[clang-format] option to control bin-packing keyworded parameters

The Q_PROPERTY declaration is almost like a function declaration, but uses keywords as parameter separators. This allows users to provide list of those keywords to be used to control bin-packing of the macro parameters.

Author: Eugene Shalygin

clang/docs/ClangFormatStyleOptions.rst

@@ -4740,6 +4740,38 @@ the configuration (without a prefix: ``Auto``).
  replaced with a single newline and form feed followed by the remaining
  newlines.
 
+.. _KeywordedFunctionLikeMacros:
+
+**KeywordedFunctionLikeMacros** (``List of KeywordedFunctionLikeMacros``) :versionbadge:`clang-format 21` :ref:`¶ <KeywordedFunctionLikeMacros>`
+ Allows to format function-like macros with keyworded parameters according
+ to the BinPackParameters setting, treating keywords as parameter
+ separators.
+
+ Q_PROPERTY is an example of such a macro:
+
+ .. code-block:: c++
+
+ Q_PROPERTY(int name READ name WRITE setName NOTIFY nameChanged)
+
+ With ``BinPackParameters`` set to ``OnePerLine`` (or
+ ``AlwaysOnePerLine``) and
+
+ .. code-block:: yaml
+
+ KeywordedFunctionLikeMacros:
+ - Name: "Q_PROPERTY"
+ Keywords: ['READ', 'WRITE', 'MEMBER', 'RESET', 'NOTIFY']
+
+ the line above will be split on these keywords:
+
+ .. code-block:: c++
+
+ Q_PROPERTY(
+ int name
+ READ name
+ WRITE setName
+ NOTIFY nameChanged)
+
 .. _LambdaBodyIndentation:
 
 **LambdaBodyIndentation** (``LambdaBodyIndentationKind``) :versionbadge:`clang-format 13` :ref:`¶ <LambdaBodyIndentation>`

clang/docs/tools/dump_format_style.py

@@ -462,6 +462,7 @@ class State:
  "std::string",
  "std::vector<std::string>",
  "std::vector<IncludeCategory>",
+ "std::vector<KeywordedFunctionLikeMacro>",
  "std::vector<RawStringFormat>",
  "std::optional<unsigned>",
  "deprecated",

clang/docs/tools/plurals.txt

@@ -1,3 +1,4 @@
 Strings
 IncludeCategories
+KeywordedFunctionLikeMacros
 RawStringFormats

clang/include/clang/Format/Format.h

@@ -3276,6 +3276,45 @@ struct FormatStyle {
  /// \version 20
  bool KeepFormFeed;
 
+ /// Function-like declaration with keyworded parameters.
+ /// Lists possible keywords for a named function-like macro.
+ struct KeywordedFunctionLikeMacro {
+ std::string Name;
+ std::vector<std::string> Keywords;
+
+ bool operator==(const KeywordedFunctionLikeMacro &Other) const {
+ return Name == Other.Name && Keywords == Other.Keywords;
+ }
+ };
+
+ /// Allows to format function-like macros with keyworded parameters according
+ /// to the BinPackParameters setting, treating keywords as parameter
+ /// separators.
+ ///
+ /// Q_PROPERTY is an example of such a macro:
+ /// \code
+ /// Q_PROPERTY(int name READ name WRITE setName NOTIFY nameChanged)
+ /// \endcode
+ ///
+ /// With ``BinPackParameters`` set to ``OnePerLine`` (or
+ /// ``AlwaysOnePerLine``) and
+ /// \code{.yaml}
+ /// KeywordedFunctionLikeMacros:
+ /// - Name: "Q_PROPERTY"
+ /// Keywords: ['READ', 'WRITE', 'MEMBER', 'RESET', 'NOTIFY']
+ /// \endcode
+ ///
+ /// the line above will be split on these keywords:
+ /// \code
+ /// Q_PROPERTY(
+ /// int name
+ /// READ name
+ /// WRITE setName
+ /// NOTIFY nameChanged)
+ /// \endcode
+ /// \version 21
+ std::vector<KeywordedFunctionLikeMacro> KeywordedFunctionLikeMacros;
+
  /// Indentation logic for lambda bodies.
  enum LambdaBodyIndentationKind : int8_t {
  /// Align lambda body relative to the lambda signature. This is the default.
@@ -5352,6 +5391,7 @@ struct FormatStyle {
  JavaScriptWrapImports == R.JavaScriptWrapImports &&
  KeepEmptyLines == R.KeepEmptyLines &&
  KeepFormFeed == R.KeepFormFeed && Language == R.Language &&
+ KeywordedFunctionLikeMacros == R.KeywordedFunctionLikeMacros &&
  LambdaBodyIndentation == R.LambdaBodyIndentation &&
  LineEnding == R.LineEnding && MacroBlockBegin == R.MacroBlockBegin &&
  MacroBlockEnd == R.MacroBlockEnd && Macros == R.Macros &&

