aboutsummaryrefslogtreecommitdiffstats
path: root/src/glu/sgi/libtess/geom.c
blob: 7d3b8d8a6a57abfb4811ead281cffbfc178863fa (plain)
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
/*
 * SGI FREE SOFTWARE LICENSE B (Version 2.0, Sept. 18, 2008)
 * Copyright (C) 1991-2000 Silicon Graphics, Inc. All Rights Reserved.
 *
 * Permission is hereby granted, free of charge, to any person obtaining a
 * copy of this software and associated documentation files (the "Software"),
 * to deal in the Software without restriction, including without limitation
 * the rights to use, copy, modify, merge, publish, distribute, sublicense,
 * and/or sell copies of the Software, and to permit persons to whom the
 * Software is furnished to do so, subject to the following conditions:
 *
 * The above copyright notice including the dates of first publication and
 * either this permission notice or a reference to
 * http://oss.sgi.com/projects/FreeB/
 * shall be included in all copies or substantial portions of the Software.
 *
 * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS
 * OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
 * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
 * SILICON GRAPHICS, INC. BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY,
 * WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF
 * OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
 * SOFTWARE.
 *
 * Except as contained in this notice, the name of Silicon Graphics, Inc.
 * shall not be used in advertising or otherwise to promote the sale, use or
 * other dealings in this Software without prior written authorization from
 * Silicon Graphics, Inc.
 */
/*
** Author: Eric Veach, July 1994.
**
*/

#include "gluos.h"
#include <assert.h>
#include "mesh.h"
#include "geom.h"

int __gl_vertLeq( GLUvertex *u, GLUvertex *v )
{
  /* Returns TRUE if u is lexicographically <= v. */

  return VertLeq( u, v );
}

GLdouble __gl_edgeEval( GLUvertex *u, GLUvertex *v, GLUvertex *w )
{
  /* Given three vertices u,v,w such that VertLeq(u,v) && VertLeq(v,w),
   * evaluates the t-coord of the edge uw at the s-coord of the vertex v.
   * Returns v->t - (uw)(v->s), ie. the signed distance from uw to v.
   * If uw is vertical (and thus passes thru v), the result is zero.
   *
   * The calculation is extremely accurate and stable, even when v
   * is very close to u or w.  In particular if we set v->t = 0 and
   * let r be the negated result (this evaluates (uw)(v->s)), then
   * r is guaranteed to satisfy MIN(u->t,w->t) <= r <= MAX(u->t,w->t).
   */
  GLdouble gapL, gapR;

  assert( VertLeq( u, v ) && VertLeq( v, w ));
  
  gapL = v->s - u->s;
  gapR = w->s - v->s;

  if( gapL + gapR > 0 ) {
    if( gapL < gapR ) {
      return (v->t - u->t) + (u->t - w->t) * (gapL / (gapL + gapR));
    } else {
      return (v->t - w->t) + (w->t - u->t) * (gapR / (gapL + gapR));
    }
  }
  /* vertical line */
  return 0;
}

GLdouble __gl_edgeSign( GLUvertex *u, GLUvertex *v, GLUvertex *w )
{
  /* Returns a number whose sign matches EdgeEval(u,v,w) but which
   * is cheaper to evaluate.  Returns > 0, == 0 , or < 0
   * as v is above, on, or below the edge uw.
   */
  GLdouble gapL, gapR;

  assert( VertLeq( u, v ) && VertLeq( v, w ));
  
  gapL = v->s - u->s;
  gapR = w->s - v->s;

  if( gapL + gapR > 0 ) {
    return (v->t - w->t) * gapL + (v->t - u->t) * gapR;
  }
  /* vertical line */
  return 0;
}


/***********************************************************************
 * Define versions of EdgeSign, EdgeEval with s and t transposed.
 */

GLdouble __gl_transEval( GLUvertex *u, GLUvertex *v, GLUvertex *w )
{
  /* Given three vertices u,v,w such that TransLeq(u,v) && TransLeq(v,w),
   * evaluates the t-coord of the edge uw at the s-coord of the vertex v.
   * Returns v->s - (uw)(v->t), ie. the signed distance from uw to v.
   * If uw is vertical (and thus passes thru v), the result is zero.
   *
   * The calculation is extremely accurate and stable, even when v
   * is very close to u or w.  In particular if we set v->s = 0 and
   * let r be the negated result (this evaluates (uw)(v->t)), then
   * r is guaranteed to satisfy MIN(u->s,w->s) <= r <= MAX(u->s,w->s).
   */
  GLdouble gapL, gapR;

  assert( TransLeq( u, v ) && TransLeq( v, w ));
  
  gapL = v->t - u->t;
  gapR = w->t - v->t;

  if( gapL + gapR > 0 ) {
    if( gapL < gapR ) {
      return (v->s - u->s) + (u->s - w->s) * (gapL / (gapL + gapR));
    } else {
      return (v->s - w->s) + (w->s - u->s) * (gapR / (gapL + gapR));
    }
  }
  /* vertical line */
  return 0;
}

