Passing parent id reference when creating child object through REST api - spring-boot

I am using spring boot (version - 2.1.1). I have a one to many database model exposed for CRUD operations through rest api's. The model looks as below. How do I configure the POST /departments api (that creates a department object) to accept just the organization id in the input json body?
#PostMapping
public Long createDepartment(#RequestBody Department Department) {
Department d = departmentService.save(Department);
return d.getId();
}
Note - I do not want to allow creating organization object when creating a department.
Model object mapping
#Entity
#Table(name="ORGANIZATIONS")
public class Organization{
#Id
#GeneratedValue
Private long id;
#Column(unique=true)
Private String name;
#OneToMany(mappedBy = "organization", fetch = FetchType.EAGER)
private List<Department> departments;
}
#Entity
#Table(name="DEPARTMENTS")
Public class Department{
#Id
#GeneratedValue
Private long id;
#Column(unique=true)
Private String name;
#ManyToOne(fetch = FetchType.EAGER)
private Organization organization;
}
Thanks!

The easiest and most sane way in my opinion is to utilize the DTO (Data Transfer Object) pattern.
Create a class that represent the model you want to get as your input:
public class CreateDepartmentRequest {
private long id;
// getters and setters
}
Then use it in your controller:
#PostMapping
public Long createDepartment(#RequestBody CreateDepartmentRequest request) {
Department d = new Department();
d.setId(request.getId());
Department d = departmentService.save(d);
return d.getId();
}
Side note, its better to ALWAYS return JSON through REST API (unless you use some other format across your APIs) so you can also utilize the same pattern as I mentioned above to return a proper model as a result of the POST operation or a simple Map if you don't want to create to many models.

Related

Spring Data JPA save child object with the ID of parent object

I have two objects, one parent and one child as follows :
#Entity
#Table(name="category")
public class CategoryModel {
private #Id #GeneratedValue Long id;
private String name;
#OneToMany(mappedBy="category", cascade=CascadeType.PERSIST)
private List<AttributeModel> attributes;
}
#Entity
#Table(name="attribute")
public class AttributeModel {
private #Id #GeneratedValue Long id;
private String name;
#ManyToOne(cascade = CascadeType.ALL)
#JoinColumn(name="category_id")
private CategoryModel category;
}
I also have dtos which maps to these model objects but I ommited them.
When I try to save a category object with this payload Attribute values are also created in the attribute table but with null category ids.
{
"name":"Chemicals",
"attributes":[
{"name": "volume"}, {"name":"humidity"}
]
}
What can I do to have my attribute values persisted into the database with the category id which is created before them?
First of all, this problem is not a "Spring Data JPA" problem, it is a JPA (probably Hibernate) problem.
Analysis
Since you left out the code for the controller and the JSON mapping, I have to guess a bit:
fact 1: The relationship between category and attributes is controlled by the attribute AttributeModel.category but not by CategoryModel.attributes. (That is how JPA works).
observation 2: Your JSON object define CategoryModel.attributes (i.e. opposite to how JPA works).
Without knowing your JSON mapping configuration and controller code, I would guess that the problem is: that your JSON mapper does not set the AttributeModel.category field when it deserialises the JSON object.
Solution
So you need to instruct the JSON mapper to set the AttributeModel.category field during deserialisation. If you use Jackson, you could use:
#JsonManagedReference and
#JsonBackReference
#Entity
#Table(name="category")
public class CategoryModel {
...
#JsonManagedReference
#OneToMany(mappedBy="category", cascade=CascadeType.PERSIST)
private List<AttributeModel> attributes;
}
#Entity
#Table(name="attribute")
public class AttributeModel {
...
#JsonBackReference
#ManyToOne(cascade = CascadeType.ALL)
#JoinColumn(name="category_id")
private CategoryModel category;
}
I solved this by manually setting child object's reference to the parent object as follows :
public Long createCategory(CategoryDto categoryDto) {
CategoryModel categoryModel = categoryDto.toModel(true,true);
categoryModel.getAttributes().forEach(a -> a.setCategory(categoryModel));
return categoryRepository.save(categoryModel).getId();
}

