Fix save to not force committing into git if reference dataset is pure git (not git-annex) #7355
Conversation
otherwise there is no point of having two tests - one for GitRepo and one for AnnexRepo
…aset.create otherwise it becomes an AnnexRepo and thus behaving pretty much the same way as in AnnexRepo test and actually test running the same duration. With this change we get 11.45s call datalad/support/tests/test_repo_save.py::test_annexrepo_save_all 8.14s call datalad/support/tests/test_repo_save.py::test_gitrepo_save_all thus attesting to the fact that we get different testing for GitRepo case
Although the file is about .repo.save and not ds.save, there is already a good number of other tests which pretty much test ds.save. It is also the only place which already uses that convoluted scenario so I decided it would be good place for an addition of another testing thus not introducing much of extra run time as if I was to create an independent new test for this problem.
yarikoptic
added
semver-patch
github-actions
bot
removed
the
CHANGELOG-missing
Codecov ReportPatch coverage:
Additional details and impacted files@@ Coverage Diff @@
## maint #7355 +/- ##
==========================================
+ Coverage 88.73% 90.70% +1.97%
==========================================
Files 327 327
Lines 44567 44569 +2
Branches 5906 5906
==========================================
+ Hits 39545 40428 +883
+ Misses 5007 4126 -881
Partials 15 15
... and 4 files with indirect coverage changes Help us with your feedback. Take ten seconds to tell us how you rate us. Have a feature suggestion? Share it here. ☔ View full report in Codecov by Sentry. |
|
ha -- #7353 lacked codecov report so might be due to that but here we even got |
|
I would even say that we as well release after this fix since it would already be useful to some folks I know ;-) |
|
It is as green as you can ever see a PR here, with even +1.97% boost in coverage -- what could be better? ;) I will merge unless objections voiced on Sunday. |
yarikoptic
added
the
merge-if-ok
|
PR released in |
Sits on top of #7353 to reuse existing testing setup. Actual fix is few character change in
save.pyand is present in the last non-changelog commit.Closes #7351