%---------------------------------------------------------------------------% % vim: ft=mercury ts=4 sw=4 et %---------------------------------------------------------------------------% % Copyright (C) 2006-2009 The University of Melbourne. % Copyright (C) 2013-2016 Opturion Pty Ltd. % Copyright (C) 2017-2019, 2022 The Mercury team. % This file is distributed under the terms specified in COPYING.LIB. %---------------------------------------------------------------------------% % % File: ranges.m. % Authors: Mark Brown. % Stability: medium. % % This module defines the ranges abstract type. % %---------------------------------------------------------------------------% :- module ranges. :- interface. :- import_module list. :- import_module set. %---------------------------------------------------------------------------% % Range lists represent sets of integers. Each contiguous block % of integers in the set is stored as a range which specifies % the bounds of the block, and these ranges are kept in a list-like % structure. % :- type ranges. % empty returns the empty set. % :- func empty = ranges. % is_empty(Set): % Succeeds iff Set is the empty set. % :- pred is_empty(ranges::in) is semidet. % is_non_empty(Set): % Succeeds iff Set is not the empty set. % :- pred is_non_empty(ranges::in) is semidet. % universe returns the largest set that can be handled by this module. % This is the set of integers (min_int+1)..max_int. Note that min_int % cannot be represented in any set. % :- func universe = ranges. % range(Min, Max) is the set of all integers from Min to Max inclusive. % :- func range(int, int) = ranges. % split(D, L, H, Rest) is true iff L..H is the first range % in D, and Rest is the domain D with this range removed. % :- pred split(ranges::in, int::out, int::out, ranges::out) is semidet. % is_contiguous(R, L, H) <=> split(R, L, H, empty): % Test if the set is a contiguous set of integers and return the endpoints % of this set if this is the case. % :- pred is_contiguous(ranges::in, int::out, int::out) is semidet. % Add an integer to the set. % :- func insert(int, ranges) = ranges. :- pred insert(int::in, ranges::in, ranges::out) is det. % Delete an integer from the set. % :- func delete(int, ranges) = ranges. % Return the number of distinct integers which are in the ranges % (as opposed to the number of ranges). % :- func size(ranges) = int. % Returns the median value of the set. In case of a tie, returns % the lower of the two options. % :- func median(ranges) = int. % least(A, N) is true iff N is the least element of A. % :- pred least(ranges::in, int::out) is semidet. % greatest(A, N) is true iff N is the greatest element of A. % :- pred greatest(ranges::in, int::out) is semidet. % next(A, N0, N) is true iff N is the least element of A greater % than N0. % :- pred next(ranges::in, int::in, int::out) is semidet. % Test set membership. % :- pred member(int::in, ranges::in) is semidet. % Nondeterministically produce each range. % :- pred range_member(int::out, int::out, ranges::in) is nondet. % Nondeterministically produce each element. % :- pred nondet_member(int::out, ranges::in) is nondet. % subset(A, B) is true iff every value in A is in B. % :- pred subset(ranges::in, ranges::in) is semidet. % disjoint(A, B) is true iff A and B have no values in common. % :- pred disjoint(ranges::in, ranges::in) is semidet. % union(A, B) contains all the integers in either A or B. % :- func union(ranges, ranges) = ranges. % intersection(A, B) contains all the integers in both A and B. % :- func