B-Tree Data Structure

This project implements a disk-based B-Tree data structure that simulates how databases manage indexing when nodes must be persisted to disk instead of memory. The tree supports efficient insert(key, value) and find(key) operations while maintaining balanced height, sorted keys, and configurable branching factors (M and L). Each node is stored in its own disk block through a custom DISK interface, introducing strict constraints that resemble real storage systems.

• Managing a fully disk-backed data structure introduces complexity: every update requires explicit writes, and stale in-memory references easily lead to structural corruption.
• Node splitting is non-trivial: handling medians, redistributing children, and maintaining invariants across multiple recursive levels requires precise logic, especially when no node can exist purely in memory.
• Debugging is harder than in-memory trees because errors often arise from incorrect disk writes, missing updates, or reading outdated nodes—issues that normal pointer-based trees don’t encounter.
• Testing reveals subtle structural bugs (off-by-one errors, incorrect child pointer movement, unbalanced splits), which were identified and resolved using extensive Pytest cases and custom visualizations.
• Overall, the project demands careful coordination between algorithm design and simulated storage constraints, making it significantly more challenging than standard B-Tree implementations.