Selenium Documentation

Release 1.0

Selenium Project

November 05, 2010

CONTENTS

1 2

Note to the Reader Introducing Selenium 2.1 To Automate or Not to Automate? That is the Question! 2.2 Test Automation for Web Applications . . . . . . . . . 2.3 Introducing Selenium . . . . . . . . . . . . . . . . . . 2.4 Selenium Components . . . . . . . . . . . . . . . . . . 2.5 Supported Browsers . . . . . . . . . . . . . . . . . . . 2.6 Flexibility and Extensibility . . . . . . . . . . . . . . . 2.7 About this Book . . . . . . . . . . . . . . . . . . . . . 2.8 The Documentation Team . . . . . . . . . . . . . . . . Selenium Basics 3.1 Getting Started Choosing Your Selenium Tool 3.2 Introducing Selenium Commands . . . . . . . . 3.3 Test Suites . . . . . . . . . . . . . . . . . . . . 3.4 Commonly Used Selenium Commands . . . . . 3.5 Summary . . . . . . . . . . . . . . . . . . . . .

3 5 5 5 6 6 7 8 8 9 11 11 11 13 14 14 15 15 15 18 18 21 24 24 25 27 28 29 29 29

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

Selenium-IDE 4.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . 4.2 Installing the IDE . . . . . . . . . . . . . . . . . . . . . 4.3 Opening the IDE . . . . . . . . . . . . . . . . . . . . . . 4.4 IDE Features . . . . . . . . . . . . . . . . . . . . . . . . 4.5 Building Test Cases . . . . . . . . . . . . . . . . . . . . 4.6 Running Test Cases . . . . . . . . . . . . . . . . . . . . 4.7 Using Base URL to Run Test Cases in Different Domains 4.8 Debugging . . . . . . . . . . . . . . . . . . . . . . . . . 4.9 Writing a Test Suite . . . . . . . . . . . . . . . . . . . . 4.10 User Extensions . . . . . . . . . . . . . . . . . . . . . . 4.11 Format . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.12 Executing Selenium-IDE Tests on Different Browsers . . 4.13 Troubleshooting . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

Selenium Commands 35 5.1 Verifying Page Elements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35 5.2 Locating Elements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37 5.3 Matching Text Patterns . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42

5.4 5.5 5.6 5.7 5.8 5.9 5.10 6

The AndWait Commands . . . . . . . . . . The waitFor Commands in AJAX applications Sequence of Evaluation and Flow Control . . . Store Commands and Selenium Variables . . . JavaScript and Selenese Parameters . . . . . . echo - The Selenese Print Command . . . . . . Alerts, Popups, and Multiple Windows . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

44 44 44 45 46 47 47 51 51 51 53 56 59 65 67 68 71 75 75 79 80 80 87 87 87 89 90 93 93 95 96 99 101 101 101 101 102 103 103 107 107 108 108 110 114 118 119

Selenium-RC 6.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . 6.2 How Selenium-RC Works . . . . . . . . . . . . . . . . . . . 6.3 Installation . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.4 From Selenese to a Program . . . . . . . . . . . . . . . . . . 6.5 Programming Your Test . . . . . . . . . . . . . . . . . . . . 6.6 Learning the API . . . . . . . . . . . . . . . . . . . . . . . . 6.7 Reporting Results . . . . . . . . . . . . . . . . . . . . . . . 6.8 Adding Some Spice to Your Tests . . . . . . . . . . . . . . . 6.9 Server Options . . . . . . . . . . . . . . . . . . . . . . . . . 6.10 Specifying the Path to a Specic Browser . . . . . . . . . . . 6.11 Selenium-RC Architecture . . . . . . . . . . . . . . . . . . . 6.12 Handling HTTPS and Security Popups . . . . . . . . . . . . 6.13 Supporting Additional Browsers and Browser Congurations 6.14 Troubleshooting Common Problems . . . . . . . . . . . . . . Test Design ConsIDerations 7.1 Introducing Test Design . . . 7.2 What to Test? . . . . . . . . . 7.3 Verifying Results . . . . . . . 7.4 Choosing a Location Strategy 7.5 Testing Ajax Applications . . 7.6 UI Mapping . . . . . . . . . 7.7 Bitmap Comparison . . . . . 7.8 Handling Errors . . . . . . . Selenium-Grid User-Extensions 9.1 Introduction . . . . . . . . . . . . . . . . 9.2 Actions . . . . . . . . . . . . . . . . . . . 9.3 Accessors/Assertions . . . . . . . . . . . . 9.4 Locator Strategies . . . . . . . . . . . . . 9.5 Using User-Extensions With Selenium-IDE 9.6 Using User-Extensions With Selenium RC

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

8 9

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

10 Selenium 2.0 and WebDriver 10.1 What is WebDriver? . . . . . . . . 10.2 When to Use WebDriver? . . . . . 10.3 The 5 Minute Getting Started Guide 10.4 Next Steps For Using WebDriver . 10.5 WebDriver Implementations . . . . 10.6 Emulating Selenium RC . . . . . . 10.7 Tips and Tricks . . . . . . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

ii

10.8 10.9 10.10 10.11

How XPATH Works in WebDriver Getting and Using WebDriver . . Roadmap . . . . . . . . . . . . . Further Resources . . . . . . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

120 121 122 122 123

11 .NET client driver conguration

12 Java Client Driver Conguration 127 12.1 Conguring Selenium-RC With Eclipse . . . . . . . . . . . . . . . . . . . . . . . . . . 127 12.2 Conguring Selenium-RC With Intellij . . . . . . . . . . . . . . . . . . . . . . . . . . 143 13 Python Client Driver Conguration 155

14 Locating Techniques 159 14.1 Useful XPATH patterns . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 159 14.2 Starting to use CSS instead of XPATH . . . . . . . . . . . . . . . . . . . . . . . . . . . 159

iii

iv

Selenium Documentation, Release 1.0

Contents:

CONTENTS

Selenium Documentation, Release 1.0

CONTENTS

CHAPTER

ONE

NOTE TO THE READER

Hello, and welcome to Selenium! The Documentation Team would like to welcome you, and to thank you for being interested in Selenium. We have worked very, very hard on this document. Why? We absolutely believe this is the best tool for web-application testing. We feel its extensibility and exibility, along with its tight integration with the browser, is unmatched by available proprietary tools. We are very excited to promote Selenium and, hopefully, to expand its user community. In short, we really want to get the word out about Selenium. We truly believe you will be similarly excited once you learn how Selenium approaches test automation. Its quite different from other tools. Whether you are brand-new to Selenium, or have been using it for awhile, we believe this documentation will truly help to spread the knowledge around. Also, we have aimed to write so that those completely new to test automation will be able to use this document as a stepping stone. No doubt, experienced users and newbies will benet from our Selenium Users Guide. Please realize that this document is a work in progress. There are planned areas we havent written yet. However, we have written the beginning chapters rst so newcomers can get started more smoothly. We have also already added some valuable information that more experienced users will appreciate. This document will be a live document on the SeleniumHQ website where frequent updates will occur as we complete the additional planned documentation. We really want to hear about any problems you have with the documentation, or suggestions you have for improving it. For errors, please le a bug report. Improvement suggestions should be led as an enhancement request. Thanks very much for reading. the Selenium Documentation Team

Selenium Documentation, Release 1.0

Chapter 1. Note to the Reader

CHAPTER

TWO

INTRODUCING SELENIUM
2.1 To Automate or Not to Automate? That is the Question!
Is automation always advantageous? When should one decide to automate test cases? It is not always advantageous to automate test cases. There are times when manual testing may be more appropriate. For instance, if the applications user interface will change considerably in the near future, then any automation would need to be rewritten. Also, sometimes there simply is not enough time to build test automation. For the short term, manual testing may be more effective. If an application has a very tight deadline, there is currently no test automation available, and its imperative that the testing get done within that time frame, then manual testing is the best solution. However, automation has specic advantages for improving the long-term efciency of a software teams testing processes. Test automation supports: Frequent regression testing Rapid feedback to developers during the development process Virtually unlimited iterations of test case execution Customized reporting of application defects Support for Agile and eXtreme development methodologies Disciplined documentation of test cases Finding defects missed by manual testing

2.2 Test Automation for Web Applications

Many, perhaps most, software applications today are written as web-based applications to be run in an Internet browser. The effectiveness of testing these applications varies widely among companies and organizations. In an era of continuously improving software processes, such as eXtreme programming (XP) and Agile, it can be argued that disciplined testing and quality assurance practices are still underdeveloped in many organizations. Software testing is often conducted manually. At times, this is effective; however there are alternatives to manual testing that many organizations are unaware of, or lack the skills to perform. Utilizing these alternatives would in most cases greatly improve the efciency of their software development by adding efciencies to their testing. Test automation is often the answer. Test automation means using a tool to run repeatable tests against the target application whenever necessary. 5

Selenium Documentation, Release 1.0

There are many advantages to test automation. Most are related to the repeatability of the tests and the speed at which the tests can be executed. There are a number of commercial and open source tools available for assisting with the development of test automation. Selenium is possibly the most widelyused open source solution. This users guide will assist both new and experienced Selenium users in learning effective techniques in building test automation for web applications. This guide introduces Selenium, teaches its most widely used features, and provides useful advice in best practices accumulated from the Selenium community. Many examples are provided. Also, technical information on the internal structure of Selenium and recommended uses of Selenium are provided as contributed by a consortium of experienced Selenium users. It is our hope that this guide will get additional new users excited about using Selenium for test automation. We hope this guide will assist in getting the word out that quality assurance and software testing have many options beyond what is currently practiced. We hope this users guide and Selenium itself provide a valuable aid to boosting the readers efciency in his or her software testing processes.

2.3 Introducing Selenium

Selenium is a robust set of tools that supports rapid development of test automation for web-based applications. Selenium provides a rich set of testing functions specically geared to the needs of testing of a web application. These operations are highly exible, allowing many options for locating UI elements and comparing expected test results against actual application behavior. One of Seleniums key features is the support for executing ones tests on multiple browser platforms.

2.4 Selenium Components

Selenium is composed of three major tools. Each one has a specic role in aiding the development of web application test automation.

2.4.1 Selenium-IDE
Selenium-IDE is the Integrated Development Environment for building Selenium test cases. It operates as a Firefox add-on and provides an easy-to-use interface for developing and running individual test cases or entire test suites. Selenium-IDE has a recording feature, which will keep account of user actions as they are performed and store them as a reusable script to play back. It also has a context menu (right-click) integrated with the Firefox browser, which allows the user to pick from a list of assertions and verications for the selected location. Selenium-IDE also offers full editing of test cases for more precision and control. Although Selenium-IDE is a Firefox only add-on, tests created in it can also be run against other browsers by using Selenium-RC and specifying the name of the test suite on the command line.

2.4.2 Selenium-RC (Remote Control)

Selenium-RC allows the test automation developer to use a programming language for maximum exibility and extensibility in developing test logic. For instance, if the application under test returns a result set, and if the automated test program needs to run tests on each element in the result set, the programming languages iteration support can be used to iterate through the result set, calling Selenium commands to run tests on each item.

Chapter 2. Introducing Selenium

Selenium Documentation, Release 1.0

Selenium-RC provides an API (Application Programming Interface) and library for each of its supported languages: HTML, Java, C#, Perl, PHP, Python, and Ruby. This ability to use Selenium-RC with a highlevel programming language to develop test cases also allows the automated testing to be integrated with a projects automated build environment.

2.4.3 Selenium-Grid
Selenium-Grid allows the Selenium-RC solution to scale for large test suites or test suites that must be run in multiple environments. With Selenium-Grid, multiple instances of Selenium-RC are running on various operating system and browser congurations; Each of these when launching register with a hub. When tests are sent to the hub they are then redirected to an available Selenium-RC, which will launch the browser and run the test. This allows for running tests in parallel, with the entire test suite theoretically taking only as long to run as the longest individual test.

2.5 Supported Browsers

Browser Firefox 3.x Firefox 3 Firefox 2 IE 8 IE 7 IE 6 Safari 4 Safari 3 Safari 2 Opera 10 Opera 9 Opera 8 Google Chrome Others Selenium-IDE Record and playback tests Record and playback tests Record and playback tests Test execution only via Selenium-RC* Test execution only via Selenium-RC* Test execution only via Selenium-RC* Test execution only via Selenium-RC Test execution only via Selenium-RC Test execution only via Selenium-RC Test execution only via Selenium-RC Test execution only via Selenium-RC Test execution only via Selenium-RC Test execution only via Selenium-RC (Windows) Test execution only via Selenium-RC Selenium-RC Start browser, run tests Start browser, run tests Start browser, run tests Start browser, run tests Start browser, run tests Start browser, run tests Start browser, run tests Start browser, run tests Start browser, run tests Start browser, run tests Start browser, run tests Start browser, run tests Start browser, run tests Partial support possible** Operating Systems Windows, Linux, Mac Windows, Linux, Mac Windows, Linux, Mac Windows Windows Windows Mac Mac Mac Windows, Linux, Mac Windows, Linux, Mac Windows, Linux, Mac Windows As applicable

* Tests developed on Firefox via Selenium-IDE can be executed on any other supported browser via a simple Selenium-RC command line. ** Selenium-RC server can start any executable, but depending on browser security settings there may 2.5. Supported Browsers 7

Selenium Documentation, Release 1.0

be technical limitations that would limit certain features.

2.6 Flexibility and Extensibility

Youll nd that Selenium is highly exible. There are multiple ways in which one can add functionality to Seleniums framework to customize test automation for ones specic testing needs. This is, perhaps, Seleniums strongest characteristic when compared with proprietary test automation tools and other open source solutions. Selenium-RC support for multiple programming and scripting languages allows the test writer to build any logic they need into their automated testing and to use a preferred programming or scripting language of ones choice. Selenium-IDE allows for the addition of user-dened user-extensions for creating additional commands customized to the users needs. Also, it is possible to re-congure how the Selenium-IDE generates its Selenium-RC code. This allows users to customize the generated code to t in with their own test frameworks. Finally, Selenium is an Open Source project where code can be modied and enhancements can be submitted for contribution.

2.7 About this Book

This reference documentation targets both new users of Selenium and those who have been using Selenium and are seeking additional knowledge. It introduces the novice to Selenium test automation. We do not assume the reader has experience in testing beyond the basics. The experienced Selenium user will also nd this reference valuable. It compiles in one place a set of useful Selenium techniques and best practices by drawing from the knowledge of multiple experienced Selenium QA professionals. The remaining chapters of the reference present: Selenium Basics Introduces Selenium by describing how to select the Selenium component most appropriate for your testing tasks. Also provides a general description of Selenium commands and syntax. This section allows you to get a general feel for how Selenium approaches test automation and helps you decide where to begin. Selenium-IDE Teaches how to build test cases using the Selenium Integrated Development Environment. This chapter also describes useful techniques for making your scripts more readable when interpreting defects caught by your Selenium tests. We explain how your test script can be exported to the programming language of your choice. Finally, this section describes some congurations available for extending and customizing how the Selenium-IDE supports test case development. Selenium Commands Describes a subset of the most useful Selenium commands in detail. This chapter shows what types of actions, verications and assertions can be made against a web application. Selenium-RC Explains how to develop an automated test program using the Selenium-RC API. Many examples are presented in both a programming language and a scripting language. Also, the installation and setup of Selenium-RC is covered here. The various modes, or congurations, that Selenium-RC supports are described, along with their trade-offs and limitations. Architecture diagrams are provided to help illustrate these points. A number of solutions to problems which are often difcult for the new user are described in this chapter. This includes handling Security Certicates, https requests, pop-ups, and the opening of new windows.

Chapter 2. Introducing Selenium

Selenium Documentation, Release 1.0

Test Design Considerations Presents many useful techniques for using Selenium efciently. This includes scripting techniques and programming techniques for use with Selenium-RC. We cover examples of source code showing how to report defects in the application under test. We also cover techniques commonly asked about in the user group such as how to implement data-driven tests (tests where one can vary the data between different test passes). Selenium-Grid This chapter is not yet developed. User extensions Presents all the information required for easily extending Selenium.

2.8 The Documentation Team

2.8.1 The Original Authors
Dave Hunt Paul Grandjean Santiago Suarez Ordonez Tarun Kumar The original authors who kickstarted this document are listed in alphabetical order. Each of us contributed signicantly by taking a leadership role in specic areas. Each chapter originally had a primary author who kicked off the intial writing, but in the end, each of us made signicant contributions to each chapter throughout the project.

2.8.2 Current Authors

Mary Ann May-Pumphrey Peter Newhook In addition to the original team members who are still involved (May 09), Mary Ann, and Peter have recently made major contributions. Their reviewing and editorial contributions proved invaluable. Mary Ann is actively writing new subsections and has provided editorial assistance throughout the document. Peter has provided assistance with restructuring our most difcult chapter and has provided valuable advice on topics to include. Their enthusiasm and dedication has been incredibly helpful. We hope they continue to be involved.

2.8.3 Acknowledgements
A huge special thanks goes to Patrick Lightbody. As an administrator of the SeleniumHQ website, his support has been invaluable. Patrick has helped us understand the Selenium communityour audience. He also set us up with everything we needed on the SeleniumHQ website for developing and releasing this users guide. His enthusiasm and encouragement denitely helped drive this project. Also thanks goes to Andras Hatvani for his advice on publishing solutions, and to Amit Kumar for participating in our discussions and for assisting with reviewing the document. And of course, we must recognize the Selenium Developers. They have truly designed an amazing tool. Without the vision of the original designers, and the continued efforts of the current developers, we would not have such a great tool to pass on to you, the reader. 2.8. The Documentation Team 9

Selenium Documentation, Release 1.0

10

Chapter 2. Introducing Selenium

CHAPTER

THREE

SELENIUM BASICS
3.1 Getting Started Choosing Your Selenium Tool
Most people get started with Selenium-IDE. This is what we recommend. Its an easy way to get familiar with Selenium commands quickly. You can develop your rst script in just a few minutes. Selenium-IDE is also very easy to install. See the chapter on Selenium-IDE for specics. You may also run your scripts from the Selenium-IDE. Its simple to use and is recommended for lesstechnical users. The IDE allows developing and running tests without the need for programming skills as required by Selenium-RC. The Selenium-IDE can serve as an excellent way to train junior-level employees in test automation. Anyone who understands how to conduct manual testing of a website can easily transition to using the Selenium-IDE for running and developing tests. Some testing tasks are too complex though for the Selenium-IDE. When programming logic is required Selenium-RC must be used. For example, any tests requiring iteration, such as testing each element of a variable length list requires running the script from a programming language. Selenium-IDE does not support iteration or condition statements. Finally, Selenium-Core is another way of running tests. One can run test scripts from a web-browser using the HTML interface TestRunner.html. This is the original method for running Selenium commands. It has limitations though; similar to Selenium-IDE, it does not support iteration. Selenium-Core also cannot switch between http and https protocols. Since the development of SeleniumIDE and Selenium-RC, more people are using these tools rather than Selenium-Core. At the time of writing (April 09) it is still available and may be convenient for some. However, the Selenium community is encouraging the use of Selenium-IDE and RC and discouraging the use of Selenium-Core. Support for Selenium-Core is becoming less available and it may even be deprecated in a future release.

3.2 Introducing Selenium Commands

3.2.1 Selenium Commands Selenese
Selenium provides a rich set of commands for fully testing your web-app in virtually any way you can imagine. The command set is often called selenese. These commands essentially create a testing language. In selenese, one can test the existence of UI elements based on their HTML tags, test for specic content, test for broken links, input elds, selection list options, submitting forms, and table data among other things. In addition Selenium commands support testing of window size, mouse position, alerts, Ajax functionality, pop up windows, event handling, and many other web-application features. The Command Reference lists all the available commands. 11

Selenium Documentation, Release 1.0

A command is what tells Selenium what to do. Selenium commands come in three avors: Actions, Accessors and Assertions. Actions are commands that generally manipulate the state of the application. They do things like click this link and select that option. If an Action fails, or has an error, the execution of the current test is stopped. Many Actions can be called with the AndWait sufx, e.g. clickAndWait. This sufx tells Selenium that the action will cause the browser to make a call to the server, and that Selenium should wait for a new page to load. Accessors examine the state of the application and store the results in variables, e.g. storeTitle. They are also used to automatically generate Assertions. Assertions are like Accessors, but they verify that the state of the application conforms to what is expected. Examples include make sure the page title is X and verify that this checkbox is checked. All Selenium Assertions can be used in 3 modes: assert, verify, and waitFor. For example, you can assertText, verifyText and waitForText. When an assert fails, the test is aborted. When a verify fails, the test will continue execution, logging the failure. This allows a single assert to ensure that the application is on the correct page, followed by a bunch of verify assertions to test form eld values, labels, etc. waitFor commands wait for some condition to become true (which can be useful for testing Ajax applications). They will succeed immediately if the condition is already true. However, they will fail and halt the test if the condition does not become true within the current timeout setting (see the setTimeout action below).

3.2.2 Script Syntax

Selenium commands are simple, they consist of the command and two parameters. For example: verifyText //div//a[2] Login

The parameters are not always required; it depends on the command. In some cases both are required, in others one parameter is required, and in still others the command may take no parameters at all. Here are a couple more examples: goBackAndWait verifyTextPresent type type Welcome to My Home Page (555) 666-7066 ${myVariableAddress}

id=phone id=address1

The command reference describes the parameter requirements for each command. Parameters vary, however they are typically: a locator for identifying a UI element within a page. a text pattern for verifying or asserting expected page content a text pattern or a selenium variable for entering text in an input eld or for selecting an option from an option list. Locators, text patterns, selenium variables, and the commands themselves are described in considerable detail in the section on Selenium Commands. 12 Chapter 3. Selenium Basics

Selenium Documentation, Release 1.0

Selenium scripts that will be run from Selenium-IDE will be be stored in an HTML text le format. This consists of an HTML table with three columns. The rst column identies the Selenium command, the second is a target, and the nal column contains a value. The second and third columns may not require values depending on the chosen Selenium command, but they should be present. Each table row represents a new Selenium command. Here is an example of a test that opens a page, asserts the page title and then veries some content on the page:
<table> <tr><td>open</td><td></td><td>/download/</td></tr> <tr><td>assertTitle</td><td></td><td>Downloads</td></tr> <tr><td>verifyText</td><td>//h2</td><td>Downloads</td></tr> </table>

Rendered as a table in a browser this would look like the following: open assertTitle verifyText /download/ Downloads Downloads

//h2

The Selenese HTML syntax can be used to write and run tests without requiring knowledge of a programming language. With a basic knowledge of selenese and Selenium-IDE you can quickly produce and run testcases.

3.3 Test Suites

A test suite is a collection of tests. Often one will run all the tests in a test suite as one continuous batch-job. When using Selenium-IDE, test suites also can be dened using a simple HTML le. The syntax again is simple. An HTML table denes a list of tests where each row denes the lesystem path to each test. An example tells it all.
<html> <head> <title>Test Suite Function Tests - Priority 1</title> </head> <body> <table> <tr><td><b>Suite Of Tests</b></td></tr> <tr><td><a href= "./Login.html" >Login</a></td></tr> <tr><td><a href= "./SearchValues.html" >Test Searching for Values</a></td></tr> <tr><td><a href= "./SaveValues.html" >Test Save</a></td></tr> </table> </body> </html>

A le similar to this would allow running the tests all at once, one after another, from the Selenium-IDE. Test suites can also be maintained when using Selenium-RC. This is done via programming and can be done a number of ways. Commonly Junit is used to maintain a test suite if one is using Selenium-RC with Java. Additionally, if C# is the chosen language, Nunit could be employed. If using an interpreted language like Python with Selenium-RC than some simple programming would be involved in setting up a test suite. Since the whole reason for using Sel-RC is to make use of programming logic for your testing this usually isnt a problem.