intersection(ranges, ranges) = ranges. % difference(A, B) contains all the integers which are in A but % not in B. % :- func difference(ranges, ranges) = ranges. % restrict_min(Min, A) contains all integers in A which are greater % than or equal to Min. % :- func restrict_min(int, ranges) = ranges. % restrict_max(Max, A) contains all integers in A which are less than % or equal to Max. % :- func restrict_max(int, ranges) = ranges. % restrict_range(Min, Max, A) contains all integers I in A which % satisfy Min =< I =< Max. % :- func restrict_range(int, int, ranges) = ranges. % prune_to_next_non_member(A0, A, N0, N): % % N is the smallest integer larger than or equal to N0 which is not % in A0. A is the set A0 restricted to values greater than N. % :- pred prune_to_next_non_member(ranges::in, ranges::out, int::in, int::out) is det. % prune_to_prev_non_member(A0, A, N0, N): % % N is the largest integer smaller than or equal to N0 which is not % in A0. A is the set A0 restricted to values less than N. % :- pred prune_to_prev_non_member(ranges::in, ranges::out, int::in, int::out) is det. % Negate all numbers: A in R <=> -A in negate(R) % :- func negate(ranges) = ranges. % The sum of two ranges. % :- func plus(ranges, ranges) = ranges. % Shift a range by const C. % :- func shift(ranges, int) = ranges. % Dilate a range by const C. % :- func dilation(ranges, int) = ranges. % Contract a range by const C. % :- func contraction(ranges, int) = ranges. %---------------------------------------------------------------------------% % Convert to a sorted list of integers. % :- func to_sorted_list(ranges) = list(int). % Convert from a list of integers. % :- func from_list(list(int)) = ranges. % Convert from a set of integers. % :- func from_set(set(int)) = ranges. %---------------------------------------------------------------------------% % Compare the sets of integers given by the two ranges using lexicographic % ordering on the sorted set form. % :- pred compare_lex(comparison_result::uo, ranges::in, ranges::in) is det. %---------------------------------------------------------------------------% :- pred foldl(pred(int, A, A), ranges, A, A). :- mode foldl(pred(in, in, out) is det, in, in, out) is det. :- mode foldl(pred(in, mdi, muo) is det, in, mdi, muo) is det. :- mode foldl(pred(in, di, uo) is det, in, di, uo) is det. :- mode foldl(pred(in, in, out) is semidet, in, in, out) is semidet. :- mode foldl(pred(in, mdi, muo) is semidet, in, mdi, muo) is semidet. :- mode foldl(pred(in, di, uo) is semidet, in, di, uo) is semidet. :- pred foldl2(pred(int, A, A, B, B), ranges, A, A, B, B). :- mode foldl2(pred(in, in, out, in, out) is det, in, in, out, in, out) is det. :- mode foldl2(pred(in, in, out, mdi, muo) is det, in, in, out, mdi, muo) is det. :- mode foldl2(pred(in, in, out, di, uo) is det, in, in, out, di, uo) is det. :- mode foldl2(pred(in, in, out, in, out) is semidet, in, in, out, in, out) is semidet. :- mode foldl2(pred(in, in, out, mdi, muo) is semidet, in, in, out, mdi, muo) is semidet. :- mode foldl2(pred(in, in, out, di, uo) is semidet, in, in, out, di, uo) is semidet. :- pred foldl3(pred(int, A, A, B, B, C, C), ranges, A, A, B, B, C, C). :- mode foldl3(pred(in, in, out, in, out, in, out) is det, in, in, out, in, out, in, out) is det. :- mode foldl3(pred(in, in, out, in, out, mdi, muo) is det, in, in, out, in, out, mdi, muo) is det. :- mode foldl3(pred(in, in, out, in, out, di, uo) is det, in, in, out, in, out, di, uo) is det. :- mode foldl3(pred(in, in, out, in, out, di, uo) is