obfuscated C++ Translation: float&, two for loops

Question

I was reading Hacker News and this article came up. It contains a raytracer that the code is written on the back of a business card. I decided it would be a good academic challenge to translate the c++ to python, but there's a few concepts I'm stuck on.

First, this function comes up: i T(v o,v d,f& t,v& n){...} Which is translated to int Tracer(vector o, vector d, float& t, vector& n){...} What does the float& mean? I know that in other places & is used as a == is that the case here? Can you do that in c++?

Second, I noticed these three lines:

for(i k=19;k--;)  //For each columns of objects
for(i j=9;j--;)   //For each line on that columns
if(G[j]&1<<k){

I know the << is a the bit shift, and I assume the & is ==. Are the for loops just like one for loop in an other?

Finally, this line: v p(13,13,13); I am not quite sure what it does. Does it create a class labeled by p that extends v (vector) with the defaults of 13,13,13?

These are probably dumb questions, but I want to see if I can understand this and my searching didn't come up with anything. Thank you in advance!

Answer 1

What does the float& mean?

Here, & means "reference", so the argument is passed by reference.

I know that in other places & is used as a == is that the case here?

& means various things in various contexts, but it never means ==. In this case, it's not an operator either; it's part of a type specification, meaning that it's a reference type.

I know the << is a the bit shift, and I assume the & is ==

No, it's a bitwise and operator. The result has its bits set where a bit is set in both operands. Here, with 1<<k as one operand, the result is the k^th bit of G[j]; so this tests whether that bit is set.

Are the for loops just like one for loop in an other?

Yes. If you don't use braces around a for-loop's body, then the body is a single statement. So in this case, the body of the first loop is the second loop. To make this clear, I would recommend indenting the body of the loop, and using braces whether or not they are strictly necessary. But of course, I don't write (deliberately) obfuscated code.

Finally, this line: v p(13,13,13);

v is a class with a constructor taking three arguments. This declares an variable called p, of type v, initialised using that constructor; i.e. the three co-ordinates are initialised to 13.

Answer 2

When you seeVector& n it is referencing the vector passed into the function. This means that you can change n inside of this function without having to copy it to another Vector or without returning the Vector. This previous answer should be helpful to you.