React's Hooks API but for standard web components and lit-html or hyperHTML.
<html lang="en">
<my-counter></my-counter>
<script type="module">
import { html } from 'https://unpkg.com/lit-html/lit-html.js';
import { component, useState } from 'https://unpkg.com/haunted/haunted.js';
function Counter() {
const [count, setCount] = useState(0);
return html`
<div id="count">${count}</div>
<button type="button" @click=${() => setCount(count + 1)}>
Increment
</button>
`;
}
customElements.define('my-counter', component(Counter));
</script>
</html>
A starter app is available on codesandbox and also can be cloned from this repo. This app gives you the basics of how to use Haunted and build components.
npm install haunted
For Internet Explorer 11, you'll need to use a proxy polyfill to use Haunted, in addition to the usual webcomponentsjs polyfills.
<script src="https://cdn.jsdelivr.net/npm/proxy-polyfill@0.3.0/proxy.min.js"></script>
Here is a full example of a web app that uses Haunted built for Internet Explorer 11. You can also use Custom Elements without the Shadow DOM if you need to:
component(MyComponent, { useShadowDOM: false }));
Haunted can be imported just like any other library when using a bundler of your choice:
import {
html,
component,
useState,
useMemo,
useCallback,
useEffect,
useLayoutEffect,
useReducer,
useRef,
useController,
useContext
} from 'haunted';
The main entry point is intended for lit-html users.
If you are using lighterhtml or hyperHTML then instead import haunted/core. This export gives you a function that creates Hooks that work with any template library.
import haunted, { useState } from 'haunted/core';
import { html, render } from 'lighterhtml';
const { component } = haunted({
render(what, where) {
render(where, () => what);
}
});
function App() {
const [count, setCount] = useState(0);
return html`Using lighterhtml! Count: ${count}`;
}
customElements.define('my-app', component(App));
Haunted can work directly in the browser without using any build tools. Simply import the haunted.js bundle. You can use the unpkg or pika CDNs. This works great for demo pages and small apps. Here's an example with unpkg:
import { html } from 'https://unpkg.com/lit-html/lit-html.js';
import { component, useState } from 'https://unpkg.com/haunted/haunted.js';
If using pika then use the html export from Haunted, as pika bundles everything together:
import { html, component, useState } from 'https://cdn.pika.dev/haunted';
If you install Haunted locally this build is located at /node_modules/haunted/haunted.js. And if you're using PikaPkg (@pika/web) then you'll import it from /web_modules/haunted.js.
Haunted is all about writing plain functions that can contain their own state. The documentation below is separated into two sections: creating components (the functions) and using hooks to manage state.
Components are functions that contain state and return HTML via lit-html or hyperHTML. Through the component() and virtual() they become connected to a lifecycle that keeps the HTML up-to-date when state changes.
Using Haunted you can create custom elements or virtual components (components that contain state but have no element tag).
A custom element can be defined via haunted by passing your component you defined (e.g. function App() {}) to haunted's component function. You then take the result of your call to component and pass it to customElements.define like so:
function App({ name }) {
return html`Hello ${name}`;
}
customElements.define('my-app', component(App));
You can now use <my-app></my-app> anywhere you use HTML (directly in a .html file, in JSX, in lit-html templates, wherever).
Another way to write your component is with the arrow function syntax, or fat arrow. It offers a more concise and clean way to write your code, however, it does come with one drawback:
There's no way to use the this keyword to refer to the instance of your web component that is running.
const App = ({ name }) => {
console.log(this); // => 'undefined'
return html`Hello ${name}!`;
};
Now that we have access to <my-app> anywhere we can write HTML, what if you want to render it from JavaScript?
In this instance, we're using lit-html which comes with a render function. You can utilize it by passing in your HTML template and the element you wish to render the template into:
import { render, html } from 'lit-html';
render(
html`<my-app name="world"></my-app>`,
document.body
);
In custom elements, attributes must be pre-defined. To define what attributes your component supports, set the observedAttributes property on the function you defined. Note that attributes use kebab case in templates and are converted into camel case for use in your component's code.
function App({ firstName }) {
return `Hello ${firstName}!`;
}
App.observedAttributes = ['first-name'];
customElements.define('my-app', component(App));
Alternatively, you can pass observedAttributes as an option to component():
component(App, { observedAttributes: ['first-name'] });
Once your custom element is defined you can then pass in attributes as you would with any other HTML element. Just like any other HTML attribute only strings are accepted, anything else will be converted into a string.
