go-math

A Go library for converting floating-point values between IEEE 754 and legacy binary formats used by IBM mainframes, VAX, and Cray supercomputers.

Formats

Format	Variable	Float32	Float64
IEEE 754	IEEE	standard	standard
IBM Hexadecimal	IBMHex	single-precision	double-precision
VAX	Vax	F format	G format
Cray NDR	Cray	IEEE big-endian	Cray native

Install

go get github.com/oiweiwei/go-math

Usage

Each format implements the FloatFormat interface:

type FloatFormat interface {
    Float32bits(float32) uint32
    Float32frombits(uint32) float32
    Float64bits(float64) uint64
    Float64frombits(uint64) float64
}

Convert a float32 to IBM Hexadecimal bits

import math "github.com/oiweiwei/go-math"

bits := math.IBMHex.Float32bits(123.45)       // float32 -> IBM hex uint32
val  := math.IBMHex.Float32frombits(bits)     // IBM hex uint32 -> float32

Convert a float64 to VAX G format

bits := math.Vax.Float64bits(3.141592653589793)   // float64 -> VAX G uint64
val  := math.Vax.Float64frombits(bits)            // VAX G uint64 -> float64

Low-level functions

Individual conversion functions are also exported:

// IBM Hexadecimal
math.IBMHexfloat32bits(v float32) uint32
math.IBMHexfloat32frombits(b uint32) float32
math.IBMHexfloat64bits(v float64) uint64
math.IBMHexfloat64frombits(b uint64) float64

// VAX F (single) and G (double)
math.VaxFfloat32bits(v float32) uint32
math.VaxFfloat32frombits(b uint32) float32
math.VaxGfloat64bits(v float64) uint64
math.VaxGfloat64frombits(b uint64) float64

// Cray NDR
math.CrayFloat32bits(v float32) uint32
math.CrayFloat32frombits(b uint32) float32
math.CrayFloat64bits(v float64) uint64
math.CrayFloat64frombits(b uint64) float64

License

MIT