The Spring Data forum is a message board for all Spring Data (not just Document) users to share information and help each other. Note that registration is needed only for posting.

Professional, from-the-source support, with guaranteed response time, is available from SpringSource, the company behind Spring Data and Spring.

As a very first step you define a domain class specific repository interface. It's got to extend Repository and be typed to the domain class and an ID type. If you want to expose CRUD methods for that domain type, extend CrudRepository instead of Repository.

interface MyBaseRepository<T, ID extends Serializable> extends Repository<T, ID> {
  T findOne(ID id);
  T save(T entity);
}

interface UserRepository extends MyBaseRepository<User, Long> {

  User findByEmailAddress(EmailAddress emailAddress);
}

So now the question is how to create instances and bean definitions for the repository interfaces defined.

<?xml version="1.0" encoding="UTF-8"?>
<beans:beans xmlns:beans="http://www.springframework.org/schema/beans"
  xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
  xmlns="http://www.springframework.org/schema/data/jpa"
  xsi:schemaLocation="http://www.springframework.org/schema/beans
    http://www.springframework.org/schema/beans/spring-beans.xsd
    http://www.springframework.org/schema/data/jpa
    http://www.springframework.org/schema/data/jpa/spring-jpa.xsd">

  <repositories base-package="com.acme.repositories" />

</beans:beans>

<repositories base-package="com.acme.repositories">
  <context:exclude-filter type="regex" expression=".*SomeRepository" />
</repositories>

<repositories base-package="com.acme.repositories">
  <repository id="userRepository" />
</repositories>

RepositoryFactorySupport factory = … // Instantiate factory here
UserRepository repository = factory.getRepository(UserRepository.class);

Often it is necessary to provide a custom implementation for a few repository methods. Spring Data repositories easily allow you to provide custom repository code and integrate it with generic CRUD abstraction and query method functionality. To enrich a repository with custom functionality you have to define an interface and an implementation for that functionality first and let the repository interface you provided so far extend that custom interface.

interface UserRepositoryCustom {

  public void someCustomMethod(User user);
}

class UserRepositoryImpl implements UserRepositoryCustom {

  public void someCustomMethod(User user) {
    // Your custom implementation
  }
}

public interface UserRepository extends CrudRepository<User, Long>, UserRepositoryCustom {

  // Declare query methods here
}

<repositories base-package="com.acme.repository">
  <repository id="userRepository" />
</repositories>

<repositories base-package="com.acme.repository" repository-impl-postfix="FooBar">
  <repository id="userRepository" />
</repositories>

<repositories base-package="com.acme.repository">
  <repository id="userRepository" />
</repositories>

<beans:bean id="userRepositoryImpl" class="…">
  <!-- further configuration -->
</beans:bean>

<repositories base-package="com.acme.repository">
  <repository id="userRepository" repository-impl-ref="customRepositoryImplementation" />
</repositories>

<bean id="customRepositoryImplementation" class="…">
  <!-- further configuration -->
</bean>

In other cases you might want to add a single method to all of your repository interfaces. So the approach just shown is not feasible. The first step to achieve this is adding and intermediate interface to declare the shared behaviour

public interface MyRepository<T, ID extends Serializable>
  extends JpaRepository<T, ID> {

  void sharedCustomMethod(ID id);
}

Now your individual repository interfaces will extend this intermediate interface to include the functionality declared. The second step is to create an implementation of this interface that extends the persistence technology specific repository base class which will act as custom base class for the repository proxies then.

Note

If you're using automatic repository interface detection using the Spring namespace using the interface just as is will cause Spring to create an instance of MyRepository. This is of course not desired as it just acts as intermediary between Repository and the actual repository interfaces you want to define for each entity. To exclude an interface extending Repository from being instantiated as repository instance annotate it with @NoRepositoryBean.

public class MyRepositoryImpl<T, ID extends Serializable>
  extends SimpleJpaRepository<T, ID> implements MyRepository<T, ID> {

  public void sharedCustomMethod(ID id) {
    // implementation goes here
  }
}

The last step to get this implementation used as base class for Spring Data repositories is replacing the standard RepositoryFactoryBean with a custom one using a custom RepositoryFactory that in turn creates instances of your MyRepositoryImpl class.

public class MyRepositoryFactoryBean<T extends JpaRepository<?, ?>
  extends JpaRepositoryFactoryBean<T> {

  protected RepositoryFactorySupport getRepositoryFactory(…) {
    return new MyRepositoryFactory(…);
  }

  private static class MyRepositoryFactory extends JpaRepositoryFactory{

    public MyRepositoryImpl getTargetRepository(…) {
      return new MyRepositoryImpl(…);
    }

    public Class<? extends RepositorySupport> getRepositoryClass() {
      return MyRepositoryImpl.class;
    }
  }
}

Finally you can either declare beans of the custom factory directly or use the factory-class attribute of the Spring namespace to tell the repository infrastructure to use your custom factory implementation.

<repositories base-package="com.acme.repository"
  factory-class="com.acme.MyRepositoryFactoryBean" />

Given you are developing a Spring MVC web applications you typically have to resolve domain class ids from URLs. By default it's your task to transform that request parameter or URL part into the domain class to hand it layers below then or execute business logic on the entities directly. This should look something like this:

@Controller
@RequestMapping("/users")
public class UserController {

  private final UserRepository userRepository;

  public UserController(UserRepository userRepository) {
    userRepository = userRepository;
  }

  @RequestMapping("/{id}")
  public String showUserForm(@PathVariable("id") Long id, Model model) {
    
    // Do null check for id
    User user = userRepository.findOne(id);
    // Do null check for user
    // Populate model
    return "user";
  }
}

First you pretty much have to declare a repository dependency for each controller to lookup the entity managed by the controller or repository respectively. Beyond that looking up the entity is boilerplate as well as it's always a findOne(…) call. Fortunately Spring provides means to register custom converting components that allow conversion between a String value to an arbitrary type.

