Re: UTF-8 and LIKE vs =

On Mon, 23 Aug 2004 12:41:30 -0700, David Wheeler <david(at)kineticode(dot)com> wrote:
(...)
> bric=# select version();
> version
> ------------------------------------------------------------------------
> ---------------------------------
> PostgreSQL 7.4.2 on i686-pc-linux-gnu, compiled by GCC gcc (GCC) 3.2.2
> 20030222 (Red Hat Linux 3.2.2-5)
> (1 row)
>
> bric=# select * from keyword where name = '북한의';
> id | name | screen_name | sort_name | active
> ------+--------+-------------+-----------+--------
> 1218 | 국방비 | 국방비 | 국방비 | 1
> (1 row)

er, the characters in "name" don't seem to match the characters in the
query - '국방비' vs. '북한의' - does that have any bearing?

FWIW (on 7.4.3):

test=# select * from t1 ;
id | value
----+--------
1 | 日本
2 | 日本語
3 | 北海道
(3 rows)

test=# select * from t1 where value ~ '日';
id | value
----+--------
1 | 日本
2 | 日本語
(2 rows)

test=# select * from t1 where value like '日%';
id | value
----+--------
1 | 日本
2 | 日本語

test=# select * from t1 where value like '北海%';
id | value
----+--------
3 | 北海道
(1 row)

Ian Barwick
barwick(at)gmail(dot)net