I have a JPA query written like this:
public interface MyDataRepository extends CrudRepository<MyData, String> {
#Query("select md " +
"from MyData md " +
"join fetch md.child c " +
"where c.date = :date")
List<MyData> getMyDataOfDate(#NotNull LocalDate date);
#Query("select md " +
"from MyData md " +
"join fetch md.child c " +
"where c.name = :name")
List<MyData> getMyDataOfName(#NotNull String name);
#Query("select md " +
"from MyData md " +
"join fetch md.child c " +
"where md.type = :type")
List<MyData> getMyDataOfType(#NotNull String type);
}
Class MyData and Child are defined as:
class MyData {
String id;
String type;
#ManyToOne
#JoinColumn(name = "CHILD_ID", referencedColumnName = "ID", nullable = false, updatable = false)
Child child;
}
class Child {
String id;
String name;
LocalDate date;
}
The problem is that whenever I call the getMyDataOfDate method or getMyDataOfName method, they always return ALL rows rather than the rows that matches the where condition, as if the where clause never exists.
However, the getMyDataOfType method works fine. The difference of this method is that the where condition is on a property of md, not c.
What did I do wrong?
JPA does not allow filtering on join fetches. Reasons being is that when you specify a join fetch, you are telling JPA to fetch the parent and all its children defined by that relationship in the managed entities it returns. If filtering were allowed, the list of children, the relationship in the parent, might not reflect what is actually in the database. Take the case of Parent with many children
"Select parents from Parent p fetch join p.children c where c.firstName = 'Bob'"
For such a query, when you get a list of parents and calling getChildren on them, do you expect to see all their children or a list that only contains children named Bob? If the later (which is the only way to do so), how should JPA handle changes to a parents children list, and know what to do with the not-fetched children?
This is why JPA doesn't allow filtering over fetch joins, and they restrict it across all relationships to be consistent. If you want the parents who have children with the firstName of 'Bob', it would look like:
"Select parents from Parent p join p.children c where c.firstName = 'Bob'"
Every parent returned will be a complete representation of its state in the database based on its mappings; so accessing parent.getChildren will return the current state of its children list and not something affected by the way it was fetched.
How can I read list of #Embeddable objects from MongoDB with Hibernate OGM after aggregation.
I have entity like this
#javax.persistence.Entity
public class MySession implements Serializable {
#Id
#GeneratedValue(strategy = GenerationType.IDENTITY)
#Type(type = "objectid")
private String id;
private Date start;
private Date end;
#ElementCollection
private List<MySessionEvent> events;
}
and #Embeddable object
#javax.persistence.Embeddable
public class MySessionEvent implements Serializable {
private Long time;
private String name;
}
I stuck with mapping Embeddable objects from native query
String queryString = "db.MySession.aggregate([" +
" { '$match': { 'events.name': { '$regex': 'Abrakadabra'} }}, " +
" { '$unwind': '$events' }, " +
" { '$replaceRoot': { 'newRoot': '$events'}} " +
"])";
List<MySessionEvent> objects = em.createNativeQuery(queryString, MySessionEvent.class).getResultList();
I get an error Caused by: org.hibernate.MappingException: Unknown entity
Copying your comment here because it adds some details:
I have data like this [ {id:'s1', events: [{name: 'one'},{name:
'two'}]}, {id:'s2', events: [{name: 'three'},{name: 'four'}]} ] and I
want result like this [{name: 'one'},{name: 'two'},{name:
'three'},{name: 'four'}]
The query you are running returns the following type of results if I run it on MongoDB natively (I populated it with some random data):
{ "name" : "Event 3", "time" : NumberLong(3) }
{ "name" : "Abrakadabra", "time" : NumberLong(5) }
This is not enough to rebuild an entity and that's the reason you are seeing the exception.
Considering that you only want the list of events, this should work:
List<Object[]> poems = em.createNativeQuery( queryString ).getResultList();
Hibernate OGM will convert the previous result in a list of arrays.
Each element of the List is an array where the firs value of the array is the name of the event and the second one is the time.
For supported cases like this I think HQL queries are better. You could rewrite the same example with the following:
String queryString =
"SELECT e.name " +
"FROM MySession s JOIN s.events e " +
"WHERE e.name LIKE 'Abrakadabra'";
List<Object[]> events = em.createQuery( queryString ).getResultList();
Note that I decided not to return the time because in your comment you didn't request it, but this will work as well:
String queryString =
"SELECT e.time, e.name " +
"FROM MySession s JOIN s.events e " +
"WHERE e.name LIKE 'Abrakadabra'";
List<Object[]> events = em.createQuery( queryString ).getResultList();
It is not recognizing the entity, make sure all your entities are in persistence.xml Embeddable objects too
<class>org.example.package.MySessionEvent</class>
Following up on a question I posted yesterday: How to populate POJO class from custom Hibernate query?
