#!/usr/bin/env python3
"""
补充下载缺失的 510300.SSE 沪深300ETF 日线数据
"""

import akshare as ak
import pandas as pd
import os
from pathlib import Path

# 配置
BASE_DIR = Path("/Users/chufeng/nas/stock-data/sanguo_quant_live/zhaoyun-data/data/raw/daily")
BASE_DIR.mkdir(parents=True, exist_ok=True)

# 下载沪深300ETF 510300 - 指数ETF，在上海交易所，代码格式: sh510300
symbol = "sh510300"
code = "510300"

print(f"🚀 开始下载 {symbol} 日线数据...")

try:
    # 尝试 fund_etf_hist_sina - 新浪ETF历史行情，不需要start_date/end_date，直接全部下载
    print("🔍 尝试 fund_etf_hist_sina 接口...")
    df = ak.fund_etf_hist_sina(symbol=symbol)
    
    print(f"✅ fund_etf_hist_sina: {len(df)} 行")
    print(f"📋 列名: {list(df.columns)}")
    
    if df.empty:
        # 如果不行，试试 fund_etf_hist_em
        print("\n🔍 尝试 fund_etf_hist_em 接口...")
        df = ak.fund_etf_hist_em(symbol=symbol, period="daily", start_date="20100101", end_date="20260330", adjust="hfq")
        
        print(f"✅ fund_etf_hist_em: {len(df)} 行")
        print(f"📋 列名: {list(df.columns)}")
    
    # 格式化日期
    if "日期" in df.columns:
        df["trade_date"] = pd.to_datetime(df["日期"])
    elif "date" in df.columns:
        df["trade_date"] = pd.to_datetime(df["date"])
    else:
        df["trade_date"] = pd.to_datetime(df.index)
    
    # 格式化列名符合vnpy要求
    column_mapping = {
        "开盘": "open",
        "最高": "high",
        "最低": "low",
        "收盘": "close",
        "成交量": "volume",
        "成交额": "amount",
        "open": "open",
        "high": "high",
        "low": "low",
        "close": "close",
        "volume": "volume",
        "amount": "amount",
    }
    
    df = df.rename(columns=column_mapping)
    
    # 过滤日期 >= 2010-01-01
    df = df[df["trade_date"] >= pd.to_datetime("2010-01-01")]
    
    # 检查列
    required_columns = ["trade_date", "open", "high", "low", "close", "volume", "amount"]
    for col in required_columns:
        if col not in df.columns:
            print(f"⚠️ 缺失列: {col}")
    
    # 保存
    output_file = BASE_DIR / f"{symbol}_daily.parquet"
    df.to_parquet(output_file, compression="snappy", index=False)
    
    if not df.empty:
        print(f"\n✅ {symbol}: 下载成功，{len(df)} 条记录")
        print(f"📦 保存到: {output_file}")
        print(f"📊 数据日期范围: {df['trade_date'].min()} → {df['trade_date'].max()}")
    else:
        print(f"\n❌ {symbol}: 数据仍然为空")
    
except Exception as e:
    import traceback
    traceback.print_exc()
    print(f"\n❌ {symbol}: 下载失败 - {str(e)}")