Remote server authentication to read Googlesheets from R script not using service accounts?

问题

I'm automating a process in R that pulls, pushes, analyses & stores data in a certain way and all in all, juggles data between Googlesheets, Postgresql, and salesforce. Till now, I had it running in my system but I will need to run the scripts from my remote login ssh server on aws.

The thing is though, I use 'googlesheets' package in R to auth & read google sheets but that needs an interactive environment to register/setup. I've read other answers on SO about this issue but they all suggest opening a google service account.

However, I dont want to do this because all I need is to read those google sheets which by itself is free. I have no GCP, google compute, bigquery or any of those right now and I certainly cant get pay for anything now.

Is there a way to read google sheets from a non interactive environment (linux terminal remote login interface)? I'm open to trying other libraries or any other hack that you may have.

Anything guys?

回答1:

I was struggling with something similar for many hours, so even if this an old question, I think some other people might find this solution useful. My problem was the authentication in a non interactive way to read and modify a google sheet through a Shiny App. The app use to always trigger the dance auth process no matter if I would have saved it in a cache folder embedded in the app.

I have used the following guides and questions to guide myself through the process:

https://cran.r-project.org/web/packages/gargle/vignettes/non-interactive-auth.html
https://community.rstudio.com/t/shinyapps-io-non-interactive-auth-fails-for-googledrive/46504
https://docs.rstudio.com/shinyapps.io/getting-started.html#deploying-applications

Try the following reproducible example:

library("googledrive")
library("googlesheets4") # I am using the developing version 0.1.0.9000
library("shiny")

# You want to deploy an app in Shinyapps.io or other server
# FIRST STEP----
# Get the token an store it in a cache folder embedded in your app directory
# designate project-specific cache
options(gargle_oauth_cache = ".secrets")
# options(gargle_quiet = FALSE) # So you can know what is happening
# Authenticate in interactive mode (run the app) ONCE and check if the token 
# has been stored inside the .secrets folder, after that just comment this line
#drive_auth() # Authenticate to produce the token in the cache folder
# Grant permission to googlesheets to access to the token produced
#sheets_auth(token = drive_token())

# SECOND STEP----
# Comment lines 10, 13 and 15 and uncomment lines 21 and 22
# You tell gargle to search the token in the secrets folder and to look
# for an auth given to a certain email (enter your email linked to googledrive!)
drive_auth(cache = ".secrets", email = "enter_your_email@here")
sheets_auth(token = drive_token())

# THIRD STEP---
# Now you can deploy your app in shinyapps.io!!
# Test if your app runs properly in the local version
# Authenticate in ShinyApps.io
# rsconnect::setAccountInfo(name="<ACCOUNT>", token="<TOKEN>", secret="<SECRET>")
# setwd() in your App directory
# library(rsconnect)
# deployApp()
# Enjoy your new App!!

ui <- # Define UI for application that plots random distributions 
  fluidPage(

    # Application title
    titlePanel("Hello Shiny!"),

    # Sidebar with a slider input for number of observations
    sidebarLayout(
      sidebarPanel(
        sliderInput("obs", 
                    "Number of observations:", 
                    min = 1, 
                    max = 1000, 
                    value = 500),
        actionButton(
          "add",
          "Add new entry")
      ),

      # Show a plot of the generated distribution
      mainPanel(
        "Check your googlesheet!!"
      )
    )
  )

server <- function(input, output, session) {
  # Expression that generates a plot of the distribution. The expression
  # is wrapped in a call to renderPlot to indicate that:
  #
  #  1) It is "reactive" and therefore should be automatically 
  #     re-executed when inputs change
  #  2) Its output type is a plot 
  #
  observeEvent(input$add, {
    # You should have or create a googlesheets through google drive with
    # the name "example_shiny"
    wb <- drive_get("example_shiny")
    dt <- read_sheet(wb)
    new_entry <-
      data.frame(ID = tail(dt$ID, 1) + 1, NAME = "new",
                 OBSERVATION = input$obs)
    sheets_append(new_entry, wb)
  })
}

shinyApp(ui, server)

回答2:

I think this bit of documentation from Rstudio shows how to do it, if I understand your question: https://support.rstudio.com/hc/en-us/articles/217952868-Generating-OAuth-tokens-from-a-server

来源：https://stackoverflow.com/questions/44980757/remote-server-authentication-to-read-googlesheets-from-r-script-not-using-servic

标签

remote-access

interactive

google-sheets-api

r-googlesheets