awelless commented on code in PR #11537:
URL: https://github.com/apache/nifi/pull/11537#discussion_r3811528828
##########
nifi-commons/nifi-record/src/test/java/org/apache/nifi/serialization/record/TestDataTypeUtils.java:
##########
@@ -1392,4 +1397,278 @@ public void
testMergeDataTypesWithManyDistinctRecordSchemasCompletesQuickly() {
assertTrue(finalSchema.getField("field_0").isPresent());
assertTrue(finalSchema.getField("field_4999").isPresent());
}
+
+ @Test
+ public void testMergeSchemasMakesSingleSideFieldsNullable() {
+ final RecordSchema schemaA = new SimpleRecordSchema(List.of(
+ new RecordField("id", RecordFieldType.STRING.getDataType(),
false)));
+ final RecordSchema schemaB = new SimpleRecordSchema(List.of(
+ new RecordField("name", RecordFieldType.STRING.getDataType(),
false)));
+
+ final RecordSchema merged = DataTypeUtils.merge(schemaA, schemaB);
+
+ assertEquals(2, merged.getFieldCount());
+ assertTrue(merged.getField("id").orElseThrow().isNullable());
+ assertTrue(merged.getField("name").orElseThrow().isNullable());
+ }
+
+ @Test
+ public void testMergeSchemasKeepsSharedNonNullableFieldNonNullable() {
+ final RecordSchema schemaA = new SimpleRecordSchema(List.of(
+ new RecordField("id", RecordFieldType.STRING.getDataType(),
false)));
+ final RecordSchema schemaB = new SimpleRecordSchema(List.of(
+ new RecordField("id", RecordFieldType.STRING.getDataType(),
false)));
+
+ final RecordSchema merged = DataTypeUtils.merge(schemaA, schemaB);
+
+ assertEquals(1, merged.getFieldCount());
+ assertFalse(merged.getField("id").orElseThrow().isNullable());
+ }
+
+ @Test
+ public void testMergeSchemasPreservesNullableWhenLeftIsNullable() {
+ final RecordSchema schemaA = new SimpleRecordSchema(List.of(
+ new RecordField("id", RecordFieldType.STRING.getDataType(),
true)));
+ final RecordSchema schemaB = new SimpleRecordSchema(List.of(
+ new RecordField("id", RecordFieldType.STRING.getDataType(),
false)));
+
+ final RecordSchema merged = DataTypeUtils.merge(schemaA, schemaB);
+
+ assertTrue(merged.getField("id").orElseThrow().isNullable());
+ }
+
+ @Test
+ public void testMergeSchemasPreservesNullableWhenRightIsNullable() {
+ final RecordSchema schemaA = new SimpleRecordSchema(List.of(
+ new RecordField("id", RecordFieldType.STRING.getDataType(),
false)));
+ final RecordSchema schemaB = new SimpleRecordSchema(List.of(
+ new RecordField("id", RecordFieldType.STRING.getDataType(),
true)));
+
+ final RecordSchema merged = DataTypeUtils.merge(schemaA, schemaB);
+
+ assertTrue(merged.getField("id").orElseThrow().isNullable());
+ }
+
+ @Test
+ public void testMergeSchemasWidensTypesAndNullifiesSingleSideFields() {
+ final RecordSchema schemaA = new SimpleRecordSchema(List.of(
+ new RecordField("id", RecordFieldType.INT.getDataType(), false),
+ new RecordField("onlyA", RecordFieldType.STRING.getDataType(),
false)));
+ final RecordSchema schemaB = new SimpleRecordSchema(List.of(
+ new RecordField("id", RecordFieldType.LONG.getDataType(), true),
+ new RecordField("onlyB", RecordFieldType.STRING.getDataType(),
Set.of("bAlias"), false)));
+
+ final RecordSchema merged = DataTypeUtils.merge(schemaA, schemaB);
+
+ assertEquals(3, merged.getFieldCount());
+ final RecordField idField = merged.getField("id").orElseThrow();
+ assertEquals(RecordFieldType.LONG,
idField.getDataType().getFieldType());
+ assertTrue(idField.isNullable());
