Day 6: Trash Compactor
Megathread guidelines
- Keep top level comments as only solutions, if you want to say something other than a solution put it in a new post. (replies to comments can be whatever)
- You can send code in code blocks by using three backticks, the code, and then three backticks or use something such as https://topaz.github.io/paste/ if you prefer sending it through a URL
FAQ
- What is this?: Here is a post with a large amount of details: https://programming.dev/post/6637268
- Where do I participate?: https://adventofcode.com/
- Is there a leaderboard for the community?: We have a programming.dev leaderboard with the info on how to join in this post: https://programming.dev/post/6631465
Managed to keep it compact, but boy, do I hate cephalopod math >_<
Python
from csv import reader from functools import reduce from itertools import pairwise from operator import mul from pathlib import Path from typing import Any, List, Sequence def _calc(values: List[str]) -> int: match values[-1]: case "+": return sum(map(int, values[:-1])) case "*": return reduce(mul, map(int, values[:-1])) case _: return 0 def _transpose(values: Sequence[Sequence[Any]]) -> List[List[Any]]: return [[values[row][col] for row in range(len(values))] for col in range(len(values[0]))] def part_one(input: str) -> int: def _parse_input(input: str) -> List[List[str]]: return _transpose(list(map(lambda r: list(filter(None, r)), reader(input.splitlines(), delimiter=" ")))) return sum(map(_calc, _parse_input(input))) def part_two(input: str) -> int: def _parse_input(input: str) -> List[List[str]]: data = list(input.splitlines()) columns = [t[0] for t in filter(lambda t: t[1] != " ", enumerate(data[-1]))] + [len(data[0])] numbers = [[line[a:b] for line in data[:-1]] for a, b in pairwise(columns)] numbers = [list(filter(None, ["".join(num).strip() for num in column])) for column in map(_transpose, numbers)] return list(map(lambda t: t[0] + [t[1]], zip(numbers, list(filter(None, data[-1].split(" ")))))) return sum(map(_calc, _parse_input(input))) if __name__ == "__main__": input = Path("_2025/_6/input").read_text("utf-8") print(part_one(input)) print(part_two(input))Uiua
I’m new to Uiua, so probably not the best way to express the solution!
ParseRows ← ( ⊜∘⊸≠@\n ⍜⇌°⊂ # Get the row of operations. ) ParseOperations ← ⊜⊢⊸≠@\s SumCalculations ← ( ≡( ⍣(◇/×°@* | ◇/+°@+ ) ) /+ ) Part₁ ← ( ParseRows ⊓(ParseOperations | ⍉ ≡(⊜⋕⊸≠@\s) # Parse columns. ) SumCalculations ) Part₂ ← ( ParseRows ⊓(ParseOperations | ⊜(□≡(⋕▽⊸≠@\s)) ≡/↥⊸≠@\s⍉ # Parse numbers. ) SumCalculations ) &fras "6.txt" ⊃Part₁ Part₂
( ͡° ͜ʖ├┬┴┬┴
Looks good. I like how you managed to use the same structure for Parts 1 and 2; that’s more than I did. And yours is faster than mine too.
One of us! One of us!
Nice use of the inversion, I always forget that’s a thing
Haskell
There’s probably a really clever way of abstracting just the difference between the two layouts.
import Data.Char (isSpace) import Data.List (transpose) import Data.List.Split (splitWhen) op '+' = sum op '*' = product part1 = sum . map ((op . head . last) <*> (map read . init)) . (transpose . map words . lines) part2 = sum . map ((op . last . last) <*> map (read . init)) . (splitWhen (all isSpace) . reverse . transpose . lines) main = do input <- readFile "input06" print $ part1 input print $ part2 inputUlua probably has a single character that rotates the input -90 degrees…
Your code really reads like your explaining the solution out loud. That’s so elegant!
Thanks! I try to write code to be readable by humans above all else.
nushell
I was afk when the puzzle went up so I had another go at doing it on my phone in Turmux with my shell’s scripting language. It’s quite nice how your shell is also a REPL so you can build up the answer in pieces, although I wrote a file for the second part.

