International Journal of Foundations of Computer Science

Černý's conjecture asserts the existence of a synchronizing word of length at most (n - 1)² for any synchronized n-state deterministic automaton. We prove a quadratic upper bound on the length of a synchronizing word for any synchronized n-state deterministic automaton satisfying the following additional property: there is a letter a such that for any pair of states p, q, one has p·a^r = q·a^s for some integers r, s (for a state p and a word w, we denote by p·w the state reached from p by the path labeled w). As a consequence, we show that for any finite synchronized prefix code with an n-state decoder, there is a synchronizing word of length O(n²). This applies in particular to Huffman codes.

We study the problem of testing whether a context-free language is included in a fixed set L₀, where L₀ is the language of words reducing to the empty word in the monoid defined by a complete string rewrite system. We prove that, if the monoid is cancellative, then our inclusion problem is polynomially reducible to the problem of testing equivalence of straight-line programs in the same monoid. As an application, we obtain a polynomial time algorithm for testing if a context-free language is included in a Dyck language (the best previous algorithm for this problem was doubly exponential).

A famous theorem of Kuratowski states that, in a topological space, at most 14 distinct sets can be produced by repeatedly applying the operations of closure and complement to a given set. We re-examine this theorem in the setting of formal languages, where by "closure" we mean either Kleene closure or positive closure. We classify languages according to the structure of the algebras they generate under iterations of complement and closure. There are precisely 9 such algebras in the case of positive closure, and 12 in the case of Kleene closure. We study how the properties of being open and closed are preserved under concatenation. We investigate analogues, in formal languages, of the separation axioms in topological spaces; one of our main results is that there is a clopen partition separating two words if and only if the words do not commute. We can decide in quadratic time if the language specified by a DFA is closed, but if the language is specified by an NFA, the problem is PSPACE-complete.

In this paper we prove that it is decidable whether the set pow(L), which we get by taking all the powers of all the words in some regular language L, is regular or not. The problem was originally posed by Calbrix and Nivat in 1995. Partial solutions have been given by Cachat for unary languages and by Horváth et al. for various kinds of exponent sets for the powers and regular languages which have primitive roots satisfying certain properties. We show that the regular languages which have a regular power are the ones which are 'almost' equal to their Kleene-closure.

A number α, in the range from n to 2ⁿ, is magic for n with respect to a given alphabet size s, if there is no minimal nondeterministic finite automaton of n states and s input symbols whose equivalent minimal deterministic finite automaton has α states. We show that in the case of a ternary alphabet, there are no magic numbers. For all n and α satisfying n ⩽ α ⩽ 2ⁿ, we define an n-state nondeterministic finite automaton with a three-letter input alphabet that requires exactly α deterministic states.

We consider properties of the solution set of a word equation with one unknown. We prove that the solution set of a word equation possessing infinite number of solutions is of the form (pq)*p where pq is primitive. Next, we prove that a word equation with at most four occurrences of the unknown possesses either infinitely many solutions or at most two solutions. We show that there are equations with at most four occurrences of the unknown possessing exactly two solutions. Finally, we prove that a word equation with at most 2k occurrences of the unknown possesses either infinitely many solutions or at most 8 log k + O(1) solutions. Hence, if we consider a class ε_k of equations with at most 2k occurrences of the unknown, then each equation in this class possesses either infinitely many solutions or O(log k) number of solutions. Our considerations allow to construct the first alphabet independent linear time algorithm for computing the solution set of an equation in a nontrivial class of equations.

Systems of equations of the form X = Y + Z and X = C, in which the unknowns are sets of natural numbers, "+" denotes elementwise sum of sets S + T = {m + n | m ∈ S, n ∈ T}, and C is an ultimately periodic constant, have recently been proved to be computationally universal (Jeż, Okhotin, "Equations over sets of natural numbers with addition only", STACS 2009). This paper establishes some limitations of such systems. A class of sets of numbers that cannot be represented by unique, least or greatest solutions of systems of this form is defined, and a particular set in this class is constructed. The argument is then extended to equations over sets of integers.

