buffer_platform/app/api/futures_analysis.py

"""
期货智析接口 - 提供期货分析数据
"""
import json
import logging
from pathlib import Path
from typing import Optional

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

from app.database import get_db
from app.analysis_db import get_analysis_db
from app.analysis_models import FuturesAnalysis, WatchedSymbol, AIModelConfig, AnalysisSettings
from app.services.cache import get_cached_data, get_latest_cached

logger = logging.getLogger(__name__)
router = APIRouter(prefix="/futures", tags=["期货智析"])

CONFIG_DIR = Path(__file__).resolve().parent.parent.parent / "config"
SYMBOLS_CONFIG_FILE = CONFIG_DIR / "symbols_config.json"


def _load_symbols_config() -> dict:
    """加载品种配置文件"""
    if not SYMBOLS_CONFIG_FILE.exists():
        return {"futures": {}, "stock": {}}
    with open(SYMBOLS_CONFIG_FILE, "r", encoding="utf-8") as f:
        return json.load(f)


@router.get("/list")
def get_futures_list(db: Session = Depends(get_db)):
    """获取所有期货品种列表及摘要信息（从symbols_config.json读取）"""
    config = _load_symbols_config()
    futures_config = config.get("futures", {})
    
    if not futures_config:
        return {"success": True, "data": []}

    futures_data = []
    for name, symbol_code in futures_config.items():
        cached = get_cached_data(db, symbol_code, "futures")
        if cached and cached.get("timeframes"):
            all_candles = []
            for period, candles in cached.get("timeframes", {}).items():
                all_candles.extend(candles)
            
            if all_candles:
                latest_candle = all_candles[-1]
                open_price = float(latest_candle.get("open", 0))
                close_price = float(latest_candle.get("close", 0))
                high_price = float(latest_candle.get("high", 0))
                low_price = float(latest_candle.get("low", 0))
                
                change = close_price - open_price
                change_pct = (change / open_price * 100) if open_price > 0 else 0
                
                futures_data.append({
                    "symbol": symbol_code,
                    "name": name,
                    "price": close_price,
                    "change": round(change, 2),
                    "changePct": round(change_pct, 2),
                    "suggestion": _get_suggestion(close_price, open_price, change_pct),
                    "suggestionType": "up" if change >= 0 else "down",
                    "periods": _get_period_trends(all_candles),
                    "successRate": _calc_success_rate(all_candles),
                    "trendScore": _calc_trend_score(all_candles),
                    "resistance": round(high_price * 1.02, 2),
                    "support": round(low_price * 0.98, 2),
                    "open": open_price,
                    "high": high_price,
                    "low": low_price,
                    "volume": sum(float(c.get("volume", 0)) for c in all_candles)
                })
        else:
            futures_data.append({
                "symbol": symbol_code,
                "name": name,
                "price": 0,
                "change": 0,
                "changePct": 0,
                "suggestion": "等待数据",
                "suggestionType": "neutral",
                "periods": {"5": "neutral", "15": "neutral", "30": "neutral", "60": "neutral"},
                "successRate": 0,
                "trendScore": 0,
                "resistance": 0,
                "support": 0,
                "open": 0,
                "high": 0,
                "low": 0,
                "volume": 0
            })

    return {"success": True, "data": futures_data}


@router.get("/detail/{symbol}")
def get_futures_detail(symbol: str, db: Session = Depends(get_db)):
    """获取指定期货品种的详细分析数据"""
    cached = get_cached_data(db, symbol, "futures")
    if not cached:
        raise HTTPException(status_code=404, detail=f"未找到 {symbol} 的缓存数据")

    all_candles = []
    for period, candles in cached.get("timeframes", {}).items():
        all_candles.extend(candles)

    if not all_candles:
        raise HTTPException(status_code=404, detail=f"未找到 {symbol} 的K线数据")

