Re: [Exim] Email Aliases DB?

Top Page
Delete this message
Reply to this message
Author: Chris Knipe
Date:  
To: Philip Hazel
CC: exim-users
Subject: Re: [Exim] Email Aliases DB?
Hi Philip,

> On Sun, 3 Sep 2000, Chris Knipe wrote:
>
> > Firstly, (just for the record) what are all the valid characters that an
> > alias or login account can consist of (on a db and unix platform

please)?
> > I know about [A-Z, a-z, 0-9], what else?
>
> RFC 822 allows almost any characters you like in a local part, provided
> they are quoted correctly in header lines, and RFC 821 allows for
> quoting most things in envelope local parts.


Ta for the info.

> > Does Exim support MSSQL, and if not, Philip, will you be able to perhaps
> > code support for SQL Server if I am able to provide you with the data
> > structures / headers for the MSSQL Server? (I think I have both the

socket
> > and TCP/IP (Port 1433) headers available should it be needed)
>
> I do not like coding anything that I cannot test myself. It is highly
> improbable that I would be able to test this. Also, I am not really an
> SQL person at all.


Hmmmm.... The SQL Strucured language is the same. ie: select value from
table, or update table value (table entries)....

> The code for both MySQL and pgsql is quite short and either could be
> used as a template for another SQL API (unless, of course, MS has chosen
> some totally different way of doing things).


If you don't mind giving me some more info on this.... What connection
"type" is exim using to physical connect to the database server? The MySQL
socket call? or is it a direct tcp connection or what...

See, the thing is just that well, I do know very little about c / c++
programming, but the bit of what I know, is still enough to more than likely
be able to hack something in the lines of this, and get it working
(obviously testing it aswell), if push comes to pull.

Perhaps I can have a look at some of the structures and just do it on my own
or something like that... The main problem which I have at the moment is
just that I don't know how these DBs work over raw TCP stacks. ie: if I
telnet say to the SQL server's port, I don't know how to use the connection
without interfacing through something like ODBC. Given the API is very
similar, perhaps I can get some tips out of it on how to "talk" to the
server, in which case, I can even write a perl router or something to use
perl based scripts on Linux and query the MS Database without any problems.

I hope I understand this correctly :)

Regards,
Chris Knipe
Cell (083) 430-8151