Search for: All records

For a length n text over an alphabet of size σ, we can encode the suffix tree data structure in 𝒪(nlog σ) bits of space. It supports suffix array (SA), inverse suffix array (ISA), and longest common extension (LCE) queries in 𝒪(log^ε_σ n) time, which enables efficient pattern matching; here ε > 0 is an arbitrarily small constant. Further improvements are possible for LCE queries, where 𝒪(1) time queries can be achieved using an index of space 𝒪(nlog σ) bits. However, compactly indexing a two-dimensional text (i.e., an n× n matrix) has been a major open problem. We show progress in this direction by first presenting an 𝒪(n²log σ)-bit structure supporting LCE queries in near 𝒪((log_σ n)^{2/3}) time. We then present an 𝒪(n²log σ + n²log log n)-bit structure supporting ISA queries in near 𝒪(log n ⋅ (log_σ n)^{2/3}) time. Within a similar space, achieving SA queries in poly-logarithmic (even strongly sub-linear) time is a significant challenge. However, our 𝒪(n²log σ + n²log log n)-bit structure can support SA queries in 𝒪(n²/(σ log n)^c) time, where c is an arbitrarily large constant, which enables pattern matching in time faster than what is possible without preprocessing. We then design a repetition-aware data structure. The δ_2D compressibility measure for two-dimensional texts was recently introduced by Carfagna and Manzini [SPIRE 2023]. The measure ranges from 1 to n², with smaller δ_2D indicating a highly compressible two-dimensional text. The current data structure utilizing δ_2D allows only element access. We obtain the first structure based on δ_2D for LCE queries. It takes 𝒪^{~}(n^{5/3} + n^{8/5}δ_2D^{1/5}) space and answers queries in 𝒪(log n) time. 

We study the problem of indexing a text T[1..n] to support pattern matching with wildcards. The input of a query is a pattern P[1..m] containing h ∈ [0, k] wildcard (a.k.a. don't care) characters and the output is the set of occurrences of P in T (i.e., starting positions of substrings of T that matches P), where k = o(log n) is fixed at index construction. A classic solution by Cole et al. [STOC 2004] provides an index with space complexity O(n ⋅ (clog n)^k/k!)) and query time O(m+2^h log log n+occ), where c > 1 is a constant, and occ denotes the number of occurrences of P in T. We introduce a new data structure that significantly reduces space usage for highly repetitive texts while maintaining efficient query processing. Its space (in words) and query time are as follows: O(δ log (n/δ)⋅ c^k (1+(log^k (δ log n))/k!)) and O((m+2^h +occ)log n)) The parameter δ, known as substring complexity, is a recently introduced measure of repetitiveness that serves as a unifying and lower-bounding metric for several popular measures, including the number of phrases in the LZ77 factorization (denoted by z) and the number of runs in the Burrows-Wheeler Transform (denoted by r). Moreover, O(δ log (n/δ)) represents the optimal space required to encode the data in terms of n and δ, helping us see how close our space is to the minimum required. In another trade-off, we match the query time of Cole et al.’s index using O(n+δ log (n/δ) ⋅ (clogδ)^{k+ε}/k!) space, where ε > 0 is an arbitrarily small constant. We also demonstrate how these techniques can be applied to a more general indexing problem, where the query pattern includes k-gaps (a gap can be interpreted as a contiguous sequence of wildcard characters). 

We revisit the following version of the Gapped String Indexing problem, where the goal is to preprocess a text T[1..n] to enable efficient reporting of all occ occurrences of a gapped pattern P=P1[α..β]P2 in T. An occurrence of P in T is defined as a pair (i,j) where substrings T[i..i+|P1|) and T[j..j+|P2|) match P1 and P2, respectively, with a gap j−(i+|P1|) lying within the interval [α..β]. This problem has significant applications in computational biology and text mining. A hardness result on this problem suggests that any index with polylogarithmic query time must occupy near quadratic space. In a recent study [STACS 2024], Bille et al. presented a sub-quadratic space index using space O˜(n2−δ/3), where 0≤δ≤1 is a parameter fixed at the time of index construction. Its query time is O˜(|P1|+|P2|+nδ·(1+occ)), which is sub-linear per occurrence when δ<1. We show how to achieve a gap-sensitive query time of O˜(|P1|+|P2|+nδ·(1+occ1−δ)+∑g∈[α..β]occg·gδ) using the same space, where occg denotes the number of occurrences with gap g. This is faster when there are many occurrences with small gaps.