mirror of
https://github.com/g-truc/glm.git
synced 2024-11-10 12:41:54 +00:00
96 lines
2.2 KiB
C++
96 lines
2.2 KiB
C++
///////////////////////////////////////////////////////////////////////////////////////////////////
|
|
// OpenGL Mathematics Copyright (c) 2005 - 2011 G-Truc Creation (www.g-truc.net)
|
|
///////////////////////////////////////////////////////////////////////////////////////////////////
|
|
// Created : 2010-09-16
|
|
// Updated : 2011-05-25
|
|
// Licence : This source is under MIT licence
|
|
// File : test/gtc/quaternion.cpp
|
|
///////////////////////////////////////////////////////////////////////////////////////////////////
|
|
|
|
#include <glm/glm.hpp>
|
|
#include <glm/gtc/quaternion.hpp>
|
|
#include <glm/gtx/epsilon.hpp>
|
|
|
|
int test_quat_precision()
|
|
{
|
|
int Error = 0;
|
|
|
|
Error += sizeof(glm::lowp_quat) <= sizeof(glm::mediump_quat) ? 0 : 1;
|
|
Error += sizeof(glm::mediump_quat) <= sizeof(glm::highp_quat) ? 0 : 1;
|
|
|
|
return Error;
|
|
}
|
|
|
|
int test_quat_type()
|
|
{
|
|
glm::quat A;
|
|
glm::dquat B;
|
|
|
|
return 0;
|
|
}
|
|
|
|
int test_quat_slerp()
|
|
{
|
|
int Error = 0;
|
|
|
|
glm::quat A(glm::vec3(0, 0, 1));
|
|
glm::quat B(glm::vec3(0, 1, 0));
|
|
glm::quat C = glm::mix(A, B, 0.5f);
|
|
glm::quat D(glm::normalize(glm::vec3(0, 1, 1)));
|
|
|
|
Error += C == D ? 0 : 1;
|
|
|
|
return Error;
|
|
}
|
|
|
|
int test_quat_length()
|
|
{
|
|
int Error = 0;
|
|
|
|
float A = glm::length(glm::quat(45.0f, glm::vec3(0, 0, 1)));
|
|
Error += A == 1.0f ? 0 : 1;
|
|
float B = glm::length(glm::quat(90.0f, glm::vec3(0, 0, 2)));
|
|
Error += B == 2.0f ? 0 : 1;
|
|
|
|
return Error;
|
|
}
|
|
|
|
int test_quat_normalize()
|
|
{
|
|
int Error = 0;
|
|
|
|
{
|
|
glm::quat Q(45.0f, glm::vec3(0, 0, 1));
|
|
glm::quat N = glm::normalize(Q);
|
|
float L = glm::length(N);
|
|
Error += L == 1.0f ? 0 : 1;
|
|
}
|
|
{
|
|
glm::quat Q(45.0f, glm::vec3(0, 0, 2));
|
|
glm::quat N = glm::normalize(Q);
|
|
float L = glm::length(N);
|
|
Error += L == 1.0f ? 0 : 1;
|
|
}
|
|
{
|
|
glm::quat Q(45.0f, glm::vec3(1, 2, 3));
|
|
glm::quat N = glm::normalize(Q);
|
|
float L = glm::length(N);
|
|
Error += L == 1.0f ? 0 : 1;
|
|
}
|
|
|
|
return Error;
|
|
}
|
|
|
|
int main()
|
|
{
|
|
int Error = 0;
|
|
|
|
Error += test_quat_precision();
|
|
Error += test_quat_type();
|
|
Error += test_quat_slerp();
|
|
Error += test_quat_length();
|
|
Error += test_quat_normalize();
|
|
|
|
return Error;
|
|
}
|