JPA OneToOne and shared primary key need manual assignment - spring-boot

I'm using Springboot and JPA to create two tables sharing the same primary key.
For the first table I write:
public class UserAccount implements Serializable
{
#Id
#GeneratedValue(strategy = GenerationType.IDENTITY)
private Long id;
#OneToOne(mappedBy ="user", cascade = {CascadeType.REMOVE, CascadeType.MERGE,
CascadeType.REFRESH}, fetch=FetchType.LAZY)
#PrimaryKeyJoinColumn
private UserLogin login;
}
For the second table I write:
public class UserLogin implements Serializable
{
#Id
private Long user_id;
#OneToOne(cascade = {CascadeType.MERGE, CascadeType.REFRESH},
fetch=FetchType.LAZY)
#MapsId("user_id")
#JoinColumn(name = "user_id", referencedColumnName = "id")
#Setter(AccessLevel.NONE)
private UserAccount user;
public void setUser(UserAccount user)
{
this.user = user;
this.user_id = user.getId();
}
}
Other stuff are omitted for conciseness. The code works because I manually set the id of UserLogin by writing the statement
this.user_id = user.getId();
otherwise I get the error:
Hibernate error: ids for this class must be manually assigned before calling save():
I guess that the ids can be manually managed but I cannot get the right configuration.
UPDATE:
I found the solution thanks (see the accepted answer). Now I just would get rid of the findById() when setting the user login.
//these methods are defined within a dedicated #Service
#Transactional
public void createLoginInfo(UserAccount user)
{
UserLogin userlogin=new UserLogin();
this.addLoginToUser(userlogin,user);
loginService.save(userlogin);
}
#Transactional
public void addLoginToUser(UserLogin login, UserAccount account)
{
//whit this commented line works
//UserAccount acc= this.findById(account.getId());
login.setUser(account);
account.setLogin(login);
}
//In a transactional test method I first create the user then I call
userService.save(theuser);
userService.createLoginInfo(theuser);

You have a bidirectional relationship, but have mapped it with a few competing options that don't work well together. First, in UserAccount, it isn't clear why you have an ID that is generated, yet try to also map it with the relationship (specifically using a PrimaryKeyJoinColumn). If you want it generated, it can't also be a foreign key value in a reference - and you've already got this relationship setup as the 'other' side via the 'mappedBy' setting. It should just be:
public class UserAccount implements Serializable {
#Id
#GeneratedValue(strategy = GenerationType.IDENTITY)
private Long id;
#OneToOne(mappedBy ="user", cascade = {CascadeType.REMOVE, CascadeType.MERGE,
CascadeType.REFRESH}, fetch=FetchType.LAZY)
private UserLogin login;
}
User login then should just be:
public class UserLogin implements Serializable {
#Id
private Long user_id;
#OneToOne(cascade = {CascadeType.MERGE, CascadeType.REFRESH},
fetch=FetchType.LAZY)
#MapsId("user_id")
#JoinColumn(name = "user_id", referencedColumnName = "id")
#Setter(AccessLevel.NONE)
private UserAccount user;
public void setUser(UserAccount user) {
this.user = user;
}
}
Note because you have the mapsId annotation on the user relationship, JPA will set the user_id property for you once the ID is assigned - there is no need to manually set it yourself. You can, but if you do you must insure it was assigned previously - which requires a save/flush on the UserAccount. If you don't actually use the Long user_id property, you don't really even need to map it; you can just mark the user property as the ID:
public class UserLogin implements Serializable {
#Id
#OneToOne(cascade = {CascadeType.MERGE, CascadeType.REFRESH},
fetch=FetchType.LAZY)
#JoinColumn(name = "user_id", referencedColumnName = "id")
#Setter(AccessLevel.NONE)
private UserAccount user;
public void setUser(UserAccount user) {
this.user = user;
}
}
The Long ID from UserAccount then can be used to lookup UesrAccounts and UserLogin instances.

