Oracle ADF与Hibernate集成问询：从Hibernate/JPA实体读取数据至ADF视图对象

Great question—integrating Hibernate/JPA with Oracle ADF is totally feasible without relying on EJB, and I’ve worked through a few practical approaches that get the job done. Let’s break down how to pull data from your Hibernate entities into ADF View Objects (VOs) directly:

Core Approach: Custom Programmatic View Object #

Instead of using ADF’s standard Entity Object (EO)-backed VOs, you’ll create a custom VO that leverages Hibernate’s SessionFactory (or JPA’s EntityManager) to fetch and map data. Here’s a step-by-step guide:

1. Set Up Hibernate/JPA in Your ADF Project #

• Add Hibernate dependencies (like hibernate-core, hibernate-entitymanager) to your ADF project. In JDeveloper, use the Libraries and Classpath settings to include these JARs, or add them to your pom.xml if using Maven.
• Configure your hibernate.cfg.xml (or persistence.xml for JPA) with entity mappings, database connection details, and session factory settings. Ensure this file lives in your project’s classpath (e.g., src/main/resources).

2. Create a Custom View Object Class #

Create a new ADF View Object (select Programmatic mode, no Entity Object attached) and override key methods to integrate Hibernate:

import org.hibernate.Session;
import org.hibernate.SessionFactory;
import org.hibernate.cfg.Configuration;
import oracle.jbo.server.ViewObjectImpl;
import oracle.jbo.server.ViewRowImpl;
import java.util.List;

public class HibernateVOImpl extends ViewObjectImpl {
    private SessionFactory sessionFactory;

    @Override
    protected void create() {
        super.create();
        // Initialize Hibernate SessionFactory
        sessionFactory = new Configuration().configure().buildSessionFactory();
    }

    @Override
    protected void executeQuery() {
        super.executeQuery();
        // Clear existing rows to avoid duplicates
        getRowSetIterator().closeRowSetIterator();
        removeAllRows();

        // Fetch data from Hibernate entity
        try (Session session = sessionFactory.openSession()) {
            List<YourHibernateEntity> entities = session.createQuery(
                "from YourHibernateEntity", YourHibernateEntity.class
            ).list();
            
            // Map entity fields to VO attributes
            for (YourHibernateEntity entity : entities) {
                ViewRowImpl row = (ViewRowImpl) createRow();
                row.setAttribute("EntityId", entity.getId());
                row.setAttribute("EntityName", entity.getName());
                // Add mappings for other fields as needed
                insertRow(row);
            }
        } catch (Exception e) {
            throw new RuntimeException("Failed to fetch data from Hibernate: " + e.getMessage(), e);
        }
    }
}

3. Align VO Attributes with Hibernate Entity Fields #

• In your VO’s definition, create attributes that match the fields in your Hibernate entity (e.g., EntityId for Long id, EntityName for String name).
• Ensure attribute data types match exactly to avoid mapping errors.

4. Integrate with ADF Application Module and View Layer #

• Add your custom VO to your ADF Application Module.
• Bind it to your JSF page like any standard ADF VO—use components like af:table or af:inputText to display the data from your Hibernate entities.

Alternative: JPA EntityManager Approach #

If you prefer JPA over raw Hibernate, replace the Hibernate Session code with JPA’s EntityManager:

import jakarta.persistence.EntityManager;
import jakarta.persistence.EntityManagerFactory;
import jakarta.persistence.Persistence;
// ... other imports

@Override
protected void executeQuery() {
    super.executeQuery();
    removeAllRows();

    EntityManagerFactory emf = Persistence.createEntityManagerFactory("your-persistence-unit-name");
    EntityManager em = emf.createEntityManager();
    
    List<YourJpaEntity> entities = em.createQuery(
        "SELECT e FROM YourJpaEntity e", YourJpaEntity.class
    ).getResultList();
    
    // Map to VO rows same as above
    em.close();
    emf.close();
}

Key Considerations #

• Transaction Management: Since you’re bypassing ADF BC’s built-in transactions, handle Hibernate/JPA transactions manually (use session.beginTransaction()/commit() or em.getTransaction() for write operations).
• Performance: For large datasets, add pagination with setFirstResult() and setMaxResults() in your Hibernate/JPA queries to avoid loading all data into memory.
• Caching: Use Hibernate’s second-level cache wisely to boost performance, but ensure it’s configured to avoid stale data in your ADF views.
• Maintenance: Keep in mind you’ll be maintaining two persistence layers (ADF BC and Hibernate/JPA). If possible, evaluate if standardizing on one layer makes sense for your project.

内容的提问来源于stack exchange，提问作者Atul