Re: Suggested update for spamass-milter

[Cassandra Lynette Brockett]

That sounds great.  The message suggests we can access cvs trees on
savannah at the moment, if so I'll grab a copy.

I'm am not in a rush for it at the moment (we need to get the replacement
hosting server built yet, and I've been lazy, we've mainly been doing
testing to verify things will work the way we expect them too...), so I
can wait a few more days for it if I can't get it today, I was expecting
it to take a week or so for the addition to happen anyway :-)

--
Regards,
   Cassandra

Dan Nelson said:
> In the last episode (Dec 19), Cassandra Lynette Brockett said:
>> At the moment a username is sent to spam[cd] as the localpart only,
>> this causes a few issues if you are handling virtualdomains on the
>> mailserver.  Here's an example :-
>>
>> User 1 - address@hidden (it's a fake testing domain)
>> User 2 - address@hidden (again, another fake domain).
> [...]
>> Therefore is there a way an option could be added to spamass-milter
>> (even if it is --work-with-cyrus or something like that if all the
>> single-letters are taken *grin*), which would hand over full
>> recipient usernames to the local spam[cd] process.
>
> The CVS source already does that, with the -e flag.  Of course, with
> savannah down, you can't get it :(  As soon as it comes back up I'll
> roll another release.  Their current message says it'll be up today,
> but I wouldn't be surprised if it took a couple more days.
>
> Until then, using this in place of the regular local_user function in
> spamass-milter.cpp should do what you want:
>
> string
> SpamAssassin::local_user()
> {
>   // assuming we have a recipient in the form:
> <address@hidden>
>   // we return 'address@hidden'
>   if (_rcpt[0]=='<')
>     return _rcpt.substr(1,_rcpt.find('>')-1);
>   else
>     return _rcpt;
> }
>
> --
>       Dan Nelson
>       address@hidden
>