Gitoxide: A peek into the future…

I started working on GitPython in 2009, back in the days when Python was 'my thing' and I had great plans with it. Of course, back in the days, I didn't really know what I was doing and this shows in many places. Somewhat similar to Python this happens to be 'good enough', but at the same time is deeply flawed and broken beyond repair.

By now, GitPython is widely used and I am sure there is a good reason for that, it's something to be proud of and happy about. The community is maintaining the software and is keeping it relevant for which I am absolutely grateful. For the time to come I am happy to continue maintaining GitPython, remaining hopeful that one day it won't be needed anymore.

More than 15 years after my first meeting with 'git' I am still in excited about it, and am happy to finally have the tools and probably the skills to scratch that itch of mine: implement git in a way that makes tool creation a piece of cake for most.

If you like the idea and want to learn more, please head over to gitoxide, an implementation of 'git' in Rust.

GitPython

GitPython is a python library used to interact with git repositories, high-level like git-porcelain, or low-level like git-plumbing.

It provides abstractions of git objects for easy access of repository data often backed by calling the git command-line program.

DEVELOPMENT STATUS

This project is in maintenance mode, which means that

• …there will be no feature development, unless these are contributed
• …there will be no bug fixes, unless they are relevant to the safety of users, or contributed
• …issues will be responded to with waiting times of up to a month

The project is open to contributions of all kinds, as well as new maintainers.

REQUIREMENTS

GitPython needs the git executable to be installed on the system and available in your PATH for most operations. If it is not in your PATH, you can help GitPython find it by setting the GIT_PYTHON_GIT_EXECUTABLE=<path/to/git> environment variable.

• Git (1.7.x or newer)
• Python >= 3.7

The list of dependencies are listed in ./requirements.txt and ./test-requirements.txt. The installer takes care of installing them for you.

INSTALL

GitPython and its required package dependencies can be installed in any of the following ways, all of which should typically be done in a virtual environment.

From PyPI

To obtain and install a copy from PyPI, run:

pip install GitPython

(A distribution package can also be downloaded for manual installation at the PyPI page.)

From downloaded source code

If you have downloaded the source code, run this from inside the unpacked GitPython directory:

pip install .

By cloning the source code repository

To clone the the GitHub repository from source to work on the code, you can do it like so:

git clone https://github.com/gitpython-developers/GitPython
cd GitPython
./init-tests-after-clone.sh

On Windows, ./init-tests-after-clone.sh can be run in a Git Bash shell.

If you are cloning your own fork, then replace the above git clone command with one that gives the URL of your fork. Or use this gh command (assuming you have gh and your fork is called GitPython):

gh repo clone GitPython

Having cloned the repo, create and activate your virtual environment.

Then make an editable install:

pip install -e ".[test]"

In the less common case that you do not want to install test dependencies, pip install -e . can be used instead.

With editable dependencies (not preferred, and rarely needed)

In rare cases, you may want to work on GitPython and one or both of its gitdb and smmap dependencies at the same time, with changes in your local working copy of gitdb or smmap immediatley reflected in the behavior of your local working copy of GitPython. This can be done by making editable installations of those dependencies in the same virtual environment where you install GitPython.

If you want to do that and you want the versions in GitPython's git submodules to be used, then pass -e git/ext/gitdb and/or -e git/ext/gitdb/gitdb/ext/smmap to pip install. This can be done in any order, and in separate pip install commands or the same one, so long as -e appears before each path. For example, you can install GitPython, gitdb, and smmap editably in the currently active virtual environment this way:

pip install -e ".[test]" -e git/ext/gitdb -e git/ext/gitdb/gitdb/ext/smmap

The submodules must have been cloned for that to work, but that will already be the case if you have run ./init-tests-after-clone.sh. You can use pip list to check which packages are installed editably and which are installed normally.

To reiterate, this approach should only rarely be used. For most development it is preferable to allow the gitdb and smmap dependencices to be retrieved automatically from PyPI in their latest stable packaged versions.

Limitations

Leakage of System Resources

GitPython is not suited for long-running processes (like daemons) as it tends to leak system resources. It was written in a time where destructors (as implemented in the __del__ method) still ran deterministically.

In case you still want to use it in such a context, you will want to search the codebase for __del__ implementations and call these yourself when you see fit.

