Quantcast
Channel: Baeldung
Viewing all articles
Browse latest Browse all 4535

Introduction to Morphia – Java ODM for MongoDB

$
0
0

1. Overview

In this tutorial, we’ll understand how to use Morphia, an Object Document Mapper (ODM) for MongoDB in Java.

In the process, we’ll also understand what is an ODM and how it facilitates working with MongoDB.

2. What is an ODM?

For those uninitiated in this area, MongoDB is a document-oriented database built to be distributed by nature. Document-oriented databases, in simple terms, manage documents, which are nothing but a schema-less way of organizing semi-structured data. They fall under a broader and loosely defined umbrella of NoSQL databases, named after their apparent departure from traditional organization of SQL databases.

MongoDB provides drivers for almost all popular programming languages like Java. These drivers offer a layer of abstraction for working with MongoDB so that we aren’t working with Wire Protocol directly. Think of this as Oracle providing an implementation of JDBC driver for their relational database.

However, if we recall our days working with JDBC directly, we can appreciate how messy it can get — especially in an object-oriented paradigm. Fortunately, we have Object Relational Mapping (ORM) frameworks like Hibernate to our rescue. It isn’t very different for MongoDB.

While we can certainly work with the low-level driver, it requires a lot more boilerplate to accomplish the task. Here, we’ve got a similar concept to ORM called Object Document Mapper (ODM). Morphia exactly fills that space for the Java programming language and works on top of the Java driver for MongoDB.

3. Setting up Dependencies

We’ve seen enough theory to get us into some code. For our examples, we’ll model a library of books and see how we can manage it in MongoDB using Morphia.

But before we begin, we’ll need to set-up some of the dependencies.

3.1. MongoDB

We need to have a running instance of MongoDB to work with. There are several ways to get this, and the simplest is to download and install community edition on our local machine.

We should leave all default configurations as-is, including the port on which MongoDB runs.

3.2. Morphia

We can download the pre-built JARs for Morphia from Maven Central and use them in our Java project.

However, the simplest way is to use a dependency management tool like Maven:

<dependency>
    <groupId>dev.morphia.morphia</groupId>
    <artifactId>core</artifactId>
    <version>1.5.3</version>
</dependency>

4. How to Connect using Morphia?

Now that we have MongoDB installed and running and have set-up Morphia in our Java project, we’re ready to connect to MongoDB using Morphia.

Let’s see how we can accomplish that:

Morphia morphia = new Morphia();
morphia.mapPackage("com.baeldung.morphia");
Datastore datastore = morphia.createDatastore(new MongoClient(), "library");
datastore.ensureIndexes();

That’s pretty much it! Let’s understand this better. We need two things for our mapping operations to work:

  1. A Mapper: This is responsible for mapping our Java POJOs to MongoDB Collections. In our code snippet above, Morphia is the class responsible for that. Note how we’re configuring the package where it should look for our POJOs.
  2. A Connection: This is the connection to a MongoDB database on which the mapper can execute different operations. The class Datastore takes as a parameter an instance of MongoClient (from the Java MongoDB driver) and the name of the MongoDB database, returning an active connection to work with.

So, we’re all set to use this Datastore and work with our entities.

5. How to Work with Entities?

Before we can use our freshly minted Datastore, we need to define some domain entities to work with.

5.1. Simple Entity

Let’s begin by defining a simple Book entity with some attributes:

@Entity("Books")
public class Book {
    @Id
    private String isbn;
    private String title;
    private String author;
    @Property("price")
    private double cost;
    // constructors, getters, setters and hashCode, equals, toString implementations
}

There are a couple of interesting things to note here:

  • Notice the annotation @Entity that qualifies this POJO for ODM mapping by Morphia
  • Morphia, by default, maps an entity to a collection in MongoDB by the name of its class, but we can explicitly override this (like we’ve done for the entity Book here)
  • Morphia, by default, maps the variables in an entity to the keys in a MongoDB collection by the name of the variable, but again we can override this (like we’ve done for the variable cost here)
  • Lastly, we need to mark a variable in the entity to act as the primary key by the annotation @Id (like we’re using ISBN for our book here)

5.2. Entities with Relationships

In the real world, though, entities are hardly as simple as they look and have complex relationships with each other. For instance, our simple entity Book can have a Publisher and can reference other companion books. How do we model them?

