fix: get codepoint range rather than bytes (#45)

denoland · Jul 26, 2024 · d3f6a05 · d3f6a05
1 parent 6b09f74
commit d3f6a05
Show file tree

Hide file tree

Showing 2 changed files with 14 additions and 3 deletions.
diff --git a/src/constructor_parser.rs b/src/constructor_parser.rs
@@ -230,11 +230,13 @@ impl<'a> ConstructorStringParser<'a> {
     assert!(self.token_index < self.token_list.len());
     let token = &self.token_list[self.token_index];
     let component_start_index = self.get_safe_token(self.component_start).index;
+
     self
       .input
-      .get(component_start_index..token.index) // TODO: check & codepoint
-      .unwrap()
-      .to_string()
+      .chars()
+      .skip(component_start_index)
+      .take(token.index - component_start_index)
+      .collect()
   }
 
   // Ref: https://wicg.github.io/urlpattern/#rewind-and-set-state

diff --git a/src/lib.rs b/src/lib.rs
@@ -954,6 +954,15 @@ mod tests {
     .unwrap();
   }
 
+  #[test]
+  fn issue46() {
+    quirks::process_construct_pattern_input(
+      quirks::StringOrInit::String(":café://:foo".to_owned()),
+      None,
+    )
+    .unwrap();
+  }
+
   #[test]
   fn has_regexp_group() {
     let pattern = <UrlPattern>::parse(UrlPatternInit {