summaryrefslogtreecommitdiff
path: root/indra/llmath/m3math.cpp
diff options
context:
space:
mode:
Diffstat (limited to 'indra/llmath/m3math.cpp')
-rw-r--r--indra/llmath/m3math.cpp720
1 files changed, 360 insertions, 360 deletions
diff --git a/indra/llmath/m3math.cpp b/indra/llmath/m3math.cpp
index c7777164c0..472d340af5 100644
--- a/indra/llmath/m3math.cpp
+++ b/indra/llmath/m3math.cpp
@@ -1,25 +1,25 @@
-/**
+/**
* @file m3math.cpp
* @brief LLMatrix3 class implementation.
*
* $LicenseInfo:firstyear=2000&license=viewerlgpl$
* Second Life Viewer Source Code
* Copyright (C) 2010, Linden Research, Inc.
- *
+ *
* This library is free software; you can redistribute it and/or
* modify it under the terms of the GNU Lesser General Public
* License as published by the Free Software Foundation;
* version 2.1 of the License only.
- *
+ *
* This library 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
* Lesser General Public License for more details.
- *
+ *
* You should have received a copy of the GNU Lesser General Public
* License along with this library; if not, write to the Free Software
* Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
- *
+ *
* Linden Research, Inc., 945 Battery Street, San Francisco, CA 94111 USA
* $/LicenseInfo$
*/
@@ -36,7 +36,7 @@
// LLMatrix3
-// ji
+// ji
// LLMatrix3 = |00 01 02 |
// |10 11 12 |
// |20 21 22 |
@@ -51,266 +51,266 @@
LLMatrix3::LLMatrix3(const LLQuaternion &q)
{
- setRot(q);
+ setRot(q);
}
LLMatrix3::LLMatrix3(const F32 angle, const LLVector3 &vec)
{
- LLQuaternion quat(angle, vec);
- setRot(quat);
+ LLQuaternion quat(angle, vec);
+ setRot(quat);
}
LLMatrix3::LLMatrix3(const F32 angle, const LLVector3d &vec)
{
- LLVector3 vec_f;
- vec_f.setVec(vec);
- LLQuaternion quat(angle, vec_f);
- setRot(quat);
+ LLVector3 vec_f;
+ vec_f.setVec(vec);
+ LLQuaternion quat(angle, vec_f);
+ setRot(quat);
}
LLMatrix3::LLMatrix3(const F32 angle, const LLVector4 &vec)
{
- LLQuaternion quat(angle, vec);
- setRot(quat);
+ LLQuaternion quat(angle, vec);
+ setRot(quat);
}
LLMatrix3::LLMatrix3(const F32 roll, const F32 pitch, const F32 yaw)
{
- setRot(roll,pitch,yaw);
+ setRot(roll,pitch,yaw);
}
// From Matrix and Quaternion FAQ
void LLMatrix3::getEulerAngles(F32 *roll, F32 *pitch, F32 *yaw) const
{
- F64 angle_x, angle_y, angle_z;
- F64 cx, cy, cz; // cosine of angle_x, angle_y, angle_z
- F64 sx, sz; // sine of angle_x, angle_y, angle_z
+ F64 angle_x, angle_y, angle_z;
+ F64 cx, cy, cz; // cosine of angle_x, angle_y, angle_z
+ F64 sx, sz; // sine of angle_x, angle_y, angle_z
- angle_y = asin(llclamp(mMatrix[2][0], -1.f, 1.f));
- cy = cos(angle_y);
+ angle_y = asin(llclamp(mMatrix[2][0], -1.f, 1.f));
+ cy = cos(angle_y);
- if (fabs(cy) > 0.005) // non-zero
- {
- // no gimbal lock
- cx = mMatrix[2][2] / cy;
- sx = - mMatrix[2][1] / cy;
+ if (fabs(cy) > 0.005) // non-zero
+ {
+ // no gimbal lock
+ cx = mMatrix[2][2] / cy;
+ sx = - mMatrix[2][1] / cy;
- angle_x = (F32) atan2(sx, cx);
+ angle_x = (F32) atan2(sx, cx);
- cz = mMatrix[0][0] / cy;
- sz = - mMatrix[1][0] / cy;
+ cz = mMatrix[0][0] / cy;
+ sz = - mMatrix[1][0] / cy;
- angle_z = (F32) atan2(sz, cz);
- }
- else
- {
- // yup, gimbal lock
- angle_x = 0;
+ angle_z = (F32) atan2(sz, cz);
+ }
+ else
+ {
+ // yup, gimbal lock
+ angle_x = 0;
- // some tricky math thereby avoided, see article
