Spring Data Jpa Many to Many Query - jpa

entity :
Class User{
private String id;
#ManyToMany
private List<Role> roles;
}
Class Role{
private String id;
#ManyToMany(mappedBy = "roles")
private List<User> users;
}
UserRepository :
public interface UserRepository extends PagingAndSortingRepository<User, Serializable>{
#Query(" ??? ")
public User getUser(String id);
}
public class InMemorySecurityConfiguration extends WebSecurityConfigurerAdapter {
#Autowired
private UserRepository userRepository;
#Override
protected void configure(AuthenticationManagerBuilder auth) throws Exception {
auth
.userDetailsService(username -> {
com.casnetvi.cloud.domain.login.User employee = userRepository.findByEmail(username);
if (employee == null) {
throw new BadCredentialsException("User not found");
}
List<Role> roles1 = employee.getRoles();
System.out.println(roles1);
Print exception occurred :
failed to lazily initialize a collection of role: com.casnetvi.cloud.domain.login.User.roles, could not initialize proxy - no Session"
Want to find out what role(use the getRoles() method to get The Role object) , #Query how to write
( Not #Transactional , )

The default fetch type for manyToMany is Lazy.
so in order to get the list of roles you need to do one of the two :
1. execute the configure method inside a transaction.
2. set the relationship with EAGER fetch type:
#ManyToMany(fetch=FetchType.EAGER)
private List<Role> roles;

Related

Disable additional criteria only in some entity relations

I am making an application based on JPA/EclipseLink, I implemented a soft delete functionality using #AdditionalCriteria in the root class of the hierarchy of entities (all entities inherit from this).
My problem is that now, I need to create a special entity that contains multiple relationships with other entities; and I need recover all relationed entities, including soft deleted ones. It is possible disabled #AdditionalCriteria only in the relations of this special entity with EclipseLink? If not, what is the best option to do this? My code looks like the following:
////Example of my top entity class (all others inherit from this)
#AdditionalCriteria("this.deleted = false")
public abstract class TopHierarchyClass {
···
#Column(name = "deleted")
protected boolean deleted = false;
···
}
//Example of entity that needs recover all relationed entities including soft deleted
#Entity
#Table(name = "special_entities")
public class SpecialEntity extends EntityBase {
···
#JoinColumn(name = "iditem", referencedColumnName = "id")
#ManyToOne(fetch = FetchType.LAZY)
private Item item;
···
}
#Entity
#Table(name = "items")
public class Item extends EntityBase {
···
}
Thanks in advance
Create a new entity for the same table without the #AdditionalCriteria. This way you can retrieve all records from that table without applying the additional filter.
For example:
public interface Person {
Long getId();
void setId(Long id);
Date getDeletedAt();
void setDeletedAt(Date deletedAt);
}
#MappedSupperclass
public class PersonEntity implements Person {
#Id
private Long id;
private Date deletedAt;
public Long getId() { return id; }
public void setId(Long id) { this.id = id; }
public Date getDeletedAt() { return deletedAt; }
public void setDeletedAt(Date deletedAt) { this.deletedAt = deletedAt; }
}
#Entity
#Table(name = "people")
#AdditionalCriteria("this.deletedAt is null")
public class ActivePersonEntity extends PersonEntity {
}
#Entity
#Table(name = "people")
public class RawPersonEntity extends PersonEntity {
}
public class PeopleRepository {
#PersistenceContext
private EntityManager em;
public List<Person> all() {
return all(false);
}
public List<Person> all(boolean includeSoftDeleted) {
if (!includeSoftDeleted) {
return em.createQuery("select p from ActivePersonEntity p", ActivePersonEntity.class).getResultList();
} else {
return em.createQuery("select p from RawPersonEntity p", RawPersonEntity.class).getResultList();
}
}
}
Also, if your #AdditionalCriteria is in a super class, you may override it by declaring a new empty #AdditionalCriteria in a sub class:
You can define additional criteria on entities or mapped superclass.
When specified at the mapped superclass level, the additional criteria
definition applies to all inheriting entities, unless those entities
define their own additional criteria, in which case those defined for
the mapped superclass are ignored.
#AdditionalCriteria doc

