Ecosyste.ms: Awesome
An open API service indexing awesome lists of open source software.
https://github.com/jl-/tupe
Generic unit-testing runner for front-end.➰
https://github.com/jl-/tupe
ava front-end puppeteer test testing unit unit-testing
Last synced: 22 days ago
JSON representation
Generic unit-testing runner for front-end.➰
- Host: GitHub
- URL: https://github.com/jl-/tupe
- Owner: jl-
- Created: 2018-10-11T01:05:07.000Z (about 6 years ago)
- Default Branch: master
- Last Pushed: 2018-10-18T06:07:41.000Z (about 6 years ago)
- Last Synced: 2024-11-16T15:05:49.595Z (27 days ago)
- Topics: ava, front-end, puppeteer, test, testing, unit, unit-testing
- Language: JavaScript
- Homepage:
- Size: 867 KB
- Stars: 2
- Watchers: 2
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
- awesome-puppeteer - tupe - A generic unit-testing runner for front-end. (Testing)
README
# Tupe
[](https://www.npmjs.com/package/tupe)
[](https://www.npmjs.com/package/tupe)
##### [Features](#features) | [Motivations](#motivations) | [Configuration](#configuration) | [API](#api) | [Examples](#examples) | [FAQ](#faq)
> A generic unit-testing runner for front-end.➰
## Features
- 🚀 Blazing fast: Tupe use the great [Parcel](https://github.com/parcel-bundler/parcel) as bundler, and [Puppeteer](https://github.com/GoogleChrome/puppeteer) as runtime environment under the hook, making it fast on both bundle-time and runtime.
- 🌓 Isolation & 🎎 Concurrency: Each test file is bundled and run as a isolated new tab on puppeteer concurrently.
- 🎨 Descriptive assertion: bonus with [Power-Assert](https://github.com/power-assert-js/power-assert) builtin.
- 🐦 Configuration-Free: almost no configuration, except for some basic babel setup, which might agrees to your project setup already
- ❄️ Simplicity. use the standard assert interface.
## Motivations
There are already tons of unit-testing runner/framework for JavaScript.
- some requires a bit configuration, like karma, not simple enough.
- when it comes to front-end code testing that requires a browser environment, many of them utilize jsdom, which is a subset emulation of a web browser for use in nodejs environment. That's awesome, but would it be great and more reliable if our code runs in a real-world browser?
- features like concurrency and isolation are great. You should definitely try out AVA and Jest, they are awesome. But again, they don't seems to bundle for a real-world browser? Correct me if I'm wrong, thanks~
- Puppeteer is born, which seems to be a perfect fit for browser-side JavaScript testing. Maybe we can try it out and learn how to build a unit test runner, And have all the fun along the way, that's the whole point..
## Install
```bash
yarn add --dev tupe
```
## CLI
```txt
tupe
Run Tests
Positionals:
files path or glob for test files [string]
Options:
-h, --help Show help [boolean]
-v, --verbose [default: false]
--watch watch mode [default: false]
--port server port [default: 1234]
--tmpdir temporary directory [default: ".tmp"]
--fail-fast exit on first fail [boolean] [default: true]
```
## Configuration
You can configure Tupe by specifying a `tupe` field at your project's `package.json`
- `files`: file or glob for test files
- `tmpdir`: this is a temporary folder used by [Parcel](https://parceljs.org/api.html) for compiled files
- `fail-fast`: in a test file, global `before` or `after` hooks failure will always terminate the runner as failure; For each individual case, if a `beforeEach` hook failed, this case itself and its `afterEach` hooks will not be executed, a case-failure event will araise and the runner will continue to the next case. However, by setting this option to `true`, any case failure will terminate the runner of that test file.
- `coverage`: options for istanbul, defaults to [this](https://github.com/istanbuljs/istanbuljs/blob/master/packages/istanbul-api/lib/config.js#L14)
```json
{
"tupe": {
"files": [
"test/**/*.js"
],
"tmpdir": ".tmp",
"fail-fast": false,
"coverage": {
"dir": "test/.coverage",
"reporters": ["text"],
"check": {
}
}
}
}
```
#### .babelrc
- to enable power-assert feature, add `power-assert` preset and set patterns for `babel-plugin-espower`
- to enable code coverage, add `istanbul` plugin, recommend only set for a test env
```json
{
"presets": ["@babel/env", "power-assert"],
"plugins": [
["babel-plugin-espower", {
"patterns": [
"t(value, [message])",
"t.ok(value, [message])",
"t.equal(actual, expected, [message])",
"t.deepEqual(actual, expected, [message])",
"t.notEqual(actual, expected, [message])",
"t.strictEqual(actual, expected, [message])",
"t.notStrictEqual(actual, expected, [message])",
"t.notDeepEqual(actual, expected, [message])",
"t.deepStrictEqual(actual, expected, [message])",
"t.notDeepStrictEqual(actual, expected, [message])",
"t.throws(block, [error], [message])"
]
}]
],
"env": {
"test": {
"plugins": [
["istanbul", {
"include": [
"src/**/*.js"
]
}]
]
}
}
}
```
## API
- `test.before([title], fn(t, [callback]))`
- `test.beforeEach([title], fn(t, [callback]))`
- `test.afterEach([title], fn(t, [callback]))`
- `test.after([title], fn(t, [callback]))`
- `test(title, fn(t, [callback]))`
- `t` is a enhenced assert function, added with a additional `t.context` abject as a isolated context for each case, live through `beforeEach` due `afterEach` hook
## Examples
```javascript
import test from 'tupe';
test('hello tupe!', t => {
const who = { name: 'Tupe!' };
t(who.name === 'Tupe!');
});
test('async waiting...', async t => {
const answer = true;
await new Promise(r => setTimeout(r, 1500));
t(answer === true);
});
test('misspell my name ?', t => {
const who = { name: 'Hello Tupe!' };
t(who.name === 'Hel' + 'o Tupe~');
});
test('use callback ?', (t, done) => {
setTimeout(() => {
t('dump' === 'dump');
done();
}, 1500);
});
test('array assert', t => {
const runners = ['AVA', 'Tupe'];
t(runners.indexOf('Tupe') === 0);
});
```
## FAQ