feat: get plan bulk with cache

[tomerqodo]

Benchmark PR from qodo-benchmark#445

[greptile-apps]

Greptile Summary

This PR adds a new get_plan_bulk_with_cache method to BillingService that reduces billing API load by caching tenant subscription plans in Redis with a 10-minute TTL. The implementation uses mget for batch cache reads and pipelines for batch writes.

• Added Redis caching layer for bulk plan retrieval with cache-aside pattern
• Added error handling for individual tenant plan validation failures in get_plan_bulk
• Comprehensive integration and unit tests for cache hit/miss scenarios

Issues requiring attention:

• Dead code: __str__ method in SubscriptionPlan TypedDict will never be called (TypedDict instances are plain dicts at runtime)
• Type safety violation: get_plan_bulk assigns None to results but return type is dict[str, SubscriptionPlan]
• Test-code mismatch: Unit test assertions for invalid tenant handling contradict actual code behavior

Confidence Score: 2/5

• This PR has type safety violations and test-code mismatches that should be resolved before merging.
• Score reflects the TypedDict method bug (dead code), type annotation mismatch that violates the return type contract, and unit test assertions that will fail against the actual code behavior.
• api/services/billing_service.py needs fixes for the TypedDict __str__ method and the type annotation mismatch. api/tests/unit_tests/services/test_billing_service.py needs assertions updated to match code behavior.

Important Files Changed

Filename	Overview
api/services/billing_service.py	Adds new get_plan_bulk_with_cache method with Redis caching. Contains critical bugs: dead __str__ method in TypedDict and type annotation mismatch where None values violate return type.
api/tests/test_containers_integration_tests/services/test_billing_service.py	New integration tests for get_plan_bulk_with_cache covering cache hit/miss, Redis failures, and TTL expiration scenarios.
api/tests/unit_tests/services/test_billing_service.py	New unit test for invalid plan handling has assertion that contradicts actual code behavior - test expects key not in result, but code adds key with None value.

Sequence Diagram

sequenceDiagram
    participant Client
    participant BillingService
    participant Redis
    participant BillingAPI

    Client->>BillingService: get_plan_bulk_with_cache(tenant_ids)
    BillingService->>Redis: mget(cache_keys)
    
    alt All Cache Hits
        Redis-->>BillingService: cached plans
        BillingService-->>Client: tenant_plans
    else Partial/Full Cache Miss
        Redis-->>BillingService: partial/empty results
        BillingService->>BillingAPI: get_plan_bulk(missing_ids)
        BillingAPI-->>BillingService: subscription plans
        BillingService->>Redis: pipeline.setex(plans, TTL=600)
        BillingService-->>Client: merged tenant_plans
    else Redis Failure
        Redis-->>BillingService: Exception
        BillingService->>BillingAPI: get_plan_bulk(all_tenant_ids)
        BillingAPI-->>BillingService: subscription plans
        BillingService->>Redis: pipeline.setex(plans, TTL=600)
        BillingService-->>Client: tenant_plans
    end

Loading

[greptile-apps]

_{3 files reviewed, 3 comments}

_{Edit Code Review Agent Settings | Greptile}

[greptile-apps]

logic: TypedDict is just a dict at runtime - methods defined in the class body are ignored. This __str__ method will never be called on SubscriptionPlan instances.

Suggested change

	def __str__(self) -> str:
	"""Return a human-readable string representation for debugging."""
	return f"Plan: {self['plan']}, Expiration: {self['expiration_date']}"

Prompt To Fix With AI

This is a comment left during a code review.
Path: api/services/billing_service.py
Line: 28:30

Comment:
**logic:** `TypedDict` is just a dict at runtime - methods defined in the class body are ignored. This `__str__` method will never be called on `SubscriptionPlan` instances.

```suggestion
```

How can I resolve this? If you propose a fix, please make it concise.

[greptile-apps]

logic: Assigning None violates the return type dict[str, SubscriptionPlan]. Either update the return type to dict[str, SubscriptionPlan | None] or skip adding the key entirely by using continue instead.

Suggested change

	results[tenant_id] = None
	continue

Prompt To Fix With AI

This is a comment left during a code review.
Path: api/services/billing_service.py
Line: 292:292

Comment:
**logic:** Assigning `None` violates the return type `dict[str, SubscriptionPlan]`. Either update the return type to `dict[str, SubscriptionPlan | None]` or skip adding the key entirely by using `continue` instead.

```suggestion
                        continue
```

How can I resolve this? If you propose a fix, please make it concise.

[greptile-apps]

logic: These assertions contradict the actual code behavior. The code at line 292 sets results[tenant_id] = None for invalid tenants, so len(result) will be 3 (not 2) and "tenant-invalid" will be in the result (with None value). Fix the production code or update these assertions.

Prompt To Fix With AI

This is a comment left during a code review.
Path: api/tests/unit_tests/services/test_billing_service.py
Line: 1316:1319

Comment:
**logic:** These assertions contradict the actual code behavior. The code at line 292 sets `results[tenant_id] = None` for invalid tenants, so `len(result)` will be 3 (not 2) and `"tenant-invalid"` will be in the result (with `None` value). Fix the production code or update these assertions.

How can I resolve this? If you propose a fix, please make it concise.