IllegalArgumentException: To search for element by tag, use By.tag() instead of tag as CSS selector.

[javydreamercsw]

Looks like this needs to be updated to use By.tag() instead.

at de.retest.web.selenium.TestHealer.retrieveUsableCssSelector(TestHealer.java:161)
	at de.retest.web.selenium.TestHealer.findElementByCssSelector(TestHealer.java:139)
	at de.retest.web.selenium.TestHealer.findElement(TestHealer.java:66)
	at de.retest.web.selenium.TestHealer.findElement(TestHealer.java:49)
	at de.retest.web.selenium.UnbreakableDriver.findElement(UnbreakableDriver.java:81)
	at de.retest.web.selenium.AutocheckingRecheckDriver.findElement(AutocheckingRecheckDriver.java:107)
	at jdk.internal.reflect.GeneratedMethodAccessor6.invoke(Unknown Source)
	at java.base/jdk.internal.reflect.DelegatingMethodAccessorImpl.invoke(DelegatingMethodAccessorImpl.java:43)
	at java.base/java.lang.reflect.Method.invoke(Method.java:566)
	at org.openqa.selenium.support.events.EventFiringWebDriver.lambda$new$1(EventFiringWebDriver.java:105)
	at com.sun.proxy.$Proxy17.findElement(Unknown Source)
	at org.openqa.selenium.support.events.EventFiringWebDriver.findElement(EventFiringWebDriver.java:194)

[javydreamercsw]

This failure only happened for me on Mac 10.14 Safari, Windows 10 Firefox

[roesslerj]

Thanks for raising this issue. I addressed it in #304. However, it will only work for simple CSS selectors, not for complex ones...

[javydreamercsw]

Did this made it into 1.5.0 or 1.5.1? Still seeing this issue on 1.5.1.

[diba1013]

We proposed a fix with the PR mentioned above. If this does not fix your problem, I kindly ask you to describe your setup.

Apparently you are trying to find an element by a css selector. However, we currently support only basic selectors, starting with # or . (e.g. #foo, .bar) and not contain additional selectors. If this is not the case, we throw above exception.

For now, I would suggest that you convert your css selector into a xpath selector or use a different By if possible.

[roesslerj]

Or could you just give the complete findElement expression that triggers that error?

[roesslerj]

... to properly answer your question: The fix (#304) made it into the release. Therefore if it still occurs, we need to refine that by better understanding your concrete problem.

[javydreamercsw]

This only occurs when using the AutocheckingRecheckDriver. Basically this is the code at fault:

@Override
  public Object scrollIntoView(WebElement element) {
    try {
      return ((JavascriptExecutor) getWebDriver())
          .executeScript("arguments[0].scrollIntoView(true);", element);
    } catch (Exception ex) {
      LOG.debug("Unable to scroll into view: " + element, ex);
    }
    return null;
  }

The return line leads to the following stack trace:

java.lang.IllegalArgumentException: To search for element by tag, use `By.tagName()` instead of `tag` as CSS selector.
	at de.retest.web.selenium.TestHealer.retrieveUsableCssSelector(TestHealer.java:177)
	at de.retest.web.selenium.TestHealer.findElementByCssSelector(TestHealer.java:155)
	at de.retest.web.selenium.TestHealer.findElement(TestHealer.java:75)
	at de.retest.web.selenium.TestHealer.findElement(TestHealer.java:51)
	at de.retest.web.selenium.UnbreakableDriver.findElement(UnbreakableDriver.java:84)
	at de.retest.web.selenium.AutocheckingRecheckDriver.findElement(AutocheckingRecheckDriver.java:136)
	at jdk.internal.reflect.GeneratedMethodAccessor1.invoke(Unknown Source)
	at java.base/jdk.internal.reflect.DelegatingMethodAccessorImpl.invoke(DelegatingMethodAccessorImpl.java:43)
	at java.base/java.lang.reflect.Method.invoke(Method.java:566)
	at org.openqa.selenium.support.events.EventFiringWebDriver.lambda$new$1(EventFiringWebDriver.java:105)
	at com.sun.proxy.$Proxy31.findElement(Unknown Source)
	at org.openqa.selenium.support.events.EventFiringWebDriver.findElement(EventFiringWebDriver.java:194)
	at com.github.webdriverextensions.internal.WebDriverExtensionElementLocator.findElement(WebDriverExtensionElementLocator.java:37)
	at org.openqa.selenium.support.pagefactory.internal.LocatingElementHandler.invoke(LocatingElementHandler.java:38)
	at com.sun.proxy.$Proxy34.toString(Unknown Source)
	at java.base/java.lang.String.valueOf(String.java:2951)
	at java.base/java.lang.StringBuilder.append(StringBuilder.java:168)