clang/lib/Format/ContinuationIndenter.cpp

@@ -349,6 +349,10 @@ bool ContinuationIndenter::canBreak(const LineState &State) {
  }
  }
 
+ // Don't break between function parameter keywords and parameter names.
+ if (Previous.is(TT_FunctionParameterKeyword) && Current.is(TT_StartOfName))
+ return false;
+
  // Don't allow breaking before a closing brace of a block-indented braced list
  // initializer if there isn't already a break.
  if (Current.is(tok::r_brace) && Current.MatchingParen &&

clang/lib/Format/Format.cpp

@@ -28,6 +28,7 @@
 
 using clang::format::FormatStyle;
 
+LLVM_YAML_IS_SEQUENCE_VECTOR(FormatStyle::KeywordedFunctionLikeMacro)
 LLVM_YAML_IS_SEQUENCE_VECTOR(FormatStyle::RawStringFormat)
 
 namespace llvm {
@@ -399,6 +400,14 @@ template <> struct MappingTraits<FormatStyle::KeepEmptyLinesStyle> {
  }
 };
 
+template <> struct MappingTraits<FormatStyle::KeywordedFunctionLikeMacro> {
+ static void mapping(IO &IO,
+ FormatStyle::KeywordedFunctionLikeMacro &Function) {
+ IO.mapOptional("Name", Function.Name);
+ IO.mapOptional("Keywords", Function.Keywords);
+ }
+};
+
 template <> struct ScalarEnumerationTraits<FormatStyle::LanguageKind> {
  static void enumeration(IO &IO, FormatStyle::LanguageKind &Value) {
  IO.enumCase(Value, "C", FormatStyle::LK_C);
@@ -1072,6 +1081,8 @@ template <> struct MappingTraits<FormatStyle> {
  IO.mapOptional("JavaScriptWrapImports", Style.JavaScriptWrapImports);
  IO.mapOptional("KeepEmptyLines", Style.KeepEmptyLines);
  IO.mapOptional("KeepFormFeed", Style.KeepFormFeed);
+ IO.mapOptional("KeywordedFunctionLikeMacros",
+ Style.KeywordedFunctionLikeMacros);
  IO.mapOptional("LambdaBodyIndentation", Style.LambdaBodyIndentation);
  IO.mapOptional("LineEnding", Style.LineEnding);
  IO.mapOptional("MacroBlockBegin", Style.MacroBlockBegin);

clang/lib/Format/FormatToken.cpp

@@ -331,6 +331,8 @@ bool startsNextParameter(const FormatToken &Current, const FormatStyle &Style) {
  }
  if (Style.Language == FormatStyle::LK_Proto && Current.is(TT_SelectorName))
  return true;
+ if (Current.is(TT_FunctionParameterKeyword))
+ return true;
  return Previous.is(tok::comma) && !Current.isTrailingComment() &&
  ((Previous.isNot(TT_CtorInitializerComma) ||
  Style.BreakConstructorInitializers !=

clang/lib/Format/FormatToken.h

@@ -84,6 +84,7 @@ namespace format {
  TYPE(FunctionDeclarationLParen) \
  TYPE(FunctionLBrace) \
  TYPE(FunctionLikeOrFreestandingMacro) \
+ TYPE(FunctionParameterKeyword) \
  TYPE(FunctionTypeLParen) \
  /* The colons as part of a C11 _Generic selection */ \
  TYPE(GenericSelectionColon) \

clang/lib/Format/TokenAnnotator.cpp

@@ -116,6 +116,16 @@ static bool isCppAttribute(bool IsCpp, const FormatToken &Tok) {
  return AttrTok && AttrTok->startsSequence(tok::r_square, tok::r_square);
 }
 
