octave-maintainers
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: MinGW


From: John W. Eaton
Subject: Re: MinGW
Date: Wed, 26 Apr 2006 12:45:35 -0400

On 25-Apr-2006, David Bateman wrote:

| >|   DEFCONST (sepchar, SEPCHAR_STR,
| >|     "-*- texinfo -*-\n\
| >| @defvr {Built-in Variable} sepchar\n\
| >| The character used to separate directories in the path. The\n\
| >| value of this variable is system dependent.\n\
| >| @seealso{filesep,dir, ls}\n\
| >| @end defvr");
| >| 
| >| There are several problem functions for sepchar, including path.m,
| >| addpath.m and rmpath.m  that hardcodes the ":" as sepchar, while it
| >| should use sepchar.
| >
| >I'm trying to remove DEFCONST/DEFVAR so this should probably be a
| >function.
| >  
| >
| Ok, with me, the change to the above is minor

Will sepchar be used all that much?  Perhaps we should simply make it
available in octave_config_info and then provide a function like the
following (renamed to pathsep for compatibility).

  ## -*- texinfo -*-
  ## @deftypefn {Function} {} pathsep ()
  ## The character used to separate directories in the path. The
  ## value of this variable is system dependent.
  ## @seealso{filesep, dir, ls}
  ## @end deftypefn

  function c = pathsep ()
    c = octave_config_info ("pathsep");
  endfunction

For consistency, we should probably also change the name in the
configure script and makefiles to be pathsep.

Unless there are objections, I can add this function and fix
octave_config_info, configure, Makeconf, etc.

jwe


reply via email to

[Prev in Thread] Current Thread [Next in Thread]