pg4n

PostgreSQL for novices

📄 Documentation

This README is meant for developers of the project, and not for end users. For end users, please see the documentation linked above.

• PostgreSQL for novices
  • Notes for developers
    • Poetry
      • Versioning
    • Imports
    • Running tests
      • Using docker
    • Building documents
    • Linters and formatters
      • Githooks

Notes for developers

Poetry

This project uses Poetry for packaging. Although one should refer to Poetry docs for a thorough introduction, here's a short summary of the intended workflow with Poetry:

• To install all dependencies and the application, type poetry install. After installation, if the Python scripts folder is in your PATH, you should be able to invoke main.main() with pg4n.
• To make VS Code use Poetry's virtual environment, type poetry env info, copy virtual environment executable path, press F1 and type Python: Select Interpreter > Enter interpreter path... > paste path and press <ENTER>.
• To add/remove a dependency, type poetry add <dep>/poetry remove <dep>.
• To execute a command from within virtual environment shell, type poetry run <cmd>.
• To enter a shell session within the Poetry virtual environment, type poetry shell.

Versioning

You can bump the version number automatically with poetry version patch, poetry version minor, etc. See poetry version -h.

See version history here.

Imports

During development, you must run the program as a module, e.g., poetry run python -m src.pg4n.main, so that the imports work.

Running tests

Having PostgreSQL running on port 5432, do poetry run pytest.

You may need to provide environment variables that match your config:

Variable	Default value	Description
PGHOST	127.0.0.1	Hostname of the PostgreSQL server.
PGPORT	5432	Port to an active PostgreSQL instance.
PGUSER	postgres	The user that will be used to manage the test database.
PGPASSWORD		Password, in case password authentication is used.
PGDBNAME	test_database	Database name.

For example, if PostgreSQL is on port 5433, just do PGPORT=5433 poetry run pytest (Bash syntax).

Using docker

To get a similar PostgreSQL instance as with GitHub Actions workflow:
docker run --rm -P -p 127.0.0.1:5432:5432 --name pg -e POSTGRES_PASSWORD=postgres -d postgres:14.5-alpine

You'll need to tell pytest the password: PGPASSWORD=postgres poetry run pytest.

Building documents

• If docs/api is not up-to-date or doesn't exist, run:
  poetry run sphinx-apidoc -f -o docs/api src/pg4n '*/test*'
• To generate the documentation:
  poetry run sphinx-build -b html docs docs/build

Note that the GitHub Pages site is only updated on pushes to main branch.

Linters and formatters

For linting, the following tools are used:

• black for formatting
• pylint for linting
• mypy for static type checking
• isort for sorting imports

To get a grade that the CI/CD pipeline would give you, you can do poetry run scripts/ci-grade.sh to run all the checks. The output is possibly long, so pipe it to a file perusal filter such as less to scroll through it and search for things of concern, e.g., summary to see scores.

Githooks

This project uses poetry-githooks to run automatic formatting on each commit. To set this up, run:

poetry run githooks setup

This needs to be re-run each time the [tool.githooks] section is modified in the pyproject.toml file.

One can skip pre-commit hooks by running git commit with the --no-verify flag.