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

java.lang.Object
  org.apache.derby.impl.sql.execute.BasicNoPutResultSetImpl

All Implemented Interfaces:

NoPutResultSet, ResultSet, RowLocationRetRowSource, RowSource

Direct Known Subclasses:

NoPutResultSetImpl

abstract class BasicNoPutResultSetImpl
extends java.lang.Object
implements NoPutResultSet

Abstract ResultSet for for operations that return rows but do not allow the caller to put data on output pipes. This basic implementation does not include support for an Activiation. See NoPutResultSetImpl.java for an implementaion with support for an activiation.

This abstract class does not define the entire ResultSet interface, but leaves the 'get' half of the interface for subtypes to implement. It is package-visible only, with its methods being public for exposure by its subtypes.

Field Summary

protected Activation	activation
private int[]	baseColumnMap
long	beginTime
long	closeTime
(package private) ExecRow	compactRow
long	constructorTime
protected ExecRow	currentRow
protected long	endExecutionTime
protected boolean	finished
protected boolean	isOpen
protected boolean	isTopResultSet
long	nextTime
int	numOpens
long	openTime
double	optimizerEstimatedCost
double	optimizerEstimatedRowCount
(package private) ResultDescription	resultDescription
int	rowsFiltered
int	rowsSeen
protected long	startExecutionTime
private StatementContext	statementContext
private boolean	statisticsTimingOn
NoPutResultSet[]	subqueryTrackingArray
private TransactionController	tc
private java.sql.SQLWarning	warnings

Fields inherited from interface org.apache.derby.iapi.sql.execute.NoPutResultSet

ABSOLUTE, FIRST, LAST, NEXT, PREVIOUS, RELATIVE

Fields inherited from interface org.apache.derby.iapi.sql.ResultSet

CURRENT_RESULTSET_ONLY, ENTIRE_RESULTSET_TREE, ISAFTERLAST, ISBEFOREFIRST, ISFIRST, ISLAST

Constructor Summary

BasicNoPutResultSetImpl(ResultDescription resultDescription, Activation activation, double optimizerEstimatedRowCount, double optimizerEstimatedCost)
Constructor.

