KT-49721 Avoid matching from the same position after a zero-width match

In JS, RegExp can return a match before the specified start index,
if it has matched at that index, but it is in the middle of a surrogate
pair. Account for that when advancing to the next position after
a zero-width match so that it doesn't get to the middle of SP.

kotlin-native/backend.native/tests/stdlib_external/utils.kt

@@ -30,4 +30,6 @@ public actual val isFloat32RangeEnforced: Boolean get() = true
 
 public actual val supportsSuppressedExceptions: Boolean get() = true
 
-public actual val supportsNamedCapturingGroup: Boolean get() = false
+public actual val supportsNamedCapturingGroup: Boolean get() = false
+
+public actual val regexSplitUnicodeCodePointHandling: Boolean get() = true

libraries/stdlib/common/test/testUtils.kt

@@ -18,4 +18,6 @@ public expect val isFloat32RangeEnforced: Boolean
 
 public expect val supportsSuppressedExceptions: Boolean
 
-public expect val supportsNamedCapturingGroup: Boolean
+public expect val supportsNamedCapturingGroup: Boolean
+
+public expect val regexSplitUnicodeCodePointHandling: Boolean

libraries/stdlib/js/src/kotlin/text/regex.kt

@@ -359,7 +359,20 @@ private fun RegExp.findNext(input: String, from: Int, nextPattern: RegExp): Matc
             }
 
         override fun next(): MatchResult? =
-            nextPattern.findNext(input, if (range.isEmpty()) range.start + 1 else range.endInclusive + 1, nextPattern)
+            nextPattern.findNext(input, if (range.isEmpty()) advanceToNextCharacter(range.start) else range.endInclusive + 1, nextPattern)
+
+        private fun advanceToNextCharacter(index: Int): Int {
+            if (index < input.lastIndex) {
+                val code1 = input.asDynamic().charCodeAt(index).unsafeCast<Int>()
+                if (code1 in 0xD800..0xDBFF) {
+                    val code2 = input.asDynamic().charCodeAt(index + 1).unsafeCast<Int>()
+                    if (code2 in 0xDC00..0xDFFF) {
+                        return index + 2
+                    }
+                }
+            }
+            return index + 1
+        }
     }
 }
 

libraries/stdlib/js/test/core/testUtils.kt

@@ -28,4 +28,6 @@ public actual val isFloat32RangeEnforced: Boolean = false
 actual val supportsSuppressedExceptions: Boolean get() = true
 
 // TODO: implement named group reference in replacement expression
-public actual val supportsNamedCapturingGroup: Boolean get() = false
+public actual val supportsNamedCapturingGroup: Boolean get() = false
+
+public actual val regexSplitUnicodeCodePointHandling: Boolean get() = true

libraries/stdlib/jvm/test/testUtilsJVM.kt

@@ -41,4 +41,6 @@ public actual val isFloat32RangeEnforced: Boolean = true
 
 public actual val supportsSuppressedExceptions: Boolean get() = !isJava6
 
-public actual val supportsNamedCapturingGroup: Boolean get() = !isJava6
+public actual val supportsNamedCapturingGroup: Boolean get() = !isJava6
+
+public actual val regexSplitUnicodeCodePointHandling: Boolean get() = false

libraries/stdlib/test/text/RegexTest.kt

@@ -7,6 +7,7 @@
 
 package test.text
 
+import test.regexSplitUnicodeCodePointHandling
 import test.supportsNamedCapturingGroup
 import kotlin.test.*
 
@@ -354,6 +355,11 @@ class RegexTest {
 
         testSplitEquals("".split(""), "", emptyMatch)
 
+        testSplitEquals(
+            if (regexSplitUnicodeCodePointHandling) listOf("", "\uD83D\uDE04", "\uD801", "") else listOf("", "\uD83D", "\uDE04", "\uD801", ""),
+            "\uD83D\uDE04\uD801", emptyMatch
+        )
+
         val emptyMatchBeforeT = "(?=t)".toRegex()
 
         testSplitEquals(listOf("", "tes", "t"), input, emptyMatchBeforeT)
@@ -397,4 +403,16 @@ class RegexTest {
         assertFailsWith<NoSuchElementException> { splits.next() }
     }
 
+    @Test fun findAllEmoji() {
+        val input = "\uD83D\uDE04\uD801x"
+        val regex = ".".toRegex()
+
+        val matches = regex.findAll(input).toList()
+        val values = matches.map { it.value }
+        val ranges = matches.map { it.range }
+
+        assertEquals(listOf("\uD83D\uDE04", "\uD801", "x"), values)
+        assertEquals(listOf(0..1, 2..2, 3..3), ranges)
+    }
+
 }

libraries/stdlib/wasm/test/testUtils.kt

@@ -31,3 +31,5 @@ actual val supportsSuppressedExceptions: Boolean get() = false
 
 // TODO: implement named group reference in replacement expression
 public actual val supportsNamedCapturingGroup: Boolean get() = false
+
+public actual val regexSplitUnicodeCodePointHandling: Boolean get() = TODO()