net.snowflake.client.jdbc.cloud.storage.StorageObjectSummary Maven / Gradle / Ivy
/*
* Copyright (c) 2012-2018 Snowflake Computing Inc. All right reserved.
*/
package net.snowflake.client.jdbc.cloud.storage;
import com.amazonaws.services.s3.model.S3ObjectSummary;
import com.amazonaws.util.Base64;
import com.microsoft.azure.storage.StorageException;
import com.microsoft.azure.storage.blob.BlobProperties;
import com.microsoft.azure.storage.blob.CloudBlob;
import com.microsoft.azure.storage.blob.ListBlobItem;
import java.net.URISyntaxException;
/**
* Storage platform agnostic class that encapsulates remote storage object properties
*
* @author lgiakoumakis
*/
public class StorageObjectSummary
{
private String location; // location translates to "bucket" for S3
private String key;
private String md5;
private long size;
/**
* Contructs a StorageObjectSummary object from the S3 equivalent S3ObjectSummary
* @param
*/
private StorageObjectSummary(String location, String key, String md5, long size)
{
this.location = location;
this.key = key;
this.md5 = md5;
this.size = size;
}
/**
* Contructs a StorageObjectSummary object from the S3 equivalent S3ObjectSummary
* @param objSummary the AWS S3 ObjectSummary object to copy from
* @return the ObjectSummary object created
*/
public static StorageObjectSummary createFromS3ObjectSummary(S3ObjectSummary objSummary)
{
return new StorageObjectSummary(
objSummary.getBucketName(),
objSummary.getKey(),
// S3 ETag is not always MD5, but since this code path is only
// used in skip duplicate files in PUT command, It's not
// critical to guarantee that it's MD5
objSummary.getETag(),
objSummary.getSize()
);
}
/**
* Contructs a StorageObjectSummary object from Azure BLOB properties
* Using factory methods to create these objects since Azure can throw,
* while retrieving the BLOB properties
* @param listBlobItem an Azure ListBlobItem object
* @return the ObjectSummary object created
*/
public static StorageObjectSummary createFromAzureListBlobItem(ListBlobItem listBlobItem)
throws StorageProviderException
{
String location, key, md5;
long size;
// Retrieve the BLOB properties that we need for the Summary
// Azure Storage stores metadata inside each BLOB, therefore the listBlobItem
// will point us to the underlying BLOB and will get the properties from it
// During the process the Storage Client could fail, hence we need to wrap the
// get calls in try/catch and handle possible exceptions
try
{
location = listBlobItem.getContainer().getName();
CloudBlob cloudBlob = (CloudBlob) listBlobItem;
key = cloudBlob.getName();
BlobProperties blobProperties = cloudBlob.getProperties();
// the content md5 property is not always the actual md5 of the file. But for here, it's only
// used for skipping file on PUT command, hense is ok.
md5 = convertBase64ToHex(blobProperties.getContentMD5());
size = blobProperties.getLength();
}
catch (URISyntaxException | StorageException ex)
{
// This should only happen if somehow we got here with and invalid URI (it should never happen)
// ...or there is a Storage service error. Unlike S3, Azure fetches metadata from the BLOB itself,
// and its a lazy operation
throw new StorageProviderException(ex);
}
return new StorageObjectSummary(location, key, md5, size);
}
private static String convertBase64ToHex(String base64String){
try{
byte[] bytes = Base64.decode(base64String);
final StringBuilder builder = new StringBuilder();
for(byte b : bytes) {
builder.append(String.format("%02x", b));
}
return builder.toString();
// return empty string if input is not a valid Base64 string
}catch(Exception e){
return "";
}
}
/**
* @return returns the location of the object
*/
public String getLocation()
{
return location;
}
/**
* @return returns the key property of the object
*/
public String getKey() {
return key;
}
/**
* @return returns the MD5 hash of the object
*/
public String getMD5()
{
return md5;
}
/**
* @return returns the size property of the object
*/
public long getSize()
{
return size;
}
}