    latest_candle = all_candles[-1]
    open_price = float(latest_candle.get("open", 0))
    close_price = float(latest_candle.get("close", 0))
    high_price = float(latest_candle.get("high", 0))
    low_price = float(latest_candle.get("low", 0))
    
    change = close_price - open_price
    change_pct = (change / open_price * 100) if open_price > 0 else 0
    
    resistance1 = round(high_price * 1.01, 2)
    resistance2 = round(high_price * 1.03, 2)
    resistance3 = round(high_price * 1.05, 2)
    support1 = round(low_price * 0.99, 2)
    support2 = round(low_price * 0.97, 2)
    support3 = round(low_price * 0.95, 2)
    
    suggestion = _get_suggestion(close_price, open_price, change_pct)
    suggestion_type = "up" if change >= 0 else "down"
    trend_score = _calc_trend_score(all_candles)
    
    data = {
        "symbol": symbol,
        "name": _get_futures_name(symbol),
        "price": close_price,
        "change": round(change, 2),
        "changePct": round(change_pct, 2),
        "suggestion": suggestion,
        "suggestionType": suggestion_type,
        "suggestionReason": _get_suggestion_reason(symbol, suggestion),
        "open": open_price,
        "high": high_price,
        "low": low_price,
        "volume": sum(float(c.get("volume", 0)) for c in all_candles),
        "entryPrice": round(close_price * 0.995, 2) if change >= 0 else round(close_price * 1.005, 2),
        "targetPrice": resistance1 if change >= 0 else support1,
        "stopLoss": support1 if change >= 0 else resistance1,
        "riskLevel": "低" if trend_score >= 80 else "中" if trend_score >= 60 else "高",
        "macd": _calc_macd(all_candles),
        "rsi": _calc_rsi(all_candles),
        "boll": _calc_boll(all_candles),
        "kdj": _calc_kdj(all_candles),
        "resistances": [resistance1, resistance2, resistance3],
        "supports": [support1, support2, support3],
        "periodConsistency": _get_period_trends(all_candles)
    }

    return {"success": True, "data": data}


@router.get("/kline/{symbol}")
def get_kline_data(symbol: str, period: str = "15", db: Session = Depends(get_db)):
    """获取指定品种和周期的K线数据"""
    period_map = {
        "5": "5min",
        "15": "15min",
        "30": "30min",
        "60": "60min",
        "1440": "daily",
        "daily": "daily"
    }
    db_period = period_map.get(period, f"{period}min")
    
    cached = get_cached_data(db, symbol, "futures", [db_period])
    if not cached or not cached.get("timeframes"):
        raise HTTPException(status_code=404, detail=f"未找到 {symbol} {db_period} 的缓存数据")

    candles = cached["timeframes"].get(db_period, [])
    kline_data = []
    for c in candles:
        time_str = c.get("datetime", c.get("time", ""))
        if time_str and len(time_str) >= 16:
            time_str = time_str[:16].replace("T", " ")
        kline_data.append([
            time_str,
            str(c.get("open", 0)),
            str(c.get("close", 0)),
            str(c.get("low", 0)),
            str(c.get("high", 0)),
            str(int(c.get("volume", 0)))
        ])