Using Pageable to query a collection

I have two entities. A NewsCategory and a NewsItem which have a one-to-many relationship.
NewsCategory
#Entity
public class NewsCategory extends AbstractEntity<Long> {
private String name;
#OneToMany(cascade = CascadeType.ALL)
private List<NewsItem> items = new ArrayList<>();
public String getName() {
return name;
}
public void setName(final String name) {
this.name = name;
}
public List<NewsItem> getItems() {
return items;
}
}
NewsItem
#Entity
public class NewsItem extends AbstractEntity<Long> {
private String title;
private LocalDate startDate;
private LocalDate endDate;
private String resource;
#Column(columnDefinition = "text")
private String content;
// getters and setters...
}
Repository interface
I would like to have the items collection to be pageable but I'm having some difficulties with defining the repository interface for it.
This interface does not work like expected.
public interface NewsCategoryRepository extends JpaRepository<NewsCategory, Long> {
#Query("SELECT e.items FROM #{#entityName} e WHERE e = ?1")
public List<NewsItem> findItems(NewsCategory category, Pageable pageable);
}
When executing findItems() the following exception is thrown.
Caused by: org.hibernate.QueryException: illegal attempt to dereference collection [newscatego0_.id.items] with element property reference [startDate] [SELECT e.items FROM NewsCategory e WHERE e = ?1 order by e.items.startDate asc]
How can I modify the above interface so it will return a portion of the items property using Spring Data and Pageable?

Why entityManager.contains returns different results?

This is in JPA2 (EclipseLink) and JSF2.
I have an entity class Student:
#Entity
public class Student implements Serializable {
private static final long serialVersionUID = 1L;
#Id
#GeneratedValue(strategy = GenerationType.AUTO)
private Long id;
private String firstname;
private String lastname;
private int age;
public Student(String firstname, String lastname, int age) {
this.firstname = firstname;
this.lastname = lastname;
this.age = age;
}
public Student() {
}
// accessors and mutators here
}
Session bean StudentFacade that inherits AbstractFacade:
public abstract class AbstractFacade<T> {
private Class<T> entityClass;
public AbstractFacade(Class<T> entityClass) {
this.entityClass = entityClass;
}
protected abstract EntityManager getEntityManager();
public void create(T entity) {
getEntityManager().persist(entity);
}
public T edit(T entity) {
return getEntityManager().merge(entity);
}
public void remove(T entity) {
getEntityManager().remove(getEntityManager().merge(entity));
}
public T find(Object id) {
return getEntityManager().find(entityClass, id);
}
// other methods: findAll, findRange, count
}
#Stateless
public class StudentFacade extends AbstractFacade<Student> {
#PersistenceContext(unitName = "jpa2testsPU")
private EntityManager em;
#Override
protected EntityManager getEntityManager() {
return em;
}
public StudentFacade() {
super(Student.class);
}
public boolean contains(Student s) {
return getEntityManager().contains(s);
}
public void testContains() {
Student s = find(1L);
boolean isContains = getEntityManager().contains(s);
}
}
This is my JSF Managed Bean:
#ManagedBean
#RequestScoped
public class IndexController {
#EJB
private StudentFacade studentFacade;
/**
* Creates a new instance of IndexController
*/
public IndexController() {
}
public String test() {
Student s = new Student("John", "Doe", 20);
studentFacade.create(s);
Student s1 = studentFacade.find(1L); // This works because table only has 1 record
boolean isContains = studentFacade.contains(s);
return null;
}
}
When I run test() from managed bean, isContains is false. But when testContains() in StudentFacade is called, isContains is true. Why is this?
StudentFacade is a Stateless Session Bean (SSB). The contents of its instance variables are not guaranteed to be preserved across method calls (reference). It's like having a different instance of EntityManager created for each method invocation.
When you run your test from the managed bean, you invoke two different methods on the SSB, therefore a different EntityManager instance is created for each call, and the second one does not contain the Student instance because it has not been loaded yet.
But when you run your test inside a method of the SSB itself, the same EntityManager is used for the scope of the entire method, therefore the call to contains() returns true.

