advent-of-code/aoc2020/src/day16.rs

241 lines
6.5 KiB
Rust
Raw Normal View History

2020-12-17 00:31:36 +01:00
use std::collections::HashMap;
2020-12-16 09:28:45 +01:00
use std::fmt::Write;
use std::ops::RangeInclusive;
use anyhow::{Context, Result};
const INPUT: &str = include_str!("../input/day16.txt");
pub fn run() -> Result<String> {
let mut res = String::with_capacity(128);
writeln!(res, "part 1: {}", part1(INPUT)?)?;
2020-12-17 00:31:36 +01:00
writeln!(res, "part 2: {}", part2(INPUT)?)?;
2020-12-16 09:28:45 +01:00
Ok(res)
}
fn part1(input: &str) -> Result<u64> {
let (fields, _, tickets) = parse_input(input)?;
2020-12-17 00:31:36 +01:00
let fields_vec = fields.values().collect::<Vec<_>>();
2020-12-16 09:28:45 +01:00
Ok(tickets
.iter()
2020-12-17 00:31:36 +01:00
.map(|t| t.invalid_values(&fields_vec))
2020-12-16 09:28:45 +01:00
.flatten()
.sum())
}
2020-12-17 00:31:36 +01:00
fn assign_field_positions(
fields: HashMap<&str, Field>,
tickets: Vec<Ticket>,
) -> HashMap<usize, &str> {
let fields_vec = fields.values().collect::<Vec<_>>();
let tickets: Vec<Ticket> = tickets
.into_iter()
.filter(|t| t.invalid_values(&fields_vec).count() == 0)
.collect();
let num_values = tickets[0].values.len();
// build list of all possibilities for each field
let mut possibilities = fields
.iter()
.map(|(&name, field)| {
let possibilities = (0..num_values)
.into_iter()
.filter(|i| tickets.iter().all(|t| t.valid_field(field, *i)))
.collect();
(name, possibilities)
})
.collect::<HashMap<&str, Vec<usize>>>();
let mut fields_to_assign: Vec<&str> = fields.keys().copied().collect();
let mut field_indices: HashMap<usize, &str> = HashMap::new();
for _ in 0..fields.len() {
// get field with only one possibility, and assign it
let field = fields_to_assign
.iter()
.min_by_key(|&name| possibilities[name].len())
.copied()
.expect("fields_to_assign should never be empty in this loop");
assert_eq!(possibilities[field].len(), 1);
let possibility = possibilities[field][0];
// remove position from other fields' possibilities
possibilities.values_mut().for_each(|list| {
if let Some(pos) = list.iter().position(|idx| *idx == possibility) {
list.swap_remove(pos);
}
});
field_indices.insert(possibility, field);
fields_to_assign.swap_remove(fields_to_assign.iter().position(|f| *f == field).unwrap());
}
field_indices
}
fn part2(input: &str) -> Result<u64> {
let (fields, my_ticket, tickets) = parse_input(input)?;
let field_pos_matches = assign_field_positions(fields, tickets);
Ok(my_ticket
.values
.iter()
.enumerate()
.filter(|(idx, _)| {
let field = field_pos_matches[idx];
field.starts_with("departure")
})
.map(|(_, val)| val)
.product())
}
fn parse_input(input: &str) -> Result<(HashMap<&str, Field>, Ticket, Vec<Ticket>)> {
2020-12-16 09:28:45 +01:00
let mut parts = input.split("\n\n");
let fields_part = parts.next().context("no fields specification found")?;
let my_ticket_part = parts.next().context("no personal ticket found")?;
let tickets_part = parts.next().context("no list of tickets found")?;
let fields = fields_part
.lines()
2020-12-17 00:31:36 +01:00
.map(|line| {
let mut parts = line.split(": ");
let name = parts.next().context("no name found")?;