semidet, in, in, out, in, out, di, uo) is semidet. :- pred foldr(pred(int, A, A), ranges, A, A). :- mode foldr(pred(in, in, out) is det, in, in, out) is det. :- mode foldr(pred(in, mdi, muo) is det, in, mdi, muo) is det. :- mode foldr(pred(in, di, uo) is det, in, di, uo) is det. :- mode foldr(pred(in, in, out) is semidet, in, in, out) is semidet. :- mode foldr(pred(in, mdi, muo) is semidet, in, mdi, muo) is semidet. :- mode foldr(pred(in, di, uo) is semidet, in, di, uo) is semidet. %---------------------------------------------------------------------------% % For each range, call the predicate, passing it the lower and % upper bound and threading through an accumulator. % :- pred range_foldl(pred(int, int, A, A), ranges, A, A). :- mode range_foldl(pred(in, in, in, out) is det, in, in, out) is det. :- mode range_foldl(pred(in, in, mdi, muo) is det, in, mdi, muo) is det. :- mode range_foldl(pred(in, in, di, uo) is det, in, di, uo) is det. :- mode range_foldl(pred(in, in, in, out) is semidet, in, in, out) is semidet. :- mode range_foldl(pred(in, in, mdi, muo) is semidet, in, mdi, muo) is semidet. :- mode range_foldl(pred(in, in, di, uo) is semidet, in, di, uo) is semidet. % As above, but with two accumulators. % :- pred range_foldl2(pred(int, int, A, A, B, B), ranges, A, A, B, B). :- mode range_foldl2(pred(in, in, in, out, in, out) is det, in, in, out, in, out) is det. :- mode range_foldl2(pred(in, in, in, out, mdi, muo) is det, in, in, out, mdi, muo) is det. :- mode range_foldl2(pred(in, in, in, out, di, uo) is det, in, in, out, di, uo) is det. :- mode range_foldl2(pred(in, in, in, out, in, out) is semidet, in, in, out, in, out) is semidet. :- mode range_foldl2(pred(in, in, in, out, mdi, muo) is semidet, in, in, out, mdi, muo) is semidet. :- mode range_foldl2(pred(in, in, in, out, di, uo) is semidet, in, in, out, di, uo) is semidet. :- pred range_foldr(pred(int, int, A, A), ranges, A, A). :- mode range_foldr(pred(in, in, in, out) is det, in, in, out) is det. :- mode range_foldr(pred(in, in, mdi, muo) is det, in, mdi, muo) is det. :- mode range_foldr(pred(in, in, di, uo) is det, in, di, uo) is det. :- mode range_foldr(pred(in, in, in, out) is semidet, in, in, out) is semidet. :- mode range_foldr(pred(in, in, mdi, muo) is semidet, in, mdi, muo) is semidet. :- mode range_foldr(pred(in, in, di, uo) is semidet, in, di, uo) is semidet. %---------------------------------------------------------------------------% % % C interface to ranges. % % This section describes the C interface to the ranges/0 type that is exported % by this module. % % In C the ranges/0 type is represented by the ML_Ranges type. % The following operations are exported and may be called from C or C++ % code. % % ML_Ranges ML_ranges_empty(void); % Return the empty set. % % ML_Ranges ML_ranges_universe(void); % Return the set of integers from (min_int+1)..max_int. % % ML_Ranges ML_ranges_range(MR_Integer l, MR_Integer h); % Return the set of integers from `l' to `h' inclusive. % % int ML_ranges_is_empty(ML_Ranges r); % Return true iff `r` is the empty set. % % MR_Integer ML_ranges_size(ML_Ranges r); % Return the number of distinct integers in `r'. % % int ML_ranges_split(ML_Ranges d, MR_Integer *l, MR_Integer *h, % ML_Ranges *rest); % Return true if `d' is not the empty set, setting `l' and `h' to the % lower and upper bound of the first range in `d', and setting `rest' % to `d' with the first range removed. % Return false if `d' is the empty set. % % ML_Ranges ML_ranges_insert(MR_Integer i, ML_ranges r); % Return the ranges value that is the