Saturday, March 20, 2010

Ant script for generating a Junit report for Selenium Test execution

We know the advantages of using Selenium RC. But Selenium RC by default does not provide us with any reports for the test execution done through it. We need to explicitly implement a reporting framework for test reporting in RC. For implementing a Reporting framework you can write your own custom code or can use an already existing reporting framework available.

There are mainly two reporting framework available for Selenium RC they are Junit and Test-NG. There are two more frameworks Report-NG and testng-xslt but they are implemented over the Test-NG framework.

In this blog I will mention about generating a Junit Report for your Selenium RC test execution using Ant script. I had written a “build.xml” for Ant for executing Selenium cases and generating a Junit report for the said selenium execution. I will be discussing about the same in this blog. If someone doesn’t have an idea about Ant can follow my steps and can generate a Junit report for Selenium Execution.


Explaining the “build.xml”

  1. In Build.xml I am mainly defining 5 properties

o RELEASE_ROOT this has been set to the Java Project folder.

o SRC which defines the source folder of the project

o LIB which defines the lib folder of the project

o BIN which defines the binary folder where the classes needs to be pasted after compiling

o REPORT folder which defines the report folder where the test execution reports will be generated.

  1. There are 4 targets or tasks of ant inside the build.xml

o “init’ which deletes the binary folder and creates it again.

o “compile” which depends upon the “init” task and which compiles the java files under the source folder and paste the classes onto the binary folder.

o “run-single-test” which depends on “compile” and runs only a single Selenium Junit test mentioned in it. This task also generates a Junit html report under the report/html folder of Java project after Selenium Test execution is done.

o “run-batch-test” which depends on “compile” and runs multiple Selenium Junit test mentioned in it. This task also generates a Junit html report under the report/html folder of Java project after Selenium Test execution is done.

  1. In each of the compile, run-single-test and run-batch-test I am setting the classpath of the binary folder, the library folder and the source folder in case of “run-batch-test”.
  2. In targets “run-single-test” and “run-batch-test” I am using the “junit” task of test execution. This task provides two types of execution

o Single class execution which I had used in “run-single-test” target. There is a subtask “test” of “junit” that can used for single test class execution. It is written as follows:


<test name="test.TestClass" haltonfailure="no" todir="${REPORT}/xml" outfile="TEST-result">

<formatter type="xml" />

</test>


Here “name” is the test class name. “todir” is the directory where initial xml report needs to be generated. “outfile” is the xml file name for initial result xml file. “formatter” is the initial file format type.


o Batch test class execution which I had used in “run-batch-test” target. There is a subtask “batchtest” of “junit” that can used for single test class execution. It is written as follows:


<formatter type="xml" usefile="true" />

<batchtest fork="yes" todir="${REPORT}/xml">

<fileset dir="${SRC}">

<include name="**/Test*.java" />

</fileset>

</batchtest>


Here I am telling the “batchtest” what all test classes to include using the “fileset” and “include” task. Along with this if you want you can use the “exclude” task to exclude particular type of classes. In the following example the Junit will include all the test class starting from “Test” and exclude all the test class having “Base” word in their name.


<formatter type="xml" usefile="true" />

<batchtest fork="yes" todir="${REPORT}/xml">

<fileset dir="${SRC}">

<include name="**/Test*.java" />

<exclude name="**/*Base*.java" />

</fileset>

</batchtest>


5. For generating the test report I am using the “junitreport” task of ant. It takes all the xml file from the “report/xml” and generates an html report at “report/html”.


Setting up a Java Project for test execution using Ant.


In this I am just telling how to setup a basic java project for using the build.xml file attached with this blog and running the selenium test using Ant.


Preconditions:

1. Install Ant from apache.

2. Install Java JDK (not JRE) and add its bin folder to your “PATH” environment variable of your system. For checking JDK is mapped check whether "javac" is recognized in your system by typing "javac" in your command prompt.