+ assertTrue(merged.getField("onlyA").orElseThrow().isNullable());
+ assertTrue(merged.getField("onlyB").orElseThrow().isNullable());
+ }
+
+ @Test
+ public void testMergeSchemasIdenticalSchemasRetainNullability() {
+ final RecordSchema schemaA = new SimpleRecordSchema(List.of(
+ new RecordField("required", RecordFieldType.STRING.getDataType(),
false),
+ new RecordField("optional", RecordFieldType.INT.getDataType(),
true)));
+ final RecordSchema schemaB = new SimpleRecordSchema(List.of(
+ new RecordField("required", RecordFieldType.STRING.getDataType(),
false),
+ new RecordField("optional", RecordFieldType.INT.getDataType(),
true)));
+
+ final RecordSchema merged = DataTypeUtils.merge(schemaA, schemaB);
+
+ assertEquals(2, merged.getFieldCount());
+ assertFalse(merged.getField("required").orElseThrow().isNullable());
+ assertTrue(merged.getField("optional").orElseThrow().isNullable());
+ }
+
+ @Test
+ public void testMergeSchemasMakesNestedSingleSideFieldsNullable() {
+ final RecordSchema nestedA = new SimpleRecordSchema(List.of(
+ new RecordField("street", RecordFieldType.STRING.getDataType(),
false),
+ new RecordField("city", RecordFieldType.STRING.getDataType(),
false)));
+ final RecordSchema nestedB = new SimpleRecordSchema(List.of(
+ new RecordField("street", RecordFieldType.STRING.getDataType(),
false),
+ new RecordField("zip", RecordFieldType.STRING.getDataType(),
false)));
+
+ final RecordSchema schemaA = new SimpleRecordSchema(List.of(
+ new RecordField("address",
RecordFieldType.RECORD.getRecordDataType(nestedA), false)));
+ final RecordSchema schemaB = new SimpleRecordSchema(List.of(
+ new RecordField("address",
RecordFieldType.RECORD.getRecordDataType(nestedB), false)));
+
+ final RecordSchema merged = DataTypeUtils.merge(schemaA, schemaB);
+ final RecordField addressField =
merged.getField("address").orElseThrow();
+ assertFalse(addressField.isNullable());
+
+ final RecordSchema nestedMerged = ((RecordDataType)
addressField.getDataType()).getChildSchema();
+ assertEquals(3, nestedMerged.getFieldCount());
+
assertFalse(nestedMerged.getField("street").orElseThrow().isNullable());
+ assertTrue(nestedMerged.getField("city").orElseThrow().isNullable());
+ assertTrue(nestedMerged.getField("zip").orElseThrow().isNullable());
+ }
+
+ @Test
+ public void testMergeSchemasMakesArrayOfRecordSingleSideFieldsNullable() {
Review Comment:
Added `testMergeSchemasWhenOneSchemaIsStrictSubset`
##########
nifi-extension-bundles/nifi-kafka-bundle/nifi-kafka-processors/src/main/java/org/apache/nifi/kafka/processors/ConsumeKafka.java:
##########
@@ -274,6 +278,17 @@ public class ConsumeKafka extends AbstractProcessor
implements VerifiableProcess
.dependsOn(PROCESSING_STRATEGY, ProcessingStrategy.RECORD)
.build();
+ static final PropertyDescriptor SCHEMA_CONFLICT_RESOLUTION = new
PropertyDescriptor.Builder()
+ .name("Schema Conflict Resolution")
+ .description("Specifies how to handle records with different
schemas within the same topic and partition. "
Review Comment:
I changed the description to `Specifies how records with different schemas
are grouped into FlowFiles`.
I'm not sure we want to put `records with different schemas within the same
output FlowFile`, as it isn't necessary that records will be in the same output
FlowFile. Perhaps we could've used `records with different schemas within the
same batch`, but batch is something known internally only.
