woly

Exhaustive component library for React

Usage no npm install needed!

<script type="module">
  import woly from 'https://cdn.skypack.dev/woly';
</script>

README

🐣 Woly UI development kit

Storybook Commitizen friendly Conventional Commits NPM Package CI

Create your own UI/UX library.

Usage

  • yarn commit - run commit wizard, all commits should be created with this command
  • yarn start - start dev server on http://localhost:8000
  • yarn story-build - build storybook
  • yarn lint - runs eslint, prettier and stylelint and fixes source code

Building

Script yarn build runs build scripts from ./build directory.

Env variables:

  • DEBUG=woly:* - shows debug lines
  • NEXT=true - publishes version to the next npm dist-tag
  • DRY_RUN=true - runs npm publish with --dry-run parameter
  • PACKAGE=name - build only name package
  • PACKAGES=first,second - build only first and second packages

Example:

DEBUG=woly:* NEXT=true DRY_RUN=true yarn build

Publish process

  1. Review draft release in GitHub
  2. Set version from draft release to versions.json
  3. Commit and push
  4. At the Publish Package CI press "Run workflow"
  5. Set "yes" for packages that need to be published, and press "Run workflow"

Screenshot testing

Description

Screenshot testing is the automated process of comparing the visible output of a component against a baseline image. It helps to prevent unwanted visual changes and to make sure that it works as intended.

All examples are based on the woly package.

Writing and Organizing Tests

Folder Structure

Below you can see recommended folder structure. Test files MUST be inside __screenshot-test__ directory, which directly inside the component's folder.

button/
├─ __screenshot-test__
│  ├─ config.json
│  ├─ index.tsx
├─ index.tsx
├─ usage.mdx
├─ specification.mdx

Test files

  • config.js – config file for test-runner, which describes how to capture component's states
  • index.tsx – a React component, that renders all combinations of the tested component via StateMap component from lib/state-map
config.json
Name Type Default value Description
name string null Component's name
selector string null Selector, which test-runner uses to find a component in a test page
screenshotSize { width: number, height: number} null Screenshot's size. Final snapshot's width will be equal to states amount x screenshot's width
states State[] [] States to capture
State

A state can be described by a simple string like a static | press | hover | focus. The test-runner will bring a component to that state.

However, when it comes to more complex components, we need a flexible way to reproduce a state we want to capture. In this case, pass an object instead of a string with the structure described below:

Name Type Default value Description
name string null State's name
actions function null An async function that will bring component to the desired state

A actions function gets the following parameters:

  • el – the actual component (see methods)
  • elWrapper – a component's wrapper. The test-runner makes a screenshot of this element. Has the same methods as el
  • disabled – if a component is disabled, some playwright methods can't be triggered on component and it will throw an error (see actionability checks). Check for this boolean before invocing methods like fill on input elements
  • page – a test page (see methods)

Attention: When using a function for describing a state, be aware that test-runner unable to reset the state after capturing it, a state will be just passed unchanged to the next one. If you don't want this behavior, you can return a reset function from state that will be called before moving to the next state.

For example:

{
  name: 'text-filled',
  actions: async ({ el, elWrapper, disabled, page }) => {
    const input = await el.$('input[type="password"]');
    await input.type('qwerty');

    // reset function
    return async () => {
      if(!disabled) {
        await input.fill(''); // remove the text in the input
      }
      await elWrapper.focus(); // remove the focus from the input
    }
  },
},
index.tsx

The example is based on the Button component.

import React from 'react';
import { IconPlus } from 'static/icons';
import { Sizes, StateMap, Priorities } from 'dev/state-map';
import { block } from 'dev/block';

import { Button } from '../index';

export const ButtonStateMap = () => {
  return (
    <StateMap
      // all `buttons` props variations
      propVariations={{
        disabled: [true, false],
        icon: [true, false],
        outlined: [true, false],
        size: Sizes,
        priority: Priorities,
      }}
      // prop, by wich the variants will be grouped
      groupByProp="variant"
      render={({ size, icon, priority, outlined, disabled }) => {
        const SizeBlock = block[size];

        return (
          <SizeBlock>
            <Button
              // provide classname for component and match it
              // in configs 'selector' option
              className="button-st"
              text="button"
              icon={icon ? <IconPlus /> : undefined}
              priority={priority}
              outlined={outlined}
              disabled={disabled}
            />
          </SizeBlock>
        );
      }}
    />
  );
};



Local testing

For local testing, run the command yarn test:screenshot

Env variables:

  • DEBUG=screenshot* - shows debug lines
  • INCLUDE=first,second - include only components first and second to testing. Has higher priority than EXCLUDE.
  • EXCLUDE=first,second - exclude components first and second from testing

If you want to manually send snapshots to percy, you must pass PERCY_TOKEN env variable with a token as the value (grab it in the persy.io's Project settings section) and run the test by the command yarn percy:ci, e.g. PERCY_TOKEN=***** yarn percy:ci.