Files
python/daily.py
T

165 lines
4.1 KiB
Python

import tushare as ts
import pymysql
import pandas as pd
from datetime import datetime
import time
import argparse
from stock_config import DB_CONFIG, TUSHARE_TOKEN
# ===== 初始化 =====
ts.set_token(TUSHARE_TOKEN)
pro = ts.pro_api()
conn = None
cursor = None
def get_connection():
global conn, cursor
if conn is None or not conn.open:
conn = pymysql.connect(**DB_CONFIG)
cursor = conn.cursor()
return conn, cursor
# ===== 建表 =====
def create_tables():
db_conn, db_cursor = get_connection()
sql = """
CREATE TABLE IF NOT EXISTS stock_daily (
id BIGINT AUTO_INCREMENT PRIMARY KEY,
ts_code VARCHAR(20) NOT NULL,
symbol VARCHAR(10) NOT NULL,
trade_date DATE NOT NULL,
open DECIMAL(16,4),
high DECIMAL(16,4),
low DECIMAL(16,4),
close DECIMAL(16,4),
pre_close DECIMAL(16,4),
change_val DECIMAL(16,4),
pct_chg DECIMAL(10,4),
vol BIGINT,
amount BIGINT,
UNIQUE KEY uniq_code_date (ts_code, trade_date)
);
"""
db_cursor.execute(sql)
db_conn.commit()
print("表检查/创建完成")
# ===== 日期转换 =====
def str_to_date(date_str):
return datetime.strptime(date_str, "%Y%m%d").date()
def mysql_value(value):
if pd.isna(value):
return None
if hasattr(value, "item"):
return value.item()
return value
# ===== 获取交易日 =====
def get_trade_dates(start_date, end_date):
_, db_cursor = get_connection()
sql = """
SELECT cal_date
FROM trade_calendar
WHERE cal_date BETWEEN %s AND %s
AND is_open = 1
ORDER BY cal_date
"""
db_cursor.execute(sql, (str_to_date(start_date), str_to_date(end_date)))
results = db_cursor.fetchall()
return [row[0].strftime("%Y%m%d") for row in results]
# ===== 插入数据 =====
def insert_data(df):
db_conn, db_cursor = get_connection()
if df is None or df.empty:
return 0, 0
sql = """
INSERT IGNORE INTO stock_daily
(ts_code, symbol, trade_date, open, high, low, close, pre_close, change_val, pct_chg, vol, amount)
VALUES (%s,%s,%s,%s,%s,%s,%s,%s,%s,%s,%s,%s)
"""
data = []
for _, row in df.iterrows():
ts_code = row['ts_code']
symbol = ts_code.split('.')[0] # ✅ 只改这里
data.append((
ts_code,
symbol,
str_to_date(row['trade_date']), # 保持 DATE 类型
mysql_value(row['open']),
mysql_value(row['high']),
mysql_value(row['low']),
mysql_value(row['close']),
mysql_value(row['pre_close']),
mysql_value(row['change']),
mysql_value(row['pct_chg']),
mysql_value(row['vol']),
mysql_value(row['amount'])
))
db_cursor.executemany(sql, data)
db_conn.commit()
inserted = db_cursor.rowcount
skipped = len(data) - inserted
return inserted, skipped
# ===== 主逻辑 =====
def run(start_date, end_date):
create_tables()
trade_dates = get_trade_dates(start_date, end_date)
print(f"需要处理 {len(trade_dates)} 个交易日")
total_inserted = 0
total_skipped = 0
for trade_date in trade_dates:
print(f"\n处理交易日: {trade_date}")
try:
df = pro.daily(trade_date=trade_date)
inserted, skipped = insert_data(df)
total_inserted += inserted
total_skipped += skipped
print(f"新增: {inserted}, 跳过: {skipped}")
time.sleep(0.2)
except Exception as e:
print(f"失败: {trade_date}, 错误: {e}")
print("\n====== 汇总 ======")
print(f"总新增: {total_inserted}")
print(f"总跳过: {total_skipped}")
return True
# ===== 输入参数 =====
if __name__ == "__main__":
parser = argparse.ArgumentParser(description="同步日行情")
parser.add_argument("--start-date", default="20260513", help="开始日期,格式 YYYYMMDD")
parser.add_argument("--end-date", default="20260515", help="结束日期,格式 YYYYMMDD")
args = parser.parse_args()
run(args.start_date, args.end_date)