Use the code point iterator when computing 16-bit string hash codes.

runtime/vm/object.cc

 // Copyright (c) 2012, the Dart project authors.  Please see the AUTHORS file
 // for details. All rights reserved. Use of this source code is governed by a
 // BSD-style license that can be found in the LICENSE file.
 
 #include "vm/object.h"
 
 #include "include/dart_api.h"
 #include "platform/assert.h"
 #include "vm/assembler.h"
 #include "vm/bigint_operations.h"
@@ skipping 9975 matching lines @@
   this->SetHash(result);
   return result;
 }
 
 
 intptr_t String::Hash(const String& str, intptr_t begin_index, intptr_t len) {
   ASSERT(begin_index >= 0);
   ASSERT(len >= 0);
   ASSERT((begin_index + len) <= str.Length());
   StringHasher hasher;
-  for (intptr_t i = 0; i < len; i++) {
-    hasher.Add(str.CharAt(begin_index + i));
+  CodePointIterator it(str, begin_index, len);
+  while (it.Next()) {
+    hasher.Add(it.Current());
   }
   return hasher.Finalize(String::kHashBits);
 }
 
 
 template<typename T>
 static intptr_t HashImpl(const T* characters, intptr_t len) {
   ASSERT(len >= 0);
   StringHasher hasher;
   for (intptr_t i = 0; i < len; i++) {
     hasher.Add(characters[i]);
   }
   return hasher.Finalize(String::kHashBits);
 }
 
 
 intptr_t String::Hash(const uint8_t* characters, intptr_t len) {
   return HashImpl(characters, len);
 }
 
 
 intptr_t String::Hash(const uint16_t* characters, intptr_t len) {
-  return HashImpl(characters, len);
+  StringHasher hasher;
+  intptr_t i = 0;
+  while (i < len) {
+    hasher.Add(Utf16::Next(characters, &i, len));
+  }
+  return hasher.Finalize(String::kHashBits);
 }
 
 
 intptr_t String::Hash(const int32_t* characters, intptr_t len) {
   return HashImpl(characters, len);
 }
 
 
 int32_t String::CharAt(intptr_t index) const {
   intptr_t class_id = raw()->GetClassId();
@@ skipping 604 matching lines @@
 
 
 RawString* String::ToLowerCase(const String& str, Heap::Space space) {
   // TODO(cshapiro): create a fast-path for OneByteString instances.
   return Transform(CaseMapping::ToLower, str, space);
 }
 
 
 bool String::CodePointIterator::Next() {
   ASSERT(index_ >= -1);
-  ASSERT(index_ < str_.Length());
-  int d = Utf16::Length(ch_);
-  if (index_ == (str_.Length() - d)) {
-    return false;
+  intptr_t length = Utf16::Length(ch_);
+  if (index_ < (end_ - length)) {
+    index_ += length;
+    ch_ = str_.CharAt(index_);
+    if (Utf16::IsLeadSurrogate(ch_) && (index_ < (end_ - 1))) {
+      int32_t ch2 = str_.CharAt(index_ + 1);
+      if (Utf16::IsTrailSurrogate(ch2)) {
+        ch_ = Utf16::Decode(ch_, ch2);
+      }
+    }
+    return true;
+  } else {

[siva, 2012/11/30 02:00:19]

is this else necessary?
it could just be:
  }
  index_ = end_;
  return false;
}

[cshapiro, 2012/11/30 02:28:22]

Strictly speaking, it is not.  I will remove it.

+    index_ = end_;
   }
-  index_ += d;
-  ch_ = str_.CharAt(index_);
-  if (Utf16::IsLeadSurrogate(ch_) && (index_ != (str_.Length() - 1))) {
-    int32_t ch2 = str_.CharAt(index_ + 1);
-    if (Utf16::IsTrailSurrogate(ch2)) {
-      ch_ = Utf16::Decode(ch_, ch2);
-    }
-  }
-  return true;
+  return false;
 }
 
 
 RawOneByteString* OneByteString::EscapeSpecialCharacters(const String& str,
                                                          bool raw_str) {
   intptr_t len = str.Length();
   if (len > 0) {
     intptr_t num_escapes = 0;
     intptr_t index = 0;
     for (intptr_t i = 0; i < len; i++) {
@@ skipping 1541 matching lines @@
   }
   return result.raw();
 }
 
 
 const char* WeakProperty::ToCString() const {
   return "_WeakProperty";
 }
 
 }  // namespace dart