STATIC - QB64Official/qb64 GitHub Wiki

The STATIC keyword is used in declaration statements to control where variables are stored.

Syntax

STATIC variableName[()] [AS dataType][, ...] STATIC [AS dataType] variableName[()][, ...]

Syntax

{SUB|FUNCTION} procedureName [(parameterList)] STATIC

Description

  • A STATIC list can be used in SUB and FUNCTION procedures to designate one or more variables to retain their values.
  • Variables and arrays with static storage will retain their values in between procedure calls. The values may also be used recursively.
    • variableName may include a type suffix or use AS to specify a type other than the default SINGLE type.
    • Arrays with static storage are declared by specifying empty parenthesis following the array name. See Arrays
  • STATIC can be used after the name of a SUB or FUNCTION in the procedure to force all variables to retain their values.
  • Recursive procedures may be required to be STATIC to avoid a Stack Overflow error. QB64 programs may just close.
  • $STATIC defined program arrays cannot be REDIM or use _PRESERVE.

Example(s)

*Example 1: Finding the binary bit settings from a 32 bit LONG register return using recursion.


INPUT "Enter a numerical value to see binary value: ", num&
PRINT Bin$(num&)

END

FUNCTION Bin$ (n&) STATIC 'comment out STATIC to see what happens!
  DIM p%, s$
  IF 2 ^ p% > n& THEN
    p% = 0
  ELSE
    IF n& AND 2 ^ p% THEN s$ = "1" + s$ ELSE s$ = "0" + s$
    IF n& > 2 ^ p% THEN
      p% = p% + 1
      s$ = Bin$(n&) 'recursive call to itself
    ELSE: p% = 0
    END IF
  END IF
  IF s$ = "" THEN Bin$ = "0" ELSE Bin$ = s$
END FUNCTION

Explanation: The FUNCTION above returns a STRING value representing the bits ON in an INTEGER value. The string can be printed to the screen to see what is happening in a port register. STATIC keeps the function from overloading the memory "Stack" and is normally REQUIRED when recursive calls are used in QBasic! QB64 procedures will close without warning or error!

Using a static array to cache factorials, speeding up repeated calculations:


PRINT Factorial(0)
PRINT Factorial(5)
PRINT Factorial(50)

FUNCTION Factorial# ( n AS DOUBLE )
    CONST maxNToCache = 50
    STATIC resultCache() AS DOUBLE
    STATIC firstCall AS INTEGER
    
    ' The lookup table is initially empty, so re-size it..
    IF firstCall = 0 THEN
        firstCall = -1
        REDIM resultCache(maxNToCache) AS DOUBLE
        
        ' ..and pre-calculate some factorials.
        resultCache(0) = 1
        resultCache(1) = 1
        resultCache(2) = 2
    END IF
    
    ' See if we have the result cached. If so, we're done.
    IF n <= maxNToCache THEN
        IF resultCache(n) <> 0 THEN
            Factorial = resultCache(n)
            EXIT FUNCTION
        END IF
    END IF
    
    ' If not, we use recursion to calculate the result, then cache it for later use:
    resultCache(n) = INT(n) * Factorial(INT(n) - 1)
    Factorial = resultCache(n)
END FUNCTION




 1
 120
 3.041409320171338D+64

See Also