<my-app first-name="World"></my-app>
If you haven't used lit-html before you're probably wondering what the differences between properties and attributes are. As stated above, attributes can only have string values, this is because all attributes go through Element#setAttribute. Properties do not go through setAttribute, instead they are properties on the custom element itself. This allows you to pass in any value instead of just strings.
To bind to a property in lit-html, you can use the . prefix before the property name:
function Profile({ userData }) {
return html`
<section>
<h2>Profile</h2>
<article>
<figure>
<img src=${userData.portrait} alt="user portait" />
<figcaption>${userData.name}</figcaption>
</figure>
<p>${userData.bio}</p>
</article>
</section>
`;
}
customElements.define('my-profile', component(Profile));
function App() {
const userData = useFictitiousUser();
return html`
<my-profile .userData=${userData}></my-profile>
`;
}
Notice that it is encouraged to use camel case when writing your template instead of kebab case as these aren't attributes. They're properties so they won't be automatically rewritten in camel case for you.
There are a few steps you have to take in order to dispatch an event from your custom element:
this, the instance of your custom element.this. You can do this by using the fat arrow syntax.new CustomEvent and pass that to this.dispatchEvent.Here are a couple of examples of dispatching events from a haunted custom element:
function Product({ name, price, productId }) {
const buyProduct = () => {
const event = new CustomEvent('buy-product', {
bubbles: true, // this let's the event bubble up through the DOM
composed: true, // this let's the event cross the Shadow DOM boundary
detail: { productId } // all data you wish to pass must be in `detail`
});
this.dispatchEvent(event);
}
return html`
<article>
<h3>${name}</h3>
<p>Price: ${price} USD</p>
<button @click=${buyProduct}>Purchase</button>
</article>
`;
}
Product.observedAttributes = ['name', 'price', 'product-id'];
customElements.define('store-product', component(Product));
With this, you can now listen for the buy-product event either on an instance of <store-product> itself or higher up in the DOM. Here are examples of both of these instances:
/* Example of listening on an element (can be <store-product> or anything above it) */
function Store() {
return html`
<store-product
name="T-Shirt"
price="10.00"
product-id="0001"
@buy-product=${event => {
console.log('We can listen to the event on <store-product> itself 🙂')
}}
></store-product>
`;
}
customElements.define('my-store', component(Store));
/* Example of listening higher up in the DOM on `this` */
function App() {
useEffect(() => {
const handleReportProduct = event => {
console.log('Because the event bubbled all the way up here,');
console.log('we can listen directly on this web component itself! 🎉');
};
this.addEventListener('report-product', handleReportProduct);
return () => {
// very important to remove the event listener!
this.removeEventListener('report-product', handleReportProduct);
}
}, []); // make sure you list all dependencies
return html`<my-store></my-store>`;
}
If you want to look more into firing events, here are some links:
CustomEvent — MDNEvent — MDNdispatchEvent — MDNHaunted also has the concept of virtual components. These are components that are not defined as a tag. Instead they're defined as functions that can be called from within another template. They have their own state and will rerender when that state changes without causing any parent components to rerender.
The following is an example of using virtual components:
const Counter = virtual(() => {
const [count, setCount] = useState(0);
return html`
<button type="button" @click=${() => setCount(count + 1)}>
${count}
</button>
`;
});
function App() {
return html`
<main>
<h1>My app</h1>
${Counter()}
</main>
`;
}
customElements.define('my-app', component(App));
Notice that we have Counter, a virtual component, and App, a custom element. You can use virtual components within custom elements and custom elements within virtual components.
The only difference is that custom elements are utilized via their tag name (e.g. <my-app>) and virtual components are called as functions (e.g. ${Counter()}).
If you wanted you could create an entire app of virtual components.
thisYou'll notice that in the above examples, we're using the fat arrow syntax. This is due to virtual components not being attached to any actual custom elements, therefore, this never points to anything which negates the purpose of using the function syntax.
Haunted supports the same API as React Hooks. The hope is that by doing so you can reuse hooks available on npm simply by aliasing package names in your bundler's config.
Currently Haunted supports the following hooks:
Returns an array of two values. The first value is the immutable state that is initialized to the argument you pass in. The second value is the function that is used to change the first value (the setter).
