#include "rs_core.rsh" /* Implementation of Core Runtime */ ///////////////////////////////////////////////////// // Quaternion ops ///////////////////////////////////////////////////// #if (defined(RS_VERSION) && (RS_VERSION >= UNRELEASED)) extern void __attribute__((overloadable)) rsQuaternionAdd(rs_quaternion* q, const rs_quaternion* rhs) { q->w += rhs->w; q->x += rhs->x; q->y += rhs->y; q->z += rhs->z; } extern void __attribute__((overloadable)) rsQuaternionConjugate(rs_quaternion* q) { q->x = -q->x; q->y = -q->y; q->z = -q->z; } extern float __attribute__((overloadable)) rsQuaternionDot(const rs_quaternion* q0, const rs_quaternion* q1) { return q0->w*q1->w + q0->x*q1->x + q0->y*q1->y + q0->z*q1->z; } extern void __attribute__((overloadable)) rsQuaternionGetMatrixUnit(rs_matrix4x4* m, const rs_quaternion* q) { float xx = q->x * q->x; float xy = q->x * q->y; float xz = q->x * q->z; float xw = q->x * q->w; float yy = q->y * q->y; float yz = q->y * q->z; float yw = q->y * q->w; float zz = q->z * q->z; float zw = q->z * q->w; m->m[0] = 1.0f - 2.0f * ( yy + zz ); m->m[4] = 2.0f * ( xy - zw ); m->m[8] = 2.0f * ( xz + yw ); m->m[1] = 2.0f * ( xy + zw ); m->m[5] = 1.0f - 2.0f * ( xx + zz ); m->m[9] = 2.0f * ( yz - xw ); m->m[2] = 2.0f * ( xz - yw ); m->m[6] = 2.0f * ( yz + xw ); m->m[10] = 1.0f - 2.0f * ( xx + yy ); m->m[3] = m->m[7] = m->m[11] = m->m[12] = m->m[13] = m->m[14] = 0.0f; m->m[15] = 1.0f; } extern void __attribute__((overloadable)) rsQuaternionLoadRotateUnit(rs_quaternion* q, float rot, float x, float y, float z) { rot *= (float)(M_PI / 180.0f) * 0.5f; float c = cos(rot); float s = sin(rot); q->w = c; q->x = x * s; q->y = y * s; q->z = z * s; } extern void __attribute__((overloadable)) rsQuaternionSet(rs_quaternion* q, float w, float x, float y, float z) { q->w = w; q->x = x; q->y = y; q->z = z; } extern void __attribute__((overloadable)) rsQuaternionSet(rs_quaternion* q, const rs_quaternion* rhs) { q->w = rhs->w; q->x = rhs->x; q->y = rhs->y; q->z = rhs->z; } extern void __attribute__((overloadable)) rsQuaternionLoadRotate(rs_quaternion* q, float rot, float x, float y, float z) { const float len = x*x + y*y + z*z; if (len != 1) { const float recipLen = 1.f / sqrt(len); x *= recipLen; y *= recipLen; z *= recipLen; } rsQuaternionLoadRotateUnit(q, rot, x, y, z); } extern void __attribute__((overloadable)) rsQuaternionNormalize(rs_quaternion* q) { const float len = rsQuaternionDot(q, q); if (len != 1) { const float recipLen = 1.f / sqrt(len); q->w *= recipLen; q->x *= recipLen; q->y *= recipLen; q->z *= recipLen; } } extern void __attribute__((overloadable)) rsQuaternionMultiply(rs_quaternion* q, float scalar) { q->w *= scalar; q->x *= scalar; q->y *= scalar; q->z *= scalar; } extern void __attribute__((overloadable)) rsQuaternionMultiply(rs_quaternion* q, const rs_quaternion* rhs) { rs_quaternion qtmp; rsQuaternionSet(&qtmp, q); q->w = qtmp.w*rhs->w - qtmp.x*rhs->x - qtmp.y*rhs->y - qtmp.z*rhs->z; q->x = qtmp.w*rhs->x + qtmp.x*rhs->w + qtmp.y*rhs->z - qtmp.z*rhs->y; q->y = qtmp.w*rhs->y + qtmp.y*rhs->w + qtmp.z*rhs->x - qtmp.x*rhs->z; q->z = qtmp.w*rhs->z + qtmp.z*rhs->w + qtmp.x*rhs->y - qtmp.y*rhs->x; rsQuaternionNormalize(q); } extern void __attribute__((overloadable)) rsQuaternionSlerp(rs_quaternion* q, const rs_quaternion* q0, const rs_quaternion* q1, float t) { if (t <= 0.0f) { rsQuaternionSet(q, q0); return; } if (t >= 1.0f) { rsQuaternionSet(q, q1); return; } rs_quaternion tempq0, tempq1; rsQuaternionSet(&tempq0, q0); rsQuaternionSet(&tempq1, q1); float angle = rsQuaternionDot(q0, q1); if (angle < 0) { rsQuaternionMultiply(&tempq0, -1.0f); angle *= -1.0f; } float scale, invScale; if (angle + 1.0f > 0.05f) { if (1.0f - angle >= 0.05f) { float theta = acos(angle); float invSinTheta = 1.0f / sin(theta); scale = sin(theta * (1.0f - t)) * invSinTheta; invScale = sin(theta * t) * invSinTheta; } else { scale = 1.0f - t; invScale = t; } } else { rsQuaternionSet(&tempq1, tempq0.z, -tempq0.y, tempq0.x, -tempq0.w); scale = sin(M_PI * (0.5f - t)); invScale = sin(M_PI * t); } rsQuaternionSet(q, tempq0.w*scale + tempq1.w*invScale, tempq0.x*scale + tempq1.x*invScale, tempq0.y*scale + tempq1.y*invScale, tempq0.z*scale + tempq1.z*invScale); } #endif // (defined(RS_VERSION) && (RS_VERSION >= UNRELEASED))