Class GroupedAggregateResultSet

java.lang.Object

org.apache.derby.impl.sql.execute.BasicNoPutResultSetImpl

org.apache.derby.impl.sql.execute.NoPutResultSetImpl

org.apache.derby.impl.sql.execute.GenericAggregateResultSet

org.apache.derby.impl.sql.execute.GroupedAggregateResultSet

All Implemented Interfaces:

CursorResultSet, NoPutResultSet, ResultSet, RowLocationRetRowSource, RowSource

Direct Known Subclasses:

DistinctGroupedAggregateResultSet

class GroupedAggregateResultSet
extends GenericAggregateResultSet
implements CursorResultSet

This ResultSet evaluates grouped, non distinct aggregates. It will scan the entire source result set and calculate the grouped aggregates when scanning the source during the first call to next(). The implementation is capable of computing multiple levels of grouping in a single result set (this is requested using GROUP BY ROLLUP). This implementation has 3 variations, which it chooses according to the following rules: - If the data are guaranteed to arrive already in sorted order, we make a single pass over the data, computing the aggregates in-line as the data are read. - If the statement requests either multiple ROLLUP levels, or a DISTINCT grouping, then the data are first sorted, then we make a single pass over the data as above. - Otherwise, the data are sorted, and a SortObserver is used to compute the aggregations inside the sort, and the results are read back directly from the sorter. Note that, as of the introduction of the ROLLUP support, we no longer ALWAYS compute the aggregates using a SortObserver, which is an arrangement by which the sorter calls back into the aggregates during the sort process each time it consolidates two rows with the same sort key. Using aggregate sort observers is an efficient technique, but it was complex to extend it to the ROLLUP case, so to simplify the code we just have one path for both already-sorted and un-sorted data sources in the ROLLUP case.

Nested Class Summary

Nested classes/interfaces inherited from class BasicNoPutResultSetImpl

BasicNoPutResultSetImpl.FieldComparator

Field Summary

Fields

Modifier and Type	Field	Description
private List<List<Set<DataValueDescriptor>>>	distinctValues
private List<ExecRow>	finishedResults
private long	genericSortId
boolean	hasDistinctAggregate
boolean	isInSortedOrder
private int	maxRowSize
private int	numDistinctAggs
private ColumnOrdering[]	order
private ExecIndexRow[]	resultRows
private boolean	resultsComplete
private boolean	rollup
int	rowsInput
int	rowsReturned
private ScanController	scanController
Properties	sortProperties
private ExecIndexRow	sortResultRow
private ExecIndexRow	sourceExecIndexRow
private TransactionController	tc
private boolean	usingAggregateObserver

Fields inherited from class GenericAggregateResultSet

aggInfoList, aggregates, originalSource, source

Fields inherited from class NoPutResultSetImpl

checkNullCols, clonedExecRow, cncLen, resultSetNumber, targetResultSet

Fields inherited from class BasicNoPutResultSetImpl

activation, beginTime, closeTime, compactRow, constructorTime, currentRow, endExecutionTime, finished, isOpen, isTopResultSet, nextTime, numOpens, openTime, optimizerEstimatedCost, optimizerEstimatedRowCount, resultDescription, rowsFiltered, rowsSeen, startExecutionTime, subqueryTrackingArray

Fields inherited from interface NoPutResultSet

ABSOLUTE, FIRST, LAST, NEXT, PREVIOUS, RELATIVE, TEMPORARY_RESULT_SET_NUMBER

Fields inherited from interface ResultSet

CURRENT_RESULTSET_ONLY, ENTIRE_RESULTSET_TREE, ISAFTERLAST, ISBEFOREFIRST, ISFIRST, ISLAST

Constructor Summary

Constructors

Constructor	Description
GroupedAggregateResultSet(NoPutResultSet s, boolean isInSortedOrder, int aggregateItem, int orderingItem, Activation a, int ra, int maxRowSize, int resultSetNumber, double optimizerEstimatedRowCount, double optimizerEstimatedCost, boolean isRollup)	Constructor

