ChromaDB Example Project

A Python project demonstrating ChromaDB integration for vector similarity search and document storage, with support for loading and querying Markdown files.

Features

• Load and index Markdown documents from a directory
• Support for both persistent and in-memory database modes
• Multiple client setup options (local, server-based, ephemeral)
• Document search with metadata support
• Automatic file opening for search results
• Development mode support

Prerequisites

• Python 3.10 or higher
• Poetry for dependency management

Installation

1. Clone this repository
2. Install dependencies using Poetry:

   poetry install

Project Structure

.
├── src/
│   ├── __init__.py
│   ├── main.py           # Main application entry point
│   ├── chroma_setups.py  # ChromaDB client configuration
│   ├── data_loader.py    # Document loading utilities
│   ├── utils.py          # General utility functions
│   └── config.py         # Application configuration
├── input_data/           # Default directory for markdown files
├── poetry.lock
└── pyproject.toml

Usage

Basic Usage

Run the example application:

poetry run python src/main.py

Development Mode

Set the DEV_MODE environment variable to enable development features:

DEV_MODE=true poetry run python src/main.py

Client Types

The project supports three types of ChromaDB clients:

1. Persistent Client: Default mode, stores data on disk
2. Ephemeral Client: In-memory storage, useful for testing
3. Server Client: Connects to a remote ChromaDB server

Configuration

• Default input directory: input_data/
• ChromaDB server URL: http://localhost:8000 (for server mode)
• File types supported: Markdown (.md)