🚀 go-pugleaf

/*****************************************************************************/
/* finding the minimum and the maximum from a list of numbers */
/* call list_min_max(+ListofNumbers,-TheMinimum,-TheMaximum)  */

list_min_max([Number|RL],Min,Max):-           /* if look at the first      */
     find_min_max(RL,Number,Number,Min,Max).  /* number, it is min and max */

/* (1a) if the list is empty you have the solutions */
find_min_max([],Min,Max,Min,Max).

/* (1b) you must search each number in the list (loop) */
find_min_max([Number|RL],OldMin,OldMax,Min,Max):-
     is_min_or_max(Number,OldMin,OldMax,NewMin,NewMax), /* decision */
     find_min_max(RL,NewMin,NewMax,Min,Max).

/* (2) decision */
/* is the number a new minimum or new maximum? */
is_min_or_max(Number,Min,Max,Number,Max):- Number < Min, !. /* number is min */
is_min_or_max(Number,Min,Max,Min,Number):- Number > Max, !. /* number is max */
is_min_or_max(Number,Min,Max,Min,Max).                      /* nothing       */

/*****************************************************************************/
Benefits of this implementation:

 .) declarative approach
 .) same structure as member/2  (ease to understand)
 .) tailrecursive
 .) 'search-time' increases linear with length of list O(N)
 .) runs in all Prolog implementations as fast as possible with
    as few space (memory) as needed

Martin