Re: unexpected (to me) sorting order

Поиск
Список
Период
Сортировка
Искать
От
Alberto Cabello Sánchez
Тема
Re: unexpected (to me) sorting order
Дата
Msg-id
20150408115616.0414b63cf88f430139c35187@unex.es
Ответ на
Список
Дерево обсуждения
unexpected (to me) sorting order Björn Lundin <b.f.lundin@gmail.com>
Re: unexpected (to me) sorting order Chris Mair <chris@1006.org>
Re: unexpected (to me) sorting order Alberto Cabello Sánchez <alberto@unex.es>
Re: unexpected (to me) sorting order Björn Lundin <b.f.lundin@gmail.com>
Re: unexpected (to me) sorting order Glyn Astill <glynastill@yahoo.co.uk>
Re: unexpected (to me) sorting order Björn Lundin <b.f.lundin@gmail.com>
Re: unexpected (to me) sorting order Glyn Astill <glynastill@yahoo.co.uk>
Re: unexpected (to me) sorting order Björn Lundin <b.f.lundin@gmail.com>
Re: unexpected (to me) sorting order Scott Marlowe <scott.marlowe@gmail.com>
Re: unexpected (to me) sorting order Glyn Astill <glynastill@yahoo.co.uk>
On Wed, 08 Apr 2015 11:36:01 +0200
Chris Mair  wrote:

> PostreSQL relies on the OS's C lib. So this kind
> of ordering problems depend on the OS' idea about
> collations.
> 
> I don't know what's the rationale behin this,
> but it looks like Linux ignores the . when doing the sort.

Not only '.'. Sorting by VARCHAR is not ASCII order, but "alphabetical" (whatever
it means), which causes some behaviour to be more real-life oriented than
logically coherent, like this one:


select * from ejemplo order by texto;

 id |  texto   
----+----------
  2 | Lalin
  1 | La Palma
  3 | Lasarte


It is pretty obvious that ' ' is not between 'l' and 's', but this makes more
sense than 

  1 | La Palma
  2 | Lalin
  3 | Lasarte


-- 
Alberto Cabello Sánchez


В списке pgsql-general по дате отправления
От: Björn Lundin
Дата:
От: Filipe Pina
Дата:
FAQ