Many resources are needed to download a project. Please understand that we have to compensate our server costs. Thank you in advance. Project price only 1 $
You can buy this project and download/modify it how often you want.
/**
* 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.hadoop.hive.ql.optimizer.calcite;
import java.util.ArrayList;
import java.util.HashMap;
import java.util.HashSet;
import java.util.List;
import java.util.Map;
import java.util.Set;
import java.util.concurrent.atomic.AtomicInteger;
import org.apache.calcite.plan.RelOptAbstractTable;
import org.apache.calcite.plan.RelOptSchema;
import org.apache.calcite.plan.RelOptUtil.InputFinder;
import org.apache.calcite.rel.RelCollation;
import org.apache.calcite.rel.RelCollationTraitDef;
import org.apache.calcite.rel.RelDistribution;
import org.apache.calcite.rel.RelFieldCollation;
import org.apache.calcite.rel.RelFieldCollation.Direction;
import org.apache.calcite.rel.RelNode;
import org.apache.calcite.rel.logical.LogicalTableScan;
import org.apache.calcite.rel.type.RelDataType;
import org.apache.calcite.rel.type.RelDataTypeField;
import org.apache.calcite.rex.RexNode;
import org.apache.calcite.util.ImmutableBitSet;
import org.apache.commons.logging.Log;
import org.apache.commons.logging.LogFactory;
import org.apache.hadoop.hive.common.StatsSetupConst;
import org.apache.hadoop.hive.conf.HiveConf;
import org.apache.hadoop.hive.metastore.api.FieldSchema;
import org.apache.hadoop.hive.metastore.api.Order;
import org.apache.hadoop.hive.ql.exec.ColumnInfo;
import org.apache.hadoop.hive.ql.metadata.HiveException;
import org.apache.hadoop.hive.ql.metadata.Partition;
import org.apache.hadoop.hive.ql.metadata.Table;
import org.apache.hadoop.hive.ql.metadata.VirtualColumn;
import org.apache.hadoop.hive.ql.optimizer.calcite.translator.ExprNodeConverter;
import org.apache.hadoop.hive.ql.optimizer.ppr.PartitionPruner;
import org.apache.hadoop.hive.ql.parse.BaseSemanticAnalyzer;
import org.apache.hadoop.hive.ql.parse.PrunedPartitionList;
import org.apache.hadoop.hive.ql.plan.ColStatistics;
import org.apache.hadoop.hive.ql.plan.ExprNodeDesc;
import org.apache.hadoop.hive.ql.plan.Statistics;
import org.apache.hadoop.hive.ql.stats.StatsUtils;
import com.google.common.collect.ImmutableList;
import com.google.common.collect.ImmutableMap;
import com.google.common.collect.Lists;
public class RelOptHiveTable extends RelOptAbstractTable {
private final Table hiveTblMetadata;
private final ImmutableList hiveNonPartitionCols;
private final ImmutableList hivePartitionCols;
private final ImmutableMap hiveNonPartitionColsMap;
private final ImmutableMap hivePartitionColsMap;
private final ImmutableList hiveVirtualCols;
private final int noOfNonVirtualCols;
final HiveConf hiveConf;
private double rowCount = -1;
Map hiveColStatsMap = new HashMap();
PrunedPartitionList partitionList;
Map partitionCache;
AtomicInteger noColsMissingStats;
protected static final Log LOG = LogFactory
.getLog(RelOptHiveTable.class
.getName());
public RelOptHiveTable(RelOptSchema calciteSchema, String qualifiedTblName,
RelDataType rowType, Table hiveTblMetadata, List hiveNonPartitionCols,
List hivePartitionCols, List hiveVirtualCols, HiveConf hconf,
Map partitionCache, AtomicInteger noColsMissingStats) {
super(calciteSchema, qualifiedTblName, rowType);
this.hiveTblMetadata = hiveTblMetadata;
this.hiveNonPartitionCols = ImmutableList.copyOf(hiveNonPartitionCols);
this.hiveNonPartitionColsMap = HiveCalciteUtil.getColInfoMap(hiveNonPartitionCols, 0);
this.hivePartitionCols = ImmutableList.copyOf(hivePartitionCols);
