File: surfaces.cpp

package info (click to toggle)
regina-normal 7.4.1-1.1
  • links: PTS
  • area: main
  • in suites: forky, sid
  • size: 154,244 kB
  • sloc: cpp: 295,026; xml: 9,992; sh: 1,344; python: 1,225; perl: 616; ansic: 138; makefile: 26
file content (616 lines) | stat: -rw-r--r-- 25,096 bytes parent folder | download
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

/**************************************************************************
 *                                                                        *
 *  Regina - A Normal Surface Theory Calculator                           *
 *  Computational Engine                                                  *
 *                                                                        *
 *  Copyright (c) 1999-2025, Ben Burton                                   *
 *  For further details contact Ben Burton (bab@debian.org).              *
 *                                                                        *
 *  This program is free software; you can redistribute it and/or         *
 *  modify it under the terms of the GNU General Public License as        *
 *  published by the Free Software Foundation; either version 2 of the    *
 *  License, or (at your option) any later version.                       *
 *                                                                        *
 *  As an exception, when this program is distributed through (i) the     *
 *  App Store by Apple Inc.; (ii) the Mac App Store by Apple Inc.; or     *
 *  (iii) Google Play by Google Inc., then that store may impose any      *
 *  digital rights management, device limits and/or redistribution        *
 *  restrictions that are required by its terms of service.               *
 *                                                                        *
 *  This program is distributed in the hope that it will be useful, but   *
 *  WITHOUT ANY WARRANTY; without even the implied warranty of            *
 *  MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU     *
 *  General Public License for more details.                              *
 *                                                                        *
 *  You should have received a copy of the GNU General Public License     *
 *  along with this program. If not, see <https://www.gnu.org/licenses/>. *
 *                                                                        *
 **************************************************************************/

#include "enumerate/treetraversal.h"
#include "packet/container.h"
#include "surface/normalsurfaces.h"
#include "triangulation/dim3.h"
#include <stack>

namespace regina {

/**
 * When testing 0-efficiency, to prove that a normal 2-sphere must occur
 * at a vertex we use Euler characteristic arguments.  One issue that
 * arises for non-orientable 3-manifolds is whether a non-vertex normal
 * 2-sphere can be decomposed into two-sided projective planes and other
 * surfaces of non-positive Euler characteristic.  On this issue, Jaco
 * writes:
 *
 *  "Remember that in any 3-manifold, regular curves of intersection between
 *   normal surfaces are orientation preserving; thus if you add a two-sided
 *   projective plane to any other surface, the curves of intersection must
 *   be trivial curves on the projective plane - thus the result must be
 *   nonorientable."
 */

/**
 * In the general case, 0-efficiency must be tested for in standard
 * triangle-quad coordinates.  For example, the triangulation with
 * isosig dLQacccbnjk (which is ideal with one torus cusp) is
 * not 0-efficient, but the non-trivial sphere does not appear as a
 * vertex in quad coordinates.
 */

/**
 * Splitting surfaces must also be tested for in standard triangle-quad
 * coordinates.  See the triangulation J_{1|3,-5} (chained triangular
 * solid torus of major type) of S^3 / Q_32 x Z_3 an an example of a
 * triangulation with a splitting surface having chi=-1 that can be
 * decomposed in quad space as the sum of two vertex normal tori minus a
 * vertex link.
 */

std::optional<NormalSurface> Triangulation<3>::nonTrivialSphereOrDisc() const {
    // Get the empty triangulation out of the way now.
    if (simplices_.empty())
        return std::nullopt;

    // Do we already know the answer?
    if (prop_.zeroEfficient_.has_value() && *prop_.zeroEfficient_)
        return std::nullopt;

    // Use combinatorial optimisation if we can.
    if (isValid() && countVertices() == 1) {
        // For now, just use the safe arbitrary-precision Integer type.
        TreeSingleSoln<LPConstraintEulerPositive> tree(*this,
            NormalCoords::Standard);
        if (tree.find()) {
            NormalSurface s = tree.buildSurface();
            if (! ((! s.hasRealBoundary()) &&
                    (s.eulerChar() == 1) && s.isTwoSided()))
                return s;
            // Looks like we've found a two-sided projective plane.
            // Fall through to a full enumeration of vertex surfaces.
        } else
            return std::nullopt;
    }

