BF: make logic more consistent for files=[] argument (which is False but not None)

[yarikoptic]

Similar problem was resolved previously at higher level in
http://github.com/datalad/datalad/commit/ec0243c92822f36ada5e87557eb9f5f53929c9ff
for get_content_info. But IMHO it could and should be resolved at the
lowest/common point of invocation, that is why I added it to
_generator_call_git which is underlying most (if not all, hard to tell now ;) )
git invocations.

I also fixed a few other invocations where it was pure if files or if paths.

This commit was initially proposed as a part of the #6974 but there it failed and later I realized that approach I proposed in that PR for get_modules_ was flawed, so this PR is dedicated to only this change (although at once across different commands/levels). Want to see if we could distill it independently.

[codecov]

Codecov Report

Merging #6976 (f7734c6) into maint (29b42b4) will increase coverage by 0.47%.
The diff coverage is 100.00%.

❗ Current head f7734c6 differs from pull request most recent head a6814ba. Consider uploading reports for the commit a6814ba to get more accurate results

@@            Coverage Diff             @@
##            maint    #6976      +/-   ##
==========================================
+ Coverage   90.13%   90.61%   +0.47%     
==========================================
  Files         354      354              
  Lines       46326    46328       +2     
  Branches     6613        0    -6613     
==========================================
+ Hits        41755    41978     +223     
+ Misses       4554     4350     -204     
+ Partials       17        0      -17     

Impacted Files	Coverage Δ
datalad/dataset/gitrepo.py	96.84% <100.00%> (+0.02%)	⬆️
datalad/support/annexrepo.py	90.35% <100.00%> (-0.17%)	⬇️
datalad/support/gitrepo.py	91.02% <100.00%> (-1.72%)	⬇️
datalad/cmdline/__init__.py	0.00% <0.00%> (-100.00%)	⬇️
datalad/interface/run_procedure.py	0.00% <0.00%> (-100.00%)	⬇️
datalad/cmdline/helpers.py	0.00% <0.00%> (-78.27%)	⬇️
datalad/support/nda_.py	0.00% <0.00%> (-54.55%)	⬇️
datalad/__main__.py	39.65% <0.00%> (-29.32%)	⬇️
datalad/support/collections.py	84.37% <0.00%> (-9.38%)	⬇️
datalad/distribution/create_sibling.py	65.24% <0.00%> (-4.56%)	⬇️
... and 31 more

Help us with your feedback. Take ten seconds to tell us how you rate us. Have a feature suggestion? Share it here.

[yarikoptic]

wow -- "all green" . The only confusing item is a flaky test failure on Travis - confused since I thought that we xfail'ed that test_ExtractedArchive recently in https://github.com/datalad/datalad/pull/6912/files . And code in traceback shows that try but not clear how the heck if 'nfsmount' in fpath: did not match or pytest.xfail( didn't raise its own exception? I dont' know. I have restarted that job. And will take this PR our of draft since I think it is the right thing to do/fix here although we might run into cases where it would show changed behavior (but may be "for good")

[yarikoptic]

eh, last commit adjusting for "early decision making" did introduce regression!

        assert_false(
            ds.repo.get_content_annexinfo(paths=[], ref="HEAD", init=None))
        # ... where whatever was passed for init will be returned as is.
>       assert_equal(
            ds.repo.get_content_annexinfo(
                paths=[], ref="HEAD", init={"random": {"entry": "a"}}),
            {"random": {"entry": "a"}})
../datalad/support/tests/test_fileinfo.py:323: 
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ 
first = OrderedDict(), second = {'random': {'entry': 'a'}}, msg = None
    def assert_equal(first, second, msg=None):
        if msg is None:
>           assert first == second
E           AssertionError: assert OrderedDict() == {'random': {'entry': 'a'}}
E             Right contains 1 more item:
E             {'random': {'entry': 'a'}}
E             Full diff:
E             - {'random': {'entry': 'a'}}
E             + OrderedDict()
../datalad/tests/utils_pytest.py:107: AssertionError

because there is apparently some extra semantic of init for dict-like.

        init : 'git' or dict-like or None
          If set to 'git' annex content info will amend the output of
          GitRepo.get_content_info(), otherwise the dict-like object
          supplied will receive this information and the present keys will
          limit the report of annex properties. Alternatively, if `None`
          is given, no initialization is done, and no limit is in effect.

which is imho likely something to be done outside or with a helper... Hence I reverted that shortcut return change, leaving only the portion for that "heavily debated" if paths to stay as is.

[yarikoptic]

appveyor fail is unrelated

FAILED ../datalad/support/tests/test_parallel.py::test_gracefull_death - assert 20 >= 23

Let's proceed with the release and someone could subsequently push on more of similar unification or adding shortcut returns for some performance benefit.

[github-actions]

🚀 PR was released in 0.17.4 🚀