Fix an error when incomplete S3Client is destroyed

[bryanwweber]

If an error occurs during initialization of the S3Client instance, the initializer for the Client superclass is never called. This could happen, for example, when the S3Client has a bad profile name passed which botocore doesn't understand. The specific error is that the file_cache_mode attribute is not set because the superclass initializer is not called in the S3Client initializer if an error occurs during S3 session setup. This commit moves the superclass initializer to the top of the S3Client initializer to ensure that all superclass attributes are defined if/when any errors occur.

The same fix is applied for the Google client and the local client.

Closes #372

---

Contributor checklist:

• I have read and understood CONTRIBUTING.md
• Confirmed an issue exists for the PR, and the text Closes #issue appears in the PR summary (e.g., Closes #123).
• Confirmed PR is rebased onto the latest base
• Confirmed failure before change and success after change
• Any generic new functionality is replicated across cloud providers if necessary
• Tested manually against live server backend for at least one provider
• Added tests for any new functionality
• Linting passes locally
• Tests pass locally
• Updated HISTORY.md with the issue that is addressed and the PR you are submitting. If the top section is not `## UNRELEASED``, then you need to add a new section to the top of the document for your change.

[codecov]

Codecov Report

Merging #373 (bd14bca) into master (ffbdd1c) will decrease coverage by 0.5%.
The diff coverage is 100.0%.

Additional details and impacted files

@@           Coverage Diff            @@
##           master    #373     +/-   ##
========================================
- Coverage    93.7%   93.2%   -0.5%     
========================================
  Files          22      22             
  Lines        1563    1563             
========================================
- Hits         1465    1458      -7     
- Misses         98     105      +7     

Files	Coverage Δ
cloudpathlib/client.py	89.6% <100.0%> (ø)

... and 2 files with indirect coverage changes

[bryanwweber]

I don't have a Windows machine available and all the tests pass locally for me on my Mac. I think I'll need some help to debug and fix that.

[pjbull]

@bryanwweber I have a preference to do this with getattr in the __del__ method, which will be a little more resilient to future changes and not depend on order of execution.

E.g.,

    def __del__(self) -> None:
        # remove containing dir, even if a more aggressive strategy
        # removed the actual files
        if getattr(self, "file_cache_mode", None) in [
            FileCacheMode.tmp_dir,
            FileCacheMode.close_file,
            FileCacheMode.cloudpath_object,
        ]:
            self.clear_cache()

            if self._local_cache_dir.exists():
                self._local_cache_dir.rmdir()

[bryanwweber]

🎉 great idea @pjbull! I'll make that change

[bryanwweber]

@pjbull Change is made, sorry for the delay!

[pjbull]

@bryanwweber I don't think the Windows failure is due to your change. Going to merge into a repo-local branch to see if that passes, and if not tweak the test on that branch.