Message-ID: <3F01ED9F.3060704@joeconway.com>
Date: Tue, 01 Jul 2003 13:22:55 -0700
From: Joe Conway <mail@joeconway.com>
User-Agent: Mozilla/5.0 (Windows; U; Windows NT 5.0; en-US;
	rv:1.3) Gecko/20030312
MIME-Version: 1.0
To: Greg Stark <gsstark@mit.edu>
Cc: Josh Berkus <josh@agliodbs.com>, Stefan Bill <sjb26@yahoo.com>,
	pgsql-sql@postgresql.org
Subject: Re: LEAST and GREATEST functions?
References: <20030630201057.14507.qmail@web13905.mail.yahoo.com>
	<200306302032.23113.josh@agliodbs.com>
	<87d6gu2otf.fsf@stark.dyndns.tv>
In-Reply-To: <87d6gu2otf.fsf@stark.dyndns.tv>
Content-Type: text/plain; charset=us-ascii; format=flowed
Content-Transfer-Encoding: 7bit

Greg Stark wrote:
> SELECT greatest(a,b) FROM bar
> 
> would return one tuple for every record in the table with a single value
> representing the greater of bar.a and bar.b.
> 
> You could define your own functions to do this but it would be tiresome to
> define one for every datatype.
> 

In 7.4devel (just starting beta) you can do this:

create or replace function greatest(anyelement, anyelement) returns 
anyelement as 'select case when $1 > $2 then $1 else $2 end' language 'sql';

regression=# select greatest(1, 2);
  greatest
----------
         2
(1 row)

regression=# select greatest('b'::text, 'a');
  greatest
----------
  b
(1 row)

regression=# select greatest(now(), 'yesterday');
            greatest
-------------------------------
  2003-07-01 13:21:56.506106-07
(1 row)

The cast to text is needed because 'a' and 'b' are really typed as 
unknown, and with polymorphic functions, you need a well defined data type.

So if you had a table:
create table g(f1 text, f2 text);
insert into g values ('a','b');
insert into g values ('c','b');
regression=# select greatest(f1, f2) from g;
  greatest
----------
  b
  c
(2 rows)

Doesn't help for 7.3.x, but at least you know help is on the way ;-)
Of course, you could always just use the case statement.

Joe