[clang][Tooling] Fix getFileRange returning a range spanning across macro arguments (llvm#169757)

When the start and end token are both spelled in macro arguments, we
still want to reject the range if they come from two separate macro
arguments, as the original specified range is not precisely spelled in a
single sequence of characters in source.

Author: tJener

clang/lib/Tooling/Transformer/SourceCode.cpp

@@ -86,18 +86,39 @@ llvm::Error clang::tooling::validateEditRange(const CharSourceRange &Range,
   return validateRange(Range, SM, /*AllowSystemHeaders=*/false);
 }
 
-static bool spelledInMacroDefinition(SourceLocation Loc,
-                                     const SourceManager &SM) {
+// Returns the location of the top-level macro argument that is the spelling for
+// the expansion `Loc` is from. If `Loc` is spelled in the macro definition,
+// returns an invalid `SourceLocation`.
+static SourceLocation getMacroArgumentSpellingLoc(SourceLocation Loc,
+                                                  const SourceManager &SM) {
+  assert(Loc.isMacroID() && "Location must be in a macro");
   while (Loc.isMacroID()) {
     const auto &Expansion = SM.getSLocEntry(SM.getFileID(Loc)).getExpansion();
     if (Expansion.isMacroArgExpansion()) {
       // Check the spelling location of the macro arg, in case the arg itself is
       // in a macro expansion.
       Loc = Expansion.getSpellingLoc();
     } else {
-      return true;
+      return {};
     }
   }
+  return Loc;
+}
+
+static bool spelledInMacroDefinition(CharSourceRange Range,
+                                     const SourceManager &SM) {
+  if (Range.getBegin().isMacroID() && Range.getEnd().isMacroID()) {
+    // Check whether the range is entirely within a single macro argument.
+    auto B = getMacroArgumentSpellingLoc(Range.getBegin(), SM);
+    auto E = getMacroArgumentSpellingLoc(Range.getEnd(), SM);
+    return B.isInvalid() || B != E;
+  }
+
+  if (Range.getBegin().isMacroID())
+    return getMacroArgumentSpellingLoc(Range.getBegin(), SM).isInvalid();
+  if (Range.getEnd().isMacroID())
+    return getMacroArgumentSpellingLoc(Range.getEnd(), SM).isInvalid();
+
   return false;
 }
 
@@ -158,8 +179,7 @@ static CharSourceRange getRange(const CharSourceRange &EditRange,
     Range = Lexer::makeFileCharRange(EditRange, SM, LangOpts);
   } else {
     auto AdjustedRange = getRangeForSplitTokens(EditRange, SM, LangOpts);
-    if (spelledInMacroDefinition(AdjustedRange.getBegin(), SM) ||
-        spelledInMacroDefinition(AdjustedRange.getEnd(), SM))
+    if (spelledInMacroDefinition(AdjustedRange, SM))
       return {};
 
     auto B = SM.getSpellingLoc(AdjustedRange.getBegin());

clang/unittests/Tooling/SourceCodeTest.cpp

@@ -510,10 +510,12 @@ TEST(SourceCodeTest, EditInvolvingExpansionIgnoringExpansionShouldFail) {
 #define M1(x) x(1)
 #define M2(x, y) x ## y
 #define M3(x) foobar(x)
+#define M4(x, y) x y
 int foobar(int);
 int a = M1(foobar);
 int b = M2(foo, bar(2));
 int c = M3(3);
+int d = M4(foobar, (4));
 )cpp");
 
   CallsVisitor Visitor;