Skip to content
/ voby Public

A high-performance framework with fine-grained observable-based reactivity for building rich applications.

License

Notifications You must be signed in to change notification settings

vobyjs/voby

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Voby's Banner

Join The Discord Chat Open The Playground Donate With Open Collective

A high-performance framework with fine-grained observable/signal-based reactivity for building rich applications.

Features

This works similarly to Solid, but without a custom Babel transform and with a different API.

  • No VDOM: there's no VDOM overhead, the framework deals with raw DOM nodes directly.
  • No stale closures: functions are always executed afresh, no need to worry about previous potential executions of the current function, ever.
  • No rules of hooks: hooks are just regular functions, which you can nest indefinitely, call conditionally, use outside components, whatever you want.
  • No dependencies arrays: the framework is able to detect what depends on what else automatically, no need to specify dependencies manually.
  • No props diffing: updates are fine grained, there's no props diffing, whenever an attribute/property/class/handler/etc. should be updated it's updated directly and immediately.
  • No key prop: you can just map over arrays, or use the For component with an array of unique values, no need to specify keys explicitly.
  • No Babel: there's no need to use Babel with this framework, it works with plain old JS (plus JSX if you are into that). As a consequence we have 0 transform function bugs, because we don't have a transform function.
  • No magic: what you see is what you get, your code is not transformed to actually do something different than what you write, there are no surprises.
  • No server support: for the time being this framework is focused on local-first rich applications, most server-related features are not implemented: no hydration, no server components, no streaming etc.
  • Observable-based: observables, also known as "signals", are at the core of our reactivity system. The way it works is very different from a React-like system, it may be more challenging to learn, but it's well worth the effort.
  • Work in progress: this is probably beta software, I'm working on it because I need something with great performance for Notable, I'm allergic to third-party dependencies, I'd like something with an API that resonates with me, and I wanted to deeply understand how the more solid Solid, which you should also check out, works.

Demos

You can find some demos and benchmarks below, more demos are contained inside the repository.

APIs

Methods Components Hooks core Hooks web Types Extras
$ Dynamic useBoolean useAbortController Context Contributing
$$ ErrorBoundary useCleanup useAbortSignal Directive Globals
batch For useContext useAnimationFrame DirectiveOptions JSX
createContext Fragment useDisposed useAnimationLoop EffectOptions Tree Shaking
createDirective If useEffect useEventListener FunctionMaybe TypeScript
createElement KeepAlive useMemo useFetch MemoOptions
h Portal usePromise useIdleCallback Observable
hmr Suspense useReadonly useIdleLoop ObservableLike
html Switch useResolved useInterval ObservableReadonly
isBatching Ternary useResource useMicrotask ObservableReadonlyLike
isObservable useRoot useTimeout ObservableMaybe
isServer useSelector ObservableOptions
isStore useSuspended Resource
lazy useUntracked StoreOptions
render
renderToString
resolve
store
template
tick
untrack

Usage

This framework is simply a view layer built on top of the Observable library oby, knowing how that works is necessary to understand how this works.

This framework basically re-exports everything that oby exports, sometimes with a slightly different interface, adjusted for usage as components or hooks, plus some additional functions.

Methods

The following top-level functions are provided.

$

This function is just the default export of oby, it can be used to wrap a value in an observable.

No additional methods are attached to this function. Everything that oby attaches to it is instead exported as components and hooks.

Read upstream documentation.

Interface:

function $ <T> (): Observable<T | undefined>;
function $ <T> ( value: undefined, options?: ObservableOptions<T | undefined> ): Observable<T | undefined>;
function $ <T> ( value: T, options?: ObservableOptions<T> ): Observable<T>;

Usage:

import {$} from 'voby';

// Create an observable without an initial value

$<number> ();

// Create an observable with an initial value

$(1);

// Create an observable with an initial value and a custom equality function

const equals = ( value, valuePrev ) => Object.is ( value, valuePrev );

const o = $( 1, { equals } );

// Create an observable with an initial value and a special "false" equality function, which is a shorthand for `() => false`, which causes the observable to always emit when its setter is called

const oFalse = $( 1, { equals: false } );

// Getter

o (); // => 1

// Setter

o ( 2 ); // => 2

// Setter via a function, which gets called with the current value

o ( value => value + 1 ); // => 3

// Setter that sets a function, it has to be wrapped in another function because the above form exists

const noop = () => {};

o ( () => noop );

$$

This function unwraps a potentially observable value.

Read upstream documentation.

