refactor(codegen): print string literals containing lone surrogates without reference to raw (#10044)

#10041 changed how lone surrogates are handled in `StringLiteral`s.

`StringLiteral`s which include lone surrogates now have the `lone_surrogates` flag set, and `value` encodes lone surrogates as `\u{FFFD}XXXX`, where `XXXX` is the code unit encoded as hex.

Codegen check the `lone_surrogates` flag and decode the lone surrogates if they're present. This means that:

1. A `StringLiteral` no longer needs to have `raw` field populated, so you can (if you choose to for some reason) create a new `StringLiteral` containing lone surrogates.

2. `StringLiteral`s containing lone surrogates now have any other characters escaped same as how `StringLiteral`s without lone surrogates are printed.

Author: overlookmotel

crates/oxc_codegen/src/lib.rs

@@ -242,6 +242,12 @@ impl<'a> Codegen<'a> {
         self.code.print_str(s);
     }
 
+    /// Push `char` into the buffer.
+    #[inline]
+    pub fn print_char(&mut self, ch: char) {
+        self.code.print_char(ch);
+    }
+
     /// Print a single [`Expression`], adding it to the code generator's
     /// internal buffer. Unlike [`Codegen::build`], this does not consume `self`.
     #[inline]
@@ -578,14 +584,7 @@ impl<'a> Codegen<'a> {
 
     fn print_string_literal(&mut self, s: &StringLiteral<'_>, allow_backtick: bool) {
         self.add_source_mapping(s.span);
-        if s.lone_surrogates {
-            self.print_str(s.raw.unwrap().as_str());
-            return;
-        }
-        self.print_quoted_utf16(s, allow_backtick);
-    }
 
-    fn print_quoted_utf16(&mut self, s: &StringLiteral<'_>, allow_backtick: bool) {
         let quote = if self.options.minify {
             let mut single_cost: i32 = 0;
             let mut double_cost: i32 = 0;
@@ -680,6 +679,26 @@ impl<'a> Codegen<'a> {
                     }
                     self.print_ascii_byte(b'$');
                 }
+                '\u{FFFD}' if s.lone_surrogates => {
+                    // If `lone_surrogates` is set, string contains lone surrogates which are escaped
+                    // using the lossy replacement character (U+FFFD) as an escape marker.
+                    // The lone surrogate is encoded as `\u{FFFD}XXXX` where `XXXX` is the code point as hex.
+                    let hex1 = chars.next().unwrap();
+                    let hex2 = chars.next().unwrap();
+                    let hex3 = chars.next().unwrap();
+                    let hex4 = chars.next().unwrap();
+                    if [hex1, hex2, hex3, hex4] == ['f', 'f', 'f', 'd'] {
+                        // Actual lossy replacement character
+                        self.print_char('\u{FFFD}');
+                    } else {
+                        // Lossy replacement character representing a lone surrogate
+                        self.print_str("\\u");
+                        self.print_char(hex1);
+                        self.print_char(hex2);
+                        self.print_char(hex3);
+                        self.print_char(hex4);
+                    }
+                }
                 _ => self.print_str(c.encode_utf8([0; 4].as_mut())),
             }
         }

crates/oxc_codegen/tests/integration/esbuild.rs

@@ -363,15 +363,15 @@ fn test_string() {
     test("let x = '\\U000123AB'", "let x = \"U000123AB\";\n");
     test("let x = '\\u{123AB}'", "let x = \"\u{123ab}\";\n");
     test("let x = '\\uD808\\uDFAB'", "let x = \"\u{123ab}\";\n");
-    test("let x = '\\uD808'", "let x = '\\uD808';\n"); // lone surrogate
-    test("let x = '\\uD808X'", "let x = '\\uD808X';\n");
-    test("let x = '\\uDFAB'", "let x = '\\uDFAB';\n");
-    test("let x = '\\uDFABX'", "let x = '\\uDFABX';\n");
+    test("let x = '\\uD808'", "let x = \"\\ud808\";\n"); // lone surrogate
+    test("let x = '\\uD808X'", "let x = \"\\ud808X\";\n");
+    test("let x = '\\uDFAB'", "let x = \"\\udfab\";\n");
+    test("let x = '\\uDFABX'", "let x = \"\\udfabX\";\n");
 
     test("let x = '\\x80'", "let x = \"\u{80}\";\n");
     test("let x = '\\xFF'", "let x = \"ÿ\";\n");
     test("let x = '\\xF0\\x9F\\x8D\\x95'", "let x = \"ð\u{9f}\u{8d}\u{95}\";\n");
-    test("let x = '\\uD801\\uDC02\\uDC03\\uD804'", "let x = '\\uD801\\uDC02\\uDC03\\uD804';\n"); // lossy
+    test("let x = '\\uD801\\uDC02\\uDC03\\uD804'", "let x = \"𐐂\\udc03\\ud804\";\n"); // surrogates
 }
 
 #[test]

crates/oxc_codegen/tests/integration/unit.rs

@@ -143,7 +143,7 @@ fn unicode_escape() {
     test("console.log('こんにちは');", "console.log(\"こんにちは\");\n");
     test("console.log('안녕하세요');", "console.log(\"안녕하세요\");\n");
     test("console.log('🧑‍🤝‍🧑');", "console.log(\"🧑‍🤝‍🧑\");\n");
-    test("console.log(\"\\uD800\\uD801\")", "console.log(\"\\uD800\\uD801\");\n");
+    test("console.log(\"\\uD800\\uD801\")", "console.log(\"\\ud800\\ud801\");\n");
 }
 
 #[test]