On Tue, Sep 22, 2009 at 08:47:23AM +0200, Jan Johansson wrote:
> >Probably not the most efficient, but possibly one of the simplest, ways of
> >doing this is as follows:
> >
> >- create lists of local-parts, one per domain, e.g.
>
> That is exactly what I want to avoid, because we are talking about _many_ domains and a heck-load of users.
>
> > e.g. each file would contain lines like "bob", not bob@???
>
> Yeah, which once again would require me to re-parse the data.
>
> I am looking for short-cuts here :)
OK. I'm not sure if you're still seeking advice from this list, but if you
are, you might want to explain your requirements more fully, because saying
"many domains" and "re-parse the data" and "short-cuts" doesn't adequately
explain what to me what you're seeking.
For example: how many domains? How many local-parts per domain? Are the
lists of local-parts in fact identical for all domains? What is it about the
method I proposed which won't work for you (you say "because we are talking
about many domains", but so what?). How many lookups a second do you expect?
How often do the lists change? Where do the lists come from? and so on ...
--
Dave Evans
http://djce.org.uk/
http://djce.org.uk/pgpkey
