Spring Data JPA map the native query result to Non-Entity POJO - spring

I have a Spring Data repository method with a native query
#Query(value = "SELECT g.*, gm.* FROM group g LEFT JOIN group_members gm ON g.group_id = gm.group_id and gm.user_id = :userId WHERE g.group_id = :groupId", nativeQuery = true)
GroupDetails getGroupDetails(#Param("userId") Integer userId, #Param("groupId") Integer groupId);
and I'd like to map the result to Non-Entity POJO GroupDetails.
Is it possible and if so, could you please provide an example ?

I think the easiest way to do that is to use so called projection. It can map query results to interfaces. Using SqlResultSetMapping is inconvienient and makes your code ugly :).
An example right from spring data JPA source code:
public interface UserRepository extends JpaRepository<User, Integer> {
#Query(value = "SELECT firstname, lastname FROM SD_User WHERE id = ?1", nativeQuery = true)
NameOnly findByNativeQuery(Integer id);
public static interface NameOnly {
String getFirstname();
String getLastname();
}
}
You can also use this method to get a list of projections.
Check out this spring data JPA docs entry for more info about projections.
Note 1:
Remember to have your User entity defined as normal - the fields from projected interface must match fields in this entity. Otherwise field mapping might be broken (getFirstname() might return value of last name et cetera).
Note 2:
If you use SELECT table.column ... notation always define aliases matching names from entity. For example this code won't work properly (projection will return nulls for each getter):
#Query(value = "SELECT user.firstname, user.lastname FROM SD_User user WHERE id = ?1", nativeQuery = true)
NameOnly findByNativeQuery(Integer id);
But this works fine:
#Query(value = "SELECT user.firstname AS firstname, user.lastname AS lastname FROM SD_User user WHERE id = ?1", nativeQuery = true)
NameOnly findByNativeQuery(Integer id);
In case of more complex queries I'd rather use JdbcTemplate with custom repository instead.

Assuming GroupDetails as in orid's answer have you tried JPA 2.1 #ConstructorResult?
#SqlResultSetMapping(
name="groupDetailsMapping",
classes={
#ConstructorResult(
targetClass=GroupDetails.class,
columns={
#ColumnResult(name="GROUP_ID"),
#ColumnResult(name="USER_ID")
}
)
}
)
#NamedNativeQuery(name="getGroupDetails", query="SELECT g.*, gm.* FROM group g LEFT JOIN group_members gm ON g.group_id = gm.group_id and gm.user_id = :userId WHERE g.group_id = :groupId", resultSetMapping="groupDetailsMapping")
and use following in repository interface:
GroupDetails getGroupDetails(#Param("userId") Integer userId, #Param("groupId") Integer groupId);
According to Spring Data JPA documentation, spring will first try to find named query matching your method name - so by using #NamedNativeQuery, #SqlResultSetMapping and #ConstructorResult you should be able to achieve that behaviour

I think Michal's approach is better. But, there is one more way to get the result out of the native query.
#Query(value = "SELECT g.*, gm.* FROM group g LEFT JOIN group_members gm ON g.group_id = gm.group_id and gm.user_id = :userId WHERE g.group_id = :groupId", nativeQuery = true)
String[][] getGroupDetails(#Param("userId") Integer userId, #Param("groupId") Integer groupId);
Now, you can convert this 2D string array into your desired entity.

You can write your native or non-native query the way you want, and you can wrap JPQL query results with instances of custom result classes.
Create a DTO with the same names of columns returned in query and create an all argument constructor with same sequence and names as returned by the query.
Then use following way to query the database.
#Query("SELECT NEW example.CountryAndCapital(c.name, c.capital.name) FROM Country AS c")
Create DTO:
package example;
public class CountryAndCapital {
public String countryName;
public String capitalName;
public CountryAndCapital(String countryName, String capitalName) {
this.countryName = countryName;
this.capitalName = capitalName;
}
}

This is my solution for converting to Map and then to custom Object
private ObjectMapper objectMapper;
public static List<Map<String, Object>> convertTuplesToMap(List<?> tuples) {
List<Map<String, Object>> result = new ArrayList<>();
tuples.forEach(object->{
if(object instanceof Tuple single) {
Map<String, Object> tempMap = new HashMap<>();
for (TupleElement<?> key : single.getElements()) {
tempMap.put(key.getAlias(), single.get(key));
}
result.add(tempMap);
}else{
throw new RuntimeException("Query should return instance of Tuple");
}
});
return result;
}
public <T> List<T> parseResult(List<?> list, Class<T> clz){
List<T> result = new ArrayList<>();
convertTuplesToMap(list).forEach(map->{
result.add(objectMapper.convertValue(map, clz));
});
return result;
}
public static class CustomDTO{
private String param1;
private Integer param2;
private OffsetDateTime param3;
}
public List<CustomDTO> doSomeQuery(){
Query query = entityManager.createNativeQuery("SELECT param1, param2 param3 ... ", Tuple.class);
return parseResult(query.getResultList(), CustomDTO.class);
}

