Search-Based Testing of Ajax Web Applications

Ajax is an emerging Web engineering technology that supports advanced interaction features that go beyond Web page navigation. The Ajax technology is based on asynchronous communication with the Web server and direct manipulation of the GUI, taking advantage of reflection. Correspondingly, new classes of Web faults are associated with Ajax applications.In previous work, we investigated a state-based testing approach, based on semantically interacting events. The main drawback of this approach is that exhaustive generation of semantically interacting event sequences limits quite severely the maximum achievable length, while longer sequences would have higher fault exposing capability. In this paper we investigate a search-based algorithm for the exploration of the huge space of long interaction sequences, in order to select those that are most promising, based on a measure of test case diversity