Issues I had:

• TargetedWebDriverFactory was redundant with TargetLocator interface
• WebDriverAlert worked differently than Browser and Elements (though you could argue it was simpler if only less consistent)
• CachingTargetLocator did not cache an alert target.

Talking with @jvervaec this morning, the discussion of Selenium WebDriver version bumps came up. Occasionally there is a PR for bumping the version of WebDriver that gets merged in hopes of fixing some issues a particular consumer may be having with such things as what versions of Firefox a certain WebDriver build supports. Sometimes the most recent WebDriver version is backwards compatible with a certain number of Firefox versions but other times, WebDriver cuts support for older versions of Firefox that consumers may still be using reliably.

The main issue here is that when the WebDriver dependency is bumped and merged, the darcy-webdriver version remains the same even though I can see that making some sense since it's SNAPSHOT'd. However, if someone was doing just fine on something like darcy-webdriver 0.3.2-SNAPSHOT that used selenium-webdriver 2.50.0 but then WebDriver was bumped and darcy-webdriver remained 0.3.2-SNAPSHOT they may run into issues when using the latest artifact.

In my opinion, I think that there should be some sort of darcy-webdriver version bump as well when the WebDriver dependency is bumped so as to avoid consumers running into possible compatibility issues that may arise unknowingly. I was taking a look at semver.org to try and get an idea of what should be bumped and my guess would be either the MINOR or PATCH version. Either way, I believe there should be a bump in there somewhere.

Given Synq.after(someElement::click) .expect(transition().to(new View()).inNewContext())
.waitUpTo(10, ChronoUnit.SECONDS);

Throws com.redhat.darcy.ui.DarcyException: Cannot find contexts of type: interface com.redhat.darcy.ui.api.FindableElementContext at com.redhat.darcy.webdriver.internal.TargetedWebDriverParentContext.findByView(TargetedWebDriverParentContext.java:112) at com.redhat.darcy.webdriver.internal.DelegatingWebContext.findByView(DelegatingWebContext.java:356) at com.redhat.darcy.webdriver.WebDriverBrowser.lambda$29(WebDriverBrowser.java:317) at com.redhat.darcy.webdriver.WebDriverBrowser$$Lambda$64/1060042118.get(Unknown Source) at com.redhat.darcy.webdriver.WebDriverBrowser.attemptAndGet(WebDriverBrowser.java:342) at com.redhat.darcy.webdriver.WebDriverBrowser.findByView(WebDriverBrowser.java:317) at com.redhat.darcy.ui.By$ByView.find(By.java:520) at com.redhat.darcy.web.internal.AbstractWebSelection.contextOfType(AbstractWebSelection.java:43) at com.redhat.darcy.ui.api.ContextSelection.elementContext(ContextSelection.java:44) at com.redhat.darcy.ui.internal.NewContextTransitionEvent.(NewContextTransitionEvent.java:36) at com.redhat.darcy.ui.internal.SimpleTransitionEvent.inNewContext(SimpleTransitionEvent.java:57) at com.redhat.darcy.ui.internal.SimpleTransitionEvent.inNewContext(SimpleTransitionEvent.java:47)

#72

FYI, it looks a number of the BrowserDriver constructors have been deprecated in selenium 3. Probably should also deprecate those in the darcy browser factories, but let me know what you think.

This allows us to use FF47.

The SessionNotFoundException I removed was changed to extend NoSuchSessionException and was deprecated. It's replacement NoSuchSessionException is a subclass of NotFoundException.

MultiEvents spin up separate threads for each event. WebDriver is not thread safe. This could be a big problem if both events are interacting with the driver.

Because there is a lot of element caching going on in the webdriver module, it is conceivable that StaleElementReferenceExceptions could become an issue.

We need to evaluate if it's an issue worth solving, and if so how best to solve it.

Inspired from the Dart style guide,

If you really find yourself wanting lines longer than 80 characters, our experience is that your code is likely too verbose and could be a little more compact. Do you really need to call that class AbstractWidgetFactoryManagerBuilder?