Try this :
public class UserLogin implements Serializable
{
#Id
private Long user_id;
#OneToOne(fetch=FetchType.LAZY)
#MapsId
#JoinColumn(name = "user_id")
private UserAccount user;
public UserAccount getUser() {
return user;
}
public void setUser(UserAccount user) {
this.user = user;
}
}
public class UserAccount implements Serializable
{
#Id
#GeneratedValue(strategy = GenerationType.IDENTITY)
private Long id;
}
To persist UserLogin :
EntityManager em;
UserAccount user = em.find(UserAccount.class, 1L)
UserLogin login = new UserLogin();
login.setUser(user);
em.persist(login);

Related

Confused why getting a User from Repository fixed "failed to lazily initialize a collection of role" compared to using SecurityContextHolder

My goal was to pass a List of Businesses to the model from the controller to display it in a view and I have succeeded, but have a bit of confusion.
When I initially tried using:
public User getCurrentAuthenticatedUser() {
UserDetailsImpl user = (UserDetailsImpl) SecurityContextHolder.getContext().getAuthentication().getPrincipal();
return user.getUser();
}
#GetMapping("")
public String list(Model model) {
model.addAttribute("businesses", userService.getCurrentAuthenticatedUser().getBusinesses());
return "business/list";
}
I got this error: "failed to lazily initialize a collection of role: com.xyz.User.businesses could not initialize proxy - no Session"
Then I tried:
#GetMapping("")
public String list(Model model) {
int userId = userService.getCurrentAuthenticatedUser().getId();
User user = userService.getById(userId); // gets User using Spring Data JPA UserRepository
List<Business> businesses = user.getBusinesses();
model.addAttribute("businesses", businesses);
return "business/list";
}
And this worked perfectly fine.
What was the issue using the first method. It seemed more simple rather than calling a User from the UserRepository. I've seen some posts that say you should use EAGER fetching, but that's just seems like a bandaid solution.
From the beginner's understanding: Since fetch type is LAZY the businesses don't exist yet in the User but are fetched on demand later on so there shouldn't be an issue.
Edit: After more thought I remembered that with basic Hibernate you would have to create Transactions and commit transactions. I'm assuming that User is not within a Transaction that's why I can't get businesses using the 1st method.
What would be a better solution to fetch the current Authenticated user? And that user's attributes such as a list of businesses.
Model Classes:
Business:
#Entity
#Table(name = "businesses")
public class Business {
#Id
#GeneratedValue(strategy = GenerationType.IDENTITY)
private Long id;
private String name;
private String description;
private LocalDate date;
#ManyToOne(cascade={CascadeType.MERGE})
#JoinColumn(name="user_id")
private User user;
public Business() {
}
public Business(String name, String description, LocalDate date, User user) {
...
}
public Business(Long id, String name, String description, LocalDate date, User user) {
...
}
... getters/setters
}
USER:
#Entity
#Table(name = "users")
public class User {
#Id
#GeneratedValue(strategy = GenerationType.IDENTITY)
private int id;
private String username;
private String password;
private boolean enabled;
#ManyToMany(cascade = CascadeType.ALL, fetch = FetchType.EAGER)
#JoinTable( name = "users_roles",
joinColumns = #JoinColumn(name = "user_id"),
inverseJoinColumns = #JoinColumn(name = "role_id"))
private Set<Role> roles = new HashSet<>();
#OneToMany(fetch = FetchType.LAZY, mappedBy="user", cascade={CascadeType.MERGE})
private List<Business> businesses;
... getters/setters
}

Spring JPA Unable To Find Composite Foreign Key Target Column (Non-PK)