When the setter is called, the element in which the state resides is rerendered.
const [count, setCount] = useState(0);
Additionally if you provide a function as the argument to useState, the function is called to initialize the first state, but never called again.
const [count, setCount] = useState(() => {
return expensiveComputation(42);
});
Used to run a side-effect when the component rerenders or when a dependency changes. To run your side-effect only when the component rerenders, only pass in your side-effect function and nothing else:
function Counter() {
const [count, setCount] = useState(0);
useEffect(() => {
// on every render, a new number will be in the title
document.title = `A random number ${Math.random()}`;
});
return html`
<div id="count">${count}</div>
<button type="button" @click=${() => setCount(count + 1)}>
Cause rerender
</button>
`;
}
What happens when your code begins to rely on dependencies that can change (state, refs, props)? We no longer want it to rerun on every rerender so we need to ensure that the code we're running, or its result, doesn't become stale. To do this, you should always state all of the dependencies you're using in an array as the second argument to useEffect:
const [name, setName] = useState('Dracula');
useEffect(() => {
// This only occurs when `name` changes and on the initial render.
document.title = `Hello ${name}`;
}, [name]);
A dependency is anything that your side-effect relies on that can change between renders (e.g. state, refs, props). This does not include setName or other setters from useState because they will never change between renders.
Generally, you won't want to do this but rather make sure you're actually listing all of your dependencies in the dependency array. If you don't have any dependencies then your code will run on mount and clean up on unmount.
Here is an example of only running an effect once as opposed to every rerender:
useEffect(() => {
document.title = 'I will stay like this until someone changes me';
}, []); // note that you must pass the empty array
Since effects are used for side-effectual things and might run many times in the lifecycle of a component, useEffect supports returning a teardown function.
An example of when you might use this is if you are setting up an event listener:
const [name, setName] = useState('Wolf Man');
useEffect(() => {
function updateNameFromWorker(event) {
setName(event.data);
}
worker.addEventListener('message', updateNameFromWorker);
return () => {
worker.removeEventListener('message', updateNameFromWorker);
};
}, []); // note that it is safe to exclude `setName` from the dependencies because it will never change
The function signature is the same as useEffect, but the callback is being called synchronously after rendering. Therefore, updates scheduled inside useLayoutEffect will be flushed synchronously before the browser has a chance to paint.
Most of the time, it is preferable to use useEffect to avoid blocking visual updates.
Similarly to useState, useReducer will return an array of two values, the first one being the state. The second one, however, is dispatch. This is a function that takes an action. The action is then passed to your reducer (the first argument) and your reducer will determine the new state and return it.
The following is an example of useReducer being used to handle incrementing and decrementing a count:
// doesn't have to be an object, could just be `initialState = 0`
const initialState = { count: 0 };
function reducer(state, action) {
switch (action.type) {
case 'reset':
return initialState;
case 'increment':
return { count: state.count + 1 };
case 'decrement':
return { count: state.count - 1 };
}
}
function Counter() {
const [state, dispatch] = useReducer(reducer, initialState);
return html`
Count: ${state.count}
<button @click=${() => dispatch({ type: 'reset' })}>
Reset
</button>
<button @click=${() => dispatch({ type: 'increment' })}>+</button>
<button @click=${() => dispatch({ type: 'decrement' })}>-</button>
`;
}
This might look familiar as redux uses a similar concept.
Create a memoized state value. Only reruns the function when dependent values have changed.
function fibonacci(num) {
if (num <= 1) return 1;
return fibonacci(num - 1) + fibonacci(num - 2);
}
function App() {
const [value, setVal] = useState(12);
const fib = useMemo(() => fibonacci(value), [value]);
return html`
<h1>Fibonacci</h1>
<input
type="text"
@change=${event => setVal(parseInt(event.target.value, 10))}
value=${value}
/>
<div>Fibonacci <strong>${fib}</strong></div>
`;
}
Very similar to useMemo but instead of writing a function that returns your memoized value, your function is the memoized value. This and useMemo are often overused so try to only use this when your callback has dependencies and it is itself a dependency for something else (like useEffect) .
const submit = useCallback(event => {
event.preventDefault();
api.post('submit', { email, password });
}, [email, password]);
useEffect(() => {
this.addEventListener('submit', submit);
return () => this.removeEventListener('submit', submit);
}, [submit]);
Creates and returns a mutable object (a 'ref') whose .current property is initialized to the passed argument.
