Parsing of Hyperedge Replacement Grammars with Graph Parser Combinators

Authors

  • Steffen Mazanek
  • Mark Minas

DOI:

https://doi.org/10.14279/tuj.eceasst.10.143

Abstract

Graph parsing is known to be computationally expensive. For this reason the construction of special-purpose parsers may be beneficial for particular graph languages. In the domain of string languages so-called parser combinators are very popular for writing efficient parsers. Inspired by this approach, we have proposed graph parser combinators in a recent paper, a framework for the rapid development of special-purpose graph parsers. Our basic idea has been to define primitive graph parsers for elementary graph components and a set of combinators for the flexible construction of more advanced graph parsers. Following this approach, a declarative, but also more operational description of a graph language can be given that is a parser at the same time. In this paper we address the question how the process of writing correct parsers on top of our framework can be simplified by demonstrating the translation of hyperedge replacement grammars into graph parsers. The result are recursive descent parsers as known from string parsing with some additional nondeterminism.

Downloads

Published

2008-07-14

How to Cite

[1]
S. Mazanek and M. Minas, “Parsing of Hyperedge Replacement Grammars with Graph Parser Combinators”, eceasst, vol. 10, Jul. 2008.