Interface:

function $$ <T> ( value: T ): (T extends ObservableReadonly<infer U> ? U : T);

Usage:

import {$$} from 'voby';

// Getting the value out of an observable

const o = $(123);

$$ ( o ); // => 123

// Getting the value out of a function

$$ ( () => 123 ); // => 123

// Getting the value out of an observable but not out of a function

$$ ( o, false ); // => 123
$$ ( () => 123, false ); // => () => 123

// Getting the value out of a non-observable and non-function

$$ ( 123 ); // => 123

batch

This function prevents effects from firing until the function passed to it resolves. It's largely only useful when the passed function is asynchronous, as otherwise the reactivity system is lazy so effects won't be over-executed anyway.

Read upstream documentation.

Interface:

function batch <T> ( fn: () => Promise<T> | T ): Promise<Awaited<T>>;
function batch <T> ( value: T ): Promise<Awaited<T>>;

Usage:

import {batch} from 'voby';

batch // => Same as require ( 'oby' ).batch

createContext

This function creates a context object, optionally with a default value, which can later be used to provide a new value for the context or to read the current value.

A context's Provider will register the value of context with its children.

Interface:

type ContextProvider<T> = ( props: { value: T, children: JSX.Element } ) => JSX.Element;
type Context<T> = { Provider: ContextProvider<T> };

function createContext <T> ( defaultValue?: T ): Context<T>;

Usage:

import {createContext, useContext} from 'voby';

const App = () => {
  const Context = createContext ( 123 );
  return (
    <>
      {() => {
        const value = useContext ( Context );
        return <p>{value}</p>;
      }}
      <Context.Provider value={312}>
        {() => {
          const value = useContext ( Context );
          return <p>{value}</p>;
        }}
      </Context.Provider>
    </>
  );
};

createDirective

This function creates a directive provider, which can be used to register a directive with its children.

A directive is a function that always receives an Element as its first argument, which is basically a ref to the target element, and arbitrary user-provided arguments after that.

Each directive has a unique name and it can be called by simply writing use:directivename={[arg1, arg2, ...argN]]} in the JSX.

Directives internally are registered using context providers, so you can also override directives for a particular scope just by registering another directive with the same name closer to where you are reading it.

A directive's Provider will register the directive with its children, which is always what you want, but it can lead to messy code due to nesting.

A directive's register function will register the directive with the current parent observer, which is usually only safe to do at the root level, but it will lead to very readable code.

Interface:

type DirectiveFunction = <T extends unknown[]> ( ref: Element, ...args: T ) => void;
type DirectiveProvider = ( props: { children: JSX.Element } ) => JSX.Element;
type DirectiveRef<T extends unknown[]> = ( ...args: T ) => (( ref: Element ) => void);
type DirectiveRegister = () => void;
type Directive = { Provider: DirectiveProvider, ref: DirectiveRef, register: DirectiveRegister };

function createDirective <T extends unknown[] = []> ( name: string, fn: DirectiveFunction<T>, options?: DirectiveOptions ): Directive;

Usage:

import {createDirective, useEffect} from 'voby';

// First of all if you are using TypeScript you should extend the "JSX.Directives" interface, so that TypeScript will know about your new directive

namespace JSX {
  interface Directives {
    tooltip: [title: string] // Mapping the name of the directive to the array of arguments it accepts
  }
}

// Then you should create a directive provider

const TooltipDirective = createDirective ( 'tooltip', ( ref, title: string ) => {

  useEffect ( () => {

    if ( !ref () ) return; // The element may not be available yet, or it might have been unmounted

    // Code that implements a tooltip for the given element here...

  });

});

// Then you can use the new "tooltip" directive anywhere inside the "TooltipDirective.Provider"

const App = () => {
  return (
    <TooltipDirective.Provider>
      <input value="Placeholder..." use:tooltip={['This is a tooltip!']} />
    </TooltipDirective.Provider>
  );
};

// You can also use directives directly by padding them along as refs

const App = () => {
  return <input ref={TooltipDirective.ref ( 'This is a tooltip!' )} value="Placeholder..." />;
};

createElement

This is the internal function that will make DOM nodes and call/instantiate components, it will be called for you automatically via JSX.

Interface:

function createElement <P = {}> ( component: JSX.Component<P>, props: P | null, ...children: JSX.Element[] ): () => JSX.Element);

Usage:

import {createElement} from 'voby';

const element = createElement ( 'div', { class: 'foo' }, 'child' ); // => () => HTMLDivElement

h

This function is just an alias for the createElement function, it's more convenient to use if you want to use Voby in hyperscript mode just because it has a much shorter name.

