@svelte-drama/swr v4.0.2

@svelte-suspense/swr

This is a data management/fetching library written for Svelte, inspired by SWR, and with built in integrations for Suspense. By keeping all requests in cache, views can be rendered instantly while refetching any potentially stale data in the background.

See it in action

Installation

npm install --save @svelte-suspense/swr

Requirements

SWR requires BroadcastChannel and LockManager.

Full support for those interfaces is included natively in:

• Chrome 69
• Edge 79
• FireFox 96
• Opera 56
• Safari 15.4

Usage

swr

import { swr } from '@svelte-drama/swr'

const model = swr<ID, MODEL>({
  key(id: ID) {
    return `/api/endpoint/${id}`
  },
  async fetcher(key: string, id: ID) {
    const request = fetch(key)
    return request.json() as MODEL
  },
})

ID may be of type any. MODEL must be an object that can be cloned via the structured clone algorithm.

Options

• key(id: ID) => string

  A function to create a unique cache when given the user defined id. Typically, this is the API path this data would be fetched from.

• fetcher(key: string, id: ID) => MODEL | Promise<MODEL>

  A function to retrieve data from the server. It is passed key, the result of the key function and the same id passed to the key function.

• maxAge: number

  Only use cached values that are no older than maxAge in milliseconds.

• name: string

  Segment the cache using this as a key. Models with the same name share the same cache, so key collision must be kept in mind.

Methods

The returned object model has several functions for fetching data.

• model.clear() => Promise<void>

  Clear all data from this cache.

• model.clearErrors() => Promise<void>

  Clear all errors from this cache.

• model.delete(id: ID) => Promise<void>

  Delete item from cache.

• model.fetch(id: ID) => Promise<MODEL>

  Returns data from cache if less than maxAge or performs a request using the provided fetcher

• model.get(id: ID | undefined) => MODEL | undefined

  Returns data from cache if less than maxAge or performs a request using the provided fetcher. Use inside of $effect or $derived in order to update as the underlying data changes.

• model.keys() => Promise<string[]>

  Returns all currently cached keys, regardless of age.

• model.refresh(id: ID) => Promise<MODEL>

  Performs a request using the provided fetcher. Always makes a request, regradless of current cache status.

• model.update(id: ID, data: MODEL) => Promise<MODEL>
model.update(id: ID, fn: (data: MODEL) => MODEL | Promise<MODEL>) => Promise<MODEL>

  Update data in the cache.

clear

import { clear } from '@svelte-drama/swr'

clear()

Remove all data from all caches.

clearErrors

import { clearErrors } from '@svelte-drama/swr'

clearErrors()

Remove all errors from all caches.