this.hivePartitionColsMap = HiveCalciteUtil.getColInfoMap(hivePartitionCols, hiveNonPartitionColsMap.size());
this.noOfNonVirtualCols = hiveNonPartitionCols.size() + hivePartitionCols.size();
this.hiveVirtualCols = ImmutableList.copyOf(hiveVirtualCols);
this.hiveConf = hconf;
this.partitionCache = partitionCache;
this.noColsMissingStats = noColsMissingStats;
}
public RelOptHiveTable copy(RelDataType newRowType) {
// 1. Build map of column name to col index of original schema
// Assumption: Hive Table can not contain duplicate column names
Map nameToColIndxMap = new HashMap();
for (RelDataTypeField f : this.rowType.getFieldList()) {
nameToColIndxMap.put(f.getName(), f.getIndex());
}
// 2. Build nonPart/Part/Virtual column info for new RowSchema
List newHiveNonPartitionCols = new ArrayList();
List newHivePartitionCols = new ArrayList();
List newHiveVirtualCols = new ArrayList();
Map virtualColInfoMap = HiveCalciteUtil.getVColsMap(this.hiveVirtualCols,
this.noOfNonVirtualCols);
Integer originalColIndx;
ColumnInfo cInfo;
VirtualColumn vc;
for (RelDataTypeField f : newRowType.getFieldList()) {
originalColIndx = nameToColIndxMap.get(f.getName());
if ((cInfo = hiveNonPartitionColsMap.get(originalColIndx)) != null) {
newHiveNonPartitionCols.add(new ColumnInfo(cInfo));
} else if ((cInfo = hivePartitionColsMap.get(originalColIndx)) != null) {
newHivePartitionCols.add(new ColumnInfo(cInfo));
} else if ((vc = virtualColInfoMap.get(originalColIndx)) != null) {
newHiveVirtualCols.add(vc);
} else {
throw new RuntimeException("Copy encountered a column not seen in original TS");
}
}
// 3. Build new Table
return new RelOptHiveTable(this.schema, this.name, newRowType,
this.hiveTblMetadata, newHiveNonPartitionCols, newHivePartitionCols, newHiveVirtualCols,
this.hiveConf, this.partitionCache, this.noColsMissingStats);
}
@Override
public boolean isKey(ImmutableBitSet arg0) {
return false;
}
@Override
public RelNode toRel(ToRelContext context) {
return new LogicalTableScan(context.getCluster(), this);
}
@Override
public T unwrap(Class arg0) {
return arg0.isInstance(this) ? arg0.cast(this) : null;
}
@Override
public List getCollationList() {
ImmutableList.Builder collationList = new ImmutableList.Builder();
for (Order sortColumn : this.hiveTblMetadata.getSortCols()) {
for (int i=0; i()
.add(RelCollationTraitDef.INSTANCE.canonize(
new HiveRelCollation(collationList.build())))
.build();
}
@Override
public RelDistribution getDistribution() {
ImmutableList.Builder columnPositions = new ImmutableList.Builder();
for (String bucketColumn : this.hiveTblMetadata.getBucketCols()) {
for (int i=0; i rowCounts = StatsUtils.getBasicStatForPartitions(hiveTblMetadata,
partitionList.getNotDeniedPartns(), StatsSetupConst.ROW_COUNT);
rowCount = StatsUtils.getSumIgnoreNegatives(rowCounts);
} else {
rowCount = StatsUtils.getNumRows(hiveTblMetadata);
}
}
if (rowCount == -1)
noColsMissingStats.getAndIncrement();
return rowCount;
}
public Table getHiveTableMD() {
return hiveTblMetadata;
}
private String getColNamesForLogging(Set colLst) {
StringBuffer sb = new StringBuffer();
boolean firstEntry = true;
for (String colName : colLst) {
if (firstEntry) {
sb.append(colName);
firstEntry = false;
} else {
sb.append(", " + colName);
}
}
return sb.toString();
}
public void computePartitionList(HiveConf conf, RexNode pruneNode) {
try {
if (!hiveTblMetadata.isPartitioned() || pruneNode == null
|| InputFinder.bits(pruneNode).length() == 0) {
// there is no predicate on partitioning column, we need all partitions
// in this case.