##########
nifi-extension-bundles/nifi-kafka-bundle/nifi-kafka-3-integration/src/test/java/org/apache/nifi/kafka/processors/ConsumeKafkaMergeSchemaIT.java:
##########
@@ -0,0 +1,310 @@
+/*
+ * 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.nifi.kafka.processors;
+
+import com.fasterxml.jackson.databind.JsonNode;
+import com.fasterxml.jackson.databind.node.ArrayNode;
+import org.apache.kafka.clients.admin.AdminClient;
+import org.apache.kafka.clients.admin.AdminClientConfig;
+import org.apache.kafka.clients.admin.NewTopic;
+import org.apache.kafka.clients.producer.ProducerRecord;
+import org.apache.kafka.common.header.Header;
+import org.apache.nifi.kafka.processors.consumer.ProcessingStrategy;
+import org.apache.nifi.kafka.service.api.consumer.AutoOffsetReset;
+import org.apache.nifi.kafka.shared.attribute.KafkaFlowFileAttribute;
+import org.apache.nifi.kafka.shared.property.OutputStrategy;
+import org.apache.nifi.kafka.shared.property.SchemaConflictResolution;
+import org.apache.nifi.reporting.InitializationException;
+import org.apache.nifi.util.MockFlowFile;
+import org.apache.nifi.util.TestRunner;
+import org.apache.nifi.util.TestRunners;
+import org.junit.jupiter.api.BeforeEach;
+import org.junit.jupiter.api.Test;
+
+import java.io.IOException;
+import java.util.List;
+import java.util.Map;
+import java.util.UUID;
+import java.util.concurrent.ExecutionException;
+
+import static org.junit.jupiter.api.Assertions.assertEquals;
+import static org.junit.jupiter.api.Assertions.assertInstanceOf;
+import static org.junit.jupiter.api.Assertions.assertTrue;
+
+class ConsumeKafkaMergeSchemaIT extends AbstractConsumeKafkaIT {
+
+ private static final int FIRST_PARTITION = 0;
+
+ private static final String RECORD_WITH_ID = """
+ { "id": 1 }
+ """;
+
+ private static final String RECORD_WITH_NAME = """
+ { "name": "Alice" }
+ """;
+
+ private static final String INVALID_RECORD = "not-valid-json";
+
+ private TestRunner runner;
+
+ @BeforeEach
+ void setRunner() throws InitializationException {
+ runner = TestRunners.newTestRunner(ConsumeKafka.class);
+ addKafkaConnectionService(runner);
+ runner.setProperty(ConsumeKafka.CONNECTION_SERVICE,
CONNECTION_SERVICE_ID);
+ addRecordReaderService(runner);
+ addRecordWriterService(runner);
+ }
+
+ @Test
+ void testMergedSchemaProducesSingleFlowFile() throws ExecutionException,
InterruptedException, IOException {
+ final String topic = UUID.randomUUID().toString();
+ final String groupId = topic.substring(0, topic.indexOf("-"));
+
+ runner.setProperty(ConsumeKafka.GROUP_ID, groupId);
+ runner.setProperty(ConsumeKafka.TOPICS, topic);
+ runner.setProperty(ConsumeKafka.PROCESSING_STRATEGY,
ProcessingStrategy.RECORD.getValue());
+ runner.setProperty(ConsumeKafka.OUTPUT_STRATEGY,
OutputStrategy.USE_VALUE.getValue());
+ runner.setProperty(ConsumeKafka.SCHEMA_CONFLICT_RESOLUTION,
SchemaConflictResolution.CONTINUE_WITH_MERGED_SCHEMA.getValue());
+ runner.setProperty(ConsumeKafka.AUTO_OFFSET_RESET,
AutoOffsetReset.EARLIEST.getValue());
+
+ runner.run(1, false, true);
+
+ produce(topic, List.of(
+ new ProducerRecord<>(topic, FIRST_PARTITION, (String) null,
RECORD_WITH_ID, List.<Header>of()),
+ new ProducerRecord<>(topic, FIRST_PARTITION, (String) null,
RECORD_WITH_NAME, List.<Header>of())));
+
+ while
(runner.getFlowFilesForRelationship(ConsumeKafka.SUCCESS).isEmpty()) {
Review Comment:
I added `runUntil` method in [the abstract consumer test
class](https://github.com/apache/nifi/pull/11537/changes#diff-47c8f4fb3eae98f6963918a9e5db7036a25116e82e2a04c0ead0df3054f25de6R83).
I switched the tests I added to this method, but other tests still use
infinite `while` loop.
I didn't change them in this PR, but can do a follow up if needed.
