Re: index type for indexing long texts

> Aleksander Kmetec <aleksander.kmetec@intera.si> writes:
> > I'm looking for a solution for indexing long TEXT columns. We're currently using a HASH index, which can handle
most
> > situations, but every now and then we need support for even longer texts.
>
> > One solution would be to create a functional index which would only use the first N chars of mycol, but then we'd
have
> > to change several hundred occurences of "mycol = someval" with "(mycol = someval AND firstN(mycol) =
firstN(someval))",
> > as well as update some SQL generators...
>
> > That's why I'd be interested to know if there are any index types available which store only the first N chars or
use
> > some highly compressed form for storing index data, and then recheck any potential hits against the main table. And
if
> > something like that does not exist yet, how difficult would it be to construct such a solution out of many "spare
parts"
> > that come with PG?
>

Try moving where the hash takes place - ie, use your own hash function to
create the key.

RT



--
Richard Troy, Chief Scientist
Science Tools Corporation
510-924-1363 or 202-747-1263
rtroy@ScienceTools.com, http://ScienceTools.com/