Re: [GENERAL] why isn't this subquery wrong?

Поиск
Список
Период
Сортировка
От Steve Crawford
Тема Re: [GENERAL] why isn't this subquery wrong?
Дата
Msg-id CAEfWYyy5DaKqZCBhArEXJ1zhYr_EBViU97ivvO_Kq-pWgsKqhg@mail.gmail.com
обсуждение исходный текст
Ответ на Re: [GENERAL] why isn't this subquery wrong?  (jonathan vanasco <postgres@2xlp.com>)
Список pgsql-general


On Thu, Apr 20, 2017 at 3:56 PM, jonathan vanasco <postgres@2xlp.com> wrote:
thanks all!

On Apr 20, 2017, at 6:42 PM, David G. Johnston wrote:

​Subqueries can see all columns of the parent.  When the subquery actually uses one of them it is called a "correlated subquery".

i thought a correlated subquery had to note that table/alias, not a raw column.   I guess i've just been adhering to good form.


On Apr 20, 2017, at 6:43 PM, Tom Lane wrote:

Cautious SQL programmers qualify all references inside sub-selects to avoid getting caught by this accidentally.

is there a syntax to qualify a reference to lock a subquery to the current scope (disable looking at the parents)?  that's how I got caught on this by accident.

Like Tom said, "qualify all references":

...(SELECT example_a__rollup.bar_id FROM example_a__rollup)...

Or shortened with alises:

...(SELECT x.bar_id FROM example_a__rollup x)...

Cheers,
Steve

В списке pgsql-general по дате отправления:

Предыдущее
От: jonathan vanasco
Дата:
Сообщение: Re: [GENERAL] why isn't this subquery wrong?
Следующее
От: "David G. Johnston"
Дата:
Сообщение: Re: [GENERAL] why isn't this subquery wrong?