mirror of
https://github.com/UberGames/GtkRadiant.git
synced 2024-11-26 13:51:38 +00:00
780 lines
17 KiB
C
780 lines
17 KiB
C
/*
|
|
Copyright (C) 1999-2007 id Software, Inc. and contributors.
|
|
For a list of contributors, see the accompanying CONTRIBUTORS file.
|
|
|
|
This file is part of GtkRadiant.
|
|
|
|
GtkRadiant 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.
|
|
|
|
GtkRadiant 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 GtkRadiant; if not, write to the Free Software
|
|
Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA
|
|
*/
|
|
|
|
// mathlib.c -- math primitives
|
|
#include "mathlib.h"
|
|
// we use memcpy and memset
|
|
#include <memory.h>
|
|
|
|
vec3_t vec3_origin = {0.0f,0.0f,0.0f};
|
|
|
|
/*
|
|
================
|
|
VectorIsOnAxis
|
|
================
|
|
*/
|
|
qboolean VectorIsOnAxis( vec3_t v ){
|
|
int i, zeroComponentCount;
|
|
|
|
zeroComponentCount = 0;
|
|
for ( i = 0; i < 3; i++ )
|
|
{
|
|
if ( v[i] == 0.0 ) {
|
|
zeroComponentCount++;
|
|
}
|
|
}
|
|
|
|
if ( zeroComponentCount > 1 ) {
|
|
// The zero vector will be on axis.
|
|
return qtrue;
|
|
}
|
|
|
|
return qfalse;
|
|
}
|
|
|
|
/*
|
|
================
|
|
VectorIsOnAxialPlane
|
|
================
|
|
*/
|
|
qboolean VectorIsOnAxialPlane( vec3_t v ){
|
|
int i;
|
|
|
|
for ( i = 0; i < 3; i++ )
|
|
{
|
|
if ( v[i] == 0.0 ) {
|
|
// The zero vector will be on axial plane.
|
|
return qtrue;
|
|
}
|
|
}
|
|
|
|
return qfalse;
|
|
}
|
|
|
|
/*
|
|
================
|
|
MakeNormalVectors
|
|
|
|
Given a normalized forward vector, create two
|
|
other perpendicular vectors
|
|
================
|
|
*/
|
|
void MakeNormalVectors( vec3_t forward, vec3_t right, vec3_t up ){
|
|
float d;
|
|
|
|
// this rotate and negate guarantees a vector
|
|
// not colinear with the original
|
|
right[1] = -forward[0];
|
|
right[2] = forward[1];
|
|
right[0] = forward[2];
|
|
|
|
d = DotProduct( right, forward );
|
|
VectorMA( right, -d, forward, right );
|
|
VectorNormalize( right, right );
|
|
CrossProduct( right, forward, up );
|
|
}
|
|
|
|
vec_t VectorLength( vec3_t v ){
|
|
int i;
|
|
float length;
|
|
|
|
length = 0.0f;
|
|
for ( i = 0 ; i < 3 ; i++ )
|
|
length += v[i] * v[i];
|
|
length = (float)sqrt( length );
|
|
|
|
return length;
|
|
}
|
|
|
|
qboolean VectorCompare( vec3_t v1, vec3_t v2 ){
|
|
int i;
|
|
|
|
for ( i = 0 ; i < 3 ; i++ )
|
|
if ( fabs( v1[i] - v2[i] ) > EQUAL_EPSILON ) {
|
|
return qfalse;
|
|
}
|
|
|
|
return qtrue;
|
|
}
|
|
|
|
/*
|
|
// FIXME TTimo this implementation has to be particular to radiant
|
|
// through another name I'd say
|
|
vec_t Q_rint (vec_t in)
