Обсуждение: String trim function - possible bug?

I am seeing weirdness using the trim function on a string:

This works as expected:

SELECT 'dhct:bn', trim(leading 'dhct:' from 'dhct:bn');
 ?column? | ltrim
----------+-------
 dhct:bn  | bn
(1 row)

However it fails for these cases:

SELECT 'dhct:dn', trim(leading 'dhct:' from 'dhct:dn');
 ?column? | ltrim
----------+-------
 dhct:dn  | n
(1 row)

SELECT 'dhct:cn', trim(leading 'dhct:' from 'dhct:cn');
 ?column? | ltrim
----------+-------
 dhct:cn  | n
(1 row)

The second case truncates an extra character (which is unwanted or
unexpected).

It fails for both 8.1.4 and 8.2.3.  It seems if the first character after
the 'dhct:' is a 'd' or a 'c' it truncates it as well.  I have not tested
every letter of the alphabet yet.

Any suggestions?

Thanks,
Woody

----------------------------------------
iGLASS Networks
211-A S. Salem St
Apex NC 27502
(919) 387-3550 x813
www.iglass.net

On Wed, Jun 06, 2007 at 12:21:33PM -0400, Woody Woodring wrote:
> I am seeing weirdness using the trim function on a string:

<snip>

You are missunderstanding the trim function.

> However it fails for these cases:
>
> SELECT 'dhct:dn', trim(leading 'dhct:' from 'dhct:dn');
>  ?column? | ltrim
> ----------+-------
>  dhct:dn  | n
> (1 row)

The bit after the leading is the set of *characters* to be removed.
Like so:

# select trim( leading 'dhct:' from 'dhdhda');
 ltrim
-------
 a
(1 row)

It's matching characters, not the exact string...

Have a nice day,
--
Martijn van Oosterhout   <kleptog@svana.org>   http://svana.org/kleptog/
> From each according to his ability. To each according to his ability to litigate.

On Wed, 6 Jun 2007, Woody Woodring wrote:

> I am seeing weirdness using the trim function on a string:
>
> This works as expected:
>
> SELECT 'dhct:bn', trim(leading 'dhct:' from 'dhct:bn');
>  ?column? | ltrim
> ----------+-------
>  dhct:bn  | bn
> (1 row)
>
> However it fails for these cases:
>
> SELECT 'dhct:dn', trim(leading 'dhct:' from 'dhct:dn');
>  ?column? | ltrim
> ----------+-------
>  dhct:dn  | n
> (1 row)

The 8.2 docs give this as the description in the table:
"Remove the longest string containing only the characters (a space by
default) from the start/end/both ends of the string"

That implies that with characters 'dhct:' the string to remove is 'dhct:d'
because that's the longest leading string made up of those characters.

Maybe a form using something like regexp_replace might work better for
you.

My fault, I apologize.  I did realize my mistake while I was at lunch but
had responses before I could post to fix my error.  I can use the
"replace()" function instead.

Thanks,
Woody

-----Original Message-----
From: pgsql-general-owner@postgresql.org
[mailto:pgsql-general-owner@postgresql.org] On Behalf Of Stephan Szabo
Sent: Wednesday, June 06, 2007 1:33 PM
To: Woody Woodring
Cc: pgsql-general@postgresql.org
Subject: Re: [GENERAL] String trim function - possible bug?


On Wed, 6 Jun 2007, Woody Woodring wrote:

> I am seeing weirdness using the trim function on a string:
>
> This works as expected:
>
> SELECT 'dhct:bn', trim(leading 'dhct:' from 'dhct:bn');  ?column? |
> ltrim
> ----------+-------
>  dhct:bn  | bn
> (1 row)
>
> However it fails for these cases:
>
> SELECT 'dhct:dn', trim(leading 'dhct:' from 'dhct:dn');  ?column? |
> ltrim
> ----------+-------
>  dhct:dn  | n
> (1 row)

The 8.2 docs give this as the description in the table:
"Remove the longest string containing only the characters (a space by
default) from the start/end/both ends of the string"

That implies that with characters 'dhct:' the string to remove is 'dhct:d'
because that's the longest leading string made up of those characters.

Maybe a form using something like regexp_replace might work better for you.

---------------------------(end of broadcast)---------------------------
TIP 1: if posting/reading through Usenet, please send an appropriate
       subscribe-nomail command to majordomo@postgresql.org so that your
       message can get through to the mailing list cleanly