+ // some tricky math thereby avoided, see article
- cz = mMatrix[1][1];
- sz = mMatrix[0][1];
+ cz = mMatrix[1][1];
+ sz = mMatrix[0][1];
- angle_z = atan2(sz, cz);
- }
+ angle_z = atan2(sz, cz);
+ }
- *roll = (F32)angle_x;
- *pitch = (F32)angle_y;
- *yaw = (F32)angle_z;
+ *roll = (F32)angle_x;
+ *pitch = (F32)angle_y;
+ *yaw = (F32)angle_z;
}
-
+
// Clear and Assignment Functions
-const LLMatrix3& LLMatrix3::setIdentity()
+const LLMatrix3& LLMatrix3::setIdentity()
{
- mMatrix[0][0] = 1.f;
- mMatrix[0][1] = 0.f;
- mMatrix[0][2] = 0.f;
+ mMatrix[0][0] = 1.f;
+ mMatrix[0][1] = 0.f;
+ mMatrix[0][2] = 0.f;
- mMatrix[1][0] = 0.f;
- mMatrix[1][1] = 1.f;
- mMatrix[1][2] = 0.f;
+ mMatrix[1][0] = 0.f;
+ mMatrix[1][1] = 1.f;
+ mMatrix[1][2] = 0.f;
- mMatrix[2][0] = 0.f;
- mMatrix[2][1] = 0.f;
- mMatrix[2][2] = 1.f;
- return (*this);
+ mMatrix[2][0] = 0.f;
+ mMatrix[2][1] = 0.f;
+ mMatrix[2][2] = 1.f;
+ return (*this);
}
-const LLMatrix3& LLMatrix3::clear()
+const LLMatrix3& LLMatrix3::clear()
{
- mMatrix[0][0] = 0.f;
- mMatrix[0][1] = 0.f;
- mMatrix[0][2] = 0.f;
+ mMatrix[0][0] = 0.f;
+ mMatrix[0][1] = 0.f;
+ mMatrix[0][2] = 0.f;
- mMatrix[1][0] = 0.f;
- mMatrix[1][1] = 0.f;
- mMatrix[1][2] = 0.f;
+ mMatrix[1][0] = 0.f;
+ mMatrix[1][1] = 0.f;
+ mMatrix[1][2] = 0.f;
- mMatrix[2][0] = 0.f;
- mMatrix[2][1] = 0.f;
- mMatrix[2][2] = 0.f;
- return (*this);
+ mMatrix[2][0] = 0.f;
+ mMatrix[2][1] = 0.f;
+ mMatrix[2][2] = 0.f;
+ return (*this);
}
-const LLMatrix3& LLMatrix3::setZero()
+const LLMatrix3& LLMatrix3::setZero()
{
- mMatrix[0][0] = 0.f;
- mMatrix[0][1] = 0.f;
- mMatrix[0][2] = 0.f;
+ mMatrix[0][0] = 0.f;
+ mMatrix[0][1] = 0.f;
+ mMatrix[0][2] = 0.f;
- mMatrix[1][0] = 0.f;
- mMatrix[1][1] = 0.f;
- mMatrix[1][2] = 0.f;
+ mMatrix[1][0] = 0.f;
+ mMatrix[1][1] = 0.f;
+ mMatrix[1][2] = 0.f;
- mMatrix[2][0] = 0.f;
- mMatrix[2][1] = 0.f;
- mMatrix[2][2] = 0.f;
- return (*this);
+ mMatrix[2][0] = 0.f;
+ mMatrix[2][1] = 0.f;
+ mMatrix[2][2] = 0.f;
+ return (*this);
}
// various useful mMatrix functions
-const LLMatrix3& LLMatrix3::transpose()
+const LLMatrix3& LLMatrix3::transpose()
{
- // transpose the matrix
- F32 temp;
- temp = mMatrix[VX][VY]; mMatrix[VX][VY] = mMatrix[VY][VX]; mMatrix[VY][VX] = temp;
- temp = mMatrix[VX][VZ]; mMatrix[VX][VZ] = mMatrix[VZ][VX]; mMatrix[VZ][VX] = temp;
- temp = mMatrix[VY][VZ]; mMatrix[VY][VZ] = mMatrix[VZ][VY]; mMatrix[VZ][VY] = temp;
- return *this;
+ // transpose the matrix
+ F32 temp;
+ temp = mMatrix[VX][VY]; mMatrix[VX][VY] = mMatrix[VY][VX]; mMatrix[VY][VX] = temp;
+ temp = mMatrix[VX][VZ]; mMatrix[VX][VZ] = mMatrix[VZ][VX]; mMatrix[VZ][VX] = temp;
+ temp = mMatrix[VY][VZ]; mMatrix[VY][VZ] = mMatrix[VZ][VY]; mMatrix[VZ][VY] = temp;
+ return *this;
}
-F32 LLMatrix3::determinant() const
+F32 LLMatrix3::determinant() const
{
- // Is this a useful method when we assume the matrices are valid rotation
- // matrices throughout this implementation?
- return mMatrix[0][0] * (mMatrix[1][1] * mMatrix[2][2] - mMatrix[1][2] * mMatrix[2][1]) +
- mMatrix[0][1] * (mMatrix[1][2] * mMatrix[2][0] - mMatrix[1][0] * mMatrix[2][2]) +
- mMatrix[0][2] * (mMatrix[1][0] * mMatrix[2][1] - mMatrix[1][1] * mMatrix[2][0]);
+ // Is this a useful method when we assume the matrices are valid rotation
+ // matrices throughout this implementation?
