Grant: Collaborative Research and Development Grants

We wish to investigate if in large, long-lived industrial software systems, such as those developed or purchased by CA, code cloning, and code forking is a fairly common practice, as has already determined is the case in open source systems. We propose to use one or more systems owned by CA as case studies for this research. The code cloning that is detected in this proprietary software, then needs to be categorized according to why it was introduced into the system, and on the software engineering merits of this deliberate or unintentional code cloning. Having developed a taxonomy to classify the different ways in which code cloning arises in commercial systems, we then wish to extend and enhance existing tools developed by us, that when presented with large software systems perform an assessment of the degree of code duplication within these systems, and the merits of that duplication. We also wish to develop tools that automate the reduction of source code, so as to remove cloned code, without altering the external behaviour of that code when this cloned code is deemed to be undesirable.