Merge pull request #13 from faturita/testcase23

Changed the way the turret is controlled.  The azimuth and inclinatio…

carrier.cpp

@@ -591,6 +591,8 @@ int main(int argc, char** argv) {
     // Initialize ODE, create islands, structures and populate the world.
     if (argc>1 && strcmp(argv[1],"-test")==0)
         initWorldModelling(atoi(argv[2]));
+    else if (argc>2 && strcmp(argv[2],"-test")==0)
+        initWorldModelling(atoi(argv[3]));
     else
         initWorldModelling();
 

main.cpp

@@ -28,6 +28,7 @@
 
 #include "camera.h"
 
+#include "sounds.h"
 
 
 float modAngleY=0, modAngleX = 0, modAngleZ=0;

math/splines.hpp

@@ -0,0 +1,110 @@
+#ifndef SPLINES_HPP
+#define SPLINES_HPP
+
+#include <vector>
+
+// =============================================================================
+namespace spline {
+// =============================================================================
+
+enum Node_e {
+    eUNIFORM,
+    eOPEN_UNIFORM ///< Connected to the first and last control points
+};
+
+}// END ESPline ================================================================
+
+/**
+ * @class Spline
+ *
+ * @brief Handling spline curves of arbitrary dimensions
+ * @note This class use the efficient blossom algorithm to compute a position on
+ * the curve.
+ *
+ * @tparam Point_t : type of a point operators such as '+' '*' must be correctly
+ * overloaded. The default constructor must be defined to return the
+ * null vector (0, 0 ,0 ...)
+ * @tparam Real_t ; floating point reprensentation of the points
+ * (float, double etc.)
+ */
+template<typename Point_t, typename Real_t>
+class Spline {
+public:
+
+    /// Type of the nodal vector
+    /// @param k : order of the spline (minimum is two)
+    /// @param node_type : nodal vector type (uniform, open_uniform)
+    /// This will define the behavior of the spline with its control points
+    /// as well as its speed according to its parameter.
+    Spline(int k = 2, spline::Node_e node_type = spline::eOPEN_UNIFORM);
+
+    /// Set the position of the spline control points.
+    void set_ctrl_points(const std::vector<Point_t>& point);
+
+    /// Get the control points of the spline
+    void get_ctrl_points(std::vector<Point_t>& points) const;
+
+    /// The the nodal vector type
+    void set_node_type( spline::Node_e type);
+
+    /// Evaluate position of the spline
+    /// @param u : curve parameter ranging from [0; 1]
+    Point_t eval_f(Real_t u) const;
+
+    /// Evaluate speed of the spline
+    Point_t eval_df(Real_t u) const;
+
+    int get_order() const { return _k; }
+
+private:
+    // -------------------------------------------------------------------------
+    /// @name Class tools
+    // -------------------------------------------------------------------------
+
+    void assert_splines() const;
+
+    /// set value and size of the nodal vector depending on the current number
+    /// of control points
+    void set_nodal_vector();
+
+    /// Set values of the nodal vector to be uniform
+    void set_node_to_uniform();
+
+    /// Set values of the nodal vector to be open uniform
+    void set_node_to_open_uniform();
+
+    /// Evaluate the equation of a splines using the blossom algorithm
+    /// @param u : the curve parameter which range from the values
+    /// [node[k-1]; node[point.size()]]
+    /// @param point : the control points which size must be at least equal to
+    /// the order of the spline (point.size() >= k)
+    /// @param k : the spline order (degree == k-1)
+    /// @param node : the nodal vector which defines the speed of the spline
+    /// parameter u. The nodal vector size must be equal to (k + point.size())
+    /// @param off : offset to apply to the nodal vector 'node' before reading
+    /// from it. this is useful to compute derivatives.
+    Point_t eval(Real_t u,
+                 const std::vector<Point_t>& point,
+                 int k,
+                 const std::vector<Real_t>& node,
+                 int off = 0) const;
+
+    Point_t eval_rec(Real_t u,
+                     std::vector<Point_t> p_in,
+                     int k,
+                     std::vector<Real_t> node_in) const;
+
+    // -------------------------------------------------------------------------
+    /// @name attributes
+    // -------------------------------------------------------------------------
+
+    spline::Node_e _node_type;    ///< Nodal vector type
+    int _k;                        ///< spline order
+    std::vector<Point_t> _point;   ///< Control points
+    std::vector<Point_t> _vec;     ///< Control points differences
+    std::vector<Real_t>  _node;    ///< Nodal vector
+};
+
+#include "splines.inl"
+
+#endif // SPLINES_HPP

