martong marked 14 inline comments as done. martong added a comment. > Do you plan to enable this functionality for AST import checking?
Yes. We'd like to test the structural equivalency independently from ASTImporter, because in certain cases it may have faulty behavior. This can be very handy also when we further extend structural equivalency. ================ Comment at: unittests/AST/StructuralEquivalenceTest.cpp:119 + *cast<ClassTemplateDecl>(get<1>(t))->spec_begin(); + ASSERT_TRUE(Spec0 != nullptr); + ASSERT_TRUE(Spec1 != nullptr); ---------------- a.sidorin wrote: > Should we assert for `spec_begin() != spec_end()` instead or within these > assertions? I think that is not needed, because the previous `cast` in line 116 and 118 would assert if `spec_begin() == spec_end()` was true. ================ Comment at: unittests/AST/StructuralEquivalenceTest.cpp:174 + )"; + auto t = makeNamedDecls( Code0, Code0, Lang_CXX); + ---------------- a.sidorin wrote: > 1. It's better to use more meaningful names than `t`. DeclTuple? > 2. The space after `(` is redundant. Renamed `t` to `Decls`. Repository: rC Clang https://reviews.llvm.org/D46867 _______________________________________________ cfe-commits mailing list cfe-commits@lists.llvm.org http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits