feathers-reactive


> Reactive API extensions for Feathers

About

feathers-reactive adds a watch() method to services. The returned object implements all service methods as RxJS v7 observables that automatically update on real-time events.

Options

The following options are supported:

- idField (mandatory): The id property field of your services. Depends on your service/database. Usually 'id' (SQL, Rethinkdb, …) or '_id' (MongoDB, NeDB, … ).
- dataField (default: data): The data property field in paginated responses
- listStrategy (default: smart): The strategy to use for streaming the data. Can be smart, always or never. __Avoid using always whenever possible__.
- sorter (function(query, options) {}): A function that returns a sorting function for the given query and option including pagination and limiting. Does not need to be customized unless there is a sorting mechanism other than Feathers standard in place.
- matcher (function(query)): A function that returns a function which returns whether an item matches the original query or not.
- pipe (operator | operator[]) One or multiple rxjs operators of the form function(observable) => observable like you would pass them to an Observable's .pipe method. The supplied operators are applied to any Observable created by feathers-reactive. options.pipe: tap(data => console.log(data)) would log every emitted value to the console.

#### Application level

``js
import feathers from '@feathersjs/feathers';
import { rx } from 'feathers-reactive';

const app = feathers().configure(rx(options));
`

#### Service level

With feathers-reactive configured on the application individual options can be set at the service level with service.rx:

`js
// Set a different id field
app.service('todos').rx({
idField: '_id'
});
`

#### Method call level

Each method call can also pass its own options via params.rx:

`js
// Never update data for this method call
app.service('todos').watch({ listStrategy: 'never' }).find();
`

$3

List strategies are used to determine how a data stream behaves. Currently there are three strategies:

- never - Returns a stream from the service promise that only emits the method call data and never updates after that
- smart (default) - Returns a stream that smartly emits updated list data based on the services real-time events. It does not re-query any new data (but does not cover some cases in which the always strategy can be used). When using smart list strategy, an additional method reset is available to get fresh data from the server.
- always - Re-runs the original query to always get fresh data from the server on any matching real-time event. __Avoid this list strategy if possible__ since it will put a higher load on the server than necessary.

Usage

`js
import {feathers} from '@feathersjs/feathers';
import memory from 'feathers-memory';
import { rx } from 'feathers-reactive';

const app = feathers()
.configure(rx({
idField: 'id'
}))
.use('/messages', memory());

const messages = app.service('messages');

messages.create({
text: 'A test message'
}).then(() => {
// Get a specific message with id 0. Emit the message data once it resolves
// and every time it changes e.g. through an updated or patched event
messages.watch().get(0).subscribe(message => console.log('My message', message));

// Find all messages and emit a new list every time anything changes
messages.watch().find().subscribe(messages => console.log('Message list', messages));

setTimeout(() => {
messages.create({ text: 'Another message' }).then(() =>
setTimeout(() => messages.patch(0, { text: 'Updated message' }), 1000)
);
}, 1000);
});
`

Will output:

`console
My message { text: 'A test message', id: 0 }
Message list [ { text: 'A test message', id: 0 } ]
Message list [ { text: 'A test message', id: 0 },
{ text: 'Another message', id: 1 } ]
My message { text: 'Updated message', id: 0 }
Message list [ { text: 'Updated message', id: 0 },
{ text: 'Another message', id: 1 } ]
`

Frameworks

Let's assume a simple Feathers Socket.io server in app.js like this:

> npm install @feathersjs/feathers @feathersjs/socketio feathers-memory

`js
import {feathers} from '@feathersjs/feathers';
import socketio from '@feathersjs/socketio';
import memory from 'feathers-memory';

const app = feathers()
.configure(socketio())
.use('/todos', memory());

app.on('connection', connection => app.channel('everybody').join(connection));
app.publish(() => app.channel('everybody'));

app.listen(3030).on('listening', () =>
console.log('Feathers Socket.io server running on localhost:3030')
);
`

$3

For an ES5 compatible version on the client (e.g. when using create-react-app) you can import feathers-reactive/dist/feathers-reactive. In client.js:

`js
import io from 'socket.io-client';
import feathers from '@feathersjs/client';
import rx from 'feathers-reactive/dist/feathers-reactive';

const socket = io('http://localhost:3030');
const app = feathers()
.configure(feathers.socketio(socket))
.configure(rx({
idField: 'id'
}));

export default app;
`

$3

A real-time ReactJS Todo application (with Bootstrap styles) can look like this (see the examples/react-todos folder for a working example);

`js
import React, { Component } from 'react';
import client from './client';

class App extends Component {
constructor (props) {
super(props);
this.state = {
todos: [],
text: ''
};
}

componentDidMount () {
this.todos = client.service('todos').watch()
.find().subscribe(todos => this.setState(todos));
}

componentWillUnmount () {
this.todos.unsubscribe();
}

updateText (ev) {
this.setState({ text: ev.target.value });
}

createTodo (ev) {
client.service('todos').create({
text: this.state.text,
complete: false
});
this.setState({ text: '' });
ev.preventDefault();
}

updateTodo (todo, ev) {
todo.complete = ev.target.checked;
client.service('todos').patch(todo.id, todo);
}

deleteTodo (todo) {
client.service('todos').remove(todo.id);
}

render () {
const renderTodo = todo =>