Reword error message in wheel and wininst read

[deveshks]

Fixes and closes #598

Note: I will also update https://github.com/pypa/twine/pull/596/files#diff-b1bd28fb732b18d9d7e161b08847044fR78 with the updated error message, after this PR gets merged and I pull master into that PR branch, or vice versa

[bhrutledge]

Looks good, and the I think checks failed due to making changes to files that are missing test coverage. With that in mind, I'm going to wait on merging this until #596 has been merged.

[deveshks]

Agreed, but that will only cover wheel.py. I couldn't find any tests for wininst.py , so would we need to create unit tests for it as well to get this merged?

[bhrutledge]

I couldn't find any tests for wininst.py , so would we need to create unit tests for it as well to get this merged?

I don't think that's necessary, unless you want to. 😉

With the merge of #596, I think this should be rebased on master, and the tests updated to reflect the new message.

[deveshks]

With the merge of #596, I think this should be rebased on master, and the tests updated to reflect the new message.

I did that, but then the codecov/patch only passed for wheel.py, and not wininst.py, since the added code in wininst.py isn't covered by any unit tests

Edit: I can try adding unit tests to wininst.py, but unfortunately I don't have access to a windows machine

[bhrutledge]

I can try adding unit tests to wininst.py, but unfortunately I don't have access to a windows machine

I don't think you need one; the code looks OS-agnostic to me. I suspect that a test_wininst.py could start from a copy of test_wheel.py, replacing instances of .whl with .exe.

But, I'm going to merge this as-is.

[deveshks]

I don't think you need one; the code looks OS-agnostic to me. I suspect that a test_wininst.py could start from a copy of test_wheel.py, replacing instances of .whl with .exe.

But, I'm going to merge this as-is.

The latest checks run after merging it to master also failed on codecov/patch, but If I understand correctly, future merges will only run codecov/patch on the code changes made to those specific commits, so this failure won't have any effect in the longer run?

[bhrutledge]

future merges will only run codecov/patch on the code changes made to those specific commits, so this failure won't have any effect in the longer run?

Correct. So let's get another PR merged. 😉

[bhrutledge]

@deveshks This was released in 3.2.0. Thanks again for your work on this, and all of the testing and typing work.