Add `CharLiteral` to SyntaxTree

Subscribers: cfe-commits

Tags: #clang

Differential Revision: https://reviews.llvm.org/D82312

clang/include/clang/Tooling/Syntax/Nodes.h

@@ -46,6 +46,7 @@ enum class NodeKind : uint16_t {
   CxxNullPtrExpression,
   IntegerLiteralExpression,
   BoolLiteralExpression,
+  CharacterLiteralExpression,
   IdExpression,
 
   // Statements.
@@ -255,6 +256,17 @@ public:
   syntax::Leaf *nullPtrKeyword();
 };
 
+/// Expression for character literals. C++ [lex.ccon]
+class CharacterLiteralExpression final : public Expression {
+public:
+  CharacterLiteralExpression()
+      : Expression(NodeKind::CharacterLiteralExpression) {}
+  static bool classof(const Node *N) {
+    return N->kind() == NodeKind::CharacterLiteralExpression;
+  }
+  syntax::Leaf *literalToken();
+};
+
 /// Expression for integer literals.
 class IntegerLiteralExpression final : public Expression {
 public:

clang/lib/Tooling/Syntax/BuildTree.cpp

@@ -661,6 +661,13 @@ public:
     return true;
   }
 
+  bool WalkUpFromCharacterLiteral(CharacterLiteral *S) {
+    Builder.markChildToken(S->getLocation(), syntax::NodeRole::LiteralToken);
+    Builder.foldNode(Builder.getExprRange(S),
+                     new (allocator()) syntax::CharacterLiteralExpression, S);
+    return true;
+  }
+
   bool WalkUpFromCXXNullPtrLiteralExpr(CXXNullPtrLiteralExpr *S) {
     Builder.markChildToken(S->getLocation(), syntax::NodeRole::LiteralToken);
     Builder.foldNode(Builder.getExprRange(S),

clang/lib/Tooling/Syntax/Nodes.cpp

@@ -24,6 +24,8 @@ llvm::raw_ostream &syntax::operator<<(llvm::raw_ostream &OS, NodeKind K) {
     return OS << "IntegerLiteralExpression";
   case NodeKind::BoolLiteralExpression:
     return OS << "BoolLiteralExpression";
+  case NodeKind::CharacterLiteralExpression:
+    return OS << "CharacterLiteralExpression";
   case NodeKind::PrefixUnaryOperatorExpression:
     return OS << "PrefixUnaryOperatorExpression";
   case NodeKind::PostfixUnaryOperatorExpression:
@@ -207,6 +209,11 @@ syntax::Leaf *syntax::BoolLiteralExpression::literalToken() {
       findChild(syntax::NodeRole::LiteralToken));
 }
 
+syntax::Leaf *syntax::CharacterLiteralExpression::literalToken() {
+  return llvm::cast_or_null<syntax::Leaf>(
+      findChild(syntax::NodeRole::LiteralToken));
+}
+
 syntax::Leaf *syntax::CxxNullPtrExpression::nullPtrKeyword() {
   return llvm::cast_or_null<syntax::Leaf>(
       findChild(syntax::NodeRole::LiteralToken));

clang/unittests/Tooling/Syntax/TreeTest.cpp

@@ -73,6 +73,10 @@ struct TestClangConfig {
     return Language == Lang_C89 || Language == Lang_C99;
   }
 
+  bool isCXX17OrLater() const {
+    return Language == Lang_CXX17 || Language == Lang_CXX20;
+  }
+
   bool supportsCXXDynamicExceptionSpecification() const {
     return Language == Lang_CXX03 || Language == Lang_CXX11 ||
            Language == Lang_CXX14;
@@ -1232,6 +1236,135 @@ void test() {
 )txt"));
 }
 
+TEST_P(SyntaxTreeTest, CharacterLiteral) {
+  EXPECT_TRUE(treeDumpEqual(
+      R"cpp(
+void test() {
+  'a';
+  '\n';
+  '\x20';
+  '\0';
+  L'a';
+  L'α';
+}
+)cpp",
+      R"txt(
+*: TranslationUnit
+`-SimpleDeclaration
+  |-void
+  |-SimpleDeclarator
+  | |-test
+  | `-ParametersAndQualifiers
+  |   |-(
+  |   `-)
+  `-CompoundStatement
+    |-{
+    |-ExpressionStatement
+    | |-CharacterLiteralExpression
+    | | `-'a'
+    | `-;
+    |-ExpressionStatement
+    | |-CharacterLiteralExpression
+    | | `-'\n'
+    | `-;
+    |-ExpressionStatement
+    | |-CharacterLiteralExpression
+    | | `-'\x20'
+    | `-;
+    |-ExpressionStatement
+    | |-CharacterLiteralExpression
+    | | `-'\0'
+    | `-;
+    |-ExpressionStatement
+    | |-CharacterLiteralExpression
+    | | `-L'a'
+    | `-;
+    |-ExpressionStatement
+    | |-CharacterLiteralExpression
+    | | `-L'α'
+    | `-;
+    `-}
+)txt"));
+}
+
+TEST_P(SyntaxTreeTest, CharacterLiteralUtf) {
+  if (!GetParam().isCXX11OrLater()) {
+    return;
+  }
+  EXPECT_TRUE(treeDumpEqual(
+      R"cpp(
+void test() {
+  u'a';
+  u'構';
+  U'a';
+  U'🌲';
+}
+)cpp",
+      R"txt(
+*: TranslationUnit
+`-SimpleDeclaration
+  |-void
+  |-SimpleDeclarator
+  | |-test
+  | `-ParametersAndQualifiers
+  |   |-(
+  |   `-)
+  `-CompoundStatement
+    |-{
+    |-ExpressionStatement
+    | |-CharacterLiteralExpression
+    | | `-u'a'
+    | `-;
+    |-ExpressionStatement
+    | |-CharacterLiteralExpression
+    | | `-u'構'
+    | `-;
+    |-ExpressionStatement
+    | |-CharacterLiteralExpression
+    | | `-U'a'
+    | `-;
+    |-ExpressionStatement
+    | |-CharacterLiteralExpression
+    | | `-U'🌲'
+    | `-;
+    `-}
+)txt"));
+}
+
+TEST_P(SyntaxTreeTest, CharacterLiteralUtf8) {
+  if (!GetParam().isCXX17OrLater()) {
+    return;
+  }
+  EXPECT_TRUE(treeDumpEqual(
+      R"cpp(
+void test() {
+  u8'a';
+  u8'\x7f';
+}
+)cpp",
+      R"txt(
+*: TranslationUnit
+`-SimpleDeclaration
+  |-void
+  |-SimpleDeclarator
+  | |-test
+  | `-ParametersAndQualifiers
+  |   |-(
+  |   `-)
+  `-CompoundStatement
+    |-{
+    |-ExpressionStatement
+    | |-CharacterLiteralExpression
+    | | `-u8'a'
+    | `-;
+    |-ExpressionStatement
+    | |-CharacterLiteralExpression
+    | | `-u8'\x7f'
+    | `-;
+    `-}
+)txt"));
+}
+
 TEST_P(SyntaxTreeTest, BoolLiteral) {
   if (GetParam().hasBoolType()) {
     return;