High quality software can be obtained by means of rigorous testing of all the components of the software. This research work has proposed an automated software testing framework that performs a mutant based components impact analysis to identify the higher critical components from the Software Under Test (SUT). In this work, the mutants are automatically generated by injecting faults in the original program and they are used to identify the impact over the other components in the SUT. The generated mutants are executed using a suite of test cases to identify their impact over the other components of the system. Based on their impact level, the critical components are identified and then rigorously verified using the test cases generated using Genetic Algorithm (GA) based approach with branch coverage and mutation score based test adequacy criterion as the fitness functions. For unit testing, the branch coverage based test case adequacy criteria is used to test whether all the branches have been covered or not. In integration testing, the components are tested against the test cases generated using GA by means of identifying the execution trace of each method and each intermediate results is compared against the expected output stored in the repository. The testing tool named as “JImpact Arbiter” developed as part of this work has carried out all these tasks in an automated way and has generated various graphs for the purpose of visualization.
D. Jeyamala, K. Sabari Nathan, S. Balamurugan, A. Jalila
Thiagarajar College of Engineering, India
Critical Components, Impact Analysis, Mutation Analysis, Genetic Algorithm (GA), Branch Coverage Value (BCV) | Published By :
ICTACT
Published In :
ICTACT Journal on Soft Computing
( Volume: 3 , Issue: 4 )
Date of Publication :
July 2013
|