Handle illegal escape sequences with �

Author: stasm

fluent.syntax/src/main/kotlin/org/projectfluent/syntax/processor/Processor.kt

@@ -1,7 +1,6 @@
 package org.projectfluent.syntax.processor
 
 import org.projectfluent.syntax.ast.* // ktlint-disable no-wildcard-imports
-import java.lang.Exception
 import java.lang.StringBuilder
 
 /**
@@ -161,21 +160,27 @@ class Processor {
         if (uni4 != "") {
             val codepoint = uni4.substring(1).toInt(16)
             if (Character.isBmpCodePoint(codepoint)) {
-                return codepoint.toChar().toString()
+                val char = codepoint.toChar()
+                if (!Character.isSurrogate(char)) {
+                    return char.toString()
+                }
             }
         }
 
         val uni6 = matches.next()
         if (uni6 != "") {
             val codepoint = uni6.substring(1).toInt(16)
             if (Character.isValidCodePoint(codepoint)) {
-                val builder = StringBuilder()
-                builder.append(Character.highSurrogate(codepoint))
-                builder.append(Character.lowSurrogate(codepoint))
-                return builder
+                val char = codepoint.toChar()
+                if (!Character.isSurrogate(char)) {
+                    val builder = StringBuilder()
+                    builder.append(Character.highSurrogate(codepoint))
+                    builder.append(Character.lowSurrogate(codepoint))
+                    return builder
+                }
             }
         }
 
-        throw Exception("Unexpected")
+        return "�"
     }
 }

fluent.syntax/src/test/kotlin/org/projectfluent/syntax/processor/ProcessorTest.kt

@@ -81,6 +81,30 @@ internal class ProcessorTest {
             processor.unescapeLiteralsToText(pattern)
         )
 
+        pattern.elements.clear()
+        pattern.elements.addAll(
+            arrayOf(
+                TextElement("Illegal escape sequence: "),
+                Placeable(expression = StringLiteral("""\ud800"""))
+            )
+        )
+        assertEquals(
+            Pattern(TextElement("Illegal escape sequence: �")),
+            processor.unescapeLiteralsToText(pattern)
+        )
+
+        pattern.elements.clear()
+        pattern.elements.addAll(
+            arrayOf(
+                TextElement("Illegal escape sequence: "),
+                Placeable(expression = StringLiteral("""\U00d800"""))
+            )
+        )
+        assertEquals(
+            Pattern(TextElement("Illegal escape sequence: �")),
+            processor.unescapeLiteralsToText(pattern)
+        )
+
         pattern.elements.clear()
         pattern.elements.addAll(
             arrayOf(