[clang] [RFC] Initial implementation of P2719 (PR #113510)

[Oliver Hunt via cfe-commits]

================
@@ -3134,6 +3134,31 @@ bool Type::isStdByteType() const {
   return false;
 }
 
+bool Type::isDestroyingDeleteT() const {
+  auto *RD = getAsCXXRecordDecl();
+  return RD && RD->isInStdNamespace() && RD->getIdentifier() &&
+         RD->getIdentifier()->isStr("destroying_delete_t");
+}
+
+TemplateDecl *Type::getSpecializedTemplateDecl() const {
+  auto UnderlyingType = getCanonicalTypeInternal();
+  if (auto *TST = UnderlyingType->getAs<TemplateSpecializationType>())
+    return TST->getTemplateName().getAsTemplateDecl();
+  if (auto *RT = UnderlyingType->getAsCXXRecordDecl()) {
+    if (auto *CTSD = dyn_cast<ClassTemplateSpecializationDecl>(RT))
+      return CTSD->getSpecializedTemplate();
+  }
+  return nullptr;
+}
+
+bool Type::isTypeIdentitySpecialization() const {
+  const TemplateDecl *STDecl = getSpecializedTemplateDecl();
+  if (!STDecl)
+    return false;
+  IdentifierInfo *II = STDecl->getIdentifier();
+  return STDecl->isInStdNamespace() && II->isStr("type_identity");
+}
+
----------------
ojhunt wrote:

@cor3ntin  Ok, I worked out how I ended up here - I originally had multiple "isX" methods, one on Sema one on the type or decl, but reduced it to a single shared implementation in response to one of the first reviews - they're modeled after the existing Type::isFoo methods above. As these functions are not always queried when there is a Sema object available, we'd need to thread a Sema reference through a bunch of methods on FunctionDecl and MethodDecl which does not seem to be something that is done on any of the other methods.

https://github.com/llvm/llvm-project/pull/113510