open input.txt | str replace --all --regex ' +' ' ' | lines | each { $in | str trim } | to text | from csv --noheaders --separator ' ' | reverse | transpose --ignore-titles | each { |list| transpose | skip 1 | if $list.column0 == '+' { math sum } else { math product } } | math sumPart 2
let input = open input.txt | lines | each { $in | split chars } let last_row = ($input | length) - 1 let last_col = ($input | first | length) - 1 mut op = ' ' mut numbers = [] mut grand_tot = 0 for x in $last_col..0 { if $op == '=' { $op = ' ' continue } let n = 0..($last_row - 1) | each { |y| $input | get $y | get $x } | str join | into int $numbers = ($numbers | append $n) $op = $input | get $last_row | get $x if $op != ' ' { $grand_tot += $numbers | if $op == '+' { math sum } else { math product } $numbers = [] $op = '=' } } $grand_totUiua
This was fun :D
I had a fun experience just throwing the strings with both numbers and spaces at the parse function. In the online pad, everything worked out fine but running the same code on my input locally gave me a “invalid float literal” error.
I thought I’d missed some edge case in the real input again, like is often the case.
Turns out that the Uiua version I used locally had a bug that’s fixed in the latest build. For once it wasn’t directly my fault ^^Code
$ 123 328 51 64 $ 45 64 387 23 $ 6 98 215 314 $ * + * + # &fras "input-6.txt" ◌ Calc ← ( ↘₂⊛⊂"+*"/◇⊂ ≡◇⨬(/+|/×) /+ ) P₁ ← ( ⊜(⊜□⊸≠@ )⊸≠@\n ⊃⊣↘₋₁ ⊙(⍉≡₀◇⋕) Calc ) P₂ ← ( ⊜∘⊸≠@\n ⟜⧻ ⊓⍉(˜↯@ ) ⊜(⊙(□≡⋕)⍜⍉(⊃(⊢⊣)↘₋₁))¬⤚≡⌟≍ Calc ) 1_2 [⊃P₁P₂] ≡(&p &pf $"Part _: ")Kotlin
I’m not fully happy with my parsing today, but oh well. I also thought about just plain building the grid and then rotating it, but “normal input parsing” works too.
Solution
class Day06 : Puzzle { val numsPartOne = mutableListOf<MutableList<Long>>() val numsPartTwo = mutableListOf<List<Long>>() val ops = mutableListOf<(Long, Long) -> Long>() override fun readFile() { val input = readInputFromFile("src/main/resources/a2025/day06.txt") val lines = input.lines().filter { it.isNotBlank() } // parse part1 input for (line in lines.dropLast(1)) { for ((c, num) in line.trim().split(" +".toRegex()).withIndex()) { if (numsPartOne.getOrNull(c) == null) numsPartOne.add(mutableListOf()) numsPartOne[c].add(num.toLong()) } } // parse part2 input var numList = mutableListOf<Long>() for (c in 0..<lines.maxOf { it.length }) { var numStr = "" for (r in 0..<lines.size - 1) { numStr += lines[r].getOrElse(c) { ' ' } } if (numStr.isBlank()) { numsPartTwo.add(numList) numList = mutableListOf() } else { numList.add(numStr.trim().toLong()) } } numsPartTwo.add(numList) // parse operators ops.addAll( lines.last().split(" +".toRegex()) .map { it.trim()[0] } .map { when (it) { '*' -> { a: Long, b: Long -> a * b } '+' -> { a: Long, b: Long -> a + b } else -> throw IllegalArgumentException("Unknown operator: $it") } } ) } override fun solvePartOne(): String { return numsPartOne.mapIndexed { c, list -> list.reduce { a, b -> ops[c](a, b) } }.sum().toString() } override fun solvePartTwo(): String { return numsPartTwo.mapIndexed { c, list -> list.reduce { a, b -> ops[c](a, b) } }.sum().toString() } }full code on Codeberg
I also thought about trying to rotate, but not for very long. Mine would be a bit simpler if I’d done what you did and build the number string and then check if it’s blank.