##########
nifi-commons/nifi-record/src/test/java/org/apache/nifi/serialization/record/TestDataTypeUtils.java:
##########
@@ -1392,4 +1397,278 @@ public void
testMergeDataTypesWithManyDistinctRecordSchemasCompletesQuickly() {
assertTrue(finalSchema.getField("field_0").isPresent());
assertTrue(finalSchema.getField("field_4999").isPresent());
}
+
+ @Test
+ public void testMergeSchemasMakesSingleSideFieldsNullable() {
+ final RecordSchema schemaA = new SimpleRecordSchema(List.of(
+ new RecordField("id", RecordFieldType.STRING.getDataType(),
false)));
+ final RecordSchema schemaB = new SimpleRecordSchema(List.of(
+ new RecordField("name", RecordFieldType.STRING.getDataType(),
false)));
+
+ final RecordSchema merged = DataTypeUtils.merge(schemaA, schemaB);
+
+ assertEquals(2, merged.getFieldCount());
+ assertTrue(merged.getField("id").orElseThrow().isNullable());
+ assertTrue(merged.getField("name").orElseThrow().isNullable());
+ }
+
+ @Test
+ public void testMergeSchemasKeepsSharedNonNullableFieldNonNullable() {
+ final RecordSchema schemaA = new SimpleRecordSchema(List.of(
+ new RecordField("id", RecordFieldType.STRING.getDataType(),
false)));
+ final RecordSchema schemaB = new SimpleRecordSchema(List.of(
+ new RecordField("id", RecordFieldType.STRING.getDataType(),
false)));
+
+ final RecordSchema merged = DataTypeUtils.merge(schemaA, schemaB);
+
+ assertEquals(1, merged.getFieldCount());
+ assertFalse(merged.getField("id").orElseThrow().isNullable());
+ }
+
+ @Test
+ public void testMergeSchemasPreservesNullableWhenLeftIsNullable() {
+ final RecordSchema schemaA = new SimpleRecordSchema(List.of(
+ new RecordField("id", RecordFieldType.STRING.getDataType(),
true)));
+ final RecordSchema schemaB = new SimpleRecordSchema(List.of(
+ new RecordField("id", RecordFieldType.STRING.getDataType(),
false)));
+
+ final RecordSchema merged = DataTypeUtils.merge(schemaA, schemaB);
+
+ assertTrue(merged.getField("id").orElseThrow().isNullable());
+ }
+
+ @Test
+ public void testMergeSchemasPreservesNullableWhenRightIsNullable() {
+ final RecordSchema schemaA = new SimpleRecordSchema(List.of(
+ new RecordField("id", RecordFieldType.STRING.getDataType(),
false)));
+ final RecordSchema schemaB = new SimpleRecordSchema(List.of(
+ new RecordField("id", RecordFieldType.STRING.getDataType(),
true)));
+
+ final RecordSchema merged = DataTypeUtils.merge(schemaA, schemaB);
+
+ assertTrue(merged.getField("id").orElseThrow().isNullable());
+ }
+
+ @Test
+ public void testMergeSchemasWidensTypesAndNullifiesSingleSideFields() {
+ final RecordSchema schemaA = new SimpleRecordSchema(List.of(
+ new RecordField("id", RecordFieldType.INT.getDataType(), false),
+ new RecordField("onlyA", RecordFieldType.STRING.getDataType(),
false)));
+ final RecordSchema schemaB = new SimpleRecordSchema(List.of(
+ new RecordField("id", RecordFieldType.LONG.getDataType(), true),
+ new RecordField("onlyB", RecordFieldType.STRING.getDataType(),
Set.of("bAlias"), false)));
+
+ final RecordSchema merged = DataTypeUtils.merge(schemaA, schemaB);
+
+ assertEquals(3, merged.getFieldCount());
+ final RecordField idField = merged.getField("id").orElseThrow();
+ assertEquals(RecordFieldType.LONG,
idField.getDataType().getFieldType());
+ assertTrue(idField.isNullable());
+ assertTrue(merged.getField("onlyA").orElseThrow().isNullable());
+ assertTrue(merged.getField("onlyB").orElseThrow().isNullable());
+ }
+
+ @Test
+ public void testMergeSchemasIdenticalSchemasRetainNullability() {
+ final RecordSchema schemaA = new SimpleRecordSchema(List.of(
+ new RecordField("required", RecordFieldType.STRING.getDataType(),
false),
+ new RecordField("optional", RecordFieldType.INT.getDataType(),
true)));
+ final RecordSchema schemaB = new SimpleRecordSchema(List.of(
