Détail du package

mvt

doesdev40MIT4.2.1

Minimum Viable Testing framework

test, testing, lite, no dependencies

readme

mvt NPM version js-standard-style Build Status

Minimum Viable Testing framework

A Minimalist Take on AVA's Approach to Testing

Because AVA is awesome. Security alerts on dev dependencies are not awesome. Especially when you use the same test library across dozens of projects. No matter how well maintained a project is, when it contains 300+ dependencies security alerts are going to occur.

But It's Just a Dev Dependency...

What's good about it

  • It has 0 dependencies (1 dev dep, and that's ava :smirk:)
  • It can be called via the mvt cli or by simply calling node [test-file].js
  • It doesn't transpile your code (the code you write is the code we test)

What it lacks (the most notable items)

  • Concurrency
    • That's not a thing here (likely never will be)
  • Transpilation
    • Also not a thing here (definitely never will be)
    • Actually, maybe I should add this to "What's good..."
    • I added it...
  • TAP Reporter
    • Happy to add it if there is any demand
  • A community and product maturity
    • The most crucial elements
    • And the primary reason you may want to stick to AVA

Table of Contents

Install

# Install globally
$ npm install --global mvt

# Install for project
$ npm install --save-dev mvt

Usage

mvt
# OR
mvt test
# OR
mvt test.js test2.js
# OR
mvt --verbose
# OR
mvt test --verbose
# OR
node test.js --verbose
# etc...
const test = require('mvt')

test.setup({ verbose: true })

test.after(() => console.log('test.after invoked'))

test.before(() => console.log('test.before invoked'))

test('assert.is works', (assert) => {
  assert.is(1, 1)
})

test.failing('test.failing and assert.fail works', (assert) => {
  assert.fail()
})

test.todo('test.todo works')

test.skip('test.skip works', (assert) => assert.truthy('skipped'))

test.only('test.only works', (assert) => assert.truthy('only'))

test.bench('test.bench works', (assert) => {
  return new Promise((resolve, reject) => {
    setTimeout(() => resolve(), 200)
  })
}, { samples: 5, max: 300 })

Output

API

Test Function

The only thing this module exports.

test ( message, testFunction )

Main function, give it a message and a test function. Test function receives the assert object (see below).

  • message: (String) Description of test
  • testFunction: ([Async]Function) Description of test

Setup and Teardown

test.setup ( opts )

Use this to configure your tests.

  • opts: (Object)
    • verbose (Boolean) - Print every test if true

test.before ( callback )

Run this before we start running any tests. [callback can be async]

test.after ( callback )

Run this after we run all tests. [callback can be async]

Test Modifiers

test.only ( message, testFunction )

Tests will only be run on any tests run with this modifier.

test.skip ( message, testFunction )

Skip that test (logical enough).

test.failing ( message, testFunction )

This test must fail. If it passes, we'll fail your whole test suite. Goteem.

test.todo ( message )

This is just a placeholder for your good intentions.

Special Tests

test.bench ( message, testFunction, opts )

Run the testFunction opts.samples || 10 times. If average run duration is more than opts.max || 100 milliseconds fail the test.

  • opts: (Object)
    • samples (Number) - How many times we should run the testFunction
    • max (Number [in ms]) - Maximum average duration threshhold
    • parallel (Boolean) - If Async Func run in parallel, default is false
    • cb (Function) - Called with { msTotal, msAvg } on bench completion

Assertions

Methods available on assert object passed to testFunction

  • is ( a, b ) - a and b must be identical
  • not ( a, b ) - a and b must not be identical
  • pass () - Passes errydamntime
  • fail () - Fails errydamntime
  • true ( a ) - a must be strictly true
  • false ( a ) - a must be strictly false
  • truthy ( a ) - a must be truthy
  • falsy ( a ) - a must be falsy
  • contains ( a, b ) - JSON.stringify(a) must contain (String)b
  • doesNotContain ( a, b ) - JSON.stringify(a) must not contain (String)b
  • lessThan ( a, b ) - a must be less than b
  • greaterThan ( a, b ) - a must be greater than b
  • deepEqual ( a, b ) - a must be deepEqual to b
  • notDeepEqual ( a, b ) - a must not be deepEqual to b
  • throws ( a ) - a must be a function, and it must throw
  • notThrows ( a ) - a must be a function, and it must not throw
  • throwsAsync ( a ) - a must be an async function, and it must throw
  • notThrowsAsync ( a ) - a must be an async function, and it must not throw

Notes

  • If your test file is called with the --verbose flag it will list all passed tests

  • It fails fast and hard with process.exit(1)

  • If you have diff installed as a peer dependency, we'll use that for string diffs. To make them more readable and what not.

License

MIT © Andrew Carpenter

changelog

Changelog

All notable changes to this project will be documented in this file.

The format is (loosely) based on Keep a Changelog, and this project adheres to Semantic Versioning.

4.2.1

2021-07-10

Changed

  • Fix #9, published with CRLF line endings

4.2.0

2021-07-10

Changed

  • Fix #10, caused by circular require
  • Update Travis to use node 16, 14, 12
  • Update ava

4.1.1

2020-07-16

Changed

  • Change emoji support check logic
  • Update ava

4.1.0

2020-03-10

Summary

Refactored into components. Fixed CLI bug (#2), fixed bug causing after not to be called on failing tests (#4). Various enhancements.

Fixes

  • Bugfix #2: CLI failed with exit code 0 on require failure (aka tests passed)
  • Bugfix #5: test.after never called on failing tests
  • Bugfix: Fix potential hang in cli-char-supported check

Changed

  • Refactored code base into components

Added

  • Asserts: lessThan, greaterThan, doesNotContain
  • Bench options parallel and cb
  • Allow only, failing, skip, todo to be chainable

4.0.0

2019-07-30

Summary

Version 4.0.0 is a complete rewrite. There is no overlap with the previous incarnation except for it's name. If you were using it before (unlikely) you'll need to rewrite your tests against the new API.

Added

  • Everything.

Changed

  • Everything.

Removed

  • Everything.