partitionList = PartitionPruner.prune(hiveTblMetadata, null, conf, getName(),
partitionCache);
return;
}
// We have valid pruning expressions, only retrieve qualifying partitions
ExprNodeDesc pruneExpr = pruneNode.accept(new ExprNodeConverter(getName(), getRowType(),
HiveCalciteUtil.getInputRefs(pruneNode), this.getRelOptSchema().getTypeFactory()));
partitionList = PartitionPruner.prune(hiveTblMetadata, pruneExpr, conf, getName(),
partitionCache);
} catch (HiveException he) {
throw new RuntimeException(he);
}
}
private void updateColStats(Set projIndxLst, boolean allowNullColumnForMissingStats) {
List nonPartColNamesThatRqrStats = new ArrayList();
List nonPartColIndxsThatRqrStats = new ArrayList();
List partColNamesThatRqrStats = new ArrayList();
List partColIndxsThatRqrStats = new ArrayList();
Set colNamesFailedStats = new HashSet();
// 1. Separate required columns to Non Partition and Partition Cols
ColumnInfo tmp;
for (Integer pi : projIndxLst) {
if (hiveColStatsMap.get(pi) == null) {
if ((tmp = hiveNonPartitionColsMap.get(pi)) != null) {
nonPartColNamesThatRqrStats.add(tmp.getInternalName());
nonPartColIndxsThatRqrStats.add(pi);
} else if ((tmp = hivePartitionColsMap.get(pi)) != null) {
partColNamesThatRqrStats.add(tmp.getInternalName());
partColIndxsThatRqrStats.add(pi);
} else {
noColsMissingStats.getAndIncrement();
String logMsg = "Unable to find Column Index: " + pi + ", in "
+ hiveTblMetadata.getCompleteName();
LOG.error(logMsg);
throw new RuntimeException(logMsg);
}
}
}
if (null == partitionList) {
// We could be here either because its an unpartitioned table or because
// there are no pruning predicates on a partitioned table.
computePartitionList(hiveConf, null);
}
// 2. Obtain Col Stats for Non Partition Cols
if (nonPartColNamesThatRqrStats.size() > 0) {
List hiveColStats;
if (!hiveTblMetadata.isPartitioned()) {
// 2.1 Handle the case for unpartitioned table.
hiveColStats = StatsUtils.getTableColumnStats(hiveTblMetadata, hiveNonPartitionCols,
nonPartColNamesThatRqrStats);
// 2.1.1 Record Column Names that we needed stats for but couldn't
if (hiveColStats == null) {
colNamesFailedStats.addAll(nonPartColNamesThatRqrStats);
} else if (hiveColStats.size() != nonPartColNamesThatRqrStats.size()) {
Set setOfFiledCols = new HashSet(nonPartColNamesThatRqrStats);
Set setOfObtainedColStats = new HashSet();
for (ColStatistics cs : hiveColStats) {
setOfObtainedColStats.add(cs.getColumnName());
}
setOfFiledCols.removeAll(setOfObtainedColStats);
colNamesFailedStats.addAll(setOfFiledCols);
}
} else {
// 2.2 Obtain col stats for partitioned table.
