Improve pop_grapheme perf on JS

2 files changed, 6 insertions, 0 deletions

diff --git a/CHANGELOG.md b/CHANGELOG.md
index 2422e7e..d553c2a 100644
--- a/CHANGELOG.md
+++ b/CHANGELOG.md
@@ -2,6 +2,7 @@
 
 ## Unreleased
 
+- Improved performance of `string.to_graphemes` on JavaScript.
 - The `iterator` module gains the `map2` function.
 - The `list` module gains the `key_filter` function.
 
diff --git a/src/gleam/string.gleam b/src/gleam/string.gleam
index 39a69d3..5a85cc3 100644
--- a/src/gleam/string.gleam
+++ b/src/gleam/string.gleam
@@ -679,6 +679,10 @@ fn do_trim_right(string string: String) -> String
 /// Splits a non-empty `String` into its first element (head) and rest (tail).
 /// This lets you pattern match on `String`s exactly as you would with lists.
 ///
+/// Note on JavaScript using the function to iterate over a string will likely
+/// be slower than using `to_graphemes` due to string slicing being more
+/// expensive on JavaScript than Erlang.
+///
 /// ## Examples
 ///
 /// ```gleam
@@ -707,6 +711,7 @@ fn do_pop_grapheme(string string: String) -> Result(#(String, String), Nil)
 /// ["a", "b", "c"]
 /// ```
 ///
+@external(javascript, "../gleam_stdlib.mjs", "graphemes")
 pub fn to_graphemes(string: String) -> List(String) {
   do_to_graphemes(string, [])
   |> list.reverse