1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55
//! # Tuning Trouble
//!
//! One solution to this problem is to use the [`windows`] method to slide over groups of the desired
//! size, then construct a [`HashSet`] from the characters. If the [`HashSet`] is the same size
//! as the window then we know that all characters are unique, as sets contain no duplicate elements.
//!
//! We'll use a faster approach that minimizes the work needed. Instead of creating a set for each
//! window, we'll maintain the last position seen of each character. As we advance character by
//! character we lookup the previous position. If this is within the packet size, then we advance
//! the start of the packet to exclude that character. Once the packet has reached the desired
//! size then we return the current index.
//!
//! [`windows`]: slice::windows
//! [`HashSet`]: std::collections::HashSet
/// Return the input directly.
pub fn parse(input: &str) -> &str {
input
}
/// Find the first unique set of size 4
pub fn part1(input: &str) -> usize {
find(input, 4)
}
/// Find the first unique set of size 14
pub fn part2(input: &str) -> usize {
find(input, 14)
}
/// The cardinality of the input is only 26 so a fixed size array can store the last position
/// of each character.
fn find(input: &str, marker: usize) -> usize {
let mut start = 0;
let mut seen = [0; 26];
for (i, b) in input.bytes().enumerate() {
// Use the character as an index into the array.
let index = (b - b'a') as usize;
let previous = seen[index];
// Positions are 1-based.
seen[index] = i + 1;
// There's a duplicate so advance the start of the window one character past it.
if previous > start {
start = previous;
}
// We've reached the desired packet size with no duplicates so finish.
if i + 1 - start == marker {
return i + 1;
}
}
unreachable!()
}