This tool goes through a list of tasks with URLs or click paths, makes screenshots of them and compares them.
You need to install selenium and the browsers you want to test.
npm install burdamagazinorg/sharpeye
Configure your project as described in this document
Execute./node_modules/.bin/sharpeye
once to create reference images.
If you execute it again, the new screenshots will be compared to the reference.
(You can also add sharpeye
to the scripts section in you package.json
)
If there is a difference, the diff image will be placed in the screenshot subdirectory diff
.
Create a file sharpeye.conf.js
and sharpeye.tasks.js
in your project directory, with following contents:
sharpeye.conf.js:
// Options for sharpeye
exports.options = {
// The base URL of the website.
baseUrl: 'http://thunder.dd:8083',
// Specify directories, in which screenshots should be saved.
// They will get a postfix of '/screen', '/reference' and '/diff', respectively.
screenBaseDirectory: process.cwd() + '/screenshots',
referenceBaseDirectory: process.cwd() + '/screenshots',
diffBaseDirectory: process.cwd() + '/screenshots'
}
// Webdriver.io config overwrites.
// See the sharpeye.conf.js file in the root of the sharpeye module
exports.config = {
// Define, which browser you want to use
// See: https://github.com/SeleniumHQ/selenium/wiki/DesiredCapabilities
capabilities: [
{
browserName: 'firefox'
},
{
browserName: 'chrome'
}
]
};
sharpeye.tasks.js:
For taking a full-page screenshot of a URL, just specify the path as string. e.g.:
/my/path
To click/fill in some elements and take a screenshot afterwards, you can specify an object, with following properties:
name
: The name of the actions package. Will be used for the screenshot filenamepath
: The URL path to start fromelement
: (optional) The element, from which a screenshot should be takenviewport
: (optional) Whether the viewport should be captured, instead of the whole pagenoScreenshot
: (optional) Set to true, to prevent taking a screenshot. Might be used e.g. for loginsactions
: An array of objects, which specify, where to click and for what to wait for
The action objects have following properties:
$
: The selector for the DOM element, on which should be clicked, or which should be selected. Also works for options in select lists.wait
: (optinal) The element, which should be waited for, after clickingfill
: (optional) A string, which should be filled into the selected elementwaitBefore
: (optional) time in milliseconds, to be waited, before clickingoffset
: (optional) an offset in y direction, to be scrolled (useful, when elements are hidden behind floating elements)
The actions array can also contain an object, which switches the context to another frame. This object can have following properties:
switchToFrame
: ID of the (i)frame, to switch to, ornull
, to switch back to the default framewait
: After switching, element to wait for, before continuing
module.exports = [
'/admin/content',
{ name: 'name of the click path',
path: '/node/add/article',
[element: 'selector for DOM-element from which the screenshot should be made'],
[viewport: true // makes screenshot of the viewport, instead of the whole page],
[noScreenshot: true // Makes no screenshot at all]
actions: [
{ $: 'DOM selector', wait: 'DOM selector to wait until visible', [waitBefore: 'optional time in milliseconds, which should pass, before clicking']},
{ $: 'DOM selector', fill: 'My value to be filled into the DOM element' },
{ switchToFrame: 'ID of frame', wait: 'DOM selector to wait until visible inside frame' },
{ switchToFrame: null } // switch to default context
]},
]
Different types of selectors can be used: http://webdriver.io/guide/usage/selectors.html