A Logical Approach to Data-Aware Automated Sequence Generation

Automated sequence generation can be loosely defined as the algorithmic construction of a sequence of objects satisfying a set of constraints formulated declaratively. A variety of scenarios, ranging from self-configuration of network devices to automated testing of web services, can be described as automated sequence generation problems. In all these scenarios, the sequence of valid objects and their data contents are interdependent. Despite these similarities, most existing solutions for these scenarios consist of

ad hoc

, domain-specific tools. This paper stems from the observation that, when such “data-aware” constraints are expressed using mathematical logic, automated sequence generation becomes a case of satisfiability solving. This approach presents the advantage that, for many logical languages, existing satisfiability solvers can be used off-the-shelf. The paper surveys three logics suitable to express real-world data-aware constraints and discusses the practical implications, with respect to automated sequence generation, of some of their theoretical properties.