Modifier and Type	Class and Description
private static class	FloatingDecimal.ASCIIToBinaryBuffer
public static class	FloatingDecimal.BinaryToASCIIBuffer

Field Summary

Modifier and Type	Field and Description
private static final FloatingDecimal.BinaryToASCIIBuffer	B2AC_NEGATIVE_ZERO
private static final FloatingDecimal.BinaryToASCIIBuffer	B2AC_POSITIVE_ZERO
private static final int	BINARY_16_IX
private static final int	BINARY_32_IX
private static final int	BINARY_64_IX
private static final int[]	E_THR_I
private static final int[]	E_THR_Z
private static final long	EXP_ONE
private static final int	EXP_SHIFT
private static final int	FLOG_10_MAX_LONG
private static final long	FRACT_HOB
private static final int[]	HEX_COUNT
private static final String	INFINITY_REP
private static final int	L_HALF
private static final int	MAX_DEC_DIGITS
private static final int	MAX_OUT
private static final int	MAX_SMALL_BIN_EXP
private static final int	MIN_SMALL_BIN_EXP
private static final String	NAN_REP
private static final String	OMITTED
private static final int[]	P
private static final int[]	Q_MIN
private static final int[]	QP_MAX
private static final int[]	QP_MIN
private static final int	R_HALF
private static final ThreadLocal<FloatingDecimal.BinaryToASCIIBuffer>	threadLocalBinaryToASCIIBuffer

Constructor Summary

Access	Constructor and Description
public	FloatingDecimal()

Method Summary

Modifier and Type	Method and Description
private static long	appendDecDigit(long v, int ch)
private static double	buildDouble(int ssign, int q, long c)
private static float	buildFloat(int ssign, int q, long c)
private static void	check(String in, boolean expected)
private static int	checkMultiplePoints(int pt, int i)
private static void	copyDigits(String in, byte[] d, int len, int i)
private static int	digitFor(int ch)
private static FloatingDecimal.BinaryToASCIIBuffer	getBinaryToASCIIBuffer()
public static FloatingDecimal.BinaryToASCIIBuffer	Returns: The converter. getBinaryToASCIIConverter(double The double precision value to convert. d, boolean compatibility with releases < JDK 21 compat) Returns a `BinaryToASCIIConverter` for a `double`.
private static FloatingDecimal.BinaryToASCIIBuffer	getBinaryToASCIIConverter(double d)
private static FloatingDecimal.BinaryToASCIIBuffer	getCompatBinaryToASCIIConverter(double d)
private static boolean	isDigit(int ch, boolean isDec)
public static double	Returns: The double precision value. parseDouble(String The `String` to convert. s) Converts a `String` to a double precision floating point value.
public static double	Returns: The double-precision value of the conversion parseDoubleSignlessDigits(int The decimal exponent of the value to generate decExp, byte[] The digits of the significand. d, int Number of digits to use length) Converts a sequence of digits ('0'-'9') as well as an exponent to a positive double value
public static float	Returns: The single precision value. parseFloat(String The `String` to convert. s) Converts a `String` to a single precision floating point value.
private static double	Returns: an appropriate binary converter readJavaFormatString(String the non-null input in, int one of the `BINARY_<S>_IX` constants, where `<S>` is one of 16, 32, 64 ix) The input must match the `rules described here`, about leading and trailing whitespaces, and the grammar.
private static void	scanSymbolic(String in, int i, String sub)
private static int	skipWhitespaces(String in, int i)
private static int	toLowerCase(int ch)

Inherited from java.lang.Object:: clone equals finalize getClass hashCode notify notifyAll toString wait wait wait

Field Detail

B2AC_NEGATIVE_ZERO	back to summary
private static final FloatingDecimal.BinaryToASCIIBuffer B2AC_NEGATIVE_ZERO

B2AC_POSITIVE_ZERO	back to summary
private static final FloatingDecimal.BinaryToASCIIBuffer B2AC_POSITIVE_ZERO

BINARY_16_IX	back to summary
private static final int BINARY_16_IX

BINARY_32_IX	back to summary
private static final int BINARY_32_IX

BINARY_64_IX	back to summary
private static final int BINARY_64_IX

E_THR_I	back to summary
private static final int[] E_THR_I Annotations @Stable

E_THR_Z	back to summary
private static final int[] E_THR_Z Annotations @Stable

EXP_ONE	back to summary
private static final long EXP_ONE

EXP_SHIFT	back to summary
private static final int EXP_SHIFT

FLOG_10_MAX_LONG	back to summary
private static final int FLOG_10_MAX_LONG

FRACT_HOB	back to summary
private static final long FRACT_HOB

HEX_COUNT	back to summary
private static final int[] HEX_COUNT Annotations @Stable

