leonardBang commented on code in PR #4387: URL: https://github.com/apache/flink-cdc/pull/4387#discussion_r3308121695
########## flink-cdc-connect/flink-cdc-source-connectors/flink-connector-postgres-cdc/src/main/java/org/apache/flink/cdc/connectors/postgres/source/reader/PostgresSourceStreamFetcher.java: ########## @@ -0,0 +1,58 @@ +/* + * 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 org.apache.flink.cdc.connectors.postgres.source.reader; + +import org.apache.flink.cdc.connectors.base.source.reader.external.FetchTask; +import org.apache.flink.cdc.connectors.base.source.reader.external.IncrementalSourceStreamFetcher; + +import io.debezium.data.Envelope; +import org.apache.kafka.connect.data.Struct; +import org.apache.kafka.connect.source.SourceRecord; + +/** + * Stream fetcher for Postgres. When {@code includeLogicalMessages} is enabled, lets {@code + * pg_logical_emit_message} records (op="m") bypass table-based watermark filtering, since logical + * messages are not bound to a table. + */ +public class PostgresSourceStreamFetcher extends IncrementalSourceStreamFetcher { + + private final boolean includeLogicalMessages; + + public PostgresSourceStreamFetcher( + FetchTask.Context taskContext, int subtaskId, boolean includeLogicalMessages) { + super(taskContext, subtaskId); + this.includeLogicalMessages = includeLogicalMessages; + } + + @Override + protected boolean shouldEmit(SourceRecord sourceRecord) { + if (includeLogicalMessages && isLogicalMessage(sourceRecord)) { + return true; + } + return super.shouldEmit(sourceRecord); + } + + static boolean isLogicalMessage(SourceRecord record) { + if (record.value() instanceof Struct) { + Struct struct = (Struct) record.value(); Review Comment: Regarding the configuration design: if providing logical message prefixes is mandatory for PG logical decoding to work correctly (i.e., every job that writes pg_logical_emit_message must specify a prefix to differentiate subscriptions), I'd suggest we can remove *scan.logical-message.enabled* entirely and use the presence of *scan.logical-message.prefixes* as the implicit enablement signal. When *scan.logical-message.prefixes* is set (non-empty), the feature is implicitly enabled — the user clearly intends to consume logical messages When *scan.logical-message.prefixes* is empty or absent, no logical messages are consumed — default safe behavior preserved This eliminates the risk of misconfiguration where enabled=true but prefixes is accidentally left empty (which would consume ALL logical messages across subscriptions — a potentially dangerous default) It removes one config knob without losing expressiveness. ```java // PostgresSourceFetchTaskContext.shouldEmit() if (isLogicalMessage(record)) { // Feature is implicitly enabled when prefixes are configured return logicalMessagePrefixes.isEmpty() ? false : logicalMessagePrefixes.stream().anyMatch(prefix -> matchesPrefix(record, prefix)); } ``` The PostgresSourceBuilder API could expose a single method: ```java builder.includeLogicalMessages(List<String> prefixes) ``` where passing a non-empty list enables the feature, and omitting the call or passing an empty list keeps it disabled. WDYT? @loserwang1024 @eskabetxe -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: [email protected] For queries about this service, please contact Infrastructure at: [email protected]
