releascenify 0.3.0

A regex-based Python library to parse scene release names and extract technical metadata

Releascenify

Parse Scene and P2P release names into structured metadata and compare release quality automatically. Built with a rule/regex engine.

Live Demo: https://dmachard.github.io/releascenify/docs/

Why?

Release names follow a strict grammar inherited from the Scene and P2P groups. Once decoded, you can predict the quality, source, language, and release group just from the filename. This library is an alternative to PTN (Parse Torrent Name).

Installation

pip install releascenify

Usage

from releascenify import parse_filename
from releascenify.comparator import get_quality_score, is_better_release

filename = "Gladiator.II.2024.MULTi.2160p.WEB-DL.DV.HDR.H265-GROUP"
parsed = parse_filename(filename)

print(parsed)
# {
#   "title": "Gladiator II",
#   "year": 2024,
#   "resolution": "2160P",
#   "quality": "WEB-DL",
#   "v_quality": "HDR DV",
#   "codec": "H265",
#   "languages": ["MULTI"]
#   ...
# }

# Calculate quality score
score = get_quality_score(parsed)
print(f"Quality Score: {score}")

# Compare two releases
rel_a = parse_filename("Gladiator.II.2024.1080p.BluRay.x264-GROUP")
rel_b = parse_filename("Gladiator.II.2024.2160p.WEB-DL.x265-GROUP")

if is_better_release(rel_b, rel_a):
    print("Release B is better than Release A")

Quality Scoring

For details on how files are rated, scoring weights, and priority factors, see SCORING.md.

Development & Tests

Setup Local Virtual Environment

To set up the development environment, create a virtual environment and install the package in editable mode with development dependencies:

virtualenv .venv
source .venv/bin/activate
pip install -e .[dev]

Run Tests

Run the test suite using pytest:

pytest tests/ -v

Run Website Locally

You need to run a local web server from the repository root:

# Start a local HTTP server
python3 -m http.server 8000

Then navigate to: http://localhost:8000/ (which redirects to /docs/) or directly to http://localhost:8000/docs/.