Investigating the applicability of generating test cases for web applications based on traditional graph coverage

Abstract

Web applications provide services to hundreds of millions of people throughout the world. However, developers face a range of problems and challenges in testing them, including the fact that web applications run on diverse and heterogeneous platforms and are written using diverse programming languages. Moreover, they can be dynamic, with their contents and structures determined by inputs from users, so they need to be tested to ensure their validity. In this paper, we investigate the ability to generate a set of test cases for web applications based on traditional graph coverage criteria. First, we extracted the in-link and out-link from given web applications in order to draw a web graph, before extracting the prime paths from the graph. After that, the invalid transitions were built from the prime paths. Finally, all the invalid transitions were extended with valid transitions. We evaluated our investigation process by using different sizes of web applications. Two cases studies were used in this paper, the first a small size application and the second a medium size. The results show how difficult it is to run a huge number of test cases generated manually using graph coverage criteria, even for a small web application.