Using JPA without Cascade Persist

Question

I have two Classes Error and MainError (* - * association with MainError is the owner of the association)

I want to persist the data in my database and I cannot use Persist Cascade because I have many duplicate objects and I want to recognize the duplicate objects and associate them with the persisted ones (The name of Error and MainError is unique)

Error error1 = new Error ("ERR1");
Error error2 = new Error ("ERR2");
Error error3 = new Error ("ERR3");
Error error4 = new Error ("ERR4");
Error error5 = new Error ("ERR5");
Error error6 = new Error ("ERR1");

MainError mainError1 = new MainError("MAIN1");
MainError mainError2 = new MainError("MAIN2");
MainError mainError3 = new MainError("MAIN2");

mainError1.addError(error1);
mainError1.addError(error2);
mainError1.addError(error3);
mainError1.addError(error6);
mainError2.addError(error1);
mainError2.addError(error4);
mainError3.addError(error5);

//persisting Error and MainError

for example if I already persisted Error error1 = new Error ("ERR1") in my database and then I want to persist Error error6 = new Error ("ERR1"); I want that my application recognize that this is already persisted and associate "ERR1" to the corresponding MainError. I guess I will need to create a method findByName to know if the name of Error/MainError is already persisted and if so, returning this object, working with it and merge it ?

I hope that my question is not confusing Thank you very much

score 1 · Accepted Answer · answered Aug 11 '15 at 17:08

As you mentioned in your question, you can manually check for data exist in DB or not? For that you can also set that column as a uniq constraint.

Alternative solution :

In your case, the database design should be something like this :

error :

id varchar (PK)

mainerror :

id varchar (PK)
error_id varchar (FK)

and the entity classes should be something like this :

Error.java

@Entity
public class Error {

    @Id
    @Basic(optional = false)
    @NotNull
    @Size(min = 1, max = 2147483647)
    @Column(nullable = false, length = 2147483647)
    private String id;

    @OneToMany(mappedBy = "errorId", fetch = FetchType.LAZY)
    private Collection<MainError> mainErrorCollection;

    public Error() {
    }

    public Error(String id) {
        this.id = id;
    }

    //getter and setter methods

}

MainError.java

@Entity
@Table(name = "main_error")
public class MainError {

    @Id
    @Basic(optional = false)
    @NotNull
    @Size(min = 1, max = 2147483647)
    @Column(nullable = false, length = 2147483647)
    private String id;

    @JoinColumn(name = "error_id", referencedColumnName = "id")
    @ManyToOne(fetch = FetchType.LAZY)
    private Error errorId;

    public MainError() {
    }

    public MainError(String id) {
        this.id = id;
    }

    //getter and setter methods
}

You should prevent adding duplicate record by making joint PK of id and error_id for main_error table.

Thank you for your answer but I haven't understood your solution, what is the purpose of JoinColumn? (the constraint of FK is already made when I specify OneToMany(mappedBy..) right? — Learthgz, Aug 11 '15 at 19:03
I just want to guide you that either you can check manually before every persist, method like `findByName` or let the database handle for duplicate record by making that column as a `primary key` or `uniq key`. Java classes in the answer just to demostrate structure.. You can change it according to your requirement.. also refer [jpa-joincolumn-vs-mappedby](http://stackoverflow.com/questions/11938253/jpa-joincolumn-vs-mappedby) — Darshan Patel, Aug 12 '15 at 05:05

Using JPA without Cascade Persist

1 Answers1