Package com.pervasive.datarush.functions

Provides classes and interfaces related to defining functions on records.

See: Description

Interface Summary

Interface	Description
ArgumentConverter	An ArgumentConverter is used by the DefaultFunctionInvoker to convert ScalarValuedFunction arguments to the correct type.
EvaluatorFactory	A factory used to construct evaluators for ScalarValuedFunctions.
FunctionContext	Information about the context in which a function is evaluated.
FunctionEvaluator	An executor of a function.
FunctionInvoker	A FunctionInvoker is used by the FunctionRegistry to invoke a function method and by the AggregationInvoker to invoke an aggregation method. The implementing class essentially translates an array of ScalarValuedFunction arguments into the proper format for a given function method. A function method can specify its invoker using the Invoker annotation.
FunctionProvider	A provider of registered classes for the FunctionRegistry.
FunctionTyper	Determines the upper bound and result type of a ScalarValuedFunction.
NondeterministicFunctionEvaluator	An executor of a function which is non-deterministic.
ScalarValuedFunction	The description of a function taking record data and producing a scalar result.

Class Summary

Class	Description
Actions	Provides implementations of action functions.
Arithmetic	Provides implementations of common arithmetic functions.
Conditionals	Provides implementations of common conditional evaluation functions.
ConstantArgumentConverter	An abstract argument converter for converting String constants to the correct object type. Child classes should implement the convert(String) object to produce an object.
ConstantReference	A reference to a known value.
Conversions	Provides implementations of common type conversion functions.
CoreFunctionProvider	Registers core classes containing ScalarValuedFunction methods.
DateFormatArgumentConverter
DateTime	Provides implementations of common date/time functions.
DateTimeValueArgumentConverter	Converts an argument to a DateTimeValue enum value.
DefaultArgumentConverter	The default argument converter.
DefaultFunctionInvoker	A default function invoker, which passes arguments directly to the method using an ArgumentConverter. Arguments may specify a converter using the FunctionArgument annotation.
EvaluationContext	An object describing the context in which a subexpression of a larger expression is being evaluated.
FieldReference	A reference to a field in a record schema.
FixedType	A typer for functions which return a fixed, known type.
FormatFunctionInvoker	A specialized function invoker for format.
Formatting	Provides implementations of common formatting functions.
FunctionDescription	A description of a function, including its name, description, and argument names.
FunctionRegistry	Stores a list of functions registered by the FunctionTypeResoultionProvider, and provides a method for calling those functions.
FunctionUtil	A collection of useful utilities for implementing classes related to functions.
GranularityArgumentConverter	Converts an argument to a Granularity enum value.
InputMappedFactory	A factory mapped by the widest type encompassing one or more input values.
ListFunctionInvoker	A function invoker which passes all of its arguments as a single list to the method.
MappedEvaluatorFactory	A base class for factories which choose an evaluator based on a token type.
MappedEvaluatorFactory.Mapping	Defines a mapping between data type and an evaluator class.
Math	Provides implementations of common math functions found in the Math library.
NumberFormatArgumentConverter	Converts an argument to a NumberFormat object.
Predicates	Provides implementations of common boolean and logical functions.
ReflectiveFactory	A factory creating a fixed class of function evaluator.
ResultMappedFactory	A factory mapped by the expected type of the result.
ScalarFunctionDescriptor	Describes an implementation of a function.
ScalarFunctionDescriptor.SubexpressionBound	Specifies a bound on subexpressions used as arguments to a function.
SimpleFunctionProvider	Convenient base class for implementations of FunctionProvider.
Statistics	Provides implementations of common statistical functions.
Strings	Provides implementations of common string functions.
TimeZoneArgumentConverter	Converts an argument to a TimeZone object.
WavgFunctionInvoker	A specialized function invoker for wavg.
WidestInputTyper	A typer determining the output type of a function based on the types of one or more inputs.

Enum Summary

Enum	Description
DateTime.DateTimeValue	Definitions of date and time values that can be extracted from date and timestamp types.
DateTime.Granularity	The desired granularity of the output.
Strings.NullConcatenation	Specifies behavior of null values for string concatenation.

Exception Summary

Exception	Description
InvalidFunctionArgumentException	Exception thrown when an invalid argument is passed to a function within an expression.
InvalidOperandTypeException	An exception indicating the data type of a subexpression is not valid for use in the containing expression.

Package com.pervasive.datarush.functions Description

Provides classes and interfaces related to defining functions on records.

Many operators on records can be expressed as functions. These take a record from the input, perform some computation on the records, and emit a new record on the output. This can further be broken down into a sequence of functions producing each output field. These take values from one or more (or even no) fields from the record, and compute a single value representing the field's value.

Because functions encompass a much smaller range of functionality than do operators, they can generally be implemented much more quickly and easily. Functions, like operators, are composable - multiple functions can be combined to build more complex processing. Additionally, multiple functions can be applied to a record, yielding multiple values which can be appended or overlaid on the original records.

Due to this composability, functions enjoy some performance advantages over operators as they can combine the effects of several simple operators into one, reducing the total overhead related to the execution of operators. It should be noted that not all operators can be expressed as functions; only operators which produce one record of output per each record of input can be expressed in this way.

See Also:

DeriveFields, FilterRows, FieldCopier