Temporal Graph Queries to Support Software Evolution

Graph transformation techniques have already been used successfully by several research groups to support re-engineering of large legacy systems. Where others often aim at transforming the system to improve it, we advocate an evolutionary approach that embeds transformations within the ordinary development process and provides tool support to monitor the ongoing progress regularly. In this paper, we discuss how temporal graph queries based on Fujaba story diagrams can provide a natural means to express trend-oriented metrics and consistency rules that we identified in our industrial case studies. To this end, we discuss a first-order logic rather than operational interpretation of a graph queries and show how well-known temporal logic operators can be added to express rules over consecutive states of the same instance graph