How to get latest object in PageFactory without/with doing InitElements? - selenium-webdriver

I am using PageFactory model in selenium framework (WebDriver, Java).
Test: Create an entity in application and wait until it will get complete ( To verify state of entity (completed/in-process) , I have to verify status image in-front of created entity).
Issues/Queries:
** How should I wait for specific status should get change from In-process to Complete? (As whole page is not getting refresh after completion, only that image is getting refresh – AJAX/JSON Call)
***How should I get latest page objects using PageFactory? I have tried using PageFactory.InitElements(driver, classname.class), but it’s not working.
**Is any way, we can refresh only PageFactory Cache? Like SilkTest having ‘BrowserPage.FlushCache()’?

The WebElements defined with the #FindBy annotation are effectively proxies. Every time you perform an action with the WebElement it uses the locator in the #FindBy annotation to find that element on the page again.
PageFactory.InitElements(driver, classname.class) just sets up these proxies, it doesn't actually find the elements so you don't need to keep calling it again and again.
You can cache individual WebElements by using the #CacheLookup annotation, if you use this the WebElement will be lazily evaluated the first time that you utilise it, but from that point onwards it will not be looked up again (which means you could start getting StaleElementExceptions if the element changes).
The latest information is kept in the Selenium Wiki:
http://code.google.com/p/selenium/wiki/PageFactory

Related

Why explicit wait is waiting for entire wait time even though element is visible/clickable before the wait time is over

I have given a wait time for 60 sec, and i am using wait until element to be clickable/visible but script is waiting for whole 30 secs even though the element is visible on UI and also clickable?
I tried using latest selenium version, tried using different waits also using different locators. but it did not work
The reasons could be in:
The element belongs to iframe so you need to switch to the iframe prior to attempting locating anything inside it
The element belongs to Shadow DOM so you need to locate ShadowRoot object, cast it to the WebElement and find the sub-element you want to click
Your locator is not correct, try getting the page source and saving it to a file. Once done use your favourite browser developer tools to locate the elemen
The syntax of your Explicit Wait is not correct. Check out Explicit Waits and How to use Selenium to test web applications using AJAX technology for code examples
Going forward consider adding page source and your code to your question as the chance you will get the comprehensive answer will be much higher, otherwise we have to go for "blind shots"

Oracle ADF: Refresh Form data

I am developing a web app using Oracle ADF. I have a bounded task flow. In that I have a search page like below.
I have created the above two forms using view object data controls.
Searching is performing well. But my problem is when I go some where else in my application using menus provided left side and come back to the search page , the page is not getting refreshed. I am getting a search page that contains old search results. At this point of time if I am trying to make any changes am getting some error called "Another user with this id already modifed data ....". After this error my app is not running. Means what ever am trying to do its showing the same error.
So I need to make this: "When ever the user come to this form, He should get fresh form. It should not contain old search results.
Please help me. How do I achieve this.
Thank you.
There are 2 ways of doing it:
1) Set your task flow as ISOLATED, from Task Flow Overview tab -> Behaviour -> Share Data Control with calling task flow -> unchecked (or isolated, if you are using JDev 12c)
This will ensure you always start FRESH when accessing the page, but it will potentially create a performance overhead because entire View Object cache will be recreated (requeried) on page load. Nevertheless, it is the quickest solution.
2) You may create a default Method Call Activity in your task flow from where you may call a AM's custom method that resets the view criteria. The method will be placed on application module's implementation class and it may look like this:
public void initTaskFlow() {
this.getViewObject1().executeEmptyRowSet();
}
This will clean the result data. If you want to reset the querying parameters as well, you can use this example:
http://www.jobinesh.com/2011/04/programmatically-resetting-and-search.html
When you made any changes to any viewObject then excute this viewObject to match entity state and viewState , i think excuting viewObject will solve your issue
Ashish

