valid

VALID - VisuAL Image Diff --- visual regression testing with ease

Note that this project is WIP and not yet reliable for use in actual testing. Proposals and PRs are welcome! :D

Installation

Note that node version 7.10.1 or newer is required to use valid.

node install -g valid
or
yarn global add valid

Usage

1. Setup your config JSON file if you haven't (see example config below)
2. Run valid config.json
   a. Optionally you can set --checkMobile and --mobileLandscape parameters which are false by default. It is suggested that you use different config files if you want to run tests for desktop and mobile viewports.

Proposed Flow

1. Pass in an array of links/routes to visit.
2. The lib will then loop through these, visit each link and take a screenshot.
3. Screenshots are saved in an "expected" and "tmp" folder.
4. The generated diff image will use a naming pattern "output_[i].png" where [i] equals the current iteration index.
5. An image-diff lib will then be used to detect visual changes.
6. If a difference is found the link will be added to a "faulty" array.
7. If no difference is found the "output_[i].png" will be deleted.
8. [OPTIONAL] After each passed iteration the taken screenshot from "tmp" will be moved to replace the old one in "expected". This might also be configurable on a per-item basis.
9. After the loop finishes an object will be returned with the following design

   {
     didPass: BOOLEAN,
     faultyPages: [{
       url: 'LINK',
       image: '[PATH TO DIFF IMG]',
       ...
     }]
   }

Notes on implementation

The input data should be passable as JSON file or a JSON compliant object. This could look like the following:

[{
 url: 'https://www.example.com/about',
 image: './regression-testing/expected/about.png',
 didChange: false
}]

The didChange property implies that there should be no difference between the captured and expected image. If a difference is found it will be treated as an error case.
If this is on the other hand set to true, we expect a difference and won't treat this as an error. The captured image will then (if point 8 in the Propose Flow list above is implemented) replace the one in the expected folder