Book Image

Soar with Haskell

By : Tom Schrijvers
Book Image

Soar with Haskell

By: Tom Schrijvers

Overview of this book

With software systems reaching new levels of complexity and programmers aiming for the highest productivity levels, software developers and language designers are turning toward functional programming because of its powerful and mature abstraction mechanisms. This book will help you tap into this approach with Haskell, the programming language that has been leading the way in pure functional programming for over three decades. The book begins by helping you get to grips with basic functions and algebraic datatypes, and gradually adds abstraction mechanisms and other powerful language features. Next, you’ll explore recursion, formulate higher-order functions as reusable templates, and get the job done with laziness. As you advance, you’ll learn how Haskell reconciliates its purity with the practical need for side effects and comes out stronger with a rich hierarchy of abstractions, such as functors, applicative functors, and monads. Finally, you’ll understand how all these elements are combined in the design and implementation of custom domain-specific languages for tackling practical problems such as parsing, as well as the revolutionary functional technique of property-based testing. By the end of this book, you’ll have mastered the key concepts of functional programming and be able to develop idiomatic Haskell solutions.
Table of Contents (23 chapters)
Free Chapter
1
Part 1:Basic Functional Programming
6
Part 2: Haskell-Specific Features
11
Part 3: Functional Design Patterns
16
Part 4: Practical Programming

An abstraction for structural recursion

In the previous chapter, we studied recursive functions and presented structural recursion schemes as a useful template to write such functions. Thanks to HOF, we can capture these structural recursion schemes in reusable functions.

Folding lists

The first structural recursion scheme we encountered in the previous chapter is that for lists:

f :: [A] -> B
f []     = n
f (x:xs) = c x (f xs)

By replacing the underlined elements in the preceding template with concrete names, we obtain different functions, such as sum and and:

Prelude
sum :: [Integer] -> Integer
sum []     = 0
sum (x:xs) = x + sum xs
and :: [Bool] -> Bool
and []     = True
and (x:xs) = x && and xs

Thanks to the ability to parameterize over functions, we can actually do better than this. Indeed, we can capture the recursion scheme in a HOF called foldr:

Prelude
foldr :: (a...