From 8d6218ca98f6d80e5f63a5aa9839f0397d8a50ac Mon Sep 17 00:00:00 2001
From: jhkim <jhkim@apache.org>
Date: Tue, 9 Dec 2014 12:21:34 +0900
Subject: [PATCH] TAJO-1236: Remove slow 'new String' operation in parquet
 format

---
 .../storage/parquet/TajoRecordConverter.java  | 34 ++++++++-----------
 .../storage/parquet/TajoWriteSupport.java     | 23 ++++++-------
 2 files changed, 25 insertions(+), 32 deletions(-)

diff --git a/tajo-storage/src/main/java/org/apache/tajo/storage/parquet/TajoRecordConverter.java b/tajo-storage/src/main/java/org/apache/tajo/storage/parquet/TajoRecordConverter.java
index 7c3d79d778..a091eac29d 100644
--- a/tajo-storage/src/main/java/org/apache/tajo/storage/parquet/TajoRecordConverter.java
+++ b/tajo-storage/src/main/java/org/apache/tajo/storage/parquet/TajoRecordConverter.java
@@ -18,29 +18,23 @@
 
 package org.apache.tajo.storage.parquet;
 
-import com.google.protobuf.Message;
 import com.google.protobuf.InvalidProtocolBufferException;
-
-import java.nio.ByteBuffer;
-
-import parquet.io.api.GroupConverter;
-import parquet.io.api.Converter;
-import parquet.io.api.PrimitiveConverter;
-import parquet.io.api.Binary;
-import parquet.schema.Type;
-import parquet.schema.GroupType;
-
+import com.google.protobuf.Message;
+import org.apache.tajo.catalog.Column;
+import org.apache.tajo.catalog.Schema;
 import org.apache.tajo.common.TajoDataTypes;
 import org.apache.tajo.common.TajoDataTypes.DataType;
-import org.apache.tajo.catalog.Schema;
-import org.apache.tajo.catalog.Column;
+import org.apache.tajo.datum.*;
 import org.apache.tajo.storage.Tuple;
 import org.apache.tajo.storage.VTuple;
-import org.apache.tajo.datum.DatumFactory;
-import org.apache.tajo.datum.Datum;
-import org.apache.tajo.datum.BlobDatum;
-import org.apache.tajo.datum.NullDatum;
-import org.apache.tajo.datum.ProtobufDatumFactory;
+import parquet.io.api.Binary;
+import parquet.io.api.Converter;
+import parquet.io.api.GroupConverter;
+import parquet.io.api.PrimitiveConverter;
+import parquet.schema.GroupType;
+import parquet.schema.Type;
+
+import java.nio.ByteBuffer;
 
 /**
  * Converter to convert a Parquet record into a Tajo Tuple.
@@ -222,7 +216,7 @@ public FieldCharConverter(ParentValueContainer parent) {
 
     @Override
     final public void addBinary(Binary value) {
-      parent.add(DatumFactory.createChar(value.toStringUsingUTF8()));
+      parent.add(DatumFactory.createChar(value.getBytes()));
     }
   }
 
@@ -343,7 +337,7 @@ public FieldTextConverter(ParentValueContainer parent) {
 
     @Override
     final public void addBinary(Binary value) {
-      parent.add(DatumFactory.createText(value.toStringUsingUTF8()));
+      parent.add(DatumFactory.createText(value.getBytes()));
     }
   }
 
diff --git a/tajo-storage/src/main/java/org/apache/tajo/storage/parquet/TajoWriteSupport.java b/tajo-storage/src/main/java/org/apache/tajo/storage/parquet/TajoWriteSupport.java
index 35165de791..86511315df 100644
--- a/tajo-storage/src/main/java/org/apache/tajo/storage/parquet/TajoWriteSupport.java
+++ b/tajo-storage/src/main/java/org/apache/tajo/storage/parquet/TajoWriteSupport.java
@@ -18,10 +18,12 @@
 
 package org.apache.tajo.storage.parquet;
 
-import java.util.Map;
-import java.util.HashMap;
-import java.util.List;
-
+import org.apache.hadoop.conf.Configuration;
+import org.apache.tajo.catalog.Column;
+import org.apache.tajo.catalog.Schema;
+import org.apache.tajo.common.TajoDataTypes;
+import org.apache.tajo.datum.Datum;
+import org.apache.tajo.storage.Tuple;
 import parquet.hadoop.api.WriteSupport;
 import parquet.io.api.Binary;
 import parquet.io.api.RecordConsumer;
@@ -29,12 +31,9 @@
 import parquet.schema.MessageType;
 import parquet.schema.Type;
 
-import org.apache.hadoop.conf.Configuration;
-import org.apache.tajo.catalog.Schema;
-import org.apache.tajo.catalog.Column;
-import org.apache.tajo.common.TajoDataTypes;
-import org.apache.tajo.storage.Tuple;
-import org.apache.tajo.datum.Datum;
+import java.util.HashMap;
+import java.util.List;
+import java.util.Map;
 
 /**
  * Tajo implementation of {@link WriteSupport} for {@link Tuple}s.
@@ -116,7 +115,7 @@ private void writeRecordFields(GroupType schema, Schema tajoSchema,
   private void writeValue(Type fieldType, Column column, Datum datum) {
     switch (column.getDataType().getType()) {
       case BOOLEAN:
-        recordConsumer.addBoolean((Boolean) datum.asBool());
+        recordConsumer.addBoolean(datum.asBool());
         break;
       case BIT:
       case INT2:
@@ -134,7 +133,7 @@ private void writeValue(Type fieldType, Column column, Datum datum) {
         break;
       case CHAR:
       case TEXT:
-        recordConsumer.addBinary(Binary.fromString(datum.asChars()));
+        recordConsumer.addBinary(Binary.fromByteArray(datum.asTextBytes()));
         break;
       case PROTOBUF:
       case BLOB: