I struggled a lot more when doing list slices that I would’ve liked to

Haskell

import Data.List qualified as List

collectDiagonal :: [String] -> Int -> Int -> String
collectDiagonal c y x
        | length c > y && length (c !! y) > x = c !! y !! x : collectDiagonal c (y+1) (x+1)
        | otherwise = []

part1 c = do
        let forwardXMAS  = map (length . filter (List.isPrefixOf "XMAS") . List.tails) $ c
        let backwardXMAS = map (length . filter (List.isPrefixOf "XMAS") . List.tails . reverse) $ c
        let downwardXMAS  = map (length . filter (List.isPrefixOf "XMAS") . List.tails ) . List.transpose $ c
        let upwardXMAS = map (length . filter (List.isPrefixOf "XMAS") . List.tails . reverse ) . List.transpose $ c
        let leftSideDiagonals = map (\ y -> collectDiagonal c y 0) [0..length c]
        let leftTopDiagonals = map (\ x -> collectDiagonal c 0 x) [1..(length . List.head $ c)]
        let leftDiagonals = leftSideDiagonals ++ leftTopDiagonals
        let rightSideDiagonals = map (\ y -> collectDiagonal (map List.reverse c) y 0) [0..length c]
        let rightTopDiagonals = map (\ x -> collectDiagonal (map List.reverse c) 0 x) [1..(length . List.head $ c)]
        let rightDiagonals = rightSideDiagonals ++ rightTopDiagonals
        let diagonals = leftDiagonals ++ rightDiagonals

        let diagonalXMAS = map (length . filter (List.isPrefixOf "XMAS") . List.tails) $ diagonals
        let reverseDiagonalXMAS = map (length . filter (List.isPrefixOf "XMAS") . List.tails . reverse) $ diagonals

        print . sum $ [sum forwardXMAS, sum backwardXMAS, sum downwardXMAS, sum upwardXMAS, sum diagonalXMAS, sum reverseDiagonalXMAS]
        return ()

getBlock h w c y x = map (take w . drop x) . take h . drop y $ c

isXBlock b = do
        let diagonal1 = collectDiagonal b 0 0
        let diagonal2 = collectDiagonal (map List.reverse b) 0 0

        diagonal1 `elem` ["SAM", "MAS"] && diagonal2 `elem` ["SAM", "MAS"]

part2 c = do
        
        let lineBlocks = List.map (getBlock 3 3 c) [0..length c - 1]
        let groupedBlocks = List.map (flip List.map [0..(length . head $ c) - 1]) lineBlocks

        print . sum . map (length . filter isXBlock) $ groupedBlocks

        return ()

main = do
        c <- lines <$> getContents

        part1 c
        part2 c

        return ()

Nim

Could be done more elegantly, but I haven’t bothered yet.

proc solve(input: string): AOCSolution[int, int] =
  var lines = input.splitLines()

  block p1:
    # horiz
    for line in lines:
      for i in 0..line.high-3:
        if line[i..i+3] in ["XMAS", "SAMX"]:
          inc result.part1

    for y in 0..lines.high-3:
      #vert
      for x in 0..lines[0].high:
        let word = collect(for y in y..y+3: lines[y][x])
        if word in [@"XMAS", @"SAMX"]:
          inc result.part1

      #diag \
      for x in 0..lines[0].high-3:
        let word = collect(for d in 0..3: lines[y+d][x+d])
        if word in [@"XMAS", @"SAMX"]:
          inc result.part1

      #diag /
      for x in 3..lines[0].high:
        let word = collect(for d in 0..3: lines[y+d][x-d])
        if word in [@"XMAS", @"SAMX"]:
          inc result.part1

  block p2:
    for y in 0..lines.high-2:
      for x in 0..lines[0].high-2:
        let diagNW = collect(for d in 0..2: lines[y+d][x+d])
        let diagNE = collect(for d in 0..2: lines[y+d][x+2-d])
        if diagNW in [@"MAS", @"SAM"] and diagNE in [@"MAS", @"SAM"]:
          inc result.part2

Codeberg repo

Haskell

Popular language this year :)

