[clang-format] Handle common C++ non-keyword types as such (#83709)

Fixes #83400.

7 files changed, 83 insertions, 41 deletions

diff --git a/clang/lib/Format/FormatToken.cpp b/clang/lib/Format/FormatToken.cpp
index 56a7b2d..4fb70ff 100644
--- a/clang/lib/Format/FormatToken.cpp
+++ b/clang/lib/Format/FormatToken.cpp
@@ -71,8 +71,22 @@ bool FormatToken::isSimpleTypeSpecifier() const {
   }
 }
 
-bool FormatToken::isTypeOrIdentifier() const {
-  return isSimpleTypeSpecifier() || Tok.isOneOf(tok::kw_auto, tok::identifier);
+// Sorted common C++ non-keyword types.
+static SmallVector<StringRef> CppNonKeywordTypes = {
+    "clock_t",  "int16_t",   "int32_t", "int64_t",   "int8_t",
+    "intptr_t", "ptrdiff_t", "size_t",  "time_t",    "uint16_t",
+    "uint32_t", "uint64_t",  "uint8_t", "uintptr_t",
+};
+
+bool FormatToken::isTypeName(bool IsCpp) const {
+  return is(TT_TypeName) || isSimpleTypeSpecifier() ||
+         (IsCpp && is(tok::identifier) &&
+          std::binary_search(CppNonKeywordTypes.begin(),
+                             CppNonKeywordTypes.end(), TokenText));
+}
+
+bool FormatToken::isTypeOrIdentifier(bool IsCpp) const {
+  return isTypeName(IsCpp) || isOneOf(tok::kw_auto, tok::identifier);
 }
 
 bool FormatToken::isBlockIndentedInitRBrace(const FormatStyle &Style) const {
diff --git a/clang/lib/Format/FormatToken.h b/clang/lib/Format/FormatToken.h
index 3124549..f4566e4 100644
--- a/clang/lib/Format/FormatToken.h
+++ b/clang/lib/Format/FormatToken.h
@@ -676,7 +676,9 @@ public:
   /// Determine whether the token is a simple-type-specifier.
   [[nodiscard]] bool isSimpleTypeSpecifier() const;
 
-  [[nodiscard]] bool isTypeOrIdentifier() const;
+  [[nodiscard]] bool isTypeName(bool IsCpp) const;
+
+  [[nodiscard]] bool isTypeOrIdentifier(bool IsCpp) const;
 
   bool isObjCAccessSpecifier() const {
     return is(tok::at) && Next &&
diff --git a/clang/lib/Format/QualifierAlignmentFixer.cpp b/clang/lib/Format/QualifierAlignmentFixer.cpp
index 0c63d33..c263530 100644
--- a/clang/lib/Format/QualifierAlignmentFixer.cpp
+++ b/clang/lib/Format/QualifierAlignmentFixer.cpp
@@ -268,11 +268,13 @@ const FormatToken *LeftRightQualifierAlignmentFixer::analyzeRight(
   if (isPossibleMacro(TypeToken))
     return Tok;
 
+  const bool IsCpp = Style.isCpp();
+
   // The case `const long long int volatile` -> `long long int const volatile`
   // The case `long const long int volatile` -> `long long int const volatile`
   // The case `long long volatile int const` -> `long long int const volatile`
   // The case `const long long volatile int` -> `long long int const volatile`
-  if (TypeToken->isSimpleTypeSpecifier()) {
+  if (TypeToken->isTypeName(IsCpp)) {
     // The case `const decltype(foo)` -> `const decltype(foo)`
     // The case `const typeof(foo)` -> `const typeof(foo)`
     // The case `const _Atomic(foo)` -> `const _Atomic(foo)`
@@ -280,8 +282,10 @@ const FormatToken *LeftRightQualifierAlignmentFixer::analyzeRight(
       return Tok;
 
     const FormatToken *LastSimpleTypeSpecifier = TypeToken;
-    while (isQualifierOrType(LastSimpleTypeSpecifier->getNextNonComment()))
+    while (isQualifierOrType(LastSimpleTypeSpecifier->getNextNonComment(),
+                             IsCpp)) {
       LastSimpleTypeSpecifier = LastSimpleTypeSpecifier->getNextNonComment();
+    }
 