[roesslerj]

Hey, thanks for providing more input. Sorry for still not understanding ... where is that scrollIntoView method from? Is this recheck code? And how and when is it called for leading to that stack trace?

[javydreamercsw]

This is part of my platform code. Seems that using a WebElement using Selenium exclusive methods in something like this:

private WebElement waitForClickable(int seconds, WebElement webElement) {
    WebDriverWait wait = new WebDriverWait(getWebDriver(), seconds);
    return wait.until(ExpectedConditions.elementToBeClickable(webElement));
  }

Triggers the error when AutocheckingRecheckDriver takes over. The WebElements are triggering the stack trace above defined like this:

@FindBy(css = "[data-id=\"id\"]")

[javydreamercsw]

FYI this showed up during the monitoring phase of the tests implemented using the tool and it's currently blocking any progress towards a decision in using the tool as part of our tests.

[diba1013]

We are working on a workaround in 1.6.0, so that no exception will be thrown and your test will continue. We will try—however, I cannot promise—to have a solution for the next release that makes your element unbreakable. So to say, the exception should not be thrown with the next release.

In the meantime you may use another By like By.id that is currently supported, if your element allows so.

[javydreamercsw]

I understand but it is a strict requirement to use data-id tag.

…

On Thu, Oct 24, 2019, 1:10 PM Bastian Diehl ***@***.***> wrote: We are working on a workaround in 1.6.0, so that *no* exception will be thrown and your test will continue. We will try—however, I cannot promise—to have a solution for the next release that makes your element unbreakable. So to say, the exception should not be thrown with the next release. In the meantime you may use another By like By.id that is currently supported, if your element allows so. — You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub <#303?email_source=notifications&email_token=AAVP2P2C5HM5YZNVU7ATHQLQQHQI3A5CNFSM4IIKHIMKYY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOECF55LA#issuecomment-546037420>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AAVP2PYGBZTZMY2OVGU2OXLQQHQI3ANCNFSM4IIKHIMA> .

[roesslerj]

As mentioned, we fixed the exception being thrown. Supporting more complex CSS attributes will be tracked in #389 and #390.

[diba1013]

We partially implemented this in recheck-web-1.6.0 for which we have a new release available. If you find any issues regarding this problem, please reopen this ticket and post the exception and/or log here.

Please us the following versions:

• recheck-web-1.6.0 with either
• recheck.cli-1.6.0 (Open Source) or
• review-1.6.0 or review-1.6.0-win (Paid)

Edit: Updated to most up to date releases.

[roesslerj]

By "partially implemented", diba1013 meant that it should work for you case specifically, and for most cases that use CSS selectors ... but not for each edge-case of possible CSS selectors... Please report any remaining problems on that. Thanks

[javydreamercsw]

Haven't had a chance to try it. Was not a drop in replacement as some classes were missing. Is there a recheck-web beta with this changes?

[diba1013]

We released a full release 1.6.0 already which should be on maven central by now. However, it is still not visible in maven repository. Sadly, we have no influence on when the artifact is available, as we have no open release in sonatype.

Could you describe what classes are missing? If those are essential classes (like RecheckImpl or RecheckOptions) this is most likely that the dependency could not be resolved (i.e. missing). Could you check in your respective library view in eclipse or IntelliJ?

[javydreamercsw]

This morning seems it was resolving fine. Still need to do some more tests but at least seems to be resolve now.