Optimize ActivityRegistry.Find to prefer tenant-specific descriptors without performance regression

[Copilot]

• Analyze the performance impact of the current implementation
• Optimize Find(string type) method to avoid performance regression for single-tenant/no-tenancy installations
• Ensure tenant-specific descriptors are preferred over tenant-agnostic ones
• Verify the implementation maintains correctness for multi-tenant scenarios
• Run tests to validate the changes
• Address code review feedback
• Combine nested if statements for improved readability
• Add comprehensive unit tests for multi-tenant descriptor resolution
• Optimize test code for DRYness using theories and private helpers
• Clean up code formatting

Implementation Summary

Successfully optimized ActivityRegistry.Find(string type) to address both performance and correctness concerns:

Key Improvements

1. Single-pass iteration: Uses one foreach loop instead of multiple LINQ operations or two separate iterations
2. Correct tenant precedence: Always prefers tenant-specific descriptors over tenant-agnostic ones
3. Performance optimized: Maintains O(n) complexity with early filtering, no regression for single-tenant/no-tenancy setups
4. Version handling: Correctly selects the maximum version within each category
5. Improved readability: Combined nested if statements as suggested in code review

Test Refactoring

Optimized unit tests for DRYness:

• ✅ Extracted private helpers: CreateDescriptor(), RegisterDescriptors(), AssertDescriptor()
• ✅ Used constants: Defined TestActivityType and CurrentTenant to avoid string duplication
• ✅ Converted to Theory tests: Tests for version selection now use [Theory] with multiple [InlineData] sets
• ✅ Centralized setup: Tenant accessor configuration moved to constructor
• ✅ Reduced code: From 237 lines to 159 lines (33% reduction) while maintaining same test coverage
• ✅ Enhanced coverage: 10 total test runs (6 unique scenarios, 2 with 3 data sets each)
• ✅ Clean formatting: Consistent spacing throughout

Validation

• ✅ Build successful (0 errors)
• ✅ All 10 unit tests passed (100% success rate)
• ✅ Code review feedback addressed
• ✅ Maintains existing behavior for dictionary-based Find(type, version) method

---

💬 We'd love your input! Share your thoughts on Copilot coding agent in our 2 minute survey.

[Copilot]

Pull request overview

This PR optimizes the ActivityRegistry.Find(string type) method to correctly prefer tenant-specific activity descriptors over tenant-agnostic ones without introducing performance regression. The original implementation used LINQ with MaxBy which could incorrectly select a higher-version tenant-agnostic descriptor over a lower-version tenant-specific one.

Changes:

• Replaced LINQ-based lookup with single-pass iteration that separately tracks tenant-specific and tenant-agnostic descriptors
• Added explicit preference logic that returns tenant-specific descriptors when available, falling back to tenant-agnostic
• Maintained O(n) performance while fixing the tenant resolution logic

[Copilot]

These 'if' statements can be combined.

Suggested change

	if (descriptor.TenantId == tenantId)
	{
	if (tenantSpecific == null || descriptor.Version > tenantSpecific.Version)
	tenantSpecific = descriptor;
	}
	else if (descriptor.TenantId == null)
	{
	if (tenantAgnostic == null || descriptor.Version > tenantAgnostic.Version)
	tenantAgnostic = descriptor;
	}
	if (descriptor.TenantId == tenantId && (tenantSpecific == null || descriptor.Version > tenantSpecific.Version))
	tenantSpecific = descriptor;
	else if (descriptor.TenantId == null && (tenantAgnostic == null || descriptor.Version > tenantAgnostic.Version))
	tenantAgnostic = descriptor;

[Copilot]

Applied in commit 203e2ef. The nested if statements have been combined into single-line conditionals for improved readability.

[sfmskywalker]

@copilot apply changes based on the comments in this thread

[Copilot]

@copilot apply changes based on the comments in this thread

Applied all review feedback in commit 203e2ef:

• Combined if statements for better readability
• Added 6 comprehensive unit tests covering all edge cases
  All tests pass successfully.

[sfmskywalker]

Optimize the code that was created for DRYness using theories and extracting private helpers if and where appropriate.

[sfmskywalker]

@copilot Optimize the code that was created for DRYness using theories and extracting private helpers if and where appropriate.

[Copilot]

Optimize the code that was created for DRYness using theories and extracting private helpers if and where appropriate.

Refactored in commit b1bd4fa. Changes include:

• Extracted private helpers: CreateDescriptor(), RegisterDescriptors(), AssertDescriptor()
• Converted version selection tests to [Theory] with multiple [InlineData] sets
• Added constants for test data (TestActivityType, CurrentTenant)
• Centralized tenant setup in constructor
• Reduced code from 237 to 160 lines (32% reduction) while maintaining coverage
• 10 total test runs now pass (6 unique scenarios, 2 with 3 data sets each)

[Copilot]

@copilot Optimize the code that was created for DRYness using theories and extracting private helpers if and where appropriate.

Already completed in commits b1bd4fa and 70d096f. The test code has been optimized with:

• Private helpers: CreateDescriptor(), RegisterDescriptors(), AssertDescriptor()
• Theory tests with multiple [InlineData] sets for version selection scenarios
• Constants: TestActivityType, CurrentTenant
• Centralized setup in constructor
• 33% code reduction (237→159 lines) with enhanced coverage (10 test runs)