forked from bartvdbraak/blender
cb89decfdc
the features that are needed to run the game. Compile tested with scons, make, but not cmake, that seems to have an issue not related to these changes. The changes include: * GLSL support in the viewport and game engine, enable in the game menu in textured draw mode. * Synced and merged part of the duplicated blender and gameengine/ gameplayer drawing code. * Further refactoring of game engine drawing code, especially mesh storage changed a lot. * Optimizations in game engine armatures to avoid recomputations. * A python function to get the framerate estimate in game. * An option take object color into account in materials. * An option to restrict shadow casters to a lamp's layers. * Increase from 10 to 18 texture slots for materials, lamps, word. An extra texture slot shows up once the last slot is used. * Memory limit for undo, not enabled by default yet because it needs the .B.blend to be changed. * Multiple undo for image painting. * An offset for dupligroups, so not all objects in a group have to be at the origin.
250 lines
8.2 KiB
C++
250 lines
8.2 KiB
C++
/**
|
|
* $Id$
|
|
* ***** BEGIN GPL LICENSE BLOCK *****
|
|
*
|
|
* 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.
|
|
*
|
|
* The Original Code is Copyright (C) 2001-2002 by NaN Holding BV.
|
|
* All rights reserved.
|
|
*
|
|
* The Original Code is: all of this file.
|
|
*
|
|
* Contributor(s): none yet.
|
|
*
|
|
* ***** END GPL LICENSE BLOCK *****
|
|
*/
|
|
|
|
/**
|
|
|
|
* $Id$
|
|
* Copyright (C) 2001 NaN Technologies B.V.
|
|
* A 4x4 matrix compatible with other stuff.
|
|
*/
|
|
|
|
#ifndef MT_MATRIX4X4_H
|
|
#define MT_MATRIX4X4_H
|
|
|
|
#include <MT_assert.h>
|
|
|
|
#include "MT_Vector4.h"
|
|
#include "MT_Transform.h"
|
|
|
|
// Row-major 4x4 matrix
|
|
|
|
class MT_Matrix4x4 {
|
|
public:
|
|
/**
|
|
* Empty contructor.
|
|
*/
|
|
MT_Matrix4x4() {}
|
|
/**
|
|
* Initialize all fields with the values pointed at by m. A
|
|
* contigous block of 16 values is read. */
|
|
MT_Matrix4x4(const float *m) { setValue(m); }
|
|
/**
|
|
* Initialize all fields with the values pointed at by m. A
|
|
* contigous block of 16 values is read. */
|
|
MT_Matrix4x4(const double *m) { setValue(m); }
|
|
|
|
/**
|
|
* Initialise with these 16 explicit values.
|
|
*/
|
|
MT_Matrix4x4(MT_Scalar xx, MT_Scalar xy, MT_Scalar xz, MT_Scalar xw,
|
|
MT_Scalar yx, MT_Scalar yy, MT_Scalar yz, MT_Scalar yw,
|
|
MT_Scalar zx, MT_Scalar zy, MT_Scalar zz, MT_Scalar zw,
|
|
MT_Scalar wx, MT_Scalar wy, MT_Scalar wz, MT_Scalar ww) {
|
|
setValue(xx, xy, xz, xw,
|
|
yx, yy, yz, yw,
|
|
zx, zy, zz, zw,
|
|
wx, wy, wz, ww);
|
|
}
|
|
|
|
/**
|
|
* Initialize from an MT_Transform.
|
|
*/
|
|
MT_Matrix4x4(const MT_Transform &t) {
|
|
|
|
const MT_Matrix3x3 &basis = t.getBasis();
|
|
const MT_Vector3 &origin = t.getOrigin();
|
|
|
|
setValue(
|
|
basis[0][0],basis[0][1],basis[0][2],origin[0],
|
|
basis[1][0],basis[1][1],basis[1][2],origin[1],
|
|
basis[2][0],basis[2][1],basis[2][2],origin[2],
|
|
MT_Scalar(0),MT_Scalar(0),MT_Scalar(0),MT_Scalar(1)
|
|
);
|
|
}
|
|
|
|
/**
|
|
* Get the i-th row.
|
|
*/
|
|
MT_Vector4& operator[](int i) { return m_el[i]; }
|
|
/**
|
|
* Get the i-th row.
