make a_star function use an IndexMap like the pathfinding crate

azalea/Cargo.toml

@@ -34,6 +34,7 @@ bevy_tasks = { workspace = true, features = ["multi_threaded"] }
 derive_more = { workspace = true, features = ["deref", "deref_mut"] }
 futures = { workspace = true }
 futures-lite = { workspace = true }
+indexmap = "2.7.0"
 nohash-hasher = { workspace = true }
 num-format = { workspace = true }
 num-traits = { workspace = true }

azalea/src/pathfinder/astar.rs

@@ -2,12 +2,13 @@ use std::{
     cmp::{self},
     collections::BinaryHeap,
     fmt::Debug,
-    hash::Hash,
+    hash::{BuildHasherDefault, Hash},
     time::{Duration, Instant},
 };
 
+use indexmap::IndexMap;
 use num_format::ToFormattedString;
-use rustc_hash::FxHashMap;
+use rustc_hash::FxHasher;
 use tracing::{debug, trace, warn};
 
 pub struct Path<P, M>
@@ -37,6 +38,12 @@ pub enum PathfinderTimeout {
     Nodes(usize),
 }
 
+type FxIndexMap<K, V> = IndexMap<K, V, BuildHasherDefault<FxHasher>>;
+
+// Sources:
+// - https://en.wikipedia.org/wiki/A*_search_algorithm
+// - https://github.com/evenfurther/pathfinding/blob/main/src/directed/astar.rs
+// - https://github.com/cabaletta/baritone/blob/1.19.4/src/main/java/baritone/pathing/calc/AbstractNodeCostSearch.java
 pub fn a_star<P, M, HeuristicFn, SuccessorsFn, SuccessFn>(
     start: P,
     heuristic: HeuristicFn,
@@ -52,77 +59,100 @@ where
 {
     let start_time = Instant::now();
 
-    let mut open_set = BinaryHeap::<WeightedNode<P>>::new();
-    open_set.push(WeightedNode(start, 0.));
-    let mut nodes: FxHashMap<P, Node<P, M>> = FxHashMap::default();
+    let mut open_set = BinaryHeap::<WeightedNode>::new();
+    open_set.push(WeightedNode {
+        g_score: 0.,
+        f_score: 0.,
+        index: 0,
+    });
+    let mut nodes: FxIndexMap<P, Node<M>> = IndexMap::default();
     nodes.insert(
         start,
         Node {
-            position: start,
             movement_data: None,
-            came_from: None,
-            g_score: f32::default(),
-            f_score: f32::INFINITY,
+            came_from: usize::MAX,
+            g_score: 0.,
         },
     );
 
-    let mut best_paths: [P; 7] = [start; 7];
+    let mut best_paths: [usize; 7] = [0; 7];
     let mut best_path_scores: [f32; 7] = [heuristic(start); 7];
 
     let mut num_nodes = 0;
 
-    while let Some(WeightedNode(current_node, _)) = open_set.pop() {
+    while let Some(WeightedNode { index, g_score, .. }) = open_set.pop() {
         num_nodes += 1;
-        if success(current_node) {
+
+        let (&node, node_data) = nodes.get_index(index).unwrap();
+        if success(node) {
             debug!("Nodes considered: {num_nodes}");
 
             return Path {
-                movements: reconstruct_path(nodes, current_node),
+                movements: reconstruct_path(nodes, index),
                 partial: false,
             };
         }
 
-        let current_g_score = nodes
-            .get(&current_node)
-            .map(|n| n.g_score)
-            .unwrap_or(f32::INFINITY);
-
-        for neighbor in successors(current_node) {
-            let tentative_g_score = current_g_score + neighbor.cost;
-            let neighbor_g_score = nodes
-                .get(&neighbor.movement.target)
-                .map(|n| n.g_score)
-                .unwrap_or(f32::INFINITY);
-            if neighbor_g_score - tentative_g_score > MIN_IMPROVEMENT {
-                let heuristic = heuristic(neighbor.movement.target);
-                let f_score = tentative_g_score + heuristic;
-                nodes.insert(
-                    neighbor.movement.target,
-                    Node {
-                        position: neighbor.movement.target,
+        if g_score > node_data.g_score {
+            continue;
+        }
+
+        for neighbor in successors(node) {
+            let tentative_g_score = g_score + neighbor.cost;
+            // let neighbor_heuristic = heuristic(neighbor.movement.target);
+            let neighbor_heuristic;
+            let neighbor_index;
+
+            // skip neighbors that don't result in a big enough improvement
+            if tentative_g_score - g_score < MIN_IMPROVEMENT {
+                continue;
+            }
+
+            match nodes.entry(neighbor.movement.target) {
+                indexmap::map::Entry::Occupied(mut e) => {
+                    if e.get().g_score > tentative_g_score {
+                        neighbor_heuristic = heuristic(*e.key());
+                        neighbor_index = e.index();
+                        e.insert(Node {
+                            movement_data: Some(neighbor.movement.data),
+                            came_from: index,
+                            g_score: tentative_g_score,
+                        });
+                    } else {
+                        continue;
+                    }
+                }
+                indexmap::map::Entry::Vacant(e) => {
+                    neighbor_heuristic = heuristic(*e.key());
+                    neighbor_index = e.index();
+                    e.insert(Node {
                         movement_data: Some(neighbor.movement.data),
-                        came_from: Some(current_node),
+                        came_from: index,
                         g_score: tentative_g_score,
-                        f_score,
-                    },
-                );
-                open_set.push(WeightedNode(neighbor.movement.target, f_score));
-
-                for (coefficient_i, &coefficient) in COEFFICIENTS.iter().enumerate() {
-                    let node_score = heuristic + tentative_g_score / coefficient;
-                    if best_path_scores[coefficient_i] - node_score > MIN_IMPROVEMENT {
-                        best_paths[coefficient_i] = neighbor.movement.target;
-                        best_path_scores[coefficient_i] = node_score;
-                    }
+                    });
+                }
+            }
+
+            open_set.push(WeightedNode {
+                index: neighbor_index,
+                g_score: tentative_g_score,
+                f_score: tentative_g_score + neighbor_heuristic,
+            });
+
+            for (coefficient_i, &coefficient) in COEFFICIENTS.iter().enumerate() {
+                let node_score = neighbor_heuristic + tentative_g_score / coefficient;
+                if best_path_scores[coefficient_i] - node_score > MIN_IMPROVEMENT {
+                    best_paths[coefficient_i] = neighbor_index;
+                    best_path_scores[coefficient_i] = node_score;
                 }
             }
         }
 
