All Classes and Interfaces

Class
Description
Implement of ExecNodeVisitor.
Strategy to parse statement to Operation by regex.
A CallContext backed by Calcite abstractions.
Helper interface for abstracting accessing literals.
AbstractSqlShowConverter<T extends org.apache.flink.sql.parser.dql.SqlShowCall>
An abstract class for SHOW converters.
A ExecNode which support to convert to adaptive join node.
Implementation class for AdaptiveJoin.
A ExecNodeGraphProcessor which replace the qualified join nodes into adaptive join nodes.
A sub-class of SourceAbilitySpec that can not only serialize/deserialize the aggregation to/from JSON, but also can push the local aggregate into a SupportsAggregatePushDown.
Planner rule that reduces unless grouping columns.
Rule configuration.
Converter to convert SqlAlterTableSchema with source table to generate new Schema.
A SqlOperandCountRange backed by ArgumentCount.
Defines split rules for async calc nodes.
Splits nested call <- asyncCall chains so that nothing is immediately waiting on an async call in a single calc.
Splits async calls if there are multiple across projections, so that there's one per calc.
State object used to keep track of whether a match has been found yet.
An Async implementation of RemoteCalcCallFinder which finds uses of AsyncScalarFunction.
Generates an AsyncFunction which can be used to evaluate calc projections from an async scalar function.
Contains utilities for AsyncScalarFunction.
Options for configuring async behavior.
Converts an expression for a group window function (e.g.
Simple implementation of AuxiliaryConverter.
built-in avg aggregate function.
Built-in Byte Avg aggregate function.
Built-in Decimal Avg aggregate function.
Built-in Double Avg aggregate function.
Built-in Float Avg aggregate function.
Built-in Integer Avg aggregate function.
Built-in Long Avg aggregate function.
Built-in Short Avg aggregate function.
BatchExecNode for adaptive join.
Batch ExecNode to connect a given bounded DataStream and consume data from it.
Batch ExecNode for Calc.
Batch exec node which matches along with join a Java/Scala user defined table function.
Batch ExecNode that collects inputs and builds DynamicFilteringData, and then sends the DynamicFilteringEvent to the source coordinator.
This ExecNode represents a change of partitioning of the input elements for batch.
Batch ExecNode for ExecutionOrderEnforcer.
Batch ExecNode that can expand one row to multiple rows based on given projects.
Batch ExecNode for global runtime filter builder.
Batch ExecNode for hash-based aggregate operator.
BatchExecNode for Hash Join.
Batch ExecNode for hash-based window aggregate operator.
Batch ExecNode for multiple operator fusion codegen input, it is adapter source node.
Batch ExecNode to write data into an external sink defined by a TableSink.
Batch ExecNode to read data from an external source defined by a bounded StreamTableSource.
Batch ExecNode for Limit.
Batch ExecNode for local runtime filter builder.
BatchExecNode for temporal table join that implemented by lookup.
Batch ExecNode which matches along with MATCH_RECOGNIZE.
Batch ExecNode for multiple input which contains a sub-graph of ExecNodes.
BatchExecNode for Nested-loop Join.
Base class for batch ExecNode.
Batch ExecNode for sort-based over window aggregate.
Batch ExecNode base class for sort-based over window aggregate.
OverWindowMode describes the mode of a group in over window.
Batch ExecNode for Python ScalarFunctions.
Batch exec node which matches along with join a Python user defined table function.
Batch ExecNode for Python unbounded group aggregate.
Batch ExecNode for group widow aggregate (Python user defined aggregate function).
Batch ExecNode for sort-based over window aggregate (Python user defined aggregate function).
BatchExecNode for Rank.
Batch ExecNode for runtime filter.
Batch ExecNode for ScripTransform.
Batch ExecNode to write data into an external sink defined by a DynamicTableSink.
BatchExecNode for Sort without limit.
Batch ExecNode for (global) sort-based aggregate operator.
BatchExecNode for Sort with limit.
BatchExecNode for Sort Merge Join.
Batch ExecNode for sort-based window aggregate operator.
Batch ExecNode to read data from an external source defined by a bounded ScanTableSource.
Batch ExecNode that is not a physical node and just union the inputs' records into one node.
Batch ExecNode that read records from given values.
Batch ExecNode for window table-valued function.
Converts FlinkLogicalTableFunctionScan with constant RexCall to
Batch physical RelNode responsible for aggregating all received filters into a global filter.
Batch physical RelNode responsible for building a local runtime filter based on its local data.
Batch physical RelNode which matches along with MATCH_RECOGNIZE.
The physical rule is responsible for convert FlinkLogicalMatch to BatchPhysicalMatch.
The physical rule which is responsible for converting FlinkLogicalAggregate to BatchPhysicalPythonGroupAggregate.
The physical rule is responsible for convert FlinkLogicalCorrelate to BatchPhysicalPythonCorrelate.
The physical rule is responsible for convert FlinkLogicalWindowAggregate to BatchPhysicalPythonGroupWindowAggregate.
Batch physical RelNode responsible for filtering probe side data, see FlinkRuntimeFilterProgram for more info.
Bridges FunctionDefinition to Calcite's representation of an aggregating function (either a system or user-defined function).
Bridges FunctionDefinition to Calcite's representation of a scalar or table function (either a system or user-defined function).
Special flavor of BridgingSqlFunction to indicate a table function to Calcite.
Bridges Procedure to Calcite's representation of a function.
No properties.
SQL version of BuiltInFunctionDefinition in cases where BridgingSqlFunction does not apply.
Builder for fluent definition of built-in functions.
SQL version of BuiltInFunctionDefinition in cases where BridgingSqlFunction does not apply.
Thin wrapper around SqlParser that does exception conversion and SqlNode casting.
This class is used to create a CalciteSchema with a given Schema as the root.
Rule will transpose the conditions after the Python correlate node if the join type is inner join.
Planner rule that transposes FlinkLogicalCalc past FlinkLogicalRank to reduce rank input fields.
Rule configuration.
Transpose FlinkLogicalCalc past into FlinkLogicalSnapshot.
Rule configuration.
A CallContext backed by SqlCallBinding.
Rule to convert CallExpression.
Planner expression resolver for UnresolvedCallExpression.
A shuttle to capitalize all query hints on corresponding nodes.
Generated cast code block result.
A CastRule provides the logic to create a CastExecutor starting from the input and the target types.
Casting context.
In order to apply a CastRule, the runtime checks if a particular rule matches the tuple of input and target type using this class.
Builder for the CastRulePredicate.
This class resolves CastRule using the input and the target type.
A mapping between Flink's catalog and Calcite's schema.
Bridge between the CatalogManager and the Schema.
Represents a wrapper for CatalogBaseTable in Schema.
A FlinkPreparingTableBase implementation which defines the interfaces required to translate the Calcite RelOptTable to the Flink specific TableSourceTable.
Utility class for converting CatalogTableStatistics and CatalogColumnStatistics to TableStats.
Strategy to parse statement to ClearOperation.
A shuttle to remove query block alias hint.
Clear the invalid query hints in the unmatched nodes.
Due to Calcite will expand the whole SQL RelNode tree that contains query block, query hints (including join hints and state ttl hints) will be propagated from root to leaves in the whole RelNode tree.
Cast rule that has code generation capabilities.
Context for code generation.
Table sink for TableResult.collect().
A finder used to look up referenced column name in a ResolvedExpression.
