#!/usr/bin/env python3
"""market_watch.py — 行業熱點數據採集，寫入 dashboard data/market.json

數據源優先級：
  後端A：東方財富 push2 API（首選，有板塊代碼+實時指數）
  後端B：同花順 THS / akshare（降級，有漲跌家數+資金流向）

注意：當前服務器無法連通東方財富API（已被封禁/域名不可達），
實際運行時自動降級到同花順 THS 後端。THS 提供90+行業板塊的
實時漲跌、上漲/下跌家數、淨流入資金等數據，足以滿足需求。

輸出：data/market.json → MoFin Dashboard 市場數據展示
"""

import json
from datetime import datetime
from pathlib import Path

from mofin_db import get_conn, init_all_tables, write_market_snapshot

DATA_DIR = Path(__file__).parent / "data"


# ── 後端A：東方財富 push2 API（首選，有板塊代碼+實時指數） ──

def _fetch_em(url):
    """通用 EM API 請求"""
    import urllib.request
    req = urllib.request.Request(
        url,
        headers={"User-Agent": "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36"}
    )
    resp = urllib.request.urlopen(req, timeout=10)
    return json.loads(resp.read().decode("utf-8"))


def fetch_sector_em():
    """東方財富行業板塊"""
    try:
        data = _fetch_em(
            "https://push2.eastmoney.com/api/qt/clist/get?"
            "pn=1&pz=60&po=1&np=1&fields=f2,f3,f4,f12,f14&fs=m:90+t:2"
        )
        return [{
            "name": i["f14"],
            "code": i["f12"],
            "price": i.get("f2", 0),
            "change": i.get("f3", 0),
        } for i in data.get("data", {}).get("diff", [])]
    except Exception:
        return None


def fetch_concept_em():
    """東方財富概念板塊"""
    try:
        data = _fetch_em(
            "https://push2.eastmoney.com/api/qt/clist/get?"
            "pn=1&pz=30&po=1&np=1&fields=f2,f3,f4,f12,f14&fs=m:90+t:3"
        )
        return [{
            "name": i["f14"],
            "code": i["f12"],
            "change": i.get("f3", 0),
        } for i in data.get("data", {}).get("diff", [])]
    except Exception:
        return None


# ── 後端B：同花順 THS / akshare（降級） ──

def fetch_sector_ths():
    """THS 行業板塊（含漲跌家數、資金流向、領漲股）"""
    try:
        import akshare as ak
        df = ak.stock_board_industry_summary_ths()
        return [{
            "name": r["板块"],
            "code": "",
            "price": 0,
            "change": float(r.get("涨跌幅", 0)),
            "volume": float(r.get("总成交量", 0)),
            "turnover": float(r.get("总成交额", 0)),
            "net_inflow": float(r.get("净流入", 0)),
            "up_count": int(r.get("上涨家数", 0)),
            "down_count": int(r.get("下跌家数", 0)),
            "avg_price": float(r.get("均价", 0)),
            "lead_stock": r.get("领涨股", ""),
            "lead_stock_change": float(r.get("领涨股-涨跌幅", 0)),
        } for _, r in df.iterrows()]
    except Exception as e:
        print(f"THS行業失敗: {e}", flush=True)
        return []


def fetch_concept_ths():
    """THS 概念板塊（僅名稱，無實時漲跌）"""
    try:
        import akshare as ak
        df = ak.stock_board_concept_name_ths()
        return [{
            "name": r["name"],
            "code": str(r.get("code", "")),
            "change": 0,
        } for _, r in df.iterrows()]
    except Exception as e:
        print(f"THS概念失敗: {e}", flush=True)
        return []


