A cheatsheet containing ES2015 [ES6] tips, tricks, best practices and code snippet examples for your day to day workflow. Contributions are welcome!
- var versus let / const
- Replacing IIFEs with Blocks
- Arrow Functions
- Strings
- Destructuring
- Modules
- Parameters
- Classes
- Symbols
- Maps
- WeakMaps
- Promises
- Generators
- Async Await
- Getter/Setter functions
- Sample Code
Besides
var
, we now have access to two new identifiers for storing values —let
andconst
. Unlikevar
,let
andconst
statements are not hoisted to the top of their enclosing scope.
An example of using var
:
var snack = 'Meow Mix';
function getFood(food) {
if (food) {
var snack = 'Friskies';
return snack;
}
return snack;
}
getFood(false); // undefined
However, observe what happens when we replace var
using let
:
let snack = 'Meow Mix';
function getFood(food) {
if (food) {
let snack = 'Friskies';
return snack;
}
return snack;
}
getFood(false); // 'Meow Mix'
This change in behavior highlights that we need to be careful when refactoring
legacy code which uses var
. Blindly replacing instances of var
with let
may lead to unexpected behavior.
Note:
let
andconst
are block scoped. Therefore, referencing block-scoped identifiers before they are defined will produce aReferenceError
.
console.log(x); // ReferenceError: x is not defined
let x = 'hi';
Best Practice: Leave
var
declarations inside of legacy code to denote that it needs to be carefully refactored. When working on a new codebase, uselet
for variables that will change their value over time, andconst
for variables which cannot be reassigned.
A common use of Immediately Invoked Function Expressions is to enclose values within its scope. In ES6, we now have the ability to create block-based scopes and therefore are not limited purely to function-based scope.
(function () {
var food = 'Meow Mix';
}());
console.log(food); // Reference Error
Using ES6 Blocks:
{
let food = 'Meow Mix';
}
console.log(food); // Reference Error
Often times we have nested functions in which we would like to preserve the
context of this
from its lexical scope. An example is shown below:
function Person(name) {
this.name = name;
}
Person.prototype.prefixName = function (arr) {
return arr.map(function (character) {
return this.name + character; // Cannot read property 'name' of undefined
});
};
One common solution to this problem is to store the context of this
using
a variable:
function Person(name) {
this.name = name;
}
Person.prototype.prefixName = function (arr) {
var that = this; // Store the context of this
return arr.map(function (character) {
return that.name + character;
});
};
We can also pass in the proper context of this
:
function Person(name) {
this.name = name;
}
Person.prototype.prefixName = function (arr) {
return arr.map(function (character) {
return this.name + character;
}, this);
};
As well as bind the context:
function Person(name) {
this.name = name;
}
Person.prototype.prefixName = function (arr) {
return arr.map(function (character) {
return this.name + character;
}.bind(this));
};
Using Arrow Functions, the lexical value of this
isn't shadowed and we
can re-write the above as shown:
function Person(name) {
this.name = name;
}
Person.prototype.prefixName = function (arr) {
return arr.map(character => this.name + character);
};
Best Practice: Use Arrow Functions whenever you need to preserve the lexical value of
this
.
Arrow Functions are also more concise when used in function expressions which simply return a value:
var squares = arr.map(function (x) { return x * x }); // Function Expression
const arr = [1, 2, 3, 4, 5];
const squares = arr.map(x => x * x); // Arrow Function for terser implementation
Best Practice: Use Arrow Functions in place of function expressions when possible.
With ES6, the standard library has grown immensely. Along with these changes
are new methods which can be used on strings, such as .includes()
and
.repeat()
.
var string = 'food';
var substring = 'foo';
console.log(string.indexOf(substring) > -1);
Instead of checking for a return value > -1
to denote string containment,
we can simply use .includes()
which will return a boolean:
const string = 'food';
const substring = 'foo';
console.log(string.includes(substring)); // true
function repeat(string, count) {
var strings = [];
while(strings.length < count) {
strings.push(string);
}
return strings.join('');
}
In ES6, we now have access to a terser implementation:
// String.repeat(numberOfRepetitions)
'meow'.repeat(3); // 'meowmeowmeow'
Using Template Literals, we can now construct strings that have special characters in them without needing to escape them explicitly.
