Project1: a better way to perform ceiled division

[Pokryton]

In Project 1 of this book, when calculating the size of an encoded or decoded string, ceiled division of usize is done by casting to f64 and calling the @ceil function.

zig-book/Chapters/01-base64.qmd

Lines 260 to 279 in 5879b29

	The `_calc_encode_length()` function below encapsulates this logic. Notice that we convert
	the `.len` property of the array, which is always a integer (more precisely, an `usize` value),
	into a floating point number of 64 bits (`f64`). We do this, because the ceiling function (`@ceil()`) works
	only with floating point numbers. So, we convert it so that the division with the number `3.0` results in
	a floating point number. Then, after the ceiling process, we can convert the result back into an
	integer value (with the `@intFromFloat()` function).
	```{zig}
	#| auto_main: false
	#| build_type: "lib"
	fn _calc_encode_length(input: []const u8) u64 {
	if (input.len < 3) {
	const n_output: u64 = 4;
	return n_output;
	}
	const len_as_float: f64 = @floatFromInt(input.len);
	const n_output: u64 = @intFromFloat(@ceil(len_as_float / 3.0) * 4.0);
	return n_output;
	}
	```

While this approach works, it may lost precision in extreme cases (unlikely though), and floating point arithmetic is generally much more expensive than integer arithmetic. A better approach would be to use std.math.divCeil(usize, input.len, 3) or simply (input.len + 2) / 3. This is also easier to read IMO.

For reference, here's how the same functionality implemented in the standard library.

https://github.com/ziglang/zig/blob/218cf059dd215282aa96d6b4715e68d533a4238e/lib/std/base64.zig#L93-L100

[pedropark99]

Hey @Pokryton ! First of all, thank you for your comment!
It is a good suggestion. I will make it happen 😉