[Feature] Adding waitFor Locator Helpers that do not fail after timeout

[sustainjane98]

I am currently testing a Web Application, which has a lot of optional processes in it (f.e. an Box just appears in a process in special cases, but takes some time)

For this I wanted to use the waitFor Locator Function, but it always ends the test with an error, like it should. But maybe it would be handy to support a flag on the existing waitFor, waitForSelector etc Functions to just keep the element as optional. So that I can do something like this:

test("Example for WaitFor", async ({page})=>{

 const exampleNotAlwaysAvailableElement = page.getByTestId("example-elem").waitFor({timeout: 5000,  noFail: true});

 if(exampleNotAlwaysAvailableElement) {
   // Do whatever needed to handle the case when this element somehow appears in time
   
}

// Continue here when not


})

[dgozman]

@sustainjane98 You can try/catch to achieve the same:

test("Example for WaitFor", async ({page})=>{
  try {
    await page.getByTestId("example-elem").waitFor({ timeout: 5000 });
    // Do whatever needed to handle the case when this element somehow appears in time
  } catch {
    // Continue here when not
  }
})

[sustainjane98]

I tried this, but it seems like the catched error is still visible in trace viewer.

So it is more like something has an error but continue but I need more like an if it exists else cause. So it is no error when it doesn't exist.

Also a try catch without error handling is bad code in general.

[larsheylen]

I typically handle this situation by using a Promise.race() construction like the following:

await Promise.race([
   page.getByTestId("example-elem").waitFor({ timeout: 6000 }),
   page.waitForTimeout(5000)
]);

This way either the element you are waiting for resolves within 5s or the timeout gets resolved first and the test continues.

I do agree however that it would be useful to have this as an option of the waitFor function.

[subpardaemon]

chiming in regarding this: i have a few situations where either A, B or C could occur, and whether any of them do is outside of the test's jurisdiction. i do the Promise.race solution but the downside is that it puts a lot of garbage in the log (failures which don't affect the overall success of the test). i would really welcome a throw: false option or something like that in the .waitFor() method.

[Arsou33]

You can wait for one of several locators. This can be useful in some situations. For my case which is to wait for either a locator or the closing of the popup, I have not yet found an adequate solution.

// Wait for a locator
assertThat(accountButton.or(emailInput).or(userAnotherAccountButton).first()).isVisible();

[giangtb7798]

async isElementVisible(locator: Locator, timeout: number, interval: number = 1000): Promise<boolean> {
        while (timeout > 0) {
            if (await locator.isVisible()) {
                return true;
            }
            await this.page.waitForTimeout(interval);
            timeout -= interval;
        }
        return false;
}

Just implement this function it would skip if element not found and also step would not been mark as failed step in both trace viewer and allure report