I finally managed to get some time to wrap up the code coverage for the library - sorry for the long delay to get this done.
I'd also like to suggest a bit of an aggressive approach: we could turn on a pytest flag that will fail the build if coverage goes below 100%, so that anyone contributing to the project has to write automated tests to prove that the contribution was tested. This doesn't guarantee that the library becomes bug-proof, but it does help reducing chances of bugs creeping in. Let me know what you think, and, if you agree, I could even add this in this PR (or the next one if you prefer).
Next on: with the full project covered, I want to start doing some bigger changes, like starting to support different JSON codecs (ujson/orjson), make the API more pythonic (there's a few places where we don't follow PEP-8 for code style) etc.
Hey everyone!
I finally managed to get some time to wrap up the code coverage for the library - sorry for the long delay to get this done.
I'd also like to suggest a bit of an aggressive approach: we could turn on a pytest flag that will fail the build if coverage goes below 100%, so that anyone contributing to the project has to write automated tests to prove that the contribution was tested. This doesn't guarantee that the library becomes bug-proof, but it does help reducing chances of bugs creeping in. Let me know what you think, and, if you agree, I could even add this in this PR (or the next one if you prefer).
Next on: with the full project covered, I want to start doing some bigger changes, like starting to support different JSON codecs (ujson/orjson), make the API more pythonic (there's a few places where we don't follow PEP-8 for code style) etc.