+ return mMatrix[0][0] * (mMatrix[1][1] * mMatrix[2][2] - mMatrix[1][2] * mMatrix[2][1]) +
+ mMatrix[0][1] * (mMatrix[1][2] * mMatrix[2][0] - mMatrix[1][0] * mMatrix[2][2]) +
+ mMatrix[0][2] * (mMatrix[1][0] * mMatrix[2][1] - mMatrix[1][1] * mMatrix[2][0]);
}
// inverts this matrix
void LLMatrix3::invert()
{
- // fails silently if determinant is zero too small
- F32 det = determinant();
- const F32 VERY_SMALL_DETERMINANT = 0.000001f;
- if (fabs(det) > VERY_SMALL_DETERMINANT)
- {
- // invertiable
- LLMatrix3 t(*this);
- mMatrix[VX][VX] = ( t.mMatrix[VY][VY] * t.mMatrix[VZ][VZ] - t.mMatrix[VY][VZ] * t.mMatrix[VZ][VY] ) / det;
- mMatrix[VY][VX] = ( t.mMatrix[VY][VZ] * t.mMatrix[VZ][VX] - t.mMatrix[VY][VX] * t.mMatrix[VZ][VZ] ) / det;
- mMatrix[VZ][VX] = ( t.mMatrix[VY][VX] * t.mMatrix[VZ][VY] - t.mMatrix[VY][VY] * t.mMatrix[VZ][VX] ) / det;
- mMatrix[VX][VY] = ( t.mMatrix[VZ][VY] * t.mMatrix[VX][VZ] - t.mMatrix[VZ][VZ] * t.mMatrix[VX][VY] ) / det;
- mMatrix[VY][VY] = ( t.mMatrix[VZ][VZ] * t.mMatrix[VX][VX] - t.mMatrix[VZ][VX] * t.mMatrix[VX][VZ] ) / det;
- mMatrix[VZ][VY] = ( t.mMatrix[VZ][VX] * t.mMatrix[VX][VY] - t.mMatrix[VZ][VY] * t.mMatrix[VX][VX] ) / det;
- mMatrix[VX][VZ] = ( t.mMatrix[VX][VY] * t.mMatrix[VY][VZ] - t.mMatrix[VX][VZ] * t.mMatrix[VY][VY] ) / det;
- mMatrix[VY][VZ] = ( t.mMatrix[VX][VZ] * t.mMatrix[VY][VX] - t.mMatrix[VX][VX] * t.mMatrix[VY][VZ] ) / det;
- mMatrix[VZ][VZ] = ( t.mMatrix[VX][VX] * t.mMatrix[VY][VY] - t.mMatrix[VX][VY] * t.mMatrix[VY][VX] ) / det;
- }
+ // fails silently if determinant is zero too small
+ F32 det = determinant();
+ const F32 VERY_SMALL_DETERMINANT = 0.000001f;
+ if (fabs(det) > VERY_SMALL_DETERMINANT)
+ {
+ // invertiable
+ LLMatrix3 t(*this);
+ mMatrix[VX][VX] = ( t.mMatrix[VY][VY] * t.mMatrix[VZ][VZ] - t.mMatrix[VY][VZ] * t.mMatrix[VZ][VY] ) / det;
+ mMatrix[VY][VX] = ( t.mMatrix[VY][VZ] * t.mMatrix[VZ][VX] - t.mMatrix[VY][VX] * t.mMatrix[VZ][VZ] ) / det;
+ mMatrix[VZ][VX] = ( t.mMatrix[VY][VX] * t.mMatrix[VZ][VY] - t.mMatrix[VY][VY] * t.mMatrix[VZ][VX] ) / det;
+ mMatrix[VX][VY] = ( t.mMatrix[VZ][VY] * t.mMatrix[VX][VZ] - t.mMatrix[VZ][VZ] * t.mMatrix[VX][VY] ) / det;
+ mMatrix[VY][VY] = ( t.mMatrix[VZ][VZ] * t.mMatrix[VX][VX] - t.mMatrix[VZ][VX] * t.mMatrix[VX][VZ] ) / det;
+ mMatrix[VZ][VY] = ( t.mMatrix[VZ][VX] * t.mMatrix[VX][VY] - t.mMatrix[VZ][VY] * t.mMatrix[VX][VX] ) / det;
+ mMatrix[VX][VZ] = ( t.mMatrix[VX][VY] * t.mMatrix[VY][VZ] - t.mMatrix[VX][VZ] * t.mMatrix[VY][VY] ) / det;
+ mMatrix[VY][VZ] = ( t.mMatrix[VX][VZ] * t.mMatrix[VY][VX] - t.mMatrix[VX][VX] * t.mMatrix[VY][VZ] ) / det;
+ mMatrix[VZ][VZ] = ( t.mMatrix[VX][VX] * t.mMatrix[VY][VY] - t.mMatrix[VX][VY] * t.mMatrix[VY][VX] ) / det;
+ }
}
// does not assume a rotation matrix, and does not divide by determinant, assuming results will be renormalized
-const LLMatrix3& LLMatrix3::adjointTranspose()
+const LLMatrix3& LLMatrix3::adjointTranspose()
{
- LLMatrix3 adjoint_transpose;
- adjoint_transpose.mMatrix[VX][VX] = mMatrix[VY][VY] * mMatrix[VZ][VZ] - mMatrix[VY][VZ] * mMatrix[VZ][VY] ;
- adjoint_transpose.mMatrix[VY][VX] = mMatrix[VY][VZ] * mMatrix[VZ][VX] - mMatrix[VY][VX] * mMatrix[VZ][VZ] ;
- adjoint_transpose.mMatrix[VZ][VX] = mMatrix[VY][VX] * mMatrix[VZ][VY] - mMatrix[VY][VY] * mMatrix[VZ][VX] ;
- adjoint_transpose.mMatrix[VX][VY] = mMatrix[VZ][VY] * mMatrix[VX][VZ] - mMatrix[VZ][VZ] * mMatrix[VX][VY] ;
