코드베이스를 향상시키는 5가지 TypeScript 기능

Let's consider the following example.

function getNumberOrDefault(value?: number) {
  return value || 22;
}

In the mentioned example, we would like to return a default value, if the provided value is undefined. However, this code has a severe pitfall — javascript treats 0 as a falsy value.

console.log(getNumberOrDefault(10)) // Output 10
console.log(getNumberOrDefault(0)) // Output 22

Returning 22 is probably, not something we expect this function to do. Thanks to the nullish coalescing we don't need to refactor the code and add check for 0 , instead we can use a shorthand syntax with ?? .

function getNumberOrDefaultTs(value?: number) {
    return value ?? 22;
}

console.log(getNumberOrDefault(10)) // Output 10
console.log(getNumberOrDefault(0)) // Output 0

Let's imagine we have two types, Truck and F1 , which extend the same interface Car . Although they share some common properties, they also possess a set of unique attributes like load and acceleration respectively.

interface Car {
  model: string;
}

interface Truck extends Car {
  load: number;
}

interface F1 extends Car {
  acceleration: number;
}

function scanCar(car: Truck | F1) {
    ...some code
}

Also, we have a function scanCar which accepts either a Truck or a F1 and performs some type-specific actions.

Type guard is a great way to let typescript know with which type we are currently dealing.

function carIsTruck(car: Truck | F1): car is Truck {
  return 'load' in car;
}

function scanCar(car: Truck | F1) {
  if(carIsTruck(car)) {
        // Only truck properties will be suggested
    console.log(`Truck has a load of ${car.load}`)
  } else {
        // Only F1 properties will be suggested
    console.log(`F1 has acceleration of ${car.acceleration}`)
  }
}

Type guard is a function that returns a boolean, which is often used as part of the conditional statement, to let typescript know which type is assumed in the current if block.

There are some cases when we want to enforce object keys being certain values only. And here is how we can achieve it using enum and Record .

enum Size {
  M,
  L,
  XL
}

function getPlainTextSizes(): Record<Size, string> {
  return {
    [Size.M]: "medium",
    [Size.L]: "large",
    [Size.XL]: "extra large",
    10: "small", // error
    "size": "xs" // error
  }
}

Record is a generic type util, which allows defining types for key-value maps.

interface Item<T> {
  id: number,
  value: T,
}

interface Post {
  title: string;
}

const item1: Item<number> = { id: 1, value: 10 }
const item2: Item<Post> = { id: 1, value: { title: "Post name" } }

Generics in typescript significantly improve code reusability.

interface Book {
  id: number;
  author: string;
}

interface Recipe {
  id: number;
  cookingTime: number;
}

function mapById<T extends { id: number }>(array: T[]): {[key: number]: T} {
  return array.reduce((map, item) => ({ ...map, [item.id]: item }), {})
}

const books: Book[] = [{ id: 1, author: "A" }];
const recipies: Recipe[] = [{ id: 1, cookingTime: 10 }];

mapById(books);
mapById(recipies);

It's often the case that we have util functions which require only some field from the interface. In the example above you can see a function which takes an array as input and returns values mapped by id. The only important thing is that the items of the array have an id . When you pass a parameter to the function typescript will:

1. Infere the type of the provided parameter
2. Check if the parameter matches specified condition (must have id of type number) So now you are save to use this helper for various interfaces.