Riese

DISCLAIMER: This gem doesn't claim to be a mathematic correct implementation. If you need a implementation of fraction in ruby, checkout the Rational library shipped with ruby.

Introduction

See the story behind it

To play with it, just grab a ruby version (only tested with MRI Ruby 1.9) and run the following commands:

gem install bundler
bundle install --path vendor/local

To execute the Minitest Test file run:

bundle exec rake test

To execute the RSpec examples run:

bundle exec rake spec

To start the file watcher, run:

bundle exec guard

Purpose

This library is solely for the purpose to demonstrate a couple of programming paradigms to newcomers:

• Basic OOP
• Mutability/Immutability
• TDD using Minitest
• BDD using RSpec

Object Oriented Programming

This library only consists of one class Riese::Fraction. This class is a representation of a fraction. You can initialize it with a numerator and a denominator. It implements the following (public) instance methods:

• • (addition)
• • (subtraction)
• • (multiplication)
• / (division)
• inverse
• to_s
• ==

It has one class method:

• init (makes sure the inputted value is converted to fraction)

Additionally there are 2 private methods:

• normalize (takes care of reducing the fraction and makes sure that negative fractions only have a negative numerator)
• gcd (Euclidean algorithm for determing the greatest common divisor)

If used for explanation these 2 methods can be treated as a black box.

Mutability/Immutability

The underlying instance variables are only set once. All actions on the instance will never change these variables and always return a new instance of a fraction.

TDD using Minitest

For testing the functionality of the Fraction minitest/unit is used. I believe that for a total beginner the assert-Syntax is easier to understand. So all the functionality is tested using this framework.

BDD using RSpec

The same tests that have been implemented using minitest, are also available as examples in RSpec. For people that already have a background in OOP or Test::Unit syntax, this might be useful.

Limitations

• Currently it is not possible to normalize fractions with a float numerator or denominator
• the used algorithm for determing the greatest common divisor produces a SystemStackError very fast (try initalize the fraction 1/10000)