- adjoint_transpose.mMatrix[VY][VY] = mMatrix[VZ][VZ] * mMatrix[VX][VX] - mMatrix[VZ][VX] * mMatrix[VX][VZ] ;
- adjoint_transpose.mMatrix[VZ][VY] = mMatrix[VZ][VX] * mMatrix[VX][VY] - mMatrix[VZ][VY] * mMatrix[VX][VX] ;
- adjoint_transpose.mMatrix[VX][VZ] = mMatrix[VX][VY] * mMatrix[VY][VZ] - mMatrix[VX][VZ] * mMatrix[VY][VY] ;
- adjoint_transpose.mMatrix[VY][VZ] = mMatrix[VX][VZ] * mMatrix[VY][VX] - mMatrix[VX][VX] * mMatrix[VY][VZ] ;
- adjoint_transpose.mMatrix[VZ][VZ] = mMatrix[VX][VX] * mMatrix[VY][VY] - mMatrix[VX][VY] * mMatrix[VY][VX] ;
+ LLMatrix3 adjoint_transpose;
+ adjoint_transpose.mMatrix[VX][VX] = mMatrix[VY][VY] * mMatrix[VZ][VZ] - mMatrix[VY][VZ] * mMatrix[VZ][VY] ;
+ adjoint_transpose.mMatrix[VY][VX] = mMatrix[VY][VZ] * mMatrix[VZ][VX] - mMatrix[VY][VX] * mMatrix[VZ][VZ] ;
+ adjoint_transpose.mMatrix[VZ][VX] = mMatrix[VY][VX] * mMatrix[VZ][VY] - mMatrix[VY][VY] * mMatrix[VZ][VX] ;
+ adjoint_transpose.mMatrix[VX][VY] = mMatrix[VZ][VY] * mMatrix[VX][VZ] - mMatrix[VZ][VZ] * mMatrix[VX][VY] ;
+ adjoint_transpose.mMatrix[VY][VY] = mMatrix[VZ][VZ] * mMatrix[VX][VX] - mMatrix[VZ][VX] * mMatrix[VX][VZ] ;
+ adjoint_transpose.mMatrix[VZ][VY] = mMatrix[VZ][VX] * mMatrix[VX][VY] - mMatrix[VZ][VY] * mMatrix[VX][VX] ;
+ adjoint_transpose.mMatrix[VX][VZ] = mMatrix[VX][VY] * mMatrix[VY][VZ] - mMatrix[VX][VZ] * mMatrix[VY][VY] ;
+ adjoint_transpose.mMatrix[VY][VZ] = mMatrix[VX][VZ] * mMatrix[VY][VX] - mMatrix[VX][VX] * mMatrix[VY][VZ] ;
+ adjoint_transpose.mMatrix[VZ][VZ] = mMatrix[VX][VX] * mMatrix[VY][VY] - mMatrix[VX][VY] * mMatrix[VY][VX] ;
- *this = adjoint_transpose;
- return *this;
+ *this = adjoint_transpose;
+ return *this;
}
// SJB: This code is correct for a logicly stored (non-transposed) matrix;
-// Our matrices are stored transposed, OpenGL style, so this generates the
-// INVERSE quaternion (-x, -y, -z, w)!
-// Because we use similar logic in LLQuaternion::getMatrix3,
-// we are internally consistant so everything works OK :)
-LLQuaternion LLMatrix3::quaternion() const
-{
- LLQuaternion quat;
- F32 tr, s, q[4];
- U32 i, j, k;
- U32 nxt[3] = {1, 2, 0};
-
- tr = mMatrix[0][0] + mMatrix[1][1] + mMatrix[2][2];
-
- // check the diagonal
- if (tr > 0.f)
- {
- s = (F32)sqrt (tr + 1.f);
- quat.mQ[VS] = s / 2.f;
- s = 0.5f / s;
- quat.mQ[VX] = (mMatrix[1][2] - mMatrix[2][1]) * s;
- quat.mQ[VY] = (mMatrix[2][0] - mMatrix[0][2]) * s;
- quat.mQ[VZ] = (mMatrix[0][1] - mMatrix[1][0]) * s;
- }
- else
- {
- // diagonal is negative
- i = 0;
- if (mMatrix[1][1] > mMatrix[0][0])
- i = 1;
- if (mMatrix[2][2] > mMatrix[i][i])
- i = 2;
-
- j = nxt[i];
- k = nxt[j];
-
-
- s = (F32)sqrt ((mMatrix[i][i] - (mMatrix[j][j] + mMatrix[k][k])) + 1.f);
-
- q[i] = s * 0.5f;
-
- if (s != 0.f)
- s = 0.5f / s;
-
- q[3] = (mMatrix[j][k] - mMatrix[k][j]) * s;
- q[j] = (mMatrix[i][j] + mMatrix[j][i]) * s;
- q[k] = (mMatrix[i][k] + mMatrix[k][i]) * s;
-
- quat.setQuat(q);
- }
- return quat;
-}
-
-const LLMatrix3& LLMatrix3::setRot(const F32 angle, const LLVector3 &vec)
-{
- setRot(LLQuaternion(angle, vec));
- return *this;
-}
-
-const LLMatrix3& LLMatrix3::setRot(const F32 roll, const F32 pitch, const F32 yaw)
-{
- // Rotates RH about x-axis by 'roll' then
- // rotates RH about the old y-axis by 'pitch' then
- // rotates RH about the original z-axis by 'yaw'.
