Package org.apache.druid.math.expr.vector

Class FallbackVectorProcessor<T>

java.lang.Object

org.apache.druid.math.expr.vector.FallbackVectorProcessor<T>

All Implemented Interfaces:

ExprVectorProcessor<T>

public abstract class FallbackVectorProcessor<T>
extends Object
implements ExprVectorProcessor<T>

Implementation of ExprVectorProcessor that adapts non-vectorized Expr.eval(Expr.ObjectBinding). This allows non-vectorized expressions to participate in vectorized queries.

Field Summary

Fields

Modifier and Type	Field	Description
protected final ExpressionType	outputType

Method Summary

Modifier and Type	Method	Description
static boolean	canFallbackVectorize(List<Expr> args, ExpressionType outputType, Expr.InputBindingInspector inspector)	Returns whether create(Function, List, Expr.VectorInputBindingInspector) can be used to make a fallback vectorized processor for a function with the given output type and arguments.
static boolean	canFallbackVectorize(LambdaExpr lambdaExpr, List<Expr> args, ExpressionType outputType, Expr.InputBindingInspector inspector)	Returns whether create(ApplyFunction, LambdaExpr, List, Expr.VectorInputBindingInspector) can be used to make a fallback vectorized processor for a function with the given output type and arguments.
static <T> ExprVectorProcessor<T>	create(ApplyFunction function, LambdaExpr lambdaExpr, List<Expr> args, Expr.VectorInputBindingInspector inspector)	Create a processor for a non-vectorizable ApplyFunction.
static <T> ExprVectorProcessor<T>	create(ExprMacroTable.ExprMacro macro, List<Expr> args, Expr.VectorInputBindingInspector inspector)	Create a processor for a non-vectorizable ExprMacroTable.ExprMacro.
static <T> ExprVectorProcessor<T>	create(Function function, List<Expr> args, Expr.VectorInputBindingInspector inspector)	Create a processor for a non-vectorizable Function.
ExpressionType	getOutputType()

Methods inherited from class java.lang.Object

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

Methods inherited from interface org.apache.druid.math.expr.vector.ExprVectorProcessor

evalVector, maxVectorSize

Field Details

outputType

protected final ExpressionType outputType

Method Details

create

public static <T> ExprVectorProcessor<T> create(Function function,
 List<Expr> args,
 Expr.VectorInputBindingInspector inspector)

Create a processor for a non-vectorizable Function.

create

public static <T> ExprVectorProcessor<T> create(ApplyFunction function,
 LambdaExpr lambdaExpr,
 List<Expr> args,
 Expr.VectorInputBindingInspector inspector)

Create a processor for a non-vectorizable ApplyFunction.

create

public static <T> ExprVectorProcessor<T> create(ExprMacroTable.ExprMacro macro,
 List<Expr> args,
 Expr.VectorInputBindingInspector inspector)

Create a processor for a non-vectorizable ExprMacroTable.ExprMacro.

canFallbackVectorize

public static boolean canFallbackVectorize(List<Expr> args,
 @Nullable
 ExpressionType outputType,
 Expr.InputBindingInspector inspector)

Returns whether create(Function, List, Expr.VectorInputBindingInspector) can be used to make a fallback vectorized processor for a function with the given output type and arguments.

canFallbackVectorize

public static boolean canFallbackVectorize(LambdaExpr lambdaExpr,
 List<Expr> args,
 @Nullable
 ExpressionType outputType,
 Expr.InputBindingInspector inspector)

Returns whether create(ApplyFunction, LambdaExpr, List, Expr.VectorInputBindingInspector) can be used to make a fallback vectorized processor for a function with the given output type and arguments.

getOutputType

public ExpressionType getOutputType()

Specified by:

getOutputType in interface ExprVectorProcessor<T>