Java Automation Job Top
We’re looking for a Backend Java/Spring Developer with Integration Experience: Read More
announcement-icon.png

JPA can behave very differently depending on the exact circumstances under which it is used. Code that works in our local environment or in staging performs very poorly (or even flat out fails) when thrown against real-scale databases in production environments.

Debugging these JPA issues in production is pretty difficult - existing APMs don’t provide enough granular insights at the code level, and tracking every single place someone queried entities one by one instead of in bulk can be a grueling, time-consuming task.

Lightrun is a new approach to debugging in production. Using Lightrun’s Logs and Snapshots, you can now get debugger-level granularity in production without opening inbound ports, redeploying, restarting, or even stropping the running application.

In addition, instrumenting Lightrun Metrics at runtime allows you to track down persistence issues securely and in real-time. Want to see it in action? Check out our 2-minute tutorial for debugging JPA performance issues in production using Lightrun:

>> Debugging Spring Persistence and JPA Issues Using Lightrun

1. Overview

In this tutorial, we're looking at building a query language for a REST API using Spring Data JPA and Querydsl.

In the first two articles of this series, we built the same search/filtering functionality using JPA Criteria and Spring Data JPA Specifications.

So – why a query language? Because – for any complex enough API – searching/filtering your resources by very simple fields is simply not enough. A query language is more flexible, and allows you to filter down to exactly the resources you need.

2. Querydsl Configuration

First – let's see how to configure our project to use Querydsl.

We need to add the following dependencies to pom.xml:

<dependency> 
    <groupId>com.querydsl</groupId> 
    <artifactId>querydsl-apt</artifactId> 
    <version>4.2.2</version>
    </dependency>
<dependency> 
    <groupId>com.querydsl</groupId> 
    <artifactId>querydsl-jpa</artifactId> 
    <version>4.2.2</version> 
</dependency>

We also need to configure the APT – Annotation processing tool – plugin as follows:

<plugin>
    <groupId>com.mysema.maven</groupId>
    <artifactId>apt-maven-plugin</artifactId>
    <version>1.1.3</version>
    <executions>
        <execution>
            <goals>
                <goal>process</goal>
            </goals>
            <configuration>
                <outputDirectory>target/generated-sources/java</outputDirectory>
                <processor>com.mysema.query.apt.jpa.JPAAnnotationProcessor</processor>
            </configuration>
        </execution>
    </executions>
</plugin>

This will generate the Q-types for our entities. 

3. The MyUser Entity

Next – let's take a look at the “MyUser” entity which we are going to use in our Search API:

@Entity
public class MyUser {
    @Id
    @GeneratedValue(strategy = GenerationType.AUTO)
    private Long id;

    private String firstName;
    private String lastName;
    private String email;

    private int age;
}

4. Custom Predicate With PathBuilder

Now – let's create a custom Predicate based on some arbitrary constraints.

We're using PathBuilder here instead of the automatically generated Q-types because we need to create paths dynamically for more abstract usage:

public class MyUserPredicate {

    private SearchCriteria criteria;

    public BooleanExpression getPredicate() {
        PathBuilder<MyUser> entityPath = new PathBuilder<>(MyUser.class, "user");

        if (isNumeric(criteria.getValue().toString())) {
            NumberPath<Integer> path = entityPath.getNumber(criteria.getKey(), Integer.class);
            int value = Integer.parseInt(criteria.getValue().toString());
            switch (criteria.getOperation()) {
                case ":":
                    return path.eq(value);
                case ">":
                    return path.goe(value);
                case "<":
                    return path.loe(value);
            }
        } 
        else {
            StringPath path = entityPath.getString(criteria.getKey());
            if (criteria.getOperation().equalsIgnoreCase(":")) {
                return path.containsIgnoreCase(criteria.getValue().toString());
            }
        }
        return null;
    }
}

Note how the implementation of the predicate is generically dealing with multiple types of operations. This is because the query language is by definition an open language where you can potentially filter by any field, using any supported operation.

To represent that kind of open filtering criteria, we're using a simple but quite flexible implementation – SearchCriteria:

public class SearchCriteria {
    private String key;
    private String operation;
    private Object value;
}

The SearchCriteria holds the details we need to represent a constraint:

  • key: the field name – for example: firstName, age, … etc
  • operation: the operation – for example: Equality, less than, … etc
  • value: the field value – for example: john, 25, … etc

5. MyUserRepository

Now – let’s take a look at our MyUserRepository.

We need our MyUserRepository to extend QuerydslPredicateExecutor so that we can use Predicates later to filter search results:

public interface MyUserRepository extends JpaRepository<MyUser, Long>, 
  QuerydslPredicateExecutor<MyUser>, QuerydslBinderCustomizer<QMyUser> {
    @Override
    default public void customize(
      QuerydslBindings bindings, QMyUser root) {
        bindings.bind(String.class)
          .first((SingleValueBinding<StringPath, String>) StringExpression::containsIgnoreCase);
        bindings.excluding(root.email);
      }
}

Note that we're using here the generated Q-type for the MyUser entity, which will be named QMyUser.

6. Combine Predicates

Next– let’s take a look at combining Predicates to use multiple constraints in results filtering.

