Repository navigation
Getting Started
- Python 3.12
-
Poetry (for
tp2andtp4, which manage dependencies viapyproject.toml) - Docker & Docker Compose (for
tp5) - DVC and MLflow (for
tp7) - A Jenkins instance and a SonarQube server if you want to exercise
tp3/tp4beyond their pytest suites — these are reference/integration pieces, not required to run the code itself
There is no single project-wide requirements.txt: each lab pins or documents its own dependencies (pyproject.toml for tp2/tp4, requirements.txt for tp5, plain pip install commands for tp7).
# tp2 - pytest suite (unit / integration / functional)
cd tp2 && poetry install && poetry run pytest -v --cov=src
# tp4 - SonarQube-scanned calculator
cd tp4/test-sonarqube && poetry install && poetry run pytest -v --cov=src
# tp5 - Flask + Docker web app
cd tp5 && pip install -r requirements.txt && python app.py
# or: docker-compose up --build
# tp7 - DVC + MLflow pipeline
cd tp7/tp-dvc-mlflow && pip install pandas scikit-learn mlflow joblib pyyaml && dvc reprotp1_git_intro is a plain Git exercise (no code to run). tp3/Jenkinsfile documents a pipeline meant for a self-hosted Jenkins runner rather than something you execute locally. tp6 is a written Kubernetes lab specification (minikube, kubectl, namespaces and pods) — reference material, not an implemented lab.
.github/workflows/ci.yml runs on every push to main and does three things:
- Executes the real
tp2pytest suite (unit, integration, functional). - Executes the real
tp4pytest suite for the SonarQube-scannedCalculatormodule. - Runs a Python syntax/import sanity check across the remaining labs (
tp1,tp5,tp7) that don't have their own automated test suite in CI.
CI does not spin up Jenkins, SonarQube, Docker, or a Kubernetes cluster — those require live infrastructure this repository doesn't ship. tp3/Jenkinsfile and the SonarQube scan config in tp4 document how those pieces would run against a real server.