|
|
{
|
|
if (g_PrefsDlg.m_bNoClamp)
|
|
return in;
|
|
else
|
|
return (float)floor (in + 0.5);
|
|
}
|
|
*/
|
|
|
|
void VectorMA( const vec3_t va, vec_t scale, const vec3_t vb, vec3_t vc ){
|
|
vc[0] = va[0] + scale * vb[0];
|
|
vc[1] = va[1] + scale * vb[1];
|
|
vc[2] = va[2] + scale * vb[2];
|
|
}
|
|
|
|
void _CrossProduct( vec3_t v1, vec3_t v2, vec3_t cross ){
|
|
cross[0] = v1[1] * v2[2] - v1[2] * v2[1];
|
|
cross[1] = v1[2] * v2[0] - v1[0] * v2[2];
|
|
cross[2] = v1[0] * v2[1] - v1[1] * v2[0];
|
|
}
|
|
|
|
vec_t _DotProduct( vec3_t v1, vec3_t v2 ){
|
|
return v1[0] * v2[0] + v1[1] * v2[1] + v1[2] * v2[2];
|
|
}
|
|
|
|
void _VectorSubtract( vec3_t va, vec3_t vb, vec3_t out ){
|
|
out[0] = va[0] - vb[0];
|
|
out[1] = va[1] - vb[1];
|
|
out[2] = va[2] - vb[2];
|
|
}
|
|
|
|
void _VectorAdd( vec3_t va, vec3_t vb, vec3_t out ){
|
|
out[0] = va[0] + vb[0];
|
|
out[1] = va[1] + vb[1];
|
|
out[2] = va[2] + vb[2];
|
|
}
|
|
|
|
void _VectorCopy( vec3_t in, vec3_t out ){
|
|
out[0] = in[0];
|
|
out[1] = in[1];
|
|
out[2] = in[2];
|
|
}
|
|
|
|
vec_t VectorNormalize( const vec3_t in, vec3_t out ) {
|
|
|
|
#if MATHLIB_VECTOR_NORMALIZE_PRECISION_FIX
|
|
|
|
// The sqrt() function takes double as an input and returns double as an
|
|
// output according the the man pages on Debian and on FreeBSD. Therefore,
|
|
// I don't see a reason why using a double outright (instead of using the
|
|
// vec_accu_t alias for example) could possibly be frowned upon.
|
|
|
|
double x, y, z, length;
|
|
|
|
x = (double) in[0];
|
|
y = (double) in[1];
|
|
z = (double) in[2];
|
|
|
|
length = sqrt( ( x * x ) + ( y * y ) + ( z * z ) );
|
|
if ( length == 0 ) {
|
|
VectorClear( out );
|
|
return 0;
|
|
}
|
|
|
|
out[0] = (vec_t) ( x / length );
|
|
out[1] = (vec_t) ( y / length );
|
|
out[2] = (vec_t) ( z / length );
|
|
|
|
return (vec_t) length;
|
|
|
|
#else
|
|
|
|
vec_t length, ilength;
|
|
|
|
length = (vec_t)sqrt( in[0] * in[0] + in[1] * in[1] + in[2] * in[2] );
|
|
if ( length == 0 ) {
|
|
VectorClear( out );
|
|
return 0;
|
|
}
|
|
|
|
ilength = 1.0f / length;
|
|
out[0] = in[0] * ilength;
|
|
out[1] = in[1] * ilength;
|
|
out[2] = in[2] * ilength;
|
|
|
|
return length;
|
|
|
|
#endif
|
|
|
|
}
|
|
|
|
vec_t ColorNormalize( const vec3_t in, vec3_t out ) {
|
|
float max, scale;
|
|
|
|
max = in[0];
|
|
if ( in[1] > max ) {
|
|
max = in[1];
|
|
}
|
|
if ( in[2] > max ) {
|
|
max = in[2];
|
|
}
|
|
|
|
if ( max == 0 ) {
|
|
out[0] = out[1] = out[2] = 1.0;
|
|
return 0;
|
|
}
|
|
|
|
scale = 1.0f / max;
|
|
|
|
VectorScale( in, scale, out );
|
|
|
|
return max;
|
|
}
|
|
|
|
