mirror of
https://github.com/DrBeef/QuakeQuest.git
synced 2024-11-24 13:01:51 +00:00
305 lines
17 KiB
C
305 lines
17 KiB
C
/*
|
|
Copyright (C) 1996-1997 Id Software, Inc.
|
|
|
|
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.
|
|
|
|
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, write to the Free Software
|
|
Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA.
|
|
|
|
*/
|
|
// mathlib.h
|
|
|
|
#ifndef MATHLIB_H
|
|
#define MATHLIB_H
|
|
|
|
#include "qtypes.h"
|
|
|
|
#ifndef M_PI
|
|
#define M_PI 3.14159265358979323846 // matches value in gcc v2 math.h
|
|
#endif
|
|
|
|
struct mplane_s;
|
|
extern vec3_t vec3_origin;
|
|
|
|
#define float_nanmask (0x7F800000)
|
|
#define double_nanmask (0x7FF8000000000000)
|
|
#define FLOAT_IS_NAN(x) (((*(int *)&x)&float_nanmask)==float_nanmask)
|
|
#define DOUBLE_IS_NAN(x) (((*(long long *)&x)&double_nanmask)==double_nanmask)
|
|
|
|
#ifdef VEC_64
|
|
#define VEC_IS_NAN(x) DOUBLE_IS_NAN(x)
|
|
#else
|
|
#define VEC_IS_NAN(x) FLOAT_IS_NAN(x)
|
|
#endif
|
|
|
|
#ifdef PRVM_64
|
|
#define PRVM_IS_NAN(x) DOUBLE_IS_NAN(x)
|
|
#else
|
|
#define PRVM_IS_NAN(x) FLOAT_IS_NAN(x)
|
|
#endif
|
|
|
|
#define bound(min,num,max) ((num) >= (min) ? ((num) < (max) ? (num) : (max)) : (min))
|
|
|
|
#ifndef min
|
|
#define min(A,B) ((A) < (B) ? (A) : (B))
|
|
#define max(A,B) ((A) > (B) ? (A) : (B))
|
|
#endif
|
|
|
|
/// LordHavoc: this function never returns exactly MIN or exactly MAX, because
|
|
/// of a QuakeC bug in id1 where the line
|
|
/// self.nextthink = self.nexthink + random() * 0.5;
|
|
/// can result in 0 (self.nextthink is 0 at this point in the code to begin
|
|
/// with), causing "stone monsters" that never spawned properly, also MAX is
|
|
/// avoided because some people use random() as an index into arrays or for
|
|
/// loop conditions, where hitting exactly MAX may be a fatal error
|
|
#define lhrandom(MIN,MAX) (((double)(rand() + 0.5) / ((double)RAND_MAX + 1)) * ((MAX)-(MIN)) + (MIN))
|
|
|
|
#define invpow(base,number) (log(number) / log(base))
|
|
|
|
/// returns log base 2 of "n"
|
|
/// \WARNING: "n" MUST be a power of 2!
|
|
#define log2i(n) ((((n) & 0xAAAAAAAA) != 0 ? 1 : 0) | (((n) & 0xCCCCCCCC) != 0 ? 2 : 0) | (((n) & 0xF0F0F0F0) != 0 ? 4 : 0) | (((n) & 0xFF00FF00) != 0 ? 8 : 0) | (((n) & 0xFFFF0000) != 0 ? 16 : 0))
|
|
|
|
/// \TODO: what is this function supposed to do?
|
|
#define bit2i(n) log2i((n) << 1)
|
|
|
|
/// boolean XOR (why doesn't C have the ^^ operator for this purpose?)
