Package com.apple.foundationdb.record.query.plan.plans

Class RecordQueryCoveringIndexPlan

java.lang.Object

com.apple.foundationdb.record.query.plan.plans.RecordQueryCoveringIndexPlan

All Implemented Interfaces:

PlanHashable, QueryPlan<FDBQueriedRecord<Message>>, RecordQueryPlan, RecordQueryPlanWithNoChildren, Bindable, Correlated<RelationalExpression>, PlannerGraphRewritable, RelationalExpression

@API(INTERNAL)
public class RecordQueryCoveringIndexPlan
extends Object
implements RecordQueryPlanWithNoChildren

A query plan that reconstructs records from the entries in a covering index.

Nested Class Summary

Nested classes/interfaces inherited from interface com.apple.foundationdb.record.query.plan.temp.Correlated

Correlated.BoundEquivalence

Nested classes/interfaces inherited from interface com.apple.foundationdb.record.PlanHashable

PlanHashable.PlanHashKind

Nested classes/interfaces inherited from interface com.apple.foundationdb.record.query.plan.temp.RelationalExpression

RelationalExpression.CombineFunction<R,S>, RelationalExpression.CombinePredicate

Field Summary

Fields inherited from interface com.apple.foundationdb.record.query.plan.plans.QueryPlan

UNKNOWN_MAX_CARDINALITY

Constructor Summary

Constructors

Constructor	Description
RecordQueryCoveringIndexPlan(RecordQueryPlanWithIndex indexPlan, String recordTypeName, AvailableFields availableFields, IndexKeyValueToPartialRecord toRecord)

Method Summary

Modifier and Type	Method	Description
boolean	equals(Object other)
boolean	equalsWithoutChildren(RelationalExpression otherExpression, AliasMap equivalencesMap)
<M extends Message> RecordCursor<FDBQueriedRecord<M>>	execute(FDBRecordStoreBase<M> store, EvaluationContext context, byte[] continuation, ExecuteProperties executeProperties)	Execute this query plan.
AvailableFields	getAvailableFields()
int	getComplexity()	Returns an integer representing the "complexity" of the generated plan.
Set<CorrelationIdentifier>	getCorrelatedTo()	Returns the set of CorrelationIdentifiers this entity is correlated to.
String	getIndexName()
RecordQueryPlanWithIndex	getIndexPlan()
List<? extends Quantifier>	getQuantifiers()	Return an iterator of references to the children of this planner expression.
IndexScanType	getScanType()
Set<String>	getUsedIndexes()	Returns a set of names of the indexes used by this plan (and its sub-plans).
boolean	hasFullRecordScan()	Indicates whether this plan (or one of its components) must perform a scan over all records in the store directly rather than going through a secondary index.
int	hashCode()
int	hashCodeWithoutChildren()
boolean	hasIndexScan(String indexName)	Indicates whether this plan scans the given index.
boolean	hasLoadBykeys()	Indicates whether this plan (or one of its components) loads records by their primary key.
boolean	hasRecordScan()	Indicates whether this plan (or one of its components) scans at least a subset of the record range directly rather than going through a secondary index.
<M extends Message> Function<IndexEntry,FDBQueriedRecord<M>>	indexEntryToQueriedRecord(FDBRecordStoreBase<M> store)
boolean	isReverse()	Indicates whether this plan will return values in "reverse" order from the natural order of results of this plan.
boolean	isStrictlySorted()	Indicates whether the ordering of records from this plan is fully accounted for by its requested sort.
void	logPlanStructure(StoreTimer timer)	Adds one to an appropriate StoreTimer counter for each plan and subplan of this plan, allowing tracking of which plans are being chosen (e.g.
int	maxCardinality(RecordMetaData metaData)	Indicates how many records this plan could possibly return.
int	planHash(PlanHashable.PlanHashKind hashKind)	Return a hash similar to hashCode, but with the additional guarantee that is is stable across JVMs.
RecordQueryCoveringIndexPlan	rebase(AliasMap translationMap)	Rebases this and all other objects this objects is composed of using a given translation map.
PlannerGraph	rewritePlannerGraph(List<? extends PlannerGraph> childGraphs)	Method to rewrite the planner graph.
RecordQueryCoveringIndexPlan	strictlySorted()	Return a copy of this plan that has the QueryPlan.isStrictlySorted() property.
String	toString()

Methods inherited from class java.lang.Object

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

Methods inherited from interface com.apple.foundationdb.record.PlanHashable

planHash

Methods inherited from interface com.apple.foundationdb.record.query.plan.plans.QueryPlan

execute, execute

Methods inherited from interface com.apple.foundationdb.record.query.plan.plans.RecordQueryPlan

accept, execute, execute, execute, getQueryPlanChildren, structuralEquals, structuralEquals, structuralHashCode