     rotateTokens(SourceMgr, Fixes, Tok, LastSimpleTypeSpecifier,
                  /*Left=*/false);
@@ -291,7 +295,7 @@ const FormatToken *LeftRightQualifierAlignmentFixer::analyzeRight(
   // The case  `unsigned short const` -> `unsigned short const`
   // The case:
   // `unsigned short volatile const` -> `unsigned short const volatile`
-  if (PreviousCheck && PreviousCheck->isSimpleTypeSpecifier()) {
+  if (PreviousCheck && PreviousCheck->isTypeName(IsCpp)) {
     if (LastQual != Tok)
       rotateTokens(SourceMgr, Fixes, Tok, LastQual, /*Left=*/false);
     return Tok;
@@ -408,11 +412,11 @@ const FormatToken *LeftRightQualifierAlignmentFixer::analyzeLeft(
   // The case `volatile long long const int` -> `const volatile long long int`
   // The case `const long long volatile int` -> `const volatile long long int`
   // The case `long volatile long int const` -> `const volatile long long int`
-  if (TypeToken->isSimpleTypeSpecifier()) {
+  if (const bool IsCpp = Style.isCpp(); TypeToken->isTypeName(IsCpp)) {
     const FormatToken *LastSimpleTypeSpecifier = TypeToken;
     while (isConfiguredQualifierOrType(
         LastSimpleTypeSpecifier->getPreviousNonComment(),
-        ConfiguredQualifierTokens)) {
+        ConfiguredQualifierTokens, IsCpp)) {
       LastSimpleTypeSpecifier =
           LastSimpleTypeSpecifier->getPreviousNonComment();
     }
@@ -610,16 +614,16 @@ void prepareLeftRightOrderingForQualifierAlignmentFixer(
   }
 }
 
-bool LeftRightQualifierAlignmentFixer::isQualifierOrType(
-    const FormatToken *const Tok) {
-  return Tok && (Tok->isSimpleTypeSpecifier() || Tok->is(tok::kw_auto) ||
-                 isQualifier(Tok));
+bool LeftRightQualifierAlignmentFixer::isQualifierOrType(const FormatToken *Tok,
+                                                         bool IsCpp) {
+  return Tok &&
+         (Tok->isTypeName(IsCpp) || Tok->is(tok::kw_auto) || isQualifier(Tok));
 }
 
 bool LeftRightQualifierAlignmentFixer::isConfiguredQualifierOrType(
-    const FormatToken *const Tok,
-    const std::vector<tok::TokenKind> &Qualifiers) {
-  return Tok && (Tok->isSimpleTypeSpecifier() || Tok->is(tok::kw_auto) ||
+    const FormatToken *Tok, const std::vector<tok::TokenKind> &Qualifiers,
+    bool IsCpp) {
+  return Tok && (Tok->isTypeName(IsCpp) || Tok->is(tok::kw_auto) ||
                  isConfiguredQualifier(Tok, Qualifiers));
 }
 
diff --git a/clang/lib/Format/QualifierAlignmentFixer.h b/clang/lib/Format/QualifierAlignmentFixer.h
index e922d80..e1cc27e 100644
--- a/clang/lib/Format/QualifierAlignmentFixer.h
+++ b/clang/lib/Format/QualifierAlignmentFixer.h
@@ -71,10 +71,11 @@ public:
                                  tok::TokenKind QualifierType);
 
   // Is the Token a simple or qualifier type
-  static bool isQualifierOrType(const FormatToken *Tok);
+  static bool isQualifierOrType(const FormatToken *Tok, bool IsCpp = true);
   static bool
   isConfiguredQualifierOrType(const FormatToken *Tok,
-                              const std::vector<tok::TokenKind> &Qualifiers);
+                              const std::vector<tok::TokenKind> &Qualifiers,
+                              bool IsCpp = true);
 
