On Sep 12, 2008, at 11:31, Tom Lane wrote:
> "David E. Wheeler" <david@kineticode.com> writes:
>> Oh, and text_name seems to give me this error:
>
>> ERROR: compressed data is corrupt
>
>> That's when I have this cast:
>
>> CREATE OR REPLACE FUNCTION citext(name)
>> RETURNS citext
>> AS 'text_name'
>> LANGUAGE internal IMMUTABLE STRICT;
>
> I think you've got the direction backwards.
Oh. Duh.
> BTW, I removed the "Limitations" entry about I/O casting not working
> with citext; we fixed that, no?
Yes, we did. Thanks for the catch.
I've got another patch I'm working on adding support for "char" (and
tests for char). Just to fill out a gap I saw in the casting coverage.
I'm trying to get it done now. With that, AFAIK, citext will work just
like text.
Best,
David