Ytukyg

I have a parent class Contact which has one to one relationship with ContactType.

@OneToOne(cascade=CascadeType.ALL, orphanRemoval = true)

@JoinColumn(name = "type")

private ContactType contactType;

I am trying to create a new contact using ContactRepository.save() which extends JPARepository. I am getting the following error.

Violation of PRIMARY KEY constraint type. Cannot insert duplicate key in object ContactType

If I change the contact type declaration to below:

@OneToOne(cascade=CascadeType.MERGE, orphanRemoval = true)

@JoinColumn(name = "type")

private ContactType contactType;

I get the following error:

object references an unsaved transient instance - save the transient instance before flushing

Code used to create a contact (It just calls JPARepository.save()):

contactsRepository.save(contact);

How to use merge and persist the data at the same time.

Your comments mentioned:

The primary Key is type string. I want to insert a new row to type if
the user enters new type Ex:"Type2" when creating a new contact. If
the user enters already existing type then I dont want to insert a new
row.

This makes sense then why you're getting a Cannot insert duplicate key error.

The reality is there is not a one to one relationship between Contact and Contact Type as there is not one contact type for every contact. As you said, contact types are reused. What you should be doing is using a many to one between Contact and Contact Type as one contact can have only one Contact Type but one Contact Type can apply to more than one Contact. Iow, the same Contact Type can be one more than one Contact.

So you make it a many to one and then before you save you'll need to look up the Contact Type matching the given one and insert that one if it exists, if not, populate it and let the cascade save the new Contact Type.

Calling contactsRepository.save(contact) with:

@OneToOne(cascade=CascadeType.ALL, orphanRemoval = true)

@JoinColumn(name = "type")

private ContactType contactType;

raises exception because the persistence context cascades the persist operation and sees contactType as transient with primary key set, ready to be persisted. A row with the same PK already exists, hence the error.

The second case:

@OneToOne(cascade=CascadeType.MERGE, orphanRemoval = true)

@JoinColumn(name = "type")

private ContactType contactType;

The operation is persist (not merge) therefore not cascaded. The persistent context sees contactType as transient and cannot go further with the persist because one of the dependencies is in transient state.

Solution

Get rid of cascade:

@OneToOne

@JoinColumn(name = "type")

private ContactType contactType;

Before calling contactsRepository.save(contact); make sure the contactType is in managed state. You can do it this way:

contact.setContactType( entityManager.getReference(ContactType.class, contactType.getId()));

Make sure you replace getId() with the primary key getter.

Merging contactType into the persistence context with contact.setContactType(contactTypeRepository.merge(contactType)); is also valid.