Origami-Cloudless-AI / TinyMLaaS-2023-winter

Run Hello World of TensorFlow Lite for micro automatically in Docker
https://Origami-TinyML.github.io/tflm_hello_world
Apache License 2.0
1 stars 2 forks source link

Unit tests: code coverage ratio > 70% #113

Open ArttuLe opened 1 year ago

ArttuLe commented 1 year ago

One of the KPIs while development to ensure code quality

WoW

  1. A developer should submit a PR for UT along with its feature implementation (FT).
  2. A developer could submit a PR for failing UTs first, alternatively.
  3. A developer would grow UT & FT incrementally over PRs.
  4. A developer should implement as much code as possible in a jupyter notebook as a python package so that he could keep the streamlit UI code as little as possible, referring to those package APIs.
  5. A python package code could be tested with unit tests, UI coulde could be covered with AT mostly.

If this WoW prevents to pass CI / CD, we could temporarily disable to fail a CI / CD but just showing failing messages as a place holder. AT being accepted first is important since AT is an acceptance criteria, where we could avoid unnecessary misunderstandings.

Acceptance criteria

lifeofborna commented 1 year ago

I did some robot framework tests for the UI me and Matias worked on. can be found at

https://github.com/OhtuProjTinyML/tflm_hello_world_staging/tree/robot_test_streamlit/pages/tests

doyu commented 1 year ago

Looks enough descriptive for the spec (i.e. Acceptrance criteria), promising :)

doyu commented 1 year ago

And your staging seems to be set public visibility, though.......

ArttuLe commented 1 year ago

Staging is set to private now