master-port 51846 and 52969

[mchugh19]

Ports #51846 and #52969 to master.

Likely needed to be merged before #53307 can be rebased.

@gtmanfred @max-arnold

[Akm0d]

unit.test_state.StateCompilerTestCase.test_render_requisite_require_disabled and unit.test_state.StateCompilerTestCase.test_render_requisite_require_in_disabled need to be passing before this can be merged

  File "/tmp/kitchen/testing/tests/unit/test_state.py", line 102, in test_render_requisite_require_disabled
    self.assertEqual(run_num, 0)
AssertionError: 1 != 0

Traceback (most recent call last):
  File "/tmp/kitchen/testing/tests/unit/test_state.py", line 130, in test_render_requisite_require_in_disabled
    self.assertEqual(run_num, 0)
AssertionError: 1 != 0```

[mchugh19]

@Akm0d I think I might need some help here. At first glance the failing tests don't seem to be calling anything in the modified code. I added a few log lines to be sure, and I don't see the output. If all went well, the only functions modified in state.py are _run_check_unless and _run_check_onlyif and the failing tests don't use those modules.

Is it possible there is something else impacting those tests?

[Akm0d]

@mchugh19 The failing tests were added to the test suite with this PR. If they don't touch the actual code that was changed then maybe they can be removed? Debugging it yesterday I found that 'step_one' and 'step_two' in high_data caused __run_num__ to be incremented twice, which seems to be the correct behavior after stepping through with the debugger. The tests will pass consistently if self.assertEqual(run_num, 0) is changed to self.assertEqual(run_num, 1) which again, seems like the correct assertion to make, but I don't see how it tests what the test description says it is testing...

[mchugh19]

@Akm0d Interesting! Thanks for that! Those tests were not in the original PRs, so I didn't notice those tests in there. I don't want to over complicate things, so I've removed them here. (Looks like they were a left over from @garethgreenaway in #49955)

Thanks much!

[dwoz]

@mchugh19 Can you please re-base this on top of the current master?

[mchugh19]

Bam! Done!

[dwoz]

@mchugh19 Looks like the test failure are related, please have a look.

[mchugh19]

@dwoz tests corrected to not rely on local files (and be Debian specific)

[dwoz]

The windows tests passed but an issue in the build is causing it to go red, we are working on resolving that issue for windows.

OK (total=9298, skipped=1212, passed=8086, failures=0, errors=0)

Merging