Support utf8mb3 charset

[GMHDBJD]

Feature Request

Is your feature request related to a problem? Please describe:

mysql 8.0.24+ and mariadb 10.6.1+, create table with utf8

create table tb(id int) DEFAULT CHARACTER SET=utf8;

show create table will return utf8mb3

show create table tb;
CREATE TABLE `tb` (
  `id` int(11) DEFAULT NULL
) ENGINE=InnoDB DEFAULT CHARSET=utf8mb3

TiDB doesn't support utf8mb3

CREATE TABLE `tb` (
    ->   `id` int(11) DEFAULT NULL
    -> ) ENGINE=InnoDB DEFAULT CHARSET=utf8mb3;
ERROR 1115 (42000): Unknown character set: 'utf8mb3'

Describe the feature you'd like:

Describe alternatives you've considered:

Teachability, Documentation, Adoption, Migration Strategy:

[wojiushixiaobai]

关注下，遇到了同样的问题

[niubell]

/assign WizardXiao

[WizardXiao]

When use dm sync data, you can use binlog filter or handle-error as a workaround. And also, you can change a charset like utf8mb4_bin ref doc.
However, support utf8mb3 is the long-term solution.

[dveeden]

MySQL 8.0.30:

mysql> CREATE TABLE t1(id char(1) collate utf8_general_ci);
Query OK, 0 rows affected, 1 warning (0.05 sec)

mysql> CREATE TABLE t2(id char(1) collate utf8mb3_general_ci);
Query OK, 0 rows affected, 1 warning (0.05 sec)

mysql> show warnings;
+---------+------+---------------------------------------------------------------------------------------------------------------------------------------------------+
| Level   | Code | Message                                                                                                                                           |
+---------+------+---------------------------------------------------------------------------------------------------------------------------------------------------+
| Warning | 3778 | 'utf8mb3_general_ci' is a collation of the deprecated character set UTF8MB3. Please consider using UTF8MB4 with an appropriate collation instead. |
+---------+------+---------------------------------------------------------------------------------------------------------------------------------------------------+
1 row in set (0.00 sec)

TiDB v6.2.0:

sql> CREATE TABLE t1(id char(1) collate utf8_general_ci);
Query OK, 0 rows affected (0.1375 sec)

sql> CREATE TABLE t2(id char(1) collate utf8mb3_general_ci);
ERROR: 1273 (HY000): Unknown collation: 'utf8mb3_general_ci'

I think this should be part of #7968

[dveeden]

To me it looks like we should:

1. Rename utf8_general_ci to utf8mb3_general_ci and the same for other 3-byte UTF-8 collations.
2. Map utf8 to utf8mb3 in the parser, and the same for the related collations
3. Consider sending warnings for the use of 3-byte UTF-8 character sets and collations.

[Fanduzi]

I currently understand that this is a problem with tidb parser, which does not recognize utf8mb3, and the SQL audit tool we implemented based on tidb parser also encountered this problem. I hope tidb can fix it!