void VectorInverse( vec3_t v ){
|
|
v[0] = -v[0];
|
|
v[1] = -v[1];
|
|
v[2] = -v[2];
|
|
}
|
|
|
|
/*
|
|
void VectorScale (vec3_t v, vec_t scale, vec3_t out)
|
|
{
|
|
out[0] = v[0] * scale;
|
|
out[1] = v[1] * scale;
|
|
out[2] = v[2] * scale;
|
|
}
|
|
*/
|
|
|
|
void VectorRotate( vec3_t vIn, vec3_t vRotation, vec3_t out ){
|
|
vec3_t vWork, va;
|
|
int nIndex[3][2];
|
|
int i;
|
|
|
|
VectorCopy( vIn, va );
|
|
VectorCopy( va, vWork );
|
|
nIndex[0][0] = 1; nIndex[0][1] = 2;
|
|
nIndex[1][0] = 2; nIndex[1][1] = 0;
|
|
nIndex[2][0] = 0; nIndex[2][1] = 1;
|
|
|
|
for ( i = 0; i < 3; i++ )
|
|
{
|
|
if ( vRotation[i] != 0 ) {
|
|
float dAngle = vRotation[i] * Q_PI / 180.0f;
|
|
float c = (vec_t)cos( dAngle );
|
|
float s = (vec_t)sin( dAngle );
|
|
vWork[nIndex[i][0]] = va[nIndex[i][0]] * c - va[nIndex[i][1]] * s;
|
|
vWork[nIndex[i][1]] = va[nIndex[i][0]] * s + va[nIndex[i][1]] * c;
|
|
}
|
|
VectorCopy( vWork, va );
|
|
}
|
|
VectorCopy( vWork, out );
|
|
}
|
|
|
|
void VectorRotateOrigin( vec3_t vIn, vec3_t vRotation, vec3_t vOrigin, vec3_t out ){
|
|
vec3_t vTemp, vTemp2;
|
|
|
|
VectorSubtract( vIn, vOrigin, vTemp );
|
|
VectorRotate( vTemp, vRotation, vTemp2 );
|
|
VectorAdd( vTemp2, vOrigin, out );
|
|
}
|
|
|
|
void VectorPolar( vec3_t v, float radius, float theta, float phi ){
|
|
v[0] = (float)( radius * cos( theta ) * cos( phi ) );
|
|
v[1] = (float)( radius * sin( theta ) * cos( phi ) );
|
|
v[2] = (float)( radius * sin( phi ) );
|
|
}
|
|
|
|
void VectorSnap( vec3_t v ){
|
|
int i;
|
|
for ( i = 0; i < 3; i++ )
|
|
{
|
|
v[i] = (vec_t)floor( v[i] + 0.5 );
|
|
}
|
|
}
|
|
|
|
void VectorISnap( vec3_t point, int snap ){
|
|
int i;
|
|
for ( i = 0 ; i < 3 ; i++ )
|
|
{
|
|
point[i] = (vec_t)floor( point[i] / snap + 0.5 ) * snap;
|
|
}
|
|
}
|
|
|
|
void VectorFSnap( vec3_t point, float snap ){
|
|
int i;
|
|
for ( i = 0 ; i < 3 ; i++ )
|
|
{
|
|
point[i] = (vec_t)floor( point[i] / snap + 0.5 ) * snap;
|
|
}
|
|
}
|
|
|
|
void _Vector5Add( vec5_t va, vec5_t vb, vec5_t out ){
|
|
out[0] = va[0] + vb[0];
|
|
out[1] = va[1] + vb[1];
|
|
out[2] = va[2] + vb[2];
|
|
out[3] = va[3] + vb[3];
|
|
out[4] = va[4] + vb[4];
|
|
}
|
|
|
|
void _Vector5Scale( vec5_t v, vec_t scale, vec5_t out ){
|
|
out[0] = v[0] * scale;
|
|
out[1] = v[1] * scale;
|
|
out[2] = v[2] * scale;
|
|
out[3] = v[3] * scale;
|
|
out[4] = v[4] * scale;
|
|
}
|
|
|
|
void _Vector53Copy( vec5_t in, vec3_t out ){
|
|
out[0] = in[0];
|
|
out[1] = in[1];
|
|
out[2] = in[2];
|
|
}
|
|
|
|
// NOTE: added these from Ritual's Q3Radiant
|
|
void ClearBounds( vec3_t mins, vec3_t maxs ){
|
|
mins[0] = mins[1] = mins[2] = 99999;