   // Is the Token likely a Macro
   static bool isPossibleMacro(const FormatToken *Tok);
diff --git a/clang/lib/Format/TokenAnnotator.cpp b/clang/lib/Format/TokenAnnotator.cpp
index 04f0374..3a55106 100644
--- a/clang/lib/Format/TokenAnnotator.cpp
+++ b/clang/lib/Format/TokenAnnotator.cpp
@@ -562,7 +562,7 @@ private:
           (CurrentToken->is(tok::l_paren) && CurrentToken->Next &&
            CurrentToken->Next->isOneOf(tok::star, tok::amp, tok::caret));
       if ((CurrentToken->Previous->isOneOf(tok::kw_const, tok::kw_auto) ||
-           CurrentToken->Previous->isSimpleTypeSpecifier()) &&
+           CurrentToken->Previous->isTypeName(IsCpp)) &&
           !(CurrentToken->is(tok::l_brace) ||
             (CurrentToken->is(tok::l_paren) && !ProbablyFunctionTypeLParen))) {
         Contexts.back().IsExpression = false;
@@ -2573,7 +2573,7 @@ private:
       return true;
 
     // MyClass a;
-    if (PreviousNotConst->isSimpleTypeSpecifier())
+    if (PreviousNotConst->isTypeName(IsCpp))
       return true;
 
     // type[] a in Java
@@ -2704,9 +2704,9 @@ private:
     }
 
     // Heuristically try to determine whether the parentheses contain a type.
-    auto IsQualifiedPointerOrReference = [](FormatToken *T) {
+    auto IsQualifiedPointerOrReference = [this](FormatToken *T) {
       // This is used to handle cases such as x = (foo *const)&y;
-      assert(!T->isSimpleTypeSpecifier() && "Should have already been checked");
+      assert(!T->isTypeName(IsCpp) && "Should have already been checked");
       // Strip trailing qualifiers such as const or volatile when checking
       // whether the parens could be a cast to a pointer/reference type.
       while (T) {
@@ -2738,7 +2738,7 @@ private:
     bool ParensAreType =
         !Tok.Previous ||
         Tok.Previous->isOneOf(TT_TemplateCloser, TT_TypeDeclarationParen) ||
-        Tok.Previous->isSimpleTypeSpecifier() ||
+        Tok.Previous->isTypeName(IsCpp) ||
         IsQualifiedPointerOrReference(Tok.Previous);
     bool ParensCouldEndDecl =
         Tok.Next->isOneOf(tok::equal, tok::semi, tok::l_brace, tok::greater);
@@ -3595,7 +3595,8 @@ static bool isFunctionDeclarationName(bool IsCpp, const FormatToken &Current,
   if (!Current.Tok.getIdentifierInfo())
     return false;
 
