Spring Boot 集成 AmazonS3 存储服务教程

原文链接：https://www.changxuan.top/?p=963

环境准备

Java环境：JDK6以上版本
SpringBoot 2.2.5
存储服务的 accessKey、secreKey 以及地址:端口

使用

1. 引入依赖

在 pom.xml 文件中添加下列依赖：

<dependency><groupId>com.amazonaws</groupId><artifactId>aws-java-sdk</artifactId><version>1.11.433</version>
</dependency>

2. 初始化

> 简单模式

String accessKey = "your-accesskey";
String secretKey = "your-secretKey";AWSCredentials credentials = new BasicAWSCredentials(accessKey, secretKey);
AmazonS3 conn = AmazonS3ClientBuilder.standard().withCredentials(new AWSStaticCredentialsProvider(credentials)).withEndpointConfiguration(new AwsClientBuilder.EndpointConfiguration("endpoint","region")).build();

示例：

String accessKey = "TPDDEA5PCT9C8RUPKAWW";
String secretKey = "ryesPMSiSfOzaP1TkFe9TIOBnxpdrA2sw6isFwDZ";AWSCredentials credentials = new BasicAWSCredentials(accessKey, secretKey);
AmazonS3 conn = AmazonS3ClientBuilder.standard().withCredentials(new AWSStaticCredentialsProvider(credentials)).withEndpointConfiguration(new   AwsClientBuilder.EndpointConfiguration("http://10.200.100.37:7480","region")).build();

> 配置模式

import com.amazonaws.ClientConfiguration;
import com.amazonaws.Protocol;
import com.amazonaws.auth.AWSCredentials;
import com.amazonaws.auth.AWSStaticCredentialsProvider;
import com.amazonaws.auth.BasicAWSCredentials;
import com.amazonaws.client.builder.AwsClientBuilder;
import com.amazonaws.services.s3.AmazonS3;
import com.amazonaws.services.s3.AmazonS3ClientBuilder;String accessKey = "your-accesskey";
String secretKey = "your-secretKey ";
AWSCredentials credentials = new BasicAWSCredentials(accessKey,secretKey);
ClientConfiguration conf = new ClientConfiguration();
// 设置AmazonS3使用的最大连接数
conf.setMaxConnections(200);
// 设置socket超时时间
conf.setSocketTimeout(10000);
// 设置失败请求重试次数
conf.setMaxErrorRetry(2);
// 如果要用https协议，请加上下面语句
conf.setProtocol(Protocol.HTTPS);//AmazonS3 s3Client = new AmazonS3Client(credentials,clientConfiguration);
//s3Client.setEndpoint(endPoint);
AmazonS3 s3Client = AmazonS3ClientBuilder.standard().withCredentials(new AWSStaticCredentialsProvider(credentials)).withEndpointConfiguration(new AwsClientBuilder.EndpointConfiguration("endpoint","region")).withClientConfiguration(conf).build();//endpoint,region请指定为NOS支持的