    // Fall back to a slow-but-general method: enumerate all vertex surfaces.
    // For valid, non-ideal triangulations we can do this in quad
    // coordinates (where a non-trivial sphere or disc is guaranteed to
    // appear as a vertex surface).  Otherwise fall back to standard coords.
    NormalSurfaces surfaces(*this, (isValid() && ! isIdeal()) ?
        NormalCoords::Quad : NormalCoords::Standard);
    for (const NormalSurface& s : surfaces) {
        // These are vertex surfaces, so we know they must be connected.
        // Because we are either (i) using standard coordinates, or
        // (ii) working with a non-ideal triangulation; we know the
        // vertex surfaces are compact also.

        if (s.isVertexLinking())
            continue;

        // Now they are compact, connected and non-vertex-linking.
        // We just need to pick out spheres and discs.
        if (s.eulerChar() == 2) {
            // Must be a sphere; no bounded surface has chi=2.
            NormalSurface ans = s;
            return ans;
        } else if (s.eulerChar() == 1) {
            if (s.hasRealBoundary()) {
                // Must be a disc.
                return s;
            } else if (! s.isTwoSided()) {
                // A projective plane that doubles to a sphere.
                return s * 2;
            }
        }
    }

    return std::nullopt;
}

std::optional<NormalSurface> Triangulation<3>::octagonalAlmostNormalSphere()
        const {
    // Get the empty triangulation out of the way now.
    if (simplices_.empty())
        return std::nullopt;

    // Use combinatorial optimisation if we can.
    // This is good for large problems, but for small problems a full
    // enumeration is usually faster.  Still, the big problems are the
    // ones we need to be more fussy about.
    if (countVertices() == 1) {
        // For now, just use the safe arbitrary-precision Integer type.
        TreeSingleSoln<LPConstraintEulerPositive> tree(*this,
            NormalCoords::AlmostNormal);
        if (tree.find()) {
            // Since our preconditions ensure the triangulation is
            // closed, orientable and 0-efficient, there are no
            // non-vertex-linking normal surfaces with positive Euler
            // characteristic.  Our optimisation asks for (Euler - #octs) > 0,
            // which then implies that our surface here is almost normal
            // with exactly 1 octagon and Euler = 2.  This is exactly
            // what we're looking for.
            return tree.buildSurface();
        } else
            return std::nullopt;
    }

    // Fall back to a slow-but-general method: enumerate all vertex surfaces.
    // Given our preconditions, we can do this in quadrilateral-octagon
    // coordinates; for details see "Quadrilateral-octagon coordinates for
    // almost normal surfaces", B.B., Experiment. Math. 19 (2010), 285-315.
    NormalSurfaces surfaces(*this, NormalCoords::QuadOct);

    // Our vertex surfaces are guaranteed to be in smallest possible
    // integer coordinates, with at most one non-zero octagonal coordinate.
    unsigned long tet;
    unsigned oct;
    bool found, broken;
    LargeInteger coord;
    for (const NormalSurface& s : surfaces) {
        // These are vertex surfaces, so we know they must be connected.
        // Because we are working with a non-ideal triangulation, we know the
        // vertex surfaces are compact.

        // Hunt for spheres with exactly one octagon.
        // Note that 1-sided projective planes are no good here,
        // since when doubled they give too many octagonal discs.
        if (s.eulerChar() == 2) {
            // Euler char = 2 implies no real boundary.
            found = false; // At least one octagon found so far?
            broken = false; // More than one octagon found so far?
            for (tet = 0; tet < simplices_.size() && ! broken; ++tet)
                for (oct = 0; oct < 3; ++oct) {
                    coord = s.octs(tet, oct);
                    if (coord > 1) {
                        broken = true;
                        break;
                    } else if (coord == 1) {
                        if (found) {
                            broken = true;
                            break;
                        } else
                            found = true;
                    }
                }
            if (found && ! broken) {
                // This is it!
                return s;
            }
        }
    }

    return std::nullopt;
}

bool Triangulation<3>::isZeroEfficient() const {
    if (! prop_.zeroEfficient_.has_value()) {
        if (hasTwoSphereBoundaryComponents())
            prop_.zeroEfficient_ = false;
        else if (nonTrivialSphereOrDisc()) {
            prop_.zeroEfficient_ = false;
        } else {
            prop_.zeroEfficient_ = true;

