Unexpected behaviors for utf8mb4 in MariaDbClob.

In the MariaDbClob.getSubString use java.lang.String.substring but MariaDbClob.length uses codePointCount.

It means String will be truncated unexpectedly in next code if String contains 4 byte UTF-8. because Java uses 2 chars for 4 byte unicode.

clob.getSubString(1L, (int)clob.length());

It makes issue with mybatis ClobTypeHandler.
https://github.com/mybatis/mybatis-3/blob/master/src/main/java/org/apache/ibatis/type/ClobTypeHandler.java