tragicjun commented on a change in pull request #6495: [FLINK-10061] [table]
[kafka] Fix unsupported reconfiguration in KafkaTableSink
URL: https://github.com/apache/flink/pull/6495#discussion_r207925584
##########
File path:
flink-connectors/flink-connector-kafka-base/src/main/java/org/apache/flink/streaming/connectors/kafka/KafkaTableSink.java
##########
@@ -166,8 +167,17 @@ public void emitDataStream(DataStream<Row> dataStream) {
@Override
public KafkaTableSink configure(String[] fieldNames,
TypeInformation<?>[] fieldTypes) {
if (schema.isPresent()) {
- // a fixed schema is defined so reconfiguration is not
supported
- throw new
UnsupportedOperationException("Reconfiguration of this sink is not supported.");
+ if (getFieldTypes().length != fieldTypes.length ||
!Arrays.deepEquals(getFieldTypes(), fieldTypes)) {
+ String fixedFieldTypes =
Arrays.deepToString(Arrays.stream(getFieldTypes())
+ .map(s ->
s.getTypeClass().getSimpleName()).toArray());
Review comment:
Agreed, I was thinking too much, `Arrays.toString()` should be fine. Please
review the latest file changes.
As for the checking of the field names, I would like to keep consistent with
`TableEnvironment.insertInto()` which only checks field types between source
table and table sink, considering field names are really irrelevant.
----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
[email protected]
With regards,
Apache Git Services
