libvaxdata: VAX Data Format Conversion Routines

ABSTRACT

libvaxdata provides a collection of routines for converting numeric data—integer and floating-point—to and from the formats used on a Digital Equipment Corporation¹ (DEC) VAX 32-bit minicomputer (Brunner, 1991). Since the VAX numeric data formats are inherited from those used on a DEC PDP-11 16-bit minicomputer, these routines can be used to convert PDP–11 data as well. VAX numeric data formats are also the default data formats used on DEC Alpha 64-bit minicomputers running OpenVMS

The libvaxdata routines are callable from Fortran or C. They require that the caller use two's-complement format for integer data and IEEE 754 format (ANSI/IEEE, 1985) for floating-point data. They also require that the "natural" size of a C int type (integer) is 32 bits. That is the case for most modern 32-bit and 64-bit computer systems. Nevertheless, you may wish to consult the Fortran or C compiler documentation on your system to be sure.

Some Fortran compilers support conversion of VAX numeric data on-the-fly when reading or writing unformatted files, either as a compiler option or a run-time I/O option. This feature may be easier to use than the libvaxdata routines. Consult the Fortran compiler documentation on your system to determine if this alternative is available to you.

¹Later Compaq Computer Corporation, now Hewlett-Packard Company