loQL

A light, modular npm package for performant client-side GraphQL caching with Service Workers and IndexedDB. More detailed information about installing and configuring loQL can be found here.

Installation

Install via npm

npm install loql-cache

Or with Yarn

yarn add loql-cache

The service worker must also be included in your build folder. With webpack:

const path = require('path');

module.exports = {
  entry: {
    bundle: './client/index.js',
    loQL: './node_modules/loql-cache/loQL.js', // Add this line!
  },
  output: {
  path: path.resolve(__dirname, 'public'),
    filename: '[name].js',
    clean: true,
  },
  devServer: {
    static: './client',
  },
};

Register the service worker

import { register } from "loql-cache";
register({ gqlEndpoints: ["https://foo.com"] });

Settings

gqlEndpoints: string[] Required

Enable caching for specific GraphQL endpoint URLs. Network calls from the browser to any URL not listed here will be ignored by the service worker and the response data will not be cached.

useMetrics: boolean Optional

Enable metrics collection.

cacheMethod: string Optional

Desired caching strategy. The loql-cache package supports both "cache-first" and "cache-network" policies.

cacheExpirationLimit: Integer Optional

The interval, in milliseconds, after which cached data is considered stale.

doNotCacheGlobal: string[] Optional

Fields on a GraphQL query that will prevent the query from being cached, no matter the endpoint.

doNotCacheCustom:{ [url]: string[] } Optional

This setting is like doNotCacheGlobal, but can be used on a per-endpoint basis.

Example Configuration

const loQLConfiguration = {
  gqlEndpoints: ['http://localhost:<###>/api/graphql', 'https://<abc>.com/graphql'],
  useMetrics: false,
  cacheExpirationLimit: 20000,
  cacheMethod: 'cache-network',
  doNotCacheGlobal: [],
  doNotCacheCustom: {
     'http://localhost:<###>/api/graphql': ['password'],
     'https://<abc>.com/graphql': ['account', 'real_time_data'];
  }
};

register(loqlConfiguration);

Features

• Enables offline use: IndexedDB storage provides high-capacity and persistent storage, while keeping reads/writes asynchronous
• Minimum-dependency: No server-side component, avoid the use of large libraries
• Cache validation: Keep data fresh with shorter expiration limits, cache-network strategy, or both!
• Easy-to-use: Install package, register and configure service worker, start caching
• Flexible: Works with GQL queries made as both fetch POST and GET requests
• Easily exempt types of queries from being cached at the global or endpoint-specific level

Usage Notes

• Caching is currently only supported for query-type operations. Mutations, subscriptions, etc will still run, but will not be cached.
• Cached data normalization feature is disabled.

Contributing

Contributions are welcome. Please read CONTRIBUTE.md prior to making a Pull Request.

License

MIT