Re: replacing a function called "isnull" reports an error - Mailing list pgsql-bugs

From Tom Lane
Subject Re: replacing a function called "isnull" reports an error
Date
Msg-id 7061.1093377552@sss.pgh.pa.us
Whole thread Raw
In response to Re: replacing a function called "isnull" reports an error  (Peter Eisentraut <peter_e@gmx.net>)
Responses Re: replacing a function called "isnull" reports an error  (César Arnold <cesararnold@yahoo.com.br>)
List pgsql-bugs
Peter Eisentraut <peter_e@gmx.net> writes:
> César Arnold wrote:
>> Hi, I created a function called "isnull" for test
>> purposes and when trying to replace it returns the
>> message belows :

> ISNULL is a reserved word.  You need to quote the name or pick a
> different one.

I think what he's unhappy about is this inconsistency:

regression=# create function isnull(int) returns bool as 'select $1 is null' language sql;
CREATE FUNCTION
regression=# drop function isnull(int);
DROP FUNCTION
regression=# create function public.isnull(int) returns bool as 'select $1 is null' language sql;
ERROR:  syntax error at or near "isnull" at character 24
LINE 1: create function public.isnull(int) returns bool as 'select $...
                               ^
regression=#

ISNULL is a func_name_keyword, so it's legal standing on its own, but we
only allow ColId in a qualified name.  Possibly that could be relaxed.

            regards, tom lane

pgsql-bugs by date:

Previous
From: Tom Lane
Date:
Subject: Re: vm/swap used until exhausted
Next
From: Alvaro Herrera
Date:
Subject: Re: server crash in very big transaction [postgresql 8.0beta1]