Document:     floating

Purpose:      Describes the routines which handle floating point
              conversion.

Category:     DATA

File:         floating.c

Author:       K.G. Begeman

Description:  The routines which deal with single and double precision
              floating point types are:

              Routine      Function
              SPFPFL       convert from foreign single precision
                           representation to local
              SPFPLF       convert from local single precision
                           representation to foreign
              DPFPFL       convert from foreign double precision
                           representation to local
              DPFPLF       convert from local double precision
                           representation to foreign
              FBLANK       tests whether argument is a (local)
                           universal single precision blank
              SETFBLANK    sets argument to (local) universal
                           single precision blank
              SETNFBLANK   sets an array to (local) universal
                           single precision blank
              DBLANK       tests whether argument is a (local)
                           universal double precision blank
              SETDBLANK    sets argument to (local) universal
                           double precision blank
              SETNDBLANK   sets an array to (local) universal
                           double precision blank

              The local type of floating point must be defined in "osdef.h"
              via the OS_FLOATING_TYPE macro. The following types are
              recognized:

              OS_FLOATING_TYPE          Type of floating point
                     0                  IEEE High_Endian (BLANK=-Inf)
                     1                  IEEE Low_Endian (BLANK=-Inf)
                     2                  CONVEX Native
                     3                  VMS D_Floating
                     4                  VMS G_Floating
                     5                  IEEE High_Endian (BLANK=-FLT_MAX)
                     6                  IEEE Low_Endian (BLANK=-FLT_MAX)

              The routines mentioned above are separately documented.

Updates:      Dec 10, 1991: Document created.