- // .
- // /|\ yaw axis
- // | __.
- // ._ ___| /| pitch axis
- // _||\ \\ |-. /
- // \|| \_______\_|__\_/_______
- // | _ _ o o o_o_o_o o /_\_ ________\ roll axis
- // // /_______/ /__________> /
- // /_,-' // /
- // /__,-'
-
- F32 cx, sx, cy, sy, cz, sz;
- F32 cxsy, sxsy;
+// Our matrices are stored transposed, OpenGL style, so this generates the
+// INVERSE quaternion (-x, -y, -z, w)!
+// Because we use similar logic in LLQuaternion::getMatrix3,
+// we are internally consistant so everything works OK :)
+LLQuaternion LLMatrix3::quaternion() const
+{
+ LLQuaternion quat;
+ F32 tr, s, q[4];
+ U32 i, j, k;
+ U32 nxt[3] = {1, 2, 0};
+
+ tr = mMatrix[0][0] + mMatrix[1][1] + mMatrix[2][2];
+
+ // check the diagonal
+ if (tr > 0.f)
+ {
+ s = (F32)sqrt (tr + 1.f);
+ quat.mQ[VS] = s / 2.f;
+ s = 0.5f / s;
+ quat.mQ[VX] = (mMatrix[1][2] - mMatrix[2][1]) * s;
+ quat.mQ[VY] = (mMatrix[2][0] - mMatrix[0][2]) * s;
+ quat.mQ[VZ] = (mMatrix[0][1] - mMatrix[1][0]) * s;
+ }
+ else
+ {
+ // diagonal is negative
+ i = 0;
+ if (mMatrix[1][1] > mMatrix[0][0])
+ i = 1;
+ if (mMatrix[2][2] > mMatrix[i][i])
+ i = 2;
+
+ j = nxt[i];
+ k = nxt[j];
+
+
+ s = (F32)sqrt ((mMatrix[i][i] - (mMatrix[j][j] + mMatrix[k][k])) + 1.f);
+
+ q[i] = s * 0.5f;
+
+ if (s != 0.f)
+ s = 0.5f / s;
+
+ q[3] = (mMatrix[j][k] - mMatrix[k][j]) * s;
+ q[j] = (mMatrix[i][j] + mMatrix[j][i]) * s;
+ q[k] = (mMatrix[i][k] + mMatrix[k][i]) * s;
+
+ quat.setQuat(q);
+ }
+ return quat;
+}
+
+const LLMatrix3& LLMatrix3::setRot(const F32 angle, const LLVector3 &vec)
+{
+ setRot(LLQuaternion(angle, vec));
+ return *this;
+}
+
+const LLMatrix3& LLMatrix3::setRot(const F32 roll, const F32 pitch, const F32 yaw)
+{
+ // Rotates RH about x-axis by 'roll' then
+ // rotates RH about the old y-axis by 'pitch' then
+ // rotates RH about the original z-axis by 'yaw'.
+ // .
+ // /|\ yaw axis
+ // | __.
+ // ._ ___| /| pitch axis
+ // _||\ \\ |-. /
+ // \|| \_______\_|__\_/_______
+ // | _ _ o o o_o_o_o o /_\_ ________\ roll axis
+ // // /_______/ /__________> /
+ // /_,-' // /
+ // /__,-'
+
+ F32 cx, sx, cy, sy, cz, sz;
+ F32 cxsy, sxsy;
cx = (F32)cos(roll); //A
sx = (F32)sin(roll); //B
@@ -320,7 +320,7 @@ const LLMatrix3& LLMatrix3::setRot(const F32 roll, const F32 pitch, const F32 ya
sz = (F32)sin(yaw); //F
cxsy = cx * sy; //AD
- sxsy = sx * sy; //BD
+ sxsy = sx * sy; //BD
mMatrix[0][0] = cy * cz;
mMatrix[1][0] = -cy * sz;
@@ -331,118 +331,118 @@ const LLMatrix3& LLMatrix3::setRot(const F32 roll, const F32 pitch, const F32 ya
mMatrix[0][2] = -cxsy * cz + sx * sz;
mMatrix[1][2] = cxsy * sz + sx * cz;
mMatrix[2][2] = cx * cy;
- return *this;
+ return *this;
}
-const LLMatrix3& LLMatrix3::setRot(const LLQuaternion &q)
+const LLMatrix3& LLMatrix3::setRot(const LLQuaternion &q)
{
- *this = q.getMatrix3();
- return *this;
+ *this = q.getMatrix3();
+ return *this;
}
-const LLMatrix3& LLMatrix3::setRows(const LLVector3 &fwd, const LLVector3 &left, const LLVector3 &up)
+const LLMatrix3& LLMatrix3::setRows(const LLVector3 &fwd, const LLVector3 &left, const LLVector3 &up)
{
- mMatrix[0][0] = fwd.mV[0];
- mMatrix[0][1] = fwd.mV[1];
- mMatrix[0][2] = fwd.mV[2];
+ mMatrix[0][0] = fwd.mV[0];
+ mMatrix[0][1] = fwd.mV[1];
+ mMatrix[0][2] = fwd.mV[2];
- mMatrix[1][0] = left.mV[0];
- mMatrix[1][1] = left.mV[1];
- mMatrix[1][2] = left.mV[2];
+ mMatrix[1][0] = left.mV[0];
+ mMatrix[1][1] = left.mV[1];
+ mMatrix[1][2] = left.mV[2];
- mMatrix[2][0] = up.mV[0];
- mMatrix[2][1] = up.mV[1];
- mMatrix[2][2] = up.mV[2];
+ mMatrix[2][0] = up.mV[0];
+ mMatrix[2][1] = up.mV[1];
+ mMatrix[2][2] = up.mV[2];
- return *this;
+ return *this;
}
const LLMatrix3& LLMatrix3::setRow( U32 rowIndex, const LLVector3& row )
{
- llassert( rowIndex >= 0 && rowIndex < NUM_VALUES_IN_MAT3 );
