The application of metaheuristic search techniques, such as evolutionary algorithms, to the problem of automatically generating software test data has been a burgeoning interest for many researchers in recent years. To date, work in applying search techniques to structural test data generation has largely focused on generating inputs for test objects with input-output behaviour. This thesis aims to extend the approach for test objects with state behaviour. This presents several challenges, not least because test goals with state-based test objects may require input sequences to be generated. Another problem includes generating test data in the presence of internal variables such ... [more] as flags, enumerations and counters. Such variables are often responsible for managing the “state” of the test object. However, their use can lead to information loss with regards to the original input conditions that lead to the fulfilment of certain test goals. Consequently the search receives less guidance, and may fail to find test data. This thesis proposes an extended evolutionary structural test data generation approach that allows input sequences to be generated, and tackles internal variable problems through hybridization of the method with an extended chaining approach. The basic idea of the chaining approach is to find a sequence of statements, involving internal variables, which need to be executed prior to the test goal. By requiring that these statements are executed, information previously unavailable to the search can be made use of, possibly guiding it into potentially promising and unexplored areas of the test object’s input domain. A number of experiments show the value of the approach for both test objects with states and test objects with input-output behaviour. For all test objects considered, higher levels of branch coverage are obtained

Bibtex Entry

@phdthesis{McMinn2005,
  author      = "McMinn, Phil",
  title       = "Evolutionary Search for Test Data in the Presence of State Behaviour",
  year        = "2005",
  institution = "The University of Sheffield"
}