move is_null_datetimelike to nattype for self-containment #21692
Conversation
Codecov Report
@@ Coverage Diff @@
## master #21692 +/- ##
=======================================
Coverage 91.9% 91.9%
=======================================
Files 154 154
Lines 49659 49659
=======================================
Hits 45640 45640
Misses 4019 4019
Continue to review full report at Codecov.
|
pandas/_libs/missing.pxd
Outdated
| @@ -1,6 +1,12 @@ | |||
| # -*- coding: utf-8 -*- | |||
| # cython: profile=False | |||
|
|
|||
| cdef bint is_null_datetimelike(object val) | |||
| from tslibs.nattype cimport is_null_datetimelike | |||
| # pass-through so other modules can cimport is_null_datetimelike from here | |||
There was a problem hiding this comment.
why is this comment needed? this is for api simplicity yes? if so i don’t think really necessary for a comment
|
needs a rebase |
jreback
added
Datetime
| return True | ||
| elif val is NaT: | ||
| return True | ||
| elif util.is_timedelta64_object(val): |
There was a problem hiding this comment.
why import this from util? Didn't you move it to tslibs.missing ?
There was a problem hiding this comment.
Not sure I understand the question. This PR is moving this function from _libs.missing to tslibs.nattype because doing so simplifies the dependency structure.
There was a problem hiding this comment.
and by moving the function should not change the import?
There was a problem hiding this comment.
these util functions of basically c versions of isinstance checks, so this is ok.
|
thanks @jbrockmendel |
ATM
is_null_datetimelikeis used only intslibs.period. The dependency structure is simplified nicely if we put that function istslibs.nattypeinstead of in_libs.missing.This also takes a couple of missing-themed functions out of inference.pyx
Finally, moves the libperiod block in setup.py to down with the rest of the tslibs blocks to restore alphabetical order.