Code to Docker

Processor to convert source code into a Artefacts-ready Dockerfile, following guidance from a configuration file.

Setup and development

• Python 3.11

Setting Up a Dev Environment

It is best practice to setup a virtual environment. One way to do this is by using virtualenv which can be installed with pip install virtualenv.

Once done, setup your environment for this repo by running the following commands

python3.11 -m venv env
source env/bin/activate
pip install -r requirements.txt
# and for dev environments
pip install -r requirements-dev.txt

You will note that running pip list will only have the packages required for this repo (and not everything you might have installed on your computer).

(The env folder created is already noted in .gitignore file so wont be pushed up to github)

Publication

The package is published on a private repository hosted on AWS. Publication is automated:

./scripts/publish [AWS profile name]

If your account is default, no parameter needed. The script relies on Twine (in the development requirements) to package and deploy to AWS CodeArtifact.

Usage

Generating the DockerFile

The repository contains fake but valid configuration files in examples/. A simple trial is to execute:

PYTHONPATH=src python src/c2d/cli.py . -c examples/basic/config.yaml -o examples/basic/

(or whichever file you wish)

The command results in a Dockerfile generated under examples/basic, ready to build with Docker.

You can also pass the additional flag --as_text to simply output the dockerfile as a string rather than creating an actual file.

Building the Dockerfile

docker build -t <tagName> .

Running the Dockerfile (using Artefacts)

• Create your project on artefacts.com, generate an api key and keep a note of it.
• The project name should match the project in the warp.yaml file.
• Place the dockerfile and artefacts.yaml in the root of your project, then run

docker run --env ARTEFACTS_KEY=<generatedApiKey>  --env ARTEFACTS_JOB_NAME=<jobNameSetInConfigFile> <tagName>

(For the dev environment you will want to add this additional env --env ARTEFACTS_API_URL=https://ui.artefacts.com/api)

The auto-generated dockerfiles by c2d are designed to run headless, even if the project requires a display. Examples can be seen when running turtlesim examples. If you do however want to use a display for your tests, you can add the following flags to your docker run command --network=host --env DISPLAY=$DISPLAY