聚寬是個非常棒的平臺，回測用的數據洗的非常乾淨，沒有未來信息的問題。最近聽說他們發新產品了，祝賀!
如果經常用聚寬的話，會經常需要把聚寬研究環境中自己分析數據的結果下載下來，我做了個小工具，能批量自動下載聚寬上保存的文件。
用到了selenium模擬瀏覽器，不會用的可以搜一下。
github：https://github.com/QuantGin/Quant_For_All
from selenium.common.exceptions import *
import requests
import time
import os
import traceback
class joinquant:
def __init__(self,d):
self.d=d
def login(self,username,password):
d=self.d
d.get('https://www.joinquant.com/research')
d1=d.find_element_by_class_name('phone')
d1.clear()
d1.send_keys(username)
d1=d.find_element_by_class_name('jq-login__password')
d1.clear()
d1.send_keys(password)
self.wait()
d.find_element_by_class_name('login-submit').click()
for i in range(60):
try:
d.switch_to.default_content()
d.switch_to.frame("research")
d.find_element_by_class_name('fa-folder')
if d.current_url=='https://www.joinquant.com/default/research/index?target=self&url=/default/research/redirect':
self._get_cookie()
print('Login was successful!')
self.wait(3)
return 1
except Exception as e:
pass
self.wait()
return 0
def wait(self,n=1):
time.sleep(1*n)
def _get_cookie(self):
d=self.d
cookie =[item["name"] + ":" + item["value"] for item in d.get_cookies()]
cookiestr = ';'.join(item for item in cookie)
cook_map = {}
for item in cookie :
str = item.split(':')
cook_map[str[0]] = str[1]
cookies = requests.utils.cookiejar_from_dict(cook_map, cookiejar=None, overwrite=True)
self.cookies=cookies
def _research(self):
d=self.d
d.switch_to.default_content()
d.execute_script("document.getElementById('kk_nav').style.display='none'")
d.switch_to.frame("research")
def download(self,url,file_dir):
file=file_dir+url.split('/')[-1]
if os.path.exists(file):
print('Download',file)
return 1
else:
print('Downloading',file)
self._get_cookie()
r=requests.get(url,cookies=self.cookies)
fout=open(file,'wb')
fout.write(r.content)
fout.close
return 1
#輸出當前路徑
@property
def pwd(self):
d=self.d
self._research()
d1=d.find_element_by_class_name('breadcrumb')
s=d1.find_elements_by_xpath(".//*")[-2].get_attribute('innerHTML')
print(s.split('"')[1].split('/tree')[-1])
return s.split('"')[1].split('/tree')[-1]
@property
def ls(self):
d=self.d
self._research()
d1=d.find_element_by_id('notebook_list')
l=[key.get_attribute('innerHTML') for key in d1.find_elements_by_class_name('item_name')]
ll=[]
for key in l:
if key!='..':
ll.append(key)
print(ll[:3],ll[-3:])
return ll
def _clear(self):
d=self.d
self._research()
for i in range(3):
try:
d.find_element_by_class_name('modal-content').find_element_by_class_name('btn,btn-default,btn-sm,btn-primary').click()
break
except:
pass
def click(self,name):
d=self.d
self._research()
d1=d.find_element_by_id('notebook_list')
d1.find_element_by_link_text(name).click()
print(name,'clicked')
def home(self):
d=self.d
self._research()
for i in range(3):
self._research()
try:
for ii in range(10):
self._research()
d.find_element_by_class_name('fa-folder').click()
self.wait()
if self.pwd=='':
print('cd /')
return 1
except:
d.get('https://www.joinquant.com/research')
self.wait()
return 0
def cd(self,s_dir):
d=self.d
for ii in range(3):
try:
self.home()
_dir=''
if len(s_dir)>0:
for s in s_dir.split('/'):
if len(s)>0:
_dir+='/'+s
for i in range(10):
try:
self.click(s)
except:pass
if self.pwd==_dir:
break
self.wait()
if _dir==self.pwd:
print('cd',_dir)
return 1
except:pass
self.wait(3)
return 0
def get_url(self,name):
d=self.d
self._research()
d2=d.find_element_by_id('notebook_list')
