Trace-baed Error Prevention and Detection of Consistency Maintenance via Bidirectional Graph Transformations

Abstract

Bidirectional transformations are used to maintain consistency between various artifacts. They are studied across different research fields such as databases, programming languages, category theories and software engineering. The consistency is represented in terms of transformations and consistency is maintained by propagating changes through transformations in forward and backward directions when one of the artifacts are changed. In the authors' previous work, the users of bidirectional graph transformations are informed before backward transformation the possibility of errors by explaining the reason using automatically-generated traceability information. In this paper, we recaptulate the previous work by regarding bidirectional transformations as specifications of consistency relation and errors as violation of consistency, and discuss the essence of the mechanism focusing on the data flow in the bidirectional transformation languages.