# ngrok ssh
https://dashboard.ngrok.com/cloud-edge/endpoints
https://learningsky.io/access-raspberry-pi-from-outside-network/
- error fix
./ngrok authtoken 1rCuDQZ1WuhfxxxxxxxxxxxxxxxxxxxxxxDjrNr1ANYwKLL3cEsxTE
- result
# screen command
`sudo apt-get install screen`下載screen
`screen -S ngrok_back`啟動一個名為ngrok_back的screen
`./ngrok tcp 22`啟動ngrok連線
`ctrl+a+d`暫時離開當前screen,使用此方法維持NGROK運作
`screen -ls`列出正在運作的screen
`screen -r 16100`進入ID為16100的sheel
link:
- http://mufeng-blog.logdown.com/posts/158421-linux-under-the-screen-instructions
- https://blog.csdn.net/x_zcl/article/details/50622201
# google sheet update url

https://www.learncodewithmike.com/2021/06/pandas-and-google-sheets.html
https://www.maxlist.xyz/2018/09/25/python_googlesheet_crud/
- example code

import subprocess
import time
import pyngrok
import gspread
from oauth2client.service_account import ServiceAccountCredentials

# 啟動ngrok
ngrok_process = subprocess.Popen(['./path/to/ngrok', 'http', '5000'])  # 5000是你本地Web服務器的端口號
time.sleep(5)  # 等待ngrok啟動，可以根據需要調整等待時間

# 獲取ngrok公開的URL
ngrok_url = pyngrok.ngrok.get_ngrok_tunnel().public_url

# 連接到Google試算表
scope = ['https://spreadsheets.google.com/feeds', 'https://www.googleapis.com/auth/drive']
credentials = ServiceAccountCredentials.from_json_keyfile_name('path/to/credentials.json', scope)
client = gspread.authorize(credentials)
spreadsheet = client.open('你的試算表名稱')
worksheet = spreadsheet.sheet1

# 在試算表中寫入URL
worksheet.update_cell(1, 1, ngrok_url)

# 等待一段時間，確保URL已傳輸到Google試算表
time.sleep(5)

# 停止ngrok進程
ngrok_process.terminate()

- google sheet test code(done)

from google.oauth2.service_account import Credentials
import gspread
import pandas as pd
 
scope = ['https://www.googleapis.com/auth/spreadsheets']

creds = Credentials.from_service_account_file("robotfishngrokurl.json", scopes=scope)

gs = gspread.authorize(creds)
 
sheet = gs.open_by_url('https://docs.google.com/spreadsheets/d/1ae-90q2kxzyYXTdymz-ZNUurXK0H7FY1Gb4xCz2qfs0/edit?usp=sharing')
worksheet = sheet.get_worksheet(0)

df = pd.read_csv('hand_detect_output.csv')
# df.fillna(0)

worksheet.update([df.columns.values.tolist()] + df.values.tolist())