Can someone show me an example of how to code the following SQL in Hibernate, and get the results correctly?
SQL:
select firstName, lastName
from Employee
What I'd like to do, if it's possible in Hibernate, is to put the results in their own base class:
class Results {
private firstName;
private lastName;
// getters and setters
}
I believe it's possible in JPA (using EntityManager), but I haven't figured out how to do it in Hibernate (using SessionFactory and Session).
I'm trying to learn Hibernate better, and even this "simple" query is proving confusing to know what form Hibernate returns the results, and how to map the results into my own (base) class. So at the end of the DAO routine, I'd do:
List<Results> list = query.list();
returning a List of Results (my base class).
select firstName, lastName from Employee
query.setResultTransformer(Transformers.aliasToBean(MyResults.class));
You can't use above code with Hibernate 5 and Hibernate 4 (at least Hibernate 4.3.6.Final), because of an exception
java.lang.ClassCastException: com.github.fluent.hibernate.request.persistent.UserDto cannot be cast to java.util.Map
at org.hibernate.property.access.internal.PropertyAccessMapImpl$SetterImpl.set(PropertyAccessMapImpl.java:102)
The problem is that Hibernate converts aliases for column names to upper case — firstName becomes FIRSTNAME. And it try to find a getter with name getFIRSTNAME(), and setter setFIRSTNAME() in the DTO using such strategies
PropertyAccessStrategyChainedImpl propertyAccessStrategy = new PropertyAccessStrategyChainedImpl(
PropertyAccessStrategyBasicImpl.INSTANCE,
PropertyAccessStrategyFieldImpl.INSTANCE,
PropertyAccessStrategyMapImpl.INSTANCE
);
Only PropertyAccessStrategyMapImpl.INSTANCE suits, in opinion of Hibernate, well. So after that it tries to do conversion (Map)MyResults.
public void set(Object target, Object value, SessionFactoryImplementor factory) {
( (Map) target ).put( propertyName, value );
}
Don't know, it is a bug or feature.
How to solve
Using aliases with quotes
public class Results {
private String firstName;
private String lastName;
public String getFirstName() {
return firstName;
}
public String getLastName() {
return lastName;
}
public void setFirstName(String firstName) {
this.firstName = firstName;
}
public void setLastName(String lastName) {
this.lastName = lastName;
}
}
String sql = "select firstName as \"firstName\",
lastName as \"lastName\" from Employee";
List<Results> employees = session.createSQLQuery(sql).setResultTransformer(
Transformers.aliasToBean(Results.class)).list();
Using a custom result transformer
Another way to solve the problem — using a result transformer that ignores method names case (treat getFirstName() as getFIRSTNAME()). You can write your own or use FluentHibernateResultTransformer. You will not need to use quotes and aliases (if you have column names equal to DTO names).
Just download the library from the project page (it doesn't need additional jars): fluent-hibernate.
String sql = "select firstName, lastName from Employee";
List<Results> employees = session.createSQLQuery(sql)
.setResultTransformer(new FluentHibernateResultTransformer(Results.class))
.list();
This transformer can be used for nested projections too: How to transform a flat result set using Hibernate
See AliasToBeanResultTransformer:
Result transformer that allows to transform a result to a user specified class which will be populated via setter methods or fields matching the alias names.
List resultWithAliasedBean = s.createCriteria(Enrolment.class)
.createAlias("student", "st")
.createAlias("course", "co")
.setProjection( Projections.projectionList()
.add( Projections.property("co.description"), "courseDescription" )
)
.setResultTransformer( new AliasToBeanResultTransformer(StudentDTO.class) )
.list();
StudentDTO dto = (StudentDTO)resultWithAliasedBean.get(0);
Your modified code:
List resultWithAliasedBean = s.createCriteria(Employee.class, "e")
.setProjection(Projections.projectionList()
.add(Projections.property("e.firstName"), "firstName")
.add(Projections.property("e.lastName"), "lastName")
)
.setResultTransformer(new AliasToBeanResultTransformer(Results.class))
.list();
Results dto = (Results) resultWithAliasedBean.get(0);
For native SQL queries see Hibernate documentation:
13.1.5. Returning non-managed entities
It is possible to apply a ResultTransformer to native SQL queries, allowing it to return non-managed entities.
sess.createSQLQuery("SELECT NAME, BIRTHDATE FROM CATS")
.setResultTransformer(Transformers.aliasToBean(CatDTO.class))
This query specified:
the SQL query string
a result transformer
The above query will return a list of CatDTO which has been instantiated and injected the values of NAME and BIRTHNAME into its corresponding properties or fields.
