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
|
/*
---------------------------------------------------------------------------
Open Asset Import Library (assimp)
---------------------------------------------------------------------------
Copyright (c) 2006-2022, assimp team
All rights reserved.
Redistribution and use of this software in source and binary forms,
with or without modification, are permitted provided that the following
conditions are met:
* Redistributions of source code must retain the above
copyright notice, this list of conditions and the
following disclaimer.
* Redistributions in binary form must reproduce the above
copyright notice, this list of conditions and the
following disclaimer in the documentation and/or other
materials provided with the distribution.
* Neither the name of the assimp team, nor the names of its
contributors may be used to endorse or promote products
derived from this software without specific prior
written permission of the assimp team.
THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
"AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
(INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
---------------------------------------------------------------------------
*/
/** @file TriangulateProcess.cpp
* @brief Implementation of the post processing step to split up
* all faces with more than three indices into triangles.
*
*
* The triangulation algorithm will handle concave or convex polygons.
* Self-intersecting or non-planar polygons are not rejected, but
* they're probably not triangulated correctly.
*
* DEBUG SWITCHES - do not enable any of them in release builds:
*
* AI_BUILD_TRIANGULATE_COLOR_FACE_WINDING
* - generates vertex colors to represent the face winding order.
* the first vertex of a polygon becomes red, the last blue.
* AI_BUILD_TRIANGULATE_DEBUG_POLYS
* - dump all polygons and their triangulation sequences to
* a file
*/
#ifndef ASSIMP_BUILD_NO_TRIANGULATE_PROCESS
#include "PostProcessing/TriangulateProcess.h"
#include "PostProcessing/ProcessHelper.h"
#include "Common/PolyTools.h"
#include <memory>
#include <cstdint>
//#define AI_BUILD_TRIANGULATE_COLOR_FACE_WINDING
//#define AI_BUILD_TRIANGULATE_DEBUG_POLYS
#define POLY_GRID_Y 40
#define POLY_GRID_X 70
#define POLY_GRID_XPAD 20
#define POLY_OUTPUT_FILE "assimp_polygons_debug.txt"
using namespace Assimp;
namespace {
/**
* @brief Helper struct used to simplify NGON encoding functions.
*/
struct NGONEncoder {
NGONEncoder() : mLastNGONFirstIndex((unsigned int)-1) {}
/**
* @brief Encode the current triangle, and make sure it is recognized as a triangle.
*
* This method will rotate indices in tri if needed in order to avoid tri to be considered
* part of the previous ngon. This method is to be used whenever you want to emit a real triangle,
* and make sure it is seen as a triangle.
*
* @param tri Triangle to encode.
*/
void ngonEncodeTriangle(aiFace * tri) {
ai_assert(tri->mNumIndices == 3);
// Rotate indices in new triangle to avoid ngon encoding false ngons
// Otherwise, the new triangle would be considered part of the previous NGON.
if (isConsideredSameAsLastNgon(tri)) {
std::swap(tri->mIndices[0], tri->mIndices[2]);
std::swap(tri->mIndices[1], tri->mIndices[2]);
}
mLastNGONFirstIndex = tri->mIndices[0];
}
/**
* @brief Encode a quad (2 triangles) in ngon encoding, and make sure they are seen as a single ngon.
*
* @param tri1 First quad triangle
* @param tri2 Second quad triangle
*
* @pre Triangles must be properly fanned from the most appropriate vertex.
*/
void ngonEncodeQuad(aiFace *tri1, aiFace *tri2) {
ai_assert(tri1->mNumIndices == 3);
ai_assert(tri2->mNumIndices == 3);
ai_assert(tri1->mIndices[0] == tri2->mIndices[0]);
// If the selected fanning vertex is the same as the previously
// emitted ngon, we use the opposite vertex which also happens to work
// for tri-fanning a concave quad.
// ref: https://github.com/assimp/assimp/pull/3695#issuecomment-805999760
if (isConsideredSameAsLastNgon(tri1)) {
// Right-rotate indices for tri1 (index 2 becomes the new fanning vertex)
std::swap(tri1->mIndices[0], tri1->mIndices[2]);
std::swap(tri1->mIndices[1], tri1->mIndices[2]);
// Left-rotate indices for tri2 (index 2 becomes the new fanning vertex)
std::swap(tri2->mIndices[1], tri2->mIndices[2]);
std::swap(tri2->mIndices[0], tri2->mIndices[2]);
ai_assert(tri1->mIndices[0] == tri2->mIndices[0]);
}
mLastNGONFirstIndex = tri1->mIndices[0];
}
/**
* @brief Check whether this triangle would be considered part of the lastly emitted ngon or not.
