Man page for UNIXCRYPT-BREAKER(1)

NAME

SYNOPSIS

unixcrypt-breaker [options] -s statfile -c corpusfile...
unixcrypt-breaker [options] -s statfile cipherfile...

DESCRIPTION

unixcrypt-breaker is an automated tool for breaking the encryption of the old unix crypt(1) utility. It can guess substantial portions of the plaintext without previous knowledge of the key.

It is well-known that unix crypt provides only weak security. Existing software for breaking it includes the Crypt Breaker's Workbench (http://axion.physics.ubc.ca/cbw.html). Unlike the Crypt Breaker's Workbench, unixcrypt-breaker is fully automatic and does not require user interaction. Rather than relying on the user to guess likely plaintexts, unixcrypt-breaker uses a statistical model for guessing plausible plaintexts.

To guess the plaintext correctly, unixcrypt-breaker must first be primed with a body of sample data (the "corpus") that is presumed to be similar to the unknown plaintext. For example, if the plaintext is assumed to be HTML data, then unixcrypt-breaker should be primed with a corpus of HTML data.

Note that unixcrypt-breaker can be used to extract a partial plaintext, but cannot currently be used to extract the actual encryption key used.

OPERATION

unixcrypt-breaker operates in two distinct steps: sampling and decryption. During the sampling step (specified with the -c option, corresponding to the first form of usage above), data is read from a corpus of one or more sample plaintext files, and the resulting statistics are written to a file specified with the -s option. This file is then used during the decryption step (the second form of usage above) to partially decrypt one or more ciphertext files. If more than one file is specified for decryption, they must have been encrypted with the same key.

It is important to note that unixcrypt-breaker only produces an approximation to the plaintext. The percentage of bytes correctly guessed depends on many factors, such as the file type, and the quality of the corpus of sample data. Performance will be best for file contents that have high redundancy (such as text).

USAGE EXAMPLE

Suppose file.cpt is a file encrypted with crypt(1), and plain1, plain2, plain3 are large representative plaintext files of a similar type. First, sample the plaintext files using the command

    unixcrypt-breaker -s /tmp/stats.txt -c plain1 plain2 plain3

This will write some statistical information about the plaintext files to the file /tmp/stats.txt. Then, decrypt file.cpt with the following command:

    unixcrypt-breaker -s /tmp/stats.txt file.cpt

The output is written to file.cpt.guess.

OPTIONS

-h, --help

Help. Print usage information and exit.

-V, --version

Print version info and exit.

-L, --license

Print license info and exit.

-c, --corpus

Sampling mode: generate statistics from a corpus of sample data. The corpus is read from one or more filenames specified on the command line, or from stdin if no files are specified. The statistics are written to a file specified with the -s option. For best performance, the corpus should be as large as possible (several megabytes would be great), and should be as similar as possible to the likely plaintext.

-s file, --statfile file

Specify the file to which statistical data is written (in sampling mode) or from which it is read (in decryption mode). This option is mandatory.

-o file, --output file

Specify a file to write the plaintext guess to. The default is to append the extension ".guess" to the ciphertext file.

-d mode, --display mode

Specify the display mode for progress information. Possible choices are: color (default), mono, or none. The default is to show a color display that is compatible with color X-Terminals. mono selects a black-and-white display, and none displays no progress information at all.

-w file, --wirefile file

Output optional wiring information to a file. The wirings are internal to the crypt algorithm and are of interest to experts.

-m n, --maxdisplay n

Specify how many blocks can fit on the display. This is used when decrypting multiple files simultanesouly. Regardless of the number of files given on the command line, at most n blocks will be displayed simultaneously. The default is 1.

--

End of options. Any remaining arguments are interpreted as filenames.

VERSION

0.2

AUTHOR

Peter Selinger

COPYRIGHT

Copyright (C) 2006-2013 Peter Selinger

This program is free software; you can redistribute it and/or modify it under the terms of the GNU General Public License as published by the Free Software Foundation; either version 2 of the License, or (at your option) any later version.

This program is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for more details.

You should have received a copy of the GNU General Public License along with this program; if not, write to the Free Software Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307, USA. See also http://www.gnu.org/.