Receipt App

A small command-line Python application that processes a folder of receipt images and uses the OpenAI API to extract structured information such as date, total amount, vendor, and category.

The program outputs a single JSON dictionary mapping each receipt filename to its extracted fields.

---

Project Structure

lab_1_repo/
├── src/
│   └── receipt_app/
│       ├── __init__.py
│       ├── main.py
│       ├── gpt.py
│       └── file_io.py
├── receipts/
│   └── sample receipt images
├── pyproject.toml
├── Makefile
└── README.md

---

Requirements

• Python 3.10+
• An OpenAI API key
• Conda or virtualenv recommended

---

Setup

1. Create and activate your environment (conda or venv).

2. Install the project and dependencies:

pip install -e .

3. Set your OpenAI API key:

Windows (cmd / Anaconda Prompt)

set OPENAI_API_KEY=YOUR_API_KEY_HERE

macOS / Linux

export OPENAI_API_KEY=YOUR_API_KEY_HERE

---

Usage

Run the application on a folder of receipt images:

python -m receipt_app.main receipts --print

If you added the console script entry point:

receipt-app receipts

The program prints JSON output to stdout.

---

Output Format

Example:

{
  "receipt_1.jpg": {
    "date": "2024-01-12",
    "vendor": "Walmart",
    "amount": 23.45,
    "category": "groceries"
  }
}

If a field cannot be determined, the value will be null.

---

Visualization (Expenses by Category)

The --plot option generates a pie chart saved to expenses_by_category.png.

Readability choices:

• Start angle set to 90° so slices begin at the top.
• Categories sorted by total spent (largest first).
• The largest slice is slightly separated (“explode”) to draw attention.
• A legend is placed outside the pie to avoid cluttering labels on the chart.

---

Notes

• Receipt images are not included in version control.
• This project uses the OpenAI API and requires internet access.
• The codebase follows a standard src/ Python package layout.

---

License

MIT License