This differs from useState in that state is immutable and can only be changed via setState which will cause a rerender. That rerender will allow you to be able to see the updated state value. A ref, on the other hand, can only be changed via .current and since changes to it are mutations, no rerender is required to view the updated value in your component's code (e.g. listeners, callbacks, effects).
function App() {
const myRef = useRef(0);
return html`${myRef.current}`;
}
Grabs the context value from the closest provider above and updates your component, the consumer, whenever the provider changes the value.
useContext currently only works with custom element components, track the issue here.
const ThemeContext = createContext('dark');
customElements.define('theme-provider', ThemeContext.Provider);
customElements.define('theme-consumer', ThemeContext.Consumer);
function Consumer() {
const context = useContext(ThemeContext);
return context;
}
customElements.define('my-consumer', component(Consumer));
function App() {
const [theme, setTheme] = useState("light");
return html`
<select value=${theme} @change=${event => setTheme(event.target.value)}>
<option value="dark">Dark</option>
<option value="light">Light</option>
</select>
<theme-provider .value=${theme}>
<my-consumer></my-consumer>
<!-- creates context with inverted theme -->
<theme-provider .value=${theme === 'dark' ? 'light' : 'dark'}>
<theme-consumer
.render=${value =>
html`<h1>${value}</h1>`
}
></theme-consumer>
</theme-provider>
</theme-provider>
`;
}
Reuse existing ReactiveControllers in your haunted component without adding @lit/reactive-element to your bundle.
import { component, html, useController } from 'haunted'
// See example at https://lit.dev/docs/composition/controllers/#external-inputs
import { MouseController } from './mouse-controller';
const useMouseController = useController(host => new MouseController(host));
function TrackMouse() {
const { pos } = useMouseController();
return html`
<dl>
<dt>x</dt> <dd>${pos.x}</dd>
<dt>y</dt> <dd>${pos.y}</dd>
</dl>
`;
}
Most functionality can be achieved with the provided hooks above, but you can also create your own hooks for custom functionality like so:
import { hook, Hook } from 'haunted';
const useMyHook = hook(class extends Hook {
// everything after `state` is an argument passed to your hook
constructor(id, state, your, hooks, args) {
super(id, state);
// you can trigger an update by calling `this.state.update()`
}
update(your, hooks, args) {
return ['the value returned from', your, 'hook'];
}
teardown() { /* called when the hook is to be torn down */ }
});
At its heart, Haunted is a container for state derived from hooks. The component and virtual signatures build on top of this state container.
In order to use Haunted outside of its component types, such as to extend another custom element base class, you can use the State constructor.
It has a signature of: new State(update, [ hostElement ]).
Note that the second argument
hostElementis optional. If you want to use theuseContexthook you will need to provide a host element, however.
Here's an example how it can be used to run hooks code:
import { State, useState } from 'haunted';
let state = new State(() => {
update();
});
function update() {
state.run(() => {
const [count, setCount] = useState(0);
console.log('count is', count);
setTimeout(() => setCount(count + 1), 3000);
});
}
update();
The above will result in the count being incremented every 3 seconds and the current count being logged.
A more practical example is integration with a custom element base class. Here's a simple integration with LitElement:
import { LitElement } from 'lit-element';
import { State } from 'haunted';
export default class LitHauntedElement extends LitElement {
constructor() {
super();
this.hauntedState = new State(() => this.requestUpdate(), this);
}
update(changedProperties) {
this.hauntedState.run(() => super.update(changedProperties));
this.hauntedState.runEffects();
}
disconnectedCallback() {
super.disconnectedCallback();
this.hauntedState.teardown();
}
}
More example integrations can be found in this gist.
component(renderer, options): Element
component(renderer, BaseElement, options): Element
(element) => html`...`HTMLElement{baseElement: HTMLElement, observedAttributes: [], useShadowDOM: true}virtual(renderer): directive
(element) => html`...`BSD-2-Clause
FAQs
Hooks for web components
The npm package haunted receives a total of 0 weekly downloads. As such, haunted popularity was classified as not popular.
We found that haunted demonstrated a not healthy version release cadence and project activity because the last version was released a year ago. It has 2 open source maintainers collaborating on the project.
Did you know?
Socket for GitHub automatically highlights issues in each pull request and monitors the health of all your open source dependencies. Discover the contents of your packages and block harmful activity before you install or update your dependencies.
Security News
GitHub removed 27 malicious pull requests attempting to inject harmful code across multiple open source repositories, in another round of low-effort attacks.
Security News
RubyGems.org has added a new "maintainer" role that allows for publishing new versions of gems. This new permission type is aimed at improving security for gem owners and the service overall.
Security News
Node.js will be enforcing stricter semver-major PR policies a month before major releases to enhance stability and ensure reliable release candidates.