SD-Regular Transducer Expressions for Aperiodic Transformations

TL;DR

This paper introduces SD-regular transducer expressions (SDRTEs) that precisely characterize first order definable functions over finite words, extending classical results on aperiodic languages and providing a new formalism for aperiodic transformations.

Contribution

It generalizes the SF=AP and SD=AP results by defining SDRTEs that capture all first order definable functions via a novel restricted Kleene star construction.

Findings

SDRTEs characterize first order definable functions over finite words.

Aperiodic languages are captured by unambiguous, stabilising SD-regular expressions.

Constructive method for translating aperiodic two-way transducers into SDRTEs.

Abstract

FO transductions, aperiodic deterministic two-way transducers, as well as aperiodic streaming string transducers are all equivalent models for first order definable functions. In this paper, we solve the long standing open problem of expressions capturing first order definable functions, thereby generalizing the seminal SF=AP (star free expressions = aperiodic languages) result of Sch\"utzenberger. Our result also generalizes a lesser known characterization by Sch\"utzenberger of aperiodic languages by SD-regular expressions (SD=AP). We show that every first order definable function over finite words captured by an aperiodic deterministic two-way transducer can be described with an SD-regular transducer expression (SDRTE). An SDRTE is a regular expression where Kleene stars are used in a restricted way: they can appear only on aperiodic languages which are prefix codes of bounded synchronization delay. SDRTEs are constructed from simple functions using the combinators unambiguous sum (deterministic choice), Hadamard product, and unambiguous versions of the Cauchy product and the k-chained Kleene-star, where the star is restricted as mentioned. In order to construct an SDRTE associated with an aperiodic deterministic two-way transducer, (i) we concretize Sch\"utzenberger's SD=AP result, by proving that aperiodic languages are captured by SD-regular expressions which are unambiguous and stabilising; (ii) by structural induction on the unambiguous, stabilising SD-regular expressions describing the domain of the transducer, we construct SDRTEs. Finally, we also look at various formalisms equivalent to SDRTEs which use the function composition, allowing to trade the k-chained star for a 1-star.