gh-139482: Improve performance of os.environ.clear()

[eendebakpt]

Implementation based on a suggestion from @gukoff.

When the environ is mutated during the execution of os.environ.clear() (via another thread, or via destructors of the keys removed), the result can be a non-empty environment after the clear(). This is also the case for the current implementation, so this seems acceptable. The exact results can be different, but this seems unavoidable (in particular in the case of threads).
A variation could be to add super().clear() at the end of the definition of clear() (this would not impact performance complexity)

• Issue: os.environ.clear() has a quadratic time complexity #139482

[picnixz]

Actually we already have a PR #139483 so I'm going to close this one as a duplicate. Sorry!

Note: the complexity is still quadratic because unsetenv() rewrites the environment every time it's called. This is unfortunately a drawback from glibc.