Open AccessComplexity of Acyclic Term Graph Rewriting
Author(s) -
Martin Avanzini,
Georg Moser
Publication year - 2016
Publication title -
hal (le centre pour la communication scientifique directe)
Language(s) - English
Resource type - Conference proceedings
DOI - 10.4230/lipics.fscd.2016
Subject(s) - rewriting , term (time) , computer science , confluence , directed acyclic graph , graph , graph rewriting , theoretical computer science , programming language , algorithm , physics , quantum mechanics
International audienceTerm rewriting has been used as a formal model to reason about the complexity of logic, functional , and imperative programs. In contrast to term rewriting, term graph rewriting permits sharing of common sub-expressions, and consequently is able to capture more closely reasonable implementations of rule based languages. However, the automated complexity analysis of term graph rewriting has received little to no attention. With this work, we provide first steps towards overcoming this situation. We present adap-tions of two prominent complexity techniques from term rewriting, viz, the interpretation method and dependency tuples. Our adaptions are non-trivial, in the sense that they can observe not only term but also graph structures, i.e. take sharing into account. In turn, the developed methods allow us to more precisely estimate the runtime complexity of programs where sharing of sub-expressions is essential