Method Summary

void	addWarning(java.sql.SQLWarning w) Add a warning to this result set.
protected void	attachStatementContext() Attach this result set to the top statement context on the stack.
void	checkCancellationFlag() Checks whether the currently executing statement has been cancelled.
boolean	checkRowPosition(int isType) Determine if the cursor is before the first row in the result set.
void	cleanUp() Clean up on error
protected java.lang.String	dumpTimeStats(java.lang.String indent, java.lang.String subIndent) Dump out the time information for run time stats.
void	finish() Tells the system that there will be no more access to any database information via this result set; in particular, no more calls to open().
protected void	finishAndRTS()
ExecRow	getAbsoluteRow(int row) Returns the row at the absolute position from the query, and returns NULL when there is no such position.
Activation	getActivation()
ResultSet	getAutoGeneratedKeysResultset() ResultSet for rows inserted into the table (contains auto-generated keys columns only)
java.sql.Timestamp	getBeginExecutionTimestamp() Get the Timestamp for the beginning of execution.
protected ExecRow	getCompactRow(ExecRow candidate, FormatableBitSet accessedCols, boolean isKeyed) Get a compacted version of the candidate row according to the columns specified in the bit map.
protected long	getCurrentTimeMillis() Return the current time in milliseconds, if DEBUG and RunTimeStats is on, else return 0.
protected long	getElapsedMillis(long beginTime) Return the elapsed time in milliseconds, between now and the beginTime, if DEBUG and RunTimeStats is on, else return 0.
java.sql.Timestamp	getEndExecutionTimestamp() Get the Timestamp for the end of execution.
double	getEstimatedRowCount() Get the estimated row count from this result set.
long	getExecuteTime() Get the execution time in milliseconds.
(package private) ExecutionFactory	getExecutionFactory() Get a execution factory
ExecRow	getFirstRow() Returns the first row from the query, and returns NULL when there are no rows.
protected LanguageConnectionContext	getLanguageConnectionContext() Cache the language connection context.
ExecRow	getLastRow() Returns the last row from the query, and returns NULL when there are no rows.
ExecRow	getNextRow() Return the requested values computed from the next row (if any) for which the restriction evaluates to true.
abstract ExecRow	getNextRowCore() Return the requested values computed from the next row (if any) for which the restriction evaluates to true.
int	getPointOfAttachment() Return the point of attachment for this subquery.
ExecRow	getPreviousRow() Returns the previous row from the query, and returns NULL when there are no more previous rows.
ExecRow	getRelativeRow(int row) Returns the row at the relative position from the current cursor position, and returns NULL when there is no such position.
ResultDescription	getResultDescription() Returns the description of the table's rows
int	getRowNumber() Returns the row number of the current row.
int	getScanIsolationLevel() Return the isolation level of the scan in the result set.
NoPutResultSet[]	getSubqueryTrackingArray(int numSubqueries) Get the subquery ResultSet tracking array from the top ResultSet.
(package private) TransactionController	getTransactionController() Get the current transaction controller.
java.sql.SQLWarning	getWarnings() Return the set of warnings generated during the execution of this result set.
boolean	isClosed() Report if closed.
boolean	isForUpdate() Is this ResultSet or it's source result set for update This method will be overriden in the inherited Classes if it is true
void	markAsTopResultSet() Mark the ResultSet as the topmost one in the ResultSet tree.
long	modifiedRowCount() Returns the number of rows affected by the statement.
void	open() open a scan on the table. scan parameters are evaluated at each open, so there is probably some way of altering their values...
protected void	recordConstructorTime() Allow sub-classes to record the total time spent in their constructor time.
void	reopenCore() This is the default implementation of reopenCore().
boolean	requiresRelocking() Do we need to relock the row when going to the heap.
int	resultSetNumber() Get the number of this ResultSet, which is guaranteed to be unique within a statement.
boolean	returnsRows() Returns true.
ExecRow	setAfterLastRow() Sets the current position to after the last row and returns NULL because there is no current row.
ExecRow	setBeforeFirstRow() Sets the current position to before the first row and returns NULL because there is no current row.
protected ExecRow	setCompactRow(ExecRow candidateRow, ExecRow compactRow) Copy columns from the candidate row from the store to the given compact row.
protected void	setCompatRow(ExecRow compactRow, DataValueDescriptor[] sourceRow)

Methods inherited from class java.lang.Object

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

Methods inherited from interface org.apache.derby.iapi.sql.execute.NoPutResultSet

markRowAsDeleted, openCore, positionScanAtRowLocation, setCurrentRow, setNeedsRowLocation, setTargetResultSet, updateRow

Methods inherited from interface org.apache.derby.iapi.sql.ResultSet

clearCurrentRow, close, getCursorName, getTimeSpent

Methods inherited from interface org.apache.derby.iapi.store.access.RowLocationRetRowSource

