stdlib.run: Fix missing executable

[pirat89]

Currently, if the executable required in run does not exist (or is not executable), the child process's code is not replaced by os.execvpe function and it raises the OSError instead. However, the parent process does not get this OSError. It consumes exit code, stderr, ... of the child process. So in case the code does something like this:

```
try:
    result = run(['non-executable'])
except OSError:
    pass
except CalledProcessError:
    # do something..
```

the child process passes and do whatever.. In case it ends with zero exit code, the obtained result is usually something totally different than expected in actors. Also there could be problems with non-idempotent code, when some actions could be done twice (once executed by the child, send time executed by the parent [current] process).

We have realized that number of existing leapp actors for in-place upgrades already count with the raise of OSError when executable cannot be used. So we choosed for now to check whether executables are present and raise OSError if not, so we are sure that only one process leave the function really.

Also applied another seatbelt into the child process - if the OSError is raised anyway despite our checks (e.g. SELinux prevents the execution) let's just kill the process instead giving it a possibility to continue. In such a case, always print a msg to stderr of the child process and exit with ecode 1.

Note: To check an executable we use distutils.spawn.find_executable which is deprecated in Python 3 and will be dropped in Python 3.12. However it exists now for Python 2 & 3, so we use this one for now and will replace it in future by shutils.which when the time comes.

Additional changes:

• Make pylint happy (set noqa for E721 for the check the Fields is not
  created directly). The "isinstance" function has a different behaviour
  in this case than we want.

• Fix imports to make pylint happy

• Set result to empty dict instead of None: in case a ValueError
  or TypeError is raised inside the _call function, it's expected
  to copy the result content inside the final block, however in
  case the results is None, it fails and raise additional exception
  that covers the original one.

• Update unit tests to cover issues with missing executable

Missing:

• update unit-tests

[github-actions]

Thank you for contributing to the Leapp project!

Please note that every PR needs to comply with the Leapp Guidelines and must pass all tests in order to be mergeable.
If you want to request a review or rebuild a package in copr, you can use following commands as a comment:

• review please @oamg/developers to notify leapp developers of the review request
• /packit copr-build to submit a public copr build using packit

To launch regression testing public members of oamg organization can leave the following comment:

• /rerun to schedule basic regression tests using this pr build and leapp-repository*master* as artifacts
• /rerun 42 to schedule basic regression tests using this pr build and leapp-repository*PR42* as artifacts
• /rerun-sst to schedule sst tests using this pr build and leapp-repository*master* as artifacts
• /rerun-sst 42 to schedule sst tests using this pr build and leapp-repository*PR42* as artifacts

Please open ticket in case you experience technical problem with the CI. (RH internal only)

Note: In case there are problems with tests not being triggered automatically on new PR/commit or pending for a long time, please consider rerunning the CI by commenting leapp-ci build (might require several comments). If the problem persists, contact leapp-infra.

[fernflower]

Can you add a unit test please? Currently the description reads pretty scary and not that easy to grasp, a repro can help explain what's it actually about imho

[pirat89]

@fernflower seems the current impementatino of the fix is wrong as bunch of checks (some expected by unit tests..) are executed inside the _call function.

I will think whether to duplicate some checks or whether to move some functionality deeper.

[pirat89]

/rerun

[github-actions]

Copr build succeeded: https://copr.fedorainfracloud.org/coprs/build/6234502

[github-actions]

Testing Farm request for RHEL-8.6-rhui/6224797;6234502 regression testing has been created.
Once finished, results should be available here.
Full pipeline log.

[github-actions]

Testing Farm request for RHEL-7.9-rhui/6224797;6234502 regression testing has been created.
Once finished, results should be available here.
Full pipeline log.

[pirat89]

@fernflower btw, just realized the unit-tests is invalid. updating it now

[fernflower]

/rerun

[github-actions]

Copr build succeeded: https://copr.fedorainfracloud.org/coprs/build/6237236

[github-actions]

Testing Farm request for RHEL-8.6-rhui/6234661;6237236 regression testing has been created.
Once finished, results should be available here.
Full pipeline log.

[github-actions]

Testing Farm request for RHEL-7.9-rhui/6234661;6237236 regression testing has been created.
Once finished, results should be available here.
Full pipeline log.

[fernflower]

lgtm, thanks @pirat89

[pirat89]

@fernflower rebased to fix the conflict. can you reapprove/rereview it and merge?

UPDATE: give me a minute. tests started to fail. fixed