<bean class="….web.servlet.mvc.annotation.AnnotationMethodHandlerAdapter">
  <property name="webBindingInitializer">
    <bean class="….web.bind.support.ConfigurableWebBindingInitializer">
      <property name="propertyEditorRegistrars">
        <bean class="org.springframework.data.repository.support.DomainClassPropertyEditorRegistrar" />
      </property>
    </bean>
  </property>
</bean>

@Controller
@RequestMapping("/users")
public class UserController {

  @RequestMapping("/{id}")
  public String showUserForm(@PathVariable("id") User user, Model model) {

    // Do null check for user
    // Populate model
    return "userForm";
  }
}

<mvc:annotation-driven conversion-service="conversionService" />

<bean id="conversionService" class="….context.support.ConversionServiceFactoryBean">
  <property name="converters">
    <list>
      <bean class="org.springframework.data.repository.support.DomainClassConverter">
        <constructor-arg ref="conversionService" />
      </bean>
    </list>
  </property>
</bean>

@Controller
@RequestMapping("/users")
public class UserController {

  // DI code omitted

  @RequestMapping
  public String showUsers(Model model, HttpServletRequest request) {

    int page = Integer.parseInt(request.getParameter("page"));
    int pageSize = Integer.parseInt(request.getParameter("pageSize"));
    model.addAttribute("users", userService.getUsers(pageable));
    return "users";
  }
}

As you can see the naive approach requires the method to contain an HttpServletRequest parameter that has to be parsed manually. We even omitted an appropriate failure handling which would make the code even more verbose. The bottom line is that the controller actually shouldn't have to handle the functionality of extracting pagination information from the request. So we include a PageableArgumentResolver that will do the work for you.

<bean class="….web.servlet.mvc.annotation.AnnotationMethodHandlerAdapter">
  <property name="customArgumentResolvers">
    <list>
      <bean class="org.springframework.data.web.PageableArgumentResolver" />
    </list>
  </property>
</bean>

This configuration allows you to simplify controllers down to something like this:

@Controller
@RequestMapping("/users")
public class UserController {

  @RequestMapping
  public String showUsers(Model model, Pageable pageable) {

    model.addAttribute("users", userDao.readAll(pageable));
    return "users";
  }
}

The PageableArgumentResolver will automatically resolve request parameters to build a PageRequest instance. By default it will expect the following structure for the request parameters:

In case you need multiple Pageables to be resolved from the request (for multiple tables e.g.) you can use Spring's @Qualifier annotation to distinguish one from another. The request parameters then have to be prefixed with ${qualifier}_. So a method signature like this:

public String showUsers(Model model, 
      @Qualifier("foo") Pageable first,
      @Qualifier("bar") Pageable second) { … }

you'd have to populate foo_page and bar_page and the according subproperties.

public String showUsers(Model model, 
  @PageableDefaults(pageNumber = 0, value = 30) Pageable pageable) { … }

There were several API changes introduced in the M3 release. To upgrade from M2 to M3 you will need to make. For a full listing of API changes please refer to this JDiff Report.

The major changes are with respect to MongoTemplate

An example of using Java based bean metadata to register an instance of a com.mongodb.Mongo is shown below

@Configuration
public class AppConfig {

  /*
   * Use the standard Mongo driver API to create a com.mongodb.Mongo instance.
   */
   public @Bean Mongo mongo() throws UnknownHostException {
       return new Mongo("localhost");
   }
}      

This approach allows you to use the standard com.mongodb.Mongo API that you may already be used to using but also pollutes the code with the UnknownHostException checked exception. The use of the checked exception is not desirable as Java based bean metadata uses methods as a means to set object dependencies, making the calling code cluttered.

An alternative is to register an instance of com.mongodb.Mongo instance with the container using Spring's MongoFactoryBean. As compared to instantiating a com.mongodb.Mongo instance directly, the FactoryBean approach does not throw a checked exception and has the added advantage of also providing the container with an ExceptionTranslator implementation that translates MongoDB exceptions to exceptions in Spring's portable DataAccessException hierarchy for data access classes annoated with the @Repository annotation. This hierarchy and use of @Repository is described in Spring's DAO support features.

An example of a Java based bean metadata that supports exception translation on @Repository annotated classes is shown below:

@Configuration
public class AppConfig {

    /*
     * Factory bean that creates the com.mongodb.Mongo instance
     */
     public @Bean MongoFactoryBean mongo() {
          MongoFactoryBean mongo = new MongoFactoryBean();
          mongo.setHost("localhost");
          return mongo;
     }
}

While you can use Spring's traditional <beans/> XML namespace to register an instance of com.mongodb.Mongo with the container, the XML can be quite verbose as it is general purpose. XML namespaces are a better alternative to configuring commonly used objects such as the Mongo instance. The mongo namespace alows you to create a Mongo instance server location, replica-sets, and options.

To use the Mongo namespace elements you will need to reference the Mongo schema:

<?xml version="1.0" encoding="UTF-8"?>
<beans xmlns="http://www.springframework.org/schema/beans"
          xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
          xmlns:context="http://www.springframework.org/schema/context"
          xmlns:mongo="http://www.springframework.org/schema/data/mongo"
          xsi:schemaLocation=
          "http://www.springframework.org/schema/context
          http://www.springframework.org/schema/context/spring-context-3.0.xsd
          http://www.springframework.org/schema/data/mongo
          http://www.springframework.org/schema/data/mongo/spring-mongo-1.0.xsd
          http://www.springframework.org/schema/beans
          http://www.springframework.org/schema/beans/spring-beans-3.0.xsd">

    <!-- Default bean name is 'mongo' -->
    <mongo:mongo host="localhost" port="27017"/>

</beans>
        

A more advanced configuration with MongoOptions is shown below (note these are not recommended values)

<beans>

  <mongo:mongo host="localhost" port="27017">
    <mongo:options connections-per-host="8"
                   threads-allowed-to-block-for-connection-multiplier="4"
                   connect-timeout="1000"
                   max-wait-time="1500}"
                   auto-connect-retry="true"
                   socket-keep-alive="true"
                   socket-timeout="1500"
                   slave-ok="true"
                   write-number="1"
                   write-timeout="0"
                   write-fsync="true"/>
  </mongo:mongo/>

</beans>
        

A configuration using replica sets is shown below.

