| From: | jtp <john(at)akadine(dot)com> | 
|---|---|
| To: | Nick Fankhauser <nickf(at)ontko(dot)com> | 
| Cc: | pgsql-sql(at)postgresql(dot)org | 
| Subject: | Re: Resources | 
| Date: | 2002-01-11 16:23:45 | 
| Message-ID: | Pine.BSF.4.21.0201111115590.6416-100000@db.akadine.com | 
| Views: | Whole Thread | Raw Message | Download mbox | Resend email | 
| Thread: | |
| Lists: | pgsql-sql | 
On Fri, 11 Jan 2002, Nick Fankhauser wrote:
> 
> > I don't beleive indexes will improve SELECT using LIKE.
> 
> I wondered about that too, so I did a test using a database I'm working
> with. The results indicate that it helps substantially- here is my terminal
> log file:
[snip]
an interesting (?) addendum to this.  Yes it does help substantially,
unless your wildcharacter has characters after it.
> staging=# explain select count(*) from actor where actor_full_name like
> 'A%';
try: explain select count(*) from actor where actor_full_name like '%a';
not too relevant for searches here, but occasionally it could be.  my
situation was having a merged last name + zipcode field, instead of having
two separately indexed fields.  Found out i couldn't do the job i wanted
that way.
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Frank Bax | 2002-01-11 16:27:58 | Re: Resources | 
| Previous Message | Nick Fankhauser | 2002-01-11 16:13:58 | Re: Resources |