How to Upload Files to Amazon S3 in Spring Boot

January 25, 2021

Amazon simple storage (Amazon S3) is a service offered by Amazon web services that offers scalable, secure, and well performing object storage. This article will go over how to upload files into Amazon S3 using Spring Boot.

Prerequisites

Amazon web services account

Before we start creating our application, head over to Amazon console, and create an account. You will be given 12 months of free access to various Amazon web services that you can use to test various Amazon services.

After signing up, head over to Amazon console and search for Amazon S3 in the search box provided in the console.

Amazon search console

Amazon S3 bucket

After selecting Amazon S3 from the step above, create a new S3 bucket that we will use to store the files we will be uploading from our application.

Amazon creating S3 bucket

Name the bucket as spring-amazon-storage and leave all other settings as default then create the bucket.

Access and secret keys

Create a new access key from My Security Credentials navigation menu as shown in the image below. Copy the access and the secret key generated as we will be using them to access the bucket from the application we will be creating.

Amazon secret key

Creating the application

We will be using spring initializr to create our application. Head over to spring initializr and create a new Spring Boot application adding h2, spring boot dev tools, spring data jpa and spring web as dependencies then generate the project.

Spring Boot create application

Unzip the downloaded project and open it in your favorite IDE.

Adding Amazon SDK dependency

Amazon SDK makes it possible to interact with various Amazon services from our applications. In the pom.xml file add the Amazon SDK dependency as shown below.

    <!-- https://mvnrepository.com/artifact/com.amazonaws/aws-java-sdk -->
    <dependency>
        <groupId>com.amazonaws</groupId>
        <artifactId>aws-java-sdk</artifactId>
        <version>1.11.931</version>
    </dependency>

Project structure

config/
|--- AmazonConfig.java
|--- BucketName.java
controllers/
|--- TodoController.java
domain/
|--- Todo.java
repositories/
|--- TodoRepository.java
service/
|--- FileStore.java
|--- TodoService.java
|--- TodoServiceImpl.java
SpringAmazonApplication.java

Configuration package

In the configuration package, we have two Java files, one that authenticates with Amazon S3 and the other which contains the bucket name.

import com.amazonaws.auth.AWSCredentials;
import com.amazonaws.auth.AWSStaticCredentialsProvider;
import com.amazonaws.auth.BasicAWSCredentials;
import com.amazonaws.services.s3.AmazonS3;
import com.amazonaws.services.s3.AmazonS3ClientBuilder;
import org.springframework.context.annotation.Bean;
import org.springframework.context.annotation.Configuration;

@Configuration
public class AmazonConfig {
    @Bean
    public AmazonS3 s3() {
        AWSCredentials awsCredentials =
                new BasicAWSCredentials("accessKey", "secretKey");
        return AmazonS3ClientBuilder
                .standard()
                .withRegion("ap-south-1")
                .withCredentials(new AWSStaticCredentialsProvider(awsCredentials))
                .build();

    }
}

The AmazonConfig class above is annotated with @Configuration annotation to make it available to the Spring context as a configuration class. With the Amazon credentials we got from the Amazon console earlier we will authenticate to S3 using the AmazonS3ClientBuilder available in the Amazon-SDK that we added to our pom.xml ealier.

import lombok.AllArgsConstructor;
import lombok.Getter;

@AllArgsConstructor
@Getter
public enum BucketName {
    TODO_IMAGE("spring-amazon-storage");
    private final String bucketName;
}

In the BucketName enum above we pass in the bucket name that we created in the Amazon console earlier. The bucket will be used to store all of our file uploads.

  • @AllArgsConstructor annotation generates a constructor with the bucketName variable in the enum.
  • @Getter annotation generates a getter for the bucketName variable in the enum.

Domain package

In this package we have the Todo model that represents our Todo in the database.

@Data
@AllArgsConstructor
@NoArgsConstructor
@Builder
@Entity
public class Todo {
    @Id
    @GeneratedValue
    private Long id;
    private String title;
    private String description;
    private String imagePath;
    private String imageFileName;

}
  • @Data annotation generates getters, setters, toString and equals methods for the Todo class.
  • @AllArgsConstructor annotation generates a constructor with all the arguments for the Todo class.
  • @NoArgsConstructor annotation generates a constructor with no arguments for the Todo class.
  • @Builder annotation creates a builder pattern for the Todo class.
  • @Entity annotation makes the Todo class a database entity.
  • @Id annotation marks the id field as a primary key in the database.
  • @GeneratedValue annotation makes the id field auto-increment whenever a new todo is saved into the database.

Repository package

In this package, we have the repository class that extends the JPA CrudRepository interface that makes it possible to perform various database queries.

public interface TodoRepository extends CrudRepository<Todo, Long> {
    Todo findByTitle(String title);
}

Service package

@AllArgsConstructor
@Service
public class FileStore {
    private final AmazonS3 amazonS3;

    public void upload(String path,
                       String fileName,
                       Optional<Map<String, String>> optionalMetaData,
                       InputStream inputStream) {
        ObjectMetadata objectMetadata = new ObjectMetadata();
        optionalMetaData.ifPresent(map -> {
            if (!map.isEmpty()) {
                map.forEach(objectMetadata::addUserMetadata);
            }
        });
        try {
            amazonS3.putObject(path, fileName, inputStream, objectMetadata);
        } catch (AmazonServiceException e) {
            throw new IllegalStateException("Failed to upload the file", e);
        }
    }

    public byte[] download(String path, String key) {
        try {
            S3Object object = amazonS3.getObject(path, key);
            S3ObjectInputStream objectContent = object.getObjectContent();
            return IOUtils.toByteArray(objectContent);
        } catch (AmazonServiceException | IOException e) {
            throw new IllegalStateException("Failed to download the file", e);
        }
    }

}

In the FileStore class above, we have the logic used to upload and download files from Amazon S3.