|
|
maxs[0] = maxs[1] = maxs[2] = -99999;
|
|
}
|
|
|
|
void AddPointToBounds( vec3_t v, vec3_t mins, vec3_t maxs ){
|
|
int i;
|
|
vec_t val;
|
|
|
|
for ( i = 0 ; i < 3 ; i++ )
|
|
{
|
|
val = v[i];
|
|
if ( val < mins[i] ) {
|
|
mins[i] = val;
|
|
}
|
|
if ( val > maxs[i] ) {
|
|
maxs[i] = val;
|
|
}
|
|
}
|
|
}
|
|
|
|
#define PITCH 0 // up / down
|
|
#define YAW 1 // left / right
|
|
#define ROLL 2 // fall over
|
|
#ifndef M_PI
|
|
#define M_PI 3.14159265358979323846f // matches value in gcc v2 math.h
|
|
#endif
|
|
|
|
void AngleVectors( vec3_t angles, vec3_t forward, vec3_t right, vec3_t up ){
|
|
float angle;
|
|
static float sr, sp, sy, cr, cp, cy;
|
|
// static to help MS compiler fp bugs
|
|
|
|
angle = angles[YAW] * ( M_PI * 2.0f / 360.0f );
|
|
sy = (vec_t)sin( angle );
|
|
cy = (vec_t)cos( angle );
|
|
angle = angles[PITCH] * ( M_PI * 2.0f / 360.0f );
|
|
sp = (vec_t)sin( angle );
|
|
cp = (vec_t)cos( angle );
|
|
angle = angles[ROLL] * ( M_PI * 2.0f / 360.0f );
|
|
sr = (vec_t)sin( angle );
|
|
cr = (vec_t)cos( angle );
|
|
|
|
if ( forward ) {
|
|
forward[0] = cp * cy;
|
|
forward[1] = cp * sy;
|
|
forward[2] = -sp;
|
|
}
|
|
if ( right ) {
|
|
right[0] = -sr * sp * cy + cr * sy;
|
|
right[1] = -sr * sp * sy - cr * cy;
|
|
right[2] = -sr * cp;
|
|
}
|
|
if ( up ) {
|
|
up[0] = cr * sp * cy + sr * sy;
|
|
up[1] = cr * sp * sy - sr * cy;
|
|
up[2] = cr * cp;
|
|
}
|
|
}
|
|
|
|
void VectorToAngles( vec3_t vec, vec3_t angles ){
|
|
float forward;
|
|
float yaw, pitch;
|
|
|
|
if ( ( vec[ 0 ] == 0 ) && ( vec[ 1 ] == 0 ) ) {
|
|
yaw = 0;
|
|
if ( vec[ 2 ] > 0 ) {
|
|
pitch = 90;
|
|
}
|
|
else
|
|
{
|
|
pitch = 270;
|
|
}
|
|
}
|
|
else
|
|
{
|
|
yaw = (vec_t)atan2( vec[ 1 ], vec[ 0 ] ) * 180 / M_PI;
|
|
if ( yaw < 0 ) {
|
|
yaw += 360;
|
|
}
|
|
|
|
forward = ( float )sqrt( vec[ 0 ] * vec[ 0 ] + vec[ 1 ] * vec[ 1 ] );
|
|
pitch = (vec_t)atan2( vec[ 2 ], forward ) * 180 / M_PI;
|
|
if ( pitch < 0 ) {
|
|
pitch += 360;
|
|
}
|
|
}
|
|
|
|
angles[ 0 ] = pitch;
|
|
angles[ 1 ] = yaw;
|
|
angles[ 2 ] = 0;
|
|
}
|
|
|
|
/*
|
|
=====================
|
|
PlaneFromPoints
|
|
|
|
Returns false if the triangle is degenrate.
|
|
The normal will point out of the clock for clockwise ordered points
|
|
=====================
|
|
*/
|
|
qboolean PlaneFromPoints( vec4_t plane, const vec3_t a, const vec3_t b, const vec3_t c ) {
|
|
vec3_t d1, d2;
|
|
|
|
VectorSubtract( b, a, d1 );
|
|
VectorSubtract( c, a, d2 );
|
|
CrossProduct( d2, d1, plane );
|
|
if ( VectorNormalize( plane, plane ) == 0 ) {
|
|
return qfalse;
|
|
}
|
|
|
|
plane[3] = DotProduct( a, plane );
|
|
return qtrue;
