| With the development of mobile device, CPU processing power and storage capacity are greatly improved which makes it's possible to deploy Web Service on mobile devices. Deploying RESTful Web Service on smart mobile devices to collect and analyze data from individual devices could achieve real-time data acquisition and sharing, and optimized data mining. RESTful Web Service application based on Representational State Transfer (REST) use Web Application Description Language (WADL) to describe service interface. Thus, In order to effectively validate whether the mobile Web Service system could provide correct and available Service, we need to do a lot of testing to ensure the accuracy and reliability of application before large-scale deployment of the Web Service.In this paper, we propose a new method of test data generation for RESTful Web Service base on mutation testing. Our main work include:(1) We extend the data type constraints of WADL for RESTful Web Service and automatically generate preliminary test data set using equivalence partitioning, boundary analysis, error guessing and random generation method(2) We generate mutants using the constraint of WADL based on mutation testing technology. Then eliminate the equivalent mutants and implement mutant set reduction using clustering technology based on genetic algorithm. Finally we select the efficient and non-redundant test data set using greedy algorithm.(3) We develop a automatic test data generation tool based on hadoop to improve efficiency of large-scale test data generation process.We select a RESTful Web Service of online shopping system as a case study to design a series of experiments. The proposed method which proved the method of generate test data could reduce the size of test data set effectively, prevent the input combinatorial explosion, accelerates the testing process using the capability of distributed processing and increasing the efficiency of test data generation for RESTful Web Service system. |
PDF Full Text Request