Updating for coverage v4.4

[jonrkarr]

Coverage v4.4 changes how file names are reported in XML report files. For example

<class ... filename="karr_lab_build_utils/init.py" ... name="init.py">

is changed to

<class ... filename="init.py" ... name="init.py">

These commits update the test reporter to accommodate this change. The changes are compatible with v4.4 as well as earlier versions.

[dblandin]

Hey @jonrkarr!

This PR looks great. Thanks for the fix!

Would you mind addressing the Code Climate issues? I'd be fine with removing the inline comments. The added lines seem sufficiently intention-revealing to me.

Is the new version of 0.2.1a0 considered an alpha release by pip? Since we'll release a new version, would 0.2.2 or 0.2.2a0 be more appropriate? In either case, we usually bump versions in separate PRs to keep the version history independent from changes merged into master.

[jonrkarr]

Hi Devon, After I created the pull request I noticed there's already an issue about this. It would be best not to require a specific version of coverage because that would preclude the use of future versions. The coveragepy discussion sounds like the original author isn't going to revert the change that broke python-test-reporter, so it sounds like the solution is to generate complete file paths by combining the sources/source paths and the paths in packages/package/classes/class. For cases with multiple source paths, the author author will need to add an id to sources/source so they can be linked to specific classes and file paths can be unambiguously be resolved. The changes are only affect a couple lines. Feel free to incorporate the changes as you best see fit. Regards Jonathan

…

On Thu, May 11, 2017 at 1:05 PM, Devon Blandin ***@***.***> wrote: Hey @jonrkarr <https://github.com/jonrkarr>! This PR looks great. Thanks for the fix! Would you mind addressing the Code Climate issues? I'd be fine with removing the inline comments. The added lines seem sufficiently intention-revealing to me. Is the new version of 0.2.1a0 considered an alpha release by pip? Since we'll release a new version, would 0.2.2 or 0.2.2a0 be more appropriate? In either case, we usually bump versions in separate PRs to keep the version history independent from changes merged into master. — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub <#41 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/ACt2KXiKi2aRue8pvyTxXhKJDfQ353XRks5r4z_FgaJpZM4NYQSd> .

[dblandin]

For cases with multiple source paths, the author author will need to add an id to sources/source so they can be linked to specific classes and file paths can be unambiguously be resolved.

That sounds like an important addition to make before supporting coverage v4.4 . Would you suggest locking down to <4.4 in the meantime while we figure that out?

We're also working on a new unified test reporter which will support python so I'm less concerned with the long-term maintainability of this particular repo.

https://github.com/codeclimate/test-reporter

[jonrkarr]

Using a specific version of coverage is fine for us for now. I didn't read the coveragepy discussion until after I made the changes. Had I read that first, I would have just specified a specific version for now until coveragepy sorts out this issue.

…

On Thu, May 11, 2017 at 2:55 PM, Devon Blandin ***@***.***> wrote: For cases with multiple source paths, the author author will need to add an id to sources/source so they can be linked to specific classes and file paths can be unambiguously be resolved. That sounds like an important addition to make before supporting coverage v4.4 . Would you suggest locking down to <4.4 in the meantime while we figure that out? We're also working on a new unified test reporter which will support python so I'm less concerned with the long-term maintainability of this particular repo. https://github.com/codeclimate/test-reporter — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub <#41 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/ACt2KaopnQZKjP0sqkH8IWE27L9QXx_Bks5r41mugaJpZM4NYQSd> .

[dblandin]

Had I read that first, I would have just specified a specific version for now until coveragepy sorts out this issue.

Yep, that sounds reasonable to me too. Going to close this PR for now.

Thanks!

[jonrkarr]

Thanks for the quick response on this!

…

On Thu, May 11, 2017 at 3:05 PM, Devon Blandin ***@***.***> wrote: Had I read that first, I would have just specified a specific version for now until coveragepy sorts out this issue. Yep, that sounds reasonable to me too. Going to close this PR for now. Thanks! — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub <#41 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/ACt2KUGSoHUyaI_zEgWzrT-MNcLAM2C3ks5r41wHgaJpZM4NYQSd> .