|
|
#define boolxor(a,b) (!(a) != !(b))
|
|
|
|
/// returns the smallest integer greater than or equal to "value", or 0 if "value" is too big
|
|
unsigned int CeilPowerOf2(unsigned int value);
|
|
|
|
#define DEG2RAD(a) ((a) * ((float) M_PI / 180.0f))
|
|
#define RAD2DEG(a) ((a) * (180.0f / (float) M_PI))
|
|
#define ANGLEMOD(a) ((a) - 360.0 * floor((a) / 360.0))
|
|
|
|
#define DotProduct2(a,b) ((a)[0]*(b)[0]+(a)[1]*(b)[1])
|
|
#define Vector2Clear(a) ((a)[0]=(a)[1]=0)
|
|
#define Vector2Compare(a,b) (((a)[0]==(b)[0])&&((a)[1]==(b)[1]))
|
|
#define Vector2Copy(a,b) ((b)[0]=(a)[0],(b)[1]=(a)[1])
|
|
#define Vector2Negate(a,b) ((b)[0]=-((a)[0]),(b)[1]=-((a)[1]))
|
|
#define Vector2Set(a,b,c) ((a)[0]=(b),(a)[1]=(c))
|
|
#define Vector2Scale(in, scale, out) ((out)[0] = (in)[0] * (scale),(out)[1] = (in)[1] * (scale))
|
|
#define Vector2Normalize2(v,dest) {float ilength = (float) sqrt(DotProduct2((v),(v)));if (ilength) ilength = 1.0f / ilength;dest[0] = (v)[0] * ilength;dest[1] = (v)[1] * ilength;}
|
|
|
|
#define DotProduct4(a,b) ((a)[0]*(b)[0]+(a)[1]*(b)[1]+(a)[2]*(b)[2]+(a)[3]*(b)[3])
|
|
#define Vector4Clear(a) ((a)[0]=(a)[1]=(a)[2]=(a)[3]=0)
|
|
#define Vector4Compare(a,b) (((a)[0]==(b)[0])&&((a)[1]==(b)[1])&&((a)[2]==(b)[2])&&((a)[3]==(b)[3]))
|
|
#define Vector4Copy(a,b) ((b)[0]=(a)[0],(b)[1]=(a)[1],(b)[2]=(a)[2],(b)[3]=(a)[3])
|
|
#define Vector4Negate(a,b) ((b)[0]=-((a)[0]),(b)[1]=-((a)[1]),(b)[2]=-((a)[2]),(b)[3]=-((a)[3]))
|
|
#define Vector4Set(a,b,c,d,e) ((a)[0]=(b),(a)[1]=(c),(a)[2]=(d),(a)[3]=(e))
|
|
#define Vector4Normalize2(v,dest) {float ilength = (float) sqrt(DotProduct4((v),(v)));if (ilength) ilength = 1.0f / ilength;dest[0] = (v)[0] * ilength;dest[1] = (v)[1] * ilength;dest[2] = (v)[2] * ilength;dest[3] = (v)[3] * ilength;}
|
|
#define Vector4Subtract(a,b,c) ((c)[0]=(a)[0]-(b)[0],(c)[1]=(a)[1]-(b)[1],(c)[2]=(a)[2]-(b)[2],(c)[3]=(a)[3]-(b)[3])
|
|
#define Vector4Add(a,b,c) ((c)[0]=(a)[0]+(b)[0],(c)[1]=(a)[1]+(b)[1],(c)[2]=(a)[2]+(b)[2],(c)[3]=(a)[3]+(b)[3])
|
|
#define Vector4Scale(in, scale, out) ((out)[0] = (in)[0] * (scale),(out)[1] = (in)[1] * (scale),(out)[2] = (in)[2] * (scale),(out)[3] = (in)[3] * (scale))
|
|
#define Vector4Multiply(a,b,c) ((c)[0]=(a)[0]*(b)[0],(c)[1]=(a)[1]*(b)[1],(c)[2]=(a)[2]*(b)[2],(c)[3]=(a)[3]*(b)[3])
|
|
#define Vector4MA(a, scale, b, c) ((c)[0] = (a)[0] + (scale) * (b)[0],(c)[1] = (a)[1] + (scale) * (b)[1],(c)[2] = (a)[2] + (scale) * (b)[2],(c)[3] = (a)[3] + (scale) * (b)[3])
