From b478bfaf9db85b139e11ce6216a065ccc3065922 Mon Sep 17 00:00:00 2001
From: Ishan Jain <contact@ishanjain.me>
Date: Tue, 5 Dec 2023 16:38:21 +0530
Subject: [PATCH] day5: wip optimize

---
 src/day5/1.rs |  18 ++++-----
 src/day5/2.rs | 104 +++++++++++++++++++++++++++++++++++++-------------
 2 files changed, 86 insertions(+), 36 deletions(-)

diff --git a/src/day5/1.rs b/src/day5/1.rs
index 77867e0..7fea732 100644
--- a/src/day5/1.rs
+++ b/src/day5/1.rs
@@ -7,24 +7,22 @@ extern crate test;
 
 const INPUTS: [&str; 2] = [include_str!("./sample.txt"), include_str!("./input.txt")];
 
-fn process(data: &str) -> u64 {
+fn process(data: &str) -> i64 {
     let mut lines = data.split("\n\n");
 
-    let seeds: Vec<u64> = lines
+    let seeds = lines
         .next()
-        .map(|x| x.split(' ').skip(1).map(|y| y.parse::<u64>().unwrap()))
-        .unwrap()
-        .collect();
-
-    let maps: Vec<Vec<(Range<u64>, Range<u64>)>> = lines
+        .map(|x| x.split(' ').skip(1).map(|y| y.parse::<i64>().unwrap()))
+        .unwrap();
+    let maps: Vec<Vec<(Range<i64>, Range<i64>)>> = lines
         .map(|lines| {
             lines
                 .lines()
                 .skip(1)
                 .map(|line| {
-                    let y: Vec<u64> = line
+                    let y: Vec<i64> = line
                         .split_ascii_whitespace()
-                        .map(|y| y.parse::<u64>().unwrap())
+                        .map(|y| y.parse::<i64>().unwrap())
                         .collect();
                     let size = y[2];
 
@@ -34,7 +32,7 @@ fn process(data: &str) -> u64 {
         })
         .collect();
 
-    let mut answer = std::u64::MAX;
+    let mut answer = std::i64::MAX;
 
     for mut src in seeds {
         for map in maps.iter() {
diff --git a/src/day5/2.rs b/src/day5/2.rs
index 20ede3f..20e26a0 100644
--- a/src/day5/2.rs
+++ b/src/day5/2.rs
@@ -1,4 +1,4 @@
-#![feature(slice_split_once)]
+#![feature(iter_array_chunks)]
 #![feature(test)]
 
 use std::ops::Range;
@@ -7,59 +7,111 @@ extern crate test;
 
 const INPUTS: [&str; 2] = [include_str!("./sample.txt"), include_str!("./input.txt")];
 
-fn process(data: &str) -> u64 {
+fn process(data: &str) -> i64 {
     let mut lines = data.split("\n\n");
 
-    let seeds: Vec<u64> = lines
+    let mut seeds: Vec<Range<i64>> = lines
         .next()
         .map(|x| {
-            let v: Vec<u64> = x
+            let v: Vec<i64> = x
                 .split(' ')
                 .skip(1)
-                .map(|y| y.parse::<u64>().unwrap())
+                .map(|y| y.parse::<i64>().unwrap())
                 .collect();
 
             v.chunks(2)
-                .flat_map(|x| (x[0]..x[0] + x[1]).collect::<Vec<u64>>())
-                .collect::<Vec<u64>>()
+                .map(|x| (x[0]..x[0] + x[1]))
+                .collect::<Vec<Range<i64>>>()
         })
         .unwrap();
 
-    let maps: Vec<Vec<(Range<u64>, Range<u64>)>> = lines
+    let maps: Vec<Vec<(Range<i64>, Range<i64>)>> = lines
         .map(|lines| {
             lines
                 .lines()
                 .skip(1)
                 .map(|line| {
-                    let y: Vec<u64> = line
-                        .split_ascii_whitespace()
-                        .map(|y| y.parse::<u64>().unwrap())
-                        .collect();
-                    let size = y[2];
+                    line.split_ascii_whitespace()
+                        .map(|y| y.parse::<i64>().unwrap())
+                        .array_chunks::<2>()
+                        .map(|y| {
+                            let size = y[2];
 
-                    (y[1]..y[1] + size, y[0]..y[0] + size)
+                            (y[1]..y[1] + size, y[0]..y[0] + size)
+                        })
+                        .collect()
                 })
                 .collect()
         })
         .collect();
+    for map in maps {
+        let mut out = vec![];
 
-    let mut answer = std::u64::MAX;
+        for seed in seeds.drain(..) {
+            //  split seed based on this rangemap
+            let splitted = split(&map, seed);
 
-    for mut src in seeds {
-        for map in maps.iter() {
-            let dst = map
-                .iter()
-                .find(|x| x.0.contains(&src))
-                .map(|x| src + x.1.start - x.0.start)
-                .unwrap_or(src);
-
-            src = dst;
+            out.extend(splitted);
         }
 
-        answer = std::cmp::min(answer, src);
+        seeds.extend_from_slice(&out);
     }
 
-    answer
+    seeds.into_iter().map(|x| x.start).min().unwrap()
+}
+
+fn split(map: &[(Range<i64>, Range<i64>)], node: Range<i64>) -> Vec<Range<i64>> {
+    let mut out = vec![];
+
+    let mut stack = vec![node];
+
+    while let Some(node) = stack.pop() {
+        let mut found_match = false;
+
+        for (src, dst) in map.iter() {
+            let overlap = std::cmp::max(0, node.end.min(src.end) - node.start.max(src.start));
+
+            if overlap == 0 {
+                continue;
+            } else if node.start > src.start && node.end < src.end {
+                let offset = std::cmp::max(0, node.start - src.start);
+                let dst = dst.start + offset..dst.start + offset + overlap;
+                found_match = true;
+                out.push(dst);
+            } else if node.start < src.start && node.end > src.end {
+                let r1_non = node.start..src.start;
+                let r2_non = src.end..node.end;
+
+                found_match = true;
+
+                out.push(dst.clone());
+                stack.push(r1_non);
+                stack.push(r2_non);
+            } else {
+                // Partial overlap
+
+                let offset = std::cmp::max(0, node.start - src.start);
+
+                let dst_range = dst.start + offset..dst.start + offset + overlap;
+
+                found_match = true;
+                let new_range = if node.start < src.start {
+                    node.start..src.start
+                } else {
+                    src.end..node.end
+                };
+
+                stack.push(new_range);
+                out.push(dst_range);
+            }
+        }
+
+        if !found_match {
+            out.push(node);
+        }
+    }
+
+    out
 }
 
 fn main() {