Constraint Logic Programming

点击翻译按钮获取中文摘要

The long-term goal of this research is to support established paradigms in programming language research in order to reduce programming effort and shorten development time. Constraint logic programming (CLP) is a new class of programming languages whose basic operation is constraint solving over a computation domain. CLP languages are best used for combinatorial problems that often require considerable expertise and development time. Typical examples include scheduling and formal verification of circuits. CLP languages presently support partial enumeration (subsuming branch and bound and backtracking) as the main paradigm. A successful representative of the first generation of these languages is CHIP which is now widely used in industry (e.g., at Siemens, Bull and ICL) for tasks like the scheduling of ships in Hong Kong harbor. This project will develop a second-generation CLP language, extending the support for combinatorial problems in three directions: 1. Constraint handling: new reasoning and pruning techniques will be incorporated by providing new primitive operations such as constraint entailment and new ways of combining primitive constraints; 2. Execution models: dynamic programming and local search will be investigated in order to incorporate them into CLP languages; 3. Computation domains: the project will investigate new application areas such as computational geometry and parsing where constraints play a fundamental role in order to identify new computation domains.