Github user xccui commented on a diff in the pull request:
https://github.com/apache/flink/pull/5662#discussion_r173879267
--- Diff:
flink-connectors/flink-connector-kafka-base/src/test/java/org/apache/flink/streaming/connectors/kafka/KafkaJsonTableSourceFactoryTestBase.java
---
@@ -89,9 +94,10 @@ private void testTableSource(FormatDescriptor format) {
// construct table source using a builder
final Map<String, String> tableJsonMapping = new HashMap<>();
+ tableJsonMapping.put("name", "name");
--- End diff --
Well, according to the current implementation, you are right. But I still
feel uncomfortable about that since we actually mix the physical schema (format
schema) and the logical schema (table schema) into the same map. Do you think
it's necessary to make some changes here?
---