needsRowLocation, rowLocation

Methods inherited from interface org.apache.derby.iapi.store.access.RowSource

closeRowSource, getNextRowFromRowSource, getValidColumns, needsToClone

Field Detail

isOpen

protected boolean isOpen

finished

protected boolean finished

currentRow

protected ExecRow currentRow

isTopResultSet

protected boolean isTopResultSet

warnings

private java.sql.SQLWarning warnings

numOpens

public int numOpens

rowsSeen

public int rowsSeen

rowsFiltered

public int rowsFiltered

startExecutionTime

protected long startExecutionTime

endExecutionTime

protected long endExecutionTime

beginTime

public long beginTime

constructorTime

public long constructorTime

openTime

public long openTime

nextTime

public long nextTime

closeTime

public long closeTime

optimizerEstimatedRowCount

public double optimizerEstimatedRowCount

optimizerEstimatedCost

public double optimizerEstimatedCost

statementContext

private StatementContext statementContext

subqueryTrackingArray

public NoPutResultSet[] subqueryTrackingArray

compactRow

ExecRow compactRow

activation

protected final Activation activation

statisticsTimingOn

private final boolean statisticsTimingOn

resultDescription

ResultDescription resultDescription

tc

private transient TransactionController tc

baseColumnMap

private int[] baseColumnMap

Constructor Detail

BasicNoPutResultSetImpl

BasicNoPutResultSetImpl(ResultDescription resultDescription,
                        Activation activation,
                        double optimizerEstimatedRowCount,
                        double optimizerEstimatedCost)

Constructor.
Sets beginTime for all children to use to measue constructor time.

Parameters:

resultDescription - the result description. May be null.

activation - The activation

optimizerEstimatedRowCount - The optimizer's estimate of the total number of rows for this result set

optimizerEstimatedCost - The optimizer's estimated cost for this result set

Method Detail

recordConstructorTime

protected final void recordConstructorTime()

Allow sub-classes to record the total time spent in their constructor time.

getActivation

public final Activation getActivation()

Specified by:

getActivation in interface ResultSet

reopenCore

public void reopenCore()
                throws StandardException

This is the default implementation of reopenCore(). It simply does a close() followed by an open(). If there are optimizations to be made (caching, etc), this is a good place to do it -- this will be overridden by a number of resultSet imlementations. and SHOULD be overridden by any node that can get between a base table and a join.

Specified by:

reopenCore in interface NoPutResultSet

Throws:

StandardException - thrown if cursor finished.

See Also:

NoPutResultSet.openCore()

getNextRowCore

public abstract ExecRow getNextRowCore()
                                throws StandardException

Description copied from interface: NoPutResultSet

Return the requested values computed from the next row (if any) for which the restriction evaluates to true.

restriction and projection parameters are evaluated for each row.

Specified by:

getNextRowCore in interface NoPutResultSet

Returns:

the next row in the result

Throws:

StandardException - thrown if cursor finished.

See Also:

NoPutResultSet.getNextRowCore()

getPointOfAttachment

public int getPointOfAttachment()

Description copied from interface: NoPutResultSet

Return the point of attachment for this subquery. (Only meaningful for Any and Once ResultSets, which can and will only be at the top of a ResultSet for a subquery.)

Specified by:

getPointOfAttachment in interface NoPutResultSet

Returns:

int Point of attachment (result set number) for this subquery. (-1 if not a subquery - also Sanity violation)

See Also:

NoPutResultSet.getPointOfAttachment()

markAsTopResultSet

public void markAsTopResultSet()

Mark the ResultSet as the topmost one in the ResultSet tree. Useful for closing down the ResultSet on an error.

Specified by:

markAsTopResultSet in interface NoPutResultSet

getScanIsolationLevel

public int getScanIsolationLevel()

Description copied from interface: NoPutResultSet

Return the isolation level of the scan in the result set. Only expected to be called for those ResultSets that contain a scan.

Specified by:

getScanIsolationLevel in interface NoPutResultSet

Returns:

The isolation level of the scan (in TransactionController constants).

See Also:

NoPutResultSet.getScanIsolationLevel()

getEstimatedRowCount

public double getEstimatedRowCount()

Description copied from interface: NoPutResultSet

Get the estimated row count from this result set.

Specified by:

getEstimatedRowCount in interface NoPutResultSet

Returns:

The estimated row count (as a double) from this result set.

See Also:

NoPutResultSet.getEstimatedRowCount()

requiresRelocking

public boolean requiresRelocking()

Description copied from interface: NoPutResultSet

Do we need to relock the row when going to the heap.

Specified by:

requiresRelocking in interface NoPutResultSet

Returns:

Whether or not we need to relock the row when going to the heap.

