Ways of Creating Single Case Discriminated Unions in F#

In this post, I will go through a number of the approaches that I have seen of creating single case discriminated unions in F#.

There are quite a few ways of creating single case discriminated unions in F# and this makes them popular for wrapping primatives. In this post, I will go through a number of the approaches that I have seen. The inspiration for this post is Twitter threads from the likes of @McCrews, @asp_net and @jordan_n_marr. It is in no way exhaustive but should give you plenty of ideas.

Basics

We start with a simple single case discriminated union:

type CustomerId = CustomerId of Guid

To create a CustomerId and deconstruct it to get the value, we can do the following:

let id = CustomerId (Guid.NewGuid())
let (CustomerId value) = id

We can add a private modifier to it which means that we do not have access to the constructor and cannot create a CustomerId outside of the module the type is defined in:

type CustomerId = private CustomerId of Guid

// Problem outside of current module
let id = CustomerId (Guid.NewGuid())
let (CustomerId value) = id

This is easy to solve in a number of ways, firstly using a module.

Using a Module

If we create a module with the same name as the type definition, we can add functions to create and extract the value:

type CustomerId = private CustomerId of Guid

module CustomerId =
    let New() = CustomerId (Guid.NewGuid())
    let Value (CustomerId value) = id

We can now create a CustomerId and extract the value from it like this:

let id = CustomerId.New()
let value = CustomerId.Value id

It's also possible to do this without the need for a module.

Without Using a Module

We can add an instance member for extracting the value and a static member to create an instance of CustomerId:

type CustomerId = private CustomerId of Guid
    with
        member this.Value =
            let (CustomerId value) = this
            value
        static member New() = CustomerId (Guid.NewGuid())

This now makes extracting the value much cleaner:

let id = CustomerId.New()
let value = id.Value

If you think that the two lines for the Value function are one too many, we can write it in one line. This is the first version using the in keyword:

type CustomerId = private CustomerId of Guid
    with
        member this.Value = let (CustomerId value) = this in value
        static member New() = CustomerId (Guid.NewGuid())

It doesn't impact how we create and extract the data:

let id = CustomerId.New()
let value = id.Value

Another way is to use an anonymous function:

type CustomerId = private CustomerId of Guid
    with
        member this.Value = this |> fun (CustomerId value) -> value
        static member New() = CustomerId (Guid.NewGuid())

Again the usage is the same as before:

let id = CustomerId.New()
let value = id.Value

Having a separate New function means that we could add logic to ensure that it cannot be created with an invalid value. If you don't need that, we can remove the private access modifier and the New function:

type CustomerId = CustomerId of Guid with member this.Value = this |> fun (CustomerId value) -> value

Usage then looks like this:

let id = CustomerId (Guid.NewGuid())
let value = id.Value

Alternate Solution

After I'd posted this article, @Savlambda suggested an alternate approach using an active pattern:

module Identifier =
    type CustomerId = private CustomerId of Guid
        with
            static member New() = CustomerId (Guid.NewGuid())

    let (|CustomerId|) (CustomerId value) = value

Using the new module looks like this:

module OtherModule =
    open Identifier

    let id = CustomerId.New()
    let (CustomerId value) = id

So we have quite a few styles and I'm not going to suggest the superiority of one over the others. In one of the Twitter threads that led to this post, @pblasucci said that you can do the same with records as well!
 

Using a Record Type

Let's create a simple record type:

type CustomerId = { CustomerId : Guid }

Creation and value extraction look like this:

let id = { CustomerId = Guid.NewGuid() }
let value = id.CustomerId

That's not too different! We can even add member functions to record types:

type CustomerId = { CustomerId : Guid }
    with 
        member this.Value = this.CustomerId
        static member New() = { CustomerId = Guid.NewGuid() }

This means that we can create and extract the value in exactly the same way that we did with single case discriminated unions:

let id = CustomerId.New()
let value = id.Value

I don't know what impact using records has on performance, maybe that will be another post?