3. Download the Selenium RC.

4. Test cases are already converted to Selenium Junit cases.


Steps:

  1. Create a Java project in your Java IDE. I am using Eclipse so I will refer to it.
  2. By default there will be the “src” folder inside your Java Project. Right click on the “src” folder and create a new package under it say”test”.
  3. Now copy your Selenium Junit classes to this new “test” package.
  4. Create a new folder called “lib” under your Java Project.
  5. Copy selenium-java-client-driver.jar, junit-4.5.jar(if you have another version of Junit 4 you can use that) and ant-junit.jar(can be found under the “lib” folder of your ant installation) to your “lib” folder of Java Project.
  6. Now create a folder called “build” under your Java Project and copy the “build.xml” file attached with this blog to the build folder you created.
  7. Now open the build.xml file for edit.
  8. Search for “run-single-test” in build.xml. Go to the “<test” under the “run-single-test” target .Replace the values for “name=” variable with your Selenium Junit class name. For ex if your class name is TestSelenium.java and it is under package “test” under “src” then replace the value with “test.TestSelenium” then save the build.xml file.
  9. Now start the Selenium RC server.
  10. Now go to the build folder of the Java Project using command prompt and run the command “ant run-single-test”
  11. It should execute your Selenium cases on selenium server and generate an html report onto the “report/html” folder of your Java project.


Similarly you can modify the “run-batch-test” target of build.xml and execute multiple Selenium Tests together.

Download build.xml from Here

Thursday, March 11, 2010

Defining your own Environment and Using it in Selenium Grid


If you want to use an environment something like “Firefox on Windows” or “Firefox on Linux”.
You can use that by starting the Remote control with the said environment name. This will register the Remote control to the hub with the said environment name.
For ex. Start the hub normally but while starting the Remote control replace the -Denvironment value with your own particular name.

ant -Dport=5555 -Dhost=192.168.1.0 -DhubURL=http://192.168.1.2:4444 -Denvironment=”Firefox on Linux” launch-remote-control


If you check the Hub console it will look something like


Also if you want to use such an environment for executing your Test you need to pass the environment name to the selenium object in your code.
For Java you need to use the following function while creating the selenium object.


new DefaultSelenium("localhost", 4444, “Firefox on Windows”, 'http://google.com');

Now if you go through the above things you will think about how to use the above function and define the selenium object since it is already been declared and defined when you get it from selenium IDE.
This happens because when you copy the code from the Selenium Ide, the class that is defined in the code extends the “SeleneseTestCase” class where this “selenium ” object has already been defined.


For using the custom Environment with the Selenium Grid do the following thing:

1.Remove the “SeleneseTestCase” class which is extended by the Test class.

2.Define your own Selenese object as shown below

Selenium selenium;
selenium =new DefaultSelenium("localhost", 4444, " ", "http://www.google.com");


3.Start the selenium using the below command.

selenium.start();

This will start your selenium session. Then using this selenium object you run your test cases.

4.After your test is done use the following command to stop the selenium.


selenium.stop();

When you call the selenium grid by giving some particular Environment name. The Selenium Hub will make sure to provide the test with a remote control that registered with this specific Hub environment.

Friday, March 5, 2010

Setting up Selenium-Grid on your System

In this blog I will tell you how to set-up Selenium Grid on your system.


  1. Check that Java is installed in your system by using the command “java -version”. If Java is installed it will give the output about Java Version that is installed in your system. Else download the Java from Java website (http://www.java.com/en/download/index.jsp).

  2. Check ANT is installed in your system by using the command “ant -version”.If Ant is installed it will give the output about Ant Version that is installed in your system. Else download the Apache Ant from Apache website (http://ant.apache.org/).

  3. Download Selenium Grid from Selenium Grid website(http://selenium-grid.seleniumhq.org/download.html) and unpack it.

  4. Go to the Unpacked directory of the Selenium grid using command prompt and type the command “ant launch-hub”. This will launch the Selenium Grid Hub on a default port “4444” to check whether the selenium grid is running. Type the URL “http://localhost:4444/console ” on your browser and check that you are able to see the selenium hub console on the browser.

  5. Now go to the same Selenium Grid directory using command prompt and type the command “ant launch-remote-control”. This will start the Selenium Grid Remote control on default port “5555” and environment “*firefox” and register it to the Hub which you had earlier started.

  6. Now after starting the selenium remote control we will test the example provided with Selenium Grid for sequence testing in single machine. For this go to the Selnium grid folder from command prompt and type the command “ant run-in-demo-sequence”. This will execute the test example provided with selenium grid to be executed in sequence.


Now you may had noticed in the setting up steps that I am running the test in sequence. You may think in your mind that, what's special in that? The same things can be achieved when we use a Selenium RC.Then here comes the important part of Selenium Grid.


