Why?

It's annoying to lookup new repos on GitHub only to see instructions like:

pip install -r requirements.txt

Where am I supposed to install all these packages into?

Possibly there will be an instruction like:

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

But if the first line's executable python is a wrong Python version? How can I change that? It's just so annoying.

Or they might write python3.11 -m venv venv. But if my system does not have python3.11?

Don't you wish there would be an easier way to download not only Python packages separately for each GitHub repo you clone, but also manage their versions, and the version of Python itself, in an easy out-of-the-box way?

Enter: python-deployer

python-deployer is an easy answer to the above questions. In short, just do something like:

git clone https://github.com/dimitri-rusin/python-deployer.git

git clone https://github.com/ferdinandvanwyk/pyfilm # this can be any other GitHub repo with an annoying `pip install -r requirements.txt` instruction
cd pyfilm

# This is the magic command that manages a Python environment for you
../python-deployer/DEPLOY .

# Adapt .python/pip.txt
.python/BUILD
# Adapt .python/RUN
.python/RUN

The Python environment that you create with ../python-deployer/DEPLOY . consists of:

• Miniconda3 that lets you adapt your Python version, just go to the file .python/Miniconda3.yaml
• A pip.txt file, where you can adapt your Python packages and their versions.
• Everything is completely contained within a repo-private .python folder, just download all the Python packages again for another GitHub repo (so they can have different versions of numpy etc.)

You have downloaded a fresh Miniconda3 binary just for this GitHub repo. If you want to activate the repo using this Miniconda3 binary, go:

source .python/ACTIVATE_BASH # if you use bash
source .python/ACTIVATE_FISH # for those awesome folks knowing how to fish!

Forgot your Python package version? Just run python .python/align_pip.py and current package versions will be written into the pip.txt.

GO NUTS!! :)

More features

You can also do

./python-deployer/BUILD --verbose

This will get you a lot of information about what's going on during the installation of the Python binary and packages.

Or:

./python-deployer/RUN --verbose