Converting pyproject.toml Back to requirements.txt

While pyproject.toml is the modern standard, there are still valid reasons to generate a requirements.txt: Docker multi-stage builds, legacy CI systems, deployment platforms that only support requirements.txt, or simply sharing a dependency list without the full project structure.

pyproject.toml input:

[project]
name = "my-app"
version = "1.0.0"
requires-python = ">=3.10"

dependencies = [
    "fastapi>=0.104",
    "uvicorn[standard]>=0.24",
    "sqlalchemy[asyncio]>=2.0",
    "alembic>=1.13",
    "pydantic>=2.5",
    "httpx>=0.25",
]

[project.optional-dependencies]
dev = [
    "pytest>=7.4",
    "ruff>=0.1.6",
]

Generated requirements.txt:

# Main dependencies
fastapi>=0.104
uvicorn[standard]>=0.24
sqlalchemy[asyncio]>=2.0
alembic>=1.13
pydantic>=2.5
httpx>=0.25

# Optional: dev
pytest>=7.4
ruff>=0.1.6

When you need requirements.txt from pyproject.toml:

1. Docker builds -- pip install -r requirements.txt in a Dockerfile is a common pattern for caching layers
2. Heroku / PaaS -- some platforms auto-detect requirements.txt but not pyproject.toml
3. pip-tools -- pip-compile pyproject.toml generates a locked requirements.txt
4. Legacy CI -- older Jenkins or TeamCity setups may expect requirements.txt
5. Sharing -- a flat text file is easier to share in chat or email

Note: For production deployments, consider using pip-compile (from pip-tools) to generate a fully resolved, pinned requirements.txt from your pyproject.toml. This gives you the best of both worlds: flexible constraints in pyproject.toml and reproducible installs from requirements.txt.