ENH: Complain like Git if to-be-removed sibling is unknown #4257
Conversation
Codecov Report
@@ Coverage Diff @@
## master #4257 +/- ##
==========================================
- Coverage 89.43% 88.74% -0.69%
==========================================
Files 275 282 +7
Lines 36182 36891 +709
==========================================
+ Hits 32358 32739 +381
- Misses 3824 4152 +328
Continue to review full report at Codecov.
|
|
As far as I can see, the failing tests don't seem to be related to the change in this PR... |
datalad/distribution/siblings.py
Outdated
| @@ -301,6 +301,10 @@ def __call__( | |||
| @staticmethod | |||
| def custom_result_renderer(res, **kwargs): | |||
| from datalad.ui import ui | |||
| # should we attempt to remove an unknown sibling, complain like Git does | |||
| if res['status'] == 'notneeded' and res['action'] == 'remove-sibling': | |||
| ui.message('fatal: No such remote: {name}'.format(**dict(res))) | |||
There was a problem hiding this comment.
Thanks, this looks like an improvement to me. I have the following suggestions, some nit-picks:
-
You could just use
**resinstead of passingrestodict. -
While I think we should complain in the spirit of
git remote rm, I think it'd be better to not complain exactly like it. In particular, I'd say we should drop "fatal: " and replace "remote" with "sibling". -
The message should include some indication of the path so that recursive operation gives a clearer message:
% datalad subdatasets subdataset(ok): subds (dataset) % datalad siblings -r remove -s nonexistent fatal: No such remote: nonexistent fatal: No such remote: nonexistent
There was a problem hiding this comment.
Thanks for these comments, I think I have addressed them now.
Yep, this is unlikely to cause issues given that functionally this is just adding a |
|
uhhh, cool, color! |
|
Travis failure is unrelated and there is no chance that the type of change I made would fail in this test run only. |
fixes #4249