people. Let us say there is a
class that consumes a DataHolder object containing String . Here is the
code for such a class:
public class SomeClass {

public static void consumeData(DataHolder<String>
stringData){
String s = stringData.getData(); //no cast re
quired
//do something with s
//Integer i = (Integer) stringData.getData();

//will not compile
}
}
Observe the usage of <String> while declaring stringData parameter.

Here, String is the parameter (a type argument , to be precise) that is being
passed to the parameterized class DataHolder . In technical parlance, we call
it "typing" DataHolder to String . It tells the compiler that only a
DataHolder object containing String can be passed to this method. With
this declaration, the compiler gets all the information that it needs to prevent a
programmer from making wrong assumption about what the argument contains.
Note that this typing applies only to the stringData variable and not to all
DataHolder objects in the program. Only the DataHolder object referred
to by the stringData variable is being guaranteed to contain String here.
A program could certainly use another DataHolder variable typed to a
different type.
There are three important points to understand in the above code.
First, no cast is required to assign the value returned by getData to s . The

compiler knows that the DataHolder object pointed to by stringData has
been typed to String . So, stringData.getData() will always return a
String .
Second, the compiler will not let you cast the value returned by
stringData.getData() to any type that is not a String . The reason is
the same. The compiler knows that stringData.getData() can never
return anything other than a String and so casting the return value to
Integer is a programming error.
Finally, third, the compiler will not let you invoke the consumeData method
with anything other than a DataHolder typed to String . For example, the
following code in yet another class will not compile:

DataHolder<Integer> dh = new DataHolder<Intege
r>();
SomeClass.processData(dh);//will not compile
}
}
The compiler knows that SomeClass.processData expects a

DataHolder containing String . So, it will not let the programmer make the
mistake of invoking this method with a DataHolder typed to Integer .
Observe that the compiler is able to guarantee that the call to

stringData.getData() inside processData() method will return a
String only because it prevents an invocation of processData method with
a DataHolder typed to anything other than String from compiling in the
first place! No loose ends, right?
You may wonder at this point how the above code is better than simply declaring
the data variable as Object . Something like this:
public class DataHolder{

Object data;
public Object getData(){ return data; }
public void setData(Object e){ data = e; }
}
After all, the above code also allows a DataHolder to contain any type of
object. But there lies the problem. We don't know the type of the object that a
given DataHolder object contains. In other words, we do want the
DataHolder object to be able to contain any type of object but not without us
knowing what kind of object it contains at any given point. Furthermore, we
don't want DataHolder object to contain any other type of object once we
decide what type of object it should contain.
Parameterized Methods ☝
Parameterized methods are similar to parameterized classes. The only difference
is that the type parameter is valid only for that method instead of the whole class.
Here is an example:
public class ListProcessor{

public static <T> T processList(List<T> listOfT){
return listOfT.get(0);
}
}
In the above code, T is used as a placeholder for the actual type in the
processList method. This method takes a List of T s and returns a T .
Unlike the placeholder that was used in the DataHolder class, T is valid only
within this method. So, if you pass a List<String> to this method, it will
return a String .
Java 5 parameterized the code for all the classes/interfaces in the Collections
API so that you could use type safe collections in your code. That is why you are
able to type ArrayList to any type you want using the < > syntax. You may
go ahead and look at the source code of any of the Collection API classes to get
a feel of how exactly it has been parameterized.
Using parameterized types ☝
You will not be tested on the syntax of defining a parameterized class or a

parameterized method but you should be aware of it because such code may
appear in the real exam questions. Furthermore, even though defining
parameterized types is not on the part 1 exam (it is on the part 2 exam), using a
parameterized type or a method is. The code snippets presented in exam
questions use parameterized types quite liberally. So, you need to know the
following ways in which you can declare type safe variables and instantiate
parameterized types to understand what the code presented in a question is trying
to do:
1. A type safe variable can be declared by specifying the type within < and > .
For example,
ArrayList<String> stringArray;
List<Integer> iList;
ArrayList<com.acme.Foo> fooList;
List<> iList; //invalid, type must be specified in

side <> while declaring a variable
2. A parameterized class can be instantiated similarly by specifying the type

within < and > . For example,
new ArrayList<String>();
new ArrayList<Integer>();
new ArrayList<com.acme.Foo>();
3. If you are instantiating a type safe ArrayList while assigning it to a type

safe variable in the same statement, then you may omit the type name from
the instantiation as shown below:
ArrayList<String> al = new ArrayList<>
();
Upon encountering <> , the compiler infers the type of the ArrayList
object from the type of the variable. The <> operator is known as the
"diamond operator ". It can be used only in a new statement. Thus, the
statement ArrayList<> al; will not compile but new
ArrayList<>(); will. This operator was introduced in Java 7 with the
sole purpose of saving you from typing a few extra keystrokes.
Remember that you cannot apply the < > syntax to just about any class. You can
apply it only to classes that have been parameterized.
Generics and Type Erasure ☝
Generics were added quite late to Java (only in 1.5) and so backward
compatibility was an important concern. Furthermore, since it was a fairly
complicated feature, it was expected that their wide spread usage will take some
time. Thus, it was deemed necessary that code that uses Generics must work
with code that doesn't use Generics. Basically, there are two scenarios that had to
be supported. Generic aware Java source code should be able use preexisting
classes that did not use generics and non-generics source code should be able to
use new generic aware libraries.
To achieve this design goal, the mechanism of type erasure was used, where all
the generic information is stripped from a class at run time. In other words, the
JVM does not get to know any of the type information that is present in a class
file in the form of generics. This information is used only by the compiler to
perform its type checking of the source code at the time of compilation. Thus,
even if you write ArrayList<String> al = new ArrayList<>();
in your code, the JVM will only see ArrayList al = new
ArrayList(); . The type information is effectively "erased" from a class for
the JVM. This implies that the Java code String str = al.get(0); gets
translated to String str = (String) al.get(0); in the class file.
Observe that this is completely opposite to the behavior displayed by arrays. As
discussed in the Arrays chapter, arrays are reified, which means the type
information stays with an array at all times.
Type erasure eliminates the need to create separate class files for differently
typed usage of a parameterized class. This ensures backward compatibility
because there is essentially no difference between generics aware code and
regular code from the JVM's perspective.
Type erasure has a serious impact on overloading and overriding of methods.

Although not mentioned explicitly, the exam has questions that touch upon this
aspect. So, let's dig a little deeper into it.
Impact of Type Erasure on Method Overloading ☝
As explained in the previous chapters, overloaded methods are methods that

differ in their signature. Recall that a method signature is made up of method
name plus parameter types. Since, due to type erasure, generic information is
removed at run time, it follows that generic information cannot be a part of the
signature. Thus, two methods that differ only in their type parameters are not
valid overloads. The compiler will refuse to compile such code because it knows
that the JVM will not be able to distinguish between them at run time. For
example, the following code won't compile:

public void processData(DataHolder<String> stringDa
ta){
String s = stringData.getData();
}

public void processData(DataHolder<Integer> intData
){
Integer i = intData.getData();
}
}
The error message will be error: name clash:

processData(DataHolder<Integer>) and
processData(DataHolder<String>) have the same erasure .
Indeed, once the generic information is removed, signature of both the methods
are same and there is no way for the JVM to bind an invocation of this method to
any one version.
Impact of Type Erasure on Method Overriding ☝
Overriding a method that uses generic types has the same issue that I talked
about above. The following code fails to compile:
class Base{
public void processData(DataHolder<String> stringDa
ta){
String s = stringData.getData();
}
}
public class SubClass extends Base{
public void processData(DataHolder<Integer> intData
){
Integer i = intData.getData();
}
}
The error message in this case is: error: name clash:

processData(DataHolder<Integer>) in SubClass and
processData(DataHolder<String>) in Base have the same
erasure, yet neither overrides the other .
In this case, the overriding method is trying to break the contract promised by
the base class. The base class's processData version expects a
DataHolder containing String , while the subclass's version expects a
DataHolder containing Integer . Although after erasing the type
information, the subclass method seems to correctly override the base class's
method. However, the compiler notices that the cast to Integer in the
subclass's method will certainly fail at run time because
intData.getData() will return a String ! Therefore, it refuses to
compile the code.
The presence of generics can throw up complicated situations with

respect to Overloading and Overriding. However, the exam does not
grill you on this topic. You are only expected to identify simple cases
of overloading and overriding involving generics. It is important to not
panic upon seeing such questions. You will be able to answer it by
applying the rules explained above.
There are still several important concepts about
Generics such as mixing generic and non-generic code,
multiple type parameters, and bounded type parameters
that I have not discussed in this book because they are
way beyond the scope of the part 1 exam. However,
you should read about them if you are going to face
technical interviews.
13.3.3 Quiz ☝
Given the following two classes:
class Base{
public <T> List<T> processList(List<T> list){ retur
n null; }
}
and

//add method here
Which of the following options are true if the method is added to SubClass ?
A. public <T> List<T> processList(List<T> list){ return

null; } correctly overrides processList of Base .
B. public <T> Collection<T> processList(List<T> list){

return null; } correctly overrides processList of Base .
C. public List<String> processList(List<String> list){
D. . public <T> ArrayList<T> processList(List<T> list){

E. public <T> List<T> processList(ArrayList<T> list){

return null; } correctly overloads processList of Base .
F. public <T> List<T> processList(List<T> list){ return

null; } correctly overloads processList of Base .
G. public <T> List<T> acceptList(List<T> list){ return

null; } correctly overloads processList of Base .
Correct answer is A, D, E
Remember that when a subclass contains a method with the same signature as a
method in the super class, it is called an override and when a class contains (or
inherits) multiple methods with the same name but different parameter types, it
is called an overload .
Option A is obviously correct because the method signature of a method in the

subclass matches exactly to the method signature of a method in the super class.
It is a valid override.
Option B could have been a valid override but the return type of the overriding
method cannot return a wider type. Recall the rule of covariant returns . So, it
is incorrect .
Option C is incorrect because of a mismatch in the generic types of the method

parameters and the return type as explained in this section.
Option D is correct because the method signatures match and the subclass
method follows the rule of covariant returns. ArrayList is a narrower type
than List because ArrayList implements List .
Option E is correct because the parameter type of the method in subclass is

different. The subclass, therefore, has two methods (one of its own and one
inherited) with the same name but different parameter types. Thus, it is a valid
overload.
Option F is incorrect because it is a valid override and not a valid overload.
Option G is incorrect because the method name doesn't match. It is neither an

overload nor an override. It is a valid method though.
13.3.4 Collection and List API ☝
Collection API ☝
java.util.Collection is the root interface in the collections hierarchy. It abstracts

the concept of a collection in general. It declares only the methods that are
applicable to all sorts of collections and leaves the methods that deal with
specialized features such as ordering or uniqueness of elements to sub-interfaces
such as List and Set. For example, Collection declares add(Object e)
and remove(Object e) methods but does not declare add(int index,
Object e) and remove(int index) because the concept of ordering is
not applicable to all collections. We use the Collection interface when we don't
want to make any assumption about the characteristics of the group of objects. In
other words, if we get a Collection object, all we know is that it contains a bunch
of elements. We don't know whether the elements are ordered in any way,
whether the group contains duplicate elements, or what is the data structure used
to keeps the elements in that collection.
The following figure shows the hierarchy of interfaces and classes that you will
see on the exam. The rectangles with dashed borders are interfaces and the
rectangles with solid borders are classes.

Although the exam does not expect you to remember which method is declared
in Collection and which is not, it is still a good idea to go through the
JavaDoc API description of this interface and check out the methods declared in
this interface. The method names are quiet descriptive and the methods do what
their names suggest. Some of the methods that you should pay attention to are:
add , addAll , remove , removeAll , removeIf , contains ,
containsAll , isEmpty , and clear .
List API ☝
java.util.List interface defines the behavior of collections that keep objects in an

order. Elements of a list can be accessed using an index. It also allows elements
to be inserted in or removed from any given index. Duplicate elements as well as
nulls are permitted. Since it is just an interface, it doesn't specify how the
functionality is implemented. Actual implementation classes such as
ArrayList and LinkedList use different mechanisms to implement the
behavior. As you will see later with ArrayList , implementation classes may
provide additional features on top of the features defined in List .
Again, you should go through all of the methods of the List interface from the
JavaDoc. The exam doesn't trick you on method names and doesn't require you
to recall whether a given method exists in List or not. The exam questions are
based on the understanding of how a given method behaves. You are expected to
know what happens to the existing list of objects after the invocation of a
method. The trick is in knowing "boundary conditions". For example, what
happens if you try to remove a non-existing object or a null from a list or what
happens if you try to add null.
I will list the important methods of List interface below. Remember that index
is zero based, which means, the index of the first element is zero.
1. E get(int index) : Returns the element at index.

2. E set(int index, E e) : Replaces the existing element at index
with the passed object and returns the original element that was replaced.
3. void add(E e) : Adds an object at the end of the list.
4. void add(int index, E e) : Adds an object at the specified index
and shifts the existing elements at a higher index to the right.
5. void addAll(Collection c) : Adds all the elements of the given
collection to this list at the end.
6. void addAll(int index, Collection c) : Adds all the
elements of the given collection at the given index.
7. E remove(int index) : Removes an object from the specified index
and shifts the existing elements at a higher index to the left. It returns the
original element that was removed from the list.
8. boolean remove(E e) : Removes the first occurrence of given object
from the list. Returns true if the object was found. The passed object is
matched with the objects in the list using the equals method.
9. void removeAll(Collection c) : Removes from this list all the
elements that are present in the passed collection.
10. void retainAll(Collection c) : Removes from this list all the
elements that are not in the passed collection.
11. void clear() : Removes all the elements from this list.
12. int size() : Returns the number of elements in this list.
13. default void forEach(Consumer<? super T> action) :
Performs the given action for each element of the Iterable until all elements
have been processed or the action throws an exception.
You will see this method used with lambda expressions a lot.
The index value must lie between 0 and list's current size (both inclusive)
otherwise an IndexOutOfBoundsException is thrown. A typical question
that tests your understanding of the above methods would be the predict to
output of the following code:
List<Integer> list = new ArrayList<Integer>();

list.add(0, 1);
list.add(0, 2);
list.add(0, 3);
System.out.println(list);
If you think the output is [1, 2, 3] , you have been tricked. The output is
[3, 2, 1] because the elements are being inserted at the 0 th index.
Here are few methods that help you inspect a list:
1. boolean isEmpty() : Returns true if the list has no elements.

2. boolean contain(Object o) : Returns true of the list contains the
given object.
3. boolean containsAll(Collection c) : Returns true of the list
contains all of the elements present in the given collection.
4. List subList(int fromIndex, int toIndex) : Returns a
view of the portion of this list between the specified fromIndex, inclusive,
and toIndex, exclusive. If fromIndex and toIndex are equal, the returned list
is empty.
5. int indexOfâ€‹(Object o) : Returns the index at which the given
object is found in the list. Returns -1 if it is not found.
6. int lastIndexOfâ€‹(Object o) : Returns the last index at which
the given object is found in the list. Returns -1 if it is not found.
7. Object[] toArray() : Returns a array of Objects containing all the
elements of this list.
8. <T> T[] toArray(T[] a) : Returns an array of the type specified in
the argument containing all the elements of this list. The objects of the list
must be of the type passed in the argument. This is method is helpful when
you want to convert a list into an array of a specific type. For example,
String[] strArray = listOfString.toArray(new
String[0]); will return a String[] containing all the elements of the list.
The subList(int fromIndex, int toIndex) method in the above

list of methods is particularly interesting. It does not return a new independent
list but just a view into the original list. Thus, any change you do to the view is
reflected in the original list as shown the example below:
List<String> al = new ArrayList<String>();

al.addAll(Arrays.asList( new String[]{"a", "b", "c", "
d", "e" } ));
List<String> al2 = al.subList(2, 4);
System.out.println(al2); //prints [c, d]
al2.add("x");
System.out.println(al2); //prints [c, d, x]
System.out.println(al); //prints [a, b, c, d, x, e]
al.add("y"); //structural modification to the original

list
System.out.println(al2); //throws java.util.Concurrent

ModificationException
Observe the output of the last two lines of the output. x was added to the end of
the list pointed to by al2 . However, since al2 is just a view of a portion of the
original list pointed to by al , x is visible in the middle of this list. Furthermore,
all structural modifications must be done through the view. If you structurally
modify the backing list and then try to use the view, the results will be
unpredictable. Structural modification means any change in size of the list. That
is why, the last line of the above code throws a
java.util.ConcurrentModificationException exception.
List.of and List.copyOf methods ☝
Java has had the java.util.List interface ever since the Collections API
was introduced in Java 1.2. It was enhanced in Java 9 and 10 with several static
methods and the exam expects you to know them. These are the twelve
overloaded of (added in Java 9) methods and the copyOf method (added in
Java 10)..
The overloaded of methods are merely convenience methods that take 0 to 10

parameters and return an unmodifiable List. For example, if you want to create
a list of 3 Strings, you could do List<String> list1 =
List.of("a", "b", "c"); .
Similarly, the copyOf method is a convenience method to create an

unmodifiable list using the elements of an existing collection. For example,
List<String> list2 = List.copyOf(list1); .
You will see the usage of these methods all over exam questions even when the
question is not about lists. Since understanding the behavior of the lists returned
by these methods is very important for the exam, let me first jot down their
characteristics as given in JavaDoc:
1. They are unmodifiable . Elements cannot be added, removed, or replaced.

Calling any mutator method on the List will always cause
UnsupportedOperationException to be thrown. However, if the
contained elements are themselves mutable, this may cause the List's
contents to appear to change.
2. They disallow null elements. Attempts to create them with null elements
result in NullPointerException .
3. They are serializable if all elements are serializable. (Not important for the
Part 1 exam)
4. The order of elements in the list is the same as the order of the provided
arguments, or of the elements in the provided array.
5. They are value-based. Callers should make no assumptions about the
identity of the returned instances. Factories are free to create new instances
or reuse existing ones. Therefore, identity-sensitive operations on these
instances (reference equality (==), identity hash code, and synchronization)
are unreliable and should be avoided.
Here is the same code that I showed above but with a small modification:
List<Integer> list = List.of();

list.add(0, 1);
list.add(0, 2);
list.add(0, 3);
Before jumping on to answer [3, 2, 1] this time, you need to appreciate the
fact that the list returned by List.of is unmodifiable! So, the code will
compile but the first call to list.add will throw an
java.lang.UnsupportedOperationException exception at run time.
13.3.5 ArrayList API ☝
As discussed earlier, ArrayList is one of the implementation classes of the List

interface. It has three constructors .
1. ArrayList() : Constructs an empty list with an initial capacity of 10.

