Lex: add a callback for `#pragma mark`

[compnerd]

Summary

Allow a preprocessor observer to be notified of mark pragmas. Although
this does not impact code generation in any way, it is useful for other
clients, such as clangd, to be able to identify any marked regions.

[dgoldman]

clang/unittests/Lex/PPCallbacksTest.cpp

119

should this be an std::string? not sure of the lifetime guarantees here

[compnerd]

Correct lifetime of data in the tests.

clang/include/clang/Lex/PPCallbacks.h

virtual void PragmaDirective(SourceLocation Loc,
                                PragmaIntroducerKind Introducer) {
   }
 
   /// Callback invoked when a \#pragma comment directive is read.
   virtual void PragmaComment(SourceLocation Loc, const IdentifierInfo *Kind,
                              StringRef Str) {
   }
 
+  /// Callback invoked when a \#pragma mark comment is read.
+  virtual void PragmaMark(SourceLocation Loc, StringRef Trivia) {
+  }
+
   /// Callback invoked when a \#pragma detect_mismatch directive is
   /// read.
   virtual void PragmaDetectMismatch(SourceLocation Loc, StringRef Name,
                                     StringRef Value) {
   }
 
   /// Callback invoked when a \#pragma clang __debug directive is read.
   /// \param Loc The location of the debug directive.

clang/include/clang/Lex/Preprocessor.h

private:
   void HandleElifFamilyDirective(Token &ElifToken, const Token &HashToken,
                                  tok::PPKeywordKind Kind);
 
   // Pragmas.
   void HandlePragmaDirective(PragmaIntroducer Introducer);
 
 public:
   void HandlePragmaOnce(Token &OnceTok);
-  void HandlePragmaMark();
+  void HandlePragmaMark(Token &MarkTok);
   void HandlePragmaPoison();
   void HandlePragmaSystemHeader(Token &SysHeaderTok);
   void HandlePragmaDependency(Token &DependencyTok);
   void HandlePragmaPushMacro(Token &Tok);
   void HandlePragmaPopMacro(Token &Tok);
   void HandlePragmaIncludeAlias(Token &Tok);
   void HandlePragmaModuleBuild(Token &Tok);
   void HandlePragmaHdrstop(Token &Tok);

clang/lib/Lex/Pragma.cpp

if (isInPrimaryFile() && TUKind != TU_Prefix && !getLangOpts().IsHeaderFile) {
     return;
   }
 
   // Get the current file lexer we're looking at.  Ignore _Pragma 'files' etc.
   // Mark the file as a once-only file now.
   HeaderInfo.MarkFileIncludeOnce(getCurrentFileLexer()->getFileEntry());
 }
 
-void Preprocessor::HandlePragmaMark() {
+void Preprocessor::HandlePragmaMark(Token &MarkTok) {
   assert(CurPPLexer && "No current lexer?");
-  CurLexer->ReadToEndOfLine();
+
+  SmallString<64> Buffer;
+  CurLexer->ReadToEndOfLine(&Buffer);
+  if (Callbacks)
+    Callbacks->PragmaMark(MarkTok.getLocation(), Buffer);
 }
 
 /// HandlePragmaPoison - Handle \#pragma GCC poison.  PoisonTok is the 'poison'.
 void Preprocessor::HandlePragmaPoison() {
   Token Tok;
 
   while (true) {
     // Read the next token to poison.  While doing this, pretend that we are

 
 /// PragmaMarkHandler - "\#pragma mark ..." is ignored by the compiler, and the
 /// rest of the line is not lexed.
 struct PragmaMarkHandler : public PragmaHandler {
   PragmaMarkHandler() : PragmaHandler("mark") {}
 
   void HandlePragma(Preprocessor &PP, PragmaIntroducer Introducer,
                     Token &MarkTok) override {
-    PP.HandlePragmaMark();
+    PP.HandlePragmaMark(MarkTok);
   }
 };
 
 /// PragmaPoisonHandler - "\#pragma poison x" marks x as not usable.
 struct PragmaPoisonHandler : public PragmaHandler {
   PragmaPoisonHandler() : PragmaHandler("poison") {}
 
   void HandlePragma(Preprocessor &PP, PragmaIntroducer Introducer,

clang/unittests/Lex/PPCallbacksTest.cpp

public:
   }
 
   SourceLocation NameLoc;
   SmallString<16> Name;
   SourceLocation StateLoc;
   unsigned State;
 };
 
