distribution.subdatasets -> local.subdatasets
#3429
Conversation
Codecov Report
@@ Coverage Diff @@
## master #3429 +/- ##
===========================================
- Coverage 91.3% 36.33% -54.97%
===========================================
Files 265 263 -2
Lines 34488 34460 -28
===========================================
- Hits 31488 12520 -18968
- Misses 3000 21940 +18940
Continue to review full report at Codecov.
|
Codecov Report
@@ Coverage Diff @@
## master #3429 +/- ##
==========================================
+ Coverage 91.2% 91.28% +0.08%
==========================================
Files 265 267 +2
Lines 34489 34486 -3
==========================================
+ Hits 31457 31482 +25
+ Misses 3032 3004 -28
Continue to review full report at Codecov.
|
This would happen on the cmdline when `subdatasets` is executed in a directory. It will now report all subdatasets below to curdir, and not any subdatasets of the dataset we are in.
Following the plan in datalad#3192
mih
changed the title
subdatasetsdistribution.subdatasets -> local.subdatasets
|
This is now ready for review. |
| # no constraints given -> query subdatasets under curdir | ||
| if path is None and dataset is None: | ||
| path = os.curdir | ||
| paths = [rev_resolve_path(p, dataset) for p in assure_list(path)] \ |
There was a problem hiding this comment.
Could be simplified once #3435 is merged, but no requirement to get it done before this PR can be merged.
|
Remaining test failure is the already-known neuroimaging extension test run failure, due to a missing config procedure. |
| if parser is not None: | ||
| # release parser lock manually, auto-cleanup is not reliable in PY3 | ||
| parser.release() | ||
| # Import command class to ease 3rd-party transitions |
There was a problem hiding this comment.
Should we issue a warning (probably with warnings.warn() rather than logging.warn()) when this module is imported?
There was a problem hiding this comment.
I don't think it is needed. Even in our own code base we did not import it directly.
|
I did upload fresh -neuroimaging with the fix, and restarted one sample run which was failing -- https://travis-ci.org/datalad/datalad/builds/534685969?utm_source=github_status&utm_medium=notification -- all green now back then a few commits back before ci skips |
|
All green. @yarikoptic didnt update his review, but I think I successfully addressed his requests. |
This PR implements the transition of the
subdatasetscommand to the "local" API section #3192 . It aligns arguments, behavior, and result structure with the rest "core.local" API.Python API breakage is caused by the necessary introduction of positional
PATHarguments (like the other commands have) in conjunction with the requirement for uniform argument order.--containsstatus/diff/savecurdirstatus/diff/save, e.g.revision->gitshasum(butrevisionis kept as a duplicate for compatibility).gitmodulesanymorelocal/subdatasets.py