Class RealPos

Class Methods

Overrides: Position class

make an XuReal given an IEEE floating point number of whatever precision on this platform is able to hold all the real numbers currently representable by an XuReal. Currently this is IEEE64 (double precision), but may be redeclared as a larger IEEE precision in the future. See comment in XuReal::makeIEEE64

See comment in XuReal::makeIEEE64

Returns an XuReal which exactly represents the same real number that is represented by 'value'. BLASTs if value doesn't represent a real (ie., no NANs or inifinities). Negative 0 will be silently converted to positive zero

See comment in XuReal::makeIEEE64

Instance Methods

Overrides: Position

Returns the value as IEEE basic data type is big enough to hold any value which can be put into an XuReal. Currently this is an IEEE64 (double precision). In future releases of this API, the return type of this method may be changed to IEEE128 (quad precision). Once we support other ways of representing real numbers, there may not be an all-inclusive IEEE type, in which case this message will BLAST.

The only IEEE values which this will return are those that represent real numbers. Ie., no NANs, no inifinities, no negative zero.

Overridden by: IEEE32Pos IEEE64Pos IEEE8Pos

Returns the value as IEEE64 (double precision).
The only IEEE values which this will return are those that represent real numbers. Ie., no NANs, no inifinities, no negative zero.

Overridden by: IEEE32Pos IEEE64Pos IEEE8Pos

Overrides: Position

Overrides: Position

Overridden by: IEEE32Pos IEEE64Pos IEEE8Pos

Overrides: Position

Overrides: Position

Whether the real number that this object represents is exactly representable in an available IEEE precision. Currently the answer is always TRUE, and the available precisions are 8 (stupid precision), 32 (single precision), and 64 (double precision). If the answer is FALSE, the meaning of the messages 'precision' and 'asIEEE' remain to be defined.

This number represents exactly this->mantissa() * 2 ^ this->exponent(). Should we eventually support real numbers which cannot be expressed exactly with integral mantissa and exponent, then this message (and 'exponent') will BLAST for such numbers.

What precision is it, in terms of the number of bits used to represent it. In the interests of efficiency, this may return a number larger than that *needed* to represent it. However, the precision reported must be at least that needed to represent this number. It is assumed that the format of the number satisfies the IEEE radix independent floating point spec. Should we represent real numbers other that those representable in IEEE, the meaning of this message will be more fully specified.

The fact that this message is allowed to overestimate precision doesn't interfere with equality: a->isEqual(b) exactly when they represent that same real number, even if one of them happens to overestimate precision more that the other.

Overridden by: IEEE32Pos IEEE64Pos IEEE8Pos

Overridden by: IEEE32Pos IEEE64Pos IEEE8Pos

Essential. Return the number as a PrimFloat object from which you can get it in a variety of representations.

Overridden by: IEEE32Pos IEEE64Pos IEEE8Pos