Another way assure proper cleanup of resources is to factor out GitPython into a separate process which can be dropped periodically.

Windows support

See Issue #525.

RUNNING TESTS

Important: Right after cloning this repository, please be sure to have executed the ./init-tests-after-clone.sh script in the repository root. Otherwise you will encounter test failures.

Install test dependencies

Ensure testing libraries are installed. This is taken care of already if you installed with:

pip install -e ".[test]"

Otherwise, you can run:

pip install -r test-requirements.txt

Test commands

To test, run:

pytest

To lint, and apply automatic code formatting, run:

pre-commit run --all-files

• Linting without modifying code can be done with: make lint
• Auto-formatting without other lint checks can be done with: black .

To typecheck, run:

mypy -p git

CI (and tox)

The same linting, and running tests on all the different supported Python versions, will be performed:

• Upon submitting a pull request.
• On each push, if you have a fork with GitHub Actions enabled.
• Locally, if you run tox (this skips any Python versions you don't have installed).

Configuration files

Specific tools:

• Configurations for mypy, pytest, coverage.py, and black are in ./pyproject.toml.
• Configuration for flake8 is in the ./.flake8 file.

Orchestration tools:

• Configuration for pre-commit is in the ./.pre-commit-config.yaml file.
• Configuration for tox is in ./tox.ini.
• Configuration for GitHub Actions (CI) is in files inside ./.github/workflows/.

Contributions

Please have a look at the contributions file.

INFRASTRUCTURE

• User Documentation
• Questions and Answers
• Please post on Stack Overflow and use the gitpython tag
• Issue Tracker
  • Post reproducible bugs and feature requests as a new issue. Please be sure to provide the following information if posting bugs:
    • GitPython version (e.g. import git; git.__version__)
    • Python version (e.g. python --version)
    • The encountered stack-trace, if applicable
    • Enough information to allow reproducing the issue

How to make a new release

1. Update/verify the version in the VERSION file.
2. Update/verify that the doc/source/changes.rst changelog file was updated. It should include a link to the forthcoming release page: https://github.com/gitpython-developers/GitPython/releases/tag/<version>
3. Commit everything.
4. Run git tag -s <version> to tag the version in Git.
5. Optionally create and activate a virtual environment. (Then the next step can install build and twine.)
6. Run make release.
7. Go to GitHub Releases and publish a new one with the recently pushed tag. Generate the changelog.

How to verify a release (DEPRECATED)

Note that what follows is deprecated and future releases won't be signed anymore. More details about how it came to that can be found in this issue.

---

Please only use releases from pypi as you can verify the respective source tarballs.

This script shows how to verify the tarball was indeed created by the authors of this project:

curl https://files.pythonhosted.org/packages/09/bc/ae32e07e89cc25b9e5c793d19a1e5454d30a8e37d95040991160f942519e/GitPython-3.1.8-py3-none-any.whl > gitpython.whl
curl https://files.pythonhosted.org/packages/09/bc/ae32e07e89cc25b9e5c793d19a1e5454d30a8e37d95040991160f942519e/GitPython-3.1.8-py3-none-any.whl.asc >  gitpython-signature.asc
gpg --verify gitpython-signature.asc gitpython.whl

which outputs

gpg: Signature made Fr  4 Sep 10:04:50 2020 CST
gpg:                using RSA key 27C50E7F590947D7273A741E85194C08421980C9
gpg: Good signature from "Sebastian Thiel (YubiKey USB-C) <[email protected]>" [ultimate]
gpg:                 aka "Sebastian Thiel (In Rust I trust) <[email protected]>" [ultimate]

You can verify that the keyid indeed matches the release-signature key provided in this repository by looking at the keys details:

gpg --list-packets ./release-verification-key.asc

You can verify that the commit adding it was also signed by it using:

git show --show-signature  ./release-verification-key.asc

If you would like to trust it permanently, you can import and sign it:

gpg --import ./release-verification-key.asc
gpg --edit-key 4C08421980C9

> sign
> save

Projects using GitPython

• PyDriller
• Kivy Designer
• Prowl
• Python Taint
• Buster
• git-ftp
• Git-Pandas
• PyGitUp
• PyJFuzz
• Loki
• Omniwallet
• GitViper
• Git Gud

LICENSE

3-Clause BSD License, also known as the New BSD License. See the LICENSE file.