Comparing Python Project Build and Dependency Management Tools: CookieCutter, PyScaffold, PyBuilder, and Poetry

Classic Python Project Structure with CookieCutter

<code>$ pip install cookiecutter
$ cookiecutter gh:audreyr/cookiecutter-pypackage   # answer prompts to generate a project
... 
project_name [Python Boilerplate]: sample
...</code>

CookieCutter creates a conventional Python package layout, including AUTHORS.rst , README.rst , setup.py , a src package, tests , and documentation under docs . The generated Makefile provides common targets such as make test , make coverage , and make docs .

Typical make commands include clean , clean-build , lint , test , coverage , docs , dist , and install . Required auxiliary packages (e.g., tox , wheel , coverage , sphinx , flake8 ) are installed via pip and invoked through the Makefile.

Creating a Project with PyScaffold

PyScaffold is a scaffolding tool that places source files in a src directory instead of the top‑level package folder.

<code>$ pip install pyscaffold
$ putup sample</code>

The resulting layout mirrors CookieCutter's but moves the package under src . Building and testing are driven by tox , which creates isolated virtual environments and runs the defined tasks.

<code>$ tox -av
default environments:
  default -> Invoke pytest to run automated tests
additional environments:
  build   -> Build the package in isolation (PEP517)
  clean   -> Remove old distribution files
  docs    -> Build Sphinx documentation
  publish -> Publish the package to PyPI (default uses TestPyPI)</code>

Project Generation with PyBuilder

PyBuilder follows a Maven‑like approach, generating a directory structure similar to Java projects.

<code>$ pip install pybuilder
$ mkdir sample && cd sample   # create project directory manually
$ pyb --start-project        # answer prompts to create files</code>

The generated tree includes build.py , src , docs , and tests . Build tasks are defined as plugins in build.py and can be listed with pyb -t .

<code>$ pyb -t sample
Tasks found for project "sample":
  analyze, clean, compile_sources, coverage, install, package, prepare,
  publish, run_integration_tests, run_unit_tests, upload, verify</code>

Dependencies are declared inside build.py using project.depends_on() and project.build_depends_on() . The --no-venvs flag can skip virtual‑environment creation.

Managing Projects with Poetry

Poetry is a modern, opinionated tool that centralises configuration in pyproject.toml and provides robust dependency resolution.

<code>$ pip install poetry
$ poetry new sample</code>

The scaffold is minimal, containing README.rst , pyproject.toml , the package directory, and a tests folder. Adding dependencies is as simple as poetry add boto3 , and the lock file ensures reproducible builds.

<code># Build distribution files
poetry build
# Open a shell in the virtual environment
poetry shell
# Run tests with pytest
poetry run pytest
# Export requirements.txt
poetry export --without-hashes --output requirements.txt</code>

Custom scripts can be defined in [tool.poetry.scripts] and executed via poetry run my-script . Unlike the other tools, Poetry does not generate documentation or linting configurations automatically; those must be added manually.

Overall, the complexity of the generated project structure decreases from CookieCutter → PyScaffold → PyBuilder → Poetry, while the learning curve follows the same order.