-        // check for timeout every ~20ms
+        // check for timeout every ~10ms
         if num_nodes % 10000 == 0 {
             let timed_out = match timeout {
-                PathfinderTimeout::Time(max_duration) => start_time.elapsed() > max_duration,
-                PathfinderTimeout::Nodes(max_nodes) => num_nodes > max_nodes,
+                PathfinderTimeout::Time(max_duration) => start_time.elapsed() >= max_duration,
+                PathfinderTimeout::Nodes(max_nodes) => num_nodes >= max_nodes,
             };
             if timed_out {
                 // timeout, just return the best path we have so far
@@ -132,7 +162,7 @@ where
         }
     }
 
-    let best_path = determine_best_path(&best_paths, &start);
+    let best_path = determine_best_path(best_paths, 0);
 
     debug!(
         "A* ran at {} nodes per second",
@@ -146,48 +176,46 @@ where
     }
 }
 
-fn determine_best_path<P>(best_paths: &[P; 7], start: &P) -> P
-where
-    P: Eq + Hash + Copy + Debug,
-{
+fn determine_best_path(best_paths: [usize; 7], start: usize) -> usize {
     // this basically makes sure we don't create a path that's really short
 
-    for node in best_paths.iter() {
+    for node in best_paths {
         if node != start {
-            return *node;
+            return node;
         }
     }
     warn!("No best node found, returning first node");
     best_paths[0]
 }
 
-fn reconstruct_path<P, M>(mut nodes: FxHashMap<P, Node<P, M>>, current: P) -> Vec<Movement<P, M>>
+fn reconstruct_path<P, M>(
+    mut nodes: FxIndexMap<P, Node<M>>,
+    mut current_index: usize,
+) -> Vec<Movement<P, M>>
 where
     P: Eq + Hash + Copy + Debug,
 {
     let mut path = Vec::new();
-    let mut current = current;
-    while let Some(node) = nodes.remove(&current) {
-        if let Some(came_from) = node.came_from {
-            current = came_from;
-        } else {
+    while let Some((&node_position, node)) = nodes.get_index_mut(current_index) {
+        if node.came_from == usize::MAX {
             break;
         }
+
+        current_index = node.came_from;
+
         path.push(Movement {
-            target: node.position,
-            data: node.movement_data.unwrap(),
+            target: node_position,
+            data: node.movement_data.take().unwrap(),
         });
     }
     path.reverse();
     path
 }
 
-pub struct Node<P, M> {
-    pub position: P,
+pub struct Node<M> {
     pub movement_data: Option<M>,
-    pub came_from: Option<P>,
+    pub came_from: usize,
     pub g_score: f32,
-    pub f_score: f32,
 }
 
 pub struct Edge<P: Hash + Copy, M> {
@@ -218,16 +246,30 @@ impl<P: Hash + Copy + Clone, M: Clone> Clone for Movement<P, M> {
 }
 
 #[derive(PartialEq)]
-pub struct WeightedNode<P: PartialEq>(P, f32);
+pub struct WeightedNode {
+    index: usize,
+    g_score: f32,
+    f_score: f32,
+}
 
-impl<P: PartialEq> Ord for WeightedNode<P> {
+impl Ord for WeightedNode {
     fn cmp(&self, other: &Self) -> cmp::Ordering {
         // intentionally inverted to make the BinaryHeap a min-heap
-        other.1.partial_cmp(&self.1).unwrap_or(cmp::Ordering::Equal)
+        match other
+            .f_score
+            .partial_cmp(&self.f_score)
+            .unwrap_or(cmp::Ordering::Equal)
+        {
+            cmp::Ordering::Equal => self
+                .g_score
+                .partial_cmp(&other.g_score)
+                .unwrap_or(cmp::Ordering::Equal),
+            s => s,
+        }
     }
 }
-impl<P: PartialEq> Eq for WeightedNode<P> {}
-impl<P: PartialEq> PartialOrd for WeightedNode<P> {
+impl Eq for WeightedNode {}
+impl PartialOrd for WeightedNode {
     fn partial_cmp(&self, other: &Self) -> Option<cmp::Ordering> {
         Some(self.cmp(other))
     }

azalea/src/pathfinder/mod.rs

@@ -778,7 +778,7 @@ pub fn check_for_path_obstruction(
                             new_path
                                 .extend(executing_path.path.iter().skip(patch_end_index).cloned());
                             is_patch_complete = true;
-                            debug!("the obstruction patch is not partial");
+                            debug!("the obstruction patch is not partial :)");
                         } else {
                             debug!(
                                 "the obstruction patch is partial, throwing away rest of path :("