如何在 Spring Data JPA 中保存期间避免选择语句

Question

I have a composite primary key (made of 3 columns) for an entity.

Out of 2 columns from the composite key, 1 is auto increment and another is getting populated from another entity.

when I save the entity, it executes select statement and then insert statement. I want to ignore/avoid execution of select statement as I know that the entity are unique and simply want to insert into the table.

I read about having a field with @Version (like @Version Long version ) in the entity however, in that case, I will have to have a column on the table as well which I don't want.

The execution of select statement taking lots of time especially when I know that the records I am inserting are unique.

[EDITED]

my entity class is:

@Entity
public class Table implements Serializable {
    private static final long serialVersionUID = 1L;

    @EmbeddedId
    private TablePK id;

    private String Name;

    private String country;
}

@Embeddable
public class TablePK implements Serializable {
    //default serial version id, required for serializable classes.
    private static final long serialVersionUID = 1L;

    @GeneratedValue(strategy=GenerationType.AUTO)
    private Long id;

    private Integer age;
}

and the log is:

10:39:57.650 [main] DEBUG org.hibernate.SQL - select 
10:39:57.684 [main] DEBUG org.hibernate.loader.Loader - Done entity load
10:39:57.693 [main] DEBUG o.h.e.i.AbstractSaveEventListener - Generated identifier: 
10:39:57.705 [main] DEBUG o.h.e.t.internal.TransactionImpl - committing
10:39:57.706 [main] DEBUG o.h.e.i.AbstractFlushingEventListener - Processing flush-time cascades
10:39:57.707 [main] DEBUG o.h.e.i.AbstractFlushingEventListener - Dirty checking collections
10:39:57.709 [main] DEBUG o.h.e.i.AbstractFlushingEventListener - Flushed: 1 insertions, 0 updates, 0 deletions to 1 objects
10:39:57.709 [main] DEBUG o.h.e.i.AbstractFlushingEventListener - Flushed: 0 (re)creations, 0 updates, 0 removals to 0 collections
10:39:57.710 [main] DEBUG o.h.internal.util.EntityPrinter - Listing entities:
10:39:57.726 [main] DEBUG org.hibernate.SQL - insert into Table

Can anyone please suggest.

Answer 1

The entity mapping design is wrong. Why do you need a composite identifier when you already want to use an auto-incremented column?

So, just remove the @EmbededId use a simple auto-incremented identifier and map the age as a basic attribute.

But there are more questionable design decisions, like:

I read about having a field with @Version (like @Version Long version) in the entity however in that case I will have to have a column in the table as well which I dont want.

The reason of the @Version annotation is to prevent lost updates in long-running transactions or weaker isolation levels. I'm not sure why do you think it has anything to do with your current problem.

out of 3 column from composite key, 1 is auto increment and other two are getting populated from another entity.

You say 3 columns, but the composite id has only 2 columns.

I suggest you read this comprehensive JPA and Hibernate tutorial so that you understand how Hibernate works and how you should be using it.

Answer 2

I completely Agree with Vlad's answer. However for you or someone else who has a scenario similar to yours and wants to avoid the initial SELECT statement, you can make your Entity implement Persistable

@Entity
public class Table implements Persistable<TablePK> {

  @EmbeddedId
  private TablePK id;

  private String Name;

  private String country;

  @Override
  public boolean isNew() {
    return true; // this makes every entity a new one
  }
}

NOTE: Use this if you can make the guarantee that your every save() will be a unique entity like the author has mentioned in the question. It will throw exceptions in case of conflicts.