            // Things implied by 0-efficiency:
            if (isValid() && isClosed() && isConnected())
                prop_.irreducible_ = true;
        }
    }
    return *prop_.zeroEfficient_;
}

bool Triangulation<3>::isOneEfficient() const {
    // Check the preconditions before examining the cached value, since it's
    // possible the 1-efficiency value was cached from a newer calculation
    // engine that supports 1-efficiency testing in more settings.
    if (! isValid()) {
        throw FailedPrecondition(
            "1-efficiency testing requires a valid triangulation");
    }
    if (! isIdeal()) {
        // The empty triangulation is eliminated here.
        throw FailedPrecondition(
            "1-efficiency testing requires an ideal triangulation");
    }
    for (auto v : vertices()) {
        if (v->linkType() != Vertex<3>::Link::Torus &&
                v->linkType() != Vertex<3>::Link::KleinBottle)
            throw FailedPrecondition(
                "1-efficiency testing requires a triangulation whose "
                "vertex links are all tori and/or Klein bottles");
    }

    if (prop_.oneEfficient_.has_value())
        return *prop_.oneEfficient_;

    // We might already know the answer from the 0-efficiency property,
    // since for the settings in which we are able to test 1-efficiency,
    // 0-efficiency is a necessary condition.
    if (prop_.zeroEfficient_.has_value() && ! *prop_.zeroEfficient_)
        return *(prop_.oneEfficient_ = false);

    // For now we just enumerate and test all vertex normal surfaces.
    // In the future we should consider fetching one surface at a time using
    // the tree traversal method, which will allow us to fail early if a bad
    // surface is discovered.

    // TODO: Do this in quad closed coordinates if we can.

    for (const auto& s : NormalSurfaces(*this, NormalCoords::Standard)) {
        auto euler = s.eulerChar();
        if (euler > 0) {
            // Note: due to the preconditions on the triangulation, this
            // surface cannot be vertex linking.
            if (isOrientable() || euler == 2 || ! s.isTwoSided()) {
                // Either this is a non-trivial sphere, or its double cover is.
                // Record the 0-efficiency failure also.
                // Note: the only way this test can _fail_ is if we have a
                // two-sided projective plane in a non-orientable manifold
                // (which does not actually violate our definition of
                // 0-efficiency, but which _does_ violate our definition of
                // 1-efficiency).
                prop_.zeroEfficient_ = false;
            }
            return *(prop_.oneEfficient_ = false);
        } else if (euler == 0 && ! s.isVertexLinking()) {
            return *(prop_.oneEfficient_ = false);
        }
    }

    return *(prop_.oneEfficient_ = true);
}

bool Triangulation<3>::hasSplittingSurface() const {
    if (prop_.splittingSurface_.has_value())
        return *prop_.splittingSurface_;

    if (isEmpty())
        return *(prop_.splittingSurface_ = false);

    // In the main loop of this procedure, we assume the triangulation is
    // connected.  If it isn't connected, we see instead if each component
    // has a splitting surface.

    if (!isConnected()) {
        for (const Triangulation<3>& comp : triangulateComponents())
            if (! comp.hasSplittingSurface())
                return *(prop_.splittingSurface_ = false);
        return *(prop_.splittingSurface_ = true);
    }

    // Now we can assume the triangulation is connected.

    // We keep track of whether an edge has been assumed to be disjoint
    // or not from a putative splitting surface.
    enum class EdgeState {
        Unknown = 0,
        Disjoint = 1,
        Intersecting = 2
    };
    auto* state = new EdgeState[countEdges()];

    // We also keep track of each edge e that is not yet assumed disjoint but
    // that is a candidate for this assumption.
    std::deque<Edge<3>*> candidate_disjoint;

    // At the outset, we may regard any edge as a candidate.
    // We will do so for each of the three edges in a triangle of the
    // triangulation.
    // The triangulation is connected.
    // So these exhaust the possibilities for a splitting surface.
    auto tri = triangle(0);

    for (int i = 0; i < 3; i++){
        candidate_disjoint.clear();
        std::fill(state, state + countEdges(), EdgeState::Unknown);

        // Outset
        candidate_disjoint.push_back(tri->edge(i));

        // Main inner loop
        bool broken = false;
        while (!candidate_disjoint.empty()){
            Edge<3>* e = candidate_disjoint.front();
            candidate_disjoint.pop_front();

            for (auto& emb : *e){
                Tetrahedron<3>* tet_e = emb.simplex();
                Perm<4> v_perm = emb.vertices();