<mongo:mongo id="replicaSetMongo" replica-set="127.0.0.1:27017,localhost:27018"/>	   

While com.mongodb.Mongo is the entry point to the MongoDB driver API, connecting to a specific MongoDB database instance requires additional information such as the database name and an optional username and password. With that information you can obtain a com.mongodb.DB object and access all the functionality of a specific MongoDB database instance. Spring provides the org.springframework.data.mongodb.core.MongoDbFactory interface shown below to bootstrap connectivity to the database.

public interface MongoDbFactory {

  DB getDb() throws DataAccessException;

  DB getDb(String dbName) throws DataAccessException;

}

The following sections show how you can use the contiainer with either Java or the XML based metadata to configure an instance of the MongoDbFactory interface. In turn, you can use the MongoDbFactory instance to configure MongoTemplate.

The class org.springframework.data.mongodb.core.SimpleMongoDbFactory provides implements the MongoDbFactory interface and is created with a standard com.mongodb.Mongo instance, the database name and an optional org.springframework.data.authentication.UserCredentials constructor argument.

Instead of using the IoC container to create an instance of MongoTemplate, you can just use them in standard Java code as shown below.

public class MongoApp {

  private static final Log log = LogFactory.getLog(MongoApp.class);

  public static void main(String[] args) throws Exception {

    MongoOperations mongoOps = new MongoTemplate(new SimpleMongoDbFactory(new Mongo(), "database"));

    mongoOps.insert(new Person("Joe", 34));

    log.info(mongoOps.findOne(new Query(where("name").is("Joe")), Person.class));

    mongoOps.dropCollection("person");
  }
}

The code in bold highlights the use of SimpleMongoDbFactory and is the only difference between the listing shown in the getting started section.

To register a MongoDbFactory instance with the container, you write code much like what was highlighted in the previous code listing. A simple example is shown below

@Configuration
public class MongoConfiguration {
  
  public @Bean MongoDbFactory mongoDbFactory() throws Exception {
    return new SimpleMongoDbFactory(new Mongo(), "database");
  }

}

To define the username and password create an instance of org.springframework.data.authentication.UserCredentials and pass it into the constructor as shown below. This listing also shows using MongoDbFactory register an instance of MongoTemplate with the container.

@Configuration
public class MongoConfiguration {
  
  public @Bean MongoDbFactory mongoDbFactory() throws Exception {
    UserCredentials userCredentials = new UserCredentials("joe", "secret");
    return new SimpleMongoDbFactory(new Mongo(), "database", userCredentials);
  }

  public @Bean MongoTemplate mongoTemplate() throws Exception {
    return new MongoTemplate(mongoDbFactory());
  }
}

The mongo namespace provides a convient way to create a SimpleMongoDbFactory as compared to using the<beans/> namespace. Simple usage is shown below

<mongo:db-factory dbname="database">

In the above example a com.mongodb.Mongo instance is created using the default host and port number. The SimpleMongoDbFactory registered with the container is identified by the id 'mongoDbFactory' unless a value for the id attribute is specified.

You can also provide the host and port for the underlying com.mongodb.Mongo instance as shown below, in addition to username and password for the database.

<mongo:db-factory id="anotherMongoDbFactory"
                  host="localhost"
                  port="27017"
                  dbname="database"
                  username="joe"
                  password="secret"/>

If you need to configure additional options on the com.mongodb.Mongo instance that is used to create a SimpleMongoDbFactory you can refer to an existing bean using the mongo-ref attribute as shown below. To show another common usage pattern, this listing show the use of a property placeholder to parameterise the configuration and creating MongoTemplate.

<context:property-placeholder location="classpath:/com/myapp/mongodb/config/mongo.properties"/>

<mongo:mongo host="${mongo.host}" port="${mongo.port}">
  <mongo:options
     connections-per-host="${mongo.connectionsPerHost}"
     threads-allowed-to-block-for-connection-multiplier="${mongo.threadsAllowedToBlockForConnectionMultiplier}"
     connect-timeout="${mongo.connectTimeout}"
     max-wait-time="${mongo.maxWaitTime}"
     auto-connect-retry="${mongo.autoConnectRetry}"
     socket-keep-alive="${mongo.socketKeepAlive}"
     socket-timeout="${mongo.socketTimeout}"
     slave-ok="${mongo.slaveOk}"
     write-number="1"
     write-timeout="0"
     write-fsync="true"/>
</mongo:mongo>

<mongo:db-factory dbname="database" mongo-ref="mongo"/>

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

You can use Java to create and register an instance of MongoTemplate as shown below.

@Configuration
public class AppConfig {

    public @Bean Mongo mongo() throws Exception {
        return new Mongo("localhost");
    }

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

There are several overloaded constructors of MongoTemplate. These are

You can also configure a MongoTemplate using Spring's XML <beans/> schema.

  <mongo:mongo host="localhost" port="27017"/>
  
  <bean id="mongoTemplate" class="org.springframework.data.mongodb.core.MongoTemplate">
    <constructor-arg ref="mongo"/>
    <constructor-arg name="databaseName" value="geospatial"/>
  </bean>

Other optional properties that you might like to set when creating a MongoTemplate are the default WriteResultCheckingPolicy, WriteConcern, and ReadPreference.

	Note
	The preferred way to reference the operations on MongoTemplate instance is via its interface MongoOperations.

public interface WriteConcernResolver {
  WriteConcern resolve(MongoAction action);
}

private class MyAppWriteConcernResolver implements WriteConcernResolver {

  public WriteConcern resolve(MongoAction action) {
    if (action.getEntityClass().getSimpleName().contains("Audit")) {
      return WriteConcern.NONE;
    } else if (action.getEntityClass().getSimpleName().contains("Metadata")) {
      return WriteConcern.JOURNAL_SAFE;
    }
    return action.getDefaultWriteConcern();
  }
}

MongoDB requires that you have an '_id' field for all documents. If you don't provide one the driver will assign a ObjectId with a generated value. When using the MongoMappingConverter there are certain rules that govern how properties from the Java class is mapped to this '_id' field.

