Parallel Bottom-up Tomita parsing

Abstract

In the previous chapter we have derived the parsing schema LR(0) and concluded that the differences with Earley are trivial details. Hence there is a structural correspondence between Earley chart parser and generalized LR parsers. This correspondence can be used to cross-fertilize different variants of either kind algorithm. A particularly interesting example that we will discuss here is the Parallel Bottom-up Tomita (PBT) algorithm [Lankhorst and Sikkel, 1991], [Sikkel and Lankhorst, 1992], where the conventional parallelization of Earley’s algorithm is applied to the Tomita parser.