Fix input and output to handle UTF16 surrogate pairs.

test/cctest/test-parsing.cc

Index: test/cctest/test-parsing.cc
===================================================================
--- test/cctest/test-parsing.cc	(revision 10944)
+++ test/cctest/test-parsing.cc	(working copy)
@@ -63,7 +63,7 @@
     int length = i::StrLength(key_token.keyword);
     CHECK(static_cast<int>(sizeof(buffer)) >= length);
     {
-      i::Utf8ToUC16CharacterStream stream(keyword, length);
+      i::Utf8ToUtf16CharacterStream stream(keyword, length);
       i::Scanner scanner(&unicode_cache);
       // The scanner should parse Harmony keywords for this test.
       scanner.SetHarmonyScoping(true);
@@ -74,7 +74,7 @@
     }
     // Removing characters will make keyword matching fail.
     {
-      i::Utf8ToUC16CharacterStream stream(keyword, length - 1);
+      i::Utf8ToUtf16CharacterStream stream(keyword, length - 1);
       i::Scanner scanner(&unicode_cache);
       scanner.Initialize(&stream);
       CHECK_EQ(i::Token::IDENTIFIER, scanner.Next());
@@ -85,7 +85,7 @@
     for (int j = 0; j < static_cast<int>(ARRAY_SIZE(chars_to_append)); ++j) {
       memmove(buffer, keyword, length);
       buffer[length] = chars_to_append[j];
-      i::Utf8ToUC16CharacterStream stream(buffer, length + 1);
+      i::Utf8ToUtf16CharacterStream stream(buffer, length + 1);
       i::Scanner scanner(&unicode_cache);
       scanner.Initialize(&stream);
       CHECK_EQ(i::Token::IDENTIFIER, scanner.Next());
@@ -95,7 +95,7 @@
     {
       memmove(buffer, keyword, length);
       buffer[length - 1] = '_';
-      i::Utf8ToUC16CharacterStream stream(buffer, length);
+      i::Utf8ToUtf16CharacterStream stream(buffer, length);
       i::Scanner scanner(&unicode_cache);
       scanner.Initialize(&stream);
       CHECK_EQ(i::Token::IDENTIFIER, scanner.Next());
@@ -255,7 +255,7 @@
   uintptr_t stack_limit = i::Isolate::Current()->stack_guard()->real_climit();
   for (int i = 0; programs[i]; i++) {
     const char* program = programs[i];
-    i::Utf8ToUC16CharacterStream stream(
+    i::Utf8ToUtf16CharacterStream stream(
         reinterpret_cast<const i::byte*>(program),
         static_cast<unsigned>(strlen(program)));
     i::CompleteParserRecorder log;
@@ -291,7 +291,7 @@
   uintptr_t stack_limit = i::Isolate::Current()->stack_guard()->real_climit();
   for (int i = 0; programs[i]; i++) {
     const char* program = programs[i];
-    i::Utf8ToUC16CharacterStream stream(
+    i::Utf8ToUtf16CharacterStream stream(
         reinterpret_cast<const i::byte*>(program),
         static_cast<unsigned>(strlen(program)));
     i::CompleteParserRecorder log;
@@ -326,8 +326,9 @@
   // and then used the invalid currently scanned literal. This always
   // failed in debug mode, and sometimes crashed in release mode.
 
-  i::Utf8ToUC16CharacterStream stream(reinterpret_cast<const i::byte*>(program),
-                                      static_cast<unsigned>(strlen(program)));
+  i::Utf8ToUtf16CharacterStream stream(
+      reinterpret_cast<const i::byte*>(program),
+      static_cast<unsigned>(strlen(program)));
   i::ScriptDataImpl* data =
       i::ParserApi::PreParse(&stream, NULL, false);
   CHECK(data->HasError());
@@ -392,7 +393,7 @@
 
   uintptr_t stack_limit = i::Isolate::Current()->stack_guard()->real_climit();
 
-  i::Utf8ToUC16CharacterStream stream(
+  i::Utf8ToUtf16CharacterStream stream(
       reinterpret_cast<const i::byte*>(*program),
       static_cast<unsigned>(kProgramSize));
   i::CompleteParserRecorder log;
@@ -449,10 +450,10 @@
   i::Handle<i::String> uc16_string(
       FACTORY->NewExternalStringFromTwoByte(&resource));
 
-  i::ExternalTwoByteStringUC16CharacterStream uc16_stream(
+  i::ExternalTwoByteStringUtf16CharacterStream uc16_stream(
       i::Handle<i::ExternalTwoByteString>::cast(uc16_string), start, end);
-  i::GenericStringUC16CharacterStream string_stream(ascii_string, start, end);
-  i::Utf8ToUC16CharacterStream utf8_stream(
+  i::GenericStringUtf16CharacterStream string_stream(ascii_string, start, end);
+  i::Utf8ToUtf16CharacterStream utf8_stream(
       reinterpret_cast<const i::byte*>(ascii_source), end);
   utf8_stream.SeekForward(start);
 