The following outlines what property will be mapped to the '_id' document field:

The following outlines what type conversion, if any, will be done on the property mapped to the _id document field when using the MappingMongoConverter, the default for MongoTemplate.

If no field or property specified above is present in the Java class then an implicit '_id' file will be generated by the driver but not mapped to a property or field of the Java class.

When querying and updating MongoTemplate will use the converter to handle conversions of the Query and Update objects that correspond to the above rules for saving documents so field names and types used in your queries will be able to match what is in your domain classes.

As MongoDB collections can contain documents that represent instances of a variety of types. A great example here is if you store a hierarchy of classes or simply have a class with a property of type Object. In the latter case the values held inside that property have to be read in correctly when retrieving the object. Thus we need a mechanism to store type information alongside the actual document.

To achieve that the MappingMongoConverter uses a MongoTypeMapper abstraction with DefaultMongoTypeMapper as it's main implementation. It's default behaviour is storing the fully qualified classname under _class inside the document for the top-level document as well as for every value if it's a complex type and a subtype of the property type declared.

public class Sample {
  Contact value;
}

public abstract class Contact { … }

public class Person extends Contact { … }

Sample sample = new Sample();
sample.value = new Person();

mongoTemplate.save(sample);

{ "_class" : "com.acme.Sample",
  "value" : { "_class" : "com.acme.Person" }
}

As you can see we store the type information for the actual root class persistet as well as for the nested type as it is complex and a subtype of Contact. So if you're now using mongoTemplate.findAll(Object.class, "sample") we are able to find out that the document stored shall be a Sample instance. We are also able to find out that the value property shall be a Person actually.

There are several convenient methods on MongoTemplate for saving and inserting your objects. To have more fine grained control over the conversion process you can register Spring converters with the MappingMongoConverter, for example Converter<Person, DBObject> and Converter<DBObject, Person>.

	Note
	The difference between insert and save operations is that a save operation will perform an insert if the object is not already present.

The simple case of using the save operation is to save a POJO. In this case the collection name will be determined by name (not fully qualfied) of the class. You may also call the save operation with a specific collection name. The collection to store the object can be overriden using mapping metadata.

When inserting or saving, if the Id property is not set, the assumption is that its value will be autogenerated by the database. As such, for autogeneration of an ObjectId to succeed the type of the Id property/field in your class must be either a String, ObjectId, or BigInteger.

Here is a basic example of using the save operation and retrieving its contents.

import static org.springframework.data.mongodb.core.query.Criteria.where;
import static org.springframework.data.mongodb.core.query.Criteria.query;

…

Person p = new Person("Bob", 33);
mongoTemplate.insert(p);

Person qp = mongoTemplate.findOne(query(where("age").is(33)), Person.class);        

The insert/save operations available to you are listed below.

A similar set of insert operations is listed below

For updates we can elect to update the first document found using MongoOperation's method updateFirst or we can update all documents that were found to match the query using the method updateMulti. Here is an example of an update of all SAVINGS accounts where we are adding a one time $50.00 bonus to the balance using the $inc operator.

import static org.springframework.data.mongodb.core.query.Criteria.where;
import static org.springframework.data.mongodb.core.query.Query;
import static org.springframework.data.mongodb.core.query.Update;

          ...

  WriteResult wr = mongoTemplate.updateMulti(new Query(where("accounts.accountType").is(Account.Type.SAVINGS)),
                                                            new Update().inc("accounts.$.balance", 50.00),
                                                            Account.class);
        

In addition to the Query discussed above we provide the update definition using an Update object. The Update class has methods that match the update modifiers available for MongoDB.

As you can see most methods return the Update object to provide a fluent style for the API.

Related to perfomring an updateFirst operations, you can also perform an upsert operation which will perform an insert if no document is found that matches the query. The document that is inserted is a combination of the query document and the update document. Here is an example

template.upsert(query(where("ssn").is(1111).and("firstName").is("Joe").and("Fraizer").is("Update")), update("address", addr), Person.class);

The findAndModify(…) method on DBCollection can update a document and return either the old or newly updated document in a single operation. MongoTemplate provides a findAndModify method that takes Query and Update classes and converts from DBObject to your POJOs. Here are the methods

<T> T findAndModify(Query query, Update update, Class<T> entityClass);

<T> T findAndModify(Query query, Update update, Class<T> entityClass, String collectionName);

<T> T findAndModify(Query query, Update update, FindAndModifyOptions options, Class<T> entityClass);

<T> T findAndModify(Query query, Update update, FindAndModifyOptions options, Class<T> entityClass, String collectionName);

As an example usage, we will insert of few Person objects into the container and perform a simple findAndUpdate operation

mongoTemplate.insert(new Person("Tom", 21));
mongoTemplate.insert(new Person("Dick", 22));
mongoTemplate.insert(new Person("Harry", 23));

Query query = new Query(Criteria.where("firstName").is("Harry"));
Update update = new Update().inc("age", 1);
Person p = mongoTemplate.findAndModify(query, update, Person.class); // return's old person object

assertThat(p.getFirstName(), is("Harry"));
assertThat(p.getAge(), is(23));
p = mongoTemplate.findOne(query, Person.class);
assertThat(p.getAge(), is(24));


// Now return the newly updated document when updating
p = template.findAndModify(query, update, new FindAndModifyOptions().returnNew(true), Person.class);
assertThat(p.getAge(), is(25));

The FindAndModifyOptions lets you set the options of returnNew, upsert, and remove. An example extending off the previous code snippit is shown below

Query query2 = new Query(Criteria.where("firstName").is("Mary"));
p = mongoTemplate.findAndModify(query2, update, new FindAndModifyOptions().returnNew(true).upsert(true), Person.class);
assertThat(p.getFirstName(), is("Mary"));
assertThat(p.getAge(), is(1));

You can use several overloaded methods to remove an object from the database.