See Also:

NoPutResultSet.requiresRelocking()

open

public final void open()
                throws StandardException

open a scan on the table. scan parameters are evaluated at each open, so there is probably some way of altering their values... NOTE: This method should only be called on the top ResultSet of a ResultSet tree to ensure that the entire ResultSet tree gets closed down on an error. the openCore() method will be called for all other ResultSets in the tree.

Specified by:

open in interface ResultSet

Throws:

StandardException - thrown if cursor finished.

getAbsoluteRow

public ExecRow getAbsoluteRow(int row)
                       throws StandardException

Returns the row at the absolute position from the query, and returns NULL when there is no such position. (Negative position means from the end of the result set.) Moving the cursor to an invalid position leaves the cursor positioned either before the first row (negative position) or after the last row (positive position). NOTE: An exception will be thrown on 0.

Specified by:

getAbsoluteRow in interface ResultSet

Parameters:

row - The position.

Returns:

The row at the absolute position, or NULL if no such position.

Throws:

StandardException - Thrown on failure

See Also:

Row

getRelativeRow

public ExecRow getRelativeRow(int row)
                       throws StandardException

Returns the row at the relative position from the current cursor position, and returns NULL when there is no such position. (Negative position means toward the beginning of the result set.) Moving the cursor to an invalid position leaves the cursor positioned either before the first row (negative position) or after the last row (positive position). NOTE: 0 is valid. NOTE: An exception is thrown if the cursor is not currently positioned on a row.

Specified by:

getRelativeRow in interface ResultSet

Parameters:

row - The position.

Returns:

The row at the relative position, or NULL if no such position.

Throws:

StandardException - Thrown on failure

See Also:

Row

setBeforeFirstRow

public ExecRow setBeforeFirstRow()
                          throws StandardException

Sets the current position to before the first row and returns NULL because there is no current row.

Specified by:

setBeforeFirstRow in interface ResultSet

Returns:

NULL.

Throws:

StandardException - Thrown on failure

See Also:

Row

checkRowPosition

public boolean checkRowPosition(int isType)
                         throws StandardException

Determine if the cursor is before the first row in the result set.

Specified by:

checkRowPosition in interface ResultSet

Returns:

true if before the first row, false otherwise. Returns false when the result set contains no rows.

Throws:

StandardException - Thrown on error.

getRowNumber

public int getRowNumber()

Returns the row number of the current row. Row numbers start from 1 and go to 'n'. Corresponds to row numbering used to position current row in the result set (as per JDBC).

Specified by:

getRowNumber in interface ResultSet

Returns:

the row number, or 0 if not on a row

getFirstRow

public ExecRow getFirstRow()
                    throws StandardException

Returns the first row from the query, and returns NULL when there are no rows.

Specified by:

getFirstRow in interface ResultSet

Returns:

The first row, or NULL if no rows.

Throws:

StandardException - Thrown on failure

See Also:

Row

getNextRow

public final ExecRow getNextRow()
                         throws StandardException

Return the requested values computed from the next row (if any) for which the restriction evaluates to true.

restriction and projection parameters are evaluated for each row. NOTE: This method should only be called on the top ResultSet of a ResultSet tree to ensure that the entire ResultSet tree gets closed down on an error. the getNextRowCore() method will be called for all other ResultSets in the tree.

Specified by:

getNextRow in interface ResultSet

Returns:

the next row in the result

Throws:

StandardException - thrown on failure.

StandardException - ResultSetNotOpen thrown if not yet open.

See Also:

Row

getPreviousRow

public ExecRow getPreviousRow()
                       throws StandardException

Returns the previous row from the query, and returns NULL when there are no more previous rows.

Specified by:

getPreviousRow in interface ResultSet

Returns:

The previous row, or NULL if no more previous rows.

Throws:

StandardException - Thrown on failure

See Also:

Row

getLastRow

public ExecRow getLastRow()
                   throws StandardException

Returns the last row from the query, and returns NULL when there are no rows.