Just like you saw with the StringBuilder class, capacity is simply the
size of the initial array that is used to store the objects. As soon as you add
the 11th object, a new array with bigger capacity will be allocated and all
the existing objects will be transferred to the new array.
2. ArrayList(Collection c) : Constructs a list containing the

elements of the specified collection.
3. ArrayList(int initialCapacity) : Constructs an empty list with

the specified initial capacity. This constructor is helpful when you know the
approximate number of objects that you want to add to the list. Specifying
an initial capacity that is greater than the number of objects that the list will
hold improves performance by avoiding the need to allocate a new array
every time it uses up its existing capacity.
It is possible to increase the capacity of an ArrayList even after it has

been created by invoking ensureCapacity(int n) on that
ArrayList instance. Calling this method with an appropriate number
before inserting a large number of elements in the ArrayList improves
performance of the add operation by reducing the need for incremental
reallocation of the internal array. The opposite of ensureCapacity is
the trimToSize() method, which gets rid of all the unused space by
reducing its capacity to the match the number of elements in the list.
Here are a few declarations that you may see on the exam:
List<String> list = new ArrayList<>(); //ok because Ar

rayList implements List
var al = new ArrayList<Integer>(50); //initial capacit

y is 50
ArrayList<String> al2 = new ArrayList<>(list); //copyi

ng an existing list, observe the diamond operator
var list1 = new ArrayList<>(); //ok, list1 is of type

ArrayList<Object>, observe the combination of var decl
aration and the diamond operator
List list2 = new List(list); //will not compile becaus

e List is an interface, it cannot be instantiated
Important methods of ArrayList ☝
ArrayList has quite a lot of methods. However, since ArrayList

implements List (which, in turn, extends Collection ), several of
ArrayList 's methods are declared in List and Collection interfaces.
The exam does not expect you to make a distinction between the methods
inherited from List /Collection and the methods declared in ArrayList
.
The following are the ones that you need to know for the exam:
1. String toString() : Well, toString is not really the most

important method of ArrayList but since we will be depending on its
output in our examples, it is better to know about it anyway. ArrayList 's
toString first gets a string representation for each of its elements (by
invoking toString on them) and then combines into a single string that
starts with [ and ends with ] . For example, the following code prints [a,
b, c] :
var al = new ArrayList<String>();

al.add("a");
al.add("b");
al.add("c");
Observe the order of the elements in the output. It is the same as the order
in which they were added in the list. Calling toString on an empty
ArrayList gets you [ ] . I will use the same format to show the
contents of a list in code samples.
Methods that add elements to an ArrayList :
1. boolean add(E e) : Appends the specified element to the end of this

list. As discussed in the Generics section, E is just a place holder for
whichever type you specify while creating the ArrayList . For example,
if you create an ArrayList of Strings, i.e., ArrayList<String> , E
stands for String .
This method is actually declared in Collection interface and the return

value is used to convey whether the collection was changed as a result of
calling this method. In case of an ArrayList , the add method always
adds the given element to the list (even if the element is null), which means
it changes the collection every time it is invoked. Therefore, it always
returns true .
2. void add(int index, E element) : Inserts the specified element

at the specified position in this list. The indexing starts from 0 . Therefore,
if you call add(0, "hello") on an list of Strings , "hello" will
be inserted at the first position.
3. boolean addAll(Collection<? extends E> c) : Appends all

of the elements in the specified collection to the end of this list, in the order
that they are returned by the specified collection's Iterator. Again, for the
purpose of the Part 1 exam, you don't need to worry about the "?
extends E" or the Iterator part. You just need to know that you can add
all the elements of one list to another list using this method. For example:
ArrayList<String> sList1 = new ArrayList<>(); //ob

serve the usage of the diamond operator
sList1.add("a"); //[a]
ArrayList<String> sList2 = new ArrayList<>();

sList2.add("b"); //[b]
sList2.addAll(sList1); //sList2 now contains [b, a]
4. boolean addAll(int index, Collection<? extends E>

c) : This method is similar to the one above except that it inserts the
elements of the passed list in the specified collection into this list, starting at
the specified position. For example:

sList1.add("a"); //[a]

sList3.add("b"); //[b]
sList3.addAll(0, sList1); //sList3 now contains [a

, b]
Methods that remove elements from an ArrayList :
1. E remove(int index) : Removes the element at the specified

position in this list. For example,
ArrayList<String> list = ... // an ArrayList conta

ining [a, b, c]
String s = list.remove(1); //list now has [a, c]
It returns the element that has been removed from the list. Therefore, s will
be assigned the element that was removed, i.e., "b" . If you pass an invalid
int value as an argument (such as a negative value or a value that is beyond
the range of the list), an IndexOutOfBoundsException will be
thrown.
2. boolean remove(Object o) : Removes the first occurrence of the

specified element from this list, if it is present. For example,
ArrayList<String> list = ... // an ArrayList conta
ining [a, b, a]
list.remove("a"); //[b, a]
Observe that only the first a is removed.

You have to pay attention while using this method on an ArrayList of
Integers . Can you guess what the following code will print?
ArrayList<Integer> list = new ArrayList<>(Arrays.a

sList( new Integer[]{1, 2, 3 } ));
list.remove(1);
list = new ArrayList<>(Arrays.asList( new Integer[
]{1, 2, 3 } ));
list.remove(new Integer(1));
The output is:
[1, 3]
[2, 3]
Recall the rules of method selection In case of overloaded methods. When

you call remove(1) , the argument is an int and since a remove method
with int parameter is available, this method will be preferred over the
other remove method with Object parameter because invoking the other
method requires boxing 1 into an Integer .
This method returns true if an element was actually removed from the list
as a result of this call. In other words, if there is no element in the list that
matches the argument, the method will return false .
3. boolean removeAll(Collection<?> c) : Removes from this list
all of its elements that are contained in the specified collection. For
example the following code prints [ c ] :
ArrayList<String> al1 = new ArrayList<>(Arrays.asL

ist( new String[]{"a", "b", "c", "a" } ));
ArrayList<String> al2 = new ArrayList<>(Arrays.asL
ist( new String[]{"a", "b" } ));
al1.removeAll(al2);
System.out.println(al1); //prints [ c ]
Observe that unlike the remove(Object obj) method, which removes

only the first occurrence of an element, removeAll removes all
occurrences of an element.
This method returns true if an element was actually removed from the
list as a result of this call.
4. void clear() : Removes all of the elements from this list.
Methods that replace an element in an ArrayList :
1. E set(int index, E element) : Replaces the element at the

specified position in this list with the specified element. It returns the
element that was replaced.
Example:
ArrayList<String> al = ... // create a list contai

ning [a, b, c]
String oldVal = al.set(1, "x");

System.out.println(al); //prints [a, x, c]
System.out.println(oldVal); //prints b
Methods that read an ArrayList without modifying it:
1. boolean contains(Object o) : The object passed in the argument

is compared with each element in the list using the equals method. A true
is returned as soon as a matches is found, a false is returned otherwise.
Here are a couple of examples:
ArrayList<String> al = new ArrayList<>();

al.addAll(Arrays.asList( new String[]{"a", null, "
b", "c" } ));
System.out.println(al.contains("c")); //prints true
System.out.println(al.contains("z")); //prints fal

se
System.out.println(al.contains(null)); //prints tr
ue
Observe that it does not throw a NullPointerException even if you

pass it a null . In fact, a null argument matches a null element.
2. E get(int index) : Returns the element at the specified position in

this list. It throws an IndexOutOfBoundsException if an invalid
value is passed as an argument.
3. int indexOf(Object o) : The object passed in the argument is
compared with each element in the list using the equals method. The index
of the first element that matches is returned. If this list does not contain a
matching element, -1 is returned. Here are a couple of examples:
ArrayList<String> al = new ArrayList<>();

al.addAll(Arrays.asList( new String[]{"a", null, "
b", "c", null } ));
System.out.println(al.indexOf("c")); //prints 3
System.out.println(al.indexOf("z")); //prints -1
System.out.println(al.indexOf(null)); //prints 1
Observe that just like contains , indexOf does not throw a

NullPointerException either even if you pass it a null . A null
argument matches a null element.
4. boolean isEmpty() : Returns true if this list contains no elements.
5. int size() : Returns the number of elements in this list. Recall that to
get the number of elements in a simple array, you use the variable named
length of that array.
The examples that I have given above are meant to illustrate only a single
method. In the exam, however, you will see code that uses multiple methods.
Here are a few points that you should remember for the exam:
1. Adding nulls : ArrayList supports null elements.

2. Adding duplicates : ArrayList supports duplicate elements.
3. Exceptions : None of the ArrayList methods throw
NullPointerException . They throw
IndexOutOfBoundsException if you try to access an element
beyond the range of the list.
4. Method chaining : Unlike StringBuilder , none of the ArrayList
methods return a reference to the same ArrayList object. Therefore, it is
not possible to chain method calls.
Here are a few examples of the kind of code you will see in the exam. Try to
determine the output of the following code snippets when they are compiled and
executed:
1. var al = new ArrayList<Integer>(); //observe that

the type specification is on the right side
al.add(1).add(2);
2. ArrayList<String> al = new ArrayList<>();//observe

the usage of the diamond operator
if( al.add("a") ){
if( al.contains("a") ){
al.add(al.indexOf("a"), "b");
}
}
3. ArrayList<String> al = new ArrayList<>();

al.add("a"); al.add("b");
al.add(al.size(), "x");
4. var list1 = new ArrayList<String>();

var list2 = new ArrayList<String>();
list1.add( "a"); list1.add("b");
list2.add("b"); list2.add("c"); list2.add("d");
list1.addAll(list2);
list1.remove("b");
System.out.println(list1);
5. ArrayList<String> list1 = new ArrayList<>();
ArrayList<String> list2 = new ArrayList<>();
list1.remove("b");
6. ArrayList<String> list1 = new ArrayList<>();

ArrayList<String> list2 = new ArrayList<>();
list1.removeAll("b");
Size vs Capacity ☝
Size is the number of elements that are currently stored

in the ArrayList. While capacity is the length of the
internal array that an ArrayList uses to store its
elements.
You know that in Java, arrays are of fixed length. You

must specify the length of an array while creating it
and you can never change this length. So, what
happens when you add more elements to an ArrayList
than its capacity? Simple. A new array with bigger
length is allocated and elements are transferred from
the old array to the new array. Since all this is
managed internally by the ArrayList, you don't have to
worry about it most of the time.
However, it has implications on the performance of an

ArrayList. Since an ArrayList has no idea how many
elements are going to be added to it, it starts with a
capacity of 10 and keeps incrementing the capacity as
soon as it is full (the exact algorithm for incrementing
the capacity is not important here). Now, imagine if you
wanted to add a thousand elements to an ArrayList.
This could potentially require the ArrayList to perform
the allocation and transfer operation, which is quite
expensive, several times. This will reduce its
performance.
This is where the ArrayList(int capacity)

constructor and ensureCapacity(int capacity)
method come in handy. If you know the number of
elements you are going to add to the ArrayList, you
should use this constructor/method to make sure that
the internal array of the ArrayList is big enough to hold
all the elements.
Remember that insertion of an element depends the

size (and not capacity ) of the ArrayList. For example,
if the size of an ArrayList is 5, you can't insert an
element at index 6 even if the capacity of that ArrayList
is 10. It will throw an IndexOutOfBoundsException.
13.3.6 ArrayList vs array ☝
You may get a few theoretical questions in the exam about the advantages and
disadvantages of an ArrayList over an array. You have already seen all that we
can do with ArrayLists and arrays, so, I am just going to summarize their
advantages and disadvantages here.
Advantages of ArrayList ☝
1. Dynamic sizing - An ArrayList can grow in size as required. The

programmer doesn't have to worry about the length of the ArrayList while
adding elements to it.
2. Type safety - An ArrayList can be made type safe using generics.
3. Readymade features - ArrayList provides methods for searching and for
inserting elements anywhere in the list.
Disadvantages of ArrayList ☝
1. Higher memory usage - An ArrayList generally requires more space than

is necessary to hold the same number of elements in an array.
2. No type safety - Without generics, an ArrayList is not type safe at all.
3. No support for primitive values - ArrayLists cannot store primitive values
while arrays can. This disadvantage has been mitigated somewhat with the
introduction of autoboxing in Java 5, which makes it is possible to pass
primitive values to various methods of an ArrayList. However, autoboxing
does impact performance.
Similarities between ArrayLists and arrays ☝
1. Ordering - Both maintain the order of their elements.

2. Duplicates - Both allow duplicate elements to be stored.
3. nulls - Both allows nulls to be stored.
4. Performance - Since an ArrayList is backed by an array internally, there is
no difference in performance of various operations such as searching on an
ArrayList and on an array.
5. Thread safety - Neither of them are thread safe. Don't worry, thread safety
is not on the exam, but you should be aware of the fact that accessing either
of them from multiple threads, may produce incorrect results in certain
situations.
13.3.7 Map and HashMap ☝
Map is data structure that allows you to lookup one piece of information, i.e.,
value, using another piece of information, i.e., key. For example, a map of
country codes and country names will allow you to look up a country name
using the country code, a map of student id and student name will allow you to
look up a student name using the student id, and so on. Here, countrycode and
student id are the keys and country name and student name are the values. In that
sense, a Map is a collection of key-value pairs.
The java.util.Map interface of the Collections API captures this behavior.

There are several implementation classes that implement the Map interface. The
one that is used the most is java.util.HashMap .
It is important to know that Map does not extend Collection interface. Map
is the root of a separate hierarchy of classes and interfaces that is unrelated to the
Collection hierarchy. Although the exam does not mention Map and HashMap
explicitly, we have seen questions on the exam that require you to know the
following methods of Map.
1. V get(Object key): Returns the value to which the specified key is

mapped, or null if this map contains no mapping for the key.
2. V put(K key, V value) : Associates the specified value with the
specified key in this map. If the map previously contained a mapping for
the key, the old value is replaced by the specified value and the old value is
returned. It returns null if there was no mapping for key.
3. V remove(Object key) : Removes the mapping for a key from this
map if it is present. Returns the value or null, if not present.
4. Set<K> keySet() : Returns a Set of keys present in the map. Observe
that keys are unique, so the return type is Set (not Collection or List).
5. Collection<V> values() : Returns a Collection of the values
present in the map. Observe that values may be duplicate, so, the return
type is Collection.
6. void clear() : Removes all entries stored in the map.
7. int size() : Returns the number of entries stored in the map.
8. default void forEach(BiConsumer<? super K, ? super
V> action): Performs the given action for each entry in this map until
all entries have been processed or the action throws an exception. Note:
This is way out of scope for the Part 1 exam, yet, we have seen a question
on lambda expressions that refers to this method in the exam. It is possible
that it came up in the Part 1 exam by mistake.
The above methods and their descriptions are enough for the exam but there are
several interesting and useful methods in Map and it would be a good idea to go
through the JavaDoc to learn more.
The following code illustrates how a map is typically used:
import java.util.*;
static Map<String, String> idNameMap = new HashMap<>
();
static List<String> keys = new ArrayList(List.of("0"
, "1", "2"));
static List<String> values = List.of("a", "b", "c");
public static void buildMap(){

//add key-value pairs in the map
for(int i=0; i<keys.size(); i++){

idNameMap.put(keys.get(i), values.get(i));
}
System.out.println(idNameMap); //prints {0=a, 1=b,
2=c}
}
buildMap();
String value = idNameMap.get("2");
System.out.println(value);//prints c
//Integer ivalue = idNameMap.get("2"); //will not

compile
keys.clear(); //removes all elements from keys Lis

t, doesn't affect the map
System.out.println(keys.size()); //prints 0, keys
is now empty
System.out.println(idNameMap.size());//prints 3, m
ap still has the key value entries
idNameMap.remove("1"); //remove the key 1 and its

associated value from the map
idNameMap.clear(); //remove all entries from the m

ap
System.out.println(idNameMap); //prints {}
}
}
Following are a couple of points that you should observe in the above code.
1. Map and HashMap are parameterized classes. However, unlike a single

type parameter that you saw in the case List and ArrayList , they use two
type parameters. One for the type of key (denoted by K) and one for the
type of value (denoted by V). In the above program, the type of the key and
the value is the same, i.e., String .
2. Since the Map's key and value have been typed to String, the compiler will
neither allow you to put anything else in it nor will allow you to retrieve
anything other than a String from it.
Map and HashMap are used a lot in Java development

and are, therefore, a favorite topic in technical
discussions. You should definitely read about them if
you are going to face technical interviews.
13.3.8 Quiz ☝
Given the following two classes:
class Base{
public void process(Collection c){ }
}
and

public void process(Map<Integer, String> map){
//...code here
}
}
Which of the following options are correct?
A. The process method in SubClass correctly overrides the process

method in Base .
B. The process method in SubClass correctly overloads the process

method in Base .
C. A call to super.process(map.values()); can be inserted in the

process method of SubClass .
D. A call to super.process(map.keySet()); can be inserted in the

process method of SubClass .
E. A call to process(map); can be inserted in the process method of

SubClass .
Correct answer is B, C, D .
Since the type of the arguments of the process methods do not match, this is
not a case of overriding. It is overloading. Thus, option A is wrong and option B
is correct. Since values() and keySet() methods of Map
return Collection and Set respectively (Set is a Collection ), options
C and D are correct. Since Collection and Map do not have a parent child
relationship, a Map is not a Collection and so, option E is incorrect.
13.4 Exercise ☝
1. Given the following two classes:

Document d = new PdfDocument();
System.out.println(d.getType()); //should prin
t "pdf"
}
}
class Document{
private String type = "dummy";
private byte[] data;
//insert appropriate getters and setters
The above code refers to a class named PdfDocument . Write code for
this class such that TestClass will print "pdf" when executed.
class Radio {
private double frequency=1.1;
//insert appropriate getter and setter
}
class TV {
private int channel = 5;
//insert appropriate getter and setter
}
TV t = new TV();
Radio r = new Radio();
reset(t);
reset(r);
System.out.println(r.getFrequency()); //sho
uld print 0.0
System.out.println(t.getChannel()); //shoul
d print 0
}
}
Write code for the reset method in TestClass such that TestClass
will print 0.0 and 0 when executed.
3. You are expected to reset several electronic devices in future. Refactor the
code given above such that TestClass's reset method is able to reset any
new device without requiring any change in the method code.
class Pie{
public void makePie(){
System.out.println("making pie");
}
}
Create two classes PumpkinPie and ApplePie that extend Pie .

Override the makePie method in these classes. Ensure that the overridden
method is also invoked whenever makePie is invoked on objects of these
classes.
5. Add a static method named getCalories in Pie :
public static int getCalories(){

return 100;
}
Create a new class named Nutritionist with a method named

printCalories . This method should take any kind of Pie and print
the correct number of calories as applicable for the given kind of Pie .
6. Given the following two classes:
class XMLTransformer {
public String transform(String Data){ return "xm
ldata"; }
}
class NetworkTransformer {
public String transform(String Data) throws IOEx
ception { return "data from network"; }
}
Refactor the two classes by adding an abstract super type named

Transformer .
7. Create a class named TransformerFactory with a method named
getTransformer . This method should return different types of
transformers based on an argument. Use this factory class to get different
types of transformers and invoke the transform method on them.
8. Create a method that takes an array of strings and returns an ArrayList
containing the same strings.
9. Update the above method to remove the duplicate elements from the
ArrayList before returning.
10. Create a method with the signature switchIt(ArrayList al, int
a, int b) . This method should return the same list but after switching
the elements at positions a and b .
Chapter 14 Lambda Expressions
Understanding Lambda Expressions
14.1 Understanding Lambda Expressions

14.1.1 Lambda Expressions ☝
While taking training sessions, I have observed that many Java beginners feel
intimidated by lambda expressions. Most of the time this fear is because of the
name "lambda". The word lambda itself does not create any useful mental
picture and that makes it a bit difficult to relate to. On top of that, most books
and tutorials start explaining lambda expression as a way to do "functional"
programming, which is another buzz word that is hard to relate to when a person
is just learning "regular" programming!
Well, I can tell you that there is no need to be afraid. Lambda Expressions are
actually very simple. While writing test programs, have you ever been frustrated
while typing the words "public static void main(String[] args)" over and over
again? I am sure you have been. If you are using an IDE such as NetBeans or
Eclipse, then you are probably aware of shortcuts such as psvm . You type
psvm , hit the tab key and the IDE replaces psvm with the text "public static
void main(String[] args)" immediately. A lambda expression is just like that but
for the compiler. Think of a lambda expression as a shortcut for the compiler that
does two things - defines a class with a method and instantiates that class. As
soon as the compiler sees a lambda expression, it expands the expression into a
class definition and a statement that instantiates that class. If you think of it as a
tool that saves you from typing a lot of keystrokes, you will start loving it. You
will love it so much that you will look for opportunities to use it as much as
possible. Let me show you how cool it is.
Imagine that you are working on an application for a car shop. You have a list of
cars and you need to filter that list by various criteria. You may need to filter it
by company, you may need to filter it by price, or by any other property that the
users may want. The following code shows how one might do it. You should
actually copy all of the following code in a single file named
TestClass.java so that you can try it out. Just add import
java.util.*; at the top.
class Car
{
String company; int year; double price; String typ
e;
Car(String c, int y, double p, String t){
this.company = c; this.year = y;
this.price = p; this.type = t;
}
public String toString(){ return "("+company+" "+
year+")"; }
}
The Car class represents a car with a few properties. Agreed, Car is not well
encapsulated. Ideally, Car should have had private fields and public accessors.
But since encapsulation is not relevant here, I haven't shown these methods to
conserve space.
class CarMall
{
List<Car> cars = new ArrayList<>();

CarMall(){
cars.add(new Car("Honda", 2012, 9000.0, "HATCH
"));
cars.add(new Car("Honda", 2018, 17000.0, "SEDA
N"));
cars.add(new Car("Toyota", 2014, 19000.0, "SUV
"));
cars.add(new Car("Ford", 2014, 13000.0, "SPORT
S"));
cars.add(new Car("Nissan", 2017, 8000.0, "SUV"
));
}

List<Car> showCars(CarFilter
cf){
ArrayList<Car> carsToShow = new ArrayList<>();
for(Car c : cars){
if(cf.showCar(c)) carsToShow.add(c);
}
return carsToShow;
}
}
interface CarFilter
{
boolean showCar(Car c);
}
CarMall represents the shop. It creates a list of a few Car objects. This list
contains details of all the cars that the shop has. It has a showCars method that
returns a list of cars based on any given criteria. Instead of specifying the actual
criteria for filtering the cars inside the showCars method, it uses a CarFilter
instance to determine whether a car needs to be listed or not. The CarFilter
interface declares just the basic structure of a filter. Note that it is not really
possible to code the actual filtering criteria inside the showCars method because
the criterion is determined by the user of the CarMall class. By accepting an
interface as an argument, the showCars method lets the caller decide the
criterion.

