Towards a Lightweight Standard Search Language

In this paper we propose a lightweight search language for MiniZinc that allows a user to direct the search process by employing explicitly defined search templates. A challenge in providing a standard search language is to balance expressiveness with ease of implementation. Previous work mainly introduced general frameworks that facilitate userdefined search strategies at virtually arbitrary degrees of freedom (e.g., programmable search). The goal of this paper is to propose a specific set of explicit search templates that is solver-independent, does not impose high requirements on underlying constraint solvers, and enables users to construct search strategies to tackle problems successfully at a sufficiently abstract level by using predefined search templates. We believe a standard for search should start small and expand as the community of solver writers agree to implement it. The ultimate goal of this work is to define a standard set of search templates that form a language that can be employed across various constraint solving systems.