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

kangaroo

v0.0.2

Published

"JSON API testing without the fuss"

Downloads

4

Readme

Kangaroo

JSON API testing without the fuss.

Example Test

suite("Invites", function() {

  before(function(driver) {
    driver
      .as("admin")
      .POST("/user", {handle: "mia", password: "abc123"})
      .wait()
      .introduce("mia")
      .POST("/auth", {handle: "mia", password:"abc123"});
  });

  test("Send an invite and respond to it",

    step("Mia sends an invite to Ben", function(driver) {
      driver
        .as("mia")
        .POST("/invites", {email: "[email protected]"})
        .expect(200, {
          email: "[email protected]",
          code: /[a-f1-9]{32}/,
          status: "pending"})
        .stash("invite");
    }),

    step("Mia can't send the same invite again", function(driver) {
      driver
        .as("mia")
        .POST("/invites", {to: ":invite.email"})
        .expect(400);
    }),

    branch(
      step("Ben accepts the invite", function(driver) {
        driver
          .introduce("ben")
          .POST("/invites/:invite.code/accept")
          .expect(200)
          .as("mia")
          .GET("/invites?status=accepted")
          .until(200, {$length: 1})
      })
    ),

    branch(
      step("Ben declines the invite", function(driver) {
        driver
          .introduce("ben")
          .POST("/invites/:invite.code/decline"})
          .expect(200)
          .as("mia")
          .GET("/invites?status=accepted")
          .never(200, {$length: 1})
      })
    ),

  ));

  // more invite tests can be added to the suite here...

});

Get Started

$ npm install -g kangaroo
$ roo --help
$ cat > kangaroo_setup.js
var roo = require("kangaroo")
roo.before( function(driver) {
  driver
    .config({
      requestEndpoint: "http://localhost:3000",
    })
});
$ roo tests/*

Generate a trace to use as API documentation

$ roo tests/* -o ../docs/transcripts

This dumps a JSON trace of all activity, organized by the test and step descriptions. Our team uses some trivial templates to render it, but that's not comitted to this repo yet.

Another testing framework? Why?

Kangaroo is only for testing JSON APIs. That's it. This focus has some benefits.

Streamlined API handling

All tests are passed a driver that manages cookies for multiple users and makes stringing together many API calls while checking expectations pretty easy. The driver also helps you deal with lags or eventual consistency: you just replace expect(...) with until(...) and the driver will poll instead of doing a single request.

Generated documentation

Kangaroo can trace your API requests, dumping a JSON document containing all API activity organized by tests. It is easy to render this dump and our team uses it as our API documentation.

Long tests, organized

Similar to other automated test harnesses, kangaroo allows you to break your tests up using the suite and test key words. But kangaroo adds two more directives: step and branch.

Step and branch are inspired by how use cases are structured. Like use cases, high level integration tests tend to be made up of several steps, where later steps are dependent on the success of earlier steps. This is quite different from unit tests, which, ideally, are short and test exactly one thing in isolation.

So with kangaroo you can, optionally, break your test into steps. (Branches are a bit experimental. When you use branches, kangaroo identifies all the unique paths through the test and runs each path in isolation, i.e., creates a fresh driver for each.)

Configuration example:

A config file named kangaroo_setup.js should be place somewhere above the folder that contains your test files. kangaroo.js starts in the folder of your test files, then traverses to root looking for it. Here's an example from a real project:

"use strict";

var roo = require("kangaroo"),
  _ = require("lodash"),
  assert = require("assert");

// Make these global for convenience.  Not required.
global.test = roo.test;
global.step = roo.step;
global.branch = roo.branch;
global.before = roo.before;
global.after = roo.after;

require("http").globalAgent.maxSockets = 20;

// If no expectation is specified, default to expecting a 2xx code
function defaultExpectation(result) {
  assert(
    [200, 201, 202, 203, 204].indexOf(result.statusCode) !== -1,
    "Expectd 2xx status code by default, but got: " + result.statusCode +
    "\nResponse Body:\n"+JSON.stringify(result.json, null, 4)
  );
  return true;
}