I got embarrassingly stuck on this one trying to be clever with list operations. Then I realized I should just use an array…

import Data.Array.Unboxed (UArray)
import Data.Array.Unboxed qualified as A
import Data.Bifunctor

readInput :: String -> UArray (Int, Int) Char
readInput s =
  let rows = lines s
      n = length rows
   in A.listArray ((1, 1), (n, n)) $ concat rows

s1 `eq` s2 = s1 == s2 || s1 == reverse s2

part1 arr = length $ filter isXmas $ concatMap lines $ A.indices arr
  where
    isXmas ps = all (A.inRange $ A.bounds arr) ps && map (arr A.!) ps `eq` "XMAS"
    lines p = [take 4 $ iterate (bimap (+ di) (+ dj)) p | (di, dj) <- [(1, 0), (0, 1), (1, 1), (1, -1)]]

part2 arr = length $ filter isXmas innerPoints
  where
    innerPoints =
      let ((i1, j1), (i2, j2)) = A.bounds arr
       in [(i, j) | i <- [i1 + 1 .. i2 - 1], j <- [j1 + 1 .. j2 - 1]]
    isXmas p = up p `eq` "MAS" && down p `eq` "MAS"
    up (i, j) = map (arr A.!) [(i + 1, j - 1), (i, j), (i - 1, j + 1)]
    down (i, j) = map (arr A.!) [(i - 1, j - 1), (i, j), (i + 1, j + 1)]

main = do
  input <- readInput <$> readFile "input04"
  print $ part1 input
  print $ part2 input

J

Unsurprisingly this is the kind of problem that J is really good at. The dyadic case (table) of the adverb / is doing all the heavy lifting here: it makes a higher rank tensor by traversing items of the specified rank on each side and combining them according to the remaining frame of each side’s shape. The hard part is arranging the arguments so that your resulting matrix has its axes in the correct order.

data_file_name =: '4.data'

NB. cutopen yields boxed lines, so unbox them and ravel items to make a letter matrix
grid =: ,. > cutopen fread data_file_name
NB. pad the grid on every side with #'XMAS' - 1 spaces
hpadded_grid =: (('   ' & ,) @: (, & '   '))"1 grid
padded_grid =: (3 1 $ ' ') , hpadded_grid , (3 1 $ ' ')
NB. traversal vectors
directions =: 8 2 $ 1 0 1 1 0 1 _1 1 _1 0 _1 _1 0 _1 1 _1
NB. rpos cpos matches rdir cdir if the string starting at rpos cpos in
NB. direction rdir cdir is the string we want
matches =: 4 : 0
*/ ,'XMAS' -: padded_grid {~ <"1 x +"1 y *"1 0 i. 4
)"1
positions =: (3 + i. 0 { $ grid) ,"0/ (3 + i. 1 { $ grid)
result1 =: +/, positions matches/ directions

NB. pairs of traversal vectors
x_directions =: 4 2 2 $ 1 1 _1 1 1 1 1 _1 _1 _1 _1 1 _1 _1 1 _1
NB. rpos cpos x_matches 2 2 $ rdir1 cdir1 rdir2 cdir2 if there is an 'A' at
NB. rpos cpos and the string in each of dir1 and dir2 centered at rpos cpos
NB. is the string we want
x_matches =: 4 : 0
NB. (2 2 $ rdir1 cdir1 rdir2 cdir2) *"1 0/ (_1 + i.3) yields a matrix
NB. 2 3 $ (_1 * dir1) , (0 * dir1) , (1 * dir1) followed by the same for dir2
*/ ,'MAS' -:"1 padded_grid {~ <"1 x +"1 y *"1 0/ _1 + i. 3
)"1 2
result2 =: +/, positions x_matches/ x_directions

Uiua

Just part1 for now as I need to walk the dog :-)

[edit] Part 2 now added, and a nicer approach than Part 1 in my opinion, if you’re able to keep that many dimensions straight in your head :-)

[edit 2] Tightened it up a bit more.

