A complete solution for building a React/Redux application: routing, page preloading, (optional) server-side rendering, asynchronous HTTP requests, document metadata, internationalization, etc.
Switch branches/tags
Nothing to show
Clone or download
Latest commit 8890eda Sep 18, 2018

README.md

react-website

npm version npm downloads npm downloads

A complete solution for building a React/Redux application

  • Routing
  • Page preloading
  • (optional) Code splitting
  • (optional) Server-side rendering
  • Asynchronous HTTP requests
  • Easy and simplified Redux (no boilerplate code)
  • Document metadata (<title/>, <meta/>, social network sharing)
  • Webpack "hot reload"
  • HTTP Cookies
  • etc

2.x

These are the docs for the latest version 3.x. For version 2.x (legacy projects) docs see the README of the 2.x branch.

Introduction

Getting started

First, install Redux:

$ npm install redux react-redux --save

Then, install react-website:

$ npm install react-website --save

Start by creating react-website configuration file.

./src/react-website.js

import routes from './routes'

// Redux reducers, which will be combined into
// a single Redux reducer via `combineReducers()`.
import * as reducers from './redux/index'

export default {
  routes,
  reducers
}

The routes:

./src/routes.js

import React from 'react'
import { Route } from 'react-website'

import App from '../pages/App'
import Home from '../pages/Home'
import About from '../pages/About'

export default (
  <Route path="/" component={ App }>
    <Route component={ Home }/>
    <Route path="about" component={ About }/>
  </Route>
)

./src/pages/App.js

import React from 'react'
import { Link } from 'react-website'

export default ({ children }) => (
  <div>
    <h1> Web Application </h1>
    <ul>
      <li> <Link exact to="/"> Home </Link> </li>
      <li> <Link to="/about"> About </Link> </li>
    </ul>
    { children }
  </div>
)

./src/pages/Home.js

import React from 'react'

export default () => <div> This is a home page </div>

./src/pages/About.js

import React from 'react'

export default () => <div> Made using `react-website` </div>

The reducers:

./src/redux/index.js

// For those who're unfamiliar with Redux,
// a reducer is a function `(state, action) => state`.
export { default as homePage } from './homePageReducer'
export { default as aboutPage } from './aboutPageReducer'
...

Then call render() in the main client-side javascript file.

./src/index.js

import { render } from 'react-website'
import settings from './react-website'

// Render the page in web browser
render(settings)

And the index.html would look like this:

<html>
  <head>
    <title>Example</title>
  </head>
  <body>
    <script src="/bundle.js"></script>
  </body>
</html>

Where bundle.js is the ./src/index.js file built with Webpack (or you could use any other javascript bundler).

Now, index.html and bundle.js files must be served over HTTP(S).

If you're using Webpack then add HtmlWebpackPlugin to generate index.html, and run webpack-serve with historyApiFallback to serve the generated index.html and bundle.js files over HTTP on localhost:8080.

If you're using Parcel then it's much simpler than Webpack: see the basic example project for the setup required in order to generate and serve index.html and bundle.js files over HTTP on localhost:1234.

So now the website should be fully working.

The website (index.html, bundle.js, CSS stylesheets and images, etc) can now be deployed as-is in a cloud (e.g. on Amazon S3) and served statically for a very low price. The API can be hosted "serverlessly" in a cloud (e.g. Amazon Lambda) which is also considered cheap. No running Node.js server is required. Yes, it's not a Server-Side Rendered approach because a user is given a blank page first, then bundle.js script is loaded by the web browser, then bundle.js script is executed fetching some data from the API via an HTTP request, and only when that HTTP request comes back — only then the page is rendered (in the browser). Google won't index such websites, but if searchability is not a requirement (at all or yet) then that would be the way to go (e.g. startup "MVP"s or "internal applications"). Server-Side Rendering can be easily added to such setup should the need arise.

Server Side Rendering

Search engines

Search engine crawlers like Google bot won't wait for a page to make its asynchronous HTTP calls to an API server for data: they would simply abort all asynchronous javascript and index the page as is. Don't mistake it for web crawlers not being able to execute javascript — they're perfectly fine with doing that (watch out though for using the latest javascript language features and always use polyfills for the older browsers since web crawlers may be using those under the hood).

So the only thing preventing a dynamic website from being indexed by a crawler is asynchronous HTTP queries for data, not javascript itself. This therefore brings two solutions: one is to perform everything (routing, data fetching, rendering) on the server side and the other is to perform routing and data fetching on the server side leaving rendering to the client's web browser. Both these approaches work with web crawlers. And this is what this library provides.

While the first approach is more elegant and pure, while also delivering the fastest "time to first byte", currently it is a CPU intensive task to render a complex React page (takes about 30 milliseconds of blocking CPU single core time for complex pages having more than 1000 components, as of 2017). Therefore one may prefer the second approach: performing routing and page preloading on the server side while leaving page rendering to the client. This means that the user won't see any content until the javascript bundle is downloaded (which takes some time, especially with large applications not using "code splitting"), but it also means that the server's CPU is freed from rendering React. This mode is activated by passing renderContent: false flag to the rendering server.

Page loading time

Another argument in favour of Server-Side Rendering is that even if a website doesn't need search engine indexing it could still benefit from saving that additional asynchronous HTTP roundtrip from the web browser to the API server for fetching the page's data. And no matter how fast the API server is, latency is unbeatable being about 100ms. So, by performing routing and page preloading on the server side one can speed up website loading by about 100ms.

Adding server-side rendering

Not everyone needs server-side rendering for their apps. E.g. if search engine indexing is not a priority, or if a website is a "static" one, like a "promosite" or a "personal portfolio" (just build it with a bundler and host it as a bunch of files in a cloud).

