Skip to content

moderakh/azure-documentdb-java

 
 

Repository files navigation

#Microsoft Azure DocumentDB Java SDK

This project provides a client library in Java that makes it easy to interact with Azure DocumentDB. For documentation please see the Microsoft Azure Java Developer Center and the JavaDocs.

##Download ###Option 1: Via Maven

To get the binaries of this library as distributed by Microsoft, ready for use within your project, you can use Maven.

<dependency>
	<groupId>com.microsoft.azure</groupId>
	<artifactId>azure-documentdb</artifactId>
	<version>1.8.1</version>
</dependency>

###Option 2: Source Via Git

To get the source code of the SDK via git just type:

git clone git://github.com/Azure/azure-documentdb-java.git

###Option 3: Source Zip

To download a copy of the source code, click "Download ZIP" on the right side of the page or click here.

##Minimum Requirements

  • Java Development Kit 7
  • (Optional) Maven

Dependencies

  • Apache Commons Lang 3.3.2 (org.apache.commons / commons-lang3 / 3.3.2)
  • Apache HttpClient 4.2.5 (org.apache.httpcomponents / httpclient / 4.2.5)
  • Apache HttpCore 4.2.5 (org.apache.httpcomponents / httpcore / 4.2.5)
  • Jackson Data Mapper 1.8 (org.codehaus.jackson / jackson-mapper-asl / 1.8.5)
  • JSON 20140107 (org.json / json / 20140107)
  • JUnit 4.12 (junit / junit / 4.12)
  • mockito 1.10.19 (org.mockito / mockito-core / 1.10.19)

Dependencies will be added automatically if Maven is used. Otherwise, please download the jars and add them to your build path.

##Usage

To use this SDK to call Azure DocumentDB, you need to first create an account.

You can follow this tutorial to help you get started.

import java.io.IOException;
import java.util.List;

import com.google.gson.Gson;
import com.microsoft.azure.documentdb.ConnectionPolicy;
import com.microsoft.azure.documentdb.ConsistencyLevel;
import com.microsoft.azure.documentdb.Database;
import com.microsoft.azure.documentdb.Document;
import com.microsoft.azure.documentdb.DocumentClient;
import com.microsoft.azure.documentdb.DocumentClientException;
import com.microsoft.azure.documentdb.DocumentCollection;
import com.microsoft.azure.documentdb.RequestOptions;

public class HelloWorld {
    // Replace with your DocumentDB end point and master key.
    private static final String END_POINT = "[YOUR_ENDPOINT_HERE]";
    private static final String MASTER_KEY = "[YOUR_KEY_HERE]";
    
    // Define an id for your database and collection
    private static final String DATABASE_ID = "TestDB";
    private static final String COLLECTION_ID = "TestCollection";

    // We'll use Gson for POJO <=> JSON serialization for this sample.
    // Codehaus' Jackson is another great POJO <=> JSON serializer.
    private static Gson gson = new Gson();

    public static void main(String[] args) throws DocumentClientException,
            IOException {
        // Instantiate a DocumentClient w/ your DocumentDB Endpoint and AuthKey.
        DocumentClient documentClient = new DocumentClient(END_POINT,
                MASTER_KEY, ConnectionPolicy.GetDefault(),
                ConsistencyLevel.Session);

        // Start from a clean state (delete database in case it already exists).
        documentClient.deleteDatabase("dbs/" + DATABASE_ID, null);
        
        // Define a new database using the id above.
        Database myDatabase = new Database();
        myDatabase.setId(DATABASE_ID);

        // Create a new database.
        myDatabase = documentClient.createDatabase(myDatabase, null)
                .getResource();

        System.out.println("Created a new database:");
        System.out.println(myDatabase.toString());
        System.in.read();

        // Define a new collection using the id above.
        DocumentCollection myCollection = new DocumentCollection();
        myCollection.setId(COLLECTION_ID);

        // Set the provisioned throughput for this collection to be 1000 RUs.
        RequestOptions requestOptions = new RequestOptions();
        requestOptions.setOfferThroughput(1000);

        // Create a new collection.
        myCollection = documentClient.createCollection(
                myDatabase.getSelfLink(), myCollection, requestOptions)
                .getResource();

        System.out.println("Created a new collection:");
        System.out.println(myCollection.toString());
        System.in.read();

        // Create an object, serialize it in to JSON, and wrap it in to a
        // document.
        SomePojo andrewPojo = new SomePojo("123", "Andrew Liu", "[email protected]");
        String andrewJson = gson.toJson(andrewPojo);
        Document andrewDocument = new Document(andrewJson);

