Day 4Scratchcards

View puzzle description on Advent of Code

This was relatively straightforward. I was able to reuse a lot of code between the two parts. The majority of the code is just parsing the numbers from the string, the implementation is pretty simple.

I'm sure there must be a faster way to do part 2. I'd imagine it can all be done mathematically without the need for nested loops, given the list of scores. Figuring out how to do that is beyond my skill though.

import {
  filter,
  intersection,
  map,
  match,
  nth,
  pipe,
  split,
  sum,
  trim,
} from 'rambda';


const getFirstMatch = (regex: RegExp) =>
  pipe<[string], string[], string>(match(regex), nth(1));


const oneOrMoreSpacesRegex = /\s+/g;
const parseNumbersFromString = pipe(
  split(oneOrMoreSpacesRegex),
  map(pipe(trim, Number)),
);


const winningNumbersRegex = /:\s*(.*?)\s*\|/;
const parseWinningNumbers = pipe(
  getFirstMatch(winningNumbersRegex),
  parseNumbersFromString,
);


const myNumbersRegex = /\|\s*(.*?)\s*$/;
const parseMyNumbers = pipe(
  getFirstMatch(myNumbersRegex),
  parseNumbersFromString,
);


const calculateScore = (x: number) => Math.pow(2, x - 1);


export default pipe(
  split('\n'),
  map(line => {
    const myNumbers = parseMyNumbers(line);
    const winningNumbers = parseWinningNumbers(line);


    return intersection(myNumbers, winningNumbers).length;
  }),
  filter(x => x > 0),
  map(calculateScore),
  sum,
);