Struct Tree 

pub struct Tree<T, C = i32> {
    pub arena: Vec<Node<T, C>>,
    pub free_list: Vec<usize>,
    pub root: Option<usize>,
    pub item_count: i32,
    pub dead_count: i32,
    pub extent: KdBox<C>,
    pub delete_flip: bool,
}

A 3D KD-Tree implementation using arena allocation.

The tree stores items of type T associated with a 6-element bounding box [left, bottom, floor, right, top, ceil]. It uses a Vec-backed arena for node storage to improve performance and cache locality.

Fields

arena: Vec<Node<T, C>>

free_list: Vec<usize>

root: Option<usize>

item_count: i32

dead_count: i32

extent: KdBox<C>

delete_flip: bool

Implementations

impl<T: PartialEq + Clone, C: Coord> Tree<T, C>

pub fn new() -> Self

Creates a new, empty 3D KD-Tree.

pub fn with_capacity(capacity: usize) -> Self

Creates a new 3D KD-Tree with a pre-allocated capacity for the node arena.

pub fn insert(&mut self, item: T, size: KdBox<C>)

Inserts an item into the tree with the given 3D bounding box.

pub fn is_member(&self, item: &T, size: &KdBox<C>) -> bool

Checks if the given item is stored in the tree with the specified bounding box.

pub fn delete(&mut self, item: &T, size: &KdBox<C>) -> bool

Marks an item as deleted (soft delete).

pub fn hard_delete(&mut self, item: &T, size: &KdBox<C>) -> bool

Physically deletes an item from the tree and restructures the nodes (hard delete).

pub fn count(&self) -> i32

pub fn badness(&self)

impl<T: PartialEq + Clone, C: Coord> Tree<T, C>

pub fn start(&self, area: KdBox<C>) -> Generator<'_, T, C>

Returns an iterator over all items that intersect with the given 3D bounding box.

pub fn nearest(&self, x: C, y: C, z: C, m: usize) -> Vec<Priority<T>>

Finds the m nearest neighbors to the point (x, y, z).

pub fn really_delete( &mut self, data: &T, old_size: &KdBox<C>, ) -> (Status, i32, i32)

Deletes structurally an item from the tree (really delete).