math/splines.inl

@@ -0,0 +1,192 @@
+#include "splines.hpp"
+
+#include <cassert>
+#include <algorithm>
+
+template<typename Point_t, typename Real_t>
+Spline<Point_t, Real_t>::Spline(
+        int k,
+        spline::Node_e node_type)
+    :
+      _node_type(node_type),
+      _k(k),
+      _point( _k ),
+      _vec( _k-1 ),
+      _node( _k + _point.size() )
+{
+    assert_splines();
+}
+
+// -----------------------------------------------------------------------------
+
+template<typename Point_t, typename Real_t>
+void Spline<Point_t, Real_t>::set_ctrl_points(const std::vector<Point_t>& point)
+{
+    _point = point;
+    _vec.resize(_point.size() - 1);
+    for(int i = 0; i < (int)_vec.size(); ++i)
+        _vec[i] = _point[i + 1] - _point[i];
+    set_nodal_vector();
+    assert_splines();
+
+    for(int i = 0; i < (int)_vec.size(); ++i)
+        _vec[i] /= _node[_k+i] - _node[i+1];
+}
+
+// -----------------------------------------------------------------------------
+
+template<typename Point_t, typename Real_t>
+void Spline<Point_t, Real_t>::get_ctrl_points(std::vector<Point_t>& points) const
+{
+    points = _point;
+}
+
+// -----------------------------------------------------------------------------
+
+/// The the nodal vector type
+template<typename Point_t, typename Real_t>
+void Spline<Point_t, Real_t>::set_node_type( spline::Node_e type)
+{
+    _node_type = type;
+    set_nodal_vector();
+    assert_splines();
+}
+
+// -----------------------------------------------------------------------------
+
+template<typename Point_t, typename Real_t>
+Point_t Spline<Point_t, Real_t>::eval_f(Real_t u) const
+{
+    u = std::max(std::min(u, (Real_t)1), (Real_t)0); // clamp between [0 1]
+    return eval(u, _point, _k, _node);
+}
+
+// -----------------------------------------------------------------------------
+
+template<typename Point_t, typename Real_t>
+Point_t Spline<Point_t, Real_t>::eval_df(Real_t u) const
+{
+    u = std::max(std::min(u, (Real_t)1), (Real_t)0); // clamp between [0 1]
+    return eval(u, _vec, (_k-1), _node, 1) * (Real_t)(_k-1);
+}
+
+// -----------------------------------------------------------------------------
+
+template<typename Point_t, typename Real_t>
+void Spline<Point_t, Real_t>::assert_splines() const
+{
+    assert( _k > 1);
+    assert((int)_point.size() >= _k );
+    assert(_node. size() == (_k + _point.size()) );
+    assert(_point.size() == (_vec.size()    + 1) );
+}
+
+// -----------------------------------------------------------------------------
+
+template<typename Point_t, typename Real_t>
+void Spline<Point_t, Real_t>::set_nodal_vector()
+{
+    if( _node_type == spline::eOPEN_UNIFORM)
+        set_node_to_open_uniform();
+    else if( _node_type == spline::eUNIFORM )
+        set_node_to_uniform();
+}
+
+// -----------------------------------------------------------------------------
+
+template<typename Point_t, typename Real_t>
+void Spline<Point_t, Real_t>::set_node_to_uniform()
+{
+    const int n = _point.size() - 1;
+    _node.resize( _k + n + 1 );
+
+    Real_t step = (Real_t)1 / (Real_t)(n-_k+2);
+    for (int i = 0; i < (int)_node.size(); ++i){
+        _node[i] = ((Real_t)i) * step  - step * (Real_t)(_k-1);
+    }
+}
+
+// -----------------------------------------------------------------------------
+
+template<typename Point_t, typename Real_t>
+void Spline<Point_t, Real_t>::set_node_to_open_uniform()
+{
+    _node.resize( _k + _point.size() );
+
+    int acc = 1;
+    for (int i = 0; i < (int)_node.size(); ++i)
+    {
+        if(i < _k)
+            _node[i] = 0.;
+        else if( i >= ((int)_point.size() + 1) )
+            _node[i] = 1.;
+        else{
+            _node[i] = (Real_t)acc / (Real_t)(_point.size() + 1 - _k);
+            acc++;
+        }
+    }
+}
+
+// -----------------------------------------------------------------------------
+
+template<typename Point_t, typename Real_t>
+Point_t  Spline<Point_t, Real_t>::
+
+eval(Real_t u,
+     const std::vector<Point_t>& point,
+     int k,
+     const std::vector<Real_t>& node,
+     int off) const
+{
+    assert( k > 1);
+    assert((int)point.size() >= k );
+    assert_splines();
+    int dec = 0;
+    // TODO: better search with dychotomi ?
+    // TODO: check for overflow
+    while( u > node[dec + k + off] )
+        dec++;
+
+    // TODO: use buffers in attributes for better performances ?
+    std::vector<Point_t> p_rec(k, Point_t());
+    for(int i = dec, j = 0; i < (dec + k); ++i, ++j)
+        p_rec[j] = point[i];
+
+    std::vector<Real_t> node_rec(k + k - 2, (Real_t)0);
+    for(int i = (dec + 1), j = 0; i < (dec + k + k - 1); ++i, ++j)
+        node_rec[j] = node[i + off];
+
+    return eval_rec(u, p_rec, k, node_rec);
+}
+
+// -----------------------------------------------------------------------------
+
+template<typename Point_t, typename Real_t>
+Point_t Spline<Point_t, Real_t>::
+
+eval_rec(Real_t u,
+         std::vector<Point_t> p_in,
+         int k,
+         std::vector<Real_t> node_in) const
+{
+    if(p_in.size() == 1)
+        return p_in[0];
+
+    // TODO: use buffers in attributes for better performances ?
+    std::vector<Point_t> p_out(k - 1,  Point_t());
+    for(int i = 0; i < (k - 1); ++i)
+    {
+        const Real_t n0 = node_in[i + k - 1];
+        const Real_t n1 = node_in[i];
+        const Real_t f0 = (n0 - u) / (n0 - n1);
+        const Real_t f1 = (u - n1) / (n0 - n1);
+
+        p_out[i] = p_in[i] *  f0 + p_in[i + 1] * f1;
+    }
+
+    std::vector<Real_t> node_out(node_in.size() - 2);
+    for(int i = 1, j = 0; i < ((int)node_in.size()-1); ++i, ++j)
+        node_out[j] = node_in[i];
+
+    return eval_rec(u, p_out, (k - 1), node_out);
+}

