Question

My domain has a Category entity which has a biderectional relationship on itself. Each category can have a parent and children.

@Entity
public class Category implements DomainObject {

    private Long id;
    private Integer version;
    private String name;
    private Category parent;
    private Set<Category> children;

    @Override
    @Id
    @GeneratedValue
    public final Long getId() {
        return id;
    }

    @Version
    public Integer getVersion() {
        return version;
    }

    public void setVersion(Integer version) {
        this.version = version;
    }

    public void setId(Long id) {
        this.id = id;
    }

    @Column(unique=true, nullable=false)
    public String getName() {
        return name;
    }

    public void setName(String name) {
        this.name = name;
    }

    @ManyToOne
    public Category getParent() {
        return parent;
    }

    public void setParent(Category parent) {
        this.parent = parent;
    } 

    @OneToMany
    @JoinColumn(name = "parent_id")
    public Set<Category> getChildren() {
        return children;
    }

    public void setChildren(Set<Category> children) {
        this.children = children;
    }
}

I have created the following query to fetch the “root” categories with their direct (level 1) children.

select distinct c from Category c left join fetch c.children where c.parent is null order by c.name

This actually works. My question is: why do I need the “JoinColumn” annotation on getChildren() to make this work and why can’t I just make a “foin fetch” query, without “distinct”? If I remove “distinct” I get a multiplication. For each child of a parent, the entire parent is copied in the result set.

Is there a better way to do this? It just feels… a bit crappy.

Answer 1

In JPA you need to set distinct when you join a OneToMany, otherwise it will return duplicates.
This is required.

The JPA spec requires this, but it is an odd default, but relates to what happens in database joins.