Reengineering with Reflexion Models: A Case Study

• Authors: Gail C. Murphy;David Notkin
• Venue: Computer
• Year: 1997

Reengineering large and complex software systems is often very costly. Reflexion models let software engineers begin with a structural high-level model that they can selectively refine to rapidly gain task-specific knowledge about the source code. In this article we give an overview of our technique and then relate how a Microsoft engineer used it to aid an experimental reengineering of Excel-a product that comprises about 1.2 million lines of C code. Our technique begins with a high-level model, which users define on the basis of a desired software engineering task. Thus, the next steps in our technique are to extract a model of the source, define a map, and, through a set of computation tools, compare the two models. This lets software engineers effectively validate their high-level reasoning with information from the source code. To satisfy the need for a quick and inexpensive method, we made the technique "lightweight" and iterative. The user can easily and rapidly access the structural information of interest and can balance the cost of refinement with the potential benefits of a more complete and accurate model. The engineer in our case study-a developer with 10-plus years at Microsoft-specified and computed an initial reflexion model of Excel in a day and then spent four weeks iteratively refining it. He estimated that gaining the same degree of familiarity with the Excel source code might have taken up to two years with other available approaches. The experimental reengineering case study shows that our technique has practical application. First, the engineer chose to use the technique even when facing extreme pressure. Second, he continued to use the technique beyond the original time (one month) to refine additional parts of the reflexion model for Excel and to compute reflexion models for successive Excel versions. Finally, the engineer stated that the slowdowns he did encounter while performing the experimental reengineering were often due to a lack of up-front understanding. Had the reflexion model technique been used more during planning, he felt that he might have understood the product better to begin with. We believe our technique was successful in large part because it uses approximation. This ensures a smooth feedback from the time invested in applying the technique to the results. The more time the Microsoft engineer spent refining the map, the more information he derived. Although this curve is not completely smooth, the engineer was able to gauge the accuracy of the results and use that information to manage the time and effort invested in using the technique.