Interface:

function h <P = {}> ( component: JSX.Component<P>, props: P | null, ...children: JSX.Element[] ): () => JSX.Element);

Usage:

import {h} from 'voby';

const element = h ( 'div', { class: 'foo' }, 'child' ); // => () => HTMLDivElement

hmr

This function wraps a component and makes it HMR-aware, for implementations of HMR like Vite's, this makes the component refresh itself and its children without requiring a reload of the whole page.

For an automated way to make all your components HMR-aware check out voby-vite instead.

Interface:

function hmr <T extends Function> ( accept: Function, component: T ): T;

Usage:

import {hmr} from 'voby';

// Define a component

const Counter = ({ value }): JSX.Element => {
  // Return something...
};

// Optionally attach components and other values to it

Counter.Button = ({ onClick }): JSX.Element => {
  // Return something...
};

Counter.INITIAL_VALUE = 0;

// Lastly export it as "default", wrapped in "hmr"
// Only components exported as "default" are supported

export default hmr ( import.meta.hot?.accept?.bind ( import.meta.hot ), Counter );

html

This function provides an alternative way to use the framework, without writing JSX or using the h function manually, it instead allows you to write your markup as tagged template literals.

htm is used under the hood, read its documentation.

Interface:

function html ( strings: TemplateStringsArray, ...values: any[] ): JSX.Element;

Usage:

import {html, If} from 'voby';

const Counter = (): JSX.Element => {
  const value = $(0);
  const increment = () => value ( prev => prev + 1 );
  const decrement = () => value ( prev => prev - 1 );
  return html`
    <h1>Counter</h1>
    <p>${value}</p>
    <button onClick=${increment}>+</button>
    <button onClick=${decrement}>-</button>
  `;
};

// Using a custom component without registering it

const NoRegistration = (): JSX.Element => {
  return html`
    <${If} when=${true}>
      <p>content</p>
    </${If}>
  `;
};

// Using a custom component after registering it, so you won't need to interpolate it anymore

html.register ({ If });

const NoRegistration = (): JSX.Element => {
  return html`
    <If when=${true}>
      <p>content</p>
    </If>
  `;
};

isBatching

This function tells you if batching is currently active or not.

Interface:

function isBatching (): boolean;

Usage:

import {batch, isBatching} from 'voby';

// Checking if currently batching

isBatching (); // => false

batch ( () => {

  isBatching (); // => true

});

isBatching (); // => false

isObservable

This function tells you if a variable is an observable or not.

Interface:

function isObservable <T = unknown> ( value: unknown ): value is Observable<T> | ObservableReadonly<T>;

Usage:

import {$, isObservable} from 'voby';

isObservable ( 123 ); // => false
isObservable ( $(123) ); // => true

isServer

This function tells you if your code is executing in a browser environment or not.

Interface:

function isServer (): boolean;

Usage:

import {isServer} from 'voby';

isServer (); // => true or false

isStore

This function tells you if a variable is a store or not.

Interface:

function isStore ( value: unknown ): boolean;

Usage:

import {store, isStore} from 'voby';

isStore ( {} ); // => false
isStore ( store ( {} ) ); // => true

lazy

This function creates a lazy component, which is loaded via the provided function only when/if needed.

This function uses useResource internally, so it's significant for Suspense too.

Interface:

type LazyComponent<P = {}> = ( props: P ) => ObservableReadonly<Child>;
type LazyFetcher<P = {}> = () => Promise<{ default: JSX.Component<P> } | JSX.Component<P>>;
type LazyResult<P = {}> = LazyComponent<P> & ({ preload: () => Promise<void> });

function lazy <P = {}> ( fetcher: LazyFetcher<P> ): LazyResult<P>;

Usage:

import {lazy} from 'voby';

const LazyComponent = lazy ( () => import ( './component' ) );

render

This function mounts a component inside a provided DOM element and returns a disposer function for unmounting it and stopping all reactivity inside it.

Interface:

function render ( child: JSX.Element, parent?: HTMLElement | null ): Disposer;

Usage:

import {render} from 'voby';

const App = () => <p>Hello, World!</p>;

const dispose = render ( <App />, document.body );

dispose (); // Unmounted and all reactivity inside it stopped

renderToString

This works just like render, but it returns a Promise to the HTML representation of the rendered component.

