Mini DB

Mini DB is an educational, relational database built from scratch in C++17. The primary goal is to provide a clean, understandable implementation of core database concepts such as storage management, SQL parsing, and query execution.

🚀 Features

• Slotted Page Storage: Professional page layout with a slot directory for efficient variable-length record management.
• SQL Interface: Supports core SQL operations:
  • CREATE TABLE with INT and TEXT types.
  • INSERT INTO records.
  • SELECT * with optional WHERE filters and JOIN.
  • DELETE records with space reclamation support.
  • UPDATE existing records.
• Indexing:
  • Hash Index: O(1) equality lookups for fast retrieval.
  • B-Tree Index: Ordered indexing supporting range queries (>, <, >=, <=).
• Networking: TCP Client/Server architecture for remote SQL execution.
• Persistence: Fixed-size 4KB pages in a single database file; table schemas persist across restarts.
• Robust Testing: 100+ unit and integration tests using GoogleTest.
• Benchmarking: Built-in utility to compare scan vs. index performance.

🏗️ Architecture

Mini DB follows a clean layered architecture:

1. Storage (src/storage/): Manages the Pager (File I/O), Page layout (Slotted Pages), and Record serialization.
2. Catalog (src/catalog/): Handles metadata, table schemas, and persistence of definitions.
3. Parser (src/parser/): A custom Lexer and recursive-descent Parser that converts SQL strings into an AST.
4. Execution (src/execution/): A dispatcher that bridges the Parser and Storage/Catalog layers to run queries.
5. Network (src/network/): TCP server/client implementation for remote interactions.
6. REPL (src/repl/): An interactive command-line shell.

🛠️ Getting Started

Prerequisites

• CMake (3.14+)
• C++17 Compatible Compiler (GCC 7+, Clang 5+)
• Git

Build Instructions

# Clone the repository (if not already done)
git clone <repo-url>
cd database

# Configure and build
cmake -S . -B build
cmake --build build -j$(nproc)

Running Tests

ctest --test-dir build --output-on-failure

Running the REPL

./build/src/minidb_repl

Running the Server

# Start the server in one terminal
./build/src/minidb_server

# Connect using the client in another terminal
./build/src/minidb_client 127.0.0.1 "SELECT * FROM users;"

📝 Example Usage

-- Create a table
CREATE TABLE users (id INT, name TEXT);

-- Create an index for fast lookups
CREATE INDEX id_idx ON users (id);

-- Insert data
INSERT INTO users VALUES (1, 'Alice');
INSERT INTO users VALUES (2, 'Bob');

-- Query with filtering
SELECT * FROM users WHERE id = 1;

-- Update data
UPDATE users SET name = 'Alicia' WHERE id = 1;

-- Relational Join
CREATE TABLE orders (order_id INT, user_id INT);
INSERT INTO orders VALUES (101, 1);
SELECT * FROM users JOIN orders ON users.id = orders.user_id;

-- Delete data
DELETE FROM users WHERE id = 2;

📜 Roadmap

• Milestone 1-7: Core Engine (Pager to REPL)
• Milestone 8: Catalog Persistence
• Milestone 9: Basic Indexing (Hash)
• Milestone 10: Extensive Testing & Benchmarking
• Milestone 11: Networking
• Milestone 12: B-Tree Indexes (Range Queries)
• Milestone 13: Slotted Pages
• Milestone 14: UPDATE & Joins
• Milestone 15: Transactions & WAL (Future)
• Milestone 16: Aggregations (Future)

⚖️ License

This project is for educational purposes.