You need to use a constructor and in the hql use new. I let you the code example taken from this question: hibernate HQL createQuery() list() type cast to model directly
class Result {
private firstName;
private lastName;
public Result (String firstName, String lastName){
this.firstName = firstName;
this.lastName = lastName;
}
}
then your hql
select new com.yourpackage.Result(employee.firstName,employee.lastName)
from Employee
and your java (using Hibernate)
List<Result> results = session.createQuery("select new com.yourpackage.Result(employee.firstName,employee.lastName) from Employee").list();
YMMV but I've found that the key factor is you must make sure to alias every field in your SELECT clause with the SQL "AS" keyword. I've never had to use quotes around the alias names. Also, in your SELECT clause use the case and punctuation of the actual columns in your database and in the aliases use the case of the fields in your POJO. This has worked for me in Hibernate 4 and 5.
#Autowired
private SessionFactory sessionFactory;
...
String sqlQuery = "SELECT firstName AS firstName," +
"lastName AS lastName from Employee";
List<Results> employeeList = sessionFactory
.getCurrentSession()
.createSQLQuery(sqlQuery)
.setResultTransformer(Transformers.aliasToBean(Results.class))
.list();
If you have multiple tables you can use table aliases in the SQL as well. This contrived example with an additional table named "Department" uses more traditional lower case and underscores in database field names with camel case in the POJO field names.
String sqlQuery = "SELECT e.first_name AS firstName, " +
"e.last_name AS lastName, d.name as departmentName" +
"from Employee e, Department d" +
"WHERE e.department_id - d.id";
List<Results> employeeList = sessionFactory
.getCurrentSession()
.createSQLQuery(sqlQuery)
.setResultTransformer(Transformers.aliasToBean(Results.class))
.list();
java.lang.ClassCastException: "CustomClass" cannot be cast to java.util.Map.
This issue appears when the columns specified in SQL Query doesn't match with the columns of the mapping class.
It may be due to:
Non-matching casing of column name or
The column names are not matching or
column exist in query but missing in class.
JPQL case from hibernate 5.4:
Query<Employee> queryList = session.createQuery("select new xxx.xxx.Employee(e.firstName,e.lastName) from Employee e", Employee.class);
List<Employee> list = queryList.list();
Query<Long> queryCount = session.createQuery("select count(*) from Employee", Long.class);
Long count = queryCount.getSingleResult();
The select statement in JPQL is exactly the same as for HQL except that JPQL requires a select_clause, whereas HQL does not.
setResultTransformer #Deprecated It should not be used
more on Hibernate_User_Guide.html#hql-select
In case you have a native query, all answers here use deprecated methods for newer versions of Hibernate, so if you are using 5.1+ this is the way to go:
// Note this is a org.hibernate.query.NativeQuery NOT Query.
NativeQuery query = getCurrentSession()
.createNativeQuery(
"SELECT {y.*} , {x.*} from TableY y left join TableX x on x.id = y.id");
// This maps the results to entities.
query.addEntity("x", TableXEntity.class);
query.addEntity("y", TableYEntity.class);
query.list()
Below is a result transformer that ignores case:
package org.apec.abtc.dao.hibernate.transform;
import java.lang.reflect.Field;
import java.util.Arrays;
import java.util.List;
import org.hibernate.HibernateException;
import org.hibernate.property.access.internal.PropertyAccessStrategyBasicImpl;
import org.hibernate.property.access.internal.PropertyAccessStrategyChainedImpl;
import org.hibernate.property.access.internal.PropertyAccessStrategyFieldImpl;
import org.hibernate.property.access.internal.PropertyAccessStrategyMapImpl;
import org.hibernate.property.access.spi.Setter;
import org.hibernate.transform.AliasedTupleSubsetResultTransformer;
/**
* IgnoreCaseAlias to BeanResult Transformer
*
* #author Stephen Gray
*/
public class IgnoreCaseAliasToBeanResultTransformer extends AliasedTupleSubsetResultTransformer
{
/** The serialVersionUID field. */
