Math:  export_int,  export_real,  export_real32,  import_int, import_real, import_real32: conversion to and from  external representation     include "math.m";    math := load Math Math->PATH;     export_int:    fn(b: array of byte, x: array of int);    export_real32: fn(b: array of byte, x: array of real);    export_real:   fn(b: array of byte, x: array of real);        import_int:    fn(b: array of byte, x: array of int);    import_real32: fn(b: array of byte, x: array of real);    import_real:   fn(b: array of byte, x: array of real); These  routines  convert between Limbo's internal representa‐tion of integer and  floating-point  values,  and  equivalentexternal  representations  as arrays of bytes, allowing effi‐cient input/output of binary representations. Export_int(b, x) converts each integer  element  of  array  xinto a sequence of 4 bytes in array b. Export_real(b, x)  converts  each  double-precision floating-point element of array x into a sequence of 8 bytes in  arrayb. Export_real32(b, x)  converts each double-precision floating-point element of array x to a  single-precision  value,  thenencodes it as a sequence of 4 bytes in array b. Each  import_T  operation  reverses the transformation of thecorresponding export_T  operation,  converting  an  array  ofbytes  containing a sequence of external representations intoan array of values of the appropriate internal form. Values are encoded in big-endian order (most significant bytefirst), with floating-point values represented in IEEE 32-bitor 64-bit form (again, most significant byte first). == source/libinterp/math.c == see ALSOmath-intro(2)