We saw how to retrieve a single document using the findOne and findById methods on MongoTemplate in previous sections which return a single domain object. We can also query for a collection of documents to be returned as a list of domain objects. Assuming that we have a number of Person objects with name and age stored as documents in a collection and that each person has an embedded account document with a balance. We can now run a query using the following code.

import static org.springframework.data.mongodb.core.query.Criteria.where;
import static org.springframework.data.mongodb.core.query.Query.query;

…

List<Person> result = mongoTemplate.find(query(where("age").lt(50)
                                                .and("accounts.balance").gt(1000.00d)), Person.class);        

All find methods take a Query object as a parameter. This object defines the criteria and options used to perform the query. The criteria is specified using a Criteria object that has a static factory method named where used to instantiate a new Criteria object. We recommend using a static import for org.springframework.data.mongodb.core.query.Criteria.where and Query.query to make the query more readable.

This query should return a list of Person objects that meet the specified criteria. The Criteria class has the following methods that correspond to the operators provided in MongoDB.

As you can see most methods return the Criteria object to provide a fluent style for the API.

There are also methods on the Criteria class for geospatial queries. Here is a listing but look at the section on GeoSpatial Queries to see them in action.

The Query class has some additional methods used to provide options for the query.

The query methods need to specify the target type T that will be returned and they are also overloaded with an explicit collection name for queries that should operate on a collection other than the one indicated by the return type.

MongoDB supports GeoSpatial queries through the use of operators such as $near, $within, and $nearSphere. Methods specific to geospatial queries are available on the Criteria class. There are also a few shape classes, Box, Circle, and Point that are used in conjunction with geospatial related Criteria methods.

To understand how to perform GeoSpatial queries we will use the following Venue class taken from the integration tests.which relies on using the rich MappingMongoConverter.

@Document(collection="newyork")
public class Venue {
    
  @Id
  private String id;
  private String name;
  private double[] location;
  
  @PersistenceConstructor
  Venue(String name, double[] location) {
    super();
    this.name = name;
    this.location = location;    
  }
  
  public Venue(String name, double x, double y) {
    super();
    this.name = name;
    this.location = new double[] { x, y };    
  }

  public String getName() {
    return name;
  }

  public double[] getLocation() {
    return location;
  }

  @Override
  public String toString() {
    return "Venue [id=" + id + ", name=" + name + ", location="
        + Arrays.toString(location) + "]";
  } 
}

To find locations within a Circle, the following query can be used.

Circle circle = new Circle(-73.99171, 40.738868, 0.01);
List<Venue> venues = 
    template.find(new Query(Criteria.where("location").withinCenter(circle)), Venue.class);

To find venues within a Circle using spherical coordinates the following query can be used

Circle circle = new Circle(-73.99171, 40.738868, 0.003712240453784);
List<Venue> venues = 
    template.find(new Query(Criteria.where("location").withinCenterSphere(circle)), Venue.class);

To find venues within a Box the following query can be used

//lower-left then upper-right
Box box = new Box(new Point(-73.99756, 40.73083), new Point(-73.988135, 40.741404));  
List<Venue> venues = 
    template.find(new Query(Criteria.where("location").withinBox(box)), Venue.class);

To find venues near a Point, the following query can be used

Point point = new Point(-73.99171, 40.738868);
List<Venue> venues = 
    template.find(new Query(Criteria.where("location").near(point).maxDistance(0.01)), Venue.class);

To find venues near a Point using spherical coordines the following query can be used

Point point = new Point(-73.99171, 40.738868);
List<Venue> venues = 
    template.find(new Query(
        Criteria.where("location").nearSphere(point).maxDistance(0.003712240453784)), 
        Venue.class);

Point location = new Point(-73.99171, 40.738868);
NearQuery query = NearQuery.near(location).maxDistance(new Distance(10, Metrics.MILES));

GeoResults<Restaurant> = operations.geoNear(query, Restaurant.class);

To understand how to perform Map-Reduce operations an example from the book 'MongoDB - The definitive guide' is used. In this example we will create three documents that have the values [a,b], [b,c], and [c,d] respectfully. The values in each document are associated with the key 'x' as shown below. For this example assume these documents are in the collection named "jmr1".

{ "_id" : ObjectId("4e5ff893c0277826074ec533"), "x" : [ "a", "b" ] }
{ "_id" : ObjectId("4e5ff893c0277826074ec534"), "x" : [ "b", "c" ] }
{ "_id" : ObjectId("4e5ff893c0277826074ec535"), "x" : [ "c", "d" ] }

A map function that will count the occurance of each letter in the array for each document is shown below

function () {
    for (var i = 0; i < this.x.length; i++) {
        emit(this.x[i], 1);
    }
}

The reduce function that will sum up the occurance of each letter across all the documents is shown below

function (key, values) {
    var sum = 0;
    for (var i = 0; i < values.length; i++)
        sum += values[i];
    return sum;
}

Executing this will result in a collection as shown below.

{ "_id" : "a", "value" : 1 }
{ "_id" : "b", "value" : 2 }
{ "_id" : "c", "value" : 2 }
{ "_id" : "d", "value" : 1 }

Assuming that the map and reduce functions are located in map.js and reduce.js and bundled in your jar so they are available on the classpath, you can execute a map-reduce operation and obtain the results as shown below

MapReduceResults<ValueObject> results = mongoOperations.mapReduce("jmr1", "classpath:map.js", "classpath:reduce.js", ValueObject.class);
for (ValueObject valueObject : results) {
  System.out.println(valueObject);
}

The output of the above code is

ValueObject [id=a, value=1.0]
ValueObject [id=b, value=2.0]
ValueObject [id=c, value=2.0]
ValueObject [id=d, value=1.0]

The MapReduceResults class implements Iterable and provides access to the raw output, as well as timing and count statistics. The ValueObject class is simply

public class ValueObject {

  private String id;
  private float value;

  public String getId() {
    return id;
  }

  public float getValue() {
    return value;
  }

  public void setValue(float value) {
    this.value = value;
  }

  @Override
  public String toString() {
    return "ValueObject [id=" + id + ", value=" + value + "]";
  }
}

By default the output type of INLINE is used so you don't have to specify an output collection. To specify additional map-reduce options use an overloaded method that takes an additional MapReduceOptions argument. The class MapReduceOptions has a fluent API so adding additional options can be done in a very compact syntax. Here an example that sets the output collection to "jmr1_out". Note that setting only the output collection assumes a default output type of REPLACE.