Adding server-side rendering to the setup is quite simple though requiring a Node.js process running which increases hosting costs and maintenance complexity.

In case of server-side rendering index.html is being generated on-the-fly by page rendering server for each incoming HTTP request, so the index.html file may be deleted as it's of no use now.

./rendering-server.js

import webpageServer from 'react-website/server'
import settings from './react-website'

// Create webpage rendering server
const server = webpageServer(settings, {
  // Pass `secure: true` for HTTPS.
  //
  // These are the URLs of the "static" javascript and CSS files
  // which are injected in the resulting Html webpage
  // as <script src="..."/> and <link rel="style" href="..."/>.
  // (this is for the main application JS and CSS bundles only,
  //  for injecting 3rd party JS and CSS use `html` settings instead:
  //  https://github.com/catamphetamine/react-website/blob/master/README-ADVANCED.md#all-webpage-rendering-server-options)
  assets() {
    return {
      // Assuming that it's being tested on a local computer first
      // therefore using "localhost" URLs.
      javascript: 'http://localhost:8080/bundle.js',
      // (optional) If using a separate CSS bundle:
      style: 'http://localhost:8080/bundle.css'
    }
  }
})

// Start webpage rendering server on port 3000
// (`server.listen(port, [host], [callback])`)
server.listen(3000, function(error) {
  if (error) {
    throw error
  }
  console.log(`Webpage rendering server is listening at http://localhost:3000`)
})

Run the rendering server:

$ npm install npx --global
$ npm install babel-cli
$ npx babel-node rendering-server.js

Now disable javascript in Chrome DevTools, go to localhost:3000 and the server should respond with a fully server-side-rendered page.

Conclusion

This concludes the introductory part of the README and the rest is the description of the various tools and techniques which come prepackaged with this library.

A working example illustrating Server-Side Rendering and all other things can be found here: webpack-react-redux-isomorphic-render-example.

A much simpler and smaller example (using Parcel instead of Webpack) can be found here: react-website-basic-example.

Documentation

Preloading pages

For page preloading use the @preload() decorator to load the neccessary data before the page is rendered.

import { connect } from 'react-redux'
import { preload } from 'react-website'

// Redux "asynchronous action",
// explained later in this document.
function fetchUsers() {
  return {
    promise: http => http.get('/api/users'),
    events: ['FETCH_USERS_PENDING', 'FETCH_USERS_SUCCESS', 'FETCH_USERS_FAILURE']
  }
}

@preload(async ({ dispatch }) => {
  // Send HTTP request and wait for response
  await dispatch(fetchUsers())
})
@connect(
  (state) => ({ users: state.usersPage.users }),
  // Calls `bindActionCreators()`
  // for the specified Redux action creators.
  { fetchUsers }
)
export default class UsersPage extends Component {
  render() {
    const { users, fetchUsers } = this.props
    return (
      <div>
        <ul> { users.map(user => <li> { user.name } </li>) } </ul>
        <button onClick={ fetchUsers }> Refresh </button>
      </div>
    )
  }
}

In this example the @preload() decorator is used to preload a page before it is displayed, i.e. before the page is rendered (both on server side and on client side).

@preload() decorator takes an async/await function which takes an object of arguments:

@preload(async (preloadArguments) => {
  const = {
    // Redux `dispatch()`
    dispatch,
    // Get Redux state
    getState,
    // Current page location
    location,
    // `react-router` URL `params`
    // (e.g. '/users/:id')
    params,
    // Is this server side rendering
    server
  }
  = preloadArguments

  // Send HTTP request and wait for response.
  await dispatch(fetchPageData(params.id))
})
The decorator also receives an optional `options` argument (advanced topic)
@preload(async () => { ... }, options)

The available options are:

  • blocking — If false then child <Route/>'s @preload()s will not wait for this @preload() to finish in order to get executed (blocking is true by default).

  • blockingSibling — If true then all further adjacent (sibling) @preload()s for the same <Route/>'s component will wait for this @preload() to finish in order to get executed. (is true by default).

  • client — If true then the @preload() will be executed only on client side. If false then this @preload() will be executed normally: if part of initial page preloading then on server side and if part of subsequent preloading (e.g. navigation) then on client side. false is the default value unless overridden by preload.client configuration parameter.

  • server — If true then the @preload() will be executed only on server side. If false then this @preload() will be executed normally: if part of initial page preloading then on server side and if part of subsequent preloading (e.g. navigation) then on client side. false is the default value unless overridden by preload.client configuration parameter.

Note: transform-decorators-legacy Babel plugin is needed at the moment to make decorators work with Babel:

npm install babel-plugin-transform-decorators-legacy --save

.babelrc

{
  ...
  "plugins": [
    "transform-decorators-legacy",
    ...
  ]
}

On the client side, in order for @preload to work all <Link/>s imported from react-router must be instead imported from react-website. Upon a click on a <Link/> first it waits for the next page to preload, and then, when the next page is fully loaded, react-router navigation itself takes place.

`@preload` also works for Back/Forward navigation. To disable page `@preload` on Back navigation pass `instantBack` property to a ``.

For example, consider a search results page preloading some data (could be search results themselves, could be anything else unrelated). A user navigates to this page, waits for @preload to finish and then sees a list of items. Without instantBack if the user clicks on an item he's taken to the item's page. Then the user clicks "Back" and is taken back to the search results page but has to wait for that @preload again. With instantBack though the "Back" transition occurs instantly without having to wait for that @preload again. Same goes then for the reverse "Forward" navigation from the search results page back to the item's page, but that's just a small complementary feature. The main benefit is the instantaneous "Back" navigation creating a much better UX where a user can freely explore a list of results without getting penalized for it with a waiting period on each click.