Grid ← ⊜∘⊸≠@\n "MMMSXXMASM\nMSAMXMSMSA\nAMXSXMAAMM\nMSAMASMSMX\nXMASAMXAMM\nXXAMMXXAMA\nSMSMSASXSS\nSAXAMASAAA\nMAMMMXMMMM\nMXMXAXMASX"

≡⍉⍉×⇡4¤[1_0 0_1 1_1 1_¯1]         # Use core dirs to build sets of 4-offsets.
↯∞_2⇡△ Grid                       # Get all possible starting points.
&p/+♭⊞(+∩(≍"XMAS")⇌.⬚@.⊡:Grid≡+¤) # Part 1. Join the two into a table, use to pick 4-elements, check, count.

Diags   ← [[¯. 1_1] [¯. 1_¯1]]
BothMas ← /×≡(+∩(≍"MS")⇌.)⬚@.⊡≡+Diags¤¤ # True if both diags here are MAS.
&p/+≡BothMas⊚="A"⟜¤Grid                 # Part 2. For all "A"s in grid, check diags, count where good.

Haskell

import Control.Arrow
import Data.Array.Unboxed
import Data.List

type Pos = (Int, Int)
type Board = Array Pos Char
data Dir = N | NE | E | SE | S | SW | W | NW

target = "XMAS"

parse s = listArray ((1, 1), (n, m)) [l !! i !! j | i <- [0 .. n - 1], j <- [0 .. m - 1]]
  where
    l = lines s
    (n, m) = (length $ head l, length l)

move N = first pred
move S = first succ
move E = second pred
move W = second succ
move NW = move N . move W
move SW = move S . move W
move NE = move N . move E
move SE = move S . move E

check :: Board -> Pos -> Int -> Dir -> Bool
check b p i d =
    i >= length target
        || ( inRange (bounds b) p
                && (b ! p) == (target !! i)
                && check b (move d p) (succ i) d
           )

checkAllDirs :: Board -> Pos -> Int
checkAllDirs b p = length . filter (check b p 0) $ [N, NE, E, SE, S, SW, W, NW]

check2 :: Board -> Pos -> Bool
check2 b p =
    all (inRange (bounds b)) moves && ((b ! p) == 'A') && ("SSMM" `elem` rotations)
  where
    rotations = rots $ (b !) <$> moves
    moves = flip move p <$> [NE, SE, SW, NW]

    rots xs = init $ zipWith (++) (tails xs) (inits xs)

part1 b = sum $ checkAllDirs b <$> indices b
part2 b = length . filter (check2 b) $ indices b

main = getContents >>= print . (part1 &&& part2) . parse

Factor

: get-input ( -- rows )
  "vocab:aoc-2024/04/input.txt" utf8 file-lines ;

: verticals ( rows -- lines )
  [ dimension last [0..b) ] keep cols ;

: slash-origins ( rows -- coords )
  dimension
  [ first [0..b) [ 0 2array ] map ] [
    first2 [ 1 - ] [ 1 (a..b] ] bi*
    [ 2array ] with map
  ] bi append ;

: backslash-origins ( rows -- coords )
  dimension first2
  [ [0..b) [ 0 2array ] map ]
  [ 1 (a..b] [ 0 swap 2array ] map ] bi* append ;

: slash ( rows origin -- line )
  first2
  [ 0 [a..b] ]
  [ pick dimension last [a..b) ] bi* zip
  swap matrix-nths ;

: backslash ( rows origin -- line )
  [ dup dimension ] dip first2
  [ over first [a..b) ]
  [ pick last [a..b) ] bi* zip nip
  swap matrix-nths ;

: slashes ( rows -- lines )
  dup slash-origins
  [ slash ] with map ;

: backslashes ( rows -- lines )
  dup backslash-origins
  [ backslash ] with map ;

: word-count ( line word -- n )
  dupd [ reverse ] dip
  '[ _ subseq-indices length ] bi@ + ;

: part1 ( -- n )
  get-input
  { [ ] [ verticals ] [ slashes ] [ backslashes ] } cleave-array concat
  [ "XMAS" word-count ] map-sum ;

: origin-adistances ( rows origins line-quot: ( rows origin -- line ) -- origin-adistances-assoc )
  with zip-with
  "MAS" "SAM" [ '[ [ _ subseq-indices ] map-values ] ] bi@ bi append
  harvest-values
  [ [ 1 + ] map ] map-values ; inline

: a-coords ( origin-adistances coord-quot: ( adistance -- row-delta col-delta ) -- coords )
  '[ first2 [ @ 2array v+ ] with map ] map-concat ; inline

: slash-a-coords ( rows -- coords )
  dup slash-origins [ slash ] origin-adistances
  [ [ 0 swap - ] keep ] a-coords ;

: backslash-a-coords ( rows -- coords )
  dup backslash-origins [ backslash ] origin-adistances
  [ dup ] a-coords ;

: part2 ( -- n )
  get-input [ slash-a-coords ] [ backslash-a-coords ] bi
  intersect length ;

Better viewed on GitHub.

