A scraper I am creating with intention to create a dataset of rental and property values by region. With impunity on the lower mainland.
I am trying to have the tests be independent of db state. The best way to ensure I don't have tests that accidentally use my local db that is in use when debugging, I set the project name to clscraper_test.
This basically creates all different containers (and maybe images) from the standard project that docker creates (based off of the folder name that docker-compose.yml resides in).
So I have a helper script to run tests bin/runtests which uses the project name when calling docker-compose. So if I need to clean out the volumes I need to make sure to run docker-compose -p clscraper_test.
I am storing alembic outside of the clscraper package, this makes sense to me to disconnect db state management from the scraper domain code, however by doing so if I try to run the alembic script I end up creating __pycache__ everywhere. In order to avoid this I am using a helper script bin/alembic that runs python -B. The -B prevents cache file creation.
By creating a directory with all the python liraries inside it inside my repo folder, I can use goto definition with vscode. This is helpful because I often end up reading the source code of libraries used to understand library processes.