Metadata-Version: 2.4
Name: paulstretch
Version: 0.1.3
Summary: A modern Python implementation of Paul Nasca's extreme time stretch algorithm with minimal dependencies.
Author-email: Moritz Schaller <kontakt@moritz-schaller.de>
License-File: LICENSE
Requires-Python: >=3.9
Requires-Dist: numpy>=2
Description-Content-Type: text/markdown

# Paulstretch

A modern Python implementation of [Paul Nasca](https://github.com/paulnasca)'s [extreme sound stretch algorithm](https://www.paulnasca.com/algorithms-created-by-me#h.4c6i2abbt3xk). It allows for extremely high stretch factors like 10 or 1000 without introducing the unpleasant grainy artifacts of other time stretch algorithms. It does smear all the transients though.

There was no modern and well maintained python package for this algorithm available on the package index, so I decided to make one. This is a fun DSP algorithm with a permissive license, so everyone should enjoy it!

### Installation

If you want to use this package in your own project, install it from the [python package index](https://pypi.org/project/paulstretch/):

`pip install paulstretch`

If you want to work with the [github repository](https://github.com/MoritzSchaller/paulstretch/), install the latest version of the [uv package manager](https://docs.astral.sh/uv/). A virtual environment will automatically be installed when you run or build the project with uv.

### Run Tests

The project uses pytest as its testing framework. Run them like this:

`uv run pytest`

### Build & Publish Package

To build and publish this package on the [python package index](https://pypi.org), this is what you (or rather I) should do:

1. Increment Version if appropriate

    The package uses semantic versioning, so bump the version depending on the extent of changes.

    `uv version --bump patch` or 

    `uv version --bump minor` or 

    `uv version --bump major`


2. Build both the source distribution and the wheels

    `uv build`


3. Publish the package:

    `uv publish`

    Enter your pypi credentials when prompted.
