Add support for surrogates when serializing and deserializing for native ports

runtime/vm/unicode.cc

Index: runtime/vm/unicode.cc
diff --git a/runtime/vm/unicode.cc b/runtime/vm/unicode.cc
index 3129a06787edb5cbc96dee61444c63c96a862cc1..25e18e897badb2317ec71da8e6017498b49fe535 100644
--- a/runtime/vm/unicode.cc
+++ b/runtime/vm/unicode.cc
@@ -1,4 +1,4 @@
-// Copyright (c) 2011, the Dart project authors.  Please see the AUTHORS file
+// 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.
 
@@ -59,7 +59,7 @@ static bool IsTrailByte(uint8_t code_unit) {
 
 
 static bool IsLatin1SequenceStart(uint8_t code_unit) {
-  // Check is codepoint is <= U+00FF
+  // Check is codepoint is <= U+00FF.
   return (code_unit <= Utf8::kMaxOneByteChar);
 }
 
@@ -108,7 +108,8 @@ intptr_t Utf8::CodePointCount(const uint8_t* utf8_array,
 
 
 // Returns true if str is a valid NUL-terminated UTF-8 string.
-bool Utf8::IsValid(const uint8_t* utf8_array, intptr_t array_len) {
+static bool IsValidUtf8(
+    const uint8_t* utf8_array, intptr_t array_len, bool allow_surrogates) {
   intptr_t i = 0;
   while (i < array_len) {
     uint32_t ch = utf8_array[i] & 0xFF;
@@ -130,7 +131,7 @@ bool Utf8::IsValid(const uint8_t* utf8_array, intptr_t array_len) {
             (j == num_trail_bytes) &&
             !IsOutOfRange(ch) &&
             !IsNonShortestForm(ch, j) &&
-            !Utf16::IsSurrogate(ch))) {
+            (!Utf16::IsSurrogate(ch) || allow_surrogates))) {
         return false;
       }
     }
@@ -140,6 +141,17 @@ bool Utf8::IsValid(const uint8_t* utf8_array, intptr_t array_len) {
 }
 
 
+bool Utf8::IsValid(const uint8_t* utf8_array, intptr_t array_len) {
+  return IsValidUtf8(utf8_array, array_len, false);
+}
+
+
+bool Utf8::IsValidAllowSurrogates(
+    const uint8_t* utf8_array, intptr_t array_len) {
+  return IsValidUtf8(utf8_array, array_len, true);
+}
+
+
 intptr_t Utf8::Length(int32_t ch) {
   if (ch <= kMaxOneByteChar) {
     return 1;
@@ -206,9 +218,10 @@ intptr_t Utf8::Encode(const String& src, char* dst, intptr_t len) {
 }
 
 
-intptr_t Utf8::Decode(const uint8_t* utf8_array,
-                      intptr_t array_len,
-                      int32_t* dst) {
+static intptr_t DecodeUTF8(const uint8_t* utf8_array,
+                           intptr_t array_len,
+                           int32_t* dst,
+                           bool allow_surrogates) {
   uint32_t ch = utf8_array[0] & 0xFF;
   intptr_t i = 1;
   if (ch >= 0x80) {
@@ -229,7 +242,7 @@ intptr_t Utf8::Decode(const uint8_t* utf8_array,
           (i == num_trail_bytes) &&
           !IsOutOfRange(ch) &&
           !IsNonShortestForm(ch, i) &&
-          !Utf16::IsSurrogate(ch))) {
+          (!Utf16::IsSurrogate(ch) || allow_surrogates))) {
       *dst = -1;
       return 0;
     }
@@ -239,6 +252,20 @@ intptr_t Utf8::Decode(const uint8_t* utf8_array,
 }
 
 
+intptr_t Utf8::Decode(const uint8_t* utf8_array,
+                      intptr_t array_len,
+                      int32_t* dst) {
+  return DecodeUTF8(utf8_array, array_len, dst, false);
+}
+
+
+intptr_t Utf8::DecodeAllowSurrogates(const uint8_t* utf8_array,
+                      intptr_t array_len,
+                      int32_t* dst) {
+  return DecodeUTF8(utf8_array, array_len, dst, true);
+}
+
+
 bool Utf8::DecodeToLatin1(const uint8_t* utf8_array,
                           intptr_t array_len,
                           uint8_t* dst,
@@ -251,31 +278,33 @@ bool Utf8::DecodeToLatin1(const uint8_t* utf8_array,
     ASSERT(IsLatin1SequenceStart(utf8_array[i]));
     num_bytes = Utf8::Decode(&utf8_array[i], (array_len - i), &ch);
     if (ch == -1) {
-      return false;  // invalid input
+      return false;  // Invalid input.
     }
     ASSERT(ch <= 0xff);
     dst[j] = ch;
   }
   if ((i < array_len) && (j == len)) {
-    return false;  // output overflow
+    return false;  // Output overflow.
   }
-  return true;  // success
+  return true;  // Success.
 }
 
 
-bool Utf8::DecodeToUTF16(const uint8_t* utf8_array,
-                         intptr_t array_len,
-                         uint16_t* dst,
-                         intptr_t len) {
+bool DecodeUTF8ToUTF16(const uint8_t* utf8_array,
+                       intptr_t array_len,
+                       uint16_t* dst,
+                       intptr_t len,
+                       bool allow_surrogates) {
   intptr_t i = 0;
   intptr_t j = 0;
   intptr_t num_bytes;
   for (; (i < array_len) && (j < len); i += num_bytes, ++j) {
     int32_t ch;
     bool is_supplementary = IsSupplementarySequenceStart(utf8_array[i]);
-    num_bytes = Utf8::Decode(&utf8_array[i], (array_len - i), &ch);
+    num_bytes = DecodeUTF8(
+        &utf8_array[i], (array_len - i), &ch, allow_surrogates);
     if (ch == -1) {
-      return false;  // invalid input
+      return false;  // Invalid input.
     }
     if (is_supplementary) {
       Utf16::Encode(ch, &dst[j]);
@@ -285,9 +314,25 @@ bool Utf8::DecodeToUTF16(const uint8_t* utf8_array,
     }
   }
   if ((i < array_len) && (j == len)) {
-    return false;  // output overflow
+    return false;  // Output overflow.
   }
-  return true;  // success
+  return true;  // Success.
+}
+
+
+bool Utf8::DecodeToUTF16(const uint8_t* utf8_array,
+                         intptr_t array_len,
+                         uint16_t* dst,
+                         intptr_t len) {
+  return DecodeUTF8ToUTF16(utf8_array, array_len, dst, len, false);
+}
+
+
+bool Utf8::DecodeToUTF16AllowSurrogates(const uint8_t* utf8_array,
+                                        intptr_t array_len,
+                                        uint16_t* dst,
+                                        intptr_t len) {
+  return DecodeUTF8ToUTF16(utf8_array, array_len, dst, len, true);
 }
 
 
@@ -302,14 +347,14 @@ bool Utf8::DecodeToUTF32(const uint8_t* utf8_array,
     int32_t ch;
     num_bytes = Utf8::Decode(&utf8_array[i], (array_len - i), &ch);
     if (ch == -1) {
-      return false;  // invalid input
+      return false;  // Invalid input.
     }
     dst[j] = ch;
   }
   if ((i < array_len) && (j == len)) {
-    return false;  // output overflow
+    return false;  // Output overflow.
   }
-  return true;  // success
+  return true;  // Success.
 }