Use the default method in the interface and get the EntityManager to get the opportunity to set the ResultTransformer, then you can return the pure POJO, like this:
final String sql = "SELECT g.*, gm.* FROM group g LEFT JOIN group_members gm ON g.group_id = gm.group_id and gm.user_id = ? WHERE g.group_id = ?";
default GroupDetails getGroupDetails(Integer userId, Integer groupId) {
return BaseRepository.getInstance().uniqueResult(sql, GroupDetails.class, userId, groupId);
}
And the BaseRepository.java is like this:
#PersistenceContext
public EntityManager em;
public <T> T uniqueResult(String sql, Class<T> dto, Object... params) {
Session session = em.unwrap(Session.class);
NativeQuery q = session.createSQLQuery(sql);
if(params!=null){
for(int i=0,len=params.length;i<len;i++){
Object param=params[i];
q.setParameter(i+1, param);
}
}
q.setResultTransformer(Transformers.aliasToBean(dto));
return (T) q.uniqueResult();
}
This solution does not impact any other methods in repository interface file.

USE JPA PROJECTIONS
In your case it may be desirable to retrieve data as objects of customized types. These types reflect partial views of the root class, containing only properties we care about. This is where projections come in handy.
first declare Entity as #immutable
#Entity
#Immutable
public class Address {
#Id
private Long id;
set your Repository
public interface AddressView {
String getZipCode();
}
Then use it in a repository interface:
public interface AddressRepository extends Repository<Address, Long> {
#Query("EXEC SP_GETCODE ?1")
List<AddressView> getAddressByState(String state);
}

If you are looking for running a custom SQL query in spring boot with #repository and #service structures. Please have a look.
https://stackoverflow.com/a/71501509/4735043

You can do something like
#NamedQuery(name="IssueDescriptor.findByIssueDescriptorId" ,
query=" select new com.test.live.dto.IssuesDto (idc.id, dep.department, iss.issueName,
cat.issueCategory, idc.issueDescriptor, idc.description)
from Department dep
inner join dep.issues iss
inner join iss.category cat
inner join cat.issueDescriptor idc
where idc.id in(?1)")
And there must be Constructor like
public IssuesDto(long id, String department, String issueName, String issueCategory, String issueDescriptor,
String description) {
super();
this.id = id;
this.department = department;
this.issueName = issueName;
this.issueCategory = issueCategory;
this.issueDescriptor = issueDescriptor;
this.description = description;
}

Related

Pagination of an complex object into DTO. JPA #Query

I have the following query to take some data regarding two entities in the same time and I receive an error.
#Query(value = "select new base.models.HRTableEntity( yr.user.gid, yr.user.id, yr.user.lastName || ' ' || yr.user.firstName, yr.user.position, yr.user.created,yr.genericField1,yr.genericField2) from YearlyReview yr where yr.year = :yr and yr.user.realDepartment = :dep and yr.user.city = :ct",
countQuery = "select count(yr.id) from YearlyReview yr where yr.year = :yr and yr.user.realDepartment = :dep and yr.user.city = :ct",
nativeQuery = false)
Page<HRTableEntity> getAllTableEntity(Pageable pageRequest, #Param("yr") int year, #Param("dep") String department, #Param("ct") String location);
I call this cunction with default Sort (gid: ASC) and receive the following error
org.hibernate.QueryException: could not resolve property: gid of: base.entities.YearlyReview
Repo interface:
public interface PageableYearlyReview extends CrudRepository<YearlyReview, UUID>
Yearly review have a member (user) of type ApplicationUser and I want to put information into DTO from yr.user.gid into HRTableEntry.gid.
What is the right way to do that ?
EDIT:
function call:
crunRepoYearTable.getAllTableEntity(PageRequest.of(pageNo - 1, pageSize, sort), year, realDepartment, user.getCity())
sort building
sortDir.equalsIgnoreCase(Sort.Direction.ASC.name()) ? Sort.by(sortedField).ascending() : Sort.by(sortedField).descending();
Entity structure:
public class YearlyReview {
#Id
private UUID id;
private int year;
#OneToOne
private ApplicationUser user;
....
}
public class ApplicationUser {
#Id
private String id;
private String gid;
.....
}
Problem solved:
for sort by gid need to send from the fronted like this user.gid. User is required to refer at AppUser and gid to access information. And change interface like this
public interface PageableYearlyReview extends CrudRepository<ApplicationUser, String>

Spring specification search inside JSON array of String

I am using Postgres version 12 database and ORM Hibernate mapping, and Assuming that i have next Entity:
#Entity
public class MyEntity {
#Type(type = "json")
#Column(name = "names", columnDefinition = "json")
private List<String> names;
// Getters and Setters
}
So the data will be persisted like that inside the table column (names):
["Sara", "Anton", "Lars"]
and i want to add a specification to search on that json from postgres database, i tried next one but it didn't work :(
private static Specification<MyEntity> withNames(String name) {
return (root, query, builder) -> {
Expression<String> function = builder.function("json_array_elements", String.class, root.get("names"));
return StringUtils.isBlank(name) ? builder.conjunction() : builder.like(function, name + "%");
};
}
Any suggestions of how to get it work?

Spring Specification Criteria Multiple Joins ? How?

