Add rename_identical_headers option to get_all_records

[ChrisJr404]

Implements #1436.

get_all_records() raises when the header row has duplicate or blank entries, since they can't all be dictionary keys. This adds an opt-in rename_identical_headers flag so those headers are made unique instead, similar to how pandas reads a sheet like that:

# header row: A1, faff, C3, faff
records = worksheet.get_all_records(rename_identical_headers=True)
# keys become: A1, faff, C3, faff_1

The renaming itself lives in a new utils.make_headers_unique() helper (as suggested in the issue), so it's also usable on its own alongside to_records(). Blank headers get named after their column position, duplicates get a numbered suffix, and both patterns are overridable via a str.format template.

Default is False, so existing behavior is unchanged. Added unit tests for the helper and an integration test for the new flag, plus a note in the user guide.