ifc-lite-geometry 4.1.0

Geometry processing and mesh generation for IFC models
Documentation
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998
999
1000
1001
1002
1003
1004
1005
1006
1007
1008
1009
1010
1011
1012
1013
1014
1015
1016
1017
1018
1019
1020
1021
1022
1023
1024
1025
1026
1027
1028
1029
1030
1031
1032
1033
1034
1035
1036
1037
1038
1039
1040
1041
1042
1043
1044
1045
1046
1047
1048
1049
1050
1051
1052
1053
1054
1055
1056
1057
1058
1059
1060
1061
1062
1063
1064
1065
1066
1067
1068
1069
1070
1071
1072
1073
1074
1075
1076
1077
1078
1079
1080
1081
1082
1083
1084
1085
1086
1087
1088
1089
1090
1091
1092
1093
1094
1095
1096
1097
1098
1099
1100
1101
1102
1103
1104
1105
1106
1107
1108
1109
1110
1111
1112
1113
1114
1115
1116
1117
1118
1119
1120
1121
1122
1123
1124
1125
1126
1127
1128
1129
1130
1131
1132
1133
1134
1135
1136
1137
1138
1139
1140
1141
1142
1143
1144
1145
1146
1147
1148
1149
1150
1151
1152
1153
1154
1155
1156
1157
1158
1159
1160
1161
1162
1163
1164
1165
1166
1167
1168
1169
1170
1171
1172
1173
1174
1175
1176
1177
1178
1179
1180
1181
1182
1183
1184
1185
1186
1187
1188
1189
1190
1191
1192
1193
1194
1195
1196
1197
1198
1199
1200
1201
1202
1203
1204
1205
1206
1207
1208
1209
1210
1211
1212
1213
1214
1215
1216
1217
1218
1219
1220
1221
1222
1223
1224
1225
1226
1227
1228
1229
1230
1231
1232
1233
1234
1235
1236
1237
1238
1239
1240
1241
1242
1243
1244
1245
1246
1247
1248
1249
1250
1251
1252
1253
1254
1255
1256
1257
1258
1259
1260
1261
1262
1263
1264
1265
1266
1267
1268
1269
1270
1271
1272
1273
1274
1275
1276
1277
1278
1279
1280
1281
1282
1283
1284
1285
1286
1287
1288
1289
1290
1291
1292
1293
1294
1295
1296
1297
1298
1299
1300
1301
1302
1303
1304
1305
1306
1307
1308
1309
1310
1311
1312
1313
1314
1315
1316
1317
1318
1319
1320
1321
1322
1323
1324
1325
1326
1327
1328
1329
1330
1331
1332
1333
1334
1335
1336
1337
1338
1339
1340
1341
1342
1343
1344
1345
1346
1347
1348
1349
1350
1351
1352
1353
1354
1355
1356
1357
1358
1359
1360
1361
1362
1363
1364
1365
1366
1367
1368
// This Source Code Form is subject to the terms of the Mozilla Public
// License, v. 2.0. If a copy of the MPL was not distributed with this
// file, You can obtain one at https://mozilla.org/MPL/2.0/.

//! Mesh arrangement — the conforming intersection of two operand meshes.
//!
//! For every pair of triangles (one per operand) that cross, the intersection
//! segment becomes a constraint for BOTH triangles (it lies on both planes), and
//! each crossed triangle is re-triangulated with its accumulated constraints.
//! All re-triangulations share ONE interner, so a vertex created on the
//! intersection of A's triangle and B's triangle gets the SAME symbolic Vid in
//! both surfaces — i.e. the two operands' meshes CONFORM along the intersection.
//! The result is a single intersection-free complex, ready for L4 winding
//! classification.
//!
//! This increment: all-pairs broadphase (AABB cull) + the proper-crossing
//! `Segment` case. BVH broadphase, coplanar overlap, and vertex/edge `Touches`
//! degeneracies are later increments.

use super::coplanar::coplanar_clip;
use super::interner::{Interner, Vid};
use super::predicates::{cmp_lex, orient2d_any, orient3d};
use super::rational::point_of;
use super::retriangulate::{projection_axis, triangulate, Constraint, RetriInput};
use super::tritri::{tri_tri_intersection, TriTri};
use super::{DropAxis, ImplicitPoint, Sign, Tpi};
use num_traits::ToPrimitive;
use std::cmp::Ordering;

pub type Tri = [[f64; 3]; 3];

/// Boolean operation.
#[derive(Clone, Copy, PartialEq, Eq, Debug)]
pub enum BoolOp {
    /// `A − B`.
    Difference,
    /// `A ∪ B`.
    Union,
    /// `A ∩ B`.
    Intersection,
}

/// The conforming arrangement of two operand meshes over a shared interner.
pub struct Arrangement {
    pub interner: Interner,
    /// Operand A's conforming sub-triangles (interned Vids).
    pub tris_a: Vec<[Vid; 3]>,
    /// Operand B's conforming sub-triangles.
    pub tris_b: Vec<[Vid; 3]>,
    /// Total constraint sub-segments the per-triangle re-triangulations could
    /// not force as edges (`Mesh2d::unrecovered`). Non-zero ⇒ the operands do
    /// not fully conform along their intersection; centroid classification of
    /// straddling sub-triangles is unreliable. The N-ary batched difference
    /// treats this as a hard reject; the binary boolean keeps its historical
    /// graceful-degrade behavior.
    pub unrecovered: usize,
    /// Per-sub-triangle (parallel to `tris_a`) flag: this sub-triangle's PARENT
    /// face had a COPLANAR overlap with the other operand — so a sub-triangle on
    /// it may lie on a shared interface plane and needs the [`solid_side`] regime-2
    /// classification (`boolean_vids`). `false` ⇒ the parent face is transversal to
    /// the other operand, so the plain centroid ray-cast suffices (the common case;
    /// keeps box−box etc. on the fast single-cast path).
    pub coplanar_a: Vec<bool>,
    /// Per-sub-triangle (parallel to `tris_b`) coplanar-parent flag.
    pub coplanar_b: Vec<bool>,
}

/// A raw intersection segment on a triangle, tagged with the cutter triangle that
/// produced it (needed to locate seg×seg crossing points as triple points).
struct RawSeg {
    a: ImplicitPoint,
    b: ImplicitPoint,
    cutter: Tri,
}

#[inline]
fn tri_plane(t: &Tri) -> [[f64; 3]; 3] {
    *t
}

/// Do segments `(a1,b1)` and `(a2,b2)` (in `T`'s plane, projected by `axis`)
/// properly cross at an interior point? (Shared endpoints don't count.)
fn segments_cross(a1: &ImplicitPoint, b1: &ImplicitPoint, a2: &ImplicitPoint, b2: &ImplicitPoint, axis: super::DropAxis) -> bool {
    let s1 = orient2d_any(a1, b1, a2, axis);
    let s2 = orient2d_any(a1, b1, b2, axis);
    let s3 = orient2d_any(a2, b2, a1, axis);
    let s4 = orient2d_any(a2, b2, b1, axis);
    s1 != Sign::Zero && s2 != Sign::Zero && s1 != s2 && s3 != Sign::Zero && s4 != Sign::Zero && s3 != s4
}

/// seg×seg pre-pass for one triangle `t`: split every pair of crossing
/// constraint segments at their crossing point — `TPI(t.plane, cutter_i.plane,
/// cutter_j.plane)` (a valid triple point: crossing ⇒ the two cutter lines are
/// non-parallel in `t`'s plane ⇒ the three planes meet at a point). The result
/// is a crossing-free constraint set the re-triangulation can handle directly.
fn split_crossings(t: &Tri, raws: &[RawSeg]) -> Vec<Constraint> {
    let axis = match projection_axis(t) {
        Some((a, _)) => a,
        None => return Vec::new(),
    };
    let n = raws.len();
    let mut splits: Vec<Vec<ImplicitPoint>> = vec![Vec::new(); n];
    for k in 0..n {
        for l in (k + 1)..n {
            if segments_cross(&raws[k].a, &raws[k].b, &raws[l].a, &raws[l].b, axis) {
                let x = ImplicitPoint::Tpi(Tpi {
                    planes: [tri_plane(t), tri_plane(&raws[k].cutter), tri_plane(&raws[l].cutter)],
                });
                splits[k].push(x.clone());
                splits[l].push(x);
            }
        }
    }
    let mut out = Vec::new();
    for k in 0..n {
        let mut chain = vec![raws[k].a.clone()];
        chain.append(&mut splits[k]);
        chain.push(raws[k].b.clone());
        // order along the segment (collinear ⇒ lex order = line order) + dedup coincident
        chain.sort_by(|p, q| match cmp_lex(p, q) {
            Sign::Negative => Ordering::Less,
            Sign::Positive => Ordering::Greater,
            Sign::Zero => Ordering::Equal,
        });
        chain.dedup_by(|p, q| cmp_lex(p, q) == Sign::Zero);
        for w in chain.windows(2) {
            out.push(Constraint { a: w[0].clone(), b: w[1].clone() });
        }
    }
    out
}

