Fix cmn-hans-cn-t-ca-u-ca-x_t-u (#6001)

The issue here is that once the parser was in extensions mode, it would only look at the first character of the `x_t` subtag (the produced locale was `cmn-hans-cn-t-ca-u-ca-x-u`). #5943 (comment)
unicode-org · Jan 15, 2025 · 1bc501a · 1bc501a
1 parent 0959d99
commit 1bc501a
Show file tree

Hide file tree

Showing 3 changed files with 21 additions and 6 deletions.
diff --git a/components/locale_core/src/extensions/mod.rs b/components/locale_core/src/extensions/mod.rs
@@ -264,26 +264,31 @@ impl Extensions {
             if subtag.is_empty() {
                 return Err(ParseError::InvalidExtension);
             }
-            match subtag.first().map(|b| ExtensionType::try_from_byte(*b)) {
-                Some(Ok(ExtensionType::Unicode)) => {
+
+            let &[subtag] = subtag else {
+                return Err(ParseError::InvalidExtension);
+            };
+
+            match ExtensionType::try_from_byte(subtag) {
+                Ok(ExtensionType::Unicode) => {
                     if unicode.is_some() {
                         return Err(ParseError::DuplicatedExtension);
                     }
                     unicode = Some(Unicode::try_from_iter(iter)?);
                 }
-                Some(Ok(ExtensionType::Transform)) => {
+                Ok(ExtensionType::Transform) => {
                     if transform.is_some() {
                         return Err(ParseError::DuplicatedExtension);
                     }
                     transform = Some(Transform::try_from_iter(iter)?);
                 }
-                Some(Ok(ExtensionType::Private)) => {
+                Ok(ExtensionType::Private) => {
                     if private.is_some() {
                         return Err(ParseError::DuplicatedExtension);
                     }
                     private = Some(Private::try_from_iter(iter)?);
                 }
-                Some(Ok(ExtensionType::Other(ext))) => {
+                Ok(ExtensionType::Other(ext)) => {
                     if other.iter().any(|o: &Other| o.get_ext_byte() == ext) {
                         return Err(ParseError::DuplicatedExtension);
                     }

diff --git a/components/locale_core/tests/fixtures/invalid-extensions.json b/components/locale_core/tests/fixtures/invalid-extensions.json
@@ -1,4 +1,14 @@
 [
+  {
+    "input": {
+      "type": "Locale",
+      "identifier": "cmn-hans-cn-t-ca-u-ca-x_t-u"
+    },
+    "output": {
+      "error": "InvalidExtension",
+      "text": "unused"
+    }
+  },
   {
     "input": {
       "type": "Locale",

diff --git a/components/locale_core/tests/locale.rs b/components/locale_core/tests/locale.rs
@@ -47,7 +47,7 @@ fn test_locale_parsing() {
 }
 
 #[test]
-fn test_langid_invalid() {
+fn test_locale_invalid() {
     let data = serde_json::from_str(include_str!("fixtures/invalid-extensions.json"))
         .expect("Failed to read a fixture");