private static final long serialVersionUID = -3779317531110592988L;
/** The resultClass field. */
#SuppressWarnings("rawtypes")
private final Class resultClass;
/** The setters field. */
private Setter[] setters;
/** The fields field. */
private Field[] fields;
private String[] aliases;
/**
* #param resultClass
*/
#SuppressWarnings("rawtypes")
public IgnoreCaseAliasToBeanResultTransformer(final Class resultClass)
{
if (resultClass == null)
{
throw new IllegalArgumentException("resultClass cannot be null");
}
this.resultClass = resultClass;
this.fields = this.resultClass.getDeclaredFields();
}
#Override
public boolean isTransformedValueATupleElement(String[] aliases, int tupleLength) {
return false;
}
/**
* {#inheritDoc}
*/
#Override
public Object transformTuple(final Object[] tuple, final String[] aliases)
{
Object result;
try
{
if (this.setters == null)
{
this.aliases = aliases;
setSetters(aliases);
}
result = this.resultClass.newInstance();
for (int i = 0; i < aliases.length; i++)
{
if (this.setters[i] != null)
{
this.setters[i].set(result, tuple[i], null);
}
}
}
catch (final InstantiationException | IllegalAccessException e)
{
throw new HibernateException("Could not instantiate resultclass: " + this.resultClass.getName(), e);
}
return result;
}
private void setSetters(final String[] aliases)
{
PropertyAccessStrategyChainedImpl propertyAccessStrategy = new PropertyAccessStrategyChainedImpl(
PropertyAccessStrategyBasicImpl.INSTANCE,
PropertyAccessStrategyFieldImpl.INSTANCE,
PropertyAccessStrategyMapImpl.INSTANCE
);
this.setters = new Setter[aliases.length];
for (int i = 0; i < aliases.length; i++)
{
String alias = aliases[i];
if (alias != null)
{
for (final Field field : this.fields)
{
final String fieldName = field.getName();
if (fieldName.equalsIgnoreCase(alias))
{
alias = fieldName;
break;
}
}
setters[i] = propertyAccessStrategy.buildPropertyAccess( resultClass, alias ).getSetter();
}
}
}
/**
* {#inheritDoc}
*/
#Override
#SuppressWarnings("rawtypes")
public List transformList(final List collection)
{
return collection;
}
#Override
public boolean equals(Object o) {
if ( this == o ) {
return true;
}
if ( o == null || getClass() != o.getClass() ) {
return false;
}
IgnoreCaseAliasToBeanResultTransformer that = ( IgnoreCaseAliasToBeanResultTransformer ) o;
if ( ! resultClass.equals( that.resultClass ) ) {
return false;
}
if ( ! Arrays.equals( aliases, that.aliases ) ) {
return false;
}
return true;
}
#Override
public int hashCode() {
int result = resultClass.hashCode();
result = 31 * result + ( aliases != null ? Arrays.hashCode( aliases ) : 0 );
return result;
}
}
Writing (exist this type of Challenges working with hibernate)
Custom Queries
Custom Queries with Optional Parameters
Mapping Hibernate Custom query results to Custom class.
I am not saying about custom EntityRepository interface which extends JpaRepository on SpringBoot which you can write custom Query with #Query -> here you can't write query with optional params
e.g. if param is null don't append it in query string. And you can use Criteria api of hibernate but it not recommended in their documentation because of performance issue...
But exist simple and error prone and performance good way...
Write your own QueryService class which are methods will get
string(answer for first and second problem) sql and will map result to
Custom class (third problem) with it's any association #OneToMany, #ManyToOne ....
#Service
#Transactional
public class HibernateQueryService {
private final Logger log = LoggerFactory.getLogger(HibernateQueryService.class);
private JpaContext jpaContext;
public HibernateQueryService(JpaContext jpaContext) {
this.jpaContext = jpaContext;
}
public List executeJPANativeQuery(String sql, Class entity){
log.debug("JPANativeQuery executing: "+sql);
EntityManager entityManager = jpaContext.getEntityManagerByManagedType(Article.class);
return entityManager.createNativeQuery(sql, entity).getResultList();
}
/**
* as annotation #Query -> we can construct here hibernate dialect
* supported query and fetch any type of data
* with any association #OneToMany and #ManyToOne.....