/// Compute the conforming arrangement of operand meshes `a` and `b`.
pub fn arrange(a: &[Tri], b: &[Tri]) -> Arrangement {
    // 1. accumulate raw intersection segments (Segment pairs) + coplanar overlaps
    let mut raw_a: Vec<Vec<RawSeg>> = (0..a.len()).map(|_| Vec::new()).collect();
    let mut raw_b: Vec<Vec<RawSeg>> = (0..b.len()).map(|_| Vec::new()).collect();
    let mut cop_a: Vec<Vec<Constraint>> = (0..a.len()).map(|_| Vec::new()).collect();
    let mut cop_b: Vec<Vec<Constraint>> = (0..b.len()).map(|_| Vec::new()).collect();
    let mut pt_a: Vec<Vec<ImplicitPoint>> = (0..a.len()).map(|_| Vec::new()).collect();
    let mut pt_b: Vec<Vec<ImplicitPoint>> = (0..b.len()).map(|_| Vec::new()).collect();
    let pairs = super::broadphase::candidate_pairs(a, b);
    for (i, j) in pairs {
        // Deterministic escalation guardrail (#1109): stop accumulating
        // intersections once this boolean has blown its BigRational budget. The
        // partial arrangement is discarded by the caller (csg.rs checks
        // `budget::tripped()` and falls back), so an early break is safe; it just
        // bounds the wasted work to ~one more triangle pair.
        if super::budget::tripped() {
            break;
        }
        let (ta, tb) = (&a[i], &b[j]);
        match tri_tri_intersection(ta, tb) {
            TriTri::Segment([s, t]) => {
                raw_a[i].push(RawSeg { a: s.clone(), b: t.clone(), cutter: *tb });
                raw_b[j].push(RawSeg { a: s, b: t, cutter: *ta });
            }
            TriTri::Coplanar => {
                cop_a[i].extend(coplanar_clip(ta, tb).into_iter().map(|(a, b)| Constraint { a, b }));
                cop_b[j].extend(coplanar_clip(tb, ta).into_iter().map(|(a, b)| Constraint { a, b }));
            }
            TriTri::Point(p) => {
                // Tangential touch: BOTH triangles must intern the touch point
                // as a conformity vertex. The triangle on whose EDGE the point
                // lies otherwise never splits that edge, while its neighbor
                // (which sees a full crossing segment ending there) does — a
                // T-junction that opens the surface (the flush-corner-on-
                // diagonal family; see `RetriInput::points`).
                pt_a[i].push(p.clone());
                pt_b[j].push(p);
            }
            TriTri::None => {}
        }
    }
    // 2. seg×seg pre-pass on the segment constraints, then append the coplanar ones
    let build = |tris: &[Tri], raw: &[Vec<RawSeg>], cop: &mut [Vec<Constraint>]| -> Vec<Vec<Constraint>> {
        (0..tris.len())
            .map(|i| {
                let mut c = split_crossings(&tris[i], &raw[i]);
                c.append(&mut cop[i]);
                c
            })
            .collect()
    };
    // Per-ORIGINAL-triangle "had a coplanar overlap with the other operand" flag,
    // captured BEFORE `build` drains the coplanar-constraint accumulators.
    let cop_parent_a: Vec<bool> = cop_a.iter().map(|c| !c.is_empty()).collect();
    let cop_parent_b: Vec<bool> = cop_b.iter().map(|c| !c.is_empty()).collect();
    let ca = build(a, &raw_a, &mut cop_a);
    let cb = build(b, &raw_b, &mut cop_b);
    // 3. re-triangulate each operand over the SHARED interner ⇒ conforming surfaces
    let mut interner = Interner::new();
    let mut unrecovered = 0usize;
    let (tris_a, coplanar_a) =
        retriangulate_each(a, &ca, &pt_a, &cop_parent_a, &mut interner, &mut unrecovered);
    let (tris_b, coplanar_b) =
        retriangulate_each(b, &cb, &pt_b, &cop_parent_b, &mut interner, &mut unrecovered);
    Arrangement { interner, tris_a, tris_b, coplanar_a, coplanar_b, unrecovered }
}

/// The conforming arrangement of `n` operand meshes over one shared interner.
/// `subtris[k]` is mesh `k`'s conforming sub-triangles; `owner[k][t]` is the index
/// in mesh `k`'s sub-tri list (unused externally, kept implicit by position).
pub struct MultiArrangement {
    pub interner: Interner,
    /// `subtris[k]` = mesh `k`'s conforming sub-triangles (interned Vids).
    pub subtris: Vec<Vec<[Vid; 3]>>,
}

/// Conforming arrangement of `meshes` (N operands) over ONE interner.
///
/// Every ORDERED pair of distinct meshes is intersected (segment + coplanar)
/// exactly as the binary [`arrange`] does, and the accumulated constraints
/// re-triangulate each mesh over the shared interner — so a vertex created on a
/// shared edge/seam interns to the SAME Vid in every mesh that touches it (full
/// N-way conformity). This is what lets [`union_all`] classify each sub-triangle
/// against all OTHER solids WITHOUT ever re-meshing an intermediate union (the
/// pairwise-accumulation step that compounds coplanar T-junctions).
pub fn arrange_many(meshes: &[&[Tri]]) -> MultiArrangement {
    let n = meshes.len();
    // per-mesh, per-triangle constraint accumulators
    let mut raw: Vec<Vec<Vec<RawSeg>>> =
        meshes.iter().map(|m| (0..m.len()).map(|_| Vec::new()).collect()).collect();
    let mut cop: Vec<Vec<Vec<Constraint>>> =
        meshes.iter().map(|m| (0..m.len()).map(|_| Vec::new()).collect()).collect();
    let mut pts: Vec<Vec<Vec<ImplicitPoint>>> =
        meshes.iter().map(|m| (0..m.len()).map(|_| Vec::new()).collect()).collect();
    // intersect every unordered mesh pair (i<j); push constraints to BOTH.
    for i in 0..n {
        if super::budget::tripped() {
            break;
        }
        for j in (i + 1)..n {
            let pairs = super::broadphase::candidate_pairs(meshes[i], meshes[j]);
            for (ti, tj) in pairs {
                // Escalation guardrail (#1109) — see `arrange`.
                if super::budget::tripped() {
                    break;
                }
                let (ta, tb) = (&meshes[i][ti], &meshes[j][tj]);
                match tri_tri_intersection(ta, tb) {
                    TriTri::Segment([s, t]) => {
                        raw[i][ti].push(RawSeg { a: s.clone(), b: t.clone(), cutter: *tb });
                        raw[j][tj].push(RawSeg { a: s, b: t, cutter: *ta });
                    }
                    TriTri::Coplanar => {
                        cop[i][ti].extend(
                            coplanar_clip(ta, tb).into_iter().map(|(a, b)| Constraint { a, b }),
                        );
                        cop[j][tj].extend(
                            coplanar_clip(tb, ta).into_iter().map(|(a, b)| Constraint { a, b }),
                        );
                    }
                    TriTri::Point(p) => {
                        // tangential touch — conformity vertex on both sides
                        pts[i][ti].push(p.clone());
                        pts[j][tj].push(p);
                    }
                    TriTri::None => {}
                }
            }
        }
    }
    let mut interner = Interner::new();
    let mut subtris = Vec::with_capacity(n);
    for k in 0..n {
        let cop_parent: Vec<bool> = cop[k].iter().map(|c| !c.is_empty()).collect();
        let cons: Vec<Vec<Constraint>> = (0..meshes[k].len())
            .map(|t| {
                let mut c = split_crossings(&meshes[k][t], &raw[k][t]);
                c.append(&mut cop[k][t]);
                c
            })
            .collect();
        // `union_all` classifies every sub-triangle against each other mesh via the
        // off-plane `solid_side` probe, so it needs no per-sub coplanar flag here.
        let mut _unrecovered = 0usize;
        let (tris, _coplanar) =
            retriangulate_each(meshes[k], &cons, &pts[k], &cop_parent, &mut interner, &mut _unrecovered);
        subtris.push(tris);
    }
    MultiArrangement { interner, subtris }
}

