Re: [HACKERS] make more use of RoleSpec struct

Поиск
Список
Период
Сортировка
От Stephen Frost
Тема Re: [HACKERS] make more use of RoleSpec struct
Дата
Msg-id 20161228153826.GK18360@tamriel.snowman.net
обсуждение исходный текст
Ответ на Re: [HACKERS] make more use of RoleSpec struct  (Alvaro Herrera <alvherre@2ndquadrant.com>)
Список pgsql-hackers
* Alvaro Herrera (alvherre@2ndquadrant.com) wrote:
> Stephen Frost wrote:
> > * Alvaro Herrera (alvherre@2ndquadrant.com) wrote:
>
> > > The only functional issue might be the removal of the IsA() checks.  If
> > > we don't cast any Node before passing it to any of those functions,
> > > there should be no problem because any misfeasance will be reported as a
> > > compile-time warning.  Perhaps it's worth adding IsA() checks in loops
> > > such as the one in roleSpecsToIds().
> >
> > Maybe inside of an Assert() though..?
>
> Yeah, I was of two minds when writing that para.  Maybe roleSpecsToIds
> (and all similar functions, if any others exist) has a limited enough
> caller base that it's trivial to audit that no bogus callsite exists.

Well, I think an Assert() is fine to make sure someone doesn't add a new
call that passes in something else where we don't have the compile-time
type checking happening.

Thanks!

Stephen

В списке pgsql-hackers по дате отправления:

Предыдущее
От: Alvaro Herrera
Дата:
Сообщение: Re: [HACKERS] Reporting planning time with EXPLAIN
Следующее
От: Fabien COELHO
Дата:
Сообщение: Re: [HACKERS] proposal: session server side variables