Re: Expression indexes ignore typmod of expression

Hi,

On 2014-04-25 17:05:26 -0400, Tom Lane wrote:
> A Salesforce colleague asked me why, for something like
> ....
> the index column ends up as "bpchar" and not "char(15)".  The CASE
> expression does get resolved as char(15), but it turns out that
> index.c just ignores that.

I've seen that before but never looked what's the origin. +1 for fixing
it.

>  I think this is just a hangover from
> before we paid much attention to expression typmods at all, and
> propose the attached patch.
> 
> Comments?

Any chance it could cause problems with stored trees being different
from newly generated ones due to it? I.e. is it something that can be
done without a catversion bump?

Greetings,

Andres Freund

-- Andres Freund                       http://www.2ndQuadrant.com/PostgreSQL Development, 24x7 Support, Training &
Services