*
* @param tri Current triangle.
* @return true If used as is, this triangle will be part of last ngon.
* @return false If used as is, this triangle is not considered part of the last ngon.
*/
bool isConsideredSameAsLastNgon(const aiFace * tri) const {
ai_assert(tri->mNumIndices == 3);
return tri->mIndices[0] == mLastNGONFirstIndex;
}
private:
unsigned int mLastNGONFirstIndex;
};
}
// ------------------------------------------------------------------------------------------------
// Constructor to be privately used by Importer
TriangulateProcess::TriangulateProcess()
{
// nothing to do here
}
// ------------------------------------------------------------------------------------------------
// Destructor, private as well
TriangulateProcess::~TriangulateProcess()
{
// nothing to do here
}
// ------------------------------------------------------------------------------------------------
// Returns whether the processing step is present in the given flag field.
bool TriangulateProcess::IsActive( unsigned int pFlags) const
{
return (pFlags & aiProcess_Triangulate) != 0;
}
// ------------------------------------------------------------------------------------------------
// Executes the post processing step on the given imported data.
void TriangulateProcess::Execute( aiScene* pScene)
{
ASSIMP_LOG_DEBUG("TriangulateProcess begin");
bool bHas = false;
for( unsigned int a = 0; a < pScene->mNumMeshes; a++)
{
if (pScene->mMeshes[ a ]) {
if ( TriangulateMesh( pScene->mMeshes[ a ] ) ) {
bHas = true;
}
}
}
if ( bHas ) {
ASSIMP_LOG_INFO( "TriangulateProcess finished. All polygons have been triangulated." );
} else {
ASSIMP_LOG_DEBUG( "TriangulateProcess finished. There was nothing to be done." );
}
}
// ------------------------------------------------------------------------------------------------
// Triangulates the given mesh.
bool TriangulateProcess::TriangulateMesh( aiMesh* pMesh)
{
// Now we have aiMesh::mPrimitiveTypes, so this is only here for test cases
if (!pMesh->mPrimitiveTypes) {
bool bNeed = false;
for( unsigned int a = 0; a < pMesh->mNumFaces; a++) {
const aiFace& face = pMesh->mFaces[a];
if( face.mNumIndices != 3) {
bNeed = true;
}
}
if (!bNeed)
return false;
}
else if (!(pMesh->mPrimitiveTypes & aiPrimitiveType_POLYGON)) {
return false;
}
// Find out how many output faces we'll get
uint32_t numOut = 0, max_out = 0;
bool get_normals = true;
for( unsigned int a = 0; a < pMesh->mNumFaces; a++) {
aiFace& face = pMesh->mFaces[a];
if (face.mNumIndices <= 4) {
get_normals = false;
}
if( face.mNumIndices <= 3) {
numOut++;
}
else {
numOut += face.mNumIndices-2;
max_out = std::max(max_out,face.mNumIndices);
}
}
// Just another check whether aiMesh::mPrimitiveTypes is correct
ai_assert(numOut != pMesh->mNumFaces);
aiVector3D *nor_out = nullptr;
// if we don't have normals yet, but expect them to be a cheap side
// product of triangulation anyway, allocate storage for them.
if (!pMesh->mNormals && get_normals) {
// XXX need a mechanism to inform the GenVertexNormals process to treat these normals as preprocessed per-face normals
// nor_out = pMesh->mNormals = new aiVector3D[pMesh->mNumVertices];
}
// the output mesh will contain triangles, but no polys anymore
pMesh->mPrimitiveTypes |= aiPrimitiveType_TRIANGLE;
pMesh->mPrimitiveTypes &= ~aiPrimitiveType_POLYGON;
// The mesh becomes NGON encoded now, during the triangulation process.
pMesh->mPrimitiveTypes |= aiPrimitiveType_NGONEncodingFlag;
aiFace* out = new aiFace[numOut](), *curOut = out;
std::vector<aiVector3D> temp_verts3d(max_out+2); /* temporary storage for vertices */
std::vector<aiVector2D> temp_verts(max_out+2);
NGONEncoder ngonEncoder;
// Apply vertex colors to represent the face winding?
