hi stephen/fred,
>> but, i've not (yet) found an example of dbm usage for mixed-type lists, as
>> above.
>
> As has already been noted, it's a single exact key match.
got it.
and, now, here's me, currently snuggled up with a freshly _printed_ spec and a
highlighter. us old folks are a little slow when reading online, and just like
the feel of paper ...
;-)
>> any example/reference would be appreciated ....
> Grab this script:
> http://www.megacity.org/software_downloads/cidrexpand
>
> and do a 2 pass run over your file - first convert it with the abve
> script and then build the db.
>
> HTH,
in does in principle. thx much for the script/example!
> Interesting, that would work. But I would think it would be typical
> for the resulting DB file to have at least two or three orders of
> magnitude more entries than the original CIDR file. Is a dbm or cdb
> search really so much faster that this would be a win?
from a performance perspective, i certainly agree that's a valid question.
what remains is the 'advantage' -- perceived or real is up for discussion -- of
having just one format of lists ...
from my _current_ perspective, with sub-100 entries in the list, it's probably
easier to stay with net-iplsearch.
_might_ be worth a benchmark comparison, tho ...
cheers,
richard
