Flatc parser support for nullable scalars (#6026)

* Parser support for nullable scalars

* Use older C++ features

* use default element

* Add a test for json, flexbuffers, and null

* test comments and names

Co-authored-by: Casper Neo <cneo@google.com>

3 files changed, 108 insertions, 2 deletions

diff --git a/include/flatbuffers/idl.h b/include/flatbuffers/idl.h
index 8c02f282..70a2b787 100644
--- a/include/flatbuffers/idl.h
+++ b/include/flatbuffers/idl.h
@@ -296,6 +296,7 @@ struct FieldDef : public Definition {
         shared(false),
         native_inline(false),
         flexbuffer(false),
+        nullable(false),
         nested_flatbuffer(NULL),
         padding(0) {}
 
@@ -314,6 +315,8 @@ struct FieldDef : public Definition {
   bool native_inline;  // Field will be defined inline (instead of as a pointer)
                        // for native tables if field is a struct.
   bool flexbuffer;     // This field contains FlexBuffer data.
+  bool nullable;       // If True, this field is Null (as opposed to default
+                       // valued).
   StructDef *nested_flatbuffer;  // This field contains nested FlatBuffer data.
   size_t padding;                // Bytes to always pad after this field.
 };
@@ -927,6 +930,7 @@ class Parser : public ParserState {
 
   bool SupportsAdvancedUnionFeatures() const;
   bool SupportsAdvancedArrayFeatures() const;
+  bool SupportsNullableScalars() const;
   Namespace *UniqueNamespace(Namespace *ns);
 
   FLATBUFFERS_CHECKED_ERROR RecurseError();
diff --git a/src/idl_parser.cpp b/src/idl_parser.cpp
index 6faca7ac..5ccd690d 100644
--- a/src/idl_parser.cpp
+++ b/src/idl_parser.cpp
@@ -721,7 +721,7 @@ CheckedError Parser::ParseField(StructDef &struct_def) {
     // Only cpp, js and ts supports the union vector feature so far.
     if (!SupportsAdvancedUnionFeatures()) {
       return Error(
-          "Vectors of unions are not yet supported in all "
+          "Vectors of unions are not yet supported in at least one of "
           "the specified programming languages.");
     }
     // For vector of union fields, add a second auto-generated vector field to
@@ -743,6 +743,19 @@ CheckedError Parser::ParseField(StructDef &struct_def) {
       return Error(
           "default values currently only supported for scalars in tables");
   }
+
+  // Mark the nullable scalars. Note that a side effect of ParseSingleValue is
+  // fixing field->value.constant to null.
+  if (IsScalar(type.base_type)) {
+    field->nullable = (field->value.constant == "null");
+    if (field->nullable && !SupportsNullableScalars()) {
+      return Error(
+        "Nullable scalars are not yet supported in at least one the of "
+        "the specified programming languages."
+      );
+    }
+  }
+
   // Append .0 if the value has not it (skip hex and scientific floats).
   // This suffix needed for generated C++ code.
   if (IsFloat(type.base_type)) {
@@ -1758,6 +1771,12 @@ CheckedError Parser::ParseSingleValue(const std::string *name, Value &e,
         TRY_ECHECK(kTokenStringOrIdent, IsBool(in_type), BASE_TYPE_BOOL);
       }
     }
+    // Check for optional scalars.
+    if (!match && IsScalar(in_type) && attribute_ == "null") {
+      e.constant = "null";
+      NEXT();
+      match = true;
+    }
     // Check if this could be a string/identifier enum value.
     // Enum can have only true integer base type.
     if (!match && IsInteger(in_type) && !IsBool(in_type) &&
@@ -1811,7 +1830,8 @@ CheckedError Parser::ParseSingleValue(const std::string *name, Value &e,
   // This flag forces to check default scalar values or metadata of field.
   // For JSON parser the flag should be false.
   // If it is set for JSON each value will be checked twice (see ParseTable).
-  if (check_now && IsScalar(match_type)) {
+  // Special case 'null' since atot can't handle that.
+  if (check_now && IsScalar(match_type) && e.constant != "null") {
     // clang-format off
     switch (match_type) {
     #define FLATBUFFERS_TD(ENUM, IDLTYPE, CTYPE, ...) \
@@ -2236,6 +2256,11 @@ CheckedError Parser::CheckClash(std::vector<FieldDef *> &fields,
   return NoError();
 }
 