    return {"success": True, "data": kline_data}


def _get_futures_name(symbol: str) -> str:
    """根据合约代码获取品种名称"""
    name_map = {
        "AU": "黄金", "AG": "白银", "CU": "铜", "AL": "铝",
        "ZN": "锌", "NI": "镍", "SN": "锡", "PB": "铅",
        "RB": "螺纹钢", "HC": "热卷", "I": "铁矿石", "J": "焦炭",
        "JM": "焦煤", "ZC": "动力煤", "MA": "甲醇", "TA": "PTA",
        "EG": "乙二醇", "PP": "聚丙烯", "L": "塑料", "V": "PVC",
        "M": "豆粕", "RM": "菜粕", "C": "玉米", "CS": "淀粉",
        "A": "豆一", "B": "豆二", "Y": "豆油", "P": "棕榈油",
        "OI": "菜油", "CF": "棉花", "SR": "白糖", "AP": "苹果",
        "JD": "鸡蛋", "LH": "生猪", "FU": "燃料油", "LU": "低硫燃油",
        "SC": "原油", "EC": "集运指数", "BU": "沥青", "RU": "橡胶",
        "NR": "20号胶", "SP": "纸浆", "SS": "不锈钢", "SA": "纯碱",
        "FG": "玻璃", "UR": "尿素", "SF": "硅铁", "SM": "锰硅",
        "IF": "沪深300", "IC": "中证500", "IH": "上证50", "IM": "中证1000",
        "T": "10年期国债", "TF": "5年期国债", "TS": "2年期国债", "TL": "30年期国债",
    }
    return name_map.get(symbol, symbol)


def _get_suggestion(close: float, open: float, change_pct: float) -> str:
    """根据价格走势给出操作建议"""
    if change_pct > 2:
        return "逢低做多"
    elif change_pct > 0.5:
        return "逢低做多"
    elif change_pct > -0.5:
        return "观望等待"
    elif change_pct > -2:
        return "逢高做空"
    else:
        return "逢高做空"


def _get_suggestion_reason(symbol: str, suggestion: str) -> str:
    """获取建议理由"""
    reasons = {
        "逢低做多": "技术面突破，趋势明确，建议逢低介入",
        "逢高做空": "技术面走弱，下行压力增大",
        "观望等待": "多空力量均衡，等待方向明确"
    }
    return reasons.get(suggestion, "等待进一步信号")


def _get_period_trends(candles: list) -> dict:
    """计算各周期趋势 - 根据不同周期取不同长度的K线计算"""
    period_config = {
        "5": {"bars": 10, "threshold": 0.003},
        "15": {"bars": 15, "threshold": 0.005},
        "30": {"bars": 20, "threshold": 0.008},
        "60": {"bars": 30, "threshold": 0.01}
    }
    
    result = {}
    
    for period, cfg in period_config.items():
        bars = cfg["bars"]
        threshold = cfg["threshold"]
        
        if len(candles) < bars:
            result[period] = "neutral"
            continue
        
        recent = candles[-bars:]
        first_close = float(recent[0].get("close", 0))
        last_close = float(recent[-1].get("close", 0))
        
        if first_close <= 0:
            result[period] = "neutral"
            continue
        
        change_pct = (last_close - first_close) / first_close
        
        if change_pct > threshold:
            result[period] = "up"
        elif change_pct < -threshold:
            result[period] = "down"
        else:
            result[period] = "neutral"
    
    return result


def _calc_success_rate(candles: list) -> int:
    """计算交易成功率（简化版）"""
    if len(candles) < 10:
        return 50
    
    wins = 0
    for i in range(1, len(candles)):
        prev_close = float(candles[i-1].get("close", 0))
        curr_close = float(candles[i].get("close", 0))
        if curr_close >= prev_close:
            wins += 1
    
    return int(wins / (len(candles) - 1) * 100)


def _calc_trend_score(candles: list) -> int:
    """计算趋势评分（0-100）"""
    if len(candles) < 5:
        return 50
    
    recent = candles[-10:]
    closes = [float(c.get("close", 0)) for c in recent]
    
    if len(closes) < 2:
        return 50
    
    up_count = sum(1 for i in range(1, len(closes)) if closes[i] >= closes[i-1])
    score = int(up_count / (len(closes) - 1) * 100)
    
    return max(0, min(100, score))


def _calc_ema(data: list, period: int) -> list:
    """计算EMA，返回与输入等长的列表，前面用None填充"""
    ema = [None] * len(data)
    multiplier = 2 / (period + 1)
    
    if len(data) < period:
        return ema
    
    ema[period - 1] = sum(data[:period]) / period
    
    for i in range(period, len(data)):
        ema[i] = (data[i] - ema[i-1]) * multiplier + ema[i-1]
    