3.3. Test Suites

13

Selenium Documentation, Release 1.0

3.4 Commonly Used Selenium Commands

To conclude our introduction of Selenium, well show you a few typical Selenium commands. These are probably the most commonly used commands for building tests. open opens a page using a URL. click/clickAndWait performs a click operation, and optionally waits for a new page to load. verifyTitle/assertTitle veries an expected page title. verifyTextPresent veries expected text is somewhere on the page. verifyElementPresent veries an expected UI element, as dened by its HTML tag, is present on the page. verifyText veries expected text and its corresponding HTML tag are present on the page. verifyTable veries a tables expected contents. waitForPageToLoad pauses execution until an expected new page loads. Called automatically when clickAndWait is used. waitForElementPresent pauses execution until an expected UI element, as dened by its HTML tag, is present on the page.

3.5 Summary
Now that youve seen an introduction to Selenium, youre ready to start writing your rst scripts. We recommend beginning with the Selenium IDE and its context-sensitive, right-click, menu. This will allow you to get familiar with the most common Selenium commands quickly, and you can have a simple script done in just a minute or two. Chapter 3 gets you started and then guides you through all the features of the Selenium-IDE.

14

Chapter 3. Selenium Basics

CHAPTER

FOUR

SELENIUM-IDE
4.1 Introduction
The Selenium-IDE (Integrated Development Environment) is the tool you use to develop your Selenium test cases. Its an easy-to-use Firefox plug-in and is generally the most efcient way to develop test cases. It also contains a context menu that allows you to rst select a UI element from the browsers currently displayed page and then select from a list of Selenium commands with parameters pre-dened according to the context of the selected UI element. This is not only a time-saver, but also an excellent way of learning Selenium script syntax. This chapter is all about the Selenium IDE and how to use it effectively.

4.2 Installing the IDE

Using Firefox, rst, download the IDE from the SeleniumHQ downloads page When downloading from Firefox, youll be presented with the following window.

15

Selenium Documentation, Release 1.0

Select Install Now. The Firefox Add-ons window pops up, rst showing a progress bar, and when the download is complete, displays the following.

16

Chapter 4. Selenium-IDE

Selenium Documentation, Release 1.0

Restart Firefox. After Firefox reboots you will nd the Selenium-IDE listed under the Firefox Tools menu.

4.2. Installing the IDE

17

Selenium Documentation, Release 1.0

4.3 Opening the IDE

To run the Selenium-IDE, simply select it from the Firefox Tools menu. It opens as follows with an empty script-editing window and a menu for loading, or creating new test cases.

4.4 IDE Features

4.4.1 Menu Bar
The File menu allows you to create, open, and save test case and test suite les. The Edit menu allows copy, paste, delete, undo, and select all operations for editing the commands in your test case. The 18 Chapter 4. Selenium-IDE

Selenium Documentation, Release 1.0

Options menu allows the changing of settings. You can set the timeout value for certain commands, add user-dened user extensions to the base set of Selenium commands, and specify the format (language) used when saving your test cases. The Help menu is the standard Firefox Help menu; only one item on this menuUI-Element Documentationpertains to Selenium-IDE.

4.4.2 Toolbar
The toolbar contains buttons for controlling the execution of your test cases, including a step feature for debugging your test cases. The right-most button, the one with the red-dot, is the record button.

Speed Control: controls how fast your test case runs. Run All: Runs the entire test suite when a test suite with multiple test cases is loaded. Run: Runs the currently selected test. When only a single test is loaded this button and the Run All button have the same effect. Pause/Resume: Allows stopping and re-starting of a running test case. Step: Allows you to step through a test case by running it one command at a time. Use for debugging test cases. TestRunner Mode: Allows you to run the test case in a browser loaded with the Selenium-Core TestRunner. The TestRunner is not commonly used now and is likely to be deprecated. This button is for evaluating test cases for backwards compatibility with the TestRunner. Most users will probably not need this button. Apply Rollup Rules: This advanced feature allows repetitive sequences of Selenium commands to be grouped into a single action. Detailed documentation on rollup rules can be found in the UI-Element Documentation on the Help menu. Record: Records the users browser actions.

4.4.3 Test Case Pane

Your script is displayed in the test case pane. It has two tabs, one for displaying the command and their parameters in a readable table format.

4.4. IDE Features

19

Selenium Documentation, Release 1.0

The other tab - Source displays the test case in the native format in which the le will be stored. By default, this is HTML although it can be changed to a programming language such as Java or C#, or a scripting language like Python. See the Options menu for details. The Source view also allows one to edit the test case in its raw form, including copy, cut and paste operations. The Command, Target, and Value entry elds display the currently selected command along with its parameters. These are entry elds where you can modify the currently selected command. The rst parameter specied for a command in the Reference tab of the bottom pane always goes in the Target eld. If a second parameter is specied by the Reference tab, it always goes in the Value eld.

If you start typing in the Command eld, a drop-down list will be populated based on the rst characters you type; you can then select your desired command from the drop-down.

4.4.4 Log/Reference/UI-Element/Rollup Pane

The bottom pane is used for four different functionsLog, Reference, UI-Element, and Rollup depending on which tab is selected. Log When you run your test case, error messages and information messages showing the progress are displayed in this pane automatically, even if you do not rst select the Log tab. These messages are often useful for test case debugging. Notice the Clear button for clearing the Log. Also notice the Info button is a drop-down allowing selection of different levels of information to log.

20

Chapter 4. Selenium-IDE

Selenium Documentation, Release 1.0

Reference The Reference tab is the default selection whenever you are entering or modifying Selenese commands and parameters in Table mode. In Table mode, the Reference pane will display documentation on the current command. When entering or modifying commands, whether from Table or Source mode, it is critically important to ensure that the parameters specied in the Target and Value elds match those specied in the parameter list in the Reference pane. The number of parameters provided must match the number specied, the order of parameters provided must match the order specied, and the type of parameters provided must match the type specied. If there is a mismatch in any of these three areas, the command will not run correctly.

While the Reference tab is invaluable as a quick reference, it is still often necessary to consult the Selenium Reference document. UI-Element and Rollup Detailed information on these two panes (which cover advanced features) can be found in the UIElement Documentation on the Help menu of Selenium-IDE.

4.5 Building Test Cases

There are three primary methods for developing test cases. Frequently, a test developer will require all three techniques.

4.5.1 Recording
Many rst-time users begin by recording a test case from their interactions with a website. When Selenium-IDE is rst opened, the record button is ON by default. Note: This can be set to OFF as a default with an available user extension. 4.5. Building Test Cases 21

Selenium Documentation, Release 1.0

During recording, Selenium-IDE will automatically insert commands into your test case based on your actions. Typically, this will include: clicking a link - click or clickAndWait commands entering values - type command selecting options from a drop-down listbox - select command clicking checkboxes or radio buttons - click command Here are some gotchas to be aware of: The type command may require clicking on some other area of the web page for it to record. Following a link usually records a click command. You will often need to change this to clickAndWait to ensure your test case pauses until the new page is completely loaded. Otherwise, your test case will continue running commands before the page has loaded all its UI elements. This will cause unexpected test case failures.

4.5.2 Adding Verications and Asserts With the Context Menu

Your test cases will also need to check the properties of a web-page. This requires assert and verify commands. We wont describe the specics of these commands here; that is in the chapter on Selenese Selenium Commands. Here well simply describe how to add them to your test case. With Selenium-IDE recording, go to the browser displaying your test application and right click anywhere on the page. You will see a context menu showing verify and/or assert commands. The rst time you use Selenium, there may only be one Selenium command listed. As you use the IDE however, you will nd additional commands will quickly be added to this menu. Selenium-IDE will attempt to predict what command, along with the parameters, you will need for a selected UI element on the current web-page. Lets see how this works. Open a web-page of your choosing and select a block of text on the page. A paragraph or a heading will work ne. Now, right-click the selected text. The context menu should give you a verifyTextPresent command and the suggested parameter should be the text itself. Also, notice the Show All Available Commands menu option. This shows many, many more commands, again, along with suggested parameters, for testing your currently selected UI element. Try a few more UI elements. Try right-clicking an image, or a user control like a button or a checkbox. You may need to use Show All Available Commands to see options other than verifyTextPresent. Once you select these other options, the more commonly used ones will show up on the primary context menu. For example, selecting verifyElementPresent for an image should later cause that command to be available on the primary context menu the next time you select an image and right-click. Again, these commands will be explained in detail in the chapter on Selenium commands. For now though, feel free to use the IDE to record and select commands into a test case and then run it. You can learn a lot about the Selenium commands simply by experimenting though the IDE.

22

Chapter 4. Selenium-IDE

Selenium Documentation, Release 1.0

4.5.3 Editing
Insert Command

Table View
Select the point in your test case where you want to insert the command. To do this, in the Test Case Pane, left-click on the line where you want to insert a new command. Right-click and select Insert Command; the IDE will add a blank line just ahead of the line you selected. Now use the command editing text elds to enter your new command and its parameters.

Source View
Select the point in your test case where you want to insert the command. To do this, in the Test Case Pane, left-click between the commands where you want to insert a new command, and enter the HTML tags needed to create a 3-column row containing the Command, rst parameter (if one is required by the Command), and second parameter (again, if one is required). Be sure to save your test before switching back to Table view. Insert Comment Comments may be added to make your test case more readable. These comments are ignored when the test case is run. Comments may also be used to add vertical white space (one or more blank lines) in your tests; just create empty comments. An empty command will cause an error during execution; an empty comment wont.

Table View
Select the line in your test case where you want to insert the comment. Right-click and select Insert Comment. Now use the Command eld to enter the comment. Your comment will appear in purple font.

Source View
Select the point in your test case where you want to insert the comment. Add an HTML-style comment, i.e., <! your comment here >. Edit a Command or Comment

Table View
Simply select the line to be changed and edit it using the Command, Target, and Value elds.

Source View
Since Source view provides the equivalent of a WYSIWYG editor, simply modify which line you wish command, parameter, or comment. 4.5. Building Test Cases 23

Selenium Documentation, Release 1.0

4.5.4 Opening and Saving a Test Case

The File=>Open, Save and Save As menu commands behave similarly to opening and saving les in most other programs. When you open an existing test case, Selenium-IDE displays its Selenium commands in the Test Case Pane. Test suite les can also be opened and saved via the File menu. However, such operations have their own menu entries near the bottom; the Open, Save, and Save As items are only for les. Note: At the time of this writing, theres a bug, where at times, when the IDE is rst opened and then you select File=>Open, nothing happens. If you see this, close down the IDE and restart it (you dont need to close the browser itself). This will x the problem.

4.6 Running Test Cases

The IDE allows many options for running your test case. You can run a test case all at once, stop and start it, run it one line at a time, run a single command you are currently developing, and you can do a batch run of an entire test suite. Execution of test cases is very exible in the IDE. Run a Test Case Click the Run button to run the currently displayed test case. Run a Test Suite Click the Run All button to run all the test cases in the currently loaded test suite. Stop and Start The Pause button can be used to stop the test case while it is running. The icon of this button then changes to indicate the Resume button. To continue click Resume. Stop in the Middle You can set a breakpoint in the test case to cause it to stop on a particular command. This is useful for debugging your test case. To set a breakpoint, select a command, right-click, and from the context menu select Toggle Breakpoint. Start from the Middle You can tell the IDE to begin running from a specic command in the middle of the test case. This also is used for debugging. To set a startpoint, select a command, right-click, and from the context menu select Set/Clear Start Point. Run Any Single Command Double-click any single command to run it by itself. This is useful when writing a single command. It lets you immediately test a command you are constructing, when you are not sure if it is correct. You can double-click it to see if it runs correctly. This is also available from the context menu.

4.7 Using Base URL to Run Test Cases in Different Domains

The Base URL eld at the top of the Selenium-IDE window is very useful for allowing test cases to be run across different domains. Suppose that a site named http://news.portal.com had an in-house beta site named http://beta.news.portal.com. Any test cases for these sites that begin with an open statement should specify a relative URL as the argument to open rather than an absolute URL (one starting with a protocol such as http: or https:). Selenium-IDE will then create an absolute URL by appending the open commands argument onto the end of the value of Base URL. For example, the test case below would be run against http://news.portal.com/about.html:

24

Chapter 4. Selenium-IDE

Selenium Documentation, Release 1.0

This same test case with a modied http://beta.news.portal.com/about.html:

Base

URL

setting

would

be

run

against

4.8 Debugging
Debugging means nding and xing errors in your test case. This is a normal part of test case development. We wont teach debugging here as most new users to Selenium will already have some basic experience with debugging. If this is new to you, we recommend you ask one of the developers in your organization.

4.8.1 Breakpoints and Startpoints

The Sel-IDE supports the setting of breakpoints and the ability to start and stop the running of a test case, from any point within the test case. That is, one can run up to a specic command in the middle of the test case and inspect how the test case behaves at that point. To do this, set a breakpoint on the command just before the one to be examined. To set a breakpoint, select a command, right-click, and from the context menu select Toggle Breakpoint. Then click the Run button to run your test case from the beginning up to the breakpoint. It is also sometimes useful to run a test case from somewhere in the middle to the end of the test case or up to a breakpoint that follows the starting point. For example, suppose your test case rst logs into the website and then performs a series of tests and you are trying to debug one of those tests. However, you only need to login once, but you need to keep rerunning your tests as you are developing them. You can login once, then run your test case from a startpoint placed after the login portion of your test case. That will prevent you from having to manually logout each time you rerun your test case. To set a startpoint, select a command, right-click, and from the context menu select Set/Clear Start Point.

4.8. Debugging

25

Selenium Documentation, Release 1.0

Then click the Run button to execute the test case beginning at that startpoint.

4.8.2 Stepping Through a Testcase

To execute a test case one command at a time (step through it), follow these steps: 1. Start the test case running with the Run button from the toolbar.

1. Immediately pause the executing test case with the Pause button.

1. Repeatedly select the Step button.

4.8.3 Find Button

The Find button is used to see which UI element on the currently displayed webpage (in the browser) is used in the currently selected Selenium command. This is useful when building a locator for a commands rst parameter (see the section on locators in the Selenium Commands chapter). It can be used with any command that identies a UI element on a webpage, i.e. click, clickAndWait, type, and certain assert and verify commands, among others. From Table view, select any command that has a locator parameter. Click the Find button. Now look on the webpage: There should be a bright green rectangle enclosing the element specied by the locator parameter.

4.8.4 Page Source for Debugging

Often, when debugging a test case, you simply must look at the page source (the HTML for the webpage youre trying to test) to determine a problem. Firefox makes this easy. Simply right-click the webpage and select View->Page Source. The HTML opens in a separate window. Use its Search feature (Edit=>Find) to search for a keyword to nd the HTML for the UI element youre trying to test. Alternatively, select just that portion of the webpage for which you want to see the source. Then rightclick the webpage and select View Selection Source. In this case, the separate HTML window will contain just a small amount of source, with highlighting on the portion representing your selection.

4.8.5 Locator Assistance

Whenever Selenium-IDE records a locator-type argument, it stores additional information which allows the user to view other possible locator-type arguments that could be used instead. This feature can be very useful for learning more about locators, and is often needed to help one build a different type of locator than the type that was recorded.

26

Chapter 4. Selenium-IDE

Selenium Documentation, Release 1.0

This locator assistance is presented on the Selenium-IDE window as a drop-down list accessible at the right end of the Target eld (only when the Target eld contains a recorded locator-type argument). Below is a snapshot showing the contents of this drop-down for one command. Note that the rst column of the drop-down provides alternative locators, whereas the second column indicates the type of each alternative.

4.9 Writing a Test Suite

A test suite is a collection of test cases which is displayed in the leftmost pane in the IDE. The test suite pane can be manually opened or closed via selecting a small dot halfway down the right edge of the pane (which is the left edge of the entire Selenium-IDE window if the pane is closed). The test suite pane will be automatically opened when an existing test suite is opened or when the user selects the New Test Case item from the File menu. In the latter case, the new test case will appear immediately below the previous test case. Selenium-IDE does not yet support loading pre-existing test cases into a test suite. Users who want to create or modify a test suite by adding pre-existing test cases must manually edit a test suite le. A test suite le is an HTML le containing a one-column table. Each cell of each row in the <tbody> section contains a link to a test case. The example below is of a test suite containing four test cases:
<html> <head> <meta http-equiv= "Content-Type" content= "text/html; charset=UTF-8" > <title>Sample Selenium Test Suite</title> </head> <body> <table cellpadding= "1" cellspacing= "1" border= "1" > <thead> <tr><td>Test Cases for De Anza A-Z Directory Links</td></tr> </thead>

4.9. Writing a Test Suite

27

Selenium Documentation, Release 1.0

<tbody> <tr><td><a <tr><td><a <tr><td><a <tr><td><a </tbody> </table> </body> </html>

href= href= href= href=

"./a.html" "./b.html" "./c.html" "./d.html"

>A >B >C >D

Links</a></td></tr> Links</a></td></tr> Links</a></td></tr> Links</a></td></tr>

Note: Test case les should not have to be co-located with the test suite le that invokes them. And on Mac OS and Linux systems, that is indeed the case. However, at the time of this writing, a bug prevents Windows users from being able to place the test cases elsewhere than with the test suite that invokes them.

4.10 User Extensions

User extensions are JavaScript les that allow one to create his or her own customizations and features to add additional functionality. Often this is in the form of customized commands although this extensibility is not limited to additional commands. There are a number of useful extensions created by users. Perhaps the most popular of all Selenium-IDE extensions is one which provides ow control in the form of while loops and primitive conditionals. This extension is the goto_sel_ide.js. For an example of how to use the functionality provided by this extension, look at the page created by its author. To install this extension, put the pathname to its location on your computer in the Selenium Core extensions eld of Selenium-IDEs Options=>Options=>General tab.

After selecting the OK button, you must close and reopen Selenium-IDE in order for the extensions le to be read. Any change you make to an extension will also require you to close and reopen SeleniumIDE. Information on writing your own extensions can be found near the bottom of the Selenium Reference document.

28

Chapter 4. Selenium-IDE

Selenium Documentation, Release 1.0

4.11 Format
Format, under the Options menu, allows you to select a language for saving and displaying the test case. The default is HTML. If you will be using Selenium-RC to run your test cases, this feature is used to translate your test case into a programming language. Select the language, i.e. Java, PHP, you will be using with Selenium-RC for developing your test programs. Then simply save the test case using File=>Save. Your test case will be translated into a series of functions in the language you choose. Essentially, program code supporting your test is generated for you by Selenium-IDE. Also, note that if the generated code does not suit your needs, you can alter it by editing a conguration le which denes the generation process. Each supported language has conguration settings which are editable. This is under the Options=>Options=>Format tab. Note: At the time of this writing, this feature is not yet supported by the Selenium developers. However the author has altered the C# format in a limited manner and it has worked well.

4.12 Executing Selenium-IDE Tests on Different Browsers

While Selenium-IDE can only run tests against Firefox, tests developed with Selenium-IDE can be run against other browsers, using a simple command-line interface that invokes the Selenium-RC server. This topic is covered in the Run Selenese tests section on Selenium-RC chapter. The -htmlSuite command-line option is the particular feature of interest.

4.13 Troubleshooting
Below is a list of image/explanation pairs which describe frequent sources of problems with SeleniumIDE:

4.11. Format

29

Selenium Documentation, Release 1.0

This problem occurs occasionally when Selenium IDE is rst brought up. The solution is to close and reopen Selenium IDE. The bug has been led as SIDE-230.

30

Chapter 4. Selenium-IDE

Selenium Documentation, Release 1.0

Youve used File=>Open to try to open a test suite le. Use File=>Open Test Suite instead.

This type of error may indicate a timing problem, i.e., the element specied by a locator in your command wasnt fully loaded when the command was executed. Try putting a pause 5000 before the command to determine whether the problem is indeed related to timing. If so, investigate using an appropriate waitFor* or *AndWait command immediately before the failing command.

Whenever your attempt to use variable substitution fails as is the case for the open command above, it indicates that you havent actually created the variable whose value youre trying to access. This is sometimes due to putting the variable in the Value eld when it should be in the Target eld or vice versa. In the example above, the two parameters for the store command have been erroneously placed in the reverse order of what is required. For any Selenese command, the rst required parameter must go in the Target eld, and the second required parameter (if one exists) must go in the Value eld.

4.13. Troubleshooting

31

Selenium Documentation, Release 1.0

One of the test cases in your test suite cannot be found. Make sure that the test case is indeed located where the test suite indicates it is located. Also, make sure that your actual test case les have the .html extension both in their lenames, and in the test suite le where they are referenced.

Selenium-IDE is very space-sensitive! An extra space before or after a command will cause it to be unrecognizable.

Your extension les contents have not been read by Selenium-IDE. Be sure you have specied the proper pathname to the extensions le via Options=>Options=>General in the Selenium Core extensions eld. Also, Selenium-IDE must be restarted after any change to either an extensions le or to the contents of the Selenium Core extensions eld.

32

Chapter 4. Selenium-IDE

Selenium Documentation, Release 1.0

This type of error message makes it appear that Selenium-IDE has generated a failure where there is none. However, Selenium-IDE is correct that the actual value does not match the value specied in such test cases. The problem is that the log le error messages collapse a series of two or more spaces into a single space, which is confusing. In the example above, note that the parameter for verifyTitle has two spaces between the words System and Division. The pages actual title has only one space between these words. Thus, Selenium-IDE is correct to generate an error, but is misleading in the nature of the error.

4.13. Troubleshooting

33

Selenium Documentation, Release 1.0

34

Chapter 4. Selenium-IDE

CHAPTER

FIVE

SELENIUM COMMANDS
Selenium commands, often called selenese, are the set of commands that run your tests. A sequence of these commands is a test script. Here we explain those commands in detail, and we present the many choices you have in testing your web application when using Selenium.

5.1 Verifying Page Elements

Verifying UI elements on a web page is probably the most common feature of your automated tests. Selenese allows multiple ways of checking for UI elements. It is important that you understand these different methods because these methods dene what you are actually testing. For example, will you test that... 1. an element is present somewhere on the page? 2. specic text is somewhere on the page? 3. specic text is at a specic location on the page? For example, if you are testing a text heading, the text and its position at the top of the page are probably relevant for your test. If, however, you are testing for the existence of an image on the home page, and the web designers frequently change the specic image le along with its position on the page, then you only want to test that an image (as opposed to the specic image le) exists somewhere on the page.

5.1.1 Assertion or Verication?

Choosing between assert and verify comes down to convenience and management of failures. Theres very little point checking that the rst paragraph on the page is the correct one if your test has already failed when checking that the browser is displaying the expected page. If youre not on the correct page, youll probably want to abort your test case so that you can investigate the cause and x the issue(s) promptly. On the other hand, you may want to check many attributes of a page without aborting the test case on the rst failure as this will allow you to review all failures on the page and take the appropriate action. Effectively an assert will fail the test and abort the current test case, whereas a verify will fail the test and continue to run the test case. The best use of this feature is to logically group your test commands, and start each group with an assert followed by one or more verify test commands. An example follows:

35

Selenium Documentation, Release 1.0

Command open assertTitle verifyText assertTable verifyTable verifyTable

Target /download/ Downloads //h2 1.2.1 1.2.2 1.2.3

Value

Downloads Selenium IDE June 3, 2008 1.0 beta 2

The above example rst opens a page and then asserts that the correct page is loaded by comparing the title with the expected value. Only if this passes will the following command run and verify that the text is present in the expected location. The test case then asserts the rst column in the second row of the rst table contains the expected value, and only if this passed will the remaining cells in that row be veried.

5.1.2 verifyTextPresent
The command verifyTextPresent is used to verify specic text exists somewhere on the page. It takes a single argumentthe text pattern to be veried. For example: Command verifyTextPresent Target Marketing Analysis Value

This would cause Selenium to search for, and verify, that the text string Marketing Analysis appears somewhere on the page currently being tested. Use verifyTextPresent when you are interested in only the text itself being present on the page. Do not use this when you also need to test where the text occurs on the page.