JPA Mapping, parent to children and child to parent with two classes and abstract class

I have have two classes that inherit from an abstract class and have a parent-children relation.
So I use annotation OneToMany and ManyToOne but the parent entity in child class is always null.
Can Someone help me please, I have spend several hours to googling and test many conf without success.
These are code from my classes :
public #Table(name="flowentity") #Entity abstract class FlowEntity {
final static Logger log = LoggerFactory.getLogger(FlowEntity.class);
//Globals informations concerning the flow state
private #Id #GeneratedValue(strategy = GenerationType.IDENTITY) Integer flowId = 0;
private String flowName;
private #OneToMany(fetch=FetchType.EAGER, cascade=CascadeType.ALL)
Set<PeopleEntity> actorSet = new HashSet<>();
//Global parameters for most of flows
//Organizational parameters
private #OneToOne(fetch=FetchType.EAGER, cascade=CascadeType.ALL)
#JoinColumn(name="organisationalEntity_Id")
OrganisationalEntity organisationalEntity;
...
public #Table(name="ams_newCPEntity") #Entity class NewMultiCPEntity extends FlowEntity {
private #OneToMany(targetEntity=NewCPEntity.class, fetch=FetchType.EAGER, cascade=CascadeType.ALL,mappedBy="parent")
Set<NewCPEntity> cpList = new HashSet<NewCPEntity>();
//Constructor
public NewMultiCPEntity(){
setFlowName(EnumFlow.N_CP_M.getFlowAcronym());
}
...
public #Table(name="ams_newCPEntity") #Entity class NewCPEntity extends FlowEntity {
final static Logger log = LoggerFactory.getLogger(NewCPEntity.class);
private boolean formNCPValidated;
private #ManyToOne #JoinColumn(name="parent_Id", nullable=false)
NewMultiCPEntity parent;
public NewCPEntity(){
log.debug("Instanciation of a new CP");
setFlowName(EnumFlow.N_CP.getFlowAcronym());
}
public #Override OrganisationalEntity getOrganisationalEntity(){
return parent.getOrganisationalEntity();
}
...
If I don't add the #JoinColumn annotation, JPA create an association table but is not able to retrieve the parent whereas the association can be done directly by requesting in database.
Thankyou very much to help.
Regards,
Thank you Chris for your comment, you are right, I forget to change the name of the table. I don't think it was the problem because the inheritance mapping is in one table flowentity with a DTYPE discriminator column.
Finally I resolve my problem by setting parent attributs when adding a new child like this :
public #Table #Entity class NewMultiCPEntity extends FlowEntity {
private #OneToMany(targetEntity=NewCPEntity.class, fetch=FetchType.EAGER, cascade=CascadeType.ALL)
List<NewCPEntity> cpList = new ArrayList<>();
//Constructor
public NewMultiCPEntity(){
setOrganisationalEntity(new OrganisationalEntity());
setFlowName(EnumFlow.N_CP_M.getFlowAcronym());
}
public List<NewCPEntity> getNCPList(){
if(cpList == null){
cpList = new ArrayList<>();
}
if(cpList.isEmpty()){
addCPEntity(new NewCPEntity());
}
return Collections.unmodifiableList(cpList);}
public boolean removeCPEntity(NewCPEntity entity){
return cpList.remove(entity);
}
public boolean addCPEntity(NewCPEntity entity){
entity.setParent(this);
entity.setOrganisationalEntity(this.getOrganisationalEntity());
return cpList.add(entity);
}
And I remove the override of getOrganizationalEntity in the child :
public #Table #Entity class NewCPEntity extends FlowEntity {
final static Logger log = LoggerFactory.getLogger(NewCPEntity.class);
private #ManyToOne(targetEntity=NewMultiCPEntity.class,cascade=CascadeType.ALL)
NewMultiCPEntity parent;
public NewCPEntity(){
log.debug("Instanciation of a new CP");
setFlowName(EnumFlow.N_CP.getFlowAcronym());
}
public NewMultiCPEntity getParent() {
return parent;
}
public void setParent(NewMultiCPEntity parent){
this.parent = parent;
}
Regards,

