Closed wfehr closed 4 months ago
Attention: Patch coverage is 0%
with 4 lines
in your changes are missing coverage. Please review.
Project coverage is 31.18%. Comparing base (
7a1a005
) to head (437a123
).
Files | Patch % | Lines |
---|---|---|
djangocms_transfer/forms.py | 0.00% | 4 Missing :warning: |
:umbrella: View full report in Codecov by Sentry.
:loudspeaker: Have feedback on the report? Share it here.
@wfehr Thanks for this PR!
I just had a quick look. Does with also recover gracefully on re-import? Would it be possible to have a test for that?
@wfehr Thanks for this PR!
I just had a quick look. Does with also recover gracefully on re-import? Would it be possible to have a test for that?
I'm not sure I understand those questions correctly.
As for testing: Besides "migrations" there aren't really tests for anything, which the codecov-bot also "knows". I think for general testing of the project a new issue or something would be better suited.
The changes done here would currently fix an issue introduced with other latest changes.
Sorry, if my question was not clear: Have you tested (manually) to reimport the resulting output?
I understand that tests are close to not existing here. Would it be feasible to create a simple test that creates some plugins (including in a static placeholder), exports them and reimports to verify the process works? If that is not feasible now, it might be a great separate PR.
Sorry, if my question was not clear: Have you tested (manually) to reimport the resulting output?
I understand that tests are close to not existing here. Would it be feasible to create a simple test that creates some plugins (including in a static placeholder), exports them and reimports to verify the process works? If that is not feasible now, it might be a great separate PR.
I tested export/import with this changes, yes. Since this change only affects the filename, the import (if it would be broken so far) wouldn't be affected otherwise since the filename is irrelevant for that.
As for testing: I'll create a new issue for that so this will be done seperately. A new PR can reference that issue then.
Both cases result in an AttributeError for trying 'None.get_slug()'.
This error was introduced with the changes in #15.
With the changes of this PR the occuring
AttributeError
is avoided and a fallback filename used if no "page" is present.