De: hubert depesz lubaczewski [mailto:[EMAIL PROTECTED]
>>On Fri, Aug 10, 2007 at 04:40:34PM -0300, Fernando Hevia wrote:
>> Found your query is shorter and clearer, problem is I couldn't have it
use
>> an index. Thought it was a locale issue but adding a 2nd index with
>> varchar_pattern_ops ma
On Fri, Aug 10, 2007 at 08:13:46PM -0500, Rodrigo De León wrote:
> On 8/10/07, hubert depesz lubaczewski <[EMAIL PROTECTED]> wrote:
> > unfortunatelly this query will be hard to optimize.
>
> Uh, how about
>
> SELECT MAX(t1)
> FROM t1
> WHERE '9849' LIKE t1 || '%';
it will not help much as th
On 8/10/07, hubert depesz lubaczewski <[EMAIL PROTECTED]> wrote:
> unfortunatelly this query will be hard to optimize.
Uh, how about
SELECT MAX(t1)
FROM t1
WHERE '9849' LIKE t1 || '%';
---(end of broadcast)---
TIP 6: explain analyze is your frie
On Fri, Aug 10, 2007 at 04:40:34PM -0300, Fernando Hevia wrote:
> Found your query is shorter and clearer, problem is I couldn't have it use
> an index. Thought it was a locale issue but adding a 2nd index with
> varchar_pattern_ops made no difference.
> In result, it turned out to be too slow in c
ing the explain plan. It has ~31k rows.
Any hindsight will be greatly appreciated.
Regards,
Fernando.
-Mensaje original-
De: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED]
En nombre de hubert depesz lubaczewski
Enviado el: Viernes, 10 de Agosto de 2007 05:00
Para: Kiran
CC: pgsql-sql@postgresql.o
On Mon, Aug 06, 2007 at 01:57:07AM -0700, Kiran wrote:
> Could anyone help me in writing Best Fit SQL statement.
> Suppose we have table t1 with coloumn t1 (text) with following rows.
> 98456
> 98457
> 9845
> 9846
> 984
> 985
> 98
> 99
> and if I query on 98456 the result must be 98456,
> However
On Aug 6, 3:57 am, Kiran <[EMAIL PROTECTED]> wrote:
> and if I query on 98456 the result must be 98456,
> However if I query on 98455 the result must be 9845
> and If I query 9849 the result must be 984
SELECT MAX(t1)
FROM t1
WHERE '9849' LIKE t1 || '%';
---(end of bro