|
|
}
|
|
|
|
/*
|
|
** NormalToLatLong
|
|
**
|
|
** We use two byte encoded normals in some space critical applications.
|
|
** Lat = 0 at (1,0,0) to 360 (-1,0,0), encoded in 8-bit sine table format
|
|
** Lng = 0 at (0,0,1) to 180 (0,0,-1), encoded in 8-bit sine table format
|
|
**
|
|
*/
|
|
void NormalToLatLong( const vec3_t normal, byte bytes[2] ) {
|
|
// check for singularities
|
|
if ( normal[0] == 0 && normal[1] == 0 ) {
|
|
if ( normal[2] > 0 ) {
|
|
bytes[0] = 0;
|
|
bytes[1] = 0; // lat = 0, long = 0
|
|
}
|
|
else {
|
|
bytes[0] = 128;
|
|
bytes[1] = 0; // lat = 0, long = 128
|
|
}
|
|
}
|
|
else {
|
|
int a, b;
|
|
|
|
a = (int)( RAD2DEG( atan2( normal[1], normal[0] ) ) * ( 255.0f / 360.0f ) );
|
|
a &= 0xff;
|
|
|
|
b = (int)( RAD2DEG( acos( normal[2] ) ) * ( 255.0f / 360.0f ) );
|
|
b &= 0xff;
|
|
|
|
bytes[0] = b; // longitude
|
|
bytes[1] = a; // lattitude
|
|
}
|
|
}
|
|
|
|
/*
|
|
=================
|
|
PlaneTypeForNormal
|
|
=================
|
|
*/
|
|
int PlaneTypeForNormal( vec3_t normal ) {
|
|
if ( normal[0] == 1.0 || normal[0] == -1.0 ) {
|
|
return PLANE_X;
|
|
}
|
|
if ( normal[1] == 1.0 || normal[1] == -1.0 ) {
|
|
return PLANE_Y;
|
|
}
|
|
if ( normal[2] == 1.0 || normal[2] == -1.0 ) {
|
|
return PLANE_Z;
|
|
}
|
|
|
|
return PLANE_NON_AXIAL;
|
|
}
|
|
|
|
/*
|
|
================
|
|
MatrixMultiply
|
|
================
|
|
*/
|
|
void MatrixMultiply( float in1[3][3], float in2[3][3], float out[3][3] ) {
|
|
out[0][0] = in1[0][0] * in2[0][0] + in1[0][1] * in2[1][0] +
|
|
in1[0][2] * in2[2][0];
|
|
out[0][1] = in1[0][0] * in2[0][1] + in1[0][1] * in2[1][1] +
|
|
in1[0][2] * in2[2][1];
|
|
out[0][2] = in1[0][0] * in2[0][2] + in1[0][1] * in2[1][2] +
|
|
in1[0][2] * in2[2][2];
|
|
out[1][0] = in1[1][0] * in2[0][0] + in1[1][1] * in2[1][0] +
|
|
in1[1][2] * in2[2][0];
|
|
out[1][1] = in1[1][0] * in2[0][1] + in1[1][1] * in2[1][1] +
|
|
in1[1][2] * in2[2][1];
|
|
out[1][2] = in1[1][0] * in2[0][2] + in1[1][1] * in2[1][2] +
|
|
in1[1][2] * in2[2][2];
|
|
out[2][0] = in1[2][0] * in2[0][0] + in1[2][1] * in2[1][0] +
|
|
in1[2][2] * in2[2][0];
|
|
out[2][1] = in1[2][0] * in2[0][1] + in1[2][1] * in2[1][1] +
|
|
in1[2][2] * in2[2][1];
|
|
out[2][2] = in1[2][0] * in2[0][2] + in1[2][1] * in2[1][2] +
|
|
in1[2][2] * in2[2][2];
|
|
}
|
|
|
|
void ProjectPointOnPlane( vec3_t dst, const vec3_t p, const vec3_t normal ){
|
|
float d;
|
|
vec3_t n;
|
|
float inv_denom;
|
|
|
|
inv_denom = 1.0F / DotProduct( normal, normal );
|
|
|
|
d = DotProduct( normal, p ) * inv_denom;
|
|
|
|
n[0] = normal[0] * inv_denom;
|
|
n[1] = normal[1] * inv_denom;
|
|
n[2] = normal[2] * inv_denom;
|
|
|
|
dst[0] = p[0] - d * n[0];
|
|
dst[1] = p[1] - d * n[1];
|
|
dst[2] = p[2] - d * n[2];
|
|
}
|
|
|
|
/*
|
|
** assumes "src" is normalized
|
|
*/
|
|
void PerpendicularVector( vec3_t dst, const vec3_t src ){
|
|
int pos;
|
|
int i;
|
|
vec_t minelem = 1.0F;
|
|
vec3_t tempvec;
|
|
|
|
/*
|
|
** find the smallest magnitude axially aligned vector
|
|
*/
|
|
for ( pos = 0, i = 0; i < 3; i++ )
|
|
{
|
|
if ( fabs( src[i] ) < minelem ) {
|
|
pos = i;
|
|
minelem = (vec_t)fabs( src[i] );
|
|
}
|
|
}
|
|
tempvec[0] = tempvec[1] = tempvec[2] = 0.0F;
|
|
tempvec[pos] = 1.0F;
|
|
|
|
/*
|
|
** project the point onto the plane defined by src
|
|
*/
|
|
ProjectPointOnPlane( dst, tempvec, src );
|
|
|
|
/*
|
|
** normalize the result
|
|
*/
|
|
VectorNormalize( dst, dst );
|
|
}
|
|
|
|
/*
|
|
===============
|
|
RotatePointAroundVector
|
|
|
|
This is not implemented very well...