var text = "This string contains \"double quotes\" which are escaped.";
let text = `This string contains "double quotes" which don't need to be escaped anymore.`;
Template Literals also support interpolation, which makes the task of concatenating strings and values:
var name = 'Tiger';
var age = 13;
console.log('My cat is named ' + name + ' and is ' + age + ' years old.');
Much simpler:
const name = 'Tiger';
const age = 13;
console.log(`My cat is named ${name} and is ${age} years old.`);
In ES5, we handled new lines as follows:
var text = (
'cat\n' +
'dog\n' +
'nickelodeon'
);
Or:
var text = [
'cat',
'dog',
'nickelodeon'
].join('\n');
Template Literals will preserve new lines for us without having to explicitly place them in:
let text = ( `cat
dog
nickelodeon`
);
Template Literals can accept expressions, as well:
let today = new Date();
let text = `The time and date is ${today.toLocaleString()}`;
Destructuring allows us to extract values from arrays and objects (even deeply nested) and store them in variables with a more convenient syntax.
var arr = [1, 2, 3, 4];
var a = arr[0];
var b = arr[1];
var c = arr[2];
var d = arr[3];
let [a, b, c, d] = [1, 2, 3, 4];
console.log(a); // 1
console.log(b); // 2
var luke = { occupation: 'jedi', father: 'anakin' };
var occupation = luke.occupation; // 'jedi'
var father = luke.father; // 'anakin'
let luke = { occupation: 'jedi', father: 'anakin' };
let {occupation, father} = luke;
console.log(occupation); // 'jedi'
console.log(father); // 'anakin'
Prior to ES6, we used libraries such as Browserify to create modules on the client-side, and require in Node.js. With ES6, we can now directly use modules of all types (AMD and CommonJS).
module.exports = 1;
module.exports = { foo: 'bar' };
module.exports = ['foo', 'bar'];
module.exports = function bar () {};
With ES6, we have various flavors of exporting. We can perform Named Exports:
export let name = 'David';
export let age = 25;
As well as exporting a list of objects:
function sumTwo(a, b) {
return a + b;
}
function sumThree(a, b, c) {
return a + b + c;
}
export { sumTwo, sumThree };
We can also export functions, objects and values (etc.) simply by using the export
keyword:
export function sumTwo(a, b) {
return a + b;
}
export function sumThree(a, b, c) {
return a + b + c;
}
And lastly, we can export default bindings:
function sumTwo(a, b) {
return a + b;
}
function sumThree(a, b, c) {
return a + b + c;
}
let api = {
sumTwo,
sumThree
};
export default api;
/* Which is the same as
* export { api as default };
*/
Best Practices: Always use the
export default
method at the end of the module. It makes it clear what is being exported, and saves time by having to figure out what name a value was exported as. More so, the common practice in CommonJS modules is to export a single value or object. By sticking to this paradigm, we make our code easily readable and allow ourselves to interpolate between CommonJS and ES6 modules.
ES6 provides us with various flavors of importing. We can import an entire file:
import 'underscore';
It is important to note that simply importing an entire file will execute all code at the top level of that file.
Similar to Python, we have named imports:
import { sumTwo, sumThree } from 'math/addition';
We can also rename the named imports:
import {
sumTwo as addTwoNumbers,
sumThree as sumThreeNumbers
} from 'math/addition';
In addition, we can import all the things (also called namespace import):
import * as util from 'math/addition';
Lastly, we can import a list of values from a module:
import * as additionUtil from 'math/addition';
const { sumTwo, sumThree } = additionUtil;
Importing from the default binding like this:
import api from 'math/addition';
// Same as: import { default as api } from 'math/addition';
While it is better to keep the exports simple, but we can sometimes mix default import and mixed import if needed. When we are exporting like this:
// foos.js
export { foo as default, foo1, foo2 };
We can import them like the following:
import foo, { foo1, foo2 } from 'foos';
When importing a module exported using commonjs syntax (such as React) we can do:
import React from 'react';
const { Component, PropTypes } = React;
This can also be simplified further, using:
import React, { Component, PropTypes } from 'react';
Note: Values that are exported are bindings, not references. Therefore, changing the binding of a variable in one module will affect the value within the exported module. Avoid changing the public interface of these exported values.