In the following example – we work with a builder – MyUserPredicatesBuilder – to combine Predicates:

public class MyUserPredicatesBuilder {
    private List<SearchCriteria> params;

    public MyUserPredicatesBuilder() {
        params = new ArrayList<>();
    }

    public MyUserPredicatesBuilder with(
      String key, String operation, Object value) {
  
        params.add(new SearchCriteria(key, operation, value));
        return this;
    }

    public BooleanExpression build() {
        if (params.size() == 0) {
            return null;
        }

        List predicates = params.stream().map(param -> {
            MyUserPredicate predicate = new MyUserPredicate(param);
            return predicate.getPredicate();
        }).filter(Objects::nonNull).collect(Collectors.toList());
        
        BooleanExpression result = Expressions.asBoolean(true).isTrue();
        for (BooleanExpression predicate : predicates) {
            result = result.and(predicate);
        }        
        return result;
    }
}

7. Test the Search Queries

Next – let’s test our Search API.

We'll start by initializing the database with a few users – to have these ready and available for testing:

@RunWith(SpringJUnit4ClassRunner.class)
@ContextConfiguration(classes = { PersistenceConfig.class })
@Transactional
@Rollback
public class JPAQuerydslIntegrationTest {

    @Autowired
    private MyUserRepository repo;

    private MyUser userJohn;
    private MyUser userTom;

    @Before
    public void init() {
        userJohn = new MyUser();
        userJohn.setFirstName("John");
        userJohn.setLastName("Doe");
        userJohn.setEmail("[email protected]");
        userJohn.setAge(22);
        repo.save(userJohn);

        userTom = new MyUser();
        userTom.setFirstName("Tom");
        userTom.setLastName("Doe");
        userTom.setEmail("[email protected]");
        userTom.setAge(26);
        repo.save(userTom);
    }
}

Next, let’s see how to find users with given last name:

@Test
public void givenLast_whenGettingListOfUsers_thenCorrect() {
    MyUserPredicatesBuilder builder = new MyUserPredicatesBuilder().with("lastName", ":", "Doe");

    Iterable<MyUser> results = repo.findAll(builder.build());
    assertThat(results, containsInAnyOrder(userJohn, userTom));
}

Now, let’s see how to find a user with given both first and last name:

@Test
public void givenFirstAndLastName_whenGettingListOfUsers_thenCorrect() {
    MyUserPredicatesBuilder builder = new MyUserPredicatesBuilder()
      .with("firstName", ":", "John").with("lastName", ":", "Doe");

    Iterable<MyUser> results = repo.findAll(builder.build());

    assertThat(results, contains(userJohn));
    assertThat(results, not(contains(userTom)));
}

Next, let’s see how to find user with given both last name and minimum age

@Test
public void givenLastAndAge_whenGettingListOfUsers_thenCorrect() {
    MyUserPredicatesBuilder builder = new MyUserPredicatesBuilder()
      .with("lastName", ":", "Doe").with("age", ">", "25");

    Iterable<MyUser> results = repo.findAll(builder.build());

    assertThat(results, contains(userTom));
    assertThat(results, not(contains(userJohn)));
}

Now, let’s see how to search for MyUser that doesn’t actually exist:

@Test
public void givenWrongFirstAndLast_whenGettingListOfUsers_thenCorrect() {
    MyUserPredicatesBuilder builder = new MyUserPredicatesBuilder()
      .with("firstName", ":", "Adam").with("lastName", ":", "Fox");

    Iterable<MyUser> results = repo.findAll(builder.build());
    assertThat(results, emptyIterable());
}

Finally – let’s see how to find a MyUser given only part of the first name – as in the following example:

@Test
public void givenPartialFirst_whenGettingListOfUsers_thenCorrect() {
    MyUserPredicatesBuilder builder = new MyUserPredicatesBuilder().with("firstName", ":", "jo");

    Iterable<MyUser> results = repo.findAll(builder.build());

    assertThat(results, contains(userJohn));
    assertThat(results, not(contains(userTom)));
}

8. UserController

Finally, let's put everything together and build the REST API.

We're defining a UserController that defines a simple method findAll() with a “search“ parameter to pass in the query string:

@Controller
public class UserController {

    @Autowired
    private MyUserRepository myUserRepository;

    @RequestMapping(method = RequestMethod.GET, value = "/myusers")
    @ResponseBody
    public Iterable<MyUser> search(@RequestParam(value = "search") String search) {
        MyUserPredicatesBuilder builder = new MyUserPredicatesBuilder();

        if (search != null) {
            Pattern pattern = Pattern.compile("(\w+?)(:|<|>)(\w+?),");
            Matcher matcher = pattern.matcher(search + ",");
            while (matcher.find()) {
                builder.with(matcher.group(1), matcher.group(2), matcher.group(3));
            }
        }
        BooleanExpression exp = builder.build();
        return myUserRepository.findAll(exp);
    }
}

Here is a quick test URL example:

http://localhost:8080/myusers?search=lastName:doe,age>25

And the response:

[{
    "id":2,
    "firstName":"tom",
    "lastName":"doe",
    "email":"[email protected]",
    "age":26
}]

9. Conclusion

This third article covered the first steps of building a query language for a REST API, making good use of the Querydsl library.

The implementation is of course early on, but it can easily be evolved to support additional operations.

The full implementation of this article can be found in the GitHub project – this is a Maven-based project, so it should be easy to import and run as it is.

Next »
REST Query Language – Advanced Search Operations
« Previous
REST Query Language with Spring Data JPA Specifications
REST bottom

Get started with Spring 5 and Spring Boot 2, through the Learn Spring course :

>> CHECK OUT THE COURSE
Persistence bottom
Get started with Spring Data JPA through the reference Learn Spring Data JPA course: >> CHECK OUT THE COURSE
REST footer banner
6 Comments
Oldest
Newest
Inline Feedbacks
View all comments
Comments are closed on this article!