#ifdef AI_BUILD_TRIANGULATE_COLOR_FACE_WINDING
if (!pMesh->mColors[0])
pMesh->mColors[0] = new aiColor4D[pMesh->mNumVertices];
else
new(pMesh->mColors[0]) aiColor4D[pMesh->mNumVertices];
aiColor4D* clr = pMesh->mColors[0];
#endif
#ifdef AI_BUILD_TRIANGULATE_DEBUG_POLYS
FILE* fout = fopen(POLY_OUTPUT_FILE,"a");
#endif
const aiVector3D* verts = pMesh->mVertices;
// use std::unique_ptr to avoid slow std::vector<bool> specialiations
std::unique_ptr<bool[]> done(new bool[max_out]);
for( unsigned int a = 0; a < pMesh->mNumFaces; a++) {
aiFace& face = pMesh->mFaces[a];
unsigned int* idx = face.mIndices;
int num = (int)face.mNumIndices, ear = 0, tmp, prev = num-1, next = 0, max = num;
// Apply vertex colors to represent the face winding?
#ifdef AI_BUILD_TRIANGULATE_COLOR_FACE_WINDING
for (unsigned int i = 0; i < face.mNumIndices; ++i) {
aiColor4D& c = clr[idx[i]];
c.r = (i+1) / (float)max;
c.b = 1.f - c.r;
}
#endif
aiFace* const last_face = curOut;
// if it's a simple point,line or triangle: just copy it
if( face.mNumIndices <= 3)
{
aiFace& nface = *curOut++;
nface.mNumIndices = face.mNumIndices;
nface.mIndices = face.mIndices;
face.mIndices = nullptr;
// points and lines don't require ngon encoding (and are not supported either!)
if (nface.mNumIndices == 3) ngonEncoder.ngonEncodeTriangle(&nface);
continue;
}
// optimized code for quadrilaterals
else if ( face.mNumIndices == 4) {
// quads can have at maximum one concave vertex. Determine
// this vertex (if it exists) and start tri-fanning from
// it.
unsigned int start_vertex = 0;
for (unsigned int i = 0; i < 4; ++i) {
const aiVector3D& v0 = verts[face.mIndices[(i+3) % 4]];
const aiVector3D& v1 = verts[face.mIndices[(i+2) % 4]];
const aiVector3D& v2 = verts[face.mIndices[(i+1) % 4]];
const aiVector3D& v = verts[face.mIndices[i]];
aiVector3D left = (v0-v);
aiVector3D diag = (v1-v);
aiVector3D right = (v2-v);
left.Normalize();
diag.Normalize();
right.Normalize();
const float angle = std::acos(left*diag) + std::acos(right*diag);
if (angle > AI_MATH_PI_F) {
// this is the concave point
start_vertex = i;
break;
}
}
const unsigned int temp[] = {face.mIndices[0], face.mIndices[1], face.mIndices[2], face.mIndices[3]};
aiFace& nface = *curOut++;
nface.mNumIndices = 3;
nface.mIndices = face.mIndices;
nface.mIndices[0] = temp[start_vertex];
nface.mIndices[1] = temp[(start_vertex + 1) % 4];
nface.mIndices[2] = temp[(start_vertex + 2) % 4];
aiFace& sface = *curOut++;
sface.mNumIndices = 3;
sface.mIndices = new unsigned int[3];
sface.mIndices[0] = temp[start_vertex];
sface.mIndices[1] = temp[(start_vertex + 2) % 4];
sface.mIndices[2] = temp[(start_vertex + 3) % 4];
// prevent double deletion of the indices field
face.mIndices = nullptr;
ngonEncoder.ngonEncodeQuad(&nface, &sface);
continue;
}
else
{
// A polygon with more than 3 vertices can be either concave or convex.
// Usually everything we're getting is convex and we could easily
// triangulate by tri-fanning. However, LightWave is probably the only
// modeling suite to make extensive use of highly concave, monster polygons ...
// so we need to apply the full 'ear cutting' algorithm to get it right.