Summary

In this post, we have seen a few styles of single case discriminated unions and records to wrap up primatives. It is pretty trivial to use any of the approaches suggested in this artlcle and will give you additional type safety over that given by a raw staticaly typed primative, particularly when you have rules that define the type.
As always, let me know what you think about this post or any suggestions about future posts. https://twitter.com/ijrussell

Blog 12/3/21

Using Discriminated Union Labelled Fields

A few weeks ago, I re-discovered labelled fields in discriminated unions. Despite the fact that they look like tuples, they are not.

Blog 3/11/21

Introduction to Web Programming in F# with Giraffe – Part 2

In this series we are investigating web programming with Giraffe and the Giraffe View Engine plus a few other useful F# libraries.

Blog 3/22/23

Introduction to Functional Programming in F# – Part 8

Discover Units of Measure and Type Providers in F#. Enhance data management and type safety in your applications with these powerful tools.

Blog 3/12/21

Introduction to Web Programming in F# with Giraffe – Part 3

In this series we are investigating web programming with Giraffe and the Giraffe View Engine plus a few other useful F# libraries.

Blog 3/22/23

Introduction to Functional Programming in F# – Part 9

Explore Active Patterns and Computation Expressions in F#. Enhance code clarity and functionality with these advanced techniques.

Blog 8/7/20

Understanding F# Type Aliases

In this post, we discuss the difference between F# types and aliases that from a glance may appear to be the same thing.

Blog 3/10/21

Introduction to Web Programming in F# with Giraffe – Part 1

In this series we are investigating web programming with Giraffe and the Giraffe View Engine plus a few other useful F# libraries.

Blog 10/1/22

Introduction to Functional Programming in F# – Part 4

Unlock F# collections and pipelines. Manage data efficiently and streamline your functional programming workflow with these powerful tools.

Blog 12/22/22

Introduction to Functional Programming in F# – Part 6

Learn error handling in F# with option types. Improve code reliability using F#'s powerful error-handling techniques.

Blog 5/18/22

Introduction to Functional Programming in F#

Dive into functional programming with F# in our introductory series. Learn how to solve real business problems using F#'s functional programming features. This first part covers setting up your environment, basic F# syntax, and implementing a simple use case. Perfect for developers looking to enhance their skills in functional programming.

Blog 5/17/23

Introduction to Functional Programming in F# – Part 10

Discover Agents and Mailboxes in F#. Build responsive applications using these powerful concurrency tools in functional programming.

Blog 8/8/23

Introduction to Functional Programming in F# – Part 12

Explore reflection and meta-programming in F#. Learn how to dynamically manipulate code and enhance flexibility with advanced techniques.

Blog 10/11/22

Introduction to Functional Programming in F# – Part 5

Master F# asynchronous workflows and parallelism. Enhance application performance with advanced functional programming techniques.

Blog 9/13/22

Introduction to Functional Programming in F# – Part 2

Explore functions, types, and modules in F#. Enhance your skills with practical examples and insights in this detailed guide.

Blog 9/15/22

Introduction to Functional Programming in F# – Part 3

Dive into F# data structures and pattern matching. Simplify code and enhance functionality with these powerful features.

Blog 11/30/22

Introduction to Partial Function Application in F#

Partial Function Application is one of the core functional programming concepts that everyone should understand as it is widely used in most F# codebases.In this post I will introduce you to the grace and power of partial application. We will start with tupled arguments that most devs will recognise and then move onto curried arguments that allow us to use partial application.

Blog 7/21/20

Understanding F# applicatives and custom operators

In this post, Jonathan Channon, a newcomer to F#, discusses how he learnt about a slightly more advanced functional concept — Applicatives.

Blog 12/22/22

Introduction to Functional Programming in F# – Part 7

Explore LINQ and query expressions in F#. Simplify data manipulation and enhance your functional programming skills with this guide.

Blog 6/27/23

Boosting speed of scikit-learn regression algorithms

The purpose of this blog post is to investigate the performance and prediction speed behavior of popular regression algorithms, i.e. models that predict numerical values based on a set of input variables.

Blog 7/12/23

Introduction to Functional Programming in F# – Part 11

Learn type inference and generic functions in F#. Boost efficiency and flexibility in your code with these essential programming concepts.

Bleiben Sie mit dem TIMETOACT GROUP Newsletter auf dem Laufenden!