QuaDB
This project is in the very early stages of development and is nowhere near ready for use.
Description
QuadDB is an asynchronous octree manager implemented over redb, designed for efficient storage and retrieval of spatial data from disk. Binary encoding is handled using bincode for compactness and speed.
Project Vision
The goal of QuadDB is to provide a robust and efficient solution for managing and updating large spatial datasets, particularly in simulations. The use-case envisioned is for graduate students working on simulations that need to manipulate huge datasets involving particle systems or fluid dynamics where memory size is a constraint and access to a supercomputer is not feasible.
| Feature | Status |
|---|---|
| $2^n$-Tree Backend | :construction: In progress |
| Basic CRUD Operations | :construction: In progress |
| Leaf Streaming | :turtle: Planned |
| Documentation | :turtle: Planned |
| $kd$-Tree Backend | :turtle: Planned |
$2^n$-Tree Backend: An implementation of a generalised octree structure.
Basic CRUD operations: Creating, reading, updating and deleting entities in the database file.
Leaf Streaming: Dynamically retrieving leaf nodes from the database as requests are made through the CRUD interface.
Documentation: Robust documentation once the API is frozen.
$kd$-Tree Backend: An implementation of tree structures that do not form new leaf nodes by equally bipartitioning the space. Allows for custom bipartitioning schemes.
License
This work is distributed under the MIT License. Dependencies either direct or indirect may have different licenses, so all of them have been reproduced in the THIRD-PARTY-LICENSES.md file at the root of this repository.