Integrating Mutation Techniques to Keep Specification and Source Code in Sync

Most software development processes rely on testing for assuring that a software's source code is kept in sync with its specification during software evolution. In testing, the degree of conformance between source code and specification is measured by the test success rate. However, the practical value of this metric depends on the quality of the test suite, i.e., its completeness wrt. the specification and source code. Mutation testing has proved to be effective in assessing test suite completeness; however related work either considers mutating the source code or the specification. Our research will investigate how approaches based on mutating code and specification can be integrated to keep both in sync during software evolution. We aim to contribute (1) a concise method to measuring the degree of conformance between specification and code with each modification, (2) a technique for the identification and visualization of fine-granular trace links between them, and (3) an approach for suggesting updates of specification and code based on detected deviations between them.