+ new RecordField("required", RecordFieldType.STRING.getDataType(),
false),
+ new RecordField("optional", RecordFieldType.INT.getDataType(),
true)));
+
+ final RecordSchema merged = DataTypeUtils.merge(schemaA, schemaB);
+
+ assertEquals(2, merged.getFieldCount());
+ assertFalse(merged.getField("required").orElseThrow().isNullable());
+ assertTrue(merged.getField("optional").orElseThrow().isNullable());
+ }
+
+ @Test
+ public void testMergeSchemasMakesNestedSingleSideFieldsNullable() {
+ final RecordSchema nestedA = new SimpleRecordSchema(List.of(
+ new RecordField("street", RecordFieldType.STRING.getDataType(),
false),
+ new RecordField("city", RecordFieldType.STRING.getDataType(),
false)));
+ final RecordSchema nestedB = new SimpleRecordSchema(List.of(
+ new RecordField("street", RecordFieldType.STRING.getDataType(),
false),
+ new RecordField("zip", RecordFieldType.STRING.getDataType(),
false)));
+
+ final RecordSchema schemaA = new SimpleRecordSchema(List.of(
+ new RecordField("address",
RecordFieldType.RECORD.getRecordDataType(nestedA), false)));
+ final RecordSchema schemaB = new SimpleRecordSchema(List.of(
+ new RecordField("address",
RecordFieldType.RECORD.getRecordDataType(nestedB), false)));
+
+ final RecordSchema merged = DataTypeUtils.merge(schemaA, schemaB);
+ final RecordField addressField =
merged.getField("address").orElseThrow();
+ assertFalse(addressField.isNullable());
+
+ final RecordSchema nestedMerged = ((RecordDataType)
addressField.getDataType()).getChildSchema();
+ assertEquals(3, nestedMerged.getFieldCount());
+
assertFalse(nestedMerged.getField("street").orElseThrow().isNullable());
+ assertTrue(nestedMerged.getField("city").orElseThrow().isNullable());
+ assertTrue(nestedMerged.getField("zip").orElseThrow().isNullable());
+ }
+
+ @Test
+ public void testMergeSchemasMakesArrayOfRecordSingleSideFieldsNullable() {
+ final RecordSchema elementA = new SimpleRecordSchema(List.of(
+ new RecordField("x", RecordFieldType.INT.getDataType(), false),
+ new RecordField("shared", RecordFieldType.STRING.getDataType(),
false)));
+ final RecordSchema elementB = new SimpleRecordSchema(List.of(
+ new RecordField("y", RecordFieldType.STRING.getDataType(), false),
+ new RecordField("shared", RecordFieldType.STRING.getDataType(),
false)));
+
+ final RecordSchema schemaA = new SimpleRecordSchema(List.of(
+ new RecordField("items",
RecordFieldType.ARRAY.getArrayDataType(RecordFieldType.RECORD.getRecordDataType(elementA)),
false)));
+ final RecordSchema schemaB = new SimpleRecordSchema(List.of(
+ new RecordField("items",
RecordFieldType.ARRAY.getArrayDataType(RecordFieldType.RECORD.getRecordDataType(elementB)),
false)));
+
+ final RecordSchema merged = DataTypeUtils.merge(schemaA, schemaB);
+ final RecordField itemsField = merged.getField("items").orElseThrow();
+ assertFalse(itemsField.isNullable());
+
+ final DataType elementType = ((ArrayDataType)
itemsField.getDataType()).getElementType();
+ final RecordSchema elementMerged = ((RecordDataType)
elementType).getChildSchema();
+ assertEquals(3, elementMerged.getFieldCount());
+
assertFalse(elementMerged.getField("shared").orElseThrow().isNullable());
+ assertTrue(elementMerged.getField("x").orElseThrow().isNullable());
+ assertTrue(elementMerged.getField("y").orElseThrow().isNullable());
+ }
+
+ @Test
+ public void
testMergeSchemasNullifiesTopLevelRecordFieldPresentOnOnlyOneSide() {
+ final RecordSchema nested = new SimpleRecordSchema(List.of(
+ new RecordField("value", RecordFieldType.STRING.getDataType(),
false)));
+
+ final RecordSchema schemaA = new SimpleRecordSchema(List.of(
+ new RecordField("id", RecordFieldType.STRING.getDataType(), false),
+ new RecordField("details",
RecordFieldType.RECORD.getRecordDataType(nested), false)));
