OK, I *really* wanted numerical comparison, but no one else seemed that
interested. After looking at the filter code, it looked like the code
*really* wanted to do a regex rather than a numerical comparison.
While I agree that your regex is pretty friendly, it's not something that I
can unleash on someone who doesn't even know what regex means.
So here is what I've come up with to help the real newbies along:
# provided to as a -do not touch- feature in an included file
macro SPAM_LEVEL(min,max) headers '^X-Spam-Level:\s\*{min,max}'
# in the users' filters (drop 20-100, confirm 3-19, ok 0-2)
SPAM_LEVEL(20,100) drop
SPAM_LEVEL(3,19) confirm
SPAM_LEVEL(0,2) ok
This is close to what I want, but I'd still really like to have more
granularity at the lower scores.
----- Original Message -----
From: "Ray Heasman" <[EMAIL PROTECTED]>
To: "Marcus Williams" <[EMAIL PROTECTED]>
Cc: <[EMAIL PROTECTED]>
Sent: Tuesday, October 28, 2003 10:48 AM
Subject: Re: Filtering based on the numerical value of a header
> On Tue, 2003-10-28 at 02:41, Marcus Williams wrote:
>
> > For spamassassin use, you dont need to sort on the score. You can get
> > close enough by using the method I suggest in:
> >
> > http://mla.libertine.org/tmda-users/2003-09/msg00285.html
>
> I think the regex I use is more "newbie friendly" :-) :
> http://mla.libertine.org/tmda-users/2003-10/msg00246.html
_____________________________________________
tmda-users mailing list ([EMAIL PROTECTED])
http://tmda.net/lists/listinfo/tmda-users