5.1.3 verifyElementPresent
Use this command when you must test for the presence of a specic UI element, rather then its content. This verication does not check the text, only the HTML tag. One common use is to check for the presence of an image. Command verifyElementPresent Target //div/p/img Value

This command veries that an image, specied by the existence of an <img> HTML tag, is present on the page, and that it follows a <div> tag and a <p> tag. The rst (and only) parameter is a locator for telling the Selenese command how to nd the element. Locators are explained in the next section. verifyElementPresent can be used to check the existence of any HTML tag within the page. You can check the existence of links, paragraphs, divisions <div>, etc. Here are a few more examples. Command verifyElementPresent verifyElementPresent verifyElementPresent verifyElementPresent verifyElementPresent verifyElementPresent Target //div/p //div/a id=Login link=Go to Marketing Research //a[2] //head/title Value

These examples illustrate the variety of ways a UI element may be tested. Again, locators are explained in the next section.

36

Chapter 5. Selenium Commands

Selenium Documentation, Release 1.0

5.1.4 verifyText
Use verifyText when both the text and its UI element must be tested. verifyText must use a locator. If you choose an XPath or DOM locator, you can verify that specic text appears at a specic location on the page relative to other UI components on the page. Command verifyText Target //table/tr/td/div/p Value This is my text and it occurs right after the div inside the table.

5.2 Locating Elements

For many Selenium commands, a target is required. This target identies an element in the content of the web application, and consists of the location strategy followed by the location in the format locatorType=location. The locator type can be omitted in many cases. The various locator types are explained below with examples for each.

5.2.1 Locating by Identier

This is probably the most common method of locating elements and is the catch-all default when no recognised locator type is used. With this strategy, the rst element with the id attribute value matching the location will be used. If no element has a matching id attribute, then the rst element with a name attribute matching the location will be used. For instance, your page source could have id and name attributes as follows:
1 2 3 4 5 6 7 8 9

<html> <body> <form id= "loginForm" > <input name= "username" type= "text" /> <input name= "password" type= "password" /> <input name= "continue" type= "submit" value= "Login" /> </form> </body> <html>

The following locator strategies would return the elements from the HTML snippet above indicated by line number: identifier=loginForm (3) identifier=username (4) identifier=continue (5) continue (5) Since the identifier type of locator is the default, the identifier= in the rst three examples above is not necessary.

5.2.2 Locating by Id
This type of locator is more limited than the identier locator type, but also more explicit. Use this when you know an elements id attribute. 5.2. Locating Elements 37

Selenium Documentation, Release 1.0

1 2 3 4 5 6 7 8 9 10

<html> <body> <form id= "loginForm" > <input name= "username" <input name= "password" <input name= "continue" <input name= "continue" </form> </body> <html>

type= type= type= type=

"text" /> "password" /> "submit" value= "Login" /> "button" value= "Clear" />

id=loginForm (3)

5.2.3 Locating by Name

The name locator type will locate the rst element with a matching name attribute. If multiple elements have the same value for a name attribute, then you can use lters to further rene your location strategy. The default lter type is value (matching the value attribute).
1 2 3 4 5 6 7 8 9 10

<html> <body> <form id= "loginForm" > <input name= "username" <input name= "password" <input name= "continue" <input name= "continue" </form> </body> <html>

type= type= type= type=

"text" /> "password" /> "submit" value= "Login" /> "button" value= "Clear" />

name=username (4) name=continue value=Clear (7) name=continue Clear (7) name=continue type=button (7) Note: Unlike some types of XPath and DOM locators, the three types of locators above allow Selenium to test a UI element independent of its location on the page. So if the page structure and organization is altered, the test will still pass. You may or may not want to also test whether the page structure changes. In the case where web designers frequently alter the page, but its functionality must be regression tested, testing via id and name attributes, or really via any HTML property, becomes very important.

5.2.4 Locating by XPath

XPath is the language used for locating nodes in an XML document. As HTML can be an implementation of XML (XHTML), Selenium users can leverage this powerful language to target elements in their web applications. XPath extends beyond (as well as supporting) the simple methods of locating by id

38

Chapter 5. Selenium Commands

Selenium Documentation, Release 1.0

or name attributes, and opens up all sorts of new possibilities such as locating the third checkbox on the page. One of the main reasons for using XPath is when you dont have a suitable id or name attribute for the element you wish to locate. You can use XPath to either locate the element in absolute terms (not advised), or relative to an element that does have an id or name attribute. XPath locators can also be used to specify elements via attributes other than id and name. Absolute XPaths contain the location of all elements from the root (html) and as a result are likely to fail with only the slightest adjustment to the application. By nding a nearby element with an id or name attribute (ideally a parent element) you can locate your target element based on the relationship. This is much less likely to change and can make your tests more robust. Since only xpath locators start with //, it is not necessary to include the xpath= label when specifying an XPath locator.
1 2 3 4 5 6 7 8 9 10

<html> <body> <form id= "loginForm" > <input name= "username" <input name= "password" <input name= "continue" <input name= "continue" </form> </body> <html>

type= type= type= type=

"text" /> "password" /> "submit" value= "Login" /> "button" value= "Clear" />

xpath=/html/body/form[1] (3) - Absolute path (would break if the HTML was changed only slightly) //form[1] (3) - First form element in the HTML xpath=//form[@id=loginForm] (3) - The form element with attribute named id and the value loginForm xpath=//form[input/\@name=username] (4) - First form element with an input child element with attribute named name and the value username //input[@name=username] (4) - First input element with attribute named name and the value username //form[@id=loginForm]/input[1] (4) - First input child element of the form element with attribute named id and the value loginForm //input[@name=continue][@type=button] (7) - Input with attribute named name and the value continue and attribute named type and the value button //form[@id=loginForm]/input[4] (7) - Fourth input child element of the form element with attribute named id and value loginForm These examples cover some basics, but in order to learn more, the following references are recommended: W3Schools XPath Tutorial W3C XPath Recommendation 5.2. Locating Elements 39

Selenium Documentation, Release 1.0

XPath Tutorial - with interactive examples. There are also a couple of very useful Firefox Add-ons that can assist in discovering the XPath of an element: XPath Checker - suggests XPath and can be used to test XPath results. Firebug - XPath suggestions are just one of the many powerful features of this very useful add-on.

5.2.5 Locating Hyperlinks by Link Text

This is a simple method of locating a hyperlink in your web page by using the text of the link. If two links with the same text are present, then the rst match will be used.
1 2 3 4 5 6 7

<html> <body> <p>Are you sure you want to do this?</p> <a href= "continue.html" >Continue</a> <a href= "cancel.html" >Cancel</a> </body> <html>

link=Continue (4) link=Cancel (5)

5.2.6 Locating by DOM

The Document Object Model represents an HTML document and can be accessed using JavaScript. This location strategy takes JavaScript that evaluates to an element on the page, which can be simply the elements location using the hierarchical dotted notation. Since only dom locators start with document, it is not necessary to include the dom= label when specifying a DOM locator.
1 2 3 4 5 6 7 8 9 10

<html> <body> <form id= "loginForm" > <input name= "username" <input name= "password" <input name= "continue" <input name= "continue" </form> </body> <html>

type= type= type= type=

"text" /> "password" /> "submit" value= "Login" /> "button" value= "Clear" />

dom=document.getElementById(loginForm) (3) dom=document.forms[loginForm] (3) dom=document.forms[0] (3)

40

Chapter 5. Selenium Commands

Selenium Documentation, Release 1.0

document.forms[0].username (4) document.forms[0].elements[username] (4) document.forms[0].elements[0] (4) document.forms[0].elements[3] (7) You can use Selenium itself as well as other sites and extensions to explore the DOM of your web application. A good reference exists on W3Schools.

5.2.7 Locating by CSS

CSS (Cascading Style Sheets) is a language for describing the rendering of HTML and XML documents. CSS uses Selectors for binding style properties to elements in the document. These Selectors can be used by Selenium as another locating strategy.
1 2 3 4 5 6 7 8 9 10

<html> <body> <form id= "loginForm" > <input class= "required" name= "username" type= "text" /> <input class= "required passfield" name= "password" type= "password" /> <input name= "continue" type= "submit" value= "Login" /> <input name= "continue" type= "button" value= "Clear" /> </form> </body> <html>

css=form#loginForm (3) css=input[name="username"] (4) css=input.required[type="text"] (4) css=input.passfield (5) css=#loginForm input[type="button"] (4) css=#loginForm input:nth-child(2) (5) For more information about CSS Selectors, the best place to go is the W3C publication. Youll nd additional references there. Note: Most experienced Selenium users recommend CSS as their locating strategy of choice as its considerably faster than XPath and can nd the most complicated objects in an intrinsic HTML document.

5.2.8 Implicit Locators

You can choose to omit the locator type in the following situations: Locators without an explicitly dened locator strategy will default to using the identier locator strategy. See Locating by Identier. 5.2. Locating Elements 41

Selenium Documentation, Release 1.0

Locators starting with // will use the XPath locator strategy. See Locating by XPath. Locators starting with document will use the DOM locator strategy. See Locating by DOM

5.3 Matching Text Patterns

Like locators, patterns are a type of parameter frequently required by Selenese commands. Examples of commands which require patterns are verifyTextPresent, verifyTitle, verifyAlert, assertConrmation, verifyText, and verifyPrompt. And as has been mentioned above, link locators can utilize a pattern. Patterns allow you to describe, via the use of special characters, what text is expected rather than having to specify that text exactly. There are three types of patterns: globbing, regular expressions, and exact.

5.3.1 Globbing Patterns

Most people are familiar with globbing as it is utilized in lename expansion at a DOS or Unix/Linux command line such as ls *.c. In this case, globbing is used to display all the les ending with a .c extension that exist in the current directory. Globbing is fairly limited. Only two special characters are supported in the Selenium implementation: * which translates to match anything, i.e., nothing, a single character, or many characters. [ ] (character class) which translates to match any single character found inside the square brackets. A dash (hyphen) can be used as a shorthand to specify a range of characters (which are contiguous in the ASCII character set). A few examples will make the functionality of a character class clear: [aeiou] matches any lowercase vowel [0-9] matches any digit [a-zA-Z0-9] matches any alphanumeric character In most other contexts, globbing includes a third special character, the ?. However, Selenium globbing patterns only support the asterisk and character class. To specify a globbing pattern parameter for a Selenese command, you can prex the pattern with a glob: label. However, because globbing patterns are the default, you can also omit the label and specify just the pattern itself. Below is an example of two commands that use globbing patterns. The actual link text on the page being tested was Film/Television Department; by using a pattern rather than the exact text, the click command will work even if the link text is changed to Film & Television Department or Film and Television Department. The glob patterns asterisk will match anything or nothing between the word Film and the word Television. Command click verifyTitle Target link=glob:Film*Television Department glob:*Film*Television* Value

The actual title of the page reached by clicking on the link was De Anza Film And Television Department - Menu. By using a pattern rather than the exact text, the verifyTitle will pass as long as the two words Film and Television appear (in that order) anywhere in the pages title. For example, if the pages owner should shorten the title to just Film & Television Department, the test would still pass. Using a pattern for both a link and a simple test that the link worked (such as the verifyTitle above does) can greatly reduce the maintenance for such test cases. 42 Chapter 5. Selenium Commands

Selenium Documentation, Release 1.0

5.3.2 Regular Expression Patterns

Regular expression patterns are the most powerful of the three types of patterns that Selenese supports. Regular expressions are also supported by most high-level programming languages, many text editors, and a host of tools, including the Linux/Unix command-line utilities grep, sed, and awk. In Selenese, regular expression patterns allow a user to perform many tasks that would be very difcult otherwise. For example, suppose your test needed to ensure that a particular table cell contained nothing but a number. regexp: [0-9]+ is a simple pattern that will match a decimal number of any length. Whereas Selenese globbing patterns support only the * and [ ] (character class) features, Selenese regular expression patterns offer the same wide array of special characters that exist in JavaScript. Below are a subset of those special characters: PATTERN . [] * + ? {1,5} | () MATCH any single character character class: any single character that appears inside the brackets quantier: 0 or more of the preceding character (or group) quantier: 1 or more of the preceding character (or group) quantier: 0 or 1 of the preceding character (or group) quantier: 1 through 5 of the preceding character (or group) alternation: the character/group on the left or the character/group on the right grouping: often used with alternation and/or quantier

Regular expression patterns in Selenese need to be prexed with either regexp: or regexpi:. The former is case-sensitive; the latter is case-insensitive. A few examples will help clarify how regular expression patterns can be used with Selenese commands. The rst one uses what is probably the most commonly used regular expression pattern.* (dot star). This two-character sequence can be translated as 0 or more occurrences of any character or more simply, anything or nothing. It is the equivalent of the one-character globbing pattern * (a single asterisk). Command click verifyTitle Target link=regexp:Film.*Television Department regexp:.*Film.*Television.* Value

The example above is functionally equivalent to the earlier example that used globbing patterns for this same test. The only differences are the prex (regexp: instead of glob:) and the anything or nothing pattern (.* instead of just *). The more complex example below tests that the Yahoo! Weather page for Anchorage, Alaska contains info on the sunrise time: Command open verifyTextPresent Target http://weather.yahoo.com/forecast/USAK0012.html regexp:Sunrise: *[0-9]{1,2}:[0-9]{2} [ap]m Value

Lets examine the regular expression above one part at a time: Sunrise: * [0-9]{1,2} : [0-9]{2} [ap]m The string Sunrise: followed by 0 or more spaces 1 or 2 digits (for the hour of the day) The character : (no special characters involved) 2 digits (for the minutes) followed by a space a or p followed by m (am or pm)

5.3. Matching Text Patterns

43

Selenium Documentation, Release 1.0

5.3.3 Exact Patterns

The exact type of Selenium pattern is of marginal usefulness. It uses no special characters at all. So, if you needed to look for an actual asterisk character (which is special for both globbing and regular expression patterns), the exact pattern would be one way to do that. For example, if you wanted to select an item labeled Real * from a dropdown, the following code might work or it might not. The asterisk in the glob:Real * pattern will match anything or nothing. So, if there was an earlier select option labeled Real Numbers, it would be the option selected rather than the Real * option. select //select glob:Real *

In order to ensure that the Real * item would be selected, the exact: prex could be used to create an exact pattern as shown below: select //select exact:Real *

But the same effect could be achieved via escaping the asterisk in a regular expression pattern: select //select regexp:Real \*

Its rather unlikely that most testers will ever need to look for an asterisk or a set of square brackets with characters inside them (the character class for globbing patterns). Thus, globbing patterns and regular expression patterns are sufcient for the vast majority of us.

5.4 The AndWait Commands

The difference between a command and its AndWait alternative is that the regular command (e.g. click) will do the action and continue with the following command as fast as it can, while the AndWait alternative (e.g. clickAndWait) tells Selenium to wait for the page to load after the action has been done. The AndWait alternative is always used when the action causes the browser to navigate to another page or reload the present one. Be aware, if you use an AndWait command for an action that does not trigger a navigation/refresh, your test will fail. This happens because Selenium will reach the AndWaits timeout without seeing any navigation or refresh being made, causing Selenium to raise a timeout exception.

5.5 The waitFor Commands in AJAX applications

In AJAX driven web applications, data is retrieved from server without refreshing the page. Using andWait commands will not work as the page is not actually refreshed. Pausing the test execution for a certain period of time is also not a good approach as web element might appear later or earlier than the stipulated period depending on the systems responsiveness, load or other uncontrolled factors of the moment, leading to test failures. The best approach would be to wait for the needed element in a dynamic period and then continue the execution as soon as the element is found. This is done using waitFor commands, as waitForElementPresent or waitForVisible, which wait dynamically, checking for the desired condition every second and continuing to the next command in the script as soon as the condition is met.

5.6 Sequence of Evaluation and Flow Control

When a script runs, it simply runs in sequence, one command after another. 44 Chapter 5. Selenium Commands

Selenium Documentation, Release 1.0

Selenese, by itself, does not support condition statements (if-else, etc.) or iteration (for, while, etc.). Many useful tests can be conducted without ow control. However, for a functional test of dynamic content, possibly involving multiple pages, programming logic is often needed. When ow control is needed, there are three options: 1. Run the script using Selenium-RC and a client library such as Java or PHP to utilize the programming languages ow control features. 2. Run a small JavaScript snippet from within the script using the storeEval command. 3. Install the goto_sel_ide.js extension. Most testers will export the test script into a programming language le that uses the Selenium-RC API (see the Selenium-IDE chapter). However, some organizations prefer to run their scripts from SeleniumIDE whenever possible (for instance, when they have many junior-level people running tests for them, or when programming skills are lacking). If this is your case, consider a JavaScript snippet or the goto_sel_ide.js extension.

5.7 Store Commands and Selenium Variables

You can use Selenium variables to store constants at the beginning of a script. Also, when combined with a data-driven test design (discussed in a later section), Selenium variables can be used to store values passed to your test program from the command-line, from another program, or from a le. The plain store command is the most basic of the many store commands and can be used to simply store a constant value in a selenium variable. It takes two parameters, the text value to be stored and a selenium variable. Use the standard variable naming conventions of only alphanumeric characters when choosing a name for your variable. Command store Target [email protected] Value userName

Later in your script, youll want to use the stored value of your variable. To access the value of a variable, enclose the variable in curly brackets ({}) and precede it with a dollar sign like this. Command verifyText Target //div/p Value ${userName}

A common use of variables is for storing input for an input eld. Command type Target id=login Value ${userName}

Selenium variables can be used in either the rst or second parameter and are interpreted by Selenium prior to any other operations performed by the command. A Selenium variable may also be used within a locator expression. An equivalent store command exists for each verify and assert command. Here are a couple more commonly used store commands.

5.7.1 storeElementPresent
This corresponds to verifyElementPresent. It simply stores a boolean valuetrue or falsedepending on whether the UI element is found.

5.7. Store Commands and Selenium Variables

45

Selenium Documentation, Release 1.0

5.7.2 storeText
StoreText corresponds to verifyText. It uses a locater to identify specic page text. The text, if found, is stored in the variable. StoreText can be used to extract text from the page being tested.

5.7.3 storeEval
This command takes a script as its rst parameter. Embedding JavaScript within Selenese is covered in the next section. StoreEval allows the test to store the result of running the script in a variable.

5.8 JavaScript and Selenese Parameters

JavaScript can be used with two types of Selenese parameters: script and non-script (usually expressions). In most cases, youll want to access and/or manipulate a test case variable inside the JavaScript snippet used as a Selenese parameter. All variables created in your test case are stored in a JavaScript associative array. An associative array has string indexes rather than sequential numeric indexes. The associative array containing your test cases variables is named storedVars. Whenever you wish to access or manipulate a variable within a JavaScript snippet, you must refer to it as storedVars[yourVariableName].

5.8.1 JavaScript Usage with Script Parameters

Several Selenese commands specify a script parameter including assertEval, verifyEval, storeEval, and waitForEval. These parameters require no special syntax. A Selenium-IDE user would simply place a snippet of JavaScript code into the appropriate eld, normally the Target eld (because a script parameter is normally the rst or only parameter). The example below illustrates how a JavaScript snippet can be used to perform a simple numerical calculation: Command store storeXpathCount storeEval Target 10 //blockquote storedVars[hits]-storedVars[blockquotes] Value hits blockquotes paragraphs

This next example illustrates how a JavaScript snippet can include calls to methods, in this case the JavaScript String objects toUpperCase method and toLowerCase method. Command store storeEval storeEval Target Edith Wharton storedVars[name].toUpperCase() storedVars[name].toLowerCase() Value name uc lc

5.8.2 JavaScript Usage with Non-Script Parameters

JavaScript can also be used to help generate values for parameters, even when the parameter is not specied to be of type script. However, in this case, special syntax is requiredthe JavaScript snippet must be enclosed inside curly braces and preceded by the label javascript, as in javascript {*yourCodeHere*}. Below is an example in which the type commands second parameter value is generated via JavaScript code using this special syntax:

46

Chapter 5. Selenium Commands

Selenium Documentation, Release 1.0

Command store type

Target league of nations q

Value searchString javascript{storedVars[searchString].toUpperCase()}

5.9 echo - The Selenese Print Command

Selenese has a simple command that allows you to print text to your tests output. This is useful for providing informational progress notes in your test which display on the console as your test is running. These notes also can be used to provide context within your test result reports, which can be useful for nding where a defect exists on a page in the event your test nds a problem. Finally, echo statements can be used to print the contents of Selenium variables. Command echo echo Target Testing page footer now. Username is ${userName} Value

5.10 Alerts, Popups, and Multiple Windows

Suppose that you are testing a page that looks like this.
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33

<!DOCTYPE HTML> <html> <head> <script type= "text/javascript" > function output(resultText){ document.getElementById( output ).childNodes[0].nodeValue=resultText; } function show_confirm(){ var confirmation=confirm( "Chose an option." ); if (confirmation==true){ output( "Confirmed." ); } else{ output( "Rejected!" ); } } function show_alert(){ alert( "Im blocking!" ); output( "Alert is gone." ); } function show_prompt(){ var response = prompt( "Whats the best web QA tool?" , "Selenium" ); output(response); } function open_window(windowName){ window.open( "newWindow.html" ,windowName); } </script> </head> <body>

5.9. echo - The Selenese Print Command

47

Selenium Documentation, Release 1.0

34 35 36 37 38 39 40 41 42 43 44 45

<input type= "button" id= <input type= "button" id= <input type= "button" id= <a href= "newWindow.html" <input type= "button" id= <input type= "button" id= <br /> <span id= "output" > </span> </body> </html>

"btnConfirm" onclick= "show_confirm()" value= "Show confirm b "btnAlert" onclick= "show_alert()" value= "Show alert" /> "btnPrompt" onclick= "show_prompt()" value= "Show prompt" /> id= "lnkNewWindow" target= "_blank" >New Window Link</a> "btnNewNamelessWindow" onclick= "open_window()" value= "Open "btnNewNamedWindow" onclick= "open_window(Mike)" value= "Op

The user must respond to alert/conrm boxes, as well as moving focus to newly opened popup windows. Fortunately, Selenium can cover JavaScript pop-ups. But before we begin covering alerts/conrms/prompts in individual detail, it is helpful to understand the commonality between them. Alerts, conrmation boxes and prompts all have variations of the following Command assertFoo(pattern) assertFooPresent assertFooNotPresent storeFoo(variable) storeFooPresent(variable) Description throws error if pattern doesnt match the text of the pop-up throws error if pop-up is not available throws error if any pop-up is present stores the text of the pop-up in a variable stores the text of the pop-up in a variable and returns true or false

When running under Selenium, JavaScript pop-ups will not appear. This is because the function calls are actually being overridden at runtime by Seleniums own JavaScript. However, just because you cannot see the pop-up doesnt mean you dont have do deal with it. To handle a pop-up, you must call its assertFoo(pattern) function. If you fail to assert the presence of a pop-up your next command will be blocked and you will get an error similar to the following [error] Error: There was an unexpected Confirmation! [Chose an option.]

5.10.1 Alerts
Lets start with asserts because they are the simplest pop-up to handle. To begin, open the HTML sample above in a browser and click on the Show alert button. Youll notice that after you close the alert the text Alert is gone. is displayed on the page. Now run through the same steps with Selenium IDE recording, and verify the text is added after you close the alert. Your test will look something like this: Command open click assertAlert verifyTextPresent Target / btnAlert Im blocking Alert is gone. Value

You may be thinking Thats odd, I never tried to assert that alert. But this is Selenium-IDE handling and closing the alert for you. If you remove that step and replay the test you will get the following error [error] Error: There was an unexpected Alert! [Im blocking!]. You must include an assertion of the alert to acknowledge its presence. If you just want to assert that an alert is present but either dont know or dont care what text it contains, you can use assertAlertPresent. This will return true or false, with false halting the test.