User.java
#Entity
#Table(name = "users")
public class User implements Serializable {
#Id
#GeneratedValue(strategy = GenerationType.IDENTITY)
private Integer id;
#ManyToOne(fetch = FetchType.LAZY)
#JoinColumn(name = "user_role_id", referencedColumnName = "id")
private UserRole userRole;
}
UserRole.java
#Data
#Entity
#Table(name = "user_roles")
public class UserRole implements Serializable {
#Id
#GeneratedValue(strategy = GenerationType.IDENTITY)
private Integer id;
}
Client.java
#Data
#Entity
#Table(name = "clients")
public class Client implements Serializable {
#Id
#GeneratedValue(strategy = GenerationType.IDENTITY)
private Integer id;
#OneToOne(cascade = CascadeType.ALL)
#JoinColumns({ #JoinColumn(name = "user_id", referencedColumnName = "id"),
#JoinColumn(name = "user_role_id", referencedColumnName = "user_role_id") })
private User user;
}
Error
org.springframework.beans.factory.BeanCreationException: Error creating bean with name 'entityManagerFactory' defined in class path resource [org/springframework/boot/autoconfigure/orm/jpa/HibernateJpaConfiguration.class]: Invocation of init method failed; nested exception is org.hibernate.MappingException: Unable to find column with logical name: user_role_id in users
In RDBMS, users.(id, user_role_id) is unique so clients table can refer to that.
Last time, I was using insertable = false, updatable = false on user_role_id, but when I want to add records of new client, I always need to add user_role_id manually user.setUserRoleId(userRole.getId()) after user.setUserRole(userRole) and I think that is bad practice of ORM (it should be added automatically when I set user.setUserRole(userRole))
#Column(name = "user_role_id", insertable = false, updatable = false)
private Integer userRoleId;
What should I do so the relation can be mapped in Spring JPA? and what is the best practice?
In other words, this is also mean how to reference to foreign key generated logical name column?
OK! Please try following configuration:
Below is a important code part and under this link you may find repository with working example
UserRole.java
#Data
#Entity
#Table(name = "user_roles")
public class UserRole implements Serializable {
#Id
#GeneratedValue(strategy = GenerationType.IDENTITY)
#Column(name = "role_id")
private Integer roleId;
}
User.java
#Data
#Entity
#Table(name = "users")
public class User implements Serializable {
#Id
#GeneratedValue(strategy = GenerationType.IDENTITY)
#Column(name = "user_id")
private Integer userId;
#ManyToOne(fetch = FetchType.LAZY, cascade = CascadeType.ALL)
#JoinColumn(name = "user_role_id", referencedColumnName = "role_id")
private UserRole userRole;
}
Client.java
#Data
#Entity
#Table(name = "clients")
public class Client implements Serializable {
#Id
#GeneratedValue(strategy = GenerationType.AUTO)
#Column(name = "client_id")
private Integer clientId;
#OneToOne(fetch = FetchType.LAZY)
#JoinColumns(
value = {
#JoinColumn(name = "client_role_id", referencedColumnName = "user_role_id"),
#JoinColumn(name = "client_user_id", referencedColumnName = "user_id"),
}
,
foreignKey = #ForeignKey(
name = "FK_user_with_role",
foreignKeyDefinition = "FOREIGN KEY (client_user_id, client_role_id)\n" +
" REFERENCES users \n" +
" (user_id, user_role_id) \n" +
" ON UPDATE CASCADE\n" +
" ON DELETE CASCADE")
)
private User user;
}
Please note that beside adding a foreignKey in the Client implementation, you MUST keep the sequence of #JoinColum annotations.. I don't know what is the reason behind, but if you flip those lines you'll still get your error as it was before :)
EDIT: I've added another answer which fits best in my opinion. I'm leaving this one as well to see the other steps I tried.
Though the solution is not elegant and not using JPA as requested. Just in case anything in here would be helpful
If I understand the main issue correctly - you want to bind Client entity with Role entity via User entity, by first setting User's Role and then transfer that "property" by using only UserId instead setting additionally RoleId while creating Client.
Basically after playing for a while with your model I think the main issue is to assign data to each other within a #Transactional methods. That seems to be caused ba Lazy fetch strategy.
My proposal for solution that binds all your Entities according expectations differs only from yours with ommiting the RoleId JoinColumn in Clients table. I have checked that when calling a service that would have #Transactional methods, you can assign a Role to the User and User to the Client with simple user.setRole(roleEntity) followed by client.setUser(userEntity).
