1. 概述
Spring Data provides many ways to define a query that we can execute. One of these is the @Query annotation.
In this tutorial, we’ll demonstrate how to use the @Query annotation in Spring Data JPA to execute both JPQL and native SQL queries.
We’ll also show how to build a dynamic query when the @Query annotation is not enough.
2. Select 查询
In order to define SQL to execute for a Spring Data repository method, we can annotate the method with the @Query annotation — its value attribute contains the JPQL or SQL to execute.
The @Query annotation takes precedence over named queries, which are annotated with @NamedQuery or defined in an orm.xml file.
It’s a good approach to place a query definition just above the method inside the repository rather than inside our domain model as named queries. The repository is responsible for persistence, so it’s a better place to store these definitions.
2.1. JPQL
By default, the query definition uses JPQL.
Let’s look at a simple repository method that returns active User entities from the database:
@Query("SELECT u FROM User u WHERE u.status = 1")
Collection<User> findAllActiveUsers();
2.2. Native
We can use also native SQL to define our query. All we have to do is set the value of the nativeQuery attribute to true and define the native SQL query in the value attribute of the annotation:
@Query(
value = "SELECT * FROM USERS u WHERE u.status = 1",
nativeQuery = true)
Collection<User> findAllActiveUsersNative();
3. 排序操作
We can pass an additional parameter of type Sort to a Spring Data method declaration that has the @Query annotation. It’ll be translated into the ORDER BY clause that gets passed to the database.
3.1. Sorting for JPA Provided and Derived Methods
For the methods we get out of the box such as findAll(Sort) or the ones that are generated by parsing method signatures, we can only use object properties to define our sort:
userRepository.findAll(Sort.by(Sort.Direction.ASC, "name"));
Now imagine that we want to sort by the length of a name property:
userRepository.findAll(Sort.by("LENGTH(name)"));
When we execute the above code, we’ll receive an exception:
org.springframework.data.mapping.PropertyReferenceException: No property LENGTH(name) found for type User!
3.2. JPQL
When we use JPQL for a query definition, then Spring Data can handle sorting without any problem — all we have to do is add a method parameter of type Sort:
@Query(value = "SELECT u FROM User u")
List<User> findAllUsers(Sort sort);
We can call this method and pass a Sort parameter, which will order the result by the name property of the User object:
userRepository.findAllUsers(Sort.by("name"));
And because we used the @Query annotation, we can use the same method to get the sorted list of Users by the length of their names:
userRepository.findAllUsers(JpaSort.unsafe("LENGTH(name)"));
It’s crucial that we use JpaSort.unsafe() to create a Sort object instance.
When we use:
Sort.by("LENGTH(name)");
then we’ll receive the same exception as we saw above for the findAll() method.
When Spring Data discovers the unsafe Sort order for a method that uses the @Query annotation, then it just appends the sort clause to the query — it skips checking whether the property to sort by belongs to the domain model.
3.3. Native
When the @Query annotation uses native SQL, then it’s not possible to define a Sort.
If we do, we’ll receive an exception:
org.springframework.data.jpa.repository.query.InvalidJpaQueryMethodException: Cannot use native queries with dynamic sorting and/or pagination
As the exception says, the sort isn’t supported for native queries. The error message gives us a hint that pagination will cause an exception too.
However, there is a workaround that enables pagination, and we’ll cover it in the next section.
4. Join 表
When we use join clauses in queries in the @Query annotation, there are multiple ways to define the select clause when handling columns from multiple tables.
4.1. JPQL
When using JPQL queries, we need to create projections/DTOs to return the required fields from the joining tables. For example:
@Query(value = "SELECT new ResultDTO(c.id, o.id, p.id, c.name, c.email, o.orderDate, p.productName, p.price) "
+ " from Customer c, CustomerOrder o ,Product p "
+ " where c.id=o.customer.id "
+ " and o.id=p.customerOrder.id "
+ " and c.id=?1 ")
List<ResultDTO> findResultDTOByCustomer(Long id);
The DTO and ResultDTO classes are defined as follows:
class DTO {
private Long customer_id;
private Long order_id;
private Long product_id;
public DTO(Long customer_id, Long order_id, Long product_id) {
this.customer_id = customer_id;
this.order_id = order_id;
this.product_id = product_id;
}
}
@Entity
@IdClass(DTO.class)
public class ResultDTO {
@Id
private Long customer_id;
@Id
private Long order_id;
@Id
private Long product_id;
private String customerName;
private String customerEmail;
private LocalDate orderDate;
private String productName;
private Double productPrice;
// getters, setters, constructors etc
}
As can be seen, the ResultDTO class has a composite primary key which is defined in a separate class DTO and referenced via @IdClass annotation. This is only required if the repository method returns a DTO with Id as composite keys.