math/yamathutil.cpp

@@ -57,8 +57,8 @@ void Normalize(float v[3][3], float out[3])
  * @param dx
  * @param dy
  * @param dz
- * @param azimuth           In degrees.
- * @param inclination       In degrees.
+ * @param azimuth           In degrees.  glRotatef(-Structure::azimuth,0.0f,1.0f,0.0f);
+ * @param inclination       In degrees.  glRotatef(-Structure::inclination,0.0f,0.0f,1.0f);
  * @return
  */
 Vec3f toVectorInFixedSystem(float dx, float dy, float dz,float azimuth, float inclination)
@@ -93,6 +93,7 @@ float getAzimuth(Vec3f aim)
 }
 
 /**
+ * @FIXME No funca
  * @brief getInclination This is the inverse operation of the one above.  Given a vector, it returs the inclination of the given vector.
  * @param aim
  * @return The inclination in degrees.  -90 is the cenit, 0 is the horizon, positive looking down towards the floor and 90 is your feet.

samplerandom.c

@@ -0,0 +1,18 @@
+#include "stdio.h"
+#include "math.h"
+
+int main(int argc, char *argv[])
+{
+    srand(atoi(argv[1]));
+
+
+    for(int i=0;i<38;i++)
+    {
+        float x = (rand() % 3550 + 1); x -= 1800;
+        float z = (rand() % 3550 + 1); z -= 1800;
+
+        double rfs = rand();
+
+        printf ("rand %10.5f   x %10.5f ,z %10.5f\n", rfs, x, z);
+    }
+}