-  auto skipOperatorName = [](const FormatToken *Next) -> const FormatToken * {
+  auto skipOperatorName =
+      [IsCpp](const FormatToken *Next) -> const FormatToken * {
     for (; Next; Next = Next->Next) {
       if (Next->is(TT_OverloadedOperatorLParen))
         return Next;
@@ -3614,7 +3615,7 @@ static bool isFunctionDeclarationName(bool IsCpp, const FormatToken &Current,
         Next = Next->Next;
         continue;
       }
-      if ((Next->isSimpleTypeSpecifier() || Next->is(tok::identifier)) &&
+      if ((Next->isTypeName(IsCpp) || Next->is(tok::identifier)) &&
           Next->Next && Next->Next->isPointerOrReference()) {
         // For operator void*(), operator char*(), operator Foo*().
         Next = Next->Next;
@@ -3712,9 +3713,8 @@ static bool isFunctionDeclarationName(bool IsCpp, const FormatToken &Current,
       Tok = Tok->MatchingParen;
       continue;
     }
-    if (Tok->is(tok::kw_const) || Tok->isSimpleTypeSpecifier() ||
-        Tok->isOneOf(TT_PointerOrReference, TT_StartOfName, tok::ellipsis,
-                     TT_TypeName)) {
+    if (Tok->is(tok::kw_const) || Tok->isTypeName(IsCpp) ||
+        Tok->isOneOf(TT_PointerOrReference, TT_StartOfName, tok::ellipsis)) {
       return true;
     }
     if (Tok->isOneOf(tok::l_brace, TT_ObjCMethodExpr) || Tok->Tok.isLiteral())
@@ -4376,7 +4376,7 @@ bool TokenAnnotator::spaceRequiredBetween(const AnnotatedLine &Line,
     if (Left.Tok.isLiteral())
       return true;
     // for (auto a = 0, b = 0; const auto & c : {1, 2, 3})
-    if (Left.isTypeOrIdentifier() && Right.Next && Right.Next->Next &&
+    if (Left.isTypeOrIdentifier(IsCpp) && Right.Next && Right.Next->Next &&
         Right.Next->Next->is(TT_RangeBasedForLoopColon)) {
       return getTokenPointerOrReferenceAlignment(Right) !=
              FormatStyle::PAS_Left;
@@ -4419,8 +4419,8 @@ bool TokenAnnotator::spaceRequiredBetween(const AnnotatedLine &Line,
     if (Right.is(tok::l_brace) && Right.is(BK_Block))
       return true;
     // for (auto a = 0, b = 0; const auto& c : {1, 2, 3})
-    if (Left.Previous && Left.Previous->isTypeOrIdentifier() && Right.Next &&
-        Right.Next->is(TT_RangeBasedForLoopColon)) {
+    if (Left.Previous && Left.Previous->isTypeOrIdentifier(IsCpp) &&
+        Right.Next && Right.Next->is(TT_RangeBasedForLoopColon)) {
       return getTokenPointerOrReferenceAlignment(Left) !=
              FormatStyle::PAS_Right;
     }
@@ -4458,7 +4458,7 @@ bool TokenAnnotator::spaceRequiredBetween(const AnnotatedLine &Line,
   if (Right.isPointerOrReference()) {
     const FormatToken *Previous = &Left;
     while (Previous && Previous->isNot(tok::kw_operator)) {
-      if (Previous->is(tok::identifier) || Previous->isSimpleTypeSpecifier()) {
+      if (Previous->is(tok::identifier) || Previous->isTypeName(IsCpp)) {
         Previous = Previous->getPreviousNonComment();
         continue;
       }
@@ -4647,7 +4647,7 @@ bool TokenAnnotator::spaceRequiredBetween(const AnnotatedLine &Line,
   if (!Style.isVerilog() &&
       (Left.isOneOf(tok::identifier, tok::greater, tok::r_square,
                     tok::r_paren) ||
-       Left.isSimpleTypeSpecifier()) &&
+       Left.isTypeName(IsCpp)) &&
       Right.is(tok::l_brace) && Right.getNextNonComment() &&
       Right.isNot(BK_Block)) {
     return false;
diff --git a/clang/lib/Format/UnwrappedLineParser.cpp b/clang/lib/Format/UnwrappedLineParser.cpp
index 2ce291d..e3f0af1 100644
--- a/clang/lib/Format/UnwrappedLineParser.cpp
+++ b/clang/lib/Format/UnwrappedLineParser.cpp
@@ -1865,8 +1865,7 @@ void UnwrappedLineParser::parseStructuralElement(
     case tok::caret:
       nextToken();
       // Block return type.
-      if (FormatTok->Tok.isAnyIdentifier() ||
-          FormatTok->isSimpleTypeSpecifier()) {
+      if (FormatTok->Tok.isAnyIdentifier() || FormatTok->isTypeName(IsCpp)) {
         nextToken();
         // Return types: pointers are ok too.
         while (FormatTok->is(tok::star))
@@ -2222,7 +2221,7 @@ bool UnwrappedLineParser::tryToParseLambda() {
   bool InTemplateParameterList = false;
 
