IFC4X4 issues

[iegorychev]

Hello, found following:

1. Double semicolon appears in current IFC4X4 schema sometimes in such places:
   Values : ARRAY [1:SELF\IfcVoxelData.GridSize] OF IfcInteger := IfcListToExpandedArray(ValueData,1,SELF\IfcVoxelData.GridSize,SELF\IfcProduct.Representation.Representations[1].Items[1]\IfcVoxelGrid.Voxels);;
   This looks incorrect from the point of view of grammar.

2. Voxels : ARRAY [1:NumberOfVoxelsXNumberOfVoxelsYNumberOfVoxelsZ] OF IfcBoolean;
   Again, NumberOfVoxelsX (or NumberOfVoxelsY and NumberOfVoxelsZ) are unset, so ARRAY can not be instantiated as upper bound is indeterminate.
   Looks like LIST or combination of LIST+ARRAY.

3. CorrectGridSize : SIZEOF(SELF\Voxels) = NumberOfVoxelsX * NVL(NumberOfVoxelsY, NumberOfVoxelsX) * NVL(NumberOfVoxelsZ, NumberOfVoxelsX);

SIZEOF(SELF\Voxels) - doesn't look as qualified_attribute (SELF\entity_ref.attribute_ref), seems should be just SIZEOF(Voxels) as this isn't redeclared (as in other places within schemas).

[SergejMuhic]

To 1, that semicolon slipped through because I tested the expression only and IfcDoc when generating the schema automatically ads it. Fixes here.
To 2, this is allowed according to grammar. Checked and validated with an author of EXPRESS:

there will be an update to this part though, if you mean the optional version of "NumberOfVoxelsY and NumberOfVoxelsZ". NumberOfVoxelsX is set since it is not optional. Could go either way here, calling NVL or making them mandatory. I lean towards the latter i. e. making them mandatory.
To 3, Indeed, have we not had this issue already? Apparently with the merges it got reverted.

Thanks. Will correct and commit. But please wait for the merges as two PRs are still open. After I merge, I will also run the intepreter.