kalabalaDB

A lightweight Java database engine simulation that demonstrates how a storage manager, query execution layer, and indexing structures work together to process data.

What this project does

• Implements a basic database engine with table creation, record insertion, deletion, update, and selection.
• Uses a clustering key to organize table rows and manage paged storage on disk.
• Supports B+ tree indexes for fast equality and range queries.
• Supports R-tree spatial indexing for polygon-shaped data.
• Demonstrates query execution through a simple SQLTerm API.

Theory behind the design

This project models the core database engine concepts:

• Storage management: tables are split into pages and persisted to disk, with the engine managing page layout and record placement.
• Clustering key: one column is chosen as the table's clustering key, and rows are ordered by that key for efficient range access.
• B+ tree indexing: used for indexed columns to accelerate search and range operations while keeping sorted leaf pages.
• R-tree indexing: used for geometric data types such as Polygon to allow spatial queries.
• Query execution: conditions are expressed as SQLTerm objects and combined with boolean operators to filter rows using available indexes when possible.

How to run

1. Open a terminal in the Database_Engine folder.

2. Build the project:

   make all

3. Run the main test harness:

   make run-DBApp

If make is not available, compile and run with Java directly:

javac src/kalabalaDB/*.java libs/BPTree/*.java libs/General/*.java libs/RTree/*.java -d classes/
java -classpath classes/ kalabalaDB.DBAppTest

How to test

• The DBAppTest class is the entry point used by the build script and contains example database initialization and query code.

• Running make run-DBApp executes that test harness.

• Use make clean to remove generated class files and temporary data:

  make clean

Project structure

• src/kalabalaDB: core engine code and test harness.
• libs/BPTree: B+ tree implementation.
• libs/RTree: R-tree spatial index implementation.
• classes: compiled bytecode output.
• Makefile: build and run commands.

Notes

• This repository is intended as an educational engine prototype, not a production database.
• It focuses on storage and indexing algorithms rather than a full SQL parser or transaction system.