    return ema


def _calc_macd(candles: list) -> dict:
    """计算MACD指标"""
    if len(candles) < 26:
        return {"signal": "中性", "detail": "数据不足"}
    
    closes = [float(c.get("close", 0)) for c in candles]
    ema12 = _calc_ema(closes, 12)
    ema26 = _calc_ema(closes, 26)
    
    dif_list = []
    for i in range(len(closes)):
        if ema12[i] is not None and ema26[i] is not None:
            dif_list.append(ema12[i] - ema26[i])
        else:
            dif_list.append(None)
    
    # 只对有效DIF值计算DEA，避免None替换为0导致计算错误
    dif_valid = [d for d in dif_list if d is not None]
    if dif_valid:
        dea_valid = _calc_ema(dif_valid, 9)
        dea_list = [None] * (len(dif_list) - len(dif_valid)) + dea_valid
    else:
        dea_list = [None] * len(dif_list)
    
    dif = dif_list[-1]
    dea = dea_list[-1]
    
    if dif is not None and dea is not None:
        if dif > dea:
            signal = "金叉"
        elif dif < dea:
            signal = "死叉"
        else:
            signal = "中性"
    else:
        signal = "中性"
    
    return {"signal": signal, "detail": f"DIF: {dif:.4f}"}


def _calc_rsi(candles: list) -> dict:
    """计算RSI指标"""
    if len(candles) < 15:
        return {"value": 50, "status": "正常"}
    
    closes = [float(c.get("close", 0)) for c in candles[-15:]]
    gains = []
    losses = []
    
    for i in range(1, len(closes)):
        diff = closes[i] - closes[i-1]
        gains.append(max(0, diff))
        losses.append(max(0, -diff))
    
    avg_gain = sum(gains) / len(gains) if gains else 0
    avg_loss = sum(losses) / len(losses) if losses else 0
    
    if avg_loss == 0:
        rsi = 100
    else:
        rs = avg_gain / avg_loss
        rsi = 100 - (100 / (1 + rs))
    
    rsi = int(rsi)
    if rsi > 70:
        status = "超买"
    elif rsi < 30:
        status = "超卖"
    else:
        status = "正常"
    
    return {"value": rsi, "status": status}


def _calc_boll(candles: list) -> dict:
    """计算布林带"""
    if len(candles) < 20:
        return {"signal": "中轨", "detail": "区间: --"}
    
    closes = [float(c.get("close", 0)) for c in candles[-20:]]
    ma = sum(closes) / len(closes)
    std = (sum((c - ma) ** 2 for c in closes) / len(closes)) ** 0.5
    
    upper = ma + 2 * std
    lower = ma - 2 * std
    current = closes[-1]
    
    if current > upper:
        signal = "上轨外"
    elif current < lower:
        signal = "下轨外"
    elif current > ma:
        signal = "中轨上"
    else:
        signal = "中轨"
    
    return {"signal": signal, "detail": f"区间: {lower:.0f}-{upper:.0f}"}


def _calc_kdj(candles: list) -> dict:
    """计算KDJ指标"""
    if len(candles) < 9:
        return {"signal": "中性", "detail": "K: -- D: --"}
    
    highs = [float(c.get("high", 0)) for c in candles[-9:]]
    lows = [float(c.get("low", 0)) for c in candles[-9:]]
    closes = [float(c.get("close", 0)) for c in candles[-9:]]
    
    highest = max(highs)
    lowest = min(lows)
    current = closes[-1]
    
    if highest == lowest:
        rsv = 50
    else:
        rsv = (current - lowest) / (highest - lowest) * 100
    
    k = int(rsv * 2 / 3 + 50 / 3)
    d = int(k * 2 / 3 + 50 / 3)
    
    if k > d:
        signal = "偏多"
    elif k < d:
        signal = "偏空"
    else:
        signal = "中性"
    