INFINITY_REP	back to summary
private static final String INFINITY_REP

L_HALF	back to summary
private static final int L_HALF

MAX_DEC_DIGITS	back to summary
private static final int MAX_DEC_DIGITS

MAX_OUT	back to summary
private static final int MAX_OUT

MAX_SMALL_BIN_EXP	back to summary
private static final int MAX_SMALL_BIN_EXP

MIN_SMALL_BIN_EXP	back to summary
private static final int MIN_SMALL_BIN_EXP

NAN_REP	back to summary
private static final String NAN_REP

OMITTED	back to summary
private static final String OMITTED

P	back to summary
private static final int[] P Annotations @Stable

Q_MIN	back to summary
private static final int[] Q_MIN Annotations @Stable

QP_MAX	back to summary
private static final int[] QP_MAX Annotations @Stable

QP_MIN	back to summary
private static final int[] QP_MIN Annotations @Stable

R_HALF	back to summary
private static final int R_HALF

threadLocalBinaryToASCIIBuffer	back to summary
private static final ThreadLocal<FloatingDecimal.BinaryToASCIIBuffer> threadLocalBinaryToASCIIBuffer

Constructor Detail

FloatingDecimal	back to summary
public FloatingDecimal()

Method Detail

appendDecDigit	back to summary
private static long appendDecDigit(long v, int ch)

buildDouble	back to summary
private static double buildDouble(int ssign, int q, long c)

buildFloat	back to summary
private static float buildFloat(int ssign, int q, long c)

check	back to summary
private static void check(String in, boolean expected)

checkMultiplePoints	back to summary
private static int checkMultiplePoints(int pt, int i)

copyDigits	back to summary
private static void copyDigits(String in, byte[] d, int len, int i)

digitFor	back to summary
private static int digitFor(int ch)

getBinaryToASCIIBuffer	back to summary
private static FloatingDecimal.BinaryToASCIIBuffer getBinaryToASCIIBuffer()

getBinaryToASCIIConverter back to summary

getBinaryToASCIIConverter	back to summary
public static FloatingDecimal.BinaryToASCIIBuffer getBinaryToASCIIConverter(double d, boolean compat) Returns a `BinaryToASCIIConverter` for a `double`. The returned object is a `ThreadLocal` variable of this class. Parameters d:double The double precision value to convert. compat:boolean compatibility with releases < JDK 21 Returns:FloatingDecimal.BinaryToASCIIBuffer The converter.

public static FloatingDecimal.BinaryToASCIIBuffer getBinaryToASCIIConverter(double d, boolean compat)

Returns a BinaryToASCIIConverter for a double. The returned object is a ThreadLocal variable of this class.

Parameters

d:double: The double precision value to convert.
compat:boolean: compatibility with releases < JDK 21

Returns:FloatingDecimal.BinaryToASCIIBuffer

The converter.

getBinaryToASCIIConverter	back to summary
private static FloatingDecimal.BinaryToASCIIBuffer getBinaryToASCIIConverter(double d)

getCompatBinaryToASCIIConverter	back to summary
private static FloatingDecimal.BinaryToASCIIBuffer getCompatBinaryToASCIIConverter(double d)

isDigit	back to summary
private static boolean isDigit(int ch, boolean isDec)

parseDouble back to summary

parseDouble	back to summary
public static double parseDouble(String s) throws NumberFormatException Converts a `String` to a double precision floating point value. Parameters s:String The `String` to convert. Returns:double The double precision value. Exceptions NumberFormatException: If the `String` does not represent a properly formatted double precision value.

public static double parseDouble(String s) throws NumberFormatException

Converts a String to a double precision floating point value.

Parameters

s:String: The String to convert.

Returns:double

The double precision value.

