[FLINK-36084] Optimize parquet binary getBytes with getBytesUnsafe to…

… avoid copy cost (apache#25213)
Myracle · Aug 19, 2024 · adb8368 · adb8368
1 parent 5f6200c
commit adb8368
Show file tree

Hide file tree

Showing 3 changed files with 3 additions and 3 deletions.
diff --git a/...link-parquet/src/main/java/org/apache/flink/formats/parquet/vector/ParquetDictionary.java b/...link-parquet/src/main/java/org/apache/flink/formats/parquet/vector/ParquetDictionary.java
@@ -62,7 +62,7 @@ public double decodeToDouble(int id) {
 
     @Override
     public byte[] decodeToBinary(int id) {
-        return dictionary.decodeToBinary(id).getBytes();
+        return dictionary.decodeToBinary(id).getBytesUnsafe();
     }
 
     @Override

diff --git a/...rquet/src/main/java/org/apache/flink/formats/parquet/vector/reader/BytesColumnReader.java b/...rquet/src/main/java/org/apache/flink/formats/parquet/vector/reader/BytesColumnReader.java
@@ -74,7 +74,7 @@ protected void readBatchFromDictionaryIds(
             int rowId, int num, WritableBytesVector column, WritableIntVector dictionaryIds) {
         for (int i = rowId; i < rowId + num; ++i) {
             if (!column.isNullAt(i)) {
-                byte[] bytes = dictionary.decodeToBinary(dictionaryIds.getInt(i)).getBytes();
+                byte[] bytes = dictionary.decodeToBinary(dictionaryIds.getInt(i)).getBytesUnsafe();
                 column.appendBytes(i, bytes, 0, bytes.length);
             }
         }

diff --git a/...c/main/java/org/apache/flink/formats/parquet/vector/reader/FixedLenBytesColumnReader.java b/...c/main/java/org/apache/flink/formats/parquet/vector/reader/FixedLenBytesColumnReader.java
@@ -101,7 +101,7 @@ protected void readBatchFromDictionaryIds(
             WritableBytesVector bytesVector = (WritableBytesVector) column;
             for (int i = rowId; i < rowId + num; ++i) {
                 if (!bytesVector.isNullAt(i)) {
-                    byte[] v = dictionary.decodeToBinary(dictionaryIds.getInt(i)).getBytes();
+                    byte[] v = dictionary.decodeToBinary(dictionaryIds.getInt(i)).getBytesUnsafe();
                     bytesVector.appendBytes(i, v, 0, v.length);
                 }
             }