Christian Schmidt wrote:

> Hello Philip,
> 
> Philip Hazel, 22.04.2006 (d.m.y):
> 
> 
>>On Fri, 21 Apr 2006, Christian Schmidt wrote:
>>
>>
>>>How can I make exim (or better: the appendfile transport) decide if a
>>>mail shall be dropped into an existing mbox file or a maildir?
>>
>>The file_format option of the appendfile transport allows for different 
>>formats of single-file mailbox. You might be able to coerce it into 
>>doing the right thing, but then again it might not work. I suspect 
>>you'll have to make the decision at routing time (e.g. by checking for 
>>the existence of appropriate files/directories).
> 
> 
> Thanks, Philip.
> 
> But as I already mentioned, I haven't thought of problems occurring
> with the IMAP/POP server - so I'd choose a one-step-migration solution
> when it comes to migrate to maildir.
> 
> Regards,
> Christian
> 

You *can* do it.  Dovecot-pgsql or -msql are easiest, as the 
location and type can be user-specific.

- Something like:


user_query = SELECT pg_mailroot||'/'||pg_login_name||'/' 
||pg_mailstore AS mail, pg_uid AS uid, pg_gid AS gid FROM 
mailprof WHERE pg_recv_login||'@'||pg_domain_id ='%u' AND 
pg_active LIMIT 1

Where:

mailprof(ile) is the [EMAIL PROTECTED] preferences table

pg_<variable> are the DB attributes, reasonably self-documenting

pg_active is a boolean flag for active/suspended accounts

'mail' is Dovecots variable for user-specific storage

%u is Dovecot's variable for <[EMAIL PROTECTED]>

The concatenation builds a path to the mailstore location and 
type, i.e:

/data/mail/domain/user/Maildir or /data/mail/domain/user/mbox


Courier-IMAP/POP can be handled in a similar manner.

And the same effect can be had with lighter-weight lookups than 
SQL - CDB, for example.

That said, as with an abcessed tooth, a one-step migration is 
probably more easily managed.

Acute pain, rather than chronic...  ;-)

Bill


-- 
## List details at http://www.exim.org/mailman/listinfo/exim-users 
## Exim details at http://www.exim.org/
## Please use the Wiki with this list - http://www.exim.org/eximwiki/

Reply via email to