Re: [WIP] collation support revisited (phase 1)

Martijn van Oosterhout napsal(a):
> On Thu, Jul 10, 2008 at 11:24:29PM +0200, Radek Strnad wrote:
>> Hi,
>>
>> after long discussion with Mr. Kotala, we've decided to redesign our
>> collation support proposal.
>> For those of you who aren't familiar with my WIP patch and comments from
>> other hackers here's the original mail:
>> http://archives.postgresql.org/pgsql-hackers/2008-07/msg00019.php
>
> <snip>
>> phase 1
>> Implement "sort of framework" so the PostgreSQL will have basic guts
>> (pg_collation & pg_charset catalogs, CREATE COLLATION, add collation support
>> for each type needed) and will support collation at database level. This
>> phase has been accepted as a Google Summer of Code project.
>
> Why bother with pg_charset? I don't think I've seen anyone even asking
> for multiple charsets in a single DB and certainly no actual use case.

The example is when you have translation data (vocabulary) in database. But the
reason is that ANSI specify (chapter 4.2) charset as a part of string
descriptor. See below:

— The length or maximum length in characters of the character string type.
— The catalog name, schema name, and character set name of the character set of
the character string type.
— The catalog name, schema name, and collation name of the collation of the
character string type.

If I think about it, the question is if we need both information in pg_type and
so on, because collation automatically define charset.

any idea?

Zdenek

--
Zdenek Kotala Sun Microsystems
Prague, Czech Republic http://sun.com/postgresql