C# Proto2 feature : Field presence and default values (#4642)

* Compiler changes

* Generated code changes

* Library changes

* Compiler style changes

* Generated style changes

* Fix Windows build errors

* Implement changes from review

* Reintroduce proto2 check

* Compiler changes (required handling review)

* Generated code changes (required handling review)

* Library changes (required handling review

* Field presence rewrite (compiler changes)

* Field presence rewrite (generated code changes)

* Compiler comment

* IFieldAccessor.HasValue library implementation

* Remove Clear methods and default values from proto3 code (Compiler)

* Remove Clear methods and default values from proto3 code (Generated)

* Remove Clear methods and default values from proto3 code (Library)

* Fix distcheck error

* Rewrite default string values to use base64 and convert

* Library changes (IMessage2)

* Compiler changes (IMessage2)

* Generated changes (IMessage2)

* Rebased and regenerated

* Compiler changes (initialized extension)

* Generated changes (initialized extension)

* Library changes (initialized extension)

* Refactor MessageExtensions.IsRequired

* Move string default value creator and bytes default value creator back to seperate methods

* Dead code cleanup

* Fixed segmentation fault
Removed unused header method declarations

1 files changed, 40 insertions, 18 deletions

diff --git a/src/google/protobuf/compiler/csharp/csharp_message.cc b/src/google/protobuf/compiler/csharp/csharp_message.cc
index 8a4307f1..1daae6f5 100644
--- a/src/google/protobuf/compiler/csharp/csharp_message.cc
+++ b/src/google/protobuf/compiler/csharp/csharp_message.cc
@@ -61,20 +61,27 @@ bool CompareFieldNumbers(const FieldDescriptor* d1, const FieldDescriptor* d2) {
 MessageGenerator::MessageGenerator(const Descriptor* descriptor,
                                    const Options* options)
     : SourceGeneratorBase(descriptor->file(), options),
-      descriptor_(descriptor) {
-
-  // sorted field names
-  for (int i = 0; i < descriptor_->field_count(); i++) {
-    field_names_.push_back(descriptor_->field(i)->name());
-  }
-  std::sort(field_names_.begin(), field_names_.end());
-
+      descriptor_(descriptor),
+      has_bit_field_count_(0) {
   // fields by number
   for (int i = 0; i < descriptor_->field_count(); i++) {
     fields_by_number_.push_back(descriptor_->field(i));
   }
   std::sort(fields_by_number_.begin(), fields_by_number_.end(),
             CompareFieldNumbers);
+
+  if (IsProto2(descriptor_->file())) {
+    int primitiveCount = 0;
+    for (int i = 0; i < descriptor_->field_count(); i++) {
+      const FieldDescriptor* field = descriptor_->field(i);
+      if (!IsNullable(field)) {
+        primitiveCount++;
+        if (has_bit_field_count_ == 0 || (primitiveCount % 32) == 0) {
+          has_bit_field_count_++;
+        }
+      }
+    }
+  }
 }
 
 MessageGenerator::~MessageGenerator() {
@@ -88,10 +95,6 @@ std::string MessageGenerator::full_class_name() {
   return GetClassName(descriptor_);
 }
 
-const std::vector<std::string>& MessageGenerator::field_names() {
-  return field_names_;
-}
-
 const std::vector<const FieldDescriptor*>& MessageGenerator::fields_by_number() {
   return fields_by_number_;
 }
@@ -123,6 +126,12 @@ void MessageGenerator::Generate(io::Printer* printer) {
   printer->Print(
       "private pb::UnknownFieldSet _unknownFields;\n");
 
+  for (int i = 0; i < has_bit_field_count_; i++) {
+    // don't use arrays since all arrays are heap allocated, saving allocations
+    // use ints instead of bytes since bytes lack bitwise operators, saving casts
+    printer->Print("private int _hasBits$i$;\n", "i", SimpleItoa(i));
+  }
+
   WriteGeneratedCodeAttributes(printer);
 
   printer->Print(
@@ -288,6 +297,9 @@ void MessageGenerator::GenerateCloningCode(io::Printer* printer) {
     vars,
     "public $class_name$($class_name$ other) : this() {\n");
   printer->Indent();
+  for (int i = 0; i < has_bit_field_count_; i++) {
+    printer->Print("_hasBits$i$ = other._hasBits$i$;\n", "i", SimpleItoa(i));
+  }
   // Clone non-oneof fields first
   for (int i = 0; i < descriptor_->field_count(); i++) {
     if (!descriptor_->field(i)->containing_oneof()) {
@@ -559,19 +571,29 @@ void MessageGenerator::GenerateMergingMethods(io::Printer* printer) {
   printer->Print("}\n\n"); // method
 }
 
-int MessageGenerator::GetFieldOrdinal(const FieldDescriptor* descriptor) {
-  for (int i = 0; i < field_names().size(); i++) {
-    if (field_names()[i] == descriptor->name()) {
-      return i;
+// it's a waste of space to track presence for all values, so we only track them if they're not nullable
+int MessageGenerator::GetPresenceIndex(const FieldDescriptor* descriptor) {
+  if (IsNullable(descriptor) || !IsProto2(descriptor_->file())) {
+    return -1;
+  }
+
+  int index = 0;
+  for (int i = 0; i < fields_by_number().size(); i++) {
+    const FieldDescriptor* field = fields_by_number()[i];
+    if (field == descriptor) {
+      return index;
+    }
+    if (!IsNullable(field)) {
+      index++;
     }
   }
-  GOOGLE_LOG(DFATAL)<< "Could not find ordinal for field " << descriptor->name();
+  GOOGLE_LOG(DFATAL)<< "Could not find presence index for field " << descriptor->name();
   return -1;
 }
 
 FieldGeneratorBase* MessageGenerator::CreateFieldGeneratorInternal(
     const FieldDescriptor* descriptor) {
-  return CreateFieldGenerator(descriptor, GetFieldOrdinal(descriptor), this->options());
+  return CreateFieldGenerator(descriptor, GetPresenceIndex(descriptor), this->options());
 }
 
 }  // namespace csharp