4.2. Native
In the case when we want to use native queries to join tables and fetch results using Spring Data JPA API, we cannot specify a subset of columns to be selected for the objects in the select clause, rather, we have to create repository methods like below:
@Query(value = "SELECT c.*, o.*, p.* "
+ " from Customer c, CustomerOrder o ,Product p "
+ " where c.id=o.customer_id "
+ " and o.id=p.customerOrder_id "
+ " and c.id=?1 "
, nativeQuery = true)
List<Map<String, Object>> findByCustomer(Long id);
We can see that the return type of the method findByCustomer is a list of maps where the keys in the map correspond to the column names in the nativeQuery. The list itself corresponds to the list of objects returned by the method.
In the nativeQuery approach, we do not need to create any DTO classes, but at the same time, nativeQuery fetches all columns from all the tables being joined in the @Query.
5. 分页操作
Pagination allows us to return just a subset of a whole result in a Page. This is useful, for example, when navigating through several pages of data on a web page.
Another advantage of pagination is that the amount of data sent from server to client is minimized. By sending smaller pieces of data, we can generally see an improvement in performance.
5.1. JPQL
Using pagination in the JPQL query definition is straightforward:
@Query(value = "SELECT u FROM User u ORDER BY id")
Page<User> findAllUsersWithPagination(Pageable pageable);
We can pass a PageRequest parameter to get a page of data.
Pagination is also supported for native queries but requires a little bit of additional work.
5.2. Native
We can enable pagination for native queries by declaring an additional attribute countQuery.
This defines the SQL to execute to count the number of rows in the whole result:
@Query(
value = "SELECT * FROM Users ORDER BY id",
countQuery = "SELECT count(*) FROM Users",
nativeQuery = true)
Page<User> findAllUsersWithPagination(Pageable pageable);
5.3. Spring Data JPA Versions Prior to 2.0.4
The above solution for native queries works fine for Spring Data JPA versions 2.0.4 and later.
Prior to that version, when we try to execute such a query, we’ll receive the same exception we described in the previous section on sorting.
We can overcome this by adding an additional parameter for pagination inside our query:
@Query(
value = "SELECT * FROM Users ORDER BY id \n-- #pageable\n",
countQuery = "SELECT count(*) FROM Users",
nativeQuery = true)
Page<User> findAllUsersWithPagination(Pageable pageable);
In the above example, we add
\n-- #pageable\n
as the placeholder for the pagination parameter. This tells Spring Data JPA how to parse the query and inject the pageable parameter. This solution works for the H2 database.
We’ve covered how to create simple select queries via JPQL and native SQL. Next, we’ll show how to define additional parameters.
6. 查询参数 - 索引方式
There are two possible ways that we can pass method parameters to our query: indexed and named parameters.
In this section, we’ll cover indexed parameters.
6.1. JPQL
For indexed parameters in JPQL, Spring Data will pass method parameters to the query in the same order they appear in the method declaration:
@Query("SELECT u FROM User u WHERE u.status = ?1")
User findUserByStatus(Integer status);
@Query("SELECT u FROM User u WHERE u.status = ?1 and u.name = ?2")
User findUserByStatusAndName(Integer status, String name);
For the above queries, the status method parameter will be assigned to the query parameter with index 1, and the name method parameter will be assigned to the query parameter with index 2.
6.2. Native
Indexed parameters for the native queries work exactly in the same way as for JPQL:
@Query(
value = "SELECT * FROM Users u WHERE u.status = ?1",
nativeQuery = true)
User findUserByStatusNative(Integer status);
In the next section, we’ll show a different approach: passing parameters via name.
7. 查询参数 - 命名变量
We can also pass method parameters to the query using named parameters. We define these using the @Param annotation inside our repository method declaration.
Each parameter annotated with @Param must have a value string matching the corresponding JPQL or SQL query parameter name. A query with named parameters is easier to read and is less error-prone in case the query needs to be refactored.
7.1. JPQL
As mentioned above, we use the @Param annotation in the method declaration to match parameters defined by name in JPQL with parameters from the method declaration:
@Query("SELECT u FROM User u WHERE u.status = :status and u.name = :name")
User findUserByStatusAndNameNamedParams(
@Param("status") Integer status,
@Param("name") String name);
Note that in the above example, we defined our SQL query and method parameters to have the same names, but it’s not required as long as the value strings are the same:
@Query("SELECT u FROM User u WHERE u.status = :status and u.name = :name")
User findUserByUserStatusAndUserName(@Param("status") Integer userStatus,
@Param("name") String userName);
7.2. Native
For the native query definition, there is no difference in how we pass a parameter via the name to the query in comparison to JPQL — we use the @Param annotation:
@Query(value = "SELECT * FROM Users u WHERE u.status = :status and u.name = :name",
nativeQuery = true)
User findUserByStatusAndNameNamedParamsNative(
@Param("status") Integer status, @Param("name") String name);
8. Collection 类型参数
Let’s consider the case when the where clause of our JPQL or SQL query contains the IN (or NOT IN) keyword:
SELECT u FROM User u WHERE u.name IN :names
In this case, we can define a query method that takes Collection as a parameter:
@Query(value = "SELECT u FROM User u WHERE u.name IN :names")
List<User> findUserByNameList(@Param("names") Collection<String> names);
As the parameter is a Collection, it can be used with List, HashSet, etc.
