X-Git-Url: https://juplo.de/gitweb/?a=blobdiff_plain;f=src%2Fmain%2Fjava%2Fde%2Fjuplo%2Fkafka%2Fwordcount%2Fsplitter%2FSplitterStreamProcessor.java;h=66188a719e2815887ae4b9585c02bc25611a9103;hb=520528638e7487c845d7fb0f39066ce7d249297b;hp=3eca8f4cef62a7584b2a930ff11e81db2f56312d;hpb=e8eee1b8a20b4aeadcb5384dd8473bc8a543a259;p=demos%2Fkafka%2Fwordcount

diff --git a/src/main/java/de/juplo/kafka/wordcount/splitter/SplitterStreamProcessor.java b/src/main/java/de/juplo/kafka/wordcount/splitter/SplitterStreamProcessor.java
index 3eca8f4..66188a7 100644
--- a/src/main/java/de/juplo/kafka/wordcount/splitter/SplitterStreamProcessor.java
+++ b/src/main/java/de/juplo/kafka/wordcount/splitter/SplitterStreamProcessor.java
@@ -9,6 +9,9 @@ import org.apache.kafka.streams.StreamsConfig;
 import org.apache.kafka.streams.kstream.KStream;
 import org.springframework.boot.SpringApplication;
 import org.springframework.context.ConfigurableApplicationContext;
+import org.springframework.kafka.support.serializer.JsonDeserializer;
+import org.springframework.kafka.support.serializer.JsonSerde;
+import org.springframework.kafka.support.serializer.JsonSerializer;
 import org.springframework.stereotype.Component;
 
 import jakarta.annotation.PostConstruct;
@@ -25,7 +28,7 @@ import static org.apache.kafka.streams.errors.StreamsUncaughtExceptionHandler.St
 @Component
 public class SplitterStreamProcessor
 {
-	final static Pattern PATTERN = Pattern.compile("\\W+");
+	final static Pattern PATTERN = Pattern.compile("[^\\p{IsAlphabetic}]+");
 
 	public final KafkaStreams streams;
 
@@ -36,16 +39,23 @@ public class SplitterStreamProcessor
 	{
 		StreamsBuilder builder = new StreamsBuilder();
 
-		KStream<String, String> source = builder.stream(properties.getInputTopic());
+		KStream<String, Recording> source = builder.stream(properties.getInputTopic());
+
 		source
-				.flatMapValues(sentence -> Arrays.asList(PATTERN.split(sentence)))
+				.flatMapValues(recording -> Arrays
+							.stream(PATTERN.split(recording.getSentence()))
+							.map(word -> Word.of(recording.getUser(), word))
+							.toList())
 				.to(properties.getOutputTopic());
 
 		Properties props = new Properties();
 		props.put(StreamsConfig.APPLICATION_ID_CONFIG, properties.getApplicationId());
 		props.put(StreamsConfig.BOOTSTRAP_SERVERS_CONFIG, properties.getBootstrapServer());
-		props.put(StreamsConfig.DEFAULT_KEY_SERDE_CLASS_CONFIG, Serdes.String().getClass().getName());
-		props.put(StreamsConfig.DEFAULT_VALUE_SERDE_CLASS_CONFIG, Serdes.String().getClass().getName());
+		props.put(StreamsConfig.DEFAULT_KEY_SERDE_CLASS_CONFIG, Serdes.StringSerde.class.getName());
+		props.put(StreamsConfig.DEFAULT_VALUE_SERDE_CLASS_CONFIG, JsonSerde.class.getName());
+		props.put(JsonDeserializer.TRUSTED_PACKAGES, Recording.class.getName()	);
+		props.put(JsonDeserializer.VALUE_DEFAULT_TYPE, Recording.class.getName());
+		props.put(JsonSerializer.ADD_TYPE_INFO_HEADERS, false);
 		props.put(ConsumerConfig.AUTO_OFFSET_RESET_CONFIG, "earliest");
 
 		streams = new KafkaStreams(builder.build(), props);