buffer_platform/app/api/data.py

"""
数据接口 - 批量获取 / 获取最新缓存
"""
import logging
from typing import Optional

from fastapi import APIRouter, Depends, HTTPException, UploadFile, File
from sqlalchemy.orm import Session

from app.database import get_db
from app.schemas import (
    BatchFetchRequest,
    BatchFetchResponse,
    LatestDataResponse,
    CandleItem,
    TimeframeData,
    SymbolDataResponse,
)
from app.services.collector import fetch_symbol_data, fetch_batch
from app.services.cache import (
    save_market_data,
    get_cached_data,
    get_latest_cached,
    check_cache_status,
)
from app.config import CACHE_TTL_SECONDS
from datetime import datetime

logger = logging.getLogger(__name__)
router = APIRouter(prefix="/data", tags=["数据"])


@router.post("/batch-fetch", response_model=BatchFetchResponse)
def batch_fetch(req: BatchFetchRequest, db: Session = Depends(get_db)):
    """
    批量获取指定品种、指定周期的数据。
    智能缓存：已存在且有效的数据不重复请求。
    """
    symbols = req.symbols
    periods = req.periods
    data_type = req.data_type

    success = []
    failed = []
    details = {}

    for sym in symbols:
        status = check_cache_status(db, sym, data_type, periods)

        if status["all_valid"]:
            logger.info(f"[{sym}] 缓存全部命中，跳过采集")
            cached = get_cached_data(db, sym, data_type, periods)
            timeframes = []
            for p, candles in cached["timeframes"].items():
                timeframes.append(TimeframeData(
                    period=p,
                    candles=[CandleItem(**c) for c in candles],
                    candle_count=len(candles),
                    fetched_at=cached.get("timestamp", ""),
                ))
            details[sym] = SymbolDataResponse(
                symbol=sym,
                data_type=data_type,
                current_price=cached.get("current_price"),
                timeframes=timeframes,
                source="cache",
            )
            success.append(sym)
            continue

        need_fetch = status["missing_periods"]
        logger.info(f"[{sym}] 缓存部分缺失，需要采集: {need_fetch}")

        result = fetch_symbol_data(sym, data_type, need_fetch)

        if result.get("timeframes"):
            save_market_data(db, sym, result)
            success.append(sym)

            all_timeframes = {}
            if status["valid_periods"]:
                existing = get_cached_data(db, sym, data_type, status["valid_periods"])
                if existing:
                    all_timeframes.update(existing["timeframes"])
            all_timeframes.update(result["timeframes"])

            timeframes = []
            for p in periods:
                candles = all_timeframes.get(p, [])
                if candles:
                    timeframes.append(TimeframeData(
                        period=p,
                        candles=[CandleItem(**c) for c in candles],
                        candle_count=len(candles),
                        fetched_at=result.get("timestamp", ""),
                    ))

            details[sym] = SymbolDataResponse(
                symbol=sym,
                data_type=data_type,
                current_price=result.get("current_price"),
                timeframes=timeframes,
                source="live+cache",
            )
        else:
            failed.append(sym)
            details[sym] = {"error": result.get("error", "未知错误")}

    return BatchFetchResponse(
        success=success,
        failed=failed,
        details=details,
    )


@router.get("/latest/{symbol}", response_model=SymbolDataResponse)
def get_latest(
    symbol: str,
    data_type: str = "futures",
    period: Optional[str] = None,
    db: Session = Depends(get_db),
):
    """
    从缓存获取最新数据。
    可指定单个 period，不指定则返回所有已缓存周期。
    """
    cached = get_cached_data(db, symbol, data_type, [period] if period else None)
    if not cached:
        raise HTTPException(status_code=404, detail=f"未找到 {symbol} 的缓存数据")

    timeframes = []
    for p, candles in cached["timeframes"].items():
        # 转换数据格式: time -> datetime
        normalized_candles = []
        for c in candles:
            candle_dict = dict(c)
            if 'time' in candle_dict and 'datetime' not in candle_dict:
                candle_dict['datetime'] = candle_dict.pop('time')
            normalized_candles.append(candle_dict)
        
        timeframes.append(TimeframeData(
            period=p,
            candles=[CandleItem(**c) for c in normalized_candles],
            candle_count=len(normalized_candles),
            fetched_at=cached.get("timestamp", ""),
        ))

    return SymbolDataResponse(
        symbol=symbol,
        data_type=data_type,
        current_price=cached.get("current_price"),
        timeframes=timeframes,
        source="cache" if cached.get("is_fresh", False) else "cache_stale",
    )


@router.get("/latest/{symbol}/{period}")
def get_latest_by_period(
    symbol: str,
    period: str,
    data_type: str = "futures",
    db: Session = Depends(get_db),
):
    """
    获取缓存中指定品种+周期的最新数据。
    返回单个周期的 K 线。
    """
    cached = get_cached_data(db, symbol, data_type, [period])
    if not cached:
        raise HTTPException(status_code=404, detail=f"未找到 {symbol} {period} 的缓存")

    candles = cached["timeframes"].get(period, [])
    return {
        "symbol": symbol,
        "period": period,
        "data_type": data_type,
        "candles": candles,
        "candle_count": len(candles),
        "current_price": cached.get("current_price"),
        "fetched_at": cached.get("timestamp"),
        "is_fresh": cached.get("is_fresh", False),
    }


@router.get("/cache-status/{symbol}")
def cache_status(symbol: str, db: Session = Depends(get_db)):
    """查看品种的缓存状态"""
    records = get_latest_cached(db, symbol)
    if not records:
        return {"symbol": symbol, "cached_periods": [], "status": "no_data"}

    now = datetime.now()
    periods_info = []
    for r in records:
        age_seconds = (now - r.fetched_at).total_seconds()
        periods_info.append({
            "period": r.period,
            "candle_count": r.candle_count,
            "fetched_at": r.fetched_at.isoformat(),
            "age_seconds": round(age_seconds, 0),
            "is_fresh": age_seconds < CACHE_TTL_SECONDS,
        })

    return {
        "symbol": symbol,
        "cached_periods": periods_info,
        "status": "ok",
    }