示例：

Amazon3 是线程安全的，因此可以将其注入到 Spring 容器中进行管理。

@AmazonS3Config.java

import com.amazonaws.ClientConfiguration;
import com.amazonaws.Protocol;
import com.amazonaws.auth.AWSCredentials;
import com.amazonaws.auth.AWSStaticCredentialsProvider;
import com.amazonaws.auth.BasicAWSCredentials;
import com.amazonaws.client.builder.AwsClientBuilder;
import com.amazonaws.services.s3.AmazonS3;
import com.amazonaws.services.s3.AmazonS3ClientBuilder;
import com.amazonaws.services.s3.model.CreateBucketRequest;
import com.amazonaws.services.s3.transfer.TransferManager;
import com.amazonaws.services.s3.transfer.TransferManagerBuilder;
import org.springframework.beans.factory.annotation.Autowired;
import org.springframework.boot.context.properties.EnableConfigurationProperties;
import org.springframework.context.annotation.Bean;
import org.springframework.context.annotation.Configuration;/*** @Author: ChangXuan* @Decription: 初始化AmazonS3实例* @Date: 17:33 2020/6/3**/
@Configuration
@EnableConfigurationProperties(UploadConfig.class)
public class AmazonS3Config {@Autowiredprivate UploadConfig uploadConfig;@Bean(name = "amazonS3")public AmazonS3 getAmazonS3(){String accessKey = uploadConfig.getCeph().getAccessKey();String secretKey = uploadConfig.getCeph().getSecretKey();AWSCredentials credentials = new BasicAWSCredentials(accessKey,secretKey);ClientConfiguration conf = new ClientConfiguration();// 设置AmazonS3使用的最大连接数conf.setMaxConnections(uploadConfig.getCeph().getAmazonS3MaxConnections());// 设置socket超时时间conf.setSocketTimeout(uploadConfig.getCeph().getAmazonS3SocketTimeout());// 设置失败请求重试次数conf.setMaxErrorRetry(uploadConfig.getCeph().getAmazonS3MaxErrorRetry());// 设置协议if (!"blank".equals(uploadConfig.getCeph().getAmazonS3Protocol())){switch (uploadConfig.getCeph().getAmazonS3Protocol()){case "https":conf.setProtocol(Protocol.HTTPS);break;case "http":conf.setProtocol(Protocol.HTTP);break;default:break;}}AmazonS3 s3Client = AmazonS3ClientBuilder.standard().withCredentials(new AWSStaticCredentialsProvider(credentials)).withEndpointConfiguration(new AwsClientBuilder.EndpointConfiguration(uploadConfig.getCeph().getHosts().get(0),uploadConfig.getCeph().getRegion())).withClientConfiguration(conf).build();checkAndCreateBucket(s3Client);return s3Client;}@Bean(name = "transferManager")public TransferManager getTransferManager(){return TransferManagerBuilder.standard().withS3Client(getAmazonS3()).build();}/*** 检查桶是否存在，不存在则创建创建* @param s3Client*/private void checkAndCreateBucket(AmazonS3 s3Client){boolean exists = s3Client.doesBucketExistV2(uploadConfig.getCeph().getBucketName());if (!exists){CreateBucketRequest request = new CreateBucketRequest(uploadConfig.getCeph().getBucketName());s3Client.createBucket(request);}}

@UploadConfig.java

import lombok.Data;
import org.springframework.boot.context.properties.ConfigurationProperties;/*** @Author: ChangXuan* @Decription: 上传 配置* @Date: 17:25 2020/6/3**/
@ConfigurationProperties(prefix = "upload")
@Data
public class UploadConfig {private CephConfig ceph;
}

@CephConfig.java

import com.google.common.collect.Lists;
import lombok.Data;import java.util.List;/*** @Author: ChangXuan* @Decription:配置* @Date: 17:28 2020/6/3**/
@Data
public class CephConfig {/*** 集群ip:port*/private List<String> hosts = Lists.newArrayList();private String accessKey;private String secretKey;private String region;/*** 连接协议*/private String amazonS3Protocol;/*** 失败请求重试次数*/private Integer amazonS3MaxErrorRetry;/*** 超时时间*/private Integer amazonS3SocketTimeout;/*** 最大连接数*/private Integer amazonS3MaxConnections;/*** 桶名称*/private String bucketName;}

@application.yml

upload:ceph:bucketName: ic-storage-devregion: regionhosts:- "10.200.100.37:7480"- "10.200.100.38:7480"- "10.200.100.39:7480"accessKey: TPDDEA5PCT9C8RUPKAWWsecretKey: ryesPMSiSfOzaP1TkFe9TIWQnxpdrA2sw6isFwDZamazonS3MaxConnections: 200amazonS3SocketTimeout: 10000amazonS3MaxErrorRetry: 2amazonS3Protocol: http