All the data is then consistent. No need to call further like getters and setters as you mentioned in the second part of your question. Question is if for any reason you need to have RoleId as well in your Clients Table, then this soultion would have to be enhanced by additional column?
UserRole.java
#Entity
#Table(name = "user_roles")
public class UserRole implements Serializable {
#Id
#GeneratedValue(strategy = GenerationType.IDENTITY)
#Column(name = "role_id")
private Integer roleId;
//getters and setters and toString
}
User.java
#Entity
#Table(name = "users")
public class User implements Serializable {
#Id
#GeneratedValue(strategy = GenerationType.IDENTITY)
#Column(name = "user_id")
private Integer userId;
#ManyToOne(fetch = FetchType.LAZY)
#JoinColumn(name = "user_role_id", referencedColumnName = "role_id")
private UserRole userRole;;
//getters and setters and toString;
}
Client.java
#Entity
#Table(name = "clients")
public class Client implements Serializable {
#Id
#GeneratedValue(strategy = GenerationType.AUTO)
#Column(name = "client_id")
private Integer clientId;
#OneToOne(fetch = FetchType.LAZY)
#JoinColumns({
#JoinColumn(name = "client_user_id", referencedColumnName = "user_id"),
})
private User user;
#Column(name = "client_role_id")
private Integer roleId;
#PrePersist
#PreUpdate
private void prePersist(){
try {
roleId = getUser().getUserRole().getRoleId();
} catch (NullPointerException e){
roleId = null;
}
}
//getters and setters and toString
}
UserService.java
#Service
public class UserService {
UserRepo userRepo;
public UserService(UserRepo userRepo) {
this.userRepo = userRepo;
}
#Transactional
public void save(User user) {
userRepo.save(user);
}
#Transactional
public User getReferenceById(int i) {
return userRepo.getReferenceById(i);
}
}
ClientService.java
#Service
public class ClientService {
private ClientRepo clientRepo;
private UserService userService;
public ClientService(ClientRepo clientRepo, UserService userService) {
this.clientRepo = clientRepo;
this.userService = userService;
}
#Transactional
public Client save(Client client){
return clientRepo.save(client);
}
#Transactional
public Client getReferenceById(int i) {
return clientRepo.getReferenceById(i);
}
#Transactional
public void printClient(Client client){
client = clientRepo.getReferenceById(client.getClientId());
System.out.println(client);
}
#Transactional
public void bindUserToClient(int userId, int clientId) {
Client entity = clientRepo.findById(clientId).orElseGet(Client::new);
entity.setUser(userService.getReferenceById(userId));
}
#Transactional
public void printClient(int i) {
clientRepo.findById(i).ifPresentOrElse(this::printClient, EntityNotFoundException::new);
}
}
This configuration after running this commandLineRunner:
#Configuration
public class Config {
#Bean
#Transactional
public CommandLineRunner commandLineRunner(
#Autowired UserRoleRepo roleRepo,
#Autowired UserService userService,
#Autowired ClientService clientService
) {
return args -> {
for (int i = 0; i < 5; i++) {
roleRepo.save(new UserRole());
}
for (int i = 5; i > 0; i--) {
User user = new User();
user.setUserRole(roleRepo.getReferenceById(i));
userService.save(user);
}
Client client = new Client();
client.setUser(userService.getReferenceById(2));
client = clientService.save(client);
clientService.printClient(client);
client = new Client();
client.setClientId(1);
clientService.printClient(client);
int userId = 5;
clientService.bindUserToClient(userId, 1);
clientService.printClient(1);
};
}
}
gave me correct output in the console:
Client{id=1, user=User{id=2, userRole=UserRole{id=4}}}
Client{id=1, user=User{id=2, userRole=UserRole{id=4}}}
Client{id=1, user=User{id=5, userRole=UserRole{id=1}}}
WORKAROUND
I tried to reach the goal by use of Spring JPA but could'nt.
The workaround that keeps the referential integrity was by creating a constrains through DB like below and add #PrePersist and #PreUpdate annotated method which is updating the client's roleId as intended.
create table clients
(
client_id integer not null,
client_user_id integer,
client_role_id integer,
primary key (client_id)
);
create table user_roles
(
role_id integer generated by default as identity,
primary key (role_id)
);
create table users
(
user_id integer generated by default as identity,
user_role_id integer,
primary key (user_id),
CONSTRAINT User_Role UNIQUE (user_id, user_role_id)
);
alter table users
add constraint FK_role_id foreign key (user_role_id) references user_roles (role_id);
alter table clients
add constraint FK_user_id foreign key (client_user_id, client_role_id) references users (user_id, user_role_id) on update cascade ;
Thanks to that I could for instance update userRole in user entity, and the change was reflected in the clients table as well without any further actions

