Architecture
Data Flow Architecture
How data moves from the official Helldivers 1 API through validation, storage, and normalization to the frontend. Click any node for details.
Key Concepts
Two-Table Strategy
Raw API responses are stored in rebroadcast tables (faithful JSON), while normalized h1_* tables enable filtering, joining, and aggregation. Both coexist to avoid trade-offs.
Worker Thread
A dedicated thread polls the official API every 5-15 seconds using setTimeout (not setInterval) to prevent overlapping requests. Validates with Zod before any database writes.
Confirm Pattern
Each update cycle creates an unconfirmed season row, writes all child data, then confirms by setting last_updated. This ensures partial writes are detectable.
On-Demand Fetching
Missing seasons are fetched from the official API on first request. The /archives page derives available seasons from the current season number, not a database query.