/**
* 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.drill.exec.store.easy.text.compliant;
import io.netty.buffer.DrillBuf;
import io.netty.util.internal.PlatformDependent;
import java.util.ArrayList;
import java.util.Arrays;
import java.util.Collection;
import java.util.Collections;
import java.util.List;
import org.apache.drill.common.exceptions.ExecutionSetupException;
import org.apache.drill.common.expression.FieldReference;
import org.apache.drill.common.expression.SchemaPath;
import org.apache.drill.common.types.TypeProtos;
import org.apache.drill.common.types.Types;
import org.apache.drill.exec.exception.SchemaChangeException;
import org.apache.drill.exec.physical.impl.OutputMutator;
import org.apache.drill.exec.record.MaterializedField;
import org.apache.drill.exec.vector.RepeatedVarCharVector;
import com.google.common.base.Preconditions;
/**
* Class is responsible for generating record batches for text file inputs. We generate
* a record batch with a single vector of type repeated varchar vector. Each record is a single
* value within the vector containing all the fields in the record as individual array elements.
*/
class RepeatedVarCharOutput extends TextOutput {
static final org.slf4j.Logger logger = org.slf4j.LoggerFactory.getLogger(RepeatedVarCharOutput.class);
static final String COL_NAME = "columns";
static final SchemaPath COLUMNS = SchemaPath.getSimplePath("columns");
public static final int MAXIMUM_NUMBER_COLUMNS = 64 * 1024;
// output vector
private final RepeatedVarCharVector vector;
// mutator for the output vector
private final RepeatedVarCharVector.Mutator mutator;
// boolean array indicating which fields are selected (if star query entire array is set to true)
private final boolean[] collectedFields;
// pointer to keep track of the offsets per record
private long repeatedOffset;
// pointer to keep track of the original offsets per record
private long repeatedOffsetOriginal;
// pointer to end of the offset buffer
private long repeatedOffsetMax;
// pointer to the start of the actual data buffer
private long characterDataOriginal;
// pointer to the current location of the data buffer
private long characterData;
// pointer to the end of the data buffer
private long characterDataMax;
// current pointer into the buffer that keeps track of the length of individual fields
private long charLengthOffset;
// pointer to the start of the length buffer
private long charLengthOffsetOriginal;
// pointer to the end of length buffer
private long charLengthOffsetMax;
// pointer to the beginning of the record
private long recordStart;
// total number of records processed (across batches)
private long recordCount;
// number of records processed in this current batch
private int batchIndex;
// current index of the field being processed within the record
private int fieldIndex = -1;
/* boolean to indicate if we are currently appending data to the output vector
* Its set to false when we have hit out of memory or we are not interested in
* the particular field
*/
private boolean collect;
// are we currently appending to a field
private boolean fieldOpen;
// maximum number of fields/columns
private final int maxField;
/**
* We initialize and add the repeated varchar vector to the record batch in this
* constructor. Perform some sanity checks if the selected columns are valid or not.
* @param outputMutator Used to create/modify schema in the record batch
* @param columns List of columns selected in the query
* @param isStarQuery boolean to indicate if all fields are selected or not
* @throws SchemaChangeException
*/
public RepeatedVarCharOutput(OutputMutator outputMutator, Collection<SchemaPath> columns, boolean isStarQuery) throws SchemaChangeException {
super();
MaterializedField field = MaterializedField.create(COL_NAME, Types.repeated(TypeProtos.MinorType.VARCHAR));
this.vector = outputMutator.addField(field, RepeatedVarCharVector.class);
this.mutator = vector.getMutator();
{ // setup fields
List<Integer> columnIds = new ArrayList<Integer>();
if (!isStarQuery) {
String pathStr;
for (SchemaPath path : columns) {
assert path.getRootSegment().isNamed() : "root segment should be named";
pathStr = path.getRootSegment().getPath();
Preconditions.checkArgument(pathStr.equals(COL_NAME) || (pathStr.equals("*") && path.getRootSegment().getChild() == null),
String.format("Selected column '%s' must have name 'columns' or must be plain '*'", pathStr));
if (path.getRootSegment().getChild() != null) {
Preconditions.checkArgument(path.getRootSegment().getChild().isArray(),
String.format("Selected column '%s' must be an array index", pathStr));
int index = path.getRootSegment().getChild().getArraySegment().getIndex();
