# Automated Testing

Test automation is an efficient way of validating that your application code works as intended.
While Electron doesn't actively maintain its own testing solution, this guide will go over a couple
ways you can run end-to-end automated tests on your Electron app.

## Using the WebDriver interface

From [ChromeDriver - WebDriver for Chrome][chrome-driver]:

> WebDriver is an open source tool for automated testing of web apps across many
> browsers. It provides capabilities for navigating to web pages, user input,
> JavaScript execution, and more. ChromeDriver is a standalone server which
> implements WebDriver's wire protocol for Chromium. It is being developed by
> members of the Chromium and WebDriver teams.

There are a few ways that you can set up testing using WebDriver.

### With WebdriverIO

[WebdriverIO](https://webdriver.io/) (WDIO) is a test automation framework that provides a
Node.js package for testing with WebDriver. Its ecosystem also includes various plugins
(e.g. reporter and services) that can help you put together your test setup.

#### Install the testrunner

First you need to run the WebdriverIO starter toolkit in your project root directory:

```sh npm2yarn
npx wdio . --yes
```

This installs all necessary packages for you and generates a `wdio.conf.js` configuration file.

#### Connect WDIO to your Electron app

Update the capabilities in your configuration file to point to your Electron app binary:

```javascript title='wdio.conf.js'
export.config = {
  // ...
  capabilities: [{
    browserName: 'chrome',
    'goog:chromeOptions': {
      binary: '/path/to/your/electron/binary', // Path to your Electron binary.
      args: [/* cli arguments */] // Optional, perhaps 'app=' + /path/to/your/app/
    }
  }]
  // ...
}
```

#### Run your tests

To run your tests:

```sh
$ npx wdio run wdio.conf.js
```

[chrome-driver]: https://sites.google.com/chromium.org/driver/

### With Selenium

[Selenium](https://www.selenium.dev/) is a web automation framework that
exposes bindings to WebDriver APIs in many languages. Their Node.js bindings
are available under the `selenium-webdriver` package on NPM.

#### Run a ChromeDriver server

In order to use Selenium with Electron, you need to download the `electron-chromedriver`
binary, and run it:

```sh npm2yarn
npm install --save-dev electron-chromedriver
./node_modules/.bin/chromedriver
Starting ChromeDriver (v2.10.291558) on port 9515
Only local connections are allowed.
```

Remember the port number `9515`, which will be used later.

#### Connect Selenium to ChromeDriver

Next, install Selenium into your project:

```sh npm2yarn
npm install --save-dev selenium-webdriver
```

Usage of `selenium-webdriver` with Electron is the same as with
normal websites, except that you have to manually specify how to connect
ChromeDriver and where to find the binary of your Electron app:

```js title='test.js'
const webdriver = require('selenium-webdriver')
const driver = new webdriver.Builder()
  // The "9515" is the port opened by ChromeDriver.
  .usingServer('http://localhost:9515')
  .withCapabilities({
    'goog:chromeOptions': {
      // Here is the path to your Electron binary.
      binary: '/Path-to-Your-App.app/Contents/MacOS/Electron'
    }
  })
  .forBrowser('chrome') // note: use .forBrowser('electron') for selenium-webdriver <= 3.6.0
  .build()
driver.get('http://www.google.com')
driver.findElement(webdriver.By.name('q')).sendKeys('webdriver')
driver.findElement(webdriver.By.name('btnG')).click()
driver.wait(() => {
  return driver.getTitle().then((title) => {
    return title === 'webdriver - Google Search'
  })
}, 1000)
driver.quit()
```

## Using a custom test driver

It's also possible to write your own custom driver using Node.js' built-in IPC-over-STDIO.
Custom test drivers require you to write additional app code, but have lower overhead and let you
expose custom methods to your test suite.

To create a custom driver, we'll use Node.js' [`child_process`](https://nodejs.org/api/child_process.html) API.
The test suite will spawn the Electron process, then establish a simple messaging protocol:

```js title='testDriver.js'
const childProcess = require('child_process')
const electronPath = require('electron')

// spawn the process
const env = { /* ... */ }
const stdio = ['inherit', 'inherit', 'inherit', 'ipc']
const appProcess = childProcess.spawn(electronPath, ['./app'], { stdio, env })

// listen for IPC messages from the app
appProcess.on('message', (msg) => {
  // ...
})

// send an IPC message to the app
appProcess.send({ my: 'message' })
```

From within the Electron app, you can listen for messages and send replies using the Node.js
[`process`](https://nodejs.org/api/process.html) API:

```js title='main.js'
// listen for messages from the test suite
process.on('message', (msg) => {
  // ...
})

// send a message to the test suite
process.send({ my: 'message' })
```

We can now communicate from the test suite to the Electron app using the `appProcess` object.

For convenience, you may want to wrap `appProcess` in a driver object that provides more
high-level functions. Here is an example of how you can do this. Let's start by creating
a `TestDriver` class:

```js title='testDriver.js'
class TestDriver {
  constructor ({ path, args, env }) {
    this.rpcCalls = []

    // start child process
    env.APP_TEST_DRIVER = 1 // let the app know it should listen for messages
    this.process = childProcess.spawn(path, args, { stdio: ['inherit', 'inherit', 'inherit', 'ipc'], env })

    // handle rpc responses
    this.process.on('message', (message) => {
      // pop the handler
      const rpcCall = this.rpcCalls[message.msgId]
      if (!rpcCall) return
      this.rpcCalls[message.msgId] = null
      // reject/resolve
      if (message.reject) rpcCall.reject(message.reject)
      else rpcCall.resolve(message.resolve)
    })

    // wait for ready
    this.isReady = this.rpc('isReady').catch((err) => {
      console.error('Application failed to start', err)
      this.stop()
      process.exit(1)
    })
  }

  // simple RPC call
  // to use: driver.rpc('method', 1, 2, 3).then(...)
  async rpc (cmd, ...args) {
    // send rpc request
    const msgId = this.rpcCalls.length
    this.process.send({ msgId, cmd, args })
    return new Promise((resolve, reject) => this.rpcCalls.push({ resolve, reject }))
  }

  stop () {
    this.process.kill()
  }
}

module.exports = { TestDriver };
```

In your app code, can then write a simple handler to receive RPC calls:

```js title='main.js'
const METHODS = {
  isReady () {
    // do any setup needed
    return true
  }
  // define your RPC-able methods here
}

const onMessage = async ({ msgId, cmd, args }) => {
  let method = METHODS[cmd]
  if (!method) method = () => new Error('Invalid method: ' + cmd)
  try {
    const resolve = await method(...args)
    process.send({ msgId, resolve })
  } catch (err) {
    const reject = {
      message: err.message,
      stack: err.stack,
      name: err.name
    }
    process.send({ msgId, reject })
  }
}

if (process.env.APP_TEST_DRIVER) {
  process.on('message', onMessage)
}
```

Then, in your test suite, you can use your `TestDriver` class with the test automation
framework of your choosing. The following example uses
[`ava`](https://www.npmjs.com/package/ava), but other popular choices like Jest
or Mocha would work as well:

```js title='test.js'
const test = require('ava')
const electronPath = require('electron')
const { TestDriver } = require('./testDriver')

const app = new TestDriver({
  path: electronPath,
  args: ['./app'],
  env: {
    NODE_ENV: 'test'
  }
})
test.before(async t => {
  await app.isReady
})
test.after.always('cleanup', async t => {
  await app.stop()
})
```