This is currently implemented in a way that works only inside a browser-like environement, so you'll need to use JSDOM or similar for this to work server-side, but it can work server-side too potentially.

This function automatically waits for all Suspense boundaries to resolve before returning.

Interface:

function renderToString ( child: JSX.Element ): Promise<string>;

Usage:

import {renderToString} from 'voby';

const App = () => <p>Hello, World!</p>;

const html = await renderToString ( <App /> );

resolve

This function basically resolves any reactivity inside the passed argument, basically replacing every function it finds with a memo to the value of that function.

You may never need to use this function yourself, but it's necessary internally at times to make sure that a child value is properly tracked by its parent computation.

Read upstream documentation.

Interface:

type ResolvablePrimitive = null | undefined | boolean | number | bigint | string | symbol;
type ResolvableArray = Resolvable[];
type ResolvableObject = { [Key in string | number | symbol]?: Resolvable };
type ResolvableFunction = () => Resolvable;
type Resolvable = ResolvablePrimitive | ResolvableObject | ResolvableArray | ResolvableFunction;

function resolve <T> ( value: T ): T extends Resolvable ? T : never;

Usage:

import {resolve} from 'voby';

resolve // => Same as require ( 'oby' ).resolve

store

This function returns a deeply reactive version of the passed object, where property accesses and writes are automatically interpreted as Observables reads and writes for you.

Read upstream documentation.

Interface:

function store <T> ( value: T, options?: StoreOptions ): T;

Usage:

import {store} from 'voby';

store // => Same as require ( 'oby' ).store

template

This function enables constructing elements with Solid-level performance without using the Babel transform, but also without the convenience of that.

It basically works like sinuous's template function, but with a cleaner API, since you don't have to access your props any differently inside the template here.

Basically you can use this to wrap a component that doesn't directly create any observables or call any hooks to significanly improve performance when instantiating that component.

Interface:

function template <P = {}> ( fn: (( props: P ) => JSX.Element) ): (( props: P ) => () => Element);

Usage:

import {template} from 'voby';

const Row = template ( ({ id, cls, label, onSelect, onRemove }) => { // Now Row is super fast to instantiate
  return (
    <tr class={cls}>
      <td class="col-md-1">{id}</td>
      <td class="col-md-4">
        <a onClick={onSelect}>{label}</a>
      </td>
      <td class="col-md-1">
        <a onClick={onRemove}>
          <span class="glyphicon glyphicon-remove" ariaHidden={true}></span>
        </a>
      </td>
      <td class="col-md-6"></td>
    </tr>
  );
});

const Table = () => {
  const rows = [ /* props for all your rows here */ ];
  return rows.map ( row => <Row {...row}> );
};

tick

This function forces effects scheduled for execution to be executed immediately, bypassing automatic or manual batching.

Read upstream documentation.

Interface:

function tick (): void;

Usage:

import {tick} from 'voby';

tick // => Same as require ( 'oby' ).tick

untrack

This function executes the provided function without creating dependencies on observables retrieved inside it.

Read upstream documentation.

Interface:

function untrack <T> ( fn: () => T ): T;
function untrack <T> ( value: T ): T;

Usage:

import {untrack} from 'voby';

untrack // => Same as require ( 'oby' ).untrack

Components

The following components are provided.

Crucially some components are provided for control flow, since regular JavaScript control flow primitives are not reactive, and we need to have reactive alternatives to them to have great performance.

Dynamic

This component is just an alternative to createElement that can be used in JSX, it's useful to create a new element dynamically.

Interface:

function Dynamic <P = {}> ( props: { component: ObservableMaybe<JSX.Component<P>, props?: FunctionMaybe<P | null>, children?: JSX.Element }): JSX. Element;

Usage:

import {Dynamic} from 'voby';

const App = () => {
  const heading = 'h2';
  return (
    <Dynamic component={heading}>
      Some content
    </Dynamic>
  );
};

ErrorBoundary

The error boundary catches errors thrown inside it, and renders a fallback component when that happens.

Interface:

function ErrorBoundary ( props: { fallback: JSX.Element | (( props: { error: Error, reset: Callback } ) => JSX.Element), children: JSX.Element }): ObservableReadonly<JSX.Element>;

Usage:

import {ErrorBoundary} from 'voby';

const Fallback = ({ reset, error }: { reset: () => void, error: Error }) => {
  return (
    <>
      <p>Error: {error.message}</p>
      <button onClick={error}>Recover</button>
    </>
  );
};

const SomeComponentThatThrows = () => {
  throw 'whatever';
};

const App = () => {
  return (
    <ErrorBoundary fallback={Fallback}>
      <SomeComponentThatThrows />
    </ErrorBoundary>
  );
};

For

This component is the reactive alternative to natively mapping over an array.

