Infinite Recursion with Jackson JSON and Hibernate

Question

When trying to convert a JPA object that has a bi-directional association into JSON, I keep getting

org.codehaus.jackson.map.JsonMappingException: Infinite recursion (StackOverflowError)

All I found is this thread which basically concludes with recommending to avoid bi-directional associations. Does anyone have an idea for a workaround for this spring bug?

------ EDIT 2010-07-24 16:26:22 -------

Codesnippets:

Business Object 1:

@Entity
@Table(name = "ta_trainee", uniqueConstraints = {@UniqueConstraint(columnNames = {"id"})})
public class Trainee extends BusinessObject {

    @Id
    @GeneratedValue(strategy = GenerationType.TABLE)
    @Column(name = "id", nullable = false)
    private Integer id;

    @Column(name = "name", nullable = true)
    private String name;

    @Column(name = "surname", nullable = true)
    private String surname;

    @OneToMany(mappedBy = "trainee", fetch = FetchType.EAGER, cascade = CascadeType.ALL)
    @Column(nullable = true)
    private Set<BodyStat> bodyStats;

    @OneToMany(mappedBy = "trainee", fetch = FetchType.EAGER, cascade = CascadeType.ALL)
    @Column(nullable = true)
    private Set<Training> trainings;

    @OneToMany(mappedBy = "trainee", fetch = FetchType.EAGER, cascade = CascadeType.ALL)
    @Column(nullable = true)
    private Set<ExerciseType> exerciseTypes;

    public Trainee() {
        super();
    }

    ... getters/setters ...

Business Object 2:

import javax.persistence.*;
import java.util.Date;

@Entity
@Table(name = "ta_bodystat", uniqueConstraints = {@UniqueConstraint(columnNames = {"id"})})
public class BodyStat extends BusinessObject {

    @Id
    @GeneratedValue(strategy = GenerationType.TABLE)
    @Column(name = "id", nullable = false)
    private Integer id;

    @Column(name = "height", nullable = true)
    private Float height;

    @Column(name = "measuretime", nullable = false)
    @Temporal(TemporalType.TIMESTAMP)
    private Date measureTime;

    @ManyToOne(fetch = FetchType.EAGER, cascade = CascadeType.ALL)
    @JoinColumn(name="trainee_fk")
    private Trainee trainee;

Controller:

import org.slf4j.Logger;
import org.slf4j.LoggerFactory;
import org.springframework.beans.factory.annotation.Autowired;
import org.springframework.stereotype.Controller;
import org.springframework.ui.Model;
import org.springframework.web.bind.annotation.RequestBody;
import org.springframework.web.bind.annotation.RequestMapping;
import org.springframework.web.bind.annotation.RequestMethod;
import org.springframework.web.bind.annotation.ResponseBody;

import javax.servlet.http.HttpServletResponse;
import javax.validation.ConstraintViolation;
import java.util.*;
import java.util.concurrent.ConcurrentHashMap;

@Controller
@RequestMapping(value = "/trainees")
public class TraineesController {

    final Logger logger = LoggerFactory.getLogger(TraineesController.class);

    private Map<Long, Trainee> trainees = new ConcurrentHashMap<Long, Trainee>();

    @Autowired
    private ITraineeDAO traineeDAO;

    /**
     * Return json repres. of all trainees
     */
    @RequestMapping(value = "/getAllTrainees", method = RequestMethod.GET)
    @ResponseBody        
    public Collection getAllTrainees() {
        Collection allTrainees = this.traineeDAO.getAll();

        this.logger.debug("A total of " + allTrainees.size() + "  trainees was read from db");

        return allTrainees;
    }    
}

JPA-implementation of the trainee DAO:

@Repository
@Transactional
public class TraineeDAO implements ITraineeDAO {

    @PersistenceContext
    private EntityManager em;

    @Transactional
    public Trainee save(Trainee trainee) {
        em.persist(trainee);
        return trainee;
    }

    @Transactional(readOnly = true)
    public Collection getAll() {
        return (Collection) em.createQuery("SELECT t FROM Trainee t").getResultList();
    }
}

persistence.xml

<persistence xmlns="http://java.sun.com/xml/ns/persistence"
             xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
             xsi:schemaLocation="http://java.sun.com/xml/ns/persistence http://java.sun.com/xml/ns/persistence/persistence_1_0.xsd"
             version="1.0">
    <persistence-unit name="RDBMS" transaction-type="RESOURCE_LOCAL">
        <exclude-unlisted-classes>false</exclude-unlisted-classes>
        <properties>
            <property name="hibernate.hbm2ddl.auto" value="validate"/>
            <property name="hibernate.archive.autodetection" value="class"/>
            <property name="dialect" value="org.hibernate.dialect.MySQL5InnoDBDialect"/>
            <!-- <property name="dialect" value="org.hibernate.dialect.HSQLDialect"/>         -->
        </properties>
    </persistence-unit>
</persistence>

Answer 1

This worked perfectly fine for me. Add the annotation @JsonIgnore on the child class where you mention the reference to the parent class.

