Designing and Implementing a class

Then testing the class

Step 1: The specifications

• Find out what you are asked to do with an object of the class (i.e. start with what needs to be done)

• Suppose you are asked to implement a CashRegister class that will be used for customers in a store

• Make a list, in plain English, of the operations that an object of the class should carry out:

• Think of the object of the class as the financial record for a customer in the store

– Ring up the price for each item a customer buys

– Enter the amount of payment (may be more than one payment)

– Calculate the amount of change due to the customer

Step 2: Specify the public interface

• How will the user of the class (the client) want to use it?

– In your mind, you need to separate the writer of the class from the user or client of the class

• Turn the list in Step 1 into a set of methods the client can call

– Each method must have a name, a return type, and arguments

– Give some thought to the names; they should reflect what the method will do

– Information flow from/to user of class

• Decide what information must come from the user (arguments),

• what information must be returned to the user (return type)

Step 3: Determine the instance variables

• What information needs to be kept for each object?

– In this case, an object is the financial record of one customer

• Decide on the data type for each instance variables

Step 4: Implement the methods

• This includes a constructor and the methods we have designed

– I also included a toString method to make it easier to test the class

• Mutators and accessors if they are needed

• The instance variables should be private, while the methods should be public

Step 5: Test your class

• Usually this is done in a separate class with a main method

• The tester is often called the client or the driver

• Every method should be called, to be sure it is doing what it should

• If your class has a toString method, outputting the entire object is easier.

Formatting floats with printf

• printf(%f) lets you specify two things

– The width of the field

– The number of digits after the decimal point (called the precision)

Numbers in the computer

Primitive types

Four integers type

byte -- 8 bit signed integral value

short -- 16-bit signed

int-- 32 bit signed (most commonly used)

long -- 64 bits

Floats

float – 32 bits (6 to 7 significant decimal digits , exponent to 38)

double – 64 bits (15 significant decimal digits, exponent to 308)
Floats are stored in the computer differently than integers

– Floating point calculations are much slower than integer calculations

Problems with numbers

• Overflow with ints

• Rounding errors with floats or doubles