try {
if (partitionList.getNotDeniedPartns().isEmpty()) {
// no need to make a metastore call
rowCount = 0;
hiveColStats = new ArrayList();
for (String c : nonPartColNamesThatRqrStats) {
// add empty stats object for each column
hiveColStats.add(new ColStatistics(c, null));
}
colNamesFailedStats.clear();
} else {
Statistics stats = StatsUtils.collectStatistics(hiveConf, partitionList,
hiveTblMetadata, hiveNonPartitionCols, nonPartColNamesThatRqrStats,
nonPartColNamesThatRqrStats, true, true);
rowCount = stats.getNumRows();
hiveColStats = new ArrayList();
for (String c : nonPartColNamesThatRqrStats) {
ColStatistics cs = stats.getColumnStatisticsFromColName(c);
if (cs != null) {
hiveColStats.add(cs);
} else {
colNamesFailedStats.add(c);
}
}
}
} catch (HiveException e) {
String logMsg = "Collecting stats failed.";
LOG.error(logMsg, e);
throw new RuntimeException(logMsg, e);
}
}
if (hiveColStats != null && hiveColStats.size() == nonPartColNamesThatRqrStats.size()) {
for (int i = 0; i < hiveColStats.size(); i++) {
hiveColStatsMap.put(nonPartColIndxsThatRqrStats.get(i), hiveColStats.get(i));
}
}
}
// 3. Obtain Stats for Partition Cols
if (colNamesFailedStats.isEmpty() && !partColNamesThatRqrStats.isEmpty()) {
ColStatistics cStats = null;
for (int i = 0; i < partColNamesThatRqrStats.size(); i++) {
cStats = new ColStatistics(partColNamesThatRqrStats.get(i),
hivePartitionColsMap.get(partColIndxsThatRqrStats.get(i)).getTypeName());
cStats.setCountDistint(getDistinctCount(partitionList.getPartitions(),
partColNamesThatRqrStats.get(i)));
hiveColStatsMap.put(partColIndxsThatRqrStats.get(i), cStats);
}
}
// 4. Warn user if we could get stats for required columns
if (!colNamesFailedStats.isEmpty()) {
String logMsg = "No Stats for " + hiveTblMetadata.getCompleteName() + ", Columns: "
+ getColNamesForLogging(colNamesFailedStats);
noColsMissingStats.getAndAdd(colNamesFailedStats.size());
if (allowNullColumnForMissingStats) {
LOG.warn(logMsg);
} else {
LOG.error(logMsg);
throw new RuntimeException(logMsg);
}
}
}
private int getDistinctCount(Set partitions, String partColName) {
Set distinctVals = new HashSet(partitions.size());
for (Partition partition : partitions) {
distinctVals.add(partition.getSpec().get(partColName));
}
return distinctVals.size();
}
public List getColStat(List projIndxLst) {
return getColStat(projIndxLst, false);
}
public List getColStat(List projIndxLst, boolean allowNullColumnForMissingStats) {
List colStatsBldr = Lists.newArrayList();
if (projIndxLst != null) {
updateColStats(new HashSet(projIndxLst), allowNullColumnForMissingStats);
for (Integer i : projIndxLst) {
colStatsBldr.add(hiveColStatsMap.get(i));
}
} else {
List pILst = new ArrayList();
for (Integer i = 0; i < noOfNonVirtualCols; i++) {
pILst.add(i);
}
updateColStats(new HashSet(pILst), allowNullColumnForMissingStats);
for (Integer pi : pILst) {
colStatsBldr.add(hiveColStatsMap.get(pi));
}
}
return colStatsBldr;
}
/*
* use to check if a set of columns are all partition columns. true only if: -
* all columns in BitSet are partition columns.
*/
public boolean containsPartitionColumnsOnly(ImmutableBitSet cols) {
for (int i = cols.nextSetBit(0); i >= 0; i++, i = cols.nextSetBit(i + 1)) {
if (!hivePartitionColsMap.containsKey(i)) {
return false;
}
}
return true;
}
public List getVirtualCols() {
return this.hiveVirtualCols;
}
public List getPartColumns() {
return this.hivePartitionCols;
}
public List getNonPartColumns() {
return this.hiveNonPartitionCols;
}
public int getNoOfNonVirtualCols() {
return noOfNonVirtualCols;
}
public Map getPartColInfoMap() {
return hivePartitionColsMap;
}
public Map getNonPartColInfoMap() {
return hiveNonPartitionColsMap;
}
}