npm package discovery and stats viewer.

Discover Tips

  • General search

    [free text search, go nuts!]

  • Package details

    pkg:[package-name]

  • User packages

    @[username]

Sponsor

Optimize Toolset

I’ve always been into building performant and accessible sites, but lately I’ve been taking it extremely seriously. So much so that I’ve been building a tool to help me optimize and monitor the sites that I build to make sure that I’m making an attempt to offer the best experience to those who visit them. If you’re into performant, accessible and SEO friendly sites, you might like it too! You can check it out at Optimize Toolset.

About

Hi, 👋, I’m Ryan Hefner  and I built this site for me, and you! The goal of this site was to provide an easy way for me to check the stats on my npm packages, both for prioritizing issues and updates, and to give me a little kick in the pants to keep up on stuff.

As I was building it, I realized that I was actually using the tool to build the tool, and figured I might as well put this out there and hopefully others will find it to be a fast and useful way to search and browse npm packages as I have.

If you’re interested in other things I’m working on, follow me on Twitter or check out the open source projects I’ve been publishing on GitHub.

I am also working on a Twitter bot for this site to tweet the most popular, newest, random packages from npm. Please follow that account now and it will start sending out packages soon–ish.

Open Software & Tools

This site wouldn’t be possible without the immense generosity and tireless efforts from the people who make contributions to the world and share their work via open source initiatives. Thank you 🙏

© 2024 – Pkg Stats / Ryan Hefner

external

v1.0.0

Published

Fittings for BigPipe which allows you to use the BigPipe as a third party/external content provider.

Downloads

371

Readme

external

External is a dual purpose library. It ships with a client-side framework renders third party or external pages in the most optimal way as possible. This is done using various of techniques:

  • The payload is downloaded using a fully async streaming XHR request. This way we can continuously update and render our placeholder while data flows and therefor reducing the time to render.
  • All assets of the page are loaded async, this includes the CSS.
  • The received client code is wrapped before execution so client code can re-use our dependencies while keeping a sandboxed approach.
  • While the client was specifically written for the BigPipe framework it should work against any back-end as long as it returns the same data structure.
  • Templates are rendered using React so it's easy to compose and update.

But we also ship with a server-side framework implementation for BigPipe which makes it possible to serve the client and automatically format all the output in the expected HTML structure.

Table of Contents

Installation

The client-side component is composed from various of tiny modules and can be build using Browserify. It can be build-in to other browserify components by simply requiring the external module in your client-code.

The server side part of this framework can be installed through npm:

npm install external

In addition to providing a browserify-able client-side script there is also a compiled version of this code which lives in the dist folder called extern.js. This pre-compiled library exposes it self using the Extern global and therefor does not introduced require statement as globals. In all the code examples in documentation we assume that you have an Extern global. If you use the dist build you can skip the following example:

var Extern = require('extern');

Building

If you want to generate new stand alone bundles of the Extern library you can run our prepublish and dev scripts using the npm run command. These commands do assume that you've installed the devDependencies of this project. To generate a new production build, dist/extern.min.js run:

npm run prepublish

As this is a prepublish script, it means that every release to npm will have the dist/extern.js included. So if browserify isn't your think, you can just include the extern/dist/extern.min.js instead.

To generate an un-minified build for development purposes you can run:

npm run dev

This will generate a new dist/extern.dev.js file.

Serving

Now that you know how to install it and what type of bundles there are you can decide how to serve the library. When this module is used as plugin in BigPipe it will automatically serve the browserify and plugin combined bundle from:

http(s)://domain.com/extern.js

We also mount our dist folder on the server so the static assets in this folder can also be served:

http(s)://domain.com/extern.min.js

Now that you've picked your build, and know how the files are served you can simply put the script tag in your page and your ready to display external pages/apps.

<script src="https//yourdomain.com/extern.min.js"></script>
<script>
var extern = new Extern(document.body, 'http://whateverurlyouwantousehere.com/path/name');
</script>

Listening

The easiest way to have Extern load your remote pages is by using the Extern.listen method in combination with the rel="extern" attributes on <a> elements:

<a href="http://my-remote-server.com/optional/path" rel="extern">Remote</a>

The Extern.listen method will gather all <a> elements and search for a rel that is set to extern and uses the set href of the element as URL that needs to be remotely loaded.

