Re: BUG #14885: mistake in sorting win1251 chars

Kalin:

1st, if you follow on top-posting the flow is going to become dificult
to understand.

2nd:
On Fri, Nov 3, 2017 at 12:18 PM, Kalin Daskalov
<k(dot)daskalov(dot)911(at)gmail(dot)com> wrote:
> It's the proper locale. It's set to Bulgarian everywhere.

> - Windows settings:
> format: Bulgarian (Bulgaria)
> Location: Bulgaria
> Current language for non-Unicode programs: Bulgarian (Bulgaria)

I assume this is correct, I don't do windows.

> - PostgreSQL Database settings:
> ENCODING = 'UTF8'
> LC_COLLATE = 'Bulgarian_Bulgaria.1251'
> LC_CTYPE = 'Bulgarian_Bulgaria.1251'

This sounds fishy. Specially the .1251, which hints at a collation for
win1251 encoded bulgarian.

In my computers ( linux, debian and redhat derived most of them ) I
have LC_COLLATE set to en_US.UTF-8, and they seem to sort utf-8
properly. Maybe your .1251 does some strange character mapping utf-8
=> 1251 before sorting. The collations are, IIRC, derived from the c
library collaiton routines anyway.

Francisco Olarte.