Refine string case conversion helpers (#593)

Fixes #592

Author: arnaudgallou

NEWS.md

@@ -10,7 +10,7 @@
 
 * New `vignette("locale-sensitive")` about locale sensitive functions (@kylieainslie, #404)
 * New `str_ilike()` that follows the conventions of the SQL ILIKE operator (@edward-burn, #543).
-* New `str_to_camel()`, `str_to_snake()`, and `str_to_kebab()` for changing "programming" case (@librill, #573).
+* New `str_to_camel()`, `str_to_snake()`, and `str_to_kebab()` for changing "programming" case (@librill, #573 + @arnaudgallou, #593).
 
 ## Minor bug fies and improvements
 

R/case.R

@@ -85,7 +85,7 @@ str_to_camel <- function(string, first_upper = FALSE) {
   string <- string |>
     to_words() |>
     str_to_title() |>
-    str_remove_all(pattern = "\\s+")
+    str_remove_all(pattern = fixed(" "))
 
   if (!first_upper) {
     str_sub(string, 1, 1) <- str_to_lower(str_sub(string, 1, 1))
@@ -97,26 +97,35 @@ str_to_camel <- function(string, first_upper = FALSE) {
 #' @rdname str_to_camel
 str_to_snake <- function(string) {
   check_character(string)
-  string |>
-    to_words() |>
-    str_replace_all(pattern = "\\s+", replacement = "_")
+  to_separated_case(string, sep = "_")
 }
 #' @export
 #' @rdname str_to_camel
 str_to_kebab <- function(string) {
   check_character(string)
-  string |>
-    to_words() |>
-    str_replace_all(pattern = "\\s+", replacement = "-")
+  to_separated_case(string, sep = "-")
+}
+
+to_separated_case <- function(string, sep) {
+  out <- to_words(string)
+  str_replace_all(out, fixed(" "), sep)
 }
 
 to_words <- function(string) {
-  string |>
-    str_replace_all("([a-z])([A-Z])", "\\1 \\2") |>
-    str_replace_all("([a-zA-Z])([0-9])", "\\1 \\2") |>
-    str_replace_all("([0-9])([a-zA-Z])", "\\1 \\2") |>
-    str_replace_all("([A-Z]+)([A-Z][a-z])", "\\1 \\2") |>
-    str_to_lower() |>
-    str_replace_all(pattern = "[:punct:]", replacement = " ") |>
-    str_trim()
+  breakpoints <- paste(
+    # non-word characters
+    "[^\\p{L}\\p{N}]+",
+    # lowercase followed by uppercase
+    "(?<=\\p{Ll})(?=\\p{Lu})",
+    # letter followed by number
+    "(?<=\\p{L})(?=\\p{N})",
+    # number followed by letter
+    "(?<=\\p{N})(?=\\p{L})",
+    # uppercase followed uppercase then lowercase (i.e. end of acronym)
+    "(?<=\\p{Lu})(?=\\p{Lu}\\p{Ll})",
+    sep = "|"
+  )
+  out <- str_replace_all(string, breakpoints, " ")
+  out <- str_to_lower(out)
+  str_trim(out)
 }

tests/testthat/test-case.R

@@ -24,19 +24,29 @@ test_that("case conversions preserve names", {
 
 test_that("to_camel can control case of first argument", {
   expect_equal(str_to_camel("my_variable"), "myVariable")
+  expect_equal(str_to_camel("my$variable"), "myVariable")
+  expect_equal(str_to_camel(" my    variable  "), "myVariable")
   expect_equal(str_to_camel("my_variable", first_upper = TRUE), "MyVariable")
 })
 
 test_that("to_kebab converts to kebab case", {
   expect_equal(str_to_kebab("myVariable"), "my-variable")
   expect_equal(str_to_kebab("MyVariable"), "my-variable")
-  expect_equal(str_to_kebab("MyVariable1"), "my-variable-1")
+  expect_equal(str_to_kebab("1MyVariable1"), "1-my-variable-1")
+  expect_equal(str_to_kebab("My$Variable"), "my-variable")
+  expect_equal(str_to_kebab(" My   Variable  "), "my-variable")
+  expect_equal(str_to_kebab("testABCTest"), "test-abc-test")
+  expect_equal(str_to_kebab("IlÉtaitUneFois"), "il-était-une-fois")
 })
 
 test_that("to_snake converts to snake case", {
   expect_equal(str_to_snake("myVariable"), "my_variable")
   expect_equal(str_to_snake("MyVariable"), "my_variable")
-  expect_equal(str_to_snake("MyVariable1"), "my_variable_1")
+  expect_equal(str_to_snake("1MyVariable1"), "1_my_variable_1")
+  expect_equal(str_to_snake("My$Variable"), "my_variable")
+  expect_equal(str_to_snake(" My   Variable  "), "my_variable")
+  expect_equal(str_to_snake("testABCTest"), "test_abc_test")
+  expect_equal(str_to_snake("IlÉtaitUneFois"), "il_était_une_fois")
 })
 
 test_that("to_words handles common compound cases", {