Extern.listen(document.body, {});

Extern

The following options are supported:

  • timeout Timeout for dependency loading. If assets take longer we should render and error template instead. The timeout is in milliseconds.
  • document Reference to the document global can be useful if assets need to be loaded in iframes instead of the global document.
  • className If a link has this className we will automatically load it in the placeholder. This className will also automatically be add and removed once the link is clicked. Defaults to extern-loads.
var extern = new Extern('http://my.example.com/page', document.body, {
  timeout: 10000
});

Events

The returned extern instance is actually an EventEmitter3 instance so you can listen to the various of events that we're emitting:

  • error Emitted when something went so horribly wrong that we decided to show the error template. This event receives the actual error as argument.
  • done The streaming XHR is finished with loading.
  • name:render Called when a fragment is about to render in to the placeholder. The name part in the event should be name of the fragment you want to listen for.
  • name:loaded All the assets are loaded for the given placeholder name.

Wrapping

The client code for each fragments are loaded through an XHR connection. This way we can safely executed third party code by wrapping the execution in a try/catch statement. But not only does this allow us to wrap code, it also allows us to introduce variables in the function. The following variables are introduced as "globals":

  • React, This is the react/addons reference.
  • require, Reference to our require statement so you can re-use all the bundled things.

API

The following properties and methods are exposed on the Extern instance.

Extern.listen

Exposed on the constructor

Scan the current document for all <a rel="extern"> elements and attach click listeners to it so we can automatically update the supplied placeholder with the contents of the set URL. This method accepts one argument and that is the placeholder DOM element where all pages should loaded in

Extern.listen(document.body);

Extern.merge

Exposed on the constructor

Merge the object of the second argument in to the first argument. It returns the fully merged first argument.

var x = Extern.merge({ foo: 'foo' }, { bar: 'bar' });

Extern.requests

Exposed on the constructor

A reference to the requests module that we're using for our XHR requests.

var requests = Extern.requests.

See unshiftio/requests for more information.

BigPipe

This library ships with a custom Fittings framework implementation for BigPipe which allows us to control how everything is processed inside of BigPipe. Adding it to your BigPipe instance is just as simple as passing a custom framework option while creating a new instance:

'use strict';

var BigPipe = require('bigpipe')
  , Extern = require('external');

var app = BigPipe.createServer({
  framework: Extern,
  port: 8080
});

But the framework can also be set after the construction using the framework method:

app.framework(Extern);

Please do note that the current Fittings implentation is in the BigPipe master branch but will out in the release that follows 0.9

Once the fittings are installed on the application, it will start spitting out responses based on the specified Wire Format below. The processing instructions can be found in the instructions folder in the root of this repository. But before fiddling with these files I would suggest giving the README.md of Fittings a read so you know how the data formatting works.

Wire Format

In order to have the broadest support within this framework we came up with a dedicated wire-format in order to have the server-side and client-side components interact with each other. While this wire-format is mostly catered to the needs of an application that is build using the BigPipe framework it should be relatively easy to produce exactly the same output in different frameworks and programming languages. This wire format is also required in order to make streaming data as simple as possible as we can trigger buffer flushes based on this.

The format that we're using is \u1337 separated JSON. Every time we encounter the \u1337 character on the client-side we assume it's the end of chunk that requires processing. The JSON payload that is send should contain the following properties:

  • _id A unique id for the payload that is flushed.
  • name Name of the payload that is flushed. This is used to track potential child->parent references throughout the flushed payload.
  • details An object that contains:
    • js Array with path names for the JavaScript files that need to be loaded on the page. We will automatically prepend the server address to these assets.
    • css Array with path names for the CSS files that need to be loaded on the page. We will automatically prepend the server address to these assets.
  • state Additional state that will be spread on the component when we render it.
  • template An initial HTML template that should be rendered in the given placeholder.

CSS Assets

In order to be able to load CSS assets fully async in every browser we need to know when the styles are applied. This is done by Extern client by adding a DOM element to the page that has an id attribute which contains _ and the filename of the asset that is being downloaded (#_yourfilename). We therefor require that the CSS file contains CSS selector and sets the height property to 42px. This allows us to poll the element for height changes to know when the CSS is fully loaded. So if we have a file called 1aFafa801jz09.css it should have the following selector in the source:

#_1aFafa801jz09 { height: 42px }

License

This project has been released under the MIT license, see [LICENSE].