From a8a6104149011bf1b7f3cc4104bb7a14f1fcbb1f Mon Sep 17 00:00:00 2001
From: Yuchao Liang <yuchao.dev@gmail.com>
Date: Wed, 25 Dec 2024 02:55:42 +0900
Subject: [PATCH] fix(formatjs): Failed to parse unicode

Closes #372, closes #367
---
 .../src/parser.rs                             | 22 ++++++++--
 .../tests/fixtures/unicode_1                  | 43 +++++++++++++++++++
 .../tests/run_parser_e2e.rs                   |  1 +
 3 files changed, 63 insertions(+), 3 deletions(-)
 create mode 100644 crates/swc_icu_messageformat_parser/tests/fixtures/unicode_1

diff --git a/crates/swc_icu_messageformat_parser/src/parser.rs b/crates/swc_icu_messageformat_parser/src/parser.rs
index c7d3feec..2f35211a 100644
--- a/crates/swc_icu_messageformat_parser/src/parser.rs
+++ b/crates/swc_icu_messageformat_parser/src/parser.rs
@@ -832,7 +832,16 @@ impl<'s> Parser<'s> {
             self.bump();
         }
 
-        &self.message[start_offset..self.offset()]
+        let length = self.offset() - start_offset;
+        #[cfg(feature = "utf16")]
+        let (start_offset, length) = (
+            self.message_utf16[..start_offset].to_string().len(),
+            self.message_utf16[start_offset..start_offset + length]
+                .to_string()
+                .len(),
+        );
+
+        &self.message[start_offset..start_offset + length]
     }
 
     fn parse_literal(&self, nesting_level: usize, parent_arg_type: &str) -> Result<AstElement> {
@@ -1735,9 +1744,16 @@ impl<'s> Parser<'s> {
         if self.is_eof() {
             return None;
         }
-        self.message[self.offset() + self.char().len_utf8()..]
+
+        #[cfg(feature = "utf16")]
+        return self.message_utf16[self.offset() + self.char().len_utf16()..]
             .chars()
-            .next()
+            .next();
+
+        #[cfg(not(feature = "utf16"))]
+        return self.message[self.offset() + self.char().len_utf8()..]
+            .chars()
+            .next();
     }
 
     /// Returns true if the next call to `bump` would return false.
diff --git a/crates/swc_icu_messageformat_parser/tests/fixtures/unicode_1 b/crates/swc_icu_messageformat_parser/tests/fixtures/unicode_1
new file mode 100644
index 00000000..0d4408eb
--- /dev/null
+++ b/crates/swc_icu_messageformat_parser/tests/fixtures/unicode_1
@@ -0,0 +1,43 @@
+<a>ö🚀</a>
+---
+{}
+---
+{
+  "err": null,
+  "val": [
+    {
+      "children": [
+        {
+          "location": {
+            "end": {
+              "column": 6,
+              "line": 1,
+              "offset": 6
+            },
+            "start": {
+              "column": 4,
+              "line": 1,
+              "offset": 3
+            }
+          },
+          "type": 0,
+          "value": "ö🚀"
+        }
+      ],
+      "location":  {
+        "end":  {
+          "column": 10,
+          "line": 1,
+          "offset": 10
+        },
+        "start":  {
+          "column": 1,
+          "line": 1,
+          "offset": 0
+        }
+      },
+      "type": 8,
+      "value": "a"
+    }
+  ]
+}
\ No newline at end of file
diff --git a/crates/swc_icu_messageformat_parser/tests/run_parser_e2e.rs b/crates/swc_icu_messageformat_parser/tests/run_parser_e2e.rs
index ac3e0fef..2845569f 100644
--- a/crates/swc_icu_messageformat_parser/tests/run_parser_e2e.rs
+++ b/crates/swc_icu_messageformat_parser/tests/run_parser_e2e.rs
@@ -38,6 +38,7 @@ fn read_sections(file: PathBuf) -> TestFixtureSections {
     fixture("tests/fixtures/treat_unicode_nbsp_as_whitespace")
 )]
 #[cfg_attr(feature = "utf16", fixture("tests/fixtures/trivial_2"))]
+#[cfg_attr(feature = "utf16", fixture("tests/fixtures/unicode_1"))]
 #[fixture("tests/fixtures/uppercase_tag_1")]
 #[fixture("tests/fixtures/expect_number_arg_skeleton_token_1")]
 #[fixture("tests/fixtures/self_closing_tag_1")]