/// `∪ meshes` as a watertight triangle list — the N-ary union.
///
/// Computed in ONE conforming arrangement ([`arrange_many`]), so it never re-meshes
/// an intermediate union (which is what makes left-deep pairwise accumulation tear
/// along coplanar seams shared by 3+ operands — the #960 segmented-roof cutters).
///
/// A sub-triangle of mesh `k` is on `∂(∪meshes)` iff its OUTER side (`+n`) lies
/// outside every other mesh. Identical co-oriented faces shared by several meshes
/// (e.g. duplicated cutter prisms) are kept once, owned by the LOWEST mesh index.
pub fn union_all(meshes: &[&[Tri]]) -> Vec<Tri> {
    if meshes.is_empty() {
        return Vec::new();
    }
    if meshes.len() == 1 {
        return meshes[0].to_vec();
    }
    use std::collections::HashMap;
    let arr = arrange_many(meshes);
    let exts: Vec<f64> = meshes.iter().map(|m| operand_extent(m)).collect();
    // Owner map: oriented (winding-preserving) Vid key → lowest mesh index that
    // KEEPS that face. A later mesh's identical co-oriented copy is dropped.
    let mut owner: HashMap<[Vid; 3], usize> = HashMap::new();
    // First pass: decide keep per sub-triangle (boundary of the union), recording
    // the canonical owner of each kept oriented face.
    let mut keep: Vec<Vec<bool>> = Vec::with_capacity(meshes.len());
    for (k, sub) in arr.subtris.iter().enumerate() {
        let mut kk = Vec::with_capacity(sub.len());
        for &tri in sub {
            let c = centroid_multi(&arr, tri);
            let n = tri_normal_multi(&arr, tri);
            // outer side just past the face along +n
            let outer = offset_point(c, n, exts[k]);
            // on the union boundary iff the outer side is outside ALL other meshes
            let on_boundary = (0..meshes.len())
                .filter(|&m| m != k)
                .all(|m| !point_inside(outer, meshes[m], exts[m]));
            let mut keep_this = on_boundary;
            if keep_this {
                let key = rotate_min_first(tri);
                match owner.get(&key) {
                    Some(&o) if o != k => keep_this = false, // duplicate; earlier mesh owns it
                    _ => {
                        owner.entry(key).or_insert(k);
                    }
                }
            }
            kk.push(keep_this);
        }
        keep.push(kk);
    }
    // Materialize kept sub-triangles to f64.
    let mut out = Vec::new();
    for (k, sub) in arr.subtris.iter().enumerate() {
        for (t, &tri) in sub.iter().enumerate() {
            if keep[k][t] {
                out.push([
                    point_via_interner(&arr.interner, tri[0]),
                    point_via_interner(&arr.interner, tri[1]),
                    point_via_interner(&arr.interner, tri[2]),
                ]);
            }
        }
    }
    out
}

/// Step `step·n̂` off `c` along the unit-normalised `dir`. Deterministic FMA-free
/// f64 (mirrors [`solid_side`]'s probe).
fn offset_point(c: [f64; 3], dir: [f64; 3], far_l: f64) -> [f64; 3] {
    let len = (dir[0] * dir[0] + dir[1] * dir[1] + dir[2] * dir[2]).sqrt();
    if len == 0.0 {
        return c;
    }
    let step = far_l * (1.0 / 1_048_576.0);
    [
        c[0] + dir[0] / len * step,
        c[1] + dir[1] / len * step,
        c[2] + dir[2] / len * step,
    ]
}

/// Centroid of a multi-arrangement sub-triangle (interner lookup).
fn centroid_multi(arr: &MultiArrangement, tri: [Vid; 3]) -> [f64; 3] {
    let c = [
        point_via_interner(&arr.interner, tri[0]),
        point_via_interner(&arr.interner, tri[1]),
        point_via_interner(&arr.interner, tri[2]),
    ];
    [
        (c[0][0] + c[1][0] + c[2][0]) / 3.0,
        (c[0][1] + c[1][1] + c[2][1]) / 3.0,
        (c[0][2] + c[1][2] + c[2][2]) / 3.0,
    ]
}

fn tri_normal_multi(arr: &MultiArrangement, tri: [Vid; 3]) -> [f64; 3] {
    let (a, b, c) = (
        point_via_interner(&arr.interner, tri[0]),
        point_via_interner(&arr.interner, tri[1]),
        point_via_interner(&arr.interner, tri[2]),
    );
    cross3(sub_f64(b, a), sub_f64(c, a))
}

#[inline]
fn point_via_interner(it: &Interner, v: Vid) -> [f64; 3] {
    let pt = it.get(v);
    if let Some(f) = super::fixed::point_to_f64(pt) {
        return f;
    }
    let p = point_of(pt);
    [p[0].to_f64().unwrap(), p[1].to_f64().unwrap(), p[2].to_f64().unwrap()]
}

/// Re-triangulate each original triangle over the shared interner. Returns the
/// conforming sub-triangles AND a parallel `coplanar` flag (each sub-triangle
/// inherits its parent original triangle's `cop_parent[i]` — "had a coplanar
/// overlap with the other operand"). The flag drives the [`solid_side`] regime-2
/// classification in `boolean_vids` only where it can matter.
fn retriangulate_each(
    tris: &[Tri],
    cons: &[Vec<Constraint>],
    pts: &[Vec<ImplicitPoint>],
    cop_parent: &[bool],
    it: &mut Interner,
    unrecovered: &mut usize,
) -> (Vec<[Vid; 3]>, Vec<bool>) {
    let mut out = Vec::new();
    let mut coplanar = Vec::new();
    for (i, t) in tris.iter().enumerate() {
        // Escalation guardrail (#1109): the seam retriangulation runs exact
        // orient2d per ear and is the other heavy escalation site. Stop once the
        // budget is blown — the caller discards the partial arrangement.
        if super::budget::tripped() {
            break;
        }
        let parent_cop = cop_parent.get(i).copied().unwrap_or(false);
        let passthrough = |it: &mut Interner| {
            [
                it.intern(ImplicitPoint::Explicit(t[0])),
                it.intern(ImplicitPoint::Explicit(t[1])),
                it.intern(ImplicitPoint::Explicit(t[2])),
            ]
        };
        let before = out.len();
        let tri_pts = pts.get(i).cloned().unwrap_or_default();
        if cons[i].is_empty() && tri_pts.is_empty() {
            out.push(passthrough(it));
        } else if let Some(mesh) = triangulate(
            &RetriInput { tri: *t, constraints: cons[i].clone(), points: tri_pts },
            it,
        ) {
            *unrecovered += mesh.unrecovered;
            out.extend(mesh.tris);
        } else {
            out.push(passthrough(it)); // degenerate triangle — pass through
        }
        coplanar.resize(out.len(), parent_cop);
        debug_assert!(coplanar.len() == out.len() && before <= out.len());
    }
    (out, coplanar)
}

// --- winding classification + boolean extraction --------------------------

fn cross3(a: [f64; 3], b: [f64; 3]) -> [f64; 3] {
    [a[1] * b[2] - a[2] * b[1], a[2] * b[0] - a[0] * b[2], a[0] * b[1] - a[1] * b[0]]
}
fn dot3(a: [f64; 3], b: [f64; 3]) -> f64 {
    a[0] * b[0] + a[1] * b[1] + a[2] * b[2]
}

#[inline]
fn e(p: [f64; 3]) -> ImplicitPoint {
    ImplicitPoint::Explicit(p)
}

/// EXACT segment–triangle intersection via `orient3d` (no epsilon): the segment
/// `q1→q2` crosses triangle `t` iff its endpoints straddle `t`'s plane AND the
/// line passes the same side of all three edges. A grazing hit (`orient3d == 0`)
/// is rejected — the fixed generic ray direction makes those vanishingly rare.
fn exact_seg_hits_tri(q1: [f64; 3], q2: [f64; 3], t: &Tri) -> bool {
    let s1 = orient3d(&e(t[0]), &e(t[1]), &e(t[2]), &e(q1));
    let s2 = orient3d(&e(t[0]), &e(t[1]), &e(t[2]), &e(q2));
    if s1 == Sign::Zero || s2 == Sign::Zero || s1 == s2 {
        return false;
    }
    let ea = orient3d(&e(q1), &e(q2), &e(t[0]), &e(t[1]));
    let eb = orient3d(&e(q1), &e(q2), &e(t[1]), &e(t[2]));
    let ec = orient3d(&e(q1), &e(q2), &e(t[2]), &e(t[0]));
    ea != Sign::Zero && ea == eb && eb == ec
}

/// Ray-cast "far" distance: just past the operand's actual extent. Critically NOT
/// a huge constant (1e7) — that blows the orient3d float-filter error bound so
/// EVERY predicate escalates to BigRational (≈5000× slower). Sized to the operand,
/// the float filter resolves the common case and only true grazing escalates.
fn operand_extent(tris: &[Tri]) -> f64 {
    let mut hi = 1.0f64;
    for t in tris {
        for v in t {
            for &c in v {
                hi = hi.max(c.abs());
            }
        }
    }
    2.0 * hi + 1.0
}

/// The fixed generic ray direction for parity casts. No two components
/// near-equal and no pairwise ratio near a simple architectural slope (1:1
/// roofs, axis planes). The previous direction had x≈y and dz/dx≈1 — nearly
/// PARALLEL to 45° roof slopes/ridge edges, so a roof-clipped wall's ray grazed
/// the roof and edge-crossings (rejected, not counted) miscounted parity → the
/// sub-ridge gable triangle was wrongly judged inside the cutter and removed
/// (the "missing wall" over-clip). Shared by [`point_inside`] and the
/// per-component AABB ray prefilter so they can never disagree.
fn ray_dir() -> [f64; 3] {
    [0.301_511_3, 0.557_328_1, 0.773_890_1]
}