Spring JPA bidirectional relation on multiple nested entities

I know there has been multiple questions on bidirectional relations using spring jpa in the past but my case is a little bit different because i am using 3 entities with 2 relationships to implement a medical system
I have 3 entities : doctor/patient/appointment
here is the code for the 3 entities
please note all setters , getters and constructors implemented but ommited here for clarity
Patient class
#Entity
public class resPatient {
#Id
#GeneratedValue( strategy= GenerationType.IDENTITY )
private long code;
private String name;
private String gender;
private String email;
private String mobile;
private int age;
private String notes;
#OneToMany(mappedBy = "patient")
List<resPackageMembership> memberships;
#OneToMany(mappedBy = "patient")
List<resAppointment> appointments;
#OneToMany(fetch = FetchType.LAZY,mappedBy = "patient")
List<resMedImage> medImages;
Doctor class
#Entity
public class resDoctor {
#Id
#GeneratedValue( strategy= GenerationType.IDENTITY )
private long code;
private String name;
private String mobile;
private String email;
private String gender;
private int age;
private String speciality;
#OneToMany(mappedBy = "doctor")
List<resAppointment> appointments;
Appointment class
#Entity
public class resAppointment {
#Id
#GeneratedValue( strategy= GenerationType.IDENTITY )
private long code;
private String speciality;
#Basic
#Temporal(TemporalType.TIMESTAMP)
private Date dateCreated;
#Basic
#Temporal(TemporalType.TIMESTAMP)
private Date dateToVisit;
private String status;
private String notes;
#ManyToOne(fetch = FetchType.LAZY)
#JoinColumn(name = "doctorCode")
private resDoctor doctor;
#ManyToOne(fetch = FetchType.LAZY)
#JoinColumn(name = "patientCode")
private resPatient patient;
the way my medical system should work is that when i get a patient using my restful controller i want all the patient data including his appointments but this leads to an infinite loop as the appointment has the doctor which also has appointments and so on.
i cannot user #JSONIGNORE as there are 2 relationships i want to get the patient with his appointments which should have the doctor without the appointments array and should not have any patient data as i already am in the patient object
As a general best-practice, it's recommended to separate the entities from the data transfer objects used for the rest controllers. With DTO's in place, you have more control on which data to include and serialize within them to avoid the circlular references.
If you like check out https://bootify.io, it generates the DTOs from your database schema, but the custom endpoint you still need to define/build.
I develop an annotation processor called beanknife recently, it support generate DTO from any class. You need config by annotation. But you don't need change the original class. This library support configuring on a separate class. Of course you can choose which property you want and which you not need. And you can add new property by the static method in the config class. For your question:
// this will generate a DTO class named "resPatientView".
// You can change this name using genName attribute.
#ViewOf(value=resPatient.class, includePattern = ".*")
public class PatientViewConfigure {
// here tell the processor to automatically convert the property appointments from List<resAppointment> to List<resAppointmentWithoutPatient>.
// resAppointmentWithoutPatient is the generated class configured at the following.
// Note, although at this moment it not exists and your idea think it is an error.
// this code really can be compiled, and after compiled, all will ok.
#OverrideViewProperty("appointments")
private List<resAppointmentWithoutPatient> appointments;
}
// here generated a class named resAppointmentWithoutPatient whick has all properties of resAppointment except patient
#ViewOf(value=resAppointment.class, genName="resAppointmentWithoutPatient", includePattern = ".*", excludes={"patient"})
public class AppointmentWithoutPatientViewConfigure {
// the doctor property will be converted to its dto version which defined by the configure class DoctorWithoutAppointmentsViewConfigure.
#OverrideViewProperty("doctor")
private resDoctorWithoutAppointments doctor;
}
// here we generate a class which has all properties of resDoctor except appointments
#ViewOf(value=resDoctor.class, genName="resDoctorWithoutAppointments", includePattern = ".*", excludes={"appointments"})
public class DoctorWithoutAppointmentsViewConfigure {}
// in you rest controller. return the dto instead of the entities.
resPatient patient = ...
resPatientView dto = resPatientView.read(patient);
List<resPatient> patients = ...
List<resPatientView> dto = resPatientView.read(patients);
At the end, the class resPatientView will has the same shap with resPatient except its appointments not having patient property and its doctor property is replaced with a version without appointments property.
Here are more examples.
The version 1.10 is ready. Will fix some bug and support the configure bean to be managed by spring.

Spring Boot - creating endpoints that involve more than one database table - what is the best practice