                // The splitting surface should intersect the other edges
                // of the candidate edge in the given embedding meeting
                // the candidate edge---lateral edges.
                Edge<3>* lat02 = tet_e->edge(v_perm[0],v_perm[2]);
                Edge<3>* lat03 = tet_e->edge(v_perm[0],v_perm[3]);
                Edge<3>* lat12 = tet_e->edge(v_perm[1],v_perm[2]);
                Edge<3>* lat13 = tet_e->edge(v_perm[1],v_perm[3]);
                state[lat02->index()] = EdgeState::Intersecting;
                state[lat03->index()] = EdgeState::Intersecting;
                state[lat12->index()] = EdgeState::Intersecting;
                state[lat13->index()] = EdgeState::Intersecting;
            }

            // Now we check for a local obstruction to a splitting surface
            // opposite e.
            if (state[e->index()] == EdgeState::Intersecting) {
                broken = true;
                break;
            } else {
                state[e->index()] = EdgeState::Disjoint;
                // Regard the edges opposite e as candidates if they're not
                // already assumed disjoint.
                for (auto& emb : *e) {
                    Tetrahedron<3>* tet_e = emb.simplex();
                    Perm<4> v_perm = emb.vertices();
                    Edge<3>* opp = tet_e->edge(v_perm[2],v_perm[3]);
                    if (state[opp->index()] == EdgeState::Intersecting) {
                        broken = true;
                        break;
                    }
                    if (state[opp->index()] != EdgeState::Disjoint)
                        candidate_disjoint.push_back(opp);
                }
            }
        } // End main inner loop

        if (! broken) {
            // We partitioned the edges into disjoint and intersecting,
            // with two opposite disjoint edges per tetrahedron.
            // Thus there is a splitting surface.
            delete[] state;
            return *(prop_.splittingSurface_ = true);
        }
    } // End search for splitting surfaces along each edge of tri.

    // We found no splitting surfaces; there is none.
    delete[] state;
    return *(prop_.splittingSurface_ = false);
}

template <int subdim>
std::pair<NormalSurface, bool> Triangulation<3>::linkingSurface(
        const Face<3, subdim>& face) const {
    static_assert(0 <= subdim && subdim < 3,
        "Triangulation<3>::linkingSurface() requires a face of dimension "
        "0, 1 or 2.");

    Vector<LargeInteger> coords(7 * size());
    bool thin = true;

    if constexpr (subdim == 0) {
        // Vertex links are trivial to construct.
        for (auto& emb : face)
            coords[7 * emb.simplex()->index() + emb.vertex()] = 1;
    } else {
        // In general, edge and triangle links can require normalisation.
        //
        // However, this normalisation always involves expanding the face
        // into a larger subcomplex using the following rules:
        // 1) at least two edges of a triangle -> absorb the full triangle;
        // 2) at least two triangles of a tetrahedron -> absorb the full
        //    tetrahedron.
        //
        // The resulting face link is then the frontier of a regular
        // neighbourhood of the resulting subcomplex.
        //
        // So: our main job is to build the subcomplex.
        // We track the subcomplex with an array of booleans for each facial
        // dimension, indicating which of the faces is currently included.

        bool* use0 = new bool[countVertices()];
        bool* use1 = new bool[countEdges()];
        bool* use2 = new bool[countTriangles()];
        bool* use3 = new bool[size()];

        std::fill(use0, use0 + countVertices(), false);
        std::fill(use1, use1 + countEdges(), false);
        std::fill(use2, use2 + countTriangles(), false);
        std::fill(use3, use3 + size(), false);

        if constexpr (subdim == 1) {
            use1[face.index()] = true;
            use0[face.vertex(0)->index()] = true;
            use0[face.vertex(1)->index()] = true;
        } else {
            use2[face.index()] = true;
            for (int i = 0; i < 3; ++i)
                use1[face.edge(i)->index()] = true;
            for (int i = 0; i < 3; ++i)
                use0[face.vertex(i)->index()] = true;
        }