C

What can I say, bunch of for loops! I add a 3 cell border to avoid having to do bounds checking in the inner loops.

#include "common.h"
#define GZ 146

int main(int argc, char **argv) {
	static char g[GZ][GZ];
	static const char w[] = "XMAS";
	int p1=0,p2=0, x,y, m,i;

	if (argc > 1) DISCARD(freopen(argv[1], "r", stdin));
	for (y=3; y<GZ && fgets(g[y]+3, GZ-3, stdin); y++) ;

	for (y=3; y<GZ-3; y++)
	for (x=3; x<GZ-3; x++) {
		for (m=1,i=0; i<4; i++) {m &= g[y+i][x]==w[i];} p1+=m;
		for (m=1,i=0; i<4; i++) {m &= g[y-i][x]==w[i];} p1+=m;
		for (m=1,i=0; i<4; i++) {m &= g[y][x+i]==w[i];} p1+=m;
		for (m=1,i=0; i<4; i++) {m &= g[y][x-i]==w[i];} p1+=m;
		for (m=1,i=0; i<4; i++) {m &= g[y+i][x+i]==w[i];} p1+=m;
		for (m=1,i=0; i<4; i++) {m &= g[y-i][x-i]==w[i];} p1+=m;
		for (m=1,i=0; i<4; i++) {m &= g[y+i][x-i]==w[i];} p1+=m;
		for (m=1,i=0; i<4; i++) {m &= g[y-i][x+i]==w[i];} p1+=m;

		p2 += g[y+1][x+1]=='A' &&
		      ((g[y][x]  =='M' && g[y+2][x+2]=='S')  ||
		       (g[y][x]  =='S' && g[y+2][x+2]=='M')) &&
		      ((g[y+2][x]=='M' && g[y][x+2]  =='S')  ||
		       (g[y+2][x]=='S' && g[y][x+2]  =='M'));
	}

	printf("04: %d %d\n", p1, p2);
}

https://github.com/sjmulder/aoc/blob/master/2024/c/day04.c

Nim

import ../aoc, strutils

type
  Cell* = tuple[x,y:int]

#the 8 grid direction
const directions : array[8, Cell] = [
  (1, 0), (-1, 0),
  (0, 1), ( 0,-1),
  (1, 1), (-1,-1),
  (1,-1), (-1, 1)
]

const xmas = "XMAS"

#part 1
proc searchXMAS*(grid:seq[string], x,y:int):int =
  #search in all 8 directions (provided we can find a full match in that direction)
  let w = grid[0].len
  let h = grid.len
  
  for dir in directions:
    # check if XMAS can even fit
    let xEnd = x + dir.x * 3
    let yEnd = y + dir.y * 3
    if xEnd < 0 or xEnd >= w or
       yEnd < 0 or yEnd >= h:
      continue;
    
    #step along direction
    var matches = 0
    for s in 0..3:
      if grid[y + dir.y * s][x + dir.x * s] == xmas[s]:
        inc matches
        
    if matches == xmas.len:
      inc result

#part 2
proc isMAS(grid:seq[string], c, o:Cell):bool=
  let ca : Cell = (c.x+o.x, c.y+o.y)
  let cb : Cell = (c.x-o.x, c.y-o.y)
  let a = grid[ca.y][ca.x]
  let b = grid[cb.y][cb.x]
  (a == 'M' and b == 'S') or (a == 'S' and b == 'M')

proc searchCrossMAS*(grid:seq[string], x,y:int):bool =
  grid[y][x] == 'A' and
  grid.isMAS((x,y), (1,1)) and
  grid.isMAS((x,y), (1,-1))

proc solve*(input:string): array[2,int] =
  let grid = input.splitLines
  let w = grid[0].len
  let h = grid.len
  