/// Is point `p` inside the closed mesh `tris`? Exact ray-cast parity to a far
/// point (`far_l` past the extent) along a fixed generic direction; each crossing
/// tested by the exact predicate above.
fn point_inside(p: [f64; 3], tris: &[Tri], far_l: f64) -> bool {
    let dir = ray_dir();
    let far = [p[0] + dir[0] * far_l, p[1] + dir[1] * far_l, p[2] + dir[2] * far_l];
    tris.iter().filter(|t| exact_seg_hits_tri(p, far, t)).count() % 2 == 1
}

fn to_f64_pt(arr: &Arrangement, v: Vid) -> [f64; 3] {
    let pt = arr.interner.get(v);
    // Fast path: materialize via the fixed-width lambda. Falls back to the exact
    // BigRational `point_of` only for off-grid/overflow points (rare).
    if let Some(f) = super::fixed::point_to_f64(pt) {
        return f;
    }
    let p = point_of(pt);
    [p[0].to_f64().unwrap(), p[1].to_f64().unwrap(), p[2].to_f64().unwrap()]
}

fn sub_f64(a: [f64; 3], b: [f64; 3]) -> [f64; 3] {
    [a[0] - b[0], a[1] - b[1], a[2] - b[2]]
}

fn centroid(arr: &Arrangement, tri: [Vid; 3]) -> [f64; 3] {
    let c = [to_f64_pt(arr, tri[0]), to_f64_pt(arr, tri[1]), to_f64_pt(arr, tri[2])];
    [
        (c[0][0] + c[1][0] + c[2][0]) / 3.0,
        (c[0][1] + c[1][1] + c[2][1]) / 3.0,
        (c[0][2] + c[1][2] + c[2][2]) / 3.0,
    ]
}

fn tri_normal(arr: &Arrangement, tri: [Vid; 3]) -> [f64; 3] {
    let (a, b, c) = (to_f64_pt(arr, tri[0]), to_f64_pt(arr, tri[1]), to_f64_pt(arr, tri[2]));
    cross3(sub_f64(b, a), sub_f64(c, a))
}

fn drop_axis_of(n: [f64; 3]) -> DropAxis {
    let an = [n[0].abs(), n[1].abs(), n[2].abs()];
    if an[0] >= an[1] && an[0] >= an[2] {
        DropAxis::X
    } else if an[1] >= an[2] {
        DropAxis::Y
    } else {
        DropAxis::Z
    }
}

/// If point `c` lies exactly on a triangle of `others` (coplanar AND inside it),
/// return that triangle's f64 normal — i.e. detect that a sub-triangle whose
/// centroid is `c` sits on a coplanar SHARED face of the other operand.
fn on_surface_normal(c: [f64; 3], others: &[Tri]) -> Option<[f64; 3]> {
    for t in others {
        if orient3d(&e(t[0]), &e(t[1]), &e(t[2]), &e(c)) != Sign::Zero {
            continue; // c not on t's plane
        }
        let n = cross3(sub_f64(t[1], t[0]), sub_f64(t[2], t[0]));
        let axis = drop_axis_of(n);
        let w0 = orient2d_any(&e(t[0]), &e(t[1]), &e(t[2]), axis);
        if w0 == Sign::Zero {
            continue; // degenerate t
        }
        let inside = |u: [f64; 3], v: [f64; 3]| orient2d_any(&e(u), &e(v), &e(c), axis) != w0.flip();
        if inside(t[0], t[1]) && inside(t[1], t[2]) && inside(t[2], t[0]) {
            return Some(n);
        }
    }
    None
}

/// Power-of-two grid `mesh_bridge` snaps both operands to (metres); the
/// `near_on_surface_normal` band below is sized to the scatter this snap
/// leaves on a TILTED flush face.
use super::mesh_bridge::SNAP_GRID;

/// The NEAR-coplanar analogue of [`on_surface_normal`], used ONLY for a
/// sub-triangle whose parent face had a near-coplanar overlap with the other
/// operand (`coplanar_a/b[i]` set). Returns the covering `other` face's f64 normal
/// when `c` sits within the snap-scatter band of that face's plane AND projects
/// strictly inside it.
///
/// WHY this exists — the flush-cap defect (#1007 host #1112 openings #2150/#2154):
/// real IFC is f32, so an opening cap authored EXACTLY flush with a TILTED roof
/// surface is NOT exactly coplanar after `mesh_bridge`'s per-axis snap — each
/// operand scatters up to `SNAP_GRID·√3` off the shared tilted plane. The exact
/// [`on_surface_normal`] needs `orient3d == 0` (centroid bit-exactly on the other
/// face's plane), so it MISSES the flush cap; the sub-triangle then falls to the
/// fragile `solid_side` probe (regime 2) or the on-boundary centroid ray-cast
/// (regime 3), both undefined at a µm-flush interface ⇒ the inside-footprint host
/// sub-triangle is wrongly KEPT and bridges the opening.
///
/// DETERMINISM: the only inexactness vs. [`on_surface_normal`] is the perpendicular
/// plane-gap admitted (the `band` test); the in-plane containment still uses the
/// EXACT `orient2d_any`. `band` is an absolute power-of-two multiple of `SNAP_GRID`
/// (≈ the 2-operand scatter envelope) widened only for far-from-origin operands
/// (coarser f32 import) — always THREE orders below the smallest real feature edge
/// (~0.2 m), so a genuinely-distinct parallel face (a thin slab's two surfaces)
/// can never be within it. All FMA-free f64 over input coords ⇒ byte-identical
/// native==wasm. GATED on the near-coplanar-parent flag, so a transversal cut
/// (every pinned box−box manifest face) never reaches it.
fn near_on_surface_normal(c: [f64; 3], others: &[Tri]) -> Option<[f64; 3]> {
    let mut extent = 1.0f64;
    for &x in &c {
        extent = extent.max(x.abs());
    }
    for t in others {
        for v in t {
            for &x in v {
                extent = extent.max(x.abs());
            }
        }
    }
    // band: max perpendicular plane-gap a flush f32 face leaves after the 2-operand
    // snap, widened for far-from-origin coords (8·SNAP_GRID ≈ 0.12 mm; the
    // extent·2^-22 term only dominates past ~32 km from origin).
    let band = (8.0 * SNAP_GRID).max(extent * (1.0 / 4_194_304.0));
    let band2 = band * band;
    for t in others {
        let n = cross3(sub_f64(t[1], t[0]), sub_f64(t[2], t[0]));
        let nn = n[0] * n[0] + n[1] * n[1] + n[2] * n[2];
        if nn <= 0.0 || !nn.is_finite() {
            continue; // degenerate t
        }
        // squared perpendicular distance of `c` to t's plane = (d·n)² / |n|²
        let d = dot3(sub_f64(c, t[0]), n);
        if (d * d) / nn > band2 {
            continue; // c not within the snap band of t's plane
        }
        let axis = drop_axis_of(n);
        let w0 = orient2d_any(&e(t[0]), &e(t[1]), &e(t[2]), axis);
        if w0 == Sign::Zero {
            continue; // degenerate t in projection
        }
        let inside = |u: [f64; 3], v: [f64; 3]| orient2d_any(&e(u), &e(v), &e(c), axis) != w0.flip();
        if inside(t[0], t[1]) && inside(t[1], t[2]) && inside(t[2], t[0]) {
            return Some(n);
        }
    }
    None
}

/// Is the OTHER operand's solid present just off `c` along `±dir`?
///
/// `c` is a sub-triangle centroid that may lie EXACTLY on a shared interface plane
/// where the on-plane ray-cast is undefined. We probe an infinitesimal step to
/// EACH side along the (unit-ish) face normal `dir` and ray-cast each probe point
/// — the probe points are strictly off the plane, so the parity test is
/// well-defined. Returns `(in_plus, in_minus)` = solid present on the `+dir` /
/// `−dir` side.
///
/// DETERMINISM: every operation here is FMA-free f64 over input coordinates, and
/// the ray-cast uses the all-`Explicit` `geometry_predicates::orient3d` path (no
/// implicit points, no BigRational escalation, const float error bounds) — so the
/// keep/drop verdict is byte-identical native==wasm, exactly like the existing
/// on-plane centroid classification.
fn solid_side(c: [f64; 3], dir: [f64; 3], other: &[Tri], far_l: f64) -> (bool, bool) {
    // Unit-normalise `dir` so the step magnitude is plane-independent, then step a
    // small fraction of the operand extent off the plane — far enough that the
    // float predicate resolves the probe vs. the plane, near enough that no other
    // feature is crossed.
    let len = (dir[0] * dir[0] + dir[1] * dir[1] + dir[2] * dir[2]).sqrt();
    if len == 0.0 {
        return (false, false);
    }
    // far_l = 2*extent + 1; a step of far_l * 2^-20 ≈ extent * 2^-19 is ~µm at
    // building scale — well inside a face yet resolvable by orient3d.
    let step = far_l * (1.0 / 1_048_576.0);
    let u = [dir[0] / len * step, dir[1] / len * step, dir[2] / len * step];
    let p_plus = [c[0] + u[0], c[1] + u[1], c[2] + u[2]];
    let p_minus = [c[0] - u[0], c[1] - u[1], c[2] - u[2]];
    (point_inside(p_plus, other, far_l), point_inside(p_minus, other, far_l))
}

