Smart Results for JS

Introduction

This is just a small library to make error handling somewhat joy in TypeScript. It is heavily inspired by Rust and its great concepts about error handling.

How to Use?

Currently there are two types you can mak use of:

• Option<T> which is a union type of Some<T> and None<T>
• Result<T, E> which is a union type of Ok<T, E> and Err<T, E>

The type Option<T> comes with two helper functions, some(value) and none(). You should use these functions to create an Option<T>. The type Result<T, E> follows the same principle except that you will have ok(value) and err(value) to create it.

The following code provides a brief overview which helps understanding the concepts.

import { Option, some, none, Result, ok, err } from '@cakasim/smart-results';

// Just a simple type for us to play with
type Animal = {
  name: string;
};

// Another type which holds metrics about an animal
type Metric = {
  minSize: number;
  maxSize: number;
  isCarnivorous: boolean;
  cutenessFactor: number;
};

// A pre-filled list of animals
const animals = new Map<number, Animal>([
  [1, { name: 'Ape' }],
  [2, { name: 'Dog' }],
  [3, { name: 'Cat' }],
  [4, { name: 'Pig' }],
  [5, { name: 'Bird' }],
])

/**
 * Finds an animal in the list and returns an Option<Animal>
 * which holds a valid animal if we have found an entry or nothing
 * if there is no entry for the provided ID.
 *
 * @param id Animal ID.
 */
function findAnimal(id: number): Option<Animal>
{
  // Get animal for provided ID, or undefined if ID is invalid
  const animal = animals.get(id);

  // Check if animal is not undefined and return either Some or None
  return animal
    ? some(animal)
    : none();
}

/**
 * Fetches a metric dataset about the provided animal from an HTTP Rest API.
 * The return value is a Promise that (always) resolves to a Result<Metric, Error>.
 * If a metric could be loaded the result contains a valid Metric value.
 * Otherwise the result contains an Error value.
 *
 * @param animal The animal.
 */
async function fetchAnimalMetric(animal: Animal): Promise<Result<Metric, Error>>
{
  // Assume we are fetching actual data here
  const response = await fetchAnimalMetricFromRestApi(animal);

  // Check if the response code is good and return either Ok or Err
  return response.statusCode === 200
    ? ok(response.data as Metric)
    : err(new Error(`Could not fetch from HTTP API, go status code ${response.statusCode}.`));
}

/**
 * This is the function that gets actually executed by someone.
 *
 * @param animalId Animal ID.
 */
async function printAnimalMetric(animalId: number)
{
  // Search for the animal, *expect* a None result and handle it by throwing the provided Error.
  // If the result is Some no error will be thrown. Please note that `expect()` should be used
  // carefully only if the exception case can be handled properly.
  const animal = findAnimal(animalId).expect(() => new Error(`Cannot find animal for ID ${animalId}.`));

  // metric is a Result<Metric, Error>, at this point we are forced to check for a valid result
  const metric = await fetchAnimalMetric(animal);

  if (metric.isErr()) {
    // Handle the error case
    console.error(`Failed to fetch data for animal ID ${animalId} with name ${animal.name} from REST API: ${metric.value.message}`);
  } else {
    // At this point metric.value is safe to use
    console.log(`The metric for animal ID ${animalId} with name ${animal.name} is:`);
    console.log(metric.value);
  }
}