sa-learn, --remove-markup and mbox format

[Sam Halliday <fo...@yahoo.ie>]

hi there,

i have 2 questions:

 - when running `sa-learn` on an mbox file, must i first strip all
   the spamassassin markups? or does it deal with those automatically?

 - is there any way i can run `spamassassin --remove-markup` on an mbox
   file? when i run it on an mbox file, it treats it as a single email.
   there is no --mbox flag for spamassassin (just for sa-learn). a lot of
   people on the web seem to think that it will work on an mbox file,
   but from personal eperience, i know it does not.

thanks in advance!

cheers,
Sam
-- 
Free High School Science Texts
  http://savannah.nongnu.org/projects/fhsst
Sam's Homepages
  http://fommil.homeunix.org/~samuel
  http://www.ma.hw.ac.uk/~samuel

Re: sa-learn, --remove-markup and mbox format

[Bob George <ma...@ttlexceeded.com>]

Sam Halliday <fo...@yahoo.ie> wrote:
> >  - when running `sa-learn` on an mbox file, must i first strip
>    all the spamassassin markups? or does it deal with those
> automatically?

Per the manpage: Messages learnt as spam will have SpamAssassin markup removed,
on the fly.

>  - is there any way i can run `spamassassin --remove-markup`
>    on an mbox file? when i run it on an mbox file, it treats
>    it as a single email. there is no --mbox flag for
>    spamassassin (just for sa-learn). a lot of people on the
>    web seem to think that it will work on an mbox file, but
>    from personal eperience, i know it does not.

I've written a generic filter script (cleanspam) that calls
spamassassin --remove-markup, then pipes the message through formail to remove
locally-added headers:

cat - | spamassassin -d | formail \
-I X-Local-Header1: \
[... other unwanted headers ...]
-I X-Local-Header2:

I routinely do things like:

mv mbox mbox.old
formail -s cleanspam < mbox.old >> mbox

It seems to work very well.

- Bob