[regexp] back refs must not start/end in the middle of a surrogate pair

src/regexp/regexp-macro-assembler.h

 // Copyright 2012 the V8 project authors. All rights reserved.
 // Use of this source code is governed by a BSD-style license that can be
 // found in the LICENSE file.
 
 #ifndef V8_REGEXP_REGEXP_MACRO_ASSEMBLER_H_
 #define V8_REGEXP_REGEXP_MACRO_ASSEMBLER_H_
 
 #include "src/assembler.h"
 #include "src/regexp/regexp-ast.h"
 
 namespace v8 {
 namespace internal {
 
+static const uc32 kLeadSurrogateStart = 0xd800;
+static const uc32 kLeadSurrogateEnd = 0xdbff;
+static const uc32 kTrailSurrogateStart = 0xdc00;
+static const uc32 kTrailSurrogateEnd = 0xdfff;
+static const uc32 kNonBmpStart = 0x10000;
+static const uc32 kNonBmpEnd = 0x10ffff;
+
 struct DisjunctDecisionRow {
   RegExpCharacterClass cc;
   Label* on_match;
 };
 
 
 class RegExpMacroAssembler {
  public:
   // The implementation must be able to handle at least:
   static const int kMaxRegister = (1 << 16) - 1;
@@ skipping 121 matching lines @@
   virtual void WriteCurrentPositionToRegister(int reg, int cp_offset) = 0;
   virtual void ClearRegisters(int reg_from, int reg_to) = 0;
   virtual void WriteStackPointerToRegister(int reg) = 0;
 
   // Compares two-byte strings case insensitively.
   // Called from generated RegExp code.
   static int CaseInsensitiveCompareUC16(Address byte_offset1,
                                         Address byte_offset2,
                                         size_t byte_length, Isolate* isolate);
 
+  // Check that we are not in the middle of a surrogate pair.
+  void CheckNotInSurrogatePair(int cp_offset, Label* on_failure);
+
   // Controls the generation of large inlined constants in the code.
   void set_slow_safe(bool ssc) { slow_safe_compiler_ = ssc; }
   bool slow_safe() { return slow_safe_compiler_; }
 
   enum GlobalMode { NOT_GLOBAL, GLOBAL, GLOBAL_NO_ZERO_LENGTH_CHECK };
   // Set whether the regular expression has the global flag.  Exiting due to
   // a failure in a global regexp may still mean success overall.
   inline void set_global_mode(GlobalMode mode) { global_mode_ = mode; }
   inline bool global() { return global_mode_ != NOT_GLOBAL; }
   inline bool global_with_zero_length_check() {
@@ skipping 74 matching lines @@
                         int output_size,
                         Isolate* isolate);
 };
 
 #endif  // V8_INTERPRETED_REGEXP
 
 }  // namespace internal
 }  // namespace v8
 
 #endif  // V8_REGEXP_REGEXP_MACRO_ASSEMBLER_H_