@preload(async () => await fetchSomeData())
class SearchResultsPage extends Component {
  render() {
    return (
      <ul>
        { results.map((item) => (
          <li>
            <Link to="/items/{item.id}" instantBack>
              {item.name}
            </Link>
          </li>
        ))) }
      </ul>
    )
  }
}

One can also use the exported wasInstantNavigation() function (on client side) to find out if the current page was navigated to "instantly". This can be used, for example, for Algolia "Instant Search" component to reset cached search results if it's not an instant "Back" navigation.

@preload() indicator

Sometimes preloading a page can take some time so one may want to (and actually should) add some kind of a "spinner" to inform the user that the application isn't frozen and that the navigation process needs some more time to finish. This can be achieved by adding the built-in <Loading/> component on a page:

import { Loading } from 'react-website'
// Using Webpack CSS loader
import 'react-website/components/Loading.css'
import 'react-website/components/LoadingIndicator.css'

export default function Application() {
  return (
    <div>
      ....
      <Loading/>
    </div>
  )
}

The <Loading/> component takes an optional indicator property which can be a React component accepting a className property and which is a white circular spinner by default.

Asynchronous actions

Implementing synchronous actions in Redux is straightforward. But what about asynchronous actions like HTTP requests? Redux itself doesn't provide any built-in solution for that leaving it to 3rd party middlewares. Therefore this library provides one.

Pure Promises

This is the lowest-level approach to asynchronous actions. It is described here just for academic purposes and most likely won't be used directly in any app.

If a Redux "action creator" returns an object with a promise (function) and events (array) then dispatch()ing such an action results in the following steps:

  • An event of type = events[0] is dispatched
  • promise function gets called and returns a Promise
  • If the Promise succeeds then an event of type = events[1] is dispatched having result property set to the Promise result
  • If the Promise fails then an event of type = events[2] is dispatched having error property set to the Promise error
function asynchronousAction() {
  return {
    promise: () => Promise.resolve({ success: true }),
    events: ['PROMISE_PENDING', 'PROMISE_SUCCESS', 'PROMISE_ERROR']
  }
}

dispatch(asynchronousAction()) call returns the Promise itself:

@preload(async ({ dispatch }) => {
  await dispatch(asynchronousAction())
})

HTTP utility

Because in almost all cases dispatching an "asynchronous action" means "making an HTTP request", the promise function described above always takes an http argument: promise: http => ....

The http utility has the following methods:

  • head
  • get
  • post
  • put
  • patch
  • delete

Each of these methods returns a Promise and takes three arguments:

  • the url of the HTTP request
  • data object (e.g. HTTP GET query or HTTP POST body)
  • options (described further)

So, API endpoints can be queried using http and ES6 async/await syntax like so:

function fetchFriends(personId, gender) {
  return {
    promise: http => http.get(`/api/person/${personId}/friends`, { gender }),
    events: ['GET_FRIENDS_PENDING', 'GET_FRIENDS_SUCCESS', 'GET_FRIENDS_FAILURE']
  }
}

The possible options (the third argument of all http methods) are

  • headers — HTTP Headers JSON object.
  • authentication — Set to false to disable sending the authentication token as part of the HTTP request. Set to a String to pass it as an Authorization: Bearer ${token} token (no need to supply the token explicitly for every http method call, it is supposed to be set globally, see below).
  • progress(percent, event) — Use for tracking HTTP request progress (e.g. file upload).
  • onResponseHeaders(headers) – Use for examining HTTP response headers (e.g. Amazon S3 file upload).

Redux module

Once one starts writing a lot of promise/http Redux actions it becomes obvious that there's a lot of copy-pasting and verbosity involved. To reduce those tremendous amounts of copy-pasta "redux module" tool may be used which:

  • Gives access to http.
  • Autogenerates Redux action status events (${actionName}_PENDING, ${actionName}_SUCCESS, ${actionName}_ERROR).
  • Automatically adds Redux reducers for the action status events.
  • Automatically populates the corresponding action status properties (${actionName}Pending: true/false, ${actionName}Error: Error) in Redux state.

For example, the fetchFriends() action from the previous section can be rewritten as:

Before:

// ./actions/friends.js
function fetchFriends(personId, gender) {
  return {
    promise: http => http.get(`/api/person/${personId}/friends`, { gender }),
    events: ['FETCH_FRIENDS_PENDING', 'FETCH_FRIENDS_SUCCESS', 'FETCH_FRIENDS_FAILURE']
  }
}

// ./reducers/friends.js
export default function(state = {}, action = {}) {
  switch (action.type) {
    case 'FETCH_FRIENDS_PENDING':
      return {
        ...state,
        fetchFriendsPending: true,
        fetchFriendsError: null
      }
    case 'FETCH_FRIENDS_SUCCESS':
      return {
        ...state,
        fetchFriendsPending: false,
        friends: action.value
      }
    case 'FETCH_FRIENDS_ERROR':
      return {
        ...state,
        fetchFriendsPending: false,
        fetchFriendsError: action.error
      }
    default
      return state
  }
}

After:

import { ReduxModule } from 'react-website'

const redux = new ReduxModule('FRIENDS')

export const fetchFriends = redux.action(
  'FETCH_FRIENDS',
  (personId, gender) => http => {
    return http.get(`/api/person/${personId}/friends`, { gender })
  },
  // The fetched friends list will be placed
  // into the `friends` Redux state property.
  'friends'
  //
  // Or write it like this:
  // { friends: result => result }
  //
  // Or write it as a Redux reducer:
  // (state, result) => ({ ...state, friends: result })
)

// This is the Redux reducer which now
// handles the asynchronous action defined above.
export default redux.reducer()

Much cleaner.

