Package edu.caltech.nanodb.plannodes

Class ThetaJoinNode

java.lang.Object

edu.caltech.nanodb.plannodes.PlanNode

edu.caltech.nanodb.plannodes.ThetaJoinNode

All Implemented Interfaces:

java.lang.Cloneable

Direct Known Subclasses:

NestedLoopJoinNode, SortMergeJoinNode

public abstract class ThetaJoinNode
extends PlanNode

PlanNode representing the FROM clause in a SELECT operation. This is the relational algebra ThetaJoin operator.

Field Summary

Fields

Modifier and Type	Field	Description
JoinType	joinType	The type of the join operation to perform.
protected Schema	leftSchema	The cached schema of the left subplan, used for join-predicate evaluation.
protected java.util.ArrayList<ColumnStats>	leftStats	The cached statistics of the left subplan, used for cost estimation.
Expression	predicate	Join condition.
protected Schema	rightSchema	The cached schema of the right subplan, used for join-predicate evaluation.
protected java.util.ArrayList<ColumnStats>	rightStats	The cached statistics of the right subplan, used for cost estimation.
protected boolean	schemaSwapped	True if the output schema of this node is swapped.

Fields inherited from class edu.caltech.nanodb.plannodes.PlanNode

cost, environment, leftChild, rightChild, schema, stats

Constructor Summary

Constructors

Constructor	Description
ThetaJoinNode(PlanNode leftChild, PlanNode rightChild, JoinType joinType, Expression predicate)	Constructs a ThetaJoinNode that joins the tuples from the left and right subplans, using the specified join type and join predicate.

Method Summary

Modifier and Type	Method	Description
void	initialize()	Do initialization for the join operation.
boolean	isSwapped()	Returns true if the schema is swapped in this theta join node, false otherwise.
protected Tuple	joinTuples(Tuple left, Tuple right)	Combine the left tuple and the right tuple.
protected void	prepareSchemaStats()	This helper method can be used by the PlanNode.prepare() method in subclasses, to compute the output schema and initial stats of the join operation.
void	swap()	Swaps the left child and right child subtrees.

Methods inherited from class edu.caltech.nanodb.plannodes.PlanNode

addParentEnvironmentToPlanTree, cleanUp, clone, duplicate, equals, getCost, getEnvironment, getNextTuple, getSchema, getStats, hashCode, markCurrentPosition, prepare, printNodeTree, printNodeTree, printNodeTree, printNodeTreeToString, printNodeTreeToString, requiresLeftMarking, requiresRightMarking, resetToLastMark, resultsOrderedBy, setEnvironment, supportsMarking, toString

Methods inherited from class java.lang.Object

finalize, getClass, notify, notifyAll, wait, wait, wait

Field Detail

joinType

public JoinType joinType

The type of the join operation to perform. Note that not all join implementations can necessarily implement all possible join types.

predicate

public Expression predicate

Join condition.

leftSchema

protected Schema leftSchema

The cached schema of the left subplan, used for join-predicate evaluation.

leftStats

protected java.util.ArrayList<ColumnStats> leftStats

The cached statistics of the left subplan, used for cost estimation.

rightSchema

protected Schema rightSchema

The cached schema of the right subplan, used for join-predicate evaluation.

rightStats

protected java.util.ArrayList<ColumnStats> rightStats

The cached statistics of the right subplan, used for cost estimation.

schemaSwapped

protected boolean schemaSwapped

True if the output schema of this node is swapped. If this flag is true then tuples from the left subplan will be on the right of joined results and so forth, but if the flag is false then the tuples from the left subplan will be on the left of the joined results. The schema and the statistics values reflect the state of this flag, after PlanNode.prepare() has been called.

Constructor Detail

ThetaJoinNode

public ThetaJoinNode(PlanNode leftChild,
                     PlanNode rightChild,
                     JoinType joinType,
                     Expression predicate)

Constructs a ThetaJoinNode that joins the tuples from the left and right subplans, using the specified join type and join predicate.

Parameters:

leftChild - the left relation

rightChild - the right relation

joinType - the type of join operation to perform

predicate - the join condition

Method Detail

joinTuples

protected Tuple joinTuples(Tuple left,
                           Tuple right)

Combine the left tuple and the right tuple. If schemaSwapped is set to true, the tuples are copied in the opposite order. This can only happen if swap() was called an odd number of times, switching the left and right subtrees.

Parameters:

left - the left tuple

right - the right tuple

Returns:

the combined tuple

initialize

public void initialize()

Do initialization for the join operation. Resets state variables. Initialize both children.

Overrides:

initialize in class PlanNode

prepareSchemaStats

protected void prepareSchemaStats()

This helper method can be used by the PlanNode.prepare() method in subclasses, to compute the output schema and initial stats of the join operation. This method is provided because it takes the schemaSwapped flag into account when ordering the schema and stats.

swap

public void swap()

Swaps the left child and right child subtrees. Ensures that the schema of the node does not change in the swap, so that this is still a valid query plan.

isSwapped

public boolean isSwapped()

Returns true if the schema is swapped in this theta join node, false otherwise.

Returns:

true if the schema is swapped in this theta join node, false otherwise.