@ManyToOne
@JoinColumn(name = "ID", nullable = false, updatable = false)
@JsonIgnore
private Member member;

Answer 2

Be sure you use com.fasterxml.jackson everywhere. I spent much time to find it out.

<properties>
  <fasterxml.jackson.version>2.9.2</fasterxml.jackson.version>
</properties>

<!-- https://mvnrepository.com/artifact/com.fasterxml.jackson.core/jackson-annotations -->
<dependency>
  <groupId>com.fasterxml.jackson.core</groupId>
    <artifactId>jackson-annotations</artifactId>
    <version>${fasterxml.jackson.version}</version>
</dependency>

<!-- https://mvnrepository.com/artifact/com.fasterxml.jackson.core/jackson-databind -->
<dependency>
  <groupId>com.fasterxml.jackson.core</groupId>
    <artifactId>jackson-databind</artifactId>
    <version>${fasterxml.jackson.version}</version>
</dependency>

Then use @JsonManagedReference and @JsonBackReference.

Finally, you can serialize your model to JSON:

import com.fasterxml.jackson.databind.ObjectMapper;

ObjectMapper mapper = new ObjectMapper();
String json = mapper.writeValueAsString(model);

Answer 3

For me the best solution is to use @JsonView and create specific filters for each scenario. You could also use @JsonManagedReference and @JsonBackReference, however it is a hardcoded solution to only one situation, where the owner always references the owning side, and never the opposite. If you have another serialization scenario where you need to re-annotate the attribute differently, you will not be able to.

Problem

Lets use two classes, Company and Employee where you have a cyclic dependency between them:

public class Company {

    private Employee employee;

    public Company(Employee employee) {
        this.employee = employee;
    }

    public Employee getEmployee() {
        return employee;
    }
}

public class Employee {

    private Company company;

    public Company getCompany() {
        return company;
    }

    public void setCompany(Company company) {
        this.company = company;
    }
}

And the test class that tries to serialize using ObjectMapper (Spring Boot):

@SpringBootTest
@RunWith(SpringRunner.class)
@Transactional
public class CompanyTest {

    @Autowired
    public ObjectMapper mapper;

    @Test
    public void shouldSaveCompany() throws JsonProcessingException {
        Employee employee = new Employee();
        Company company = new Company(employee);
        employee.setCompany(company);

        String jsonCompany = mapper.writeValueAsString(company);
        System.out.println(jsonCompany);
        assertTrue(true);
    }
}

If you run this code, you'll get the:

org.codehaus.jackson.map.JsonMappingException: Infinite recursion (StackOverflowError)

Solution Using `@JsonView`

@JsonView enables you to use filters and choose what fields should be included while serializing the objects. A filter is just a class reference used as a identifier. So let's first create the filters:

public class Filter {

    public static interface EmployeeData {};

    public static interface CompanyData extends EmployeeData {};

} 

Remember, the filters are dummy classes, just used for specifying the fields with the @JsonView annotation, so you can create as many as you want and need. Let's see it in action, but first we need to annotate our Company class:

public class Company {

    @JsonView(Filter.CompanyData.class)
    private Employee employee;

    public Company(Employee employee) {
        this.employee = employee;
    }

    public Employee getEmployee() {
        return employee;
    }
}

and change the Test in order for the serializer to use the View:

@SpringBootTest
@RunWith(SpringRunner.class)
@Transactional
public class CompanyTest {

    @Autowired
    public ObjectMapper mapper;

    @Test
    public void shouldSaveCompany() throws JsonProcessingException {
        Employee employee = new Employee();
        Company company = new Company(employee);
        employee.setCompany(company);

        ObjectWriter writter = mapper.writerWithView(Filter.CompanyData.class);
        String jsonCompany = writter.writeValueAsString(company);

        System.out.println(jsonCompany);
        assertTrue(true);
    }
}

Now if you run this code, the Infinite Recursion problem is solved, because you have explicitly said that you just want to serialize the attributes that were annotated with @JsonView(Filter.CompanyData.class).