Also, when the namespace or the action name argument is omitted it is autogenerated, so this

const redux = new ReduxModule('FRIENDS')
...
redux.action('FETCH_ITEM', id => http => http.get(`/items/${id}`), 'item')

could be written as

const redux = new ReduxModule()
...
redux.action(id => http => http.get(`/items/${id}`), 'item')

and in this case redux will autogenerate the namespace and the action name, something like REACT_WEBSITE_12345 and REACT_WEBSITE_ACTION_12345.

Here's a more complex example: a comments section for a blog post page.

redux/blogPost.js

import { ReduxModule } from 'react-website'

const redux = new ReduxModule('BLOG_POST')

// Post comment Redux "action creator"
export const postComment = redux.action(
  // 'POST_COMMENT',
  (userId, blogPostId, commentText) => async http => {
    // The original action call looks like:
    // `dispatch(postComment(1, 12345, 'bump'))`
    return await http.post(`/blog/posts/${blogPostId}/comment`, {
      userId: userId,
      text: commentText
    })
  }
)

// Get comments Redux "action creator"
export const getComments = redux.action(
  // 'GET_COMMENTS',
  (blogPostId) => async http => {
    return await http.get(`/blog/posts/${blogPostId}/comments`)
  },
  // The fetched comments will be placed
  // into the `comments` Redux state property.
  'comments'
  //
  // Or write it like this:
  // { comments: result => result }
  //
  // Or write it as a Redux reducer:
  // (state, result) => ({ ...state, comments: result })
)

// A developer can listen to any event.
// If two string arguments are passed
// then the first one is namespace
// and the second one is the event name
// and the listener will be called "on success".
// If only one string argument is passed
// then it is a raw Redux `action.type`.
redux.on('BLOG_POST', 'CUSTOM_EVENT', (state, action) => ({
  ...state,
  reduxStateProperty: action.value
}))

// This is the Redux reducer which now
// handles the asynchronous actions defined above
// (and also the `handler.on()` events).
// Export it as part of the "main" reducer.
export default redux.reducer()

redux/index.js

// The "main" reducer composed of various reducers.
export { default as blogPost } from './blogPost'
...

The React Component would look like this

import React, { Component } from 'react'
import { connect } from 'react-redux'
import { preload } from 'react-website'
import { connectComments, getComments, postComment } from './redux/blogPost'

// Preload comments before showing the page
// (see "Page preloading" section of this document)
@preload(async ({ dispatch, params }) => {
  // `params` are the URL parameters populated by `react-router`:
  // `<Route path="/blog/:blogPostId"/>`.
  await dispatch(getComments(params.blogPostId))
})
// See `react-redux` documentation on `@connect()` decorator
@connect((state) => ({
  userId: state.user.id,
  comments: state.blogPost.comments
}), {
  postComment
})
export default class BlogPostPage extends Component {
  render() {
    const {
      comments
    } = this.props

    return (
      <div>
        <ul>
          { comments.map(comment => <li>{comment}</li>) }
        </ul>
        {this.renderPostCommentForm()}
      </div>
    )
  }

  renderPostCommentForm() {
    // `params` are the URL parameters:
    // `<Route path="/blog/:blogPostId"/>`.
    const {
      userId,
      params,
      postComment
    } = this.props

    return (
      <button onClick={() => postComment(userId, params.blogPostId, 'text')}>
        Post comment
      </button>
    )
  }
}

Redux module can also handle synchronous actions along with asynchronous ones, should the need arise.

See how
import { ReduxModule } from 'react-website'

const redux = new ReduxModule('NOTIFICATIONS')

// Displays a notification.
//
// The Redux "action" creator is gonna be:
//
// function(text) {
//   return {
//     type    : 'NOTIFICATIONS:NOTIFY',
//     message : formatMessage(text)
//   }
// }
//
// And the corresponding reducer is gonna be:
//
// case 'NOTIFICATIONS:NOTIFY':
//   return {
//     ...state,
//     message: action.message
//   }
//
// Call it as `dispatch(notify(text))`.
//
export const notify = redux.simpleAction(
  'NOTIFY',
  // The action
  (text) => formatMessage(text),
  // The Redux state reducer for the action's return value
  (state, message) => ({ ...state, message }),
  // The Redux reducer above could be also defined as
  // 'message'
)

// This is the Redux reducer which now
// handles the actions defined above.
export default redux.reducer()

HTTP authentication

In order for http utility to send an authentication token as part of an HTTP request (the Authorization: Bearer ${token} HTTP header) the authentication.accessToken() function must be specified in react-website.js.

{
  authentication: {
    accessToken({ getState, path, url, getCookie }) {
      // Optionally check the `path` (or `url`) to make sure that the access token
      // is not leaked to a third party: only send it to own servers.
      // For example, at some point someone may use `http` utility
      // to get some data from a 3rd party endpoint, hence the `url` check.
      // `path` is the URL supplied to `http` utility method call.
      // `url` is the result of `http.transformURL(path)` (if configured).
      // An example of `path`: "/api/method".
      // An example of `url`: "https://server.com/api/method".
      // if (path.indexOf('/') !== 0) {
      //   return
      // }
      return localStorage.getItem('accessToken')
      return getCookie('accessToken')
      return getState().authentication.accessToken
    }
  }
}
Authentication and authorization using access tokens

The accessToken is initially obtained when a user signs in: the web browser sends HTTP POST request to /sign-in API endpoint with { email, password } parameters and gets { userInfo, accessToken } as a response, which is then stored in localStorage (or in Redux state, or in a cookie) and all subsequent HTTP requests use that accessToken to call the API endpoints. The accessToken itself is usually a JSON Web Token signed on the server side and holding the list of the user's priviliges ("roles"). Hence authentication and authorization are completely covered. Refresh tokens are also supported.