Specified by:

getLastRow in interface ResultSet

Returns:

The last row, or NULL if no rows.

Throws:

StandardException - Thrown on failure

See Also:

Row

setAfterLastRow

public ExecRow setAfterLastRow()
                        throws StandardException

Sets the current position to after the last row and returns NULL because there is no current row.

Specified by:

setAfterLastRow in interface ResultSet

Returns:

NULL.

Throws:

StandardException - Thrown on failure

See Also:

Row

returnsRows

public boolean returnsRows()

Returns true.

Specified by:

returnsRows in interface ResultSet

Returns:

TRUE if the statement returns rows, FALSE if not.

modifiedRowCount

public final long modifiedRowCount()

Description copied from interface: ResultSet

Returns the number of rows affected by the statement. Only valid of returnsRows() returns false. For other DML statements, it returns the number of rows modified by the statement. For statements that do not affect rows (like DDL statements), it returns zero.

Specified by:

modifiedRowCount in interface ResultSet

Returns:

The number of rows affect by the statement, so far.

cleanUp

public void cleanUp()
             throws StandardException

Clean up on error

Specified by:

cleanUp in interface ResultSet

Throws:

StandardException - Thrown on failure

isClosed

public boolean isClosed()

Report if closed.

Specified by:

isClosed in interface ResultSet

Returns:

true if the ResultSet has been closed.

finish

public void finish()
            throws StandardException

Description copied from interface: ResultSet

Tells the system that there will be no more access to any database information via this result set; in particular, no more calls to open(). Will close the result set if it is not already closed.

Specified by:

finish in interface ResultSet

Throws:

StandardException - on error

finishAndRTS

protected final void finishAndRTS()
                           throws StandardException

Throws:

StandardException - on error

getResultDescription

public ResultDescription getResultDescription()

Returns the description of the table's rows

Specified by:

getResultDescription in interface ResultSet

Returns:

A ResultDescription describing the results of the statement.

getExecuteTime

public long getExecuteTime()

Get the execution time in milliseconds.

Specified by:

getExecuteTime in interface ResultSet

Returns:

long The execution time in milliseconds.

getBeginExecutionTimestamp

public java.sql.Timestamp getBeginExecutionTimestamp()

Get the Timestamp for the beginning of execution.

Specified by:

getBeginExecutionTimestamp in interface ResultSet

Returns:

Timestamp The Timestamp for the beginning of execution.

getEndExecutionTimestamp

public java.sql.Timestamp getEndExecutionTimestamp()

Get the Timestamp for the end of execution.

Specified by:

getEndExecutionTimestamp in interface ResultSet

Returns:

Timestamp The Timestamp for the end of execution.

getSubqueryTrackingArray

public final NoPutResultSet[] getSubqueryTrackingArray(int numSubqueries)

Description copied from interface: ResultSet

Get the subquery ResultSet tracking array from the top ResultSet. (Used for tracking open subqueries when closing down on an error.)

Specified by:

getSubqueryTrackingArray in interface ResultSet

Parameters:

numSubqueries - The size of the array (For allocation on demand.)

Returns:

NoPutResultSet[] Array of NoPutResultSets for subqueries.

See Also:

ResultSet.getSubqueryTrackingArray(int)

getCurrentTimeMillis

protected final long getCurrentTimeMillis()

Return the current time in milliseconds, if DEBUG and RunTimeStats is on, else return 0. (Only pay price of system call if need to.)

Returns:

long Current time in milliseconds.

getAutoGeneratedKeysResultset

public ResultSet getAutoGeneratedKeysResultset()

Description copied from interface: ResultSet

ResultSet for rows inserted into the table (contains auto-generated keys columns only)

Specified by:

getAutoGeneratedKeysResultset in interface ResultSet

Returns:

NoPutResultSet NoPutResultSets for rows inserted into the table.

