Fresh Data Framework

Fresh Data Framework is a declarative data API framework for JavaScript apps.

Caveat

Fresh Data is new. Very new. As such it should not be used in production just yet! This code will be changing still.

Try it out on something noncritical and provide some feedback!

Installation

npm install --save @fresh-data/framework

Benefits

• Keep data in your web application current, without writing any application code to do it.
• Avoid fetching data that you already have in browser state.
• Works with any way data can be fetched (REST, GraphQL, WebSockets, Offline, etc.)
• Automatically clear out old data that hasn't been used for a while (coming soon)

How it works

1. Applications declare the data they need and how they need it.
2. APIs define the way data is stored and accessed.

There is support for React applications using Redux for an internal cache available from @fresh-data/react-redux

Creating a Fresh Data API Module

Each API Specification can be kept in your application or a separate module.

import { compact, startsWith } from 'lodash';

const URL_ROOT = 'http://example.com/';

const get( endpointPath, params ) => {
	const uri = URL_ROOT + endpointPath.join( '/' );
	const queryString = qs.stringify( params );
	return fetch( `${ uri }?${ query }` );
}

const put( endpointPath, params ) => {
	const uri = URL_ROOT + endpointPath.join( '/' );
	const { data } = params;
	return fetch( uri, { method: 'PUT', body: JSON.stringify( data ) } );
}

const apiSpec = {
	operations: {
		read: ( resourceNames ) => {
			return compact( resourceNames.map( resourceName => {
				if ( startsWith( resourceName, 'thing:' ) ) {
					const thingNumber = resourceName.substr( resourceName.indexOf( ':' ) + 1 );

					const request = get( [ 'things' ] ).then( responseData => {
						return { [ resourceName ]: { data: responseData } };
					} );
					return request;
				}
			} ) );
		},
		update: ( resourceNames, resourceData ) => {
			return compact( resourceNames.map( resourceName => {
				if ( startsWith( resourceName, 'thing:' ) ) {
					const thingNumber = resourceName.substr( resourceName.indexOf( ':' ) + 1 );
					const data = resourceData[ resourceName ];

					const request = put( [ 'things' ], { data } ).then( responseData => {
						return { [ resourceName ]: { data: responseData } };
					} );
					return request;
				}
			} ) );
		}
	},
	mutations: {
		updateThing: ( operations ) => ( thingId, data ) => {
			const resourceName = `thing:${ thingId }`;
			const resourceNames = [ resourceName ];
			const resourceData = { [ resourceName ]: data };
			return operations.update( resourceNames, resourceData );
		}
	},
	selectors: {
		getThing: ( getResource, requireResource ) => ( requirement, thingId ) => {
			const resourceName = `thing:${ thingId }`;
			return requireResource( requirement, resourceName );
		}
	}
};

export apiSpec;

Your own API depends on the operations, methods, and selectors you define.

• Operations: The operations you can perform on your data (e.g. read, update, create, delete )
• Mutations: Functions you provide to application developers can call to perform operations on your data.
• Selectors: Functions you provide to application developers to access data in their preferred format.

Still to be completed

Fresh Data is functional, but still a work in progress. Here's what's next on the list:

• More examples:
  • GitHub API
  • GraphQL
  • WebSockets
• Feature: Fetch on first mount (regardless of freshness)
• Feature: Clearing out old data
  • Detecting when data was last rendered
  • Unlinking data over a threshold