Re: Limits of SQL

Поиск
Список
Период
Сортировка
Искать
От
Andreas Seltenreich
Тема
Re: Limits of SQL
Дата
Msg-id
87ll5osd12.fsf@gate450.dyndns.org
Ответ на
Re: Limits of SQL (Joachim Zobel)
Список
Дерево обсуждения
Limits of SQL Joachim Zobel <jzobel@heute-morgen.de>
Re: Limits of SQL Sean Davis <sdavis2@mail.nih.gov>
Re: Limits of SQL Scott Ribe <scott_ribe@killerbytes.com>
Re: Limits of SQL Philip Hallstrom <postgresql@philip.pjkh.com>
Re: Limits of SQL Oleg Bartunov <oleg@sai.msu.su>
Re: Limits of SQL Ben <bench@silentmedia.com>
Re: Limits of SQL Joachim Zobel <jzobel@heute-morgen.de>
Re: Limits of SQL Bruno Wolff III <bruno@wolff.to>
Re: Limits of SQL Joachim Zobel <jzobel@heute-morgen.de>
Re: Limits of SQL Bruno Wolff III <bruno@wolff.to>
Re: Limits of SQL Joachim Zobel <jzobel@heute-morgen.de>
Re: Limits of SQL Andreas Seltenreich <seltenreich@gmx.de>
Joachim Zobel schrob:

> Am Samstag, den 04.06.2005, 15:22 -0500 schrieb Bruno Wolff III:
>> On Sat, Jun 04, 2005 at 21:53:24 +0200,
>>   Joachim Zobel  wrote:
>> > So WITH will allow recursion so I can walk the graph, right? Does this
>> > mean I can recursively join until a terminating condition is reached?
>> 
>> It can be used to compute transitive closures, which I think is what
>> you are really looking for.
>
> There aren't any plans to implement grouping by a user defined
> equivalence relation? This is the other thing I am missing. But OK, WITH

Isn't this already possible by representing the relation through its
canonical mapping (i.e. f(a)=f(b) <=> a relates to b)? You could then
use GROUP BY f(x) to group data into its equivalence classes.

regards,
Andreas
В списке pgsql-general по дате отправления
От: Richard Hayward
Дата:
От: Tom Lane
Дата:
FAQ