It must be called with an array, or a function that returns an array, of unique values, and each of them are passed to the child function to render something.

It can be used to map over values either with a keyed (default) or unkeyed (opt-in) strategy. Read this for some details about the differences between those, and the upstream documentation.

Interface:

function For <T> ( props: { values?: FunctionMaybe<readonly T[]>, fallback?: JSX.ELement, children: (( value: T, index: FunctionMaybe<number> ) => JSX.Element) }): ObservableReadonly<JSX.Element>;
function For <T> ( props: { values?: FunctionMaybe<readonly T[]>, fallback?: JSX.ELement, pooled?: true, unkeyed?: true, children: (( value: ObservableReadonly<T>, index: FunctionMaybe<number> ) => JSX.Element) }): ObservableReadonly<JSX.Element>;

Usage:

import {For} from 'voby';

const App = () => {
  const numbers = [1, 2, 3, 4, 5];
  return (
    <For values={numbers}>
      {( value ) => {
        return <p>Value: {value}</p>
      }}
    </For>
  );
};

Fragment

This is just the internal component used for rendering fragments: <></>, you probably would never use this directly even if you are not using JSX, since you can return plain arrays from your components anyway.

Interface:

function Fragment ( props: { children: JSX.Element }): JSX.Element;

Usage:

import {Fragment} from 'voby';

const App = () => {
  return (
    <Fragment>
      <p>child 1</p>
      <p>child 2</p>
    </Fragment>
  );
};

If

This component is the reactive alternative to the native if.

If a function is passed as the children then it will be called with a read-only observable that contains the current, always truthy, value of the "when" condition.

Interface:

type Truthy<T = unknown> = Extract<T, number | bigint | string | true | object | symbol | Function>;

function If <T> ( props: { when: FunctionMaybe<T>, fallback?: JSX.Element, children: JSX.Element | (( value: (() => Truthy<T>) ) => JSX.Element) }): ObservableReadonly<JSX.Element>;

Usage:

import {If} from 'voby';

const App = () => {
  const visible = $(false);
  const toggle = () => visible ( !visible () );
  return (
    <>
      <button onClick={toggle}>Toggle</button>
      <If when={visible}>
        <p>Hello!</p>
      </If>
    </>
  );
};

KeepAlive

This component allows you to create singleton instances of other components that survive their parent components being disposed, and can therefore be reused cheaply.

Components rendered inside a KeepAlive are detached from the rest of the reactivity graph, so they don't inherit any context provided outside of their parent KeepAlive wrapper.

Components rendered inside a KeepAlive are kept in memory until the wrapper KeepAlive is unmounted and ttl milliseconds have passed without another KeepAlive with the same id being mounted. Or never, if no ttl prop is provided.

Interface:

function KeepAlive ( props: { id: FunctionMaybe<string>, ttl?: FunctionMaybe<number>, children: JSX.Element } ): ObservableReadonly<JSX.Element>;

Usage:

import {KeepAlive} from 'voby';

// Render some expensive component inside a KeepAlive

const App = () => {
  return (
    <KeepAlive id="some-unique-id" ttl={60_000}>
      <SomeExpensiveComponent />
    </KeepAlive>
  );
};

Portal

This component mounts its children inside a provided DOM element, or inside document.body otherwise.

The mount prop can also be an observable, if its value changes the portal is reparented.

The when prop can be used to apply the portal conditionally, if it explicitly resolves to false then children are mounted normally, as if they weren't wrapped in a portal.

Events will propagate natively, according to the resulting DOM hierarchy, not the components hierarchy.

Interface:

function Portal ( props: { when: boolean, mount?: JSX.Element, wrapper?: JSX.Element, children: JSX.Element }): (() => JSX.Element | null) & ({ metadata: { portal: HTMLDivElement } });

Usage:

import {Portal} from 'voby';

const Modal = () => {
  // Some modal component maybe...
};

const App = () => {
  return (
    <Portal mount={document.body}>
      <Modal />
    </Portal>
  );
};

Suspense

This component is like If, the reactive alternative to the native if, but the fallback branch is shown automatically while there are some resources loading in the main branch, and the main branch is kept alive under the hood.

So this can be used to show some fallback content while the actual content is loading in the background.