Method Summary

Modifier and Type	Method	Description
void	close()	If the result set has been opened, close the open scan.
void	closeSource()	Close the source of whatever we have been scanning.
private ExecRow	finalizeResults()
ExecRow	getCurrentRow()	This result set has its row from the last fetch done.
ExecRow	getNextRowCore()	Return the next row.
private ExecIndexRow	getNextRowFromRS()	Get the next output row for processing
private ExecIndexRow	getRowFromResultSet()	Get a row from the input result set.
private ExecIndexRow	getRowFromSorter()	Get a row from the sorter.
RowLocation	getRowLocation()	This result set has its row location from the last fetch done.
long	getTimeSpent(int type)	Return the total amount of time spent in this ResultSet
private void	initializeDistinctMaps(int r, boolean allocate)
private void	initializeVectorAggregation(ExecRow row)	Run the aggregator initialization method for each aggregator in the row.
private ScanController	loadSorter()	Load up the sorter.
private ExecRow	makeCurrent(Object row)
private void	mergeVectorAggregates(ExecRow newRow, ExecRow currRow, int level)	Run the aggregator merge method for each aggregator in the row.
private int	numGCols()	Return the number of grouping columns.
void	openCore()	Open the scan.
private int	sameGroupingValues(ExecRow currRow, ExecRow newRow)	Return whether or not the new row has the same values for the grouping columns as the current row.
private void	setRollupColumnsToNull(ExecRow row, int resultNum)

Methods inherited from class GenericAggregateResultSet

finish, finishAggregation, getRowTemplate, getSortAggregators, toXML

Methods inherited from class NoPutResultSetImpl

clearCurrentRow, clearOrderableCache, closeRowSource, getCursorName, getNextRowFromRowSource, getResultDescription, getValidColumns, isForUpdate, markRowAsDeleted, needsRowLocation, needsRowLocationForDeferredCheckConstraints, needsToClone, offendingRowLocation, positionScanAtRowLocation, printQualifiers, resultSetNumber, rowLocation, setCurrentRow, setHasDeferrableChecks, setNeedsRowLocation, setTargetResultSet, skipRow, skipScan, unpackHashValue, updateRow

Methods inherited from class BasicNoPutResultSetImpl

addWarning, attachStatementContext, checkCancellationFlag, checkRowPosition, childrenToXML, cleanUp, dumpTimeStats, finishAndRTS, getAbsoluteRow, getActivation, getAutoGeneratedKeysResultset, getBeginExecutionTimestamp, getCompactRow, getCurrentTimeMillis, getElapsedMillis, getEndExecutionTimestamp, getEstimatedRowCount, getExecuteTime, getExecutionFactory, getFirstRow, getLanguageConnectionContext, getLastRow, getNextRow, getPointOfAttachment, getPreviousRow, getRelativeRow, getRowNumber, getScanIsolationLevel, getSubqueryTrackingArray, getTransactionController, getWarnings, isClosed, isXplainOnlyMode, markAsTopResultSet, modifiedRowCount, open, recordConstructorTime, reopenCore, requiresRelocking, returnsRows, setAfterLastRow, setBeforeFirstRow, setCompactRow, setCompatRow, toXML

Methods inherited from class Object

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

Methods inherited from interface ResultSet

addWarning, checkRowPosition, cleanUp, clearCurrentRow, finish, getAbsoluteRow, getActivation, getAutoGeneratedKeysResultset, getBeginExecutionTimestamp, getCursorName, getEndExecutionTimestamp, getExecuteTime, getFirstRow, getLastRow, getNextRow, getPreviousRow, getRelativeRow, getResultDescription, getRowNumber, getSubqueryTrackingArray, getWarnings, isClosed, modifiedRowCount, open, returnsRows, setAfterLastRow, setBeforeFirstRow, toXML

Field Details

rowsInput

public int rowsInput

rowsReturned

public int rowsReturned

order

private ColumnOrdering[] order

hasDistinctAggregate

public boolean hasDistinctAggregate

isInSortedOrder

public boolean isInSortedOrder

numDistinctAggs

private int numDistinctAggs

maxRowSize

private int maxRowSize

scanController

private ScanController scanController

sourceExecIndexRow

private ExecIndexRow sourceExecIndexRow

sortResultRow

private ExecIndexRow sortResultRow

resultsComplete

private boolean resultsComplete

finishedResults

private List<ExecRow> finishedResults

resultRows

private ExecIndexRow[] resultRows

distinctValues

private List<List<Set<DataValueDescriptor>>> distinctValues

rollup

private boolean rollup

usingAggregateObserver

private boolean usingAggregateObserver

genericSortId

private long genericSortId

tc

private TransactionController tc

sortProperties

public Properties sortProperties

Constructor Details

GroupedAggregateResultSet

GroupedAggregateResultSet(NoPutResultSet s,
 boolean isInSortedOrder,
 int aggregateItem,
 int orderingItem,
 Activation a,
 int ra,
 int maxRowSize,
 int resultSetNumber,
 double optimizerEstimatedRowCount,
 double optimizerEstimatedCost,
 boolean isRollup)
                   throws StandardException

Constructor

Parameters:

s - input result set

isInSortedOrder - true if the source results are in sorted order

aggregateItem - indicates the number of the SavedObject off of the PreparedStatement that holds the AggregatorInfoList used by this routine.

orderingItem - indicates the number of the SavedObject off of the PreparedStatement that holds the ColumOrdering array used by this routine

a - activation

ra - saved object that builds an empty output row

maxRowSize - approx row size, passed to sorter

resultSetNumber - The resultSetNumber for this result set

Throws:

StandardException - Thrown on error

Method Details

openCore

public void openCore()
              throws StandardException

Open the scan. Load the sorter and prepare to get rows from it.

