npx-compress

Script to compress NPX binary files. It is a thin wrapper around mtscomp. Mtscomp can compress one file at a time. This app simply scans a given directory with subdirectories and passes all found files to mtscomp, one at a time.

Install

A preferred method of installation on target machine is via pipx. Once you have pipx installed, run

pipx install git+https://github.com/kavli-ntnu/npx-compress.git
npxcompress --help

It is also possible to use the tool via Docker.

Docker

# Build image with (use whatever tag name you want)
docker build -t npx .
# Run tool via:
docker run --rm npx --help

Development

Make sure to setup git config to handle the formatting/whitespaces correctly. To be able to work cross-platform, we upload all files in LF format. If you work on a windows machine, the default format is CRLF.

# On Windows:
git config --global core.autocrlf true
# On Linux/Mac:
git config --global core.autocrlf input

To turn off the warning (not the functionality), run:

$ git config --global core.safecrlf false

Install the project in editable mode and with development requirements:

pip install -e .[dev]

Initialize pre-comit git hooks (on first setup)

This is only required once when you start to work on the code.

pre-commit install

Versioning

Versioning is done via continuous integration (CI) pipelines also known as GitHub Actions. Currently, the pipeline is triggered on every commit. It reads the version from version.json file (that must be in format MAJOR.MINOR) and adds a build number to it, so it becomes MAJOR.MINOR.BUILD. The build number is simply the counter of pipeline runs. This means that if you re-run the pipeline, you will get a version increase.

GitHub Actions

There is one pipeline/action that runs on every commit. It increases a version number of the package, builds wheel and msi distributables. These distributable files are available in form of an artifact. It is possible to download them as a single .zip archive. See for example this.

Making standalone executables

cx_Freeze may be used to create standalone executable packages for different platforms. It's up to you to create virtual environment with cx_Freeze in it. See original documentation for details.

In general, you should prepare the python environment on your build machine before using cx_Freeze. Run in Python >=3.7:

python -m pip install --upgrade pip build cx_freeze trove-classifiers
pip install .

For example, one may create .msi package for usage under Windows:

python setup_cx.py bdist_msi

Created .msi file will be in dist folder. You are encouraged to use github actions in order to build .msi packages.