https://github.com/mischasan/aho-corasick

A-C implementation in "C". Tight-packed (interleaved) state-transition matrix -- as fast as it gets, as small as it gets.
https://github.com/mischasan/aho-corasick

Last synced: about 2 months ago
JSON representation

A-C implementation in "C". Tight-packed (interleaved) state-transition matrix -- as fast as it gets, as small as it gets.

• Host: GitHub
• URL: https://github.com/mischasan/aho-corasick
• Owner: mischasan
• License: lgpl-3.0
• Created: 2013-08-28T19:31:00.000Z (about 11 years ago)
• Default Branch: master
• Last Pushed: 2021-01-16T00:31:07.000Z (over 3 years ago)
• Last Synced: 2024-07-23T10:02:02.814Z (2 months ago)
• Language: C
• Homepage:
• Size: 1.72 MB
• Stars: 145
• Watchers: 17
• Forks: 40
• Open Issues: 19
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

README

        
aho-corasick

==

Aho-Corasick parallel string search, using interleaved arrays.

Mischa Sandberg [email protected]

ACISM is an implementation of Aho-Corasick parallel string search,

using an Interleaved State-transition Matrix.

It combines the fastest possible Aho-Corasick implementation,

with the smallest possible data structure (!).

FEATURES

--------

* Fast. No hashing, no tree traversal; just a straight look-up equivalent to

    matrix[state, input-byte] per input character.

* Tiny. On average, the whole data structure (mostly the array) takes about 2-3 bytes per

    input pattern byte. The original set of pattern strings can be reverse-generated from the machine.

* Shareable. The state machine contains no pointers, so it can be compiled once,

    then memory-mapped by many processes.

* Searches byte vectors, not null-terminated strings.

    Suitable for searching machine code as much as searching text.

* DOS-proof. Well, that's an attribute of Aho-Corasick,

    so no real points for that.

* Stream-ready. The state can be saved between calls to search data.

DOCUMENTATION

-------------

The GoogleDocs description is at http://goo.gl/lE6zG

I originally called it "psearch", but found that name was overused by other authors.

LICENSE

-------

LGPL v3

GETTING STARTED

---------------

Download the source, type "gmake".

"gmake install" exports lib/libacism.a, include/acism.h and bin/acism_x.

"acism_x.c" is a good example of calling acism_create and acism_scan/acism_more.

(If you're interested in the GNUmakefile and rules.mk,

 check my blog posts on non-recursive make, at mischasan.wordpress.com.)

 

HISTORY

-------

The interleaved-array approach was tried and discarded in the late 70's, because the compile time was O(n^2). 

acism_create beats the problem with a "hint" array that tracks the restart points for searches.

That, plus discarding the original idea of how to get maximal density, resulted in the tiny-fast win-win.

ACKNOWLEDGEMENTS

----------------

I'd like to thank Mike Shannon, who wanted to see a machine built to make best use of L1/L2 cache.

The change to do that doubled performance on hardware with a much larger cache than the matrix.

Go figure.