Re: [Exim] spam acl condition: cannot parse spamd output

Top Page
Delete this message
Reply to this message
Author: Andreas Metzler
Date:  
To: exim-users
Subject: Re: [Exim] spam acl condition: cannot parse spamd output
On Sat, Oct 11, 2003 at 04:45:52PM +0200, Michael Haardt wrote:
> > Could you make your patch more upgrade-friendly? That is, change it in
> > a way that applying your patch would not require _everybody_ to change
> > their config, making '127.0.0.1:783:' in the spam condition optional,
> > defaulting to the value of spamd_address.
>
> That would make the argument ambiguous.


Hello,
Afaict, no, either spam has one or two arguments (old syntax) or three
or four arguments, you'll only get problems if you make the port
optional, too.

spam =                     | Interpretation
---------------------------+---------------
foo                        | scan as user foo
foo:true                   | scan as user foo, return success
foo:(string != true)       | scan as user foo, return success[1]
foo:bar:true               | freak out(), username missing.
foo:bar:(string != true)   | spamd on foo port bar with username (string
                           | != true)
foo:bar:blah:true          | spamd on foo port bar with username blah
                           | always return success
foo:bar:blah:(str != true) | freak out


If this is too messy to implement or you think it is too ugly please
introduce spam++ with your semantics instead of changing "spam". -
Invalidating every current configfile should only be done if there is
absolutely no way around it.
             cu andreas


[1] Should return an error but afaict currently exiscan just ignores
any second argument != true. Minimizes breakage.
--
"See, I told you they'd listen to Reason," [SPOILER] Svfurlr fnlf,
fuhggvat qbja gur juveyvat tha.
Neal Stephenson in "Snow Crash"