[PATCH] [MSVC] Better lookup into dependent bases for unspecified declaration (also fixes http://llvm.org/PR22066)

[Alexey Bataev]

Reid, thanks for the review!


================
Comment at: lib/Sema/SemaDecl.cpp:139-142
@@ +138,6 @@
+  FoundNotSpecifiedKind,
+  /// \brief There are no members is found, but class has a base of class
+  /// TemplateTypeParmType and we are allowed to suppose that the specified
+  /// identifier may be a member of this base.
+  SupposeFoundSpecifiedKind,
+  /// \brief Found a member of the expected type.
----------------
rnk wrote:
> OK, this is sort of the case being described:
>   template <typename T> struct Foo : T { void f(int a) { IdFromBase(a); } };
> 
> I think we need better terminology to talk about how this case differs from this case:
>   template <typename T> struct Foo : Bar<T> { ... };
> 
> Both of these are classes with dependent bases, but in the second case, we can go searching through the template pattern for Bar to try to see if it has type or non-type members. I think if we can come up with a name for one of these two cases, we can make this code a lot clearer.
> 
> We should get Richard's input on this, but here are some ideas:
> - Fully dependent / partially dependent
> - Dependent base with known pattern / unknown dependent base
> 
> So far, I like using the "known / unknown pattern" terminology best, so I might rename this enumerator to NotFoundWithUnknownBase. This indicates that after doing extended lookup into dependent bases with known patterns, no members with this name were found, but we hit an unknown base that we couldn't look through.
Yep, you're right. I also prefer "known / unknown pattern". I'll rename the enumerator.

================
Comment at: lib/Sema/SemaDecl.cpp:170
@@ +169,3 @@
+  }
+  if (FoundDecl == IdNameLookupResult::NotFound) {
+    for (const auto &Base : RD->bases()) {
----------------
rnk wrote:
> This can be an early return:
>   if (FoundDecl != IdNameLookupResult::NotFound)
>     return FoundDecl;
Ok, fixed

================
Comment at: lib/Sema/SemaDecl.cpp:195
@@ +194,3 @@
+      }
+      if (BaseRD) {
+        switch (lookupIdNameInDependentClass(
----------------
rnk wrote:
> This can also continue early:
>   if (!BaseRD)
>     continue;
Fixed

================
Comment at: lib/Sema/SemaDecl.cpp:220
@@ -196,1 +219,3 @@
+                       const IdentifierInfo &II, bool SupposeExistInBaseParm,
+                       const std::function<bool(NamedDecl *)> &CheckKind) {
   // Lookup in the parent class template context, if any.
----------------
rnk wrote:
> This is still `std::function`
Fixed

http://reviews.llvm.org/D8100

EMAIL PREFERENCES
  http://reviews.llvm.org/settings/panel/emailpreferences/