[Scilab-users] strange %h_set overloading

[Stéphane Mottelet]

Hello,

I have a question concerning the overloading function %h_set and 
concerning overloading mechanism in general.

Here is an example where the original %h_set is redefined in order to 
display "in %h_set" when it is called:

function hdl=%h_set(varargin)
     disp("in %h_set")
     if (size(varargin) == 3)
         hdl = varargin(1)
         str = varargin(2)
         val = varargin(3)
         for i=1:size(hdl,"*")
             set(hdl(i),str,val)
         end,
     elseif (size(varargin) == 2)
         str = varargin(1)
         val = varargin(2)
         hdl=[]
         set(str,val)
     end,
endfunction

set(gcf(),'tag','foo');
set([gcf(),gcf()],'tag','foo');

--> set(gcf(),'tag','foo');

--> set([gcf(),gcf()],'tag','foo');

  in %h_set

%h_set is not called wen a 1x1 handle is used, but only when a vector of 
more than 2 handles is used. I suppose this mechanism is implemented i 
the built-in "set" ?

My other question/remark is about the naming convention of built-in 
functions like "set". When macros do exist wit the same name but in 
different libraries,  the library name used as a prefix plays the role 
of a namespace and allows to disambiguate calls: e.g."graphicslib.plot" 
vs "plolib.plot". However, there is no namespace for built-in functions.

The reason for these questions is that I am currently re-engineering the 
plotlib and need a smooth mechanism to emulate new properties/fields for 
*native* Scilab handles.

S.

-- 
Stéphane Mottelet
Ingénieur de recherche
EA 4297 Transformations Intégrées de la Matière Renouvelable
Département Génie des Procédés Industriels
Sorbonne Universités - Université de Technologie de Compiègne
CS 60319, 60203 Compiègne cedex
Tel : +33(0)344234688
http://www.utc.fr/~mottelet