Fixed 0AB1 input argument count verification for string arrays.

[MiranDMC]

Enabled using string array argument as module name in 0AB1.

[x87]

Why IsString includes variables and IsInteger does not? IsVariable should be part of IsInteger

[MiranDMC]

Because there is no DT_VAR_INT.
DT_VAR actually can contain multiple types.

[x87]

Maybe if you call it IsNumber then it would better reflect what it does

[MiranDMC]

How should it be called then? DT_VAR can in fact be string too (pointer to text).

[x87]

Make 2 functions: IsNumber and IsString.

If IsNumber returns true, it's a label, one path. If IsString, it's a module, another path. Otherwise, error

[MiranDMC]

These functions reflect type to text function

static const char* ToKindStr(eDataType type)
{
	switch (type)
	{
		case DT_BYTE:
		case DT_WORD:
		case DT_DWORD:
			return "int"; break;

		case DT_FLOAT:
			return "float"; break;

		case DT_STRING:
		case DT_TEXTLABEL:
		case DT_LVAR_TEXTLABEL:
		case DT_LVAR_TEXTLABEL_ARRAY:
		case DT_LVAR_STRING:
		case DT_LVAR_STRING_ARRAY:
		case DT_VAR_TEXTLABEL:
		case DT_VAR_TEXTLABEL_ARRAY:
		case DT_VAR_STRING:
		case DT_VAR_STRING_ARRAY:
		case DT_VARLEN_STRING:
			return "string"; break;

		case DT_VAR:
		case DT_VAR_ARRAY:
		case DT_LVAR:
		case DT_LVAR_ARRAY:
			return "variable"; break;

		case DT_END:
			return "varArgEnd"; break;

		default:
			return "corrupted"; break;
	}
}

Motivation was to just make simpler verification of DATA TYPE byte without need for switches, where as already proved few times, it is to easy to forget to include some cases.

These utilities are meant to verify eDataType sub-category only, in abstraction of actual in code use (as mentioned it would be possible to deduce int-float value type within arrays).

It would perhaps make more sense if it was DataType::IsInteger(type), but that would require C++ features set.

[x87]

I still don't get why we separate int literal and int variable while combine string literal and string variable. should be some consistency at least. we either combine literals and variables (for int, float, and string types) under the same check, or separate them out (for all three types).

In some cases, like mentioned 0AB1's arguments count which is compile constant too, so makes no sense to be variable.

there is nothing that prevents you to use a variable in this position, though.

[MiranDMC]

As mentioned because there is DT_LVAR_STRING but no DT_LVAR_INT or DT_LVAR_FLOAT. I just organized DT_x types together in groups of same flavour.

[MiranDMC]

I'm splitting then the functions to separate immediate and variable values.

[x87]

Agreed!