# ── 輔助函數 ──

def get_market_mood(sectors):
    if not sectors:
        return "unknown"
    ratio = sum(1 for s in sectors if s.get("change", 0) > 0) / len(sectors)
    return "bullish" if ratio > 0.7 else "neutral" if ratio > 0.4 else "bearish"


def get_market_verdict(up_ratio, mood, sectors):
    """Return (verdict, reason) based on sector data."""
    if not sectors:
        return "unknown", "数据不足"
    if up_ratio < 25:
        return "弱势", f"仅{up_ratio}%板块上涨，{mood}"
    elif up_ratio < 40:
        return "偏弱", f"{up_ratio}%板块上涨，结构分化"
    elif up_ratio < 60:
        return "均衡", f"{up_ratio}%板块上涨，涨跌均衡"
    else:
        return "强势", f"{up_ratio}%板块上涨，整体走强"


def get_hot_sectors(sectors, top_n=3):
    """Return sectors with highest positive change as hot sectors."""
    hot = [s for s in sectors if s.get("change", 0) > 1.0]
    hot.sort(key=lambda s: s.get("change", 0), reverse=True)
    return [{
        "name": s["name"],
        "change": s.get("change", 0),
        "reason": f"板块涨{s.get('change',0):.1f}%"
    } for s in hot[:top_n]]


def get_danger_sectors(sectors, top_n=3):
    """Return sectors with lowest (negative) change as danger sectors."""
    danger = [s for s in sectors if s.get("change", 0) < -1.0]
    danger.sort(key=lambda s: s.get("change", 0))
    return [{
        "name": s["name"],
        "change": s.get("change", 0),
        "reason": f"板块跌{s.get('change',0):.1f}%"
    } for s in danger[:top_n]]


# ── 主流程 ──

def main():
    # 行業板塊：EM → THS → 兜底
    sectors = fetch_sector_em()
    source = "eastmoney"
    if sectors is None:
        sectors = fetch_sector_ths()
        source = "ths"

    # 概念板塊：EM → THS → 空
    concepts = fetch_concept_em()
    concept_source = "eastmoney"
    if concepts is None:
        concepts = fetch_concept_ths()
        concept_source = "ths"
    if not concepts:
        concepts = []
        concept_source = "unavailable"

    # 排序
    sorted_sectors = sorted(sectors, key=lambda s: s.get("change", 0), reverse=True)
    top_gainers = [s for s in sorted_sectors if s.get("change", 0) > 0][:5]
    top_losers = [s for s in reversed(sorted_sectors) if s.get("change", 0) < 0][:3]

    # 计算大盘数据
    up_ratio = round(
        sum(1 for s in sectors if s.get("change", 0) > 0) / max(len(sectors), 1) * 100, 1
    )
    mood = get_market_mood(sectors)
    verdict, verdict_reason = get_market_verdict(up_ratio, mood, sectors)

    market_data = {
        "timestamp": datetime.now().strftime("%Y-%m-%d %H:%M"),
        "source": source,
        "concept_source": concept_source,
        "total_sectors": len(sectors),
        "up_ratio": up_ratio,
        "mood": mood,
        "market_verdict": verdict,
        "verdict_reason": verdict_reason,
        "hot_sectors": get_hot_sectors(sectors),
        "danger_sectors": get_danger_sectors(sectors),
        "top_gainers": top_gainers,
        "top_losers": top_losers,
        "sectors": sectors,
        "concepts": concepts,
    }

    DATA_DIR.mkdir(parents=True, exist_ok=True)
    with open(DATA_DIR / "market.json", "w", encoding="utf-8") as f:
        json.dump(market_data, f, ensure_ascii=False, indent=2)

    # ── SQLite 双写 ──
    conn = get_conn()
    init_all_tables(conn)
    ok, msg, sid = write_market_snapshot(conn, market_data)
    if ok:
        print(f"[DB] {msg}", flush=True)
    else:
        print(f"[DB] 写入失败（JSON 不受影响）: {msg}", flush=True)
    conn.close()

    # 靜默：只寫文件，不輸出到stdout，避免cron推送


if __name__ == "__main__":
    main()