TwinCAT 3 using Union in function argument-CodePudding

I made union, which allow me to use same data as REAL or 4 bytes (Module in profibus device has 4 BYTE registers to write REAL floating point type value).

The union:
TYPE U_4Bytes2Real :
UNION
    abDataBytes     :       ARRAY[0..3] OF BYTE;
    rDataFloat      :       REAL;
END_UNION
END_TYPE

When I want to get access to this variable like REAL, I write:

U_4Bytes2Real.rDataFloat

When I want to get access to this variable like 4 BYTE ARRAY, I write:

U_4Bytes2Real.abDataBytes

I want to have a function, which gets REAL value, and inside it, I want to write it to the registers as an ARRAY of BYTES.

How to tell my function, that argument is REAL?

I' using function like that:

bFunResult := F_SetMod22(bDataGroup := 3, bChannel := 3, bDataFloat := 20.0, nTimeout := 100);

and i get error

Cannot convert type 'LREAL' to type 'U_4Bytes2Real'

Do I have to convert it INTO function, or there is some method to use union in function argument?

CodePudding user response：

Answer

UNION is a STRUCT with a shared dataspace

This means anything that you can do with a STRUCT, you can do with a UNION. Whether that is use for inputs/outputs to a method, function, function block.

VAR
   Data : U_4Bytes2Real;
END_VAR

 F_Function( RealValue := Data.rDataFloat );

Example

Very simplest program to demonstrate this principle. It doesn't matter whether you are using a UNION or a STRUCT, they are accessed exactly the same.

TYPE st_Struct :
STRUCT
    Value                       : REAL;
    Packed                      : ARRAY [ 0..3 ] OF BYTE;    
END_STRUCT
END_TYPE

TYPE u_Union :
UNION
    Value                       : REAL;
    Packed                      : ARRAY[0..3] OF BYTE;
END_UNION
END_TYPE

FUNCTION f_Sum : REAL
VAR_INPUT
    A, B                        : REAL;
END_VAR 

f_Sum := A   B;

PROGRAM MAIN
VAR
    Str                         : st_Struct;
    Uni                         : u_Union;
    Result                      : REAL;
END_VAR

Result := f_Sum( Str.Value, Uni.Value );

CodePudding user response：

For this answer I assume from your incomplete information:

Declaration assumption:

VAR
bFunResult : U_4Bytes2Real;


METHOD F_SetMod22 : LREAL

Then you can write:

bFunResult.rDataFloat = TO_REAL(F_SetMod22(...));

And later use:

bFunResult.abDataBytes

BUT !!!:
Be aware, that TO_REAL will cut off information if the returned LREAL does not fit into a REAL.