|
|
===============
|
|
*/
|
|
void RotatePointAroundVector( vec3_t dst, const vec3_t dir, const vec3_t point,
|
|
float degrees ) {
|
|
float m[3][3];
|
|
float im[3][3];
|
|
float zrot[3][3];
|
|
float tmpmat[3][3];
|
|
float rot[3][3];
|
|
int i;
|
|
vec3_t vr, vup, vf;
|
|
float rad;
|
|
|
|
vf[0] = dir[0];
|
|
vf[1] = dir[1];
|
|
vf[2] = dir[2];
|
|
|
|
PerpendicularVector( vr, dir );
|
|
CrossProduct( vr, vf, vup );
|
|
|
|
m[0][0] = vr[0];
|
|
m[1][0] = vr[1];
|
|
m[2][0] = vr[2];
|
|
|
|
m[0][1] = vup[0];
|
|
m[1][1] = vup[1];
|
|
m[2][1] = vup[2];
|
|
|
|
m[0][2] = vf[0];
|
|
m[1][2] = vf[1];
|
|
m[2][2] = vf[2];
|
|
|
|
memcpy( im, m, sizeof( im ) );
|
|
|
|
im[0][1] = m[1][0];
|
|
im[0][2] = m[2][0];
|
|
im[1][0] = m[0][1];
|
|
im[1][2] = m[2][1];
|
|
im[2][0] = m[0][2];
|
|
im[2][1] = m[1][2];
|
|
|
|
memset( zrot, 0, sizeof( zrot ) );
|
|
zrot[0][0] = zrot[1][1] = zrot[2][2] = 1.0F;
|
|
|
|
rad = DEG2RAD( degrees );
|
|
zrot[0][0] = (vec_t)cos( rad );
|
|
zrot[0][1] = (vec_t)sin( rad );
|
|
zrot[1][0] = (vec_t)-sin( rad );
|
|
zrot[1][1] = (vec_t)cos( rad );
|
|
|
|
MatrixMultiply( m, zrot, tmpmat );
|
|
MatrixMultiply( tmpmat, im, rot );
|
|
|
|
for ( i = 0; i < 3; i++ ) {
|
|
dst[i] = rot[i][0] * point[0] + rot[i][1] * point[1] + rot[i][2] * point[2];
|
|
}
|
|
}
|
|
|
|
|
|
////////////////////////////////////////////////////////////////////////////////
|
|
// Below is double-precision math stuff. This was initially needed by the new
|
|
// "base winding" code in q3map2 brush processing in order to fix the famous
|
|
// "disappearing triangles" issue. These definitions can be used wherever extra
|
|
// precision is needed.
