Gitlab2Graph

A Pipeline processor to extract data from Gitlab and transform it to a graph representation.

Setup

Install

Gitlab2Graph requires Python >= 3.6.5. We recommend a virtual environment to work with it.

# Create virtual environment
python3.6 -m venv env

# Activate environment
source env/bin/activate

# Install dependencies
pip install -U pip
pip install -r requirements.txt

Obtain private access token for Gitlab

Go to https://YOUR-GITLAB/profile/personal_access_tokens and claim a personal access token. The necessary scopes are api and read_user.

Configure project

Put the obtained token into a new configuration file e.g. named project1.ini placed under the configurations folder. An example file on how to configure a project can be found here.

Add the connection parameters for the Gitlab and the Neo4j database, you like to use. The project_id defines the Gitlab project the pipelines shall be executed for.

[GITLAB]
token = deadc0ffee

[NEO4J]
hostname = localhost
;bolt or http
protocol = bolt
port = 7687
db = db/data/
user = neo4j
password = neo4jneo4j

[GENERAL]
project_id = 23

[ProjectPipeline]
my_extra_parameter = 42

Usage

gitlab2graph.py [-h] configuration [configuration ...]

    Executes a pipelines for one or more Gitlab projects defined by
    one or more configuration files.

    positional arguments:
        configuration  Configuration file(s) in INI format

Note: Configuration files are expected to be in the configurations directory

Examples

# With activated environment
python gitlab2graph.py project1.ini

or

# With activated environment
python gitlab2graph.py project1.ini project2.ini

How to cite?

If you use this work in a scientific publication, please cite the specific version that you have used as follows:

Stoffers, Martin. (2019, October 13). Gitlab2Graph (Version 1.0.0). Zenodo. http://doi.org/10.5281/zenodo.3469386

You can find information about the release number and the publication date in the CHANGELOG.

License

The script code and the accompanying material is licensed under the terms of the MIT License.