bert/adventofcode
1
0
Fork 0
mirror of https://github.com/bertptrs/adventofcode.git synced 2025-12-25 12:50:32 +01:00
Code Issues Projects Releases Wiki Activity

Use running masks instead of computing one by one

Browse Source
This commit is contained in:
Bert Peters
2022-01-06 23:05:41 +01:00
parent 2dab7342f8
commit ad0b4a4659
1 changed files with 11 additions and 12 deletions
Download Patch File Download Diff File Expand all files Collapse all files

23
2021/src/day20.rs
View File

@@ -50,24 +50,23 @@ impl Field {
let mut new_finite = BitSet::with_capacity(new_width * new_height);
// Now we can just do a normal loop
for y in 0..new_height {
let mut mask = if self.infinity { INDEX_MASK } else { 0 };
for x in 0..new_width {
let mut mask = if self.infinity { INDEX_MASK } else { 0 };
const COLUMN_MASK: usize = 0b001001001;
let mut submask = if self.infinity { COLUMN_MASK } else { 0 };
for y in y.saturating_sub(2)..=y {
if x < 2 {
for _ in 0..(2 - x) {
mask = self.infinity as usize | (mask << 1);
}
}
for x in x.saturating_sub(2)..=x {
mask = (mask << 1) | (self[(x, y)] as usize);
}
submask = (submask << 3) | (self[(x, y)] as usize);
}
if translation[mask & INDEX_MASK] {
mask <<= 1;
mask &= !COLUMN_MASK;
mask |= submask;
mask &= INDEX_MASK;
if translation[mask] {
let index = x + y * new_width;
new_finite.insert(index);
}