Add tests for feature manager

[linglingye001]

NewFeatureManager initialization:

• Nil provider returns error
• Custom filters are registered and used correctly
• Nil filter in options doesn't cause a panic

GetFeatureNames:

• Returns all feature flag names correctly
• Returns empty slice for empty provider
• Returns nil when provider errors

Unknown/missing filter:

• Feature with unregistered filter returns disabled

RequirementType logic:

• All — enabled when all filters return true
• All — disabled when any filter returns false
• Any — enabled when at least one filter returns true
• Any — disabled when all filters return false

Filter error handling:

• Filter returning an error propagates the error

StatusOverride:

• Variant with StatusOverride: "Enabled" keeps feature enabled

GetVariant edge cases:

• Non-existent feature returns error
• Feature with no variants returns nil variant
• Feature with no allocation returns nil variant
• User allocation overrides default_when_enabled
• Pointer *TargetingContext works the same as value

IsEnabledWithAppContext:

• Non-TargetingContext app context still works
• Provider error propagates correctly

Validation errors:

• Empty feature flag ID
• Invalid RequirementType
• Variant missing name
• Invalid StatusOverride value
• Invalid percentile range (negative From)
• User allocation with empty users list
• Group allocation with empty groups list
• Client filter missing name
• Percentile allocation missing variant
• User allocation missing variant
• Group allocation missing variant

Provider errors:

• IsEnabled, IsEnabledWithAppContext, and GetVariant all return errors when provider fails

FeatureManagement struct:

• JSON deserialization into the FeatureManagement type works correctly

[Copilot]

Pull request overview

This PR adds a comprehensive set of unit tests around the FeatureManager API to validate initialization behavior, filter evaluation semantics, variant selection edge cases, and validation/provider error paths.

Changes:

• Adds tests for NewFeatureManager initialization scenarios (nil provider, custom filters, nil filters in options).
• Adds tests for GetFeatureNames, missing/unknown filters, requirement type evaluation (Any/All), and filter error propagation.
• Adds tests covering variant assignment edge cases, provider error propagation across APIs, and JSON deserialization into FeatureManagement.

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

[Copilot]

TestGetFeatureNames asserts a fixed ordering of returned names, but GetFeatureNames does not document any ordering guarantee (it just forwards whatever order the provider returns). This can make the test flaky or overly constraining for alternative providers; consider sorting both slices or comparing as sets instead of by index.

[Copilot]

The result of json.Unmarshal is ignored here. If the JSON is malformed (or the schema changes), the test will fail later with a less actionable error; capture and assert the unmarshal error so failures point to the real cause.

Suggested change

	json.Unmarshal([]byte(jsonData), &fm)
	if err := json.Unmarshal([]byte(jsonData), &fm); err != nil {
	t.Fatalf("Failed to unmarshal test JSON: %v", err)
	}

[Copilot]

The result of json.Unmarshal is ignored here. Please check and fail the test on unmarshal error to avoid masking the root cause of failures.

Suggested change

	json.Unmarshal([]byte(jsonData), &fm)
	if err := json.Unmarshal([]byte(jsonData), &fm); err != nil {
	t.Fatalf("Failed to unmarshal test JSON: %v", err)
	}

[Copilot]

This subtest ignores the error returned from IsEnabled. If evaluation unexpectedly errors (e.g., validation or filter issues), the assertion may report an incorrect cause; assert err == nil before checking enabled.

Suggested change

	enabled, _ := fm.IsEnabled("AnyFilterMixed")
	if !enabled {
	t.Error("Expected feature to be enabled when one filter returns true with RequirementTypeAny")
	}
	})
	t.Run("All filters false", func(t *testing.T) {
	enabled, _ := fm.IsEnabled("AnyFilterAllFalse")
	enabled, err := fm.IsEnabled("AnyFilterMixed")
	if err != nil {
	t.Fatalf("Expected no error evaluating feature AnyFilterMixed: %v", err)
	}
	if !enabled {
	t.Error("Expected feature to be enabled when one filter returns true with RequirementTypeAny")
	}
	})
	t.Run("All filters false", func(t *testing.T) {
	enabled, err := fm.IsEnabled("AnyFilterAllFalse")
	if err != nil {
	t.Fatalf("Expected no error evaluating feature AnyFilterAllFalse: %v", err)
	}

[Copilot]

This subtest ignores the error returned from IsEnabled. Please assert the error is nil before asserting on the enabled state.

Suggested change

	enabled, _ := fm.IsEnabled("AnyFilterMixed")
	if !enabled {
	t.Error("Expected feature to be enabled when one filter returns true with RequirementTypeAny")
	}
	})
	t.Run("All filters false", func(t *testing.T) {
	enabled, _ := fm.IsEnabled("AnyFilterAllFalse")
	enabled, err := fm.IsEnabled("AnyFilterMixed")
	if err != nil {
	t.Fatalf("Expected no error from IsEnabled: %v", err)
	}
	if !enabled {
	t.Error("Expected feature to be enabled when one filter returns true with RequirementTypeAny")
	}
	})
	t.Run("All filters false", func(t *testing.T) {
	enabled, err := fm.IsEnabled("AnyFilterAllFalse")
	if err != nil {
	t.Fatalf("Expected no error from IsEnabled: %v", err)
	}