extend get_bunch_o_tweets to apply to all three datasets and ad an n_…

…tweets == 'all' option

R/format_submissions.R

@@ -5,21 +5,21 @@
 #'
 #' @export
 format_submissions <- function(subs, tips) {
+  maxid <- max(tips$id, na.rm = T)
 
- maxid <- max(tips$id, na.rm = T)
+  x <- subs %>%
+    mutate(
+      id = seq(maxid + 1, maxid + nrow(subs)),
+      Tip = `Suggested Tweet`,
+      Author = `Your Name or Twitter Handle`,
+      Last.Sent = "",
+      Category = "Uncategorized"
+    ) %>%
+    select(id, Tip, Author, Last.Sent, Category)
 
- x <- subs %>%
-  mutate(id = seq(maxid+1, maxid+nrow(subs)),
-         Tip = `Suggested Tweet`, 
-         Author = `Your Name or Twitter Handle`, 
-         Last.Sent = "",
-         Category = "Uncategorized") %>% 
-  select(id, Tip, Author, Last.Sent, Category) 
-
- colnames(x) <- NULL
-
- readr::write_csv(x, file("clipboard"))
-
+  colnames(x) <- NULL
+
+  readr::write_csv(x, file("clipboard"))
 }
 
 utils::globalVariables(c("Suggested Tweet", "Your Name or Twitter Handle", "id", "Tip", "Author", "Last.Sent", "Category"))

R/get_tweets.R

@@ -27,15 +27,18 @@ save_tweet_number <- function(val) {
 
 #' Get Tweets
 #'
+#' @param n_tweets_to_grab Number of tweets to grab, or "all"
 #' @param save_number Whether to save the numbers of the tweets or not
 #' @export
-get_tweets <- function(save_number = TRUE) {
-  n_tweets_to_grab <-
-    get_tweet_number()
+get_tweets <- function(save_number = TRUE, n_tweets_to_grab = "all") {
+  if (n_tweets_to_grab == "all") {
+    n_tweets_to_grab <-
+      get_tweet_number()
+  }
 
   tbl <- rtweet::get_timeline("RLangTip", n = n_tweets_to_grab)
 
-  if (save_number) {
+  if (save_number && n_tweets_to_grab == "all") {
     n_tweets <- nrow(tbl)
     save_tweet_number(n_tweets)
   }

R/rtip.R

@@ -20,29 +20,29 @@ read_tips <- function() {
 #' @param cowsay if TRUE, a random cowsay animal will present the tip
 #' @param color if TRUE, use a colorful cowsay display. Ignored if cowsay=FALSE
 #' @param excluded categories excluded for random tip selection
-#' @param keyword character vector keywords contained in tips to search for. 
+#' @param keyword character vector keywords contained in tips to search for.
 #'
 #' @return the tip, as a 3-element vector (tip number and category; tip text; author and date)
 #' @export
-rtip <- function(id, cowsay = TRUE, color = FALSE, 
+rtip <- function(id, cowsay = TRUE, color = FALSE,
                  excluded = c("deprecated", "Uncategorized"),
                  keyword = NULL) {
   ## Print a random tweet from tips.csv
   tips <- read_tips()
-  
+
   if (!is.null(keyword)) {
     if (!is.character(keyword)) stop("keyword must be of class character.")
-    
-    keyword <- keyword %>% 
-      unique() %>% 
+
+    keyword <- keyword %>%
+      unique() %>%
       tolower()
-    
+
     keyword <- stringr::str_c(keyword, collapse = "|")
-    
-    tips <- tips %>% 
+
+    tips <- tips %>%
       dplyr::filter(str_detect(tolower(Tip), keyword))
   }
-  
+
   N <- NROW(tips)
   if (missing(id)) {
     candidates <- (1:N)[!(tips$Category %in% excluded)]
@@ -56,7 +56,7 @@ rtip <- function(id, cowsay = TRUE, color = FALSE,
 
   ### Word-wrap the tip to fit the terminal
   wrappedTip <- strwrap(tiprow$Tip)
-    
+
   display <- c(
     paste0("Tip #", tiprow$id, " in category ", tiprow$Category, sep = ""),
     wrappedTip,
@@ -79,18 +79,17 @@ rtip <- function(id, cowsay = TRUE, color = FALSE,
 
     who <- sample(who_pool, 1)
 