result of inserting the integer % `i' into the ranges value `r'. %---------------------------------------------------------------------------% % % Java interface to ranges. % % This section describes the Java interface to the ranges/0 type that is % exported by this module. % % In Java the ranges/0 type is represented by the ranges.Ranges_0 class. % The following operations are exported as public static methods of the ranges % module and may be called from Java code. % % ranges.Ranges_0 empty(); % Return the empty set. % % ranges.Ranges_0 universe(); % Return the set of integers from (min_int+1)..max_int. % % ranges.Ranges_0 range(int l, int, h); % Return the set of integers from `l' to `h' inclusive. % % boolean is_empty(ranges.Ranges_0 r); % Return true iff `r' is the empty set. % % int size(ranges.Ranges_0 r); % Return the number of distinct integers in `r'. % % boolean split(ranges.Ranges_0 d, % jmercury.runtime.Ref l, % jmercury.runtime.Ref h, % jmercury.runtime.Ref rest); % Return true if `d' is not the empty set, setting `l' and `h' to the % lower and upper bound of the first range in `d', and setting `rest' % to `d' with the first range removed. % Return false if `d' is the empty set. % % ranges.Ranges_0 insert(int i, ranges.Ranges_0 r); % Return the ranges value that is the result of inserting the integer % `i' into the ranges value `r'. %---------------------------------------------------------------------------% %---------------------------------------------------------------------------% :- implementation. :- import_module int. :- import_module require. %---------------------------------------------------------------------------% % Values of this type represent finite sets of integers. % They are interpreted in the following way. % % S[[ nil ]] = {} % S[[ range(L, H, Rest) ]] = {N | L < N =< H} \/ S[[ Rest ]] % % For example, `range(1, 4, nil)' is interpreted as {2, 3, 4}. % % The invariants on this type are: % % 1) Each range must be non-empty (i.e., L < H). % 2) The ranges must not overlap or abut (i.e. for any % value `range(_, H1, range(L2, _, _)' we must have H1 < L2). % 3) The ranges must be in sorted order. % % These invariants ensure that the representation is canonical. % % Note that it is not possible to represent a set containing min_int. % Attempting to create such a set will result in an exception being thrown. % :- type ranges ---> nil ; range(int, int, ranges). %---------------------------------------------------------------------------% :- pragma foreign_decl("C", " typedef MR_Word ML_Ranges; "). :- pragma foreign_export("C", ranges.empty = out, "ML_ranges_empty"). :- pragma foreign_export("Java", ranges.empty = out, "empty"). empty = nil. :- pragma foreign_export("C", ranges.is_empty(in), "ML_ranges_is_empty"). :- pragma foreign_export("Java", ranges.is_empty(in), "is_empty"). is_empty(nil). is_non_empty(range(_, _, _)). :- pragma foreign_export("C", universe = out, "ML_ranges_universe"). :- pragma foreign_export("Java", universe = out, "universe"). universe = range(min_int, max_int, nil). :- pragma foreign_export("C", range(in, in) = out, "ML_ranges_range"). :- pragma foreign_export("Java", range(in, in) = out, "range"). range(Min, Max) = Ranges :- ( if Min = min_int then error($pred, "cannot represent min_int") else if Min > Max then Ranges = nil else Ranges = range(Min - 1, Max, nil) ). :- pragma foreign_export("C", ranges.split(in, out, out, out), "ML_ranges_split"). :- pragma foreign_export("Java", ranges.split(in, out, out, out), "split"). split(range(Min1, Max, Rest), Min1 + 1, Max, Rest). is_contiguous(Range, Min + 1, Max) :- Range = range(Min, Max, nil). :- pragma foreign_export("C", ranges.insert(in, in) = out, "ML_ranges_insert"). :- pragma foreign_export("Java", ranges.insert(in, in) = out, "insert"). insert(N, As) = union(As, range(N, N)). insert(N, As, Bs) :- Bs = insert(N, As). delete(N, As) = difference(As, range(N, N)). %---------------------------------------------------------------------------% :- pragma foreign_export("C", size(in) = out, "ML_ranges_size"). :- pragma foreign_export("Java", size(in) = out, "size"). size(nil) = 0. size(range(L, U, Rest)) = (U - L) + size(Rest). median(As) = N :- Size = size(As), ( if Size > 0 then MiddleIndex = (Size + 1) / 2 else error($pred, "empty set") ), N = element_index(As, MiddleIndex). % element_index(Intervals, I) returns the I'th largest value in the set % represented by Intervals (the least item in the set having index 1). % :- func element_index(ranges, int) = int. element_index(nil, _) = func_error($pred, "index out of range"). element_index(range(L, U, Rest), I) = N :- N0 = L + I, ( if N0 =< U then N = N0 else N = element_index(Rest, N0 - U) ). %---------------------------------------------------------------------------% least(range(L, _, _), L + 1). greatest(range(_, U0, As), U) :- greatest_2(U0, As, U). :- pred greatest_2(int::in, ranges::in, int::out) is det. greatest_2(U, nil, U). greatest_2(_, range(_, U0, As), U) :- greatest_2(U0, As, U). next(range(L, U, As), N0, N) :- ( if N0 < U then N = int.max(L, N0) + 1 else ranges.next(As, N0, N) ). %---------------------------------------------------------------------------% member(N, range(L, U, As)) :- ( N > L, N =< U ; ranges.member(N, As) ). range_member(L, U, range(A0, A1, As)) :- ( L = A0 + 1, U = A1 ; range_member(L, U, As) ). nondet_member(N, As) :- range_member(L, U, As), int.nondet_int_in_range(L, U, N). %---------------------------------------------------------------------------% subset(A, B) :- % XXX Should implement this more efficiently. ranges.difference(A, B) = nil. disjoint(A, B) :- % XXX Should implement this more efficiently. ranges.intersection(A, B) = nil. %---------------------------------------------------------------------------% % union(A, B) = A \/ B % union(nil, Bs) = Bs. union(As @ range(_, _, _), nil) = As. union(As @ range(LA, UA, As0), Bs @ range(LB, UB, Bs0)) = Result :- compare(R, LA, LB), ( R = (<), Result = n_diff_na_b(LA, UA, As0, Bs) ; R = (=), Result = n_int_na_nb(LA, UA, As0, UB, Bs0) ; R = (>), Result = n_diff_na_b(LB, UB, Bs0, As) ). % n_union_a_nb(L, A, U, B) = % {X | X > L} \ (A \/ ({Y | Y > U} \ B)) % % assuming L < min(A), L < U and U < min(B). % :- func n_union_a_nb(int, ranges, int, ranges) = ranges. n_union_a_nb(L, nil, U, Bs) = range(L, U, Bs). n_union_a_nb(L, As @ range(LA, UA, As0), UB, Bs0) = Result :- compare(R, LA, UB), ( R = (<), Result = range(L, LA, diff_na_nb(UA, As0, UB, Bs0)) ; R = (=), Result = range(L, LA, int_na_b(UA, As0, Bs0)) ; R = (>), Result = range(L, UB, ranges.difference(Bs0, As)) ). % n_union_na_b(L, U, A, B) = % {X | X > L} \ (({Y | Y > U} \ A) \/ B) % % assuming L < U, U < min(A) and L < min(B). % :- func n_union_na_b(int, int, ranges, ranges) = ranges. n_union_na_b(L, U, As, nil) = range(L, U, As). n_union_na_b(L, UA, As0, Bs @ range(LB, UB, Bs0)) = Result :- compare(R, UA, LB), ( R = (<), Result = range(L, UA, ranges.difference(As0, Bs)) ; R = (=), Result = range(L, UA, int_a_nb(As0, UB, Bs0)) ; R = (>), Result = range(L, LB, diff_na_nb(UB, Bs0, UA, As0)) ). % n_union_na_nb(L, UA, A, UB, B) = % {X | X > L} \ (({Y | Y > UA} \ A) \/ ({Z | Z > UB} \ B)) % % assuming L < UA, UA < min(A), L < UB and UB < min(B). % :- func n_union_na_nb(int, int, ranges, int, ranges) = ranges. n_union_na_nb(L, UA, As0, UB, Bs0) = Result :- compare(R, UA, UB), ( R = (<), Result = range(L, UA, diff_a_nb(As0, UB, Bs0)) ; R = (=), Result = range(L, UA, ranges.intersection(As0, Bs0)) ; R = (>), Result = range(L, UB, diff_a_nb(Bs0, UA, As0)) ). % intersection(A, B) = A /\ B % intersection(nil, _) = nil. intersection(range(_, _, _), nil) = nil. intersection(As @ range(LA, UA, As0), Bs @ range(LB, UB, Bs0)) = Result :- compare(R, LA, LB), ( R = (<), Result = diff_a_nb(Bs, UA, As0) ; R = (=), Result = n_union_na_nb(LA, UA, As0, UB, Bs0) ; R = (>), Result = diff_a_nb(As, UB, Bs0) ). % int_na_b(U, A, B) = ({X | X > U} \ A) /\ B % % assuming U < min(A). % :- func int_na_b(int, ranges, ranges) = ranges. int_na_b(_, _, nil) = nil. int_na_b(UA, As0, Bs @ range(LB, UB, Bs0)) = Result :- compare(R, UA, LB), ( R = (<), Result = ranges.difference(Bs, As0) ; R = (=), Result = n_union_a_nb(UA, As0, UB, Bs0) ; R = (>), Result = diff_na_nb(UA, As0, UB, Bs0) ). % n_int_na_nb(L, UA, A, UB, B) = % {X | X > L} (({Y | Y > UA} \ A) /\ ({Z | Z > UB} \ B)) % % assuming L < UA, UA < min(A), L < UB and UB < min(B). % :- func n_int_na_nb(int, int, ranges, int, ranges) = ranges. n_int_na_nb(L, UA, As0, UB, Bs0) = Result :- compare(R, UA, UB), ( R = (<), Result = n_diff_na_b(L, UB, Bs0, As0) ; R = (=), Result = range(L, UA, ranges.union(As0, Bs0)) ; R = (>), Result = n_diff_na_b(L, UA, As0, Bs0) ). % int_a_nb(A, U, B) = A /\ ({X | X > U} \ B) % % assuming U < min(B). % :- func int_a_nb(ranges, int, ranges) = ranges. int_a_nb(nil, _, _) = nil. int_a_nb(As @ range(LA, UA, As0), UB, Bs0) = Result :- compare(R, LA, UB), ( R = (<), Result = diff_na_nb(UB, Bs0, UA, As0) ; R = (=), Result = n_union_na_b(LA, UA, As0, Bs0) ; R = (>), Result = ranges.difference(As, Bs0) ). % difference(A, B) = A \ B % difference(nil, _) = nil. difference(As @ range(_, _, _), nil) = As. difference(As @ range(LA, UA, As0), Bs @ range(LB, UB, Bs0)) = Result :- compare(R, LA, LB), ( R = (<), Result = n_union_na_b(LA, UA, As0, Bs) ; R = (=), Result = diff_na_nb(UB, Bs0, UA, As0) ; R = (>), Result = int_a_nb(As, UB, Bs0) ). % n_diff_na_b(L, U, A, B) = {X | X > L} \ (({Y | Y > U} \ A) \ B) % % assuming L < U, U < min(A) and L < min(B). % :- func n_diff_na_b(int, int, ranges, ranges) = ranges. n_diff_na_b(L, U, As, nil) = range(L, U, As). n_diff_na_b(L, UA, As0, Bs @ range(LB, UB, Bs0)) = Result :- compare(R, UA, LB), ( R = (<), Result = range(L, UA, ranges.union(As0, Bs)) ; R = (=), Result = n_diff_na_b(L, UB, Bs0, As0) ; R = (>), Result = n_int_na_nb(L, UA, As0, UB, Bs0) ). % diff_a_nb(A, U, B) = A \ ({X | X > U} \ B) % % assuming U < min(B). % :- func diff_a_nb(ranges, int, ranges) = ranges. diff_a_nb(nil, _, _) = nil. diff_a_nb(As @ range(LA, UA, As0), UB, Bs0) = Result :- compare(R, LA, UB), ( R = (<), Result = n_union_na_nb(LA, UA, As0, UB, Bs0) ; R = (=), Result = diff_a_nb(Bs0, UA, As0) ; R = (>), Result = ranges.intersection(As, Bs0) ). % diff_na_nb(UA, A, UB, B) = ({X | X > UA} \ A) \ ({Y | Y > UB} \ B) % % assuming UA < min(A) and UB < min(B). % :- func diff_na_nb(int, ranges, int, ranges) = ranges. diff_na_nb(UA, As0, UB, Bs0) = Result :- compare(R, UA, UB), ( R = (<), Result = n_union_a_nb(UA, As0, UB, Bs0) ; R = (=), Result = ranges.difference(Bs0, As0) ; R = (>), Result = int_na_b(UA, As0, Bs0) ). %---------------------------------------------------------------------------% restrict_min(_, nil) = nil. restrict_min(Min, As0 @ range(L, U, As1)) = As :- ( if Min =< L then As = As0 else if Min =< U then As = range(Min - 1, U, As1) else As = restrict_min(Min, As1) ). restrict_max(_, nil) = nil. restrict_max(Max, range(L, U, As0)) = As :- ( if Max =< L then As = nil else if Max =< U then As = range(L, Max, nil) else As = range(L, U, restrict_max(Max, As0)) ). restrict_range(Min, Max, As) = ranges.intersection(range(Min - 1, Max, nil), As). %---------------------------------------------------------------------------% prune_to_next_non_member(nil, nil, !N). prune_to_next_non_member(As @ range(LA, UA, As0), Result, !N) :- ( if !.N =< LA then Result = As else if !.N =< UA then !:N = UA + 1, Result = As0 else prune_to_next_non_member(As0, Result, !N) ). prune_to_prev_non_member(nil, nil, !N). prune_to_prev_non_member(range(LA, UA, As0), Result, !N) :- ( if !.N =< LA then Result = nil else if !.N =< UA then !:N = LA, Result = nil else prune_to_prev_non_member(As0, Result0, !N), Result = range(LA, UA, Result0) ). negate(As) = negate_aux(As, nil). :- func negate_aux(ranges::in, ranges::in) = (ranges::out) is det. negate_aux(nil, As) = As. negate_aux(range(L, U, As), A) = negate_aux(As, range(-U-1, -L-1, A)). plus(nil, nil) = nil. plus(nil, range(_,_,_)) = nil. plus(range(_,_,_), nil) = nil. plus(range(La, Ha, nil), range(L, H, nil)) = range(La + L + 1, Ha + H, nil). plus(range(Lx0, Hx0, Xs0 @ range(Lx1, Hx1, Xs1)), range(L, H, nil)) = Result :- ( if Lx1 - Hx0 < H - L then Result = plus(range(Lx0, Hx1, Xs1), range(L, H, nil)) else Result = range(Lx0 + L + 1, Hx0 + H, plus(Xs0, range(L, H, nil))) ). plus(range(Lx, Hx, Xs), range(L, H, S @ range(_,_,_))) = Result :- A = plus(range(Lx, Hx, Xs), range(L, H, nil)), B = plus(range(Lx, Hx, Xs), S), Result = union(A,B). shift(nil, _) = nil. shift(As @ range(L, H, As0), C) = Result :- ( if C = 0 then Result = As else Result = range(L + C, H + C, shift(As0, C)) ). dilation(nil, _) = nil. dilation(A @ range(_,_,_) , C) = Result :- ( if C < 0 then Result = dilation(negate(A), -C) else if C = 0 then Result = range(-1, 0, nil) else if C = 1 then Result = A else L = to_sorted_list(A), list.map(*(C), L) = L0, Result = from_list(L0) ). contraction(nil, _) = nil. contraction(A @ range(L, H, As), C) = Result :- ( if C < 0 then Result = contraction(negate(A), -C) else if C = 0 then Result = range(-1, 0, nil) else if C = 1 then Result = A else L0 = div_up_xp(L + 1, C) - 1, H0 = div_down_xp(H, C), Result = contraction_0(L0, H0, As, C) ). :- func contraction_0(int, int, ranges, int) = ranges. contraction_0(L0, H0, nil, _) = range(L0, H0, nil). contraction_0(L0, H0, range(L1, H1, As), C) = Result :- L1N = div_up_xp(L1 + 1, C) - 1, H1N = div_down_xp(H1, C), ( if H0 >= L1N then Result = contraction_0(L0, H1N, As, C) else Result = range(L0, H0, contraction_0(L1N, H1N, As, C)) ). % 0 < B. Round up. % :- func div_up_xp(int::in, int::in) = (int::out) is det. div_up_xp(A, B) = (A > 0 -> div_up_pp(A, B) ; div_up_np(A, B)). % 0 < A,B. Round up. % :- func div_up_pp(int::in, int::in) = (int::out) is det. div_up_pp(A, B) = int.unchecked_quotient(A + B - 1, B). % A < 0 < B. Round up. % :- func div_up_np(int::in, int::in) = (int::out) is det. div_up_np(A, B) = int.unchecked_quotient(A, B). % 0 < B. Round down. % :- func div_down_xp(int::in, int::in) = (int::out) is det. div_down_xp(A, B) = (A > 0 -> div_down_pp(A, B) ; div_down_np(A, B)). % 0 < A,B. Round down. % :- func div_down_pp(int::in, int::in) = (int::out) is