Relational Database in Rust

Relational Database in Rust is a from‑scratch implementation of a relational database engine written in pure Rust, using only the Rust standard library. It's a project designed to explore how relational database management systems (RDBMS) work under the hood. From schemas and row storage to indexing, constraints, and persistence.

This is not a wrapper over SQLite or an ORM. This is a ground‑up exploration of database internals.

How Relational Databases work under the hood

What do we generally know about Relational Database systems such as MySQL?

• A database is a collection of Tables
• Data is presented as a collection of Tables, each of which have rows and columns
• Users have the ability to manipulate data in tabular form.

Core Components

Storage Layer

• database.rs
  • create_table
  • update_table_name
  • delete_table
  • get_table
  • get_table_mut
• table.rs
  • new
  • add_row
  • delete_row
  • edit_row
  • get_row
• schema.rs
  • new
  • get_column_by_name
  • get_column_by_index
  • get_column_index
• row.rs
  • new
• column.rs
  • new
  • not_null
  • unique
  • default
  • index
  • build
• constraint_state.rs
  • ConstraintKind (enum)
  • Constraint (enum)
  • ConstraintState (struct)
  • from_schema

SQL Pipeline

• tokenizer.rs - converts SQL string to tokens
• parser.rs - builds Abstract Syntax Tree from tokens
• executor.rs - executes queries against the database
• repl.rs - interactive SQL shell

Installation

# Clone the repository
git clone https://github.com/philbotar/relational-database-rust
cd rust-db/rust-database

# Build the project
cargo build

How to Run

# Run the REPL
cd rust-db/rust-database
cargo run

REPL

Start the interactive REPL with cargo run. The REPL supports:

SQL Statements

• SELECT: Query data with column projection, WHERE clauses, ORDER BY, and LIMIT
• INSERT INTO: Add new rows to tables
• UPDATE ... SET ... WHERE: Modify existing rows
• DELETE FROM ... WHERE: Remove rows from tables
• CREATE TABLE: Define new tables with schemas

WHERE Clause Operators

• Comparison: =, !=, >, <, >=, <=
• Logical: AND, OR

Column Constraints

• NOT NULL
• UNIQUE
• DEFAULT
• INDEX

Data Types

• STRING - text values
• INTEGER - 64-bit integers
• BOOLEAN - true/false values

Special Commands

am

• EXIT or QUIT - exit the REPL
• HELP - show available commands
• TABLES or .TABLES - list tables

Examples

-- Create a table with constraints
CREATE TABLE users (id INTEGER NOT NULL UNIQUE, name STRING, age INTEGER);

-- Insert data
INSERT INTO users VALUES (1, 'Alice', 30);
INSERT INTO users VALUES (2, 'Bob', 25);

-- Query data
SELECT * FROM users;
SELECT name, age FROM users WHERE age > 20;
SELECT * FROM users ORDER BY age LIMIT 1;

-- Update data
UPDATE users SET age = 31 WHERE name = 'Alice';

-- Delete data
DELETE FROM users WHERE id = 1;

To-Do

• Be able to create a schema
• Be able to create a database
• Be able to create Tables. Each table has a name, schema and metadata against it.
• Be able to add rows to a table
• Add ability to have column constraints
• Apply Constraints to each row.
• Add "Check" Constraint (Expressions)
• Add the ability to have foreign_keys.
• Persist the data locally.
• Create a Tokenizer to split Queries into tokens
• Create a Parser to convert tokens to an Abstract Syntax Tree
• Be able to perform queries on the data
• Add ability to have indexes against the data