Topic: -={ Challenge }=- Nested Summing

[It's Alive!]

I tried a C# solution, but it's not very elegant 

A work of art... I see you caught the point2d -3d too 

[gile]

A work of art...

Thanks, but I think it's very exaggerated...

[Lee Mac]

Another way:

Superb Gile - thinking outside the box 

[LE3]

I will try with this old style school and mickey mouse approach and solution (and without looking into any of already code posted), compared to you masters of lisp 
... and don't laugh..... ok you can 

(setq lst '( 1 2 ( 3 4 ( 1 ))))
Command: tst
Total: 11

(setq lst '( 1 2 ( 3 4 ( 1 )) ( 3 4 ( 1 ))))
Command: tst
Total: 19

Code: [Select]

(defun sum (lst / lth cont)
(setq lth (length lst) cont 0)
(while (< cont lth)
  (setq item (nth cont lst))
  (cond
  ((eq (type item) 'list) (sum item))
  ((eq (type item) 'int) (setq total (+ item total))))
  (setq cont (1+ cont))))

(defun summer (/ lth cont total)
(setq lth (length lst) cont 0 total 0)
(while (< cont lth)
  (setq item (nth cont lst))
  (cond
  ((eq (type item) 'list) (sum item))
  ((eq (type item) 'int) (setq total (+ item total))))
  (setq cont (1+ cont))
  total))

(defun c:tst () (princ "\nTotal: ") (princ (summer)) (princ))

[Lee Mac]

Thanks for joining in the fun Luis 

Try with a dotted pair 

[LE3]

let's see...

(setq lst '( 1 2 ( 3 4 ( 1 ) (3 . 4))))
Command: tst
Total: 18

Code: [Select]

(defun isDottedPair (item) (and (car item) (cdr item) (atom (car item)) (atom (cdr item))))
   
(defun sum (lst / lth cont)
(setq lth (length lst) cont 0)
(while (< cont lth)
  (setq item (nth cont lst))
  (cond
  ((and (eq (type item) 'list) (isDottedPair item)) (setq total (+ (car item) (cdr item) total)))
  ((eq (type item) 'list) (sum item))
  ((eq (type item) 'int) (setq total (+ item total))))
  (setq cont (1+ cont))))

(defun summer (/ lth cont total)
(setq lth (length lst) cont 0 total 0)
(while (< cont lth)
  (setq item (nth cont lst))
  (cond
  ((and (eq (type item) 'list) (isDottedPair item)) (setq total (+ (car item) (cdr item) total)))
  ((eq (type item) 'list) (sum item))
  ((eq (type item) 'int) (setq total (+ item total))))
  (setq cont (1+ cont))
  total))


Thanks for joining in the fun Luis 

Try with a dotted pair 

[pkohut]

Lee,

I always liked this one from the Compuserve days:
-David

Ah, yes. 100 virtual points for anyone that can remember the moderator's name?

[LE3]

Hi Gile,
Just tried this, and does not return the right sum... ?

(1 2 3)
_$ (SUMLIST lst)
6 ;; ok

(1 2 (3 4 (1)))
_$ (SUMLIST lst)
6 ;; ?

Another way:

Code: [Select]

(defun sumlist (l)
  (eval
    (read
      (strcat "(+ "
      (vl-string-translate "(.)" "   " (vl-princ-to-string l1))
      ")"
      )
    )
  )
)

[gile]

Oops !!!

I forgot to change l1 for l after debugging...
Try with corrected code.

[LE3]

I'll check... Je sais que c'était quelque chose de simple

Oops !!!

I forgot to change l1 for l after debugging...
Try with corrected code.

[pkohut]

Slightly reworked version of D's. Call to acedGetArgs eliminated. pRbTmp != NULL eliminated. RTPOINT and RT3DPOINT reworked. Got rid of unneeded curly brackets in each of the case statements (not needed unless a local variable needs to be assigned defined at case scope).

Code: [Select]

static int ads_nsum(resbuf * pArgs)
{
    double ans = 0;
    for(resbuf *pRbTmp = pArgs; pRbTmp; pRbTmp=pRbTmp->rbnext)
    {
        switch(pRbTmp->restype)
        {
        case RTSHORT: ans+=pRbTmp->resval.rint; break;
        case RTLONG: ans+=pRbTmp->resval.rlong; break;
        case RTREAL: ans+=pRbTmp->resval.rreal; break;
        case RTPOINT: ans += pRbTmp->resval.rpoint[0] +
                          pRbTmp->resval.rpoint[1]; break;
        case RT3DPOINT: ans+=pRbTmp->resval.rpoint[0] +
                            pRbTmp->resval.rpoint[1] +
                            pRbTmp->resval.rpoint[2]; break;
        }
    }
    acedRetReal(ans);
    return (RSRSLT) ;
}


[David Bethel]

Lee,

I always liked this one from the Compuserve days:
-David

Ah, yes. 100 virtual points for anyone that can remember the moderator's name?

Anne Brown


She was into quilting or something.  Not a CAd person at all.

Some of the threadmatsers that I remember were

Jon Fleming
Bill Townsend
Tony Tanzillio ( for a while )
Steve Johhnson
Ditmar Rudolph

[pkohut]

Anne Brown

She was into quilting or something.  Not a CAd person at all.

Yep. You can collect your points at the CompuServe front desk. 

Some of the threadmatsers that I remember were

Jon Fleming
Bill Townsend
Tony Tanzillio ( for a while )
Steve Johhnson
Ditmar Rudolph

Wow, scored bonus on those. I'd forgotten.

[It's Alive!]

Slightly reworked version of D's. Call to acedGetArgs eliminated. pRbTmp != NULL eliminated. RTPOINT and RT3DPOINT reworked. Got rid of unneeded curly brackets in each of the case statements (not needed unless a local variable needs to be assigned defined at case scope).

Code: [Select]

static int ads_nsum(resbuf * pArgs)
{
    double ans = 0;
    for(resbuf *pRbTmp = pArgs; pRbTmp; pRbTmp=pRbTmp->rbnext)
    {
        switch(pRbTmp->restype)
        {
        case RTSHORT: ans+=pRbTmp->resval.rint; break;
        case RTLONG: ans+=pRbTmp->resval.rlong; break;
        case RTREAL: ans+=pRbTmp->resval.rreal; break;
        case RTPOINT: ans += pRbTmp->resval.rpoint[0] +
                          pRbTmp->resval.rpoint[1]; break;
        case RT3DPOINT: ans+=pRbTmp->resval.rpoint[0] +
                            pRbTmp->resval.rpoint[1] +
                            pRbTmp->resval.rpoint[2]; break;
        }
    }
    acedRetReal(ans);
    return (RSRSLT) ;
}


so your adding another function call for acedGetArgs() somewhere?

[pkohut]

so your adding another function call for acedGetArgs() somewhere?

No, check the signature of ads_nsum.