import requests
import time
import pandas as pd
if __name__ == '__main__':
for j in range(1, 49):
url = f'http://fund.eastmoney.com/Data/Fund_JJJZ_Data.aspx?t=1&lx=1&letter=&gsid=&text=&sort=zdf,desc&page={j},200&dt=1597126258333&atfc=&onlySale=0'
headers = {
'User-Agent': 'Mozilla/5.0 (Windows NT 6.1; WOW64) AppleWebKit/537.1 (KHTML, like Gecko) Chrome/21.0.1180.71 Safari/537.1 LBBROWSER'
}
resp = requests.get(url, headers=headers).text
str_ = resp[102:]
list1 = eval(str_.split(",count")[0])
print(f'正在爬取第{j}頁')
print(f'本頁爬取{len(list1)}條資料')
num = []
name = []
today_price = []
yesterday_price = []
day_value = []
day_value_rate = []
subscription_status = []
redemption_status = []
service_charge = []
for i in range(len(list1)):
# 1、基金程式碼號
num.append(list1[i][0])
# 2、股票名稱
name.append(list1[i][1])
# 3、今日基金淨額
today_price.append(list1[i][3])
# 4、昨日基金淨額
yesterday_price.append(list1[i][5])
# 5、日增長值
day_value.append(list1[i][7])
# 6、日增長率
day_value_rate.append(list1[i][8])
# 7、申購狀態
subscription_status.append(list1[i][9])
# 8、贖回狀態
redemption_status.append(list1[i][10])
# 9、手續費
service_charge.append(list1[i][17])
df = pd.DataFrame()
df['基金程式碼'] = num
df['基金名稱'] = name
df['2020-08-12\n單位淨值'] = today_price
df['2020-08-11\n單位淨值'] = yesterday_price
df['日增長值'] = day_value
df['日增長率\n%'] = day_value_rate
df['申購狀態'] = subscription_status
df['贖回狀態'] = redemption_status
df['手續費'] = service_charge
try:
df.to_excel(f'基金{j}.xlsx', '基金資訊', index=None, encoding='utf-8')
except Exception as e:
print(e)
time.sleep(1)