Marco,
> > That used to be the case for early integration of dspam into amavisd,
> > but has been changed with the introduction of @spam_scanners list with
> > amavisd-new-2.6.3. See release notes, searching for '@spam_scanners'.
> >
> > Now the dspam or crm114 header fields are only inserted into a
> > passed or quarantined message, but are not visible from one
> > spam scanner to another.
>
> Hi Mark,
> thank you, now what happens it is clear.
>
> Maybe, using spamassassin rules I could make more complexes
> evaluations, as suggested by dspam.pm plugin:
> http://eric.lubow.org/projects/dspam-spamassassin-plugin/
>
> For example, I can combine confidence and probability in spam score.
Will make available these headers fields to SA with the next release
of amavisd-new, thanks for the suggestion:
- if running other spam scanners besides SpamAssassin through a @spam_scanners
mechanism (such as DSPAM or CRM114), make header fields produced by them
visible to SpamAssassin too, so that its rules can benefit from additional
information. Note that in order for SpamAssassin to be able to see such
header fields from other scanners, they must be listed in the @spam_scanners
list *before* the 'SpamAssassin' entry.
> I've written little patch for amavisd 2.6.4:
>
> 21621,21626c21621
> < # $spam_score = $dspam_result eq 'Spam' ? 10 : -1; # fabricated
> < my $dspam_prob = $header_field{lc('X-DSPAM-Probability')};
> < my $dspam_conf = $header_field{lc('X-DSPAM-Confidence')};
> < chomp($dspam_prob); chomp($dspam_conf);
> < $spam_score = ($dspam_prob-0.5)*$dspam_conf*20;
> < if ($header_field{lc('X-DSPAM-Class')} eq 'Whitelisted')
> {$spam_score = -10;}
> ---
>
> > $spam_score = $dspam_result eq 'Spam' ? 10 : -1; # fabricated
>
> This is what I mean.
> It could use a more complicated spread function on spam score, above
> spam score is equally likely from -10 to 10.
>
> Can I add above code on my amavisd.conf without modify official amavisd
> code?
No, you can't have it in amavisd.conf or in a custom hook.
For a complex formula you'd still need to hack the code.
For simpler cases (like the 'Whitelisted' above, or based on
a few bucketized results, this could be done with SA rules.
Mark
------------------------------------------------------------------------------
Download Intel® Parallel Studio Eval
Try the new software tools for yourself. Speed compiling, find bugs
proactively, and fine-tune applications for parallel performance.
See why Intel Parallel Studio got high marks during beta.
http://p.sf.net/sfu/intel-sw-dev
_______________________________________________
AMaViS-user mailing list
AMaViS-user@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/amavis-user
AMaViS-FAQ:http://www.amavis.org/amavis-faq.php3
AMaViS-HowTos:http://www.amavis.org/howto/
