OUTDATED: do not use

Bootstrap env

• install pyenv first to able to use pyenv install 3.10.3
• then use pyenv shell 3.10.3 and pip install poetry
• poetry install to init your env
• poetry run pre-commit run -a to autoformat your code

Project commands [inside app/bin.py]

• poetry run test -> runs docker compose containers and then runs tests
• poetry run server -> runs docker compose and starts dev server with auto reload
• poetry run recreatemigrations -> recreates initial migration, we can use this when service not deployed
• poetry run makemigration {description} -> create migration for db
• poetry run migrate -> upgrades to heads migration

Manage packages

• use poetry add {name} to add package
• use poetry add -D {name} to add package only for development

Code styleguide

• for first level module variables use UPPER_CASE = 1 API etc.
• all new features should be placed at features module
• db tables always lower case some_table_name, values upper case, for example `CONTENT_TABLE = 'content', not 'contents'

Helpful urls

• REST https://devhints.io/rest-api

TODO:

• branch management: -> production branch (last production version with tag of latest release) -> auto commit creates a tag (not pushed into prod, only manually from github workflow -> workflow creates pull request into main rebase) -> hotfix/someshing needs to be fixed -> main branch (all stuff goes here) -> (feature or bugfix)/example-button branch from develop branch (service will create feautre-env -> all from develop + service for feature, all other services auto updated) -> could be fast forward
• add pipeline (python build with deps -> gitlab ci runner + cache)
• init git branch production

Feature development

• New features located at app.features module -> to create new feature you can use command poetry run feature <name>

• This will create new folder with stub API and wired module inside app/init.py

• Feature structure:

  • __init__.py - includes global instances that could not be initialized by container, like API of FastAPI APIRouter instance
  • models.py - defines pydantic models for endpoints, DB objects and general object validation
  • tables.py - defines DB tables with keys, relations, indexes etc.
  • services.py - defines incapsulated part of reused or isolated part of logic
  • managers.py - defines aggregated flows using services,resources into business logic
  • containers.py - defines construction of all dependencies (services, managers, tables, etc.)
  • api.py - defines endpoints with injected container dependencies

• We need coverage report, mypylint