Re: searching archives should be a weeeee bit faster ... - Mailing list pgsql-general

From Marc G. Fournier
Subject Re: searching archives should be a weeeee bit faster ...
Date
Msg-id 20030904203636.X51587@ganymede.hub.org
Whole thread Raw
In response to Re: searching archives should be a weeeee bit faster ...  (Sean Chittenden <sean@chittenden.org>)
List pgsql-general

On Thu, 4 Sep 2003, Sean Chittenden wrote:

> > yes it is/was ...I've just removed it from the list of stop words,
> > but am not going to force a re-index of 250k URLs ;( So, any new
> > index'ng will pull in index as a valid search word, and when the
> > current URLs expire/re-index,they will get pulled in then also ...
>
> Hrm... if I search for "correlation," I get some results.  If I search
> for "index," I don't get anything (expected), but if I search for
> "index correlation," I get nothing.  Why isn't it returning the same
> results as when I searched for "correlation?"
>
> Just a scratch that I finally had to scratch and ask about...

my understanding of how it searches is it defaults to an 'and' ... so it
would have to find all docs that have both index (0) *and* correlation (n)
... which would equal zero results ...


pgsql-general by date:

Previous
From: Sean Chittenden
Date:
Subject: Re: searching archives should be a weeeee bit faster ...
Next
From: Tom Lane
Date:
Subject: Re: descending Indexes