Message-ID: <47CEDEB1.3050902@denterprises.org>
Date: Wed, 05 Mar 2008 12:56:01 -0500
From: Colin Wetherbee <cww@denterprises.org>
User-Agent: Thunderbird 2.0.0.12 (Windows/20080213)
MIME-Version: 1.0
To: Jeff Frost <jeff@frostconsultingllc.com>
CC: pgsql-sql@postgresql.org
Subject: Re: finding columns that have three or fewer distinct
 characters
References: 
 <Pine.LNX.4.64.0803050912570.4804@discord.home.frostconsultingllc.com>
In-Reply-To: 
 <Pine.LNX.4.64.0803050912570.4804@discord.home.frostconsultingllc.com>
Content-Type: text/plain; charset=UTF-8; format=flowed
Content-Transfer-Encoding: 7bit

Jeff Frost wrote:
> I've got an interesting one...I'm trying to find columns that have three 
> or fewer distinct characters (for example, "aaaaaaaaaa").  Wondering if 
> I need to write a function or if someone has an idea how to do it with 
> built in functions and/or pattern matching?
> 
> I think the thing to do would be to lowercase everything, then remove 
> all duplicate chars and spaces, then use length() on that, but it's not 
> obvious to me how I might remove the duplicate chars with the pattern 
> matching support in the docs.

It's interesting, indeed.

Here's how you might do it with a PL/Perl function. :)


CREATE OR REPLACE FUNCTION remove_duplicates(TEXT) RETURNS TEXT AS
$$
   my ($text) = @_;
   while ($text =~ s/(.)(.*)\1/$1$2/g != 0) {};
   return $text;
$$ LANGUAGE plperl;


cww=# SELECT 
remove_duplicates('ffffoooooooobbbbaaaarrrr.!@#$.foobar-baz-qux');
  remove_duplicates
-------------------
  fobar.!@#$-zqux
(1 row)


Colin