    return {"signal": signal, "detail": f"K: {k} D: {d}"}


# ==================== 期货智析数据管理接口 ====================

@router.get("/analysis/history/{symbol}")
def get_analysis_history(symbol: str, limit: int = 10, adb: Session = Depends(get_analysis_db)):
    """获取品种历史分析记录"""
    records = adb.query(FuturesAnalysis).filter(
        FuturesAnalysis.symbol == symbol
    ).order_by(
        FuturesAnalysis.analysis_time.desc()
    ).limit(limit).all()
    
    return {
        "success": True,
        "data": [{
            "id": r.id,
            "symbol": r.symbol,
            "analysis_time": r.analysis_time.isoformat(),
            "suggestion": r.suggestion,
            "suggestion_type": r.suggestion_type,
            "trend_score": r.trend_score,
            "entry_price": r.entry_price,
            "target_price": r.target_price,
            "stop_loss": r.stop_loss,
            "risk_level": r.risk_level
        } for r in records]
    }


@router.post("/analysis/save")
def save_analysis_record(data: dict, adb: Session = Depends(get_analysis_db)):
    """保存分析记录到数据库"""
    try:
        record = FuturesAnalysis(
            symbol=data.get("symbol"),
            suggestion=data.get("suggestion"),
            suggestion_type=data.get("suggestion_type"),
            entry_price=data.get("entry_price"),
            target_price=data.get("target_price"),
            stop_loss=data.get("stop_loss"),
            risk_level=data.get("risk_level"),
            macd_signal=data.get("macd", {}).get("signal") if data.get("macd") else None,
            rsi_value=data.get("rsi", {}).get("value") if data.get("rsi") else None,
            boll_signal=data.get("boll", {}).get("signal") if data.get("boll") else None,
            kdj_signal=data.get("kdj", {}).get("signal") if data.get("kdj") else None,
            trend_score=data.get("trend_score"),
            success_rate=data.get("success_rate"),
            resistance_levels=data.get("resistances"),
            support_levels=data.get("supports"),
            period_trends=data.get("periodConsistency")
        )
        adb.add(record)
        adb.commit()
        return {"success": True, "message": "分析记录已保存", "id": record.id}
    except Exception as e:
        adb.rollback()
        logger.error(f"保存分析记录失败: {e}")
        return {"success": False, "message": str(e)}


# ==================== 关注品种管理 ====================

@router.get("/watched")
def get_watched_symbols(adb: Session = Depends(get_analysis_db)):
    """获取关注的品种列表"""
    symbols = adb.query(WatchedSymbol).order_by(WatchedSymbol.created_at.desc()).all()
    return {
        "success": True,
        "data": [{
            "id": s.id,
            "symbol": s.symbol,
            "name": s.name,
            "note": s.note,
            "created_at": s.created_at.isoformat()
        } for s in symbols]
    }


@router.post("/watched")
def add_watched_symbol(data: dict, adb: Session = Depends(get_analysis_db)):
    """添加关注品种"""
    try:
        symbol = data.get("symbol")
        existing = adb.query(WatchedSymbol).filter(WatchedSymbol.symbol == symbol).first()
        if existing:
            return {"success": False, "message": "该品种已关注"}
        
        new_symbol = WatchedSymbol(
            symbol=symbol,
            name=data.get("name"),
            note=data.get("note")
        )
        adb.add(new_symbol)
        adb.commit()
        return {"success": True, "message": "已添加关注", "id": new_symbol.id}
    except Exception as e:
        adb.rollback()
        return {"success": False, "message": str(e)}


@router.delete("/watched/{symbol}")
def remove_watched_symbol(symbol: str, adb: Session = Depends(get_analysis_db)):
    """取消关注品种"""
    try:
        record = adb.query(WatchedSymbol).filter(WatchedSymbol.symbol == symbol).first()
        if not record:
            return {"success": False, "message": "未找到该品种"}
        adb.delete(record)
        adb.commit()
        return {"success": True, "message": "已取消关注"}
    except Exception as e:
        adb.rollback()
        return {"success": False, "message": str(e)}