/// The B operand as PAIRWISE-DISJOINT closed components (disjoint-cutter
/// batching). The binary boolean is the 1-component special case. Classification
/// against the union of disjoint closed solids decomposes per component —
/// "inside the union" ⇔ "inside exactly one component" — which (a) caps the ray
/// parity scan at the component the ray can actually reach (the per-component
/// AABB prefilter removes the O(|subA|·Σ|B_k|) blowup of a soup-wide cast), and
/// (b) structurally eliminates cross-component graze-parity fragility (a graze
/// on a FAR component can no longer flip the verdict for a near one).
struct BComponents<'a> {
    comps: &'a [&'a [Tri]],
    /// Per-component AABB, inflated by `pad` (conservative for both the ray
    /// prefilter and the coincident-face band test).
    aabbs: Vec<([f64; 3], [f64; 3])>,
    /// Per-component ray-cast far length (`operand_extent`), exactly as the
    /// binary path computes it for its single operand.
    exts: Vec<f64>,
}

impl<'a> BComponents<'a> {
    fn new(comps: &'a [&'a [Tri]]) -> Self {
        let exts: Vec<f64> = comps.iter().map(|c| operand_extent(c)).collect();
        // Inflation: ≥ 2× the near-coplanar band envelope (`8·SNAP_GRID`,
        // widened far from origin), so a centroid within the coincident-face
        // band of a component face is always inside that component's inflated
        // AABB; also dwarfs any f64 rounding in the slab test. Deterministic
        // FMA-free f64.
        let max_ext = exts.iter().cloned().fold(1.0f64, f64::max);
        let pad = 4.0 * (8.0 * SNAP_GRID).max(max_ext * (1.0 / 4_194_304.0));
        let aabbs = comps
            .iter()
            .map(|c| {
                let mut lo = [f64::MAX; 3];
                let mut hi = [f64::MIN; 3];
                for t in c.iter() {
                    for v in t {
                        for k in 0..3 {
                            lo[k] = lo[k].min(v[k]);
                            hi[k] = hi[k].max(v[k]);
                        }
                    }
                }
                for k in 0..3 {
                    lo[k] -= pad;
                    hi[k] += pad;
                }
                (lo, hi)
            })
            .collect();
        Self { comps, aabbs, exts }
    }

    /// Conservative "could the parity segment from `p` hit component `k`?" —
    /// a slab test of `[p, p + dir·ext_k]` against the inflated AABB. Sound:
    /// the component's triangles lie inside the (un-inflated) AABB, so a
    /// segment that misses the inflated AABB has parity 0 there.
    fn ray_may_hit(&self, k: usize, p: [f64; 3]) -> bool {
        let dir = ray_dir();
        let far_l = self.exts[k];
        let q = [p[0] + dir[0] * far_l, p[1] + dir[1] * far_l, p[2] + dir[2] * far_l];
        let (lo, hi) = (&self.aabbs[k].0, &self.aabbs[k].1);
        let (mut t0, mut t1) = (0.0f64, 1.0f64);
        for i in 0..3 {
            let d = q[i] - p[i];
            if d == 0.0 {
                if p[i] < lo[i] || p[i] > hi[i] {
                    return false;
                }
                continue;
            }
            let (a, b) = ((lo[i] - p[i]) / d, (hi[i] - p[i]) / d);
            let (a, b) = if a <= b { (a, b) } else { (b, a) };
            t0 = t0.max(a);
            t1 = t1.min(b);
            if t0 > t1 {
                return false;
            }
        }
        true
    }

    #[inline]
    fn point_in_aabb(&self, k: usize, p: [f64; 3]) -> bool {
        let (lo, hi) = (&self.aabbs[k].0, &self.aabbs[k].1);
        (0..3).all(|i| p[i] >= lo[i] && p[i] <= hi[i])
    }

    /// Is `p` inside the union of the components? (Pairwise-disjoint closed
    /// solids ⇒ inside exactly one ⇒ a per-component exact ray parity, each
    /// prefiltered by its AABB.)
    fn inside(&self, p: [f64; 3]) -> bool {
        self.comps
            .iter()
            .enumerate()
            .any(|(k, comp)| self.ray_may_hit(k, p) && point_inside(p, comp, self.exts[k]))
    }

    /// Regime-1 coincident-face probe across the components: the EXACT
    /// [`on_surface_normal`] on every AABB-near component first, then the
    /// snap-band [`near_on_surface_normal`] analogue (same priority order as
    /// the binary path). Disjointness makes at most one component eligible.
    fn surface_normal(&self, c: [f64; 3]) -> Option<[f64; 3]> {
        for (k, comp) in self.comps.iter().enumerate() {
            if self.point_in_aabb(k, c) {
                if let Some(n) = on_surface_normal(c, comp) {
                    return Some(n);
                }
            }
        }
        for (k, comp) in self.comps.iter().enumerate() {
            if self.point_in_aabb(k, c) {
                if let Some(n) = near_on_surface_normal(c, comp) {
                    return Some(n);
                }
            }
        }
        None
    }

    /// Multi-component [`solid_side`]: probe an infinitesimal step to each side
    /// of the face along `dir` and test the union-inside on each probe. The
    /// step is sized to the LARGEST component extent (== the binary behavior
    /// for one component).
    fn solid_side(&self, c: [f64; 3], dir: [f64; 3]) -> (bool, bool) {
        let len = (dir[0] * dir[0] + dir[1] * dir[1] + dir[2] * dir[2]).sqrt();
        if len == 0.0 {
            return (false, false);
        }
        let max_ext = self.exts.iter().cloned().fold(1.0f64, f64::max);
        let step = max_ext * (1.0 / 1_048_576.0);
        let u = [dir[0] / len * step, dir[1] / len * step, dir[2] / len * step];
        let p_plus = [c[0] + u[0], c[1] + u[1], c[2] + u[2]];
        let p_minus = [c[0] - u[0], c[1] - u[1], c[2] - u[2]];
        (self.inside(p_plus), self.inside(p_minus))
    }
}

/// Regime-2 classifier (see [`boolean_vids`]): a sub-triangle whose centroid `c`
/// lies on the OTHER solid's boundary surface — without a COINCIDENT face there
/// (regime 1 already handled that), so the on-plane centroid ray-cast is undefined.
///
/// Applies ONLY when the other solid's boundary actually passes through `c` —
/// detected by [`solid_side`] reporting DIFFERENT inside/outside verdicts on the
/// two sides of the face (`op_plus != op_minus`). That is exactly the degenerate
/// case the plain ray-cast can't resolve (the #960 seam end-region: the other
/// solid abuts via a NON-coplanar face, so its solid is present on one side of
/// this coplanar sub-triangle but it has no coincident face there). When the two
/// sides AGREE, `c` is strictly inside or outside the other solid → the plain
/// ray-cast is reliable → return `None` so the caller uses it (preserving every
/// existing in/out classification, incl. the pinned box−box manifests).
///
/// `c` is on the boundary, so the other solid occupies exactly one side. With the
/// outward normal `n` (own-solid on `−n`):
/// * Union — keep iff the other solid is NOT outside (`!op_plus`): an interface
///   with the other solid on the `+n` side is interior to `A∪B` → drop.
/// * Intersection — keep iff the other solid is on the inner side: `op_minus`.
/// * Difference — keep iff the other solid is NOT on the inner side: `!op_minus`.
fn on_interface_keep(
    arr: &Arrangement,
    tri: [Vid; 3],
    c: [f64; 3],
    other: &[Tri],
    ext_other: f64,
    op: BoolOp,
    _a_side: bool,
) -> Option<bool> {
    let n = tri_normal(arr, tri);
    let (op_plus, op_minus) = solid_side(c, n, other, ext_other);
    if op_plus == op_minus {
        // `c` is strictly inside/outside the other solid — NOT on its boundary;
        // the plain centroid ray-cast is well-defined, so defer to it. This is the
        // common case for every non-coplanar sub-triangle (incl. the pinned
        // box−box cut faces), so regime 2 perturbs nothing there.
        return None;
    }
    Some(match op {
        BoolOp::Union => !op_plus,
        BoolOp::Intersection => op_minus,
        BoolOp::Difference => !op_minus,
    })
}

