Tsundere, but useful bot (it's like wasabi!)
Nanachan only supports python version 3.12.
You can use pyenv to manage multiple version of python in the same environment.
Nanachan project uses poetry v1.x as virtual environment manager. You can install it either via your package manager (be sure to install version 1), or using the install script (see the docs).
In case of using the install script, be sure to follow the instructions displayed during installation to
update your PATH.
If you had a previous version of poetry, just update it by running poetry self:update or poetry self update.
More info in the installation instruction.
To setup your development environment, just run poetry install.
Follow this getting started guide to setup your Visual Studio Code and Docker environment. (Note: the Remote - Containers extension is sufficient, no need to download the full Remote Development extension pack.)
Do not forget to correctly setup your git credentials to be able to perform git operations directly inside the container.
Just open the cloned repository in VS Code: a notification will appear in the bottom right corner, select the Reopen in Container option. If it is not the case, click on the green button on the left of the Status Bar and select Remote-Containers: Reopen in Container.
The provided devcontainer.json and Dockerfile in .devcontainer will tell VS Code how to build the Docker image and prepare the container.
A few minutes later your container will be ready and Dev Container: nanachan will appear in the left of the Status Bar. You can now start coding inside.
To simplify these operations in VS Code, launch.json and tasks.json are provided in the .vscode folder.
So just go to in the Run section of the Side Bar or press F5 to start debugging nanachan.
To check your changes before committing you can use the provided pre-commit hook:
git config core.hookspath hooksor
cp hooks/pre-commit .git/hooksBy default it will run ruff check on the codebase and pyright if the
HOOKS_PYRIGHT_CHECK environment variable is set (because pyright is very
slow).
In any case the CI will run the same commands and report the errors just the same.