Introduction to Spring Data MongoDB

1. Overview

This article will be a quick and practical introduction to Spring Data MongoDB.

We’ll go over the basics using both the MongoTemplate as well as MongoRepository using practical examples to illustrate each operation.

2. MongoTemplate and MongoRepository

The MongoTemplate follows the standard template pattern in Spring and provides a ready to go, basic API to the underlying persistence engine.

The repository follows the Spring Data-centric approach and comes with more flexible and complex API operations, based on the well-known access patterns in all Spring Data projects.

For both, we need to start by defining the dependency – for example, in the pom.xml, with Maven:

<dependency>
    <groupId>org.springframework.data</groupId>
    <artifactId>spring-data-mongodb</artifactId>
    <version>3.0.3.RELEASE</version>
</dependency>

To check if any new version of the library has been released – track the releases here.

3. Configuration for MongoTemplate

3.1. XML Configuration

Let’s start with the simple XML configuration for the Mongo template:

<mongo:mongo-client id="mongoClient" host="localhost" />
<mongo:db-factory id="mongoDbFactory" dbname="test" mongo-client-ref="mongoClient" />

First, we need to define the factory bean responsible for creating Mongo instances.

Next – we need to actually define (and configure) the template bean:

<bean id="mongoTemplate" class="org.springframework.data.mongodb.core.MongoTemplate"> 
    <constructor-arg ref="mongoDbFactory"/> 
</bean>

And finally, we need to define a post processor to translate any MongoExceptions thrown in @Repository annotated classes:

<bean class=
  "org.springframework.dao.annotation.PersistenceExceptionTranslationPostProcessor"/>

3.2. Java Configuration

Let’s now create a similar configuration using Java config by extending the base class for MongoDB configuration AbstractMongoConfiguration:

@Configuration
public class MongoConfig extends AbstractMongoClientConfiguration {
 
    @Override
    protected String getDatabaseName() {
        return "test";
    }
 
    @Override
    public MongoClient mongoClient() {
        ConnectionString connectionString = new ConnectionString("mongodb://localhost:27017/test");
        MongoClientSettings mongoClientSettings = MongoClientSettings.builder()
            .applyConnectionString(connectionString)
            .build();
        
        return MongoClients.create(mongoClientSettings);
    }
 
    @Override
    public Collection getMappingBasePackages() {
        return Collections.singleton("com.maixuanviet");
    }
}

Note: We didn’t need to define MongoTemplate bean in the previous configuration as it’s already defined in AbstractMongoClientConfiguration.

We can also use our configuration from scratch without extending AbstractMongoClientConfiguration – as follows:

@Configuration
public class SimpleMongoConfig {
 
    @Bean
    public MongoClient mongo() {
        ConnectionString connectionString = new ConnectionString("mongodb://localhost:27017/test");
        MongoClientSettings mongoClientSettings = MongoClientSettings.builder()
          .applyConnectionString(connectionString)
          .build();
        
        return MongoClients.create(mongoClientSettings);
    }

    @Bean
    public MongoTemplate mongoTemplate() throws Exception {
        return new MongoTemplate(mongo(), "test");
    }
}

4. Configuration for MongoRepository

4.1. XML Configuration

To make use of custom repositories (extending the MongoRepository) – we need to continue the configuration from section 3.1 and set up the repositories:

<mongo:repositories 
  base-package="com.maixuanviet.repository" mongo-template-ref="mongoTemplate"/>

4.2. Java Configuration

Similarly, we’ll build on the configuration we already created in section 3.2 and add a new annotation into the mix:

@EnableMongoRepositories(basePackages = "com.maixuanviet.repository")

4.3. Create the Repository

Now, after the configuration, we need to create a repository – extending the existing MongoRepository interface:

public interface UserRepository extends MongoRepository<User, String> {
    // 
}

Now we can auto-wire this UserRepository and use operations from MongoRepository or add custom operations.

5. Using MongoTemplate

5.1. Insert

Let’s start with the insert operation; let’s also start with an empty database:

{
}

Now if we insert a new user:

User user = new User();
user.setName("Jon");
mongoTemplate.insert(user, "user");

The database will look like this:

{
    "_id" : ObjectId("55b4fda5830b550a8c2ca25a"),
    "_class" : "com.maixuanviet.model.User",
    "name" : "Jon"
}

5.2. Save – Insert

The save operation has save-or-update semantics: if an id is present, it performs an update, if not – it does an insert.

Let’s look at the first semantic – the insert; here’s the initial state of the database:

{
}

When we now save a new user:

User user = new User();
user.setName("Albert"); 
mongoTemplate.save(user, "user");

The entity will be inserted in the database:

{
    "_id" : ObjectId("55b52bb7830b8c9b544b6ad5"),
    "_class" : "com.maixuanviet.model.User",
    "name" : "Albert"
}

Next, we’ll look at the same operation – save – with update semantics.