3. 依赖注入

@Autowired
private AmazonS3 amazonS3;//或@Autowired
private TransferManager transferManager;

4. 文件操作

文件上传

1. 直接内容上传

//要上传文件内容
String content = "Object content";
try {amazonS3.putObject("your-bucketname","your-objectname",content);
}catch (Exception e){System.out.println(e.getMessage());
}

2. 本地文件普通上传

对于小对象可以使用putObject接口进行上传，putObject上传支持的最大文件大小为100M，如果上传大于100M的文件需要使用分块上传。本地文件普通上传的示例代码如下：

//要上传文件的路径
String filePath = "your-local-file-path";
try {amazonS3.putObject("your-bucketname","your-objectname", new File(filePath));
}catch (Exception e){System.out.println(e.getMessage());
}

3. 上传文件时设置文件元数据信息

String filePath = "your-local-file-path";
ObjectMetadata objectMetadata = new ObjectMetadata();
//设置Content-Type
objectMetadata.setContentType("application/xml");
//设置标准http消息头（元数据）
objectMetadata.setHeader("Cache-Control", "no-cache");
//设置用户自定义元数据信息
Map<String, String> userMeta = new HashMap<String, String>();
userMeta.put("ud", "test");
objectMetadata.setUserMetadata(userMeta);
PutObjectRequest putObjectRequest = new PutObjectRequest("your-bucketname","your-objectname", new File(filePath));
putObjectRequest.setMetadata(objectMetadata);
amazonS3.putObject(putObjectRequest);

4. 流式上传

try {ObjectMetadata objectMetadata = new ObjectMetadata();//设置流的长度，您还可以设置其他文件元数据信息objectMetadata.setContentLength(streamLength);amazonS3.putObject("your-bucketname","your-objectname", inputStream, objectMetadata)
}catch (Exception e){System.out.println(e.getMessage());
}

文件下载

1. 流式下载

S3Object fileObject = amazonS3.getObject("your-bucketname","your-objectname");
//可以通过getObjectMetadata方法获取对象的ContentType等元数据信息
String contentType = fileObject.getObjectMetadata().getContentType();
//流式获取文件内容
InputStream in = fileObject.getObjectContent();
BufferedReader reader = new BufferedReader(new InputStreamReader(in));
while (true) {String line;try {line = reader.readLine();if (line == null) break;System.out.println("\n" + line);} catch (IOException e) {e.printStackTrace();}
}
try {reader.close();
} catch (IOException e) {e.printStackTrace();
}

2. 下载到本地文件

String destinationFile = "your-local-filepath";
GetObjectRequest getObjectRequest = new GetObjectRequest("your-bucketname","your-objectname");
ObjectMetadata objectMetadata = amazonS3.getObject(getObjectRequest, new File(destinationFile));

3. Range 下载

GetObjectRequest getObjectRequest = new GetObjectRequest("your-bucketname","your-objectname");
getObjectRequest.setRange(0, 100);
S3Object nosObject = amazonS3.getObject(getObjectRequest);
BufferedReader reader = new BufferedReader(new InputStreamReader(in));
while (true) {String line;try {line = reader.readLine();if (line == null) break;System.out.println("\n" + line);} catch (IOException e) {e.printStackTrace();}
}
try {reader.close();
} catch (IOException e) {e.printStackTrace();
}

文件管理

判断文件是否存在

您可以通过AmazonS3.doesObjectExist判断文件是否存在。

boolean isExist = amazonS3.doesObjectExist("your-bucketname","your-objectname");

文件删除

您可以通过AmazonS3.deleteObject删除单个文件

amazonS3.deleteObject("your-bucketname","your-objectname");

