|
| 1 | +/** |
| 2 | +* Copyright (C) 2018 Lightbend Inc. <https://www.lightbend.com> |
| 3 | +*/ |
| 4 | + |
| 5 | +package com.lightbend.kafka.scala.streams |
| 6 | + |
| 7 | +import java.util.Properties |
| 8 | +import java.util.regex.Pattern |
| 9 | + |
| 10 | +import com.lightbend.kafka.scala.server.{KafkaLocalServer, MessageListener, MessageSender, RecordProcessorTrait} |
| 11 | +import minitest.TestSuite |
| 12 | +import org.apache.kafka.clients.consumer.ConsumerRecord |
| 13 | +import org.apache.kafka.common.serialization._ |
| 14 | +import org.apache.kafka.streams.{KafkaStreams, KeyValue, StreamsConfig} |
| 15 | +import ImplicitConversions._ |
| 16 | +import com.typesafe.scalalogging.LazyLogging |
| 17 | + |
| 18 | +object KafkaStreamsMergeTest extends TestSuite[KafkaLocalServer] with WordCountMergeTestData with LazyLogging { |
| 19 | + |
| 20 | +override def setup(): KafkaLocalServer = { |
| 21 | +val s = KafkaLocalServer(cleanOnStart = true, Some(localStateDir)) |
| 22 | +s.start() |
| 23 | +s |
| 24 | +} |
| 25 | + |
| 26 | +override def tearDown(server: KafkaLocalServer): Unit = { |
| 27 | +server.stop() |
| 28 | +} |
| 29 | + |
| 30 | +test("should count words") { server => |
| 31 | + |
| 32 | +server.createTopic(inputTopic1) |
| 33 | +server.createTopic(inputTopic2) |
| 34 | +server.createTopic(outputTopic) |
| 35 | + |
| 36 | +// |
| 37 | +// Step 1: Configure and start the processor topology. |
| 38 | +// |
| 39 | +import DefaultSerdes._ |
| 40 | + |
| 41 | +val streamsConfiguration = new Properties() |
| 42 | +streamsConfiguration.put(StreamsConfig.APPLICATION_ID_CONFIG, s"wordcount-${scala.util.Random.nextInt(100)}") |
| 43 | +streamsConfiguration.put(StreamsConfig.CLIENT_ID_CONFIG, "wordcountgroup") |
| 44 | + |
| 45 | +streamsConfiguration.put(StreamsConfig.BOOTSTRAP_SERVERS_CONFIG, brokers) |
| 46 | +streamsConfiguration.put(StreamsConfig.STATE_DIR_CONFIG, localStateDir) |
| 47 | + |
| 48 | +val builder = new StreamsBuilderS() |
| 49 | + |
| 50 | +val textLines1 = builder.stream[String, String](inputTopic1) |
| 51 | +val textLines2 = builder.stream[String, String](inputTopic2) |
| 52 | + |
| 53 | +val textLines = textLines1.merge(textLines2) |
| 54 | + |
| 55 | +val pattern = Pattern.compile("\\W+", Pattern.UNICODE_CHARACTER_CLASS) |
| 56 | + |
| 57 | +val wordCounts: KTableS[String, Long] = |
| 58 | +textLines.flatMapValues(v => pattern.split(v.toLowerCase)) |
| 59 | +.groupBy((k, v) => v) |
| 60 | +.count() |
| 61 | + |
| 62 | +wordCounts.toStream.to(outputTopic) |
| 63 | + |
| 64 | +val streams = new KafkaStreams(builder.build(), streamsConfiguration) |
| 65 | +streams.start() |
| 66 | + |
| 67 | +// |
| 68 | +// Step 2: Produce some input data to the input topics. |
| 69 | +// |
| 70 | +val sender = MessageSender[String, String](brokers, classOf[StringSerializer].getName, classOf[StringSerializer].getName) |
| 71 | +val mvals1 = sender.batchWriteValue(inputTopic1, inputValues) |
| 72 | +val mvals2 = sender.batchWriteValue(inputTopic2, inputValues) |
| 73 | + |
| 74 | +// |
| 75 | +// Step 3: Verify the application's output data. |
| 76 | +// |
| 77 | +val listener = MessageListener(brokers, outputTopic, "wordcountgroup", |
| 78 | +classOf[StringDeserializer].getName, |
| 79 | +classOf[LongDeserializer].getName, |
| 80 | +new RecordProcessor |
| 81 | +) |
| 82 | + |
| 83 | +val l = listener.waitUntilMinKeyValueRecordsReceived(expectedWordCounts.size, 30000) |
| 84 | + |
| 85 | +assertEquals(l.sortBy(_.key), expectedWordCounts.sortBy(_.key)) |
| 86 | + |
| 87 | +streams.close() |
| 88 | +} |
| 89 | + |
| 90 | +class RecordProcessor extends RecordProcessorTrait[String, Long] { |
| 91 | +override def processRecord(record: ConsumerRecord[String, Long]): Unit = { |
| 92 | +// logger.info(s"Get Message $record") |
| 93 | +} |
| 94 | +} |
| 95 | + |
| 96 | +} |
| 97 | + |
| 98 | +trait WordCountMergeTestData { |
| 99 | +val inputTopic1 = s"inputTopic1.${scala.util.Random.nextInt(100)}" |
| 100 | +val inputTopic2 = s"inputTopic2.${scala.util.Random.nextInt(100)}" |
| 101 | +val outputTopic = s"outputTpic.${scala.util.Random.nextInt(100)}" |
| 102 | +val brokers = "localhost:9092" |
| 103 | +val localStateDir = "local_state_data" |
| 104 | + |
| 105 | +val inputValues = List( |
| 106 | +"Hello Kafka Streams", |
| 107 | +"All streams lead to Kafka", |
| 108 | +"Join Kafka Summit", |
| 109 | +"И теперь пошли русские слова" |
| 110 | +) |
| 111 | + |
| 112 | +val expectedWordCounts: List[KeyValue[String, Long]] = List( |
| 113 | +new KeyValue("hello", 2L), |
| 114 | +new KeyValue("all", 2L), |
| 115 | +new KeyValue("streams", 4L), |
| 116 | +new KeyValue("lead", 2L), |
| 117 | +new KeyValue("to", 2L), |
| 118 | +new KeyValue("join", 2L), |
| 119 | +new KeyValue("kafka", 6L), |
| 120 | +new KeyValue("summit", 2L), |
| 121 | +new KeyValue("и", 2L), |
| 122 | +new KeyValue("теперь", 2L), |
| 123 | +new KeyValue("пошли", 2L), |
| 124 | +new KeyValue("русские", 2L), |
| 125 | +new KeyValue("слова", 2L) |
| 126 | +) |
| 127 | +} |
| 128 | + |
| 129 | + |
0 commit comments