In ES5, we had varying ways to handle functions which needed default values, indefinite arguments, and named parameters. With ES6, we can accomplish all of this and more using more concise syntax.
function addTwoNumbers(x, y) {
x = x || 0;
y = y || 0;
return x + y;
}
In ES6, we can simply supply default values for parameters in a function:
function addTwoNumbers(x=0, y=0) {
return x + y;
}
addTwoNumbers(2, 4); // 6
addTwoNumbers(2); // 2
addTwoNumbers(); // 0
In ES5, we handled an indefinite number of arguments like so:
function logArguments() {
for (var i=0; i < arguments.length; i++) {
console.log(arguments[i]);
}
}
Using the rest operator, we can pass in an indefinite amount of arguments:
function logArguments(...args) {
for (let arg of args) {
console.log(arg);
}
}
One of the patterns in ES5 to handle named parameters was to use the options object pattern, adopted from jQuery.
function initializeCanvas(options) {
var height = options.height || 600;
var width = options.width || 400;
var lineStroke = options.lineStroke || 'black';
}
We can achieve the same functionality using destructuring as a formal parameter to a function:
function initializeCanvas(
{ height=600, width=400, lineStroke='black'}) {
// Use variables height, width, lineStroke here
}
If we want to make the entire value optional, we can do so by destructuring an empty object:
function initializeCanvas(
{ height=600, width=400, lineStroke='black'} = {}) {
// ...
}
In ES5, we could find the max of values in an array by using the apply
method on Math.max
like this:
Math.max.apply(null, [-1, 100, 9001, -32]); // 9001
In ES6, we can now use the spread operator to pass an array of values to be used as parameters to a function:
Math.max(...[-1, 100, 9001, -32]); // 9001
We can concat array literals easily with this intuitive syntax:
let cities = ['San Francisco', 'Los Angeles'];
let places = ['Miami', ...cities, 'Chicago']; // ['Miami', 'San Francisco', 'Los Angeles', 'Chicago']
Prior to ES6, we implemented Classes by creating a constructor function and adding properties by extending the prototype:
function Person(name, age, gender) {
this.name = name;
this.age = age;
this.gender = gender;
}
Person.prototype.incrementAge = function () {
return this.age += 1;
};
And created extended classes by the following:
function Personal(name, age, gender, occupation, hobby) {
Person.call(this, name, age, gender);
this.occupation = occupation;
this.hobby = hobby;
}
Personal.prototype = Object.create(Person.prototype);
Personal.prototype.constructor = Personal;
Personal.prototype.incrementAge = function () {
Person.prototype.incrementAge.call(this);
this.age += 20;
console.log(this.age);
};
ES6 provides much needed syntactic sugar for doing this under the hood. We can create Classes directly:
class Person {
constructor(name, age, gender) {
this.name = name;
this.age = age;
this.gender = gender;
}
incrementAge() {
this.age += 1;
}
}
And extend them using the extends
keyword:
class Personal extends Person {
constructor(name, age, gender, occupation, hobby) {
super(name, age, gender);
this.occupation = occupation;
this.hobby = hobby;
}
incrementAge() {
super.incrementAge();
this.age += 20;
console.log(this.age);
}
}
Best Practice: While the syntax for creating classes in ES6 obscures how implementation and prototypes work under the hood, it is a good feature for beginners and allows us to write cleaner code.
Symbols have existed prior to ES6, but now we have a public interface to using them directly. Symbols are immutable and unique and can be used as keys in any hash.
Calling Symbol()
or Symbol(description)
will create a unique symbol that cannot be looked up
globally. A Use case for Symbol()
is to patch objects or namespaces from third parties with your own
logic, but be confident that you won't collide with updates to that library. For example,
if you wanted to add a method refreshComponent
to the React.Component
class, and be certain that
you didn't trample a method they add in a later update:
const refreshComponent = Symbol();
React.Component.prototype[refreshComponent] = () => {
// do something
}
Symbol.for(key)
will create a Symbol that is still immutable and unique, but can be looked up globally.
Two identical calls to Symbol.for(key)
will return the same Symbol instance. NOTE: This is not true for
Symbol(description)
:
Symbol('foo') === Symbol('foo') // false
Symbol.for('foo') === Symbol('foo') // false
Symbol.for('foo') === Symbol.for('foo') // true
A common use case for Symbols, and in particular with Symbol.for(key)
is for interoperability. This can be
achieved by having your code look for a Symbol member on object arguments from third parties that contain some
known interface. For example:
function reader(obj) {
const specialRead = Symbol.for('specialRead');
if (obj[specialRead]) {
const reader = obj[specialRead]();
// do something with reader
} else {
throw new TypeError('object cannot be read');
}
}
And then in another library:
const specialRead = Symbol.for('specialRead');
class SomeReadableType {
[specialRead]() {
const reader = createSomeReaderFrom(this);
return reader;
}
}
A notable example of Symbol use for interoperability is
Symbol.iterator
which exists on all iterable types in ES6: Arrays, strings, generators, etc. When called as a method it returns an object with an Iterator interface.