This kind of an authentication and authorization scheme is self-sufficient and doesn't require "restricting" any routes: if a route's @preload() uses http utility for querying an API endpoint then this API endpoint must check if the user is signed in and if the user has the necessary priviliges. If yes then the route is displayed. If not then the user is redirected to either a "Sign In Required" page or "Access Denied" page.

A real-world (advanced) example for handling "Unauthenticated"/"Unauthorized" errors happening in @preload()s and during http calls:

./react-website.js

{
  ...,
  onError(error, { path, url, redirect, getState, server }) {
    // Not authenticated
    if (error.status === 401) {
      return handleUnauthenticatedError(error, url, redirect);
    }
    // Not authorized
    if (error.status === 403) {
      return redirect('/unauthorized');
    }
    // Not found
    if (error.status === 404) {
      return redirect('/not-found');
    }
    // Redirect to a generic error page in production
    if (process.env.NODE_ENV === 'production') {
      // Prevents infinite redirect to the error page
      // in case of overall page rendering bugs, etc.
      if (path !== '/error') {
        // Redirect to a generic error page
        return redirect(`/error?url=${encodeURIComponent(url)}`);
      }
    } else {
      // Report the error
      console.error(`Error while preloading "${url}"`);
      console.error(error);
    }
  },
  http: {
    onError(error, { path, url, redirect, dispatch, getState }) {
      // JWT token expired, the user needs to relogin.
      if (error.status === 401) {
        return handleUnauthenticatedError(error, url, redirect);
      }
    },
    ...
  }
}

function handleUnauthenticatedError(error, url, redirect) {
  // Prevent double redirection to `/unauthenticated`.
  // (e.g. when two parallel `Promise`s load inside `@preload()`
  //  and both get Status 401 HTTP Response)
  if (typeof window !== 'undefined' && window.location.pathname === '/unauthenticated') {
    return;
  }
  let unauthenticatedURL = '/unauthenticated';
  let parametersDelimiter = '?';
  if (url !== '/') {
    unauthenticatedURL += `${parametersDelimiter}url=${encodeURIComponent(url)}`;
    parametersDelimiter = '&';
  }
  switch (error.message) {
    case 'TokenExpiredError':
      return redirect(`${unauthenticatedURL}${parametersDelimiter}expired=✔`);
    case 'AuthenticationError':
      return redirect(`${unauthenticatedURL}`);
    default:
      return redirect(unauthenticatedURL);
  }
}

HTTP request URLs

All URLs queried via http utility are supposed to be relative ones (e.g. /api/users/list) for convenience. In order to transform these convenient relative URLs into real ones there are two approaches built-in.

The old-school approach is for people using a proxy server.

In this case all client-side HTTP requests will still query relative URLs which are gonna hit the proxy server and the proxy server will route them to the API service. And on server side it's gonna query the proxy server by an absolute URL (there is no notion of "relative URLs" on the server side) therefore the proxy host and port need to be configured in webpage rendering service options.

const server = webpageServer(settings, {
  proxy: {
    host: '192.168.0.1',
    port: 3000,
    // (enable for HTTPS protocol)
    // secure: true
  }
})

The modern approach is to query API by an absolute URL (through CORS) in a cloud. In this case all URLs are transformed from relative ones into absolute ones by the http.transformURL(path) parameter configured in react-website.js.

{
  http: {
    transformURL: path => `https://my-api.cloud.com${path}`
  }
}

File upload

The http utility will also upload files if they're passed as part of data (see example below). The files passed inside data must have one of the following types:

  • In case of a File it will be a single file upload.
  • In case of a FileList with a single File inside it would be treated as a single File.
  • In case of a FileList with multiple Files inside a multiple file upload will be performed.
  • In case of an <input type="file"/> DOM element all its .files will be taken as a FileList parameter.

File upload progress can be metered by passing progress option as part of the options .

See example
// React component
class ItemPage extends React.Component {
  render() {
    return (
      <div>
        ...
        <input type="file" onChange={this.onFileSelected}/>
      </div>
    )
  }

  // Make sure to `.bind()` this handler
  onFileSelected(event) {
    const file = event.target.files[0]

    // Could also pass just `event.target.files` as `file`
    dispatch(uploadItemPhoto(itemId, file))

    // Reset the selected file
    // so that onChange would trigger again
    // even with the same file.
    event.target.value = null
  }
}

// Redux action creator
function uploadItemPhoto(itemId, file) {
  return {
    promise: http => http.post(
      '/item/photo',
      { itemId, file },
      { progress(percent) { console.log(percent) } }
    ),
    events: ['UPLOAD_ITEM_PHOTO_PENDING', 'UPLOAD_ITEM_PHOTO_SUCCESS', 'UPLOAD_ITEM_PHOTO_FAILURE']
  }
}

JSON Date parsing

By default, when using http utility all JSON responses get parsed for javascript Dates which are then automatically converted from Strings to Dates. This is convenient, and also safe because such date Strings have to be in a very specific ISO format in order to get parsed (year-month-dayThours:minutes:seconds[timezone], e.g. 2017-12-22T23:03:48.912Z), but if someone still prefers to disable this feature and have their stringified dates untouched then there's the parseDates: false flag in the configuration to opt-out of this feature.

