Re: Selectivity estimation for intarray with @@

Поиск
Список
Период
Сортировка
От Heikki Linnakangas
Тема Re: Selectivity estimation for intarray with @@
Дата
Msg-id 55AE8912.3050809@iki.fi
обсуждение исходный текст
Ответ на Re: Selectivity estimation for intarray with @@  (Alexander Korotkov <a.korotkov@postgrespro.ru>)
Список pgsql-hackers
On 07/21/2015 07:28 PM, Alexander Korotkov wrote:
> On Tue, Jul 21, 2015 at 6:49 PM, Heikki Linnakangas <hlinnaka@iki.fi> wrote:
>
> In this version of patch it's not checked if variable is actually and int[]
> not query_int. See following test case.
>
> # create table test2 as (select '1'::query_int val from
> generate_series(1,1000000));
> # analyze test2;
>
> # explain select * from test2 where '{1}'::int[] @@ val;
> ERROR:  unrecognized int query item type: 0
>
> I've added this check.
>
> * Also use the estimators for the obsolete @ and ~ operators. Not that I
>> care much about those as they are obsolete, but seems strange not to, as
>> it's a trivial matter of setting the right estimator function.
>>
>> * I added an ANALYZE in the regression test. It still won't systematically
>> test all the cost estimation code, and there's nothing to check that the
>> estimates make sense, but at least more of the code will now run.
>
> You also forgot to include intarray--1.0--1.1.sql into patch. I've also
> added it.

Thanks, committed!

- Heikki




В списке pgsql-hackers по дате отправления:

Предыдущее
От: Paul Ramsey
Дата:
Сообщение: Re: [PATCH] postgres_fdw extension support
Следующее
От: Jim Nasby
Дата:
Сообщение: Re: [PROPOSAL] VACUUM Progress Checker.