Time for docs!

deno.json

@@ -2,7 +2,7 @@
 	"name": "@ltgc/kd-tree",
 	"version": "0.0.2",
 	"exports": {
-		"./kd-tree": "./dist/kd-tree.mjs"
+		"./kd-tree": "./dist/kd-tree.d.mts"
 	},
 	"publish": {
 		"include": [
@@ -26,4 +26,4 @@
 	},
 	"nodeModulesDir": false,
 	"npmRegistry": "http://127.0.0.1:80/"
-}
+}

sh/publish.sh

@@ -2,5 +2,5 @@
 echo "Building..."
 shx build
 echo "Publishing to JSR..."
-deno publish --config ./deno.json
+deno publish --allow-dirty --config ./deno.json
 exit

src/kd-tree/index.d.mts

@@ -0,0 +1,98 @@
+/**
+ * k-d Tree JavaScript - v1.1
+ *
+ * https://github.com/ubilabs/kd-tree-javascript
+ * https://github.com/ltgcgo/kd-tree
+ * https://codeberg.org/ltgc/kd-tree
+ *
+ * @author Mircea Pricop <pricop@ubilabs.net>, 2012
+ * @author Martin Kleppe <kleppe@ubilabs.net>, 2012
+ * @author Ubilabs http://ubilabs.net, 2012
+ * @author Lumière Élevé <ble-m@ltgc.cc>, 2025
+ * @author Lightingale Community https://ltgc.cc, 2025
+ * @license MIT License <http://www.opensource.org/licenses/mit-license.php>
+ */
+
+/**
+ * A k-dimensional tree.
+ * ```js
+ * import {
+ * 	KDTree
+ * } from "./kd-tree.mjs";
+ *
+ * const points = [
+ * 	[0, 0, 0],
+ * 	[11, 45, 14],
+ * 	[191, 98, 10],
+ * 	[255, 255, 255]
+ * ];
+ * const prop3Dim = [0, 1, 2];
+ * const dist3Dim = (a, b) => {
+ * 	let sum = 0;
+ * 	for (let e of prop3Dim) {
+ * 		let diff = a[e] - b[e];
+ * 		sum += diff * diff;
+ * 	};
+ * 	return sum;
+ * };
+ *
+ * let exampleTree = new KDTree(points, dist3Dim, prop3Dim);
+ * ```
+ @module
+ */
+
+/*
+ * Node object of a k-dimensional tree.
+ */
+export class TreeNode {
+	/** The left side of the node. */
+	left: TreeNode;
+	/** The right side of the node. */
+	right: TreeNode;
+	/** The referred-to object. */
+	obj: any;
+	/** The parent node of the node. */
+	parent: TreeNode;
+	/** The dimension specifier of this node. */
+	dimension: Number;
+	/**
+	 @param object The referred-to object holding actual data.
+	 @param dimension The dimension specifier of the node.
+	 @param parent The parent of the node.
+	 */
+	constructor(object: any, dimension: Number, parent?: TreeNode);
+}
+
+/*
+ * The k-dimensional tree.
+ */
+export class KDTree {
+	/** The root of the k-d tree. */
+	root: TreeNode;
+	/** The dimension specifier array of the k-d tree. */
+	get dimensions(): ArrayLike<Number>;
+	/** The function used to measure distance. */
+	get metric(): Function;
+	/** Returns the serialized internal structure of the k-d tree. */
+	toJSON(src?: TreeNode): object;
+	/** Insert a point. */
+	insert(point: ArrayLike<Number>): void;
+	/** Remove a point. */
+	remove(point: ArrayLike<Number>): void;
+	/** Returns the nearest N points around the specified coordinate. */
+	nearest(point: ArrayLike<Number>, maxNodes: Number, maxDist: Number): ArrayLike<ArrayLike<any>>;
+	/** Returns the balance factor of the tree. */
+	balanceFactor(): Number;
+	/**
+	* @param points All of the points to be put in the tree.
+	* @param metric The distance measuring function.
+	* @param dimensions The array containing the dimension specifiers.
+	*/
+	constructor(points: ArrayLike<ArrayLike<Number>>, metric: Function, dimensions: ArrayLike<Number>);
+}
+
+/*
+ * Binary heap implementation from http://eloquentjavascript.net/appendix2.html . Rewritten into modern syntax.
+ * I don't know how this specific implemention work yet, so no docs.
+ */
+export class KDBinaryHeap {}

src/kd-tree/index.mjs

@@ -398,5 +398,6 @@ let KDTree = class KDTree {
 
 export {
 	KDTree,
+	TreeNode,
 	KDBinaryHeap
 };