Snapshotting

Server-Side Rendering is good for search engine indexing but it's also heavy on CPU not to mention the bother of setting up a Node.js server itself and keeping it running.

In many cases data on a website is "static" (doesn't change between redeployments), e.g. a personal blog or a portfolio website, so in these cases it will be beneficial (much cheaper and faster) to host a statically generated version a website on a CDN as opposed to hosting a Node.js application just for the purpose of real-time webpage rendering. In such cases one should generate a static version of the website by snapshotting it on a local machine and then host the snapshotted pages in a cloud (e.g. Amazon S3) for a very low price.

Snapshotting instructions

First run the website in production mode (for example, on localhost).

Then run the following Node.js script which is gonna snapshot the currently running website and put it in a folder which can then be hosted anywhere.

# If the website will be hosted on Amazon S3
npm install s3 --save
import path from 'path'

import {
  // Snapshots website pages.
  snapshot,
  // Uploads files.
  upload,
  // Uploads files to Amazon S3.
  S3Uploader,
  // Copies files/folders into files/folders.
  // Same as Linux `cp [from] [to]`.
  copy,
  // Downloads data from a URL into an object
  // of shape `{ status: Number, content: String }`.
  download
} from 'react-website/static-site-generator'

import configuration from '../configuration'

// Temporary generated files path.
const generatedSitePath = path.resolve(__dirname, '../static-site')

async function run() {
  // Snapshot the website.
  await snapshot({
    // The host and port on which the website
    // is currently running in production mode.
    // E.g. `localhost` and `3000`.
    host: configuration.host,
    port: configuration.port,
    pages: await generatePageList(),
    outputPath: generatedSitePath
  })

  // Copy assets (built by Webpack).
  await copy(path.resolve(__dirname, '../build/assets'), path.resolve(generatedSitePath, 'assets'))
  await copy(path.resolve(__dirname, '../robots.txt'), path.resolve(generatedSitePath, 'robots.txt'))

  // Upload the website to Amazon S3.
  await upload(generatedSitePath, S3Uploader({
    bucket: confiugration.s3.bucket,
    accessKeyId: configuration.s3.accessKeyId,
    secretAccessKey: configuration.s3.secretAccessKey,
    region: configuration.s3.region
  }))

  console.log('Done');
}

run().catch((error) => {
  console.error(error)
  process.exit(1)
})

// Get the list of all page URLs.
async function generatePageList() {
  const pages = [
    '/',
    '/about',
    // Error pages need a `status` property
    // to indicate that it shouldn't throw on such errors
    // and should proceed with snapshotting the next pages.
    { url: '/unauthenticated', status: 401 },
    { url: '/unauthorized', status: 403 },
    { url: '/not-found', status: 404 },
    { url: '/error', status: 500 }
  ]

  // (optional) Add some dynamic page URLs, like `/items/123`.

  // Query the database for the list of items.
  const { status, content } = JSON.parse(await download(`https://example.com/api/items`))

  if (status !== 200) {
    throw new Error('Couldn\'t load items')
  }

  // Add item page URLs.
  const items = JSON.parse(content)
  return pages.concat(items.map(item => `/items/${item.id}`))
}

The snapshot() function snapshots the list of pages to .html files and then the upload() function uploads them to the cloud (in this case to Amazon S3). The snapshot() function also snapshots a special base.html page which is an empty page that should be used as the "fallback", i.e. the cloud should respond with base.html file contents when the file for the requested URL is not found: in this case base.html will see the current URL and perform all the routing neccessary on the client side to show the correct page. If the snapshot() function isn't passed the list of pages to snapshot (e.g. if pages argument is null or undefined) then it will only snapshot base.html. The static website will work with just base.html, the only point of snapshotting other pages is for Google indexing.

The snapshotting approach works not only for classical web "documents" (a blog, a book, a portfolio, a showcase) but also for dynamic applications. Consider an online education portal where users (students) can search for online courses and the prices are different for each user (student) based on their institution. Now, an online course description itself is static (must be indexed by Google) and the actual course price is dynamic (must not be indexed by Google).

The solution is to add two @preload()s for the course page: one for static data (which runs while snapshotting) and another for dynamic data (which runs only in a user's web browser).
import React, { Component } from 'react'
import { preload } from 'react-website'

@preload(async ({ dispatch }) => await dispatch(loadCourseInfo()))
@preload(async ({ dispatch }) => await dispatch(loadCoursePrice()), { client: true })
export default class Course extends Component {
  ...
}

In this example loadCourseInfo() will be executed while snapshotting and therefore course info will be present on the snapshotted page. But course price won't be present on the snapshotted page because it's being loaded inside @preloadClient() which only gets called in a user's web browser. When a user opens the course page in his web browser it will show the snapshotted page with course info with a "loading" spinner on top of it as it is loading the course price. After the course price has been loaded the "loading" spinner disappears and the user sees the fully rendered course page.

