netcdf-c/libcf/nfconfig.in
2010-06-03 13:24:43 +00:00

84 lines
2.3 KiB
Plaintext

#if 0
$Id: nfconfig.in,v 1.1.1.1 2009/07/06 15:06:29 ed Exp $
#endif
#ifndef UD_NETCDF_CPP_INC
#define UD_NETCDF_CPP_INC
#if 0
Do not have C-style comments in here because this file is processed
by both the FORTRAN compiler (for the nf_test/ stuff) and the C
compiler (for the FORTRAN-callable interface routines) and some
FORTRAN preprocessors do not understand the /*...*/ syntax.
#endif
#if 0
The following macros define the supplementary FORTRAN arithmetic
datatypes beyond the standard INTEGER, REAL, and DOUBLEPRECISION --
ostensibly corresponding to 8-bit and 16-bit integers, respectively.
For example:
#define NF_INT1_T byte
#define NF_INT2_T integer*2
These are the types of the relevant arguments in the NF_*_INT1() and
NF_*_INT2() netCDF FORTRAN function calls. The word "ostensibly"
is used advisedly: on some systems an "integer*2" datatype,
nevertheless, occupies 64 bits (we are not making this up).
If your FORTRAN system does not have the respective supplementary
datatype, then do not define the corresponding macro.
#endif
#undef NF_INT1_T
#undef NF_INT2_T
#if 0
Define the following NF_*_IS_C_* macros appropriatly for your system.
The "INT1", "INT2" and "INT" after the "NF_" refer to the NF_INT1_T
FORTRAN datatype, the NF_INT2_T FORTRAN datatype, and the INTEGER
FORTRAN datatype, respectively. If the respective FORTRAN datatype
does not exist, then do not define the corresponding macro.
#endif
#undef NF_INT1_IS_C_SIGNED_CHAR
#undef NF_INT1_IS_C_SHORT
#undef NF_INT1_IS_C_INT
#undef NF_INT1_IS_C_LONG
#undef NF_INT2_IS_C_SHORT
#undef NF_INT2_IS_C_INT
#undef NF_INT2_IS_C_LONG
#undef NF_INT_IS_C_INT
#undef NF_INT_IS_C_LONG
#undef NF_REAL_IS_C_FLOAT
#undef NF_REAL_IS_C_DOUBLE
#undef NF_DOUBLEPRECISION_IS_C_DOUBLE
#undef NF_DOUBLEPRECISION_IS_C_FLOAT
#if 0
Whether the system uses something besides the IEEE floating-point
format to represent floating-point values.
#endif
#undef NO_IEEE_FLOAT
#if 0
END OF CUSTOMIZATION
#endif
#if 0
FORTRAN data types corresponding to netCDF version 2 "byte" and "short"
data types (e.g. INTEGER*1, INTEGER*2). See file "ftest.F" for usage.
#endif
#if !defined(NO_NETCDF_2)
# undef NCBYTE_T
# undef NCSHORT_T
#endif
#endif