The unary operator (.byte) is the byte coersion operator. It instructs the XCASM compiler to use 8 bit quantities during computation of its operand (i.e the tagged part of the expression).
Consider the following example (where D, G and H are defined as word variables)
The expression:D = G + Hwould generate0006 08 1F movf G+1,w 0007 00 9D movwf D+1 0008 08 1E movf G,w 0009 07 20 addwf H,w 000A 00 9C movwf D 000B 18 03 btfsc STATUS,C 000C 0A 9D incf D+1 000D 08 21 movf H+1,w 000E 07 9D addwf D+1Here you can see that the computation is carried out using 16 bit addition.
However if instead we used:D = (.byte)(G + H)this would generate000F 08 1E movf G,w 0010 07 20 addwf H,w 0011 01 9D clrf D+1 0012 00 9C movwf DHere you can see that the computation is carried out using 8 bit addition even though the quantities involved were 16 bits. This is because we told the compiler to do so by using the (.byte) operator.
To be exact, what the (.byte) operator did was to tell the compiler that the sub-expression (G + H) should yield an 8 bit value, and the compiler produced only sufficent code to produce an 8 bit result. Any code necessary to extend the 8 bit result to a 16 bit result was discarded.
Note how using the (.byte) operator is far more efficient than masking out the most significant 8 bits. e.g.D = (G + H) & 0xffwould generate:0013 08 1F movf G+1,w 0014 00 9D movwf D+1 0015 08 1E movf G,w 0016 07 20 addwf H,w 0017 00 9C movwf D 0018 18 03 btfsc STATUS,C 0019 0A 9D incf D+1 001A 08 21 movf H+1,w 001B 07 9D addwf D+1 001C 01 9D clrf D+1