Guided Mutation Testing for JavaScript Web Applications

摘要: Mutation testing is an effective test adequacy assessment technique. However, there a high computational cost in executing the suite against potentially large pool of generated mutants. Moreover, much effort involved filtering out equivalent Prior work has mainly focused on detecting mutants after mutation generation phase, which computationally expensive and limited efficiency. We propose algorithm to select variables branches for as well metric, called $FunctionRank$ , rank functions according their relative importance from application’s behaviour point view. present technique that leverages static dynamic analysis guide process towards parts code are more likely influence program’s output. Further, we focus JavaScript language, set operators specific web applications. implement our approach tool Mutandis . The results empirical evaluation show (1) than 93 percent non-equivalent, (2) 75 surviving non-equivalent top 30 ranked functions.