det. div_down_pp(A, B) = int.unchecked_quotient(A, B). % A < 0 < B. Round down. % :- func div_down_np(int::in, int::in) = (int::out) is det. div_down_np(A, B) = int.unchecked_quotient(A - B + 1, B). %---------------------------------------------------------------------------% to_sorted_list(nil) = []. to_sorted_list(range(L, H, Rest)) = to_sorted_list_2(L, H, to_sorted_list(Rest)). :- func to_sorted_list_2(int, int, list(int)) = list(int). to_sorted_list_2(L, H, Ints) = ( if H = L then Ints else to_sorted_list_2(L, H-1, [H | Ints]) ). from_list(List) = list.foldl(ranges.insert, List, ranges.empty). from_set(Set) = ranges.from_list(set.to_sorted_list(Set)). %---------------------------------------------------------------------------% compare_lex(Result, A, B) :- ( A = nil, B = nil, Result = (=) ; A = nil, B = range(_, _, _), Result = (<) ; A = range(_, _, _), B = nil, Result = (>) ; A = range(LBA, UBA, APrime), B = range(LBB, UBB, BPrime), % NOTE: when we unpack a range/3 constructor we must add one % to the first argument since that is the lowest value in that % subset. compare_lex_2(Result, LBA + 1, UBA, LBB + 1, UBB, APrime, BPrime) ). :- pred compare_lex_2(comparison_result::uo, int::in, int::in, int::in, int::in, ranges::in, ranges::in) is det. compare_lex_2(Result, !.LBA, !.UBA, !.LBB, !.UBB, !.NextA, !.NextB) :- compare(LBResult, !.LBA, !.LBB), ( ( LBResult = (<) ; LBResult = (>) ), Result = LBResult ; LBResult = (=), compare(UBResult, !.UBA, !.UBB), ( UBResult = (=), compare_lex(Result, !.NextA, !.NextB) ; ( UBResult = (<) ; UBResult = (>) ), ( if !.LBA = !.UBA, !.LBB = !.UBB then compare_lex(Result, !.NextA, !.NextB) else if !.LBA < !.UBA, !.LBB = !.UBB then !:LBA = !.LBA + 1, ( !.NextB = nil, Result = (>) ; !.NextB = range(!:LBB, !:UBB, !:NextB), compare_lex_2(Result, !.LBA, !.UBA, !.LBB + 1, !.UBB, !.NextA, !.NextB) ) else if !.LBA = !.UBA, !.LBB < !.UBB then !:LBB = !.LBB + 1, ( !.NextA = nil, Result = (<) ; !.NextA = range(!:LBA, !:UBA, !:NextA), compare_lex_2(Result, !.LBA + 1, !.UBA, !.LBB, !.UBB, !.NextA, !.NextB) ) else !:LBA = !.LBA + 1, !:LBB = !.LBB + 1, disable_warning [suspicious_recursion] ( compare_lex_2(Result, !.LBA, !.UBA, !.LBB, !.UBB, !.NextA, !.NextB) ) ) ) ). %---------------------------------------------------------------------------% foldl(P, Ranges, !Acc) :- ( Ranges = nil ; Ranges = range(L, U, Rest), int.fold_up(P, L + 1, U, !Acc), foldl(P, Rest, !Acc) ). foldl2(P, Ranges, !Acc1, !Acc2) :- ( Ranges = nil ; Ranges = range(L, U, Rest), int.fold_up2(P, L + 1, U, !Acc1, !Acc2), foldl2(P, Rest, !Acc1, !Acc2) ). foldl3(P, Ranges, !Acc1, !Acc2, !Acc3) :- ( Ranges = nil ; Ranges = range(L, U, Rest), int.fold_up3(P, L + 1, U, !Acc1, !Acc2, !Acc3), foldl3(P, Rest, !Acc1, !Acc2, !Acc3) ). foldr(P, Ranges, !Acc) :- ( Ranges = nil ; Ranges = range(L, H, Rest), foldr(P, Rest, !Acc), int.fold_down(P, L + 1, H, !Acc) ). %---------------------------------------------------------------------------% range_foldl(_, nil, !Acc). range_foldl(P, range(L, U, Rest), !Acc) :- P(L + 1, U, !Acc), range_foldl(P, Rest, !Acc). range_foldl2(_, nil, !Acc1, !Acc2). range_foldl2(P, range(L, U, Rest), !Acc1, !Acc2) :- P(L + 1, U, !Acc1, !Acc2), range_foldl2(P, Rest, !Acc1, !Acc2). range_foldr(_, nil, !Acc). range_foldr(P, range(L, U, Rest), !Acc) :- range_foldr(P, Rest, !Acc), P(L + 1, U, !Acc). %---------------------------------------------------------------------------% :- end_module ranges. %---------------------------------------------------------------------------%