Daemon for Echolalia

A Python-based daemon that periodically checks and schedules tasks from Echolalia

Development

We use poetry for package and dependency management. You can install poetry system-wide with pipx:

pipx install poetry

Conda does not always play well with other package managers, so conda may need to be deactivated (not only the conda environment) during development.

To lock and install dependencies run

poetry install
poetry lock

To enter the Python virtual environment run

eval $(poetry env activate)

Note this doesn't spawn a subshell, so exit will close your shell entirely. You could use the poetry shell plugin for more control. Finally, it is important to enter the virtual environment before installing dependencies.

To build the project just run

poetry build

And the source and binary distributions will appear in the dist/ folder.

Running the Daemon

Create a virtual environment somewhere, or use your Conda environment, and just install the project

pip install git@github.com:LAAC-LSCP/analysis-daemon.git
run-daemon

Note you need to have an ssh key associated with LAAC.

Testing

Run pytest as usual. In the root of the project run

pytest

To run the tests in various fresh virtual environments you can use tox. You can install tox via pipx pipx install tox.

And tox tests in python 3.13 can be run with, say

tox -e py313 -- --randomly-seed=1234

The seed is optional, and will shuffle the order of the tests and is good practice. To run the full suite with linting, formatting and type-checking, you will need to install black, isort, autoflake, flake8 and mypy with pipx, and run tox.

Working with the Daemon and source maps

Within your environment, Conda or otherwise, use the --editable option to install the Daemon with the source maps

pip install -e [path]

Then your dependency updates with the changes made locally, and even debugging and IntelliSense will work.

Lint and Typecheck Locally

Install black, isort, autoflake, flake8, mypy, system-wide with pipx or in your environment with pip. Go to the repository root and run black ., isort ., autoflake ., flake8 . and mypy . to lint/format/type-check.

Commits and Semantic Versioning

We bump our releases and update our changelog automatically, but this requires commits to follow the conventional commits scheme. We use a combination of release-please and commitlint.

We recommend using squash-merge for pull requests for many reasons. Rebase-merge works too, but if you're doing something like red/green development, or did not validate all your individual commits against the actions, the main branch may not be clean after a rebase-merge (in the sense that every snapshot be clean).

Migrations

The database may change over time, whether for reasons of data or schema. Either way, we depend on alembic to handle migrations. We initialise the database through a migration. We have wrapped our alembic logic in a command

run-migrations