48

Chapter 5. Selenium Commands

Selenium Documentation, Release 1.0

5.10.2 Conrmations
Conrmations behave in much the same way as alerts, with assertConfirmation and assertConfirmationPresent offering the same characteristics as their alert counterparts. However, by default Selenium will select OK when a conrmation pops up. Try recording clicking on the Show conrm box button in the sample page, but click on the Cancel button in the popup, then assert the output text. Your test may look something like this: Command open click chooseCancelOnNextConrmation assertConrmation verifyTextPresent Target / btnConrm Choose and option. Rejected Value

The chooseCancelOnNextConfirmation function tells Selenium that all following conrmation should return false. It can be reset by calling chooseOkOnNextConrmation. You may notice that you cannot replay this test, because Selenium complains that there is an unhandled conrmation. This is because the order of events Selenium-IDE records causes the click and chooseCancelOnNextConrmation to be put in the wrong order (it makes sense if you think about it, Selenium cant know that youre cancelling before you open a conrmation) Simply switch these two commands and your test will run ne.

5.10. Alerts, Popups, and Multiple Windows

49

Selenium Documentation, Release 1.0

50

Chapter 5. Selenium Commands

CHAPTER

SIX

SELENIUM-RC
6.1 Introduction
Selenium-RC is the solution for tests that need more than simple browser actions and linear execution. Selenium-RC uses the full power of programming languages to create more complex tests like reading and writing les, querying a database, and emailing test results. Youll want to use Selenium-RC whenever your test requires logic not supported by Selenium-IDE. What logic could this be? For example, Selenium-IDE does not directly support: condition statements iteration logging and reporting of test results error handling, particularly unexpected errors database testing test case grouping re-execution of failed tests test case dependency screenshot capture of test failures Although these tasks are not supported by Selenium directly, all of them can be achieved by using programming techniques with a language-specic Selenium-RC client library. In the Adding Some Spice to Your Tests section, youll nd examples that demonstrate the advantages of using a programming language for your tests.

6.2 How Selenium-RC Works

First, we will describe how the components of Selenium-RC operate and the role each plays in running your test scripts.

51

Selenium Documentation, Release 1.0

6.2.1 RC Components
Selenium-RC components are: The Selenium Server which launches and kills browsers, interprets and runs the Selenese commands passed from the test program, and acts as an HTTP proxy, intercepting and verifying HTTP messages passed between the browser and the AUT. Client libraries which provide the interface between each programming language and the Selenium-RC Server. Here is a simplied architecture diagram....

The diagram shows the client libraries communicate with the Server passing each Selenium command for execution. Then the server passes the Selenium command to the browser using Selenium-Core JavaScript commands. The browser, using its JavaScript interpreter, executes the Selenium command. This runs the Selenese action or verication you specied in your test script.

52

Chapter 6. Selenium-RC

Selenium Documentation, Release 1.0

6.2.2 Selenium Server

Selenium Server receives Selenium commands from your test program, interprets them, and reports back to your program the results of running those tests. The RC server bundles Selenium Core and automatically injects it into the browser. This occurs when your test program opens the browser (using a client library API function). Selenium-Core is a JavaScript program, actually a set of JavaScript functions which interprets and executes Selenese commands using the browsers built-in JavaScript interpreter. The Server receives the Selenese commands from your test program using simple HTTP GET/POST requests. This means you can use any programming language that can send HTTP requests to automate Selenium tests on the browser.

6.2.3 Client Libraries

The client libraries provide the programming support that allows you to run Selenium commands from a program of your own design. There is a different client library for each supported language. A Selenium client library provides a programming interface (API), i.e., a set of functions, which run Selenium commands from your own program. Within each interface, there is a programming function that supports each Selenese command. The client library takes a Selenese command and passes it to the Selenium Server for processing a specic action or test against the application under test (AUT). The client library also receives the result of that command and passes it back to your program. Your program can receive the result and store it into a program variable and report it as a success or failure, or possibly take corrective action if it was an unexpected error. So to create a test program, you simply write a program that runs a set of Selenium commands using a client library API. And, optionally, if you already have a Selenese test script created in the SeleniumIDE, you can generate the Selenium-RC code. The Selenium-IDE can translate (using its Export menu item) its Selenium commands into a client-drivers API function calls. See the Selenium-IDE chapter for specics on exporting RC code from Selenium-IDE.

6.3 Installation
After downloading the Selenium-RC zip le from the downloads page, youll notice it has several subfolders. These folders have all the components you need for using Selenium-RC with the programming language of your choice. Once youve chosen a language to work with, you simply need to: Install the Selenium-RC Server. Set up a programming project using a language specic client driver.

6.3.1 Installing Selenium Server

The Selenium-RC server is simply a Java jar le (selenium-server.jar), which doesnt require any special installation. Just downloading the zip le and extracting the server in the desired directory is sufcient.

6.3. Installation

53

Selenium Documentation, Release 1.0

6.3.2 Running Selenium Server

Before starting any tests you must start the server. Go to the directory where Selenium-RCs server is located and run the following from a command-line console. java -jar selenium-server.jar This can be simplied by creating a batch or shell executable le (.bat on Windows and .sh on Linux) containing the command above. Then make a shortcut to that executable on your desktop and simply double-click the icon to start the server. For the server to run youll need Java installed and the PATH environment variable correctly congured to run it from the console. You can check that you have Java correctly installed by running the following on a console:
java -version

If you get a version number (which needs to be 1.5 or later), youre ready to start using Selenium-RC.

6.3.3 Using the Java Client Driver

Download Selenium-RC from the SeleniumHQ downloads page. Extract the le selenium-java-client-driver.jar. Open your desired Java IDE (Eclipse, NetBeans, IntelliJ, Netweaver, etc.) Create a new project. Add the selenium-java-client-driver.jar les to your project as references. Add to your project classpath the le selenium-java-client-driver.jar. From Selenium-IDE, export a script to a Java le and include it in your Java project, or write your Selenium test in Java using the selenium-java-client API. The API is presented later in this chapter. You can either use JUnit, or TestNg to run your test, or you can write your own simple main() program. These concepts are explained later in this section. Run Selenium server from the console. Execute your test from the Java IDE or from the command-line. For details on Java test project conguration, see the Appendix sections Conguring Selenium-RC With Eclipse and Conguring Selenium-RC With Intellij.

6.3.4 Using the Python Client Driver

Download Selenium-RC from the SeleniumHQ downloads page Extract the le selenium.py Either write your Selenium test in Python or export a script from Selenium-IDE to a python le. Add to your tests path the le selenium.py

54

Chapter 6. Selenium-RC

Selenium Documentation, Release 1.0

Run Selenium server from the console Execute your test from a console or your Python IDE For details on Python client driver conguration, see the appendix Python Client Driver Conguration.

6.3.5 Using the .NET Client Driver

Download Selenium-RC from the SeleniumHQ downloads page Extract the folder Download and install NUnit ( Note: You can use NUnit as your test engine. If youre not familiar yet with NUnit, you can also write a simple main() function to run your tests; however NUnit is very useful as a test engine.) Open your desired .Net IDE (Visual Studio, SharpDevelop, MonoDevelop) Create a class library (.dll) Add references to the following DLLs: nmock.dll, nunit.core.dll, nunit. framework.dll, ThoughtWorks.Selenium.Core.dll, ThoughtWorks.Selenium.IntegrationTests.dll and ThoughtWorks.Selenium.UnitTests.dll Write your Selenium test in a .Net language (C#, VB.Net), or export a script from Selenium-IDE to a C# le and copy this code into the class le you just created. Write your own simple main() program or you can include NUnit in your project for running your test. These concepts are explained later in this chapter. Run Selenium server from console Run your test either from the IDE, from the NUnit GUI or from the command line For specic details on .NET client driver conguration with Visual Studio, see the appendix .NET client driver conguration.

6.3.6 Using the Ruby Client Driver

If you do not already have RubyGems, install it from RubyForge Run gem install selenium-client At the top of your test script, add require "selenium/client" Write your test script using any Ruby test harness (eg Test::Unit, Mini::Test or RSpec). Run Selenium-RC server from the console. Execute your test in the same way you would run any other Ruby script. For details on Ruby client driver conguration, see the Selenium-Client documentation

6.3. Installation

55

Selenium Documentation, Release 1.0

6.4 From Selenese to a Program

The primary task for using Selenium-RC is to convert your Selenese into a programming language. In this section, we provide several different language-specic examples.

6.4.1 Sample Test Script

Lets start with an example Selenese test script. Imagine recording the following test with Seleniumopen / type q selenium rc IDE. clickAndWait btnG assertTextPresent Results * for selenium rc Note: This example would work with the Google search page http://www.google.com

6.4.2 Selenese as Programming Code

Here is the test script exported (via Selenium-IDE) to each of the supported programming languages. If you have at least basic knowledge of an object- oriented programming language, you will understand how Selenium runs Selenese commands by reading one of these examples. To see an example in a specic language, select one of these buttons. In C#:
using using using using using using System; System.Text; System.Text.RegularExpressions; System.Threading; NUnit.Framework; Selenium;

namespace SeleniumTests { [TestFixture] public class NewTest { private ISelenium selenium; private StringBuilder verificationErrors;

[SetUp] public void SetupTest() { selenium = new DefaultSelenium( "localhost" , 4444, "*firefox" , "http://www. selenium.Start(); verificationErrors = new StringBuilder(); } [TearDown] public void TeardownTest() { try { selenium.Stop(); } catch (Exception)

56

Chapter 6. Selenium-RC

Selenium Documentation, Release 1.0

{ // Ignore errors if unable to close the browser } Assert.AreEqual( "" , verificationErrors.ToString()); } [Test] public void TheNewTest() { selenium.Open( "/" ); selenium.Type( "q" , "selenium rc" ); selenium.Click( "btnG" ); selenium.WaitForPageToLoad( "30000" ); Assert.AreEqual( "selenium rc - Google Search" , selenium.GetTitle()); } } }

In Java:
/** Add JUnit framework to your classpath if not already there * for this example to work / * package com.example.tests; import com.thoughtworks.selenium.*; import java.util.regex.Pattern; public class NewTest extends SeleneseTestCase { public void setUp() throws Exception { setUp( "http://www.google.com/" , "*firefox" ); } public void testNew() throws Exception { selenium.open( "/" ); selenium.type( "q" , "selenium rc" ); selenium.click( "btnG" ); selenium.waitForPageToLoad( "30000" ); assertTrue(selenium.isTextPresent( "Results * for selenium rc" )); } }

In Perl:
use use use use use use strict; warnings; Time::HiRes qw( sleep ) ; Test::WWW::Selenium; Test::More "no_plan" ; Test::Exception;

my $sel = Test::WWW::Selenium->new( host => "localhost" , port => 4444, browser => "*firefox" , browser_url => "http://www.google.com/" );

6.4. From Selenese to a Program

57

Selenium Documentation, Release 1.0

$sel->open_ok( "/" ); $sel->type_ok( "q" , "selenium rc" ); $sel->click_ok( "btnG" ); $sel->wait_for_page_to_load_ok( "30000" ); $sel->is_text_present_ok( "Results * for selenium rc" );

In PHP:
<?php require_once PHPUnit/Extensions/SeleniumTestCase.php ; class Example extends PHPUnit_Extensions_SeleniumTestCase { function setUp() { $this->setBrowser( " *firefox " ); $this->setBrowserUrl( " http://www.google.com/ " ); } function testMyTestCase() { $this->open( " / " ); $this->type( " q " , " selenium rc " ); $this->click( " btnG " ); $this->waitForPageToLoad( " 30000 " ); $this->assertTrue($this->isTextPresent( " Results * for selenium rc " )); } } ?>

in Python:
from selenium import selenium import unittest, time, re class NewTest(unittest.TestCase): def setUp(self): self.verificationErrors = [] self.selenium = selenium( " localhost " , 4444, " *firefox " , " http://www.google.com/ " ) self.selenium.start() def test_new(self): sel = self.selenium sel.open( " / " ) sel.type( " q " , " selenium rc " ) sel.click( " btnG " ) sel.wait_for_page_to_load( " 30000 " ) self.failUnless(sel.is_text_present( " Results * for selenium rc " )) def tearDown(self): self.selenium.stop() self.assertEqual([], self.verificationErrors)

58

Chapter 6. Selenium-RC

Selenium Documentation, Release 1.0

in Ruby:
require " selenium " require " test/unit "

class NewTest < Test::Unit::TestCase def setup @verification_errors = [] if $selenium @selenium = $selenium else @selenium = Selenium::SeleniumDriver.new( " localhost " , 4444, " *firefox " , " ht @selenium.start end @selenium.set_context( " test_new " ) end def teardown @selenium.stop unless $selenium assert_equal [], @verification_errors end def test_new @selenium.open " / " @selenium.type " q " , " selenium rc " @selenium.click " btnG " @selenium.wait_for_page_to_load " 30000 " assert @selenium.is_text_present( " Results * for selenium rc " ) end end

In the next section well explain how to build a test program using the generated code.

6.5 Programming Your Test

Now well illustrate how to program your own tests using examples in each of the supported programming languages. There are essentially two tasks: Generate your script into a programming language from Selenium-IDE, optionally modifying the result. Write a very simple main program that executes the generated code. Optionally, you can adopt a test engine platform like JUnit or TestNG for Java, or NUnit for .NET if you are using one of those languages. Here, we show language-specic examples. The language-specic APIs tend to differ from one to another, so youll nd a separate explanation for each. Java C# 6.5. Programming Your Test 59

Selenium Documentation, Release 1.0

Python Ruby Perl, PHP

6.5.1 Java
For Java, people use either JUnit or TestNG as the test engine. Some development environments like Eclipse have direct support for these via plug-ins. This makes it even easier. Teaching JUnit or TestNG is beyond the scope of this document however materials may be found online and there are publications available. If you are already a java-shop chances are your developers will already have some experience with one of these test frameworks. You will probably want to rename the test class from NewTest to something of your own choosing. Also, you will need to change the browser-open parameters in the statement:
selenium = new DefaultSelenium("localhost", 4444, "*iehta", "http://www.google.com/");

The Selenium-IDE generated code will look like this. This example has comments added manually for additional clarity.
package com.example.tests; // We specify the package of our tests import com.thoughtworks.selenium.*; // This is the drivers import. Youll use this for instantiating a // browser and making it do what you need. import java.util.regex.Pattern; // Selenium-IDE add the Pattern module because its sometimes used for // regex validations. You can remove the module if its not used in your // script. public class NewTest extends SeleneseTestCase { // We create our Selenium test case public void setUp() throws Exception { setUp( "http://www.google.com/" , "*firefox" ); // We instantiate and start the browser } public void testNew() throws Exception { selenium.open( "/" ); selenium.type( "q" , "selenium rc" ); selenium.click( "btnG" ); selenium.waitForPageToLoad( "30000" ); assertTrue(selenium.isTextPresent( "Results * for selenium rc" )); // These are the real test steps } }

60

Chapter 6. Selenium-RC

Selenium Documentation, Release 1.0

6.5.2 C#
The .NET Client Driver works with Microsoft.NET. It can be used with any .NET testing framework like NUnit or the Visual Studio 2005 Team System. Selenium-IDE assumes you will use NUnit as your testing framework. You can see this in the generated code below. It includes the using statement for NUnit along with corresponding NUnit attributes identifying the role for each member function of the test class. You will probably have to rename the test class from NewTest to something of your own choosing. Also, you will need to change the browser-open parameters in the statement:
selenium = new DefaultSelenium("localhost", 4444, "*iehta", "http://www.google.com/");

The generated code will look similar to this.

using using using using using using System; System.Text; System.Text.RegularExpressions; System.Threading; NUnit.Framework; Selenium;

namespace SeleniumTests { [TestFixture] public class NewTest { private ISelenium selenium; private StringBuilder verificationErrors; [SetUp] public void SetupTest() { selenium = new DefaultSelenium( "localhost" , 4444, "*iehta" , "http://www.google.com/" ); selenium.Start(); verificationErrors = new StringBuilder(); } [TearDown] public void TeardownTest() { try { selenium.Stop(); }

6.5. Programming Your Test

61

Selenium Documentation, Release 1.0

catch (Exception) { // Ignore errors if unable to close the browser } Assert.AreEqual( "" , verificationErrors.ToString()); } [Test] public void TheNewTest() { // Open Google search engine. selenium.Open( "http://www.google.com/" ); // Assert Title of page. Assert.AreEqual( "Google" , selenium.GetTitle()); // Provide search term as "Selenium OpenQA" selenium.Type( "q" , "Selenium OpenQA" ); // Read the keyed search term and assert it. Assert.AreEqual( "Selenium OpenQA" , selenium.GetValue( "q" )); // Click on Search button. selenium.Click( "btnG" ); // Wait for page to load. selenium.WaitForPageToLoad( "5000" ); // Assert that "www.openqa.org" is available in search results. Assert.IsTrue(selenium.IsTextPresent( "www.openqa.org" )); // Assert that page title is - "Selenium OpenQA - Google Search" Assert.AreEqual( "Selenium OpenQA - Google Search" , selenium.GetTitle()); } } }

You can allow NUnit to manage the execution of your tests. Or alternatively, you can write a simple main() program that instantiates the test object and runs each of the three methods, SetupTest(), TheNewTest(), and TeardownTest() in turn.

6.5.3 Python
Pyunit is the test framework to use for Python. To learn Pyunit refer to its ofcial documentation <http://docs.python.org/library/unittest.html>_. The basic test structure is:
from selenium import selenium # This is the drivers import. Youll use this class for instantiating a # browser and making it do what you need. import unittest, time, re # This are the basic imports added by Selenium-IDE by default.

62

Chapter 6. Selenium-RC

Selenium Documentation, Release 1.0

# You can remove the modules if they are not used in your script. class NewTest(unittest.TestCase): # We create our unittest test case def setUp(self): self.verificationErrors = [] # This is an empty array where we will store any verification errors # we find in our tests self.selenium = selenium( " localhost " , 4444, " *firefox " , " http://www.google.com/ " ) self.selenium.start() # We instantiate and start the browser def test_new(self): # This is the test code. Here you should put the actions you need # the browser to do during your test. sel = self.selenium # We assign the browser to the variable "sel" (just to save us from # typing "self.selenium" each time we want to call the browser). sel.open( " / " ) sel.type( " q " , " selenium rc " ) sel.click( " btnG " ) sel.wait_for_page_to_load( " 30000 " ) self.failUnless(sel.is_text_present( " Results * for selenium rc " )) # These are the real test steps def tearDown(self): self.selenium.stop() # we close the browser (Id recommend you to comment this line while # you are creating and debugging your tests) self.assertEqual([], self.verificationErrors) # And make the test fail if we found that any verification errors # were found

6.5.4 Ruby
Selenium-IDE generates reasonable Ruby, but requires the old Selenium gem. This is a problem because the ofcial Ruby driver for Selenium is the Selenium-Client gem, not the old Selenium gem. In fact, the Selenium gem is no longer even under active development. Therefore, it is advisable to update any Ruby scripts generated by the IDE as follows: 1. On line 1, change require "selenium" to require "selenium/client" 2. On line 11, change Selenium::Client::Driver.new Selenium::SeleniumDriver.new to

You probably also want to change the class name to something more informative than Untitled, and change the test methods name to something other than test_untitled. Here is a simple example created by modifying the Ruby code generated by Selenium IDE, as described

6.5. Programming Your Test

63

Selenium Documentation, Release 1.0

above.
# load the Selenium-Client gem require " selenium/client " # Load Test::Unit, Ruby 1.8s default test framework. # If you prefer RSpec, see the examples in the Selenium-Client # documentation. require " test/unit " class Untitled < Test::Unit::TestCase # The setup method is called before each test. def setup # This array is used to capture errors and display them at the # end of the test run. @verification_errors = [] # Create a new instance of the Selenium-Client driver. @selenium = Selenium::Client::Driver.new \ :host => " localhost " , :port => 4444, :browser => " *chrome " , :url => " http://www.google.com/ " , :timeout_in_second => 60 # Start the browser session @selenium.start # Print a message in the browser-side log and status bar # (optional). @selenium.set_context( " test_untitled " ) end # The teardown method is called after each test. def teardown # Stop the browser session. @selenium.stop # Print the array of error messages, if any. assert_equal [], @verification_errors end # This is the main body of your test. def test_untitled # Open the root of the site we specified when we created the # new driver instance, above. @selenium.open " / " # Type selenium rc into the field named q @selenium.type " q " , " selenium rc " # Click the button named "btnG" @selenium.click " btnG "

64

Chapter 6. Selenium-RC

Selenium Documentation, Release 1.0

# Wait for the search results page to load. # Note that we dont need to set a timeout here, because that # was specified when we created the new driver instance, above. @selenium.wait_for_page_to_load begin # Test whether the search results contain the expected text. # Notice that the star (*) is a wildcard that matches any # number of characters. assert @selenium.is_text_present( " Results * for selenium rc " ) rescue Test::Unit::AssertionFailedError # If the assertion fails, push it onto the array of errors. @verification_errors << $! end end end

6.5.5 Perl, PHP

The members of the documentation team have not used Selenium-RC with Perl or PHP. If you are using Selenium-RC with either of these two languages please contact the Documentation Team (see the chapter on contributing). We would love to include some examples from you and your experiences, to support Perl and PHP users.

6.6 Learning the API

The Selenium-RC API uses naming conventions that, assuming you understand Selenese, much of the interface will be self-explanatory. Here, however, we explain the most critical and possibly less obvious aspects.

6.6.1 Starting the Browser

In C#:

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

In Java:
setUp( "http://www.google.com/" , "*firefox" );

In Perl:
my $sel = Test::WWW::Selenium->new( host => "localhost" , port => 4444,

6.6. Learning the API

65

Selenium Documentation, Release 1.0

browser => "*firefox" , browser_url => "http://www.google.com/" );

In PHP:
$this->setBrowser("*firefox"); $this->setBrowserUrl("http://www.google.com/");

In Python:
self.selenium = selenium( " localhost " , 4444, " *firefox " , " http://www.google.com/ " ) self.selenium.start()

In Ruby:

@selenium = Selenium::ClientDriver.new( " localhost " , 4444, " *firefox " , " http://www @selenium.start

Each of these examples opens the browser and represents that browser by assigning a browser instance to a program variable. This program variable is then used to call methods from the browser. These methods execute the Selenium commands, i.e. like open or type or the verify commands. The parameters required when creating the browser instance are: host Species the IP address of the computer where the server is located. Usually, this is the same machine as where the client is running, so in this case localhost is passed. In some clients this is an optional parameter. port Species the TCP/IP socket where the server is listening waiting for the client to establish a connection. This also is optional in some client drivers. browser The browser in which you want to run the tests. This is a required parameter. url The base url of the application under test. This is required by all the client libs and is integral information for starting up the browser-proxy-AUT communication. Note that some of the client libraries require the browser to be started explicitly by calling its start() method.

6.6.2 Running Commands

Once you have the browser initialized and assigned to a variable (generally named selenium) you can make it run Selenese commands by calling the respective methods from the browser variable. For example, to call the type method of the selenium object:
selenium.type( " field-id " , " string to type " )

In the background the browser will actually perform a type operation, essentially identical to a user typing input into the browser, by using the locator and the string you specied during the method call.

66

Chapter 6. Selenium-RC

Selenium Documentation, Release 1.0

6.7 Reporting Results

Selenium-RC does not have its own mechanism for reporting results. Rather, it allows you to build your reporting customized to your needs using features of your chosen programming language. Thats great, but what if you simply want something quick thats already done for you? Often an existing library or test framework can meet your needs faster than developing your own test reporting code.

6.7.1 Test Framework Reporting Tools

Test frameworks are available for many programming languages. These, along with their primary function of providing a exible test engine for executing your tests, include library code for reporting results. For example, Java has two commonly used test frameworks, JUnit and TestNG. .NET also has its own, NUnit. We wont teach the frameworks themselves here; thats beyond the scope of this user guide. We will simply introduce the framework features that relate to Selenium along with some techniques you can apply. There are good books available on these test frameworks however along with information on the internet.

