[hashcash] New procmail recipe

  • From: "Todd A. Jacobs" <nospam@xxxxxxxxxxxxx>
  • To: hashcash@xxxxxxxxxxxxx
  • Date: Tue, 14 Nov 2006 04:06:08 -0800

# Capture the hostname, and escape any periods so the value can be used
# in regular expressions.
MYDOMAIN=`hostname -d | sed 's%\.%\\.%g'`

# Process hashcash-stamped emails. Don't just validate it; include the
# results in the email headers. This is partly for debugging, but it
# also provides more tokens for Bayesian filters such as bogofilter to
# use in differentiating spam vs. ham.
* $^X-Hashcash: .*\/[^:]+@${MYDOMAIN}
    # Hashcash sends its output to stderr, so we capture it and then
    # ensure that every line but the first is indented in order to
    # comply with RFC 822.
    RESULT=`hashcash -cvMXdp 1d -b20 -f $HOME/hashcash.db -r $MATCH 2>&1 | \
            sed -e '1!s/^/\t\t   /'`

    # Filter successful mail through formail, and append the results.
    | formail -i "X-Hashcash-Validated: yes" \
      -i "X-Hashcash-Result: $RESULT"

    # If the hashcash stamp is invalid, append the results and let further
    # filtering take place.
    | formail -i "X-Hashcash-Validated: no" \
      -i "X-Hashcash-Result: $RESULT"

    # Deliver the mail, bypassing further filtering. Disable this if you
    # still want to do further processing (e.g. Bayesian filtering)
    # before delivery.

Unabashedly littering the information superhighway with detritus like
this for over 15 years now.

Other related posts:

  • » [hashcash] New procmail recipe