通过ByteBuffer&;将Java对象序列化到Cassandra 1.2;CQL3
我拼凑了下面的代码,不做任何复杂的事情——只需创建一个byte[]变量,将其写入Cassandra中的blob字段(v1.2,通过新的Datastax CQL库),然后再次读取 当我把它放进去的时候,它有3个元素长,当我把它读回来的时候,它有84个元素长。。。!这意味着我实际尝试做的事情(序列化Java对象)失败,再次尝试反序列化时出现通过ByteBuffer&;将Java对象序列化到Cassandra 1.2;CQL3,java,serialization,cassandra,Java,Serialization,Cassandra,我拼凑了下面的代码,不做任何复杂的事情——只需创建一个byte[]变量,将其写入Cassandra中的blob字段(v1.2,通过新的Datastax CQL库),然后再次读取 当我把它放进去的时候,它有3个元素长,当我把它读回来的时候,它有84个元素长。。。!这意味着我实际尝试做的事情(序列化Java对象)失败,再次尝试反序列化时出现org.apache.commons.lang.SerializationException:Java.io.StreamCorruptedException:i
org.apache.commons.lang.SerializationException:Java.io.StreamCorruptedException:invalid stream header:81000008
错误
下面是一些示例代码,演示了我的问题:
import java.nio.ByteBuffer;
import org.apache.commons.lang.SerializationUtils;
import com.datastax.driver.core.BoundStatement;
import com.datastax.driver.core.Cluster;
import com.datastax.driver.core.Host;
import com.datastax.driver.core.Metadata;
import com.datastax.driver.core.PreparedStatement;
import com.datastax.driver.core.ResultSet;
import com.datastax.driver.core.Row;
import com.datastax.driver.core.Session;
public class TestCassandraSerialization {
private Cluster cluster;
private Session session;
public TestCassandraSerialization(String node) {
connect(node);
}
private void connect(String node) {
cluster = Cluster.builder().addContactPoint(node).build();
Metadata metadata = cluster.getMetadata();
System.out.printf("Connected to %s\n", metadata.getClusterName());
for (Host host: metadata.getAllHosts()) {
System.out.printf("Datacenter: %s; Host: %s; Rack: %s\n",
host.getDatacenter(), host.getAddress(), host.getRack());
}
session = cluster.connect();
}
public void setUp() {
session.execute("CREATE KEYSPACE test_serialization WITH replication = {'class':'SimpleStrategy', 'replication_factor':1};");
session.execute("CREATE TABLE test_serialization.test_table (id text PRIMARY KEY, data blob)");
}
public void tearDown() {
session.execute("DROP KEYSPACE test_serialization");
}
public void insertIntoTable(String key, byte[] data) {
PreparedStatement statement = session.prepare("INSERT INTO test_serialization.test_table (id,data) VALUES (?, ?)");
BoundStatement boundStatement = new BoundStatement(statement);
session.execute(boundStatement.bind(key,ByteBuffer.wrap(data)));
}
public byte[] readFromTable(String key) {
String q1 = "SELECT * FROM test_serialization.test_table WHERE id = '"+key+"';";
ResultSet results = session.execute(q1);
for (Row row : results) {
ByteBuffer data = row.getBytes("data");
return data.array();
}
return null;
}
public static boolean compareByteArrays(byte[] one, byte[] two) {
if (one.length > two.length) {
byte[] foo = one;
one = two;
two = foo;
}
// so now two is definitely the longer array
for (int i=0; i<one.length; i++) {
//System.out.printf("%d: %s\t%s\n", i, one[i], two[i]);
if (one[i] != two[i]) {
return false;
}
}
return true;
}
public static void main(String[] args) {
TestCassandraSerialization tester = new TestCassandraSerialization("localhost");
try {
tester.setUp();
byte[] dataIn = new byte[]{1,2,3};
tester.insertIntoTable("123", dataIn);
byte[] dataOut = tester.readFromTable("123");
System.out.println(dataIn);
System.out.println(dataOut);
System.out.println(dataIn.length); // prints "3"
System.out.println(dataOut.length); // prints "84"
System.out.println(compareByteArrays(dataIn, dataOut)); // prints false
String toSave = "Hello, world!";
dataIn = SerializationUtils.serialize(toSave);
tester.insertIntoTable("toSave", dataIn);
dataOut = tester.readFromTable("toSave");
System.out.println(dataIn.length); // prints "20"
System.out.println(dataOut.length); // prints "104"
// The below throws org.apache.commons.lang.SerializationException: java.io.StreamCorruptedException: invalid stream header: 81000008
String hasLoaded = (String) SerializationUtils.deserialize(dataOut);
System.out.println(hasLoaded);
} finally {
tester.tearDown();
}
}
}
因此,在读取而不是写入二进制数据时,它看起来像是一个错误。有人能告诉我我做错了什么吗?问题几乎可以肯定,因为ByteBuffer.array()返回的数组是完全支持数组,但数据可能只包含在其中的一部分 返回的有效数据从ByteBuffer.arrayOffset()开始,长度为ByteBuffer.remaining()。要获取仅包含有效数据的字节数组,请在readFromTable中使用以下代码:
byte[] result = new byte[data.remaining()];
data.get(result);
然后您的数据就是结果,您可以返回该结果。因为您已经使用了DataStax Java驱动程序,所以在
com.DataStax.Driver.core.utils中也有一个类似于:
byte[] result = Bytes.getArray(data)
谢谢你,理查德。你也救了我一天:)回答得好,你救了我一晚!我刚碰到这个。谢谢你和wtf?他们在设计什么样的API?
byte[] result = Bytes.getArray(data)