Enhancement of Branch Coverage using Java Program Code Transformer

Abstract

In traditional concolic testing branch coverage is low. Automated technique appears as a promising technique to reduce test time and eort. In this project we used a code transformation technique. We take input a simple java program and transform it using various algorithms. We have used four algorithms to transform the given code into a transformed code. We used Quine Mc-cluskey method and Petric methods to achieve the transformed code. After transforming the code, we pass it through a tool called Cobertura which gives the branch coverage of that transformed code. Here we observe that the percentage of branch coverage using the transformed code is greater than the coverage of original code. Hence, our technique helps to achieve a significant increase in branch coverage in comparison to traditional techniques.