Methods inherited from interface com.apple.foundationdb.record.query.plan.plans.RecordQueryPlanWithNoChildren

getChildren

Methods inherited from interface com.apple.foundationdb.record.query.plan.temp.RelationalExpression

acceptPropertyVisitor, adjustMatch, bindIdentities, bindTo, canCorrelate, compensate, enumerateUnboundCorrelatedTo, exactlySubsumedBy, findMatches, getResultValues, hasIncompatibleBoundQuantifiers, hasUnboundQuantifiers, match, match, semanticEquals, semanticEquals, semanticHashCode, show, subsumedBy

Constructor Detail

RecordQueryCoveringIndexPlan

public RecordQueryCoveringIndexPlan(@Nonnull
                                    RecordQueryPlanWithIndex indexPlan,
                                    @Nonnull
                                    String recordTypeName,
                                    @Nonnull
                                    AvailableFields availableFields,
                                    @Nonnull
                                    IndexKeyValueToPartialRecord toRecord)

Method Detail

execute

@Nonnull
public <M extends Message> RecordCursor<FDBQueriedRecord<M>> execute(@Nonnull
                                                                     FDBRecordStoreBase<M> store,
                                                                     @Nonnull
                                                                     EvaluationContext context,
                                                                     @Nullable
                                                                     byte[] continuation,
                                                                     @Nonnull
                                                                     ExecuteProperties executeProperties)

Description copied from interface: RecordQueryPlan

Execute this query plan.

Specified by:

execute in interface RecordQueryPlan

Type Parameters:

M - type used to represent stored records

Parameters:

store - record store from which to fetch records

context - evaluation context containing parameter bindings

continuation - continuation from a previous execution of this same plan

executeProperties - limits on execution

Returns:

a cursor of records that match the query criteria

indexEntryToQueriedRecord

@Nonnull
@API(INTERNAL)
public <M extends Message> Function<IndexEntry,FDBQueriedRecord<M>> indexEntryToQueriedRecord(@Nonnull
                                                                                                    FDBRecordStoreBase<M> store)

getIndexPlan

@Nonnull
public RecordQueryPlanWithIndex getIndexPlan()

getIndexName

@Nonnull
public String getIndexName()

getScanType

@Nonnull
public IndexScanType getScanType()

isReverse

public boolean isReverse()

Description copied from interface: QueryPlan

Indicates whether this plan will return values in "reverse" order from the natural order of results of this plan. The return value is true if the plan returns elements in descending order and false if the elements are returned in ascending order.

Specified by:

isReverse in interface QueryPlan<FDBQueriedRecord<Message>>

Returns:

true if this plan returns elements in reverse order

hasRecordScan

public boolean hasRecordScan()

Description copied from interface: QueryPlan

Indicates whether this plan (or one of its components) scans at least a subset of the record range directly rather than going through a secondary index. A plan may only scan over a subset of all records. Someone might not use any secondary indexes explicitly, but they might make use of the primary key index. For example, if they had a compound primary key, and they issued a query for all records that had some value for the first element of their primary key, the planner will produce a plan which scans over a subset of all records.

Specified by:

hasRecordScan in interface QueryPlan<FDBQueriedRecord<Message>>

Returns:

true if this plan (or one of its components) scans at least a subset of the records directly

hasFullRecordScan

public boolean hasFullRecordScan()

Description copied from interface: QueryPlan

Indicates whether this plan (or one of its components) must perform a scan over all records in the store directly rather than going through a secondary index. See hasRecordScan for the comparison between two methods.

Specified by:

hasFullRecordScan in interface QueryPlan<FDBQueriedRecord<Message>>

Returns:

true if this plan (or one of its components) must perform a scan over all records in the store directly

hasIndexScan

public boolean hasIndexScan(@Nonnull
                            String indexName)

Description copied from interface: QueryPlan

Indicates whether this plan scans the given index.

Specified by:

hasIndexScan in interface QueryPlan<FDBQueriedRecord<Message>>

Parameters:

indexName - the name of the index to check for

Returns:

true if this plan (or one of its children) scans the given index

getUsedIndexes

@Nonnull
public Set<String> getUsedIndexes()

Description copied from interface: QueryPlan

Returns a set of names of the indexes used by this plan (and its sub-plans).

Specified by:

getUsedIndexes in interface QueryPlan<FDBQueriedRecord<Message>>

Returns:

an set of indexes used by this plan

maxCardinality

public int maxCardinality(@Nonnull
                          RecordMetaData metaData)

Description copied from interface: QueryPlan

Indicates how many records this plan could possibly return.

Specified by:

maxCardinality in interface QueryPlan<FDBQueriedRecord<Message>>

Parameters:

metaData - meta-data to use to determine things like index uniqueness

Returns:

the maximum number of records or QueryPlan.UNKNOWN_MAX_CARDINALITY if not known

isStrictlySorted

public boolean isStrictlySorted()

Description copied from interface: QueryPlan

Indicates whether the ordering of records from this plan is fully accounted for by its requested sort. That is, each new record will have a different value for that sort key.

Specified by:

isStrictlySorted in interface QueryPlan<FDBQueriedRecord<Message>>

Returns:

true if this plan is fully sorted according to the query definition

strictlySorted

public RecordQueryCoveringIndexPlan strictlySorted()

Description copied from interface: QueryPlan

Return a copy of this plan that has the QueryPlan.isStrictlySorted() property.