Exceptions

NumberFormatException:: If the String does not represent a properly formatted double precision value.

parseDoubleSignlessDigits	back to summary
public static double parseDoubleSignlessDigits(int decExp, byte[] d, int length) Converts a sequence of digits ('0'-'9') as well as an exponent to a positive double value Parameters decExp:int The decimal exponent of the value to generate d:byte[] The digits of the significand. length:int Number of digits to use Returns:double The double-precision value of the conversion

parseDoubleSignlessDigits

back to summary

public static double parseDoubleSignlessDigits(int decExp, byte[] d, int length)

Converts a sequence of digits ('0'-'9') as well as an exponent to a positive double value

Parameters

decExp:int: The decimal exponent of the value to generate
d:byte[]: The digits of the significand.
length:int: Number of digits to use

Returns:double

The double-precision value of the conversion

parseFloat back to summary

parseFloat	back to summary
public static float parseFloat(String s) throws NumberFormatException Converts a `String` to a single precision floating point value. Parameters s:String The `String` to convert. Returns:float The single precision value. Exceptions NumberFormatException: If the `String` does not represent a properly formatted single precision value.

public static float parseFloat(String s) throws NumberFormatException

Converts a String to a single precision floating point value.

Parameters

s:String: The String to convert.

Returns:float

The single precision value.

Exceptions

NumberFormatException:: If the String does not represent a properly formatted single precision value.

readJavaFormatString back to summary

readJavaFormatString	back to summary
private static double readJavaFormatString(String in, int ix) The input must match the `rules described here`, about leading and trailing whitespaces, and the grammar. Parameters in:String the non-null input ix:int one of the `BINARY_<S>_IX` constants, where `<S>` is one of 16, 32, 64 Returns:double an appropriate binary converter Exceptions NullPointerException: if the input is null NumberFormatException: if the input is malformed

private static double readJavaFormatString(String in, int ix)

The input must match the rules described here, about leading and trailing whitespaces, and the grammar.

Parameters

in:String: the non-null input
ix:int: one of the BINARY_<S>_IX constants, where <S> is one of 16, 32, 64

Returns:double

an appropriate binary converter

Exceptions

NullPointerException:: if the input is null
NumberFormatException:: if the input is malformed

scanSymbolic	back to summary
private static void scanSymbolic(String in, int i, String sub)

skipWhitespaces	back to summary
private static int skipWhitespaces(String in, int i)

toLowerCase	back to summary
private static int toLowerCase(int ch)

jdk.internal.math back to summary

private final Class FloatingDecimal.ASCIIToBinaryBuffer

Class Inheritance

extends Object

java.lang.Object
jdk.internal.math.FloatingDecimal.ASCIIToBinaryBuffer

Field Summary

Modifier and Type	Field and Description
private final byte[]	d
private final int	e
private final boolean	isNegative
private static final int	MAX_SMALL_TEN
private final int	n
private static final int	SINGLE_MAX_SMALL_TEN
private static final float[]	SINGLE_SMALL_10_POW
private static final double[]	SMALL_10_POW

Constructor Summary

Access	Constructor and Description
private	ASCIIToBinaryBuffer(boolean isNegative, int e, byte[] d, int n)

Method Summary

Modifier and Type	Method and Description
private double	doubleValue()
private float	floatValue()
private double	signed(double v)
private float	signed(float v)
private long	toLong(int n)

Inherited from java.lang.Object:: clone equals finalize getClass hashCode notify notifyAll toString wait wait wait

Field Detail

d	back to summary
private final byte[] d

e	back to summary
private final int e

isNegative	back to summary
private final boolean isNegative

MAX_SMALL_TEN	back to summary
private static final int MAX_SMALL_TEN

n	back to summary
private final int n

SINGLE_MAX_SMALL_TEN	back to summary
private static final int SINGLE_MAX_SMALL_TEN

SINGLE_SMALL_10_POW	back to summary
private static final float[] SINGLE_SMALL_10_POW Annotations @Stable

SMALL_10_POW	back to summary
private static final double[] SMALL_10_POW Annotations @Stable

Constructor Detail

ASCIIToBinaryBuffer	back to summary
private ASCIIToBinaryBuffer(boolean isNegative, int e, byte[] d, int n)

Method Detail

doubleValue	back to summary
private double doubleValue()

floatValue	back to summary
private float floatValue()

signed	back to summary
private double signed(double v)

signed	back to summary
private float signed(float v)

toLong	back to summary
private long toLong(int n)

jdk.internal.math back to summary

public final Class FloatingDecimal.BinaryToASCIIBuffer

Class Inheritance

extends Object

java.lang.Object
jdk.internal.math.FloatingDecimal.BinaryToASCIIBuffer

Field Summary

Modifier and Type	Field and Description
private int	decExponent
private boolean	decimalDigitsRoundedUp
private final byte[]	digits
private boolean	exactDecimalConversion
private int	firstDigitIndex
private static final int[]	insignificantDigitsNumber If insignificant==(1L << ixd) i = insignificantDigitsNumber[idx] is the same as: int i; for ( i = 0; insignificant >= 10L; i++ ) insignificant /= 10L;
private static final int[]	N_5_BITS
private int	nDigits

Constructor Summary

Access	Constructor and Description
private	BinaryToASCIIBuffer() Default constructor; used for non-zero values, `BinaryToASCIIBuffer` may be thread-local and reused
private	BinaryToASCIIBuffer(byte[] digits) Creates a specialized value (positive and negative zeros).