Page HTTP response status code

To set a custom HTTP response status code for a specific route set the status property of that <Route/>.

export default (
  <Route path="/" Component={Layout}>
    <Route Component={Home}/>
    <Route path="blog"  Component={Blog}/>
    <Route path="about" Component={About}/>
    <Route path="*"     Component={PageNotFound} status={404}/>
  </Route>
)

Setting <title/> and tags

Use @meta(state) => ...) decorator for adding <title/> and <meta/> tags:

import { meta } from 'react-website'

@meta(state) => ({
  // `<meta property="og:site_name" .../>`
  site_name: 'International Bodybuilders Club',

  // Webpage `<title/>` will be replaced with this one
  // and also `<meta property="og:title" .../>` will be added.
  title: `${state.user.name}`,

  // `<meta property="og:description" .../>`
  description: 'Muscles',

  // `<meta property="og:image" .../>`
  image: 'https://cdn.google.com/logo.png',

  // `<meta property="og:audio" .../>`
  audio: '...',

  // `<meta property="og:video" .../>`
  video: '...',

  // `<meta property="og:locale" content="ru_RU"/>`
  locale: state.user.locale,

  // `<meta property="og:locale:alternate" content="en_US"/>`
  // `<meta property="og:locale:alternate" content="fr_FR"/>`
  locales: ['ru_RU', 'en_US', 'fr_FR'],

  // `<meta property="og:url" .../>`
  url: 'https://google.com/',

  // `<meta property="og:type" .../>`
  type: 'profile',

  // `<meta charset="utf-8"/>` tag is added automatically.
  // The default "utf-8" encoding can be changed
  // by passing custom `charset` parameter.
  charset: 'utf-16',

  // `<meta name="viewport" content="width=device-width, initial-scale=1.0"/>`
  // tag is added automatically
  // (prevents downscaling on mobile devices).
  // This default behaviour can be changed
  // by passing custom `viewport` parameter.
  viewport: '...',

  // All other properties will be transformed directly to
  // either `<meta property="{property_name}" content="{property_value}/>`
  // or `<meta name="{property_name}" content="{property_value}/>`
}))
export default class Page extends React.Component {
  ...
}

@meta() decorator discards all other <meta/> set by any other means, e.g. if there are any <meta/> tags in index.html template then all of them will be dicarded if using @meta() decorator so don't mix @meta() decorator with <meta/> tags inserted manually into index.html.

Get current location

Inside @preload(): use the location parameter.

Anywhere in a React component: use the found property in Redux state.

@connect(({ found }) => ({
  location: found.resolvedMatch.location,
  params: found.resolvedMatch.params
}))

Changing current location

Dispatch goto/redirect Redux action to change current location (both on client and server).

import { goto, redirect } from 'react-website'
import { connect } from 'react-redux'

// Usage example
// (`goto` navigates to a URL while adding a new entry in browsing history,
//  `redirect` does the same replacing the current entry in browsing history)
@connect(state = {}, { goto, redirect })
class Page extends Component {
  handleClick(event) {
    const { goto, redirect } = this.props
    goto('/items/1?color=red')
    // redirect('/somewhere')
  }
}

If the current location needs to be changed while still staying at the same page (e.g. a checkbox has been ticked and the corresponding URL query parameter must be added), then use dispatch(pushLocation(location)) or dispatch(replaceLocation(location)) Redux actions.

import { pushLocation, replaceLocation } from 'react-website'

@connect(() => ({
  ...
}), {
  pushLocation
})
class Page extends Component {
  onSearch(query) {
    const { pushLocation } = this.props

    pushLocation({
      pathname: '/'
      query: {
        query
      }
    })
  }
}

To go "Back"

import { goBack } from 'react-website'

@connect(() => ({
  ...
}), {
  goBack
})
class Page extends Component {
  render() {
    const { goBack } = this.props

    return (
      <button onClick={goBack}>
        Back
      </button>
    )
  }
}

If someone prefers interacting with found router directly instead then it is available on all pages as a router property, or via @withRouter decorator.

import React from 'react'
import { withRouter } from 'found'

@withRouter
export default class Component extends React.Component {
  render() {
    const { router } = this.props
    ...
  }
}

Monitoring

For each page being rendered stats are reported if stats() parameter is passed as part of the rendering service settings.

{
  ...

  stats({ url, route, time: { initialize, preload, total } }) {
    if (total > 1000) { // in milliseconds
      db.query('insert into server_side_rendering_stats ...')
    }
  }
}

The arguments for the stats() function are:

  • url — the requested URL (without the protocol://host:port part)
  • route — the route path (e.g. /user/:userId/post/:postId)
  • time.initialize — server side initialize() function execution time (if defined)
  • time.preload — page preload time
  • time.total — total time spent preloading and rendering the page

Rendering a complex React page (having more than 1000 components) takes about 30ms (as of 2017).

One could also set up overall Server Side Rendering performance monitoring using, for example, StatsD
{
  ...

  stats({ url, route, time: { initialize, preload, total } }) {
    statsd.increment('count')

    statsd.timing('initialize', initialize)
    statsd.timing('@preload()', preload)
    statsd.timing('total', total)

    if (total > 1000) { // in milliseconds
      db.query('insert into server_side_rendering_stats ...')
    }
  }
}

Where the metrics collected are

  • count — rendered pages count
  • initialize — server side initialize() function execution time (if defined)
  • preload — page preload time
  • time - total time spent preloading and rendering the page

Speaking of StatsD itself, one could either install the conventional StatsD + Graphite bundle or, for example, use something like Telegraf + InfluxDB + Grafana.

Telegraf starter example:

# Install Telegraf (macOS).
brew install telegraf
# Generate Telegraf config.
telegraf -input-filter statsd -output-filter file config > telegraf.conf
# Run Telegraf.
telegraf -config telegraf.conf
# Request a webpage and see rendering stats being output to the terminal.

Webpack HMR

Webpack's Hot Module Replacement (aka Hot Reload) works for React components and Redux reducers and Redux action creators (it just doesn't work for page @preload()s).