|
|
#define Vector4Lerp(v1,lerp,v2,c) ((c)[0] = (v1)[0] + (lerp) * ((v2)[0] - (v1)[0]), (c)[1] = (v1)[1] + (lerp) * ((v2)[1] - (v1)[1]), (c)[2] = (v1)[2] + (lerp) * ((v2)[2] - (v1)[2]), (c)[3] = (v1)[3] + (lerp) * ((v2)[3] - (v1)[3]))
|
|
|
|
#define VectorNegate(a,b) ((b)[0]=-((a)[0]),(b)[1]=-((a)[1]),(b)[2]=-((a)[2]))
|
|
#define VectorSet(a,b,c,d) ((a)[0]=(b),(a)[1]=(c),(a)[2]=(d))
|
|
#define VectorClear(a) ((a)[0]=(a)[1]=(a)[2]=0)
|
|
#define DotProduct(a,b) ((a)[0]*(b)[0]+(a)[1]*(b)[1]+(a)[2]*(b)[2])
|
|
#define VectorSubtract(a,b,c) ((c)[0]=(a)[0]-(b)[0],(c)[1]=(a)[1]-(b)[1],(c)[2]=(a)[2]-(b)[2])
|
|
#define VectorAdd(a,b,c) ((c)[0]=(a)[0]+(b)[0],(c)[1]=(a)[1]+(b)[1],(c)[2]=(a)[2]+(b)[2])
|
|
#define VectorCopy(a,b) ((b)[0]=(a)[0],(b)[1]=(a)[1],(b)[2]=(a)[2])
|
|
#define VectorMultiply(a,b,c) ((c)[0]=(a)[0]*(b)[0],(c)[1]=(a)[1]*(b)[1],(c)[2]=(a)[2]*(b)[2])
|
|
#define CrossProduct(a,b,c) ((c)[0]=(a)[1]*(b)[2]-(a)[2]*(b)[1],(c)[1]=(a)[2]*(b)[0]-(a)[0]*(b)[2],(c)[2]=(a)[0]*(b)[1]-(a)[1]*(b)[0])
|
|
#define VectorNormalize(v) {float ilength = (float) sqrt(DotProduct((v),(v)));if (ilength) ilength = 1.0f / ilength;(v)[0] *= ilength;(v)[1] *= ilength;(v)[2] *= ilength;}
|
|
#define VectorNormalize2(v,dest) {float ilength = (float) sqrt(DotProduct((v),(v)));if (ilength) ilength = 1.0f / ilength;dest[0] = (v)[0] * ilength;dest[1] = (v)[1] * ilength;dest[2] = (v)[2] * ilength;}
|
|
#define VectorNormalizeDouble(v) {double ilength = sqrt(DotProduct((v),(v)));if (ilength) ilength = 1.0 / ilength;(v)[0] *= ilength;(v)[1] *= ilength;(v)[2] *= ilength;}
|
|
#define VectorDistance2(a, b) (((a)[0] - (b)[0]) * ((a)[0] - (b)[0]) + ((a)[1] - (b)[1]) * ((a)[1] - (b)[1]) + ((a)[2] - (b)[2]) * ((a)[2] - (b)[2]))
|
|
#define VectorDistance(a, b) (sqrt(VectorDistance2(a,b)))
|
|
#define VectorLength(a) (sqrt((double)DotProduct(a, a)))
|
|
#define VectorLength2(a) (DotProduct(a, a))
|
|
#define VectorScale(in, scale, out) ((out)[0] = (in)[0] * (scale),(out)[1] = (in)[1] * (scale),(out)[2] = (in)[2] * (scale))
|
|
#define VectorScaleCast(in, scale, outtype, out) ((out)[0] = (outtype) ((in)[0] * (scale)),(out)[1] = (outtype) ((in)[1] * (scale)),(out)[2] = (outtype) ((in)[2] * (scale)))
|
|
#define VectorCompare(a,b) (((a)[0]==(b)[0])&&((a)[1]==(b)[1])&&((a)[2]==(b)[2]))
|
|
#define VectorMA(a, scale, b, c) ((c)[0] = (a)[0] + (scale) * (b)[0],(c)[1] = (a)[1] + (scale) * (b)[1],(c)[2] = (a)[2] + (scale) * (b)[2])
