[Plplot-devel] integer and logical precision for our Fortran bindings and examples

[Alan W. I. <ir...@be...>]

On 2014-09-09 06:44-0000 Arjen Markus wrote:

> As for the kinds of integers and logicals, I suggest:
>
>
>
> integer, parameter :: plint = kind(0)   ! The default integer, which is in 99% of the cases a 4-bytes integer
> integer, parameter :: plbool = kind(.false.)

These are indefinite types, but note the C equivalents (which these
fortran types are supposed to match exactly) are exact precision in
include/plplot.h rather than indefinite precision:

typedef int32_t        PLINT;
typedef PLINT    PLBOOL;

So isn't

integer, parameter :: plint  = 4
integer, parameter :: plbool  = 4

the exactly correct thing to do instead of the indefinite types you
have suggested using above?

If you are convinced now by this "equivalent to what we do in C"
argument that the above two explicit definitions of plint=4 and
plbool=4 should go into bindings/f95/plflt.inc.in, then my plan is to
make that change, then sed-edit all our Fortran binding and examples
code to transform

integer ==> integer(kind=plint)
logical ==> logical(kind=plbool)

Alan
__________________________
Alan W. Irwin

Astronomical research affiliation with Department of Physics and Astronomy,
University of Victoria (astrowww.phys.uvic.ca).

Programming affiliations with the FreeEOS equation-of-state
implementation for stellar interiors (freeeos.sf.net); the Time
Ephemerides project (timeephem.sf.net); PLplot scientific plotting
software package (plplot.sf.net); the libLASi project
(unifont.org/lasi); the Loads of Linux Links project (loll.sf.net);
and the Linux Brochure Project (lbproject.sf.net).
__________________________

Linux-powered Science
__________________________