/// The boolean result as ORIENTED Vid triangles.
///
/// A sub-triangle's centroid is classified against the OTHER operand in one of
/// three regimes:
/// 1. **On a coincident SHARED face** (`on_surface_normal` finds a covering,
///    coplanar triangle of the other operand): classify by NORMAL AGREEMENT —
///    keep the A-copy for a co-oriented face on a union/intersection, or an
///    opposite face on a difference; drop the B-copy (dedup). This is the
///    original exact handling and is unchanged.
/// 2. **On a shared INTERFACE plane WITHOUT a coincident face** (the segmented-
///    roof seam end-regions, #960 — the other solid abuts via a NON-coplanar
///    face, so regime 1 misses it and the on-plane ray-cast is undefined):
///    classify by the other solid's presence on the `−n` (inner) side, probed an
///    infinitesimal step off the plane ([`solid_side`]). This is the fix: such a
///    sub-triangle is an interior interface for union (drop) but was previously
///    mis-routed to the undefined ray-cast and wrongly kept.
/// 3. **Strictly off any shared plane**: the exact centroid ray-cast, as before.
///
/// For union/intersection, a co-oriented DUPLICATE face (same patch, same winding
/// — e.g. unioning identical/overlapping coplanar solids) survives regime 1's
/// keep test on BOTH operands; because the arrangement conforms over one interner
/// it is the SAME oriented Vid triangle on both, so the B-copy is dropped when its
/// rotation-canonical key already appears among the kept A-copies. (A back-to-back
/// interface shares the unordered vertex set but has OPPOSITE winding, so it never
/// collides; both its copies are dissolved by regime 2.)
fn boolean_vids(arr: &Arrangement, a: &[Tri], b: &[Tri], op: BoolOp) -> Vec<[Vid; 3]> {
    boolean_vids_components(arr, a, &BComponents::new(&[b]), op)
}

/// [`boolean_vids`] with the B operand as pairwise-disjoint closed components
/// (see [`BComponents`]). For a single component this is verdict-identical to
/// the historical binary classifier — the AABB gates are pure prefilters — so
/// the pinned boolean manifest is unperturbed.
fn boolean_vids_components(
    arr: &Arrangement,
    a: &[Tri],
    bc: &BComponents,
    op: BoolOp,
) -> Vec<[Vid; 3]> {
    use std::collections::HashSet;
    let ext_a = operand_extent(a);
    let dedup = matches!(op, BoolOp::Union | BoolOp::Intersection);
    let mut a_kept: HashSet<[Vid; 3]> = HashSet::new();
    let mut out = Vec::new();
    for (i, &tri) in arr.tris_a.iter().enumerate() {
        let c = centroid(arr, tri);
        let cop_parent = arr.coplanar_a.get(i).copied().unwrap_or(false);
        let keep;
        // regime 1: coincident shared face → classify by normal agreement. Tried
        // EXACT first, then the snap-band-flush analogue (`near_on_surface_normal`)
        // which catches a face left a few µm off a TILTED shared plane by per-axis
        // import snapping (the #1007 flush roof-opening cap). The near test is
        // ungated (like the exact one) because a coincident-shared-face DROP/keep is
        // unconditionally correct, and its centroid-inside + µm-perp requirements
        // never match a transversal cut face (every pinned box−box manifest face).
        if let Some(n_other) = bc.surface_normal(c) {
            let co_oriented = dot3(tri_normal(arr, tri), n_other) > 0.0;
            keep = match op {
                BoolOp::Union | BoolOp::Intersection => co_oriented,
                BoolOp::Difference => !co_oriented,
            };
        } else if let Some(k) = cop_parent
            .then(|| {
                // regime 2 (only on coplanar-overlap parents): shared interface
                // plane, no coincident face → solid-side probe (see
                // `on_interface_keep` for the keep table rationale).
                let n = tri_normal(arr, tri);
                let (op_plus, op_minus) = bc.solid_side(c, n);
                if op_plus == op_minus {
                    return None; // strictly in/out — the plain ray-cast decides
                }
                Some(match op {
                    BoolOp::Union => !op_plus,
                    BoolOp::Intersection => op_minus,
                    BoolOp::Difference => !op_minus,
                })
            })
            .flatten()
        {
            keep = k;
        } else {
            // regime 3: strictly off-plane → centroid ray-cast (original)
            let inside_b = bc.inside(c);
            keep = match op {
                BoolOp::Intersection => inside_b,
                _ => !inside_b,
            };
        }
        if keep {
            if dedup {
                a_kept.insert(rotate_min_first(tri));
            }
            out.push(tri);
        }
    }
    for (i, &tri) in arr.tris_b.iter().enumerate() {
        // dedup a true co-oriented duplicate of a kept A face (keep the A-copy)
        if dedup && a_kept.contains(&rotate_min_first(tri)) {
            continue;
        }
        let c = centroid(arr, tri);
        let cop_parent = arr.coplanar_b.get(i).copied().unwrap_or(false);
        // A coincident-shared B face is dropped (the A-copy is the kept one). Tried
        // EXACT first, then the snap-band-flush analogue — ungated, because a B cap
        // fully CONTAINED in a larger host face has NO coplanar constraint of its
        // own (the host imposes none on it) so `coplanar_b` is unset for it, yet it
        // is still a coincident shared face that must drop (the #1007 flush roof cap
        // — without the near drop here it survives and bridges the opening).
        if on_surface_normal(c, a).is_some() || near_on_surface_normal(c, a).is_some() {
            continue; // coplanar-shared B-copy: dropped (the A-copy is the kept one)
        }
        if cop_parent {
            if let Some(keep) = on_interface_keep(arr, tri, c, a, ext_a, op, false) {
                // regime 2 for B (coplanar-overlap parent only).
                if keep {
                    let flip = matches!(op, BoolOp::Difference);
                    out.push(if flip { [tri[0], tri[2], tri[1]] } else { tri });
                }
                continue;
            }
        }
        let inside_a = point_inside(c, a, ext_a);
        let (keep, flip) = match op {
            BoolOp::Difference => (inside_a, true),
            BoolOp::Union => (!inside_a, false),
            BoolOp::Intersection => (inside_a, false),
        };
        if keep {
            out.push(if flip { [tri[0], tri[2], tri[1]] } else { tri });
        }
    }
    out
}

/// Compute the boolean `op` of operand meshes `a` and `b`, materialised to f64
/// triangles. `A−B = (A outside B) ∪ flip(B inside A)`;
/// `A∪B = (A outside B) ∪ (B outside A)`; `A∩B = (A inside B) ∪ (B inside A)`.
pub fn boolean(a: &[Tri], b: &[Tri], op: BoolOp) -> Vec<Tri> {
    let arr = arrange(a, b);
    let vids = boolean_vids(&arr, a, b, op);
    vids.into_iter()
        .map(|t| [to_f64_pt(&arr, t[0]), to_f64_pt(&arr, t[1]), to_f64_pt(&arr, t[2])])
        .collect()
}

/// `a − (∪ comps)` for PAIRWISE-DISJOINT, per-component-closed, OUTWARD-wound
/// cutter components, computed in ONE conforming arrangement (disjoint-cutter
/// batching).
///
/// Soundness: `arrange` only intersects A×B pairs, and disjoint B components
/// have no B×B intersections to miss; classification decomposes per component
/// (`BComponents`). The #2176 lesson is a PRECONDITION here: every component
/// must be closed and outward-wound on its own — the caller (`mesh_bridge::
/// subtract_many`) orients each component before concatenation, and the router
/// only admits per-component-watertight cutters to a batch.
///
/// vs. sequential per-cutter subtraction this avoids re-arranging the (growing)
/// host once per cutter — each intermediate round-trip through f32 + the snap
/// grid re-jitters carve vertices off shared planes, so cut N+1 re-cracks what
/// cut N reconciled (many-void walls' compounding open edges) — and is
/// ~N× cheaper on N box cutters.
pub fn difference_all(a: &[Tri], comps: &[&[Tri]]) -> Option<Vec<Tri>> {
    if comps.is_empty() {
        return Some(a.to_vec());
    }
    let b_all: Vec<Tri> = comps.iter().flat_map(|c| c.iter().copied()).collect();
    let arr = arrange(a, &b_all);
    // Hard conformity gate (unlike the binary `boolean`'s graceful degrade):
    // an unrecovered constraint means some sub-triangle STRADDLES a cutter
    // boundary and its centroid classification can silently over/under-cut
    // (the #559171 batched-door family: the jamb never carved into the back
    // face ⇒ −0.43 m³ and an open rim). The caller falls back to sequential
    // per-cutter subtraction, which carves few constraints per arrangement and
    // avoids the degenerate channel.
    if arr.unrecovered > 0 {
        return None;
    }
    let bc = BComponents::new(comps);
    let vids = boolean_vids_components(&arr, a, &bc, BoolOp::Difference);
    Some(
        vids.into_iter()
            .map(|t| [to_f64_pt(&arr, t[0]), to_f64_pt(&arr, t[1]), to_f64_pt(&arr, t[2])])
            .collect(),
    )
}