您还可以通过AmazonS3.deleteObjects一次删除多个文件

try {DeleteObjectsResult result = amazonS3.deleteObjects(deleteObjectsRequest);List<DeletedObject>  deleteObjects = result.getDeletedObjects();//print the delete resultsfor (DeletedObject items: deleteObjects){System.out.println(items.getKey());}
// 部分对象删除失败
} catch (MultiObjectDeleteException e) { List<DeleteError> deleteErrors = e.getErrors();for (DeleteError error : deleteErrors) {System.out.println(error.getKey());}
} catch (AmazonServiceException  e) {//捕捉服务器异常错误
} catch (AmazonClientException ace) {//捕捉客户端错误
}

获取文件元数据信息

您可以通过AmazonS3.getObjectMetadata获取文件元数据信息

amazonS3.getObjectMetadata("your-bucketname","your-objectname");

文件复制(copy)

您可以通过AmazonS3.copyObject接口实现文件拷贝功能。

amazonS3.copyObject("source-bucket", "source-object", "dst-bucket", "dst-object");

列举桶内文件

您可以通过AmazonS3.listObjects列出桶里的文件。listObjects接口如果调用成功，会返回一个ObjectListing对象，列举的结果保持在该对象中。

ObjectListing的具体信息如下表所示：

方法	含义
List getObjectSummaries()	返回的文件列表（包含文件的名称、Etag的元数据信息）
String getPrefix()	本次查询的文件名前缀
String getDelimiter()	文件分界符
String getMarker()	这次List Objects的起点
int getMaxKeys()	响应请求内返回结果的最大数目
String getNextMarker()	下一次List Object的起点
boolean isTruncated()	是否截断，如果因为设置了limit导致不是所有的数据集都返回，则该值设置为true
List getCommonPrefixes()	如果请求中指定了delimiter参数，则返回的包含CommonPrefixes元素。该元素标明以delimiter结尾，并有共同前缀的对象的集合

AmazonS3.listObjects接口提供两种调用方式：简单列举、通过ListObjectsRequest列举

简单列举

简单列举只需指定需要列举的桶名，最多返回100条对象记录，建议桶内对象数较少时（小于100）使用。

ObjectListing objectListing = amazonS3.listObjects("your-bucketname");
List<S3ObjectSummary> sums = objectListing.getObjectSummaries();
for (S3ObjectSummary s : sums) {System.out.println("\t" + s.getKey());
}

通过ListObjectsRequest列举

您还可以通过设置ListObjectsReques参数实现各种灵活的查询功能。ListObjectsReques的可设置的参数如下：

设置方法	作用
setPrefix(String prefix)	限定返回的object key必须以prefix作为前缀
setDelimiter(String delimiter)	是一个用于对Object名字进行分组的字符。所有名字包含指定的前缀且第一次出现delimiter字符之间的object作为一组元素——CommonPrefixes
setMarker(String marker)	字典序的起始标记，只列出该标记之后的部分
setMaxKeys(Integer maxKeys)	限定返回的数量，返回的结果小于或等于该值(默认值为100)

1、分页列举桶内的所有文件：

List<S3ObjectSummary> listResult = new ArrayList<S3ObjectSummary>();
ListObjectsRequest listObjectsRequest = new ListObjectsRequest();
listObjectsRequest.setBucketName("your-bucketname");
listObjectsRequest.setMaxKeys(50);
ObjectListing listObjects = amazonS3.listObjects(listObjectsRequest);
do {listResult.addAll(listObjects.getObjectSummaries());if (listObjects.isTruncated()) {ListObjectsRequest request = new ListObjectsRequest();request.setBucketName(listObjectsRequest.getBucketName());request.setMarker(listObjects.getNextMarker());listObjects =  amazonS3.listObjects(request);} else {break;}
} while (listObjects != null);

