I have been stuck up in this scenario.
My application is in C++ which connects to MySQL database 5.5.34 through MySQL odbc connector v5.2 Unicode Driver.
My tables were using the character set and collation properties 'utf-8'. To insert supplementary unicode characters, i changed it via
My table was initially created:
CREATE TABLE mytable (SAMPLECOLUMN text) ENGINE=InnoDB DEFAULTCHARSET=utf8;
changed to,
ALTER TABLE mytable CONVERT TO CHARACTER SET utf8mb4 COLLATE utf8mb4_unicode_ci;
After this, I am able to directly (not through my app) insert the supplementary characters into the column holding 'TEXT' data type.
The odbc connector crashes when it encounters supplementary characters.
The following are some of my my.ini configurations.
[mysql]
default-character-set=utf8
[mysqld]
character-set-server=utf8
default-storage-engine=INNODB
The query that i am trying to insert:
INSERT INTO mytable (SAMPLECOLUMN) VALUES ("乕乭
The workaround here is to use MariaDB odbc connector to connect to MySQL database. https://mariadb.com/my_portal/download/connector-odbc/1.0
Make sure you set this in ODBC dsn and keep Connection Character Set empty
Starting v5.2 of MyODBC,
SET NAMES
has been deprecated. Hence the need to use MariaDB connector. Hope the supplementary character issue is fixed in MySQL odbc connector.You need to issue
SET NAMES utf8mb4
after connecting to MySQL. Otherwise, the connection cannot handle the 4-byte characters.If that is not sufficient, then chase down
SQLExecDirectW
-- see if there is an update to it.