5.3. Save – Update

Let’s now look at save with update semantics, operating on an existing entity:

{
    "_id" : ObjectId("55b52bb7830b8c9b544b6ad5"),
    "_class" : "com.maixuanviet.model.User",
    "name" : "Jack"
}

Now, when we save the existing user – we will update it:

user = mongoTemplate.findOne(
  Query.query(Criteria.where("name").is("Jack")), User.class);
user.setName("Jim");
mongoTemplate.save(user, "user");

As you can see, in this particular example, save uses the semantics of update, because we use an object with given _id.

5.4. UpdateFirst

updateFirst updates the very first document that matches the query.

Let’s start with the initial state of the database:

[
    {
        "_id" : ObjectId("55b5ffa5511fee0e45ed614b"),
        "_class" : "com.maixuanviet.model.User",
        "name" : "Alex"
    },
    {
        "_id" : ObjectId("55b5ffa5511fee0e45ed614c"),
        "_class" : "com.maixuanviet.model.User",
        "name" : "Alex"
    }
]

When we now run the updateFirst:

Query query = new Query();
query.addCriteria(Criteria.where("name").is("Alex"));
Update update = new Update();
update.set("name", "James");
mongoTemplate.updateFirst(query, update, User.class);

Only the first entry will be updated:

[
    {
        "_id" : ObjectId("55b5ffa5511fee0e45ed614b"),
        "_class" : "com.maixuanviet.model.User",
        "name" : "James"
    },
    {
        "_id" : ObjectId("55b5ffa5511fee0e45ed614c"),
        "_class" : "com.maixuanviet.model.User",
        "name" : "Alex"
    }
]

5.5. UpdateMulti

pdateMulti updates all document that matches the given query.

First – here’s the state of the database before doing the updateMulti:

[
    {
        "_id" : ObjectId("55b5ffa5511fee0e45ed614b"),
        "_class" : "com.maixuanviet.model.User",
        "name" : "Eugen"
    },
    {
        "_id" : ObjectId("55b5ffa5511fee0e45ed614c"),
        "_class" : "com.maixuanviet.model.User",
        "name" : "Eugen"
    }
]

Now, let’s now run the updateMulti operation:

Query query = new Query();
query.addCriteria(Criteria.where("name").is("Eugen"));
Update update = new Update();
update.set("name", "Victor");
mongoTemplate.updateMulti(query, update, User.class);

Both existing objects will be updated in the database:

[
    {
        "_id" : ObjectId("55b5ffa5511fee0e45ed614b"),
        "_class" : "com.maixuanviet.model.User",
        "name" : "Victor"
    },
    {
        "_id" : ObjectId("55b5ffa5511fee0e45ed614c"),
        "_class" : "com.maixuanviet.model.User",
        "name" : "Victor"
    }
]

5.6. FindAndModify

This operation works like updateMulti, but it returns the object before it was modified.

First – the state of the database before calling findAndModify:

{
    "_id" : ObjectId("55b5ffa5511fee0e45ed614b"),
    "_class" : "com.maixuanviet.model.User",
    "name" : "Markus"
}

Let’s look at the actual operation code:

Query query = new Query();
query.addCriteria(Criteria.where("name").is("Markus"));
Update update = new Update();
update.set("name", "Nick");
User user = mongoTemplate.findAndModify(query, update, User.class);

The returned user object has the same values as the initial state in the database.

However, the new state in the database is:

{
    "_id" : ObjectId("55b5ffa5511fee0e45ed614b"),
    "_class" : "com.maixuanviet.model.User",
    "name" : "Nick"
}

5.7. Upsert

The upsert works operate on the find and modify else create semantics: if the document is matched, update it, else create a new document by combining the query and update object.

Let’s start with the initial state of the database:

{
    "_id" : ObjectId("55b5ffa5511fee0e45ed614b"),
    "_class" : "com.maixuanviet.model.User",
    "name" : "Markus"
}

Now – let’s run the upsert:

Query query = new Query();
query.addCriteria(Criteria.where("name").is("Markus"));
Update update = new Update();
update.set("name", "Nick");
mongoTemplate.upsert(query, update, User.class);

Here’s the state of the database after the operation:

{
    "_id" : ObjectId("55b5ffa5511fee0e45ed614b"),
    "_class" : "com.maixuanviet.model.User",
    "name" : "Nick"
}

5.8. Remove

The state of the database before calling remove:

{
    "_id" : ObjectId("55b5ffa5511fee0e45ed614b"),
    "_class" : "com.maixuanviet.model.User",
    "name" : "Benn"
}

Let’s now run remove:

mongoTemplate.remove(user, "user");

The result will be as expected:

{
}

6. Using MongoRepository

6.1. Insert

First – the state of the database before running the insert:

{
}

Now, when we insert a new user:

User user = new User();
user.setName("Jon");
userRepository.insert(user);

Here’s the end state of the database:

{
    "_id" : ObjectId("55b4fda5830b550a8c2ca25a"),
    "_class" : "com.maixuanviet.model.User",
    "name" : "Jon"
}

Note how the operation works the same as the insert in the MongoTemplate API.

