From commits-return-7559-archive-asf-public=cust-asf.ponee.io@pulsar.incubator.apache.org Wed May 2 09:20:40 2018 Return-Path: X-Original-To: archive-asf-public@cust-asf.ponee.io Delivered-To: archive-asf-public@cust-asf.ponee.io Received: from mail.apache.org (hermes.apache.org [140.211.11.3]) by mx-eu-01.ponee.io (Postfix) with SMTP id B752518065D for ; Wed, 2 May 2018 09:20:39 +0200 (CEST) Received: (qmail 22546 invoked by uid 500); 2 May 2018 07:20:38 -0000 Mailing-List: contact commits-help@pulsar.incubator.apache.org; run by ezmlm Precedence: bulk List-Help: List-Unsubscribe: List-Post: List-Id: Reply-To: dev@pulsar.incubator.apache.org Delivered-To: mailing list commits@pulsar.incubator.apache.org Received: (qmail 22521 invoked by uid 99); 2 May 2018 07:20:38 -0000 Received: from ec2-52-202-80-70.compute-1.amazonaws.com (HELO gitbox.apache.org) (52.202.80.70) by apache.org (qpsmtpd/0.29) with ESMTP; Wed, 02 May 2018 07:20:38 +0000 From: GitBox To: commits@pulsar.apache.org Subject: [GitHub] sijie commented on a change in pull request #1708: Refactor functions to use Sink interface Message-ID: <152524563809.6450.3133097993455456274.gitbox@gitbox.apache.org> Date: Wed, 02 May 2018 07:20:38 -0000 Content-Type: text/plain; charset=utf-8 Content-Transfer-Encoding: 8bit sijie commented on a change in pull request #1708: Refactor functions to use Sink interface URL: https://github.com/apache/incubator-pulsar/pull/1708#discussion_r185410288 ########## File path: pulsar-functions/instance/src/main/java/org/apache/pulsar/functions/sink/PulsarSink.java ########## @@ -18,5 +18,249 @@ */ package org.apache.pulsar.functions.sink; -public class PulsarSink { +import com.google.common.annotations.VisibleForTesting; +import lombok.AccessLevel; +import lombok.Getter; +import lombok.extern.slf4j.Slf4j; +import net.jodah.typetools.TypeResolver; +import org.apache.pulsar.client.api.Consumer; +import org.apache.pulsar.client.api.ConsumerEventListener; +import org.apache.pulsar.client.api.Message; +import org.apache.pulsar.client.api.MessageBuilder; +import org.apache.pulsar.client.api.Producer; +import org.apache.pulsar.client.api.PulsarClient; +import org.apache.pulsar.client.api.PulsarClientException; +import org.apache.pulsar.connect.core.RecordContext; +import org.apache.pulsar.functions.api.SerDe; +import org.apache.pulsar.functions.api.utils.DefaultSerDe; +import org.apache.pulsar.functions.instance.InstanceUtils; +import org.apache.pulsar.functions.instance.producers.AbstractOneOuputTopicProducers; +import org.apache.pulsar.functions.instance.producers.MultiConsumersOneOuputTopicProducers; +import org.apache.pulsar.functions.instance.producers.Producers; +import org.apache.pulsar.functions.source.PulsarRecord; +import org.apache.pulsar.functions.utils.FunctionConfig; + +import java.util.Base64; +import java.util.Map; +import java.util.concurrent.CompletableFuture; + +@Slf4j +public class PulsarSink implements RuntimeSink { + + private PulsarClient client; + private PulsarSinkConfig pulsarSinkConfig; + private SerDe outputSerDe; + + private PulsarSinkProcessor pulsarSinkProcessor; + + private interface PulsarSinkProcessor { + void initializeOutputProducer(String outputTopic) throws Exception; + + void sendOutputMessage(MessageBuilder outputMsgBuilder, + PulsarRecord pulsarRecord) throws Exception; + + void close() throws Exception; + } + + private class PulsarSinkAtMostOnceProcessor implements PulsarSinkProcessor { + private Producer producer; + + @Override + public void initializeOutputProducer(String outputTopic) throws Exception { + this.producer = AbstractOneOuputTopicProducers.createProducer( + client, pulsarSinkConfig.getTopic()); + } + + @Override + public void sendOutputMessage(MessageBuilder outputMsgBuilder, + PulsarRecord pulsarRecord) throws Exception { + Message outputMsg = outputMsgBuilder.build(); + this.producer.sendAsync(outputMsg); + } + + @Override + public void close() throws Exception { + if (null != producer) { + try { + producer.close(); + } catch (PulsarClientException e) { + log.warn("Fail to close producer for processor {}", pulsarSinkConfig.getTopic(), e); + } + } + } + } + + private class PulsarSinkAtLeastOnceProcessor implements PulsarSinkProcessor { + private Producer producer; + + @Override + public void initializeOutputProducer(String outputTopic) throws Exception { + this.producer = AbstractOneOuputTopicProducers.createProducer( + client, pulsarSinkConfig.getTopic()); + } + + @Override + public void sendOutputMessage(MessageBuilder outputMsgBuilder, + PulsarRecord pulsarRecord) throws Exception { + Message outputMsg = outputMsgBuilder.build(); + this.producer.sendAsync(outputMsg).thenAccept(messageId -> pulsarRecord.ack()); + } + + @Override + public void close() throws Exception { + if (null != producer) { + try { + producer.close(); + } catch (PulsarClientException e) { + log.warn("Fail to close producer for processor {}", pulsarSinkConfig.getTopic(), e); + } + } + } + } + + private class PulsarSinkEffectivelyOnceProcessor implements PulsarSinkProcessor, ConsumerEventListener { + + @Getter(AccessLevel.PACKAGE) + protected Producers outputProducer; + + @Override + public void initializeOutputProducer(String outputTopic) throws Exception { + outputProducer = new MultiConsumersOneOuputTopicProducers(client, outputTopic); + outputProducer.initialize(); + } + + @Override + public void sendOutputMessage(MessageBuilder outputMsgBuilder, PulsarRecord pulsarRecord) + throws Exception { + + // assign sequence id to output message for idempotent producing + outputMsgBuilder = outputMsgBuilder + .setSequenceId(pulsarRecord.getRecordSequence()); + + // currently on PulsarRecord + Producer producer = outputProducer.getProducer(pulsarRecord.getTopicName(), + Integer.parseInt(pulsarRecord.getPartitionId())); Review comment: I think we can change `MultiConsumerOneOutputTopicProducers` to use `Map`, rather than converting string back to int ---------------------------------------------------------------- This is an automated message from the Apache Git Service. To respond to the message, please log on GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: users@infra.apache.org With regards, Apache Git Services