English

Forest languages defined by counting maximal paths

Formal Languages and Automata Theory 2021-06-15 v2

Abstract

A leaf path language is a Boolean combination of sets of the form mEkL\mathsf{{}^mE}^k L, with k1k \ge 1 and LL a regular word language, which consist of those forests where the node labels in at least kk leaf-to-root paths make up a word that belongs to LL. We look at the class D\mathsf{*D} of the languages recognized by iterated wreath products of syntactic algebras of leaf path languages. We prove the existence of an algorithm that, given a regular forest language, returns in finite time a sequence of such algebras; their wreath product is divided by the language's syntactic algebra if, and only if this language belongs to D\mathsf{*D}, which makes membership in this class a decidable question. The result also applies to the subclasses PDL\mathsf{PDL} and CTL\mathsf{CTL^*}.

Keywords

Cite

@article{arxiv.2105.09970,
  title  = {Forest languages defined by counting maximal paths},
  author = {Martin Beaudry},
  journal= {arXiv preprint arXiv:2105.09970},
  year   = {2021}
}

Comments

The proof of the main Lemma (3.11, section 3.4) is incomplete: in the middle of page 22, the fact that $\gamma$ is weakly distributive is not sufficient to justify the chain of two inclusions used to invoke Proposition 2.1

R2 v1 2026-06-24T02:19:02.859Z