In Spring data – MongoDB, you can use following methods to update documents.

  1. save – Update the whole object, if “_id” is present, perform an update, else insert it.
  2. updateFirst – Updates the first document that matches the query.
  3. updateMulti – Updates all documents that match the query.
  4. Upserting – If no document that matches the query, a new document is created by combining the query and update object.
  5. findAndModify – Same with updateMulti, but it has an extra option to return either the old or newly updated document.

P.S All examples are tested under mongo-java-driver-2.11.0.jar and spring-data-mongodb-1.2.0.RELEASE.jar

1. saveOrUpdate – part 1 example

Assume below json data is inserted into MongoDB.

{ 
	"_id" : ObjectId("id"), 
	"ic" : "1001", 
	"name" : "appleA", 
	"age" : 20, 
	"createdDate" : ISODate("2013-04-06T23:17:35.530Z") 
}

Find the document, modify and update it with save() method.

	Query query = new Query();
	query.addCriteria(Criteria.where("name").is("appleA"));
 
	User userTest1 = mongoOperation.findOne(query, User.class);
 
	System.out.println("userTest1 - " + userTest1);
 
	//modify and update with save()
	userTest1.setAge(99);
	mongoOperation.save(userTest1);
 
	//get the updated object again
	User userTest1_1 = mongoOperation.findOne(query, User.class);
 
	System.out.println("userTest1_1 - " + userTest1_1);

Output

userTest1 - User [id=id, ic=1001, name=appleA, age=20, createdDate=Sat Apr 06 23:17:35 MYT 2013]
userTest1_1 - User [id=id, ic=1001, name=appleA, age=99, createdDate=Sat Apr 06 23:17:35 MYT 2013]
Note
See example 2, it shows a common mistake made by most of the developers.

2. saveOrUpdate - part 2 example

This is a failed example, read carefully, a really common mistake.

Assume below json data is inserted into MongoDB.

{ 
	"_id" : ObjectId("id"), 
	"ic" : "1002", 
	"name" : "appleB", 
	"age" : 20, 
	"createdDate" : ISODate("2013-04-06T15:22:34.530Z") 
}

In Query, you get the document returned with a single "name" field value only, it did happened often to save the object returned size. The returned "User" object has null value in the fields : age, ic and createdDate, if you modify the 'age' field and update it, it will override everything instead of update the modified field - 'age'.

		Query query = new Query();
		query.addCriteria(Criteria.where("name").is("appleB"));
		query.fields().include("name");
 
		User userTest2 = mongoOperation.findOne(query, User.class);
		System.out.println("userTest2 - " + userTest2);
 
		userTest2.setAge(99);
 
		mongoOperation.save(userTest2);
 
		// ooppss, you just override everything, it caused ic=null and
		// createdDate=null
 
		Query query1 = new Query();
		query1.addCriteria(Criteria.where("name").is("appleB"));
 
		User userTest2_1 = mongoOperation.findOne(query1, User.class);
		System.out.println("userTest2_1 - " + userTest2_1);

Output

userTest2 - User [id=51603dba3004d7fffc202391, ic=null, name=appleB, age=0, createdDate=null]
userTest2_1 - User [id=51603dba3004d7fffc202391, ic=null, name=appleB, age=99, createdDate=null]

After the save(), the field 'age' is updated correctly, but ic and createdDate are both set to null, the entire "user" object is updated. To update a single field / key value, don't use save(), use updateFirst() or updateMulti() instead.

3. updateFirst example

Updates the first document that matches the query. In this case, only the single field "age" is updated.

{ 
	"_id" : ObjectId("id"), 
	"ic" : "1003", 
	"name" : "appleC", 
	"age" : 20, 
	"createdDate" : ISODate("2013-04-06T23:22:34.530Z") 
}
		//returns only 'name' field
		Query query = new Query();
		query.addCriteria(Criteria.where("name").is("appleC"));
		query.fields().include("name");
 
		User userTest3 = mongoOperation.findOne(query, User.class);
		System.out.println("userTest3 - " + userTest3);
 
		Update update = new Update();
		update.set("age", 100);
 
		mongoOperation.updateFirst(query, update, User.class);
 
		//returns everything
		Query query1 = new Query();
		query1.addCriteria(Criteria.where("name").is("appleC"));
 
		User userTest3_1 = mongoOperation.findOne(query1, User.class);
		System.out.println("userTest3_1 - " + userTest3_1);

Output

userTest3 - User [id=id, ic=null, name=appleC, age=0, createdDate=null]
userTest3_1 - User [id=id, ic=1003, name=appleC, age=100, createdDate=Sat Apr 06 23:22:34 MYT 2013]

4. updateMulti example

Updates all documents that matches the query.

