For any of you who are wondering what all this does - I'm gathering up a
bunch of information about DNS relating to the message and adding it in
a header. I then run it through spam assassin and this extra information
gives the bayesian filter more facts to chew on. I then remove the
header coming out of the filter. The more info you feed bayes - the
better it works.
David S. Madole wrote:
> From: "Marc Perkel" <marc@???>
>
> # cat test
> #Exim Filter
>
> testprint "${sg{192x168x231x193}\
> {\\N([0-9]{1,3})\.([0-9]{1,3})\.([0-9]{1,3}\)\.([0-9]{1,3}\)\\N}\
> {\\N$4.$3.$2.$1.in-addr.arpa\\N}}"
>
> # exim -bf test < /dev/null
> Testprint: 193.231.168.192.in-addr.arpa
>
> David
>
>