Base class for exec Async Calc.
Base class for exec Calc.
Base ExecNode which matches along with join a Java/Scala user defined table function.
Base class for exec Exchange.
Base ExecNode that can expand one row to multiple rows based on given projects.
Base ExecNode to write data into an external sink defined by a TableSink.
Base ExecNode to read data from an external source defined by a StreamTableSource.
Base ExecNode for temporal table join which shares most methods.
Common ExecNode which matches along with MATCH_RECOGNIZE.
Base class for exec Python Calc.
Base ExecNode which matches along with join a Python user defined table function.
Base ExecNode to write data to an external sink defined by a DynamicTableSink.
Base ExecNode to read data from an external source defined by a ScanTableSource.
Base ExecNode that is not a physical node and just union the inputs' records into one node.
Base ExecNode that read records from given values.
Base ExecNode for window table-valued function.
Base physical RelNode which matches along with MATCH_RECOGNIZE.
The physical rule is responsible for converting FlinkLogicalMatch to physical Match rel.
A utility class used in PyFlink.
Default implementation of CompositeOperandTypeChecker, the class was copied over because of current Calcite issue CALCITE-5380.
Custom filtering for Configuration used by ExecNodeBase.getPersistedConfig() to avoid serializing null or empty configurations.
Utility for deciding whether than expression supports constant folding or not.
Planner rule that removes the output column of rank number iff there is an equality condition for the rank column.
Rule configuration.
Planner rule that rewrites sort correlation to a Rank.
Rule configuration.
This count1 aggregate function returns the count1 of values which go into it like [[CountAggFunction]].
built-in count aggregate function.
built-in CUME_DIST agg function.
Logical representation of a cumulative window specification.
Customized converter used by CustomizedConvertRule.
Registry of customized converters used by CustomizedConvertRule.
Customized CallExpressionConvertRule, Functions conversion here all require special logic, and there may be some special rules, such as needing get the literal values of inputs, such as converting to combinations of functions, to convert to RexNode of calcite.
Utilities to deal with DataViews.
Specification for a special MapView for deduplication.
A ExecNodeGraphProcessor that finds out all deadlocks in the ExecNodeGraph and resolves them.
Abstract class to resolve the expressions in DeclarativeAggregateFunction.
A context for default dialect with calcite related interfaces.
Default implementation of Executor.
Factory to create the default implementation of an Executor for TableEnvironment.
A Parser factory that creates ParserImpl.
Factory for the default Planner.
A utility class for delete push down.
built-in dense_rank aggregate function.
A CallExpressionConvertRule that performs a simple one-to-one mapping between FunctionDefinition and a corresponding SqlOperator.
This is dummy StreamExecutionEnvironment, which holds a real StreamExecutionEnvironment, shares all configurations of the real environment, and disables all configuration setting methods.
This processor future checks each dynamic filter source to see if it is chained with a multiple input operator.
Planner utils for Dynamic partition Pruning.
Utilities for dealing with DynamicTableSink.
Utilities for dealing with DynamicTableSource.
DynamicTableSourceSpec describes how to serialize/deserialize dynamic table sink table and create DynamicTableSink from the deserialization result.
DynamicTableSourceSpec describes how to serialize/deserialize dynamic table source table and create DynamicTableSource from the deserialization result.
Rule that converts an EnumerableTableScan into a LogicalTableScan.
Rule configuration.
Traverses an event time temporal table join RelNode tree and update the right child to set FlinkLogicalTableSourceScan's eventTimeSnapshot property to true which will prevent it generating a new StreamPhysicalChangelogNormalize later.
The representation of an edge connecting two ExecNodes.
Builder of the ExecEdge.
Records are shuffled by hash when passing through this edge.
The ExecEdge.Shuffle defines how to exchange the records between ExecNodes.
Enumeration which describes the shuffle type for records when passing this edge.
The representation of execution information for a FlinkPhysicalRel.
Base class for ExecNode.
Configuration view which is used combine the PlannerBase#getTableConfig() with the ExecNodeBase.getPersistedConfig() configuration.
Helper class that holds the necessary identifier fields that are used for JSON plan serialization and deserialization.
The ExecNodeGraph representing the ExecNode topology.
A generator that generates a ExecNode graph from a graph of FlinkPhysicalRels.
Implementation of CompiledPlan backed by an ExecNodeGraph.
ExecNodeGraphProcessor plugin, use it can change or update the given ExecNodeGraph.
This class validates the ExecNodeGraph.
Annotation to be used for ExecNodes to keep necessary metadata when serializing/deserializing them in a plan.
Utility class for ExecNodeMetadata related functionality.
Helper Pojo used as a tuple for the ExecNodeMetadataUtil.LOOKUP_MAP.
An utility class for converting an exec node plan to a string as a tree style.
An ExecNodeTranslator is responsible for translating an ExecNode to Transformations.
An Utility class that helps translating ExecNode to Transformation.
Visitor pattern for traversing a dag of ExecNode objects.
Implement of ExecNodeVisitor.
Utility that expand SQL identifiers in a SQL query.
Result of expanding.
A common subclass for all tables that are expanded to sub-query (views).
Builder for Expressions.
Cast rule that is able to generate a single expression containing all the casting logic.
Visit expression to generator RexNode.
ExtendedParser is used for parsing some special command which can't supported by CalciteParser, e.g.
Strategy to parse statement to Operation. parsing some special command which can't supported by CalciteParser, e.g.
Compiler-checked resources similar to CalciteResource in Calcite project.
A sub-class of SourceAbilitySpec that can not only serialize/deserialize the filter to/from JSON, but also can push the filter into a SupportsFilterPushDown.
Planner rule that expands distinct aggregates (such as COUNT(DISTINCT x)) from a Aggregate.
Planner rule that pushes an Aggregate past a Join.
Extends calcite's AggregateProjectMergeRule, modification: consider the timeAttribute field when analyzing interesting fields for the LogicalWindowAggregate node while normal Aggregate node needn't.
Planner rule that removes a Aggregate if its aggregate functions are SUM, MIN, MAX, AUXILIARY_GROUP with no filterArgs, and the underlying relational expression is already distinct.
Flink bushy join reorder rule, which will convert MultiJoin to a bushy join tree.
Rule configuration.
Flink specific CalciteCatalogReader that changes the RelOptTable which wrapped a CatalogSchemaTable to a FlinkPreparingTableBase.
Extends FlinkCalciteCatalogReader to allow to read a snapshot of the CalciteSchema.
Extends Calcite's SqlValidator by Flink-specific behavior.
Planner rule that merges a Calc onto a Calc.
Rule configuration.
Custom Flink SqlRexConvertletTable to add custom SqlNode to RexNode conversions.
The Flink CURRENT_DATE function differs from the parent SqlCurrentDateFunction which is aware of whether it is used in batch mode, if true it will act totally same as the parent SqlCurrentDateFunction, but will be a non-deterministic function if not in batch mode.
The function CURRENT_ROW_TIMESTAMP() in Flink which supports configuring the return type and the precision of return type.
Planner program that tries to do partition prune in the execution phase, which can translate a BatchPhysicalTableSourceScan to a BatchPhysicalDynamicFilteringTableSourceScan whose source is a partition source.
Extends calcite's FilterCalcMergeRule, modification: only merge the two neighbouring Filter and Calc if each non-deterministic RexNode of bottom Calc should appear at most once in the implicit project list and condition of top Filter.
