Re: Update on tables when the row doesn't change

Поиск
Список
Период
Сортировка
От Sebastian Böck
Тема Re: Update on tables when the row doesn't change
Дата
Msg-id 42943BFE.304@freenet.de
обсуждение исходный текст
Ответ на Re: Update on tables when the row doesn't change  (Dawid Kuroczko <qnex42@gmail.com>)
Ответы Re: Update on tables when the row doesn't change  (Dawid Kuroczko <qnex42@gmail.com>)
Список pgsql-general
Dawid Kuroczko wrote:
> On 5/24/05, Sebastian Böck <sebastianboeck@freenet.de> wrote:
>
>>/* 3rd way of separating updates
>>   con: unnecessary updates on tables
>>   pro: view gets evaluated only 1 time
>>
>>   Not adressing the problem of unnecessary updates, but the view
>>   gets only evaluated one time.
>>
>>*/
>>
>>CREATE OR REPLACE FUNCTION upd (view_test) RETURNS VOID AS $$
>>        DECLARE
>>                NEW ALIAS FOR $1;
>>        BEGIN
>>                RAISE NOTICE 'UPDATE';
>>                UPDATE test SET test = NEW.test WHERE id = OLD.id;
>>                UPDATE join1 SET text1 = NEW.text1 WHERE id = OLD.id;
>>                UPDATE join2 SET text2 = NEW.text2 WHERE id = OLD.id;
>>                UPDATE join3 SET text3 = NEW.text3 WHERE id = OLD.id;
>>                RETURN;
>>        END;
>>$$ LANGUAGE plpgsql;
>
>
> Control question, I didn't check it, but would it be enough to change from:
>    UPDATE join1 SET text1 = NEW.text1 WHERE id = OLD.id;
> to:
>    UPDATE join1 SET text1 = NEW.text1 WHERE id = OLD.id AND text1 <> NEW.text1?
>
> ...  I may be wrong. :)

Yes, thats more elegant then my other (4th) solution.
Was late yesterday evening ;)

Thanks

Sebastian

В списке pgsql-general по дате отправления:

Предыдущее
От: "Dweck Nir"
Дата:
Сообщение: postmaster fails to start
Следующее
От: Martijn van Oosterhout
Дата:
Сообщение: Re: Update on tables when the row doesn't change