6.7.2 Test Report Libraries

Also available are third-party libraries specically created for reporting test results in your chosen programming language. These often support a variety of formats such as HTML or PDF.

6.7.3 Whats The Best Approach?

Most people new to the testing frameworks will being with the frameworks built-in reporting features. From there most will examine any available libraries as thats less time consuming than developing your own. As you begin to use Selenium no doubt you will start putting in your own print statements for reporting progress. That may gradually lead to you developing your own reporting, possibly in parallel to using a library or test framework. Regardless, after the initial, but short, learning curve you will naturally develop what works best for your own situation.

6.7.4 Test Reporting Examples

To illustrate, well direct you to some specic tools in some of the other languages supported by Selenium. The ones listed here are commonly used and have been used extensively (and therefore recommended) by the authors of this guide. Test Reports in Java If Selenium Test cases are developed using JUnit then JUnit Report can be used to generate test reports. Refer to JUnit Report for specics. If Selenium Test cases are developed using TestNG then no external task is required to generate test reports. The TestNG framework generates an HTML report which list details of tests. See TestNG Report for more.

6.7. Reporting Results

67

Selenium Documentation, Release 1.0

ReportNG is a HTML reporting plug-in for the TestNG framework. It is intended as a replacement for the default TestNG HTML report. ReportNG provides a simple, colour-coded view of the test results. See ReportNG for more. Also, for a very nice summary report try using TestNG-xslt. A TestNG-xslt Report looks like this.

See TestNG-xslt for more. Logging the Selenese Commands Logging Selenium can be used to generate a report of all the Selenese commands in your test along with the success or failure of each. Logging Selenium extends the Java client driver to add this Selenese logging ability. Please refer to Logging Selenium. Test Reports for Python When using Python Client Driver then HTMLTestRunner can be used to generate a Test Report. See HTMLTestRunner. Test Reports for Ruby If RSpec framework is used for writing Selenium Test Cases in Ruby then its HTML report can be used to generate a test report. Refer to RSpec Report for more. Note: If you are interested in a language independent log of whats going on, take a look at Selenium Server Logging

6.8 Adding Some Spice to Your Tests

Now well get to the whole reason for using Selenium-RC, adding programming logic to your tests. Its the same as for any program. Program ow is controlled using condition statements and iteration. In addition you can report progress information using I/O. In this section well show some examples of how programming language constructs can be combined with Selenium to solve common testing problems. You will nd as you transition from the simple tests of the existence of page elements to tests of dynamic functionality involving multiple web-pages and varying data that you will require programming logic for verifying expected results. Basically, the Selenium-IDE does not support iteration and standard condition statements. You can do some conditions by embedding javascript in Selenese parameters, 68 Chapter 6. Selenium-RC

Selenium Documentation, Release 1.0

however iteration is impossible, and most conditions will be much easier in a programming language. In addition, you may need exception handling for error recovery. For these reasons and others, we have written this section to illustrate the use of common programming techniques to give you greater verication power in your automated testing. The examples in this section are written in C# and Java, although the code is simple and can be easily adapted to the other supported languages. If you have some basic knowledge of an object-oriented programming language you shouldnt have difculty understanding this section.

6.8.1 Iteration
Iteration is one of the most common things people need to do in their tests. For example, you may want to to execute a search multiple times. Or, perhaps for verifying your test results you need to process a result set returned from a database. Using the same Google search example we used earlier, lets check the Selenium search results. This test could use the Selenese: open type clickAndWait assertTextPresent type clickAndWait assertTextPresent type clickAndWait assertTextPresent / q btnG Results * for selenium rc q btnG Results * for selenium ide q btnG Results * for selenium grid selenium rc

selenium ide

selenium grid

The code has been repeated to run the same steps 3 times. But multiple copies of the same code is not good program practice because its more work to maintain. By using a programming language, we can iterate over the search results for a more exible and maintainable solution. In C#:
// Collection of String values. String[] arr = { "ide" , "rc" , "grid" }; // Execute loop for each String in array arr. foreach (String s in arr) { sel.open( "/" ); sel.type( "q" , "selenium " +s); sel.click( "btnG" ); sel.waitForPageToLoad( "30000" ); assertTrue( "Expected text: " +s+ " is missing on page." , sel.isTextPresent( "Results * for selenium " + s)); }

6.8.2 Condition Statements

To illustrate using conditions in tests well start with an example. A common problem encountered while running Selenium tests occurs when an expected element is not available on page. For example, when running the following line:

6.8. Adding Some Spice to Your Tests

69

Selenium Documentation, Release 1.0

selenium.type( "q" , "selenium " +s);

If element q is not on the page then an exception is thrown:

com.thoughtworks.selenium.SeleniumException: ERROR: Element q not found

This can cause your test to abort. For some tests thats what you want. But often that is not desirable as your test script has many other subsequent tests to perform. A better approach is to rst validate if the element is really present and then take alternatives when it it is not. Lets look at this using Java.
// If element is available on page then perform type operation. if(selenium.isElementPresent( "q" )) { selenium.type( "q" , "Selenium rc" ); } else { System.out.printf( "Element: " +q+ " is not available on page." ) }

The advantage of this approach is to continue with test execution even if some UI elements are not available on page.

6.8.3 Executing JavaScript from Your Test

JavaScript comes very handy in exercising an application which is not directly supported by selenium. The getEval method of selenium API can be used to execute JavaScript from selenium RC. Consider an application having check boxes with no static identiers. In this case one could evaluate JavaScript from selenium RC to get ids of all check boxes and then exercise them.

public static String[] getAllCheckboxIds () { String script = "var inputId = new Array();" ;// Create array in java scri script += "var cnt = 0;" ; // Counter for check box ids. script += "var inputFields = new Array();" ; // Create array in java scrip script += "inputFields = window.document.getElementsByTagName(input);" ; script += "for(var i=0; i<inputFields.length; i++) {" ; // Loop through the script += "if(inputFields[i].id !=null " + "&& inputFields[i].id !=undefined " + "&& inputFields[i].getAttribute(type) == checkbox) {" ; // If input fie script += "inputId[cnt]=inputFields[i].id ;" + // Save check box id to inp "cnt++;" + // increment the counter. "}" + // end of if. "}" ; // end of for. script += "inputId.toString();" ;// Convert array in to string. String[] checkboxIds = selenium.getEval(script).split( "," ); // Split the s return checkboxIds; }

To count number of images on a page:

70

Chapter 6. Selenium-RC

Selenium Documentation, Release 1.0

selenium.getEval( "window.document.images.length;" );

Remember to use window object in case of DOM expressions as by default selenium window is referred to, not the test window.

6.9 Server Options

When the server is launched, command line options can be used to change the default server behaviour. Recall, the server is started by running the following.
$ java -jar selenium-server.jar

To see the list of options, run the server with the -h option.
$ java -jar selenium-server.jar -h

Youll see a list of all the options you can use with the server and a brief description of each. The provided descriptions will not always be enough, so weve provided explanations for some of the more important options.

6.9.1 Proxy Conguration

If your AUT is behind an HTTP proxy which requires authentication then you should congure http.proxyHost, http.proxyPort, http.proxyUser and http.proxyPassword using the following command.

$ java -jar selenium-server.jar -Dhttp.proxyHost=proxy.com -Dhttp.proxyPort=8080 -Dhttp.

6.9.2 Multi-Window Mode

If you are using Selenium 1.0 you can probably skip this section, since multiwindow mode is the default behavior. However, prior to version 1.0, Selenium by default ran the application under test in a sub frame as shown here.

6.9. Server Options

71

Selenium Documentation, Release 1.0

Some applications didnt run correctly in a sub frame, and needed to be loaded into the top frame of the window. The multi-window mode option allowed the AUT to run in a separate window rather than in the default frame where it could then have the top frame it required.

72

Chapter 6. Selenium-RC

Selenium Documentation, Release 1.0

For older versions of Selenium you must specify multiwindow mode explicitly with the following option:
-multiwindow

As of Selenium-RC 1.0, if you want to run your test within a single frame (i.e. using the standard for earlier Selenium versions) you can state this to the Selenium Server using the option
-singlewindow

6.9.3 Specifying the Firefox Prole

Firefox will not run two instances simultaneously unless you specify a separate prole for each instance. Selenium-RC 1.0 and later runs in a separate prole automatically, so if you are using Selenium 1.0, you can probably skip this section. However, if youre using an older version of Selenium or if you need to use a specic prole for your tests (such as adding an https certicate or having some addons installed), you will need to explicitly specify the prole. First, to create a separate Firefox prole, follow this procedure. Open the Windows Start menu, select Run, then type and enter one of the following:
firefox.exe -profilemanager

6.9. Server Options

73

Selenium Documentation, Release 1.0

firefox.exe -P

Create the new prole using the dialog. Then when you run Selenium Server, tell it to use this new Firefox prole with the server command-line option -refoxProleTemplate and specify the path to the prole using its lename and directory path.
-firefoxProfileTemplate "path to the profile"

Warning: Be sure to put your prole in a new folder separate from the default!!! The Firefox prole manager tool will delete all les in a folder if you delete a prole, regardless of whether they are prole les or not. More information about Firefox proles can be found in Mozillas Knowledge Base

6.9.4 Run Selenese Directly Within the Server Using -htmlSuite

You can run Selenese html les directly within the Selenium Server by passing the html le to the servers command line. For instance:

java -jar selenium-server.jar -htmlSuite "*firefox" "http://www.google.com" "c:\absolute

This will automatically launch your HTML suite, run all the tests and save a nice HTML report with the results. Note: When using this option, the server will start the tests and wait for a specied number of seconds for the test to complete; if the test doesnt complete within that amount of time, the command will exit with a non-zero exit code and no results le will be generated. This command line is very long so be careful when you type it. Note this requires you to pass in an HTML Selenese suite, not a single test. Also be aware the -htmlSuite option is incompatible with -interactive You cannot run both at the same time.

6.9.5 Selenium Server Logging

Server-Side Logs When launching selenium server the -log option can be used to record valuable debugging information reported by the Selenium Server to a text le.
java -jar selenium-server.jar -log selenium.log

This log le is more verbose than the standard console logs (it includes DEBUG level logging messages). The log le also includes the logger name, and the ID number of the thread that logged the message. For example:
20:44:25 DEBUG [12] org.openqa.selenium.server.SeleniumDriverResourceHandler Browser 465828/:top frame1 posted START NEW

The message format is

74

Chapter 6. Selenium-RC

Selenium Documentation, Release 1.0

TIMESTAMP(HH:mm:ss) LEVEL [THREAD] LOGGER - MESSAGE

This message may be multiline. Browser-Side Logs JavaScript on the browser side (Selenium Core) also logs important messages; in many cases, these can be more useful to the end-user than the regular Selenium Server logs. To access browser-side logs, pass the -browserSideLog argument to the Selenium Server.
java -jar selenium-server.jar -browserSideLog

-browserSideLog must be combined with the -log argument, to log browserSideLogs (as well as all other DEBUG level logging messages) to a le.

6.10 Specifying the Path to a Specic Browser

You can specify to Selenium-RC a path to a specic browser. This is useful if you have different versions of the same browser and you wish to use a specic one. Also, this is used to allow your tests to run against a browser not directly supported by Selenium-RC. When specifying the run mode, use the *custom specier followed by the full path to the browsers executable:
*custom <path to browser>

6.11 Selenium-RC Architecture

Note: This topic tries to explain the technical implementation behind Selenium-RC. Its not fundamental for a Selenium user to know this, but could be useful for understanding some of the problems you might nd in the future. To understand in detail how Selenium-RC Server works and why it uses proxy injection and heightened privilege modes you must rst understand the same origin policy.

6.11.1 The Same Origin Policy

The main restriction that Selenium faces is the Same Origin Policy. This security restriction is applied by every browser in the market and its objective is to ensure that a sites content will never be accessible by a script from another site. The Same Origin Policy dictates that any code loaded within the browser can only operate within that websites domain. It cannot perform functions on another website. So for example, if the browser loads JavaScript code when it loads www.mysite.com, it cannot run that loaded code against www.mysite2.comeven if thats another of your sites. If this were possible, a script placed on any website you open would be able to read information on your bank account if you had the account page opened on other tab. This is called XSS (Cross-site Scripting). To work within this policy, Selenium-Core (and its JavaScript commands that make all the magic happen) must be placed in the same origin as the Application Under Test (same URL).

6.10. Specifying the Path to a Specic Browser

75

Selenium Documentation, Release 1.0

Historically, Selenium-Core was limited by this problem since it was implemented in JavaScript. Selenium-RC is not, however, restricted by the Same Origin Policy. Its use of the Selenium Server as a proxy avoids this problem. It, essentially, tells the browser that the browser is working on a single spoofed website that the Server provides. Note: You can nd additional information about this topic on Wikipedia pages about Same Origin Policy and XSS.

6.11.2 Proxy Injection

The rst method Selenium used to avoid the The Same Origin Policy was Proxy Injection. In Proxy Injection Mode, the Selenium Server acts as a client-congured 1 HTTP proxy 2 , that sits between the browser and the Application Under Test. It then masks the AUT under a ctional URL (embedding Selenium-Core and the set of tests and delivering them as if they were coming from the same origin). Here is an architectural diagram.
The proxy is a third person in the middle that passes the ball between the two parts. It acts as a web server that delivers the AUT to the browser. Being a proxy gives Selenium Server the capability of lying about the AUTs real URL. 2 The browser is launched with a conguration prole that has set localhost:4444 as the HTTP proxy, this is why any HTTP request that the browser does will pass through Selenium server and the response will pass through it and not from the real server.
1

76

Chapter 6. Selenium-RC

Selenium Documentation, Release 1.0

As a test suite starts in your favorite language, the following happens: 1. The client/driver establishes a connection with the selenium-RC server. 2. Selenium-RC server launches a browser (or reuses an old one) with a URL that injects SeleniumCores JavaScript into the browser-loaded web page. 3. The client-driver passes a Selenese command to the server. 4. The Server interprets the command and then triggers the corresponding JavaScript execution to execute that command within the browser. 5. Selenium-Core instructs the browser to act on that rst instruction, typically opening a page of the AUT. 6. The browser receives the open request and asks for the websites content from the Selenium-RC server (set as the HTTP proxy for the browser to use). 7. Selenium-RC server communicates with the Web server asking for the page and once it receives it, it sends the page to the browser masking the origin to look like the page comes from the same server as Selenium-Core (this allows Selenium-Core to comply with the Same Origin Policy). 6.11. Selenium-RC Architecture 77

Selenium Documentation, Release 1.0

8. The browser receives the web page and renders it in the frame/window reserved for it.

6.11.3 Heightened Privileges Browsers

This workow in this method is very similar to Proxy Injection but the main difference is that the browsers are launched in a special mode called Heightened Privileges, which allows websites to do things that are not commonly permitted (as doing XSS, or lling le upload inputs and pretty useful stuff for Selenium). By using these browser modes, Selenium Core is able to directly open the AUT and read/interact with its content without having to pass the whole AUT through the Selenium-RC server. Here is the architectural diagram.

As a test suite starts in your favorite language, the following happens: 1. The client/driver establishes a connection with the selenium-RC server. 2. Selenium-RC server launches a browser (or reuses an old one) with a URL that will load SeleniumCore in the web page. 3. Selenium-Core gets the rst instruction from the client/driver (via another HTTP request made to the Selenium-RC Server). 78 Chapter 6. Selenium-RC

Selenium Documentation, Release 1.0

4. Selenium-Core acts on that rst instruction, typically opening a page of the AUT. 5. The browser receives the open request and asks the Web Server for the page. Once the browser receives the web page, renders it in the frame/window reserved for it.

6.12 Handling HTTPS and Security Popups

Many applications switch from using HTTP to HTTPS when they need to send encrypted information such as passwords or credit card information. This is common with many of todays web applications. Selenium-RC supports this. To ensure the HTTPS site is genuine, the browser will need a security certicate. Otherwise, when the browser accesses the AUT using HTTPS, it will assume that application is not trusted. When this occurs the browser displays security popups, and these popups cannot be closed using Selenium-RC. When dealing with HTTPS in a Selenium-RC test, you must use a run mode that supports this and handles the security certicate for you. You specify the run mode when your test program initializes Selenium. In Selenium-RC 1.0 beta 2 and later use *refox or *iexplore for the run mode. In earlier versions, including Selenium-RC 1.0 beta 1, use *chrome or *iehta, for the run mode. Using these run modes, you will not need to install any special security certicates; Selenium-RC will handle it for you. In version 1.0 the run modes *refox or *iexplore are recommended. However, there are additional run modes of *iexploreproxy and *refoxproxy. These are provided for backwards compatibility only, and should not be used unless required by legacy test programs. Their use will present limitations with security certicate handling and with the running of multiple windows if your application opens additional browser windows. In earlier versions of Selenium-RC, *chrome or *iehta were the run modes that supported HTTPS and the handling of security popups. These were considered experimental modes although they became quite stable and many people used them. If you are using Selenium 1.0 you do not need, and should not use, these older run modes.

6.12.1 Security Certicates Explained

Normally, your browser will trust the application you are testing by installing a security certicate which you already own. You can check this in your browsers options or Internet properties (if you dont know your AUTs security certicate ask your system administrator). When Selenium loads your browser it injects code to intercept messages between the browser and the server. The browser now thinks untrusted software is trying to look like your application. It responds by alerting you with popup messages. To get around this, Selenium-RC, (again when using a run mode that support this) will install its own security certicate, temporarily, to your client machine in a place where the browser can access it. This tricks the browser into thinking its accessing a site different from your AUT and effectively suppresses the popups. Another method used with earlier versions of Selenium was to install the Cybervillians security certicate provided with your Selenium installation. Most users should no longer need to do this however; if you are running Selenium-RC in proxy injection mode, you may need to explicitly install this security certicate.

6.12. Handling HTTPS and Security Popups

79

Selenium Documentation, Release 1.0

6.13 Supporting Additional Browsers and Browser Congurations

The Selenium API supports running against multiple browsers in addition to Internet Explorer and Mozilla Firefox. See the SeleniumHQ.org website for supported browsers. In addition, when a browser is not directly supported, you may still run your Selenium tests against a browser of your choosing by using the *custom run-mode (i.e. in place of *refox or *iexplore) when your test application starts the browser. With this, you pass in the path to the browsers executable within the API call. This can also be done from the Server in interactive mode.

cmd=getNewBrowserSession&1=*custom c: \P rogram Files \M ozilla Firefox \M yBrowser.exe&2=h

6.13.1 Running Tests with Different Browser Congurations

Normally Selenium-RC automatically congures the browser, but if you launch the browser using the *custom run mode, you can force Selenium RC to launch the browser as-is, without using an automatic conguration. For example, you can launch Firefox with a custom conguration like this:

cmd=getNewBrowserSession&1=*custom c: \P rogram Files \M ozilla Firefox \f irefox.exe&2=htt

Note that when launching the browser this way, you must manually congure the browser to use the Selenium Server as a proxy. Normally this just means opening your browser preferences and specifying localhost:4444 as an HTTP proxy, but instructions for this can differ radically from browser to browser. Consult your browsers documentation for details. Be aware that Mozilla browsers can vary in how they start and stop. One may need to set the MOZ_NO_REMOTE environment variable to make Mozilla browsers behave a little more predictably. Unix users should avoid launching the browser using a shell script; its generally better to use the binary executable (e.g. refox-bin) directly.

6.14 Troubleshooting Common Problems

When getting started with Selenium-RC theres a few potential problems that are commonly encountered. We present them along with their solutions here.

6.14.1 Unable to Connect to Server

When your test program cannot connect to the Selenium Server, an exception will be thrown in your test program. It should display this message or a similar one:
"Unable to connect to remote server....Inner Exception Message: No connection could be made because the target machine actively refused it...." (using .NET and XP Service Pack 2)

If you see a message like this, be sure you started the Selenium Server. If so, then there is a problem with the connectivity between the Selenium Client Library and the Selenium Server.

80

Chapter 6. Selenium-RC

Selenium Documentation, Release 1.0

When starting with Selenium-RC, most people begin by running thier test program (with a Selenium Client Library) and the Selenium Server on the same machine. To do this use localhost as your connection parameter. We recommend beginning this way since it reduces the inuence of potential networking problems which youre getting started. Assuming your operating system has typical networking and TCP/IP settings you should have little difculty. In truth, many people choose to run the tests this way. If, however, you do want to run Selenium Server on a remote machine, the connectivity should be ne assuming you have valid TCP/IP connectivity between the two machines. If you have difculty connecting, you can use common networking tools like ping, telnet, ifcong(Unix)/ipcong (Windows), etc to ensure you have a valid network connection. If unfamilar with these, your system administrator can assist you.

6.14.2 Unable to Load the Browser

Ok, not a friendly error message, sorry, but if the Selenium Server cannot load the browser you will likley see this error.
(500) Internal Server Error

This could be caused by Firefox (prior to Selenium 1.0) cannot start because the browser is already open and you did not specify a separate prole. See the section on Firefox proles under Server Options. The run mode youre using doesnt match any browser on your machine. Check the parameters you passed to Selenium when you program opens the browser. You specied the path to the browser explicitly (using *customsee above) but the path is incorrect. Check to be sure the path is correct. Also check the user group to be sure there are no known issues with your browser and the *custom parameters.

6.14.3 Selenium Cannot Find the AUT

If your test program starts the browser successfully, but the browser doesnt display the website youre testing, the most likely cause is your test program is not using the correct URL. This can easily happen. When you use Selenium-IDE to export your script, it inserts a dummy URL. You must manually change the URL to the correct one for your application to be tested.

6.14.4 Firefox Refused Shutdown While Preparing a Prole

This most often occurs when your run your Selenium-RC test program against Firefox, but you already have a Firefox browser session running and, you didnt specify a separate prole when you started the Selenium Server. The error from the test program looks like this:
Error: java.lang.RuntimeException: Firefox refused shutdown while preparing a profile

Heres the complete error message from the server:

6.14. Troubleshooting Common Problems

81

Selenium Documentation, Release 1.0

16:20:03.919 INFO - Preparing Firefox profile... 16:20:27.822 WARN - GET /selenium-server/driver/?cmd=getNewBrowserSession&1=*fir efox&2=https%3a%2f%2f2.zoppoz.workers.dev%3a443%2fhttp%2fsage-webapp1.qa.idc.com HTTP/1.1 java.lang.RuntimeException: Firefox refused shutdown while preparing a profile at org.openqa.selenium.server.browserlaunchers.FirefoxCustomProfileLaunc her.waitForFullProfileToBeCreated(FirefoxCustomProfileLauncher.java:277) ......................... Caused by: org.openqa.selenium.server.browserlaunchers.FirefoxCustomProfileLaunc her$FileLockRemainedException: Lock file still present! C:\DOCUME~1\jsvec\LOCALS ~1\Temp\customProfileDir203138\parent.lock

To resolve this, see the section on Specifying a Separate Firefox Prole

6.14.5 Versioning Problems

Make sure your version of Selenium supports the version of your browser. For example, Selenium-RC 0.92 does not support Firefox 3. At times you may be lucky (I was). But dont forget to check which browser versions are supported by the version of Selenium you are using. When in doubt, use the latest release version of Selenium with the most widely used version of your browser.

6.14.6 Error message: (Unsupported major.minor version 49.0) while starting server
This error says youre not using a correct version of Java. The Selenium Server requires Java 1.5 or higher. To check double-check your java version, run this from the command line.
java -version

You should see a message showing the Java version.

java version "1.5.0_07" Java(TM) 2 Runtime Environment, Standard Edition (build 1.5.0_07-b03) Java HotSpot(TM) Client VM (build 1.5.0_07-b03, mixed mode)

If you see a lower version number, you may need to update the JRE, or you may simply need to add it to your PATH environment variable.