  #part 1
  for y in 0..<h:
    for x in 0..<w:
      result[0] += grid.searchXMAS(x, y)
  
  #part 2, skipping borders
  for y in 1..<h-1:
    for x in 1..<w-1:
      result[1] += (int)grid.searchCrossMAS(x, y)

Part 1 was done really quickly. Part 2 as well, but the result was not accepted…

Turns out +MAS isn’t actually a thing :P

python

import aoc

def setup():
    return (aoc.get_lines(4, padded=(True, '.', 3)), 0)

def one():
    lines, acc = setup()
    for row, l in enumerate(lines):
        for col, c in enumerate(l):
            if c == 'X':
                w = l[col - 3:col + 1]
                e = l[col:col + 4]
                n = c + lines[row - 1][col] + \
                    lines[row - 2][col] + lines[row - 3][col]
                s = c + lines[row + 1][col] + \
                    lines[row + 2][col] + lines[row + 3][col]
                nw = c + lines[row - 1][col - 1] + \
                    lines[row - 2][col - 2] + lines[row - 3][col - 3]
                ne = c + lines[row - 1][col + 1] + \
                    lines[row - 2][col + 2] + lines[row - 3][col + 3]
                sw = c + lines[row + 1][col - 1] + \
                    lines[row + 2][col - 2] + lines[row + 3][col - 3]
                se = c + lines[row + 1][col + 1] + \
                    lines[row + 2][col + 2] + lines[row + 3][col + 3]
                for word in [w, e, n, s, nw, ne, sw, se]:
                    if word in ['XMAS', 'SAMX']:
                        acc += 1
    print(acc)

def two():
    lines, acc = setup()
    for row, l in enumerate(lines):
        for col, c in enumerate(l):
            if c == 'A':
                l = lines[row - 1][col - 1] + c + lines[row + 1][col + 1]
                r = lines[row + 1][col - 1] + c + lines[row - 1][col + 1]
                if l in ['MAS', 'SAM'] and r in ['MAS', 'SAM']:
                    acc += 1
    print(acc)

one()
two()

Rust

Ugh. Spent way too long on today’s. Should have just used my own grid structure from last year. I will likely refactor to use that. Even though it’s likely a super slow implementation, the convenience of dealing with it is better than shoehorning in the grid::Grid<T> from that crate.

use grid::Grid;

use crate::shared::{
    grid2d::{iter_diag_nesw, iter_diag_nwse, Point},
    util::read_lines,
};

fn parse_grid(input: &[String]) -> Grid<u8> {
    let cols = input.first().unwrap().len();
    Grid::from_vec(
        input
            .iter()
            .flat_map(|row| row.chars().map(|c| c as u8).collect::<Vec<u8>>())
            .collect(),
        cols,
    )
}

fn part1(grid: &Grid<u8>) -> usize {
    let mut xmas_count = 0;
    let rows = grid
        .iter_rows()
        .map(|d| String::from_utf8(d.copied().collect()).unwrap());
    let cols = grid
        .iter_cols()
        .map(|d| String::from_utf8(d.copied().collect()).unwrap());
    for diag in iter_diag_nesw(grid)
        .chain(iter_diag_nwse(grid))
        .filter_map(|d| {
            if d.len() >= 4 {
                Some(String::from_utf8(d.clone()).unwrap())
            } else {
                None
            }
        })
        .chain(rows)
        .chain(cols)
    {
        xmas_count += diag.matches("XMAS").count() + diag.matches("SAMX").count()
    }
    xmas_count
}

fn part2(grid: &Grid<u8>) -> usize {
    let mut xmas_count = 0;
    let valid = [
        [b'M', b'M', b'S', b'S'],
        [b'M', b'S', b'S', b'M'],
        [b'S', b'M', b'M', b'S'],
        [b'S', b'S', b'M', b'M'],
    ];
    for x in 1..grid.cols() - 1 {
        for y in 1..grid.rows() - 1 {
            if grid.get(y, x) == Some(&b'A')
                && valid.contains(
                    &(Point::new(x as isize, y as isize)
                        .diagonal_neighbors(grid)
                        .map(|i| i.unwrap_or(0))),
                )
            {
                xmas_count += 1;
            }
        }
    }
    xmas_count
}

pub fn solve() {
    let input = read_lines("inputs/day04.txt");
    let grid = parse_grid(&input);
    println!("Part 1: {}", part1(&grid));
    println!("Part 2: {}", part2(&grid));
}

