Common Functional Programming Algebraic Data Types for JavaScript

Functional

Common Functional Programming Algebraic data types for JavaScript that is compatible with most modern browsers and Deno.

Usage

This example uses the Ramda library - for simplification - but you should be able to use any library that implements the Fantasy-land specifications.

import { compose, converge, lift, map, prop } from "https://x.nest.land/ramda@0.27.0/source/index.js";
import Either from "https://deno.land/x/functional@v0.5.4/Either.js";
import Task from "https://deno.land/x/functional@v0.5.4/Task.js";

const fetchUser = userID => Task.wrap(_ => fetch(`${URL}/users/${userID}`).then(response => response.json()));

const sayHello = compose(
  converge(
    lift((username, email) => `Hello ${username} (${email})!`),
    [
      map(prop("username")),
      map(prop("email"))
    ]
  ),
  fetchUser
);

// Calling `sayHello` results in an instance of `Task` keeping the function pure.
assert(Task.is(sayHello(userID)));

// Finally, calling `Task#run` will call `fetch` and return a promise
sayHello(userID).run()
  .then(container => {
    // The returned value should be an instance of `Either.Right` or `Either.Left`
    assert(Either.Right.is(container));
    // Forcing to coerce the container to string will show that the final value is our message.
    assert(container.toString(), `Either.Right("Hello johndoe (johndoe@gmail.com)!")`);
  });

// await sayHello(userID).run() === Either.Right(String)

Using the bundle

As a convenience, when using Functional in the browser, you can use the unminified bundled copy.

import { compose, converge, lift, map, prop } from "https://x.nest.land/ramda@0.27.0/source/index.js";
import { Either, Task } from "https://deno.land/x/functional@v0.5.4/functional.js";

const fetchUser = userID => Task.wrap(_ => fetch(`${URL}/users/${userID}`).then(response => response.json()));

const sayHello = compose(
  converge(
    lift((username, email) => `Hello ${username} (${email})!`),
    [
      map(prop("username")),
      map(prop("email"))
    ]
  ),
  fetchUser
);

Type factory

The Type factory can be used to build complex data structure.

import { factorizeType } from "https://deno.land/x/functional@v0.5.4/SumType.js";

const Coordinates = factorizeType("Coordinates", [ "x", "y" ]);
const vector = Coordinates(150, 200);
// vector.x === 150
// vector.y === 200

Type.from

from :: Type ~> Object -> t

Create an instance of Type using an object representation.

const vector = Coordinates.from({ x: 150, y: 200 });
// vector.x === 150
// vector.y === 200

Type.is

is :: Type ~> Type t -> Boolean

Assert that an instance is of the same Type.

Coordinates.is(vector);
// true

Type.toString

toString :: Type ~> () -> String

Serialize the Type Representation into a string.

Coordinates.toString();
// "Coordinates"

Type(a).toString

toString :: Type t => t ~> () -> String

Serialize the instance into a string.

vector.toString();
// "Coordinates(150, 200)"

Type Sum factory

import { factorizeSumType } from "https://deno.land/x/functional@v0.5.4/SumType.js";

const Shape = factorizeSumType(
  "Shape",
  {
    // Square :: (Coord, Coord) -> Shape
    Square: [ "topLeft", "bottomRight" ],
    // Circle :: (Coord, Number) -> Shape
    Circle: [ "center", "radius" ]
  }
);

SumType.from

from :: SumType ~> Object -> t

Create an instance of Type using an object representation.

const oval = Shape.Circle.from(
  {
    center: Coordinates.from({ x: 150, y: 200 }),
    radius: 200
  }
);
// oval.center === Coordinates(150, 200)
// oval.radius === 200

SumType.is

is :: SumType ~> SumType t -> Boolean

Assert that an instance is of the same Sum Type.

Shape.Circle.is(oval);
// true

SumType#fold

Shape.prototype.translate =
  function (x, y, z) {
    return this.fold({
      Square: (topleft, bottomright) =>
        Shape.Square(
          topLeft.translate(x, y, z),
          bottomRight.translate(x, y, z)
        ),

      Circle: (centre, radius) =>
        Shape.Circle(
          centre.translate(x, y, z),
          radius
        )
    })
  };

SumType(a).toString

toString :: SumType t => t ~> () -> String

Serialize the instance into a string.

oval.toString();
// "Shape.Circle(Coordinates(150, 200), 200)"

Example of writing a binary tree with Sum Types

import { factorizeSumType } from "https://deno.land/x/functional@v0.5.4/SumType.js";

const BinaryTree = factorizeSumType('BinaryTree', {
  Node: ['left', 'x', 'right'],
  Leaf: []
});

