A Source Code Change Impact Analysis Algorithm for Iterative Software Development

CHA-AS is a source code change impact analysis algorithm for Java programs. CHA-AS differs from other algorithms in that it does not require the program versions it compares to be whole programs with a welldefined program entry point. The need for such an algorithm is evident in iterative software development projects and projects involving the development of code libraries and frameworks—all of which may not have a well-defined program entry point at the time when change impact analysis needs to be performed. The CHA-AS algorithm supports the development of Decision Support Systems for software development managers and programmers working on iterative software development projects, or projects to develop source code libraries and frameworks. This paper describes the CHA-AS algorithm and demonstrates it to be efficient and effective in calculating source code change impact. DOI: 10.4018/jdsst.2012100104 International Journal of Decision Support System Technology, 4(4), 60-75, October-December 2012 61 Copyright © 2012, IGI Global. Copying or distributing in print or electronic forms without written permission of IGI Global is prohibited. calls. Although fully-functional software programs can reasonably be expected to have such a program entry point, other software cannot. Software being developed using iterative software development methods, such as TestDriven Development (TDD) (Beck, 2002), may not have a well-defined program entry point at the time when change impact analysis needs to be performed; and code libraries and software development frameworks may never have such an entry point. Source code change impact analysis algorithms may be used to form the foundation of Decision Support Systems that provide critical information to software development managers and programmers. The most obvious use of these change impact based Decision Support Systems is to obtain information about how project schedules and budgets will be affected by software change requests. This information can be used as input for project scheduling, feasibility analysis, and client negotiation decisions. These Decision Support Systems can also be used by software developers to evaluate different alternatives for implementing change requests in existing software. Other decisions that these systems can support include determining which if any unit tests need to be reexecuted after development of a new iteration (Parashar, Bhatia, & Kalia, 2011; Ren, Shah, Tip, Ryder, & Chesley, 2004), which code needs to be re-analyzed when using automated test data generation (DeMillo & Offutt, 1991) on iterative projects, and which software mutants need to be generated and executed when using mutation analysis (DeMillo, Krauser, & Mathur, 1991) on iterative projects. This paper promotes the development of change impact based Decision Support Systems by describing and evaluating the Class Hierarchy Analysis, Arbitrary Start (CHA-AS) change impact analysis algorithm. CHA-AS is a source code change impact analysis algorithm for Java programs, which can identify the impact of a set of source code changes in situations where a clearly defined program entry point does not exist at the time when change impact analysis needs to be performed. A prior version of the algorithm was presented at the 2010 Annual Meeting of the Decision Sciences Institute (Wilkerson, 2010).