mummy

Tools for using Zombie.js with Hapi.

npm install mummy

Overview

mummy is a browser extension for the Zombie.js headless browser testing framework that allows automatic binding of Browser instances to Hapi servers.

var Browser = require("zombie");
var Mummy   = require("mummy");
var browser;

Browser.extend(new Mummy(server));
browser = new Browser();

browser.visit("/")
.then(function () {
	. . .
});

Extending the Browser API

Zombie provides an extension API that allows all new Browser objects to be augmented with additional functionality. Using this approach, mummy will cause all Browser objects to direct their requests to the wrapped Hapi server. Only requests with URLs matching the hostname and port of one of the connections on the server will be injected. All other requests will be processed normally.

Wrapping a Single Browser

Alternatively, mummy can wrap a single Browser instance as follows:

var Browser = require("zombie");
var Mummy   = require("mummy");

var browser = new Browser();
Mummy.embalm(server, browser);

Event Loop Management

Zombie does not continually evaluate the browser event loop. Rather, it runs the event loop in relatively short bursts at keep times to drive browser functionality. In many cases this is fine. However, sometimes it is useful for test cases to queue events after the initial page load or separate from a user interaction (i.e. browser.pressButton(...)). In these cases it is useful to run the event loop for the entire duration of the test. This can be done with something like the following:

describe("an async interaction", function () {
	before(function() {
		browser.runner.start();
		browser.window.doSomethingInTheFuture();
	});

	after(function () {
		browser.runner.stop();
	});

	it("eventually does something", ...);
});

Raw HTTP Requests

mummy also provides the ability to make "raw" HTTP requests to wrapped servers. This can be useful for testing REST APIs. For example:

var browser = new Browser();

browser.http({ method : "GET", url : "/" }).then(function (response) {
	expect(response.statusCode).to.equal(200);
});

API

Mummy(server)

• server -- a Hapi Server instance to create a Browser extension for.

Returns a Browser extension suitable for passing to Zombie's Browser.extend().

Mummy.embalm(server, browser)

• server -- a Hapi Server instance to inject requests into.
• browser -- a Browser instance to augment with request redirection.

Returns the original Browser instance after it has been augmented to redirect requests to the server.

browser.credentials.set(credentials)

• credentials -- an object containing simulated authentication information

Update the browser state to bypass the normal authentication strategies when requests are sent to Hapi. See the Hapi documentation for more details.

browser.credentials.clear()

Clear any browser credentials. This will cause normal authentication flows to be used for requests sent to Hapi.

browser.runner.start()

Causes Zombie to continually execute the event loop until manually stopped. This is useful for test suites that schedule events after the initial page load. Returns a promise that is resolved when the event loop has stopped.

browser.runner.stop()

Stops the event loop when it has been started with browser.runner.start().

browser.http(options, [callback])

• options -- an object representing the request.
• callback -- Optional a callback function receiving arguments of the form (error, response) depending on if the response is successful. If not provided, a promise is returned.

The options hash can include the following:

• method -- the HTTP request method. Defaults to "GET".
• url -- the path or URL to request. Defaults to "/".
• headers -- an object defining request headers.

Performs a "raw" HTTP request. The server.inject() method from Hapi is used for requests to wrapped servers while Request is used for requests to URLs not belonging to wrapped servers. Additional options are supported for both (see the respective docs for details).