When it reaches the back reference for company in the Employee, it checks that it's not annotated and ignore the serialization. You also have a powerful and flexible solution to choose which data you want to send through your REST APIs.

With Spring you can annotate your REST Controllers methods with the desired @JsonView filter and the serialization is applied transparently to the returning object.

Here are the imports used in case you need to check:

import static org.junit.Assert.assertTrue;

import javax.transaction.Transactional;

import org.junit.Test;
import org.junit.runner.RunWith;
import org.springframework.beans.factory.annotation.Autowired;
import org.springframework.boot.test.context.SpringBootTest;
import org.springframework.test.context.junit4.SpringRunner;

import com.fasterxml.jackson.core.JsonProcessingException;
import com.fasterxml.jackson.databind.ObjectMapper;
import com.fasterxml.jackson.databind.ObjectWriter;

import com.fasterxml.jackson.annotation.JsonView;

Answer 4

In case you are using Spring Data Rest, issue can be resolved by creating Repositories for every Entity involved in cyclical references.

Answer 5

I also met the same problem. I used @JsonIdentityInfo's ObjectIdGenerators.PropertyGenerator.class generator type.

That's my solution:

@Entity
@Table(name = "ta_trainee", uniqueConstraints = {@UniqueConstraint(columnNames = {"id"})})
@JsonIdentityInfo(generator = ObjectIdGenerators.PropertyGenerator.class, property = "id")
public class Trainee extends BusinessObject {
...

Answer 6

JsonIgnoreProperties [2017 Update]:

You can now use JsonIgnoreProperties to suppress serialization of properties (during serialization), or ignore processing of JSON properties read (during deserialization). If this is not what you're looking for, please keep reading below.

(Thanks to As Zammel AlaaEddine for pointing this out).

---

JsonManagedReference and JsonBackReference

Since Jackson 1.6 you can use two annotations to solve the infinite recursion problem without ignoring the getters/setters during serialization: @JsonManagedReference and @JsonBackReference.

Explanation

For Jackson to work well, one of the two sides of the relationship should not be serialized, in order to avoid the infite loop that causes your stackoverflow error.

So, Jackson takes the forward part of the reference (your Set<BodyStat> bodyStats in Trainee class), and converts it in a json-like storage format; this is the so-called marshalling process. Then, Jackson looks for the back part of the reference (i.e. Trainee trainee in BodyStat class) and leaves it as it is, not serializing it. This part of the relationship will be re-constructed during the deserialization (unmarshalling) of the forward reference.

You can change your code like this (I skip the useless parts):

Business Object 1:

@Entity
@Table(name = "ta_trainee", uniqueConstraints = {@UniqueConstraint(columnNames = {"id"})})
public class Trainee extends BusinessObject {

    @OneToMany(mappedBy = "trainee", fetch = FetchType.EAGER, cascade = CascadeType.ALL)
    @Column(nullable = true)
    @JsonManagedReference
    private Set<BodyStat> bodyStats;

Business Object 2:

@Entity
@Table(name = "ta_bodystat", uniqueConstraints = {@UniqueConstraint(columnNames = {"id"})})
public class BodyStat extends BusinessObject {

    @ManyToOne(fetch = FetchType.EAGER, cascade = CascadeType.ALL)
    @JoinColumn(name="trainee_fk")
    @JsonBackReference
    private Trainee trainee;

Now it all should work properly.

If you want more informations, I wrote an article about Json and Jackson Stackoverflow issues on Keenformatics, my blog.

EDIT:

Another useful annotation you could check is @JsonIdentityInfo: using it, everytime Jackson serializes your object, it will add an ID (or another attribute of your choose) to it, so that it won't entirely "scan" it again everytime. This can be useful when you've got a chain loop between more interrelated objects (for example: Order -> OrderLine -> User -> Order and over again).

In this case you've got to be careful, since you could need to read your object's attributes more than once (for example in a products list with more products that share the same seller), and this annotation prevents you to do so. I suggest to always take a look at firebug logs to check the Json response and see what's going on in your code.

Sources:

• Keenformatics - How To Solve JSON infinite recursion Stackoverflow (my blog)
• Jackson References
• Personal experience