Re: Re: Re: [GENERAL] can postgresql supported utf8mb4 character sets?

[Arjen Nienhuis <a.g.nienhuis@xxxxxxxxx>]

On Fri, Mar 6, 2015 at 3:55 AM, lsliang <lsliang@xxxxxxxxxxxxxxx> wrote:

 

 

 

 

2015-03-06

 

---

发件人：Adrian Klaver

发送时间：2015-03-05 21:31:39

收件人：lsliang; pgsql-general

抄送：

主题：Re: can postgresql supported utf8mb4 character sets?

 

On 03/05/2015 01:45 AM, lsliang wrote:

> can  postgresql supported   utf8mb4  character set?

> today   mobile  apps support   4-byte  character   and  utf8 can only

> support  1-3 bytes character

The docs would seem to indicate otherwise:

http://www.postgresql.org/docs/9.3/interactive/multibyte.html

http://en.wikipedia.org/wiki/UTF-8

> if   load  string  to database which  contain  a  4-byte character

> will failed  .

Have you actually tried to load strings in to Postgres?

If so and it failed what was the method you used and what was the error?

> mysql   since  5.5.3 support utf8mb4 character sets

> I don't  find  some information about  postgresql

>   thanks

-- 

Adrian Klaver

adrian.klaver@xxxxxxxxxxx

 

>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>

thanks   for  your help .

 

 postgresql   can support   4-byte  character  

 

test=> select * from utf8mb4_test ;

ERROR:  character with byte sequence 0xf0 0x9f 0x98 0x84 in encoding "UTF8" has no equivalent in encoding "GB18030"

test=> \encoding utf8 

test=> select * from utf8mb4_test ;

 content 

---------

 ðŸ˜„

 ðŸ˜„

 

pcauto=> 

 

 

UTF-8 support works fine. The 3 byte limit was something mysql invented. But it only works if your client encoding is UTF-8. In your example, your terminal is not set to UTF-8.

create table test (glyph text);
insert into test values ('A'), ('馬'), ('𐁀'), ('😄'), ('🇪🇸');

select glyph, convert_to(glyph, 'utf-8'), length(glyph) FROM test;
 glyph |     convert_to     | length
-------+--------------------+--------
 A     | \x41               |      1
 馬    | \xe9a6ac           |      1
 𐁀     | \xf0908180         |      1
 😄     | \xf09f9884         |      1
 🇪🇸    | \xf09f87aaf09f87b8 |      2
(5 rows)

What doesn't work is GB18030:

select glyph, convert_to(glyph, 'GB18030'), length(glyph) FROM test;
ERROR:  character with byte sequence 0xf0 0x90 0x81 0x80 in encoding "UTF8" has no equivalent in encoding "GB18030"

I think that is a bug.

Gr. Arjen