And here’s a link to the Github if you care to see the gross supporting code :D

Rust

I had a hunch about part two that didn’t pay off, so I over-coded this instead of just using an array of arrays.

use std::{fs, str::FromStr};

use color_eyre::eyre::{Report, Result};

#[derive(Debug, Copy, Clone)]
enum Direction {
    N,
    NE,
    E,
    SE,
    S,
    SW,
    W,
    NW,
}

impl Direction {
    fn all() -> &'static [Direction] {
        &[
            Direction::N,
            Direction::NE,
            Direction::E,
            Direction::SE,
            Direction::S,
            Direction::SW,
            Direction::W,
            Direction::NW,
        ]
    }
}

#[derive(Debug, PartialEq, Eq)]
struct WordSearch {
    grid: Vec<char>,
    width: usize,
    height: usize,
}

impl FromStr for WordSearch {
    type Err = Report;

    fn from_str(s: &str) -> Result<Self, Self::Err> {
        let grid: Vec<_> = s.chars().filter(|&ch| ch != '\n').collect();
        let width = s
            .chars()
            .position(|ch| ch == '\n')
            .ok_or_else(|| Report::msg("grid width cannot be zero, or one line"))?;
        let height = grid.len() / width;
        Ok(Self {
            grid,
            width,
            height,
        })
    }
}

impl WordSearch {
    fn neighbour(&self, i: usize, dir: Direction) -> Option<usize> {
        let width = self.width;
        let length = self.grid.len();
        use Direction::*;
        match dir {
            N if i >= width => Some(i - width),
            NE if i >= width && i % width != width - 1 => Some(i - width + 1),
            E if i % width != width - 1 => Some(i + 1),
            SE if i + width + 1 < length && i % width != width - 1 => Some(i + width + 1),
            S if i + width < length => Some(i + width),
            SW if i + width - 1 < length && i % width != 0 => Some(i + width - 1),
            W if i % width != 0 => Some(i - 1),
            NW if i >= width && i % width != 0 => Some(i - width - 1),
            _ => None,
        }
    }

    fn word_count(&self, word: &str) -> Result<usize> {
        let mut found = 0;
        for i in 0..self.grid.len() {
            for dir in Direction::all() {
                if self.word_present(word, i, *dir) {
                    found += 1;
                }
            }
        }
        Ok(found)
    }

    fn x_count(&self) -> Result<usize> {
        let mut found = 0;
        for i in 0..self.grid.len() {
            if self.x_present(i) {
                found += 1;
            }
        }
        Ok(found)
    }

    fn word_present(&self, word: &str, location: usize, dir: Direction) -> bool {
        let mut next = Some(location);
        for ch in word.chars() {
            let i = if let Some(i) = next {
                i
            } else {
                // Off the edge
                return false;
            };

            if self.grid[i] != ch {
                return false;
            }
            next = self.neighbour(i, dir);
        }
        true
    }

    fn x_present(&self, location: usize) -> bool {
        if self.grid.get(location) != Some(&'A') {
            return false;
        }
        let diags = [
            (Direction::NE, Direction::SW),
            (Direction::NW, Direction::SE),
        ];
        diags.iter().all(|(dir_a, dir_b)| {
            let Some(a_idx) = self.neighbour(location, *dir_a) else {
                return false;
            };
            let Some(b_idx) = self.neighbour(location, *dir_b) else {
                return false;
            };
            let a = self.grid[a_idx];
            let b = self.grid[b_idx];
            (a == 'M' && b == 'S') || (b == 'M' && a == 'S')
        })
    }
}

fn part1(filepath: &str) -> Result<usize> {
    let input = fs::read_to_string(filepath)?;
    let grid = WordSearch::from_str(&input)?;
    grid.word_count("XMAS")
}

fn part2(filepath: &str) -> Result<usize> {
    let input = fs::read_to_string(filepath)?;
    let grid = WordSearch::from_str(&input)?;
    grid.x_count()
}

fn main() -> Result<()> {
    color_eyre::install()?;

    println!("Part 1: {}", part1("d04/input.txt")?);
    println!("Part 2: {}", part2("d04/input.txt")?);
    Ok(())
}

Uiua

This one was nice. The second part seemed quite daunting at first but wasn’t actually that hard in the end.

