This is part 1 of this subject. Part 2 will be about writing values to the arrays.

One of the cool things with the latest version of Cube Voyager is multi-dimensional arrays. However, it appears behind the scenes (or at least to C++) that the multi-dimensional arrays are a wrapper over a single-dimension array.

The easiest way to show this is to make a random array and send it to the print file. Making the random array in Cube is simple:

RUN PGM=MATRIX PRNFILE="C:\TEMP\DTMAT00B.PRN"
FILEO PRINTO[1] = "C:\TEMP\DEBUG.PRN"

PAR ZONES=1

ARRAY MDARRAY=5,5

LOOP _c=1,5
  LOOP _r=1,5
    MDARRAY[_c][_r]=RAND()
    PRINT PRINTO=1 LIST='MDARRAY[',_c(1.0),'][',_r(1.0),']=',MDARRAY[_c][_r](8.6)
  ENDLOOP
ENDLOOP

CALL DLL=DLLFILE(TableReader)
ENDRUN

Then, in C++ we can pull 25 (!) array values from this:

int TableReader (Callstack* Stack){
	double* TableRecord;
	char message[100];

	TableRecord=(double*)Stack->pfFindVar("MDARRAY",0,1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,
             16,17,18,19,20,21,22,23,24);
	for(int x=0;x<=24;x++){
	if(&TableRecord!=0){
		sprintf(message,"TableRecord=%f",TableRecord[x]);
		Stack->pfPrnLine(1,message);
		}
	}
	return 0;
}

For fixed size multi-dimensional arrays, this isn’t really an issue. It would be very easy to wrap the Stack->pfFindVar line in a set of loops that fills a multi-dimensional array.