*/
public List executeHibernateQuery(String sql, Class entity){
log.debug("HibernateNativeQuery executing: "+sql);
Session session = jpaContext.getEntityManagerByManagedType(Article.class).unwrap(Session.class);
return session.createQuery(sql, entity).getResultList();
}
public <T> List<T> executeGenericHibernateQuery(String sql, Class<T> entity){
log.debug("HibernateNativeQuery executing: "+sql);
Session session = jpaContext.getEntityManagerByManagedType(Article.class).unwrap(Session.class);
return session.createQuery(sql, entity).getResultList();
}
}
Use case - you can write any type condition about query params
#Transactional(readOnly = true)
public List<ArticleDTO> findWithHibernateWay(SearchFiltersVM filter){
Long[] stores = filter.getStores();
Long[] categories = filter.getCategories();
Long[] brands = filter.getBrands();
Long[] articles = filter.getArticles();
Long[] colors = filter.getColors();
String query = "select article from Article article " +
"left join fetch article.attributeOptions " +
"left join fetch article.brand " +
"left join fetch article.stocks stock " +
"left join fetch stock.color " +
"left join fetch stock.images ";
boolean isFirst = true;
if(!isArrayEmptyOrNull(stores)){
query += isFirst ? "where " : "and ";
query += "stock.store.id in ("+ Arrays.stream(stores).map(store -> store.toString()).collect(Collectors.joining(", "))+") ";
isFirst = false;
}
if(!isArrayEmptyOrNull(brands)){
query += isFirst ? "where " : "and ";
query += "article.brand.id in ("+ Arrays.stream(brands).map(store -> store.toString()).collect(Collectors.joining(", "))+") ";
isFirst = false;
}
if(!isArrayEmptyOrNull(articles)){
query += isFirst ? "where " : "and ";
query += "article.id in ("+ Arrays.stream(articles).map(store -> store.toString()).collect(Collectors.joining(", "))+") ";
isFirst = false;
}
if(!isArrayEmptyOrNull(colors)){
query += isFirst ? "where " : "and ";
query += "stock.color.id in ("+ Arrays.stream(colors).map(store -> store.toString()).collect(Collectors.joining(", "))+") ";
}
List<Article> articles = hibernateQueryService.executeHibernateQuery(query, Article.class);
/**
* MapStruct [http://mapstruct.org/][1]
*/
return articles.stream().map(articleMapper::toDto).collect(Collectors.toList());
}
Is there any way to get this query with QueryDSL?
select
person.name,
count(neighbors.*)
from person as neighbors
where person.address = neighbor.address
group by person.name
where address is not a FK.
The key to the first problem is to use aliases to be able to make the self-join:
// static instance uses default alias "person"
QPerson person = QPerson.person;
QPerson neighbor = new QPerson("neighbor");
JPAQuery query = new JPAQuery(em)
.from(person, neighbor)
.where(person.adress.eq(neighbor.adress))
.groupBy(person.name);
To get the results you can simply use the Tuple class:
List<Tuple> results = query.list(person.name, neighbor.count());
for (Tuple row : result) {
System.out.println("name: " + row.get(person.name));
System.out.println("count: " + row.get(neighbor.count()));
}
Or you can use a ConstructorExpression to do something like this:
List<MyResult> results = query.list(ConstructorExpression.create(
MyResult.class, person.name, neighbor.count()));
public class MyResult {
private String name;
private long count;
public MyResult(String name, long count) {
this.name = name;
this.count = count;
}
}
I get a ClassCastException when trying to query my JPA entity class. I only want json to show two columns. That is name and address. How do I only show selected columns in JPA? From debugging it has to be the for loop. So List is the object and I need to make the right side an object instead of a list correct?
Entity
#Entity
#Table(name = "Personnel")
public class User implements Serializable {
private String id;
private String name;
private String address;
public User(String id, String name, String address)
{
this.id = id;
this.name = name;
this.address = address;
}
#Id
#Column(name = "name", unique = true, nullable = false)
public String getName() {
return this.name;
}....
//setters getters
Query/Impl
public List<User> getRecords(User ent){
String sql = "select "
+ " usr.name, usr.address "
+ " from User usr"
+ " where usr.id = '1' ";
List<User> records = this.getSession().createQuery(sql).list();
for ( User event : records ) {
System.out.println( "Records (" + event.getName + ") );
}
return records;
}
Update
This is my attempt to declare the result object as List. Does the method have to be an object instead of ?
public List<User> getRecords(User ent){
String sql = "select "
+ " usr.name, usr.address "
+ " from User usr"
+ " where usr.id = '1' ";
Map<String, String> results = new HashMap<String, String>();
List<Object[]> resultList = this.getSession().createQuery(sql).list();
// Place results in map
for (Object[] items: resultList) {
results.put((String)items[0], (String)items[1]);
results.toString();
}
return (List<User>) results;
You can pass a DTO class to the SELECT clause like this:
List<UserDTO> resultList = this.getSession().createQuery("""
select
new my.package.UserDTO(usr.name, usr.address)
from User usr
where usr.id = :userId
""")
.setParameter("userId", userId)
.getResultList();
You should read the complete object:
String sql = " from User usr"
+ " where usr.id = '1' ";
Or declare the result object as List<Object[]>
You can use generic(I assure you are using java 1.5 and above) and and one you get result, you do type check and downcast to desired type.
If You don't want to read the complete object you have to use Criteria and Projection on specific properties.
See this answer it will help