depsight 0.5.0

A modular dependency analysis framework

Depsight - Dependency Analysis Framework

Usage

Installation

Via pip

pip install depsight

Via uv

uv tool install depsight

Via Docker

docker pull <your-dockerhub-user>/depsight

Getting Started

Read the docs.

Development

System Requirements

• IDE with DevContainer support — either of the following:
  • Visual Studio Code with the Dev Containers extension
  • JetBrains Gateway (supports Dev Containers via the Remote Development plugin)
• Container manager — any one of the following:
  • Docker Desktop (macOS, Windows, Linux) (RECOMMENDED)
  • Docker Engine (Linux)
  • Podman with the Podman Desktop or CLI

Setup Locally

• Open Visual Studio Code at the project root directory
• When prompted, click Reopen in Container (or use Command Palette: Dev Containers: Reopen in Container)
• Wait for the containers to build and start
• Once ready, you'll have a fully configured development environment with all dependencies installed
• Open a terminal inside the DevContainer and run depsight --help

Build Container Image

• Open a terminal inside the DevContainer (Docker-in-Docker is available)

• Build the image:

  docker build -t depsight .
  

• Optionally override the Python or uv version via build args:

  docker build -t depsight --build-arg PYTHON_VERSION=3.13 --build-arg UV_VERSION=0.10.9 .
  

• Verify the image works:

  docker run --rm depsight --help
  

Run Tests

• Open a terminal inside the DevContainer

• Activate the virtual environment:

  source .venv/bin/activate
  

• Run all tests:

  pytest tests/ -v
  

Lint & Type Check

• Run the Ruff linter:

  ruff check src/ tests/
  

• Run the Mypy type checker:

  mypy src/
  

Build Docs

• Open a terminal inside the DevContainer

• Activate the virtual environment:

  source .venv/bin/activate
  

• Serve the documentation locally with live reload:

  mkdocs serve
  

  The site will be available at http://127.0.0.1:8000.

• Alternatively, build the static site as a site/ directory:

  mkdocs build
  

Build

Pre-release

Use this to trigger a build manually at any time without creating a release.

• Navigate to your repository on GitHub and click the Actions tab
• Select the Manual Dispatch workflow from the left sidebar
• Click Run workflow
• Enter the depsight_version (must match the version field in pyproject.toml, e.g. 0.1.0)
• Select the desired python_version (3.12 or 3.13, defaults to 3.12)
• Optionally set uv_version (defaults to 0.10.9)
• Optionally check Upload wheel as artifact to download the built .whl after the run
• Click Run workflow
• Once the run completes, the wheel is available under the run's Artifacts section (if enabled)

Release

Use this to publish an official versioned build. The wheel is automatically attached to the GitHub Release.

• Bump the version field in pyproject.toml to the desired version (e.g. 1.2.3)
• Commit and push the change to main
• Navigate to your repository on GitHub and click Releases → Draft a new release
• Create a new tag matching the version in pyproject.toml exactly (e.g. 1.2.3)
• Click Publish release
• The On Release workflow triggers automatically, runs the full CI pipeline, and uploads the .whl to the release assets