Re: INDEX possible for reverse wildcards?
From: Mikito Harakiri <mikharakiri_at_iahu.com>
Date: Fri, 4 Jun 2004 15:52:28 -0700
Message-ID: <Qd7wc.36$924.32_at_news.oracle.com>
Date: Fri, 4 Jun 2004 15:52:28 -0700
Message-ID: <Qd7wc.36$924.32_at_news.oracle.com>
<ctcgag_at_hotmail.com> wrote in message
news:20040604183510.277$gk_at_newsreader.com...
> Well, you would only need to look at index entries that start with 'w',
> '%', or '_'. And if the first character is a 'w' or a '_', you only need
> to look at ones where the second is 'w', '%', or '_'.
And if the first character happens to be '%', then?
> And so on. Kind of
> like a skip scan. (or course, in the example given, this wouldn't gain
you
> anything, as they all either start with 'w' or a wild-card.)
Received on Sat Jun 05 2004 - 00:52:28 CEST