Re: [gforth] my vsum function is suspiciously non-local

[Carsten Strotmann (private)]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Hello Terence,

my solution would be:

: vsum ( addr u -- n )
  CELLS      ( addr u      ) \ adjust u to size of cell
  OVER       ( addr u addr )
  +          ( addr addr2  )
  SWAP       ( addr2 addr  )
  0          ( addr2 addr 0 )
  -ROT       ( 0 addr2 addr )
  DO         ( 0            )
    I        ( 0 i          ) \ i = index = address
    @        ( 0 u          ) \ u = value of cell i
    +        ( n            ) \ n = result
    1        ( n 1          )
    CELLS    ( n <c>        ) \ <c> = size of a cell in byte
                              \ this is architecture dep.
  +LOOP      ( n            )
;            ( n            )



condensed:

: vsum ( addr u -- n ) CELLS OVER + SWAP  0 -ROT  DO  I @ +  1 CELLS +LOOP ;

No local variables needed.

- -- Carsten

On 3/7/10 6:53 PM, Terrence Brannon wrote:
> I used a "global variable" named tally to handle this tutorial
> assignment, but I think there must be a way to avoid the use of such a
> variable in solving this problem. Any help is appreciated:
> 
> \    Assignment: Write a definition vsum ( addr u -- n ) that computes
> the sum of u cells, with the first of these cells at addr, the next
> one at addr cell+ etc.
> 
> ( Usage:
>      create vx 2 , 2 , 2 , 2 , 2 ,
>      vx 5 vsum
> )
> 
> : loop-range ( number-of-cells -- number-of-cells zero )  0 ;
> : get-value-at-address-offset ( address offset -- v ) cells + @ ;
> : increase-value ( n addr -- v+n ) TUCK @ + SWAP ! ;
> VARIABLE tally
> : vsum { address u }
>     u loop-range U+DO
>       address i get-value-at-address-offset
>       tally increase-value
>     LOOP
>     tally @ ;
> 
> 

-----BEGIN PGP SIGNATURE-----
Version: GnuPG/MacGPG2 v2.0.14 (Darwin)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/

iEYEARECAAYFAkuT7lkACgkQsUJ3c+pomYFG4ACg48Bqt/EqPOymMou0WpfFjRCz
SEgAoNOrgqzuqLBgcvCy0dCC1FvL7L/X
=pNtG
-----END PGP SIGNATURE-----