|
|
#define VectorM(scale1, b1, c) ((c)[0] = (scale1) * (b1)[0],(c)[1] = (scale1) * (b1)[1],(c)[2] = (scale1) * (b1)[2])
|
|
#define VectorMAM(scale1, b1, scale2, b2, c) ((c)[0] = (scale1) * (b1)[0] + (scale2) * (b2)[0],(c)[1] = (scale1) * (b1)[1] + (scale2) * (b2)[1],(c)[2] = (scale1) * (b1)[2] + (scale2) * (b2)[2])
|
|
#define VectorMAMAM(scale1, b1, scale2, b2, scale3, b3, c) ((c)[0] = (scale1) * (b1)[0] + (scale2) * (b2)[0] + (scale3) * (b3)[0],(c)[1] = (scale1) * (b1)[1] + (scale2) * (b2)[1] + (scale3) * (b3)[1],(c)[2] = (scale1) * (b1)[2] + (scale2) * (b2)[2] + (scale3) * (b3)[2])
|
|
#define VectorMAMAMAM(scale1, b1, scale2, b2, scale3, b3, scale4, b4, c) ((c)[0] = (scale1) * (b1)[0] + (scale2) * (b2)[0] + (scale3) * (b3)[0] + (scale4) * (b4)[0],(c)[1] = (scale1) * (b1)[1] + (scale2) * (b2)[1] + (scale3) * (b3)[1] + (scale4) * (b4)[1],(c)[2] = (scale1) * (b1)[2] + (scale2) * (b2)[2] + (scale3) * (b3)[2] + (scale4) * (b4)[2])
|
|
#define VectorRandom(v) do{(v)[0] = lhrandom(-1, 1);(v)[1] = lhrandom(-1, 1);(v)[2] = lhrandom(-1, 1);}while(DotProduct(v, v) > 1)
|
|
#define VectorLerp(v1,lerp,v2,c) ((c)[0] = (v1)[0] + (lerp) * ((v2)[0] - (v1)[0]), (c)[1] = (v1)[1] + (lerp) * ((v2)[1] - (v1)[1]), (c)[2] = (v1)[2] + (lerp) * ((v2)[2] - (v1)[2]))
|
|
#define VectorReflect(a,r,b,c) do{double d;d = DotProduct((a), (b)) * -(1.0 + (r));VectorMA((a), (d), (b), (c));}while(0)
|
|
#define BoxesOverlap(a,b,c,d) ((a)[0] <= (d)[0] && (b)[0] >= (c)[0] && (a)[1] <= (d)[1] && (b)[1] >= (c)[1] && (a)[2] <= (d)[2] && (b)[2] >= (c)[2])
|
|
#define BoxInsideBox(a,b,c,d) ((a)[0] >= (c)[0] && (b)[0] <= (d)[0] && (a)[1] >= (c)[1] && (b)[1] <= (d)[1] && (a)[2] >= (c)[2] && (b)[2] <= (d)[2])
|
|
#define TriangleBBoxOverlapsBox(a,b,c,d,e) (min((a)[0], min((b)[0], (c)[0])) < (e)[0] && max((a)[0], max((b)[0], (c)[0])) > (d)[0] && min((a)[1], min((b)[1], (c)[1])) < (e)[1] && max((a)[1], max((b)[1], (c)[1])) > (d)[1] && min((a)[2], min((b)[2], (c)[2])) < (e)[2] && max((a)[2], max((b)[2], (c)[2])) > (d)[2])
|
|
|
|
#define TriangleNormal(a,b,c,n) ( \
|
|
(n)[0] = ((a)[1] - (b)[1]) * ((c)[2] - (b)[2]) - ((a)[2] - (b)[2]) * ((c)[1] - (b)[1]), \
|
|
(n)[1] = ((a)[2] - (b)[2]) * ((c)[0] - (b)[0]) - ((a)[0] - (b)[0]) * ((c)[2] - (b)[2]), \
|
|
(n)[2] = ((a)[0] - (b)[0]) * ((c)[1] - (b)[1]) - ((a)[1] - (b)[1]) * ((c)[0] - (b)[0]) \
|
|
)
|
|
|
|
/*! Fast PointInfrontOfTriangle.