fun main() { val input = getInput(6) val output = parseInput2(input) var total = 0L for ((numbers, operator) in output) { when (operator) { '+' -> { total += numbers.sum() } '*' -> { total += numbers.reduce { acc, number -> acc * number }} } } println(getElapsedTime()) println(total) } fun parseInput2(input: String): List<Pair<List<Long>, Char>> { val rows = input.lines() .filter { it.isNotBlank() } .map { it.toCharArray() } val output: MutableList<Pair<List<Long>, Char>> = mutableListOf() val numberRowCount = rows.size - 1 var isNewProblem = true var currentNumbers: MutableList<Long> = mutableListOf() var operator = ' ' for (column in rows[0].indices) { if (!isNewProblem && isColumnEmpty(rows, column)) { isNewProblem = true output.add(currentNumbers to operator) continue } if (isNewProblem) { isNewProblem = false currentNumbers = mutableListOf() operator = rows.last()[column] } var number = "" for (row in 0..<numberRowCount) { if (rows[row][column] != ' ') { number += rows[row][column] } } currentNumbers.add(number.toLong()) } if (!isNewProblem) { output.add(currentNumbers to operator) } return output } fun isColumnEmpty(rows: List<CharArray>, column: Int): Boolean { for (i in rows.indices) { if (rows[i][column] != ' ') { return false } } return true }
Ironically for Lisp, a good chunk of the work here is type conversion, because strings, vectors, multidimensional arrays, characters, and numbers don’t have implicit conversions between them; you have to specify what you want explicitly. I also found it easier to manually transpose the character array for part 2 rather than traverse in column-major order, because that makes the relationship between input and output data structure more transparent.
(ql:quickload :str) (ql:quickload :array-operations) (defun parse-line-1 (line) (let ((broken-line (str:split " " (str:collapse-whitespaces (str:trim line))))) (mapcar #'(lambda (s) (cond ((equal s "+") #'+) ((equal s "*") #'*) (t (parse-integer s)))) broken-line))) (defun read-inputs-1 (filename) (let* ((input-lines (uiop:read-file-lines filename))) (mapcar #'parse-line-1 input-lines))) (defun main-1 (filename) (let* ((problems (read-inputs-1 filename)) (arguments (apply #'mapcar #'list (butlast problems)))) (reduce #'+ (mapcar #'apply (car (last problems)) arguments)))) (defun parse-operands-2 (lines) (let* ((initial-rows (length lines)) (initial-cols (length (car lines))) (flat-chars (make-array (list (* initial-rows initial-cols)) :initial-contents (apply #'concatenate 'string lines))) (box-chars (make-array (list initial-rows initial-cols) :displaced-to flat-chars)) (transposed-chars (aops:each-index (i j) (aref box-chars j i)))) (loop for cv across (aops:split transposed-chars 1) for s = (str:trim (coerce cv 'string)) collect (if (zerop (length s)) nil (parse-integer s))))) (defun list-split (xs sep &optional (predicate #'equal)) (let ((current nil) (result nil)) (loop for x in xs do (if (funcall predicate x sep) (progn (setf result (cons (reverse current) result)) (setf current nil)) (setf current (cons x current))) finally (setf result (cons (reverse current) result))) (reverse result))) (defun main-2 (filename) (let* ((lines (uiop:read-file-lines filename)) (operators (parse-line-1 (car (last lines)))) (operands (parse-operands-2 (butlast lines)))) (loop for rator in operators for rands in (list-split operands nil) sum (apply rator rands))))Haskell
import Control.Arrow import Data.Char import Data.List import Text.ParserCombinators.ReadP op "*" = product op "+" = sum part1 s = sum $ zipWith ($) (op <$> a) (transpose $ fmap read <$> as) where (a : as) = reverse . fmap words . lines $ s parseGroups = fst . last . readP_to_S (sepBy (endBy int eol) eol) . filter (/= ' ') where eol = char '\n' int = read <$> munch1 isDigit :: ReadP Int part2 s = sum $ zipWith ($) (op <$> words a) (parseGroups . unlines $ reverse <$> transpose as) where (a : as) = reverse $ lines s main = getContents >>= print . (part1 &&& part2)Go
Damn, I actually reeaally enjoyed this one! I didn’t expect the twist of part 2, but somehow it wasn’t that hard to manage.