This component relies on useResource to understand if there's a resource loading or not.

This component also supports a manual "when" prop for manually deciding whether the fallback branch should be rendered or not.

Interface:

function Suspense ( props: { when?: FunctionMaybe<unknown>, fallback?: JSX.Element, children: JSX.Element }): ObservableReadonly<JSX.Element>;

Usage:

import {Suspense} from 'voby';

const App = () => {
  const Content = () => {
    const resource = useResource ( () => makeSomePromise () );
    return (
      <If when={() => !resource ().pending && !resource ().error}>
        {resource ().value}
      </If>
    );
  };
  const Spinner = () => {
    return <p>Loading...</p>;
  };
  return (
    <Suspense fallback={<Spinner />}>
      <Content />
    </Suspense>
  );
};

Switch

This component is the reactive alternative to the native switch.

Interface:

function Switch <T> ( props: { when: FunctionMaybe<T>, fallback?: JSX.Element, children: JSX.Element }): ObservableReadonly<JSX.Element>;

Switch.Case = function <T> ( props: { when: T, children: JSX.Element } ): (() => JSX.Element) & ({ metadata: [T, JSX.Element] });
Switch.Default = function ( props: { children: JSX.Element } ): (() => JSX.Element) & ({ metadata: [JSX.Element] });

Usage:

import {Switch} from 'voby';

const App = () => {
  const value = $(0);
  const increment = () => value ( value () + 1 );
  const decrement = () => value ( value () - 1 );
  return (
    <>
      <Switch when={value}>
        <Switch.Case when={0}>
          <p>0, the boundary between positives and negatives! (?)</p>
        </Switch.Case>
        <Switch.Case when={1}>
          <p>1, the multiplicative identity!</p>
        </Switch.Case>
        <Switch.Default>
          <p>{value}, I don't have anything interesting to say about that :(</p>
        </Switch.Default>
      </Switch>
      <button onClick={increment}>+</button>
      <button onClick={decrement}>-</button>
    </>
  );
};

Ternary

This component is the reactive alternative to the native ternary operator.

The first child will be rendered when the condition is truthy, otherwise the second child will be rendered.

Interface:

function Ternary ( props: { when: FunctionMaybe<unknown>, children: [JSX.Element, JSX.Element] } ): ObservableReadonly<JSX.Element>;

Usage:

import {Ternary} from 'voby';

const App = () => {
  const visible = $(false);
  const toggle = () => visible ( !visible () );
  return (
    <>
      <button onClick={toggle}>Toggle</button>
      <Ternary when={visible}>
        <p>Visible :)</p>
        <p>Invisible :(</p>
      </Ternary>
    </>
  );
};

Hooks core

The following core hooks are provided.

Most of these are just functions that oby provides, re-exported as use* functions.

useBoolean

This hook is like the reactive equivalent of the !! operator, it returns you a boolean, or a function to a boolean, depending on the input that you give it.

Read upstream documentation.

Interface:

function useBoolean ( value: FunctionMaybe<unknown> ): FunctionMaybe<boolean>;

Usage:

import {useBoolean} from 'voby';

useBoolean // => Same as require ( 'oby' ).boolean

useCleanup

This hook registers a function to be called when the parent computation is disposed.

Read upstream documentation.

Interface:

function useCleanup ( fn: () => void ): void;

Usage:

import {useCleanup} from 'voby';

useCleanup // => Same as require ( 'oby' ).cleanup

useContext

This hook retrieves the value out of a context object.

Interface:

function useContext <T> ( context: Context<T> ): T | undefined;

Usage:

import {createContext, useContext} from 'voby';

const App = () => {
  const ctx = createContext ( 123 );
  const value = useContext ( ctx );
  return <p>{value}</p>;
};

useDisposed

This hook returns a boolean read-only observable that is set to true when the parent computation gets disposed of.

Read upstream documentation.

Interface:

function useDisposed (): ObservableReadonly<boolean>;

Usage:

import {useDisposed} from 'voby';

useDisposed // => Same as require ( 'oby' ).disposed

useEffect

This hook registers a function to be called when any of its dependencies change. If a function is returned it's automatically registered as a cleanup function.

Read upstream documentation.

Interface:

function useEffect ( fn: () => (() => void) | void ): (() => void);

Usage:

import {useEffect} from 'voby';

useEffect // => Same as require ( 'oby' ).effect

useMemo

This hook is the crucial other ingredient that we need, other than observables themselves, to have a powerful reactive system that can track dependencies and re-execute computations when needed.