CarMall cm = new CarMall();
CarFilter cf = new CompanyFilter("Honda");
List<Car> carsByCompany = cm.showCars(cf);

System.out.println(carsByCompany);
}
}
class CompanyFilter implements CarFilter
{
private String company;
public CompanyFilter(String c){
this.company = c;
}
public boolean showCar(Car c){
return company.equals(c.company);
}
}
TestClass represents a third party class that uses CarMall . It wants to get
the details of all cars from a particular company, say, Honda . To do that, it
defines a CompanyFilter class that contains the actual logic for filtering cars
based on company name. At run time, it creates a CompanyFilter object and
passes it to CarMall 's showCars method, which returns a filtered list of cars.
Now look at the following code for TestClass that uses a lambda expression:

CarMall cm = new CarMall();
List<Car> carsByCompany = cm.showCars(c -> c.c
ompany.equals("Honda")
);
System.out.println(carsByCompany);
}
}
Observe that there is no separate class that implements CarFilter and there
is no explicit instantiation of a CarFilter object either. Both of these tasks
have been replaced by a very short statement - c ->
c.company.equals("Honda") . That's it. We have actually eliminated 10
lines of code with that change! Go ahead, count them :)
As I said before, the lambda expression used above is just a shortcut for the
compiler. The compiler actually expands this expression into a fully-fledged
class definition plus code for instantiating an object of that class. Once you
understand how the compiler is able to do this expansion, lambda expressions
will seem like a piece of cake to you.
In fact, all the information that is required to do the expansion is available in the
context of cm.showCars(...) method call already. The compiler knows that
it must pass an object of a class that implements CarFilter to the
showCars(CarFilter cf) method. It knows that this class must
implement the boolean showCar(Car c) method because that is the only
abstract method declared in the CarFilter interface. The compiler gathers
from the signature of this method that this method receives an argument of type
Car and returns a boolean . From this information, it can easily generate the
following code on its own:
//This class is created by the compiler. It can give a

ny random name to the class here!
class XYZ implements CarFilter{

public boolean showCar(Car <<parameterName>>
){
return <<an expression that returns a boolean mu
st appear here>>
;
}
}
The only thing that the compiler cannot generate on its own is the boolean
expression that goes inside the showCar method. It is not a coincidence that
that is exactly what our lambda expression c ->
c.company.equals("Honda") contains. The compiler simply takes the
variable name from the left-hand side of -> of the lambda expression, i.e., c ,
plugs in the expression given on the right-hand side of -> i.e.
c.company.equals("Honda") , into the body of the above method and it
has the complete code for a class that implements CarFilter ! Finally, it
throws in an instantiation expression new XYZ(); as a bonus! These are the
same two things that are needed to invoke cm.showCars(CarFilter cf)
method, i.e., code for a class that implements CarFilter and an expression
that instantiates an object of that class!
I suggest you go through the above discussion a couple of times to absorb the
steps that the compiler takes for expanding a lambda expression into a fully-
fledged class before moving forward. As an exercise, try to expand the lambda
expression x -> x.price>10000 into a class that implements CarFilter
.
Observe that the lambda expression does not specify the method name, the
parameter types, and the return type. The compiler infers these three pieces of
information from the context in which you put the lambda expression. Thus, a
lambda expression must exist within a context that can supply all this
information. Furthermore, you know that in Java, a method cannot exist on its
own. It can exist only as a member of a class (or an interface, for that matter).
This implies that the method generated by the compiler for a lambda expression
cannot exist on it own either. The compiler must create a class as well to contain
the method. Since a lambda expression has no name, the class must be such that
it requires implementation of exactly one abstract method whose name the
compiler can use for the generated the method. The only way this can happen is
if the class generated by the compiler implements an interface with exactly one
abstract method or extends an abstract class with exactly one abstract method. If
the interface or the class has more than one abstract method or no abstract
method, the compiler wouldn't know which method the code in the lambda
expression belongs to.
Java language designers decided not to allow lambda expressions for abstract
classes to reduce complexity. Thus, the only question remaining is which
interface should the generated class implement. That depends on the context.
The type that the context expects is described as the "target type" and is the
interface that the class generated by the compiler must implement. In the above
code, the showCars method expects a CarFilter object (i.e. object of a
class that implements CarFilter ). Therefore, the generated class must
implement CarFilter . In technical terms, CarFilter is the target type of
the lambda expression that is passed as an argument to showCars method.
You may wonder at this point why the compiler can't provide a made up name to
the generated method just like it did for the generated class. Well, it could
provide a made up name to the method but what would that achieve? Since the
programmer wouldn't know that made up name, how would they call this
method? Remember that the programmer doesn't need to know the name of the
class because the compiler passes an object of this class and the programmer
knows the name of the parent class or the interface. You have learnt in the
Polymorphism section that a reference of type parent class/interface can be used
to invoke a method on a subclass. That is exactly what is happening here. The
receiver of the lambda expression doesn't care about the type of the actual object
that it receives because the receiving code invokes a method on that object using
an interface reference. Take a look at the showCars method of CarMall
again. It uses a reference of type CarFilter to invoke the showCar method.
It doesn't matter to this code what name the compiler gives to the generated
class. This code only cares about the name of the method.
From the above discussion, it should be clear that a lambda expression can be
written only where the target type is an interface with exactly one abstract
method. Java has a special name for such an interface: Functional Interface . I
will get to Functional interfaces soon, but first, you need to know the various
ways in which you can write a lambda expression.
14.1.2 Parts of a Lambda expression ☝
You have seen that a lambda expression is basically code for a method in
compact form. It has two parts separated by the "arrow" operator, i.e., ->. The
left side is for variable declarations and the right side is for the code that you
want executed. Just like a method, a lambda expression also can have any
number of parameters and can return (or not return) a value. Java allows lambda
expressions to be written in several different ways. The reason for having so
many ways is to cut out as much redundant code as possible. The exam expects
you to know all these ways. You will be asked to identify valid and invalid
lambda expressions in the exam. From this perspective, I have categorized the
variations into two categories - variations on the parameter section and
variations on the body section.
There are three possibilities for the parameters section:
1. No parameter - If a lambda expression takes no parameters, the parameter

part of the expression must have an empty set of brackets, i.e., ( ). For
example,
() -> true //valid

-> 1 //invalid, missing variable declaration part
2. One parameter - If a lambda expression takes exactly one parameter, the

parameter name may be specified within brackets, i.e., ( pName ) or without
the brackets, i.e., pName. For example,
a -> a*a //valid
(a) -> a*a //valid
You may also include the parameter type for the parameter name if you
want but then you will need to use brackets. For example:
(int a) -> a*a //valid

int a -> a*a //invalid
3. More than one parameters - If a lambda expression takes more than one
parameter, all the parameter names must be specified within the brackets,
i.e., ( pName1, pName2, pName3 ). For example,
(a, b, c) -> a + b + c //valid
a, b -> a+b //invalid, parameters must be within (

)
Again, parameter types are optional. For example.
(int a, int b, int c) -> a + b + c //valid

If you are specifying parameter types, you must specify them for all the
parameters. Thus, (int a, int b, c) -> a + b + c would be
invalid because it does not specify the type of c
Java 11 has added one more way to declare the parameters. You can use the var
type for declaring the parameters like this:
(var a) -> a*a //valid. Observe the brackets, they are

required if you are using var.
Normally, there is no need to specify the type for the lambda parameters.
However, it is not possible to apply annotations on the parameters unless you
specify the type. For example, the following will not compile because
annotations (a topic not on the part 1 exam) cannot be specified directly on a
variable:(@NotNull a) -> a*a .
To be able to apply an annotation, you need to have a type for the variable, like
this: (@NotNull Integer a) -> a*a . So, if you want to use an
annotation on a lambda parameter but don't want to specify the type, you can use
the var declaration, like this: (@NotNull var a) -> a*a .
The syntax of the code part of a lambda expression is simple. It can either be an
expression or a block of code contained within curly braces. Given that the body
may or may not return a value, there are four possibilities:
1. Expression with or without a return value - This is the most common use
case and is therefore, the smallest. You can simply put an expression on the
right side of -> . If the expression has a return value and if the lambda
expression is supposed to return a value, the compiler will figure out that
the value generated by the expression is to be returned and will insert a
return statement on it own. You must not write the return keyword. For
example,
a -> a + 2 //valid
a - > return a + 2 //invalid, must not have return
keyword
Similarly, an expression that doesn't return any value can also be used
directly as the body of the lambda expression. For example,
(a, b) -> System.out.println(a+b)

//method call is a valid expression
2. Block of code with or without a return value - If you have multiple lines
of code, you must write them within curly braces, i.e., { }. If the expression
is supposed to return a value, you must use a return statement to return the
desired value.
This is pretty much the same as writing a method body with or without a
return value. You can use this syntax even if you have just one statement in
the body. For example, here is Lambda expression that returns a value:
(a) -> {
int x = 2;
int y = x+a;
return y;
}
and here is one that doesn't:
() -> {
int x = 2;
int y = 3;
System.out.println(x+y);
}
Observe that unlike lambdas with just an expression as their bodies, the
statements within the block end with a semi-colon. This is just like a regular
code block. All the rules that apply to code within a method body apply to
code within a lambda expression's code block as well. After all, the
compiler uses this code block to generate a method body.
The OCP Java 11 Part 1 exam does not try to trick you with complicated lambda
expressions. If you learn the basic rules shown above, you will not have any
trouble identifying valid and invalid lambda expressions.
14.1.3 Using Predicate ☝
Let's take a look at the CarFilter interface that we defined in our CarMall
example again:
interface CarFilter{
boolean showCar(Car c);
}
The whole purpose of this interface is to let you check whether a Car satisfies a
given criteria so that you could filter a list of Cars. Filtering through a list of
objects is such a common requirement in Java application that the Java standard
library includes a generic interface for this purpose -
java.util.function.Predicate . It looks like this:
interface Predicate<T>{
boolean test(T t);
}
The <T> part means that this interface can be typed for any class as explained in
the Generics section of the previous chapter. If the name of that class a
Predicate is typed to is T , then the method test will accept an object of
type T and return a boolean .
Let's change the code for CarMall 's showCars method to use Predicate
interface:
List<Car> showCars(Predicate<Car> cp){

ArrayList<Car> carsToShow = new ArrayList<>();
for(Car c : cars){
if(cp.test(c)) carsToShow.add(c);
}
return carsToShow;
}
Observe that we have typed Predicate to Car in the above code. Apart from
that, the above code is the same as previous one. But by using the Predicate
interface instead of writing a custom interface, we have eliminated another three
lines of code.
There is no change in the code that calls showCars method. The lambda
expression that we used earlier, i.e., cm.showCars(c ->
c.company.equals("Honda")) works for this new method as well. It
works because the lambda expression never required us to use the name of any
interface or method. Therefore, the lambda expression was not tied to a
particular interface or method. It was only tied to a particular behavior -, i.e., to a
method that takes Car as an argument and returns a boolean . We relied on
the compiler to produce an appropriate class with an appropriate method. We
supplied only the raw code for the method body. Earlier the compiler generated a
class that implemented the CarFilter interface and it now generates a class
that implements the Predicate<Car> interface using the same code that we
wrote in the lambda expression! In fact, the change is so subtle that if you have
both the versions of showCars method in CarMall , the compiler will reject
the line cm.showCars(c -> c.company.equals("Honda")) with
the error message, "reference to showCars is ambiguous. Both
method showCars(java.util.function.Predicate<Car>) in
CarMall and method showCars(CarFilter) in CarMall
match ".
I showed only one method in Predicate interface but it actually has three
default methods and one static method in addition to the abstract test
method. I didn't mention them before because they have nothing to do with
lambda expressions. You will notice that these methods are basically just helpful
utility methods. We haven't ever seen anyone getting any question on these but,
it is better to be aware of these methods.
1. default Predicate<T> and(Predicate<? super T>

other) : Returns a composed predicate that represents a short-circuiting
logical AND of this predicate and another. This is helpful when you have
more than one checks to be performed. For example, the following code
checks whether a Car satisfies two predicates using two separate
invocations of test() :
Predicate<Car> p1 = c -> c.company.equals("Honda")

;
Predicate<Car> p2 = c -> c.price>(20000.0);
Car c = ...
if(p1.test(c) && p2.test(c)
) System.out.println("yes");
Instead of making two calls to test() , you could combine the two
predicates into one and use only one call to test, like this:
Predicate<Car> p3 = p1.and(p2)
;
Car c = ...
if( p3.test(c)
) System.out.println("yes");
You do not need to worry about the ? super T part. It is related to

"bounded types" in generics, which is not on the part 1 exam.
2. default Predicate<T> negate() : Returns a predicate that

represents the logical negation of this predicate. For example, if you have
Predicate<Car> p = c -> c.price<20000; then
Predicate<Car> notP = p.negate(); represents a predicate
where price is >= 20000 .
3. default Predicate<T> or(Predicate<? super T>

other) : Returns a composed predicate that represents a short-circuiting
logical OR of this predicate and another. For example, if you have
Predicate<Car> isHonda = c ->
c.company.equals("Honda"); and Predicate<Car>
isToyota = c -> c.company.equals("Toyota"); then
Predicate<Car> isHondaOrToyota =
isHonda().or(isToyota); represents a predicate where the
company name is "Honda" or "Toyota" .
4. static <T> Predicate<T> isEqual(Object targetRef) :

Returns a predicate that tests if two arguments are equal according to
Objects.equals(Object, Object) . This method is a way to
convert a class's equals method into a Predicate . For example,
normally, you would compare two Car objects using c1.equals(c2) .
You could create a Predicate out of the equals method like this -
Predicate equals = Predicate.isEquals(c1); . and then
compare c1 with other Car objects using this Predicate, i.e.,
equals.test(c2) .
14.1.4 Functional Interfaces ☝
The Predicate interface that you saw just now is one among several
interfaces defined in java.util.function package. The Predicate
interface captures the use case for testing a condition. In the same way, Java
designers have identified several use cases that are encountered in day to day
Java development and have created readymade interfaces for them so that every
project doesn't have to define the same kind of interfaces again and again. These
interfaces belong to a category of interfaces known as "Functional interfaces".
Don't be alarmed by the fancy sounding name. Any interface that has exactly one
abstract method is a functional interface. It may have other static or default
methods but it must have one and only one abstract method. This abstract
method is also called the "functional method" of that functional interface. For
example, the functional method of the Predicate interface is boolean
test(T t) . Having exactly one abstract method is an important property
because it allows you to implement such interfaces using Lambda expressions.
The reason they are called functional interfaces is that they represent a single
function without maintaining any state. A functional interface is meant to do
exactly one thing. For example, if you take a look at the Predicate interface, you
will notice that it is meant to check a condition. But the logic for the actual
condition is not there in the functional method (because it is an abstract method).
The logic is provided by the code that implements the interface, for example,
through a lambda expression. The benefit of this approach is that it allows you to
separate the logic from the data upon which that logic is to be applied. That is
why it is called "functional programming". Observe that this is exactly opposite
to the approach adopted by Object-Oriented Programming, in which an object
contains data as well as the logic to operate on that data.
OCP Java 11 Part 1 exam does not mention the topic of Functional interfaces
explicitly but it contains questions that refer to a few functional interfaces
defined in the java.util.function package, namely, Predicate , Consumer
, and Supplier . The exam expects you to know the basic usage of these
interfaces without going into the details too much. You have already seen the
usage of the Predicate interface. I will talk about the rest next.
java.util.function.Consumer ☝
A consumer is meant to "consume" an object. By consume, we mean that we

want to do something with the given object. The functional method of the
Consumer interface is named accept(T t) . You just need to write the code
for what you want to do with the object passed in this method. For example:
//creating a Consumer that consumes a String.
Consumer<String> strConsumer = s -> System.out.printl

n(s.length());
strConsumer.accept("hello"); //prints 5
In the above code, the lambda expression contains only the logic for consuming
the String argument. It doesn't contain the actual data on which this logic is to be
applied. The actual data comes in the next statement.
java.util.function.Supplier ☝
A Supplier supplies an object whenever invoked. Its functional method is T

get() . Here is an example that illustrates its usage:
public List<Car> getCars(Supplier<Car> carSupplier){

List<Car> cars = new ArrayList<>();
for(int i=0; i<10; i++) cars.add(carSupplier.get()
);
return cars;
}
The above method could be invoked like this:
List<Car> cars = getCars(()->new Car()); //assuming Ca

r has a default constructor.
14.1.5 Using Functional Interfaces with Collections API ☝
Generics made Java collections type safe, while Lambda expressions together
with functional interfaces gave them super powers. Most of the classes and
interfaces of the collections API were updated in Java 8 to include methods that
accept functional interfaces so that programmers can invoke those methods with
lambda expressions. Things such as iterating, filtering, and replacing elements,
that took several lines of code can now be done by half a line of code.
Since the Part 1 exam explicitly mentions List and ArrayList, let's see a few
examples of how functional interfaces and Lists are used together.
The forEach method ☝
As you have seen before, iterating through a collection is very common

requirement. Prior to Java 8, a common way to iterate through a collection was
to use a regular for loop. The Collection interface actually extends
java.util.Iterable interface and so, it was also possible to use the
enhanced for loop (aka the for-each loop) for this purpose.
With Java 8, a default method named forEach(Consumer<E>

consumer) was added to the Iterable interface, which made it an absolute
delight to iterate through the elements of any collection. The following snippet
shows the old and the new way to do the same:
List<String> list = List.of("a", "b", "c");

//old way
for(String s : list){
}
//new way
list.forEach(s->System.out.println(s));
The new way has pretty much taken over completely because of its simplicity.
Similarly, the Map interface defines a default forEach method that takes a
BiConsumer instead of Consumer . A BiConsumer is similar to
Consumer except that it take two arguments instead of one. Its functional
method is void accept(T, U) . Here is an example of how it is used to
process the elements of a Map:
BiConsumer<String, Integer> bc = (s, i) -> System.out.