6.14.7 404 error when running the getNewBrowserSession command

If youre getting a 404 error while attempting to open a page on http://www.google.com/seleniumserver/, then it must be because the Selenium Server was not correctly congured as a proxy. The selenium-server directory doesnt exist on google.com; it only appears to exist when the proxy is properly congured. Proxy Conguration highly depends on how the browser is launched with *refox, *iexplore, *opera, or *custom. *iexplore: If the browser is launched using *iexplore, you could be having a problem with Internet Explorers proxy settings. Selenium Server attempts To congure the global proxy settings in the Internet Options Control Panel. You must make sure that those are correctly congured when Selenium Server launches the browser. Try 82 Chapter 6. Selenium-RC

Selenium Documentation, Release 1.0

looking at your Internet Options control panel. Click on the Connections tab and click on LAN Settings. If you need to use a proxy to access the application you want to test, youll need to start Selenium Server with -Dhttp.proxyHost; see the Proxy Conguration for more details. You may also try conguring your proxy manually and then launching the browser with *custom, or with *iehta browser launcher. *custom: When using *custom you must congure the proxy correctly(manually), otherwise youll get a 404 error. Double-check that youve congured your proxy settings correctly. To check whether youve congured the proxy correctly is to attempt to intentionally congure the browser incorrectly. Try conguring the browser to use the wrong proxy server hostname, or the wrong port. If you had successfully congured the browsers proxy settings incorrectly, then the browser will be unable to connect to the Internet, which is one way to make sure that one is adjusting the relevant settings. For other browsers (*refox, *opera) we automatically hard-code the proxy for you, and so ther are no known issues with this functionality. If youre encountering 404 errors and have followed this user guide carefully post your results to user group for some help from the user community.

6.14.8 Permission Denied Error

The most common reason for this error is that your session is attempting to violate the same-origin policy by crossing domain boundaries (e.g., accesses a page from http://domain1 and then accesses a page from http://domain2) or switching protocols (moving from http://domainX to https://domainX). This error can also occur when JavaScript attempts to nd UI objects which are not yet available (before the page has completely loaded), or are no longer available (after the page has started to be unloaded). This is most typically encountered with AJAX pages which are working with sections of a page or subframes that load and/or reload independently of the larger page. This error can be intermittent. Often it is impossible to reproduce the problem with a debugger because the trouble stems from race conditions which are not reproducible when the debuggers overhead is added to the system. Permission issues are covered in some detail in the tutorial. Read the section about the The Same Origin Policy, Proxy Injection carefully.

6.14.9 Handling Browser Popup Windows

There are several kinds of Popups that you can get during a Selenium test. You may not be able to close these popups by running selenium commands if they are initiated by the browser and not your AUT. You may need to know how to manage these. Each type of popup needs to be addressed differently. HTTP basic authentication dialogs: These dialogs prompt for a username/password to login to the site. To login to a site that requires HTTP basic authentication, use a username and password in the URL, as described in RFC 1738, like this: open(http://myusername:[email protected]/blah/blah/blah). SSL certicate warnings: Selenium RC automatically attempts to spoof SSL certicates when it is enabled as a proxy; see more on this in the section on HTTPS. If your browser is congured correctly, you should never see SSL certicate warnings, but you may need to congure your 6.14. Troubleshooting Common Problems 83

Selenium Documentation, Release 1.0

browser to trust our dangerous CyberVillains SSL certicate authority. Again, refer to the HTTPS section for how to do this. modal JavaScript alert/conrmation/prompt dialogs: Selenium tries to conceal those dialogs from you (by replacing window.alert, window.conrm and window.prompt) so they wont stop the execution of your page. If youre seeing an alert pop-up, its probably because it red during the page load process, which is usually too early for us to protect the page. Selenese contains commands for asserting or verifying alert and conrmation popups. See the sections on these topics in Chapter 4.

6.14.10 On Linux, why isnt my Firefox browser session closing?

On Unix/Linux you must invoke refox-bin directly, so make sure that executable is on the path. If executing Firefox through a shell script, when it comes time to kill the browser Selenium RC will kill the shell script, leaving the browser running. You can specify the path to refox-bin directly, like this.

cmd=getNewBrowserSession&1=*firefox /usr/local/firefox/firefox-bin&2=http://www.google.c

6.14.11 Firefox *chrome doesnt work with custom prole

Check Firefox prole folder -> prefs.js -> user_pref(browser.startup.page, 0); Comment this line like this: //user_pref(browser.startup.page, 0); and try again.

6.14.12 Is it ok to load a custom pop-up as the parent page is loading (i.e., before the parent pages javascript window.onload() function runs)?
No. Selenium relies on interceptors to determine window names as they are being loaded. These interceptors work best in catching new windows if the windows are loaded AFTER the onload() function. Selenium may not recognize windows loaded before the onload function.

6.14.13 Problems With Verify Commands

If you export your tests from Selenium-IDE, you may nd yourself getting empty verify strings from your tests (depending on the programming language used). Note: This section is not yet developed.

6.14.14 Safari and MultiWindow Mode

Note: This section is not yet developed.

6.14.15 Firefox on Linux

On Unix/Linux, versions of Selenium before 1.0 needed to invoke refox-bin directly, so if you are using a previous version, make sure that the real executable is on the path. On most Linux distributions, the real refox-bin is located on: 84 Chapter 6. Selenium-RC

Selenium Documentation, Release 1.0

/usr/lib/firefox-x.x.x/

Where the x.x.x is the version number you currently have. So, to add that path to the users path. you will have to add the following to your .bashrc le:
export PATH= "$PATH:/usr/lib/firefox-x.x.x/"

If necessary, you can specify the path to refox-bin directly in your test, like this:
" *firefox /usr/lib/firefox-x.x.x/firefox-bin "

6.14.16 IE and Style Attributes

If you are running your tests on Internet Explorer and you cannot locate elements using their style attribute. For example:
//td[@style="background-color:yellow"]

This would work perfectly in Firefox, Opera or Safari but not with IE. IE interprets the keys in @style as uppercase. So, even if the source code is in lowercase, you should use:
//td[@style="BACKGROUND-COLOR:yellow"]

This is a problem if your test is intended to work on multiple browsers, but you can easily code your test to detect the situation and try the alternative locator that only works in IE.

6.14.17 Where can I Ask Questions that Arent Answered Here?

Try our user group

6.14. Troubleshooting Common Problems

85

Selenium Documentation, Release 1.0

86

Chapter 6. Selenium-RC

CHAPTER

SEVEN

TEST DESIGN CONSIDERATIONS

NOTE: Some sections of this chapter are not yet complete.

7.1 Introducing Test Design

In this subsection we describe a few types of different tests you can do with Selenium. This may not be new to you, but we provide this as a framework for relating Selenium test automation to the decisions a quality assurance professional will make when deciding what tests to perform, the priority for each of those tests, and whether to automate those tests or not.

7.2 What to Test?

What parts of your application should you test? That depends on aspects of your project: user expectations, time allowed for the project, priorities set by the project manager and so on. Once the project boundaries are dened though, you, the tester, will certainly make many decisions on what to test. We will dene some terms here to help us categorize the types of testing typical for a web-application. These terms are by no means standard, although the concepts we present here are typical for webapplication testing. Weve created a few terms here of our own for the purposes of categorizing the types of test you may perform on your web application.

7.2.1 Testing for Expected Content

The simplest type of test, a content test, is a simple test for the existence of a static, non-changing, UI element. For instance Does each page have its expected page title? This can be used to verify your test found an expected page after following a link. Does the applications home page contain an image expected to be at the top of the page? Does each page of the website contain a footer area with links to the company contact page, privacy policy, and trademarks information? Does each page begin with heading text using the <h1> tag? And, does each page have the correct text within that header? You may or may not need content tests. If your page content is not likely to be affected then it may be more efcient to test page content manually. If, however, your application will be undergoing platform changes, or les will likely be moved to different locations, content tests may prove valuable. 87

Selenium Documentation, Release 1.0

7.2.2 Testing Links

A frequent source of errors for web-sites is broken links and missing pages behind those broken links. Testing for these involves clicking each link and verifying the expected page behind that link loads correctly. Need to include a description of how to design this test and a simple example. Should that go in this section or in a separate section?

7.2.3 Function Tests

These would be tests of a specic function within your application, requiring some type of user input, and returning some type of results. Often a function test will involve multiple pages with a formbased input page containing a collection of input elds, Submit and Cancel operations, and one or more response pages. User input can be via text-input elds, checkboxes, drop-down lists, or any other browser-supported input.

7.2.4 Testing Dynamic Elements

Every web page element has an Identier, a name that can be use to locate the element. These names can be static - for example, when you enter a search term into Google, the text eld in which you type your query is always named q. These names can also be dynamic - generated fresh, new, and unique for every instance of a web page. For example, some web servers might name a text entry eld q1 on the rst instance of a page, and q2 on the next instance of the same page, etc. This means your test script could be looking at a different eld name every time you run the script. Dynamic names are usually on a result page of some given function. An example would be a result set of data returned to the user. Suppose each data result, in, say for example a list of documents, had a unique Identier for each specic document. So, for a particular search, the search results page returns a data set with one set of documents and their correponding Identiers. Then, in a different search, the search results page returns a different data set where each document in the result set uses different Identiers. An example will help. Dynamic content involve UI elements who Identifying properties change each time you open the page displaying them. For example, Dynamic HTML of an object might look as:
<input type= "checkbox" value= "true" id= "addForm:_ID74:_ID75:0:_ID79:0: checkBox" name= "addForm:_ID74:_ID75:0:_ID79:0:checkBox" />

This is HTML snippet for a check box. Its ID and name (addForm:_ID74:_ID75:0:_ID79:0:checkBox) are same and both are dynamic (they will change the next time you open the application). In this case

7.2.5 Ajax Tests

Ajax is a technology which supports dynamic real-time UI elements such as animation and RSS feeds. In Ajax-driven web applications, data is retrieved from the application server without refreshing the page.

88

Chapter 7. Test Design ConsIDerations

Selenium Documentation, Release 1.0

7.3 Verifying Results

7.3.1 Assert vs. Verify
When should you use an assert command and when should you use a verify command? This is up to you. The difference is in what you want to happen when the check fails. Do you want your test to terminate or continue and record that the check failed? Heres the tradeoff. If you use an assert, the test will stop at that point and not run any subsequent checks. Sometimes, perhaps often, that is what you want. If the test fails you will immediately know the test did not pass. Test engines such as TestNG and JUnit have plugins for commonly used development environments (Chap 5) which conveniently ag these tests as failed tests. The advantage: you have an immediate visual of whether the checks (those using asserts anyway) passed. The disadvantage: when a check does fail, there are other checks which were never performed, so you have no information on their status. In contrast, verify commands will not terminate the test. If your test uses only verify commands you are guaranteed (assuming no unexpected exceptions) the test will run to completion whether the checks nd defects in the AUT or not. The disavantage: you have to do more work to examine your test results. That is, you wont get feedback from TestNG or JUnit. Rather, you will need to look at the results of a console printout or a log output by your test application. And you will need to take the time to look through this output everytime you run your test. For Java, Logging Selenium (Chap 5) is a convenient logging utility for recording the results of verify commands, however you still need to open the logs and examine the results. If you are running hundreds of tests, each with its own log, this will be time-consuming.

7.3.2 Trade-offs: assertTextPresent, assertElementPresent, assertText

You should now be familiar with these commands, and the mechanics of using them. If not, please refer to Chapter 4 rst. When constructing your tests, you will need to decide Do I only check that the text exists on the page? (verify/assertTextPresent) Do I only check that the HTML element exists on the page? That is, the text, image, or other content is not to be checked, only the HTML tag is what is relevant. (verify/assertElementPresent) Must I test both, the element and its text content? (verify/assertText) There is no right answer. It depends on the requirements for your test. Which, of course, depend on the requirements for the application youre testing. If in doubt, and if the requirements are not clear, you can go with your best guess and can always change the test later. Most of these are easily changed in either Sel-IDE or Sel-RC. Realize that verify/assertText is the most specic test. This can fail if either the HTML element (tag) OR the text is not what your test is expecting. Sometimes, for instance if HTML changes frequently by your programmers, verifyTextPresent makes more sense. It can check for the content, but will pass the test when the programmers change the HTML used to present that text. Alternatively, perhaps your webdesigners are frequently changing the page and you dont want your test to fail everytime they do this because the changes themeselves are expected periodically. However, assume you still need to check that something is on the page, say a paragraph, or heading text, or an image. In this case you can use verify/assertElementPresent. It will ensure that a particular type of element exists (and if using Xpath can ensure it exists relative to other objects within the page). But you dont care what the content is, that is, a specic image le, or specic text. You only care that some type of image exists.

7.3. Verifying Results

89

Selenium Documentation, Release 1.0

Getting a feel for these types of decisions will come with time and a little experience. They are easy concepts, and easy to change in your test, but they do depend do depend on the requirements of your AUT. For some projects the requirements are clear and therefore your tests will be clear. For others, not so much, and you will have to give it your best guess. The purpose of this subsection is to help you anticipate your needs so you can make these decisions more efciently.

7.4 Choosing a Location Strategy

You know from the Selenese section there are multiple ways of selecting an object on a page. But what are the tradeoffs of each of these locator types? Recall we can locate an object using the element ID the element name attribute an Xpath statement document object model (DOM) Generally, using an ID locator is more efcient as it makes your test code more readable, assuming the ID used by the AUTs page source is a meaningful one. Using the name attribute also has similar advantages. These also give the best performance. Xpath statements have been known to be slow in Internet Explorer due to limations of IEs Xpath processor. Sometimes though, you must use an Xpath locator. If the page source does not have an ID or name attribute you have no choice but to use an Xpath or DOM locator. It appears at the time of writing that DOM locators are not commonly used now, and Xpath appears to the preferred choice, possibly because Xpath provides a rich set of possibilities for Identifying an objectit is quite exible. There is an advantage to using Xpath or DOM that locating via ID or name attributes do not have. With Xpath and DOM you can locate an object with respect to another object on the page. For example, if there is a link that must occur within the second paragragh within a <div> section, you can use Xpath or DOM to specify this. With ID and name locators, you can only specify that they occur on the page somewhere on the page. If you must test that an image displaying the company logo appears at the top of the page within a header section Xpath may be the better locator.

7.4.1 Locating Dynamic Objects

First, you must understand what a dynamic object is, and to do so, we will contrast that with a static object. Until now, all the AUT page elements we have been considering have been static objects. These are objects whos html page source is the same each time the page is loaded in the browser. For example,
<a class= "button" id= "adminHomeForm" onclick= "return oamSubmitForm(adminHomeForm, adminHomeForm:_ID38);" href= "#" >View Archived Allocation Events</a>

This is HTML anchor tag dening an button with an ID attribute of adminHomeForm. Its a fairly complex anchor tag when compared to most HTML tags, but it is still a static tag. The HTML will be the same each time this page is loaded in the browser. Its ID remains constant within all instances of this page. That is, when this page is displayed, this UI element will always have this Identier. So, for your test script to click this button you simply need to use the following selenium command. 90 Chapter 7. Test Design ConsIDerations

Selenium Documentation, Release 1.0

click

adminHomeForm

Or, in Selenium-RC
selenium.click( "adminHomeForm" );

Your application, however, may generate HTML with IDs that are generated dynamically and therefore the ID itself varies on different instances of the webpage under test. For instance, HTML for a dynamic page element might look like this.
<input type= "checkbox" value= "true" id= "addForm:_ID74:_ID75:0:_ID79:0:checkBox" name= "addForm:_ID74:_ID75:0:_ID79:0:checkBox" />

This denes a checkbox. Its ID and name attributes (both addForm:_ID74:_ID75:0:_ID79:0:checkBox) are dynamically generated values. In this case, using a standard locator would look something like the following.
click addForm:_ID74:_ID75:0:_ID79:0:checkBox

Or, again in Selelenium-RC

selenium.click("addForm:_ID74:_ID75:0:_ID79:0:checkBox);

Given the dynamically generated Identier, this approach would not work. The next time this page is loaded the Identier will be a different value from the one used in the Selenium command and therefore, will not be found. The click operation will fail with an element not found error. To correct this, a simple solution would be to just use an XPath locator rather than trying to use an ID locator. So, for the checkbox you can simply use
click //input

Or, if it is not the rst input element on the page (which it likely is not) try a more detailed Xpath statement.
click //input[3]

Or
click //div/p[2]/input[3]

If however, you do need to use the ID to locate the element, a programmed solution is required. Another solution is to capture this ID from the website itself, before you need to use it in a Selenium command. It can be done like this.
String[] checkboxids = selenium.getAllFields(); // Collect all input IDs on page. if(!GenericValidator.IsBlankOrNull(checkboxids[i])) // If collected ID is not null. { // If the ID starts with addForm if(checkboxids[i].indexOf( "addForm" ) > -1) { selenium.click(checkboxids[i]);

7.4. Choosing a Location Strategy

91

Selenium Documentation, Release 1.0

} }

This approach will work if there is only one eld whose ID has the text addForm appended to it. Consider one more example of a Dynamic object; a page with two links having the same name and the same html name. Now if href is used to click the link, it would always be clicking on the rst element. Clicking on the second link can be achieved as follows.
// Flag for second appearance of link. boolean isSecondInstanceLink = false; // Desired link. String editInfo = null; // Collect all links. String[] links = selenium.getAllLinks(); // Loop through collected links. for(String linkID: links) { // If retrieved link is not null if(!GenericValidator.isBlankOrNull(linkID))

// Find the inner HTML of link. String editTermSectionInfo = selenium.getEval ( "window.document.getElementByID(" +linkID+ ").innerHTML" ); // If retrieved link is expected link. if(editTermSectionInfo.equalsIgnoreCase( "expectedlink" )) { // If it is second appearance of link then save the link ID and break the loop. if(isSecondInstanceLink) { editInfo = linkID; break; } // Set the second appearance of Autumn term link to true as isSecondInstanceLink = true; } } } // Click on link. selenium.click(editInfo);

7.4.2 How can I avoid using complex xpath expressions in my test?

Where possible, element IDs should be explicitly created by the application designer, not automatically generated. Automatically generated (non-descriptive) element IDs (i.e. id_147) tend to cause two problems: rst, each time the application is deployed, different element IDs could be generated. If the element ID changes then your test will fail. Second, a non-descriptive element id makes it hard for automation testers to keep track of and determine which element IDs to use (which ID is associated with which GUI object). 92 Chapter 7. Test Design ConsIDerations

Selenium Documentation, Release 1.0

You might consider trying the UI-Element extension in this situation. http://wiki.openqa.org/display/SIDE/Contributed+Extensions+and+Formats#ContributedExtensionsandFormatsUIElementLocator

7.4.3 Performance Considerations for Locators 7.4.4 Custom Locators

This section is not yet developed.

7.5 Testing Ajax Applications

We introduced the special characteristics of AJAX technology earlier in this chapter. Basically, a page element implemented with Ajax is an element that can be dynamically refreshed without having to refresh the entire page.

7.5.1 Waiting for an Ajax Element

For an Ajax element using Seleniums waitForPageToLoad wouldnt work since the page is not actually loaded to refresh the Ajax element. Pausing the test execution for a specied period of time is also not good because the web element might appear later than expected leading to invalid test failures (reported failures that arent actually failures). A better approach is to wait for a predened period and then continue execution as soon as the element is found. Consider a page which brings a link (link=ajaxLink) on click of a button on page (without refreshing the page) This could be handled by Selenium using a for loop.
// Loop initialization. for (int second = 0;; second++) { // If loop is reached 60 seconds then break the loop. if (second >= 60) break;

// Search for element "link=ajaxLink" and if available then break loop. try { if (selenium.isElementPresent( "link=ajaxLink" )) break; } catch (Exception e) // Pause for 1 second. Thread.sleep(1000); }

This certainly isnt the only solution. Ajax is a common topic in the user group and we suggest searching previous discussions to see what others have done along with the questions they have posted.

7.6 UI Mapping
A UI map is a mechanism that stores Identiers, or in our case, locators, for an applications UI elements. The test script then uses the UI Map for locating the elements to be tested. Basically, a UI map is a repository of test script objects that correspond to UI elements of the application being tested. 7.5. Testing Ajax Applications 93

Selenium Documentation, Release 1.0

What makes a UI map heplful? Its primary purpose for making test script management much easier. When a locator needs to be edited, there is a central location for easily nding that object, rather than having to search through test script code. Also, it allows changing the Identifer in a single place, rather than having to make the change in multiple places within a test script, or for that matter, in multiple test scripts. To summarize, a UI map has two signicant advantages. Using a centralized location for UI objects instead of having them scattered throughout the script. This makes script maintanence more efcient. Cryptic HTML Identiers and names can be given more human-readable names improving the readability of test scripts. Consider the following example (in java) of selenium tests for a website:
public void testNew() throws Exception { selenium.open( "http://www.test.com" ); selenium.type( "loginForm:tbUsername" , "xxxxxxxx" ); selenium.click( "loginForm:btnLogin" ); selenium.click( "adminHomeForm:_activitynew" ); selenium.waitForPageToLoad( "30000" ); selenium.click( "addEditEventForm:_IDcancel" ); selenium.waitForPageToLoad( "30000" ); selenium.click( "adminHomeForm:_activityold" ); selenium.waitForPageToLoad( "30000" ); }

This script is incomprehensible to anyone who isnt very familiar with the AUTs page source. Even regular users of the application would have difculty understanding what the script does. A better script would be:
public void testNew() throws Exception { selenium.open( "http://www.test.com" ); selenium.type(admin.username, "xxxxxxxx" ); selenium.click(admin.loginbutton); selenium.click(admin.events.createnewevent); selenium.waitForPageToLoad( "30000" ); selenium.click(admin.events.cancel); selenium.waitForPageToLoad( "30000" ); selenium.click(admin.events.viewoldevents); selenium.waitForPageToLoad( "30000" ); }

There are no comments provided but it is more comprehensible because of the keywords used in scripts. (please be aware that UI Map is NOT a replacement for comments! Comments are still important for documenting an automated test.) An even better test script could look like this.
public void testNew() throws Exception { // Open app url. selenium.open( "http://www.test.com" );

94

Chapter 7. Test Design ConsIDerations

Selenium Documentation, Release 1.0

// Provide admin username. selenium.type(admin.username, "xxxxxxxx" ); // Click on Login button. selenium.click(admin.loginbutton); // Click on Create New Event button. selenium.click(admin.events.createnewevent); selenium.waitForPageToLoad( "30000" ); // Click on Cancel button. selenium.click(admin.events.cancel); selenium.waitForPageToLoad( "30000" ); // Click on View Old Events button. selenium.click(admin.events.viewoldevents); selenium.waitForPageToLoad( "30000" ); }

The idea is to have a centralized location for objects and using comprehensible names for those objects. To achieve this, properties les can be used in java. A properties le contains key/value pairs, where each key and value are strings. Consider a property le prop.properties which assigns as aliases easily understood Identiers for the HTML objects used earlier.
admin.username = loginForm:tbUsername admin.loginbutton = loginForm:btnLogin admin.events.createnewevent = adminHomeForm:_activitynew admin.events.cancel = addEditEventForm:_IDcancel admin.events.viewoldevents = adminHomeForm:_activityold

The locators will still refer to html objects, but we have introduced a layer of abstraction between the test script and the UI elements. Values are read from the properties le and used in the Test Class to implement the UI Map. For more on Properties les refer to the following link.

7.7 Bitmap Comparison

This section has not been developed yet.