This hook registers a function to be called when any of its dependencies change, and the return of that function is wrapped in a read-only observable and returned.

Read upstream documentation.

Interface:

function useMemo <T> ( fn: () => T, options?: MemoOptions<T | undefined> ): ObservableReadonly<T>;

Usage:

import {useMemo} from 'voby';

useMemo // => Same as require ( 'oby' ).memo

usePromise

This hook wraps a promise in an observable, so that you can be notified when it resolves or rejects.

This hook uses useResource internally, so it's significant for Suspense too.

Interface:

function usePromise <T> ( promise: FunctionMaybe<Promise<T>> ): ObservableReadonly<Resource<T>>;

Usage:

import {usePromise} from 'voby';

const App = () => {
  const request = fetch ( 'https://my.api' ).then ( res => res.json ( 0 ) );
  const resource = usePromise ( request );
  return () => {
    const state = resource ();
    if ( state.pending ) return <p>pending...</p>;
    if ( state.error ) return <p>{state.error.message}</p>;
    return <p>{JSON.stringify ( state.value )}</p>
  };
};

useReadonly

This hook creates a read-only observable out of another observable.

Read upstream documentation.

Interface:

function useReadonly <T> ( observable: Observable<T> | ObservableReadonly<T> ): ObservableReadonly<T>;

Usage:

import {useReadonly} from 'voby';

useReadonly // => Same as require ( 'oby' ).readonly

useResolved

This hook receives a value, or an array of values, potentially wrapped in functions and/or observables, and unwraps it/them.

If no callback is used then it returns the unwrapped value, otherwise it returns whatever the callback returns.

This is useful for handling reactive and non reactive values the same way. Usually if the value is a function, or always for convenience, you'd want to wrap the useResolved call in a useMemo, to maintain reactivity.

This is potentially a more convenient version of $$, made especially for handling nicely arguments passed that your hooks receive that may or may not be observables.

Interface:

The precise interface for this function is insane, you can find it here: https://github.com/fabiospampinato/voby/blob/master/src/hooks/use_resolved.ts

Usage:

import {$, useResolved} from 'voby';

useResolved ( 123 ); // => 123

useResolved ( $(123) ); // => 123

useResolved ( () => 123 ); // => 123

useResolved ( () => 123, false ); // => () => 123

useResolved ( $(123), value => 321 ); // => 321

useResolved ( [$(123), () => 123], ( a, b ) => 321 ); // => 321

useResource

This hook wraps the result of a function call with an observable, handling the cases where the function throws, the result is an observable, the result is a promise or an observale that resolves to a promise, and the promise rejects, so that you don't have to worry about these issues.

This basically provides a unified way to handle sync and async results, observable and non observable results, and functions that throw and don't throw.

This function is also the mechanism through which Suspense understands if there are things loading under the hood or not.

When the value property is read while fetching, or when the latest property is read the first time, or after an error, while fetching, then Suspense boundaries will be triggered.

When the value property or the latest property are read after the fetch errored they will throw, triggering ErrorBoundary.

The passed function is tracked and it will be automatically re-executed whenever any of the observables it reads change.

Interface:

function useResource <T> ( fetcher: (() => ObservableMaybe<PromiseMaybe<T>>) ): Resource<T>;

Usage:

import {useResource} from 'voby';

const fetcher = () => fetch ( 'https://my.api' );

const resource = useResource ( fetcher );

useRoot

This hook creates a new computation root, detached from any parent computation.

Read upstream documentation.

Interface:

function useRoot <T> ( fn: ( dispose: () => void ) => T ): T;

Usage:

import {useRoot} from 'voby';

useRoot // => Same as require ( 'oby' ).root

useSelector

This hook massively optimizes isSelected kind of workloads.

Read upstream documentation.

Interface:

type SelectorFunction<T> = ( value: T ) => ObservableReadonly<boolean>;

function useSelector <T> ( source: () => T | ObservableReadonly<T> ): SelectorFunction<T>;

Usage:

import {useSelector} from 'voby';

useSelector // => Same as require ( 'oby' ).selector

useSuspended

This hook returns a read-only observable that tells you if the closest suspense boundary is currently suspended or not.

Read upstream documentation.

Interface:

function useSuspended (): ObservableReadonly<boolean>;

Usage:

import {useSuspended} from 'voby';

useSuspended // => Same as require ( 'oby' ).suspended

useUntracked

This hook returns an untracked version of a value.

Read upstream documentation.