I got stuck using a Spring Project with Spring Data + specification + criteria api.
I will try to simulate the situation with general entities we used write to get easy example.
The Entities:
Consider all attributes of the each entity is passed on the constructor showed below
Country(Long id, String name, String iso)
State(Long id, String name, String iso)
City(Long id, String name, String iso)
This is my repository:
public interface CityRepository extends PagingAndSortingRepository<City, Integer>, JpaSpecificationExecutor<City> {
}
As you can see, I don't need to implement anything on the repository
This is my service
#Service
#Transactional
public class CityService {
#Autowired
private CityRepository cityRepository;
#Transactional(readOnly = true)
public CityListVO findByNameLike(String name, PageRequest pageRequest) {
name = "%" + name + "%";
if (pageRequest == null) {
List<City> result = cityRepository.findAll(fillGridCriteria(name));
return new CityListVO(1, result.size(), result);
} else {
Page<City> result = cityRepository. findAll(fillGridCriteria(name), pageRequest);
return new CityListVO(result.getTotalPages(), result.getTotalElements(), result.getContent());
}
}
private static Specification<City> fillGridCriteria(String name) {
return new Specification<City>() {
#Override
public Predicate toPredicate(
Root<City> root,
CriteriaQuery<?> query,
CriteriaBuilder builder) {
/*
The current return I can do a like by name, and it works fine.
My problem is if for any reason I need to do multiple joins like the folow jpql:
select ci FROM City ci, State st, Country co where ci.st = st AND st.co = co AND co.name = 'Canada';
How to do this from here ? Inside this method.
How is gonna be the return for this method ?
*/
return builder.like(root.get("name"), name.trim());
}
};
}
}
Let's assume you want all the cities that their country's name like name and you have a relational Model in which :
Country(Long id, String name, String iso)
State(Long id,Long country, String name, String iso)
City(Long id, Long state, String name, String iso)
Predicate:
private static Specification<City> fillGridCriteria(String name) {
return new Specification<City>() {
#Override
public Predicate toPredicate(
Root<City> root,
CriteriaQuery<?> query,
CriteriaBuilder builder) {
return
builder.like(root.get("state").get("country").get("name"), name.trim());
}
};
}

spring data neo4j crud - many optional param?

I use Spring-data-neo4j with one CrudRepository
#Repository
public interface PersonRepository extends GraphRepository<Person> {}
I have a Html form with 3 inputs FirstName, Name, Age, so I have possible a multiple criteria choose : All, FirstName, FirstName + Name, FirstName + Age etc....
I would like to make a "multiple criteria find" with Map or other stuff. Is it possible?
I try this in my CRUD:
List<Person> findByFirstnameAndNameAndAge(String firstname, String name, int age);
but it's not work if one or all parameters is null.
Try to use a map and a #Query annotation
#Query("MATCH (u:Person) WHERE u.name = {param}.name OR u.age = {param}.age RETURN u")
List<Person> findDynamic(#Param("param") Map params);
Hi #Michael Hunger Thank you for your response. It's not exactly what I expected but you delivered me some fine search stuff
finally I do this :
import org.apache.commons.collections.map.HashedMap;
import com.google.common.collect.Lists;
(...)
#Autowired
private EventRepository eventRepository; //#Repository extends GraphRepository<Event>
(...)
public List<Event> findByDynamicParam(HashedMap params) {
String query = "match (event)-[:user]-(user), (event)-[:action]-(action)";
if (!params.isEmpty()) {
query += " where";
}
if (params.containsKey("actionId")) {
query += " id(action) = {actionId} and";
}
if (params.containsKey("userId")) {
query += " id(user) = {userId} and";
}
if (!params.isEmpty()) {
query = query.substring(0, query.length() - 4);
}
query += " return (event)";
return Lists.newArrayList(eventRepository.query(query, params));
}
client's caller :
HashedMap params = new HashedMap();
if (actionId != null) {
params.put("actionId", actionId);
}
if (userId != null) {
params.put("actionId", userId);
}
List<Event> events = eventService.findByDynamicParam(params);
What do you think? Is it possible to optimize this function?
Regards
Olivier from Paris

Is it possible to avoid typecast on custom #Query?

Imagine that we have an entity:
#Entity
public class Person implements Serializable {
#Id
private String name;
private Long age;
private Boolean isMad;
...
}
And a repository with a trivial (and unnecessary) example for a custom query:
#Repository
public interface PersonRepository extends PagingAndSortingRepository<Info, String> {
#Query("select p.isMad, count(*) from Person p group by p.isMad")
List<Object> aggregateByMadness();
}
Now to parse this List we need to do something like this:
for (Object element : list) {
Object[] result = (Object[]) element;
Boolean isMad = (Boolean) result[0];
Long count = (Long) result[1];
}
which is a pain, can we cast the result of the query directly to List of a POJO?
Yes, you could use the JPQL construction expression:
package com.foo;
public class Madness {
public Madness(boolean isMad, Number count) { /* ...*/ }
}
And in your repository:
#Query("select new com.foo.Madness(p.isMad, count(*)) from Person p group by p.isMad")
List<Madness> aggregateByMadness();

Resources