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]

Reply via email to