Futuristic test runner

Even though JavaScript is single-threaded, IO in Node.js can happen in parallel due to its async nature. AVA takes advantage of this and runs your tests concurrently, which is especially beneficial for IO heavy tests. In addition, test files are run in parallel as separate processes, giving you even better performance and an isolated environment for each test file. Switching from Mocha to AVA in Pageres brought the test time down from 31 sec to 11 sec. Having tests run concurrently forces you to write atomic tests, meaning tests don't depend on global state or the state of other tests, which is a great thing!

• Usage
• CLI Usage
• Documentation
• API
• Assertions
• FAQ

• Minimal and fast
• Simple test syntax
• Runs tests concurrently
• Enforces writing atomic tests
• No implicit globals
• Isolated environment for each test file
• Write your tests in ES2015
• Promise support
• Generator function support
• Async function support
• Enhanced asserts

import test from 'ava';

test(t => {
	t.same([1, 2], [1, 2]);
	t.end();
});

Install AVA globally $ npm install --global ava and run $ ava --init (with any options) to add AVA to your package.json or create one.

{
	"name": "awesome-package",
	"scripts": {
		"test": "ava"
	},
	"devDependencies": {
		"ava": "^0.3.0"
	}
}

import test from 'ava';

test('foo', t => {
	t.pass();
	t.end();
});

test('bar', t => {
	t.plan(2);

	setTimeout(() => {
		t.is('bar', 'bar');
		t.same(['a', 'b'], ['a', 'b']);
	}, 100);
});

$ npm test

$ ava --help

  Usage
    ava [<file|folder|glob> ...]

  Options
    --init       Add AVA to your project
    --fail-fast  Stop after first test failure
    --serial     Run tests serially

  Examples
    ava
    ava test.js test2.js
    ava test-*.js
    ava --init
    ava --init foo.js

  Default patterns when no arguments:
  test.js test-*.js test/*.js

Files starting with _ are ignored. This can be useful for having helpers in the same directory as your test files.

Tests are run asynchronously and require you to either set planned assertions t.plan(1), explicitly end the test when done t.end(), or return a promise. Async functions already returns a promise implicitly, so no need for you to explicitly return a promise in that case.

You have to define all tests synchronously, meaning you can't define a test in the next tick, e.g. inside a setTimeout.

Test files are run from their current directory, so process.cwd() is always the same as __dirname. You can just use relative paths instead of doing path.join(__dirname, 'relative/path').

To create a test, you call the test function you required from AVA and pass in an optional test name and a function containing the test execution. The passed function is given the context as the first argument, where you can call the different AVA methods and assertions.

test('name', t => {
	t.pass();
	t.end();
});

Naming a test is optional, but you're recommended to use one if you have more than one test.

test(t => {
	t.end();
});

You can also choose to use a named function instead:

test(function name(t) {
	t.end();
});

Planned assertions are useful for being able to assert that all async actions happened. It also comes with the benefit of not having to manually end the test.

This will result in a passed test:

test(t => {
	t.plan(1);

	setTimeout(() => {
		t.pass();
	}, 100);
});

While concurrency is awesome, there are some things that can't be done concurrently. In these rare cases, you can call test.serial, which will force those tests to run serially before the concurrent ones.

test.serial(t => {
	t.end();
});

Skipped tests are shown in the output as skipped, but never run.

test.skip('unicorn', t => {
	t.end();
});

When setup and/or teardown is required, you can use test.before() and test.after(), used in the same manner as test(). The test function given to test.before() and test.after() is called before/after all tests. You can also use test.beforeEach() and test.afterEach() if you need setup/teardown for each test. Hooks are run serially in the test file. Add as many of these as you want. You can optionally specify a title that is shown on failure.

test.before(t => {
	// this runs before all tests
	t.end();
});

test.before(t => {
	// this runs after the above, but before tests
	t.end();
});

test.after('cleanup', t => {
	// this runs after all tests
	t.end();
});

test.beforeEach(t => {
	// this runs before each test
	t.end();
});

test.afterEach(t => {
	// this runs after each test
	t.end();
});

test(t => {
	// regular test
	t.end();
});

The beforeEach & afterEach hooks can share context with the test:

test.beforeEach(t => {
	t.context.data = generateUniqueData();
	t.end();
});

test(t => {
	t.is(t.context.data + 'bar', 'foobar');
	t.end();
});

The context is by default an object, but it can also be directly assigned:

test.beforeEach(t => {
	t.context = 'unicorn';
	t.end();
});

test(t => {
	t.is(t.context, 'unicorn');
	t.end();
});

You can use any assertion module instead or in addition to the one that comes with AVA, but you won't be able to use the .plan() method, yet.

import assert from 'assert';

test(t => {
	assert(true);
	t.end();
});

AVA comes with builtin support for ES2015 through Babel.