Planner rule that pushes filters above and within a join node into the join node and/or its children nodes.
Rule configuration.
Rule that tries to push filter expressions into a join condition and into the inputs of the join.
Rule configuration.
Rule that pushes parts of the join condition to its inputs.
Rule configuration.
Predicate that returns whether a filter is valid in the ON clause of a join for this particular kind of join.
Extends calcite's FilterProjectTransposeRule for streaming scenario, modification: does not transpose the top filter references field which generated by non-deterministic function.
Utility class for Flink hints.
A collection of Flink style HintStrategys.
Flink join reorder rule, which can change the join order of input relNode tree.
Rule configuration.
Flink Planner rule to flatten a tree of LogicalJoins into a single MultiJoin with N inputs.
Rule configuration.
Rule to remove trivial FlinkLogicalCalc.
Contains the interfaces for several specified metadata of flink.
Metadata about the interval of given column from a specified relational expression.
Handler API.
Metadata about the null count of given column from a specified relational expression.
Handler API.
Origin null count, looking until source.
Handler API.
Metadata about the interval of given column under the given filter argument from a specified relational expression.
Handler API.
Metadata about how a relational expression is distributed.
Handler API.
Metadata about the modified property of a RelNode.
Handler API.
Metadata about the (minimum) unique groups of the given columns from a specified relational expression.
Handler API.
Metadata about which combinations of columns are upsert identifiers.
Handler API.
Metadata about the window properties of a RelNode.
Handler API.
A Flink Prepare.AbstractPreparingTable implementation for the purposes of the sql-to-rel conversion and planner.
Extends calcite's ProjectMergeRule, modification: only merge the two neighbouring Project and Calc if each non-deterministic RexNode of bottom Calc should appear at most once in the project list of top Project.
Planner rule that pushes a Project past a Join by splitting the projection into a projection on top of each child of the join.
Extends calcite's ProjectMergeRule, modification: only merge the two neighbouring Projects if each non-deterministic RexNode of bottom Project should appear at most once in the project list of top Project.
A FlinkOptimizeProgram that recompute statistics after partition pruning and filter push down.
Flink-specific RelBuilder.
FlinkRelMdCollation supplies a default implementation of RelMetadataQuery.collations(org.apache.calcite.rel.RelNode) for the standard logical algebra.
A RelMetadataQuery that defines extended metadata handler in Flink, e.g ColumnInterval, ColumnNullCount.
Utilities for RelNode.
Type inference in Flink.
A slim extension over a RexBuilder.
Planner program that tries to inject runtime filter for suitable join to improve join performance.
Base class for flink Schema, which provides some default implementations.
Planner rule that pushes semi-join down in a tree past a Filter.
Planner rule that pushes a org.apache.calcite.rel.core.SemiJoin down in a tree past a Join in order to trigger other rules that will convert SemiJoins.
Planner rule that pushes a org.apache.calcite.rel.core.SemiJoin down in a tree past a Project.
Binding supports to rewrite the DEFAULT operator.
A workaround for adjusting types of ROW nested fields.
Operator table that contains only Flink-specific functions and operators.
A util method to create SqlParserImplFactory according to SqlConformance.
A collection of PlanAnalyzers for org.apache.flink.table.planner.delegation.StreamPlanner.
Function that used to define SQL time functions like LOCALTIME, CURRENT_TIME(these are all dynamic functions in Calcite's SqlStdOperatorTable) in Flink, the difference from the parent SqlAbstractTimeFunction is this function class be aware of whether it is used in batch mode, if true it will act totally same as the parent SqlAbstractTimeFunction, but will be a non-deterministic function if not in batch mode.
Function that used to define SQL time function like LOCALTIMESTAMP, CURRENT_TIMESTAMP, NOW() in Flink, the function support configuring the return type and the * precision of return type.
Custom type system for Flink.
A ExecNodeGraphProcessor which finds all ExecNodes that require hash distribution, but its input has no hash Exchange node.
Thin adapter between SqlOperatorTable and FunctionCatalog.
A call expression converter rule that converts calls to user defined functions.
A ExecNode which support operator fusion codegen.
An implementation of PlanAnalyzer to analyze whether group aggregation can be optimized to local-global two_phase aggregation by tuning table configurations.
Utility for generating hash join operator factory, including HashJoinOperator and the codegen version of LongHashJoinOperator generated by LongHashJoinGenerator.
Strategy to parse statement to HelpOperation.
Logical representation of a hopping window specification.
A namespace for the statement with FOR SYSTEM_TIME AS OF TIMESTAMP clause.
ImmutableAggregateReduceGroupingRule contains immutable implementation classes generated from abstract value types defined as nested inside AggregateReduceGroupingRule.
Builds instances of type AggregateReduceGroupingRuleConfig.
ImmutableBatchPhysicalConstantTableFunctionScanRule contains immutable implementation classes generated from abstract value types defined as nested inside BatchPhysicalConstantTableFunctionScanRule.
ImmutableCalcRankTransposeRule contains immutable implementation classes generated from abstract value types defined as nested inside CalcRankTransposeRule.
Builds instances of type CalcRankTransposeRuleConfig.
ImmutableCalcSnapshotTransposeRule contains immutable implementation classes generated from abstract value types defined as nested inside CalcSnapshotTransposeRule.
Builds instances of type CalcSnapshotTransposeRuleConfig.
ImmutableConstantRankNumberColumnRemoveRule contains immutable implementation classes generated from abstract value types defined as nested inside ConstantRankNumberColumnRemoveRule.
ImmutableCorrelateSortToRankRule contains immutable implementation classes generated from abstract value types defined as nested inside CorrelateSortToRankRule.
Builds instances of type CorrelateSortToRankRuleConfig.
ImmutableEnumerableToLogicalTableScan contains immutable implementation classes generated from abstract value types defined as nested inside EnumerableToLogicalTableScan.
Builds instances of type EnumerableToLogicalTableScanConfig.
ImmutableEventTimeTemporalJoinRewriteRule contains immutable implementation classes generated from abstract value types defined as nested inside EventTimeTemporalJoinRewriteRule.
Immutable implementation of EventTimeTemporalJoinRewriteRule.Config.
Builds instances of type Config.
ImmutableFlinkBushyJoinReorderRule contains immutable implementation classes generated from abstract value types defined as nested inside FlinkBushyJoinReorderRule.
Immutable implementation of FlinkBushyJoinReorderRule.Config.
Builds instances of type Config.
ImmutableFlinkCalcMergeRule contains immutable implementation classes generated from abstract value types defined as nested inside FlinkCalcMergeRule.
Immutable implementation of FlinkCalcMergeRule.FlinkCalcMergeRuleConfig.
Builds instances of type FlinkCalcMergeRuleConfig.
Builds instances of type ImmutableFlinkFilterIntoJoinRuleConfig.
Builds instances of type ImmutableFlinkFilterJoinRuleConfig.
ImmutableFlinkJoinReorderRule contains immutable implementation classes generated from abstract value types defined as nested inside FlinkJoinReorderRule.
Immutable implementation of FlinkJoinReorderRule.Config.
Builds instances of type Config.
ImmutableFlinkJoinToMultiJoinRule contains immutable implementation classes generated from abstract value types defined as nested inside FlinkJoinToMultiJoinRule.
Immutable implementation of FlinkJoinToMultiJoinRule.Config.
Builds instances of type Config.
