Java-SDK之如何实现管理文件（六）？

简单拷贝

1. // endpoint以杭州为例，其它region请按实际情况填写
2. String endpoint = "http://oss-cn-hangzhou.aliyuncs.com";
3. // 云账号AccessKey有所有API访问权限，建议遵循阿里云安全最佳实践，创建并使用RAM子账号进行API访问或日常运维，请登录 https://ram.console.aliyun.com 创建
4. String accessKeyId = "<yourAccessKeyId>";
5. String accessKeySecret = "<yourAccessKeySecret>";
7. // 创建OSSClient实例
8. OSSClient ossClient = new OSSClient(endpoint, accessKeyId, accessKeySecret);
9. // 拷贝Object
10. CopyObjectResult result = ossClient.copyObject("<srcBucketName>", "<srcKey>", "<destBucketName>", "<destKey>");
11. System.out.println("ETag: " + result.getETag() + " LastModified: " + result.getLastModified());
12. // 关闭client
13. ossClient.shutdown();

1. // endpoint以杭州为例，其它region请按实际情况填写
2. String endpoint = "http://oss-cn-hangzhou.aliyuncs.com";
3. // 云账号AccessKey有所有API访问权限，建议遵循阿里云安全最佳实践，创建并使用RAM子账号进行API访问或日常运维，请登录 https://ram.console.aliyun.com 创建
4. String accessKeyId = "<yourAccessKeyId>";
5. String accessKeySecret = "<yourAccessKeySecret>";
7. // 创建OSSClient实例
8. OSSClient ossClient = new OSSClient(endpoint, accessKeyId, accessKeySecret);
9. // 创建CopyObjectRequest对象
10. CopyObjectRequest copyObjectRequest = new CopyObjectRequest(srcBucketName, srcKey, destBucketName, destKey);
11. // 设置新的Metadata
12. ObjectMetadata meta = new ObjectMetadata();
13. meta.setContentType("text/html");
14. copyObjectRequest.setNewObjectMetadata(meta);
15. // 复制Object
16. CopyObjectResult result = ossClient.copyObject(copyObjectRequest);
17. System.out.println("ETag: " + result.getETag() + " LastModified: " + result.getLastModified());
18. // 关闭client
19. ossClient.shutdown();

• 初始化分片拷贝任务OSSClient.initiateMultipartUpload；
• 分片拷贝OSSClient.uploadPartCopy，除最后一个分片外，其它的分片大小都要大于100KB；
• 提交分片拷贝任务OSSClient.completeMultipartUpload。

提示：

• 分片拷贝的完整代码请参考：GitHub

1. String sourceBucketName = "<sourceBucketName>";
2. String sourceKey = "<sourceKey>";
3. String targetBucketName = "<targetBucketName>";
4. String targetKey = "<targetKey>";
6. // 得到被拷贝object大小
7. ObjectMetadata objectMetadata = ossClient.getObjectMetadata(sourceBucketName, sourceKey);
8. long contentLength = objectMetadata.getContentLength();
9. // 分片大小，10MB
10. long partSize = 1024 * 1024 * 10;
11. // 计算分块数目
12. int partCount = (int) (contentLength / partSize);
13. if (contentLength % partSize != 0) {
14.     partCount++;
15. }
16. System.out.println("total part count:" + partCount);
17. // 初始化拷贝任务
18. InitiateMultipartUploadRequest initiateMultipartUploadRequest = new InitiateMultipartUploadRequest(targetBucketName, targetKey);
19. InitiateMultipartUploadResult initiateMultipartUploadResult = ossClient.initiateMultipartUpload(initiateMultipartUploadRequest);
20. String uploadId = initiateMultipartUploadResult.getUploadId();
21. // 分片拷贝
22. List<PartETag> partETags = new ArrayList<PartETag>();
23. for (int i = 0; i < partCount; i++) {
24.      // 计算每个分块的大小
25.     long skipBytes = partSize * i;
26.     long size = partSize < contentLength - skipBytes ? partSize : contentLength - skipBytes;
27.     // 创建UploadPartCopyRequest
28.     UploadPartCopyRequest uploadPartCopyRequest =
29.         new UploadPartCopyRequest(sourceBucketName, sourceKey, targetBucketName, targetKey);
30.     uploadPartCopyRequest.setUploadId(uploadId);
31.     uploadPartCopyRequest.setPartSize(size);
32.     uploadPartCopyRequest.setBeginIndex(skipBytes);
33.     uploadPartCopyRequest.setPartNumber(i + 1);
34.     UploadPartCopyResult uploadPartCopyResult = ossClient.uploadPartCopy(uploadPartCopyRequest);
35.     // 将返回的PartETag保存到List中
36.     partETags.add(uploadPartCopyResult.getPartETag());
37. }
38. // 提交分片拷贝任务
39. CompleteMultipartUploadRequest completeMultipartUploadRequest = new CompleteMultipartUploadRequest(
40.                     targetBucketName, targetKey, uploadId, partETags);
41. ossClient.completeMultipartUpload(completeMultipartUploadRequest);

提示：

• 分片拷贝时，可以指定目标object的元信息，通过InitiateMultipartUploadRequest指定；
• 分片拷贝也支持限定条件，通过UploadPartCopyRequest指定。

• 归档类型的文件初始时处于冷冻状态；
• 提交解冻（Restore）操作后，服务端执行解冻，文件处于解冻中；
• 完成解冻后，可以读取Object；
• 解冻状态默认持续1天，最多延长7天，之后文件又回到冷冻状态。

1. ObjectMetadata objectMetadata = ossClient.getObjectMetadata(bucketName, key);
3. // check whether the object is archive class
4. StorageClass storageClass = objectMetadata.getObjectStorageClass();
5. if (storageClass == StorageClass.Archive) {
6.     // restore object
7.     ossClient.restoreObject(bucketName, key);
8.     // wait for restore completed
9.     do {
10.         Thread.sleep(1000);
11.         objectMetadata = ossClient.getObjectMetadata(bucketName, key);
12.     } while (!objectMetadata.isRestoreCompleted());
13. }
14. // get restored object
15. OSSObject ossObject = ossClient.getObject(bucketName, key);
16. ossObject.getObjectContent().close();

提示：

• 上述示例的完整代码请参看 GitHUb；
• 非归档类型的文件，不要调用 restoreObject 接口。