// REQUIREMENT: polygon is expected to be simple and *nearly* planar.
// We project it onto a plane to get a 2d triangle.
// Collect all vertices of of the polygon.
for (tmp = 0; tmp < max; ++tmp) {
temp_verts3d[tmp] = verts[idx[tmp]];
}
// Get newell normal of the polygon. Store it for future use if it's a polygon-only mesh
aiVector3D n;
NewellNormal<3,3,3>(n,max,&temp_verts3d.front().x,&temp_verts3d.front().y,&temp_verts3d.front().z);
if (nor_out) {
for (tmp = 0; tmp < max; ++tmp)
nor_out[idx[tmp]] = n;
}
// Select largest normal coordinate to ignore for projection
const float ax = (n.x>0 ? n.x : -n.x);
const float ay = (n.y>0 ? n.y : -n.y);
const float az = (n.z>0 ? n.z : -n.z);
unsigned int ac = 0, bc = 1; /* no z coord. projection to xy */
float inv = n.z;
if (ax > ay) {
if (ax > az) { /* no x coord. projection to yz */
ac = 1; bc = 2;
inv = n.x;
}
}
else if (ay > az) { /* no y coord. projection to zy */
ac = 2; bc = 0;
inv = n.y;
}
// Swap projection axes to take the negated projection vector into account
if (inv < 0.f) {
std::swap(ac,bc);
}
for (tmp =0; tmp < max; ++tmp) {
temp_verts[tmp].x = verts[idx[tmp]][ac];
temp_verts[tmp].y = verts[idx[tmp]][bc];
done[tmp] = false;
}
#ifdef AI_BUILD_TRIANGULATE_DEBUG_POLYS
// plot the plane onto which we mapped the polygon to a 2D ASCII pic
aiVector2D bmin,bmax;
ArrayBounds(&temp_verts[0],max,bmin,bmax);
char grid[POLY_GRID_Y][POLY_GRID_X+POLY_GRID_XPAD];
std::fill_n((char*)grid,POLY_GRID_Y*(POLY_GRID_X+POLY_GRID_XPAD),' ');
for (int i =0; i < max; ++i) {
const aiVector2D& v = (temp_verts[i] - bmin) / (bmax-bmin);
const size_t x = static_cast<size_t>(v.x*(POLY_GRID_X-1)), y = static_cast<size_t>(v.y*(POLY_GRID_Y-1));
char* loc = grid[y]+x;
if (grid[y][x] != ' ') {
for(;*loc != ' '; ++loc);
*loc++ = '_';
}
*(loc+::ai_snprintf(loc, POLY_GRID_XPAD,"%i",i)) = ' ';
}
for(size_t y = 0; y < POLY_GRID_Y; ++y) {
grid[y][POLY_GRID_X+POLY_GRID_XPAD-1] = '\0';
fprintf(fout,"%s\n",grid[y]);
}
fprintf(fout,"\ntriangulation sequence: ");
#endif
//
// FIXME: currently this is the slow O(kn) variant with a worst case
// complexity of O(n^2) (I think). Can be done in O(n).
while (num > 3) {
// Find the next ear of the polygon
int num_found = 0;
for (ear = next;;prev = ear,ear = next) {
// break after we looped two times without a positive match
for (next=ear+1;done[(next>=max?next=0:next)];++next);
if (next < ear) {
if (++num_found == 2) {
break;
}
}
const aiVector2D* pnt1 = &temp_verts[ear],
*pnt0 = &temp_verts[prev],
*pnt2 = &temp_verts[next];
// Must be a convex point. Assuming ccw winding, it must be on the right of the line between p-1 and p+1.
if (OnLeftSideOfLine2D(*pnt0,*pnt2,*pnt1)) {
continue;
}
// and no other point may be contained in this triangle
for ( tmp = 0; tmp < max; ++tmp) {
// We need to compare the actual values because it's possible that multiple indexes in
// the polygon are referring to the same position. concave_polygon.obj is a sample
//
// FIXME: Use 'epsiloned' comparisons instead? Due to numeric inaccuracies in
// PointInTriangle() I'm guessing that it's actually possible to construct
// input data that would cause us to end up with no ears. The problem is,
// which epsilon? If we chose a too large value, we'd get wrong results
const aiVector2D& vtmp = temp_verts[tmp];
if ( vtmp != *pnt1 && vtmp != *pnt2 && vtmp != *pnt0 && PointInTriangle2D(*pnt0,*pnt1,*pnt2,vtmp)) {
break;
}
}
if (tmp != max) {
continue;
}
// this vertex is an ear
break;
}
if (num_found == 2) {
// Due to the 'two ear theorem', every simple polygon with more than three points must
// have 2 'ears'. Here's definitely something wrong ... but we don't give up yet.