ImmutableJoinConditionTypeCoerceRule contains immutable implementation classes generated from abstract value types defined as nested inside JoinConditionTypeCoerceRule.
Builds instances of type JoinConditionTypeCoerceRuleConfig.
ImmutableJoinDeriveNullFilterRule contains immutable implementation classes generated from abstract value types defined as nested inside JoinDeriveNullFilterRule.
Builds instances of type JoinDeriveNullFilterRuleConfig.
ImmutableJoinTableFunctionScanToCorrelateRule contains immutable implementation classes generated from abstract value types defined as nested inside JoinTableFunctionScanToCorrelateRule.
Immutable implementation of JoinTableFunctionScanToCorrelateRule.Config.
Builds instances of type Config.
ImmutableLogicalUnnestRule contains immutable implementation classes generated from abstract value types defined as nested inside LogicalUnnestRule.
Immutable implementation of LogicalUnnestRule.LogicalUnnestRuleConfig.
Builds instances of type LogicalUnnestRuleConfig.
ImmutableProjectSemiAntiJoinTransposeRule contains immutable implementation classes generated from abstract value types defined as nested inside ProjectSemiAntiJoinTransposeRule.
Builds instances of type ProjectSemiAntiJoinTransposeRuleConfig.
ImmutableProjectSnapshotTransposeRule contains immutable implementation classes generated from abstract value types defined as nested inside ProjectSnapshotTransposeRule.
Immutable implementation of ProjectSnapshotTransposeRule.Config.
Builds instances of type Config.
ImmutablePushCalcPastChangelogNormalizeRule contains immutable implementation classes generated from abstract value types defined as nested inside PushCalcPastChangelogNormalizeRule.
Immutable implementation of PushCalcPastChangelogNormalizeRule.Config.
Builds instances of type Config.
ImmutablePushFilterIntoLegacyTableSourceScanRule contains immutable implementation classes generated from abstract value types defined as nested inside PushFilterIntoLegacyTableSourceScanRule.
ImmutablePushLimitIntoLegacyTableSourceScanRule contains immutable implementation classes generated from abstract value types defined as nested inside PushLimitIntoLegacyTableSourceScanRule.
ImmutablePushProjectIntoTableSourceScanRule contains immutable implementation classes generated from abstract value types defined as nested inside PushProjectIntoTableSourceScanRule.
Immutable implementation of PushProjectIntoTableSourceScanRule.Config.
Builds instances of type Config.
ImmutableRemoveRedundantLocalRankRule contains immutable implementation classes generated from abstract value types defined as nested inside RemoveRedundantLocalRankRule.
Builds instances of type RemoveRedundantLocalRankRuleConfig.
ImmutableRemoveUnreachableCoalesceArgumentsRule contains immutable implementation classes generated from abstract value types defined as nested inside RemoveUnreachableCoalesceArgumentsRule.
Builds instances of type Config.
ImmutableReplaceIntersectWithSemiJoinRule contains immutable implementation classes generated from abstract value types defined as nested inside ReplaceIntersectWithSemiJoinRule.
Builds instances of type ReplaceIntersectWithSemiJoinRuleConfig.
ImmutableReplaceMinusWithAntiJoinRule contains immutable implementation classes generated from abstract value types defined as nested inside ReplaceMinusWithAntiJoinRule.
Builds instances of type ReplaceMinusWithAntiJoinRuleConfig.
ImmutableRewriteIntersectAllRule contains immutable implementation classes generated from abstract value types defined as nested inside RewriteIntersectAllRule.
Builds instances of type RewriteIntersectAllRuleConfig.
ImmutableRewriteMinusAllRule contains immutable implementation classes generated from abstract value types defined as nested inside RewriteMinusAllRule.
Builds instances of type RewriteMinusAllRuleConfig.
ImmutableRewriteMultiJoinConditionRule contains immutable implementation classes generated from abstract value types defined as nested inside RewriteMultiJoinConditionRule.
Builds instances of type RewriteMultiJoinConditionRuleConfig.
ImmutableSimplifyJoinConditionRule contains immutable implementation classes generated from abstract value types defined as nested inside SimplifyJoinConditionRule.
Builds instances of type SimplifyJoinConditionRuleConfig.
ImmutableSplitPythonConditionFromCorrelateRule contains immutable implementation classes generated from abstract value types defined as nested inside SplitPythonConditionFromCorrelateRule.
ImmutableStreamPhysicalConstantTableFunctionScanRule contains immutable implementation classes generated from abstract value types defined as nested inside StreamPhysicalConstantTableFunctionScanRule.
ImmutableStreamPhysicalWindowJoinRule contains immutable implementation classes generated from abstract value types defined as nested inside StreamPhysicalWindowJoinRule.
Builds instances of type StreamPhysicalWindowJoinRuleConfig.
ImmutableTemporalJoinRewriteWithUniqueKeyRule contains immutable implementation classes generated from abstract value types defined as nested inside TemporalJoinRewriteWithUniqueKeyRule.
ImmutableTwoStageOptimizedAggregateRule contains immutable implementation classes generated from abstract value types defined as nested inside TwoStageOptimizedAggregateRule.
Builds instances of type TwoStageOptimizedAggregateRuleConfig.
ImmutableUncollectToTableFunctionScanRule contains immutable implementation classes generated from abstract value types defined as nested inside UncollectToTableFunctionScanRule.
Builds instances of type UncollectToTableFunctionScanRuleConfig.
ImmutableWatermarkAssignerChangelogNormalizeTransposeRule contains immutable implementation classes generated from abstract value types defined as nested inside WatermarkAssignerChangelogNormalizeTransposeRule.
Builds instances of type Config.
ImmutableWindowGroupReorderRule contains immutable implementation classes generated from abstract value types defined as nested inside WindowGroupReorderRule.
Builds instances of type WindowGroupReorderRuleConfig.
Builds instances of type ImmutableWindowPropertiesRuleConfig.
ImmutableWrapJsonAggFunctionArgumentsRule contains immutable implementation classes generated from abstract value types defined as nested inside WrapJsonAggFunctionArgumentsRule.
Immutable implementation of WrapJsonAggFunctionArgumentsRule.Config.
Builds instances of type Config.
This class contains algorithm to detect and resolve input priority conflict in an ExecNode graph.
The InputProperty describes the input information of the ExecNode.
Builder of the InputProperty.
Enumeration which describes how an input record may trigger the output behavior of the target ExecNode.
Enumeration which describes the type of the input data distribution.
The input will read the records whose keys hash to a particular hash value.
A special distribution which indicators the data distribution is the same as its input. '
The required input data distribution for records when they are read in.
A BatchExecNode which does not sort the input data within the operator, but requires the input data is already sorted.
This class holds internal configuration constants used by Flink's table module.
Deprecated.
Dictionary of function definitions for all internal used functions.
IntervalJoinSpec describes how two tables will be joined in interval join.
WindowBounds describes the time range condition of a Interval Join.
Planner rule that coerces the both sides of EQUALS(`=`) operator in Join condition to the same type while sans nullability.
Rule configuration.
Planner rule that filters null values before join if the count null value from join input is greater than null filter threshold.
Rule configuration.
JoinSpec describes how two tables will be joined.
Currently available join strategies and corresponding join hint names.
Rule that rewrites Join on TableFunctionScan to Correlate.
Utility for FlinkJoinType.
A utility class that provide abilities for JSON serialization and deserialization.
Utility for KeySelector.
