Leftrb is a Left-Leaning Red-Black (LLRB) implementation of 2–3 balanced binary search trees in Python. Port of the code presented by Robert Sedgewick in his paper.
Python
Latest commit a35f54e Feb 16, 2015 @peterhil Merge pull request #1 from dpsanders/patch-1
Two (rather crucial!) typos in the instructions to clone the repo.

README.md

Leftrb/LLRB

Leftrb is a Left-Leaning Red-Black (LLRB) implementation of 2–3 balanced binary search trees in Python.

This is a straightforward port of the Java code presented by Robert Sedgewick in [his paper]((http://www.cs.princeton.edu/~rs/talks/LLRB/LLRB.pdf) and in the book Algorithms, 4th Edition, which is written by Robert Sedgewick and Kevin Wayne. By their permission, the original GPL v3 licensed Java code is licensed as LGPL v3, and ported to Python.

Overview

A balanced binary search tree (BBST) maintains elements in sorted order under dynamic updates (inserts and deletes) and can support various order-specific queries.

Red-black trees are the de facto standard BBST algorithms, and are the underlying data structure for symbol-table implementations within C++, Java, Python, BSD Unix, Linux and many other modern systems.

All red–black trees are based on implementing 2-3 or 2-3-4 trees within a binary tree, using red links to bind together internal nodes into 3-nodes or 4-nodes. Search, insert and delete operations are O(log n) and space requirements are O(n).

However, many traditional implementations have lots of repetitive code on the symmetric branches of rotation and deletion operations. So they are not easy to reason about and augment with other properties, which is what BBST's are often used for: They are used to implement other common data structures like Priority Queues and Interval Trees.

The LLRB method of implementing 2-3 trees is a recent improvement over the traditional implementation — it maintains an additional invariant that all red links must lean left except during inserts and deletes. Because of this, they can be implemented by adding just a few lines of code to standard BST algorithms.

The LLRB tree is based on combining three ideas:

  • Use a recursive implementation.
  • Require that all 3-nodes lean left.
  • Perform rotations on the way up the tree (after the recursive calls).

The LLRB approach was discovered relatively recently (in 2008) by Robert Sedgewick of Princeton University. For original code and more information read the paper "Left-leaning Red-Black Trees" at http://www.cs.princeton.edu/~rs/talks/LLRB/LLRB.pdf

Installation

From Python package index:

pip install leftrb

or from Github source:

git clone https://github.com/peterhil/leftrb.git
cd leftrb 
python setup.py install

About

Leftrb/LLRB was written by Peter Hillerström.
Follow me on Twitter @peterhil!