+ llassert( rowIndex >= 0 && rowIndex < NUM_VALUES_IN_MAT3 );
- mMatrix[rowIndex][0] = row[0];
- mMatrix[rowIndex][1] = row[1];
- mMatrix[rowIndex][2] = row[2];
+ mMatrix[rowIndex][0] = row[0];
+ mMatrix[rowIndex][1] = row[1];
+ mMatrix[rowIndex][2] = row[2];
- return *this;
+ return *this;
}
const LLMatrix3& LLMatrix3::setCol( U32 colIndex, const LLVector3& col )
{
- llassert( colIndex >= 0 && colIndex < NUM_VALUES_IN_MAT3 );
+ llassert( colIndex >= 0 && colIndex < NUM_VALUES_IN_MAT3 );
- mMatrix[0][colIndex] = col[0];
- mMatrix[1][colIndex] = col[1];
- mMatrix[2][colIndex] = col[2];
+ mMatrix[0][colIndex] = col[0];
+ mMatrix[1][colIndex] = col[1];
+ mMatrix[2][colIndex] = col[2];
- return *this;
+ return *this;
}
-const LLMatrix3& LLMatrix3::rotate(const F32 angle, const LLVector3 &vec)
+const LLMatrix3& LLMatrix3::rotate(const F32 angle, const LLVector3 &vec)
{
- LLMatrix3 mat(angle, vec);
- *this *= mat;
- return *this;
+ LLMatrix3 mat(angle, vec);
+ *this *= mat;
+ return *this;
}
-const LLMatrix3& LLMatrix3::rotate(const F32 roll, const F32 pitch, const F32 yaw)
+const LLMatrix3& LLMatrix3::rotate(const F32 roll, const F32 pitch, const F32 yaw)
{
- LLMatrix3 mat(roll, pitch, yaw);
- *this *= mat;
- return *this;
+ LLMatrix3 mat(roll, pitch, yaw);
+ *this *= mat;
+ return *this;
}
-const LLMatrix3& LLMatrix3::rotate(const LLQuaternion &q)
+const LLMatrix3& LLMatrix3::rotate(const LLQuaternion &q)
{
- LLMatrix3 mat(q);
- *this *= mat;
- return *this;
+ LLMatrix3 mat(q);
+ *this *= mat;
+ return *this;
}
void LLMatrix3::add(const LLMatrix3& other_matrix)
{
- for (S32 i = 0; i < 3; ++i)
- {
- for (S32 j = 0; j < 3; ++j)
- {
- mMatrix[i][j] += other_matrix.mMatrix[i][j];
- }
- }
+ for (S32 i = 0; i < 3; ++i)
+ {
+ for (S32 j = 0; j < 3; ++j)
+ {
+ mMatrix[i][j] += other_matrix.mMatrix[i][j];
+ }
+ }
}
-LLVector3 LLMatrix3::getFwdRow() const
+LLVector3 LLMatrix3::getFwdRow() const
{
- return LLVector3(mMatrix[VX]);
+ return LLVector3(mMatrix[VX]);
}
-LLVector3 LLMatrix3::getLeftRow() const
+LLVector3 LLMatrix3::getLeftRow() const
{
- return LLVector3(mMatrix[VY]);
+ return LLVector3(mMatrix[VY]);
}
-LLVector3 LLMatrix3::getUpRow() const
+LLVector3 LLMatrix3::getUpRow() const
{
- return LLVector3(mMatrix[VZ]);
+ return LLVector3(mMatrix[VZ]);
}
-const LLMatrix3& LLMatrix3::orthogonalize()
+const LLMatrix3& LLMatrix3::orthogonalize()
{
- LLVector3 x_axis(mMatrix[VX]);
- LLVector3 y_axis(mMatrix[VY]);
- LLVector3 z_axis(mMatrix[VZ]);
+ LLVector3 x_axis(mMatrix[VX]);
+ LLVector3 y_axis(mMatrix[VY]);
+ LLVector3 z_axis(mMatrix[VZ]);
- x_axis.normVec();
- y_axis -= x_axis * (x_axis * y_axis);
- y_axis.normVec();
- z_axis = x_axis % y_axis;
- setRows(x_axis, y_axis, z_axis);
- return (*this);
+ x_axis.normVec();
+ y_axis -= x_axis * (x_axis * y_axis);
+ y_axis.normVec();
+ z_axis = x_axis % y_axis;
+ setRows(x_axis, y_axis, z_axis);
+ return (*this);
}
@@ -450,141 +450,141 @@ const LLMatrix3& LLMatrix3::orthogonalize()
LLMatrix3 operator*(const LLMatrix3 &a, const LLMatrix3 &b)
{
- U32 i, j;
- LLMatrix3 mat;
- for (i = 0; i < NUM_VALUES_IN_MAT3; i++)
- {
- for (j = 0; j < NUM_VALUES_IN_MAT3; j++)
- {
- mat.mMatrix[j][i] = a.mMatrix[j][0] * b.mMatrix[0][i] +
- a.mMatrix[j][1] * b.mMatrix[1][i] +
- a.mMatrix[j][2] * b.mMatrix[2][i];
- }
- }
- return mat;
-}
-
-/* Not implemented to help enforce code consistency with the syntax of
+ U32 i, j;
+ LLMatrix3 mat;
+ for (i = 0; i < NUM_VALUES_IN_MAT3; i++)
+ {
+ for (j = 0; j < NUM_VALUES_IN_MAT3; j++)
+ {
+ mat.mMatrix[j][i] = a.mMatrix[j][0] * b.mMatrix[0][i] +
+ a.mMatrix[j][1] * b.mMatrix[1][i] +
+ a.mMatrix[j][2] * b.mMatrix[2][i];
+ }
+ }
+ return mat;
+}
+
+/* Not implemented to help enforce code consistency with the syntax of
row-major notation. This is a Good Thing.