columnIds.add(index);
}
}
Collections.sort(columnIds);
}
boolean[] fields = new boolean[MAXIMUM_NUMBER_COLUMNS];
int maxField = fields.length;
if(isStarQuery){
Arrays.fill(fields, true);
}else{
for(Integer i : columnIds){
maxField = 0;
maxField = Math.max(maxField, i);
fields[i] = true;
}
}
this.collectedFields = fields;
this.maxField = maxField;
}
}
/**
* Start a new record batch. Resets all the offsets and pointers that
* store buffer addresses
*/
@Override
public void startBatch() {
this.recordStart = characterDataOriginal;
this.fieldOpen = false;
this.batchIndex = 0;
this.fieldIndex = -1;
this.collect = true;
loadRepeatedOffsetAddress();
loadVarCharOffsetAddress();
loadVarCharDataAddress();
}
private void loadRepeatedOffsetAddress(){
@SuppressWarnings("resource")
DrillBuf buf = vector.getOffsetVector().getBuffer();
checkBuf(buf);
this.repeatedOffset = buf.memoryAddress() + 4;
this.repeatedOffsetOriginal = buf.memoryAddress() + 4;
this.repeatedOffsetMax = buf.memoryAddress() + buf.capacity();
}
private void loadVarCharDataAddress(){
@SuppressWarnings("resource")
DrillBuf buf = vector.getDataVector().getBuffer();
checkBuf(buf);
this.characterData = buf.memoryAddress();
this.characterDataOriginal = buf.memoryAddress();
this.characterDataMax = buf.memoryAddress() + buf.capacity();
}
private void loadVarCharOffsetAddress(){
@SuppressWarnings("resource")
DrillBuf buf = vector.getDataVector().getOffsetVector().getBuffer();
checkBuf(buf);
this.charLengthOffset = buf.memoryAddress() + 4;
this.charLengthOffsetOriginal = buf.memoryAddress() + 4; // add four as offsets conceptually start at 1. (first item is 0..1)
this.charLengthOffsetMax = buf.memoryAddress() + buf.capacity();
}
private void expandVarCharOffsets(){
vector.getDataVector().getOffsetVector().reAlloc();
long diff = charLengthOffset - charLengthOffsetOriginal;
loadVarCharOffsetAddress();
charLengthOffset += diff;
}
private void expandVarCharData(){
vector.getDataVector().reAlloc();
long diff = characterData - characterDataOriginal;
loadVarCharDataAddress();
characterData += diff;
}
private void expandRepeatedOffsets(){
vector.getOffsetVector().reAlloc();
long diff = repeatedOffset - repeatedOffsetOriginal;
loadRepeatedOffsetAddress();
repeatedOffset += diff;
}
/**
* Helper method to check if the buffer we are accessing
* has a minimum reference count and has not been deallocated
* @param b working drill buffer
*/
private void checkBuf(DrillBuf b){
if(b.refCnt() < 1){
throw new IllegalStateException("Cannot access a dereferenced buffer.");
}
}
@Override
public void startField(int index) {
fieldIndex = index;
collect = collectedFields[index];
fieldOpen = true;
}
@Override
public boolean endField() {
fieldOpen = false;
if(charLengthOffset >= charLengthOffsetMax){
expandVarCharOffsets();
}
int newOffset = (int) (characterData - characterDataOriginal);
PlatformDependent.putInt(charLengthOffset, newOffset);
charLengthOffset += 4;
return fieldIndex < maxField;
}
@Override
public boolean endEmptyField() {
return endField();
}
@Override
public void append(byte data) {
if(!collect){
return;
}
if(characterData >= characterDataMax){
expandVarCharData();
}
PlatformDependent.putByte(characterData, data);
characterData++;
}
@Override
public long getRecordCount() {
return recordCount;
}
@Override
public boolean rowHasData() {
return this.recordStart < characterData;
}
@Override
public void finishRecord() {
this.recordStart = characterData;
if(fieldOpen){
endField();
}
if(repeatedOffset >= repeatedOffsetMax){
expandRepeatedOffsets();
}
int newOffset = ((int) (charLengthOffset - charLengthOffsetOriginal))/4;
PlatformDependent.putInt(repeatedOffset, newOffset);
repeatedOffset += 4;
// if there were no defined fields, skip.
if(fieldIndex > -1){
batchIndex++;
recordCount++;
}
}
/**
* This method is a helper method added for DRILL-951
* TextRecordReader to call this method to get field names out
* @return array of field data strings
*/
public String [] getTextOutput () throws ExecutionSetupException {
if (recordCount == 0 || fieldIndex == -1) {
return null;
}
if (this.recordStart != characterData) {
throw new ExecutionSetupException("record text was requested before finishing record");
}
//Currently only first line header is supported. Return only first record.
int retSize = fieldIndex+1;
String [] out = new String [retSize];
RepeatedVarCharVector.Accessor a = this.vector.getAccessor();
for (int i=0; i<retSize; i++){
out[i] = a.getSingleObject(0,i).toString();
}
return out;
}
// Sets the record count in this batch within the value vector
@Override
public void finishBatch() {
mutator.setValueCount(batchIndex);
}
}