LEAD and LAG aggregate functions return the value of given expression evaluated at given offset.
BooleanLeadLagAggFunction.
ByteLeadLagAggFunction.
CharLeadLagAggFunction.
DateLeadLagAggFunction.
DecimalLeadLagAggFunction.
DoubleLeadLagAggFunction.
FloatLeadLagAggFunction.
IntLeadLagAggFunction.
LongLeadLagAggFunction.
ShortLeadLagAggFunction.
StringLeadLagAggFunction.
TimeLeadLagAggFunction.
TimestampLeadLagAggFunction.
CallExpressionConvertRule to convert ScalarFunctionDefinition.
A sub-class of SourceAbilitySpec that can not only serialize/deserialize the limit value to/from JSON, but also can push the limit value into a LimitPushDownSpec.
built-in listagg aggregate function.
LogicalDistribution is used to represent the expected distribution of the data, similar to Hive's SORT BY, DISTRIBUTE BY, and CLUSTER BY semantics.
Symmetric converter between LogicalType and RelDataType.
LogicalScriptTransform is used to represent the sql semantic "TRANSFORM c1, c2, xx USING 'script'".
Sub-class of Snapshot not targeted at any particular engine or calling convention.
A LogicalTableScan reads all the rows from a RelOptTable.
Planner rule that rewrites UNNEST to explode function.
Rule configuration.
The CallContext of a LookupTableSource runtime function.
Utilities for lookup joins using LookupTableSource.
AsyncLookupOptions includes async related options.
A LookupJoinUtil.LookupKey whose value is constant.
A LookupJoinUtil.LookupKey whose value comes from the left table field.
A field used as an equal condition when querying content from a dimension table.
RetryOptions includes retry lookup related options.
ShuffleLookupOptions includes shuffle related options.
The function used to access a rowtime attribute with TIMESTAMP or TIMESTAMP_LTZ type from MATCH_RECOGNIZE clause.
MatchSpec describes the MATCH_RECOGNIZE info, see Match.
The utils for materialized table.
built-in max aggregate function.
Built-in Boolean Max aggregate function.
Built-in Byte Max aggregate function.
Built-in Date Max aggregate function.
Built-in Decimal Max aggregate function.
Built-in Double Max aggregate function.
Built-in Float Max aggregate function.
Built-in Int Max aggregate function.
Built-in Long Max aggregate function.
Built-in Short Max aggregate function.
Built-in String Max aggregate function.
Built-in Time Max aggregate function.
Built-in TimestampLtz Max aggregate function.
Built-in Timestamp Max aggregate function.
A utility class with logic for handling the CREATE TABLE ... AS SELECT clause.
built-in min aggregate function.
Built-in Boolean Min aggregate function.
Built-in Byte Min aggregate function.
Built-in Date Min aggregate function.
Built-in Decimal Min aggregate function.
Built-in Double Min aggregate function.
Built-in Float Min aggregate function.
Built-in Int Min aggregate function.
Built-in Long Min aggregate function.
Built-in Short Min aggregate function.
Built-in String Min aggregate function.
Built-in Time Min aggregate function.
Built-in TimestampLtz Min aggregate function.
Built-in Timestamp Min aggregate function.
The pojo class that describes the mini-batch interval and mini-batch mode.
The type of mini-batch interval: rowtime or proctime.
Utility class for mini-batch related config.
Lists all kinds of modify operations that happen in a changelog.
The set of modify operations contained in a changelog.
Builder for configuring and creating instances of ModifyKindSet.
Helper annotation to enable multiple ExecNodeMetadata annotations on an ExecNode class.
A ExecNodeGraphProcessor which organize ExecNodes into multiple input nodes.
An ExecNodeTranslator that will generate ZERO, ONE or more Transformations.
An implementation of PlanAnalyzer to analyze the potential risk of non-deterministic update when OptimizerConfigOptions.TABLE_OPTIMIZER_NONDETERMINISTIC_UPDATE_STRATEGY is ignored.
built-in NTILE agg function.
Parameter type-checking strategy where all operand types except first one must be numeric type.
Utils methods for converting sql to operations.
A CallContext backed by SqlOperatorBinding.
Some dedicated operator type which is used in ExecutionConfigOptions.TABLE_EXEC_DISABLED_OPERATORS.
An interface for those physical operators that support operator fusion codegen.
OpFusionCodegenSpecGenerator is used to operator fusion codegen that generate the fusion code, it has multiple inputs and outputs, then form a DAG.
A OpFusionContext contains information about the context in which OpFusionCodegenSpec needed to do operator fusion codegen.
A CallExpressionConvertRule that converts BuiltInFunctionDefinitions.OVER.
OverSpec describes a set of over aggregates.
OverSpec.GroupSpec describes group of over aggregate calls that have the same window specification.
A sub-class of SinkAbilitySpec that can not only serialize/deserialize the overwrite flag to/from JSON, but also can overwrite existing data for SupportsOverwrite.
Implementation of Parser that uses Calcite.
Enumerations for partial final aggregate types.
A sub-class of SinkAbilitySpec that can not only serialize/deserialize the partition to/from JSON, but also can write partitioned data for SupportsPartitioning.
A sub-class of SourceAbilitySpec that can not only serialize/deserialize the partitions to/from JSON, but also can push the partitions into a SupportsPartitionPushDown.
PartitionSpec describes how data is partitioned in Rank.
built-in percent_rank aggregate function.
Plain POJO for advice provided by PlanAnalyzer.
Categorize the semantics of a PlanAdvice.
Categorize the scope of a PlanAdvice.
Plan analyzer analyzes the optimized physical plan and provide potential risk warnings and optimization advice.
The analyzed PlanAdvice with a list of applicable RelNode ids.
Wrapper for valid call procedure operation generated by Planner.
Utility class to create RelBuilder or FrameworkConfig used to create a corresponding Planner.
Wrapper for valid logical plans generated by Planner.
Namespace whose contents are defined by the result of a call to a user-defined procedure.
Context for ExecNodeGraphProcessor to process ExecNodeGraph.
A sub-class of SourceAbilitySpec that can not only serialize/deserialize the projection to/from JSON, but also can push the projection into a SupportsProjectionPushDown.
Planner rule that pushes a Project down in a tree past a semi/anti Join by splitting the projection into a projection on top of left child of the Join.
Rule configuration.
Transpose LogicalProject past into LogicalSnapshot.
Configuration for ProjectSnapshotTransposeRule.
Transpose between the LogicalWatermarkAssigner and LogicalProject.
Planner rule that pushes a LogicalProject into a LogicalTableFunctionScan which contains a Window table function call by splitting the projection into a projection on top of child of the TableFunctionScan.
Pushes primary key filters and used fields project through a ChangelogNormalize operator to reduce its state size.
Pushes a filter condition from the FlinkLogicalCalc and into a FlinkLogicalTableSourceScan.
Planner rule that tries to push a filter into a FilterableTableSource.
Rule configuration.
Base class for rules that push down filters into table scan.
Planner rule that tries to push a filter into a LogicalTableScan, which table is a TableSourceTable.
Planner rule that tries to push limit into a LimitableTableSource.
Planner rule that tries to push limit into a FlinkLogicalTableSourceScan, which table is a TableSourceTable.
Planner rule that tries to push a local aggregator into an BatchPhysicalTableSourceScan whose table is a TableSourceTable with a source supporting SupportsAggregatePushDown.