MongoDB offers two mechanisms to build relationships — Referencing and Embedding. As the name suggests, with referencing, MongoDB stores related data as a separate document in the same or a different collection and just references it using its id.

On the contrary, with embedding, MongoDB stores or rather embeds the relation within the parent document itself.

Let’s see how we can use them. Let’s begin by embedding Publisher in our Book:

@Embedded
private Publisher publisher;

Simple enough. Now let’s go ahead and add references to other books:

@Reference
private List<Book> companionBooks;

That’s it — Morphia provides convenient annotations to model relationships as supported by MongoDB. The choice of referencing vs embedding, however, should draw from data model complexity, redundancy, and consistency amongst other considerations.

The exercise is similar to normalization in relational databases.

Now, we’re ready to perform some operations on Book using Datastore.

6. Some Basic Operations

Let’s see how to work with some of the basic operations using Morphia.

6.1. Save

Let’s begin with the simplest of the operations, creating an instance of Book in our MongoDB database library:

Publisher publisher = new Publisher(new ObjectId(), "Awsome Publisher");

Book book = new Book("9781565927186", "Learning Java", "Tom Kirkman", 3.95, publisher);
Book companionBook = new Book("9789332575103", "Java Performance Companion", 
  "Tom Kirkman", 1.95, publisher);

book.addCompanionBooks(companionBook);

datastore.save(companionBook);
datastore.save(book);

This is enough to let Morphia create a collection in our MongoDB database, if it does not exist, and perform an upsert operation.

6.2. Query

Let’s see if we’re able to query the book we just created in MongoDB:

List<Book> books = datastore.createQuery(Book.class)
  .field("title")
  .contains("Learning Java")
  .find()
  .toList();

assertEquals(1, books.size());

assertEquals(book, books.get(0));

Querying a document in Morphia begins with creating a query using Datastore and then declaratively adding filters, to the delight of those in love with functional programming!

Morphia supports much more complex query construction with filters and operators. Moreover, Morphia allows for limiting, skipping, and ordering of results in the query.

What’s more, Morphia allows us to use raw queries written with the Java driver for MongoDB for more control, should that be needed.

6.3. Update

Although a save operation can handle updates if the primary key matches, Morphia provides ways to selectively update documents:

Query<Book> query = datastore.createQuery(Book.class)
  .field("title")
  .contains("Learning Java");

UpdateOperations<Book> updates = datastore.createUpdateOperations(Book.class)
  .inc("price", 1);

datastore.update(query, updates);

List<Book> books = datastore.createQuery(Book.class)
  .field("title")
  .contains("Learning Java")
  .find()
  .toList();

assertEquals(4.95, books.get(0).getCost());

Here, we’re building a query and an update operation to increase by one the price of all books returned by the query.

6.4. Delete

Finally, that which has been created must be deleted! Again, with Morphia, it’s quite intuitive:

Query<Book> query = datastore.createQuery(Book.class)
  .field("title")
  .contains("Learning Java");

datastore.delete(query);

List<Book> books = datastore.createQuery(Book.class)
  .field("title")
  .contains("Learning Java")
  .find()
  .toList();

assertEquals(0, books.size());

We create the query quite similarly as before and run the delete operation on the Datastore.

7. Advanced Usage

MongoDB has some advanced operations like Aggregation, Indexing, and many others. While it isn’t possible to perform all of that using Morphia, it’s certainly possible to achieve some of that. For others, sadly, we’ll have to fall back to the Java driver for MongoDB.

