I plan to use ByteArrayOutputStream to write Objects to byte array, then write the byte array to a file. I keep the position of each Object. Then I read from the file, using FileInputStream.getChannel().position() to relocate the position, and call ObjectInputStream.readObject().
Here is my test code:
public static void main(String[] args) {
try {
HashBucketTest bucket1 = new HashBucketTest();
bucket1.putAddress("0", 30);
HashBucketTest bucket2 = new HashBucketTest(); // bucket2 key is the same as bucket3
bucket2.putAddress("22313", 40);
HashBucketTest bucket3 = new HashBucketTest();
bucket3.putAddress("22313", 50);
ByteArrayOutputStream baos = new ByteArrayOutputStream();
ObjectOutputStream oops = new ObjectOutputStream(baos);
FileOutputStream fout = new FileOutputStream("objects.txt");
BufferedOutputStream bs = new BufferedOutputStream(fout);
int position1 = baos.size(); // the first object position
bs.write(baos.toByteArray()); // write head
baos.reset();
oops.writeObject(bucket1); // write the first object to byte array
int position2 = position1 + baos.size(); // the second object position
bs.write(baos.toByteArray()); // write first object to file
baos.reset();
oops.writeObject(bucket2); // write the second object to byte array
int position3 = position2 + baos.size(); // the third object position
bs.write(baos.toByteArray()); // write the second object to file
baos.reset();
oops.writeObject(bucket3); // write the third object to byte array
int position4 = position3 + baos.size(); // the fourth object position
bs.write(baos.toByteArray()); // write the third object to file
baos.reset();
bs.flush();
bs.close();
fout.flush();
fout.close();
// read according to the position x
HashBucketTest readCase1 = null;
HashBucketTest readCase2 = null;
HashBucketTest readCase3 = null;
FileInputStream fis = new FileInputStream("objects.txt");
ObjectInputStream ois = new ObjectInputStream(
fis);
// success case
readCase1 = (HashBucketTest) ois.readObject(); // read the first object, success
fis.getChannel().position(position2);
readCase2 = (HashBucketTest) ois.readObject(); // read the second object, success
fis.getChannel().position(position3);
readCase3 = (HashBucketTest) ois.readObject(); // read the third object, success
// failed case!!!!
//readCase1 = (HashBucketTest) ois.readObject(); // read the first object, success
//fis.getChannel().position(position3);
//readCase3 = (HashBucketTest) ois.readObject(); // read the third object, failed!!
} catch (Exception e) {
// TODO Auto-generated catch block
e.printStackTrace();
}
}
public static class HashBucketTest implements Serializable {
private static final long serialVersionUID = 3610182543890121796L;
Map<String, Integer> values = null; // need to write to disk
public HashBucketTest() {
values = new TreeMap<String, Integer>();
}
public void putAddress(String key, int number) {
values.put(key, number);
}
}
My problem is, when the key of "bucket2" is the same as "bucket3"( "22313" in this case). I can't read bucket3 skippting bucket2
java.io.StreamCorruptedException: invalid handle value: 007E000C
at java.io.ObjectInputStream.readHandle(ObjectInputStream.java:1456)
at java.io.ObjectInputStream.readObject0(ObjectInputStream.java:1331)
at java.io.ObjectInputStream.readObject(ObjectInputStream.java:371)
at java.util.TreeMap.buildFromSorted(TreeMap.java:2563)
at java.util.TreeMap.buildFromSorted(TreeMap.java:2504)
at java.util.TreeMap.readObject(TreeMap.java:2450)
at sun.reflect.NativeMethodAccessorImpl.invoke0(Native Method)
at sun.reflect.NativeMethodAccessorImpl.invoke(NativeMethodAccessorImpl.java:62)
at sun.reflect.DelegatingMethodAccessorImpl.invoke(DelegatingMethodAccessorImpl.java:43)
at java.lang.reflect.Method.invoke(Method.java:497)
at java.io.ObjectStreamClass.invokeReadObject(ObjectStreamClass.java:1017)
at java.io.ObjectInputStream.readSerialData(ObjectInputStream.java:1896)
at java.io.ObjectInputStream.readOrdinaryObject(ObjectInputStream.java:1801)
at java.io.ObjectInputStream.readObject0(ObjectInputStream.java:1351)
at java.io.ObjectInputStream.defaultReadFields(ObjectInputStream.java:1993)
at java.io.ObjectInputStream.readSerialData(ObjectInputStream.java:1918)
at java.io.ObjectInputStream.readOrdinaryObject(ObjectInputStream.java:1801)
at java.io.ObjectInputStream.readObject0(ObjectInputStream.java:1351)
at java.io.ObjectInputStream.readObject(ObjectInputStream.java:371)
at com.alibaba.middleware.benchmark.Test.main(Test.java:146)
But if the key is different, I can do this successfully. It seems bucket3's key refer to the bucket2's? that's very strange. And in any case, sequential reading is ok, but I dont want sequential reading.
I suspect that ObjectInputStream
simply isn't designed for this sort of use. In order to maintain referential integrity, I suspect there's an internal counter for "object number X that I've written to the stream" which is increased as objects are read - then when there's a reference to that object, it can just be written as "reference number X". When you read the objects in a different order to the order in which they were written, that would mess things up significantly.
I would strongly advise you to basically not try to do this. (I'd also advise you to avoid regular Java binary serialization if at all possible, but that's a different matter.)
See more on this question at Stackoverflow