| name | ngat-tradingview-backend |
| description | Maintain the FastAPI service in tradingview_dashboard/backend for candles, indicators, and latest timestamp endpoints backed by RedisTimeSeries. Use when changing API contracts, performance, indicator response shape, or Redis query behavior for frontend consumers. |
Ngat Tradingview Backend
Overview
Provide read APIs over Redis-stored OHLC and computed indicator data for chart consumers.
Submodules
tradingview_dashboard/backend/main.py: /candles, /indicator, /indicators, /candles/latest.tradingview_dashboard/backend/pyproject.toml: backend dependency constraints.
Pros
- Keep API slim and data-focused.
- Use label-driven RedisTimeSeries filters for efficient field retrieval.
- Expose indicator endpoints directly for frontend overlays.
Cons
- Keep duplicate endpoint families (
/candles and /_candles) with overlapping purpose.
- Compute indicators on request without caching or batching.
- Include hardcoded
COUNT 500 in queries, not dynamically tied to limit.
Caveats
- Keep indicator output schemas stable; frontend assumes specific lowercased column names.
- Guard against missing field series; endpoint currently hard-fails when any OHLCV field is missing.
- Verify pandas-ta column names for KC; upstream naming may differ from
KC_Upper/Middle/Lower.
Forward Features
- Add pagination/windowing contracts with cursor or explicit candle index.
- Add server-side indicator caching keyed by symbol/timeframe/window/config hash.
- Add pydantic response models and OpenAPI examples for frontend contract safety.
