Wow! This was great! So hard to find good Playwright videos like there are for Selenium (Java, C#, Python). A few videos are out there for creating setup and teardown, but they are in the same file as the testcases are. Your video went at a good face pace and while watching it I started to notice it was about to end. You didn't yet show how to put the code in a separate file so I thought it wasn't going to be covered , but you covered it! Thanks so much!
My God this is what I was looking for! That's funny because I didn't find this clip at once, I stumbled on the clip about parallelism because google pointed out me that topic for my question. Thank you very much, your examples are very clear and straightforward
The color theme is Yi Dark and if you're after the custom font it's described in this article 👉 www.stefanjudis.com/blog/how-to-enable-beautiful-cursive-fonts-in-your-vs-code-theme/
Hi, I have started exploring playwright very recently and found your videos on the topic . Kudos for creating these :-). I have a question. If there are more than one reusable functions to use inside the test, how do you do it then?
Custom fixtures aren't bound to a single one. You could add more if you wanted to. :) Add more properties to the `extend` object and you're good to go. :)
Nice tutorial, thank you. Have one question. Currently on project we have POM and UtilClass for common methods we use. And usually I just create method in POM and then I can re use it in my test files. Looks like creating custom fixtures is more or less the same as storing method in Page Object file and then execute it if needed. Or are there any other advantages of using custom fixtures? Thanks
Yes, this is correct. There's nothing particularly wrong with importing utils/poms within your test files. But I found that by relying on fixtures a project becomes more structure and does things "the playwright way". One advantage is that if you're using classes that need to be initialized you can hide this and don't neet to care about it. :) 👇 ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-k488kAtT-Pw.html
@@ChecklyHQ Thank you for your reply, appreciate it! Speaking about initializing classes... we already hide them with pageManager. There is a function, that returns class object, so we call that function in test file. I agree about "the playwright way". Looks like there are few ways to achieve same result. Agree, that fixtures look nice. And thanks for sharing your tips and playwright insides. Before fixtures were unclear to me... but now I understand them a bit better. Thanks for your effort!
@@glebmirosnikovs6722 Oh for sure - there are plenty of ways to abstract things. :) Playwright is also just JS/TS at the end. Happy to hear that the videos have been valuable. If you have more questions or comments, I'm always looking for new ideas. ;)
What’s the advantage of using a fixture as opposed to a regular function? You could simply define const login = async (page) => { // Do all the stuff } and then simply call `await login(page)` at the beginning of the test.
That's right Guillaume and a great question. In this simple example scenario, a helper method definitely works. 👍 Fixtures enable you to follow the PW conventions and group helper/utility code like the framework does. I consider this to be a great advantage. Additionally, and considering your example: if you'd want to perform a teardown action (maybe you want to log out after) you'd already have two methods to pull in every test case and the number of util functions grows. With a fixture-based approach, you could put setup (before the 'use' call), the action and teardowns (after the 'use' call) into a single place. Hope this helps. :)
Very nice channel with good and simple explanations! How does this compare to a Page Object method? Is it better to use one over the other or maybe in combination?
Thanks Aleksandar. I don't think it's an either-or decision. Fixtures play nicely with POMs. I see their main advantage in being to use "native PW functionality" and the ability to have setups / teardowns next to the code running in your tests. So if you're POMs have some sort of required clean-up steps, i'd def go for a fixture setup. Hope this helps!
@@stefanjudis I am using fixtures to create POM objects, e.g. myAccountPage: async ({ page }, use) => { await use(new MyAccountPage(page)); }, so that I don't repear this useless code every time in test script
Great video, Is it possible to create a base fixture that is common across all your tests but then extend the base fixture with something more specific that can target a subset of tests?
@@ChecklyHQ Thanks I found a solution where I create a baseFixture that holds everything that is common to all my tests and then I have fixtures that extend the base fixture that targets a subset of tests and its working perfectly. I did try the test.use() but in my use case it did not work
@@ChecklyHQ I was getting conflicts in the api request where 1 of the 2 requests was bad as they conflicted, so iv separated the conflicts into 2 new fixtures that both extend from the base so i dont have to duplicate the common stuff
Great stuff , Thanks , but how can we utilize this is page object model and maintain the same page state as page instance will be different hence different instances will be launched , can u plz show , thanks
How would you see that as different from just using .before that will already run before every test. (Or login session is better put in config where playwright has options for putting it per project/browser or for default of all. Saving in a json file. )
It depends on the case I'd say. If you have a single `spec` file before/after does the trick for sure. And also, when it comes to saving the login logic, you also make a valid case buy saving the browser state. All that said, I often prefer fixtures, because they "hide" functionality from my spec files. I don't have to worry about fiddling with before/after or anything. Additionally, I can also configure them with fixture options. Implementing fixtures just feel like a "native Playwright approach". And I like this a lot. But, of course and as always, it's a classical "it depends" and a matter of taste. ;)
thanks the videos Is there possible that we use different storagestate inside one test between test.steps? e.g.: test.step.use({storageState: '1.json '})... test.step.use({storageState: '2.json '})?
Hey Dániel, I just had a look and I don't think that's possible. But if you're testing your sites with two different states (such as a regular and admin one) I'd also favor fixtures. `test('something', ({page, AdminPage})` Hope this helps. :)
@@stefanjudis thanks, i am doing this now, my test login admin... logout admin login editor... logout editor login admin, that i can delete the item what do u suggest how could i avoid so much login,logut?
@@danielkovacs3664 Hey Daniel, in this case I'd login admin and editor so that you don't have to switch and log in/out all the time. Similar to this: `test('something', ({page, AdminPage})` That said though, here at checkly we have a good experience with cleaning up created resources on the API level. blog.checklyhq.com/how-low-level-api-calls-stabilize-your-end-to-end-tests/
is there a way we can enable and disable the fixture based on run time values like flag to enable or disable? Scenario: if the login/SSO is required for higher env's and not for lower env
Enabling and disabling fixtures doesn't feel right to me. I'd rather make the fixture behave differently depending on run time flags. :) Hope this helps.
You could always structure your project to use different fixture with different exported `test` object. And depending on the fixture differences you could also make fixtures configurable with fixture options. It's a bit tough to give advice without knowing specific. :)
Hello, How can you return something form Webapp fixure? I am trying to copy text from text field on clipboard and then eceluate it and return it to my test,
This doesn't sound like a fixture case for me. I'd instead extract the logic into a separate function for the case of evaluating text in the clipboard. That said, you could always change the fixture to be a nested object or a function. It's up to you what to call `use` with. :)
You can create a similar fixture called "accessToken", do whatever needs to happen to access the token and then pass it to your tests via "use(token)". Hope this helps. :)
@@ChecklyHQ ok, Thanks, That was a quick help. at the same time another question I have json(playload) stored in a file and I want one Field (RefNumber) has a random value every time How can I provide a random value every time to the RefNumber in json? for exmple in follwing workOrderPayload is the work load and I was to provide RefNumber as a random to the Json everytime const response = await request.post(workOrderUrl, { headers:{ "Accept" : "application/json, "Content-Type" : "application/json "Authorization" : `Bearer ${process.env.TOKEN}` }, data: workOrderPayload })
@@salikhanan I'm not sure I understand. Could you use `Math.random()` and alter `workOrder` to include a different `RefNumber`? That said though, let's stay focused on this video here. :) For questions around synthetic monitoring with Playwright and Checkly we're always happy to help at checklyhq.com/slack/. :)
It's FiraCode iScript and if you're looking for the same editor setup it's described on my blog: www.stefanjudis.com/blog/how-to-enable-beautiful-cursive-fonts-in-your-vs-code-theme/