Here is my modern solution:
day06.go
package main import ( "aoc/utils" "fmt" "regexp" "slices" "strconv" "strings" ) type operation int func (o operation) compute(values []int) int { switch o { case add: sum := 0 for _, val := range values { sum += val } return sum case mul: product := 1 for _, val := range values { product *= val } return product } return 0 } const ( add operation = iota mul ) var allOperationSymbols = []string{"+", "*"} func operationFromSymbol(sym string) operation { switch sym { case "+": return add case "*": return mul default: panic(fmt.Sprintf("wtf is a %s?", sym)) } } type problems struct { values [][]int operations []operation } func (p *problems) feed(column string) { last := string(column[len(column)-1]) done := false if slices.Contains(allOperationSymbols, last) { p.operations = append(p.operations, operationFromSymbol(last)) column = column[:len(column)-1] done = true } val, _ := strconv.Atoi(strings.TrimSpace(column)) idx := len(p.values) - 1 p.values[idx] = append(p.values[idx], val) if done { p.values = append(p.values, []int{}) } } func (p *problems) addLine(line string) (done bool) { parts := strings.Split(line, " ") parts = slices.DeleteFunc(parts, func(elem string) bool { return elem == "" }) if slices.Contains(allOperationSymbols, parts[0]) { p.operations = make([]operation, len(parts)) for idx, sym := range parts { p.operations[idx] = operationFromSymbol(sym) } done = true } else { if len(p.values) == 0 { lenparts := len(parts) p.values = make([][]int, lenparts) for idx := range lenparts { p.values[idx] = []int{} } } for idx, part := range parts { num, _ := strconv.Atoi(part) p.values[idx] = append(p.values[idx], num) } done = false } return done } func (p problems) solve() []int { solutions := make([]int, len(p.values)) for idx, values := range p.values { op := p.operations[idx] solutions[idx] = op.compute(values) } return solutions } func stepOne(input chan string) (int, error) { modernProblems := problems{} for line := range input { done := modernProblems.addLine(line) if done { break } } modernSolutions := modernProblems.solve() sum := 0 for _, solution := range modernSolutions { sum += solution } return sum, nil } func transposeInputChan(input chan string) []string { lines := [][]rune{} for line := range input { lines = append(lines, []rune(line)) } linecount := len(lines) columncount := len(lines[0]) transposed := make([][]rune, columncount) for idx := range transposed { transposed[idx] = make([]rune, linecount) } for row, line := range lines { for col, char := range line { transposed[col][row] = char } } columns := make([]string, len(transposed)) for idx, col := range transposed { columns[idx] = string(col) } return columns } func stepTwo(input chan string) (int, error) { transposedInput := transposeInputChan(input) slices.Reverse(transposedInput) // problem-set with one empty problem. modernProblems := problems{ values: [][]int{[]int{}}, } for _, column := range transposedInput { if matched, _ := regexp.MatchString("^\\s*$", column); matched { continue } modernProblems.feed(column) } // Remove last useless empty problem. modernProblems.values = modernProblems.values[:len(modernProblems.values)-1] modernSolutions := modernProblems.solve() sum := 0 for _, solution := range modernSolutions { sum += solution } return sum, nil } func main() { inputFile := utils.FilePath("day06.txt") utils.RunStep(utils.ONE, inputFile, stepOne) utils.RunStep(utils.TWO, inputFile, stepTwo) }Go
Part 2: Read the whole input in a rune matrix. Scan it column by column, store the numbers as you go, ignoring all spaces, and store the operand when you find it. When you hit an empty column or the end, do the operation and add it to the total.