Planner rule that tries to push a local hash aggregate which without sort into a BatchPhysicalTableSourceScan whose table is a TableSourceTable with a source supporting SupportsAggregatePushDown.
Planner rule that tries to push a local hash aggregate which with calc into a BatchPhysicalTableSourceScan whose table is a TableSourceTable with a source supporting SupportsAggregatePushDown.
Planner rule that tries to push a local sort aggregate which without sort into a BatchPhysicalTableSourceScan whose table is a TableSourceTable with a source supporting SupportsAggregatePushDown.
Planner rule that tries to push a local sort aggregate which without sort into a BatchPhysicalTableSourceScan whose table is a TableSourceTable with a source supporting SupportsAggregatePushDown.
Planner rule that tries to push a local sort aggregate which with sort and calc into a BatchPhysicalTableSourceScan whose table is a TableSourceTable with a source supporting SupportsAggregatePushDown.
Planner rule that tries to push a local sort aggregate which with sort into a BatchPhysicalTableSourceScan whose table is a TableSourceTable with a source supporting SupportsAggregatePushDown.
Planner rule that tries to push partition evaluated by filter condition into a LogicalTableScan.
Pushes a LogicalProject into a LogicalTableScan.
Rule to push the FlinkLogicalWatermarkAssigner across the FlinkLogicalCalc to the FlinkLogicalTableSourceScan.
Rule to push the FlinkLogicalWatermarkAssigner into the FlinkLogicalTableSourceScan.
Base rule for interface SupportsWatermarkPushDown and SupportsSourceWatermark.
Rule will split the Python FlinkLogicalTableFunctionScan with Java calls or the Java FlinkLogicalTableFunctionScan with Python calls into a FlinkLogicalCalc which will be the left input of the new FlinkLogicalCorrelate and a new FlinkLogicalTableFunctionScan.
Rule will merge Python FlinkLogicalCalc used in Map operation, Flatten FlinkLogicalCalc and Python FlinkLogicalCalc used in Map operation together.
Rule which renames the field names of the Flatten FlinkLogicalCalc which is right after a FlinkLogicalCalc representing a Python Map operation to the output names of the map function.
Utility for Python.
An abstract shuttle for each shuttle used for query hint.
Resolve and validate the query hints.
A bridge between a Flink's specific QueryOperationCatalogView and a Calcite's RelOptTable.
Converter from Flink's specific relational representation: QueryOperation to Calcite's specific relational representation: RelNode.
Operation to parse statement to QuitOperation.
built-in rank aggregate function.
built-in rank like aggregate function, e.g. rank, dense_rank
Base class of Strategy to choose different rank process function.
A strategy which only works when input only contains insertion changes.
A strategy which works when input contains update or deletion changes.
A placeholder strategy which will be inferred after FlinkChangelogModeInferenceProgram.
A strategy which only works when input shouldn't contains deletion changes and input should have the given RankProcessStrategy.UpdateFastStrategy.primaryKeys and should be monotonic on the order by field.
The RelDataType representation of a RawType.
A sub-class of SourceAbilitySpec that can not only serialize/deserialize the metadata columns to/from JSON, but also can read the metadata columns from SupportsReadingMetadata.
Planner rule that removes the output column of rank number iff the rank number column is not used by successor calc.
Copied from calcite to workaround CALCITE-4668
Information necessary to create a call to an aggregate function.
Configuration of RelBuilder.
Information necessary to create the GROUP BY clause of an Aggregate.
Call to a windowed aggregate function.
RelDecorrelator replaces all correlated expressions (corExp) in a relational expression (RelNode) tree with non-correlated expressions that are produced from joining the RelNode that produces the corExp with the RelNode that references it.
Planner rule that adjusts projects when counts are added.
Rule configuration.
Base configuration for rules that are non-static in a RelDecorrelator.
A map of the locations of Correlate in a tree of RelNodes.
Planner rule that removes correlations for scalar aggregates.
Rule configuration.
Planner rule that removes correlations for scalar projects.
Rule configuration.
Rule to remove an Aggregate with SINGLE_VALUE.
Rule configuration.
Converts an RelNode to string with only the information from the RelNode itself without the information from its inputs.
Utility to infer Predicates that are applicable above a RelNode.
Traverses a RelNode tree and converts fields with TimeIndicatorRelDataType type.
It describes the information of window properties of a RelNode.
Finds whether a RexNode contains remote calls or not.
Planner rule that matches a global BatchPhysicalRank on a local BatchPhysicalRank, and merge them into a global BatchPhysicalRank.
Configuration for RemoveRedundantLocalRankRule.
Removes unreachable BuiltInFunctionDefinitions.COALESCE arguments.
Planner rule that replaces distinct Intersect with a distinct Aggregate on a SEMI Join.
Rule configuration.
Planner rule that replaces distinct Minus (SQL keyword: EXCEPT) with a distinct Aggregate on an ANTI Join.
Rule configuration.
Replace CommonPhysicalTableSourceScan with Calc.
Strategy to parse statement to ResetOperation.
Find reusable sources.
Replaces logical Intersect operator using a combination of union all, aggregate and table function.
Rule configuration.
Replaces logical Minus operator using a combination of union all, aggregate and table function.
Rule configuration.
Planner rule to apply transitive closure on MultiJoin for equi-join predicates.
Rule configuration.
Planner internal factory for parsing/translating to RexNode.
Access to a field of a row-expression.
Utilities to work with RexLiteral.
Wrapper for a ResolvedExpression that originated from a RexNode.
A special RexCall that represents a table argument in a signature of StaticArguments.
Default implementation of RexUtil, the class was copied over because of current Calcite way of inferring constants from IS NOT DISTINCT FROM clashes with filter push down.
Deprecated.
Visitor which builds a bitmap of the inputs used by an expression.
Shuttle that fixes up an expression to match changes in nullability of input fields.
Visitor that tells whether a node matching a particular description exists in a tree.
Visitor that throws Util.FoundOne if applied to an expression that contains a RexSubQuery.
A TableSourceQueryOperation with FlinkStatistic and qualifiedName.
RowData to String converter using CastRule.
A sub-class of SinkAbilitySpec that can not only serialize/deserialize the row-level delete mode & required physical column indices to/from JSON, but also can delete existing data for SupportsRowLevelDelete.
Utils for putting/getting RowLevelModificationScanContext/SupportsRowLevelModificationScan.RowLevelModificationType in current thread, it enables to put/get them in different methods.
A sub-class of SinkAbilitySpec that can not only serialize/deserialize the row-level update mode, columns & required physical column indices to/from JSON, but also can update existing data for SupportsRowLevelUpdate.
built-in row_number aggregate function.
Utils for deriving row types of RelNodes.
Reuse sources.
Utils for ScanReuser.
A utility class for MergeTableAsUtil and MergeTableLikeUtil classes.
A context to allow the store user-defined data within ExecNode serialization and deserialization.
Logical representation of a session window specification.
Strategy to parse statement to SetOperation.
Util class that rewrite SetOp.
Utilities for quick access of commonly used instances (like FlinkTypeFactory) without long chains of getters or casting like (FlinkTypeFactory) agg.getCluster.getTypeFactory().
Planner rule that apply various simplifying transformations on join condition. e.g. reduce same expressions: a=b AND b=a -> a=b, simplify boolean expressions: x = 1 AND FALSE -> FALSE, simplify cast expressions: CAST('123' as integer) -> 123
Rule configuration.
