Fix notice_error logic for non-iterable exceptions.

[umaannamalai]

In some cases, caught exceptions passed to notice_error() are singular class errors and not iterables like the notice_error() logic currently expects. This PR updates logic to check if the errors passed to the API exist and are fully populated iterables before grabbing sys.exc_info() instead.

[github-actions]

✅MegaLinter analysis: Success

Descriptor	Linter	Files	Fixed	Errors	Warnings	Elapsed time
✅ ACTION	actionlint	7		0	0	1.06s
✅ MARKDOWN	markdownlint	7	0	0	0	1.32s
✅ PYTHON	ruff	948	0	0	0	1.0s
✅ PYTHON	ruff-format	948	0	0	0	0.4s
✅ YAML	prettier	15	0	0	0	1.54s
✅ YAML	v8r	15		0	0	4.61s
✅ YAML	yamllint	15		0	0	0.69s

See detailed reports in MegaLinter artifacts

[codecov-commenter]

Codecov Report

❌ Patch coverage is 91.66667% with 1 line in your changes missing coverage. Please review.
✅ Project coverage is 81.78%. Comparing base (2dd463d) to head (71703bc).

Files with missing lines	Patch %	Lines
newrelic/core/stats_engine.py	83.33%	0 Missing and 1 partial ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##             main    #1564      +/-   ##
==========================================
+ Coverage   81.76%   81.78%   +0.01%     
==========================================
  Files         207      207              
  Lines       23953    23961       +8     
  Branches     3799     3799              
==========================================
+ Hits        19585    19596      +11     
+ Misses       3101     3098       -3     
  Partials     1267     1267              

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[TimPansino]

This line should only ever deal with the output of sys.exc_info, which is guaranteed to be a tuple of (None, None, None) or (exc, val, tb). This line is adding a bunch of unnecessary complexity when the original code here should have been fine.

[TimPansino]

If the intent of this change is to make it possible to pass in an instance of BaseException to be able to report it then this change is overcomplicated and unfortunately looks to be introducing a bug.

Take for example this code:

try:
    raise ValueError()
except ValueError as exc1:
    pass

try:
    raise TypeError()
except TypeError as exc2:
    notice_error(exc1)

Previously the exception instance was not allowed as an input, and would have caused an error that would crash.

With this PR's changes instead, it silently reports the wrong exception. The user is trying to pass in an exception to report, but as written this will trigger a pull from sys.exc_info() which will yield back the tuple for exc2, instead of exc1 which was passed in.

There's also a lot of complexity in these if statements now that I don't care for, as well as a bunch of new function calls that add overhead. Instead, I propose we construct a tuple in the form we expect from the exception instance if and only if the error passed in is an instance of BaseException and it has a non None __traceback__ attribute.