Re: [Urgent] Regexp_replace question

On Tue, Sep 25, 2007 at 01:36:26PM +0800, Phoenix Kiula wrote:
> How can I remove characters that form a part of regular expressions?

Why do you want to do that?

> I would like to remove all instances of the following characters:
>
> [
> ]
> \
> +

test=> select id, t, regexp_replace(t, e'[\\\\[\\]+]', '', 'g') from foo;
 id |     t      | regexp_replace
----+------------+----------------
  1 | foo[]+\bar | foobar
(1 row)

test=> select id, t, translate(t, e'[]\\+', '') from foo;
 id |     t      | translate
----+------------+-----------
  1 | foo[]+\bar | foobar
(1 row)

--
Michael Fuhr