How to convert pyproject toml to setup py or setup.cfg?

To convert pyproject toml to setup py: Parse the [project] table defined in PEP 621 or [tool.poetry] in pyproject toml. Extract metadata like name, version, authors, and dependencies into setuptools.setup() arguments such as install_requires, package_dir, and packages=find_packages().

Should modern Python packages use setup py or pyproject toml?

Choosing between setup py or pyproject toml: The Python packaging ecosystem (PyPA) strongly recommends pyproject toml as the declarative standard (PEP 517, 518, 621). While legacy projects relied on setup.py or setup cfg pyproject toml configurations, modern build backends (Hatch, Flit, Poetry, setuptools>=61) utilize pyproject.toml exclusively for declarative package definition.

How to declare packages pyproject toml and pyproject toml dependencies?

To declare pyproject toml dependencies, specify dependencies = ['requests>=2.28.0', 'fastapi'] inside the [project] section according to the pyproject toml schema. For packages pyproject toml discovery, modern setuptools automatically finds your modules or can be configured via [tool.setuptools.packages.find].

Convert pyproject.toml to setup.py Converter

Python Packaging Evolution: setup py or pyproject toml?

For over two decades, the Python community used setup.py for package configuration and distribution. However, running arbitrary Python code during the build process caused serious dependency resolution and security bottlenecks. In response, PEP 518 and PEP 621 introduced pyproject toml as the universal, static declarative standard.

When deciding whether to use setup py or pyproject toml, PyPA recommends adopting pyproject.toml for all new projects. However, legacy deployment pipelines, enterprise packaging mirrors, or air-gapped systems still require standard setup.py files. This tool bridges both worlds by automating the process to convert pyproject toml to setup py.

Declaring Dependencies in pyproject toml

Handling pyproject toml dependencies under the PEP 621 specification uses a standardized array of PEP 508 requirement strings:

[project]
dependencies = [
  "requests >= 2.28.0",
  "pydantic >= 2.0.0",
  "click >= 8.1.0"
]

Our parser automatically translates these into the legacy install_requires list in setup.py or install_requires key in setup.cfg.

Packages Discovery & setup cfg pyproject toml Mapping

When configuring packages pyproject toml, modern Setuptools 61+ supports automatic package discovery. When transitioning between setup cfg pyproject toml configurations, Setuptools finds all Python modules under the src/ directory or project root.

This tool also validates compliance with the official pyproject toml schema, ensuring your metadata fields (like description, authors, license, and scripts) align seamlessly.

Real-World pyproject toml Example

Here is a comprehensive pyproject toml example showcasing the PEP 621 declarative specification:

[build-system]
requires = ["setuptools>=61.0"]
build-backend = "setuptools.build_meta"

[project]
name = "my-awesome-tool"
version = "1.0.0"
description = "A high-performance Python analytics utility."
readme = "README.md"
requires-python = ">=3.9"
authors = [
  { name = "Developer", email = "dev@example.com" }
]
dependencies = [
  "httpx>=0.24.0",
  "rich>=13.0.0"
]

[project.scripts]
mytool = "my_awesome_tool.cli:main"

Frequently Asked Questions

Can I still install a package if it only has a pyproject.toml and no setup.py? ▼

Yes! Modern pip (version 19.0+) fully supports PEP 517 build systems. Running `pip install .` or `pip install -e .` will automatically build the package using the backend defined in your pyproject.toml [build-system].

What build backends are supported by pyproject.toml? ▼

Common build backends include Setuptools (setuptools.build_meta), Poetry (poetry.core.masonry.api), Hatch (hatchling.build), Flit (flit_core.buildapi), and Meson-Python.

How does this converter handle packages pyproject toml and find_packages? ▼

In setup.py, it generates packages=find_packages(), which inspects directories for __init__.py files, replicating modern declarative package discovery.