{ 
	"_id" : ObjectId("id"), 
	"ic" : "1004", 
	"name" : "appleD", 
	"age" : 20, 
	"createdDate" : ISODate("2013-04-06T15:22:34.530Z") 
}
{ 
	"_id" : ObjectId("id"), 
	"ic" : "1005", 
	"name" : "appleE", 
	"age" : 20, 
	"createdDate" : ISODate("2013-04-06T15:22:34.530Z") 
}
		//show the use of $or operator
		Query query = new Query();
		query.addCriteria(Criteria
				.where("name").exists(true)
				.orOperator(Criteria.where("name").is("appleD"),
						Criteria.where("name").is("appleE")));
		Update update = new Update();
 
		//update age to 11
		update.set("age", 11);
 
		//remove the createdDate field
		update.unset("createdDate");
 
		// if use updateFirst, it will update 1004 only.
		// mongoOperation.updateFirst(query4, update4, User.class);
 
		// update all matched, both 1004 and 1005
		mongoOperation.updateMulti(query, update, User.class);
 
		System.out.println(query.toString());
 
		List<User> usersTest4 = mongoOperation.find(query4, User.class);
 
		for (User userTest4 : usersTest4) {
			System.out.println("userTest4 - " + userTest4);
		}

Output

Query: { "name" : { "$exists" : true} , 
	"$or" : [ { "name" : "appleD"} , { "name" : "appleE"}]}, Fields: null, Sort: null
 
userTest4 - User [id=id, ic=1004, name=appleD, age=11, createdDate=null]
userTest4 - User [id=id, ic=1005, name=appleE, age=11, createdDate=null]

5. Upsert example

If document is matched, update it, else create a new document by combining the query and update object, it's works likefindAndModifyElseCreate() :)

{ 
	//no data
}
		//search a document that doesn't exist
		Query query = new Query();
		query.addCriteria(Criteria.where("name").is("appleZ"));
 
		Update update = new Update();
		update.set("age", 21);
 
		mongoOperation.upsert(query, update, User.class);
 
		User userTest5 = mongoOperation.findOne(query, User.class);
		System.out.println("userTest5 - " + userTest5);

Output, a new document is created by combining both query and update object.

userTest5 - User [id=id, ic=null, name=appleZ, age=21, createdDate=null]

6. findAndModify example

Find and modify and get the newly updated object from a single operation.

{ 
	"_id" : ObjectId("id"), 
	"ic" : "1006", 
	"name" : "appleF", 
	"age" : 20, 
	"createdDate" : ISODate("2013-04-07T13:11:34.530Z") 
}
		Query query6 = new Query();
		query6.addCriteria(Criteria.where("name").is("appleF"));
 
		Update update6 = new Update();
		update6.set("age", 101);
		update6.set("ic", 1111);
 
		//FindAndModifyOptions().returnNew(true) = newly updated document
		//FindAndModifyOptions().returnNew(false) = old document (not update yet)
		User userTest6 = mongoOperation.findAndModify(
				query6, update6, 
				new FindAndModifyOptions().returnNew(true), User.class);
		System.out.println("userTest6 - " + userTest6);

Output

userTest6 - User [id=id, ic=1111, name=appleF, age=101, createdDate=Sun Apr 07 13:11:34 MYT 2013]

7. Full example

Full application to combine everything from example 1 to 6.

package com.mkyong.core;
 
import java.util.ArrayList;
import java.util.Date;
import java.util.List;
 
import org.springframework.context.ApplicationContext;
import org.springframework.context.annotation.AnnotationConfigApplicationContext;
import org.springframework.data.mongodb.core.FindAndModifyOptions;
import org.springframework.data.mongodb.core.MongoOperations;
import org.springframework.data.mongodb.core.query.Criteria;
import org.springframework.data.mongodb.core.query.Query;
import org.springframework.data.mongodb.core.query.Update;
 
import com.mkyong.config.SpringMongoConfig;
import com.mkyong.model.User;
 
public class UpdateApp {
 
	public static void main(String[] args) {
		// For Annotation
		ApplicationContext ctx = 
			new AnnotationConfigApplicationContext(SpringMongoConfig.class);
		MongoOperations mongoOperation = 
			(MongoOperations) ctx.getBean("mongoTemplate");
 
		// insert 6 users for testing
		List<User> users = new ArrayList<User>();
 
		User user1 = new User("1001", "appleA", 20, new Date());
		User user2 = new User("1002", "appleB", 20, new Date());
		User user3 = new User("1003", "appleC", 20, new Date());
		User user4 = new User("1004", "appleD", 20, new Date());
		User user5 = new User("1005", "appleE", 20, new Date());
		User user6 = new User("1006", "appleF", 20, new Date());
		users.add(user1);
		users.add(user2);
		users.add(user3);
		users.add(user4);
		users.add(user5);
		users.add(user6);
		mongoOperation.insert(users, User.class);
 
		// Case 1 ... find and update
		System.out.println("Case 1");
 
		Query query1 = new Query();
		query1.addCriteria(Criteria.where("name").is("appleA"));
 