7.7.1 Data Driven Testing

This section needs an introduction and it has not been completed yet. In Python:
# Collection of String values source = open( " input_file.txt " , " r " ) values = source.readlines() source.close() # Execute For loop for each String in the values array for search in values: sel.open( " / " )

7.7. Bitmap Comparison

95

Selenium Documentation, Release 1.0

sel.type( " q " , search) sel.click( " btnG " ) sel.waitForPageToLoad( " 30000 " ) self.failUnless(sel.is_text_present( " Results * for

" + search))

Why would we want a separate le with data in it for our tests? One important method of testing is to run the same test repetitively with different data values. This is called Data Driven Testing and is a very common testing task. Test automation tools, Selenium included, generally handle this as its a common reason for test automation. The Python script above opens a text le. This le contains a different search string on each line. The code then saves this in an array of strings, and iterates over the array doing a search and assert on each string. This is a very basic example of what you can do, but the Idea is to show you things that can easily be done with either a programming or scripting language when theyre difcult or even impossible to do using Selenium-IDE. Refer to the Selnium RC wiki for examples of reading data from spread sheet or using data provider capabilities of TestNG with java client driver.

7.8 Handling Errors

Note: This section is not yet developed.

7.8.1 Error Reporting 7.8.2 Recovering From Failure

A quick note thoughrecognize that your programming languages exception- handling support can be used for error handling and recovery. This section has not been developed yet.

7.8.3 Database Validations

Since you can also do database queries from your favorite programming language, assuming you have database support functions, why not use them for some data validations/retrieval on the Application Under Test? Consider the example of a registration process where a registered email address is to be retrieved from the database. An example of establishing a DB connection and retrieving data from the DB would be: In Java:
// Load Microsoft SQL Server JDBC driver. Class.forName( "com.microsoft.sqlserver.jdbc.SQLServerDriver" ); // Prepare connection url. String url = "jdbc:sqlserver://192.168.1.180:1433;DatabaseName=TEST_DB" ; // Get connection to DB.

96

Chapter 7. Test Design ConsIDerations

Selenium Documentation, Release 1.0

public static Connection con = DriverManager.getConnection(url, "username" , "password" ); // Create statement object which would be used in writing DDL and DML // SQL statement. public static Statement stmt = con.createStatement(); // Send SQL SELECT statements to the database via the Statement.executeQuery // method which returns the requested information as rows of data in a // ResultSet object. ResultSet result = stmt.executeQuery ( "select top 1 email_address from user_register_table" ); // Fetch value of "email_address" from "result" object. String emailaddress = result.getString( "email_address" ); // Use the fetched value to login to application. selenium.type( "userID" , emailaddress);

This is a very simple example of data retrieval from a DB in Java. A more complex test could be to validate that inactive users are not able to login to the application. This wouldnt take too much work from what youve already seen.

7.8. Handling Errors

97

Selenium Documentation, Release 1.0

98

Chapter 7. Test Design ConsIDerations

CHAPTER

EIGHT

SELENIUM-GRID
Please refer to the Selenium Grid website http://selenium-grid.seleniumhq.org/how_it_works.html This section is not yet developed. If there is a member of the community who is experienced in SeleniumGrid, and would like to contribute, please contact the Documentation Team. We would love to have you contribute.

99

Selenium Documentation, Release 1.0

100

Chapter 8. Selenium-Grid

CHAPTER

NINE

USER-EXTENSIONS
NOTE: This section is close to completion, but it has not been reviewed and edited.

9.1 Introduction
It can be quite simple to extend Selenium, adding your own actions, assertions and locator-strategies. This is done with JavaScript by adding methods to the Selenium object prototype, and the PageBot object prototype. On startup, Selenium will automatically look through methods on these prototypes, using name patterns to recognize which ones are actions, assertions and locators. The following examples give an indication of how Selenium can be extended with JavaScript.

9.2 Actions
All methods on the Selenium prototype beginning with do are added as actions. For each action foo there is also an action fooAndWait registered. An action method can take up to two parameters, which will be passed the second and third column values in the test. Example: Add a typeRepeated action to Selenium, which types the text twice into a text box.
Selenium.prototype.doTypeRepeated = function(locator, text) { // All locator-strategies are automatically handled by "findElement" var element = this.page().findElement(locator); // Create the text to type var valueToType = text + text; // Replace the element text with the new text this.page().replaceText(element, valueToType); };

9.3 Accessors/Assertions
All getFoo and isFoo methods on the Selenium prototype are added as accessors (storeFoo). For each accessor there is an assertFoo, verifyFooa nd waitForFoo registered. An assert method can take up to 2 parameters, which will be passed the second and third column values in the test. You can also dene your own assertions literally as simple assert methods, which will also auto-generate verify and waitFor commands. Example: Add a valueRepeated assertion, that makes sure that the element

101

Selenium Documentation, Release 1.0

value consists of the supplied text repeated. The 2 commands that would be available in tests would be assertValueRepeated and verifyValueRepeated.
Selenium.prototype.assertValueRepeated = function(locator, text) { // All locator-strategies are automatically handled by "findElement" var element = this.page().findElement(locator); // Create the text to verify var expectedValue = text + text; // Get the actual element value var actualValue = element.value; // Make sure the actual value matches the expected Assert.matches(expectedValue, actualValue); };

9.3.1 Automatic availability of storeFoo, assertFoo, assertNotFoo, waitForFoo and waitForNotFoo for every getFoo
All getFoo and isFoo methods on the Selenium prototype automatically result in the availability of storeFoo, assertFoo, assertNotFoo, verifyFoo, verifyNotFoo, waitForFoo, and waitForNotFoo commands. Example, if you add a getTextLength() method, the following commands will automatically be available: storeTextLength, assertTextLength, assertNotTextLength, verifyTextLength, verifyNotTextLength, waitForTextLength, and waitForNotTextLength commands.
Selenium.prototype.getTextLength = function(locator, text) { return this.getText(locator).length; };

Also note that the assertValueRepeated method described above could have been implemented using isValueRepeated, with the added benet of also automatically getting assertNotValueRepeated, storeValueRepeated, waitForValueRepeated and waitForNotValueRepeated.

9.4 Locator Strategies

All locateElementByFoo methods on the PageBot prototype are added as locator-strategies. A locator strategy takes 2 parameters, the rst being the locator string (minus the prex), and the second being the document in which to search. Example: Add a valuerepeated= locator, that nds the rst element a value attribute equal to the the supplied value repeated.
// The "inDocument" is a the document you are searching. PageBot.prototype.locateElementByValueRepeated = function(text, inDocument) { // Create the text to search for var expectedValue = text + text; // Loop through all elements, looking for ones that have // a value === our expected value var allElements = inDocument.getElementsByTagName( "*" ); for (var i = 0; i < allElements.length; i++) { var testElement = allElements[i];

102

Chapter 9. User-Extensions

Selenium Documentation, Release 1.0

if (testElement.value && testElement.value === expectedValue) { return testElement; } } return null; };

9.5 Using User-Extensions With Selenium-IDE

User-extensions are very easy to use with the selenium IDE. 1. Create your user extension and save it as user-extensions.js. While this name isnt technically necessary, its good practice to keep things consistent. 2. Open Firefox and open Selenium-IDE. 3. Click on Tools, Options 4. In Selenium Core Extensions click on Browse and nd the user-extensions. js le. Click on OK. 5. Your user-extension will not yet be loaded, you must close and restart Selenium-IDE. 6. In your empty test, create a new command, your user-extension should now be an options in the Commands dropdown.

9.6 Using User-Extensions With Selenium RC

If you Google Selenium RC user-extension ten times you will nd ten different approaches to using this feature. Below, is the ofcial Selenium suggested approach.

9.6.1 Example
C# 1. Place your user extension in the same directory as your Selenium Server. 2. If you are using client code generated by the Selenium-IDE you will need to make a couple small edits. First, you will need to create an HttpCommandProcessor object with class scope (outside the SetupTest method, just below private StringBuilder verificationErrors;)
HttpCommandProcessor proc;

1. Next, instantiate that HttpCommandProcessor object DefaultSelenium object. This can be done in the test setup.

as

you

would

the

proc = new HttpCommandProcessor( "localhost" , 4444, "*iexplore" , "http://google.ca/" );

9.5. Using User-Extensions With Selenium-IDE

103

Selenium Documentation, Release 1.0

1. Instantiate the DefaultSelenium object using the HttpCommandProcessor object you created.
selenium = new DefaultSelenium(proc);

1. Within your test code, execute your user-extension by calling it with the DoCommand() method of HttpCommandProcessor. This method takes two arguments: a string to identify the userextension method you want to use and string array to pass arguments. Notice that the rst letter of your function is lower case, regardless of the capitalization in your user-extension. Selenium automatically does this to keep common JavaScript naming conventions. Because JavaScript is case sensitive, your test will fail if you begin this command with a capital. inputParams is the array of arguments you want to pass to the JavaScript user-extension. In this case there is only one string in the array because there is only one parameter for our user extension, but a longer array will map each index to the corresponding user-extension parameter. Remember that user extensions designed for Selenium-IDE will only take two arguments.
string[] inputParams = { "Hello World" }; proc.DoCommand( "alertWrapper" , inputParams);

1. Start the test server using the -userExtensions argument and pass in your user-extensinos.js le.
java -jar selenium-server.jar -userExtensions user-extensions.js

using using using using using using

System; System.Text; System.Text.RegularExpressions; System.Threading; NUnit.Framework; Selenium;

namespace SeleniumTests { [TestFixture] public class NewTest { private ISelenium selenium; private StringBuilder verificationErrors; private HttpCommandProcessor proc;

[SetUp] public void SetupTest() { proc = new HttpCommandProcessor( "localhost" , 4444, "*iexplore" selenium = new DefaultSelenium(proc); //selenium = new DefaultSelenium("localhost", 4444, "*iexplore", selenium.Start(); verificationErrors = new StringBuilder(); }

104

Chapter 9. User-Extensions

Selenium Documentation, Release 1.0

[TearDown] public void TeardownTest() { try { selenium.Stop(); } catch (Exception) { // Ignore errors if unable to close the browser } Assert.AreEqual( "" , verificationErrors.ToString()); } [Test] public void TheNewTest() { selenium.Open( "/" ); string[] inputParams = { "Hello World" ,}; proc.DoCommand( "alertWrapper" , inputParams); } } }

9.6. Using User-Extensions With Selenium RC

105

Selenium Documentation, Release 1.0

106

Chapter 9. User-Extensions

CHAPTER

TEN

SELENIUM 2.0 AND WEBDRIVER

The Selenium developers are working towards a Selenium 2.0 release. The primary new feature will be the integration of the WebDriver API into Selenium-RC. This will address a number of Selenium 1.0 limitations along with providing an alternative programming interface. The goal is to develop a standardized Selenium API that provides additional support for a larger number of browsers along with improved support for advanced web-app testing problems. The best explanation for why WebDriver and Selenium are merging was detailed by Simon Stewart, the creator of WebDriver, in a joint email to the WebDriver and Selenium community on August 6, 2009. Why are the projects merging? Partly because webdriver addresses some shortcomings in selenium (by being able to bypass the JS sandbox, for example. And weve got a gorgeous API), partly because selenium addresses some shortcomings in webdriver (such as supporting a broader range of browsers) and partly because the main selenium contributors and I felt that it was the best way to offer users the best possible framework. PLEASE NOTE: Selenium 2.0 is currently under development. The WebDriver integration is currently undergoing testing. We encourage advanced Selenium users to try it out. Users new to Selenium, should wait until Selenium 2.0 is ofcially released.

10.1 What is WebDriver?

WebDriver uses a different underlying framework from Seleniums javascript Selenium-Core. It also provides an alternative API with functionality not supported in Selenium-RC. WebDriver does not depend on a javascript core embedded within the browser, therefore it is able to avoid some long-running Selenium limitations. WebDrivers goal is to provide an API that establishes A well-designed standard programming interface for web-app testing. Improved consistency between browsers. Additional functionality addressing testing problems not well-supported in Selenium 1.0. The Selenium developers strive to continuously improve Selenium. Integrating WebDriver is another step in that process. The developers of Selenium and of WebDriver felt they could make signicant gains for the Open Source test automation community be combining forces and merging their ideas and technologies. Integrating WebDriver into Selenium is the current result of those efforts.

107

Selenium Documentation, Release 1.0

10.2 When to Use WebDriver?

One should use WebDriver when requiring improved support for Mult-browser testing including improved functionality for browsers not well-supported by Selenium-1.0. Handling multiple frames, multiple browser windows, popups, and alerts. Page navigation. Drag-and-drop. AJAX-based UI elements.

10.3 The 5 Minute Getting Started Guide

WebDriver is a tool for automating testing web applications, and in particular to verify that they work as expected. It aims to provide a friendly API thats easy to explore and understand, which will help make your tests easier to read and maintain. Its not tied to any particular test framework, so it can be used equally well with JUnit, TestNG or from a plain old main method. This Getting Started guide introduces you to WebDrivers Java API and helps get you started becoming familiar with it. Start by Downloading the latest binaries and unpack them into a directory. From now on, well refer to that as $WEBDRIVER_HOME. Now, open your favourite IDE and: Start a new Java project in your favourite IDE Add all the JAR les under $WEBDRIVER_HOME to the CLASSPATH You can see that WebDriver acts just as a normal Java library does: its entirely self-contained, and you dont need to remember to start any additional processes or run any installers before using it. Youre now ready to write some code. An easy way to get started is this example, which searches for the term Cheese on Google and then outputs the result pages title to the console. Youll start by using the HtmlUnit Driver. This is a pure Java driver that runs entirely in-memory. Because of this, you wont see a new browser window open.
package org.openqa.selenium.example; import import import import org.openqa.selenium.By; org.openqa.selenium.WebDriver; org.openqa.selenium.WebElement; org.openqa.selenium.htmlunit.HtmlUnitDriver;

public class Example { public static void main(String[] args) { // Create a new instance of the html unit driver // Notice that the remainder of the code relies on the interface, // not the implementation. WebDriver driver = new HtmlUnitDriver(); // And now use this to visit Google driver.get( "http://www.google.com" );

108

Chapter 10. Selenium 2.0 and WebDriver

Selenium Documentation, Release 1.0

// Find the text input element by its name WebElement element = driver.findElement(By.name( "q" )); // Enter something to search for element.sendKeys( "Cheese!" ); // Now submit the form. WebDriver will find the form for us from the element element.submit(); // Check the title of the page System.out.println( "Page title is: " + driver.getTitle()); } }

Compile and run this. You should see a line with the title of the Google search results as output on the console. Congratulations, youve managed to get started with WebDriver! In this next example, you shall use a page that requires Javascript to work properly, such as Google Suggest. You will also be using the Firefox Driver. Make sure that Firefox is installed on your machine and is in the normal location for your OS. Once thats done, create a new class called GoogleSuggest, which looks like:
package org.openqa.selenium.example; import java.util.List; import import import import import org.openqa.selenium.By; org.openqa.selenium.WebDriver; org.openqa.selenium.RenderedWebElement; org.openqa.selenium.WebElement; org.openqa.selenium.firefox.FirefoxDriver;

public class GoogleSuggest { public static void main(String[] args) throws Exception { // The Firefox driver supports javascript WebDriver driver = new FirefoxDriver(); // Go to the Google Suggest home page driver.get( "http://www.google.com/webhp?complete=1&hl=en" ); // Enter the query string "Cheese" WebElement query = driver.findElement(By.name( "q" )); query.sendKeys( "Cheese" );

// Sleep until the div we want is visible or 5 seconds is over long end = System.currentTimeMillis() + 5000; while (System.currentTimeMillis() < end) { // Browsers which render content (such as Firefox and IE) // return "RenderedWebElements" RenderedWebElement resultsDiv = (RenderedWebElement) driver.findElement(By.c // If results have been returned, // the results are displayed in a drop down. if (resultsDiv.isDisplayed()) { break;

10.3. The 5 Minute Getting Started Guide

109

Selenium Documentation, Release 1.0

} }

// And now list the suggestions List<WebElement> allSuggestions = driver.findElements(By.xpath( "//td[@class=gac for (WebElement suggestion : allSuggestions) { System.out.println(suggestion.getText()); } } }

When you run this program, youll see the list of suggestions being printed to the console. Thats all there is to using WebDriver! Hopefully, this will have whet your appetite for more. In the Next Steps section you will learn more about how to use WebDriver for things such as navigating forward and backward in your browsers history, and how to use frames and windows. It also provides a more complete discussion of the examples than The 5 Minute Getting Started Guide. If youre ready, lets take the Next Steps!

10.4 Next Steps For Using WebDriver

10.4.1 Which Implementation of WebDriver Should I Use?
WebDriver is the name of the key interface against which tests should be written, but there are several implementations. These are: Name of driver HtmlUnit Driver Firefox Driver Internet Explorer Driver Chrome Driver Available on which OS? All All Windows All Class to instantiate org.openqa.selenium.htmlunit.HtmlUnitDriver org.openqa.selenium.refox.FirefoxDriver org.openqa.selenium.ie.InternetExplorerDriver org.openqa.selenium.chrome.ChromeDriver

You can nd out more information about each of these by following the links in the table. Which you use depends on what you want to do. For sheer speed, the HtmlUnit Driver is great, but its not graphical, which means that you cant watch whats happening. As a developer you may be comfortable with this, but sometimes its good to be able to test using a real browser, especially when youre showing a demo of your application (or running the tests) for an audience. Often, this idea is referred to as safety, and it falls into two parts. Firstly, theres actual safety, which refers to whether or not the tests work as they should. This can be measured and quantied. Secondly, theres perceived safety, which refers to whether or not an observer believes the tests work as they should. This varies from person to person, and will depend on their familiarity with the application under test, WebDriver, and your testing framework. To support higher perceived safety, you may wish to choose a driver such as the Firefox Driver. This has the added advantage that this driver actually renders content to a screen, and so can be used to detect information such as the position of an element on a page, or the CSS properties that apply to it. However, this additional exibility comes at the cost of slower overall speed. By writing your tests against the WebDriver interface, it is possible to pick the most appropriate driver for a given test. To keep things simple, lets start with the HtmlUnit Driver:
WebDriver driver = new HtmlUnitDriver();

110

Chapter 10. Selenium 2.0 and WebDriver

Selenium Documentation, Release 1.0

10.4.2 Navigating
The rst thing youll want to do with WebDriver is navigate to a page. The normal way to do this is by calling get:
driver.get( "http://www.google.com" );

WebDriver will wait until the page has fully loaded (that is, the onload event has red) before returning control to your test or script. Its worth noting that if your page uses a lot of AJAX on load then WebDriver may not know when it has completely loaded. If you need to ensure such pages are fully loaded then you can use waits.

10.4.3 Interacting With the Page

Just being able to go to places isnt terribly useful. What wed really like to do is to interact with the pages, or, more specically, the HTML elements within a page. First of all, we need to nd one. WebDriver offers a number of ways of nding elements. For example, given an element dened as:
<input type= "text" name= "passwd" id= "passwd-id" />

you could nd it using any of:

WebElement element; element = driver.findElement(By.id( "passwd-id" )); element = driver.findElement(By.name( "passwd" )); element = driver.findElement(By.xpath( "//input[@id=passwd-id]" ));

You can also look for a link by its text, but be careful! The text must be an exact match! You should also be careful when using XPATH in WebDriver. If theres more than one element that matches the query, then only the rst will be returned. If nothing can be found, a NoSuchElementException will be thrown. WebDriver has an Object-based API; we represent all types of elements using the same interface: Web Element. This means that although you may see a lot of possible methods you could invoke when you hit your IDEs auto-complete key combination, not all of them will make sense or be valid. Dont worry! WebDriver will attempt to do the Right Thing, and if you call a method that makes no sense (setSelected() on a meta tag, for example) an exception will be thrown. So, youve got an element. What can you do with it? First of all, you may want to enter some text into a text eld:
element.sendKeys( "some text" );

You can simulate pressing the arrow keys by using the Keys class:
element.sendKeys( " and some" , Keys.ARROW_DOWN);

It is possible to call sendKeys on any element, which makes it possible to test keyboard shortcuts such as those used on GMail. A side-effect of this is that typing something into a text eld wont automatically clear it. Instead, what you type will be appended to whats already there. You can easily clear the contents of a text eld or textarea:

10.4. Next Steps For Using WebDriver

111

Selenium Documentation, Release 1.0

element.clear();

10.4.4 Filling In Forms

Weve already seen how to enter text into a textarea or text eld, but what about the other elements? You can toggle the state of checkboxes, and you can use setSelected to set something like an OPTION tag selected. Dealing with SELECT tags isnt too bad:
WebElement select = driver.findElement(By.xpath( "//select" )); List<WebElement> allOptions = select.findElements(By.tagName( "option" )); for (WebElement option : allOptions) { System.out.println(String.format( "Value is: %s" , option.getValue())); option.setSelected(); }

This will nd the rst SELECT element on the page, and cycle through each of its OPTIONs in turn, printing out their values, and selecting each in turn. As you can see, this isnt the most efcient way of dealing with SELECT elements. WebDrivers support classes include one called Select, which provides useful methods for interacting with these.
Select select = new Select(driver.findElement(By.xpath( "//select" ))); select.deselectAll(); select.selectByVisibleText( "Edam" );

This will deselect all OPTIONs from the rst SELECT on the page, and then select the OPTION with the displayed text of Edam. Once youve nished lling out the form, you probably want to submit it. One way to do this would be to nd the submit button and click it:
driver.findElement(By.id( "submit" )).click(); // Assume the button has the ID "submit" :)

Alternatively, WebDriver has the convenience method submit on every element. If you call this on an element within a form, WebDriver will walk up the DOM until it nds the enclosing form and then calls submit on that. If the element isnt in a form, then the NoSuchElementException will be thrown:
element.submit();

10.4.5 Getting Visual Information And Drag And Drop

Sometimes you want to extract some visual information out of an element, perhaps to see if its visible or where it is on screen. You can nd out this information by casting the element to a RenderedWebElement:
WebElement plain = driver.findElement(By.name( "q" )); RenderedWebElement element = (RenderedWebElement) element;

112

Chapter 10. Selenium 2.0 and WebDriver

Selenium Documentation, Release 1.0

Not all drivers render their content to the screen (such as the HtmlUnit Driver), so its not safe to assume that the cast will work, but if it does you can gather additional information such as the size and location of the element. In addition, you can use drag and drop, either moving an element by a certain amount, or on to another element:

RenderedWebElement element = (RenderedWebElement) driver.findElement(By.name( "source" )) RenderedWebElement target = (RenderedWebElement) driver.findElement(By.name( "target" )); element.dragAndDropOn(target);

10.4.6 Moving Between Windows and Frames

Its rare for a modern web application not to have any frames or to be constrained to a single window. WebDriver supports moving between named windows using the switchTo method:
driver.switchTo().window( "windowName" );

All calls to driver will now be interpreted as being directed to the particular window. But how do you know the windows name? Take a look at the javascript or link that opened it:
<a href= "somewhere.html" target= "windowName" >Click here to open a new window</a>

Alternatively, you can pass a window handle to the switchTo().window() method. Knowing this, its possible to iterate over every open window like so:
for (String handle : driver.getWindowHandles()) { driver.switchTo().window(handle); }

You can also swing from frame to frame (or into iframes):
driver.switchTo().frame( "frameName" );

Its possible to access subframes by separating the path with a dot, and you can specify the frame by its index too. That is:
driver.switchTo().frame( "frameName.0.child" );

would go to the frame named child of the rst subframe of the frame called frameName. All frames are evaluated as if from *top*.

10.4.7 Navigation: History and Location

Earlier, we covered navigating to a page using the get command ( driver.get("http://www.example.com")) As youve seen, WebDriver has a number of smaller, task-focused interfaces, and navigation is a useful task. Because loading a page is such a fundamental requirement, the method to do this lives on the main WebDriver interface, but its simply a synonym to:

10.4. Next Steps For Using WebDriver

113

Selenium Documentation, Release 1.0

driver.navigate().to( "http://www.example.com" );

To reiterate: navigate().to() and get() do exactly the same thing. Ones just a lot easier to type than the other! The navigate interface also exposes the ability to move backwards and forwards in your browsers history:
driver.navigate().forward(); driver.navigate().back();

Please be aware that this functionality depends entirely on the underlying browser. Its just possible that something unexpected may happen when you call these methods if youre used to the behaviour of one browser over another.

