- the function queries the database and aggregates the ad-hoc orders
by pixel and time steps into a demand time series
- implement "heavy" integration tests for `aggregate_orders()`
- make `pandas` a package dependency
- streamline the `Config`
- reorganize `urban_meal_delivery.console` into a sub-package
- move `tests.db.conftest` fixtures into `tests.conftest`
=> some integration tests regarding CLI scripts need a database
- add `urban_meal_delivery.console.decorators.db_revision` decorator
to ensure the database is at a certain state before a CLI script runs
- refactor the `urban_meal_delivery.db.grids.Grid.gridify()` constructor:
- bug fix: even empty `Pixel`s end up in the database temporarily
=> create `Pixel` objects only if an `Address` is to be assigned
to it
- streamline code and docstring
- add further test cases
As a lot of the integration tests populate the database with test data,
it is deemed safer to run the tests in random order to uncover potential
dependencies between distinct test cases.
Because of how the `db_session` fixture is designed, this should already
be taken care of.
- dependencies used to run the Jupyter Lab environment that are not
required by the `urban-meal-delivery` package itself are put into
an installation extra called "research"
- this allows to NOT install the requirements, for example, when
testing the package in an isolated environment
- use Alembic to migrate the PostgreSQL database
+ create initial migration script to set up the database,
as an alternative to db.Base.metadata.create_all()
+ integrate Alembic into the test suite; the db_engine fixture
now has two modes:
* create the latest version of tables all at once
* invoke `alembic upgrade head`
=> the "e2e" tests are all run twice, once in each mode; this
ensures that the migration scripts re-create the same database
schema as db.Base.metadata.create_all() would
* in both modes, a temporary PostgreSQL schema is used to create the
tables in
=> could now run "e2e" tests against production database and still
have isolation
- make the configuration module public (to be used by Alembic)
- adjust linting rules for Alembic
- use SQLAlchemy (and PostgreSQL) to model the ORM layer
- add the following models:
+ Address => modelling all kinds of addresses
+ City => model the three target cities
+ Courier => model the UDP's couriers
+ Customer => model the UDP's customers
+ Order => model the orders received by the UDP
+ Restaurant => model the restaurants active on the UDP
- so far, the emphasis lies on expression the Foreign Key
and Check Constraints that are used to validate the assumptions
inherent to the cleanded data
- provide database-independent unit tests with 100% coverage
- provide additional integration tests ("e2e") that commit data to
a PostgreSQL instance to validate that the constraints work
- adapt linting rules a bit
- add the following file:
+ src/urban_meal_delivery/_config.py
- a config module is created holding two sets of configurations:
+ production => against the real database
+ testing => against a database with test data
- the module is "protected" (i.e., underscore) and imported at the
top level via a proxy-like object `config` that detects in which of
the two environments the package is being run
- use sphinx to document the developed package
- create a nox session "docs" that builds the docs
- include a skeleton in the docs/ folder
+ how to install the package
+ how to use nox
+ license
- add the following file:
+ src/urban_meal_delivery/console.py => click-based CLI tools
+ tests/test_console.py => tests for the module above
- add a CLI entry point `umd`:
+ implement the --version / -V option
to show the installed package's version
+ rework to package's top-level:
* add a __pkg_name__ variable to parameterize the package name
+ add unit and integration tests
- fix that pylint cannot know the proper order of imports in the
isolated nox session
- add pre-commit and pre-merge hooks:
+ run `poetry run nox -s pre-commit` on staged *.py files
+ run common pre-commit hooks for validations that could not be
achieved with tools in the develop environment so easily
+ run `poetry run nox -s pre-merge` before merges and pushes
- implement the "pre-commit" and "pre-merge" sessions in nox
+ include a little hack to deal with the positional arguments
passed by the pre-commit framework
- provide more documentation on the nox sessions
- use pytest as the base, measure coverage with pytest-cov
+ configure coverage to include branches and specify source locations
+ configure pytest to enforce explicit markers
- add a package for the test suite under tests/
- add a `__version__` identifier at the package's root
+ it is dynamically assigned the version of the installed package
+ the version is PEP440 compliant and follows a strict subset of
semantic versioning: x.y.z[.devN] where x, y, z, and N are all
non-negative integers
+ add module with tests for the __version__
- add a nox session "test" that runs the test suite
- use flake8 to lint pytest for consistent style
- use flake8 as the main and pylint as the auxiliary linter
- install flake8 with the following plug-ins:
+ flake8-annotations => enforce type annotations for functions/classes
+ flake8-black => ensure black would not make any changes
+ flake8-expression-complexity
+ wemake-python-styleguide, which packages the following:
* darglint * flake8-bandit * flake8-broken-line
* flake8-bugbear * flake8-commas * flake8-comprehensions
* flake8-debugger * flake8-docstrings * flake8-eradicate
* flake8-isort * flake8-rst-docstrings * flake8-string-format
* flake8-quotes * pep8-naming
- configure flake8 & friends in a rather explicit and strict way
- isort needed to be downgraded to ^4.3.21 due to a conflict with
pylint and wemake-python-styleguide:
+ provide TODO's to remove the parts that "fix" isort
- use mypy for static type checking
- add a nox session "lint" that runs flake8, mypy, and pylint
- lint all source files
- (auto-)format code with:
+ autoflake => * remove unused imports and variables
* remove duplicate dict keys
* expand star imports
+ black => enforce an uncompromising code style
+ isort => enforce a consistent import style
(complying with Google's Python Style Guide)
- implement the nox session "format" that runs all these tools
- add the following file:
+ setup.cfg => holds configurations for the develop tools
- set up skeletons for all planned nox sessions
- provide a base configuration for all nox sessions
- create two utility functions:
+ _begin() => run commands common to all sessions
+ _install_packages() => install dependencies in nox sessions
respecting the versions pinned in poetry.lock
- add the following file:
+ noxfile.py
- add the following files:
+ .gitignore => ignore poetry's and pyenv's artifacts
+ LICENSE.txt => MIT license as the project is scientific research
+ README.md => rough description of the project
+ poetry.lock
+ pyproject.toml
+ src/urban_meal_delivery/__init__.py => source code package
- use a "src" layout structure:
+ ensure that pytest runs the tests against a packaged and pip
installed version of the source code, not the *.py files in the
project directory
+ more info: https://hynek.me/articles/testing-packaging/
