I recently came across a very helpful Python package on Github, that I wanted to use for the project I’m currently working on. The package was also available on the Python Package Index PyPI, allowing me to integrate it into my project using pip
. However, I needed to change the package’s code. One way to use the package with the changed code is to integrate it as a git
submodule into my current project and not use pip
. However, I wanted to keep the convenience of pip
such that my co-workers don’t have to mess with git
submodules and can just run:
1
python -m pip install -r requirements.txt
And pip allows you to install packages from remote git repositories. If you want to install the package foobar
through pip
you would normally execute the following command:
1
python -m pip install foobar
This command installs the most recent version available for your Python version. However, you can also install the foobar
package from its GitHub repository and even specify the commit you would like to install by running the following command:
1
python -m pip install git+https://github.com/foobar/foobar.git@91f9ad741b03d8587a5b52612c805c658e4f1d84
This command uses https
as a protocol to connect to GitHub. But you can also use the ssh
protocol, which might be necessary if you got a private GitLab instance, for example. The next command shows you how to install a Python package through pip using ssh:
1
python -m pip install git+ssh://[email protected]/foobar/foobar.git@91f9ad741b03d8587a5b52612c805c658e4f1d84
To successfully run this command, you need to set up an ssh-key before authenticating yourself. Finally, to add such a git repository to your requirements.txt
you insert the ssh
or https
link into it like this:
1
2
3
matplotlib==3.4.3
numpy>=1.21.2
git+ssh://[email protected]/foobar/foobar.git@91f9ad741b03d8587a5b52612c805c658e4f1d84
And this is how you can use pip to install packages directly from git repositories.
If you want to learn how to set up and run multiple Python versions on your system, check out this article: How to Install Multiple Python Versions on your Computer and use them with VSCode together with the videos on Youtube.