org.apache.druid.query.groupby.epinephelinae

Class GroupByQueryEngineV2.HashAggregateIterator

java.lang.Object

org.apache.druid.query.groupby.epinephelinae.GroupByQueryEngineV2.GroupByEngineIterator<ByteBuffer>

org.apache.druid.query.groupby.epinephelinae.GroupByQueryEngineV2.HashAggregateIterator

All Implemented Interfaces:

Closeable, AutoCloseable, Iterator<ResultRow>

Enclosing class:

GroupByQueryEngineV2

public static class GroupByQueryEngineV2.HashAggregateIterator
extends GroupByQueryEngineV2.GroupByEngineIterator<ByteBuffer>

Field Summary

Fields

Modifier and Type	Field and Description
protected ByteBuffer	keyBuffer

Fields inherited from class org.apache.druid.query.groupby.epinephelinae.GroupByQueryEngineV2.GroupByEngineIterator

allowMultiValueGrouping, allSingleValueDims, buffer, cursor, delegate, dims, keySerde, maxSelectorFootprint, query, querySpecificConfig, timestamp

Constructor Summary

Constructors

Constructor and Description
HashAggregateIterator(GroupByQuery query, GroupByQueryConfig querySpecificConfig, DruidProcessingConfig processingConfig, Cursor cursor, ByteBuffer buffer, org.joda.time.DateTime fudgeTimestamp, GroupByColumnSelectorPlus[] dims, boolean allSingleValueDims)

Method Summary

Modifier and Type	Method and Description
protected void	aggregateMultiValueDims(Grouper<ByteBuffer> grouper) Grouping dimensions can be multi-valued, and thus the given grouper should handle them properly during aggregation.
protected void	aggregateSingleValueDims(Grouper<ByteBuffer> grouper) Grouping dimensions are all single-valued, and thus the given grouper don't have to worry about multi-valued dimensions.
protected Grouper<ByteBuffer>	newGrouper() Create a new grouper.
protected void	putToRow(ByteBuffer key, ResultRow resultRow) Add the key to the result row.

Methods inherited from class org.apache.druid.query.groupby.epinephelinae.GroupByQueryEngineV2.GroupByEngineIterator

checkIfMultiValueGroupingIsAllowed, close, getSingleValue, hasNext, next, remove

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Methods inherited from interface java.util.Iterator

forEachRemaining

Field Detail

keyBuffer

protected final ByteBuffer keyBuffer

Constructor Detail

HashAggregateIterator

public HashAggregateIterator(GroupByQuery query,
                             GroupByQueryConfig querySpecificConfig,
                             DruidProcessingConfig processingConfig,
                             Cursor cursor,
                             ByteBuffer buffer,
                             @Nullable
                             org.joda.time.DateTime fudgeTimestamp,
                             GroupByColumnSelectorPlus[] dims,
                             boolean allSingleValueDims)

Method Detail

newGrouper

protected Grouper<ByteBuffer> newGrouper()

Description copied from class: GroupByQueryEngineV2.GroupByEngineIterator

Create a new grouper.

Specified by:

newGrouper in class GroupByQueryEngineV2.GroupByEngineIterator<ByteBuffer>

aggregateSingleValueDims

protected void aggregateSingleValueDims(Grouper<ByteBuffer> grouper)

Description copied from class: GroupByQueryEngineV2.GroupByEngineIterator

Grouping dimensions are all single-valued, and thus the given grouper don't have to worry about multi-valued dimensions.

Specified by:

aggregateSingleValueDims in class GroupByQueryEngineV2.GroupByEngineIterator<ByteBuffer>

aggregateMultiValueDims

protected void aggregateMultiValueDims(Grouper<ByteBuffer> grouper)

Description copied from class: GroupByQueryEngineV2.GroupByEngineIterator

Grouping dimensions can be multi-valued, and thus the given grouper should handle them properly during aggregation.

Specified by:

aggregateMultiValueDims in class GroupByQueryEngineV2.GroupByEngineIterator<ByteBuffer>

putToRow

protected void putToRow(ByteBuffer key,
                        ResultRow resultRow)

Description copied from class: GroupByQueryEngineV2.GroupByEngineIterator

Add the key to the result row. Some pre-processing like deserialization might be done for the key before adding to the map.

Specified by:

putToRow in class GroupByQueryEngineV2.GroupByEngineIterator<ByteBuffer>