An ExecNodeTranslator that will generate only ONE Transformation.
Base class for built-in single value aggregate function.
Built-in boolean single value aggregate function.
Built-in byte single value aggregate function.
Built-in char single value aggregate function.
Built-in date single value aggregate function.
Built-in decimal single value aggregate function.
Built-in double single value aggregate function.
Built-in float single value aggregate function.
Built-in int single value aggregate function.
Built-in long single value aggregate function.
Built-in short single value aggregate function.
Built-in string single value aggregate function.
Built-in time single value aggregate function.
Built-in timestampLtz single value aggregate function.
Built-in timestamp single value aggregate function.
An interface that can not only serialize/deserialize the sink abilities to/from JSON, but also can apply the abilities to a DynamicTableSink.
Some functions like CUME_DIST/PERCENT_RANK/NTILE need the size of current window for calculation.
A windowing strategy that gets windows from input columns as window slice have been assigned and attached to the physical columns.
Represents the name-resolution context for expressions in an FOR SYSTEM_TIME AS OF TIMESTAMP clause.
Utility for SortMergeJoinOperator.
SortSpec describes how the data will be sorted.
Sort info for a Field.
SortSpec builder.
A sub-class of FlinkContext which contains the information for SourceAbilitySpec to apply the following abilities to DynamicTableSource.
An interface that can not only serialize/deserialize the source abilities to/from JSON, but also can apply the abilities to a DynamicTableSource.
Base class for SourceAbilitySpec.
A sub-class of SourceAbilitySpec that can not only serialize/deserialize the boolean flag whether a source watermark should be used to/from JSON, but can also apply it to SupportsSourceWatermark.
Rule will split a FlinkLogicalCalc which is the upstream of a FlinkLogicalCorrelate and contains Python Functions in condition into two FlinkLogicalCalcs.
Rule configuration.
The class to get SqlAggFunctionVisitor of CallExpression.
A converter for SqlAlterCatalogComment.
A converter for SqlAlterCatalogOptions.
A converter for SqlAlterCatalogReset.
A converter for SqlAlterMaterializedTableAsQuery.
A converter for SqlAlterMaterializedTableRefresh.
A converter for SqlAlterMaterializedTableResume.
A converter for SqlAlterMaterializedTableSuspend.
A converter for SqlAddPartitions.
A converter for SqlDropPartitions.
A converter for SqlAlterViewAs.
A converter for SqlAlterViewProperties.
A converter for SqlAlterViewRename.
SqlOperator for ARRAY, which makes explicit casting if the element type not equals the derived component type.
An internal [[SqlAggFunction]] to represents auxiliary group keys which will not be computed as key and does not also affect the correctness of the final result.
SqlCastFunction.
A bridge between a Flink's specific CatalogView and a Calcite's RelOptTable.
A converter for SqlCreateCatalog.
A converter for SqlCreateMaterializedTable.
A converter for SqlCreateView.
SqlCumulateTableFunction implements an operator for cumulative.
Marker for optional arguments in a signature of StaticArguments inserted by FlinkSqlCallBinding.
A converter for SqlDescribeCatalog.
A converter for SqlRichDescribeFunction.
A converter for SqlDescribeJob.
DESCRIPTOR appears as an argument in a function.
A converter for SqlDropMaterializedTable.
FIRST_VALUE and LAST_VALUE aggregate functions return the first or the last value in a list of values that are input to the function.
SQL function that computes keys by which rows can be partitioned and aggregated.
SqlHopTableFunction implements an operator for hopping.
LISTAGG aggregate function returns the concatenation of a list of values that are input to the function.
SqlOperator for MAP, which makes explicit casting if the element type not equals the derived component type.
An implementation of SqlNodeConverter.ConvertContext.
SqlNodeConverter<S extends org.apache.calcite.sql.SqlNode>
A converter to convert SqlNode instance into Operation.
Context of SqlNodeConverter.
Registry of SqlNode converters.
Mix-in tool class for SqlNode that allows DDL commands to be converted to Operation.
A converter for call procedure node.
A converter for query SqlNode, e.g., SqlSelect, SqlOrderBy, SqlTableRef.
A converter for SqlReplaceTableAs.
Copied to keep null semantics of table api and sql in sync.
SqlSessionTableFunction implements an operator for per-key sessionization.
A converter for SqlShowCatalogs.
A converter for SqlShowCreateCatalog.
A converter for SqlShowDatabases.
A converter for SqlShowFunctions.
A converter for SqlShowPartitions.
A converter for SqlShowProcedures.
 
 
Marker for table arguments in a signature of StaticArguments inserted by FlinkConvertletTable.
Function used to throw an exception, only used internally.
The TIMESTAMPADD function, which adds an interval to a datetime (TIMESTAMP, TIME or DATE).
Converts a SQL parse tree (consisting of SqlNode objects) into a relational algebra expression (consisting of RelNode objects).
Interface to define the configuration for a SqlToRelConverter.
Visitor that looks for an SqlIdentifier inside a tree of SqlNode objects and return Boolean.TRUE when it finds one.
Converts SQL expressions to RexNode.
A converter for SqlTruncateTable.
This class implements the TRY_CAST built-in, essentially delegating all the method invocations, whenever is possible, to Calcite's SqlCastFunction.
SqlTumbleTableFunction implements an operator for tumbling.
Default implementation SqlTypeFactoryImpl, the class was copied over because of FLINK-31350.
Default implementation of SqlValidator, the class was copied over because of CALCITE-4554.
Common base class for DML statement namespaces.
Utility object used to maintain information about the parameters in a function call.
Information about an identifier in a particular scope.
Validation status.
Utility methods related to SQL validation.
Extends SqlValidatorImpl to allow to read the schema with a specific version.
Base class for a table-valued function that computes windows.
Partial implementation of operand type checker.
Standard implementation of SqlRexConvertletTable.
It is used to describe the state metadata of a stateful operator, which is serialized/deserialized into/from those StreamExecNodes that can generate stateful operators.
Hint strategy to configure different ExecutionConfigOptions.IDLE_STATE_RETENTION for stream joins.
Utility class to load job-wide exchange mode.
The base class for aggregate ExecNode.
Stream ExecNode for AsyncScalarFunction.
Stream ExecNode for Calc.
Stream ExecNode which normalizes a changelog stream which maybe an upsert stream or a changelog stream containing duplicate events.
Stream ExecNode which matches along with join a Java/Scala user defined table function.
Stream ExecNode to connect a given DataStream and consume data from it.
Stream ExecNode which deduplicate on keys and keeps only first row or last row.
Stream ExecNode which will drop the UPDATE_BEFORE messages.
This ExecNode represents a change of partitioning of the input elements for stream.
Stream ExecNode that can expand one row to multiple rows based on given projects.
Stream ExecNode for unbounded global group aggregate.
Stream ExecNode for window table-valued based global aggregate.
Stream ExecNode for unbounded group aggregate.
Stream ExecNode for unbounded java/scala group table aggregate.
Stream ExecNode for either group window aggregate or group window table aggregate.
Stream ExecNode for unbounded incremental group aggregate.
StreamExecNode for a time interval stream join.
StreamExecNode for regular Joins.
Stream ExecNode to write data into an external sink defined by a TableSink.
Stream ExecNode to read data from an external source defined by a StreamTableSource.
Stream ExecNode for Limit.
Stream ExecNode for unbounded local group aggregate.
Stream ExecNode for window table-valued based local aggregate.