Next, we’ll show how to modify data with the @Modifying annotation.
9. 使用 @Modifying 执行更新操作
We can use the @Query annotation to modify the state of the database by also adding the @Modifying annotation to the repository method.
9.1. JPQL
The repository method that modifies the data has two differences in comparison to the select query — it has the @Modifying annotation and, of course, the JPQL query uses update instead of select:
@Modifying
@Query("update User u set u.status = :status where u.name = :name")
int updateUserSetStatusForName(@Param("status") Integer status,
@Param("name") String name);
The return value defines how many rows are updated by the execution of the query. Both indexed and named parameters can be used inside update queries.
9.2. Native
We can modify the state of the database also with a native query. We just need to add the @Modifying annotation:
@Modifying
@Query(value = "update Users u set u.status = ? where u.name = ?",
nativeQuery = true)
int updateUserSetStatusForNameNative(Integer status, String name);
9.3. 数据插入
To perform an insert operation, we have to both apply @Modifying and use a native query since INSERT is not a part of the JPA interface:
@Modifying
@Query(
value =
"insert into Users (name, age, email, status) values (:name, :age, :email, :status)",
nativeQuery = true)
void insertUser(@Param("name") String name, @Param("age") Integer age,
@Param("status") Integer status, @Param("email") String email);
10. 动态查询语句
Often, we’ll encounter the need for building SQL statements based on conditions or data sets whose values are only known at runtime. And in those cases, we can’t just use a static query.
10.1. Example of a Dynamic Query
For example, let’s imagine a situation where we need to select all the users whose email is LIKE one from a set defined at runtime — email1, email2, …, emailn:
SELECT u FROM User u WHERE u.email LIKE '%email1%'
or u.email LIKE '%email2%'
...
or u.email LIKE '%emailn%'
Since the set is dynamically constructed, we can’t know at compile-time how many LIKE clauses to add.
In this case, we can’t just use the @Query annotation since we can’t provide a static SQL statement.
Instead, by implementing a custom composite repository, we can extend the base JpaRepository functionality and provide our own logic for building a dynamic query. Let’s take a look at how to do this.
10.2. Custom Repositories and the JPA Criteria API
Luckily for us, Spring provides a way for extending the base repository through the use of custom fragment interfaces. We can then link them together to create a composite repository.
We’ll start by creating a custom fragment interface:
public interface UserRepositoryCustom {
List<User> findUserByEmails(Set<String> emails);
}
And then we’ll implement it:
public class UserRepositoryCustomImpl implements UserRepositoryCustom {
@PersistenceContext
private EntityManager entityManager;
@Override
public List<User> findUserByEmails(Set<String> emails) {
CriteriaBuilder cb = entityManager.getCriteriaBuilder();
CriteriaQuery<User> query = cb.createQuery(User.class);
Root<User> user = query.from(User.class);
Path<String> emailPath = user.get("email");
List<Predicate> predicates = new ArrayList<>();
for (String email : emails) {
predicates.add(cb.like(emailPath, email));
}
query.select(user)
.where(cb.or(predicates.toArray(new Predicate[predicates.size()])));
return entityManager.createQuery(query)
.getResultList();
}
}
As shown above, we leveraged the JPA Criteria API to build our dynamic query.
Also, we need to make sure to include the Impl postfix in the class name. Spring will search the UserRepositoryCustom implementation as UserRepositoryCustomImpl. Since fragments are not repositories by themselves, Spring relies on this mechanism to find the fragment implementation.
10.3. 继承现有Repository
Notice that all the query methods from section 2 through section 7 are in the UserRepository.
So, now we’ll integrate our fragment by extending the new interface in the UserRepository:
public interface UserRepository extends JpaRepository<User, Integer>, UserRepositoryCustom {
// query methods from section 2 - section 7
}
10.4. Using the Repository
And finally, we can call our dynamic query method:
Set<String> emails = new HashSet<>();
// filling the set with any number of items
userRepository.findUserByEmails(emails);
We’ve successfully created a composite repository and called our custom method.
11. 总结
In this article, we covered several ways of defining queries in Spring Data JPA repository methods using the @Query annotation.
We also learned how to implement a custom repository and create a dynamic query.
As always, the complete code examples used in this article are available over on GitHub.