Build complete Market Data backend

[Musa24]

@claude
Read all the documents in the planning directory, then build the complete market data backend in the backend/directory. This should include:

The Massive API interface
The unified market data interface
The market data simulator

Include full unit tests

[claude]

Claude finished @Musa24's task in 5m 28s —— View job • claude/issue-4-20260406-1338 • Create PR ➔

---

Building Market Data Backend

• Gather context and read planning docs
• Explore existing backend structure (no backend existed — built from scratch)
• Create backend/ project scaffold (pyproject.toml, src/, tests/)
• Create abstract market data interface (models.py, cache.py, interface.py)
• Implement GBM-based market data simulator (simulator_engine.py, simulator.py)
• Implement Massive API client (massive.py)
• Create factory and package exports (factory.py, __init__.py)
• Add SSE streaming endpoint (sse.py)
• Wire up FastAPI app (main.py)
• Write unit tests (simulator engine, cache, massive parsing, SSE, factory)
• Commit and push changes

---

What was built

19 files, 1,425 lines implementing the full market data subsystem per planning/MARKET_DATA_DESIGN.md:

Core modules (backend/src/market/)

File	Description
models.py	PriceTick dataclass with slots=True
cache.py	PriceCache — in-memory, asyncio-safe dict
interface.py	MarketDataSource ABC (start/stop/get_prices/add_ticker/remove_ticker/get_tickers)
simulator_engine.py	MarketSimulator — GBM with Cholesky correlated sectors, random events, vectorized numpy
simulator.py	SimulatorDataSource — async wrapper, 500ms background tick loop
massive.py	MassiveDataSource — REST poller (Polygon.io compatible), 15s free tier
factory.py	create_market_source() — selects simulator or Massive from MASSIVE_API_KEY env var
sse.py	SSE endpoint /api/stream/prices — pushes all ticker prices every 500ms
__init__.py	Clean public exports: PriceTick, PriceCache, MarketDataSource, create_market_source

FastAPI app (backend/src/main.py)

• Lifespan wiring: creates PriceCache and MarketDataSource, starts on startup, stops on shutdown
• Exposes /api/stream/prices (SSE) and /api/health

Unit tests (backend/tests/) — 6 test files

• test_simulator_engine.py — GBM math, determinism with seed, correlation matrix, add/remove tickers, edge cases
• test_cache.py — CRUD, copy isolation, remove, empty state
• test_massive.py — response parsing, prev price tracking, HTTP error handling, async start/stop (with respx mocking)
• test_simulator.py — async start/stop, cache population, ticker management
• test_factory.py — env-var selection (no key → simulator, key set → Massive, whitespace → simulator)
• test_sse.py — 200 status, text/event-stream content type, price_update event, field validation, empty cache behavior
  |