Exclude hidden worksheets by default in XLSX conversion

[Siddlab (SiddarthaK16)]

Summary

• Exclude hidden and very hidden worksheets from XLSX conversion by default.
• Add a Python option and CLI flag to include hidden worksheets when needed.
• Document the opt-in and cover default and opt-in behavior with tests.

Closes #1073.

PR #2456 also addresses hidden worksheets, but keeps them included by default for compatibility. This PR follows issue #1073's request for visible-only output by default, with an explicit opt-in for the previous behavior.

[Siddlab (SiddarthaK16)]

@microsoft-github-policy-service agree

[Umesh Patel (UmeshCode1)]

Very clean implementation and well-aligned with #1073.

A few notable positive aspects of this design:

1. Stream Hygiene: In _get_visible_sheet_names, opening with read_only=True avoids loading entire cell matrices just to inspect sheet visibility metadata, and ensuring workbook.close() + file_stream.seek(start_pos) inside finally prevents stream pointer corruption for downstream converters or repair passes.
2. Empty Edge-Case Safety: The if not sheet_names: return {} guard cleanly avoids pd.read_excel throwing when passed an empty list if a workbook has all sheets marked hidden.
3. CLI & Python Option Symmetry: Passing **conversion_options to both convert() and convert_stream() keeps CLI and programmatic usage uniform.
4. Test Coverage: Explicitly asserting both hidden and veryHidden sheets in the test suite verifies the complete OpenXML visibility spec.

LGTM!