Preserve surrogate code points in CSV dialect names - #9639

Summary

Store CSV dialect registry names as WTF-8 instead of requiring valid UTF-8. Python strings may contain lone surrogate code points; CPython accepts such names when registering, selecting, listing, and unregistering a dialect. Lookups borrow the string key and do not allocate a new key.

The existing registry is process-global and is not a Python dict. Its insertion order, key identity, and custom string-subclass or non-string hash/equality behavior remain outside this change. General WTF-8 representation limitations and support for surrogate characters in CSV data or dialect options are also not addressed.

This combines cleanly at file level with #9635 and #9636; no merge was performed.

Validation

• Compared with CPython 3.15.0's Unicode-name registration and reproduced register/get/list/unregister plus reader and writer selection with lone-surrogate names.
• Independent source review of the patch and its composition with Require callable write method for CSV writers #9635/Handle CSV writer callbacks safely #9636, normal commit hooks, and diff checks passed.
• RustPython was not built or executed locally. No tests were added or changed; hosted CI can run the existing test_csv suite.