Immediate Error Detection in Strong LL(1) Parsers

Abstract

An algorithm is presented which guarantees that no erroneous actions are performed by a Strong LL(1) parser while parsing an incorrect input. The class of Strong LL(1) grammars for which this algorithm is usable appears to closely coincide with grammars actually used in practice. Further any Strong LL(1) grammar can be algorithmically transformed into a form suitable for use with this algorithm.