+
+bool Parser::SupportsNullableScalars() const {
+  return opts.lang_to_generate == 0;  // No support yet.
+}
+
 bool Parser::SupportsAdvancedUnionFeatures() const {
   return opts.lang_to_generate != 0 &&
          (opts.lang_to_generate &
diff --git a/tests/test.cpp b/tests/test.cpp
index b7ae7672..de86f02b 100644
--- a/tests/test.cpp
+++ b/tests/test.cpp
@@ -3416,6 +3416,81 @@ void TestEmbeddedBinarySchema() {
           0);
 }
 
+void NullableScalarsTest() {
+  // Simple schemas and a "has nullable scalar" sentinal.
+  std::vector<std::string> schemas;
+  schemas.push_back("table Monster { mana : int; }");
+  schemas.push_back("table Monster { mana : int = 42; }");
+  schemas.push_back("table Monster { mana : int =  null; }");
+  schemas.push_back("table Monster { mana : long; }");
+  schemas.push_back("table Monster { mana : long = 42; }");
+  schemas.push_back("table Monster { mana : long = null; }");
+  schemas.push_back("table Monster { mana : float; }");
+  schemas.push_back("table Monster { mana : float = 42; }");
+  schemas.push_back("table Monster { mana : float = null; }");
+  schemas.push_back("table Monster { mana : double; }");
+  schemas.push_back("table Monster { mana : double = 42; }");
+  schemas.push_back("table Monster { mana : double = null; }");
+  schemas.push_back("table Monster { mana : bool; }");
+  schemas.push_back("table Monster { mana : bool = 42; }");
+  schemas.push_back("table Monster { mana : bool = null; }");
+
+  // Check the FieldDef is correctly set.
+  for (auto schema = schemas.begin(); schema < schemas.end(); schema++) {
+    const bool has_null = schema->find("null") != std::string::npos;
+    flatbuffers::Parser parser;
+    TEST_ASSERT(parser.Parse(schema->c_str()));
+    const auto *mana = parser.structs_.Lookup("Monster")->fields.Lookup("mana");
+    TEST_EQ(mana->nullable, has_null);
+  }
+
+  // Test if nullable scalars are allowed for each language.
+  const int kNumLanguages = 17;
+  for (int lang=0; lang<kNumLanguages; lang++) {
+    flatbuffers::IDLOptions opts;
+    opts.lang_to_generate |= 1 << lang;
+    for (auto schema = schemas.begin(); schema < schemas.end(); schema++) {
+      const bool has_null = schema->find("null") != std::string::npos;
+      flatbuffers::Parser parser(opts);
+      // Its not supported in any language yet so has_null means error.
+      TEST_EQ(parser.Parse(schema->c_str()), !has_null);
+    }
+  }
+}
+
+void ParseFlexbuffersFromJsonWithNullTest() {
+  // Test nulls are handled appropriately through flexbuffers to exercise other
+  // code paths of ParseSingleValue in the nullable scalars change.
+  // TODO(cneo): Json -> Flatbuffers test once some language can generate code
+  // with nullable scalars.
+  {
+    char json[] = "{\"opt_field\": 123 }";
+    flatbuffers::Parser parser;
+    flexbuffers::Builder flexbuild;
+    parser.ParseFlexBuffer(json, nullptr, &flexbuild);
+    auto root = flexbuffers::GetRoot(flexbuild.GetBuffer());
+    TEST_EQ(root.AsMap()["opt_field"].AsInt64(), 123);
+  }
+  {
+    char json[] = "{\"opt_field\": 123.4 }";
+    flatbuffers::Parser parser;
+    flexbuffers::Builder flexbuild;
+    parser.ParseFlexBuffer(json, nullptr, &flexbuild);
+    auto root = flexbuffers::GetRoot(flexbuild.GetBuffer());
+    TEST_EQ(root.AsMap()["opt_field"].AsDouble(), 123.4);
+  }
+  {
+    char json[] = "{\"opt_field\": null }";
+    flatbuffers::Parser parser;
+    flexbuffers::Builder flexbuild;
+    parser.ParseFlexBuffer(json, nullptr, &flexbuild);
+    auto root = flexbuffers::GetRoot(flexbuild.GetBuffer());
+    TEST_ASSERT(!root.AsMap().IsTheEmptyMap());
+    TEST_ASSERT(root.AsMap()["opt_field"].IsNull());
+    TEST_EQ(root.ToString(), std::string("{ opt_field: null }"));
+  }
+}
+
 int FlatBufferTests() {
   // clang-format off
 
@@ -3503,6 +3578,8 @@ int FlatBufferTests() {
   TestMonsterExtraFloats();
   FixedLengthArrayTest();
   NativeTypeTest();
+  NullableScalarsTest();
+  ParseFlexbuffersFromJsonWithNullTest();
   return 0;
 }