In-Reply-To: <479D5F54.3040107@obviously.com>
References: <479D3E63.2080601@obviously.com> <16843.1201489840@sss.pgh.pa.us>
	<479D5F54.3040107@obviously.com>
Mime-Version: 1.0 (Apple Message framework v753)
Content-Type: text/plain; charset=US-ASCII; delsp=yes; format=flowed
Message-Id: <EA81A141-5E8B-4C82-8B9E-209817E62EC9@seespotcode.net>
Cc: pgsql-sql@postgresql.org
Content-Transfer-Encoding: 7bit
From: Michael Glaesemann <grzm@seespotcode.net>
Subject: Re: extract or date_part on an interval? How many e
Date: Mon, 28 Jan 2008 00:19:04 -0500
To: Bryce Nesbitt <bryce1@obviously.com>


On Jan 27, 2008, at 23:51 , Bryce Nesbitt wrote:

> Sigh. Ok, I settled on:
> select '1987-01-29'::timestamp + interval '21 years' > now();
> Which is closer to what I wanted anyway (this was all about  
> determining who was under 21 years old).  This at least should be  
> robust over leap years.

I think this should work for you for your particular case:

EXTRACT(year FROM AGE(born_on)) >= 21

  SELECT born_on, current_date, age(born_on), EXTRACT(year FROM AGE 
(born_on)) AS age_in_years, EXTRACT(year FROM age(born_on)) >= 21 as  
old_enough_to_drink FROM (SELECT CAST('1987-01-25' AS DATE) + i AS  
born_on FROM generate_series(0,5) as the(i)) foo;
   born_on   |    date    |           age            | age_in_years |  
old_enough_to_drink
------------+------------+--------------------------+-------------- 
+---------------------
  1987-01-25 | 2008-01-28 | 21 years 3 days          |           21 | t
  1987-01-26 | 2008-01-28 | 21 years 2 days          |           21 | t
  1987-01-27 | 2008-01-28 | 21 years 1 day           |           21 | t
  1987-01-28 | 2008-01-28 | 21 years                 |           21 | t
  1987-01-29 | 2008-01-28 | 20 years 11 mons 30 days |           20 | f
  1987-01-30 | 2008-01-28 | 20 years 11 mons 29 days |           20 | f
(6 rows)

Michael Glaesemann
grzm seespotcode net