Random Popups While Automating the Webpage

I was attending an interview and he gave me the following scenarios . If I could get an hint as I could not answer the questions.
Assume that there is an application and popups keep coming up all the time. These are not times, its just random. You never know when they are going to come. How to deal with it?
Assume that the script you wrote is fine. But due to network issues the objects in the page are really slow to load or the page itself is taking long time. How do you deal with such a scenario?
Assume that I have 5-6 pages in the application. In all the pages we have certain text fields. In page 1 and Page 5 there is an object which is a text box. I see that what ever whatever identification method (css, xpath, id etc) you take, the values are same. That is duplicates. How do you deal with this scenario?
What is the basic purpose of "data provider" annotation in TestNG. In genral, what is the purpose of testng annotations?
Thanks.
Assume that the script you wrote is fine. But due to network issues the objects in the page are really slow to load or the page itself is taking long time.
How do you deal with such a scenario. In such situation, You should Wait property of Selenium. Implicit Wait or Explicit wait.
Implicit Wait -- Used for setting Timeout for Webpage loading
Driverobject.manage().timeouts().PageLoadtimeOut(units,TimeUnit.SECONDS);
Explicit Wait-- Used for setting Timeout for particular
Webelement FirefoxDriver f = new FirefoxDriver();
WebDriverWait ww = new WebDriverWait(f,Units);
ww.until(ExpectedConditions.CONDITION);
For second question, Anubhav has answered it.
For third, even if elements are same for the page1 and page5, they can be differentiated. First, switch to page to, whose text field you want to interact with, and then interact with that text field.
For forth, dataprovider is an annotation in TestNG using which you can do data driven testing and using TestNG annotations, you can manage test execution flow of your tests. For more details of dataprovider and TestNG annotation, please go here
For third, If you open 5-6 the pages in different tabs of single browser you will get such a duplication problem. That time only one page is visible to the end user. So we can differentiate that element by visibility and can interact with that element using webdriver
List<WebElement> el=driver.findElements(By.xpath("xpath of that text element"));//you can use other than xpath too to identify the elements
for(int i=0;i<el.size();i++)
{
if(el.get(i).isDisplayed())
el.get(i).sendKeys("text you want to send");//any other action you want to perform
break;
}

how to resolve race conditions in webdriver

I have an app that uses a mixture of angular and asp.net. My issue is that the home page is redirected by setting window.location and then the required data and page is requested form the server.
Previously this was not the case and all routing was done via the angular app. However due to requirements the applications routing had to be changed to what it is now.
Now because the application requires a server side request/response (I believe) this is causing a race condition in my tests as I only receive the expected result once in every 5 tests.
At present I am not able to provide code to explain my situation. However, I will be albe to provide some code in a edit later today.
The only code I can provide for now would be the test that is being run. Although I do not believe this would help without the code running the application.
Recent frameworks have this feature where it sets the document.readState of browser to complete and the content is loaded afterwards. Due to this the test may fails as we will be expecting an element to be present.
For such conditions you have to use explicit wait for the element to be present for which you want to take an action after the page is loaded or changed.
Here is an example how we wait for elements in our project (The application is also angularjs, we use Java for webdriver):
In our Webdriver implementation we added:
private WebDriverWait iWait(int timeoutInSeconds) {
return new WebDriverWait(webDriver, timeoutInSeconds);
}
we want to wait for an element to be visible ("Visibility means that the element is not only displayed but also has a height and width that is greater than 0"):
public void waitForElementToAppear(By by, int timeoutInSeconds) {
iWait(timeoutInSeconds).ignoring(StaleElementReferenceException.class).until(ExpectedConditions.visibilityOfElementLocated(by));
}
the ExpectedConditions class provides many other out of the box condition, here are some:
elementToBeClickable
textToBePresentInElement
titleContains
elementSelectionStateToBe
for more, please look at the ExpectedConditions Javadoc
if you need to create your own conditions, you can use ExoectedCondition (no 's') class
ExpectedCondition Javadoc

<f:view> beforePhase method called twice when page loads

I am trying to capture url parameters when page is loaded. I am using the beforePhase attribute inside the tag to call the backing bean method which handles the PhaseEvent. I am able to read the parameters but there seems to be some problem. The backing bean method is executed twice. First time I am getting all the url paramters but second time I am getting null.
This is the output in the console I get (The line "BeforePhase ONLOAD" is my SOP statement inside the method. As you can see the SOP is printed twice (below is the console output) -
* <_checkTimestamp> Apache Trinidad is running with time-stamp checking enabled. This should not be used in a production environment. See the org.apache.myfaces.trinidad.CHECK_FILE_MODIFICATION property in WEB-INF/web.xml
beforePhase ONLOAD
<_isBeanValidationAvailable> A Bean Validation provider is not present, therefore bean validation is disabled
beforePhase ONLOAD***
I have to use these url parameters to bind it to the view object and retrieve the data for the page. But if the method is called twice like it has been called here then the second will cause some database error or ultimately i will not see any records on screen.
Is there a way to handle this ?
I actually followed the Approach 1 example seen on this page - http://jneelmani.blogspot.com/2013/01/adf-how-to-call-method-on-page-load.html. In this example, he shows how to avoid beforePhase being called multiple times, I have done that but still the method is called twice.
Any suggestions here will be helpful.
Thanks.
Of course you know ADF supports a mechanism already for bookmarkable pages & URL parameters?
http://myadfnotebook.blogspot.pt/2010/11/bookmarking-with-adf.html

Resources