Dullahan

Work in progress, lots of (currently) private code is being transferred bit by bit :-)

Browser Support

We test against a wide range of browsers to make sure Dullahan doesn't accidentally break something that used to be work. However, we cannot test against every browser version, as that would either be crazy expensive or take weeks to complete between every release. For example, all releases are tested to work with Edge 12 and Edge 18, but versions in between are not tested. If both Edge 12 and Edge 18 work, we assume Edge 13 through 17 will also work.

The versions listed in this table are the versions that Dullahan is guaranteed to work with.

	Selenium 3	Selenium 4	Puppeteer	Playwright
Internet Explorer	11	11	X	X
Edge (EdgeHTML)	12 - 18	12 - 18	X	X
Edge (Chromium)	?? - ??	?? - ??	?	?
Firefox	?? - latest	?? - latest	nightly	nightly
Chrome	?? - latest	?? - latest	latest	latest
Chromium	?? - ??	?? - ??	latest	latest
Safari	?? - ??	?? - latest	X	X
WebKit	?? - ??	?? - ??	X	latest
Opera (Presto)	?? - ??	?? - ??	X	X
Opera (Chromium)	?? - ??	?? - ??	?	?

There are a few considerations to take into account:

Headless mode is not supported by Internet Explorer 11;
Offline simulation is not supported by Selenium 3 and Selenium 4;
Firefox support for Puppeteer and Playwright is still experimental and should be considered unstable;

Adapters: Selenium, Puppeteer and Playwright

Selenium supports an incredibly wide range of browsers, but not all browsers support the same Selenium versions and the drivers for each browser are horrible inconsistent in behaviour. Puppeteer and Playwright are a lot more consistent and slightly faster, but do not support as many browsers and also don't play well with tools like Browserstack.

Instead of creating yet another protocol, Dullahan simply wraps around these existing protocols with a custom API. Each function in Dullahan's API is tested to behave exactly the same for each of these protocols, allowing you to switch between them at will, without having to change any of your code.

More information can be found in each adapter's readme file.

Runners: Standard, Development and AWS-Lambda

Dullahan uses runners to determine how tests will be run. The simplest of the runners is the Standard runner, which runs all of your tests (if they match the criteria from your config) and then simply stops. It can run them in series, or in parallel, but in the end it still just means "start - run all tests - stop".

That's nice if you want to run tests, but when you're writing a new test that can be incredibly annoying. In this case, you'll likely want to use the Development runner, which doesn't run anything until you change a test and then runs only that test. The flow with this runner can be described as "start - run a test whenever it changes - stop on user request"

When it comes to running hundreds, or even thousands of tests, the Standard runner likely isn't enough. Imagine opening 100 instances of Google Chrome at the same time on your laptop, it will probably crash. There's a limit to concurrency and that's generally tied to the device running the tests. This is where more specialized runners can save the day: the AWS Lambda runner can easily handle hundreds of tests all running at the same time, as each test will be executed on its own temporary "server". Thousands of tests can be completed within a few minutes, mostly depending on how many active users your website can handle at any given time.

More information can be found in each runner's readme file.

Plugins: AWS S3

@k2g/dullahan-plugin-aws-s3

... TODO

Plugins: HTML, Markdown, JSON