Re: immutable functions vs. join for lookups ?

От: Tom Lane
Тема: Re: immutable functions vs. join for lookups ?
Дата: ,
Msg-id: 25151.1113853826@sss.pgh.pa.us
(см: обсуждение, исходный текст)
Ответ на: Re: immutable functions vs. join for lookups ?  (Greg Stark)
Ответы: Re: immutable functions vs. join for lookups ?  (Enrico Weigelt)
Список: pgsql-performance

Скрыть дерево обсуждения

immutable functions vs. join for lookups ?  (Enrico Weigelt, )
 Re: immutable functions vs. join for lookups ?  (Tom Lane, )
  Re: immutable functions vs. join for lookups ?  (Enrico Weigelt, )
   Re: immutable functions vs. join for lookups ?  (Jaime Casanova, )
    Re: immutable functions vs. join for lookups ?  (Enrico Weigelt, )
     Re: immutable functions vs. join for lookups ?  (Dawid Kuroczko, )
 Re: immutable functions vs. join for lookups ?  (Dawid Kuroczko, )
 Re: immutable functions vs. join for lookups ?  ("Merlin Moncure", )
  Re: immutable functions vs. join for lookups ?  (Dawid Kuroczko, )
   Re: immutable functions vs. join for lookups ?  ("Jim C. Nasby", )
    Re: immutable functions vs. join for lookups ?  (Dawid Kuroczko, )
     Re: immutable functions vs. join for lookups ?  (Christopher Kings-Lynne, )
     Re: immutable functions vs. join for lookups ?  (Christopher Kings-Lynne, )
  Re: immutable functions vs. join for lookups ?  (Tom Lane, )
   Re: immutable functions vs. join for lookups ?  (Greg Stark, )
    Re: immutable functions vs. join for lookups ?  (Tom Lane, )
     Re: immutable functions vs. join for lookups ?  (Enrico Weigelt, )
      Re: immutable functions vs. join for lookups ?  (Dawid Kuroczko, )

Greg Stark <> writes:
> Tom Lane <> writes:
>> It would be interesting sometime to try to teach the planner about
>> inlining SQL-language functions to become joins.  That is, given

> The Inlining of the function is presumably a side-issue. I have tons of
> queries that use subqueries in the select list for which the same behaviour
> would be appropriate.

Yeah, I was actually thinking about a two-step process: inline the
function to produce somethig equivalent to a handwritten scalar
sub-SELECT, and then try to convert sub-SELECTs into joins.

            regards, tom lane


В списке pgsql-performance по дате сообщения:

От: Richard van den Berg
Дата:
Сообщение: Foreign key slows down copy/insert
От: Bruce Momjian
Дата:
Сообщение: Re: How to improve db performance with $7K?