lib(fd_sets)


    This is a solver for constraints over the domain of finite integer sets.
    
    (Ground) integer sets are represented simply as sorted, duplicate-free
    lists of integers e.g.
    
    	SetOfThree = [1,3,7]
    	EmptySet = []
    

Set Variables
    Set variables are variables which can eventually take a ground integer
    set as their value. They are characterized by a lower bound (the set 
    of elements that are definitely in the set) and an upper bound (the
    set of elements that may be in the set). A set variable can be declared
    as follows:
    
    	SetVar in_set_range []..[1,2,3,4,5,6,7],
    
    Since the lower bound is the empty set, this can be written as
    
    	SetVar subset [1,2,3,4,5,6,7],
    
    If the lower bound is the empty set and the upper bound is a set
    of consecutive integers, you can also write
    
    	intset(SetVar, 1, 7)
    

Set Constraints
    Most of the usual set operations/relations are provided as constraints:
    
    membership
    non-membership
    inclusion (subset)
    equality
    intersection
    union
    difference
    symmetric difference
    disjointness
    cardinality
    
    as well as a constraint on set weight.  Note that there is no
    complement-constraint because the library has no concept of a set
    universe and cannot represent infinite sets.
    
    On most argument positions where sets are expected, set expressions
    are allowed, e.g.
    
    Set1 /\ Set2       % intersection
    Set1 \/ Set2       % union
    Set1 \ Set2        % difference
    
    as well as references to array elements like Set[3].
    

Search
    The insetdomain/4 predicate can be used to enumerate all ground
    instantiations of a set variable, much like indomain/1 in the
    finite-domain case.

Cooperation with a finite domain solver
    This library comes in two flavours: lib(fd_sets) which cooperates with
    lib(fd), and lib(ic_sets) which cooperates with lib(ic). This is relevant
    only for those constraints which involve integer variables, e.g. the
    cardinality argument in #/2, the weight argument in weight/3 and the
    booleans in membership_booleans/2. These will be represented as fd-
    or ic-variables respectively.


