Dec 15, 2023
Advent of Code 2023 - Day 15: Lens Library
Solving the 15th day of Advent of Code 2023 in Kotlin
Welcome back to today's Advent of Code challenge! We are already at day 15 which means puzzles are not going to be that easy anymore. So let's see what they prepared for us today.
The Problem
Today's challenge can be found here
The following is the example input which is provided:
rn=1,cm-,qp=3,cm=2,qp-,pc=4,ot=9,ab=5,pc-,pc=6,ot=7
It's just a single line of letters, numbers and symbols.
For now the only thing we have to consider is that these are multiple pairs of characters split by a comma ,
.
Our task for today is to hash each of these pairs using a different algorithm and then summing up the results.
The algorithm, named appendix1A, is described by multiplying the hash of the preceding characters plus the code of the current character by
17
and then taking the remainder of dividing it by 256
.
That's pretty much all to explain for now, let's solve part one!
⭐ Part 1
Let's start by creating the hash function which takes a sequence of characters or just a String
as its parameter.
private fun appendix1A(str: String): Int {
return str
.map(Char::code)
.fold(0) { acc, next -> (acc + next) * 17 % 256 }
}
I'm just mapping each character of the String to its ASCII code and then applying the hash algorithm by using the fold function.
Now we just have to apply this function to each pair of our input, so let's quickly do that:
override fun partOne(input: List<String>): Int {
return input
.first()
.split(",")
.sumOf(::appendix1A)
}
When running this it returns 1320
for the example input and 507291
for mine.
Both of these values are correct, and so we collect our first star ⭐ of today.
⭐ Part 2
In this part things become a lot more complicated and the numbers and symbols in our input start to have a meaning. However, our hashing algorithm is still correct. But rather than summing up all hashes we now have to treat each pair as an instruction.
Also, we now have 256
individual boxes which we have to fill with values depending on the instructions.
Let's take a look on how this instructions work. Let's just take a look at the first two instructions of the example input for this purpose: `rn=1,cm-
Each instruction starts with a label followed by either an =
or a -
.
If there is a =
it is also followed by a number from 1 to 9, which is called a focal length.
I'm going to explain the meaning of those symbols in just a second, let's first set some things up:
val boxes = mutableMapOf<Int, MutableList<String>>()
val focalLengths = mutableMapOf<String, Int>()
You could also use an array to store the boxes, but I went with a map instead because not every box of those 256
will actually be used. I also store the focal length for each able, so we can access them easier later on.
Now let's get to the instructions part:
input.first().split(",").forEach { instruction ->
if (instruction.contains('-')) {
val label = instruction.split("-").first()
val index = appendix1A(label)
boxes[index]?.remove(label)
} else {
val (label, focalLength) = instruction.split("=")
val index = appendix1A(label)
val box = boxes.getOrPut(index) { mutableListOf() }
if (!box.contains(label)) box.add(label)
focalLengths[label] = focalLength.toInt()
}
}
If the instruction contains a -
we are supposed to remove the label out of the corresponding box.
The index of the box can be calculated by using the algorithm from part one.
Since we take the remainder of dividing it by 256
the index can never be greater than that.
Otherwise, when there is a =
in the instruction we have to set the label inside the corresponding box.
We can calculate the index just like above. And also I will store the focal length for that label.
To wrap this all up we need to calculate the final result.
This is done by summing up the focusing power
of each box which contains one or more labels.
The focusing power
is calculated by (boxIndex + 1) * (slotIndex + 1) * focalLenghts[label]
.
Since we don't store empty boxes this is pretty simple:
return boxes.toList().sumOf { (box, values) ->
values.withIndex().sumOf { (slot, label) ->
(box + 1) * (slot + 1) * focalLengths[label]!!
}
}
Running this with the example input gives 145
and with my own it returns 296921
.
And again both of these values are correct.
Second star ⭐ acquired!
Final Words
I find today's problem super enjoyable and actually quite easy. The hard part was to read the huge amounts of text which they provided for today, but once you actually understand what you have to do it should be simple.
You can check out my full solution here.