Run with example input here

Row    ← ⌕ "XMAS"
RevRow ← ⌕"SAMX"
Sum    ← /+/+
Count  ← +∩Sum⊃Row RevRow

PartOne ← (
  &rs ∞ &fo "input-4.txt"
  ⊜∘≠@\n.
  ⊙+⟜∩Count⟜⍉ # horizontal and vertical search
  ⟜(/+⧈(Count⍉≡⬚@ ↻⇡⧻.)4)
  /+⧈(Count⍉≡⬚@ ↻¯⇡⧻.)4
  ++
)

Mask ← °⊚×2⇡5
# Create variations of X-MAS
Vars ← (
  ["M S"
   " A "
   "M S"]
  ≡♭[∩⟜⍉]≡⇌.
  Mask
  ⊏0⊞▽¤
)

PartTwo ← (
  &rs ∞ &fo "input-4.txt"
  ⊜∘≠@\n.
  ⧈(/+♭⊞≍⊙¤Vars▽Mask♭)3_3
  Sum
)

&p "Day 4:"
&pf "Part 1: "
&p PartOne
&pf "Part 2: "
&p PartTwo

C#

public class Day04 : Solver
{
  private int width, height;
  private char[,] data;

  public void Presolve(string input) {
    var lines = input.Trim().Split("\n").ToList();
    height = lines.Count;
    width = lines[0].Length;
    data = new char[height, width];
    for (int i = 0; i < height; i++) {
      for (int j = 0; j < width; j++) {
        data[i, j] = lines[i][j];
      }
    }
  }

  private static readonly string word = "XMAS";

  public string SolveFirst()
  {
    int counter = 0;
    for (int start_i = 0; start_i < height; start_i++) {
      for (int start_j = 0; start_j < width; start_j++) {
        if (data[start_i, start_j] != word[0]) continue;
        for (int di = -1; di <= 1; di++) {
          for (int dj = -1; dj <= 1; dj++) {
            if (di == 0 && dj == 0) continue;
            int end_i = start_i + di * (word.Length - 1);
            int end_j = start_j + dj * (word.Length - 1);
            if (end_i < 0 || end_j < 0 || end_i >= height || end_j >= width) continue;
            for (int k = 1; k < word.Length; k++) {
              if (data[start_i + di * k, start_j + dj * k] != word[k]) break;
              if (k == word.Length - 1) counter++;
            }
          }
        }
      }
    }
    return counter.ToString();
  }

  public string SolveSecond()
  {
    int counter = 0;
    for (int start_i = 1; start_i < height - 1; start_i++) {
      for (int start_j = 1; start_j < width - 1; start_j++) {
        if (data[start_i, start_j] != 'A') continue;
        int even_mas_starts = 0;
        for (int di = -1; di <= 1; di++) {
          for (int dj = -1; dj <= 1; dj++) {
            if (di == 0 && dj == 0) continue;
            if ((di + dj) % 2 != 0) continue;
            if (data[start_i + di, start_j + dj] != 'M') continue;
            if (data[start_i - di, start_j - dj] != 'S') continue;
            even_mas_starts++;
          }
        }
        if (even_mas_starts == 2) counter++;
      }
    }
    return counter.ToString();
  }
}

Raku

Oof, my struggle to make custom index walking paths for part 1 did not pay off for part 2.

