# pm-jaube-prg-spamprobe.rc -- Interface to Spamprobe program
# $Id: pm-jaube-prg-spamprobe.rc,v 2.11 2004/10/18 06:58:37 jaalto Exp $
#
#   File id
#
#       .Copyright (C)  Jari Aalto
#       .$Keywords:     procmail, subroutine, UBE, spam, filter $
#
#       This code is free software in terms of GNU Gen. pub. Lic. v2 or later
#       Refer to http://www.gnu.org/copyleft/gpl.html
#
#   Warning
#
#       Put all your Unsolicited Bulk Emacil (aka spam) filters towards the
#       end of your `~/.procmailrc'. The idea is that valid messages are filed
#       first (mailing lists, your work and private mail, bounces) and only
#       the uncategorized messages are checked last.
#
#       YOU CANNOT USE THIS PROCMAIL SUBROUTINE UNLESS YOU HAVE TRAINED THE
#       BAYESIAN PROGRAM FIRST!
#
#       To train:
#
#           $ spamprobe -8 good good.msg ...
#           $ spamprobe -8 spam spam.msg ...
#
#	Make sure there are no stale lock files, or the spamprobe and this
#	subroutine will hang infinitely:
#
#	    $ rm -f ~/.spamprobe/lock 
#
#   Overview of features
#
#       o   Implements interface to http://freshmeat.net/projects/spamprobe/
#           project.
#       o   variable `ERROR' is set if the message was UBE.
#       o   Results are available by default in header 
#           `X-Spam-Spamprobe-Status'.
#
#   Description
#
#       There are several Bayesian based statistical analysis programs that
#       study the message's tokens and then classify it into two categories:
#       good or bad, or if you like, ham and spam. All the Bayesian programs
#       are not the same, so if you want to achive magic 99.99% probability
#       the only methodology to do that is to chain several programs in
#       serially. There is no single program that can solve the UBE detection.
#       This procmail subroutine implements call interface to program
#       `spamprobe', which must already have been installed.
#
#   About bouncing message back
#
#       The general consensus is, that you should not send bounces. The UBE
#       sender is not there, because the address forged. Do not increase
#       the network traffic; you will not do any good to anybody by
#       bouncing messgas -- you just increase mail traffic even more.
#       Instead save the messages to folders and periodically periodically
#       check their contents.
#
#   Required settings
#
#       If `spamprobe' program is available, define this variable in your
#       `~/.procmailrc'. Use absolute path to make the external shell
#       quick; it'll save server load considerably.
#
#           JA_UBE_SPAMPROBE_PRG = /usr/bin/spamprobe
#
#       If you _do_ _not_ have program installed, do not leave the
#       variable lying aroung, because it will keep this subroutine active.
#       Calling a non existing program is not a good idea, so it better to
#       empty the variable if the program is not available.
#
#   Required settings
#
#       None. No dependencies to other procmail modules.
#
#   Call arguments (variables to set before calling)
#
#       o   `JA_UBE_SPAMPROBE_PRG', path to the program.
#       o   `JA_UBE_SPAMPROBE_HEADER', the header name where the results
#           are put. If not defined, no header is added.
#           Defaults to *X-Spam-Spamprobe-Status*
#       o   `JA_UBE_SPAMPROBE_FORCE', if set to _yes_ then call program no matter
#           what. Normally if there already is *X-Spam-Spamprobe-Status* header,
#           it is assumed that the message has already been checked
#           and no new checking is needed.
#
#   Return values
#
#       o   `ERROR', is set to the return value of `spamprobe' program.
#
#   Usage example
#
#           PMSRC                  = $HOME/procmail # procmail recipe dir
#
#           <other checks, mailing lists, work mail etc.>
#
#           JA_UBE_SPAMPROBE_PRG   = "/usr/bin/nice -n 5 /usr/bin/spamprobe"
#           INCLUDERC              = $PMSRC/pm-jaube-prg-spamprobe.rc
#
#           #   The ERROR will contains word "yes" if message was spam
#
#           :0 :
#           * ERROR ?? yes
#           junk.mbox
#
#   File layout
#
#       The layout of this file is managed by Emacs packages tinyprocmail.el
#       and tinytab.el for the 4 tab text placement.
#       See project http://tiny-tools.sourceforge.net/
#
#   Change Log
#
#       None

dummy = "
========================================================================
pm-jaube-prg-spamprobe.rc: init:"

# ................................................... User variables ...

#   You must define program path, because we don't know
#   if it has been installed in this system or not

JA_UBE_SPAMPROBE_PRG           = ${JA_UBE_SPAMPROBE_PRG:-""}

#   Command "score" examines the message
#   You could use command "train" is you have good database.

JA_UBE_SPAMPROBE_PRG_OPT       = ${JA_UBE_SPAMPROBE_PRG_OPT:-"score"}

#   The header name with no colon at the end. If this variable
#   is empty, then external formail call is saved - results can still
#   be checked from variable ERROR. There is no trailing colon in this
#   variable.

JA_UBE_SPAMPROBE_HEADER = ${JA_UBE_SPAMPROBE_HEADER:-"\
X-Spam-Spamprobe-Status"}

# Should we check even if there already was header
# JA_UBE_SPAMPROBE_HEADER

JA_UBE_SPAMPROBE_FORCE = ${JA_UBE_SPAMPROBE_FORCE:-"no"}

# ............................................................ do it ...

ERROR   # Killl variable

:0
*                   JA_UBE_SPAMPROBE_PRG     ??  [a-z]
*$ ! 9876543210^0  ^$JA_UBE_SPAMPROBE_HEADER
*    9876543210^0   JA_UBE_SPAMPROBE_FORCE   ?? yes
{
    :0 w : spamprobe$LOCKEXT
    jaubeSpamprobeScore=|$JA_UBE_SPAMPROBE_PRG ${SPAMPROBE_OPT:-"score"}

    ERROR = $jaubeSpamprobeScore

    # X-Spam-Spamprobe-Status: GOOD 0.0000273 e5af974acad8028bea8a10f1f08b966e
    # X-Spam-Spamprobe-Status: SPAM 0.9742932 07a3b71c9c96968641bbfa497a458d8e

    :0 fw
    * ! jaubeSpamprobeScore     ?? ^^^^
    * ! JA_UBE_SPAMPROBE_HEADER ?? ^^^^
    | ${FORMAIL:-formail} -I "$JA_UBE_SPAMPROBE_HEADER: $jaubeSpamprobeScore"
}

dummy = "pm-jaube-prg-spamprobe.rc: end: $ERROR"

# End of file