spoiler
func part2() { // file, _ := os.Open("sample.txt") file, _ := os.Open("input.txt") defer file.Close() scanner := bufio.NewScanner(file) chars := [][]rune{} for scanner.Scan() { chars = append(chars, []rune(scanner.Text())) } m := len(chars) n := len(chars[0]) var op rune nums := []int{} total := 0 for j := range n { current := []rune{} for i := range m { if chars[i][j] == '+' || chars[i][j] == '*' { op = chars[i][j] } else if chars[i][j] != ' ' { current = append(current, chars[i][j]) } } if len(current) > 0 { x, _ := strconv.Atoi(string(current)) nums = append(nums, x) } if len(current) == 0 || j == n-1 { result := 0 if op == '*' { result = 1 } for _, x := range nums { if op == '+' { result = result + x } else { result = result * x } } total += result nums = []int{} } } fmt.Println(total) }Ruby
I decided to rotate the entire input character-by-character, then parse the numbers (see the full source here)
grid = input.lines.map(&:chomp).map {|l| l.each_char.map.to_a }.to_a transposed = Array.new(grid[0].length) { Array.new(grid.length) } grid.each_with_index do |row, y| row.each_with_index do |col, x| transposed[x][y] = col end end vals = [] ops = [] temp_vals = [] transposed.each do |row| l = row.join("").strip temp_vals << l.scan(/\d+/).map(&:to_i).to_a[0] /[+*]/.match(l) { |m| ops << m.to_s.to_sym } if l == "" vals << temp_vals.compact temp_vals = [] end end vals << temp_vals.compact unless temp_vals.empty? vals.each_with_index.sum do |v, i| v.inject(ops[i]) endRust
Finally having some fun with iters. I think part 2 came out nicely once I figured I can rotate the whole input to get numbers that look like ours.
Q: Anyone have any tips how to reason about the intermediate types in a long iter chain? I’m currently placing dummy maps, and on occasion, even printing the values from them when I get too stuck.
the code
use std::fs::File; use std::io::{BufReader, Lines}; #[allow(dead_code)] pub fn part1(input: Lines<BufReader<File>>) { let mut input = input .map_while(Result::ok) .map(|line| { line.split_ascii_whitespace() .map(|s| s.to_string()) .collect::<Vec<String>>() }) .collect::<Vec<Vec<String>>>(); let ops = input.pop().unwrap(); let values = input .iter() .map(|v| { v.iter() .map(|v| v.parse::<i64>().unwrap()) .collect::<Vec<i64>>() }) .collect::<Vec<Vec<i64>>>(); let transposed: Vec<Vec<i64>> = (0..values[0].len()) .map(|i| values.iter().map(|row| row[i]).collect()) .collect(); let mut sum = 0; for i in 0..ops.len() { let op: &str = &ops[i]; match op { "+" => { sum += transposed[i].iter().sum::<i64>(); } "*" => { sum += transposed[i].iter().product::<i64>(); } _ => panic!("Invalid operation"), } } println!("sum = {}", sum) } #[allow(dead_code)] pub fn part2(input: Lines<BufReader<File>>) { let mut input = input .map_while(Result::ok) .map(|line| line.chars().collect::<Vec<char>>()) .collect::<Vec<Vec<char>>>(); let ops = input .pop() .unwrap() .iter() .map(|c| c.to_string()) .filter(|s| s != " ") .collect::<Vec<String>>(); let transposed: Vec<i64> = (0..input[0].len()) .map(|i| input.iter().map(|row| row[i]).collect()) .map(|vec: String| vec.trim().to_string()) .map(|s| { if s.len() == 0 { 0 } else { s.parse::<i64>().unwrap() } }) .collect(); let groups = transposed .into_iter() .fold(Vec::new(), |mut acc: Vec<Vec<i64>>, num| { if num == 0 { if let Some(last) = acc.last_mut() { if !last.is_empty() { acc.push(Vec::new()); } } else { acc.push(Vec::new()); } } else { if acc.is_empty() { acc.push(Vec::new()); } acc.last_mut().unwrap().push(num); } acc }); let mut sum = 0; for i in 0..ops.len() { let op: &str = &ops[i]; match op { "+" => { sum += groups[i].iter().sum::<i64>(); } "*" => { sum += groups[i].iter().product::<i64>(); } _ => panic!("Invalid operation"), } } println!("sum = {}", sum) }(Browser-based) Javascript
I got lazy and lucky writing the first part; hard-coded the number of lines / operands and it worked on the first try! I didn’t factor in the final block not being padded by spaces on both sides for part 2, and so needed to test on the example code – which has 3 lines of operands instead of the problem input’s 4 lines, so I had to properly solve the problem across all possible amounts of lines of input 😩.