println(s+" is mapped to "+i);
Map<String, Integer> map = new HashMap<>();
map.put("One", 1);map.put("Two", 2);
map.forEach(bc);
//you could use the lambda expression directly as an a

rgument
map.forEach((s, i) -> System.out.println(s+" is mapped

to "+i));
You should get familiar with the usage of the forEach method because you
will see it used on the exam. There are no trick question in the exam on this
method though.
The removeIf method ☝

Another common operation performed with collections is filtering. The
Collection interface has a default removeIf(Predicate<? super
E> filter) method for this purpose. This method removes all of the
elements of this list that satisfy the given predicate. Here is a simple example:
List<Integer> iList = List.of(1, 2, 3, 4, 5, 6);

Predicate<Integer> p = x->x%2==0;
iList.removeIf(p);
System.out.println(iList);
The above code prints [1, 3, 5] . As you have probably guessed, the lambda
expression returns true if an element is even. The removeIf method
executes this lambda expression for each of the elements in the list and removes
the element if the expression returns true for that element.
Again, we didn't really have to use the variable p in the above code. We could
have passed the lambda expression directly to the removeIf method, i.e.,
iList.removeIf( x->x%2==0 );
The sort method ☝
A collection has no notion of order but a list does. It makes sense, therefore, that
List interface has a default sort(Comparator<? super E>
comparator) method which allows you to sort the elements of this list using
the sorting order determined by the comparator. The
java.util.Comparator interface has been around since Java 1.2 but it has
been made a functional interface in Java 8. Its functional method is int
compare(T o1, T o2) , which compares its two arguments and returns a
negative integer, zero, or a positive integer if the first argument is less than,
equal to, or greater than the second. This helps us sort a list in a single line of
code as the following example shows:
List<String> games = new ArrayList<>(List.of("football

", "cricket", "baseball", "tennis"));
games.sort( (a, b)->a.compareTo(b));
games.forEach(s->System.out.println(s));
The output is:
baseball
cricket
football
tennis
If you want to sort the list in the reverse order, you can change the lambda
expression to, (a, b)->-a.compareTo(b) . Observe the minus sign in
front of a.compareTo(b) .
None of the methods above are mentioned explicitly in

the official exam objectives. However, we have seen
questions on the exam that refer to these methods. The
questions are not tricky and you should be able to
answer them with the information given above.
14.1.6 Scope of variables in a lambda expression ☝
There is just one more thing that you need to know about lambda expressions.
The variables that you define in the variable section of a lambda expression exist
in the same scope as which the lambda expression itself exists. This means, you
cannot redefine the variables that already exist in that scope. This is illustrated in
the following code:
List<String> names = Arrays.asList(new String[]{ "alex

", "bob", "casy", "abel"});
for(String n : names){
Predicate p = n->n.startsWith("a"); //will not co
mpile
if(p.test(n)) {
System.out.println(n);
}
}
The above code will fail to compile with an error message that says,
"variable n is already defined" . This is because n is already in
scope within the for loop as well as in the lambda expression. So, when the
lambda expression tries to define n , the compiler complains.
It is possible to access a variable that is in scope of the lambda expression from

within the lambda expression's body but only if that variable is declared as final
or is "effectively final". Effectively final means that even though the variable is
not explicitly declared as final, its value is not changed though out the scope in
which it exists, and so, the compiler assumes it as final. Here is an example:
List<String> names = Arrays.asList(new String[]{ "alex

", "bob", "casy", "abel"});
int x = 0;
for(String n : names){
Predicate<String> p = k->{
System.out.println(n); //valid
//System.out.println(x); //will not compile
return k.startsWith("a");
};
if(p.test(n)) {
System.out.println(n);
}
}
x = 1; //x is being changed here
In the above code, the variable n is never changed after it is assigned a value.
Thus, n is effectively final. But x is not because it is assigned a new value later
in the code.
14.1.7 Quiz ☝
Q. Given the following class:
public class Item{

String name; double price;
public Item(String name, double price){
this.name = name; this.price = price;
}
}
and the following code appearing in another class
double price = 100.0;

Supplier<Item> itemSupplier = ()-> new Item("dummy", p
rice);
price = 200.0;
System.out.println(itemSupplier.get().price);
What will be the output?
A. 100.0
B. 200.0
C. Nothing is printed.
D. Compilation failure.
E. An exception is thrown at run time.
Correct answer is D.
Observe that the variable price is not effectively final because its value is
being changed in the code. Therefore, it cannot be used in the lambda
expression.
14.2 Exercise ☝
1. Given the following lambda expressions, define appropriate interfaces that
can be implemented using these lambda expressions.
() -> true
k -> k>5
2. Given the following interfaces, create lambda expressions that can be used
to capture these interfaces.
interface Shape{
double computeArea();
}
interface Operation{
void operate(String name, double[] params);
}
3. Write a method that takes a list of Image objects and a Predicate as

arguments, and returns another list containing only those Images that satisfy
the predicate.
4. Assuming that the Image class has width and height properties,
invoke the above method that filters out images that are smaller than 100 x
100.
Chapter 15 Handling Exceptions
checked, unchecked exceptions, and Errors
Create try-catch blocks and determine how exceptions alter program flow
15.1 Create try-catch blocks and determine how

exceptions alter program flow
15.1.1 Java exception handling ☝
Exceptions are for managing exceptional situations. For a file copy program, the
normal course of action could be - open file A, read the contents of file A, create
file B, and write the contents to file B. But what if the program is not able to
open file A? What if the program is not able to create or write to file B? There
could be many reasons for such failure such as a typo in the file name, lack of
permission, no space on disk, or even disk failure. Now, you don't expect these
problems to occur all the time but it is reasonable to expect them some times.
Since we don't expect these problems to occur regularly in normal course of
operation, we call them "exceptional".
In exceptional situations, you may want to give the user a feedback about the
error or you may even want to take an alternative course of action. For example,
you may want to let the user to input another source file name if the input file is
not found or another target location if the given location is out of space. Even if
you don't want to take any special action upon such situations, you should at
least want your program to end gracefully instead of crashing unexpectedly at
run time. This means, you should provide a path for the program to take in such
situations. One way would be to check for each situation before proceeding to
copy. Something like this:
if(checkFileAccess(file1)){
if(checkWritePermission(targetDirectory){
//code for normal course of action
}else{
System.out.println("Unable to create file2");
}
}else{
System.out.println("Unable to read file1");
}
You can see where this is going. You will end up having a lot of if-else
statements. Not only is it cumbersome to code, it will be a nightmare to read and
maintain later on.
There is another serious problem with the above approach. It doesn't provide
way to write code for situations that you don't even know about at the time of
writing the code. For example, what if the user runs this code in an environment
that requires file names to follow a particular format? So, now, you have two
kinds of "exceptional situations". One that you know about at the time of writing
the code, and one that you don't know anything about.
Java exception mechanism is designed to help you write code that covers all
possible execution paths that a program may take - 1. path for normal operation
2. paths for known exceptional situations, and 3. path for unknown exceptional
situations. Here is how the above mentioned program can be written:
try{
//code for normal course of action
}catch(SecurityException se){
//code for known exceptional situation
System.out.println("No permission!");
}
catch(Throwable t){
//code for unknown exceptional situations
System.out.println("Some problem in copying: "+t.ge

tMessage());
t.printStackTrace();
}
Observe that the perspective is reversed here. In the if/else approach, you check
for each exceptional situation and proceed to copy if everything is good, while in
the try/catch approach, you assume everything is good and only if you encounter
a problem, you decide what to do depending on the problem. The benefits of the
try/catch approach are obvious. It provides a clean separation between code for
normal execution and code for exceptional situations, which makes the code
easier to read and maintain. It allows an alternative path to the program to
proceed even in cases where the programmer has not anticipated the problem.
Exception handling in Java is not a bag of all goodies.

There has been a good amount of debate on what
constitutes "exceptional situations" and what should be
the right approach to handle such situations. My
objective in this chapter is to teach you Java's
approach to exception handling and so, I will not go
into an academic discussion on whether it is good or
bad as compared to other languages. However, it is a
very important topic of discussion in technical
interviews. I suggest you google criticism of Java's
exception handling and compare it with C#'s.
15.1.2 Fundamentals of the try/catch approach ☝
When you think of developing code as developing a component, you will realize
that there are always two stakeholders involved - the provider/developer of the
component and the client or the user of the component. For the client to be able
to use the component, it is imperative for the provider to tell the client about
"how" his component works. The how not only includes the input/output details
of the component but also the information about exceptional situations, i.e.,
about situations the component knows may occur but does not deal with.
For example, let's say you are developing a method that copies the contents of
one file to another and that this method is to be used by a developer in another
team. You must convey to the other developer that you are aware of the
Input/Output issues associated with reading and writing a file but you won't do
anything about them. In other words, you must convey that your method will try
its best to copy the file but if there is an I/O issue, it will abandon the attempt
and let her know about the failure so that she can deal with it however she want
to. This is done through the use of a "throws" clause in method declaration:
public void copyFile(String inputPath, String outputPa

th) throws IOException
{
//code to copy file
In the above code, if the copyFile method encounters any I/O issue while
copying a file, it will simply abort the copying and throw an IOException to
the caller. If it does not encounter any I/O issue, the method will end
successfully.
On the other side of the component is the user of that component, who uses the
information provided by the component provider to develop her code. The user
has to decide how she wants to handle the exceptional situation. If she believes
that she has the ability to "resolve" the situation, she will put the usage of that
component in a "try" block with an associated "catch" block that contains
code for "resolution" of the problem. When that exceptional situation actually
occurs, the control goes to the catch block instead of proceeding to the next
statement after the method call that threw the exception.
If the user decides that she cannot handle the exceptional situation either, she
propagates the exception to the caller of her component.
For example, a program that creates backup of a file may use the file copy
program internally to copy a file. If the file copy program throws an
IOException , the backup creator program may catch that exception and
show a message to the user.
public void createBackup(String input) {

String output = input+".backup";
try{
copyFile(input, output);
System.out.println("backup successful");
}catch(IOException ioe
){
System.out.println("backup failure");
}
}
If the copyFile method completes without any issue, the control will go to the
println statement. If the copyFile method throws an IOException , the
control will go to the catch block, thus, providing an opportunity to take a
different course of action. Here, the code may show a failure message to the user.
From the perspective of the developer of the backup program, showing the error
message to the user is the resolution of the I/O problem. One can also write code
to have the user specify another file or directory to create a backup.
An exception is considered "handled" when it is caught in a catch block. If the

backup creator method doesn't know what to do in case the copyFile is not
successful, then it should let the exception propagate to its caller by using a
throws clause of its own:
public void createBackup(String input) throws IOExcept

ion
{
String output = input+".backup";
copyFile(input, output);
System.out.println("backup successful");
}
In the above code, if the copyFile method completes without any issue, the
control will go to the println statement. But if the copyFile method throws
an IOException , the createBackup method will end immediately and the
caller will receive an IOException . Note that this will be the same
IOException that it receives from the copyFile method. This is how an
exception propagates from one method to the other. There could be a long chain
of method calls through which an exception bubbles up before it is handled. If an
exception is not caught anywhere while it is bubbling up the call chain, it
ultimately reaches the JVM code. Since the JVM has no idea about the business
logic of the program that it is executing, it "handles" the exception by killing(aka
terminating) the program (actually, the JVM kills only the thread that was used
to invoke the chain of method calls but threading is not on the OCP Java 11 Part
1 exam, so you can assume that a program is composed of only one thread and
killing of that thread is the same as killing the program).
15.1.3 Pieces of the exception handling puzzle ☝
Java exception handling mechanism comprises several moving parts. What

makes this mechanism a bit complicated is that you need to put each part in its
right place to make the whole thing work. So let me introduce these parts first
briefly.
The java.lang.Throwable object - ☝
Throwable is the root class of all exceptions. It captures the details of the
program and its surroundings at the time it is created. Among other things, a
Throwable object includes the chain of the method calls that led to the
exception (known as the "stack trace") and any informational message specified
by the programmer while creating that exception. This information is helpful in
determining the location and the cause of the exception while debugging a
program. You may have seen crazy looking output on the console containing
method names upon a program crash. This output is actually the stack trace
contained in the Throwable object.
Throwable has two subclasses - java.lang.Error and

java.lang.Exception , and a huge number of grand child classes. Each
class is meant for a specific situation. For example, a
java.lang.NullPointerException is thrown when the code tries to
access a null reference or an
java.lang.ArrayIndexOutOfBoundsException is thrown when you
try to access an array beyond its size. You can also create your own subclasses
by extending any of these classes if the existing classes don't represent the
exceptional situation appropriately. For example, an accounting application
could define a LowBalanceException for a situation where more money is
being withdrawn from an account than what the account has.
What is called "exception" (i.e. exception with a lower case 'e') in common
parlance, is in reality is an object of class java.lang.Throwable or one of
its subclasses.
One subclass of java.lang.Exception that is particularly important is

java.lang.RuntimeException . RuntimeException and its
subclasses belong to a category of exception classes called "unchecked
exceptions" . You will see their significance soon.
The throw statement ☝
The throw statement is used by a programmer to "throw an exception" or "raise

an exception" explicitly. A programmer may decide to throw an exception upon
encountering a condition that makes continuing the execution of the code futile.
For example, if, while executing a method, you find that the value of a required
parameter is invalid, you may throw an IllegalArgumentException
using a throw statement like this:
public double computeSimpleInterest(double p, double r

, double t){
if( t<0) {
IllegalArgumentException iae = new IllegalArgume
ntException("time is less than 0");
throw iae;
}
//other code
Usually, an exception object is created only to be "thrown" and so there is no

need store its reference in a variable. That is why it is often created and thrown
in the same statement as shown below:

, double t){
if( t<0) throw new IllegalArgumentException("time is
less than 0");
//other code
Throwing an exception implies that the code has encountered an unexpected

situation with which it does not want to deal. The code shown above, for
example, expects the time argument to be greater than zero. For this code, time
being less than zero is an unexpected situation. It does not want to deal with this
situation (probably because the programmer is not sure what to do in this case)
and so it throws an exception in such a situation. This also means that this
method passes on the responsibility of determining what should be done in case
time is less than zero to the user of this method.
Note that only an instance of Throwable (or its subclasses) can be thrown using
the throw statement. You cannot do something like throw new
Object(); or throw "bad situation";
Explicitly throwing an exception using the throw statement is not the only way
an exception can be thrown. JVM may also decide to throw an exception if the
code tries to do some bad thing like calling a method on a null reference. For
example:
public void printLength(String str){

System.out.println(str.length());
}
If you pass null to the above method, the JVM will create a new instance of
NullPointerException and throw that instance when it tries to execute
str.length() .
The throws clause ☝
Java requires that you list the exceptions that a method might throw in the
throws clause of that method. This ties back to Java's design goal of letting the
user know of the complete behavior of a method. It wants to make sure that if
the method encounters an exceptional situation, then the method either deals
with that situation itself or lets the caller know about that situation. The throws
clause is used for this purpose. It conveys to the user of a method that this
method may throw the exception mentioned in the throws clause. For
example:

, double t) throws Exception
{
if( t<0) throw new Exception("time is less than 0");
//other code
Now, anyone who uses the above method knows that this method may throw an
exception instead of returning the interest. This helps the user write appropriate
code to deal with the exceptional situation if it arises.
The try statement ☝
A try statement gives the programmer an opportunity to recover from and/or

salvage an exceptional situation that may arise while executing a block of code.
A try statement consists of a try block , zero or more catch blocks , and an
optional finally block . Its syntax is as follows:
try {
//code that might throw exceptions
}catch(<ExceptionClass1> e1){
//code to execute if code in try throws exception 1
}
catch(<ExceptionClass2> e2){
//code to execute if code in try throws exception 2
}
catch(<ExceptionClassN> en){
//code to execute if code in try throws exception N
}
finally{
//code to execute after the try block and catch bloc
k finish execution
Note that curly braces for the try , catch , and finally blocks are required
even if there is a single statement in these blocks (compare that to if , while ,
do/while and for blocks where curly braces are not required if there is only
one statement in the block). Furthermore, a try block must follow with at least
one catch block or the finally block . A try block that follows with neither a
catch block nor a finally block will not compile.
The try statement is the counterpart of the throw statement because putting a
piece of code within a try block signifies that the programmer wants do
something in case that piece of code throws an exception. In other words, a try
block lets the programmer deal with an exceptional situation as opposed to the
throw statement, which lets the programmer avoid dealing with it.
While a try block contains code for normal operation of the program, a catch
block is the location where the programmer tries to recover from an exceptional
situation that arises in the try block. If the code in the try block throws an
exception, the normal flow of execution in that try block is aborted (i.e. no
further code in the try block is executed) and the control goes to the catch block.
Here, the programmer can take alternate approach to finish the processing of the
method. For example, the programmer may decide to just show a popup message
to the user about the exception and move on to the next statement after the try
statement.
A catch block is associated with a catch clause , which specifies the exception
that the catch block is meant to handle. For example, a catch block with the
catch clause as catch(IllegalArgumentException e) is meant to
handle IllegalArgumentException and its subclasses. Thus, this catch
block will be executed only if the code in the try block throws an
IllegalArgumentException or its subclass exception. You can specify
any valid exception class (including java.lang.Throwable ) in the catch
clause.
A finally block is the location where the programmer tries to salvage the
situation or control the damage so to say, without attempting to recover from an
exception. For example, a program that tries to copy a file may want to close any
open files irrespective of whether the copy operation is successful or not. The
programmer can do this in the finally block. You may think of a finally block as
the step where a car mechanic reassembles the parts back irrespective of whether
he was able to fix the car or not!
Here is a method that calls the computeSimpleInterest method shown

above within a try statement:
public void doInterest(){

try{
double interest = computeSimpleInterest(1000.0,
10.0, -1);
System.out.println("Computed interest "+interest
);
System.out.println("Problem in computing interes
t:"+e.getMessage());
}finally{
System.out.println("all done");
}

}
In the above code, the call to computeSimpleInterest throws an

IllegalArgumentException because t is negative. Thus, the println
statement after the method call is not executed. The exception is caught in the
catch block because its catch clause, i.e.,
catch(IllegalArgumentException ) matches with the exception that
is thrown by the try block and the control goes to the first statement in this catch
block. It prints the exception's message and the stack trace on the console.
Finally, the control goes to the code in the finally block, where it prints "all
done" . If you omit the catch block, the control will go directly to the finally
block after the invocation of computeSimpleInterest method. After the
execution of the code in the finally block, the caller of doInterest method
will receive the same IllegalArgumentException .
Note that a finally block, if present, always executes irrespective of what

happens in the try block or the catch block. Even if the try block throws an
exception and there is no catch block to catch that exception, the JVM will
execute the finally block. It will throw the exception to the caller only after the
finally block finishes. The only case where the finally block is not executed is if
the code in the try or the catch block forces the JVM to shut down using a call to
System.exit() method.
The following is a complete program that illustrates how an exception alters the
normal program flow:

tc.doInterest();
}

public double computeSimpleInterest(double p, double
r, double t){
if( t<0) throw new IllegalArgumentException("time
is less than 0");
return p*r*t/100;
}
public void doInterest(){

try{
double interest = computeSimpleInterest(1000.0
, 10.0, -1);
System.out.println("Computed interest "+intere
st);
}catch(IllegalArgumentException iae){
System.out.println("Problem in computing inter
est:"+iae.getMessage());
iae.printStackTrace();
}finally{
System.out.println("all done");
}
}
It generates the following output on the console :

Problem in computing interest:time is less than 0
java.lang.IllegalArgumentException: time is less than
0
at TestClass.computeSimpleInterest(TestClass.j
ava:8)
at TestClass.doInterest(TestClass.java:14)
at TestClass.main(TestClass.java:4)
all done
You should observe the following points in the above code:
1. The return statement in computeSimpleInterest is not executed

because the previous statement throws an exception.
2. The println statement in the try block of doInterest is not executed
because the call to computeSimpleInterest ends with an exception
instead of a return value. Control goes to the catch block directly after the
method call.
3. The catch block prints the details captured in the exception object. It
shows the sequence of the method invocations in reverse order from the
point where the IllegalArgumentException object was created.
You should try removing the catch block and see the output.
4. Once the catch block is finished, the control goes to the finally block.
5. The doInterest method returns after the execution of the finally
block.
6. There is no throws clause in computeSimpleInterest method even
though it throws an exception. The reason will be clear in the next section
where I talk about checked and unchecked exceptions.
15.2 Differentiate among checked, unchecked

exceptions, and Errors
15.2.1 Checked and Unchecked exceptions ☝
As discussed earlier, exceptions thrown by a method are part of the contract

between the method and the user of that method. If the user is not aware of the
exceptions that a method might throw, she will be blindsided during run time
because her code would not be prepared to handle that exception. The throws
clause of a method is meant to list all exceptions that the method might throw. If
an exception is listed in the throws clause, the user of the method will know that
she needs to somehow handle that situation.
But what if a developer forgets to list an exception in the throws clause of a

method? In that case, we are back to the same problem of blindsiding the user of
that method at run time. This is where the compiler plays an important role.
While compiling a method, the compiler checks for the possibility of any
exception that might get thrown by the method to the caller. If that exception is
not listed in the throws clause of that method, it refuses to compile the method.
This sounds like a good solution, but the problem here is that the compiler does
not check for all kinds of exceptions thrown by a piece of code. It checks for
only a certain kind of exceptions called "checked exceptions" and forces you
to list the only those exceptions that belong to this category of exceptions in the
throws clause.
Exceptions that do not belong to the category of checked exceptions are called
"unchecked exceptions" . They are called unchecked because the compiler
doesn't care whether a piece of code throws such an exception or not. Listing
unchecked exceptions in the throws clause is optional.
Finding out whether an exception is a checked exception or not is easy. Java

language designers have postulated that any exception that extends
java.lang.Throwable but does not extend
java.lang.RuntimeException or java.lang.Error is a checked
exception. The rest (i.e. java.lang.Error
, java.lang.RuntimeException , and their subclasses) are unchecked.
The following figure illustrates this grouping of exceptions.

