[clang-format] Merge name and colon into a single token for C# named arguments

[jbcoe]

Summary

Merge 'argumentName' and ':' into a single token in foo(argumentName: bar).

Add C# named argument as a token type.

[krasimir]

clang/lib/Format/FormatTokenLexer.cpp

189

Please add a comment with an example how is this merge expected to work (like token followed by : is merged etc.)

190

did you mean Tokens.size() < 3? Below, you look at the last 3 tokens from Tokens.

194

You might need to skip over comment while searching for CommaOrLeftParen.
I'm not 100% sure of that, because comment tokens might be handled in a separate token sequence etc.

Could you please check if your code works on this variation of the test case:

  verifyFormat(R"(//
PrintOrderDetails(orderNum: 31, productName: "Red Mug", // comment
                  sellerName: "Gift Shop");)",
               Style);

[jbcoe]

Fix token-length check as 3 tokens are read not 2.

[jbcoe]

Handle comments in named argument lists.

[jbcoe]

remove commented out code

[jbcoe]

clang/lib/Format/FormatTokenLexer.cpp

190

We only take two tokens now without checking that we've not run out.

[krasimir]

Thank you!

clang/lib/Format/FormatTokenLexer.cpp

204

great comment!

clang/lib/Format/FormatToken.h

#define LIST_TOKEN_TYPES                                                       \
   TYPE(ProtoExtensionLSquare)                                                  \
   TYPE(TrailingAnnotation)                                                     \
   TYPE(TrailingReturnArrow)                                                    \
   TYPE(TrailingUnaryOperator)                                                  \
   TYPE(TypenameMacro)                                                          \
   TYPE(UnaryOperator)                                                          \
   TYPE(CSharpStringLiteral)                                                    \
   TYPE(CSharpNullCoalescing)                                                   \
+  TYPE(CSharpNamedArgument)                                                    \
   TYPE(Unknown)
 
 enum TokenType {
 #define TYPE(X) TT_##X,
   LIST_TOKEN_TYPES
 #undef TYPE
       NUM_TOKEN_TYPES
 };

clang/lib/Format/FormatTokenLexer.h

private:
   bool tryMergeNSStringLiteral();
   bool tryMergeJSPrivateIdentifier();
   bool tryMergeCSharpStringLiteral();
   bool tryMergeCSharpKeywordVariables();
   bool tryMergeCSharpNullConditionals();
   bool tryMergeCSharpDoubleQuestion();
   bool tryTransformCSharpForEach();
   bool tryMergeCSharpAttributeAndTarget();
+  bool tryMergeCSharpNamedArgument();
 
   bool tryMergeTokens(ArrayRef<tok::TokenKind> Kinds, TokenType NewType);
 
   // Returns \c true if \p Tok can only be followed by an operand in JavaScript.
   bool precedesOperand(FormatToken *Tok);
 
   bool canPrecedeRegexLiteral(FormatToken *Prev);
 

clang/lib/Format/FormatTokenLexer.cpp

void FormatTokenLexer::tryMergePreviousTokens() {
   if (tryMerge_TMacro())
     return;
   if (tryMergeConflictMarkers())
     return;
   if (tryMergeLessLess())
     return;
 
   if (Style.isCSharp()) {
+    if (tryMergeCSharpNamedArgument())
+      return;
     if (tryMergeCSharpAttributeAndTarget())
       return;
     if (tryMergeCSharpKeywordVariables())
       return;
     if (tryMergeCSharpStringLiteral())
       return;
     if (tryMergeCSharpDoubleQuestion())
       return;
Hash->TokenText =
       StringRef(Hash->TokenText.begin(),
                 Identifier->TokenText.end() - Hash->TokenText.begin());
   Hash->ColumnWidth += Identifier->ColumnWidth;
   Hash->Type = TT_JsPrivateIdentifier;
   Tokens.erase(Tokens.end() - 1);
   return true;
 }
 
+// Merge 'argName' and ':' into a single token in `foo(argName: bar)`.
+bool FormatTokenLexer::tryMergeCSharpNamedArgument() {
+  if (Tokens.size() < 2)
+    return false;
+  auto &Colon = *(Tokens.end() - 1);
+  if (!Colon->is(tok::colon))
+    return false;
+
+  auto &Name = *(Tokens.end() - 2);
+  if (!Name->is(tok::identifier))
+    return false;
+    
+  const FormatToken *CommaOrLeftParen = nullptr;
+  for (auto I = Tokens.rbegin() + 2, E = Tokens.rend(); I != E; ++I) {
+    // NB: Because previous pointers are not initialized yet, this cannot use
+    // Token.getPreviousNonComment.
+    if ((*I)->isNot(tok::comment)) {
+      CommaOrLeftParen = *I;
+      break;
+    }
+  }
+
+  if (!CommaOrLeftParen || !CommaOrLeftParen->isOneOf(tok::l_paren, tok::comma))
+    return false;
+
+  Name->TokenText = StringRef(Name->TokenText.begin(),
+                              Colon->TokenText.end() - Name->TokenText.begin());
+  Name->ColumnWidth += Colon->ColumnWidth;
+  Name->Type = TT_CSharpNamedArgument;
+  Tokens.erase(Tokens.end() - 1);
+  return true;
+}
+
 // Search for verbatim or interpolated string literals @"ABC" or
 // $"aaaaa{abc}aaaaa" i and mark the token as TT_CSharpStringLiteral, and to
 // prevent splitting of @, $ and ".
 // Merging of multiline verbatim strings with embedded '"' is handled in
 // handleCSharpVerbatimAndInterpolatedStrings with lower-level lexing.
 bool FormatTokenLexer::tryMergeCSharpStringLiteral() {
   if (Tokens.size() < 2)
     return false;

clang/unittests/Format/FormatTestCSharp.cpp

Code = R"(Console.WriteLine($"{string.Join(@",
     ", values)}");)";
 
   EXPECT_EQ(Code, format(Code, Style));
 }
 
 TEST_F(FormatTestCSharp, CSharpObjectInitializers) {
   FormatStyle Style = getGoogleStyle(FormatStyle::LK_CSharp);
 
-  // Start code fragemnts with a comment line so that C++ raw string literals
+  // Start code fragments with a comment line so that C++ raw string literals
   // as seen are identical to expected formatted code.
 
   verifyFormat(R"(//
 Shape[] shapes = new[] {
     new Circle {
         Radius = 2.7281,
         Colour = Colours.Red,
     },

 Shape[] shapes = new[] {new Circle {Radius = 2.7281, Colour = Colours.Red},
                         new Square {
                             Side = 101.1,
                             Colour = Colours.Yellow,
                         }};)",
                Style);
 }
 
+TEST_F(FormatTestCSharp, CSharpNamedArguments) {
+  FormatStyle Style = getGoogleStyle(FormatStyle::LK_CSharp);
+
+  verifyFormat(R"(//
+PrintOrderDetails(orderNum: 31, productName: "Red Mug",
+                  sellerName: "Gift Shop");)",
+               Style);
+
+  // Ensure that trailing comments do not cause problems.
+  verifyFormat(R"(//
+PrintOrderDetails(orderNum: 31, productName: "Red Mug",  // comment
+                  sellerName: "Gift Shop");)",
+               Style);
+}
+
 } // namespace format
 } // end namespace clang