Implementation of term rewritings with the evaluation strategy

Abstract

The evaluation strategy (the E-strategy) is more flexible than lazy, eager or any other fixed order of evaluation because each operation can have its own local strategy. An efficient implementation of the E-strategy based on strategy lists is proposed. A strategy list for a term t is basically a sequence of all eager positions (reachable nodes) in t. The order of the strategy list corresponds to the order of the evaluation of t. The E-strategy can control parallel rewritings as well as sequential ones. The parallel extension of the E-strategy is also described.