Tools to Make a Stack Overflow Twitter Bot

The aim of this package is to allow for easy creation of Stack Overflow twitter bots. It provides a wrapper around functionality from rtweet and stackr. It was inspired by the tidyverse tweets bot by Jake Thompson, from which large amounts of code has been adapted. For a working example of a bot built with this package see h2o tweets bot, or here for the code.

Installation

Install the development version from GitHub:

# install.packages("devtools")
devtools::install_github("seabbs/StackTweetBot")

To schedule the bot you will either need to install cronR (if on Linux or Mac) with,

install.packages("cronR")

or taskscheduler (if on windows) with,

install.packages("taskscheduleR")

Quick start

  1. Load the package
library(StackTweetBot)

  1. Set up your bots Twitter account - making it clear it is a bot!

  2. Set up access to the Twitter API with,

add_twitter_api()
#> A Twitter token is required to add posts to your twitter account. 
#> The first step is to create a twitter app, enter the following code for instructions. 
#> 
#> library(rtweet); vignette("auth") 
#> 
#> Follow the instructions to create a Twitter app. 
#> Copy the Consumer key and secret and create a token with the following code: 
#> 
#> twitter_token <- rtweet::create_token(app = "your_appname_here",
#>                                             consumer_key = "key_here",
#>                                             consumer_secret = "secret_here") 
#> 
#> Note if setting up the bot on a remote server do this locally and then upload 
#> 
#> Now save this token to your home directory with the following code: 
#> 
#> saveRDS(twitter_token, file = "~/twitter_token.rds"") 
#> 
#> Now save the token as an enviroment variable: 
#> 
#> Sys.setenv(TWITTER_PAT = "~/twitter_token.rds") 
#> 
#> If using Rstudio server you made need to manually add the TWITTER_PAT to the .Renviron.
  1. Set up access to the Stack Overflow API. This is required to increase your daily rate limit from 300 to 10,000.
add_stack_api()
#> An API key is required to increase your daily rate limit from 300 to 10,000. 
#> The first step is to sign up for an API key here: https://stackapps.com/apps/oauth/register 
#> Now add it as a enviroment variable using the following code (replacing *** with your API key): 
#> 
#> Sys.setenv(STACK_EXCHANGE_KEY = "***") 
#> 
#> If using Rstudio server you may need to manually add the STACK_EXCHANGE_KEY to the .Renviron.
  1. Check you have correctly set up your API access (restart your r session first).
add_twitter_api()
add_stack_api()
  1. Set up and schedule the twitter bot, settting the tags to look for and to exclude. Schedule the bot using schedule = TRUE, with the arguements to use to define the update time being platform dependent, see ?set_up_stack_tweet_bot for details. To allow posting set post = TRUE, and to save the bot to a permanent directory (so that it will run after your R session has ended) set the directory with dir.
set_up_stack_tweet_bot(extracted_tags = "ggplot2",
                       excluded_tags = "python",
                       time_window = 60,
                       add_process_fn = NULL, 
                       hashtags = "rstats",
                       run = TRUE, 
                       schedule = FALSE,
                       save = TRUE, 
                       post = FALSE,
                       dir = NULL,
                       verbose = TRUE,
                       frequency = "hourly")
#> No directory has been supplied for saving the twitter bot,
#>            defaulting to saving to the temporary directory. This directory will not be
#>              preserved once the r session has ended.
#> Saving the following code to: /tmp/Rtmp3Pi6J8/stack_tweet_bot.R
#> Code: 
#> 
#> library(StackTweetBot);
#> 
#> message('Getting Stack Overflow questions at ', Sys.time());
#> 
#> questions <- get_stack_questions(extracted_tags = 'ggplot2',
#>                                  excluded_tags = 'python',
#>                                  time_window = 60,
#>                                  add_process_fn = NULL);
#> 
#> message('Extracted ', length(questions$title), ' questions');
#> 
#> message('Posting questions as tweets');
#> 
#> posts <- post_stack_tweets(questions, hashtags = 'rstats',
#>                            post = FALSE);
#> Running bot..
#> Getting Stack Overflow questions at 2018-10-24 15:58:27
#> Extracted 3 questions
#> Posting questions as tweets
#> library(StackTweetBot);
#> 
#> message('Getting Stack Overflow questions at ', Sys.time());
#> 
#> questions <- get_stack_questions(extracted_tags = 'ggplot2',
#>                                  excluded_tags = 'python',
#>                                  time_window = 60,
#>                                  add_process_fn = NULL);
#> 
#> message('Extracted ', length(questions$title), ' questions');
#> 
#> message('Posting questions as tweets');
#> 
#> posts <- post_stack_tweets(questions, hashtags = 'rstats',
#>                            post = FALSE);
  1. Check your bot is running.
## If using cronR on mac or linux
cronR::cron_ls()
## If using taskscheduleR on windows
taskscheduleR::taskscheduler_ls()

Additional Functionality

File an issue here if there is a feature, that you think is missing from the package, or better yet submit a pull request!

Docker

This package has been developed in docker based on the rocker/tidyverse image, to access the development environment enter the following at the command line (with an active docker daemon running),

docker pull seabbs/stacktweetbot
docker run -d -p 8787:8787 -p 1410:1410 -e USER=StackTweetBot -e PASSWORD=StackTweetBot --name StackTweetBot seabbs/stacktweetbot

The rstudio client can be accessed on port 8787 at localhost (or your machines ip). The default username isStackTweetBot and the default password is StackTweetBot. The docker file comes reloaded with cronR and therefore may also be used as a staging environment for your twitter bot.