javascript tests and fixes for string.reverse() / string.slice() / string.to_graphemes() (#298)

3 files changed, 93 insertions, 23 deletions

diff --git a/CHANGELOG.md b/CHANGELOG.md
index 02b3a65..2457140 100644
--- a/CHANGELOG.md
+++ b/CHANGELOG.md
@@ -5,6 +5,8 @@
 - The `float` module gains the `divide` function.
 - The `int` module gains the `divide`, `power`, and `square_root` functions.
 - The `string` module gains the `first`, `last`, and `capitalise` functions.
+- Fixed a bug where `string.reverse` would break utf8 strings on target JavaScript.
+- Fixed a bug where `string.slice` would break utf8 strings on target JavaScript.
 
 ## v0.21.0 - 2022-04-24
 
diff --git a/src/gleam/string.gleam b/src/gleam/string.gleam
index 37baed4..45d4d31 100644
--- a/src/gleam/string.gleam
+++ b/src/gleam/string.gleam
@@ -73,10 +73,25 @@ if javascript {
 /// ```
 ///
 pub fn reverse(string: String) -> String {
-  string
-  |> string_builder.from_string
-  |> string_builder.reverse
-  |> string_builder.to_string
+  do_reverse(string)
+}
+
+if erlang {
+  fn do_reverse(string: String) -> String {
+    string
+    |> string_builder.from_string
+    |> string_builder.reverse
+    |> string_builder.to_string
+  }
+}
+
+if javascript {
+  fn do_reverse(string: String) -> String {
+    string
+    |> to_graphemes
+    |> list.reverse
+    |> concat
+  }
 }
 
 /// Creates a new `String` by replacing all occurrences of a given substring.
@@ -233,8 +248,13 @@ if erlang {
 }
 
 if javascript {
-  external fn do_slice(String, Int, Int) -> String =
-    "../gleam_stdlib.mjs" "slice_string"
+  fn do_slice(string: String, idx: Int, len: Int) -> String {
+    string
+    |> to_graphemes
+    |> list.drop(idx)
+    |> list.take(len)
+    |> concat
+  }
 }
 
 /// Drops contents of the first `String` that occur before the second `String`.
diff --git a/test/gleam/string_test.gleam b/test/gleam/string_test.gleam
index bb9aff0..4c73b41 100644
--- a/test/gleam/string_test.gleam
+++ b/test/gleam/string_test.gleam
@@ -25,13 +25,34 @@ pub fn uppercase_test() {
 }
 
 pub fn reverse_test() {
-  string.reverse("Gleam")
+  "Gleam"
+  |> string.reverse
   |> should.equal("maelG")
-}
 
-pub fn unicode_reverse_test() {
-  string.reverse("👍 OK")
+  " Gleam"
+  |> string.reverse
+  |> should.equal("maelG ")
+
+  "👍 OK"
+  |> string.reverse
   |> should.equal("KO 👍")
+
+  "👍"
+  |> string.reverse
+  |> should.equal("👍")
+
+  "ÅÄÖ"
+  |> string.reverse
+  |> should.equal("ÖÄÅ")
+
+  "👶🏿"
+  |> string.reverse
+  |> should.equal("👶🏿")
+
+  "👶🏿"
+  |> string.reverse
+  |> string.reverse
+  |> should.equal("👶🏿")
 }
 
 pub fn split_test() {
@@ -132,19 +153,19 @@ pub fn join_test() {
 
 pub fn trim_test() {
   "  hats  \n"
-  |> string.trim()
+  |> string.trim
   |> should.equal("hats")
 }
 
 pub fn trim_left_test() {
   "  hats  \n"
-  |> string.trim_left()
+  |> string.trim_left
   |> should.equal("hats  \n")
 }
 
 pub fn trim_right_test() {
   "  hats  \n"
-  |> string.trim_right()
+  |> string.trim_right
   |> should.equal("  hats")
 }
 
@@ -208,6 +229,10 @@ pub fn slice_test() {
   "gleam"
   |> string.slice(at_index: 2, length: -3)
   |> should.equal("")
+
+  "👶🏿"
+  |> string.slice(at_index: 0, length: 3)
+  |> should.equal("👶🏿")
 }
 
 pub fn crop_test() {
@@ -313,30 +338,53 @@ pub fn pad_right_test() {
 
 pub fn pop_grapheme_test() {
   "gleam"
-  |> string.pop_grapheme()
+  |> string.pop_grapheme
   |> should.equal(Ok(#("g", "leam")))
 
   "g"
-  |> string.pop_grapheme()
+  |> string.pop_grapheme
   |> should.equal(Ok(#("g", "")))
 
   ""
-  |> string.pop_grapheme()
+  |> string.pop_grapheme
   |> should.equal(Error(Nil))
 }
 
 pub fn to_graphemes_test() {
-  "abc"
-  |> string.to_graphemes()
-  |> should.equal(["a", "b", "c"])
+  ""
+  |> string.to_graphemes
+  |> should.equal([])
+
+  "\n\t\r\"\\"
+  |> string.to_graphemes
+  |> should.equal(["\n", "\t", "\r", "\"", "\\"])
 
   "a"
-  |> string.to_graphemes()
+  |> string.to_graphemes
   |> should.equal(["a"])
 
-  ""
-  |> string.to_graphemes()
-  |> should.equal([])
+  "abc"
+  |> string.to_graphemes
+  |> should.equal(["a", "b", "c"])
+
+  "🌷🎁💩😜👍🏳️‍🌈"
+  |> string.to_graphemes
+  |> should.equal(["🌷", "🎁", "💩", "😜", "👍", "🏳️‍🌈"])
+
+  "Ĺo͂řȩm̅"
+  |> string.to_graphemes
+  |> should.equal(["Ĺ", "o͂", "ř", "ȩ", "m̅"])
+
+  "뎌쉐"
+  |> string.to_graphemes
+  |> should.equal(["뎌", "쉐"])
+
+  "Z͑ͫ̓ͪ̂ͫ̽͏̴̙̤̞͉͚̯̞̠͍A̴̵̜̰͔ͫ͗͢L̠ͨͧͩ͘G̴̻͈͍͔̹̑͗̎̅͛́Ǫ̵̹̻̝̳͂̌̌͘!͖̬̰̙̗̿̋ͥͥ̂ͣ̐́́͜͞"
+  |> string.to_graphemes
+  |> should.equal([
+    "Z͑ͫ̓ͪ̂ͫ̽͏̴̙̤̞͉͚̯̞̠͍", "A̴̵̜̰͔ͫ͗͢", "L̠ͨͧͩ͘",
+    "G̴̻͈͍͔̹̑͗̎̅͛́", "Ǫ̵̹̻̝̳͂̌̌͘", "!͖̬̰̙̗̿̋ͥͥ̂ͣ̐́́͜͞",
+  ])
 }
 
 pub fn utf_codepoint_test() {