BinaryTree.prototype.reduce = function (f, accumulator) {

  return this.fold(
    {
      Node: (l, x, r) => {
        const left = l.reduce(f, accumulator);
        const leftAndMiddle = f(left, x);

        return r.reduce(f, leftAndMiddle);
      },
      Leaf: () => accumulator
    }
  );
};

const tree =
  BinaryTree.Node(
    BinaryTree.Node(
      BinaryTree.Leaf,
      1,
      BinaryTree.Node(
        BinaryTree.Leaf,
        2,
        BinaryTree.Leaf
      )
    ),
    3,
    BinaryTree.Node(
      BinaryTree.Node(
        BinaryTree.Leaf,
        4,
        BinaryTree.Leaf
      ),
      5,
      BinaryTree.Leaf
    )
  );

// tree.reduce((x, y) => x + y, 0) === 15

Maybe type

The Maybe type represents potentially Just a value or Nothing.

import Maybe from "https://deno.land/x/functional@v0.5.4/Maybe.js";

const container = Maybe.Just(42);

const serialize = (container) =>
  container.fold({
    Nothing: () => "There is no value.",
    Just: value => `The value is ${value}.`
  });

// serialize(container) === "The value is 42."

This implementation of Maybe is a valid Filterable, Functor, Applicative, Alternative, Foldable, Traversable and Monad.

Either type

The Either type represents the possibility of two values; either an a or a b.

import Either from "https://deno.land/x/functional@v0.5.4/Either.js";

const container = Either.Right(42);

const serialize = (container) =>
  container.fold({
    Left: value => `An error occured: ${value}.`,
    Right: value => `The value is ${value}.`
  });

// serialize(container) === "The value is 42."

This implementation of Either is a valid Functor, Applicative, Alternative, Foldable, Traversable and Monad.

IO type

The IO type represents a function that access IO. It will be lazily executed when the #run method is called.

import IO from "https://deno.land/x/functional@v0.5.4/IO.js";

// Eventually 42
const container = IO(_ => Promise.resolve(42));

const multiply = container.map(promise => promise.then(x => x * x));
const add = container.map(promise => promise.then(x => x + x));

// multiply === IO(Function)
// add === IO(Function)

const multiplyThenAdd = multiply.map(promise => promise.then(x => x + x));

// multiply.run() === Promise(1764)
// add.run() === Promise(84)
// multiplyThenAdd.run() === Promise(3528)

This implementation of IO is a valid Functor, Applicative and Monad.

Task type

The Task type represents a function that access IO. It will be lazily executed when the #run method is called. Unlike IO, the Task type also abstract away the promise making for a more intuitive experience.
Note that the function must return an instance of Either; Either.Right to represent a success and Either.Left to represent a failure. Also check-out the Task.wrap method.

If the runtime throws an error, the final value will be Either.Left(error).

import Either from "https://deno.land/x/functional@v0.5.4/Either.js";
import Task from "https://deno.land/x/functional@v0.5.4/Task.js";

// Eventually 42
const container = Task(_ => Promise.resolve(Either.Right(42)));

const multiply = container.map(x => x * x);
const add = container.map(x => x + x);

// multiply === Task(Function)
// add === Task(Function)

const multiplyThenAdd = multiply.map(x => x + x);

// await multiply.run() === Either.Right(1764)
// await add.run() === Either.Right(84)
// await multiplyThenAdd.run() === Either.Right(3528)

Task.wrap

Create a wrapped instance of Task. An instance of Task made using the wrap method is different in two ways:

1. The result of the function call is memoized;
2. If the function call was successful, the value will automatically be an instance of Either.Right;

import Task from "https://deno.land/x/functional@v0.5.4/Task.js";

let count = 0;
const fetchUser = userID => Task.wrap(
  _ => ++count && fetch(`${URL}/users/${userID}`).then(response => response.json())
);

const user = fetchUser(userID);
const username = user.map(({ username }) => username);
const email = user.map(({ email }) => email);

// await user.run() === Either.Right({ email: "johndoe@gmail.com", username: "johndoe" })
// await username.run() === Either.Right("johndoe")
// await email.run() === Either.Right("johndoe@gmail.com")
// count === 1

This implementation of Task is a valid Functor, Applicative, Alternative and Monad.

TypeScript

I will try to publish TypeScript type hint files for those who needs it.
So far, I’ve only implemented the Type factory functions.

// @deno-types="https://deno.land/x/functional@v0.5.4/SumType.d.ts"
import { factorizeType, factorizeSumType } from "https://deno.land/x/functional@v0.5.4/SumType.js";

Deno

This codebase uses the assertion library from Deno.

Download Details:

Author: sebastienfilion

Source Code: https://github.com/sebastienfilion/functional

#deno #nodejs #node #javascript