I recently opened an issue for the open source Python grafeas client.
And was planning on working on it myself. But before I throw myself into this...
Is this project upstream of the GCP platform? Is there any effort made to keep things compatible at the API layer between the two or have they forked?
There appears to be an open-sourced, well-maintained Python client for the GCP offering at
https://github.com/googleapis/python-grafeas, created from the protocol buffers by Google tooling instead of the swagger/OpenAPI files.
If the goal is to keep both this and the GCP offering in sync, is working on a separate client Python SDK worth doing, since Google's should work with both?
I also notice that the googleapi one is in PyPI, and what you get if you naively pip install from there.
Assuming that there is a reason to maintain this separate client, a few further questions come to mind:
1. Is there a desire to package this client (under a different name) in PyPI as well? Note that a recent change to make the Python client package version synonymous with the api version is not compatible with
https://www.python.org/dev/peps/pep-0440/2. Should I keep the alpha version of the api in the package, such that it's not the default, but still reachable, or should I just eliminate it?
3. Swagger generates test stubs, but they've not been filled in. Is this because nobody has had time to work on it, or do the project owners feel that it's not needed?
4. If tests are wanted, preference for (obvs highly mocked) unit tests, or e2e tests with the Dockerized Grafeas server first?
5. Is there a general road map doc for the server and clients? Is this generally a do-ocracy or should I seek guidance somewhere before haring off and doing what I think is right?