Genetically Programmed Pattern Matching for Overlapping Patterns

Abstract

Pattern matching is a fundamental feature in many applications such as functional programming, logic programming, theorem proving, term rewriting and rule-based expert systems. Usually, patterns size is not constrained and ambiguous patterns are allowed. This generality leads to a clear and concise programming style. However, it yields challenging problems in compiling of such programming languages. Generally, patterns are pre-processed into a deterministic finite automaton. With ambiguous or overlapping patterns a subject term may be an instance of more than one pattern. In this case, pattern matching order in lazy evaluation affects the size of the matching automaton and the matching time. Furthermore, it may even affect the termination properties of term evaluations. In this paper, we engineer good traversal orders that allow one to design an efficient adaptive pattern-matchers that visit necessary positions only. We do so using genetic programming to evolve the most adequate traversal order given the set of allowed patterns. Hence, we improve time and space requirements of pattern-matching as well as termination properties of term evaluation