+class PragmaMarkCallbacks : public PPCallbacks {
+public:
+  struct Mark {
+    SourceLocation Location;
+    std::string Trivia;
+  };
+
+  std::vector<Mark> Marks;
+
+  void PragmaMark(SourceLocation Loc, StringRef Trivia) override {
+    Marks.emplace_back(Mark{Loc, Trivia.str()});
+  }
+};
+
 // PPCallbacks test fixture.
 class PPCallbacksTest : public ::testing::Test {
 protected:
   PPCallbacksTest()
       : InMemoryFileSystem(new llvm::vfs::InMemoryFileSystem),
         FileMgr(FileSystemOptions(), InMemoryFileSystem),
         DiagID(new DiagnosticIDs()), DiagOpts(new DiagnosticOptions()),
         Diags(DiagID, DiagOpts.get(), new IgnoringDiagConsumer()),
while (true) {
       PP.Lex(Tok);
       if (Tok.is(tok::eof))
         break;
     }
 
     return Callbacks->Results;
   }
 
+  std::vector<PragmaMarkCallbacks::Mark>
+  PragmaMarkCall(const char *SourceText) {
+    std::unique_ptr<llvm::MemoryBuffer> SourceBuf =
+        llvm::MemoryBuffer::getMemBuffer(SourceText, "test.c");
+    SourceMgr.setMainFileID(SourceMgr.createFileID(std::move(SourceBuf)));
+
+    HeaderSearch HeaderInfo(std::make_shared<HeaderSearchOptions>(), SourceMgr,
+                            Diags, LangOpts, Target.get());
+    TrivialModuleLoader ModLoader;
+
+    Preprocessor PP(std::make_shared<PreprocessorOptions>(), Diags, LangOpts,
+                    SourceMgr, HeaderInfo, ModLoader, /*IILookup=*/nullptr,
+                    /*OwnsHeaderSearch=*/false);
+    PP.Initialize(*Target);
+
+    auto *Callbacks = new PragmaMarkCallbacks;
+    PP.addPPCallbacks(std::unique_ptr<PPCallbacks>(Callbacks));
+
+    // Lex source text.
+    PP.EnterMainSourceFile();
+    while (true) {
+      Token Tok;
+      PP.Lex(Tok);
+      if (Tok.is(tok::eof))
+        break;
+    }
+
+    return Callbacks->Marks;
+  }
+
   PragmaOpenCLExtensionCallbacks::CallbackParameters
   PragmaOpenCLExtensionCall(const char *SourceText) {
     LangOptions OpenCLLangOpts;
     OpenCLLangOpts.OpenCL = 1;
 
     std::unique_ptr<llvm::MemoryBuffer> SourceBuf =
         llvm::MemoryBuffer::getMemBuffer(SourceText, "test.cl");
     SourceMgr.setMainFileID(SourceMgr.createFileID(std::move(SourceBuf)));
TEST_F(PPCallbacksTest, OpenCLExtensionPragmaDisabled) {
   PragmaOpenCLExtensionCallbacks::CallbackParameters Parameters =
     PragmaOpenCLExtensionCall(Source);
 
   ASSERT_EQ("cl_khr_fp16", Parameters.Name);
   unsigned ExpectedState = 0;
   ASSERT_EQ(ExpectedState, Parameters.State);
 }
 
+TEST_F(PPCallbacksTest, CollectMarks) {
+  const char *Source =
+    "#pragma mark\n"
+    "#pragma mark\r\n"
+    "#pragma mark - trivia\n"
+    "#pragma mark - trivia\r\n";
+
+  auto Marks = PragmaMarkCall(Source);
+
+  ASSERT_EQ(4u, Marks.size());
+  ASSERT_TRUE(Marks[0].Trivia.empty());
+  ASSERT_TRUE(Marks[1].Trivia.empty());
+  ASSERT_FALSE(Marks[2].Trivia.empty());
+  ASSERT_FALSE(Marks[3].Trivia.empty());
+  ASSERT_EQ(" - trivia", Marks[2].Trivia);
+  ASSERT_EQ(" - trivia", Marks[3].Trivia);
+}
+
 TEST_F(PPCallbacksTest, DirectiveExprRanges) {
   const auto &Results1 = DirectiveExprRange("#if FLUZZY_FLOOF\n#endif\n");
   EXPECT_EQ(Results1.size(), 1U);
   EXPECT_EQ(
       GetSourceStringToEnd(CharSourceRange(Results1[0].ConditionRange, false)),
       "FLUZZY_FLOOF");
 
   const auto &Results2 = DirectiveExprRange("#if 1 + 4 < 7\n#endif\n");