        // Edges/triangles that were recently incorporated into the subcomplex,
        // for which we need to now check for any follow-up triangles/tetrahedra
        // that will need to be incorporated also as a result:
        std::stack<const Edge<3>*> process1;
        std::stack<const Triangle<3>*> process2;

        if constexpr (subdim == 1) {
            process1.push(std::addressof(face));
        } else {
            process2.push(std::addressof(face));
            for (int i = 0; i < 3; ++i)
                process1.push(face.edge(i));
        }

        while (true) {
            if (! process2.empty()) {
                const Triangle<3>* t = process2.top();
                process2.pop();

                for (const auto& emb : *t) {
                    const Tetrahedron<3>* tet = emb.tetrahedron();
                    if (use3[tet->index()])
                        continue;

                    int found = 0;
                    for (int j = 0; j < 4; ++j)
                        if (use2[tet->triangle(j)->index()]) {
                            ++found;
                            if (found == 2)
                                break;
                        }

                    if (found == 2) {
                        // Absorb the entire tetrahedron.
                        thin = false;
                        use3[tet->index()] = true;

                        for (int j = 0; j < 4; ++j) {
                            const Triangle<3>* next = tet->triangle(j);
                            if (! use2[next->index()]) {
                                use2[next->index()] = true;
                                process2.push(next);
                            }
                        }

                        for (int j = 0; j < 6; ++j) {
                            const Edge<3>* next = tet->edge(j);
                            if (! use1[next->index()]) {
                                use1[next->index()] = true;
                                process1.push(next);
                            }
                        }
                    }
                }

                continue;
            }

            if (! process1.empty()) {
                const Edge<3>* e = process1.top();
                process1.pop();

                bool bothSides = e->isBoundary();
                for (const auto& emb : *e) {
                    for (int side = 0; side < 2; ++side) {
                        if (side == 1) {
                            if (! bothSides)
                                continue; // side 0 is enough

                            // We do both sides for this embedding (which is
                            // the first in the list), but *only* this
                            // embedding.
                            bothSides = false;
                        }

                        const Triangle<3>* triangle = emb.tetrahedron()->
                            triangle(emb.vertices()[side == 0 ? 2 : 3]);
                        if (use2[triangle->index()])
                            continue;

                        int found = 0;
                        for (int j = 0; j < 3; ++j)
                            if (use1[triangle->edge(j)->index()]) {
                                ++found;
                                if (found == 2)
                                    break;
                            }

                        if (found == 2) {
                            // Absorb the entire triangle.
                            thin = false;
                            use2[triangle->index()] = true;
                            process2.push(triangle);

                            for (int j = 0; j < 3; ++j) {
                                const Edge<3>* next = triangle->edge(j);
                                if (! use1[next->index()]) {
                                    use1[next->index()] = true;
                                    process1.push(next);
                                }
                            }
                        }
                    }
                }

                continue;
            }

            break;
        }

        size_t tetIndex = 0;
        for (const Tetrahedron<3>* tet : tetrahedra()) {
            if (use3[tetIndex])
                goto doneTet;

            for (int j = 0; j < 4; ++j)
                if (use2[tet->triangle(j)->index()]) {
                    if (use0[tet->vertex(j)->index()])
                        coords[7 * tetIndex + j] = 2;
                    else
                        coords[7 * tetIndex + j] = 1;

                    goto doneTet;
                }

            for (int j = 0; j < 6; ++j)
                if (use1[tet->edge(j)->index()]) {
                    // Note: quad type i does not intersect edge i.
                    int quad = (j < 3 ? j : 5 - j);
                    if (use1[tet->edge(5 - j)->index()])
                        coords[7 * tetIndex + 4 + quad] = 2;
                    else {
                        coords[7 * tetIndex + 4 + quad] = 1;

                        for (int k = 0; k < 2; ++k) {
                            int v = Edge<3>::edgeVertex[5 - j][k];
                            if (use0[tet->vertex(v)->index()])
                                coords[7 * tetIndex + v] = 1;
                        }
                    }

                    goto doneTet;
                }

            for (int j = 0; j < 4; ++j)
                if (use0[tet->vertex(j)->index()])
                    coords[7 * tetIndex + j] = 1;

doneTet:
            ++tetIndex;
        }

        delete[] use0;
        delete[] use1;
        delete[] use2;
        delete[] use3;
    }

    return { NormalSurface(*this, NormalCoords::Standard, std::move(coords)),
        thin };
}

// Instantiate linkingSurface() for all possible template arguments,
// so that the implementation can stay out of the headers.
template std::pair<NormalSurface, bool> Triangulation<3>::linkingSurface<0>(
    const Face<3, 0>&) const;
template std::pair<NormalSurface, bool> Triangulation<3>::linkingSurface<1>(
    const Face<3, 1>&) const;
template std::pair<NormalSurface, bool> Triangulation<3>::linkingSurface<2>(
    const Face<3, 2>&) const;

} // namespace regina