Cascading of persist, does not create identifiaction

Having this entities:
User.java:
#Entity
#NoArgsConstructor
#Getter
#Setter
public class User {
#Id #GeneratedValue
private Long id;
private String username;
#OneToMany(mappedBy = "owner", cascade = CascadeType.ALL, fetch = FetchType.EAGER)
#MapKey(name = "friend_id")
private Map<User, Friendship> friends = new HashMap<>();
}
Friendship.java:
#Entity
#Data
#IdClass(Friendship.class)
public class Friendship implements Serializable {
#Id
private Long owner_id;
#Id
private Long friend_id;
private String level;
#ManyToOne(cascade = CascadeType.ALL)
#MapsId("owner_id")
private User owner;
#ManyToOne(cascade = CascadeType.ALL)
#MapsId("friend_id")
private User friend;
}
and DemoApplication.java:
#Bean
public CommandLineRunner loadData(UserRepository userRepo){
return new CommandLineRunner() {
#Override
public void run(String... args) throws Exception {
User owner = new User();
owner.setUsername("owner");
User f1 = new User();
f1.setUsername("f1");
User f2 = new User();
f2.setUsername("f2");
Friendship fs1 = new Friendship();
fs1.setOwner(owner);
fs1.setFriend(f1);
Friendship fs2 = new Friendship();
fs2.setOwner(owner);
fs2.setFriend(f2);
owner.getFriends().put(f1, fs1);
owner.getFriends().put(f2, fs2);
userRepo.saveAndFlush(owner);
}
};
}
I get error:
A different object with the same identifier value was already associated with the session : [com.example.demo.model.Friendship#Friendship(owner_id=null, friend_id=null, level=null, owner=com.example.demo.model.User#2b036135, friend=com.example.demo.model.User#a9e28af9)]
Which means both Users f1 and f2, are having null in Long id. The indeed have, when the object is created, but I thought the mapping had specified CascadeType.ALL and #GeneratedValue so the if should be created.
But I had try to set the ids myself:
...
f1.setUsername("f1");
f1.setId(1L);
User f2 = new User();
f2.setUsername("f2");
f2.setId(2L);
...
But now I got
detached entity passed to persist: com.example.demo.model.User
So I guess I should let the creation of primary keys on JPA. But as you saw from above, it does not that even with Cascading. So what now?
Try adding this under your #Id annotation
#Id
#GeneratedValue(strategy=GenerationType.IDENTITY)

