JSART: javascript assertion-based regression testing

摘要: Web 2.0 applications rely heavily on JavaScript and client-side runtime manipulation of the DOM tree. One way to provide assurance about correctness such highly evolving dynamic is through regression testing. However, loosely typed, dynamic, notoriously challenging analyze test. We propose an automated technique for testing, which based on-the-fly source code instrumentation analysis infer invariant assertions. These obtained assertions are injected back into uncover faults in subsequent revisions web application under Our approach implemented a tool called Jsart. present our case study conducted nine open evaluate proposed approach. The results show that able effectively generate stable detect with high degree accuracy minimal performance overhead.