We the introduction of the ".toml" installation file we now have the version specified in two seperate locations (in the toml file and that version.py file). We should consider unifying these or at the very least checking that they are the same in the test system. @tylerjereddy pointed to the helpful example below.
@clevelam clevelam 52 minutes ago
This looks great. Thanks @tylerjereddy! Do you think I should remove my "version.py" file as a follow on to this. I was mostly using this to track if I'm trying to append existing dictionary files that might have been parsed by pervious version of the code (probably overkill). Or should I just pull the version from this toml file in the python version.py that I currently have?
Member
Author
@tylerjereddy tylerjereddy 39 minutes ago
Good question, I wasn't immediately sure, and checked what NumPy and SciPy are doing for version. It looks like both of them do a bit of work with it in their init.py files--NumPy example here: https://github.com/numpy/numpy/blob/main/numpy/__init__.py#L701
Both those projects have regression tests for correct version handling too. So, I'm a bit hesitant to suggest changing anything right away unless there is already test coverage for whatever you use versioning for.
We the introduction of the ".toml" installation file we now have the version specified in two seperate locations (in the toml file and that version.py file). We should consider unifying these or at the very least checking that they are the same in the test system. @tylerjereddy pointed to the helpful example below.
@clevelam clevelam 52 minutes ago
This looks great. Thanks @tylerjereddy! Do you think I should remove my "version.py" file as a follow on to this. I was mostly using this to track if I'm trying to append existing dictionary files that might have been parsed by pervious version of the code (probably overkill). Or should I just pull the version from this toml file in the python version.py that I currently have?
Member
Author
@tylerjereddy tylerjereddy 39 minutes ago
Good question, I wasn't immediately sure, and checked what NumPy and SciPy are doing for version. It looks like both of them do a bit of work with it in their init.py files--NumPy example here: https://github.com/numpy/numpy/blob/main/numpy/__init__.py#L701
Both those projects have regression tests for correct version handling too. So, I'm a bit hesitant to suggest changing anything right away unless there is already test coverage for whatever you use versioning for.