Maps is a much needed data structure in JavaScript. Prior to ES6, we created hash maps through objects:
var map = new Object();
map[key1] = 'value1';
map[key2] = 'value2';
However, this does not protect us from accidentally overriding functions with specific property names:
> getOwnProperty({ hasOwnProperty: 'Hah, overwritten'}, 'Pwned');
> TypeError: Property 'hasOwnProperty' is not a function
Actual Maps allow us to set
, get
and search
for values (and much more).
let map = new Map();
> map.set('name', 'david');
> map.get('name'); // david
> map.has('name'); // true
The most amazing part of Maps is that we are no longer limited to just using strings. We can now use any type as a key, and it will not be type-cast to a string.
let map = new Map([
['name', 'david'],
[true, 'false'],
[1, 'one'],
[{}, 'object'],
[function () {}, 'function']
]);
for (let key of map.keys()) {
console.log(typeof key);
// > string, boolean, number, object, function
}
Note: Using non-primitive values such as functions or objects won't work when testing equality using methods such as
map.get()
. As such, stick to primitive values such as Strings, Booleans and Numbers.
We can also iterate over maps using .entries()
:
for (let [key, value] of map.entries()) {
console.log(key, value);
}
In order to store private data versions < ES6, we had various ways of doing this. One such method was using naming conventions:
class Person {
constructor(age) {
this._age = age;
}
_incrementAge() {
this._age += 1;
}
}
But naming conventions can cause confusion in a codebase and are not always going to be upheld. Instead, we can use WeakMaps to store our values:
let _age = new WeakMap();
class Person {
constructor(age) {
_age.set(this, age);
}
incrementAge() {
let age = _age.get(this) + 1;
_age.set(this, age);
if (age > 50) {
console.log('Midlife crisis');
}
}
}
The cool thing about using WeakMaps to store our private data is that their
keys do not give away the property names, which can be seen by using
Reflect.ownKeys()
:
> const person = new Person(50);
> person.incrementAge(); // 'Midlife crisis'
> Reflect.ownKeys(person); // []
A more practical example of using WeakMaps is to store data which is associated to a DOM element without having to pollute the DOM itself:
let map = new WeakMap();
let el = document.getElementById('someElement');
// Store a weak reference to the element with a key
map.set(el, 'reference');
// Access the value of the element
let value = map.get(el); // 'reference'
// Remove the reference
el.parentNode.removeChild(el);
el = null;
// map is empty, since the element is destroyed
As shown above, once the object is destroyed by the garbage collector, the WeakMap will automatically remove the key-value pair which was identified by that object.
Note: To further illustrate the usefulness of this example, consider how jQuery stores a cache of objects corresponding to DOM elements which have references. Using WeakMaps, jQuery can automatically free up any memory that was associated with a particular DOM element once it has been removed from the document. In general, WeakMaps are very useful for any library that wraps DOM elements.
Promises allow us to turn our horizontal code (callback hell):
func1(function (value1) {
func2(value1, function (value2) {
func3(value2, function (value3) {
func4(value3, function (value4) {
func5(value4, function (value5) {
// Do something with value 5
});
});
});
});
});
Into vertical code:
func1(value1)
.then(func2)
.then(func3)
.then(func4)
.then(func5, value5 => {
// Do something with value 5
});
Prior to ES6, we used bluebird or Q. Now we have Promises natively:
new Promise((resolve, reject) =>
reject(new Error('Failed to fulfill Promise')))
.catch(reason => console.log(reason));
Where we have two handlers, resolve (a function called when the Promise is fulfilled) and reject (a function called when the Promise is rejected).
Benefits of Promises: Error Handling using a bunch of nested callbacks can get chaotic. Using Promises, we have a clear path to bubbling errors up and handling them appropriately. Moreover, the value of a Promise after it has been resolved/rejected is immutable - it will never change.
Here is a practical example of using Promises:
var request = require('request');
return new Promise((resolve, reject) => {
request.get(url, (error, response, body) => {
if (body) {
resolve(JSON.parse(body));
} else {
resolve({});
}
});
});
We can also parallelize Promises to handle an array of asynchronous
operations by using Promise.all()
:
let urls = [
'/api/commits',
'/api/issues/opened',
'/api/issues/assigned',
'/api/issues/completed',
'/api/issues/comments',
'/api/pullrequests'
];
let promises = urls.map((url) => {
return new Promise((resolve, reject) => {
$.ajax({ url: url })
.done((data) => {
resolve(data);
});
});
});
Promise.all(promises)
.then((results) => {
// Do something with results of all our promises
});
Similar to how Promises allow us to avoid callback hell, Generators allow us to flatten our code - giving our asynchronous code a synchronous feel. Generators are essentially functions which we can pause their execution and subsequently return the value of an expression.