Note that Java has a deep rooted exception class hierarchy, which means there
are several classes, subclasses, and subclasses of subclasses in the tree of
exceptions. So just because an exception is a RuntimeException , does not
mean that that exception directly extends RuntimeException . It could even
be a grand child of RuntimeException . For example,
ArrayIndexOutOfBoundsException actually extends
IndexOutOfBoundsException , which in turn extends
RuntimeException . Similarly, just because an exception is a checked
exception does not mean that it directly extends Exception. It may extend any
subclass of Exception (except RuntimeException , of course).
Rationale behind checked and unchecked exceptions ☝
Recall our discussion on exceptional situations where I talked about two kinds of
exceptional situations - ones that a developer knows about and ones that a
developer doesn't expect to occur at all. While a developer may want to provide
an alternate path of execution in case of a situation that is known to occur but
there is no point in providing an alternate path of execution for a situation that is
never expected to occur.
For example, if a piece of code tries to write to a file, the developer may want to
take a different approach if he is not able to write to the file system. But if the
data array that it is trying to write is null, there is nothing much he can do. Such
unexpected situations occur mostly due to badly written code. In other words, if
a code gets itself into an unexpected situation, it is most likely because of a
programming error, i.e., a bug in the code. Such issues should be fixed during
development itself. But if they do occur in production, they should rather be
handled at a much higher level than at the component level.
Unchecked exceptions are for such unexpected situations. Java language

designers believe that unchecked exceptions need not be declared in a method
declaration because there is nothing to gain by forcing the caller to catch them.
Only checked exceptions need to be declared because the caller of the method
may have a plan to recover from them. There are two kinds of unchecked
exceptions - exceptions that extend java.lang.RuntimeException (aka
runtime exceptions ) and exceptions that extend java.lang.Error (aka
errors ).
Runtime exceptions and errors ☝
Characterizing a situation as expected or unexpected is a design decision that

depends on the business purpose of the method. One method may expect to
receive a null argument and may work well if it gets a null argument, while
another may not expect a null argument and may end up throwing a
NullPointerException when it tries to access that null. In the second
case, passing a null to that method would be considered a bug in the code, which
must be identified and fixed during testing. Such exceptions that signify the
presence of a bug in the code are categorized as runtime exceptions .
Here, the word runtime in RuntimeException does not imply that only
exceptions that extend RuntimeException can be thrown at run time. All
exceptions are thrown only when the program is executed, i.e., at run time. It
refers to the fact that the developer comes to know of the occurrence of the
situation that results in a RuntimeException only when the program is run, i.e.,
during run time. Had the programmer anticipated the occurrence of that situation
during compile time, he would have fixed the code, and in which case the
exception would not have been thrown upon running the program. For the same
reason, runtime exceptions are usually not thrown explicitly using the throw
statement. Indeed, why would you write the code to throw an exception if you
don't even expect that situation to occur. The JVM throws runtime exceptions on
its own when it encounters an unexpected situation. For example, if you try to
access a null reference, the JVM will throw a NullPointerException , or
if you try to access an array beyond its size, the JVM will throw an
ArrayIndexOutOfBoundsException . It is possible to recover from
runtime exceptions but ideally, since they indicate bugs in the code, you should
not attempt to catch them and recover from them. A well written program should
not cause the JVM to throw runtime exceptions.
The case of errors is similar to runtime exceptions . The difference is that

errors are reserved for situations where the operation of the JVM itself is in
jeopardy. For example, a badly written code may consume so much memory that
there is no free memory left. Once that happens, the JVM may end up throwing
OutOfMemoryError . Similarly, a bad recursion may cause
StackOverflowError from which no recovery is possible. Errors signify
serious issues in the interaction between the code and the JVM and are thrown
exclusively by the JVM. It is never a good idea to throw them explicitly or to try
to recover from them because the code will likely not work as expected anyway
once the JVM starts throwing Errors.
Although the exam does not focus on the reason for

categorizing exception between checked and unchecked
exceptions, it is actually a very important topic to
understand for a professional developer. You should
also compare Java's exception handling with C#'s.
Identifying exceptions as checked or unchecked ☝
Java follows a convention in naming exception classes. This convention is

sometimes helpful in determining the kind of exception you are dealing with.
The name of any class that extends Error ends with Error and the name of any
class than extends Exception ends with Exception . For example,
OutOfMemoryError and StackOverflowError are Errors while
IOException , SecurityException ,
IndexOutOfBoundsException are Exceptions.
However, there is no way to distinguish between unchecked exceptions that

extend RuntimeException and checked exceptions just by looking at their
names. It is therefore, important to memorize the names of a few important
runtime exception classes, namely - NullPointerException ,
ArrayIndexOutOfBoundsException , ArithmeticException ,
ClassCastException , and SecurityException .
15.2.2 Commonly used exception classes ☝
The Java standard library contains a huge number of exception classes but the
exam expects you to know about only a few of them, which I will cover now.
These exception classes are important because you will encounter them quite
often while working with Java.
You may have read about the classification of exceptions based on whether they
are thrown by the programmer or by the JVM. It is true that there are a few
Exceptions that are thrown by the JVM on its own. However, there is no
Exception that is thrown exclusively by the JVM. In fact, many methods of the
Java standard library classes throw the same exceptions explicitly using the
throw statement.
I have omitted the package name from the classes below for brevity but note that
all of the following exception classes belong to the java.lang package.
Common Exceptions that are usually thrown by the JVM ☝
1. ArithmeticException extends RuntimeException

The JVM throws this exception when you try to divide a number by zero.
Example :
public class X {
static int k = 0;
k = 10/0; //ArithmeticException
}
}
2. ClassCastException extends RuntimeException

The JVM throws this exception when you try to cast a reference variable to
a type that fails the IS-A test. Don't worry I will discuss this in detail in the
next chapter.
Example :
Object s = "asdf";
StringBuffer sb = (StringBuffer) s; //ClassCastExc

eption because s refers to a String and not a Stri
ngBuffer.
3. IndexOutOfBoundsException extends RuntimeException

This exception is a common superclass of exceptions that are thrown where
an invalid index is used to access a value that supports indexed access.
For example, the JVM throws its subclass

ArrayIndexOutOfBoundsException when you attempt to access an
array with an invalid index value such as a negative value or a value that is
greater than the length (minus one, of course) of the array. Methods of
String class throw another of its subclass
StringIndexOutOfBoundsException when you try to access a
character at an invalid index.
Example :
int[] ia = new int[]{ 1, 2, 3}; // ia is of length

3.
System.out.println(ia[-1]); //ArrayIndexOutOfBound
sException
System.out.println(ia[3]); //ArrayIndexOutOfBounds
Exception
System.out.println("0123".charAt(4));//StringIndex
OutOfBoundsException
4. NullPointerException extends RuntimeException

The JVM throws this exception when you attempt to call a method or
access a field using a reference variable that is pointing to null.
Example :
String s = null;
System.out.println(s.length()); //NullPointerExcep
tion because s is null.
Common Errors usually thrown by the JVM ☝
1. ExceptionInInitializerError extends Error

The JVM throws this Error when any exception is thrown while initializing
a static variable or a static block.
Example :
public class X {
static int k = 0;
static{
k = 10/0; //throws ArithmeticException but i
t is wrapped into an ExceptionInInitializationError
}
}
2. OutOfMemoryError extends Error

The JVM throws this Error when it runs out of memory. This usually
happens when a programs creates too many objects.
Example:
public static void main (String[] args) {

StringBuilder sb = new StringBuilder("a long st
ring");
for(int i=0; i<Integer.MAX_VALUE; i++){
sb.append(sb.toString());
}
}
3. StackOverflowError extends Error

The JVM throws this Error when the thread executing the method runs out
of stack space. This usually happens when a method calls itself recursively
and there is no boundary condition to stop the recursion.
Example :
public void m1(int k){

m1(k++); //StackOverflowError
Exceptions thrown by Application Programmer ☝
All instances of Exception and its subclasses (except RuntimeException

s) are generally thrown by the application programmer using the throw
statement. In some cases, application programmer may throw
RuntimeException s as well.
1. IllegalArgumentException extends RuntimeException

This exception is thrown when a method receives an argument that the
programmer has determined is not legal.
Example:
public void processData(byte[] data, int datatype)

{
if(datatype != 1 && datatype != 2)
throw new IllegalArgumentException("Invali
d datatype "+datatype);
else System.out.println("Data Processed.");
}
2. NoClassDefFoundError extends Error

Although it is an Error but it is not thrown by the JVM. It is thrown by a
class loader (which is just another class in Java standard library) when it is
not able to find the definition of a class that it is trying to load. Beginners
get this error often while trying to run their program. For example, if your
class has a package statement but you have not put the class file in its
proper directory structure, the system class loader will not be able to find
that class when you try to run it and will throw this error.
3. NumberFormatException extends
IllegalArgumentException
This exception is thrown when a method that converts a String to a
number receives a String that it cannot convert.
Example:
int i = Integer.parseInt("asdf");//a NumberFormatE

xception will be thrown by the parseInt method
4. SecurityException extends RuntimeException
This exception is thrown if the Security Manager refuses to permit the
requested operation due to restrictions placed by the JVM. For example,
when a Java program runs in a sandbox (such as an applet) and tries to use
prohibited APIs such as File I/O, the security manager throws this
exception. Since this exception is explicitly thrown using the new keyword
by a security manager class, it can be considered to be thrown by the
application programmer.
15.3 Create and invoke a method that throws an

exception
15.3.1 Creating a method that throws an exception ☝
Now that you know about the throws clause and the types of exceptions, let us
look at the rules for creating a method that throws an exception. Actually, there
is just one rule - If there is a possibility of a checked exception getting thrown
out of a method, then that exception or its superclass exception must be declared
in the throws clause of the method. The following are examples of a few valid
methods that illustrate this rule:
1. void foo(int x) throws Exception
{
if(x == 2) throw new Exception(); //throws Exc
eption only if x==2
else return;
}
It doesn't matter whether the code throws an exception every time it runs or
only some times. If there is a path of execution in which an exception will
be throws, the method must list that exception in the throws clause.
2. void foo() { //no throws clause necessary
if(someCondition) throw new RuntimeException()

;
else throw new Error();
}
RuntimeException and Error are unchecked exceptions and are

therefore, exempt from being declared in the throws clause. Declaring them
in the throws clause is valid though.
3. void foo() throws Exception
{
if(someCondition) throw new java.io.IOExceptio
n(); //throwing a sub-exception
else return;
}
It is ok to declare a superclass exception in the throws clause and throw a

subclass exception in the method. But the reverse is not acceptable:
void foo() throws java.io.IOException
{ //will not compile
if(someCondition) throw new Exception(); //thr

owing a super-exception
else return;
}
Remember that a throws clause is a commitment that you give to the user of
this method that this method may throw only those exceptions that are listed
in the throws clause. If you commit that you may only throw an
IOException but then throw an Exception instead, then the caller
will have a problem because the caller is only prepared to handle an
IOException and not an Exception . Throwing an Exception will
break the caller's code.
Declaring a broader exception (i.e. a superclass exception) and throwing a

narrower exception (i.e. a subclass exception) is ok because if the caller is
already prepared to handle the broader exception it can handle the narrower
one without changing any code. For example, an IOException is
an Exception and by throwing IOException , you are not breaking
your commitment if you have committed that you may throw an
Exception .
4. void foo() throws java.io.IOException, java.sql.SQ

LException
//can throw a common superclass exception as well.
{
n();
else throw new java.sql.SQLException();
}
If a method throws multiple exceptions, you can either list them

individually or list a common superclass. The logic behind this is the same
as before. As long as the method is true to what is has committed in its
declaration, it is good!
5. void foo() throws Exception, java.io.IOException

{//specifying IOException is redundant because IOE
xcception is a subclaclass of Exception

n();
else throw new java.sql.SQLException();
}
It is ok for a method to declare a superclass exception as well as a subclass

exception even though adding the subclass exception to the list when a
superclass is present is redundant.
6. void foo() throws Exception
{ //declaring Exception in the throws clause even

though it is not thrown by the method body
}
A method can declare any exception in its throws clause irrespective of

whether that method actually throws that exception or not. It is sometimes
useful to "future-proof" a method by declaring Exception in the throws
clause if you believe that the method's implementation may change later. By
declaring Exception in the throws clause, the users of the method will not
have to change their code if the method actually starts throwing Exception
or any of Exception subclasses later.
7. void foo1() {
try{
if(someCondition) throw new Exception(); //
will be caught by the catch block
else return;

}
}
The requirement to list an exception in the throws clause is applicable only

when an exception is thrown out of the method to the caller. If the code
inside a method throws an exception but that exception is caught within the
method itself, there is no need to declare it in the throws clause of the
method.
Remember that an exception can only be caught by a catch block and not by
a finally block. Therefore, the throws clause is necessary in the following
method:
void foo2() throws Exception
{
try{
if(someCondition) throw new Exception();
else return;
}finally {
System.out.println("in finally"); //will be
executed but the exception is not caught here
}
}
8. void foo() throws Throwable
{
if(true) throw new Exception();
else return;
}
As you know, Throwable is the root of all exception classes, and

therefore, if you declare Throwable in the throws clause, you can throw
any exception in the method.
Although not important for the exam, deciding which

exception to throw and which to declare is an important
matter.
Declaring exceptions ☝
As shown in point number 8 above, declaring a broad

exception class in the throws clause is an easy way to
get rid of any compilation errors with the method if the
method throws different kinds of exceptions from
different parts of its code. However, this is considered
a bad practice because this burdens the user of the
method with dealing with a broad range of exceptions.
On the other hand, listing specific exception classes

individually restricts the future modifiability of a method
because throwing new exceptions later will break other
people's code.
It is recommended to be balanced in your approach

towards listing exceptions in the throws clause. You
should try to be only as specific as is possible without
compromising the modifiability of the method. For
example, I/O related methods may encounter different
kinds of issues while reading a file. It may not be
possible to identify all such issues while writing the
method. New issues may be discovered later and you
may have to modify your method to accommodate
those. Therefore, it is better to declare a common
superclass IOException in the throws clause instead of
individual subclasses such as FileNotFoundException
or EOFException .
Throwing exceptions ☝
You should always throw the most specific exception

possible. For example, if you don't find the file while
trying to open it, you should throw a
FileNotFoundException instead of an IOException
or Exception . By throwing the most specific
exception, you give more information to the caller about
the problem. This helps the caller in determining the
most suitable resolution of the problem.
15.3.2 Throwing exceptions from initializers and constructors ☝
Throwing exceptions from static initializers ☝
The JVM executes static initializers automatically when the it loads a class.
Although the JVM loads a class due some action taken by the application code
such as creating an object of the class or invoking a static method of the class,
there is no direct invocation of a static initializer from the application. Therefore,
if a static initializer ends up throwing an exception, there is no way for the
application to handle that exception and to recover from it. For this reason, a
static initializer is not allowed to throw any checked exception. If the compiler
sees a possibility of a checked exception getting thrown out of a static initializer,
it will generate an error. For example, the following code will not compile:

static int i = 5;
static{
if(i == 0) throw new Exception();
}
}
The error generated by the compiler message says, "Error: unreported

exception java.lang.Exception; must be caught or
declared to be thrown" , which is a bit misleading because there is no
where to declare the exception!
Throwing exceptions from instance initializers ☝
Instance initializers are always executed when the application code tries to create
an instance of the class. Thus, it is possible for the application code to catch an
exception thrown by an instance initializer and for the same reason, instance
initializers are allowed to throw checked exceptions. However, an instance
initializer has the same problem as a static initializer - there is no way to specify
a throws clause for an instance initializer. Recall that an instance initializer is
executed no matter which constructor of the class is invoked. In that sense, an
instance initializer is kind of a part of each constructor of the class. Therefore, an
exception thrown from an instance initializer can be thought of as an exception
thrown by every constructor of the class. Thus, if we declare an exception
thrown from an instance initializer in the throws clause of each constructor of
the class, we should be good. That is exactly what Java mandates. Here is an
example:

int i = 5;
{
}
TestClass() throws Exception{

}
TestClass(int x) throws Exception{

}
}
In the above code, since the instance initializer of TestClass throws

Exception , each of its two constructors must declare Exception in their
throws clauses. Now, can you guess whether the following code will compile or
not?

int i = 5;
{
}
Observe that TestClass does not define any constructor explicitly. Therefore,
the compiler will provide the default no-args constructor for this class. However,
the default constructor does not have any throws clause. Therefore, the exception
thrown by the instance initializer is not getting declared by that constructor.
Hence, the above code will not compile.
Throwing exceptions from constructors ☝
Just like a method, a constructor is allowed to throw any exception as long as it

declares that exception in its throws clause. However, there is one important
difference between a method and a constructor. Recall that the first line of every
constructor is always a call to a constructor of its super class or to another
constructor of the same class. Thus, if a constructor decides to throw an
exception, it has an impact on the subclass because that exception will be
propagated to the subclass constructor as well. Therefore, if a subclass
constructor invokes a superclass constructor that throws an exception, that
subclass constructor must also declare that exception in its throws clause. Here is
an example:
class Fruit{
Fruit() throws Exception{
if(Math.random()>0.5) throws Exception; //throws
an exception randomly
}
Fruit(int calories){
}

}
The Fruit class above has two constructors - one with a throws clause and one
without. Now let's see some subclasses of Fruit :
class Apple extends Fruit{ //will NOT compile.
Notice that since Apple doesn't define any constructor explicitly, the compiler
will provide the default no-args constructor to this class. Furthermore, the
compiler will insert a call to super(); in the default constructor. This default
constructor will cause the Apple class to fail compilation because it has neither
a throws clause nor a wrapping try/catch block that could catch the exception
thrown by the call to super(); , which is nothing but a call to Fruit 's no-
args constructor. To make it compile, you must provide the no-args constructor
with an appropriate throws clause yourself:

Apple() throws Exception{
}
Apple(int calories) { //no throws clause needed
super(calories);
}
}
Observe that the int constructor of Apple does not require any throws clause
because it invokes Fruit 's int constructor, which does not throw any
exceptions. You could also change Apple 's no-args constructor to invoke
Apple 's int constructor using this(100); to avoid having a throws clause:
Apple() {
this(100);
}
Apple(int calories) { //no throws clause needed
super(calories);
}
}
A subclass constructor is free to throw any new exception along with the
exceptions listed in the throws clause of a super class constructor.
Note that it is not possible for a subclass constructor to catch an exception

thrown by the superclass constructor because to do that you would have to put
the call to super(...) in a try block, in which case the call to super(...)
will not be at the first line of the constructor!
15.3.3 Invoking a method that throws an exception ☝

The rules for invoking a method that throws an exception are similar to the ones
for creating a method that throws an exception. As far as the compiler is
concerned, there is no difference between using a throw statement to throw an
exception and invoking a method that throws an exception to throw an
exception. In both the cases, the compiler forces you to either declare the
exception in your throws clause or handle the exception using a try/catch
block. Of course, as discussed before, the compiler is only concerned with
checked exceptions . The following example illustrates this point:
public static void foo() throws Exception{

if(true) throw new Exception();
else return;
}
public static void main(String[] args) throws Except

ion {
foo();
}
In the above code, the method foo() throws an Exception using the throw
statement. Since Exception is a checked exception, the compiler forces it to
be declared in the throws clause. On the other hand, instead of throwing an
exception explicitly using the throw statement, main invokes foo . But the
compiler knows that invoking foo may result in an Exception being thrown
(because it is mentioned in the throws clause of foo ) and so the compiler
forces main to declare Exception in its throws clause as well. The other
option for main is, of course, to handle the exception itself:
public static void main(String[] args) { //no throws

clause necessary
try{
foo();
}
}
It is ok to use a catch block with a broader exception class to catch a narrower

exception. For example, if you are calling a method that declares Exception
in its throws clause, you can use catch(Throwable e) to handle that
exception:
public static void foo() throws Exception{

//some code that throws an exception
}

try{
foo();
}catch(Throwable e)
{
}
}
The reverse, i.e., using a catch block with a narrower exception to catch a
broader exception is not acceptable. Thus, the following will not compile:

try{
foo();
}catch(java.io.IOException e)
{ //catch clause is too narrow to catch Exception
}
}
You can think of a catch clause as a basket of a particular size. You can use a
bigger basket to catch a smaller exception, but you cannot use a smaller basket
to catch a bigger exception. Note that I am using the word "bigger" here in the
sense that a bigger exception has a bigger tree of subclasses than a smaller
exception. For example, Throwable has a lot more subclasses than
Exception because the tree of Throwable includes the tree of Exception
as well as the tree of Error . Similarly, Exception is bigger than
IOException because the tree of Exception includes the tree of
IOException along with a lot of other exception subclasses.
If a method doesn't want to catch the exception then it must declare that
exception (or its superclass) in its throws clause. This is no different from the
rule that you have seen before while creating a method that throws an exception.
For example, assuming that the method foo throws Exception , here are
valid throws clauses for a method that invokes foo :
//declare the same exception class
public static void bar() throws Exception
{
foo();
}
//declare a super class of the exception class throw

n by foo
public static void bar() throws Throwable
{
foo();
}
Of course, bar is not limited to throwing just the exceptions declared in the
throws clause of foo . It can add its own exceptions to the throws clause
irrespective of whether the code inside the method throws them or not.
To catch or to throw ☝
The decision to catch an exception or to let it propagate to the caller depends on

whether you can resolve the problem that resulted in the exception being thrown
or not. Consider the following code for a method that computes simple interest:
public static double computeInterest(double p, double
r, int t) throws Exception{
if(t<0) throw new Exception("t must be > 0");
else return p*r*t;
}
and the following code that uses the above method:

double interest = 0.0;
try{
computeInterest(100, 0.1, -1);
}
}
Upon execution, the main method prints interest as 0.0 even though the
computeInterest method did not really compute interest at all. It threw an
exception because t was less than 0. However, as a user of the program, you
won't know that there was actually a failure during the computation of interest.
While the computeInterest method did its job of telling the main method
of a problem in computation by throwing an exception, the main method
sweeped this problem silently under the rug by using an empty catch block.
This is called "swallowing the exception" and is a bad practice.
The purpose of a catch block is to resolve the problem and not to cover up the
problem. By covering up the problem, the program keeps running but starts
producing illogical results. Ideally, main should not have caught the exception
but declared the exception in its throws clause because it is in no position to
resolve the problem. It would have been appropriate for a program with GUI to
catch the exception and ask the user to input valid arguments.
Sometimes, it becomes necessary to catch an exception even though no

resolution is possible at that point. The right approach in such a case is to log the
exception to the console so that the program can be easily debugged later by
inspecting the logs.
double interest = 0.0;
try{
computeInterest(100, 0.1, -1);
//or System.out.println(e);
}
}
15.3.4 Using multiple catch blocks ☝
The Java standard library has a huge number of exception classes. On top of it,
Java allows you to create your own exception classes as well. The purpose of
having so many exception classes instead of having a few generic exception
classes is to capture as many details of an exceptional situation as possible.
These details help the caller of a method resolve the problem. For example,
instead of throwing a general IOException when a file is not found, you
should throw a more specific FileNotFoundException because it
describes the problem more accurately to the caller. If a method expects to
encounter different issues, it should throw a different exception for each issue.
The caller can use multiple catch blocks to catch these exceptions and take
action based on which exception is thrown by a method. Here is how -
import java.io.*;
//assume that CharConversionException, FileNotFound

Exception, and EOFException extend IOException

void foo(int x) throws IOException{
if(x == 0) throw new CharConversionException();
else if(x == 1) throw new FileNotFoundException(
);
else throw new EOFException();
}

try{
tc.foo(2);
}catch(EOFException eofe){
System.out.println("End of file reached");
}
catch(CharConversionException cce){
System.out.println("Some problem with file sy
stem");
}
catch(FileNotFoundException fnfe){
System.out.println("No such file found");
}
catch(IOException ioe){
System.out.println("Unknown I/O Exception");
}
}
}
In the above code, if the call to foo throws an exception, the control will check
each catch clause one by one to see if it is able to catch the exception. The
control will enter the first catch block that is able to handle the exception. So for
example, if the call to foo throws FileNotFoundException , the first
catch block that is able to handle this exception is the third one, i.e.,
the catch(FileNotFoundException fnfe) block. The last catch
clause, i.e., catch(IOException ioe) is also able to handle a
FileNotFoundException but the control will not bother with it because it
has already found a matching catch block before it reaches
the catch(IOException ) clause.
Unreachable catch blocks ☝
The above program makes for an interesting problem. What if you move the
catch(IOException ioe) block before the
catch(FileNotFoundException fnfe) block? Well, the code will fail
to compile. The catch(IOException ioe) clause will always satisfy a
FileNotFoundException and so the control will never get to enter
catch(FileNotFoundException fnfe) block. Therefore, the compiler
will deem the catch(FileNotFoundException fnfe) block as
unreachable. It is like putting a bigger basket above a smaller one. The smaller
basket will never be able to catch anything because of the bigger one sitting
above it!
A similar issue will occur if you change foo 's throws clause in the above
program from throws IOException to throws
CharConversionException, FileNotFoundException,
EOFException . In this case, the compiler will realize that foo cannot throw
an IOException (because it is not listed in foo 's throws clause anymore)
and all three exceptions that it can actually throw are being caught already by the
three catch blocks and therefore, the catch(IOException ioe) block will
never be executed. Ideally, the compiler should refuse to compile the code but
the JLS allows this and only requires a compiler to generate a warning.
Wait, there is more. What if you change the catch(IOException ioe)

clause in the main method to catch(Exception e) . You know that foo
doesn't throw Exception . So, what you do think the compiler will do? Well,
actually, the compiler won't have an issue with it. Remember that one of the
subclasses of Exception is RuntimeException , which is an unchecked
exception. Any method is free to throw a RuntimeException without
declaring it in its throws clause. Therefore, the compiler needs to consider the
possibility that foo may throw a RuntimeException and the
catch(Exception e) block will get executed if that happens. Thus, the
compiler has no choice but to accept the code.
Nested try statements ☝
It is possible to nest try statements inside other try , catch , or finally

blocks. The rules for such try statements are exactly the same as those for
regular try statements. But remember that a catch block associated with a
nested try statement is cannot catch an exception thrown by the outer try
block. Here is an example:
void foo() throws EOFException {

try{
if(true) throw new FileNotFoundException();
else throw new EOFException();
}catch(FileNotFoundException fnfe){
try{
throw new EOFException();
eofe.printStackTrace();
}
}
}
In the above code, the catch(EOFException eofe) block is associated

with the try statement that is nested inside the
catch(FileNotFoundException fnfe) block. Therefore, it cannot
catch an EOFException thrown from the outer try block. In fact, if the outer
try block throws EOFException , the control will not even enter
the catch(FileNotFoundException ) block because
catch(FileNotFoundException fnfe) clause does not satisfy
EOFException . Thus, the EOFException thrown from the outer try
block will remain unhandled, which means it must be declared in the throws
clause of the method.
Throwing an exception from catch or finally ☝
It is possible to rethrow the same exception (or throw a new one) from the catch
block. You may want to do this if you want to do something upon receiving an
exception before letting it propagate up the call chain. Of course, if it is a
checked exception you would have to declare it in the throws clause of the
method. Here is an example:
void bar() throws EOFException {
try{
foo();
//do something here and then rethrow the same ex
ception again.
throw eofe;
}
}
The same effect can be achieved with a finally block:
void bar() throws EOFException {

try{
foo();
}finally{
//do something here
}
}
Since there is no catch block to catch the EOFException thrown by foo , the
exception will automatically be thrown out of the bar() method to the caller of
bar() , but only after the code in the finally block is executed.
Rethrowing an exception explicitly from a catch block is helpful only if you

have multiple catch blocks and you want to do a different thing in
every catch block.
It is also possible to throw an entirely new exception from the catch and finally
blocks. Here is an example:
void bar() throws Exception {

try{
foo();
}
}
The EOFException thrown by foo will be caught by

catch(EOFException eofe) block. Since the catch block throws
Exception , bar will end up throwing Exception to its caller. Thus, bar
must list Exception in its throws clause.
If you throw an exception from the finally block, then the exception thrown from
the try block or the catch block is ignored, and the exception thrown from the
finally block is what gets thrown out of the method. The following example
illustrates this point:
void bar() throws IOException {

try{
foo();
}
finally{
throw new IOException();
}
}
Even though the catch block throws Exception , the JVM doesn't throw it to
the caller of bar . It waits until the finally block finishes execution. However,
the finally block throws a new IOException . So the JVM ignores the
Exception that it was about to throw to the caller and throws the
IOException instead. As far as the compiler is concerned, it realizes that bar
can only throw an IOException and not Exception to the caller and
therefore, it is ok with listing just IOException in bar 's throws clause.
15.4 Exercise ☝
1. Create a method named countVowels that takes an array of characters as
input and returns the number of vowels in the array. Furthermore, the
method should throw a checked exception if the array contains the letter
'x' .
2. Invoke the countVowels method from main in a loop and print its return
value for each command line argument. Observe what happens in the
following situations: there is no command line argument, there are multiple
arguments, there are multiple arguments but the first argument contains an
'x' . (Use String's toCharArray method to get an array of characters
from the string.)
3. Ensure that your main method prints the number of vowels in other
command line arguments even if one argument contains an 'x' .
4. Pass null to the countVowels method and observe the output.
5. Modify countVowels method to throw an

IllegalArgumentExcception if it is passed a null .
6. Modify countVowels method to return -1 , if the input array is null

and 0 , if the input array length is less than 10 . Do not use an if
statement.
Chapter 16 Understanding Modules
16.1 Module Basics

16.1.1 What are modules? ☝
In the Kickstarter for Beginners chapter, you saw how to package individual
classes of an application into a jar file and how to run an application using that
jar file. The jar file approach has been a great way to package multiple files into
a single file and to deliver applications as well as class libraries to users.
However, Jar files are not without problems. Packaging classes into a jar files
without a well thought out plan gives rise to unwieldy applications that are
difficult to update and/or reuse. Jar files also make it difficult to use just a part of
an application without having the whole set of jar files. If you are a Java
beginner or haven't been involved in professional level Java development much,
you may find it hard to believe that the jar file approach also led the Java
community to something called "Jar Hell". Jar Hell is a term coined to describe
hard to debug behavior of applications due to the presence of multiple versions
of a class library on the classpath and/or due to the use of different versions of a
library used by different parts of the applications. For example, what if an
application uses Spring and Hibernate and these libraries require a different
versions of the same jdbc driver or a logging library?
Although the JDK did not provide any out of the box solution to such issues,
various Java community projects such as Ant, Maven, and Gradle have
attempted to provide some way of managing these issues.
With Java 9, the Java language designers came up with a new approach called
the Java Module System, also known as Project Jigsaw, to packaging Java
classes that attempts to solve at least some of the problems mentioned above. It
envisions a Java application (or a library) to be composed not of classes or jar
files but of modules, where each module of an application clearly specifies what
other modules it depends on and what features it provides to other modules. A
module, in turn, is composed of classes. A module, therefore, is a much coarser
grained component of an application than a class. In other words, classes are
packaged in a jar file to form a module, and modules are mixed and matched at
run time to form an application. The following diagram illustrates how classes
and modules come together to make an application.
The above diagram shows three jar files named moduleA.jar, moduleB.jar, and
moduleX.jar. These module jar files are used together to create an application.
Note that an application is just a loosely coupled collection of modules and is not
packaged as a jar file itself. You will soon see that an application is created by
specifying the required modules from different teams, groups, or companies on
the command line at run time.
It is not like professional application teams were not already grouping

independent parts of an application into separate jar files. Indeed, even before
the advent of modules, Java applications were composed of jars collected from
various teams. But the process of grouping the classes and, more importantly,
documenting what a jar file requires and provides, varied from team to team and
from build tool to build tool. Java 9 takes the best of the practices prevalent in
the industry and formalizes them into the Module System.
Besides formally specifying a structural design for applications, Java 9 also

includes new tools and enhances existing tools to help developers package their
code as modules.
Package vs Module ☝
You may be wondering at this point where packages fit into this picture. After
all, we have been using packages to organize our Java code since ages. Well,
packages still play the same role. They are still used to organize our code but at a
finer level.
Conceptually, a package creates a namespace where all of the code required to

implement a particular functionality resides. It is a means to reduce the size of
the code contained in a class file by separating the logic into multiple closely
related classes. This helps in reading, understanding, and navigating the code.
But this is from the perspective of the developer of the functionality and not
from the perspective of the users of that functionality. The users of that
functionality are not interested in how the code is organized. They are interested
in knowing what it takes to use that functionality. Users simply want to use the
functionality as a black box. Unfortunately, packages do not provide this
information. Developers have been relying on informal methods such as naming
jar files according to their own conventions and adding release notes to deliver
this information.
This is where modules come in. A module provides a way to deliver

functionality in a single artifact that includes not just the code but also the
information on the services that the code provides and any dependencies that it
may have.
Thus, if you are a developer of some functionality, you would still organize your
code into packages. But you would then also package all of your classes (and
any other resources such as image files, xml files, or property files) that are used
to implement this functionality into a single deliverable with a description of
your functionality and its dependencies in a prescribed format in the form of a
module. In other words, modules are not an alternative to packages but are on
top of packages.
The Java module system is a fairly large topic with dedicated books written to
cover it. However, the OCP Java SE 11 Part 1 exam focuses only on the basics of
the Java module system.
There has been a fair bit of criticism of the Java

Module system in the industry. Even though Java
Modules is an advanced topic, Oracle has included it in
the Part 1 Exam of the OCP certification, which tells
me that they are quite serious about pushing it. For
this reason, I will focus mostly on how to work with it
from the exam perspective instead of getting into the
analysis of its merits and demerits. However, if you are
going to face technical interviews, you should go
through online discussions and articles that talk about
the problems associated with the Java Module system.
16.1.2 Declaring a module ☝
Since a module is a logical unit of functionality from the users' perspective, it is

best to first identify the functionality that we want to deliver as a module. For
example, let us say we are developing an application for finance and one of its
functions is to compute simple interest. We want to deliver this simple interest
calculator as a module.
Naming a module ☝
A module name follows the same rules and conventions that you saw for naming
packages and classes earlier. A valid package name would therefore, be a valid
module name. In short, the name must be a valid Java identifier (so, special
characters such as dash and slash are out but underscore and dot are in) and it
should follow the reverse domain name pattern to avoid name clash. Ideally, the
name should be descriptive enough to tell the user the purpose and/or
functionality of the module instead of being too generic such as tools or
utils . So, com.abc.finance.calculators.simpleinterest
would be a good name for our module. However, for the purpose of this chapter,
I am going to name the module as just simpleinterest to reduce clutter and
to make it easy to try out the code that I am going to present here. Let us also
decide that the FQCN of our class will be
simpleinterest.SimpleInterestCalculator instead of
com.abc.finance.calculators.simpleinterest.SimpleInterestCalcul
for the same reason. Note that there is no relation between the module name and
the package name(s) of the class(es) contained in that module. However, since
both follow the reverse domain name convention, they may very well be the
same.
The module descriptor ☝
Remember I talked about the lack of any formal way to describe a package? This
is where a module differs from a package. Module descriptor is the formal way
to describe a module. Every module must have a module descriptor that specifies
the name of the module, what it provides, and what it requires. A module
descriptor is nothing but a file by the name module-info.java . Yes, the
extension of the file is java . It is a Java file that will be compiled to generate a
class file by the name module-info.class . When we talk of a module, it is
this descriptor that we generally refer to. Indeed, the module itself is a black box
and the module descriptor is all the user should worry about. In our case, the
contents of the module descriptor will be as follows:
//in file module-info.java
module simpleinterest{
}
There isn't much in this module descriptor. All it says is that this is a module
named simpleinterest . But it is a valid module descriptor nonetheless and
it illustrates the basic syntax for defining a module. You have the module
keyword and the module name, followed by the opening and closing braces. We
will be adding more information in this descriptor as we go along. At this time,
however, it is a good start!
The module-info.java file cannot be empty.
16.1.3 Directory structure of a module ☝