for d1 in d2.find_elements_by_class_name('item_link'):
_name=d1.find_element_by_class_name('item_name').get_attribute('innerHTML')
if _name==name:
url=d1.get_attribute('href')
return url
def select(self,name):
d=self.d
self._research()
d2=d.find_element_by_id('notebook_list')
for d1 in d2.find_elements_by_class_name('col-md-12')[:]:
_name=d1.find_element_by_class_name('item_name').get_attribute('innerHTML')
if _name==name:
d1.find_element_by_class_name('item_icon,file_icon,icon-fixed-width').click()
print(name,'selected')
break
def click_delete(self):
d=self.d
self._research()
for i in range(10):
try:
d.find_element_by_class_name('fa-trash').click()
break
except:self.wait(1)
for i in range(10):
try:
d.find_element_by_class_name('modal-content').find_element_by_class_name('btn-danger').click()
except:self.wait(1)
self._clear()
self.wait(1)
def _download_delete(self,s_dir,file_dir):
d=self.d
self.cd(s_dir)
self.wait(10)
l=self.ls
while len(l)>0:
for name in l[:20]:
url=self.get_url(name).replace('/edit/','/files/')
name0=url.split('.')[-1]
if not name0 in ['ipynb','py'] and url.find('.')!=0 and url.find(s_dir)!=-1:
if self.download(url,file_dir):
self.select(name)
self.click_delete()
l=self.ls
return 1
def download_delete(self,s_dir,file_dir):
while 1:
try:
if self._download_delete(s_dir,file_dir):
return 1
except ElementClickInterceptedException:pass
except StaleElementReferenceException:pass
except ElementNotInteractableException:pass
except NoSuchElementException:pass
def _stop_run(self):
self._clear()
#停止並運行
d=self.d
self._research()
d.find_element_by_class_name('fa-stop,fa').click()
self.wait(5)
d.find_element_by_link_text('單元格').click()
self.wait()
d.find_element_by_link_text('運行所有').click()
self.wait()
def _click_run(self):
self._clear()
d=self.d
self._research()
self.wait()
d.find_element_by_link_text('單元格').click()
self.wait()
d.find_element_by_link_text('運行所有').click()
self.wait()
def run(self,name):
d=self.d
self._research()
self.click(name)
self.wait()
d.switch_to.window(d.window_handles[-1])
self._research()
self._stop_run()
for i in range(60):
self.wait()
d1=d.find_elements_by_class_name('output_subarea,output_text,output_result')
if len(d1)<1:
continue
d1=d1[0]
html=d1.get_attribute('innerHTML')
if html.find('!!success!!')!=-1:
d.close()
self.wait()
d.switch_to.window(d.window_handles[0])
self.wait()
return 1
errors=d.find_elements_by_class_name('output_subarea,output_text,output_error')
for error in errors:
if error.get_attribute('innerHTML').find('Traceback')!=-1:
error=error.get_attribute('innerHTML')
raise(ValueError(error))
def keep_running(self,name):
d=self.d
self._research()
self.click(name)
self.wait()
d.switch_to.window(d.window_handles[-1])
self._research()
while 1:
try:
self._click_run()
except:
print(traceback.format_exc())
print('click')
self.wait(10)
if __name__ == "__main__":
#配置虛擬瀏覽器
from selenium import webdriver
from selenium.common.exceptions import *
chrome_options = webdriver.ChromeOptions()
chrome_options.add_argument("--disable-popup-blocking")
# chrome_options.add_argument('--headless')
# chrome_options.add_argument('--no-sandbox')
d = webdriver.Chrome(options=chrome_options)
jq=joinquant(d)
# 登錄
jq.login('username','password')
# 進入指定路徑
jq.cd('price_daily')
# 列出路徑下文件
print(jq.ls)
# 保持指定notebook運行
self.keep_running('XXX.ipynb')
# 下載並在網站上刪除指定路徑下的所有文件XXX
jq.download_delete('XXX','../XXX/XXX/')