I received a question:

 

…I've always been curious about code like this:

IF wParam=ASC("a")
    wParam=ASC("b")
ENDIF

Is it more appropriate to think of ASC("a") as the Fox equivalent of a C 'a', or is it an actual function call every time through?

 

It’s easy to test this.

Try running this code:

 

 

num=1e7     && 1e7 is 10 * 1e6 (3,6,9 are thousand, million, billion)  = 10 * 1 million

wParam=ASC("a")

ns=SECONDS()

FOR i  = 1 TO num

      IF wParam=ASC("a")

            wParam=ASC("b")

      ENDIF

ENDFOR

 

nd1=SECONDS()-ns

?nd1

 

ns=SECONDS()

FOR i  = 1 TO num

      IF wParam=97

            wParam=98

      ENDIF

ENDFOR

 

nd2=SECONDS()-ns

?nd2,100*(nd1-nd2)/nd2

 

 

 

On my machine, the 2nd loop is 20-25% faster.

 

Of course an optimizing compiler could transform the 1st loop to the 2nd. In fact, a really good optimizing compiler would remove the loop completely!

 

In C and C++, the literal ‘a’ is exactly equal to 97, even at runtime, whereas the literal “a” (notice the single vs double quotes) is a string represented as a 2 byte array of 97 followed by a terminating 0.

 

The ASC() function takes any string expression as a parameter, including possibly a function call, so it cannot be optimized out in general.