Skip to content
This repository has been archived by the owner. It is now read-only.
The low-level SDK used by other SDKs to communicate with the Visual Grid
Branch: master
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Permalink
Type Name Latest commit message Commit time
Failed to load latest commit information.
.vscode
docs
scripts
src
test
.babelrc
.eslintignore Now abort is stopping tests and then waiting for open and checkWindow… Feb 4, 2019
.eslintrc.json
.gitignore
History.md
LICENSE
README.md
package-lock.json
package.json
test-iframe.cdt.json

README.md

visual-grid-client

A library that drives the visual grid with dom snapshot rendering.

Installing

npm install @applitools/visual-grid-client

Using the package

const {makeVisualGridClient} = require('@applitools/visual-grid-client')

See below for the full API.

API

makeVisualGridClient

  • To create a visualGridClient, call makeVisualGridClient:
const {makeVisualGridClient} = require('@applitools/visual-grid-client')
const visualGridClient = makeVisualGridClient()

The visualGridClient, returned by makeVisualGridClient, is an object with the following function:

  • openEyes(configOverride): to start a set of tests, where each step is a set of renderings according to the browser stuff in the configuration. This function will return an object with functions (see below) allowing you to create renderings (or "steps" in Applitools parlance) for the test.

openEyes

Async function openEyes will create a test. Actually, it will create a series of tests, one for each browser configuration defined in the browser property of the configuraion.

  • openEyes accepts a configuration object that will override the default configuration found by makeVisualGridClient, per this test.

  • Returns a promise to an object with the following functions:

  • checkWindow(...): creates a "step" that checks the window according to the baseline. Note that this function will not fail, and you need to await the promises returned from close() to wait for the failure or success of a batch of steps in the test.

  • close(): async closes the test (or series of tests) created by openEyes.

  • abort(): if you want to abort this test (or series of tests). Async.

checkWindow(...)

checkWindow receives an object with the following parameters:

  • tag: the name of the step, as seen in Applitools Eyes.
  • url: the URL appearing in the address bar of the browser. All relative URLs in the CDT will be relative to it.
  • cdt: the HTML and set and resources, in the x-applitools-html/cdt format (see below). you can use domNodesToCdt to create a CDT from a document.
  • sizeMode: the target of the rendering. Can be one of viewport, full-page, selector, region
  • selector: if the sizeMode is selector, this is the selector we are targetting.
  • region: if the sizeMode is region, this is the region we are targetting. This is an object with x, y, width, height properties.
  • ignore: TBD
  • floating: TBD
  • strict: TBD
  • layout: TBD
  • sendDom: TBD
  • scriptHooks: a set of scripts to be run by the browser during the rendering. An object with the following properties:
    • beforeCaptureScreenshot: a script that runs after the page is loaded but before taking the screenshot.
  • resourceUrls: By default, an empty array. Additional resource URLs not found in the CDT.
  • resourceContents: a map of all resource values (buffers). The keys are URLs (relative to the url property). The value is an object with the following properties:
    • url: yes, again.
    • type: the content type of the resource.
    • value: a Buffer of the resource content.
  • matchLevel: The method to use when comparing two screenshots, which expresses the extent to which the two images are expected to match.

close()

close receives no parameters, and returns a promise.

  • The promise will be resolved (with undefined as value) if all tests defined in the openEyes passed.
  • The promise will be rejected (with DiffsFoundError as defined in Applitools Eyes SDK Core) if there were differences found in some tests defined in the openEyes.

The CDT format

{
  domNodes: [
    {
      nodeType: number, // like in the DOM Standard
      nodeName:...’ , // for elements and DocumentType
      nodeValue:...’, // for text nodes
      attributes: [{name, value}, ...],
      childNodeIndexes: [index, index, ...]
    },
    //...
  ],
  resources: [
    {
      hashFormat: 'sha256', // currently the only hash format allowed
      hash: '....', // the hash of the resource.
      contentType: '...', // the mime type of the resource.
    },
    //...
  ]
}

domNodesToCdt

Accepts a document object conforming to the DOM specification (browser document is fine, as is the JSDOM document). Returns a cdt, ready to be passed to checkWindow

Configuration

  • See Eyes Cypress configuration for a list of properties in the configuration and to understand how the visual grid client reads the configuration.

Logging

???

Example

Example Mocha test that uses the visual grid client:

const path = require('path')
const fs = require('fs')
const {makeVisualGridClient} = require('@applitools/visual-grid-client')
const {getProcessPageAndSerializeScript} = require('@applitools/dom-snapshot')
const puppeteer = require('puppeteer')

describe('visual-grid-client test', function() {
  let visualGridClient
  let closePromises = []
  let processPageAndSerialize
  let browser
  let page

  before(async () => {
    browser = await puppeteer.launch()
    page = await browser.newPage()

    visualGridClient = makeVisualGridClient({
      showLogs: true,
    })

    processPageAndSerialize = `(${await getProcessPageAndSerializeScript()})()`
  })

  after(() => {
    await browser.close()
    return Promise.all(closePromises)
  })

  let checkWindow, close
  beforeEach(async () => {
    ;({checkWindow, close} = await visualGridClient.openEyes({
      appName: 'visual grid client with a cat',
      testName: 'visual-grid-client test',
    }))
  })
  afterEach(() => closePromises.push(close()))

  it('should work', async () => {
    await page.goto('index.html')
    const {cdt, url, resourceUrls, blobs, frames} = await page.evaluate(processPageAndSerialize)
    const resourceContents = blobs.map(({url, type, value}) => ({
      url,
      type,
      value: Buffer.from(value, 'base64'),
    }));
    checkWindow({
      tag: 'first test',
      sizeMode: 'viewport',
      url,
      cdt,
      resourceUrls,
      resourceContents,
      frames
    })
  })
})

Contributing

Generating a changelog

The best way is to run npm run changelog. The prerequisite for that is to have jq installed, and also define the following in git configuration:

git config changelog.format "* %s - %an [[%h](https://github.com/applitools/visual-grid-client/commit/%H)]"
You can’t perform that action at this time.