Mtest.search extensions for test code generation

Abstract

Software testing focuses on detecting defects or failures during code execution. Testing is a challenging creative task, requiring automation. The adoption of good practices and testing strategies contributes to increasing the efficiency of software development companies. MTest.search model for automatic unit test generation has defined domain model extension, test, and execution mechanisms. In this work, mechanisms to extend the search-based reduction model are presented. The proposed extensions take into account the objects and sets involved in the source code, and enhance the detection of defects or failures based on the significance of the values ​​and paths / scenarios involved in the test. To validate the proposal, three case studies were defined using classical methods and real projects.