A hungry etl process for humans
This is a quick mockup of an idea. Can you take a code first approach to ETL similar to how EntityFramework in .Net works?
Goals: A user should be able to add a model to the list of models and have the app spin up a new table, detect relationships and handle CRUD through:
- an API
- a website that allows searching, editing, etc
- filewatchers for datafiles with good headers
Security should be handled at the model level in code.
Forked applications should be able to facade the main application for storage to present a coherent whole.
Currently this consists of a demo application running both a website and an API on Flask.
If you are on a system that runs
Make, there's a
Makefile with all the commands to run.
make init to install dependencies.
Then you can
make test to run the tests or
make run to run the application.
If you don't have Make:
Firstly, make sure you have pipenv -
pip install pipenv.
pipenv install --dev