Implementation patterns to extend the search-based reduction Model of MTest.search

Resumen

The process of generating test cases contributes to the quality of software products, detecting as many errors as possible. The design stage of test cases from test values is a challenging task, so its automation is necessary to increase the effectiveness to detect errors and reduce cost. Metaheuristic algorithms such as genetic algorithms, simulated annealing, and tabu search have been successfully applied to solve the combinatorial explosion of automatic test suite generation. The MTest.search model for automatic unit test generation has domain model extension, test, and execution mechanisms defined. In this work, implementation patterns are presented to extend the search-based reduction model. This proposal allows incorporating new configurations of the optimization problem to generate combinations of test values obtaining reduced test suites. To validate the proposed solution, case studies were defined based on the analysis of extensions implemented following the patterns defined in this work and the resulting test cases. In the implemented extensions, the significance criterion of the test values and of the paths/scenarios was introduced to obtain reduced test suites with greater potential to detect errors.