且构网

分享程序员开发的那些事...
且构网 - 分享程序员编程开发的那些事

如何在非主键列上连接表?

更新时间:2023-11-24 13:31:34

您不能两次映射双向关联。一方必须使用 mappedBy 属性标记为Many方的逆向:

  @OneToMany(fetch = FetchType.LAZY,mappedBy =foo)
@Fetch(value = FetchMode.SUBSELECT)
@Cascade(value = {CascadeType.ALL})
私人设置< Bar> bar = new LinkedHashSet< Bar>(0);

...

@ManyToOne(fetch = FetchType.LAZY)
@JoinColumn(name =FOOID,referencedColumnName =FOOID)
私人Foo foo;

没有任何理由告诉Hibernate两次该关联是由连接列FOOID映射的。这样做实际上是一个错误,因为它定义了两种不同的单向关联,而不是一种双向关联。



编辑



HHH-4284 。

为了避免这个问题,由于 FOOID 足以确保唯一性,所以解决方法是删除 @Id 注释来自所有者ID和 @IdClass 注释。


I have an issue with join tables on an object in an ORM class hierarchy where the join column is NOT the primary key of the base class due a lagacy database structure. Here is an example of the table design:

CREATE TABLE "SCH"."FOO"
(
        "OWNERID"       NUMBER(10,0) NOT NULL ENABLE,
        "FOOID"         NUMBER(10,0) NOT NULL ENABLE,
        CONSTRAINT "FOO_PK" PRIMARY KEY ("OWNERID", "FOOID")
        CONSTRAINT "FOO_FK1" FOREIGN KEY ("OWNERID") REFERENCES "SCH"."OWNERS" ("OWNERID") ENABLE
)

CREATE TABLE "SCH"."BAR"
(
        "BARID"             NUMBER(10,0) NOT NULL ENABLE,
        "FOOID"             NUMBER(10,0)
        CONSTRAINT "BAR_PK" PRIMARY KEY ("BARID")
)

And here are the mappings (unesessary infomation removed)

@Entity
@IdClass(FooId.class)
@Table(name = "FOO")
public class Foo implements java.io.Serializable
{
    @Id
    @Column(name = "OWNERID")
    private BigInteger ownerId;

    @Id
    @SequenceGenerator(name = "FOO_GENERATOR", sequenceName = "SEQ_FOO")
    @GeneratedValue(generator = "FOO_GENERATOR")
    @Column(name = "FOOID")
    private BigInteger id;

    @OneToMany(fetch = FetchType.LAZY)
    @JoinColumn(name = "FOOID", referencedColumnName = "FOOID")
    @Fetch(value = FetchMode.SUBSELECT)
    @Cascade(value = {CascadeType.ALL})
    private Set<Bar> bar = new LinkedHashSet<Bar>(0);
}


@Entity
@Table(name = "BAR")
public class Bar implements java.io.Serializable
{
    @Id
    @Column(name = "BARID")
    private BigInteger id;

    @ManyToOne(fetch = FetchType.LAZY)
    @JoinColumn(name = "FOOID", referencedColumnName = "FOOID")
    private Foo foo;
}

This fails with an exception:

Caused by: org.hibernate.AnnotationException: referencedColumnNames(FOOID) of com.package.Bar.foo referencing com.package.Foo not mapped to a single property
    at org.hibernate.cfg.BinderHelper.createSyntheticPropertyReference(BinderHelper.java:204)
    at org.hibernate.cfg.ToOneFkSecondPass.doSecondPass(ToOneFkSecondPass.java:114)
    at org.hibernate.cfg.Configuration.processEndOfQueue(Configuration.java:1580)
    at org.hibernate.cfg.Configuration.processFkSecondPassInOrder(Configuration.java:1503)
    at org.hibernate.cfg.Configuration.secondPassCompile(Configuration.java:1419)
    at org.hibernate.cfg.Configuration.buildMappings(Configuration.java:1375)

Could you please help with a solution?

You must not map the bidirectional association twice. The One side must be marked as the inverse of the Many side, using the mappedBy attribute:

@OneToMany(fetch = FetchType.LAZY, mappedBy = "foo")
@Fetch(value = FetchMode.SUBSELECT)
@Cascade(value = {CascadeType.ALL})
private Set<Bar> bar = new LinkedHashSet<Bar>(0);

...

@ManyToOne(fetch = FetchType.LAZY)
@JoinColumn(name = "FOOID", referencedColumnName = "FOOID")
private Foo foo;

There is no reason to tell Hibernate twice that the association is mapped by the join column FOOID. And doing it is actually an error, because it defines two different unidirectional associations rather than one bidirectional association.

EDIT

The above should work, but doesn't due to the following Hibernate bug: It's a Hibernate bug. See HHH-4284.

To circumvent this problem, since the FOOID is enough to ensure uniqueness, a solution would be to remove the @Id annotation from the owner ID and the @IdClass annotation.