voonhous commented on code in PR #17581:
URL: https://github.com/apache/hudi/pull/17581#discussion_r2641768612
##########
hudi-common/src/test/java/org/apache/hudi/common/schema/TestHoodieSchemaUtils.java:
##########
@@ -1539,4 +1540,202 @@ public void
testConvertValueForSpecificDataTypes_UnionWithNull() {
assertTrue(result instanceof LocalDate);
assertEquals(LocalDate.of(2023, 1, 1), result);
}
+
+ @Test
+ void testResolveUnionSchemaWithNonUnionSchema() {
+ // Non-union schemas should be returned as-is
+ HoodieSchema stringSchema = HoodieSchema.create(HoodieSchemaType.STRING);
+ HoodieSchema result = HoodieSchemaUtils.resolveUnionSchema(stringSchema,
"any");
+
+ assertSame(stringSchema, result);
+ }
+
+ @Test
+ void testResolveUnionSchemaWithSimpleNullableUnion() {
+ // Simple nullable union: ["null", "string"] should return the non-null
type efficiently
+ HoodieSchema nullableString =
HoodieSchema.createNullable(HoodieSchema.create(HoodieSchemaType.STRING));
+ HoodieSchema result = HoodieSchemaUtils.resolveUnionSchema(nullableString,
"string");
+
+ assertEquals(HoodieSchemaType.STRING, result.getType());
+ }
+
+ @Test
+ void testResolveUnionSchemaWithSimpleNullableRecord() {
+ // Test with nullable record type
+ HoodieSchema personSchema = HoodieSchema.createRecord(
+ "Person",
+ null,
+ null,
+ Collections.singletonList(
+ HoodieSchemaField.of("name",
HoodieSchema.create(HoodieSchemaType.STRING))
+ )
+ );
+
+ HoodieSchema nullablePerson = HoodieSchema.createNullable(personSchema);
+ HoodieSchema result = HoodieSchemaUtils.resolveUnionSchema(nullablePerson,
"Person");
+
+ assertEquals(HoodieSchemaType.RECORD, result.getType());
+ assertEquals("Person", result.getName());
+ assertFalse(result.isNullable());
+ }
+
+ @Test
+ void testResolveUnionSchemaWithComplexUnionMatchingFullName() {
+ // Complex union with 3+ types, matching by fullName
+ String unionSchemaJson = "{"
+ + "\"type\":\"record\","
+ + "\"name\":\"Container\","
+ + "\"fields\":[{"
+ + " \"name\":\"data\","
+ + " \"type\":[\"null\","
+ + "
{\"type\":\"record\",\"name\":\"PersonRecord\",\"fields\":[{\"name\":\"name\",\"type\":\"string\"}]},"
+ + "
{\"type\":\"record\",\"name\":\"CompanyRecord\",\"fields\":[{\"name\":\"companyName\",\"type\":\"string\"}]}"
+ + " ]"
+ + "}]}";
+
+ HoodieSchema containerSchema = HoodieSchema.parse(unionSchemaJson);
+ HoodieSchema dataFieldSchema =
containerSchema.getField("data").get().schema();
+
+ // Resolve to PersonRecord
+ HoodieSchema personResult =
HoodieSchemaUtils.resolveUnionSchema(dataFieldSchema, "PersonRecord");
+ assertEquals(HoodieSchemaType.RECORD, personResult.getType());
+ assertEquals("PersonRecord", personResult.getName());
+ assertFalse(personResult.isNullable());
+ assertTrue(personResult.getField("name").isPresent());
+
+ // Resolve to CompanyRecord
+ HoodieSchema companyResult =
HoodieSchemaUtils.resolveUnionSchema(dataFieldSchema, "CompanyRecord");
+ assertEquals(HoodieSchemaType.RECORD, companyResult.getType());
+ assertEquals("CompanyRecord", companyResult.getName());
+ assertFalse(companyResult.isNullable());
+ assertTrue(companyResult.getField("companyName").isPresent());
+ }
+
+ @Test
+ void testResolveUnionSchemaWithNonNullableTwoTypeUnion() {
+ // Union of two non-nullable types should use the complex resolution path
+ String unionSchemaJson = "{"
+ + "\"type\":\"record\","
+ + "\"name\":\"Container\","
+ + "\"fields\":[{"
+ + " \"name\":\"data\","
+ + " \"type\":["
+ + "
{\"type\":\"record\",\"name\":\"TypeA\",\"fields\":[{\"name\":\"fieldA\",\"type\":\"string\"}]},"
+ + "
{\"type\":\"record\",\"name\":\"TypeB\",\"fields\":[{\"name\":\"fieldB\",\"type\":\"int\"}]}"
+ + " ]"
+ + "}]}";
+
+ HoodieSchema containerSchema = HoodieSchema.parse(unionSchemaJson);
+ HoodieSchema dataFieldSchema =
containerSchema.getField("data").get().schema();
+
+ // Resolve to TypeA
+ HoodieSchema typeAResult =
HoodieSchemaUtils.resolveUnionSchema(dataFieldSchema, "TypeA");
+ assertEquals(HoodieSchemaType.RECORD, typeAResult.getType());
+ assertEquals("TypeA", typeAResult.getName());
+ assertFalse(typeAResult.isNullable());
+
+ // Resolve to TypeB
+ HoodieSchema typeBResult =
HoodieSchemaUtils.resolveUnionSchema(dataFieldSchema, "TypeB");
+ assertEquals(HoodieSchemaType.RECORD, typeBResult.getType());
+ assertEquals("TypeB", typeBResult.getName());
+ assertFalse(typeAResult.isNullable());
+ }
+
+ @Test
+ void testResolveUnionSchemaThrowsExceptionWhenNoMatch() {
+ // Complex union where the requested fullName doesn't match any type
+ String unionSchemaJson = "{"
+ + "\"type\":\"record\","
+ + "\"name\":\"Container\","
+ + "\"fields\":[{"
+ + " \"name\":\"data\","
+ + " \"type\":[\"null\","
+ + "
{\"type\":\"record\",\"name\":\"PersonRecord\",\"fields\":[{\"name\":\"name\",\"type\":\"string\"}]},"
+ + "
{\"type\":\"record\",\"name\":\"CompanyRecord\",\"fields\":[{\"name\":\"companyName\",\"type\":\"string\"}]}"
+ + " ]"
+ + "}]}";
+
+ HoodieSchema containerSchema = HoodieSchema.parse(unionSchemaJson);
+ HoodieSchema dataFieldSchema =
containerSchema.getField("data").get().schema();
+
+ // Try to resolve to a type that doesn't exist in the union
+ org.apache.hudi.internal.schema.HoodieSchemaException exception =
assertThrows(
Review Comment:
Done
--
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]