		User userTest1 = mongoOperation.findOne(query1, User.class);
 
		System.out.println("userTest1 - " + userTest1);
 
		userTest1.setAge(99);
		mongoOperation.save(userTest1);
 
		User userTest1_1 = mongoOperation.findOne(query1, User.class);
 
		System.out.println("userTest1_1 - " + userTest1_1);
 
		// Case 2 ... select single field only
		System.out.println("\nCase 2");
 
		Query query2 = new Query();
		query2.addCriteria(Criteria.where("name").is("appleB"));
		query2.fields().include("name");
 
		User userTest2 = mongoOperation.findOne(query2, User.class);
		System.out.println("userTest2 - " + userTest2);
 
		userTest2.setAge(99);
 
		mongoOperation.save(userTest2);
 
		// ooppss, you just override everything, it caused ic=null and
		// createdDate=null
 
		Query query2_1 = new Query();
		query2_1.addCriteria(Criteria.where("name").is("appleB"));
 
		User userTest2_1 = mongoOperation.findOne(query2_1, User.class);
		System.out.println("userTest2_1 - " + userTest2_1);
 
		System.out.println("\nCase 3");
		Query query3 = new Query();
		query3.addCriteria(Criteria.where("name").is("appleC"));
		query3.fields().include("name");
 
		User userTest3 = mongoOperation.findOne(query3, User.class);
		System.out.println("userTest3 - " + userTest3);
 
		Update update3 = new Update();
		update3.set("age", 100);
 
		mongoOperation.updateFirst(query3, update3, User.class);
 
		Query query3_1 = new Query();
		query3_1.addCriteria(Criteria.where("name").is("appleC"));
 
		User userTest3_1 = mongoOperation.findOne(query3_1, User.class);
		System.out.println("userTest3_1 - " + userTest3_1);
 
		System.out.println("\nCase 4");
		Query query4 = new Query();
		query4.addCriteria(Criteria
				.where("name")
				.exists(true)
				.orOperator(Criteria.where("name").is("appleD"),
						Criteria.where("name").is("appleE")));
		Update update4 = new Update();
		update4.set("age", 11);
		update4.unset("createdDate");
 
		// update 1004 only.
		// mongoOperation.updateFirst(query4, update4, User.class);
 
		// update all matched
		mongoOperation.updateMulti(query4, update4, User.class);
 
		System.out.println(query4.toString());
 
		List<User> usersTest4 = mongoOperation.find(query4, User.class);
 
		for (User userTest4 : usersTest4) {
			System.out.println("userTest4 - " + userTest4);
		}
 
		System.out.println("\nCase 5");
		Query query5 = new Query();
		query5.addCriteria(Criteria.where("name").is("appleZ"));
 
		Update update5 = new Update();
		update5.set("age", 21);
 
		mongoOperation.upsert(query5, update5, User.class);
 
		User userTest5 = mongoOperation.findOne(query5, User.class);
		System.out.println("userTest5 - " + userTest5);
 
		System.out.println("\nCase 6");
		Query query6 = new Query();
		query6.addCriteria(Criteria.where("name").is("appleF"));
 
		Update update6 = new Update();
		update6.set("age", 101);
		update6.set("ic", 1111);
 
		User userTest6 = mongoOperation.findAndModify(query6, update6,
				new FindAndModifyOptions().returnNew(true), User.class);
		System.out.println("userTest6 - " + userTest6);
 
		mongoOperation.dropCollection(User.class);
 
	}
 
}

Output

Case 1
userTest1 - User [id=id, ic=1001, name=appleA, age=20, createdDate=Sun Apr 07 13:22:48 MYT 2013]
userTest1_1 - User [id=id, ic=1001, name=appleA, age=99, createdDate=Sun Apr 07 13:22:48 MYT 2013]
 
Case 2
userTest2 - User [id=id, ic=null, name=appleB, age=0, createdDate=null]
userTest2_1 - User [id=id, ic=null, name=appleB, age=99, createdDate=null]
 
Case 3
userTest3 - User [id=id, ic=null, name=appleC, age=0, createdDate=null]
userTest3_1 - User [id=id, ic=1003, name=appleC, age=100, createdDate=Sun Apr 07 13:22:48 MYT 2013]
 
Case 4
Query: { "name" : { "$exists" : true} , "$or" : [ { "name" : "appleD"} , { "name" : "appleE"}]}, Fields: null, Sort: null
userTest4 - User [id=id, ic=1004, name=appleD, age=11, createdDate=null]
userTest4 - User [id=id, ic=1005, name=appleE, age=11, createdDate=null]
 
Case 5
userTest5 - User [id=id, ic=null, name=appleZ, age=21, createdDate=null]
 
Case 6
userTest6 - User [id=id, ic=1006, name=appleF, age=20, createdDate=Sun Apr 07 13:22:48 MYT 2013]