Generalize and harmonize sub-expression traversal

[kimgr]

Summary

CastExpr::getSubExprAsWritten and getConversionFunction used to have
disparate implementations to traverse the sub-expression chain and skip
so-called "implicit temporaries" (which are really implicit nodes added by
Sema to represent semantic details in the AST).

There's some friction in these algorithms that makes it hard to extend and
change them:

• skipImplicitTemporary is order-dependent; it can skip a CXXBindTemporaryExpr nested inside a MaterializeTemporaryExpr, but not vice versa
• skipImplicitTemporary only runs one pass, it does not traverse multiple nested sequences of MTE/CBTE/MTE/CBTE, for example

Both of these weaknesses are void at this point, because this kind of
out-of-order multi-level nesting does not exist in the current AST.

Adding a new implicit expression to skip exacerbates the problem, however,
since a node X might show up in any and all locations between the existing.

Thus;

• Harmonize the form of getSubExprAsWritten and getConversionFunction so they both use a for loop
• Use the IgnoreExprNodes machinery to skip multiple nodes
• Rename skipImplicitTemporary to ignoreImplicitSemaNodes to generalize
• Update ignoreImplicitSemaNodes so it only skips one level per call, to mirror existing Ignore functions and work better with IgnoreExprNodes

This is a functional change, but one without visible effect.

[kimgr]

@aaron.ballman First refactor to generalize and harmonize getSubExprAsWritten and getConversionFunction. As mentioned in the commit message, this is strictly speaking a functional change, but it should have no visible effect.

[kimgr]

clang/lib/AST/Expr.cpp

1903

Typo: IgnoreExpr*e*Nodes

[kimgr]

Fix typo in comment.

[kimgr]

Gentle weekly ping

[kimgr]

@aaron.ballman Friendly Monday ping.

[aaron.ballman]

Sorry for the delay in reviewing this, but this looks correct to me. I don't think the precommit CI failures on Debian relate to your patch, either, so this LGTM

[kimgr]

Thanks! Could you help me land it? Author: Kim Gräsman <kim.grasman at gmail.com>.

[aaron.ballman]

In D119476#3394189, @kimgr wrote:

Thanks! Could you help me land it? Author: Kim Gräsman <kim.grasman at gmail.com>.

Happy to!

[aaron.ballman]

In D119476#3394189, @kimgr wrote:

Thanks! Could you help me land it? Author: Kim Gräsman <kim.grasman at gmail.com>.

I've commit on your behalf in 276d2143148fd1519af8eef124dabc41eabb1bb0

clang/lib/AST/Expr.cpp

const char *CastExpr::getCastKindName(CastKind CK) {
   switch (CK) {
 #define CAST_OPERATION(Name) case CK_##Name: return #Name;
 #include "clang/AST/OperationKinds.def"
   }
   llvm_unreachable("Unhandled cast kind!");
 }
 
 namespace {
-  const Expr *skipImplicitTemporary(const Expr *E) {
-    // Skip through reference binding to temporary.
-    if (auto *Materialize = dyn_cast<MaterializeTemporaryExpr>(E))
-      E = Materialize->getSubExpr();
+// Skip over implicit nodes produced as part of semantic analysis.
+// Designed for use with IgnoreExprNodes.
+Expr *ignoreImplicitSemaNodes(Expr *E) {
+    return Materialize->getSubExpr();
 
-    // Skip any temporary bindings; they're implicit.
-    if (auto *Binder = dyn_cast<CXXBindTemporaryExpr>(E))
-      E = Binder->getSubExpr();
+    return Binder->getSubExpr();
 
-    return E;
-  }
-}
+} // namespace
 
 Expr *CastExpr::getSubExprAsWritten() {
   const Expr *SubExpr = nullptr;
-  const CastExpr *E = this;
-  do {
-    SubExpr = skipImplicitTemporary(E->getSubExpr());
+
+  for (const CastExpr *E = this; E; E = dyn_cast<ImplicitCastExpr>(SubExpr)) {
+    SubExpr = IgnoreExprNodes(E->getSubExpr(), ignoreImplicitSemaNodes);
 
     // Conversions by constructor and conversion functions have a
     // subexpression describing the call; strip it off.
-    if (E->getCastKind() == CK_ConstructorConversion)
-      SubExpr =
-        skipImplicitTemporary(cast<CXXConstructExpr>(SubExpr->IgnoreImplicit())->getArg(0));
-    else if (E->getCastKind() == CK_UserDefinedConversion) {
+    if (E->getCastKind() == CK_ConstructorConversion) {
+      SubExpr = IgnoreExprNodes(
+          cast<CXXConstructExpr>(SubExpr->IgnoreImplicit())->getArg(0),
+          ignoreImplicitSemaNodes);
+    } else if (E->getCastKind() == CK_UserDefinedConversion) {
       SubExpr = SubExpr->IgnoreImplicit();
-      assert((isa<CXXMemberCallExpr>(SubExpr) ||
-              isa<BlockExpr>(SubExpr)) &&
+      assert((isa<CXXMemberCallExpr>(SubExpr) || isa<BlockExpr>(SubExpr)) &&
              "Unexpected SubExpr for CK_UserDefinedConversion.");
       if (auto *MCE = dyn_cast<CXXMemberCallExpr>(SubExpr))
         SubExpr = MCE->getImplicitObjectArgument();
     }
-
-    // If the subexpression we're left with is an implicit cast, look
-    // through that, too.
-  } while ((E = dyn_cast<ImplicitCastExpr>(SubExpr)));
+  }
 
-  return const_cast<Expr*>(SubExpr);
+  return const_cast<Expr *>(SubExpr);
 }
 
 NamedDecl *CastExpr::getConversionFunction() const {
   const Expr *SubExpr = nullptr;
 
   for (const CastExpr *E = this; E; E = dyn_cast<ImplicitCastExpr>(SubExpr)) {
-    SubExpr = skipImplicitTemporary(E->getSubExpr());
+    SubExpr = IgnoreExprNodes(E->getSubExpr(), ignoreImplicitSemaNodes);
 
     if (E->getCastKind() == CK_ConstructorConversion)
       return cast<CXXConstructExpr>(SubExpr)->getConstructor();
 
     if (E->getCastKind() == CK_UserDefinedConversion) {
       if (auto *MCE = dyn_cast<CXXMemberCallExpr>(SubExpr))
         return MCE->getMethodDecl();
     }