Par2: Dynamically Dimensioned Arrays 2002-05-25 -- Randy Goodhew Don't dismiss Routines as useless.
They certainly have their place.
Here's an example that uses a Routine to
create Dynamically Dimensioned Arrays.
Normally the DIM() value must be a constant,
but in the context of a Routine, the DIM()
value can be a variable, e.g.:
----------------------
DatHash PROCEDURE(STRING password)! ,STRING
Dim LONG,AUTO ! Ushorts array dim
Sum USHORT,AUTO ! sum of Ushorts
SumBytes BYTE,DIM(2),OVER(Sum) ! parse bytes from Ushort
Result STRING(16),AUTO ! two char password
CODE
Dim = LEN(CLIP(password)) / 2 ! calc array dimensions
DO R_Calc
Result = '''<<' & SumBytes[1] & ',' & SumBytes[2] & '>'''
RETURN(Result) ! string like '<99,88>'
R_Calc ROUTINE
DATA ! Routine-level dynamic data
Ushorts USHORT,DIM(Dim),OVER(password) ! dynamic array
J LONG,AUTO ! loop counter
CODE
Sum = 0 ! init Sum
LOOP J = 1 TO Dim ! loop thru password as Ushorts
Sum += Ushorts[J] ! sum of Ushorts
END !loop
EXIT
------------------
BTW, this little algorithm will calculate the 16 bit Relative Password
Hash for a Clarion DAT file (if the password is known).
Note: Regardless of how large a password is for a Clarion DAT file,
they can be reduced to one of 64K possible two character clones.
Thus, the longer the password, the more coincidental dups there
will be, essentially one in every 64K of character combinations.
Routines are also useful in writing recursive algorithms, where
the Routine can call itself, i.e. classical QuickSort algorithm.
Printed May 4, 2024, 5:00 am This article has been viewed/printed 35117 times. |