add test to verify if sheet names are used for tab labels

[TomTasche]

Reproduces #452

[TomTasche]

was there for HTTP mode already, but missing for file-mode

[claude]

Claude finished @TomTasche's task —— View job

---

Review Tasks:

• Analyze CoreTest.java changes
• Review core_wrapper.cpp modifications
• Assess test coverage
• Check for potential bugs or issues
• Review code quality and best practices
• Evaluate performance considerations
• Check for security concerns

Code Review

I've reviewed this PR that adds test coverage for sheet name functionality in ODS files. Here's my comprehensive feedback:

✅ Strengths

1. Good test coverage: The new test testSpreadsheetSheetNames() properly validates that sheet names from ODS files are correctly extracted and exposed through the CoreResult.pageNames list.

2. Proper test structure: The test follows existing patterns in the test class, including proper setup/teardown of test files and resource management.

3. Clear assertions: Test assertions are descriptive and verify both the count and specific names of sheets.

4. Consistent implementation: The C++ filtering logic was already present in the HTTP mode (hostFileNative) and has now been correctly added to the file mode (parseNative), ensuring consistent behavior across both modes.

🔍 Code Quality & Best Practices

1. Code duplication: The filtering logic is now duplicated between parseNative (lines 224-235) and hostFileNative (lines 454-464). Consider extracting this into a helper function to maintain DRY principles:

static bool shouldSkipView(const odr::DecodedFile& file, const std::string& viewName) {
    if (!file.is_document_file()) return false;
    
    auto docType = file.as_document_file().document_type();
    return ((docType == odr::DocumentType::presentation || 
             docType == odr::DocumentType::drawing) && viewName != "document") ||
           (docType == odr::DocumentType::spreadsheet && viewName == "document");
}

2. Resource naming: The test file naming is consistent (spreadsheet-test.ods follows the pattern of test.odt and password-test.odt).

⚡ Performance Considerations

• The filtering happens during iteration, which is efficient
• No performance concerns with the current implementation

🔒 Security Concerns

• No security issues identified
• Test files are properly isolated in the test assets directory
• File operations use proper resource management with try-with-resources

🐛 Potential Issues

1. Missing binary file: The PR shows spreadsheet-test.ods as added but it's a binary file. Ensure this test asset is properly committed and contains the expected sheet structure (3 sheets named "hey", "ho", and "Sheet3").

2. Error handling: The test assumes the file will parse successfully. Consider adding a test case for malformed ODS files to ensure proper error handling.

📝 Suggestions for Improvement

1. Consider adding a comment in the C++ code explaining why certain views are filtered out for different document types
2. The complex boolean logic in the filtering condition could be simplified for better readability
3. Consider adding more edge cases to the test (empty sheet names, special characters, very long names)

✅ Overall Assessment

This is a solid PR that addresses the issue of sheet names not being exposed for spreadsheets in file mode. The implementation is correct and follows existing patterns. The main improvement would be to reduce code duplication in the C++ filtering logic.