The JaCoCo XML object didn't have any Type information. Instead we basically treated it as any at every point. This was causing a lot of errors in the code.
So now, we will convert the XML to a Report which is Typed properly and we can safely access it. If the XML report itself is faulty, we will throw an error.

PS: There are still places in code, where we are doing calculations that could cause NaN. But I will fix these in another PR. For now, I kept the code as is.