// Run this before every path of every suite.
roo.before( function(driver) {
  driver
    .config({
      requestEndpoint: "http://localhost:3100/api",
      defaultExpectation: defaultExpectation
    })
    .introduce("admin")
    .GET("/test/reset_elastic_search")
    .GET("/test/reset_database")
    .GET("/test/reset_caches/")
    .wait()
    .POST("/auth/form", {"handle": "roboto", "password": "abc1234"})
    .stash("admin");
});

Running kangaroo

$ roo --help
$ roo tests/*

Reference

kangaroo

I recommend globalizing the kangaroo functions in your kangaroo_setup.js.

suite("description", function() {

  before(function(driver) {
    //...
  });

  after(function(driver) {
    //...
  });

  //simple test
  test("description", function(driver) {
    //...
  });

  //multi step test
  test("description",
    step( "description", function(driver) {
      //...
    }),
    step( "description", function(driver) {
      //...
    })
  );
});

driver

driver github repo

The driver makes it easy to call your api and check expectations.

Additionally, a driver manages two very useful pieces of state:

  1. user sessions: http cookie collections tied to user aliases
  2. the stash: responses you save to use in later requests

.introduce(name)

Introduce an actor. Under the hood, creates a new cookie collection, assigns it to that name, and sets it as the current cookie colleciton for subsequent requests.

.as(name)

Switch the current actor. Under the hood, this just switches the current cookie collection. Must intrdoduce an actor first.

http methods

   .GET(url, headers)
.DELETE(url, headers)
  .HEAD(url, headers)
   .PUT(url, body, headers)
 .PATCH(url, body, headers)
  .POST(url, body, headers)

.stash()

Any result can be stashed, e.g.:

.stash("invite");

You can stash only part of a result if you like:

.stash("inviteCode", function(result) { return result.json.code; });

Anything you've stashed can be retrieved by passing in a name preceded by a ":". You can also destash a nested attribute like this: ":invite.code".

You can use these ":" names in urls, request bodies, and expectations.

The stash is also a nice way to ensure that an operation does not run until some result it needs is available. An operation just waits until the stashed result has been fulfilled.

.wait([millis])

By default, driver executes all of your requests in parallel. A request will automatically block if it depends on a stashed value. However, when you need to wait for some previous request to complete but you are not depent upon a returned value, you can use wait(). Requests after a wait() won't fire untill all previous requests have returned. You can also specify an additional number of millis to wait for, but this is generally a brittle approach to handling lags (see until() below).

.expect([statusCode], [fn | jsonExpression]);

fn(result)

If using a custom fn, it must return a truthy to pass, and return a falsey or throw an exception to fail. The result obect has the following keys:

  • json: http response body parsed as json
  • text: http response body as a string
  • headers: http response headers
  • statusCode: http status code
jsonExpressions
  • The default behavior for a json expression is to check that the response has at least the specified values, i.e. the expectation does not need to include all of the responses values
  • $unordered: Replace an [1,2,3] with {$unordered: [1, 2, 3]} if you do not care about the order of the result
  • $length: Replace [1,2,3] with {$length: 3} if all you care about is length
  • {key: "$not-exists"} and {key: "$exists"}: insure the specified field is not present or is present
  • $int: require any integer
  • $date: require any iso date
  • $gt, $gte, $lt, $lte
  • Check out expector.js to find all the special '$' keywords.

.until([statusCode], [fn | jsonExpression], [millis])

Works exactly like .expect(...), only it will repeat the previous api call until the stated condition is met, or give up after 10 seconds or the specified time. It's not recommended to specify millis, except when initially setting up the test.

.never([statusCode], [fn | jsonExpression], [millis])

Ensure that an expectation "never" comes to be, or at least doesn't happen for a while... By default it waits 10 seconds. (This can make testing take painfully long... Would be nice if never tests could run in the background and not block the next test, but this won't work if you are clearing your server state between tests.)