Cavy

Cavy is a cross-platform integration test framework for React Native, by Pixie Labs.

How does it work?

Cavy (ab)uses React ref generating functions to give you the ability to refer to, and simulate actions upon, deeply nested components within your application. Unlike a tool like enzyme which uses a simulated renderer, Cavy runs within your live application as it is running on a host device (e.g. your Android or iOS simulator).

This allows you to do far more accurate integration testing than if you run your React app within a simulated rendering environment.

Where does it fit in?

We built Cavy because, at the time of writing, React Native had only a handful of testing approaches available:

Unit testing components (Jest).
Shallow-render testing components (enzyme).
Testing within your native environment, using native JS hooks (Appium).
Testing completely within your native environment (XCTest).

Cavy fits in between shallow-render testing and testing within your native environment.

Cavy's components

Cavy provides 3 tools to let you run integration tests:

A store of 'test hooks'; key-value pairs between a string identifier and a component somewhere in your app component tree.
A set of helper functions to write spec files.
A <Tester> component you wrap around your entire app to make the test hook store available, and autorun your test cases on boot.

Installation

To get started using Cavy, install it using yarn:

yarn add cavy --dev

or npm:

npm i --save-dev cavy

Basic usage

Check out the sample app for example usage. Here it is running:

Hook up components for testing

Add 'hooks' to any components you want to test by adding a ref and using the generateTestHook function.

generateTestHook takes a string as its first argument - this is the identifier to be used in tests. It takes an optional second argument in case you want to set your own ref generating function.

Stateless functional components cannot be assigned a ref since they don't have instances. Use the wrap function to wrap them inside a non-stateless component.

import React, { Component } from 'react';
import { TextInput } from 'react-native';
import { FuncComponent } from 'somewhere';

import { hook, wrap } from 'cavy';

class Scene extends Component {
  render() {
    const WrappedComponent = wrap(FuncComponent);
    return (
      <View>
        <TextInput
          ref={this.props.generateTestHook('Scene.TextInput')}
          onChangeText={...}
        />
        <WrappedComponent
          ref={this.props.generateTestHook('Scene.Component')}
          onPress={...}
        />
      </View>      
    );
  }
}

const TestableScene = hook(Scene);
export default TestableScene;

Write your test cases

Using your component identifiers, write your spec functions. We suggest saving these in a spec folder, naming them something like ./specs/AppSpec.js.

export default function(spec) {
  spec.describe('My feature', function() {
    spec.it('works', async function() {
      await spec.fillIn('Scene.TextInput', 'some string')
      await spec.press('Scene.button');
      await spec.exists('NextScene')
    });
  });
}

See below for a list of currently available spec helper functions.

Set up your test wrapper

Import Tester, TestHookStore and your specs in your top-level JS file (typically this is your index.{ios,android}.js files), and instantiate a new TestHookStore.

Wrap your app in a Tester component, passing in the TestHookStore and an array containing your imported spec functions.

Optional props:

waitTime - Integer, the time in milliseconds that your tests should wait to find specified 'hooked' components. Set to 2000 (2 seconds) by default.

startDelay - Integer, the time in milliseconds before test execution begins. Set to 0 by default.

clearAsyncStorage - Boolean, set this to true to clear AsyncStorage between each test e.g. to remove a logged in user. Set to false by default.

import React, { Component } from 'react';
import { AppRegistry } from 'react-native';
import { Tester, TestHookStore } from 'cavy';
import AppSpec from './specs/AppSpec';
import App from './app';

const testHookStore = new TestHookStore();

export default class AppWrapper extends Component {
  render() {
    return (
      <Tester specs={[AppSpec]} store={testHookStore} waitTime={4000}>
        <App />
      </Tester>
    );
  }
}

AppRegistry.registerComponent('AppWrapper', () => AppWrapper);

Congratulations! You are now all set up to start testing your app with Cavy.

Your tests will run automatically when you run your app using either:

$ react-native run-ios

$ react-native run-android

Available spec helpers

fillIn(identifier, str) - fills in the identified 'TextInput'-compatible component with the provided string (str). Your component must respond to the property onChangeText.

press(identifier) - presses the identified component. Your component must respond to the property onPress.

pause(integer) - pauses the running test for the length of time, specified in milliseconds (integer). This is useful if you need to allow time for a response to be received before progressing.

exists(identifier) - returns true if the component can be identified (i.e. is currently on screen).

notExists(identifier) - as above, but checks for the absence of the component.

findComponent(identifier) - returns the identified component. This function should be used if your testable component does not respond to either onChangeText or onPress, for example:

picker = await spec.findComponent('Scene.modalPicker');
picker.open();

FAQs

How does Cavy compare to Appium? What is the benefit?

Cavy is a comparable tool to Appium. The key difference is that Appium uses native hooks to access components (accessibility IDs), wheras Cavy uses React Native refs. This means that Cavy sits directly within your React Native environment (working identically with both Android and iOS builds), making it easy to integrate into your application very quickly, without much overhead.

What does this allow me to do that Jest does not?

Jest is a useful tool for unit testing individual React Native components, whereas Cavy is an integration testing tool allowing you to run end-to-end user interface tests.

How about supporting stateless components?

We'd love for Cavy to be better compatible with stateless functional components and would be more than happy to see its reliance on refs replaced with something better suited to the task... What that looks like specifically, we're not 100% sure yet. We're very happy to discuss possible alternatives!

Contributing

Check out the latest master to make sure the feature hasn't been implemented or the bug hasn't been fixed yet.
Check out the issue tracker to make sure someone already hasn't requested it and/or contributed it.
Fork the project.
Start a feature/bugfix branch.
Commit and push until you are happy with your contribution.
Please try not to mess with the package.json, version, or history. If you want to have your own version, or is otherwise necessary, that is fine, but please isolate to its own commit so we can cherry-pick around it.

rajit/cavy