See Also:

ResultSet.getAutoGeneratedKeysResultset()

getElapsedMillis

protected final long getElapsedMillis(long beginTime)

Return the elapsed time in milliseconds, between now and the beginTime, if DEBUG and RunTimeStats is on, else return 0. (Only pay price of system call if need to.)

Returns:

long Elapsed time in milliseconds.

dumpTimeStats

protected final java.lang.String dumpTimeStats(java.lang.String indent,
                                               java.lang.String subIndent)

Dump out the time information for run time stats.

Returns:

Nothing.

attachStatementContext

protected void attachStatementContext()
                               throws StandardException

Attach this result set to the top statement context on the stack. Result sets can be directly read from the JDBC layer. The JDBC layer will push and pop a statement context around each ResultSet.getNext(). There's no guarantee that the statement context used for the last getNext() will be the context used for the current getNext(). The last statement context may have been popped off the stack and so will not be available for cleanup if an error occurs. To make sure that we will be cleaned up, we always attach ourselves to the top context. The fun and games occur in nested contexts: using JDBC result sets inside user code that is itself invoked from queries or CALL statements.

Throws:

StandardException - thrown if cursor finished.

getLanguageConnectionContext

protected final LanguageConnectionContext getLanguageConnectionContext()

Cache the language connection context. Return it.

Returns:

the language connection context

resultSetNumber

public int resultSetNumber()

Description copied from interface: NoPutResultSet

Get the number of this ResultSet, which is guaranteed to be unique within a statement.

Specified by:

resultSetNumber in interface NoPutResultSet

See Also:

NoPutResultSet.resultSetNumber()

getExecutionFactory

final ExecutionFactory getExecutionFactory()

Get a execution factory

Returns:

the execution factory

getTransactionController

final TransactionController getTransactionController()

Get the current transaction controller.

getCompactRow

protected ExecRow getCompactRow(ExecRow candidate,
                                FormatableBitSet accessedCols,
                                boolean isKeyed)
                         throws StandardException

Get a compacted version of the candidate row according to the columns specified in the bit map. Share the holders between rows. If there is no bit map, use the candidate row as the compact row. Also, create an array of ints mapping base column positions to compact column positions, to make it cheaper to copy columns to the compact row, if we ever have to do it again.

Parameters:

candidate - The row to get the columns from

accessedCols - A bit map of the columns that are accessed in the candidate row

isKeyed - Tells whether to return a ValueRow or an IndexRow

Returns:

A compact row.

Throws:

StandardException

setCompactRow

protected ExecRow setCompactRow(ExecRow candidateRow,
                                ExecRow compactRow)

Copy columns from the candidate row from the store to the given compact row. If there is no column map, just use the candidate row. This method assumes the above method (getCompactRow()) was called first. getCompactRow() sets up the baseColumnMap.

Parameters:

candidateRow - The candidate row from the store

compactRow - The compact row to fill in

Returns:

The compact row to use

setCompatRow

protected final void setCompatRow(ExecRow compactRow,
                                  DataValueDescriptor[] sourceRow)

isForUpdate

public boolean isForUpdate()

Is this ResultSet or it's source result set for update This method will be overriden in the inherited Classes if it is true

Specified by:

isForUpdate in interface NoPutResultSet

Returns:

Whether or not the result set is for update.

checkCancellationFlag

public void checkCancellationFlag()
                           throws StandardException

Checks whether the currently executing statement has been cancelled. This is done by checking the statement's allocated StatementContext object.

Throws:

StandardException

See Also:

StatementContext

addWarning

public final void addWarning(java.sql.SQLWarning w)

Description copied from interface: ResultSet

Add a warning to this result set.

Specified by:

addWarning in interface ResultSet

Parameters:

w - the warning to add

getWarnings

public final java.sql.SQLWarning getWarnings()

Description copied from interface: ResultSet

Return the set of warnings generated during the execution of this result set. The warnings are cleared once this call returns.

Specified by:

getWarnings in interface ResultSet