Specified by:

openCore in interface NoPutResultSet

Throws:

StandardException - thrown if cursor finished.

loadSorter

private ScanController loadSorter()
                           throws StandardException

Load up the sorter. Feed it every row from the source scan. When done, close the source scan and open the sort. Return the sort scan controller.

Returns:

the sort controller

Throws:

StandardException - thrown on failure.

numGCols

private int numGCols()

Return the number of grouping columns. Since some additional sort columns may have been included in the sort for DISTINCT aggregates, this function is used to ignore those columns when computing the grouped results.

getNextRowCore

public ExecRow getNextRowCore()
                       throws StandardException

Return the next row.

Specified by:

getNextRowCore in interface NoPutResultSet

Specified by:

getNextRowCore in class BasicNoPutResultSetImpl

Returns:

the next row in the result

Throws:

StandardException - thrown on failure.

StandardException - ResultSetNotOpen thrown if not yet open.

See Also:

• NoPutResultSet.getNextRowCore()

makeCurrent

private ExecRow makeCurrent(Object row)
                     throws StandardException

Throws:

StandardException

finalizeResults

private ExecRow finalizeResults()
                         throws StandardException

Throws:

StandardException

sameGroupingValues

private int sameGroupingValues(ExecRow currRow,
 ExecRow newRow)
                        throws StandardException

Return whether or not the new row has the same values for the grouping columns as the current row. (This allows us to process in-order group bys without a sorter.)

Parameters:

currRow - The current row.

newRow - The new row.

Returns:

The order index number which first distinguished these rows, or order.length if the rows match.

Throws:

StandardException - thrown on failure to get row location

close

public void close()
           throws StandardException

If the result set has been opened, close the open scan.

Specified by:

close in interface ResultSet

Overrides:

close in class NoPutResultSetImpl

Throws:

StandardException - thrown on error

getTimeSpent

public long getTimeSpent(int type)

Return the total amount of time spent in this ResultSet

Specified by:

getTimeSpent in interface ResultSet

Parameters:

type - CURRENT_RESULTSET_ONLY - time spent only in this ResultSet ENTIRE_RESULTSET_TREE - time spent in this ResultSet and below.

Returns:

long The total amount of time spent (in milliseconds).

getRowLocation

public RowLocation getRowLocation()
                           throws StandardException

This result set has its row location from the last fetch done. If the cursor is closed, a null is returned.

Specified by:

getRowLocation in interface CursorResultSet

Returns:

the row location of the current cursor row.

Throws:

StandardException - thrown on failure to get row location

See Also:

• CursorResultSet

getCurrentRow

public ExecRow getCurrentRow()
                      throws StandardException

This result set has its row from the last fetch done. If the cursor is closed, a null is returned.

Specified by:

getCurrentRow in interface CursorResultSet

Returns:

the last row returned;

Throws:

StandardException - thrown on failure.

See Also:

• CursorResultSet

getNextRowFromRS

private ExecIndexRow getNextRowFromRS()
                               throws StandardException

Get the next output row for processing

Throws:

StandardException

getRowFromResultSet

private ExecIndexRow getRowFromResultSet()
                                  throws StandardException

Get a row from the input result set.

Throws:

StandardException

setRollupColumnsToNull

private void setRollupColumnsToNull(ExecRow row,
 int resultNum)
                             throws StandardException

Throws:

StandardException

getRowFromSorter

private ExecIndexRow getRowFromSorter()
                               throws StandardException

Get a row from the sorter. Side effects: sets currentRow.

Throws:

StandardException

closeSource

public void closeSource()
                 throws StandardException

Close the source of whatever we have been scanning.

Throws:

StandardException - thrown on error

initializeVectorAggregation

private void initializeVectorAggregation(ExecRow row)
                                  throws StandardException

Run the aggregator initialization method for each aggregator in the row. Accumulate the input column. WARNING: initializiation performs accumulation -- no need to accumulate a row that has been passed to initialization.

Parameters:

row - the row to initialize

mergeVectorAggregates

private void mergeVectorAggregates(ExecRow newRow,
 ExecRow currRow,
 int level)
                            throws StandardException

Run the aggregator merge method for each aggregator in the row.

Parameters:

newRow - the row to merge

currRow - the row to merge into

initializeDistinctMaps

private void initializeDistinctMaps(int r,
 boolean allocate)
                             throws StandardException

Throws:

StandardException