Automated Software Test Optimization using Test Language Processing

摘要

The delivery of error free software has become a major challenge for software practitioner since many past years. In order to deliver an error free software testers spends 40-50 % software design life cycle cost during testing, which further get incremented with changing user demands. The Large existence of test cases for a particular functionality is possible and some of them may cause software fails. Thus it raises a demand to automate existing approach of manual testing which can minimize execution efforts while maintaining the quality of testing. In this paper, a regression framework based on keyword oriented data-driven approach has been proposed for generation and execution of test cases. The methodology for the developed framework is based on Test Language Processing (TLP) which acts as a comprehensive approach to design and execution of test cases. The framework is tested on an open source web application called Vtiger-Customer Relationship Management (CRM) version 5. The framework is compared against manual testing in terms of test suite execution and their optimization. Based on our experiments it is concluded that (1) Test execution time using TLP based framework is significantly low and (2) a test suite optimization of 83.78% is achieved through the proposed TLP framework.