MapReduceResults<ValueObject> results = mongoOperations.mapReduce("jmr1", "classpath:map.js", "classpath:reduce.js", 
                                                                     new MapReduceOptions().outputCollection("jmr1_out"), ValueObject.class);

There is also a static import import static org.springframework.data.mongodb.core.mapreduce.MapReduceOptions.options; that can be used to make the syntax slightly more compact

MapReduceResults<ValueObject> results = mongoOperations.mapReduce("jmr1", "classpath:map.js", "classpath:reduce.js", 
                                                                     options().outputCollection("jmr1_out"), ValueObject.class);

You can also specify a query to reduce the set of data that will be used to feed into the map-reduce operation. This will remove the document that contains [a,b] from consideration for map-reduce operations.

Query query = new Query(where("x").ne(new String[] { "a", "b" }));
MapReduceResults<ValueObject> results = mongoOperations.mapReduce(query, "jmr1", "classpath:map.js", "classpath:reduce.js", 
                                                                     options().outputCollection("jmr1_out"), ValueObject.class);

Note that you can specify additional limit and sort values as well on the query but not skip values.

In order to understand how group operations work the following example is used, which is somewhat artifical. For a more realistic example consult the book 'MongoDB - The definitive guide'. A collection named "group_test_collection" created with the following rows.

{ "_id" : ObjectId("4ec1d25d41421e2015da64f1"), "x" : 1 }
{ "_id" : ObjectId("4ec1d25d41421e2015da64f2"), "x" : 1 }
{ "_id" : ObjectId("4ec1d25d41421e2015da64f3"), "x" : 2 }
{ "_id" : ObjectId("4ec1d25d41421e2015da64f4"), "x" : 3 }
{ "_id" : ObjectId("4ec1d25d41421e2015da64f5"), "x" : 3 }
{ "_id" : ObjectId("4ec1d25d41421e2015da64f6"), "x" : 3 }

We would like to group by the only field in each row, the 'x' field and aggregate the number of times each specific value of 'x' occurs. To do this we need to create an initial document that contains our count variable and also a reduce function which will increment it each time it is encountered. The Java code to execute the group operation is shown below

GroupByResults<XObject> results = mongoTemplate.group("group_test_collection", 
                                                      GroupBy.key("x").initialDocument("{ count: 0 }").reduceFunction("function(doc, prev) { prev.count += 1 }"), 
                                                      XObject.class);

The first argument is the name of the collection to run the group operation over, the second is a fluent API that specifies properties of the group operation via a GroupBy class. In this example we are using just the intialDocument and reduceFunction methods. You can also specify a key-function, as well as a finalizer as part of the fluent API. If you have multiple keys to group by, you can pass in a comma separated list of keys.

The raw results of the group operation is a JSON document that looks like this

{ 
  "retval" : [ { "x" : 1.0 , "count" : 2.0} , 
               { "x" : 2.0 , "count" : 1.0} , 
               { "x" : 3.0 , "count" : 3.0} ] , 
  "count" : 6.0 , 
  "keys" : 3 , 
  "ok" : 1.0
}

The document under the "retval" field is mapped onto the third argument in the group method, in this case XObject which is shown below.

public class XObject {

  private float x;

  private float count;


  public float getX() {
    return x;
  }

  public void setX(float x) {
    this.x = x;
  }

  public float getCount() {
    return count;
  }

  public void setCount(float count) {
    this.count = count;
  }

  @Override
  public String toString() {
    return "XObject [x=" + x + " count = " + count + "]";
  }
}

You can also obtain tha raw result as a DbObject by calling the method getRawResults on the GroupByResults class.

There is an additional method overload of the group method on MongoOperations which lets you specify a Criteria object for selecting a subset of the rows. An example which uses a Criteria object, with some syntax sugar using static imports, as well as referencing a key-function and reduce function javascript files via a Spring Resource string is shown below.

import static org.springframework.data.mongodb.core.mapreduce.GroupBy.keyFunction;
import static org.springframework.data.mongodb.core.query.Criteria.where;

GroupByResults<XObject> results = mongoTemplate.group(where("x").gt(0), 
                                        "group_test_collection", 
                                        keyFunction("classpath:keyFunction.js").initialDocument("{ count: 0 }").reduceFunction("classpath:groupReduce.js"), XObject.class);

An example implementation of the Converter that converts from a Person object to a com.mongodb.DBObject is shown below

import org.springframework.core.convert.converter.Converter;

import com.mongodb.BasicDBObject;
import com.mongodb.DBObject;

public class PersonWriteConverter implements Converter<Person, DBObject> {

  public DBObject convert(Person source) {
    DBObject dbo = new BasicDBObject();
    dbo.put("_id", source.getId());
    dbo.put("name", source.getFirstName());
    dbo.put("age", source.getAge());
    return dbo;
  }
}

An example implemention of a Converter that converts from a DBObject ot a Person object is shownn below

public class PersonReadConverter implements Converter<DBObject, Person> {

  public Person convert(DBObject source) {
    Person p = new Person((ObjectId) source.get("_id"), (String) source.get("name"));
    p.setAge((Integer) source.get("age"));
    return p;
  }
}

The Mongo Spring namespace provides a convenience way to register Spring Converters with the MappingMongoConverter. The configuration snippet below shows how to manually register converter beans as well as configuring the wrapping MappingMongoConverter into a MongoTemplate.

