RF: de-wrapt

[yarikoptic]

A quick chat with @mih inspired me to look into it.... in a nutshell (more to come if feasible):

• use inspect.signature instead of inspect.getargspec since that one manages to get args of functools.wrapsed functions. Source of information: https://stackoverflow.com/questions/147816/preserving-signatures-of-decorated-functions#comment68410839_147878

• remaining gotcha (TEMPorarily disabled that test ) -- not yet "binding" __doc__ to Dataset.method -- @build_doc assigns __doc__ to __call__ later than @datasetmethod is invoked. So I guess wrapt does some magic for __doc__ to be always be queried for the wrapted functions. -- I think I came up with a good enough workaround

• reworked getargspec is based on the one in ENH/RF: Interface-s - add * to explicitly separate posargs from kwargs #6176 but redone on top - will conflict. If we go with this -- this one should take over ;-)

oh well -- let's see if anything else breaks ;-)

[yarikoptic]

Woohoo, not yet sure what docs are complaining about, but the rest seems to be happy, so it might be viable

[mih]

YES! I knew it. This is really cool. Thanks for taking it on!

[yarikoptic]

Ha - that test said fails on Windows and Mac so may be it wasn't really wrapt to blame!

======================================================================
ERROR: datalad.runner.tests.test_nonasyncrunner.test_popen_invocation
----------------------------------------------------------------------
Traceback (most recent call last):
  File "C:\Python39-x64\lib\site-packages\nose\case.py", line 198, in runTest
    self.test(*self.arg)
  File "c:\projects\datalad\datalad\tests\utils.py", line 856, in _wrap_with_tempfile
    return t(*(arg + (filename,)), **kw)
  File "c:\projects\datalad\datalad\tests\utils.py", line 856, in _wrap_with_tempfile
    return t(*(arg + (filename,)), **kw)
  File "c:\projects\datalad\datalad\runner\tests\test_nonasyncrunner.py", line 180, in test_popen_invocation
    fetching_data.start()
  File "C:\Python39-x64\lib\multiprocessing\process.py", line 121, in start
    self._popen = self._Popen(self)
  File "C:\Python39-x64\lib\multiprocessing\context.py", line 224, in _Popen
    return _default_context.get_context().Process._Popen(process_obj)
  File "C:\Python39-x64\lib\multiprocessing\context.py", line 327, in _Popen
    return Popen(process_obj)
  File "C:\Python39-x64\lib\multiprocessing\popen_spawn_win32.py", line 93, in __init__
    reduction.dump(process_obj, to_child)
  File "C:\Python39-x64\lib\multiprocessing\reduction.py", line 60, in dump
    ForkingPickler(file, protocol).dump(obj)
TypeError: cannot pickle '_thread.lock' object

[yarikoptic]

ok, dropped that last TEMP optimistic commit (so test will remain skipped on OSX and windows). Fixed up docstring so it builds. Should be ready for tries/review/ and even merge unless some oddity detected

[codecov]

Codecov Report

Merging #6190 (e5505ac) into master (d8270f6) will decrease coverage by 0.65%.
The diff coverage is 87.89%.

❗ Current head e5505ac differs from pull request most recent head 95ac32f. Consider uploading reports for the commit 95ac32f to get more accurate results

@@            Coverage Diff             @@
##           master    #6190      +/-   ##
==========================================
- Coverage   89.25%   88.60%   -0.66%     
==========================================
  Files         317      319       +2     
  Lines       41366    41699     +333     
==========================================
+ Hits        36920    36946      +26     
- Misses       4446     4753     +307     

Impacted Files	Coverage Δ
datalad/interface/__init__.py	100.00% <ø> (ø)
datalad/interface/add_archive_content.py	0.00% <0.00%> (-87.61%)	⬇️
datalad/interface/clean.py	0.00% <0.00%> (-83.34%)	⬇️
datalad/support/external_versions.py	94.87% <ø> (ø)
datalad/local/clean.py	82.85% <82.85%> (ø)
datalad/tests/test_utils.py	96.07% <84.84%> (-0.69%)	⬇️
datalad/distribution/dataset.py	96.26% <85.71%> (-0.38%)	⬇️
datalad/local/add_archive_content.py	87.98% <87.98%> (ø)
datalad/customremotes/archives.py	49.49% <100.00%> (ø)
datalad/interface/base.py	89.51% <100.00%> (+0.05%)	⬆️
... and 94 more

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update d8270f6...95ac32f. Read the comment docs.

[mih]

I think this should fly. Let's try it it! Thx!