LLVector3 operator*(const LLMatrix3 &a, const LLVector3 &b)
{
- LLVector3 vec;
- // matrix operates "from the left" on column vector
- vec.mV[VX] = a.mMatrix[VX][VX] * b.mV[VX] +
- a.mMatrix[VX][VY] * b.mV[VY] +
- a.mMatrix[VX][VZ] * b.mV[VZ];
-
- vec.mV[VY] = a.mMatrix[VY][VX] * b.mV[VX] +
- a.mMatrix[VY][VY] * b.mV[VY] +
- a.mMatrix[VY][VZ] * b.mV[VZ];
-
- vec.mV[VZ] = a.mMatrix[VZ][VX] * b.mV[VX] +
- a.mMatrix[VZ][VY] * b.mV[VY] +
- a.mMatrix[VZ][VZ] * b.mV[VZ];
- return vec;
+ LLVector3 vec;
+ // matrix operates "from the left" on column vector
+ vec.mV[VX] = a.mMatrix[VX][VX] * b.mV[VX] +
+ a.mMatrix[VX][VY] * b.mV[VY] +
+ a.mMatrix[VX][VZ] * b.mV[VZ];
+
+ vec.mV[VY] = a.mMatrix[VY][VX] * b.mV[VX] +
+ a.mMatrix[VY][VY] * b.mV[VY] +
+ a.mMatrix[VY][VZ] * b.mV[VZ];
+
+ vec.mV[VZ] = a.mMatrix[VZ][VX] * b.mV[VX] +
+ a.mMatrix[VZ][VY] * b.mV[VY] +
+ a.mMatrix[VZ][VZ] * b.mV[VZ];
+ return vec;
}
*/
LLVector3 operator*(const LLVector3 &a, const LLMatrix3 &b)
{
- // matrix operates "from the right" on row vector
- return LLVector3(
- a.mV[VX] * b.mMatrix[VX][VX] +
- a.mV[VY] * b.mMatrix[VY][VX] +
- a.mV[VZ] * b.mMatrix[VZ][VX],
-
- a.mV[VX] * b.mMatrix[VX][VY] +
- a.mV[VY] * b.mMatrix[VY][VY] +
- a.mV[VZ] * b.mMatrix[VZ][VY],
-
- a.mV[VX] * b.mMatrix[VX][VZ] +
- a.mV[VY] * b.mMatrix[VY][VZ] +
- a.mV[VZ] * b.mMatrix[VZ][VZ] );
+ // matrix operates "from the right" on row vector
+ return LLVector3(
+ a.mV[VX] * b.mMatrix[VX][VX] +
+ a.mV[VY] * b.mMatrix[VY][VX] +
+ a.mV[VZ] * b.mMatrix[VZ][VX],
+
+ a.mV[VX] * b.mMatrix[VX][VY] +
+ a.mV[VY] * b.mMatrix[VY][VY] +
+ a.mV[VZ] * b.mMatrix[VZ][VY],
+
+ a.mV[VX] * b.mMatrix[VX][VZ] +
+ a.mV[VY] * b.mMatrix[VY][VZ] +
+ a.mV[VZ] * b.mMatrix[VZ][VZ] );
}
LLVector3d operator*(const LLVector3d &a, const LLMatrix3 &b)
{
- // matrix operates "from the right" on row vector
- return LLVector3d(
- a.mdV[VX] * b.mMatrix[VX][VX] +
- a.mdV[VY] * b.mMatrix[VY][VX] +
- a.mdV[VZ] * b.mMatrix[VZ][VX],
-
- a.mdV[VX] * b.mMatrix[VX][VY] +
- a.mdV[VY] * b.mMatrix[VY][VY] +
- a.mdV[VZ] * b.mMatrix[VZ][VY],
-
- a.mdV[VX] * b.mMatrix[VX][VZ] +
- a.mdV[VY] * b.mMatrix[VY][VZ] +
- a.mdV[VZ] * b.mMatrix[VZ][VZ] );
+ // matrix operates "from the right" on row vector
+ return LLVector3d(
+ a.mdV[VX] * b.mMatrix[VX][VX] +
+ a.mdV[VY] * b.mMatrix[VY][VX] +
+ a.mdV[VZ] * b.mMatrix[VZ][VX],
+
+ a.mdV[VX] * b.mMatrix[VX][VY] +
+ a.mdV[VY] * b.mMatrix[VY][VY] +
+ a.mdV[VZ] * b.mMatrix[VZ][VY],
+
+ a.mdV[VX] * b.mMatrix[VX][VZ] +
+ a.mdV[VY] * b.mMatrix[VY][VZ] +
+ a.mdV[VZ] * b.mMatrix[VZ][VZ] );
}