Setting up Parallel execution on a single machine

  1. Start the hub and one remote control in the same way as in the earlier setting up steps.

  2. Now go to the Selenium Grid folder using command prompt and type the command

“ant -Dport =5556 launch-remote-control”

This will start another remote control on port “5556” and register it to the local hub. If you want to register more remote control to the Hub. You can do that by starting multiple remote control but with different ports.

  1. To know the no. of registered remote control onto a hub go to the URL “http://localhost:4444/console”. It will give you a view as follows:


  1. Now to run the test in parallel. Go to the selenium grid directory using command prompt and type the command “ant run-demo-in-parallel”. This will run the example tests given with selenium grid in parallel.


Setting up parallel execution for different machines and different environments

In this case start the Hub normally as you do in any of the system.

For starting the remote control use the following command.


ant -Dport=<port> -Dhost=<hostname> -DhubURL=<hub> -Denvironment=<environment> launch-remote-control


Here

“-Dport” should be used to set the port of the remote control.

“-Dhost” should be used to set the host(where the remote control has been started) IP. This is the IP which the hub will use to talk to the remote control.

“-DhubURL” should be used to set the Hub Url where the current remote control will register itself.

“-Denvironment” should be used to set an environment name that remote control represents to the hub.

Now start 4 different remote controls from 2 different machines and register them to the hub.

After this check the Hub console. It will look similar to the table as shown below depending upon the remote control that are registered to the Hub.


Then to verify your set-up go to the Selenium Grid folder from the command prompt and type the command “ant run-demo-in-parallel”


This will run an example test accompanied with the Selenium Grid to run in parallel.


Advantages of Selenium Grid

  1. Remote Execution of the selenium tests can be done using Selenium Grid without actually copying your test code to to remote system.

  2. Selenium Grid can be used to save time by executing independent testcases in parallel.

  3. Load Testing can be done if we register a lot of remote control to a single hub and execute testcase in parallel to simulate multiple transactions at a single timeperiod.

Monday, March 1, 2010

Useful Functions and Notes while writing cases in Selenium RC

Useful Functions of Selenium

You have a huge list of selenium functions that we can use for writing test cases. Some of the main functions are:

isElementPresent(location) – Returns true or false based on whether “location” of the element passed is present on the page or not. Verifies that the specified element is somewhere on the page. This function can be used while you are working on an Ajax based system or in a system where you need to identify whether the element is present before we execute any action on the said "location". It is good to check if the location is present or not if your test case include clicking on some link or page refresh and then taking some action.

getValue(location)- Returns the data stored in the said “location” of the element that is passed to the function. Gets the (whitespace-trimmed) value of an input field (or anything else with a value parameter). For checkbox/radio elements, the value will be "on" or "off" depending on whether the element is checked or not.This function is useful when you need to get some data from some field or a particular location and need to use it in you program or testcase.

isTextPresent(text)- Returns true or false based on the specified text pattern appears somewhere on the rendered page shown to the user.Used to verify the whether a text present on the page or not.

click(location) - Clicks on a link, button, checkbox or radio button. If the click action causes a new page to load (like a link usually does), call "waitForPageToLoad" function.

waitForPageToLoad(timeout) - Waits for a new page to load. Selenium constantly keeps track of new pages loading, and sets a "newPageLoaded" flag when it first notices a page load. Running any other Selenium command after turns the flag to false. Hence, if you want to wait for a page to load, you must wait immediately after a Selenium command that caused a page-load.

type(locator,value) - Sets the value of an input field, as though you typed it in.

select(locator,option locator) - Select an option from a drop-down using an option locator. Using to select values from combo boxes.


Useful things while writing cases in Selenium

  1. While testing any Ajax based application use “waitForElementPresent” or “isElementPresent” before trying to do any action over the element.

  2. While writing your own case always remember to extend the “SeleneseTestCase” class.This create a “selenium” object and stores the browser session onto the created object.

  3. To set the browser on which you need to do execution of your test case add the following function to your test case class

public void setUp() throws Exception {

setUp(URL,browser name);

}

List of browsers that can used is “*firefox” -for firefox, “*iexplore” -for IE, “*safari” - for Safari, “*opera” - For Opera.This need to be placed in place of "browser name" while calling the "setUp" function.

  1. If you are internally going to call another function from your test case and going to execute some actions under it. Don't forget to pass the “selenium” object to the function otherwise your test case will fail. This happens because when you are executing a case ,the browser session details is stored with the “selenium” object and if you call a function without passing the selenium object to the function it wont be having any session details with it and hence the case will Fail.