When solving Advent of Code puzzles, sometimes you will come across an input data, that consist of two parts: configuration and instructions:

    [S] [C]         [Z]            
[F] [J] [P]         [T]     [N]    
[G] [H] [G] [Q]     [G]     [D]    
[V] [V] [D] [G] [F] [D]     [V]    
[R] [B] [F] [N] [N] [Q] [L] [S]    
[J] [M] [M] [P] [H] [V] [B] [B] [D]
[L] [P] [H] [D] [L] [F] [D] [J] [L]
[D] [T] [V] [M] [J] [N] [F] [M] [G]
 1   2   3   4   5   6   7   8   9 

move 3 from 4 to 6
move 1 from 5 to 8
move 3 from 7 to 3
move 4 from 5 to 7
move 1 from 7 to 8
move 3 from 9 to 4

While it's perfectly fine to write some kind of parser for it, very often I manually make the data easier to work with, which can be super fast thanks to text editor features, like multiline-cursor, or select all occurrences.

For example, with this particular puzzle, I thought that it might be a good idea to store the first part of the input as an object, so I manually typed the stack for each column, and then used multiline-cursor to append split part of the code. It took me only couple of seconds, definitely much faster than writing the whole parser for it.

let MAP = {
  1: 'DLJRVGF'.split(''),
  2: 'TPMBVHJS'.split(''),
  3: 'VHMFDGPC'.split(''),
  4: 'MDPNGQ'.split(''),
  5: 'JLHNF'.split(''),
  6: 'NFVQDGTZ'.split(''),
  7: 'FDBL'.split(''),
  8: 'MJBSVDN'.split(''),
  9: 'GLD'.split(''),
}

Second part, wouldn't require that much parsing, but still, instead of doing that, I removed the text by selecting all occurrences, and left only important bits of information:

3 4 6
1 5 8
3 7 3
4 5 7
1 7 8
3 9 4
2 8 2

With that, I could iterate over list of steps pretty easily:

let instructions = input
  .split('\n')
  .map(i => i.split(' ')
  .map(Number))

for (let [quantity, from, to] of instructions) {
   ...
}

---

Advent of Code - 2022 Day 5