GLdouble __gl_transSign( GLUvertex *u, GLUvertex *v, GLUvertex *w )
{
  /* Returns a number whose sign matches TransEval(u,v,w) but which
   * is cheaper to evaluate.  Returns > 0, == 0 , or < 0
   * as v is above, on, or below the edge uw.
   */
  GLdouble gapL, gapR;

  assert( TransLeq( u, v ) && TransLeq( v, w ));
  
  gapL = v->t - u->t;
  gapR = w->t - v->t;

  if( gapL + gapR > 0 ) {
    return (v->s - w->s) * gapL + (v->s - u->s) * gapR;
  }
  /* vertical line */
  return 0;
}


int __gl_vertCCW( GLUvertex *u, GLUvertex *v, GLUvertex *w )
{
  /* For almost-degenerate situations, the results are not reliable.
   * Unless the floating-point arithmetic can be performed without
   * rounding errors, *any* implementation will give incorrect results
   * on some degenerate inputs, so the client must have some way to
   * handle this situation.
   */
  return (u->s*(v->t - w->t) + v->s*(w->t - u->t) + w->s*(u->t - v->t)) >= 0;
}

/* Given parameters a,x,b,y returns the value (b*x+a*y)/(a+b),
 * or (x+y)/2 if a==b==0.  It requires that a,b >= 0, and enforces
 * this in the rare case that one argument is slightly negative.
 * The implementation is extremely stable numerically.
 * In particular it guarantees that the result r satisfies
 * MIN(x,y) <= r <= MAX(x,y), and the results are very accurate
 * even when a and b differ greatly in magnitude.
 */
#define RealInterpolate(a,x,b,y)			\
  (a = (a < 0) ? 0 : a, b = (b < 0) ? 0 : b,		\
  ((a <= b) ? ((b == 0) ? ((x+y) / 2)			\
                        : (x + (y-x) * (a/(a+b))))	\
            : (y + (x-y) * (b/(a+b)))))

#ifndef FOR_TRITE_TEST_PROGRAM
#define Interpolate(a,x,b,y)	RealInterpolate(a,x,b,y)
#else

/* Claim: the ONLY property the sweep algorithm relies on is that
 * MIN(x,y) <= r <= MAX(x,y).  This is a nasty way to test that.
 */
#include <stdlib.h>
extern int RandomInterpolate;

GLdouble Interpolate( GLdouble a, GLdouble x, GLdouble b, GLdouble y)
{
printf("*********************%d\n",RandomInterpolate);
  if( RandomInterpolate ) {
    a = 1.2 * drand48() - 0.1;
    a = (a < 0) ? 0 : ((a > 1) ? 1 : a);
    b = 1.0 - a;
  }
  return RealInterpolate(a,x,b,y);
}

#endif

#define Swap(a,b)	if (1) { GLUvertex *t = a; a = b; b = t; } else

void __gl_edgeIntersect( GLUvertex *o1, GLUvertex *d1,
			 GLUvertex *o2, GLUvertex *d2,
			 GLUvertex *v )
/* Given edges (o1,d1) and (o2,d2), compute their point of intersection.
 * The computed point is guaranteed to lie in the intersection of the
 * bounding rectangles defined by each edge.
 */
{
  GLdouble z1, z2;

  /* This is certainly not the most efficient way to find the intersection
   * of two line segments, but it is very numerically stable.
   *
   * Strategy: find the two middle vertices in the VertLeq ordering,
   * and interpolate the intersection s-value from these.  Then repeat
   * using the TransLeq ordering to find the intersection t-value.
   */

  if( ! VertLeq( o1, d1 )) { Swap( o1, d1 ); }
  if( ! VertLeq( o2, d2 )) { Swap( o2, d2 ); }
  if( ! VertLeq( o1, o2 )) { Swap( o1, o2 ); Swap( d1, d2 ); }

  if( ! VertLeq( o2, d1 )) {
    /* Technically, no intersection -- do our best */
    v->s = (o2->s + d1->s) / 2;
  } else if( VertLeq( d1, d2 )) {
    /* Interpolate between o2 and d1 */
    z1 = EdgeEval( o1, o2, d1 );
    z2 = EdgeEval( o2, d1, d2 );
    if( z1+z2 < 0 ) { z1 = -z1; z2 = -z2; }
    v->s = Interpolate( z1, o2->s, z2, d1->s );
  } else {
    /* Interpolate between o2 and d2 */
    z1 = EdgeSign( o1, o2, d1 );
    z2 = -EdgeSign( o1, d2, d1 );
    if( z1+z2 < 0 ) { z1 = -z1; z2 = -z2; }
    v->s = Interpolate( z1, o2->s, z2, d2->s );
  }

  /* Now repeat the process for t */

  if( ! TransLeq( o1, d1 )) { Swap( o1, d1 ); }
  if( ! TransLeq( o2, d2 )) { Swap( o2, d2 ); }
  if( ! TransLeq( o1, o2 )) { Swap( o1, o2 ); Swap( d1, d2 ); }

  if( ! TransLeq( o2, d1 )) {
    /* Technically, no intersection -- do our best */
    v->t = (o2->t + d1->t) / 2;
  } else if( TransLeq( d1, d2 )) {
    /* Interpolate between o2 and d1 */
    z1 = TransEval( o1, o2, d1 );
    z2 = TransEval( o2, d1, d2 );
    if( z1+z2 < 0 ) { z1 = -z1; z2 = -z2; }
    v->t = Interpolate( z1, o2->t, z2, d1->t );
  } else {
    /* Interpolate between o2 and d2 */
    z1 = TransSign( o1, o2, d1 );
    z2 = -TransSign( o1, d2, d1 );
    if( z1+z2 < 0 ) { z1 = -z1; z2 = -z2; }
    v->t = Interpolate( z1, o2->t, z2, d2->t );
  }
}