Selenium 3 has been out for awhile now. It appears to have lots of useful new features. Selenium 2 also seems to have stopped supporting Firefox versions greater than 47.

Feel free to assign this one to me. Will open a PR shortly.

This pull request is focused on resolving occurrences of Sonar rules squid:S1155 - Collection.isEmpty() should be used to test for emptiness. squid:S2293 - The diamond operator ("<>") should be used. squid:S2259- Null pointers should not be dereferenced. You can find more information about the issue here: https://dev.eclipse.org/sonar/rules/show/squid:S1155 https://dev.eclipse.org/sonar/rules/show/squid:S2293 https://dev.eclipse.org/sonar/rules/show/squid:S2259

Please let me know if you have any questions.

Faisal Hameed

This pull request is focused on resolving occurrences of Sonar rule squid:S1213 - The members of an interface declaration or class should appear in a pre-defined order. You can find more information about the issue here: https://dev.eclipse.org/sonar/rules/show/squid:S1213

Please let me know if you have any questions.

Faisal Hameed

This pull request is focused on resolving occurrences of Sonar rule squid:S1206- "equals(Object obj)" and "hashCode()" should be overridden in pairs You can find more information about the issue here: https://dev.eclipse.org/sonar/rules/show/squid:S1206

Please let me know if you have any questions.

Faisal Hameed

Thank you for open sourcing such a awesome project. There is so much to learn from project,

It would be great if you can add some sample project that anyone can follow along.

Thank you and keep up the great work!

Happy New Year!!

The result can produce exceptions which are a little hairy to understand:

FAILED: [REDACTED]
com.redhat.darcy.ui.FindableNotPresentException: Can't interact with something if it is not present.
Tried to interact with a A WebDriverLink backed by, An element found nested within another element.
The parent element is a A WebDriverElement backed by, An element found nested within another element.
The parent element is a A WebDriverElement backed by, An element found nested within another element.
The parent element is a A WebDriverElement backed by, An element found via, By.xpath: //div[contains(concat(' ', normalize-space(@class), ' '),' bRelatedList ')][.//td[@class='pbTitle']//*[text() = '[REDACTED]']]
in search context, com.redhat.darcy.webdriver.internal.ForwardingTargetedWebDriver@4729d1.
The locator to find the child element is ByXPath: {xpath='.//div[@class='pbBody']/table'}.
The locator to find the child element is ByXPath: {xpath='./tbody/tr[2]/td[1]'}.
The locator to find the child element is ByLinkText: {linkText='Del'}
        at com.redhat.darcy.webdriver.elements.WebDriverElement.webElement(WebDriverElement.java:205)
        at com.redhat.darcy.webdriver.elements.WebDriverElement.attempt(WebDriverElement.java:158)
        at com.redhat.darcy.webdriver.elements.WebDriverElement.click(WebDriverElement.java:87)
        at com.redhat.darcy.salesforce.StandardActions$$Lambda$135/1517758613.run(Unknown Source)
        at com.redhat.synq.ActionEvent.waitUpTo(ActionEvent.java:42)
        at com.redhat.synq.ActionEvent.waitUpTo(ActionEvent.java:31)
        at com.redhat.synq.SequentialEvent.waitUpTo(SequentialEvent.java:48)
        at com.redhat.synq.Event.waitUpTo(Event.java:131)
        at com.redhat.darcy.salesforce.StandardActions.lambda$delete$1(StandardActions.java:45)

FirefoxDriver won't type something into a file select box that isn't an absolute path. I'm not sure if other drivers work this way, but we should investigate and throw an appropriate exception in cases of unsupported input, instead of simply doing nothing (which is webdriver's default behavior).

I'm not incredibly happy with the unit tests.

Besides that many had to just be eliminated due to refactorings and have not yet been replaced, I think we need functional tests too. This can present edge cases more naturally and actually catch real bugs before they are found by real users.

Need...

• Decide on a WebDriver implementation to test against. Many of them? Firefox? PhantomJS? HTMLUnit?
• Configuration to run real browsers - Travis can integrate with SauceLabs.
• Test pages
• Code abstractions for accessing these test resources, starting and closing browsers
• The tests