Specified by:

strictlySorted in interface QueryPlan<FDBQueriedRecord<Message>>

Returns:

a copy of this plan

getAvailableFields

@Nonnull
public AvailableFields getAvailableFields()

Specified by:

getAvailableFields in interface RecordQueryPlan

hasLoadBykeys

public boolean hasLoadBykeys()

Description copied from interface: QueryPlan

Indicates whether this plan (or one of its components) loads records by their primary key.

Specified by:

hasLoadBykeys in interface QueryPlan<FDBQueriedRecord<Message>>

Returns:

true if this plan (or one of its components) loads records by their primary key

toString

@Nonnull
public String toString()

Overrides:

toString in class Object

getCorrelatedTo

@Nonnull
public Set<CorrelationIdentifier> getCorrelatedTo()

Description copied from interface: Correlated

Returns the set of CorrelationIdentifiers this entity is correlated to. This means that without a process that binds these correlations to values, the RelationalExpression (or others that implement this interface) cannot even in principle produce a meaningful result. As often times entities that implement this interface use composition and effectively describe trees or DAGs, this method should be the set union of the CorrelationIdentifiers this object is correlated to as well as all children, constituent parts, etc.

Specified by:

getCorrelatedTo in interface Correlated<RelationalExpression>

Returns:

the set of CorrelationIdentifiers this entity is correlated to

rebase

@Nonnull
public RecordQueryCoveringIndexPlan rebase(@Nonnull
                                           AliasMap translationMap)

Description copied from interface: Correlated

Rebases this and all other objects this objects is composed of using a given translation map.

Specified by:

rebase in interface Correlated<RelationalExpression>

Parameters:

translationMap - a map defining a translation from CorrelationIdentifiers ids to CorrelationIdentifiers ids'. After the rebase, every correlation to an id contained ids that is contained or referred to directly or indirectly by this must have been transformed to use the mapped counterpart of id id' in ids'. IDs not contained in the translation map must remain unmodified by the rebase operation.

Returns:

a new entity that has been rebased

equalsWithoutChildren

@Nonnull
public boolean equalsWithoutChildren(@Nonnull
                                     RelationalExpression otherExpression,
                                     @Nonnull
                                     AliasMap equivalencesMap)

Specified by:

equalsWithoutChildren in interface RelationalExpression

equals

public boolean equals(Object other)

Overrides:

equals in class Object

hashCode

public int hashCode()

Overrides:

hashCode in class Object

hashCodeWithoutChildren

public int hashCodeWithoutChildren()

Specified by:

hashCodeWithoutChildren in interface RelationalExpression

logPlanStructure

public void logPlanStructure(StoreTimer timer)

Description copied from interface: QueryPlan

Adds one to an appropriate StoreTimer counter for each plan and subplan of this plan, allowing tracking of which plans are being chosen (e.g. index scan vs. full scan).

Specified by:

logPlanStructure in interface QueryPlan<FDBQueriedRecord<Message>>

Parameters:

timer - the counters to increment

getComplexity

public int getComplexity()

Description copied from interface: QueryPlan

Returns an integer representing the "complexity" of the generated plan. Currently, this should simply be the number of plans in the plan tree with this plan as the root (i.e. the number of descendants of this plan, including itself).

Specified by:

getComplexity in interface QueryPlan<FDBQueriedRecord<Message>>

Returns:

the complexity of this plan

planHash

public int planHash(@Nonnull
                    PlanHashable.PlanHashKind hashKind)

Description copied from interface: PlanHashable

Return a hash similar to hashCode, but with the additional guarantee that is is stable across JVMs.

Specified by:

planHash in interface PlanHashable

Parameters:

hashKind - the "kind" of hash to calculate. Each kind of hash has a particular logic with regards to included and excluded items.

Returns:

a stable hash code

getQuantifiers

@Nonnull
public List<? extends Quantifier> getQuantifiers()

Description copied from interface: RelationalExpression

Return an iterator of references to the children of this planner expression. The iterators returned by different calls are guaranteed to be independent (i.e., advancing one will not advance another). However, they might point to the same object, as when Collections.emptyIterator() is returned. The returned iterator should be treated as an immutable object and may throw an exception if Iterator.remove() is called. The iterator must return its elements in a consistent order.

Specified by:

getQuantifiers in interface RecordQueryPlanWithNoChildren

Specified by:

getQuantifiers in interface RelationalExpression

Returns:

an iterator of references to the children of this planner expression

rewritePlannerGraph

@Nonnull
public PlannerGraph rewritePlannerGraph(@Nonnull
                                        List<? extends PlannerGraph> childGraphs)

Description copied from interface: PlannerGraphRewritable

Method to rewrite the planner graph.

Specified by:

rewritePlannerGraph in interface PlannerGraphRewritable

Parameters:

childGraphs - planner graphs of children expression that already have been computed

Returns:

a new planner graph that can combine the childGraphs in a meaningful way. Note that there is no obligation to use the childGraphs at all, this method can create a new independent planner graph completely from scratch.