ARROW-12869: [R] Bindings for utf8_reverse and ascii_reverse

This adds tests of the ascii_reverse kernel and a binding and tests for the stri_reverse function which calls the utf8_reverse kernel Closes apache#10589 from thisisnic/ARROW-12869_str_reverse Lead-authored-by: Nic Crane <[email protected]> Co-authored-by: Ian Cook <[email protected]> Signed-off-by: Ian Cook <[email protected]>
davidhao3300 · Jun 24, 2021 · c4a20e9 · c4a20e9
1 parent 5275e72
commit c4a20e9
Show file tree

Hide file tree

Showing 3 changed files with 41 additions and 3 deletions.
diff --git a/r/DESCRIPTION b/r/DESCRIPTION
@@ -48,6 +48,7 @@ Suggests:
     pkgload,
     reticulate,
     rmarkdown,
+    stringi,
     stringr,
     testthat,
     tibble,

diff --git a/r/R/expression.R b/r/R/expression.R
@@ -30,6 +30,7 @@
   "str_length" = "utf8_length",
   "str_to_lower" = "utf8_lower",
   "str_to_upper" = "utf8_upper",
+  "str_reverse" = "utf8_reverse",
   # str_trim is defined in dplyr-functions.R
   "year" = "year",
   "isoyear" = "iso_year",

diff --git a/r/tests/testthat/test-dplyr-string-functions.R b/r/tests/testthat/test-dplyr-string-functions.R
@@ -20,6 +20,7 @@ skip_if_not_available("utf8proc")
 
 library(dplyr)
 library(stringr)
+library(stringi)
 
 test_that("paste, paste0, and str_c", {
   df <- tibble(
@@ -712,7 +713,6 @@ test_that("strptime", {
     tstamp,
     check.tzone = FALSE
   )
-
 })
 
 test_that("errors in strptime", {
@@ -725,6 +725,43 @@ test_that("errors in strptime", {
   )
 })
 
+test_that("stri_reverse and arrow_ascii_reverse functions", {
+
+  df_ascii <- tibble(x = c("Foo\nand bar", "baz\tand qux and quux"))
+
+  df_utf8 <- tibble(x = c("Foo\u00A0\u0061nd\u00A0bar", "\u0062az\u00A0and\u00A0qux\u3000and\u00A0quux"))
+
+  expect_dplyr_equal(
+    input %>%
+      mutate(x = stri_reverse(x)) %>%
+      collect(),
+    df_utf8
+  )
+
+  expect_dplyr_equal(
+    input %>%
+      mutate(x = stri_reverse(x)) %>%
+      collect(),
+    df_ascii
+  )
+
+  expect_equivalent(
+    df_ascii %>%
+      Table$create() %>%
+      mutate(x = arrow_ascii_reverse(x)) %>%
+      collect(),
+    tibble(x = c("rab dna\nooF", "xuuq dna xuq dna\tzab"))
+  )
+
+  expect_error(
+    df_utf8 %>%
+      Table$create() %>%
+      mutate(x = arrow_ascii_reverse(x)) %>%
+      collect(),
+    "Invalid: Non-ASCII sequence in input"
+  )
+})
+
 test_that("str_like", {
 
   df <- tibble(x = c("Foo and bar", "baz and qux and quux"))
@@ -783,7 +820,6 @@ test_that("str_like", {
     input %>%
       mutate(x = str_like(x, "%baz%")) %>%
       collect(),
-    df,
+    df
   )
-
 })