JPA SQL Server No Dialect mapping for JDBC type: -9

Question

Iam writing a native query like

Query query = 
  entityManagerUtil.getEntityManager().createNativeQuery(\"SELECT c.NodeID,c.Code,c.Name FROM COM_Location c\"         


        

Answer 1

First you have to define what is your dialect class name that you want to use: in hibernate.cfg.xml add your own class address

<property name="hibernate.dialect">com.nhl.dao.SQlServerDBDialect</property>

then create new class same below

package com.nhl.dao;
import java.sql.Types;
import org.hibernate.dialect.SQLServerDialect;
import org.hibernate.type.StandardBasicTypes;

public class SQlServerDBDialect extends SQLServerDialect {

    public SQlServerDBDialect() {
        super();
        registerHibernateType(Types.NCHAR, StandardBasicTypes.CHARACTER.getName()); 
        registerHibernateType(Types.NCHAR, 1, StandardBasicTypes.CHARACTER.getName());
        registerHibernateType(Types.NCHAR, 255, StandardBasicTypes.STRING.getName());
        registerHibernateType(Types.NVARCHAR, StandardBasicTypes.STRING.getName());
        registerHibernateType(Types.LONGNVARCHAR, StandardBasicTypes.TEXT.getName());
        registerHibernateType(Types.NCLOB, StandardBasicTypes.CLOB.getName());

    }
}

Answer 2

I also face the same issue, So i do R&D and got solution as Do explicit cast like cast(t2.name as varchar), It works for me.

cast(t2.name as varchar)

Check this url. http://www.coderanch.com/t/565413/ORM/databases/Dialect-mapping-JDBC-type

Answer 3

The type -9 is java.sql.Types.NVARCHAR. Looking at the sources of the SQLServerDialect variants on https://github.com/hibernate/hibernate-orm/tree/master/hibernate-core/src/main/java/org/hibernate/dialect there is no mapping for nvarchar columns.

You might want to try to define your own dialect that also registers various NVARCHAR-like definitions:

public class SQLServer2008DialectWithNvarchar extends SQLServer2008Dialect {
    public SQLServer2008DialectWithNvarchar () {
        registerColumnType( Types.NCLOB, "nvarchar(MAX)" );
        registerColumnType( Types.LONGNVARCHAR, "nvarchar(MAX)" );
        registerColumnType( Types.NVARCHAR, "nvarchar(MAX)" );
        registerColumnType( Types.NVARCHAR, 4000, "nvarchar($1)" );
    }
}

I based this on the definition for VARCHAR in the SQLServer2005Dialect. You may need to put this class in the org.hibernate.dialect package (or at least I seem to remember there are issues if you don't).

NOTE: I haven't actually tested this!

Answer 4

You have to extend the dialect, and register the appropriate Hibernate types for the N* JDBC types, in the constructor:

public class MyDialect extends SomeOfTheProvidedDialects {

    public MyDialect() {
        registerHibernateType(Types.NCHAR, StandardBasicTypes.CHARACTER.getName());
        registerHibernateType(Types.NCHAR, 1, StandardBasicTypes.CHARACTER.getName());
        registerHibernateType(Types.NCHAR, 255, StandardBasicTypes.STRING.getName());
        registerHibernateType(Types.NVARCHAR, StandardBasicTypes.STRING.getName());
        registerHibernateType(Types.LONGNVARCHAR, StandardBasicTypes.TEXT.getName());
        registerHibernateType(Types.NCLOB, StandardBasicTypes.CLOB.getName());
    }
}

Tested, works fine.

They should have added these to the Hibernate source code already (in the org.hibernate.dialect.Dialect class)...

Answer 5

I had a similar issue, my sql is this

Query nmspQuery = em.createNativeQuery("select aster from MyTable where my_column = 1001")

The datatype for PRODUCT_NODE_ONIX_TYPE in the Oracle 11g is nvarachar2

I fixed it by casting the value astr to to_char. This worked fine

Query nmspQuery = em.createNativeQuery("select TO_CHAR(aster) from MyTable where my_column = 1001")