bool operator==(const LLMatrix3 &a, const LLMatrix3 &b)
{
- U32 i, j;
- for (i = 0; i < NUM_VALUES_IN_MAT3; i++)
- {
- for (j = 0; j < NUM_VALUES_IN_MAT3; j++)
- {
- if (a.mMatrix[j][i] != b.mMatrix[j][i])
- return false;
- }
- }
- return true;
+ U32 i, j;
+ for (i = 0; i < NUM_VALUES_IN_MAT3; i++)
+ {
+ for (j = 0; j < NUM_VALUES_IN_MAT3; j++)
+ {
+ if (a.mMatrix[j][i] != b.mMatrix[j][i])
+ return false;
+ }
+ }
+ return true;
}
bool operator!=(const LLMatrix3 &a, const LLMatrix3 &b)
{
- U32 i, j;
- for (i = 0; i < NUM_VALUES_IN_MAT3; i++)
- {
- for (j = 0; j < NUM_VALUES_IN_MAT3; j++)
- {
- if (a.mMatrix[j][i] != b.mMatrix[j][i])
- return true;
- }
- }
- return false;
+ U32 i, j;
+ for (i = 0; i < NUM_VALUES_IN_MAT3; i++)
+ {
+ for (j = 0; j < NUM_VALUES_IN_MAT3; j++)
+ {
+ if (a.mMatrix[j][i] != b.mMatrix[j][i])
+ return true;
+ }
+ }
+ return false;
}
const LLMatrix3& operator*=(LLMatrix3 &a, const LLMatrix3 &b)
{
- U32 i, j;
- LLMatrix3 mat;
- for (i = 0; i < NUM_VALUES_IN_MAT3; i++)
- {
- for (j = 0; j < NUM_VALUES_IN_MAT3; j++)
- {
- mat.mMatrix[j][i] = a.mMatrix[j][0] * b.mMatrix[0][i] +
- a.mMatrix[j][1] * b.mMatrix[1][i] +
- a.mMatrix[j][2] * b.mMatrix[2][i];
- }
- }
- a = mat;
- return a;
+ U32 i, j;
+ LLMatrix3 mat;
+ for (i = 0; i < NUM_VALUES_IN_MAT3; i++)
+ {
+ for (j = 0; j < NUM_VALUES_IN_MAT3; j++)
+ {
+ mat.mMatrix[j][i] = a.mMatrix[j][0] * b.mMatrix[0][i] +
+ a.mMatrix[j][1] * b.mMatrix[1][i] +
+ a.mMatrix[j][2] * b.mMatrix[2][i];
+ }
+ }
+ a = mat;
+ return a;
}
const LLMatrix3& operator*=(LLMatrix3 &a, F32 scalar )
{
- for( U32 i = 0; i < NUM_VALUES_IN_MAT3; ++i )
- {
- for( U32 j = 0; j < NUM_VALUES_IN_MAT3; ++j )
- {
- a.mMatrix[i][j] *= scalar;
- }
- }
+ for( U32 i = 0; i < NUM_VALUES_IN_MAT3; ++i )
+ {
+ for( U32 j = 0; j < NUM_VALUES_IN_MAT3; ++j )
+ {
+ a.mMatrix[i][j] *= scalar;
+ }
+ }
- return a;
+ return a;
}
-std::ostream& operator<<(std::ostream& s, const LLMatrix3 &a)
+std::ostream& operator<<(std::ostream& s, const LLMatrix3 &a)
{
- s << "{ "
- << a.mMatrix[VX][VX] << ", " << a.mMatrix[VX][VY] << ", " << a.mMatrix[VX][VZ] << "; "
- << a.mMatrix[VY][VX] << ", " << a.mMatrix[VY][VY] << ", " << a.mMatrix[VY][VZ] << "; "
- << a.mMatrix[VZ][VX] << ", " << a.mMatrix[VZ][VY] << ", " << a.mMatrix[VZ][VZ]
- << " }";
- return s;
+ s << "{ "
+ << a.mMatrix[VX][VX] << ", " << a.mMatrix[VX][VY] << ", " << a.mMatrix[VX][VZ] << "; "
+ << a.mMatrix[VY][VX] << ", " << a.mMatrix[VY][VY] << ", " << a.mMatrix[VY][VZ] << "; "
+ << a.mMatrix[VZ][VX] << ", " << a.mMatrix[VZ][VY] << ", " << a.mMatrix[VZ][VZ]
+ << " }";
+ return s;
}