//
// Instead we're continuing with the standard tri-fanning algorithm which we'd
// use if we had only convex polygons. That's life.
ASSIMP_LOG_ERROR("Failed to triangulate polygon (no ear found). Probably not a simple polygon?");
#ifdef AI_BUILD_TRIANGULATE_DEBUG_POLYS
fprintf(fout,"critical error here, no ear found! ");
#endif
num = 0;
break;
/*curOut -= (max-num); // undo all previous work
for (tmp = 0; tmp < max-2; ++tmp) {
aiFace& nface = *curOut++;
nface.mNumIndices = 3;
if (!nface.mIndices)
nface.mIndices = new unsigned int[3];
nface.mIndices[0] = 0;
nface.mIndices[1] = tmp+1;
nface.mIndices[2] = tmp+2;
}
num = 0;
break;*/
}
aiFace& nface = *curOut++;
nface.mNumIndices = 3;
if (!nface.mIndices) {
nface.mIndices = new unsigned int[3];
}
// setup indices for the new triangle ...
nface.mIndices[0] = prev;
nface.mIndices[1] = ear;
nface.mIndices[2] = next;
// exclude the ear from most further processing
done[ear] = true;
--num;
}
if (num > 0) {
// We have three indices forming the last 'ear' remaining. Collect them.
aiFace& nface = *curOut++;
nface.mNumIndices = 3;
if (!nface.mIndices) {
nface.mIndices = new unsigned int[3];
}
for (tmp = 0; done[tmp]; ++tmp);
nface.mIndices[0] = tmp;
for (++tmp; done[tmp]; ++tmp);
nface.mIndices[1] = tmp;
for (++tmp; done[tmp]; ++tmp);
nface.mIndices[2] = tmp;
}
}
#ifdef AI_BUILD_TRIANGULATE_DEBUG_POLYS
for(aiFace* f = last_face; f != curOut; ++f) {
unsigned int* i = f->mIndices;
fprintf(fout," (%i %i %i)",i[0],i[1],i[2]);
}
fprintf(fout,"\n*********************************************************************\n");
fflush(fout);
#endif
for(aiFace* f = last_face; f != curOut; ) {
unsigned int* i = f->mIndices;
// drop dumb 0-area triangles - deactivated for now:
//FindDegenerates post processing step can do the same thing
//if (std::fabs(GetArea2D(temp_verts[i[0]],temp_verts[i[1]],temp_verts[i[2]])) < 1e-5f) {
// ASSIMP_LOG_VERBOSE_DEBUG("Dropping triangle with area 0");
// --curOut;
// delete[] f->mIndices;
// f->mIndices = nullptr;
// for(aiFace* ff = f; ff != curOut; ++ff) {
// ff->mNumIndices = (ff+1)->mNumIndices;
// ff->mIndices = (ff+1)->mIndices;
// (ff+1)->mIndices = nullptr;
// }
// continue;
//}
i[0] = idx[i[0]];
i[1] = idx[i[1]];
i[2] = idx[i[2]];
// IMPROVEMENT: Polygons are not supported yet by this ngon encoding + triangulation step.
// So we encode polygons as regular triangles. No way to reconstruct the original
// polygon in this case.
ngonEncoder.ngonEncodeTriangle(f);
++f;
}
delete[] face.mIndices;
face.mIndices = nullptr;
}
#ifdef AI_BUILD_TRIANGULATE_DEBUG_POLYS
fclose(fout);
#endif
// kill the old faces
delete [] pMesh->mFaces;
// ... and store the new ones
pMesh->mFaces = out;
pMesh->mNumFaces = (unsigned int)(curOut-out); /* not necessarily equal to numOut */
return true;
}
#endif // !! ASSIMP_BUILD_NO_TRIANGULATE_PROCESS
|