+static bool isParametersKeyword(
+ const FormatToken &Tok,
+ const FormatStyle::KeywordedFunctionLikeMacro *declaration) {
+ if (!declaration)
+ return false;
+
+ return std::find(declaration->Keywords.begin(), declaration->Keywords.end(),
+ Tok.TokenText) != declaration->Keywords.end();
+}
+
 /// A parser that gathers additional information about tokens.
 ///
 /// The \c TokenAnnotator tries to match parenthesis and square brakets and
@@ -148,6 +158,32 @@ class AnnotatingParser {
  }
  }
 
+ const FormatStyle::KeywordedFunctionLikeMacro *
+ findKeywordedFunctionLikeMacro() const {
+ const FormatToken *TokBeforeFirstLParent = nullptr;
+ for (const FormatToken *T = Line.First; T != Line.Last; T = T->Next) {
+ if (T->Tok.is(tok::l_paren)) {
+ TokBeforeFirstLParent = T->getPreviousNonComment();
+ break;
+ }
+ }
+
+ // Unknown if line ends with ';', FunctionLikeOrFreestandingMacro otherwise.
+ if (!TokBeforeFirstLParent ||
+ !TokBeforeFirstLParent->isOneOf(TT_FunctionLikeOrFreestandingMacro,
+ TT_Unknown)) {
+ return nullptr;
+ }
+ auto I = std::find_if(
+ Style.KeywordedFunctionLikeMacros.begin(),
+ Style.KeywordedFunctionLikeMacros.end(),
+ [TokBeforeFirstLParent](
+ const FormatStyle::KeywordedFunctionLikeMacro &Declaration) {
+ return TokBeforeFirstLParent->TokenText == Declaration.Name;
+ });
+ return I != Style.KeywordedFunctionLikeMacros.end() ? &*I : nullptr;
+ }
+
  bool parseAngle() {
  if (!CurrentToken)
  return false;
@@ -2415,8 +2451,12 @@ class AnnotatingParser {
  Current.setType(TT_BinaryOperator);
  } else if (isStartOfName(Current) &&
  (!Line.MightBeFunctionDecl || Current.NestingLevel != 0)) {
- Contexts.back().FirstStartOfName = &Current;
- Current.setType(TT_StartOfName);
+ if (isParametersKeyword(Current, findKeywordedFunctionLikeMacro())) {
+ Current.setType(TT_FunctionParameterKeyword);
+ } else {
+ Contexts.back().FirstStartOfName = &Current;
+ Current.setType(TT_StartOfName);
+ }
  } else if (Current.is(tok::semi)) {
  // Reset FirstStartOfName after finding a semicolon so that a for loop
  // with multiple increment statements is not confused with a for loop
@@ -6231,7 +6271,8 @@ bool TokenAnnotator::canBreakBefore(const AnnotatedLine &Line,
  Right.Next->isOneOf(TT_FunctionDeclarationName, tok::kw_const)));
  }
  if (Right.isOneOf(TT_StartOfName, TT_FunctionDeclarationName,
- TT_ClassHeadName, tok::kw_operator)) {
+ TT_FunctionParameterKeyword, TT_ClassHeadName,
+ tok::kw_operator)) {
  return true;
  }
  if (Right.isAttribute())

clang/unittests/Format/ConfigParseTest.cpp

@@ -1105,6 +1105,16 @@ TEST(ConfigParseTest, ParsesConfiguration) {
  FormatStyle::SDS_Leave);
  CHECK_PARSE("SeparateDefinitionBlocks: Never", SeparateDefinitionBlocks,
  FormatStyle::SDS_Never);
+
+ Style.KeywordedFunctionLikeMacros.clear();
+ std::vector<FormatStyle::KeywordedFunctionLikeMacro> ExpectedFunctions = {
+ {"MACRO_A", {"PARAM1", "KEYWORD", "KW_2"}}, {"macro", {"mKW1", "mKW2"}}};
+ CHECK_PARSE("KeywordedFunctionLikeMacros:\n"
+ " - Name: MACRO_A\n"
+ " Keywords: ['PARAM1', 'KEYWORD', 'KW_2']\n"
+ " - Name: macro\n"
+ " Keywords: [ \"mKW1\", \"mKW2\" ]\n",
+ KeywordedFunctionLikeMacros, ExpectedFunctions);
 }
 
 TEST(ConfigParseTest, ParsesConfigurationWithLanguages) {