Transform docker images into knowledge graphs.

Installation

For using this repo we asume that you have installed:

• docker
• syft
• python 3.10.4

Create a virtual environment and install from source: pip install -r requirements.txt

CLI Usage

To create your knowledge graph from your Docker container you must run one of these commands. (add -s to the command to save the json files in the output_path)

To create the graph from a local image: ({name} can be Image ID or image tag)

py CLI.py image -i {name} -o {output_path}

To create the graph from a file with a list of local images (the file must have this format:

"mongo:latest

python:3.10

mysql:latest")

py CLI.py image -p {file_path} -o {output_path}

To create the graph from a DockerHub image:

py CLI.py dockerhub -i {name} -o {output_path}

To create the graph from a file with a list of DockerHub images (the file must have this format:

"mongo:latest

python:3.10

mysql:latest")

py CLI.py dockerhub -p {file_path} -o {output_path}

To create the graph from a DockerFile:

py CLI.py dockerhub -p {dockerhub_path} -o {output_path}

How it works

We inspect your image to get enviroment properties. Then we use syft to obtain the image dependencies. After that we have to transform a bit the output.

Once we have these data sources, we have to create a knowledge graph. For this purpose we will use morphkgc

Ontology Documentation

Available here: https://osoc-es.github.io/c2t/onto_documentation/index-en.html

Landing Page

Available here: https://osoc-es.github.io/c2t/website

Demo Page

Available here: https://c2t.linkeddata.es/

Assuming you have the sparql endpoint up, you can launch the demo with docker run -it -p 3000:3000 c2t_demo. The Dockerfile for creating the image is under the web-demo folder

Ontology diagram