+ final RecordSchema schemaB = new SimpleRecordSchema(List.of(
+ new RecordField("id", RecordFieldType.STRING.getDataType(),
false)));
+
+ final RecordSchema merged = DataTypeUtils.merge(schemaA, schemaB);
+
+ assertFalse(merged.getField("id").orElseThrow().isNullable());
+ assertTrue(merged.getField("details").orElseThrow().isNullable());
+ final RecordSchema detailsSchema = ((RecordDataType)
merged.getField("details").orElseThrow().getDataType()).getChildSchema();
+
assertFalse(detailsSchema.getField("value").orElseThrow().isNullable());
+ }
+
+ @Test
+ public void testMergeSchemasMakesDeeplyNestedSingleSideFieldsNullable() {
+ final RecordSchema leafA = new SimpleRecordSchema(List.of(
+ new RecordField("aOnly", RecordFieldType.STRING.getDataType(),
false),
+ new RecordField("common", RecordFieldType.INT.getDataType(),
false)));
+ final RecordSchema leafB = new SimpleRecordSchema(List.of(
+ new RecordField("bOnly", RecordFieldType.STRING.getDataType(),
false),
+ new RecordField("common", RecordFieldType.INT.getDataType(),
false)));
+
+ final RecordSchema midA = new SimpleRecordSchema(List.of(
+ new RecordField("leaf",
RecordFieldType.RECORD.getRecordDataType(leafA), false)));
+ final RecordSchema midB = new SimpleRecordSchema(List.of(
+ new RecordField("leaf",
RecordFieldType.RECORD.getRecordDataType(leafB), false)));
+
+ final RecordSchema schemaA = new SimpleRecordSchema(List.of(
+ new RecordField("mid",
RecordFieldType.RECORD.getRecordDataType(midA), false)));
+ final RecordSchema schemaB = new SimpleRecordSchema(List.of(
+ new RecordField("mid",
RecordFieldType.RECORD.getRecordDataType(midB), false)));
+
+ final RecordSchema merged = DataTypeUtils.merge(schemaA, schemaB);
+ final RecordSchema midMerged = ((RecordDataType)
merged.getField("mid").orElseThrow().getDataType()).getChildSchema();
+ final RecordSchema leafMerged = ((RecordDataType)
midMerged.getField("leaf").orElseThrow().getDataType()).getChildSchema();
+
+ assertFalse(leafMerged.getField("common").orElseThrow().isNullable());
+ assertTrue(leafMerged.getField("aOnly").orElseThrow().isNullable());
+ assertTrue(leafMerged.getField("bOnly").orElseThrow().isNullable());
+ }
+
+ @Test
+ public void testMergeSchemasMatchesFieldsByAliasWithoutNullifying() {
+ final RecordSchema schemaA = new SimpleRecordSchema(List.of(
+ new RecordField("id", RecordFieldType.STRING.getDataType(),
false)));
+ final RecordSchema schemaB = new SimpleRecordSchema(List.of(
+ new RecordField("id", RecordFieldType.STRING.getDataType(),
Set.of("identifier"), false)));
+
+ final RecordSchema merged = DataTypeUtils.merge(schemaA, schemaB);
+
+ assertEquals(1, merged.getFieldCount());
+ final RecordField mergedField = merged.getField("id").orElseThrow();
+ assertFalse(mergedField.isNullable());
+ assertTrue(mergedField.getAliases().contains("identifier"));
+ }
+
+ @Test
+ public void testMergeSchemasWithEmptySchemaReturnsOtherUnchanged() {
+ final RecordSchema schemaA = new SimpleRecordSchema(List.of(
+ new RecordField("id", RecordFieldType.STRING.getDataType(),
false)));
+ final RecordSchema empty = new SimpleRecordSchema(List.of());
+
+ final RecordSchema mergedWithEmptyOther = DataTypeUtils.merge(schemaA,
empty);
+ final RecordSchema mergedWithEmptyThis = DataTypeUtils.merge(empty,
schemaA);
+
+
assertFalse(mergedWithEmptyOther.getField("id").orElseThrow().isNullable());
+
assertFalse(mergedWithEmptyThis.getField("id").orElseThrow().isNullable());
Review Comment:
Originally empty schemas were treated as missing schemas. Agree, with the
nullability logic in-place we should treat them normally.
I removed the special handling for empty schemas, now they follow the normal
path.
--
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]
