Github user shubham-pathak22 commented on a diff in the pull request:
https://github.com/apache/incubator-apex-malhar/pull/157#discussion_r48948897
--- Diff:
library/src/main/java/com/datatorrent/lib/async/AbstractAsyncProcessor.java ---
@@ -0,0 +1,357 @@
+/**
+ * Licensed to the Apache Software Foundation (ASF) under one
+ * or more contributor license agreements. See the NOTICE file
+ * distributed with this work for additional information
+ * regarding copyright ownership. The ASF licenses this file
+ * to you 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.datatorrent.lib.async;
+
+import java.util.Iterator;
+import java.util.Queue;
+import java.util.concurrent.ExecutorService;
+import java.util.concurrent.Executors;
+import java.util.concurrent.Future;
+
+import org.slf4j.Logger;
+import org.slf4j.LoggerFactory;
+
+import com.google.common.collect.Lists;
+
+import com.datatorrent.api.Context;
+import com.datatorrent.api.DefaultInputPort;
+import com.datatorrent.api.Operator.IdleTimeHandler;
+import com.datatorrent.api.annotation.InputPortFieldAnnotation;
+import com.datatorrent.common.util.BaseOperator;
+
+/**
+ * This base operator having a single input and tuple coming from input
port are enqueued for processing asynchronously
+ * by Thread Executors.
+ *
+ * User can configure number of of threads that are required for
processing. Defaulting to single thread which is
+ * different than operator thread.
+ * The user will also configure the timeout for the operations to happen.
+ *
+ * <b>Workflow is as follows:</b>
+ * <ol>
+ * <li>Tuple of type INPUT is received on input port.</li>
+ * <li>Tuple gets enqueued into Executors by calling
<i>enqueueTupleForProcessing</i>.
+ * Once can override <i>processTuple</i> method in which case, its
users responsibility to enqueue tuple for
+ * processing by calling <i>enqueueTupleForProcessing</i>.</li>
+ * <li>When tuple gets its turn of execution, <i>processTupleAsync</i>
gets called from thread other than
+ * operator thread.</li>
+ * <li>Based on maintainTupleOrder parameter, the results are
consolidated at end of every window and respective call
+ * to <i>handleProcessedTuple</i> will be made. This call will
happen in operator thread.</li>
+ * </ol>
+ *
+ * This operator can be implemented to do async and parallel operations.
+ *
+ * Use case examples:
+ * <ul>
+ * <li>Parallel reads from external datastore/database system.</li>
+ * <li>Doing any operations which are long running tasks per tuple but
DAG io should not be blocked.</li>
+ * </ul>
+ *
+ * @param <INPUT> input type of tuple
+ * @param <RESULT> Result of async processing of tuple.
+ * @since 3.3.0
+ */
+public abstract class AbstractAsyncProcessor<INPUT, RESULT> extends
BaseOperator implements IdleTimeHandler
+{
+ private static final Logger logger =
LoggerFactory.getLogger(AbstractAsyncProcessor.class);
+
+ private transient ExecutorService executor;
+ private int numProcessors = 1;
+ private boolean maintainTupleOrder = false;
+ private long processTimeoutMs = 0;
+
+ /**
+ * This Queue will hold all the tuples that are in process, has
completed processing but yet to be notified for
+ * processing completion.
+ */
+ private Queue<ProcessTuple> waitingTuples = Lists.newLinkedList();
+
+ @InputPortFieldAnnotation(optional = true)
+ public final transient DefaultInputPort<INPUT> input = new
DefaultInputPort<INPUT>()
+ {
+ @Override public void process(INPUT input)
+ {
+ processTuple(input);
+ }
+
+ @Override public void setup(Context.PortContext context)
+ {
+ setupInputPort(context);
+ }
+ };
+
+ /**
+ * This method is exposed for implementing class so as to take care of
input port based initialization.
+ * For eg. Retrieve TUPLE_CLASS attribute from input port.
+ *
+ * @param context PortContext of input port
+ */
+ protected void setupInputPort(Context.PortContext context)
+ {
+ // Do nothing. Concrete class may choose to do override this. For eg.
reading TUPLE_CLASS attribute.
+ }
+
+ /**
+ * Setup method of this operator will initialize the various types of
executors possible.
+ *
+ * @param context OperatorContext
+ */
+ @Override public void setup(Context.OperatorContext context)
+ {
+ if (numProcessors == 1) {
+ this.executor = Executors.newSingleThreadExecutor();
+ } else if (numProcessors == 0) {
+ this.executor = Executors.newCachedThreadPool();
+ } else if (numProcessors > 0) {
+ this.executor = Executors.newFixedThreadPool(numProcessors);
--- End diff --
Do we need to distinguish between Executors.newSingleThreadExecutor() and
Executors.newFixedThreadPool(numProcessors) with numProcessors = 1 ? Would it
be functionally different ?
---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and wishes so, or if the feature is enabled but not working, please
contact infrastructure at [email protected] or file a JIRA ticket
with INFRA.
---
