Re: BUG #7846: Documentation: Should FOUND be documented as an output of UPDATE statement?

"mike(at)fontling(dot)com" <mike(at)fontling(dot)com> wrote:

> Documentation for the SQL UPDATE statement doesn't mention that
> it sets the FOUND variable. (Refer to sql-update.html,
> "Outputs".) That behavior *is* documented, but only in
> plpgsql-statements.html (Basic Statements, section 39.5.5).
>
> I understand the difference between a SQL statement and a
> PL/PGSQL statement. But if I were looking for the effects and
> side-effects of an UPDATE statement, I'd go to the docs for
> UPDATE, not for PL/PGSQL basic statements. I think that's what
> most people would do, especially if they didn't have prior
> knowledge of the FOUND variable.

Outside of plpgsql there isn't any FOUND variable to set.  If we
did what you suggest, I think we'd get bug reports about this:

test=# update t set amt = 2 where id = 1;
UPDATE 1
test=# select found;
ERROR:  column "found" does not exist
LINE 1: select found;
               ^

plpgsql creates a FOUND variable and sets it based on the rows
affected by various statements. Besides that, with all the various
PLs that are supported in PostgreSQL, littering the documention of
each SQL statement with details of how that statement is supported
in each PL would get really messy really quickly.

-Kevin