很无聊的一次作业，代码框架也挺糟糕。rayTriangleIntersect 直接抄 PPT 的公式，Render 的难度全在代码注释不清晰上。咱就直接放代码了。
首先是 rayTriangleIntersect：
bool rayTriangleIntersect(const Vector3f& v0, const Vector3f& v1, const Vector3f& v2, const Vector3f& orig,
const Vector3f& dir, float& tNear, float& u, float& v)
{
auto e1 = v1 - v0;
auto e2 = v2 - v0;
auto s = orig - v0;
auto s1 = crossProduct(dir, e2);
auto s2 = crossProduct(s, e1);
auto coefficient = 1.0 / dotProduct(s1, e1);
tNear = coefficient * dotProduct(s2, e2);
u = coefficient * dotProduct(s1, s);
v = coefficient * dotProduct(s2, dir);
return (tNear >= 0) && (u >= 0) && (v >= 0) && ((1 - u - v) >= 0);
}然后是 Render 的关键部分：
for (int i = 0; i < scene.width; ++i)
{
// generate primary ray direction
float x;
float y;
// I don't understand what are the guiding comments talking about.
// Anyway, the code is assuming the distance between eye and screen is one, since abs(dir.z) == 1
// With this assumption we can compute screen's width and height
// Then we map x from [0, scene.width - 1] to [-screen_width / 2, screenwidth / 2]
// and map y from [0, scene.height - 1] to [screen_height / 2, -screen_height / 2]
float screen_height = 2 * scale;
float screen_width = imageAspectRatio * screen_height;
x = (screen_width / (scene.width - 1)) * i - (screen_width / 2.0) - eye_pos.x;
y = (-screen_height / (scene.height - 1)) * j + (screen_height / 2.0) - eye_pos.y;
Vector3f dir = normalize(Vector3f(x, y, -1)); // Don't forget to normalize this direction!
framebuffer[m++] = castRay(eye_pos, dir, scene, 0);
}看看结果，我感觉这个反射也不太对啊……靠近观察者的球对地板的反射怎么在顶部而非底部？它对靠后的球的反射怎么在前方而非后方？总之就这样吧，咱也不太想看这个代码框架……
