Book Image

Learn Type-Driven Development

By : Yawar Amin, Kamon Ayeva
Book Image

Learn Type-Driven Development

By: Yawar Amin, Kamon Ayeva

Overview of this book

Type-driven development is an approach that uses a static type system to achieve results including safety and efficiency. Types are used to express relationships and other assumptions directly in the code, and these assumptions are enforced by the compiler before the code is run. Learn Type-Driven Development covers how to use these type systems to check the logical consistency of your code. This book begins with the basic idea behind type-driven development. You’ll learn about values (or terms) and how they contrast with types. As you progress through the chapters, you’ll cover how to combine types and values inside modules and build structured types out of simpler ones. You’ll then understand how to express choices or alternatives directly in the type system using variants, polymorphic variants, and generalized algebraic data types. You’ll also get to grips with sum types, build sophisticated data types from generics, and explore functions that express change in the types of values. In the concluding chapters, you’ll cover advanced techniques for code reuse, such as parametric polymorphism and subtyping. By end of this book, you will have learned how to iterate through a type-driven process of solving coding problems using static types, together with dynamic behavior, to obtain more safety and speed.
Table of Contents (12 chapters)

Forcing a difference with phantom types

Because we can declare types that can slot in any type parameters, that includes type parameters that the types don't actually use. These are called phantom type parameters, or more informally phantom types.

A common use case for phantom types is in a kind of type-safe builder pattern. (The builder pattern is a piece of code that helps us to construct an object according to specific rules.) For example, we might want to construct syntactically valid SQL statements. One way to do that is to have a validator function that takes an input SQL statement and decides whether it follows SQL syntax rules or not at runtime. This function might try to parse the input statement and build an expression tree. If the tree can be built, the statement is valid. Otherwise, it's invalid.

Another way to approach this is to provide a set of functions...