-    display_cat <- display %>% paste(collapse = "\n")
+    display <- display %>% paste(collapse = "\n")
 
     if (color) {
-      cowsay::say(display_cat, by_color = "rainbow", by = who, type = "string") %>%
+      cowsay::say(display, by_color = "rainbow", by = who, type = "string") %>%
         cat()
     } else {
-      cowsay::say(display_cat, by = who, type = "string") %>%
+      cowsay::say(display, by = who, type = "string") %>%
         cat()
     }
   } else {
     cat(display, sep = "\n")
-
   }
 
   invisible(display)

R/run_pipleine.R

@@ -14,20 +14,44 @@ save_tweets <- function(tbl) {
 #' @export
 run_tweet_pipeline <- function() {
   get_tweets() %>%
-   join_tips() %>%
-   save_tweets()
+    join_tips() %>%
+    save_tweets()
 }
 
 
 #' Get a bunch of tweets
 #'
-#' @param n_tweets Number of tweets to choose
-#' @param weighted Should tweets with higher scores be chosen more frequently?
+#' @param from Which data source should be used. "canonical" for tips CSV, "twitter" for full Twitter data, "joined" for canonical joined on Twitter data and deduped
+#' @param n_tweets Number of tweets to choose, or "all"
+#' @param weighted Should tweets with higher scores be chosen more frequently? Will be false if from is "canonical"
 #' @export
 #' @importFrom dplyr sample_n
-get_bunch_o_tweets <- function(n_tweets = 5, weighted = TRUE) {
- joined <- readr::read_csv(here::here("data-raw", "joined.csv"))
-
- joined %>% 
-  dplyr::sample_n(n_tweets, weight = score)
-} 
+get_bunch_o_tweets <- function(from = "joined", n_tweets = "all", weighted = TRUE) {
+  stopifnot(from %in% c("canonical", "twitter", "joined"))
+  stopifnot(is.numeric(n_tweets) || n_tweets == "all")
+
+  suppressMessages({
+    if (from == "twitter") {
+      tbl <- get_tweets(save_number = FALSE, n_tweets_to_grab = n_tweets) %>%
+        score_tweets()
+    } else if (from == "joined") {
+      tbl <- readr::read_csv(joined_path)
+    } else if (from == "canonical") {
+      tbl <- readr::read_csv(tips_path)
+    }
+  })
+
+  if (n_tweets == "all") {
+    return(tbl)
+  }
+
+  if (from %in% c("joined", "twitter") && weighted) {
+    tbl <- tbl %>%
+      dplyr::sample_n(n_tweets, weight = score)
+  } else if (from %in% c("joined", "twitter") && !weighted || from == "canonical") {
+    tbl <- tbl %>%
+      dplyr::sample_n(n_tweets)
+  }
+
+  tbl
+}

R/zzz.R

@@ -1,5 +1,6 @@
 n_tweets_path <- here::here("n_tweets.txt")
-tips_path <- here::here("tips.csv")
+tips_path <- here::here("inst", "extdata", "tips.csv")
+joined_path <- here::here("data-raw", "joined.csv")
 
 utils::globalVariables(c("consumer_key", "consumer_secret", "access_token", "access_secret"))
 save_twitter_token <- function() {

tests/testthat/test-sample_tweet.R

@@ -8,10 +8,13 @@ testthat::test_that("rtip produces string", {
 
 testthat::test_that("rtip produces expected string selected by id", {
   tip <- rtip(id = 1, cowsay = FALSE)
-  expect_equal(tip,
-    c("Tip #1 in category statistics",
+  expect_equal(
+    tip,
+    c(
+      "Tip #1 in category statistics",
       "Use as.dendrogram() to prepare hierarchical clusters or trees for plotting #rstats http://bit.ly/1sshY0I",
-      "      -- @RStudioJoe, 2010-06-09")
+      "      -- @RStudioJoe, 2010-06-09"
+    )
   )
 })
 
@@ -22,5 +25,5 @@ testthat::test_that("rtip produces an error if id doesn't exist", {
 })
 
 testthat::test_that("rtip is expected length", {
- expect_length(rtip(), 3)
+  expect_length(rtip(), 3)
 })