hug-api-sqlalchemy-example
A little sandbox project to explore possibility of using hug/falcon for future API development instead of Djanog Rest Framework
Features implemented so far ...
- sqlalchemy extension/middleware (need to tidy up the way it initialise)
- user login/signup with JWT
- alembic integration with sqlalchemy and models used in API
- Unit test base class to create test database and db schema per class
Installation
If you really want to try this...good luck..
Prerequisite
PostgreSQL installed, where
- user: postgres
- password: postgres
and rights to create database
install from repo
git clone [email protected]:jimmyho/hug-api-sqlalchemy-example.git
pip install -r requirements.txt
gunicorn --reload api.app:__hug_wsgi__
# launch browser to http://127.0.0.1:8000
My Take on Hug/Falcon (vs Django Rest Framework and Flask)
The experience have been pleasant, Hug is great as a micro-framework.
Most of the pain was mainly due to the missing extension/features which DRF and Flask eco-system have.
But that's the whole point of a micro-framework!
What I like
- Python 3 annotation to validate inputs
- CLI support
- simple test client
- I know what's going on underneath
What I don't like
- Lacks extensions for common stuff, eg CORS, db, auth.
- Hard to find documentation. I had to dig into hug's source code
Will I use it for serious projects?
- YES! But only/mostly if...
- I am using non-relational backends, eg, firebase, cassandra, etc.
- authentication and authorisation is simple/custom
- rapid development of micro-services
- potential pypy'ing