The directory structures for the sources and for the compiled classes of a module
are almost exactly the same as that of a package. The only difference is that
while a package may reside in any folder, a module resides in a folder by the
same name as the module. This rule applies to the source code as well as the
class files. If the module directory contains the source code, it is called "source
module definition" and if the module directory contains compiled classes, it is
called "module definition".
Thus, in our example, the source code and compiled classes of our module must
reside in folders named simpleinterest . The following image illustrates
the directory structure that I will use to develop and execute the code for our
module:
My base directory is c:\ocp11\moduletest . This is the directory where I

am going to develop and run all the code presented in this chapter. If you are
using *nix, you might want to create a moduletest directory under
/home/<username> . (Remember to flip \ to / in directory paths if you are
using *nix.)
Under the base directory, I have a src directory, where I will keep the source
code for all of my modules. Since this particular module is named
simpleinterest , I have a directory by the same name under src . The
directory tree rooted at simpleinterest is therefore, the source module
definition of the simpleinterest module. Furthermore, since our
SimpleInterestCalculator class belongs to the simpleinterest
package, I have kept SimpleInterestCalculator.java under
simpleinterest\simpleinterest .
Note that it is not necessary to follow the package structure for organizing Java
source files. You may keep all your Java source file(s) directly under the module
root folder as well. However, organizing the source code as per their packages
has a couple of advantages. Besides making the source code easy to navigate, it
is understood well by the Java compiler. The compiler can automatically locate
and compile the code present in a source module definition.
The contents of SimpleInterestCalculator.java are as follows:
package simpleinterest;
public class SimpleInterestCalculator{
public double calculate(double principle, double ra
te, double time){
return principle*rate*time;
}
System.out.println(new SimpleInterestCalculator()
.calculate(100, .05, 2));
}
}
16.2 Describe how a modular project is compiled and

run
16.2.1 Compiling a module ☝
Now that our directory structure and the code base for the module is ready, let us
compile it. Open a command/shell prompt, cd to
the c:\ocp11\moduletest directory and run the following command:
c:\ocp11\moduletest>javac -d out --module-source-path

src --module simpleinterest
The above command uses three command line switches: -d , --module-

source-path , and --module .
You have used the -d switch before to direct the output of the Java compiler to a
particular directory. It works the same way here. So, -d out will cause the
compiler to produce its output in the out directory. The compiler will create the
out directory if it does not already exist.
The --module-source-path switch tells the compiler the location of the

source module definition. Since we have kept our source code in a directory
under the src directory, that is what we have specified here as well. Observe
the double dash in this switch.
Since the name of the module is simpleinterest , the compiler will create a
simpleinterest directory under out . The out\simpleinterst
directory is the simpleinterest module's "root directory" and the tree
rooted under simpleinterest is the "module definition" of the
simpleinterest module.
The compiler will save the class files of this module inside
out\simpleinterest as per the package driven directory structure of the
classes of this module. For example, since SimpleInterestCalculator
class is in simpleinterest package, the compiler will put
SimpleInterestCalculator.class under
out\simpleinterest\simpleinterest directory.
Note that I am using relative paths instead of absolute paths (i.e. out instead of
c:\ocp11\moduletest\out ). Relative paths are relative to the current
directory, and since our current directory is c:\ocp11\moduletest , out
resolves to c:\ocp11\moduletest\out . You may use absolute paths and
relative paths interchangeably in these commands but it is better to use relative
paths because they will work even if your base directory something different
from c:\ocp11\moduletest .
The --module switch specifies the name of the module that we want to
compile. The compiler tries to locate this module in the paths specified in
module-source-path . In this case, we want to compile the
simpleinterest module. Again, observe the double dash.
Based on the above options, the compiler will look for module-info.java
under src\simpleinterest directory and will compile all the Java files
under all the directories that are present inside src\simpleinterest
directory.
Upon successful execution of the above command, you should see the directory
structure as shown below:
Observe the directory structure generated by the compiler
under out\simpleinterest . It is the same as what you get after compiling
non-modular Java source files.
I suggest that you play around with the above command by changing the
particulars of this set up. For example, rename the module source directory to
something else or move the Java source file to another location within the
module source directory and observe the error message generated by the
compiler.
Compiling individual module source files ☝
It is possible to compile module source files without using the --module

option by listing the files (either individually or using a wildcard) that you want
to compile. For example, the following old style command will produce the
same result as produced by the command used above:
c:\ocp11\moduletest>javac -d out\simpleinterest
src\simpleinterest\module-info.java
src\simpleinteres\simpleinterest
SimpleInterestCalculator.java
Observe that you have to specify the module root directory component
simpleinterest in the output path explicitly. Without this, the compiler will
save the output under out instead of out\simpleinterest . Listing
individual files for compilation is not a preferred way to compile a module and is
seldom used. You will not be tested on this in the exam either. For the purpose of
the exam, you are expected to know how to compile a module using the --
module switch only.
16.2.2 Running a module ☝
Running a module means executing the main method of a class contained in that
module. Thus, the class that we want to execute must have the standard main
method. The following command shows how to run our simpleinterest
module that we just compiled.
c:\ocp11\moduletest>java --module-path out --module

simpleinterest/simpleinterest.SimpleInterestCalculator
This command uses two switches: --module-path and --module .
The module-path switch is used to tell Java the location of the module
definition. This is the directory where the module's root directory is located or, if
the module is packaged as a jar file, module-path is the directory where that jar
is located. In our case, we have specified out in module-path because our
module's root directory, i.e., simpleinterest , is located in out .
The module switch is used to tell Java the name of the module that we want to
execute. Observe the value specified for --module . It is in the format
<module name>/<main class name> because, as of now, our module
is present in an exploded format and does not contain any information about the
main class. So, we need to tell the JVM about the class that we want to run as
well. Once we package our module into a jar, we won't need to specify the main
class on the command line.
Packaging a module ☝
The process of packaging classes of a module into a jar file is the same as the
one we saw for packaging classes into a regular jar. The only thing special about
a module jar is that the name of the jar is, by convention, the same as the name
of the module (with .jar extension, of course). Inside the jar file, class files
must still exist in their package driven directory structure just like before.
Observe that a module also has a module-info.class in its root folder.
This class must be in the root folder of the module's jar as well. Since the
structure of the files in a modular jar file mimics the directory structured of the
compiled module, a modular jar is also considered as the "module definition" of
a module.
Here is the command that will package our module into a jar file:
c:\ocp11\moduletest>jar --create --file

simpleinterest.jar --main-class
simpleinterest.SimpleInterestCalculator -C
out\simpleinterest .
This should create simpleinterest.jar in the moduletest directory.

Once
The --create switch tells the jar tool that we want to create something and
the --file and the subsequent value specify the jar file that we want to create.
The --main-class switch makes the jar tool add a Main-Class entry in
the resulting jar file's manifest.
The -C switch needs some explanation. We want the jar file to mimic the
directory structure of the module, which means the structure inside the jar file
should be the same as the structure inside the out\simpleinterest
directory. But we are executing the command from the moduletest directory.
There are two extra levels under moduletest (out and simpleinterest )
that we don't want reflected in the jar file. The -C option is to make the jar tool
change its working directory before adding files to the jar. We want it to step
down to out\simpleinterest directory and then include the files from
there.
The dot at the end of the command means that we want to include everything in
its current directory (which is now out\simpleinterest due to the use of
the -C switch) to be included in the jar file.
Running this module is now a piece of cake:
c:\ocp11\moduletest>java --module-path .
--module com.abc.finance.calculators.simple
interest
Observe that, since we added the Main-Class attribute in module jar's manifest,
we are able to "run" the module.
You need not worry too much about the various switches used in
the jar command. You won't be tested on how to create a jar file in the exam.
However, executing a module packaged in a jar fiIe is on the exam. So, it would
be good if you understand how a modular jar file is created.
Understanding module-path ☝
Another interesting thing in the above command is the value that we have
specified for module-path . We have specified only a dot for the module-
path because our module jar is in the current directory. A module-path contains
all the locations where you want the JVM to search for module definitions.
While trying to load a module, the JVM looks for that module in all the locations
specified in the module-path . It expects to find exactly one of the two things
- a directory with the same name as the name of the module or a jar file
containing module-info.class for that module. The name of the jar file is
not important because the JVM takes the name of the module from the
module-info.class contained in the jar file. The same module should not
be present more than once on the module-path, otherwise the JVM will
complain.
In the command used above, while searching for the module

named simpleinterest , the JVM will look for a jar file that contains
module-info.class with the module name as simpleinterest at the
root or a directory by the name of simpleinterest containing appropriate
module-info.class .
Compare this with the -classpath switch, which requires all jar files that you
want on the class path to be listed explicitly instead of requiring just the
directory that contains the jar file(s).
16.3 Declare modules and enable access between

modules
16.3.1 Enabling access between modules ☝
Let us improve the design of our simpleinterest module a little bit. It is a

good design practice to define functionality in the form of an interface and let
the actual implementation implement that interface. This ensures that the user of
the functionality is able to replace one implementation with another without
much impact on their code. Java modules allow us to take this technique even
further by defining functionality separately in a module of its own. Separating
the interface and the implementation into separate modules allows us to build an
application by mixing and matching modules without the need to bundle classes
that are not required for the application.
In our case, we will create a calculators.InterestCalculator

interface in calculators module. We will then have our
SimpleInterestCalculator class implement this interface. Here is the
interface definition:
//in file calculators\calculators\InterestCalculator.j

ava
package calculators;
public interface InterestCalculator{
te, double time);
}
and the following is the module descriptor for the new module:
//in file calculators\module-info.java
module calculators{
exports calculators;
}
The only thing new in this module definition is the exports clause. A module
is an insular unit of packages. Members of a module are normally not accessible
to code belonging to other modules. To make a package accessible to other
modules, the package must be "exported" explicitly in the module descriptor.
Making something accessible only through an explicit exports clause ensures
that code from other modules does not access code that is internal to this module
inadvertently. Note that I am talking about packages instead of classes because
only packages can be exported and not individual classes. When you export a
package, all of the public types contained in that package are made accessible to
other modules. In module parlance, accessing a module is called "reading" the
module. So, the exports clause in the above code enables classes in other
modules to "read" the InterestCalculator interface. This is also known
as module readability .
The directory structure of this module is as per the source module definition
format explained earlier.
The c:\ocp11\moduletest\src\calculators directory is the source
directory for this module. It contains module-info.java and
calculators\InterestCalculator.java files. Let us run the
following command to compile this module:

src --module calculators
This should create appropriate output in the out directory.
Let us now move our attention to the simpleinterest module. Since we

want SimpleInterestCalculator to implement
InterestCalculator interface, we need to first make our intention clear by
stating that our simpleinterest module "requires" access to the
calculators module by modifying simpleinterest 's module
descriptor as follows:
module simpleinterest{
requires calculators;
}
The requires clause is the counterpart of the exports clause. The purpose
of having a requires clause is to make the dependencies of a module
explicitly clear to the users. The users of a module need not go through the
import statements of each class of a module to know which other modules does
this module depend on. All they need is to check the module-info.
We can now modify SimpleInterestCalculator.java as follows:

package simpleinterest;
import calculators.InterestCalculator;
public class SimpleInterestCalculator implements Inter
estCalculator{
te, double time){
return principle*rate*time;
}
InterestCalculator ic = new SimpleInterestCalcu
lator();
System.out.println(ic.calculate(100, .05, 2));
}
}
The following command can be used to compile this module:

src --module simpleinterest
After the compilation of this module, our directory structure looks like this:

We can now execute the simpleinterest module as we did before using the
following command:
c:\ocp11\moduletest>java --module-path out --module
simpleinterest/simpleinterest.SimpleInterestCalculator
16.3.2 Qualified exports ☝
The exports clause opens up a package to access from all other modules.
Once a package is exported, you cannot stop any other module from accessing it.
It is similar to making a member of a class public , and just like the public
modifier, the exports clause makes a module less encapsulated.
The Java module systems allows you to fine tune access to a module only to
specific modules using a variation of the exports clause. It is called the exports-
to clause or the qualified exports clause. The following is how it is used:
module <modulename>{
exports <packagename> to <modulename(s)>;
}
For example, in the following module-info, the com.abc.datatransfer

package is being made accessible only to db and service modules:
module datatransfer{
exports com.abc.datatransfer to db, service;
}
The modules for which a package is made readable are called "friendly"
modules. Java standard library uses this feature heavily for its internal modules.
For example, the java.base module of the JDK has several packages which
are supposed to be used only by a selected few internal modules. It uses
qualified exports to achieve the same. Here is a partial code snippet from
java.base module's module-info:
module java.base {
...
exports jdk.internal.logger to
java.logging;
exports jdk.internal.perf to
java.desktop,
java.management,
jdk.jvmstat;
...
}
There is no particular order in which unqualified and qualified exports have to

appear in a module info. Conventionally, however, qualified exports are listed
after all the unqualified exports.
Using qualified exports provides protection from creating inadvertent

dependencies on packages that are not meant to used by everyone. Therefore, it
is a good idea to use qualified exports as much as possible. Unqualified exports
should be used only when you are absolutely sure that it is ok for the package to
used by any other module.
16.3.3 Transitive dependencies ☝
A module A may read another module B by using a requires clause in its

module-info. But what if the module B reads yet another module C ? Does that
mean module A reads the module C as well? Normally, no. Module dependencies
are not transitive and if module A specifies that it requires only module B , it
cannot read module C just because B reads C .
Now, let's see how this pans out in the following situation. What if a class in
module A needs to invoke a method of a class in module B , whose return type is
defined in module C ? Something like this:
module ui{
requires hr;
}
module hr{
requires valueobjects;
exports hrservice;
}
//code appearing in a class in ui module
HRService hrService = new HRService(); //HRService is

defined in hr module
Employee e = hrService.getEmployee(employeeId); //Empl

oyee is defined in valueObjects module
Since the ui module does not have a requires valuesobjects; clause,
the ui module cannot access the Employee class from the valuesobjects
module. Thus, the above code will fail to compile.
The obvious solution is to add a requires valuesobjects; clause in the

ui module. But this solution has a serious problem. What if the hr module has
several requires clauses? How would the ui module know which of the hr
module's requires clauses should be added to its module-info? Only multiple
compilation failures can make this information known to the ui module!
This is where the requires transitive clause comes into picture. Since the hr
module knows that whichever module requires hrservice package, will
require classes in valueobjects module as well, it documents this
information using the requires transitive clause, like this:
module hr{
requires transitive valueobjects;
exports hrservice;
}
This basically says that if you require the hr module, then you also require the
valueobjects module. But more than merely documenting this fact, the
requires transitive clause also eliminates the need for adding an extra
requires valuesobjects; clause in the ui module. The requires
hr; clause in the ui module automatically makes all the modules transitively
required by hr module, readable to the ui module. This is called "implied
readability".
It is not clear whether Qualified Exports and requires

transitive are on the OCP Java 11 Part 1 exam or not.
Some candidates have reported seeing options that use
these clauses, so, it is a good idea to understand what
they do.
16.4 Advanced module compilation and execution ☝
Seeing all the new module related command line switches feels overwhelming at
first. But once you understand the underlying theme behind these options, you
will realize that they are not much different from the -classpath (and -
sourcepath ) options that you have used so far. The module-path and
module-source-path switches are to a modular project what classpath
and sourcepath switches are to a regular project. Earlier, you used to
specify just the FQCN of the main class while launching an application, now,
you specify the module name using the --module option. That is basically it.
I have observed that many students get confused due to the fact that
classpath and module-path are not mutually exclusive. It is possible (and
sometimes necessary) to use both the switches simultaneously in a command
while compiling as well as executing a module. They are used together in
projects that make use of modular as well as non-modular libraries. However, a
discussion on this usage is not in scope for the OCP Java 11 Part 1 exam.
Another cause of confusion is the presence of multiple switches that mean and
do the same things. For example, --module-path is the same as -p and --
module is the same as -m .
In this section, I will summarize all the options that you need to use while
compiling and executing a module.
16.4.1 Compiling multiple modules at once ☝
In the previous example, we compiled the calculators and

simpleinterest modules separately one after another. We could have easily
compiled both of them together at the same time by listing both of them in the -
-module switch:

src --module calculators,simpleinterest
Observe that the module names are separated using a comma. In fact, Java will
compile the calculators module even without us specifying it explicitly in
the --module switch because javac will notice that simpleinterest
requires calculators and so, it will compile calculators first. However,
this is possible only when the source code of the module is organized correctly
in the package driven directory structure. If the source files are organized
differently, javac will not be able to locate them on its own.
Using a module without source code ☝
While developing a module, you may be required to use third party modules
whose source code is not available. For example, what if the calculators
module was developed by another team and you just had a
calculators.jar for it?
Well, we can use the --module-path switch for telling javac the location
of a module just like we used it while executing a module. Let's say we put
calculators.jar in c:\ocp11\moduletest\modulejars directory.
We can now use the following command to compile the simpleinterest
module:
c:\ocp11\moduletest>javac --module-path modulejars -d

out --module-source-path src --module simpleinterest
Observe that we didn't specify the actual jar name in the module path (although
we could have done that as well). We specified just the directory that contains it.
We can specify as many directories as we need separated by the path separator (;
on windows and : on *nix). The difference between specifying a directory and
specifying a jar is that when you specify a directory, all the jars as well as
exploded module definitions present in that directory become available to the
compiler (or to the jvm, in the case of execution), while in the other case, only
the jar that you specify on the module path becomes available. It is alright to
specify a directory on the module-path while learning but it is always better to
specify the jars while developing a professional application. The command to
run simpleinterest with explicitly listed jar files in the module-path looks
like this:
c:\ocp11\moduletest>java --module-path
modulejars\simpleinterest.jar;modulejars\calculato
rs.jar --module simpleinterest
16.4.2 Module jar vs regular jar ☝
As you saw before, a modular jar is no different from a regular non-modular jar
from in terms of its structure. In fact, it is possible to use a modular jar just like a
non-modular jar. If you are wondering whether you can run the main class of a
modular jar file the old way, then yes, you can. For example, assuming that you
have packaged the two modules into two jars as explained in the previous
section and put them in c:\ocp11\moduletest directory, you can execute
our SimpleInterestCalculator class with the following command:
c:\ocp11\moduletest>java -classpath
calculators.jar;simpleinterest.jar
simpleinterest.SimpleInterestCalculator
However, when you run a class using the -classpath or -jar option, you
lose all the benefits of modules. Which means, the JVM will not enforce the
access rules specified in module descriptors. A class from any jar can access a
class from another jar. This option is helpful when a team has not moved to the
modular structure for their project but wants to use a modular jar produced by
another team.
Similarly, if you are running a modular project, it is possible to use a non-

modular jar as a modular jar by simply putting it on the module-path instead
of the classpath . Once you put a non-modular jar on the module-path, it is
deemed to be an "automatic module". A discussion on automatic modules is
beyond the scope of this exam and so, I will not talk about it any further in this
book.
16.4.3 Summary of command line switches used for compilation
☝
For compiling a Java class and/or module, you need to remember the following
five command line options:
1. --module-source-path : This switch is used to specify the location of

the module source files. It should point to the directory that contains source
module definition. In other words it should point to the parent directory of the
directory where module-info.java of the module is stored. For example, if
your module name is moduleA , then the module-info.java for this
module would be in moduleA directory and if moduleA directory exists in
src directory, then src should be specified in the --module-source-
path option, i.e., --module-source-path src
If moduleA depends on another module named moduleB , and if moduleB

directory exists in src2 directory, you can add this directory in --module-
source-path as well, i.e., --module-source-path src;src2 .
2. -d : This switch is required when you compile a module. It is used to

specify the output directory. This is the directory where javac will generate the
module and package driven directory structure and the class files for the sources.
For example, if you specify out as the output directory, javac will create a
directory under out with the same name as the name of the module and will
create class files with appropriate package driven directory structure under that
directory.
3. --module or -m : This switch is used when you want to compile all the
source files of a particular module. This option is helpful when you want to
compile all the files at once without listing any of the source files of a module
individually in the command.
For example, if you have two Java files in moduleA , stored under
moduleA\a\A1.java and moduleA\a\A2.java , you can compile both
of them at the same time using the command: javac --module-source-
path src -d out --module moduleA
Javac will find out all the Java source files under moduleA and compile all of
them. It will create the class files under the output directory specified in -d
option, i.e., out . Thus, the out directory will now have two class files -
moduleA/a/A1.class and moduleA/a/A2.class .
4. --module-path or -p : This option specifies the location(s) of any other

module upon which the module to be compiled depends. You can specify the
module directories or jar files containing the module classes here. For example,
if you want to compile moduleA and if it depends on some other module named
abc.util , then you can add this to your javac command like this:
javac --module-source-path src --module-path
thirdpartymodules/abc.util.jar -d out --module moduleA
5. -classpath : This option is used for compilation of non-modular code. If

you are compiling regular non-modular code but that code depends on some
classes, then you can put those classes/jars on the classpath using -classpath
option.
Note: This option is not helpful for compilation of modular code because classes
in modular code cannot read classes present on classpath. Modular code can only
see other modular code. That is why, non-modular classes have to be converted
into "automatic modules" and specified on --module-path as explained in
the note below.
The following is not required for the part 1 but is good

to know anyway because it is on the part 2 exam.
If your module depends on a non-modular third party

jar, you need to do two things -
1. Put that third party jar in --module-path .