Interface:

function useUntracked <T> ( fn: () => T ): () => T;
function useUntracked <T> ( value: T ): () => T;

Usage:

import {useUntracked} from 'voby';

useUntracked // => Same as require ( 'oby' ).untracked

Hooks web

The following web hooks are provided.

Most of these are just reactive alternatives to native web APIs.

useAbortController

This hook is just an alternative to new AbortController () that automatically aborts itself when the parent computation is disposed.

Interface:

function useAbortController ( signals?: ArrayMaybe<AbortSignal> ): AbortController;

Usage:

import {useAbortController} from 'voby';

const controller = useAbortController ();

useAbortSignal

This hook is just a convenient alternative to useAbortController, if you are only interested in its signal, which is automatically aborted when the parent computation is disposed.

Interface:

function useAbortSignal ( signals?: ArrayMaybe<AbortSignal> ): AbortSignal;

Usage:

import {useAbortSignal} from 'voby';

const signal = useAbortSignal ();

useAnimationFrame

This hook is just an alternative to requestAnimationFrame that automatically clears itself when the parent computation is disposed.

Interface:

function useAnimationFrame ( callback: ObservableMaybe<FrameRequestCallback> ): Disposer;

Usage:

import {useAnimationFrame} from 'voby';

useAnimationFrame ( () => console.log ( 'called' ) );

useAnimationLoop

This hook is just a version of useAnimationFrame that loops until the parent computation is disposed.

Interface:

function useAnimationLoop ( callback: ObservableMaybe<FrameRequestCallback> ): Disposer;

Usage:

import {useAnimationLoop} from 'voby';

useAnimationLoop ( () => console.log ( 'called' ) );

useEventListener

This hook is just an alternative to addEventListener that automatically clears itself when the parent computation is disposed.

Interface:

function useEventListener ( target: FunctionMaybe<EventTarget>, event: FunctionMaybe<string>, handler: ObservableMaybe<( event: Event ) => void>, options?: FunctionMaybe<true | AddEventListenerOptions> ): Disposer;

Usage:

import {useEventListener} from 'voby';

useEventListener ( document, 'click', console.log );

useFetch

This hook wraps the output of a fetch request in an observable, so that you can be notified when it resolves or rejects. The request is also aborted automatically when the parent computation gets disposed of.

This hook uses useResource internally, so it's significant for Suspense too.

Interface:

function useFetch ( request: FunctionMaybe<RequestInfo>, init?: FunctionMaybe<RequestInit> ): ObservableReadonly<Resource<Response>>;

Usage:

import {useFetch} from 'voby';

const App = () => {
  const resource = useFetch ( 'https://my.api' );
  return () => {
    const state = resource ();
    if ( state.pending ) return <p>pending...</p>;
    if ( state.error ) return <p>{state.error.message}</p>;
    return <p>Status: {state.value.status}</p>
  };
};

useIdleCallback

This hook is just an alternative to requestIdleCallback that automatically clears itself when the parent computation is disposed.

Interface:

function useIdleCallback ( callback: ObservableMaybe<IdleRequestCallback>, options?: FunctionMaybe<IdleRequestOptions> ): Disposer;

Usage:

import {useIdleCallback} from 'voby';

useIdleCallback ( () => console.log ( 'called' ) );

useIdleLoop

This hook is just a version of useIdleCallback that loops until the parent computation is disposed.

Interface:

function useIdleLoop ( callback: ObservableMaybe<IdleRequestCallback>, options?: FunctionMaybe<IdleRequestOptions> ): Disposer;

Usage:

import {useIdleLoop} from 'voby';

useIdleLoop ( () => console.log ( 'called' ) );

useInterval

This hook is just an alternative to setInterval that automatically clears itself when the parent computation is disposed.

Interface:

function useInterval ( callback: ObservableMaybe<Callback>, ms?: FunctionMaybe<number> ): Disposer;

Usage:

import {useInterval} from 'voby';

useInterval ( () => console.log ( 'called' ), 1000 );

useMicrotask

This hook is just an alternative to queueMicrotask that automatically clears itself when the parent computation is disposed, and that ensures things like contexts, error boundaries etc. keep working inside the microtask.

Interface:

function useMicrotask ( fn: () => void ): void;

Usage:

import {useMicrotask} from 'voby';

useMicrotask ( () => console.log ( 'called' ) );

useTimeout

This hook is just an alternative to setTimeout that automatically clears itself when the parent computation is disposed.

Interface:

function useTimeout (