#[inline]
fn rotate_min_first(t: [Vid; 3]) -> [Vid; 3] {
    let i = (0..3).min_by_key(|&k| t[k]).unwrap();
    [t[i], t[(i + 1) % 3], t[(i + 2) % 3]]
}

/// Topology fingerprint of a boolean result: each oriented Vid triangle rotated
/// min-first (canonical start vertex, winding preserved), the list sorted,
/// FNV-1a-hashed. Platform-stable — Vids are deterministic symbolic identities
/// and the ray-cast classification is FMA-free f64.
pub fn boolean_topology_hash(a: &[Tri], b: &[Tri], op: BoolOp) -> u64 {
    let arr = arrange(a, b);
    let mut tris: Vec<[Vid; 3]> =
        boolean_vids(&arr, a, b, op).into_iter().map(rotate_min_first).collect();
    tris.sort_unstable();
    let mut h: u64 = 0xcbf2_9ce4_8422_2325;
    for t in tris {
        for v in t {
            h ^= v as u64;
            h = h.wrapping_mul(0x0000_0100_0000_01b3);
        }
    }
    h
}

/// Axis-aligned box `[lo,hi]` as 12 outward-wound triangles.
pub fn box_mesh(lo: [f64; 3], hi: [f64; 3]) -> Vec<Tri> {
    let p = [
        [lo[0], lo[1], lo[2]], [hi[0], lo[1], lo[2]], [hi[0], hi[1], lo[2]], [lo[0], hi[1], lo[2]],
        [lo[0], lo[1], hi[2]], [hi[0], lo[1], hi[2]], [hi[0], hi[1], hi[2]], [lo[0], hi[1], hi[2]],
    ];
    let idx = [
        [0, 3, 2], [0, 2, 1], [4, 5, 6], [4, 6, 7],
        [0, 4, 7], [0, 7, 3], [1, 2, 6], [1, 6, 5],
        [0, 1, 5], [0, 5, 4], [3, 7, 6], [3, 6, 2],
    ];
    idx.iter().map(|f| [p[f[0]], p[f[1]], p[f[2]]]).collect()
}

/// Axis-aligned cube `[lo,hi]³`.
pub fn cube_mesh(lo: f64, hi: f64) -> Vec<Tri> {
    box_mesh([lo, lo, lo], [hi, hi, hi])
}

/// Cross-platform full-boolean determinism manifest: `cube[0,2]³ − cube[1,3]³`.
pub fn boolean_manifest() -> u64 {
    boolean_topology_hash(&cube_mesh(0.0, 2.0), &cube_mesh(1.0, 3.0), BoolOp::Difference)
}

#[cfg(test)]
mod tests {
    use super::*;
    use std::collections::BTreeSet;

    fn cube(lo: f64, hi: f64) -> Vec<Tri> {
        super::cube_mesh(lo, hi)
    }

    /// Signed volume of a closed mesh (divergence theorem): (1/6)Σ v0·(v1×v2).
    fn volume(m: &[Tri]) -> f64 {
        m.iter().map(|t| dot3(t[0], cross3(t[1], t[2]))).sum::<f64>() / 6.0
    }

    /// 552611 regression — coplanar host face × cutter face complex where one
    /// cutter corner's entire triangulation fan is swallowed by a constraint
    /// channel: the tiny middle-quad diagonal (5.027,3.800)→(5.142,3.5) crosses
    /// every fan spoke of the adjacent through-rectangle corner (5.142,3.800),
    /// and `recover_subsegment`'s pocket rebuild used to DESTROY that vertex
    /// (and every constraint edge through it). The host sub-triangulation then
    /// overlapped the cutter footprint and the difference over-cut the wall 4×
    /// (advanced_model wall #552611 / opening #552651). The conforming
    /// arrangement must keep every cutter corner as a host vertex.
    #[test]
    fn coplanar_channel_swallowed_corner_survives_552611() {
        let x = 7.7642822265625;
        let host: Vec<Tri> = vec![
            [[x, 7.7840423583984375, 0.09417724609375], [x, -0.6199951171875, 0.09417724609375], [x, -0.6199951171875, 10.600006103515625]],
            [[x, 7.7840423583984375, 0.09417724609375], [x, -0.6199951171875, 10.600006103515625], [x, 7.7840423583984375, 10.600006103515625]],
        ];
        let cutter: Vec<Tri> = vec![
            [[x, -0.160003662109375, 3.8000030517578125], [x, 5.02655029296875, 3.8000030517578125], [x, 5.02655029296875, 3.5]],
            [[x, -0.160003662109375, 3.8000030517578125], [x, 5.02655029296875, 3.5], [x, -0.160003662109375, 3.5]],
            [[x, 5.02655029296875, 3.8000030517578125], [x, 5.14154052734375, 3.8000030517578125], [x, 5.14154052734375, 3.5]],
            [[x, 5.02655029296875, 3.8000030517578125], [x, 5.14154052734375, 3.5], [x, 5.02655029296875, 3.5]],
            [[x, 5.14154052734375, 3.8000030517578125], [x, 7.7840423583984375, 3.8000030517578125], [x, 7.7840423583984375, 3.5]],
            [[x, 5.14154052734375, 3.8000030517578125], [x, 7.7840423583984375, 3.5], [x, 5.14154052734375, 3.5]],
        ];
        let arr = arrange(&host, &cutter);
        // Every cutter corner must be a vertex of the host's conforming
        // sub-triangulation (they all lie strictly inside / on the host face).
        for ct in &cutter {
            for corner in ct {
                let found = arr.tris_a.iter().flatten().any(|&v| {
                    let p = to_f64_pt(&arr, v);
                    p[1] == corner[1] && p[2] == corner[2]
                });
                assert!(
                    found,
                    "cutter corner ({}, {}) lost from the host conforming triangulation",
                    corner[1], corner[2]
                );
            }
        }
    }

    #[test]
    fn arrange_two_crossing_triangles_conform_along_the_intersection() {
        // Two triangles that skewer each other (from the tri-tri tests).
        let ta: Tri = [[-2., 0., -1.], [2., 0., -1.], [0., 0., 2.]]; // plane y=0
        let tb: Tri = [[1., -2., 1.], [1., 2., 1.], [1., 0.5, -3.]]; // plane x=1
        let arr = arrange(&[ta], &[tb]);
        // both operands were subdivided
        assert!(arr.tris_a.len() >= 2, "operand A not subdivided");
        assert!(arr.tris_b.len() >= 2, "operand B not subdivided");
        // CONFORMITY: the intersection segment's two endpoints are shared vertices
        // (same Vid) of BOTH operands' sub-meshes — they have no coincident corners,
        // so the only shared vertices are the intersection endpoints.
        let va: BTreeSet<Vid> = arr.tris_a.iter().flatten().copied().collect();
        let vb: BTreeSet<Vid> = arr.tris_b.iter().flatten().copied().collect();
        let shared: Vec<Vid> = va.intersection(&vb).copied().collect();
        assert_eq!(
            shared.len(),
            2,
            "operands must share exactly the 2 intersection-segment vertices (conformity)"
        );
    }

    #[test]
    fn arrange_disjoint_meshes_are_untouched() {
        let ta: Tri = [[0., 0., 0.], [1., 0., 0.], [0., 1., 0.]];
        let tb: Tri = [[0., 0., 10.], [1., 0., 10.], [0., 1., 10.]]; // far away
        let arr = arrange(&[ta], &[tb]);
        assert_eq!(arr.tris_a.len(), 1, "disjoint A should pass through");
        assert_eq!(arr.tris_b.len(), 1, "disjoint B should pass through");
    }

    #[test]
    fn cube_helper_has_outward_winding() {
        assert!((volume(&cube(0., 2.)) - 8.0).abs() < 1e-9, "cube volume wrong (winding?)");
    }

    #[test]
    fn boolean_containment_cases_have_exact_volumes() {
        // A entirely inside B — no surface intersection, so this exercises the
        // classification + extraction WITHOUT the arrangement / seg×seg crossings.
        let a = cube(1., 2.); // vol 1, strictly inside B
        let b = cube(0., 3.); // vol 27
        let diff = boolean(&a, &b, BoolOp::Difference);
        assert!(volume(&diff).abs() < 1e-9, "A−B should be empty, vol={}", volume(&diff));
        let inter = boolean(&a, &b, BoolOp::Intersection);
        assert!((volume(&inter) - 1.0).abs() < 1e-9, "A∩B should be A (vol 1), got {}", volume(&inter));
        let uni = boolean(&a, &b, BoolOp::Union);
        assert!((volume(&uni) - 27.0).abs() < 1e-9, "A∪B should be B (vol 27), got {}", volume(&uni));
    }