I am very jealous of all y’all who have a transpose function available in your language of choice.
Code
function part1(inputText) { const [firstArgs, secondArgs, thirdArgs, fourthArgs, operators] = inputText.trim().split('\n').map(line => line.trim().split(/\s+/)); let totalSum = 0; for (let i = 0; i < firstArgs.length; i++) { if (operators[i] === '+') { totalSum += Number.parseInt(firstArgs[i], 10) + Number.parseInt(secondArgs[i], 10) + Number.parseInt(thirdArgs[i], 10) + Number.parseInt(fourthArgs[i], 10); } else if (operators[i] === '*') { totalSum += Number.parseInt(firstArgs[i], 10) * Number.parseInt(secondArgs[i], 10) * Number.parseInt(thirdArgs[i], 10) * Number.parseInt(fourthArgs[i], 10); } } return totalSum } { const start = performance.now(); const result = part1(document.body.textContent); const end = performance.now(); console.info({ day: 6, part: 1, time: end - start, result }) } function part2(inputText) { const lines = inputText.trimEnd().split('\n'); const interBlockIndices = []; for (let i = 0; i < lines[0].length; i++) { let allEmpty = true; for (let j = 0; j < lines.length; j++) { if (lines[j][i] !== ' ') { allEmpty = false; break; } } if (allEmpty) { interBlockIndices.push(i); } } interBlockIndices.push(lines[0].length); let totalSum = 0; let blockStart = 0; for (const interBlockIndex of interBlockIndices) { // compute calculation of block [blockStart, interBlockIndex - 1] const operands = []; for (let i = interBlockIndex - 1; i >= blockStart; i--) { // parse operands let operand = 0; for (let j = 0; j < lines.length - 1; j++) { if (lines[j][i] !== ' ') { operand *= 10; operand += Number.parseInt(lines[j][i], 10); } } operands.push(operand) } if (lines.at(-1)[blockStart] === '+') { totalSum += operands.reduce((accu, next) => accu + next, 0); } else if (lines.at(-1)[blockStart] === '*') { totalSum += operands.reduce((accu, next) => accu * next, 1); } // console.debug({ totalSum, operands, blockStart, interBlockIndex }); blockStart = interBlockIndex + 1; } return totalSum; } { const example = `123 328 51 64 45 64 387 23 6 98 215 314 * + * + `; const start = performance.now(); const result = part2(document.body.textContent); // const result = part2(example); const end = performance.now(); console.info({ day: 6, part: 2, time: end - start, result }) }Nim
The hardest part was reading the part 2 description. I literally looked at it for minutes trying to understand where the problem numbers come from and how they’re related to the example input. But then it clicked.
The next roadblock was that my template was stripping whitespace at the end of the last line, making parsing a lot harder. I’ve replaced
strip()withstrip(chars={'\n'})to keep the trailing space intact.Runtime:
1.4 ms618 μsview code
type AOCSolution[T,U] = tuple[part1: T, part2: U] proc solve(input: string): AOCSolution[int, int] = let lines = input.splitLines() let numbers = lines[0..^2] let ops = lines[^1] block p1: let numbers = numbers.mapIt(it.splitWhiteSpace().mapIt(parseInt it)) let ops = ops.splitWhitespace() for x in 0 .. numbers[0].high: var res = numbers[0][x] for y in 1 .. numbers.high: case ops[x] of "*": res *= numbers[y][x] of "+": res += numbers[y][x] result.part1 += res block p2: var problems: seq[(char, Slice[int])] var ind = 0 while ind < ops.len: let len = ops.skipWhile({' '}, ind+1) problems.add (ops[ind], ind .. ind + len - (if ind+len < ops.high: 1 else: 0)) ind += len + 1 for (op, cols) in problems: var res = 0 for x in cols: var num = "" for y in 0 .. numbers.high: num &= numbers[y][x] if res == 0: res = parseInt num.strip else: case op of '*': res *= parseInt num.strip of '+': res += parseInt num.strip else: discard result.part2 += resFull solution at Codeberg: solution.nim