2、使用Delimiter模拟文件夹功能

假设桶内有如下对象：a/1.jpg、a/2.jpg、a/b/1.txt、a/b/2.txt，列举a文件夹下的文件及子文件夹的示例代码如下：

ListObjectsRequest listObjectsRequest = new ListObjectsRequest();
listObjectsRequest.setBucketName("your-bucketname");
listObjectsRequest.setDelimiter("/");
listObjectsRequest.setPrefix("a/");
ObjectListing listing = amazonS3.listObjects(listObjectsRequest);
// 遍历所有Object
System.out.println("Objects:");
for (S3ObjectSummary objectSummary : listing.getObjectSummaries()) {System.out.println(objectSummary.getKey());
}
// 遍历所有CommonPrefix
System.out.println("CommonPrefixs:");
for (String commonPrefix : listing.getCommonPrefixes()) {System.out.println(commonPrefix);
}

示例代码的输出如下：

Objects:

a/1.jpg

a/2.jpg

CommonPrefixs:

a/b/

生成私有对象可下载的URL链接

AWS Java SDK支持生成可下载私有对象的URL连接，您可以将该链接提供给第三方进行文件下载：

GeneratePresignedUrlRequest generatePresignedUrlRequest = new GeneratePresignedUrlRequest(bucketName, key);
// 设置可下载URL的过期时间为1天后
generatePresignedUrlRequest.setExpiration(new Date(System.currentTimeMillis()+3600*1000*24));
URL url = amazonS3.generatePresignedUrl(generatePresignedUrlRequest);//生成URL
System.out.println(url);//可以用这个url来下载文件

文件上传下载工具类 TransferManager

前文提到的是 Java SDK提供的基础接口，为方便用户进行文件上传下载，Java SDK提供了封装更好、使用更方便的工具类：TransferManager。

TransferManager的初始化

注：在 SpringBoot 项目中 TransferManager 的初始化参考上述 [配置模式](#### 配置模式)

//先实例化一个AmazonS3
String accessKey = "your-accesskey";
String secretKey = "your-secretKey ";
Credentials credentials = new BasicCredentials(accessKey, secretKey);
AmazonS3 amazonS3 = new AmazonS3(credentials);
amazonS3.setEndpoint(endPoint);
//然后通过AmazonS3对象来初始化TransferManager
TransferManager transferManager = new TransferManager(amazonS3);
Download download = transferManager.download(TestConfig.bucketName,key,new File("localFilePath"));
try {download.waitForCompletion();
} catch (InterruptedException e) {e.printStackTrace();
}

使用TransferManager进行文件上传

TransferManager会根据文件大小，选择是否进行分块上传。当文件小于等于16M时，TransferManager会自动调用PutObject接口，否则TransferManager会自动对文件进行分块上传。

1、上传本地文件：

如果指定上传的本地文件大于16M，TransferManager会自动对文件进行分块，并发调用分块上传接口进行上传，大大提高上传文件的速度。

//上传文件
Upload upload = transferManager.upload("your-bucketname", "your-objectname", new File("your-file"));
try {upload.waitForUploadResult();
} catch (InterruptedException e) {e.printStackTrace();
}

2、流式上传：

您也可以使用TransferManager进行流式上传，但是相比本地文件上传，流式上传无法做到多个分块并发上传，只能一个分块一个分块顺序上传。

//流式上传文件
ObjectMetadata objectMetadata = new ObjectMetadata();
objectMetadata.setContentLength(file.length());
Upload upload = transferManager.upload("your-bucketname", "your-objectname", inputStream, objectMetadata);
UploadResult  result = upload.waitForUploadResult();

3、上传目录

您可以使用TransferManager将某个目录下的文件全部上传到NOS,对象名即文件名

MultipleFileUpload result = transferManager.uploadDirectory("your-buckename", null, new File("dirPath"), false);
result.waitForCompletion();

3.2 支持多级目录，会递归的上传目录下的所有文件

MultipleFileUpload result = transferManager.uploadDirectory("your-buckename", null, new File("dirPath"), true);
result.waitForCompletion();

4、下载文件

File file = new  File("your-destFile");
Download download = transferManager.download("your-bucketname", "your-objectname", file);
download.waitForCompletion();

参考资料

[1] 网易数帆S3 Java SDK 手册