Terminology plays an important role in understanding TypeScript and JavaScript. Terminology helps us to describe a complex context and prevents us from confusion when talking about code.

As important as it is to know the control structures of a programming language, it is just as important to be able to name their contexts and surroundings. Using the right vocabulary is particularly effective in code reviews as it supports us to put our thoughts into words.

Algorithms & Functions

An algorithm is a set of instructions to solve specific problems or to perform a computation. In TypeScript algorithms can be implemented with functions.

Algorithm Characteristics

Algorithms have defining characteristics such as:

Imperative Programming

Imperative programming is a programming paradigm that uses statements to change an application’s state. In a nutshell, imperative programming defines a set of instructions from start to finish.

Example:

Declarative Programming

Declarative programming is a programming paradigm that defines the desired state of an application without explicitly listing statements that must be executed. In a nutshell, declarative programming defines an application’s execution from finish to start.

Example:

Deterministic Functions

A deterministic function will always produce the same output given the same input. This makes the output of a deterministic function predictable as it does not rely on a dynamic state.

Pure Functions

Pure functions are a subset of deterministic functions. A pure function always produces the same result given a particular input. In addition, it does not cause side effects by avoiding I/O operations like printing to the console or writing to the disk.

A pure function does not mutate its passed parameters and is referentially transparent.

Referentially Transparent Expressions

An expression is referentially transparent when it can be replaced with its return value.

Example:

1
2
3
4
5
function sayHello(): string {
  return `Hello!`;
}

const message = sayHello();

The sayHello function always returns the same text, so we can safely replace our expression with const message = Hello! which makes it referentially transparent.

Identity Function

An identity function returns the identical value that was given to it:

Example:

1
2
3
function identityFunction(text: string): string {
  return text;
}

Referentially Opaque Expressions

An expression is referentially transparent when it cannot be replaced with its return value.

Example:

1
2
3
4
5
function today(): string {
  return new Date().toISOString();
}

const isoDate = today();

At the time of writing the execution of today() returned '2021-09-22T12:45:25.657Z'. This result will change over time, so we cannot replace const isoDate = today() with const isoDate = '2021-09-22T12:45:25.657Z' which makes this expression referentially opaque.

Function Declaration

A function declaration gets hoisted and is written in the following way:

1
2
3
function myFunction(): number {
  return 1337;
}

Function Expression

A function expression is part of an assignment and does not get hoisted:

1
2
3
const myFunction = function(): number {
  return 1337;
}

Function Scope

By default, JavaScript is function scoped, which means that variables are acccessible within a function:

1
2
3
4
5
6
7
8
9
function myFunction(): void {
  if (true) {
    var myNumber = 1337;
  }
  // Variable `myNumber` can be accessed within the function although it is being used outside of its conditional if-block
  return myNumber;
}

console.log(myFunction()); // 1337

Block Scope

To enforce block-scoping in JavaScript, the let keyword can be used. It makes variables unaccessible from the outside of their blocks:

1
2
3
4
5
6
7
8
9
function myFunction(): void {
  if (true) {
    let myNumber = 1337;
  }
  // Will throw a `ReferenceError` because `myNumber` is not defined
  return myNumber;
}

console.log(myFunction()); // Causes an uncaught `ReferenceError`

TypeScript

Ambient Context

By default, the TypeScript compiler does not know in which runtime environment (for instance Node.js v16, Electron v16, Chrome v94) our code will be executed later. That’s why we can help the compiler knowing that by defining an ambience / ambient context.

Example: If you run your code in an environment where there is a “world” object that TypeScript does not know about, you can define that context using declare var:

1
2
3
4
5
declare var world: {
  name: string;
};

console.log(world.name);

Union Types

A union type describes a collection of types (i.e. string and number):

1
2
3
4
type MyUnionType = string | number;

const myName: MyUnionType = 'Benny';
const myAge: MyUnionType = 34;

It is called a union because it unites the amount of possible types. The term union comes from set theory where it is used when two (or more) sets are combined.

