CustomFloat

Allows user code to define custom floating-point formats. These formats are for storage only; all operations on them are performed by first implicitly extracting them to real first. After the operation is completed the result can be stored in a custom floating-point value via assignment.

template CustomFloat(uint bits)
template CustomFloat(uint precision, uint exponentWidth, CustomFloatFlags flags = CustomFloatFlags.ieee)
struct CustomFloat(uint precision, uint exponentWidth, CustomFloatFlags flags, uint bias)
struct CustomFloat (
uint precision
uint exponentWidth
CustomFloatFlags flags
uint bias
) if (
isCorrectCustomFloat(precision, exponentWidth, flags)
) {
string bitfields() via import std.bitmanip : bitfields;;
template staticIndexOf(args...) via import std.meta : staticIndexOf;;

static if(precision == 64)
static if(isIEEEQuadruple)
enum ulong significand_max;
version(LittleEndian)

static if(precision == 64)
static if(isIEEEQuadruple)
ulong significand;
version(!LittleEndian)

static if(precision == 64)
static if(isIEEEQuadruple)
ulong significand;
static if(precision == 64)
static if(!(isIEEEQuadruple))
ulong significand;
static if(precision == 64)
static if(!(isIEEEQuadruple))
enum ulong significand_max;
static if(flags & Flags.infinity)

static CustomFloat infinity [@property getter];
static if(flags & Flags.nan)

static CustomFloat nan [@property getter];
static size_t dig [@property getter];
static CustomFloat epsilon [@property getter];
enum mant_dig;
static int max_10_exp [@property getter];
enum max_exp;
static int min_10_exp [@property getter];
enum min_exp;
static CustomFloat max [@property getter];
static CustomFloat min_normal [@property getter];
CustomFloat re [@property getter];
static CustomFloat im [@property getter];
this(F input);
void opAssign(F input);
void opAssign(F input);
F get [@property getter];
alias opCast = get;
real opUnary();
real opBinary(T b);
real opBinary(T b);
real opBinaryRight(T a);
int opCmp(T b);
void opOpAssign(T b);
template toString();
}

Constructors

this this(F input): Initialize from any real compatible type.

Members

Aliases

opCast alias opCast = get: Fetches the stored value either as a float, double or real.

Functions

opAssign void opAssign(F input): Self assignment
opAssign void opAssign(F input): Assigns from any real compatible type.
opBinary real opBinary(T b)
opBinaryRight real opBinaryRight(T a)
opCmp int opCmp(T b)
opOpAssign void opOpAssign(T b): Convert the CustomFloat to a real and perform the relevant operator on the result
opUnary real opUnary(): Convert the CustomFloat to a real and perform the relevant operator on the result

Manifest constants

mant_dig enum mant_dig;: the number of bits in mantissa
max_exp enum max_exp;: maximum int value such that 2<sup>max_exp-1</sup> is representable
min_exp enum min_exp;: minimum int value such that 2<sup>min_exp-1</sup> is representable as a normalized value

Properties

dig size_t dig [@property getter]
epsilon CustomFloat epsilon [@property getter]
get F get [@property getter]: Fetches the stored value either as a float, double or real.
im CustomFloat im [@property getter]
infinity CustomFloat infinity [@property getter]
max CustomFloat max [@property getter]
max_10_exp int max_10_exp [@property getter]
min_10_exp int min_10_exp [@property getter]
min_normal CustomFloat min_normal [@property getter]
nan CustomFloat nan [@property getter]
re CustomFloat re [@property getter]

Templates

toString template toString(): Convert the CustomFloat to a real and perform the relevant operator on the result