twitterBackend.r

getTweets <- function(hashtag, sampleSize) {
  # load credentials from the json file
  credentials <- fromJSON(file="apikey.json", simplify = TRUE)
  
  # create an oauth token from the loaded credentials
  oauthToken <- create_token(app = as.character(credentials["app_name"]), 
                             consumer_key = as.character(credentials["api_key"]),
                             consumer_secret = as.character(credentials["api_key_secret"]),
                             access_token = as.character(credentials["access_token"]),
                             access_secret = as.character(credentials["access_token_secret"]))
  
  df <- search_tweets(paste("#", hashtag, sep = ""), n=sampleSize)
  df$cleanText <- as.character(df$text)
  df$cleanText <- gsub("(ftp|http(s)*)://.*", " ", df$cleanText) #remove urls
  df$cleanText <- removePunctuation(df$cleanText)
  df$cleanText <- replace_internet_slang(df$cleanText)
  df$cleanText <- replace_emoji(df$cleanText)
  df$cleanText <- replace_contraction(df$cleanText)
  df$cleanText <- removeWords(df$cleanText, stopwords())
  df$cleanText <- gsub(pattern="\\d", replace=" ", df$cleanText) # remove all the numbers
  df$cleanText <- gsub("[[:punct:]]", " ", df$cleanText) # remove all the special characters
  df$cleanText <- gsub(pattern="\\b[a-z]\\b{1}", "", df$cleanText) # remove all single chars
  df$cleanText <- tolower(df$cleanText)
  df$cleanText <- stripWhitespace(df$cleanText)
  df$cleanText <- removeWords(df$cleanText, c(hashtag))
  return(df)
}