In the upload method we pass in:

  • path is the path on the Amazon S3 bucket where the file will be stored.
  • fileName is the actual name of the file being uploaded. It will be used as the key when downloading the file from S3.
  • optionalMetaData map contains the details of the file i.e file type and file size.
  • inputStream contains the actual file that should be saved to Amazon S3.
      ObjectMetadata objectMetadata = new ObjectMetadata();
      optionalMetaData.ifPresent(map -> {
            if (!map.isEmpty()) {
                map.forEach(objectMetadata::addUserMetadata);
            }
        });

The above code block loops through the optionalMetaData map adding all of the file information to the S3 objectMetaData.

  • amazonS3.putObject(path, fileName, inputStream, objectMetadata); saves the file to Amazon S3 bucket.

In the download method:

  • S3Object object = amazonS3.getObject(path, key); downloads the file from the path passed in and with the file name similar to the key passed in the getObject method.
  • S3ObjectInputStream objectContent = object.getObjectContent(); gets an inputStream from the object returned from Amazon S3.
  • IOUtils.toByteArray(objectContent) converts the input stream to byteArray that can be sent over Restful APIs.
public interface TodoService {
    Todo saveTodo(String title, String description, MultipartFile file);

    byte[] downloadTodoImage(Long id);

    List<Todo> getAllTodos();
}

The TodoService interface above contains various methods that we will implement to be able to save and get todos.

@Service
@AllArgsConstructor
public class TodoServiceImpl implements TodoService {
    private final FileStore fileStore;
    private final TodoRepository repository;

    @Override
    public Todo saveTodo(String title, String description, MultipartFile file) {
        //check if the file is empty
        if (file.isEmpty()) {
            throw new IllegalStateException("Cannot upload empty file");
        }
        //Check if the file is an image
        if (!Arrays.asList(IMAGE_PNG.getMimeType(),
                IMAGE_BMP.getMimeType(),
                IMAGE_GIF.getMimeType(),
                IMAGE_JPEG.getMimeType()).contains(file.getContentType())) {
            throw new IllegalStateException("FIle uploaded is not an image");
        }
        //get file metadata
        Map<String, String> metadata = new HashMap<>();
        metadata.put("Content-Type", file.getContentType());
        metadata.put("Content-Length", String.valueOf(file.getSize()));
        //Save Image in S3 and then save Todo in the database
        String path = String.format("%s/%s", BucketName.TODO_IMAGE.getBucketName(), UUID.randomUUID());
        String fileName = String.format("%s", file.getOriginalFilename());
        try {
            fileStore.upload(path, fileName, Optional.of(metadata), file.getInputStream());
        } catch (IOException e) {
            throw new IllegalStateException("Failed to upload file", e);
        }
        Todo todo = Todo.builder()
                .description(description)
                .title(title)
                .imagePath(path)
                .imageFileName(fileName)
                .build();
        repository.save(todo);
        return repository.findByTitle(todo.getTitle());
    }

    @Override
    public byte[] downloadTodoImage(Long id) {
        Todo todo = repository.findById(id).get();
        return fileStore.download(todo.getImagePath(), todo.getImageFileName());
    }

    @Override
    public List<Todo> getAllTodos() {
        List<Todo> todos = new ArrayList<>();
        repository.findAll().forEach(todos::add);
        return todos;
    }
}

In the TodoServiceImpl above, we provide the implementation for the methods for saving and getting all todos.

Controllers package

In this package, we have TodoController class which handles the incoming HTTP requests.


@RestController
@RequestMapping("api/v1/todo")
@AllArgsConstructor
@CrossOrigin("*")
public class TodoController {
    TodoService service;

    @GetMapping
    public ResponseEntity<List<Todo>> getTodos() {
        return new ResponseEntity<>(service.getAllTodos(), HttpStatus.OK);
    }

    @PostMapping(
            path = "",
            consumes = MediaType.MULTIPART_FORM_DATA_VALUE,
            produces = MediaType.APPLICATION_JSON_VALUE
    )
    public ResponseEntity<Todo> saveTodo(@RequestParam("title") String title,
                                         @RequestParam("description") String description,
                                         @RequestParam("file") MultipartFile file) {
        return new ResponseEntity<>(service.saveTodo(title, description, file), HttpStatus.OK);
    }

    @GetMapping(value = "{id}/image/download")
    public byte[] downloadTodoImage(@PathVariable("id") Long id) {
        return service.downloadTodoImage(id);
    }
}

Testing our uploads and downloads from the S3 bucket

Amazon S3 upload

Conclusion

Congratulations! Now that you learned how to upload and download files from Amazon S3, go ahead and implement the logic for uploading multiple files to Amazon S3.

Find the source code here.

Happy Coding!


Peer Review Contributions by: Linus Muema


About the author

Odhiambo Paul

Odhiambo Paul is a second-year undergraduate student who develops Python, Java and Android applications. Paul has a great passion for writing clean and optimized code.

This article was contributed by a student member of Section's Engineering Education Program. Please report any errors or innaccuracies to enged@section.io.