    #[test]
    fn box_minus_box_real_cut_has_exact_volume() {
        // Two overlapping cubes — a real surface cut, exercising seg×seg crossings
        // (each cut face gets a closed constraint loop with corner X-junctions).
        let a = cube(0., 2.); // vol 8
        let b = cube(1., 3.); // vol 8, overlap [1,2]³ = vol 1
        let diff = volume(&boolean(&a, &b, BoolOp::Difference));
        assert!((diff - 7.0).abs() < 1e-6, "A−B volume = {diff}, expected 7");
        let inter = volume(&boolean(&a, &b, BoolOp::Intersection));
        assert!((inter - 1.0).abs() < 1e-6, "A∩B volume = {inter}, expected 1");
        let uni = volume(&boolean(&a, &b, BoolOp::Union));
        assert!((uni - 15.0).abs() < 1e-6, "A∪B volume = {uni}, expected 15");
    }

    #[test]
    fn abutting_boxes_union_is_manifold_and_correct_volume() {
        use num_rational::BigRational;
        use num_traits::ToPrimitive;
        use std::collections::BTreeMap;
        // two unit cubes sharing the x=1 face (a coplanar SHARED-FACE degeneracy)
        let a = box_mesh([0., 0., 0.], [1., 1., 1.]);
        let b = box_mesh([1., 0., 0.], [2., 1., 1.]);
        let arr = arrange(&a, &b);
        let result = boolean_vids(&arr, &a, &b, BoolOp::Union);
        assert!(!result.is_empty(), "union is empty");
        // manifold: every undirected edge used exactly twice (no doubled face)
        let mut edges: BTreeMap<(Vid, Vid), u32> = BTreeMap::new();
        for t in &result {
            for (u, v) in [(t[0], t[1]), (t[1], t[2]), (t[2], t[0])] {
                *edges.entry(if u < v { (u, v) } else { (v, u) }).or_insert(0) += 1;
            }
        }
        assert!(
            edges.values().all(|&c| c == 2),
            "abutting union is non-manifold (the shared x=1 face was not deduped)"
        );
        // volume == 2 (the merged [0,2]×[0,1]×[0,1] box)
        let co = |v: Vid| {
            let p = point_of(arr.interner.get(v));
            [
                BigRational::to_f64(&p[0]).unwrap(),
                BigRational::to_f64(&p[1]).unwrap(),
                BigRational::to_f64(&p[2]).unwrap(),
            ]
        };
        let vol: f64 = result
            .iter()
            .map(|t| dot3(co(t[0]), cross3(co(t[1]), co(t[2]))))
            .sum::<f64>()
            / 6.0;
        assert!((vol - 2.0).abs() < 1e-6, "abutting-boxes union volume = {vol}, expected 2");
    }

    /// Edge-use audit on a materialised f64 triangle list: returns
    /// `(boundary, nonmanifold)` = count of undirected edges used exactly ONCE
    /// (an open boundary) and MORE THAN twice (a non-manifold fin). Both zero ⇒
    /// the surface is closed and 2-manifold. Vertices are keyed on the snap grid.
    fn edge_audit(tris: &[Tri]) -> (usize, usize) {
        use std::collections::BTreeMap;
        let key = |p: [f64; 3]| {
            (
                (p[0] * 65536.0).round() as i64,
                (p[1] * 65536.0).round() as i64,
                (p[2] * 65536.0).round() as i64,
            )
        };
        let mut edges: BTreeMap<((i64, i64, i64), (i64, i64, i64)), u32> = BTreeMap::new();
        for t in tris {
            let k = [key(t[0]), key(t[1]), key(t[2])];
            for (u, v) in [(k[0], k[1]), (k[1], k[2]), (k[2], k[0])] {
                *edges.entry(if u < v { (u, v) } else { (v, u) }).or_insert(0) += 1;
            }
        }
        (
            edges.values().filter(|&&c| c == 1).count(),
            edges.values().filter(|&&c| c > 2).count(),
        )
    }

    fn bounds(tris: &[Tri]) -> ([f64; 3], [f64; 3]) {
        let mut lo = [f64::MAX; 3];
        let mut hi = [f64::MIN; 3];
        for t in tris {
            for v in t {
                for k in 0..3 {
                    lo[k] = lo[k].min(v[k]);
                    hi[k] = hi[k].max(v[k]);
                }
            }
        }
        (lo, hi)
    }

    #[test]
    fn abutting_boxes_union_is_watertight_combined_hull() {
        // The MINIMAL coplanar-union repro (the #960 root cause, distilled): two
        // unit boxes sharing the x=1 face union to ONE watertight 1×1×2 hull —
        // every edge shared by exactly two faces, no open boundary, bounds span
        // BOTH boxes. Pure geometry ⇒ must always pass.
        let a = box_mesh([0., 0., 0.], [1., 1., 1.]);
        let b = box_mesh([1., 0., 0.], [2., 1., 1.]);
        let u = boolean(&a, &b, BoolOp::Union);
        let (boundary, nonmanifold) = edge_audit(&u);
        assert_eq!(boundary, 0, "abutting-boxes union has {boundary} open boundary edges (torn)");
        assert_eq!(nonmanifold, 0, "abutting-boxes union has {nonmanifold} non-manifold edges (the shared x=1 face was not dissolved)");
        let (lo, hi) = bounds(&u);
        assert_eq!(lo, [0., 0., 0.], "union lower bound should span both boxes");
        assert_eq!(hi, [2., 1., 1.], "union upper bound should span both boxes");
    }

    #[test]
    fn partial_coplanar_seam_union_is_watertight() {
        // Two boxes sharing the x=1 plane but with only PARTIAL face overlap in Y
        // (A: y∈[0,2], B: y∈[1,3]); each box's x=1 face extends past the shared
        // band. This is the #960 seam END-REGION case: the seam face of one box
        // reaches where the other has NO coincident face (its solid abuts via a
        // perpendicular wall). The end-region must still dissolve, not tear.
        let a = box_mesh([0., 0., 0.], [1., 2., 1.]);
        let b = box_mesh([1., 1., 0.], [2., 3., 1.]);
        let u = boolean(&a, &b, BoolOp::Union);
        let (boundary, nonmanifold) = edge_audit(&u);
        assert_eq!(boundary, 0, "partial-seam union has {boundary} open boundary edges");
        assert_eq!(nonmanifold, 0, "partial-seam union has {nonmanifold} non-manifold edges");
    }

    #[test]
    fn identical_solids_union_is_the_single_solid() {
        // Union of a box with an exact COPY of itself = that box (every face is a
        // co-oriented coincident duplicate; exactly one copy survives). The
        // sloped/rotated variant is the case the narrow coincident-face dedup
        // missed before the Vid-key dedup.
        let a = box_mesh([0., 0., 0.], [2., 2., 2.]);
        let u = boolean(&a, &a, BoolOp::Union);
        let (boundary, nonmanifold) = edge_audit(&u);
        assert_eq!((boundary, nonmanifold), (0, 0), "self-union is not watertight: b={boundary} nm={nonmanifold}");
        assert_eq!(u.len(), 12, "self-union should be the single 12-triangle box, got {}", u.len());
    }

    #[test]
    fn nary_union_of_abutting_and_duplicate_prisms_is_watertight() {
        // The #960 cutter-union shape (distilled to axis-aligned): a strip of
        // three abutting unit boxes PLUS an exact duplicate of the middle one —
        // unioned in ONE conforming arrangement (`union_all`), never re-meshing an
        // intermediate. Result = the 3×1×1 strip, watertight, duplicate dissolved.
        let b0 = box_mesh([0., 0., 0.], [1., 1., 1.]);
        let b1 = box_mesh([1., 0., 0.], [2., 1., 1.]);
        let b2 = box_mesh([2., 0., 0.], [3., 1., 1.]);
        let b1_dup = b1.clone();
        let meshes: Vec<&[Tri]> = vec![&b0, &b1, &b2, &b1_dup];
        let u = super::union_all(&meshes);
        let (boundary, nonmanifold) = edge_audit(&u);
        assert_eq!(boundary, 0, "N-ary union has {boundary} open boundary edges");
        assert_eq!(nonmanifold, 0, "N-ary union has {nonmanifold} non-manifold edges");
        let (lo, hi) = bounds(&u);
        assert_eq!((lo, hi), ([0., 0., 0.], [3., 1., 1.]), "N-ary union bounds wrong");
    }

    #[test]
    fn boolean_manifest_is_pinned() {
        // The full-boolean topology fingerprint (cube[0,2]³ − cube[1,3]³),
        // byte-identical across x86_64/aarch64/wasm (re-pin + re-run the wasm
        // cross-check if the boolean logic legitimately changes).
        const PINNED: u64 = 0x0465_b83a_5fdb_8b2b;
        let m = super::boolean_manifest();
        assert_eq!(m, PINNED, "boolean topology manifest changed: 0x{m:016x}");
    }
}