10.4.8 Cookies
Before we leave these next steps, you may be interested in understanding how to use cookies. First of all, you need to be on the domain that the cookie will be valid for:
// Go to the correct domain driver.get( "http://www.example.com" ); // Now set the cookie. This ones valid for the entire domain Cookie cookie = new Cookie( "key" , "value" ); driver.manage().addCookie(cookie);

// And now output all the available cookies for the current URL Set<Cookie> allCookies = driver.manage().getCookies(); for (Cookie loadedCookie : allCookies) { System.out.println(String.format( "%s -> %s" , loadedCookie.getName(), loadedCookie.g }

10.4.9 Next, Next Steps!

This has been a high level walkthrough of WebDriver and some of its key capabilities. You may want to look at the Test Design Considerations chapter to get some ideas about how you can reduce the pain of maintaining your tests and how to make your code more modular.

10.5 WebDriver Implementations

10.5.1 HtmlUnit Driver
This is currently the fastest and most lightweight implementation of WebDriver. As the name suggests, this is based on HtmlUnit. Pros Fastest implementation of WebDriver 114 Chapter 10. Selenium 2.0 and WebDriver

Selenium Documentation, Release 1.0

A pure Java solution and so it is platform independent. Supports JavaScript Cons Emulates other browsers JavaScript behaviour (see below) JavaScript in the HtmlUnit Driver None of the popular browsers uses the JavaScript engine used by HtmlUnit (Rhino). If you test JavaScript using HtmlUnit the results may differ signicantly from those browsers. When we say JavaScript we actually mean JavaScript and the DOM. Although the DOM is dened by the W3C each browser out there has its own quirks and differences in their implementation of the DOM and in how JavaScript interacts with it. HtmlUnit has an impressively complete implementation of the DOM and has good support for using JavaScript, but it is no different from any other browser: it has its own quirks and differences from both the W3C standard and the DOM implementations of the major browsers, despite its ability to mimic other browsers. With WebDriver, we had to make a choice; do we enable HtmlUnits JavaScript capabilities and run the risk of teams running into problems that only manifest themselves there, or do we leave JavaScript disabled, knowing that there are more and more sites that rely on JavaScript? We took the conservative approach, and by default have disabled support when we use HtmlUnit. With each release of both WebDriver and HtmlUnit, we reassess this decision: we hope to enable JavaScript by default on the HtmlUnit at some point. Enabling JavaScript If you cant wait, enabling JavaScript support is very easy:
HtmlUnitDriver driver = new HtmlUnitDriver(); driver.setJavascriptEnabled(true);

This will cause the HtmlUnit Driver to emulate Internet Explorers JavaScript handling by default.

10.5.2 Firefox Driver

Pros Runs in a real browser and supports JavaScript Faster than the Internet Explorer Driver Cons Slower than the HtmlUnit Driver

10.5. WebDriver Implementations

115

Selenium Documentation, Release 1.0

Before Going Any Further The Firefox Driver contains everything it needs in the JAR le. If youre just interested in using this driver, then all you need to do is put the webdriver-firefox.jar or webdriver-all.jar on your CLASSPATH, and WebDriver will do everything else for you. If you want to dig deeper, though, carry on reading! Important System Properties The following system properties (read using System.getProperty() and set using System.setProperty() in Java code or the -DpropertyName=value command line ag) are used by the Firefox Driver: Property What it means webThe location of the binary used to control Firefox. driver.refox.bin webThe name of the prole to use when starting Firefox. This defaults to driver.refox.prole WebDriver creating an anonymous prole webShould be true if temporary les and proles should not be deleted driver.reap_prole Normally the Firefox binary is assumed to be in the default location for your particular operating system: OS Expected Location of Firefox Linux refox (found using which) Mac /Applications/Firefox.app/Contents/MacOS/refox Windows XP %PROGRAMFILES%\Mozilla Firefox\refox.exe Windows Vista \Program Files (x86)\Mozilla Firefox\refox.exe By default, the Firefox driver creates an anonymous prole Installing a Downloaded Binary The wedriver-all.zip which may be downloaded from the website, contains all the dependencies (including the common library) required to run the Firefox Driver. In order to use it: Copy all the jar les on to your CLASSPATH.

10.5.3 Internet Explorer Driver

This driver has been tested with Internet Explorer 6, 7 and 8 on XP. It has also been successfully tested on Vista. Pros Runs in a real browser and supports JavaScript Cons Obviously the Internet Explorer Driver will only work on Windows! Comparatively slow (though still pretty snappy :) 116 Chapter 10. Selenium 2.0 and WebDriver

Selenium Documentation, Release 1.0

Installing Simply add webdriver-all.jar to your CLASSPATH. You do not need to run an installer before using the Internet Explorer Driver, though some conguration is required. Required Conguration Add every site you intend to visit to your Trusted Sites If you do not do this, then you will not be able to interact with the page.

10.5.4 Chrome Driver

See below for instructions on how to install the Chrome Driver. Note that Chrome Driver is one of the newest drivers. Please report any problems through the issue tracker. Pros Runs in a real browser and supports JavaScript Because Chrome is a Webkit-based browser, the Chrome Driver may allow you to verify that your site works in Safari. Note that since Chrome uses its own V8 JavaScript engine rather than Safaris Nitro engine, JavaScript execution may differ. Cons Slower than the HtmlUnit Driver Before Going Any Further The Chrome Driver contains everything it needs in the JAR le. If youre just interested in using this driver, then all you need to do is put webdriver-all.jar on your CLASSPATH, and WebDriver will do everything else for you. The Chrome Driver_ works with Google Chrome version 4.0 and above. Important System Properties The following system properties (read using System.getProperty() and set using System.setProperty() in Java code or the -DpropertyName=value command line ag) are used by the Chrome Driver: Property webdriver.chrome.bin webdriver.reap_prole What it means The location of the binary used to control Chrome. Should be true if temporary les and proles should not be deleted

Normally the Chrome binary is assumed to be in the default location for your particular operating system:

10.5. WebDriver Implementations

117

Selenium Documentation, Release 1.0

OS Linux Mac Windows XP Windows Vista

Expected Location of Chrome /usr/bin/google-chrome /Applications/Google Chrome.app/Contents/MacOS/GoogleChrome or /User/:username/:as_to_the_left %HOMEPATH%\Local Settings\Application Data\Google\Chrome\Application\chrome.exe C:\Users%USERNAME%\AppData\Local\Google\Chrome\Application\chrome.exe

Installing a Downloaded Binary The wedriver-all.zip which may be downloaded from the website, contains all the dependencies required to run the Chrome Driver. In order to use it, copy all the jar les on to your CLASSPATH.

10.6 Emulating Selenium RC

The Java version of WebDriver provides an implementation of the Selenium RC API. It is used like so:
// You may use any WebDriver implementation. Firefox is used here as an example WebDriver driver = new FirefoxDriver(); // A "base url", used by selenium to resolve relative URLs String baseUrl = "http://www.google.com" ; // Create the Selenium implementation Selenium selenium = new WebDriverBackedSelenium(driver, baseUrl); // Perform actions with selenium selenium.open( "http://www.google.com" ); selenium.type( "name=q" , "cheese" ); selenium.click( "name=btnG" );

// And get the underlying WebDriver implementation back. This will refer to the // same WebDriver instance as the "driver" variable above. WebDriver driverInstance = ((WebDriverBackedSelenium) selenium).getUnderlyingWebDriver()

10.6.1 Pros
Allows for the WebDriver and Selenium APIs to live side-by-side Provides a simple mechanism for a managed migration from the Selenium RC API to WebDrivers Does not require the standalone Selenium RC server to be run

10.6.2 Cons
Does not implement every method More advanced Selenium usage (using browserbot or other built-in JavaScript methods from Selenium Core) may not work 118 Chapter 10. Selenium 2.0 and WebDriver

Selenium Documentation, Release 1.0

Some methods may be slower due to underlying implementation differences

10.6.3 Backing WebDriver with Selenium

WebDriver doesnt support as many browsers as Selenium RC does, so in order to provide that support while still using the WebDriver API, you can make use of the SeleneseCommandExecutor It is done like this:

Capabilities capabilities = new DesiredCapabilities() capabilities.setBrowserName( "safari" ); CommandExecutor executor = new SeleneseCommandExecutor( "http:localhost:4444/" , "http:// WebDriver driver = new RemoteWebDriver(executor, capabilities);

There are currently some major limitations with this approach, notably that ndElements doesnt work as expected. Also, because were using Selenium Core for the heavy lifting of driving the browser, you are limited by the JavaScript sandbox.

10.7 Tips and Tricks

10.7.1 Using Drag and Drop
It may not be immediately obvious, but if youre using a browser that supports it you can cast a WebElement to RenderedWebElement and then its easy to do drag and drop:
// Note the casts RenderedWebElement from = (RenderedWebElement) driver.findElement(By.id( "one" )); RenderedWebElement to = (RenderedWebElement) driver.findElement(By.id( "two" )); from.dragAndDropOn(to);

Currently, only the Firefox Driver supports this, but you should also expect support for the Internet Explorer Driver too.

10.7.2 Changing the user agent

This is easy with the Firefox Driver:
FirefoxProfile profile = new FirefoxProfile(); profile.addAdditionalPreference( "general.useragent.override" , "some UA string" ); WebDriver driver = new FirefoxDriver(profile);

10.7.3 Tweaking an existing Firefox prole

Suppose that you wanted to modify the user agent string (as above), but youve got a tricked out Firefox prole that contains dozens of useful extensions. There are two ways to obtain this prole. Assuming that the prole has been created using Firefoxs prole manager (firefox -ProfileManager):

10.7. Tips and Tricks

119

Selenium Documentation, Release 1.0

ProfileIni allProfiles = new ProfilesIni(); FirefoxProfile profile = allProfiles.getProfile( "WebDriver" ); profile.setPreferences( "foo.bar" , 23); WebDriver driver = new FirefoxDriver(profile);

Alternatively, if the prole isnt already registered with Firefox:

File profileDir = new File( "path/to/top/level/of/profile" ); FirefoxProfile profile = new FirefoxProfile(profileDir); profile.addAdditionalPreferences(extraPrefs); WebDriver driver = new FirefoxDriver(profile); Enabling features that might not be wise to use in Firefox

As we develop features in the Firefox Driver, we expose the ability to use them. For example, until we feel native events are stable on Firefox for Linux, they are disabled by default. To enable them:
FirefoxProfile profile = new FirefoxProfile(); profile.setEnableNativeEvents(true); WebDriver driver = new FirefoxDriver(profile);

10.8 How XPATH Works in WebDriver

At a high level, WebDriver uses a browsers native XPath capabilities wherever possible. On those browsers that dont have native XPath support, we have provided our own implementation. This can lead to some unexpected behaviour unless you are aware of the differences in the various xpath engines. Driver HtmlUnit Driver Internet Explorer Driver Firefox Driver Tag and Attribute Name Lower-cased Lower-cased Case insensitive Attribute Values As they appear in the HTML As they appear in the HTML As they appear in the HTML Native XPath Support Yes No Yes

This is a little abstract, so for the following piece of HTML:

<input type= "text" name= "example" /> <INPUT type= "text" name= "other" />

The following number of matches will be found XPath expression //input //INPUT HtmlUnit Driver 1 (example) 0 Firefox Driver 2 2 Internet Explorer Driver 2 0

10.8.1 Matching Implicit Attributes

Sometimes HTML elements do not need attributes to be explicitly declared because they will default to known values. For example, the input tag does not require the type attribute because it defaults to

120

Chapter 10. Selenium 2.0 and WebDriver

Selenium Documentation, Release 1.0

text. The rule of thumb when using xpath in WebDriver is that you should not expect to be able to match against these implicit attributes.

10.9 Getting and Using WebDriver

10.9.1 From a New Download
Unpack the webdriver-all.zip you can download from the site, and add all the JARs to your CLASSPATH. This will give you the Chrome Driver, Firefox Driver, HtmlUnit Driver, Internet Explorer Driver, Remote Web Driver client and the support packages. The support packages give you useful helper classes, such as the LiftStyleApi and the PageFactory.

10.9.2 With Maven

If you want to use the HtmlUnit Driver, add the following dependency to your pom.xml:
<dependency> <groupId>org.seleniumhq.webdriver</groupId> <artifactId>webdriver-htmlunit</artifactId> <version>0.9.7376</version> </dependency>

If you want to use the Firefox Driver, you need to add the following dependency to your pom.xml:
<dependency> <groupId>org.seleniumhq.webdriver</groupId> <artifactId>webdriver-firefox</artifactId> <version>0.9.7376</version> </dependency>

If you want to use the Internet Explorer Driver, you need to add the following dependency to your pom.xml:
<dependency> <groupId>org.seleniumhq.webdriver</groupId> <artifactId>webdriver-ie</artifactId> <version>0.9.7376</version> </dependency>

If you want to use the Chrome Driver, you need to add the following dependency to your pom.xml:
<dependency> <groupId>org.seleniumhq.webdriver</groupId> <artifactId>webdriver-chrome</artifactId> <version>0.9.7376</version> </dependency>

Finally, if you like to use any of our support classes, you should add the following dependency to your pom.xml:

10.9. Getting and Using WebDriver

121

Selenium Documentation, Release 1.0

<dependency> <groupId>org.seleniumhq.webdriver</groupId> <artifactId>webdriver-support</artifactId> <version>0.9.7376</version> </dependency>

10.10 Roadmap
The roadmap for WebDriver is available here

10.11 Further Resources

You can nd further resources for WebDriver in WebDrivers wiki Appendixes:

122

Chapter 10. Selenium 2.0 and WebDriver

CHAPTER

ELEVEN

.NET CLIENT DRIVER CONFIGURATION

.NET client Driver can be used with Microsoft Visual Studio. To Congure it with Visual do as Following. Launch Visual Studio and navigate to File > New > Project.

Select Visual C# > Class Library > Name your project > Click on OK button.

123

Selenium Documentation, Release 1.0

A Class (.cs) is created. Rename it as appropriate.

Under right hand pane of Solution Explorer right click on References > Add References.

124

Chapter 11. .NET client driver conguration

Selenium Documentation, Release 1.0

Select following dll les - nmock.dll, nunit.core.dll, nunit.framework.dll,ThoughtWorks. Selenium.Core.dll, ThoughtWorks.Selenium.IntegrationTests.dll, ThoughtWorks.Selenium.UnitTests.dll and click on Ok button

125

Selenium Documentation, Release 1.0

With This Visual Studio is ready for Selenium Test Cases.

126

Chapter 11. .NET client driver conguration

CHAPTER

TWELVE

JAVA CLIENT DRIVER CONFIGURATION

In General conguration of Selenium-RC with any java IDE would have following steps: Download Selenium-RC from the SeleniumHQ downloads page Start any java IDE Create new project Add to your project classpath selenium-java-client-driver.jar Record your test from Selenium-IDE and translate it to java code (Selenium IDE has automatic translation feature to generate tests in variety of languages) Run selenium server from console Run your test in the IDE These points have been delineated below with reference to Eclipse and IntelliJ:

12.1 Conguring Selenium-RC With Eclipse

Eclipse is a multi-language software development platform comprising an IDE and a plug-in system to extend it. It is written primarily in Java and is used to develop applications in this language and, by means of the various plug-ins, in other languages as well as C/C++, Cobol, Python, Perl, PHP and more. Following lines describes conguration of Selenium-RC with Eclipse - Version: 3.3.0. (Europa Release). It should not be too different for higher versions of Eclipse Launch Eclipse. Select File > New > Other.

127

Selenium Documentation, Release 1.0

Java > Java Project > Next

128

Chapter 12. Java Client Driver Conguration

Selenium Documentation, Release 1.0

Provide Name to your project, Select JDK in Use a project Specic JRE option (JDK 1.5 selected in this example) > click Next

12.1. Conguring Selenium-RC With Eclipse

129

Selenium Documentation, Release 1.0

Keep JAVA Settings intact in next window. Project specic libraries can be added here. (This described in detail in later part of document.)

130

Chapter 12. Java Client Driver Conguration

Selenium Documentation, Release 1.0

Click Finish > Click on Yes in Open Associated Perspective pop up window.

12.1. Conguring Selenium-RC With Eclipse

131

Selenium Documentation, Release 1.0

This would create Project Google in Package Explorer/Navigator pane.

132

Chapter 12. Java Client Driver Conguration

Selenium Documentation, Release 1.0

Right click on src folder and click on New > Folder

12.1. Conguring Selenium-RC With Eclipse

133

Selenium Documentation, Release 1.0

Name this folder as com and click on Finish button. This should get com package insider src folder.

134

Chapter 12. Java Client Driver Conguration

Selenium Documentation, Release 1.0

Following the same steps create core folder inside com

12.1. Conguring Selenium-RC With Eclipse

135

Selenium Documentation, Release 1.0

SelTestCase class can be kept inside core package. Create one more package inside src folder named testscripts. This is a place holder for test scripts. Please notice this is about the organization of project and it entirely depends on individuals choice / organizations standards. Test scripts package can further be segregated depending upon the project requirements.

136

Chapter 12. Java Client Driver Conguration

Selenium Documentation, Release 1.0

Create a folder called lib inside project Google. Right click on Project name > New > Folder. This is a place holder for jar les to project (i.e. Selenium client driver, selenium server etc)

12.1. Conguring Selenium-RC With Eclipse

137

Selenium Documentation, Release 1.0

This would create lib folder in Project directory.

138

Chapter 12. Java Client Driver Conguration

Selenium Documentation, Release 1.0

Right click on lib folder > Build Path > Congure build Path

12.1. Conguring Selenium-RC With Eclipse

139

Selenium Documentation, Release 1.0

Under Library tab click on Add External Jars to navigate to directory where jar les are saved. Select the jar les which are to be added and click on Open button.

140

Chapter 12. Java Client Driver Conguration

Selenium Documentation, Release 1.0

After having added jar les click on OK button.

12.1. Conguring Selenium-RC With Eclipse

141

Selenium Documentation, Release 1.0

Added libraries would appear in Package Explorer as following:

142

Chapter 12. Java Client Driver Conguration

Selenium Documentation, Release 1.0

12.2 Conguring Selenium-RC With Intellij

IntelliJ IDEA is a commercial Java IDE by the company JetBrains. Intellij provides a set of integrated refactoring tools that allow programmers to quickly redesign their code. IntelliJ IDEA provides close integration with popular open source development tools such as CVS, Subversion, Apache Ant and JUnit. Following lines describes conguration of Selenium-RC with IntelliJ 6.0 It should not be very different for higher version of intelliJ. Open a New Project in IntelliJ IDEA.

12.2. Conguring Selenium-RC With Intellij

143

Selenium Documentation, Release 1.0

Provide name and location to Project.

Click Next and provide compiler output path.

144

Chapter 12. Java Client Driver Conguration

Selenium Documentation, Release 1.0

Click Next and select the JDK to be used.

Click Next and select Single Module Project.

12.2. Conguring Selenium-RC With Intellij

145

Selenium Documentation, Release 1.0

Click Next and select Java module.

Click Next and provide Module name and Module content root.

Click Next and select Source directory.

146

Chapter 12. Java Client Driver Conguration

Selenium Documentation, Release 1.0

At last click Finish. This will launch the Project Pan.

Adding Libraries to Project: Click on Settings button in the Project Tool bar.

Click on Project Structure in Settings pan.

12.2. Conguring Selenium-RC With Intellij

147

Selenium Documentation, Release 1.0

Select Module in Project Structure and browse to Dependencies tab.

148

Chapter 12. Java Client Driver Conguration

Selenium Documentation, Release 1.0

Click on Add button followed by click on Module Library.

Browse to the Selenium directory and select selenium-java-client-driver.jar and seleniumserver.jar. (Multiple Jars can be selected b holding down the control key.).

12.2. Conguring Selenium-RC With Intellij

149

Selenium Documentation, Release 1.0

Select both jar les in project pan and click on Apply button.

150

Chapter 12. Java Client Driver Conguration

Selenium Documentation, Release 1.0

Now click ok on Project Structure followed by click on Close on Project Settings pan. Added jars would appear in project Library as following.

12.2. Conguring Selenium-RC With Intellij

151

Selenium Documentation, Release 1.0

Create the directory structure in src folder as following.

152

Chapter 12. Java Client Driver Conguration

Selenium Documentation, Release 1.0

Note: This is not hard and fast convention and might very from project to project. Herein core contains the SelTestCase class which is used to create Selenium object and re up the browser. testscripts package contains the test classes which extend the SelTestCase class. Hence extended structure would look as following.

12.2. Conguring Selenium-RC With Intellij

153

Selenium Documentation, Release 1.0

154

Chapter 12. Java Client Driver Conguration

CHAPTER

THIRTEEN

PYTHON CLIENT DRIVER CONFIGURATION

Download Selenium-RC from the SeleniumHQ downloads page Extract the le selenium.py Either write your Selenium test in Python or export a script from Selenium-IDE to a python le. Add to your tests path the le selenium.py Run Selenium server from the console Execute your test from a console or your Python IDE The following steps describe the basic installation procedure. After following this, the user can start using the desired IDE, (even write tests in a text processor and run them from command line!) without any extra work (at least on the Selenium side). Installing Python Note: This will cover python installation on Windows and Mac only, as in most linux distributions python is already pre-installed by default. Windows 1. Download Active pythons installer from ActiveStates ofcial site: http://activestate.com/Products/activepython/index.mhtml 2. Run the installer downloaded (ActivePython-x.x.x.x-win32-x86.msi)

155

Selenium Documentation, Release 1.0

Mac The latest Mac OS X version (Leopard at this time) comes with Python pre-installed. To install an extra Python, get a universal binary at http://www.pythonmac.org/ (packages for Python 2.5.x). 156 Chapter 13. Python Client Driver Conguration

Selenium Documentation, Release 1.0

You will get a .dmg le that you can mount. It contains a .pkg le that you can launch.

Installing the Selenium driver client for python 1. Download the last version of Selenium Remote Control from the downloads page 2. Extract the content of the downloaded zip le 3. Copy the module with the Seleniums driver for Python (selenium.py) in the folder C:/Python25/Lib (this will allow you to import it directly in any script you write). You will nd the module in the extracted folder, its located inside seleniumpython-driver-client.

Congratulations, youre done! Now any python script that you create can import selenium and start interacting with the browsers.

157

Selenium Documentation, Release 1.0

158

Chapter 13. Python Client Driver Conguration

CHAPTER

FOURTEEN

LOCATING TECHNIQUES
14.1 Useful XPATH patterns
14.1.1 text
Not yet written - locate elements based on the text content of the node.

14.1.2 starts-with
Many sites use dynamic values for elements id attributes, which can make them difcult to locate. One simple solution is to use XPath functions and base the location on what you do know about the element. For example, if your dynamic ids have the format <input id="text-12345" /> where 12345 is a dynamic number you could use the following XPath: //input[starts-with(@id, text-)]

14.1.3 contains
If an element can be located by a value that could be surrounded by other text, the contains function can be used. To demonstrate, the element <span class="top heading bold"> can be located based on the heading class without having to couple it with the top and bold classes using the following XPath: //span[contains(@class, heading)]. Incidentally, this would be much neater (and probably faster) using the CSS locator strategy css=span.heading

14.1.4 siblings
Not yet written - locate elements based on their siblings. Useful for forms and tables.

14.2 Starting to use CSS instead of XPATH

14.2.1 Locating elements based on class
In order to locate an element based on associated class in XPath you must consider that the element could have multiple classes and dened in any order, however with CSS locators this is much simpler (and faster). XPath: //div[contains(@class, article-heading)]

159

Selenium Documentation, Release 1.0

CSS: css=div.article-heading

160

Chapter 14. Locating Techniques