[Gllug] How do I do this?

[Nix]

On Thu, 20 Jan 2005, Adrian McMenamin uttered the following:
> On Mon, 2005-01-17 at 18:16 +0000, Martin A. Brooks wrote:
> 
>> The script works by going through every message in the designated IMAP 
>> folder, retrieving it and then feeding the message to sa-learn with "no 
>> sync" set. Once all of the messages have been learned if then deletes 
>> all the messages, one by one, and finally syncs the bayesian db.
>> 
> It is this aspect - one file at a time - that makes it run slow.

Yes, definitely. The perl startup delay is nasty for massive runs
like that. That's why it accepts more than one filename. :)

Note that as of SA 3.0, you can feed whole nnmh-style directory names to
`spamassassin -r', as well, which can be useful if like me you
mass-report hundreds of spams a day. (You can feed maildir directories
and Gnus nnml directories to it, as well; it skips files starting with a
dot to let the mailers save metadata there, as nnml does.)

-- 
`Blish is clearly in love with language. Unfortunately,
 language dislikes him intensely.' --- Russ Allbery
-- 
Gllug mailing list  -  Gllug at gllug.org.uk
http://lists.gllug.org.uk/mailman/listinfo/gllug