From eba984724dda47188b1be2c0473761c79f14d373 Mon Sep 17 00:00:00 2001
From: "Sung Yun (CODE SIGNING KEY)" <sungwy@apache.org>
Date: Tue, 19 May 2026 21:22:53 -0400
Subject: [PATCH] fix int96 timestamp offset in arrow dictionary decode

---
 ...ectorizedParquetDefinitionLevelReader.java |   2 +-
 ...ectorizedParquetDefinitionLevelReader.java | 115 ++++++++++++++++++
 2 files changed, 116 insertions(+), 1 deletion(-)
 create mode 100644 arrow/src/test/java/org/apache/iceberg/arrow/vectorized/parquet/TestVectorizedParquetDefinitionLevelReader.java

diff --git a/arrow/src/main/java/org/apache/iceberg/arrow/vectorized/parquet/VectorizedParquetDefinitionLevelReader.java b/arrow/src/main/java/org/apache/iceberg/arrow/vectorized/parquet/VectorizedParquetDefinitionLevelReader.java
index 1ca3bfe809c0..d093d4c97989 100644
--- a/arrow/src/main/java/org/apache/iceberg/arrow/vectorized/parquet/VectorizedParquetDefinitionLevelReader.java
+++ b/arrow/src/main/java/org/apache/iceberg/arrow/vectorized/parquet/VectorizedParquetDefinitionLevelReader.java
@@ -539,7 +539,7 @@ protected void nextDictEncodedVal(
                   .toByteBuffer()
                   .order(ByteOrder.LITTLE_ENDIAN);
           long timestampInt96 = ParquetUtil.extractTimestampInt96(buffer);
-          vector.getDataBuffer().setLong(idx, timestampInt96);
+          vector.getDataBuffer().setLong((long) idx * typeWidth, timestampInt96);
           break;
         default:
           throw new UnsupportedOperationException(
diff --git a/arrow/src/test/java/org/apache/iceberg/arrow/vectorized/parquet/TestVectorizedParquetDefinitionLevelReader.java b/arrow/src/test/java/org/apache/iceberg/arrow/vectorized/parquet/TestVectorizedParquetDefinitionLevelReader.java
new file mode 100644
index 000000000000..10bf3e64766c
--- /dev/null
+++ b/arrow/src/test/java/org/apache/iceberg/arrow/vectorized/parquet/TestVectorizedParquetDefinitionLevelReader.java
@@ -0,0 +1,115 @@
+/*
+ * 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.iceberg.arrow.vectorized.parquet;
+
+import static org.assertj.core.api.Assertions.assertThat;
+
+import java.nio.ByteBuffer;
+import java.nio.ByteOrder;
+import org.apache.arrow.memory.RootAllocator;
+import org.apache.arrow.vector.BigIntVector;
+import org.apache.parquet.column.Dictionary;
+import org.apache.parquet.column.Encoding;
+import org.apache.parquet.io.api.Binary;
+import org.junit.jupiter.api.Test;
+
+public class TestVectorizedParquetDefinitionLevelReader {
+  private static final int UNIX_EPOCH_JULIAN_DAY = 2_440_588;
+
+  @Test
+  public void timestampInt96ReaderPackedDictionaryDecodeDecodesRowsCorrectly() {
+    try (RootAllocator allocator = new RootAllocator();
+        BigIntVector vector = new BigIntVector("ts", allocator)) {
+      vector.allocateNew(2);
+      vector.set(0, -1L);
+      vector.set(1, -1L);
+
+      VectorizedParquetDefinitionLevelReader definitionReader =
+          new VectorizedParquetDefinitionLevelReader(1, 1, false);
+      VectorizedDictionaryEncodedParquetValuesReader dictionaryReader =
+          new VectorizedDictionaryEncodedParquetValuesReader(1, false);
+
+      dictionaryReader.mode = BaseVectorizedParquetValuesReader.Mode.PACKED;
+      dictionaryReader.currentCount = 2;
+      dictionaryReader.packedValuesBuffer[0] = 0;
+      dictionaryReader.packedValuesBuffer[1] = 1;
+
+      Dictionary dictionary =
+          new Dictionary(Encoding.PLAIN_DICTIONARY) {
+            @Override
+            public int getMaxId() {
+              return 1;
+            }
+
+            @Override
+            public Binary decodeToBinary(int id) {
+              if (id == 0) {
+                return int96Binary(111_111L);
+              } else if (id == 1) {
+                return int96Binary(222_222L);
+              }
+
+              throw new IllegalArgumentException("Unexpected dictionary id: " + id);
+            }
+          };
+
+      VectorizedParquetDefinitionLevelReader.TimestampInt96Reader timestampReader =
+          definitionReader.timestampInt96Reader();
+
+      timestampReader.nextDictEncodedVal(
+          vector,
+          0,
+          dictionaryReader,
+          dictionary,
+          BaseVectorizedParquetValuesReader.Mode.PACKED,
+          1,
+          null,
+          Long.BYTES);
+      timestampReader.nextDictEncodedVal(
+          vector,
+          1,
+          dictionaryReader,
+          dictionary,
+          BaseVectorizedParquetValuesReader.Mode.PACKED,
+          1,
+          null,
+          Long.BYTES);
+
+      vector.setValueCount(2);
+
+      assertThat(vector.get(0))
+          .as("row 0 should receive the first decoded timestamp")
+          .isEqualTo(111_111L);
+      assertThat(vector.get(1))
+          .as("row 1 should receive the second decoded timestamp")
+          .isEqualTo(222_222L);
+    }
+  }
+
+  private static Binary int96Binary(long micros) {
+    long timeOfDayNanos = micros * 1_000L;
+    byte[] bytes =
+        ByteBuffer.allocate(12)
+            .order(ByteOrder.LITTLE_ENDIAN)
+            .putLong(timeOfDayNanos)
+            .putInt(UNIX_EPOCH_JULIAN_DAY)
+            .array();
+    return Binary.fromConstantByteArray(bytes);
+  }
+}