espresso-py

Educational Python reimplementation of the Espresso two-level Boolean logic minimizer.

Overview

This project provides a small, readable implementation of the classic Espresso flow (expand → reduce → irredundant) for Boolean function minimization over PLA inputs. It prints the original and optimized truth tables, the Boolean expressions (sum-of-products), and compares literal counts before and after optimization.

Focused on clarity and correctness for small PLAs; not intended for large-scale/industrial use.

Project Structure

espresso-py/
├── src/
│   ├── pla.py              # PLA data structures and parsing
│   ├── visualization.py    # Truth table + SOP printing and metrics
│   └── espresso.py         # Expand/Reduce/Irredundant + simple iterative loop
├── examples/               # Example PLA files
├── main.py                 # CLI entry point
├── requirements.txt        # Minimal dependencies
└── README.md

Installation

python -m venv .venv && source .venv/bin/activate
pip install -r requirements.txt

Usage

# Run with an example file
python main.py examples/simple.pla

# Run with your own PLA file
python main.py path/to/your/file.pla

The program prints the original/optimized truth tables and SOP expressions, and a literal count comparison.

Features

• PLA parsing and validation
• Truth table rendering (pandas DataFrame)
• SOP expression printing (single and multi-output)
• Simple Espresso-style cycle: expand → reduce → irredundant with a small iterative improvement loop
• Literal count comparison before/after optimization

Limitations

• Educational heuristic, not a full Espresso reimplementation
• No explicit handling of separate on/off/don’t-care sets
• Aimed at small inputs; no performance tuning

Example

python main.py examples/4x1_parity.pla

Don't-care demo (shows a clear reduction by leveraging - rows):

python main.py examples/dont_care.pla

Requirements

• Python 3.8+
• pandas

License

MIT — see LICENSE.

Credits

Inspired by the original Espresso implementation (UC Berkeley).