<mongo:db-factory dbname="database"/>

<mongo:mapping-converter>
  <mongo:custom-converters>
    <mongo:converter ref="readConverter"/>
    <mongo:converter>
      <bean class="org.springframework.data.mongodb.test.PersonWriteConverter"/>
    </mongo:converter>
  </mongo:custom-converters>
</mongo:mapping-converter>

<bean id="readConverter" class="org.springframework.data.mongodb.test.PersonReadConverter"/>

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

You can also use the base-package attribute of the custom-converters element to enable classpath scanning for all Converter and GenericConverter implementations below the given package.

<mongo:mapping-converter>
  <mongo:custom-converters base-package="com.acme.**.converters" />
</mongo:mapping-converter>

Generally we inspect the Converter implementations for the source and target types they convert from and to. Depending on whether one of those is a type MongoDB can handle natively we will register the converter instance as reading or writing one. Have a look at the following samples:

// Write converter as only the target type is one Mongo can handle natively
class MyConverter implements Converter<Person, String> { … }

// Read converter as only the source type is one Mongo can handle natively
class MyConverter implements Converter<String, Person> { … }

In case you write a Converter whose source and target type are native Mongo types there's no way for us to determine whether we should consider it as reading or writing converter. Registering the converter instance as both might lead to unwanted results then. E.g. a Converter<String, Long> is ambiguous although it probably does not make sense to try to convert all Strings into Longs when writing. To be generally able to force the infrastructure to register a converter for one way only we provide @ReadingConverter as well as @WritingConverter to be used at the converter implementation.

We can create an index on a collection to improve query performance.

mongoTemplate.indexOps(Person.class).ensureIndex(new Index().on("name",Order.ASCENDING));        

You can create both standard indexes and geospatial indexes using the classes IndexDefinition and GeoSpatialIndex respectfully. For example, given the Venue class defined in a previous section, you would declare a geospatial query as shown below

mongoTemplate.indexOps(Venue.class).ensureIndex(new GeospatialIndex("location"));

The IndexOperations interface has the method getIndexInfo that returns a list of IndexInfo objects. This contains all the indexes defined on the collectcion. Here is an example that defines an index on the Person class that has age property.

template.indexOps(Person.class).ensureIndex(new Index().on("age", Order.DESCENDING).unique(Duplicates.DROP));

List<IndexInfo> indexInfoList = template.indexOps(Person.class).getIndexInfo();

// Contains
// [IndexInfo [fieldSpec={_id=ASCENDING}, name=_id_, unique=false, dropDuplicates=false, sparse=false], 
//  IndexInfo [fieldSpec={age=DESCENDING}, name=age_-1, unique=true, dropDuplicates=true, sparse=false]]

It's time to look at some code examples showing how to use the MongoTemplate. First we look at creating our first collection.

DBCollection collection = null;
if (!mongoTemplate.getCollectionNames().contains("MyNewCollection")) {
    collection = mongoTemplate.createCollection("MyNewCollection");
}

mongoTemplate.dropCollection("MyNewCollection");        

As you've just seen there are a few keywords triggering geo-spatial operations within a MongoDB query. The Near keyword allows some further modification. Let's have look at some examples:

public interface PersonRepository extends MongoRepository<Person, String>

  // { 'location' : { '$near' : [point.x, point.y], '$maxDistance' : distance}}
  List<Person> findByLocationNear(Point location, Distance distance);
}

Adding a Distance parameter to the query method allows restricting results to those within the given distance. If the Distance was set up containing a Metric we will transparently use $nearSphere instead of $code.

Point point = new Point(43.7, 48.8);
Distance distance = new Distance(200, Metrics.KILOMETERS);
… = repository.findByLocationNear(point, distance);
// {'location' : {'$nearSphere' : [43.7, 48.8], '$maxDistance' : 0.03135711885774796}}

As you can see using a Distance equipped with a Metric causes $nearSphere clause to be added instead of a plain $near. Beyond that the actual distance gets calculated according to the Metrics used.

public interface PersonRepository extends MongoRepository<Person, String>

  // {'geoNear' : 'location', 'near' : [x, y] }
  GeoResults<Person> findByLocationNear(Point location);

  // No metric: {'geoNear' : 'person', 'near' : [x, y], maxDistance : distance }
  // Metric: {'geoNear' : 'person', 'near' : [x, y], 'maxDistance' : distance, 
  //          'distanceMultiplier' : metric.multiplier, 'spherical' : true }
  GeoResults<Person> findByLocationNear(Point location, Distance distance);

  // {'geoNear' : 'location', 'near' : [x, y] }
  GeoResults<Person> findByLocationNear(Point location);
}

By adding the annotation org.springframework.data.mongodb.repository.Query repository finder methods you can specify a MongoDB JSON query string to use instead of having the query derived from the method name. For example

public interface PersonRepository extends MongoRepository<Person, String>

  @Query("{ 'firstname' : ?0 }")
  List<Person> findByThePersonsFirstname(String firstname);

}

The placeholder ?0 lets you substitute the value from the method arguments into the JSON query string.

You can also use the filter property to restrict the set of properties that will be mapped into the Java object. For example,

public interface PersonRepository extends MongoRepository<Person, String>

  @Query(value="{ 'firstname' : ?0 }", fields="{ 'firstname' : 1, 'lastname' : 1}")
  List<Person> findByThePersonsFirstname(String firstname);

}

This will return only the firstname, lastname and Id properties of the Person objects. The age property, a java.lang.Integer, will not be set and its value will therefore be null.

MongoDB repository support integrates with the QueryDSL project which provides a means to perform type-safe queries in Java. To quote from the project description, "Instead of writing queries as inline strings or externalizing them into XML files they are constructed via a fluent API." It provides the following features

Please refer to the QueryDSL documentation which describes how to bootstrap your environment for APT based code generation using Maven or using Ant.

Using QueryDSL you will be able to write queries as shown below

QPerson person = new QPerson("person");
List<Person> result = repository.findAll(person.address.zipCode.eq("C0123"));

Page<Person> page = repository.findAll(person.lastname.contains("a"), 
                                       new PageRequest(0, 2, Direction.ASC, "lastname"));

QPerson is a class that is generated (via the Java annotation post processing tool) which is a Predicate that allows you to write type safe queries. Notice that there are no strings in the query other than the value "C0123".