StreamExecNode for temporal table join that implemented by lookup.
Stream ExecNode which matches along with MATCH_RECOGNIZE.
Stream ExecNode which injects a mini-batch event in the streaming data.
Stream ExecNode for multiple input which contains a sub-graph of ExecNodes.
Base class for stream ExecNode.
Stream ExecNode for time-based over operator.
Stream ExecNode for Python ScalarFunctions.
Stream exec node which matches along with join a Python user defined table function.
Stream ExecNode for Python unbounded group aggregate.
Stream ExecNode for unbounded python group table aggregate.
Stream ExecNode for group widow aggregate (Python user defined aggregate function).
Stream ExecNode for python time-based over operator.
Stream ExecNode for Rank.
Stream ExecNode to write data into an external sink defined by a DynamicTableSink.
StreamExecNode for Sort.
StreamExecNode for Sort with limit.
Stream ExecNode to read data from an external source defined by a ScanTableSource.
StreamExecNode for temporal table join (FOR SYSTEM_TIME AS OF) and temporal TableFunction join (LATERAL TemporalTableFunction(o.proctime)).
StreamExecNode for time-ascending-order Sort without `limit`.
Stream ExecNode that is not a physical node and just union the inputs' records into one node.
Stream ExecNode that read records from given values.
Stream ExecNode which generates watermark based on the input elements.
Stream ExecNode for window table-valued based aggregate.
The base class for window aggregate ExecNode.
Stream ExecNode for Window Deduplicate.
StreamExecNode for WindowJoin.
Stream ExecNode for WindowRank.
Stream ExecNode which acts as a table-valued function to assign a window for each row of the input relation.
The StreamNonDeterministicPhysicalPlanResolver tries to resolve the correctness issue caused by 'Non-Deterministic Updates' (NDU) in a changelog pipeline.
An inner visitor to validate if there's any NDU problems which may cause wrong result and try to rewrite lookup join node with materialization (to eliminate the non-deterministic result generated by lookup join node only).
Stream physical RelNode for AsyncScalarFunction.
A physical rule for identifying logical async Calcs and converting it to a physical RelNode.
Rule to remove trivial StreamPhysicalCalc.
Converts FlinkLogicalTableFunctionScan with constant RexCall.
The physical rule is responsible for convert FlinkLogicalCorrelate to StreamPhysicalPythonCorrelate.
Rule to convert a FlinkLogicalAggregate into a StreamPhysicalPythonGroupAggregate.
Rule to convert a FlinkLogicalTableAggregate into a StreamPhysicalPythonGroupTableAggregateRule.
The physical rule is responsible for converting FlinkLogicalWindowAggregate to StreamPhysicalPythonGroupWindowAggregate.
The physical rule is responsible for converting FlinkLogicalOverAggregate to StreamPhysicalPythonOverAggregate.
Rule to convert a FlinkLogicalJoin into a StreamPhysicalWindowJoin.
Configuration for StreamPhysicalWindowJoinRule.
The RelDataType representation of a StructuredType.
SubQueryDecorrelator finds all correlated expressions in a SubQuery, and gets an equivalent non-correlated relational expression tree and correlation conditions.
Result describing the relational expression after decorrelation and where to find the equivalent non-correlated expressions and correlated conditions.
Pull out all correlation conditions from a given subquery to top level, and rebuild the subquery rel tree without correlation conditions.
built-in sum0 aggregate function.
Built-in Byte Sum0 aggregate function.
Built-in Decimal Sum0 aggregate function.
Built-in Double Sum0 aggregate function.
Built-in Float Sum0 aggregate function.
Built-in Int Sum0 aggregate function.
Built-in Long Sum0 aggregate function.
Built-in Short Sum0 aggregate function.
built-in sum aggregate function.
Built-in Byte Sum aggregate function.
Built-in Decimal Sum aggregate function.
Built-in Double Sum aggregate function.
Built-in Float Sum aggregate function.
Built-in Int Sum aggregate function.
Built-in Long Sum aggregate function.
Built-in Short Sum aggregate function.
built-in sum aggregate function with retraction.
Built-in Byte Sum with retract aggregate function.
Built-in Decimal Sum with retract aggregate function.
Built-in Double Sum with retract aggregate function.
Built-in Float Sum with retract aggregate function.
Built-in Int Sum with retract aggregate function.
Built-in Long Sum with retract aggregate function.
Built-in Short Sum with retract aggregate function.
Utilities to map between symbols from both Calcite and Flink.
Serializable representation of a symbol that can be used for persistence.
Utility class for TableConfig related helper functions.
Basic table lineage dataset which has catalog context and table in it.
Implementation for TableLineageDataSet.
Util class for building table lineage graph.
Sink lineage vertex for table.
Implementation of TableSinkLineageVertex.
Source lineage vertex for table.
Implementation of TableSourceLineageVertex.
Planner rule that rewrites temporal join with extracted primary key, Event-time temporal table join requires primary key and row time attribute of versioned table.
Rule configuration.
Utility for temporal table join which will gradually replace the scala class TemporalJoinUtil.
TemporalTableSpec describes how the right tale of lookupJoin ser/des.
A windowing strategy that gets windows by calculating on time attribute column.
The implementation of SchemaVersion to specify the snapshot at the specific time.
Utility functions for calcite's TimestampString.
This POJO is meant to hold some metadata information about operators, which usually needs to be passed to "factory" methods for Transformation.
Provider that produces a Transformation as a runtime implementation for ScanTableSource.
Provider that produces a Transformation as a runtime implementation for DynamicTableSink.
Logical representation of a tumbling window specification.
Rule that matches StreamPhysicalGroupAggregate on StreamPhysicalExchange with the following condition: 1. mini-batch is enabled in given TableConfig, 2. two-phase aggregation is enabled in given TableConfig, 3. all aggregate functions are mergeable, 4. the input of exchange does not satisfy the shuffle distribution,
Rule configuration.
Rule that matches StreamPhysicalWindowAggregate on StreamPhysicalExchange with following condition: the applied windowing is not on processing-time, because processing-time should be materialized in a single node.
A SqlOperandTypeChecker backed by TypeInference.
A SqlOperandTypeInference backed by TypeInference.
A SqlReturnTypeInference backed by TypeInference.
Planner rule that converts Uncollect values to TableFunctionScan.
 
Exception for unwanted method calling on unresolved expression.
Lists all kinds of ModifyKind.UPDATE operation.
Utility for upsertKey which represented as a Set of ImmutableBitSet.
Transpose StreamPhysicalWatermarkAssigner past into StreamPhysicalChangelogNormalize.
Rule configuration.
A sub-class of SourceAbilitySpec that can not only serialize/deserialize the watermark to/from JSON, but also can push the watermark into a SupportsWatermarkPushDown.
Rule to convert complex aggregation functions into simpler ones.
A windowing strategy that gets windows from input columns as windows have been assigned and attached to the physical columns.
Planner rule that makes the over window groups which have the same shuffle keys and order keys together.
Rule configuration.
Logical representation of a windowing strategy.
 
 
Rule configuration.
 
Rule configuration.
Logical representation of a window specification.
Utilities for Window Table Function.
Transforms JSON aggregation functions by wrapping operands into BuiltInFunctionDefinitions.JSON_STRING.
A sub-class of SinkAbilitySpec that can not only serialize/deserialize the metadata columns to/from JSON, but also can write the metadata columns for SupportsWritingMetadata.