Re: [patch] [doc] Minor variable related cleanup and rewording of plpgsql docs

Hi

pá 12. 3. 2021 v 21:08 odesílatel Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us> napsal:

> I looked over the v2 patch. Parts of it seem like improvements but
> other parts definitely don't. In particular, I thought you introduced
> a great deal of confusion in 43.5.2 (Executing a Command with No Result).
> The statement that you can write a non-result-returning SQL command as-is
> is true in general, and ought not be confused with the question of whether
> you can insert variable values into it. Also, starting with a spongy
> definition of "utility command" and then contrasting with that does not
> seem to me to add clarity.
>
> I attach a v3 that I like better, although there's room to disagree
> about that. I've always felt that the separation between 43.5.2 and
> 43.5.3 was rather artificial --- it's okay I guess for describing
> how to handle command output, but we end up with considerable
> duplication when it comes to describing how to insert values into a
> command. It's tempting to try re-splitting it to separate optimizable
> from non-optimizable statements; but maybe that'd just end with
> different duplication.
>

I am not sure if people can understand the "optimizable command" term. More
common categories are DML, DDL and SELECT. Maybe it is easier to say. DDL
statements don't support parametrizations, and then the variables cannot be
used there.

> regards, tom lane
>
>