Output details
11 - Computer Science and Informatics
University of York
Class Schema Evolution for Persistent Object-Oriented Software : Model, Empirical Study, and Automated Support
<08>Evolutionary changes to the schemas of persistent objects meant that "old" objects may not be retrievable, or may be retrieved incorrectly. The approach to this issue is normally ad hoc; this paper presents the first automated transformational solution to this problem, based on canonical schema modification operators (SMOs). Empirical assessments on the evolution of large code bases over several generations, e.g. TomCat, Eiffel, show that 17.7% of changes affect the schemas of persistent objects; the approach detected 93% of the SMOs, in the code, and thus can remove cost and sources of error across a range of object-oriented programming languages.