        // Create the 1st document.
        andrewDocument = documentClient.createDocument(
                myCollection.getSelfLink(), andrewDocument, null, false)
                .getResource();

        System.out.println("Created 1st document:");
        System.out.println(andrewDocument.toString());
        System.in.read();

        // Create another object, serialize it in to JSON, and wrap it in to a
        // document.
        SomePojo mimiPojo = new SomePojo("456", "Mimi Gentz",
                "[email protected]");
        String somePojoJson = gson.toJson(mimiPojo);
        Document mimiDocument = new Document(somePojoJson);

        // Create the 2nd document.
        mimiDocument = documentClient.createDocument(
                myCollection.getSelfLink(), mimiDocument, null, false)
                .getResource();

        System.out.println("Created 2nd document:");
        System.out.println(mimiDocument.toString());
        System.in.read();

        // Query documents
        List<Document> results = documentClient
                .queryDocuments(
                        myCollection.getSelfLink(),
                        "SELECT * FROM myCollection WHERE myCollection.email = '[email protected]'",
                        null).getQueryIterable().toList();

        System.out.println("Query document where e-mail address = '[email protected]':");
        System.out.println(results.toString());
        System.in.read();

        // Replace Document Andrew with Shireesh
        andrewPojo = gson.fromJson(results.get(0).toString(), SomePojo.class);
        andrewPojo.setName("Shireesh Thota");
        andrewPojo.setEmail("[email protected]");

        andrewDocument = documentClient.replaceDocument(
                andrewDocument.getSelfLink(),
                new Document(gson.toJson(andrewPojo)), null)
                .getResource();

        System.out.println("Replaced Andrew's document with Shireesh's contact information");
        System.out.println(andrewDocument.toString());
        System.in.read();

        // Delete Shireesh's Document

        documentClient.deleteDocument(andrewDocument.getSelfLink(), null);

        System.out.println("Deleted Shireesh's document");
        System.in.read();

        // Delete Database
        documentClient.deleteDatabase(myDatabase.getSelfLink(), null);

        System.out.println("Deleted database");
        System.in.read();

    }
}

The sample code above depends on a sample Plain Old Java Object (POJO) class, which is defined as follows:

class SomePojo {
    private String id;
    private String name;
    private String email;

    public SomePojo(String id, String name, String email) {
          super();
          this.id = id;
          this.name = name;
          this.email = email;
    }

    public String getEmail() {
          return email;
    }

    public String getId() {
          return id;
    }

    public String getName() {
          return name;
    }

    public void setEmail(String email) {
          this.email = email;
    }

    public void setId(String id) {
          this.id = id;
    }

    public void setName(String name) {
          this.name = name;
    }
}

The following code Illustrates how to create a partitioned collection and use the partition key to access documents:

// Create a partition key definition that specifies the path to the property
// within a document that is used as the partition key.          
PartitionKeyDefinition partitionKeyDef = new PartitionKeyDefinition();
ArrayList<String> paths = new ArrayList<String>();
paths.add("/id");
partitionKeyDef.setPaths(paths);

// Create a collection with the partition key definition and set the offer throughput
// to 10100 RU per second.
DocumentCollection myPartitionedCollection = new DocumentCollection();
myPartitionedCollection.setId(COLLECTION_ID_PARTITIONED);
myPartitionedCollection.setPartitionKey(partitionKeyDef);
                      
RequestOptions options = new RequestOptions();
options.setOfferThroughput(10100);
myPartitionedCollection = documentClient.createCollection(
    myDatabase.getSelfLink(), myCollection, options).getResource();

// Insert a document into the created collection.
String document = "{ 'id': 'document1', 'description': 'this is a test document.' }";
Document newDocument = new Document(document);
newDocument = documentClient.createDocument(myPartitionedCollection.getSelfLink(),
        newDocument, null, false).getResource();
        
 // Read the created document, specifying the required partition key in RequestOptions.
options = new RequestOptions();
options.setPartitionKey(new PartitionKey("document1"));
newDocument = documentClient.readDocument(newDocument.getSelfLink(), options).getResource();

Additional samples are provided in the unit tests.

##Need Help?

Be sure to check out the Microsoft Azure Developer Forums on MSDN or the Developer Forums on Stack Overflow if you have trouble with the provided code.

##Contribute Code or Provide Feedback

If you would like to become an active contributor to this project please follow the instructions provided in Azure Projects Contribution Guidelines.

If you encounter any bugs with the library please file an issue in the Issues section of the project.

##Learn More

About

No description, website, or topics provided.

Resources

License

Code of conduct

Security policy

Stars

Watchers

Forks

Packages

No packages published

Languages

  • Java 100.0%