[HUDI-8004] avoid unnecessary record rewriting for merge handler

apache · Jan 21, 2025 · b5de509 · b5de509
1 parent 9969b44
commit b5de509
Show file tree

Hide file tree

Showing 2 changed files with 38 additions and 1 deletion.
diff --git a/hudi-common/src/main/java/org/apache/hudi/avro/AvroSchemaUtils.java b/hudi-common/src/main/java/org/apache/hudi/avro/AvroSchemaUtils.java
@@ -174,7 +174,18 @@ private static boolean isAtomicSchemasCompatible(Schema oneAtomicType, Schema an
    * </ol>
    */
   public static boolean isStrictProjectionOf(Schema sourceSchema, Schema targetSchema) {
-    return isProjectionOfInternal(sourceSchema, targetSchema, Objects::equals);
+    return isProjectionOfInternal(sourceSchema, targetSchema, AvroSchemaUtils::isAtomicTypeStrictProject);
+  }
+
+  private static boolean isAtomicTypeStrictProject(Schema source, Schema target) {
+    // ignore name/namespace for FIXED type
+    if (source.getType() == Schema.Type.FIXED && target.getType() == Schema.Type.FIXED) {
+      return source.getLogicalType().equals(target.getLogicalType())
+          && source.getFixedSize() == target.getFixedSize()
+          && source.getObjectProps().equals(target.getObjectProps());
+    } else {
+      return Objects.equals(source, target);
+    }
   }
 
   private static boolean isProjectionOfInternal(Schema sourceSchema,

diff --git a/hudi-common/src/test/java/org/apache/hudi/avro/TestAvroSchemaUtils.java b/hudi-common/src/test/java/org/apache/hudi/avro/TestAvroSchemaUtils.java
@@ -46,6 +46,19 @@ public class TestAvroSchemaUtils {
       + "      \"name\": \"number\",\n"
       + "      \"type\": [\"null\", \"int\"]\n"
       + "    },\n"
+      + "    {\n" +
+      "        \"name\" : \"f1\",\n" +
+      "        \"type\" : [ \"null\", {\n" +
+      "           \"type\" : \"fixed\",\n" +
+      "           \"name\" : \"f1\",\n" +
+      "           \"namespace\" : \"\",\n" +
+      "           \"size\" : 5,\n" +
+      "           \"logicalType\" : \"decimal\",\n" +
+      "           \"precision\" : 10,\n" +
+      "           \"scale\" : 2\n" +
+      "           }],\n" +
+      "       \"default\" : null\n" +
+      "      },\n"
       + "    {\n"
       + "      \"name\": \"nested_record\",\n"
       + "      \"type\": {\n"
@@ -75,6 +88,19 @@ public class TestAvroSchemaUtils {
       + "      \"name\": \"number\",\n"
       + "      \"type\": [\"null\", \"int\"]\n"
       + "    },\n"
+      + "    {\n" +
+      "        \"name\" : \"f1\",\n" +
+      "        \"type\" : [ \"null\", {\n" +
+      "           \"type\" : \"fixed\",\n" +
+      "           \"name\" : \"fixed\",\n" +
+      "           \"namespace\" : \"example.schema.source.f1\",\n" +
+      "           \"size\" : 5,\n" +
+      "           \"logicalType\" : \"decimal\",\n" +
+      "           \"precision\" : 10,\n" +
+      "           \"scale\" : 2\n" +
+      "           }],\n" +
+      "       \"default\" : null\n" +
+      "      },\n"
       + "    {\n"
       + "      \"name\": \"nested_record\",\n"
       + "      \"type\": {\n"