irgx 2.1.1

Linear-time regex engine for Rust - no catastrophic backtracking, no ReDoS - plus the shared analytic substrate (row protocol, transports, contracts).
Documentation 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55

irregex — the irregular-expression engine
Copyright 2026 The Billy Company

This product includes software developed at The Billy Company
(https://billylives.com/).

This product bundles the following third-party components, which carry their
own licenses and copyright notices. Those notices are reproduced with the
component and are not superseded by the Apache License, Version 2.0 above.

  * PCRE2 — Perl-Compatible Regular Expressions (vendor/pcre2/)
    BSD-3-Clause WITH PCRE2-exception
    Copyright (c) 1997-2024 University of Cambridge
    Copyright (c) 2009-2024 Zoltan Herczeg
    The backend behind `-P`. See vendor/pcre2/LICENCE.md for the full text.

  * libsais — linear-time suffix array construction (vendor/libsais/)
    Apache-2.0
    Copyright (c) 2021-2025 Ilya Grebnov
    The suffix sort under the FM-index. See vendor/libsais/LICENSE for the
    full text.

  * WHATWG Encoding Standard index tables (tools/whatwg/index-*.txt,
    tools/whatwg/encodings.json)
    CC0-1.0 (public-domain dedication)
    Sourced from https://encoding.spec.whatwg.org/ — used to generate the
    legacy-encoding decode tables.

  * Unicode Character Database, version 16.0.0 (tools/ucd/*.txt)
    Unicode License v3
    Copyright (c) 2024 Unicode, Inc. Unicode and the Unicode Logo are
    registered trademarks of Unicode, Inc. in the U.S. and other countries.
    Pinned upstream data files from https://www.unicode.org/Public/16.0.0/ucd/
    (emoji-data.txt from .../ucd/emoji/, DerivedGeneralCategory.txt from
    .../ucd/extracted/) — used to generate the Unicode property, case-folding,
    script, emoji-property, and property-alias tables.
    See tools/ucd/LICENSE.txt for the full copyright and permission notice, and
    tools/ucd/README.md for the per-file pin and what each one lowers into.

  * ripgrep integration-test corpus, mined
    (the exact face package's bench/conformance/rgsuite/spec.json)
    Unlicense OR MIT
    Copyright (c) 2015 Andrew Gallant
    Test names, arguments, and haystack bytes mechanically extracted from
    ripgrep's own `rgtest!` suite (https://github.com/BurntSushi/ripgrep) so
    that ripgrep — not a hardcoded expectation — is the correctness oracle for
    this engine's `rg` drop-in parity. Some haystacks are Project Gutenberg
    public-domain texts carried through from that suite. The implementation is
    an independent reimplementation; only this test corpus is derived.

Algorithms and interface conventions implemented from published descriptions
rather than from borrowed source — among them Aho–Corasick (1975), Hyperscan's
Teddy prefilter, Cox's trigram-index construction, and ripgrep's CLI and
ignore-file conventions — carry no obligation here. They are credited where
they are used, in the module headers and in research/*/PRIOR_ART.md.