A Benchmark Production Tool for Regular Expressions

Abstract

material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed. The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use. The publisher, the authors and the editors are safe to assume that the advice and information in this book are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or the editors give a warranty, expressed or implied, with respect to the material contained herein or for any errors or omissions that may have been made. The publisher remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

YARB: a Methodology to Characterize Regular Expression Matching on Heterogeneous Systems

“…For instance, Becchi et al devised a REs generator called Regex Processor (RP) [27], but it is limited to Deep-Packet Inspection (DPI). Borsotti et al proposed an interesting benchmark generator called REgen to evaluate advanced and ambiguous REs [28], but they evaluate mainly the tool performance for ambiguity rather than execution engines. Wadden et al devised ANMLZoo [23], a multidomain benchmark suite to address the literature gap in benchmarking.…”

“…These suites are composed of REs and NFA represented in ANML (MNRL) ADL(s), along with different sizes of input stimuli. We natively integrate REgen [28] as a possible source of benchmarks in YARB. We slightly modify REgen to decouple REs from the input data and use its output immediately.…”

“…On the one hand, researchers focused on RE generators for specific domains, such as intrusion detection systems [27], which lack generality. On the other hand, utterly general-purpose benchmark generators may focus on purely theoretical aspects without strictly applying any real-case scenario [28]. The absence of a transparent approach to evaluate the REs domain hinders a clear evaluation of past, present, and future engines.…”

“…We designed YARB to be modular and extensible by others. YARB can exploit state-of-the-art REs and data generators [27], [28] along with proposed literature ready-to-use benchmarks [23], [24]. Since we adopt REs as universal language, we devise a novel Automata to REs translator that can optionally apply different minimization algorithms and explore solutions' performance.…”