Hi HN! I built a data structures library for JavaScript/TypeScript that I've been refining for 2+ years.
Why another one?
Native JS only gives us Array, Map, Set, Object. When you need a sorted map, priority queue, or graph algorithms, you're stuck with either: - Half-baked npm packages with poor TS support - Porting Java/C++ code yourself - Libraries that expose too many internals
What's different:
1. Native API feel - TreeMap/TreeSet work like Map/Set, just sorted:
2. Java-style navigation - floor(), ceiling(), lower(), higher(), rangeSearch()
3. Full TypeScript - No any types, proper generics throughout
4. Performance - Red-Black Tree backed, ~2x faster iteration than js-sdsl
Hi HN! I built a data structures library for JavaScript/TypeScript that I've been refining for 2+ years.
*Why another one?*
Native JS only gives us Array, Map, Set, Object. When you need a sorted map, priority queue, or graph algorithms, you're stuck with either:
- Half-baked npm packages with poor TS support
- Porting Java/C++ code yourself
- Libraries that expose too many internals
*What's different:*
1. *Native API feel* - TreeMap/TreeSet work like Map/Set, just sorted:
```ts
const map = new TreeMap([['b', 2], ['a', 1], ['c', 3]]);
[...map.keys()] // ['a', 'b', 'c'] - sorted!
```
Happy to answer questions about implementation details, performance tradeoffs, or TypeScript typing challenges!
---
## Alternative shorter version:
*Title:* Show HN: TreeMap, Heap, Graph for TypeScript – native API feel, full type safety
*Text:*
After 2 years of refinement, sharing my TypeScript data structures library.
TreeMap/TreeSet work like native Map/Set but sorted. Includes Java-style navigation (floor, ceiling, rangeSearch). Also: Heap, PriorityQueue, Graph with Dijkstra/Bellman-Ford/Tarjan, Trie, and more.
Red-Black Tree backed, ~2x faster iteration than js-sdsl.
Would love feedback, especially on API design choices!
Why another one? Native JS only gives us Array, Map, Set, Object. When you need a sorted map, priority queue, or graph algorithms, you're stuck with either: - Half-baked npm packages with poor TS support - Porting Java/C++ code yourself - Libraries that expose too many internals What's different:
1. Native API feel - TreeMap/TreeSet work like Map/Set, just sorted: 2. Java-style navigation - floor(), ceiling(), lower(), higher(), rangeSearch()
3. Full TypeScript - No any types, proper generics throughout
4. Performance - Red-Black Tree backed, ~2x faster iteration than js-sdsl
Included: - Binary Trees: BST, AVL, Red-Black, TreeMap, TreeSet, TreeMultiMap - Heaps: Min/Max Heap, Priority Queue - Graphs: Directed/Undirected with Dijkstra, Bellman-Ford, Tarjan, etc. - Linear: Queue, Deque, Stack, LinkedList - Others: Trie, HashMap
npm: npm install data-structure-typed
Happy to answer questions about implementation details, performance tradeoffs, or TypeScript typing challenges!