6.2. Save – Insert

Similarly – save works the same as the save operation in the MongoTemplate API.

Let’s start by looking at the insert semantics of the operation; here’s the initial state of the database:

{
}

Now – we execute the save operation:

User user = new User();
user.setName("Aaron");
userRepository.save(user);

This results in the user being added to the database:

{
    "_id" : ObjectId("55b52bb7830b8c9b544b6ad5"),
    "_class" : "com.maixuanviet.model.User",
    "name" : "Aaron"
}

Note again how, in this example, save works with insert semantics, because we are inserting a new object.

6.3. Save – Update

Let’s now look at the same operation but with update semantics.

First – here’s the state of the database before running the new save:

{
    "_id" : ObjectId("55b52bb7830b8c9b544b6ad5"),
    "_class" : "com.maixuanviet.model.User",
    "name" : "Jack"81*6
}

Now – we execute the operation:

user = mongoTemplate.findOne(
  Query.query(Criteria.where("name").is("Jack")), User.class);
user.setName("Jim");
userRepository.save(user);

Finally, here is the state of the database:

{
    "_id" : ObjectId("55b52bb7830b8c9b544b6ad5"),
    "_class" : "com.maixuanviet.model.User",
    "name" : "Jim"
}

Note again how, in this example, save works with update semantics, because we are using an existing object.

6.4. Delete

The state of the database before calling delete:

{
    "_id" : ObjectId("55b5ffa5511fee0e45ed614b"),
    "_class" : "com.maixuanviet.model.User",
    "name" : "Benn"
}

Let’s run delete:

userRepository.delete(user);

The result will simply be:

{
}

6.5. FindOne

The state of the database when findOne is called:

{
    "_id" : ObjectId("55b5ffa5511fee0e45ed614b"),
    "_class" : "com.maixuanviet.model.User",
    "name" : "Chris"
}

Let’s now execute the findOne:

userRepository.findOne(user.getId())

The result which will return the existing data:

{
    "_id" : ObjectId("55b5ffa5511fee0e45ed614b"),
    "_class" : "com.maixuanviet.model.User",
    "name" : "Chris"
}

6.6. Exists

The state of the database before calling exists:

{
    "_id" : ObjectId("55b5ffa5511fee0e45ed614b"),
    "_class" : "com.maixuanviet.model.User",
    "name" : "Harris"
}

Now, let’s run exists:

boolean isExists = userRepository.exists(user.getId());

Which of course will return true.

6.7. FindAll With Sort

The state of the database before calling findAll:

[
    {
        "_id" : ObjectId("55b5ffa5511fee0e45ed614b"),
        "_class" : "com.maixuanviet.model.User",
        "name" : "Brendan"
    },
    {
       "_id" : ObjectId("67b5ffa5511fee0e45ed614b"),
       "_class" : "com.maixuanviet.model.User",
       "name" : "Adam"
    }
]

Let’s now run findAll with Sort:

List<User> users = userRepository.findAll(Sort.by(Sort.Direction.ASC, "name"));

The result will be sorted by name in ascending order:

[
    {
        "_id" : ObjectId("67b5ffa5511fee0e45ed614b"),
        "_class" : "com.maixuanviet.model.User",
        "name" : "Adam"
    },
    {
        "_id" : ObjectId("55b5ffa5511fee0e45ed614b"),
        "_class" : "com.maixuanviet.model.User",
        "name" : "Brendan"
    }
]

6.8. FindAll With Pageable

The state of the database before calling findAll:

[
    {
        "_id" : ObjectId("55b5ffa5511fee0e45ed614b"),
        "_class" : "com.maixuanviet.model.User",
        "name" : "Brendan"
    },
    {
        "_id" : ObjectId("67b5ffa5511fee0e45ed614b"),
        "_class" : "com.maixuanviet.model.User",
        "name" : "Adam"
    }
]

Let’s now execute findAll with a pagination request:

Pageable pageableRequest = PageRequest.of(0, 1);
Page<User> page = userRepository.findAll(pageableRequest);
List<User> users = pages.getContent();

The resulting users list will be only one user:

{
    "_id" : ObjectId("55b5ffa5511fee0e45ed614b"),
    "_class" : "com.maixuanviet.model.User",
    "name" : "Brendan"
}

7. Annotations

Finally, let’s also go over the simple annotations that Spring Data uses to drive these API operations.

@Id
private String id;

The field level @Id annotation can decorate any type, including long and string.

If the value of the @Id field is not null, it’s stored in the database as-is; otherwise, the converter will assume you want to store an ObjectId in the database (either ObjectId, String or BigInteger work).

Next – @Document:

@Document
public class User {
    //
}

This annotation simply marks a class as being a domain object that needs to be persisted to the database, along with allowing us to choose the name of the collection to be used.

8. Conclusion

This article was a quick but comprehensive introduction to using MongoDB with Spring Data, both via the MongoTemplate API as well as making use of MongoRepository.

The implementation of all these examples and code snippets can be found over on Github.