To: pgsql-sql@postgresql.org
From: Thomas Kellerer <spam_eater@gmx.net>
Subject: Re: Removing whitespace using regexp_replace
Date: Sun, 28 Oct 2007 13:15:20 +0100
Lines: 43
Message-ID: <fg1ugo$m6v$1@ger.gmane.org>
References: <fg1ra3$ehj$1@ger.gmane.org> <20071028114247.GA1327@KanotixBox>
Mime-Version: 1.0
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit
User-Agent: Mozilla/5.0 (Windows; U; Windows NT 5.1; de;
	rv:1.8.1.6) Gecko/20070728 Thunderbird/2.0.0.6 Mnenhy/0.7.5.666
In-Reply-To: <20071028114247.GA1327@KanotixBox>

Andreas Kretschmer wrote on 28.10.2007 12:42:
>> I have a column with the datatype "text" that may contain leading 
>> whitespace (tabs, spaces newlines, ...) and I would like to remove them all 
>> (ideally leading and trailing).
> 
> You can use trim() for that:
> 
> select 'x' || trim(both '\t' from trim(both ' ' from ' \t\tfoo bar  ')) || 'x';
> 
> (for testing with 'x' around the result)
Yes I was thinking about a solution like that as well, but wouldn't that only 
work if the order in which spaces and tabs appear is always the same?
The above would replace ' \t' but not '\t ', right?


> For regexp_replace() you need an extra parameter 'g' like below:
Cool, works like a charm.
Didn't see that parameter when first reading that chapter.

But it seems my problem was actually caused by something else:

SELECT regexp_replace(myfield, '\s*', '', 'g')
FROM mytable;

does not replace anything, but

SELECT regexp_replace(myfield, '[ \t\n\r]*', '', 'g')
FROM mytable;

does replace all whitespaces (as I expected). And subsequently

SELECT regexp_replace(myfield, '^[ \t\n\r]*', '', 'g')
FROM mytable;

replaces only the whitespace at the beginning.

I thought \s is a "shortcut" for "whitespace", which in my understanding is the 
same as [ \t\r\n]. Am I wrong here?

Cheers
Thomas