Module

In TypeScript a module is a file with at least one top level import or export statement.

Template literal

1
const templateLiteralString = `1 + 1 = ${1 + 1}`;

Tagged template

1
2
3
4
5
6
7
8
9
10
import gql from 'graphql-tag';

const query = gql`
  {
    user(id: 5) {
      firstName
      lastName
    }
  }
`

Array Destructuring

1
const [bart, homer, marge] = ['Bart', 'Homer', 'Marge'];

Type Annotation

A type annotation is when you explicitly define the type that a variable can take on:

1
const x: number = 10;

Type Inference

When there is no explicit type annotation then TypeScript will infer the type for you:

1
2
/** TypeScript infers `number` because "x" is initialized with a number and can be reassigned. */
let x = 10; 
1
2
/** TypeScript infers `10` because "x" is a constant and cannot be reassigned. */
const x = 10;
1
2
/** TypeScript infers an array of `(string | number)` value types. */
const x = [10, '11'];

Type Argument / Type Parameter

In the following code, <string> is the type argument:

1
const array = new Array<string>();

Type Argument Inference

TypeScript can infer types but also type arguments. This usually happens when the arguments of your generic functions are connected to the generic type of your function. In this case TypeScript can infer the type of your type variable by inspecting the input values of your function call:

1
2
3
4
5
6
function yourGenericFunction<T>(input: T[]): number {
  return input.length;
}

/** By inspecting the input of this function call, TypeScript will infer `yourGenericFunction<string>`. */
yourGenericFunction(['Benny']);

Type Variable

A type variable is the placeholder for a generic type in your generic code:

1
2
3
function yourGenericFunction<MyTypeVariable>(input: MyTypeVariable[]): number {
  return input.length;
}

Type variables are written by using the angle brackets and defining a name for the variable (e.g. <T>). This construct is often referred to as the diamond operator because the angle brackets look like a diamond (<>, 💎).

String Literal Type

A string literal is a specific string (e.g. “click”) and a string literal type is the type describing a specific string:

1
const action: 'click' = 'click';

Template Literal Type

A template literal type is a combination of Template literal and string literal type. It is capable of resolving string interpolations within a template string to use it for strong typing:

1
2
3
4
5
type Action = 'click';

type ClickEvent = `${Action}Event`;

const myEvent: ClickEvent = 'clickEvent';

Tuple Types

A tuple can contain values of different data types and works like an array with a fixed number of elements:

1
2
type MyTupleType = [string, number];
const benny: MyTupleType = ['Benny', 34];

Note: A tuple (e.g. [number]) is different to an array (e.g. number[]).

Primitive Data Types

All primitives are passed by value:

  1. bigint
  2. boolean
  3. null
  4. number
  5. string
  6. symbol
  7. undefined

Non-primitive data types / Composite data types

All non-primitive types are passed by reference:

Compiler

Compilers transform high-level programming language code to low-level machine code or some low-level intermediate representation (e.g. Bytecode in Java).

Transpiler

Transpilers transform high-level programming language code (e.g. TypeScript) into another high-level programming language code (e.g. JavaScript).

Discriminated Unions

Discrimination Unions are a form of type guards that narrow down a type based on a shared property.

In the following example, the type of Dog and Person have a shared property called type. Depending on the value of this property, TypeScript can narrow down the type from Dog | Person to either Dog or Person:

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
type Dog = {
  age: number;
  name: string;
  bark: () => void;
  type: 'dog'
}

type Person = {
  age: number;
  name: string;
  shout: () => void;
  type: 'person'
}

function makeNoise(dogOrPerson: Dog | Person): void {
  switch (dogOrPerson.type) {
    case 'dog':
      // Type is narrowed down to "Dog", so we can "bark":
      dogOrPerson.bark();
      break;
    case 'person':
      // Type is narrowed down to "Person", so we can "shout":
      dogOrPerson.shout();
      break;
  }
}

Because this allows the type to be discriminated, this concept is referred to as Discriminated Unions. This concept also exists in F# (Discriminated Unions in F#).