A Novel Method for Test Path Prioritization using Centrality Measures

Introduction

A major problem that the software industry encounters is to develop bug free software. As per latest IBM reports “31% of the projects get cancelled before they are completed, 53% over-run their cost estimates by an average of 189% and for every 100 projects, there are 94 restarts” (Aggarwal & Singh, 2006). Therefore, there is an urgent need to adopt better software engineering concepts and strategies in order to improve software quality. Software testing is of utmost importance amongst the phases of project development since it aims at demonstrating that errors are not present in the program under consideration. The purpose of testing is to show that a program performs its intended functions correctly and establish confidence that it does what it is supposed to do. The two major types of testing include structural and functional testing (Paul et al., 2011). In functional testing, various functions are tested for quality assurance but the internal structure of the program is rarely considered. It usually describes what the system does. On the other hand, structural testing aims at testing the internal structure of the software. The technique that has been discussed in this paper is Path Testing which is a type of structural testing (Tutorials Point, 2017). It is the name given to a group of testing techniques that are based on judiciously selecting a set of test paths through the program. A path in a graph is a finite or infinite sequence of edges which connect a sequence of vertices which, by most definitions, are all distinct from one another. If the set of paths is properly chosen then it means that one has achieved some measure of test thoroughness. For example, picking enough paths to assure that every source statement is executed at least once. Hence it becomes mandatory to have the complete knowledge of the program's structure. This type of testing involves generating a set of test paths that will cover every branch in the program as well as finding a set of test cases that will execute every path in this set of program paths.

An important aspect of path testing is path optimization which does the task of decreasing the number of paths to be tested by prioritizing the independent paths. Till date very little work is done in optimization of path testing. However, Cyclomatic Complexity is one of the approaches that were introduced to directly measure the number of linearly independent paths through a program's source code. This minimized the number of paths to be tested in comparison to normal path testing where all dependent and independent paths are tested. Researchers identified that the effectiveness of path testing rapidly deteriorates as the size of software under test increases (Goldberg & Harrelson, 2005; Malhotra & Manju, 2013). As the size of software increases it becomes difficult to test every possible path in the program. Hence there should be a technique for path prioritization so that number of paths to be tested can be minimized according to their prioritized value. Since the introduction of Cyclomatic Complexity, very little work has been done in the field of path prioritization.