Research And Implementation Of Automatic Testing Of Web Application Software Based On Watir

With the continuous development of Internet technology, users work on regression testing Web-based applications use the increasing demand, the size of Web applications has been expanded, the traditional manual testing has been unable to meet the large and complex Web systems. How to improve testing efficiency has become an urgent need to address the issue. Web automated testing can be executed 24 hours around the clock, and can also perform monotonous repetition of the test procedure. During execution, automation will not ignore human subjective factors test point and test details. Automated testing in limited time, the number of test cases executed by long time is artificial. Therefore, the introduction of Web automated testing technology is very necessary.Research background, the company has a Web-based IT Service Management System Service Manager, in the development process in order to meet the increasing demand for regression testing, we select Watir(Web Automation Test in Ruby) as an automated test framework, and the framework Watir secondary development, developed a new Watir automated testing framework named SAFIR(Service Manager Automation Framework in Ruby), into the product regression testing work.The main work of this paper is to study the theory of automated testing, test the boundaries of today’s popular automation framework tool to compare and choose Watir as a basic test framework to study the principles of Watir through Html Dom tree to locate and operating elements on the Web. We analyzed the characteristics of Service Manager, and used Watir framework for secondary development, developed a new SAFIR framework. SAFIR framework i s divided into three layers: system maintenance layer, page encapsulation layer and test scripts layer. We added a test log, error screenshots, page package and other functions, to make up for the lack of Watir framework, so that the new framework of the preparation of test scripts, test execution and analysis of test results is more convenient and more perfect. Finally, the method according to the results of SAFIR, analyzes test log and identify the cause of the failure.The new framework will be put to the test in the actual product, the results of the automated and manual execution results were compared with data from a script time, test execution time and defect analysis time to draw conclusions, in large-scale regression testing, although automation test defect analysis time is greater than the manual test, but in the configuration and implementation of test cases automated environment aspects of time, automated tests can improve the efficiency of more than 70% of the tests, the number of defects found a n increase of 15% compared to manual testing. With the increase of the test execution times, they save more time cost, bring greater benefits.