com.hazelcast.query.impl

Class Numbers

java.lang.Object

com.hazelcast.query.impl.Numbers

public final class Numbers
extends Object

Provides utilities which compare and canonicalize Number instances.

Method Summary

Modifier and Type	Method and Description
static double	asDoubleExactly(Number number) Represents the given Number exactly as a double value without any magnitude and precision losses; if that's not possible, fails by throwing an exception.
static int	asIntExactly(Number number) Represents the given Number exactly as an int value without any magnitude and precision losses; if that's not possible, fails by throwing an exception.
static long	asLongExactly(Number number) Represents the given Number exactly as a long value without any magnitude and precision losses; if that's not possible, fails by throwing an exception.
static Comparable	canonicalizeForHashLookup(Comparable value) Canonicalizes the given Number value for the purpose of a hash-based lookup.
static int	compare(Comparable lhs, Comparable rhs) Compares the provided Number instances.
static boolean	equal(Number lhs, Number rhs) Checks the provided Number instances for equality.
static boolean	equalDoubles(double lhs, double rhs) Checks the provided double values for equality exactly as Double.equals(java.lang.Object) does it, avoiding the boxing-unboxing cycle.
static boolean	equalFloats(float lhs, float rhs) Checks the provided float values for equality exactly as Float.equals(java.lang.Object) does it, avoiding the boxing-unboxing cycle.
static boolean	equalLongAndDouble(long l, double d) Checks the provided long and double values for equality.
static boolean	isDoubleRepresentable(Class clazz)
static boolean	isLongRepresentable(Class clazz)

Methods inherited from class java.lang.Object

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

Method Detail

equal

public static boolean equal(Number lhs,
                            Number rhs)

Checks the provided Number instances for equality.

Special numeric comparison logic is used for Double, Long, Float, Integer, Short and Byte: two values are considered equal, if both represent a number of the same magnitude and sign, disregarding the actual underlying types.

Parameters:

lhs - the left-hand side Number. Can't be null.

rhs - the right-hand side Number. Can't be null.

Returns:

true if the provided numbers are equal, false otherwise.

compare

public static int compare(Comparable lhs,
                          Comparable rhs)

Compares the provided Number instances.

The method accepts Comparable instances because to compare something it must be comparable, but Number is not comparable.

Special numeric comparison logic is used for Double, Long, Float, Integer, Short and Byte: comparison is performed by taking into account only the magnitudes and signs of the passed numbers, disregarding the actual underlying types.

Parameters:

lhs - the left-hand side Number. Can't be null.

rhs - the right-hand side Number. Can't be null.

Returns:

a negative integer, zero, or a positive integer as the left-hand side Number is less than, equal to, or greater than the right-hand side Number.

canonicalizeForHashLookup

public static Comparable canonicalizeForHashLookup(Comparable value)

Canonicalizes the given Number value for the purpose of a hash-based lookup.

The method accepts a Comparable instance and returns one because query engine acts on comparable values, but Number is not comparable.

Special numeric canonicalization logic is used for Double, Long, Float, Integer, Short and Byte: all whole numbers in the [Long.MIN_VALUE, Long.MAX_VALUE] range are represented as Long, all other whole and non-whole numbers are presented as Double. That logic allows to mix numeric types while performing hash-based lookups, e.g. while using one of the standard Map implementations.

Parameters:

value - the Number to canonicalize.

Returns:

a canonical representation of the given Number or the original Number if there is no special canonical representation for it.

equalDoubles

public static boolean equalDoubles(double lhs,
                                   double rhs)

Checks the provided double values for equality exactly as Double.equals(java.lang.Object) does it, avoiding the boxing-unboxing cycle.

Parameters:

lhs - the left-hand side double value.

rhs - the right-hand side double value.

Returns:

true if the given values are equal, false otherwise.

equalFloats

public static boolean equalFloats(float lhs,
                                  float rhs)

Checks the provided float values for equality exactly as Float.equals(java.lang.Object) does it, avoiding the boxing-unboxing cycle.

Parameters:

lhs - the left-hand side float value.

rhs - the right-hand side float value.

Returns:

true if the given values are equal, false otherwise.

asDoubleExactly

public static double asDoubleExactly(Number number)

Represents the given Number exactly as a double value without any magnitude and precision losses; if that's not possible, fails by throwing an exception.

Parameters:

number - the number to represent as a double value.

Returns:

a double representation of the given number.

Throws:

IllegalArgumentException - if no exact representation exists.

asLongExactly

public static long asLongExactly(Number number)

Represents the given Number exactly as a long value without any magnitude and precision losses; if that's not possible, fails by throwing an exception.

Parameters:

number - the number to represent as a long value.

Returns:

a long representation of the given number.

Throws:

IllegalArgumentException - if no exact representation exists.

asIntExactly

public static int asIntExactly(Number number)

Represents the given Number exactly as an int value without any magnitude and precision losses; if that's not possible, fails by throwing an exception.

Parameters:

number - the number to represent as an int value.

Returns:

an int representation of the given number.

Throws:

IllegalArgumentException - if no exact representation exists.

isDoubleRepresentable

public static boolean isDoubleRepresentable(Class clazz)

Returns:

true if instances of the given class can be represented as double values without any magnitude and precision losses, false otherwise.

isLongRepresentable

public static boolean isLongRepresentable(Class clazz)

Returns:

true if instances of the given class can be represented as long values without any magnitude and precision losses, false otherwise.

equalLongAndDouble

public static boolean equalLongAndDouble(long l,
                                         double d)

Checks the provided long and double values for equality.

The method fully avoids magnitude and precision losses while performing the equality check, e.g. (double) Long.MAX_VALUE is not equal to Long.MAX_VALUE - 1.

Parameters:

l - the long value to check the equality of.

d - the double value to check the equality of.

Returns:

true if the provided values are equal, false otherwise.