Zorki 📸

The command-line utility for visual-regression testing.

NOTE: Currently the package is in Alpha, the API will be changing...

Internally it uses AWS S3 for storing baseline images, so they can be accessed from different machines. This is an alternative to storing them into a local folder and commiting to your git repository. The latter solution is not bad, but it can easily bloat the size of your git-stored project, so Zorki takes a different approach.

Requirements

You need an account on AWS and need to create a special bucket which the application will use for storing baseline images.

Installation

yarn add zorki

Configuration

Zorki expects a JSON file with some configuration. Example:

{
  "urls": [
    "http://example.com"
  ],
  "storeDir": "./tmp",
  "screenWidths": [320, 1920],
  "accessKeyId": "AWS_S3_ACCESS_KEY_ID",
  "secretAccessKey": "AWS_S3_SECRET_ACCESS_KEY",
  "bucket": "AWS_S3_BUCKET"
}

This will run tests against http://example.com with screen widths of 320 and 1920 pixels. Directory ./tmp will be used as an intermediate storage. AWS_S3_BUCKET will be the bucket used in a cloud.

Usage

To run tests:

zorki run --config=/path/to/your/config.json

To clean up all stored baseline images from the remote storage:

zorki cleanup --config=/path/to/your/config.json

To list all baseline images already stored for the project:

zorki cleanup --config=/path/to/your/config.json

Gotchas

If your page has some animations they can introduce some flakiness when running your tests. The simplest approach would be to disable them manually. Zorki has no clear control over pages which you are testing, so this would be your goal to make them look consistent between tests.

TODO

• Write comprehensive README