Enhancing the LR Parsing Strategy Using Incremental GPLR Parsing Method

Abstract

Parsers in modern integrated development environments (IDEs) for generalpurpose languages are virtually all of ad hoc, recursive descent variety. While such
parsers have many disadvantages when compared with machine-generated LALR(1)
parsers but they have two major good qualities: they are not restricted to any finite of
lookahead, and in IDE, they can re-parse parts of a file as they change rather than reparsing the entire file. Theoretically, both of these two capabilities can be achieved
through variations of the traditional LR parsing techniques, but the traditional LR
parsing methods still suffer two irresolvable problems; which are shift-reduce and
reduce-reduce conflicts. In this research, we are trying to solve these two drawbacks
with preserving the capabilities of the traditional LR parsing techniques. This has
been achieved by employing the generalized piecewise LR parsing (GPLR) technique
instead of the traditional LR techniques.