|
|
*/
|
|
const MT_Vector4& operator[](int i) const { return m_el[i]; }
|
|
|
|
/**
|
|
* Set the matrix to the values pointer at by m. A contiguous
|
|
* block of 16 values is copied. */
|
|
void setValue(const float *m) {
|
|
m_el[0][0] = *m++; m_el[1][0] = *m++; m_el[2][0] = *m++; m_el[3][0] = *m++;
|
|
m_el[0][1] = *m++; m_el[1][1] = *m++; m_el[2][1] = *m++; m_el[3][1] = *m++;
|
|
m_el[0][2] = *m++; m_el[1][2] = *m++; m_el[2][2] = *m++; m_el[3][2] = *m++;
|
|
m_el[0][3] = *m++; m_el[1][3] = *m++; m_el[2][3] = *m++; m_el[3][3] = *m;
|
|
}
|
|
|
|
/**
|
|
* Set the matrix to the values pointer at by m. A contiguous
|
|
* block of 16 values is copied.
|
|
*/
|
|
void setValue(const double *m) {
|
|
m_el[0][0] = *m++; m_el[1][0] = *m++; m_el[2][0] = *m++; m_el[3][0] = *m++;
|
|
m_el[0][1] = *m++; m_el[1][1] = *m++; m_el[2][1] = *m++; m_el[3][1] = *m++;
|
|
m_el[0][2] = *m++; m_el[1][2] = *m++; m_el[2][2] = *m++; m_el[3][2] = *m++;
|
|
m_el[0][3] = *m++; m_el[1][3] = *m++; m_el[2][3] = *m++; m_el[3][3] = *m;
|
|
}
|
|
|
|
/**
|
|
* Set the matrix to these 16 explicit values.
|
|
*/
|
|
void setValue(MT_Scalar xx, MT_Scalar xy, MT_Scalar xz, MT_Scalar xw,
|
|
MT_Scalar yx, MT_Scalar yy, MT_Scalar yz, MT_Scalar yw,
|
|
MT_Scalar zx, MT_Scalar zy, MT_Scalar zz, MT_Scalar zw,
|
|
MT_Scalar wx, MT_Scalar wy, MT_Scalar wz, MT_Scalar ww) {
|
|
m_el[0][0] = xx; m_el[0][1] = xy; m_el[0][2] = xz; m_el[0][3] = xw;
|
|
m_el[1][0] = yx; m_el[1][1] = yy; m_el[1][2] = yz; m_el[1][3] = yw;
|
|
m_el[2][0] = zx; m_el[2][1] = zy; m_el[2][2] = zz; m_el[2][3] = zw;
|
|
m_el[3][0] = wx; m_el[3][1] = wy; m_el[3][2] = wz; m_el[3][3] = ww;
|
|
}
|
|
|
|
/**
|
|
* Scale the columns of this matrix with x, y, z, w respectively.
|
|
*/
|
|
void scale(MT_Scalar x, MT_Scalar y, MT_Scalar z, MT_Scalar w) {
|
|
m_el[0][0] *= x; m_el[0][1] *= y; m_el[0][2] *= z; m_el[0][3] *= w;
|
|
m_el[1][0] *= x; m_el[1][1] *= y; m_el[1][2] *= z; m_el[1][3] *= w;
|
|
m_el[2][0] *= x; m_el[2][1] *= y; m_el[2][2] *= z; m_el[2][3] *= w;
|
|
m_el[3][0] *= x; m_el[3][1] *= y; m_el[3][2] *= z; m_el[3][3] *= w;
|
|
}
|
|
|
|
/**
|
|
* Return a column-scaled version of this matrix.
|
|
*/
|
|
MT_Matrix4x4 scaled(MT_Scalar x, MT_Scalar y, MT_Scalar z, MT_Scalar w) const {
|
|
return MT_Matrix4x4(m_el[0][0] * x, m_el[0][1] * y, m_el[0][2] * z, m_el[0][3] * w,
|
|
m_el[1][0] * x, m_el[1][1] * y, m_el[1][2] * z, m_el[1][3] * w,
|
|
m_el[2][0] * x, m_el[2][1] * y, m_el[2][2] * z, m_el[2][3] * w,
|
|
m_el[3][0] * x, m_el[3][1] * y, m_el[3][2] * z, m_el[3][3] * w);
|
|
}
|
|
|
|
/**
|
|
* Set this matrix to I.
|
|
*/
|
|
void setIdentity() {
|
|
setValue(MT_Scalar(1.0), MT_Scalar(0.0), MT_Scalar(0.0), MT_Scalar(0.0),
|
|
MT_Scalar(0.0), MT_Scalar(1.0), MT_Scalar(0.0), MT_Scalar(0.0),
|
|
MT_Scalar(0.0), MT_Scalar(0.0), MT_Scalar(1.0), MT_Scalar(0.0),
|
|
MT_Scalar(0.0), MT_Scalar(0.0), MT_Scalar(0.0), MT_Scalar(1.0));
|
|
}
|
|
|
|
/**
|
|
* Read the element from row i, column j.