Let’s focus on some of these advanced operations that we can perform through Morphia.

7.1. Aggregation

Aggregation in MongoDB allows us to define a series of operations in a pipeline that can operate on a set of documents and produce aggregated output.

Morphia has an API to support such an aggregation pipeline.

Let’s assume we wish to aggregate our library data in such a manner that we have all the books grouped by their author:

Iterator<Author> iterator = datastore.createAggregation(Book.class)
  .group("author", grouping("books", push("title")))
  .out(Author.class);

So, how does this work? We begin by creating an aggregation pipeline using the same old Datastore. We have to provide the entity on which we wish to perform aggregation operations, for instance, Book here.

Next, we want to group documents by “author” and aggregate their “title” under a key called “books”. Finally, we’re working with an ODM here. So, we have to define an entity to collect our aggregated data — in our case, it’s Author.

Of course, we have to define an entity called Author with a variable called books:

@Entity
public class Author {
    @Id
    private String name;
    private List<String> books;
    // other necessary getters and setters
}

This, of course, just scratches the surface of a very powerful construct provided by MongoDB and can be explored further for details.

7.2. Projection

Projection in MongoDB allows us to select only the fields we want to fetch from documents in our queries. In case document structure is complex and heavy, this can be really useful when we need only a few fields.

Let’s suppose we only need to fetch books with their title in our query:

List<Book> books = datastore.createQuery(Book.class)
  .field("title")
  .contains("Learning Java")
  .project("title", true)
  .find()
  .toList();
 
assertEquals("Learning Java", books.get(0).getTitle());
assertNull(books.get(0).getAuthor());

Here, as we can see, we only get back the title in our result and not the author and other fields. We should, however, be careful in using the projected output in saving back to MongoDB. This may result in data loss!

7.3. Indexing

Indexes play a very important role in query optimization with databases — relational as well as many non-relational ones.

MongoDB defines indexes at the level of the collection with a unique index created on the primary key by default. Moreover, MongoDB allows indexes to be created on any field or sub-field within a document. We should choose to create an index on a key depending on the query we wish to create.

For instance, in our example, we may wish to create an index on the field “title” of Book as we often end up querying on it:

@Indexes({
  @Index(
    fields = @Field("title"),
    options = @IndexOptions(name = "book_title")
  )
})
public class Book {
    // ...
    @Property
    private String title;
    // ...
}

Of course, we can pass additional indexing options to tailor the nuances of the index that gets created. Note that the field should be annotated by @Property to be used in an index.

Moreover, apart from the class-level index, Morphia has an annotation to define a field-level index as well.

7.4. Schema Validation

We’ve got an option to provide data validation rules for a collection that MongoDB can use while performing an update or insert operation. Morphia supports this through their APIs.

Let’s say that we don’t want to insert a book without a valid price. We can leverage schema validation to achieve this:

@Validation("{ price : { $gt : 0 } }")
public class Book {
    // ...
    @Property("price")
    private double cost;
    // ...
}

There is a rich set of validations provided by MongoDB that can be employed here.

8. Alternative MongoDB ODMs

Morphia is not the only available MongoDB ODM for Java. There are several others that we can consider to use in our applications. A discussion on comparison with Morphia is not possible here, but it is always useful to know our options:

  • Spring Data: Provides a Spring-based programming model for working with MongoDB
  • MongoJack: Provides direct mapping from JSON to MongoDB objects

This is not a complete list of MongoDB ODMs for Java, but there are some interesting alternates available!

9. Conclusion

In this article, we understood the basic details of MongoDB and the use of an ODM to connect and operate on MongoDB from a programing language like Java. We further explored Morphia as a MongoDB ODM for Java and the various capabilities it has.

As always, the code can be found over on GitHub.


Viewing all articles
Browse latest Browse all 4535

Trending Articles



<script src="https://jsc.adskeeper.com/r/s/rssing.com.1596347.js" async> </script>