Recap: Containers, CI, and Testing Workflows

Goals for the recap¶

• Rebuild the full toolchain from scratch, end-to-end

• Three example stacks: LaTeX, Python+pytest, C++/googletest

• CI workflows per stack: build/push + tests inside GHCR images

• Emphasize reproducibility and automation

Lab 09 folder layout¶

codes/lab09/
├── python/
│   ├── docker/
│   ├── tests/
│   └── stats.py
├── cpp/
│   ├── docker/
│   ├── include/
│   ├── src/
│   └── tests/
└── latex/
    ├── docker/
    └── main.tex

• Each subproject is self-contained and buildable with Docker.

Pattern to repeat across stacks¶

1. Minimal code + tests

2. Docker image that can run the tests

3. GitHub Actions workflow to build/push the image

4. GitHub Actions workflow to run tests/compile in container.image

Python stack¶

• stats.py exposes a small numerical helper

• tests/ uses pytest to validate mean/std outputs

• Docker image installs numpy + pytest and runs tests

docker build -t ghcr.io/<owner>/<repo>:latest-python -f docker/Dockerfile .
docker push ghcr.io/<owner>/<repo>:latest-python

C++ stack¶

• CMakeLists.txt fetches googletest and builds tests

• src/ + include/ define a small library

• tests/ verifies mean/std behavior

cmake -S . -B build
cmake --build build
ctest --test-dir build --output-on-failure

LaTeX stack¶

• main.tex is a tiny document for CI compilation

• Docker image uses TeX Live + latexmk

• Workflow compiles the PDF and uploads it as an artifact

docker build -t ghcr.io/<owner>/<repo>:latest-latex -f docker/Dockerfile .
docker push ghcr.io/<owner>/<repo>:latest-latex

GitHub Actions: build and push images¶

• One workflow per stack (manual trigger)

• Uses docker/build-push-action with GHCR login

• Tags images consistently (latest-python, latest-cpp, latest-latex)

name: Lab09 Python - Build Image
on: { workflow_dispatch: {} }

GitHub Actions: run tests¶

• Run tests inside the job container image

• C++ uses ctest, Python uses pytest, LaTeX compiles

• LaTeX workflow uploads the generated PDF

jobs:
  test:
    runs-on: ubuntu-latest
    container:
      image: ghcr.io/${{ github.repository }}:latest-python
    steps:
      - uses: actions/checkout@v4
      - run: pytest -q

Artifacts in CI¶

• actions/upload-artifact for build outputs

• Useful for PDFs, logs, or compiled binaries

• Makes CI feedback tangible and shareable

Devcontainers¶

• .devcontainer/devcontainer.json per stack

• Uses the same GHCR images as CI

• Aligns local dev with remote testing

Wrap-up checklist¶

• Can every stack build and test in a clean container?

• Are workflows minimal and readable?

• Do CI outputs (logs/artifacts) help debugging?

• Is the repo structure consistent across stacks?