[Scilab-users] overloading size for tlist

[Samuel Gougeon]

Hello Philippe,

In Scilab 5, size() is overloadable only for mlist(), not for tlist().
In Scilab 6, it is overloadable for foth.
So, to be compatible with both Scilab 5 and 6, you shall define your 
object as a mlist instead of tlist.

Best regards
Samuel

Le 23/07/2017 à 11:43, philippe a écrit :
> Hi to all,
>
> I've created a new type "mytype", represented by tlist, and another type
> "matrixmytype", for matrix of "mytype" object, represented by mlist .
> I've created the function %matrixmytype_size to overload the function
> "size" for "matrixmytype" objects (it works fine) . For compatibility
> reason I would like that size(x), where typeof(x)=='mytype', retrieve
> the value [1,1], but actually size(x) return the size of the tlist for x
>   an "mytype" .
>
>   Let's take a minimal example :
>
> //*****************************************************
> function x=mytype()
>      x=tlist(['mytype' 'tab' 'pol' 'str'],..
>              grand(2,3,'uin',1,3), ..          //x.tab
>              poly(0,'x'), ..                   //x.pol
>              'scilab')                         //x.str
> endfunction
>
> function [p,n]=%mytype_size(x)
>      [p,n]=size(x.tab)
> endfunction
> //******************************************************
>
> then in scilab console :
>
>
> //***********************************
> -->x=mytype()
>   x  =
>
>
>         x(1)
>
> !mytype  tab  pol  str  !
>
>         x(2)
>
>      1.    3.    1.
>      2.    1.    2.
>
>         x(3)
>
>      x
>
>         x(4)
>
>   scilab
>
>
> -->[p,n]=%mytype_size(x)  // this is the expected result for size(x)
>   n  =
>
>      3.
>   p  =
>
>      2.
>
> -->[p,n]=size(x)
>                !--error 39
> Nombre erroné d'arguments d'entrée.
>
>
> -->size(x)
>   ans  =
>
>      4.
>
> //****************************************************
>
>
> My goal is only to ensure overloading compatibility between mytype and
> matrixmytype  like for "double"  size(x) returns [1,1] if x is a scalar
> and not a matrix.
>
> Thank's for reading !
>
> Philippe
>
> _______________________________________________
> users mailing list
> users at lists.scilab.org
> http://lists.scilab.org/mailman/listinfo/users