HMR setup for Redux reducers is as simple as adding store.hotReload() (as shown below). For enabling HMR on React Components (and Redux action creators) use react-hot-loader:

application.js

import { render } from 'react-website'
import settings from './react-website'

render(settings).then(({ store, rerender }) => {
  if (module.hot) {
    module.hot.accept('./react-website', () => {
      rerender()
      // Update reducer
      store.hotReload(settings.reducers)
    })
  }
})

Container.js

import React from 'react'
import { Provider } from 'react-redux'
import { hot } from 'react-hot-loader'

function Container({ store, children }) {
  return (
    <Provider store={ store }>
      { children }
    </Provider>
  )
}

export default hot(module)(Container)

.babelrc

{
  "presets": [
    "react",
    ["env", { modules: false }],
  ],

  "plugins": [
    // `react-hot-loader` Babel plugin
    "react-hot-loader/babel"
  ]
}

./src/index.js

// An ES6 polyfill is required for `react-hot-loader`.
require('babel-polyfill')
...

Then start webpack-dev-server or webpack-serve with --hot option.

WebSocket

websocket() helper sets up a WebSocket connection.

import { render } from 'react-website'
import websocket from 'react-website/websocket'

render(settings).then(({ store }) => {
  websocket({
    host: 'localhost',
    port: 80,
    // secure: true,
    store,
    token: localStorage.getItem('token')
  })
})

If token parameter is specified then it will be sent as part of every message (providing support for user authentication).

How to use WebSocket

WebSocket will autoreconnect (with "exponential backoff") emitting open event every time it does.

After the websocket() call a global websocket variable is created exposing the following methods:

  • listen(eventName, function(event, store))
  • onOpen(function(event, store)) – is called on open event
  • onClose(function(event, store)) – is called on close event
  • onError(function(event, store)) – is called on error event (close event always follows the corresponding error event)
  • onMessage(function(message, store))
  • send(message)
  • close()

The store argument can be used to dispatch() Redux "actions".

websocket.onMessage((message, store) => {
  if (message.command) {
    switch (message.command) {
      case 'initialized':
        store.dispatch(connected())
        return console.log('Realtime service connected', message)
      case 'notification':
        return alert(message.text)
      default:
        return console.log('Unknown message type', message)
    }
  }
})

websocket.onOpen((event, store) => {
  websocket.send({ command: 'initialize' })
})

websocket.onClose((event, store) => {
  store.dispatch(disconnected())
})

The global websocket object also exposes the socket property which is the underlying robust-websocket object (for advanced use cases).

As for the server-side counterpart I can recommend using uWebSockets

import WebSocket from 'uws'

const server = new WebSocket.Server({ port: 8888 })

const userConnections = {}

server.on('connection', (socket) => {
  console.log('Incoming WebSocket connection')

  socket.sendMessage = (message) => socket.send(JSON.stringify(message))

  socket.on('close', async () => {
    console.log('Client disconnected')

    if (socket.userId) {
      userConnections[socket.userId].remove(socket)
    }
  })

  socket.on('message', async (message) => {
    try {
      message = JSON.parse(message)
    } catch (error) {
      return console.error(error)
    }

    try {
      switch (message.command) {
        case 'initialize':
          // If a user connected (not a guest)
          // then store `userId` for push notifications.
          // Using an authentication token here
          // instead of simply taking `userId` out of the `message`
          // because the input can't be trusted (could be a hacker).
          if (message.userAuthenticationToken) {
            // (make sure `socket.userId` is a `String`)
            // The token could be a JWT token (jwt.io)
            // and `authenticateUserByToken` function could
            // check the token's authenticity (by verifying its signature)
            // and then extract `userId` out of the token payload.
            socket.userId = authenticateUserByToken(message.userAuthenticationToken)

            if (!userConnections[socket.userId]) {
              userConnections[socket.userId] = []
            }

            userConnections[socket.userId].push(socket)
          }

          return socket.sendMessage({
            command: 'initialized',
            data: ...
          })

        default:
          return socket.sendMessage({
            status: 404,
            error: `Unknown command: ${message.command}`
          })
      }
    } catch (error) {
      console.error(error)
    }
  })
})

server.on('error', (error) => {
  console.error(error)
})

// Also an HTTP server is started and a REST API endpoint is exposed
// which can be used for pushing notifications to clients via WebSocket.
// The HTTP server must only be accessible from the inside
// (i.e. not listening on an external IP address, not proxied to)
// otherwise an attacker could push any notifications to all users.
// Therefore, only WebSocket connections should be proxied (e.g. using NginX).
httpServer().handle('POST', '/notification', ({ to, text }) => {
  if (userConnections[to]) {
    for (const socket of userConnections[to]) {
      socket.sendMessage({
        command: 'notification',
        text
      })
    }
  }
})

Feature: upon receiving a message (on the client side) having a type property defined such a message is dispatch()ed as a Redux "action" (this can be disabled via autoDispatch option). For example, if { type: 'PRIVATE_MESSAGE', content: 'Testing', from: 123 } is received on a websocket connection then it is automatically dispatch()ed as a Redux "action". Therefore, the above example could be rewritten as

// Server side (REST API endpoint)
socket.sendMessage({
  type: 'DISPLAY_NOTIFICATION',
  text
})

// Client side (Redux reducer)
function reducer(state, action) {
  switch (action.type) {
    case 'DISPLAY_NOTIFICATION':
      return {
        ...state,
        notifications: state.notifications.concat([action.text])
      }
    default:
      return state
  }
}

Server-Side Rendering and bundlers

If the application is being built with a bundler (most likely Webpack) and Server-Side Rendering is enabled then make sure to build the server-side code with the bundler too so that require() calls for assets (images, styles, fonts, etc) inside React components don't break (see universal-webpack, for example).

Code splitting

Code splitting is supported. See README-CODE-SPLITTING

Advanced

At some point in time this README became huge so I extracted some less relevant parts of it into README-ADVANCED (including the list of all possible settings and options). If you're a first timer then just skip that one - you don't need it for sure.

License

MIT