I am doing a spring boot tutorial for fun. I understand that each entity can be defined by a model with a repository that the controller for that entity uses to create endpoints. I am wondering what is the best practice to create an endpoint in a controller that involves more than one model. I.e. we have a session model and speaker model below and i want to have a endpoint that returns all sessions with a speaker of a certain last name (many to many). Do I have to use both repositories? Is it even good practice to put this in session's controller since it also involves speaker? See the example below. Sorry if the question is stupid, I am trying to learn Spring-Boot and I am completely new to it.
//Session class
#Entity(name="sessions")
#JsonIgnoreProperties({"hibernateLazyInitializer", "handler"})
public class Session {
#Id
#GeneratedValue(strategy = GenerationType.IDENTITY)
private Long session_id;
private String session_name;
private String session_description;
private Integer session_length;
#ManyToMany
#JoinTable(
name = "session_speakers",
joinColumns = #JoinColumn(name="session_id"),
inverseJoinColumns = #JoinColumn(name= "speaker_id"))
private List<Speaker> speakers;
}
//Speaker
#Entity(name="speakers")
#JsonIgnoreProperties({"hibernateLazyInitializer", "handler"})
public class Speaker {
#Id
#GeneratedValue(strategy = GenerationType.IDENTITY)
private Long speaker_id;
private String first_name;
private String last_name;
private String title;
private String company;
private String speaker_bio;
#ManyToMany(mappedBy = "speakers")
#JsonIgnore
private List<Session> sessions;
#RestController
#RequestMapping("/api/v1/sessions")
public class SessionsController {
#Autowired
private SessionRepository sessionRepository;
#GetMapping
#RequestMapping("{id}")
public Session getSessionsBySpeakerName(#PathVariable String lname{
//want to implement
}
Controllers are supposed to be very light,so it would be better to have seperate controllers.
Usualy it is good to have seperate Repositories also so that later on if you want to add a new table you would not be touching the existing stable code, it sould be loosly coupled as much as possible
Here are few blog posts which I found useful.
https://www.e4developer.com/2018/08/06/spring-boot-best-practices/
https://dzone.com/articles/spring-boot-best-practices-for-microservices
https://www.javaguides.net/2019/03/spring-boot-best-practices.html

Fetch specific property in hibernate One-to-many relationship

I have two pojo classes with one-to-many relationship in hibernate
CustomerAccountEnduserOrderDetails.class
#Entity #Table(name="customer_account_enduser_order_details")
public class CustomerAccountEnduserOrderDetails implements Serializable{
private static final long serialVersionUID = 1L;
#Id
#GeneratedValue(strategy=GenerationType.AUTO)
#Column(name="id")
private Long id;
#ManyToOne(fetch=FetchType.EAGER)
#JoinColumn(name = "product_id", insertable = false, updatable = false)
private CustomerCmsProduct customerCmsProduct;
}
Second is CustomerCmsProduct.class
#Entity
#Table(name="customer_cms_product")
#JsonIgnoreProperties(ignoreUnknown = true)
public class CustomerCmsProduct {
#Id
#GeneratedValue(strategy = GenerationType.AUTO)
#Column(name="id")
private Long id;
#Column(name="name")
private String name;
#Column(name="offer_price")
private String offerPrice;
#Column(name="original_price")
private String originalPrice;
#Column(name="discount")
private String discount;
}
Here if I fetch the object of CustomerAccountEnduserOrderDetails class,then i will get the CustomerCmsProduct class also , my problem is that here i want the specific column of CustomerCmsProduct table (not all by default i am getting all) like only id and originalPrice.
How i can do like that projection here?
In the service layer or at a webservice layer( if this is a web project) Create two different classes other than #Entity as DTO(Data Transfer Objects) which helps is data transfer from one layer to the other.
public class CustomerAccountEnduserOrderDetailsPojo {
private List<CustomerCmsProductPojo> productPojoList = new ArrayList<> ();
// getter and setter
}
public class CustomerCmsProductPojo {}
Follow the below steps
Retrieve the #Entity class data by executing the query from service layer.
Iterate over all the fields and copy only required fields to pojo layer
Expose the data to other layers using service method.
This way, we can avoid changing the custom hibernate behavior as it is linked with many parameters like cache, one to many queries that are fired per iteration.
And also, do any customization that you want in this layer. Hope this is multi layered project where you have different layers which servers different purpose.

JPQL Special Query

I have two entity bean :
#Entity
public class User {
#Id
#GeneratedValue
private Long id;
#OneToMany(mappedBy="user", cascade = CascadeType.ALL)
private List<Comment> comments = new ArrayList<Comment>();
//SOME OTHER CLASS VARIABLES
//GETTERS AND SETTERS
}
and my Comment class is like this :
#Entity
public class Comment {
#Id
#GeneratedValue
private Long id;
private String title;
private String content;
#ManyToOne
private User user
//SOME OTHER CLASS VARIABLES
//GETTERS AND SETTERS
}
now I know that I can get the User Object from session and set the user for my comment like this in order to be able to use the join feature in JPA:
commentObject.setUser(TheSessionGrabedUserObject/UserObjectWhichHasFetchedFromDbUsingUserId);
but as long as I have the userId for my user Object I do not need to do this.
I'm looking for a way to insert this foreignKey into my comment table without getting the User Object from session or maybe query to database to fetch it first !
how I'm gonna do it using JPQL ?
You can use the entityManager.getReference() method. In your case:
entityManager.getReference(User.class, userId);
This will not perform any DB query, but will give you a User instance with only the ID populated, and you can pass that to commentObject.setUser().

Resources