# ==================== AI模型配置管理 ====================

@router.get("/ai-models")
def get_ai_models(adb: Session = Depends(get_analysis_db)):
    """获取AI模型配置列表"""
    models = adb.query(AIModelConfig).order_by(AIModelConfig.created_at.desc()).all()
    settings = adb.query(AnalysisSettings).filter(
        AnalysisSettings.key == "analysis_settings"
    ).first()
    
    return {
        "success": True,
        "data": {
            "models": [{
                "id": m.id,
                "provider": m.provider,
                "model_name": m.model_name,
                "api_base": m.api_base,
                "model_id": m.model_id,
                "temperature": m.temperature,
                "max_tokens": m.max_tokens,
                "enabled": m.enabled,
                "is_active": m.is_active,
                "created_at": m.created_at.isoformat()
            } for m in models],
            "analysis_settings": settings.value if settings else {
                "enable_technical_analysis": True,
                "enable_fundamental_analysis": False,
                "enable_sentiment_analysis": False,
                "risk_tolerance": "medium",
                "max_position_pct": 10
            }
        }
    }


@router.post("/ai-models")
def save_ai_model(data: dict, adb: Session = Depends(get_analysis_db)):
    """保存AI模型配置"""
    try:
        if data.get("action") == "save_settings":
            settings = adb.query(AnalysisSettings).filter(
                AnalysisSettings.key == "analysis_settings"
            ).first()
            if settings:
                settings.value = data.get("settings", {})
            else:
                settings = AnalysisSettings(
                    key="analysis_settings",
                    value=data.get("settings", {})
                )
                adb.add(settings)
            adb.commit()
            return {"success": True, "message": "分析设置已保存"}
        
        model_data = data.get("model", {})
        model = AIModelConfig(
            provider=model_data.get("provider", "custom"),
            model_name=model_data.get("model_name", ""),
            api_key=model_data.get("api_key", ""),
            api_base=model_data.get("api_base"),
            model_id=model_data.get("model_id"),
            temperature=model_data.get("temperature", 0.7),
            max_tokens=model_data.get("max_tokens", 2000),
            enabled=model_data.get("enabled", True),
            is_active=model_data.get("is_active", False)
        )
        
        if model.is_active:
            adb.query(AIModelConfig).update({"is_active": False})
        
        adb.add(model)
        adb.commit()
        return {"success": True, "message": "AI模型已保存", "id": model.id}
    except Exception as e:
        adb.rollback()
        return {"success": False, "message": str(e)}


@router.put("/ai-models/{model_id}")
def update_ai_model(model_id: int, data: dict, adb: Session = Depends(get_analysis_db)):
    """更新AI模型配置"""
    try:
        model = adb.query(AIModelConfig).filter(AIModelConfig.id == model_id).first()
        if not model:
            return {"success": False, "message": "模型不存在"}
        
        if "is_active" in data and data["is_active"]:
            adb.query(AIModelConfig).update({"is_active": False})
            model.is_active = True
        else:
            for key, value in data.items():
                if hasattr(model, key):
                    setattr(model, key, value)
        
        adb.commit()
        return {"success": True, "message": "模型已更新"}
    except Exception as e:
        adb.rollback()
        return {"success": False, "message": str(e)}


@router.delete("/ai-models/{model_id}")
def delete_ai_model(model_id: int, adb: Session = Depends(get_analysis_db)):
    """删除AI模型配置"""
    try:
        model = adb.query(AIModelConfig).filter(AIModelConfig.id == model_id).first()
        if not model:
            return {"success": False, "message": "模型不存在"}
        adb.delete(model)
        adb.commit()
        return {"success": True, "message": "模型已删除"}
    except Exception as e:
        adb.rollback()
        return {"success": False, "message": str(e)}