JPA中DomesticViolence、Person、PersonState三实体关联映射咨询

你的核心问题是误用了@ManyToMany注解——因为业务要求每对DomesticViolence-Person关联对应多个PersonState，而@ManyToMany的中间表只能存储关联关系，无法承载额外的多状态数据。正确的做法是取消@ManyToMany，将PersonState作为关联DomesticViolence和Person的实体，通过双向@OneToMany/@ManyToOne来映射关系。

问题分析 #

原代码中@ManyToMany生成的中间表domestic-violence-person只能记录“某DomesticViolence关联某Person”，但无法为这对关联存储多个状态（如drunk、drugged）。而PersonState本身就是关联关系的载体，同时需要保存状态信息，因此需要调整实体映射逻辑。

修正后的实体代码 #

1. DomesticViolence实体 #

移除@ManyToMany，改为关联PersonState的@OneToMany：

@Entity
@Table(name = "domestic_violence")
@Getter
@Setter
public class DomesticViolence {
    @Id
    @GeneratedValue(strategy = GenerationType.IDENTITY)
    private Long id;

    // 关联该家暴事件下的所有人员状态记录
    @OneToMany(mappedBy = "domesticViolence", cascade = CascadeType.ALL, orphanRemoval = true)
    private Set<PersonState> personStates = new HashSet<>();

    // 辅助方法：快速添加人员与状态的关联
    public void addPersonState(Person person, String state) {
        PersonState personState = new PersonState();
        personState.setDomesticViolence(this);
        personState.setPerson(person);
        personState.setState(state);
        personStates.add(personState);
        person.getPersonStates().add(personState);
    }
}

2. Person实体 #

同样移除@ManyToMany，改为关联PersonState的@OneToMany：

@Entity
@Table(name = "person")
@Getter
@Setter
public class Person {
    @Id
    @GeneratedValue(strategy = GenerationType.IDENTITY)
    private Long id;

    // 关联该人员参与的所有家暴事件状态记录
    @OneToMany(mappedBy = "person", cascade = CascadeType.ALL, orphanRemoval = true)
    private Set<PersonState> personStates = new HashSet<>();
}

3. PersonState实体（两种主键方案） #

PersonState作为关联实体，需要保证DomesticViolence-Person-State的组合唯一，提供两种实现方案：

方案一：复合主键（推荐，贴合业务唯一性）

先定义复合主键类：

@Embeddable
@Getter
@Setter
@EqualsAndHashCode
public class PersonStateId implements Serializable {
    @Column(name = "domestic_violence_id")
    private Long domesticViolenceId;

    @Column(name = "person_id")
    private Long personId;

    @Column(name = "state")
    private String state;
}

然后实现PersonState实体：

@Entity
@Table(name = "person_state")
@Getter
@Setter
public class PersonState {
    @EmbeddedId
    private PersonStateId id;

    @ManyToOne(fetch = FetchType.LAZY, optional = false)
    @MapsId("domesticViolenceId")
    @JoinColumn(name = "domestic_violence_id")
    private DomesticViolence domesticViolence;

    @ManyToOne(fetch = FetchType.LAZY, optional = false)
    @MapsId("personId")
    @JoinColumn(name = "person_id")
    private Person person;
}

方案二：自增主键+唯一约束

如果不想使用复合主键，可通过唯一约束保证业务唯一性：

@Entity
@Table(name = "person_state", uniqueConstraints = {
    @UniqueConstraint(columnNames = {"domestic_violence_id", "person_id", "state"})
})
@Getter
@Setter
public class PersonState {
    @Id
    @GeneratedValue(strategy = GenerationType.IDENTITY)
    private Long id;

    @ManyToOne(fetch = FetchType.LAZY, optional = false)
    @JoinColumn(name = "domestic_violence_id")
    private DomesticViolence domesticViolence;

    @ManyToOne(fetch = FetchType.LAZY, optional = false)
    @JoinColumn(name = "person_id")
    private Person person;

    @Column(nullable = false)
    private String state; // 存储drunk、drugged、sober等状态值
}

额外注意点 #

• 原代码中peopele是拼写错误，应为people；
• 使用cascade = CascadeType.ALL和orphanRemoval = true可以简化关联数据的增删操作；
• FetchType.LAZY是默认值，可避免不必要的关联数据加载，提升性能。

内容的提问来源于stack exchange，提问作者HDonev