Putting a non-modular jar in --module-path causes
that jar to be loaded as an "automatic module". The
name of this module is assumed to be the same as
the name of the jar minus any version numbers. For
example, if you put mysql-driver-6.0.jar in --
module-path , it will be loaded as an automatic
module with name mysql.driver . Name derivation
is explained in detail in
java.lang.module.ModuleFinder JavaDoc but for the
exam, just remember that hyphens are converted
into dots and the version number and extension part
is removed.
It is also possible for a non-modular jar to specify
its module name using Automatic-Module-Name:
<module name> entry to the jar's MANIFEST.MF .
2. Add a requires <module-name>; clause in
module-info of your module.
16.4.4 Summary of command line switches used for execution ☝
You need to know about the following three command line switches for running
a class that is contained in a module:
1. --module-path or -p : This switch tells the location of the module

definition. If the module is packaged in a jar file, then you can either specify the
path to the jar, or specify the path to the directory where the jar file stored
(relative to the current directory or an absolute path). For example, --module-
path c:\javatest\modulejars
You can also specify the location where the compiled module exists in the
exploded form. For example, if your module is named abc.math.utils and
this module is stored in c:\javatest\output , then you can use: --
module-path c:/javatest/output . Remember that
c:\javatest\output directory must contain abc.math.utils
directory and the module files (including module-info.class ) must be
present in their appropriate directory structure under abc.math.utils
directory.
You can specify as many module locations separated by path separator (; on

windows and : on *nix) as required.
2. --module or -m : This switch is used to specify the module that you want to
run. For example, if you want to run abc.utils.Main class of
abc.math.utils module, you should write --module
abc.math.utils/abc.utils.Main
If a module jar's manifest contains the Main-Class property, you can omit the
main class from the command. For example, you can write, --module
abc.math.utils instead of --module
abc.math.utils/abc.utils.Main .
3. -classpath or --class-path or -cp : This switch is used while

executing non-modular code. It is also used while executing a modular code to
specify non-modular jars that are required by the project.
You are not required to know the following information

for the part 1 exam. It is good to know anyway
because it is required for the part 2 exam.
It is possible to treat modular code as non-modular by

ignoring module options altogether. For example, if you
want to run the same class using the older classpath
option, you can do it like this:
java -classpath mathutils.jar abc.utils.Main
Remember that modular code cannot access code

present on the -classpath but "automatic modules" are
an exception to this rule. When a non-modular jar is
put on --module-path , it becomes an "automatic
module" but it can still access all the modular as well
as non-modular code. In other words, a class from an
automatic module can access classes present on --
module-path as well as on -classpath without having
any "requires" clause (remember that there is no
module-info in automatic modules).
Thus, if your modular jar A depends on a non-modular
jar B, you have to put that non-modular jar B on --
module-path . You must also add appropriate requires
clause in your module A's module-info otherwise
compilation of your module will not succeed.
Furthermore, if the non-modular jar B depends on
another non-modular jar C, then the non-modular jar C
may be put on the classpath or on the module-path.
16.5 Describe the modular JDK

16.5.1 Modular JDK ☝
In the previous sections of this chapter, I showed you how to create and use
modules. It was important to discuss the basic concepts of modules first because
Java 9 has modularized all the classes and packages of the JDK by following the
same concepts. Although there are still a few more topics related to modules left
that one needs to understand before they can fully appreciate the amount of
restructuring done in JDK 9, those topics are not on the Java 11 part 1
certification exam. So, I will try to explain the modular JDK without relying too
much on such topics.
If you look at the size of the Java runtime library, it is hard to believe that Java
was originally designed to run on even the smallest of devices. Now, twenty five
years later, Java has gone far beyond its goal. It has been enormously useful in
building large mission critical applications as well as small desktop applications.
One of the reasons for its success is the constant evolution of the features, tools,
and libraries that Java comes bundled with.
Another key promise of Java was WORA, "Write Once Run Anywhere". Java
came with a single set of API that you could use to write an application that
would work the same on all platforms. Java achieved this feat by developing a
run time environment that hid the differences in the underlying hardware and
provided a common set of features.
Over the years, the types and capabilities of devices differed vastly and it
become impossible to hide over the differences in hardware capabilities of so
many platforms any longer. But if you have a Java runtime for a device and if
that runtime is not able to run a Java application developed for another device,
then that would amount to abandoning the WORA promise. Java did finally
abandon WORA when it came up with multiple versions of the runtime such as
J2SE, J2EE, and J2ME. J2ME was small footprint version of the Java runtime
that could run on initial generation of mobile phones. It was vastly different from
J2SE and J2EE in terms of capabilities. However, the core Java library still
remained the same.
By Java 8, the standard Java library, with over four thousand five hundred
classes, had become a big monolithic pile of interdependent classes and
packages that was impossible to split into smaller independent pieces. Even if an
application used a very small part of the runtime library, it still had to be bundled
with the whole library. You might have heard about a jar file named rt.jar
that every Java installation comes with. This 60 MB jar file contains the
complete Java SE library. If you ever wanted to distribute your Java application
comprising just a single class, you would have to include this huge jar file along
with it.
Modular JDK is an effort to solve this problem. It divided the JDK into several
loosely coupled modules that can be combined at compile time, build time, as
well as run time, into a variety of configurations as per the needs of the
applications. For example, a large service side application can use the full JDK
but a micro-service can be packaged with only the modules that are absolutely
essential. It also allows applications that are bundled along with the runtime, to
be shipped with smaller profiles, thus, increasing the reach of Java applications
to more devices and platforms.
16.5.2 Organization of the modular JDK ☝
The packages, or API packages, as they are called in technical jargon, contained
in the Java platform are distributed into various modules. These modules can be
categorized into two broad categories - standard modules and non-standard
modules. The modules that are governed by the Java Community Process (JCP)
are called Standard modules. Their names start with java . For example,
java.base , java.sql , and java.logging are standard modules. All
other modules that are specific to a JDK are called non-standard modules. Their
names start with jdk . For example, jdk.jdeps , jdk.rmic ,
and jdk.javadoc are non-standard modules.
Similarly, standard packages have names starting with java or javax such as
java.lang , java.sql , javax.crypto and javax.net , and the JDK
specific non-standard packages, generally, have names starting with jdk such as
jdk.internal.perf , jdk.internal.logger , and
jdk.internal.util . Depending on the provider of the JDK, it may have
packages reflecting the company name as well such as sun.net or
sun.utils.resource .
It is important to know that a standard module may contain non-standard

packages but a non-standard module does not contain any standard package.
The standard and non-standard modules are then combined to make a variety of
Java configurations such as the full Java SE Platform , the full JRE , and the
full JDK .
For the purpose of the exam, we need to focus on the Java SE platform.
The Java SE Platform ☝
The Java Platform, Standard Edition (Java SE) is the core Java platform for
general-purpose computing. It is composed only of the standard modules (i.e.
whose names start with java ) and is required to be supported by all Java
implementations. You do not need to remember all the modules available in this
platform but some of the most commonly used modules in this platform are
java.base , java.desktop , java.logging , java.sql ,
java.prefs , java.desktop , and java.net.http . All of the
programs that you have written or will write for preparing for the Java
certification exam will require just this platform.
Note that this does not imply that the Java SE platform includes all of the
standard modules. For example, java.cardio is a standard module (it is
governed by the JCP) but is not a part of the Java SE Platform. Neither does it
imply that this platform itself will not make use of classes from non-standard
packages or modules. As I mentioned before, a standard module may contain
non-standard packages. All it means is that none of the modules included in the
Java SE Platform "export" any non-standard package. In other words, if your
module "requires" only the modules contained in the Java SE platform, it will
depend only on the standard Java packages and will be portable to all
Implementations of the Java SE Platform.
The java.base Module ☝
This module defines the foundational APIs of the Java SE platform. By

foundational, we mean that it lies at the core of all Java platforms. This module
doesn't require any other module but every other module depends on this
module. Again, you need not remember the complete list of packages contained
in this module but here are a few important packages that you should be aware
of - java.lang , java.lang.annotation , java.io , java.nio ,
java.net , java.util , java.util.concurrent , java.security
, and java.time .
All of the Java API classes and packages that we have used in this book belong
to this module. Every class that we have written in this book depends on a class
from this module. Of course, java.lang.Object belongs to this module,
after all! But did you notice that we never wrote requires java.base; in
the module declarations of the modules that we developed in the previous
section? The reason is that since everything in Java requires packages exported
from this module, the Java compiler does not require it to be specified explicitly.
In other words, much like the java.lang package is not required to be
imported in a class explicitly, the java.base module is not required to be
specified in a requires clause of a module explicitly. It is always assumed to
be "required".
Note that you may see the use of phrases such as "core APIs" or
"core packages" in problem statements or options of the questions in
the exam. You should mark such options as incorrect. On the other
hand, the phrase "foundational APIs" is used by the official API
description and is, therefore, correct. So, for the purpose of the
exam, java.base defines all the "foundational APIs" and not all the
"core APIs" of the Java platform.
There is a "core Java platform" though. It is defined by Java
Platform, Standard Edition (Java SE) APIs, which are defined in
multiple standard modules (and not just in the java.base module).
16.5.3 Benefits of the modular JDK ☝
The following is a list of benefits of a modular JDK as described by Java

Enhancement Proposal (JEP) 200. I am quoting important points from this
document because the exam has questions containing statements from it. If you
have time, you should go through this document, otherwise, the following list
should suffice.
1. It makes the Java platform more easily scalable down to small devices.
Since Java classes and packages are distributed into modules, it is possible
to create a configuration with a small footprint by packaging just the
modules that are required. For example, if an application does not require
any GUI, there is no need to package the java.desktop module with it.
2. It improves security and maintainability. Before modules, all members of
all classes, irrespective of their access modifiers, were accessible through
reflection API. Now, by encapsulating packages into modules, the JVM is
able to restrict reflective access only to packages that explicitly allow it.
The use of explicit exports and requires clauses prevent inadvertent
dependence on classes that are not meant to be used outside a module.
3. It improves application performance by not requiring unnecessary classes to
be packaged and/or loaded.
4. It is backward compatible. It is not required for an application to be
structured as modular to be run on the modular JDK. A non-modular
application will still run just like before on the modular JDK using the -
classpath option.
5. Customization is possible. It is possible to make the packages that are not
exported explicitly in the module descriptor accessible through the use of
appropriate command line options. For example, if a module does not
export a package, you can still use the --add-exports option to make it
accessible to other modules. Of course, this option breaks the encapsulation
mechanism and is undesirable. It should be used only in exceptional
circumstances.
As I mentioned earlier, Java modules is a fairly large topic. There are several
important concepts such as module graphs, dependency analysis, and using non-
modular jars that I have not talked about. The reason is that these concepts
require a lot of time to discuss, while the OCP Java 11 Certification Part 1 exam
merely scratches the surface of this topic. You will, at the most, get one question
on it, which you will be able to answer based on the information given in this
chapter. For this reason, it is best to ignore it at this time but you will be required
to study this topic thoroughly for the Part 2 exam.
16.6 Exercise ☝
1. Create a module named compoundinterest containing a class named
CompoundInterestCalculator similar to the simpleinterest
module described in this chapter.
2. Compile the source code of compoundinterest module with and
without using the --module switch. Use -d option to direct the output to
your output directory.
3. Run the CompoundInterestCalculator class with and without
using the --module-path switch.
4. Enhance your module by making the
CompoundInterestCalculator class implement the
InterestCalculator interface of the calculators module.
5. Package compoundinterest module into a jar and run this module
using the --module-path switch.
6. Create a module finance with a class TestClass . It should have a main
method that computes simple interest as well as compound interest. Make
appropriate modifications to simpleinterest and
compundinterest modules for this purpose.
Chapter 17 Understanding Java
Technology and environment
Describe Java Technology and the Java development
17.1 Java Technology and key features of the Java

language ☝
After going through the previous chapters, you already know all you need to
know about this objective. I will just summarize the important points here.
Key points on Java Technology and the Java Development
1. Platform Independence - Java code is compiled into Java bytecode,

which is interpreted by a virtual machine called the Java Virtual Machine
(JVM). JVM is available for multitude of platforms (CPU+OS
architectures). This means that the bytecode can run on all those platforms
without any change. Thus, you do not need a Java compiler for every
platform on which you want to run a Java class. The a class file produced
on one platform will run without on any other platform as it is, if there is a
JVM for that platform. If there is no JVM for a particular platform (for
example, Android or iPhone), you cannot run a Java program on that
platform.
2. Java Installation - Java is not a part of the Operating System. It is an

application itself and is installed separately on top of the Operating System.
A separate installable is available for most main stream desktop OSs such
as Windows, MacOS, and Linux. As of now, the Java platform is not
available for Android and iOS.
3. JRE vs JDK - The Java Runtime Environment (JRE) includes just the class
libraries and executables that are required to run a Java program while the
Java Development Kit (JDK) includes the tools such as the Java compiler
and the Java debugger that are required to develop a Java program in
addition to a JRE. Thus, it is not necessary to install the JDK if you just
want to run a Java program but installation of the JDK requires the
installation of the JRE.
4. Development tools - Java Development Kit (JDK) comes bundled with

several applications. Some are pretty much required such as the compiler
(javac) and some are useful while development such as debugger (), class
inspector (javap), and, documentation generator (javadoc). It also comes
with the JVM (java) for most common platforms.
5. IDE - An Integrated Development Environment (IDE) is an additional third

party tool that makes Java development easier. It contains several tools on
top of the tools available in the JDK such as code editor, syntax highlighter,
and source organizer. However, an IDE is not essential for developing a
Java program. Netbeans, Eclipse, and IntelliJ are some commonly used
IDEs.
6. Java Platform configurations - Due to modularization of the JDK, it is

now possible to split the Java runtime as well as library into configurations
of varying features and sizes. While the core Java features are bundled in a
readymade configuration named as the Java SE platform, a minimal JRE
image for your application can also be created using just the modules that
are required by your application.
Key Features of Java -
1. Object-Oriented - Java has features such as classes, objects, and access

control, that allow you to do object-oriented development. It eliminates
some non-OO features such as standalone functions. The following are
some of the "object-oriented features" of Java:
1. Encapsulation ensures that classes can be designed so that only

certain fields and methods of an object are accessible from other
objects. Java allows precise access control by marking data members
as public/protected/private (or default), which promotes encapsulation.
2. Java allows a class to extend at most one class but allows a class to
implement more than one interfaces. Java supports multiple
inheritance of type but does not support multiple inheritance of
state and implementation .
3. Polymorphism ensures that at run time the method to be executed

depends on the actual object referred to by a reference. If a subclass
overrides a method of a base class and if the object referred to by a
variable is of type subclass, then the subclass's version of the method
is used even if the declared type of the variable is of base class. This is
also called dynamic binding. Java supports dynamic binding and
polymorphism.
2. Huge standard library - Java Runtime Environment includes a huge set of

readymade classes are useful for a wide range of applications such as
networking, file and database I/O, text processing, data structures, and so
on. It includes implementation for several networking protocols such as
HTTP and protocols for communicating with a database server. It does not
include a database engine though.
3. Less Complex - Java eliminates a lot of complicated programming

constructs to make it the code less prone to errors. For example, Java does
not have pointers, multiple class inheritance, operator overloading, goto,
and pragmas.
4. Garbage collection - Java frees the developers from actively coding for
garbage collection. It performs checks in the background that identify
unused object and cleans them up. Garbage collection is done on low
priority and so, it does not affect performance.
5. Secure - A Java application can be run with a security manager. This

security manager can be customized to allow precisely only those
operations that you want to allow for an application. Third party Java
applications downloaded from the internet can be run within a sandbox.
This limits the operations that a program can do on the host machine.
6. Multithreading - Java makes developing multithreaded applications a lot

easier than other languages. It offers lower level language constructs such
as monitors as well as higher level multithreading API such as Fork/Join
framework that helps you develop multithreaded and concurrent programs.
However, Java does not parallelize code execution automatically on its
own.
7. High Performance - Java used to be criticized a lot at one point in the past
for its poor performance. However, that was a long time ago. Java
interpreters are now highly optimized and deliver almost similar
performance as a native application. One of the most important features of
Java interpreters that contributes to the high performance is their ability to
monitor and optimize code blocks that are executed frequently in an
application.
Here are a few links worth going through for a more detailed discussion.
Platform Independence - http://stackoverflow.com/questions/2748910/how-

is-java-platform-independent-when-it-needs-a-jvm-to-run
and http://functionspace.org/topic/1689/Why-is-java--called-a-platform-
independent-language-
Object-Oriented Concepts
- http://docs.oracle.com/javase/tutorial/java/concepts/
Encapsulation - http://en.wikipedia.org/wiki/Encapsulation_(object-
oriented_programming)
ENTHUWARE END USER LICENSE AGREEMENT
Go to www.enthuware.com/eula to access Enthuware ebook EULA.

Deshmukh - Java SE 11 Programmer I Study Guide 2019

Uploaded by

Copyright:

Available Formats

Deshmukh - Java SE 11 Programmer I Study Guide 2019

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Deshmukh - Java SE 11 Programmer I Study Guide 2019

Uploaded by

Copyright:

Available Formats

Oracle©

Certified Professional Java SE 11 Programmer I

(Exam Code 1Z0-815)

18th Sep 2019 Build 1.7

Ещё больше книг в нашей группе:

Cover Design: Kino A Lockhart, LOXarts Development, http://www.loxarts.com .

Indian Institute of Technology, Varanasi

I am also thankful to countless CodeRanch.com and Enthuware.com forum

Hanumant Deshmukh is a professional Java architect, author, and director of

Hanumant achieved his Bachelor of Technology from Institute of Technology,

You may reach him at [email protected]

Describe Java Technology and the Java development environment

Working With Java Primitive Data Types and String APIs

Declare and initialize variables (including casting and promoting primitive

Working with Java Arrays

Declare, instantiate, initialize and use a one-dimensional array

Creating and Using Methods

Create methods and constructors with arguments and return values

Reusing Implementations Through Inheritance

Create and use subclasses and superclasses

Describe the advantages of Exception handling and differentiate among

Creating a Simple Java Program

Create an executable Java program with a main class

Using Operators and Decision Constructs

Use Java operators including the use of parenthesis to override operator

Describing Objects and Classes

Declare and instantiate Java objects, and explain objects' lifecycles

Apply access modifiers

Programming Abstractly Through Interfaces

Create and implement interfaces

Describe the Modular JDK

Chapter 1 Kickstarter for Beginners

Chapter 2 Creating a Simple Java Program

Chapter 3 Working With Java Primitive Data Types

Chapter 4 Describing and Using Objects and Classes

Chapter 5 Working with String APIs

Chapter 6 Using Operators

Chapter 7 Using Decision Constructs

Chapter 8 Using Loop Constructs

Chapter 9 Creating and Using Arrays

Chapter 10 Creating and Using Methods

Chapter 12 Reusing Implementations Through Inheritance

Chapter 13 Programming Abstractly Through Interfaces

Chapter 14 Lambda Expressions

Chapter 15 Handling Exceptions

Chapter 16 Understanding Modules

Chapter 17 Understanding Java Technology and environment

About the mock exams

1. Do you know what OS, RAM, and CPU are?

Conventions used in this book

Asking for clarification

0.4 Will this be asked in the exam? ☝

Since we, at Enthuware, conduct classroom training as well, we get to

0.5 General tips for the exam ☝

ArrayList al = new ArrayList();

0.6 Official Exam Details and Exam Objectives ☝

Duration : 180 Minutes

Passing Score: 63%

Format: Multiple Choice