|
|
////////////////////////////////////////////////////////////////////////////////
|
|
|
|
/*
|
|
=================
|
|
VectorLengthAccu
|
|
=================
|
|
*/
|
|
vec_accu_t VectorLengthAccu( const vec3_accu_t v ){
|
|
return (vec_accu_t) sqrt( ( v[0] * v[0] ) + ( v[1] * v[1] ) + ( v[2] * v[2] ) );
|
|
}
|
|
|
|
/*
|
|
=================
|
|
DotProductAccu
|
|
=================
|
|
*/
|
|
vec_accu_t DotProductAccu( const vec3_accu_t a, const vec3_accu_t b ){
|
|
return ( a[0] * b[0] ) + ( a[1] * b[1] ) + ( a[2] * b[2] );
|
|
}
|
|
|
|
/*
|
|
=================
|
|
VectorSubtractAccu
|
|
=================
|
|
*/
|
|
void VectorSubtractAccu( const vec3_accu_t a, const vec3_accu_t b, vec3_accu_t out ){
|
|
out[0] = a[0] - b[0];
|
|
out[1] = a[1] - b[1];
|
|
out[2] = a[2] - b[2];
|
|
}
|
|
|
|
/*
|
|
=================
|
|
VectorAddAccu
|
|
=================
|
|
*/
|
|
void VectorAddAccu( const vec3_accu_t a, const vec3_accu_t b, vec3_accu_t out ){
|
|
out[0] = a[0] + b[0];
|
|
out[1] = a[1] + b[1];
|
|
out[2] = a[2] + b[2];
|
|
}
|
|
|
|
/*
|
|
=================
|
|
VectorCopyAccu
|
|
=================
|
|
*/
|
|
void VectorCopyAccu( const vec3_accu_t in, vec3_accu_t out ){
|
|
out[0] = in[0];
|
|
out[1] = in[1];
|
|
out[2] = in[2];
|
|
}
|
|
|
|
/*
|
|
=================
|
|
VectorScaleAccu
|
|
=================
|
|
*/
|
|
void VectorScaleAccu( const vec3_accu_t in, vec_accu_t scaleFactor, vec3_accu_t out ){
|
|
out[0] = in[0] * scaleFactor;
|
|
out[1] = in[1] * scaleFactor;
|
|
out[2] = in[2] * scaleFactor;
|
|
}
|
|
|
|
/*
|
|
=================
|
|
CrossProductAccu
|
|
=================
|
|
*/
|
|
void CrossProductAccu( const vec3_accu_t a, const vec3_accu_t b, vec3_accu_t out ){
|
|
out[0] = ( a[1] * b[2] ) - ( a[2] * b[1] );
|
|
out[1] = ( a[2] * b[0] ) - ( a[0] * b[2] );
|
|
out[2] = ( a[0] * b[1] ) - ( a[1] * b[0] );
|
|
}
|
|
|
|
/*
|
|
=================
|
|
Q_rintAccu
|
|
=================
|
|
*/
|
|
vec_accu_t Q_rintAccu( vec_accu_t val ){
|
|
return (vec_accu_t) floor( val + 0.5 );
|
|
}
|
|
|
|
/*
|
|
=================
|
|
VectorCopyAccuToRegular
|
|
=================
|
|
*/
|
|
void VectorCopyAccuToRegular( const vec3_accu_t in, vec3_t out ){
|
|
out[0] = (vec_t) in[0];
|
|
out[1] = (vec_t) in[1];
|
|
out[2] = (vec_t) in[2];
|
|
}
|
|
|
|
/*
|
|
=================
|
|
VectorCopyRegularToAccu
|
|
=================
|
|
*/
|
|
void VectorCopyRegularToAccu( const vec3_t in, vec3_accu_t out ){
|
|
out[0] = (vec_accu_t) in[0];
|
|
out[1] = (vec_accu_t) in[1];
|
|
out[2] = (vec_accu_t) in[2];
|
|
}
|
|
|
|
/*
|
|
=================
|
|
VectorNormalizeAccu
|
|
=================
|
|
*/
|
|
vec_accu_t VectorNormalizeAccu( const vec3_accu_t in, vec3_accu_t out ){
|
|
// The sqrt() function takes double as an input and returns double as an
|
|
// output according the the man pages on Debian and on FreeBSD. Therefore,
|
|
// I don't see a reason why using a double outright (instead of using the
|
|
// vec_accu_t alias for example) could possibly be frowned upon.
|
|
|
|
vec_accu_t length;
|
|
|
|
length = (vec_accu_t) sqrt( ( in[0] * in[0] ) + ( in[1] * in[1] ) + ( in[2] * in[2] ) );
|
|
if ( length == 0 ) {
|
|
VectorClear( out );
|
|
return 0;
|
|
}
|
|
|
|
out[0] = in[0] / length;
|
|
out[1] = in[1] / length;
|
|
out[2] = in[2] / length;
|
|
|
|
return length;
|
|
}
|