#IdClass JPA Annotation

Here is a composite primary key for one of my entities.
public class GroupMembershipPK implements Serializable{
private static final long serialVersionUID = 7816433876820743311L;
private User user;
private Group group;
public GroupMembershipPK(){
}
public boolean equals(Object o){
if(o==null){
return false;
}
if(o instanceof GroupMembershipPK){
final GroupMembershipPK groupMembershipPK=(GroupMembershipPK)o;
return groupMembershipPK.group.getGroupName().equals(this.group.getGroupName()) &&
groupMembershipPK.user.getName().equals(this.user.getName());
}
return false;
}
public int hashCode(){
return super.hashCode();
}
}
Heres my entity(part of) using the above as a Composite Primary Key.
#Entity
#IdClass(GroupMembershipPK.class)
public class GroupMembership extends AbstractModelElementVersionOther{
private static final long serialVersionUID = 9188465566607060376L;
private String memType;
private Group group;
private User user;
public GroupMembership(){
super();
}
#Column(nullable=false)
public String getMemType(){
return this.memType;
}
public void setMemType(String memType){
this.memType=memType;
}
#Id
#ManyToOne
#JoinColumn(name="groupId")
public Group getGroup(){
return this.group;
}
public void setGroup(Group group){
this.group=group;
}
#Id
#ManyToOne
#JoinColumn(name="userId")
public User getUser(){
return this.user;
}
public void setUser(User user){
this.user=user;
}
#Override
public boolean equals(Object o) {
//
I am a little confused on what should be the equals method implementation for the above entity. How can I compare two composite primary keys?
Any comments on other parts of my code are welcome too.
It is not a good idea you store entities as primary key. There is some limitations when using query language and JPA 1.0 does not support. Besides it there is no need to use entities as primary key. Think about it.If you want, take a special look at the following question
A class that behaves like #Entity and #Embeddable
Answer one
Comment about answer one
You will see that using a entity as primary key is not necessary.
Instead of
public class GroupMembershipPK implements Serializable {
private User user;
private Group group;
}
Use
public class GroupMembershipPK implements Serializable {
private Integer userId;
private Integer groupId;
}
equals implementation is important because JPA compares two entities by using it (JPA checks whether an entity is in persistence context by using equals implementation). So you can implement according to
public boolean equals(Object o) {
if(o == null)
return false;
if(!(o instanceof GroupMembershipPK))
return false;
GroupMembershipPK other = (GroupMembershipPK) o;
if(!(getUserId().equals(other.getUserId()))
return false;
if(!(getGroupId().equals(other.getGroupId()))
return false;
return true;
}
Advice: it is a good idea you use property access instead of field access because, at some times, JPA implementation uses a proxy object because of performance issues. A proxy object makes use of property access because it allows JPA implementation hits the database when necessary.
How to save an object that uses a composite primary key ?
User user = new user();
Group group = new Group();
entityManager.save(user);
entityManager.save(group);
entityManager.flush();
UserGroup userGroup = new UserGroup();
userGroup.setId(new UserGroup.UserGroupId(user.getId(), group.getId()));
entityManager.save(userGroup);
Do you want to know how to implement UserGroup ?
public class UserGroup {
private UserGroupId id;
// You can create UserGroupId outside UserGroup class
// Feel free to choice your best approach
#Embeddable
public static class UserGroupId implements Serializable {
private Integer userId;
private Integer groupId;
// required no-arg constructor
public UserGroupId() {}
public UserGroupId(Integer userId, Integer groupId) {
this.userId = userId;
this.roupId = groupId;
}
// getter's and setter's
// equals and hashcode as shown above
}
#EmbeddedId
public UserGroupId getId() {
return this.id;
}
public setId(UserGroupId id) {
this.id = id;
}
}
Another approach to use composite primary key is IdClass. See IdClass
regards,
It was mentioned partially, anyway:
When implementing the equals method, you should use instanceof to allow comparing with subclasses. If Hibernate lazy loads a one to one or many to one relation, you will have a proxy for the class instead of the plain class. A proxy is a subclass. Comparing the class names would fail.
More technically, you should follow the Liskovs Substitution Principle and ignore symmetry.
The next pitfall is using something like name.equals(that.name) instead of name.equals(that.getName()). The first will fail, if that is a proxy.
http://www.laliluna.de/jpa-hibernate-guide/ch06s06.html