Script streaming: fix split UTF-8 character handling.

src/scanner-character-streams.cc

 // Copyright 2011 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.
 
 #include "src/v8.h"
 
 #include "src/scanner-character-streams.h"
 
 #include "include/v8.h"
 #include "src/handles.h"
@@ skipping 393 matching lines @@
     *data_in_buffer += new_chars_in_buffer;
     // Make sure we used all the data.
     DCHECK(new_offset == utf8_split_char_buffer_length_);
     DCHECK(*data_in_buffer <= kBufferSize);
 
     utf8_split_char_buffer_length_ = 0;
   }
 
   // Move bytes which are part of an incomplete character from the end of the
   // current chunk to utf8_split_char_buffer_. They will be converted when the
-  // next data chunk arrives.
+  // next data chunk arrives. Note that all valid UTF-8 characters are at most 4
+  // bytes long, but if the data is invalid, we can have character values bigger
+  // than unibrow::Utf8::kMaxOneByteChar for more than 4 consecutive bytes.
   while (current_data_length_ > current_data_offset_ &&
          (c = current_data_[current_data_length_ - 1]) >
-             unibrow::Utf8::kMaxOneByteChar) {
+             unibrow::Utf8::kMaxOneByteChar &&
+         utf8_split_char_buffer_length_ < 4) {
     --current_data_length_;
     ++utf8_split_char_buffer_length_;
   }
+  CHECK(utf8_split_char_buffer_length_ <= 4);
   for (unsigned i = 0; i < utf8_split_char_buffer_length_; ++i) {
     utf8_split_char_buffer_[i] = current_data_[current_data_length_ + i];
   }
 }
 
 
 // ----------------------------------------------------------------------------
 // ExternalTwoByteStringUtf16CharacterStream
 
 ExternalTwoByteStringUtf16CharacterStream::
     ~ExternalTwoByteStringUtf16CharacterStream() { }
 
 
 ExternalTwoByteStringUtf16CharacterStream
     ::ExternalTwoByteStringUtf16CharacterStream(
         Handle<ExternalTwoByteString> data,
         int start_position,
         int end_position)
     : Utf16CharacterStream(),
       source_(data),
       raw_data_(data->GetTwoByteData(start_position)) {
   buffer_cursor_ = raw_data_,
   buffer_end_ = raw_data_ + (end_position - start_position);
   pos_ = start_position;
 }
 
 } }  // namespace v8::internal