Gimie (GIt Meta Information Extractor) is a python library and command line tool to extract structured metadata from git repositories.

Context

Scientific code repositories contain valuable metadata which can be used to enrich existing catalogues, platforms or databases. This tool aims to easily extract structured metadata from a generic git repositories. It can extract extract metadata from the Git provider (GitHub or GitLab) or from the git index itself.

---

Using Gimie: easy peasy, it's a 3 step process.

1: Installation

To install the stable version on PyPI:

pip install gimie

To install the dev version from github:

pip install git+https://github.com/sdsc-ordes/gimie.git@main#egg=gimie

Gimie is also available as a docker container hosted on the Github container registry:

docker pull ghcr.io/sdsc-ordes/gimie:latest

# The access token can be provided as an environment variable
docker run -e GITHUB_TOKEN=$GITHUB_TOKEN ghcr.io/sdsc-ordes/gimie:latest gimie data <repo>

2 : Set your credentials

In order to access the github api, you need to provide a github token with the read:org scope.

A. Create access tokens

New to access tokens? Or don't know how to get your Github / Gitlab token ?

Have no fear, see here for Github tokens and here for Gitlab tokens. (Note: tokens are as precious as passwords! Treat them as such.)

B. Set your access tokens via the Terminal

Gimie will use your access tokens to gather information for you. If you want info about a Github repo, Gimie needs your Github token; if you want info about a Gitlab Project then Gimie needs your Gitlab token.

Add your tokens one by one in your terminal: your Github token:

export GITHUB_TOKEN=

and/or your Gitlab token:

export GITLAB_TOKEN=

3: GIMIE info ! Run Gimie

As a command line tool

gimie data https://github.com/numpy/numpy

(want a Gitlab project instead? Just replace the URL in the command line)

As a python library

from gimie.project import Project
proj = Project("https://github.com/numpy/numpy")

# To retrieve the rdflib.Graph object
g = proj.extract()

# To retrieve the serialized graph
g_in_ttl = g.serialize(format='ttl')
print(g_in_ttl)

For more advanced use see the documentation.

Outputs

The default output is Turtle, a textual syntax for RDF data model. We follow the schema recommended by codemeta. Supported formats are turtle, json-ld and n-triples (by specifying the --format argument in your call i.e. gimie data https://github.com/numpy/numpy --format 'ttl').

With no specifications, Gimie will print results in the terminal. Want to save Gimie output to a file? Add your file path to the end : gimie data https://github.com/numpy/numpy > path_to_output/gimie_output.ttl

---

Contributing

All contributions are welcome. New functions and classes should have associated tests and docstrings following the numpy style guide.

The code formatting standard we use is black, with --line-length=79 to follow PEP8 recommendations. We use pytest as our testing framework. This project uses pyproject.toml to define package information, requirements and tooling configuration.

For development:

activate a conda or virtual environment with Python 3.8 or higher

git clone https://github.com/sdsc-ordes/gimie && cd gimie
make install

run tests:

make test

run checks:

make check

for an easier use Github/Gitlab APIs, place your access tokens in the .env file: (and don't worry, the .gitignore will ignore them when you push to GitHub)

cp .env.dist .env

build documentation:

make doc

Releases and Publishing on Pypi

Releases are done via github release

• a release will trigger a github workflow to publish the package on Pypi
• Make sure to update to a new version in pyproject.toml before making the release
• It is possible to test the publishing on Pypi.test by running a manual workflow: go to github actions and run the Workflow: 'Publish on Pypi Test'