Many resources are needed to download a project. Please understand that we have to compensate our server costs. Thank you in advance. Project price only 1 $
You can buy this project and download/modify it how often you want.
/*
* Copyright (c) 2008-2020, Hazelcast, Inc. All Rights Reserved.
*
* Licensed under the Apache License, Version 2.0 (the "License");
* you may not use this file except in compliance with the License.
* You may obtain a copy of the License at
*
* http://www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an "AS IS" BASIS,
* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*/
package com.hazelcast.internal.journal;
import com.hazelcast.nio.ObjectDataInput;
import com.hazelcast.nio.ObjectDataOutput;
import com.hazelcast.nio.serialization.IdentifiedDataSerializable;
import com.hazelcast.ringbuffer.impl.ReadResultSetImpl;
import com.hazelcast.spi.impl.operationservice.BlockingOperation;
import com.hazelcast.internal.services.DistributedObjectNamespace;
import com.hazelcast.spi.impl.operationservice.Operation;
import com.hazelcast.spi.impl.operationservice.PartitionAwareOperation;
import com.hazelcast.spi.impl.operationservice.ReadonlyOperation;
import com.hazelcast.spi.impl.operationservice.WaitNotifyKey;
import java.io.IOException;
/**
* Reads from the map event journal in batches. You may specify the start sequence,
* the minumum required number of items in the response, the maximum number of items
* in the response, a predicate that the events should pass and a projection to
* apply to the events in the journal.
* If the event journal currently contains less events than the required minimum, the
* call will wait until it has sufficient items.
* The predicate, filter and projection may be {@code null} in which case all elements are returned
* and no projection is applied.
*
* @param the return type of the projection. It is equal to the journal event type
* if the projection is {@code null} or it is the identity projection
* @param journal event type
* @since 3.9
*/
public abstract class EventJournalReadOperation extends Operation
implements IdentifiedDataSerializable, PartitionAwareOperation, BlockingOperation, ReadonlyOperation {
protected String name;
protected int minSize;
protected int maxSize;
protected long startSequence;
protected transient ReadResultSetImpl resultSet;
protected transient long sequence;
protected transient DistributedObjectNamespace namespace;
private WaitNotifyKey waitNotifyKey;
public EventJournalReadOperation() {
}
public EventJournalReadOperation(String name, long startSequence, int minSize, int maxSize) {
this.name = name;
this.minSize = minSize;
this.maxSize = maxSize;
this.startSequence = startSequence;
}
/**
* {@inheritDoc}
* Checks the precondition that the start sequence is already
* available (in the event journal) or is the sequence of the
* next event to be added into the journal.
*/
@Override
public void beforeRun() {
namespace = new DistributedObjectNamespace(getServiceName(), name);
final EventJournal journal = getJournal();
if (!journal.hasEventJournal(namespace)) {
throw new UnsupportedOperationException(
"Cannot subscribe to event journal because it is either not configured or disabled for " + namespace);
}
final int partitionId = getPartitionId();
journal.cleanup(namespace, partitionId);
startSequence = clampToBounds(journal, partitionId, startSequence);
journal.isAvailableOrNextSequence(namespace, partitionId, startSequence);
// we'll store the wait notify key because ICache destroys the record store
// and the cache config is unavailable at the time operations are being
// cancelled. Hence, we cannot create the journal and fetch it's wait notify
// key
waitNotifyKey = journal.getWaitNotifyKey(namespace, partitionId);
}
/**
* {@inheritDoc}
* On every invocation this method reads from the event journal until
* it has collected the minimum required number of response items.
* Returns {@code true} if there are currently not enough
* elements in the response and the operation should be parked.
*
* @return if the operation should wait on the wait/notify key
*/
@Override
public boolean shouldWait() {
if (resultSet == null) {
resultSet = createResultSet();
sequence = startSequence;
}
final EventJournal journal = getJournal();
final int partitionId = getPartitionId();
journal.cleanup(namespace, partitionId);
sequence = clampToBounds(journal, partitionId, sequence);
if (minSize == 0) {
if (!journal.isNextAvailableSequence(namespace, partitionId, sequence)) {
readMany(journal, partitionId);
}
return false;
}
if (resultSet.isMinSizeReached()) {
// enough items have been read, we are done.
return false;
}
if (journal.isNextAvailableSequence(namespace, partitionId, sequence)) {
// the sequence is not readable
return true;
}
readMany(journal, partitionId);
return !resultSet.isMinSizeReached();
}
private void readMany(EventJournal journal, int partitionId) {
sequence = journal.readMany(namespace, partitionId, sequence, resultSet);
resultSet.setNextSequenceToReadFrom(sequence);
}
@Override
public void run() throws Exception {
// no-op; we already did the work in the shouldWait method.
}
@Override
public Object getResponse() {
return resultSet;
}
@Override
public WaitNotifyKey getWaitKey() {
return waitNotifyKey;
}
@Override
public void onWaitExpire() {
}
@Override
protected void writeInternal(ObjectDataOutput out) throws IOException {
super.writeInternal(out);
out.writeUTF(name);
out.writeInt(minSize);
out.writeInt(maxSize);
out.writeLong(startSequence);
}
@Override
protected void readInternal(ObjectDataInput in) throws IOException {
super.readInternal(in);
name = in.readUTF();
minSize = in.readInt();
maxSize = in.readInt();
startSequence = in.readLong();
}
public abstract String getServiceName();
protected abstract ReadResultSetImpl createResultSet();
protected abstract EventJournal getJournal();
/**
* Checks if the provided {@code requestedSequence} is within bounds of the
* oldest and newest sequence in the event journal. If the
* {@code requestedSequence} is too old or too new, it will return the
* current oldest or newest journal sequence.
* This method can be used for a loss-tolerant reader when trying to avoid a
* {@link com.hazelcast.ringbuffer.StaleSequenceException}.
*
* @param journal the event journal
* @param partitionId the partition ID to read
* @param requestedSequence the requested sequence to read
* @return the bounded journal sequence
*/
private long clampToBounds(EventJournal journal, int partitionId, long requestedSequence) {
final long oldestSequence = journal.oldestSequence(namespace, partitionId);
final long newestSequence = journal.newestSequence(namespace, partitionId);
// fast forward if late and no store is configured
if (requestedSequence < oldestSequence && !journal.isPersistenceEnabled(namespace, partitionId)) {
return oldestSequence;
}
// jump back if too far in future
if (requestedSequence > newestSequence + 1) {
return newestSequence + 1;
}
return requestedSequence;
}
}
