Fast ssdeep comparison library
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Permalink
Failed to load latest commit information.
examples/internal
ext
m4
.gitattributes
.gitignore
COPYING
COPYING.Boost
COPYING.GPLv2
Doxyfile.example
Makefile.am
NEWS
README
bootstrap.sh
configure.ac
ffuzzy.h
ffuzzy_blocksize.c
ffuzzy_blocksize.h
ffuzzy_compare.c
ffuzzy_digest.c
ffuzzy_digest_conv.c
ffuzzy_digest_unnorm.c
ffuzzy_parse.c
ffuzzy_parse.h
ffuzzy_parse_unnorm.c
str_base64.h
str_common_substr.h
str_edit_dist.h
str_hash_rolling.h
util.h

README

libffuzzy : Fast ssdeep comparison library
===========================================

libffuzzy is a fuzzy hash comparison library compatible with
ssdeep (<http://ssdeep.sourceforge.net/>) version 2.10.

This library is designed to be fast and thread-safe. It does not even
allocate memory at run time (which may increase performance
on parallel computation).

The another purpose to write this library is to find implementation
issues in ssdeep. During this re-implementation, the author found
several issues which may affect robustness and portability of ssdeep.


Installation
-------------

If you don't have `configure` script (e.g. on git work-tree),
run `./bootstrap.sh` or `autoreconf -i` first.
Then you can install the program by `./configure`, `make`
and then `make install`.


Functions
----------

Global functions are named like ffuzzy_SOMETHING.
You can use ffuzzy_compare function just like fuzzy_compare.

You can also hold ssdeep digest after parsing.
Use ffuzzy_read_digest to parse digest and use ffuzzy_compare_digest
to compare against another digest (after parsing).


Performance
------------

Version 1.1 achieved expected performance gain.
libffuzzy is now faster than libfuzzy in most cases.

In version 2.0, interfaces for specialized comparison and clustering
are added. Clustering applications will get faster if you use
this optimized implementation correctly.

In the large clustering job (about 100k hashes),
libffuzzy 2.0 was about 60% faster than libfuzzy 2.11.