A simple example of using generators is shown below:
function* sillyGenerator() {
yield 1;
yield 2;
yield 3;
yield 4;
}
var generator = sillyGenerator();
> console.log(generator.next()); // { value: 1, done: false }
> console.log(generator.next()); // { value: 2, done: false }
> console.log(generator.next()); // { value: 3, done: false }
> console.log(generator.next()); // { value: 4, done: false }
Where next will allow us to push our generator forward and evaluate a new expression. While the above example is extremely contrived, we can utilize Generators to write asynchronous code in a synchronous manner:
// Hiding asynchronousity with Generators
function request(url) {
getJSON(url, function(response) {
generator.next(response);
});
}
And here we write a generator function that will return our data:
function* getData() {
var entry1 = yield request('http://some_api/item1');
var data1 = JSON.parse(entry1);
var entry2 = yield request('http://some_api/item2');
var data2 = JSON.parse(entry2);
}
By the power of yield
, we are guaranteed that entry1
will have the data needed to be parsed and stored
in data1
.
While generators allow us to write asynchronous code in a synchronous manner, there is no clear and easy path for error propagation. As such, as we can augment our generator with Promises:
function request(url) {
return new Promise((resolve, reject) => {
getJSON(url, resolve);
});
}
And we write a function which will step through our generator using next
which in turn will utilize our
request
method above to yield a Promise:
function iterateGenerator(gen) {
var generator = gen();
(function iterate(val) {
var ret = generator.next();
if(!ret.done) {
ret.value.then(iterate);
}
})();
}
By augmenting our Generator with Promises, we have a clear way of propagating errors through the use of our
Promise .catch
and reject
. To use our newly augmented Generator, it is as simple as before:
iterateGenerator(function* getData() {
var entry1 = yield request('http://some_api/item1');
var data1 = JSON.parse(entry1);
var entry2 = yield request('http://some_api/item2');
var data2 = JSON.parse(entry2);
});
We were able to reuse our implementation to use our Generator as before, which shows their power. While Generators and Promises allow us to write asynchronous code in a synchronous manner while retaining the ability to propagate errors in a nice way, we can actually begin to utilize a simpler construction that provides the same benefits: async-await.
While this is actually an upcoming ES2016 feature, async await
allows us to perform the same thing we accomplished
using Generators and Promises with less effort:
var request = require('request');
function getJSON(url) {
return new Promise(function(resolve, reject) {
request(url, function(error, response, body) {
resolve(body);
});
});
}
async function main() {
var data = await getJSON();
console.log(data); // NOT undefined!
}
main();
Under the hood, it performs similarly to Generators. I highly recommend using them over Generators + Promises. A great resource for getting up and running with ES7 and Babel can be found here.
ES6 has started supporting getter and setter functions. Using the following example:
class Employee {
constructor(name) {
this._name = name;
}
get name() {
if(this._name) {
return 'Mr. ' + this._name.toUpperCase();
} else {
return undefined;
}
}
set name(newName) {
if (newName == this._name) {
console.log('I already have this name.');
} else if (newName) {
this._name = newName;
} else {
return false;
}
}
}
var emp = new Employee("James Bond");
// uses the get method in the background
if (emp.name) {
console.log(emp.name); // Mr. JAMES BOND
}
// uses the setter in the background
emp.name = "Bond 007";
console.log(emp.name); // Mr. BOND 007
Latest browsers are also supporting getter/setter functions in Objects and we can use them for computed properties, adding listeners and preprocessing before setting/getting:
var person = {
firstName: 'James',
lastName: 'Bond',
get fullName() {
console.log('Getting FullName');
return this.firstName + ' ' + this.lastName;
},
set fullName (name) {
console.log('Setting FullName');
var words = name.toString().split(' ');
this.firstName = words[0] || '';
this.lastName = words[1] || '';
}
}
person.fullName; // James Bond
person.fullName = 'Bond 007';
person.fullName; // Bond 007
install babel modules
npm install --save-dev babel-cli babel-preset-es2015
create .babelrc config
{
"presets": ["es2015"]
}
update package.json as below
"scripts": {
"build": "babel src -d build"
},
"dependencies": {
"babel-cli": "~6.10.1",
"babel-preset-es2015": "~6.9.0"
}
npm run build
node build/sample.js
./node_modules/.bin/babel-node src/sample.js