   while (FormatTok->isNot(tok::l_brace)) {
-    if (FormatTok->isSimpleTypeSpecifier()) {
+    if (FormatTok->isTypeName(IsCpp)) {
       nextToken();
       continue;
     }
@@ -3415,7 +3414,7 @@ bool clang::format::UnwrappedLineParser::parseRequires() {
     break;
   }
   default:
-    if (PreviousNonComment->isTypeOrIdentifier()) {
+    if (PreviousNonComment->isTypeOrIdentifier(IsCpp)) {
       // This is a requires clause.
       parseRequiresClause(RequiresToken);
       return true;
@@ -3478,7 +3477,7 @@ bool clang::format::UnwrappedLineParser::parseRequires() {
       --OpenAngles;
       break;
     default:
-      if (NextToken->isSimpleTypeSpecifier()) {
+      if (NextToken->isTypeName(IsCpp)) {
         FormatTok = Tokens->setPosition(StoredPosition);
         parseRequiresExpression(RequiresToken);
         return false;
@@ -3962,8 +3961,8 @@ void UnwrappedLineParser::parseRecord(bool ParseAsExpr) {
       }
       if (FormatTok->is(tok::l_square)) {
         FormatToken *Previous = FormatTok->Previous;
-        if (!Previous ||
-            !(Previous->is(tok::r_paren) || Previous->isTypeOrIdentifier())) {
+        if (!Previous || (Previous->isNot(tok::r_paren) &&
+                          !Previous->isTypeOrIdentifier(IsCpp))) {
           // Don't try parsing a lambda if we had a closing parenthesis before,
           // it was probably a pointer to an array: int (*)[].
           if (!tryToParseLambda())
diff --git a/clang/unittests/Format/TokenAnnotatorTest.cpp b/clang/unittests/Format/TokenAnnotatorTest.cpp
index c736dac..fcba0e6 100644
--- a/clang/unittests/Format/TokenAnnotatorTest.cpp
+++ b/clang/unittests/Format/TokenAnnotatorTest.cpp
@@ -620,6 +620,24 @@ TEST_F(TokenAnnotatorTest, UnderstandsCasts) {
   ASSERT_EQ(Tokens.size(), 8u) << Tokens;
   EXPECT_TOKEN(Tokens[3], tok::r_paren, TT_Unknown);
   EXPECT_TOKEN(Tokens[4], tok::amp, TT_BinaryOperator);
+
+  Tokens = annotate("#define FOO(bar) foo((uint64_t)&bar)");
+  ASSERT_EQ(Tokens.size(), 15u) << Tokens;
+  EXPECT_TOKEN(Tokens[10], tok::r_paren, TT_CastRParen);
+  EXPECT_TOKEN(Tokens[11], tok::amp, TT_UnaryOperator);
+
+  Tokens = annotate("#define FOO(bar) foo((Foo) & bar)");
+  ASSERT_EQ(Tokens.size(), 15u) << Tokens;
+  EXPECT_TOKEN(Tokens[10], tok::r_paren, TT_Unknown);
+  EXPECT_TOKEN(Tokens[11], tok::amp, TT_BinaryOperator);
+
+  auto Style = getLLVMStyle();
+  Style.TypeNames.push_back("Foo");
+  Tokens = annotate("#define FOO(bar) foo((Foo)&bar)", Style);
+  ASSERT_EQ(Tokens.size(), 15u) << Tokens;
+  EXPECT_TOKEN(Tokens[9], tok::identifier, TT_TypeName);
+  EXPECT_TOKEN(Tokens[10], tok::r_paren, TT_CastRParen);
+  EXPECT_TOKEN(Tokens[11], tok::amp, TT_UnaryOperator);
 }
 
 TEST_F(TokenAnnotatorTest, UnderstandsDynamicExceptionSpecifier) {
@@ -1751,9 +1769,13 @@ TEST_F(TokenAnnotatorTest, UnderstandsFunctionDeclarationNames) {
   EXPECT_TOKEN(Tokens[3], tok::identifier, TT_Unknown);
   EXPECT_TOKEN(Tokens[4], tok::l_paren, TT_FunctionTypeLParen);
 
+  Tokens = annotate("int iso_time(time_t);");
+  ASSERT_EQ(Tokens.size(), 7u) << Tokens;
+  EXPECT_TOKEN(Tokens[1], tok::identifier, TT_FunctionDeclarationName);
+
   auto Style = getLLVMStyle();
-  Style.TypeNames.push_back("time_t");
-  Tokens = annotate("int iso_time(time_t);", Style);
+  Style.TypeNames.push_back("MyType");
+  Tokens = annotate("int iso_time(MyType);", Style);
   ASSERT_EQ(Tokens.size(), 7u) << Tokens;
   EXPECT_TOKEN(Tokens[1], tok::identifier, TT_FunctionDeclarationName);
   EXPECT_TOKEN(Tokens[3], tok::identifier, TT_TypeName);