|
|
* subtracts v1 from v0 and v2, combined into a crossproduct, combined with a
|
|
* dotproduct of the light location relative to the first point of the
|
|
* triangle (any point works, since any triangle is obviously flat), and
|
|
* finally a comparison to determine if the light is infront of the triangle
|
|
* (the goal of this statement) we do not need to normalize the surface
|
|
* normal because both sides of the comparison use it, therefore they are
|
|
* both multiplied the same amount... furthermore a subtract can be done on
|
|
* the point to eliminate one dotproduct
|
|
* this is ((p - a) * cross(a-b,c-b))
|
|
*/
|
|
#define PointInfrontOfTriangle(p,a,b,c) \
|
|
( ((p)[0] - (a)[0]) * (((a)[1] - (b)[1]) * ((c)[2] - (b)[2]) - ((a)[2] - (b)[2]) * ((c)[1] - (b)[1])) \
|
|
+ ((p)[1] - (a)[1]) * (((a)[2] - (b)[2]) * ((c)[0] - (b)[0]) - ((a)[0] - (b)[0]) * ((c)[2] - (b)[2])) \
|
|
+ ((p)[2] - (a)[2]) * (((a)[0] - (b)[0]) * ((c)[1] - (b)[1]) - ((a)[1] - (b)[1]) * ((c)[0] - (b)[0])) > 0)
|
|
|
|
#if 0
|
|
// readable version, kept only for explanatory reasons
|
|
int PointInfrontOfTriangle(const float *p, const float *a, const float *b, const float *c)
|
|
{
|
|
float dir0[3], dir1[3], normal[3];
|
|
|
|
// calculate two mostly perpendicular edge directions
|
|
VectorSubtract(a, b, dir0);
|
|
VectorSubtract(c, b, dir1);
|
|
|
|
// we have two edge directions, we can calculate a third vector from
|
|
// them, which is the direction of the surface normal (its magnitude
|
|
// is not 1 however)
|
|
CrossProduct(dir0, dir1, normal);
|
|
|
|
// compare distance of light along normal, with distance of any point
|
|
// of the triangle along the same normal (the triangle is planar,
|
|
// I.E. flat, so all points give the same answer)
|
|
return DotProduct(p, normal) > DotProduct(a, normal);
|
|
}
|
|
#endif
|
|
|
|
#define lhcheeserand() (seed = (seed * 987211u) ^ (seed >> 13u) ^ 914867)
|
|
#define lhcheeserandom(MIN,MAX) ((double)(lhcheeserand() + 0.5) / ((double)4096.0*1024.0*1024.0) * ((MAX)-(MIN)) + (MIN))
|
|
#define VectorCheeseRandom(v) do{(v)[0] = lhcheeserandom(-1, 1);(v)[1] = lhcheeserandom(-1, 1);(v)[2] = lhcheeserandom(-1, 1);}while(DotProduct(v, v) > 1)
|
|
|
|
/*
|
|
// LordHavoc: quaternion math, untested, don't know if these are correct,
|
|
// need to add conversion to/from matrices
|
|
// LordHavoc: later note: the matrix faq is useful: http://skal.planet-d.net/demo/matrixfaq.htm
|
|
// LordHavoc: these are probably very wrong and I'm not sure I care, not used by anything
|
|
|
|
// returns length of quaternion
|
|
#define qlen(a) ((float) sqrt((a)[0]*(a)[0]+(a)[1]*(a)[1]+(a)[2]*(a)[2]+(a)[3]*(a)[3]))
|
|