You can use the generated Predicate class via the interface QueryDslPredicateExecutor which is shown below

public interface QueryDslPredicateExecutor<T> {

  T findOne(Predicate predicate);

  List<T> findAll(Predicate predicate);

  List<T> findAll(Predicate predicate, OrderSpecifier<?>... orders);

  Page<T> findAll(Predicate predicate, Pageable pageable);

  Long count(Predicate predicate);
}

To use this in your repository implementation, simply inherit from it in additiion to other repository interfaces. This is shown below

public interface PersonRepository extends MongoRepository<Person, String>, QueryDslPredicateExecutor<Person> {

   // additional finder methods go here

}

We think you will find this an extremely powerful tool for writing MongoDB queries.

MongoDB requires that you have an '_id' field for all documents. If you don't provide one the driver will assign a ObjectId with a generated value. The "_id" field can be of any type the, other than arrays, so long as it is unique. The driver naturally supports all primitive types and Dates. When using the MongoMappingConverter there are certain rules that govern how properties from the Java class is mapped to this '_id' field.

The following outlines what field will be mapped to the '_id' document field:

The following outlines what type conversion, if any, will be done on the property mapped to the _id document field.

When querying and updating MongoTemplate will use the converter to handle conversions of the Query and Update objects that correspond to the above rules for saving documents so field names and types used in your queries will be able to match what is in your domain classes.

The MappingMongoConverter can use metadata to drive the mapping of objects to documents. An overview of the annotations is provided below

The mapping metadata infrastructure is defined in a seperate spring-data-commons project that is technology agnostic. Specific subclasses are using in the MongoDB support to support annotation based metadata. Other strategies are also possible to put in place if there is demand.

Here is an example of a more complex mapping.

@Document
@CompoundIndexes({
    @CompoundIndex(name = "age_idx", def = "{'lastName': 1, 'age': -1}")
})
public class Person<T extends Address> {

  @Id
  private String id;

  @Indexed(unique = true)
  private Integer ssn;

  @Field("fName")
  private String firstName;

  @Indexed
  private String lastName;

  private Integer age;

  @Transient
  private Integer accountTotal;

  @DBRef
  private List<Account> accounts;

  private T address;

  
  public Person(Integer ssn) {
    this.ssn = ssn;
  }
  
  @PersistenceConstructor
  public Person(Integer ssn, String firstName, String lastName, Integer age, T address) {
    this.ssn = ssn;
    this.firstName = firstName;
    this.lastName = lastName;
    this.age = age;
    this.address = address;
  }

  public String getId() {
    return id;
  }

 // no setter for Id.  (getter is only exposed for some unit testing)

  public Integer getSsn() {
    return ssn;
  }


// other getters/setters ommitted

Compound indexes are also supported. They are defined at the class level, rather than on indidvidual properties.

	Note
	Compound indexes are very important to improve the performance of queries that involve criteria on multiple fields

Here's an example that creates a compound index of lastName in ascending order and age in descending order:

package com.mycompany.domain;

@Document
@CompoundIndexes({
    @CompoundIndex(name = "age_idx", def = "{'lastName': 1, 'age': -1}")
})
public class Person {

  @Id
  private ObjectId id;
  private Integer age;
  private String firstName;
  private String lastName;

}
          

The mapping framework doesn't have to store child objects embedded within the document. You can also store them separately and use a DBRef to refer to that document. When the object is loaded from MongoDB, those references will be eagerly resolved and you will get back a mapped object that looks the same as if it had been stored embedded within your master document.

Here's an example of using a DBRef to refer to a specific document that exists independently of the object in which it is referenced (both classes are shown in-line for brevity's sake):

@Document
public class Account {

  @Id
  private ObjectId id;
  private Float total;

}

@Document
public class Person {

  @Id
  private ObjectId id;
  @Indexed
  private Integer ssn;
  @DBRef
  private List<Account> accounts;

}
        

There's no need to use something like @OneToMany because the mapping framework sees that you're wanting a one-to-many relationship because there is a List of objects. When the object is stored in MongoDB, there will be a list of DBRefs rather than the Account objects themselves.

	Important
	The mapping framework does not handle cascading saves. If you change an Account object that is referenced by a Person object, you must save the Account object separately. Calling save on the Person object will not automatically save the Account objects in the property accounts.

Events are fired throughout the lifecycle of the mapping process. This is described in the Lifecycle Events section.

Simply declaring these beans in your Spring ApplicationContext will cause them to be invoked whenever the event is dispatched.

When storing and querying your objects it is convenient to have a MongoConverter instance handle the mapping of all Java types to DBObjects. However, sometimes you may want the MongoConverter's do most of the work but allow you to selectivly handle the conversion for a particular type or to optimize performance.

To selectivly handle the conversion yourself, register one or more one or more org.springframework.core.convert.converter.Converter instances with the MongoConverter.

	Note
	Spring 3.0 introduced a core.convert package that provides a general type conversion system. This is described in detail in the Spring reference documentation section entitled Spring 3 Type Conversion.

The setConverters method on SimpleMongoConverter and MappingMongoConverter should be used for this purpose. The method afterMappingMongoConverterCreation in AbstractMongoConfiguration can be overriden to configure a MappingMongoConverter. The examples here at the begining of this chapter show how to perform the configuration using Java and XML.

Below is an example of a Spring Converter implementation that converts from a DBObject to a Person POJO.

public class PersonReadConverter implements Converter<DBObject, Person> {

  public Person convert(DBObject source) {
    Person p = new Person((ObjectId) source.get("_id"), (String) source.get("name"));
    p.setAge((Integer) source.get("age"));
    return p;
  }

}

Here is an example that converts from a Person to a DBObject.

public class PersonWriteConverter implements Converter<Person, DBObject> {

  public DBObject convert(Person source) {
    DBObject dbo = new BasicDBObject();
    dbo.put("_id", source.getId());
    dbo.put("name", source.getFirstName());
    dbo.put("age", source.getAge());
    return dbo;
  }

}