Fix scanEscapeSequence for multi-byte UTF-8 characters after backslash (#3911)

Co-authored-by: copilot-swe-agent[bot] <198982749+Copilot@users.noreply.github.com>
Co-authored-by: jakebailey <5341706+jakebailey@users.noreply.github.com>

Author: Copilot

internal/scanner/scanner.go

@@ -1796,12 +1796,23 @@ func (s *Scanner) scanEscapeSequence(flags EscapeSequenceScanningFlags) string {
 		}
 		fallthrough
 	case '\n':
-		// case CharacterCodes.lineSeparator !!!
-		// case CharacterCodes.paragraphSeparator !!!
 		return ""
 	default:
+		// ch was read as a single byte; for multi-byte UTF-8 characters,
+		// we need to decode the full rune and advance past all its bytes.
+		if ch >= utf8.RuneSelf {
+			s.pos-- // back up past the single-byte advance
+			var size int
+			ch, size = utf8.DecodeRuneInString(s.text[s.pos:])
+			s.pos += size
+			s.containsNonASCII = true
+		}
+		// LineContinuation: a backslash followed by a line terminator is "the empty code unit sequence".
+		if ch == '\u2028' || ch == '\u2029' {
+			return ""
+		}
 		if flags&EscapeSequenceScanningFlagsAnyUnicodeMode != 0 || flags&EscapeSequenceScanningFlagsRegularExpression != 0 && flags&EscapeSequenceScanningFlagsAnnexB == 0 && IsIdentifierPart(ch) {
-			s.errorAt(diagnostics.This_character_cannot_be_escaped_in_a_regular_expression, s.pos-2, 2)
+			s.errorAt(diagnostics.This_character_cannot_be_escaped_in_a_regular_expression, start, s.pos-start)
 		}
 		return string(ch)
 	}

testdata/baselines/reference/compiler/backslashBeforeNonSpecialChar.js

@@ -0,0 +1,13 @@
+//// [tests/cases/compiler/backslashBeforeNonSpecialChar.ts] ////
+
+//// [backslashBeforeNonSpecialChar.ts]
+const enum Currency {
+  Euro = "\€",
+}
+
+const currency = Currency.Euro;
+
+
+//// [backslashBeforeNonSpecialChar.js]
+"use strict";
+const currency = "\u20AC" /* Currency.Euro */;

testdata/baselines/reference/compiler/backslashBeforeNonSpecialChar.symbols

@@ -0,0 +1,16 @@
+//// [tests/cases/compiler/backslashBeforeNonSpecialChar.ts] ////
+
+=== backslashBeforeNonSpecialChar.ts ===
+const enum Currency {
+>Currency : Symbol(Currency, Decl(backslashBeforeNonSpecialChar.ts, 0, 0))
+
+  Euro = "\€",
+>Euro : Symbol(Currency.Euro, Decl(backslashBeforeNonSpecialChar.ts, 0, 21))
+}
+
+const currency = Currency.Euro;
+>currency : Symbol(currency, Decl(backslashBeforeNonSpecialChar.ts, 4, 5))
+>Currency.Euro : Symbol(Currency.Euro, Decl(backslashBeforeNonSpecialChar.ts, 0, 21))
+>Currency : Symbol(Currency, Decl(backslashBeforeNonSpecialChar.ts, 0, 0))
+>Euro : Symbol(Currency.Euro, Decl(backslashBeforeNonSpecialChar.ts, 0, 21))
+

testdata/baselines/reference/compiler/backslashBeforeNonSpecialChar.types

@@ -0,0 +1,17 @@
+//// [tests/cases/compiler/backslashBeforeNonSpecialChar.ts] ////
+
+=== backslashBeforeNonSpecialChar.ts ===
+const enum Currency {
+>Currency : Currency
+
+  Euro = "\€",
+>Euro : Currency.Euro
+>"\€" : "€"
+}
+
+const currency = Currency.Euro;
+>currency : Currency.Euro
+>Currency.Euro : Currency
+>Currency : typeof Currency
+>Euro : Currency
+

testdata/baselines/reference/submodule/compiler/sourceMap-LineBreaks(target=es2015).types

@@ -49,15 +49,15 @@ line 2";
 
 var stringLiteralWithLineSeparator = "line 1\
 >stringLiteralWithLineSeparator : string
->"line 1\
line 2" : "line 1�line 2"
+>"line 1\
line 2" : "line 1line 2"
 
 line 2";
 var stringLiteralWithParagraphSeparator = "line 1\
 >stringLiteralWithParagraphSeparator : string
->"line 1\
line 2" : "line 1�line 2"
+>"line 1\
line 2" : "line 1line 2"
 
 line 2";
 var stringLiteralWithNextLine = "line 1\…line 2";
 >stringLiteralWithNextLine : string
->"line 1\…line 2" : "line 1Â�line 2"
+>"line 1\…line 2" : "line 1\u0085line 2"
 

testdata/baselines/reference/submodule/compiler/sourceMap-LineBreaks(target=es2015).types.diff

@@ -74,7 +74,7 @@ STRING_CONTENT
 
 STRING_CONTENT",
     `\
->`\
` : "�"
+>`\
` : ""
 
 `,
     '

testdata/baselines/reference/submodule/conformance/allowUnescapedParagraphAndLineSeparatorsInStringLiteral.types

@@ -0,0 +1,7 @@
+// @target: esnext
+
+const enum Currency {
+  Euro = "\€",
+}
+
+const currency = Currency.Euro;