We investigate the efficiency of simulations of storages by several counters. A simulation of a pushdown store is described which is optimal in the sense that reducing the number of counters of a simulator leads to an increase in time complexity. The lower bound also establishes a tight counter hierarchy in exponential time.

Then we turn to simulations of a set of counters by a different number of counters. We improve and generalize a known simulation in polynomial time. Greibach has shown that adding s + 1 counters increases the power of machines working in time n^s. Using a new family of languages we show here a tight hierarchy result for machines with the same polynomial time-bound. We also prove hierarchies for machines with a fixed number of counters and with growing polynomial time-bounds.

For machines with one counter and an additional "store zero" instruction we establish the equivalence of real-time and linear time. If at least two counters are available, the classes of languages accepted in real-time and linear time can be separated.

This article presents approaches to the open problem of whether erasing rules can be eliminated in matrix grammars. The class of languages generated by non-erasing matrix grammars is characterized by the newly introduced linear Petri net grammars. Petri net grammars are known to be equivalent to arbitrary matrix grammars (without appearance checking). In linear Petri net grammars, the marking has to be linear in size with respect to the length of the sentential form. The characterization by linear Petri net grammars is then used to show that applying linear erasing to a Petri net language yields a language generated by a non-erasing matrix grammar. It is also shown that in Petri net grammars (with final markings and arbitrary labeling), erasing rules can be eliminated, which yields two reformulations of the problem of whether erasing rules in matrix grammars can be eliminated.

A bisemigroup is a set with two associative operations. Subsets of free bisemigroups are called bisemigroup languages. Recognizable, regular and MSO-definable bisemigroup languages have been studied earlier, and these classes are known to be equal. In this paper we prove a Kleene theorem for bisemigroup languages, namely we show that the class of recognizable bisemigroup languages is the least class which contains the finite languages and closed under the operations of union, horizontal and vertical product, horizontal and vertical iteration, ξ-substitution and a restricted version of the the ξ-iteration. We extend our result to binoid languages, i.e., to subsets of free algebras, where the two associative operations share a common identity element.

A pseudo-primitive word with respect to an antimorphic involution θ is a word which cannot be written as a catenation of occurrences of a strictly shorter word t and θ(t). Properties of pseudo-primitive words are investigated in this paper. These properties link pseudo-primitive words with essential notions in combinatorics on words such as primitive words, (pseudo)-palindromes, and (pseudo)-commutativity. Their applications include an improved solution to the extended Lyndon-Schützenberger equation u₁u₂ ⋯ u_ℓ = v₁ ⋯ v_nw₁ ⋯ w_m, where u₁, … , u_ℓ ∈ {u,θ(u)}, v₁, …, v_n ∈ {v, θ(v)}, and w₁, … , w_m ∈ {w, θ(w)} for some words u, v, w, integers ℓ, n, m ≥ 2, and an antimorphic involution θ. We prove that for ℓ ≥ 4, n, m ≥ 3, this equation implies that u, v, w can be expressed in terms of a common word t and its image θ(t). Moreover, several cases of this equation where ℓ = 3 are examined.

An instance of the (Generalized) Post Correspondence Problem is during the decision process typically reduced to one or more other instances, called its successors. In this paper we study the reduction tree of GPCP in the binary case. This entails in particular a detailed analysis of the structure of end blocks. We give an upper bound for the number of end blocks, and show that even if an instance has more than one successor, it can nevertheless be reduced to a single instance. This, in particular, implies that binary GPCP can be decided in polynomial time.

An algebraic linear ordering is a component of the initial solution of a first-order recursion scheme over the continuous categorical algebra of countable linear orderings equipped with the sum operation and the constant 1. Due to a general Mezei-Wright type result, algebraic linear orderings are exactly those isomorphic to the linear ordering of the leaves of an algebraic tree. Using Courcelle's characterization of algebraic trees, we obtain the fact that a linear ordering is algebraic if and only if it can be represented as the lexicographic ordering of a deterministic context-free language. When the algebraic linear ordering is a well-ordering, its order type is an algebraic ordinal. We prove that the Hausdorff rank of any scattered algebraic linear ordering is less than ω^ω. It follows that the algebraic ordinals are exactly those less than ω^ωω.