// returns squared length of quaternion
|
|
#define qlen2(a) ((a)[0]*(a)[0]+(a)[1]*(a)[1]+(a)[2]*(a)[2]+(a)[3]*(a)[3])
|
|
// makes a quaternion from x, y, z, and a rotation angle (in degrees)
|
|
#define QuatMake(x,y,z,r,c)\
|
|
{\
|
|
if (r == 0)\
|
|
{\
|
|
(c)[0]=(float) ((x) * (1.0f / 0.0f));\
|
|
(c)[1]=(float) ((y) * (1.0f / 0.0f));\
|
|
(c)[2]=(float) ((z) * (1.0f / 0.0f));\
|
|
(c)[3]=(float) 1.0f;\
|
|
}\
|
|
else\
|
|
{\
|
|
float r2 = (r) * 0.5 * (M_PI / 180);\
|
|
float r2is = 1.0f / sin(r2);\
|
|
(c)[0]=(float) ((x)/r2is);\
|
|
(c)[1]=(float) ((y)/r2is);\
|
|
(c)[2]=(float) ((z)/r2is);\
|
|
(c)[3]=(float) (cos(r2));\
|
|
}\
|
|
}
|
|
// makes a quaternion from a vector and a rotation angle (in degrees)
|
|
#define QuatFromVec(a,r,c) QuatMake((a)[0],(a)[1],(a)[2],(r))
|
|
// copies a quaternion
|
|
#define QuatCopy(a,c) {(c)[0]=(a)[0];(c)[1]=(a)[1];(c)[2]=(a)[2];(c)[3]=(a)[3];}
|
|
#define QuatSubtract(a,b,c) {(c)[0]=(a)[0]-(b)[0];(c)[1]=(a)[1]-(b)[1];(c)[2]=(a)[2]-(b)[2];(c)[3]=(a)[3]-(b)[3];}
|
|
#define QuatAdd(a,b,c) {(c)[0]=(a)[0]+(b)[0];(c)[1]=(a)[1]+(b)[1];(c)[2]=(a)[2]+(b)[2];(c)[3]=(a)[3]+(b)[3];}
|
|
#define QuatScale(a,b,c) {(c)[0]=(a)[0]*b;(c)[1]=(a)[1]*b;(c)[2]=(a)[2]*b;(c)[3]=(a)[3]*b;}
|
|
// FIXME: this is wrong, do some more research on quaternions
|
|
//#define QuatMultiply(a,b,c) {(c)[0]=(a)[0]*(b)[0];(c)[1]=(a)[1]*(b)[1];(c)[2]=(a)[2]*(b)[2];(c)[3]=(a)[3]*(b)[3];}
|
|
// FIXME: this is wrong, do some more research on quaternions
|
|
//#define QuatMultiplyAdd(a,b,d,c) {(c)[0]=(a)[0]*(b)[0]+d[0];(c)[1]=(a)[1]*(b)[1]+d[1];(c)[2]=(a)[2]*(b)[2]+d[2];(c)[3]=(a)[3]*(b)[3]+d[3];}
|
|
#define qdist(a,b) ((float) sqrt(((b)[0]-(a)[0])*((b)[0]-(a)[0])+((b)[1]-(a)[1])*((b)[1]-(a)[1])+((b)[2]-(a)[2])*((b)[2]-(a)[2])+((b)[3]-(a)[3])*((b)[3]-(a)[3])))
|
|
#define qdist2(a,b) (((b)[0]-(a)[0])*((b)[0]-(a)[0])+((b)[1]-(a)[1])*((b)[1]-(a)[1])+((b)[2]-(a)[2])*((b)[2]-(a)[2])+((b)[3]-(a)[3])*((b)[3]-(a)[3]))
|
|
*/
|
|
|
|
#define VectorCopy4(a,b) {(b)[0]=(a)[0];(b)[1]=(a)[1];(b)[2]=(a)[2];(b)[3]=(a)[3];}
|
|
|
|
vec_t Length (vec3_t v);
|
|
|
|
/// returns vector length
|
|
float VectorNormalizeLength (vec3_t v);
|
|
|
|
/// returns vector length
|
|
float VectorNormalizeLength2 (vec3_t v, vec3_t dest);
|
|
|
|
#define NUMVERTEXNORMALS 162
|
|
extern float m_bytenormals[NUMVERTEXNORMALS][3];
|
|
|
|
unsigned char NormalToByte(const vec3_t n);
|
|
void ByteToNormal(unsigned char num, vec3_t n);
|
|
|
|
void R_ConcatRotations (const float in1[3*3], const float in2[3*3], float out[3*3]);
|
|
void R_ConcatTransforms (const float in1[3*4], const float in2[3*4], float out[3*4]);
|
|
|
|
