DocSearch

A search application that provides quick access to the documents and text files you want on your computer by searching their contents

---

About

Searching through hundreds of documents for a specific piece of content can be tedious. DocSearch solves this by indexing all supported files into a local SQLite database and running full-text searches using the BM25 (Best Match 25) algorithm — returning the most relevant results in milliseconds.

---

Features

• BM25-based search — An advanced ranking algorithm that accounts for term frequency and document length to deliver highly relevant results
• SQLite indexing — File contents are stored in a local database, keeping searches fast and lightweight
• Wide file format support — Supports PDF, TXT, DOCX, and other text-based formats
• GTK interface — A clean, user-friendly interface compatible with GNOME desktop environments

---

Requirements

System Requirements

• Python 3.x
• SQLite (typically bundled with Python)

Python Dependencies

rank-bm25
pyPDF2
fitz
pdfplumber
PyGObject

---

Installation

1. Clone the repository

git clone https://github.com/heyderismayilli092/docsearch.git
cd docsearch

2. Install dependencies

pip install -r requirements.txt

3. Run the installer

sudo python3 ~/docsearch/src/main.py

---

Build .deb package

sudo apt install devscripts git-buildpackage
sudo mk-build-deps -ir
gbp buildpackage --git-export-dir=/tmp/build/docsearch -us -uc

---

Screenshots

---

How It Works

File System
     │
     ▼
[docextract.py]          ← Extracts text from PDF, TXT, DOCX files
     │
     ▼
[docdatabase.py]         ← Stores extracted text in SQLite database
     │
     ▼
[docsearch_functions.py] ← Runs BM25 search against the database
     │
     ▼
[GTK Interface]          ← Displays ranked results to the user

---

Author

Heydar Ismayilli

• GitHub: @heyderismayilli092
• Email: heyderismayilli092@gmail.com

NOTE: This software was prepared as part of the "Teknofest 2026 Pardus Bug Finding and Suggestion Competition"