Interacting with browser using TestCafe APIs but without TestCafe test runner

[anon46571079]

Hi there!

I am looking for a pure Node Selenium alternative and Testcafe has really caught my eye. However I am looking to interact with the browser from within my own test runner and without really using the TestCafe runner. Is this something that is possible? And if so can you point me to an entry point in your code from where I can enable this for my projects?

Thanks!

[anon55777542]

I think what you're looking for is the programming interface:

https://devexpress.github.io/testcafe/documentation/using-testcafe/programming-interface/

[anon493717]

Hi @anescobar1991,

TestCafe opens and closes browsers itself before/after running a test and interacts with the browser via Test API commands.
I didn't clearly understand what type of interaction with a browser you are going to do with your own test runner and why you need it instead of using the built-in TestCafe functionality?

[anon46571079]

What I mean is I am already using a test runner (Jest) for unit and integration tests and if I want to keep using it for functional tests then I should be able to decouple the TestCafe test runner, reporter, etc... from the browser interaction API. So that I can do:

it('should demonstrate usage of TestCafe within any test runner', async () => {
  const testCafeConfig = { ... };
  const browser = TestCafe.initiatlize(testCafeConfig);
  await browser
    .typeText('#developer-name', 'John Smith')
    .click('#submit-button');
  ...
});

Is this possible today or should I work on adding this functionality myself? I haven't spent too much time looking at the source code yet but I can do that and see where this could be possible if it isn't already.

[anon493717]

Thanks for clarification. These parts of TestCafe are closely integrated with each other so I don't see a way to use them separately.
Is there a reason why you are not satisfied with the TestCafe runner, or you just like to use runner which you used to use?

[anon46571079]

No what you guys are doing is great! I'm glad there's another option to Selenium that is purely JS.

The reason we are wanting to use our test runner as opposed to TestCafe is that we want our developer experience for testing to be frictionless and so we want to be able to use one test runner to run our unit, integration, and browser functional tests. We already have a pretty sizeable ecosystem of modules that we use for our testing that we want to reuse . I can already use selenium or even Nightmare for our browser functional test needs in our test runner but I really really would rather not have our engineers need to leave the Node ecosystem in order to have browser tests.

It would be nice to see the "browser API" part of TestCafe decoupled from the test runner so that it can be test framework agnostic. In meantime I am trying to hack something together to work for me. I can open PR if you would like once I am done (assuming I succeed).

[anon86331989]

I am interested in this feature too, let me describe the use case: I'd like to execute tests with my own runner (jest) where different browsers communicate with each other using WebRTC.

One browser executes some action and then the other browsers can expect some UI changes.

Does this make sense?
Is it feasible?

Best,
Gorka

[MarinaRukavitsyna]

@Gorka
We do not have immediate plans to support multiple browser windows and running parts of the same test in different browsers. While the following feature is on our TO-DO list, I cannot give you any ETA when it will be added:

Add the capability to perform testing in multiple browser windows

[anon86331989]

Marina, thanks for the reply.

Too bad, I have to use selenium-webdriver directly.

I actually managed to spawn a FF browser with TestCafe but was not able to control it, maybe in the future.

Thanks for the amazing software and effort!

[churkin]

Hi @Gorka,
I suggest you use the approach applied in this demo. The main idea of this is to create an empty test that will capture test runner context and will wait commands from you. It allows you to create some TestCafe instances that work with different pages and wait commands from your application. Please refer to here to see how to create mock tests and catch test runner context. Please contact us if you use this approach and have any questions.

[anon86331989]

@churkin thanks for the ideas, I am currently solving this using selenium-webdriver and opening two drivers with two pages (or more) simultaneously.