sub MAIN($input) {
    my $file = (open $input).slurp;
    my @grid is List = $file.lines».comb».list;
    my @transposedGrid is List = [Z] @grid;
    my @reversedGrid is List = @grid».reverse;
    my @transposedReversedGrid is List = @transposedGrid».reverse;

    my @horizontalScanRows is List = generateScanHorizontal(@grid);
    my @transposedHorizontalScanRows is List = generateScanHorizontal(@transposedGrid);

    my @part-one-counts = [];
    @part-one-counts.push(count-xmas(@grid, @horizontalScanRows)); # Right
    @part-one-counts.push(count-xmas(@transposedGrid, @transposedHorizontalScanRows)); # Down
    @part-one-counts.push(count-xmas(@reversedGrid, @horizontalScanRows)); # Left
    @part-one-counts.push(count-xmas(@transposedReversedGrid, @transposedHorizontalScanRows)); # Up

    my @diagonalScanRows is List = generateScanDiagonal(@grid);
    my @transposedDiagonalScanRows is List = generateScanDiagonal(@transposedGrid);
    @part-one-counts.push(count-xmas(@grid, @diagonalScanRows)); # Down Right
    @part-one-counts.push(count-xmas(@grid, @diagonalScanRows».reverse)); # Up Left
    @part-one-counts.push(count-xmas(@reversedGrid, @diagonalScanRows)); # Down Left
    @part-one-counts.push(count-xmas(@reversedGrid, @diagonalScanRows».reverse)); # Up Right

    my $part-one-solution = @part-one-counts.sum;
    say "part 1: $part-one-solution";


    my @part-two-counts = [];
    @part-two-counts.push(countGridMatches(@grid, (<M . S>,<. A .>,<M . S>)));
    @part-two-counts.push(countGridMatches(@grid, (<S . S>,<. A .>,<M . M>)));
    @part-two-counts.push(countGridMatches(@grid, (<S . M>,<. A .>,<S . M>)));
    @part-two-counts.push(countGridMatches(@grid, (<M . M>,<. A .>,<S . S>)));

    my $part-two-solution = @part-two-counts.sum;
    say "part 2: $part-two-solution";

}

sub count-xmas(@grid, @scanRows) {
    my $xmas-count = 0;
    for @scanRows -> @scanRow {
        my $xmas-pos = 0;
        for @scanRow -> @pos {
            my $char = @grid[@pos[0]][@pos[1]];
            if "X" eq $char {
                $xmas-pos = 1;
            }elsif <X M A S>[$xmas-pos] eq $char {
                if $xmas-pos == 3 {
                    $xmas-pos = 0;
                    $xmas-count += 1;
                } else {
                    $xmas-pos += 1;
                }
            } else {
                $xmas-pos = 0;
            }
        }
    }
    return $xmas-count;
}

sub generateScanHorizontal(@grid) {
    # Horizontal
    my $rows = @grid.elems;
    my $cols = @grid[0].elems;
    my @scanRows = ();
    for 0..^$rows -> $row {
        my @scanRow = ();
        for 0..^$cols -> $col {
            @scanRow.push(($row, $col));
        }
        @scanRows.push(@scanRow);
    }
    return @scanRows.List».List;
}

sub generateScanDiagonal(@grid) {
    # Down-right diagonal
    my $rows = @grid.elems;
    my $cols = @grid[0].elems;
    my @scanRows = ();
    for 0..^($rows + $cols - 1) -> $diag {
        my @scanRow = ();
        my $starting-row = max(-$cols + $diag + 1, 0);
        my $starting-col = max($rows - $diag - 1, 0);
        my $diag-len = min($rows - $starting-row, $cols - $starting-col);
        for 0..^$diag-len -> $diag-pos {
            @scanRow.push(($starting-row + $diag-pos, $starting-col + $diag-pos));
        }
        @scanRows.push(@scanRow);
    }
    return @scanRows.List».List;
}

sub countGridMatches(@grid, @needle) {
    my $count = 0;
    for 0..(@grid.elems - @needle.elems) -> $top {
        TOP-LEFT:
        for 0..(@grid[$top].elems - @needle[0].elems) -> $left {
            for 0..^@needle.elems -> $row-offset {
                for 0..^@needle[$row-offset].elems -> $col-offset {
                    my $needle-char = @needle[$row-offset][$col-offset];
                    next if $needle-char eq ".";
                    next TOP-LEFT if $needle-char ne @grid[$top+$row-offset][$left+$col-offset];
                }
            }
            $count += 1;
        }
    }
    return $count;
}

github