void AngleVectors (const vec3_t angles, vec3_t forward, vec3_t right, vec3_t up);
|
|
/// LordHavoc: proper matrix version of AngleVectors
|
|
void AngleVectorsFLU (const vec3_t angles, vec3_t forward, vec3_t left, vec3_t up);
|
|
/// LordHavoc: builds a [3][4] matrix
|
|
void AngleMatrix (const vec3_t angles, const vec3_t translate, vec_t matrix[][4]);
|
|
/// LordHavoc: calculates pitch/yaw/roll angles from forward and up vectors
|
|
void AnglesFromVectors (vec3_t angles, const vec3_t forward, const vec3_t up, qboolean flippitch);
|
|
|
|
/// LordHavoc: like AngleVectors, but taking a forward vector instead of angles, useful!
|
|
void VectorVectors(const vec3_t forward, vec3_t right, vec3_t up);
|
|
void VectorVectorsDouble(const double *forward, double *right, double *up);
|
|
|
|
void PlaneClassify(struct mplane_s *p);
|
|
int BoxOnPlaneSide(const vec3_t emins, const vec3_t emaxs, const struct mplane_s *p);
|
|
int BoxOnPlaneSide_Separate(const vec3_t emins, const vec3_t emaxs, const vec3_t normal, const vec_t dist);
|
|
void BoxPlaneCorners(const vec3_t emins, const vec3_t emaxs, const struct mplane_s *p, vec3_t outnear, vec3_t outfar);
|
|
void BoxPlaneCorners_Separate(const vec3_t emins, const vec3_t emaxs, const vec3_t normal, vec3_t outnear, vec3_t outfar);
|
|
void BoxPlaneCornerDistances(const vec3_t emins, const vec3_t emaxs, const struct mplane_s *p, vec_t *outnear, vec_t *outfar);
|
|
void BoxPlaneCornerDistances_Separate(const vec3_t emins, const vec3_t emaxs, const vec3_t normal, vec_t *outnear, vec_t *outfar);
|
|
|
|
#define PlaneDist(point,plane) ((plane)->type < 3 ? (point)[(plane)->type] : DotProduct((point), (plane)->normal))
|
|
#define PlaneDiff(point,plane) (((plane)->type < 3 ? (point)[(plane)->type] : DotProduct((point), (plane)->normal)) - (plane)->dist)
|
|
|
|
/// LordHavoc: minimal plane structure
|
|
typedef struct tinyplane_s
|
|
{
|
|
float normal[3], dist;
|
|
}
|
|
tinyplane_t;
|
|
|
|
typedef struct tinydoubleplane_s
|
|
{
|
|
double normal[3], dist;
|
|
}
|
|
tinydoubleplane_t;
|
|
|
|
void RotatePointAroundVector(vec3_t dst, const vec3_t dir, const vec3_t point, float degrees);
|
|
|
|
float RadiusFromBounds (const vec3_t mins, const vec3_t maxs);
|
|
float RadiusFromBoundsAndOrigin (const vec3_t mins, const vec3_t maxs, const vec3_t origin);
|
|
|
|
struct matrix4x4_s;
|
|
/// print a matrix to the console
|
|
void Matrix4x4_Print(const struct matrix4x4_s *in);
|
|
int Math_atov(const char *s, prvm_vec3_t out);
|
|
|
|
void BoxFromPoints(vec3_t mins, vec3_t maxs, int numpoints, vec_t *point3f);
|
|
|
|
int LoopingFrameNumberFromDouble(double t, int loopframes);
|
|
|
|
void Mathlib_Init(void);
|
|
|
|
#endif
|
|
|