EclipseLink/Examples/JPA/Migration/OpenJPA/ColumnTypes

Differences between OpenJPA and EclipseLink: Database Column Type

The columns added in the database tables created for each entity differ between OpenJPA and EclipseLink. When migrating from OpenJPA to EclipseLink, if a user wants to use existing tables created by OpenJPA, most of these differences will not be problematic since EclipseLink adapts to the existing column types when persisting new entities. The following table lists the field types that are created as a different column type in OpenJPA and EclipseLink.

Java Type	OpenJPA	EclipseLink
java.io.Serializable java.lang.Byte[], byte[] java.lang.Class java.lang.String[]	BLOB(1048576)	BLOB(64000)
java.lang.Boolean, boolean	SMALLINT	SMALLINT WITH DEFAULT 0
java.lang.Character	CHAR(254)	CHAR(1)
java.lang.Character[], char[]	VARCHAR(254)	CLOB(64000) LOGGED NOT COMPACT
java.lang.Float, float	REAL	DOUBLE
java.lang.String	VARCHAR(254)	VARCHAR(255)
java.math.BigDecimal	DOUBLE	DECIMAL(15,0)
enum	SMALLINT	INTEGER

Additional Differences: CHAR(size) Return Value

In OpenJPA, when an entity defines a CHAR column with a size specified, any persisted value shorter than the size specified will be returned with extra whitespace. However, in the same scenario, EclipseLink will return the original persisted value with no extra whitespace.
For example, for the following entity:

@Entity
public class Student {
…
	@Column (columnDefinition="CHAR(20)")
	private String firstName;
…
}

The following operations are executed:

Student student = new Student();
student.setFirstName("John");
…
em.persist(student);
…
System.out.println("Student name is (" + em.find(Student.class, id).getFirstName() + ")");

The output will be the following:

Student name is (John                )