Just write your tests in ES2015. No extra setup needed.

test(t => {
	t.pass();
	t.end();
});

You can also use your own local Babel version:

{
	"devDependencies": {
		"ava": "^0.3.0",
		"babel-core": "^5.8.0"
	}
}

If you return a promise in the test you don't need to explicitly end the test as it will end when the promise resolves.

test(t => {
	return somePromise().then(result => {
		t.is(result, 'unicorn');
	});
});

AVA comes with builtin support for generator functions.

test(function * (t) {
	const value = yield generatorFn();
	t.true(value);
});

You don't have to manually call t.end().

AVA comes with builtin support for async functions (async/await).

test(async function (t) {
	const value = await promiseFn();
	t.true(value);
});

// async arrow function
test(async t => {
	const value = await promiseFn();
	t.true(value);
});

You don't have to manually call t.end().

Type: string

Test title.

Type: function

Should contain the actual test.

Passed into the test function and contains the different AVA methods and assertions.

Plan how many assertion there are in the test. The test will fail if the actual assertion count doesn't match planned assertions. When planned assertions are used you don't need to explicitly end the test.

Be aware that this doesn't work with custom assert modules. You must then call .end() explicitly.

End the test. Use this when plan() is not used.

Assertions are mixed into the test context:

test(t => {
	t.ok('unicorn'); // assertion
	t.end();
});

Passing assertion.

Failing assertion.

Assert that value is truthy.

Assert that value is falsy.

Assert that value is true.

Assert that value is false.

Assert that value is equal to expected.

Assert that value is not equal to expected.

Assert that value is deep equal to expected.

Assert that value is not deep equal to expected.

Assert that function throws an error or promise rejects.

error can be a constructor, regex, error message or validation function.

Assert that function doesn't throw an error or promise resolves.

Assert that regex matches contents.

Assert that error is falsy.

AVA comes with power-assert builtin, giving you more descriptive assertion messages. It reads your test and tries to infer more information from the code.

The following test:

test(t => {
	const x = 'foo';
	t.ok(x === 'bar');
	t.end();
});

Would normally give the unhelpful output:

false === true

With the enhanced asserts, you'll get:

t.ok(x === 'bar')
     |
     "foo"

True, you could use t.is() in this case, and probably should, but this is just a simple example.

Let try a more advanced example:

test(t => {
	const a = /foo/;
	const b = 'bar';
	const c = 'baz';
	t.ok(a.test(b) || b === c);
	t.end();
});

And there you go:

t.ok(a.test(b) || b === c)
       |    |     |     |
       |    "bar" "bar" "baz"
       false

All the assert methods are enhanced.

Have fun!

Each test file is run in a separate Node.js process. This comes with a lot of benefits. Different test files can no longer affect each other. Like test files mocking with the global environment, overriding builtins, etc. However, it's mainly done for performance reasons. Even though Node.js can run async IO concurrently, that doesn't help much when tests are heavy on synchronous operations, which blocks the main thread. By running tests concurrently and test files in parallel we take full advantage of modern systems.

Running tests concurrently comes with some challenges, doing IO is one. Usually, serial tests just create temp directories in the current test directory and cleans it up at the end. This won't work when you run tests concurrently as tests will conflict with each other. The correct way to do it is to use a new temp directory for each test. The tempfile and temp-write modules can be helpful.

AVA runs tests concurrently by default, which is suboptimal when you need to debug something. Instead, run tests serially with the --serial option:

$ ava --serial

You can't use istanbul for code coverage as AVA spawns the test files, but you can use nyc instead, which is basically istanbul with support for subprocesses.

Mocha requires you to use implicit globals like describe and it with the default interface (which most people use), too unopinionated, bloated, synchronous by default, serial test execution, and slow. Tape and node-tap are pretty good. AVA is highly inspired by their syntax. However, they both execute tests serially and they've made TAP a first-class citizen which has IMHO made their codebases a bit convoluted and coupled. TAP output is hard to read so you always end up using an external tap reporter. AVA is highly opinionated and concurrent. It comes with a default simple reporter and will in the future support TAP through a reporter.

AVA, not Ava or ava. Pronounced /ˈeɪvə/ ay-və.

Andromeda galaxy.

Concurrency is not parallelism. It enables parallelism. To learn more about the difference, see this Stack Overflow answer.

• Stack Overflow
• Gitter chat

• gulp-ava - Run tests with gulp
• grunt-ava - Run tests with grunt

Sindre Sorhus	Kevin Mårtensson	Vadim Demedes