NITCbase Project

A mini relational database management system (RDBMS) built in C++ by following the official NITCbase roadmap.

This project demonstrates how a database works internally, starting from disk block storage up to relational algebra operations like select, project, and join.

---

Project Overview

The goal of this project was to understand and implement the core internal working of a DBMS.

The implementation starts from low-level disk operations and gradually builds higher layers such as buffer management, cache management, schema handling, and query operations.

---

Build and Run

Compile:

make clean make

Run:

./nitcbase

Screenshots ← paste here

Project Build

Projection Output

Selection Output

Join Output

Rename Output

Technologies Used

Step-by-Step Implementation

1. Project Setup

The project was created and compiled using:

make

The required readline library was installed for frontend interaction.

---

2. Disk Layer

Implemented:

• Disk::readBlock()
• Disk::writeBlock()

Purpose:

• Reads fixed-size blocks from the database file
• Writes modified blocks back to disk

This is the lowest storage layer.

---

3. Buffer Layer

Implemented:

• StaticBuffer
• BlockBuffer
• RecBuffer

Purpose:

• Loads disk blocks into memory
• Reduces repeated disk access
• Improves speed

---

4. Cache Layer

Implemented:

• RelCacheTable
• AttrCacheTable
• OpenRelTable

Purpose:

• Stores relation metadata in memory
• Stores attribute metadata in memory
• Tracks opened relations

This helps fast metadata access.

---

5. Linear Search

Implemented:

• BlockAccess::linearSearch()
• BlockAccess::search()

Purpose:

• Searches records by scanning one by one
• Finds records matching conditions

---

6. Record Insert

Implemented:

• BlockAccess::insert()

Purpose:

• Inserts records into free slots
• Updates slot map and headers

Example:

101 Saurav 95

---

7. Relation Open / Close

Implemented:

• OpenRelTable::openRel()
• OpenRelTable::closeRel()

Purpose:

• Opens relations into cache
• Closes them after use

---

8. Create / Delete Relation

Implemented:

• Schema::createRel()
• Schema::deleteRel()

Purpose:

• Adds relation metadata into RELATIONCAT
• Adds attribute metadata into ATTRIBUTECAT
• Deletes relation and its attributes

---

9. Projection

Implemented:

• BlockAccess::project()

Purpose:

• Reads records from relation

Output example:

RELATIONCAT 6 4 4 4 20

---

10. Rename Operations

Implemented:

• BlockAccess::renameRelation()
• BlockAccess::renameAttribute()

Purpose:

• Rename table names
• Rename column names

Example:

Students → Learners → Students

---

11. Selection

Implemented:

• Algebra::select()

Purpose:

• Select records using conditions

Example:

SELECT * FROM Students WHERE Name = "Saurav"

---

12. Join Operation

Implemented:

• Algebra::join()

Purpose:

• Combines two relations using common attributes

Example:

RELATIONCAT JOIN ATTRIBUTECAT ON RelName

---

Roadmap Progress

Stage	Status
Stage 0	Completed
Stage 1	Completed
Stage 2	Completed
Stage 3	Completed
Stage 4	Completed
Stage 5	Completed
Stage 6	Completed
Stage 7	Completed
Stage 8	Completed
Stage 9	Completed
Stage 10	Completed
Stage 11	Completed
Stage 12	Completed

---

Project Structure

• Algebra/ → Relational operations
• BlockAccess/ → Record access methods
• Buffer/ → Buffer management
• Cache/ → Metadata caching
• Disk_Class/ → Disk operations
• Frontend/ → User interface
• FrontendInterface/ → Command parser
• Schema/ → Schema management
• define/ → Constants and IDs

---

Build and Run

Compile:

make clean
make

Run:

./nitcbase

---

Technologies Used

• C++
• File Handling
• Data Structures
• Buffer Management
• Relational Algebra
• Database Internals

---

Future Work

• B+ Tree Indexing
• Index-based Search
• Query Optimization
• Full SQL Support

---

Author

Saurav Kumar Singh B.Tech CSE National Institute of Technology Calicut