@@ -575,12 +576,14 @@
   char buffer[kAllUtf8CharsSizeU];
   unsigned cursor = 0;
   for (int i = 0; i <= kMaxUC16Char; i++) {
-    cursor += unibrow::Utf8::Encode(buffer + cursor, i);
+    cursor += unibrow::Utf8::Encode(buffer + cursor,
+                                    i,
+                                    unibrow::Utf16::kNoPreviousCharacter);
   }
   ASSERT(cursor == kAllUtf8CharsSizeU);
 
-  i::Utf8ToUC16CharacterStream stream(reinterpret_cast<const i::byte*>(buffer),
-                                      kAllUtf8CharsSizeU);
+  i::Utf8ToUtf16CharacterStream stream(reinterpret_cast<const i::byte*>(buffer),
+                                       kAllUtf8CharsSizeU);
   for (int i = 0; i <= kMaxUC16Char; i++) {
     CHECK_EQU(i, stream.pos());
     int32_t c = stream.Advance();
@@ -610,7 +613,7 @@
 
 #undef CHECK_EQU
 
-void TestStreamScanner(i::UC16CharacterStream* stream,
+void TestStreamScanner(i::Utf16CharacterStream* stream,
                        i::Token::Value* expected_tokens,
                        int skip_pos = 0,  // Zero means not skipping.
                        int skip_to = 0) {
@@ -633,8 +636,8 @@
   v8::V8::Initialize();
 
   const char* str1 = "{ foo get for : */ <- \n\n /*foo*/ bib";
-  i::Utf8ToUC16CharacterStream stream1(reinterpret_cast<const i::byte*>(str1),
-                                       static_cast<unsigned>(strlen(str1)));
+  i::Utf8ToUtf16CharacterStream stream1(reinterpret_cast<const i::byte*>(str1),
+                                        static_cast<unsigned>(strlen(str1)));
   i::Token::Value expectations1[] = {
       i::Token::LBRACE,
       i::Token::IDENTIFIER,
@@ -652,8 +655,8 @@
   TestStreamScanner(&stream1, expectations1, 0, 0);
 
   const char* str2 = "case default const {THIS\nPART\nSKIPPED} do";
-  i::Utf8ToUC16CharacterStream stream2(reinterpret_cast<const i::byte*>(str2),
-                                       static_cast<unsigned>(strlen(str2)));
+  i::Utf8ToUtf16CharacterStream stream2(reinterpret_cast<const i::byte*>(str2),
+                                        static_cast<unsigned>(strlen(str2)));
   i::Token::Value expectations2[] = {
       i::Token::CASE,
       i::Token::DEFAULT,
@@ -683,7 +686,7 @@
   for (int i = 0; i <= 4; i++) {
      expectations3[6 - i] = i::Token::ILLEGAL;
      expectations3[5 - i] = i::Token::EOS;
-     i::Utf8ToUC16CharacterStream stream3(
+     i::Utf8ToUtf16CharacterStream stream3(
          reinterpret_cast<const i::byte*>(str3),
          static_cast<unsigned>(strlen(str3)));
      TestStreamScanner(&stream3, expectations3, 1, 1 + i);
@@ -692,7 +695,7 @@
 
 
 void TestScanRegExp(const char* re_source, const char* expected) {
-  i::Utf8ToUC16CharacterStream stream(
+  i::Utf8ToUtf16CharacterStream stream(
        reinterpret_cast<const i::byte*>(re_source),
        static_cast<unsigned>(strlen(re_source)));
   i::Scanner scanner(i::Isolate::Current()->unicode_cache());
@@ -835,6 +838,20 @@
     { "  for ", "(let x in {})\n"
       "    statement;", "\n"
       "  more;", i::BLOCK_SCOPE, i::EXTENDED_MODE },
+    // Check that 6-byte and 4-byte encodings of UTF-8 strings do not throw
+    // the preparser off in terms of byte offsets.
+    { "  'foo\355\240\201\355\260\211';\n"
+      "  (function fun", "(a,b) { infunction; }", ")();",
+      i::FUNCTION_SCOPE, i::CLASSIC_MODE },
+    { "  'foo\360\220\220\212';\n"
+      "  (function fun", "(a,b) { infunction; }", ")();",
+      i::FUNCTION_SCOPE, i::CLASSIC_MODE },
+    { "  'foo';\n"
+      "  (function fun", "(a,b) { 'bar\355\240\201\355\213'; }", ")();",
+      i::FUNCTION_SCOPE, i::CLASSIC_MODE },
+    { "  'foo';\n"
+      "  (function fun", "(a,b) { 'bar\360\220\220\214'; }", ")();",
+      i::FUNCTION_SCOPE, i::CLASSIC_MODE },
     { NULL, NULL, NULL, i::EVAL_SCOPE, i::CLASSIC_MODE }
   };
 
@@ -894,7 +911,7 @@
   // Preparse the data.
   i::CompleteParserRecorder log;
   i::Scanner scanner(i::Isolate::Current()->unicode_cache());
-  i::GenericStringUC16CharacterStream stream(source, 0, source->length());
+  i::GenericStringUtf16CharacterStream stream(source, 0, source->length());
   scanner.SetHarmonyScoping(harmony_scoping);
   scanner.Initialize(&stream);
   v8::preparser::PreParser::PreParseResult result =