OneToOne CascadeType in spring data jpa

I use OneToOne in the spring data JPA and I want to delete a record from the Address table without touching the user. But I can't.
If I remove User, in this case Address is removed, that's good.
But how can you delete an Address without touching the User?
https://github.com/myTestPercon/TestCascade
User.Java
#Entity
#Table(name = "user", schema = "testCascade")
public class User implements Serializable {
#Id
#GeneratedValue(strategy = GenerationType.IDENTITY)
#Column(name = "id")
private Long id;
#Column(name = "name")
private String name;
#OneToOne(mappedBy = "user", cascade = CascadeType.ALL)
private Address address;
// Getter and Setter ...
}
Address.java
#Entity
#Table(name = "address", schema = "testCascade")
public class Address implements Serializable {
#Id
private Long id;
#Column(name = "city")
private String city;
#OneToOne
#MapsId
#JoinColumn(name = "id")
private User user;
// Getter and Setter ...
}
DeleteController.java
#Controller
public class DeleteController {
#Autowired
ServiceJpa serviceJpa;
#GetMapping(value = "/deleteAddressById")
public String deleteAddressById () {
serviceJpa.deleteAddressById(4L);
return "redirect:/home";
}
}
You got your mapping wrong thats all is the problem .
try the below and see
User.java
#Entity
#Table(name = "user", schema = "testCascade")
public class User implements Serializable {
#Id
#GeneratedValue(strategy = GenerationType.IDENTITY)
#Column(name = "id")
private Long id;
#Column(name = "name")
private String name;
#OneToOne(cascade=CascadeType.ALL)
#JoinColumn(name="foriegn key column in user table for address example.. address_id")
private Address address;
// Getter and Setter ...
}
Address.java
#Entity
#Table(name = "address", schema = "testCascade")
public class Address implements Serializable {
#Id
private Long id;
#Column(name = "city")
private String city;
//name of the address variable in your user class
#OneToOne(mappedBy="address",
cascade={CascadeType.DETACH, CascadeType.MERGE, CascadeType.PERSIST,
CascadeType.REFRESH})
private User user;
// Getter and Setter ...
}
In order to solve this problem, you need to read the hibernate Documentation Hibernate Example 162, Example 163, Example 164.
And also I recommend to look at this is Using #PrimaryKeyJoinColumn annotation in spring data jpa
This helped me in solving this problem.
And also you need to specify the parameter orphanRemoval = true
User.java
#Entity(name = "User")
#Table(name = "user", schema = "testother")
public class User implements Serializable {
#Id
#GeneratedValue(strategy = GenerationType.IDENTITY)
#Column(name = "id")
private Long id;
#Column(name = "name")
private String name;
#OneToOne(mappedBy = "user", cascade = CascadeType.ALL, orphanRemoval = true)
private Address address;
public void addAddress(Address address) {
address.setUser( this );
this.address = address;
}
public void removeAddress() {
if ( address != null ) {
address.setUser( null );
this.address = null;
}
}
// Getter and Setter
}
Address.java
#Entity(name = "Address")
#Table(name = "address", schema = "testother")
public class Address implements Serializable {
#Id
private Long id;
#Column(name = "city")
private String city;
#OneToOne
#MapsId
#JoinColumn(name = "id")
private User user;
// Getter and Setter
}
DeleteController .java
#Controller
public class DeleteController {
#Autowired
ServiceJpa serviceJpa;
#GetMapping(value = "/deleteUser")
public String deleteUser () {
User user = serviceJpa.findUserById(2L).get();
user.removeAddress();
serviceJpa.saveUser(user);
return "/deleteUser";
}
}
Or make a custom SQL query.
#Repository
public interface DeleteAddress extends JpaRepository<Address, Long> {
#Modifying
#Query("delete from Address b where b.id=:id")
void deleteBooks(#Param("id") Long id);
}
public class Address {
#Id
private Long id;
#MapsId
#JoinColumn(name = "id")
private User user;
}
Rename #JoinColumn(name = "id") to #JoinColumn(name = "user_id")
You can't say that the column that will point to user will be the id of the Address

how to insert data in one to many relationship in spring jpa

I am new in Spring. I want to insert the data into the AUDIT table. There is a one-two-many relationship between USER and AUDIT tables.
For every login and logout data should be inserted into the AUDIT table. How can I do that?
The user entity:
#Table(name="USER")
public class User {
#Id
#GeneratedValue(strategy=GenerationType.AUTO)
#Column(name="USERID")
private Long id;
private String username;
private String email;
private Long phone;
private String password;
private int OTP;
private boolean activation_flag;
#OneToMany(cascade=CascadeType.ALL)
#JoinColumn(name="USERID")
private Set<Audit> audit;
#ManyToOne
#JoinColumn(name="ADDRESSID")
private Address address;
#ManyToMany(fetch = FetchType.EAGER)
#JoinTable(
name = "USER_AUTHORITY",
joinColumns = {#JoinColumn(name = "USER_ID", referencedColumnName = "USERID")},
inverseJoinColumns = {#JoinColumn(name = "AUTHORITY_ID", referencedColumnName = "ID")})
private List<Authority> authorities;
The audit entity:
#Entity
#Table(name="AUDIT")
public class Audit {
#Id
#GeneratedValue(strategy=GenerationType.AUTO)
#Column(name="AUDITID")
private long auditid;
#Column(name="GEOLOCATION")
private String geolocation;
#Column(name="ACTION")
private String action;
#Column(name="DATETIME")
private Date datetime;
#Column(name="DEVICEID")
private long deviceid;
#Column(name="ACTIONSUCCESS_FAIL")
private boolean actionsuccess_fail;
#Column(name="JWT_TOKEN")
#Type(type="text")
private String JWT_token;
#ManyToOne
private User user;
The audit repository:
public interface AuditRepository extends JpaRepository<Audit, Long> {
}
The user repository:
public interface UserRepository extends JpaRepository<User, Long> {
User findByUsername(String username);
User findByEmail(String email);
User findByPhone(Long phone);
User findById(Long id);
}
Add the audit to the user you want the audit associated with, then save.

Resources