Preserve the contents of Dart strings with unmatched surrogate halfs by avoiding a UTF16 -> UTF8

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/cpu.h"
@@ skipping 8507 matching lines @@
   JSONObject jsobj(stream);
   AddTypeProperties(&jsobj, "Script", JSONType(), ref);
   const String& name = String::Handle(url());
   ASSERT(!name.IsNull());
   const String& encoded_url = String::Handle(String::EncodeIRI(name));
   ASSERT(!encoded_url.IsNull());
   const Library& lib = Library::Handle(FindLibrary());
   intptr_t lib_index = (lib.IsNull()) ? -1 : lib.index();
   jsobj.AddPropertyF("id", "libraries/%" Pd "/scripts/%s",
       lib_index, encoded_url.ToCString());
-  jsobj.AddProperty("name", name.ToCString());
+  jsobj.AddPropertyS("name", name, name.Length());

[turnidge, 2014/09/10 19:55:28]

I'm thinking it would be nicer to have the length parameter provide a default
value of, say, -1.  Then we could modify AddDartString to interpret a -1 length
as no limit.  This would make it simpler to use AddPropertyS.

[rmacnak, 2014/09/10 21:38:49]

Done.

   jsobj.AddProperty("kind", GetKindAsCString());
   if (ref) {
     return;
   }
   jsobj.AddProperty("owningLibrary", lib);
   const String& source = String::Handle(Source());
-  jsobj.AddProperty("source", source.ToCString());
+  jsobj.AddPropertyS("source", source, source.Length());
 
   // Print the line number table
   {
     JSONArray tokenPosTable(&jsobj, "tokenPosTable");
 
     const GrowableObjectArray& lineNumberArray =
         GrowableObjectArray::Handle(GenerateLineNumberArray());
     Object& value = Object::Handle();
     intptr_t pos = 0;
 
@@ skipping 1259 matching lines @@
 
 
 void Library::PrintJSONImpl(JSONStream* stream, bool ref) const {
   const char* library_name = String::Handle(name()).ToCString();
   intptr_t id = index();
   ASSERT(id >= 0);
   JSONObject jsobj(stream);
   AddTypeProperties(&jsobj, "Library", JSONType(), ref);
   jsobj.AddPropertyF("id", "libraries/%" Pd "", id);
   jsobj.AddProperty("name", library_name);
-  const char* library_url = String::Handle(url()).ToCString();
-  jsobj.AddProperty("url", library_url);
+  const String& library_url = String::Handle(url());
+  jsobj.AddPropertyS("url", library_url, library_url.Length());
   if (ref) {
     return;
   }
   {
     JSONArray jsarr(&jsobj, "classes");
     ClassDictionaryIterator class_iter(*this);
     Class& klass = Class::Handle();
     while (class_iter.HasNext()) {
       klass = class_iter.GetNextClass();
       if (!klass.IsCanonicalSignatureClass() &&
@@ skipping 7893 matching lines @@
     result = OneByteString::New(length, space);
   } else {
     result = TwoByteString::New(length, space);
   }
   String::Copy(result, 0, str, begin_index, length);
   return result.raw();
 }
 
 
 const char* String::ToCString() const {
-  intptr_t length;
-  return ToCString(&length);
-}
-
-
-const char* String::ToCString(intptr_t* length) const {
   if (IsOneByteString()) {
     // Quick conversion if OneByteString contains only ASCII characters.
     intptr_t len = Length();
     if (len == 0) {
-      *length = 0;
       return "";
     }
     Zone* zone = Isolate::Current()->current_zone();
     uint8_t* result = zone->Alloc<uint8_t>(len + 1);
     NoGCScope no_gc;
     const uint8_t* original_str = OneByteString::CharAddr(*this, 0);
     for (intptr_t i = 0; i < len; i++) {
       if (original_str[i] <= Utf8::kMaxOneByteChar) {
         result[i] = original_str[i];
       } else {
         len = -1;
         break;
       }
     }
     if (len > 0) {
       result[len] = 0;
-      *length = len;
       return reinterpret_cast<const char*>(result);
     }
   }
   const intptr_t len = Utf8::Length(*this);
   Zone* zone = Isolate::Current()->current_zone();
   uint8_t* result = zone->Alloc<uint8_t>(len + 1);
   ToUTF8(result, len);
   result[len] = 0;
-  *length = len;
   return reinterpret_cast<const char*>(result);
 }
 
 
-const char* String::ToCStringTruncated(intptr_t max_len,
-                                       bool* did_truncate,
-                                       intptr_t* length) const {
-  if (Length() <= max_len) {
-    *did_truncate = false;
-    return ToCString(length);
-  }
-
-  intptr_t aligned_limit = max_len;
-  if (Utf16::IsLeadSurrogate(CharAt(max_len - 1))) {
-    // Don't let truncation split a surrogate pair.
-    aligned_limit--;
-  }
-  ASSERT(!Utf16::IsLeadSurrogate(CharAt(aligned_limit - 1)));
-
-  *did_truncate = true;
-  const String& truncated =
-      String::Handle(String::SubString(*this, 0, aligned_limit));
-  return truncated.ToCString(length);
-}
-
-
 void String::PrintJSONImpl(JSONStream* stream, bool ref) const {
   JSONObject jsobj(stream);
   if (raw() == Symbols::OptimizedOut().raw()) {
     // TODO(turnidge): This is a hack.  The user could have this
     // special string in their program.  Fixing this involves updating
     // the debugging api a bit.
     jsobj.AddProperty("type", "Sentinel");
     jsobj.AddProperty("id", "objects/optimized-out");
     jsobj.AddProperty("valueAsString", "<optimized out>");
     return;
   }
   AddTypeProperties(&jsobj, "String", JSONType(), ref);
   PrintSharedInstanceJSON(&jsobj, ref);
   ObjectIdRing* ring = Isolate::Current()->object_id_ring();
   const intptr_t id = ring->GetIdForObject(raw());
   jsobj.AddPropertyF("id", "objects/%" Pd "", id);
   if (ref) {
-    bool did_truncate = false;
-    intptr_t length = 0;
-    const char* cstr = ToCStringTruncated(128, &did_truncate, &length);
-    jsobj.AddProperty("valueAsString", cstr, length);
+    bool did_truncate = jsobj.AddPropertyS("valueAsString", *this, 128);
     if (did_truncate) {
       jsobj.AddProperty("valueAsStringIsTruncated", did_truncate);
     }
   } else {
-    intptr_t length = 0;
-    const char* cstr = ToCString(&length);
-    jsobj.AddProperty("valueAsString", cstr, length);
+    bool did_truncate = jsobj.AddPropertyS("valueAsString", *this, Length());
+    ASSERT(!did_truncate);
   }
 }
 
 
 void String::ToUTF8(uint8_t* utf8_array, intptr_t array_len) const {
   ASSERT(array_len >= Utf8::Length(*this));
   Utf8::Encode(*this, reinterpret_cast<char*>(utf8_array), array_len);
 }
 
 
@@ skipping 2523 matching lines @@
   return tag_label.ToCString();
 }
 
 
 void UserTag::PrintJSONImpl(JSONStream* stream, bool ref) const {
   Instance::PrintJSONImpl(stream, ref);
 }
 
 
 }  // namespace dart