let field = parts.next().context("no ranges found")?.parse()?;
Ok((name, field))
})
2020-12-16 09:28:45 +01:00
.collect::<Result<_>>()
.context("couldn't parse fields")?;
let my_ticket = my_ticket_part
.lines()
.skip(1)
.map(str::parse)
2020-12-16 09:28:45 +01:00
.next()
.context("no second line for ticket")??;
let tickets = tickets_part
.lines()
.skip(1)
.map(str::parse)
2020-12-16 09:28:45 +01:00
.collect::<Result<_>>()
.context("couldn't parse tickets")?;
Ok((fields, my_ticket, tickets))
}
#[derive(Debug)]
struct Ticket {
values: Vec<u64>,
}
impl Ticket {
2020-12-17 00:31:36 +01:00
fn invalid_values<'a>(&'a self, fields: &'a [&Field]) -> impl Iterator<Item = u64> + 'a {
2020-12-16 09:28:45 +01:00
self.values
.iter()
.copied()
2020-12-17 00:31:36 +01:00
.filter(move |val| !fields.iter().any(|f| f.contains(val)))
}
fn valid_field(&self, field: &Field, val_idx: usize) -> bool {
field.contains(&self.values[val_idx])
2020-12-16 09:28:45 +01:00
}
}
impl std::str::FromStr for Ticket {
type Err = anyhow::Error;
fn from_str(s: &str) -> Result<Self> {
let values = s
.split(',')
.map(|num| {
num.parse()
.with_context(|| format!("couldn't parse `{}`", num))
})
.collect::<Result<_>>()?;
Ok(Ticket { values })
}
}
#[derive(Debug)]
struct Field {
ranges: (RangeInclusive<u64>, RangeInclusive<u64>),
}
impl Field {
fn contains(&self, val: &u64) -> bool {
self.ranges.0.contains(val) || self.ranges.1.contains(val)
}
}
impl std::str::FromStr for Field {
type Err = anyhow::Error;
fn from_str(s: &str) -> Result<Self> {
2020-12-17 00:31:36 +01:00
let mut ranges = s.split(" or ");
2020-12-16 09:28:45 +01:00
let mut first_range = ranges.next().context("no first range found")?.split('-');
let first_range_start = first_range.next().context("no bound for range")?.parse()?;
let first_range_end = first_range.next().context("no bound for range")?.parse()?;
2020-12-16 09:28:45 +01:00
let mut second_range = ranges.next().context("no second range found")?.split('-');
let second_range_start = second_range.next().context("no bound for range")?.parse()?;
let second_range_end = second_range.next().context("no bound for range")?.parse()?;
2020-12-16 09:28:45 +01:00
Ok(Field {
ranges: (
first_range_start..=first_range_end,
second_range_start..=second_range_end,
),
2020-12-16 09:28:45 +01:00
})
}
}
#[cfg(test)]
mod tests {
use super::*;
2020-12-17 00:31:36 +01:00
const PROVIDED1: &str = include_str!("../input/day16_provided1.txt");
const PROVIDED2: &str = include_str!("../input/day16_provided2.txt");
2020-12-16 09:28:45 +01:00
#[test]
fn part1_provided() {
2020-12-17 00:31:36 +01:00
assert_eq!(part1(PROVIDED1).unwrap(), 71);
2020-12-16 09:28:45 +01:00
}
#[test]
fn part1_real() {
assert_eq!(part1(INPUT).unwrap(), 20013);
}
2020-12-17 00:31:36 +01:00
#[test]
fn part2_provided() {
let (fields, _, tickets) = parse_input(PROVIDED2).unwrap();
let matches = assign_field_positions(fields, tickets);
let expected = (&["row", "class", "seat"])
.iter()
.copied()
.enumerate()
.collect();
assert_eq!(matches, expected);
}
#[test]
fn part2_real() {
assert_eq!(part2(INPUT).unwrap(), 5977293343129);
}
2020-12-16 09:28:45 +01:00
}