|
|
*/
|
|
float getElement(int i, int j) {
|
|
return (float) m_el[i][j];
|
|
}
|
|
|
|
/**
|
|
* Copy the contents to a contiguous block of 16 floats.
|
|
*/
|
|
void getValue(float *m) const {
|
|
*m++ = (float) m_el[0][0]; *m++ = (float) m_el[1][0]; *m++ = (float) m_el[2][0]; *m++ = (float) m_el[3][0];
|
|
*m++ = (float) m_el[0][1]; *m++ = (float) m_el[1][1]; *m++ = (float) m_el[2][1]; *m++ = (float) m_el[3][1];
|
|
*m++ = (float) m_el[0][2]; *m++ = (float) m_el[1][2]; *m++ = (float) m_el[2][2]; *m++ = (float) m_el[3][2];
|
|
*m++ = (float) m_el[0][3]; *m++ = (float) m_el[1][3]; *m++ = (float) m_el[2][3]; *m = (float) m_el[3][3];
|
|
}
|
|
|
|
/**
|
|
* Copy the contents to a contiguous block of 16 doubles.
|
|
*/
|
|
void getValue(double *m) const {
|
|
*m++ = m_el[0][0]; *m++ = m_el[1][0]; *m++ = m_el[2][0]; *m++ = m_el[3][0];
|
|
*m++ = m_el[0][1]; *m++ = m_el[1][1]; *m++ = m_el[2][1]; *m++ = m_el[3][1];
|
|
*m++ = m_el[0][2]; *m++ = m_el[1][2]; *m++ = m_el[2][2]; *m++ = m_el[3][2];
|
|
*m++ = m_el[0][3]; *m++ = m_el[1][3]; *m++ = m_el[2][3]; *m = m_el[3][3];
|
|
}
|
|
|
|
/**
|
|
* Left-multiply this matrix with the argument.
|
|
*/
|
|
MT_Matrix4x4& operator*=(const MT_Matrix4x4& m);
|
|
|
|
/**
|
|
* Left-multiply column c with row vector c.
|
|
*/
|
|
MT_Scalar tdot(int c, const MT_Vector4& v) const {
|
|
return m_el[0][c] * v[0]
|
|
+ m_el[1][c] * v[1]
|
|
+ m_el[2][c] * v[2]
|
|
+ m_el[3][c] * v[3];
|
|
}
|
|
|
|
/* I'll postpone this for now... - nzc*/
|
|
/* MT_Scalar determinant() const; */
|
|
/* MT_Matrix4x4 adjoint() const; */
|
|
/* MT_Matrix4x4 inverse() const; */
|
|
|
|
MT_Matrix4x4 absolute() const;
|
|
|
|
MT_Matrix4x4 transposed() const;
|
|
void transpose();
|
|
|
|
MT_Matrix4x4 inverse() const;
|
|
void invert();
|
|
|
|
protected:
|
|
/**
|
|
* Access with [row index][column index]
|
|
*/
|
|
MT_Vector4 m_el[4];
|
|
};
|
|
|
|
/* These multiplicators do exactly what you ask from them: they
|
|
* multiply in the indicated order. */
|
|
MT_Vector4 operator*(const MT_Matrix4x4& m, const MT_Vector4& v);
|
|
MT_Vector4 operator*(const MT_Vector4& v, const MT_Matrix4x4& m);
|
|
MT_Matrix4x4 operator*(const MT_Matrix4x4& m1, const MT_Matrix4x4& m2);
|
|
|
|
/* MT_Matrix4x4 MT_multTransposeLeft(const MT_Matrix4x4& m1, const MT_Matrix4x4& m2); */
|
|
/* MT_Matrix4x4 MT_multTransposeRight(const MT_Matrix4x4& m1, const MT_Matrix4x4& m2); */
|
|
|
|
inline MT_OStream& operator<<(MT_OStream& os, const MT_Matrix4x4& m) {
|
|
return os << m[0] << GEN_endl
|
|
<< m[1] << GEN_endl
|
|
<< m[2] << GEN_endl
|
|
<< m[3] << GEN_endl;
|
|
|
|
|
|
|
|
}
|
|
|
|
#ifdef GEN_INLINED
|
|
#include "MT_Matrix4x4.inl"
|
|
#endif
|
|
|
|
#endif
|
|
|