Method Summary

Modifier and Type	Method and Description
public boolean	decimalDigitsExact()
private void	developLongDigits(long lvalue, int insignificantDigits) This is the easy subcase -- all the significant bits, after scaling, are held in lvalue.
public boolean	digitsRoundedUp()
private void	dtoa(int binExp, long fractBits, int nSignificantBits)
private static int	estimateDecExp(long fractBits, int binExp) Estimate decimal exponent.
public int	getDecimalExponent()
public int	getDigits(byte[] digits)
private static int	insignificantDigitsForPow2(int p2) Calculates
private void	roundup()

Inherited from java.lang.Object:: clone equals finalize getClass hashCode notify notifyAll toString wait wait wait

Field Detail

decExponent	back to summary
private int decExponent

decimalDigitsRoundedUp	back to summary
private boolean decimalDigitsRoundedUp

digits	back to summary
private final byte[] digits

exactDecimalConversion	back to summary
private boolean exactDecimalConversion

firstDigitIndex	back to summary
private int firstDigitIndex

insignificantDigitsNumber	back to summary
private static final int[] insignificantDigitsNumber If insignificant==(1L << ixd) i = insignificantDigitsNumber[idx] is the same as: int i; for ( i = 0; insignificant >= 10L; i++ ) insignificant /= 10L; Annotations @Stable

N_5_BITS	back to summary
private static final int[] N_5_BITS Annotations @Stable

nDigits	back to summary
private int nDigits

Constructor Detail

BinaryToASCIIBuffer	back to summary
private BinaryToASCIIBuffer() Default constructor; used for non-zero values, `BinaryToASCIIBuffer` may be thread-local and reused

BinaryToASCIIBuffer	back to summary
private BinaryToASCIIBuffer(byte[] digits) Creates a specialized value (positive and negative zeros).

Method Detail

decimalDigitsExact	back to summary
public boolean decimalDigitsExact()

developLongDigits

back to summary

private void developLongDigits(long lvalue, int insignificantDigits)

This is the easy subcase -- all the significant bits, after scaling, are held in lvalue. negSign and decExponent tell us what processing and scaling has already been done. Exceptional cases have already been stripped out. In particular: lvalue is a finite number (not Inf, nor NaN) lvalue > 0L (not zero, nor negative).

The only reason that we develop the digits here, rather than calling on Long.toString() is that we can do it a little faster, and besides want to treat trailing 0s specially. If Long.toString changes, we should re-evaluate this strategy!

digitsRoundedUp	back to summary
public boolean digitsRoundedUp()

dtoa	back to summary
private void dtoa(int binExp, long fractBits, int nSignificantBits)

estimateDecExp

back to summary

private static int estimateDecExp(long fractBits, int binExp)

Estimate decimal exponent. (If it is small-ish, we could double-check.)

First, scale the mantissa bits such that 1 <= d2 < 2. We are then going to estimate log10(d2) ~=~ (d2-1.5)/1.5 + log(1.5) and so we can estimate log10(d) ~=~ log10(d2) + binExp * log10(2) take the floor and call it decExp.

getDecimalExponent	back to summary
public int getDecimalExponent()

getDigits	back to summary
public int getDigits(byte[] digits)

insignificantDigitsForPow2	back to summary
private static int insignificantDigitsForPow2(int p2) Calculates

roundup	back to summary
private void roundup()

public final Class FloatingDecimal

Nested and Inner Type Summary

Field Summary

Constructor Summary

Method Summary

Field Detail

Constructor Detail

Method Detail

private final Class FloatingDecimal.ASCIIToBinaryBuffer

Field Summary

Constructor Summary

Method Summary

Field Detail

Constructor Detail

Method Detail

public final Class FloatingDecimal.BinaryToASCIIBuffer

Field Summary

Constructor Summary

Method Summary

Field Detail

Constructor Detail

Method Detail