Improving the Java PathFinder Tool for Generating Test Data by Optimizing Constraints

Test data generation is one of the most important and crucial phases in software testing. Structural test data generation techniques are based on source code to create test data, which can uncover programmer\'s faults.A test data generation technique using symbolic execution is implemented in the Java PathFinder tool to generate test data for Java programs. However, Java PathFinder has a performance limitation: execution time is quite big. In this paper, we propose applying an optimization algorithm for constraints - representing execution paths in programs - in order to reduce the execution time but ensure the coverage effectiveness. The experiment on different Java programs shows a promising result of the proposed solution