feval_w_path

FEVAL_W_PATH  Calls a function located by the specified path.
   FEVAL_W_PATH(FPATH, F, x1, ..., xn)
   [y1, ..., yn] = FEVAL_W_PATH(FPATH, F, x1, ..., xn)

   Identical to the built-in FEVAL, except that the function F need not be
   in the MATLAB/Octave path if it is defined in a file in the path specified
   by FPATH. Assumes that the current working directory is always first in
   the MATLAB/Octave path.

   Inputs:
       FPATH - string containing the path to the function to be called,
               can be absolute or relative to current working directory
       F - string containing the name of the function to be called
       x1, ..., xn - variable number of input arguments to be passed to F

   Output:
       y1, ..., yn - variable number arguments returned by F (depending on
                     the caller)

   Note that any sub-functions located in the directory specified by FPATH
   will also be available to be called by the F function.

   Examples:
       % Assume '/opt/testfunctions' is NOT in the MATLAB/Octave path, but
       % /opt/testfunctions/mytestfcn.m defines the function mytestfcn()
       % which takes 2 input arguments and outputs 1 return argument.
       y = feval_w_path('/opt/testfunctions', 'mytestfcn', x1, x2);

0001 function varargout = feval_w_path(fpath, fname, varargin)
0002 %FEVAL_W_PATH  Calls a function located by the specified path.
0003 %   FEVAL_W_PATH(FPATH, F, x1, ..., xn)
0004 %   [y1, ..., yn] = FEVAL_W_PATH(FPATH, F, x1, ..., xn)
0005 %
0006 %   Identical to the built-in FEVAL, except that the function F need not be
0007 %   in the MATLAB/Octave path if it is defined in a file in the path specified
0008 %   by FPATH. Assumes that the current working directory is always first in
0009 %   the MATLAB/Octave path.
0010 %
0011 %   Inputs:
0012 %       FPATH - string containing the path to the function to be called,
0013 %               can be absolute or relative to current working directory
0014 %       F - string containing the name of the function to be called
0015 %       x1, ..., xn - variable number of input arguments to be passed to F
0016 %
0017 %   Output:
0018 %       y1, ..., yn - variable number arguments returned by F (depending on
0019 %                     the caller)
0020 %
0021 %   Note that any sub-functions located in the directory specified by FPATH
0022 %   will also be available to be called by the F function.
0023 %
0024 %   Examples:
0025 %       % Assume '/opt/testfunctions' is NOT in the MATLAB/Octave path, but
0026 %       % /opt/testfunctions/mytestfcn.m defines the function mytestfcn()
0027 %       % which takes 2 input arguments and outputs 1 return argument.
0028 %       y = feval_w_path('/opt/testfunctions', 'mytestfcn', x1, x2);
0029 
0030 %   MATPOWER
0031 %   Copyright (c) 2016, Power Systems Engineering Research Center (PSERC)
0032 %   by Ray Zimmerman, PSERC Cornell
0033 %
0034 %   This file is part of MATPOWER.
0035 %   Covered by the 3-clause BSD License (see LICENSE file for details).
0036 %   See https://matpower.org for more info.
0037 
0038 %% check input type
0039 if ~ischar(fpath)
0040     error('feval_w_path: FPATH must be a string');
0041 end
0042 if ~ischar(fname)
0043     error('feval_w_path: FNAME must be a string');
0044 end
0045 
0046 if isempty(fpath)   %% just call feval directly, no cd necessary
0047     [varargout{1:nargout}] = feval(fname, varargin{:});
0048 else
0049     %% see if path exists
0050     if exist(fpath, 'dir') ~= 7
0051         error('feval_w_path: Sorry, ''%s'' is not a valid directory path.', fpath);
0052     end
0053 
0054     cwd = pwd;      %% save the current working dir
0055     cd(fpath);      %% switch to the dir with the mfile
0056     try
0057         [varargout{1:nargout}] = feval(fname, varargin{:});
0058         cd(cwd);    %% switch back to saved dir
0059     catch
0060         cd(cwd);    %% switch back to saved dir
0061         rethrow(lasterror);
0062     end
0063 end